From 8d836581f9b1f57ceaa2b47654754ef1260b410b Mon Sep 17 00:00:00 2001 From: Niklas Cassel Date: Wed, 16 Sep 2026 15:00:32 +0200 Subject: ata: libata-scsi: fix ata_dsm_trim_pages() kernel-doc make htmldocs fails with: Documentation/driver-api/libata:604: ./drivers/ata/libata-scsi.c:2301: ERROR: Unexpected indentation. [docutils] The Return: section of ata_dsm_trim_pages() ends its first sentence with a colon and continues with an indented bullet list. reStructuredText requires a blank line before an indented block, so docutils chokes on the list. Simply adding the missing blank line does not work either: Return: is a kernel-doc "special section", which is terminated by the first blank line, so the bullet list would end up in the Description section, detached from the sentence introducing it. Spell the two bounds out as prose instead, so that the Return: section stays self-contained. Documentation-only change. Fixes: e64e6b5dc867 ("ata: libata-scsi: scale DSM TRIM payload by MAX PAGES PER DSM COMMAND") Reported-by: Thomas Huth Closes: https://lore.kernel.org/linux-ide/b0a0b8e8-cc4f-4c7b-8bc5-fee0712d405a@redhat.com/ Reviewed-by: Damien Le Moal Reviewed-by: Thomas Huth Link: https://lore.kernel.org/r/20260916130031.29990-2-cassel@kernel.org Signed-off-by: Niklas Cassel --- drivers/mtd/ubi/Kconfig | 129 +++ drivers/mtd/ubi/Makefile | 10 + drivers/mtd/ubi/attach.c | 1925 +++++++++++++++++++++++++++++++++++++ drivers/mtd/ubi/block.c | 678 +++++++++++++ drivers/mtd/ubi/build.c | 1632 +++++++++++++++++++++++++++++++ drivers/mtd/ubi/cdev.c | 1177 +++++++++++++++++++++++ drivers/mtd/ubi/debug.c | 697 ++++++++++++++ drivers/mtd/ubi/debug.h | 387 ++++++++ drivers/mtd/ubi/eba.c | 1709 +++++++++++++++++++++++++++++++++ drivers/mtd/ubi/fastmap-wl.c | 561 +++++++++++ drivers/mtd/ubi/fastmap.c | 1648 ++++++++++++++++++++++++++++++++ drivers/mtd/ubi/gluebi.c | 500 ++++++++++ drivers/mtd/ubi/io.c | 1487 +++++++++++++++++++++++++++++ drivers/mtd/ubi/kapi.c | 848 +++++++++++++++++ drivers/mtd/ubi/misc.c | 191 ++++ drivers/mtd/ubi/nvmem.c | 191 ++++ drivers/mtd/ubi/ubi-media.h | 503 ++++++++++ drivers/mtd/ubi/ubi.h | 1266 +++++++++++++++++++++++++ drivers/mtd/ubi/upd.c | 420 ++++++++ drivers/mtd/ubi/vmt.c | 848 +++++++++++++++++ drivers/mtd/ubi/vtbl.c | 877 +++++++++++++++++ drivers/mtd/ubi/wl.c | 2160 ++++++++++++++++++++++++++++++++++++++++++ drivers/mtd/ubi/wl.h | 33 + 23 files changed, 19877 insertions(+) create mode 100644 drivers/mtd/ubi/Kconfig create mode 100644 drivers/mtd/ubi/Makefile create mode 100644 drivers/mtd/ubi/attach.c create mode 100644 drivers/mtd/ubi/block.c create mode 100644 drivers/mtd/ubi/build.c create mode 100644 drivers/mtd/ubi/cdev.c create mode 100644 drivers/mtd/ubi/debug.c create mode 100644 drivers/mtd/ubi/debug.h create mode 100644 drivers/mtd/ubi/eba.c create mode 100644 drivers/mtd/ubi/fastmap-wl.c create mode 100644 drivers/mtd/ubi/fastmap.c create mode 100644 drivers/mtd/ubi/gluebi.c create mode 100644 drivers/mtd/ubi/io.c create mode 100644 drivers/mtd/ubi/kapi.c create mode 100644 drivers/mtd/ubi/misc.c create mode 100644 drivers/mtd/ubi/nvmem.c create mode 100644 drivers/mtd/ubi/ubi-media.h create mode 100644 drivers/mtd/ubi/ubi.h create mode 100644 drivers/mtd/ubi/upd.c create mode 100644 drivers/mtd/ubi/vmt.c create mode 100644 drivers/mtd/ubi/vtbl.c create mode 100644 drivers/mtd/ubi/wl.c create mode 100644 drivers/mtd/ubi/wl.h (limited to 'drivers/mtd/ubi') diff --git a/drivers/mtd/ubi/Kconfig b/drivers/mtd/ubi/Kconfig new file mode 100644 index 000000000..e28a3af83 --- /dev/null +++ b/drivers/mtd/ubi/Kconfig @@ -0,0 +1,129 @@ +# SPDX-License-Identifier: GPL-2.0-only +menuconfig MTD_UBI + tristate "Enable UBI - Unsorted block images" + select CRC32 + help + UBI is a software layer above MTD layer which admits use of LVM-like + logical volumes on top of MTD devices, hides some complexities of + flash chips like wear and bad blocks and provides some other useful + capabilities. Please, consult the MTD web site for more details + (www.linux-mtd.infradead.org). + +if MTD_UBI + +config MTD_UBI_WL_THRESHOLD + int "UBI wear-leveling threshold" + default 4096 + range 2 65536 + help + This parameter defines the maximum difference between the highest + erase counter value and the lowest erase counter value of eraseblocks + of UBI devices. When this threshold is exceeded, UBI starts performing + wear leveling by means of moving data from eraseblock with low erase + counter to eraseblocks with high erase counter. + + The default value should be OK for SLC NAND flashes, NOR flashes and + other flashes which have eraseblock life-cycle 100000 or more. + However, in case of MLC NAND flashes which typically have eraseblock + life-cycle less than 10000, the threshold should be lessened (e.g., + to 128 or 256, although it does not have to be power of 2). + +config MTD_UBI_BEB_LIMIT + int "Maximum expected bad eraseblock count per 1024 eraseblocks" + default 20 + range 0 768 + help + This option specifies the maximum bad physical eraseblocks UBI + expects on the MTD device (per 1024 eraseblocks). If the underlying + flash does not admit of bad eraseblocks (e.g. NOR flash), this value + is ignored. + + NAND datasheets often specify the minimum and maximum NVM (Number of + Valid Blocks) for the flashes' endurance lifetime. The maximum + expected bad eraseblocks per 1024 eraseblocks then can be calculated + as "1024 * (1 - MinNVB / MaxNVB)", which gives 20 for most NANDs + (MaxNVB is basically the total count of eraseblocks on the chip). + + To put it differently, if this value is 20, UBI will try to reserve + about 1.9% of physical eraseblocks for bad blocks handling. And that + will be 1.9% of eraseblocks on the entire NAND chip, not just the MTD + partition UBI attaches. This means that if you have, say, a NAND + flash chip admits maximum 40 bad eraseblocks, and it is split on two + MTD partitions of the same size, UBI will reserve 40 eraseblocks when + attaching a partition. + + This option can be overridden by the "mtd=" UBI module parameter or + by the "attach" ioctl. + + Leave the default value if unsure. + +config MTD_UBI_FASTMAP + bool "UBI Fastmap (Experimental feature)" + default n + help + Important: this feature is experimental so far and the on-flash + format for fastmap may change in the next kernel versions + + Fastmap is a mechanism which allows attaching an UBI device + in nearly constant time. Instead of scanning the whole MTD device it + only has to locate a checkpoint (called fastmap) on the device. + The on-flash fastmap contains all information needed to attach + the device. Using fastmap makes only sense on large devices where + attaching by scanning takes long. UBI will not automatically install + a fastmap on old images, but you can set the UBI module parameter + fm_autoconvert to 1 if you want so. Please note that fastmap-enabled + images are still usable with UBI implementations without + fastmap support. On typical flash devices the whole fastmap fits + into one PEB. UBI will reserve PEBs to hold two fastmaps. + + If in doubt, say "N". + +config MTD_UBI_GLUEBI + tristate "MTD devices emulation driver (gluebi)" + help + This option enables gluebi - an additional driver which emulates MTD + devices on top of UBI volumes: for each UBI volumes an MTD device is + created, and all I/O to this MTD device is redirected to the UBI + volume. This is handy to make MTD-oriented software (like JFFS2) + work on top of UBI. Do not enable this unless you use legacy + software. + +config MTD_UBI_BLOCK + bool "Read-only block devices on top of UBI volumes" + default n + depends on BLOCK + help + This option enables read-only UBI block devices support. UBI block + devices will be layered on top of UBI volumes, which means that the + UBI driver will transparently handle things like bad eraseblocks and + bit-flips. You can put any block-oriented file system on top of UBI + volumes in read-only mode (e.g., ext4), but it is probably most + practical for read-only file systems, like squashfs. + + When selected, this feature will be built in the UBI driver. + + If in doubt, say "N". + +config MTD_UBI_FAULT_INJECTION + bool "Fault injection capability of UBI device" + default n + depends on FAULT_INJECTION_DEBUG_FS + help + This option enables fault-injection support for UBI devices for + testing purposes. + + If in doubt, say "N". + +config MTD_UBI_NVMEM + tristate "UBI virtual NVMEM" + default n + depends on NVMEM + help + This option enabled an additional driver exposing UBI volumes as NVMEM + providers, intended for platforms where UBI is part of the firmware + specification and used to store also e.g. MAC addresses or board- + specific Wi-Fi calibration data. + + If in doubt, say "N". + +endif # MTD_UBI diff --git a/drivers/mtd/ubi/Makefile b/drivers/mtd/ubi/Makefile new file mode 100644 index 000000000..4b51aaf00 --- /dev/null +++ b/drivers/mtd/ubi/Makefile @@ -0,0 +1,10 @@ +# SPDX-License-Identifier: GPL-2.0 +obj-$(CONFIG_MTD_UBI) += ubi.o + +ubi-y += vtbl.o vmt.o upd.o build.o cdev.o kapi.o eba.o io.o wl.o attach.o +ubi-y += misc.o debug.o +ubi-$(CONFIG_MTD_UBI_FASTMAP) += fastmap.o +ubi-$(CONFIG_MTD_UBI_BLOCK) += block.o + +obj-$(CONFIG_MTD_UBI_GLUEBI) += gluebi.o +obj-$(CONFIG_MTD_UBI_NVMEM) += nvmem.o diff --git a/drivers/mtd/ubi/attach.c b/drivers/mtd/ubi/attach.c new file mode 100644 index 000000000..0ce7ff740 --- /dev/null +++ b/drivers/mtd/ubi/attach.c @@ -0,0 +1,1925 @@ +// SPDX-License-Identifier: GPL-2.0-or-later +/* + * Copyright (c) International Business Machines Corp., 2006 + * + * Author: Artem Bityutskiy (Битюцкий Артём) + */ + +/* + * UBI attaching sub-system. + * + * This sub-system is responsible for attaching MTD devices and it also + * implements flash media scanning. + * + * The attaching information is represented by a &struct ubi_attach_info' + * object. Information about volumes is represented by &struct ubi_ainf_volume + * objects which are kept in volume RB-tree with root at the @volumes field. + * The RB-tree is indexed by the volume ID. + * + * Logical eraseblocks are represented by &struct ubi_ainf_peb objects. These + * objects are kept in per-volume RB-trees with the root at the corresponding + * &struct ubi_ainf_volume object. To put it differently, we keep an RB-tree of + * per-volume objects and each of these objects is the root of RB-tree of + * per-LEB objects. + * + * Corrupted physical eraseblocks are put to the @corr list, free physical + * eraseblocks are put to the @free list and the physical eraseblock to be + * erased are put to the @erase list. + * + * About corruptions + * ~~~~~~~~~~~~~~~~~ + * + * UBI protects EC and VID headers with CRC-32 checksums, so it can detect + * whether the headers are corrupted or not. Sometimes UBI also protects the + * data with CRC-32, e.g., when it executes the atomic LEB change operation, or + * when it moves the contents of a PEB for wear-leveling purposes. + * + * UBI tries to distinguish between 2 types of corruptions. + * + * 1. Corruptions caused by power cuts. These are expected corruptions and UBI + * tries to handle them gracefully, without printing too many warnings and + * error messages. The idea is that we do not lose important data in these + * cases - we may lose only the data which were being written to the media just + * before the power cut happened, and the upper layers (e.g., UBIFS) are + * supposed to handle such data losses (e.g., by using the FS journal). + * + * When UBI detects a corruption (CRC-32 mismatch) in a PEB, and it looks like + * the reason is a power cut, UBI puts this PEB to the @erase list, and all + * PEBs in the @erase list are scheduled for erasure later. + * + * 2. Unexpected corruptions which are not caused by power cuts. During + * attaching, such PEBs are put to the @corr list and UBI preserves them. + * Obviously, this lessens the amount of available PEBs, and if at some point + * UBI runs out of free PEBs, it switches to R/O mode. UBI also loudly informs + * about such PEBs every time the MTD device is attached. + * + * However, it is difficult to reliably distinguish between these types of + * corruptions and UBI's strategy is as follows (in case of attaching by + * scanning). UBI assumes corruption type 2 if the VID header is corrupted and + * the data area does not contain all 0xFFs, and there were no bit-flips or + * integrity errors (e.g., ECC errors in case of NAND) while reading the data + * area. Otherwise UBI assumes corruption type 1. So the decision criteria + * are as follows. + * o If the data area contains only 0xFFs, there are no data, and it is safe + * to just erase this PEB - this is corruption type 1. + * o If the data area has bit-flips or data integrity errors (ECC errors on + * NAND), it is probably a PEB which was being erased when power cut + * happened, so this is corruption type 1. However, this is just a guess, + * which might be wrong. + * o Otherwise this is corruption type 2. + */ + +#include +#include +#include +#include +#include +#include "ubi.h" + +static int self_check_ai(struct ubi_device *ubi, struct ubi_attach_info *ai); + +#define AV_FIND BIT(0) +#define AV_ADD BIT(1) +#define AV_FIND_OR_ADD (AV_FIND | AV_ADD) + +/** + * find_or_add_av - internal function to find a volume, add a volume or do + * both (find and add if missing). + * @ai: attaching information + * @vol_id: the requested volume ID + * @flags: a combination of the %AV_FIND and %AV_ADD flags describing the + * expected operation. If only %AV_ADD is set, -EEXIST is returned + * if the volume already exists. If only %AV_FIND is set, NULL is + * returned if the volume does not exist. And if both flags are + * set, the helper first tries to find an existing volume, and if + * it does not exist it creates a new one. + * @created: in value used to inform the caller whether it"s a newly created + * volume or not. + * + * This function returns a pointer to a volume description or an ERR_PTR if + * the operation failed. It can also return NULL if only %AV_FIND is set and + * the volume does not exist. + */ +static struct ubi_ainf_volume *find_or_add_av(struct ubi_attach_info *ai, + int vol_id, unsigned int flags, + bool *created) +{ + struct ubi_ainf_volume *av; + struct rb_node **p = &ai->volumes.rb_node, *parent = NULL; + + /* Walk the volume RB-tree to look if this volume is already present */ + while (*p) { + parent = *p; + av = rb_entry(parent, struct ubi_ainf_volume, rb); + + if (vol_id == av->vol_id) { + *created = false; + + if (!(flags & AV_FIND)) + return ERR_PTR(-EEXIST); + + return av; + } + + if (vol_id > av->vol_id) + p = &(*p)->rb_left; + else + p = &(*p)->rb_right; + } + + if (!(flags & AV_ADD)) + return NULL; + + /* The volume is absent - add it */ + av = kzalloc_obj(*av); + if (!av) + return ERR_PTR(-ENOMEM); + + av->vol_id = vol_id; + + if (vol_id > ai->highest_vol_id) + ai->highest_vol_id = vol_id; + + rb_link_node(&av->rb, parent, p); + rb_insert_color(&av->rb, &ai->volumes); + ai->vols_found += 1; + *created = true; + dbg_bld("added volume %d", vol_id); + return av; +} + +/** + * ubi_find_or_add_av - search for a volume in the attaching information and + * add one if it does not exist. + * @ai: attaching information + * @vol_id: the requested volume ID + * @created: whether the volume has been created or not + * + * This function returns a pointer to the new volume description or an + * ERR_PTR if the operation failed. + */ +static struct ubi_ainf_volume *ubi_find_or_add_av(struct ubi_attach_info *ai, + int vol_id, bool *created) +{ + return find_or_add_av(ai, vol_id, AV_FIND_OR_ADD, created); +} + +/** + * ubi_alloc_aeb - allocate an aeb element + * @ai: attaching information + * @pnum: physical eraseblock number + * @ec: erase counter of the physical eraseblock + * + * Allocate an aeb object and initialize the pnum and ec information. + * vol_id and lnum are set to UBI_UNKNOWN, and the other fields are + * initialized to zero. + * Note that the element is not added in any list or RB tree. + */ +struct ubi_ainf_peb *ubi_alloc_aeb(struct ubi_attach_info *ai, int pnum, + int ec) +{ + struct ubi_ainf_peb *aeb; + + aeb = kmem_cache_zalloc(ai->aeb_slab_cache, GFP_KERNEL); + if (!aeb) + return NULL; + + aeb->pnum = pnum; + aeb->ec = ec; + aeb->vol_id = UBI_UNKNOWN; + aeb->lnum = UBI_UNKNOWN; + + return aeb; +} + +/** + * ubi_free_aeb - free an aeb element + * @ai: attaching information + * @aeb: the element to free + * + * Free an aeb object. The caller must have removed the element from any list + * or RB tree. + */ +void ubi_free_aeb(struct ubi_attach_info *ai, struct ubi_ainf_peb *aeb) +{ + kmem_cache_free(ai->aeb_slab_cache, aeb); +} + +/** + * add_to_list - add physical eraseblock to a list. + * @ai: attaching information + * @pnum: physical eraseblock number to add + * @vol_id: the last used volume id for the PEB + * @lnum: the last used LEB number for the PEB + * @ec: erase counter of the physical eraseblock + * @to_head: if not zero, add to the head of the list + * @list: the list to add to + * + * This function allocates a 'struct ubi_ainf_peb' object for physical + * eraseblock @pnum and adds it to the "free", "erase", or "alien" lists. + * It stores the @lnum and @vol_id alongside, which can both be + * %UBI_UNKNOWN if they are not available, not readable, or not assigned. + * If @to_head is not zero, PEB will be added to the head of the list, which + * basically means it will be processed first later. E.g., we add corrupted + * PEBs (corrupted due to power cuts) to the head of the erase list to make + * sure we erase them first and get rid of corruptions ASAP. This function + * returns zero in case of success and a negative error code in case of + * failure. + */ +static int add_to_list(struct ubi_attach_info *ai, int pnum, int vol_id, + int lnum, int ec, int to_head, struct list_head *list) +{ + struct ubi_ainf_peb *aeb; + + if (list == &ai->free) { + dbg_bld("add to free: PEB %d, EC %d", pnum, ec); + } else if (list == &ai->erase) { + dbg_bld("add to erase: PEB %d, EC %d", pnum, ec); + } else if (list == &ai->alien) { + dbg_bld("add to alien: PEB %d, EC %d", pnum, ec); + ai->alien_peb_count += 1; + } else + BUG(); + + aeb = ubi_alloc_aeb(ai, pnum, ec); + if (!aeb) + return -ENOMEM; + + aeb->vol_id = vol_id; + aeb->lnum = lnum; + if (to_head) + list_add(&aeb->u.list, list); + else + list_add_tail(&aeb->u.list, list); + return 0; +} + +/** + * add_corrupted - add a corrupted physical eraseblock. + * @ai: attaching information + * @pnum: physical eraseblock number to add + * @ec: erase counter of the physical eraseblock + * + * This function allocates a 'struct ubi_ainf_peb' object for a corrupted + * physical eraseblock @pnum and adds it to the 'corr' list. The corruption + * was presumably not caused by a power cut. Returns zero in case of success + * and a negative error code in case of failure. + */ +static int add_corrupted(struct ubi_attach_info *ai, int pnum, int ec) +{ + struct ubi_ainf_peb *aeb; + + dbg_bld("add to corrupted: PEB %d, EC %d", pnum, ec); + + aeb = ubi_alloc_aeb(ai, pnum, ec); + if (!aeb) + return -ENOMEM; + + ai->corr_peb_count += 1; + list_add(&aeb->u.list, &ai->corr); + return 0; +} + +/** + * add_fastmap - add a Fastmap related physical eraseblock. + * @ai: attaching information + * @pnum: physical eraseblock number the VID header came from + * @vid_hdr: the volume identifier header + * @ec: erase counter of the physical eraseblock + * + * This function allocates a 'struct ubi_ainf_peb' object for a Fastamp + * physical eraseblock @pnum and adds it to the 'fastmap' list. + * Such blocks can be Fastmap super and data blocks from both the most + * recent Fastmap we're attaching from or from old Fastmaps which will + * be erased. + */ +static int add_fastmap(struct ubi_attach_info *ai, int pnum, + struct ubi_vid_hdr *vid_hdr, int ec) +{ + struct ubi_ainf_peb *aeb; + + aeb = ubi_alloc_aeb(ai, pnum, ec); + if (!aeb) + return -ENOMEM; + + aeb->vol_id = be32_to_cpu(vid_hdr->vol_id); + aeb->sqnum = be64_to_cpu(vid_hdr->sqnum); + list_add(&aeb->u.list, &ai->fastmap); + + dbg_bld("add to fastmap list: PEB %d, vol_id %d, sqnum: %llu", pnum, + aeb->vol_id, aeb->sqnum); + + return 0; +} + +/** + * validate_vid_hdr - check volume identifier header. + * @ubi: UBI device description object + * @vid_hdr: the volume identifier header to check + * @av: information about the volume this logical eraseblock belongs to + * @pnum: physical eraseblock number the VID header came from + * + * This function checks that data stored in @vid_hdr is consistent. Returns + * non-zero if an inconsistency was found and zero if not. + * + * Note, UBI does sanity check of everything it reads from the flash media. + * Most of the checks are done in the I/O sub-system. Here we check that the + * information in the VID header is consistent to the information in other VID + * headers of the same volume. + */ +static int validate_vid_hdr(const struct ubi_device *ubi, + const struct ubi_vid_hdr *vid_hdr, + const struct ubi_ainf_volume *av, int pnum) +{ + int vol_type = vid_hdr->vol_type; + int vol_id = be32_to_cpu(vid_hdr->vol_id); + int used_ebs = be32_to_cpu(vid_hdr->used_ebs); + int data_pad = be32_to_cpu(vid_hdr->data_pad); + + if (av->leb_count != 0) { + int av_vol_type; + + /* + * This is not the first logical eraseblock belonging to this + * volume. Ensure that the data in its VID header is consistent + * to the data in previous logical eraseblock headers. + */ + + if (vol_id != av->vol_id) { + ubi_err(ubi, "inconsistent vol_id"); + goto bad; + } + + if (av->vol_type == UBI_STATIC_VOLUME) + av_vol_type = UBI_VID_STATIC; + else + av_vol_type = UBI_VID_DYNAMIC; + + if (vol_type != av_vol_type) { + ubi_err(ubi, "inconsistent vol_type"); + goto bad; + } + + if (used_ebs != av->used_ebs) { + ubi_err(ubi, "inconsistent used_ebs"); + goto bad; + } + + if (data_pad != av->data_pad) { + ubi_err(ubi, "inconsistent data_pad"); + goto bad; + } + } + + return 0; + +bad: + ubi_err(ubi, "inconsistent VID header at PEB %d", pnum); + ubi_dump_vid_hdr(vid_hdr); + ubi_dump_av(av); + return -EINVAL; +} + +/** + * add_volume - add volume to the attaching information. + * @ai: attaching information + * @vol_id: ID of the volume to add + * @pnum: physical eraseblock number + * @vid_hdr: volume identifier header + * + * If the volume corresponding to the @vid_hdr logical eraseblock is already + * present in the attaching information, this function does nothing. Otherwise + * it adds corresponding volume to the attaching information. Returns a pointer + * to the allocated "av" object in case of success and a negative error code in + * case of failure. + */ +static struct ubi_ainf_volume *add_volume(struct ubi_attach_info *ai, + int vol_id, int pnum, + const struct ubi_vid_hdr *vid_hdr) +{ + struct ubi_ainf_volume *av; + bool created; + + ubi_assert(vol_id == be32_to_cpu(vid_hdr->vol_id)); + + av = ubi_find_or_add_av(ai, vol_id, &created); + if (IS_ERR(av) || !created) + return av; + + av->used_ebs = be32_to_cpu(vid_hdr->used_ebs); + av->data_pad = be32_to_cpu(vid_hdr->data_pad); + av->compat = vid_hdr->compat; + av->vol_type = vid_hdr->vol_type == UBI_VID_DYNAMIC ? UBI_DYNAMIC_VOLUME + : UBI_STATIC_VOLUME; + + return av; +} + +/** + * ubi_compare_lebs - find out which logical eraseblock is newer. + * @ubi: UBI device description object + * @aeb: first logical eraseblock to compare + * @pnum: physical eraseblock number of the second logical eraseblock to + * compare + * @vid_hdr: volume identifier header of the second logical eraseblock + * + * This function compares 2 copies of a LEB and informs which one is newer. In + * case of success this function returns a positive value, in case of failure, a + * negative error code is returned. The success return codes use the following + * bits: + * o bit 0 is cleared: the first PEB (described by @aeb) is newer than the + * second PEB (described by @pnum and @vid_hdr); + * o bit 0 is set: the second PEB is newer; + * o bit 1 is cleared: no bit-flips were detected in the newer LEB; + * o bit 1 is set: bit-flips were detected in the newer LEB; + * o bit 2 is cleared: the older LEB is not corrupted; + * o bit 2 is set: the older LEB is corrupted. + */ +int ubi_compare_lebs(struct ubi_device *ubi, const struct ubi_ainf_peb *aeb, + int pnum, const struct ubi_vid_hdr *vid_hdr) +{ + int len, err, second_is_newer, bitflips = 0, corrupted = 0; + uint32_t data_crc, crc; + struct ubi_vid_io_buf *vidb = NULL; + unsigned long long sqnum2 = be64_to_cpu(vid_hdr->sqnum); + + if (sqnum2 == aeb->sqnum) { + /* + * This must be a really ancient UBI image which has been + * created before sequence numbers support has been added. At + * that times we used 32-bit LEB versions stored in logical + * eraseblocks. That was before UBI got into mainline. We do not + * support these images anymore. Well, those images still work, + * but only if no unclean reboots happened. + */ + ubi_err(ubi, "unsupported on-flash UBI format"); + return -EINVAL; + } + + /* Obviously the LEB with lower sequence counter is older */ + second_is_newer = (sqnum2 > aeb->sqnum); + + /* + * Now we know which copy is newer. If the copy flag of the PEB with + * newer version is not set, then we just return, otherwise we have to + * check data CRC. For the second PEB we already have the VID header, + * for the first one - we'll need to re-read it from flash. + * + * Note: this may be optimized so that we wouldn't read twice. + */ + + if (second_is_newer) { + if (!vid_hdr->copy_flag) { + /* It is not a copy, so it is newer */ + dbg_bld("second PEB %d is newer, copy_flag is unset", + pnum); + return 1; + } + } else { + if (!aeb->copy_flag) { + /* It is not a copy, so it is newer */ + dbg_bld("first PEB %d is newer, copy_flag is unset", + pnum); + return bitflips << 1; + } + + vidb = ubi_alloc_vid_buf(ubi, GFP_KERNEL); + if (!vidb) + return -ENOMEM; + + pnum = aeb->pnum; + err = ubi_io_read_vid_hdr(ubi, pnum, vidb, 0); + if (err) { + if (err == UBI_IO_BITFLIPS) + bitflips = 1; + else { + ubi_err(ubi, "VID of PEB %d header is bad, but it was OK earlier, err %d", + pnum, err); + if (err > 0) + err = -EIO; + + goto out_free_vidh; + } + } + + vid_hdr = ubi_get_vid_hdr(vidb); + } + + /* Read the data of the copy and check the CRC */ + + len = be32_to_cpu(vid_hdr->data_size); + + mutex_lock(&ubi->buf_mutex); + err = ubi_io_read_data(ubi, ubi->peb_buf, pnum, 0, len); + if (err && err != UBI_IO_BITFLIPS && !mtd_is_eccerr(err)) + goto out_unlock; + + data_crc = be32_to_cpu(vid_hdr->data_crc); + crc = crc32(UBI_CRC32_INIT, ubi->peb_buf, len); + if (crc != data_crc) { + dbg_bld("PEB %d CRC error: calculated %#08x, must be %#08x", + pnum, crc, data_crc); + corrupted = 1; + bitflips = 0; + second_is_newer = !second_is_newer; + } else { + dbg_bld("PEB %d CRC is OK", pnum); + bitflips |= !!err; + } + mutex_unlock(&ubi->buf_mutex); + + ubi_free_vid_buf(vidb); + + if (second_is_newer) + dbg_bld("second PEB %d is newer, copy_flag is set", pnum); + else + dbg_bld("first PEB %d is newer, copy_flag is set", pnum); + + return second_is_newer | (bitflips << 1) | (corrupted << 2); + +out_unlock: + mutex_unlock(&ubi->buf_mutex); +out_free_vidh: + ubi_free_vid_buf(vidb); + return err; +} + +/** + * ubi_add_to_av - add used physical eraseblock to the attaching information. + * @ubi: UBI device description object + * @ai: attaching information + * @pnum: the physical eraseblock number + * @ec: erase counter + * @vid_hdr: the volume identifier header + * @bitflips: if bit-flips were detected when this physical eraseblock was read + * + * This function adds information about a used physical eraseblock to the + * 'used' tree of the corresponding volume. The function is rather complex + * because it has to handle cases when this is not the first physical + * eraseblock belonging to the same logical eraseblock, and the newer one has + * to be picked, while the older one has to be dropped. This function returns + * zero in case of success and a negative error code in case of failure. + */ +int ubi_add_to_av(struct ubi_device *ubi, struct ubi_attach_info *ai, int pnum, + int ec, const struct ubi_vid_hdr *vid_hdr, int bitflips) +{ + int err, vol_id, lnum; + unsigned long long sqnum; + struct ubi_ainf_volume *av; + struct ubi_ainf_peb *aeb; + struct rb_node **p, *parent = NULL; + + vol_id = be32_to_cpu(vid_hdr->vol_id); + lnum = be32_to_cpu(vid_hdr->lnum); + sqnum = be64_to_cpu(vid_hdr->sqnum); + + dbg_bld("PEB %d, LEB %d:%d, EC %d, sqnum %llu, bitflips %d", + pnum, vol_id, lnum, ec, sqnum, bitflips); + + av = add_volume(ai, vol_id, pnum, vid_hdr); + if (IS_ERR(av)) + return PTR_ERR(av); + + if (ai->max_sqnum < sqnum) + ai->max_sqnum = sqnum; + + /* + * Walk the RB-tree of logical eraseblocks of volume @vol_id to look + * if this is the first instance of this logical eraseblock or not. + */ + p = &av->root.rb_node; + while (*p) { + int cmp_res; + + parent = *p; + aeb = rb_entry(parent, struct ubi_ainf_peb, u.rb); + if (lnum != aeb->lnum) { + if (lnum < aeb->lnum) + p = &(*p)->rb_left; + else + p = &(*p)->rb_right; + continue; + } + + /* + * There is already a physical eraseblock describing the same + * logical eraseblock present. + */ + + dbg_bld("this LEB already exists: PEB %d, sqnum %llu, EC %d", + aeb->pnum, aeb->sqnum, aeb->ec); + + /* + * Make sure that the logical eraseblocks have different + * sequence numbers. Otherwise the image is bad. + * + * However, if the sequence number is zero, we assume it must + * be an ancient UBI image from the era when UBI did not have + * sequence numbers. We still can attach these images, unless + * there is a need to distinguish between old and new + * eraseblocks, in which case we'll refuse the image in + * 'ubi_compare_lebs()'. In other words, we attach old clean + * images, but refuse attaching old images with duplicated + * logical eraseblocks because there was an unclean reboot. + */ + if (aeb->sqnum == sqnum && sqnum != 0) { + ubi_err(ubi, "two LEBs with same sequence number %llu", + sqnum); + ubi_dump_aeb(aeb, 0); + ubi_dump_vid_hdr(vid_hdr); + return -EINVAL; + } + + /* + * Now we have to drop the older one and preserve the newer + * one. + */ + cmp_res = ubi_compare_lebs(ubi, aeb, pnum, vid_hdr); + if (cmp_res < 0) + return cmp_res; + + if (cmp_res & 1) { + /* + * This logical eraseblock is newer than the one + * found earlier. + */ + err = validate_vid_hdr(ubi, vid_hdr, av, pnum); + if (err) + return err; + + err = add_to_list(ai, aeb->pnum, aeb->vol_id, + aeb->lnum, aeb->ec, cmp_res & 4, + &ai->erase); + if (err) + return err; + + aeb->ec = ec; + aeb->pnum = pnum; + aeb->vol_id = vol_id; + aeb->lnum = lnum; + aeb->scrub = ((cmp_res & 2) || bitflips); + aeb->copy_flag = vid_hdr->copy_flag; + aeb->sqnum = sqnum; + + if (av->highest_lnum == lnum) + av->last_data_size = + be32_to_cpu(vid_hdr->data_size); + + return 0; + } else { + /* + * This logical eraseblock is older than the one found + * previously. + */ + return add_to_list(ai, pnum, vol_id, lnum, ec, + cmp_res & 4, &ai->erase); + } + } + + /* + * We've met this logical eraseblock for the first time, add it to the + * attaching information. + */ + + err = validate_vid_hdr(ubi, vid_hdr, av, pnum); + if (err) + return err; + + aeb = ubi_alloc_aeb(ai, pnum, ec); + if (!aeb) + return -ENOMEM; + + aeb->vol_id = vol_id; + aeb->lnum = lnum; + aeb->scrub = bitflips; + aeb->copy_flag = vid_hdr->copy_flag; + aeb->sqnum = sqnum; + + if (av->highest_lnum <= lnum) { + av->highest_lnum = lnum; + av->last_data_size = be32_to_cpu(vid_hdr->data_size); + } + + av->leb_count += 1; + rb_link_node(&aeb->u.rb, parent, p); + rb_insert_color(&aeb->u.rb, &av->root); + return 0; +} + +/** + * ubi_add_av - add volume to the attaching information. + * @ai: attaching information + * @vol_id: the requested volume ID + * + * This function returns a pointer to the new volume description or an + * ERR_PTR if the operation failed. + */ +struct ubi_ainf_volume *ubi_add_av(struct ubi_attach_info *ai, int vol_id) +{ + bool created; + + return find_or_add_av(ai, vol_id, AV_ADD, &created); +} + +/** + * ubi_find_av - find volume in the attaching information. + * @ai: attaching information + * @vol_id: the requested volume ID + * + * This function returns a pointer to the volume description or %NULL if there + * are no data about this volume in the attaching information. + */ +struct ubi_ainf_volume *ubi_find_av(const struct ubi_attach_info *ai, + int vol_id) +{ + bool created; + + return find_or_add_av((struct ubi_attach_info *)ai, vol_id, AV_FIND, + &created); +} + +static void destroy_av(struct ubi_attach_info *ai, struct ubi_ainf_volume *av, + struct list_head *list); + +/** + * ubi_remove_av - delete attaching information about a volume. + * @ai: attaching information + * @av: the volume attaching information to delete + */ +void ubi_remove_av(struct ubi_attach_info *ai, struct ubi_ainf_volume *av) +{ + dbg_bld("remove attaching information about volume %d", av->vol_id); + + rb_erase(&av->rb, &ai->volumes); + destroy_av(ai, av, &ai->erase); + ai->vols_found -= 1; +} + +/** + * early_erase_peb - erase a physical eraseblock. + * @ubi: UBI device description object + * @ai: attaching information + * @pnum: physical eraseblock number to erase; + * @ec: erase counter value to write (%UBI_UNKNOWN if it is unknown) + * + * This function erases physical eraseblock 'pnum', and writes the erase + * counter header to it. This function should only be used on UBI device + * initialization stages, when the EBA sub-system had not been yet initialized. + * This function returns zero in case of success and a negative error code in + * case of failure. + */ +static int early_erase_peb(struct ubi_device *ubi, + const struct ubi_attach_info *ai, int pnum, int ec) +{ + int err, torture = 0; + struct ubi_ec_hdr *ec_hdr; + + if ((long long)ec >= UBI_MAX_ERASECOUNTER) { + /* + * Erase counter overflow. Upgrade UBI and use 64-bit + * erase counters internally. + */ + ubi_err(ubi, "erase counter overflow at PEB %d, EC %d", + pnum, ec); + return -EINVAL; + } + + ec_hdr = kzalloc(ubi->ec_hdr_alsize, GFP_KERNEL); + if (!ec_hdr) + return -ENOMEM; + + ec_hdr->ec = cpu_to_be64(ec); + + err = ubi_io_sync_erase(ubi, pnum, &torture); + if (err < 0) + goto out_free; + + err = ubi_io_write_ec_hdr(ubi, pnum, ec_hdr); + +out_free: + kfree(ec_hdr); + return err; +} + +/** + * ubi_early_get_peb - get a free physical eraseblock. + * @ubi: UBI device description object + * @ai: attaching information + * + * This function returns a free physical eraseblock. It is supposed to be + * called on the UBI initialization stages when the wear-leveling sub-system is + * not initialized yet. This function picks a physical eraseblocks from one of + * the lists, writes the EC header if it is needed, and removes it from the + * list. + * + * This function returns a pointer to the "aeb" of the found free PEB in case + * of success and an error code in case of failure. + */ +struct ubi_ainf_peb *ubi_early_get_peb(struct ubi_device *ubi, + struct ubi_attach_info *ai) +{ + int err = 0; + struct ubi_ainf_peb *aeb, *tmp_aeb; + + if (!list_empty(&ai->free)) { + aeb = list_entry(ai->free.next, struct ubi_ainf_peb, u.list); + list_del(&aeb->u.list); + dbg_bld("return free PEB %d, EC %d", aeb->pnum, aeb->ec); + return aeb; + } + + /* + * We try to erase the first physical eraseblock from the erase list + * and pick it if we succeed, or try to erase the next one if not. And + * so forth. We don't want to take care about bad eraseblocks here - + * they'll be handled later. + */ + list_for_each_entry_safe(aeb, tmp_aeb, &ai->erase, u.list) { + if (aeb->ec == UBI_UNKNOWN) + aeb->ec = ai->mean_ec; + + err = early_erase_peb(ubi, ai, aeb->pnum, aeb->ec+1); + if (err) + continue; + + aeb->ec += 1; + list_del(&aeb->u.list); + dbg_bld("return PEB %d, EC %d", aeb->pnum, aeb->ec); + return aeb; + } + + ubi_err(ubi, "no free eraseblocks"); + return ERR_PTR(-ENOSPC); +} + +/** + * check_corruption - check the data area of PEB. + * @ubi: UBI device description object + * @vid_hdr: the (corrupted) VID header of this PEB + * @pnum: the physical eraseblock number to check + * + * This is a helper function which is used to distinguish between VID header + * corruptions caused by power cuts and other reasons. If the PEB contains only + * 0xFF bytes in the data area, the VID header is most probably corrupted + * because of a power cut (%0 is returned in this case). Otherwise, it was + * probably corrupted for some other reasons (%1 is returned in this case). A + * negative error code is returned if a read error occurred. + * + * If the corruption reason was a power cut, UBI can safely erase this PEB. + * Otherwise, it should preserve it to avoid possibly destroying important + * information. + */ +static int check_corruption(struct ubi_device *ubi, struct ubi_vid_hdr *vid_hdr, + int pnum) +{ + int err; + + mutex_lock(&ubi->buf_mutex); + memset(ubi->peb_buf, 0x00, ubi->leb_size); + + err = ubi_io_read(ubi, ubi->peb_buf, pnum, ubi->leb_start, + ubi->leb_size); + if (err == UBI_IO_BITFLIPS || mtd_is_eccerr(err)) { + /* + * Bit-flips or integrity errors while reading the data area. + * It is difficult to say for sure what type of corruption is + * this, but presumably a power cut happened while this PEB was + * erased, so it became unstable and corrupted, and should be + * erased. + */ + err = 0; + goto out_unlock; + } + + if (err) + goto out_unlock; + + if (ubi_check_pattern(ubi->peb_buf, 0xFF, ubi->leb_size)) + goto out_unlock; + + ubi_err(ubi, "PEB %d contains corrupted VID header, and the data does not contain all 0xFF", + pnum); + ubi_err(ubi, "this may be a non-UBI PEB or a severe VID header corruption which requires manual inspection"); + ubi_dump_vid_hdr(vid_hdr); + pr_err("hexdump of PEB %d offset %d, length %d", + pnum, ubi->leb_start, ubi->leb_size); + ubi_dbg_print_hex_dump(KERN_DEBUG, "", DUMP_PREFIX_OFFSET, 32, 1, + ubi->peb_buf, ubi->leb_size, 1); + err = 1; + +out_unlock: + mutex_unlock(&ubi->buf_mutex); + return err; +} + +static bool vol_ignored(int vol_id) +{ + switch (vol_id) { + case UBI_LAYOUT_VOLUME_ID: + return true; + } + +#ifdef CONFIG_MTD_UBI_FASTMAP + return ubi_is_fm_vol(vol_id); +#else + return false; +#endif +} + +/** + * scan_peb - scan and process UBI headers of a PEB. + * @ubi: UBI device description object + * @ai: attaching information + * @pnum: the physical eraseblock number + * @fast: true if we're scanning for a Fastmap + * + * This function reads UBI headers of PEB @pnum, checks them, and adds + * information about this PEB to the corresponding list or RB-tree in the + * "attaching info" structure. Returns zero if the physical eraseblock was + * successfully handled and a negative error code in case of failure. + */ +static int scan_peb(struct ubi_device *ubi, struct ubi_attach_info *ai, + int pnum, bool fast) +{ + struct ubi_ec_hdr *ech = ai->ech; + struct ubi_vid_io_buf *vidb = ai->vidb; + struct ubi_vid_hdr *vidh = ubi_get_vid_hdr(vidb); + long long ec; + int err, bitflips = 0, vol_id = -1, ec_err = 0; + + dbg_bld("scan PEB %d", pnum); + + /* Skip bad physical eraseblocks */ + err = ubi_io_is_bad(ubi, pnum); + if (err < 0) + return err; + else if (err) { + ai->bad_peb_count += 1; + return 0; + } + + err = ubi_io_read_ec_hdr(ubi, pnum, ech, 0); + if (err < 0) + return err; + switch (err) { + case 0: + break; + case UBI_IO_BITFLIPS: + bitflips = 1; + break; + case UBI_IO_FF: + ai->empty_peb_count += 1; + return add_to_list(ai, pnum, UBI_UNKNOWN, UBI_UNKNOWN, + UBI_UNKNOWN, 0, &ai->erase); + case UBI_IO_FF_BITFLIPS: + ai->empty_peb_count += 1; + return add_to_list(ai, pnum, UBI_UNKNOWN, UBI_UNKNOWN, + UBI_UNKNOWN, 1, &ai->erase); + case UBI_IO_BAD_HDR_EBADMSG: + case UBI_IO_BAD_HDR: + /* + * We have to also look at the VID header, possibly it is not + * corrupted. Set %bitflips flag in order to make this PEB be + * moved and EC be re-created. + */ + ec_err = err; + ec = UBI_UNKNOWN; + bitflips = 1; + break; + default: + ubi_err(ubi, "'ubi_io_read_ec_hdr()' returned unknown code %d", + err); + return -EINVAL; + } + + if (!ec_err) { + int image_seq; + + /* Make sure UBI version is OK */ + if (ech->version != UBI_VERSION) { + ubi_err(ubi, "this UBI version is %d, image version is %d", + UBI_VERSION, (int)ech->version); + return -EINVAL; + } + + ec = be64_to_cpu(ech->ec); + if (ec > UBI_MAX_ERASECOUNTER) { + /* + * Erase counter overflow. The EC headers have 64 bits + * reserved, but we anyway make use of only 31 bit + * values, as this seems to be enough for any existing + * flash. Upgrade UBI and use 64-bit erase counters + * internally. + */ + ubi_err(ubi, "erase counter overflow, max is %d", + UBI_MAX_ERASECOUNTER); + ubi_dump_ec_hdr(ech); + return -EINVAL; + } + + /* + * Make sure that all PEBs have the same image sequence number. + * This allows us to detect situations when users flash UBI + * images incorrectly, so that the flash has the new UBI image + * and leftovers from the old one. This feature was added + * relatively recently, and the sequence number was always + * zero, because old UBI implementations always set it to zero. + * For this reasons, we do not panic if some PEBs have zero + * sequence number, while other PEBs have non-zero sequence + * number. + */ + image_seq = be32_to_cpu(ech->image_seq); + if (!ubi->image_seq) + ubi->image_seq = image_seq; + if (image_seq && ubi->image_seq != image_seq) { + ubi_err(ubi, "bad image sequence number %d in PEB %d, expected %d", + image_seq, pnum, ubi->image_seq); + ubi_dump_ec_hdr(ech); + return -EINVAL; + } + } + + /* OK, we've done with the EC header, let's look at the VID header */ + + err = ubi_io_read_vid_hdr(ubi, pnum, vidb, 0); + if (err < 0) + return err; + switch (err) { + case 0: + break; + case UBI_IO_BITFLIPS: + bitflips = 1; + break; + case UBI_IO_BAD_HDR_EBADMSG: + if (ec_err == UBI_IO_BAD_HDR_EBADMSG) + /* + * Both EC and VID headers are corrupted and were read + * with data integrity error, probably this is a bad + * PEB, bit it is not marked as bad yet. This may also + * be a result of power cut during erasure. + */ + ai->maybe_bad_peb_count += 1; + fallthrough; + case UBI_IO_BAD_HDR: + /* + * If we're facing a bad VID header we have to drop *all* + * Fastmap data structures we find. The most recent Fastmap + * could be bad and therefore there is a chance that we attach + * from an old one. On a fine MTD stack a PEB must not render + * bad all of a sudden, but the reality is different. + * So, let's be paranoid and help finding the root cause by + * falling back to scanning mode instead of attaching with a + * bad EBA table and cause data corruption which is hard to + * analyze. + */ + if (fast) + ai->force_full_scan = 1; + + if (ec_err) + /* + * Both headers are corrupted. There is a possibility + * that this a valid UBI PEB which has corresponding + * LEB, but the headers are corrupted. However, it is + * impossible to distinguish it from a PEB which just + * contains garbage because of a power cut during erase + * operation. So we just schedule this PEB for erasure. + * + * Besides, in case of NOR flash, we deliberately + * corrupt both headers because NOR flash erasure is + * slow and can start from the end. + */ + err = 0; + else + /* + * The EC was OK, but the VID header is corrupted. We + * have to check what is in the data area. + */ + err = check_corruption(ubi, vidh, pnum); + + if (err < 0) + return err; + else if (!err) + /* This corruption is caused by a power cut */ + err = add_to_list(ai, pnum, UBI_UNKNOWN, + UBI_UNKNOWN, ec, 1, &ai->erase); + else + /* This is an unexpected corruption */ + err = add_corrupted(ai, pnum, ec); + if (err) + return err; + goto adjust_mean_ec; + case UBI_IO_FF_BITFLIPS: + err = add_to_list(ai, pnum, UBI_UNKNOWN, UBI_UNKNOWN, + ec, 1, &ai->erase); + if (err) + return err; + goto adjust_mean_ec; + case UBI_IO_FF: + if (ec_err || bitflips) + err = add_to_list(ai, pnum, UBI_UNKNOWN, + UBI_UNKNOWN, ec, 1, &ai->erase); + else + err = add_to_list(ai, pnum, UBI_UNKNOWN, + UBI_UNKNOWN, ec, 0, &ai->free); + if (err) + return err; + goto adjust_mean_ec; + default: + ubi_err(ubi, "'ubi_io_read_vid_hdr()' returned unknown code %d", + err); + return -EINVAL; + } + + vol_id = be32_to_cpu(vidh->vol_id); + if (vol_id > UBI_MAX_VOLUMES && !vol_ignored(vol_id)) { + int lnum = be32_to_cpu(vidh->lnum); + + /* Unsupported internal volume */ + switch (vidh->compat) { + case UBI_COMPAT_DELETE: + ubi_msg(ubi, "\"delete\" compatible internal volume %d:%d found, will remove it", + vol_id, lnum); + + err = add_to_list(ai, pnum, vol_id, lnum, + ec, 1, &ai->erase); + if (err) + return err; + return 0; + + case UBI_COMPAT_RO: + ubi_msg(ubi, "read-only compatible internal volume %d:%d found, switch to read-only mode", + vol_id, lnum); + ubi->ro_mode = 1; + break; + + case UBI_COMPAT_PRESERVE: + ubi_msg(ubi, "\"preserve\" compatible internal volume %d:%d found", + vol_id, lnum); + err = add_to_list(ai, pnum, vol_id, lnum, + ec, 0, &ai->alien); + if (err) + return err; + return 0; + + case UBI_COMPAT_REJECT: + ubi_err(ubi, "incompatible internal volume %d:%d found", + vol_id, lnum); + return -EINVAL; + } + } + + if (ec_err) + ubi_warn(ubi, "valid VID header but corrupted EC header at PEB %d", + pnum); + + if (ubi_is_fm_vol(vol_id)) + err = add_fastmap(ai, pnum, vidh, ec); + else + err = ubi_add_to_av(ubi, ai, pnum, ec, vidh, bitflips); + + if (err) + return err; + +adjust_mean_ec: + if (!ec_err) { + ai->ec_sum += ec; + ai->ec_count += 1; + if (ec > ai->max_ec) + ai->max_ec = ec; + if (ec < ai->min_ec) + ai->min_ec = ec; + } + + return 0; +} + +/** + * late_analysis - analyze the overall situation with PEB. + * @ubi: UBI device description object + * @ai: attaching information + * + * This is a helper function which takes a look what PEBs we have after we + * gather information about all of them ("ai" is compete). It decides whether + * the flash is empty and should be formatted of whether there are too many + * corrupted PEBs and we should not attach this MTD device. Returns zero if we + * should proceed with attaching the MTD device, and %-EINVAL if we should not. + */ +static int late_analysis(struct ubi_device *ubi, struct ubi_attach_info *ai) +{ + struct ubi_ainf_peb *aeb; + int max_corr, peb_count; + + peb_count = ubi->peb_count - ai->bad_peb_count - ai->alien_peb_count; + max_corr = peb_count / 20 ?: 8; + + /* + * Few corrupted PEBs is not a problem and may be just a result of + * unclean reboots. However, many of them may indicate some problems + * with the flash HW or driver. + */ + if (ai->corr_peb_count) { + ubi_err(ubi, "%d PEBs are corrupted and preserved", + ai->corr_peb_count); + pr_err("Corrupted PEBs are:"); + list_for_each_entry(aeb, &ai->corr, u.list) + pr_cont(" %d", aeb->pnum); + pr_cont("\n"); + + /* + * If too many PEBs are corrupted, we refuse attaching, + * otherwise, only print a warning. + */ + if (ai->corr_peb_count >= max_corr) { + ubi_err(ubi, "too many corrupted PEBs, refusing"); + return -EINVAL; + } + } + + if (ai->empty_peb_count + ai->maybe_bad_peb_count == peb_count) { + /* + * All PEBs are empty, or almost all - a couple PEBs look like + * they may be bad PEBs which were not marked as bad yet. + * + * This piece of code basically tries to distinguish between + * the following situations: + * + * 1. Flash is empty, but there are few bad PEBs, which are not + * marked as bad so far, and which were read with error. We + * want to go ahead and format this flash. While formatting, + * the faulty PEBs will probably be marked as bad. + * + * 2. Flash contains non-UBI data and we do not want to format + * it and destroy possibly important information. + */ + if (ai->maybe_bad_peb_count <= 2) { + ai->is_empty = 1; + ubi_msg(ubi, "empty MTD device detected"); + get_random_bytes(&ubi->image_seq, + sizeof(ubi->image_seq)); + } else { + ubi_err(ubi, "MTD device is not UBI-formatted and possibly contains non-UBI data - refusing it"); + return -EINVAL; + } + + } + + return 0; +} + +/** + * destroy_av - free volume attaching information. + * @av: volume attaching information + * @ai: attaching information + * @list: put the aeb elements in there if !NULL, otherwise free them + * + * This function destroys the volume attaching information. + */ +static void destroy_av(struct ubi_attach_info *ai, struct ubi_ainf_volume *av, + struct list_head *list) +{ + struct ubi_ainf_peb *aeb; + struct rb_node *this = av->root.rb_node; + + while (this) { + if (this->rb_left) + this = this->rb_left; + else if (this->rb_right) + this = this->rb_right; + else { + aeb = rb_entry(this, struct ubi_ainf_peb, u.rb); + this = rb_parent(this); + if (this) { + if (this->rb_left == &aeb->u.rb) + this->rb_left = NULL; + else + this->rb_right = NULL; + } + + if (list) + list_add_tail(&aeb->u.list, list); + else + ubi_free_aeb(ai, aeb); + } + } + kfree(av); +} + +/** + * destroy_ai - destroy attaching information. + * @ai: attaching information + */ +static void destroy_ai(struct ubi_attach_info *ai) +{ + struct ubi_ainf_peb *aeb, *aeb_tmp; + struct ubi_ainf_volume *av; + struct rb_node *rb; + + list_for_each_entry_safe(aeb, aeb_tmp, &ai->alien, u.list) { + list_del(&aeb->u.list); + ubi_free_aeb(ai, aeb); + } + list_for_each_entry_safe(aeb, aeb_tmp, &ai->erase, u.list) { + list_del(&aeb->u.list); + ubi_free_aeb(ai, aeb); + } + list_for_each_entry_safe(aeb, aeb_tmp, &ai->corr, u.list) { + list_del(&aeb->u.list); + ubi_free_aeb(ai, aeb); + } + list_for_each_entry_safe(aeb, aeb_tmp, &ai->free, u.list) { + list_del(&aeb->u.list); + ubi_free_aeb(ai, aeb); + } + list_for_each_entry_safe(aeb, aeb_tmp, &ai->fastmap, u.list) { + list_del(&aeb->u.list); + ubi_free_aeb(ai, aeb); + } + + /* Destroy the volume RB-tree */ + rb = ai->volumes.rb_node; + while (rb) { + if (rb->rb_left) + rb = rb->rb_left; + else if (rb->rb_right) + rb = rb->rb_right; + else { + av = rb_entry(rb, struct ubi_ainf_volume, rb); + + rb = rb_parent(rb); + if (rb) { + if (rb->rb_left == &av->rb) + rb->rb_left = NULL; + else + rb->rb_right = NULL; + } + + destroy_av(ai, av, NULL); + } + } + + kmem_cache_destroy(ai->aeb_slab_cache); + kfree(ai); +} + +/** + * scan_all - scan entire MTD device. + * @ubi: UBI device description object + * @ai: attach info object + * @start: start scanning at this PEB + * + * This function does full scanning of an MTD device and returns complete + * information about it in form of a "struct ubi_attach_info" object. In case + * of failure, an error code is returned. + */ +static int scan_all(struct ubi_device *ubi, struct ubi_attach_info *ai, + int start) +{ + int err, pnum; + struct rb_node *rb1, *rb2; + struct ubi_ainf_volume *av; + struct ubi_ainf_peb *aeb; + + err = -ENOMEM; + + ai->ech = kzalloc(ubi->ec_hdr_alsize, GFP_KERNEL); + if (!ai->ech) + return err; + + ai->vidb = ubi_alloc_vid_buf(ubi, GFP_KERNEL); + if (!ai->vidb) + goto out_ech; + + for (pnum = start; pnum < ubi->peb_count; pnum++) { + cond_resched(); + + dbg_gen("process PEB %d", pnum); + err = scan_peb(ubi, ai, pnum, false); + if (err < 0) + goto out_vidh; + } + + ubi_msg(ubi, "scanning is finished"); + + /* Calculate mean erase counter */ + if (ai->ec_count) + ai->mean_ec = div_u64(ai->ec_sum, ai->ec_count); + + err = late_analysis(ubi, ai); + if (err) + goto out_vidh; + + /* + * In case of unknown erase counter we use the mean erase counter + * value. + */ + ubi_rb_for_each_entry(rb1, av, &ai->volumes, rb) { + ubi_rb_for_each_entry(rb2, aeb, &av->root, u.rb) + if (aeb->ec == UBI_UNKNOWN) + aeb->ec = ai->mean_ec; + } + + list_for_each_entry(aeb, &ai->free, u.list) { + if (aeb->ec == UBI_UNKNOWN) + aeb->ec = ai->mean_ec; + } + + list_for_each_entry(aeb, &ai->corr, u.list) + if (aeb->ec == UBI_UNKNOWN) + aeb->ec = ai->mean_ec; + + list_for_each_entry(aeb, &ai->erase, u.list) + if (aeb->ec == UBI_UNKNOWN) + aeb->ec = ai->mean_ec; + + err = self_check_ai(ubi, ai); + if (err) + goto out_vidh; + + ubi_free_vid_buf(ai->vidb); + kfree(ai->ech); + + return 0; + +out_vidh: + ubi_free_vid_buf(ai->vidb); +out_ech: + kfree(ai->ech); + return err; +} + +static struct ubi_attach_info *alloc_ai(const char *slab_name) +{ + struct ubi_attach_info *ai; + + ai = kzalloc_obj(struct ubi_attach_info); + if (!ai) + return ai; + + INIT_LIST_HEAD(&ai->corr); + INIT_LIST_HEAD(&ai->free); + INIT_LIST_HEAD(&ai->erase); + INIT_LIST_HEAD(&ai->alien); + INIT_LIST_HEAD(&ai->fastmap); + ai->volumes = RB_ROOT; + ai->aeb_slab_cache = kmem_cache_create(slab_name, + sizeof(struct ubi_ainf_peb), + 0, 0, NULL); + if (!ai->aeb_slab_cache) { + kfree(ai); + ai = NULL; + } + + return ai; +} + +#ifdef CONFIG_MTD_UBI_FASTMAP + +/** + * scan_fast - try to find a fastmap and attach from it. + * @ubi: UBI device description object + * @ai: attach info object + * + * Returns 0 on success, negative return values indicate an internal + * error. + * UBI_NO_FASTMAP denotes that no fastmap was found. + * UBI_BAD_FASTMAP denotes that the found fastmap was invalid. + */ +static int scan_fast(struct ubi_device *ubi, struct ubi_attach_info **ai) +{ + int err, pnum; + struct ubi_attach_info *scan_ai; + + err = -ENOMEM; + + scan_ai = alloc_ai("ubi_aeb_slab_cache_fastmap"); + if (!scan_ai) + goto out; + + scan_ai->ech = kzalloc(ubi->ec_hdr_alsize, GFP_KERNEL); + if (!scan_ai->ech) + goto out_ai; + + scan_ai->vidb = ubi_alloc_vid_buf(ubi, GFP_KERNEL); + if (!scan_ai->vidb) + goto out_ech; + + for (pnum = 0; pnum < UBI_FM_MAX_START; pnum++) { + cond_resched(); + + dbg_gen("process PEB %d", pnum); + err = scan_peb(ubi, scan_ai, pnum, true); + if (err < 0) + goto out_vidh; + } + + ubi_free_vid_buf(scan_ai->vidb); + kfree(scan_ai->ech); + + if (scan_ai->force_full_scan) + err = UBI_NO_FASTMAP; + else + err = ubi_scan_fastmap(ubi, *ai, scan_ai); + + if (err) { + /* + * Didn't attach via fastmap, do a full scan but reuse what + * we've aready scanned. + */ + destroy_ai(*ai); + *ai = scan_ai; + } else + destroy_ai(scan_ai); + + return err; + +out_vidh: + ubi_free_vid_buf(scan_ai->vidb); +out_ech: + kfree(scan_ai->ech); +out_ai: + destroy_ai(scan_ai); +out: + return err; +} + +#endif + +/** + * ubi_attach - attach an MTD device. + * @ubi: UBI device descriptor + * @force_scan: if set to non-zero attach by scanning + * + * This function returns zero in case of success and a negative error code in + * case of failure. + */ +int ubi_attach(struct ubi_device *ubi, int force_scan) +{ + int err; + struct ubi_attach_info *ai; + + ai = alloc_ai("ubi_aeb_slab_cache"); + if (!ai) + return -ENOMEM; + +#ifdef CONFIG_MTD_UBI_FASTMAP + /* On small flash devices we disable fastmap in any case. */ + if ((int)mtd_div_by_eb(ubi->mtd->size, ubi->mtd) <= UBI_FM_MAX_START) { + ubi->fm_disabled = 1; + force_scan = 1; + } + + if (force_scan) + err = scan_all(ubi, ai, 0); + else { + err = scan_fast(ubi, &ai); + if (err > 0 || mtd_is_eccerr(err)) { + if (err != UBI_NO_FASTMAP) { + destroy_ai(ai); + ai = alloc_ai("ubi_aeb_slab_cache"); + if (!ai) + return -ENOMEM; + + err = scan_all(ubi, ai, 0); + } else { + err = scan_all(ubi, ai, UBI_FM_MAX_START); + } + } + } +#else + err = scan_all(ubi, ai, 0); +#endif + if (err) + goto out_ai; + + ubi->bad_peb_count = ai->bad_peb_count; + ubi->good_peb_count = ubi->peb_count - ubi->bad_peb_count; + ubi->corr_peb_count = ai->corr_peb_count; + ubi->max_ec = ai->max_ec; + ubi->mean_ec = ai->mean_ec; + dbg_gen("max. sequence number: %llu", ai->max_sqnum); + + err = ubi_read_volume_table(ubi, ai); + if (err) + goto out_fm; + + err = ubi_wl_init(ubi, ai); + if (err) + goto out_vtbl; + + err = ubi_eba_init(ubi, ai); + if (err) + goto out_wl; + +#ifdef CONFIG_MTD_UBI_FASTMAP + if (ubi->fm && ubi_dbg_chk_fastmap(ubi)) { + struct ubi_attach_info *scan_ai; + + scan_ai = alloc_ai("ubi_aeb_slab_cache_dbg_chk_fastmap"); + if (!scan_ai) { + err = -ENOMEM; + goto out_wl; + } + + err = scan_all(ubi, scan_ai, 0); + if (err) { + destroy_ai(scan_ai); + goto out_wl; + } + + err = self_check_eba(ubi, ai, scan_ai); + destroy_ai(scan_ai); + + if (err) + goto out_wl; + } +#endif + + destroy_ai(ai); + return 0; + +out_wl: + ubi_wl_close(ubi); +out_vtbl: + ubi_free_all_volumes(ubi); + vfree(ubi->vtbl); +out_fm: + ubi_free_fastmap(ubi); +out_ai: + destroy_ai(ai); + return err; +} + +/** + * self_check_ai - check the attaching information. + * @ubi: UBI device description object + * @ai: attaching information + * + * This function returns zero if the attaching information is all right, and a + * negative error code if not or if an error occurred. + */ +static int self_check_ai(struct ubi_device *ubi, struct ubi_attach_info *ai) +{ + struct ubi_vid_io_buf *vidb = ai->vidb; + struct ubi_vid_hdr *vidh = ubi_get_vid_hdr(vidb); + int pnum, err, vols_found = 0; + struct rb_node *rb1, *rb2; + struct ubi_ainf_volume *av; + struct ubi_ainf_peb *aeb, *last_aeb; + uint8_t *buf; + + if (!ubi_dbg_chk_gen(ubi)) + return 0; + + /* + * At first, check that attaching information is OK. + */ + ubi_rb_for_each_entry(rb1, av, &ai->volumes, rb) { + int leb_count = 0; + + cond_resched(); + + vols_found += 1; + + if (ai->is_empty) { + ubi_err(ubi, "bad is_empty flag"); + goto bad_av; + } + + if (av->vol_id < 0 || av->highest_lnum < 0 || + av->leb_count < 0 || av->vol_type < 0 || av->used_ebs < 0 || + av->data_pad < 0 || av->last_data_size < 0) { + ubi_err(ubi, "negative values"); + goto bad_av; + } + + if (av->vol_id >= UBI_MAX_VOLUMES && + av->vol_id < UBI_INTERNAL_VOL_START) { + ubi_err(ubi, "bad vol_id"); + goto bad_av; + } + + if (av->vol_id > ai->highest_vol_id) { + ubi_err(ubi, "highest_vol_id is %d, but vol_id %d is there", + ai->highest_vol_id, av->vol_id); + goto out; + } + + if (av->vol_type != UBI_DYNAMIC_VOLUME && + av->vol_type != UBI_STATIC_VOLUME) { + ubi_err(ubi, "bad vol_type"); + goto bad_av; + } + + if (av->data_pad > ubi->leb_size / 2) { + ubi_err(ubi, "bad data_pad"); + goto bad_av; + } + + last_aeb = NULL; + ubi_rb_for_each_entry(rb2, aeb, &av->root, u.rb) { + cond_resched(); + + last_aeb = aeb; + leb_count += 1; + + if (aeb->pnum < 0 || aeb->ec < 0) { + ubi_err(ubi, "negative values"); + goto bad_aeb; + } + + if (aeb->ec < ai->min_ec) { + ubi_err(ubi, "bad ai->min_ec (%d), %d found", + ai->min_ec, aeb->ec); + goto bad_aeb; + } + + if (aeb->ec > ai->max_ec) { + ubi_err(ubi, "bad ai->max_ec (%d), %d found", + ai->max_ec, aeb->ec); + goto bad_aeb; + } + + if (aeb->pnum >= ubi->peb_count) { + ubi_err(ubi, "too high PEB number %d, total PEBs %d", + aeb->pnum, ubi->peb_count); + goto bad_aeb; + } + + if (av->vol_type == UBI_STATIC_VOLUME) { + if (aeb->lnum >= av->used_ebs) { + ubi_err(ubi, "bad lnum or used_ebs"); + goto bad_aeb; + } + } else { + if (av->used_ebs != 0) { + ubi_err(ubi, "non-zero used_ebs"); + goto bad_aeb; + } + } + + if (aeb->lnum > av->highest_lnum) { + ubi_err(ubi, "incorrect highest_lnum or lnum"); + goto bad_aeb; + } + } + + if (av->leb_count != leb_count) { + ubi_err(ubi, "bad leb_count, %d objects in the tree", + leb_count); + goto bad_av; + } + + if (!last_aeb) + continue; + + aeb = last_aeb; + + if (aeb->lnum != av->highest_lnum) { + ubi_err(ubi, "bad highest_lnum"); + goto bad_aeb; + } + } + + if (vols_found != ai->vols_found) { + ubi_err(ubi, "bad ai->vols_found %d, should be %d", + ai->vols_found, vols_found); + goto out; + } + + /* Check that attaching information is correct */ + ubi_rb_for_each_entry(rb1, av, &ai->volumes, rb) { + last_aeb = NULL; + ubi_rb_for_each_entry(rb2, aeb, &av->root, u.rb) { + int vol_type; + + cond_resched(); + + last_aeb = aeb; + + err = ubi_io_read_vid_hdr(ubi, aeb->pnum, vidb, 1); + if (err && err != UBI_IO_BITFLIPS) { + ubi_err(ubi, "VID header is not OK (%d)", + err); + if (err > 0) + err = -EIO; + return err; + } + + vol_type = vidh->vol_type == UBI_VID_DYNAMIC ? + UBI_DYNAMIC_VOLUME : UBI_STATIC_VOLUME; + if (av->vol_type != vol_type) { + ubi_err(ubi, "bad vol_type"); + goto bad_vid_hdr; + } + + if (aeb->sqnum != be64_to_cpu(vidh->sqnum)) { + ubi_err(ubi, "bad sqnum %llu", aeb->sqnum); + goto bad_vid_hdr; + } + + if (av->vol_id != be32_to_cpu(vidh->vol_id)) { + ubi_err(ubi, "bad vol_id %d", av->vol_id); + goto bad_vid_hdr; + } + + if (av->compat != vidh->compat) { + ubi_err(ubi, "bad compat %d", vidh->compat); + goto bad_vid_hdr; + } + + if (aeb->lnum != be32_to_cpu(vidh->lnum)) { + ubi_err(ubi, "bad lnum %d", aeb->lnum); + goto bad_vid_hdr; + } + + if (av->used_ebs != be32_to_cpu(vidh->used_ebs)) { + ubi_err(ubi, "bad used_ebs %d", av->used_ebs); + goto bad_vid_hdr; + } + + if (av->data_pad != be32_to_cpu(vidh->data_pad)) { + ubi_err(ubi, "bad data_pad %d", av->data_pad); + goto bad_vid_hdr; + } + } + + if (!last_aeb) + continue; + + if (av->highest_lnum != be32_to_cpu(vidh->lnum)) { + ubi_err(ubi, "bad highest_lnum %d", av->highest_lnum); + goto bad_vid_hdr; + } + + if (av->last_data_size != be32_to_cpu(vidh->data_size)) { + ubi_err(ubi, "bad last_data_size %d", + av->last_data_size); + goto bad_vid_hdr; + } + } + + /* + * Make sure that all the physical eraseblocks are in one of the lists + * or trees. + */ + buf = kzalloc(ubi->peb_count, GFP_KERNEL); + if (!buf) + return -ENOMEM; + + for (pnum = 0; pnum < ubi->peb_count; pnum++) { + err = ubi_io_is_bad(ubi, pnum); + if (err < 0) { + kfree(buf); + return err; + } else if (err) + buf[pnum] = 1; + } + + ubi_rb_for_each_entry(rb1, av, &ai->volumes, rb) + ubi_rb_for_each_entry(rb2, aeb, &av->root, u.rb) + buf[aeb->pnum] = 1; + + list_for_each_entry(aeb, &ai->free, u.list) + buf[aeb->pnum] = 1; + + list_for_each_entry(aeb, &ai->corr, u.list) + buf[aeb->pnum] = 1; + + list_for_each_entry(aeb, &ai->erase, u.list) + buf[aeb->pnum] = 1; + + list_for_each_entry(aeb, &ai->alien, u.list) + buf[aeb->pnum] = 1; + + err = 0; + for (pnum = 0; pnum < ubi->peb_count; pnum++) + if (!buf[pnum]) { + ubi_err(ubi, "PEB %d is not referred", pnum); + err = 1; + } + + kfree(buf); + if (err) + goto out; + return 0; + +bad_aeb: + ubi_err(ubi, "bad attaching information about LEB %d", aeb->lnum); + ubi_dump_aeb(aeb, 0); + ubi_dump_av(av); + goto out; + +bad_av: + ubi_err(ubi, "bad attaching information about volume %d", av->vol_id); + ubi_dump_av(av); + goto out; + +bad_vid_hdr: + ubi_err(ubi, "bad attaching information about volume %d", av->vol_id); + ubi_dump_av(av); + ubi_dump_vid_hdr(vidh); + +out: + dump_stack(); + return -EINVAL; +} diff --git a/drivers/mtd/ubi/block.c b/drivers/mtd/ubi/block.c new file mode 100644 index 000000000..29c0d6941 --- /dev/null +++ b/drivers/mtd/ubi/block.c @@ -0,0 +1,678 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * Copyright (c) 2014 Ezequiel Garcia + * Copyright (c) 2011 Free Electrons + * + * Driver parameter handling strongly based on drivers/mtd/ubi/build.c + * Copyright (c) International Business Machines Corp., 2006 + * Copyright (c) Nokia Corporation, 2007 + * Authors: Artem Bityutskiy, Frank Haverkamp + */ + +/* + * Read-only block devices on top of UBI volumes + * + * A simple implementation to allow a block device to be layered on top of a + * UBI volume. The implementation is provided by creating a static 1-to-1 + * mapping between the block device and the UBI volume. + * + * The addressed byte is obtained from the addressed block sector, which is + * mapped linearly into the corresponding LEB: + * + * LEB number = addressed byte / LEB size + * + * This feature is compiled in the UBI core, and adds a 'block' parameter + * to allow early creation of block devices on top of UBI volumes. Runtime + * block creation/removal for UBI volumes is provided through two UBI ioctls: + * UBI_IOCVOLCRBLK and UBI_IOCVOLRMBLK. + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include "ubi-media.h" +#include "ubi.h" + +/* Maximum number of supported devices */ +#define UBIBLOCK_MAX_DEVICES 32 + +/* Maximum length of the 'block=' parameter */ +#define UBIBLOCK_PARAM_LEN 63 + +/* Maximum number of comma-separated items in the 'block=' parameter */ +#define UBIBLOCK_PARAM_COUNT 2 + +struct ubiblock_param { + int ubi_num; + int vol_id; + char name[UBIBLOCK_PARAM_LEN+1]; +}; + +struct ubiblock_pdu { + struct ubi_sgl usgl; +}; + +/* Numbers of elements set in the @ubiblock_param array */ +static int ubiblock_devs; + +/* MTD devices specification parameters */ +static struct ubiblock_param ubiblock_param[UBIBLOCK_MAX_DEVICES]; + +struct ubiblock { + struct ubi_volume_desc *desc; + int ubi_num; + int vol_id; + int refcnt; + int leb_size; + + struct gendisk *gd; + struct request_queue *rq; + + struct mutex dev_mutex; + struct list_head list; + struct blk_mq_tag_set tag_set; +}; + +/* Linked list of all ubiblock instances */ +static LIST_HEAD(ubiblock_devices); +static DEFINE_IDR(ubiblock_minor_idr); +/* Protects ubiblock_devices and ubiblock_minor_idr */ +static DEFINE_MUTEX(devices_mutex); +static int ubiblock_major; + +static int __init ubiblock_set_param(const char *val, + const struct kernel_param *kp) +{ + int i, ret; + size_t len; + struct ubiblock_param *param; + char buf[UBIBLOCK_PARAM_LEN]; + char *pbuf = &buf[0]; + char *tokens[UBIBLOCK_PARAM_COUNT]; + + if (!val) + return -EINVAL; + + len = strnlen(val, UBIBLOCK_PARAM_LEN); + if (len == 0) { + pr_warn("UBI: block: empty 'block=' parameter - ignored\n"); + return 0; + } + + if (len == UBIBLOCK_PARAM_LEN) { + pr_err("UBI: block: parameter \"%s\" is too long, max. is %d\n", + val, UBIBLOCK_PARAM_LEN); + return -EINVAL; + } + + strcpy(buf, val); + + /* Get rid of the final newline */ + if (buf[len - 1] == '\n') + buf[len - 1] = '\0'; + + for (i = 0; i < UBIBLOCK_PARAM_COUNT; i++) + tokens[i] = strsep(&pbuf, ","); + + param = &ubiblock_param[ubiblock_devs]; + if (tokens[1]) { + /* Two parameters: can be 'ubi, vol_id' or 'ubi, vol_name' */ + ret = kstrtoint(tokens[0], 10, ¶m->ubi_num); + if (ret < 0) + return -EINVAL; + + /* Second param can be a number or a name */ + ret = kstrtoint(tokens[1], 10, ¶m->vol_id); + if (ret < 0) { + param->vol_id = -1; + strcpy(param->name, tokens[1]); + } + + } else { + /* One parameter: must be device path */ + strcpy(param->name, tokens[0]); + param->ubi_num = -1; + param->vol_id = -1; + } + + ubiblock_devs++; + + return 0; +} + +static const struct kernel_param_ops ubiblock_param_ops = { + .set = ubiblock_set_param, +}; +module_param_cb(block, &ubiblock_param_ops, NULL, 0); +MODULE_PARM_DESC(block, "Attach block devices to UBI volumes. Parameter format: block=.\n" + "Multiple \"block\" parameters may be specified.\n" + "UBI volumes may be specified by their number, name, or path to the device node.\n" + "Examples\n" + "Using the UBI volume path:\n" + "ubi.block=/dev/ubi0_0\n" + "Using the UBI device, and the volume name:\n" + "ubi.block=0,rootfs\n" + "Using both UBI device number and UBI volume number:\n" + "ubi.block=0,0\n"); + +static struct ubiblock *find_dev_nolock(int ubi_num, int vol_id) +{ + struct ubiblock *dev; + + list_for_each_entry(dev, &ubiblock_devices, list) + if (dev->ubi_num == ubi_num && dev->vol_id == vol_id) + return dev; + return NULL; +} + +static blk_status_t ubiblock_read(struct request *req) +{ + struct ubiblock_pdu *pdu = blk_mq_rq_to_pdu(req); + struct ubiblock *dev = req->q->queuedata; + u64 pos = blk_rq_pos(req) << 9; + int to_read = blk_rq_bytes(req); + int bytes_left = to_read; + /* Get LEB:offset address to read from */ + int offset = do_div(pos, dev->leb_size); + int leb = pos; + struct req_iterator iter; + struct bio_vec bvec; + int ret; + + blk_mq_start_request(req); + + /* + * It is safe to ignore the return value of blk_rq_map_sg() because + * the number of sg entries is limited to UBI_MAX_SG_COUNT + * and ubi_read_sg() will check that limit. + */ + ubi_sgl_init(&pdu->usgl); + blk_rq_map_sg(req, pdu->usgl.sg); + + while (bytes_left) { + /* + * We can only read one LEB at a time. Therefore if the read + * length is larger than one LEB size, we split the operation. + */ + if (offset + to_read > dev->leb_size) + to_read = dev->leb_size - offset; + + ret = ubi_read_sg(dev->desc, leb, &pdu->usgl, offset, to_read); + if (ret < 0) + break; + + bytes_left -= to_read; + to_read = bytes_left; + leb += 1; + offset = 0; + } + + rq_for_each_segment(bvec, req, iter) + flush_dcache_page(bvec.bv_page); + + blk_mq_end_request(req, errno_to_blk_status(ret)); + + return BLK_STS_OK; +} + +static int ubiblock_open(struct gendisk *disk, blk_mode_t mode) +{ + struct ubiblock *dev = disk->private_data; + int ret; + + mutex_lock(&dev->dev_mutex); + if (dev->refcnt > 0) { + /* + * The volume is already open, just increase the reference + * counter. + */ + goto out_done; + } + + /* + * We want users to be aware they should only mount us as read-only. + * It's just a paranoid check, as write requests will get rejected + * in any case. + */ + if (mode & BLK_OPEN_WRITE) { + ret = -EROFS; + goto out_unlock; + } + dev->desc = ubi_open_volume(dev->ubi_num, dev->vol_id, UBI_READONLY); + if (IS_ERR(dev->desc)) { + dev_err(disk_to_dev(dev->gd), "failed to open ubi volume %d_%d", + dev->ubi_num, dev->vol_id); + ret = PTR_ERR(dev->desc); + dev->desc = NULL; + goto out_unlock; + } + +out_done: + dev->refcnt++; + mutex_unlock(&dev->dev_mutex); + return 0; + +out_unlock: + mutex_unlock(&dev->dev_mutex); + return ret; +} + +static void ubiblock_release(struct gendisk *gd) +{ + struct ubiblock *dev = gd->private_data; + + mutex_lock(&dev->dev_mutex); + dev->refcnt--; + if (dev->refcnt == 0) { + ubi_close_volume(dev->desc); + dev->desc = NULL; + } + mutex_unlock(&dev->dev_mutex); +} + +static int ubiblock_getgeo(struct gendisk *disk, struct hd_geometry *geo) +{ + /* Some tools might require this information */ + geo->heads = 1; + geo->cylinders = 1; + geo->sectors = get_capacity(disk); + geo->start = 0; + return 0; +} + +static const struct block_device_operations ubiblock_ops = { + .owner = THIS_MODULE, + .open = ubiblock_open, + .release = ubiblock_release, + .getgeo = ubiblock_getgeo, +}; + +static blk_status_t ubiblock_queue_rq(struct blk_mq_hw_ctx *hctx, + const struct blk_mq_queue_data *bd) +{ + switch (req_op(bd->rq)) { + case REQ_OP_READ: + return ubiblock_read(bd->rq); + default: + return BLK_STS_IOERR; + } +} + +static int ubiblock_init_request(struct blk_mq_tag_set *set, + struct request *req, unsigned int hctx_idx, + int numa_node) +{ + struct ubiblock_pdu *pdu = blk_mq_rq_to_pdu(req); + + sg_init_table(pdu->usgl.sg, UBI_MAX_SG_COUNT); + return 0; +} + +static const struct blk_mq_ops ubiblock_mq_ops = { + .queue_rq = ubiblock_queue_rq, + .init_request = ubiblock_init_request, +}; + +static int calc_disk_capacity(struct ubi_volume_info *vi, u64 *disk_capacity) +{ + u64 size = vi->used_bytes >> 9; + + if (vi->used_bytes % 512) { + if (vi->vol_type == UBI_DYNAMIC_VOLUME) + pr_warn("UBI: block: volume size is not a multiple of 512, last %llu bytes are ignored!\n", + vi->used_bytes - (size << 9)); + else + pr_info("UBI: block: volume size is not a multiple of 512, last %llu bytes are ignored!\n", + vi->used_bytes - (size << 9)); + } + + if ((sector_t)size != size) + return -EFBIG; + + *disk_capacity = size; + + return 0; +} + +int ubiblock_create(struct ubi_volume_info *vi) +{ + struct queue_limits lim = { + .max_segments = UBI_MAX_SG_COUNT, + }; + struct ubiblock *dev; + struct gendisk *gd; + u64 disk_capacity; + int ret; + + ret = calc_disk_capacity(vi, &disk_capacity); + if (ret) { + return ret; + } + + /* Check that the volume isn't already handled */ + mutex_lock(&devices_mutex); + if (find_dev_nolock(vi->ubi_num, vi->vol_id)) { + ret = -EEXIST; + goto out_unlock; + } + + dev = kzalloc_obj(struct ubiblock); + if (!dev) { + ret = -ENOMEM; + goto out_unlock; + } + + mutex_init(&dev->dev_mutex); + + dev->ubi_num = vi->ubi_num; + dev->vol_id = vi->vol_id; + dev->leb_size = vi->usable_leb_size; + + dev->tag_set.ops = &ubiblock_mq_ops; + dev->tag_set.queue_depth = 64; + dev->tag_set.numa_node = NUMA_NO_NODE; + dev->tag_set.flags = BLK_MQ_F_BLOCKING; + dev->tag_set.cmd_size = sizeof(struct ubiblock_pdu); + dev->tag_set.driver_data = dev; + dev->tag_set.nr_hw_queues = 1; + + ret = blk_mq_alloc_tag_set(&dev->tag_set); + if (ret) { + pr_err("ubiblock%d_%d: blk_mq_alloc_tag_set failed\n", + dev->ubi_num, dev->vol_id); + goto out_free_dev; + } + + + /* Initialize the gendisk of this ubiblock device */ + gd = blk_mq_alloc_disk(&dev->tag_set, &lim, dev); + if (IS_ERR(gd)) { + ret = PTR_ERR(gd); + goto out_free_tags; + } + + gd->fops = &ubiblock_ops; + gd->major = ubiblock_major; + gd->minors = 1; + gd->first_minor = idr_alloc(&ubiblock_minor_idr, dev, 0, 0, GFP_KERNEL); + if (gd->first_minor < 0) { + pr_err("ubiblock%d_%d: block: dynamic minor allocation failed\n", + dev->ubi_num, dev->vol_id); + ret = -ENODEV; + goto out_cleanup_disk; + } + gd->flags |= GENHD_FL_NO_PART; + gd->private_data = dev; + sprintf(gd->disk_name, "ubiblock%d_%d", dev->ubi_num, dev->vol_id); + set_capacity(gd, disk_capacity); + dev->gd = gd; + + dev->rq = gd->queue; + + list_add_tail(&dev->list, &ubiblock_devices); + + /* Must be the last step: anyone can call file ops from now on */ + ret = device_add_disk(vi->dev, dev->gd, NULL); + if (ret) + goto out_remove_minor; + + dev_info(disk_to_dev(dev->gd), "created from ubi%d:%d(%s)", + dev->ubi_num, dev->vol_id, vi->name); + mutex_unlock(&devices_mutex); + return 0; + +out_remove_minor: + list_del(&dev->list); + idr_remove(&ubiblock_minor_idr, gd->first_minor); +out_cleanup_disk: + put_disk(gd); +out_free_tags: + blk_mq_free_tag_set(&dev->tag_set); +out_free_dev: + kfree(dev); +out_unlock: + mutex_unlock(&devices_mutex); + + return ret; +} + +static void ubiblock_cleanup(struct ubiblock *dev) +{ + int id = dev->gd->first_minor; + + /* Stop new requests to arrive */ + del_gendisk(dev->gd); + /* Finally destroy the blk queue */ + dev_info(disk_to_dev(dev->gd), "released"); + put_disk(dev->gd); + blk_mq_free_tag_set(&dev->tag_set); + idr_remove(&ubiblock_minor_idr, id); +} + +int ubiblock_remove(struct ubi_volume_info *vi) +{ + struct ubiblock *dev; + int ret; + + mutex_lock(&devices_mutex); + dev = find_dev_nolock(vi->ubi_num, vi->vol_id); + if (!dev) { + ret = -ENODEV; + goto out_unlock; + } + + /* Found a device, let's lock it so we can check if it's busy */ + mutex_lock(&dev->dev_mutex); + if (dev->refcnt > 0) { + ret = -EBUSY; + goto out_unlock_dev; + } + + /* Remove from device list */ + list_del(&dev->list); + ubiblock_cleanup(dev); + mutex_unlock(&dev->dev_mutex); + mutex_unlock(&devices_mutex); + + kfree(dev); + return 0; + +out_unlock_dev: + mutex_unlock(&dev->dev_mutex); +out_unlock: + mutex_unlock(&devices_mutex); + return ret; +} + +static int ubiblock_resize(struct ubi_volume_info *vi) +{ + struct ubiblock *dev; + u64 disk_capacity; + int ret; + + /* + * Need to lock the device list until we stop using the device, + * otherwise the device struct might get released in + * 'ubiblock_remove()'. + */ + mutex_lock(&devices_mutex); + dev = find_dev_nolock(vi->ubi_num, vi->vol_id); + if (!dev) { + mutex_unlock(&devices_mutex); + return -ENODEV; + } + + ret = calc_disk_capacity(vi, &disk_capacity); + if (ret) { + mutex_unlock(&devices_mutex); + if (ret == -EFBIG) { + dev_warn(disk_to_dev(dev->gd), + "the volume is too big (%d LEBs), cannot resize", + vi->size); + } + return ret; + } + + mutex_lock(&dev->dev_mutex); + + if (get_capacity(dev->gd) != disk_capacity) { + set_capacity(dev->gd, disk_capacity); + dev_info(disk_to_dev(dev->gd), "resized to %lld bytes", + vi->used_bytes); + } + mutex_unlock(&dev->dev_mutex); + mutex_unlock(&devices_mutex); + return 0; +} + +static bool +match_volume_desc(struct ubi_volume_info *vi, const char *name, int ubi_num, int vol_id) +{ + int err, len, cur_ubi_num, cur_vol_id; + + if (ubi_num == -1) { + /* No ubi num, name must be a vol device path */ + err = ubi_get_num_by_path(name, &cur_ubi_num, &cur_vol_id); + if (err || vi->ubi_num != cur_ubi_num || vi->vol_id != cur_vol_id) + return false; + + return true; + } + + if (vol_id == -1) { + /* Got ubi_num, but no vol_id, name must be volume name */ + if (vi->ubi_num != ubi_num) + return false; + + len = strnlen(name, UBI_VOL_NAME_MAX + 1); + if (len < 1 || vi->name_len != len) + return false; + + if (strcmp(name, vi->name)) + return false; + + return true; + } + + if (vi->ubi_num != ubi_num) + return false; + + if (vi->vol_id != vol_id) + return false; + + return true; +} + +static void +ubiblock_create_from_param(struct ubi_volume_info *vi) +{ + int i, ret = 0; + struct ubiblock_param *p; + + /* + * Iterate over ubiblock cmdline parameters. If a parameter matches the + * newly added volume create the ubiblock device for it. + */ + for (i = 0; i < ubiblock_devs; i++) { + p = &ubiblock_param[i]; + + if (!match_volume_desc(vi, p->name, p->ubi_num, p->vol_id)) + continue; + + ret = ubiblock_create(vi); + if (ret) { + pr_err( + "UBI: block: can't add '%s' volume on ubi%d_%d, err=%d\n", + vi->name, p->ubi_num, p->vol_id, ret); + } + break; + } +} + +static int ubiblock_notify(struct notifier_block *nb, + unsigned long notification_type, void *ns_ptr) +{ + struct ubi_notification *nt = ns_ptr; + + switch (notification_type) { + case UBI_VOLUME_ADDED: + ubiblock_create_from_param(&nt->vi); + break; + case UBI_VOLUME_REMOVED: + ubiblock_remove(&nt->vi); + break; + case UBI_VOLUME_RESIZED: + ubiblock_resize(&nt->vi); + break; + case UBI_VOLUME_UPDATED: + /* + * If the volume is static, a content update might mean the + * size (i.e. used_bytes) was also changed. + */ + if (nt->vi.vol_type == UBI_STATIC_VOLUME) + ubiblock_resize(&nt->vi); + break; + default: + break; + } + return NOTIFY_OK; +} + +static struct notifier_block ubiblock_notifier = { + .notifier_call = ubiblock_notify, +}; + +static void ubiblock_remove_all(void) +{ + struct ubiblock *next; + struct ubiblock *dev; + + mutex_lock(&devices_mutex); + list_for_each_entry_safe(dev, next, &ubiblock_devices, list) { + /* The module is being forcefully removed */ + WARN_ON(dev->desc); + /* Remove from device list */ + list_del(&dev->list); + ubiblock_cleanup(dev); + kfree(dev); + } + mutex_unlock(&devices_mutex); +} + +int __init ubiblock_init(void) +{ + int ret; + + ubiblock_major = register_blkdev(0, "ubiblock"); + if (ubiblock_major < 0) + return ubiblock_major; + + ret = ubi_register_volume_notifier(&ubiblock_notifier, 0); + if (ret) + goto err_unreg; + return 0; + +err_unreg: + unregister_blkdev(ubiblock_major, "ubiblock"); + ubiblock_remove_all(); + return ret; +} + +void ubiblock_exit(void) +{ + ubi_unregister_volume_notifier(&ubiblock_notifier); + ubiblock_remove_all(); + unregister_blkdev(ubiblock_major, "ubiblock"); +} diff --git a/drivers/mtd/ubi/build.c b/drivers/mtd/ubi/build.c new file mode 100644 index 000000000..910e6b0ff --- /dev/null +++ b/drivers/mtd/ubi/build.c @@ -0,0 +1,1632 @@ +// SPDX-License-Identifier: GPL-2.0-or-later +/* + * Copyright (c) International Business Machines Corp., 2006 + * Copyright (c) Nokia Corporation, 2007 + * + * Author: Artem Bityutskiy (Битюцкий Артём), + * Frank Haverkamp + */ + +/* + * This file includes UBI initialization and building of UBI devices. + * + * When UBI is initialized, it attaches all the MTD devices specified as the + * module load parameters or the kernel boot parameters. If MTD devices were + * specified, UBI does not attach any MTD device, but it is possible to do + * later using the "UBI control device". + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include "ubi.h" + +/* Maximum length of the 'mtd=' parameter */ +#define MTD_PARAM_LEN_MAX 64 + +/* Maximum number of comma-separated items in the 'mtd=' parameter */ +#define MTD_PARAM_MAX_COUNT 7 + +/* Maximum value for the number of bad PEBs per 1024 PEBs */ +#define MAX_MTD_UBI_BEB_LIMIT 768 + +#ifdef CONFIG_MTD_UBI_MODULE +#define ubi_is_module() 1 +#else +#define ubi_is_module() 0 +#endif + +/** + * struct mtd_dev_param - MTD device parameter description data structure. + * @name: MTD character device node path, MTD device name, or MTD device number + * string + * @ubi_num: UBI number + * @vid_hdr_offs: VID header offset + * @max_beb_per1024: maximum expected number of bad PEBs per 1024 PEBs + * @enable_fm: enable fastmap when value is non-zero + * @need_resv_pool: reserve pool->max_size pebs when value is none-zero + * @wl_threshold: wear-leveling threshold, 0 means use CONFIG_MTD_UBI_WL_THRESHOLD + */ +struct mtd_dev_param { + char name[MTD_PARAM_LEN_MAX]; + int ubi_num; + int vid_hdr_offs; + int max_beb_per1024; + int enable_fm; + int need_resv_pool; + int wl_threshold; +}; + +/* Numbers of elements set in the @mtd_dev_param array */ +static int mtd_devs; + +/* MTD devices specification parameters */ +static struct mtd_dev_param mtd_dev_param[UBI_MAX_DEVICES]; +#ifdef CONFIG_MTD_UBI_FASTMAP +/* UBI module parameter to enable fastmap automatically on non-fastmap images */ +static bool fm_autoconvert; +static bool fm_debug; +#endif + +/* Slab cache for wear-leveling entries */ +struct kmem_cache *ubi_wl_entry_slab; + +/* UBI control character device */ +static struct miscdevice ubi_ctrl_cdev = { + .minor = MISC_DYNAMIC_MINOR, + .name = "ubi_ctrl", + .fops = &ubi_ctrl_cdev_operations, +}; + +/* All UBI devices in system */ +static struct ubi_device *ubi_devices[UBI_MAX_DEVICES]; + +/* Serializes UBI devices creations and removals */ +DEFINE_MUTEX(ubi_devices_mutex); + +/* Protects @ubi_devices, @ubi->ref_count and @ubi->is_dead */ +static DEFINE_SPINLOCK(ubi_devices_lock); + +/* "Show" method for files in '//class/ubi/' */ +/* UBI version attribute ('//class/ubi/version') */ +static ssize_t version_show(const struct class *class, const struct class_attribute *attr, + char *buf) +{ + return sprintf(buf, "%d\n", UBI_VERSION); +} +static CLASS_ATTR_RO(version); + +static struct attribute *ubi_class_attrs[] = { + &class_attr_version.attr, + NULL, +}; +ATTRIBUTE_GROUPS(ubi_class); + +/* Root UBI "class" object (corresponds to '//class/ubi/') */ +const struct class ubi_class = { + .name = UBI_NAME_STR, + .class_groups = ubi_class_groups, +}; + +static ssize_t dev_attribute_show(struct device *dev, + struct device_attribute *attr, char *buf); + +/* UBI device attributes (correspond to files in '//class/ubi/ubiX') */ +static struct device_attribute dev_eraseblock_size = + __ATTR(eraseblock_size, S_IRUGO, dev_attribute_show, NULL); +static struct device_attribute dev_avail_eraseblocks = + __ATTR(avail_eraseblocks, S_IRUGO, dev_attribute_show, NULL); +static struct device_attribute dev_total_eraseblocks = + __ATTR(total_eraseblocks, S_IRUGO, dev_attribute_show, NULL); +static struct device_attribute dev_volumes_count = + __ATTR(volumes_count, S_IRUGO, dev_attribute_show, NULL); +static struct device_attribute dev_max_ec = + __ATTR(max_ec, S_IRUGO, dev_attribute_show, NULL); +static struct device_attribute dev_reserved_for_bad = + __ATTR(reserved_for_bad, S_IRUGO, dev_attribute_show, NULL); +static struct device_attribute dev_bad_peb_count = + __ATTR(bad_peb_count, S_IRUGO, dev_attribute_show, NULL); +static struct device_attribute dev_max_vol_count = + __ATTR(max_vol_count, S_IRUGO, dev_attribute_show, NULL); +static struct device_attribute dev_min_io_size = + __ATTR(min_io_size, S_IRUGO, dev_attribute_show, NULL); +static struct device_attribute dev_bgt_enabled = + __ATTR(bgt_enabled, S_IRUGO, dev_attribute_show, NULL); +static struct device_attribute dev_mtd_num = + __ATTR(mtd_num, S_IRUGO, dev_attribute_show, NULL); +static struct device_attribute dev_ro_mode = + __ATTR(ro_mode, S_IRUGO, dev_attribute_show, NULL); + +/** + * ubi_volume_notify - send a volume change notification. + * @ubi: UBI device description object + * @vol: volume description object of the changed volume + * @ntype: notification type to send (%UBI_VOLUME_ADDED, etc) + * + * This is a helper function which notifies all subscribers about a volume + * change event (creation, removal, re-sizing, re-naming, updating). Returns + * zero in case of success and a negative error code in case of failure. + */ +int ubi_volume_notify(struct ubi_device *ubi, struct ubi_volume *vol, int ntype) +{ + int ret; + struct ubi_notification nt; + + ubi_do_get_device_info(ubi, &nt.di); + ubi_do_get_volume_info(ubi, vol, &nt.vi); + + switch (ntype) { + case UBI_VOLUME_ADDED: + case UBI_VOLUME_REMOVED: + case UBI_VOLUME_RESIZED: + case UBI_VOLUME_RENAMED: + ret = ubi_update_fastmap(ubi); + if (ret) + ubi_msg(ubi, "Unable to write a new fastmap: %i", ret); + } + + return blocking_notifier_call_chain(&ubi_notifiers, ntype, &nt); +} + +/** + * ubi_notify_all - send a notification to all volumes. + * @ubi: UBI device description object + * @ntype: notification type to send (%UBI_VOLUME_ADDED, etc) + * @nb: the notifier to call + * + * This function walks all volumes of UBI device @ubi and sends the @ntype + * notification for each volume. If @nb is %NULL, then all registered notifiers + * are called, otherwise only the @nb notifier is called. Returns the number of + * sent notifications. + */ +int ubi_notify_all(struct ubi_device *ubi, int ntype, struct notifier_block *nb) +{ + struct ubi_notification nt; + int i, count = 0; + + ubi_do_get_device_info(ubi, &nt.di); + + mutex_lock(&ubi->device_mutex); + for (i = 0; i < ubi->vtbl_slots; i++) { + /* + * Since the @ubi->device is locked, and we are not going to + * change @ubi->volumes, we do not have to lock + * @ubi->volumes_lock. + */ + if (!ubi->volumes[i]) + continue; + + ubi_do_get_volume_info(ubi, ubi->volumes[i], &nt.vi); + if (nb) + nb->notifier_call(nb, ntype, &nt); + else + blocking_notifier_call_chain(&ubi_notifiers, ntype, + &nt); + count += 1; + } + mutex_unlock(&ubi->device_mutex); + + return count; +} + +/** + * ubi_enumerate_volumes - send "add" notification for all existing volumes. + * @nb: the notifier to call + * + * This function walks all UBI devices and volumes and sends the + * %UBI_VOLUME_ADDED notification for each volume. If @nb is %NULL, then all + * registered notifiers are called, otherwise only the @nb notifier is called. + * Returns the number of sent notifications. + */ +int ubi_enumerate_volumes(struct notifier_block *nb) +{ + int i, count = 0; + + /* + * Since the @ubi_devices_mutex is locked, and we are not going to + * change @ubi_devices, we do not have to lock @ubi_devices_lock. + */ + for (i = 0; i < UBI_MAX_DEVICES; i++) { + struct ubi_device *ubi = ubi_devices[i]; + + if (!ubi) + continue; + count += ubi_notify_all(ubi, UBI_VOLUME_ADDED, nb); + } + + return count; +} + +/** + * ubi_get_device - get UBI device. + * @ubi_num: UBI device number + * + * This function returns UBI device description object for UBI device number + * @ubi_num, or %NULL if the device does not exist. This function increases the + * device reference count to prevent removal of the device. In other words, the + * device cannot be removed if its reference count is not zero. + */ +struct ubi_device *ubi_get_device(int ubi_num) +{ + struct ubi_device *ubi; + + spin_lock(&ubi_devices_lock); + ubi = ubi_devices[ubi_num]; + if (ubi && ubi->is_dead) + ubi = NULL; + + if (ubi) { + ubi_assert(ubi->ref_count >= 0); + ubi->ref_count += 1; + get_device(&ubi->dev); + } + spin_unlock(&ubi_devices_lock); + + return ubi; +} + +/** + * ubi_put_device - drop an UBI device reference. + * @ubi: UBI device description object + */ +void ubi_put_device(struct ubi_device *ubi) +{ + spin_lock(&ubi_devices_lock); + ubi->ref_count -= 1; + put_device(&ubi->dev); + spin_unlock(&ubi_devices_lock); +} + +/** + * ubi_get_by_major - get UBI device by character device major number. + * @major: major number + * + * This function is similar to 'ubi_get_device()', but it searches the device + * by its major number. + */ +struct ubi_device *ubi_get_by_major(int major) +{ + int i; + struct ubi_device *ubi; + + spin_lock(&ubi_devices_lock); + for (i = 0; i < UBI_MAX_DEVICES; i++) { + ubi = ubi_devices[i]; + if (ubi && !ubi->is_dead && MAJOR(ubi->cdev.dev) == major) { + ubi_assert(ubi->ref_count >= 0); + ubi->ref_count += 1; + get_device(&ubi->dev); + spin_unlock(&ubi_devices_lock); + return ubi; + } + } + spin_unlock(&ubi_devices_lock); + + return NULL; +} + +/** + * ubi_major2num - get UBI device number by character device major number. + * @major: major number + * + * This function searches UBI device number object by its major number. If UBI + * device was not found, this function returns -ENODEV, otherwise the UBI device + * number is returned. + */ +int ubi_major2num(int major) +{ + int i, ubi_num = -ENODEV; + + spin_lock(&ubi_devices_lock); + for (i = 0; i < UBI_MAX_DEVICES; i++) { + struct ubi_device *ubi = ubi_devices[i]; + + if (ubi && !ubi->is_dead && MAJOR(ubi->cdev.dev) == major) { + ubi_num = ubi->ubi_num; + break; + } + } + spin_unlock(&ubi_devices_lock); + + return ubi_num; +} + +/* "Show" method for files in '//class/ubi/ubiX/' */ +static ssize_t dev_attribute_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + ssize_t ret; + struct ubi_device *ubi; + + /* + * The below code looks weird, but it actually makes sense. We get the + * UBI device reference from the contained 'struct ubi_device'. But it + * is unclear if the device was removed or not yet. Indeed, if the + * device was removed before we increased its reference count, + * 'ubi_get_device()' will return -ENODEV and we fail. + * + * Remember, 'struct ubi_device' is freed in the release function, so + * we still can use 'ubi->ubi_num'. + */ + ubi = container_of(dev, struct ubi_device, dev); + + if (attr == &dev_eraseblock_size) + ret = sprintf(buf, "%d\n", ubi->leb_size); + else if (attr == &dev_avail_eraseblocks) + ret = sprintf(buf, "%d\n", ubi->avail_pebs); + else if (attr == &dev_total_eraseblocks) + ret = sprintf(buf, "%d\n", ubi->good_peb_count); + else if (attr == &dev_volumes_count) + ret = sprintf(buf, "%d\n", ubi->vol_count - UBI_INT_VOL_COUNT); + else if (attr == &dev_max_ec) + ret = sprintf(buf, "%d\n", ubi->max_ec); + else if (attr == &dev_reserved_for_bad) + ret = sprintf(buf, "%d\n", ubi->beb_rsvd_pebs); + else if (attr == &dev_bad_peb_count) + ret = sprintf(buf, "%d\n", ubi->bad_peb_count); + else if (attr == &dev_max_vol_count) + ret = sprintf(buf, "%d\n", ubi->vtbl_slots); + else if (attr == &dev_min_io_size) + ret = sprintf(buf, "%d\n", ubi->min_io_size); + else if (attr == &dev_bgt_enabled) + ret = sprintf(buf, "%d\n", ubi->thread_enabled); + else if (attr == &dev_mtd_num) + ret = sprintf(buf, "%d\n", ubi->mtd->index); + else if (attr == &dev_ro_mode) + ret = sprintf(buf, "%d\n", ubi->ro_mode); + else + ret = -EINVAL; + + return ret; +} + +static struct attribute *ubi_dev_attrs[] = { + &dev_eraseblock_size.attr, + &dev_avail_eraseblocks.attr, + &dev_total_eraseblocks.attr, + &dev_volumes_count.attr, + &dev_max_ec.attr, + &dev_reserved_for_bad.attr, + &dev_bad_peb_count.attr, + &dev_max_vol_count.attr, + &dev_min_io_size.attr, + &dev_bgt_enabled.attr, + &dev_mtd_num.attr, + &dev_ro_mode.attr, + NULL +}; +ATTRIBUTE_GROUPS(ubi_dev); + +static void dev_release(struct device *dev) +{ + struct ubi_device *ubi = container_of(dev, struct ubi_device, dev); + + kfree(ubi); +} + +/** + * kill_volumes - destroy all user volumes. + * @ubi: UBI device description object + */ +static void kill_volumes(struct ubi_device *ubi) +{ + int i; + + for (i = 0; i < ubi->vtbl_slots; i++) + if (ubi->volumes[i]) + ubi_free_volume(ubi, ubi->volumes[i]); +} + +/** + * uif_init - initialize user interfaces for an UBI device. + * @ubi: UBI device description object + * + * This function initializes various user interfaces for an UBI device. If the + * initialization fails at an early stage, this function frees all the + * resources it allocated, returns an error. + * + * This function returns zero in case of success and a negative error code in + * case of failure. + */ +static int uif_init(struct ubi_device *ubi) +{ + int i, err; + dev_t dev; + + sprintf(ubi->ubi_name, UBI_NAME_STR "%d", ubi->ubi_num); + + /* + * Major numbers for the UBI character devices are allocated + * dynamically. Major numbers of volume character devices are + * equivalent to ones of the corresponding UBI character device. Minor + * numbers of UBI character devices are 0, while minor numbers of + * volume character devices start from 1. Thus, we allocate one major + * number and ubi->vtbl_slots + 1 minor numbers. + */ + err = alloc_chrdev_region(&dev, 0, ubi->vtbl_slots + 1, ubi->ubi_name); + if (err) { + ubi_err(ubi, "cannot register UBI character devices"); + return err; + } + + ubi->dev.devt = dev; + + ubi_assert(MINOR(dev) == 0); + cdev_init(&ubi->cdev, &ubi_cdev_operations); + dbg_gen("%s major is %u", ubi->ubi_name, MAJOR(dev)); + ubi->cdev.owner = THIS_MODULE; + + dev_set_name(&ubi->dev, UBI_NAME_STR "%d", ubi->ubi_num); + err = cdev_device_add(&ubi->cdev, &ubi->dev); + if (err) + goto out_unreg; + + for (i = 0; i < ubi->vtbl_slots; i++) + if (ubi->volumes[i]) { + err = ubi_add_volume(ubi, ubi->volumes[i]); + if (err) { + ubi_err(ubi, "cannot add volume %d", i); + ubi->volumes[i] = NULL; + goto out_volumes; + } + } + + return 0; + +out_volumes: + kill_volumes(ubi); + cdev_device_del(&ubi->cdev, &ubi->dev); +out_unreg: + unregister_chrdev_region(ubi->cdev.dev, ubi->vtbl_slots + 1); + ubi_err(ubi, "cannot initialize UBI %s, error %d", + ubi->ubi_name, err); + return err; +} + +/** + * uif_close - close user interfaces for an UBI device. + * @ubi: UBI device description object + * + * Note, since this function un-registers UBI volume device objects (@vol->dev), + * the memory allocated voe the volumes is freed as well (in the release + * function). + */ +static void uif_close(struct ubi_device *ubi) +{ + kill_volumes(ubi); + cdev_device_del(&ubi->cdev, &ubi->dev); + unregister_chrdev_region(ubi->cdev.dev, ubi->vtbl_slots + 1); +} + +/** + * ubi_free_volumes_from - free volumes from specific index. + * @ubi: UBI device description object + * @from: the start index used for volume free. + */ +static void ubi_free_volumes_from(struct ubi_device *ubi, int from) +{ + int i; + + for (i = from; i < ubi->vtbl_slots + UBI_INT_VOL_COUNT; i++) { + if (!ubi->volumes[i] || ubi->volumes[i]->is_dead) + continue; + ubi_eba_replace_table(ubi->volumes[i], NULL); + ubi_fastmap_destroy_checkmap(ubi->volumes[i]); + kfree(ubi->volumes[i]); + ubi->volumes[i] = NULL; + } +} + +/** + * ubi_free_all_volumes - free all volumes. + * @ubi: UBI device description object + */ +void ubi_free_all_volumes(struct ubi_device *ubi) +{ + ubi_free_volumes_from(ubi, 0); +} + +/** + * ubi_free_internal_volumes - free internal volumes. + * @ubi: UBI device description object + */ +void ubi_free_internal_volumes(struct ubi_device *ubi) +{ + ubi_free_volumes_from(ubi, ubi->vtbl_slots); +} + +static int get_bad_peb_limit(const struct ubi_device *ubi, int max_beb_per1024) +{ + int limit, device_pebs; + uint64_t device_size; + + if (!max_beb_per1024) { + /* + * Since max_beb_per1024 has not been set by the user in either + * the cmdline or Kconfig, use mtd_max_bad_blocks to set the + * limit if it is supported by the device. + */ + limit = mtd_max_bad_blocks(ubi->mtd, 0, ubi->mtd->size); + if (limit < 0) + return 0; + return limit; + } + + /* + * Here we are using size of the entire flash chip and + * not just the MTD partition size because the maximum + * number of bad eraseblocks is a percentage of the + * whole device and bad eraseblocks are not fairly + * distributed over the flash chip. So the worst case + * is that all the bad eraseblocks of the chip are in + * the MTD partition we are attaching (ubi->mtd). + */ + device_size = mtd_get_device_size(ubi->mtd); + device_pebs = mtd_div_by_eb(device_size, ubi->mtd); + limit = mult_frac(device_pebs, max_beb_per1024, 1024); + + /* Round it up */ + if (mult_frac(limit, 1024, max_beb_per1024) < device_pebs) + limit += 1; + + return limit; +} + +/** + * io_init - initialize I/O sub-system for a given UBI device. + * @ubi: UBI device description object + * @max_beb_per1024: maximum expected number of bad PEB per 1024 PEBs + * + * If @ubi->vid_hdr_offset or @ubi->leb_start is zero, default offsets are + * assumed: + * o EC header is always at offset zero - this cannot be changed; + * o VID header starts just after the EC header at the closest address + * aligned to @io->hdrs_min_io_size; + * o data starts just after the VID header at the closest address aligned to + * @io->min_io_size + * + * This function returns zero in case of success and a negative error code in + * case of failure. + */ +static int io_init(struct ubi_device *ubi, int max_beb_per1024) +{ + dbg_gen("sizeof(struct ubi_ainf_peb) %zu", sizeof(struct ubi_ainf_peb)); + dbg_gen("sizeof(struct ubi_wl_entry) %zu", sizeof(struct ubi_wl_entry)); + + if (ubi->mtd->numeraseregions != 0) { + /* + * Some flashes have several erase regions. Different regions + * may have different eraseblock size and other + * characteristics. It looks like mostly multi-region flashes + * have one "main" region and one or more small regions to + * store boot loader code or boot parameters or whatever. I + * guess we should just pick the largest region. But this is + * not implemented. + */ + ubi_err(ubi, "multiple regions, not implemented"); + return -EINVAL; + } + + if (ubi->vid_hdr_offset < 0) + return -EINVAL; + + /* + * Note, in this implementation we support MTD devices with 0x7FFFFFFF + * physical eraseblocks maximum. + */ + + ubi->peb_size = ubi->mtd->erasesize; + ubi->peb_count = mtd_div_by_eb(ubi->mtd->size, ubi->mtd); + ubi->flash_size = ubi->mtd->size; + + if (mtd_can_have_bb(ubi->mtd)) { + ubi->bad_allowed = 1; + ubi->bad_peb_limit = get_bad_peb_limit(ubi, max_beb_per1024); + } + + if (ubi->mtd->type == MTD_NORFLASH) + ubi->nor_flash = 1; + + ubi->min_io_size = ubi->mtd->writesize; + ubi->hdrs_min_io_size = ubi->mtd->writesize >> ubi->mtd->subpage_sft; + + /* + * Make sure minimal I/O unit is power of 2. Note, there is no + * fundamental reason for this assumption. It is just an optimization + * which allows us to avoid costly division operations. + */ + if (!is_power_of_2(ubi->min_io_size)) { + ubi_err(ubi, "min. I/O unit (%d) is not power of 2", + ubi->min_io_size); + return -EINVAL; + } + + ubi_assert(ubi->hdrs_min_io_size > 0); + ubi_assert(ubi->hdrs_min_io_size <= ubi->min_io_size); + ubi_assert(ubi->min_io_size % ubi->hdrs_min_io_size == 0); + + ubi->max_write_size = ubi->mtd->writebufsize; + /* + * Maximum write size has to be greater or equivalent to min. I/O + * size, and be multiple of min. I/O size. + */ + if (ubi->max_write_size < ubi->min_io_size || + ubi->max_write_size % ubi->min_io_size || + !is_power_of_2(ubi->max_write_size)) { + ubi_err(ubi, "bad write buffer size %d for %d min. I/O unit", + ubi->max_write_size, ubi->min_io_size); + return -EINVAL; + } + + /* Calculate default aligned sizes of EC and VID headers */ + ubi->ec_hdr_alsize = ALIGN(UBI_EC_HDR_SIZE, ubi->hdrs_min_io_size); + ubi->vid_hdr_alsize = ALIGN(UBI_VID_HDR_SIZE, ubi->hdrs_min_io_size); + + dbg_gen("min_io_size %d", ubi->min_io_size); + dbg_gen("max_write_size %d", ubi->max_write_size); + dbg_gen("hdrs_min_io_size %d", ubi->hdrs_min_io_size); + dbg_gen("ec_hdr_alsize %d", ubi->ec_hdr_alsize); + dbg_gen("vid_hdr_alsize %d", ubi->vid_hdr_alsize); + + if (ubi->vid_hdr_offset == 0) + /* Default offset */ + ubi->vid_hdr_offset = ubi->vid_hdr_aloffset = + ubi->ec_hdr_alsize; + else { + ubi->vid_hdr_aloffset = ubi->vid_hdr_offset & + ~(ubi->hdrs_min_io_size - 1); + ubi->vid_hdr_shift = ubi->vid_hdr_offset - + ubi->vid_hdr_aloffset; + } + + /* + * Memory allocation for VID header is ubi->vid_hdr_alsize + * which is described in comments in io.c. + * Make sure VID header shift + UBI_VID_HDR_SIZE not exceeds + * ubi->vid_hdr_alsize, so that all vid header operations + * won't access memory out of bounds. + */ + if ((ubi->vid_hdr_shift + UBI_VID_HDR_SIZE) > ubi->vid_hdr_alsize) { + ubi_err(ubi, "Invalid VID header offset %d, VID header shift(%d)" + " + VID header size(%zu) > VID header aligned size(%d).", + ubi->vid_hdr_offset, ubi->vid_hdr_shift, + UBI_VID_HDR_SIZE, ubi->vid_hdr_alsize); + return -EINVAL; + } + + /* Similar for the data offset */ + ubi->leb_start = ubi->vid_hdr_offset + UBI_VID_HDR_SIZE; + ubi->leb_start = ALIGN(ubi->leb_start, ubi->min_io_size); + + dbg_gen("vid_hdr_offset %d", ubi->vid_hdr_offset); + dbg_gen("vid_hdr_aloffset %d", ubi->vid_hdr_aloffset); + dbg_gen("vid_hdr_shift %d", ubi->vid_hdr_shift); + dbg_gen("leb_start %d", ubi->leb_start); + + /* The shift must be aligned to 32-bit boundary */ + if (ubi->vid_hdr_shift % 4) { + ubi_err(ubi, "unaligned VID header shift %d", + ubi->vid_hdr_shift); + return -EINVAL; + } + + /* Check sanity */ + if (ubi->vid_hdr_offset < UBI_EC_HDR_SIZE || + ubi->leb_start < ubi->vid_hdr_offset + UBI_VID_HDR_SIZE || + ubi->leb_start > ubi->peb_size - UBI_VID_HDR_SIZE || + ubi->leb_start & (ubi->min_io_size - 1)) { + ubi_err(ubi, "bad VID header (%d) or data offsets (%d)", + ubi->vid_hdr_offset, ubi->leb_start); + return -EINVAL; + } + + /* + * Set maximum amount of physical erroneous eraseblocks to be 10%. + * Erroneous PEB are those which have read errors. + */ + ubi->max_erroneous = ubi->peb_count / 10; + if (ubi->max_erroneous < 16) + ubi->max_erroneous = 16; + dbg_gen("max_erroneous %d", ubi->max_erroneous); + + /* + * It may happen that EC and VID headers are situated in one minimal + * I/O unit. In this case we can only accept this UBI image in + * read-only mode. + */ + if (ubi->vid_hdr_offset + UBI_VID_HDR_SIZE <= ubi->hdrs_min_io_size) { + ubi_warn(ubi, "EC and VID headers are in the same minimal I/O unit, switch to read-only mode"); + ubi->ro_mode = 1; + } + + ubi->leb_size = ubi->peb_size - ubi->leb_start; + + if (!(ubi->mtd->flags & MTD_WRITEABLE)) { + ubi_msg(ubi, "MTD device %d is write-protected, attach in read-only mode", + ubi->mtd->index); + ubi->ro_mode = 1; + } + + /* + * Note, ideally, we have to initialize @ubi->bad_peb_count here. But + * unfortunately, MTD does not provide this information. We should loop + * over all physical eraseblocks and invoke mtd->block_is_bad() for + * each physical eraseblock. So, we leave @ubi->bad_peb_count + * uninitialized so far. + */ + + return 0; +} + +/** + * autoresize - re-size the volume which has the "auto-resize" flag set. + * @ubi: UBI device description object + * @vol_id: ID of the volume to re-size + * + * This function re-sizes the volume marked by the %UBI_VTBL_AUTORESIZE_FLG in + * the volume table to the largest possible size. See comments in ubi-header.h + * for more description of the flag. Returns zero in case of success and a + * negative error code in case of failure. + */ +static int autoresize(struct ubi_device *ubi, int vol_id) +{ + struct ubi_volume_desc desc; + struct ubi_volume *vol = ubi->volumes[vol_id]; + int err, old_reserved_pebs = vol->reserved_pebs; + + if (ubi->ro_mode) { + ubi_warn(ubi, "skip auto-resize because of R/O mode"); + return 0; + } + + /* + * Clear the auto-resize flag in the volume in-memory copy of the + * volume table, and 'ubi_resize_volume()' will propagate this change + * to the flash. + */ + ubi->vtbl[vol_id].flags &= ~UBI_VTBL_AUTORESIZE_FLG; + + if (ubi->avail_pebs == 0) { + struct ubi_vtbl_record vtbl_rec; + + /* + * No available PEBs to re-size the volume, clear the flag on + * flash and exit. + */ + vtbl_rec = ubi->vtbl[vol_id]; + err = ubi_change_vtbl_record(ubi, vol_id, &vtbl_rec); + if (err) + ubi_err(ubi, "cannot clean auto-resize flag for volume %d", + vol_id); + } else { + desc.vol = vol; + err = ubi_resize_volume(&desc, + old_reserved_pebs + ubi->avail_pebs); + if (err) + ubi_err(ubi, "cannot auto-resize volume %d", + vol_id); + } + + if (err) + return err; + + ubi_msg(ubi, "volume %d (\"%s\") re-sized from %d to %d LEBs", + vol_id, vol->name, old_reserved_pebs, vol->reserved_pebs); + return 0; +} + +/** + * ubi_attach_mtd_dev - attach an MTD device. + * @mtd: MTD device description object + * @ubi_num: number to assign to the new UBI device + * @vid_hdr_offset: VID header offset + * @max_beb_per1024: maximum expected number of bad PEB per 1024 PEBs + * @disable_fm: whether disable fastmap + * @need_resv_pool: whether reserve pebs to fill fm_pool + * @wl_threshold: wear-leveling threshold for this UBI device; 0 means use + * %CONFIG_MTD_UBI_WL_THRESHOLD; accepted range is 2-65536 + * + * This function attaches MTD device @mtd_dev to UBI and assign @ubi_num number + * to the newly created UBI device, unless @ubi_num is %UBI_DEV_NUM_AUTO, in + * which case this function finds a vacant device number and assigns it + * automatically. Returns the new UBI device number in case of success and a + * negative error code in case of failure. + * + * If @disable_fm is true, ubi doesn't create new fastmap even the module param + * 'fm_autoconvert' is set, and existed old fastmap will be destroyed after + * doing full scanning. + * + * Note, the invocations of this function has to be serialized by the + * @ubi_devices_mutex. + */ +int ubi_attach_mtd_dev(struct mtd_info *mtd, int ubi_num, + int vid_hdr_offset, int max_beb_per1024, bool disable_fm, + bool need_resv_pool, int wl_threshold) +{ + struct ubi_device *ubi; + int i, err; + + if (max_beb_per1024 < 0 || max_beb_per1024 > MAX_MTD_UBI_BEB_LIMIT) + return -EINVAL; + + if (!max_beb_per1024) + max_beb_per1024 = CONFIG_MTD_UBI_BEB_LIMIT; + + if (!wl_threshold) + wl_threshold = CONFIG_MTD_UBI_WL_THRESHOLD; + + if (wl_threshold < 2 || wl_threshold > 65536) { + pr_err("ubi: bad wear-leveling threshold %d\n", + wl_threshold); + return -EINVAL; + } + + /* + * Check if we already have the same MTD device attached. + * + * Note, this function assumes that UBI devices creations and deletions + * are serialized, so it does not take the &ubi_devices_lock. + */ + for (i = 0; i < UBI_MAX_DEVICES; i++) { + ubi = ubi_devices[i]; + if (ubi && mtd->index == ubi->mtd->index) { + pr_err("ubi: mtd%d is already attached to ubi%d\n", + mtd->index, i); + return -EEXIST; + } + } + + /* + * Make sure this MTD device is not emulated on top of an UBI volume + * already. Well, generally this recursion works fine, but there are + * different problems like the UBI module takes a reference to itself + * by attaching (and thus, opening) the emulated MTD device. This + * results in inability to unload the module. And in general it makes + * no sense to attach emulated MTD devices, so we prohibit this. + */ + if (mtd->type == MTD_UBIVOLUME) { + pr_err("ubi: refuse attaching mtd%d - it is already emulated on top of UBI\n", + mtd->index); + return -EINVAL; + } + + /* + * Both UBI and UBIFS have been designed for SLC NAND and NOR flashes. + * MLC NAND is different and needs special care, otherwise UBI or UBIFS + * will die soon and you will lose all your data. + * Relax this rule if the partition we're attaching to operates in SLC + * mode. + */ + if (mtd->type == MTD_MLCNANDFLASH && + !(mtd->flags & MTD_SLC_ON_MLC_EMULATION)) { + pr_err("ubi: refuse attaching mtd%d - MLC NAND is not supported\n", + mtd->index); + return -EINVAL; + } + + /* UBI cannot work on flashes with zero erasesize. */ + if (!mtd->erasesize) { + pr_err("ubi: refuse attaching mtd%d - zero erasesize flash is not supported\n", + mtd->index); + return -EINVAL; + } + + if (ubi_num == UBI_DEV_NUM_AUTO) { + /* Search for an empty slot in the @ubi_devices array */ + for (ubi_num = 0; ubi_num < UBI_MAX_DEVICES; ubi_num++) + if (!ubi_devices[ubi_num]) + break; + if (ubi_num == UBI_MAX_DEVICES) { + pr_err("ubi: only %d UBI devices may be created\n", + UBI_MAX_DEVICES); + return -ENFILE; + } + } else { + if (ubi_num >= UBI_MAX_DEVICES) + return -EINVAL; + + /* Make sure ubi_num is not busy */ + if (ubi_devices[ubi_num]) { + pr_err("ubi: ubi%i already exists\n", ubi_num); + return -EEXIST; + } + } + + ubi = kzalloc_obj(struct ubi_device); + if (!ubi) + return -ENOMEM; + + device_initialize(&ubi->dev); + ubi->dev.release = dev_release; + ubi->dev.class = &ubi_class; + ubi->dev.groups = ubi_dev_groups; + ubi->dev.parent = &mtd->dev; + + ubi->mtd = mtd; + ubi->ubi_num = ubi_num; + ubi->vid_hdr_offset = vid_hdr_offset; + ubi->autoresize_vol_id = -1; + ubi->wl_threshold = wl_threshold; + ubi->wl_free_max_diff = wl_threshold * 2; + +#ifdef CONFIG_MTD_UBI_FASTMAP + ubi->fm_pool.used = ubi->fm_pool.size = 0; + ubi->fm_wl_pool.used = ubi->fm_wl_pool.size = 0; + + /* + * fm_pool.max_size is 5% of the total number of PEBs but it's also + * between UBI_FM_MAX_POOL_SIZE and UBI_FM_MIN_POOL_SIZE. + */ + ubi->fm_pool.max_size = min(((int)mtd_div_by_eb(ubi->mtd->size, + ubi->mtd) / 100) * 5, UBI_FM_MAX_POOL_SIZE); + ubi->fm_pool.max_size = max(ubi->fm_pool.max_size, + UBI_FM_MIN_POOL_SIZE); + + ubi->fm_wl_pool.max_size = ubi->fm_pool.max_size / 2; + ubi->fm_pool_rsv_cnt = need_resv_pool ? ubi->fm_pool.max_size : 0; + ubi->fm_disabled = (!fm_autoconvert || disable_fm) ? 1 : 0; + if (fm_debug) + ubi_enable_dbg_chk_fastmap(ubi); + + if (!ubi->fm_disabled && (int)mtd_div_by_eb(ubi->mtd->size, ubi->mtd) + <= UBI_FM_MAX_START) { + ubi_err(ubi, "More than %i PEBs are needed for fastmap, sorry.", + UBI_FM_MAX_START); + ubi->fm_disabled = 1; + } + + ubi_msg(ubi, "default fastmap pool size: %d", ubi->fm_pool.max_size); + ubi_msg(ubi, "default fastmap WL pool size: %d", + ubi->fm_wl_pool.max_size); +#else + ubi->fm_disabled = 1; +#endif + mutex_init(&ubi->buf_mutex); + mutex_init(&ubi->ckvol_mutex); + mutex_init(&ubi->device_mutex); + spin_lock_init(&ubi->volumes_lock); + init_rwsem(&ubi->fm_protect); + init_rwsem(&ubi->fm_eba_sem); + + ubi_msg(ubi, "attaching mtd%d", mtd->index); + + err = io_init(ubi, max_beb_per1024); + if (err) + goto out_free; + + err = -ENOMEM; + ubi->peb_buf = vmalloc(ubi->peb_size); + if (!ubi->peb_buf) + goto out_free; + +#ifdef CONFIG_MTD_UBI_FASTMAP + ubi->fm_size = ubi_calc_fm_size(ubi); + ubi->fm_buf = vzalloc(ubi->fm_size); + if (!ubi->fm_buf) + goto out_free; +#endif + err = ubi_attach(ubi, disable_fm ? 1 : 0); + if (err) { + ubi_err(ubi, "failed to attach mtd%d, error %d", + mtd->index, err); + goto out_free; + } + + if (ubi->autoresize_vol_id != -1) { + err = autoresize(ubi, ubi->autoresize_vol_id); + if (err) + goto out_detach; + } + + err = uif_init(ubi); + if (err) + goto out_detach; + + err = ubi_debugfs_init_dev(ubi); + if (err) + goto out_uif; + + ubi->bgt_thread = kthread_create(ubi_thread, ubi, "%s", ubi->bgt_name); + if (IS_ERR(ubi->bgt_thread)) { + err = PTR_ERR(ubi->bgt_thread); + ubi_err(ubi, "cannot spawn \"%s\", error %d", + ubi->bgt_name, err); + goto out_debugfs; + } + + ubi_msg(ubi, "attached mtd%d (name \"%s\", size %llu MiB)", + mtd->index, mtd->name, ubi->flash_size >> 20); + ubi_msg(ubi, "PEB size: %d bytes (%d KiB), LEB size: %d bytes", + ubi->peb_size, ubi->peb_size >> 10, ubi->leb_size); + ubi_msg(ubi, "min./max. I/O unit sizes: %d/%d, sub-page size %d", + ubi->min_io_size, ubi->max_write_size, ubi->hdrs_min_io_size); + ubi_msg(ubi, "VID header offset: %d (aligned %d), data offset: %d", + ubi->vid_hdr_offset, ubi->vid_hdr_aloffset, ubi->leb_start); + ubi_msg(ubi, "good PEBs: %d, bad PEBs: %d, corrupted PEBs: %d", + ubi->good_peb_count, ubi->bad_peb_count, ubi->corr_peb_count); + ubi_msg(ubi, "user volume: %d, internal volumes: %d, max. volumes count: %d", + ubi->vol_count - UBI_INT_VOL_COUNT, UBI_INT_VOL_COUNT, + ubi->vtbl_slots); + ubi_msg(ubi, "max/mean erase counter: %d/%d, WL threshold: %d, image sequence number: %u", + ubi->max_ec, ubi->mean_ec, ubi->wl_threshold, + ubi->image_seq); + ubi_msg(ubi, "available PEBs: %d, total reserved PEBs: %d, PEBs reserved for bad PEB handling: %d", + ubi->avail_pebs, ubi->rsvd_pebs, ubi->beb_rsvd_pebs); + + /* + * The below lock makes sure we do not race with 'ubi_thread()' which + * checks @ubi->thread_enabled. Otherwise we may fail to wake it up. + */ + spin_lock(&ubi->wl_lock); + ubi->thread_enabled = 1; + wake_up_process(ubi->bgt_thread); + spin_unlock(&ubi->wl_lock); + + ubi_devices[ubi_num] = ubi; + ubi_notify_all(ubi, UBI_VOLUME_ADDED, NULL); + return ubi_num; + +out_debugfs: + ubi_debugfs_exit_dev(ubi); +out_uif: + uif_close(ubi); +out_detach: + ubi_wl_close(ubi); + ubi_free_all_volumes(ubi); + vfree(ubi->vtbl); +out_free: + vfree(ubi->peb_buf); + vfree(ubi->fm_buf); + put_device(&ubi->dev); + return err; +} + +/** + * ubi_detach_mtd_dev - detach an MTD device. + * @ubi_num: UBI device number to detach from + * @anyway: detach MTD even if device reference count is not zero + * + * This function destroys an UBI device number @ubi_num and detaches the + * underlying MTD device. Returns zero in case of success and %-EBUSY if the + * UBI device is busy and cannot be destroyed, and %-EINVAL if it does not + * exist. + * + * Note, the invocations of this function has to be serialized by the + * @ubi_devices_mutex. + */ +int ubi_detach_mtd_dev(int ubi_num, int anyway) +{ + struct ubi_device *ubi; + + if (ubi_num < 0 || ubi_num >= UBI_MAX_DEVICES) + return -EINVAL; + + ubi = ubi_get_device(ubi_num); + if (!ubi) + return -EINVAL; + + spin_lock(&ubi_devices_lock); + ubi->ref_count -= 1; + if (ubi->ref_count) { + if (!anyway) { + put_device(&ubi->dev); + spin_unlock(&ubi_devices_lock); + return -EBUSY; + } + /* This may only happen if there is a bug */ + ubi_err(ubi, "%s reference count %d, destroy anyway", + ubi->ubi_name, ubi->ref_count); + } + ubi->is_dead = true; + spin_unlock(&ubi_devices_lock); + + ubi_notify_all(ubi, UBI_VOLUME_SHUTDOWN, NULL); + + spin_lock(&ubi_devices_lock); + put_device(&ubi->dev); + ubi_devices[ubi_num] = NULL; + spin_unlock(&ubi_devices_lock); + + ubi_assert(ubi_num == ubi->ubi_num); + ubi_notify_all(ubi, UBI_VOLUME_REMOVED, NULL); + ubi_msg(ubi, "detaching mtd%d", ubi->mtd->index); +#ifdef CONFIG_MTD_UBI_FASTMAP + /* If we don't write a new fastmap at detach time we lose all + * EC updates that have been made since the last written fastmap. + * In case of fastmap debugging we omit the update to simulate an + * unclean shutdown. */ + if (!ubi_dbg_chk_fastmap(ubi)) + ubi_update_fastmap(ubi); +#endif + /* + * Before freeing anything, we have to stop the background thread to + * prevent it from doing anything on this device while we are freeing. + */ + if (ubi->bgt_thread) + kthread_stop(ubi->bgt_thread); + +#ifdef CONFIG_MTD_UBI_FASTMAP + cancel_work_sync(&ubi->fm_work); +#endif + ubi_debugfs_exit_dev(ubi); + uif_close(ubi); + + ubi_wl_close(ubi); + ubi_free_internal_volumes(ubi); + vfree(ubi->vtbl); + vfree(ubi->peb_buf); + vfree(ubi->fm_buf); + ubi_msg(ubi, "mtd%d is detached", ubi->mtd->index); + put_mtd_device(ubi->mtd); + put_device(&ubi->dev); + return 0; +} + +/** + * open_mtd_by_chdev - open an MTD device by its character device node path. + * @mtd_dev: MTD character device node path + * + * This helper function opens an MTD device by its character node device path. + * Returns MTD device description object in case of success and a negative + * error code in case of failure. + */ +static struct mtd_info * __init open_mtd_by_chdev(const char *mtd_dev) +{ + int err, minor; + struct path path; + struct kstat stat; + + /* Probably this is an MTD character device node path */ + err = kern_path(mtd_dev, LOOKUP_FOLLOW, &path); + if (err) + return ERR_PTR(err); + + err = vfs_getattr(&path, &stat, STATX_TYPE, AT_STATX_SYNC_AS_STAT); + path_put(&path); + if (err) + return ERR_PTR(err); + + /* MTD device number is defined by the major / minor numbers */ + if (MAJOR(stat.rdev) != MTD_CHAR_MAJOR || !S_ISCHR(stat.mode)) + return ERR_PTR(-EINVAL); + + minor = MINOR(stat.rdev); + + if (minor & 1) + /* + * Just do not think the "/dev/mtdrX" devices support is need, + * so do not support them to avoid doing extra work. + */ + return ERR_PTR(-EINVAL); + + return get_mtd_device(NULL, minor / 2); +} + +/** + * open_mtd_device - open MTD device by name, character device path, or number. + * @mtd_dev: name, character device node path, or MTD device device number + * + * This function tries to open and MTD device described by @mtd_dev string, + * which is first treated as ASCII MTD device number, and if it is not true, it + * is treated as MTD device name, and if that is also not true, it is treated + * as MTD character device node path. Returns MTD device description object in + * case of success and a negative error code in case of failure. + */ +static struct mtd_info * __init open_mtd_device(const char *mtd_dev) +{ + struct mtd_info *mtd; + int mtd_num; + char *endp; + + mtd_num = simple_strtoul(mtd_dev, &endp, 0); + if (*endp != '\0' || mtd_dev == endp) { + /* + * This does not look like an ASCII integer, probably this is + * MTD device name. + */ + mtd = get_mtd_device_nm(mtd_dev); + if (PTR_ERR(mtd) == -ENODEV) + /* Probably this is an MTD character device node path */ + mtd = open_mtd_by_chdev(mtd_dev); + } else + mtd = get_mtd_device(NULL, mtd_num); + + return mtd; +} + +static void ubi_notify_add(struct mtd_info *mtd) +{ + struct device_node *np = mtd_get_of_node(mtd); + int err; + + if (!of_device_is_compatible(np, "linux,ubi")) + return; + + /* + * we are already holding &mtd_table_mutex, but still need + * to bump refcount + */ + err = __get_mtd_device(mtd); + if (err) + return; + + /* called while holding mtd_table_mutex */ + mutex_lock_nested(&ubi_devices_mutex, SINGLE_DEPTH_NESTING); + err = ubi_attach_mtd_dev(mtd, UBI_DEV_NUM_AUTO, 0, 0, false, false, 0); + mutex_unlock(&ubi_devices_mutex); + if (err < 0) + __put_mtd_device(mtd); +} + +static void ubi_notify_remove(struct mtd_info *mtd) +{ + /* do nothing for now */ +} + +static struct mtd_notifier ubi_mtd_notifier = { + .add = ubi_notify_add, + .remove = ubi_notify_remove, +}; + +static int __init ubi_init_attach(void) +{ + int err, i, k; + + /* Attach MTD devices */ + for (i = 0; i < mtd_devs; i++) { + struct mtd_dev_param *p = &mtd_dev_param[i]; + struct mtd_info *mtd; + + cond_resched(); + + mtd = open_mtd_device(p->name); + if (IS_ERR(mtd)) { + err = PTR_ERR(mtd); + pr_err("UBI error: cannot open mtd %s, error %d\n", + p->name, err); + /* See comment below re-ubi_is_module(). */ + if (ubi_is_module()) + goto out_detach; + continue; + } + + mutex_lock(&ubi_devices_mutex); + err = ubi_attach_mtd_dev(mtd, p->ubi_num, + p->vid_hdr_offs, p->max_beb_per1024, + p->enable_fm == 0, + p->need_resv_pool != 0, + p->wl_threshold); + mutex_unlock(&ubi_devices_mutex); + if (err < 0) { + pr_err("UBI error: cannot attach mtd%d\n", + mtd->index); + put_mtd_device(mtd); + + /* + * Originally UBI stopped initializing on any error. + * However, later on it was found out that this + * behavior is not very good when UBI is compiled into + * the kernel and the MTD devices to attach are passed + * through the command line. Indeed, UBI failure + * stopped whole boot sequence. + * + * To fix this, we changed the behavior for the + * non-module case, but preserved the old behavior for + * the module case, just for compatibility. This is a + * little inconsistent, though. + */ + if (ubi_is_module()) + goto out_detach; + } + } + + return 0; + +out_detach: + for (k = 0; k < UBI_MAX_DEVICES; k++) + if (ubi_devices[k]) { + mutex_lock(&ubi_devices_mutex); + ubi_detach_mtd_dev(k, 1); + mutex_unlock(&ubi_devices_mutex); + } + return err; +} +#ifndef CONFIG_MTD_UBI_MODULE +late_initcall(ubi_init_attach); +#endif + +static int __init ubi_init(void) +{ + int err; + + /* Ensure that EC and VID headers have correct size */ + BUILD_BUG_ON(sizeof(struct ubi_ec_hdr) != 64); + BUILD_BUG_ON(sizeof(struct ubi_vid_hdr) != 64); + + if (mtd_devs > UBI_MAX_DEVICES) { + pr_err("UBI error: too many MTD devices, maximum is %d\n", + UBI_MAX_DEVICES); + return -EINVAL; + } + + /* Create base sysfs directory and sysfs files */ + err = class_register(&ubi_class); + if (err < 0) + return err; + + err = misc_register(&ubi_ctrl_cdev); + if (err) { + pr_err("UBI error: cannot register device\n"); + goto out; + } + + ubi_wl_entry_slab = kmem_cache_create("ubi_wl_entry_slab", + sizeof(struct ubi_wl_entry), + 0, 0, NULL); + if (!ubi_wl_entry_slab) { + err = -ENOMEM; + goto out_dev_unreg; + } + + err = ubi_debugfs_init(); + if (err) + goto out_slab; + + err = ubiblock_init(); + if (err) { + pr_err("UBI error: block: cannot initialize, error %d\n", err); + + /* See comment above re-ubi_is_module(). */ + if (ubi_is_module()) + goto out_debugfs; + } + + register_mtd_user(&ubi_mtd_notifier); + + if (ubi_is_module()) { + err = ubi_init_attach(); + if (err) + goto out_mtd_notifier; + } + + return 0; + +out_mtd_notifier: + unregister_mtd_user(&ubi_mtd_notifier); + ubiblock_exit(); +out_debugfs: + ubi_debugfs_exit(); +out_slab: + kmem_cache_destroy(ubi_wl_entry_slab); +out_dev_unreg: + misc_deregister(&ubi_ctrl_cdev); +out: + class_unregister(&ubi_class); + pr_err("UBI error: cannot initialize UBI, error %d\n", err); + return err; +} +device_initcall(ubi_init); + + +static void __exit ubi_exit(void) +{ + int i; + + ubiblock_exit(); + unregister_mtd_user(&ubi_mtd_notifier); + + for (i = 0; i < UBI_MAX_DEVICES; i++) + if (ubi_devices[i]) { + mutex_lock(&ubi_devices_mutex); + ubi_detach_mtd_dev(ubi_devices[i]->ubi_num, 1); + mutex_unlock(&ubi_devices_mutex); + } + ubi_debugfs_exit(); + kmem_cache_destroy(ubi_wl_entry_slab); + misc_deregister(&ubi_ctrl_cdev); + class_unregister(&ubi_class); +} +module_exit(ubi_exit); + +/** + * bytes_str_to_int - convert a number of bytes string into an integer. + * @str: the string to convert + * + * This function returns positive resulting integer in case of success and a + * negative error code in case of failure. + */ +static int bytes_str_to_int(const char *str) +{ + char *endp; + unsigned long result; + + result = simple_strtoul(str, &endp, 0); + if (str == endp || result >= INT_MAX) { + pr_err("UBI error: incorrect bytes count: \"%s\"\n", str); + return -EINVAL; + } + + switch (*endp) { + case 'G': + result *= 1024; + fallthrough; + case 'M': + result *= 1024; + fallthrough; + case 'K': + result *= 1024; + break; + case '\0': + break; + default: + pr_err("UBI error: incorrect bytes count: \"%s\"\n", str); + return -EINVAL; + } + + return result; +} + +/** + * ubi_mtd_param_parse - parse the 'mtd=' UBI parameter. + * @val: the parameter value to parse + * @kp: not used + * + * This function returns zero in case of success and a negative error code in + * case of error. + */ +static int ubi_mtd_param_parse(const char *val, const struct kernel_param *kp) +{ + int i, len; + struct mtd_dev_param *p; + char buf[MTD_PARAM_LEN_MAX]; + char *pbuf = &buf[0]; + char *tokens[MTD_PARAM_MAX_COUNT], *token; + + if (!val) + return -EINVAL; + + if (mtd_devs == UBI_MAX_DEVICES) { + pr_err("UBI error: too many parameters, max. is %d\n", + UBI_MAX_DEVICES); + return -EINVAL; + } + + len = strnlen(val, MTD_PARAM_LEN_MAX); + if (len == MTD_PARAM_LEN_MAX) { + pr_err("UBI error: parameter \"%s\" is too long, max. is %d\n", + val, MTD_PARAM_LEN_MAX); + return -EINVAL; + } + + if (len == 0) { + pr_warn("UBI warning: empty 'mtd=' parameter - ignored\n"); + return 0; + } + + strcpy(buf, val); + + /* Get rid of the final newline */ + if (buf[len - 1] == '\n') + buf[len - 1] = '\0'; + + for (i = 0; i < MTD_PARAM_MAX_COUNT; i++) + tokens[i] = strsep(&pbuf, ","); + + if (pbuf) { + pr_err("UBI error: too many arguments at \"%s\"\n", val); + return -EINVAL; + } + + p = &mtd_dev_param[mtd_devs]; + strcpy(&p->name[0], tokens[0]); + + token = tokens[1]; + if (token) { + p->vid_hdr_offs = bytes_str_to_int(token); + + if (p->vid_hdr_offs < 0) + return p->vid_hdr_offs; + } + + token = tokens[2]; + if (token) { + int err = kstrtoint(token, 10, &p->max_beb_per1024); + + if (err) { + pr_err("UBI error: bad value for max_beb_per1024 parameter: %s\n", + token); + return -EINVAL; + } + } + + token = tokens[3]; + if (token) { + int err = kstrtoint(token, 10, &p->ubi_num); + + if (err || p->ubi_num < UBI_DEV_NUM_AUTO) { + pr_err("UBI error: bad value for ubi_num parameter: %s\n", + token); + return -EINVAL; + } + } else + p->ubi_num = UBI_DEV_NUM_AUTO; + + token = tokens[4]; + if (token) { + int err = kstrtoint(token, 10, &p->enable_fm); + + if (err) { + pr_err("UBI error: bad value for enable_fm parameter: %s\n", + token); + return -EINVAL; + } + } else + p->enable_fm = 0; + + token = tokens[5]; + if (token) { + int err = kstrtoint(token, 10, &p->need_resv_pool); + + if (err) { + pr_err("UBI error: bad value for need_resv_pool parameter: %s\n", + token); + return -EINVAL; + } + } else + p->need_resv_pool = 0; + + token = tokens[6]; + if (token) { + int err = kstrtoint(token, 10, &p->wl_threshold); + + if (err) { + pr_err("UBI error: bad value for wl_threshold parameter: %s\n", + token); + return -EINVAL; + } + } else + p->wl_threshold = 0; + + mtd_devs += 1; + return 0; +} + +module_param_call(mtd, ubi_mtd_param_parse, NULL, NULL, 0400); +MODULE_PARM_DESC(mtd, "MTD devices to attach. Parameter format: mtd=[,[,max_beb_per1024[,ubi_num[,enable_fm[,need_resv_pool[,wl_threshold]]]]]].\n" + "Multiple \"mtd\" parameters may be specified.\n" + "MTD devices may be specified by their number, name, or path to the MTD character device node.\n" + "Optional \"vid_hdr_offs\" parameter specifies UBI VID header position to be used by UBI. (default value if 0)\n" + "Optional \"max_beb_per1024\" parameter specifies the maximum expected bad eraseblock per 1024 eraseblocks. (default value (" + __stringify(CONFIG_MTD_UBI_BEB_LIMIT) ") if 0)\n" + "Optional \"ubi_num\" parameter specifies UBI device number which have to be assigned to the newly created UBI device (assigned automatically by default)\n" + "Optional \"enable_fm\" parameter determines whether to enable fastmap during attach. If the value is non-zero, fastmap is enabled. Default value is 0.\n" + "Optional \"need_resv_pool\" parameter determines whether to reserve pool->max_size pebs during attach. If the value is non-zero, peb reservation is enabled. Default value is 0.\n" + "\n" + "Example 1: mtd=/dev/mtd0 - attach MTD device /dev/mtd0.\n" + "Example 2: mtd=content,1984 mtd=4 - attach MTD device with name \"content\" using VID header offset 1984, and MTD device number 4 with default VID header offset.\n" + "Example 3: mtd=/dev/mtd1,0,25 - attach MTD device /dev/mtd1 using default VID header offset and reserve 25*nand_size_in_blocks/1024 erase blocks for bad block handling.\n" + "\t(e.g. if the NAND *chipset* has 4096 PEB, 100 will be reserved for this UBI device).\n" + "Example 4: mtd=/dev/mtd1,0,0,5 - attach MTD device /dev/mtd1 to UBI 5 and using default values for the other fields.\n" + "example 5: mtd=1,0,0,5 mtd=2,0,0,6,1 - attach MTD device /dev/mtd1 to UBI 5 and disable fastmap; attach MTD device /dev/mtd2 to UBI 6 and enable fastmap.(only works when fastmap is enabled and fm_autoconvert=Y).\n" + "Example 6: mtd=/dev/mtd0,0,0,0,0,0,256 mtd=/dev/mtd1,0,0,0,0,0,4096 - attach MTD device /dev/mtd0 with wear-leveling threshold 256, and MTD device /dev/mtd1 with threshold 4096.\n"); +#ifdef CONFIG_MTD_UBI_FASTMAP +module_param(fm_autoconvert, bool, 0644); +MODULE_PARM_DESC(fm_autoconvert, "Set this parameter to enable fastmap automatically on images without a fastmap."); +module_param(fm_debug, bool, 0); +MODULE_PARM_DESC(fm_debug, "Set this parameter to enable fastmap debugging by default. Warning, this will make fastmap slow!"); +#endif +MODULE_VERSION(__stringify(UBI_VERSION)); +MODULE_DESCRIPTION("UBI - Unsorted Block Images"); +MODULE_AUTHOR("Artem Bityutskiy"); +MODULE_LICENSE("GPL"); diff --git a/drivers/mtd/ubi/cdev.c b/drivers/mtd/ubi/cdev.c new file mode 100644 index 000000000..54acd8394 --- /dev/null +++ b/drivers/mtd/ubi/cdev.c @@ -0,0 +1,1177 @@ +// SPDX-License-Identifier: GPL-2.0-or-later +/* + * Copyright (c) International Business Machines Corp., 2006 + * + * Author: Artem Bityutskiy (Битюцкий Артём) + */ + +/* + * This file includes implementation of UBI character device operations. + * + * There are two kinds of character devices in UBI: UBI character devices and + * UBI volume character devices. UBI character devices allow users to + * manipulate whole volumes: create, remove, and re-size them. Volume character + * devices provide volume I/O capabilities. + * + * Major and minor numbers are assigned dynamically to both UBI and volume + * character devices. + * + * Well, there is the third kind of character devices - the UBI control + * character device, which allows to manipulate by UBI devices - create and + * delete them. In other words, it is used for attaching and detaching MTD + * devices. + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include "ubi.h" + +/** + * get_exclusive - get exclusive access to an UBI volume. + * @desc: volume descriptor + * + * This function changes UBI volume open mode to "exclusive". Returns previous + * mode value (positive integer) in case of success and a negative error code + * in case of failure. + */ +static int get_exclusive(struct ubi_volume_desc *desc) +{ + int users, err; + struct ubi_volume *vol = desc->vol; + + spin_lock(&vol->ubi->volumes_lock); + users = vol->readers + vol->writers + vol->exclusive + vol->metaonly; + ubi_assert(users > 0); + if (users > 1) { + ubi_err(vol->ubi, "%d users for volume %d", users, vol->vol_id); + err = -EBUSY; + } else { + vol->readers = vol->writers = vol->metaonly = 0; + vol->exclusive = 1; + err = desc->mode; + desc->mode = UBI_EXCLUSIVE; + } + spin_unlock(&vol->ubi->volumes_lock); + + return err; +} + +/** + * revoke_exclusive - revoke exclusive mode. + * @desc: volume descriptor + * @mode: new mode to switch to + */ +static void revoke_exclusive(struct ubi_volume_desc *desc, int mode) +{ + struct ubi_volume *vol = desc->vol; + + spin_lock(&vol->ubi->volumes_lock); + ubi_assert(vol->readers == 0 && vol->writers == 0 && vol->metaonly == 0); + ubi_assert(vol->exclusive == 1 && desc->mode == UBI_EXCLUSIVE); + vol->exclusive = 0; + if (mode == UBI_READONLY) + vol->readers = 1; + else if (mode == UBI_READWRITE) + vol->writers = 1; + else if (mode == UBI_METAONLY) + vol->metaonly = 1; + else + vol->exclusive = 1; + spin_unlock(&vol->ubi->volumes_lock); + + desc->mode = mode; +} + +static int vol_cdev_open(struct inode *inode, struct file *file) +{ + struct ubi_volume_desc *desc; + int vol_id = iminor(inode) - 1, mode, ubi_num; + + ubi_num = ubi_major2num(imajor(inode)); + if (ubi_num < 0) + return ubi_num; + + if (file->f_mode & FMODE_WRITE) + mode = UBI_READWRITE; + else + mode = UBI_READONLY; + + dbg_gen("open device %d, volume %d, mode %d", + ubi_num, vol_id, mode); + + desc = ubi_open_volume(ubi_num, vol_id, mode); + if (IS_ERR(desc)) + return PTR_ERR(desc); + + file->private_data = desc; + return 0; +} + +static int vol_cdev_release(struct inode *inode, struct file *file) +{ + struct ubi_volume_desc *desc = file->private_data; + struct ubi_volume *vol = desc->vol; + + dbg_gen("release device %d, volume %d, mode %d", + vol->ubi->ubi_num, vol->vol_id, desc->mode); + + if (vol->updating) { + ubi_warn(vol->ubi, "update of volume %d not finished, volume is damaged", + vol->vol_id); + ubi_assert(!vol->changing_leb); + vol->updating = 0; + vfree(vol->upd_buf); + } else if (vol->changing_leb) { + dbg_gen("only %lld of %lld bytes received for atomic LEB change for volume %d:%d, cancel", + vol->upd_received, vol->upd_bytes, vol->ubi->ubi_num, + vol->vol_id); + vol->changing_leb = 0; + vfree(vol->upd_buf); + } + + ubi_close_volume(desc); + return 0; +} + +static loff_t vol_cdev_llseek(struct file *file, loff_t offset, int origin) +{ + struct ubi_volume_desc *desc = file->private_data; + struct ubi_volume *vol = desc->vol; + + if (vol->updating) { + /* Update is in progress, seeking is prohibited */ + ubi_err(vol->ubi, "updating"); + return -EBUSY; + } + + return fixed_size_llseek(file, offset, origin, vol->used_bytes); +} + +static int vol_cdev_fsync(struct file *file, loff_t start, loff_t end, + int datasync) +{ + struct ubi_volume_desc *desc = file->private_data; + struct ubi_device *ubi = desc->vol->ubi; + struct inode *inode = file_inode(file); + int err; + inode_lock(inode); + err = ubi_sync(ubi->ubi_num); + inode_unlock(inode); + return err; +} + + +static ssize_t vol_cdev_read(struct file *file, __user char *buf, size_t count, + loff_t *offp) +{ + struct ubi_volume_desc *desc = file->private_data; + struct ubi_volume *vol = desc->vol; + struct ubi_device *ubi = vol->ubi; + int err, lnum, off, len, tbuf_size; + size_t count_save = count; + void *tbuf; + + dbg_gen("read %zd bytes from offset %lld of volume %d", + count, *offp, vol->vol_id); + + if (vol->updating) { + ubi_err(vol->ubi, "updating"); + return -EBUSY; + } + if (vol->upd_marker) { + ubi_err(vol->ubi, "damaged volume, update marker is set"); + return -EBADF; + } + if (*offp == vol->used_bytes || count == 0) + return 0; + + if (vol->corrupted) + dbg_gen("read from corrupted volume %d", vol->vol_id); + + if (*offp + count > vol->used_bytes) + count_save = count = vol->used_bytes - *offp; + + tbuf_size = vol->usable_leb_size; + if (count < tbuf_size) + tbuf_size = ALIGN(count, ubi->min_io_size); + tbuf = vmalloc(tbuf_size); + if (!tbuf) + return -ENOMEM; + + len = count > tbuf_size ? tbuf_size : count; + lnum = div_u64_rem(*offp, vol->usable_leb_size, &off); + + do { + cond_resched(); + + if (off + len >= vol->usable_leb_size) + len = vol->usable_leb_size - off; + + err = ubi_eba_read_leb(ubi, vol, lnum, tbuf, off, len, 0); + if (err) + break; + + off += len; + if (off == vol->usable_leb_size) { + lnum += 1; + off -= vol->usable_leb_size; + } + + count -= len; + *offp += len; + + err = copy_to_user(buf, tbuf, len); + if (err) { + err = -EFAULT; + break; + } + + buf += len; + len = count > tbuf_size ? tbuf_size : count; + } while (count); + + vfree(tbuf); + return err ? err : count_save - count; +} + +/* + * This function allows to directly write to dynamic UBI volumes, without + * issuing the volume update operation. + */ +static ssize_t vol_cdev_direct_write(struct file *file, const char __user *buf, + size_t count, loff_t *offp) +{ + struct ubi_volume_desc *desc = file->private_data; + struct ubi_volume *vol = desc->vol; + struct ubi_device *ubi = vol->ubi; + int lnum, off, len, tbuf_size, err = 0; + size_t count_save = count; + char *tbuf; + + if (!vol->direct_writes) + return -EPERM; + + dbg_gen("requested: write %zd bytes to offset %lld of volume %u", + count, *offp, vol->vol_id); + + if (vol->vol_type == UBI_STATIC_VOLUME) + return -EROFS; + + lnum = div_u64_rem(*offp, vol->usable_leb_size, &off); + if (off & (ubi->min_io_size - 1)) { + ubi_err(ubi, "unaligned position"); + return -EINVAL; + } + + if (*offp + count > vol->used_bytes) + count_save = count = vol->used_bytes - *offp; + + /* We can write only in fractions of the minimum I/O unit */ + if (count & (ubi->min_io_size - 1)) { + ubi_err(ubi, "unaligned write length"); + return -EINVAL; + } + + tbuf_size = vol->usable_leb_size; + if (count < tbuf_size) + tbuf_size = ALIGN(count, ubi->min_io_size); + tbuf = vmalloc(tbuf_size); + if (!tbuf) + return -ENOMEM; + + len = count > tbuf_size ? tbuf_size : count; + + while (count) { + cond_resched(); + + if (off + len >= vol->usable_leb_size) + len = vol->usable_leb_size - off; + + err = copy_from_user(tbuf, buf, len); + if (err) { + err = -EFAULT; + break; + } + + err = ubi_eba_write_leb(ubi, vol, lnum, tbuf, off, len); + if (err) + break; + + off += len; + if (off == vol->usable_leb_size) { + lnum += 1; + off -= vol->usable_leb_size; + } + + count -= len; + *offp += len; + buf += len; + len = count > tbuf_size ? tbuf_size : count; + } + + vfree(tbuf); + return err ? err : count_save - count; +} + +static ssize_t vol_cdev_write(struct file *file, const char __user *buf, + size_t count, loff_t *offp) +{ + int err = 0; + struct ubi_volume_desc *desc = file->private_data; + struct ubi_volume *vol = desc->vol; + struct ubi_device *ubi = vol->ubi; + + if (!vol->updating && !vol->changing_leb) + return vol_cdev_direct_write(file, buf, count, offp); + + if (vol->updating) + err = ubi_more_update_data(ubi, vol, buf, count); + else + err = ubi_more_leb_change_data(ubi, vol, buf, count); + + if (err < 0) { + ubi_err(ubi, "cannot accept more %zd bytes of data, error %d", + count, err); + return err; + } + + if (err) { + /* + * The operation is finished, @err contains number of actually + * written bytes. + */ + count = err; + + if (vol->changing_leb) { + revoke_exclusive(desc, UBI_READWRITE); + return count; + } + + /* + * We voluntarily do not take into account the skip_check flag + * as we want to make sure what we wrote was correctly written. + */ + err = ubi_check_volume(ubi, vol->vol_id); + if (err < 0) + return err; + + if (err) { + ubi_warn(ubi, "volume %d on UBI device %d is corrupted", + vol->vol_id, ubi->ubi_num); + vol->corrupted = 1; + } + vol->checked = 1; + ubi_volume_notify(ubi, vol, UBI_VOLUME_UPDATED); + revoke_exclusive(desc, UBI_READWRITE); + } + + return count; +} + +static long vol_cdev_ioctl(struct file *file, unsigned int cmd, + unsigned long arg) +{ + int err = 0; + struct ubi_volume_desc *desc = file->private_data; + struct ubi_volume *vol = desc->vol; + struct ubi_device *ubi = vol->ubi; + void __user *argp = (void __user *)arg; + + switch (cmd) { + /* Volume update command */ + case UBI_IOCVOLUP: + { + int64_t bytes, rsvd_bytes; + + if (!capable(CAP_SYS_RESOURCE)) { + err = -EPERM; + break; + } + + err = copy_from_user(&bytes, argp, sizeof(int64_t)); + if (err) { + err = -EFAULT; + break; + } + + if (desc->mode == UBI_READONLY) { + err = -EROFS; + break; + } + + rsvd_bytes = (long long)vol->reserved_pebs * + vol->usable_leb_size; + if (bytes < 0 || bytes > rsvd_bytes) { + err = -EINVAL; + break; + } + + err = get_exclusive(desc); + if (err < 0) + break; + + err = ubi_start_update(ubi, vol, bytes); + if (bytes == 0) { + ubi_volume_notify(ubi, vol, UBI_VOLUME_UPDATED); + revoke_exclusive(desc, UBI_READWRITE); + } + break; + } + + /* Atomic logical eraseblock change command */ + case UBI_IOCEBCH: + { + struct ubi_leb_change_req req; + + err = copy_from_user(&req, argp, + sizeof(struct ubi_leb_change_req)); + if (err) { + err = -EFAULT; + break; + } + + if (desc->mode == UBI_READONLY || + vol->vol_type == UBI_STATIC_VOLUME) { + err = -EROFS; + break; + } + + /* Validate the request */ + err = -EINVAL; + if (!ubi_leb_valid(vol, req.lnum) || + req.bytes < 0 || req.bytes > vol->usable_leb_size) + break; + + err = get_exclusive(desc); + if (err < 0) + break; + + err = ubi_start_leb_change(ubi, vol, &req); + if (req.bytes == 0) + revoke_exclusive(desc, UBI_READWRITE); + break; + } + + /* Logical eraseblock erasure command */ + case UBI_IOCEBER: + { + int32_t lnum; + + err = get_user(lnum, (__user int32_t *)argp); + if (err) { + err = -EFAULT; + break; + } + + if (desc->mode == UBI_READONLY || + vol->vol_type == UBI_STATIC_VOLUME) { + err = -EROFS; + break; + } + + if (!ubi_leb_valid(vol, lnum)) { + err = -EINVAL; + break; + } + + dbg_gen("erase LEB %d:%d", vol->vol_id, lnum); + err = ubi_eba_unmap_leb(ubi, vol, lnum); + if (err) + break; + + err = ubi_wl_flush(ubi, UBI_ALL, UBI_ALL); + break; + } + + /* Logical eraseblock map command */ + case UBI_IOCEBMAP: + { + struct ubi_map_req req; + + err = copy_from_user(&req, argp, sizeof(struct ubi_map_req)); + if (err) { + err = -EFAULT; + break; + } + err = ubi_leb_map(desc, req.lnum); + break; + } + + /* Logical eraseblock un-map command */ + case UBI_IOCEBUNMAP: + { + int32_t lnum; + + err = get_user(lnum, (__user int32_t *)argp); + if (err) { + err = -EFAULT; + break; + } + err = ubi_leb_unmap(desc, lnum); + break; + } + + /* Check if logical eraseblock is mapped command */ + case UBI_IOCEBISMAP: + { + int32_t lnum; + + err = get_user(lnum, (__user int32_t *)argp); + if (err) { + err = -EFAULT; + break; + } + err = ubi_is_mapped(desc, lnum); + break; + } + + /* Set volume property command */ + case UBI_IOCSETVOLPROP: + { + struct ubi_set_vol_prop_req req; + + err = copy_from_user(&req, argp, + sizeof(struct ubi_set_vol_prop_req)); + if (err) { + err = -EFAULT; + break; + } + switch (req.property) { + case UBI_VOL_PROP_DIRECT_WRITE: + mutex_lock(&ubi->device_mutex); + desc->vol->direct_writes = !!req.value; + mutex_unlock(&ubi->device_mutex); + break; + default: + err = -EINVAL; + break; + } + break; + } + + /* Create a R/O block device on top of the UBI volume */ + case UBI_IOCVOLCRBLK: + { + struct ubi_volume_info vi; + + ubi_get_volume_info(desc, &vi); + err = ubiblock_create(&vi); + break; + } + + /* Remove the R/O block device */ + case UBI_IOCVOLRMBLK: + { + struct ubi_volume_info vi; + + ubi_get_volume_info(desc, &vi); + err = ubiblock_remove(&vi); + break; + } + + default: + err = -ENOTTY; + break; + } + return err; +} + +/** + * verify_mkvol_req - verify volume creation request. + * @ubi: UBI device description object + * @req: the request to check + * + * This function zero if the request is correct, and %-EINVAL if not. + */ +static int verify_mkvol_req(const struct ubi_device *ubi, + const struct ubi_mkvol_req *req) +{ + int n, err = -EINVAL; + + if (req->bytes < 0 || req->alignment < 0 || req->vol_type < 0 || + req->name_len < 0) + goto bad; + + if ((req->vol_id < 0 || req->vol_id >= ubi->vtbl_slots) && + req->vol_id != UBI_VOL_NUM_AUTO) + goto bad; + + if (req->alignment == 0) + goto bad; + + if (req->bytes == 0) + goto bad; + + if (req->vol_type != UBI_DYNAMIC_VOLUME && + req->vol_type != UBI_STATIC_VOLUME) + goto bad; + + if (req->flags & ~UBI_VOL_VALID_FLGS) + goto bad; + + if (req->flags & UBI_VOL_SKIP_CRC_CHECK_FLG && + req->vol_type != UBI_STATIC_VOLUME) + goto bad; + + if (req->alignment > ubi->leb_size) + goto bad; + + n = req->alignment & (ubi->min_io_size - 1); + if (req->alignment != 1 && n) + goto bad; + + if (!req->name[0] || !req->name_len) + goto bad; + + if (req->name_len > UBI_VOL_NAME_MAX) { + err = -ENAMETOOLONG; + goto bad; + } + + n = strnlen(req->name, req->name_len + 1); + if (n != req->name_len) + goto bad; + + return 0; + +bad: + ubi_err(ubi, "bad volume creation request"); + ubi_dump_mkvol_req(req); + return err; +} + +/** + * verify_rsvol_req - verify volume re-size request. + * @ubi: UBI device description object + * @req: the request to check + * + * This function returns zero if the request is correct, and %-EINVAL if not. + */ +static int verify_rsvol_req(const struct ubi_device *ubi, + const struct ubi_rsvol_req *req) +{ + if (req->bytes <= 0) + return -EINVAL; + + if (req->vol_id < 0 || req->vol_id >= ubi->vtbl_slots) + return -EINVAL; + + return 0; +} + +/** + * rename_volumes - rename UBI volumes. + * @ubi: UBI device description object + * @req: volumes re-name request + * + * This is a helper function for the volume re-name IOCTL which validates the + * request, opens the volume and calls corresponding volumes management + * function. Returns zero in case of success and a negative error code in case + * of failure. + */ +static int rename_volumes(struct ubi_device *ubi, + struct ubi_rnvol_req *req) +{ + int i, n, err; + struct list_head rename_list; + struct ubi_rename_entry *re, *re1; + + if (req->count < 0 || req->count > UBI_MAX_RNVOL) + return -EINVAL; + + if (req->count == 0) + return 0; + + /* Validate volume IDs and names in the request */ + for (i = 0; i < req->count; i++) { + if (req->ents[i].vol_id < 0 || + req->ents[i].vol_id >= ubi->vtbl_slots) + return -EINVAL; + if (req->ents[i].name_len < 0) + return -EINVAL; + if (req->ents[i].name_len > UBI_VOL_NAME_MAX) + return -ENAMETOOLONG; + req->ents[i].name[req->ents[i].name_len] = '\0'; + n = strlen(req->ents[i].name); + if (n != req->ents[i].name_len) + return -EINVAL; + } + + /* Make sure volume IDs and names are unique */ + for (i = 0; i < req->count - 1; i++) { + for (n = i + 1; n < req->count; n++) { + if (req->ents[i].vol_id == req->ents[n].vol_id) { + ubi_err(ubi, "duplicated volume id %d", + req->ents[i].vol_id); + return -EINVAL; + } + if (!strcmp(req->ents[i].name, req->ents[n].name)) { + ubi_err(ubi, "duplicated volume name \"%s\"", + req->ents[i].name); + return -EINVAL; + } + } + } + + /* Create the re-name list */ + INIT_LIST_HEAD(&rename_list); + for (i = 0; i < req->count; i++) { + int vol_id = req->ents[i].vol_id; + int name_len = req->ents[i].name_len; + const char *name = req->ents[i].name; + + re = kzalloc_obj(struct ubi_rename_entry); + if (!re) { + err = -ENOMEM; + goto out_free; + } + + re->desc = ubi_open_volume(ubi->ubi_num, vol_id, UBI_METAONLY); + if (IS_ERR(re->desc)) { + err = PTR_ERR(re->desc); + ubi_err(ubi, "cannot open volume %d, error %d", + vol_id, err); + kfree(re); + goto out_free; + } + + /* Skip this re-naming if the name does not really change */ + if (re->desc->vol->name_len == name_len && + !memcmp(re->desc->vol->name, name, name_len)) { + ubi_close_volume(re->desc); + kfree(re); + continue; + } + + re->new_name_len = name_len; + memcpy(re->new_name, name, name_len); + list_add_tail(&re->list, &rename_list); + dbg_gen("will rename volume %d from \"%s\" to \"%s\"", + vol_id, re->desc->vol->name, name); + } + + if (list_empty(&rename_list)) + return 0; + + /* Find out the volumes which have to be removed */ + list_for_each_entry(re, &rename_list, list) { + struct ubi_volume_desc *desc; + int no_remove_needed = 0; + + /* + * Volume @re->vol_id is going to be re-named to + * @re->new_name, while its current name is @name. If a volume + * with name @re->new_name currently exists, it has to be + * removed, unless it is also re-named in the request (@req). + */ + list_for_each_entry(re1, &rename_list, list) { + if (re->new_name_len == re1->desc->vol->name_len && + !memcmp(re->new_name, re1->desc->vol->name, + re1->desc->vol->name_len)) { + no_remove_needed = 1; + break; + } + } + + if (no_remove_needed) + continue; + + /* + * It seems we need to remove volume with name @re->new_name, + * if it exists. + */ + desc = ubi_open_volume_nm(ubi->ubi_num, re->new_name, + UBI_EXCLUSIVE); + if (IS_ERR(desc)) { + err = PTR_ERR(desc); + if (err == -ENODEV) + /* Re-naming into a non-existing volume name */ + continue; + + /* The volume exists but busy, or an error occurred */ + ubi_err(ubi, "cannot open volume \"%s\", error %d", + re->new_name, err); + goto out_free; + } + + re1 = kzalloc_obj(struct ubi_rename_entry); + if (!re1) { + err = -ENOMEM; + ubi_close_volume(desc); + goto out_free; + } + + re1->remove = 1; + re1->desc = desc; + list_add(&re1->list, &rename_list); + dbg_gen("will remove volume %d, name \"%s\"", + re1->desc->vol->vol_id, re1->desc->vol->name); + } + + mutex_lock(&ubi->device_mutex); + err = ubi_rename_volumes(ubi, &rename_list); + mutex_unlock(&ubi->device_mutex); + +out_free: + list_for_each_entry_safe(re, re1, &rename_list, list) { + ubi_close_volume(re->desc); + list_del(&re->list); + kfree(re); + } + return err; +} + +static int ubi_get_ec_info(struct ubi_device *ubi, struct ubi_ecinfo_req __user *ureq) +{ + struct ubi_ecinfo_req req; + struct ubi_wl_entry *wl; + int read_cnt; + int peb; + int end_peb; + + /* Copy the input arguments */ + if (copy_from_user(&req, ureq, sizeof(struct ubi_ecinfo_req))) + return -EFAULT; + + /* Check input arguments */ + if (req.length <= 0 || req.start < 0 || req.start >= ubi->peb_count) + return -EINVAL; + + if (check_add_overflow(req.start, req.length, &end_peb)) + return -EINVAL; + + if (end_peb > ubi->peb_count) + end_peb = ubi->peb_count; + + /* Check access rights before filling erase_counters array */ + if (!access_ok((void __user *)ureq->erase_counters, + (end_peb-req.start) * sizeof(int32_t))) + return -EFAULT; + + /* Fill erase counter array */ + read_cnt = 0; + for (peb = req.start; peb < end_peb; read_cnt++, peb++) { + int ec; + + if (ubi_io_is_bad(ubi, peb)) { + if (__put_user(UBI_UNKNOWN, ureq->erase_counters+read_cnt)) + return -EFAULT; + + continue; + } + + spin_lock(&ubi->wl_lock); + + wl = ubi->lookuptbl[peb]; + if (wl) + ec = wl->ec; + else + ec = UBI_UNKNOWN; + + spin_unlock(&ubi->wl_lock); + + if (__put_user(ec, ureq->erase_counters+read_cnt)) + return -EFAULT; + + } + + /* Return actual read length */ + req.read_length = read_cnt; + + /* Copy everything except erase counter array */ + if (copy_to_user(ureq, &req, sizeof(struct ubi_ecinfo_req))) + return -EFAULT; + + return 0; +} + +static long ubi_cdev_ioctl(struct file *file, unsigned int cmd, + unsigned long arg) +{ + int err = 0; + struct ubi_device *ubi; + struct ubi_volume_desc *desc; + void __user *argp = (void __user *)arg; + + if (!capable(CAP_SYS_RESOURCE)) + return -EPERM; + + ubi = ubi_get_by_major(imajor(file->f_mapping->host)); + if (!ubi) + return -ENODEV; + + switch (cmd) { + /* Create volume command */ + case UBI_IOCMKVOL: + { + struct ubi_mkvol_req req; + + dbg_gen("create volume"); + err = copy_from_user(&req, argp, sizeof(struct ubi_mkvol_req)); + if (err) { + err = -EFAULT; + break; + } + + err = verify_mkvol_req(ubi, &req); + if (err) + break; + + mutex_lock(&ubi->device_mutex); + err = ubi_create_volume(ubi, &req); + mutex_unlock(&ubi->device_mutex); + if (err) + break; + + err = put_user(req.vol_id, (__user int32_t *)argp); + if (err) + err = -EFAULT; + + break; + } + + /* Remove volume command */ + case UBI_IOCRMVOL: + { + int vol_id; + + dbg_gen("remove volume"); + err = get_user(vol_id, (__user int32_t *)argp); + if (err) { + err = -EFAULT; + break; + } + + desc = ubi_open_volume(ubi->ubi_num, vol_id, UBI_EXCLUSIVE); + if (IS_ERR(desc)) { + err = PTR_ERR(desc); + break; + } + + mutex_lock(&ubi->device_mutex); + err = ubi_remove_volume(desc, 0); + mutex_unlock(&ubi->device_mutex); + + /* + * The volume is deleted (unless an error occurred), and the + * 'struct ubi_volume' object will be freed when + * 'ubi_close_volume()' will call 'put_device()'. + */ + ubi_close_volume(desc); + break; + } + + /* Re-size volume command */ + case UBI_IOCRSVOL: + { + int pebs; + struct ubi_rsvol_req req; + + dbg_gen("re-size volume"); + err = copy_from_user(&req, argp, sizeof(struct ubi_rsvol_req)); + if (err) { + err = -EFAULT; + break; + } + + err = verify_rsvol_req(ubi, &req); + if (err) + break; + + desc = ubi_open_volume(ubi->ubi_num, req.vol_id, UBI_EXCLUSIVE); + if (IS_ERR(desc)) { + err = PTR_ERR(desc); + break; + } + + pebs = div_u64(req.bytes + desc->vol->usable_leb_size - 1, + desc->vol->usable_leb_size); + + mutex_lock(&ubi->device_mutex); + err = ubi_resize_volume(desc, pebs); + mutex_unlock(&ubi->device_mutex); + ubi_close_volume(desc); + break; + } + + /* Re-name volumes command */ + case UBI_IOCRNVOL: + { + struct ubi_rnvol_req *req; + + dbg_gen("re-name volumes"); + req = kmalloc_obj(struct ubi_rnvol_req); + if (!req) { + err = -ENOMEM; + break; + } + + err = copy_from_user(req, argp, sizeof(struct ubi_rnvol_req)); + if (err) { + err = -EFAULT; + kfree(req); + break; + } + + err = rename_volumes(ubi, req); + kfree(req); + break; + } + + /* Check a specific PEB for bitflips and scrub it if needed */ + case UBI_IOCRPEB: + { + int pnum; + + err = get_user(pnum, (__user int32_t *)argp); + if (err) { + err = -EFAULT; + break; + } + + err = ubi_bitflip_check(ubi, pnum, 0); + break; + } + + /* Force scrubbing for a specific PEB */ + case UBI_IOCSPEB: + { + int pnum; + + err = get_user(pnum, (__user int32_t *)argp); + if (err) { + err = -EFAULT; + break; + } + + err = ubi_bitflip_check(ubi, pnum, 1); + break; + } + + case UBI_IOCECNFO: + { + err = ubi_get_ec_info(ubi, argp); + break; + } + + default: + err = -ENOTTY; + break; + } + + ubi_put_device(ubi); + return err; +} + +static long ctrl_cdev_ioctl(struct file *file, unsigned int cmd, + unsigned long arg) +{ + int err = 0; + void __user *argp = (void __user *)arg; + + if (!capable(CAP_SYS_RESOURCE)) + return -EPERM; + + switch (cmd) { + /* Attach an MTD device command */ + case UBI_IOCATT: + { + struct ubi_attach_req req; + struct mtd_info *mtd; + + dbg_gen("attach MTD device"); + err = copy_from_user(&req, argp, sizeof(struct ubi_attach_req)); + if (err) { + err = -EFAULT; + break; + } + + if (req.mtd_num < 0 || + (req.ubi_num < 0 && req.ubi_num != UBI_DEV_NUM_AUTO)) { + err = -EINVAL; + break; + } + + mtd = get_mtd_device(NULL, req.mtd_num); + if (IS_ERR(mtd)) { + err = PTR_ERR(mtd); + break; + } + + /* + * Note, further request verification is done by + * 'ubi_attach_mtd_dev()'. + */ + mutex_lock(&ubi_devices_mutex); + err = ubi_attach_mtd_dev(mtd, req.ubi_num, req.vid_hdr_offset, + req.max_beb_per1024, !!req.disable_fm, + !!req.need_resv_pool, req.wl_threshold); + mutex_unlock(&ubi_devices_mutex); + if (err < 0) + put_mtd_device(mtd); + else + /* @err contains UBI device number */ + err = put_user(err, (__user int32_t *)argp); + + break; + } + + /* Detach an MTD device command */ + case UBI_IOCDET: + { + int ubi_num; + + dbg_gen("detach MTD device"); + err = get_user(ubi_num, (__user int32_t *)argp); + if (err) { + err = -EFAULT; + break; + } + + mutex_lock(&ubi_devices_mutex); + err = ubi_detach_mtd_dev(ubi_num, 0); + mutex_unlock(&ubi_devices_mutex); + break; + } + + default: + err = -ENOTTY; + break; + } + + return err; +} + +/* UBI volume character device operations */ +const struct file_operations ubi_vol_cdev_operations = { + .owner = THIS_MODULE, + .open = vol_cdev_open, + .release = vol_cdev_release, + .llseek = vol_cdev_llseek, + .read = vol_cdev_read, + .write = vol_cdev_write, + .fsync = vol_cdev_fsync, + .unlocked_ioctl = vol_cdev_ioctl, + .compat_ioctl = compat_ptr_ioctl, +}; + +/* UBI character device operations */ +const struct file_operations ubi_cdev_operations = { + .owner = THIS_MODULE, + .unlocked_ioctl = ubi_cdev_ioctl, + .compat_ioctl = compat_ptr_ioctl, +}; + +/* UBI control character device operations */ +const struct file_operations ubi_ctrl_cdev_operations = { + .owner = THIS_MODULE, + .unlocked_ioctl = ctrl_cdev_ioctl, + .compat_ioctl = compat_ptr_ioctl, +}; diff --git a/drivers/mtd/ubi/debug.c b/drivers/mtd/ubi/debug.c new file mode 100644 index 000000000..d2a53961d --- /dev/null +++ b/drivers/mtd/ubi/debug.c @@ -0,0 +1,697 @@ +// SPDX-License-Identifier: GPL-2.0-or-later +/* + * Copyright (c) International Business Machines Corp., 2006 + * + * Author: Artem Bityutskiy (Битюцкий Артём) + */ + +#include "ubi.h" +#include +#include +#include +#include +#include + +#ifdef CONFIG_MTD_UBI_FAULT_INJECTION +static DECLARE_FAULT_ATTR(fault_eccerr_attr); +static DECLARE_FAULT_ATTR(fault_bitflips_attr); +static DECLARE_FAULT_ATTR(fault_read_failure_attr); +static DECLARE_FAULT_ATTR(fault_write_failure_attr); +static DECLARE_FAULT_ATTR(fault_erase_failure_attr); +static DECLARE_FAULT_ATTR(fault_power_cut_attr); +static DECLARE_FAULT_ATTR(fault_io_ff_attr); +static DECLARE_FAULT_ATTR(fault_io_ff_bitflips_attr); +static DECLARE_FAULT_ATTR(fault_bad_hdr_attr); +static DECLARE_FAULT_ATTR(fault_bad_hdr_ebadmsg_attr); + +#define FAIL_ACTION(name, fault_attr) \ +bool should_fail_##name(void) \ +{ \ + return should_fail(&fault_attr, 1); \ +} + +FAIL_ACTION(eccerr, fault_eccerr_attr) +FAIL_ACTION(bitflips, fault_bitflips_attr) +FAIL_ACTION(read_failure, fault_read_failure_attr) +FAIL_ACTION(write_failure, fault_write_failure_attr) +FAIL_ACTION(erase_failure, fault_erase_failure_attr) +FAIL_ACTION(power_cut, fault_power_cut_attr) +FAIL_ACTION(io_ff, fault_io_ff_attr) +FAIL_ACTION(io_ff_bitflips, fault_io_ff_bitflips_attr) +FAIL_ACTION(bad_hdr, fault_bad_hdr_attr) +FAIL_ACTION(bad_hdr_ebadmsg, fault_bad_hdr_ebadmsg_attr) +#endif + +/** + * ubi_dump_flash - dump a region of flash. + * @ubi: UBI device description object + * @pnum: the physical eraseblock number to dump + * @offset: the starting offset within the physical eraseblock to dump + * @len: the length of the region to dump + */ +void ubi_dump_flash(struct ubi_device *ubi, int pnum, int offset, int len) +{ + int err; + size_t read; + void *buf; + loff_t addr = (loff_t)pnum * ubi->peb_size + offset; + + buf = vmalloc(len); + if (!buf) + return; + err = mtd_read(ubi->mtd, addr, len, &read, buf); + if (err && err != -EUCLEAN) { + ubi_err(ubi, "err %d while reading %d bytes from PEB %d:%d, read %zd bytes", + err, len, pnum, offset, read); + goto out; + } + + ubi_msg(ubi, "dumping %d bytes of data from PEB %d, offset %d", + len, pnum, offset); + print_hex_dump(KERN_DEBUG, "", DUMP_PREFIX_OFFSET, 32, 1, buf, len, 1); +out: + vfree(buf); + return; +} + +/** + * ubi_dump_ec_hdr - dump an erase counter header. + * @ec_hdr: the erase counter header to dump + */ +void ubi_dump_ec_hdr(const struct ubi_ec_hdr *ec_hdr) +{ + pr_err("Erase counter header dump:\n"); + pr_err("\tmagic %#08x\n", be32_to_cpu(ec_hdr->magic)); + pr_err("\tversion %d\n", (int)ec_hdr->version); + pr_err("\tec %llu\n", (long long)be64_to_cpu(ec_hdr->ec)); + pr_err("\tvid_hdr_offset %d\n", be32_to_cpu(ec_hdr->vid_hdr_offset)); + pr_err("\tdata_offset %d\n", be32_to_cpu(ec_hdr->data_offset)); + pr_err("\timage_seq %d\n", be32_to_cpu(ec_hdr->image_seq)); + pr_err("\thdr_crc %#08x\n", be32_to_cpu(ec_hdr->hdr_crc)); + pr_err("erase counter header hexdump:\n"); + print_hex_dump(KERN_DEBUG, "", DUMP_PREFIX_OFFSET, 32, 1, + ec_hdr, UBI_EC_HDR_SIZE, 1); +} + +/** + * ubi_dump_vid_hdr - dump a volume identifier header. + * @vid_hdr: the volume identifier header to dump + */ +void ubi_dump_vid_hdr(const struct ubi_vid_hdr *vid_hdr) +{ + pr_err("Volume identifier header dump:\n"); + pr_err("\tmagic %08x\n", be32_to_cpu(vid_hdr->magic)); + pr_err("\tversion %d\n", (int)vid_hdr->version); + pr_err("\tvol_type %d\n", (int)vid_hdr->vol_type); + pr_err("\tcopy_flag %d\n", (int)vid_hdr->copy_flag); + pr_err("\tcompat %d\n", (int)vid_hdr->compat); + pr_err("\tvol_id %d\n", be32_to_cpu(vid_hdr->vol_id)); + pr_err("\tlnum %d\n", be32_to_cpu(vid_hdr->lnum)); + pr_err("\tdata_size %d\n", be32_to_cpu(vid_hdr->data_size)); + pr_err("\tused_ebs %d\n", be32_to_cpu(vid_hdr->used_ebs)); + pr_err("\tdata_pad %d\n", be32_to_cpu(vid_hdr->data_pad)); + pr_err("\tsqnum %llu\n", + (unsigned long long)be64_to_cpu(vid_hdr->sqnum)); + pr_err("\thdr_crc %08x\n", be32_to_cpu(vid_hdr->hdr_crc)); + pr_err("Volume identifier header hexdump:\n"); + print_hex_dump(KERN_DEBUG, "", DUMP_PREFIX_OFFSET, 32, 1, + vid_hdr, UBI_VID_HDR_SIZE, 1); +} + +/** + * ubi_dump_vol_info - dump volume information. + * @vol: UBI volume description object + */ +void ubi_dump_vol_info(const struct ubi_volume *vol) +{ + pr_err("Volume information dump:\n"); + pr_err("\tvol_id %d\n", vol->vol_id); + pr_err("\treserved_pebs %d\n", vol->reserved_pebs); + pr_err("\talignment %d\n", vol->alignment); + pr_err("\tdata_pad %d\n", vol->data_pad); + pr_err("\tvol_type %d\n", vol->vol_type); + pr_err("\tname_len %d\n", vol->name_len); + pr_err("\tusable_leb_size %d\n", vol->usable_leb_size); + pr_err("\tused_ebs %d\n", vol->used_ebs); + pr_err("\tused_bytes %lld\n", vol->used_bytes); + pr_err("\tlast_eb_bytes %d\n", vol->last_eb_bytes); + pr_err("\tcorrupted %d\n", vol->corrupted); + pr_err("\tupd_marker %d\n", vol->upd_marker); + pr_err("\tskip_check %d\n", vol->skip_check); + + if (vol->name_len <= UBI_VOL_NAME_MAX && + strnlen(vol->name, vol->name_len + 1) == vol->name_len) { + pr_err("\tname %s\n", vol->name); + } else { + pr_err("\t1st 5 characters of name: %c%c%c%c%c\n", + vol->name[0], vol->name[1], vol->name[2], + vol->name[3], vol->name[4]); + } +} + +/** + * ubi_dump_vtbl_record - dump a &struct ubi_vtbl_record object. + * @r: the object to dump + * @idx: volume table index + */ +void ubi_dump_vtbl_record(const struct ubi_vtbl_record *r, int idx) +{ + int name_len = be16_to_cpu(r->name_len); + + pr_err("Volume table record %d dump:\n", idx); + pr_err("\treserved_pebs %d\n", be32_to_cpu(r->reserved_pebs)); + pr_err("\talignment %d\n", be32_to_cpu(r->alignment)); + pr_err("\tdata_pad %d\n", be32_to_cpu(r->data_pad)); + pr_err("\tvol_type %d\n", (int)r->vol_type); + pr_err("\tupd_marker %d\n", (int)r->upd_marker); + pr_err("\tname_len %d\n", name_len); + + if (r->name[0] == '\0') { + pr_err("\tname NULL\n"); + return; + } + + if (name_len <= UBI_VOL_NAME_MAX && + strnlen(&r->name[0], name_len + 1) == name_len) { + pr_err("\tname %s\n", &r->name[0]); + } else { + pr_err("\t1st 5 characters of name: %c%c%c%c%c\n", + r->name[0], r->name[1], r->name[2], r->name[3], + r->name[4]); + } + pr_err("\tcrc %#08x\n", be32_to_cpu(r->crc)); +} + +/** + * ubi_dump_av - dump a &struct ubi_ainf_volume object. + * @av: the object to dump + */ +void ubi_dump_av(const struct ubi_ainf_volume *av) +{ + pr_err("Volume attaching information dump:\n"); + pr_err("\tvol_id %d\n", av->vol_id); + pr_err("\thighest_lnum %d\n", av->highest_lnum); + pr_err("\tleb_count %d\n", av->leb_count); + pr_err("\tcompat %d\n", av->compat); + pr_err("\tvol_type %d\n", av->vol_type); + pr_err("\tused_ebs %d\n", av->used_ebs); + pr_err("\tlast_data_size %d\n", av->last_data_size); + pr_err("\tdata_pad %d\n", av->data_pad); +} + +/** + * ubi_dump_aeb - dump a &struct ubi_ainf_peb object. + * @aeb: the object to dump + * @type: object type: 0 - not corrupted, 1 - corrupted + */ +void ubi_dump_aeb(const struct ubi_ainf_peb *aeb, int type) +{ + pr_err("eraseblock attaching information dump:\n"); + pr_err("\tec %d\n", aeb->ec); + pr_err("\tpnum %d\n", aeb->pnum); + if (type == 0) { + pr_err("\tlnum %d\n", aeb->lnum); + pr_err("\tscrub %d\n", aeb->scrub); + pr_err("\tsqnum %llu\n", aeb->sqnum); + } +} + +/** + * ubi_dump_mkvol_req - dump a &struct ubi_mkvol_req object. + * @req: the object to dump + */ +void ubi_dump_mkvol_req(const struct ubi_mkvol_req *req) +{ + char nm[17]; + + pr_err("Volume creation request dump:\n"); + pr_err("\tvol_id %d\n", req->vol_id); + pr_err("\talignment %d\n", req->alignment); + pr_err("\tbytes %lld\n", (long long)req->bytes); + pr_err("\tvol_type %d\n", req->vol_type); + pr_err("\tname_len %d\n", req->name_len); + + memcpy(nm, req->name, 16); + nm[16] = 0; + pr_err("\t1st 16 characters of name: %s\n", nm); +} + +/* + * Root directory for UBI stuff in debugfs. Contains sub-directories which + * contain the stuff specific to particular UBI devices. + */ +static struct dentry *dfs_rootdir; + +#ifdef CONFIG_MTD_UBI_FAULT_INJECTION +static void dfs_create_fault_entry(struct dentry *parent) +{ + struct dentry *dir; + + dir = debugfs_create_dir("fault_inject", parent); + if (IS_ERR_OR_NULL(dir)) { + int err = dir ? PTR_ERR(dir) : -ENODEV; + + pr_warn("UBI error: cannot create \"fault_inject\" debugfs directory, error %d\n", + err); + return; + } + + fault_create_debugfs_attr("emulate_eccerr", dir, + &fault_eccerr_attr); + + fault_create_debugfs_attr("emulate_read_failure", dir, + &fault_read_failure_attr); + + fault_create_debugfs_attr("emulate_bitflips", dir, + &fault_bitflips_attr); + + fault_create_debugfs_attr("emulate_write_failure", dir, + &fault_write_failure_attr); + + fault_create_debugfs_attr("emulate_erase_failure", dir, + &fault_erase_failure_attr); + + fault_create_debugfs_attr("emulate_power_cut", dir, + &fault_power_cut_attr); + + fault_create_debugfs_attr("emulate_io_ff", dir, + &fault_io_ff_attr); + + fault_create_debugfs_attr("emulate_io_ff_bitflips", dir, + &fault_io_ff_bitflips_attr); + + fault_create_debugfs_attr("emulate_bad_hdr", dir, + &fault_bad_hdr_attr); + + fault_create_debugfs_attr("emulate_bad_hdr_ebadmsg", dir, + &fault_bad_hdr_ebadmsg_attr); +} +#endif + +/** + * ubi_debugfs_init - create UBI debugfs directory. + * + * Create UBI debugfs directory. Returns zero in case of success and a negative + * error code in case of failure. + */ +int ubi_debugfs_init(void) +{ + if (!IS_ENABLED(CONFIG_DEBUG_FS)) + return 0; + + dfs_rootdir = debugfs_create_dir("ubi", NULL); + if (IS_ERR_OR_NULL(dfs_rootdir)) { + int err = dfs_rootdir ? PTR_ERR(dfs_rootdir) : -ENODEV; + + pr_err("UBI error: cannot create \"ubi\" debugfs directory, error %d\n", + err); + return err; + } + +#ifdef CONFIG_MTD_UBI_FAULT_INJECTION + dfs_create_fault_entry(dfs_rootdir); +#endif + + return 0; +} + +/** + * ubi_debugfs_exit - remove UBI debugfs directory. + */ +void ubi_debugfs_exit(void) +{ + if (IS_ENABLED(CONFIG_DEBUG_FS)) + debugfs_remove(dfs_rootdir); +} + +/* Read an UBI debugfs file */ +static ssize_t dfs_file_read(struct file *file, char __user *user_buf, + size_t count, loff_t *ppos) +{ + unsigned long ubi_num = (unsigned long)file->private_data; + struct dentry *dent = file->f_path.dentry; + struct ubi_device *ubi; + struct ubi_debug_info *d; + char buf[16]; + int val; + + ubi = ubi_get_device(ubi_num); + if (!ubi) + return -ENODEV; + d = &ubi->dbg; + + if (dent == d->dfs_chk_gen) + val = d->chk_gen; + else if (dent == d->dfs_chk_io) + val = d->chk_io; + else if (dent == d->dfs_chk_fastmap) + val = d->chk_fastmap; + else if (dent == d->dfs_disable_bgt) + val = d->disable_bgt; + else if (dent == d->dfs_emulate_bitflips) + val = d->emulate_bitflips; + else if (dent == d->dfs_emulate_io_failures) + val = d->emulate_io_failures; + else if (dent == d->dfs_emulate_failures) { + snprintf(buf, sizeof(buf), "0x%04x\n", d->emulate_failures); + count = simple_read_from_buffer(user_buf, count, ppos, + buf, strlen(buf)); + goto out; + } else if (dent == d->dfs_emulate_power_cut) { + snprintf(buf, sizeof(buf), "%u\n", d->emulate_power_cut); + count = simple_read_from_buffer(user_buf, count, ppos, + buf, strlen(buf)); + goto out; + } else if (dent == d->dfs_power_cut_min) { + snprintf(buf, sizeof(buf), "%u\n", d->power_cut_min); + count = simple_read_from_buffer(user_buf, count, ppos, + buf, strlen(buf)); + goto out; + } else if (dent == d->dfs_power_cut_max) { + snprintf(buf, sizeof(buf), "%u\n", d->power_cut_max); + count = simple_read_from_buffer(user_buf, count, ppos, + buf, strlen(buf)); + goto out; + } else { + count = -EINVAL; + goto out; + } + + if (val) + buf[0] = '1'; + else + buf[0] = '0'; + buf[1] = '\n'; + buf[2] = 0x00; + + count = simple_read_from_buffer(user_buf, count, ppos, buf, 2); + +out: + ubi_put_device(ubi); + return count; +} + +/* Write an UBI debugfs file */ +static ssize_t dfs_file_write(struct file *file, const char __user *user_buf, + size_t count, loff_t *ppos) +{ + unsigned long ubi_num = (unsigned long)file->private_data; + struct dentry *dent = file->f_path.dentry; + struct ubi_device *ubi; + struct ubi_debug_info *d; + size_t buf_size; + char buf[16] = {0}; + int val; + + ubi = ubi_get_device(ubi_num); + if (!ubi) + return -ENODEV; + d = &ubi->dbg; + + buf_size = min_t(size_t, count, (sizeof(buf) - 1)); + if (copy_from_user(buf, user_buf, buf_size)) { + count = -EFAULT; + goto out; + } + + if (dent == d->dfs_emulate_failures) { + if (kstrtouint(buf, 0, &d->emulate_failures) != 0) + count = -EINVAL; + goto out; + } else if (dent == d->dfs_power_cut_min) { + if (kstrtouint(buf, 0, &d->power_cut_min) != 0) + count = -EINVAL; + goto out; + } else if (dent == d->dfs_power_cut_max) { + if (kstrtouint(buf, 0, &d->power_cut_max) != 0) + count = -EINVAL; + goto out; + } else if (dent == d->dfs_emulate_power_cut) { + if (kstrtoint(buf, 0, &val) != 0) + count = -EINVAL; + else + d->emulate_power_cut = val; + goto out; + } + + if (buf[0] == '1') + val = 1; + else if (buf[0] == '0') + val = 0; + else { + count = -EINVAL; + goto out; + } + + if (dent == d->dfs_chk_gen) + d->chk_gen = val; + else if (dent == d->dfs_chk_io) + d->chk_io = val; + else if (dent == d->dfs_chk_fastmap) + d->chk_fastmap = val; + else if (dent == d->dfs_disable_bgt) + d->disable_bgt = val; + else if (dent == d->dfs_emulate_bitflips) + d->emulate_bitflips = val; + else if (dent == d->dfs_emulate_io_failures) + d->emulate_io_failures = val; + else + count = -EINVAL; + +out: + ubi_put_device(ubi); + return count; +} + +/* File operations for all UBI debugfs files except + * detailed_erase_block_info + */ +static const struct file_operations dfs_fops = { + .read = dfs_file_read, + .write = dfs_file_write, + .open = simple_open, + .owner = THIS_MODULE, +}; + +/* As long as the position is less then that total number of erase blocks, + * we still have more to print. + */ +static void *eraseblk_count_seq_start(struct seq_file *s, loff_t *pos) +{ + struct ubi_device *ubi = s->private; + + if (*pos < ubi->peb_count) + return pos; + + return NULL; +} + +/* Since we are using the position as the iterator, we just need to check if we + * are done and increment the position. + */ +static void *eraseblk_count_seq_next(struct seq_file *s, void *v, loff_t *pos) +{ + struct ubi_device *ubi = s->private; + + (*pos)++; + + if (*pos < ubi->peb_count) + return pos; + + return NULL; +} + +static void eraseblk_count_seq_stop(struct seq_file *s, void *v) +{ +} + +static int eraseblk_count_seq_show(struct seq_file *s, void *iter) +{ + struct ubi_device *ubi = s->private; + struct ubi_wl_entry *wl; + int *block_number = iter; + int erase_count = -1; + int err; + + /* If this is the start, print a header */ + if (*block_number == 0) + seq_puts(s, "physical_block_number\terase_count\n"); + + err = ubi_io_is_bad(ubi, *block_number); + if (err) + return err; + + spin_lock(&ubi->wl_lock); + + wl = ubi->lookuptbl[*block_number]; + if (wl) + erase_count = wl->ec; + + spin_unlock(&ubi->wl_lock); + + if (erase_count < 0) + return 0; + + seq_printf(s, "%-22d\t%-11d\n", *block_number, erase_count); + + return 0; +} + +static const struct seq_operations eraseblk_count_seq_ops = { + .start = eraseblk_count_seq_start, + .next = eraseblk_count_seq_next, + .stop = eraseblk_count_seq_stop, + .show = eraseblk_count_seq_show +}; + +static int eraseblk_count_open(struct inode *inode, struct file *f) +{ + struct seq_file *s; + int err; + + err = seq_open(f, &eraseblk_count_seq_ops); + if (err) + return err; + + s = f->private_data; + s->private = ubi_get_device((unsigned long)inode->i_private); + + if (!s->private) + return -ENODEV; + else + return 0; +} + +static int eraseblk_count_release(struct inode *inode, struct file *f) +{ + struct seq_file *s = f->private_data; + struct ubi_device *ubi = s->private; + + ubi_put_device(ubi); + + return seq_release(inode, f); +} + +static const struct file_operations eraseblk_count_fops = { + .owner = THIS_MODULE, + .open = eraseblk_count_open, + .read = seq_read, + .llseek = seq_lseek, + .release = eraseblk_count_release, +}; + +/** + * ubi_debugfs_init_dev - initialize debugfs for an UBI device. + * @ubi: UBI device description object + * + * This function creates all debugfs files for UBI device @ubi. Returns zero in + * case of success and a negative error code in case of failure. + */ +int ubi_debugfs_init_dev(struct ubi_device *ubi) +{ + unsigned long ubi_num = ubi->ubi_num; + struct ubi_debug_info *d = &ubi->dbg; + umode_t mode = S_IRUSR | S_IWUSR; + int n; + + if (!IS_ENABLED(CONFIG_DEBUG_FS)) + return 0; + + n = snprintf(d->dfs_dir_name, UBI_DFS_DIR_LEN, UBI_DFS_DIR_NAME, + ubi->ubi_num); + if (n >= UBI_DFS_DIR_LEN) { + /* The array size is too small */ + return -EINVAL; + } + + d->dfs_dir = debugfs_create_dir(d->dfs_dir_name, dfs_rootdir); + + d->dfs_chk_gen = debugfs_create_file("chk_gen", mode, d->dfs_dir, + (void *)ubi_num, &dfs_fops); + + d->dfs_chk_io = debugfs_create_file("chk_io", mode, d->dfs_dir, + (void *)ubi_num, &dfs_fops); + + d->dfs_chk_fastmap = debugfs_create_file("chk_fastmap", mode, + d->dfs_dir, (void *)ubi_num, + &dfs_fops); + + d->dfs_disable_bgt = debugfs_create_file("tst_disable_bgt", mode, + d->dfs_dir, (void *)ubi_num, + &dfs_fops); + + d->dfs_emulate_bitflips = debugfs_create_file("tst_emulate_bitflips", + mode, d->dfs_dir, + (void *)ubi_num, + &dfs_fops); + + d->dfs_emulate_io_failures = debugfs_create_file("tst_emulate_io_failures", + mode, d->dfs_dir, + (void *)ubi_num, + &dfs_fops); + + d->dfs_emulate_power_cut = debugfs_create_file("tst_emulate_power_cut", + mode, d->dfs_dir, + (void *)ubi_num, + &dfs_fops); + + d->dfs_power_cut_min = debugfs_create_file("tst_emulate_power_cut_min", + mode, d->dfs_dir, + (void *)ubi_num, &dfs_fops); + + d->dfs_power_cut_max = debugfs_create_file("tst_emulate_power_cut_max", + mode, d->dfs_dir, + (void *)ubi_num, &dfs_fops); + + debugfs_create_file("detailed_erase_block_info", S_IRUSR, d->dfs_dir, + (void *)ubi_num, &eraseblk_count_fops); + +#ifdef CONFIG_MTD_UBI_FAULT_INJECTION + d->dfs_emulate_failures = debugfs_create_file("emulate_failures", + mode, d->dfs_dir, + (void *)ubi_num, + &dfs_fops); +#endif + return 0; +} + +/** + * ubi_debugfs_exit_dev - free all debugfs files corresponding to device @ubi + * @ubi: UBI device description object + */ +void ubi_debugfs_exit_dev(struct ubi_device *ubi) +{ + if (IS_ENABLED(CONFIG_DEBUG_FS)) + debugfs_remove_recursive(ubi->dbg.dfs_dir); +} + +/** + * ubi_dbg_power_cut - emulate a power cut if it is time to do so + * @ubi: UBI device description object + * @caller: Flags set to indicate from where the function is being called + * + * Returns non-zero if a power cut was emulated, zero if not. + */ +int ubi_dbg_power_cut(struct ubi_device *ubi, int caller) +{ + unsigned int range; + + if ((ubi->dbg.emulate_power_cut & caller) == 0) + return 0; + + if (ubi->dbg.power_cut_counter == 0) { + ubi->dbg.power_cut_counter = ubi->dbg.power_cut_min; + + if (ubi->dbg.power_cut_max > ubi->dbg.power_cut_min) { + range = ubi->dbg.power_cut_max - ubi->dbg.power_cut_min; + ubi->dbg.power_cut_counter += get_random_u32_below(range); + } + return 0; + } + + ubi->dbg.power_cut_counter--; + if (ubi->dbg.power_cut_counter) + return 0; + + return 1; +} diff --git a/drivers/mtd/ubi/debug.h b/drivers/mtd/ubi/debug.h new file mode 100644 index 000000000..b2fd97548 --- /dev/null +++ b/drivers/mtd/ubi/debug.h @@ -0,0 +1,387 @@ +/* SPDX-License-Identifier: GPL-2.0-or-later */ +/* + * Copyright (c) International Business Machines Corp., 2006 + * + * Author: Artem Bityutskiy (Битюцкий Артём) + */ + +#ifndef __UBI_DEBUG_H__ +#define __UBI_DEBUG_H__ + +void ubi_dump_flash(struct ubi_device *ubi, int pnum, int offset, int len); +void ubi_dump_ec_hdr(const struct ubi_ec_hdr *ec_hdr); +void ubi_dump_vid_hdr(const struct ubi_vid_hdr *vid_hdr); + +#include + +#define ubi_assert(expr) do { \ + if (unlikely(!(expr))) { \ + pr_crit("UBI assert failed in %s at %u (pid %d)\n", \ + __func__, __LINE__, current->pid); \ + dump_stack(); \ + } \ +} while (0) + +#define ubi_dbg_print_hex_dump(l, ps, pt, r, g, b, len, a) \ + print_hex_dump(l, ps, pt, r, g, b, len, a) + +#define ubi_dbg_msg(type, fmt, ...) \ + pr_debug("UBI DBG " type " (pid %d): " fmt "\n", current->pid, \ + ##__VA_ARGS__) + +/* General debugging messages */ +#define dbg_gen(fmt, ...) ubi_dbg_msg("gen", fmt, ##__VA_ARGS__) +/* Messages from the eraseblock association sub-system */ +#define dbg_eba(fmt, ...) ubi_dbg_msg("eba", fmt, ##__VA_ARGS__) +/* Messages from the wear-leveling sub-system */ +#define dbg_wl(fmt, ...) ubi_dbg_msg("wl", fmt, ##__VA_ARGS__) +/* Messages from the input/output sub-system */ +#define dbg_io(fmt, ...) ubi_dbg_msg("io", fmt, ##__VA_ARGS__) +/* Initialization and build messages */ +#define dbg_bld(fmt, ...) ubi_dbg_msg("bld", fmt, ##__VA_ARGS__) + +void ubi_dump_vol_info(const struct ubi_volume *vol); +void ubi_dump_vtbl_record(const struct ubi_vtbl_record *r, int idx); +void ubi_dump_av(const struct ubi_ainf_volume *av); +void ubi_dump_aeb(const struct ubi_ainf_peb *aeb, int type); +void ubi_dump_mkvol_req(const struct ubi_mkvol_req *req); +int ubi_self_check_all_ff(struct ubi_device *ubi, int pnum, int offset, + int len); +int ubi_debugfs_init(void); +void ubi_debugfs_exit(void); +int ubi_debugfs_init_dev(struct ubi_device *ubi); +void ubi_debugfs_exit_dev(struct ubi_device *ubi); + +/** + * The following function is a legacy implementation of UBI fault-injection + * hook. When using more powerful fault injection capabilities, the legacy + * fault injection interface should be retained. + */ +int ubi_dbg_power_cut(struct ubi_device *ubi, int caller); + +static inline int ubi_dbg_bitflip(const struct ubi_device *ubi) +{ + if (ubi->dbg.emulate_bitflips) + return !get_random_u32_below(200); + return 0; +} + +static inline int ubi_dbg_write_failure(const struct ubi_device *ubi) +{ + if (ubi->dbg.emulate_io_failures) + return !get_random_u32_below(500); + return 0; +} + +static inline int ubi_dbg_erase_failure(const struct ubi_device *ubi) +{ + if (ubi->dbg.emulate_io_failures) + return !get_random_u32_below(400); + return 0; +} + +/** + * MASK_XXX: Mask for emulate_failures in ubi_debug_info.The mask is used to + * precisely control the type and process of fault injection. + */ +/* Emulate a power cut when writing EC/VID header */ +#define MASK_POWER_CUT_EC (1 << 0) +#define MASK_POWER_CUT_VID (1 << 1) +/* Emulate a power cut when writing data*/ +#define MASK_POWER_CUT_DATA (1 << 2) +/* Emulate bit-flips */ +#define MASK_BITFLIPS (1 << 3) +/* Emulate ecc error */ +#define MASK_ECCERR (1 << 4) +/* Emulates -EIO during data read */ +#define MASK_READ_FAILURE (1 << 5) +#define MASK_READ_FAILURE_EC (1 << 6) +#define MASK_READ_FAILURE_VID (1 << 7) +/* Emulates -EIO during data write */ +#define MASK_WRITE_FAILURE (1 << 8) +/* Emulates -EIO during erase a PEB*/ +#define MASK_ERASE_FAILURE (1 << 9) +/* Return UBI_IO_FF when reading EC/VID header */ +#define MASK_IO_FF_EC (1 << 10) +#define MASK_IO_FF_VID (1 << 11) +/* Return UBI_IO_FF_BITFLIPS when reading EC/VID header */ +#define MASK_IO_FF_BITFLIPS_EC (1 << 12) +#define MASK_IO_FF_BITFLIPS_VID (1 << 13) +/* Return UBI_IO_BAD_HDR when reading EC/VID header */ +#define MASK_BAD_HDR_EC (1 << 14) +#define MASK_BAD_HDR_VID (1 << 15) +/* Return UBI_IO_BAD_HDR_EBADMSG when reading EC/VID header */ +#define MASK_BAD_HDR_EBADMSG_EC (1 << 16) +#define MASK_BAD_HDR_EBADMSG_VID (1 << 17) + +#ifdef CONFIG_MTD_UBI_FAULT_INJECTION + +extern bool should_fail_eccerr(void); +extern bool should_fail_bitflips(void); +extern bool should_fail_read_failure(void); +extern bool should_fail_write_failure(void); +extern bool should_fail_erase_failure(void); +extern bool should_fail_power_cut(void); +extern bool should_fail_io_ff(void); +extern bool should_fail_io_ff_bitflips(void); +extern bool should_fail_bad_hdr(void); +extern bool should_fail_bad_hdr_ebadmsg(void); + +static inline bool ubi_dbg_fail_bitflip(const struct ubi_device *ubi) +{ + if (ubi->dbg.emulate_failures & MASK_BITFLIPS) + return should_fail_bitflips(); + return false; +} + +static inline bool ubi_dbg_fail_write(const struct ubi_device *ubi) +{ + if (ubi->dbg.emulate_failures & MASK_WRITE_FAILURE) + return should_fail_write_failure(); + return false; +} + +static inline bool ubi_dbg_fail_erase(const struct ubi_device *ubi) +{ + if (ubi->dbg.emulate_failures & MASK_ERASE_FAILURE) + return should_fail_erase_failure(); + return false; +} + +static inline bool ubi_dbg_fail_power_cut(const struct ubi_device *ubi, + unsigned int caller) +{ + if (ubi->dbg.emulate_failures & caller) + return should_fail_power_cut(); + return false; +} + +static inline bool ubi_dbg_fail_read(const struct ubi_device *ubi, + unsigned int caller) +{ + if (ubi->dbg.emulate_failures & caller) + return should_fail_read_failure(); + return false; +} + +static inline bool ubi_dbg_fail_eccerr(const struct ubi_device *ubi) +{ + if (ubi->dbg.emulate_failures & MASK_ECCERR) + return should_fail_eccerr(); + return false; +} + +static inline bool ubi_dbg_fail_ff(const struct ubi_device *ubi, + unsigned int caller) +{ + if (ubi->dbg.emulate_failures & caller) + return should_fail_io_ff(); + return false; +} + +static inline bool ubi_dbg_fail_ff_bitflips(const struct ubi_device *ubi, + unsigned int caller) +{ + if (ubi->dbg.emulate_failures & caller) + return should_fail_io_ff_bitflips(); + return false; +} + +static inline bool ubi_dbg_fail_bad_hdr(const struct ubi_device *ubi, + unsigned int caller) +{ + if (ubi->dbg.emulate_failures & caller) + return should_fail_bad_hdr(); + return false; +} + +static inline bool ubi_dbg_fail_bad_hdr_ebadmsg(const struct ubi_device *ubi, + unsigned int caller) +{ + if (ubi->dbg.emulate_failures & caller) + return should_fail_bad_hdr_ebadmsg(); + return false; +} +#else /* CONFIG_MTD_UBI_FAULT_INJECTION */ + +#define ubi_dbg_fail_bitflip(u) false +#define ubi_dbg_fail_write(u) false +#define ubi_dbg_fail_erase(u) false +#define ubi_dbg_fail_power_cut(u, c) false +#define ubi_dbg_fail_read(u, c) false +#define ubi_dbg_fail_eccerr(u) false +#define ubi_dbg_fail_ff(u, c) false +#define ubi_dbg_fail_ff_bitflips(u, v) false +#define ubi_dbg_fail_bad_hdr(u, c) false +#define ubi_dbg_fail_bad_hdr_ebadmsg(u, c) false + +#endif + +/** + * ubi_dbg_is_power_cut - if it is time to emulate power cut. + * @ubi: UBI device description object + * + * Returns true if power cut should be emulated, otherwise returns false. + */ +static inline bool ubi_dbg_is_power_cut(struct ubi_device *ubi, + unsigned int caller) +{ + if (ubi_dbg_power_cut(ubi, caller)) + return true; + return ubi_dbg_fail_power_cut(ubi, caller); +} + +/** + * ubi_dbg_is_bitflip - if it is time to emulate a bit-flip. + * @ubi: UBI device description object + * + * Returns true if a bit-flip should be emulated, otherwise returns false. + */ +static inline bool ubi_dbg_is_bitflip(const struct ubi_device *ubi) +{ + if (ubi_dbg_bitflip(ubi)) + return true; + return ubi_dbg_fail_bitflip(ubi); +} + +/** + * ubi_dbg_is_write_failure - if it is time to emulate a write failure. + * @ubi: UBI device description object + * + * Returns true if a write failure should be emulated, otherwise returns + * false. + */ +static inline bool ubi_dbg_is_write_failure(const struct ubi_device *ubi) +{ + if (ubi_dbg_write_failure(ubi)) + return true; + return ubi_dbg_fail_write(ubi); +} + +/** + * ubi_dbg_is_erase_failure - if its time to emulate an erase failure. + * @ubi: UBI device description object + * + * Returns true if an erase failure should be emulated, otherwise returns + * false. + */ +static inline bool ubi_dbg_is_erase_failure(const struct ubi_device *ubi) +{ + if (ubi_dbg_erase_failure(ubi)) + return true; + return ubi_dbg_fail_erase(ubi); +} + +/** + * ubi_dbg_is_eccerr - if it is time to emulate ECC error. + * @ubi: UBI device description object + * + * Returns true if a ECC error should be emulated, otherwise returns false. + */ +static inline bool ubi_dbg_is_eccerr(const struct ubi_device *ubi) +{ + return ubi_dbg_fail_eccerr(ubi); +} + +/** + * ubi_dbg_is_read_failure - if it is time to emulate a read failure. + * @ubi: UBI device description object + * + * Returns true if a read failure should be emulated, otherwise returns + * false. + */ +static inline bool ubi_dbg_is_read_failure(const struct ubi_device *ubi, + unsigned int caller) +{ + return ubi_dbg_fail_read(ubi, caller); +} + +/** + * ubi_dbg_is_ff - if it is time to emulate that read region is only 0xFF. + * @ubi: UBI device description object + * + * Returns true if read region should be emulated 0xFF, otherwise + * returns false. + */ +static inline bool ubi_dbg_is_ff(const struct ubi_device *ubi, + unsigned int caller) +{ + return ubi_dbg_fail_ff(ubi, caller); +} + +/** + * ubi_dbg_is_ff_bitflips - if it is time to emulate that read region is only 0xFF + * with error reported by the MTD driver + * + * @ubi: UBI device description object + * + * Returns true if read region should be emulated 0xFF and error + * reported by the MTD driver, otherwise returns false. + */ +static inline bool ubi_dbg_is_ff_bitflips(const struct ubi_device *ubi, + unsigned int caller) +{ + return ubi_dbg_fail_ff_bitflips(ubi, caller); +} + +/** + * ubi_dbg_is_bad_hdr - if it is time to emulate a bad header + * @ubi: UBI device description object + * + * Returns true if a bad header error should be emulated, otherwise + * returns false. + */ +static inline bool ubi_dbg_is_bad_hdr(const struct ubi_device *ubi, + unsigned int caller) +{ + return ubi_dbg_fail_bad_hdr(ubi, caller); +} + +/** + * ubi_dbg_is_bad_hdr_ebadmsg - if it is time to emulate a bad header with + * ECC error. + * + * @ubi: UBI device description object + * + * Returns true if a bad header with ECC error should be emulated, otherwise + * returns false. + */ +static inline bool ubi_dbg_is_bad_hdr_ebadmsg(const struct ubi_device *ubi, + unsigned int caller) +{ + return ubi_dbg_fail_bad_hdr_ebadmsg(ubi, caller); +} + +/** + * ubi_dbg_is_bgt_disabled - if the background thread is disabled. + * @ubi: UBI device description object + * + * Returns non-zero if the UBI background thread is disabled for testing + * purposes. + */ +static inline int ubi_dbg_is_bgt_disabled(const struct ubi_device *ubi) +{ + return ubi->dbg.disable_bgt; +} + +static inline int ubi_dbg_chk_io(const struct ubi_device *ubi) +{ + return ubi->dbg.chk_io; +} + +static inline int ubi_dbg_chk_gen(const struct ubi_device *ubi) +{ + return ubi->dbg.chk_gen; +} + +static inline int ubi_dbg_chk_fastmap(const struct ubi_device *ubi) +{ + return ubi->dbg.chk_fastmap; +} + +static inline void ubi_enable_dbg_chk_fastmap(struct ubi_device *ubi) +{ + ubi->dbg.chk_fastmap = 1; +} + +#endif /* !__UBI_DEBUG_H__ */ diff --git a/drivers/mtd/ubi/eba.c b/drivers/mtd/ubi/eba.c new file mode 100644 index 000000000..f0c549e62 --- /dev/null +++ b/drivers/mtd/ubi/eba.c @@ -0,0 +1,1709 @@ +// SPDX-License-Identifier: GPL-2.0-or-later +/* + * Copyright (c) International Business Machines Corp., 2006 + * + * Author: Artem Bityutskiy (Битюцкий Артём) + */ + +/* + * The UBI Eraseblock Association (EBA) sub-system. + * + * This sub-system is responsible for I/O to/from logical eraseblock. + * + * Although in this implementation the EBA table is fully kept and managed in + * RAM, which assumes poor scalability, it might be (partially) maintained on + * flash in future implementations. + * + * The EBA sub-system implements per-logical eraseblock locking. Before + * accessing a logical eraseblock it is locked for reading or writing. The + * per-logical eraseblock locking is implemented by means of the lock tree. The + * lock tree is an RB-tree which refers all the currently locked logical + * eraseblocks. The lock tree elements are &struct ubi_ltree_entry objects. + * They are indexed by (@vol_id, @lnum) pairs. + * + * EBA also maintains the global sequence counter which is incremented each + * time a logical eraseblock is mapped to a physical eraseblock and it is + * stored in the volume identifier header. This means that each VID header has + * a unique sequence number. The sequence number is only increased an we assume + * 64 bits is enough to never overflow. + */ + +#include +#include +#include +#include "ubi.h" + +/** + * struct ubi_eba_entry - structure encoding a single LEB -> PEB association + * @pnum: the physical eraseblock number attached to the LEB + * + * This structure is encoding a LEB -> PEB association. Note that the LEB + * number is not stored here, because it is the index used to access the + * entries table. + */ +struct ubi_eba_entry { + int pnum; +}; + +/** + * struct ubi_eba_table - LEB -> PEB association information + * @entries: the LEB to PEB mapping (one entry per LEB). + * + * This structure is private to the EBA logic and should be kept here. + * It is encoding the LEB to PEB association table, and is subject to + * changes. + */ +struct ubi_eba_table { + struct ubi_eba_entry *entries; +}; + +/** + * ubi_next_sqnum - get next sequence number. + * @ubi: UBI device description object + * + * This function returns next sequence number to use, which is just the current + * global sequence counter value. It also increases the global sequence + * counter. + */ +unsigned long long ubi_next_sqnum(struct ubi_device *ubi) +{ + unsigned long long sqnum; + + spin_lock(&ubi->ltree_lock); + sqnum = ubi->global_sqnum++; + spin_unlock(&ubi->ltree_lock); + + return sqnum; +} + +/** + * ubi_get_compat - get compatibility flags of a volume. + * @ubi: UBI device description object + * @vol_id: volume ID + * + * This function returns compatibility flags for an internal volume. User + * volumes have no compatibility flags, so %0 is returned. + */ +static int ubi_get_compat(const struct ubi_device *ubi, int vol_id) +{ + if (vol_id == UBI_LAYOUT_VOLUME_ID) + return UBI_LAYOUT_VOLUME_COMPAT; + return 0; +} + +/** + * ubi_eba_get_ldesc - get information about a LEB + * @vol: volume description object + * @lnum: logical eraseblock number + * @ldesc: the LEB descriptor to fill + * + * Used to query information about a specific LEB. + * It is currently only returning the physical position of the LEB, but will be + * extended to provide more information. + */ +void ubi_eba_get_ldesc(struct ubi_volume *vol, int lnum, + struct ubi_eba_leb_desc *ldesc) +{ + ldesc->lnum = lnum; + ldesc->pnum = vol->eba_tbl->entries[lnum].pnum; +} + +/** + * ubi_eba_create_table - allocate a new EBA table and initialize it with all + * LEBs unmapped + * @vol: volume containing the EBA table to copy + * @nentries: number of entries in the table + * + * Allocate a new EBA table and initialize it with all LEBs unmapped. + * Returns a valid pointer if it succeed, an ERR_PTR() otherwise. + */ +struct ubi_eba_table *ubi_eba_create_table(struct ubi_volume *vol, + int nentries) +{ + struct ubi_eba_table *tbl; + int err = -ENOMEM; + int i; + + tbl = kzalloc_obj(*tbl); + if (!tbl) + return ERR_PTR(-ENOMEM); + + tbl->entries = kmalloc_objs(*tbl->entries, nentries); + if (!tbl->entries) + goto err; + + for (i = 0; i < nentries; i++) + tbl->entries[i].pnum = UBI_LEB_UNMAPPED; + + return tbl; + +err: + kfree(tbl); + + return ERR_PTR(err); +} + +/** + * ubi_eba_destroy_table - destroy an EBA table + * @tbl: the table to destroy + * + * Destroy an EBA table. + */ +void ubi_eba_destroy_table(struct ubi_eba_table *tbl) +{ + if (!tbl) + return; + + kfree(tbl->entries); + kfree(tbl); +} + +/** + * ubi_eba_copy_table - copy the EBA table attached to vol into another table + * @vol: volume containing the EBA table to copy + * @dst: destination + * @nentries: number of entries to copy + * + * Copy the EBA table stored in vol into the one pointed by dst. + */ +void ubi_eba_copy_table(struct ubi_volume *vol, struct ubi_eba_table *dst, + int nentries) +{ + struct ubi_eba_table *src; + int i; + + ubi_assert(dst && vol && vol->eba_tbl); + + src = vol->eba_tbl; + + for (i = 0; i < nentries; i++) + dst->entries[i].pnum = src->entries[i].pnum; +} + +/** + * ubi_eba_replace_table - assign a new EBA table to a volume + * @vol: volume containing the EBA table to copy + * @tbl: new EBA table + * + * Assign a new EBA table to the volume and release the old one. + */ +void ubi_eba_replace_table(struct ubi_volume *vol, struct ubi_eba_table *tbl) +{ + ubi_eba_destroy_table(vol->eba_tbl); + vol->eba_tbl = tbl; +} + +/** + * ltree_lookup - look up the lock tree. + * @ubi: UBI device description object + * @vol_id: volume ID + * @lnum: logical eraseblock number + * + * This function returns a pointer to the corresponding &struct ubi_ltree_entry + * object if the logical eraseblock is locked and %NULL if it is not. + * @ubi->ltree_lock has to be locked. + */ +static struct ubi_ltree_entry *ltree_lookup(struct ubi_device *ubi, int vol_id, + int lnum) +{ + struct rb_node *p; + + p = ubi->ltree.rb_node; + while (p) { + struct ubi_ltree_entry *le; + + le = rb_entry(p, struct ubi_ltree_entry, rb); + + if (vol_id < le->vol_id) + p = p->rb_left; + else if (vol_id > le->vol_id) + p = p->rb_right; + else { + if (lnum < le->lnum) + p = p->rb_left; + else if (lnum > le->lnum) + p = p->rb_right; + else + return le; + } + } + + return NULL; +} + +/** + * ltree_add_entry - add new entry to the lock tree. + * @ubi: UBI device description object + * @vol_id: volume ID + * @lnum: logical eraseblock number + * + * This function adds new entry for logical eraseblock (@vol_id, @lnum) to the + * lock tree. If such entry is already there, its usage counter is increased. + * Returns pointer to the lock tree entry or %-ENOMEM if memory allocation + * failed. + */ +static struct ubi_ltree_entry *ltree_add_entry(struct ubi_device *ubi, + int vol_id, int lnum) +{ + struct ubi_ltree_entry *le, *le1, *le_free; + + le = kmalloc_obj(struct ubi_ltree_entry, GFP_NOFS); + if (!le) + return ERR_PTR(-ENOMEM); + + le->users = 0; + init_rwsem(&le->mutex); + le->vol_id = vol_id; + le->lnum = lnum; + + spin_lock(&ubi->ltree_lock); + le1 = ltree_lookup(ubi, vol_id, lnum); + + if (le1) { + /* + * This logical eraseblock is already locked. The newly + * allocated lock entry is not needed. + */ + le_free = le; + le = le1; + } else { + struct rb_node **p, *parent = NULL; + + /* + * No lock entry, add the newly allocated one to the + * @ubi->ltree RB-tree. + */ + le_free = NULL; + + p = &ubi->ltree.rb_node; + while (*p) { + parent = *p; + le1 = rb_entry(parent, struct ubi_ltree_entry, rb); + + if (vol_id < le1->vol_id) + p = &(*p)->rb_left; + else if (vol_id > le1->vol_id) + p = &(*p)->rb_right; + else { + ubi_assert(lnum != le1->lnum); + if (lnum < le1->lnum) + p = &(*p)->rb_left; + else + p = &(*p)->rb_right; + } + } + + rb_link_node(&le->rb, parent, p); + rb_insert_color(&le->rb, &ubi->ltree); + } + le->users += 1; + spin_unlock(&ubi->ltree_lock); + + kfree(le_free); + return le; +} + +/** + * leb_read_lock - lock logical eraseblock for reading. + * @ubi: UBI device description object + * @vol_id: volume ID + * @lnum: logical eraseblock number + * + * This function locks a logical eraseblock for reading. Returns zero in case + * of success and a negative error code in case of failure. + */ +static int leb_read_lock(struct ubi_device *ubi, int vol_id, int lnum) +{ + struct ubi_ltree_entry *le; + + le = ltree_add_entry(ubi, vol_id, lnum); + if (IS_ERR(le)) + return PTR_ERR(le); + down_read(&le->mutex); + return 0; +} + +/** + * leb_read_unlock - unlock logical eraseblock. + * @ubi: UBI device description object + * @vol_id: volume ID + * @lnum: logical eraseblock number + */ +static void leb_read_unlock(struct ubi_device *ubi, int vol_id, int lnum) +{ + struct ubi_ltree_entry *le; + + spin_lock(&ubi->ltree_lock); + le = ltree_lookup(ubi, vol_id, lnum); + le->users -= 1; + ubi_assert(le->users >= 0); + up_read(&le->mutex); + if (le->users == 0) { + rb_erase(&le->rb, &ubi->ltree); + kfree(le); + } + spin_unlock(&ubi->ltree_lock); +} + +/** + * leb_write_lock - lock logical eraseblock for writing. + * @ubi: UBI device description object + * @vol_id: volume ID + * @lnum: logical eraseblock number + * + * This function locks a logical eraseblock for writing. Returns zero in case + * of success and a negative error code in case of failure. + */ +static int leb_write_lock(struct ubi_device *ubi, int vol_id, int lnum) +{ + struct ubi_ltree_entry *le; + + le = ltree_add_entry(ubi, vol_id, lnum); + if (IS_ERR(le)) + return PTR_ERR(le); + down_write(&le->mutex); + return 0; +} + +/** + * leb_write_trylock - try to lock logical eraseblock for writing. + * @ubi: UBI device description object + * @vol_id: volume ID + * @lnum: logical eraseblock number + * + * This function locks a logical eraseblock for writing if there is no + * contention and does nothing if there is contention. Returns %0 in case of + * success, %1 in case of contention, and a negative error code in case of + * failure. + */ +static int leb_write_trylock(struct ubi_device *ubi, int vol_id, int lnum) +{ + struct ubi_ltree_entry *le; + + le = ltree_add_entry(ubi, vol_id, lnum); + if (IS_ERR(le)) + return PTR_ERR(le); + if (down_write_trylock(&le->mutex)) + return 0; + + /* Contention, cancel */ + spin_lock(&ubi->ltree_lock); + le->users -= 1; + ubi_assert(le->users >= 0); + if (le->users == 0) { + rb_erase(&le->rb, &ubi->ltree); + kfree(le); + } + spin_unlock(&ubi->ltree_lock); + + return 1; +} + +/** + * leb_write_unlock - unlock logical eraseblock. + * @ubi: UBI device description object + * @vol_id: volume ID + * @lnum: logical eraseblock number + */ +static void leb_write_unlock(struct ubi_device *ubi, int vol_id, int lnum) +{ + struct ubi_ltree_entry *le; + + spin_lock(&ubi->ltree_lock); + le = ltree_lookup(ubi, vol_id, lnum); + le->users -= 1; + ubi_assert(le->users >= 0); + up_write(&le->mutex); + if (le->users == 0) { + rb_erase(&le->rb, &ubi->ltree); + kfree(le); + } + spin_unlock(&ubi->ltree_lock); +} + +/** + * ubi_eba_is_mapped - check if a LEB is mapped. + * @vol: volume description object + * @lnum: logical eraseblock number + * + * This function returns true if the LEB is mapped, false otherwise. + */ +bool ubi_eba_is_mapped(struct ubi_volume *vol, int lnum) +{ + return vol->eba_tbl->entries[lnum].pnum >= 0; +} + +/** + * ubi_eba_unmap_leb - un-map logical eraseblock. + * @ubi: UBI device description object + * @vol: volume description object + * @lnum: logical eraseblock number + * + * This function un-maps logical eraseblock @lnum and schedules corresponding + * physical eraseblock for erasure. Returns zero in case of success and a + * negative error code in case of failure. + */ +int ubi_eba_unmap_leb(struct ubi_device *ubi, struct ubi_volume *vol, + int lnum) +{ + int err, pnum, vol_id = vol->vol_id; + + if (ubi->ro_mode) + return -EROFS; + + err = leb_write_lock(ubi, vol_id, lnum); + if (err) + return err; + + pnum = vol->eba_tbl->entries[lnum].pnum; + if (pnum < 0) + /* This logical eraseblock is already unmapped */ + goto out_unlock; + + dbg_eba("erase LEB %d:%d, PEB %d", vol_id, lnum, pnum); + + down_read(&ubi->fm_eba_sem); + vol->eba_tbl->entries[lnum].pnum = UBI_LEB_UNMAPPED; + up_read(&ubi->fm_eba_sem); + err = ubi_wl_put_peb(ubi, vol_id, lnum, pnum, 0); + +out_unlock: + leb_write_unlock(ubi, vol_id, lnum); + return err; +} + +#ifdef CONFIG_MTD_UBI_FASTMAP +/** + * check_mapping - check and fixup a mapping + * @ubi: UBI device description object + * @vol: volume description object + * @lnum: logical eraseblock number + * @pnum: physical eraseblock number + * + * Checks whether a given mapping is valid. Fastmap cannot track LEB unmap + * operations, if such an operation is interrupted the mapping still looks + * good, but upon first read an ECC is reported to the upper layer. + * Normaly during the full-scan at attach time this is fixed, for Fastmap + * we have to deal with it while reading. + * If the PEB behind a LEB shows this symthom we change the mapping to + * %UBI_LEB_UNMAPPED and schedule the PEB for erasure. + * + * Returns 0 on success, negative error code in case of failure. + */ +static int check_mapping(struct ubi_device *ubi, struct ubi_volume *vol, int lnum, + int *pnum) +{ + int err; + struct ubi_vid_io_buf *vidb; + struct ubi_vid_hdr *vid_hdr; + + if (!ubi->fast_attach) + return 0; + + if (!vol->checkmap || test_bit(lnum, vol->checkmap)) + return 0; + + vidb = ubi_alloc_vid_buf(ubi, GFP_NOFS); + if (!vidb) + return -ENOMEM; + + err = ubi_io_read_vid_hdr(ubi, *pnum, vidb, 0); + if (err > 0 && err != UBI_IO_BITFLIPS) { + int torture = 0; + + switch (err) { + case UBI_IO_FF: + case UBI_IO_FF_BITFLIPS: + case UBI_IO_BAD_HDR: + case UBI_IO_BAD_HDR_EBADMSG: + break; + default: + ubi_assert(0); + } + + if (err == UBI_IO_BAD_HDR_EBADMSG || err == UBI_IO_FF_BITFLIPS) + torture = 1; + + down_read(&ubi->fm_eba_sem); + vol->eba_tbl->entries[lnum].pnum = UBI_LEB_UNMAPPED; + up_read(&ubi->fm_eba_sem); + ubi_wl_put_peb(ubi, vol->vol_id, lnum, *pnum, torture); + + *pnum = UBI_LEB_UNMAPPED; + } else if (err < 0) { + ubi_err(ubi, "unable to read VID header back from PEB %i: %i", + *pnum, err); + + goto out_free; + } else { + int found_vol_id, found_lnum; + + ubi_assert(err == 0 || err == UBI_IO_BITFLIPS); + + vid_hdr = ubi_get_vid_hdr(vidb); + found_vol_id = be32_to_cpu(vid_hdr->vol_id); + found_lnum = be32_to_cpu(vid_hdr->lnum); + + if (found_lnum != lnum || found_vol_id != vol->vol_id) { + ubi_err(ubi, "EBA mismatch! PEB %i is LEB %i:%i instead of LEB %i:%i", + *pnum, found_vol_id, found_lnum, vol->vol_id, lnum); + ubi_ro_mode(ubi); + err = -EINVAL; + goto out_free; + } + } + + set_bit(lnum, vol->checkmap); + err = 0; + +out_free: + ubi_free_vid_buf(vidb); + + return err; +} +#else +static int check_mapping(struct ubi_device *ubi, struct ubi_volume *vol, int lnum, + int *pnum) +{ + return 0; +} +#endif + +/** + * ubi_eba_read_leb - read data. + * @ubi: UBI device description object + * @vol: volume description object + * @lnum: logical eraseblock number + * @buf: buffer to store the read data + * @offset: offset from where to read + * @len: how many bytes to read + * @check: data CRC check flag + * + * If the logical eraseblock @lnum is unmapped, @buf is filled with 0xFF + * bytes. The @check flag only makes sense for static volumes and forces + * eraseblock data CRC checking. + * + * In case of success this function returns zero. In case of a static volume, + * if data CRC mismatches - %-EBADMSG is returned. %-EBADMSG may also be + * returned for any volume type if an ECC error was detected by the MTD device + * driver. Other negative error cored may be returned in case of other errors. + */ +int ubi_eba_read_leb(struct ubi_device *ubi, struct ubi_volume *vol, int lnum, + void *buf, int offset, int len, int check) +{ + int err, pnum, scrub = 0, vol_id = vol->vol_id; + struct ubi_vid_io_buf *vidb; + struct ubi_vid_hdr *vid_hdr; + uint32_t crc; + + err = leb_read_lock(ubi, vol_id, lnum); + if (err) + return err; + + pnum = vol->eba_tbl->entries[lnum].pnum; + if (pnum >= 0) { + err = check_mapping(ubi, vol, lnum, &pnum); + if (err < 0) + goto out_unlock; + } + + if (pnum == UBI_LEB_UNMAPPED) { + /* + * The logical eraseblock is not mapped, fill the whole buffer + * with 0xFF bytes. The exception is static volumes for which + * it is an error to read unmapped logical eraseblocks. + */ + dbg_eba("read %d bytes from offset %d of LEB %d:%d (unmapped)", + len, offset, vol_id, lnum); + leb_read_unlock(ubi, vol_id, lnum); + ubi_assert(vol->vol_type != UBI_STATIC_VOLUME); + memset(buf, 0xFF, len); + return 0; + } + + dbg_eba("read %d bytes from offset %d of LEB %d:%d, PEB %d", + len, offset, vol_id, lnum, pnum); + + if (vol->vol_type == UBI_DYNAMIC_VOLUME) + check = 0; + +retry: + if (check) { + vidb = ubi_alloc_vid_buf(ubi, GFP_NOFS); + if (!vidb) { + err = -ENOMEM; + goto out_unlock; + } + + vid_hdr = ubi_get_vid_hdr(vidb); + + err = ubi_io_read_vid_hdr(ubi, pnum, vidb, 1); + if (err && err != UBI_IO_BITFLIPS) { + if (err > 0) { + /* + * The header is either absent or corrupted. + * The former case means there is a bug - + * switch to read-only mode just in case. + * The latter case means a real corruption - we + * may try to recover data. FIXME: but this is + * not implemented. + */ + if (err == UBI_IO_BAD_HDR_EBADMSG || + err == UBI_IO_BAD_HDR) { + ubi_warn(ubi, "corrupted VID header at PEB %d, LEB %d:%d", + pnum, vol_id, lnum); + err = -EBADMSG; + } else { + /* + * Ending up here in the non-Fastmap case + * is a clear bug as the VID header had to + * be present at scan time to have it referenced. + * With fastmap the story is more complicated. + * Fastmap has the mapping info without the need + * of a full scan. So the LEB could have been + * unmapped, Fastmap cannot know this and keeps + * the LEB referenced. + * This is valid and works as the layer above UBI + * has to do bookkeeping about used/referenced + * LEBs in any case. + */ + if (ubi->fast_attach) { + err = -EBADMSG; + } else { + err = -EINVAL; + ubi_ro_mode(ubi); + } + } + } + goto out_free; + } else if (err == UBI_IO_BITFLIPS) + scrub = 1; + + ubi_assert(lnum < be32_to_cpu(vid_hdr->used_ebs)); + ubi_assert(len == be32_to_cpu(vid_hdr->data_size)); + + crc = be32_to_cpu(vid_hdr->data_crc); + ubi_free_vid_buf(vidb); + } + + err = ubi_io_read_data(ubi, buf, pnum, offset, len); + if (err) { + if (err == UBI_IO_BITFLIPS) + scrub = 1; + else if (mtd_is_eccerr(err)) { + if (vol->vol_type == UBI_DYNAMIC_VOLUME) + goto out_unlock; + scrub = 1; + if (!check) { + ubi_msg(ubi, "force data checking"); + check = 1; + goto retry; + } + } else + goto out_unlock; + } + + if (check) { + uint32_t crc1 = crc32(UBI_CRC32_INIT, buf, len); + if (crc1 != crc) { + ubi_warn(ubi, "CRC error: calculated %#08x, must be %#08x", + crc1, crc); + err = -EBADMSG; + goto out_unlock; + } + } + + if (scrub) + err = ubi_wl_scrub_peb(ubi, pnum); + + leb_read_unlock(ubi, vol_id, lnum); + return err; + +out_free: + ubi_free_vid_buf(vidb); +out_unlock: + leb_read_unlock(ubi, vol_id, lnum); + return err; +} + +/** + * ubi_eba_read_leb_sg - read data into a scatter gather list. + * @ubi: UBI device description object + * @vol: volume description object + * @lnum: logical eraseblock number + * @sgl: UBI scatter gather list to store the read data + * @offset: offset from where to read + * @len: how many bytes to read + * @check: data CRC check flag + * + * This function works exactly like ubi_eba_read_leb(). But instead of + * storing the read data into a buffer it writes to an UBI scatter gather + * list. + */ +int ubi_eba_read_leb_sg(struct ubi_device *ubi, struct ubi_volume *vol, + struct ubi_sgl *sgl, int lnum, int offset, int len, + int check) +{ + int to_read; + int ret; + struct scatterlist *sg; + + for (;;) { + ubi_assert(sgl->list_pos < UBI_MAX_SG_COUNT); + sg = &sgl->sg[sgl->list_pos]; + if (len < sg->length - sgl->page_pos) + to_read = len; + else + to_read = sg->length - sgl->page_pos; + + ret = ubi_eba_read_leb(ubi, vol, lnum, + sg_virt(sg) + sgl->page_pos, offset, + to_read, check); + if (ret < 0) + return ret; + + offset += to_read; + len -= to_read; + if (!len) { + sgl->page_pos += to_read; + if (sgl->page_pos == sg->length) { + sgl->list_pos++; + sgl->page_pos = 0; + } + + break; + } + + sgl->list_pos++; + sgl->page_pos = 0; + } + + return ret; +} + +/** + * try_recover_peb - try to recover from write failure. + * @vol: volume description object + * @pnum: the physical eraseblock to recover + * @lnum: logical eraseblock number + * @buf: data which was not written because of the write failure + * @offset: offset of the failed write + * @len: how many bytes should have been written + * @vidb: VID buffer + * @retry: whether the caller should retry in case of failure + * + * This function is called in case of a write failure and moves all good data + * from the potentially bad physical eraseblock to a good physical eraseblock. + * This function also writes the data which was not written due to the failure. + * Returns 0 in case of success, and a negative error code in case of failure. + * In case of failure, the %retry parameter is set to false if this is a fatal + * error (retrying won't help), and true otherwise. + */ +static int try_recover_peb(struct ubi_volume *vol, int pnum, int lnum, + const void *buf, int offset, int len, + struct ubi_vid_io_buf *vidb, bool *retry) +{ + struct ubi_device *ubi = vol->ubi; + struct ubi_vid_hdr *vid_hdr; + int new_pnum, err, vol_id = vol->vol_id, data_size; + uint32_t crc; + + *retry = false; + + new_pnum = ubi_wl_get_peb(ubi); + if (new_pnum < 0) { + err = new_pnum; + goto out_put; + } + + ubi_msg(ubi, "recover PEB %d, move data to PEB %d", + pnum, new_pnum); + + err = ubi_io_read_vid_hdr(ubi, pnum, vidb, 1); + if (err && err != UBI_IO_BITFLIPS) { + if (err > 0) + err = -EIO; + goto out_put; + } + + vid_hdr = ubi_get_vid_hdr(vidb); + ubi_assert(vid_hdr->vol_type == UBI_VID_DYNAMIC); + + mutex_lock(&ubi->buf_mutex); + memset(ubi->peb_buf + offset, 0xFF, len); + + /* Read everything before the area where the write failure happened */ + if (offset > 0) { + err = ubi_io_read_data(ubi, ubi->peb_buf, pnum, 0, offset); + if (err && err != UBI_IO_BITFLIPS) + goto out_unlock; + } + + *retry = true; + + memcpy(ubi->peb_buf + offset, buf, len); + + data_size = offset + len; + crc = crc32(UBI_CRC32_INIT, ubi->peb_buf, data_size); + vid_hdr->sqnum = cpu_to_be64(ubi_next_sqnum(ubi)); + vid_hdr->copy_flag = 1; + vid_hdr->data_size = cpu_to_be32(data_size); + vid_hdr->data_crc = cpu_to_be32(crc); + err = ubi_io_write_vid_hdr(ubi, new_pnum, vidb); + if (err) + goto out_unlock; + + err = ubi_io_write_data(ubi, ubi->peb_buf, new_pnum, 0, data_size); + +out_unlock: + mutex_unlock(&ubi->buf_mutex); + + if (!err) + vol->eba_tbl->entries[lnum].pnum = new_pnum; + +out_put: + up_read(&ubi->fm_eba_sem); + + if (!err) { + ubi_wl_put_peb(ubi, vol_id, lnum, pnum, 1); + ubi_msg(ubi, "data was successfully recovered"); + } else if (new_pnum >= 0) { + /* + * Bad luck? This physical eraseblock is bad too? Crud. Let's + * try to get another one. + */ + ubi_wl_put_peb(ubi, vol_id, lnum, new_pnum, 1); + ubi_warn(ubi, "failed to write to PEB %d", new_pnum); + } + + return err; +} + +/** + * recover_peb - recover from write failure. + * @ubi: UBI device description object + * @pnum: the physical eraseblock to recover + * @vol_id: volume ID + * @lnum: logical eraseblock number + * @buf: data which was not written because of the write failure + * @offset: offset of the failed write + * @len: how many bytes should have been written + * + * This function is called in case of a write failure and moves all good data + * from the potentially bad physical eraseblock to a good physical eraseblock. + * This function also writes the data which was not written due to the failure. + * Returns 0 in case of success, and a negative error code in case of failure. + * This function tries %UBI_IO_RETRIES before giving up. + */ +static int recover_peb(struct ubi_device *ubi, int pnum, int vol_id, int lnum, + const void *buf, int offset, int len) +{ + int err, idx = vol_id2idx(ubi, vol_id), tries; + struct ubi_volume *vol = ubi->volumes[idx]; + struct ubi_vid_io_buf *vidb; + + vidb = ubi_alloc_vid_buf(ubi, GFP_NOFS); + if (!vidb) + return -ENOMEM; + + for (tries = 0; tries <= UBI_IO_RETRIES; tries++) { + bool retry; + + err = try_recover_peb(vol, pnum, lnum, buf, offset, len, vidb, + &retry); + if (!err || !retry) + break; + + ubi_msg(ubi, "try again"); + } + + ubi_free_vid_buf(vidb); + + return err; +} + +/** + * try_write_vid_and_data - try to write VID header and data to a new PEB. + * @vol: volume description object + * @lnum: logical eraseblock number + * @vidb: the VID buffer to write + * @buf: buffer containing the data + * @offset: where to start writing data + * @len: how many bytes should be written + * + * This function tries to write VID header and data belonging to logical + * eraseblock @lnum of volume @vol to a new physical eraseblock. Returns zero + * in case of success and a negative error code in case of failure. + * In case of error, it is possible that something was still written to the + * flash media, but may be some garbage. + */ +static int try_write_vid_and_data(struct ubi_volume *vol, int lnum, + struct ubi_vid_io_buf *vidb, const void *buf, + int offset, int len) +{ + struct ubi_device *ubi = vol->ubi; + int pnum, opnum, err, err2, vol_id = vol->vol_id; + + pnum = ubi_wl_get_peb(ubi); + if (pnum < 0) { + err = pnum; + goto out_put; + } + + opnum = vol->eba_tbl->entries[lnum].pnum; + + dbg_eba("write VID hdr and %d bytes at offset %d of LEB %d:%d, PEB %d", + len, offset, vol_id, lnum, pnum); + + err = ubi_io_write_vid_hdr(ubi, pnum, vidb); + if (err) { + ubi_warn(ubi, "failed to write VID header to LEB %d:%d, PEB %d", + vol_id, lnum, pnum); + goto out_put; + } + + if (len) { + err = ubi_io_write_data(ubi, buf, pnum, offset, len); + if (err) { + ubi_warn(ubi, + "failed to write %d bytes at offset %d of LEB %d:%d, PEB %d", + len, offset, vol_id, lnum, pnum); + goto out_put; + } + } + + vol->eba_tbl->entries[lnum].pnum = pnum; + +out_put: + up_read(&ubi->fm_eba_sem); + + if (err && pnum >= 0) { + err2 = ubi_wl_put_peb(ubi, vol_id, lnum, pnum, 1); + if (err2) { + ubi_warn(ubi, "failed to return physical eraseblock %d, error %d", + pnum, err2); + } + } else if (!err && opnum >= 0) { + err2 = ubi_wl_put_peb(ubi, vol_id, lnum, opnum, 0); + if (err2) { + ubi_warn(ubi, "failed to return physical eraseblock %d, error %d", + opnum, err2); + } + } + + return err; +} + +/** + * ubi_eba_write_leb - write data to dynamic volume. + * @ubi: UBI device description object + * @vol: volume description object + * @lnum: logical eraseblock number + * @buf: the data to write + * @offset: offset within the logical eraseblock where to write + * @len: how many bytes to write + * + * This function writes data to logical eraseblock @lnum of a dynamic volume + * @vol. Returns zero in case of success and a negative error code in case + * of failure. In case of error, it is possible that something was still + * written to the flash media, but may be some garbage. + * This function retries %UBI_IO_RETRIES times before giving up. + */ +int ubi_eba_write_leb(struct ubi_device *ubi, struct ubi_volume *vol, int lnum, + const void *buf, int offset, int len) +{ + int err, pnum, tries, vol_id = vol->vol_id; + struct ubi_vid_io_buf *vidb; + struct ubi_vid_hdr *vid_hdr; + + if (ubi->ro_mode) + return -EROFS; + + err = leb_write_lock(ubi, vol_id, lnum); + if (err) + return err; + + pnum = vol->eba_tbl->entries[lnum].pnum; + if (pnum >= 0) { + err = check_mapping(ubi, vol, lnum, &pnum); + if (err < 0) + goto out; + } + + if (pnum >= 0) { + dbg_eba("write %d bytes at offset %d of LEB %d:%d, PEB %d", + len, offset, vol_id, lnum, pnum); + + err = ubi_io_write_data(ubi, buf, pnum, offset, len); + if (err) { + ubi_warn(ubi, "failed to write data to PEB %d", pnum); + if (err == -EIO && ubi->bad_allowed) + err = recover_peb(ubi, pnum, vol_id, lnum, buf, + offset, len); + } + + goto out; + } + + /* + * The logical eraseblock is not mapped. We have to get a free physical + * eraseblock and write the volume identifier header there first. + */ + vidb = ubi_alloc_vid_buf(ubi, GFP_NOFS); + if (!vidb) { + leb_write_unlock(ubi, vol_id, lnum); + return -ENOMEM; + } + + vid_hdr = ubi_get_vid_hdr(vidb); + + vid_hdr->vol_type = UBI_VID_DYNAMIC; + vid_hdr->sqnum = cpu_to_be64(ubi_next_sqnum(ubi)); + vid_hdr->vol_id = cpu_to_be32(vol_id); + vid_hdr->lnum = cpu_to_be32(lnum); + vid_hdr->compat = ubi_get_compat(ubi, vol_id); + vid_hdr->data_pad = cpu_to_be32(vol->data_pad); + + for (tries = 0; tries <= UBI_IO_RETRIES; tries++) { + err = try_write_vid_and_data(vol, lnum, vidb, buf, offset, len); + if (err != -EIO || !ubi->bad_allowed) + break; + + /* + * Fortunately, this is the first write operation to this + * physical eraseblock, so just put it and request a new one. + * We assume that if this physical eraseblock went bad, the + * erase code will handle that. + */ + vid_hdr->sqnum = cpu_to_be64(ubi_next_sqnum(ubi)); + ubi_msg(ubi, "try another PEB"); + } + + ubi_free_vid_buf(vidb); + +out: + if (err) + ubi_ro_mode(ubi); + + leb_write_unlock(ubi, vol_id, lnum); + + return err; +} + +/** + * ubi_eba_write_leb_st - write data to static volume. + * @ubi: UBI device description object + * @vol: volume description object + * @lnum: logical eraseblock number + * @buf: data to write + * @len: how many bytes to write + * @used_ebs: how many logical eraseblocks will this volume contain + * + * This function writes data to logical eraseblock @lnum of static volume + * @vol. The @used_ebs argument should contain total number of logical + * eraseblock in this static volume. + * + * When writing to the last logical eraseblock, the @len argument doesn't have + * to be aligned to the minimal I/O unit size. Instead, it has to be equivalent + * to the real data size, although the @buf buffer has to contain the + * alignment. In all other cases, @len has to be aligned. + * + * It is prohibited to write more than once to logical eraseblocks of static + * volumes. This function returns zero in case of success and a negative error + * code in case of failure. + */ +int ubi_eba_write_leb_st(struct ubi_device *ubi, struct ubi_volume *vol, + int lnum, const void *buf, int len, int used_ebs) +{ + int err, tries, data_size = len, vol_id = vol->vol_id; + struct ubi_vid_io_buf *vidb; + struct ubi_vid_hdr *vid_hdr; + uint32_t crc; + + if (ubi->ro_mode) + return -EROFS; + + if (lnum == used_ebs - 1) + /* If this is the last LEB @len may be unaligned */ + len = ALIGN(data_size, ubi->min_io_size); + else + ubi_assert(!(len & (ubi->min_io_size - 1))); + + vidb = ubi_alloc_vid_buf(ubi, GFP_NOFS); + if (!vidb) + return -ENOMEM; + + vid_hdr = ubi_get_vid_hdr(vidb); + + err = leb_write_lock(ubi, vol_id, lnum); + if (err) + goto out; + + vid_hdr->sqnum = cpu_to_be64(ubi_next_sqnum(ubi)); + vid_hdr->vol_id = cpu_to_be32(vol_id); + vid_hdr->lnum = cpu_to_be32(lnum); + vid_hdr->compat = ubi_get_compat(ubi, vol_id); + vid_hdr->data_pad = cpu_to_be32(vol->data_pad); + + crc = crc32(UBI_CRC32_INIT, buf, data_size); + vid_hdr->vol_type = UBI_VID_STATIC; + vid_hdr->data_size = cpu_to_be32(data_size); + vid_hdr->used_ebs = cpu_to_be32(used_ebs); + vid_hdr->data_crc = cpu_to_be32(crc); + + ubi_assert(vol->eba_tbl->entries[lnum].pnum < 0); + + for (tries = 0; tries <= UBI_IO_RETRIES; tries++) { + err = try_write_vid_and_data(vol, lnum, vidb, buf, 0, len); + if (err != -EIO || !ubi->bad_allowed) + break; + + vid_hdr->sqnum = cpu_to_be64(ubi_next_sqnum(ubi)); + ubi_msg(ubi, "try another PEB"); + } + + if (err) + ubi_ro_mode(ubi); + + leb_write_unlock(ubi, vol_id, lnum); + +out: + ubi_free_vid_buf(vidb); + + return err; +} + +/* + * ubi_eba_atomic_leb_change - change logical eraseblock atomically. + * @ubi: UBI device description object + * @vol: volume description object + * @lnum: logical eraseblock number + * @buf: data to write + * @len: how many bytes to write + * + * This function changes the contents of a logical eraseblock atomically. @buf + * has to contain new logical eraseblock data, and @len - the length of the + * data, which has to be aligned. This function guarantees that in case of an + * unclean reboot the old contents is preserved. Returns zero in case of + * success and a negative error code in case of failure. + * + * UBI reserves one LEB for the "atomic LEB change" operation, so only one + * LEB change may be done at a time. This is ensured by @ubi->alc_mutex. + */ +int ubi_eba_atomic_leb_change(struct ubi_device *ubi, struct ubi_volume *vol, + int lnum, const void *buf, int len) +{ + int err, tries, vol_id = vol->vol_id; + struct ubi_vid_io_buf *vidb; + struct ubi_vid_hdr *vid_hdr; + uint32_t crc; + + if (ubi->ro_mode) + return -EROFS; + + if (len == 0) { + /* + * Special case when data length is zero. In this case the LEB + * has to be unmapped and mapped somewhere else. + */ + err = ubi_eba_unmap_leb(ubi, vol, lnum); + if (err) + return err; + return ubi_eba_write_leb(ubi, vol, lnum, NULL, 0, 0); + } + + vidb = ubi_alloc_vid_buf(ubi, GFP_NOFS); + if (!vidb) + return -ENOMEM; + + vid_hdr = ubi_get_vid_hdr(vidb); + + mutex_lock(&ubi->alc_mutex); + err = leb_write_lock(ubi, vol_id, lnum); + if (err) + goto out_mutex; + + vid_hdr->sqnum = cpu_to_be64(ubi_next_sqnum(ubi)); + vid_hdr->vol_id = cpu_to_be32(vol_id); + vid_hdr->lnum = cpu_to_be32(lnum); + vid_hdr->compat = ubi_get_compat(ubi, vol_id); + vid_hdr->data_pad = cpu_to_be32(vol->data_pad); + + crc = crc32(UBI_CRC32_INIT, buf, len); + vid_hdr->vol_type = UBI_VID_DYNAMIC; + vid_hdr->data_size = cpu_to_be32(len); + vid_hdr->copy_flag = 1; + vid_hdr->data_crc = cpu_to_be32(crc); + + dbg_eba("change LEB %d:%d", vol_id, lnum); + + for (tries = 0; tries <= UBI_IO_RETRIES; tries++) { + err = try_write_vid_and_data(vol, lnum, vidb, buf, 0, len); + if (err != -EIO || !ubi->bad_allowed) + break; + + vid_hdr->sqnum = cpu_to_be64(ubi_next_sqnum(ubi)); + ubi_msg(ubi, "try another PEB"); + } + + /* + * This flash device does not admit of bad eraseblocks or + * something nasty and unexpected happened. Switch to read-only + * mode just in case. + */ + if (err) + ubi_ro_mode(ubi); + + leb_write_unlock(ubi, vol_id, lnum); + +out_mutex: + mutex_unlock(&ubi->alc_mutex); + ubi_free_vid_buf(vidb); + return err; +} + +/** + * is_error_sane - check whether a read error is sane. + * @err: code of the error happened during reading + * + * This is a helper function for 'ubi_eba_copy_leb()' which is called when we + * cannot read data from the target PEB (an error @err happened). If the error + * code is sane, then we treat this error as non-fatal. Otherwise the error is + * fatal and UBI will be switched to R/O mode later. + * + * The idea is that we try not to switch to R/O mode if the read error is + * something which suggests there was a real read problem. E.g., %-EIO. Or a + * memory allocation failed (-%ENOMEM). Otherwise, it is safer to switch to R/O + * mode, simply because we do not know what happened at the MTD level, and we + * cannot handle this. E.g., the underlying driver may have become crazy, and + * it is safer to switch to R/O mode to preserve the data. + * + * And bear in mind, this is about reading from the target PEB, i.e. the PEB + * which we have just written. + */ +static int is_error_sane(int err) +{ + if (err == -EIO || err == -ENOMEM || err == UBI_IO_BAD_HDR || + err == UBI_IO_BAD_HDR_EBADMSG || err == -ETIMEDOUT) + return 0; + return 1; +} + +/** + * ubi_eba_copy_leb - copy logical eraseblock. + * @ubi: UBI device description object + * @from: physical eraseblock number from where to copy + * @to: physical eraseblock number where to copy + * @vidb: data structure from where the VID header is derived + * + * This function copies logical eraseblock from physical eraseblock @from to + * physical eraseblock @to. The @vid_hdr buffer may be changed by this + * function. Returns: + * o %0 in case of success; + * o %MOVE_CANCEL_RACE, %MOVE_TARGET_WR_ERR, %MOVE_TARGET_BITFLIPS, etc; + * o a negative error code in case of failure. + */ +int ubi_eba_copy_leb(struct ubi_device *ubi, int from, int to, + struct ubi_vid_io_buf *vidb) +{ + int err, vol_id, lnum, data_size, aldata_size, idx; + struct ubi_vid_hdr *vid_hdr = ubi_get_vid_hdr(vidb); + struct ubi_volume *vol; + uint32_t crc; + + ubi_assert(rwsem_is_locked(&ubi->fm_eba_sem)); + + vol_id = be32_to_cpu(vid_hdr->vol_id); + lnum = be32_to_cpu(vid_hdr->lnum); + + dbg_wl("copy LEB %d:%d, PEB %d to PEB %d", vol_id, lnum, from, to); + + if (vid_hdr->vol_type == UBI_VID_STATIC) { + data_size = be32_to_cpu(vid_hdr->data_size); + aldata_size = ALIGN(data_size, ubi->min_io_size); + } else + data_size = aldata_size = + ubi->leb_size - be32_to_cpu(vid_hdr->data_pad); + + idx = vol_id2idx(ubi, vol_id); + spin_lock(&ubi->volumes_lock); + /* + * Note, we may race with volume deletion, which means that the volume + * this logical eraseblock belongs to might be being deleted. Since the + * volume deletion un-maps all the volume's logical eraseblocks, it will + * be locked in 'ubi_wl_put_peb()' and wait for the WL worker to finish. + */ + vol = ubi->volumes[idx]; + spin_unlock(&ubi->volumes_lock); + if (!vol) { + /* No need to do further work, cancel */ + dbg_wl("volume %d is being removed, cancel", vol_id); + return MOVE_CANCEL_RACE; + } + + /* + * We do not want anybody to write to this logical eraseblock while we + * are moving it, so lock it. + * + * Note, we are using non-waiting locking here, because we cannot sleep + * on the LEB, since it may cause deadlocks. Indeed, imagine a task is + * unmapping the LEB which is mapped to the PEB we are going to move + * (@from). This task locks the LEB and goes sleep in the + * 'ubi_wl_put_peb()' function on the @ubi->move_mutex. In turn, we are + * holding @ubi->move_mutex and go sleep on the LEB lock. So, if the + * LEB is already locked, we just do not move it and return + * %MOVE_RETRY. Note, we do not return %MOVE_CANCEL_RACE here because + * we do not know the reasons of the contention - it may be just a + * normal I/O on this LEB, so we want to re-try. + */ + err = leb_write_trylock(ubi, vol_id, lnum); + if (err) { + dbg_wl("contention on LEB %d:%d, cancel", vol_id, lnum); + return MOVE_RETRY; + } + + /* + * The LEB might have been put meanwhile, and the task which put it is + * probably waiting on @ubi->move_mutex. No need to continue the work, + * cancel it. + */ + if (vol->eba_tbl->entries[lnum].pnum != from) { + dbg_wl("LEB %d:%d is no longer mapped to PEB %d, mapped to PEB %d, cancel", + vol_id, lnum, from, vol->eba_tbl->entries[lnum].pnum); + err = MOVE_CANCEL_RACE; + goto out_unlock_leb; + } + + /* + * OK, now the LEB is locked and we can safely start moving it. Since + * this function utilizes the @ubi->peb_buf buffer which is shared + * with some other functions - we lock the buffer by taking the + * @ubi->buf_mutex. + */ + mutex_lock(&ubi->buf_mutex); + dbg_wl("read %d bytes of data", aldata_size); + err = ubi_io_read_data(ubi, ubi->peb_buf, from, 0, aldata_size); + if (err && err != UBI_IO_BITFLIPS) { + ubi_warn(ubi, "error %d while reading data from PEB %d", + err, from); + err = MOVE_SOURCE_RD_ERR; + goto out_unlock_buf; + } + + /* + * Now we have got to calculate how much data we have to copy. In + * case of a static volume it is fairly easy - the VID header contains + * the data size. In case of a dynamic volume it is more difficult - we + * have to read the contents, cut 0xFF bytes from the end and copy only + * the first part. We must do this to avoid writing 0xFF bytes as it + * may have some side-effects. And not only this. It is important not + * to include those 0xFFs to CRC because later the they may be filled + * by data. + */ + if (vid_hdr->vol_type == UBI_VID_DYNAMIC) + aldata_size = data_size = + ubi_calc_data_len(ubi, ubi->peb_buf, data_size); + + cond_resched(); + crc = crc32(UBI_CRC32_INIT, ubi->peb_buf, data_size); + cond_resched(); + + /* + * It may turn out to be that the whole @from physical eraseblock + * contains only 0xFF bytes. Then we have to only write the VID header + * and do not write any data. This also means we should not set + * @vid_hdr->copy_flag, @vid_hdr->data_size, and @vid_hdr->data_crc. + */ + if (data_size > 0) { + vid_hdr->copy_flag = 1; + vid_hdr->data_size = cpu_to_be32(data_size); + vid_hdr->data_crc = cpu_to_be32(crc); + } + vid_hdr->sqnum = cpu_to_be64(ubi_next_sqnum(ubi)); + + err = ubi_io_write_vid_hdr(ubi, to, vidb); + if (err) { + if (err == -EIO) + err = MOVE_TARGET_WR_ERR; + goto out_unlock_buf; + } + + cond_resched(); + + /* Read the VID header back and check if it was written correctly */ + err = ubi_io_read_vid_hdr(ubi, to, vidb, 1); + if (err) { + if (err != UBI_IO_BITFLIPS) { + ubi_warn(ubi, "error %d while reading VID header back from PEB %d", + err, to); + if (is_error_sane(err)) + err = MOVE_TARGET_RD_ERR; + } else + err = MOVE_TARGET_BITFLIPS; + goto out_unlock_buf; + } + + if (data_size > 0) { + err = ubi_io_write_data(ubi, ubi->peb_buf, to, 0, aldata_size); + if (err) { + if (err == -EIO) + err = MOVE_TARGET_WR_ERR; + goto out_unlock_buf; + } + + cond_resched(); + } + + ubi_assert(vol->eba_tbl->entries[lnum].pnum == from); + + /** + * The volumes_lock lock is needed here to prevent the expired old eba_tbl + * being updated when the eba_tbl is copied in the ubi_resize_volume() process. + */ + spin_lock(&ubi->volumes_lock); + vol->eba_tbl->entries[lnum].pnum = to; + spin_unlock(&ubi->volumes_lock); + +out_unlock_buf: + mutex_unlock(&ubi->buf_mutex); +out_unlock_leb: + leb_write_unlock(ubi, vol_id, lnum); + return err; +} + +/** + * print_rsvd_warning - warn about not having enough reserved PEBs. + * @ubi: UBI device description object + * @ai: UBI attach info object + * + * This is a helper function for 'ubi_eba_init()' which is called when UBI + * cannot reserve enough PEBs for bad block handling. This function makes a + * decision whether we have to print a warning or not. The algorithm is as + * follows: + * o if this is a new UBI image, then just print the warning + * o if this is an UBI image which has already been used for some time, print + * a warning only if we can reserve less than 10% of the expected amount of + * the reserved PEB. + * + * The idea is that when UBI is used, PEBs become bad, and the reserved pool + * of PEBs becomes smaller, which is normal and we do not want to scare users + * with a warning every time they attach the MTD device. This was an issue + * reported by real users. + */ +static void print_rsvd_warning(struct ubi_device *ubi, + struct ubi_attach_info *ai) +{ + /* + * The 1 << 18 (256KiB) number is picked randomly, just a reasonably + * large number to distinguish between newly flashed and used images. + */ + if (ai->max_sqnum > (1 << 18)) { + int min = ubi->beb_rsvd_level / 10; + + if (!min) + min = 1; + if (ubi->beb_rsvd_pebs > min) + return; + } + + ubi_warn(ubi, "cannot reserve enough PEBs for bad PEB handling, reserved %d, need %d", + ubi->beb_rsvd_pebs, ubi->beb_rsvd_level); + if (ubi->corr_peb_count) + ubi_warn(ubi, "%d PEBs are corrupted and not used", + ubi->corr_peb_count); +} + +/** + * self_check_eba - run a self check on the EBA table constructed by fastmap. + * @ubi: UBI device description object + * @ai_fastmap: UBI attach info object created by fastmap + * @ai_scan: UBI attach info object created by scanning + * + * Returns < 0 in case of an internal error, 0 otherwise. + * If a bad EBA table entry was found it will be printed out and + * ubi_assert() triggers. + */ +int self_check_eba(struct ubi_device *ubi, struct ubi_attach_info *ai_fastmap, + struct ubi_attach_info *ai_scan) +{ + int i, j, num_volumes, ret = 0; + int **scan_eba, **fm_eba; + struct ubi_ainf_volume *av; + struct ubi_volume *vol; + struct ubi_ainf_peb *aeb; + struct rb_node *rb; + + num_volumes = ubi->vtbl_slots + UBI_INT_VOL_COUNT; + + scan_eba = kmalloc_objs(*scan_eba, num_volumes); + if (!scan_eba) + return -ENOMEM; + + fm_eba = kmalloc_objs(*fm_eba, num_volumes); + if (!fm_eba) { + kfree(scan_eba); + return -ENOMEM; + } + + for (i = 0; i < num_volumes; i++) { + vol = ubi->volumes[i]; + if (!vol) + continue; + + scan_eba[i] = kmalloc_objs(**scan_eba, vol->reserved_pebs); + if (!scan_eba[i]) { + ret = -ENOMEM; + goto out_free; + } + + fm_eba[i] = kmalloc_objs(**fm_eba, vol->reserved_pebs); + if (!fm_eba[i]) { + ret = -ENOMEM; + kfree(scan_eba[i]); + goto out_free; + } + + for (j = 0; j < vol->reserved_pebs; j++) + scan_eba[i][j] = fm_eba[i][j] = UBI_LEB_UNMAPPED; + + av = ubi_find_av(ai_scan, idx2vol_id(ubi, i)); + if (!av) + continue; + + ubi_rb_for_each_entry(rb, aeb, &av->root, u.rb) + scan_eba[i][aeb->lnum] = aeb->pnum; + + av = ubi_find_av(ai_fastmap, idx2vol_id(ubi, i)); + if (!av) + continue; + + ubi_rb_for_each_entry(rb, aeb, &av->root, u.rb) + fm_eba[i][aeb->lnum] = aeb->pnum; + + for (j = 0; j < vol->reserved_pebs; j++) { + if (scan_eba[i][j] != fm_eba[i][j]) { + if (scan_eba[i][j] == UBI_LEB_UNMAPPED || + fm_eba[i][j] == UBI_LEB_UNMAPPED) + continue; + + ubi_err(ubi, "LEB:%i:%i is PEB:%i instead of %i!", + vol->vol_id, j, fm_eba[i][j], + scan_eba[i][j]); + ubi_assert(0); + } + } + } + +out_free: + while (--i >= 0) { + if (!ubi->volumes[i]) + continue; + + kfree(scan_eba[i]); + kfree(fm_eba[i]); + } + + kfree(scan_eba); + kfree(fm_eba); + return ret; +} + +/** + * ubi_eba_init - initialize the EBA sub-system using attaching information. + * @ubi: UBI device description object + * @ai: attaching information + * + * This function returns zero in case of success and a negative error code in + * case of failure. + */ +int ubi_eba_init(struct ubi_device *ubi, struct ubi_attach_info *ai) +{ + int i, err, num_volumes; + struct ubi_ainf_volume *av; + struct ubi_volume *vol; + struct ubi_ainf_peb *aeb; + struct rb_node *rb; + + dbg_eba("initialize EBA sub-system"); + + spin_lock_init(&ubi->ltree_lock); + mutex_init(&ubi->alc_mutex); + ubi->ltree = RB_ROOT; + + ubi->global_sqnum = ai->max_sqnum + 1; + num_volumes = ubi->vtbl_slots + UBI_INT_VOL_COUNT; + + for (i = 0; i < num_volumes; i++) { + struct ubi_eba_table *tbl; + + vol = ubi->volumes[i]; + if (!vol) + continue; + + cond_resched(); + + tbl = ubi_eba_create_table(vol, vol->reserved_pebs); + if (IS_ERR(tbl)) { + err = PTR_ERR(tbl); + goto out_free; + } + + ubi_eba_replace_table(vol, tbl); + + av = ubi_find_av(ai, idx2vol_id(ubi, i)); + if (!av) + continue; + + ubi_rb_for_each_entry(rb, aeb, &av->root, u.rb) { + if (aeb->lnum >= vol->reserved_pebs) { + /* + * This may happen in case of an unclean reboot + * during re-size. + */ + ubi_move_aeb_to_list(av, aeb, &ai->erase); + } else { + struct ubi_eba_entry *entry; + + entry = &vol->eba_tbl->entries[aeb->lnum]; + entry->pnum = aeb->pnum; + } + } + } + + if (ubi->avail_pebs < EBA_RESERVED_PEBS) { + ubi_err(ubi, "no enough physical eraseblocks (%d, need %d)", + ubi->avail_pebs, EBA_RESERVED_PEBS); + if (ubi->corr_peb_count) + ubi_err(ubi, "%d PEBs are corrupted and not used", + ubi->corr_peb_count); + err = -ENOSPC; + goto out_free; + } + ubi->avail_pebs -= EBA_RESERVED_PEBS; + ubi->rsvd_pebs += EBA_RESERVED_PEBS; + + if (ubi->bad_allowed) { + ubi_calculate_reserved(ubi); + + if (ubi->avail_pebs < ubi->beb_rsvd_level) { + /* No enough free physical eraseblocks */ + ubi->beb_rsvd_pebs = ubi->avail_pebs; + print_rsvd_warning(ubi, ai); + } else + ubi->beb_rsvd_pebs = ubi->beb_rsvd_level; + + ubi->avail_pebs -= ubi->beb_rsvd_pebs; + ubi->rsvd_pebs += ubi->beb_rsvd_pebs; + } + + dbg_eba("EBA sub-system is initialized"); + return 0; + +out_free: + for (i = 0; i < num_volumes; i++) { + if (!ubi->volumes[i]) + continue; + ubi_eba_replace_table(ubi->volumes[i], NULL); + } + return err; +} diff --git a/drivers/mtd/ubi/fastmap-wl.c b/drivers/mtd/ubi/fastmap-wl.c new file mode 100644 index 000000000..e24bb6b18 --- /dev/null +++ b/drivers/mtd/ubi/fastmap-wl.c @@ -0,0 +1,561 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * Copyright (c) 2012 Linutronix GmbH + * Copyright (c) 2014 sigma star gmbh + * Author: Richard Weinberger + */ + +/** + * update_fastmap_work_fn - calls ubi_update_fastmap from a work queue + * @wrk: the work description object + */ +static void update_fastmap_work_fn(struct work_struct *wrk) +{ + struct ubi_device *ubi = container_of(wrk, struct ubi_device, fm_work); + + ubi_update_fastmap(ubi); + spin_lock(&ubi->wl_lock); + ubi->fm_work_scheduled = 0; + spin_unlock(&ubi->wl_lock); +} + +/** + * find_anchor_wl_entry - find wear-leveling entry to used as anchor PEB. + * @root: the RB-tree where to look for + */ +static struct ubi_wl_entry *find_anchor_wl_entry(struct rb_root *root) +{ + struct rb_node *p; + struct ubi_wl_entry *e, *victim = NULL; + int max_ec = UBI_MAX_ERASECOUNTER; + + ubi_rb_for_each_entry(p, e, root, u.rb) { + if (e->pnum < UBI_FM_MAX_START && e->ec < max_ec) { + victim = e; + max_ec = e->ec; + } + } + + return victim; +} + +static inline void return_unused_peb(struct ubi_device *ubi, + struct ubi_wl_entry *e) +{ + wl_tree_add(e, &ubi->free); + ubi->free_count++; +} + +/** + * return_unused_pool_pebs - returns unused PEB to the free tree. + * @ubi: UBI device description object + * @pool: fastmap pool description object + */ +static void return_unused_pool_pebs(struct ubi_device *ubi, + struct ubi_fm_pool *pool) +{ + int i; + struct ubi_wl_entry *e; + + for (i = pool->used; i < pool->size; i++) { + e = ubi->lookuptbl[pool->pebs[i]]; + return_unused_peb(ubi, e); + } +} + +/** + * ubi_wl_get_fm_peb - find a physical erase block with a given maximal number. + * @ubi: UBI device description object + * @anchor: This PEB will be used as anchor PEB by fastmap + * + * The function returns a physical erase block with a given maximal number + * and removes it from the wl subsystem. + * Must be called with wl_lock held! + */ +struct ubi_wl_entry *ubi_wl_get_fm_peb(struct ubi_device *ubi, int anchor) +{ + struct ubi_wl_entry *e = NULL; + + if (!ubi->free.rb_node) + goto out; + + if (anchor) + e = find_anchor_wl_entry(&ubi->free); + else + e = find_mean_wl_entry(ubi, &ubi->free); + + if (!e) + goto out; + + self_check_in_wl_tree(ubi, e, &ubi->free); + + /* remove it from the free list, + * the wl subsystem does no longer know this erase block */ + rb_erase(&e->u.rb, &ubi->free); + ubi->free_count--; +out: + return e; +} + +/* + * wait_free_pebs_for_pool - wait until there enough free pebs + * @ubi: UBI device description object + * + * Wait and execute do_work until there are enough free pebs, fill pool + * as much as we can. This will reduce pool refilling times, which can + * reduce the fastmap updating frequency. + */ +static void wait_free_pebs_for_pool(struct ubi_device *ubi) +{ + struct ubi_fm_pool *wl_pool = &ubi->fm_wl_pool; + struct ubi_fm_pool *pool = &ubi->fm_pool; + int free, expect_free, executed; + /* + * There are at least following free pebs which reserved by UBI: + * 1. WL_RESERVED_PEBS[1] + * 2. EBA_RESERVED_PEBS[1] + * 3. fm pebs - 1: Twice fastmap size deducted by fastmap and fm_anchor + * 4. beb_rsvd_pebs: This value should be get under lock ubi->wl_lock + */ + int reserved = WL_RESERVED_PEBS + EBA_RESERVED_PEBS + + ubi->fm_size / ubi->leb_size - 1 + ubi->fm_pool_rsv_cnt; + + do { + spin_lock(&ubi->wl_lock); + free = ubi->free_count; + free += pool->size - pool->used + wl_pool->size - wl_pool->used; + expect_free = reserved + ubi->beb_rsvd_pebs; + spin_unlock(&ubi->wl_lock); + + /* + * Break out if there are no works or work is executed failure, + * given the fact that erase_worker will schedule itself when + * -EBUSY is returned from mtd layer caused by system shutdown. + */ + if (do_work(ubi, &executed) || !executed) + break; + } while (free < expect_free); +} + +/* + * left_free_count - returns the number of free pebs to fill fm pools + * @ubi: UBI device description object + * + * This helper function returns the number of free pebs (deducted + * by fastmap pebs) to fill fm_pool and fm_wl_pool. + */ +static int left_free_count(struct ubi_device *ubi) +{ + int fm_used = 0; // fastmap non anchor pebs. + + if (!ubi->free.rb_node) + return 0; + + if (!ubi->ro_mode && !ubi->fm_disabled) + fm_used = ubi->fm_size / ubi->leb_size - 1; + + return ubi->free_count - fm_used; +} + +/* + * can_fill_pools - whether free PEBs will be left after filling pools + * @ubi: UBI device description object + * @free: current number of free PEBs + * + * Return %1 if there are still left free PEBs after filling pools, + * otherwise %0 is returned. + */ +static int can_fill_pools(struct ubi_device *ubi, int free) +{ + struct ubi_fm_pool *wl_pool = &ubi->fm_wl_pool; + struct ubi_fm_pool *pool = &ubi->fm_pool; + int pool_need = pool->max_size - pool->size + + wl_pool->max_size - wl_pool->size; + + if (free - pool_need < 1) + return 0; + + return 1; +} + +/** + * ubi_refill_pools_and_lock - refills all fastmap PEB pools and takes fm locks. + * @ubi: UBI device description object + */ +void ubi_refill_pools_and_lock(struct ubi_device *ubi) +{ + struct ubi_fm_pool *wl_pool = &ubi->fm_wl_pool; + struct ubi_fm_pool *pool = &ubi->fm_pool; + struct ubi_wl_entry *e; + int enough; + + if (!ubi->ro_mode && !ubi->fm_disabled) + wait_free_pebs_for_pool(ubi); + + down_write(&ubi->fm_protect); + down_write(&ubi->work_sem); + down_write(&ubi->fm_eba_sem); + + spin_lock(&ubi->wl_lock); + + return_unused_pool_pebs(ubi, wl_pool); + return_unused_pool_pebs(ubi, pool); + + wl_pool->size = 0; + pool->size = 0; + + if (ubi->fm_anchor) { + wl_tree_add(ubi->fm_anchor, &ubi->free); + ubi->free_count++; + ubi->fm_anchor = NULL; + } + + if (!ubi->fm_disabled) + /* + * All available PEBs are in ubi->free, now is the time to get + * the best anchor PEBs. + */ + ubi->fm_anchor = ubi_wl_get_fm_peb(ubi, 1); + + for (;;) { + enough = 0; + if (pool->size < pool->max_size) { + if (left_free_count(ubi) <= 0) + break; + + e = wl_get_wle(ubi); + if (!e) + break; + + pool->pebs[pool->size] = e->pnum; + pool->size++; + } else + enough++; + + if (wl_pool->size < wl_pool->max_size) { + int left_free = left_free_count(ubi); + + if (left_free <= 0) + break; + + e = find_wl_entry(ubi, &ubi->free, ubi->wl_free_max_diff, + !can_fill_pools(ubi, left_free)); + self_check_in_wl_tree(ubi, e, &ubi->free); + rb_erase(&e->u.rb, &ubi->free); + ubi->free_count--; + + wl_pool->pebs[wl_pool->size] = e->pnum; + wl_pool->size++; + } else + enough++; + + if (enough == 2) + break; + } + + wl_pool->used = 0; + pool->used = 0; + + spin_unlock(&ubi->wl_lock); +} + +/** + * produce_free_peb - produce a free physical eraseblock. + * @ubi: UBI device description object + * + * This function tries to make a free PEB by means of synchronous execution of + * pending works. This may be needed if, for example the background thread is + * disabled. Returns zero in case of success and a negative error code in case + * of failure. + */ +static int produce_free_peb(struct ubi_device *ubi) +{ + int err; + + while (!ubi->free.rb_node && ubi->works_count) { + dbg_wl("do one work synchronously"); + err = do_work(ubi, NULL); + + if (err) + return err; + } + + return 0; +} + +/** + * ubi_wl_get_peb - get a physical eraseblock. + * @ubi: UBI device description object + * + * This function returns a physical eraseblock in case of success and a + * negative error code in case of failure. + * Returns with ubi->fm_eba_sem held in read mode! + */ +int ubi_wl_get_peb(struct ubi_device *ubi) +{ + int ret, attempts = 0; + struct ubi_fm_pool *pool = &ubi->fm_pool; + struct ubi_fm_pool *wl_pool = &ubi->fm_wl_pool; + +again: + down_read(&ubi->fm_eba_sem); + spin_lock(&ubi->wl_lock); + + /* We check here also for the WL pool because at this point we can + * refill the WL pool synchronous. */ + if (pool->used == pool->size || wl_pool->used == wl_pool->size) { + spin_unlock(&ubi->wl_lock); + up_read(&ubi->fm_eba_sem); + ret = ubi_update_fastmap(ubi); + if (ret) { + ubi_msg(ubi, "Unable to write a new fastmap: %i", ret); + down_read(&ubi->fm_eba_sem); + return -ENOSPC; + } + down_read(&ubi->fm_eba_sem); + spin_lock(&ubi->wl_lock); + } + + if (pool->used == pool->size) { + spin_unlock(&ubi->wl_lock); + attempts++; + if (attempts == 10) { + ubi_err(ubi, "Unable to get a free PEB from user WL pool"); + ret = -ENOSPC; + goto out; + } + up_read(&ubi->fm_eba_sem); + ret = produce_free_peb(ubi); + if (ret < 0) { + down_read(&ubi->fm_eba_sem); + goto out; + } + goto again; + } + + ubi_assert(pool->used < pool->size); + ret = pool->pebs[pool->used++]; + prot_queue_add(ubi, ubi->lookuptbl[ret]); + spin_unlock(&ubi->wl_lock); +out: + return ret; +} + +/** + * next_peb_for_wl - returns next PEB to be used internally by the + * WL sub-system. + * + * @ubi: UBI device description object + * @need_fill: whether to fill wear-leveling pool when no PEBs are found + */ +static struct ubi_wl_entry *next_peb_for_wl(struct ubi_device *ubi, + bool need_fill) +{ + struct ubi_fm_pool *pool = &ubi->fm_wl_pool; + int pnum; + + if (pool->used == pool->size) { + if (need_fill && !ubi->fm_work_scheduled) { + /* + * We cannot update the fastmap here because this + * function is called in atomic context. + * Let's fail here and refill/update it as soon as + * possible. + */ + ubi->fm_work_scheduled = 1; + schedule_work(&ubi->fm_work); + } + return NULL; + } + + pnum = pool->pebs[pool->used]; + return ubi->lookuptbl[pnum]; +} + +/** + * need_wear_leveling - checks whether to trigger a wear leveling work. + * UBI fetches free PEB from wl_pool, we check free PEBs from both 'wl_pool' + * and 'ubi->free', because free PEB in 'ubi->free' tree maybe moved into + * 'wl_pool' by ubi_refill_pools(). + * + * @ubi: UBI device description object + */ +static bool need_wear_leveling(struct ubi_device *ubi) +{ + int ec; + struct ubi_wl_entry *e; + + if (!ubi->used.rb_node) + return false; + + e = next_peb_for_wl(ubi, false); + if (!e) { + if (!ubi->free.rb_node) + return false; + e = find_wl_entry(ubi, &ubi->free, ubi->wl_free_max_diff, 0); + ec = e->ec; + } else { + ec = e->ec; + if (ubi->free.rb_node) { + e = find_wl_entry(ubi, &ubi->free, ubi->wl_free_max_diff, 0); + ec = max(ec, e->ec); + } + } + e = rb_entry(rb_first(&ubi->used), struct ubi_wl_entry, u.rb); + + return ec - e->ec >= ubi->wl_threshold; +} + +/* get_peb_for_wl - returns a PEB to be used internally by the WL sub-system. + * + * @ubi: UBI device description object + */ +static struct ubi_wl_entry *get_peb_for_wl(struct ubi_device *ubi) +{ + struct ubi_fm_pool *pool = &ubi->fm_wl_pool; + int pnum; + + ubi_assert(rwsem_is_locked(&ubi->fm_eba_sem)); + + if (pool->used == pool->size) { + /* We cannot update the fastmap here because this + * function is called in atomic context. + * Let's fail here and refill/update it as soon as possible. */ + if (!ubi->fm_work_scheduled) { + ubi->fm_work_scheduled = 1; + schedule_work(&ubi->fm_work); + } + return NULL; + } + + pnum = pool->pebs[pool->used++]; + return ubi->lookuptbl[pnum]; +} + +/** + * ubi_ensure_anchor_pebs - schedule wear-leveling to produce an anchor PEB. + * @ubi: UBI device description object + */ +int ubi_ensure_anchor_pebs(struct ubi_device *ubi) +{ + struct ubi_work *wrk; + struct ubi_wl_entry *anchor; + + spin_lock(&ubi->wl_lock); + + /* Do we already have an anchor? */ + if (ubi->fm_anchor) { + spin_unlock(&ubi->wl_lock); + return 0; + } + + /* See if we can find an anchor PEB on the list of free PEBs */ + anchor = ubi_wl_get_fm_peb(ubi, 1); + if (anchor) { + ubi->fm_anchor = anchor; + spin_unlock(&ubi->wl_lock); + return 0; + } + + ubi->fm_do_produce_anchor = 1; + /* No luck, trigger wear leveling to produce a new anchor PEB. */ + if (ubi->wl_scheduled) { + spin_unlock(&ubi->wl_lock); + return 0; + } + ubi->wl_scheduled = 1; + spin_unlock(&ubi->wl_lock); + + wrk = kmalloc_obj(struct ubi_work, GFP_NOFS); + if (!wrk) { + spin_lock(&ubi->wl_lock); + ubi->wl_scheduled = 0; + spin_unlock(&ubi->wl_lock); + return -ENOMEM; + } + + wrk->func = &wear_leveling_worker; + __schedule_ubi_work(ubi, wrk); + return 0; +} + +/** + * ubi_wl_put_fm_peb - returns a PEB used in a fastmap to the wear-leveling + * sub-system. + * see: ubi_wl_put_peb() + * + * @ubi: UBI device description object + * @fm_e: physical eraseblock to return + * @lnum: the last used logical eraseblock number for the PEB + * @torture: if this physical eraseblock has to be tortured + */ +int ubi_wl_put_fm_peb(struct ubi_device *ubi, struct ubi_wl_entry *fm_e, + int lnum, int torture) +{ + struct ubi_wl_entry *e; + int vol_id, pnum = fm_e->pnum; + + dbg_wl("PEB %d", pnum); + + ubi_assert(pnum >= 0); + ubi_assert(pnum < ubi->peb_count); + + spin_lock(&ubi->wl_lock); + e = ubi->lookuptbl[pnum]; + + /* This can happen if we recovered from a fastmap the very + * first time and writing now a new one. In this case the wl system + * has never seen any PEB used by the original fastmap. + */ + if (!e) { + e = fm_e; + ubi_assert(e->ec >= 0); + ubi->lookuptbl[pnum] = e; + } + + spin_unlock(&ubi->wl_lock); + + vol_id = lnum ? UBI_FM_DATA_VOLUME_ID : UBI_FM_SB_VOLUME_ID; + return schedule_erase(ubi, e, vol_id, lnum, torture, true); +} + +/** + * ubi_is_erase_work - checks whether a work is erase work. + * @wrk: The work object to be checked + */ +int ubi_is_erase_work(struct ubi_work *wrk) +{ + return wrk->func == erase_worker; +} + +static void ubi_fastmap_close(struct ubi_device *ubi) +{ + return_unused_pool_pebs(ubi, &ubi->fm_pool); + return_unused_pool_pebs(ubi, &ubi->fm_wl_pool); + + if (ubi->fm_anchor) { + return_unused_peb(ubi, ubi->fm_anchor); + ubi->fm_anchor = NULL; + } + + ubi_free_fastmap(ubi); +} + +/** + * may_reserve_for_fm - tests whether a PEB shall be reserved for fastmap. + * See find_mean_wl_entry() + * + * @ubi: UBI device description object + * @e: physical eraseblock to return + * @root: RB tree to test against. + */ +static struct ubi_wl_entry *may_reserve_for_fm(struct ubi_device *ubi, + struct ubi_wl_entry *e, + struct rb_root *root) { + if (e && !ubi->fm_disabled && !ubi->fm && !ubi->fm_anchor && + e->pnum < UBI_FM_MAX_START) + e = rb_entry(rb_next(root->rb_node), + struct ubi_wl_entry, u.rb); + + return e; +} diff --git a/drivers/mtd/ubi/fastmap.c b/drivers/mtd/ubi/fastmap.c new file mode 100644 index 000000000..3f9fb7c18 --- /dev/null +++ b/drivers/mtd/ubi/fastmap.c @@ -0,0 +1,1648 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * Copyright (c) 2012 Linutronix GmbH + * Copyright (c) 2014 sigma star gmbh + * Author: Richard Weinberger + */ + +#include +#include +#include "ubi.h" + +/** + * init_seen - allocate memory for used for debugging. + * @ubi: UBI device description object + */ +static inline unsigned long *init_seen(struct ubi_device *ubi) +{ + unsigned long *ret; + + if (!ubi_dbg_chk_fastmap(ubi)) + return NULL; + + ret = bitmap_zalloc(ubi->peb_count, GFP_NOFS); + if (!ret) + return ERR_PTR(-ENOMEM); + + return ret; +} + +/** + * free_seen - free the seen logic integer array. + * @seen: integer array of @ubi->peb_count size + */ +static inline void free_seen(unsigned long *seen) +{ + bitmap_free(seen); +} + +/** + * set_seen - mark a PEB as seen. + * @ubi: UBI device description object + * @pnum: The PEB to be makred as seen + * @seen: integer array of @ubi->peb_count size + */ +static inline void set_seen(struct ubi_device *ubi, int pnum, unsigned long *seen) +{ + if (!ubi_dbg_chk_fastmap(ubi) || !seen) + return; + + set_bit(pnum, seen); +} + +/** + * self_check_seen - check whether all PEB have been seen by fastmap. + * @ubi: UBI device description object + * @seen: integer array of @ubi->peb_count size + */ +static int self_check_seen(struct ubi_device *ubi, unsigned long *seen) +{ + int pnum, ret = 0; + + if (!ubi_dbg_chk_fastmap(ubi) || !seen) + return 0; + + for (pnum = 0; pnum < ubi->peb_count; pnum++) { + if (!test_bit(pnum, seen) && ubi->lookuptbl[pnum]) { + ubi_err(ubi, "self-check failed for PEB %d, fastmap didn't see it", pnum); + ret = -EINVAL; + } + } + + return ret; +} + +/** + * ubi_calc_fm_size - calculates the fastmap size in bytes for an UBI device. + * @ubi: UBI device description object + */ +size_t ubi_calc_fm_size(struct ubi_device *ubi) +{ + size_t size; + + size = sizeof(struct ubi_fm_sb) + + sizeof(struct ubi_fm_hdr) + + sizeof(struct ubi_fm_scan_pool) + + sizeof(struct ubi_fm_scan_pool) + + (ubi->peb_count * sizeof(struct ubi_fm_ec)) + + ((sizeof(struct ubi_fm_eba) + + sizeof(struct ubi_fm_volhdr)) * + (UBI_MAX_VOLUMES + UBI_INT_VOL_COUNT)) + + (ubi->peb_count * sizeof(__be32)); + return roundup(size, ubi->leb_size); +} + + +/** + * new_fm_vbuf() - allocate a new volume header for fastmap usage. + * @ubi: UBI device description object + * @vol_id: the VID of the new header + * + * Returns a new struct ubi_vid_hdr on success. + * NULL indicates out of memory. + */ +static struct ubi_vid_io_buf *new_fm_vbuf(struct ubi_device *ubi, int vol_id) +{ + struct ubi_vid_io_buf *new; + struct ubi_vid_hdr *vh; + + new = ubi_alloc_vid_buf(ubi, GFP_NOFS); + if (!new) + goto out; + + vh = ubi_get_vid_hdr(new); + vh->vol_type = UBI_VID_DYNAMIC; + vh->vol_id = cpu_to_be32(vol_id); + + /* UBI implementations without fastmap support have to delete the + * fastmap. + */ + vh->compat = UBI_COMPAT_DELETE; + +out: + return new; +} + +/** + * add_aeb - create and add a attach erase block to a given list. + * @ai: UBI attach info object + * @list: the target list + * @pnum: PEB number of the new attach erase block + * @ec: erease counter of the new LEB + * @scrub: scrub this PEB after attaching + * + * Returns 0 on success, < 0 indicates an internal error. + */ +static int add_aeb(struct ubi_attach_info *ai, struct list_head *list, + int pnum, int ec, int scrub) +{ + struct ubi_ainf_peb *aeb; + + aeb = ubi_alloc_aeb(ai, pnum, ec); + if (!aeb) + return -ENOMEM; + + aeb->lnum = -1; + aeb->scrub = scrub; + aeb->copy_flag = aeb->sqnum = 0; + + ai->ec_sum += aeb->ec; + ai->ec_count++; + + if (ai->max_ec < aeb->ec) + ai->max_ec = aeb->ec; + + if (ai->min_ec > aeb->ec) + ai->min_ec = aeb->ec; + + list_add_tail(&aeb->u.list, list); + + return 0; +} + +/** + * add_vol - create and add a new volume to ubi_attach_info. + * @ai: ubi_attach_info object + * @vol_id: VID of the new volume + * @used_ebs: number of used EBS + * @data_pad: data padding value of the new volume + * @vol_type: volume type + * @last_eb_bytes: number of bytes in the last LEB + * + * Returns the new struct ubi_ainf_volume on success. + * NULL indicates an error. + */ +static struct ubi_ainf_volume *add_vol(struct ubi_attach_info *ai, int vol_id, + int used_ebs, int data_pad, u8 vol_type, + int last_eb_bytes) +{ + struct ubi_ainf_volume *av; + + av = ubi_add_av(ai, vol_id); + if (IS_ERR(av)) + return av; + + av->data_pad = data_pad; + av->last_data_size = last_eb_bytes; + av->compat = 0; + av->vol_type = vol_type; + if (av->vol_type == UBI_STATIC_VOLUME) + av->used_ebs = used_ebs; + + dbg_bld("found volume (ID %i)", vol_id); + return av; +} + +/** + * assign_aeb_to_av - assigns a SEB to a given ainf_volume and removes it + * from it's original list. + * @ai: ubi_attach_info object + * @aeb: the to be assigned SEB + * @av: target scan volume + */ +static void assign_aeb_to_av(struct ubi_attach_info *ai, + struct ubi_ainf_peb *aeb, + struct ubi_ainf_volume *av) +{ + struct ubi_ainf_peb *tmp_aeb; + struct rb_node **p = &av->root.rb_node, *parent = NULL; + + while (*p) { + parent = *p; + + tmp_aeb = rb_entry(parent, struct ubi_ainf_peb, u.rb); + if (aeb->lnum != tmp_aeb->lnum) { + if (aeb->lnum < tmp_aeb->lnum) + p = &(*p)->rb_left; + else + p = &(*p)->rb_right; + + continue; + } else + break; + } + + list_del(&aeb->u.list); + av->leb_count++; + + rb_link_node(&aeb->u.rb, parent, p); + rb_insert_color(&aeb->u.rb, &av->root); +} + +/** + * update_vol - inserts or updates a LEB which was found a pool. + * @ubi: the UBI device object + * @ai: attach info object + * @av: the volume this LEB belongs to + * @new_vh: the volume header derived from new_aeb + * @new_aeb: the AEB to be examined + * + * Returns 0 on success, < 0 indicates an internal error. + */ +static int update_vol(struct ubi_device *ubi, struct ubi_attach_info *ai, + struct ubi_ainf_volume *av, struct ubi_vid_hdr *new_vh, + struct ubi_ainf_peb *new_aeb) +{ + struct rb_node **p = &av->root.rb_node, *parent = NULL; + struct ubi_ainf_peb *aeb, *victim; + int cmp_res; + + while (*p) { + parent = *p; + aeb = rb_entry(parent, struct ubi_ainf_peb, u.rb); + + if (be32_to_cpu(new_vh->lnum) != aeb->lnum) { + if (be32_to_cpu(new_vh->lnum) < aeb->lnum) + p = &(*p)->rb_left; + else + p = &(*p)->rb_right; + + continue; + } + + /* This case can happen if the fastmap gets written + * because of a volume change (creation, deletion, ..). + * Then a PEB can be within the persistent EBA and the pool. + */ + if (aeb->pnum == new_aeb->pnum) { + ubi_assert(aeb->lnum == new_aeb->lnum); + ubi_free_aeb(ai, new_aeb); + + return 0; + } + + cmp_res = ubi_compare_lebs(ubi, aeb, new_aeb->pnum, new_vh); + if (cmp_res < 0) + return cmp_res; + + /* new_aeb is newer */ + if (cmp_res & 1) { + victim = ubi_alloc_aeb(ai, aeb->pnum, aeb->ec); + if (!victim) + return -ENOMEM; + + list_add_tail(&victim->u.list, &ai->erase); + + if (av->highest_lnum == be32_to_cpu(new_vh->lnum)) + av->last_data_size = + be32_to_cpu(new_vh->data_size); + + dbg_bld("vol %i: AEB %i's PEB %i is the newer", + av->vol_id, aeb->lnum, new_aeb->pnum); + + aeb->ec = new_aeb->ec; + aeb->pnum = new_aeb->pnum; + aeb->copy_flag = new_vh->copy_flag; + aeb->scrub = new_aeb->scrub; + aeb->sqnum = new_aeb->sqnum; + ubi_free_aeb(ai, new_aeb); + + /* new_aeb is older */ + } else { + dbg_bld("vol %i: AEB %i's PEB %i is old, dropping it", + av->vol_id, aeb->lnum, new_aeb->pnum); + list_add_tail(&new_aeb->u.list, &ai->erase); + } + + return 0; + } + /* This LEB is new, let's add it to the volume */ + + if (av->highest_lnum <= be32_to_cpu(new_vh->lnum)) { + av->highest_lnum = be32_to_cpu(new_vh->lnum); + av->last_data_size = be32_to_cpu(new_vh->data_size); + } + + if (av->vol_type == UBI_STATIC_VOLUME) + av->used_ebs = be32_to_cpu(new_vh->used_ebs); + + av->leb_count++; + + rb_link_node(&new_aeb->u.rb, parent, p); + rb_insert_color(&new_aeb->u.rb, &av->root); + + return 0; +} + +/** + * process_pool_aeb - we found a non-empty PEB in a pool. + * @ubi: UBI device object + * @ai: attach info object + * @new_vh: the volume header derived from new_aeb + * @new_aeb: the AEB to be examined + * + * Returns 0 on success, < 0 indicates an internal error. + */ +static int process_pool_aeb(struct ubi_device *ubi, struct ubi_attach_info *ai, + struct ubi_vid_hdr *new_vh, + struct ubi_ainf_peb *new_aeb) +{ + int vol_id = be32_to_cpu(new_vh->vol_id); + struct ubi_ainf_volume *av; + + if (vol_id == UBI_FM_SB_VOLUME_ID || vol_id == UBI_FM_DATA_VOLUME_ID) { + ubi_free_aeb(ai, new_aeb); + + return 0; + } + + /* Find the volume this SEB belongs to */ + av = ubi_find_av(ai, vol_id); + if (!av) { + ubi_err(ubi, "orphaned volume in fastmap pool!"); + ubi_free_aeb(ai, new_aeb); + return UBI_BAD_FASTMAP; + } + + ubi_assert(vol_id == av->vol_id); + + return update_vol(ubi, ai, av, new_vh, new_aeb); +} + +/** + * unmap_peb - unmap a PEB. + * If fastmap detects a free PEB in the pool it has to check whether + * this PEB has been unmapped after writing the fastmap. + * + * @ai: UBI attach info object + * @pnum: The PEB to be unmapped + */ +static void unmap_peb(struct ubi_attach_info *ai, int pnum) +{ + struct ubi_ainf_volume *av; + struct rb_node *node, *node2; + struct ubi_ainf_peb *aeb; + + ubi_rb_for_each_entry(node, av, &ai->volumes, rb) { + ubi_rb_for_each_entry(node2, aeb, &av->root, u.rb) { + if (aeb->pnum == pnum) { + rb_erase(&aeb->u.rb, &av->root); + av->leb_count--; + ubi_free_aeb(ai, aeb); + return; + } + } + } +} + +/** + * scan_pool - scans a pool for changed (no longer empty PEBs). + * @ubi: UBI device object + * @ai: attach info object + * @pebs: an array of all PEB numbers in the to be scanned pool + * @pool_size: size of the pool (number of entries in @pebs) + * @max_sqnum: pointer to the maximal sequence number + * @free: list of PEBs which are most likely free (and go into @ai->free) + * + * Returns 0 on success, if the pool is unusable UBI_BAD_FASTMAP is returned. + * < 0 indicates an internal error. + */ +static int scan_pool(struct ubi_device *ubi, struct ubi_attach_info *ai, + __be32 *pebs, int pool_size, unsigned long long *max_sqnum, + struct list_head *free) +{ + struct ubi_vid_io_buf *vb; + struct ubi_vid_hdr *vh; + struct ubi_ec_hdr *ech; + struct ubi_ainf_peb *new_aeb; + int i, pnum, err, ret = 0; + + ech = kzalloc(ubi->ec_hdr_alsize, GFP_KERNEL); + if (!ech) + return -ENOMEM; + + vb = ubi_alloc_vid_buf(ubi, GFP_KERNEL); + if (!vb) { + kfree(ech); + return -ENOMEM; + } + + vh = ubi_get_vid_hdr(vb); + + dbg_bld("scanning fastmap pool: size = %i", pool_size); + + /* + * Now scan all PEBs in the pool to find changes which have been made + * after the creation of the fastmap + */ + for (i = 0; i < pool_size; i++) { + int scrub = 0; + int image_seq; + + pnum = be32_to_cpu(pebs[i]); + + if (ubi_io_is_bad(ubi, pnum)) { + ubi_err(ubi, "bad PEB in fastmap pool!"); + ret = UBI_BAD_FASTMAP; + goto out; + } + + err = ubi_io_read_ec_hdr(ubi, pnum, ech, 0); + if (err && err != UBI_IO_BITFLIPS) { + ubi_err(ubi, "unable to read EC header! PEB:%i err:%i", + pnum, err); + ret = err > 0 ? UBI_BAD_FASTMAP : err; + goto out; + } else if (err == UBI_IO_BITFLIPS) + scrub = 1; + + /* + * Older UBI implementations have image_seq set to zero, so + * we shouldn't fail if image_seq == 0. + */ + image_seq = be32_to_cpu(ech->image_seq); + + if (image_seq && (image_seq != ubi->image_seq)) { + ubi_err(ubi, "bad image seq: 0x%x, expected: 0x%x", + be32_to_cpu(ech->image_seq), ubi->image_seq); + ret = UBI_BAD_FASTMAP; + goto out; + } + + err = ubi_io_read_vid_hdr(ubi, pnum, vb, 0); + if (err == UBI_IO_FF || err == UBI_IO_FF_BITFLIPS) { + unsigned long long ec = be64_to_cpu(ech->ec); + unmap_peb(ai, pnum); + dbg_bld("Adding PEB to free: %i", pnum); + + if (err == UBI_IO_FF_BITFLIPS) + scrub = 1; + + ret = add_aeb(ai, free, pnum, ec, scrub); + if (ret) + goto out; + continue; + } else if (err == 0 || err == UBI_IO_BITFLIPS) { + dbg_bld("Found non empty PEB:%i in pool", pnum); + + if (err == UBI_IO_BITFLIPS) + scrub = 1; + + new_aeb = ubi_alloc_aeb(ai, pnum, be64_to_cpu(ech->ec)); + if (!new_aeb) { + ret = -ENOMEM; + goto out; + } + + new_aeb->lnum = be32_to_cpu(vh->lnum); + new_aeb->sqnum = be64_to_cpu(vh->sqnum); + new_aeb->copy_flag = vh->copy_flag; + new_aeb->scrub = scrub; + + if (*max_sqnum < new_aeb->sqnum) + *max_sqnum = new_aeb->sqnum; + + err = process_pool_aeb(ubi, ai, vh, new_aeb); + if (err) { + ret = err > 0 ? UBI_BAD_FASTMAP : err; + goto out; + } + } else { + /* We are paranoid and fall back to scanning mode */ + ubi_err(ubi, "fastmap pool PEBs contains damaged PEBs!"); + ret = err > 0 ? UBI_BAD_FASTMAP : err; + goto out; + } + + } + +out: + ubi_free_vid_buf(vb); + kfree(ech); + return ret; +} + +/** + * count_fastmap_pebs - Counts the PEBs found by fastmap. + * @ai: The UBI attach info object + */ +static int count_fastmap_pebs(struct ubi_attach_info *ai) +{ + struct ubi_ainf_peb *aeb; + struct ubi_ainf_volume *av; + struct rb_node *rb1, *rb2; + int n = 0; + + list_for_each_entry(aeb, &ai->erase, u.list) + n++; + + list_for_each_entry(aeb, &ai->free, u.list) + n++; + + ubi_rb_for_each_entry(rb1, av, &ai->volumes, rb) + ubi_rb_for_each_entry(rb2, aeb, &av->root, u.rb) + n++; + + return n; +} + +/** + * ubi_attach_fastmap - creates ubi_attach_info from a fastmap. + * @ubi: UBI device object + * @ai: UBI attach info object + * @fm: the fastmap to be attached + * + * Returns 0 on success, UBI_BAD_FASTMAP if the found fastmap was unusable. + * < 0 indicates an internal error. + */ +static int ubi_attach_fastmap(struct ubi_device *ubi, + struct ubi_attach_info *ai, + struct ubi_fastmap_layout *fm) +{ + struct list_head used, free; + struct ubi_ainf_volume *av; + struct ubi_ainf_peb *aeb, *tmp_aeb, *_tmp_aeb; + struct ubi_fm_sb *fmsb; + struct ubi_fm_hdr *fmhdr; + struct ubi_fm_scan_pool *fmpl, *fmpl_wl; + struct ubi_fm_ec *fmec; + struct ubi_fm_volhdr *fmvhdr; + struct ubi_fm_eba *fm_eba; + int ret, i, j, pool_size, wl_pool_size; + size_t fm_pos = 0, fm_size = ubi->fm_size; + unsigned long long max_sqnum = 0; + void *fm_raw = ubi->fm_buf; + + INIT_LIST_HEAD(&used); + INIT_LIST_HEAD(&free); + ai->min_ec = UBI_MAX_ERASECOUNTER; + + fmsb = (struct ubi_fm_sb *)(fm_raw); + ai->max_sqnum = fmsb->sqnum; + fm_pos += sizeof(struct ubi_fm_sb); + if (fm_pos >= fm_size) + goto fail_bad; + + fmhdr = (struct ubi_fm_hdr *)(fm_raw + fm_pos); + fm_pos += sizeof(*fmhdr); + if (fm_pos >= fm_size) + goto fail_bad; + + if (be32_to_cpu(fmhdr->magic) != UBI_FM_HDR_MAGIC) { + ubi_err(ubi, "bad fastmap header magic: 0x%x, expected: 0x%x", + be32_to_cpu(fmhdr->magic), UBI_FM_HDR_MAGIC); + goto fail_bad; + } + + fmpl = (struct ubi_fm_scan_pool *)(fm_raw + fm_pos); + fm_pos += sizeof(*fmpl); + if (fm_pos >= fm_size) + goto fail_bad; + if (be32_to_cpu(fmpl->magic) != UBI_FM_POOL_MAGIC) { + ubi_err(ubi, "bad fastmap pool magic: 0x%x, expected: 0x%x", + be32_to_cpu(fmpl->magic), UBI_FM_POOL_MAGIC); + goto fail_bad; + } + + fmpl_wl = (struct ubi_fm_scan_pool *)(fm_raw + fm_pos); + fm_pos += sizeof(*fmpl_wl); + if (fm_pos >= fm_size) + goto fail_bad; + if (be32_to_cpu(fmpl_wl->magic) != UBI_FM_POOL_MAGIC) { + ubi_err(ubi, "bad fastmap WL pool magic: 0x%x, expected: 0x%x", + be32_to_cpu(fmpl_wl->magic), UBI_FM_POOL_MAGIC); + goto fail_bad; + } + + pool_size = be16_to_cpu(fmpl->size); + wl_pool_size = be16_to_cpu(fmpl_wl->size); + fm->max_pool_size = be16_to_cpu(fmpl->max_size); + fm->max_wl_pool_size = be16_to_cpu(fmpl_wl->max_size); + + if (pool_size > UBI_FM_MAX_POOL_SIZE || pool_size < 0) { + ubi_err(ubi, "bad pool size: %i", pool_size); + goto fail_bad; + } + + if (wl_pool_size > UBI_FM_MAX_POOL_SIZE || wl_pool_size < 0) { + ubi_err(ubi, "bad WL pool size: %i", wl_pool_size); + goto fail_bad; + } + + + if (fm->max_pool_size > UBI_FM_MAX_POOL_SIZE || + fm->max_pool_size < 0) { + ubi_err(ubi, "bad maximal pool size: %i", fm->max_pool_size); + goto fail_bad; + } + + if (fm->max_wl_pool_size > UBI_FM_MAX_POOL_SIZE || + fm->max_wl_pool_size < 0) { + ubi_err(ubi, "bad maximal WL pool size: %i", + fm->max_wl_pool_size); + goto fail_bad; + } + + /* read EC values from free list */ + for (i = 0; i < be32_to_cpu(fmhdr->free_peb_count); i++) { + fmec = (struct ubi_fm_ec *)(fm_raw + fm_pos); + fm_pos += sizeof(*fmec); + if (fm_pos >= fm_size) + goto fail_bad; + + ret = add_aeb(ai, &ai->free, be32_to_cpu(fmec->pnum), + be32_to_cpu(fmec->ec), 0); + if (ret) + goto fail; + } + + /* read EC values from used list */ + for (i = 0; i < be32_to_cpu(fmhdr->used_peb_count); i++) { + fmec = (struct ubi_fm_ec *)(fm_raw + fm_pos); + fm_pos += sizeof(*fmec); + if (fm_pos >= fm_size) + goto fail_bad; + + ret = add_aeb(ai, &used, be32_to_cpu(fmec->pnum), + be32_to_cpu(fmec->ec), 0); + if (ret) + goto fail; + } + + /* read EC values from scrub list */ + for (i = 0; i < be32_to_cpu(fmhdr->scrub_peb_count); i++) { + fmec = (struct ubi_fm_ec *)(fm_raw + fm_pos); + fm_pos += sizeof(*fmec); + if (fm_pos >= fm_size) + goto fail_bad; + + ret = add_aeb(ai, &used, be32_to_cpu(fmec->pnum), + be32_to_cpu(fmec->ec), 1); + if (ret) + goto fail; + } + + /* read EC values from erase list */ + for (i = 0; i < be32_to_cpu(fmhdr->erase_peb_count); i++) { + fmec = (struct ubi_fm_ec *)(fm_raw + fm_pos); + fm_pos += sizeof(*fmec); + if (fm_pos >= fm_size) + goto fail_bad; + + ret = add_aeb(ai, &ai->erase, be32_to_cpu(fmec->pnum), + be32_to_cpu(fmec->ec), 1); + if (ret) + goto fail; + } + + ai->mean_ec = div_u64(ai->ec_sum, ai->ec_count); + ai->bad_peb_count = be32_to_cpu(fmhdr->bad_peb_count); + + /* Iterate over all volumes and read their EBA table */ + for (i = 0; i < be32_to_cpu(fmhdr->vol_count); i++) { + fmvhdr = (struct ubi_fm_volhdr *)(fm_raw + fm_pos); + fm_pos += sizeof(*fmvhdr); + if (fm_pos >= fm_size) + goto fail_bad; + + if (be32_to_cpu(fmvhdr->magic) != UBI_FM_VHDR_MAGIC) { + ubi_err(ubi, "bad fastmap vol header magic: 0x%x, expected: 0x%x", + be32_to_cpu(fmvhdr->magic), UBI_FM_VHDR_MAGIC); + goto fail_bad; + } + + av = add_vol(ai, be32_to_cpu(fmvhdr->vol_id), + be32_to_cpu(fmvhdr->used_ebs), + be32_to_cpu(fmvhdr->data_pad), + fmvhdr->vol_type, + be32_to_cpu(fmvhdr->last_eb_bytes)); + + if (IS_ERR(av)) { + if (PTR_ERR(av) == -EEXIST) + ubi_err(ubi, "volume (ID %i) already exists", + fmvhdr->vol_id); + + goto fail_bad; + } + + ai->vols_found++; + if (ai->highest_vol_id < be32_to_cpu(fmvhdr->vol_id)) + ai->highest_vol_id = be32_to_cpu(fmvhdr->vol_id); + + fm_eba = (struct ubi_fm_eba *)(fm_raw + fm_pos); + fm_pos += sizeof(*fm_eba); + fm_pos += (sizeof(__be32) * be32_to_cpu(fm_eba->reserved_pebs)); + if (fm_pos >= fm_size) + goto fail_bad; + + if (be32_to_cpu(fm_eba->magic) != UBI_FM_EBA_MAGIC) { + ubi_err(ubi, "bad fastmap EBA header magic: 0x%x, expected: 0x%x", + be32_to_cpu(fm_eba->magic), UBI_FM_EBA_MAGIC); + goto fail_bad; + } + + for (j = 0; j < be32_to_cpu(fm_eba->reserved_pebs); j++) { + int pnum = be32_to_cpu(fm_eba->pnum[j]); + + if (pnum < 0) + continue; + + aeb = NULL; + list_for_each_entry(tmp_aeb, &used, u.list) { + if (tmp_aeb->pnum == pnum) { + aeb = tmp_aeb; + break; + } + } + + if (!aeb) { + ubi_err(ubi, "PEB %i is in EBA but not in used list", pnum); + goto fail_bad; + } + + aeb->lnum = j; + + if (av->highest_lnum <= aeb->lnum) + av->highest_lnum = aeb->lnum; + + assign_aeb_to_av(ai, aeb, av); + + dbg_bld("inserting PEB:%i (LEB %i) to vol %i", + aeb->pnum, aeb->lnum, av->vol_id); + } + } + + ret = scan_pool(ubi, ai, fmpl->pebs, pool_size, &max_sqnum, &free); + if (ret) + goto fail; + + ret = scan_pool(ubi, ai, fmpl_wl->pebs, wl_pool_size, &max_sqnum, &free); + if (ret) + goto fail; + + if (max_sqnum > ai->max_sqnum) + ai->max_sqnum = max_sqnum; + + list_for_each_entry_safe(tmp_aeb, _tmp_aeb, &free, u.list) + list_move_tail(&tmp_aeb->u.list, &ai->free); + + list_for_each_entry_safe(tmp_aeb, _tmp_aeb, &used, u.list) + list_move_tail(&tmp_aeb->u.list, &ai->erase); + + ubi_assert(list_empty(&free)); + + /* + * If fastmap is leaking PEBs (must not happen), raise a + * fat warning and fall back to scanning mode. + * We do this here because in ubi_wl_init() it's too late + * and we cannot fall back to scanning. + */ + if (WARN_ON(count_fastmap_pebs(ai) != ubi->peb_count - + ai->bad_peb_count - fm->used_blocks)) + goto fail_bad; + + return 0; + +fail_bad: + ret = UBI_BAD_FASTMAP; +fail: + list_for_each_entry_safe(tmp_aeb, _tmp_aeb, &used, u.list) { + list_del(&tmp_aeb->u.list); + ubi_free_aeb(ai, tmp_aeb); + } + list_for_each_entry_safe(tmp_aeb, _tmp_aeb, &free, u.list) { + list_del(&tmp_aeb->u.list); + ubi_free_aeb(ai, tmp_aeb); + } + + return ret; +} + +/** + * find_fm_anchor - find the most recent Fastmap superblock (anchor) + * @ai: UBI attach info to be filled + */ +static int find_fm_anchor(struct ubi_attach_info *ai) +{ + int ret = -1; + struct ubi_ainf_peb *aeb; + unsigned long long max_sqnum = 0; + + list_for_each_entry(aeb, &ai->fastmap, u.list) { + if (aeb->vol_id == UBI_FM_SB_VOLUME_ID && aeb->sqnum > max_sqnum) { + max_sqnum = aeb->sqnum; + ret = aeb->pnum; + } + } + + return ret; +} + +static struct ubi_ainf_peb *clone_aeb(struct ubi_attach_info *ai, + struct ubi_ainf_peb *old) +{ + struct ubi_ainf_peb *new; + + new = ubi_alloc_aeb(ai, old->pnum, old->ec); + if (!new) + return NULL; + + new->vol_id = old->vol_id; + new->sqnum = old->sqnum; + new->lnum = old->lnum; + new->scrub = old->scrub; + new->copy_flag = old->copy_flag; + + return new; +} + +/** + * ubi_scan_fastmap - scan the fastmap. + * @ubi: UBI device object + * @ai: UBI attach info to be filled + * @scan_ai: UBI attach info from the first 64 PEBs, + * used to find the most recent Fastmap data structure + * + * Returns 0 on success, UBI_NO_FASTMAP if no fastmap was found, + * UBI_BAD_FASTMAP if one was found but is not usable. + * < 0 indicates an internal error. + */ +int ubi_scan_fastmap(struct ubi_device *ubi, struct ubi_attach_info *ai, + struct ubi_attach_info *scan_ai) +{ + struct ubi_fm_sb *fmsb, *fmsb2; + struct ubi_vid_io_buf *vb; + struct ubi_vid_hdr *vh; + struct ubi_ec_hdr *ech; + struct ubi_fastmap_layout *fm; + struct ubi_ainf_peb *aeb; + int i, used_blocks, pnum, fm_anchor, ret = 0; + size_t fm_size; + __be32 crc, tmp_crc; + unsigned long long sqnum = 0; + + fm_anchor = find_fm_anchor(scan_ai); + if (fm_anchor < 0) + return UBI_NO_FASTMAP; + + /* Copy all (possible) fastmap blocks into our new attach structure. */ + list_for_each_entry(aeb, &scan_ai->fastmap, u.list) { + struct ubi_ainf_peb *new; + + new = clone_aeb(ai, aeb); + if (!new) + return -ENOMEM; + + list_add(&new->u.list, &ai->fastmap); + } + + down_write(&ubi->fm_protect); + memset(ubi->fm_buf, 0, ubi->fm_size); + + fmsb = kmalloc_obj(*fmsb); + if (!fmsb) { + ret = -ENOMEM; + goto out; + } + + fm = kzalloc_obj(*fm); + if (!fm) { + ret = -ENOMEM; + kfree(fmsb); + goto out; + } + + ret = ubi_io_read_data(ubi, fmsb, fm_anchor, 0, sizeof(*fmsb)); + if (ret && ret != UBI_IO_BITFLIPS) + goto free_fm_sb; + else if (ret == UBI_IO_BITFLIPS) + fm->to_be_tortured[0] = 1; + + if (be32_to_cpu(fmsb->magic) != UBI_FM_SB_MAGIC) { + ubi_err(ubi, "bad super block magic: 0x%x, expected: 0x%x", + be32_to_cpu(fmsb->magic), UBI_FM_SB_MAGIC); + ret = UBI_BAD_FASTMAP; + goto free_fm_sb; + } + + if (fmsb->version != UBI_FM_FMT_VERSION) { + ubi_err(ubi, "bad fastmap version: %i, expected: %i", + fmsb->version, UBI_FM_FMT_VERSION); + ret = UBI_BAD_FASTMAP; + goto free_fm_sb; + } + + used_blocks = be32_to_cpu(fmsb->used_blocks); + if (used_blocks > UBI_FM_MAX_BLOCKS || used_blocks < 1) { + ubi_err(ubi, "number of fastmap blocks is invalid: %i", + used_blocks); + ret = UBI_BAD_FASTMAP; + goto free_fm_sb; + } + + fm_size = ubi->leb_size * used_blocks; + if (fm_size != ubi->fm_size) { + ubi_err(ubi, "bad fastmap size: %zi, expected: %zi", + fm_size, ubi->fm_size); + ret = UBI_BAD_FASTMAP; + goto free_fm_sb; + } + + ech = kzalloc(ubi->ec_hdr_alsize, GFP_KERNEL); + if (!ech) { + ret = -ENOMEM; + goto free_fm_sb; + } + + vb = ubi_alloc_vid_buf(ubi, GFP_KERNEL); + if (!vb) { + ret = -ENOMEM; + goto free_hdr; + } + + vh = ubi_get_vid_hdr(vb); + + for (i = 0; i < used_blocks; i++) { + int image_seq; + + pnum = be32_to_cpu(fmsb->block_loc[i]); + + if (ubi_io_is_bad(ubi, pnum)) { + ret = UBI_BAD_FASTMAP; + goto free_hdr; + } + + if (i == 0 && pnum != fm_anchor) { + ubi_err(ubi, "Fastmap anchor PEB mismatch: PEB: %i vs. %i", + pnum, fm_anchor); + ret = UBI_BAD_FASTMAP; + goto free_hdr; + } + + ret = ubi_io_read_ec_hdr(ubi, pnum, ech, 0); + if (ret && ret != UBI_IO_BITFLIPS) { + ubi_err(ubi, "unable to read fastmap block# %i EC (PEB: %i)", + i, pnum); + if (ret > 0) + ret = UBI_BAD_FASTMAP; + goto free_hdr; + } else if (ret == UBI_IO_BITFLIPS) + fm->to_be_tortured[i] = 1; + + image_seq = be32_to_cpu(ech->image_seq); + if (!ubi->image_seq) + ubi->image_seq = image_seq; + + /* + * Older UBI implementations have image_seq set to zero, so + * we shouldn't fail if image_seq == 0. + */ + if (image_seq && (image_seq != ubi->image_seq)) { + ubi_err(ubi, "wrong image seq:%d instead of %d", + be32_to_cpu(ech->image_seq), ubi->image_seq); + ret = UBI_BAD_FASTMAP; + goto free_hdr; + } + + ret = ubi_io_read_vid_hdr(ubi, pnum, vb, 0); + if (ret && ret != UBI_IO_BITFLIPS) { + ubi_err(ubi, "unable to read fastmap block# %i (PEB: %i)", + i, pnum); + goto free_hdr; + } + + if (i == 0) { + if (be32_to_cpu(vh->vol_id) != UBI_FM_SB_VOLUME_ID) { + ubi_err(ubi, "bad fastmap anchor vol_id: 0x%x, expected: 0x%x", + be32_to_cpu(vh->vol_id), + UBI_FM_SB_VOLUME_ID); + ret = UBI_BAD_FASTMAP; + goto free_hdr; + } + } else { + if (be32_to_cpu(vh->vol_id) != UBI_FM_DATA_VOLUME_ID) { + ubi_err(ubi, "bad fastmap data vol_id: 0x%x, expected: 0x%x", + be32_to_cpu(vh->vol_id), + UBI_FM_DATA_VOLUME_ID); + ret = UBI_BAD_FASTMAP; + goto free_hdr; + } + } + + if (sqnum < be64_to_cpu(vh->sqnum)) + sqnum = be64_to_cpu(vh->sqnum); + + ret = ubi_io_read_data(ubi, ubi->fm_buf + (ubi->leb_size * i), + pnum, 0, ubi->leb_size); + if (ret && ret != UBI_IO_BITFLIPS) { + ubi_err(ubi, "unable to read fastmap block# %i (PEB: %i, " + "err: %i)", i, pnum, ret); + goto free_hdr; + } + } + + kfree(fmsb); + fmsb = NULL; + + fmsb2 = (struct ubi_fm_sb *)(ubi->fm_buf); + tmp_crc = be32_to_cpu(fmsb2->data_crc); + fmsb2->data_crc = 0; + crc = crc32(UBI_CRC32_INIT, ubi->fm_buf, fm_size); + if (crc != tmp_crc) { + ubi_err(ubi, "fastmap data CRC is invalid"); + ubi_err(ubi, "CRC should be: 0x%x, calc: 0x%x", + tmp_crc, crc); + ret = UBI_BAD_FASTMAP; + goto free_hdr; + } + + fmsb2->sqnum = sqnum; + + fm->used_blocks = used_blocks; + + ret = ubi_attach_fastmap(ubi, ai, fm); + if (ret) { + if (ret > 0) + ret = UBI_BAD_FASTMAP; + goto free_hdr; + } + + for (i = 0; i < used_blocks; i++) { + struct ubi_wl_entry *e; + + e = kmem_cache_alloc(ubi_wl_entry_slab, GFP_KERNEL); + if (!e) { + while (i--) + kmem_cache_free(ubi_wl_entry_slab, fm->e[i]); + + ret = -ENOMEM; + goto free_hdr; + } + + e->pnum = be32_to_cpu(fmsb2->block_loc[i]); + e->ec = be32_to_cpu(fmsb2->block_ec[i]); + fm->e[i] = e; + } + + ubi->fm = fm; + ubi->fm_pool.max_size = ubi->fm->max_pool_size; + ubi->fm_wl_pool.max_size = ubi->fm->max_wl_pool_size; + ubi_msg(ubi, "attached by fastmap"); + ubi_msg(ubi, "fastmap pool size: %d", ubi->fm_pool.max_size); + ubi_msg(ubi, "fastmap WL pool size: %d", + ubi->fm_wl_pool.max_size); + ubi->fm_disabled = 0; + ubi->fast_attach = 1; + + ubi_free_vid_buf(vb); + kfree(ech); +out: + up_write(&ubi->fm_protect); + if (ret == UBI_BAD_FASTMAP) + ubi_err(ubi, "Attach by fastmap failed, doing a full scan!"); + return ret; + +free_hdr: + ubi_free_vid_buf(vb); + kfree(ech); +free_fm_sb: + kfree(fmsb); + kfree(fm); + goto out; +} + +int ubi_fastmap_init_checkmap(struct ubi_volume *vol, int leb_count) +{ + struct ubi_device *ubi = vol->ubi; + + if (!ubi->fast_attach) + return 0; + + vol->checkmap = bitmap_zalloc(leb_count, GFP_KERNEL); + if (!vol->checkmap) + return -ENOMEM; + + return 0; +} + +void ubi_fastmap_destroy_checkmap(struct ubi_volume *vol) +{ + bitmap_free(vol->checkmap); +} + +/** + * ubi_write_fastmap - writes a fastmap. + * @ubi: UBI device object + * @new_fm: the to be written fastmap + * + * Returns 0 on success, < 0 indicates an internal error. + */ +static int ubi_write_fastmap(struct ubi_device *ubi, + struct ubi_fastmap_layout *new_fm) +{ + size_t fm_pos = 0; + void *fm_raw; + struct ubi_fm_sb *fmsb; + struct ubi_fm_hdr *fmh; + struct ubi_fm_scan_pool *fmpl, *fmpl_wl; + struct ubi_fm_ec *fec; + struct ubi_fm_volhdr *fvh; + struct ubi_fm_eba *feba; + struct ubi_wl_entry *wl_e; + struct ubi_volume *vol; + struct ubi_vid_io_buf *avbuf, *dvbuf; + struct ubi_vid_hdr *avhdr, *dvhdr; + struct ubi_work *ubi_wrk; + struct rb_node *tmp_rb; + int ret, i, j, free_peb_count, used_peb_count, vol_count; + int scrub_peb_count, erase_peb_count; + unsigned long *seen_pebs; + + fm_raw = ubi->fm_buf; + memset(ubi->fm_buf, 0, ubi->fm_size); + + avbuf = new_fm_vbuf(ubi, UBI_FM_SB_VOLUME_ID); + if (!avbuf) { + ret = -ENOMEM; + goto out; + } + + dvbuf = new_fm_vbuf(ubi, UBI_FM_DATA_VOLUME_ID); + if (!dvbuf) { + ret = -ENOMEM; + goto out_free_avbuf; + } + + avhdr = ubi_get_vid_hdr(avbuf); + dvhdr = ubi_get_vid_hdr(dvbuf); + + seen_pebs = init_seen(ubi); + if (IS_ERR(seen_pebs)) { + ret = PTR_ERR(seen_pebs); + goto out_free_dvbuf; + } + + spin_lock(&ubi->volumes_lock); + spin_lock(&ubi->wl_lock); + + fmsb = (struct ubi_fm_sb *)fm_raw; + fm_pos += sizeof(*fmsb); + ubi_assert(fm_pos <= ubi->fm_size); + + fmh = (struct ubi_fm_hdr *)(fm_raw + fm_pos); + fm_pos += sizeof(*fmh); + ubi_assert(fm_pos <= ubi->fm_size); + + fmsb->magic = cpu_to_be32(UBI_FM_SB_MAGIC); + fmsb->version = UBI_FM_FMT_VERSION; + fmsb->used_blocks = cpu_to_be32(new_fm->used_blocks); + /* the max sqnum will be filled in while *reading* the fastmap */ + fmsb->sqnum = 0; + + fmh->magic = cpu_to_be32(UBI_FM_HDR_MAGIC); + free_peb_count = 0; + used_peb_count = 0; + scrub_peb_count = 0; + erase_peb_count = 0; + vol_count = 0; + + fmpl = (struct ubi_fm_scan_pool *)(fm_raw + fm_pos); + fm_pos += sizeof(*fmpl); + fmpl->magic = cpu_to_be32(UBI_FM_POOL_MAGIC); + fmpl->size = cpu_to_be16(ubi->fm_pool.size); + fmpl->max_size = cpu_to_be16(ubi->fm_pool.max_size); + + for (i = 0; i < ubi->fm_pool.size; i++) { + fmpl->pebs[i] = cpu_to_be32(ubi->fm_pool.pebs[i]); + set_seen(ubi, ubi->fm_pool.pebs[i], seen_pebs); + } + + fmpl_wl = (struct ubi_fm_scan_pool *)(fm_raw + fm_pos); + fm_pos += sizeof(*fmpl_wl); + fmpl_wl->magic = cpu_to_be32(UBI_FM_POOL_MAGIC); + fmpl_wl->size = cpu_to_be16(ubi->fm_wl_pool.size); + fmpl_wl->max_size = cpu_to_be16(ubi->fm_wl_pool.max_size); + + for (i = 0; i < ubi->fm_wl_pool.size; i++) { + fmpl_wl->pebs[i] = cpu_to_be32(ubi->fm_wl_pool.pebs[i]); + set_seen(ubi, ubi->fm_wl_pool.pebs[i], seen_pebs); + } + + ubi_for_each_free_peb(ubi, wl_e, tmp_rb) { + fec = (struct ubi_fm_ec *)(fm_raw + fm_pos); + + fec->pnum = cpu_to_be32(wl_e->pnum); + set_seen(ubi, wl_e->pnum, seen_pebs); + fec->ec = cpu_to_be32(wl_e->ec); + + free_peb_count++; + fm_pos += sizeof(*fec); + ubi_assert(fm_pos <= ubi->fm_size); + } + fmh->free_peb_count = cpu_to_be32(free_peb_count); + + ubi_for_each_used_peb(ubi, wl_e, tmp_rb) { + fec = (struct ubi_fm_ec *)(fm_raw + fm_pos); + + fec->pnum = cpu_to_be32(wl_e->pnum); + set_seen(ubi, wl_e->pnum, seen_pebs); + fec->ec = cpu_to_be32(wl_e->ec); + + used_peb_count++; + fm_pos += sizeof(*fec); + ubi_assert(fm_pos <= ubi->fm_size); + } + + ubi_for_each_protected_peb(ubi, i, wl_e) { + fec = (struct ubi_fm_ec *)(fm_raw + fm_pos); + + fec->pnum = cpu_to_be32(wl_e->pnum); + set_seen(ubi, wl_e->pnum, seen_pebs); + fec->ec = cpu_to_be32(wl_e->ec); + + used_peb_count++; + fm_pos += sizeof(*fec); + ubi_assert(fm_pos <= ubi->fm_size); + } + fmh->used_peb_count = cpu_to_be32(used_peb_count); + + ubi_for_each_scrub_peb(ubi, wl_e, tmp_rb) { + fec = (struct ubi_fm_ec *)(fm_raw + fm_pos); + + fec->pnum = cpu_to_be32(wl_e->pnum); + set_seen(ubi, wl_e->pnum, seen_pebs); + fec->ec = cpu_to_be32(wl_e->ec); + + scrub_peb_count++; + fm_pos += sizeof(*fec); + ubi_assert(fm_pos <= ubi->fm_size); + } + fmh->scrub_peb_count = cpu_to_be32(scrub_peb_count); + + + list_for_each_entry(ubi_wrk, &ubi->works, list) { + if (ubi_is_erase_work(ubi_wrk)) { + wl_e = ubi_wrk->e; + ubi_assert(wl_e); + + fec = (struct ubi_fm_ec *)(fm_raw + fm_pos); + + fec->pnum = cpu_to_be32(wl_e->pnum); + set_seen(ubi, wl_e->pnum, seen_pebs); + fec->ec = cpu_to_be32(wl_e->ec); + + erase_peb_count++; + fm_pos += sizeof(*fec); + ubi_assert(fm_pos <= ubi->fm_size); + } + } + fmh->erase_peb_count = cpu_to_be32(erase_peb_count); + + for (i = 0; i < UBI_MAX_VOLUMES + UBI_INT_VOL_COUNT; i++) { + vol = ubi->volumes[i]; + + if (!vol) + continue; + + vol_count++; + + fvh = (struct ubi_fm_volhdr *)(fm_raw + fm_pos); + fm_pos += sizeof(*fvh); + ubi_assert(fm_pos <= ubi->fm_size); + + fvh->magic = cpu_to_be32(UBI_FM_VHDR_MAGIC); + fvh->vol_id = cpu_to_be32(vol->vol_id); + fvh->vol_type = vol->vol_type; + fvh->used_ebs = cpu_to_be32(vol->used_ebs); + fvh->data_pad = cpu_to_be32(vol->data_pad); + fvh->last_eb_bytes = cpu_to_be32(vol->last_eb_bytes); + + ubi_assert(vol->vol_type == UBI_DYNAMIC_VOLUME || + vol->vol_type == UBI_STATIC_VOLUME); + + feba = (struct ubi_fm_eba *)(fm_raw + fm_pos); + fm_pos += sizeof(*feba) + (sizeof(__be32) * vol->reserved_pebs); + ubi_assert(fm_pos <= ubi->fm_size); + + for (j = 0; j < vol->reserved_pebs; j++) { + struct ubi_eba_leb_desc ldesc; + + ubi_eba_get_ldesc(vol, j, &ldesc); + feba->pnum[j] = cpu_to_be32(ldesc.pnum); + } + + feba->reserved_pebs = cpu_to_be32(j); + feba->magic = cpu_to_be32(UBI_FM_EBA_MAGIC); + } + fmh->vol_count = cpu_to_be32(vol_count); + fmh->bad_peb_count = cpu_to_be32(ubi->bad_peb_count); + + avhdr->sqnum = cpu_to_be64(ubi_next_sqnum(ubi)); + avhdr->lnum = 0; + + spin_unlock(&ubi->wl_lock); + spin_unlock(&ubi->volumes_lock); + + dbg_bld("writing fastmap SB to PEB %i", new_fm->e[0]->pnum); + ret = ubi_io_write_vid_hdr(ubi, new_fm->e[0]->pnum, avbuf); + if (ret) { + ubi_err(ubi, "unable to write vid_hdr to fastmap SB!"); + goto out_free_seen; + } + + for (i = 0; i < new_fm->used_blocks; i++) { + fmsb->block_loc[i] = cpu_to_be32(new_fm->e[i]->pnum); + set_seen(ubi, new_fm->e[i]->pnum, seen_pebs); + fmsb->block_ec[i] = cpu_to_be32(new_fm->e[i]->ec); + } + + fmsb->data_crc = 0; + fmsb->data_crc = cpu_to_be32(crc32(UBI_CRC32_INIT, fm_raw, + ubi->fm_size)); + + for (i = 1; i < new_fm->used_blocks; i++) { + dvhdr->sqnum = cpu_to_be64(ubi_next_sqnum(ubi)); + dvhdr->lnum = cpu_to_be32(i); + dbg_bld("writing fastmap data to PEB %i sqnum %llu", + new_fm->e[i]->pnum, be64_to_cpu(dvhdr->sqnum)); + ret = ubi_io_write_vid_hdr(ubi, new_fm->e[i]->pnum, dvbuf); + if (ret) { + ubi_err(ubi, "unable to write vid_hdr to PEB %i!", + new_fm->e[i]->pnum); + goto out_free_seen; + } + } + + for (i = 0; i < new_fm->used_blocks; i++) { + ret = ubi_io_write_data(ubi, fm_raw + (i * ubi->leb_size), + new_fm->e[i]->pnum, 0, ubi->leb_size); + if (ret) { + ubi_err(ubi, "unable to write fastmap to PEB %i!", + new_fm->e[i]->pnum); + goto out_free_seen; + } + } + + ubi_assert(new_fm); + ubi->fm = new_fm; + + ret = self_check_seen(ubi, seen_pebs); + dbg_bld("fastmap written!"); + +out_free_seen: + free_seen(seen_pebs); +out_free_dvbuf: + ubi_free_vid_buf(dvbuf); +out_free_avbuf: + ubi_free_vid_buf(avbuf); + +out: + return ret; +} + +/** + * invalidate_fastmap - destroys a fastmap. + * @ubi: UBI device object + * + * This function ensures that upon next UBI attach a full scan + * is issued. We need this if UBI is about to write a new fastmap + * but is unable to do so. In this case we have two options: + * a) Make sure that the current fastmap will not be usued upon + * attach time and contine or b) fall back to RO mode to have the + * current fastmap in a valid state. + * Returns 0 on success, < 0 indicates an internal error. + */ +static int invalidate_fastmap(struct ubi_device *ubi) +{ + int ret; + struct ubi_fastmap_layout *fm; + struct ubi_wl_entry *e; + struct ubi_vid_io_buf *vb = NULL; + struct ubi_vid_hdr *vh; + + if (!ubi->fm) + return 0; + + ubi->fm = NULL; + + ret = -ENOMEM; + fm = kzalloc_obj(*fm, GFP_NOFS); + if (!fm) + goto out; + + vb = new_fm_vbuf(ubi, UBI_FM_SB_VOLUME_ID); + if (!vb) + goto out_free_fm; + + vh = ubi_get_vid_hdr(vb); + + ret = -ENOSPC; + e = ubi_wl_get_fm_peb(ubi, 1); + if (!e) + goto out_free_fm; + + /* + * Create fake fastmap such that UBI will fall back + * to scanning mode. + */ + vh->sqnum = cpu_to_be64(ubi_next_sqnum(ubi)); + ret = ubi_io_write_vid_hdr(ubi, e->pnum, vb); + if (ret < 0) { + ubi_wl_put_fm_peb(ubi, e, 0, 0); + goto out_free_fm; + } + + fm->used_blocks = 1; + fm->e[0] = e; + + ubi->fm = fm; + +out: + ubi_free_vid_buf(vb); + return ret; + +out_free_fm: + kfree(fm); + goto out; +} + +/** + * return_fm_pebs - returns all PEBs used by a fastmap back to the + * WL sub-system. + * @ubi: UBI device object + * @fm: fastmap layout object + */ +static void return_fm_pebs(struct ubi_device *ubi, + struct ubi_fastmap_layout *fm) +{ + int i; + + if (!fm) + return; + + for (i = 0; i < fm->used_blocks; i++) { + if (fm->e[i]) { + ubi_wl_put_fm_peb(ubi, fm->e[i], i, + fm->to_be_tortured[i]); + fm->e[i] = NULL; + } + } +} + +/** + * ubi_update_fastmap - will be called by UBI if a volume changes or + * a fastmap pool becomes full. + * @ubi: UBI device object + * + * Returns 0 on success, < 0 indicates an internal error. + */ +int ubi_update_fastmap(struct ubi_device *ubi) +{ + int ret, i, j; + struct ubi_fastmap_layout *new_fm, *old_fm; + struct ubi_wl_entry *tmp_e; + + ubi_refill_pools_and_lock(ubi); + + if (ubi->ro_mode || ubi->fm_disabled) { + up_write(&ubi->fm_eba_sem); + up_write(&ubi->work_sem); + up_write(&ubi->fm_protect); + return 0; + } + + new_fm = kzalloc_obj(*new_fm, GFP_NOFS); + if (!new_fm) { + up_write(&ubi->fm_eba_sem); + up_write(&ubi->work_sem); + up_write(&ubi->fm_protect); + return -ENOMEM; + } + + new_fm->used_blocks = ubi->fm_size / ubi->leb_size; + old_fm = ubi->fm; + ubi->fm = NULL; + + if (new_fm->used_blocks > UBI_FM_MAX_BLOCKS) { + ubi_err(ubi, "fastmap too large"); + ret = -ENOSPC; + goto err; + } + + for (i = 1; i < new_fm->used_blocks; i++) { + spin_lock(&ubi->wl_lock); + tmp_e = ubi_wl_get_fm_peb(ubi, 0); + spin_unlock(&ubi->wl_lock); + + if (!tmp_e) { + if (old_fm && old_fm->e[i]) { + ret = ubi_sync_erase(ubi, old_fm->e[i], + &old_fm->to_be_tortured[i]); + if (ret < 0) { + ubi_err(ubi, "could not erase old fastmap PEB"); + + for (j = 1; j < i; j++) { + ubi_wl_put_fm_peb(ubi, new_fm->e[j], + j, 0); + new_fm->e[j] = NULL; + } + goto err; + } + new_fm->e[i] = old_fm->e[i]; + old_fm->e[i] = NULL; + } else { + ubi_err(ubi, "could not get any free erase block"); + + for (j = 1; j < i; j++) { + ubi_wl_put_fm_peb(ubi, new_fm->e[j], j, 0); + new_fm->e[j] = NULL; + } + + ret = -ENOSPC; + goto err; + } + } else { + new_fm->e[i] = tmp_e; + + if (old_fm && old_fm->e[i]) { + ubi_wl_put_fm_peb(ubi, old_fm->e[i], i, + old_fm->to_be_tortured[i]); + old_fm->e[i] = NULL; + } + } + } + + /* Old fastmap is larger than the new one */ + if (old_fm && new_fm->used_blocks < old_fm->used_blocks) { + for (i = new_fm->used_blocks; i < old_fm->used_blocks; i++) { + ubi_wl_put_fm_peb(ubi, old_fm->e[i], i, + old_fm->to_be_tortured[i]); + old_fm->e[i] = NULL; + } + } + + spin_lock(&ubi->wl_lock); + tmp_e = ubi->fm_anchor; + ubi->fm_anchor = NULL; + spin_unlock(&ubi->wl_lock); + + if (old_fm) { + /* no fresh anchor PEB was found, reuse the old one */ + if (!tmp_e) { + ret = ubi_sync_erase(ubi, old_fm->e[0], + &old_fm->to_be_tortured[0]); + if (ret < 0) { + ubi_err(ubi, "could not erase old anchor PEB"); + + for (i = 1; i < new_fm->used_blocks; i++) { + ubi_wl_put_fm_peb(ubi, new_fm->e[i], + i, 0); + new_fm->e[i] = NULL; + } + goto err; + } + new_fm->e[0] = old_fm->e[0]; + old_fm->e[0] = NULL; + } else { + /* we've got a new anchor PEB, return the old one */ + ubi_wl_put_fm_peb(ubi, old_fm->e[0], 0, + old_fm->to_be_tortured[0]); + new_fm->e[0] = tmp_e; + old_fm->e[0] = NULL; + } + } else { + if (!tmp_e) { + ubi_err(ubi, "could not find any anchor PEB"); + + for (i = 1; i < new_fm->used_blocks; i++) { + ubi_wl_put_fm_peb(ubi, new_fm->e[i], i, 0); + new_fm->e[i] = NULL; + } + + ret = -ENOSPC; + goto err; + } + new_fm->e[0] = tmp_e; + } + + ret = ubi_write_fastmap(ubi, new_fm); + + if (ret) + goto err; + +out_unlock: + up_write(&ubi->fm_eba_sem); + up_write(&ubi->work_sem); + up_write(&ubi->fm_protect); + kfree(old_fm); + + ubi_ensure_anchor_pebs(ubi); + + return ret; + +err: + ubi_warn(ubi, "Unable to write new fastmap, err=%i", ret); + + ret = invalidate_fastmap(ubi); + if (ret < 0) { + ubi_err(ubi, "Unable to invalidate current fastmap!"); + ubi_ro_mode(ubi); + } else { + return_fm_pebs(ubi, old_fm); + return_fm_pebs(ubi, new_fm); + ret = 0; + } + + kfree(new_fm); + goto out_unlock; +} diff --git a/drivers/mtd/ubi/gluebi.c b/drivers/mtd/ubi/gluebi.c new file mode 100644 index 000000000..e3891ae82 --- /dev/null +++ b/drivers/mtd/ubi/gluebi.c @@ -0,0 +1,500 @@ +// SPDX-License-Identifier: GPL-2.0-or-later +/* + * Copyright (c) International Business Machines Corp., 2006 + * + * Author: Artem Bityutskiy (Битюцкий Артём), Joern Engel + */ + +/* + * This is a small driver which implements fake MTD devices on top of UBI + * volumes. This sounds strange, but it is in fact quite useful to make + * MTD-oriented software (including all the legacy software) work on top of + * UBI. + * + * Gluebi emulates MTD devices of "MTD_UBIVOLUME" type. Their minimal I/O unit + * size (@mtd->writesize) is equivalent to the UBI minimal I/O unit. The + * eraseblock size is equivalent to the logical eraseblock size of the volume. + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include "ubi-media.h" + +#define err_msg(fmt, ...) \ + pr_err("gluebi (pid %d): %s: " fmt "\n", \ + current->pid, __func__, ##__VA_ARGS__) + +/** + * struct gluebi_device - a gluebi device description data structure. + * @mtd: emulated MTD device description object + * @refcnt: gluebi device reference count + * @desc: UBI volume descriptor + * @ubi_num: UBI device number this gluebi device works on + * @vol_id: ID of UBI volume this gluebi device works on + * @list: link in a list of gluebi devices + */ +struct gluebi_device { + struct mtd_info mtd; + int refcnt; + struct ubi_volume_desc *desc; + int ubi_num; + int vol_id; + struct list_head list; +}; + +/* List of all gluebi devices */ +static LIST_HEAD(gluebi_devices); +static DEFINE_MUTEX(devices_mutex); + +/** + * find_gluebi_nolock - find a gluebi device. + * @ubi_num: UBI device number + * @vol_id: volume ID + * + * This function seraches for gluebi device corresponding to UBI device + * @ubi_num and UBI volume @vol_id. Returns the gluebi device description + * object in case of success and %NULL in case of failure. The caller has to + * have the &devices_mutex locked. + */ +static struct gluebi_device *find_gluebi_nolock(int ubi_num, int vol_id) +{ + struct gluebi_device *gluebi; + + list_for_each_entry(gluebi, &gluebi_devices, list) + if (gluebi->ubi_num == ubi_num && gluebi->vol_id == vol_id) + return gluebi; + return NULL; +} + +/** + * gluebi_get_device - get MTD device reference. + * @mtd: the MTD device description object + * + * This function is called every time the MTD device is being opened and + * implements the MTD get_device() operation. Returns zero in case of success + * and a negative error code in case of failure. + */ +static int gluebi_get_device(struct mtd_info *mtd) +{ + struct gluebi_device *gluebi; + int ubi_mode = UBI_READONLY; + + if (mtd->flags & MTD_WRITEABLE) + ubi_mode = UBI_READWRITE; + + gluebi = container_of(mtd, struct gluebi_device, mtd); + mutex_lock(&devices_mutex); + if (gluebi->refcnt > 0) { + /* + * The MTD device is already referenced and this is just one + * more reference. MTD allows many users to open the same + * volume simultaneously and do not distinguish between + * readers/writers/exclusive/meta openers as UBI does. So we do + * not open the UBI volume again - just increase the reference + * counter and return. + */ + gluebi->refcnt += 1; + mutex_unlock(&devices_mutex); + return 0; + } + + /* + * This is the first reference to this UBI volume via the MTD device + * interface. Open the corresponding volume in read-write mode. + */ + gluebi->desc = ubi_open_volume(gluebi->ubi_num, gluebi->vol_id, + ubi_mode); + if (IS_ERR(gluebi->desc)) { + mutex_unlock(&devices_mutex); + return PTR_ERR(gluebi->desc); + } + gluebi->refcnt += 1; + mutex_unlock(&devices_mutex); + return 0; +} + +/** + * gluebi_put_device - put MTD device reference. + * @mtd: the MTD device description object + * + * This function is called every time the MTD device is being put. Returns + * zero in case of success and a negative error code in case of failure. + */ +static void gluebi_put_device(struct mtd_info *mtd) +{ + struct gluebi_device *gluebi; + + gluebi = container_of(mtd, struct gluebi_device, mtd); + mutex_lock(&devices_mutex); + gluebi->refcnt -= 1; + if (gluebi->refcnt == 0) + ubi_close_volume(gluebi->desc); + mutex_unlock(&devices_mutex); +} + +/** + * gluebi_read - read operation of emulated MTD devices. + * @mtd: MTD device description object + * @from: absolute offset from where to read + * @len: how many bytes to read + * @retlen: count of read bytes is returned here + * @buf: buffer to store the read data + * + * This function returns zero in case of success and a negative error code in + * case of failure. + */ +static int gluebi_read(struct mtd_info *mtd, loff_t from, size_t len, + size_t *retlen, unsigned char *buf) +{ + int err = 0, lnum, offs, bytes_left; + struct gluebi_device *gluebi; + + gluebi = container_of(mtd, struct gluebi_device, mtd); + lnum = div_u64_rem(from, mtd->erasesize, &offs); + bytes_left = len; + while (bytes_left) { + size_t to_read = mtd->erasesize - offs; + + if (to_read > bytes_left) + to_read = bytes_left; + + err = ubi_read(gluebi->desc, lnum, buf, offs, to_read); + if (err) + break; + + lnum += 1; + offs = 0; + bytes_left -= to_read; + buf += to_read; + } + + *retlen = len - bytes_left; + return err; +} + +/** + * gluebi_write - write operation of emulated MTD devices. + * @mtd: MTD device description object + * @to: absolute offset where to write + * @len: how many bytes to write + * @retlen: count of written bytes is returned here + * @buf: buffer with data to write + * + * This function returns zero in case of success and a negative error code in + * case of failure. + */ +static int gluebi_write(struct mtd_info *mtd, loff_t to, size_t len, + size_t *retlen, const u_char *buf) +{ + int err = 0, lnum, offs, bytes_left; + struct gluebi_device *gluebi; + + gluebi = container_of(mtd, struct gluebi_device, mtd); + lnum = div_u64_rem(to, mtd->erasesize, &offs); + + if (len % mtd->writesize || offs % mtd->writesize) + return -EINVAL; + + bytes_left = len; + while (bytes_left) { + size_t to_write = mtd->erasesize - offs; + + if (to_write > bytes_left) + to_write = bytes_left; + + err = ubi_leb_write(gluebi->desc, lnum, buf, offs, to_write); + if (err) + break; + + lnum += 1; + offs = 0; + bytes_left -= to_write; + buf += to_write; + } + + *retlen = len - bytes_left; + return err; +} + +/** + * gluebi_erase - erase operation of emulated MTD devices. + * @mtd: the MTD device description object + * @instr: the erase operation description + * + * This function calls the erase callback when finishes. Returns zero in case + * of success and a negative error code in case of failure. + */ +static int gluebi_erase(struct mtd_info *mtd, struct erase_info *instr) +{ + int err, i, lnum, count; + struct gluebi_device *gluebi; + + if (mtd_mod_by_ws(instr->addr, mtd) || mtd_mod_by_ws(instr->len, mtd)) + return -EINVAL; + + lnum = mtd_div_by_eb(instr->addr, mtd); + count = mtd_div_by_eb(instr->len, mtd); + gluebi = container_of(mtd, struct gluebi_device, mtd); + + for (i = 0; i < count - 1; i++) { + err = ubi_leb_unmap(gluebi->desc, lnum + i); + if (err) + goto out_err; + } + /* + * MTD erase operations are synchronous, so we have to make sure the + * physical eraseblock is wiped out. + * + * Thus, perform leb_erase instead of leb_unmap operation - leb_erase + * will wait for the end of operations + */ + err = ubi_leb_erase(gluebi->desc, lnum + i); + if (err) + goto out_err; + + return 0; + +out_err: + instr->fail_addr = (long long)lnum * mtd->erasesize; + return err; +} + +/** + * gluebi_create - create a gluebi device for an UBI volume. + * @di: UBI device description object + * @vi: UBI volume description object + * + * This function is called when a new UBI volume is created in order to create + * corresponding fake MTD device. Returns zero in case of success and a + * negative error code in case of failure. + */ +static int gluebi_create(struct ubi_device_info *di, + struct ubi_volume_info *vi) +{ + struct gluebi_device *gluebi, *g; + struct mtd_info *mtd; + + gluebi = kzalloc_obj(struct gluebi_device); + if (!gluebi) + return -ENOMEM; + + mtd = &gluebi->mtd; + mtd->name = kmemdup(vi->name, vi->name_len + 1, GFP_KERNEL); + if (!mtd->name) { + kfree(gluebi); + return -ENOMEM; + } + + gluebi->vol_id = vi->vol_id; + gluebi->ubi_num = vi->ubi_num; + mtd->type = MTD_UBIVOLUME; + if (!di->ro_mode) + mtd->flags = MTD_WRITEABLE; + mtd->owner = THIS_MODULE; + mtd->writesize = di->min_io_size; + mtd->erasesize = vi->usable_leb_size; + mtd->_read = gluebi_read; + mtd->_write = gluebi_write; + mtd->_erase = gluebi_erase; + mtd->_get_device = gluebi_get_device; + mtd->_put_device = gluebi_put_device; + + /* + * In case of dynamic a volume, MTD device size is just volume size. In + * case of a static volume the size is equivalent to the amount of data + * bytes. + */ + if (vi->vol_type == UBI_DYNAMIC_VOLUME) + mtd->size = (unsigned long long)vi->usable_leb_size * vi->size; + else + mtd->size = vi->used_bytes; + + /* Just a sanity check - make sure this gluebi device does not exist */ + mutex_lock(&devices_mutex); + g = find_gluebi_nolock(vi->ubi_num, vi->vol_id); + if (g) + err_msg("gluebi MTD device %d form UBI device %d volume %d already exists", + g->mtd.index, vi->ubi_num, vi->vol_id); + mutex_unlock(&devices_mutex); + + if (mtd_device_register(mtd, NULL, 0)) { + err_msg("cannot add MTD device"); + kfree(mtd->name); + kfree(gluebi); + return -ENFILE; + } + + mutex_lock(&devices_mutex); + list_add_tail(&gluebi->list, &gluebi_devices); + mutex_unlock(&devices_mutex); + return 0; +} + +/** + * gluebi_remove - remove a gluebi device. + * @vi: UBI volume description object + * + * This function is called when an UBI volume is removed and it removes + * corresponding fake MTD device. Returns zero in case of success and a + * negative error code in case of failure. + */ +static int gluebi_remove(struct ubi_volume_info *vi) +{ + int err = 0; + struct mtd_info *mtd; + struct gluebi_device *gluebi; + + mutex_lock(&devices_mutex); + gluebi = find_gluebi_nolock(vi->ubi_num, vi->vol_id); + if (!gluebi) { + err_msg("got remove notification for unknown UBI device %d volume %d", + vi->ubi_num, vi->vol_id); + err = -ENOENT; + } else if (gluebi->refcnt) + err = -EBUSY; + else + list_del(&gluebi->list); + mutex_unlock(&devices_mutex); + if (err) + return err; + + mtd = &gluebi->mtd; + err = mtd_device_unregister(mtd); + if (err) { + err_msg("cannot remove fake MTD device %d, UBI device %d, volume %d, error %d", + mtd->index, gluebi->ubi_num, gluebi->vol_id, err); + mutex_lock(&devices_mutex); + list_add_tail(&gluebi->list, &gluebi_devices); + mutex_unlock(&devices_mutex); + return err; + } + + kfree(mtd->name); + kfree(gluebi); + return 0; +} + +/** + * gluebi_updated - UBI volume was updated notifier. + * @vi: volume info structure + * + * This function is called every time an UBI volume is updated. It does nothing + * if te volume @vol is dynamic, and changes MTD device size if the + * volume is static. This is needed because static volumes cannot be read past + * data they contain. This function returns zero in case of success and a + * negative error code in case of error. + */ +static int gluebi_updated(struct ubi_volume_info *vi) +{ + struct gluebi_device *gluebi; + + mutex_lock(&devices_mutex); + gluebi = find_gluebi_nolock(vi->ubi_num, vi->vol_id); + if (!gluebi) { + mutex_unlock(&devices_mutex); + err_msg("got update notification for unknown UBI device %d volume %d", + vi->ubi_num, vi->vol_id); + return -ENOENT; + } + + if (vi->vol_type == UBI_STATIC_VOLUME) + gluebi->mtd.size = vi->used_bytes; + mutex_unlock(&devices_mutex); + return 0; +} + +/** + * gluebi_resized - UBI volume was re-sized notifier. + * @vi: volume info structure + * + * This function is called every time an UBI volume is re-size. It changes the + * corresponding fake MTD device size. This function returns zero in case of + * success and a negative error code in case of error. + */ +static int gluebi_resized(struct ubi_volume_info *vi) +{ + struct gluebi_device *gluebi; + + mutex_lock(&devices_mutex); + gluebi = find_gluebi_nolock(vi->ubi_num, vi->vol_id); + if (!gluebi) { + mutex_unlock(&devices_mutex); + err_msg("got update notification for unknown UBI device %d volume %d", + vi->ubi_num, vi->vol_id); + return -ENOENT; + } + gluebi->mtd.size = vi->used_bytes; + mutex_unlock(&devices_mutex); + return 0; +} + +/** + * gluebi_notify - UBI notification handler. + * @nb: registered notifier block + * @l: notification type + * @ns_ptr: pointer to the &struct ubi_notification object + */ +static int gluebi_notify(struct notifier_block *nb, unsigned long l, + void *ns_ptr) +{ + struct ubi_notification *nt = ns_ptr; + + switch (l) { + case UBI_VOLUME_ADDED: + gluebi_create(&nt->di, &nt->vi); + break; + case UBI_VOLUME_REMOVED: + gluebi_remove(&nt->vi); + break; + case UBI_VOLUME_RESIZED: + gluebi_resized(&nt->vi); + break; + case UBI_VOLUME_UPDATED: + gluebi_updated(&nt->vi); + break; + default: + break; + } + return NOTIFY_OK; +} + +static struct notifier_block gluebi_notifier = { + .notifier_call = gluebi_notify, +}; + +static int __init ubi_gluebi_init(void) +{ + return ubi_register_volume_notifier(&gluebi_notifier, 0); +} + +static void __exit ubi_gluebi_exit(void) +{ + struct gluebi_device *gluebi, *g; + + list_for_each_entry_safe(gluebi, g, &gluebi_devices, list) { + int err; + struct mtd_info *mtd = &gluebi->mtd; + + err = mtd_device_unregister(mtd); + if (err) + err_msg("error %d while removing gluebi MTD device %d, UBI device %d, volume %d - ignoring", + err, mtd->index, gluebi->ubi_num, + gluebi->vol_id); + kfree(mtd->name); + kfree(gluebi); + } + ubi_unregister_volume_notifier(&gluebi_notifier); +} + +module_init(ubi_gluebi_init); +module_exit(ubi_gluebi_exit); +MODULE_DESCRIPTION("MTD emulation layer over UBI volumes"); +MODULE_AUTHOR("Artem Bityutskiy, Joern Engel"); +MODULE_LICENSE("GPL"); diff --git a/drivers/mtd/ubi/io.c b/drivers/mtd/ubi/io.c new file mode 100644 index 000000000..d70d1313a --- /dev/null +++ b/drivers/mtd/ubi/io.c @@ -0,0 +1,1487 @@ +// SPDX-License-Identifier: GPL-2.0-or-later +/* + * Copyright (c) International Business Machines Corp., 2006 + * Copyright (c) Nokia Corporation, 2006, 2007 + * + * Author: Artem Bityutskiy (Битюцкий Артём) + */ + +/* + * UBI input/output sub-system. + * + * This sub-system provides a uniform way to work with all kinds of the + * underlying MTD devices. It also implements handy functions for reading and + * writing UBI headers. + * + * We are trying to have a paranoid mindset and not to trust to what we read + * from the flash media in order to be more secure and robust. So this + * sub-system validates every single header it reads from the flash media. + * + * Some words about how the eraseblock headers are stored. + * + * The erase counter header is always stored at offset zero. By default, the + * VID header is stored after the EC header at the closest aligned offset + * (i.e. aligned to the minimum I/O unit size). Data starts next to the VID + * header at the closest aligned offset. But this default layout may be + * changed. For example, for different reasons (e.g., optimization) UBI may be + * asked to put the VID header at further offset, and even at an unaligned + * offset. Of course, if the offset of the VID header is unaligned, UBI adds + * proper padding in front of it. Data offset may also be changed but it has to + * be aligned. + * + * About minimal I/O units. In general, UBI assumes flash device model where + * there is only one minimal I/O unit size. E.g., in case of NOR flash it is 1, + * in case of NAND flash it is a NAND page, etc. This is reported by MTD in the + * @ubi->mtd->writesize field. But as an exception, UBI admits use of another + * (smaller) minimal I/O unit size for EC and VID headers to make it possible + * to do different optimizations. + * + * This is extremely useful in case of NAND flashes which admit of several + * write operations to one NAND page. In this case UBI can fit EC and VID + * headers at one NAND page. Thus, UBI may use "sub-page" size as the minimal + * I/O unit for the headers (the @ubi->hdrs_min_io_size field). But it still + * reports NAND page size (@ubi->min_io_size) as a minimal I/O unit for the UBI + * users. + * + * Example: some Samsung NANDs with 2KiB pages allow 4x 512-byte writes, so + * although the minimal I/O unit is 2K, UBI uses 512 bytes for EC and VID + * headers. + * + * Q: why not just to treat sub-page as a minimal I/O unit of this flash + * device, e.g., make @ubi->min_io_size = 512 in the example above? + * + * A: because when writing a sub-page, MTD still writes a full 2K page but the + * bytes which are not relevant to the sub-page are 0xFF. So, basically, + * writing 4x512 sub-pages is 4 times slower than writing one 2KiB NAND page. + * Thus, we prefer to use sub-pages only for EC and VID headers. + * + * As it was noted above, the VID header may start at a non-aligned offset. + * For example, in case of a 2KiB page NAND flash with a 512 bytes sub-page, + * the VID header may reside at offset 1984 which is the last 64 bytes of the + * last sub-page (EC header is always at offset zero). This causes some + * difficulties when reading and writing VID headers. + * + * Suppose we have a 64-byte buffer and we read a VID header at it. We change + * the data and want to write this VID header out. As we can only write in + * 512-byte chunks, we have to allocate one more buffer and copy our VID header + * to offset 448 of this buffer. + * + * The I/O sub-system does the following trick in order to avoid this extra + * copy. It always allocates a @ubi->vid_hdr_alsize bytes buffer for the VID + * header and returns a pointer to offset @ubi->vid_hdr_shift of this buffer. + * When the VID header is being written out, it shifts the VID header pointer + * back and writes the whole sub-page. + */ + +#include +#include +#include +#include "ubi.h" + +static int self_check_not_bad(const struct ubi_device *ubi, int pnum); +static int self_check_peb_ec_hdr(const struct ubi_device *ubi, int pnum); +static int self_check_ec_hdr(const struct ubi_device *ubi, int pnum, + const struct ubi_ec_hdr *ec_hdr); +static int self_check_peb_vid_hdr(const struct ubi_device *ubi, int pnum); +static int self_check_vid_hdr(const struct ubi_device *ubi, int pnum, + const struct ubi_vid_hdr *vid_hdr); +static int self_check_write(struct ubi_device *ubi, const void *buf, int pnum, + int offset, int len); + +/** + * ubi_io_read - read data from a physical eraseblock. + * @ubi: UBI device description object + * @buf: buffer where to store the read data + * @pnum: physical eraseblock number to read from + * @offset: offset within the physical eraseblock from where to read + * @len: how many bytes to read + * + * This function reads data from offset @offset of physical eraseblock @pnum + * and stores the read data in the @buf buffer. The following return codes are + * possible: + * + * o %0 if all the requested data were successfully read; + * o %UBI_IO_BITFLIPS if all the requested data were successfully read, but + * correctable bit-flips were detected; this is harmless but may indicate + * that this eraseblock may become bad soon (but do not have to); + * o %-EBADMSG if the MTD subsystem reported about data integrity problems, for + * example it can be an ECC error in case of NAND; this most probably means + * that the data is corrupted; + * o %-EIO if some I/O error occurred; + * o other negative error codes in case of other errors. + */ +int ubi_io_read(const struct ubi_device *ubi, void *buf, int pnum, int offset, + int len) +{ + int err, retries = 0; + size_t read; + loff_t addr; + + dbg_io("read %d bytes from PEB %d:%d", len, pnum, offset); + + ubi_assert(pnum >= 0 && pnum < ubi->peb_count); + ubi_assert(offset >= 0 && offset + len <= ubi->peb_size); + ubi_assert(len > 0); + + err = self_check_not_bad(ubi, pnum); + if (err) + return err; + + /* + * Deliberately corrupt the buffer to improve robustness. Indeed, if we + * do not do this, the following may happen: + * 1. The buffer contains data from previous operation, e.g., read from + * another PEB previously. The data looks like expected, e.g., if we + * just do not read anything and return - the caller would not + * notice this. E.g., if we are reading a VID header, the buffer may + * contain a valid VID header from another PEB. + * 2. The driver is buggy and returns us success or -EBADMSG or + * -EUCLEAN, but it does not actually put any data to the buffer. + * + * This may confuse UBI or upper layers - they may think the buffer + * contains valid data while in fact it is just old data. This is + * especially possible because UBI (and UBIFS) relies on CRC, and + * treats data as correct even in case of ECC errors if the CRC is + * correct. + * + * Try to prevent this situation by changing the first byte of the + * buffer. + */ + *((uint8_t *)buf) ^= 0xFF; + + addr = (loff_t)pnum * ubi->peb_size + offset; +retry: + err = mtd_read(ubi->mtd, addr, len, &read, buf); + if (err) { + const char *errstr = mtd_is_eccerr(err) ? " (ECC error)" : ""; + + if (mtd_is_bitflip(err)) { + /* + * -EUCLEAN is reported if there was a bit-flip which + * was corrected, so this is harmless. + * + * We do not report about it here unless debugging is + * enabled. A corresponding message will be printed + * later, when it is has been scrubbed. + */ + ubi_msg(ubi, "fixable bit-flip detected at PEB %d", + pnum); + ubi_assert(len == read); + return UBI_IO_BITFLIPS; + } + + if (retries++ < UBI_IO_RETRIES) { + ubi_warn(ubi, "error %d%s while reading %d bytes from PEB %d:%d, read only %zd bytes, retry", + err, errstr, len, pnum, offset, read); + yield(); + goto retry; + } + + ubi_err(ubi, "error %d%s while reading %d bytes from PEB %d:%d, read %zd bytes", + err, errstr, len, pnum, offset, read); + dump_stack(); + + /* + * The driver should never return -EBADMSG if it failed to read + * all the requested data. But some buggy drivers might do + * this, so we change it to -EIO. + */ + if (read != len && mtd_is_eccerr(err)) { + ubi_assert(0); + err = -EIO; + } + } else { + ubi_assert(len == read); + + if (ubi_dbg_is_bitflip(ubi)) { + dbg_gen("bit-flip (emulated)"); + return UBI_IO_BITFLIPS; + } + + if (ubi_dbg_is_read_failure(ubi, MASK_READ_FAILURE)) { + ubi_warn(ubi, "cannot read %d bytes from PEB %d:%d (emulated)", + len, pnum, offset); + return -EIO; + } + + if (ubi_dbg_is_eccerr(ubi)) { + ubi_warn(ubi, "ECC error (emulated) while reading %d bytes from PEB %d:%d, read %zd bytes", + len, pnum, offset, read); + return -EBADMSG; + } + } + + return err; +} + +/** + * ubi_io_write - write data to a physical eraseblock. + * @ubi: UBI device description object + * @buf: buffer with the data to write + * @pnum: physical eraseblock number to write to + * @offset: offset within the physical eraseblock where to write + * @len: how many bytes to write + * + * This function writes @len bytes of data from buffer @buf to offset @offset + * of physical eraseblock @pnum. If all the data were successfully written, + * zero is returned. If an error occurred, this function returns a negative + * error code. If %-EIO is returned, the physical eraseblock most probably went + * bad. + * + * Note, in case of an error, it is possible that something was still written + * to the flash media, but may be some garbage. + */ +int ubi_io_write(struct ubi_device *ubi, const void *buf, int pnum, int offset, + int len) +{ + int err; + size_t written; + loff_t addr; + + dbg_io("write %d bytes to PEB %d:%d", len, pnum, offset); + + ubi_assert(pnum >= 0 && pnum < ubi->peb_count); + ubi_assert(offset >= 0 && offset + len <= ubi->peb_size); + ubi_assert(offset % ubi->hdrs_min_io_size == 0); + ubi_assert(len > 0 && len % ubi->hdrs_min_io_size == 0); + + if (ubi->ro_mode) { + ubi_err(ubi, "read-only mode"); + return -EROFS; + } + + err = self_check_not_bad(ubi, pnum); + if (err) + return err; + + /* The area we are writing to has to contain all 0xFF bytes */ + err = ubi_self_check_all_ff(ubi, pnum, offset, len); + if (err) + return err; + + if (offset >= ubi->leb_start) { + /* + * We write to the data area of the physical eraseblock. Make + * sure it has valid EC and VID headers. + */ + err = self_check_peb_ec_hdr(ubi, pnum); + if (err) + return err; + err = self_check_peb_vid_hdr(ubi, pnum); + if (err) + return err; + } + + if (ubi_dbg_is_write_failure(ubi)) { + ubi_err(ubi, "cannot write %d bytes to PEB %d:%d (emulated)", + len, pnum, offset); + dump_stack(); + return -EIO; + } + + addr = (loff_t)pnum * ubi->peb_size + offset; + err = mtd_write(ubi->mtd, addr, len, &written, buf); + if (err) { + ubi_err(ubi, "error %d while writing %d bytes to PEB %d:%d, written %zd bytes", + err, len, pnum, offset, written); + dump_stack(); + ubi_dump_flash(ubi, pnum, offset, len); + } else + ubi_assert(written == len); + + if (!err) { + err = self_check_write(ubi, buf, pnum, offset, len); + if (err) + return err; + + /* + * Since we always write sequentially, the rest of the PEB has + * to contain only 0xFF bytes. + */ + offset += len; + len = ubi->peb_size - offset; + if (len) + err = ubi_self_check_all_ff(ubi, pnum, offset, len); + } + + return err; +} + +/** + * do_sync_erase - synchronously erase a physical eraseblock. + * @ubi: UBI device description object + * @pnum: the physical eraseblock number to erase + * + * This function synchronously erases physical eraseblock @pnum and returns + * zero in case of success and a negative error code in case of failure. If + * %-EIO is returned, the physical eraseblock most probably went bad. + */ +static int do_sync_erase(struct ubi_device *ubi, int pnum) +{ + int err, retries = 0; + struct erase_info ei; + + dbg_io("erase PEB %d", pnum); + ubi_assert(pnum >= 0 && pnum < ubi->peb_count); + + if (ubi->ro_mode) { + ubi_err(ubi, "read-only mode"); + return -EROFS; + } + +retry: + memset(&ei, 0, sizeof(struct erase_info)); + + ei.addr = (loff_t)pnum * ubi->peb_size; + ei.len = ubi->peb_size; + + err = mtd_erase(ubi->mtd, &ei); + if (err) { + if (retries++ < UBI_IO_RETRIES) { + ubi_warn(ubi, "error %d while erasing PEB %d, retry", + err, pnum); + yield(); + goto retry; + } + ubi_err(ubi, "cannot erase PEB %d, error %d", pnum, err); + dump_stack(); + return err; + } + + err = ubi_self_check_all_ff(ubi, pnum, 0, ubi->peb_size); + if (err) + return err; + + if (ubi_dbg_is_erase_failure(ubi)) { + ubi_err(ubi, "cannot erase PEB %d (emulated)", pnum); + return -EIO; + } + + return 0; +} + +/* Patterns to write to a physical eraseblock when torturing it */ +static uint8_t patterns[] = {0xa5, 0x5a, 0x0}; + +/** + * torture_peb - test a supposedly bad physical eraseblock. + * @ubi: UBI device description object + * @pnum: the physical eraseblock number to test + * + * This function returns %-EIO if the physical eraseblock did not pass the + * test, a positive number of erase operations done if the test was + * successfully passed, and other negative error codes in case of other errors. + */ +static int torture_peb(struct ubi_device *ubi, int pnum) +{ + int err, i, patt_count; + + ubi_msg(ubi, "run torture test for PEB %d", pnum); + patt_count = ARRAY_SIZE(patterns); + ubi_assert(patt_count > 0); + + mutex_lock(&ubi->buf_mutex); + for (i = 0; i < patt_count; i++) { + err = do_sync_erase(ubi, pnum); + if (err) + goto out; + + /* Make sure the PEB contains only 0xFF bytes */ + err = ubi_io_read(ubi, ubi->peb_buf, pnum, 0, ubi->peb_size); + if (err) + goto out; + + err = ubi_check_pattern(ubi->peb_buf, 0xFF, ubi->peb_size); + if (err == 0) { + ubi_err(ubi, "erased PEB %d, but a non-0xFF byte found", + pnum); + err = -EIO; + goto out; + } + + /* Write a pattern and check it */ + memset(ubi->peb_buf, patterns[i], ubi->peb_size); + err = ubi_io_write(ubi, ubi->peb_buf, pnum, 0, ubi->peb_size); + if (err) + goto out; + + memset(ubi->peb_buf, ~patterns[i], ubi->peb_size); + err = ubi_io_read(ubi, ubi->peb_buf, pnum, 0, ubi->peb_size); + if (err) + goto out; + + err = ubi_check_pattern(ubi->peb_buf, patterns[i], + ubi->peb_size); + if (err == 0) { + ubi_err(ubi, "pattern %x checking failed for PEB %d", + patterns[i], pnum); + err = -EIO; + goto out; + } + } + + err = patt_count; + ubi_msg(ubi, "PEB %d passed torture test, do not mark it as bad", pnum); + +out: + mutex_unlock(&ubi->buf_mutex); + if (err == UBI_IO_BITFLIPS || mtd_is_eccerr(err)) { + /* + * If a bit-flip or data integrity error was detected, the test + * has not passed because it happened on a freshly erased + * physical eraseblock which means something is wrong with it. + */ + ubi_err(ubi, "read problems on freshly erased PEB %d, must be bad", + pnum); + err = -EIO; + } + return err; +} + +/** + * nor_erase_prepare - prepare a NOR flash PEB for erasure. + * @ubi: UBI device description object + * @pnum: physical eraseblock number to prepare + * + * NOR flash, or at least some of them, have peculiar embedded PEB erasure + * algorithm: the PEB is first filled with zeroes, then it is erased. And + * filling with zeroes starts from the end of the PEB. This was observed with + * Spansion S29GL512N NOR flash. + * + * This means that in case of a power cut we may end up with intact data at the + * beginning of the PEB, and all zeroes at the end of PEB. In other words, the + * EC and VID headers are OK, but a large chunk of data at the end of PEB is + * zeroed. This makes UBI mistakenly treat this PEB as used and associate it + * with an LEB, which leads to subsequent failures (e.g., UBIFS fails). + * + * This function is called before erasing NOR PEBs and it zeroes out EC and VID + * magic numbers in order to invalidate them and prevent the failures. Returns + * zero in case of success and a negative error code in case of failure. + */ +static int nor_erase_prepare(struct ubi_device *ubi, int pnum) +{ + int err; + size_t written; + loff_t addr; + uint32_t data = 0; + struct ubi_ec_hdr ec_hdr; + struct ubi_vid_io_buf vidb; + + /* + * Note, we cannot generally define VID header buffers on stack, + * because of the way we deal with these buffers (see the header + * comment in this file). But we know this is a NOR-specific piece of + * code, so we can do this. But yes, this is error-prone and we should + * (pre-)allocate VID header buffer instead. + */ + struct ubi_vid_hdr vid_hdr; + + /* + * If VID or EC is valid, we have to corrupt them before erasing. + * It is important to first invalidate the EC header, and then the VID + * header. Otherwise a power cut may lead to valid EC header and + * invalid VID header, in which case UBI will treat this PEB as + * corrupted and will try to preserve it, and print scary warnings. + */ + addr = (loff_t)pnum * ubi->peb_size; + err = ubi_io_read_ec_hdr(ubi, pnum, &ec_hdr, 0); + if (err != UBI_IO_BAD_HDR_EBADMSG && err != UBI_IO_BAD_HDR && + err != UBI_IO_FF){ + err = mtd_write(ubi->mtd, addr, 4, &written, (void *)&data); + if(err) + goto error; + } + + ubi_init_vid_buf(ubi, &vidb, &vid_hdr); + ubi_assert(&vid_hdr == ubi_get_vid_hdr(&vidb)); + + err = ubi_io_read_vid_hdr(ubi, pnum, &vidb, 0); + if (err != UBI_IO_BAD_HDR_EBADMSG && err != UBI_IO_BAD_HDR && + err != UBI_IO_FF){ + addr += ubi->vid_hdr_aloffset; + err = mtd_write(ubi->mtd, addr, 4, &written, (void *)&data); + if (err) + goto error; + } + return 0; + +error: + /* + * The PEB contains a valid VID or EC header, but we cannot invalidate + * it. Supposedly the flash media or the driver is screwed up, so + * return an error. + */ + ubi_err(ubi, "cannot invalidate PEB %d, write returned %d", pnum, err); + ubi_dump_flash(ubi, pnum, 0, ubi->peb_size); + return -EIO; +} + +/** + * ubi_io_sync_erase - synchronously erase a physical eraseblock. + * @ubi: UBI device description object + * @pnum: the physical eraseblock number to erase + * @torture: if this physical eraseblock has to be tortured; cleared to zero + * once the torture test has completed successfully so that a retry + * of the erase does not torture the physical eraseblock again + * + * This function synchronously erases physical eraseblock @pnum. If @torture + * flag is not zero, the physical eraseblock is checked by means of writing + * different patterns to it and reading them back. If the torturing is enabled, + * the physical eraseblock is erased more than once. + * + * This function returns the number of erasures made in case of success, %-EIO + * if the erasure failed or the torturing test failed, and other negative error + * codes in case of other errors. Note, %-EIO means that the physical + * eraseblock is bad. + */ +int ubi_io_sync_erase(struct ubi_device *ubi, int pnum, int *torture) +{ + int err, ret = 0; + + ubi_assert(pnum >= 0 && pnum < ubi->peb_count); + + err = self_check_not_bad(ubi, pnum); + if (err != 0) + return err; + + if (ubi->ro_mode) { + ubi_err(ubi, "read-only mode"); + return -EROFS; + } + + /* + * If the flash is ECC-ed then we have to erase the ECC block before we + * can write to it. But the write is in preparation to an erase in the + * first place. This means we cannot zero out EC and VID before the + * erase and we just have to hope the flash starts erasing from the + * start of the page. + */ + if (ubi->nor_flash && ubi->mtd->writesize == 1) { + err = nor_erase_prepare(ubi, pnum); + if (err) + return err; + } + + if (*torture) { + ret = torture_peb(ubi, pnum); + if (ret < 0) + return ret; + *torture = 0; + } + + err = do_sync_erase(ubi, pnum); + if (err) + return err; + + return ret + 1; +} + +/** + * ubi_io_is_bad - check if a physical eraseblock is bad. + * @ubi: UBI device description object + * @pnum: the physical eraseblock number to check + * + * This function returns a positive number if the physical eraseblock is bad, + * zero if not, and a negative error code if an error occurred. + */ +int ubi_io_is_bad(const struct ubi_device *ubi, int pnum) +{ + struct mtd_info *mtd = ubi->mtd; + + ubi_assert(pnum >= 0 && pnum < ubi->peb_count); + + if (ubi->bad_allowed) { + int ret; + + ret = mtd_block_isbad(mtd, (loff_t)pnum * ubi->peb_size); + if (ret < 0) + ubi_err(ubi, "error %d while checking if PEB %d is bad", + ret, pnum); + else if (ret) + dbg_io("PEB %d is bad", pnum); + return ret; + } + + return 0; +} + +/** + * ubi_io_mark_bad - mark a physical eraseblock as bad. + * @ubi: UBI device description object + * @pnum: the physical eraseblock number to mark + * + * This function returns zero in case of success and a negative error code in + * case of failure. + */ +int ubi_io_mark_bad(const struct ubi_device *ubi, int pnum) +{ + int err; + struct mtd_info *mtd = ubi->mtd; + + ubi_assert(pnum >= 0 && pnum < ubi->peb_count); + + if (ubi->ro_mode) { + ubi_err(ubi, "read-only mode"); + return -EROFS; + } + + if (!ubi->bad_allowed) + return 0; + + err = mtd_block_markbad(mtd, (loff_t)pnum * ubi->peb_size); + if (err) + ubi_err(ubi, "cannot mark PEB %d bad, error %d", pnum, err); + return err; +} + +/** + * validate_ec_hdr - validate an erase counter header. + * @ubi: UBI device description object + * @ec_hdr: the erase counter header to check + * + * This function returns zero if the erase counter header is OK, and %1 if + * not. + */ +static int validate_ec_hdr(const struct ubi_device *ubi, + const struct ubi_ec_hdr *ec_hdr) +{ + long long ec; + int vid_hdr_offset, leb_start; + + ec = be64_to_cpu(ec_hdr->ec); + vid_hdr_offset = be32_to_cpu(ec_hdr->vid_hdr_offset); + leb_start = be32_to_cpu(ec_hdr->data_offset); + + if (ec_hdr->version != UBI_VERSION) { + ubi_err(ubi, "node with incompatible UBI version found: this UBI version is %d, image version is %d", + UBI_VERSION, (int)ec_hdr->version); + goto bad; + } + + if (vid_hdr_offset != ubi->vid_hdr_offset) { + ubi_err(ubi, "bad VID header offset %d, expected %d", + vid_hdr_offset, ubi->vid_hdr_offset); + goto bad; + } + + if (leb_start != ubi->leb_start) { + ubi_err(ubi, "bad data offset %d, expected %d", + leb_start, ubi->leb_start); + goto bad; + } + + if (ec < 0 || ec > UBI_MAX_ERASECOUNTER) { + ubi_err(ubi, "bad erase counter %lld", ec); + goto bad; + } + + return 0; + +bad: + ubi_err(ubi, "bad EC header"); + ubi_dump_ec_hdr(ec_hdr); + dump_stack(); + return 1; +} + +/** + * ubi_io_read_ec_hdr - read and check an erase counter header. + * @ubi: UBI device description object + * @pnum: physical eraseblock to read from + * @ec_hdr: a &struct ubi_ec_hdr object where to store the read erase counter + * header + * @verbose: be verbose if the header is corrupted or was not found + * + * This function reads erase counter header from physical eraseblock @pnum and + * stores it in @ec_hdr. This function also checks CRC checksum of the read + * erase counter header. The following codes may be returned: + * + * o %0 if the CRC checksum is correct and the header was successfully read; + * o %UBI_IO_BITFLIPS if the CRC is correct, but bit-flips were detected + * and corrected by the flash driver; this is harmless but may indicate that + * this eraseblock may become bad soon (but may be not); + * o %UBI_IO_BAD_HDR if the erase counter header is corrupted (a CRC error); + * o %UBI_IO_BAD_HDR_EBADMSG is the same as %UBI_IO_BAD_HDR, but there also was + * a data integrity error (uncorrectable ECC error in case of NAND); + * o %UBI_IO_FF if only 0xFF bytes were read (the PEB is supposedly empty) + * o a negative error code in case of failure. + */ +int ubi_io_read_ec_hdr(struct ubi_device *ubi, int pnum, + struct ubi_ec_hdr *ec_hdr, int verbose) +{ + int err, read_err; + uint32_t crc, magic, hdr_crc; + + dbg_io("read EC header from PEB %d", pnum); + ubi_assert(pnum >= 0 && pnum < ubi->peb_count); + + read_err = ubi_io_read(ubi, ec_hdr, pnum, 0, UBI_EC_HDR_SIZE); + if (read_err) { + if (read_err != UBI_IO_BITFLIPS && !mtd_is_eccerr(read_err)) + return read_err; + + /* + * We read all the data, but either a correctable bit-flip + * occurred, or MTD reported a data integrity error + * (uncorrectable ECC error in case of NAND). The former is + * harmless, the later may mean that the read data is + * corrupted. But we have a CRC check-sum and we will detect + * this. If the EC header is still OK, we just report this as + * there was a bit-flip, to force scrubbing. + */ + } + + magic = be32_to_cpu(ec_hdr->magic); + if (magic != UBI_EC_HDR_MAGIC) { + if (mtd_is_eccerr(read_err)) + return UBI_IO_BAD_HDR_EBADMSG; + + /* + * The magic field is wrong. Let's check if we have read all + * 0xFF. If yes, this physical eraseblock is assumed to be + * empty. + */ + if (ubi_check_pattern(ec_hdr, 0xFF, UBI_EC_HDR_SIZE)) { + /* The physical eraseblock is supposedly empty */ + if (verbose) + ubi_warn(ubi, "no EC header found at PEB %d, only 0xFF bytes", + pnum); + dbg_bld("no EC header found at PEB %d, only 0xFF bytes", + pnum); + if (!read_err) + return UBI_IO_FF; + else + return UBI_IO_FF_BITFLIPS; + } + + /* + * This is not a valid erase counter header, and these are not + * 0xFF bytes. Report that the header is corrupted. + */ + if (verbose) { + ubi_warn(ubi, "bad magic number at PEB %d: %08x instead of %08x", + pnum, magic, UBI_EC_HDR_MAGIC); + ubi_dump_ec_hdr(ec_hdr); + } + dbg_bld("bad magic number at PEB %d: %08x instead of %08x", + pnum, magic, UBI_EC_HDR_MAGIC); + return UBI_IO_BAD_HDR; + } + + crc = crc32(UBI_CRC32_INIT, ec_hdr, UBI_EC_HDR_SIZE_CRC); + hdr_crc = be32_to_cpu(ec_hdr->hdr_crc); + + if (hdr_crc != crc) { + if (verbose) { + ubi_warn(ubi, "bad EC header CRC at PEB %d, calculated %#08x, read %#08x", + pnum, crc, hdr_crc); + ubi_dump_ec_hdr(ec_hdr); + } + dbg_bld("bad EC header CRC at PEB %d, calculated %#08x, read %#08x", + pnum, crc, hdr_crc); + + if (!read_err) + return UBI_IO_BAD_HDR; + else + return UBI_IO_BAD_HDR_EBADMSG; + } + + /* And of course validate what has just been read from the media */ + err = validate_ec_hdr(ubi, ec_hdr); + if (err) { + ubi_err(ubi, "validation failed for PEB %d", pnum); + return -EINVAL; + } + + /* + * If there was %-EBADMSG, but the header CRC is still OK, report about + * a bit-flip to force scrubbing on this PEB. + */ + if (read_err) + return UBI_IO_BITFLIPS; + + if (ubi_dbg_is_read_failure(ubi, MASK_READ_FAILURE_EC)) { + ubi_warn(ubi, "cannot read EC header from PEB %d (emulated)", + pnum); + return -EIO; + } + + if (ubi_dbg_is_ff(ubi, MASK_IO_FF_EC)) { + ubi_warn(ubi, "bit-all-ff (emulated)"); + return UBI_IO_FF; + } + + if (ubi_dbg_is_ff_bitflips(ubi, MASK_IO_FF_BITFLIPS_EC)) { + ubi_warn(ubi, "bit-all-ff with error reported by MTD driver (emulated)"); + return UBI_IO_FF_BITFLIPS; + } + + if (ubi_dbg_is_bad_hdr(ubi, MASK_BAD_HDR_EC)) { + ubi_warn(ubi, "bad_hdr (emulated)"); + return UBI_IO_BAD_HDR; + } + + if (ubi_dbg_is_bad_hdr_ebadmsg(ubi, MASK_BAD_HDR_EBADMSG_EC)) { + ubi_warn(ubi, "bad_hdr with ECC error (emulated)"); + return UBI_IO_BAD_HDR_EBADMSG; + } + + return 0; +} + +/** + * ubi_io_write_ec_hdr - write an erase counter header. + * @ubi: UBI device description object + * @pnum: physical eraseblock to write to + * @ec_hdr: the erase counter header to write + * + * This function writes erase counter header described by @ec_hdr to physical + * eraseblock @pnum. It also fills most fields of @ec_hdr before writing, so + * the caller do not have to fill them. Callers must only fill the @ec_hdr->ec + * field. + * + * This function returns zero in case of success and a negative error code in + * case of failure. If %-EIO is returned, the physical eraseblock most probably + * went bad. + */ +int ubi_io_write_ec_hdr(struct ubi_device *ubi, int pnum, + struct ubi_ec_hdr *ec_hdr) +{ + int err; + uint32_t crc; + + dbg_io("write EC header to PEB %d", pnum); + ubi_assert(pnum >= 0 && pnum < ubi->peb_count); + + ec_hdr->magic = cpu_to_be32(UBI_EC_HDR_MAGIC); + ec_hdr->version = UBI_VERSION; + ec_hdr->vid_hdr_offset = cpu_to_be32(ubi->vid_hdr_offset); + ec_hdr->data_offset = cpu_to_be32(ubi->leb_start); + ec_hdr->image_seq = cpu_to_be32(ubi->image_seq); + crc = crc32(UBI_CRC32_INIT, ec_hdr, UBI_EC_HDR_SIZE_CRC); + ec_hdr->hdr_crc = cpu_to_be32(crc); + + err = self_check_ec_hdr(ubi, pnum, ec_hdr); + if (err) + return err; + + if (ubi_dbg_is_power_cut(ubi, MASK_POWER_CUT_EC)) { + ubi_warn(ubi, "emulating a power cut when writing EC header"); + ubi_ro_mode(ubi); + return -EROFS; + } + + memset((char *)ec_hdr + UBI_EC_HDR_SIZE, 0xFF, ubi->ec_hdr_alsize - UBI_EC_HDR_SIZE); + + err = ubi_io_write(ubi, ec_hdr, pnum, 0, ubi->ec_hdr_alsize); + return err; +} + +/** + * validate_vid_hdr - validate a volume identifier header. + * @ubi: UBI device description object + * @vid_hdr: the volume identifier header to check + * + * This function checks that data stored in the volume identifier header + * @vid_hdr. Returns zero if the VID header is OK and %1 if not. + */ +static int validate_vid_hdr(const struct ubi_device *ubi, + const struct ubi_vid_hdr *vid_hdr) +{ + int vol_type = vid_hdr->vol_type; + int copy_flag = vid_hdr->copy_flag; + int vol_id = be32_to_cpu(vid_hdr->vol_id); + int lnum = be32_to_cpu(vid_hdr->lnum); + int compat = vid_hdr->compat; + int data_size = be32_to_cpu(vid_hdr->data_size); + int used_ebs = be32_to_cpu(vid_hdr->used_ebs); + int data_pad = be32_to_cpu(vid_hdr->data_pad); + int data_crc = be32_to_cpu(vid_hdr->data_crc); + int usable_leb_size = ubi->leb_size - data_pad; + + if (copy_flag != 0 && copy_flag != 1) { + ubi_err(ubi, "bad copy_flag"); + goto bad; + } + + if (vol_id < 0 || lnum < 0 || data_size < 0 || used_ebs < 0 || + data_pad < 0) { + ubi_err(ubi, "negative values"); + goto bad; + } + + if (vol_id >= UBI_MAX_VOLUMES && vol_id < UBI_INTERNAL_VOL_START) { + ubi_err(ubi, "bad vol_id"); + goto bad; + } + + if (vol_id < UBI_INTERNAL_VOL_START && compat != 0) { + ubi_err(ubi, "bad compat"); + goto bad; + } + + if (vol_id >= UBI_INTERNAL_VOL_START && compat != UBI_COMPAT_DELETE && + compat != UBI_COMPAT_RO && compat != UBI_COMPAT_PRESERVE && + compat != UBI_COMPAT_REJECT) { + ubi_err(ubi, "bad compat"); + goto bad; + } + + if (vol_type != UBI_VID_DYNAMIC && vol_type != UBI_VID_STATIC) { + ubi_err(ubi, "bad vol_type"); + goto bad; + } + + if (data_pad >= ubi->leb_size / 2) { + ubi_err(ubi, "bad data_pad"); + goto bad; + } + + if (data_size > ubi->leb_size) { + ubi_err(ubi, "bad data_size"); + goto bad; + } + + if (vol_type == UBI_VID_STATIC) { + /* + * Although from high-level point of view static volumes may + * contain zero bytes of data, but no VID headers can contain + * zero at these fields, because they empty volumes do not have + * mapped logical eraseblocks. + */ + if (used_ebs == 0) { + ubi_err(ubi, "zero used_ebs"); + goto bad; + } + if (data_size == 0) { + ubi_err(ubi, "zero data_size"); + goto bad; + } + if (lnum < used_ebs - 1) { + if (data_size != usable_leb_size) { + ubi_err(ubi, "bad data_size"); + goto bad; + } + } else if (lnum > used_ebs - 1) { + ubi_err(ubi, "too high lnum"); + goto bad; + } + } else { + if (copy_flag == 0) { + if (data_crc != 0) { + ubi_err(ubi, "non-zero data CRC"); + goto bad; + } + if (data_size != 0) { + ubi_err(ubi, "non-zero data_size"); + goto bad; + } + } else { + if (data_size == 0) { + ubi_err(ubi, "zero data_size of copy"); + goto bad; + } + } + if (used_ebs != 0) { + ubi_err(ubi, "bad used_ebs"); + goto bad; + } + } + + return 0; + +bad: + ubi_err(ubi, "bad VID header"); + ubi_dump_vid_hdr(vid_hdr); + dump_stack(); + return 1; +} + +/** + * ubi_io_read_vid_hdr - read and check a volume identifier header. + * @ubi: UBI device description object + * @pnum: physical eraseblock number to read from + * @vidb: the volume identifier buffer to store data in + * @verbose: be verbose if the header is corrupted or wasn't found + * + * This function reads the volume identifier header from physical eraseblock + * @pnum and stores it in @vidb. It also checks CRC checksum of the read + * volume identifier header. The error codes are the same as in + * 'ubi_io_read_ec_hdr()'. + * + * Note, the implementation of this function is also very similar to + * 'ubi_io_read_ec_hdr()', so refer commentaries in 'ubi_io_read_ec_hdr()'. + */ +int ubi_io_read_vid_hdr(struct ubi_device *ubi, int pnum, + struct ubi_vid_io_buf *vidb, int verbose) +{ + int err, read_err; + uint32_t crc, magic, hdr_crc; + struct ubi_vid_hdr *vid_hdr = ubi_get_vid_hdr(vidb); + void *p = vidb->buffer; + + dbg_io("read VID header from PEB %d", pnum); + ubi_assert(pnum >= 0 && pnum < ubi->peb_count); + + read_err = ubi_io_read(ubi, p, pnum, ubi->vid_hdr_aloffset, + ubi->vid_hdr_shift + UBI_VID_HDR_SIZE); + if (read_err && read_err != UBI_IO_BITFLIPS && !mtd_is_eccerr(read_err)) + return read_err; + + magic = be32_to_cpu(vid_hdr->magic); + if (magic != UBI_VID_HDR_MAGIC) { + if (mtd_is_eccerr(read_err)) + return UBI_IO_BAD_HDR_EBADMSG; + + if (ubi_check_pattern(vid_hdr, 0xFF, UBI_VID_HDR_SIZE)) { + if (verbose) + ubi_warn(ubi, "no VID header found at PEB %d, only 0xFF bytes", + pnum); + dbg_bld("no VID header found at PEB %d, only 0xFF bytes", + pnum); + if (!read_err) + return UBI_IO_FF; + else + return UBI_IO_FF_BITFLIPS; + } + + if (verbose) { + ubi_warn(ubi, "bad magic number at PEB %d: %08x instead of %08x", + pnum, magic, UBI_VID_HDR_MAGIC); + ubi_dump_vid_hdr(vid_hdr); + } + dbg_bld("bad magic number at PEB %d: %08x instead of %08x", + pnum, magic, UBI_VID_HDR_MAGIC); + return UBI_IO_BAD_HDR; + } + + crc = crc32(UBI_CRC32_INIT, vid_hdr, UBI_VID_HDR_SIZE_CRC); + hdr_crc = be32_to_cpu(vid_hdr->hdr_crc); + + if (hdr_crc != crc) { + if (verbose) { + ubi_warn(ubi, "bad CRC at PEB %d, calculated %#08x, read %#08x", + pnum, crc, hdr_crc); + ubi_dump_vid_hdr(vid_hdr); + } + dbg_bld("bad CRC at PEB %d, calculated %#08x, read %#08x", + pnum, crc, hdr_crc); + if (!read_err) + return UBI_IO_BAD_HDR; + else + return UBI_IO_BAD_HDR_EBADMSG; + } + + err = validate_vid_hdr(ubi, vid_hdr); + if (err) { + ubi_err(ubi, "validation failed for PEB %d", pnum); + return -EINVAL; + } + + if (read_err) + return UBI_IO_BITFLIPS; + + if (ubi_dbg_is_read_failure(ubi, MASK_READ_FAILURE_VID)) { + ubi_warn(ubi, "cannot read VID header from PEB %d (emulated)", + pnum); + return -EIO; + } + + if (ubi_dbg_is_ff(ubi, MASK_IO_FF_VID)) { + ubi_warn(ubi, "bit-all-ff (emulated)"); + return UBI_IO_FF; + } + + if (ubi_dbg_is_ff_bitflips(ubi, MASK_IO_FF_BITFLIPS_VID)) { + ubi_warn(ubi, "bit-all-ff with error reported by MTD driver (emulated)"); + return UBI_IO_FF_BITFLIPS; + } + + if (ubi_dbg_is_bad_hdr(ubi, MASK_BAD_HDR_VID)) { + ubi_warn(ubi, "bad_hdr (emulated)"); + return UBI_IO_BAD_HDR; + } + + if (ubi_dbg_is_bad_hdr_ebadmsg(ubi, MASK_BAD_HDR_EBADMSG_VID)) { + ubi_warn(ubi, "bad_hdr with ECC error (emulated)"); + return UBI_IO_BAD_HDR_EBADMSG; + } + + return 0; +} + +/** + * ubi_io_write_vid_hdr - write a volume identifier header. + * @ubi: UBI device description object + * @pnum: the physical eraseblock number to write to + * @vidb: the volume identifier buffer to write + * + * This function writes the volume identifier header described by @vid_hdr to + * physical eraseblock @pnum. This function automatically fills the + * @vidb->hdr->magic and the @vidb->hdr->version fields, as well as calculates + * header CRC checksum and stores it at vidb->hdr->hdr_crc. + * + * This function returns zero in case of success and a negative error code in + * case of failure. If %-EIO is returned, the physical eraseblock probably went + * bad. + */ +int ubi_io_write_vid_hdr(struct ubi_device *ubi, int pnum, + struct ubi_vid_io_buf *vidb) +{ + struct ubi_vid_hdr *vid_hdr = ubi_get_vid_hdr(vidb); + int err; + uint32_t crc; + void *p = vidb->buffer; + + dbg_io("write VID header to PEB %d", pnum); + ubi_assert(pnum >= 0 && pnum < ubi->peb_count); + + err = self_check_peb_ec_hdr(ubi, pnum); + if (err) + return err; + + vid_hdr->magic = cpu_to_be32(UBI_VID_HDR_MAGIC); + vid_hdr->version = UBI_VERSION; + crc = crc32(UBI_CRC32_INIT, vid_hdr, UBI_VID_HDR_SIZE_CRC); + vid_hdr->hdr_crc = cpu_to_be32(crc); + + err = self_check_vid_hdr(ubi, pnum, vid_hdr); + if (err) + return err; + + if (ubi_dbg_is_power_cut(ubi, MASK_POWER_CUT_VID)) { + ubi_warn(ubi, "emulating a power cut when writing VID header"); + ubi_ro_mode(ubi); + return -EROFS; + } + + if (ubi->vid_hdr_shift) { + memset((char *)p, 0xFF, ubi->vid_hdr_shift); + memset((char *)p + ubi->vid_hdr_shift + UBI_VID_HDR_SIZE, 0xFF, + ubi->vid_hdr_alsize - (ubi->vid_hdr_shift + UBI_VID_HDR_SIZE)); + } else { + memset((char *)p + UBI_VID_HDR_SIZE, 0xFF, ubi->vid_hdr_alsize - UBI_VID_HDR_SIZE); + } + + err = ubi_io_write(ubi, p, pnum, ubi->vid_hdr_aloffset, + ubi->vid_hdr_alsize); + return err; +} + +/** + * self_check_not_bad - ensure that a physical eraseblock is not bad. + * @ubi: UBI device description object + * @pnum: physical eraseblock number to check + * + * This function returns zero if the physical eraseblock is good, %-EINVAL if + * it is bad and a negative error code if an error occurred. + */ +static int self_check_not_bad(const struct ubi_device *ubi, int pnum) +{ + int err; + + if (!ubi_dbg_chk_io(ubi)) + return 0; + + err = ubi_io_is_bad(ubi, pnum); + if (!err) + return err; + + ubi_err(ubi, "self-check failed for PEB %d", pnum); + dump_stack(); + return err > 0 ? -EINVAL : err; +} + +/** + * self_check_ec_hdr - check if an erase counter header is all right. + * @ubi: UBI device description object + * @pnum: physical eraseblock number the erase counter header belongs to + * @ec_hdr: the erase counter header to check + * + * This function returns zero if the erase counter header contains valid + * values, and %-EINVAL if not. + */ +static int self_check_ec_hdr(const struct ubi_device *ubi, int pnum, + const struct ubi_ec_hdr *ec_hdr) +{ + int err; + uint32_t magic; + + if (!ubi_dbg_chk_io(ubi)) + return 0; + + magic = be32_to_cpu(ec_hdr->magic); + if (magic != UBI_EC_HDR_MAGIC) { + ubi_err(ubi, "bad magic %#08x, must be %#08x", + magic, UBI_EC_HDR_MAGIC); + goto fail; + } + + err = validate_ec_hdr(ubi, ec_hdr); + if (err) { + ubi_err(ubi, "self-check failed for PEB %d", pnum); + goto fail; + } + + return 0; + +fail: + ubi_dump_ec_hdr(ec_hdr); + dump_stack(); + return -EINVAL; +} + +/** + * self_check_peb_ec_hdr - check erase counter header. + * @ubi: UBI device description object + * @pnum: the physical eraseblock number to check + * + * This function returns zero if the erase counter header is all right and + * a negative error code if not or if an error occurred. + */ +static int self_check_peb_ec_hdr(const struct ubi_device *ubi, int pnum) +{ + int err; + uint32_t crc, hdr_crc; + struct ubi_ec_hdr *ec_hdr; + + if (!ubi_dbg_chk_io(ubi)) + return 0; + + ec_hdr = kzalloc(ubi->ec_hdr_alsize, GFP_NOFS); + if (!ec_hdr) + return -ENOMEM; + + err = ubi_io_read(ubi, ec_hdr, pnum, 0, UBI_EC_HDR_SIZE); + if (err && err != UBI_IO_BITFLIPS && !mtd_is_eccerr(err)) + goto exit; + + crc = crc32(UBI_CRC32_INIT, ec_hdr, UBI_EC_HDR_SIZE_CRC); + hdr_crc = be32_to_cpu(ec_hdr->hdr_crc); + if (hdr_crc != crc) { + ubi_err(ubi, "bad CRC, calculated %#08x, read %#08x", + crc, hdr_crc); + ubi_err(ubi, "self-check failed for PEB %d", pnum); + ubi_dump_ec_hdr(ec_hdr); + dump_stack(); + err = -EINVAL; + goto exit; + } + + err = self_check_ec_hdr(ubi, pnum, ec_hdr); + +exit: + kfree(ec_hdr); + return err; +} + +/** + * self_check_vid_hdr - check that a volume identifier header is all right. + * @ubi: UBI device description object + * @pnum: physical eraseblock number the volume identifier header belongs to + * @vid_hdr: the volume identifier header to check + * + * This function returns zero if the volume identifier header is all right, and + * %-EINVAL if not. + */ +static int self_check_vid_hdr(const struct ubi_device *ubi, int pnum, + const struct ubi_vid_hdr *vid_hdr) +{ + int err; + uint32_t magic; + + if (!ubi_dbg_chk_io(ubi)) + return 0; + + magic = be32_to_cpu(vid_hdr->magic); + if (magic != UBI_VID_HDR_MAGIC) { + ubi_err(ubi, "bad VID header magic %#08x at PEB %d, must be %#08x", + magic, pnum, UBI_VID_HDR_MAGIC); + goto fail; + } + + err = validate_vid_hdr(ubi, vid_hdr); + if (err) { + ubi_err(ubi, "self-check failed for PEB %d", pnum); + goto fail; + } + + return err; + +fail: + ubi_err(ubi, "self-check failed for PEB %d", pnum); + ubi_dump_vid_hdr(vid_hdr); + dump_stack(); + return -EINVAL; + +} + +/** + * self_check_peb_vid_hdr - check volume identifier header. + * @ubi: UBI device description object + * @pnum: the physical eraseblock number to check + * + * This function returns zero if the volume identifier header is all right, + * and a negative error code if not or if an error occurred. + */ +static int self_check_peb_vid_hdr(const struct ubi_device *ubi, int pnum) +{ + int err; + uint32_t crc, hdr_crc; + struct ubi_vid_io_buf *vidb; + struct ubi_vid_hdr *vid_hdr; + void *p; + + if (!ubi_dbg_chk_io(ubi)) + return 0; + + vidb = ubi_alloc_vid_buf(ubi, GFP_NOFS); + if (!vidb) + return -ENOMEM; + + vid_hdr = ubi_get_vid_hdr(vidb); + p = vidb->buffer; + err = ubi_io_read(ubi, p, pnum, ubi->vid_hdr_aloffset, + ubi->vid_hdr_alsize); + if (err && err != UBI_IO_BITFLIPS && !mtd_is_eccerr(err)) + goto exit; + + crc = crc32(UBI_CRC32_INIT, vid_hdr, UBI_VID_HDR_SIZE_CRC); + hdr_crc = be32_to_cpu(vid_hdr->hdr_crc); + if (hdr_crc != crc) { + ubi_err(ubi, "bad VID header CRC at PEB %d, calculated %#08x, read %#08x", + pnum, crc, hdr_crc); + ubi_err(ubi, "self-check failed for PEB %d", pnum); + ubi_dump_vid_hdr(vid_hdr); + dump_stack(); + err = -EINVAL; + goto exit; + } + + err = self_check_vid_hdr(ubi, pnum, vid_hdr); + +exit: + ubi_free_vid_buf(vidb); + return err; +} + +/** + * self_check_write - make sure write succeeded. + * @ubi: UBI device description object + * @buf: buffer with data which were written + * @pnum: physical eraseblock number the data were written to + * @offset: offset within the physical eraseblock the data were written to + * @len: how many bytes were written + * + * This functions reads data which were recently written and compares it with + * the original data buffer - the data have to match. Returns zero if the data + * match and a negative error code if not or in case of failure. + */ +static int self_check_write(struct ubi_device *ubi, const void *buf, int pnum, + int offset, int len) +{ + int err, i; + size_t read; + void *buf1; + loff_t addr = (loff_t)pnum * ubi->peb_size + offset; + + if (!ubi_dbg_chk_io(ubi)) + return 0; + + buf1 = __vmalloc(len, GFP_NOFS); + if (!buf1) { + ubi_err(ubi, "cannot allocate memory to check writes"); + return 0; + } + + err = mtd_read(ubi->mtd, addr, len, &read, buf1); + if (err && !mtd_is_bitflip(err)) + goto out_free; + + for (i = 0; i < len; i++) { + uint8_t c = ((uint8_t *)buf)[i]; + uint8_t c1 = ((uint8_t *)buf1)[i]; + int dump_len; + + if (c == c1) + continue; + + ubi_err(ubi, "self-check failed for PEB %d:%d, len %d", + pnum, offset, len); + ubi_msg(ubi, "data differ at position %d", i); + dump_len = max_t(int, 128, len - i); + ubi_msg(ubi, "hex dump of the original buffer from %d to %d", + i, i + dump_len); + print_hex_dump(KERN_DEBUG, "", DUMP_PREFIX_OFFSET, 32, 1, + buf + i, dump_len, 1); + ubi_msg(ubi, "hex dump of the read buffer from %d to %d", + i, i + dump_len); + print_hex_dump(KERN_DEBUG, "", DUMP_PREFIX_OFFSET, 32, 1, + buf1 + i, dump_len, 1); + dump_stack(); + err = -EINVAL; + goto out_free; + } + + vfree(buf1); + return 0; + +out_free: + vfree(buf1); + return err; +} + +/** + * ubi_self_check_all_ff - check that a region of flash is empty. + * @ubi: UBI device description object + * @pnum: the physical eraseblock number to check + * @offset: the starting offset within the physical eraseblock to check + * @len: the length of the region to check + * + * This function returns zero if only 0xFF bytes are present at offset + * @offset of the physical eraseblock @pnum, and a negative error code if not + * or if an error occurred. + */ +int ubi_self_check_all_ff(struct ubi_device *ubi, int pnum, int offset, int len) +{ + size_t read; + int err; + void *buf; + loff_t addr = (loff_t)pnum * ubi->peb_size + offset; + + if (!ubi_dbg_chk_io(ubi)) + return 0; + + buf = __vmalloc(len, GFP_NOFS); + if (!buf) { + ubi_err(ubi, "cannot allocate memory to check for 0xFFs"); + return 0; + } + + err = mtd_read(ubi->mtd, addr, len, &read, buf); + if (err && !mtd_is_bitflip(err)) { + ubi_err(ubi, "err %d while reading %d bytes from PEB %d:%d, read %zd bytes", + err, len, pnum, offset, read); + goto error; + } + + err = ubi_check_pattern(buf, 0xFF, len); + if (err == 0) { + ubi_err(ubi, "flash region at PEB %d:%d, length %d does not contain all 0xFF bytes", + pnum, offset, len); + goto fail; + } + + vfree(buf); + return 0; + +fail: + ubi_err(ubi, "self-check failed for PEB %d", pnum); + ubi_msg(ubi, "hex dump of the %d-%d region", offset, offset + len); + print_hex_dump(KERN_DEBUG, "", DUMP_PREFIX_OFFSET, 32, 1, buf, len, 1); + err = -EINVAL; +error: + dump_stack(); + vfree(buf); + return err; +} diff --git a/drivers/mtd/ubi/kapi.c b/drivers/mtd/ubi/kapi.c new file mode 100644 index 000000000..9626990f6 --- /dev/null +++ b/drivers/mtd/ubi/kapi.c @@ -0,0 +1,848 @@ +// SPDX-License-Identifier: GPL-2.0-or-later +/* + * Copyright (c) International Business Machines Corp., 2006 + * + * Author: Artem Bityutskiy (Битюцкий Артём) + */ + +/* This file mostly implements UBI kernel API functions */ + +#include +#include +#include +#include +#include +#include +#include "ubi.h" + +/** + * ubi_do_get_device_info - get information about UBI device. + * @ubi: UBI device description object + * @di: the information is stored here + * + * This function is the same as 'ubi_get_device_info()', but it assumes the UBI + * device is locked and cannot disappear. + */ +void ubi_do_get_device_info(struct ubi_device *ubi, struct ubi_device_info *di) +{ + di->ubi_num = ubi->ubi_num; + di->leb_size = ubi->leb_size; + di->leb_start = ubi->leb_start; + di->min_io_size = ubi->min_io_size; + di->max_write_size = ubi->max_write_size; + di->ro_mode = ubi->ro_mode; + di->cdev = ubi->cdev.dev; +} +EXPORT_SYMBOL_GPL(ubi_do_get_device_info); + +/** + * ubi_get_device_info - get information about UBI device. + * @ubi_num: UBI device number + * @di: the information is stored here + * + * This function returns %0 in case of success, %-EINVAL if the UBI device + * number is invalid, and %-ENODEV if there is no such UBI device. + */ +int ubi_get_device_info(int ubi_num, struct ubi_device_info *di) +{ + struct ubi_device *ubi; + + if (ubi_num < 0 || ubi_num >= UBI_MAX_DEVICES) + return -EINVAL; + ubi = ubi_get_device(ubi_num); + if (!ubi) + return -ENODEV; + ubi_do_get_device_info(ubi, di); + ubi_put_device(ubi); + return 0; +} +EXPORT_SYMBOL_GPL(ubi_get_device_info); + +/** + * ubi_do_get_volume_info - get information about UBI volume. + * @ubi: UBI device description object + * @vol: volume description object + * @vi: the information is stored here + */ +void ubi_do_get_volume_info(struct ubi_device *ubi, struct ubi_volume *vol, + struct ubi_volume_info *vi) +{ + vi->vol_id = vol->vol_id; + vi->ubi_num = ubi->ubi_num; + vi->size = vol->reserved_pebs; + vi->used_bytes = vol->used_bytes; + vi->vol_type = vol->vol_type; + vi->corrupted = vol->corrupted; + vi->upd_marker = vol->upd_marker; + vi->alignment = vol->alignment; + vi->usable_leb_size = vol->usable_leb_size; + vi->name_len = vol->name_len; + vi->name = vol->name; + vi->cdev = vol->cdev.dev; + vi->dev = &vol->dev; +} + +/** + * ubi_get_volume_info - get information about UBI volume. + * @desc: volume descriptor + * @vi: the information is stored here + */ +void ubi_get_volume_info(struct ubi_volume_desc *desc, + struct ubi_volume_info *vi) +{ + ubi_do_get_volume_info(desc->vol->ubi, desc->vol, vi); +} +EXPORT_SYMBOL_GPL(ubi_get_volume_info); + +/** + * ubi_open_volume - open UBI volume. + * @ubi_num: UBI device number + * @vol_id: volume ID + * @mode: open mode + * + * The @mode parameter specifies if the volume should be opened in read-only + * mode, read-write mode, or exclusive mode. The exclusive mode guarantees that + * nobody else will be able to open this volume. UBI allows to have many volume + * readers and one writer at a time. + * + * If a static volume is being opened for the first time since boot, it will be + * checked by this function, which means it will be fully read and the CRC + * checksum of each logical eraseblock will be checked. + * + * This function returns volume descriptor in case of success and a negative + * error code in case of failure. + */ +struct ubi_volume_desc *ubi_open_volume(int ubi_num, int vol_id, int mode) +{ + int err; + struct ubi_volume_desc *desc; + struct ubi_device *ubi; + struct ubi_volume *vol; + + dbg_gen("open device %d, volume %d, mode %d", ubi_num, vol_id, mode); + + if (ubi_num < 0 || ubi_num >= UBI_MAX_DEVICES) + return ERR_PTR(-EINVAL); + + if (mode != UBI_READONLY && mode != UBI_READWRITE && + mode != UBI_EXCLUSIVE && mode != UBI_METAONLY) + return ERR_PTR(-EINVAL); + + /* + * First of all, we have to get the UBI device to prevent its removal. + */ + ubi = ubi_get_device(ubi_num); + if (!ubi) + return ERR_PTR(-ENODEV); + + if (vol_id < 0 || vol_id >= ubi->vtbl_slots) { + err = -EINVAL; + goto out_put_ubi; + } + + desc = kmalloc_obj(struct ubi_volume_desc); + if (!desc) { + err = -ENOMEM; + goto out_put_ubi; + } + + err = -ENODEV; + if (!try_module_get(THIS_MODULE)) + goto out_free; + + spin_lock(&ubi->volumes_lock); + vol = ubi->volumes[vol_id]; + if (!vol || vol->is_dead) + goto out_unlock; + + err = -EBUSY; + switch (mode) { + case UBI_READONLY: + if (vol->exclusive) + goto out_unlock; + vol->readers += 1; + break; + + case UBI_READWRITE: + if (vol->exclusive || vol->writers > 0) + goto out_unlock; + vol->writers += 1; + break; + + case UBI_EXCLUSIVE: + if (vol->exclusive || vol->writers || vol->readers || + vol->metaonly) + goto out_unlock; + vol->exclusive = 1; + break; + + case UBI_METAONLY: + if (vol->metaonly || vol->exclusive) + goto out_unlock; + vol->metaonly = 1; + break; + } + get_device(&vol->dev); + vol->ref_count += 1; + spin_unlock(&ubi->volumes_lock); + + desc->vol = vol; + desc->mode = mode; + + mutex_lock(&ubi->ckvol_mutex); + if (!vol->checked && !vol->skip_check) { + /* This is the first open - check the volume */ + err = ubi_check_volume(ubi, vol_id); + if (err < 0) { + mutex_unlock(&ubi->ckvol_mutex); + ubi_close_volume(desc); + return ERR_PTR(err); + } + if (err == 1) { + ubi_warn(ubi, "volume %d on UBI device %d is corrupted", + vol_id, ubi->ubi_num); + vol->corrupted = 1; + } + vol->checked = 1; + } + mutex_unlock(&ubi->ckvol_mutex); + + return desc; + +out_unlock: + spin_unlock(&ubi->volumes_lock); + module_put(THIS_MODULE); +out_free: + kfree(desc); +out_put_ubi: + ubi_err(ubi, "cannot open device %d, volume %d, error %d", + ubi_num, vol_id, err); + ubi_put_device(ubi); + return ERR_PTR(err); +} +EXPORT_SYMBOL_GPL(ubi_open_volume); + +/** + * ubi_open_volume_nm - open UBI volume by name. + * @ubi_num: UBI device number + * @name: volume name + * @mode: open mode + * + * This function is similar to 'ubi_open_volume()', but opens a volume by name. + */ +struct ubi_volume_desc *ubi_open_volume_nm(int ubi_num, const char *name, + int mode) +{ + int i, vol_id = -1, len; + struct ubi_device *ubi; + struct ubi_volume_desc *ret; + + dbg_gen("open device %d, volume %s, mode %d", ubi_num, name, mode); + + if (!name) + return ERR_PTR(-EINVAL); + + len = strnlen(name, UBI_VOL_NAME_MAX + 1); + if (len > UBI_VOL_NAME_MAX) + return ERR_PTR(-EINVAL); + + if (ubi_num < 0 || ubi_num >= UBI_MAX_DEVICES) + return ERR_PTR(-EINVAL); + + ubi = ubi_get_device(ubi_num); + if (!ubi) + return ERR_PTR(-ENODEV); + + spin_lock(&ubi->volumes_lock); + /* Walk all volumes of this UBI device */ + for (i = 0; i < ubi->vtbl_slots; i++) { + struct ubi_volume *vol = ubi->volumes[i]; + + if (vol && len == vol->name_len && !strcmp(name, vol->name)) { + vol_id = i; + break; + } + } + spin_unlock(&ubi->volumes_lock); + + if (vol_id >= 0) + ret = ubi_open_volume(ubi_num, vol_id, mode); + else + ret = ERR_PTR(-ENODEV); + + /* + * We should put the UBI device even in case of success, because + * 'ubi_open_volume()' took a reference as well. + */ + ubi_put_device(ubi); + return ret; +} +EXPORT_SYMBOL_GPL(ubi_open_volume_nm); + +/** + * ubi_get_num_by_path - get UBI device and volume number from device path + * @pathname: volume character device node path + * @ubi_num: pointer to UBI device number to be set + * @vol_id: pointer to UBI volume ID to be set + * + * Returns 0 on success and sets ubi_num and vol_id, returns error otherwise. + */ +int ubi_get_num_by_path(const char *pathname, int *ubi_num, int *vol_id) +{ + int error; + struct path path; + struct kstat stat; + + error = kern_path(pathname, LOOKUP_FOLLOW, &path); + if (error) + return error; + + error = vfs_getattr(&path, &stat, STATX_TYPE, AT_STATX_SYNC_AS_STAT); + path_put(&path); + if (error) + return error; + + if (!S_ISCHR(stat.mode)) + return -EINVAL; + + *ubi_num = ubi_major2num(MAJOR(stat.rdev)); + *vol_id = MINOR(stat.rdev) - 1; + + if (*vol_id < 0 || *ubi_num < 0) + return -ENODEV; + + return 0; +} + +/** + * ubi_open_volume_path - open UBI volume by its character device node path. + * @pathname: volume character device node path + * @mode: open mode + * + * This function is similar to 'ubi_open_volume()', but opens a volume the path + * to its character device node. + */ +struct ubi_volume_desc *ubi_open_volume_path(const char *pathname, int mode) +{ + int error, ubi_num, vol_id; + + dbg_gen("open volume %s, mode %d", pathname, mode); + + if (!pathname || !*pathname) + return ERR_PTR(-EINVAL); + + error = ubi_get_num_by_path(pathname, &ubi_num, &vol_id); + if (error) + return ERR_PTR(error); + + return ubi_open_volume(ubi_num, vol_id, mode); +} +EXPORT_SYMBOL_GPL(ubi_open_volume_path); + +/** + * ubi_close_volume - close UBI volume. + * @desc: volume descriptor + */ +void ubi_close_volume(struct ubi_volume_desc *desc) +{ + struct ubi_volume *vol = desc->vol; + struct ubi_device *ubi = vol->ubi; + + dbg_gen("close device %d, volume %d, mode %d", + ubi->ubi_num, vol->vol_id, desc->mode); + + spin_lock(&ubi->volumes_lock); + switch (desc->mode) { + case UBI_READONLY: + vol->readers -= 1; + break; + case UBI_READWRITE: + vol->writers -= 1; + break; + case UBI_EXCLUSIVE: + vol->exclusive = 0; + break; + case UBI_METAONLY: + vol->metaonly = 0; + break; + } + vol->ref_count -= 1; + spin_unlock(&ubi->volumes_lock); + + kfree(desc); + put_device(&vol->dev); + ubi_put_device(ubi); + module_put(THIS_MODULE); +} +EXPORT_SYMBOL_GPL(ubi_close_volume); + +/** + * leb_read_sanity_check - does sanity checks on read requests. + * @desc: volume descriptor + * @lnum: logical eraseblock number to read from + * @offset: offset within the logical eraseblock to read from + * @len: how many bytes to read + * + * This function is used by ubi_leb_read() and ubi_leb_read_sg() + * to perform sanity checks. + */ +static int leb_read_sanity_check(struct ubi_volume_desc *desc, int lnum, + int offset, int len) +{ + struct ubi_volume *vol = desc->vol; + struct ubi_device *ubi = vol->ubi; + int vol_id = vol->vol_id; + + if (vol_id < 0 || vol_id >= ubi->vtbl_slots || lnum < 0 || + lnum >= vol->used_ebs || offset < 0 || len < 0 || + offset + len > vol->usable_leb_size) + return -EINVAL; + + if (vol->vol_type == UBI_STATIC_VOLUME) { + if (vol->used_ebs == 0) + /* Empty static UBI volume */ + return 0; + if (lnum == vol->used_ebs - 1 && + offset + len > vol->last_eb_bytes) + return -EINVAL; + } + + if (vol->upd_marker) + return -EBADF; + + return 0; +} + +/** + * ubi_leb_read - read data. + * @desc: volume descriptor + * @lnum: logical eraseblock number to read from + * @buf: buffer where to store the read data + * @offset: offset within the logical eraseblock to read from + * @len: how many bytes to read + * @check: whether UBI has to check the read data's CRC or not. + * + * This function reads data from offset @offset of logical eraseblock @lnum and + * stores the data at @buf. When reading from static volumes, @check specifies + * whether the data has to be checked or not. If yes, the whole logical + * eraseblock will be read and its CRC checksum will be checked (i.e., the CRC + * checksum is per-eraseblock). So checking may substantially slow down the + * read speed. The @check argument is ignored for dynamic volumes. + * + * In case of success, this function returns zero. In case of failure, this + * function returns a negative error code. + * + * %-EBADMSG error code is returned: + * o for both static and dynamic volumes if MTD driver has detected a data + * integrity problem (unrecoverable ECC checksum mismatch in case of NAND); + * o for static volumes in case of data CRC mismatch. + * + * If the volume is damaged because of an interrupted update this function just + * returns immediately with %-EBADF error code. + */ +int ubi_leb_read(struct ubi_volume_desc *desc, int lnum, char *buf, int offset, + int len, int check) +{ + struct ubi_volume *vol = desc->vol; + struct ubi_device *ubi = vol->ubi; + int err, vol_id = vol->vol_id; + + dbg_gen("read %d bytes from LEB %d:%d:%d", len, vol_id, lnum, offset); + + err = leb_read_sanity_check(desc, lnum, offset, len); + if (err < 0) + return err; + + if (len == 0) + return 0; + + err = ubi_eba_read_leb(ubi, vol, lnum, buf, offset, len, check); + if (err && mtd_is_eccerr(err) && vol->vol_type == UBI_STATIC_VOLUME) { + ubi_warn(ubi, "mark volume %d as corrupted", vol_id); + vol->corrupted = 1; + } + + return err; +} +EXPORT_SYMBOL_GPL(ubi_leb_read); + + +/** + * ubi_leb_read_sg - read data into a scatter gather list. + * @desc: volume descriptor + * @lnum: logical eraseblock number to read from + * @sgl: UBI scatter gather list to store the read data + * @offset: offset within the logical eraseblock to read from + * @len: how many bytes to read + * @check: whether UBI has to check the read data's CRC or not. + * + * This function works exactly like ubi_leb_read_sg(). But instead of + * storing the read data into a buffer it writes to an UBI scatter gather + * list. + */ +int ubi_leb_read_sg(struct ubi_volume_desc *desc, int lnum, struct ubi_sgl *sgl, + int offset, int len, int check) +{ + struct ubi_volume *vol = desc->vol; + struct ubi_device *ubi = vol->ubi; + int err, vol_id = vol->vol_id; + + dbg_gen("read %d bytes from LEB %d:%d:%d", len, vol_id, lnum, offset); + + err = leb_read_sanity_check(desc, lnum, offset, len); + if (err < 0) + return err; + + if (len == 0) + return 0; + + err = ubi_eba_read_leb_sg(ubi, vol, sgl, lnum, offset, len, check); + if (err && mtd_is_eccerr(err) && vol->vol_type == UBI_STATIC_VOLUME) { + ubi_warn(ubi, "mark volume %d as corrupted", vol_id); + vol->corrupted = 1; + } + + return err; +} +EXPORT_SYMBOL_GPL(ubi_leb_read_sg); + +/** + * ubi_leb_write - write data. + * @desc: volume descriptor + * @lnum: logical eraseblock number to write to + * @buf: data to write + * @offset: offset within the logical eraseblock where to write + * @len: how many bytes to write + * + * This function writes @len bytes of data from @buf to offset @offset of + * logical eraseblock @lnum. + * + * This function takes care of physical eraseblock write failures. If write to + * the physical eraseblock write operation fails, the logical eraseblock is + * re-mapped to another physical eraseblock, the data is recovered, and the + * write finishes. UBI has a pool of reserved physical eraseblocks for this. + * + * If all the data were successfully written, zero is returned. If an error + * occurred and UBI has not been able to recover from it, this function returns + * a negative error code. Note, in case of an error, it is possible that + * something was still written to the flash media, but that may be some + * garbage. + * + * If the volume is damaged because of an interrupted update this function just + * returns immediately with %-EBADF code. + */ +int ubi_leb_write(struct ubi_volume_desc *desc, int lnum, const void *buf, + int offset, int len) +{ + struct ubi_volume *vol = desc->vol; + struct ubi_device *ubi = vol->ubi; + int vol_id = vol->vol_id; + + dbg_gen("write %d bytes to LEB %d:%d:%d", len, vol_id, lnum, offset); + + if (vol_id < 0 || vol_id >= ubi->vtbl_slots) + return -EINVAL; + + if (desc->mode == UBI_READONLY || vol->vol_type == UBI_STATIC_VOLUME) + return -EROFS; + + if (!ubi_leb_valid(vol, lnum) || offset < 0 || len < 0 || + offset + len > vol->usable_leb_size || + offset & (ubi->min_io_size - 1) || len & (ubi->min_io_size - 1)) + return -EINVAL; + + if (vol->upd_marker) + return -EBADF; + + if (len == 0) + return 0; + + return ubi_eba_write_leb(ubi, vol, lnum, buf, offset, len); +} +EXPORT_SYMBOL_GPL(ubi_leb_write); + +/* + * ubi_leb_change - change logical eraseblock atomically. + * @desc: volume descriptor + * @lnum: logical eraseblock number to change + * @buf: data to write + * @len: how many bytes to write + * + * This function changes the contents of a logical eraseblock atomically. @buf + * has to contain new logical eraseblock data, and @len - the length of the + * data, which has to be aligned. The length may be shorter than the logical + * eraseblock size, ant the logical eraseblock may be appended to more times + * later on. This function guarantees that in case of an unclean reboot the old + * contents is preserved. Returns zero in case of success and a negative error + * code in case of failure. + */ +int ubi_leb_change(struct ubi_volume_desc *desc, int lnum, const void *buf, + int len) +{ + struct ubi_volume *vol = desc->vol; + struct ubi_device *ubi = vol->ubi; + int vol_id = vol->vol_id; + + dbg_gen("atomically write %d bytes to LEB %d:%d", len, vol_id, lnum); + + if (vol_id < 0 || vol_id >= ubi->vtbl_slots) + return -EINVAL; + + if (desc->mode == UBI_READONLY || vol->vol_type == UBI_STATIC_VOLUME) + return -EROFS; + + if (!ubi_leb_valid(vol, lnum) || len < 0 || + len > vol->usable_leb_size || len & (ubi->min_io_size - 1)) + return -EINVAL; + + if (vol->upd_marker) + return -EBADF; + + if (len == 0) + return 0; + + return ubi_eba_atomic_leb_change(ubi, vol, lnum, buf, len); +} +EXPORT_SYMBOL_GPL(ubi_leb_change); + +/** + * ubi_leb_erase - erase logical eraseblock. + * @desc: volume descriptor + * @lnum: logical eraseblock number + * + * This function un-maps logical eraseblock @lnum and synchronously erases the + * correspondent physical eraseblock. Returns zero in case of success and a + * negative error code in case of failure. + * + * If the volume is damaged because of an interrupted update this function just + * returns immediately with %-EBADF code. + */ +int ubi_leb_erase(struct ubi_volume_desc *desc, int lnum) +{ + struct ubi_volume *vol = desc->vol; + struct ubi_device *ubi = vol->ubi; + int err; + + dbg_gen("erase LEB %d:%d", vol->vol_id, lnum); + + if (desc->mode == UBI_READONLY || vol->vol_type == UBI_STATIC_VOLUME) + return -EROFS; + + if (!ubi_leb_valid(vol, lnum)) + return -EINVAL; + + if (vol->upd_marker) + return -EBADF; + + err = ubi_eba_unmap_leb(ubi, vol, lnum); + if (err) + return err; + + return ubi_wl_flush(ubi, vol->vol_id, lnum); +} +EXPORT_SYMBOL_GPL(ubi_leb_erase); + +/** + * ubi_leb_unmap - un-map logical eraseblock. + * @desc: volume descriptor + * @lnum: logical eraseblock number + * + * This function un-maps logical eraseblock @lnum and schedules the + * corresponding physical eraseblock for erasure, so that it will eventually be + * physically erased in background. This operation is much faster than the + * erase operation. + * + * Unlike erase, the un-map operation does not guarantee that the logical + * eraseblock will contain all 0xFF bytes when UBI is initialized again. For + * example, if several logical eraseblocks are un-mapped, and an unclean reboot + * happens after this, the logical eraseblocks will not necessarily be + * un-mapped again when this MTD device is attached. They may actually be + * mapped to the same physical eraseblocks again. So, this function has to be + * used with care. + * + * In other words, when un-mapping a logical eraseblock, UBI does not store + * any information about this on the flash media, it just marks the logical + * eraseblock as "un-mapped" in RAM. If UBI is detached before the physical + * eraseblock is physically erased, it will be mapped again to the same logical + * eraseblock when the MTD device is attached again. + * + * The main and obvious use-case of this function is when the contents of a + * logical eraseblock has to be re-written. Then it is much more efficient to + * first un-map it, then write new data, rather than first erase it, then write + * new data. Note, once new data has been written to the logical eraseblock, + * UBI guarantees that the old contents has gone forever. In other words, if an + * unclean reboot happens after the logical eraseblock has been un-mapped and + * then written to, it will contain the last written data. + * + * This function returns zero in case of success and a negative error code in + * case of failure. If the volume is damaged because of an interrupted update + * this function just returns immediately with %-EBADF code. + */ +int ubi_leb_unmap(struct ubi_volume_desc *desc, int lnum) +{ + struct ubi_volume *vol = desc->vol; + struct ubi_device *ubi = vol->ubi; + + dbg_gen("unmap LEB %d:%d", vol->vol_id, lnum); + + if (desc->mode == UBI_READONLY || vol->vol_type == UBI_STATIC_VOLUME) + return -EROFS; + + if (!ubi_leb_valid(vol, lnum)) + return -EINVAL; + + if (vol->upd_marker) + return -EBADF; + + return ubi_eba_unmap_leb(ubi, vol, lnum); +} +EXPORT_SYMBOL_GPL(ubi_leb_unmap); + +/** + * ubi_leb_map - map logical eraseblock to a physical eraseblock. + * @desc: volume descriptor + * @lnum: logical eraseblock number + * + * This function maps an un-mapped logical eraseblock @lnum to a physical + * eraseblock. This means, that after a successful invocation of this + * function the logical eraseblock @lnum will be empty (contain only %0xFF + * bytes) and be mapped to a physical eraseblock, even if an unclean reboot + * happens. + * + * This function returns zero in case of success, %-EBADF if the volume is + * damaged because of an interrupted update, %-EBADMSG if the logical + * eraseblock is already mapped, and other negative error codes in case of + * other failures. + */ +int ubi_leb_map(struct ubi_volume_desc *desc, int lnum) +{ + struct ubi_volume *vol = desc->vol; + struct ubi_device *ubi = vol->ubi; + + dbg_gen("map LEB %d:%d", vol->vol_id, lnum); + + if (desc->mode == UBI_READONLY || vol->vol_type == UBI_STATIC_VOLUME) + return -EROFS; + + if (!ubi_leb_valid(vol, lnum)) + return -EINVAL; + + if (vol->upd_marker) + return -EBADF; + + if (ubi_eba_is_mapped(vol, lnum)) + return -EBADMSG; + + return ubi_eba_write_leb(ubi, vol, lnum, NULL, 0, 0); +} +EXPORT_SYMBOL_GPL(ubi_leb_map); + +/** + * ubi_is_mapped - check if logical eraseblock is mapped. + * @desc: volume descriptor + * @lnum: logical eraseblock number + * + * This function checks if logical eraseblock @lnum is mapped to a physical + * eraseblock. If a logical eraseblock is un-mapped, this does not necessarily + * mean it will still be un-mapped after the UBI device is re-attached. The + * logical eraseblock may become mapped to the physical eraseblock it was last + * mapped to. + * + * This function returns %1 if the LEB is mapped, %0 if not, and a negative + * error code in case of failure. If the volume is damaged because of an + * interrupted update this function just returns immediately with %-EBADF error + * code. + */ +int ubi_is_mapped(struct ubi_volume_desc *desc, int lnum) +{ + struct ubi_volume *vol = desc->vol; + + dbg_gen("test LEB %d:%d", vol->vol_id, lnum); + + if (!ubi_leb_valid(vol, lnum)) + return -EINVAL; + + if (vol->upd_marker) + return -EBADF; + + return ubi_eba_is_mapped(vol, lnum); +} +EXPORT_SYMBOL_GPL(ubi_is_mapped); + +/** + * ubi_sync - synchronize UBI device buffers. + * @ubi_num: UBI device to synchronize + * + * The underlying MTD device may cache data in hardware or in software. This + * function ensures the caches are flushed. Returns zero in case of success and + * a negative error code in case of failure. + */ +int ubi_sync(int ubi_num) +{ + struct ubi_device *ubi; + + ubi = ubi_get_device(ubi_num); + if (!ubi) + return -ENODEV; + + mtd_sync(ubi->mtd); + ubi_put_device(ubi); + return 0; +} +EXPORT_SYMBOL_GPL(ubi_sync); + +BLOCKING_NOTIFIER_HEAD(ubi_notifiers); + +/** + * ubi_register_volume_notifier - register a volume notifier. + * @nb: the notifier description object + * @ignore_existing: if non-zero, do not send "added" notification for all + * already existing volumes + * + * This function registers a volume notifier, which means that + * 'nb->notifier_call()' will be invoked when an UBI volume is created, + * removed, re-sized, re-named, or updated. The first argument of the function + * is the notification type. The second argument is pointer to a + * &struct ubi_notification object which describes the notification event. + * Using UBI API from the volume notifier is prohibited. + * + * This function returns zero in case of success and a negative error code + * in case of failure. + */ +int ubi_register_volume_notifier(struct notifier_block *nb, + int ignore_existing) +{ + int err; + + err = blocking_notifier_chain_register(&ubi_notifiers, nb); + if (err != 0) + return err; + if (ignore_existing) + return 0; + + /* + * We are going to walk all UBI devices and all volumes, and + * notify the user about existing volumes by the %UBI_VOLUME_ADDED + * event. We have to lock the @ubi_devices_mutex to make sure UBI + * devices do not disappear. + */ + mutex_lock(&ubi_devices_mutex); + ubi_enumerate_volumes(nb); + mutex_unlock(&ubi_devices_mutex); + + return err; +} +EXPORT_SYMBOL_GPL(ubi_register_volume_notifier); + +/** + * ubi_unregister_volume_notifier - unregister the volume notifier. + * @nb: the notifier description object + * + * This function unregisters volume notifier @nm and returns zero in case of + * success and a negative error code in case of failure. + */ +int ubi_unregister_volume_notifier(struct notifier_block *nb) +{ + return blocking_notifier_chain_unregister(&ubi_notifiers, nb); +} +EXPORT_SYMBOL_GPL(ubi_unregister_volume_notifier); diff --git a/drivers/mtd/ubi/misc.c b/drivers/mtd/ubi/misc.c new file mode 100644 index 000000000..1794d66b6 --- /dev/null +++ b/drivers/mtd/ubi/misc.c @@ -0,0 +1,191 @@ +// SPDX-License-Identifier: GPL-2.0-or-later +/* + * Copyright (c) International Business Machines Corp., 2006 + * + * Author: Artem Bityutskiy (Битюцкий Артём) + */ + +/* Here we keep miscellaneous functions which are used all over the UBI code */ + +#include "ubi.h" + +/** + * ubi_calc_data_len - calculate how much real data is stored in a buffer. + * @ubi: UBI device description object + * @buf: a buffer with the contents of the physical eraseblock + * @length: the buffer length + * + * This function calculates how much "real data" is stored in @buf and returnes + * the length. Continuous 0xFF bytes at the end of the buffer are not + * considered as "real data". + */ +int ubi_calc_data_len(const struct ubi_device *ubi, const void *buf, + int length) +{ + int i; + + ubi_assert(!(length & (ubi->min_io_size - 1))); + + for (i = length - 1; i >= 0; i--) + if (((const uint8_t *)buf)[i] != 0xFF) + break; + + /* The resulting length must be aligned to the minimum flash I/O size */ + length = ALIGN(i + 1, ubi->min_io_size); + return length; +} + +/** + * ubi_check_volume - check the contents of a static volume. + * @ubi: UBI device description object + * @vol_id: ID of the volume to check + * + * This function checks if static volume @vol_id is corrupted by fully reading + * it and checking data CRC. This function returns %0 if the volume is not + * corrupted, %1 if it is corrupted and a negative error code in case of + * failure. Dynamic volumes are not checked and zero is returned immediately. + */ +int ubi_check_volume(struct ubi_device *ubi, int vol_id) +{ + void *buf; + int err = 0, i; + struct ubi_volume *vol = ubi->volumes[vol_id]; + + if (vol->vol_type != UBI_STATIC_VOLUME) + return 0; + + buf = vmalloc(vol->usable_leb_size); + if (!buf) + return -ENOMEM; + + for (i = 0; i < vol->used_ebs; i++) { + int size; + + cond_resched(); + + if (i == vol->used_ebs - 1) + size = vol->last_eb_bytes; + else + size = vol->usable_leb_size; + + err = ubi_eba_read_leb(ubi, vol, i, buf, 0, size, 1); + if (err) { + if (mtd_is_eccerr(err)) + err = 1; + break; + } + } + + vfree(buf); + return err; +} + +/** + * ubi_update_reserved - update bad eraseblock handling accounting data. + * @ubi: UBI device description object + * + * This function calculates the gap between current number of PEBs reserved for + * bad eraseblock handling and the required level of PEBs that must be + * reserved, and if necessary, reserves more PEBs to fill that gap, according + * to availability. Should be called with ubi->volumes_lock held. + */ +void ubi_update_reserved(struct ubi_device *ubi) +{ + int need = ubi->beb_rsvd_level - ubi->beb_rsvd_pebs; + + if (need <= 0 || ubi->avail_pebs == 0) + return; + + need = min_t(int, need, ubi->avail_pebs); + ubi->avail_pebs -= need; + ubi->rsvd_pebs += need; + ubi->beb_rsvd_pebs += need; + ubi_msg(ubi, "reserved more %d PEBs for bad PEB handling", need); +} + +/** + * ubi_calculate_reserved - calculate how many PEBs must be reserved for bad + * eraseblock handling. + * @ubi: UBI device description object + */ +void ubi_calculate_reserved(struct ubi_device *ubi) +{ + /* + * Calculate the actual number of PEBs currently needed to be reserved + * for future bad eraseblock handling. + */ + ubi->beb_rsvd_level = ubi->bad_peb_limit - ubi->bad_peb_count; + if (ubi->beb_rsvd_level < 0) { + ubi->beb_rsvd_level = 0; + ubi_warn(ubi, "number of bad PEBs (%d) is above the expected limit (%d), not reserving any PEBs for bad PEB handling, will use available PEBs (if any)", + ubi->bad_peb_count, ubi->bad_peb_limit); + } +} + +/** + * ubi_check_pattern - check if buffer contains only a certain byte pattern. + * @buf: buffer to check + * @patt: the pattern to check + * @size: buffer size in bytes + * + * This function returns %1 in there are only @patt bytes in @buf, and %0 if + * something else was also found. + */ +int ubi_check_pattern(const void *buf, uint8_t patt, int size) +{ + int i; + + for (i = 0; i < size; i++) + if (((const uint8_t *)buf)[i] != patt) + return 0; + return 1; +} + +/* Normal UBI messages */ +void ubi_msg(const struct ubi_device *ubi, const char *fmt, ...) +{ + struct va_format vaf; + va_list args; + + va_start(args, fmt); + + vaf.fmt = fmt; + vaf.va = &args; + + pr_notice(UBI_NAME_STR "%d: %pV\n", ubi->ubi_num, &vaf); + + va_end(args); +} + +/* UBI warning messages */ +void ubi_warn(const struct ubi_device *ubi, const char *fmt, ...) +{ + struct va_format vaf; + va_list args; + + va_start(args, fmt); + + vaf.fmt = fmt; + vaf.va = &args; + + pr_warn(UBI_NAME_STR "%d warning: %ps: %pV\n", + ubi->ubi_num, __builtin_return_address(0), &vaf); + + va_end(args); +} + +/* UBI error messages */ +void ubi_err(const struct ubi_device *ubi, const char *fmt, ...) +{ + struct va_format vaf; + va_list args; + + va_start(args, fmt); + + vaf.fmt = fmt; + vaf.va = &args; + + pr_err(UBI_NAME_STR "%d error: %ps: %pV\n", + ubi->ubi_num, __builtin_return_address(0), &vaf); + va_end(args); +} diff --git a/drivers/mtd/ubi/nvmem.c b/drivers/mtd/ubi/nvmem.c new file mode 100644 index 000000000..2dade9727 --- /dev/null +++ b/drivers/mtd/ubi/nvmem.c @@ -0,0 +1,191 @@ +// SPDX-License-Identifier: GPL-2.0-or-later +/* + * Copyright (c) 2023 Daniel Golle + */ + +/* UBI NVMEM provider */ +#include "ubi.h" +#include + +/* List of all NVMEM devices */ +static LIST_HEAD(nvmem_devices); +static DEFINE_MUTEX(devices_mutex); + +struct ubi_nvmem { + struct nvmem_device *nvmem; + int ubi_num; + int vol_id; + int usable_leb_size; + struct list_head list; +}; + +static int ubi_nvmem_reg_read(void *priv, unsigned int from, + void *val, size_t bytes) +{ + size_t to_read, bytes_left = bytes; + struct ubi_nvmem *unv = priv; + struct ubi_volume_desc *desc; + uint32_t offs; + uint32_t lnum; + int err = 0; + + desc = ubi_open_volume(unv->ubi_num, unv->vol_id, UBI_READONLY); + if (IS_ERR(desc)) + return PTR_ERR(desc); + + offs = from % unv->usable_leb_size; + lnum = from / unv->usable_leb_size; + while (bytes_left) { + to_read = unv->usable_leb_size - offs; + + if (to_read > bytes_left) + to_read = bytes_left; + + err = ubi_read(desc, lnum, val, offs, to_read); + if (err) + break; + + lnum += 1; + offs = 0; + bytes_left -= to_read; + val += to_read; + } + ubi_close_volume(desc); + + if (err) + return err; + + return 0; +} + +static int ubi_nvmem_add(struct ubi_volume_info *vi) +{ + struct device_node *np = dev_of_node(vi->dev); + struct nvmem_config config = {}; + struct ubi_nvmem *unv; + int ret; + + if (!np) + return 0; + + if (!of_get_child_by_name(np, "nvmem-layout")) + return 0; + + if (WARN_ON_ONCE(vi->usable_leb_size <= 0) || + WARN_ON_ONCE(vi->size <= 0)) + return -EINVAL; + + unv = kzalloc_obj(struct ubi_nvmem); + if (!unv) + return -ENOMEM; + + config.id = NVMEM_DEVID_NONE; + config.dev = vi->dev; + config.name = dev_name(vi->dev); + config.owner = THIS_MODULE; + config.priv = unv; + config.reg_read = ubi_nvmem_reg_read; + config.size = vi->usable_leb_size * vi->size; + config.word_size = 1; + config.stride = 1; + config.read_only = true; + config.root_only = true; + config.ignore_wp = true; + config.of_node = np; + + unv->ubi_num = vi->ubi_num; + unv->vol_id = vi->vol_id; + unv->usable_leb_size = vi->usable_leb_size; + unv->nvmem = nvmem_register(&config); + if (IS_ERR(unv->nvmem)) { + ret = dev_err_probe(vi->dev, PTR_ERR(unv->nvmem), + "Failed to register NVMEM device\n"); + kfree(unv); + return ret; + } + + mutex_lock(&devices_mutex); + list_add_tail(&unv->list, &nvmem_devices); + mutex_unlock(&devices_mutex); + + return 0; +} + +static void ubi_nvmem_remove(struct ubi_volume_info *vi) +{ + struct ubi_nvmem *unv_c, *unv = NULL; + + mutex_lock(&devices_mutex); + list_for_each_entry(unv_c, &nvmem_devices, list) + if (unv_c->ubi_num == vi->ubi_num && unv_c->vol_id == vi->vol_id) { + unv = unv_c; + break; + } + + if (!unv) { + mutex_unlock(&devices_mutex); + return; + } + + list_del(&unv->list); + mutex_unlock(&devices_mutex); + nvmem_unregister(unv->nvmem); + kfree(unv); +} + +/** + * nvmem_notify - UBI notification handler. + * @nb: registered notifier block + * @l: notification type + * @ns_ptr: pointer to the &struct ubi_notification object + */ +static int nvmem_notify(struct notifier_block *nb, unsigned long l, + void *ns_ptr) +{ + struct ubi_notification *nt = ns_ptr; + + switch (l) { + case UBI_VOLUME_RESIZED: + ubi_nvmem_remove(&nt->vi); + fallthrough; + case UBI_VOLUME_ADDED: + ubi_nvmem_add(&nt->vi); + break; + case UBI_VOLUME_SHUTDOWN: + ubi_nvmem_remove(&nt->vi); + break; + default: + break; + } + return NOTIFY_OK; +} + +static struct notifier_block nvmem_notifier = { + .notifier_call = nvmem_notify, +}; + +static int __init ubi_nvmem_init(void) +{ + return ubi_register_volume_notifier(&nvmem_notifier, 0); +} + +static void __exit ubi_nvmem_exit(void) +{ + struct ubi_nvmem *unv, *tmp; + + mutex_lock(&devices_mutex); + list_for_each_entry_safe(unv, tmp, &nvmem_devices, list) { + nvmem_unregister(unv->nvmem); + list_del(&unv->list); + kfree(unv); + } + mutex_unlock(&devices_mutex); + + ubi_unregister_volume_notifier(&nvmem_notifier); +} + +module_init(ubi_nvmem_init); +module_exit(ubi_nvmem_exit); +MODULE_DESCRIPTION("NVMEM layer over UBI volumes"); +MODULE_AUTHOR("Daniel Golle"); +MODULE_LICENSE("GPL"); diff --git a/drivers/mtd/ubi/ubi-media.h b/drivers/mtd/ubi/ubi-media.h new file mode 100644 index 000000000..2c9cd3b64 --- /dev/null +++ b/drivers/mtd/ubi/ubi-media.h @@ -0,0 +1,503 @@ +/* SPDX-License-Identifier: (GPL-2.0+ OR BSD-3-Clause) */ +/* + * Copyright (C) International Business Machines Corp., 2006 + * Authors: Artem Bityutskiy (Битюцкий Артём) + * Thomas Gleixner + * Frank Haverkamp + * Oliver Lohmann + * Andreas Arnez + * + * This file defines the layout of UBI headers and all the other UBI on-flash + * data structures. + */ + +#ifndef __UBI_MEDIA_H__ +#define __UBI_MEDIA_H__ + +#include + +/* The version of UBI images supported by this implementation */ +#define UBI_VERSION 1 + +/* The highest erase counter value supported by this implementation */ +#define UBI_MAX_ERASECOUNTER 0x7FFFFFFF + +/* The initial CRC32 value used when calculating CRC checksums */ +#define UBI_CRC32_INIT 0xFFFFFFFFU + +/* Erase counter header magic number (ASCII "UBI#") */ +#define UBI_EC_HDR_MAGIC 0x55424923 +/* Volume identifier header magic number (ASCII "UBI!") */ +#define UBI_VID_HDR_MAGIC 0x55424921 + +/* + * Volume type constants used in the volume identifier header. + * + * @UBI_VID_DYNAMIC: dynamic volume + * @UBI_VID_STATIC: static volume + */ +enum { + UBI_VID_DYNAMIC = 1, + UBI_VID_STATIC = 2 +}; + +/* + * Volume flags used in the volume table record. + * + * @UBI_VTBL_AUTORESIZE_FLG: auto-resize this volume + * @UBI_VTBL_SKIP_CRC_CHECK_FLG: skip the CRC check done on a static volume at + * open time. Should only be set on volumes that + * are used by upper layers doing this kind of + * check. Main use-case for this flag is + * boot-time reduction + * + * %UBI_VTBL_AUTORESIZE_FLG flag can be set only for one volume in the volume + * table. UBI automatically re-sizes the volume which has this flag and makes + * the volume to be of largest possible size. This means that if after the + * initialization UBI finds out that there are available physical eraseblocks + * present on the device, it automatically appends all of them to the volume + * (the physical eraseblocks reserved for bad eraseblocks handling and other + * reserved physical eraseblocks are not taken). So, if there is a volume with + * the %UBI_VTBL_AUTORESIZE_FLG flag set, the amount of available logical + * eraseblocks will be zero after UBI is loaded, because all of them will be + * reserved for this volume. Note, the %UBI_VTBL_AUTORESIZE_FLG bit is cleared + * after the volume had been initialized. + * + * The auto-resize feature is useful for device production purposes. For + * example, different NAND flash chips may have different amount of initial bad + * eraseblocks, depending of particular chip instance. Manufacturers of NAND + * chips usually guarantee that the amount of initial bad eraseblocks does not + * exceed certain percent, e.g. 2%. When one creates an UBI image which will be + * flashed to the end devices in production, he does not know the exact amount + * of good physical eraseblocks the NAND chip on the device will have, but this + * number is required to calculate the volume sized and put them to the volume + * table of the UBI image. In this case, one of the volumes (e.g., the one + * which will store the root file system) is marked as "auto-resizable", and + * UBI will adjust its size on the first boot if needed. + * + * Note, first UBI reserves some amount of physical eraseblocks for bad + * eraseblock handling, and then re-sizes the volume, not vice-versa. This + * means that the pool of reserved physical eraseblocks will always be present. + */ +enum { + UBI_VTBL_AUTORESIZE_FLG = 0x01, + UBI_VTBL_SKIP_CRC_CHECK_FLG = 0x02, +}; + +/* + * Compatibility constants used by internal volumes. + * + * @UBI_COMPAT_DELETE: delete this internal volume before anything is written + * to the flash + * @UBI_COMPAT_RO: attach this device in read-only mode + * @UBI_COMPAT_PRESERVE: preserve this internal volume - do not touch its + * physical eraseblocks, don't allow the wear-leveling + * sub-system to move them + * @UBI_COMPAT_REJECT: reject this UBI image + */ +enum { + UBI_COMPAT_DELETE = 1, + UBI_COMPAT_RO = 2, + UBI_COMPAT_PRESERVE = 4, + UBI_COMPAT_REJECT = 5 +}; + +/* Sizes of UBI headers */ +#define UBI_EC_HDR_SIZE sizeof(struct ubi_ec_hdr) +#define UBI_VID_HDR_SIZE sizeof(struct ubi_vid_hdr) + +/* Sizes of UBI headers without the ending CRC */ +#define UBI_EC_HDR_SIZE_CRC (UBI_EC_HDR_SIZE - sizeof(__be32)) +#define UBI_VID_HDR_SIZE_CRC (UBI_VID_HDR_SIZE - sizeof(__be32)) + +/** + * struct ubi_ec_hdr - UBI erase counter header. + * @magic: erase counter header magic number (%UBI_EC_HDR_MAGIC) + * @version: version of UBI implementation which is supposed to accept this + * UBI image + * @padding1: reserved for future, zeroes + * @ec: the erase counter + * @vid_hdr_offset: where the VID header starts + * @data_offset: where the user data start + * @image_seq: image sequence number + * @padding2: reserved for future, zeroes + * @hdr_crc: erase counter header CRC checksum + * + * The erase counter header takes 64 bytes and has a plenty of unused space for + * future usage. The unused fields are zeroed. The @version field is used to + * indicate the version of UBI implementation which is supposed to be able to + * work with this UBI image. If @version is greater than the current UBI + * version, the image is rejected. This may be useful in future if something + * is changed radically. This field is duplicated in the volume identifier + * header. + * + * The @vid_hdr_offset and @data_offset fields contain the offset of the + * volume identifier header and user data, relative to the beginning of the + * physical eraseblock. These values have to be the same for all physical + * eraseblocks. + * + * The @image_seq field is used to validate a UBI image that has been prepared + * for a UBI device. The @image_seq value can be any value, but it must be the + * same on all eraseblocks. UBI will ensure that all new erase counter headers + * also contain this value, and will check the value when attaching the flash. + * One way to make use of @image_seq is to increase its value by one every time + * an image is flashed over an existing image, then, if the flashing does not + * complete, UBI will detect the error when attaching the media. + */ +struct ubi_ec_hdr { + __be32 magic; + __u8 version; + __u8 padding1[3]; + __be64 ec; /* Warning: the current limit is 31-bit anyway! */ + __be32 vid_hdr_offset; + __be32 data_offset; + __be32 image_seq; + __u8 padding2[32]; + __be32 hdr_crc; +} __packed; + +/** + * struct ubi_vid_hdr - on-flash UBI volume identifier header. + * @magic: volume identifier header magic number (%UBI_VID_HDR_MAGIC) + * @version: UBI implementation version which is supposed to accept this UBI + * image (%UBI_VERSION) + * @vol_type: volume type (%UBI_VID_DYNAMIC or %UBI_VID_STATIC) + * @copy_flag: if this logical eraseblock was copied from another physical + * eraseblock (for wear-leveling reasons) + * @compat: compatibility of this volume (%0, %UBI_COMPAT_DELETE, + * %UBI_COMPAT_IGNORE, %UBI_COMPAT_PRESERVE, or %UBI_COMPAT_REJECT) + * @vol_id: ID of this volume + * @lnum: logical eraseblock number + * @padding1: reserved for future, zeroes + * @data_size: how many bytes of data this logical eraseblock contains + * @used_ebs: total number of used logical eraseblocks in this volume + * @data_pad: how many bytes at the end of this physical eraseblock are not + * used + * @data_crc: CRC checksum of the data stored in this logical eraseblock + * @padding2: reserved for future, zeroes + * @sqnum: sequence number + * @padding3: reserved for future, zeroes + * @hdr_crc: volume identifier header CRC checksum + * + * The @sqnum is the value of the global sequence counter at the time when this + * VID header was created. The global sequence counter is incremented each time + * UBI writes a new VID header to the flash, i.e. when it maps a logical + * eraseblock to a new physical eraseblock. The global sequence counter is an + * unsigned 64-bit integer and we assume it never overflows. The @sqnum + * (sequence number) is used to distinguish between older and newer versions of + * logical eraseblocks. + * + * There are 2 situations when there may be more than one physical eraseblock + * corresponding to the same logical eraseblock, i.e., having the same @vol_id + * and @lnum values in the volume identifier header. Suppose we have a logical + * eraseblock L and it is mapped to the physical eraseblock P. + * + * 1. Because UBI may erase physical eraseblocks asynchronously, the following + * situation is possible: L is asynchronously erased, so P is scheduled for + * erasure, then L is written to,i.e. mapped to another physical eraseblock P1, + * so P1 is written to, then an unclean reboot happens. Result - there are 2 + * physical eraseblocks P and P1 corresponding to the same logical eraseblock + * L. But P1 has greater sequence number, so UBI picks P1 when it attaches the + * flash. + * + * 2. From time to time UBI moves logical eraseblocks to other physical + * eraseblocks for wear-leveling reasons. If, for example, UBI moves L from P + * to P1, and an unclean reboot happens before P is physically erased, there + * are two physical eraseblocks P and P1 corresponding to L and UBI has to + * select one of them when the flash is attached. The @sqnum field says which + * PEB is the original (obviously P will have lower @sqnum) and the copy. But + * it is not enough to select the physical eraseblock with the higher sequence + * number, because the unclean reboot could have happen in the middle of the + * copying process, so the data in P is corrupted. It is also not enough to + * just select the physical eraseblock with lower sequence number, because the + * data there may be old (consider a case if more data was added to P1 after + * the copying). Moreover, the unclean reboot may happen when the erasure of P + * was just started, so it result in unstable P, which is "mostly" OK, but + * still has unstable bits. + * + * UBI uses the @copy_flag field to indicate that this logical eraseblock is a + * copy. UBI also calculates data CRC when the data is moved and stores it at + * the @data_crc field of the copy (P1). So when UBI needs to pick one physical + * eraseblock of two (P or P1), the @copy_flag of the newer one (P1) is + * examined. If it is cleared, the situation is simple and the newer one is + * picked. If it is set, the data CRC of the copy (P1) is examined. If the CRC + * checksum is correct, this physical eraseblock is selected (P1). Otherwise + * the older one (P) is selected. + * + * There are 2 sorts of volumes in UBI: user volumes and internal volumes. + * Internal volumes are not seen from outside and are used for various internal + * UBI purposes. In this implementation there is only one internal volume - the + * layout volume. Internal volumes are the main mechanism of UBI extensions. + * For example, in future one may introduce a journal internal volume. Internal + * volumes have their own reserved range of IDs. + * + * The @compat field is only used for internal volumes and contains the "degree + * of their compatibility". It is always zero for user volumes. This field + * provides a mechanism to introduce UBI extensions and to be still compatible + * with older UBI binaries. For example, if someone introduced a journal in + * future, he would probably use %UBI_COMPAT_DELETE compatibility for the + * journal volume. And in this case, older UBI binaries, which know nothing + * about the journal volume, would just delete this volume and work perfectly + * fine. This is similar to what Ext2fs does when it is fed by an Ext3fs image + * - it just ignores the Ext3fs journal. + * + * The @data_crc field contains the CRC checksum of the contents of the logical + * eraseblock if this is a static volume. In case of dynamic volumes, it does + * not contain the CRC checksum as a rule. The only exception is when the + * data of the physical eraseblock was moved by the wear-leveling sub-system, + * then the wear-leveling sub-system calculates the data CRC and stores it in + * the @data_crc field. And of course, the @copy_flag is %in this case. + * + * The @data_size field is used only for static volumes because UBI has to know + * how many bytes of data are stored in this eraseblock. For dynamic volumes, + * this field usually contains zero. The only exception is when the data of the + * physical eraseblock was moved to another physical eraseblock for + * wear-leveling reasons. In this case, UBI calculates CRC checksum of the + * contents and uses both @data_crc and @data_size fields. In this case, the + * @data_size field contains data size. + * + * The @used_ebs field is used only for static volumes and indicates how many + * eraseblocks the data of the volume takes. For dynamic volumes this field is + * not used and always contains zero. + * + * The @data_pad is calculated when volumes are created using the alignment + * parameter. So, effectively, the @data_pad field reduces the size of logical + * eraseblocks of this volume. This is very handy when one uses block-oriented + * software (say, cramfs) on top of the UBI volume. + */ +struct ubi_vid_hdr { + __be32 magic; + __u8 version; + __u8 vol_type; + __u8 copy_flag; + __u8 compat; + __be32 vol_id; + __be32 lnum; + __u8 padding1[4]; + __be32 data_size; + __be32 used_ebs; + __be32 data_pad; + __be32 data_crc; + __u8 padding2[4]; + __be64 sqnum; + __u8 padding3[12]; + __be32 hdr_crc; +} __packed; + +/* Internal UBI volumes count */ +#define UBI_INT_VOL_COUNT 1 + +/* + * Starting ID of internal volumes: 0x7fffefff. + * There is reserved room for 4096 internal volumes. + */ +#define UBI_INTERNAL_VOL_START (0x7FFFFFFF - 4096) + +/* The layout volume contains the volume table */ + +#define UBI_LAYOUT_VOLUME_ID UBI_INTERNAL_VOL_START +#define UBI_LAYOUT_VOLUME_TYPE UBI_VID_DYNAMIC +#define UBI_LAYOUT_VOLUME_ALIGN 1 +#define UBI_LAYOUT_VOLUME_EBS 2 +#define UBI_LAYOUT_VOLUME_NAME "layout volume" +#define UBI_LAYOUT_VOLUME_COMPAT UBI_COMPAT_REJECT + +/* The maximum number of volumes per one UBI device */ +#define UBI_MAX_VOLUMES 128 + +/* The maximum volume name length */ +#define UBI_VOL_NAME_MAX 127 + +/* Size of the volume table record */ +#define UBI_VTBL_RECORD_SIZE sizeof(struct ubi_vtbl_record) + +/* Size of the volume table record without the ending CRC */ +#define UBI_VTBL_RECORD_SIZE_CRC (UBI_VTBL_RECORD_SIZE - sizeof(__be32)) + +/** + * struct ubi_vtbl_record - a record in the volume table. + * @reserved_pebs: how many physical eraseblocks are reserved for this volume + * @alignment: volume alignment + * @data_pad: how many bytes are unused at the end of the each physical + * eraseblock to satisfy the requested alignment + * @vol_type: volume type (%UBI_DYNAMIC_VOLUME or %UBI_STATIC_VOLUME) + * @upd_marker: if volume update was started but not finished + * @name_len: volume name length + * @name: the volume name + * @flags: volume flags (%UBI_VTBL_AUTORESIZE_FLG) + * @padding: reserved, zeroes + * @crc: a CRC32 checksum of the record + * + * The volume table records are stored in the volume table, which is stored in + * the layout volume. The layout volume consists of 2 logical eraseblock, each + * of which contains a copy of the volume table (i.e., the volume table is + * duplicated). The volume table is an array of &struct ubi_vtbl_record + * objects indexed by the volume ID. + * + * If the size of the logical eraseblock is large enough to fit + * %UBI_MAX_VOLUMES records, the volume table contains %UBI_MAX_VOLUMES + * records. Otherwise, it contains as many records as it can fit (i.e., size of + * logical eraseblock divided by sizeof(struct ubi_vtbl_record)). + * + * The @upd_marker flag is used to implement volume update. It is set to %1 + * before update and set to %0 after the update. So if the update operation was + * interrupted, UBI knows that the volume is corrupted. + * + * The @alignment field is specified when the volume is created and cannot be + * later changed. It may be useful, for example, when a block-oriented file + * system works on top of UBI. The @data_pad field is calculated using the + * logical eraseblock size and @alignment. The alignment must be multiple to the + * minimal flash I/O unit. If @alignment is 1, all the available space of + * the physical eraseblocks is used. + * + * Empty records contain all zeroes and the CRC checksum of those zeroes. + */ +struct ubi_vtbl_record { + __be32 reserved_pebs; + __be32 alignment; + __be32 data_pad; + __u8 vol_type; + __u8 upd_marker; + __be16 name_len; + __u8 name[UBI_VOL_NAME_MAX+1]; + __u8 flags; + __u8 padding[23]; + __be32 crc; +} __packed; + +/* UBI fastmap on-flash data structures */ + +#define UBI_FM_SB_VOLUME_ID (UBI_LAYOUT_VOLUME_ID + 1) +#define UBI_FM_DATA_VOLUME_ID (UBI_LAYOUT_VOLUME_ID + 2) + +/* fastmap on-flash data structure format version */ +#define UBI_FM_FMT_VERSION 1 + +#define UBI_FM_SB_MAGIC 0x7B11D69F +#define UBI_FM_HDR_MAGIC 0xD4B82EF7 +#define UBI_FM_VHDR_MAGIC 0xFA370ED1 +#define UBI_FM_POOL_MAGIC 0x67AF4D08 +#define UBI_FM_EBA_MAGIC 0xf0c040a8 + +/* A fastmap super block can be located between PEB 0 and + * UBI_FM_MAX_START */ +#define UBI_FM_MAX_START 64 + +/* A fastmap can use up to UBI_FM_MAX_BLOCKS PEBs */ +#define UBI_FM_MAX_BLOCKS 32 + +/* 5% of the total number of PEBs have to be scanned while attaching + * from a fastmap. + * But the size of this pool is limited to be between UBI_FM_MIN_POOL_SIZE and + * UBI_FM_MAX_POOL_SIZE */ +#define UBI_FM_MIN_POOL_SIZE 8 +#define UBI_FM_MAX_POOL_SIZE 256 + +/** + * struct ubi_fm_sb - UBI fastmap super block + * @magic: fastmap super block magic number (%UBI_FM_SB_MAGIC) + * @version: format version of this fastmap + * @data_crc: CRC over the fastmap data + * @used_blocks: number of PEBs used by this fastmap + * @block_loc: an array containing the location of all PEBs of the fastmap + * @block_ec: the erase counter of each used PEB + * @sqnum: highest sequence number value at the time while taking the fastmap + * + */ +struct ubi_fm_sb { + __be32 magic; + __u8 version; + __u8 padding1[3]; + __be32 data_crc; + __be32 used_blocks; + __be32 block_loc[UBI_FM_MAX_BLOCKS]; + __be32 block_ec[UBI_FM_MAX_BLOCKS]; + __be64 sqnum; + __u8 padding2[32]; +} __packed; + +/** + * struct ubi_fm_hdr - header of the fastmap data set + * @magic: fastmap header magic number (%UBI_FM_HDR_MAGIC) + * @free_peb_count: number of free PEBs known by this fastmap + * @used_peb_count: number of used PEBs known by this fastmap + * @scrub_peb_count: number of to be scrubbed PEBs known by this fastmap + * @bad_peb_count: number of bad PEBs known by this fastmap + * @erase_peb_count: number of bad PEBs which have to be erased + * @vol_count: number of UBI volumes known by this fastmap + */ +struct ubi_fm_hdr { + __be32 magic; + __be32 free_peb_count; + __be32 used_peb_count; + __be32 scrub_peb_count; + __be32 bad_peb_count; + __be32 erase_peb_count; + __be32 vol_count; + __u8 padding[4]; +} __packed; + +/* struct ubi_fm_hdr is followed by two struct ubi_fm_scan_pool */ + +/** + * struct ubi_fm_scan_pool - Fastmap pool PEBs to be scanned while attaching + * @magic: pool magic numer (%UBI_FM_POOL_MAGIC) + * @size: current pool size + * @max_size: maximal pool size + * @pebs: an array containing the location of all PEBs in this pool + */ +struct ubi_fm_scan_pool { + __be32 magic; + __be16 size; + __be16 max_size; + __be32 pebs[UBI_FM_MAX_POOL_SIZE]; + __be32 padding[4]; +} __packed; + +/* ubi_fm_scan_pool is followed by nfree+nused struct ubi_fm_ec records */ + +/** + * struct ubi_fm_ec - stores the erase counter of a PEB + * @pnum: PEB number + * @ec: ec of this PEB + */ +struct ubi_fm_ec { + __be32 pnum; + __be32 ec; +} __packed; + +/** + * struct ubi_fm_volhdr - Fastmap volume header + * it identifies the start of an eba table + * @magic: Fastmap volume header magic number (%UBI_FM_VHDR_MAGIC) + * @vol_id: volume id of the fastmapped volume + * @vol_type: type of the fastmapped volume + * @data_pad: data_pad value of the fastmapped volume + * @used_ebs: number of used LEBs within this volume + * @last_eb_bytes: number of bytes used in the last LEB + */ +struct ubi_fm_volhdr { + __be32 magic; + __be32 vol_id; + __u8 vol_type; + __u8 padding1[3]; + __be32 data_pad; + __be32 used_ebs; + __be32 last_eb_bytes; + __u8 padding2[8]; +} __packed; + +/* struct ubi_fm_volhdr is followed by one struct ubi_fm_eba records */ + +/** + * struct ubi_fm_eba - denotes an association between a PEB and LEB + * @magic: EBA table magic number + * @reserved_pebs: number of table entries + * @pnum: PEB number of LEB (LEB is the index) + */ +struct ubi_fm_eba { + __be32 magic; + __be32 reserved_pebs; + __be32 pnum[]; +} __packed; +#endif /* !__UBI_MEDIA_H__ */ diff --git a/drivers/mtd/ubi/ubi.h b/drivers/mtd/ubi/ubi.h new file mode 100644 index 000000000..8a9ac60ff --- /dev/null +++ b/drivers/mtd/ubi/ubi.h @@ -0,0 +1,1266 @@ +/* SPDX-License-Identifier: GPL-2.0-or-later */ +/* + * Copyright (c) International Business Machines Corp., 2006 + * Copyright (c) Nokia Corporation, 2006, 2007 + * + * Author: Artem Bityutskiy (Битюцкий Артём) + */ + +#ifndef __UBI_UBI_H__ +#define __UBI_UBI_H__ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include "ubi-media.h" + +/* Maximum number of supported UBI devices */ +#define UBI_MAX_DEVICES 32 + +/* UBI name used for character devices, sysfs, etc */ +#define UBI_NAME_STR "ubi" + +struct ubi_device; + +/* Normal UBI messages */ +__printf(2, 3) +void ubi_msg(const struct ubi_device *ubi, const char *fmt, ...); + +/* UBI warning messages */ +__printf(2, 3) +void ubi_warn(const struct ubi_device *ubi, const char *fmt, ...); + +/* UBI error messages */ +__printf(2, 3) +void ubi_err(const struct ubi_device *ubi, const char *fmt, ...); + +/* Background thread name pattern */ +#define UBI_BGT_NAME_PATTERN "ubi_bgt%dd" + +/* + * This marker in the EBA table means that the LEB is um-mapped. + * NOTE! It has to have the same value as %UBI_ALL. + */ +#define UBI_LEB_UNMAPPED -1 + +/* + * In case of errors, UBI tries to repeat the operation several times before + * returning error. The below constant defines how many times UBI re-tries. + */ +#define UBI_IO_RETRIES 3 + +/* + * Length of the protection queue. The length is effectively equivalent to the + * number of (global) erase cycles PEBs are protected from the wear-leveling + * worker. + */ +#define UBI_PROT_QUEUE_LEN 10 + +/* The volume ID/LEB number/erase counter is unknown */ +#define UBI_UNKNOWN -1 + +/* + * The UBI debugfs directory name pattern and maximum name length (3 for "ubi" + * + 2 for the number plus 1 for the trailing zero byte. + */ +#define UBI_DFS_DIR_NAME "ubi%d" +#define UBI_DFS_DIR_LEN (3 + 2 + 1) + +/* Number of physical eraseblocks reserved for atomic LEB change operation */ +#define EBA_RESERVED_PEBS 1 + +/* + * Error codes returned by the I/O sub-system. + * + * UBI_IO_FF: the read region of flash contains only 0xFFs + * UBI_IO_FF_BITFLIPS: the same as %UBI_IO_FF, but also there was a data + * integrity error reported by the MTD driver + * (uncorrectable ECC error in case of NAND) + * UBI_IO_BAD_HDR: the EC or VID header is corrupted (bad magic or CRC) + * UBI_IO_BAD_HDR_EBADMSG: the same as %UBI_IO_BAD_HDR, but also there was a + * data integrity error reported by the MTD driver + * (uncorrectable ECC error in case of NAND) + * UBI_IO_BITFLIPS: bit-flips were detected and corrected + * + * Note, it is probably better to have bit-flip and ebadmsg as flags which can + * be or'ed with other error code. But this is a big change because there are + * may callers, so it does not worth the risk of introducing a bug + */ +enum { + UBI_IO_FF = 1, + UBI_IO_FF_BITFLIPS, + UBI_IO_BAD_HDR, + UBI_IO_BAD_HDR_EBADMSG, + UBI_IO_BITFLIPS, +}; + +/* + * Return codes of the 'ubi_eba_copy_leb()' function. + * + * MOVE_CANCEL_RACE: canceled because the volume is being deleted, the source + * PEB was put meanwhile, or there is I/O on the source PEB + * MOVE_SOURCE_RD_ERR: canceled because there was a read error from the source + * PEB + * MOVE_TARGET_RD_ERR: canceled because there was a read error from the target + * PEB + * MOVE_TARGET_WR_ERR: canceled because there was a write error to the target + * PEB + * MOVE_TARGET_BITFLIPS: canceled because a bit-flip was detected in the + * target PEB + * MOVE_RETRY: retry scrubbing the PEB + */ +enum { + MOVE_CANCEL_RACE = 1, + MOVE_SOURCE_RD_ERR, + MOVE_TARGET_RD_ERR, + MOVE_TARGET_WR_ERR, + MOVE_TARGET_BITFLIPS, + MOVE_RETRY, +}; + +/* + * Return codes of the fastmap sub-system + * + * UBI_NO_FASTMAP: No fastmap super block was found + * UBI_BAD_FASTMAP: A fastmap was found but it's unusable + */ +enum { + UBI_NO_FASTMAP = 1, + UBI_BAD_FASTMAP, +}; + +/** + * struct ubi_vid_io_buf - VID buffer used to read/write VID info to/from the + * flash. + * @hdr: a pointer to the VID header stored in buffer + * @buffer: underlying buffer + */ +struct ubi_vid_io_buf { + struct ubi_vid_hdr *hdr; + void *buffer; +}; + +/** + * struct ubi_wl_entry - wear-leveling entry. + * @u.rb: link in the corresponding (free/used) RB-tree + * @u.list: link in the protection queue + * @ec: erase counter + * @pnum: physical eraseblock number + * + * This data structure is used in the WL sub-system. Each physical eraseblock + * has a corresponding &struct wl_entry object which may be kept in different + * RB-trees. See WL sub-system for details. + */ +struct ubi_wl_entry { + union { + struct rb_node rb; + struct list_head list; + } u; + int ec; + int pnum; +}; + +/** + * struct ubi_ltree_entry - an entry in the lock tree. + * @rb: links RB-tree nodes + * @vol_id: volume ID of the locked logical eraseblock + * @lnum: locked logical eraseblock number + * @users: how many tasks are using this logical eraseblock or wait for it + * @mutex: read/write mutex to implement read/write access serialization to + * the (@vol_id, @lnum) logical eraseblock + * + * This data structure is used in the EBA sub-system to implement per-LEB + * locking. When a logical eraseblock is being locked - corresponding + * &struct ubi_ltree_entry object is inserted to the lock tree (@ubi->ltree). + * See EBA sub-system for details. + */ +struct ubi_ltree_entry { + struct rb_node rb; + int vol_id; + int lnum; + int users; + struct rw_semaphore mutex; +}; + +/** + * struct ubi_rename_entry - volume re-name description data structure. + * @new_name_len: new volume name length + * @new_name: new volume name + * @remove: if not zero, this volume should be removed, not re-named + * @desc: descriptor of the volume + * @list: links re-name entries into a list + * + * This data structure is utilized in the multiple volume re-name code. Namely, + * UBI first creates a list of &struct ubi_rename_entry objects from the + * &struct ubi_rnvol_req request object, and then utilizes this list to do all + * the job. + */ +struct ubi_rename_entry { + int new_name_len; + char new_name[UBI_VOL_NAME_MAX + 1]; + int remove; + struct ubi_volume_desc *desc; + struct list_head list; +}; + +struct ubi_volume_desc; + +/** + * struct ubi_fastmap_layout - in-memory fastmap data structure. + * @e: PEBs used by the current fastmap + * @to_be_tortured: if non-zero tortured this PEB + * @used_blocks: number of used PEBs + * @max_pool_size: maximal size of the user pool + * @max_wl_pool_size: maximal size of the pool used by the WL sub-system + */ +struct ubi_fastmap_layout { + struct ubi_wl_entry *e[UBI_FM_MAX_BLOCKS]; + int to_be_tortured[UBI_FM_MAX_BLOCKS]; + int used_blocks; + int max_pool_size; + int max_wl_pool_size; +}; + +/** + * struct ubi_fm_pool - in-memory fastmap pool + * @pebs: PEBs in this pool + * @used: number of used PEBs + * @size: total number of PEBs in this pool + * @max_size: maximal size of the pool + * + * A pool gets filled with up to max_size. + * If all PEBs within the pool are used a new fastmap will be written + * to the flash and the pool gets refilled with empty PEBs. + * + */ +struct ubi_fm_pool { + int pebs[UBI_FM_MAX_POOL_SIZE]; + int used; + int size; + int max_size; +}; + +/** + * struct ubi_eba_leb_desc - EBA logical eraseblock descriptor + * @lnum: the logical eraseblock number + * @pnum: the physical eraseblock where the LEB can be found + * + * This structure is here to hide EBA's internal from other part of the + * UBI implementation. + * + * One can query the position of a LEB by calling ubi_eba_get_ldesc(). + */ +struct ubi_eba_leb_desc { + int lnum; + int pnum; +}; + +/** + * struct ubi_volume - UBI volume description data structure. + * @dev: device object to make use of the Linux device model + * @cdev: character device object to create character device + * @ubi: reference to the UBI device description object + * @vol_id: volume ID + * @ref_count: volume reference count + * @readers: number of users holding this volume in read-only mode + * @writers: number of users holding this volume in read-write mode + * @exclusive: whether somebody holds this volume in exclusive mode + * @metaonly: whether somebody is altering only meta data of this volume + * + * @reserved_pebs: how many physical eraseblocks are reserved for this volume + * @vol_type: volume type (%UBI_DYNAMIC_VOLUME or %UBI_STATIC_VOLUME) + * @usable_leb_size: logical eraseblock size without padding + * @used_ebs: how many logical eraseblocks in this volume contain data + * @last_eb_bytes: how many bytes are stored in the last logical eraseblock + * @used_bytes: how many bytes of data this volume contains + * @alignment: volume alignment + * @data_pad: how many bytes are not used at the end of physical eraseblocks to + * satisfy the requested alignment + * @name_len: volume name length + * @name: volume name + * + * @upd_ebs: how many eraseblocks are expected to be updated + * @ch_lnum: LEB number which is being changing by the atomic LEB change + * operation + * @upd_bytes: how many bytes are expected to be received for volume update or + * atomic LEB change + * @upd_received: how many bytes were already received for volume update or + * atomic LEB change + * @upd_buf: update buffer which is used to collect update data or data for + * atomic LEB change + * + * @eba_tbl: EBA table of this volume (LEB->PEB mapping) + * @skip_check: %1 if CRC check of this static volume should be skipped. + * Directly reflects the presence of the + * %UBI_VTBL_SKIP_CRC_CHECK_FLG flag in the vtbl entry + * @checked: %1 if this static volume was checked + * @corrupted: %1 if the volume is corrupted (static volumes only) + * @upd_marker: %1 if the update marker is set for this volume + * @updating: %1 if the volume is being updated + * @changing_leb: %1 if the atomic LEB change ioctl command is in progress + * @direct_writes: %1 if direct writes are enabled for this volume + * + * @checkmap: bitmap to remember which PEB->LEB mappings got checked, + * protected by UBI LEB lock tree. + * + * The @corrupted field indicates that the volume's contents is corrupted. + * Since UBI protects only static volumes, this field is not relevant to + * dynamic volumes - it is user's responsibility to assure their data + * integrity. + * + * The @upd_marker flag indicates that this volume is either being updated at + * the moment or is damaged because of an unclean reboot. + */ +struct ubi_volume { + struct device dev; + struct cdev cdev; + struct ubi_device *ubi; + int vol_id; + int ref_count; + int readers; + int writers; + int exclusive; + int metaonly; + bool is_dead; + + int reserved_pebs; + int vol_type; + int usable_leb_size; + int used_ebs; + int last_eb_bytes; + long long used_bytes; + int alignment; + int data_pad; + int name_len; + char name[UBI_VOL_NAME_MAX + 1]; + + int upd_ebs; + int ch_lnum; + long long upd_bytes; + long long upd_received; + void *upd_buf; + + struct ubi_eba_table *eba_tbl; + unsigned int skip_check:1; + unsigned int checked:1; + unsigned int corrupted:1; + unsigned int upd_marker:1; + unsigned int updating:1; + unsigned int changing_leb:1; + unsigned int direct_writes:1; + +#ifdef CONFIG_MTD_UBI_FASTMAP + unsigned long *checkmap; +#endif +}; + +/** + * struct ubi_volume_desc - UBI volume descriptor returned when it is opened. + * @vol: reference to the corresponding volume description object + * @mode: open mode (%UBI_READONLY, %UBI_READWRITE, %UBI_EXCLUSIVE + * or %UBI_METAONLY) + */ +struct ubi_volume_desc { + struct ubi_volume *vol; + int mode; +}; + +/** + * struct ubi_debug_info - debugging information for an UBI device. + * + * @chk_gen: if UBI general extra checks are enabled + * @chk_io: if UBI I/O extra checks are enabled + * @chk_fastmap: if UBI fastmap extra checks are enabled + * @disable_bgt: disable the background task for testing purposes + * @emulate_bitflips: emulate bit-flips for testing purposes + * @emulate_io_failures: emulate write/erase failures for testing purposes + * @emulate_power_cut: emulate power cut for testing purposes + * @power_cut_counter: count down for writes left until emulated power cut + * @power_cut_min: minimum number of writes before emulating a power cut + * @power_cut_max: maximum number of writes until emulating a power cut + * @emulate_failures: emulate failures for testing purposes + * @dfs_dir_name: name of debugfs directory containing files of this UBI device + * @dfs_dir: direntry object of the UBI device debugfs directory + * @dfs_chk_gen: debugfs knob to enable UBI general extra checks + * @dfs_chk_io: debugfs knob to enable UBI I/O extra checks + * @dfs_chk_fastmap: debugfs knob to enable UBI fastmap extra checks + * @dfs_disable_bgt: debugfs knob to disable the background task + * @dfs_emulate_bitflips: debugfs knob to emulate bit-flips + * @dfs_emulate_io_failures: debugfs knob to emulate write/erase failures + * @dfs_emulate_power_cut: debugfs knob to emulate power cuts + * @dfs_power_cut_min: debugfs knob for minimum writes before power cut + * @dfs_power_cut_max: debugfs knob for maximum writes until power cut + * @dfs_emulate_failures: debugfs entry to control the fault injection type + */ +struct ubi_debug_info { + unsigned int chk_gen:1; + unsigned int chk_io:1; + unsigned int chk_fastmap:1; + unsigned int disable_bgt:1; + unsigned int emulate_bitflips:1; + unsigned int emulate_io_failures:1; + unsigned int emulate_power_cut:2; + unsigned int power_cut_counter; + unsigned int power_cut_min; + unsigned int power_cut_max; + unsigned int emulate_failures; + char dfs_dir_name[UBI_DFS_DIR_LEN]; + struct dentry *dfs_dir; + struct dentry *dfs_chk_gen; + struct dentry *dfs_chk_io; + struct dentry *dfs_chk_fastmap; + struct dentry *dfs_disable_bgt; + struct dentry *dfs_emulate_bitflips; + struct dentry *dfs_emulate_io_failures; + struct dentry *dfs_emulate_power_cut; + struct dentry *dfs_power_cut_min; + struct dentry *dfs_power_cut_max; + struct dentry *dfs_emulate_failures; +}; + +/** + * struct ubi_device - UBI device description structure + * @dev: UBI device object to use the Linux device model + * @cdev: character device object to create character device + * @ubi_num: UBI device number + * @ubi_name: UBI device name + * @vol_count: number of volumes in this UBI device + * @volumes: volumes of this UBI device + * @volumes_lock: protects @volumes, @rsvd_pebs, @avail_pebs, beb_rsvd_pebs, + * @beb_rsvd_level, @bad_peb_count, @good_peb_count, @vol_count, + * @vol->readers, @vol->writers, @vol->exclusive, + * @vol->metaonly, @vol->ref_count, @vol->mapping and + * @vol->eba_tbl. + * @ref_count: count of references on the UBI device + * @image_seq: image sequence number recorded on EC headers + * + * @rsvd_pebs: count of reserved physical eraseblocks + * @avail_pebs: count of available physical eraseblocks + * @beb_rsvd_pebs: how many physical eraseblocks are reserved for bad PEB + * handling + * @beb_rsvd_level: normal level of PEBs reserved for bad PEB handling + * + * @autoresize_vol_id: ID of the volume which has to be auto-resized at the end + * of UBI initialization + * @vtbl_slots: how many slots are available in the volume table + * @vtbl_size: size of the volume table in bytes + * @vtbl: in-RAM volume table copy + * @device_mutex: protects on-flash volume table and serializes volume + * creation, deletion, update, re-size, re-name and set + * property + * + * @max_ec: current highest erase counter value + * @mean_ec: current mean erase counter value + * + * @global_sqnum: global sequence number + * @ltree_lock: protects the lock tree and @global_sqnum + * @ltree: the lock tree + * @alc_mutex: serializes "atomic LEB change" operations + * + * @fm_disabled: non-zero if fastmap is disabled (default) + * @fm: in-memory data structure of the currently used fastmap + * @fm_pool: in-memory data structure of the fastmap pool + * @fm_wl_pool: in-memory data structure of the fastmap pool used by the WL + * sub-system + * @fm_protect: serializes ubi_update_fastmap(), protects @fm_buf and makes sure + * that critical sections cannot be interrupted by ubi_update_fastmap() + * @fm_buf: vmalloc()'d buffer which holds the raw fastmap + * @fm_size: fastmap size in bytes + * @fm_eba_sem: allows ubi_update_fastmap() to block EBA table changes + * @fm_work: fastmap work queue + * @fm_work_scheduled: non-zero if fastmap work was scheduled + * @fast_attach: non-zero if UBI was attached by fastmap + * @fm_anchor: The next anchor PEB to use for fastmap + * @fm_do_produce_anchor: If true produce an anchor PEB in wl + * @fm_pool_rsv_cnt: Number of reserved PEBs for filling pool/wl_pool + * + * @used: RB-tree of used physical eraseblocks + * @erroneous: RB-tree of erroneous used physical eraseblocks + * @free: RB-tree of free physical eraseblocks + * @free_count: Contains the number of elements in @free + * @scrub: RB-tree of physical eraseblocks which need scrubbing + * @pq: protection queue (contain physical eraseblocks which are temporarily + * protected from the wear-leveling worker) + * @pq_head: protection queue head + * @wl_lock: protects the @used, @free, @pq, @pq_head, @lookuptbl, @move_from, + * @move_to, @move_to_put @erase_pending, @wl_scheduled, @works, + * @erroneous, @erroneous_peb_count, @fm_work_scheduled, @fm_pool, + * and @fm_wl_pool fields + * @move_mutex: serializes eraseblock moves + * @work_sem: used to wait for all the scheduled works to finish and prevent + * new works from being submitted + * @wl_scheduled: non-zero if the wear-leveling was scheduled + * @lookuptbl: a table to quickly find a &struct ubi_wl_entry object for any + * physical eraseblock + * @move_from: physical eraseblock from where the data is being moved + * @move_to: physical eraseblock where the data is being moved to + * @move_to_put: if the "to" PEB was put + * @works: list of pending works + * @works_count: count of pending works + * @bgt_thread: background thread description object + * @thread_enabled: if the background thread is enabled + * @bgt_name: background thread name + * @wl_threshold: Maximum difference between two erase counters. If this + * threshold is exceeded, the WL sub-system starts moving + * data from used physical eraseblocks with low erase + * counter to free physical eraseblocks with high erase counter. + * @wl_free_max_diff: When a physical eraseblock is moved, the WL sub-system + * has to pick the target physical eraseblock to move to. + * The simplest way would be just to pick the one with the + * highest erase counter. But in certain workloads this + * could lead to an unlimited wear of one or few physical + * eraseblock. Indeed, imagine a situation when the picked + * physical eraseblock is constantly erased after the + * data is written to it. So, we have a constant which + * limits the highest erase counter of the free physical + * eraseblock to pick. Namely, the WL sub-system does not + * pick eraseblocks with erase counter greater than the + * lowest erase counter plus @wl_free_max_diff. + * + * @flash_size: underlying MTD device size (in bytes) + * @peb_count: count of physical eraseblocks on the MTD device + * @peb_size: physical eraseblock size + * @bad_peb_limit: top limit of expected bad physical eraseblocks + * @bad_peb_count: count of bad physical eraseblocks + * @good_peb_count: count of good physical eraseblocks + * @corr_peb_count: count of corrupted physical eraseblocks (preserved and not + * used by UBI) + * @erroneous_peb_count: count of erroneous physical eraseblocks in @erroneous + * @max_erroneous: maximum allowed amount of erroneous physical eraseblocks + * @min_io_size: minimal input/output unit size of the underlying MTD device + * @hdrs_min_io_size: minimal I/O unit size used for VID and EC headers + * @ro_mode: if the UBI device is in read-only mode + * @leb_size: logical eraseblock size + * @leb_start: starting offset of logical eraseblocks within physical + * eraseblocks + * @ec_hdr_alsize: size of the EC header aligned to @hdrs_min_io_size + * @vid_hdr_alsize: size of the VID header aligned to @hdrs_min_io_size + * @vid_hdr_offset: starting offset of the volume identifier header (might be + * unaligned) + * @vid_hdr_aloffset: starting offset of the VID header aligned to + * @hdrs_min_io_size + * @vid_hdr_shift: contains @vid_hdr_offset - @vid_hdr_aloffset + * @bad_allowed: whether the MTD device admits bad physical eraseblocks or not + * @nor_flash: non-zero if working on top of NOR flash + * @max_write_size: maximum amount of bytes the underlying flash can write at a + * time (MTD write buffer size) + * @mtd: MTD device descriptor + * + * @peb_buf: a buffer of PEB size used for different purposes + * @buf_mutex: protects @peb_buf + * @ckvol_mutex: serializes static volume checking when opening + * + * @dbg: debugging information for this UBI device + */ +struct ubi_device { + struct cdev cdev; + struct device dev; + int ubi_num; + char ubi_name[sizeof(UBI_NAME_STR)+5]; + int vol_count; + struct ubi_volume *volumes[UBI_MAX_VOLUMES+UBI_INT_VOL_COUNT]; + spinlock_t volumes_lock; + int ref_count; + int image_seq; + bool is_dead; + + int rsvd_pebs; + int avail_pebs; + int beb_rsvd_pebs; + int beb_rsvd_level; + int bad_peb_limit; + + int autoresize_vol_id; + int vtbl_slots; + int vtbl_size; + struct ubi_vtbl_record *vtbl; + struct mutex device_mutex; + + int max_ec; + /* Note, mean_ec is not updated run-time - should be fixed */ + int mean_ec; + + /* EBA sub-system's stuff */ + unsigned long long global_sqnum; + spinlock_t ltree_lock; + struct rb_root ltree; + struct mutex alc_mutex; + + /* Fastmap stuff */ + int fm_disabled; + struct ubi_fastmap_layout *fm; + struct ubi_fm_pool fm_pool; + struct ubi_fm_pool fm_wl_pool; + struct rw_semaphore fm_eba_sem; + struct rw_semaphore fm_protect; + void *fm_buf; + size_t fm_size; + struct work_struct fm_work; + int fm_work_scheduled; + int fast_attach; + struct ubi_wl_entry *fm_anchor; + int fm_do_produce_anchor; + int fm_pool_rsv_cnt; + + /* Wear-leveling sub-system's stuff */ + struct rb_root used; + struct rb_root erroneous; + struct rb_root free; + int free_count; + struct rb_root scrub; + struct list_head pq[UBI_PROT_QUEUE_LEN]; + int pq_head; + spinlock_t wl_lock; + struct mutex move_mutex; + struct rw_semaphore work_sem; + int wl_scheduled; + struct ubi_wl_entry **lookuptbl; + struct ubi_wl_entry *move_from; + struct ubi_wl_entry *move_to; + int move_to_put; + struct list_head works; + int works_count; + struct task_struct *bgt_thread; + int thread_enabled; + char bgt_name[sizeof(UBI_BGT_NAME_PATTERN)+2]; + int wl_threshold; + int wl_free_max_diff; + + /* I/O sub-system's stuff */ + long long flash_size; + int peb_count; + int peb_size; + int bad_peb_count; + int good_peb_count; + int corr_peb_count; + int erroneous_peb_count; + int max_erroneous; + int min_io_size; + int hdrs_min_io_size; + int ro_mode; + int leb_size; + int leb_start; + int ec_hdr_alsize; + int vid_hdr_alsize; + int vid_hdr_offset; + int vid_hdr_aloffset; + int vid_hdr_shift; + unsigned int bad_allowed:1; + unsigned int nor_flash:1; + int max_write_size; + struct mtd_info *mtd; + + void *peb_buf; + struct mutex buf_mutex; + struct mutex ckvol_mutex; + + struct ubi_debug_info dbg; +}; + +/** + * struct ubi_ainf_peb - attach information about a physical eraseblock. + * @ec: erase counter (%UBI_UNKNOWN if it is unknown) + * @pnum: physical eraseblock number + * @vol_id: ID of the volume this LEB belongs to + * @lnum: logical eraseblock number + * @scrub: if this physical eraseblock needs scrubbing + * @copy_flag: this LEB is a copy (@copy_flag is set in VID header of this LEB) + * @sqnum: sequence number + * @u: unions RB-tree or @list links + * @u.rb: link in the per-volume RB-tree of &struct ubi_ainf_peb objects + * @u.list: link in one of the eraseblock lists + * + * One object of this type is allocated for each physical eraseblock when + * attaching an MTD device. Note, if this PEB does not belong to any LEB / + * volume, the @vol_id and @lnum fields are initialized to %UBI_UNKNOWN. + */ +struct ubi_ainf_peb { + int ec; + int pnum; + int vol_id; + int lnum; + unsigned int scrub:1; + unsigned int copy_flag:1; + unsigned long long sqnum; + union { + struct rb_node rb; + struct list_head list; + } u; +}; + +/** + * struct ubi_ainf_volume - attaching information about a volume. + * @vol_id: volume ID + * @highest_lnum: highest logical eraseblock number in this volume + * @leb_count: number of logical eraseblocks in this volume + * @vol_type: volume type + * @used_ebs: number of used logical eraseblocks in this volume (only for + * static volumes) + * @last_data_size: amount of data in the last logical eraseblock of this + * volume (always equivalent to the usable logical eraseblock + * size in case of dynamic volumes) + * @data_pad: how many bytes at the end of logical eraseblocks of this volume + * are not used (due to volume alignment) + * @compat: compatibility flags of this volume + * @rb: link in the volume RB-tree + * @root: root of the RB-tree containing all the eraseblock belonging to this + * volume (&struct ubi_ainf_peb objects) + * + * One object of this type is allocated for each volume when attaching an MTD + * device. + */ +struct ubi_ainf_volume { + int vol_id; + int highest_lnum; + int leb_count; + int vol_type; + int used_ebs; + int last_data_size; + int data_pad; + int compat; + struct rb_node rb; + struct rb_root root; +}; + +/** + * struct ubi_attach_info - MTD device attaching information. + * @volumes: root of the volume RB-tree + * @corr: list of corrupted physical eraseblocks + * @free: list of free physical eraseblocks + * @erase: list of physical eraseblocks which have to be erased + * @alien: list of physical eraseblocks which should not be used by UBI (e.g., + * those belonging to "preserve"-compatible internal volumes) + * @fastmap: list of physical eraseblocks which relate to fastmap (e.g., + * eraseblocks of the current and not yet erased old fastmap blocks) + * @corr_peb_count: count of PEBs in the @corr list + * @empty_peb_count: count of PEBs which are presumably empty (contain only + * 0xFF bytes) + * @alien_peb_count: count of PEBs in the @alien list + * @bad_peb_count: count of bad physical eraseblocks + * @maybe_bad_peb_count: count of bad physical eraseblocks which are not marked + * as bad yet, but which look like bad + * @vols_found: number of volumes found + * @highest_vol_id: highest volume ID + * @is_empty: flag indicating whether the MTD device is empty or not + * @force_full_scan: flag indicating whether we need to do a full scan and drop + all existing Fastmap data structures + * @min_ec: lowest erase counter value + * @max_ec: highest erase counter value + * @max_sqnum: highest sequence number value + * @mean_ec: mean erase counter value + * @ec_sum: a temporary variable used when calculating @mean_ec + * @ec_count: a temporary variable used when calculating @mean_ec + * @aeb_slab_cache: slab cache for &struct ubi_ainf_peb objects + * @ech: temporary EC header. Only available during scan + * @vidh: temporary VID buffer. Only available during scan + * + * This data structure contains the result of attaching an MTD device and may + * be used by other UBI sub-systems to build final UBI data structures, further + * error-recovery and so on. + */ +struct ubi_attach_info { + struct rb_root volumes; + struct list_head corr; + struct list_head free; + struct list_head erase; + struct list_head alien; + struct list_head fastmap; + int corr_peb_count; + int empty_peb_count; + int alien_peb_count; + int bad_peb_count; + int maybe_bad_peb_count; + int vols_found; + int highest_vol_id; + int is_empty; + int force_full_scan; + int min_ec; + int max_ec; + unsigned long long max_sqnum; + int mean_ec; + uint64_t ec_sum; + int ec_count; + struct kmem_cache *aeb_slab_cache; + struct ubi_ec_hdr *ech; + struct ubi_vid_io_buf *vidb; +}; + +/** + * struct ubi_work - UBI work description data structure. + * @list: a link in the list of pending works + * @func: worker function + * @e: physical eraseblock to erase + * @vol_id: the volume ID on which this erasure is being performed + * @lnum: the logical eraseblock number + * @torture: if the physical eraseblock has to be tortured + * + * The @func pointer points to the worker function. If the @shutdown argument is + * not zero, the worker has to free the resources and exit immediately as the + * WL sub-system is shutting down. + * The worker has to return zero in case of success and a negative error code in + * case of failure. + */ +struct ubi_work { + struct list_head list; + int (*func)(struct ubi_device *ubi, struct ubi_work *wrk, int shutdown); + /* The below fields are only relevant to erasure works */ + struct ubi_wl_entry *e; + int vol_id; + int lnum; + int torture; +}; + +#include "debug.h" + +extern struct kmem_cache *ubi_wl_entry_slab; +extern const struct file_operations ubi_ctrl_cdev_operations; +extern const struct file_operations ubi_cdev_operations; +extern const struct file_operations ubi_vol_cdev_operations; +extern const struct class ubi_class; +extern struct mutex ubi_devices_mutex; +extern struct blocking_notifier_head ubi_notifiers; + +/* attach.c */ +struct ubi_ainf_peb *ubi_alloc_aeb(struct ubi_attach_info *ai, int pnum, + int ec); +void ubi_free_aeb(struct ubi_attach_info *ai, struct ubi_ainf_peb *aeb); +int ubi_add_to_av(struct ubi_device *ubi, struct ubi_attach_info *ai, int pnum, + int ec, const struct ubi_vid_hdr *vid_hdr, int bitflips); +struct ubi_ainf_volume *ubi_add_av(struct ubi_attach_info *ai, int vol_id); +struct ubi_ainf_volume *ubi_find_av(const struct ubi_attach_info *ai, + int vol_id); +void ubi_remove_av(struct ubi_attach_info *ai, struct ubi_ainf_volume *av); +struct ubi_ainf_peb *ubi_early_get_peb(struct ubi_device *ubi, + struct ubi_attach_info *ai); +int ubi_attach(struct ubi_device *ubi, int force_scan); + +/* vtbl.c */ +int ubi_change_vtbl_record(struct ubi_device *ubi, int idx, + struct ubi_vtbl_record *vtbl_rec); +int ubi_vtbl_rename_volumes(struct ubi_device *ubi, + struct list_head *rename_list); +int ubi_read_volume_table(struct ubi_device *ubi, struct ubi_attach_info *ai); + +/* vmt.c */ +int ubi_create_volume(struct ubi_device *ubi, struct ubi_mkvol_req *req); +int ubi_remove_volume(struct ubi_volume_desc *desc, int no_vtbl); +int ubi_resize_volume(struct ubi_volume_desc *desc, int reserved_pebs); +int ubi_rename_volumes(struct ubi_device *ubi, struct list_head *rename_list); +int ubi_add_volume(struct ubi_device *ubi, struct ubi_volume *vol); +void ubi_free_volume(struct ubi_device *ubi, struct ubi_volume *vol); + +/* upd.c */ +int ubi_start_update(struct ubi_device *ubi, struct ubi_volume *vol, + long long bytes); +int ubi_more_update_data(struct ubi_device *ubi, struct ubi_volume *vol, + const void __user *buf, int count); +int ubi_start_leb_change(struct ubi_device *ubi, struct ubi_volume *vol, + const struct ubi_leb_change_req *req); +int ubi_more_leb_change_data(struct ubi_device *ubi, struct ubi_volume *vol, + const void __user *buf, int count); + +/* misc.c */ +int ubi_calc_data_len(const struct ubi_device *ubi, const void *buf, + int length); +int ubi_check_volume(struct ubi_device *ubi, int vol_id); +void ubi_update_reserved(struct ubi_device *ubi); +void ubi_calculate_reserved(struct ubi_device *ubi); +int ubi_check_pattern(const void *buf, uint8_t patt, int size); + +static inline bool ubi_leb_valid(struct ubi_volume *vol, int lnum) +{ + return lnum >= 0 && lnum < vol->reserved_pebs; +} + +/* eba.c */ +struct ubi_eba_table *ubi_eba_create_table(struct ubi_volume *vol, + int nentries); +void ubi_eba_destroy_table(struct ubi_eba_table *tbl); +void ubi_eba_copy_table(struct ubi_volume *vol, struct ubi_eba_table *dst, + int nentries); +void ubi_eba_replace_table(struct ubi_volume *vol, struct ubi_eba_table *tbl); +void ubi_eba_get_ldesc(struct ubi_volume *vol, int lnum, + struct ubi_eba_leb_desc *ldesc); +bool ubi_eba_is_mapped(struct ubi_volume *vol, int lnum); +int ubi_eba_unmap_leb(struct ubi_device *ubi, struct ubi_volume *vol, + int lnum); +int ubi_eba_read_leb(struct ubi_device *ubi, struct ubi_volume *vol, int lnum, + void *buf, int offset, int len, int check); +int ubi_eba_read_leb_sg(struct ubi_device *ubi, struct ubi_volume *vol, + struct ubi_sgl *sgl, int lnum, int offset, int len, + int check); +int ubi_eba_write_leb(struct ubi_device *ubi, struct ubi_volume *vol, int lnum, + const void *buf, int offset, int len); +int ubi_eba_write_leb_st(struct ubi_device *ubi, struct ubi_volume *vol, + int lnum, const void *buf, int len, int used_ebs); +int ubi_eba_atomic_leb_change(struct ubi_device *ubi, struct ubi_volume *vol, + int lnum, const void *buf, int len); +int ubi_eba_copy_leb(struct ubi_device *ubi, int from, int to, + struct ubi_vid_io_buf *vidb); +int ubi_eba_init(struct ubi_device *ubi, struct ubi_attach_info *ai); +unsigned long long ubi_next_sqnum(struct ubi_device *ubi); +int self_check_eba(struct ubi_device *ubi, struct ubi_attach_info *ai_fastmap, + struct ubi_attach_info *ai_scan); + +/* wl.c */ +int ubi_sync_erase(struct ubi_device *ubi, struct ubi_wl_entry *e, int *torture); +int ubi_wl_get_peb(struct ubi_device *ubi); +int ubi_wl_put_peb(struct ubi_device *ubi, int vol_id, int lnum, + int pnum, int torture); +int ubi_wl_flush(struct ubi_device *ubi, int vol_id, int lnum); +int ubi_wl_scrub_peb(struct ubi_device *ubi, int pnum); +int ubi_wl_init(struct ubi_device *ubi, struct ubi_attach_info *ai); +void ubi_wl_close(struct ubi_device *ubi); +int ubi_thread(void *u); +struct ubi_wl_entry *ubi_wl_get_fm_peb(struct ubi_device *ubi, int anchor); +int ubi_wl_put_fm_peb(struct ubi_device *ubi, struct ubi_wl_entry *used_e, + int lnum, int torture); +int ubi_is_erase_work(struct ubi_work *wrk); +void ubi_refill_pools_and_lock(struct ubi_device *ubi); +int ubi_ensure_anchor_pebs(struct ubi_device *ubi); +int ubi_bitflip_check(struct ubi_device *ubi, int pnum, int force_scrub); + +/* io.c */ +int ubi_io_read(const struct ubi_device *ubi, void *buf, int pnum, int offset, + int len); +int ubi_io_write(struct ubi_device *ubi, const void *buf, int pnum, int offset, + int len); +int ubi_io_sync_erase(struct ubi_device *ubi, int pnum, int *torture); +int ubi_io_is_bad(const struct ubi_device *ubi, int pnum); +int ubi_io_mark_bad(const struct ubi_device *ubi, int pnum); +int ubi_io_read_ec_hdr(struct ubi_device *ubi, int pnum, + struct ubi_ec_hdr *ec_hdr, int verbose); +int ubi_io_write_ec_hdr(struct ubi_device *ubi, int pnum, + struct ubi_ec_hdr *ec_hdr); +int ubi_io_read_vid_hdr(struct ubi_device *ubi, int pnum, + struct ubi_vid_io_buf *vidb, int verbose); +int ubi_io_write_vid_hdr(struct ubi_device *ubi, int pnum, + struct ubi_vid_io_buf *vidb); + +/* build.c */ +int ubi_attach_mtd_dev(struct mtd_info *mtd, int ubi_num, + int vid_hdr_offset, int max_beb_per1024, + bool disable_fm, bool need_resv_pool, + int wl_threshold); +int ubi_detach_mtd_dev(int ubi_num, int anyway); +struct ubi_device *ubi_get_device(int ubi_num); +void ubi_put_device(struct ubi_device *ubi); +struct ubi_device *ubi_get_by_major(int major); +int ubi_major2num(int major); +int ubi_volume_notify(struct ubi_device *ubi, struct ubi_volume *vol, + int ntype); +int ubi_notify_all(struct ubi_device *ubi, int ntype, + struct notifier_block *nb); +int ubi_enumerate_volumes(struct notifier_block *nb); +void ubi_free_all_volumes(struct ubi_device *ubi); +void ubi_free_internal_volumes(struct ubi_device *ubi); + +/* kapi.c */ +void ubi_do_get_device_info(struct ubi_device *ubi, struct ubi_device_info *di); +void ubi_do_get_volume_info(struct ubi_device *ubi, struct ubi_volume *vol, + struct ubi_volume_info *vi); +int ubi_get_num_by_path(const char *pathname, int *ubi_num, int *vol_id); +/* scan.c */ +int ubi_compare_lebs(struct ubi_device *ubi, const struct ubi_ainf_peb *aeb, + int pnum, const struct ubi_vid_hdr *vid_hdr); + +/* fastmap.c */ +#ifdef CONFIG_MTD_UBI_FASTMAP +size_t ubi_calc_fm_size(struct ubi_device *ubi); +int ubi_update_fastmap(struct ubi_device *ubi); +int ubi_scan_fastmap(struct ubi_device *ubi, struct ubi_attach_info *ai, + struct ubi_attach_info *scan_ai); +int ubi_fastmap_init_checkmap(struct ubi_volume *vol, int leb_count); +void ubi_fastmap_destroy_checkmap(struct ubi_volume *vol); +static inline void ubi_free_fastmap(struct ubi_device *ubi) +{ + if (ubi->fm) { + int i; + + for (i = 0; i < ubi->fm->used_blocks; i++) + kmem_cache_free(ubi_wl_entry_slab, ubi->fm->e[i]); + kfree(ubi->fm); + ubi->fm = NULL; + } +} +#else +static inline int ubi_update_fastmap(struct ubi_device *ubi) { return 0; } +static inline int ubi_fastmap_init_checkmap(struct ubi_volume *vol, int leb_count) { return 0; } +static inline void ubi_fastmap_destroy_checkmap(struct ubi_volume *vol) {} +static inline void ubi_free_fastmap(struct ubi_device *ubi) { } +#endif + +/* block.c */ +#ifdef CONFIG_MTD_UBI_BLOCK +int ubiblock_init(void); +void ubiblock_exit(void); +int ubiblock_create(struct ubi_volume_info *vi); +int ubiblock_remove(struct ubi_volume_info *vi); +#else +static inline int ubiblock_init(void) { return 0; } +static inline void ubiblock_exit(void) {} +static inline int ubiblock_create(struct ubi_volume_info *vi) +{ + return -ENOSYS; +} +static inline int ubiblock_remove(struct ubi_volume_info *vi) +{ + return -ENOSYS; +} +#endif + +/* + * ubi_for_each_free_peb - walk the UBI free RB tree. + * @ubi: UBI device description object + * @e: a pointer to a ubi_wl_entry to use as cursor + * @pos: a pointer to RB-tree entry type to use as a loop counter + */ +#define ubi_for_each_free_peb(ubi, e, tmp_rb) \ + ubi_rb_for_each_entry((tmp_rb), (e), &(ubi)->free, u.rb) + +/* + * ubi_for_each_used_peb - walk the UBI used RB tree. + * @ubi: UBI device description object + * @e: a pointer to a ubi_wl_entry to use as cursor + * @pos: a pointer to RB-tree entry type to use as a loop counter + */ +#define ubi_for_each_used_peb(ubi, e, tmp_rb) \ + ubi_rb_for_each_entry((tmp_rb), (e), &(ubi)->used, u.rb) + +/* + * ubi_for_each_scub_peb - walk the UBI scub RB tree. + * @ubi: UBI device description object + * @e: a pointer to a ubi_wl_entry to use as cursor + * @pos: a pointer to RB-tree entry type to use as a loop counter + */ +#define ubi_for_each_scrub_peb(ubi, e, tmp_rb) \ + ubi_rb_for_each_entry((tmp_rb), (e), &(ubi)->scrub, u.rb) + +/* + * ubi_for_each_protected_peb - walk the UBI protection queue. + * @ubi: UBI device description object + * @i: a integer used as counter + * @e: a pointer to a ubi_wl_entry to use as cursor + */ +#define ubi_for_each_protected_peb(ubi, i, e) \ + for ((i) = 0; (i) < UBI_PROT_QUEUE_LEN; (i)++) \ + list_for_each_entry((e), &(ubi->pq[(i)]), u.list) + +/* + * ubi_rb_for_each_entry - walk an RB-tree. + * @rb: a pointer to type 'struct rb_node' to use as a loop counter + * @pos: a pointer to RB-tree entry type to use as a loop counter + * @root: RB-tree's root + * @member: the name of the 'struct rb_node' within the RB-tree entry + */ +#define ubi_rb_for_each_entry(rb, pos, root, member) \ + for (rb = rb_first(root), \ + pos = (rb ? container_of(rb, typeof(*pos), member) : NULL); \ + rb; \ + rb = rb_next(rb), \ + pos = (rb ? container_of(rb, typeof(*pos), member) : NULL)) + +/* + * ubi_move_aeb_to_list - move a PEB from the volume tree to a list. + * + * @av: volume attaching information + * @aeb: attaching eraseblock information + * @list: the list to move to + */ +static inline void ubi_move_aeb_to_list(struct ubi_ainf_volume *av, + struct ubi_ainf_peb *aeb, + struct list_head *list) +{ + rb_erase(&aeb->u.rb, &av->root); + list_add_tail(&aeb->u.list, list); +} + +/** + * ubi_init_vid_buf - Initialize a VID buffer + * @ubi: the UBI device + * @vidb: the VID buffer to initialize + * @buf: the underlying buffer + */ +static inline void ubi_init_vid_buf(const struct ubi_device *ubi, + struct ubi_vid_io_buf *vidb, + void *buf) +{ + if (buf) + memset(buf, 0, ubi->vid_hdr_alsize); + + vidb->buffer = buf; + vidb->hdr = buf + ubi->vid_hdr_shift; +} + +/** + * ubi_init_vid_buf - Allocate a VID buffer + * @ubi: the UBI device + * @gfp_flags: GFP flags to use for the allocation + */ +static inline struct ubi_vid_io_buf * +ubi_alloc_vid_buf(const struct ubi_device *ubi, gfp_t gfp_flags) +{ + struct ubi_vid_io_buf *vidb; + void *buf; + + vidb = kzalloc_obj(*vidb, gfp_flags); + if (!vidb) + return NULL; + + buf = kmalloc(ubi->vid_hdr_alsize, gfp_flags); + if (!buf) { + kfree(vidb); + return NULL; + } + + ubi_init_vid_buf(ubi, vidb, buf); + + return vidb; +} + +/** + * ubi_free_vid_buf - Free a VID buffer + * @vidb: the VID buffer to free + */ +static inline void ubi_free_vid_buf(struct ubi_vid_io_buf *vidb) +{ + if (!vidb) + return; + + kfree(vidb->buffer); + kfree(vidb); +} + +/** + * ubi_get_vid_hdr - Get the VID header attached to a VID buffer + * @vidb: VID buffer + */ +static inline struct ubi_vid_hdr *ubi_get_vid_hdr(struct ubi_vid_io_buf *vidb) +{ + return vidb->hdr; +} + +/** + * ubi_ro_mode - switch to read-only mode. + * @ubi: UBI device description object + */ +static inline void ubi_ro_mode(struct ubi_device *ubi) +{ + if (!ubi->ro_mode) { + ubi->ro_mode = 1; + ubi_warn(ubi, "switch to read-only mode"); + dump_stack(); + } +} + +/* + * This function is equivalent to 'ubi_io_read()', but @offset is relative to + * the beginning of the logical eraseblock, not to the beginning of the + * physical eraseblock. + */ +static inline int ubi_io_read_data(const struct ubi_device *ubi, void *buf, + int pnum, int offset, int len) +{ + ubi_assert(offset >= 0); + return ubi_io_read(ubi, buf, pnum, offset + ubi->leb_start, len); +} + +/* + * This function is equivalent to 'ubi_io_write()', but @offset is relative to + * the beginning of the logical eraseblock, not to the beginning of the + * physical eraseblock. + */ +static inline int ubi_io_write_data(struct ubi_device *ubi, const void *buf, + int pnum, int offset, int len) +{ + ubi_assert(offset >= 0); + + if (ubi_dbg_power_cut(ubi, MASK_POWER_CUT_DATA)) { + ubi_warn(ubi, "XXXXX emulating a power cut when writing data XXXXX"); + ubi_ro_mode(ubi); + return -EROFS; + } + return ubi_io_write(ubi, buf, pnum, offset + ubi->leb_start, len); +} + +/** + * vol_id2idx - get table index by volume ID. + * @ubi: UBI device description object + * @vol_id: volume ID + */ +static inline int vol_id2idx(const struct ubi_device *ubi, int vol_id) +{ + if (vol_id >= UBI_INTERNAL_VOL_START) + return vol_id - UBI_INTERNAL_VOL_START + ubi->vtbl_slots; + else + return vol_id; +} + +/** + * idx2vol_id - get volume ID by table index. + * @ubi: UBI device description object + * @idx: table index + */ +static inline int idx2vol_id(const struct ubi_device *ubi, int idx) +{ + if (idx >= ubi->vtbl_slots) + return idx - ubi->vtbl_slots + UBI_INTERNAL_VOL_START; + else + return idx; +} + +/** + * ubi_is_fm_vol - check whether a volume ID is a Fastmap volume. + * @vol_id: volume ID + */ +static inline bool ubi_is_fm_vol(int vol_id) +{ + switch (vol_id) { + case UBI_FM_SB_VOLUME_ID: + case UBI_FM_DATA_VOLUME_ID: + return true; + } + + return false; +} + +/** + * ubi_find_fm_block - check whether a PEB is part of the current Fastmap. + * @ubi: UBI device description object + * @pnum: physical eraseblock to look for + * + * This function returns a wear leveling object if @pnum relates to the current + * fastmap, @NULL otherwise. + */ +static inline struct ubi_wl_entry *ubi_find_fm_block(const struct ubi_device *ubi, + int pnum) +{ + int i; + + if (ubi->fm) { + for (i = 0; i < ubi->fm->used_blocks; i++) { + if (ubi->fm->e[i]->pnum == pnum) + return ubi->fm->e[i]; + } + } + + return NULL; +} + +#endif /* !__UBI_UBI_H__ */ diff --git a/drivers/mtd/ubi/upd.c b/drivers/mtd/ubi/upd.c new file mode 100644 index 000000000..962f693cf --- /dev/null +++ b/drivers/mtd/ubi/upd.c @@ -0,0 +1,420 @@ +// SPDX-License-Identifier: GPL-2.0-or-later +/* + * Copyright (c) International Business Machines Corp., 2006 + * Copyright (c) Nokia Corporation, 2006 + * + * Author: Artem Bityutskiy (Битюцкий Артём) + * + * Jan 2007: Alexander Schmidt, hacked per-volume update. + */ + +/* + * This file contains implementation of the volume update and atomic LEB change + * functionality. + * + * The update operation is based on the per-volume update marker which is + * stored in the volume table. The update marker is set before the update + * starts, and removed after the update has been finished. So if the update was + * interrupted by an unclean re-boot or due to some other reasons, the update + * marker stays on the flash media and UBI finds it when it attaches the MTD + * device next time. If the update marker is set for a volume, the volume is + * treated as damaged and most I/O operations are prohibited. Only a new update + * operation is allowed. + * + * Note, in general it is possible to implement the update operation as a + * transaction with a roll-back capability. + */ + +#include +#include +#include +#include "ubi.h" + +/** + * set_update_marker - set update marker. + * @ubi: UBI device description object + * @vol: volume description object + * + * This function sets the update marker flag for volume @vol. Returns zero + * in case of success and a negative error code in case of failure. + */ +static int set_update_marker(struct ubi_device *ubi, struct ubi_volume *vol) +{ + int err; + struct ubi_vtbl_record vtbl_rec; + + dbg_gen("set update marker for volume %d", vol->vol_id); + + if (vol->upd_marker) { + ubi_assert(ubi->vtbl[vol->vol_id].upd_marker); + dbg_gen("already set"); + return 0; + } + + vtbl_rec = ubi->vtbl[vol->vol_id]; + vtbl_rec.upd_marker = 1; + + mutex_lock(&ubi->device_mutex); + err = ubi_change_vtbl_record(ubi, vol->vol_id, &vtbl_rec); + vol->upd_marker = 1; + mutex_unlock(&ubi->device_mutex); + return err; +} + +/** + * clear_update_marker - clear update marker. + * @ubi: UBI device description object + * @vol: volume description object + * @bytes: new data size in bytes + * + * This function clears the update marker for volume @vol, sets new volume + * data size and clears the "corrupted" flag (static volumes only). Returns + * zero in case of success and a negative error code in case of failure. + */ +static int clear_update_marker(struct ubi_device *ubi, struct ubi_volume *vol, + long long bytes) +{ + int err; + struct ubi_vtbl_record vtbl_rec; + + dbg_gen("clear update marker for volume %d", vol->vol_id); + + vtbl_rec = ubi->vtbl[vol->vol_id]; + ubi_assert(vol->upd_marker && vtbl_rec.upd_marker); + vtbl_rec.upd_marker = 0; + + if (vol->vol_type == UBI_STATIC_VOLUME) { + vol->corrupted = 0; + vol->used_bytes = bytes; + vol->used_ebs = div_u64_rem(bytes, vol->usable_leb_size, + &vol->last_eb_bytes); + if (vol->last_eb_bytes) + vol->used_ebs += 1; + else + vol->last_eb_bytes = vol->usable_leb_size; + } + + mutex_lock(&ubi->device_mutex); + err = ubi_change_vtbl_record(ubi, vol->vol_id, &vtbl_rec); + vol->upd_marker = 0; + mutex_unlock(&ubi->device_mutex); + return err; +} + +/** + * ubi_start_update - start volume update. + * @ubi: UBI device description object + * @vol: volume description object + * @bytes: update bytes + * + * This function starts volume update operation. If @bytes is zero, the volume + * is just wiped out. Returns zero in case of success and a negative error code + * in case of failure. + */ +int ubi_start_update(struct ubi_device *ubi, struct ubi_volume *vol, + long long bytes) +{ + int i, err; + + dbg_gen("start update of volume %d, %llu bytes", vol->vol_id, bytes); + ubi_assert(!vol->updating && !vol->changing_leb); + vol->updating = 1; + + vol->upd_buf = vmalloc(ubi->leb_size); + if (!vol->upd_buf) + return -ENOMEM; + + err = set_update_marker(ubi, vol); + if (err) + return err; + + /* Before updating - wipe out the volume */ + for (i = 0; i < vol->reserved_pebs; i++) { + err = ubi_eba_unmap_leb(ubi, vol, i); + if (err) + return err; + } + + err = ubi_wl_flush(ubi, UBI_ALL, UBI_ALL); + if (err) + return err; + + if (bytes == 0) { + err = clear_update_marker(ubi, vol, 0); + if (err) + return err; + + vfree(vol->upd_buf); + vol->updating = 0; + return 0; + } + + vol->upd_ebs = div_u64(bytes + vol->usable_leb_size - 1, + vol->usable_leb_size); + vol->upd_bytes = bytes; + vol->upd_received = 0; + return 0; +} + +/** + * ubi_start_leb_change - start atomic LEB change. + * @ubi: UBI device description object + * @vol: volume description object + * @req: operation request + * + * This function starts atomic LEB change operation. Returns zero in case of + * success and a negative error code in case of failure. + */ +int ubi_start_leb_change(struct ubi_device *ubi, struct ubi_volume *vol, + const struct ubi_leb_change_req *req) +{ + ubi_assert(!vol->updating && !vol->changing_leb); + + dbg_gen("start changing LEB %d:%d, %u bytes", + vol->vol_id, req->lnum, req->bytes); + if (req->bytes == 0) + return ubi_eba_atomic_leb_change(ubi, vol, req->lnum, NULL, 0); + + vol->upd_bytes = req->bytes; + vol->upd_received = 0; + vol->changing_leb = 1; + vol->ch_lnum = req->lnum; + + vol->upd_buf = vmalloc(ALIGN((int)req->bytes, ubi->min_io_size)); + if (!vol->upd_buf) + return -ENOMEM; + + return 0; +} + +/** + * write_leb - write update data. + * @ubi: UBI device description object + * @vol: volume description object + * @lnum: logical eraseblock number + * @buf: data to write + * @len: data size + * @used_ebs: how many logical eraseblocks will this volume contain (static + * volumes only) + * + * This function writes update data to corresponding logical eraseblock. In + * case of dynamic volume, this function checks if the data contains 0xFF bytes + * at the end. If yes, the 0xFF bytes are cut and not written. So if the whole + * buffer contains only 0xFF bytes, the LEB is left unmapped. + * + * The reason why we skip the trailing 0xFF bytes in case of dynamic volume is + * that we want to make sure that more data may be appended to the logical + * eraseblock in future. Indeed, writing 0xFF bytes may have side effects and + * this PEB won't be writable anymore. So if one writes the file-system image + * to the UBI volume where 0xFFs mean free space - UBI makes sure this free + * space is writable after the update. + * + * We do not do this for static volumes because they are read-only. But this + * also cannot be done because we have to store per-LEB CRC and the correct + * data length. + * + * This function returns zero in case of success and a negative error code in + * case of failure. + */ +static int write_leb(struct ubi_device *ubi, struct ubi_volume *vol, int lnum, + void *buf, int len, int used_ebs) +{ + int err; + + if (vol->vol_type == UBI_DYNAMIC_VOLUME) { + int l = ALIGN(len, ubi->min_io_size); + + memset(buf + len, 0xFF, l - len); + len = ubi_calc_data_len(ubi, buf, l); + if (len == 0) { + dbg_gen("all %d bytes contain 0xFF - skip", len); + return 0; + } + + err = ubi_eba_write_leb(ubi, vol, lnum, buf, 0, len); + } else { + /* + * When writing static volume, and this is the last logical + * eraseblock, the length (@len) does not have to be aligned to + * the minimal flash I/O unit. The 'ubi_eba_write_leb_st()' + * function accepts exact (unaligned) length and stores it in + * the VID header. And it takes care of proper alignment by + * padding the buffer. Here we just make sure the padding will + * contain zeros, not random trash. + */ + memset(buf + len, 0, vol->usable_leb_size - len); + err = ubi_eba_write_leb_st(ubi, vol, lnum, buf, len, used_ebs); + } + + return err; +} + +/** + * ubi_more_update_data - write more update data. + * @ubi: UBI device description object + * @vol: volume description object + * @buf: write data (user-space memory buffer) + * @count: how much bytes to write + * + * This function writes more data to the volume which is being updated. It may + * be called arbitrary number of times until all the update data arriveis. This + * function returns %0 in case of success, number of bytes written during the + * last call if the whole volume update has been successfully finished, and a + * negative error code in case of failure. + */ +int ubi_more_update_data(struct ubi_device *ubi, struct ubi_volume *vol, + const void __user *buf, int count) +{ + int lnum, offs, err = 0, len, to_write = count; + + dbg_gen("write %d of %lld bytes, %lld already passed", + count, vol->upd_bytes, vol->upd_received); + + if (ubi->ro_mode) + return -EROFS; + + lnum = div_u64_rem(vol->upd_received, vol->usable_leb_size, &offs); + if (vol->upd_received + count > vol->upd_bytes) + to_write = count = vol->upd_bytes - vol->upd_received; + + /* + * When updating volumes, we accumulate whole logical eraseblock of + * data and write it at once. + */ + if (offs != 0) { + /* + * This is a write to the middle of the logical eraseblock. We + * copy the data to our update buffer and wait for more data or + * flush it if the whole eraseblock is written or the update + * is finished. + */ + + len = vol->usable_leb_size - offs; + if (len > count) + len = count; + + err = copy_from_user(vol->upd_buf + offs, buf, len); + if (err) + return -EFAULT; + + if (offs + len == vol->usable_leb_size || + vol->upd_received + len == vol->upd_bytes) { + int flush_len = offs + len; + + /* + * OK, we gathered either the whole eraseblock or this + * is the last chunk, it's time to flush the buffer. + */ + ubi_assert(flush_len <= vol->usable_leb_size); + err = write_leb(ubi, vol, lnum, vol->upd_buf, flush_len, + vol->upd_ebs); + if (err) + return err; + } + + vol->upd_received += len; + count -= len; + buf += len; + lnum += 1; + } + + /* + * If we've got more to write, let's continue. At this point we know we + * are starting from the beginning of an eraseblock. + */ + while (count) { + if (count > vol->usable_leb_size) + len = vol->usable_leb_size; + else + len = count; + + err = copy_from_user(vol->upd_buf, buf, len); + if (err) + return -EFAULT; + + if (len == vol->usable_leb_size || + vol->upd_received + len == vol->upd_bytes) { + err = write_leb(ubi, vol, lnum, vol->upd_buf, + len, vol->upd_ebs); + if (err) + break; + } + + vol->upd_received += len; + count -= len; + lnum += 1; + buf += len; + } + + ubi_assert(vol->upd_received <= vol->upd_bytes); + if (vol->upd_received == vol->upd_bytes) { + err = ubi_wl_flush(ubi, UBI_ALL, UBI_ALL); + if (err) + return err; + /* The update is finished, clear the update marker */ + err = clear_update_marker(ubi, vol, vol->upd_bytes); + if (err) + return err; + vol->updating = 0; + err = to_write; + vfree(vol->upd_buf); + } + + return err; +} + +/** + * ubi_more_leb_change_data - accept more data for atomic LEB change. + * @ubi: UBI device description object + * @vol: volume description object + * @buf: write data (user-space memory buffer) + * @count: how much bytes to write + * + * This function accepts more data to the volume which is being under the + * "atomic LEB change" operation. It may be called arbitrary number of times + * until all data arrives. This function returns %0 in case of success, number + * of bytes written during the last call if the whole "atomic LEB change" + * operation has been successfully finished, and a negative error code in case + * of failure. + */ +int ubi_more_leb_change_data(struct ubi_device *ubi, struct ubi_volume *vol, + const void __user *buf, int count) +{ + int err; + + dbg_gen("write %d of %lld bytes, %lld already passed", + count, vol->upd_bytes, vol->upd_received); + + if (ubi->ro_mode) + return -EROFS; + + if (vol->upd_received + count > vol->upd_bytes) + count = vol->upd_bytes - vol->upd_received; + + err = copy_from_user(vol->upd_buf + vol->upd_received, buf, count); + if (err) + return -EFAULT; + + vol->upd_received += count; + + if (vol->upd_received == vol->upd_bytes) { + int len = ALIGN((int)vol->upd_bytes, ubi->min_io_size); + + memset(vol->upd_buf + vol->upd_bytes, 0xFF, + len - vol->upd_bytes); + len = ubi_calc_data_len(ubi, vol->upd_buf, len); + err = ubi_eba_atomic_leb_change(ubi, vol, vol->ch_lnum, + vol->upd_buf, len); + if (err) + return err; + } + + ubi_assert(vol->upd_received <= vol->upd_bytes); + if (vol->upd_received == vol->upd_bytes) { + vol->changing_leb = 0; + err = count; + vfree(vol->upd_buf); + } + + return err; +} diff --git a/drivers/mtd/ubi/vmt.c b/drivers/mtd/ubi/vmt.c new file mode 100644 index 000000000..50192b95b --- /dev/null +++ b/drivers/mtd/ubi/vmt.c @@ -0,0 +1,848 @@ +// SPDX-License-Identifier: GPL-2.0-or-later +/* + * Copyright (c) International Business Machines Corp., 2006 + * + * Author: Artem Bityutskiy (Битюцкий Артём) + */ + +/* + * This file contains implementation of volume creation, deletion, updating and + * resizing. + */ + +#include +#include +#include +#include +#include "ubi.h" + +static int self_check_volumes(struct ubi_device *ubi); + +static ssize_t vol_attribute_show(struct device *dev, + struct device_attribute *attr, char *buf); + +/* Device attributes corresponding to files in '//class/ubi/ubiX_Y' */ +static struct device_attribute attr_vol_reserved_ebs = + __ATTR(reserved_ebs, S_IRUGO, vol_attribute_show, NULL); +static struct device_attribute attr_vol_type = + __ATTR(type, S_IRUGO, vol_attribute_show, NULL); +static struct device_attribute attr_vol_name = + __ATTR(name, S_IRUGO, vol_attribute_show, NULL); +static struct device_attribute attr_vol_corrupted = + __ATTR(corrupted, S_IRUGO, vol_attribute_show, NULL); +static struct device_attribute attr_vol_alignment = + __ATTR(alignment, S_IRUGO, vol_attribute_show, NULL); +static struct device_attribute attr_vol_usable_eb_size = + __ATTR(usable_eb_size, S_IRUGO, vol_attribute_show, NULL); +static struct device_attribute attr_vol_data_bytes = + __ATTR(data_bytes, S_IRUGO, vol_attribute_show, NULL); +static struct device_attribute attr_vol_upd_marker = + __ATTR(upd_marker, S_IRUGO, vol_attribute_show, NULL); + +/* + * "Show" method for files in '//class/ubi/ubiX_Y/'. + * + * Consider a situation: + * A. process 1 opens a sysfs file related to volume Y, say + * //class/ubi/ubiX_Y/reserved_ebs; + * B. process 2 removes volume Y; + * C. process 1 starts reading the //class/ubi/ubiX_Y/reserved_ebs file; + * + * In this situation, this function will return %-ENODEV because it will find + * out that the volume was removed from the @ubi->volumes array. + */ +static ssize_t vol_attribute_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + int ret; + struct ubi_volume *vol = container_of(dev, struct ubi_volume, dev); + struct ubi_device *ubi = vol->ubi; + + spin_lock(&ubi->volumes_lock); + if (!ubi->volumes[vol->vol_id] || ubi->volumes[vol->vol_id]->is_dead) { + spin_unlock(&ubi->volumes_lock); + return -ENODEV; + } + /* Take a reference to prevent volume removal */ + vol->ref_count += 1; + spin_unlock(&ubi->volumes_lock); + + if (attr == &attr_vol_reserved_ebs) + ret = sprintf(buf, "%d\n", vol->reserved_pebs); + else if (attr == &attr_vol_type) { + const char *tp; + + if (vol->vol_type == UBI_DYNAMIC_VOLUME) + tp = "dynamic"; + else + tp = "static"; + ret = sprintf(buf, "%s\n", tp); + } else if (attr == &attr_vol_name) + ret = sprintf(buf, "%s\n", vol->name); + else if (attr == &attr_vol_corrupted) + ret = sprintf(buf, "%d\n", vol->corrupted); + else if (attr == &attr_vol_alignment) + ret = sprintf(buf, "%d\n", vol->alignment); + else if (attr == &attr_vol_usable_eb_size) + ret = sprintf(buf, "%d\n", vol->usable_leb_size); + else if (attr == &attr_vol_data_bytes) + ret = sprintf(buf, "%lld\n", vol->used_bytes); + else if (attr == &attr_vol_upd_marker) + ret = sprintf(buf, "%d\n", vol->upd_marker); + else + /* This must be a bug */ + ret = -EINVAL; + + /* We've done the operation, drop volume and UBI device references */ + spin_lock(&ubi->volumes_lock); + vol->ref_count -= 1; + ubi_assert(vol->ref_count >= 0); + spin_unlock(&ubi->volumes_lock); + return ret; +} + +static struct attribute *volume_dev_attrs[] = { + &attr_vol_reserved_ebs.attr, + &attr_vol_type.attr, + &attr_vol_name.attr, + &attr_vol_corrupted.attr, + &attr_vol_alignment.attr, + &attr_vol_usable_eb_size.attr, + &attr_vol_data_bytes.attr, + &attr_vol_upd_marker.attr, + NULL +}; +ATTRIBUTE_GROUPS(volume_dev); + +/* Release method for volume devices */ +static void vol_release(struct device *dev) +{ + struct ubi_volume *vol = container_of(dev, struct ubi_volume, dev); + + ubi_eba_replace_table(vol, NULL); + ubi_fastmap_destroy_checkmap(vol); + kfree(vol); +} + +static struct fwnode_handle *find_volume_fwnode(struct ubi_volume *vol) +{ + struct fwnode_handle *fw_vols, *fw_vol; + const char *volname; + u32 volid; + + fw_vols = device_get_named_child_node(vol->dev.parent->parent, "volumes"); + if (!fw_vols) + return NULL; + + fwnode_for_each_child_node(fw_vols, fw_vol) { + if (!fwnode_property_read_string(fw_vol, "volname", &volname) && + strncmp(volname, vol->name, vol->name_len)) + continue; + + if (!fwnode_property_read_u32(fw_vol, "volid", &volid) && + vol->vol_id != volid) + continue; + + fwnode_handle_put(fw_vols); + return fw_vol; + } + fwnode_handle_put(fw_vols); + + return NULL; +} + +/** + * ubi_create_volume - create volume. + * @ubi: UBI device description object + * @req: volume creation request + * + * This function creates volume described by @req. If @req->vol_id id + * %UBI_VOL_NUM_AUTO, this function automatically assign ID to the new volume + * and saves it in @req->vol_id. Returns zero in case of success and a negative + * error code in case of failure. Note, the caller has to have the + * @ubi->device_mutex locked. + */ +int ubi_create_volume(struct ubi_device *ubi, struct ubi_mkvol_req *req) +{ + int i, err, vol_id = req->vol_id; + struct ubi_volume *vol; + struct ubi_vtbl_record vtbl_rec; + struct ubi_eba_table *eba_tbl = NULL; + + if (ubi->ro_mode) + return -EROFS; + + vol = kzalloc_obj(struct ubi_volume); + if (!vol) + return -ENOMEM; + + device_initialize(&vol->dev); + vol->dev.release = vol_release; + vol->dev.parent = &ubi->dev; + vol->dev.class = &ubi_class; + vol->dev.groups = volume_dev_groups; + + if (req->flags & UBI_VOL_SKIP_CRC_CHECK_FLG) + vol->skip_check = 1; + + spin_lock(&ubi->volumes_lock); + if (vol_id == UBI_VOL_NUM_AUTO) { + /* Find unused volume ID */ + dbg_gen("search for vacant volume ID"); + for (i = 0; i < ubi->vtbl_slots; i++) + if (!ubi->volumes[i]) { + vol_id = i; + break; + } + + if (vol_id == UBI_VOL_NUM_AUTO) { + ubi_err(ubi, "out of volume IDs"); + err = -ENFILE; + goto out_unlock; + } + req->vol_id = vol_id; + } + + dbg_gen("create device %d, volume %d, %llu bytes, type %d, name %s", + ubi->ubi_num, vol_id, (unsigned long long)req->bytes, + (int)req->vol_type, req->name); + + /* Ensure that this volume does not exist */ + err = -EEXIST; + if (ubi->volumes[vol_id]) { + ubi_err(ubi, "volume %d already exists", vol_id); + goto out_unlock; + } + + /* Ensure that the name is unique */ + for (i = 0; i < ubi->vtbl_slots; i++) + if (ubi->volumes[i] && !ubi->volumes[i]->is_dead && + ubi->volumes[i]->name_len == req->name_len && + !strcmp(ubi->volumes[i]->name, req->name)) { + ubi_err(ubi, "volume \"%s\" exists (ID %d)", + req->name, i); + goto out_unlock; + } + + /* Calculate how many eraseblocks are requested */ + vol->usable_leb_size = ubi->leb_size - ubi->leb_size % req->alignment; + vol->reserved_pebs = div_u64(req->bytes + vol->usable_leb_size - 1, + vol->usable_leb_size); + + /* Reserve physical eraseblocks */ + if (vol->reserved_pebs > ubi->avail_pebs) { + ubi_err(ubi, "not enough PEBs, only %d available", + ubi->avail_pebs); + if (ubi->corr_peb_count) + ubi_err(ubi, "%d PEBs are corrupted and not used", + ubi->corr_peb_count); + err = -ENOSPC; + goto out_unlock; + } + ubi->avail_pebs -= vol->reserved_pebs; + ubi->rsvd_pebs += vol->reserved_pebs; + spin_unlock(&ubi->volumes_lock); + + vol->vol_id = vol_id; + vol->alignment = req->alignment; + vol->data_pad = ubi->leb_size % vol->alignment; + vol->vol_type = req->vol_type; + vol->name_len = req->name_len; + memcpy(vol->name, req->name, vol->name_len); + vol->ubi = ubi; + device_set_node(&vol->dev, find_volume_fwnode(vol)); + + /* + * Finish all pending erases because there may be some LEBs belonging + * to the same volume ID. + */ + err = ubi_wl_flush(ubi, vol_id, UBI_ALL); + if (err) + goto out_acc; + + eba_tbl = ubi_eba_create_table(vol, vol->reserved_pebs); + if (IS_ERR(eba_tbl)) { + err = PTR_ERR(eba_tbl); + goto out_acc; + } + + ubi_eba_replace_table(vol, eba_tbl); + + if (vol->vol_type == UBI_DYNAMIC_VOLUME) { + vol->used_ebs = vol->reserved_pebs; + vol->last_eb_bytes = vol->usable_leb_size; + vol->used_bytes = + (long long)vol->used_ebs * vol->usable_leb_size; + } else { + vol->used_ebs = div_u64_rem(vol->used_bytes, + vol->usable_leb_size, + &vol->last_eb_bytes); + if (vol->last_eb_bytes != 0) + vol->used_ebs += 1; + else + vol->last_eb_bytes = vol->usable_leb_size; + } + + /* Make volume "available" before it becomes accessible via sysfs */ + spin_lock(&ubi->volumes_lock); + ubi->volumes[vol_id] = vol; + ubi->vol_count += 1; + spin_unlock(&ubi->volumes_lock); + + /* Register character device for the volume */ + cdev_init(&vol->cdev, &ubi_vol_cdev_operations); + vol->cdev.owner = THIS_MODULE; + + vol->dev.devt = MKDEV(MAJOR(ubi->cdev.dev), vol_id + 1); + dev_set_name(&vol->dev, "%s_%d", ubi->ubi_name, vol->vol_id); + err = cdev_device_add(&vol->cdev, &vol->dev); + if (err) { + ubi_err(ubi, "cannot add device"); + goto out_mapping; + } + + /* Fill volume table record */ + memset(&vtbl_rec, 0, sizeof(struct ubi_vtbl_record)); + vtbl_rec.reserved_pebs = cpu_to_be32(vol->reserved_pebs); + vtbl_rec.alignment = cpu_to_be32(vol->alignment); + vtbl_rec.data_pad = cpu_to_be32(vol->data_pad); + vtbl_rec.name_len = cpu_to_be16(vol->name_len); + if (vol->vol_type == UBI_DYNAMIC_VOLUME) + vtbl_rec.vol_type = UBI_VID_DYNAMIC; + else + vtbl_rec.vol_type = UBI_VID_STATIC; + + if (vol->skip_check) + vtbl_rec.flags |= UBI_VTBL_SKIP_CRC_CHECK_FLG; + + memcpy(vtbl_rec.name, vol->name, vol->name_len); + + err = ubi_change_vtbl_record(ubi, vol_id, &vtbl_rec); + if (err) + goto out_sysfs; + + ubi_volume_notify(ubi, vol, UBI_VOLUME_ADDED); + self_check_volumes(ubi); + return err; + +out_sysfs: + /* + * We have registered our device, we should not free the volume + * description object in this function in case of an error - it is + * freed by the release function. + */ + cdev_device_del(&vol->cdev, &vol->dev); +out_mapping: + spin_lock(&ubi->volumes_lock); + ubi->volumes[vol_id] = NULL; + ubi->vol_count -= 1; + spin_unlock(&ubi->volumes_lock); +out_acc: + spin_lock(&ubi->volumes_lock); + ubi->rsvd_pebs -= vol->reserved_pebs; + ubi->avail_pebs += vol->reserved_pebs; +out_unlock: + spin_unlock(&ubi->volumes_lock); + put_device(&vol->dev); + ubi_err(ubi, "cannot create volume %d, error %d", vol_id, err); + return err; +} + +/** + * ubi_remove_volume - remove volume. + * @desc: volume descriptor + * @no_vtbl: do not change volume table if not zero + * + * This function removes volume described by @desc. The volume has to be opened + * in "exclusive" mode. Returns zero in case of success and a negative error + * code in case of failure. The caller has to have the @ubi->device_mutex + * locked. + */ +int ubi_remove_volume(struct ubi_volume_desc *desc, int no_vtbl) +{ + struct ubi_volume *vol = desc->vol; + struct ubi_device *ubi = vol->ubi; + int i, err, vol_id = vol->vol_id, reserved_pebs = vol->reserved_pebs; + + dbg_gen("remove device %d, volume %d", ubi->ubi_num, vol_id); + ubi_assert(desc->mode == UBI_EXCLUSIVE); + ubi_assert(vol == ubi->volumes[vol_id]); + + if (ubi->ro_mode) + return -EROFS; + + spin_lock(&ubi->volumes_lock); + if (vol->ref_count > 1) { + /* + * The volume is busy, probably someone is reading one of its + * sysfs files. + */ + err = -EBUSY; + goto out_unlock; + } + + /* + * Mark volume as dead at this point to prevent that anyone + * can take a reference to the volume from now on. + * This is necessary as we have to release the spinlock before + * calling ubi_volume_notify. + */ + vol->is_dead = true; + spin_unlock(&ubi->volumes_lock); + + ubi_volume_notify(ubi, vol, UBI_VOLUME_SHUTDOWN); + + spin_lock(&ubi->volumes_lock); + ubi->volumes[vol_id] = NULL; + spin_unlock(&ubi->volumes_lock); + + if (!no_vtbl) { + err = ubi_change_vtbl_record(ubi, vol_id, NULL); + if (err) + goto out_err; + } + + for (i = 0; i < vol->reserved_pebs; i++) { + err = ubi_eba_unmap_leb(ubi, vol, i); + if (err) + goto out_err; + } + + cdev_device_del(&vol->cdev, &vol->dev); + put_device(&vol->dev); + + spin_lock(&ubi->volumes_lock); + ubi->rsvd_pebs -= reserved_pebs; + ubi->avail_pebs += reserved_pebs; + ubi_update_reserved(ubi); + ubi->vol_count -= 1; + spin_unlock(&ubi->volumes_lock); + + ubi_volume_notify(ubi, vol, UBI_VOLUME_REMOVED); + if (!no_vtbl) + self_check_volumes(ubi); + + return 0; + +out_err: + ubi_err(ubi, "cannot remove volume %d, error %d", vol_id, err); + spin_lock(&ubi->volumes_lock); + ubi->volumes[vol_id] = vol; +out_unlock: + spin_unlock(&ubi->volumes_lock); + return err; +} + +/** + * ubi_resize_volume - re-size volume. + * @desc: volume descriptor + * @reserved_pebs: new size in physical eraseblocks + * + * This function re-sizes the volume and returns zero in case of success, and a + * negative error code in case of failure. The caller has to have the + * @ubi->device_mutex locked. + */ +int ubi_resize_volume(struct ubi_volume_desc *desc, int reserved_pebs) +{ + int i, err, pebs; + struct ubi_volume *vol = desc->vol; + struct ubi_device *ubi = vol->ubi; + struct ubi_vtbl_record vtbl_rec; + struct ubi_eba_table *new_eba_tbl = NULL; + struct ubi_eba_table *old_eba_tbl = NULL; + int vol_id = vol->vol_id; + + if (ubi->ro_mode) + return -EROFS; + + dbg_gen("re-size device %d, volume %d to from %d to %d PEBs", + ubi->ubi_num, vol_id, vol->reserved_pebs, reserved_pebs); + + if (vol->vol_type == UBI_STATIC_VOLUME && + reserved_pebs < vol->used_ebs) { + ubi_err(ubi, "too small size %d, %d LEBs contain data", + reserved_pebs, vol->used_ebs); + return -EINVAL; + } + + /* If the size is the same, we have nothing to do */ + if (reserved_pebs == vol->reserved_pebs) + return 0; + + new_eba_tbl = ubi_eba_create_table(vol, reserved_pebs); + if (IS_ERR(new_eba_tbl)) + return PTR_ERR(new_eba_tbl); + + spin_lock(&ubi->volumes_lock); + if (vol->ref_count > 1) { + spin_unlock(&ubi->volumes_lock); + err = -EBUSY; + goto out_free; + } + spin_unlock(&ubi->volumes_lock); + + /* Reserve physical eraseblocks */ + pebs = reserved_pebs - vol->reserved_pebs; + if (pebs > 0) { + spin_lock(&ubi->volumes_lock); + if (pebs > ubi->avail_pebs) { + ubi_err(ubi, "not enough PEBs: requested %d, available %d", + pebs, ubi->avail_pebs); + if (ubi->corr_peb_count) + ubi_err(ubi, "%d PEBs are corrupted and not used", + ubi->corr_peb_count); + spin_unlock(&ubi->volumes_lock); + err = -ENOSPC; + goto out_free; + } + + ubi->avail_pebs -= pebs; + ubi->rsvd_pebs += pebs; + ubi_eba_copy_table(vol, new_eba_tbl, vol->reserved_pebs); + old_eba_tbl = vol->eba_tbl; + vol->eba_tbl = new_eba_tbl; + vol->reserved_pebs = reserved_pebs; + spin_unlock(&ubi->volumes_lock); + } + + if (pebs < 0) { + for (i = 0; i < -pebs; i++) { + err = ubi_eba_unmap_leb(ubi, vol, reserved_pebs + i); + if (err) + goto out_free; + } + spin_lock(&ubi->volumes_lock); + ubi->rsvd_pebs += pebs; + ubi->avail_pebs -= pebs; + ubi_update_reserved(ubi); + ubi_eba_copy_table(vol, new_eba_tbl, reserved_pebs); + old_eba_tbl = vol->eba_tbl; + vol->eba_tbl = new_eba_tbl; + vol->reserved_pebs = reserved_pebs; + spin_unlock(&ubi->volumes_lock); + } + + /* + * When we shrink a volume we have to flush all pending (erase) work. + * Otherwise it can happen that upon next attach UBI finds a LEB with + * lnum > highest_lnum and refuses to attach. + */ + if (pebs < 0) { + err = ubi_wl_flush(ubi, vol_id, UBI_ALL); + if (err) + goto out_acc; + } + + /* Change volume table record */ + vtbl_rec = ubi->vtbl[vol_id]; + vtbl_rec.reserved_pebs = cpu_to_be32(reserved_pebs); + err = ubi_change_vtbl_record(ubi, vol_id, &vtbl_rec); + if (err) + goto out_acc; + + if (vol->vol_type == UBI_DYNAMIC_VOLUME) { + vol->used_ebs = reserved_pebs; + vol->last_eb_bytes = vol->usable_leb_size; + vol->used_bytes = + (long long)vol->used_ebs * vol->usable_leb_size; + } + + /* destroy old table */ + ubi_eba_destroy_table(old_eba_tbl); + ubi_volume_notify(ubi, vol, UBI_VOLUME_RESIZED); + self_check_volumes(ubi); + return err; + +out_acc: + spin_lock(&ubi->volumes_lock); + vol->reserved_pebs = reserved_pebs - pebs; + ubi->rsvd_pebs -= pebs; + ubi->avail_pebs += pebs; + if (pebs > 0) + ubi_eba_copy_table(vol, old_eba_tbl, vol->reserved_pebs); + else + ubi_eba_copy_table(vol, old_eba_tbl, reserved_pebs); + vol->eba_tbl = old_eba_tbl; + spin_unlock(&ubi->volumes_lock); +out_free: + ubi_eba_destroy_table(new_eba_tbl); + return err; +} + +/** + * ubi_rename_volumes - re-name UBI volumes. + * @ubi: UBI device description object + * @rename_list: list of &struct ubi_rename_entry objects + * + * This function re-names or removes volumes specified in the re-name list. + * Returns zero in case of success and a negative error code in case of + * failure. + */ +int ubi_rename_volumes(struct ubi_device *ubi, struct list_head *rename_list) +{ + int err; + struct ubi_rename_entry *re; + + err = ubi_vtbl_rename_volumes(ubi, rename_list); + if (err) + return err; + + list_for_each_entry(re, rename_list, list) { + if (re->remove) { + err = ubi_remove_volume(re->desc, 1); + if (err) + break; + } else { + struct ubi_volume *vol = re->desc->vol; + + spin_lock(&ubi->volumes_lock); + vol->name_len = re->new_name_len; + memcpy(vol->name, re->new_name, re->new_name_len + 1); + spin_unlock(&ubi->volumes_lock); + ubi_volume_notify(ubi, vol, UBI_VOLUME_RENAMED); + } + } + + if (!err) + self_check_volumes(ubi); + return err; +} + +/** + * ubi_add_volume - add volume. + * @ubi: UBI device description object + * @vol: volume description object + * + * This function adds an existing volume and initializes all its data + * structures. Returns zero in case of success and a negative error code in + * case of failure. + */ +int ubi_add_volume(struct ubi_device *ubi, struct ubi_volume *vol) +{ + int err, vol_id = vol->vol_id; + dev_t dev; + + dbg_gen("add volume %d", vol_id); + + /* Register character device for the volume */ + cdev_init(&vol->cdev, &ubi_vol_cdev_operations); + vol->cdev.owner = THIS_MODULE; + dev = MKDEV(MAJOR(ubi->cdev.dev), vol->vol_id + 1); + err = cdev_add(&vol->cdev, dev, 1); + if (err) { + ubi_err(ubi, "cannot add character device for volume %d, error %d", + vol_id, err); + vol_release(&vol->dev); + return err; + } + + vol->dev.release = vol_release; + vol->dev.parent = &ubi->dev; + vol->dev.devt = dev; + vol->dev.class = &ubi_class; + vol->dev.groups = volume_dev_groups; + dev_set_name(&vol->dev, "%s_%d", ubi->ubi_name, vol->vol_id); + device_set_node(&vol->dev, find_volume_fwnode(vol)); + err = device_register(&vol->dev); + if (err) { + cdev_del(&vol->cdev); + put_device(&vol->dev); + return err; + } + + self_check_volumes(ubi); + return err; +} + +/** + * ubi_free_volume - free volume. + * @ubi: UBI device description object + * @vol: volume description object + * + * This function frees all resources for volume @vol but does not remove it. + * Used only when the UBI device is detached. + */ +void ubi_free_volume(struct ubi_device *ubi, struct ubi_volume *vol) +{ + dbg_gen("free volume %d", vol->vol_id); + + ubi->volumes[vol->vol_id] = NULL; + cdev_del(&vol->cdev); + device_unregister(&vol->dev); +} + +/** + * self_check_volume - check volume information. + * @ubi: UBI device description object + * @vol_id: volume ID + * + * Returns zero if volume is all right and a negative error code if not. + */ +static int self_check_volume(struct ubi_device *ubi, int vol_id) +{ + int idx = vol_id2idx(ubi, vol_id); + int reserved_pebs, alignment, data_pad, vol_type, name_len, upd_marker; + const struct ubi_volume *vol; + long long n; + const char *name; + + spin_lock(&ubi->volumes_lock); + reserved_pebs = be32_to_cpu(ubi->vtbl[vol_id].reserved_pebs); + vol = ubi->volumes[idx]; + + if (!vol) { + if (reserved_pebs) { + ubi_err(ubi, "no volume info, but volume exists"); + goto fail; + } + spin_unlock(&ubi->volumes_lock); + return 0; + } + + if (vol->reserved_pebs < 0 || vol->alignment < 0 || vol->data_pad < 0 || + vol->name_len < 0) { + ubi_err(ubi, "negative values"); + goto fail; + } + if (vol->alignment > ubi->leb_size || vol->alignment == 0) { + ubi_err(ubi, "bad alignment"); + goto fail; + } + + n = vol->alignment & (ubi->min_io_size - 1); + if (vol->alignment != 1 && n) { + ubi_err(ubi, "alignment is not multiple of min I/O unit"); + goto fail; + } + + n = ubi->leb_size % vol->alignment; + if (vol->data_pad != n) { + ubi_err(ubi, "bad data_pad, has to be %lld", n); + goto fail; + } + + if (vol->vol_type != UBI_DYNAMIC_VOLUME && + vol->vol_type != UBI_STATIC_VOLUME) { + ubi_err(ubi, "bad vol_type"); + goto fail; + } + + if (vol->upd_marker && vol->corrupted) { + ubi_err(ubi, "update marker and corrupted simultaneously"); + goto fail; + } + + if (vol->reserved_pebs > ubi->good_peb_count) { + ubi_err(ubi, "too large reserved_pebs"); + goto fail; + } + + n = ubi->leb_size - vol->data_pad; + if (vol->usable_leb_size != ubi->leb_size - vol->data_pad) { + ubi_err(ubi, "bad usable_leb_size, has to be %lld", n); + goto fail; + } + + if (vol->name_len > UBI_VOL_NAME_MAX) { + ubi_err(ubi, "too long volume name, max is %d", + UBI_VOL_NAME_MAX); + goto fail; + } + + n = strnlen(vol->name, vol->name_len + 1); + if (n != vol->name_len) { + ubi_err(ubi, "bad name_len %lld", n); + goto fail; + } + + n = (long long)vol->used_ebs * vol->usable_leb_size; + if (vol->vol_type == UBI_DYNAMIC_VOLUME) { + if (vol->corrupted) { + ubi_err(ubi, "corrupted dynamic volume"); + goto fail; + } + if (vol->used_ebs != vol->reserved_pebs) { + ubi_err(ubi, "bad used_ebs"); + goto fail; + } + if (vol->last_eb_bytes != vol->usable_leb_size) { + ubi_err(ubi, "bad last_eb_bytes"); + goto fail; + } + if (vol->used_bytes != n) { + ubi_err(ubi, "bad used_bytes"); + goto fail; + } + + if (vol->skip_check) { + ubi_err(ubi, "bad skip_check"); + goto fail; + } + } else { + if (vol->used_ebs < 0 || vol->used_ebs > vol->reserved_pebs) { + ubi_err(ubi, "bad used_ebs"); + goto fail; + } + if (vol->last_eb_bytes < 0 || + vol->last_eb_bytes > vol->usable_leb_size) { + ubi_err(ubi, "bad last_eb_bytes"); + goto fail; + } + if (vol->used_bytes < 0 || vol->used_bytes > n || + vol->used_bytes < n - vol->usable_leb_size) { + ubi_err(ubi, "bad used_bytes"); + goto fail; + } + } + + alignment = be32_to_cpu(ubi->vtbl[vol_id].alignment); + data_pad = be32_to_cpu(ubi->vtbl[vol_id].data_pad); + name_len = be16_to_cpu(ubi->vtbl[vol_id].name_len); + upd_marker = ubi->vtbl[vol_id].upd_marker; + name = &ubi->vtbl[vol_id].name[0]; + if (ubi->vtbl[vol_id].vol_type == UBI_VID_DYNAMIC) + vol_type = UBI_DYNAMIC_VOLUME; + else + vol_type = UBI_STATIC_VOLUME; + + if (alignment != vol->alignment || data_pad != vol->data_pad || + upd_marker != vol->upd_marker || vol_type != vol->vol_type || + name_len != vol->name_len || strncmp(name, vol->name, name_len)) { + ubi_err(ubi, "volume info is different"); + goto fail; + } + + spin_unlock(&ubi->volumes_lock); + return 0; + +fail: + ubi_err(ubi, "self-check failed for volume %d", vol_id); + if (vol) + ubi_dump_vol_info(vol); + ubi_dump_vtbl_record(&ubi->vtbl[vol_id], vol_id); + dump_stack(); + spin_unlock(&ubi->volumes_lock); + return -EINVAL; +} + +/** + * self_check_volumes - check information about all volumes. + * @ubi: UBI device description object + * + * Returns zero if volumes are all right and a negative error code if not. + */ +static int self_check_volumes(struct ubi_device *ubi) +{ + int i, err = 0; + + if (!ubi_dbg_chk_gen(ubi)) + return 0; + + for (i = 0; i < ubi->vtbl_slots; i++) { + err = self_check_volume(ubi, i); + if (err) + break; + } + + return err; +} diff --git a/drivers/mtd/ubi/vtbl.c b/drivers/mtd/ubi/vtbl.c new file mode 100644 index 000000000..06d87a301 --- /dev/null +++ b/drivers/mtd/ubi/vtbl.c @@ -0,0 +1,877 @@ +// SPDX-License-Identifier: GPL-2.0-or-later +/* + * Copyright (c) International Business Machines Corp., 2006 + * Copyright (c) Nokia Corporation, 2006, 2007 + * + * Author: Artem Bityutskiy (Битюцкий Артём) + */ + +/* + * This file includes volume table manipulation code. The volume table is an + * on-flash table containing volume meta-data like name, number of reserved + * physical eraseblocks, type, etc. The volume table is stored in the so-called + * "layout volume". + * + * The layout volume is an internal volume which is organized as follows. It + * consists of two logical eraseblocks - LEB 0 and LEB 1. Each logical + * eraseblock stores one volume table copy, i.e. LEB 0 and LEB 1 duplicate each + * other. This redundancy guarantees robustness to unclean reboots. The volume + * table is basically an array of volume table records. Each record contains + * full information about the volume and protected by a CRC checksum. Note, + * nowadays we use the atomic LEB change operation when updating the volume + * table, so we do not really need 2 LEBs anymore, but we preserve the older + * design for the backward compatibility reasons. + * + * When the volume table is changed, it is first changed in RAM. Then LEB 0 is + * erased, and the updated volume table is written back to LEB 0. Then same for + * LEB 1. This scheme guarantees recoverability from unclean reboots. + * + * In this UBI implementation the on-flash volume table does not contain any + * information about how much data static volumes contain. + * + * But it would still be beneficial to store this information in the volume + * table. For example, suppose we have a static volume X, and all its physical + * eraseblocks became bad for some reasons. Suppose we are attaching the + * corresponding MTD device, for some reason we find no logical eraseblocks + * corresponding to the volume X. According to the volume table volume X does + * exist. So we don't know whether it is just empty or all its physical + * eraseblocks went bad. So we cannot alarm the user properly. + * + * The volume table also stores so-called "update marker", which is used for + * volume updates. Before updating the volume, the update marker is set, and + * after the update operation is finished, the update marker is cleared. So if + * the update operation was interrupted (e.g. by an unclean reboot) - the + * update marker is still there and we know that the volume's contents is + * damaged. + */ + +#include +#include +#include +#include +#include "ubi.h" + +static void self_vtbl_check(const struct ubi_device *ubi); + +/* Empty volume table record */ +static struct ubi_vtbl_record empty_vtbl_record; + +/** + * ubi_update_layout_vol - helper for updatting layout volumes on flash + * @ubi: UBI device description object + */ +static int ubi_update_layout_vol(struct ubi_device *ubi) +{ + struct ubi_volume *layout_vol; + int i, err; + + layout_vol = ubi->volumes[vol_id2idx(ubi, UBI_LAYOUT_VOLUME_ID)]; + for (i = 0; i < UBI_LAYOUT_VOLUME_EBS; i++) { + err = ubi_eba_atomic_leb_change(ubi, layout_vol, i, ubi->vtbl, + ubi->vtbl_size); + if (err) + return err; + } + + return 0; +} + +/** + * ubi_change_vtbl_record - change volume table record. + * @ubi: UBI device description object + * @idx: table index to change + * @vtbl_rec: new volume table record + * + * This function changes volume table record @idx. If @vtbl_rec is %NULL, empty + * volume table record is written. The caller does not have to calculate CRC of + * the record as it is done by this function. Returns zero in case of success + * and a negative error code in case of failure. + */ +int ubi_change_vtbl_record(struct ubi_device *ubi, int idx, + struct ubi_vtbl_record *vtbl_rec) +{ + int err; + uint32_t crc; + + ubi_assert(idx >= 0 && idx < ubi->vtbl_slots); + + if (!vtbl_rec) + vtbl_rec = &empty_vtbl_record; + else { + crc = crc32(UBI_CRC32_INIT, vtbl_rec, UBI_VTBL_RECORD_SIZE_CRC); + vtbl_rec->crc = cpu_to_be32(crc); + } + + memcpy(&ubi->vtbl[idx], vtbl_rec, sizeof(struct ubi_vtbl_record)); + err = ubi_update_layout_vol(ubi); + + self_vtbl_check(ubi); + return err ? err : 0; +} + +/** + * ubi_vtbl_rename_volumes - rename UBI volumes in the volume table. + * @ubi: UBI device description object + * @rename_list: list of &struct ubi_rename_entry objects + * + * This function re-names multiple volumes specified in @req in the volume + * table. Returns zero in case of success and a negative error code in case of + * failure. + */ +int ubi_vtbl_rename_volumes(struct ubi_device *ubi, + struct list_head *rename_list) +{ + struct ubi_rename_entry *re; + + list_for_each_entry(re, rename_list, list) { + uint32_t crc; + struct ubi_volume *vol = re->desc->vol; + struct ubi_vtbl_record *vtbl_rec = &ubi->vtbl[vol->vol_id]; + + if (re->remove) { + memcpy(vtbl_rec, &empty_vtbl_record, + sizeof(struct ubi_vtbl_record)); + continue; + } + + vtbl_rec->name_len = cpu_to_be16(re->new_name_len); + memcpy(vtbl_rec->name, re->new_name, re->new_name_len); + memset(vtbl_rec->name + re->new_name_len, 0, + UBI_VOL_NAME_MAX + 1 - re->new_name_len); + crc = crc32(UBI_CRC32_INIT, vtbl_rec, + UBI_VTBL_RECORD_SIZE_CRC); + vtbl_rec->crc = cpu_to_be32(crc); + } + + return ubi_update_layout_vol(ubi); +} + +/** + * vtbl_check - check if volume table is not corrupted and sensible. + * @ubi: UBI device description object + * @vtbl: volume table + * + * This function returns zero if @vtbl is all right, %1 if CRC is incorrect, + * and %-EINVAL if it contains inconsistent data. + */ +static int vtbl_check(const struct ubi_device *ubi, + const struct ubi_vtbl_record *vtbl) +{ + int i, n, reserved_pebs, alignment, data_pad, vol_type, name_len; + int upd_marker, err; + uint32_t crc; + const char *name; + + for (i = 0; i < ubi->vtbl_slots; i++) { + cond_resched(); + + reserved_pebs = be32_to_cpu(vtbl[i].reserved_pebs); + alignment = be32_to_cpu(vtbl[i].alignment); + data_pad = be32_to_cpu(vtbl[i].data_pad); + upd_marker = vtbl[i].upd_marker; + vol_type = vtbl[i].vol_type; + name_len = be16_to_cpu(vtbl[i].name_len); + name = &vtbl[i].name[0]; + + crc = crc32(UBI_CRC32_INIT, &vtbl[i], UBI_VTBL_RECORD_SIZE_CRC); + if (be32_to_cpu(vtbl[i].crc) != crc) { + ubi_err(ubi, "bad CRC at record %u: %#08x, not %#08x", + i, crc, be32_to_cpu(vtbl[i].crc)); + ubi_dump_vtbl_record(&vtbl[i], i); + return 1; + } + + if (reserved_pebs == 0) { + if (memcmp(&vtbl[i], &empty_vtbl_record, + UBI_VTBL_RECORD_SIZE)) { + err = 2; + goto bad; + } + continue; + } + + if (reserved_pebs < 0 || alignment < 0 || data_pad < 0 || + name_len < 0) { + err = 3; + goto bad; + } + + if (alignment > ubi->leb_size || alignment == 0) { + err = 4; + goto bad; + } + + n = alignment & (ubi->min_io_size - 1); + if (alignment != 1 && n) { + err = 5; + goto bad; + } + + n = ubi->leb_size % alignment; + if (data_pad != n) { + ubi_err(ubi, "bad data_pad, has to be %d", n); + err = 6; + goto bad; + } + + if (vol_type != UBI_VID_DYNAMIC && vol_type != UBI_VID_STATIC) { + err = 7; + goto bad; + } + + if (upd_marker != 0 && upd_marker != 1) { + err = 8; + goto bad; + } + + if (reserved_pebs > ubi->good_peb_count) { + ubi_err(ubi, "too large reserved_pebs %d, good PEBs %d", + reserved_pebs, ubi->good_peb_count); + err = 9; + goto bad; + } + + if (name_len > UBI_VOL_NAME_MAX) { + err = 10; + goto bad; + } + + if (name[0] == '\0') { + err = 11; + goto bad; + } + + if (name_len != strnlen(name, name_len + 1)) { + err = 12; + goto bad; + } + } + + /* Checks that all names are unique */ + for (i = 0; i < ubi->vtbl_slots - 1; i++) { + for (n = i + 1; n < ubi->vtbl_slots; n++) { + int len1 = be16_to_cpu(vtbl[i].name_len); + int len2 = be16_to_cpu(vtbl[n].name_len); + + if (len1 > 0 && len1 == len2 && + !strncmp(vtbl[i].name, vtbl[n].name, len1)) { + ubi_err(ubi, "volumes %d and %d have the same name \"%s\"", + i, n, vtbl[i].name); + ubi_dump_vtbl_record(&vtbl[i], i); + ubi_dump_vtbl_record(&vtbl[n], n); + return -EINVAL; + } + } + } + + return 0; + +bad: + ubi_err(ubi, "volume table check failed: record %d, error %d", i, err); + ubi_dump_vtbl_record(&vtbl[i], i); + return -EINVAL; +} + +/** + * create_vtbl - create a copy of volume table. + * @ubi: UBI device description object + * @ai: attaching information + * @copy: number of the volume table copy + * @vtbl: contents of the volume table + * + * This function returns zero in case of success and a negative error code in + * case of failure. + */ +static int create_vtbl(struct ubi_device *ubi, struct ubi_attach_info *ai, + int copy, void *vtbl) +{ + int err, tries = 0; + struct ubi_vid_io_buf *vidb; + struct ubi_vid_hdr *vid_hdr; + struct ubi_ainf_peb *new_aeb; + + dbg_gen("create volume table (copy #%d)", copy + 1); + + vidb = ubi_alloc_vid_buf(ubi, GFP_KERNEL); + if (!vidb) + return -ENOMEM; + + vid_hdr = ubi_get_vid_hdr(vidb); + +retry: + new_aeb = ubi_early_get_peb(ubi, ai); + if (IS_ERR(new_aeb)) { + err = PTR_ERR(new_aeb); + goto out_free; + } + + vid_hdr->vol_type = UBI_LAYOUT_VOLUME_TYPE; + vid_hdr->vol_id = cpu_to_be32(UBI_LAYOUT_VOLUME_ID); + vid_hdr->compat = UBI_LAYOUT_VOLUME_COMPAT; + vid_hdr->data_size = vid_hdr->used_ebs = + vid_hdr->data_pad = cpu_to_be32(0); + vid_hdr->lnum = cpu_to_be32(copy); + vid_hdr->sqnum = cpu_to_be64(++ai->max_sqnum); + + /* The EC header is already there, write the VID header */ + err = ubi_io_write_vid_hdr(ubi, new_aeb->pnum, vidb); + if (err) + goto write_error; + + /* Write the layout volume contents */ + err = ubi_io_write_data(ubi, vtbl, new_aeb->pnum, 0, ubi->vtbl_size); + if (err) + goto write_error; + + /* + * And add it to the attaching information. Don't delete the old version + * of this LEB as it will be deleted and freed in 'ubi_add_to_av()'. + */ + err = ubi_add_to_av(ubi, ai, new_aeb->pnum, new_aeb->ec, vid_hdr, 0); + ubi_free_aeb(ai, new_aeb); + ubi_free_vid_buf(vidb); + return err; + +write_error: + if (err == -EIO && ++tries <= 5) { + /* + * Probably this physical eraseblock went bad, try to pick + * another one. + */ + list_add(&new_aeb->u.list, &ai->erase); + goto retry; + } + ubi_free_aeb(ai, new_aeb); +out_free: + ubi_free_vid_buf(vidb); + return err; + +} + +/** + * process_lvol - process the layout volume. + * @ubi: UBI device description object + * @ai: attaching information + * @av: layout volume attaching information + * + * This function is responsible for reading the layout volume, ensuring it is + * not corrupted, and recovering from corruptions if needed. Returns volume + * table in case of success and a negative error code in case of failure. + */ +static struct ubi_vtbl_record *process_lvol(struct ubi_device *ubi, + struct ubi_attach_info *ai, + struct ubi_ainf_volume *av) +{ + int err; + struct rb_node *rb; + struct ubi_ainf_peb *aeb; + struct ubi_vtbl_record *leb[UBI_LAYOUT_VOLUME_EBS] = { NULL, NULL }; + int leb_corrupted[UBI_LAYOUT_VOLUME_EBS] = {1, 1}; + + /* + * UBI goes through the following steps when it changes the layout + * volume: + * a. erase LEB 0; + * b. write new data to LEB 0; + * c. erase LEB 1; + * d. write new data to LEB 1. + * + * Before the change, both LEBs contain the same data. + * + * Due to unclean reboots, the contents of LEB 0 may be lost, but there + * should LEB 1. So it is OK if LEB 0 is corrupted while LEB 1 is not. + * Similarly, LEB 1 may be lost, but there should be LEB 0. And + * finally, unclean reboots may result in a situation when neither LEB + * 0 nor LEB 1 are corrupted, but they are different. In this case, LEB + * 0 contains more recent information. + * + * So the plan is to first check LEB 0. Then + * a. if LEB 0 is OK, it must be containing the most recent data; then + * we compare it with LEB 1, and if they are different, we copy LEB + * 0 to LEB 1; + * b. if LEB 0 is corrupted, but LEB 1 has to be OK, and we copy LEB 1 + * to LEB 0. + */ + + dbg_gen("check layout volume"); + + /* Read both LEB 0 and LEB 1 into memory */ + ubi_rb_for_each_entry(rb, aeb, &av->root, u.rb) { + leb[aeb->lnum] = vzalloc(ubi->vtbl_size); + if (!leb[aeb->lnum]) { + err = -ENOMEM; + goto out_free; + } + + err = ubi_io_read_data(ubi, leb[aeb->lnum], aeb->pnum, 0, + ubi->vtbl_size); + if (err == UBI_IO_BITFLIPS || mtd_is_eccerr(err)) + /* + * Scrub the PEB later. Note, -EBADMSG indicates an + * uncorrectable ECC error, but we have our own CRC and + * the data will be checked later. If the data is OK, + * the PEB will be scrubbed (because we set + * aeb->scrub). If the data is not OK, the contents of + * the PEB will be recovered from the second copy, and + * aeb->scrub will be cleared in + * 'ubi_add_to_av()'. + */ + aeb->scrub = 1; + else if (err) + goto out_free; + } + + err = -EINVAL; + if (leb[0]) { + leb_corrupted[0] = vtbl_check(ubi, leb[0]); + if (leb_corrupted[0] < 0) + goto out_free; + } + + if (!leb_corrupted[0]) { + /* LEB 0 is OK */ + if (leb[1]) + leb_corrupted[1] = memcmp(leb[0], leb[1], + ubi->vtbl_size); + if (leb_corrupted[1]) { + ubi_warn(ubi, "volume table copy #2 is corrupted"); + err = create_vtbl(ubi, ai, 1, leb[0]); + if (err) + goto out_free; + ubi_msg(ubi, "volume table was restored"); + } + + /* Both LEB 1 and LEB 2 are OK and consistent */ + vfree(leb[1]); + return leb[0]; + } else { + /* LEB 0 is corrupted or does not exist */ + if (leb[1]) { + leb_corrupted[1] = vtbl_check(ubi, leb[1]); + if (leb_corrupted[1] < 0) + goto out_free; + } + if (leb_corrupted[1]) { + /* Both LEB 0 and LEB 1 are corrupted */ + ubi_err(ubi, "both volume tables are corrupted"); + goto out_free; + } + + ubi_warn(ubi, "volume table copy #1 is corrupted"); + err = create_vtbl(ubi, ai, 0, leb[1]); + if (err) + goto out_free; + ubi_msg(ubi, "volume table was restored"); + + vfree(leb[0]); + return leb[1]; + } + +out_free: + vfree(leb[0]); + vfree(leb[1]); + return ERR_PTR(err); +} + +/** + * create_empty_lvol - create empty layout volume. + * @ubi: UBI device description object + * @ai: attaching information + * + * This function returns volume table contents in case of success and a + * negative error code in case of failure. + */ +static struct ubi_vtbl_record *create_empty_lvol(struct ubi_device *ubi, + struct ubi_attach_info *ai) +{ + int i; + struct ubi_vtbl_record *vtbl; + + vtbl = vzalloc(ubi->vtbl_size); + if (!vtbl) + return ERR_PTR(-ENOMEM); + + for (i = 0; i < ubi->vtbl_slots; i++) + memcpy(&vtbl[i], &empty_vtbl_record, UBI_VTBL_RECORD_SIZE); + + for (i = 0; i < UBI_LAYOUT_VOLUME_EBS; i++) { + int err; + + err = create_vtbl(ubi, ai, i, vtbl); + if (err) { + vfree(vtbl); + return ERR_PTR(err); + } + } + + return vtbl; +} + +/** + * init_volumes - initialize volume information for existing volumes. + * @ubi: UBI device description object + * @ai: scanning information + * @vtbl: volume table + * + * This function allocates volume description objects for existing volumes. + * Returns zero in case of success and a negative error code in case of + * failure. + */ +static int init_volumes(struct ubi_device *ubi, + const struct ubi_attach_info *ai, + const struct ubi_vtbl_record *vtbl) +{ + int i, err, reserved_pebs = 0; + struct ubi_ainf_volume *av; + struct ubi_volume *vol; + + for (i = 0; i < ubi->vtbl_slots; i++) { + cond_resched(); + + if (be32_to_cpu(vtbl[i].reserved_pebs) == 0) + continue; /* Empty record */ + + vol = kzalloc_obj(struct ubi_volume); + if (!vol) + return -ENOMEM; + + vol->reserved_pebs = be32_to_cpu(vtbl[i].reserved_pebs); + vol->alignment = be32_to_cpu(vtbl[i].alignment); + vol->data_pad = be32_to_cpu(vtbl[i].data_pad); + vol->upd_marker = vtbl[i].upd_marker; + vol->vol_type = vtbl[i].vol_type == UBI_VID_DYNAMIC ? + UBI_DYNAMIC_VOLUME : UBI_STATIC_VOLUME; + vol->name_len = be16_to_cpu(vtbl[i].name_len); + vol->usable_leb_size = ubi->leb_size - vol->data_pad; + memcpy(vol->name, vtbl[i].name, vol->name_len); + vol->name[vol->name_len] = '\0'; + vol->vol_id = i; + + if (vtbl[i].flags & UBI_VTBL_SKIP_CRC_CHECK_FLG) + vol->skip_check = 1; + + if (vtbl[i].flags & UBI_VTBL_AUTORESIZE_FLG) { + /* Auto re-size flag may be set only for one volume */ + if (ubi->autoresize_vol_id != -1) { + ubi_err(ubi, "more than one auto-resize volume (%d and %d)", + ubi->autoresize_vol_id, i); + kfree(vol); + return -EINVAL; + } + + ubi->autoresize_vol_id = i; + } + + ubi_assert(!ubi->volumes[i]); + ubi->volumes[i] = vol; + ubi->vol_count += 1; + vol->ubi = ubi; + reserved_pebs += vol->reserved_pebs; + + /* + * We use ubi->peb_count and not vol->reserved_pebs because + * we want to keep the code simple. Otherwise we'd have to + * resize/check the bitmap upon volume resize too. + * Allocating a few bytes more does not hurt. + */ + err = ubi_fastmap_init_checkmap(vol, ubi->peb_count); + if (err) + return err; + + /* + * In case of dynamic volume UBI knows nothing about how many + * data is stored there. So assume the whole volume is used. + */ + if (vol->vol_type == UBI_DYNAMIC_VOLUME) { + vol->used_ebs = vol->reserved_pebs; + vol->last_eb_bytes = vol->usable_leb_size; + vol->used_bytes = + (long long)vol->used_ebs * vol->usable_leb_size; + continue; + } + + /* Static volumes only */ + av = ubi_find_av(ai, i); + if (!av || !av->leb_count) { + /* + * No eraseblocks belonging to this volume found. We + * don't actually know whether this static volume is + * completely corrupted or just contains no data. And + * we cannot know this as long as data size is not + * stored on flash. So we just assume the volume is + * empty. FIXME: this should be handled. + */ + continue; + } + + if (av->leb_count != av->used_ebs) { + /* + * We found a static volume which misses several + * eraseblocks. Treat it as corrupted. + */ + ubi_warn(ubi, "static volume %d misses %d LEBs - corrupted", + av->vol_id, av->used_ebs - av->leb_count); + vol->corrupted = 1; + continue; + } + + vol->used_ebs = av->used_ebs; + vol->used_bytes = + (long long)(vol->used_ebs - 1) * vol->usable_leb_size; + vol->used_bytes += av->last_data_size; + vol->last_eb_bytes = av->last_data_size; + } + + /* And add the layout volume */ + vol = kzalloc_obj(struct ubi_volume); + if (!vol) + return -ENOMEM; + + vol->reserved_pebs = UBI_LAYOUT_VOLUME_EBS; + vol->alignment = UBI_LAYOUT_VOLUME_ALIGN; + vol->vol_type = UBI_DYNAMIC_VOLUME; + vol->name_len = sizeof(UBI_LAYOUT_VOLUME_NAME) - 1; + memcpy(vol->name, UBI_LAYOUT_VOLUME_NAME, vol->name_len + 1); + vol->usable_leb_size = ubi->leb_size; + vol->used_ebs = vol->reserved_pebs; + vol->last_eb_bytes = vol->reserved_pebs; + vol->used_bytes = + (long long)vol->used_ebs * (ubi->leb_size - vol->data_pad); + vol->vol_id = UBI_LAYOUT_VOLUME_ID; + vol->ref_count = 1; + + ubi_assert(!ubi->volumes[i]); + ubi->volumes[vol_id2idx(ubi, vol->vol_id)] = vol; + reserved_pebs += vol->reserved_pebs; + ubi->vol_count += 1; + vol->ubi = ubi; + err = ubi_fastmap_init_checkmap(vol, UBI_LAYOUT_VOLUME_EBS); + if (err) + return err; + + if (reserved_pebs > ubi->avail_pebs) { + ubi_err(ubi, "not enough PEBs, required %d, available %d", + reserved_pebs, ubi->avail_pebs); + if (ubi->corr_peb_count) + ubi_err(ubi, "%d PEBs are corrupted and not used", + ubi->corr_peb_count); + return -ENOSPC; + } + ubi->rsvd_pebs += reserved_pebs; + ubi->avail_pebs -= reserved_pebs; + + return 0; +} + +/** + * check_av - check volume attaching information. + * @vol: UBI volume description object + * @av: volume attaching information + * + * This function returns zero if the volume attaching information is consistent + * to the data read from the volume tabla, and %-EINVAL if not. + */ +static int check_av(const struct ubi_volume *vol, + const struct ubi_ainf_volume *av) +{ + int err; + + if (av->highest_lnum >= vol->reserved_pebs) { + err = 1; + goto bad; + } + if (av->leb_count > vol->reserved_pebs) { + err = 2; + goto bad; + } + if (av->vol_type != vol->vol_type) { + err = 3; + goto bad; + } + if (av->used_ebs > vol->reserved_pebs) { + err = 4; + goto bad; + } + if (av->data_pad != vol->data_pad) { + err = 5; + goto bad; + } + return 0; + +bad: + ubi_err(vol->ubi, "bad attaching information, error %d", err); + ubi_dump_av(av); + ubi_dump_vol_info(vol); + return -EINVAL; +} + +/** + * check_attaching_info - check that attaching information. + * @ubi: UBI device description object + * @ai: attaching information + * + * Even though we protect on-flash data by CRC checksums, we still don't trust + * the media. This function ensures that attaching information is consistent to + * the information read from the volume table. Returns zero if the attaching + * information is OK and %-EINVAL if it is not. + */ +static int check_attaching_info(const struct ubi_device *ubi, + struct ubi_attach_info *ai) +{ + int err, i; + struct ubi_ainf_volume *av; + struct ubi_volume *vol; + + if (ai->vols_found > UBI_INT_VOL_COUNT + ubi->vtbl_slots) { + ubi_err(ubi, "found %d volumes while attaching, maximum is %d + %d", + ai->vols_found, UBI_INT_VOL_COUNT, ubi->vtbl_slots); + return -EINVAL; + } + + if (ai->highest_vol_id >= ubi->vtbl_slots + UBI_INT_VOL_COUNT && + ai->highest_vol_id < UBI_INTERNAL_VOL_START) { + ubi_err(ubi, "too large volume ID %d found", + ai->highest_vol_id); + return -EINVAL; + } + + for (i = 0; i < ubi->vtbl_slots + UBI_INT_VOL_COUNT; i++) { + cond_resched(); + + av = ubi_find_av(ai, i); + vol = ubi->volumes[i]; + if (!vol) { + if (av) + ubi_remove_av(ai, av); + continue; + } + + if (vol->reserved_pebs == 0) { + ubi_assert(i < ubi->vtbl_slots); + + if (!av) + continue; + + /* + * During attaching we found a volume which does not + * exist according to the information in the volume + * table. This must have happened due to an unclean + * reboot while the volume was being removed. Discard + * these eraseblocks. + */ + ubi_msg(ubi, "finish volume %d removal", av->vol_id); + ubi_remove_av(ai, av); + } else if (av) { + err = check_av(vol, av); + if (err) + return err; + } + } + + return 0; +} + +/** + * ubi_read_volume_table - read the volume table. + * @ubi: UBI device description object + * @ai: attaching information + * + * This function reads volume table, checks it, recover from errors if needed, + * or creates it if needed. Returns zero in case of success and a negative + * error code in case of failure. + */ +int ubi_read_volume_table(struct ubi_device *ubi, struct ubi_attach_info *ai) +{ + int err; + struct ubi_ainf_volume *av; + + empty_vtbl_record.crc = cpu_to_be32(0xf116c36b); + + /* + * The number of supported volumes is limited by the eraseblock size + * and by the UBI_MAX_VOLUMES constant. + */ + + if (ubi->leb_size < UBI_VTBL_RECORD_SIZE) { + ubi_err(ubi, "LEB size too small for a volume record"); + return -EINVAL; + } + + ubi->vtbl_slots = ubi->leb_size / UBI_VTBL_RECORD_SIZE; + if (ubi->vtbl_slots > UBI_MAX_VOLUMES) + ubi->vtbl_slots = UBI_MAX_VOLUMES; + + ubi->vtbl_size = ubi->vtbl_slots * UBI_VTBL_RECORD_SIZE; + ubi->vtbl_size = ALIGN(ubi->vtbl_size, ubi->min_io_size); + + av = ubi_find_av(ai, UBI_LAYOUT_VOLUME_ID); + if (!av) { + /* + * No logical eraseblocks belonging to the layout volume were + * found. This could mean that the flash is just empty. In + * this case we create empty layout volume. + * + * But if flash is not empty this must be a corruption or the + * MTD device just contains garbage. + */ + if (ai->is_empty) { + ubi->vtbl = create_empty_lvol(ubi, ai); + if (IS_ERR(ubi->vtbl)) + return PTR_ERR(ubi->vtbl); + } else { + ubi_err(ubi, "the layout volume was not found"); + return -EINVAL; + } + } else { + if (av->leb_count > UBI_LAYOUT_VOLUME_EBS) { + /* This must not happen with proper UBI images */ + ubi_err(ubi, "too many LEBs (%d) in layout volume", + av->leb_count); + return -EINVAL; + } + + ubi->vtbl = process_lvol(ubi, ai, av); + if (IS_ERR(ubi->vtbl)) + return PTR_ERR(ubi->vtbl); + } + + ubi->avail_pebs = ubi->good_peb_count - ubi->corr_peb_count; + + /* + * The layout volume is OK, initialize the corresponding in-RAM data + * structures. + */ + err = init_volumes(ubi, ai, ubi->vtbl); + if (err) + goto out_free; + + /* + * Make sure that the attaching information is consistent to the + * information stored in the volume table. + */ + err = check_attaching_info(ubi, ai); + if (err) + goto out_free; + + return 0; + +out_free: + vfree(ubi->vtbl); + ubi_free_all_volumes(ubi); + return err; +} + +/** + * self_vtbl_check - check volume table. + * @ubi: UBI device description object + */ +static void self_vtbl_check(const struct ubi_device *ubi) +{ + if (!ubi_dbg_chk_gen(ubi)) + return; + + if (vtbl_check(ubi, ubi->vtbl)) { + ubi_err(ubi, "self-check failed"); + BUG(); + } +} diff --git a/drivers/mtd/ubi/wl.c b/drivers/mtd/ubi/wl.c new file mode 100644 index 000000000..5a0db928a --- /dev/null +++ b/drivers/mtd/ubi/wl.c @@ -0,0 +1,2160 @@ +// SPDX-License-Identifier: GPL-2.0-or-later +/* + * Copyright (c) International Business Machines Corp., 2006 + * + * Authors: Artem Bityutskiy (Битюцкий Артём), Thomas Gleixner + */ + +/* + * UBI wear-leveling sub-system. + * + * This sub-system is responsible for wear-leveling. It works in terms of + * physical eraseblocks and erase counters and knows nothing about logical + * eraseblocks, volumes, etc. From this sub-system's perspective all physical + * eraseblocks are of two types - used and free. Used physical eraseblocks are + * those that were "get" by the 'ubi_wl_get_peb()' function, and free physical + * eraseblocks are those that were put by the 'ubi_wl_put_peb()' function. + * + * Physical eraseblocks returned by 'ubi_wl_get_peb()' have only erase counter + * header. The rest of the physical eraseblock contains only %0xFF bytes. + * + * When physical eraseblocks are returned to the WL sub-system by means of the + * 'ubi_wl_put_peb()' function, they are scheduled for erasure. The erasure is + * done asynchronously in context of the per-UBI device background thread, + * which is also managed by the WL sub-system. + * + * The wear-leveling is ensured by means of moving the contents of used + * physical eraseblocks with low erase counter to free physical eraseblocks + * with high erase counter. + * + * If the WL sub-system fails to erase a physical eraseblock, it marks it as + * bad. + * + * This sub-system is also responsible for scrubbing. If a bit-flip is detected + * in a physical eraseblock, it has to be moved. Technically this is the same + * as moving it for wear-leveling reasons. + * + * As it was said, for the UBI sub-system all physical eraseblocks are either + * "free" or "used". Free eraseblock are kept in the @wl->free RB-tree, while + * used eraseblocks are kept in @wl->used, @wl->erroneous, or @wl->scrub + * RB-trees, as well as (temporarily) in the @wl->pq queue. + * + * When the WL sub-system returns a physical eraseblock, the physical + * eraseblock is protected from being moved for some "time". For this reason, + * the physical eraseblock is not directly moved from the @wl->free tree to the + * @wl->used tree. There is a protection queue in between where this + * physical eraseblock is temporarily stored (@wl->pq). + * + * All this protection stuff is needed because: + * o we don't want to move physical eraseblocks just after we have given them + * to the user; instead, we first want to let users fill them up with data; + * + * o there is a chance that the user will put the physical eraseblock very + * soon, so it makes sense not to move it for some time, but wait. + * + * Physical eraseblocks stay protected only for limited time. But the "time" is + * measured in erase cycles in this case. This is implemented with help of the + * protection queue. Eraseblocks are put to the tail of this queue when they + * are returned by the 'ubi_wl_get_peb()', and eraseblocks are removed from the + * head of the queue on each erase operation (for any eraseblock). So the + * length of the queue defines how may (global) erase cycles PEBs are protected. + * + * To put it differently, each physical eraseblock has 2 main states: free and + * used. The former state corresponds to the @wl->free tree. The latter state + * is split up on several sub-states: + * o the WL movement is allowed (@wl->used tree); + * o the WL movement is disallowed (@wl->erroneous) because the PEB is + * erroneous - e.g., there was a read error; + * o the WL movement is temporarily prohibited (@wl->pq queue); + * o scrubbing is needed (@wl->scrub tree). + * + * Depending on the sub-state, wear-leveling entries of the used physical + * eraseblocks may be kept in one of those structures. + * + * Note, in this implementation, we keep a small in-RAM object for each physical + * eraseblock. This is surely not a scalable solution. But it appears to be good + * enough for moderately large flashes and it is simple. In future, one may + * re-work this sub-system and make it more scalable. + * + * At the moment this sub-system does not utilize the sequence number, which + * was introduced relatively recently. But it would be wise to do this because + * the sequence number of a logical eraseblock characterizes how old is it. For + * example, when we move a PEB with low erase counter, and we need to pick the + * target PEB, we pick a PEB with the highest EC if our PEB is "old" and we + * pick target PEB with an average EC if our PEB is not very "old". This is a + * room for future re-works of the WL sub-system. + */ + +#include +#include +#include +#include +#include "ubi.h" +#include "wl.h" + +/* Number of physical eraseblocks reserved for wear-leveling purposes */ +#define WL_RESERVED_PEBS 1 + +/* + * Maximum number of consecutive background thread failures which is enough to + * switch to read-only mode. + */ +#define WL_MAX_FAILURES 32 + +static int self_check_ec(struct ubi_device *ubi, int pnum, int ec); +static int self_check_in_wl_tree(const struct ubi_device *ubi, + struct ubi_wl_entry *e, struct rb_root *root); +static int self_check_in_pq(const struct ubi_device *ubi, + struct ubi_wl_entry *e); + +/** + * wl_tree_add - add a wear-leveling entry to a WL RB-tree. + * @e: the wear-leveling entry to add + * @root: the root of the tree + * + * Note, we use (erase counter, physical eraseblock number) pairs as keys in + * the @ubi->used and @ubi->free RB-trees. + */ +static void wl_tree_add(struct ubi_wl_entry *e, struct rb_root *root) +{ + struct rb_node **p, *parent = NULL; + + p = &root->rb_node; + while (*p) { + struct ubi_wl_entry *e1; + + parent = *p; + e1 = rb_entry(parent, struct ubi_wl_entry, u.rb); + + if (e->ec < e1->ec) + p = &(*p)->rb_left; + else if (e->ec > e1->ec) + p = &(*p)->rb_right; + else { + ubi_assert(e->pnum != e1->pnum); + if (e->pnum < e1->pnum) + p = &(*p)->rb_left; + else + p = &(*p)->rb_right; + } + } + + rb_link_node(&e->u.rb, parent, p); + rb_insert_color(&e->u.rb, root); +} + +/** + * wl_entry_destroy - destroy a wear-leveling entry. + * @ubi: UBI device description object + * @e: the wear-leveling entry to add + * + * This function destroys a wear leveling entry and removes + * the reference from the lookup table. + */ +static void wl_entry_destroy(struct ubi_device *ubi, struct ubi_wl_entry *e) +{ + ubi->lookuptbl[e->pnum] = NULL; + kmem_cache_free(ubi_wl_entry_slab, e); +} + +/** + * do_work - do one pending work. + * @ubi: UBI device description object + * @executed: whether there is one work is executed + * + * This function returns zero in case of success and a negative error code in + * case of failure. If @executed is not NULL and there is one work executed, + * @executed is set as %1, otherwise @executed is set as %0. + */ +static int do_work(struct ubi_device *ubi, int *executed) +{ + int err; + struct ubi_work *wrk; + + cond_resched(); + + /* + * @ubi->work_sem is used to synchronize with the workers. Workers take + * it in read mode, so many of them may be doing works at a time. But + * the queue flush code has to be sure the whole queue of works is + * done, and it takes the mutex in write mode. + */ + down_read(&ubi->work_sem); + spin_lock(&ubi->wl_lock); + if (list_empty(&ubi->works)) { + spin_unlock(&ubi->wl_lock); + up_read(&ubi->work_sem); + if (executed) + *executed = 0; + return 0; + } + + if (executed) + *executed = 1; + wrk = list_entry(ubi->works.next, struct ubi_work, list); + list_del(&wrk->list); + ubi->works_count -= 1; + ubi_assert(ubi->works_count >= 0); + spin_unlock(&ubi->wl_lock); + + /* + * Call the worker function. Do not touch the work structure + * after this call as it will have been freed or reused by that + * time by the worker function. + */ + err = wrk->func(ubi, wrk, 0); + if (err) + ubi_err(ubi, "work failed with error code %d", err); + up_read(&ubi->work_sem); + + return err; +} + +/** + * in_wl_tree - check if wear-leveling entry is present in a WL RB-tree. + * @e: the wear-leveling entry to check + * @root: the root of the tree + * + * This function returns non-zero if @e is in the @root RB-tree and zero if it + * is not. + */ +static int in_wl_tree(struct ubi_wl_entry *e, struct rb_root *root) +{ + struct rb_node *p; + + p = root->rb_node; + while (p) { + struct ubi_wl_entry *e1; + + e1 = rb_entry(p, struct ubi_wl_entry, u.rb); + + if (e->pnum == e1->pnum) { + ubi_assert(e == e1); + return 1; + } + + if (e->ec < e1->ec) + p = p->rb_left; + else if (e->ec > e1->ec) + p = p->rb_right; + else { + ubi_assert(e->pnum != e1->pnum); + if (e->pnum < e1->pnum) + p = p->rb_left; + else + p = p->rb_right; + } + } + + return 0; +} + +/** + * in_pq - check if a wear-leveling entry is present in the protection queue. + * @ubi: UBI device description object + * @e: the wear-leveling entry to check + * + * This function returns non-zero if @e is in the protection queue and zero + * if it is not. + */ +static inline int in_pq(const struct ubi_device *ubi, struct ubi_wl_entry *e) +{ + struct ubi_wl_entry *p; + int i; + + for (i = 0; i < UBI_PROT_QUEUE_LEN; ++i) + list_for_each_entry(p, &ubi->pq[i], u.list) + if (p == e) + return 1; + + return 0; +} + +/** + * prot_queue_add - add physical eraseblock to the protection queue. + * @ubi: UBI device description object + * @e: the physical eraseblock to add + * + * This function adds @e to the tail of the protection queue @ubi->pq, where + * @e will stay for %UBI_PROT_QUEUE_LEN erase operations and will be + * temporarily protected from the wear-leveling worker. Note, @wl->lock has to + * be locked. + */ +static void prot_queue_add(struct ubi_device *ubi, struct ubi_wl_entry *e) +{ + int pq_tail = ubi->pq_head - 1; + + if (pq_tail < 0) + pq_tail = UBI_PROT_QUEUE_LEN - 1; + ubi_assert(pq_tail >= 0 && pq_tail < UBI_PROT_QUEUE_LEN); + list_add_tail(&e->u.list, &ubi->pq[pq_tail]); + dbg_wl("added PEB %d EC %d to the protection queue", e->pnum, e->ec); +} + +/** + * find_wl_entry - find wear-leveling entry closest to certain erase counter. + * @ubi: UBI device description object + * @root: the RB-tree where to look for + * @diff: maximum possible difference from the smallest erase counter + * @pick_max: pick PEB even its erase counter beyonds 'min_ec + @diff' + * + * This function looks for a wear leveling entry with erase counter closest to + * min + @diff, where min is the smallest erase counter. + */ +static struct ubi_wl_entry *find_wl_entry(struct ubi_device *ubi, + struct rb_root *root, int diff, + int pick_max) +{ + struct rb_node *p; + struct ubi_wl_entry *e; + int max; + + e = rb_entry(rb_first(root), struct ubi_wl_entry, u.rb); + max = e->ec + diff; + + p = root->rb_node; + while (p) { + struct ubi_wl_entry *e1; + + e1 = rb_entry(p, struct ubi_wl_entry, u.rb); + if (e1->ec >= max) { + if (pick_max) + e = e1; + p = p->rb_left; + } else { + p = p->rb_right; + e = e1; + } + } + + return e; +} + +/** + * find_mean_wl_entry - find wear-leveling entry with medium erase counter. + * @ubi: UBI device description object + * @root: the RB-tree where to look for + * + * This function looks for a wear leveling entry with medium erase counter, + * but not greater or equivalent than the lowest erase counter plus + * @ubi->wl_free_max_diff/2. + */ +static struct ubi_wl_entry *find_mean_wl_entry(struct ubi_device *ubi, + struct rb_root *root) +{ + struct ubi_wl_entry *e, *first, *last; + + first = rb_entry(rb_first(root), struct ubi_wl_entry, u.rb); + last = rb_entry(rb_last(root), struct ubi_wl_entry, u.rb); + + if (last->ec - first->ec < ubi->wl_free_max_diff) { + e = rb_entry(root->rb_node, struct ubi_wl_entry, u.rb); + + /* + * If no fastmap has been written and fm_anchor is not + * reserved and this WL entry can be used as anchor PEB + * hold it back and return the second best WL entry such + * that fastmap can use the anchor PEB later. + */ + e = may_reserve_for_fm(ubi, e, root); + } else + e = find_wl_entry(ubi, root, ubi->wl_free_max_diff/2, 0); + + return e; +} + +/** + * wl_get_wle - get a mean wl entry to be used by ubi_wl_get_peb() or + * refill_wl_user_pool(). + * @ubi: UBI device description object + * + * This function returns a wear leveling entry in case of success and + * NULL in case of failure. + */ +static struct ubi_wl_entry *wl_get_wle(struct ubi_device *ubi) +{ + struct ubi_wl_entry *e; + + e = find_mean_wl_entry(ubi, &ubi->free); + if (!e) { + ubi_err(ubi, "no free eraseblocks"); + return NULL; + } + + self_check_in_wl_tree(ubi, e, &ubi->free); + + /* + * Move the physical eraseblock to the protection queue where it will + * be protected from being moved for some time. + */ + rb_erase(&e->u.rb, &ubi->free); + ubi->free_count--; + dbg_wl("PEB %d EC %d", e->pnum, e->ec); + + return e; +} + +/** + * prot_queue_del - remove a physical eraseblock from the protection queue. + * @ubi: UBI device description object + * @pnum: the physical eraseblock to remove + * + * This function deletes PEB @pnum from the protection queue and returns zero + * in case of success and %-ENODEV if the PEB was not found. + */ +static int prot_queue_del(struct ubi_device *ubi, int pnum) +{ + struct ubi_wl_entry *e; + + e = ubi->lookuptbl[pnum]; + if (!e) + return -ENODEV; + + if (self_check_in_pq(ubi, e)) + return -ENODEV; + + list_del(&e->u.list); + dbg_wl("deleted PEB %d from the protection queue", e->pnum); + return 0; +} + +/** + * ubi_sync_erase - synchronously erase a physical eraseblock. + * @ubi: UBI device description object + * @e: the physical eraseblock to erase + * @torture: if the physical eraseblock has to be tortured; cleared to zero + * once the torture test has completed successfully so that a retry + * of the erase does not torture the physical eraseblock again + * + * This function returns zero in case of success and a negative error code in + * case of failure. + */ +int ubi_sync_erase(struct ubi_device *ubi, struct ubi_wl_entry *e, int *torture) +{ + int err; + struct ubi_ec_hdr *ec_hdr; + unsigned long long ec = e->ec; + + dbg_wl("erase PEB %d, old EC %llu", e->pnum, ec); + + err = self_check_ec(ubi, e->pnum, e->ec); + if (err) + return -EINVAL; + + ec_hdr = kzalloc(ubi->ec_hdr_alsize, GFP_NOFS); + if (!ec_hdr) + return -ENOMEM; + + err = ubi_io_sync_erase(ubi, e->pnum, torture); + if (err < 0) + goto out_free; + + ec += err; + if (ec > UBI_MAX_ERASECOUNTER) { + /* + * Erase counter overflow. Upgrade UBI and use 64-bit + * erase counters internally. + */ + ubi_err(ubi, "erase counter overflow at PEB %d, EC %llu", + e->pnum, ec); + err = -EINVAL; + goto out_free; + } + + dbg_wl("erased PEB %d, new EC %llu", e->pnum, ec); + + ec_hdr->ec = cpu_to_be64(ec); + + err = ubi_io_write_ec_hdr(ubi, e->pnum, ec_hdr); + if (err) + goto out_free; + + e->ec = ec; + spin_lock(&ubi->wl_lock); + if (e->ec > ubi->max_ec) + ubi->max_ec = e->ec; + spin_unlock(&ubi->wl_lock); + +out_free: + kfree(ec_hdr); + return err; +} + +/** + * serve_prot_queue - check if it is time to stop protecting PEBs. + * @ubi: UBI device description object + * + * This function is called after each erase operation and removes PEBs from the + * tail of the protection queue. These PEBs have been protected for long enough + * and should be moved to the used tree. + */ +static void serve_prot_queue(struct ubi_device *ubi) +{ + struct ubi_wl_entry *e, *tmp; + int count; + + /* + * There may be several protected physical eraseblock to remove, + * process them all. + */ +repeat: + count = 0; + spin_lock(&ubi->wl_lock); + list_for_each_entry_safe(e, tmp, &ubi->pq[ubi->pq_head], u.list) { + dbg_wl("PEB %d EC %d protection over, move to used tree", + e->pnum, e->ec); + + list_del(&e->u.list); + wl_tree_add(e, &ubi->used); + if (count++ > 32) { + /* + * Let's be nice and avoid holding the spinlock for + * too long. + */ + spin_unlock(&ubi->wl_lock); + cond_resched(); + goto repeat; + } + } + + ubi->pq_head += 1; + if (ubi->pq_head == UBI_PROT_QUEUE_LEN) + ubi->pq_head = 0; + ubi_assert(ubi->pq_head >= 0 && ubi->pq_head < UBI_PROT_QUEUE_LEN); + spin_unlock(&ubi->wl_lock); +} + +/** + * __schedule_ubi_work - schedule a work. + * @ubi: UBI device description object + * @wrk: the work to schedule + * + * This function adds a work defined by @wrk to the tail of the pending works + * list. Can only be used if ubi->work_sem is already held in read mode! + */ +static void __schedule_ubi_work(struct ubi_device *ubi, struct ubi_work *wrk) +{ + spin_lock(&ubi->wl_lock); + list_add_tail(&wrk->list, &ubi->works); + ubi_assert(ubi->works_count >= 0); + ubi->works_count += 1; + if (ubi->thread_enabled && !ubi_dbg_is_bgt_disabled(ubi)) + wake_up_process(ubi->bgt_thread); + spin_unlock(&ubi->wl_lock); +} + +/** + * schedule_ubi_work - schedule a work. + * @ubi: UBI device description object + * @wrk: the work to schedule + * + * This function adds a work defined by @wrk to the tail of the pending works + * list. + */ +static void schedule_ubi_work(struct ubi_device *ubi, struct ubi_work *wrk) +{ + down_read(&ubi->work_sem); + __schedule_ubi_work(ubi, wrk); + up_read(&ubi->work_sem); +} + +static int erase_worker(struct ubi_device *ubi, struct ubi_work *wl_wrk, + int shutdown); + +/** + * schedule_erase - schedule an erase work. + * @ubi: UBI device description object + * @e: the WL entry of the physical eraseblock to erase + * @vol_id: the volume ID that last used this PEB + * @lnum: the last used logical eraseblock number for the PEB + * @torture: if the physical eraseblock has to be tortured + * @nested: denotes whether the work_sem is already held + * + * This function returns zero in case of success and a %-ENOMEM in case of + * failure. + */ +static int schedule_erase(struct ubi_device *ubi, struct ubi_wl_entry *e, + int vol_id, int lnum, int torture, bool nested) +{ + struct ubi_work *wl_wrk; + + ubi_assert(e); + + dbg_wl("schedule erasure of PEB %d, EC %d, torture %d", + e->pnum, e->ec, torture); + + wl_wrk = kmalloc_obj(struct ubi_work, GFP_NOFS); + if (!wl_wrk) + return -ENOMEM; + + wl_wrk->func = &erase_worker; + wl_wrk->e = e; + wl_wrk->vol_id = vol_id; + wl_wrk->lnum = lnum; + wl_wrk->torture = torture; + + if (nested) + __schedule_ubi_work(ubi, wl_wrk); + else + schedule_ubi_work(ubi, wl_wrk); + return 0; +} + +static int __erase_worker(struct ubi_device *ubi, struct ubi_work *wl_wrk); +/** + * do_sync_erase - run the erase worker synchronously. + * @ubi: UBI device description object + * @e: the WL entry of the physical eraseblock to erase + * @vol_id: the volume ID that last used this PEB + * @lnum: the last used logical eraseblock number for the PEB + * @torture: if the physical eraseblock has to be tortured + * + */ +static int do_sync_erase(struct ubi_device *ubi, struct ubi_wl_entry *e, + int vol_id, int lnum, int torture) +{ + struct ubi_work wl_wrk; + + dbg_wl("sync erase of PEB %i", e->pnum); + + wl_wrk.e = e; + wl_wrk.vol_id = vol_id; + wl_wrk.lnum = lnum; + wl_wrk.torture = torture; + + return __erase_worker(ubi, &wl_wrk); +} + +static int ensure_wear_leveling(struct ubi_device *ubi, int nested); +/** + * wear_leveling_worker - wear-leveling worker function. + * @ubi: UBI device description object + * @wrk: the work object + * @shutdown: non-zero if the worker has to free memory and exit + * because the WL-subsystem is shutting down + * + * This function copies a more worn out physical eraseblock to a less worn out + * one. Returns zero in case of success and a negative error code in case of + * failure. + */ +static int wear_leveling_worker(struct ubi_device *ubi, struct ubi_work *wrk, + int shutdown) +{ + int err, scrubbing = 0, torture = 0, protect = 0, erroneous = 0; + int erase = 0, keep = 0, vol_id = -1, lnum = -1; + struct ubi_wl_entry *e1, *e2; + struct ubi_vid_io_buf *vidb; + struct ubi_vid_hdr *vid_hdr; + int dst_leb_clean = 0; + + kfree(wrk); + if (shutdown) + return 0; + + vidb = ubi_alloc_vid_buf(ubi, GFP_NOFS); + if (!vidb) + return -ENOMEM; + + vid_hdr = ubi_get_vid_hdr(vidb); + + down_read(&ubi->fm_eba_sem); + mutex_lock(&ubi->move_mutex); + spin_lock(&ubi->wl_lock); + ubi_assert(!ubi->move_from && !ubi->move_to); + ubi_assert(!ubi->move_to_put); + +#ifdef CONFIG_MTD_UBI_FASTMAP + if (!next_peb_for_wl(ubi, true) || +#else + if (!ubi->free.rb_node || +#endif + (!ubi->used.rb_node && !ubi->scrub.rb_node)) { + /* + * No free physical eraseblocks? Well, they must be waiting in + * the queue to be erased. Cancel movement - it will be + * triggered again when a free physical eraseblock appears. + * + * No used physical eraseblocks? They must be temporarily + * protected from being moved. They will be moved to the + * @ubi->used tree later and the wear-leveling will be + * triggered again. + */ + dbg_wl("cancel WL, a list is empty: free %d, used %d", + !ubi->free.rb_node, !ubi->used.rb_node); + goto out_cancel; + } + +#ifdef CONFIG_MTD_UBI_FASTMAP + e1 = find_anchor_wl_entry(&ubi->used); + if (e1 && ubi->fm_anchor && + (ubi->fm_anchor->ec - e1->ec >= ubi->wl_threshold)) { + ubi->fm_do_produce_anchor = 1; + /* + * fm_anchor is no longer considered a good anchor. + * NULL assignment also prevents multiple wear level checks + * of this PEB. + */ + wl_tree_add(ubi->fm_anchor, &ubi->free); + ubi->fm_anchor = NULL; + ubi->free_count++; + } + + if (ubi->fm_do_produce_anchor) { + if (!e1) + goto out_cancel; + e2 = get_peb_for_wl(ubi); + if (!e2) + goto out_cancel; + + self_check_in_wl_tree(ubi, e1, &ubi->used); + rb_erase(&e1->u.rb, &ubi->used); + dbg_wl("anchor-move PEB %d to PEB %d", e1->pnum, e2->pnum); + ubi->fm_do_produce_anchor = 0; + } else if (!ubi->scrub.rb_node) { +#else + if (!ubi->scrub.rb_node) { +#endif + /* + * Now pick the least worn-out used physical eraseblock and a + * highly worn-out free physical eraseblock. If the erase + * counters differ much enough, start wear-leveling. + */ + e1 = rb_entry(rb_first(&ubi->used), struct ubi_wl_entry, u.rb); + e2 = get_peb_for_wl(ubi); + if (!e2) + goto out_cancel; + + if (!(e2->ec - e1->ec >= ubi->wl_threshold)) { + dbg_wl("no WL needed: min used EC %d, max free EC %d", + e1->ec, e2->ec); + + /* Give the unused PEB back */ + wl_tree_add(e2, &ubi->free); + ubi->free_count++; + goto out_cancel; + } + self_check_in_wl_tree(ubi, e1, &ubi->used); + rb_erase(&e1->u.rb, &ubi->used); + dbg_wl("move PEB %d EC %d to PEB %d EC %d", + e1->pnum, e1->ec, e2->pnum, e2->ec); + } else { + /* Perform scrubbing */ + scrubbing = 1; + e1 = rb_entry(rb_first(&ubi->scrub), struct ubi_wl_entry, u.rb); + e2 = get_peb_for_wl(ubi); + if (!e2) + goto out_cancel; + + self_check_in_wl_tree(ubi, e1, &ubi->scrub); + rb_erase(&e1->u.rb, &ubi->scrub); + dbg_wl("scrub PEB %d to PEB %d", e1->pnum, e2->pnum); + } + + ubi->move_from = e1; + ubi->move_to = e2; + spin_unlock(&ubi->wl_lock); + + /* + * Now we are going to copy physical eraseblock @e1->pnum to @e2->pnum. + * We so far do not know which logical eraseblock our physical + * eraseblock (@e1) belongs to. We have to read the volume identifier + * header first. + * + * Note, we are protected from this PEB being unmapped and erased. The + * 'ubi_wl_put_peb()' would wait for moving to be finished if the PEB + * which is being moved was unmapped. + */ + + err = ubi_io_read_vid_hdr(ubi, e1->pnum, vidb, 0); + if (err && err != UBI_IO_BITFLIPS) { + dst_leb_clean = 1; + if (err == UBI_IO_FF) { + /* + * We are trying to move PEB without a VID header. UBI + * always write VID headers shortly after the PEB was + * given, so we have a situation when it has not yet + * had a chance to write it, because it was preempted. + * So add this PEB to the protection queue so far, + * because presumably more data will be written there + * (including the missing VID header), and then we'll + * move it. + */ + dbg_wl("PEB %d has no VID header", e1->pnum); + protect = 1; + goto out_not_moved; + } else if (err == UBI_IO_FF_BITFLIPS) { + /* + * The same situation as %UBI_IO_FF, but bit-flips were + * detected. It is better to schedule this PEB for + * scrubbing. + */ + dbg_wl("PEB %d has no VID header but has bit-flips", + e1->pnum); + scrubbing = 1; + goto out_not_moved; + } else if (ubi->fast_attach && err == UBI_IO_BAD_HDR_EBADMSG) { + /* + * While a full scan would detect interrupted erasures + * at attach time we can face them here when attached from + * Fastmap. + */ + dbg_wl("PEB %d has ECC errors, maybe from an interrupted erasure", + e1->pnum); + erase = 1; + goto out_not_moved; + } + + ubi_err(ubi, "error %d while reading VID header from PEB %d", + err, e1->pnum); + goto out_error; + } + + vol_id = be32_to_cpu(vid_hdr->vol_id); + lnum = be32_to_cpu(vid_hdr->lnum); + + err = ubi_eba_copy_leb(ubi, e1->pnum, e2->pnum, vidb); + if (err) { + if (err == MOVE_CANCEL_RACE) { + /* + * The LEB has not been moved because the volume is + * being deleted or the PEB has been put meanwhile. We + * should prevent this PEB from being selected for + * wear-leveling movement again, so put it to the + * protection queue. + */ + protect = 1; + dst_leb_clean = 1; + goto out_not_moved; + } + if (err == MOVE_RETRY) { + /* + * For source PEB: + * 1. The scrubbing is set for scrub type PEB, it will + * be put back into ubi->scrub list. + * 2. Non-scrub type PEB will be put back into ubi->used + * list. + */ + keep = 1; + dst_leb_clean = 1; + goto out_not_moved; + } + if (err == MOVE_TARGET_BITFLIPS || err == MOVE_TARGET_WR_ERR || + err == MOVE_TARGET_RD_ERR) { + /* + * Target PEB had bit-flips or write error - torture it. + */ + torture = 1; + keep = 1; + goto out_not_moved; + } + + if (err == MOVE_SOURCE_RD_ERR) { + /* + * An error happened while reading the source PEB. Do + * not switch to R/O mode in this case, and give the + * upper layers a possibility to recover from this, + * e.g. by unmapping corresponding LEB. Instead, just + * put this PEB to the @ubi->erroneous list to prevent + * UBI from trying to move it over and over again. + */ + if (ubi->erroneous_peb_count > ubi->max_erroneous) { + ubi_err(ubi, "too many erroneous eraseblocks (%d)", + ubi->erroneous_peb_count); + goto out_error; + } + dst_leb_clean = 1; + erroneous = 1; + goto out_not_moved; + } + + if (err < 0) + goto out_error; + + ubi_assert(0); + } + + /* The PEB has been successfully moved */ + if (scrubbing) + ubi_msg(ubi, "scrubbed PEB %d (LEB %d:%d), data moved to PEB %d", + e1->pnum, vol_id, lnum, e2->pnum); + ubi_free_vid_buf(vidb); + + spin_lock(&ubi->wl_lock); + if (!ubi->move_to_put) { + wl_tree_add(e2, &ubi->used); + e2 = NULL; + } + ubi->move_from = ubi->move_to = NULL; + ubi->move_to_put = ubi->wl_scheduled = 0; + spin_unlock(&ubi->wl_lock); + + err = do_sync_erase(ubi, e1, vol_id, lnum, 0); + if (err) { + if (e2) { + spin_lock(&ubi->wl_lock); + wl_entry_destroy(ubi, e2); + spin_unlock(&ubi->wl_lock); + } + goto out_ro; + } + + if (e2) { + /* + * Well, the target PEB was put meanwhile, schedule it for + * erasure. + */ + dbg_wl("PEB %d (LEB %d:%d) was put meanwhile, erase", + e2->pnum, vol_id, lnum); + err = do_sync_erase(ubi, e2, vol_id, lnum, 0); + if (err) + goto out_ro; + } + + dbg_wl("done"); + mutex_unlock(&ubi->move_mutex); + up_read(&ubi->fm_eba_sem); + return 0; + + /* + * For some reasons the LEB was not moved, might be an error, might be + * something else. @e1 was not changed, so return it back. @e2 might + * have been changed, schedule it for erasure. + */ +out_not_moved: + if (vol_id != -1) + dbg_wl("cancel moving PEB %d (LEB %d:%d) to PEB %d (%d)", + e1->pnum, vol_id, lnum, e2->pnum, err); + else + dbg_wl("cancel moving PEB %d to PEB %d (%d)", + e1->pnum, e2->pnum, err); + spin_lock(&ubi->wl_lock); + if (protect) + prot_queue_add(ubi, e1); + else if (erroneous) { + wl_tree_add(e1, &ubi->erroneous); + ubi->erroneous_peb_count += 1; + } else if (scrubbing) + wl_tree_add(e1, &ubi->scrub); + else if (keep) + wl_tree_add(e1, &ubi->used); + if (dst_leb_clean) { + wl_tree_add(e2, &ubi->free); + ubi->free_count++; + } + + ubi_assert(!ubi->move_to_put); + ubi->move_from = ubi->move_to = NULL; + ubi->wl_scheduled = 0; + spin_unlock(&ubi->wl_lock); + + ubi_free_vid_buf(vidb); + if (dst_leb_clean) { + ensure_wear_leveling(ubi, 1); + } else { + err = do_sync_erase(ubi, e2, vol_id, lnum, torture); + if (err) + goto out_ro; + } + + if (erase) { + err = do_sync_erase(ubi, e1, vol_id, lnum, 1); + if (err) + goto out_ro; + } + + mutex_unlock(&ubi->move_mutex); + up_read(&ubi->fm_eba_sem); + return 0; + +out_error: + if (vol_id != -1) + ubi_err(ubi, "error %d while moving PEB %d to PEB %d", + err, e1->pnum, e2->pnum); + else + ubi_err(ubi, "error %d while moving PEB %d (LEB %d:%d) to PEB %d", + err, e1->pnum, vol_id, lnum, e2->pnum); + spin_lock(&ubi->wl_lock); + ubi->move_from = ubi->move_to = NULL; + ubi->move_to_put = ubi->wl_scheduled = 0; + wl_entry_destroy(ubi, e1); + wl_entry_destroy(ubi, e2); + spin_unlock(&ubi->wl_lock); + + ubi_free_vid_buf(vidb); + +out_ro: + ubi_ro_mode(ubi); + mutex_unlock(&ubi->move_mutex); + up_read(&ubi->fm_eba_sem); + ubi_assert(err != 0); + return err < 0 ? err : -EIO; + +out_cancel: + ubi->wl_scheduled = 0; + spin_unlock(&ubi->wl_lock); + mutex_unlock(&ubi->move_mutex); + up_read(&ubi->fm_eba_sem); + ubi_free_vid_buf(vidb); + return 0; +} + +/** + * ensure_wear_leveling - schedule wear-leveling if it is needed. + * @ubi: UBI device description object + * @nested: set to non-zero if this function is called from UBI worker + * + * This function checks if it is time to start wear-leveling and schedules it + * if yes. This function returns zero in case of success and a negative error + * code in case of failure. + */ +static int ensure_wear_leveling(struct ubi_device *ubi, int nested) +{ + int err = 0; + struct ubi_work *wrk; + + spin_lock(&ubi->wl_lock); + if (ubi->wl_scheduled) + /* Wear-leveling is already in the work queue */ + goto out_unlock; + + /* + * If the ubi->scrub tree is not empty, scrubbing is needed, and the + * WL worker has to be scheduled anyway. + */ + if (!ubi->scrub.rb_node) { +#ifdef CONFIG_MTD_UBI_FASTMAP + if (!need_wear_leveling(ubi)) + goto out_unlock; +#else + struct ubi_wl_entry *e1; + struct ubi_wl_entry *e2; + + if (!ubi->used.rb_node || !ubi->free.rb_node) + /* No physical eraseblocks - no deal */ + goto out_unlock; + + /* + * We schedule wear-leveling only if the difference between the + * lowest erase counter of used physical eraseblocks and a high + * erase counter of free physical eraseblocks is greater than + * @ubi->wl_threshold. + */ + e1 = rb_entry(rb_first(&ubi->used), struct ubi_wl_entry, u.rb); + e2 = find_wl_entry(ubi, &ubi->free, ubi->wl_free_max_diff, 0); + + if (!(e2->ec - e1->ec >= ubi->wl_threshold)) + goto out_unlock; +#endif + dbg_wl("schedule wear-leveling"); + } else + dbg_wl("schedule scrubbing"); + + ubi->wl_scheduled = 1; + spin_unlock(&ubi->wl_lock); + + wrk = kmalloc_obj(struct ubi_work, GFP_NOFS); + if (!wrk) { + err = -ENOMEM; + goto out_cancel; + } + + wrk->func = &wear_leveling_worker; + if (nested) + __schedule_ubi_work(ubi, wrk); + else + schedule_ubi_work(ubi, wrk); + return err; + +out_cancel: + spin_lock(&ubi->wl_lock); + ubi->wl_scheduled = 0; +out_unlock: + spin_unlock(&ubi->wl_lock); + return err; +} + +/** + * __erase_worker - physical eraseblock erase worker function. + * @ubi: UBI device description object + * @wl_wrk: the work object + * + * This function erases a physical eraseblock and perform torture testing if + * needed. It also takes care about marking the physical eraseblock bad if + * needed. Returns zero in case of success and a negative error code in case of + * failure. + */ +static int __erase_worker(struct ubi_device *ubi, struct ubi_work *wl_wrk) +{ + struct ubi_wl_entry *e = wl_wrk->e; + int pnum = e->pnum; + int vol_id = wl_wrk->vol_id; + int lnum = wl_wrk->lnum; + int err, available_consumed = 0; + + dbg_wl("erase PEB %d EC %d LEB %d:%d", + pnum, e->ec, wl_wrk->vol_id, wl_wrk->lnum); + + err = ubi_sync_erase(ubi, e, &wl_wrk->torture); + if (!err) { + spin_lock(&ubi->wl_lock); + + if (!ubi->fm_disabled && !ubi->fm_anchor && + e->pnum < UBI_FM_MAX_START) { + /* + * Abort anchor production, if needed it will be + * enabled again in the wear leveling started below. + */ + ubi->fm_anchor = e; + ubi->fm_do_produce_anchor = 0; + } else { + wl_tree_add(e, &ubi->free); + ubi->free_count++; + } + + spin_unlock(&ubi->wl_lock); + + /* + * One more erase operation has happened, take care about + * protected physical eraseblocks. + */ + serve_prot_queue(ubi); + + /* And take care about wear-leveling */ + err = ensure_wear_leveling(ubi, 1); + return err; + } + + ubi_err(ubi, "failed to erase PEB %d, error %d", pnum, err); + + if (err == -EINTR || err == -ENOMEM || err == -EAGAIN || + err == -EBUSY) { + int err1; + + /* Re-schedule the LEB for erasure */ + err1 = schedule_erase(ubi, e, vol_id, lnum, wl_wrk->torture, + true); + if (err1) { + spin_lock(&ubi->wl_lock); + wl_entry_destroy(ubi, e); + spin_unlock(&ubi->wl_lock); + err = err1; + goto out_ro; + } + return err; + } + + spin_lock(&ubi->wl_lock); + wl_entry_destroy(ubi, e); + spin_unlock(&ubi->wl_lock); + if (err != -EIO) + /* + * If this is not %-EIO, we have no idea what to do. Scheduling + * this physical eraseblock for erasure again would cause + * errors again and again. Well, lets switch to R/O mode. + */ + goto out_ro; + + /* It is %-EIO, the PEB went bad */ + + if (!ubi->bad_allowed) { + ubi_err(ubi, "bad physical eraseblock %d detected", pnum); + goto out_ro; + } + + spin_lock(&ubi->volumes_lock); + if (ubi->beb_rsvd_pebs == 0) { + if (ubi->avail_pebs == 0) { + spin_unlock(&ubi->volumes_lock); + ubi_err(ubi, "no reserved/available physical eraseblocks"); + goto out_ro; + } + ubi->avail_pebs -= 1; + available_consumed = 1; + } + spin_unlock(&ubi->volumes_lock); + + ubi_msg(ubi, "mark PEB %d as bad", pnum); + err = ubi_io_mark_bad(ubi, pnum); + if (err) + goto out_ro; + + spin_lock(&ubi->volumes_lock); + if (ubi->beb_rsvd_pebs > 0) { + if (available_consumed) { + /* + * The amount of reserved PEBs increased since we last + * checked. + */ + ubi->avail_pebs += 1; + available_consumed = 0; + } + ubi->beb_rsvd_pebs -= 1; + } + ubi->bad_peb_count += 1; + ubi->good_peb_count -= 1; + ubi_calculate_reserved(ubi); + if (available_consumed) + ubi_warn(ubi, "no PEBs in the reserved pool, used an available PEB"); + else if (ubi->beb_rsvd_pebs) + ubi_msg(ubi, "%d PEBs left in the reserve", + ubi->beb_rsvd_pebs); + else + ubi_warn(ubi, "last PEB from the reserve was used"); + spin_unlock(&ubi->volumes_lock); + + return err; + +out_ro: + if (available_consumed) { + spin_lock(&ubi->volumes_lock); + ubi->avail_pebs += 1; + spin_unlock(&ubi->volumes_lock); + } + ubi_ro_mode(ubi); + return err; +} + +static int erase_worker(struct ubi_device *ubi, struct ubi_work *wl_wrk, + int shutdown) +{ + int ret; + + if (shutdown) { + struct ubi_wl_entry *e = wl_wrk->e; + + dbg_wl("cancel erasure of PEB %d EC %d", e->pnum, e->ec); + kfree(wl_wrk); + wl_entry_destroy(ubi, e); + return 0; + } + + ret = __erase_worker(ubi, wl_wrk); + kfree(wl_wrk); + return ret; +} + +/** + * ubi_wl_put_peb - return a PEB to the wear-leveling sub-system. + * @ubi: UBI device description object + * @vol_id: the volume ID that last used this PEB + * @lnum: the last used logical eraseblock number for the PEB + * @pnum: physical eraseblock to return + * @torture: if this physical eraseblock has to be tortured + * + * This function is called to return physical eraseblock @pnum to the pool of + * free physical eraseblocks. The @torture flag has to be set if an I/O error + * occurred to this @pnum and it has to be tested. This function returns zero + * in case of success, and a negative error code in case of failure. + */ +int ubi_wl_put_peb(struct ubi_device *ubi, int vol_id, int lnum, + int pnum, int torture) +{ + int err; + struct ubi_wl_entry *e; + + dbg_wl("PEB %d", pnum); + ubi_assert(pnum >= 0); + ubi_assert(pnum < ubi->peb_count); + + down_read(&ubi->fm_protect); + +retry: + spin_lock(&ubi->wl_lock); + e = ubi->lookuptbl[pnum]; + if (!e) { + /* + * This wl entry has been removed for some errors by other + * process (eg. wear leveling worker), corresponding process + * (except __erase_worker, which cannot concurrent with + * ubi_wl_put_peb) will set ubi ro_mode at the same time, + * just ignore this wl entry. + */ + spin_unlock(&ubi->wl_lock); + up_read(&ubi->fm_protect); + return 0; + } + if (e == ubi->move_from) { + /* + * User is putting the physical eraseblock which was selected to + * be moved. It will be scheduled for erasure in the + * wear-leveling worker. + */ + dbg_wl("PEB %d is being moved, wait", pnum); + spin_unlock(&ubi->wl_lock); + + /* Wait for the WL worker by taking the @ubi->move_mutex */ + mutex_lock(&ubi->move_mutex); + mutex_unlock(&ubi->move_mutex); + goto retry; + } else if (e == ubi->move_to) { + /* + * User is putting the physical eraseblock which was selected + * as the target the data is moved to. It may happen if the EBA + * sub-system already re-mapped the LEB in 'ubi_eba_copy_leb()' + * but the WL sub-system has not put the PEB to the "used" tree + * yet, but it is about to do this. So we just set a flag which + * will tell the WL worker that the PEB is not needed anymore + * and should be scheduled for erasure. + */ + dbg_wl("PEB %d is the target of data moving", pnum); + ubi_assert(!ubi->move_to_put); + ubi->move_to_put = 1; + spin_unlock(&ubi->wl_lock); + up_read(&ubi->fm_protect); + return 0; + } else { + if (in_wl_tree(e, &ubi->used)) { + self_check_in_wl_tree(ubi, e, &ubi->used); + rb_erase(&e->u.rb, &ubi->used); + } else if (in_wl_tree(e, &ubi->scrub)) { + self_check_in_wl_tree(ubi, e, &ubi->scrub); + rb_erase(&e->u.rb, &ubi->scrub); + } else if (in_wl_tree(e, &ubi->erroneous)) { + self_check_in_wl_tree(ubi, e, &ubi->erroneous); + rb_erase(&e->u.rb, &ubi->erroneous); + ubi->erroneous_peb_count -= 1; + ubi_assert(ubi->erroneous_peb_count >= 0); + /* Erroneous PEBs should be tortured */ + torture = 1; + } else { + err = prot_queue_del(ubi, e->pnum); + if (err) { + ubi_err(ubi, "PEB %d not found", pnum); + ubi_ro_mode(ubi); + spin_unlock(&ubi->wl_lock); + up_read(&ubi->fm_protect); + return err; + } + } + } + spin_unlock(&ubi->wl_lock); + + err = schedule_erase(ubi, e, vol_id, lnum, torture, false); + if (err) { + spin_lock(&ubi->wl_lock); + wl_tree_add(e, &ubi->used); + spin_unlock(&ubi->wl_lock); + } + + up_read(&ubi->fm_protect); + return err; +} + +/** + * ubi_wl_scrub_peb - schedule a physical eraseblock for scrubbing. + * @ubi: UBI device description object + * @pnum: the physical eraseblock to schedule + * + * If a bit-flip in a physical eraseblock is detected, this physical eraseblock + * needs scrubbing. This function schedules a physical eraseblock for + * scrubbing which is done in background. This function returns zero in case of + * success and a negative error code in case of failure. + */ +int ubi_wl_scrub_peb(struct ubi_device *ubi, int pnum) +{ + struct ubi_wl_entry *e; + + ubi_msg(ubi, "schedule PEB %d for scrubbing", pnum); + +retry: + spin_lock(&ubi->wl_lock); + e = ubi->lookuptbl[pnum]; + if (e == ubi->move_from || in_wl_tree(e, &ubi->scrub) || + in_wl_tree(e, &ubi->erroneous)) { + spin_unlock(&ubi->wl_lock); + return 0; + } + + if (e == ubi->move_to) { + /* + * This physical eraseblock was used to move data to. The data + * was moved but the PEB was not yet inserted to the proper + * tree. We should just wait a little and let the WL worker + * proceed. + */ + spin_unlock(&ubi->wl_lock); + dbg_wl("the PEB %d is not in proper tree, retry", pnum); + yield(); + goto retry; + } + + if (in_wl_tree(e, &ubi->used)) { + self_check_in_wl_tree(ubi, e, &ubi->used); + rb_erase(&e->u.rb, &ubi->used); + } else { + int err; + + err = prot_queue_del(ubi, e->pnum); + if (err) { + ubi_err(ubi, "PEB %d not found", pnum); + ubi_ro_mode(ubi); + spin_unlock(&ubi->wl_lock); + return err; + } + } + + wl_tree_add(e, &ubi->scrub); + spin_unlock(&ubi->wl_lock); + + /* + * Technically scrubbing is the same as wear-leveling, so it is done + * by the WL worker. + */ + return ensure_wear_leveling(ubi, 0); +} + +/** + * ubi_wl_flush - flush all pending works. + * @ubi: UBI device description object + * @vol_id: the volume id to flush for + * @lnum: the logical eraseblock number to flush for + * + * This function executes all pending works for a particular volume id / + * logical eraseblock number pair. If either value is set to %UBI_ALL, then it + * acts as a wildcard for all of the corresponding volume numbers or logical + * eraseblock numbers. It returns zero in case of success and a negative error + * code in case of failure. + */ +int ubi_wl_flush(struct ubi_device *ubi, int vol_id, int lnum) +{ + int err = 0; + int found = 1; + + /* + * Erase while the pending works queue is not empty, but not more than + * the number of currently pending works. + */ + dbg_wl("flush pending work for LEB %d:%d (%d pending works)", + vol_id, lnum, ubi->works_count); + + while (found) { + struct ubi_work *wrk, *tmp; + found = 0; + + down_read(&ubi->work_sem); + spin_lock(&ubi->wl_lock); + list_for_each_entry_safe(wrk, tmp, &ubi->works, list) { + if ((vol_id == UBI_ALL || wrk->vol_id == vol_id) && + (lnum == UBI_ALL || wrk->lnum == lnum)) { + list_del(&wrk->list); + ubi->works_count -= 1; + ubi_assert(ubi->works_count >= 0); + spin_unlock(&ubi->wl_lock); + + err = wrk->func(ubi, wrk, 0); + if (err) { + up_read(&ubi->work_sem); + return err; + } + + spin_lock(&ubi->wl_lock); + found = 1; + break; + } + } + spin_unlock(&ubi->wl_lock); + up_read(&ubi->work_sem); + } + + /* + * Make sure all the works which have been done in parallel are + * finished. + */ + down_write(&ubi->work_sem); + up_write(&ubi->work_sem); + + return err; +} + +static bool scrub_possible(struct ubi_device *ubi, struct ubi_wl_entry *e) +{ + if (in_wl_tree(e, &ubi->scrub)) + return false; + else if (in_wl_tree(e, &ubi->erroneous)) + return false; + else if (ubi->move_from == e) + return false; + else if (ubi->move_to == e) + return false; + + return true; +} + +/** + * ubi_bitflip_check - Check an eraseblock for bitflips and scrub it if needed. + * @ubi: UBI device description object + * @pnum: the physical eraseblock to schedule + * @force: don't read the block, assume bitflips happened and take action. + * + * This function reads the given eraseblock and checks if bitflips occured. + * In case of bitflips, the eraseblock is scheduled for scrubbing. + * If scrubbing is forced with @force, the eraseblock is not read, + * but scheduled for scrubbing right away. + * + * Returns: + * %EINVAL, PEB is out of range + * %ENOENT, PEB is no longer used by UBI + * %EBUSY, PEB cannot be checked now or a check is currently running on it + * %EAGAIN, bit flips happened but scrubbing is currently not possible + * %EUCLEAN, bit flips happened and PEB is scheduled for scrubbing + * %0, no bit flips detected + */ +int ubi_bitflip_check(struct ubi_device *ubi, int pnum, int force) +{ + int err = 0; + struct ubi_wl_entry *e; + + if (pnum < 0 || pnum >= ubi->peb_count) { + err = -EINVAL; + goto out; + } + + /* + * Pause all parallel work, otherwise it can happen that the + * erase worker frees a wl entry under us. + */ + down_write(&ubi->work_sem); + + /* + * Make sure that the wl entry does not change state while + * inspecting it. + */ + spin_lock(&ubi->wl_lock); + e = ubi->lookuptbl[pnum]; + if (!e) { + spin_unlock(&ubi->wl_lock); + err = -ENOENT; + goto out_resume; + } + + /* + * Does it make sense to check this PEB? + */ + if (!scrub_possible(ubi, e)) { + spin_unlock(&ubi->wl_lock); + err = -EBUSY; + goto out_resume; + } + spin_unlock(&ubi->wl_lock); + + if (!force) { + mutex_lock(&ubi->buf_mutex); + err = ubi_io_read(ubi, ubi->peb_buf, pnum, 0, ubi->peb_size); + mutex_unlock(&ubi->buf_mutex); + } + + if (force || err == UBI_IO_BITFLIPS) { + /* + * Okay, bit flip happened, let's figure out what we can do. + */ + spin_lock(&ubi->wl_lock); + + /* + * Recheck. We released wl_lock, UBI might have killed the + * wl entry under us. + */ + e = ubi->lookuptbl[pnum]; + if (!e) { + spin_unlock(&ubi->wl_lock); + err = -ENOENT; + goto out_resume; + } + + /* + * Need to re-check state + */ + if (!scrub_possible(ubi, e)) { + spin_unlock(&ubi->wl_lock); + err = -EBUSY; + goto out_resume; + } + + if (in_pq(ubi, e)) { + prot_queue_del(ubi, e->pnum); + wl_tree_add(e, &ubi->scrub); + spin_unlock(&ubi->wl_lock); + + err = ensure_wear_leveling(ubi, 1); + } else if (in_wl_tree(e, &ubi->used)) { + rb_erase(&e->u.rb, &ubi->used); + wl_tree_add(e, &ubi->scrub); + spin_unlock(&ubi->wl_lock); + + err = ensure_wear_leveling(ubi, 1); + } else if (in_wl_tree(e, &ubi->free)) { + rb_erase(&e->u.rb, &ubi->free); + ubi->free_count--; + spin_unlock(&ubi->wl_lock); + + /* + * This PEB is empty we can schedule it for + * erasure right away. No wear leveling needed. + */ + err = schedule_erase(ubi, e, UBI_UNKNOWN, UBI_UNKNOWN, + force ? 0 : 1, true); + } else { + spin_unlock(&ubi->wl_lock); + err = -EAGAIN; + } + + if (!err && !force) + err = -EUCLEAN; + } else { + err = 0; + } + +out_resume: + up_write(&ubi->work_sem); +out: + + return err; +} + +/** + * tree_destroy - destroy an RB-tree. + * @ubi: UBI device description object + * @root: the root of the tree to destroy + */ +static void tree_destroy(struct ubi_device *ubi, struct rb_root *root) +{ + struct rb_node *rb; + struct ubi_wl_entry *e; + + rb = root->rb_node; + while (rb) { + if (rb->rb_left) + rb = rb->rb_left; + else if (rb->rb_right) + rb = rb->rb_right; + else { + e = rb_entry(rb, struct ubi_wl_entry, u.rb); + + rb = rb_parent(rb); + if (rb) { + if (rb->rb_left == &e->u.rb) + rb->rb_left = NULL; + else + rb->rb_right = NULL; + } + + wl_entry_destroy(ubi, e); + } + } +} + +/** + * ubi_thread - UBI background thread. + * @u: the UBI device description object pointer + */ +int ubi_thread(void *u) +{ + int failures = 0; + struct ubi_device *ubi = u; + + ubi_msg(ubi, "background thread \"%s\" started, PID %d", + ubi->bgt_name, task_pid_nr(current)); + + set_freezable(); + for (;;) { + int err; + + if (kthread_should_stop()) + break; + + if (try_to_freeze()) + continue; + + spin_lock(&ubi->wl_lock); + if (list_empty(&ubi->works) || ubi->ro_mode || + !ubi->thread_enabled || ubi_dbg_is_bgt_disabled(ubi)) { + set_current_state(TASK_INTERRUPTIBLE); + spin_unlock(&ubi->wl_lock); + + /* + * Check kthread_should_stop() after we set the task + * state to guarantee that we either see the stop bit + * and exit or the task state is reset to runnable such + * that it's not scheduled out indefinitely and detects + * the stop bit at kthread_should_stop(). + */ + if (kthread_should_stop()) { + set_current_state(TASK_RUNNING); + break; + } + + schedule(); + continue; + } + spin_unlock(&ubi->wl_lock); + + err = do_work(ubi, NULL); + if (err) { + ubi_err(ubi, "%s: work failed with error code %d", + ubi->bgt_name, err); + if (failures++ > WL_MAX_FAILURES) { + /* + * Too many failures, disable the thread and + * switch to read-only mode. + */ + ubi_msg(ubi, "%s: %d consecutive failures", + ubi->bgt_name, WL_MAX_FAILURES); + ubi_ro_mode(ubi); + ubi->thread_enabled = 0; + continue; + } + } else + failures = 0; + + cond_resched(); + } + + dbg_wl("background thread \"%s\" is killed", ubi->bgt_name); + ubi->thread_enabled = 0; + return 0; +} + +/** + * shutdown_work - shutdown all pending works. + * @ubi: UBI device description object + */ +static void shutdown_work(struct ubi_device *ubi) +{ + while (!list_empty(&ubi->works)) { + struct ubi_work *wrk; + + wrk = list_entry(ubi->works.next, struct ubi_work, list); + list_del(&wrk->list); + wrk->func(ubi, wrk, 1); + ubi->works_count -= 1; + ubi_assert(ubi->works_count >= 0); + } +} + +/** + * erase_aeb - erase a PEB given in UBI attach info PEB + * @ubi: UBI device description object + * @aeb: UBI attach info PEB + * @sync: If true, erase synchronously. Otherwise schedule for erasure + */ +static int erase_aeb(struct ubi_device *ubi, struct ubi_ainf_peb *aeb, bool sync) +{ + struct ubi_wl_entry *e; + int err, torture = 0; + + e = kmem_cache_alloc(ubi_wl_entry_slab, GFP_KERNEL); + if (!e) + return -ENOMEM; + + e->pnum = aeb->pnum; + e->ec = aeb->ec; + ubi->lookuptbl[e->pnum] = e; + + if (sync) { + err = ubi_sync_erase(ubi, e, &torture); + if (err) + goto out_free; + + wl_tree_add(e, &ubi->free); + ubi->free_count++; + } else { + err = schedule_erase(ubi, e, aeb->vol_id, aeb->lnum, 0, false); + if (err) + goto out_free; + } + + return 0; + +out_free: + wl_entry_destroy(ubi, e); + + return err; +} + +/** + * ubi_wl_init - initialize the WL sub-system using attaching information. + * @ubi: UBI device description object + * @ai: attaching information + * + * This function returns zero in case of success, and a negative error code in + * case of failure. + */ +int ubi_wl_init(struct ubi_device *ubi, struct ubi_attach_info *ai) +{ + int err, i, reserved_pebs, found_pebs = 0; + struct rb_node *rb1, *rb2; + struct ubi_ainf_volume *av; + struct ubi_ainf_peb *aeb, *tmp; + struct ubi_wl_entry *e; + + ubi->used = ubi->erroneous = ubi->free = ubi->scrub = RB_ROOT; + spin_lock_init(&ubi->wl_lock); + mutex_init(&ubi->move_mutex); + init_rwsem(&ubi->work_sem); + ubi->max_ec = ai->max_ec; + INIT_LIST_HEAD(&ubi->works); + + sprintf(ubi->bgt_name, UBI_BGT_NAME_PATTERN, ubi->ubi_num); + + err = -ENOMEM; + ubi->lookuptbl = kcalloc(ubi->peb_count, sizeof(void *), GFP_KERNEL); + if (!ubi->lookuptbl) + return err; + + for (i = 0; i < UBI_PROT_QUEUE_LEN; i++) + INIT_LIST_HEAD(&ubi->pq[i]); + ubi->pq_head = 0; + + ubi->free_count = 0; + list_for_each_entry_safe(aeb, tmp, &ai->erase, u.list) { + cond_resched(); + + err = erase_aeb(ubi, aeb, false); + if (err) + goto out_free; + + found_pebs++; + } + + list_for_each_entry(aeb, &ai->free, u.list) { + cond_resched(); + + e = kmem_cache_alloc(ubi_wl_entry_slab, GFP_KERNEL); + if (!e) { + err = -ENOMEM; + goto out_free; + } + + e->pnum = aeb->pnum; + e->ec = aeb->ec; + ubi_assert(e->ec >= 0); + + wl_tree_add(e, &ubi->free); + ubi->free_count++; + + ubi->lookuptbl[e->pnum] = e; + + found_pebs++; + } + + ubi_rb_for_each_entry(rb1, av, &ai->volumes, rb) { + ubi_rb_for_each_entry(rb2, aeb, &av->root, u.rb) { + cond_resched(); + + e = kmem_cache_alloc(ubi_wl_entry_slab, GFP_KERNEL); + if (!e) { + err = -ENOMEM; + goto out_free; + } + + e->pnum = aeb->pnum; + e->ec = aeb->ec; + ubi->lookuptbl[e->pnum] = e; + + if (!aeb->scrub) { + dbg_wl("add PEB %d EC %d to the used tree", + e->pnum, e->ec); + wl_tree_add(e, &ubi->used); + } else { + dbg_wl("add PEB %d EC %d to the scrub tree", + e->pnum, e->ec); + wl_tree_add(e, &ubi->scrub); + } + + found_pebs++; + } + } + + list_for_each_entry(aeb, &ai->fastmap, u.list) { + cond_resched(); + + e = ubi_find_fm_block(ubi, aeb->pnum); + + if (e) { + ubi_assert(!ubi->lookuptbl[e->pnum]); + ubi->lookuptbl[e->pnum] = e; + } else { + bool sync = false; + + /* + * Usually old Fastmap PEBs are scheduled for erasure + * and we don't have to care about them but if we face + * an power cut before scheduling them we need to + * take care of them here. + */ + if (ubi->lookuptbl[aeb->pnum]) + continue; + + /* + * The fastmap update code might not find a free PEB for + * writing the fastmap anchor to and then reuses the + * current fastmap anchor PEB. When this PEB gets erased + * and a power cut happens before it is written again we + * must make sure that the fastmap attach code doesn't + * find any outdated fastmap anchors, hence we erase the + * outdated fastmap anchor PEBs synchronously here. + */ + if (aeb->vol_id == UBI_FM_SB_VOLUME_ID) + sync = true; + + err = erase_aeb(ubi, aeb, sync); + if (err) + goto out_free; + } + + found_pebs++; + } + + dbg_wl("found %i PEBs", found_pebs); + + ubi_assert(ubi->good_peb_count == found_pebs); + + reserved_pebs = WL_RESERVED_PEBS; + ubi_fastmap_init(ubi, &reserved_pebs); + + if (ubi->avail_pebs < reserved_pebs) { + ubi_err(ubi, "no enough physical eraseblocks (%d, need %d)", + ubi->avail_pebs, reserved_pebs); + if (ubi->corr_peb_count) + ubi_err(ubi, "%d PEBs are corrupted and not used", + ubi->corr_peb_count); + err = -ENOSPC; + goto out_free; + } + ubi->avail_pebs -= reserved_pebs; + ubi->rsvd_pebs += reserved_pebs; + + /* Schedule wear-leveling if needed */ + err = ensure_wear_leveling(ubi, 0); + if (err) + goto out_free; + +#ifdef CONFIG_MTD_UBI_FASTMAP + if (!ubi->ro_mode && !ubi->fm_disabled) + ubi_ensure_anchor_pebs(ubi); +#endif + return 0; + +out_free: + shutdown_work(ubi); + tree_destroy(ubi, &ubi->used); + tree_destroy(ubi, &ubi->free); + tree_destroy(ubi, &ubi->scrub); + kfree(ubi->lookuptbl); + return err; +} + +/** + * protection_queue_destroy - destroy the protection queue. + * @ubi: UBI device description object + */ +static void protection_queue_destroy(struct ubi_device *ubi) +{ + int i; + struct ubi_wl_entry *e, *tmp; + + for (i = 0; i < UBI_PROT_QUEUE_LEN; ++i) { + list_for_each_entry_safe(e, tmp, &ubi->pq[i], u.list) { + list_del(&e->u.list); + wl_entry_destroy(ubi, e); + } + } +} + +/** + * ubi_wl_close - close the wear-leveling sub-system. + * @ubi: UBI device description object + */ +void ubi_wl_close(struct ubi_device *ubi) +{ + dbg_wl("close the WL sub-system"); + ubi_fastmap_close(ubi); + shutdown_work(ubi); + protection_queue_destroy(ubi); + tree_destroy(ubi, &ubi->used); + tree_destroy(ubi, &ubi->erroneous); + tree_destroy(ubi, &ubi->free); + tree_destroy(ubi, &ubi->scrub); + kfree(ubi->lookuptbl); +} + +/** + * self_check_ec - make sure that the erase counter of a PEB is correct. + * @ubi: UBI device description object + * @pnum: the physical eraseblock number to check + * @ec: the erase counter to check + * + * This function returns zero if the erase counter of physical eraseblock @pnum + * is equivalent to @ec, and a negative error code if not or if an error + * occurred. + */ +static int self_check_ec(struct ubi_device *ubi, int pnum, int ec) +{ + int err; + long long read_ec; + struct ubi_ec_hdr *ec_hdr; + + if (!ubi_dbg_chk_gen(ubi)) + return 0; + + ec_hdr = kzalloc(ubi->ec_hdr_alsize, GFP_NOFS); + if (!ec_hdr) + return -ENOMEM; + + err = ubi_io_read_ec_hdr(ubi, pnum, ec_hdr, 0); + if (err && err != UBI_IO_BITFLIPS) { + /* The header does not have to exist */ + err = 0; + goto out_free; + } + + read_ec = be64_to_cpu(ec_hdr->ec); + if (ec != read_ec && read_ec - ec > 1) { + ubi_err(ubi, "self-check failed for PEB %d", pnum); + ubi_err(ubi, "read EC is %lld, should be %d", read_ec, ec); + dump_stack(); + err = 1; + } else + err = 0; + +out_free: + kfree(ec_hdr); + return err; +} + +/** + * self_check_in_wl_tree - check that wear-leveling entry is in WL RB-tree. + * @ubi: UBI device description object + * @e: the wear-leveling entry to check + * @root: the root of the tree + * + * This function returns zero if @e is in the @root RB-tree and %-EINVAL if it + * is not. + */ +static int self_check_in_wl_tree(const struct ubi_device *ubi, + struct ubi_wl_entry *e, struct rb_root *root) +{ + if (!ubi_dbg_chk_gen(ubi)) + return 0; + + if (in_wl_tree(e, root)) + return 0; + + ubi_err(ubi, "self-check failed for PEB %d, EC %d, RB-tree %p ", + e->pnum, e->ec, root); + dump_stack(); + return -EINVAL; +} + +/** + * self_check_in_pq - check if wear-leveling entry is in the protection + * queue. + * @ubi: UBI device description object + * @e: the wear-leveling entry to check + * + * This function returns zero if @e is in @ubi->pq and %-EINVAL if it is not. + */ +static int self_check_in_pq(const struct ubi_device *ubi, + struct ubi_wl_entry *e) +{ + if (!ubi_dbg_chk_gen(ubi)) + return 0; + + if (in_pq(ubi, e)) + return 0; + + ubi_err(ubi, "self-check failed for PEB %d, EC %d, Protect queue", + e->pnum, e->ec); + dump_stack(); + return -EINVAL; +} +#ifndef CONFIG_MTD_UBI_FASTMAP +static struct ubi_wl_entry *get_peb_for_wl(struct ubi_device *ubi) +{ + struct ubi_wl_entry *e; + + e = find_wl_entry(ubi, &ubi->free, ubi->wl_free_max_diff, 0); + self_check_in_wl_tree(ubi, e, &ubi->free); + ubi->free_count--; + ubi_assert(ubi->free_count >= 0); + rb_erase(&e->u.rb, &ubi->free); + + return e; +} + +/** + * produce_free_peb - produce a free physical eraseblock. + * @ubi: UBI device description object + * + * This function tries to make a free PEB by means of synchronous execution of + * pending works. This may be needed if, for example the background thread is + * disabled. Returns zero in case of success and a negative error code in case + * of failure. + */ +static int produce_free_peb(struct ubi_device *ubi) +{ + int err; + + while (!ubi->free.rb_node && ubi->works_count) { + spin_unlock(&ubi->wl_lock); + + dbg_wl("do one work synchronously"); + err = do_work(ubi, NULL); + + spin_lock(&ubi->wl_lock); + if (err) + return err; + } + + return 0; +} + +/** + * ubi_wl_get_peb - get a physical eraseblock. + * @ubi: UBI device description object + * + * This function returns a physical eraseblock in case of success and a + * negative error code in case of failure. + * Returns with ubi->fm_eba_sem held in read mode! + */ +int ubi_wl_get_peb(struct ubi_device *ubi) +{ + int err; + struct ubi_wl_entry *e; + +retry: + down_read(&ubi->fm_eba_sem); + spin_lock(&ubi->wl_lock); + if (!ubi->free.rb_node) { + if (ubi->works_count == 0) { + ubi_err(ubi, "no free eraseblocks"); + ubi_assert(list_empty(&ubi->works)); + spin_unlock(&ubi->wl_lock); + return -ENOSPC; + } + + err = produce_free_peb(ubi); + if (err < 0) { + spin_unlock(&ubi->wl_lock); + return err; + } + spin_unlock(&ubi->wl_lock); + up_read(&ubi->fm_eba_sem); + goto retry; + + } + e = wl_get_wle(ubi); + prot_queue_add(ubi, e); + spin_unlock(&ubi->wl_lock); + + err = ubi_self_check_all_ff(ubi, e->pnum, ubi->vid_hdr_aloffset, + ubi->peb_size - ubi->vid_hdr_aloffset); + if (err) { + ubi_err(ubi, "new PEB %d does not contain all 0xFF bytes", e->pnum); + return err; + } + + return e->pnum; +} +#else +#include "fastmap-wl.c" +#endif diff --git a/drivers/mtd/ubi/wl.h b/drivers/mtd/ubi/wl.h new file mode 100644 index 000000000..a69169c35 --- /dev/null +++ b/drivers/mtd/ubi/wl.h @@ -0,0 +1,33 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef UBI_WL_H +#define UBI_WL_H +#ifdef CONFIG_MTD_UBI_FASTMAP +static void update_fastmap_work_fn(struct work_struct *wrk); +static struct ubi_wl_entry *find_anchor_wl_entry(struct rb_root *root); +static struct ubi_wl_entry *get_peb_for_wl(struct ubi_device *ubi); +static struct ubi_wl_entry *next_peb_for_wl(struct ubi_device *ubi, + bool need_fill); +static bool need_wear_leveling(struct ubi_device *ubi); +static void ubi_fastmap_close(struct ubi_device *ubi); +static inline void ubi_fastmap_init(struct ubi_device *ubi, int *count) +{ + if (ubi->fm_disabled) + ubi->fm_pool_rsv_cnt = 0; + /* Reserve enough LEBs to store two fastmaps and to fill pools. */ + *count += (ubi->fm_size / ubi->leb_size) * 2 + ubi->fm_pool_rsv_cnt; + INIT_WORK(&ubi->fm_work, update_fastmap_work_fn); +} +static struct ubi_wl_entry *may_reserve_for_fm(struct ubi_device *ubi, + struct ubi_wl_entry *e, + struct rb_root *root); +#else /* !CONFIG_MTD_UBI_FASTMAP */ +static struct ubi_wl_entry *get_peb_for_wl(struct ubi_device *ubi); +static inline void ubi_fastmap_close(struct ubi_device *ubi) { } +static inline void ubi_fastmap_init(struct ubi_device *ubi, int *count) { } +static struct ubi_wl_entry *may_reserve_for_fm(struct ubi_device *ubi, + struct ubi_wl_entry *e, + struct rb_root *root) { + return e; +} +#endif /* CONFIG_MTD_UBI_FASTMAP */ +#endif /* UBI_WL_H */ -- cgit v1.3.1