KVM: Don't pre-reserve xarray entries when storing empty/NULL attributesgrafted

Skip the xarray reservation loop when clearing all memory attributes, as
storing NULL only erases the entry and never needs to allocate, so no
reservation (and no cleanup of a failed one) is required in that case.

Suggested-by: Sean Christopherson <seanjc@google.com>
Cc: David Ballesteros <davimaba.v@proton.me>
Signed-off-by: Zeng Chi <zengchi@kylinos.cn>
Link: https://patch.msgid.link/20260921102442.1232375-1-zeng_chi911@163.com
[sean: split to separate patch]
Signed-off-by: Sean Christopherson <seanjc@google.com>

5 files changed, 1072 insertions, 0 deletions

diff --git a/include/uapi/mtd/inftl-user.h b/include/uapi/mtd/inftl-user.h
new file mode 100644
index 000000000..8d5044e32
--- /dev/null
+++ b/include/uapi/mtd/inftl-user.h
@@ -0,0 +1,92 @@
+/* SPDX-License-Identifier: GPL-2.0 WITH Linux-syscall-note */
+/*
+ * Parts of INFTL headers shared with userspace
+ *
+ */
+
+#ifndef __MTD_INFTL_USER_H__
+#define __MTD_INFTL_USER_H__
+
+#include <linux/types.h>
+
+#define	OSAK_VERSION	0x5120
+#define	PERCENTUSED	98
+
+#define	SECTORSIZE	512
+
+/* Block Control Information */
+
+struct inftl_bci {
+	__u8 ECCsig[6];
+	__u8 Status;
+	__u8 Status1;
+} __attribute__((packed));
+
+struct inftl_unithead1 {
+	__u16 virtualUnitNo;
+	__u16 prevUnitNo;
+	__u8 ANAC;
+	__u8 NACs;
+	__u8 parityPerField;
+	__u8 discarded;
+} __attribute__((packed));
+
+struct inftl_unithead2 {
+	__u8 parityPerField;
+	__u8 ANAC;
+	__u16 prevUnitNo;
+	__u16 virtualUnitNo;
+	__u8 NACs;
+	__u8 discarded;
+} __attribute__((packed));
+
+struct inftl_unittail {
+	__u8 Reserved[4];
+	__u16 EraseMark;
+	__u16 EraseMark1;
+} __attribute__((packed));
+
+union inftl_uci {
+	struct inftl_unithead1 a;
+	struct inftl_unithead2 b;
+	struct inftl_unittail c;
+};
+
+struct inftl_oob {
+	struct inftl_bci b;
+	union inftl_uci u;
+};
+
+
+/* INFTL Media Header */
+
+struct INFTLPartition {
+	__u32 virtualUnits;
+	__u32 firstUnit;
+	__u32 lastUnit;
+	__u32 flags;
+	__u32 spareUnits;
+	__u32 Reserved0;
+	__u32 Reserved1;
+} __attribute__((packed));
+
+struct INFTLMediaHeader {
+	char bootRecordID[8];
+	__u32 NoOfBootImageBlocks;
+	__u32 NoOfBinaryPartitions;
+	__u32 NoOfBDTLPartitions;
+	__u32 BlockMultiplierBits;
+	__u32 FormatFlags;
+	__u32 OsakVersion;
+	__u32 PercentUsed;
+	struct INFTLPartition Partitions[4];
+} __attribute__((packed));
+
+/* Partition flag types */
+#define	INFTL_BINARY	0x20000000
+#define	INFTL_BDTL	0x40000000
+#define	INFTL_LAST	0x80000000
+
+#endif /* __MTD_INFTL_USER_H__ */
+
+
diff --git a/include/uapi/mtd/mtd-abi.h b/include/uapi/mtd/mtd-abi.h
new file mode 100644
index 000000000..714d55b49
--- /dev/null
+++ b/include/uapi/mtd/mtd-abi.h
@@ -0,0 +1,342 @@
+/* SPDX-License-Identifier: GPL-2.0+ WITH Linux-syscall-note */
+/*
+ * Copyright © 1999-2010 David Woodhouse <dwmw2@infradead.org> et al.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
+ *
+ */
+
+#ifndef __MTD_ABI_H__
+#define __MTD_ABI_H__
+
+#include <linux/types.h>
+
+struct erase_info_user {
+	__u32 start;
+	__u32 length;
+};
+
+struct erase_info_user64 {
+	__u64 start;
+	__u64 length;
+};
+
+struct mtd_oob_buf {
+	__u32 start;
+	__u32 length;
+	unsigned char __user *ptr;
+};
+
+struct mtd_oob_buf64 {
+	__u64 start;
+	__u32 pad;
+	__u32 length;
+	__u64 usr_ptr;
+};
+
+/**
+ * MTD operation modes
+ *
+ * @MTD_OPS_PLACE_OOB:	OOB data are placed at the given offset (default)
+ * @MTD_OPS_AUTO_OOB:	OOB data are automatically placed at the free areas
+ *			which are defined by the internal ecclayout
+ * @MTD_OPS_RAW:	data are transferred as-is, with no error correction;
+ *			this mode implies %MTD_OPS_PLACE_OOB
+ *
+ * These modes can be passed to ioctl(MEMWRITE) and ioctl(MEMREAD); they are
+ * also used internally. See notes on "MTD file modes" for discussion on
+ * %MTD_OPS_RAW vs. %MTD_FILE_MODE_RAW.
+ */
+enum {
+	MTD_OPS_PLACE_OOB = 0,
+	MTD_OPS_AUTO_OOB = 1,
+	MTD_OPS_RAW = 2,
+};
+
+/**
+ * struct mtd_write_req - data structure for requesting a write operation
+ *
+ * @start:	start address
+ * @len:	length of data buffer (only lower 32 bits are used)
+ * @ooblen:	length of OOB buffer (only lower 32 bits are used)
+ * @usr_data:	user-provided data buffer
+ * @usr_oob:	user-provided OOB buffer
+ * @mode:	MTD mode (see "MTD operation modes")
+ * @padding:	reserved, must be set to 0
+ *
+ * This structure supports ioctl(MEMWRITE) operations, allowing data and/or OOB
+ * writes in various modes. To write to OOB-only, set @usr_data == NULL, and to
+ * write data-only, set @usr_oob == NULL. However, setting both @usr_data and
+ * @usr_oob to NULL is not allowed.
+ */
+struct mtd_write_req {
+	__u64 start;
+	__u64 len;
+	__u64 ooblen;
+	__u64 usr_data;
+	__u64 usr_oob;
+	__u8 mode;
+	__u8 padding[7];
+};
+
+/**
+ * struct mtd_read_req_ecc_stats - ECC statistics for a read operation
+ *
+ * @uncorrectable_errors: the number of uncorrectable errors that happened
+ *			  during the read operation
+ * @corrected_bitflips: the number of bitflips corrected during the read
+ *			operation
+ * @max_bitflips: the maximum number of bitflips detected in any single ECC
+ *		  step for the data read during the operation; this information
+ *		  can be used to decide whether the data stored in a specific
+ *		  region of the MTD device should be moved somewhere else to
+ *		  avoid data loss.
+ */
+struct mtd_read_req_ecc_stats {
+	__u32 uncorrectable_errors;
+	__u32 corrected_bitflips;
+	__u32 max_bitflips;
+};
+
+/**
+ * struct mtd_read_req - data structure for requesting a read operation
+ *
+ * @start:	start address
+ * @len:	length of data buffer (only lower 32 bits are used)
+ * @ooblen:	length of OOB buffer (only lower 32 bits are used)
+ * @usr_data:	user-provided data buffer
+ * @usr_oob:	user-provided OOB buffer
+ * @mode:	MTD mode (see "MTD operation modes")
+ * @padding:	reserved, must be set to 0
+ * @ecc_stats:	ECC statistics for the read operation
+ *
+ * This structure supports ioctl(MEMREAD) operations, allowing data and/or OOB
+ * reads in various modes. To read from OOB-only, set @usr_data == NULL, and to
+ * read data-only, set @usr_oob == NULL. However, setting both @usr_data and
+ * @usr_oob to NULL is not allowed.
+ */
+struct mtd_read_req {
+	__u64 start;
+	__u64 len;
+	__u64 ooblen;
+	__u64 usr_data;
+	__u64 usr_oob;
+	__u8 mode;
+	__u8 padding[7];
+	struct mtd_read_req_ecc_stats ecc_stats;
+};
+
+#define MTD_ABSENT		0
+#define MTD_RAM			1
+#define MTD_ROM			2
+#define MTD_NORFLASH		3
+#define MTD_NANDFLASH		4	/* SLC NAND */
+#define MTD_DATAFLASH		6
+#define MTD_UBIVOLUME		7
+#define MTD_MLCNANDFLASH	8	/* MLC NAND (including TLC) */
+
+#define MTD_WRITEABLE		0x400	/* Device is writeable */
+#define MTD_BIT_WRITEABLE	0x800	/* Single bits can be flipped */
+#define MTD_NO_ERASE		0x1000	/* No erase necessary */
+#define MTD_POWERUP_LOCK	0x2000	/* Always locked after reset */
+#define MTD_SLC_ON_MLC_EMULATION 0x4000	/* Emulate SLC behavior on MLC NANDs */
+
+/* Some common devices / combinations of capabilities */
+#define MTD_CAP_ROM		0
+#define MTD_CAP_RAM		(MTD_WRITEABLE | MTD_BIT_WRITEABLE | MTD_NO_ERASE)
+#define MTD_CAP_NORFLASH	(MTD_WRITEABLE | MTD_BIT_WRITEABLE)
+#define MTD_CAP_NANDFLASH	(MTD_WRITEABLE)
+#define MTD_CAP_NVRAM		(MTD_WRITEABLE | MTD_BIT_WRITEABLE | MTD_NO_ERASE)
+
+/* Obsolete ECC byte placement modes (used with obsolete MEMGETOOBSEL) */
+#define MTD_NANDECC_OFF		0	/* Switch off ECC (Not recommended) */
+#define MTD_NANDECC_PLACE	1	/* Use the given placement in the structure (YAFFS1 legacy mode) */
+#define MTD_NANDECC_AUTOPLACE	2	/* Use the default placement scheme */
+#define MTD_NANDECC_PLACEONLY	3	/* Use the given placement in the structure (Do not store ecc result on read) */
+#define MTD_NANDECC_AUTOPL_USR 	4	/* Use the given autoplacement scheme rather than using the default */
+
+/* OTP mode selection */
+#define MTD_OTP_OFF		0
+#define MTD_OTP_FACTORY		1
+#define MTD_OTP_USER		2
+
+struct mtd_info_user {
+	__u8 type;
+	__u32 flags;
+	__u32 size;	/* Total size of the MTD */
+	__u32 erasesize;
+	__u32 writesize;
+	__u32 oobsize;	/* Amount of OOB data per block (e.g. 16) */
+	__u64 padding;	/* Old obsolete field; do not use */
+};
+
+struct region_info_user {
+	__u32 offset;		/* At which this region starts,
+				 * from the beginning of the MTD */
+	__u32 erasesize;	/* For this region */
+	__u32 numblocks;	/* Number of blocks in this region */
+	__u32 regionindex;
+};
+
+struct otp_info {
+	__u32 start;
+	__u32 length;
+	__u32 locked;
+};
+
+/*
+ * Note, the following ioctl existed in the past and was removed:
+ * #define MEMSETOOBSEL           _IOW('M', 9, struct nand_oobinfo)
+ * Try to avoid adding a new ioctl with the same ioctl number.
+ */
+
+/* Get basic MTD characteristics info (better to use sysfs) */
+#define MEMGETINFO		_IOR('M', 1, struct mtd_info_user)
+/* Erase segment of MTD */
+#define MEMERASE		_IOW('M', 2, struct erase_info_user)
+/* Write out-of-band data from MTD */
+#define MEMWRITEOOB		_IOWR('M', 3, struct mtd_oob_buf)
+/* Read out-of-band data from MTD */
+#define MEMREADOOB		_IOWR('M', 4, struct mtd_oob_buf)
+/* Lock a chip (for MTD that supports it) */
+#define MEMLOCK			_IOW('M', 5, struct erase_info_user)
+/* Unlock a chip (for MTD that supports it) */
+#define MEMUNLOCK		_IOW('M', 6, struct erase_info_user)
+/* Get the number of different erase regions */
+#define MEMGETREGIONCOUNT	_IOR('M', 7, int)
+/* Get information about the erase region for a specific index */
+#define MEMGETREGIONINFO	_IOWR('M', 8, struct region_info_user)
+/* Get info about OOB modes (e.g., RAW, PLACE, AUTO) - legacy interface */
+#define MEMGETOOBSEL		_IOR('M', 10, struct nand_oobinfo)
+/* Check if an eraseblock is bad */
+#define MEMGETBADBLOCK		_IOW('M', 11, __kernel_loff_t)
+/* Mark an eraseblock as bad */
+#define MEMSETBADBLOCK		_IOW('M', 12, __kernel_loff_t)
+/* Set OTP (One-Time Programmable) mode (factory vs. user) */
+#define OTPSELECT		_IOR('M', 13, int)
+/* Get number of OTP (One-Time Programmable) regions */
+#define OTPGETREGIONCOUNT	_IOW('M', 14, int)
+/* Get all OTP (One-Time Programmable) info about MTD */
+#define OTPGETREGIONINFO	_IOW('M', 15, struct otp_info)
+/* Lock a given range of user data (must be in mode %MTD_FILE_MODE_OTP_USER) */
+#define OTPLOCK			_IOR('M', 16, struct otp_info)
+/* Get ECC layout (deprecated) */
+#define ECCGETLAYOUT		_IOR('M', 17, struct nand_ecclayout_user)
+/* Get statistics about corrected/uncorrected errors */
+#define ECCGETSTATS		_IOR('M', 18, struct mtd_ecc_stats)
+/* Set MTD mode on a per-file-descriptor basis (see "MTD file modes") */
+#define MTDFILEMODE		_IO('M', 19)
+/* Erase segment of MTD (supports 64-bit address) */
+#define MEMERASE64		_IOW('M', 20, struct erase_info_user64)
+/* Write data to OOB (64-bit version) */
+#define MEMWRITEOOB64		_IOWR('M', 21, struct mtd_oob_buf64)
+/* Read data from OOB (64-bit version) */
+#define MEMREADOOB64		_IOWR('M', 22, struct mtd_oob_buf64)
+/* Check if chip is locked (for MTD that supports it) */
+#define MEMISLOCKED		_IOR('M', 23, struct erase_info_user)
+/*
+ * Most generic write interface; can write in-band and/or out-of-band in various
+ * modes (see "struct mtd_write_req"). This ioctl is not supported for flashes
+ * without OOB, e.g., NOR flash.
+ */
+#define MEMWRITE		_IOWR('M', 24, struct mtd_write_req)
+/* Erase a given range of user data (must be in mode %MTD_FILE_MODE_OTP_USER) */
+#define OTPERASE		_IOW('M', 25, struct otp_info)
+/*
+ * Most generic read interface; can read in-band and/or out-of-band in various
+ * modes (see "struct mtd_read_req"). This ioctl is not supported for flashes
+ * without OOB, e.g., NOR flash.
+ */
+#define MEMREAD			_IOWR('M', 26, struct mtd_read_req)
+
+/*
+ * Obsolete legacy interface. Keep it in order not to break userspace
+ * interfaces
+ */
+struct nand_oobinfo {
+	__u32 useecc;
+	__u32 eccbytes;
+	__u32 oobfree[8][2];
+	__u32 eccpos[32];
+};
+
+struct nand_oobfree {
+	__u32 offset;
+	__u32 length;
+};
+
+#define MTD_MAX_OOBFREE_ENTRIES	8
+#define MTD_MAX_ECCPOS_ENTRIES	64
+/*
+ * OBSOLETE: ECC layout control structure. Exported to user-space via ioctl
+ * ECCGETLAYOUT for backwards compatbility and should not be mistaken as a
+ * complete set of ECC information. The ioctl truncates the larger internal
+ * structure to retain binary compatibility with the static declaration of the
+ * ioctl. Note that the "MTD_MAX_..._ENTRIES" macros represent the max size of
+ * the user struct, not the MAX size of the internal OOB layout representation.
+ */
+struct nand_ecclayout_user {
+	__u32 eccbytes;
+	__u32 eccpos[MTD_MAX_ECCPOS_ENTRIES];
+	__u32 oobavail;
+	struct nand_oobfree oobfree[MTD_MAX_OOBFREE_ENTRIES];
+};
+
+/**
+ * struct mtd_ecc_stats - error correction stats
+ *
+ * @corrected:	number of corrected bits
+ * @failed:	number of uncorrectable errors
+ * @badblocks:	number of bad blocks in this partition
+ * @bbtblocks:	number of blocks reserved for bad block tables
+ */
+struct mtd_ecc_stats {
+	__u32 corrected;
+	__u32 failed;
+	__u32 badblocks;
+	__u32 bbtblocks;
+};
+
+/*
+ * MTD file modes - for read/write access to MTD
+ *
+ * @MTD_FILE_MODE_NORMAL:	OTP disabled, ECC enabled
+ * @MTD_FILE_MODE_OTP_FACTORY:	OTP enabled in factory mode
+ * @MTD_FILE_MODE_OTP_USER:	OTP enabled in user mode
+ * @MTD_FILE_MODE_RAW:		OTP disabled, ECC disabled
+ *
+ * These modes can be set via ioctl(MTDFILEMODE). The mode will be retained
+ * separately for each open file descriptor.
+ *
+ * Note: %MTD_FILE_MODE_RAW provides the same functionality as %MTD_OPS_RAW -
+ * raw access to the flash, without error correction or autoplacement schemes.
+ * Wherever possible, the MTD_OPS_* mode will override the MTD_FILE_MODE_* mode
+ * (e.g., when using ioctl(MEMWRITE) or ioctl(MEMREAD)), but in some cases, the
+ * MTD_FILE_MODE is used out of necessity (e.g., `write()',
+ * ioctl(MEMWRITEOOB64)).
+ */
+enum mtd_file_modes {
+	MTD_FILE_MODE_NORMAL = MTD_OTP_OFF,
+	MTD_FILE_MODE_OTP_FACTORY = MTD_OTP_FACTORY,
+	MTD_FILE_MODE_OTP_USER = MTD_OTP_USER,
+	MTD_FILE_MODE_RAW,
+};
+
+static inline int mtd_type_is_nand_user(const struct mtd_info_user *mtd)
+{
+	return mtd->type == MTD_NANDFLASH || mtd->type == MTD_MLCNANDFLASH;
+}
+
+#endif /* __MTD_ABI_H__ */
diff --git a/include/uapi/mtd/mtd-user.h b/include/uapi/mtd/mtd-user.h
new file mode 100644
index 000000000..7eca1b3f6
--- /dev/null
+++ b/include/uapi/mtd/mtd-user.h
@@ -0,0 +1,33 @@
+/* SPDX-License-Identifier: GPL-2.0+ WITH Linux-syscall-note */
+/*
+ * Copyright © 1999-2010 David Woodhouse <dwmw2@infradead.org>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
+ *
+ */
+
+#ifndef __MTD_USER_H__
+#define __MTD_USER_H__
+
+/* This file is blessed for inclusion by userspace */
+#include <mtd/mtd-abi.h>
+
+typedef struct mtd_info_user mtd_info_t;
+typedef struct erase_info_user erase_info_t;
+typedef struct region_info_user region_info_t;
+typedef struct nand_oobinfo nand_oobinfo_t;
+typedef struct nand_ecclayout_user nand_ecclayout_t;
+
+#endif /* __MTD_USER_H__ */
diff --git a/include/uapi/mtd/nftl-user.h b/include/uapi/mtd/nftl-user.h
new file mode 100644
index 000000000..b23bdc91c
--- /dev/null
+++ b/include/uapi/mtd/nftl-user.h
@@ -0,0 +1,91 @@
+/* SPDX-License-Identifier: GPL-2.0+ WITH Linux-syscall-note */
+/*
+ * Copyright © 1999-2010 David Woodhouse <dwmw2@infradead.org>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
+ *
+ */
+
+#ifndef __MTD_NFTL_USER_H__
+#define __MTD_NFTL_USER_H__
+
+#include <linux/types.h>
+
+/* Block Control Information */
+
+struct nftl_bci {
+	unsigned char ECCSig[6];
+	__u8 Status;
+	__u8 Status1;
+}__attribute__((packed));
+
+/* Unit Control Information */
+
+struct nftl_uci0 {
+	__u16 VirtUnitNum;
+	__u16 ReplUnitNum;
+	__u16 SpareVirtUnitNum;
+	__u16 SpareReplUnitNum;
+} __attribute__((packed));
+
+struct nftl_uci1 {
+	__u32 WearInfo;
+	__u16 EraseMark;
+	__u16 EraseMark1;
+} __attribute__((packed));
+
+struct nftl_uci2 {
+        __u16 FoldMark;
+        __u16 FoldMark1;
+	__u32 unused;
+} __attribute__((packed));
+
+union nftl_uci {
+	struct nftl_uci0 a;
+	struct nftl_uci1 b;
+	struct nftl_uci2 c;
+};
+
+struct nftl_oob {
+	struct nftl_bci b;
+	union nftl_uci u;
+};
+
+/* NFTL Media Header */
+
+struct NFTLMediaHeader {
+	char DataOrgID[6];
+	__u16 NumEraseUnits;
+	__u16 FirstPhysicalEUN;
+	__u32 FormattedSize;
+	unsigned char UnitSizeFactor;
+} __attribute__((packed));
+
+#define MAX_ERASE_ZONES (8192 - 512)
+
+#define ERASE_MARK 0x3c69
+#define SECTOR_FREE 0xff
+#define SECTOR_USED 0x55
+#define SECTOR_IGNORE 0x11
+#define SECTOR_DELETED 0x00
+
+#define FOLD_MARK_IN_PROGRESS 0x5555
+
+#define ZONE_GOOD 0xff
+#define ZONE_BAD_ORIGINAL 0
+#define ZONE_BAD_MARKED 7
+
+
+#endif /* __MTD_NFTL_USER_H__ */
diff --git a/include/uapi/mtd/ubi-user.h b/include/uapi/mtd/ubi-user.h
new file mode 100644
index 000000000..3538e11b5
--- /dev/null
+++ b/include/uapi/mtd/ubi-user.h
@@ -0,0 +1,514 @@
+/* SPDX-License-Identifier: GPL-2.0+ WITH Linux-syscall-note */
+/*
+ * Copyright © International Business Machines Corp., 2006
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See
+ * the GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ *
+ * Author: Artem Bityutskiy (Битюцкий Артём)
+ */
+
+#ifndef __UBI_USER_H__
+#define __UBI_USER_H__
+
+#include <linux/types.h>
+
+/*
+ * UBI device creation (the same as MTD device attachment)
+ * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+ *
+ * MTD devices may be attached using %UBI_IOCATT ioctl command of the UBI
+ * control device. The caller has to properly fill and pass
+ * &struct ubi_attach_req object - UBI will attach the MTD device specified in
+ * the request and return the newly created UBI device number as the ioctl
+ * return value.
+ *
+ * UBI device deletion (the same as MTD device detachment)
+ * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+ *
+ * An UBI device maybe deleted with %UBI_IOCDET ioctl command of the UBI
+ * control device.
+ *
+ * UBI volume creation
+ * ~~~~~~~~~~~~~~~~~~~
+ *
+ * UBI volumes are created via the %UBI_IOCMKVOL ioctl command of UBI character
+ * device. A &struct ubi_mkvol_req object has to be properly filled and a
+ * pointer to it has to be passed to the ioctl.
+ *
+ * UBI volume deletion
+ * ~~~~~~~~~~~~~~~~~~~
+ *
+ * To delete a volume, the %UBI_IOCRMVOL ioctl command of the UBI character
+ * device should be used. A pointer to the 32-bit volume ID hast to be passed
+ * to the ioctl.
+ *
+ * UBI volume re-size
+ * ~~~~~~~~~~~~~~~~~~
+ *
+ * To re-size a volume, the %UBI_IOCRSVOL ioctl command of the UBI character
+ * device should be used. A &struct ubi_rsvol_req object has to be properly
+ * filled and a pointer to it has to be passed to the ioctl.
+ *
+ * UBI volumes re-name
+ * ~~~~~~~~~~~~~~~~~~~
+ *
+ * To re-name several volumes atomically at one go, the %UBI_IOCRNVOL command
+ * of the UBI character device should be used. A &struct ubi_rnvol_req object
+ * has to be properly filled and a pointer to it has to be passed to the ioctl.
+ *
+ * UBI volume update
+ * ~~~~~~~~~~~~~~~~~
+ *
+ * Volume update should be done via the %UBI_IOCVOLUP ioctl command of the
+ * corresponding UBI volume character device. A pointer to a 64-bit update
+ * size should be passed to the ioctl. After this, UBI expects user to write
+ * this number of bytes to the volume character device. The update is finished
+ * when the claimed number of bytes is passed. So, the volume update sequence
+ * is something like:
+ *
+ * fd = open("/dev/my_volume");
+ * ioctl(fd, UBI_IOCVOLUP, &image_size);
+ * write(fd, buf, image_size);
+ * close(fd);
+ *
+ * Logical eraseblock erase
+ * ~~~~~~~~~~~~~~~~~~~~~~~~
+ *
+ * To erase a logical eraseblock, the %UBI_IOCEBER ioctl command of the
+ * corresponding UBI volume character device should be used. This command
+ * unmaps the requested logical eraseblock, makes sure the corresponding
+ * physical eraseblock is successfully erased, and returns.
+ *
+ * Atomic logical eraseblock change
+ * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+ *
+ * Atomic logical eraseblock change operation is called using the %UBI_IOCEBCH
+ * ioctl command of the corresponding UBI volume character device. A pointer to
+ * a &struct ubi_leb_change_req object has to be passed to the ioctl. Then the
+ * user is expected to write the requested amount of bytes (similarly to what
+ * should be done in case of the "volume update" ioctl).
+ *
+ * Logical eraseblock map
+ * ~~~~~~~~~~~~~~~~~~~~~
+ *
+ * To map a logical eraseblock to a physical eraseblock, the %UBI_IOCEBMAP
+ * ioctl command should be used. A pointer to a &struct ubi_map_req object is
+ * expected to be passed. The ioctl maps the requested logical eraseblock to
+ * a physical eraseblock and returns.  Only non-mapped logical eraseblocks can
+ * be mapped. If the logical eraseblock specified in the request is already
+ * mapped to a physical eraseblock, the ioctl fails and returns error.
+ *
+ * Logical eraseblock unmap
+ * ~~~~~~~~~~~~~~~~~~~~~~~~
+ *
+ * To unmap a logical eraseblock to a physical eraseblock, the %UBI_IOCEBUNMAP
+ * ioctl command should be used. The ioctl unmaps the logical eraseblocks,
+ * schedules corresponding physical eraseblock for erasure, and returns. Unlike
+ * the "LEB erase" command, it does not wait for the physical eraseblock being
+ * erased. Note, the side effect of this is that if an unclean reboot happens
+ * after the unmap ioctl returns, you may find the LEB mapped again to the same
+ * physical eraseblock after the UBI is run again.
+ *
+ * Check if logical eraseblock is mapped
+ * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+ *
+ * To check if a logical eraseblock is mapped to a physical eraseblock, the
+ * %UBI_IOCEBISMAP ioctl command should be used. It returns %0 if the LEB is
+ * not mapped, and %1 if it is mapped.
+ *
+ * Set an UBI volume property
+ * ~~~~~~~~~~~~~~~~~~~~~~~~~
+ *
+ * To set an UBI volume property the %UBI_IOCSETPROP ioctl command should be
+ * used. A pointer to a &struct ubi_set_vol_prop_req object is expected to be
+ * passed. The object describes which property should be set, and to which value
+ * it should be set.
+ *
+ * Block devices on UBI volumes
+ * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+ *
+ * To create a R/O block device on top of an UBI volume the %UBI_IOCVOLCRBLK
+ * should be used. A pointer to a &struct ubi_blkcreate_req object is expected
+ * to be passed, which is not used and reserved for future usage.
+ *
+ * Conversely, to remove a block device the %UBI_IOCVOLRMBLK should be used,
+ * which takes no arguments.
+ */
+
+/*
+ * When a new UBI volume or UBI device is created, users may either specify the
+ * volume/device number they want to create or to let UBI automatically assign
+ * the number using these constants.
+ */
+#define UBI_VOL_NUM_AUTO (-1)
+#define UBI_DEV_NUM_AUTO (-1)
+
+/* Maximum volume name length */
+#define UBI_MAX_VOLUME_NAME 127
+
+/* ioctl commands of UBI character devices */
+
+#define UBI_IOC_MAGIC 'o'
+
+/* Create an UBI volume */
+#define UBI_IOCMKVOL _IOW(UBI_IOC_MAGIC, 0, struct ubi_mkvol_req)
+/* Remove an UBI volume */
+#define UBI_IOCRMVOL _IOW(UBI_IOC_MAGIC, 1, __s32)
+/* Re-size an UBI volume */
+#define UBI_IOCRSVOL _IOW(UBI_IOC_MAGIC, 2, struct ubi_rsvol_req)
+/* Re-name volumes */
+#define UBI_IOCRNVOL _IOW(UBI_IOC_MAGIC, 3, struct ubi_rnvol_req)
+
+/* Read the specified PEB and scrub it if there are bitflips */
+#define UBI_IOCRPEB _IOW(UBI_IOC_MAGIC, 4, __s32)
+/* Force scrubbing on the specified PEB */
+#define UBI_IOCSPEB _IOW(UBI_IOC_MAGIC, 5, __s32)
+/* Read detailed device erase counter information */
+#define UBI_IOCECNFO _IOWR(UBI_IOC_MAGIC, 6, struct ubi_ecinfo_req)
+
+/* ioctl commands of the UBI control character device */
+
+#define UBI_CTRL_IOC_MAGIC 'o'
+
+/* Attach an MTD device */
+#define UBI_IOCATT _IOW(UBI_CTRL_IOC_MAGIC, 64, struct ubi_attach_req)
+/* Detach an MTD device */
+#define UBI_IOCDET _IOW(UBI_CTRL_IOC_MAGIC, 65, __s32)
+
+/* ioctl commands of UBI volume character devices */
+
+#define UBI_VOL_IOC_MAGIC 'O'
+
+/* Start UBI volume update
+ * Note: This actually takes a pointer (__s64*), but we can't change
+ *       that without breaking the ABI on 32bit systems
+ */
+#define UBI_IOCVOLUP _IOW(UBI_VOL_IOC_MAGIC, 0, __s64)
+/* LEB erasure command, used for debugging, disabled by default */
+#define UBI_IOCEBER _IOW(UBI_VOL_IOC_MAGIC, 1, __s32)
+/* Atomic LEB change command */
+#define UBI_IOCEBCH _IOW(UBI_VOL_IOC_MAGIC, 2, __s32)
+/* Map LEB command */
+#define UBI_IOCEBMAP _IOW(UBI_VOL_IOC_MAGIC, 3, struct ubi_map_req)
+/* Unmap LEB command */
+#define UBI_IOCEBUNMAP _IOW(UBI_VOL_IOC_MAGIC, 4, __s32)
+/* Check if LEB is mapped command */
+#define UBI_IOCEBISMAP _IOR(UBI_VOL_IOC_MAGIC, 5, __s32)
+/* Set an UBI volume property */
+#define UBI_IOCSETVOLPROP _IOW(UBI_VOL_IOC_MAGIC, 6, \
+			       struct ubi_set_vol_prop_req)
+/* Create a R/O block device on top of an UBI volume */
+#define UBI_IOCVOLCRBLK _IOW(UBI_VOL_IOC_MAGIC, 7, struct ubi_blkcreate_req)
+/* Remove the R/O block device */
+#define UBI_IOCVOLRMBLK _IO(UBI_VOL_IOC_MAGIC, 8)
+
+/* Maximum MTD device name length supported by UBI */
+#define MAX_UBI_MTD_NAME_LEN 127
+
+/* Maximum amount of UBI volumes that can be re-named at one go */
+#define UBI_MAX_RNVOL 32
+
+/*
+ * UBI volume type constants.
+ *
+ * @UBI_DYNAMIC_VOLUME: dynamic volume
+ * @UBI_STATIC_VOLUME:  static volume
+ */
+enum {
+	UBI_DYNAMIC_VOLUME = 3,
+	UBI_STATIC_VOLUME  = 4,
+};
+
+/*
+ * UBI set volume property ioctl constants.
+ *
+ * @UBI_VOL_PROP_DIRECT_WRITE: allow (any non-zero value) or disallow (value 0)
+ *                             user to directly write and erase individual
+ *                             eraseblocks on dynamic volumes
+ */
+enum {
+	UBI_VOL_PROP_DIRECT_WRITE = 1,
+};
+
+/**
+ * struct ubi_attach_req - attach MTD device request.
+ * @ubi_num: UBI device number to create
+ * @mtd_num: MTD device number to attach
+ * @vid_hdr_offset: VID header offset (use defaults if %0)
+ * @max_beb_per1024: maximum expected number of bad PEB per 1024 PEBs
+ * @padding: reserved for future, not used, has to be zeroed
+ * @disable_fm: whether disable fastmap
+ * @need_resv_pool: whether reserve free pebs for filling pool/wl_pool
+ *
+ * This data structure is used to specify MTD device UBI has to attach and the
+ * parameters it has to use. The number which should be assigned to the new UBI
+ * device is passed in @ubi_num. UBI may automatically assign the number if
+ * @UBI_DEV_NUM_AUTO is passed. In this case, the device number is returned in
+ * @ubi_num.
+ *
+ * Most applications should pass %0 in @vid_hdr_offset to make UBI use default
+ * offset of the VID header within physical eraseblocks. The default offset is
+ * the next min. I/O unit after the EC header. For example, it will be offset
+ * 512 in case of a 512 bytes page NAND flash with no sub-page support. Or
+ * it will be 512 in case of a 2KiB page NAND flash with 4 512-byte sub-pages.
+ *
+ * But in rare cases, if this optimizes things, the VID header may be placed to
+ * a different offset. For example, the boot-loader might do things faster if
+ * the VID header sits at the end of the first 2KiB NAND page with 4 sub-pages.
+ * As the boot-loader would not normally need to read EC headers (unless it
+ * needs UBI in RW mode), it might be faster to calculate ECC. This is weird
+ * example, but it real-life example. So, in this example, @vid_hdr_offer would
+ * be 2KiB-64 bytes = 1984. Note, that this position is not even 512-bytes
+ * aligned, which is OK, as UBI is clever enough to realize this is 4th
+ * sub-page of the first page and add needed padding.
+ *
+ * The @max_beb_per1024 is the maximum amount of bad PEBs UBI expects on the
+ * UBI device per 1024 eraseblocks.  This value is often given in an other form
+ * in the NAND datasheet (min NVB i.e. minimal number of valid blocks). The
+ * maximum expected bad eraseblocks per 1024 is then:
+ *    1024 * (1 - MinNVB / MaxNVB)
+ * Which gives 20 for most NAND devices.  This limit is used in order to derive
+ * amount of eraseblock UBI reserves for handling new bad blocks. If the device
+ * has more bad eraseblocks than this limit, UBI does not reserve any physical
+ * eraseblocks for new bad eraseblocks, but attempts to use available
+ * eraseblocks (if any). The accepted range is 0-768. If 0 is given, the
+ * default kernel value of %CONFIG_MTD_UBI_BEB_LIMIT will be used.
+ *
+ * If @disable_fm is not zero, 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.
+ *
+ * The @wl_threshold defines the maximum difference between the highest and the
+ * lowest erase counter value of eraseblocks of this UBI device. 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. If @wl_threshold is zero, the default kernel value of
+ * %CONFIG_MTD_UBI_WL_THRESHOLD is used. The accepted range is 2-65536.
+ */
+struct ubi_attach_req {
+	__s32 ubi_num;
+	__s32 mtd_num;
+	__s32 vid_hdr_offset;
+	__s16 max_beb_per1024;
+	__s8 disable_fm;
+	__s8 need_resv_pool;
+	__s32 wl_threshold;
+	__s8 padding[4];
+};
+
+/*
+ * UBI volume flags.
+ *
+ * @UBI_VOL_SKIP_CRC_CHECK_FLG: skip the CRC check done on a static volume at
+ *				open time. Only valid for static volumes and
+ *				should only be used if the volume user has a
+ *				way to verify data integrity
+ */
+enum {
+	UBI_VOL_SKIP_CRC_CHECK_FLG = 0x1,
+};
+
+#define UBI_VOL_VALID_FLGS	(UBI_VOL_SKIP_CRC_CHECK_FLG)
+
+/**
+ * struct ubi_mkvol_req - volume description data structure used in
+ *                        volume creation requests.
+ * @vol_id: volume number
+ * @alignment: volume alignment
+ * @bytes: volume size in bytes
+ * @vol_type: volume type (%UBI_DYNAMIC_VOLUME or %UBI_STATIC_VOLUME)
+ * @flags: volume flags (%UBI_VOL_SKIP_CRC_CHECK_FLG)
+ * @name_len: volume name length
+ * @padding2: reserved for future, not used, has to be zeroed
+ * @name: volume name
+ *
+ * This structure is used by user-space programs when creating new volumes. The
+ * @used_bytes field is only necessary when creating static volumes.
+ *
+ * The @alignment field specifies the required alignment of the volume logical
+ * eraseblock. This means, that the size of logical eraseblocks will be aligned
+ * to this number, i.e.,
+ *	(UBI device logical eraseblock size) mod (@alignment) = 0.
+ *
+ * To put it differently, the logical eraseblock of this volume may be slightly
+ * shortened in order to make it properly aligned. The alignment has to be
+ * multiple of the flash minimal input/output unit, or %1 to utilize the entire
+ * available space of logical eraseblocks.
+ *
+ * The @alignment field may be useful, for example, when one wants to maintain
+ * a block device on top of an UBI volume. In this case, it is desirable to fit
+ * an integer number of blocks in logical eraseblocks of this UBI volume. With
+ * alignment it is possible to update this volume using plane UBI volume image
+ * BLOBs, without caring about how to properly align them.
+ */
+struct ubi_mkvol_req {
+	__s32 vol_id;
+	__s32 alignment;
+	__s64 bytes;
+	__s8 vol_type;
+	__u8 flags;
+	__s16 name_len;
+	__s8 padding2[4];
+	char name[UBI_MAX_VOLUME_NAME + 1];
+} __packed;
+
+/**
+ * struct ubi_rsvol_req - a data structure used in volume re-size requests.
+ * @vol_id: ID of the volume to re-size
+ * @bytes: new size of the volume in bytes
+ *
+ * Re-sizing is possible for both dynamic and static volumes. But while dynamic
+ * volumes may be re-sized arbitrarily, static volumes cannot be made to be
+ * smaller than the number of bytes they bear. To arbitrarily shrink a static
+ * volume, it must be wiped out first (by means of volume update operation with
+ * zero number of bytes).
+ */
+struct ubi_rsvol_req {
+	__s64 bytes;
+	__s32 vol_id;
+} __packed;
+
+/**
+ * struct ubi_rnvol_req - volumes re-name request.
+ * @count: count of volumes to re-name
+ * @padding1:  reserved for future, not used, has to be zeroed
+ * @vol_id: ID of the volume to re-name
+ * @name_len: name length
+ * @padding2:  reserved for future, not used, has to be zeroed
+ * @name: new volume name
+ *
+ * UBI allows to re-name up to %32 volumes at one go. The count of volumes to
+ * re-name is specified in the @count field. The ID of the volumes to re-name
+ * and the new names are specified in the @vol_id and @name fields.
+ *
+ * The UBI volume re-name operation is atomic, which means that should power cut
+ * happen, the volumes will have either old name or new name. So the possible
+ * use-cases of this command is atomic upgrade. Indeed, to upgrade, say, volumes
+ * A and B one may create temporary volumes %A1 and %B1 with the new contents,
+ * then atomically re-name A1->A and B1->B, in which case old %A and %B will
+ * be removed.
+ *
+ * If it is not desirable to remove old A and B, the re-name request has to
+ * contain 4 entries: A1->A, A->A1, B1->B, B->B1, in which case old A1 and B1
+ * become A and B, and old A and B will become A1 and B1.
+ *
+ * It is also OK to request: A1->A, A1->X, B1->B, B->Y, in which case old A1
+ * and B1 become A and B, and old A and B become X and Y.
+ *
+ * In other words, in case of re-naming into an existing volume name, the
+ * existing volume is removed, unless it is re-named as well at the same
+ * re-name request.
+ */
+struct ubi_rnvol_req {
+	__s32 count;
+	__s8 padding1[12];
+	struct {
+		__s32 vol_id;
+		__s16 name_len;
+		__s8  padding2[2];
+		char    name[UBI_MAX_VOLUME_NAME + 1];
+	} ents[UBI_MAX_RNVOL];
+} __packed;
+
+/**
+ * struct ubi_ecinfo_req - a data structure used for requesting and receiving
+ * erase block counter information from a UBI device.
+ *
+ * @start: index of first physical erase block to read (in)
+ * @length: number of erase counters to read (in)
+ * @read_length: number of erase counters that was actually read (out)
+ * @padding: reserved for future, not used, has to be zeroed
+ * @erase_counters: array of erase counter values (out)
+ *
+ * This structure is used to retrieve erase counter information for a specified
+ * range of PEBs on a UBI device.
+ * Erase counters are read from @start and attempts to read @length number of
+ * erase counters.
+ * The retrieved values are stored in the @erase_counters array. It is the
+ * responsibility of the caller to allocate enough memory for storing @length
+ * elements in the @erase_counters array.
+ * If a block is bad or if the erase counter is unknown the corresponding value
+ * in the array will be set to -1.
+ * The @read_length field will indicate the number of erase counters actually
+ * read. Typically @read_length will be limited due to memory or the number of
+ * PEBs on the UBI device.
+ */
+struct ubi_ecinfo_req {
+	__s32 start;
+	__s32 length;
+	__s32 read_length;
+	__s8  padding[16];
+	__s32 erase_counters[];
+}  __packed;
+
+/**
+ * struct ubi_leb_change_req - a data structure used in atomic LEB change
+ *                             requests.
+ * @lnum: logical eraseblock number to change
+ * @bytes: how many bytes will be written to the logical eraseblock
+ * @dtype: pass "3" for better compatibility with old kernels
+ * @padding: reserved for future, not used, has to be zeroed
+ *
+ * The @dtype field used to inform UBI about what kind of data will be written
+ * to the LEB: long term (value 1), short term (value 2), unknown (value 3).
+ * UBI tried to pick a PEB with lower erase counter for short term data and a
+ * PEB with higher erase counter for long term data. But this was not really
+ * used because users usually do not know this and could easily mislead UBI. We
+ * removed this feature in May 2012. UBI currently just ignores the @dtype
+ * field. But for better compatibility with older kernels it is recommended to
+ * set @dtype to 3 (unknown).
+ */
+struct ubi_leb_change_req {
+	__s32 lnum;
+	__s32 bytes;
+	__s8  dtype; /* obsolete, do not use! */
+	__s8  padding[7];
+} __packed;
+
+/**
+ * struct ubi_map_req - a data structure used in map LEB requests.
+ * @dtype: pass "3" for better compatibility with old kernels
+ * @lnum: logical eraseblock number to unmap
+ * @padding: reserved for future, not used, has to be zeroed
+ */
+struct ubi_map_req {
+	__s32 lnum;
+	__s8  dtype; /* obsolete, do not use! */
+	__s8  padding[3];
+} __packed;
+
+
+/**
+ * struct ubi_set_vol_prop_req - a data structure used to set an UBI volume
+ *                               property.
+ * @property: property to set (%UBI_VOL_PROP_DIRECT_WRITE)
+ * @padding: reserved for future, not used, has to be zeroed
+ * @value: value to set
+ */
+struct ubi_set_vol_prop_req {
+	__u8  property;
+	__u8  padding[7];
+	__u64 value;
+}  __packed;
+
+/**
+ * struct ubi_blkcreate_req - a data structure used in block creation requests.
+ * @padding: reserved for future, not used, has to be zeroed
+ */
+struct ubi_blkcreate_req {
+	__s8  padding[128];
+}  __packed;
+
+#endif /* __UBI_USER_H__ */