From a5f7a5bb3b7f28ba7e4fa246775b29a0e5537255 Mon Sep 17 00:00:00 2001 From: Jinke Han Date: Tue, 8 Sep 2026 15:37:42 +0800 Subject: x86/kprobes: Fix crash when probing CS CALL instructions When using eBPF to probe CS CALL instructions within a function, a crash can be triggered. The eBPF tool probes offset 257 of the __hrtimer_run_queues() function: <__hrtimer_run_queues+249>: nopl 0x0(%rax,%rax,1) <__hrtimer_run_queues+254>: mov %r14,%rdi <__hrtimer_run_queues+257>: cs call <__x86_indirect_thunk_r12> <__hrtimer_run_queues+263>: mov %eax,%r12d <__hrtimer_run_queues+266>: xchg %ax,%ax <__hrtimer_run_queues+268>: mov %r13,%rdi Which triggers this crash: BUG: unable to handle page fault for address: 00000000000f41c9 #PF: supervisor write access in kernel mode #PF: error_code(0x0002) - not-present page PGD 0 P4D 0 Oops: 0002 [#1] SMP NOPTI CPU: 1 PID: 0 Comm: swapper/1 Kdump: loaded Tainted: P RIP: 0010:__hrtimer_run_queues+0x106/0x230 Note that __hrtimer_run_queues+0x106 is __hrtimer_run_queues+262, which is at the 6th byte of the above CS CALL instruction. Since the CS CALL instruction occupies 6 bytes, the exception occurred in the middle of that call instruction. The root cause is that when using eBPF tools to probe in the middle of a function, a kprobe with INT3 is used as the underlying implementation. During single-step emulation of the original CALL instruction, int3_emulate_call() assumes that the probed CALL instruction is 5 bytes long. However, the actual CS-prefixed CALL instruction occupies 6 bytes, so it constructs an incorrect exception return address. When the CPU returns from the kprobe handler, the next instruction to be executed is at the address of the last byte of that CS CALL instruction. Coincidentally, starting from that address, the CPU fetches and decodes a completely different instruction, which ultimately triggers a kernel crash. Fix the issue by using the actual instruction length obtained from the instruction decoder when constructing the exception return address, rather than relying on the hardcoded CALL_INSN_SIZE macro. [ mingo: Refined the changelog ] Fixes: 6256e668b7af ("x86/kprobes: Use int3 instead of debug trap for single-step") Suggested-by: Masami Hiramatsu (Google) Signed-off-by: Jinke Han Signed-off-by: Ingo Molnar Reviewed-by: Masami Hiramatsu (Google) Acked-by: Yafang Shao Acked-by: Borislav Petkov Cc: Peter Zijlstra Link: https://patch.msgid.link/20260908073742.GA10517@didi-ThinkCentre-M920t-N000 --- fs/udf/misc.c | 289 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 289 insertions(+) create mode 100644 fs/udf/misc.c (limited to 'fs/udf/misc.c') diff --git a/fs/udf/misc.c b/fs/udf/misc.c new file mode 100644 index 000000000..6928e378f --- /dev/null +++ b/fs/udf/misc.c @@ -0,0 +1,289 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * misc.c + * + * PURPOSE + * Miscellaneous routines for the OSTA-UDF(tm) filesystem. + * + * COPYRIGHT + * (C) 1998 Dave Boynton + * (C) 1998-2004 Ben Fennema + * (C) 1999-2000 Stelias Computing Inc + * + * HISTORY + * + * 04/19/99 blf partial support for reading/writing specific EA's + */ + +#include "udfdecl.h" + +#include +#include +#include + +#include "udf_i.h" +#include "udf_sb.h" + +struct genericFormat *udf_add_extendedattr(struct inode *inode, uint32_t size, + uint32_t type, uint8_t loc) +{ + uint8_t *ea = NULL, *ad = NULL; + int offset; + uint16_t crclen; + struct udf_inode_info *iinfo = UDF_I(inode); + + ea = iinfo->i_data; + if (iinfo->i_lenEAttr) { + ad = iinfo->i_data + iinfo->i_lenEAttr; + } else { + ad = ea; + size += sizeof(struct extendedAttrHeaderDesc); + } + + offset = inode->i_sb->s_blocksize - udf_file_entry_alloc_offset(inode) - + iinfo->i_lenAlloc; + + /* TODO - Check for FreeEASpace */ + + if (loc & 0x01 && offset >= size) { + struct extendedAttrHeaderDesc *eahd; + eahd = (struct extendedAttrHeaderDesc *)ea; + + if (iinfo->i_lenAlloc) + memmove(&ad[size], ad, iinfo->i_lenAlloc); + + if (iinfo->i_lenEAttr) { + /* check checksum/crc */ + if (eahd->descTag.tagIdent != + cpu_to_le16(TAG_IDENT_EAHD) || + le32_to_cpu(eahd->descTag.tagLocation) != + iinfo->i_location.logicalBlockNum) + return NULL; + } else { + struct udf_sb_info *sbi = UDF_SB(inode->i_sb); + + size -= sizeof(struct extendedAttrHeaderDesc); + iinfo->i_lenEAttr += + sizeof(struct extendedAttrHeaderDesc); + eahd->descTag.tagIdent = cpu_to_le16(TAG_IDENT_EAHD); + if (sbi->s_udfrev >= 0x0200) + eahd->descTag.descVersion = cpu_to_le16(3); + else + eahd->descTag.descVersion = cpu_to_le16(2); + eahd->descTag.tagSerialNum = + cpu_to_le16(sbi->s_serial_number); + eahd->descTag.tagLocation = cpu_to_le32( + iinfo->i_location.logicalBlockNum); + eahd->impAttrLocation = cpu_to_le32(0xFFFFFFFF); + eahd->appAttrLocation = cpu_to_le32(0xFFFFFFFF); + } + + offset = iinfo->i_lenEAttr; + if (type < 2048) { + if (le32_to_cpu(eahd->appAttrLocation) < + iinfo->i_lenEAttr) { + uint32_t aal = + le32_to_cpu(eahd->appAttrLocation); + memmove(&ea[offset - aal + size], + &ea[aal], offset - aal); + offset -= aal; + eahd->appAttrLocation = + cpu_to_le32(aal + size); + } + if (le32_to_cpu(eahd->impAttrLocation) < + iinfo->i_lenEAttr) { + uint32_t ial = + le32_to_cpu(eahd->impAttrLocation); + memmove(&ea[offset - ial + size], + &ea[ial], offset - ial); + offset -= ial; + eahd->impAttrLocation = + cpu_to_le32(ial + size); + } + } else if (type < 65536) { + if (le32_to_cpu(eahd->appAttrLocation) < + iinfo->i_lenEAttr) { + uint32_t aal = + le32_to_cpu(eahd->appAttrLocation); + memmove(&ea[offset - aal + size], + &ea[aal], offset - aal); + offset -= aal; + eahd->appAttrLocation = + cpu_to_le32(aal + size); + } + } + /* rewrite CRC + checksum of eahd */ + crclen = sizeof(struct extendedAttrHeaderDesc) - sizeof(struct tag); + eahd->descTag.descCRCLength = cpu_to_le16(crclen); + eahd->descTag.descCRC = cpu_to_le16(crc_itu_t(0, (char *)eahd + + sizeof(struct tag), crclen)); + eahd->descTag.tagChecksum = udf_tag_checksum(&eahd->descTag); + iinfo->i_lenEAttr += size; + return (struct genericFormat *)&ea[offset]; + } + + return NULL; +} + +struct genericFormat *udf_get_extendedattr(struct inode *inode, uint32_t type, + uint8_t subtype) +{ + struct genericFormat *gaf; + uint8_t *ea = NULL; + uint32_t offset; + struct udf_inode_info *iinfo = UDF_I(inode); + + ea = iinfo->i_data; + + if (iinfo->i_lenEAttr) { + struct extendedAttrHeaderDesc *eahd; + eahd = (struct extendedAttrHeaderDesc *)ea; + + /* check checksum/crc */ + if (eahd->descTag.tagIdent != + cpu_to_le16(TAG_IDENT_EAHD) || + le32_to_cpu(eahd->descTag.tagLocation) != + iinfo->i_location.logicalBlockNum) + return NULL; + + if (type < 2048) + offset = sizeof(struct extendedAttrHeaderDesc); + else if (type < 65536) + offset = le32_to_cpu(eahd->impAttrLocation); + else + offset = le32_to_cpu(eahd->appAttrLocation); + + while (offset + sizeof(*gaf) < iinfo->i_lenEAttr) { + uint32_t attrLength; + + gaf = (struct genericFormat *)&ea[offset]; + attrLength = le32_to_cpu(gaf->attrLength); + + /* Detect undersized elements and buffer overflows */ + if ((attrLength < sizeof(*gaf)) || + (attrLength > (iinfo->i_lenEAttr - offset))) + break; + + if (le32_to_cpu(gaf->attrType) == type && + gaf->attrSubtype == subtype) + return gaf; + else + offset += attrLength; + } + } + + return NULL; +} + +/* + * udf_read_tagged + * + * PURPOSE + * Read the first block of a tagged descriptor. + * + * HISTORY + * July 1, 1997 - Andrew E. Mileski + * Written, tested, and released. + */ +struct buffer_head *udf_read_tagged(struct super_block *sb, uint32_t block, + uint32_t location, uint16_t *ident) +{ + struct tag *tag_p; + struct buffer_head *bh = NULL; + u8 checksum; + + /* Read the block */ + if (block == 0xFFFFFFFF) + return NULL; + + bh = sb_bread(sb, block); + if (!bh) { + udf_err(sb, "read failed, block=%u, location=%u\n", + block, location); + return NULL; + } + + tag_p = (struct tag *)(bh->b_data); + + *ident = le16_to_cpu(tag_p->tagIdent); + + if (location != le32_to_cpu(tag_p->tagLocation)) { + udf_debug("location mismatch block %u, tag %u != %u\n", + block, le32_to_cpu(tag_p->tagLocation), location); + goto error_out; + } + + /* Verify the tag checksum */ + checksum = udf_tag_checksum(tag_p); + if (checksum != tag_p->tagChecksum) { + udf_err(sb, "tag checksum failed, block %u: 0x%02x != 0x%02x\n", + block, checksum, tag_p->tagChecksum); + goto error_out; + } + + /* Verify the tag version */ + if (tag_p->descVersion != cpu_to_le16(0x0002U) && + tag_p->descVersion != cpu_to_le16(0x0003U)) { + udf_err(sb, "tag version 0x%04x != 0x0002 || 0x0003, block %u\n", + le16_to_cpu(tag_p->descVersion), block); + goto error_out; + } + + /* Verify the descriptor CRC */ + if (le16_to_cpu(tag_p->descCRCLength) + sizeof(struct tag) > sb->s_blocksize) { + udf_err(sb, "block %u: CRC length %u exceeds block size\n", + block, le16_to_cpu(tag_p->descCRCLength)); + goto error_out; + } + if (le16_to_cpu(tag_p->descCRC) == crc_itu_t(0, + bh->b_data + sizeof(struct tag), + le16_to_cpu(tag_p->descCRCLength))) + return bh; + + udf_debug("Crc failure block %u: crc = %u, crclen = %u\n", block, + le16_to_cpu(tag_p->descCRC), + le16_to_cpu(tag_p->descCRCLength)); +error_out: + brelse(bh); + return NULL; +} + +struct buffer_head *udf_read_ptagged(struct super_block *sb, + struct kernel_lb_addr *loc, + uint32_t offset, uint16_t *ident) +{ + return udf_read_tagged(sb, udf_get_lb_pblock(sb, loc, offset), + loc->logicalBlockNum + offset, ident); +} + +void udf_update_tag(char *data, int length) +{ + struct tag *tptr = (struct tag *)data; + length -= sizeof(struct tag); + + tptr->descCRCLength = cpu_to_le16(length); + tptr->descCRC = cpu_to_le16(crc_itu_t(0, data + sizeof(struct tag), length)); + tptr->tagChecksum = udf_tag_checksum(tptr); +} + +void udf_new_tag(char *data, uint16_t ident, uint16_t version, uint16_t snum, + uint32_t loc, int length) +{ + struct tag *tptr = (struct tag *)data; + tptr->tagIdent = cpu_to_le16(ident); + tptr->descVersion = cpu_to_le16(version); + tptr->tagSerialNum = cpu_to_le16(snum); + tptr->tagLocation = cpu_to_le32(loc); + udf_update_tag(data, length); +} + +u8 udf_tag_checksum(const struct tag *t) +{ + u8 *data = (u8 *)t; + u8 checksum = 0; + int i; + for (i = 0; i < sizeof(struct tag); ++i) + if (i != 4) /* position of checksum */ + checksum += data[i]; + return checksum; +} -- cgit v1.3.1