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/romfs/storage.c | 287 +++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 287 insertions(+) create mode 100644 fs/romfs/storage.c (limited to 'fs/romfs/storage.c') diff --git a/fs/romfs/storage.c b/fs/romfs/storage.c new file mode 100644 index 000000000..b57b3ffcb --- /dev/null +++ b/fs/romfs/storage.c @@ -0,0 +1,287 @@ +// SPDX-License-Identifier: GPL-2.0-or-later +/* RomFS storage access routines + * + * Copyright © 2007 Red Hat, Inc. All Rights Reserved. + * Written by David Howells (dhowells@redhat.com) + */ + +#include +#include +#include +#include "internal.h" + +#if !defined(CONFIG_ROMFS_ON_MTD) && !defined(CONFIG_ROMFS_ON_BLOCK) +#error no ROMFS backing store interface configured +#endif + +#ifdef CONFIG_ROMFS_ON_MTD +#define ROMFS_MTD_READ(sb, ...) mtd_read((sb)->s_mtd, ##__VA_ARGS__) + +/* + * read data from an romfs image on an MTD device + */ +static int romfs_mtd_read(struct super_block *sb, unsigned long pos, + void *buf, size_t buflen) +{ + size_t rlen; + int ret; + + ret = ROMFS_MTD_READ(sb, pos, buflen, &rlen, buf); + return (ret < 0 || rlen != buflen) ? -EIO : 0; +} + +/* + * determine the length of a string in a romfs image on an MTD device + */ +static ssize_t romfs_mtd_strnlen(struct super_block *sb, + unsigned long pos, size_t maxlen) +{ + ssize_t n = 0; + size_t segment; + u_char buf[16], *p; + size_t len; + int ret; + + /* scan the string up to 16 bytes at a time */ + while (maxlen > 0) { + segment = min_t(size_t, maxlen, 16); + ret = ROMFS_MTD_READ(sb, pos, segment, &len, buf); + if (ret < 0) + return ret; + p = memchr(buf, 0, len); + if (p) + return n + (p - buf); + maxlen -= len; + pos += len; + n += len; + } + + return n; +} + +/* + * compare a string to one in a romfs image on MTD + * - return 1 if matched, 0 if differ, -ve if error + */ +static int romfs_mtd_strcmp(struct super_block *sb, unsigned long pos, + const char *str, size_t size) +{ + u_char buf[17]; + size_t len, segment; + int ret; + + /* scan the string up to 16 bytes at a time, and attempt to grab the + * trailing NUL whilst we're at it */ + buf[0] = 0xff; + + while (size > 0) { + segment = min_t(size_t, size + 1, 17); + ret = ROMFS_MTD_READ(sb, pos, segment, &len, buf); + if (ret < 0) + return ret; + len--; + if (memcmp(buf, str, len) != 0) + return 0; + buf[0] = buf[len]; + size -= len; + pos += len; + str += len; + } + + /* check the trailing NUL was */ + if (buf[0]) + return 0; + + return 1; +} +#endif /* CONFIG_ROMFS_ON_MTD */ + +#ifdef CONFIG_ROMFS_ON_BLOCK +/* + * read data from an romfs image on a block device + */ +static int romfs_blk_read(struct super_block *sb, unsigned long pos, + void *buf, size_t buflen) +{ + struct buffer_head *bh; + unsigned long offset; + size_t segment; + + /* copy the string up to blocksize bytes at a time */ + while (buflen > 0) { + offset = pos & (ROMBSIZE - 1); + segment = min_t(size_t, buflen, ROMBSIZE - offset); + bh = sb_bread(sb, pos >> ROMBSBITS); + if (!bh) + return -EIO; + memcpy(buf, bh->b_data + offset, segment); + brelse(bh); + buf += segment; + buflen -= segment; + pos += segment; + } + + return 0; +} + +/* + * determine the length of a string in romfs on a block device + */ +static ssize_t romfs_blk_strnlen(struct super_block *sb, + unsigned long pos, size_t limit) +{ + struct buffer_head *bh; + unsigned long offset; + ssize_t n = 0; + size_t segment; + u_char *buf, *p; + + /* scan the string up to blocksize bytes at a time */ + while (limit > 0) { + offset = pos & (ROMBSIZE - 1); + segment = min_t(size_t, limit, ROMBSIZE - offset); + bh = sb_bread(sb, pos >> ROMBSBITS); + if (!bh) + return -EIO; + buf = bh->b_data + offset; + p = memchr(buf, 0, segment); + brelse(bh); + if (p) + return n + (p - buf); + limit -= segment; + pos += segment; + n += segment; + } + + return n; +} + +/* + * compare a string to one in a romfs image on a block device + * - return 1 if matched, 0 if differ, -ve if error + */ +static int romfs_blk_strcmp(struct super_block *sb, unsigned long pos, + const char *str, size_t size) +{ + struct buffer_head *bh; + unsigned long offset; + size_t segment; + bool matched, terminated = false; + + /* compare string up to a block at a time */ + while (size > 0) { + offset = pos & (ROMBSIZE - 1); + segment = min_t(size_t, size, ROMBSIZE - offset); + bh = sb_bread(sb, pos >> ROMBSBITS); + if (!bh) + return -EIO; + matched = (memcmp(bh->b_data + offset, str, segment) == 0); + + size -= segment; + pos += segment; + str += segment; + if (matched && size == 0 && offset + segment < ROMBSIZE) { + if (!bh->b_data[offset + segment]) + terminated = true; + else + matched = false; + } + brelse(bh); + if (!matched) + return 0; + } + + if (!terminated) { + /* the terminating NUL must be on the first byte of the next + * block */ + BUG_ON((pos & (ROMBSIZE - 1)) != 0); + bh = sb_bread(sb, pos >> ROMBSBITS); + if (!bh) + return -EIO; + matched = !bh->b_data[0]; + brelse(bh); + if (!matched) + return 0; + } + + return 1; +} +#endif /* CONFIG_ROMFS_ON_BLOCK */ + +/* + * read data from the romfs image + */ +int romfs_dev_read(struct super_block *sb, unsigned long pos, + void *buf, size_t buflen) +{ + size_t limit; + + limit = romfs_maxsize(sb); + if (pos >= limit || buflen > limit - pos) + return -EIO; + +#ifdef CONFIG_ROMFS_ON_MTD + if (sb->s_mtd) + return romfs_mtd_read(sb, pos, buf, buflen); +#endif +#ifdef CONFIG_ROMFS_ON_BLOCK + if (sb->s_bdev) + return romfs_blk_read(sb, pos, buf, buflen); +#endif + return -EIO; +} + +/* + * determine the length of a string in romfs + */ +ssize_t romfs_dev_strnlen(struct super_block *sb, + unsigned long pos, size_t maxlen) +{ + size_t limit; + + limit = romfs_maxsize(sb); + if (pos >= limit) + return -EIO; + if (maxlen > limit - pos) + maxlen = limit - pos; + +#ifdef CONFIG_ROMFS_ON_MTD + if (sb->s_mtd) + return romfs_mtd_strnlen(sb, pos, maxlen); +#endif +#ifdef CONFIG_ROMFS_ON_BLOCK + if (sb->s_bdev) + return romfs_blk_strnlen(sb, pos, maxlen); +#endif + return -EIO; +} + +/* + * compare a string to one in romfs + * - the string to be compared to, str, may not be NUL-terminated; instead the + * string is of the specified size + * - return 1 if matched, 0 if differ, -ve if error + */ +int romfs_dev_strcmp(struct super_block *sb, unsigned long pos, + const char *str, size_t size) +{ + size_t limit; + + limit = romfs_maxsize(sb); + if (pos >= limit) + return -EIO; + if (size > ROMFS_MAXFN) + return -ENAMETOOLONG; + if (size + 1 > limit - pos) + return -EIO; + +#ifdef CONFIG_ROMFS_ON_MTD + if (sb->s_mtd) + return romfs_mtd_strcmp(sb, pos, str, size); +#endif +#ifdef CONFIG_ROMFS_ON_BLOCK + if (sb->s_bdev) + return romfs_blk_strcmp(sb, pos, str, size); +#endif + return -EIO; +} -- cgit v1.3.1