x86/kprobes: Fix crash when probing CS CALL instructionsgrafted

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) <mhiramat@kernel.org>
Signed-off-by: Jinke Han <jinkehan@didiglobal.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Reviewed-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
Acked-by: Yafang Shao <laoar.shao@gmail.com>
Acked-by: Borislav Petkov <bp@alien8.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://patch.msgid.link/20260908073742.GA10517@didi-ThinkCentre-M920t-N000

1 files changed, 207 insertions, 0 deletions

diff --git a/fs/exfat/misc.c b/fs/exfat/misc.c
new file mode 100644
index 000000000..6f11a96a4
--- /dev/null
+++ b/fs/exfat/misc.c
@@ -0,0 +1,207 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ *  Written 1992,1993 by Werner Almesberger
+ *  22/11/2000 - Fixed fat_date_unix2dos for dates earlier than 01/01/1980
+ *		 and date_dos2unix for date==0 by Igor Zhbanov(bsg@uniyar.ac.ru)
+ * Copyright (C) 2012-2013 Samsung Electronics Co., Ltd.
+ */
+
+#include <linux/time.h>
+#include <linux/fs.h>
+#include <linux/slab.h>
+#include <linux/buffer_head.h>
+#include <linux/blk_types.h>
+
+#include "exfat_raw.h"
+#include "exfat_fs.h"
+
+/*
+ * exfat_fs_error reports a file system problem that might indicate fa data
+ * corruption/inconsistency. Depending on 'errors' mount option the
+ * panic() is called, or error message is printed FAT and nothing is done,
+ * or filesystem is remounted read-only (default behavior).
+ * In case the file system is remounted read-only, it can be made writable
+ * again by remounting it.
+ */
+void __exfat_fs_error(struct super_block *sb, int report, const char *fmt, ...)
+{
+	struct exfat_mount_options *opts = &EXFAT_SB(sb)->options;
+	va_list args;
+	struct va_format vaf;
+
+	if (report) {
+		va_start(args, fmt);
+		vaf.fmt = fmt;
+		vaf.va = &args;
+		exfat_err(sb, "error, %pV", &vaf);
+		va_end(args);
+	}
+
+	if (opts->errors == EXFAT_ERRORS_PANIC) {
+		panic("exFAT-fs (%s): fs panic from previous error\n",
+			sb->s_id);
+	} else if (opts->errors == EXFAT_ERRORS_RO && !sb_rdonly(sb)) {
+		sb->s_flags |= SB_RDONLY;
+		exfat_err(sb, "Filesystem has been set read-only");
+	}
+}
+
+#define SECS_PER_MIN    (60)
+#define TIMEZONE_SEC(x)	((x) * 15 * SECS_PER_MIN)
+
+static void exfat_adjust_tz(struct timespec64 *ts, u8 tz_off)
+{
+	if (tz_off <= 0x3F)
+		ts->tv_sec -= TIMEZONE_SEC(tz_off);
+	else /* 0x40 <= (tz_off & 0x7F) <=0x7F */
+		ts->tv_sec += TIMEZONE_SEC(0x80 - tz_off);
+}
+
+static inline int exfat_tz_offset(struct exfat_sb_info *sbi)
+{
+	if (sbi->options.sys_tz)
+		return -sys_tz.tz_minuteswest;
+	return sbi->options.time_offset;
+}
+
+/* Convert a EXFAT time/date pair to a UNIX date (seconds since 1 1 70). */
+void exfat_get_entry_time(struct exfat_sb_info *sbi, struct timespec64 *ts,
+		u8 tz, __le16 time, __le16 date, u8 time_cs)
+{
+	u16 t = le16_to_cpu(time);
+	u16 d = le16_to_cpu(date);
+
+	ts->tv_sec = mktime64(1980 + (d >> 9), d >> 5 & 0x000F, d & 0x001F,
+			      t >> 11, (t >> 5) & 0x003F, (t & 0x001F) << 1);
+
+
+	/* time_cs field represent 0 ~ 199cs(1990 ms) */
+	if (time_cs) {
+		ts->tv_sec += time_cs / 100;
+		ts->tv_nsec = (time_cs % 100) * 10 * NSEC_PER_MSEC;
+	} else
+		ts->tv_nsec = 0;
+
+	if (tz & EXFAT_TZ_VALID)
+		/* Adjust timezone to UTC0. */
+		exfat_adjust_tz(ts, tz & ~EXFAT_TZ_VALID);
+	else
+		ts->tv_sec -= exfat_tz_offset(sbi) * SECS_PER_MIN;
+}
+
+/* Convert linear UNIX date to a EXFAT time/date pair. */
+void exfat_set_entry_time(struct exfat_sb_info *sbi, struct timespec64 *ts,
+		u8 *tz, __le16 *time, __le16 *date, u8 *time_cs)
+{
+	struct tm tm;
+	u16 t, d;
+
+	time64_to_tm(ts->tv_sec, 0, &tm);
+	t = (tm.tm_hour << 11) | (tm.tm_min << 5) | (tm.tm_sec >> 1);
+	d = ((tm.tm_year - 80) <<  9) | ((tm.tm_mon + 1) << 5) | tm.tm_mday;
+
+	*time = cpu_to_le16(t);
+	*date = cpu_to_le16(d);
+
+	/* time_cs field represent 0 ~ 199cs(1990 ms) */
+	if (time_cs)
+		*time_cs = (tm.tm_sec & 1) * 100 +
+			ts->tv_nsec / (10 * NSEC_PER_MSEC);
+
+	/*
+	 * Record 00h value for OffsetFromUtc field and 1 value for OffsetValid
+	 * to indicate that local time and UTC are the same.
+	 */
+	*tz = EXFAT_TZ_VALID;
+}
+
+/*
+ * The timestamp for access_time has double seconds granularity.
+ * (There is no 10msIncrement field for access_time unlike create/modify_time)
+ * atime also has only a 2-second resolution.
+ */
+void exfat_truncate_atime(struct timespec64 *ts)
+{
+	ts->tv_sec = round_down(ts->tv_sec, 2);
+	ts->tv_nsec = 0;
+}
+
+void exfat_truncate_inode_atime(struct inode *inode)
+{
+	struct timespec64 atime = inode_get_atime(inode);
+
+	exfat_truncate_atime(&atime);
+	inode_set_atime_to_ts(inode, atime);
+}
+
+u16 exfat_calc_chksum16(void *data, int len, u16 chksum, int type)
+{
+	int i;
+	u8 *c = (u8 *)data;
+
+	for (i = 0; i < len; i++, c++) {
+		if (unlikely(type == CS_DIR_ENTRY && (i == 2 || i == 3)))
+			continue;
+		chksum = ((chksum << 15) | (chksum >> 1)) + *c;
+	}
+	return chksum;
+}
+
+u32 exfat_calc_chksum32(void *data, int len, u32 chksum, int type)
+{
+	int i;
+	u8 *c = (u8 *)data;
+
+	for (i = 0; i < len; i++, c++) {
+		if (unlikely(type == CS_BOOT_SECTOR &&
+			     (i == 106 || i == 107 || i == 112)))
+			continue;
+		chksum = ((chksum << 31) | (chksum >> 1)) + *c;
+	}
+	return chksum;
+}
+
+int exfat_update_bh(struct buffer_head *bh, int sync)
+{
+	int err = 0;
+
+	set_buffer_uptodate(bh);
+	mark_buffer_dirty(bh);
+
+	if (sync)
+		err = sync_dirty_buffer(bh);
+
+	return err;
+}
+
+int exfat_update_bhs(struct buffer_head **bhs, int nr_bhs, int sync)
+{
+	int i, err = 0;
+
+	for (i = 0; i < nr_bhs; i++) {
+		set_buffer_uptodate(bhs[i]);
+		mark_buffer_dirty(bhs[i]);
+		if (sync)
+			write_dirty_buffer(bhs[i], REQ_SYNC);
+	}
+
+	for (i = 0; i < nr_bhs && sync; i++) {
+		wait_on_buffer(bhs[i]);
+		if (!err && !buffer_uptodate(bhs[i]))
+			err = -EIO;
+	}
+	return err;
+}
+
+void exfat_chain_set(struct exfat_chain *ec, unsigned int dir,
+		unsigned int size, unsigned char flags)
+{
+	ec->dir = dir;
+	ec->size = size;
+	ec->flags = flags;
+}
+
+void exfat_chain_dup(struct exfat_chain *dup, struct exfat_chain *ec)
+{
+	return exfat_chain_set(dup, ec->dir, ec->size, ec->flags);
+}