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, 118 insertions, 0 deletions

diff --git a/fs/befs/super.c b/fs/befs/super.c
new file mode 100644
index 000000000..7c50025c9
--- /dev/null
+++ b/fs/befs/super.c
@@ -0,0 +1,118 @@
+/*
+ * super.c
+ *
+ * Copyright (C) 2001-2002 Will Dyson <will_dyson@pobox.com>
+ *
+ * Licensed under the GNU GPL. See the file COPYING for details.
+ *
+ */
+
+#include <linux/fs.h>
+#include <asm/page.h> /* for PAGE_SIZE */
+
+#include "befs.h"
+#include "super.h"
+
+/*
+ * befs_load_sb -- Read from disk and properly byteswap all the fields
+ * of the befs superblock
+ */
+int
+befs_load_sb(struct super_block *sb, befs_super_block *disk_sb)
+{
+	struct befs_sb_info *befs_sb = BEFS_SB(sb);
+
+	/* Check the byte order of the filesystem */
+	if (disk_sb->fs_byte_order == BEFS_BYTEORDER_NATIVE_LE)
+		befs_sb->byte_order = BEFS_BYTESEX_LE;
+	else if (disk_sb->fs_byte_order == BEFS_BYTEORDER_NATIVE_BE)
+		befs_sb->byte_order = BEFS_BYTESEX_BE;
+
+	befs_sb->magic1 = fs32_to_cpu(sb, disk_sb->magic1);
+	befs_sb->magic2 = fs32_to_cpu(sb, disk_sb->magic2);
+	befs_sb->magic3 = fs32_to_cpu(sb, disk_sb->magic3);
+	befs_sb->block_size = fs32_to_cpu(sb, disk_sb->block_size);
+	befs_sb->block_shift = fs32_to_cpu(sb, disk_sb->block_shift);
+	befs_sb->num_blocks = fs64_to_cpu(sb, disk_sb->num_blocks);
+	befs_sb->used_blocks = fs64_to_cpu(sb, disk_sb->used_blocks);
+	befs_sb->inode_size = fs32_to_cpu(sb, disk_sb->inode_size);
+
+	befs_sb->blocks_per_ag = fs32_to_cpu(sb, disk_sb->blocks_per_ag);
+	befs_sb->ag_shift = fs32_to_cpu(sb, disk_sb->ag_shift);
+	befs_sb->num_ags = fs32_to_cpu(sb, disk_sb->num_ags);
+
+	befs_sb->flags = fs32_to_cpu(sb, disk_sb->flags);
+
+	befs_sb->log_blocks = fsrun_to_cpu(sb, disk_sb->log_blocks);
+	befs_sb->log_start = fs64_to_cpu(sb, disk_sb->log_start);
+	befs_sb->log_end = fs64_to_cpu(sb, disk_sb->log_end);
+
+	befs_sb->root_dir = fsrun_to_cpu(sb, disk_sb->root_dir);
+	befs_sb->indices = fsrun_to_cpu(sb, disk_sb->indices);
+	befs_sb->nls = NULL;
+
+	return BEFS_OK;
+}
+
+int
+befs_check_sb(struct super_block *sb)
+{
+	struct befs_sb_info *befs_sb = BEFS_SB(sb);
+
+	/* Check magic headers of super block */
+	if ((befs_sb->magic1 != BEFS_SUPER_MAGIC1)
+	    || (befs_sb->magic2 != BEFS_SUPER_MAGIC2)
+	    || (befs_sb->magic3 != BEFS_SUPER_MAGIC3)) {
+		befs_error(sb, "invalid magic header");
+		return BEFS_ERR;
+	}
+
+	/*
+	 * Check blocksize of BEFS.
+	 *
+	 * Blocksize of BEFS is 1024, 2048, 4096 or 8192.
+	 */
+
+	if ((befs_sb->block_size != 1024)
+	    && (befs_sb->block_size != 2048)
+	    && (befs_sb->block_size != 4096)
+	    && (befs_sb->block_size != 8192)) {
+		befs_error(sb, "invalid blocksize: %u", befs_sb->block_size);
+		return BEFS_ERR;
+	}
+
+	if (befs_sb->block_size > PAGE_SIZE) {
+		befs_error(sb, "blocksize(%u) cannot be larger "
+			   "than system pagesize(%lu)", befs_sb->block_size,
+			   PAGE_SIZE);
+		return BEFS_ERR;
+	}
+
+	/*
+	 * block_shift and block_size encode the same information
+	 * in different ways as a consistency check.
+	 */
+
+	if ((1 << befs_sb->block_shift) != befs_sb->block_size) {
+		befs_error(sb, "block_shift disagrees with block_size. "
+			   "Corruption likely.");
+		return BEFS_ERR;
+	}
+
+
+	/* ag_shift also encodes the same information as blocks_per_ag in a
+	 * different way, non-fatal consistency check
+	 */
+	if ((1 << befs_sb->ag_shift) != befs_sb->blocks_per_ag)
+		befs_error(sb, "ag_shift disagrees with blocks_per_ag.");
+
+	if (befs_sb->log_start != befs_sb->log_end ||
+	    befs_sb->flags == BEFS_DIRTY) {
+		befs_error(sb, "Filesystem not clean! There are blocks in the "
+			   "journal. You must boot into BeOS and mount this "
+			   "volume to make it clean.");
+		return BEFS_ERR;
+	}
+
+	return BEFS_OK;
+}