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, 413 insertions, 0 deletions

diff --git a/drivers/scsi/scsi_trace.c b/drivers/scsi/scsi_trace.c
new file mode 100644
index 000000000..b3baae91e
--- /dev/null
+++ b/drivers/scsi/scsi_trace.c
@@ -0,0 +1,413 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Copyright (C) 2010 FUJITSU LIMITED
+ * Copyright (C) 2010 Tomohiro Kusumi <kusumi.tomohiro@jp.fujitsu.com>
+ */
+#include <linux/kernel.h>
+#include <linux/trace_seq.h>
+#include <linux/unaligned.h>
+#include <trace/events/scsi.h>
+
+#define SERVICE_ACTION16(cdb) (cdb[1] & 0x1f)
+#define SERVICE_ACTION32(cdb) (get_unaligned_be16(&cdb[8]))
+
+static const char *
+scsi_trace_misc(struct trace_seq *, unsigned char *, int);
+
+static const char *
+scsi_trace_rw6(struct trace_seq *p, unsigned char *cdb, int len)
+{
+	const char *ret = trace_seq_buffer_ptr(p);
+	u32 lba, txlen;
+
+	lba = get_unaligned_be24(&cdb[1]) & 0x1fffff;
+	/*
+	 * From SBC-2: a TRANSFER LENGTH field set to zero specifies that 256
+	 * logical blocks shall be read (READ(6)) or written (WRITE(6)).
+	 */
+	txlen = cdb[4] ? cdb[4] : 256;
+
+	trace_seq_printf(p, "lba=%u txlen=%u", lba, txlen);
+	trace_seq_putc(p, 0);
+
+	return ret;
+}
+
+static const char *
+scsi_trace_rw10(struct trace_seq *p, unsigned char *cdb, int len)
+{
+	const char *ret = trace_seq_buffer_ptr(p);
+	u32 lba, txlen;
+
+	lba = get_unaligned_be32(&cdb[2]);
+	txlen = get_unaligned_be16(&cdb[7]);
+
+	trace_seq_printf(p, "lba=%u txlen=%u protect=%u", lba, txlen,
+			 cdb[1] >> 5);
+
+	if (cdb[0] == WRITE_SAME)
+		trace_seq_printf(p, " unmap=%u", cdb[1] >> 3 & 1);
+
+	trace_seq_putc(p, 0);
+
+	return ret;
+}
+
+static const char *
+scsi_trace_rw12(struct trace_seq *p, unsigned char *cdb, int len)
+{
+	const char *ret = trace_seq_buffer_ptr(p);
+	u32 lba, txlen;
+
+	lba = get_unaligned_be32(&cdb[2]);
+	txlen = get_unaligned_be32(&cdb[6]);
+
+	trace_seq_printf(p, "lba=%u txlen=%u protect=%u", lba, txlen,
+			 cdb[1] >> 5);
+	trace_seq_putc(p, 0);
+
+	return ret;
+}
+
+static const char *
+scsi_trace_rw16(struct trace_seq *p, unsigned char *cdb, int len)
+{
+	const char *ret = trace_seq_buffer_ptr(p);
+	u64 lba;
+	u32 txlen;
+
+	lba = get_unaligned_be64(&cdb[2]);
+	txlen = get_unaligned_be32(&cdb[10]);
+
+	trace_seq_printf(p, "lba=%llu txlen=%u protect=%u", lba, txlen,
+			 cdb[1] >> 5);
+
+	if (cdb[0] == WRITE_SAME_16)
+		trace_seq_printf(p, " unmap=%u", cdb[1] >> 3 & 1);
+
+	trace_seq_putc(p, 0);
+
+	return ret;
+}
+
+static const char *
+scsi_trace_rw32(struct trace_seq *p, unsigned char *cdb, int len)
+{
+	const char *ret = trace_seq_buffer_ptr(p), *cmd;
+	u64 lba;
+	u32 ei_lbrt, txlen;
+
+	switch (SERVICE_ACTION32(cdb)) {
+	case READ_32:
+		cmd = "READ";
+		break;
+	case VERIFY_32:
+		cmd = "VERIFY";
+		break;
+	case WRITE_32:
+		cmd = "WRITE";
+		break;
+	case WRITE_SAME_32:
+		cmd = "WRITE_SAME";
+		break;
+	default:
+		trace_seq_puts(p, "UNKNOWN");
+		goto out;
+	}
+
+	lba = get_unaligned_be64(&cdb[12]);
+	ei_lbrt = get_unaligned_be32(&cdb[20]);
+	txlen = get_unaligned_be32(&cdb[28]);
+
+	trace_seq_printf(p, "%s_32 lba=%llu txlen=%u protect=%u ei_lbrt=%u",
+			 cmd, lba, txlen, cdb[10] >> 5, ei_lbrt);
+
+	if (SERVICE_ACTION32(cdb) == WRITE_SAME_32)
+		trace_seq_printf(p, " unmap=%u", cdb[10] >> 3 & 1);
+
+out:
+	trace_seq_putc(p, 0);
+
+	return ret;
+}
+
+static const char *
+scsi_trace_unmap(struct trace_seq *p, unsigned char *cdb, int len)
+{
+	const char *ret = trace_seq_buffer_ptr(p);
+	unsigned int regions = get_unaligned_be16(&cdb[7]);
+
+	trace_seq_printf(p, "regions=%u", (regions - 8) / 16);
+	trace_seq_putc(p, 0);
+
+	return ret;
+}
+
+static const char *
+scsi_trace_service_action_in(struct trace_seq *p, unsigned char *cdb, int len)
+{
+	const char *ret = trace_seq_buffer_ptr(p), *cmd;
+	u64 lba;
+	u32 alloc_len;
+
+	switch (SERVICE_ACTION16(cdb)) {
+	case SAI_READ_CAPACITY_16:
+		cmd = "READ_CAPACITY_16";
+		break;
+	case SAI_GET_LBA_STATUS:
+		cmd = "GET_LBA_STATUS";
+		break;
+	default:
+		trace_seq_puts(p, "UNKNOWN");
+		goto out;
+	}
+
+	lba = get_unaligned_be64(&cdb[2]);
+	alloc_len = get_unaligned_be32(&cdb[10]);
+
+	trace_seq_printf(p, "%s lba=%llu alloc_len=%u", cmd, lba, alloc_len);
+
+out:
+	trace_seq_putc(p, 0);
+
+	return ret;
+}
+
+static const char *
+scsi_trace_maintenance_in(struct trace_seq *p, unsigned char *cdb, int len)
+{
+	const char *ret = trace_seq_buffer_ptr(p), *cmd;
+	u32 alloc_len;
+
+	switch (SERVICE_ACTION16(cdb)) {
+	case MI_REPORT_IDENTIFYING_INFORMATION:
+		cmd = "REPORT_IDENTIFYING_INFORMATION";
+		break;
+	case MI_REPORT_TARGET_PGS:
+		cmd = "REPORT_TARGET_PORT_GROUPS";
+		break;
+	case MI_REPORT_ALIASES:
+		cmd = "REPORT_ALIASES";
+		break;
+	case MI_REPORT_SUPPORTED_OPERATION_CODES:
+		cmd = "REPORT_SUPPORTED_OPERATION_CODES";
+		break;
+	case MI_REPORT_SUPPORTED_TASK_MANAGEMENT_FUNCTIONS:
+		cmd = "REPORT_SUPPORTED_TASK_MANAGEMENT_FUNCTIONS";
+		break;
+	case MI_REPORT_PRIORITY:
+		cmd = "REPORT_PRIORITY";
+		break;
+	case MI_REPORT_TIMESTAMP:
+		cmd = "REPORT_TIMESTAMP";
+		break;
+	case MI_MANAGEMENT_PROTOCOL_IN:
+		cmd = "MANAGEMENT_PROTOCOL_IN";
+		break;
+	default:
+		trace_seq_puts(p, "UNKNOWN");
+		goto out;
+	}
+
+	alloc_len = get_unaligned_be32(&cdb[6]);
+
+	trace_seq_printf(p, "%s alloc_len=%u", cmd, alloc_len);
+
+out:
+	trace_seq_putc(p, 0);
+
+	return ret;
+}
+
+static const char *
+scsi_trace_maintenance_out(struct trace_seq *p, unsigned char *cdb, int len)
+{
+	const char *ret = trace_seq_buffer_ptr(p), *cmd;
+	u32 alloc_len;
+
+	switch (SERVICE_ACTION16(cdb)) {
+	case MO_SET_IDENTIFYING_INFORMATION:
+		cmd = "SET_IDENTIFYING_INFORMATION";
+		break;
+	case MO_SET_TARGET_PGS:
+		cmd = "SET_TARGET_PORT_GROUPS";
+		break;
+	case MO_CHANGE_ALIASES:
+		cmd = "CHANGE_ALIASES";
+		break;
+	case MO_SET_PRIORITY:
+		cmd = "SET_PRIORITY";
+		break;
+	case MO_SET_TIMESTAMP:
+		cmd = "SET_TIMESTAMP";
+		break;
+	case MO_MANAGEMENT_PROTOCOL_OUT:
+		cmd = "MANAGEMENT_PROTOCOL_OUT";
+		break;
+	default:
+		trace_seq_puts(p, "UNKNOWN");
+		goto out;
+	}
+
+	alloc_len = get_unaligned_be32(&cdb[6]);
+
+	trace_seq_printf(p, "%s alloc_len=%u", cmd, alloc_len);
+
+out:
+	trace_seq_putc(p, 0);
+
+	return ret;
+}
+
+static const char *
+scsi_trace_zbc_in(struct trace_seq *p, unsigned char *cdb, int len)
+{
+	const char *ret = trace_seq_buffer_ptr(p), *cmd;
+	u64 zone_id;
+	u32 alloc_len;
+	u8 options;
+
+	switch (SERVICE_ACTION16(cdb)) {
+	case ZI_REPORT_ZONES:
+		cmd = "REPORT_ZONES";
+		break;
+	default:
+		trace_seq_puts(p, "UNKNOWN");
+		goto out;
+	}
+
+	zone_id = get_unaligned_be64(&cdb[2]);
+	alloc_len = get_unaligned_be32(&cdb[10]);
+	options = cdb[14] & 0x3f;
+
+	trace_seq_printf(p, "%s zone=%llu alloc_len=%u options=%u partial=%u",
+			 cmd, (unsigned long long)zone_id, alloc_len,
+			 options, (cdb[14] >> 7) & 1);
+
+out:
+	trace_seq_putc(p, 0);
+
+	return ret;
+}
+
+static const char *
+scsi_trace_zbc_out(struct trace_seq *p, unsigned char *cdb, int len)
+{
+	const char *ret = trace_seq_buffer_ptr(p), *cmd;
+	u64 zone_id;
+
+	switch (SERVICE_ACTION16(cdb)) {
+	case ZO_CLOSE_ZONE:
+		cmd = "CLOSE_ZONE";
+		break;
+	case ZO_FINISH_ZONE:
+		cmd = "FINISH_ZONE";
+		break;
+	case ZO_OPEN_ZONE:
+		cmd = "OPEN_ZONE";
+		break;
+	case ZO_RESET_WRITE_POINTER:
+		cmd = "RESET_WRITE_POINTER";
+		break;
+	default:
+		trace_seq_puts(p, "UNKNOWN");
+		goto out;
+	}
+
+	zone_id = get_unaligned_be64(&cdb[2]);
+
+	trace_seq_printf(p, "%s zone=%llu all=%u", cmd,
+			 (unsigned long long)zone_id, cdb[14] & 1);
+
+out:
+	trace_seq_putc(p, 0);
+
+	return ret;
+}
+
+static const char *
+scsi_trace_atomic_write16_out(struct trace_seq *p, unsigned char *cdb, int len)
+{
+	const char *ret = trace_seq_buffer_ptr(p);
+	unsigned int boundary_size;
+	unsigned int nr_blocks;
+	sector_t lba;
+
+	lba = get_unaligned_be64(&cdb[2]);
+	boundary_size = get_unaligned_be16(&cdb[10]);
+	nr_blocks = get_unaligned_be16(&cdb[12]);
+
+	trace_seq_printf(p, "lba=%llu txlen=%u boundary_size=%u",
+			  lba, nr_blocks, boundary_size);
+
+	trace_seq_putc(p, 0);
+
+	return ret;
+}
+
+static const char *
+scsi_trace_varlen(struct trace_seq *p, unsigned char *cdb, int len)
+{
+	switch (SERVICE_ACTION32(cdb)) {
+	case READ_32:
+	case VERIFY_32:
+	case WRITE_32:
+	case WRITE_SAME_32:
+		return scsi_trace_rw32(p, cdb, len);
+	default:
+		return scsi_trace_misc(p, cdb, len);
+	}
+}
+
+static const char *
+scsi_trace_misc(struct trace_seq *p, unsigned char *cdb, int len)
+{
+	const char *ret = trace_seq_buffer_ptr(p);
+
+	trace_seq_putc(p, '-');
+	trace_seq_putc(p, 0);
+
+	return ret;
+}
+
+const char *
+scsi_trace_parse_cdb(struct trace_seq *p, unsigned char *cdb, int len)
+{
+	switch (cdb[0]) {
+	case READ_6:
+	case WRITE_6:
+		return scsi_trace_rw6(p, cdb, len);
+	case READ_10:
+	case VERIFY:
+	case WRITE_10:
+	case WRITE_SAME:
+		return scsi_trace_rw10(p, cdb, len);
+	case READ_12:
+	case VERIFY_12:
+	case WRITE_12:
+		return scsi_trace_rw12(p, cdb, len);
+	case READ_16:
+	case VERIFY_16:
+	case WRITE_16:
+	case WRITE_SAME_16:
+		return scsi_trace_rw16(p, cdb, len);
+	case UNMAP:
+		return scsi_trace_unmap(p, cdb, len);
+	case SERVICE_ACTION_IN_16:
+		return scsi_trace_service_action_in(p, cdb, len);
+	case VARIABLE_LENGTH_CMD:
+		return scsi_trace_varlen(p, cdb, len);
+	case MAINTENANCE_IN:
+		return scsi_trace_maintenance_in(p, cdb, len);
+	case MAINTENANCE_OUT:
+		return scsi_trace_maintenance_out(p, cdb, len);
+	case ZBC_IN:
+		return scsi_trace_zbc_in(p, cdb, len);
+	case ZBC_OUT:
+		return scsi_trace_zbc_out(p, cdb, len);
+	case WRITE_ATOMIC_16:
+		return scsi_trace_atomic_write16_out(p, cdb, len);
+	default:
+		return scsi_trace_misc(p, cdb, len);
+	}
+}