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, 283 insertions, 0 deletions

diff --git a/fs/ext4/inode-test.c b/fs/ext4/inode-test.c
new file mode 100644
index 000000000..749af7ad4
--- /dev/null
+++ b/fs/ext4/inode-test.c
@@ -0,0 +1,283 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * KUnit test of ext4 inode that verify the seconds part of [a/c/m]
+ * timestamps in ext4 inode structs are decoded correctly.
+ */
+
+#include <kunit/test.h>
+#include <linux/kernel.h>
+#include <linux/time64.h>
+
+#include "ext4.h"
+
+/*
+ * For constructing the nonnegative timestamp lower bound value.
+ * binary: 00000000 00000000 00000000 00000000
+ */
+#define LOWER_MSB_0 0L
+/*
+ * For constructing the nonnegative timestamp upper bound value.
+ * binary: 01111111 11111111 11111111 11111111
+ *
+ */
+#define UPPER_MSB_0 0x7fffffffL
+/*
+ * For constructing the negative timestamp lower bound value.
+ * binary: 10000000 00000000 00000000 00000000
+ */
+#define LOWER_MSB_1 (-(UPPER_MSB_0) - 1L)  /* avoid overflow */
+/*
+ * For constructing the negative timestamp upper bound value.
+ * binary: 11111111 11111111 11111111 11111111
+ */
+#define UPPER_MSB_1 (-1L)
+/*
+ * Upper bound for nanoseconds value supported by the encoding.
+ * binary: 00111111 11111111 11111111 11111111
+ */
+#define MAX_NANOSECONDS ((1L << 30) - 1)
+
+#define CASE_NAME_FORMAT "%s: msb:%x lower_bound:%x extra_bits: %x"
+
+#define LOWER_BOUND_NEG_NO_EXTRA_BITS_CASE\
+	"1901-12-13 Lower bound of 32bit < 0 timestamp, no extra bits"
+#define UPPER_BOUND_NEG_NO_EXTRA_BITS_CASE\
+	"1969-12-31 Upper bound of 32bit < 0 timestamp, no extra bits"
+#define LOWER_BOUND_NONNEG_NO_EXTRA_BITS_CASE\
+	"1970-01-01 Lower bound of 32bit >=0 timestamp, no extra bits"
+#define UPPER_BOUND_NONNEG_NO_EXTRA_BITS_CASE\
+	"2038-01-19 Upper bound of 32bit >=0 timestamp, no extra bits"
+#define LOWER_BOUND_NEG_LO_1_CASE\
+	"2038-01-19 Lower bound of 32bit <0 timestamp, lo extra sec bit on"
+#define UPPER_BOUND_NEG_LO_1_CASE\
+	"2106-02-07 Upper bound of 32bit <0 timestamp, lo extra sec bit on"
+#define LOWER_BOUND_NONNEG_LO_1_CASE\
+	"2106-02-07 Lower bound of 32bit >=0 timestamp, lo extra sec bit on"
+#define UPPER_BOUND_NONNEG_LO_1_CASE\
+	"2174-02-25 Upper bound of 32bit >=0 timestamp, lo extra sec bit on"
+#define LOWER_BOUND_NEG_HI_1_CASE\
+	"2174-02-25 Lower bound of 32bit <0 timestamp, hi extra sec bit on"
+#define UPPER_BOUND_NEG_HI_1_CASE\
+	"2242-03-16 Upper bound of 32bit <0 timestamp, hi extra sec bit on"
+#define LOWER_BOUND_NONNEG_HI_1_CASE\
+	"2242-03-16 Lower bound of 32bit >=0 timestamp, hi extra sec bit on"
+#define UPPER_BOUND_NONNEG_HI_1_CASE\
+	"2310-04-04 Upper bound of 32bit >=0 timestamp, hi extra sec bit on"
+#define UPPER_BOUND_NONNEG_HI_1_NS_1_CASE\
+	"2310-04-04 Upper bound of 32bit>=0 timestamp, hi extra sec bit 1. 1 ns"
+#define LOWER_BOUND_NONNEG_HI_1_NS_MAX_CASE\
+	"2378-04-22 Lower bound of 32bit>= timestamp. Extra sec bits 1. Max ns"
+#define LOWER_BOUND_NONNEG_EXTRA_BITS_1_CASE\
+	"2378-04-22 Lower bound of 32bit >=0 timestamp. All extra sec bits on"
+#define UPPER_BOUND_NONNEG_EXTRA_BITS_1_CASE\
+	"2446-05-10 Upper bound of 32bit >=0 timestamp. All extra sec bits on"
+
+struct timestamp_expectation {
+	const char *test_case_name;
+	struct timespec64 expected;
+	u32 extra_bits;
+	bool msb_set;
+	bool lower_bound;
+};
+
+static const struct timestamp_expectation test_data[] = {
+	{
+		.test_case_name = LOWER_BOUND_NEG_NO_EXTRA_BITS_CASE,
+		.msb_set = true,
+		.lower_bound = true,
+		.extra_bits = 0,
+		.expected = {.tv_sec = -0x80000000LL, .tv_nsec = 0L},
+	},
+
+	{
+		.test_case_name = UPPER_BOUND_NEG_NO_EXTRA_BITS_CASE,
+		.msb_set = true,
+		.lower_bound = false,
+		.extra_bits = 0,
+		.expected = {.tv_sec = -1LL, .tv_nsec = 0L},
+	},
+
+	{
+		.test_case_name = LOWER_BOUND_NONNEG_NO_EXTRA_BITS_CASE,
+		.msb_set = false,
+		.lower_bound = true,
+		.extra_bits = 0,
+		.expected = {0LL, 0L},
+	},
+
+	{
+		.test_case_name = UPPER_BOUND_NONNEG_NO_EXTRA_BITS_CASE,
+		.msb_set = false,
+		.lower_bound = false,
+		.extra_bits = 0,
+		.expected = {.tv_sec = 0x7fffffffLL, .tv_nsec = 0L},
+	},
+
+	{
+		.test_case_name = LOWER_BOUND_NEG_LO_1_CASE,
+		.msb_set = true,
+		.lower_bound = true,
+		.extra_bits = 1,
+		.expected = {.tv_sec = 0x80000000LL, .tv_nsec = 0L},
+	},
+
+	{
+		.test_case_name = UPPER_BOUND_NEG_LO_1_CASE,
+		.msb_set = true,
+		.lower_bound = false,
+		.extra_bits = 1,
+		.expected = {.tv_sec = 0xffffffffLL, .tv_nsec = 0L},
+	},
+
+	{
+		.test_case_name = LOWER_BOUND_NONNEG_LO_1_CASE,
+		.msb_set = false,
+		.lower_bound = true,
+		.extra_bits = 1,
+		.expected = {.tv_sec = 0x100000000LL, .tv_nsec = 0L},
+	},
+
+	{
+		.test_case_name = UPPER_BOUND_NONNEG_LO_1_CASE,
+		.msb_set = false,
+		.lower_bound = false,
+		.extra_bits = 1,
+		.expected = {.tv_sec = 0x17fffffffLL, .tv_nsec = 0L},
+	},
+
+	{
+		.test_case_name = LOWER_BOUND_NEG_HI_1_CASE,
+		.msb_set = true,
+		.lower_bound = true,
+		.extra_bits =  2,
+		.expected = {.tv_sec = 0x180000000LL, .tv_nsec = 0L},
+	},
+
+	{
+		.test_case_name = UPPER_BOUND_NEG_HI_1_CASE,
+		.msb_set = true,
+		.lower_bound = false,
+		.extra_bits = 2,
+		.expected = {.tv_sec = 0x1ffffffffLL, .tv_nsec = 0L},
+	},
+
+	{
+		.test_case_name = LOWER_BOUND_NONNEG_HI_1_CASE,
+		.msb_set = false,
+		.lower_bound = true,
+		.extra_bits = 2,
+		.expected = {.tv_sec = 0x200000000LL, .tv_nsec = 0L},
+	},
+
+	{
+		.test_case_name = UPPER_BOUND_NONNEG_HI_1_CASE,
+		.msb_set = false,
+		.lower_bound = false,
+		.extra_bits = 2,
+		.expected = {.tv_sec = 0x27fffffffLL, .tv_nsec = 0L},
+	},
+
+	{
+		.test_case_name = UPPER_BOUND_NONNEG_HI_1_NS_1_CASE,
+		.msb_set = false,
+		.lower_bound = false,
+		.extra_bits = 6,
+		.expected = {.tv_sec = 0x27fffffffLL, .tv_nsec = 1L},
+	},
+
+	{
+		.test_case_name = LOWER_BOUND_NONNEG_HI_1_NS_MAX_CASE,
+		.msb_set = false,
+		.lower_bound = true,
+		.extra_bits = 0xFFFFFFFF,
+		.expected = {.tv_sec = 0x300000000LL,
+			     .tv_nsec = MAX_NANOSECONDS},
+	},
+
+	{
+		.test_case_name = LOWER_BOUND_NONNEG_EXTRA_BITS_1_CASE,
+		.msb_set = false,
+		.lower_bound = true,
+		.extra_bits = 3,
+		.expected = {.tv_sec = 0x300000000LL, .tv_nsec = 0L},
+	},
+
+	{
+		.test_case_name = UPPER_BOUND_NONNEG_EXTRA_BITS_1_CASE,
+		.msb_set = false,
+		.lower_bound = false,
+		.extra_bits = 3,
+		.expected = {.tv_sec = 0x37fffffffLL, .tv_nsec = 0L},
+	}
+};
+
+static void timestamp_expectation_to_desc(const struct timestamp_expectation *t,
+					  char *desc)
+{
+	strscpy(desc, t->test_case_name, KUNIT_PARAM_DESC_SIZE);
+}
+
+KUNIT_ARRAY_PARAM(ext4_inode, test_data, timestamp_expectation_to_desc);
+
+static time64_t get_32bit_time(const struct timestamp_expectation * const test)
+{
+	if (test->msb_set) {
+		if (test->lower_bound)
+			return LOWER_MSB_1;
+
+		return UPPER_MSB_1;
+	}
+
+	if (test->lower_bound)
+		return LOWER_MSB_0;
+	return UPPER_MSB_0;
+}
+
+
+/*
+ *  Test data is derived from the table in the Inode Timestamps section of
+ *  Documentation/filesystems/ext4/inodes.rst.
+ */
+static void inode_test_xtimestamp_decoding(struct kunit *test)
+{
+	struct timespec64 timestamp;
+
+	struct timestamp_expectation *test_param =
+			(struct timestamp_expectation *)(test->param_value);
+
+	timestamp = ext4_decode_extra_time(
+				cpu_to_le32(get_32bit_time(test_param)),
+				cpu_to_le32(test_param->extra_bits));
+
+	KUNIT_EXPECT_EQ_MSG(test,
+			    test_param->expected.tv_sec,
+			    timestamp.tv_sec,
+			    CASE_NAME_FORMAT,
+			    test_param->test_case_name,
+			    test_param->msb_set,
+			    test_param->lower_bound,
+			    test_param->extra_bits);
+	KUNIT_EXPECT_EQ_MSG(test,
+			    test_param->expected.tv_nsec,
+			    timestamp.tv_nsec,
+			    CASE_NAME_FORMAT,
+			    test_param->test_case_name,
+			    test_param->msb_set,
+			    test_param->lower_bound,
+			    test_param->extra_bits);
+}
+
+static struct kunit_case ext4_inode_test_cases[] = {
+	KUNIT_CASE_PARAM(inode_test_xtimestamp_decoding, ext4_inode_gen_params),
+	{}
+};
+
+static struct kunit_suite ext4_inode_test_suite = {
+	.name = "ext4_inode_test",
+	.test_cases = ext4_inode_test_cases,
+};
+
+kunit_test_suites(&ext4_inode_test_suite);
+
+MODULE_DESCRIPTION("KUnit test of ext4 inode timestamp decoding");
+MODULE_LICENSE("GPL v2");