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, 195 insertions, 0 deletions

diff --git a/samples/bpf/fds_example.c b/samples/bpf/fds_example.c
new file mode 100644
index 000000000..88a26f3ce
--- /dev/null
+++ b/samples/bpf/fds_example.c
@@ -0,0 +1,195 @@
+#include <linux/unistd.h>
+#include <linux/bpf.h>
+
+#include <stdio.h>
+#include <stdlib.h>
+#include <stdint.h>
+#include <unistd.h>
+#include <string.h>
+#include <assert.h>
+#include <errno.h>
+
+#include <sys/types.h>
+#include <sys/socket.h>
+
+#include <bpf/bpf.h>
+
+#include <bpf/libbpf.h>
+#include "bpf_insn.h"
+#include "sock_example.h"
+#include "bpf_util.h"
+
+#define BPF_F_PIN	(1 << 0)
+#define BPF_F_GET	(1 << 1)
+#define BPF_F_PIN_GET	(BPF_F_PIN | BPF_F_GET)
+
+#define BPF_F_KEY	(1 << 2)
+#define BPF_F_VAL	(1 << 3)
+#define BPF_F_KEY_VAL	(BPF_F_KEY | BPF_F_VAL)
+
+#define BPF_M_UNSPEC	0
+#define BPF_M_MAP	1
+#define BPF_M_PROG	2
+
+char bpf_log_buf[BPF_LOG_BUF_SIZE];
+
+static void usage(void)
+{
+	printf("Usage: fds_example [...]\n");
+	printf("       -F <file>   File to pin/get object\n");
+	printf("       -P          |- pin object\n");
+	printf("       -G          `- get object\n");
+	printf("       -m          eBPF map mode\n");
+	printf("       -k <key>    |- map key\n");
+	printf("       -v <value>  `- map value\n");
+	printf("       -p          eBPF prog mode\n");
+	printf("       -o <object> `- object file\n");
+	printf("       -h          Display this help.\n");
+}
+
+static int bpf_prog_create(const char *object)
+{
+	static struct bpf_insn insns[] = {
+		BPF_MOV64_IMM(BPF_REG_0, 1),
+		BPF_EXIT_INSN(),
+	};
+	size_t insns_cnt = ARRAY_SIZE(insns);
+	struct bpf_object *obj;
+	int err;
+
+	if (object) {
+		obj = bpf_object__open_file(object, NULL);
+		assert(!libbpf_get_error(obj));
+		err = bpf_object__load(obj);
+		assert(!err);
+		return bpf_program__fd(bpf_object__next_program(obj, NULL));
+	} else {
+		LIBBPF_OPTS(bpf_prog_load_opts, opts,
+			.log_buf = bpf_log_buf,
+			.log_size = BPF_LOG_BUF_SIZE,
+		);
+
+		return bpf_prog_load(BPF_PROG_TYPE_SOCKET_FILTER, NULL, "GPL",
+				     insns, insns_cnt, &opts);
+	}
+}
+
+static int bpf_do_map(const char *file, uint32_t flags, uint32_t key,
+		      uint32_t value)
+{
+	int fd, ret;
+
+	if (flags & BPF_F_PIN) {
+		fd = bpf_map_create(BPF_MAP_TYPE_ARRAY, NULL, sizeof(uint32_t),
+				    sizeof(uint32_t), 1024, NULL);
+		printf("bpf: map fd:%d (%s)\n", fd, strerror(errno));
+		assert(fd > 0);
+
+		ret = bpf_obj_pin(fd, file);
+		printf("bpf: pin ret:(%d,%s)\n", ret, strerror(errno));
+		assert(ret == 0);
+	} else {
+		fd = bpf_obj_get(file);
+		printf("bpf: get fd:%d (%s)\n", fd, strerror(errno));
+		assert(fd > 0);
+	}
+
+	if ((flags & BPF_F_KEY_VAL) == BPF_F_KEY_VAL) {
+		ret = bpf_map_update_elem(fd, &key, &value, 0);
+		printf("bpf: fd:%d u->(%u:%u) ret:(%d,%s)\n", fd, key, value,
+		       ret, strerror(errno));
+		assert(ret == 0);
+	} else if (flags & BPF_F_KEY) {
+		ret = bpf_map_lookup_elem(fd, &key, &value);
+		printf("bpf: fd:%d l->(%u):%u ret:(%d,%s)\n", fd, key, value,
+		       ret, strerror(errno));
+		assert(ret == 0);
+	}
+
+	return 0;
+}
+
+static int bpf_do_prog(const char *file, uint32_t flags, const char *object)
+{
+	int fd, sock, ret;
+
+	if (flags & BPF_F_PIN) {
+		fd = bpf_prog_create(object);
+		printf("bpf: prog fd:%d (%s)\n", fd, strerror(errno));
+		assert(fd > 0);
+
+		ret = bpf_obj_pin(fd, file);
+		printf("bpf: pin ret:(%d,%s)\n", ret, strerror(errno));
+		assert(ret == 0);
+	} else {
+		fd = bpf_obj_get(file);
+		printf("bpf: get fd:%d (%s)\n", fd, strerror(errno));
+		assert(fd > 0);
+	}
+
+	sock = open_raw_sock("lo");
+	assert(sock > 0);
+
+	ret = setsockopt(sock, SOL_SOCKET, SO_ATTACH_BPF, &fd, sizeof(fd));
+	printf("bpf: sock:%d <- fd:%d attached ret:(%d,%s)\n", sock, fd,
+	       ret, strerror(errno));
+	assert(ret == 0);
+
+	return 0;
+}
+
+int main(int argc, char **argv)
+{
+	const char *file = NULL, *object = NULL;
+	uint32_t key = 0, value = 0, flags = 0;
+	int opt, mode = BPF_M_UNSPEC;
+
+	while ((opt = getopt(argc, argv, "F:PGmk:v:po:")) != -1) {
+		switch (opt) {
+		/* General args */
+		case 'F':
+			file = optarg;
+			break;
+		case 'P':
+			flags |= BPF_F_PIN;
+			break;
+		case 'G':
+			flags |= BPF_F_GET;
+			break;
+		/* Map-related args */
+		case 'm':
+			mode = BPF_M_MAP;
+			break;
+		case 'k':
+			key = strtoul(optarg, NULL, 0);
+			flags |= BPF_F_KEY;
+			break;
+		case 'v':
+			value = strtoul(optarg, NULL, 0);
+			flags |= BPF_F_VAL;
+			break;
+		/* Prog-related args */
+		case 'p':
+			mode = BPF_M_PROG;
+			break;
+		case 'o':
+			object = optarg;
+			break;
+		default:
+			goto out;
+		}
+	}
+
+	if (!(flags & BPF_F_PIN_GET) || !file)
+		goto out;
+
+	switch (mode) {
+	case BPF_M_MAP:
+		return bpf_do_map(file, flags, key, value);
+	case BPF_M_PROG:
+		return bpf_do_prog(file, flags, object);
+	}
+out:
+	usage();
+	return -1;
+}