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, 231 insertions, 0 deletions

diff --git a/drivers/net/netdevsim/health.c b/drivers/net/netdevsim/health.c
new file mode 100644
index 000000000..3bd0e7a48
--- /dev/null
+++ b/drivers/net/netdevsim/health.c
@@ -0,0 +1,231 @@
+// SPDX-License-Identifier: GPL-2.0
+/* Copyright (c) 2019 Mellanox Technologies. All rights reserved */
+
+#include <linux/debugfs.h>
+#include <linux/err.h>
+#include <linux/kernel.h>
+#include <linux/slab.h>
+
+#include "netdevsim.h"
+
+static int
+nsim_dev_empty_reporter_dump(struct devlink_health_reporter *reporter,
+			     struct devlink_fmsg *fmsg, void *priv_ctx,
+			     struct netlink_ext_ack *extack)
+{
+	return 0;
+}
+
+static int
+nsim_dev_empty_reporter_diagnose(struct devlink_health_reporter *reporter,
+				 struct devlink_fmsg *fmsg,
+				 struct netlink_ext_ack *extack)
+{
+	return 0;
+}
+
+static const
+struct devlink_health_reporter_ops nsim_dev_empty_reporter_ops = {
+	.name = "empty",
+	.dump = nsim_dev_empty_reporter_dump,
+	.diagnose = nsim_dev_empty_reporter_diagnose,
+};
+
+struct nsim_dev_dummy_reporter_ctx {
+	char *break_msg;
+};
+
+static int
+nsim_dev_dummy_reporter_recover(struct devlink_health_reporter *reporter,
+				void *priv_ctx,
+				struct netlink_ext_ack *extack)
+{
+	struct nsim_dev_health *health = devlink_health_reporter_priv(reporter);
+	struct nsim_dev_dummy_reporter_ctx *ctx = priv_ctx;
+
+	if (health->fail_recover) {
+		/* For testing purposes, user set debugfs fail_recover
+		 * value to true. Fail right away.
+		 */
+		NL_SET_ERR_MSG_MOD(extack, "User setup the recover to fail for testing purposes");
+		return -EINVAL;
+	}
+	if (ctx) {
+		kfree(health->recovered_break_msg);
+		health->recovered_break_msg = kstrdup(ctx->break_msg,
+						      GFP_KERNEL);
+		if (!health->recovered_break_msg)
+			return -ENOMEM;
+	}
+	return 0;
+}
+
+static int nsim_dev_dummy_fmsg_put(struct devlink_fmsg *fmsg, u32 binary_len)
+{
+	char *binary;
+	int i;
+
+	devlink_fmsg_bool_pair_put(fmsg, "test_bool", true);
+	devlink_fmsg_u8_pair_put(fmsg, "test_u8", 1);
+	devlink_fmsg_u32_pair_put(fmsg, "test_u32", 3);
+	devlink_fmsg_u64_pair_put(fmsg, "test_u64", 4);
+	devlink_fmsg_string_pair_put(fmsg, "test_string", "somestring");
+
+	binary = kmalloc(binary_len, GFP_KERNEL | __GFP_NOWARN);
+	if (!binary)
+		return -ENOMEM;
+	get_random_bytes(binary, binary_len);
+	devlink_fmsg_binary_pair_put(fmsg, "test_binary", binary, binary_len);
+	kfree(binary);
+
+	devlink_fmsg_pair_nest_start(fmsg, "test_nest");
+	devlink_fmsg_obj_nest_start(fmsg);
+	devlink_fmsg_bool_pair_put(fmsg, "nested_test_bool", false);
+	devlink_fmsg_u8_pair_put(fmsg, "nested_test_u8", false);
+	devlink_fmsg_obj_nest_end(fmsg);
+	devlink_fmsg_pair_nest_end(fmsg);
+	devlink_fmsg_arr_pair_nest_end(fmsg);
+	devlink_fmsg_arr_pair_nest_start(fmsg, "test_u32_array");
+
+	for (i = 0; i < 10; i++)
+		devlink_fmsg_u32_put(fmsg, i);
+	devlink_fmsg_arr_pair_nest_end(fmsg);
+	devlink_fmsg_arr_pair_nest_start(fmsg, "test_array_of_objects");
+
+	for (i = 0; i < 10; i++) {
+		devlink_fmsg_obj_nest_start(fmsg);
+		devlink_fmsg_bool_pair_put(fmsg, "in_array_nested_test_bool",
+					   false);
+		devlink_fmsg_u8_pair_put(fmsg, "in_array_nested_test_u8", i);
+		devlink_fmsg_obj_nest_end(fmsg);
+	}
+	devlink_fmsg_arr_pair_nest_end(fmsg);
+
+	return 0;
+}
+
+static int
+nsim_dev_dummy_reporter_dump(struct devlink_health_reporter *reporter,
+			     struct devlink_fmsg *fmsg, void *priv_ctx,
+			     struct netlink_ext_ack *extack)
+{
+	struct nsim_dev_health *health = devlink_health_reporter_priv(reporter);
+	struct nsim_dev_dummy_reporter_ctx *ctx = priv_ctx;
+
+	if (ctx)
+		devlink_fmsg_string_pair_put(fmsg, "break_message", ctx->break_msg);
+
+	return nsim_dev_dummy_fmsg_put(fmsg, health->binary_len);
+}
+
+static int
+nsim_dev_dummy_reporter_diagnose(struct devlink_health_reporter *reporter,
+				 struct devlink_fmsg *fmsg,
+				 struct netlink_ext_ack *extack)
+{
+	struct nsim_dev_health *health = devlink_health_reporter_priv(reporter);
+
+	if (health->recovered_break_msg)
+		devlink_fmsg_string_pair_put(fmsg, "recovered_break_message",
+					     health->recovered_break_msg);
+
+	return nsim_dev_dummy_fmsg_put(fmsg, health->binary_len);
+}
+
+static const
+struct devlink_health_reporter_ops nsim_dev_dummy_reporter_ops = {
+	.name = "dummy",
+	.recover = nsim_dev_dummy_reporter_recover,
+	.dump = nsim_dev_dummy_reporter_dump,
+	.diagnose = nsim_dev_dummy_reporter_diagnose,
+};
+
+static ssize_t nsim_dev_health_break_write(struct file *file,
+					   const char __user *data,
+					   size_t count, loff_t *ppos)
+{
+	struct nsim_dev_health *health = file->private_data;
+	struct nsim_dev_dummy_reporter_ctx ctx;
+	char *break_msg;
+	int err;
+
+	if (count == 0 || count > PAGE_SIZE)
+		return -EINVAL;
+	break_msg = memdup_user_nul(data, count);
+	if (IS_ERR(break_msg))
+		return PTR_ERR(break_msg);
+
+	if (break_msg[count - 1] == '\n')
+		break_msg[count - 1] = '\0';
+
+	ctx.break_msg = break_msg;
+	err = devlink_health_report(health->dummy_reporter, break_msg, &ctx);
+	if (err)
+		goto out;
+
+out:
+	kfree(break_msg);
+	return err ?: count;
+}
+
+static const struct file_operations nsim_dev_health_break_fops = {
+	.open = simple_open,
+	.write = nsim_dev_health_break_write,
+	.llseek = generic_file_llseek,
+	.owner = THIS_MODULE,
+};
+
+int nsim_dev_health_init(struct nsim_dev *nsim_dev, struct devlink *devlink)
+{
+	struct nsim_dev_health *health = &nsim_dev->health;
+	int err;
+
+	health->empty_reporter =
+		devl_health_reporter_create(devlink,
+					    &nsim_dev_empty_reporter_ops,
+					    health);
+	if (IS_ERR(health->empty_reporter))
+		return PTR_ERR(health->empty_reporter);
+
+	health->dummy_reporter =
+		devl_health_reporter_create(devlink,
+					    &nsim_dev_dummy_reporter_ops,
+					    health);
+	if (IS_ERR(health->dummy_reporter)) {
+		err = PTR_ERR(health->dummy_reporter);
+		goto err_empty_reporter_destroy;
+	}
+
+	health->ddir = debugfs_create_dir("health", nsim_dev->ddir);
+	if (IS_ERR(health->ddir)) {
+		err = PTR_ERR(health->ddir);
+		goto err_dummy_reporter_destroy;
+	}
+
+	health->recovered_break_msg = NULL;
+	debugfs_create_file("break_health", 0200, health->ddir, health,
+			    &nsim_dev_health_break_fops);
+	health->binary_len = 16;
+	debugfs_create_u32("binary_len", 0600, health->ddir,
+			   &health->binary_len);
+	health->fail_recover = false;
+	debugfs_create_bool("fail_recover", 0600, health->ddir,
+			    &health->fail_recover);
+	return 0;
+
+err_dummy_reporter_destroy:
+	devl_health_reporter_destroy(health->dummy_reporter);
+err_empty_reporter_destroy:
+	devl_health_reporter_destroy(health->empty_reporter);
+	return err;
+}
+
+void nsim_dev_health_exit(struct nsim_dev *nsim_dev)
+{
+	struct nsim_dev_health *health = &nsim_dev->health;
+
+	debugfs_remove_recursive(health->ddir);
+	kfree(health->recovered_break_msg);
+	devl_health_reporter_destroy(health->dummy_reporter);
+	devl_health_reporter_destroy(health->empty_reporter);
+}