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, 150 insertions, 0 deletions

diff --git a/arch/um/os-Linux/execvp.c b/arch/um/os-Linux/execvp.c
new file mode 100644
index 000000000..c09a5fd5e
--- /dev/null
+++ b/arch/um/os-Linux/execvp.c
@@ -0,0 +1,150 @@
+/* Copyright (C) 2006 by Paolo Giarrusso - modified from glibc' execvp.c.
+   Original copyright notice follows:
+
+   Copyright (C) 1991,92,1995-99,2002,2004 Free Software Foundation, Inc.
+   This file is part of the GNU C Library.
+
+   The GNU C Library is free software; you can redistribute it and/or
+   modify it under the terms of the GNU Lesser General Public
+   License as published by the Free Software Foundation; either
+   version 2.1 of the License, or (at your option) any later version.
+
+   The GNU C Library is distributed in the hope that it will be useful,
+   but WITHOUT ANY WARRANTY; without even the implied warranty of
+   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+   Lesser General Public License for more details.
+
+   You should have received a copy of the GNU Lesser General Public
+   License along with the GNU C Library; if not, write to the Free
+   Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+   02111-1307 USA.  */
+#include <unistd.h>
+
+#include <stdbool.h>
+#include <stdlib.h>
+#include <string.h>
+#include <errno.h>
+#include <limits.h>
+
+#ifndef TEST
+#include <um_malloc.h>
+#else
+#include <stdio.h>
+#define um_kmalloc malloc
+#endif
+#include <os.h>
+
+/* Execute FILE, searching in the `PATH' environment variable if it contains
+   no slashes, with arguments ARGV and environment from `environ'.  */
+int execvp_noalloc(char *buf, const char *file, char *const argv[])
+{
+	if (*file == '\0') {
+		return -ENOENT;
+	}
+
+	if (strchr (file, '/') != NULL) {
+		/* Don't search when it contains a slash.  */
+		execv(file, argv);
+	} else {
+		int got_eacces;
+		size_t len, pathlen;
+		char *name, *p;
+		char *path = getenv("PATH");
+		if (path == NULL)
+			path = ":/bin:/usr/bin";
+
+		len = strlen(file) + 1;
+		pathlen = strlen(path);
+		/* Copy the file name at the top.  */
+		name = memcpy(buf + pathlen + 1, file, len);
+		/* And add the slash.  */
+		*--name = '/';
+
+		got_eacces = 0;
+		p = path;
+		do {
+			char *startp;
+
+			path = p;
+			//Let's avoid this GNU extension.
+			//p = strchrnul (path, ':');
+			p = strchr(path, ':');
+			if (!p)
+				p = strchr(path, '\0');
+
+			if (p == path)
+				/* Two adjacent colons, or a colon at the beginning or the end
+				   of `PATH' means to search the current directory.  */
+				startp = name + 1;
+			else
+				startp = memcpy(name - (p - path), path, p - path);
+
+			/* Try to execute this name.  If it works, execv will not return.  */
+			execv(startp, argv);
+
+			/*
+			if (errno == ENOEXEC) {
+			}
+			*/
+
+			switch (errno) {
+				case EACCES:
+					/* Record the we got a `Permission denied' error.  If we end
+					   up finding no executable we can use, we want to diagnose
+					   that we did find one but were denied access.  */
+					got_eacces = 1;
+					break;
+				case ENOENT:
+				case ESTALE:
+				case ENOTDIR:
+					/* Those errors indicate the file is missing or not executable
+					   by us, in which case we want to just try the next path
+					   directory.  */
+				case ENODEV:
+				case ETIMEDOUT:
+					/* Some strange filesystems like AFS return even
+					   stranger error numbers.  They cannot reasonably mean
+					   anything else so ignore those, too.  */
+				case ENOEXEC:
+					/* We won't go searching for the shell
+					 * if it is not executable - the Linux
+					 * kernel already handles this enough,
+					 * for us. */
+					break;
+
+				default:
+					/* Some other error means we found an executable file, but
+					   something went wrong executing it; return the error to our
+					   caller.  */
+					return -errno;
+			}
+		} while (*p++ != '\0');
+
+		/* We tried every element and none of them worked.  */
+		if (got_eacces)
+			/* At least one failure was due to permissions, so report that
+			   error.  */
+			return -EACCES;
+	}
+
+	/* Return the error from the last attempt (probably ENOENT).  */
+	return -errno;
+}
+#ifdef TEST
+int main(int argc, char**argv)
+{
+	char buf[PATH_MAX];
+	int ret;
+	argc--;
+	if (!argc) {
+		os_warn("Not enough arguments\n");
+		return 1;
+	}
+	argv++;
+	if (ret = execvp_noalloc(buf, argv[0], argv)) {
+		errno = -ret;
+		perror("execvp_noalloc");
+	}
+	return 0;
+}
+#endif