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, 189 insertions, 0 deletions

diff --git a/arch/sh/mm/alignment.c b/arch/sh/mm/alignment.c
new file mode 100644
index 000000000..3a76a766f
--- /dev/null
+++ b/arch/sh/mm/alignment.c
@@ -0,0 +1,189 @@
+/*
+ * Alignment access counters and corresponding user-space interfaces.
+ *
+ * Copyright (C) 2009 ST Microelectronics
+ * Copyright (C) 2009 - 2010 Paul Mundt
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ */
+#include <linux/module.h>
+#include <linux/kernel.h>
+#include <linux/seq_file.h>
+#include <linux/proc_fs.h>
+#include <linux/uaccess.h>
+#include <linux/ratelimit.h>
+#include <asm/alignment.h>
+#include <asm/processor.h>
+
+static unsigned long se_user;
+static unsigned long se_sys;
+static unsigned long se_half;
+static unsigned long se_word;
+static unsigned long se_dword;
+static unsigned long se_multi;
+/* bitfield: 1: warn 2: fixup 4: signal -> combinations 2|4 && 1|2|4 are not
+   valid! */
+static int se_usermode = UM_WARN | UM_FIXUP;
+/* 0: no warning 1: print a warning message, disabled by default */
+static int se_kernmode_warn;
+
+core_param(alignment, se_usermode, int, 0600);
+
+void inc_unaligned_byte_access(void)
+{
+	se_half++;
+}
+
+void inc_unaligned_word_access(void)
+{
+	se_word++;
+}
+
+void inc_unaligned_dword_access(void)
+{
+	se_dword++;
+}
+
+void inc_unaligned_multi_access(void)
+{
+	se_multi++;
+}
+
+void inc_unaligned_user_access(void)
+{
+	se_user++;
+}
+
+void inc_unaligned_kernel_access(void)
+{
+	se_sys++;
+}
+
+/*
+ * This defaults to the global policy which can be set from the command
+ * line, while processes can overload their preferences via prctl().
+ */
+unsigned int unaligned_user_action(void)
+{
+	unsigned int action = se_usermode;
+
+	if (current->thread.flags & SH_THREAD_UAC_SIGBUS) {
+		action &= ~UM_FIXUP;
+		action |= UM_SIGNAL;
+	}
+
+	if (current->thread.flags & SH_THREAD_UAC_NOPRINT)
+		action &= ~UM_WARN;
+
+	return action;
+}
+
+int get_unalign_ctl(struct task_struct *tsk, unsigned long addr)
+{
+	return put_user(tsk->thread.flags & SH_THREAD_UAC_MASK,
+			(unsigned int __user *)addr);
+}
+
+int set_unalign_ctl(struct task_struct *tsk, unsigned int val)
+{
+	tsk->thread.flags = (tsk->thread.flags & ~SH_THREAD_UAC_MASK) |
+			    (val & SH_THREAD_UAC_MASK);
+	return 0;
+}
+
+void unaligned_fixups_notify(struct task_struct *tsk, insn_size_t insn,
+			     struct pt_regs *regs)
+{
+	if (user_mode(regs) && (se_usermode & UM_WARN))
+		pr_notice_ratelimited("Fixing up unaligned userspace access "
+			  "in \"%s\" pid=%d pc=0x%p ins=0x%04hx\n",
+			  tsk->comm, task_pid_nr(tsk),
+			  (void *)instruction_pointer(regs), insn);
+	else if (se_kernmode_warn)
+		pr_notice_ratelimited("Fixing up unaligned kernel access "
+			  "in \"%s\" pid=%d pc=0x%p ins=0x%04hx\n",
+			  tsk->comm, task_pid_nr(tsk),
+			  (void *)instruction_pointer(regs), insn);
+}
+
+static const char *se_usermode_action[] = {
+	"ignored",
+	"warn",
+	"fixup",
+	"fixup+warn",
+	"signal",
+	"signal+warn"
+};
+
+static int alignment_proc_show(struct seq_file *m, void *v)
+{
+	seq_printf(m, "User:\t\t%lu\n", se_user);
+	seq_printf(m, "System:\t\t%lu\n", se_sys);
+	seq_printf(m, "Half:\t\t%lu\n", se_half);
+	seq_printf(m, "Word:\t\t%lu\n", se_word);
+	seq_printf(m, "DWord:\t\t%lu\n", se_dword);
+	seq_printf(m, "Multi:\t\t%lu\n", se_multi);
+	seq_printf(m, "User faults:\t%i (%s)\n", se_usermode,
+			se_usermode_action[se_usermode]);
+	seq_printf(m, "Kernel faults:\t%i (fixup%s)\n", se_kernmode_warn,
+			se_kernmode_warn ? "+warn" : "");
+	return 0;
+}
+
+static int alignment_proc_open(struct inode *inode, struct file *file)
+{
+	return single_open(file, alignment_proc_show, NULL);
+}
+
+static ssize_t alignment_proc_write(struct file *file,
+		const char __user *buffer, size_t count, loff_t *pos)
+{
+	int *data = pde_data(file_inode(file));
+	char mode;
+
+	if (count > 0) {
+		if (get_user(mode, buffer))
+			return -EFAULT;
+		if (mode >= '0' && mode <= '5')
+			*data = mode - '0';
+	}
+	return count;
+}
+
+static const struct proc_ops alignment_proc_ops = {
+	.proc_open	= alignment_proc_open,
+	.proc_read	= seq_read,
+	.proc_lseek	= seq_lseek,
+	.proc_release	= single_release,
+	.proc_write	= alignment_proc_write,
+};
+
+/*
+ * This needs to be done after sysctl_init_bases(), otherwise sys/ will be
+ * overwritten.  Actually, this shouldn't be in sys/ at all since
+ * it isn't a sysctl, and it doesn't contain sysctl information.
+ * We now locate it in /proc/cpu/alignment instead.
+ */
+static int __init alignment_init(void)
+{
+	struct proc_dir_entry *dir, *res;
+
+	dir = proc_mkdir("cpu", NULL);
+	if (!dir)
+		return -ENOMEM;
+
+	res = proc_create_data("alignment", S_IWUSR | S_IRUGO, dir,
+			       &alignment_proc_ops, &se_usermode);
+	if (!res)
+		return -ENOMEM;
+
+        res = proc_create_data("kernel_alignment", S_IWUSR | S_IRUGO, dir,
+			       &alignment_proc_ops, &se_kernmode_warn);
+        if (!res)
+                return -ENOMEM;
+
+	return 0;
+}
+fs_initcall(alignment_init);