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, 141 insertions, 0 deletions

diff --git a/drivers/accessibility/speakup/selection.c b/drivers/accessibility/speakup/selection.c
new file mode 100644
index 000000000..1713ce4e0
--- /dev/null
+++ b/drivers/accessibility/speakup/selection.c
@@ -0,0 +1,141 @@
+// SPDX-License-Identifier: GPL-2.0
+#include <linux/slab.h> /* for kmalloc */
+#include <linux/consolemap.h>
+#include <linux/interrupt.h>
+#include <linux/sched.h>
+#include <linux/device.h> /* for dev_warn */
+#include <linux/selection.h>
+#include <linux/workqueue.h>
+#include <linux/tty.h>
+#include <linux/tty_flip.h>
+#include <linux/atomic.h>
+#include <linux/console.h>
+
+#include "speakup.h"
+
+u16 spk_xs, spk_ys, spk_xe, spk_ye; /* our region points */
+struct vc_data *spk_sel_cons;
+
+struct speakup_selection_work {
+	struct work_struct work;
+	struct tiocl_selection sel;
+	struct tty_struct *tty;
+};
+
+static void __speakup_set_selection(struct work_struct *work)
+{
+	struct speakup_selection_work *ssw =
+		container_of(work, struct speakup_selection_work, work);
+
+	struct tty_struct *tty;
+	struct tiocl_selection sel;
+
+	sel = ssw->sel;
+
+	/* this ensures we copy sel before releasing the lock below */
+	rmb();
+
+	/* release the lock by setting tty of the struct to NULL */
+	tty = xchg(&ssw->tty, NULL);
+
+	if (spk_sel_cons != vc_cons[fg_console].d) {
+		spk_sel_cons = vc_cons[fg_console].d;
+		pr_warn("Selection: mark console not the same as cut\n");
+		goto unref;
+	}
+
+	console_lock();
+	clear_selection();
+	console_unlock();
+
+	set_selection_kernel(&sel, tty);
+
+unref:
+	tty_kref_put(tty);
+}
+
+static struct speakup_selection_work speakup_sel_work = {
+	.work = __WORK_INITIALIZER(speakup_sel_work.work,
+				   __speakup_set_selection)
+};
+
+int speakup_set_selection(struct tty_struct *tty)
+{
+	/* we get kref here first in order to avoid a subtle race when
+	 * cancelling selection work. getting kref first establishes the
+	 * invariant that if speakup_sel_work.tty is not NULL when
+	 * speakup_cancel_selection() is called, it must be the case that a put
+	 * kref is pending.
+	 */
+	tty_kref_get(tty);
+	if (cmpxchg(&speakup_sel_work.tty, NULL, tty)) {
+		tty_kref_put(tty);
+		return -EBUSY;
+	}
+	/* now we have the 'lock' by setting tty member of
+	 * speakup_selection_work. wmb() ensures that writes to
+	 * speakup_sel_work don't happen before cmpxchg() above.
+	 */
+	wmb();
+
+	speakup_sel_work.sel.xs = spk_xs + 1;
+	speakup_sel_work.sel.ys = spk_ys + 1;
+	speakup_sel_work.sel.xe = spk_xe + 1;
+	speakup_sel_work.sel.ye = spk_ye + 1;
+	speakup_sel_work.sel.sel_mode = TIOCL_SELCHAR;
+
+	schedule_work_on(WORK_CPU_UNBOUND, &speakup_sel_work.work);
+
+	return 0;
+}
+
+void speakup_cancel_selection(void)
+{
+	struct tty_struct *tty;
+
+	cancel_work_sync(&speakup_sel_work.work);
+	/* setting to null so that if work fails to run and we cancel it,
+	 * we can run it again without getting EBUSY forever from there on.
+	 * we need to use xchg here to avoid race with speakup_set_selection()
+	 */
+	tty = xchg(&speakup_sel_work.tty, NULL);
+	if (tty)
+		tty_kref_put(tty);
+}
+
+static void __speakup_paste_selection(struct work_struct *work)
+{
+	struct speakup_selection_work *ssw =
+		container_of(work, struct speakup_selection_work, work);
+	struct tty_struct *tty = xchg(&ssw->tty, NULL);
+
+	paste_selection(tty);
+	tty_kref_put(tty);
+}
+
+static struct speakup_selection_work speakup_paste_work = {
+	.work = __WORK_INITIALIZER(speakup_paste_work.work,
+				   __speakup_paste_selection)
+};
+
+int speakup_paste_selection(struct tty_struct *tty)
+{
+	tty_kref_get(tty);
+	if (cmpxchg(&speakup_paste_work.tty, NULL, tty)) {
+		tty_kref_put(tty);
+		return -EBUSY;
+	}
+
+	schedule_work_on(WORK_CPU_UNBOUND, &speakup_paste_work.work);
+	return 0;
+}
+
+void speakup_cancel_paste(void)
+{
+	struct tty_struct *tty;
+
+	cancel_work_sync(&speakup_paste_work.work);
+	tty = xchg(&speakup_paste_work.tty, NULL);
+	if (tty)
+		tty_kref_put(tty);
+}