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, 158 insertions, 0 deletions

diff --git a/lib/lwq.c b/lib/lwq.c
new file mode 100644
index 000000000..c1e11ba6f
--- /dev/null
+++ b/lib/lwq.c
@@ -0,0 +1,158 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Light-weight single-linked queue.
+ *
+ * Entries are enqueued to the head of an llist, with no blocking.
+ * This can happen in any context.
+ *
+ * Entries are dequeued using a spinlock to protect against multiple
+ * access.  The llist is staged in reverse order, and refreshed
+ * from the llist when it exhausts.
+ *
+ * This is particularly suitable when work items are queued in BH or
+ * IRQ context, and where work items are handled one at a time by
+ * dedicated threads.
+ */
+#include <linux/rcupdate.h>
+#include <linux/lwq.h>
+
+struct llist_node *__lwq_dequeue(struct lwq *q)
+{
+	struct llist_node *this;
+
+	if (lwq_empty(q))
+		return NULL;
+	spin_lock(&q->lock);
+	this = q->ready;
+	if (!this && !llist_empty(&q->new)) {
+		/* ensure queue doesn't appear transiently lwq_empty */
+		smp_store_release(&q->ready, (void *)1);
+		this = llist_reverse_order(llist_del_all(&q->new));
+		if (!this)
+			q->ready = NULL;
+	}
+	if (this)
+		q->ready = llist_next(this);
+	spin_unlock(&q->lock);
+	return this;
+}
+EXPORT_SYMBOL_GPL(__lwq_dequeue);
+
+/**
+ * lwq_dequeue_all - dequeue all currently enqueued objects
+ * @q:	the queue to dequeue from
+ *
+ * Remove and return a linked list of llist_nodes of all the objects that were
+ * in the queue. The first on the list will be the object that was least
+ * recently enqueued.
+ */
+struct llist_node *lwq_dequeue_all(struct lwq *q)
+{
+	struct llist_node *r, *t, **ep;
+
+	if (lwq_empty(q))
+		return NULL;
+
+	spin_lock(&q->lock);
+	r = q->ready;
+	q->ready = NULL;
+	t = llist_del_all(&q->new);
+	spin_unlock(&q->lock);
+	ep = &r;
+	while (*ep)
+		ep = &(*ep)->next;
+	*ep = llist_reverse_order(t);
+	return r;
+}
+EXPORT_SYMBOL_GPL(lwq_dequeue_all);
+
+#if IS_ENABLED(CONFIG_LWQ_TEST)
+
+#include <linux/module.h>
+#include <linux/slab.h>
+#include <linux/wait_bit.h>
+#include <linux/kthread.h>
+#include <linux/delay.h>
+struct tnode {
+	struct lwq_node n;
+	int i;
+	int c;
+};
+
+static int lwq_exercise(void *qv)
+{
+	struct lwq *q = qv;
+	int cnt;
+	struct tnode *t;
+
+	for (cnt = 0; cnt < 10000; cnt++) {
+		wait_var_event(q, (t = lwq_dequeue(q, struct tnode, n)) != NULL);
+		t->c++;
+		if (lwq_enqueue(&t->n, q))
+			wake_up_var(q);
+	}
+	while (!kthread_should_stop())
+		schedule_timeout_idle(1);
+	return 0;
+}
+
+static int lwq_test(void)
+{
+	int i;
+	struct lwq q;
+	struct llist_node *l, **t1, *t2;
+	struct tnode *t;
+	struct task_struct *threads[8];
+
+	printk(KERN_INFO "testing lwq....\n");
+	lwq_init(&q);
+	printk(KERN_INFO " lwq: run some threads\n");
+	for (i = 0; i < ARRAY_SIZE(threads); i++)
+		threads[i] = kthread_run(lwq_exercise, &q, "lwq-test-%d", i);
+	for (i = 0; i < 100; i++) {
+		t = kmalloc_obj(*t);
+		if (!t)
+			break;
+		t->i = i;
+		t->c = 0;
+		if (lwq_enqueue(&t->n, &q))
+			wake_up_var(&q);
+	}
+	/* wait for threads to exit */
+	for (i = 0; i < ARRAY_SIZE(threads); i++)
+		if (!IS_ERR_OR_NULL(threads[i]))
+			kthread_stop(threads[i]);
+	printk(KERN_INFO " lwq: dequeue first 50:");
+	for (i = 0; i < 50 ; i++) {
+		if (i && (i % 10) == 0) {
+			printk(KERN_CONT "\n");
+			printk(KERN_INFO " lwq: ... ");
+		}
+		t = lwq_dequeue(&q, struct tnode, n);
+		if (t)
+			printk(KERN_CONT " %d(%d)", t->i, t->c);
+		kfree(t);
+	}
+	printk(KERN_CONT "\n");
+	l = lwq_dequeue_all(&q);
+	printk(KERN_INFO " lwq: delete the multiples of 3 (test lwq_for_each_safe())\n");
+	lwq_for_each_safe(t, t1, t2, &l, n) {
+		if ((t->i % 3) == 0) {
+			t->i = -1;
+			kfree(t);
+			t = NULL;
+		}
+	}
+	if (l)
+		lwq_enqueue_batch(l, &q);
+	printk(KERN_INFO " lwq: dequeue remaining:");
+	while ((t = lwq_dequeue(&q, struct tnode, n)) != NULL) {
+		printk(KERN_CONT " %d", t->i);
+		kfree(t);
+	}
+	printk(KERN_CONT "\n");
+	return 0;
+}
+
+module_init(lwq_test);
+#endif /* CONFIG_LWQ_TEST*/