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, 222 insertions, 0 deletions

diff --git a/drivers/dma/idxd/submit.c b/drivers/dma/idxd/submit.c
new file mode 100644
index 000000000..03217041b
--- /dev/null
+++ b/drivers/dma/idxd/submit.c
@@ -0,0 +1,222 @@
+// SPDX-License-Identifier: GPL-2.0
+/* Copyright(c) 2019 Intel Corporation. All rights rsvd. */
+#include <linux/init.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <uapi/linux/idxd.h>
+#include "idxd.h"
+#include "registers.h"
+
+static struct idxd_desc *__get_desc(struct idxd_wq *wq, int idx, int cpu)
+{
+	struct idxd_desc *desc;
+	struct idxd_device *idxd = wq->idxd;
+
+	desc = wq->descs[idx];
+	memset(desc->hw, 0, sizeof(struct dsa_hw_desc));
+	memset(desc->completion, 0, idxd->data->compl_size);
+	desc->cpu = cpu;
+
+	if (device_pasid_enabled(idxd))
+		desc->hw->pasid = idxd->pasid;
+
+	return desc;
+}
+
+struct idxd_desc *idxd_alloc_desc(struct idxd_wq *wq, enum idxd_op_type optype)
+{
+	int cpu, idx;
+	struct idxd_device *idxd = wq->idxd;
+	DEFINE_SBQ_WAIT(wait);
+	struct sbq_wait_state *ws;
+	struct sbitmap_queue *sbq;
+
+	if (idxd->state != IDXD_DEV_ENABLED)
+		return ERR_PTR(-EIO);
+
+	sbq = &wq->sbq;
+	idx = sbitmap_queue_get(sbq, &cpu);
+	if (idx < 0) {
+		if (optype == IDXD_OP_NONBLOCK)
+			return ERR_PTR(-EAGAIN);
+	} else {
+		return __get_desc(wq, idx, cpu);
+	}
+
+	ws = &sbq->ws[0];
+	for (;;) {
+		sbitmap_prepare_to_wait(sbq, ws, &wait, TASK_INTERRUPTIBLE);
+		if (signal_pending_state(TASK_INTERRUPTIBLE, current))
+			break;
+		idx = sbitmap_queue_get(sbq, &cpu);
+		if (idx >= 0)
+			break;
+		schedule();
+	}
+
+	sbitmap_finish_wait(sbq, ws, &wait);
+	if (idx < 0)
+		return ERR_PTR(-EAGAIN);
+
+	return __get_desc(wq, idx, cpu);
+}
+EXPORT_SYMBOL_NS_GPL(idxd_alloc_desc, "IDXD");
+
+void idxd_free_desc(struct idxd_wq *wq, struct idxd_desc *desc)
+{
+	int cpu = desc->cpu;
+
+	desc->cpu = -1;
+	sbitmap_queue_clear(&wq->sbq, desc->id, cpu);
+}
+EXPORT_SYMBOL_NS_GPL(idxd_free_desc, "IDXD");
+
+static struct idxd_desc *list_abort_desc(struct idxd_wq *wq, struct idxd_irq_entry *ie,
+					 struct idxd_desc *desc)
+{
+	struct idxd_desc *d, *n;
+
+	lockdep_assert_held(&ie->list_lock);
+	list_for_each_entry_safe(d, n, &ie->work_list, list) {
+		if (d == desc) {
+			list_del(&d->list);
+			return d;
+		}
+	}
+
+	/*
+	 * At this point, the desc needs to be aborted is held by the completion
+	 * handler where it has taken it off the pending list but has not added to the
+	 * work list. It will be cleaned up by the interrupt handler when it sees the
+	 * IDXD_COMP_DESC_ABORT for completion status.
+	 */
+	return NULL;
+}
+
+static void llist_abort_desc(struct idxd_wq *wq, struct idxd_irq_entry *ie,
+			     struct idxd_desc *desc)
+{
+	struct idxd_desc *d, *t, *found = NULL;
+	struct llist_node *head;
+	LIST_HEAD(flist);
+
+	desc->completion->status = IDXD_COMP_DESC_ABORT;
+	/*
+	 * Grab the list lock so it will block the irq thread handler. This allows the
+	 * abort code to locate the descriptor need to be aborted.
+	 */
+	spin_lock(&ie->list_lock);
+	head = llist_del_all(&ie->pending_llist);
+	if (head) {
+		llist_for_each_entry_safe(d, t, head, llnode) {
+			if (d == desc) {
+				found = desc;
+				continue;
+			}
+
+			if (d->completion->status)
+				list_add_tail(&d->list, &flist);
+			else
+				list_add_tail(&d->list, &ie->work_list);
+		}
+	}
+
+	if (!found)
+		found = list_abort_desc(wq, ie, desc);
+	spin_unlock(&ie->list_lock);
+
+	if (found)
+		idxd_dma_complete_txd(found, IDXD_COMPLETE_ABORT, false,
+				      NULL, NULL);
+
+	/*
+	 * completing the descriptor will return desc to allocator and
+	 * the desc can be acquired by a different process and the
+	 * desc->list can be modified.  Delete desc from list so the
+	 * list traversing does not get corrupted by the other process.
+	 */
+	list_for_each_entry_safe(d, t, &flist, list) {
+		list_del_init(&d->list);
+		idxd_dma_complete_txd(d, IDXD_COMPLETE_ABORT, true,
+				      NULL, NULL);
+	}
+}
+
+/*
+ * ENQCMDS typically fail when the WQ is inactive or busy. On host submission, the driver
+ * has better control of number of descriptors being submitted to a shared wq by limiting
+ * the number of driver allocated descriptors to the wq size. However, when the swq is
+ * exported to a guest kernel, it may be shared with multiple guest kernels. This means
+ * the likelihood of getting busy returned on the swq when submitting goes significantly up.
+ * Having a tunable retry mechanism allows the driver to keep trying for a bit before giving
+ * up. The sysfs knob can be tuned by the system administrator.
+ */
+int idxd_enqcmds(struct idxd_wq *wq, void __iomem *portal, const void *desc)
+{
+	unsigned int retries = wq->enqcmds_retries;
+	int rc;
+
+	do {
+		rc = enqcmds(portal, desc);
+		if (rc == 0)
+			break;
+		cpu_relax();
+	} while (retries--);
+
+	return rc;
+}
+
+int idxd_submit_desc(struct idxd_wq *wq, struct idxd_desc *desc)
+{
+	struct idxd_device *idxd = wq->idxd;
+	struct idxd_irq_entry *ie = NULL;
+	u32 desc_flags = desc->hw->flags;
+	void __iomem *portal;
+	int rc;
+
+	if (idxd->state != IDXD_DEV_ENABLED)
+		return -EIO;
+
+	if (!percpu_ref_tryget_live(&wq->wq_active)) {
+		wait_for_completion(&wq->wq_resurrect);
+		if (!percpu_ref_tryget_live(&wq->wq_active))
+			return -ENXIO;
+	}
+
+	portal = idxd_wq_portal_addr(wq);
+
+	/*
+	 * Pending the descriptor to the lockless list for the irq_entry
+	 * that we designated the descriptor to.
+	 */
+	if (desc_flags & IDXD_OP_FLAG_RCI) {
+		ie = &wq->ie;
+		desc->hw->int_handle = ie->int_handle;
+		llist_add(&desc->llnode, &ie->pending_llist);
+	}
+
+	/*
+	 * The wmb() flushes writes to coherent DMA data before
+	 * possibly triggering a DMA read. The wmb() is necessary
+	 * even on UP because the recipient is a device.
+	 */
+	wmb();
+
+	if (wq_dedicated(wq)) {
+		iosubmit_cmds512(portal, desc->hw, 1);
+	} else {
+		rc = idxd_enqcmds(wq, portal, desc->hw);
+		if (rc < 0) {
+			percpu_ref_put(&wq->wq_active);
+			/* abort operation frees the descriptor */
+			if (ie)
+				llist_abort_desc(wq, ie, desc);
+			return rc;
+		}
+	}
+
+	percpu_ref_put(&wq->wq_active);
+	return 0;
+}
+EXPORT_SYMBOL_NS_GPL(idxd_submit_desc, "IDXD");