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, 305 insertions, 0 deletions

diff --git a/net/nfc/nci/data.c b/net/nfc/nci/data.c
new file mode 100644
index 000000000..4253edea5
--- /dev/null
+++ b/net/nfc/nci/data.c
@@ -0,0 +1,305 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ *  The NFC Controller Interface is the communication protocol between an
+ *  NFC Controller (NFCC) and a Device Host (DH).
+ *
+ *  Copyright (C) 2011 Texas Instruments, Inc.
+ *  Copyright (C) 2014 Marvell International Ltd.
+ *
+ *  Written by Ilan Elias <ilane@ti.com>
+ */
+
+#define pr_fmt(fmt) KBUILD_MODNAME ": %s: " fmt, __func__
+
+#include <linux/types.h>
+#include <linux/interrupt.h>
+#include <linux/wait.h>
+#include <linux/bitops.h>
+#include <linux/skbuff.h>
+
+#include "../nfc.h"
+#include <net/nfc/nci.h>
+#include <net/nfc/nci_core.h>
+#include <linux/nfc.h>
+
+/* Complete data exchange transaction and forward skb to nfc core */
+void nci_data_exchange_complete(struct nci_dev *ndev, struct sk_buff *skb,
+				__u8 conn_id, int err)
+{
+	const struct nci_conn_info *conn_info;
+	data_exchange_cb_t cb;
+	void *cb_context;
+
+	conn_info = nci_get_conn_info_by_conn_id(ndev, conn_id);
+	if (!conn_info) {
+		kfree_skb(skb);
+		clear_bit(NCI_DATA_EXCHANGE, &ndev->flags);
+		return;
+	}
+
+	cb = conn_info->data_exchange_cb;
+	cb_context = conn_info->data_exchange_cb_context;
+
+	pr_debug("len %d, err %d\n", skb ? skb->len : 0, err);
+
+	/* data exchange is complete, stop the data timer */
+	timer_delete_sync(&ndev->data_timer);
+	clear_bit(NCI_DATA_EXCHANGE_TO, &ndev->flags);
+
+	/* Claim completion atomically -- both close and rx_work may race here */
+	if (!test_and_clear_bit(NCI_DATA_EXCHANGE, &ndev->flags)) {
+		kfree_skb(skb);
+		return;
+	}
+
+	if (cb) {
+		/* forward skb to nfc core */
+		cb(cb_context, skb, err);
+	} else if (skb) {
+		pr_err("no rx callback, dropping rx data...\n");
+
+		/* no waiting callback, free skb */
+		kfree_skb(skb);
+	}
+}
+
+/* ----------------- NCI TX Data ----------------- */
+
+static inline void nci_push_data_hdr(struct nci_dev *ndev,
+				     __u8 conn_id,
+				     struct sk_buff *skb,
+				     __u8 pbf)
+{
+	struct nci_data_hdr *hdr;
+	int plen = skb->len;
+
+	hdr = skb_push(skb, NCI_DATA_HDR_SIZE);
+	hdr->conn_id = conn_id;
+	hdr->rfu = 0;
+	hdr->plen = plen;
+
+	nci_mt_set((__u8 *)hdr, NCI_MT_DATA_PKT);
+	nci_pbf_set((__u8 *)hdr, pbf);
+}
+
+int nci_conn_max_data_pkt_payload_size(struct nci_dev *ndev, __u8 conn_id)
+{
+	const struct nci_conn_info *conn_info;
+
+	conn_info = nci_get_conn_info_by_conn_id(ndev, conn_id);
+	if (!conn_info)
+		return -EPROTO;
+
+	return conn_info->max_pkt_payload_len;
+}
+EXPORT_SYMBOL(nci_conn_max_data_pkt_payload_size);
+
+static int nci_queue_tx_data_frags(struct nci_dev *ndev,
+				   __u8 conn_id,
+				   struct sk_buff *skb) {
+	const struct nci_conn_info *conn_info;
+	int total_len = skb->len;
+	const unsigned char *data = skb->data;
+	unsigned long flags;
+	struct sk_buff_head frags_q;
+	struct sk_buff *skb_frag;
+	int frag_len;
+	int rc = 0;
+
+	pr_debug("conn_id 0x%x, total_len %d\n", conn_id, total_len);
+
+	conn_info = nci_get_conn_info_by_conn_id(ndev, conn_id);
+	if (!conn_info) {
+		rc = -EPROTO;
+		goto exit;
+	}
+
+	__skb_queue_head_init(&frags_q);
+
+	while (total_len) {
+		frag_len =
+			min_t(int, total_len, conn_info->max_pkt_payload_len);
+
+		skb_frag = nci_skb_alloc(ndev,
+					 (NCI_DATA_HDR_SIZE + frag_len),
+					 GFP_ATOMIC);
+		if (skb_frag == NULL) {
+			rc = -ENOMEM;
+			goto free_exit;
+		}
+		skb_reserve(skb_frag, NCI_DATA_HDR_SIZE);
+
+		/* first, copy the data */
+		skb_put_data(skb_frag, data, frag_len);
+
+		/* second, set the header */
+		nci_push_data_hdr(ndev, conn_id, skb_frag,
+				  ((total_len == frag_len) ?
+				   (NCI_PBF_LAST) : (NCI_PBF_CONT)));
+
+		__skb_queue_tail(&frags_q, skb_frag);
+
+		data += frag_len;
+		total_len -= frag_len;
+
+		pr_debug("frag_len %d, remaining total_len %d\n",
+			 frag_len, total_len);
+	}
+
+	/* queue all fragments atomically */
+	spin_lock_irqsave(&ndev->tx_q.lock, flags);
+
+	while ((skb_frag = __skb_dequeue(&frags_q)) != NULL)
+		__skb_queue_tail(&ndev->tx_q, skb_frag);
+
+	spin_unlock_irqrestore(&ndev->tx_q.lock, flags);
+
+	/* free the original skb */
+	kfree_skb(skb);
+
+	goto exit;
+
+free_exit:
+	while ((skb_frag = __skb_dequeue(&frags_q)) != NULL)
+		kfree_skb(skb_frag);
+
+exit:
+	return rc;
+}
+
+/* Send NCI data */
+int nci_send_data(struct nci_dev *ndev, __u8 conn_id, struct sk_buff *skb)
+{
+	const struct nci_conn_info *conn_info;
+	int rc = 0;
+
+	pr_debug("conn_id 0x%x, plen %d\n", conn_id, skb->len);
+
+	conn_info = nci_get_conn_info_by_conn_id(ndev, conn_id);
+	if (!conn_info) {
+		rc = -EPROTO;
+		goto free_exit;
+	}
+
+	/* check if the packet need to be fragmented */
+	if (skb->len <= conn_info->max_pkt_payload_len) {
+		/* no need to fragment packet */
+		nci_push_data_hdr(ndev, conn_id, skb, NCI_PBF_LAST);
+
+		skb_queue_tail(&ndev->tx_q, skb);
+	} else {
+		/* fragment packet and queue the fragments */
+		rc = nci_queue_tx_data_frags(ndev, conn_id, skb);
+		if (rc) {
+			pr_err("failed to fragment tx data packet\n");
+			goto free_exit;
+		}
+	}
+
+	ndev->cur_conn_id = conn_id;
+	queue_work(ndev->tx_wq, &ndev->tx_work);
+
+	goto exit;
+
+free_exit:
+	kfree_skb(skb);
+
+exit:
+	return rc;
+}
+EXPORT_SYMBOL(nci_send_data);
+
+/* ----------------- NCI RX Data ----------------- */
+
+static void nci_add_rx_data_frag(struct nci_dev *ndev,
+				 struct sk_buff *skb,
+				 __u8 pbf, __u8 conn_id, __u8 status)
+{
+	int reassembly_len;
+	int err = 0;
+
+	if (status) {
+		err = status;
+		goto exit;
+	}
+
+	if (ndev->rx_data_reassembly) {
+		reassembly_len = ndev->rx_data_reassembly->len;
+
+		/* first, make enough room for the already accumulated data */
+		if (skb_cow_head(skb, reassembly_len)) {
+			pr_err("error adding room for accumulated rx data\n");
+
+			kfree_skb(skb);
+			skb = NULL;
+
+			kfree_skb(ndev->rx_data_reassembly);
+			ndev->rx_data_reassembly = NULL;
+
+			err = -ENOMEM;
+			goto exit;
+		}
+
+		/* second, combine the two fragments */
+		memcpy(skb_push(skb, reassembly_len),
+		       ndev->rx_data_reassembly->data,
+		       reassembly_len);
+
+		/* third, free old reassembly */
+		kfree_skb(ndev->rx_data_reassembly);
+		ndev->rx_data_reassembly = NULL;
+	}
+
+	if (pbf == NCI_PBF_CONT) {
+		/* need to wait for next fragment, store skb and exit */
+		ndev->rx_data_reassembly = skb;
+		return;
+	}
+
+exit:
+	if (ndev->nfc_dev->rf_mode == NFC_RF_TARGET) {
+		/* Data received in Target mode, forward to nfc core */
+		err = nfc_tm_data_received(ndev->nfc_dev, skb);
+		if (err)
+			pr_err("unable to handle received data\n");
+	} else {
+		nci_data_exchange_complete(ndev, skb, conn_id, err);
+	}
+}
+
+/* Rx Data packet */
+void nci_rx_data_packet(struct nci_dev *ndev, struct sk_buff *skb)
+{
+	__u8 pbf = nci_pbf(skb->data);
+	__u8 status = 0;
+	__u8 conn_id = nci_conn_id(skb->data);
+	const struct nci_conn_info *conn_info;
+
+	pr_debug("len %d\n", skb->len);
+
+	pr_debug("NCI RX: MT=data, PBF=%d, conn_id=%d, plen=%d\n",
+		 nci_pbf(skb->data),
+		 nci_conn_id(skb->data),
+		 nci_plen(skb->data));
+
+	conn_info = nci_get_conn_info_by_conn_id(ndev, nci_conn_id(skb->data));
+	if (!conn_info) {
+		kfree_skb(skb);
+		return;
+	}
+
+	/* strip the nci data header */
+	skb_pull(skb, NCI_DATA_HDR_SIZE);
+
+	if (ndev->target_active_prot == NFC_PROTO_MIFARE ||
+	    ndev->target_active_prot == NFC_PROTO_JEWEL ||
+	    ndev->target_active_prot == NFC_PROTO_FELICA ||
+	    ndev->target_active_prot == NFC_PROTO_ISO15693) {
+		/* frame I/F => remove the status byte */
+		pr_debug("frame I/F => remove the status byte\n");
+		status = skb->data[skb->len - 1];
+		skb_trim(skb, (skb->len - 1));
+	}
+
+	nci_add_rx_data_frag(ndev, skb, pbf, conn_id, nci_to_errno(status));
+}