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, 169 insertions, 0 deletions

diff --git a/drivers/misc/issei/ham.c b/drivers/misc/issei/ham.c
new file mode 100644
index 000000000..674d9733d
--- /dev/null
+++ b/drivers/misc/issei/ham.c
@@ -0,0 +1,169 @@
+// SPDX-License-Identifier: GPL-2.0
+/* Copyright (C) 2023-2026 Intel Corporation */
+#include <linux/dev_printk.h>
+#include <linux/errno.h>
+#include <linux/slab.h>
+#include <linux/string.h>
+
+#include "dma.h"
+#include "fw_client.h"
+#include "ham.h"
+#include "hw_msg.h"
+#include "issei_dev.h"
+
+static int __issei_ham_send_msg(struct issei_device *idev, u32 length, void *buf)
+{
+	struct issei_dma_data data = { };
+	int ret;
+
+	data.length = length;
+	data.buf = buf;
+	ret = issei_dma_write(idev, &data);
+	if (ret)
+		return ret;
+	return idev->ops->irq_write_generate(idev);
+}
+
+/**
+ * issei_ham_send_start_req - send start request to firmware
+ * @idev: issei device object
+ *
+ * Return: 0 on success, <0 on failures
+ */
+int issei_ham_send_start_req(struct issei_device *idev)
+{
+	struct ham_start_message_req req;
+
+	req.header.cmd = HAM_BUS_CMD_START_REQ;
+	req.supported_version = ISSEI_SUPPORTED_PROTOCOL_VER;
+	req.heci_capabilities_length = 0;
+
+	return __issei_ham_send_msg(idev, sizeof(req), &req);
+}
+
+/**
+ * issei_ham_send_clients_req - send clients request to firmware
+ * @idev: issei device object
+ *
+ * Return: 0 on success, <0 on failures
+ */
+int issei_ham_send_clients_req(struct issei_device *idev)
+{
+	struct ham_get_clients_req req;
+
+	req.header.cmd = HAM_BUS_CMD_CLIENT_REQ;
+
+	return __issei_ham_send_msg(idev, sizeof(req), &req);
+}
+
+static int issei_ham_start_rsp(struct issei_device *idev, const u8 *buf, size_t length)
+{
+	struct ham_start_message_res *res = (struct ham_start_message_res *)buf;
+	int ret;
+
+	if (idev->rst_state != ISSEI_RST_STATE_START) {
+		dev_err(&idev->dev, "Wrong state %d != %d\n",
+			idev->rst_state, ISSEI_RST_STATE_START);
+		return -EPROTO;
+	}
+
+	if (length < sizeof(*res)) {
+		dev_err(&idev->dev, "Small start response size %zu < %zu\n",
+			length, sizeof(*res));
+		return -EPROTO;
+	}
+
+	if (length - sizeof(*res) != res->heci_capabilities_length) {
+		dev_err(&idev->dev, "Wrong start response size %zu != %u\n",
+			length - sizeof(*res), res->heci_capabilities_length);
+		return -EPROTO;
+	}
+
+	memcpy(idev->fw_version, res->fw_version, sizeof(idev->fw_version));
+	idev->fw_protocol_ver = res->supported_version;
+	dev_dbg(&idev->dev, "FW protocol: %u FW version %u.%u.%u.%u", idev->fw_protocol_ver,
+		idev->fw_version[0], idev->fw_version[1],
+		idev->fw_version[2], idev->fw_version[3]);
+
+	ret = issei_ham_send_clients_req(idev);
+	if (ret == -EBUSY)
+		ret = 0;
+
+	return ret;
+}
+
+static int issei_ham_client_rsp(struct issei_device *idev, const u8 *buf, size_t length)
+{
+	struct ham_get_clients_res *res = (struct ham_get_clients_res *)buf;
+	struct ham_client_properties *client;
+
+	if (idev->rst_state != ISSEI_RST_STATE_CLIENT_ENUM) {
+		dev_err(&idev->dev, "Wrong state %d != %d\n",
+			idev->rst_state, ISSEI_RST_STATE_CLIENT_ENUM);
+		return -EPROTO;
+	}
+
+	if (length < sizeof(*res)) {
+		dev_err(&idev->dev, "Small response size %zu < %zu\n", length, sizeof(*res));
+		return -EPROTO;
+	}
+
+	if (length - sizeof(*res) != res->client_count * sizeof(struct ham_client_properties)) {
+		dev_err(&idev->dev, "Wrong response size %zu < %zu\n",
+			length - sizeof(*res),
+			res->client_count * sizeof(struct ham_client_properties));
+		return -EPROTO;
+	}
+
+	guard(mutex)(&idev->client_lock);
+
+	for (size_t i = 0; i < res->client_count; i++) {
+		client = &res->clients_props[i];
+		dev_dbg(&idev->dev, "client: id = %u ver = %u uuid = %pUb mtu = %u flags = %u",
+			client->client_number, client->protocol_ver, &client->client_uuid,
+			client->client_mtu, client->flags);
+		issei_fw_cl_create(idev, client->client_number, client->protocol_ver,
+				   &client->client_uuid, client->client_mtu, client->flags);
+	}
+	return 0;
+}
+
+static int __issei_ham_process_ham_rsp(struct issei_device *idev, const u8 *buf, size_t length)
+{
+	struct ham_bus_message *hdr = (struct ham_bus_message *)buf;
+
+	if (length < sizeof(*hdr)) {
+		dev_err(&idev->dev, "Small bus message size %zu < %zu\n",
+			length, sizeof(*hdr));
+		return -EPROTO;
+	}
+
+	switch (hdr->cmd) {
+	case HAM_BUS_CMD_START_RSP:
+		return issei_ham_start_rsp(idev, buf, length);
+
+	case HAM_BUS_CMD_CLIENT_RSP:
+		return issei_ham_client_rsp(idev, buf, length);
+
+	default:
+		dev_err(&idev->dev, "Unexpected command 0x%x", hdr->cmd);
+		return -EPROTO;
+	}
+}
+
+/**
+ * issei_ham_process_ham_rsp - process response from firmware and release buffer
+ * @idev: issei device object
+ * @buf: response buffer
+ * @length: response buffer length
+ *
+ * Return: 0 on success, <0 on failures
+ */
+int issei_ham_process_ham_rsp(struct issei_device *idev, const u8 *buf, size_t length)
+{
+	int ret;
+
+	ret = __issei_ham_process_ham_rsp(idev, buf, length);
+	kfree(buf);
+	return ret;
+}