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, 100 insertions, 0 deletions

diff --git a/drivers/usb/chipidea/udc.h b/drivers/usb/chipidea/udc.h
new file mode 100644
index 000000000..c8a47389a
--- /dev/null
+++ b/drivers/usb/chipidea/udc.h
@@ -0,0 +1,100 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * udc.h - ChipIdea UDC structures
+ *
+ * Copyright (C) 2008 Chipidea - MIPS Technologies, Inc. All rights reserved.
+ *
+ * Author: David Lopo
+ */
+
+#ifndef __DRIVERS_USB_CHIPIDEA_UDC_H
+#define __DRIVERS_USB_CHIPIDEA_UDC_H
+
+#include <linux/list.h>
+
+#define CTRL_PAYLOAD_MAX   64
+#define RX        0  /* similar to USB_DIR_OUT but can be used as an index */
+#define TX        1  /* similar to USB_DIR_IN  but can be used as an index */
+
+/* DMA layout of transfer descriptors */
+struct ci_hw_td {
+	/* 0 */
+	__le32 next;
+#define TD_TERMINATE          BIT(0)
+#define TD_ADDR_MASK          (0xFFFFFFEUL << 5)
+	/* 1 */
+	__le32 token;
+#define TD_STATUS             (0x00FFUL <<  0)
+#define TD_STATUS_TR_ERR      BIT(3)
+#define TD_STATUS_DT_ERR      BIT(5)
+#define TD_STATUS_HALTED      BIT(6)
+#define TD_STATUS_ACTIVE      BIT(7)
+#define TD_MULTO              (0x0003UL << 10)
+#define TD_IOC                BIT(15)
+#define TD_TOTAL_BYTES        (0x7FFFUL << 16)
+	/* 2 */
+	__le32 page[5];
+#define TD_CURR_OFFSET        (0x0FFFUL <<  0)
+#define TD_FRAME_NUM          (0x07FFUL <<  0)
+#define TD_RESERVED_MASK      (0x0FFFUL <<  0)
+} __attribute__ ((packed, aligned(4)));
+
+/* DMA layout of queue heads */
+struct ci_hw_qh {
+	/* 0 */
+	__le32 cap;
+#define QH_IOS                BIT(15)
+#define QH_MAX_PKT            (0x07FFUL << 16)
+#define QH_ZLT                BIT(29)
+#define QH_MULT               (0x0003UL << 30)
+#define QH_ISO_MULT(x)		((x >> 11) & 0x03)
+	/* 1 */
+	__le32 curr;
+	/* 2 - 8 */
+	struct ci_hw_td		td;
+	/* 9 */
+	__le32 RESERVED;
+	struct usb_ctrlrequest   setup;
+} __attribute__ ((packed, aligned(4)));
+
+struct td_node {
+	struct list_head	td;
+	dma_addr_t		dma;
+	struct ci_hw_td		*ptr;
+	int			td_remaining_size;
+};
+
+/**
+ * struct ci_hw_req - usb request representation
+ * @req: request structure for gadget drivers
+ * @queue: link to QH list
+ * @tds: link to TD list
+ * @sgt: hold original sglist when bounce sglist
+ */
+struct ci_hw_req {
+	struct usb_request	req;
+	struct list_head	queue;
+	struct list_head	tds;
+	struct sg_table		sgt;
+};
+
+#ifdef CONFIG_USB_CHIPIDEA_UDC
+
+int ci_hdrc_gadget_init(struct ci_hdrc *ci);
+void ci_hdrc_gadget_destroy(struct ci_hdrc *ci);
+
+#else
+
+static inline int ci_hdrc_gadget_init(struct ci_hdrc *ci)
+{
+	return -ENXIO;
+}
+
+static inline void ci_hdrc_gadget_destroy(struct ci_hdrc *ci)
+{
+
+}
+
+#endif
+
+#endif /* __DRIVERS_USB_CHIPIDEA_UDC_H */