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, 249 insertions, 0 deletions

diff --git a/drivers/acpi/ioapic.c b/drivers/acpi/ioapic.c
new file mode 100644
index 000000000..533ddd759
--- /dev/null
+++ b/drivers/acpi/ioapic.c
@@ -0,0 +1,249 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * IOAPIC/IOxAPIC/IOSAPIC driver
+ *
+ * Copyright (C) 2009 Fujitsu Limited.
+ * (c) Copyright 2009 Hewlett-Packard Development Company, L.P.
+ *
+ * Copyright (C) 2014 Intel Corporation
+ *
+ * Based on original drivers/pci/ioapic.c
+ *	Yinghai Lu <yinghai@kernel.org>
+ *	Jiang Liu <jiang.liu@intel.com>
+ */
+
+/*
+ * This driver manages I/O APICs added by hotplug after boot.
+ * We try to claim all I/O APIC devices, but those present at boot were
+ * registered when we parsed the ACPI MADT.
+ */
+
+#define pr_fmt(fmt) "ACPI: IOAPIC: " fmt
+
+#include <linux/slab.h>
+#include <linux/acpi.h>
+#include <linux/pci.h>
+#include <acpi/acpi.h>
+#include "internal.h"
+
+struct acpi_pci_ioapic {
+	acpi_handle	root_handle;
+	acpi_handle	handle;
+	u32		gsi_base;
+	struct resource	res;
+	struct pci_dev	*pdev;
+	struct list_head list;
+};
+
+static LIST_HEAD(ioapic_list);
+static DEFINE_MUTEX(ioapic_list_lock);
+
+static acpi_status setup_res(struct acpi_resource *acpi_res, void *data)
+{
+	struct resource *res = data;
+	struct resource_win win;
+
+	/*
+	 * We might assign this to 'res' later, make sure all pointers are
+	 * cleared before the resource is added to the global list
+	 */
+	memset(&win, 0, sizeof(win));
+
+	res->flags = 0;
+	if (acpi_dev_filter_resource_type(acpi_res, IORESOURCE_MEM))
+		return AE_OK;
+
+	if (!acpi_dev_resource_memory(acpi_res, res)) {
+		if (acpi_dev_resource_address_space(acpi_res, &win) ||
+		    acpi_dev_resource_ext_address_space(acpi_res, &win))
+			*res = win.res;
+	}
+	if ((res->flags & IORESOURCE_PREFETCH) ||
+	    (res->flags & IORESOURCE_DISABLED))
+		res->flags = 0;
+
+	return AE_CTRL_TERMINATE;
+}
+
+static bool acpi_is_ioapic(acpi_handle handle, char **type)
+{
+	acpi_status status;
+	struct acpi_device_info *info;
+	char *hid = NULL;
+	bool match = false;
+
+	if (!acpi_has_method(handle, "_GSB"))
+		return false;
+
+	status = acpi_get_object_info(handle, &info);
+	if (ACPI_SUCCESS(status)) {
+		if (info->valid & ACPI_VALID_HID)
+			hid = info->hardware_id.string;
+		if (hid) {
+			if (strcmp(hid, "ACPI0009") == 0) {
+				*type = "IOxAPIC";
+				match = true;
+			} else if (strcmp(hid, "ACPI000A") == 0) {
+				*type = "IOAPIC";
+				match = true;
+			}
+		}
+		kfree(info);
+	}
+
+	return match;
+}
+
+static acpi_status handle_ioapic_add(acpi_handle handle, u32 lvl,
+				     void *context, void **rv)
+{
+	acpi_status status;
+	unsigned long long gsi_base;
+	struct acpi_pci_ioapic *ioapic;
+	struct pci_dev *dev = NULL;
+	struct resource *res = NULL, *pci_res = NULL, *crs_res;
+	char *type = NULL;
+
+	if (!acpi_is_ioapic(handle, &type))
+		return AE_OK;
+
+	mutex_lock(&ioapic_list_lock);
+	list_for_each_entry(ioapic, &ioapic_list, list)
+		if (ioapic->handle == handle) {
+			mutex_unlock(&ioapic_list_lock);
+			return AE_OK;
+		}
+
+	status = acpi_evaluate_integer(handle, "_GSB", NULL, &gsi_base);
+	if (ACPI_FAILURE(status)) {
+		acpi_handle_warn(handle, "failed to evaluate _GSB method\n");
+		goto exit;
+	}
+
+	ioapic = kzalloc_obj(*ioapic);
+	if (!ioapic) {
+		pr_err("cannot allocate memory for new IOAPIC\n");
+		goto exit;
+	} else {
+		ioapic->root_handle = (acpi_handle)context;
+		ioapic->handle = handle;
+		ioapic->gsi_base = (u32)gsi_base;
+		INIT_LIST_HEAD(&ioapic->list);
+	}
+
+	if (acpi_ioapic_registered(handle, (u32)gsi_base))
+		goto done;
+
+	dev = acpi_get_pci_dev(handle);
+	if (dev && pci_resource_len(dev, 0)) {
+		if (pci_enable_device(dev) < 0)
+			goto exit_put;
+		pci_set_master(dev);
+		if (pci_request_region(dev, 0, type))
+			goto exit_disable;
+		pci_res = &dev->resource[0];
+		ioapic->pdev = dev;
+	} else {
+		pci_dev_put(dev);
+		dev = NULL;
+	}
+
+	crs_res = &ioapic->res;
+	acpi_walk_resources(handle, METHOD_NAME__CRS, setup_res, crs_res);
+	crs_res->name = type;
+	crs_res->flags |= IORESOURCE_BUSY;
+	if (crs_res->flags == 0) {
+		acpi_handle_warn(handle, "failed to get resource\n");
+		goto exit_release;
+	} else if (insert_resource(&iomem_resource, crs_res)) {
+		acpi_handle_warn(handle, "failed to insert resource\n");
+		goto exit_release;
+	}
+
+	/* try pci resource first, then "_CRS" resource */
+	res = pci_res;
+	if (!res || !res->flags)
+		res = crs_res;
+
+	if (acpi_register_ioapic(handle, res->start, (u32)gsi_base)) {
+		acpi_handle_warn(handle, "failed to register IOAPIC\n");
+		goto exit_release;
+	}
+done:
+	list_add(&ioapic->list, &ioapic_list);
+	mutex_unlock(&ioapic_list_lock);
+
+	if (dev)
+		dev_info(&dev->dev, "%s at %pR, GSI %u\n",
+			 type, res, (u32)gsi_base);
+	else
+		acpi_handle_info(handle, "%s at %pR, GSI %u\n",
+				 type, res, (u32)gsi_base);
+
+	return AE_OK;
+
+exit_release:
+	if (dev)
+		pci_release_region(dev, 0);
+	if (ioapic->res.flags && ioapic->res.parent)
+		release_resource(&ioapic->res);
+exit_disable:
+	if (dev)
+		pci_disable_device(dev);
+exit_put:
+	pci_dev_put(dev);
+	kfree(ioapic);
+exit:
+	mutex_unlock(&ioapic_list_lock);
+	*(acpi_status *)rv = AE_ERROR;
+	return AE_OK;
+}
+
+int acpi_ioapic_add(acpi_handle root_handle)
+{
+	acpi_status status, retval = AE_OK;
+
+	status = acpi_walk_namespace(ACPI_TYPE_DEVICE, root_handle,
+				     UINT_MAX, handle_ioapic_add, NULL,
+				     root_handle, (void **)&retval);
+
+	return ACPI_SUCCESS(status) && ACPI_SUCCESS(retval) ? 0 : -ENODEV;
+}
+
+void pci_ioapic_remove(struct acpi_pci_root *root)
+{
+	struct acpi_pci_ioapic *ioapic, *tmp;
+
+	mutex_lock(&ioapic_list_lock);
+	list_for_each_entry_safe(ioapic, tmp, &ioapic_list, list) {
+		if (root->device->handle != ioapic->root_handle)
+			continue;
+		if (ioapic->pdev) {
+			pci_release_region(ioapic->pdev, 0);
+			pci_disable_device(ioapic->pdev);
+			pci_dev_put(ioapic->pdev);
+		}
+	}
+	mutex_unlock(&ioapic_list_lock);
+}
+
+int acpi_ioapic_remove(struct acpi_pci_root *root)
+{
+	int retval = 0;
+	struct acpi_pci_ioapic *ioapic, *tmp;
+
+	mutex_lock(&ioapic_list_lock);
+	list_for_each_entry_safe(ioapic, tmp, &ioapic_list, list) {
+		if (root->device->handle != ioapic->root_handle)
+			continue;
+		if (acpi_unregister_ioapic(ioapic->handle, ioapic->gsi_base))
+			retval = -EBUSY;
+		if (ioapic->res.flags && ioapic->res.parent)
+			release_resource(&ioapic->res);
+		list_del(&ioapic->list);
+		kfree(ioapic);
+	}
+	mutex_unlock(&ioapic_list_lock);
+
+	return retval;
+}