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, 171 insertions, 0 deletions

diff --git a/drivers/pci/controller/pci-versatile.c b/drivers/pci/controller/pci-versatile.c
new file mode 100644
index 000000000..e9a6758fe
--- /dev/null
+++ b/drivers/pci/controller/pci-versatile.c
@@ -0,0 +1,171 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright 2004 Koninklijke Philips Electronics NV
+ *
+ * Conversion to platform driver and DT:
+ * Copyright 2014 Linaro Ltd.
+ *
+ * 14/04/2005 Initial version, colin.king@philips.com
+ */
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/of_address.h>
+#include <linux/of_pci.h>
+#include <linux/of_platform.h>
+#include <linux/pci.h>
+#include <linux/platform_device.h>
+
+#include "../pci.h"
+
+static void __iomem *versatile_pci_base;
+static void __iomem *versatile_cfg_base[2];
+
+#define PCI_IMAP(m)		(versatile_pci_base + ((m) * 4))
+#define PCI_SMAP(m)		(versatile_pci_base + 0x14 + ((m) * 4))
+#define PCI_SELFID		(versatile_pci_base + 0xc)
+
+#define VP_PCI_DEVICE_ID		0x030010ee
+#define VP_PCI_CLASS_ID			0x0b400000
+
+static u32 pci_slot_ignore;
+
+static int __init versatile_pci_slot_ignore(char *str)
+{
+	int slot;
+
+	while (get_option(&str, &slot)) {
+		if ((slot < 0) || (slot > 31))
+			pr_err("Illegal slot value: %d\n", slot);
+		else
+			pci_slot_ignore |= (1 << slot);
+	}
+	return 1;
+}
+__setup("pci_slot_ignore=", versatile_pci_slot_ignore);
+
+
+static void __iomem *versatile_map_bus(struct pci_bus *bus,
+				       unsigned int devfn, int offset)
+{
+	unsigned int busnr = bus->number;
+
+	if (pci_slot_ignore & (1 << PCI_SLOT(devfn)))
+		return NULL;
+
+	return versatile_cfg_base[1] + ((busnr << 16) | (devfn << 8) | offset);
+}
+
+static struct pci_ops pci_versatile_ops = {
+	.map_bus = versatile_map_bus,
+	.read	= pci_generic_config_read32,
+	.write	= pci_generic_config_write,
+};
+
+static int versatile_pci_probe(struct platform_device *pdev)
+{
+	struct device *dev = &pdev->dev;
+	struct resource *res;
+	struct resource_entry *entry;
+	int i, myslot = -1, mem = 1;
+	u32 val;
+	void __iomem *local_pci_cfg_base;
+	struct pci_host_bridge *bridge;
+
+	bridge = devm_pci_alloc_host_bridge(dev, 0);
+	if (!bridge)
+		return -ENOMEM;
+
+	versatile_pci_base = devm_platform_ioremap_resource(pdev, 0);
+	if (IS_ERR(versatile_pci_base))
+		return PTR_ERR(versatile_pci_base);
+
+	versatile_cfg_base[0] = devm_platform_ioremap_resource(pdev, 1);
+	if (IS_ERR(versatile_cfg_base[0]))
+		return PTR_ERR(versatile_cfg_base[0]);
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 2);
+	versatile_cfg_base[1] = devm_pci_remap_cfg_resource(dev, res);
+	if (IS_ERR(versatile_cfg_base[1]))
+		return PTR_ERR(versatile_cfg_base[1]);
+
+	resource_list_for_each_entry(entry, &bridge->windows) {
+		if (resource_type(entry->res) == IORESOURCE_MEM) {
+			writel(entry->res->start >> 28, PCI_IMAP(mem));
+			writel(__pa(PAGE_OFFSET) >> 28, PCI_SMAP(mem));
+			mem++;
+		}
+	}
+
+	/*
+	 * We need to discover the PCI core first to configure itself
+	 * before the main PCI probing is performed
+	 */
+	for (i = 0; i < 32; i++) {
+		if ((readl(versatile_cfg_base[0] + (i << 11) + PCI_VENDOR_ID) == VP_PCI_DEVICE_ID) &&
+		    (readl(versatile_cfg_base[0] + (i << 11) + PCI_CLASS_REVISION) == VP_PCI_CLASS_ID)) {
+			myslot = i;
+			break;
+		}
+	}
+	if (myslot == -1) {
+		dev_err(dev, "Cannot find PCI core!\n");
+		return -EIO;
+	}
+	/*
+	 * Do not to map Versatile FPGA PCI device into memory space
+	 */
+	pci_slot_ignore |= (1 << myslot);
+
+	dev_info(dev, "PCI core found (slot %d)\n", myslot);
+
+	writel(myslot, PCI_SELFID);
+	local_pci_cfg_base = versatile_cfg_base[1] + (myslot << 11);
+
+	val = readl(local_pci_cfg_base + PCI_COMMAND);
+	val |= PCI_COMMAND_MEMORY | PCI_COMMAND_MASTER | PCI_COMMAND_INVALIDATE;
+	writel(val, local_pci_cfg_base + PCI_COMMAND);
+
+	/*
+	 * Configure the PCI inbound memory windows to be 1:1 mapped to SDRAM
+	 */
+	writel(__pa(PAGE_OFFSET), local_pci_cfg_base + PCI_BASE_ADDRESS_0);
+	writel(__pa(PAGE_OFFSET), local_pci_cfg_base + PCI_BASE_ADDRESS_1);
+	writel(__pa(PAGE_OFFSET), local_pci_cfg_base + PCI_BASE_ADDRESS_2);
+
+	/*
+	 * For many years the kernel and QEMU were symbiotically buggy
+	 * in that they both assumed the same broken IRQ mapping.
+	 * QEMU therefore attempts to auto-detect old broken kernels
+	 * so that they still work on newer QEMU as they did on old
+	 * QEMU. Since we now use the correct (ie matching-hardware)
+	 * IRQ mapping we write a definitely different value to a
+	 * PCI_INTERRUPT_LINE register to tell QEMU that we expect
+	 * real hardware behaviour and it need not be backwards
+	 * compatible for us. This write is harmless on real hardware.
+	 */
+	writel(0, versatile_cfg_base[0] + PCI_INTERRUPT_LINE);
+
+	pci_add_flags(PCI_REASSIGN_ALL_BUS);
+
+	bridge->ops = &pci_versatile_ops;
+
+	return pci_host_probe(bridge);
+}
+
+static const struct of_device_id versatile_pci_of_match[] = {
+	{ .compatible = "arm,versatile-pci", },
+	{ },
+};
+MODULE_DEVICE_TABLE(of, versatile_pci_of_match);
+
+static struct platform_driver versatile_pci_driver = {
+	.driver = {
+		.name = "versatile-pci",
+		.of_match_table = versatile_pci_of_match,
+		.suppress_bind_attrs = true,
+	},
+	.probe = versatile_pci_probe,
+};
+module_platform_driver(versatile_pci_driver);
+
+MODULE_DESCRIPTION("Versatile PCI driver");