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, 157 insertions, 0 deletions

diff --git a/drivers/mcb/mcb-pci.c b/drivers/mcb/mcb-pci.c
new file mode 100644
index 000000000..f1353da6e
--- /dev/null
+++ b/drivers/mcb/mcb-pci.c
@@ -0,0 +1,157 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * MEN Chameleon Bus.
+ *
+ * Copyright (C) 2014 MEN Mikroelektronik GmbH (www.men.de)
+ * Author: Johannes Thumshirn <johannes.thumshirn@men.de>
+ */
+
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <linux/mcb.h>
+
+#include "mcb-internal.h"
+
+struct priv {
+	struct mcb_bus *bus;
+	phys_addr_t mapbase;
+	void __iomem *base;
+};
+
+static int mcb_pci_get_irq(struct mcb_device *mdev)
+{
+	struct mcb_bus *mbus = mdev->bus;
+	struct device *dev = mbus->carrier;
+	struct pci_dev *pdev = to_pci_dev(dev);
+
+	return pdev->irq;
+}
+
+static int mcb_pci_probe(struct pci_dev *pdev, const struct pci_device_id *id)
+{
+	struct resource *res;
+	struct priv *priv;
+	int ret, table_size;
+	unsigned long flags;
+
+	priv = devm_kzalloc(&pdev->dev, sizeof(struct priv), GFP_KERNEL);
+	if (!priv)
+		return -ENOMEM;
+
+	ret = pci_enable_device(pdev);
+	if (ret) {
+		dev_err(&pdev->dev, "Failed to enable PCI device\n");
+		return -ENODEV;
+	}
+	pci_set_master(pdev);
+
+	flags = pci_resource_flags(pdev, 0);
+	if (flags & IORESOURCE_IO) {
+		ret = -ENOTSUPP;
+		dev_err(&pdev->dev,
+			"IO mapped PCI devices are not supported\n");
+		goto out_disable;
+	}
+
+	priv->mapbase = pci_resource_start(pdev, 0);
+	if (!priv->mapbase) {
+		dev_err(&pdev->dev, "No PCI resource\n");
+		ret = -ENODEV;
+		goto out_disable;
+	}
+
+	res = devm_request_mem_region(&pdev->dev, priv->mapbase,
+				      CHAM_HEADER_SIZE,
+				      KBUILD_MODNAME);
+	if (!res) {
+		dev_err(&pdev->dev, "Failed to request PCI memory\n");
+		ret = -EBUSY;
+		goto out_disable;
+	}
+
+	priv->base = devm_ioremap(&pdev->dev, priv->mapbase, CHAM_HEADER_SIZE);
+	if (!priv->base) {
+		dev_err(&pdev->dev, "Cannot ioremap\n");
+		ret = -ENOMEM;
+		goto out_disable;
+	}
+
+	pci_set_drvdata(pdev, priv);
+
+	priv->bus = mcb_alloc_bus(&pdev->dev);
+	if (IS_ERR(priv->bus)) {
+		ret = PTR_ERR(priv->bus);
+		goto out_disable;
+	}
+
+	priv->bus->get_irq = mcb_pci_get_irq;
+
+	ret = chameleon_parse_cells(priv->bus, priv->mapbase, priv->base);
+	if (ret < 0)
+		goto out_mcb_bus;
+
+	table_size = ret;
+
+	if (table_size < CHAM_HEADER_SIZE) {
+		/* Release the previous resources */
+		devm_iounmap(&pdev->dev, priv->base);
+		devm_release_mem_region(&pdev->dev, priv->mapbase, CHAM_HEADER_SIZE);
+
+		/* Then, allocate it again with the actual chameleon table size */
+		res = devm_request_mem_region(&pdev->dev, priv->mapbase,
+						table_size,
+						KBUILD_MODNAME);
+		if (!res) {
+			dev_err(&pdev->dev, "Failed to request PCI memory\n");
+			ret = -EBUSY;
+			goto out_mcb_bus;
+		}
+
+		priv->base = devm_ioremap(&pdev->dev, priv->mapbase, table_size);
+		if (!priv->base) {
+			dev_err(&pdev->dev, "Cannot ioremap\n");
+			ret = -ENOMEM;
+			goto out_mcb_bus;
+		}
+	}
+
+	mcb_bus_add_devices(priv->bus);
+
+	return 0;
+
+out_mcb_bus:
+	mcb_release_bus(priv->bus);
+out_disable:
+	pci_disable_device(pdev);
+	return ret;
+}
+
+static void mcb_pci_remove(struct pci_dev *pdev)
+{
+	struct priv *priv = pci_get_drvdata(pdev);
+
+	mcb_release_bus(priv->bus);
+
+	pci_disable_device(pdev);
+}
+
+static const struct pci_device_id mcb_pci_tbl[] = {
+	{ PCI_DEVICE(PCI_VENDOR_ID_MEN, PCI_DEVICE_ID_MEN_CHAMELEON) },
+	{ PCI_DEVICE(PCI_VENDOR_ID_ALTERA, PCI_DEVICE_ID_MEN_CHAMELEON) },
+	{ 0 },
+};
+MODULE_DEVICE_TABLE(pci, mcb_pci_tbl);
+
+static struct pci_driver mcb_pci_driver = {
+	.name = "mcb-pci",
+	.id_table = mcb_pci_tbl,
+	.probe = mcb_pci_probe,
+	.remove = mcb_pci_remove,
+};
+
+module_pci_driver(mcb_pci_driver);
+
+MODULE_AUTHOR("Johannes Thumshirn <johannes.thumshirn@men.de>");
+MODULE_LICENSE("GPL");
+MODULE_DESCRIPTION("MCB over PCI support");
+MODULE_IMPORT_NS("MCB");