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, 122 insertions, 0 deletions

diff --git a/drivers/reset/reset-raspberrypi.c b/drivers/reset/reset-raspberrypi.c
new file mode 100644
index 000000000..fa23db554
--- /dev/null
+++ b/drivers/reset/reset-raspberrypi.c
@@ -0,0 +1,122 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Raspberry Pi 4 firmware reset driver
+ *
+ * Copyright (C) 2020 Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
+ */
+#include <linux/delay.h>
+#include <linux/device.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/platform_device.h>
+#include <linux/reset-controller.h>
+#include <soc/bcm2835/raspberrypi-firmware.h>
+#include <dt-bindings/reset/raspberrypi,firmware-reset.h>
+
+struct rpi_reset {
+	struct reset_controller_dev rcdev;
+	struct rpi_firmware *fw;
+};
+
+static inline struct rpi_reset *to_rpi(struct reset_controller_dev *rcdev)
+{
+	return container_of(rcdev, struct rpi_reset, rcdev);
+}
+
+static int rpi_reset_reset(struct reset_controller_dev *rcdev, unsigned long id)
+{
+	struct rpi_reset *priv = to_rpi(rcdev);
+	u32 dev_addr;
+	int ret;
+
+	switch (id) {
+	case RASPBERRYPI_FIRMWARE_RESET_ID_USB:
+		/*
+		 * The Raspberry Pi 4 gets its USB functionality from VL805, a
+		 * PCIe chip that implements xHCI. After a PCI reset, VL805's
+		 * firmware may either be loaded directly from an EEPROM or, if
+		 * not present, by the SoC's co-processor, VideoCore. rpi's
+		 * VideoCore OS contains both the non public firmware load
+		 * logic and the VL805 firmware blob. This triggers the
+		 * aforementioned process.
+		 *
+		 * The pci device address is expected is expected by the
+		 * firmware encoded like this:
+		 *
+		 *	PCI_BUS << 20 | PCI_SLOT << 15 | PCI_FUNC << 12
+		 *
+		 * But since rpi's PCIe is hardwired, we know the address in
+		 * advance.
+		 */
+		dev_addr = 0x100000;
+		ret = rpi_firmware_property(priv->fw, RPI_FIRMWARE_NOTIFY_XHCI_RESET,
+					    &dev_addr, sizeof(dev_addr));
+		if (ret)
+			return ret;
+
+		/* Wait for vl805 to startup */
+		usleep_range(200, 1000);
+		break;
+
+	default:
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static const struct reset_control_ops rpi_reset_ops = {
+	.reset	= rpi_reset_reset,
+};
+
+static int rpi_reset_probe(struct platform_device *pdev)
+{
+	struct device *dev = &pdev->dev;
+	struct rpi_firmware *fw;
+	struct device_node *np;
+	struct rpi_reset *priv;
+
+	np = of_get_parent(dev->of_node);
+	if (!np) {
+		dev_err(dev, "Missing firmware node\n");
+		return -ENOENT;
+	}
+
+	fw = devm_rpi_firmware_get(&pdev->dev, np);
+	of_node_put(np);
+	if (!fw)
+		return -EPROBE_DEFER;
+
+	priv = devm_kzalloc(dev, sizeof(*priv), GFP_KERNEL);
+	if (!priv)
+		return -ENOMEM;
+
+	dev_set_drvdata(dev, priv);
+
+	priv->fw = fw;
+	priv->rcdev.owner = THIS_MODULE;
+	priv->rcdev.nr_resets = RASPBERRYPI_FIRMWARE_RESET_NUM_IDS;
+	priv->rcdev.ops = &rpi_reset_ops;
+	priv->rcdev.of_node = dev->of_node;
+
+	return devm_reset_controller_register(dev, &priv->rcdev);
+}
+
+static const struct of_device_id rpi_reset_of_match[] = {
+	{ .compatible = "raspberrypi,firmware-reset" },
+	{ /* sentinel */ }
+};
+MODULE_DEVICE_TABLE(of, rpi_reset_of_match);
+
+static struct platform_driver rpi_reset_driver = {
+	.probe	= rpi_reset_probe,
+	.driver	= {
+		.name = "raspberrypi-reset",
+		.of_match_table = rpi_reset_of_match,
+	},
+};
+module_platform_driver(rpi_reset_driver);
+
+MODULE_AUTHOR("Nicolas Saenz Julienne <nsaenzjulienne@suse.de>");
+MODULE_DESCRIPTION("Raspberry Pi 4 firmware reset driver");
+MODULE_LICENSE("GPL");