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, 294 insertions, 0 deletions

diff --git a/drivers/spi/spi-oc-tiny.c b/drivers/spi/spi-oc-tiny.c
new file mode 100644
index 000000000..1cd2a934c
--- /dev/null
+++ b/drivers/spi/spi-oc-tiny.c
@@ -0,0 +1,294 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * OpenCores tiny SPI host driver
+ *
+ * https://opencores.org/project,tiny_spi
+ *
+ * Copyright (C) 2011 Thomas Chou <thomas@wytron.com.tw>
+ *
+ * Based on spi_s3c24xx.c, which is:
+ * Copyright (c) 2006 Ben Dooks
+ * Copyright (c) 2006 Simtec Electronics
+ *	Ben Dooks <ben@simtec.co.uk>
+ */
+
+#include <linux/interrupt.h>
+#include <linux/errno.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/spi/spi.h>
+#include <linux/spi/spi_bitbang.h>
+#include <linux/spi/spi_oc_tiny.h>
+#include <linux/io.h>
+#include <linux/of.h>
+
+#define DRV_NAME "spi_oc_tiny"
+
+#define TINY_SPI_RXDATA 0
+#define TINY_SPI_TXDATA 4
+#define TINY_SPI_STATUS 8
+#define TINY_SPI_CONTROL 12
+#define TINY_SPI_BAUD 16
+
+#define TINY_SPI_STATUS_TXE 0x1
+#define TINY_SPI_STATUS_TXR 0x2
+
+struct tiny_spi {
+	/* bitbang has to be first */
+	struct spi_bitbang bitbang;
+	struct completion done;
+
+	void __iomem *base;
+	int irq;
+	unsigned int freq;
+	unsigned int baudwidth;
+	unsigned int baud;
+	unsigned int speed_hz;
+	unsigned int mode;
+	unsigned int len;
+	unsigned int txc, rxc;
+	const u8 *txp;
+	u8 *rxp;
+};
+
+static inline struct tiny_spi *tiny_spi_to_hw(struct spi_device *sdev)
+{
+	return spi_controller_get_devdata(sdev->controller);
+}
+
+static unsigned int tiny_spi_baud(struct spi_device *spi, unsigned int hz)
+{
+	struct tiny_spi *hw = tiny_spi_to_hw(spi);
+
+	return min(DIV_ROUND_UP(hw->freq, hz * 2), (1U << hw->baudwidth)) - 1;
+}
+
+static int tiny_spi_setup_transfer(struct spi_device *spi,
+				   struct spi_transfer *t)
+{
+	struct tiny_spi *hw = tiny_spi_to_hw(spi);
+	unsigned int baud = hw->baud;
+
+	if (t) {
+		if (t->speed_hz && t->speed_hz != hw->speed_hz)
+			baud = tiny_spi_baud(spi, t->speed_hz);
+	}
+	writel(baud, hw->base + TINY_SPI_BAUD);
+	writel(hw->mode, hw->base + TINY_SPI_CONTROL);
+	return 0;
+}
+
+static int tiny_spi_setup(struct spi_device *spi)
+{
+	struct tiny_spi *hw = tiny_spi_to_hw(spi);
+
+	if (spi->max_speed_hz != hw->speed_hz) {
+		hw->speed_hz = spi->max_speed_hz;
+		hw->baud = tiny_spi_baud(spi, hw->speed_hz);
+	}
+	hw->mode = spi->mode & SPI_MODE_X_MASK;
+	return 0;
+}
+
+static inline void tiny_spi_wait_txr(struct tiny_spi *hw)
+{
+	while (!(readb(hw->base + TINY_SPI_STATUS) &
+		 TINY_SPI_STATUS_TXR))
+		cpu_relax();
+}
+
+static inline void tiny_spi_wait_txe(struct tiny_spi *hw)
+{
+	while (!(readb(hw->base + TINY_SPI_STATUS) &
+		 TINY_SPI_STATUS_TXE))
+		cpu_relax();
+}
+
+static int tiny_spi_txrx_bufs(struct spi_device *spi, struct spi_transfer *t)
+{
+	struct tiny_spi *hw = tiny_spi_to_hw(spi);
+	const u8 *txp = t->tx_buf;
+	u8 *rxp = t->rx_buf;
+	unsigned int i;
+
+	if (hw->irq >= 0) {
+		/* use interrupt driven data transfer */
+		hw->len = t->len;
+		hw->txp = t->tx_buf;
+		hw->rxp = t->rx_buf;
+		hw->txc = 0;
+		hw->rxc = 0;
+
+		/* send the first byte */
+		if (t->len > 1) {
+			writeb(hw->txp ? *hw->txp++ : 0,
+			       hw->base + TINY_SPI_TXDATA);
+			hw->txc++;
+			writeb(hw->txp ? *hw->txp++ : 0,
+			       hw->base + TINY_SPI_TXDATA);
+			hw->txc++;
+			writeb(TINY_SPI_STATUS_TXR, hw->base + TINY_SPI_STATUS);
+		} else {
+			writeb(hw->txp ? *hw->txp++ : 0,
+			       hw->base + TINY_SPI_TXDATA);
+			hw->txc++;
+			writeb(TINY_SPI_STATUS_TXE, hw->base + TINY_SPI_STATUS);
+		}
+
+		wait_for_completion(&hw->done);
+	} else {
+		/* we need to tighten the transfer loop */
+		writeb(txp ? *txp++ : 0, hw->base + TINY_SPI_TXDATA);
+		for (i = 1; i < t->len; i++) {
+			writeb(txp ? *txp++ : 0, hw->base + TINY_SPI_TXDATA);
+
+			if (rxp || (i != t->len - 1))
+				tiny_spi_wait_txr(hw);
+			if (rxp)
+				*rxp++ = readb(hw->base + TINY_SPI_TXDATA);
+		}
+		tiny_spi_wait_txe(hw);
+		if (rxp)
+			*rxp++ = readb(hw->base + TINY_SPI_RXDATA);
+	}
+
+	return t->len;
+}
+
+static irqreturn_t tiny_spi_irq(int irq, void *dev)
+{
+	struct tiny_spi *hw = dev;
+
+	writeb(0, hw->base + TINY_SPI_STATUS);
+	if (hw->rxc + 1 == hw->len) {
+		if (hw->rxp)
+			*hw->rxp++ = readb(hw->base + TINY_SPI_RXDATA);
+		hw->rxc++;
+		complete(&hw->done);
+	} else {
+		if (hw->rxp)
+			*hw->rxp++ = readb(hw->base + TINY_SPI_TXDATA);
+		hw->rxc++;
+		if (hw->txc < hw->len) {
+			writeb(hw->txp ? *hw->txp++ : 0,
+			       hw->base + TINY_SPI_TXDATA);
+			hw->txc++;
+			writeb(TINY_SPI_STATUS_TXR,
+			       hw->base + TINY_SPI_STATUS);
+		} else {
+			writeb(TINY_SPI_STATUS_TXE,
+			       hw->base + TINY_SPI_STATUS);
+		}
+	}
+	return IRQ_HANDLED;
+}
+
+#ifdef CONFIG_OF
+static int tiny_spi_of_probe(struct platform_device *pdev)
+{
+	struct tiny_spi *hw = platform_get_drvdata(pdev);
+	struct device_node *np = pdev->dev.of_node;
+	u32 val;
+
+	if (!np)
+		return 0;
+	if (!of_property_read_u32(np, "clock-frequency", &val))
+		hw->freq = val;
+	if (!of_property_read_u32(np, "baud-width", &val))
+		hw->baudwidth = val;
+	return 0;
+}
+#else /* !CONFIG_OF */
+static int tiny_spi_of_probe(struct platform_device *pdev)
+{
+	return 0;
+}
+#endif /* CONFIG_OF */
+
+static int tiny_spi_probe(struct platform_device *pdev)
+{
+	struct tiny_spi_platform_data *platp = dev_get_platdata(&pdev->dev);
+	struct tiny_spi *hw;
+	struct spi_controller *host;
+	int err;
+
+	host = devm_spi_alloc_host(&pdev->dev, sizeof(struct tiny_spi));
+	if (!host)
+		return -ENOMEM;
+
+	/* setup the host state. */
+	host->bus_num = pdev->id;
+	host->mode_bits = SPI_CPOL | SPI_CPHA | SPI_CS_HIGH;
+	host->setup = tiny_spi_setup;
+	host->use_gpio_descriptors = true;
+
+	hw = spi_controller_get_devdata(host);
+	platform_set_drvdata(pdev, hw);
+
+	/* setup the state for the bitbang driver */
+	hw->bitbang.ctlr = host;
+	hw->bitbang.setup_transfer = tiny_spi_setup_transfer;
+	hw->bitbang.txrx_bufs = tiny_spi_txrx_bufs;
+
+	/* find and map our resources */
+	hw->base = devm_platform_ioremap_resource(pdev, 0);
+	if (IS_ERR(hw->base))
+		return PTR_ERR(hw->base);
+	/* irq is optional */
+	hw->irq = platform_get_irq(pdev, 0);
+	if (hw->irq >= 0) {
+		init_completion(&hw->done);
+		err = devm_request_irq(&pdev->dev, hw->irq, tiny_spi_irq, 0,
+				       pdev->name, hw);
+		if (err)
+			return err;
+	}
+	/* find platform data */
+	if (platp) {
+		hw->freq = platp->freq;
+		hw->baudwidth = platp->baudwidth;
+	} else {
+		err = tiny_spi_of_probe(pdev);
+		if (err)
+			return err;
+	}
+
+	/* register our spi controller */
+	err = spi_bitbang_start(&hw->bitbang);
+	if (err)
+		return err;
+	dev_info(&pdev->dev, "base %p, irq %d\n", hw->base, hw->irq);
+
+	return 0;
+}
+
+static void tiny_spi_remove(struct platform_device *pdev)
+{
+	struct tiny_spi *hw = platform_get_drvdata(pdev);
+
+	spi_bitbang_stop(&hw->bitbang);
+}
+
+#ifdef CONFIG_OF
+static const struct of_device_id tiny_spi_match[] = {
+	{ .compatible = "opencores,tiny-spi-rtlsvn2", },
+	{},
+};
+MODULE_DEVICE_TABLE(of, tiny_spi_match);
+#endif /* CONFIG_OF */
+
+static struct platform_driver tiny_spi_driver = {
+	.probe = tiny_spi_probe,
+	.remove = tiny_spi_remove,
+	.driver = {
+		.name = DRV_NAME,
+		.pm = NULL,
+		.of_match_table = of_match_ptr(tiny_spi_match),
+	},
+};
+module_platform_driver(tiny_spi_driver);
+
+MODULE_DESCRIPTION("OpenCores tiny SPI driver");
+MODULE_AUTHOR("Thomas Chou <thomas@wytron.com.tw>");
+MODULE_LICENSE("GPL");
+MODULE_ALIAS("platform:" DRV_NAME);