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, 165 insertions, 0 deletions

diff --git a/drivers/memory/da8xx-ddrctl.c b/drivers/memory/da8xx-ddrctl.c
new file mode 100644
index 000000000..2bf34da85
--- /dev/null
+++ b/drivers/memory/da8xx-ddrctl.c
@@ -0,0 +1,165 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * TI da8xx DDR2/mDDR controller driver
+ *
+ * Copyright (C) 2016 BayLibre SAS
+ *
+ * Author:
+ *   Bartosz Golaszewski <bgolaszewski@baylibre.com>
+ */
+
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/platform_device.h>
+#include <linux/io.h>
+
+/*
+ * REVISIT: Linux doesn't have a good framework for the kind of performance
+ * knobs this driver controls. We can't use device tree properties as it deals
+ * with hardware configuration rather than description. We also don't want to
+ * commit to maintaining some random sysfs attributes.
+ *
+ * For now we just hardcode the register values for the boards that need
+ * some changes (as is the case for the LCD controller on da850-lcdk - the
+ * first board we support here). When linux gets an appropriate framework,
+ * we'll easily convert the driver to it.
+ */
+
+struct da8xx_ddrctl_config_knob {
+	const char *name;
+	u32 reg;
+	u32 mask;
+	u32 shift;
+};
+
+static const struct da8xx_ddrctl_config_knob da8xx_ddrctl_knobs[] = {
+	{
+		.name = "da850-pbbpr",
+		.reg = 0x20,
+		.mask = 0xffffff00,
+		.shift = 0,
+	},
+};
+
+struct da8xx_ddrctl_setting {
+	const char *name;
+	u32 val;
+};
+
+struct da8xx_ddrctl_board_settings {
+	const char *board;
+	const struct da8xx_ddrctl_setting *settings;
+};
+
+static const struct da8xx_ddrctl_setting da850_lcdk_ddrctl_settings[] = {
+	{
+		.name = "da850-pbbpr",
+		.val = 0x20,
+	},
+	{ }
+};
+
+static const struct da8xx_ddrctl_board_settings da8xx_ddrctl_board_confs[] = {
+	{
+		.board = "ti,da850-lcdk",
+		.settings = da850_lcdk_ddrctl_settings,
+	},
+};
+
+static const struct da8xx_ddrctl_config_knob *
+da8xx_ddrctl_match_knob(const struct da8xx_ddrctl_setting *setting)
+{
+	const struct da8xx_ddrctl_config_knob *knob;
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(da8xx_ddrctl_knobs); i++) {
+		knob = &da8xx_ddrctl_knobs[i];
+
+		if (strcmp(knob->name, setting->name) == 0)
+			return knob;
+	}
+
+	return NULL;
+}
+
+static const struct da8xx_ddrctl_setting *da8xx_ddrctl_get_board_settings(void)
+{
+	const struct da8xx_ddrctl_board_settings *board_settings;
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(da8xx_ddrctl_board_confs); i++) {
+		board_settings = &da8xx_ddrctl_board_confs[i];
+
+		if (of_machine_is_compatible(board_settings->board))
+			return board_settings->settings;
+	}
+
+	return NULL;
+}
+
+static int da8xx_ddrctl_probe(struct platform_device *pdev)
+{
+	const struct da8xx_ddrctl_config_knob *knob;
+	const struct da8xx_ddrctl_setting *setting;
+	struct resource *res;
+	void __iomem *ddrctl;
+	struct device *dev;
+	u32 reg;
+
+	dev = &pdev->dev;
+
+	setting = da8xx_ddrctl_get_board_settings();
+	if (!setting) {
+		dev_err(dev, "no settings defined for this board\n");
+		return -EINVAL;
+	}
+
+	ddrctl = devm_platform_get_and_ioremap_resource(pdev, 0, &res);
+	if (IS_ERR(ddrctl)) {
+		dev_err(dev, "unable to map memory controller registers\n");
+		return PTR_ERR(ddrctl);
+	}
+
+	for (; setting->name; setting++) {
+		knob = da8xx_ddrctl_match_knob(setting);
+		if (!knob) {
+			dev_warn(dev,
+				 "no such config option: %s\n", setting->name);
+			continue;
+		}
+
+		if (knob->reg + sizeof(u32) > resource_size(res)) {
+			dev_warn(dev,
+				 "register offset of '%s' exceeds mapped memory size\n",
+				 knob->name);
+			continue;
+		}
+
+		reg = readl(ddrctl + knob->reg);
+		reg &= knob->mask;
+		reg |= setting->val << knob->shift;
+
+		dev_dbg(dev, "writing 0x%08x to %s\n", reg, setting->name);
+
+		writel(reg, ddrctl + knob->reg);
+	}
+
+	return 0;
+}
+
+static const struct of_device_id da8xx_ddrctl_of_match[] = {
+	{ .compatible = "ti,da850-ddr-controller", },
+	{ },
+};
+
+static struct platform_driver da8xx_ddrctl_driver = {
+	.probe = da8xx_ddrctl_probe,
+	.driver = {
+		.name = "da850-ddr-controller",
+		.of_match_table = da8xx_ddrctl_of_match,
+	},
+};
+module_platform_driver(da8xx_ddrctl_driver);
+
+MODULE_AUTHOR("Bartosz Golaszewski <bgolaszewski@baylibre.com>");
+MODULE_DESCRIPTION("TI da8xx DDR2/mDDR controller driver");