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

3 files changed, 122 insertions, 0 deletions

diff --git a/drivers/clk/imgtec/Kconfig b/drivers/clk/imgtec/Kconfig
new file mode 100644
index 000000000..c965fd091
--- /dev/null
+++ b/drivers/clk/imgtec/Kconfig
@@ -0,0 +1,10 @@
+# SPDX-License-Identifier: GPL-2.0-only
+config COMMON_CLK_BOSTON
+	bool "Clock driver for MIPS Boston boards"
+	depends on MIPS || COMPILE_TEST
+	select MFD_SYSCON
+	help
+	  Enable this to support the system & CPU clocks on the MIPS Boston
+	  development board from Imagination Technologies. These are simple
+	  fixed rate clocks whose rate is determined by reading a platform
+	  provided register.
diff --git a/drivers/clk/imgtec/Makefile b/drivers/clk/imgtec/Makefile
new file mode 100644
index 000000000..d2e81426c
--- /dev/null
+++ b/drivers/clk/imgtec/Makefile
@@ -0,0 +1,2 @@
+# SPDX-License-Identifier: GPL-2.0-only
+obj-$(CONFIG_COMMON_CLK_BOSTON)		+= clk-boston.o
diff --git a/drivers/clk/imgtec/clk-boston.c b/drivers/clk/imgtec/clk-boston.c
new file mode 100644
index 000000000..f28c3ab93
--- /dev/null
+++ b/drivers/clk/imgtec/clk-boston.c
@@ -0,0 +1,110 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * Copyright (C) 2016-2017 Imagination Technologies
+ * Author: Paul Burton <paul.burton@mips.com>
+ */
+
+#define pr_fmt(fmt) "clk-boston: " fmt
+
+#include <linux/clk-provider.h>
+#include <linux/kernel.h>
+#include <linux/of.h>
+#include <linux/regmap.h>
+#include <linux/slab.h>
+#include <linux/mfd/syscon.h>
+
+#include <dt-bindings/clock/boston-clock.h>
+
+#define BOSTON_PLAT_MMCMDIV		0x30
+# define BOSTON_PLAT_MMCMDIV_CLK0DIV	(0xff << 0)
+# define BOSTON_PLAT_MMCMDIV_INPUT	(0xff << 8)
+# define BOSTON_PLAT_MMCMDIV_MUL	(0xff << 16)
+# define BOSTON_PLAT_MMCMDIV_CLK1DIV	(0xff << 24)
+
+#define BOSTON_CLK_COUNT 3
+
+static u32 ext_field(u32 val, u32 mask)
+{
+	return (val & mask) >> (ffs(mask) - 1);
+}
+
+static void __init clk_boston_setup(struct device_node *np)
+{
+	unsigned long in_freq, cpu_freq, sys_freq;
+	uint mmcmdiv, mul, cpu_div, sys_div;
+	struct clk_hw_onecell_data *onecell;
+	struct regmap *regmap;
+	struct clk_hw *hw;
+	int err;
+
+	regmap = syscon_node_to_regmap(np->parent);
+	if (IS_ERR(regmap)) {
+		pr_err("failed to find regmap\n");
+		return;
+	}
+
+	err = regmap_read(regmap, BOSTON_PLAT_MMCMDIV, &mmcmdiv);
+	if (err) {
+		pr_err("failed to read mmcm_div register: %d\n", err);
+		return;
+	}
+
+	in_freq = ext_field(mmcmdiv, BOSTON_PLAT_MMCMDIV_INPUT) * 1000000;
+	mul = ext_field(mmcmdiv, BOSTON_PLAT_MMCMDIV_MUL);
+
+	sys_div = ext_field(mmcmdiv, BOSTON_PLAT_MMCMDIV_CLK0DIV);
+	sys_freq = mult_frac(in_freq, mul, sys_div);
+
+	cpu_div = ext_field(mmcmdiv, BOSTON_PLAT_MMCMDIV_CLK1DIV);
+	cpu_freq = mult_frac(in_freq, mul, cpu_div);
+
+	onecell = kzalloc_flex(*onecell, hws, BOSTON_CLK_COUNT);
+	if (!onecell)
+		return;
+
+	onecell->num = BOSTON_CLK_COUNT;
+
+	hw = clk_hw_register_fixed_rate(NULL, "input", NULL, 0, in_freq);
+	if (IS_ERR(hw)) {
+		pr_err("failed to register input clock: %pe\n", hw);
+		goto fail_input;
+	}
+	onecell->hws[BOSTON_CLK_INPUT] = hw;
+
+	hw = clk_hw_register_fixed_rate(NULL, "sys", "input", 0, sys_freq);
+	if (IS_ERR(hw)) {
+		pr_err("failed to register sys clock: %pe\n", hw);
+		goto fail_sys;
+	}
+	onecell->hws[BOSTON_CLK_SYS] = hw;
+
+	hw = clk_hw_register_fixed_rate(NULL, "cpu", "input", 0, cpu_freq);
+	if (IS_ERR(hw)) {
+		pr_err("failed to register cpu clock: %pe\n", hw);
+		goto fail_cpu;
+	}
+	onecell->hws[BOSTON_CLK_CPU] = hw;
+
+	err = of_clk_add_hw_provider(np, of_clk_hw_onecell_get, onecell);
+	if (err) {
+		pr_err("failed to add DT provider: %d\n", err);
+		goto fail_clk_add;
+	}
+
+	return;
+
+fail_clk_add:
+	clk_hw_unregister_fixed_rate(onecell->hws[BOSTON_CLK_CPU]);
+fail_cpu:
+	clk_hw_unregister_fixed_rate(onecell->hws[BOSTON_CLK_SYS]);
+fail_sys:
+	clk_hw_unregister_fixed_rate(onecell->hws[BOSTON_CLK_INPUT]);
+fail_input:
+	kfree(onecell);
+}
+
+/*
+ * Use CLK_OF_DECLARE so that this driver is probed early enough to provide the
+ * CPU frequency for use with the GIC or cop0 counters/timers.
+ */
+CLK_OF_DECLARE(clk_boston, "img,boston-clock", clk_boston_setup);