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, 246 insertions, 0 deletions

diff --git a/drivers/sh/intc/access.c b/drivers/sh/intc/access.c
new file mode 100644
index 000000000..114390f96
--- /dev/null
+++ b/drivers/sh/intc/access.c
@@ -0,0 +1,246 @@
+/*
+ * Common INTC2 register accessors
+ *
+ * Copyright (C) 2007, 2008 Magnus Damm
+ * Copyright (C) 2009, 2010 Paul Mundt
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ */
+#include <linux/io.h>
+#include "internals.h"
+
+unsigned long intc_phys_to_virt(struct intc_desc_int *d, unsigned long address)
+{
+	struct intc_window *window;
+	int k;
+
+	/* scan through physical windows and convert address */
+	for (k = 0; k < d->nr_windows; k++) {
+		window = d->window + k;
+
+		if (address < window->phys)
+			continue;
+
+		if (address >= (window->phys + window->size))
+			continue;
+
+		address -= window->phys;
+		address += (unsigned long)window->virt;
+
+		return address;
+	}
+
+	/* no windows defined, register must be 1:1 mapped virt:phys */
+	return address;
+}
+
+unsigned int intc_get_reg(struct intc_desc_int *d, unsigned long address)
+{
+	unsigned int k;
+
+	address = intc_phys_to_virt(d, address);
+
+	for (k = 0; k < d->nr_reg; k++) {
+		if (d->reg[k] == address)
+			return k;
+	}
+
+	BUG();
+	return 0;
+}
+
+unsigned int intc_set_field_from_handle(unsigned int value,
+					unsigned int field_value,
+					unsigned int handle)
+{
+	unsigned int width = _INTC_WIDTH(handle);
+	unsigned int shift = _INTC_SHIFT(handle);
+
+	value &= ~(((1 << width) - 1) << shift);
+	value |= field_value << shift;
+	return value;
+}
+
+unsigned long intc_get_field_from_handle(unsigned int value, unsigned int handle)
+{
+	unsigned int width = _INTC_WIDTH(handle);
+	unsigned int shift = _INTC_SHIFT(handle);
+	unsigned int mask = ((1 << width) - 1) << shift;
+
+	return (value & mask) >> shift;
+}
+
+static unsigned long test_8(unsigned long addr, unsigned long h,
+			    unsigned long ignore)
+{
+	void __iomem *ptr = (void __iomem *)addr;
+	return intc_get_field_from_handle(__raw_readb(ptr), h);
+}
+
+static unsigned long test_16(unsigned long addr, unsigned long h,
+			     unsigned long ignore)
+{
+	void __iomem *ptr = (void __iomem *)addr;
+	return intc_get_field_from_handle(__raw_readw(ptr), h);
+}
+
+static unsigned long test_32(unsigned long addr, unsigned long h,
+			     unsigned long ignore)
+{
+	void __iomem *ptr = (void __iomem *)addr;
+	return intc_get_field_from_handle(__raw_readl(ptr), h);
+}
+
+static unsigned long write_8(unsigned long addr, unsigned long h,
+			     unsigned long data)
+{
+	void __iomem *ptr = (void __iomem *)addr;
+	__raw_writeb(intc_set_field_from_handle(0, data, h), ptr);
+	(void)__raw_readb(ptr);	/* Defeat write posting */
+	return 0;
+}
+
+static unsigned long write_16(unsigned long addr, unsigned long h,
+			      unsigned long data)
+{
+	void __iomem *ptr = (void __iomem *)addr;
+	__raw_writew(intc_set_field_from_handle(0, data, h), ptr);
+	(void)__raw_readw(ptr);	/* Defeat write posting */
+	return 0;
+}
+
+static unsigned long write_32(unsigned long addr, unsigned long h,
+			      unsigned long data)
+{
+	void __iomem *ptr = (void __iomem *)addr;
+	__raw_writel(intc_set_field_from_handle(0, data, h), ptr);
+	(void)__raw_readl(ptr);	/* Defeat write posting */
+	return 0;
+}
+
+static unsigned long modify_8(unsigned long addr, unsigned long h,
+			      unsigned long data)
+{
+	void __iomem *ptr = (void __iomem *)addr;
+	unsigned long flags;
+	unsigned int value;
+	local_irq_save(flags);
+	value = intc_set_field_from_handle(__raw_readb(ptr), data, h);
+	__raw_writeb(value, ptr);
+	(void)__raw_readb(ptr);	/* Defeat write posting */
+	local_irq_restore(flags);
+	return 0;
+}
+
+static unsigned long modify_16(unsigned long addr, unsigned long h,
+			       unsigned long data)
+{
+	void __iomem *ptr = (void __iomem *)addr;
+	unsigned long flags;
+	unsigned int value;
+	local_irq_save(flags);
+	value = intc_set_field_from_handle(__raw_readw(ptr), data, h);
+	__raw_writew(value, ptr);
+	(void)__raw_readw(ptr);	/* Defeat write posting */
+	local_irq_restore(flags);
+	return 0;
+}
+
+static unsigned long modify_32(unsigned long addr, unsigned long h,
+			       unsigned long data)
+{
+	void __iomem *ptr = (void __iomem *)addr;
+	unsigned long flags;
+	unsigned int value;
+	local_irq_save(flags);
+	value = intc_set_field_from_handle(__raw_readl(ptr), data, h);
+	__raw_writel(value, ptr);
+	(void)__raw_readl(ptr);	/* Defeat write posting */
+	local_irq_restore(flags);
+	return 0;
+}
+
+static unsigned long intc_mode_field(unsigned long addr,
+				     unsigned long handle,
+				     unsigned long (*fn)(unsigned long,
+						unsigned long,
+						unsigned long),
+				     unsigned int irq)
+{
+	return fn(addr, handle, ((1 << _INTC_WIDTH(handle)) - 1));
+}
+
+static unsigned long intc_mode_zero(unsigned long addr,
+				    unsigned long handle,
+				    unsigned long (*fn)(unsigned long,
+					       unsigned long,
+					       unsigned long),
+				    unsigned int irq)
+{
+	return fn(addr, handle, 0);
+}
+
+static unsigned long intc_mode_prio(unsigned long addr,
+				    unsigned long handle,
+				    unsigned long (*fn)(unsigned long,
+					       unsigned long,
+					       unsigned long),
+				    unsigned int irq)
+{
+	return fn(addr, handle, intc_get_prio_level(irq));
+}
+
+unsigned long (*intc_reg_fns[])(unsigned long addr,
+				unsigned long h,
+				unsigned long data) = {
+	[REG_FN_TEST_BASE + 0] = test_8,
+	[REG_FN_TEST_BASE + 1] = test_16,
+	[REG_FN_TEST_BASE + 3] = test_32,
+	[REG_FN_WRITE_BASE + 0] = write_8,
+	[REG_FN_WRITE_BASE + 1] = write_16,
+	[REG_FN_WRITE_BASE + 3] = write_32,
+	[REG_FN_MODIFY_BASE + 0] = modify_8,
+	[REG_FN_MODIFY_BASE + 1] = modify_16,
+	[REG_FN_MODIFY_BASE + 3] = modify_32,
+};
+
+unsigned long (*intc_enable_fns[])(unsigned long addr,
+				   unsigned long handle,
+				   unsigned long (*fn)(unsigned long,
+					    unsigned long,
+					    unsigned long),
+				   unsigned int irq) = {
+	[MODE_ENABLE_REG] = intc_mode_field,
+	[MODE_MASK_REG] = intc_mode_zero,
+	[MODE_DUAL_REG] = intc_mode_field,
+	[MODE_PRIO_REG] = intc_mode_prio,
+	[MODE_PCLR_REG] = intc_mode_prio,
+};
+
+unsigned long (*intc_disable_fns[])(unsigned long addr,
+				    unsigned long handle,
+				    unsigned long (*fn)(unsigned long,
+					     unsigned long,
+					     unsigned long),
+				    unsigned int irq) = {
+	[MODE_ENABLE_REG] = intc_mode_zero,
+	[MODE_MASK_REG] = intc_mode_field,
+	[MODE_DUAL_REG] = intc_mode_field,
+	[MODE_PRIO_REG] = intc_mode_zero,
+	[MODE_PCLR_REG] = intc_mode_field,
+};
+
+unsigned long (*intc_enable_noprio_fns[])(unsigned long addr,
+					  unsigned long handle,
+					  unsigned long (*fn)(unsigned long,
+						unsigned long,
+						unsigned long),
+					  unsigned int irq) = {
+	[MODE_ENABLE_REG] = intc_mode_field,
+	[MODE_MASK_REG] = intc_mode_zero,
+	[MODE_DUAL_REG] = intc_mode_field,
+	[MODE_PRIO_REG] = intc_mode_field,
+	[MODE_PCLR_REG] = intc_mode_field,
+};