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, 162 insertions, 0 deletions

diff --git a/drivers/acpi/acpi_lpit.c b/drivers/acpi/acpi_lpit.c
new file mode 100644
index 000000000..b8d98b1b4
--- /dev/null
+++ b/drivers/acpi/acpi_lpit.c
@@ -0,0 +1,162 @@
+// SPDX-License-Identifier: GPL-2.0-only
+
+/*
+ * acpi_lpit.c - LPIT table processing functions
+ *
+ * Copyright (C) 2017 Intel Corporation. All rights reserved.
+ */
+
+#include <linux/cpu.h>
+#include <linux/acpi.h>
+#include <asm/msr.h>
+#include <asm/tsc.h>
+#include "internal.h"
+
+struct lpit_residency_info {
+	struct acpi_generic_address gaddr;
+	u64 frequency;
+	void __iomem *iomem_addr;
+};
+
+/* Storage for an memory mapped and FFH based entries */
+static struct lpit_residency_info residency_info_mem;
+static struct lpit_residency_info residency_info_ffh;
+
+static int lpit_read_residency_counter_us(u64 *counter, bool io_mem)
+{
+	int err;
+
+	if (io_mem) {
+		u64 count = 0;
+		int error;
+
+		error = acpi_os_read_iomem(residency_info_mem.iomem_addr, &count,
+					   residency_info_mem.gaddr.bit_width);
+		if (error)
+			return error;
+
+		*counter = div64_u64(count * 1000000ULL, residency_info_mem.frequency);
+		return 0;
+	}
+
+	err = rdmsrq_safe(residency_info_ffh.gaddr.address, counter);
+	if (!err) {
+		u64 mask = GENMASK_ULL(residency_info_ffh.gaddr.bit_offset +
+				       residency_info_ffh.gaddr. bit_width - 1,
+				       residency_info_ffh.gaddr.bit_offset);
+
+		*counter &= mask;
+		*counter >>= residency_info_ffh.gaddr.bit_offset;
+		*counter = div64_u64(*counter * 1000000ULL, residency_info_ffh.frequency);
+		return 0;
+	}
+
+	return -ENODATA;
+}
+
+static ssize_t low_power_idle_system_residency_us_show(struct device *dev,
+						       struct device_attribute *attr,
+						       char *buf)
+{
+	u64 counter;
+	int ret;
+
+	ret = lpit_read_residency_counter_us(&counter, true);
+	if (ret)
+		return ret;
+
+	return sprintf(buf, "%llu\n", counter);
+}
+static DEVICE_ATTR_RO(low_power_idle_system_residency_us);
+
+static ssize_t low_power_idle_cpu_residency_us_show(struct device *dev,
+						    struct device_attribute *attr,
+						    char *buf)
+{
+	u64 counter;
+	int ret;
+
+	ret = lpit_read_residency_counter_us(&counter, false);
+	if (ret)
+		return ret;
+
+	return sprintf(buf, "%llu\n", counter);
+}
+static DEVICE_ATTR_RO(low_power_idle_cpu_residency_us);
+
+int lpit_read_residency_count_address(u64 *address)
+{
+	if (!residency_info_mem.gaddr.address)
+		return -EINVAL;
+
+	*address = residency_info_mem.gaddr.address;
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(lpit_read_residency_count_address);
+
+static void lpit_update_residency(struct lpit_residency_info *info,
+				 struct acpi_lpit_native *lpit_native)
+{
+	struct device *dev_root = bus_get_dev_root(&cpu_subsys);
+
+	/* Silently fail, if cpuidle attribute group is not present */
+	if (!dev_root)
+		return;
+
+	info->frequency = lpit_native->counter_frequency ?
+				lpit_native->counter_frequency : mul_u32_u32(tsc_khz, 1000U);
+	if (!info->frequency)
+		info->frequency = 1;
+
+	info->gaddr = lpit_native->residency_counter;
+	if (info->gaddr.space_id == ACPI_ADR_SPACE_SYSTEM_MEMORY) {
+		info->iomem_addr = ioremap(info->gaddr.address,
+						   info->gaddr.bit_width / 8);
+		if (!info->iomem_addr)
+			goto exit;
+
+		sysfs_add_file_to_group(&dev_root->kobj,
+					&dev_attr_low_power_idle_system_residency_us.attr,
+					"cpuidle");
+	} else if (info->gaddr.space_id == ACPI_ADR_SPACE_FIXED_HARDWARE) {
+		sysfs_add_file_to_group(&dev_root->kobj,
+					&dev_attr_low_power_idle_cpu_residency_us.attr,
+					"cpuidle");
+	}
+exit:
+	put_device(dev_root);
+}
+
+static void lpit_process(u64 begin, u64 end)
+{
+	while (begin + sizeof(struct acpi_lpit_native) <= end) {
+		struct acpi_lpit_native *lpit_native = (struct acpi_lpit_native *)begin;
+
+		if (!lpit_native->header.type && !lpit_native->header.flags) {
+			if (lpit_native->residency_counter.space_id == ACPI_ADR_SPACE_SYSTEM_MEMORY &&
+			    !residency_info_mem.gaddr.address) {
+				lpit_update_residency(&residency_info_mem, lpit_native);
+			} else if (lpit_native->residency_counter.space_id == ACPI_ADR_SPACE_FIXED_HARDWARE &&
+				   !residency_info_ffh.gaddr.address) {
+				lpit_update_residency(&residency_info_ffh, lpit_native);
+			}
+		}
+		begin += lpit_native->header.length;
+	}
+}
+
+void acpi_init_lpit(void)
+{
+	acpi_status status;
+	struct acpi_table_lpit *lpit;
+
+	status = acpi_get_table(ACPI_SIG_LPIT, 0, (struct acpi_table_header **)&lpit);
+	if (ACPI_FAILURE(status))
+		return;
+
+	lpit_process((u64)lpit + sizeof(*lpit),
+		     (u64)lpit + lpit->header.length);
+
+	acpi_put_table((struct acpi_table_header *)lpit);
+}