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, 153 insertions, 0 deletions

diff --git a/scripts/Makefile.dtbs b/scripts/Makefile.dtbs
new file mode 100644
index 000000000..c4e466390
--- /dev/null
+++ b/scripts/Makefile.dtbs
@@ -0,0 +1,153 @@
+# SPDX-License-Identifier: GPL-2.0-only
+
+all-dtb := $(dtb-y) $(dtb-)
+
+# If CONFIG_OF_ALL_DTBS is enabled, all DT blobs are built
+dtb-$(CONFIG_OF_ALL_DTBS) += $(dtb-)
+
+# Composite DTB (i.e. DTB constructed by overlay)
+multi-dtb-y := $(call multi-search, $(dtb-y), .dtb, -dtbs)
+# Primitive DTB compiled from *.dts
+real-dtb-y := $(call real-search, $(dtb-y), .dtb, -dtbs)
+# Base DTB that overlay is applied onto
+base-dtb-y := $(filter %.dtb, $(call real-search, $(multi-dtb-y), .dtb, -dtbs))
+
+# Ensure that any .dtbo is applied to at least one base .dtb. Otherwise, it
+# does not get validated.
+applied-dtbo := $(filter %.dtbo, \
+	$(call real-search, $(call multi-search, $(all-dtb), .dtb, -dtbs), .dtb, -dtbs))
+unapplied-dtbo := $(filter-out $(applied-dtbo),$(filter %.dtbo, $(dtb-y)))
+$(if $(unapplied-dtbo), $(warning .dtbo is not applied to any base: $(unapplied-dtbo)))
+
+dtb-y           := $(addprefix $(obj)/, $(dtb-y))
+multi-dtb-y     := $(addprefix $(obj)/, $(multi-dtb-y))
+real-dtb-y      := $(addprefix $(obj)/, $(real-dtb-y))
+
+always-y        += $(dtb-y)
+targets         += $(real-dtb-y)
+
+# dtbs-list
+# ---------------------------------------------------------------------------
+
+ifdef need-dtbslist
+subdir-dtbslist := $(addsuffix /dtbs-list, $(subdir-ym))
+dtb-y           += $(subdir-dtbslist)
+always-y        += $(obj)/dtbs-list
+endif
+
+$(subdir-dtbslist): $(obj)/%/dtbs-list: $(obj)/% ;
+
+$(obj)/dtbs-list: $(dtb-y) FORCE
+	$(call if_changed,gen_order)
+
+# Assembly file to wrap dtb(o)
+# ---------------------------------------------------------------------------
+
+builtin-dtb-section = $(if $(filter arch/$(SRCARCH)/boot/dts%, $(obj)),.dtb.init.rodata,.rodata)
+
+# Generate an assembly file to wrap the output of the device tree compiler
+quiet_cmd_wrap_S_dtb = WRAP    $@
+      cmd_wrap_S_dtb = {								\
+		symbase=__$(patsubst .%,%,$(suffix $<))_$(subst -,_,$(notdir $*));	\
+		echo '\#include <asm-generic/vmlinux.lds.h>';				\
+		echo '.section $(builtin-dtb-section),"a"';				\
+		echo '.balign STRUCT_ALIGNMENT';					\
+		echo ".global $${symbase}_begin";					\
+		echo "$${symbase}_begin:";						\
+		echo '.incbin "$<" ';							\
+		echo ".global $${symbase}_end";						\
+		echo "$${symbase}_end:";						\
+		echo '.balign STRUCT_ALIGNMENT';					\
+	} > $@
+
+$(obj)/%.dtb.S: $(obj)/%.dtb FORCE
+	$(call if_changed,wrap_S_dtb)
+
+$(obj)/%.dtbo.S: $(obj)/%.dtbo FORCE
+	$(call if_changed,wrap_S_dtb)
+
+# Schema check
+# ---------------------------------------------------------------------------
+
+ifneq ($(CHECK_DTBS),)
+DT_CHECKER ?= dt-validate
+DT_CHECKER_FLAGS ?= $(if $(DT_SCHEMA_FILES),-l $(DT_SCHEMA_FILES),-m)
+DT_BINDING_DIR := Documentation/devicetree/bindings
+DT_TMP_SCHEMA := $(objtree)/$(DT_BINDING_DIR)/processed-schema.json
+dtb-check-enabled = $(if $(filter %.dtb, $@),y)
+endif
+
+quiet_dtb_check_tag = $(if $(dtb-check-enabled),[C],   )
+cmd_dtb_check = $(if $(dtb-check-enabled),; $(DT_CHECKER) $(DT_CHECKER_FLAGS) -u $(srctree)/$(DT_BINDING_DIR) -p $(DT_TMP_SCHEMA) $@ || true)
+
+# Overlay
+# ---------------------------------------------------------------------------
+
+# NOTE:
+# Do not replace $(filter %.dtb %.dtbo, $^) with $(real-prereqs). When a single
+# DTB is turned into a multi-blob DTB, $^ will contain header file dependencies
+# recorded in the .*.cmd file.
+quiet_cmd_fdtoverlay = OVL $(quiet_dtb_check_tag) $@
+      cmd_fdtoverlay = $(objtree)/scripts/dtc/fdtoverlay -o $@ -i $(filter %.dtb %.dtbo, $^) $(cmd_dtb_check)
+
+$(multi-dtb-y): $(DT_TMP_SCHEMA) FORCE
+	$(call if_changed,fdtoverlay)
+$(call multi_depend, $(multi-dtb-y), .dtb, -dtbs)
+
+# DTC
+# ---------------------------------------------------------------------------
+
+DTC ?= $(objtree)/scripts/dtc/dtc
+DTC_FLAGS += -Wno-unique_unit_address
+
+# Disable noisy checks by default
+ifeq ($(findstring 1,$(KBUILD_EXTRA_WARN)),)
+DTC_FLAGS += -Wno-unit_address_vs_reg \
+             -Wno-avoid_unnecessary_addr_size \
+             -Wno-alias_paths \
+             -Wno-interrupt_map \
+             -Wno-simple_bus_reg
+else
+DTC_FLAGS += -Wunique_unit_address_if_enabled
+endif
+
+ifneq ($(findstring 2,$(KBUILD_EXTRA_WARN)),)
+DTC_FLAGS += -Wnode_name_chars_strict \
+             -Wproperty_name_chars_strict \
+             -Wunique_unit_address
+endif
+
+DTC_FLAGS += $(DTC_FLAGS_$(target-stem))
+
+# Set -@ if the target is a base DTB that overlay is applied onto
+DTC_FLAGS += $(if $(filter $(patsubst $(obj)/%,%,$@), $(base-dtb-y)), -@)
+
+DTC_INCLUDE := $(srctree)/scripts/dtc/include-prefixes
+
+dtc_cpp_flags = -Wp,-MMD,$(depfile).pre.tmp -nostdinc -I $(DTC_INCLUDE) -undef -D__DTS__
+
+dtc-tmp = $(subst $(comma),_,$(dot-target).dts.tmp)
+
+quiet_cmd_dtc = DTC $(quiet_dtb_check_tag) $@
+      cmd_dtc = \
+	$(HOSTCC) -E $(dtc_cpp_flags) -x assembler-with-cpp -o $(dtc-tmp) $< ; \
+	$(DTC) -o $@ -b 0 $(addprefix -i,$(dir $<) $(DTC_INCLUDE)) \
+	       $(DTC_FLAGS) -d $(depfile).dtc.tmp $(dtc-tmp) ; \
+	cat $(depfile).pre.tmp $(depfile).dtc.tmp > $(depfile) \
+	$(cmd_dtb_check)
+
+$(obj)/%.dtb: $(obj)/%.dts $(DTC) $(DT_TMP_SCHEMA) FORCE
+	$(call if_changed_dep,dtc)
+
+$(obj)/%.dtbo: $(src)/%.dtso $(DTC) FORCE
+	$(call if_changed_dep,dtc)
+
+# targets
+# ---------------------------------------------------------------------------
+
+targets += $(always-y)
+
+# %.dtb.o <- %.dtb.S <- %.dtb <- %.dts
+# %.dtbo.o <- %.dtbo.S <- %.dtbo <- %.dtso
+targets += $(call intermediate_targets, .dtb.o, .dtb.S .dtb) \
+           $(call intermediate_targets, .dtbo.o, .dtbo.S .dtbo)