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, 144 insertions, 0 deletions

diff --git a/arch/arc/lib/memset-archs.S b/arch/arc/lib/memset-archs.S
new file mode 100644
index 000000000..d0a5cec4c
--- /dev/null
+++ b/arch/arc/lib/memset-archs.S
@@ -0,0 +1,144 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Copyright (C) 2014-15 Synopsys, Inc. (www.synopsys.com)
+ */
+
+#include <linux/linkage.h>
+#include <asm/cache.h>
+
+/*
+ * The memset implementation below is optimized to use prefetchw and prealloc
+ * instruction in case of CPU with 64B L1 data cache line (L1_CACHE_SHIFT == 6)
+ * If you want to implement optimized memset for other possible L1 data cache
+ * line lengths (32B and 128B) you should rewrite code carefully checking
+ * we don't call any prefetchw/prealloc instruction for L1 cache lines which
+ * don't belongs to memset area.
+ */
+
+#if L1_CACHE_SHIFT == 6
+
+.macro PREALLOC_INSTR	reg, off
+	prealloc	[\reg, \off]
+.endm
+
+.macro PREFETCHW_INSTR	reg, off
+	prefetchw	[\reg, \off]
+.endm
+
+#else
+
+.macro PREALLOC_INSTR	reg, off
+.endm
+
+.macro PREFETCHW_INSTR	reg, off
+.endm
+
+#endif
+
+ENTRY_CFI(memset)
+	mov.f	0, r2
+;;; if size is zero
+	jz.d	[blink]
+	mov	r3, r0		; don't clobber ret val
+
+	PREFETCHW_INSTR	r0, 0	; Prefetch the first write location
+
+;;; if length < 8
+	brls.d.nt	r2, 8, .Lsmallchunk
+	mov.f	lp_count,r2
+
+	and.f	r4, r0, 0x03
+	rsub	lp_count, r4, 4
+	lpnz	@.Laligndestination
+	;; LOOP BEGIN
+	stb.ab	r1, [r3,1]
+	sub	r2, r2, 1
+.Laligndestination:
+
+;;; Destination is aligned
+	and	r1, r1, 0xFF
+	asl	r4, r1, 8
+	or	r4, r4, r1
+	asl	r5, r4, 16
+	or	r5, r5, r4
+	mov	r4, r5
+
+	sub3	lp_count, r2, 8
+	cmp     r2, 64
+	bmsk.hi	r2, r2, 5
+	mov.ls	lp_count, 0
+	add3.hi	r2, r2, 8
+
+;;; Convert len to Dwords, unfold x8
+	lsr.f	lp_count, lp_count, 6
+
+	lpnz	@.Lset64bytes
+	;; LOOP START
+	PREALLOC_INSTR	r3, 64	; alloc next line w/o fetching
+
+#ifdef CONFIG_ARC_HAS_LL64
+	std.ab	r4, [r3, 8]
+	std.ab	r4, [r3, 8]
+	std.ab	r4, [r3, 8]
+	std.ab	r4, [r3, 8]
+	std.ab	r4, [r3, 8]
+	std.ab	r4, [r3, 8]
+	std.ab	r4, [r3, 8]
+	std.ab	r4, [r3, 8]
+#else
+	st.ab	r4, [r3, 4]
+	st.ab	r4, [r3, 4]
+	st.ab	r4, [r3, 4]
+	st.ab	r4, [r3, 4]
+	st.ab	r4, [r3, 4]
+	st.ab	r4, [r3, 4]
+	st.ab	r4, [r3, 4]
+	st.ab	r4, [r3, 4]
+	st.ab	r4, [r3, 4]
+	st.ab	r4, [r3, 4]
+	st.ab	r4, [r3, 4]
+	st.ab	r4, [r3, 4]
+	st.ab	r4, [r3, 4]
+	st.ab	r4, [r3, 4]
+	st.ab	r4, [r3, 4]
+	st.ab	r4, [r3, 4]
+#endif
+.Lset64bytes:
+
+	lsr.f	lp_count, r2, 5 ;Last remaining  max 124 bytes
+	lpnz	.Lset32bytes
+	;; LOOP START
+#ifdef CONFIG_ARC_HAS_LL64
+	std.ab	r4, [r3, 8]
+	std.ab	r4, [r3, 8]
+	std.ab	r4, [r3, 8]
+	std.ab	r4, [r3, 8]
+#else
+	st.ab	r4, [r3, 4]
+	st.ab	r4, [r3, 4]
+	st.ab	r4, [r3, 4]
+	st.ab	r4, [r3, 4]
+	st.ab	r4, [r3, 4]
+	st.ab	r4, [r3, 4]
+	st.ab	r4, [r3, 4]
+	st.ab	r4, [r3, 4]
+#endif
+.Lset32bytes:
+
+	and.f	lp_count, r2, 0x1F ;Last remaining 31 bytes
+.Lsmallchunk:
+	lpnz	.Lcopy3bytes
+	;; LOOP START
+	stb.ab	r1, [r3, 1]
+.Lcopy3bytes:
+
+	j	[blink]
+
+END_CFI(memset)
+
+ENTRY_CFI(memzero)
+    ; adjust bzero args to memset args
+    mov r2, r1
+    b.d  memset    ;tail call so need to tinker with blink
+    mov r1, 0
+END_CFI(memzero)