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, 259 insertions, 0 deletions

diff --git a/lib/raid/raid6/s390/pq_arch.h b/lib/raid/raid6/s390/pq_arch.h
new file mode 100644
index 000000000..95d14c342
--- /dev/null
+++ b/lib/raid/raid6/s390/pq_arch.h
@@ -0,0 +1,15 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later */
+
+#include <linux/cpufeature.h>
+
+extern const struct raid6_calls raid6_s390vx8;
+extern const struct raid6_recov_calls raid6_recov_s390xc;
+
+static __always_inline void __init arch_raid6_init(void)
+{
+	if (cpu_has_vx())
+		raid6_algo_add(&raid6_s390vx8);
+	else
+		raid6_algo_add_default();
+	raid6_recov_algo_add(&raid6_recov_s390xc);
+}
diff --git a/lib/raid/raid6/s390/recov_s390xc.c b/lib/raid/raid6/s390/recov_s390xc.c
new file mode 100644
index 000000000..08d56896e
--- /dev/null
+++ b/lib/raid/raid6/s390/recov_s390xc.c
@@ -0,0 +1,116 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * RAID-6 data recovery in dual failure mode based on the XC instruction.
+ *
+ * Copyright IBM Corp. 2016
+ * Author(s): Martin Schwidefsky <schwidefsky@de.ibm.com>
+ */
+
+#include <linux/mm.h>
+#include <linux/raid/pq.h>
+#include "algos.h"
+
+static inline void xor_block(u8 *p1, u8 *p2)
+{
+	typedef struct { u8 _[256]; } addrtype;
+
+	asm volatile(
+		"	xc	0(256,%[p1]),0(%[p2])\n"
+		: "+m" (*(addrtype *) p1) : "m" (*(addrtype *) p2),
+		  [p1] "a" (p1), [p2] "a" (p2) : "cc");
+}
+
+/* Recover two failed data blocks. */
+static void raid6_2data_recov_s390xc(int disks, size_t bytes, int faila,
+		int failb, void **ptrs)
+{
+	u8 *p, *q, *dp, *dq;
+	const u8 *pbmul;	/* P multiplier table for B data */
+	const u8 *qmul;		/* Q multiplier table (for both) */
+	int i;
+
+	p = (u8 *)ptrs[disks-2];
+	q = (u8 *)ptrs[disks-1];
+
+	/* Compute syndrome with zero for the missing data pages
+	   Use the dead data pages as temporary storage for
+	   delta p and delta q */
+	dp = (u8 *)ptrs[faila];
+	ptrs[faila] = page_address(ZERO_PAGE(0));
+	ptrs[disks-2] = dp;
+	dq = (u8 *)ptrs[failb];
+	ptrs[failb] = page_address(ZERO_PAGE(0));
+	ptrs[disks-1] = dq;
+
+	raid6_gen_syndrome(disks, bytes, ptrs);
+
+	/* Restore pointer table */
+	ptrs[faila]   = dp;
+	ptrs[failb]   = dq;
+	ptrs[disks-2] = p;
+	ptrs[disks-1] = q;
+
+	/* Now, pick the proper data tables */
+	pbmul = raid6_gfmul[raid6_gfexi[failb-faila]];
+	qmul  = raid6_gfmul[raid6_gfinv[raid6_gfexp[faila]^raid6_gfexp[failb]]];
+
+	/* Now do it... */
+	while (bytes) {
+		xor_block(dp, p);
+		xor_block(dq, q);
+		for (i = 0; i < 256; i++)
+			dq[i] = pbmul[dp[i]] ^ qmul[dq[i]];
+		xor_block(dp, dq);
+		p += 256;
+		q += 256;
+		dp += 256;
+		dq += 256;
+		bytes -= 256;
+	}
+}
+
+/* Recover failure of one data block plus the P block */
+static void raid6_datap_recov_s390xc(int disks, size_t bytes, int faila,
+		void **ptrs)
+{
+	u8 *p, *q, *dq;
+	const u8 *qmul;		/* Q multiplier table */
+	int i;
+
+	p = (u8 *)ptrs[disks-2];
+	q = (u8 *)ptrs[disks-1];
+
+	/* Compute syndrome with zero for the missing data page
+	   Use the dead data page as temporary storage for delta q */
+	dq = (u8 *)ptrs[faila];
+	ptrs[faila] = page_address(ZERO_PAGE(0));
+	ptrs[disks-1] = dq;
+
+	raid6_gen_syndrome(disks, bytes, ptrs);
+
+	/* Restore pointer table */
+	ptrs[faila]   = dq;
+	ptrs[disks-1] = q;
+
+	/* Now, pick the proper data tables */
+	qmul  = raid6_gfmul[raid6_gfinv[raid6_gfexp[faila]]];
+
+	/* Now do it... */
+	while (bytes) {
+		xor_block(dq, q);
+		for (i = 0; i < 256; i++)
+			dq[i] = qmul[dq[i]];
+		xor_block(p, dq);
+		p += 256;
+		q += 256;
+		dq += 256;
+		bytes -= 256;
+	}
+}
+
+
+const struct raid6_recov_calls raid6_recov_s390xc = {
+	.data2 = raid6_2data_recov_s390xc,
+	.datap = raid6_datap_recov_s390xc,
+	.name = "s390xc",
+};
diff --git a/lib/raid/raid6/s390/s390vx.uc b/lib/raid/raid6/s390/s390vx.uc
new file mode 100644
index 000000000..e5cf9054b
--- /dev/null
+++ b/lib/raid/raid6/s390/s390vx.uc
@@ -0,0 +1,128 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * raid6_vx$#.c
+ *
+ * $#-way unrolled RAID6 gen/xor functions for s390
+ * based on the vector facility
+ *
+ * Copyright IBM Corp. 2016
+ * Author(s): Martin Schwidefsky <schwidefsky@de.ibm.com>
+ *
+ * This file is postprocessed using unroll.awk.
+ */
+
+#include <linux/cpufeature.h>
+#include <asm/fpu.h>
+#include "algos.h"
+
+#define NSIZE 16
+
+static __always_inline void LOAD_CONST(void)
+{
+	fpu_vrepib(24, 0x07);
+	fpu_vrepib(25, 0x1d);
+}
+
+/*
+ * The SHLBYTE() operation shifts each of the 16 bytes in
+ * vector register y left by 1 bit and stores the result in
+ * vector register x.
+ */
+#define SHLBYTE(x, y)		fpu_vab(x, y, y)
+
+/*
+ * For each of the 16 bytes in the vector register y the MASK()
+ * operation returns 0xFF if the high bit of the byte is 1,
+ * or 0x00 if the high bit is 0. The result is stored in vector
+ * register x.
+ */
+#define MASK(x, y)		fpu_vesravb(x, y, 24)
+
+#define AND(x, y, z)		fpu_vn(x, y, z)
+#define XOR(x, y, z)		fpu_vx(x, y, z)
+#define LOAD_DATA(x, ptr)	fpu_vlm(x, x + $# - 1, ptr)
+#define STORE_DATA(x, ptr)	fpu_vstm(x, x + $# - 1, ptr)
+#define COPY_VEC(x, y)		fpu_vlr(x, y)
+
+static void raid6_s390vx$#_gen_syndrome(int disks, size_t bytes, void **ptrs)
+{
+	DECLARE_KERNEL_FPU_ONSTACK32(vxstate);
+	u8 **dptr, *p, *q;
+	int d, z, z0;
+
+	kernel_fpu_begin(&vxstate, KERNEL_VXR);
+	LOAD_CONST();
+
+	dptr = (u8 **) ptrs;
+	z0 = disks - 3;		/* Highest data disk */
+	p = dptr[z0 + 1];	/* XOR parity */
+	q = dptr[z0 + 2];	/* RS syndrome */
+
+	for (d = 0; d < bytes; d += $#*NSIZE) {
+		LOAD_DATA(0,&dptr[z0][d]);
+		COPY_VEC(8+$$,0+$$);
+		for (z = z0 - 1; z >= 0; z--) {
+			MASK(16+$$,8+$$);
+			AND(16+$$,16+$$,25);
+			SHLBYTE(8+$$,8+$$);
+			XOR(8+$$,8+$$,16+$$);
+			LOAD_DATA(16,&dptr[z][d]);
+			XOR(0+$$,0+$$,16+$$);
+			XOR(8+$$,8+$$,16+$$);
+		}
+		STORE_DATA(0,&p[d]);
+		STORE_DATA(8,&q[d]);
+	}
+	kernel_fpu_end(&vxstate, KERNEL_VXR);
+}
+
+static void raid6_s390vx$#_xor_syndrome(int disks, int start, int stop,
+					size_t bytes, void **ptrs)
+{
+	DECLARE_KERNEL_FPU_ONSTACK32(vxstate);
+	u8 **dptr, *p, *q;
+	int d, z, z0;
+
+	dptr = (u8 **) ptrs;
+	z0 = stop;		/* P/Q right side optimization */
+	p = dptr[disks - 2];	/* XOR parity */
+	q = dptr[disks - 1];	/* RS syndrome */
+
+	kernel_fpu_begin(&vxstate, KERNEL_VXR);
+	LOAD_CONST();
+
+	for (d = 0; d < bytes; d += $#*NSIZE) {
+		/* P/Q data pages */
+		LOAD_DATA(0,&dptr[z0][d]);
+		COPY_VEC(8+$$,0+$$);
+		for (z = z0 - 1; z >= start; z--) {
+			MASK(16+$$,8+$$);
+			AND(16+$$,16+$$,25);
+			SHLBYTE(8+$$,8+$$);
+			XOR(8+$$,8+$$,16+$$);
+			LOAD_DATA(16,&dptr[z][d]);
+			XOR(0+$$,0+$$,16+$$);
+			XOR(8+$$,8+$$,16+$$);
+		}
+		/* P/Q left side optimization */
+		for (z = start - 1; z >= 0; z--) {
+			MASK(16+$$,8+$$);
+			AND(16+$$,16+$$,25);
+			SHLBYTE(8+$$,8+$$);
+			XOR(8+$$,8+$$,16+$$);
+		}
+		LOAD_DATA(16,&p[d]);
+		XOR(16+$$,16+$$,0+$$);
+		STORE_DATA(16,&p[d]);
+		LOAD_DATA(16,&q[d]);
+		XOR(16+$$,16+$$,8+$$);
+		STORE_DATA(16,&q[d]);
+	}
+	kernel_fpu_end(&vxstate, KERNEL_VXR);
+}
+
+const struct raid6_calls raid6_s390vx$# = {
+	.gen_syndrome	= raid6_s390vx$#_gen_syndrome,
+	.xor_syndrome	= raid6_s390vx$#_xor_syndrome,
+	.name		= "vx128x$#",
+};