Received-SPF: pass (google.com: domain of linux-kernel+bounces-126335-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:40f1:3f00::1 as permitted sender) client-ip=2604:1380:40f1:3f00::1;
From: Borislav Petkov <bp@alien8.de>
To: X86 ML <x86@kernel.org>
Cc: LKML <linux-kernel@vger.kernel.org>
Subject: [PATCH v2 1/4] x86/alternatives: Use a temporary buffer when optimizing NOPs
Date: Sun, 31 Mar 2024 20:00:39 +0200
Message-ID: <20240331180042.13933-2-bp@alien8.de>
In-Reply-To: <20240331180042.13933-1-bp@alien8.de>
References: <20240331180042.13933-1-bp@alien8.de>
Precedence: bulk
MIME-Version: 1.0
Content-Transfer-Encoding: quoted-printable

From: "Borislav Petkov (AMD)" <bp@alien8.de>

Instead of optimizing NOPs in-place, use a temporary buffer like the
usual alternatives patching flow does. This obviates the need to grab
locks when patching, see

  6778977590da ("x86/alternatives: Disable interrupts and sync when optim=
izing NOPs in place")

While at it, add nomenclature definitions clarifying and simplifying the
naming of function-local variables in the alternatives code.

Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
---
 arch/x86/include/asm/text-patching.h |  2 +-
 arch/x86/kernel/alternative.c        | 84 +++++++++++++++-------------
 arch/x86/kernel/callthunks.c         |  9 +--
 3 files changed, 49 insertions(+), 46 deletions(-)

diff --git a/arch/x86/include/asm/text-patching.h b/arch/x86/include/asm/=
text-patching.h
index 345aafbc1964..6259f1937fe7 100644
--- a/arch/x86/include/asm/text-patching.h
+++ b/arch/x86/include/asm/text-patching.h
@@ -15,7 +15,7 @@
=20
 extern void text_poke_early(void *addr, const void *opcode, size_t len);
=20
-extern void apply_relocation(u8 *buf, size_t len, u8 *dest, u8 *src, siz=
e_t src_len);
+extern void apply_relocation(u8 *buf, const u8 * const instr, size_t ins=
trlen, u8 *repl, size_t repl_len);
=20
 /*
  * Clear and restore the kernel write-protection flag on the local CPU.
diff --git a/arch/x86/kernel/alternative.c b/arch/x86/kernel/alternative.=
c
index 45a280f2161c..ec94f1359c00 100644
--- a/arch/x86/kernel/alternative.c
+++ b/arch/x86/kernel/alternative.c
@@ -124,6 +124,20 @@ const unsigned char * const x86_nops[ASM_NOP_MAX+1] =
=3D
 #endif
 };
=20
+/*
+ * Nomenclature for variable names to simplify and clarify this code and=
 ease
+ * any potential staring at it:
+ *
+ * @instr: source address of the original instructions in the kernel tex=
t as
+ * generated by the compiler.
+ *
+ * @buf: temporary buffer on which the patching operates. This buffer is
+ * eventually text-poked into the kernel image.
+ *
+ * @replacement/@repl: pointer to the opcodes which are replacing @instr=
, located
+ * in the .altinstr_replacement section.
+ */
+
 /*
  * Fill the buffer with a single effective instruction of size @len.
  *
@@ -133,28 +147,28 @@ const unsigned char * const x86_nops[ASM_NOP_MAX+1]=
 =3D
  * each single-byte NOPs). If @len to fill out is > ASM_NOP_MAX, pad wit=
h INT3 and
  * *jump* over instead of executing long and daft NOPs.
  */
-static void add_nop(u8 *instr, unsigned int len)
+static void add_nop(u8 *buf, unsigned int len)
 {
-	u8 *target =3D instr + len;
+	u8 *target =3D buf + len;
=20
 	if (!len)
 		return;
=20
 	if (len <=3D ASM_NOP_MAX) {
-		memcpy(instr, x86_nops[len], len);
+		memcpy(buf, x86_nops[len], len);
 		return;
 	}
=20
 	if (len < 128) {
-		__text_gen_insn(instr, JMP8_INSN_OPCODE, instr, target, JMP8_INSN_SIZE=
);
-		instr +=3D JMP8_INSN_SIZE;
+		__text_gen_insn(buf, JMP8_INSN_OPCODE, buf, target, JMP8_INSN_SIZE);
+		buf +=3D JMP8_INSN_SIZE;
 	} else {
-		__text_gen_insn(instr, JMP32_INSN_OPCODE, instr, target, JMP32_INSN_SI=
ZE);
-		instr +=3D JMP32_INSN_SIZE;
+		__text_gen_insn(buf, JMP32_INSN_OPCODE, buf, target, JMP32_INSN_SIZE);
+		buf +=3D JMP32_INSN_SIZE;
 	}
=20
-	for (;instr < target; instr++)
-		*instr =3D INT3_INSN_OPCODE;
+	for (;buf < target; buf++)
+		*buf =3D INT3_INSN_OPCODE;
 }
=20
 extern s32 __retpoline_sites[], __retpoline_sites_end[];
@@ -187,12 +201,12 @@ static bool insn_is_nop(struct insn *insn)
  * Find the offset of the first non-NOP instruction starting at @offset
  * but no further than @len.
  */
-static int skip_nops(u8 *instr, int offset, int len)
+static int skip_nops(u8 *buf, int offset, int len)
 {
 	struct insn insn;
=20
 	for (; offset < len; offset +=3D insn.length) {
-		if (insn_decode_kernel(&insn, &instr[offset]))
+		if (insn_decode_kernel(&insn, &buf[offset]))
 			break;
=20
 		if (!insn_is_nop(&insn))
@@ -207,7 +221,7 @@ static int skip_nops(u8 *instr, int offset, int len)
  * to the end of the NOP sequence into a single NOP.
  */
 static bool
-__optimize_nops(u8 *instr, size_t len, struct insn *insn, int *next, int=
 *prev, int *target)
+__optimize_nops(const u8 * const instr, u8 *buf, size_t len, struct insn=
 *insn, int *next, int *prev, int *target)
 {
 	int i =3D *next - insn->length;
=20
@@ -222,12 +236,12 @@ __optimize_nops(u8 *instr, size_t len, struct insn =
*insn, int *next, int *prev,
 	if (insn_is_nop(insn)) {
 		int nop =3D i;
=20
-		*next =3D skip_nops(instr, *next, len);
+		*next =3D skip_nops(buf, *next, len);
 		if (*target && *next =3D=3D *target)
 			nop =3D *prev;
=20
-		add_nop(instr + nop, *next - nop);
-		DUMP_BYTES(ALT, instr, len, "%px: [%d:%d) optimized NOPs: ", instr, no=
p, *next);
+		add_nop(buf + nop, *next - nop);
+		DUMP_BYTES(ALT, buf, len, "%px: [%d:%d) optimized NOPs: ", instr, nop,=
 *next);
 		return true;
 	}
=20
@@ -239,32 +253,22 @@ __optimize_nops(u8 *instr, size_t len, struct insn =
*insn, int *next, int *prev,
  * "noinline" to cause control flow change and thus invalidate I$ and
  * cause refetch after modification.
  */
-static void __init_or_module noinline optimize_nops(u8 *instr, size_t le=
n)
+static void __init_or_module noinline optimize_nops(const u8 * const ins=
tr, u8 *buf, size_t len)
 {
 	int prev, target =3D 0;
=20
 	for (int next, i =3D 0; i < len; i =3D next) {
 		struct insn insn;
=20
-		if (insn_decode_kernel(&insn, &instr[i]))
+		if (insn_decode_kernel(&insn, &buf[i]))
 			return;
=20
 		next =3D i + insn.length;
=20
-		__optimize_nops(instr, len, &insn, &next, &prev, &target);
+		__optimize_nops(instr, buf, len, &insn, &next, &prev, &target);
 	}
 }
=20
-static void __init_or_module noinline optimize_nops_inplace(u8 *instr, s=
ize_t len)
-{
-	unsigned long flags;
-
-	local_irq_save(flags);
-	optimize_nops(instr, len);
-	sync_core();
-	local_irq_restore(flags);
-}
-
 /*
  * In this context, "source" is where the instructions are placed in the
  * section .altinstr_replacement, for example during kernel build by the
@@ -335,11 +339,11 @@ bool need_reloc(unsigned long offset, u8 *src, size=
_t src_len)
 	return (target < src || target > src + src_len);
 }
=20
-void apply_relocation(u8 *buf, size_t len, u8 *dest, u8 *src, size_t src=
_len)
+void apply_relocation(u8 *buf, const u8 * const instr, size_t instrlen, =
u8 *repl, size_t repl_len)
 {
 	int prev, target =3D 0;
=20
-	for (int next, i =3D 0; i < len; i =3D next) {
+	for (int next, i =3D 0; i < instrlen; i =3D next) {
 		struct insn insn;
=20
 		if (WARN_ON_ONCE(insn_decode_kernel(&insn, &buf[i])))
@@ -347,7 +351,7 @@ void apply_relocation(u8 *buf, size_t len, u8 *dest, =
u8 *src, size_t src_len)
=20
 		next =3D i + insn.length;
=20
-		if (__optimize_nops(buf, len, &insn, &next, &prev, &target))
+		if (__optimize_nops(instr, buf, instrlen, &insn, &next, &prev, &target=
))
 			continue;
=20
 		switch (insn.opcode.bytes[0]) {
@@ -361,10 +365,10 @@ void apply_relocation(u8 *buf, size_t len, u8 *dest=
, u8 *src, size_t src_len)
 		case JMP8_INSN_OPCODE:
 		case JMP32_INSN_OPCODE:
 		case CALL_INSN_OPCODE:
-			if (need_reloc(next + insn.immediate.value, src, src_len)) {
+			if (need_reloc(next + insn.immediate.value, repl, repl_len)) {
 				apply_reloc(insn.immediate.nbytes,
 					    buf + i + insn_offset_immediate(&insn),
-					    src - dest);
+					    repl - instr);
 			}
=20
 			/*
@@ -372,7 +376,7 @@ void apply_relocation(u8 *buf, size_t len, u8 *dest, =
u8 *src, size_t src_len)
 			 */
 			if (insn.opcode.bytes[0] =3D=3D JMP32_INSN_OPCODE) {
 				s32 imm =3D insn.immediate.value;
-				imm +=3D src - dest;
+				imm +=3D repl - instr;
 				imm +=3D JMP32_INSN_SIZE - JMP8_INSN_SIZE;
 				if ((imm >> 31) =3D=3D (imm >> 7)) {
 					buf[i+0] =3D JMP8_INSN_OPCODE;
@@ -385,10 +389,10 @@ void apply_relocation(u8 *buf, size_t len, u8 *dest=
, u8 *src, size_t src_len)
 		}
=20
 		if (insn_rip_relative(&insn)) {
-			if (need_reloc(next + insn.displacement.value, src, src_len)) {
+			if (need_reloc(next + insn.displacement.value, repl, repl_len)) {
 				apply_reloc(insn.displacement.nbytes,
 					    buf + i + insn_offset_displacement(&insn),
-					    src - dest);
+					    repl - instr);
 			}
 		}
 	}
@@ -504,7 +508,9 @@ void __init_or_module noinline apply_alternatives(str=
uct alt_instr *start,
 		 *   patch if feature is *NOT* present.
 		 */
 		if (!boot_cpu_has(a->cpuid) =3D=3D !(a->flags & ALT_FLAG_NOT)) {
-			optimize_nops_inplace(instr, a->instrlen);
+			memcpy(insn_buff, instr, a->instrlen);
+			optimize_nops(instr, insn_buff, a->instrlen);
+			text_poke_early(instr, insn_buff, a->instrlen);
 			continue;
 		}
=20
@@ -526,7 +532,7 @@ void __init_or_module noinline apply_alternatives(str=
uct alt_instr *start,
 		for (; insn_buff_sz < a->instrlen; insn_buff_sz++)
 			insn_buff[insn_buff_sz] =3D 0x90;
=20
-		apply_relocation(insn_buff, a->instrlen, instr, replacement, a->replac=
ementlen);
+		apply_relocation(insn_buff, instr, a->instrlen, replacement, a->replac=
ementlen);
=20
 		DUMP_BYTES(ALT, instr, a->instrlen, "%px:   old_insn: ", instr);
 		DUMP_BYTES(ALT, replacement, a->replacementlen, "%px:   rpl_insn: ", r=
eplacement);
@@ -761,7 +767,7 @@ void __init_or_module noinline apply_retpolines(s32 *=
start, s32 *end)
=20
 		len =3D patch_retpoline(addr, &insn, bytes);
 		if (len =3D=3D insn.length) {
-			optimize_nops(bytes, len);
+			optimize_nops(addr, bytes, len);
 			DUMP_BYTES(RETPOLINE, ((u8*)addr),  len, "%px: orig: ", addr);
 			DUMP_BYTES(RETPOLINE, ((u8*)bytes), len, "%px: repl: ", addr);
 			text_poke_early(addr, bytes, len);
diff --git a/arch/x86/kernel/callthunks.c b/arch/x86/kernel/callthunks.c
index 30335182b6b0..771d95484453 100644
--- a/arch/x86/kernel/callthunks.c
+++ b/arch/x86/kernel/callthunks.c
@@ -185,8 +185,7 @@ static void *patch_dest(void *dest, bool direct)
 	u8 *pad =3D dest - tsize;
=20
 	memcpy(insn_buff, skl_call_thunk_template, tsize);
-	apply_relocation(insn_buff, tsize, pad,
-			 skl_call_thunk_template, tsize);
+	apply_relocation(insn_buff, pad, tsize, skl_call_thunk_template, tsize)=
;
=20
 	/* Already patched? */
 	if (!bcmp(pad, insn_buff, tsize))
@@ -308,8 +307,7 @@ static bool is_callthunk(void *addr)
 	pad =3D (void *)(dest - tmpl_size);
=20
 	memcpy(insn_buff, skl_call_thunk_template, tmpl_size);
-	apply_relocation(insn_buff, tmpl_size, pad,
-			 skl_call_thunk_template, tmpl_size);
+	apply_relocation(insn_buff, pad, tmpl_size, skl_call_thunk_template, tm=
pl_size);
=20
 	return !bcmp(pad, insn_buff, tmpl_size);
 }
@@ -327,8 +325,7 @@ int x86_call_depth_emit_accounting(u8 **pprog, void *=
func)
 		return 0;
=20
 	memcpy(insn_buff, skl_call_thunk_template, tmpl_size);
-	apply_relocation(insn_buff, tmpl_size, *pprog,
-			 skl_call_thunk_template, tmpl_size);
+	apply_relocation(insn_buff, *pprog, tmpl_size, skl_call_thunk_template,=
 tmpl_size);
=20
 	memcpy(*pprog, insn_buff, tmpl_size);
 	*pprog +=3D tmpl_size;
--=20
2.43.0