[PATCH 0/2] xor: enable auto-vectorization in Clang

[PATCH 0/2] xor: enable auto-vectorization in Clang

[Ard Biesheuvel]

Update the xor_blocks() prototypes so that the compiler understands that
the inputs always refer to distinct regions of memory. This is implied
by the existing implementations, as they use different granularities for
the load/xor/store loops.

With that, we can fix the ARM/Clang version, which refuses to SIMD
vectorize otherwise, and throws a spurious warning related to the GCC
version being incompatible.

Cc: Nick Desaulniers <ndesaulniers@google.com>
Cc: Nathan Chancellor <nathan@kernel.org>

Ard Biesheuvel (2):
  lib/xor: make xor prototypes more friendely to compiler vectorization
  crypto: arm/xor - make vectorized C code Clang-friendly

 arch/alpha/include/asm/xor.h           | 53 ++++++++----
 arch/arm/include/asm/xor.h             | 42 ++++++----
 arch/arm/lib/xor-neon.c                | 12 +--
 arch/arm64/include/asm/xor.h           | 21 +++--
 arch/arm64/lib/xor-neon.c              | 23 +++---
 arch/ia64/include/asm/xor.h            | 21 +++--
 arch/powerpc/include/asm/xor_altivec.h | 25 +++---
 arch/powerpc/lib/xor_vmx.h             | 27 ++++---
 arch/powerpc/lib/xor_vmx_glue.c        | 32 ++++----
 arch/s390/lib/xor.c                    | 21 +++--
 arch/sparc/include/asm/xor_32.h        | 21 +++--
 arch/sparc/include/asm/xor_64.h        | 42 ++++++----
 arch/x86/include/asm/xor.h             | 42 ++++++----
 arch/x86/include/asm/xor_32.h          | 42 ++++++----
 arch/x86/include/asm/xor_avx.h         | 21 +++--
 include/asm-generic/xor.h              | 84 +++++++++++++-------
 include/linux/raid/xor.h               | 21 +++--
 17 files changed, 351 insertions(+), 199 deletions(-)


base-commit: e783362eb54cd99b2cac8b3a9aeac942e6f6ac07
-- 
2.30.2

[PATCH 1/2] lib/xor: make xor prototypes more friendely to compiler vectorization

[Ard Biesheuvel]

Modern compilers are perfectly capable of extracting parallelism from
the XOR routines, provided that the prototypes reflect the nature of the
input accurately, in particular, the fact that the input vectors are
expected not to overlap. This is not documented explicitly, but is
implied by the interchangeability of the various C routines, some of
which use temporary variables while others don't: this means that these
routines only behave identically for non-overlapping inputs.

So let's decorate these input vectors with the __restrict modifier,
which informs the compiler that there is no overlap. While at it, make
the input-only vectors pointer-to-const as well.

Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
---
 arch/alpha/include/asm/xor.h           | 53 ++++++++----
 arch/arm/include/asm/xor.h             | 42 ++++++----
 arch/arm64/include/asm/xor.h           | 21 +++--
 arch/arm64/lib/xor-neon.c              | 23 +++---
 arch/ia64/include/asm/xor.h            | 21 +++--
 arch/powerpc/include/asm/xor_altivec.h | 25 +++---
 arch/powerpc/lib/xor_vmx.h             | 27 ++++---
 arch/powerpc/lib/xor_vmx_glue.c        | 32 ++++----
 arch/s390/lib/xor.c                    | 21 +++--
 arch/sparc/include/asm/xor_32.h        | 21 +++--
 arch/sparc/include/asm/xor_64.h        | 42 ++++++----
 arch/x86/include/asm/xor.h             | 42 ++++++----
 arch/x86/include/asm/xor_32.h          | 42 ++++++----
 arch/x86/include/asm/xor_avx.h         | 21 +++--
 include/asm-generic/xor.h              | 84 +++++++++++++-------
 include/linux/raid/xor.h               | 21 +++--
 16 files changed, 348 insertions(+), 190 deletions(-)

diff --git a/arch/alpha/include/asm/xor.h b/arch/alpha/include/asm/xor.h
index 5aeb4fb3cb7c..d5bc2bf37792 100644
--- a/arch/alpha/include/asm/xor.h
+++ b/arch/alpha/include/asm/xor.h
@@ -5,24 +5,43 @@
  * Optimized RAID-5 checksumming functions for alpha EV5 and EV6
  */
 
-extern void xor_alpha_2(unsigned long, unsigned long *, unsigned long *);
-extern void xor_alpha_3(unsigned long, unsigned long *, unsigned long *,
-		        unsigned long *);
-extern void xor_alpha_4(unsigned long, unsigned long *, unsigned long *,
-		        unsigned long *, unsigned long *);
-extern void xor_alpha_5(unsigned long, unsigned long *, unsigned long *,
-		        unsigned long *, unsigned long *, unsigned long *);
+extern void
+xor_alpha_2(unsigned long bytes, unsigned long * __restrict p1,
+	    const unsigned long * __restrict p2);
+extern void
+xor_alpha_3(unsigned long bytes, unsigned long * __restrict p1,
+	    const unsigned long * __restrict p2,
+	    const unsigned long * __restrict p3);
+extern void
+xor_alpha_4(unsigned long bytes, unsigned long * __restrict p1,
+	    const unsigned long * __restrict p2,
+	    const unsigned long * __restrict p3,
+	    const unsigned long * __restrict p4);
+extern void
+xor_alpha_5(unsigned long bytes, unsigned long * __restrict p1,
+	    const unsigned long * __restrict p2,
+	    const unsigned long * __restrict p3,
+	    const unsigned long * __restrict p4,
+	    const unsigned long * __restrict p5);
 
-extern void xor_alpha_prefetch_2(unsigned long, unsigned long *,
-				 unsigned long *);
-extern void xor_alpha_prefetch_3(unsigned long, unsigned long *,
-				 unsigned long *, unsigned long *);
-extern void xor_alpha_prefetch_4(unsigned long, unsigned long *,
-				 unsigned long *, unsigned long *,
-				 unsigned long *);
-extern void xor_alpha_prefetch_5(unsigned long, unsigned long *,
-				 unsigned long *, unsigned long *,
-				 unsigned long *, unsigned long *);
+extern void
+xor_alpha_prefetch_2(unsigned long bytes, unsigned long * __restrict p1,
+		     const unsigned long * __restrict p2);
+extern void
+xor_alpha_prefetch_3(unsigned long bytes, unsigned long * __restrict p1,
+		     const unsigned long * __restrict p2,
+		     const unsigned long * __restrict p3);
+extern void
+xor_alpha_prefetch_4(unsigned long bytes, unsigned long * __restrict p1,
+		     const unsigned long * __restrict p2,
+		     const unsigned long * __restrict p3,
+		     const unsigned long * __restrict p4);
+extern void
+xor_alpha_prefetch_5(unsigned long bytes, unsigned long * __restrict p1,
+		     const unsigned long * __restrict p2,
+		     const unsigned long * __restrict p3,
+		     const unsigned long * __restrict p4,
+		     const unsigned long * __restrict p5);
 
 asm("								\n\
 	.text							\n\
diff --git a/arch/arm/include/asm/xor.h b/arch/arm/include/asm/xor.h
index aefddec79286..d01b4601ba43 100644
--- a/arch/arm/include/asm/xor.h
+++ b/arch/arm/include/asm/xor.h
@@ -44,7 +44,8 @@
 		: "0" (dst), "r" (a1), "r" (a2), "r" (a3), "r" (a4))
 
 static void
-xor_arm4regs_2(unsigned long bytes, unsigned long *p1, unsigned long *p2)
+xor_arm4regs_2(unsigned long bytes, unsigned long * __restrict p1,
+	       const unsigned long * __restrict p2)
 {
 	unsigned int lines = bytes / sizeof(unsigned long) / 4;
 	register unsigned int a1 __asm__("r4");
@@ -64,8 +65,9 @@ xor_arm4regs_2(unsigned long bytes, unsigned long *p1, unsigned long *p2)
 }
 
 static void
-xor_arm4regs_3(unsigned long bytes, unsigned long *p1, unsigned long *p2,
-		unsigned long *p3)
+xor_arm4regs_3(unsigned long bytes, unsigned long * __restrict p1,
+	       const unsigned long * __restrict p2,
+	       const unsigned long * __restrict p3)
 {
 	unsigned int lines = bytes / sizeof(unsigned long) / 4;
 	register unsigned int a1 __asm__("r4");
@@ -86,8 +88,10 @@ xor_arm4regs_3(unsigned long bytes, unsigned long *p1, unsigned long *p2,
 }
 
 static void
-xor_arm4regs_4(unsigned long bytes, unsigned long *p1, unsigned long *p2,
-		unsigned long *p3, unsigned long *p4)
+xor_arm4regs_4(unsigned long bytes, unsigned long * __restrict p1,
+	       const unsigned long * __restrict p2,
+	       const unsigned long * __restrict p3,
+	       const unsigned long * __restrict p4)
 {
 	unsigned int lines = bytes / sizeof(unsigned long) / 2;
 	register unsigned int a1 __asm__("r8");
@@ -105,8 +109,11 @@ xor_arm4regs_4(unsigned long bytes, unsigned long *p1, unsigned long *p2,
 }
 
 static void
-xor_arm4regs_5(unsigned long bytes, unsigned long *p1, unsigned long *p2,
-		unsigned long *p3, unsigned long *p4, unsigned long *p5)
+xor_arm4regs_5(unsigned long bytes, unsigned long * __restrict p1,
+	       const unsigned long * __restrict p2,
+	       const unsigned long * __restrict p3,
+	       const unsigned long * __restrict p4,
+	       const unsigned long * __restrict p5)
 {
 	unsigned int lines = bytes / sizeof(unsigned long) / 2;
 	register unsigned int a1 __asm__("r8");
@@ -146,7 +153,8 @@ static struct xor_block_template xor_block_arm4regs = {
 extern struct xor_block_template const xor_block_neon_inner;
 
 static void
-xor_neon_2(unsigned long bytes, unsigned long *p1, unsigned long *p2)
+xor_neon_2(unsigned long bytes, unsigned long * __restrict p1,
+	   const unsigned long * __restrict p2)
 {
 	if (in_interrupt()) {
 		xor_arm4regs_2(bytes, p1, p2);
@@ -158,8 +166,9 @@ xor_neon_2(unsigned long bytes, unsigned long *p1, unsigned long *p2)
 }
 
 static void
-xor_neon_3(unsigned long bytes, unsigned long *p1, unsigned long *p2,
-		unsigned long *p3)
+xor_neon_3(unsigned long bytes, unsigned long * __restrict p1,
+	   const unsigned long * __restrict p2,
+	   const unsigned long * __restrict p3)
 {
 	if (in_interrupt()) {
 		xor_arm4regs_3(bytes, p1, p2, p3);
@@ -171,8 +180,10 @@ xor_neon_3(unsigned long bytes, unsigned long *p1, unsigned long *p2,
 }
 
 static void
-xor_neon_4(unsigned long bytes, unsigned long *p1, unsigned long *p2,
-		unsigned long *p3, unsigned long *p4)
+xor_neon_4(unsigned long bytes, unsigned long * __restrict p1,
+	   const unsigned long * __restrict p2,
+	   const unsigned long * __restrict p3,
+	   const unsigned long * __restrict p4)
 {
 	if (in_interrupt()) {
 		xor_arm4regs_4(bytes, p1, p2, p3, p4);
@@ -184,8 +195,11 @@ xor_neon_4(unsigned long bytes, unsigned long *p1, unsigned long *p2,
 }
 
 static void
-xor_neon_5(unsigned long bytes, unsigned long *p1, unsigned long *p2,
-		unsigned long *p3, unsigned long *p4, unsigned long *p5)
+xor_neon_5(unsigned long bytes, unsigned long * __restrict p1,
+	   const unsigned long * __restrict p2,
+	   const unsigned long * __restrict p3,
+	   const unsigned long * __restrict p4,
+	   const unsigned long * __restrict p5)
 {
 	if (in_interrupt()) {
 		xor_arm4regs_5(bytes, p1, p2, p3, p4, p5);
diff --git a/arch/arm64/include/asm/xor.h b/arch/arm64/include/asm/xor.h
index 947f6a4f1aa0..c55f862f89ec 100644
--- a/arch/arm64/include/asm/xor.h
+++ b/arch/arm64/include/asm/xor.h
@@ -16,7 +16,8 @@
 extern struct xor_block_template const xor_block_inner_neon;
 
 static void
-xor_neon_2(unsigned long bytes, unsigned long *p1, unsigned long *p2)
+xor_neon_2(unsigned long bytes, unsigned long * __restrict p1,
+	   const unsigned long * __restrict p2)
 {
 	kernel_neon_begin();
 	xor_block_inner_neon.do_2(bytes, p1, p2);
@@ -24,8 +25,9 @@ xor_neon_2(unsigned long bytes, unsigned long *p1, unsigned long *p2)
 }
 
 static void
-xor_neon_3(unsigned long bytes, unsigned long *p1, unsigned long *p2,
-		unsigned long *p3)
+xor_neon_3(unsigned long bytes, unsigned long * __restrict p1,
+	   const unsigned long * __restrict p2,
+	   const unsigned long * __restrict p3)
 {
 	kernel_neon_begin();
 	xor_block_inner_neon.do_3(bytes, p1, p2, p3);
@@ -33,8 +35,10 @@ xor_neon_3(unsigned long bytes, unsigned long *p1, unsigned long *p2,
 }
 
 static void
-xor_neon_4(unsigned long bytes, unsigned long *p1, unsigned long *p2,
-		unsigned long *p3, unsigned long *p4)
+xor_neon_4(unsigned long bytes, unsigned long * __restrict p1,
+	   const unsigned long * __restrict p2,
+	   const unsigned long * __restrict p3,
+	   const unsigned long * __restrict p4)
 {
 	kernel_neon_begin();
 	xor_block_inner_neon.do_4(bytes, p1, p2, p3, p4);
@@ -42,8 +46,11 @@ xor_neon_4(unsigned long bytes, unsigned long *p1, unsigned long *p2,
 }
 
 static void
-xor_neon_5(unsigned long bytes, unsigned long *p1, unsigned long *p2,
-		unsigned long *p3, unsigned long *p4, unsigned long *p5)
+xor_neon_5(unsigned long bytes, unsigned long * __restrict p1,
+	   const unsigned long * __restrict p2,
+	   const unsigned long * __restrict p3,
+	   const unsigned long * __restrict p4,
+	   const unsigned long * __restrict p5)
 {
 	kernel_neon_begin();
 	xor_block_inner_neon.do_5(bytes, p1, p2, p3, p4, p5);
diff --git a/arch/arm64/lib/xor-neon.c b/arch/arm64/lib/xor-neon.c
index d189cf4e70ea..64c9577fcf0f 100644
--- a/arch/arm64/lib/xor-neon.c
+++ b/arch/arm64/lib/xor-neon.c
@@ -10,8 +10,8 @@
 #include <linux/module.h>
 #include <asm/neon-intrinsics.h>
 
-void xor_arm64_neon_2(unsigned long bytes, unsigned long *p1,
-	unsigned long *p2)
+void xor_arm64_neon_2(unsigned long bytes, unsigned long * __restrict p1,
+	const unsigned long * __restrict p2)
 {
 	uint64_t *dp1 = (uint64_t *)p1;
 	uint64_t *dp2 = (uint64_t *)p2;
@@ -37,8 +37,9 @@ void xor_arm64_neon_2(unsigned long bytes, unsigned long *p1,
 	} while (--lines > 0);
 }
 
-void xor_arm64_neon_3(unsigned long bytes, unsigned long *p1,
-	unsigned long *p2, unsigned long *p3)
+void xor_arm64_neon_3(unsigned long bytes, unsigned long * __restrict p1,
+	const unsigned long * __restrict p2,
+	const unsigned long * __restrict p3)
 {
 	uint64_t *dp1 = (uint64_t *)p1;
 	uint64_t *dp2 = (uint64_t *)p2;
@@ -72,8 +73,10 @@ void xor_arm64_neon_3(unsigned long bytes, unsigned long *p1,
 	} while (--lines > 0);
 }
 
-void xor_arm64_neon_4(unsigned long bytes, unsigned long *p1,
-	unsigned long *p2, unsigned long *p3, unsigned long *p4)
+void xor_arm64_neon_4(unsigned long bytes, unsigned long * __restrict p1,
+	const unsigned long * __restrict p2,
+	const unsigned long * __restrict p3,
+	const unsigned long * __restrict p4)
 {
 	uint64_t *dp1 = (uint64_t *)p1;
 	uint64_t *dp2 = (uint64_t *)p2;
@@ -115,9 +118,11 @@ void xor_arm64_neon_4(unsigned long bytes, unsigned long *p1,
 	} while (--lines > 0);
 }
 
-void xor_arm64_neon_5(unsigned long bytes, unsigned long *p1,
-	unsigned long *p2, unsigned long *p3,
-	unsigned long *p4, unsigned long *p5)
+void xor_arm64_neon_5(unsigned long bytes, unsigned long * __restrict p1,
+	const unsigned long * __restrict p2,
+	const unsigned long * __restrict p3,
+	const unsigned long * __restrict p4,
+	const unsigned long * __restrict p5)
 {
 	uint64_t *dp1 = (uint64_t *)p1;
 	uint64_t *dp2 = (uint64_t *)p2;
diff --git a/arch/ia64/include/asm/xor.h b/arch/ia64/include/asm/xor.h
index 673051bf9d7d..117a29868ba2 100644
--- a/arch/ia64/include/asm/xor.h
+++ b/arch/ia64/include/asm/xor.h
@@ -4,13 +4,20 @@
  */
 
 
-extern void xor_ia64_2(unsigned long, unsigned long *, unsigned long *);
-extern void xor_ia64_3(unsigned long, unsigned long *, unsigned long *,
-		       unsigned long *);
-extern void xor_ia64_4(unsigned long, unsigned long *, unsigned long *,
-		       unsigned long *, unsigned long *);
-extern void xor_ia64_5(unsigned long, unsigned long *, unsigned long *,
-		       unsigned long *, unsigned long *, unsigned long *);
+extern void xor_ia64_2(unsigned long bytes, unsigned long * __restrict p1,
+		       const unsigned long * __restrict p2);
+extern void xor_ia64_3(unsigned long bytes, unsigned long * __restrict p1,
+		       const unsigned long * __restrict p2,
+		       const unsigned long * __restrict p3);
+extern void xor_ia64_4(unsigned long bytes, unsigned long * __restrict p1,
+		       const unsigned long * __restrict p2,
+		       const unsigned long * __restrict p3,
+		       const unsigned long * __restrict p4);
+extern void xor_ia64_5(unsigned long bytes, unsigned long * __restrict p1,
+		       const unsigned long * __restrict p2,
+		       const unsigned long * __restrict p3,
+		       const unsigned long * __restrict p4,
+		       const unsigned long * __restrict p5);
 
 static struct xor_block_template xor_block_ia64 = {
 	.name =	"ia64",
diff --git a/arch/powerpc/include/asm/xor_altivec.h b/arch/powerpc/include/asm/xor_altivec.h
index 6ca923510b59..9b2cb12a3145 100644
--- a/arch/powerpc/include/asm/xor_altivec.h
+++ b/arch/powerpc/include/asm/xor_altivec.h
@@ -3,17 +3,20 @@
 #define _ASM_POWERPC_XOR_ALTIVEC_H
 
 #ifdef CONFIG_ALTIVEC
-
-void xor_altivec_2(unsigned long bytes, unsigned long *v1_in,
-		   unsigned long *v2_in);
-void xor_altivec_3(unsigned long bytes, unsigned long *v1_in,
-		   unsigned long *v2_in, unsigned long *v3_in);
-void xor_altivec_4(unsigned long bytes, unsigned long *v1_in,
-		   unsigned long *v2_in, unsigned long *v3_in,
-		   unsigned long *v4_in);
-void xor_altivec_5(unsigned long bytes, unsigned long *v1_in,
-		   unsigned long *v2_in, unsigned long *v3_in,
-		   unsigned long *v4_in, unsigned long *v5_in);
+void xor_altivec_2(unsigned long bytes, unsigned long * __restrict p1,
+		   const unsigned long * __restrict p2);
+void xor_altivec_3(unsigned long bytes, unsigned long * __restrict p1,
+		   const unsigned long * __restrict p2,
+		   const unsigned long * __restrict p3);
+void xor_altivec_4(unsigned long bytes, unsigned long * __restrict p1,
+		   const unsigned long * __restrict p2,
+		   const unsigned long * __restrict p3,
+		   const unsigned long * __restrict p4);
+void xor_altivec_5(unsigned long bytes, unsigned long * __restrict p1,
+		   const unsigned long * __restrict p2,
+		   const unsigned long * __restrict p3,
+		   const unsigned long * __restrict p4,
+		   const unsigned long * __restrict p5);
 
 #endif
 #endif /* _ASM_POWERPC_XOR_ALTIVEC_H */
diff --git a/arch/powerpc/lib/xor_vmx.h b/arch/powerpc/lib/xor_vmx.h
index 5c2b0839b179..6342eafe0ef3 100644
--- a/arch/powerpc/lib/xor_vmx.h
+++ b/arch/powerpc/lib/xor_vmx.h
@@ -6,16 +6,17 @@
  * outside of the enable/disable altivec block.
  */
 
-void __xor_altivec_2(unsigned long bytes, unsigned long *v1_in,
-			     unsigned long *v2_in);
-
-void __xor_altivec_3(unsigned long bytes, unsigned long *v1_in,
-			     unsigned long *v2_in, unsigned long *v3_in);
-
-void __xor_altivec_4(unsigned long bytes, unsigned long *v1_in,
-			     unsigned long *v2_in, unsigned long *v3_in,
-			     unsigned long *v4_in);
-
-void __xor_altivec_5(unsigned long bytes, unsigned long *v1_in,
-			     unsigned long *v2_in, unsigned long *v3_in,
-			     unsigned long *v4_in, unsigned long *v5_in);
+void __xor_altivec_2(unsigned long bytes, unsigned long * __restrict p1,
+		     const unsigned long * __restrict p2);
+void __xor_altivec_3(unsigned long bytes, unsigned long * __restrict p1,
+		     const unsigned long * __restrict p2,
+		     const unsigned long * __restrict p3);
+void __xor_altivec_4(unsigned long bytes, unsigned long * __restrict p1,
+		     const unsigned long * __restrict p2,
+		     const unsigned long * __restrict p3,
+		     const unsigned long * __restrict p4);
+void __xor_altivec_5(unsigned long bytes, unsigned long * __restrict p1,
+		     const unsigned long * __restrict p2,
+		     const unsigned long * __restrict p3,
+		     const unsigned long * __restrict p4,
+		     const unsigned long * __restrict p5);
diff --git a/arch/powerpc/lib/xor_vmx_glue.c b/arch/powerpc/lib/xor_vmx_glue.c
index 80dba916c367..baae85c6733a 100644
--- a/arch/powerpc/lib/xor_vmx_glue.c
+++ b/arch/powerpc/lib/xor_vmx_glue.c
@@ -12,47 +12,51 @@
 #include <asm/xor_altivec.h>
 #include "xor_vmx.h"
 
-void xor_altivec_2(unsigned long bytes, unsigned long *v1_in,
-		   unsigned long *v2_in)
+void xor_altivec_2(unsigned long bytes, unsigned long * __restrict p1,
+		   const unsigned long * __restrict p2)
 {
 	preempt_disable();
 	enable_kernel_altivec();
-	__xor_altivec_2(bytes, v1_in, v2_in);
+	__xor_altivec_2(bytes, p1, p2);
 	disable_kernel_altivec();
 	preempt_enable();
 }
 EXPORT_SYMBOL(xor_altivec_2);
 
-void xor_altivec_3(unsigned long bytes,  unsigned long *v1_in,
-		   unsigned long *v2_in, unsigned long *v3_in)
+void xor_altivec_3(unsigned long bytes, unsigned long * __restrict p1,
+		   const unsigned long * __restrict p2,
+		   const unsigned long * __restrict p3)
 {
 	preempt_disable();
 	enable_kernel_altivec();
-	__xor_altivec_3(bytes, v1_in, v2_in, v3_in);
+	__xor_altivec_3(bytes, p1, p2, p3);
 	disable_kernel_altivec();
 	preempt_enable();
 }
 EXPORT_SYMBOL(xor_altivec_3);
 
-void xor_altivec_4(unsigned long bytes,  unsigned long *v1_in,
-		   unsigned long *v2_in, unsigned long *v3_in,
-		   unsigned long *v4_in)
+void xor_altivec_4(unsigned long bytes, unsigned long * __restrict p1,
+		   const unsigned long * __restrict p2,
+		   const unsigned long * __restrict p3,
+		   const unsigned long * __restrict p4)
 {
 	preempt_disable();
 	enable_kernel_altivec();
-	__xor_altivec_4(bytes, v1_in, v2_in, v3_in, v4_in);
+	__xor_altivec_4(bytes, p1, p2, p3, p4);
 	disable_kernel_altivec();
 	preempt_enable();
 }
 EXPORT_SYMBOL(xor_altivec_4);
 
-void xor_altivec_5(unsigned long bytes,  unsigned long *v1_in,
-		   unsigned long *v2_in, unsigned long *v3_in,
-		   unsigned long *v4_in, unsigned long *v5_in)
+void xor_altivec_5(unsigned long bytes, unsigned long * __restrict p1,
+		   const unsigned long * __restrict p2,
+		   const unsigned long * __restrict p3,
+		   const unsigned long * __restrict p4,
+		   const unsigned long * __restrict p5)
 {
 	preempt_disable();
 	enable_kernel_altivec();
-	__xor_altivec_5(bytes, v1_in, v2_in, v3_in, v4_in, v5_in);
+	__xor_altivec_5(bytes, p1, p2, p3, p4, p5);
 	disable_kernel_altivec();
 	preempt_enable();
 }
diff --git a/arch/s390/lib/xor.c b/arch/s390/lib/xor.c
index a963c3d8ad0d..a4d09fa8cca6 100644
--- a/arch/s390/lib/xor.c
+++ b/arch/s390/lib/xor.c
@@ -11,7 +11,8 @@
 #include <linux/raid/xor.h>
 #include <asm/xor.h>
 
-static void xor_xc_2(unsigned long bytes, unsigned long *p1, unsigned long *p2)
+static void xor_xc_2(unsigned long bytes, unsigned long * __restrict p1,
+		     const unsigned long * __restrict p2)
 {
 	asm volatile(
 		"	larl	1,2f\n"
@@ -32,8 +33,9 @@ static void xor_xc_2(unsigned long bytes, unsigned long *p1, unsigned long *p2)
 		: "0", "1", "cc", "memory");
 }
 
-static void xor_xc_3(unsigned long bytes, unsigned long *p1, unsigned long *p2,
-		     unsigned long *p3)
+static void xor_xc_3(unsigned long bytes, unsigned long * __restrict p1,
+		     const unsigned long * __restrict p2,
+		     const unsigned long * __restrict p3)
 {
 	asm volatile(
 		"	larl	1,2f\n"
@@ -58,8 +60,10 @@ static void xor_xc_3(unsigned long bytes, unsigned long *p1, unsigned long *p2,
 		: : "0", "1", "cc", "memory");
 }
 
-static void xor_xc_4(unsigned long bytes, unsigned long *p1, unsigned long *p2,
-		     unsigned long *p3, unsigned long *p4)
+static void xor_xc_4(unsigned long bytes, unsigned long * __restrict p1,
+		     const unsigned long * __restrict p2,
+		     const unsigned long * __restrict p3,
+		     const unsigned long * __restrict p4)
 {
 	asm volatile(
 		"	larl	1,2f\n"
@@ -88,8 +92,11 @@ static void xor_xc_4(unsigned long bytes, unsigned long *p1, unsigned long *p2,
 		: : "0", "1", "cc", "memory");
 }
 
-static void xor_xc_5(unsigned long bytes, unsigned long *p1, unsigned long *p2,
-		     unsigned long *p3, unsigned long *p4, unsigned long *p5)
+static void xor_xc_5(unsigned long bytes, unsigned long * __restrict p1,
+		     const unsigned long * __restrict p2,
+		     const unsigned long * __restrict p3,
+		     const unsigned long * __restrict p4,
+		     const unsigned long * __restrict p5)
 {
 	asm volatile(
 		"	larl	1,2f\n"
diff --git a/arch/sparc/include/asm/xor_32.h b/arch/sparc/include/asm/xor_32.h
index 3e5af37e4b9c..cfbad7a4663d 100644
--- a/arch/sparc/include/asm/xor_32.h
+++ b/arch/sparc/include/asm/xor_32.h
@@ -13,7 +13,8 @@
  */
 
 static void
-sparc_2(unsigned long bytes, unsigned long *p1, unsigned long *p2)
+sparc_2(unsigned long bytes, unsigned long * __restrict p1,
+	const unsigned long * __restrict p2)
 {
 	int lines = bytes / (sizeof (long)) / 8;
 
@@ -50,8 +51,9 @@ sparc_2(unsigned long bytes, unsigned long *p1, unsigned long *p2)
 }
 
 static void
-sparc_3(unsigned long bytes, unsigned long *p1, unsigned long *p2,
-	unsigned long *p3)
+sparc_3(unsigned long bytes, unsigned long * __restrict p1,
+	const unsigned long * __restrict p2,
+	const unsigned long * __restrict p3)
 {
 	int lines = bytes / (sizeof (long)) / 8;
 
@@ -101,8 +103,10 @@ sparc_3(unsigned long bytes, unsigned long *p1, unsigned long *p2,
 }
 
 static void
-sparc_4(unsigned long bytes, unsigned long *p1, unsigned long *p2,
-	unsigned long *p3, unsigned long *p4)
+sparc_4(unsigned long bytes, unsigned long * __restrict p1,
+	const unsigned long * __restrict p2,
+	const unsigned long * __restrict p3,
+	const unsigned long * __restrict p4)
 {
 	int lines = bytes / (sizeof (long)) / 8;
 
@@ -165,8 +169,11 @@ sparc_4(unsigned long bytes, unsigned long *p1, unsigned long *p2,
 }
 
 static void
-sparc_5(unsigned long bytes, unsigned long *p1, unsigned long *p2,
-	unsigned long *p3, unsigned long *p4, unsigned long *p5)
+sparc_5(unsigned long bytes, unsigned long * __restrict p1,
+	const unsigned long * __restrict p2,
+	const unsigned long * __restrict p3,
+	const unsigned long * __restrict p4,
+	const unsigned long * __restrict p5)
 {
 	int lines = bytes / (sizeof (long)) / 8;
 
diff --git a/arch/sparc/include/asm/xor_64.h b/arch/sparc/include/asm/xor_64.h
index 16169f3edcd5..c782c9413517 100644
--- a/arch/sparc/include/asm/xor_64.h
+++ b/arch/sparc/include/asm/xor_64.h
@@ -12,13 +12,20 @@
 
 #include <asm/spitfire.h>
 
-void xor_vis_2(unsigned long, unsigned long *, unsigned long *);
-void xor_vis_3(unsigned long, unsigned long *, unsigned long *,
-	       unsigned long *);
-void xor_vis_4(unsigned long, unsigned long *, unsigned long *,
-	       unsigned long *, unsigned long *);
-void xor_vis_5(unsigned long, unsigned long *, unsigned long *,
-	       unsigned long *, unsigned long *, unsigned long *);
+void xor_vis_2(unsigned long bytes, unsigned long * __restrict p1,
+	       const unsigned long * __restrict p2);
+void xor_vis_3(unsigned long bytes, unsigned long * __restrict p1,
+	       const unsigned long * __restrict p2,
+	       const unsigned long * __restrict p3);
+void xor_vis_4(unsigned long bytes, unsigned long * __restrict p1,
+	       const unsigned long * __restrict p2,
+	       const unsigned long * __restrict p3,
+	       const unsigned long * __restrict p4);
+void xor_vis_5(unsigned long bytes, unsigned long * __restrict p1,
+	       const unsigned long * __restrict p2,
+	       const unsigned long * __restrict p3,
+	       const unsigned long * __restrict p4,
+	       const unsigned long * __restrict p5);
 
 /* XXX Ugh, write cheetah versions... -DaveM */
 
@@ -30,13 +37,20 @@ static struct xor_block_template xor_block_VIS = {
         .do_5	= xor_vis_5,
 };
 
-void xor_niagara_2(unsigned long, unsigned long *, unsigned long *);
-void xor_niagara_3(unsigned long, unsigned long *, unsigned long *,
-		   unsigned long *);
-void xor_niagara_4(unsigned long, unsigned long *, unsigned long *,
-		   unsigned long *, unsigned long *);
-void xor_niagara_5(unsigned long, unsigned long *, unsigned long *,
-		   unsigned long *, unsigned long *, unsigned long *);
+void xor_niagara_2(unsigned long bytes, unsigned long * __restrict p1,
+		   const unsigned long * __restrict p2);
+void xor_niagara_3(unsigned long bytes, unsigned long * __restrict p1,
+		   const unsigned long * __restrict p2,
+		   const unsigned long * __restrict p3);
+void xor_niagara_4(unsigned long bytes, unsigned long * __restrict p1,
+		   const unsigned long * __restrict p2,
+		   const unsigned long * __restrict p3,
+		   const unsigned long * __restrict p4);
+void xor_niagara_5(unsigned long bytes, unsigned long * __restrict p1,
+		   const unsigned long * __restrict p2,
+		   const unsigned long * __restrict p3,
+		   const unsigned long * __restrict p4,
+		   const unsigned long * __restrict p5);
 
 static struct xor_block_template xor_block_niagara = {
         .name	= "Niagara",
diff --git a/arch/x86/include/asm/xor.h b/arch/x86/include/asm/xor.h
index 2ee95a7769e6..2bf347d3616a 100644
--- a/arch/x86/include/asm/xor.h
+++ b/arch/x86/include/asm/xor.h
@@ -57,7 +57,8 @@
 					op(i + 3, 3)
 
 static void
-xor_sse_2(unsigned long bytes, unsigned long *p1, unsigned long *p2)
+xor_sse_2(unsigned long bytes, unsigned long * __restrict p1,
+	  const unsigned long * __restrict p2)
 {
 	unsigned long lines = bytes >> 8;
 
@@ -108,7 +109,8 @@ xor_sse_2(unsigned long bytes, unsigned long *p1, unsigned long *p2)
 }
 
 static void
-xor_sse_2_pf64(unsigned long bytes, unsigned long *p1, unsigned long *p2)
+xor_sse_2_pf64(unsigned long bytes, unsigned long * __restrict p1,
+	       const unsigned long * __restrict p2)
 {
 	unsigned long lines = bytes >> 8;
 
@@ -142,8 +144,9 @@ xor_sse_2_pf64(unsigned long bytes, unsigned long *p1, unsigned long *p2)
 }
 
 static void
-xor_sse_3(unsigned long bytes, unsigned long *p1, unsigned long *p2,
-	  unsigned long *p3)
+xor_sse_3(unsigned long bytes, unsigned long * __restrict p1,
+	  const unsigned long * __restrict p2,
+	  const unsigned long * __restrict p3)
 {
 	unsigned long lines = bytes >> 8;
 
@@ -201,8 +204,9 @@ xor_sse_3(unsigned long bytes, unsigned long *p1, unsigned long *p2,
 }
 
 static void
-xor_sse_3_pf64(unsigned long bytes, unsigned long *p1, unsigned long *p2,
-	       unsigned long *p3)
+xor_sse_3_pf64(unsigned long bytes, unsigned long * __restrict p1,
+	       const unsigned long * __restrict p2,
+	       const unsigned long * __restrict p3)
 {
 	unsigned long lines = bytes >> 8;
 
@@ -238,8 +242,10 @@ xor_sse_3_pf64(unsigned long bytes, unsigned long *p1, unsigned long *p2,
 }
 
 static void
-xor_sse_4(unsigned long bytes, unsigned long *p1, unsigned long *p2,
-	  unsigned long *p3, unsigned long *p4)
+xor_sse_4(unsigned long bytes, unsigned long * __restrict p1,
+	  const unsigned long * __restrict p2,
+	  const unsigned long * __restrict p3,
+	  const unsigned long * __restrict p4)
 {
 	unsigned long lines = bytes >> 8;
 
@@ -304,8 +310,10 @@ xor_sse_4(unsigned long bytes, unsigned long *p1, unsigned long *p2,
 }
 
 static void
-xor_sse_4_pf64(unsigned long bytes, unsigned long *p1, unsigned long *p2,
-	       unsigned long *p3, unsigned long *p4)
+xor_sse_4_pf64(unsigned long bytes, unsigned long * __restrict p1,
+	       const unsigned long * __restrict p2,
+	       const unsigned long * __restrict p3,
+	       const unsigned long * __restrict p4)
 {
 	unsigned long lines = bytes >> 8;
 
@@ -343,8 +351,11 @@ xor_sse_4_pf64(unsigned long bytes, unsigned long *p1, unsigned long *p2,
 }
 
 static void
-xor_sse_5(unsigned long bytes, unsigned long *p1, unsigned long *p2,
-	  unsigned long *p3, unsigned long *p4, unsigned long *p5)
+xor_sse_5(unsigned long bytes, unsigned long * __restrict p1,
+	  const unsigned long * __restrict p2,
+	  const unsigned long * __restrict p3,
+	  const unsigned long * __restrict p4,
+	  const unsigned long * __restrict p5)
 {
 	unsigned long lines = bytes >> 8;
 
@@ -416,8 +427,11 @@ xor_sse_5(unsigned long bytes, unsigned long *p1, unsigned long *p2,
 }
 
 static void
-xor_sse_5_pf64(unsigned long bytes, unsigned long *p1, unsigned long *p2,
-	       unsigned long *p3, unsigned long *p4, unsigned long *p5)
+xor_sse_5_pf64(unsigned long bytes, unsigned long * __restrict p1,
+	       const unsigned long * __restrict p2,
+	       const unsigned long * __restrict p3,
+	       const unsigned long * __restrict p4,
+	       const unsigned long * __restrict p5)
 {
 	unsigned long lines = bytes >> 8;
 
diff --git a/arch/x86/include/asm/xor_32.h b/arch/x86/include/asm/xor_32.h
index 67ceb790e639..90600ded1491 100644
--- a/arch/x86/include/asm/xor_32.h
+++ b/arch/x86/include/asm/xor_32.h
@@ -21,7 +21,8 @@
 #include <asm/fpu/api.h>
 
 static void
-xor_pII_mmx_2(unsigned long bytes, unsigned long *p1, unsigned long *p2)
+xor_pII_mmx_2(unsigned long bytes, unsigned long * __restrict p1,
+	      const unsigned long * __restrict p2)
 {
 	unsigned long lines = bytes >> 7;
 
@@ -64,8 +65,9 @@ xor_pII_mmx_2(unsigned long bytes, unsigned long *p1, unsigned long *p2)
 }
 
 static void
-xor_pII_mmx_3(unsigned long bytes, unsigned long *p1, unsigned long *p2,
-	      unsigned long *p3)
+xor_pII_mmx_3(unsigned long bytes, unsigned long * __restrict p1,
+	      const unsigned long * __restrict p2,
+	      const unsigned long * __restrict p3)
 {
 	unsigned long lines = bytes >> 7;
 
@@ -113,8 +115,10 @@ xor_pII_mmx_3(unsigned long bytes, unsigned long *p1, unsigned long *p2,
 }
 
 static void
-xor_pII_mmx_4(unsigned long bytes, unsigned long *p1, unsigned long *p2,
-	      unsigned long *p3, unsigned long *p4)
+xor_pII_mmx_4(unsigned long bytes, unsigned long * __restrict p1,
+	      const unsigned long * __restrict p2,
+	      const unsigned long * __restrict p3,
+	      const unsigned long * __restrict p4)
 {
 	unsigned long lines = bytes >> 7;
 
@@ -168,8 +172,11 @@ xor_pII_mmx_4(unsigned long bytes, unsigned long *p1, unsigned long *p2,
 
 
 static void
-xor_pII_mmx_5(unsigned long bytes, unsigned long *p1, unsigned long *p2,
-	      unsigned long *p3, unsigned long *p4, unsigned long *p5)
+xor_pII_mmx_5(unsigned long bytes, unsigned long * __restrict p1,
+	      const unsigned long * __restrict p2,
+	      const unsigned long * __restrict p3,
+	      const unsigned long * __restrict p4,
+	      const unsigned long * __restrict p5)
 {
 	unsigned long lines = bytes >> 7;
 
@@ -248,7 +255,8 @@ xor_pII_mmx_5(unsigned long bytes, unsigned long *p1, unsigned long *p2,
 #undef BLOCK
 
 static void
-xor_p5_mmx_2(unsigned long bytes, unsigned long *p1, unsigned long *p2)
+xor_p5_mmx_2(unsigned long bytes, unsigned long * __restrict p1,
+	     const unsigned long * __restrict p2)
 {
 	unsigned long lines = bytes >> 6;
 
@@ -295,8 +303,9 @@ xor_p5_mmx_2(unsigned long bytes, unsigned long *p1, unsigned long *p2)
 }
 
 static void
-xor_p5_mmx_3(unsigned long bytes, unsigned long *p1, unsigned long *p2,
-	     unsigned long *p3)
+xor_p5_mmx_3(unsigned long bytes, unsigned long * __restrict p1,
+	     const unsigned long * __restrict p2,
+	     const unsigned long * __restrict p3)
 {
 	unsigned long lines = bytes >> 6;
 
@@ -352,8 +361,10 @@ xor_p5_mmx_3(unsigned long bytes, unsigned long *p1, unsigned long *p2,
 }
 
 static void
-xor_p5_mmx_4(unsigned long bytes, unsigned long *p1, unsigned long *p2,
-	     unsigned long *p3, unsigned long *p4)
+xor_p5_mmx_4(unsigned long bytes, unsigned long * __restrict p1,
+	     const unsigned long * __restrict p2,
+	     const unsigned long * __restrict p3,
+	     const unsigned long * __restrict p4)
 {
 	unsigned long lines = bytes >> 6;
 
@@ -418,8 +429,11 @@ xor_p5_mmx_4(unsigned long bytes, unsigned long *p1, unsigned long *p2,
 }
 
 static void
-xor_p5_mmx_5(unsigned long bytes, unsigned long *p1, unsigned long *p2,
-	     unsigned long *p3, unsigned long *p4, unsigned long *p5)
+xor_p5_mmx_5(unsigned long bytes, unsigned long * __restrict p1,
+	     const unsigned long * __restrict p2,
+	     const unsigned long * __restrict p3,
+	     const unsigned long * __restrict p4,
+	     const unsigned long * __restrict p5)
 {
 	unsigned long lines = bytes >> 6;
 
diff --git a/arch/x86/include/asm/xor_avx.h b/arch/x86/include/asm/xor_avx.h
index 0c4e5b5e3852..93f90a05f70d 100644
--- a/arch/x86/include/asm/xor_avx.h
+++ b/arch/x86/include/asm/xor_avx.h
@@ -26,7 +26,8 @@
 		BLOCK4(8) \
 		BLOCK4(12)
 
-static void xor_avx_2(unsigned long bytes, unsigned long *p0, unsigned long *p1)
+static void xor_avx_2(unsigned long bytes, unsigned long * __restrict p0,
+		      const unsigned long * __restrict p1)
 {
 	unsigned long lines = bytes >> 9;
 
@@ -52,8 +53,9 @@ do { \
 	kernel_fpu_end();
 }
 
-static void xor_avx_3(unsigned long bytes, unsigned long *p0, unsigned long *p1,
-	unsigned long *p2)
+static void xor_avx_3(unsigned long bytes, unsigned long * __restrict p0,
+		      const unsigned long * __restrict p1,
+		      const unsigned long * __restrict p2)
 {
 	unsigned long lines = bytes >> 9;
 
@@ -82,8 +84,10 @@ do { \
 	kernel_fpu_end();
 }
 
-static void xor_avx_4(unsigned long bytes, unsigned long *p0, unsigned long *p1,
-	unsigned long *p2, unsigned long *p3)
+static void xor_avx_4(unsigned long bytes, unsigned long * __restrict p0,
+		      const unsigned long * __restrict p1,
+		      const unsigned long * __restrict p2,
+		      const unsigned long * __restrict p3)
 {
 	unsigned long lines = bytes >> 9;
 
@@ -115,8 +119,11 @@ do { \
 	kernel_fpu_end();
 }
 
-static void xor_avx_5(unsigned long bytes, unsigned long *p0, unsigned long *p1,
-	unsigned long *p2, unsigned long *p3, unsigned long *p4)
+static void xor_avx_5(unsigned long bytes, unsigned long * __restrict p0,
+	     const unsigned long * __restrict p1,
+	     const unsigned long * __restrict p2,
+	     const unsigned long * __restrict p3,
+	     const unsigned long * __restrict p4)
 {
 	unsigned long lines = bytes >> 9;
 
diff --git a/include/asm-generic/xor.h b/include/asm-generic/xor.h
index b62a2a56a4d4..95b36a0f6c44 100644
--- a/include/asm-generic/xor.h
+++ b/include/asm-generic/xor.h
@@ -8,7 +8,8 @@
 #include <linux/prefetch.h>
 
 static void
-xor_8regs_2(unsigned long bytes, unsigned long *p1, unsigned long *p2)
+xor_8regs_2(unsigned long bytes, unsigned long * __restrict p1,
+	    const unsigned long * __restrict p2)
 {
 	long lines = bytes / (sizeof (long)) / 8;
 
@@ -27,8 +28,9 @@ xor_8regs_2(unsigned long bytes, unsigned long *p1, unsigned long *p2)
 }
 
 static void
-xor_8regs_3(unsigned long bytes, unsigned long *p1, unsigned long *p2,
-	    unsigned long *p3)
+xor_8regs_3(unsigned long bytes, unsigned long * __restrict p1,
+	    const unsigned long * __restrict p2,
+	    const unsigned long * __restrict p3)
 {
 	long lines = bytes / (sizeof (long)) / 8;
 
@@ -48,8 +50,10 @@ xor_8regs_3(unsigned long bytes, unsigned long *p1, unsigned long *p2,
 }
 
 static void
-xor_8regs_4(unsigned long bytes, unsigned long *p1, unsigned long *p2,
-	    unsigned long *p3, unsigned long *p4)
+xor_8regs_4(unsigned long bytes, unsigned long * __restrict p1,
+	    const unsigned long * __restrict p2,
+	    const unsigned long * __restrict p3,
+	    const unsigned long * __restrict p4)
 {
 	long lines = bytes / (sizeof (long)) / 8;
 
@@ -70,8 +74,11 @@ xor_8regs_4(unsigned long bytes, unsigned long *p1, unsigned long *p2,
 }
 
 static void
-xor_8regs_5(unsigned long bytes, unsigned long *p1, unsigned long *p2,
-	    unsigned long *p3, unsigned long *p4, unsigned long *p5)
+xor_8regs_5(unsigned long bytes, unsigned long * __restrict p1,
+	    const unsigned long * __restrict p2,
+	    const unsigned long * __restrict p3,
+	    const unsigned long * __restrict p4,
+	    const unsigned long * __restrict p5)
 {
 	long lines = bytes / (sizeof (long)) / 8;
 
@@ -93,7 +100,8 @@ xor_8regs_5(unsigned long bytes, unsigned long *p1, unsigned long *p2,
 }
 
 static void
-xor_32regs_2(unsigned long bytes, unsigned long *p1, unsigned long *p2)
+xor_32regs_2(unsigned long bytes, unsigned long * __restrict p1,
+	     const unsigned long * __restrict p2)
 {
 	long lines = bytes / (sizeof (long)) / 8;
 
@@ -129,8 +137,9 @@ xor_32regs_2(unsigned long bytes, unsigned long *p1, unsigned long *p2)
 }
 
 static void
-xor_32regs_3(unsigned long bytes, unsigned long *p1, unsigned long *p2,
-	    unsigned long *p3)
+xor_32regs_3(unsigned long bytes, unsigned long * __restrict p1,
+	     const unsigned long * __restrict p2,
+	     const unsigned long * __restrict p3)
 {
 	long lines = bytes / (sizeof (long)) / 8;
 
@@ -175,8 +184,10 @@ xor_32regs_3(unsigned long bytes, unsigned long *p1, unsigned long *p2,
 }
 
 static void
-xor_32regs_4(unsigned long bytes, unsigned long *p1, unsigned long *p2,
-	    unsigned long *p3, unsigned long *p4)
+xor_32regs_4(unsigned long bytes, unsigned long * __restrict p1,
+	     const unsigned long * __restrict p2,
+	     const unsigned long * __restrict p3,
+	     const unsigned long * __restrict p4)
 {
 	long lines = bytes / (sizeof (long)) / 8;
 
@@ -230,8 +241,11 @@ xor_32regs_4(unsigned long bytes, unsigned long *p1, unsigned long *p2,
 }
 
 static void
-xor_32regs_5(unsigned long bytes, unsigned long *p1, unsigned long *p2,
-	    unsigned long *p3, unsigned long *p4, unsigned long *p5)
+xor_32regs_5(unsigned long bytes, unsigned long * __restrict p1,
+	     const unsigned long * __restrict p2,
+	     const unsigned long * __restrict p3,
+	     const unsigned long * __restrict p4,
+	     const unsigned long * __restrict p5)
 {
 	long lines = bytes / (sizeof (long)) / 8;
 
@@ -294,7 +308,8 @@ xor_32regs_5(unsigned long bytes, unsigned long *p1, unsigned long *p2,
 }
 
 static void
-xor_8regs_p_2(unsigned long bytes, unsigned long *p1, unsigned long *p2)
+xor_8regs_p_2(unsigned long bytes, unsigned long * __restrict p1,
+	      const unsigned long * __restrict p2)
 {
 	long lines = bytes / (sizeof (long)) / 8 - 1;
 	prefetchw(p1);
@@ -320,8 +335,9 @@ xor_8regs_p_2(unsigned long bytes, unsigned long *p1, unsigned long *p2)
 }
 
 static void
-xor_8regs_p_3(unsigned long bytes, unsigned long *p1, unsigned long *p2,
-	    unsigned long *p3)
+xor_8regs_p_3(unsigned long bytes, unsigned long * __restrict p1,
+	      const unsigned long * __restrict p2,
+	      const unsigned long * __restrict p3)
 {
 	long lines = bytes / (sizeof (long)) / 8 - 1;
 	prefetchw(p1);
@@ -350,8 +366,10 @@ xor_8regs_p_3(unsigned long bytes, unsigned long *p1, unsigned long *p2,
 }
 
 static void
-xor_8regs_p_4(unsigned long bytes, unsigned long *p1, unsigned long *p2,
-	    unsigned long *p3, unsigned long *p4)
+xor_8regs_p_4(unsigned long bytes, unsigned long * __restrict p1,
+	      const unsigned long * __restrict p2,
+	      const unsigned long * __restrict p3,
+	      const unsigned long * __restrict p4)
 {
 	long lines = bytes / (sizeof (long)) / 8 - 1;
 
@@ -384,8 +402,11 @@ xor_8regs_p_4(unsigned long bytes, unsigned long *p1, unsigned long *p2,
 }
 
 static void
-xor_8regs_p_5(unsigned long bytes, unsigned long *p1, unsigned long *p2,
-	    unsigned long *p3, unsigned long *p4, unsigned long *p5)
+xor_8regs_p_5(unsigned long bytes, unsigned long * __restrict p1,
+	      const unsigned long * __restrict p2,
+	      const unsigned long * __restrict p3,
+	      const unsigned long * __restrict p4,
+	      const unsigned long * __restrict p5)
 {
 	long lines = bytes / (sizeof (long)) / 8 - 1;
 
@@ -421,7 +442,8 @@ xor_8regs_p_5(unsigned long bytes, unsigned long *p1, unsigned long *p2,
 }
 
 static void
-xor_32regs_p_2(unsigned long bytes, unsigned long *p1, unsigned long *p2)
+xor_32regs_p_2(unsigned long bytes, unsigned long * __restrict p1,
+	       const unsigned long * __restrict p2)
 {
 	long lines = bytes / (sizeof (long)) / 8 - 1;
 
@@ -466,8 +488,9 @@ xor_32regs_p_2(unsigned long bytes, unsigned long *p1, unsigned long *p2)
 }
 
 static void
-xor_32regs_p_3(unsigned long bytes, unsigned long *p1, unsigned long *p2,
-	    unsigned long *p3)
+xor_32regs_p_3(unsigned long bytes, unsigned long * __restrict p1,
+	       const unsigned long * __restrict p2,
+	       const unsigned long * __restrict p3)
 {
 	long lines = bytes / (sizeof (long)) / 8 - 1;
 
@@ -523,8 +546,10 @@ xor_32regs_p_3(unsigned long bytes, unsigned long *p1, unsigned long *p2,
 }
 
 static void
-xor_32regs_p_4(unsigned long bytes, unsigned long *p1, unsigned long *p2,
-	    unsigned long *p3, unsigned long *p4)
+xor_32regs_p_4(unsigned long bytes, unsigned long * __restrict p1,
+	       const unsigned long * __restrict p2,
+	       const unsigned long * __restrict p3,
+	       const unsigned long * __restrict p4)
 {
 	long lines = bytes / (sizeof (long)) / 8 - 1;
 
@@ -591,8 +616,11 @@ xor_32regs_p_4(unsigned long bytes, unsigned long *p1, unsigned long *p2,
 }
 
 static void
-xor_32regs_p_5(unsigned long bytes, unsigned long *p1, unsigned long *p2,
-	    unsigned long *p3, unsigned long *p4, unsigned long *p5)
+xor_32regs_p_5(unsigned long bytes, unsigned long * __restrict p1,
+	       const unsigned long * __restrict p2,
+	       const unsigned long * __restrict p3,
+	       const unsigned long * __restrict p4,
+	       const unsigned long * __restrict p5)
 {
 	long lines = bytes / (sizeof (long)) / 8 - 1;
 
diff --git a/include/linux/raid/xor.h b/include/linux/raid/xor.h
index 2a9fee8ddae3..4ddec31d18f0 100644
--- a/include/linux/raid/xor.h
+++ b/include/linux/raid/xor.h
@@ -11,13 +11,20 @@ struct xor_block_template {
         struct xor_block_template *next;
         const char *name;
         int speed;
-	void (*do_2)(unsigned long, unsigned long *, unsigned long *);
-	void (*do_3)(unsigned long, unsigned long *, unsigned long *,
-		     unsigned long *);
-	void (*do_4)(unsigned long, unsigned long *, unsigned long *,
-		     unsigned long *, unsigned long *);
-	void (*do_5)(unsigned long, unsigned long *, unsigned long *,
-		     unsigned long *, unsigned long *, unsigned long *);
+	void (*do_2)(unsigned long, unsigned long * __restrict,
+		     const unsigned long * __restrict);
+	void (*do_3)(unsigned long, unsigned long * __restrict,
+		     const unsigned long * __restrict,
+		     const unsigned long * __restrict);
+	void (*do_4)(unsigned long, unsigned long * __restrict,
+		     const unsigned long * __restrict,
+		     const unsigned long * __restrict,
+		     const unsigned long * __restrict);
+	void (*do_5)(unsigned long, unsigned long * __restrict,
+		     const unsigned long * __restrict,
+		     const unsigned long * __restrict,
+		     const unsigned long * __restrict,
+		     const unsigned long * __restrict);
 };
 
 #endif
-- 
2.30.2

Re: [PATCH 1/2] lib/xor: make xor prototypes more friendely to compiler vectorization

[kernel test robot]

[-- Attachment #1: Type: text/plain, Size: 5093 bytes --]

Hi Ard,

I love your patch! Yet something to improve:

[auto build test ERROR on e783362eb54cd99b2cac8b3a9aeac942e6f6ac07]

url:    https://github.com/0day-ci/linux/commits/Ard-Biesheuvel/xor-enable-auto-vectorization-in-Clang/20220127-161449
base:   e783362eb54cd99b2cac8b3a9aeac942e6f6ac07
config: arm64-randconfig-r015-20220127 (https://download.01.org/0day-ci/archive/20220128/202201280125.Q4T02G9Q-lkp(a)intel.com/config)
compiler: aarch64-linux-gcc (GCC) 11.2.0
reproduce (this is a W=1 build):
        wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross
        chmod +x ~/bin/make.cross
        # https://github.com/0day-ci/linux/commit/543912d52a64185845fd9bde2412680ef0156a7c
        git remote add linux-review https://github.com/0day-ci/linux
        git fetch --no-tags linux-review Ard-Biesheuvel/xor-enable-auto-vectorization-in-Clang/20220127-161449
        git checkout 543912d52a64185845fd9bde2412680ef0156a7c
        # save the config file to linux build tree
        mkdir build_dir
        COMPILER_INSTALL_PATH=$HOME/0day COMPILER=gcc-11.2.0 make.cross O=build_dir ARCH=arm64 SHELL=/bin/bash

If you fix the issue, kindly add following tag as appropriate
Reported-by: kernel test robot <lkp@intel.com>

All errors (new ones prefixed by >>):

   arch/arm64/lib/xor-neon.c:13:6: warning: no previous prototype for 'xor_arm64_neon_2' [-Wmissing-prototypes]
      13 | void xor_arm64_neon_2(unsigned long bytes, unsigned long * __restrict p1,
         |      ^~~~~~~~~~~~~~~~
   arch/arm64/lib/xor-neon.c:40:6: warning: no previous prototype for 'xor_arm64_neon_3' [-Wmissing-prototypes]
      40 | void xor_arm64_neon_3(unsigned long bytes, unsigned long * __restrict p1,
         |      ^~~~~~~~~~~~~~~~
   arch/arm64/lib/xor-neon.c:76:6: warning: no previous prototype for 'xor_arm64_neon_4' [-Wmissing-prototypes]
      76 | void xor_arm64_neon_4(unsigned long bytes, unsigned long * __restrict p1,
         |      ^~~~~~~~~~~~~~~~
   arch/arm64/lib/xor-neon.c:121:6: warning: no previous prototype for 'xor_arm64_neon_5' [-Wmissing-prototypes]
     121 | void xor_arm64_neon_5(unsigned long bytes, unsigned long * __restrict p1,
         |      ^~~~~~~~~~~~~~~~
   arch/arm64/lib/xor-neon.c: In function 'xor_neon_init':
>> arch/arm64/lib/xor-neon.c:316:43: error: assignment to 'void (*)(long unsigned int,  long unsigned int * __restrict__,  const long unsigned int * __restrict__,  const long unsigned int * __restrict__)' from incompatible pointer type 'void (*)(long unsigned int,  long unsigned int *, long unsigned int *, long unsigned int *)' [-Werror=incompatible-pointer-types]
     316 |                 xor_block_inner_neon.do_3 = xor_arm64_eor3_3;
         |                                           ^
>> arch/arm64/lib/xor-neon.c:317:43: error: assignment to 'void (*)(long unsigned int,  long unsigned int * __restrict__,  const long unsigned int * __restrict__,  const long unsigned int * __restrict__,  const long unsigned int * __restrict__)' from incompatible pointer type 'void (*)(long unsigned int,  long unsigned int *, long unsigned int *, long unsigned int *, long unsigned int *)' [-Werror=incompatible-pointer-types]
     317 |                 xor_block_inner_neon.do_4 = xor_arm64_eor3_4;
         |                                           ^
   arch/arm64/lib/xor-neon.c:318:43: error: assignment to 'void (*)(long unsigned int,  long unsigned int * __restrict__,  const long unsigned int * __restrict__,  const long unsigned int * __restrict__,  const long unsigned int * __restrict__,  const long unsigned int * __restrict__)' from incompatible pointer type 'void (*)(long unsigned int,  long unsigned int *, long unsigned int *, long unsigned int *, long unsigned int *, long unsigned int *)' [-Werror=incompatible-pointer-types]
     318 |                 xor_block_inner_neon.do_5 = xor_arm64_eor3_5;
         |                                           ^
   cc1: some warnings being treated as errors


vim +316 arch/arm64/lib/xor-neon.c

2c54b423cf85ba Ard Biesheuvel 2021-12-13  312  
2c54b423cf85ba Ard Biesheuvel 2021-12-13  313  static int __init xor_neon_init(void)
2c54b423cf85ba Ard Biesheuvel 2021-12-13  314  {
2c54b423cf85ba Ard Biesheuvel 2021-12-13  315  	if (IS_ENABLED(CONFIG_AS_HAS_SHA3) && cpu_have_named_feature(SHA3)) {
2c54b423cf85ba Ard Biesheuvel 2021-12-13 @316  		xor_block_inner_neon.do_3 = xor_arm64_eor3_3;
2c54b423cf85ba Ard Biesheuvel 2021-12-13 @317  		xor_block_inner_neon.do_4 = xor_arm64_eor3_4;
2c54b423cf85ba Ard Biesheuvel 2021-12-13  318  		xor_block_inner_neon.do_5 = xor_arm64_eor3_5;
2c54b423cf85ba Ard Biesheuvel 2021-12-13  319  	}
2c54b423cf85ba Ard Biesheuvel 2021-12-13  320  	return 0;
2c54b423cf85ba Ard Biesheuvel 2021-12-13  321  }
2c54b423cf85ba Ard Biesheuvel 2021-12-13  322  module_init(xor_neon_init);
2c54b423cf85ba Ard Biesheuvel 2021-12-13  323  

---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/kbuild-all(a)lists.01.org

Re: [PATCH 1/2] lib/xor: make xor prototypes more friendely to compiler vectorization

[Herbert Xu]

On Thu, Jan 27, 2022 at 09:12:26AM +0100, Ard Biesheuvel wrote:
>
> diff --git a/arch/arm64/lib/xor-neon.c b/arch/arm64/lib/xor-neon.c
> index d189cf4e70ea..64c9577fcf0f 100644
> --- a/arch/arm64/lib/xor-neon.c
> +++ b/arch/arm64/lib/xor-neon.c

This fails to build:

../arch/arm64/lib/xor-neon.c:13:6: warning: no previous prototype for ‘xor_arm64_neon_2’ [-Wmissing-prototypes]
   13 | void xor_arm64_neon_2(unsigned long bytes, unsigned long * __restrict p1,
      |      ^~~~~~~~~~~~~~~~
../arch/arm64/lib/xor-neon.c:40:6: warning: no previous prototype for ‘xor_arm64_neon_3’ [-Wmissing-prototypes]
   40 | void xor_arm64_neon_3(unsigned long bytes, unsigned long * __restrict p1,
      |      ^~~~~~~~~~~~~~~~
../arch/arm64/lib/xor-neon.c:76:6: warning: no previous prototype for ‘xor_arm64_neon_4’ [-Wmissing-prototypes]
   76 | void xor_arm64_neon_4(unsigned long bytes, unsigned long * __restrict p1,
      |      ^~~~~~~~~~~~~~~~
../arch/arm64/lib/xor-neon.c:121:6: warning: no previous prototype for ‘xor_arm64_neon_5’ [-Wmissing-prototypes]
  121 | void xor_arm64_neon_5(unsigned long bytes, unsigned long * __restrict p1,
      |      ^~~~~~~~~~~~~~~~
../arch/arm64/lib/xor-neon.c: In function ‘xor_neon_init’:
../arch/arm64/lib/xor-neon.c:316:29: error: assignment to ‘void (*)(long unsigned int,  long unsigned int * __restrict__,  const long unsigned int * __restrict__,  const long unsigned int * __restrict__)’ from incompatible pointer type ‘void (*)(long unsigned int,  long unsigned int *, long unsigned int *, long unsigned int *)’ [-Werror=incompatible-pointer-types]
  316 |   xor_block_inner_neon.do_3 = xor_arm64_eor3_3;
      |                             ^
../arch/arm64/lib/xor-neon.c:317:29: error: assignment to ‘void (*)(long unsigned int,  long unsigned int * __restrict__,  const long unsigned int * __restrict__,  const long unsigned int * __restrict__,  const long unsigned int * __restrict__)’ from incompatible pointer type ‘void (*)(long unsigned int,  long unsigned int *, long unsigned int *, long unsigned int *, long unsigned int *)’ [-Werror=incompatible-pointer-types]
  317 |   xor_block_inner_neon.do_4 = xor_arm64_eor3_4;
      |                             ^
../arch/arm64/lib/xor-neon.c:318:29: error: assignment to ‘void (*)(long unsigned int,  long unsigned int * __restrict__,  const long unsigned int * __restrict__,  const long unsigned int * __restrict__,  const long unsigned int * __restrict__,  const long unsigned int * __restrict__)’ from incompatible pointer type ‘void (*)(long unsigned int,  long unsigned int *, long unsigned int *, long unsigned int *, long unsigned int *, long unsigned int *)’ [-Werror=incompatible-pointer-types]
  318 |   xor_block_inner_neon.do_5 = xor_arm64_eor3_5;
      |                             ^
cc1: some warnings being treated as errors

Cheers,
-- 
Email: Herbert Xu <herbert@gondor.apana.org.au>
Home Page: http://gondor.apana.org.au/~herbert/
PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt

[PATCH 2/2] crypto: arm/xor - make vectorized C code Clang-friendly

[Ard Biesheuvel]

The ARM version of the accelerated XOR routines are simply the 8-way C
routines passed through the auto-vectorizer with SIMD codegen enabled.
This used to require GCC version 4.6 at least, but given that 5.1 is now
the baseline, this check is no longer necessary, and actually
misidentifies Clang as GCC < 4.6 as Clang defines the GCC major/minor as
well, but makes no attempt at doing this in a way that conveys feature
parity with a certain version of GCC (which would not be a great idea in
the first place).

So let's drop the version check, and make the auto-vectorize pragma
(which is based on a GCC-specific command line option) GCC-only. Since
Clang performs SIMD auto-vectorization by default at -O2, no pragma is
necessary here.

Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
---
 arch/arm/lib/xor-neon.c | 12 +++---------
 1 file changed, 3 insertions(+), 9 deletions(-)

diff --git a/arch/arm/lib/xor-neon.c b/arch/arm/lib/xor-neon.c
index b99dd8e1c93f..522510baed49 100644
--- a/arch/arm/lib/xor-neon.c
+++ b/arch/arm/lib/xor-neon.c
@@ -17,17 +17,11 @@ MODULE_LICENSE("GPL");
 /*
  * Pull in the reference implementations while instructing GCC (through
  * -ftree-vectorize) to attempt to exploit implicit parallelism and emit
- * NEON instructions.
+ * NEON instructions. Clang does this by default at O2 so no pragma is
+ * needed.
  */
-#if __GNUC__ > 4 || (__GNUC__ == 4 && __GNUC_MINOR__ >= 6)
+#ifdef CONFIG_CC_IS_GCC
 #pragma GCC optimize "tree-vectorize"
-#else
-/*
- * While older versions of GCC do not generate incorrect code, they fail to
- * recognize the parallel nature of these functions, and emit plain ARM code,
- * which is known to be slower than the optimized ARM code in asm-arm/xor.h.
- */
-#warning This code requires at least version 4.6 of GCC
 #endif
 
 #pragma GCC diagnostic ignored "-Wunused-variable"
-- 
2.30.2

Re: [PATCH 0/2] xor: enable auto-vectorization in Clang

[Nathan Chancellor]

Hi Ard,

On Thu, Jan 27, 2022 at 09:12:25AM +0100, Ard Biesheuvel wrote:
> Update the xor_blocks() prototypes so that the compiler understands that
> the inputs always refer to distinct regions of memory. This is implied
> by the existing implementations, as they use different granularities for
> the load/xor/store loops.
> 
> With that, we can fix the ARM/Clang version, which refuses to SIMD
> vectorize otherwise, and throws a spurious warning related to the GCC
> version being incompatible.
> 
> Cc: Nick Desaulniers <ndesaulniers@google.com>
> Cc: Nathan Chancellor <nathan@kernel.org>
> 
> Ard Biesheuvel (2):
>   lib/xor: make xor prototypes more friendely to compiler vectorization
>   crypto: arm/xor - make vectorized C code Clang-friendly

I tested multi_v7_defconfig + CONFIG_BTRFS=y (to get
CONFIG_XOR_BLOCKS=y) in QEMU 6.2.0 (10 boots) and the xor neon code gets
faster according to do_xor_speed():

mainline @ 626b2dda7651:

[    2.591449]    neon            :  1166 MB/sec
[    2.579454]    neon            :  1118 MB/sec
[    2.589061]    neon            :  1163 MB/sec
[    2.581827]    neon            :  1167 MB/sec
[    2.599079]    neon            :  1166 MB/sec
[    2.579252]    neon            :  1147 MB/sec
[    2.582637]    neon            :  1168 MB/sec
[    2.582872]    neon            :  1164 MB/sec
[    2.570671]    neon            :  1167 MB/sec
[    2.571830]    neon            :  1166 MB/sec

mainline @ 626b2dda7651 with series:

[    2.570227]    neon            :  1238 MB/sec
[    2.571642]    neon            :  1237 MB/sec
[    2.580370]    neon            :  1234 MB/sec
[    2.581966]    neon            :  1238 MB/sec
[    2.582313]    neon            :  1236 MB/sec
[    2.572291]    neon            :  1238 MB/sec
[    2.570625]    neon            :  1233 MB/sec
[    2.571897]    neon            :  1234 MB/sec
[    2.589616]    neon            :  1228 MB/sec
[    2.582449]    neon            :  1236 MB/sec

This series is currently broken for powerpc [1], as the functions in
arch/powerpc/lib/xor_vmx.c were not updated.

arch/powerpc/lib/xor_vmx.c:52:6: error: conflicting types for '__xor_altivec_2'
void __xor_altivec_2(unsigned long bytes, unsigned long *v1_in,
     ^
arch/powerpc/lib/xor_vmx.h:9:6: note: previous declaration is here
void __xor_altivec_2(unsigned long bytes, unsigned long * __restrict p1,
     ^
arch/powerpc/lib/xor_vmx.c:70:6: error: conflicting types for '__xor_altivec_3'
void __xor_altivec_3(unsigned long bytes, unsigned long *v1_in,
     ^
arch/powerpc/lib/xor_vmx.h:11:6: note: previous declaration is here
void __xor_altivec_3(unsigned long bytes, unsigned long * __restrict p1,
     ^
arch/powerpc/lib/xor_vmx.c:92:6: error: conflicting types for '__xor_altivec_4'
void __xor_altivec_4(unsigned long bytes, unsigned long *v1_in,
     ^
arch/powerpc/lib/xor_vmx.h:14:6: note: previous declaration is here
void __xor_altivec_4(unsigned long bytes, unsigned long * __restrict p1,
     ^
arch/powerpc/lib/xor_vmx.c:119:6: error: conflicting types for '__xor_altivec_5'
void __xor_altivec_5(unsigned long bytes, unsigned long *v1_in,
     ^
arch/powerpc/lib/xor_vmx.h:18:6: note: previous declaration is here
void __xor_altivec_5(unsigned long bytes, unsigned long * __restrict p1,
     ^
4 errors generated.

If I fix that up [2], it builds and resolves an instance of
-Wframe-larger-than= in the xor altivec code, as seen with
pmac32_defconfig.

Before this series:

arch/powerpc/lib/xor_vmx.c:119:6: error: stack frame size (1232) exceeds limit (1024) in '__xor_altivec_5' [-Werror,-Wframe-larger-than]
void __xor_altivec_5(unsigned long bytes, unsigned long *v1_in,
     ^
1 error generated.

After this patch (with CONFIG_FRAME_WARN=100 and
CONFIG_PPC_DISABLE_WERROR=y):

arch/powerpc/lib/xor_vmx.c:52:6: warning: stack frame size (128) exceeds limit (100) in '__xor_altivec_2' [-Wframe-larger-than]
void __xor_altivec_2(unsigned long bytes,
     ^
arch/powerpc/lib/xor_vmx.c:71:6: warning: stack frame size (160) exceeds limit (100) in '__xor_altivec_3' [-Wframe-larger-than]
void __xor_altivec_3(unsigned long bytes,
     ^
arch/powerpc/lib/xor_vmx.c:95:6: warning: stack frame size (144) exceeds limit (100) in '__xor_altivec_4' [-Wframe-larger-than]
void __xor_altivec_4(unsigned long bytes,
     ^
arch/powerpc/lib/xor_vmx.c:124:6: warning: stack frame size (160) exceeds limit (100) in '__xor_altivec_5' [-Wframe-larger-than]
void __xor_altivec_5(unsigned long bytes,
     ^
4 warnings generated.

There is a similar performance gain as ARM according to do_xor_speed():

Before:

   altivec         :   222 MB/sec
   altivec         :   222 MB/sec
   altivec         :   222 MB/sec
   altivec         :   219 MB/sec
   altivec         :   222 MB/sec
   altivec         :   222 MB/sec
   altivec         :   222 MB/sec
   altivec         :   222 MB/sec
   altivec         :   222 MB/sec
   altivec         :   222 MB/sec

After:

   altivec         :   278 MB/sec
   altivec         :   276 MB/sec
   altivec         :   278 MB/sec
   altivec         :   278 MB/sec
   altivec         :   278 MB/sec
   altivec         :   278 MB/sec
   altivec         :   278 MB/sec
   altivec         :   278 MB/sec
   altivec         :   278 MB/sec
   altivec         :   278 MB/sec

I did also build test arm64 and x86_64 and saw no errors. I did runtime
test arm64 for improvements and did not see any, which is good, since I
take that as meaning it was working fine before and there is no
regression.

Once the build error is fixed, consider this series:

Tested-by: Nathan Chancellor <nathan@kernel.org>

[1]: https://lore.kernel.org/r/202112310646.kuh2pXiG-lkp@intel.com/
[2]: https://github.com/ClangBuiltLinux/linux/issues/563#issuecomment-1005175153

Cheers,
Nathan