Fix SHA512 on ARM benchmarks

This was a mistake when porting from Cryptogams to Crypto++. The macros VFP_ABI_PUSH and VFP_ABI_POP needed to be defined because they save and restore SIMD register state. They were originally missing during the port. The benchmarks would hang because the doubles we used for benchmarking were blown away in sha512_block_data_order_neon.

sha.cpp

@@ -998,7 +998,7 @@ std::string SHA512_AlgorithmProvider()
     if (HasSSE2())
         return "SSE2";
 #endif
-#if CRYPTOGAMS_ARM_SHA512 && 0
+#if CRYPTOGAMS_ARM_SHA512
     if (HasNEON())
         return "NEON";
     if (HasARMv7())
@@ -1310,7 +1310,7 @@ void SHA512::Transform(word64 *state, const word64 *data)
         return;
     }
 #endif
-#if CRYPTOGAMS_ARM_SHA512 && 0
+#if CRYPTOGAMS_ARM_SHA512
     if (HasARMv7())
     {
 # if (CRYPTOPP_LITTLE_ENDIAN)

sha1_armv4.S

@@ -85,21 +85,21 @@
 
 .align	5
 sha1_block_data_order:
+.Lsha1_block_data_order:
+#if __ARM_ARCH__<7 && !defined(__thumb2__)
+	sub	r3,pc,#8		@ sha1_block_data_order
+#else
+	adr	r3,.Lsha1_block_data_order
+#endif
+
 #if __ARM_MAX_ARCH__>=7
-.Lsha1_block:
-	@ldr	r12,.LCRYPTOGAMS_armcap
 	ldr	r12,=CRYPTOGAMS_armcap_P
-# if !defined(_WIN32)
+	ldr	r12,[r12]		@ CRYPTOGAMS_armcap_P
-	adr	r3,.Lsha1_block
+
-	@ldr	r12,[r3,r12]		@ CRYPTOGAMS_armcap_P
-	ldr	r12,[r12]
-# endif
-# if defined(__APPLE__) || defined(_WIN32)
-	ldr	r12,[r12]
-# endif
 	tst	r12,#ARMV7_NEON
 	bne	.LNEON
 #endif
+
 	stmdb	sp!,{r4,r5,r6,r7,r8,r9,r10,r11,r12,lr}
 	add	r2,r1,r2,lsl#6	@ r2 to point at the end of r1
 	ldmia	r0,{r3,r4,r5,r6,r7}

sha256_armv4.S

@@ -113,19 +113,15 @@ sha256_block_data_order:
 #else
 	adr	r3,.Lsha256_block_data_order
 #endif
+
 #if __ARM_MAX_ARCH__>=7
-	@ldr	r12,.LCRYPTOGAMS_armcap
 	ldr	r12,=CRYPTOGAMS_armcap_P
-# if !defined(_WIN32)
-	@ldr	r12,[r3,r12]		@ CRYPTOGAMS_armcap_P
 	ldr	r12,[r12]		@ CRYPTOGAMS_armcap_P
-# endif
+
-# if defined(__APPLE__) || defined(_WIN32)
-	ldr	r12,[r12]
-# endif
 	tst	r12,#ARMV7_NEON
 	bne	.LNEON
 #endif
+
 	add	r2,r1,r2,lsl#6	@ len to point at the end of inp
 	stmdb	sp!,{r0,r1,r2,r4-r11,lr}
 	ldmia	r0,{r4,r5,r6,r7,r8,r9,r10,r11}

sha512_armv4.S

@@ -79,6 +79,9 @@
 # define WORD64(hi0,lo0,hi1,lo1)	.word	hi0,lo0, hi1,lo1
 #endif
 
+#define VFP_ABI_PUSH   vstmdb  sp!,{d8-d15}
+#define VFP_ABI_POP    vldmia  sp!,{d8-d15}
+
 #if defined(__thumb2__)
 .syntax	unified
 .thumb
@@ -147,19 +150,15 @@ sha512_block_data_order:
 #else
 	adr	r3,.Lsha512_block_data_order
 #endif
+
 #if __ARM_MAX_ARCH__>=7
-	@ldr	r12,.LCRYPTOGAMS_armcap
 	ldr	r12,=CRYPTOGAMS_armcap_P
-# if !defined(_WIN32)
-	@ldr	r12,[r3,r12]		@ CRYPTOGAMS_armcap_P
 	ldr	r12,[r12]		@ CRYPTOGAMS_armcap_P
-# endif
+
-# if defined(__APPLE__) || defined(_WIN32)
-	ldr	r12,[r12]
-# endif
 	tst	r12,#ARMV7_NEON
 	bne	.LNEON
 #endif
+
 	add	r2,r1,r2,lsl#7	@ len to point at the end of inp
 	stmdb	sp!,{r4,r5,r6,r7,r8,r9,r10,r11,r12,lr}
 	sub	r14,r3,#672		@ K512
@@ -554,7 +553,7 @@ sha512_block_data_order_neon:
 	dmb	@ errata #451034 on early Cortex A8
 	add	r2,r1,r2,lsl#7	@ len to point at the end of inp
 	adr	r3,K512
-	@VFP_ABI_PUSH
+	VFP_ABI_PUSH
 	vldmia	r0,{d16,d17,d18,d19,d20,d21,d22,d23}		@ load context
 .Loop_neon:
 	vshr.u64	d24,d20,#14	@ 0
@@ -1868,7 +1867,7 @@ sha512_block_data_order_neon:
 	teq	r1,r2
 	sub	r3,#640	@ rewind K512
 	bne	.Loop_neon
-	@VFP_ABI_POP
+	VFP_ABI_POP
 	bx	lr				@ .word	0xe12fff1e
 .size	sha512_block_data_order_neon,.-sha512_block_data_order_neon
 #endif