RandomX optimizations

- Optimized soft AES code, up to +30% hashrate on CPU without AES support
- Added prefetch for the first dataset access, up to +0.1% hashrate

src/crypto/randomx/asm/program_prologue_linux.inc

@@ -12,6 +12,13 @@
 	mov rcx, rdi
 	mov rbp, qword ptr [rsi]    ;# "mx", "ma"
 	mov rdi, qword ptr [rsi+8]  ;# uint8_t* dataset
+
+	;# dataset prefetch for the first iteration of the main loop
+	mov rax, rbp
+	shr rax, 32
+	and eax, RANDOMX_DATASET_BASE_MASK
+	prefetchnta byte ptr [rdi+rax]
+
 	mov rsi, rdx                ;# uint8_t* scratchpad
 
 	mov rax, rbp

src/crypto/randomx/asm/program_prologue_win64.inc

@@ -24,6 +24,13 @@
 	push rcx                    ;# RegisterFile& registerFile
 	mov rbp, qword ptr [rdx]    ;# "mx", "ma"
 	mov rdi, qword ptr [rdx+8]  ;# uint8_t* dataset
+
+	;# dataset prefetch for the first iteration of the main loop
+	mov rax, rbp
+	shr rax, 32
+	and eax, RANDOMX_DATASET_BASE_MASK
+	prefetchnta byte ptr [rdi+rax]
+
 	mov rsi, r8                 ;# uint8_t* scratchpad
 	mov rbx, r9                 ;# loop counter