src/crypto/asm/cn_main_loop.S

#ifdef __APPLE__
#   define ALIGN(x) .align 6
#else
#   define ALIGN(x) .align 64
#endif
.intel_syntax noprefix
#ifdef __APPLE__
#   define FN_PREFIX(fn) _ ## fn
.text
#else
#   define FN_PREFIX(fn) fn
.section .text
#endif
.global FN_PREFIX(cnv2_mainloop_ivybridge_asm)
.global FN_PREFIX(cnv2_mainloop_ryzen_asm)
.global FN_PREFIX(cnv2_mainloop_bulldozer_asm)
.global FN_PREFIX(cnv2_double_mainloop_sandybridge_asm)
.global FN_PREFIX(cnv2_rwz_mainloop_asm)
.global FN_PREFIX(cnv2_rwz_double_mainloop_asm)

ALIGN(64)
FN_PREFIX(cnv2_mainloop_ivybridge_asm):
	sub rsp, 48
	mov rcx, rdi
	#include "cn2/cnv2_main_loop_ivybridge.inc"
	add rsp, 48
	ret 0
	mov eax, 3735929054

ALIGN(64)
FN_PREFIX(cnv2_mainloop_ryzen_asm):
	sub rsp, 48
	mov rcx, rdi
	#include "cn2/cnv2_main_loop_ryzen.inc"
	add rsp, 48
	ret 0
	mov eax, 3735929054

ALIGN(64)
FN_PREFIX(cnv2_mainloop_bulldozer_asm):
	sub rsp, 48
	mov rcx, rdi
	#include "cn2/cnv2_main_loop_bulldozer.inc"
	add rsp, 48
	ret 0
	mov eax, 3735929054

ALIGN(64)
FN_PREFIX(cnv2_double_mainloop_sandybridge_asm):
	sub rsp, 48
	mov rcx, rdi
	mov rdx, rsi
	#include "cn2/cnv2_double_main_loop_sandybridge.inc"
	add rsp, 48
	ret 0
	mov eax, 3735929054

ALIGN(64)
FN_PREFIX(cnv2_rwz_mainloop_asm):
	sub rsp, 48
	mov rcx, rdi
	#include "cn2/cnv2_rwz_main_loop.inc"
	add rsp, 48
	ret 0
	mov eax, 3735929054

ALIGN(64)
FN_PREFIX(cnv2_rwz_double_mainloop_asm):
	sub rsp, 48
	mov rcx, rdi
	mov rdx, rsi
	#include "cn2/cnv2_rwz_double_main_loop.inc"
	add rsp, 48
	ret 0
	mov eax, 3735929054
Fixed wrong ASM code alignment on macOS, thanks @SChernykh. 2019-01-15 18:18:04 +07:00			`#ifdef __APPLE__`
			`# define ALIGN(x) .align 6`
			`#else`
			`# define ALIGN(x) .align 64`
			`#endif`
Initial compile with ASM. 2018-09-23 20:16:33 +03:00			`.intel_syntax noprefix`
Fix macOS/Clang build. 2018-09-23 20:45:10 +03:00			`#ifdef __APPLE__`
			`# define FN_PREFIX(fn) _ ## fn`
			`.text`
			`#else`
			`# define FN_PREFIX(fn) fn`
Initial compile with ASM. 2018-09-23 20:16:33 +03:00			`.section .text`
Fix macOS/Clang build. 2018-09-23 20:45:10 +03:00			`#endif`
			`.global FN_PREFIX(cnv2_mainloop_ivybridge_asm)`
			`.global FN_PREFIX(cnv2_mainloop_ryzen_asm)`
Added asm optimized code for AMD Bulldozer 2018-10-21 18:29:03 +02:00			`.global FN_PREFIX(cnv2_mainloop_bulldozer_asm)`
Added ASM code for double hash mode, thanks @SChernykh. 2018-09-25 09:25:47 +03:00			`.global FN_PREFIX(cnv2_double_mainloop_sandybridge_asm)`
CryptoNight v8 ReverseWaltz (+asm code and all test vectors) Co-Authored-By: Vitalius Parubochyi <eddragonwolf@users.noreply.github.com> 2019-03-04 19:57:33 +01:00			`.global FN_PREFIX(cnv2_rwz_mainloop_asm)`
			`.global FN_PREFIX(cnv2_rwz_double_mainloop_asm)`
Initial compile with ASM. 2018-09-23 20:16:33 +03:00
Fixed wrong ASM code alignment on macOS, thanks @SChernykh. 2019-01-15 18:18:04 +07:00			`ALIGN(64)`
Fix macOS/Clang build. 2018-09-23 20:45:10 +03:00			`FN_PREFIX(cnv2_mainloop_ivybridge_asm):`
Initial compile with ASM. 2018-09-23 20:16:33 +03:00			`sub rsp, 48`
			`mov rcx, rdi`
Rename ASM files. 2019-01-13 23:00:41 +07:00			`#include "cn2/cnv2_main_loop_ivybridge.inc"`
Initial compile with ASM. 2018-09-23 20:16:33 +03:00			`add rsp, 48`
			`ret 0`
Fixed compilation error on Windows 2019-01-16 00:24:34 +01:00			`mov eax, 3735929054`
Initial compile with ASM. 2018-09-23 20:16:33 +03:00
Fixed wrong ASM code alignment on macOS, thanks @SChernykh. 2019-01-15 18:18:04 +07:00			`ALIGN(64)`
Fix macOS/Clang build. 2018-09-23 20:45:10 +03:00			`FN_PREFIX(cnv2_mainloop_ryzen_asm):`
Initial compile with ASM. 2018-09-23 20:16:33 +03:00			`sub rsp, 48`
			`mov rcx, rdi`
Rename ASM files. 2019-01-13 23:00:41 +07:00			`#include "cn2/cnv2_main_loop_ryzen.inc"`
Initial compile with ASM. 2018-09-23 20:16:33 +03:00			`add rsp, 48`
			`ret 0`
Fixed compilation error on Windows 2019-01-16 00:24:34 +01:00			`mov eax, 3735929054`
Added ASM code for double hash mode, thanks @SChernykh. 2018-09-25 09:25:47 +03:00
Fixed wrong ASM code alignment on macOS, thanks @SChernykh. 2019-01-15 18:18:04 +07:00			`ALIGN(64)`
Added asm optimized code for AMD Bulldozer 2018-10-21 18:29:03 +02:00			`FN_PREFIX(cnv2_mainloop_bulldozer_asm):`
			`sub rsp, 48`
			`mov rcx, rdi`
Rename ASM files. 2019-01-13 23:00:41 +07:00			`#include "cn2/cnv2_main_loop_bulldozer.inc"`
Added asm optimized code for AMD Bulldozer 2018-10-21 18:29:03 +02:00			`add rsp, 48`
			`ret 0`
Fixed compilation error on Windows 2019-01-16 00:24:34 +01:00			`mov eax, 3735929054`
Added asm optimized code for AMD Bulldozer 2018-10-21 18:29:03 +02:00
Fixed wrong ASM code alignment on macOS, thanks @SChernykh. 2019-01-15 18:18:04 +07:00			`ALIGN(64)`
Added ASM code for double hash mode, thanks @SChernykh. 2018-09-25 09:25:47 +03:00			`FN_PREFIX(cnv2_double_mainloop_sandybridge_asm):`
			`sub rsp, 48`
			`mov rcx, rdi`
			`mov rdx, rsi`
Rename ASM files. 2019-01-13 23:00:41 +07:00			`#include "cn2/cnv2_double_main_loop_sandybridge.inc"`
Added ASM code for double hash mode, thanks @SChernykh. 2018-09-25 09:25:47 +03:00			`add rsp, 48`
			`ret 0`
Fixed compilation error on Windows 2019-01-16 00:24:34 +01:00			`mov eax, 3735929054`
CryptoNight v8 ReverseWaltz (+asm code and all test vectors) Co-Authored-By: Vitalius Parubochyi <eddragonwolf@users.noreply.github.com> 2019-03-04 19:57:33 +01:00
			`ALIGN(64)`
			`FN_PREFIX(cnv2_rwz_mainloop_asm):`
			`sub rsp, 48`
			`mov rcx, rdi`
			`#include "cn2/cnv2_rwz_main_loop.inc"`
			`add rsp, 48`
			`ret 0`
			`mov eax, 3735929054`

			`ALIGN(64)`
			`FN_PREFIX(cnv2_rwz_double_mainloop_asm):`
			`sub rsp, 48`
			`mov rcx, rdi`
#975 Fixed crash on Linux if used double thread. 2019-03-07 14:52:40 +07:00			`mov rdx, rsi`
CryptoNight v8 ReverseWaltz (+asm code and all test vectors) Co-Authored-By: Vitalius Parubochyi <eddragonwolf@users.noreply.github.com> 2019-03-04 19:57:33 +01:00			`#include "cn2/cnv2_rwz_double_main_loop.inc"`
			`add rsp, 48`
			`ret 0`
			`mov eax, 3735929054`