Implementation notes: ppc32, gcc1power7, crypto_hash/blake2s

Computer: gcc1power7
Architecture: ppc32
CPU ID: unknown CPU ID
SUPERCOP version: 20130126
Operation: crypto_hash
Primitive: blake2s

Time	Implementation	Compiler	Benchmark date	SUPERCOP version
26684	`regs`	`gcc -m32 -mcpu=G5 -maltivec -O2 -fomit-frame-pointer`	20130127	20130126
26859	`regs`	`gcc -m32 -mcpu=G5 -O2 -fomit-frame-pointer`	20130127	20130126
27209	`regs`	`gcc -funroll-loops -m32 -mcpu=G4 -maltivec -O2 -fomit-frame-pointer`	20130127	20130126
32704	`regs`	`gcc -m32 -mcpu=G4 -maltivec -Os -fomit-frame-pointer`	20130127	20130126
33880	`regs`	`gcc -m32 -mcpu=G4 -Os -fomit-frame-pointer`	20130127	20130126
35462	`regs`	`gcc -m32 -mcpu=G4 -O2 -fomit-frame-pointer`	20130127	20130126
40096	`regs`	`gcc -m32 -mcpu=G3 -maltivec -O2 -fomit-frame-pointer`	20130127	20130126
40180	`regs`	`gcc -m32 -mcpu=G3 -O3 -fomit-frame-pointer`	20130127	20130126
40537	`regs`	`gcc -funroll-loops -m32 -maltivec -O2 -fomit-frame-pointer`	20130127	20130126
40572	`regs`	`gcc -m32 -O3 -fomit-frame-pointer`	20130127	20130126
41104	`regs`	`gcc -funroll-loops -m32 -mcpu=G3 -O2 -fomit-frame-pointer`	20130127	20130126
41629	`regs`	`gcc -m32 -mcpu=G5 -O3 -fomit-frame-pointer`	20130127	20130126
41790	`regs`	`gcc -funroll-loops -m32 -mcpu=G5 -O2 -fomit-frame-pointer`	20130127	20130126
41825	`regs`	`gcc -m32 -mcpu=G4 -maltivec -O3 -fomit-frame-pointer`	20130127	20130126
41965	`regs`	`gcc -funroll-loops -m32 -mcpu=G4 -O3 -fomit-frame-pointer`	20130127	20130126
42119	`regs`	`gcc -funroll-loops -m32 -mcpu=G4 -maltivec -O3 -fomit-frame-pointer`	20130127	20130126
42693	`regs`	`gcc -funroll-loops -m32 -mcpu=G3 -maltivec -O2 -fomit-frame-pointer`	20130127	20130126
42714	`regs`	`gcc -funroll-loops -m32 -mcpu=G5 -maltivec -O2 -fomit-frame-pointer`	20130127	20130126
43267	`regs`	`gcc -funroll-loops -m32 -O2 -fomit-frame-pointer`	20130127	20130126
43533	`regs`	`gcc -m32 -maltivec -O2 -fomit-frame-pointer`	20130127	20130126
43855	`regs`	`gcc -funroll-loops -m32 -O3 -fomit-frame-pointer`	20130127	20130126
43876	`regs`	`gcc -m32 -maltivec -O3 -fomit-frame-pointer`	20130127	20130126
44646	`regs`	`gcc -m32 -O2 -fomit-frame-pointer`	20130127	20130126
45003	`regs`	`gcc -m32 -mcpu=G3 -maltivec -O3 -fomit-frame-pointer`	20130127	20130126
45073	`regs`	`gcc -m32 -mcpu=G5 -maltivec -O3 -fomit-frame-pointer`	20130127	20130126
45528	`regs`	`gcc -funroll-loops -m32 -mcpu=G3 -Os -fomit-frame-pointer`	20130127	20130126
45668	`regs`	`gcc -funroll-loops -m32 -maltivec -O3 -fomit-frame-pointer`	20130127	20130126
46144	`regs`	`gcc -m32 -mcpu=G4 -maltivec -O -fomit-frame-pointer`	20130127	20130126
46368	`regs`	`gcc -m32 -mcpu=G3 -maltivec -O -fomit-frame-pointer`	20130127	20130126
46480	`regs`	`gcc -funroll-loops -m32 -mcpu=G5 -maltivec -Os -fomit-frame-pointer`	20130127	20130126
46655	`regs`	`gcc -funroll-loops -m32 -mcpu=G5 -O3 -fomit-frame-pointer`	20130127	20130126
46718	`regs`	`gcc -m32 -mcpu=G4 -O -fomit-frame-pointer`	20130127	20130126
47152	`regs`	`gcc -funroll-loops -m32 -mcpu=G4 -Os -fomit-frame-pointer`	20130127	20130126
47474	`regs`	`gcc -m32 -mcpu=G5 -O -fomit-frame-pointer`	20130127	20130126
47684	`regs`	`gcc -m32 -maltivec -O -fomit-frame-pointer`	20130127	20130126
47726	`regs`	`gcc -funroll-loops -m32 -mcpu=G4 -maltivec -O -fomit-frame-pointer`	20130127	20130126
47782	`regs`	`gcc -m32 -mcpu=G4 -O3 -fomit-frame-pointer`	20130127	20130126
47789	`regs`	`gcc -m32 -mcpu=G4 -maltivec -O2 -fomit-frame-pointer`	20130127	20130126
47852	`regs`	`gcc -funroll-loops -m32 -mcpu=G3 -maltivec -O -fomit-frame-pointer`	20130127	20130126
47936	`regs`	`gcc -m32 -Os -fomit-frame-pointer`	20130127	20130126
48034	`regs`	`gcc -funroll-loops -m32 -Os -fomit-frame-pointer`	20130127	20130126
48272	`regs`	`gcc -funroll-loops -m32 -O -fomit-frame-pointer`	20130127	20130126
48272	`regs`	`gcc -m32 -mcpu=G5 -maltivec -O -fomit-frame-pointer`	20130127	20130126
48342	`regs`	`gcc -funroll-loops -m32 -mcpu=G5 -maltivec -O -fomit-frame-pointer`	20130127	20130126
48412	`regs`	`gcc -funroll-loops -m32 -mcpu=G4 -O -fomit-frame-pointer`	20130127	20130126
48503	`regs`	`gcc -funroll-loops -m32 -mcpu=G3 -O -fomit-frame-pointer`	20130127	20130126
48713	`regs`	`gcc -m32 -mcpu=G3 -Os -fomit-frame-pointer`	20130127	20130126
48720	`regs`	`gcc -m32 -O -fomit-frame-pointer`	20130127	20130126
48790	`regs`	`gcc -funroll-loops -m32 -maltivec -Os -fomit-frame-pointer`	20130127	20130126
48909	`regs`	`gcc -m32 -mcpu=G5 -maltivec -Os -fomit-frame-pointer`	20130127	20130126
49280	`regs`	`gcc -funroll-loops -m32 -maltivec -O -fomit-frame-pointer`	20130127	20130126
49756	`regs`	`gcc -funroll-loops -m32 -mcpu=G4 -maltivec -Os -fomit-frame-pointer`	20130127	20130126
49952	`regs`	`gcc -funroll-loops -m32 -mcpu=G3 -maltivec -Os -fomit-frame-pointer`	20130127	20130126
51191	`regs`	`gcc -m32 -mcpu=G3 -O -fomit-frame-pointer`	20130127	20130126
51730	`regs`	`gcc -m32 -mcpu=G3 -O2 -fomit-frame-pointer`	20130127	20130126
53536	`regs`	`gcc -funroll-loops -m32 -mcpu=G4 -O2 -fomit-frame-pointer`	20130127	20130126
53802	`regs`	`gcc -funroll-loops -m32 -mcpu=G5 -Os -fomit-frame-pointer`	20130127	20130126
55041	`regs`	`gcc -funroll-loops -m32 -mcpu=G3 -maltivec -O3 -fomit-frame-pointer`	20130127	20130126
55216	`regs`	`gcc -m32 -mcpu=G5 -Os -fomit-frame-pointer`	20130127	20130126
55244	`regs`	`gcc -funroll-loops -m32 -mcpu=G3 -O3 -fomit-frame-pointer`	20130127	20130126
57001	`regs`	`gcc -funroll-loops -m32 -mcpu=G5 -O -fomit-frame-pointer`	20130127	20130126
66234	`regs`	`gcc -m32 -maltivec -Os -fomit-frame-pointer`	20130127	20130126
66521	`regs`	`gcc -funroll-loops -m32 -mcpu=G5 -maltivec -O3 -fomit-frame-pointer`	20130127	20130126
74592	`regs`	`gcc -m32 -mcpu=G3 -maltivec -Os -fomit-frame-pointer`	20130127	20130126

Compiler output

Implementation: crypto_hash/blake2s/ref
Compiler: gcc -funroll-loops -m32 -O2 -fomit-frame-pointer

blake2s-ref.c: blake2s-ref.c: In function 'blake2s_init0':
blake2s-ref.c: blake2s-ref.c:137:3: error: 'for' loop initial declarations are only allowed in C99 mode
blake2s-ref.c: blake2s-ref.c:137:3: note: use option -std=c99 or -std=gnu99 to compile your code
blake2s-ref.c: blake2s-ref.c: In function 'blake2s_init_param':
blake2s-ref.c: blake2s-ref.c:149:3: error: 'for' loop initial declarations are only allowed in C99 mode
blake2s-ref.c: blake2s-ref.c: In function 'blake2s_compress':
blake2s-ref.c: blake2s-ref.c:215:3: error: 'for' loop initial declarations are only allowed in C99 mode
blake2s-ref.c: blake2s-ref.c:218:15: error: redefinition of 'i'
blake2s-ref.c: blake2s-ref.c:215:15: note: previous definition of 'i' was here
blake2s-ref.c: blake2s-ref.c:218:3: error: 'for' loop initial declarations are only allowed in C99 mode
blake2s-ref.c: blake2s-ref.c:262:15: error: redefinition of 'i'
blake2s-ref.c: blake2s-ref.c:218:15: note: previous definition of 'i' was here
blake2s-ref.c: blake2s-ref.c:262:3: error: 'for' loop initial declarations are only allowed in C99 mode
blake2s-ref.c: blake2s-ref.c: In function 'blake2s_final':
blake2s-ref.c: blake2s-ref.c:318:3: error: 'for' loop initial declarations are only allowed in C99 mode

Number of similar (compiler,implementation) pairs: 64, namely:

Compiler	Implementations
gcc -funroll-loops -m32 -O2 -fomit-frame-pointer	ref
gcc -funroll-loops -m32 -O3 -fomit-frame-pointer	ref
gcc -funroll-loops -m32 -O -fomit-frame-pointer	ref
gcc -funroll-loops -m32 -Os -fomit-frame-pointer	ref
gcc -funroll-loops -m32 -maltivec -O2 -fomit-frame-pointer	ref
gcc -funroll-loops -m32 -maltivec -O3 -fomit-frame-pointer	ref
gcc -funroll-loops -m32 -maltivec -O -fomit-frame-pointer	ref
gcc -funroll-loops -m32 -maltivec -Os -fomit-frame-pointer	ref
gcc -funroll-loops -m32 -mcpu=G3 -O2 -fomit-frame-pointer	ref
gcc -funroll-loops -m32 -mcpu=G3 -O3 -fomit-frame-pointer	ref
gcc -funroll-loops -m32 -mcpu=G3 -O -fomit-frame-pointer	ref
gcc -funroll-loops -m32 -mcpu=G3 -Os -fomit-frame-pointer	ref
gcc -funroll-loops -m32 -mcpu=G3 -maltivec -O2 -fomit-frame-pointer	ref
gcc -funroll-loops -m32 -mcpu=G3 -maltivec -O3 -fomit-frame-pointer	ref
gcc -funroll-loops -m32 -mcpu=G3 -maltivec -O -fomit-frame-pointer	ref
gcc -funroll-loops -m32 -mcpu=G3 -maltivec -Os -fomit-frame-pointer	ref
gcc -funroll-loops -m32 -mcpu=G4 -O2 -fomit-frame-pointer	ref
gcc -funroll-loops -m32 -mcpu=G4 -O3 -fomit-frame-pointer	ref
gcc -funroll-loops -m32 -mcpu=G4 -O -fomit-frame-pointer	ref
gcc -funroll-loops -m32 -mcpu=G4 -Os -fomit-frame-pointer	ref
gcc -funroll-loops -m32 -mcpu=G4 -maltivec -O2 -fomit-frame-pointer	ref
gcc -funroll-loops -m32 -mcpu=G4 -maltivec -O3 -fomit-frame-pointer	ref
gcc -funroll-loops -m32 -mcpu=G4 -maltivec -O -fomit-frame-pointer	ref
gcc -funroll-loops -m32 -mcpu=G4 -maltivec -Os -fomit-frame-pointer	ref
gcc -funroll-loops -m32 -mcpu=G5 -O2 -fomit-frame-pointer	ref
gcc -funroll-loops -m32 -mcpu=G5 -O3 -fomit-frame-pointer	ref
gcc -funroll-loops -m32 -mcpu=G5 -O -fomit-frame-pointer	ref
gcc -funroll-loops -m32 -mcpu=G5 -Os -fomit-frame-pointer	ref
gcc -funroll-loops -m32 -mcpu=G5 -maltivec -O2 -fomit-frame-pointer	ref
gcc -funroll-loops -m32 -mcpu=G5 -maltivec -O3 -fomit-frame-pointer	ref
gcc -funroll-loops -m32 -mcpu=G5 -maltivec -O -fomit-frame-pointer	ref
gcc -funroll-loops -m32 -mcpu=G5 -maltivec -Os -fomit-frame-pointer	ref
gcc -m32 -O2 -fomit-frame-pointer	ref
gcc -m32 -O3 -fomit-frame-pointer	ref
gcc -m32 -O -fomit-frame-pointer	ref
gcc -m32 -Os -fomit-frame-pointer	ref
gcc -m32 -maltivec -O2 -fomit-frame-pointer	ref
gcc -m32 -maltivec -O3 -fomit-frame-pointer	ref
gcc -m32 -maltivec -O -fomit-frame-pointer	ref
gcc -m32 -maltivec -Os -fomit-frame-pointer	ref
gcc -m32 -mcpu=G3 -O2 -fomit-frame-pointer	ref
gcc -m32 -mcpu=G3 -O3 -fomit-frame-pointer	ref
gcc -m32 -mcpu=G3 -O -fomit-frame-pointer	ref
gcc -m32 -mcpu=G3 -Os -fomit-frame-pointer	ref
gcc -m32 -mcpu=G3 -maltivec -O2 -fomit-frame-pointer	ref
gcc -m32 -mcpu=G3 -maltivec -O3 -fomit-frame-pointer	ref
gcc -m32 -mcpu=G3 -maltivec -O -fomit-frame-pointer	ref
gcc -m32 -mcpu=G3 -maltivec -Os -fomit-frame-pointer	ref
gcc -m32 -mcpu=G4 -O2 -fomit-frame-pointer	ref
gcc -m32 -mcpu=G4 -O3 -fomit-frame-pointer	ref
gcc -m32 -mcpu=G4 -O -fomit-frame-pointer	ref
gcc -m32 -mcpu=G4 -Os -fomit-frame-pointer	ref
gcc -m32 -mcpu=G4 -maltivec -O2 -fomit-frame-pointer	ref
gcc -m32 -mcpu=G4 -maltivec -O3 -fomit-frame-pointer	ref
gcc -m32 -mcpu=G4 -maltivec -O -fomit-frame-pointer	ref
gcc -m32 -mcpu=G4 -maltivec -Os -fomit-frame-pointer	ref
gcc -m32 -mcpu=G5 -O2 -fomit-frame-pointer	ref
gcc -m32 -mcpu=G5 -O3 -fomit-frame-pointer	ref
gcc -m32 -mcpu=G5 -O -fomit-frame-pointer	ref
gcc -m32 -mcpu=G5 -Os -fomit-frame-pointer	ref
gcc -m32 -mcpu=G5 -maltivec -O2 -fomit-frame-pointer	ref
gcc -m32 -mcpu=G5 -maltivec -O3 -fomit-frame-pointer	ref
gcc -m32 -mcpu=G5 -maltivec -O -fomit-frame-pointer	ref
gcc -m32 -mcpu=G5 -maltivec -Os -fomit-frame-pointer	ref

Compiler output

Implementation: crypto_hash/blake2s/xmm
Compiler: gcc -funroll-loops -m32 -O2 -fomit-frame-pointer

blake2s.c: In file included from blake2s.c:21:0:
blake2s.c: blake2-config.h:68:2: error: #error "This code requires at least SSE2."
blake2s.c: blake2s.c:24:23: fatal error: emmintrin.h: No such file or directory
blake2s.c: compilation terminated.

Number of similar (compiler,implementation) pairs: 64, namely:

Compiler	Implementations
gcc -funroll-loops -m32 -O2 -fomit-frame-pointer	xmm
gcc -funroll-loops -m32 -O3 -fomit-frame-pointer	xmm
gcc -funroll-loops -m32 -O -fomit-frame-pointer	xmm
gcc -funroll-loops -m32 -Os -fomit-frame-pointer	xmm
gcc -funroll-loops -m32 -maltivec -O2 -fomit-frame-pointer	xmm
gcc -funroll-loops -m32 -maltivec -O3 -fomit-frame-pointer	xmm
gcc -funroll-loops -m32 -maltivec -O -fomit-frame-pointer	xmm
gcc -funroll-loops -m32 -maltivec -Os -fomit-frame-pointer	xmm
gcc -funroll-loops -m32 -mcpu=G3 -O2 -fomit-frame-pointer	xmm
gcc -funroll-loops -m32 -mcpu=G3 -O3 -fomit-frame-pointer	xmm
gcc -funroll-loops -m32 -mcpu=G3 -O -fomit-frame-pointer	xmm
gcc -funroll-loops -m32 -mcpu=G3 -Os -fomit-frame-pointer	xmm
gcc -funroll-loops -m32 -mcpu=G3 -maltivec -O2 -fomit-frame-pointer	xmm
gcc -funroll-loops -m32 -mcpu=G3 -maltivec -O3 -fomit-frame-pointer	xmm
gcc -funroll-loops -m32 -mcpu=G3 -maltivec -O -fomit-frame-pointer	xmm
gcc -funroll-loops -m32 -mcpu=G3 -maltivec -Os -fomit-frame-pointer	xmm
gcc -funroll-loops -m32 -mcpu=G4 -O2 -fomit-frame-pointer	xmm
gcc -funroll-loops -m32 -mcpu=G4 -O3 -fomit-frame-pointer	xmm
gcc -funroll-loops -m32 -mcpu=G4 -O -fomit-frame-pointer	xmm
gcc -funroll-loops -m32 -mcpu=G4 -Os -fomit-frame-pointer	xmm
gcc -funroll-loops -m32 -mcpu=G4 -maltivec -O2 -fomit-frame-pointer	xmm
gcc -funroll-loops -m32 -mcpu=G4 -maltivec -O3 -fomit-frame-pointer	xmm
gcc -funroll-loops -m32 -mcpu=G4 -maltivec -O -fomit-frame-pointer	xmm
gcc -funroll-loops -m32 -mcpu=G4 -maltivec -Os -fomit-frame-pointer	xmm
gcc -funroll-loops -m32 -mcpu=G5 -O2 -fomit-frame-pointer	xmm
gcc -funroll-loops -m32 -mcpu=G5 -O3 -fomit-frame-pointer	xmm
gcc -funroll-loops -m32 -mcpu=G5 -O -fomit-frame-pointer	xmm
gcc -funroll-loops -m32 -mcpu=G5 -Os -fomit-frame-pointer	xmm
gcc -funroll-loops -m32 -mcpu=G5 -maltivec -O2 -fomit-frame-pointer	xmm
gcc -funroll-loops -m32 -mcpu=G5 -maltivec -O3 -fomit-frame-pointer	xmm
gcc -funroll-loops -m32 -mcpu=G5 -maltivec -O -fomit-frame-pointer	xmm
gcc -funroll-loops -m32 -mcpu=G5 -maltivec -Os -fomit-frame-pointer	xmm
gcc -m32 -O2 -fomit-frame-pointer	xmm
gcc -m32 -O3 -fomit-frame-pointer	xmm
gcc -m32 -O -fomit-frame-pointer	xmm
gcc -m32 -Os -fomit-frame-pointer	xmm
gcc -m32 -maltivec -O2 -fomit-frame-pointer	xmm
gcc -m32 -maltivec -O3 -fomit-frame-pointer	xmm
gcc -m32 -maltivec -O -fomit-frame-pointer	xmm
gcc -m32 -maltivec -Os -fomit-frame-pointer	xmm
gcc -m32 -mcpu=G3 -O2 -fomit-frame-pointer	xmm
gcc -m32 -mcpu=G3 -O3 -fomit-frame-pointer	xmm
gcc -m32 -mcpu=G3 -O -fomit-frame-pointer	xmm
gcc -m32 -mcpu=G3 -Os -fomit-frame-pointer	xmm
gcc -m32 -mcpu=G3 -maltivec -O2 -fomit-frame-pointer	xmm
gcc -m32 -mcpu=G3 -maltivec -O3 -fomit-frame-pointer	xmm
gcc -m32 -mcpu=G3 -maltivec -O -fomit-frame-pointer	xmm
gcc -m32 -mcpu=G3 -maltivec -Os -fomit-frame-pointer	xmm
gcc -m32 -mcpu=G4 -O2 -fomit-frame-pointer	xmm
gcc -m32 -mcpu=G4 -O3 -fomit-frame-pointer	xmm
gcc -m32 -mcpu=G4 -O -fomit-frame-pointer	xmm
gcc -m32 -mcpu=G4 -Os -fomit-frame-pointer	xmm
gcc -m32 -mcpu=G4 -maltivec -O2 -fomit-frame-pointer	xmm
gcc -m32 -mcpu=G4 -maltivec -O3 -fomit-frame-pointer	xmm
gcc -m32 -mcpu=G4 -maltivec -O -fomit-frame-pointer	xmm
gcc -m32 -mcpu=G4 -maltivec -Os -fomit-frame-pointer	xmm
gcc -m32 -mcpu=G5 -O2 -fomit-frame-pointer	xmm
gcc -m32 -mcpu=G5 -O3 -fomit-frame-pointer	xmm
gcc -m32 -mcpu=G5 -O -fomit-frame-pointer	xmm
gcc -m32 -mcpu=G5 -Os -fomit-frame-pointer	xmm
gcc -m32 -mcpu=G5 -maltivec -O2 -fomit-frame-pointer	xmm
gcc -m32 -mcpu=G5 -maltivec -O3 -fomit-frame-pointer	xmm
gcc -m32 -mcpu=G5 -maltivec -O -fomit-frame-pointer	xmm
gcc -m32 -mcpu=G5 -maltivec -Os -fomit-frame-pointer	xmm

Compiler output

Implementation: crypto_hash/blake2s/avxicc
Compiler: gcc -funroll-loops -m32 -O2 -fomit-frame-pointer

blake2s.s: blake2s.s: Assembler messages:
blake2s.s: blake2s.s:13: Error: Unrecognized opcode: `xorl'
blake2s.s: blake2s.s:14: Error: Unrecognized opcode: `xorl'
blake2s.s: blake2s.s:15: Error: Unrecognized opcode: `movq'
blake2s.s: blake2s.s:16: Error: Unrecognized opcode: `vmovdqu'
blake2s.s: blake2s.s:17: Error: Unrecognized opcode: `xorl'
blake2s.s: blake2s.s:18: Error: Unrecognized opcode: `vmovdqu'
blake2s.s: blake2s.s:19: Error: Unrecognized opcode: `vmovdqa'
blake2s.s: blake2s.s:20: Error: Unrecognized opcode: `vmovdqu'
blake2s.s: blake2s.s:21: Error: Unrecognized opcode: `movq'
blake2s.s: blake2s.s:22: Error: Unrecognized opcode: `vmovdqu'
blake2s.s: blake2s.s:23: Error: Unrecognized opcode: `xorl'
blake2s.s: blake2s.s:24: Error: Unrecognized opcode: `vmovdqu'
blake2s.s: blake2s.s:25: Error: Unrecognized opcode: `cmpq'
blake2s.s: blake2s.s:26: Error: Unrecognized opcode: `jbe'
blake2s.s: blake2s.s:28: Error: Unrecognized opcode: `movq'
blake2s.s: blake2s.s:29: Error: Unrecognized opcode: `lea'
blake2s.s: blake2s.s:30: Error: Unrecognized opcode: `sarq'
blake2s.s: blake2s.s:32: Error: Unrecognized opcode: `movq'
blake2s.s: blake2s.s:33: Error: Unrecognized opcode: `shrq'
blake2s.s: blake2s.s:34: Error: Unrecognized opcode: `movq'
blake2s.s: blake2s.s:35: Error: Unrecognized opcode: `lea'
blake2s.s: blake2s.s:36: Error: Unrecognized opcode: `sarq'
blake2s.s: blake2s.s:37: Error: Unrecognized opcode: `movq'
blake2s.s: blake2s.s:39: Error: Unrecognized opcode: `vmovdqu'
blake2s.s: ...

Number of similar (compiler,implementation) pairs: 64, namely:

Compiler	Implementations
gcc -funroll-loops -m32 -O2 -fomit-frame-pointer	avxicc
gcc -funroll-loops -m32 -O3 -fomit-frame-pointer	avxicc
gcc -funroll-loops -m32 -O -fomit-frame-pointer	avxicc
gcc -funroll-loops -m32 -Os -fomit-frame-pointer	avxicc
gcc -funroll-loops -m32 -maltivec -O2 -fomit-frame-pointer	avxicc
gcc -funroll-loops -m32 -maltivec -O3 -fomit-frame-pointer	avxicc
gcc -funroll-loops -m32 -maltivec -O -fomit-frame-pointer	avxicc
gcc -funroll-loops -m32 -maltivec -Os -fomit-frame-pointer	avxicc
gcc -funroll-loops -m32 -mcpu=G3 -O2 -fomit-frame-pointer	avxicc
gcc -funroll-loops -m32 -mcpu=G3 -O3 -fomit-frame-pointer	avxicc
gcc -funroll-loops -m32 -mcpu=G3 -O -fomit-frame-pointer	avxicc
gcc -funroll-loops -m32 -mcpu=G3 -Os -fomit-frame-pointer	avxicc
gcc -funroll-loops -m32 -mcpu=G3 -maltivec -O2 -fomit-frame-pointer	avxicc
gcc -funroll-loops -m32 -mcpu=G3 -maltivec -O3 -fomit-frame-pointer	avxicc
gcc -funroll-loops -m32 -mcpu=G3 -maltivec -O -fomit-frame-pointer	avxicc
gcc -funroll-loops -m32 -mcpu=G3 -maltivec -Os -fomit-frame-pointer	avxicc
gcc -funroll-loops -m32 -mcpu=G4 -O2 -fomit-frame-pointer	avxicc
gcc -funroll-loops -m32 -mcpu=G4 -O3 -fomit-frame-pointer	avxicc
gcc -funroll-loops -m32 -mcpu=G4 -O -fomit-frame-pointer	avxicc
gcc -funroll-loops -m32 -mcpu=G4 -Os -fomit-frame-pointer	avxicc
gcc -funroll-loops -m32 -mcpu=G4 -maltivec -O2 -fomit-frame-pointer	avxicc
gcc -funroll-loops -m32 -mcpu=G4 -maltivec -O3 -fomit-frame-pointer	avxicc
gcc -funroll-loops -m32 -mcpu=G4 -maltivec -O -fomit-frame-pointer	avxicc
gcc -funroll-loops -m32 -mcpu=G4 -maltivec -Os -fomit-frame-pointer	avxicc
gcc -funroll-loops -m32 -mcpu=G5 -O2 -fomit-frame-pointer	avxicc
gcc -funroll-loops -m32 -mcpu=G5 -O3 -fomit-frame-pointer	avxicc
gcc -funroll-loops -m32 -mcpu=G5 -O -fomit-frame-pointer	avxicc
gcc -funroll-loops -m32 -mcpu=G5 -Os -fomit-frame-pointer	avxicc
gcc -funroll-loops -m32 -mcpu=G5 -maltivec -O2 -fomit-frame-pointer	avxicc
gcc -funroll-loops -m32 -mcpu=G5 -maltivec -O3 -fomit-frame-pointer	avxicc
gcc -funroll-loops -m32 -mcpu=G5 -maltivec -O -fomit-frame-pointer	avxicc
gcc -funroll-loops -m32 -mcpu=G5 -maltivec -Os -fomit-frame-pointer	avxicc
gcc -m32 -O2 -fomit-frame-pointer	avxicc
gcc -m32 -O3 -fomit-frame-pointer	avxicc
gcc -m32 -O -fomit-frame-pointer	avxicc
gcc -m32 -Os -fomit-frame-pointer	avxicc
gcc -m32 -maltivec -O2 -fomit-frame-pointer	avxicc
gcc -m32 -maltivec -O3 -fomit-frame-pointer	avxicc
gcc -m32 -maltivec -O -fomit-frame-pointer	avxicc
gcc -m32 -maltivec -Os -fomit-frame-pointer	avxicc
gcc -m32 -mcpu=G3 -O2 -fomit-frame-pointer	avxicc
gcc -m32 -mcpu=G3 -O3 -fomit-frame-pointer	avxicc
gcc -m32 -mcpu=G3 -O -fomit-frame-pointer	avxicc
gcc -m32 -mcpu=G3 -Os -fomit-frame-pointer	avxicc
gcc -m32 -mcpu=G3 -maltivec -O2 -fomit-frame-pointer	avxicc
gcc -m32 -mcpu=G3 -maltivec -O3 -fomit-frame-pointer	avxicc
gcc -m32 -mcpu=G3 -maltivec -O -fomit-frame-pointer	avxicc
gcc -m32 -mcpu=G3 -maltivec -Os -fomit-frame-pointer	avxicc
gcc -m32 -mcpu=G4 -O2 -fomit-frame-pointer	avxicc
gcc -m32 -mcpu=G4 -O3 -fomit-frame-pointer	avxicc
gcc -m32 -mcpu=G4 -O -fomit-frame-pointer	avxicc
gcc -m32 -mcpu=G4 -Os -fomit-frame-pointer	avxicc
gcc -m32 -mcpu=G4 -maltivec -O2 -fomit-frame-pointer	avxicc
gcc -m32 -mcpu=G4 -maltivec -O3 -fomit-frame-pointer	avxicc
gcc -m32 -mcpu=G4 -maltivec -O -fomit-frame-pointer	avxicc
gcc -m32 -mcpu=G4 -maltivec -Os -fomit-frame-pointer	avxicc
gcc -m32 -mcpu=G5 -O2 -fomit-frame-pointer	avxicc
gcc -m32 -mcpu=G5 -O3 -fomit-frame-pointer	avxicc
gcc -m32 -mcpu=G5 -O -fomit-frame-pointer	avxicc
gcc -m32 -mcpu=G5 -Os -fomit-frame-pointer	avxicc
gcc -m32 -mcpu=G5 -maltivec -O2 -fomit-frame-pointer	avxicc
gcc -m32 -mcpu=G5 -maltivec -O3 -fomit-frame-pointer	avxicc
gcc -m32 -mcpu=G5 -maltivec -O -fomit-frame-pointer	avxicc
gcc -m32 -mcpu=G5 -maltivec -Os -fomit-frame-pointer	avxicc