SSE opcodes to TCC assembler (i386, x86_64)

patch from Anaël Seghezzi a test program: ============================ #include <stdio.h> struct fl4{ float x, y, z, w; }; void asm_test(void) { struct fl4 v1, v2, v3; v1.x = 0.1; v1.y = 0.2; v1.z = 0.4; v1.w = 0.3; v2.x = 0.11; v2.y = 0.0; v2.z = 0.01; v2.w = 0.04; asm volatile ( "movups %0, %%xmm0;" "movups %1, %%xmm1;" "addps %%xmm1, %%xmm0;" "movups %%xmm0, %2" :: "g" (v1), "g" (v2), "g" (v3) : "memory"); printf("sse fl4 add : %f %f %f %f\n", v3.x, v3.y, v3.z, v3.w); printf("expected : %f %f %f %f\n", v1.x+v2.x, v1.y+v2.y, v1.z+v2.z, v1.w+v2.w); } int main() { asm_test(); } /* sse fl4 add : 0.210000 0.200000 0.410000 0.340000 expected : 0.210000 0.200000 0.410000 0.340000 */ ============================

seyko committed 11 years ago

Commit eb870b01 parent 8f620c82 files · +54 −2Browse files

Unified · Split

2 files

i386-asm.h +27 −1
x86_64-asm.h +27 −1

i386-asm.h+27−1

@@ -464,7 +464,33 @@ ALT(DEF_ASM_OP2(psrlq, 0x0f73, 2, OPC_MODRM, OPT_IM8, OPT_MMX )) · 463 unmodified lines
464	DEF_ASM_OP2(punpcklbw, 0x0f60, 0, OPC_MODRM, OPT_EA \| OPT_MMX, OPT_MMX )	464	DEF_ASM_OP2(punpcklbw, 0x0f60, 0, OPC_MODRM, OPT_EA \| OPT_MMX, OPT_MMX )
465	DEF_ASM_OP2(punpcklwd, 0x0f61, 0, OPC_MODRM, OPT_EA \| OPT_MMX, OPT_MMX )	465	DEF_ASM_OP2(punpcklwd, 0x0f61, 0, OPC_MODRM, OPT_EA \| OPT_MMX, OPT_MMX )
466	DEF_ASM_OP2(punpckldq, 0x0f62, 0, OPC_MODRM, OPT_EA \| OPT_MMX, OPT_MMX )	466	DEF_ASM_OP2(punpckldq, 0x0f62, 0, OPC_MODRM, OPT_EA \| OPT_MMX, OPT_MMX )
467	DEF_ASM_OP2(pxor, 0x0fef, 0, OPC_MODRM, OPT_EA \| OPT_MMX, OPT_MMX ) /* must be last !OP0 */	467	DEF_ASM_OP2(pxor, 0x0fef, 0, OPC_MODRM, OPT_EA \| OPT_MMX, OPT_MMX )
		468
		469	/* sse */
		470	DEF_ASM_OP2(movups, 0x0f10, 0, OPC_MODRM, OPT_EA \| OPT_REG32, OPT_SSE )
		471	ALT(DEF_ASM_OP2(movups, 0x0f11, 0, OPC_MODRM, OPT_SSE, OPT_EA \| OPT_REG32 ))
		472	DEF_ASM_OP2(movaps, 0x0f28, 0, OPC_MODRM, OPT_EA \| OPT_REG32, OPT_SSE )
		473	ALT(DEF_ASM_OP2(movaps, 0x0f29, 0, OPC_MODRM, OPT_SSE, OPT_EA \| OPT_REG32 ))
		474	DEF_ASM_OP2(movhps, 0x0f16, 0, OPC_MODRM, OPT_EA \| OPT_REG32, OPT_SSE )
		475	ALT(DEF_ASM_OP2(movhps, 0x0f17, 0, OPC_MODRM, OPT_SSE, OPT_EA \| OPT_REG32 ))
		476	DEF_ASM_OP2(addps, 0x0f58, 0, OPC_MODRM, OPT_EA \| OPT_SSE, OPT_SSE )
		477	DEF_ASM_OP2(cvtpi2ps, 0x0f2a, 0, OPC_MODRM, OPT_EA \| OPT_MMX, OPT_SSE )
		478	DEF_ASM_OP2(cvtps2pi, 0x0f2d, 0, OPC_MODRM, OPT_EA \| OPT_SSE, OPT_MMX )
		479	DEF_ASM_OP2(cvttps2pi, 0x0f2c, 0, OPC_MODRM, OPT_EA \| OPT_SSE, OPT_MMX )
		480	DEF_ASM_OP2(divps, 0x0f5e, 0, OPC_MODRM, OPT_EA \| OPT_SSE, OPT_SSE )
		481	DEF_ASM_OP2(maxps, 0x0f5f, 0, OPC_MODRM, OPT_EA \| OPT_SSE, OPT_SSE )
		482	DEF_ASM_OP2(minps, 0x0f5d, 0, OPC_MODRM, OPT_EA \| OPT_SSE, OPT_SSE )
		483	DEF_ASM_OP2(mulps, 0x0f59, 0, OPC_MODRM, OPT_EA \| OPT_SSE, OPT_SSE )
		484	DEF_ASM_OP2(pavgb, 0x0fe0, 0, OPC_MODRM, OPT_EA \| OPT_SSE, OPT_SSE )
		485	DEF_ASM_OP2(pavgw, 0x0fe3, 0, OPC_MODRM, OPT_EA \| OPT_SSE, OPT_SSE )
		486	DEF_ASM_OP2(pmaxsw, 0x0fee, 0, OPC_MODRM, OPT_EA \| OPT_MMX, OPT_MMX )
		487	DEF_ASM_OP2(pmaxub, 0x0fde, 0, OPC_MODRM, OPT_EA \| OPT_MMX, OPT_MMX )
		488	DEF_ASM_OP2(pminsw, 0x0fea, 0, OPC_MODRM, OPT_EA \| OPT_MMX, OPT_MMX )
		489	DEF_ASM_OP2(pminub, 0x0fda, 0, OPC_MODRM, OPT_EA \| OPT_MMX, OPT_MMX )
		490	DEF_ASM_OP2(rcpss, 0x0f53, 0, OPC_MODRM, OPT_EA \| OPT_SSE, OPT_SSE )
		491	DEF_ASM_OP2(rsqrtps, 0x0f52, 0, OPC_MODRM, OPT_EA \| OPT_SSE, OPT_SSE )
		492	DEF_ASM_OP2(sqrtps, 0x0f51, 0, OPC_MODRM, OPT_EA \| OPT_SSE, OPT_SSE )
		493	DEF_ASM_OP2(subps, 0x0f5c, 0, OPC_MODRM, OPT_EA \| OPT_SSE, OPT_SSE )
468		494
469	#undef ALT	495	#undef ALT
470	#undef DEF_ASM_OP0	496	#undef DEF_ASM_OP0

x86_64-asm.h+27−1

@@ -439,7 +439,33 @@ ALT(DEF_ASM_OP2(psrlq, 0x0f73, 2, OPC_MODRM, OPT_IM8, OPT_MMX )) · 438 unmodified lines
439	DEF_ASM_OP2(punpcklbw, 0x0f60, 0, OPC_MODRM, OPT_EA \| OPT_MMX, OPT_MMX )	439	DEF_ASM_OP2(punpcklbw, 0x0f60, 0, OPC_MODRM, OPT_EA \| OPT_MMX, OPT_MMX )
440	DEF_ASM_OP2(punpcklwd, 0x0f61, 0, OPC_MODRM, OPT_EA \| OPT_MMX, OPT_MMX )	440	DEF_ASM_OP2(punpcklwd, 0x0f61, 0, OPC_MODRM, OPT_EA \| OPT_MMX, OPT_MMX )
441	DEF_ASM_OP2(punpckldq, 0x0f62, 0, OPC_MODRM, OPT_EA \| OPT_MMX, OPT_MMX )	441	DEF_ASM_OP2(punpckldq, 0x0f62, 0, OPC_MODRM, OPT_EA \| OPT_MMX, OPT_MMX )
442	DEF_ASM_OP2(pxor, 0x0fef, 0, OPC_MODRM, OPT_EA \| OPT_MMX, OPT_MMX ) /* must be last !OP0 */	442	DEF_ASM_OP2(pxor, 0x0fef, 0, OPC_MODRM, OPT_EA \| OPT_MMX, OPT_MMX )
		443
		444	/* sse */
		445	DEF_ASM_OP2(movups, 0x0f10, 0, OPC_MODRM, OPT_EA \| OPT_REG32, OPT_SSE )
		446	ALT(DEF_ASM_OP2(movups, 0x0f11, 0, OPC_MODRM, OPT_SSE, OPT_EA \| OPT_REG32 ))
		447	DEF_ASM_OP2(movaps, 0x0f28, 0, OPC_MODRM, OPT_EA \| OPT_REG32, OPT_SSE )
		448	ALT(DEF_ASM_OP2(movaps, 0x0f29, 0, OPC_MODRM, OPT_SSE, OPT_EA \| OPT_REG32 ))
		449	DEF_ASM_OP2(movhps, 0x0f16, 0, OPC_MODRM, OPT_EA \| OPT_REG32, OPT_SSE )
		450	ALT(DEF_ASM_OP2(movhps, 0x0f17, 0, OPC_MODRM, OPT_SSE, OPT_EA \| OPT_REG32 ))
		451	DEF_ASM_OP2(addps, 0x0f58, 0, OPC_MODRM, OPT_EA \| OPT_SSE, OPT_SSE )
		452	DEF_ASM_OP2(cvtpi2ps, 0x0f2a, 0, OPC_MODRM, OPT_EA \| OPT_MMX, OPT_SSE )
		453	DEF_ASM_OP2(cvtps2pi, 0x0f2d, 0, OPC_MODRM, OPT_EA \| OPT_SSE, OPT_MMX )
		454	DEF_ASM_OP2(cvttps2pi, 0x0f2c, 0, OPC_MODRM, OPT_EA \| OPT_SSE, OPT_MMX )
		455	DEF_ASM_OP2(divps, 0x0f5e, 0, OPC_MODRM, OPT_EA \| OPT_SSE, OPT_SSE )
		456	DEF_ASM_OP2(maxps, 0x0f5f, 0, OPC_MODRM, OPT_EA \| OPT_SSE, OPT_SSE )
		457	DEF_ASM_OP2(minps, 0x0f5d, 0, OPC_MODRM, OPT_EA \| OPT_SSE, OPT_SSE )
		458	DEF_ASM_OP2(mulps, 0x0f59, 0, OPC_MODRM, OPT_EA \| OPT_SSE, OPT_SSE )
		459	DEF_ASM_OP2(pavgb, 0x0fe0, 0, OPC_MODRM, OPT_EA \| OPT_SSE, OPT_SSE )
		460	DEF_ASM_OP2(pavgw, 0x0fe3, 0, OPC_MODRM, OPT_EA \| OPT_SSE, OPT_SSE )
		461	DEF_ASM_OP2(pmaxsw, 0x0fee, 0, OPC_MODRM, OPT_EA \| OPT_MMX, OPT_MMX )
		462	DEF_ASM_OP2(pmaxub, 0x0fde, 0, OPC_MODRM, OPT_EA \| OPT_MMX, OPT_MMX )
		463	DEF_ASM_OP2(pminsw, 0x0fea, 0, OPC_MODRM, OPT_EA \| OPT_MMX, OPT_MMX )
		464	DEF_ASM_OP2(pminub, 0x0fda, 0, OPC_MODRM, OPT_EA \| OPT_MMX, OPT_MMX )
		465	DEF_ASM_OP2(rcpss, 0x0f53, 0, OPC_MODRM, OPT_EA \| OPT_SSE, OPT_SSE )
		466	DEF_ASM_OP2(rsqrtps, 0x0f52, 0, OPC_MODRM, OPT_EA \| OPT_SSE, OPT_SSE )
		467	DEF_ASM_OP2(sqrtps, 0x0f51, 0, OPC_MODRM, OPT_EA \| OPT_SSE, OPT_SSE )
		468	DEF_ASM_OP2(subps, 0x0f5c, 0, OPC_MODRM, OPT_EA \| OPT_SSE, OPT_SSE )
443		469
444	#undef ALT	470	#undef ALT
445	#undef DEF_ASM_OP0	471	#undef DEF_ASM_OP0

SSE opcodes to TCC assembler (i386, x86_64)

Keyboard shortcuts

General

Issues and pull requests

Issues, pull requests and Inbox

Repository code