# HG changeset patch # User michael # Date 1187654360 0 # Node ID 2e63bd8e894ddd00009caf31bb599136f59d9d75 # Parent 9993d38f59cf0560d2eeb0608ee772d99aa453e4 Simplify and speedup code, reduce needed headroom by 2 bits in the 3rd vertical lifting step of the SSE2 code (untested) diff -r 9993d38f59cf -r 2e63bd8e894d i386/snowdsp_mmx.c --- a/i386/snowdsp_mmx.c Mon Aug 20 23:54:49 2007 +0000 +++ b/i386/snowdsp_mmx.c Mon Aug 20 23:59:20 2007 +0000 @@ -483,18 +483,17 @@ snow_vertical_compose_sse2_sub("xmm1","xmm3","xmm5","xmm7","xmm0","xmm2","xmm4","xmm6") snow_vertical_compose_sse2_store(REG_S,"xmm0","xmm2","xmm4","xmm6") "mov %2, %%"REG_a" \n\t" - snow_vertical_compose_sse2_load(REG_c,"xmm1","xmm3","xmm5","xmm7") snow_vertical_compose_sse2_add(REG_a,"xmm0","xmm2","xmm4","xmm6") - snow_vertical_compose_sse2_sll("2","xmm1","xmm3","xmm5","xmm7")\ - snow_vertical_compose_sse2_r2r_add("xmm1","xmm3","xmm5","xmm7","xmm0","xmm2","xmm4","xmm6") + snow_vertical_compose_sse2_sra("2","xmm0","xmm2","xmm4","xmm6") + snow_vertical_compose_sse2_add(REG_c,"xmm0","xmm2","xmm4","xmm6") "pcmpeqd %%xmm1, %%xmm1 \n\t" "pslld $31, %%xmm1 \n\t" - "psrld $28, %%xmm1 \n\t" + "psrld $30, %%xmm1 \n\t" "mov %1, %%"REG_S" \n\t" snow_vertical_compose_sse2_r2r_add("xmm1","xmm1","xmm1","xmm1","xmm0","xmm2","xmm4","xmm6") - snow_vertical_compose_sse2_sra("4","xmm0","xmm2","xmm4","xmm6") + snow_vertical_compose_sse2_sra("2","xmm0","xmm2","xmm4","xmm6") snow_vertical_compose_sse2_add(REG_c,"xmm0","xmm2","xmm4","xmm6") snow_vertical_compose_sse2_store(REG_c,"xmm0","xmm2","xmm4","xmm6") snow_vertical_compose_sse2_add(REG_S,"xmm0","xmm2","xmm4","xmm6")