generic-library/vpx
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

Merge "Tune SSSE3 assembly implementation to improve quantization speed"

Browse Source
This commit is contained in:
Jingning Han 2015-04-03 11:24:28 -07:00 committed by Gerrit Code Review
commit 30e9c091c0
1 changed files with 4 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
vp9/encoder/x86/vp9_quantize_ssse3_x86_64.asm
View File

@ -282,6 +282,8 @@ cglobal quantize_%1, 0, %2, 15, coeff, ncoeff, skip, zbin, round, quant, \
psignw m8, m9 psignw m8, m9
psignw m13, m10 psignw m13, m10
psrlw m0, m3, 2 psrlw m0, m3, 2
%else
psrlw m0, m3, 1
%endif %endif
mova [r4q+ncoeffq*2+ 0], m8 mova [r4q+ncoeffq*2+ 0], m8
mova [r4q+ncoeffq*2+16], m13 mova [r4q+ncoeffq*2+16], m13
@ -302,7 +304,7 @@ cglobal quantize_%1, 0, %2, 15, coeff, ncoeff, skip, zbin, round, quant, \
mova m10, [ coeffq+ncoeffq*2+16] ; m10 = c[i] mova m10, [ coeffq+ncoeffq*2+16] ; m10 = c[i]
pabsw m6, m9 ; m6 = abs(m9) pabsw m6, m9 ; m6 = abs(m9)
pabsw m11, m10 ; m11 = abs(m10) pabsw m11, m10 ; m11 = abs(m10)
%ifidn %1, fp_32x32
pcmpgtw m7, m6, m0 pcmpgtw m7, m6, m0
pcmpgtw m12, m11, m0 pcmpgtw m12, m11, m0
pmovmskb r6d, m7 pmovmskb r6d, m7
@ -310,7 +312,7 @@ cglobal quantize_%1, 0, %2, 15, coeff, ncoeff, skip, zbin, round, quant, \
or r6, r2 or r6, r2
jz .skip_iter jz .skip_iter
%endif
pcmpeqw m7, m7 pcmpeqw m7, m7
paddsw m6, m1 ; m6 += round paddsw m6, m1 ; m6 += round
@ -348,7 +350,6 @@ cglobal quantize_%1, 0, %2, 15, coeff, ncoeff, skip, zbin, round, quant, \
add ncoeffq, mmsize add ncoeffq, mmsize
jl .ac_only_loop jl .ac_only_loop
%ifidn %1, fp_32x32
jmp .accumulate_eob jmp .accumulate_eob
.skip_iter: .skip_iter:
mova [r3q+ncoeffq*2+ 0], m5 mova [r3q+ncoeffq*2+ 0], m5
@ -357,7 +358,6 @@ cglobal quantize_%1, 0, %2, 15, coeff, ncoeff, skip, zbin, round, quant, \
mova [r4q+ncoeffq*2+16], m5 mova [r4q+ncoeffq*2+16], m5
add ncoeffq, mmsize add ncoeffq, mmsize
jl .ac_only_loop jl .ac_only_loop
%endif
.accumulate_eob: .accumulate_eob:
; horizontally accumulate/max eobs and write into [eob] memory pointer ; horizontally accumulate/max eobs and write into [eob] memory pointer