ffmpeg/FFmpeg
1
0
Fork 0
mirror of https://git.ffmpeg.org/ffmpeg.git synced 2024-09-22 06:16:38 +00:00
Code Issues Packages Projects Releases Wiki Activity

avcodec/x86/hevc_mc: fix sse register counts

Browse Source 
These fix failures of --enable-xmm-clobber-test
It would be better to change the code to use fewer registers, but until
someone does the used register count must not be too small

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
This commit is contained in:
Michael Niedermayer 2014-12-10 07:48:52 +01:00
parent dbdc6e554e
commit a03f72e744
1 changed files with 14 additions and 14 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

28
libavcodec/x86/hevc_mc.asm
View File

@ -586,7 +586,7 @@ cglobal hevc_put_hevc_bi_pel_pixels%1_%2, 6, 6, 6, dst, dststride, src, srcstrid
%macro HEVC_PUT_HEVC_EPEL 2 %macro HEVC_PUT_HEVC_EPEL 2
cglobal hevc_put_hevc_epel_h%1_%2, 5, 6, 6, dst, src, srcstride, height, mx, rfilter cglobal hevc_put_hevc_epel_h%1_%2, 5, 6, 11, dst, src, srcstride, height, mx, rfilter
%assign %%stride ((%2 + 7)/8) %assign %%stride ((%2 + 7)/8)
EPEL_FILTER %2, mx, m4, m5 EPEL_FILTER %2, mx, m4, m5
.loop .loop
@ -596,7 +596,7 @@ cglobal hevc_put_hevc_epel_h%1_%2, 5, 6, 6, dst, src, srcstride, height, mx, rfi
LOOP_END dst, src, srcstride LOOP_END dst, src, srcstride
RET RET
cglobal hevc_put_hevc_uni_epel_h%1_%2, 6, 7, 7, dst, dststride, src, srcstride, height, mx, rfilter cglobal hevc_put_hevc_uni_epel_h%1_%2, 6, 7, 11, dst, dststride, src, srcstride, height, mx, rfilter
%assign %%stride ((%2 + 7)/8) %assign %%stride ((%2 + 7)/8)
movdqa m6, [pw_%2] movdqa m6, [pw_%2]
EPEL_FILTER %2, mx, m4, m5 EPEL_FILTER %2, mx, m4, m5
@ -611,7 +611,7 @@ cglobal hevc_put_hevc_uni_epel_h%1_%2, 6, 7, 7, dst, dststride, src, srcstride,
jnz .loop ; height loop jnz .loop ; height loop
RET RET
cglobal hevc_put_hevc_bi_epel_h%1_%2, 7, 8, 7, dst, dststride, src, srcstride, src2, height, mx, rfilter cglobal hevc_put_hevc_bi_epel_h%1_%2, 7, 8, 11, dst, dststride, src, srcstride, src2, height, mx, rfilter
movdqa m6, [pw_bi_%2] movdqa m6, [pw_bi_%2]
EPEL_FILTER %2, mx, m4, m5 EPEL_FILTER %2, mx, m4, m5
.loop .loop
@ -634,7 +634,7 @@ cglobal hevc_put_hevc_bi_epel_h%1_%2, 7, 8, 7, dst, dststride, src, srcstride, s
; int16_t* mcbuffer) ; int16_t* mcbuffer)
; ****************************** ; ******************************
cglobal hevc_put_hevc_epel_v%1_%2, 6, 7, 6, dst, src, srcstride, height, r3src, my, rfilter cglobal hevc_put_hevc_epel_v%1_%2, 6, 7, 11, dst, src, srcstride, height, r3src, my, rfilter
lea r3srcq, [srcstrideq*3] lea r3srcq, [srcstrideq*3]
sub srcq, srcstrideq sub srcq, srcstrideq
EPEL_FILTER %2, my, m4, m5 EPEL_FILTER %2, my, m4, m5
@ -645,7 +645,7 @@ cglobal hevc_put_hevc_epel_v%1_%2, 6, 7, 6, dst, src, srcstride, height, r3src,
LOOP_END dst, src, srcstride LOOP_END dst, src, srcstride
RET RET
cglobal hevc_put_hevc_uni_epel_v%1_%2, 7, 8, 7, dst, dststride, src, srcstride, height, r3src, my, rfilter cglobal hevc_put_hevc_uni_epel_v%1_%2, 7, 8, 11, dst, dststride, src, srcstride, height, r3src, my, rfilter
lea r3srcq, [srcstrideq*3] lea r3srcq, [srcstrideq*3]
movdqa m6, [pw_%2] movdqa m6, [pw_%2]
sub srcq, srcstrideq sub srcq, srcstrideq
@ -662,7 +662,7 @@ cglobal hevc_put_hevc_uni_epel_v%1_%2, 7, 8, 7, dst, dststride, src, srcstride,
RET RET
cglobal hevc_put_hevc_bi_epel_v%1_%2, 8, 9, 7, dst, dststride, src, srcstride, src2, height, r3src, my, rfilter cglobal hevc_put_hevc_bi_epel_v%1_%2, 8, 9, 11, dst, dststride, src, srcstride, src2, height, r3src, my, rfilter
lea r3srcq, [srcstrideq*3] lea r3srcq, [srcstrideq*3]
movdqa m6, [pw_bi_%2] movdqa m6, [pw_bi_%2]
sub srcq, srcstrideq sub srcq, srcstrideq
@ -689,7 +689,7 @@ cglobal hevc_put_hevc_bi_epel_v%1_%2, 8, 9, 7, dst, dststride, src, srcstride, s
; ****************************** ; ******************************
%macro HEVC_PUT_HEVC_EPEL_HV 2 %macro HEVC_PUT_HEVC_EPEL_HV 2
cglobal hevc_put_hevc_epel_hv%1_%2, 6, 8, 12 , dst, src, srcstride, height, mx, my, r3src, rfilter cglobal hevc_put_hevc_epel_hv%1_%2, 6, 8, 16 , dst, src, srcstride, height, mx, my, r3src, rfilter
%assign %%stride ((%2 + 7)/8) %assign %%stride ((%2 + 7)/8)
sub srcq, srcstrideq sub srcq, srcstrideq
EPEL_HV_FILTER %2 EPEL_HV_FILTER %2
@ -723,7 +723,7 @@ cglobal hevc_put_hevc_epel_hv%1_%2, 6, 8, 12 , dst, src, srcstride, height, mx,
LOOP_END dst, src, srcstride LOOP_END dst, src, srcstride
RET RET
cglobal hevc_put_hevc_uni_epel_hv%1_%2, 7, 9, 12 , dst, dststride, src, srcstride, height, mx, my, r3src, rfilter cglobal hevc_put_hevc_uni_epel_hv%1_%2, 7, 9, 16 , dst, dststride, src, srcstride, height, mx, my, r3src, rfilter
%assign %%stride ((%2 + 7)/8) %assign %%stride ((%2 + 7)/8)
sub srcq, srcstrideq sub srcq, srcstrideq
EPEL_HV_FILTER %2 EPEL_HV_FILTER %2
@ -810,7 +810,7 @@ cglobal hevc_put_hevc_bi_epel_hv%1_%2, 8, 10, 16, dst, dststride, src, srcstride
; ****************************** ; ******************************
%macro HEVC_PUT_HEVC_QPEL 2 %macro HEVC_PUT_HEVC_QPEL 2
cglobal hevc_put_hevc_qpel_h%1_%2, 5, 6, 15, dst, src, srcstride, height, mx, rfilter cglobal hevc_put_hevc_qpel_h%1_%2, 5, 6, 16, dst, src, srcstride, height, mx, rfilter
QPEL_FILTER %2, mx QPEL_FILTER %2, mx
.loop .loop
QPEL_H_LOAD %2, srcq, %1, 10 QPEL_H_LOAD %2, srcq, %1, 10
@ -822,7 +822,7 @@ cglobal hevc_put_hevc_qpel_h%1_%2, 5, 6, 15, dst, src, srcstride, height, mx, rf
LOOP_END dst, src, srcstride LOOP_END dst, src, srcstride
RET RET
cglobal hevc_put_hevc_uni_qpel_h%1_%2, 6, 7, 15 , dst, dststride, src, srcstride, height, mx, rfilter cglobal hevc_put_hevc_uni_qpel_h%1_%2, 6, 7, 16 , dst, dststride, src, srcstride, height, mx, rfilter
movdqa m9, [pw_%2] movdqa m9, [pw_%2]
QPEL_FILTER %2, mx QPEL_FILTER %2, mx
.loop .loop
@ -865,7 +865,7 @@ cglobal hevc_put_hevc_bi_qpel_h%1_%2, 7, 8, 16 , dst, dststride, src, srcstride,
; int width, int height, int mx, int my) ; int width, int height, int mx, int my)
; ****************************** ; ******************************
cglobal hevc_put_hevc_qpel_v%1_%2, 6, 8, 15, dst, src, srcstride, height, r3src, my, rfilter cglobal hevc_put_hevc_qpel_v%1_%2, 6, 8, 16, dst, src, srcstride, height, r3src, my, rfilter
lea r3srcq, [srcstrideq*3] lea r3srcq, [srcstrideq*3]
QPEL_FILTER %2, my QPEL_FILTER %2, my
.loop .loop
@ -878,7 +878,7 @@ cglobal hevc_put_hevc_qpel_v%1_%2, 6, 8, 15, dst, src, srcstride, height, r3src,
LOOP_END dst, src, srcstride LOOP_END dst, src, srcstride
RET RET
cglobal hevc_put_hevc_uni_qpel_v%1_%2, 7, 9, 15, dst, dststride, src, srcstride, height, r3src, my, rfilter cglobal hevc_put_hevc_uni_qpel_v%1_%2, 7, 9, 16, dst, dststride, src, srcstride, height, r3src, my, rfilter
movdqa m9, [pw_%2] movdqa m9, [pw_%2]
lea r3srcq, [srcstrideq*3] lea r3srcq, [srcstrideq*3]
QPEL_FILTER %2, my QPEL_FILTER %2, my
@ -924,7 +924,7 @@ cglobal hevc_put_hevc_bi_qpel_v%1_%2, 8, 10, 16, dst, dststride, src, srcstride,
; int height, int mx, int my) ; int height, int mx, int my)
; ****************************** ; ******************************
%macro HEVC_PUT_HEVC_QPEL_HV 2 %macro HEVC_PUT_HEVC_QPEL_HV 2
cglobal hevc_put_hevc_qpel_hv%1_%2, 6, 8, 12, dst, src, srcstride, height, mx, my, r3src, rfilter cglobal hevc_put_hevc_qpel_hv%1_%2, 6, 8, 16, dst, src, srcstride, height, mx, my, r3src, rfilter
lea mxq, [mxq*8-8] lea mxq, [mxq*8-8]
lea myq, [myq*8-8] lea myq, [myq*8-8]
lea r3srcq, [srcstrideq*3] lea r3srcq, [srcstrideq*3]
@ -993,7 +993,7 @@ cglobal hevc_put_hevc_qpel_hv%1_%2, 6, 8, 12, dst, src, srcstride, height, mx, m
LOOP_END dst, src, srcstride LOOP_END dst, src, srcstride
RET RET
cglobal hevc_put_hevc_uni_qpel_hv%1_%2, 7, 9, 12 , dst, dststride, src, srcstride, height, mx, my, r3src, rfilter cglobal hevc_put_hevc_uni_qpel_hv%1_%2, 7, 9, 16 , dst, dststride, src, srcstride, height, mx, my, r3src, rfilter
lea mxq, [mxq*8-8] lea mxq, [mxq*8-8]
lea myq, [myq*8-8] lea myq, [myq*8-8]
lea r3srcq, [srcstrideq*3] lea r3srcq, [srcstrideq*3]