har0ke/FFmpeg
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

ARM: simplify ff_put/avg_h264_chroma_mc4/8_neon definitions, no code change

Browse Source 
Originally committed as revision 16677 to svn://svn.ffmpeg.org/ffmpeg/trunk
This commit is contained in:
Måns Rullgård 2009-01-18 20:43:11 +00:00
parent c0a02a3e47
commit 77c4537313
1 changed files with 18 additions and 25 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

43
libavcodec/arm/h264dsp_neon.S
View File

@ -56,10 +56,11 @@
.endm .endm
/* chroma_mc8(uint8_t *dst, uint8_t *src, int stride, int h, int x, int y) */ /* chroma_mc8(uint8_t *dst, uint8_t *src, int stride, int h, int x, int y) */
.macro h264_chroma_mc8 avg=0 .macro h264_chroma_mc8 type
function ff_\type\()_h264_chroma_mc8_neon, export=1
push {r4-r7, lr} push {r4-r7, lr}
ldrd r4, [sp, #20] ldrd r4, [sp, #20]
.if \avg .ifc \type,avg
mov lr, r0 mov lr, r0
.endif .endif
pld [r1] pld [r1]
@ -103,7 +104,7 @@
vld1.64 {d6, d7}, [r5], r4 vld1.64 {d6, d7}, [r5], r4
pld [r1] pld [r1]
vrshrn.u16 d17, q9, #6 vrshrn.u16 d17, q9, #6
.if \avg .ifc \type,avg
vld1.64 {d20}, [lr,:64], r2 vld1.64 {d20}, [lr,:64], r2
vld1.64 {d21}, [lr,:64], r2 vld1.64 {d21}, [lr,:64], r2
vrhadd.u8 q8, q8, q10 vrhadd.u8 q8, q8, q10
@ -136,7 +137,7 @@
vld1.64 {d6}, [r5], r4 vld1.64 {d6}, [r5], r4
vrshrn.u16 d16, q8, #6 vrshrn.u16 d16, q8, #6
vrshrn.u16 d17, q9, #6 vrshrn.u16 d17, q9, #6
.if \avg .ifc \type,avg
vld1.64 {d20}, [lr,:64], r2 vld1.64 {d20}, [lr,:64], r2
vld1.64 {d21}, [lr,:64], r2 vld1.64 {d21}, [lr,:64], r2
vrhadd.u8 q8, q8, q10 vrhadd.u8 q8, q8, q10
@ -165,7 +166,7 @@
vext.8 d5, d4, d5, #1 vext.8 d5, d4, d5, #1
vrshrn.u16 d16, q8, #6 vrshrn.u16 d16, q8, #6
vrshrn.u16 d17, q9, #6 vrshrn.u16 d17, q9, #6
.if \avg .ifc \type,avg
vld1.64 {d20}, [lr,:64], r2 vld1.64 {d20}, [lr,:64], r2
vld1.64 {d21}, [lr,:64], r2 vld1.64 {d21}, [lr,:64], r2
vrhadd.u8 q8, q8, q10 vrhadd.u8 q8, q8, q10
@ -177,13 +178,15 @@
bgt 5b bgt 5b
pop {r4-r7, pc} pop {r4-r7, pc}
.endfunc
.endm .endm
/* chroma_mc4(uint8_t *dst, uint8_t *src, int stride, int h, int x, int y) */ /* chroma_mc4(uint8_t *dst, uint8_t *src, int stride, int h, int x, int y) */
.macro h264_chroma_mc4 avg=0 .macro h264_chroma_mc4 type
function ff_\type\()_h264_chroma_mc4_neon, export=1
push {r4-r7, lr} push {r4-r7, lr}
ldrd r4, [sp, #20] ldrd r4, [sp, #20]
.if \avg .ifc \type,avg
mov lr, r0 mov lr, r0
.endif .endif
pld [r1] pld [r1]
@ -230,7 +233,7 @@
vrshrn.u16 d16, q8, #6 vrshrn.u16 d16, q8, #6
subs r3, r3, #2 subs r3, r3, #2
pld [r1] pld [r1]
.if \avg .ifc \type,avg
vld1.32 {d20[0]}, [lr,:32], r2 vld1.32 {d20[0]}, [lr,:32], r2
vld1.32 {d20[1]}, [lr,:32], r2 vld1.32 {d20[1]}, [lr,:32], r2
vrhadd.u8 d16, d16, d20 vrhadd.u8 d16, d16, d20
@ -265,7 +268,7 @@
vadd.i16 d16, d16, d17 vadd.i16 d16, d16, d17
vadd.i16 d17, d18, d19 vadd.i16 d17, d18, d19
vrshrn.u16 d16, q8, #6 vrshrn.u16 d16, q8, #6
.if \avg .ifc \type,avg
vld1.32 {d20[0]}, [lr,:32], r2 vld1.32 {d20[0]}, [lr,:32], r2
vld1.32 {d20[1]}, [lr,:32], r2 vld1.32 {d20[1]}, [lr,:32], r2
vrhadd.u8 d16, d16, d20 vrhadd.u8 d16, d16, d20
@ -295,7 +298,7 @@
vadd.i16 d17, d18, d19 vadd.i16 d17, d18, d19
pld [r1] pld [r1]
vrshrn.u16 d16, q8, #6 vrshrn.u16 d16, q8, #6
.if \avg .ifc \type,avg
vld1.32 {d20[0]}, [lr,:32], r2 vld1.32 {d20[0]}, [lr,:32], r2
vld1.32 {d20[1]}, [lr,:32], r2 vld1.32 {d20[1]}, [lr,:32], r2
vrhadd.u8 d16, d16, d20 vrhadd.u8 d16, d16, d20
@ -309,26 +312,16 @@
bgt 5b bgt 5b
pop {r4-r7, pc} pop {r4-r7, pc}
.endfunc
.endm .endm
.text .text
.align .align
function ff_put_h264_chroma_mc8_neon, export=1 h264_chroma_mc8 put
h264_chroma_mc8 h264_chroma_mc8 avg
.endfunc h264_chroma_mc4 put
h264_chroma_mc4 avg
function ff_avg_h264_chroma_mc8_neon, export=1
h264_chroma_mc8 avg=1
.endfunc
function ff_put_h264_chroma_mc4_neon, export=1
h264_chroma_mc4
.endfunc
function ff_avg_h264_chroma_mc4_neon, export=1
h264_chroma_mc4 avg=1
.endfunc
/* H.264 loop filter */ /* H.264 loop filter */