aboutsummaryrefslogtreecommitdiffstats
path: root/libavcodec/x86
diff options
context:
space:
mode:
authorLoren Merritt <lorenm@u.washington.edu>2011-06-01 01:01:01 -0400
committerRonald S. Bultje <rsbultje@gmail.com>2011-06-02 07:07:15 -0700
commit53be7b23e9d7074d1aeee77407b008411d034e9e (patch)
tree8d1b49c9aa1c3e8a78328788ba16b3f68c0203df /libavcodec/x86
parent994c3550ffe032385833c21876fb121f59516079 (diff)
downloadffmpeg-53be7b23e9d7074d1aeee77407b008411d034e9e.tar.gz
Cosmetic changes to h264_idct_10bit.asm.
Removes redundant dword tags and whitespace changes. Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
Diffstat (limited to 'libavcodec/x86')
-rw-r--r--libavcodec/x86/h264_idct_10bit.asm139
1 files changed, 71 insertions, 68 deletions
diff --git a/libavcodec/x86/h264_idct_10bit.asm b/libavcodec/x86/h264_idct_10bit.asm
index 64089001e5..3f7cf4cefc 100644
--- a/libavcodec/x86/h264_idct_10bit.asm
+++ b/libavcodec/x86/h264_idct_10bit.asm
@@ -133,7 +133,7 @@ ADD4x4IDCT avx
%macro ADD16_OP 3
cmp byte [r4+%3], 0
jz .skipblock%2
- mov r5d, dword [r1+%2*4]
+ mov r5d, [r1+%2*4]
call add4x4_idct_%1
.skipblock%2:
%if %2<15
@@ -159,7 +159,7 @@ cglobal h264_idct_add16_10_%1, 5,6
ADD16_OP %1, 13, 7+3*8
ADD16_OP %1, 14, 6+4*8
ADD16_OP %1, 15, 7+4*8
- RET
+ REP_RET
%endmacro
INIT_XMM
@@ -201,7 +201,7 @@ IDCT_ADD16_10 avx
INIT_MMX
cglobal h264_idct_dc_add_10_mmx2,3,3
- movd m0, dword [r1]
+ movd m0, [r1]
paddd m0, [pd_32]
psrad m0, 6
lea r1, [r2*3]
@@ -215,7 +215,7 @@ cglobal h264_idct_dc_add_10_mmx2,3,3
;-----------------------------------------------------------------------------
%macro IDCT8_DC_ADD 1
cglobal h264_idct8_dc_add_10_%1,3,3,7
- mov r1d, dword [r1]
+ mov r1d, [r1]
add r1, 32
sar r1, 6
movd m0, r1d
@@ -240,9 +240,9 @@ IDCT8_DC_ADD avx
;-----------------------------------------------------------------------------
%macro AC 2
.ac%2
- mov r5d, dword [r1+(%2+0)*4]
+ mov r5d, [r1+(%2+0)*4]
call add4x4_idct_%1
- mov r5d, dword [r1+(%2+1)*4]
+ mov r5d, [r1+(%2+1)*4]
add r2, 64
call add4x4_idct_%1
add r2, 64
@@ -251,16 +251,16 @@ IDCT8_DC_ADD avx
%assign last_block 16
%macro ADD16_OP_INTRA 3
- cmp word [r4+%3], 0
+ cmp word [r4+%3], 0
jnz .ac%2
- mov r5d, dword [r2+ 0]
- or r5d, dword [r2+64]
+ mov r5d, [r2+ 0]
+ or r5d, [r2+64]
jz .skipblock%2
- mov r5d, dword [r1+(%2+0)*4]
+ mov r5d, [r1+(%2+0)*4]
call idct_dc_add_%1
.skipblock%2:
%if %2<last_block-2
- add r2, 128
+ add r2, 128
%endif
.skipadd%2:
%endmacro
@@ -288,12 +288,15 @@ cglobal h264_idct_add16intra_10_%1,5,7,8
ADD16_OP_INTRA %1, 10, 4+4*8
ADD16_OP_INTRA %1, 12, 6+3*8
ADD16_OP_INTRA %1, 14, 6+4*8
- RET
-%assign i 14
-%rep 8
- AC %1, i
-%assign i i-2
-%endrep
+ REP_RET
+ AC %1, 8
+ AC %1, 10
+ AC %1, 12
+ AC %1, 14
+ AC %1, 0
+ AC %1, 2
+ AC %1, 4
+ AC %1, 6
%endmacro
INIT_XMM
@@ -312,15 +315,15 @@ cglobal h264_idct_add8_10_%1,5,7
%ifdef ARCH_X86_64
mov r10, r0
%endif
- add r2, 1024
- mov r0, [r0]
+ add r2, 1024
+ mov r0, [r0]
ADD16_OP_INTRA %1, 16, 1+1*8
ADD16_OP_INTRA %1, 18, 1+2*8
%ifdef ARCH_X86_64
- mov r0, [r10+gprsize]
+ mov r0, [r10+gprsize]
%else
- mov r0, r0m
- mov r0, [r0+gprsize]
+ mov r0, r0m
+ mov r0, [r0+gprsize]
%endif
ADD16_OP_INTRA %1, 20, 1+4*8
ADD16_OP_INTRA %1, 22, 1+5*8
@@ -343,51 +346,51 @@ IDCT_ADD8 avx
; void h264_idct8_add(pixel *dst, dctcoef *block, int stride)
;-----------------------------------------------------------------------------
%macro IDCT8_1D 2
- SWAP 0, 1
- psrad m4, m5, 1
- psrad m1, m0, 1
- paddd m4, m5
- paddd m1, m0
- paddd m4, m7
- paddd m1, m5
- psubd m4, m0
- paddd m1, m3
-
- psubd m0, m3
- psubd m5, m3
- paddd m0, m7
- psubd m5, m7
- psrad m3, 1
- psrad m7, 1
- psubd m0, m3
- psubd m5, m7
-
- SWAP 1, 7
- psrad m1, m7, 2
- psrad m3, m4, 2
- paddd m3, m0
- psrad m0, 2
- paddd m1, m5
- psrad m5, 2
- psubd m0, m4
- psubd m7, m5
-
- SWAP 5, 6
- psrad m4, m2, 1
- psrad m6, m5, 1
- psubd m4, m5
- paddd m6, m2
-
- mova m2, %1
- mova m5, %2
- SUMSUB_BA d, 5, 2
- SUMSUB_BA d, 6, 5
- SUMSUB_BA d, 4, 2
- SUMSUB_BA d, 7, 6
- SUMSUB_BA d, 0, 4
- SUMSUB_BA d, 3, 2
- SUMSUB_BA d, 1, 5
- SWAP 7, 6, 4, 5, 2, 3, 1, 0 ; 70315246 -> 01234567
+ SWAP 0, 1
+ psrad m4, m5, 1
+ psrad m1, m0, 1
+ paddd m4, m5
+ paddd m1, m0
+ paddd m4, m7
+ paddd m1, m5
+ psubd m4, m0
+ paddd m1, m3
+
+ psubd m0, m3
+ psubd m5, m3
+ paddd m0, m7
+ psubd m5, m7
+ psrad m3, 1
+ psrad m7, 1
+ psubd m0, m3
+ psubd m5, m7
+
+ SWAP 1, 7
+ psrad m1, m7, 2
+ psrad m3, m4, 2
+ paddd m3, m0
+ psrad m0, 2
+ paddd m1, m5
+ psrad m5, 2
+ psubd m0, m4
+ psubd m7, m5
+
+ SWAP 5, 6
+ psrad m4, m2, 1
+ psrad m6, m5, 1
+ psubd m4, m5
+ paddd m6, m2
+
+ mova m2, %1
+ mova m5, %2
+ SUMSUB_BA d, 5, 2
+ SUMSUB_BA d, 6, 5
+ SUMSUB_BA d, 4, 2
+ SUMSUB_BA d, 7, 6
+ SUMSUB_BA d, 0, 4
+ SUMSUB_BA d, 3, 2
+ SUMSUB_BA d, 1, 5
+ SWAP 7, 6, 4, 5, 2, 3, 1, 0 ; 70315246 -> 01234567
%endmacro
%macro IDCT8_1D_FULL 1
@@ -523,7 +526,7 @@ IDCT8_ADD avx
%macro IDCT8_ADD4_OP 3
cmp byte [r4+%3], 0
jz .skipblock%2
- mov r0d, dword [r6+%2*4]
+ mov r0d, [r6+%2*4]
add r0, r5
call h264_idct8_add1_10_%1
.skipblock%2: