Commits · 5d972bf3f220404f7e7bd7595f8c3a804191b35c · VideoLAN / x264

22 Mar, 2008 1 commit
- special case dequant for flat matrix · 5d972bf3
  Loren Merritt authored Mar 21, 2008
  
  5d972bf3
21 Mar, 2008 3 commits
- faster dequant · f63770aa
  Loren Merritt authored Mar 21, 2008
  
  f63770aa
- simplify hpel_filter_c · 263abc67
  Loren Merritt authored Mar 20, 2008
  
  263abc67
- use x264_mc_copy_w16_sse2 in mc.copy, it was previously only in mc_luma · ead697ca
  Loren Merritt authored Mar 20, 2008
  
  ead697ca
20 Mar, 2008 7 commits
- new ssd_8x*_sse2 · 14b45a81
  Loren Merritt authored Mar 20, 2008
```
align ssd_16x*_sse2
unroll ssd_4x*_mmx
```
  14b45a81
- update altivec zigzags · 72869f76
  Manuel Rommel authored Mar 20, 2008
  
  72869f76
- r768 borked cavlc · dbfd2cc7
  Loren Merritt authored Mar 20, 2008
  
  dbfd2cc7
- cosmetics in intra predict · ac49761a
  Loren Merritt authored Mar 20, 2008
  
  ac49761a
- faster intra predict 8x8 hu/hd · 5442dafd
  Fiona Glaser authored Mar 20, 2008
  
  5442dafd
- reduce zigzag arrays from int to int16_t · 30da25a9
  Loren Merritt authored Mar 19, 2008
  
  30da25a9
- reduce the size of some arrays · 7a125e4a
  Loren Merritt authored Mar 19, 2008
  
  7a125e4a
19 Mar, 2008 6 commits
- skip intra pred+dct+quant in cases where it's redundant (analyse vs encode) · 1d56ef44
  Fiona Glaser authored Mar 19, 2008
```
large speedup with trellis=2, small speedup with trellis=0 and/or subme>=6
```
  1d56ef44
- cosmetics in asm · b10ee560
  Loren Merritt authored Mar 19, 2008
  
  b10ee560
- satd_4x4_ssse3 · a3b97adb
  Fiona Glaser authored Mar 19, 2008
  
  a3b97adb
- get_ref_sse2 · 87739884
  Fiona Glaser authored Mar 19, 2008
  
  87739884
- continue instead of crash when the threading mv constraint is violated. · 8727a01b
  Fiona Glaser authored Mar 18, 2008
```
doesn't fix the underlying bug, but hopefully less annoying until we find it.
```
  8727a01b
- remove remaining reference to clip1.h · b2d5df5f
  Loren Merritt authored Mar 18, 2008
  
  b2d5df5f
18 Mar, 2008 2 commits
- fix name mangling again. · 73b3fd48
  Loren Merritt authored Mar 18, 2008
```
apparently it's not just a convention, dll build fails if you try to export a non-prefixed name.
```
  73b3fd48
- update msvc projectfile · 1e829dbf
  Gabriel Bouvigne authored Mar 17, 2008
  
  1e829dbf
17 Mar, 2008 7 commits
- missing #ifdef HAVE_SSE3 · 08c7dd46
  Loren Merritt authored Mar 17, 2008
  
  08c7dd46
- don't define offsetof since it's standard · 0f519339
  Loren Merritt authored Mar 17, 2008
  
  0f519339
- shut up gcc warning in offsetof · cfa08dc7
  Loren Merritt authored Mar 17, 2008
  
  cfa08dc7
- increase alignment of mv arrays · e56ea086
  Håkan Hjort authored Mar 17, 2008
  
  e56ea086
- memcpy_aligned_sse2 · 5469a4ba
  Fiona Glaser authored Mar 16, 2008
  
  5469a4ba
- checkasm check whether callee-saved regs are correctly saved · 9d0c0a90
  Loren Merritt authored Mar 16, 2008
```
x86_32 only for now since x86_64 varargs are annoying
```
  9d0c0a90
- fix x86_32 ads which failed to preserve a register · 1b1e1248
  Loren Merritt authored Mar 16, 2008
  
  1b1e1248
16 Mar, 2008 3 commits
- fix some name mangling issues introduced by the merge · c82674fe
  Loren Merritt authored Mar 16, 2008
  
  c82674fe
- remove x264_mc_clip1. · 20b4106b
  Loren Merritt authored Mar 16, 2008
```
it's wrong for sufficiently perverse inputs, and clip_uint8 is faster anyway.
```
  20b4106b
- merge x86_32 and x86_64 asm, with macros to abstract calling convention and register names · c17218e8
  Loren Merritt authored Mar 16, 2008
  
  c17218e8
10 Mar, 2008 1 commit
- git compatible version script · 3445cca4
  Loren Merritt authored Mar 09, 2008
  
  3445cca4
03 Mar, 2008 2 commits
- check for broken versions of yasm · 8609ffa0
  Loren Merritt authored Mar 02, 2008
  
  8609ffa0
- increase the alignment of the i8x8 edge cache, needed for sse2 intra prediction. · 3d5beaee
  Loren Merritt authored Mar 02, 2008
```
patch by Alexander Strange.
```
  3d5beaee
02 Mar, 2008 5 commits
- .gitignore · 25fd257d
  Loren Merritt authored Mar 02, 2008
  
  25fd257d
- pic macros now keep track of which register holds the GOT, so variable access doesn't have to care · 9dce08ac
  Loren Merritt authored Mar 02, 2008
```
git-svn-id: svn://svn.videolan.org/x264/trunk@745 df754926-b1dd-0310-bc7b-ec298dee348c
```
  9dce08ac
- remove x86_64 predict_8x8_ddl_mmxext because sse2 is faster even on amd · 264f13ae
  Loren Merritt authored Mar 02, 2008
```
git-svn-id: svn://svn.videolan.org/x264/trunk@744 df754926-b1dd-0310-bc7b-ec298dee348c
```
  264f13ae
- cosmetics in dsp init · 31528574
  Loren Merritt authored Mar 02, 2008
```
git-svn-id: svn://svn.videolan.org/x264/trunk@743 df754926-b1dd-0310-bc7b-ec298dee348c
```
  31528574
- sse2 16x16 intra pred. · 564cc252
  Loren Merritt authored Mar 02, 2008
```
port the remaining intra pred functions from x86_64 to x86_32.
patch by Fiona Glaser.


git-svn-id: svn://svn.videolan.org/x264/trunk@742 df754926-b1dd-0310-bc7b-ec298dee348c
```
  564cc252
01 Mar, 2008 3 commits

some simplifications to mmx intra pred that should have been done way back... · c48882dd

Loren Merritt authored Mar 01, 2008

some simplifications to mmx intra pred that should have been done way back when we switched to constant fdec_stride.
and remove pic spills in functions that have a free caller-saved reg.
patch partly by Fiona Glaser.

git-svn-id: svn://svn.videolan.org/x264/trunk@741 df754926-b1dd-0310-bc7b-ec298dee348c

c48882dd

faster array_non_zero · 78ec7879

Loren Merritt authored Mar 01, 2008


git-svn-id: svn://svn.videolan.org/x264/trunk@740 df754926-b1dd-0310-bc7b-ec298dee348c

78ec7879

x86_32 sse2 idct8 · 405d9c66

Loren Merritt authored Mar 01, 2008

ported from ffmpeg by Fiona Glaser


git-svn-id: svn://svn.videolan.org/x264/trunk@739 df754926-b1dd-0310-bc7b-ec298dee348c

405d9c66