| Commit message (Expand) | Author | Age | Files | Lines |
* | sbcenc: add armv6 and neon asm optimizations | Aurelien Jacobs | 2018-03-07 | 4 | -0/+1067 |
* | avcodec/arm/sbrdsp_neon: Use a free register instead of putting 2 things in one | Michael Niedermayer | 2018-01-12 | 1 | -2/+2 |
* | arm/hevc_idct: fix compilation on Android | James Almer | 2017-12-09 | 1 | -59/+60 |
* | Merge commit 'b487add7ecf78efda36d49815f8f8757bd24d4cb' | James Almer | 2017-11-11 | 1 | -4/+2 |
|\ |
|
| * | arm: Remove a redundant check in fmtconvert_init_arm.c | Martin Storsjö | 2017-10-24 | 1 | -4/+2 |
* | | Merge commit '9dde6ab06c48f9447cd16f39bee33569cddb7be4' | James Almer | 2017-11-11 | 1 | -8/+12 |
|\| |
|
| * | arm: Fix SIGBUS on ARM when compiled with binutils 2.29 | Martin Storsjö | 2017-09-02 | 1 | -8/+12 |
* | | Merge commit 'd7320ca3ed10f0d35b3740fa03341161e74275ea' | James Almer | 2017-10-30 | 2 | -20/+5 |
|\| |
|
| * | arm: Avoid using .dn register aliases | Martin Storsjö | 2017-05-15 | 2 | -20/+5 |
* | | Merge commit 'ce080f47b8b55ab3d41eb00487b138d9906d114d' | James Almer | 2017-10-30 | 2 | -21/+294 |
|\| |
|
| * | hevc: Add NEON 32x32 IDCT | Alexandra Hájková | 2017-05-04 | 2 | -21/+294 |
* | | Merge commit '118dd4a321a2d67f67c21b076abd0b4d939ab642' | James Almer | 2017-10-30 | 1 | -8/+8 |
|\| |
|
| * | hevc: 16x16 NEON idct: Use the right element size for loads/stores | Alexandra Hájková | 2017-05-04 | 1 | -8/+8 |
* | | Merge commit 'edbf0fffb15dde7a1de70b05855529d5fc769f14' | James Almer | 2017-10-30 | 2 | -0/+102 |
|\| |
|
| * | hevc: Add NEON add_residual for bitdepth 10 | Alexandra Hájková | 2017-05-01 | 2 | -0/+102 |
* | | Merge commit 'e1c2453a4fac1f7116244d0d05310935c20887e6' | James Almer | 2017-10-30 | 1 | -16/+35 |
|\| |
|
| * | arm: hevc_idct: Tune the add_res_8x8 and add_res_32x32 functions | Martin Storsjö | 2017-04-28 | 1 | -16/+35 |
* | | Merge commit '0d4d43513786f1df4d561e1fac924fb0722c6700' | James Almer | 2017-10-30 | 2 | -75/+92 |
|\| |
|
| * | hevc: Add NEON add_residual for bitdepth 8 | Seppo Tomperi | 2017-04-27 | 2 | -0/+103 |
* | | Merge commit '3d69dd65c6771c28d3bf4e8e53a905aa8cd01fd9' | James Almer | 2017-10-30 | 2 | -12/+37 |
|\| |
|
| * | hevc: Add support for bitdepth 10 for IDCT DC | Alexandra Hájková | 2017-04-25 | 2 | -12/+37 |
* | | Merge commit '358adef0305618219522858e471edf7e0cb4043e' | James Almer | 2017-10-30 | 2 | -84/+84 |
|\| |
|
| * | hevc: Add NEON IDCT DC functions for bitdepth 8 | Seppo Tomperi | 2017-04-25 | 2 | -0/+88 |
* | | Merge commit '89d9869d2491d4209d707a8e7f29c58227ae5a4e' | James Almer | 2017-10-27 | 2 | -0/+201 |
|\| |
|
| * | hevc: Add NEON 16x16 IDCT | Alexandra Hájková | 2017-04-12 | 2 | -0/+201 |
| * | arm: Always build the hevcdsp_init_arm.c file | Martin Storsjö | 2017-03-28 | 1 | -2/+2 |
* | | Merge commit '0b9a237b2386ff84a6f99716bd58fa27a1b767e7' | James Almer | 2017-10-24 | 4 | -238/+217 |
|\| |
|
| * | hevc: Add NEON 4x4 and 8x8 IDCT | Alexandra Hájková | 2017-03-27 | 3 | -0/+277 |
| * | arm/aarch64: vp9: Fix vertical alignment | Martin Storsjö | 2017-03-16 | 2 | -8/+8 |
| * | arm/aarch64: vp9itxfm: Skip loading the min_eob pointer when it won't be used | Martin Storsjö | 2017-03-11 | 1 | -2/+2 |
| * | arm: vp9itxfm: Template the quarter/half idct32 function | Martin Storsjö | 2017-03-11 | 1 | -37/+20 |
| * | arm: vp9itxfm: Reorder iadst16 coeffs | Martin Storsjö | 2017-02-24 | 1 | -6/+6 |
| * | arm: vp9itxfm: Reorder the idct coefficients for better pairing | Martin Storsjö | 2017-02-24 | 1 | -62/+62 |
| * | arm: vp9itxfm: Avoid reloading the idct32 coefficients | Martin Storsjö | 2017-02-24 | 1 | -126/+120 |
| * | arm: vp9lpf: Implement the mix2_44 function with one single filter pass | Martin Storsjö | 2017-02-24 | 2 | -3/+195 |
| * | arm/aarch64: vp9lpf: Keep the comparison to E within 8 bit | Martin Storsjö | 2017-02-24 | 1 | -6/+5 |
| * | arm: vp9lpf: Interleave the start of flat8in into the calculation above | Martin Storsjö | 2017-02-11 | 1 | -2/+6 |
| * | arm: vp9lpf: Use orrs instead of orr+cmp | Martin Storsjö | 2017-02-11 | 1 | -8/+4 |
| * | arm/aarch64: vp9lpf: Calculate !hev directly | Martin Storsjö | 2017-02-11 | 1 | -3/+2 |
| * | arm: vp9itxfm: Optimize 16x16 and 32x32 idct dc by unrolling | Martin Storsjö | 2017-02-11 | 1 | -18/+36 |
| * | arm: vp9mc: Calculate less unused data in the 4 pixel wide horizontal filter | Martin Storsjö | 2017-02-11 | 1 | -11/+22 |
| * | arm: vp9itxfm: Share instructions for loading idct coeffs in the 8x8 function | Martin Storsjö | 2017-02-09 | 1 | -2/+1 |
| * | arm: vp9itxfm: Do a simpler half/quarter idct16/idct32 when possible | Martin Storsjö | 2017-02-09 | 1 | -54/+537 |
| * | arm: vp9itxfm: Move the load_add_store macro out from the itxfm16 pass2 function | Martin Storsjö | 2017-02-09 | 1 | -36/+36 |
| * | arm: vp9itxfm: Make the larger core transforms standalone functions | Martin Storsjö | 2017-02-09 | 1 | -17/+26 |
| * | arm: vp9itxfm: Avoid .irp when it doesn't save any lines | Martin Storsjö | 2017-02-05 | 1 | -12/+12 |
| * | arm: vp9mc: Fix vertical alignment of operands | Martin Storsjö | 2017-01-03 | 1 | -3/+3 |
| * | arm: vp9itxfm: Skip empty slices in the first pass of idct_idct 16x16 and 32x32 | Martin Storsjö | 2016-11-30 | 1 | -9/+66 |
| * | arm: vp9itxfm: Only reload the idct coeffs for the iadst_idct combination | Martin Storsjö | 2016-11-30 | 1 | -1/+1 |
| * | arm: vp9itxfm: Rename a macro parameter to fit better | Martin Storsjö | 2016-11-23 | 1 | -7/+7 |