shithub: dav1d

2eee4711 – Henrik Gramner <[email protected]> authored on 2018/12/17 17:48

Fix AVX2 itx crash on 64-bit Windows

08b4c9af – Henrik Gramner <[email protected]> authored on 2018/12/17 12:52

Avoid a redundant clip in 8-bit itx

a755b6e3 – Henrik Gramner <[email protected]> authored on 2018/12/15 14:02

Clip coefficients in SSSE3/AVX2 inverse transform asm

eb01bdb9 – Henrik Gramner <[email protected]> authored on 2018/12/15 14:01

Improve the fast path of IDCT 8x32 AVX2

b53a99b9 – Janne Grunau <[email protected]> authored on 2018/12/16 07:27

arm: define PIC if the compiler uses pic by default

7677c120 – Janne Grunau <[email protected]> authored on 2018/12/15 10:30

intrabc: use visible width/height in mv correction

6a95aa56 – Jean-Baptiste Kempf <[email protected]> authored on 2018/12/15 07:53

Update the THANKS file

dd961b1e – Jean-Baptiste Kempf <[email protected]> authored on 2018/12/15 07:47

Update slightly the contribution document.

6b3770ae – Jean-Baptiste Kempf <[email protected]> authored on 2018/12/15 07:43

Various update to the README, post 0.1.0 release

0c173fd1 – Jean-Baptiste Kempf <[email protected]> authored on 2018/12/15 07:29

On the road to 0.1.1

6a10a981 – Ronald S. Bultje <[email protected]> authored on 2018/12/14 10:38

Rewrite inverse transforms to prevent integer overflows

12dc2be0 – Michael Bradshaw <[email protected]> authored and Jean-Baptiste Kempf <[email protected]> committed on 2018/12/12 21:06

Document the thread context in which Dav1dPicAllocator's functions are called

41874f89 – Fred Barbier <[email protected]> authored and Ronald S. Bultje <[email protected]> committed on 2018/12/12 09:52

Allow redundant sequence header

e5bca59c – Janne Grunau <[email protected]> authored on 2018/12/12 18:07

itx: cancel 16 out of '(x * 2896) >> 12' to avoid integer overflows

1d3f6364 – Janne Grunau <[email protected]> authored on 2018/12/12 16:49

inv_identity: avoid signed int overflow for 12-bit files

c3ee7290 – David Michael Barr <[email protected]> authored and Ronald S. Bultje <[email protected]> committed on 2018/12/12 14:33

x86: Add chroma-from-luma AC 4:2:2 AVX2 asm

ec734abd – Timo Gurr <[email protected]> authored and Janne Grunau <[email protected]> committed on 2018/12/11 13:32

meson: fix installing the includes

73067e5c – Janne Grunau <[email protected]> authored on 2018/12/11 16:27

allocation fail fuzzer: use a static library and objcopy

a6b903fb – Jean-Baptiste Kempf <[email protected]> authored on 2018/12/11 10:14

Update NEWS

a3217134 – Janne Grunau <[email protected]> authored on 2018/12/11 08:48

do not build the allocation fail fuzzer when lto is enabled

c0501f1f – Jean-Baptiste Kempf <[email protected]> authored on 2018/12/10 17:34

On the way to 0.1.0

8bc73da3 – Janne Grunau <[email protected]> authored on 2018/12/08 10:09

fuzzing: add memory allocation fail fuzzer

0636633f – Francois Cartegnie <[email protected]> authored on 2018/11/30 16:13

add SSSE3 blend_h

fef13fd6 – Francois Cartegnie <[email protected]> authored on 2018/11/30 10:19

add SSSE3 blend_v

7606bb08 – Francois Cartegnie <[email protected]> authored on 2018/11/30 09:07

add SSSE3 blend

b3da18fe – Francois Cartegnie <[email protected]> authored on 2018/11/29 09:30

add SSSE3 w_mask_420

00d1f4d5 – Janne Grunau <[email protected]> authored on 2018/12/10 03:47

cdf: fix memleak in dav1d_cdf_thread_unref

0f7a97ba – Ronald S. Bultje <[email protected]> authored on 2018/12/09 04:52

Make static CDF tables source directly from rodata

6bc4acf4 – Janne Grunau <[email protected]> authored on 2018/12/09 06:21

svc: move spatial layer check out of output_picture

32908aae – James Almer <[email protected]> authored on 2018/12/08 11:57

annexb: don't use a int64_t intermediary for leb128

a0fb010f – Janne Grunau <[email protected]> authored on 2018/12/08 14:13

picture: free pic_ctx on dav1d_ref_wrap error

488374c7 – Janne Grunau <[email protected]> authored on 2018/12/08 14:09

output_picture: unref pictures on allocation errors

3746b6e1 – Janne Grunau <[email protected]> authored on 2018/12/08 14:03

memory allocation error checks

a5afcbe6 – Janne Grunau <[email protected]> authored on 2018/12/08 13:40

decode_frame: use goto error after allocation errors

c49680e6 – Janne Grunau <[email protected]> authored on 2018/12/06 16:22

checkasm: replace rand() with xorshift xor128

c82cf102 – Janne Grunau <[email protected]> authored on 2018/12/08 10:32

fix mvs_ref NULL ptr dererence after allocation error

52b575c8 – Janne Grunau <[email protected]> authored on 2018/12/08 10:28

cdf: make CdfThreadContext allocation failure safe

e72bec43 – Janne Grunau <[email protected]> authored on 2018/12/08 09:10

dav1d_open: free context completely on allocation errors

ceb6175d – Janne Grunau <[email protected]> authored and Henrik Gramner <[email protected]> committed on 2018/12/06 17:40

CI: set empty dependecies for the tests jobs

f72b1fea – James Almer <[email protected]> authored on 2018/12/07 12:04

annexb: ensure that OBU size <= frame size <= TU size

7b454ebb – James Almer <[email protected]> authored on 2018/12/07 10:52

Add Length delimited bitstream (Annex B) demuxer

0b6ee30e – Henrik Gramner <[email protected]> authored on 2018/12/05 17:49

checkasm: Try to handle crashes more gracefully

02312cae – Ronald S. Bultje <[email protected]> authored on 2018/12/07 05:38

Implement --cpumask to mask out CPU features

d3a1ebad – Ronald S. Bultje <[email protected]> authored on 2018/12/06 05:22

Fix mc.avg/w_avg/mask for x86-32

da5a5df8 – Ronald S. Bultje <[email protected]> authored on 2018/12/06 02:18

Special w=4/8 cases

87a377e9 – Liwei Wang <[email protected]> authored and Henrik Gramner <[email protected]> committed on 2018/12/04 11:23

Add SSSE3 implementation for the 4x4 blocks in itx

6f2f0188 – Xuefeng Jiang <[email protected]> authored and Ronald S. Bultje <[email protected]> committed on 2018/12/04 10:28

Add SSSE3 implementation for dav1d_ipred_h

c3980e39 – Ronald S. Bultje <[email protected]> authored on 2018/12/05 13:21

12 bits/component support

2e6c8a92 – Ronald S. Bultje <[email protected]> authored on 2018/11/30 08:55

Change type of MC intermediates from coef to int16_t

012fced2 – Ronald S. Bultje <[email protected]> authored on 2018/11/29 14:32

Fix 12-bit support in yuv4mpeg muxer

95cf7cf8 – Ronald S. Bultje <[email protected]> authored on 2018/12/05 12:34

Use adapted CDF for invisible keyframes if refresh_context=1

d6770f93 – Henrik Gramner <[email protected]> authored and Ronald S. Bultje <[email protected]> committed on 2018/12/03 15:42

Add ipred_z1 AVX2 asm

f816d5cf – Henrik Gramner <[email protected]> authored and Ronald S. Bultje <[email protected]> committed on 2018/12/03 15:41

Make some minor ipred_z simplifications

7718c464 – Janne Grunau <[email protected]> authored on 2018/12/04 18:45

frame-mt: drain delayed pictures fully before returning -EAGAIN

b735b80b – Janne Grunau <[email protected]> authored on 2018/12/04 14:43

frame-mt: output delayed frames as soon as possible

18ab22b6 – Henrik Gramner <[email protected]> authored on 2018/12/04 11:06

Correctly use movd instead of movq in blend_h_avx2.w4

7a222d46 – Liwei Wang <[email protected]> authored and Ronald S. Bultje <[email protected]> committed on 2018/11/19 10:33

Add SSSE3 implementation for dav1d_inv_txfm_add_dct_dct_4x4

1e9c428a – Janne Grunau <[email protected]> authored on 2018/12/03 17:50

film_grain: limit overlapped pixels to block boundaries

62dd32c4 – Janne Grunau <[email protected]> authored on 2018/12/03 17:05

film_grain: copy unmodified planes before applying noise

70fb01d8 – Ronald S. Bultje <[email protected]> authored on 2018/12/03 09:54

Make per-width versions of cfl_ac

e2aa2d14 – David Michael Barr <[email protected]> authored and Ronald S. Bultje <[email protected]> committed on 2018/11/28 16:00

x86: Add chroma-from-luma AC 4:2:0 AVX2 asm

4903d87b – Janne Grunau <[email protected]> authored on 2018/12/03 15:12

frame-mt: output only fully decoded frames

8cf30020 – James Almer <[email protected]> authored on 2018/12/02 16:33

obu: fix setting num_ticks_per_picture in sequence headers

36b807af – James Almer <[email protected]> authored on 2018/11/30 07:18

data: fix setting Dav1dDataProps defaults in dav1d_data_create()

5e0087d9 – James Almer <[email protected]> authored on 2018/11/29 17:53

common: make 0 the default for Dav1dDataProps.duration

6d87752b – James Almer <[email protected]> authored on 2018/11/29 17:34

common: make it clear that the default values in Dav1dDataProps mean unknown

e4ee4532 – James Almer <[email protected]> authored on 2018/11/29 17:26

common: make -1 the default for Dav1dDataProps.offset

69a0cbd9 – James Almer <[email protected]> authored on 2018/11/29 17:24

common: make Dav1dDataProps.duration signed

327b7de0 – Janne Grunau <[email protected]> authored on 2018/11/29 16:47

backup_lpf: increment row by the correct stripe_h

d5253365 – Ronald S. Bultje <[email protected]> authored on 2018/11/29 09:34

tools/dav1d: abort on unused/unknown CLI arguments

9e08ac71 – Jean-Yves Avenard <[email protected]> authored on 2018/11/29 13:16

Make Dav1dDataProps::timestamp and offset signed.

8ec9793a – Martin Storsjö <[email protected]> authored on 2018/11/29 10:53

ivf: Use _ftelli64 instead of ftello for MSVC

4874e846 – Martin Storsjö <[email protected]> authored on 2018/11/29 10:43

ivf: Use int64_t instead of off_t

be15be35 – Janne Grunau <[email protected]> authored on 2018/11/28 19:25

build: use join_paths() for doxygen paths

9587e705 – Janne Grunau <[email protected]> authored on 2018/11/28 18:25

tile-mt: reset tile tasks when the number of tile rows changes

a1fb2664 – Janne Grunau <[email protected]> authored and Ronald S. Bultje <[email protected]> committed on 2018/11/28 17:22

doc: extend dav1d_flush Doxygen

eb31119d – Janne Grunau <[email protected]> authored and Ronald S. Bultje <[email protected]> committed on 2018/11/28 17:46

add a Doxygen html target

6044a1ae – Francois Cartegnie <[email protected]> authored on 2018/11/23 06:37

add SSSE3 avg/w_avg/mask

08184fd8 – Francois Cartegnie <[email protected]> authored on 2018/11/23 06:38

checkasm: add --list tests

d4ddfdf5 – James Almer <[email protected]> authored on 2018/11/28 12:57

obu: ensure that the frame header reference is writable

a4083095 – Janne Grunau <[email protected]> authored on 2018/11/28 16:08

obu/film grain: check ref frame frame_hdr before use

aee656d7 – Janne Grunau <[email protected]> authored on 2018/11/28 15:16

clean state in dav1d_flush()

3d675c12 – Ronald S. Bultje <[email protected]> authored on 2018/11/27 12:25

Round up film grain block count

454738fa – Ronald S. Bultje <[email protected]> authored on 2018/11/27 12:25

Remove source padding in film grain

9b1d1435 – Ronald S. Bultje <[email protected]> authored on 2018/11/28 08:43

Don't use string literals in assert()

f16ffc83 – Janne Grunau <[email protected]> authored and Jean-Baptiste Kempf <[email protected]> committed on 2018/11/26 11:07

CI: use the sanitizer test setup to prevent timeouts

fcc6f66c – Janne Grunau <[email protected]> authored and Jean-Baptiste Kempf <[email protected]> committed on 2018/11/26 11:05

CI: make sure MSAN and UBSAN warnings are errors

b210e657 – Ronald S. Bultje <[email protected]> authored on 2018/11/28 06:26

Clear coefficient cache upon error

3fe0d740 – Ronald S. Bultje <[email protected]> authored on 2018/11/28 06:24

Use a single atomic integer for flush

718f9254 – Ronald S. Bultje <[email protected]> authored on 2018/11/28 06:20

ivf: store offset and timestamp in pkt.m

ae08430f – James Almer <[email protected]> authored on 2018/11/27 10:25

Add a simple API usage example to the doxy

560dc684 – Nathan E. Egge <[email protected]> authored on 2018/11/26 13:15

Reset the random seed when testing each CPU type.

06331de1 – Janne Grunau <[email protected]> authored on 2018/11/26 17:51

mc/svc: adjust frame-mt wait scanline for 8-tap filter

77bbf721 – Janne Grunau <[email protected]> authored on 2018/11/26 07:35

obu: error out on missing frame_hdrs in ref frames with frame_ref_short_signaling

a90a691d – Janne Grunau <[email protected]> authored on 2018/11/26 14:20

dav1d/md5: avoid unaligned 32-bit reads

e5e3cabe – James Almer <[email protected]> authored on 2018/11/26 09:02

dav1d: add public defines for max amount of tile and frame threads

197a19ad – Janne Grunau <[email protected]> authored on 2018/11/26 08:08

dav1d/cli: start decoding on a sequence header after '--skip'

bbcaf411 – Janne Grunau <[email protected]> authored on 2018/11/26 07:50

fuzzer: use dav1d_parse_sequence_header()

b7c64ad6 – Janne Grunau <[email protected]> authored and Jean-Baptiste Kempf <[email protected]> committed on 2018/11/26 05:49

fix memleaks with ref counted sequence and frame headers

9ce4b734 – Janne Grunau <[email protected]> authored on 2018/11/26 06:19

memset frame_hdr to fix use of uninitilized value regressions