EwoutH / rav1e

82%

master: 84%

LAST BUILD ON BRANCH master-old

branch: master-old

CHANGE BRANCH

x

Reset

• master-old
• 0.4-ci
• Arm64-CI
• cargo-tree
• master
• patch-13
• patch-16
• patch-17
• patch-19
• patch-20
• patch-21
• patch-22
• patch-23
• patch-24
• patch-25
• patch-28
• patch-6

Build # 124

Build Type

push

travis-ci

Committed by web-flow

Commit Message

Update from master

* Parallelise .travis.yml to speed up CI

This is an attempt to parallelize Travis to speed up the CI by splitting up the different tests

* Update travis.yml with names and merge build, doc & coveralls

 - Adds names to jobs
 - Merges the 3 fastest stages (Build, Doc & Coveralls) to reduce system load
 - Fixes Coveralls by running it after build script

* Splitting off Doc job again

Splitting off Doc job again since it's now the bottleneck

* Correct incorrect RU addressing introduced in PR848

RU addressing during filteirng in PR 848 works by accident; it assumes
stripes are the same size as the RU, which is not the case (but
currently true in the hardwired test code)

* Correct RestorationState setup addressing bug/removals from PR 848

PR 848 removed some of the RestorationPlane setup needed once RU != SB
In addition, the sb_log2 calculation was incorrect (but worked with current settings)

* Bypass filtering of strengh=0 cdef blocks for speed.

* For hbd use scalar code for subpel (#857)

* Clarify stripe addressing in lrf

Eliminate the passing of a PlaneConfig to stripe addressing, add a
documented field for strip height.

* Early termination for bottomup partition search (i.e. speed 0 only) (#858)

This works for speed 0 (i.e. bottomup partition search) only.

Almost no change in bdrate with more than x2 faster for large QIndex 252
with no change in its bitstream size.

* Simplify inverse txfm code (#860)

* Revert "Bypass filtering of strengh=0 cdef blocks for speed."

This reverts commit df6793dea.

* Use unreachable macro for mc asm code (#865)

* Skip rust format for quantization data (#867)

* Format some very long lines of code (#866)

Format some very long lines of code

* Format uncompressed header writing code (#868)

* Switch nasm-rs to release

* Bump dav1d-rs and bindgen versions to be the same 

Fixes #871.

* Add and... (continued)

Run Details

9550 of 11677 relevant lines covered (81.78%)

0.82 hits per line