[zlib][arm] Compression: optimize fill_window |
|||||||||||
Issue descriptionSince late 2014, Intel has an optimized fill_window function that helps in compression. We should optimize it (or alternatively the hash chain function) on ARM.
,
Jul 13
,
Aug 7
Benchmarking data: https://docs.google.com/spreadsheets/d/1Vdk4uZj18ksKESXpWKhuTjYM1Ifar0RbKxJpizn0jz4/edit#gid=941756136
,
Aug 7
The following revision refers to this bug: https://chromium.googlesource.com/chromium/src.git/+/ddbbeb05cb3f0a4c27e7a5d5d0305462db373677 commit ddbbeb05cb3f0a4c27e7a5d5d0305462db373677 Author: Adenilson Cavalcanti <adenilson.cavalcanti@arm.com> Date: Tue Aug 07 16:08:59 2018 Neon-Optimized hash chain This should help with compression of data, using NEON instructions (therefore useful for ARMv7/ARMv8). Average gains were around 4% to 5% in data compression, depending on the data entropy. Re-write of a patch contributed to Fedora Core, for reference: https://src.fedoraproject.org/rpms/zlib/c/25e9802713484882c27c1f979a6610a42414ee13?branch=master Bug: 863257 Change-Id: I76573d75843d6a63de75d8a9536da98515314543 Reviewed-on: https://chromium-review.googlesource.com/1136940 Reviewed-by: Mike Klein <mtklein@chromium.org> Reviewed-by: Chris Blume <cblume@chromium.org> Commit-Queue: Adenilson Cavalcanti <cavalcantii@chromium.org> Cr-Commit-Position: refs/heads/master@{#581241} [modify] https://crrev.com/ddbbeb05cb3f0a4c27e7a5d5d0305462db373677/third_party/zlib/BUILD.gn [add] https://crrev.com/ddbbeb05cb3f0a4c27e7a5d5d0305462db373677/third_party/zlib/contrib/optimizations/slide_hash_neon.h [modify] https://crrev.com/ddbbeb05cb3f0a4c27e7a5d5d0305462db373677/third_party/zlib/deflate.c
,
Aug 13
|
|||||||||||
►
Sign in to add a comment |
|||||||||||
Comment 1 by cavalcantii@chromium.org
, Jul 13