summaryrefslogtreecommitdiff
BranchCommit messageAuthorAge
benchmark/lowlevel-blt-benchBENCH: Benchmark most frequently used paths onlyTaekyun Kim13 years
masterARM: NEON: Fix assembly typo error in src_n_8_8888Taekyun Kim13 years
neon_bilinearARM: NEON: Instruction scheduling of bilinear over_8888_8_8888Taekyun Kim13 years
neon_src_n_8_dstARM: NEON: Standard fast path src_n_8_8Taekyun Kim13 years
normal_repeatSimple repeat: Extend too short source scanline into temporary bufferTaekyun Kim13 years
opacityARM: NEON Add opacity fast pathsTaekyun Kim13 years
over_n_xxxxARM: NEON: Better instruction scheduling of over_n_8888Taekyun Kim13 years
perfstatperfstat: Gather performance data from pixman_blt()Taekyun Kim12 years
sent_to_upstreamReplace FAST_PATH_SAMPLES_COVER_CLIP with FAST_PATH_SAMPLES_COVER_CLIP_NEARESTTaekyun Kim13 years
simple_repeatSimple repeat: Extend too short source scanlines into temporary bufferTaekyun Kim13 years
spansARM: NEON: Fix to fit in new function path tableTaekyun Kim13 years
sse2_bilinearsse2: Bilinear scaled over_8888_8_8888Taekyun Kim13 years
std_normal_repeatARM: Source scanline extension for non-scaled REPEAT_NORMALTaekyun Kim13 years
test_and_perfcomposite_bench: Beautify output message and some refinementTaekyun Kim13 years
threaded_compositeThreaded compositingTaekyun Kim13 years
wip/spansperfstat: support for spans and lerpTaekyun Kim13 years