| Commit message (Collapse) | Author | Age |
|---|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This should make each compilation unit's SkNx types distinct from each other's as far as C++ cares. This keeps us from violating the One Definition Rule with different implementations for the same function.
Here's an example I like. Sk4i SkNx_cast(Sk4b) has at least 4 different sensible implementations:
- SSE2: punpcklbw xmm, zero; punpcklbw xmm, zero
- SSSE3: load mask; pshufb xmm, mask
- SSE4.1: pmovzxbd
- AVX2: vpmovzxbd
We really want all these to inline, but if for some reason they don't (Debug build, poor inliner) and they're compiled in SkOpts.cpp, SkOpts_ssse3.cpp, SkOpts_sse41.cpp, SkOpts_hsw.cpp... boom!
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=3461
Change-Id: I0088ebfd7640c1b0de989738ed43c81b530dc0d9
Reviewed-on: https://skia-review.googlesource.com/3461
Reviewed-by: Matt Sarett <msarett@google.com>
Commit-Queue: Mike Klein <mtklein@chromium.org>
|
| |
|
|
|
|
|
|
|
|
|
| |
Notable tricks:
- v{ld,st}4q_f32 handle transposing loads and stores of floats in one step
- vcvtq_n_{f32_u32,u32_f32} let us do conversion to and from floats without shifts
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1828613002
Review URL: https://codereview.chromium.org/1828613002
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
- This drops the minimum requirement for Sk4x4f on x86 to SSE2 by
removing calls to _mm_shuffle_epi8(). Instead we use good old
shifting and masking.
- Performance is very similar to SSSE3, close enough I'm having trouble
telling which is faster. I think we should let ourselves circle back
on whether we need an SSSE3 version later. When possible it's nice
to stick to SSE2: it's most available, and performs most uniformly
across different chips.
This makes Sk4x4f fast on Windows and Linux, and may help mobile x86.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1817353005
Review URL: https://codereview.chromium.org/1817353005
|
|
|
An API for loading and storing 4 Sk4f with transpose.
This has SSSE3+ and portable versions.
SSE2 and NEON versions to follow.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1825663002
Review URL: https://codereview.chromium.org/1825663002
|
Mike Klein