| Commit message (Collapse) | Author | Age |
|
|
|
|
|
|
|
|
|
|
| |
I'm not yet caching these in the blitter, and speed is essentially unchanged in the bench where I am now building and compiling the pipeline only once. This may not be able to stay a simple std::function after I figure out caching, but for now it's a nice fit.
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=3911
Change-Id: I9545af589f73baf9f17cb4e6ace9a814c2478fe9
Reviewed-on: https://skia-review.googlesource.com/3911
Reviewed-by: Herb Derby <herb@google.com>
Commit-Queue: Mike Klein <mtklein@chromium.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The portable code now becomes entirely focused on enum+ptr descriptions, leaving the concrete implementation of the pipeline to SkOpts::run_pipeline().
As implemented, the concrete implementation is basically the same, with a little more type safety.
Speed is essentially unchanged on my laptop, and that's having run_pipeline() rebuild its concrete state every call. There's room for improvement there if we split this into a compile_pipeline() / run_pipeline() sort of thing, which is my next planned CL.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=3920
CQ_INCLUDE_TRYBOTS=master.client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
Change-Id: Ie4c554f51040426de7c5c144afa5d9d9d8938012
Reviewed-on: https://skia-review.googlesource.com/3920
Reviewed-by: Herb Derby <herb@google.com>
Commit-Queue: Mike Klein <mtklein@chromium.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
- Give body and tail functions separate types. This frees a register in body functions, especially important for Windows.
- Fill out default, SSE4.1, and HSW versions of all functions. This means we don't have to mess around with SkNf_abi... all functions come from the same compilation unit where SkNf is a single consistent type.
- Move Stage::next() into SkRasterPipeline_opts.h as a static inline function.
- Remove Stage::ctx() entirely... fCtx is literally the same thing.
This is a step along the way toward building the entire pipeline in src/opts, removing the need for all the stages to be functions living in SkOpts.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=3680
CQ_INCLUDE_TRYBOTS=master.client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot,Test-Ubuntu-Clang-GCE-CPU-AVX2-x86_64-Debug-ASAN-Trybot
Change-Id: I7de78ffebc15b9bad4eda187c9f50369cd7e5e42
Reviewed-on: https://skia-review.googlesource.com/3680
Reviewed-by: Herb Derby <herb@google.com>
Commit-Queue: Mike Klein <mtklein@chromium.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The shader leaves its color in r,g,b,a, so to implement this color filter, we move {r,g,b,a} into {dr,dg,db,da}, then load the filter's color in {r,g,b,a}, then apply the xfermode as usual.
I've left a note about how we could sometimes cut a stage for some xfermodes. Similarly we really only need to move_src_dst instead of swap_src_dst, but it seemed handy and less error prone to do a full two way swap. As usual, we can always circle back and fine-tune these things if we want.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=3243
CQ_INCLUDE_TRYBOTS=master.client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
Change-Id: I928c0fb25236eb75cf238134c6bebb53af5ddf07
Reviewed-on: https://skia-review.googlesource.com/3243
Commit-Queue: Mike Klein <mtklein@chromium.org>
Reviewed-by: Matt Sarett <msarett@google.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Original review here: https://skia-review.googlesource.com/c/2990/
Second attempt here: https://skia-review.googlesource.com/c/3064/
This is the same as the second attempt, but with the change to SkOpts_hsw.cpp left out.
That omitted part is the key piece... this just lands the refactoring.
CQ_INCLUDE_TRYBOTS=master.client.skia:Perf-Ubuntu-Clang-GCE-CPU-AVX2-x86_64-Debug-ASAN-Trybot,Perf-Ubuntu-Clang-GCE-CPU-AVX2-x86_64-Debug-GN,Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot,Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-Fast-Trybot;master.client.skia.compile:Build-Win-MSVC-x86_64-Debug-Trybot
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=3242
Change-Id: Iaafa793a4854c2c9cd7e85cca3701bf871253f71
Reviewed-on: https://skia-review.googlesource.com/3242
Reviewed-by: Mike Klein <mtklein@chromium.org>
Commit-Queue: Mike Klein <mtklein@chromium.org>
|
|
|
|
|
|
|
|
|
|
|
|
| |
This reverts commit Id0ba250037e271a9475fe2f0989d64f0aa909bae.
crbug.com/654213
Looks like Chrome Canary's picking up Haswell code on non-Haswell machines.
Change-Id: I16f976da24db86d5c99636c472ffad56db213a2a
Reviewed-on: https://skia-review.googlesource.com/3108
Commit-Queue: Mike Klein <mtklein@chromium.org>
Reviewed-by: Mike Klein <mtklein@chromium.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Original review here: https://skia-review.googlesource.com/c/2990/
Changes since:
- simpler implementations of load_tail() / store_tail(): slower, but more obviously correct to all compilers
- fleshed out math ops on Sk8i and Sk8u to make unit tests happy on -Fast bot (where we always have AVX2)
- now storing stage functions as void(*)() to avoid undefined behavior and/or linker problems. This restores 32-bit Windows.
- all AVX2 Sk8x methods are marked always-inline, to avoid linking the "wrong" version on Debug builds.
CQ_INCLUDE_TRYBOTS=master.client.skia:Perf-Ubuntu-Clang-GCE-CPU-AVX2-x86_64-Debug-ASAN-Trybot,Perf-Ubuntu-Clang-GCE-CPU-AVX2-x86_64-Debug-GN,Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot,Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-Fast-Trybot;master.client.skia.compile:Build-Win-MSVC-x86_64-Debug-Trybot
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=3064
Change-Id: Id0ba250037e271a9475fe2f0989d64f0aa909bae
Reviewed-on: https://skia-review.googlesource.com/3064
Reviewed-by: Mike Klein <mtklein@chromium.org>
Commit-Queue: Mike Klein <mtklein@chromium.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This reverts commit I1c82e5755d8e44cc0b9c6673d04b117f85d71a3a.
Reason for revert: lots of failing bots.
TBR=mtklein@chromium.org,msarett@google.com,reviews@skia.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
Change-Id: I653bed3905187f43196504f19424985fa2a765b5
Reviewed-on: https://skia-review.googlesource.com/3063
Commit-Queue: Mike Klein <mtklein@chromium.org>
Reviewed-by: Mike Klein <mtklein@chromium.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Bench runtime changes:
sRGB: 7194 -> 3735 = 1.93x faster
F16: 6531 -> 2559 = 2.55x faster
Instead of building 4x and 1-3x pipelines and then maybe 8x and 1-7x, instead build either the short ones or the long ones, but not both. If we just take care to use a compatible run_pipeline(), there's some cross-module type disagreement but everything works out in the end.
Oddly, a few places that looked like they'd be faster using SkNx_fma() or Sk4f_round()/Sk8f_round() are actually faster the long way, e.g. multiply, add 0.5, truncate. Curious! In all the other places you see here that I've used SkNx_fma(), it's been a significant speedup.
This folds in a couple refactors and cleanups that I've been meaning to do. Hope you don't mind... if find the new code considerably easier to read than the old code.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2990
CQ_INCLUDE_TRYBOTS=master.client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
Change-Id: I1c82e5755d8e44cc0b9c6673d04b117f85d71a3a
Reviewed-on: https://skia-review.googlesource.com/2990
Reviewed-by: Matt Sarett <msarett@google.com>
Commit-Queue: Mike Klein <mtklein@chromium.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
If we want to support VEX-encoded instructions (AVX, F16C, etc.) without a ridiculous slowdown, we need to make sure we're running either all VEX-encoded instructions or all non-VEX-encoded instructions. That means we cannot mix arbitrary user-defined SkRasterPipeline::Fn (never VEX) with those living in SkOpts (maybe VEX)... it's SkOpts or bust.
This ports the existing user-defined SkRasterPipeline::Fn use cases over to use stock stages from SkOpts. I rewrote the unit test to use stock stages, and moved the SkXfermode implementations to SkOpts. The code deleted for SkArithmeticMode_scalar should already be dead.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2940
CQ_INCLUDE_TRYBOTS=master.client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
Change-Id: I94dbe766b2d65bfec6e544d260f71d721f0f5cb0
Reviewed-on: https://skia-review.googlesource.com/2940
Commit-Queue: Mike Klein <mtklein@google.com>
Reviewed-by: Mike Reed <reed@google.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Haswell brought a whole slew of handy new instructions for us (AVX2, FMA, BMI1+BMI2) and also feature F16C, which came one generation earlier on Ivybridge. We work with integers often enough that we really want to target AVX2 instead of AVX, and this means it's pretty practical to ask for all those other goodies along with it.
Chrome's GN files and Google3's BUILD file will need an update, before or after this CL.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2840
CQ_INCLUDE_TRYBOTS=master.client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
Change-Id: I826daf77b5104664c5d31ddaabee347e287b87a2
Reviewed-on: https://skia-review.googlesource.com/2840
Commit-Queue: Mike Klein <mtklein@chromium.org>
Reviewed-by: Herb Derby <herb@google.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This is handy now, and becomes necessary with fancier backends:
- most code can't speak the type of AVX pipeline stages,
so indirection's definitely needed there;
- if the pipleine is entirely composed of stock stages,
these enum values become an abstract recipe that can be JITted.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2782
CQ_INCLUDE_TRYBOTS=master.client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
Change-Id: Iedd62e99ce39e94cf3e6ffc78c428f0ccc182342
Reviewed-on: https://skia-review.googlesource.com/2782
Reviewed-by: Mike Klein <mtklein@chromium.org>
Commit-Queue: Mike Klein <mtklein@chromium.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This lets them pick up runtime CPU specializations. Here I've plugged in SSE4.1. This is still one of the N prelude CLs to full 8-at-a-time AVX.
I've moved the union of the stages used by SkRasterPipelineBench and SkRasterPipelineBlitter to SkOpts... they'll all be used by the blitter eventually. Picking up SSE4.1 specialization here (even still just 4 pixels at a time) is a significant speedup, especially to store_srgb(), so much that it's no longer really interesting to compare against the fused-but-default-instruction-set version in the bench. So that's gone now.
That left the SkRasterPipeline unit test as the only other user of the EasyFn simplified interface to SkRasterPipeline. So I converted that back down to the bare-metal interface, and EasyFn and its friends became SkRasterPipeline_opts.h exclusive abbreviations (now called Kernel_Sk4f). This isn't really unexpected: SkXfermode also wanted to build up its own little abstractions, and once you build your own abstraction, the value of an additional EasyFn-like layer plummets to negative.
For simplicity I've left the SkXfermode stages alone, except srcover() which was always part of the blitter. No particular reason except keeping the churn down while I hack. These _can_ be in SkOpts, but don't have to be until we go 8-at-a-time.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2752
CQ_INCLUDE_TRYBOTS=master.client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
Change-Id: I3b476b18232a1598d8977e425be2150059ab71dc
Reviewed-on: https://skia-review.googlesource.com/2752
Reviewed-by: Mike Klein <mtklein@chromium.org>
Commit-Queue: Mike Klein <mtklein@chromium.org>
|
|
|
|
|
|
|
|
| |
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2322033002
CQ_INCLUDE_TRYBOTS=master.client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
Review-Url: https://codereview.chromium.org/2322033002
|
|
|
|
|
|
|
|
|
|
| |
This makes it easier to see the baseline CPU feature settings for a build when looking at a stack trace or profile. This fills out all the features we currently
might care about.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2318313003
Review-Url: https://codereview.chromium.org/2318313003
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
For large inputs, this runs ~11x faster than Murmur3.
My bench drops from 1µs to 88ns.
Like x86-64, this runs fastest if we work in 24 byte chunks. 16 byte chunks
run at about 0.75x this speed, 8 byte chunks at about 0.4x (which would still
be about 5x faster than Murmur3).
This'll require plumbing support for opts_crc32 into Chrome first before it can roll.
perf.skia.org charts we want to watch: https://perf.skia.org/#5490
Seach for compute_hash in these logs to see the difference:
baseline: https://luci-milo.appspot.com/swarming/task/30ba22f3dfe30e10/steps/nanobench/0/stdout
trybot: https://luci-milo.appspot.com/swarming/task/30bbc406cbf62d10/steps/nanobench/0/stdout
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2260823002
CQ_INCLUDE_TRYBOTS=master.client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
Review-Url: https://codereview.chromium.org/2260823002
|
|
|
|
|
|
|
|
|
|
|
|
| |
About 9x faster than Murmur3 for long inputs.
Most of this is a mechanical change from SkChecksum::Murmur3(...) to SkOpts::hash(...).
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2208903002
CQ_INCLUDE_TRYBOTS=master.client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot;master.client.skia.compile:Build-Ubuntu-GCC-x86_64-Release-CMake-Trybot,Build-Mac-Clang-x86_64-Release-CMake-Trybot
Review-Url: https://codereview.chromium.org/2208903002
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
(1) Performance is better or stays the same.
(2) Code is split into functions (RasterPipeline-ish
design). IMO, it's not really more or less readable.
But I think it's now much easier add capabilities,
apply optimizations, or do more refactors. Or to
actually use RasterPipeline. I help back from trying
any of these to try to keep this CL sane.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2194303002
CQ_INCLUDE_TRYBOTS=master.client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
Review-Url: https://codereview.chromium.org/2194303002
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Also changes SkColorXform to support:
RGBA->RGBA
RGBA->BGRA
Instead of:
RGBA->SkPMColor
TBR=reed@google.com
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2174493002
CQ_INCLUDE_TRYBOTS=master.client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
Committed: https://skia.googlesource.com/skia/+/73d55332e2846dd05e9efdaa2f017bcc3872884b
Review-Url: https://codereview.chromium.org/2174493002
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
(patchset #9 id:260001 of https://codereview.chromium.org/2174493002/ )
Reason for revert:
Breaking MSAN
Original issue's description:
> Add color space xform support to SkJpegCodec (includes F16!)
>
> Also changes SkColorXform to support:
> RGBA->RGBA
> RGBA->BGRA
>
> Instead of:
> RGBA->SkPMColor
>
> TBR=reed@google.com
> BUG=skia:
> GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2174493002
> CQ_INCLUDE_TRYBOTS=master.client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
>
> Committed: https://skia.googlesource.com/skia/+/73d55332e2846dd05e9efdaa2f017bcc3872884b
TBR=mtklein@google.com,reed@google.com,herb@google.com,brianosman@google.com
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=skia:
Review-Url: https://codereview.chromium.org/2195523002
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Also changes SkColorXform to support:
RGBA->RGBA
RGBA->BGRA
Instead of:
RGBA->SkPMColor
TBR=reed@google.com
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2174493002
CQ_INCLUDE_TRYBOTS=master.client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
Review-Url: https://codereview.chromium.org/2174493002
|
|
|
|
|
|
|
|
| |
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2147763002
CQ_INCLUDE_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot;master.client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
Review-Url: https://codereview.chromium.org/2147763002
|
|
|
|
|
|
|
| |
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2152583002
Review-Url: https://codereview.chromium.org/2152583002
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
If we make sure all SkOpts functions are static, we can give the namespaces any
name we like. This lets us drop the sk_ prefix and give a real indication of
the default SIMD instruction set rather than just saying sk_default.
Both of these changes help debugger, profiler, and crash report readability.
Perhaps more importantly, keeping these functions static helps prevent
accidentally linking in unused versions of functions, as you see here with
sk_avx::srcover_srgb_srgb().
This requires we update SkBlend_opts tests and benches to call SkOpts functions
through SkOpts rather than declaring the methods externally. In practice this
drops testing of the SSE2 version on machines with SSE4. If we still really
need to test/bench the compile time best SIMD level version of this method
against the runtime detected best, we can include SkBlend_opts.h into the tests
or benches directly, similar to what we do for the trivial, brute-force, or best
non-SIMD versions.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2145833002
CQ_INCLUDE_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
Review-Url: https://codereview.chromium.org/2145833002
|
|
|
|
|
|
|
|
|
|
|
| |
This refactors opt code to handle arbitrary src and dst
gammas that are specified by tables.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2130013002
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
Review-Url: https://codereview.chromium.org/2130013002
|
|
|
|
|
|
|
| |
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2087343002
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
Review-Url: https://codereview.chromium.org/2087343002
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
201295.jpg on HP z620 (300x280)
QCMS Xform 0.418 ms
Skia NEW Xform 0.378 ms
Vs QCMS 1.11x
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2078623002
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
Review-Url: https://codereview.chromium.org/2078623002
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
201295.jpg on HP z620
(300x280, most common form of sRGB profile)
QCMS Xform 0.495 ms
Skia Old Xform 0.235 ms
Skia NEW Xform 0.423 ms
Vs Old Code 0.56x
Vs QCMS 1.17x
So to summarize, we are now much slower than before,
but still a bit faster than QCMS. And now we are also
far more accurate than QCMS :).
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2060823003
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
Review-Url: https://codereview.chromium.org/2060823003
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Because we recognize commonly used gamma tables and
parameters as 2.2f, about 98% of jpegs with color profiles
will pass through this xform (assuming the dst is also
2.2f). Sample size is 10,322 jpegs.
I won't go crazy with performance numbers because this is
a work in progress, particularly in terms of correctness.
201295.jpg on HP z620
(300x280, most common form of sRGB profile)
Decode Time + QCMS Xform 1.28 ms
QCMS Xform Only 0.495 ms
Decode Time + Skia Opt Xform 1.01 ms
Skia Opt Xform Only 0.235 ms
Decode Time + Xform Speed-up 1.27x
Xform Only Speed-up 2.11x
FWIW, Skia xform time before these optimizations was
41.1 ms. But we expected that code to be slow.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2046013002
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
Review-Url: https://codereview.chromium.org/2046013002
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
When targetting iOS and using gyp to generate the build files, it is not
possible to select files to build depending on the architecture. Due to
that, the skia code was disabling all optimisation when SK_BUILD_FOR_IOS
was defined.
Since it is possible to select the correct optimised version when using
gn, this pessimisation is hurting the build. Introduce a new define to
disable the optimisation SK_BUILD_NO_OPTS. It will be used by Chromium
when building skia for iOS with gyp but not gn.
Define SK_BUILD_NO_OPTS along-side SK_BUILD_FOR_IOS for all files that
look like build configuration (Xcode projects, gyp configuration files,
public.bzl) in order to avoid introducing breakage on those builds.
BUG=607933
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2002423002
Review-Url: https://codereview.chromium.org/2002423002
|
|
|
|
|
|
|
|
| |
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1952953004
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
Review-Url: https://codereview.chromium.org/1952953004
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Herb's really starting to get serious about tweaking this, which becomes
a lot easier when you've got SkOpts' runtime CPU detection. We should be
able to optimize this usefully for SSSE3, SSE4.1, AVX, AVX2, or NEON.
(We can of course implement a subset.)
This function takes two counts to give us flexibility to write src patterns:
nsrc >= ndst -> the usual srcover function
nsrc < ndst -> repeat src until it fills dst
nsrc << ndst -> possibly preprocess src into registers
nsrc == 1 -> equivalent of blitrow_color32, srcover_1, etc.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1939783003
Review-Url: https://codereview.chromium.org/1939783003
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
- Moves CPU feature detection to its own file.
- Cleans up some redundant feature detection scattered around core/ and opts/.
- Can now detect a few new CPU features:
* F16C -> Intel f16<->f32 instructions, added between AVX and AVX2
* FMA -> Intel FMA instructions, added at the same time as AVX2
* VFP_FP16 -> ARM f16<->f32 instructions, quite common
* NEON_FMA -> ARM FMA instructions, also quite common
* SSE and SSE3... why not?
This new internal API makes it very cheap to do fine-grained runtime CPU
feature detection. Redundant calls to SkCpu::Supports() should be eliminated
and it's hoistable out of loops. It compiles away entirely when we have the
appropriate instructions available at compile time.
This means we can call it to guard even a little snippet of 1 or 2 instructions
right where needed and let inlining hoist the check (if any at all) up to
somewhere that doesn't hurt performance. I've explained how I made this work
in the private section of the new header.
Once this lands and bakes a bit, I'll start following up with CLs to use it more
and to add a bunch of those little 1-2 instruction snippets we've been wanting,
e.g. cvtps2ph, cvtph2ps, ptest, pmulld, pmovzxbd, blendvps, pshufb, roundps
(for floor) on x86, and vcvt.f32.f16, vcvt.f16.f32 on ARM.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1890483002
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
Committed: https://skia.googlesource.com/skia/+/872ea29357439f05b1f6995dd300fc054733e607
Review URL: https://codereview.chromium.org/1890483002
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Static initializers run in a confusing unspecified order,
so it's best to have as few of them as possible.
Most tools and clients I can find already call SkGraphics::Init(),
(or equivalently create an SkAutoGraphics) which calls SkOpts::Init():
- Chrome
- Chrome renderer
- Android
- DM
- nanobench
- SampleApp
- VisualBench
- the old debugger
Seems like the only thing relying on this static initializer today is
the new debugger, fixed here.
TBR=reed@google.com
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1903503002
Review URL: https://codereview.chromium.org/1903503002
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The API and implementation are very much simplified.
You may not want to bother reading the diff.
As is our trend, SkOnce now uses <atomic> directly.
Member initialization means we don't need SK_DECLARE_STATIC_ONCE.
SkSpinlock already works this same way.
All uses of the old API taking an external bool* and Lock* were pessimal,
so I have not carried this sort of API forward. It's simpler, faster,
and more space-efficient to always use this single SkOnce class interface.
SkOnce weighs 2 bytes: a done bool and an SkSpinlock, also a bool internally.
This API refactoring opens up the opportunity to fuse those into a single
three-state byte if we'd like.
No public API changes.
TBR=reed@google.com
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1894893002
Review URL: https://codereview.chromium.org/1894893002
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
of https://codereview.chromium.org/1890483002/ )
Reason for revert:
many unexpected GM diffs across GPU+CPU configs on Windows (hopefully just text masks on GPU?). seems like we pick a different srcover variant in some places.
Original issue's description:
> Move CPU feature detection to its own file.
>
> - Moves CPU feature detection to its own file.
> - Cleans up some redundant feature detection scattered around core/ and opts/.
> - Can now detect a few new CPU features:
> * F16C -> Intel f16<->f32 instructions, added between AVX and AVX2
> * FMA -> Intel FMA instructions, added at the same time as AVX2
> * VFP_FP16 -> ARM f16<->f32 instructions, quite common
> * NEON_FMA -> ARM FMA instructions, also quite common
> * SSE and SSE3... why not?
>
> This new internal API makes it very cheap to do fine-grained runtime CPU
> feature detection. Redundant calls to SkCpu::Supports() should be eliminated
> and it's hoistable out of loops. It compiles away entirely when we have the
> appropriate instructions available at compile time.
>
> This means we can call it to guard even a little snippet of 1 or 2 instructions
> right where needed and let inlining hoist the check (if any at all) up to
> somewhere that doesn't hurt performance. I've explained how I made this work
> in the private section of the new header.
>
> Once this lands and bakes a bit, I'll start following up with CLs to use it more
> and to add a bunch of those little 1-2 instruction snippets we've been wanting,
> e.g. cvtps2ph, cvtph2ps, ptest, pmulld, pmovzxbd, blendvps, pshufb, roundps
> (for floor) on x86, and vcvt.f32.f16, vcvt.f16.f32 on ARM.
>
> BUG=skia:
> GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1890483002
> CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
>
> Committed: https://skia.googlesource.com/skia/+/872ea29357439f05b1f6995dd300fc054733e607
TBR=fmalita@chromium.org,herb@google.com,reed@google.com,mtklein@chromium.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=skia:
Review URL: https://codereview.chromium.org/1892643003
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
These are implemented generically with Sk4s and don't benefit
from anything fancier than vanilla SSE/NEON.
This means there's no need to hide this code away in another
file or behind a function pointer... it's readable and we have
compile-time support for all the instructions it needs.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1872193002
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
Review URL: https://codereview.chromium.org/1872193002
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
- Moves CPU feature detection to its own file.
- Cleans up some redundant feature detection scattered around core/ and opts/.
- Can now detect a few new CPU features:
* F16C -> Intel f16<->f32 instructions, added between AVX and AVX2
* FMA -> Intel FMA instructions, added at the same time as AVX2
* VFP_FP16 -> ARM f16<->f32 instructions, quite common
* NEON_FMA -> ARM FMA instructions, also quite common
* SSE and SSE3... why not?
This new internal API makes it very cheap to do fine-grained runtime CPU
feature detection. Redundant calls to SkCpu::Supports() should be eliminated
and it's hoistable out of loops. It compiles away entirely when we have the
appropriate instructions available at compile time.
This means we can call it to guard even a little snippet of 1 or 2 instructions
right where needed and let inlining hoist the check (if any at all) up to
somewhere that doesn't hurt performance. I've explained how I made this work
in the private section of the new header.
Once this lands and bakes a bit, I'll start following up with CLs to use it more
and to add a bunch of those little 1-2 instruction snippets we've been wanting,
e.g. cvtps2ph, cvtph2ps, ptest, pmulld, pmovzxbd, blendvps, pshufb, roundps
(for floor) on x86, and vcvt.f32.f16, vcvt.f16.f32 on ARM.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1890483002
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
Review URL: https://codereview.chromium.org/1890483002
|
|
|
|
|
|
|
|
|
|
| |
This should be a pixel-for-pixel (i.e. bug-for-bug) port.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1820313002
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
Review URL: https://codereview.chromium.org/1820313002
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Most of these implementations now just say "always inline".
Let's see if we can get away with the simplicity of doing that all the time.
These inlined implementations can autovectorize easily.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1639863002
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
Review URL: https://codereview.chromium.org/1639863002
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Nothing fancy yet, just calls the serial code in a loop.
I will try to folow this up with at least some of:
- SSE2 version of serial code
- NEON version of serial code
- NEON version using vcvt.f32.f16/vcvt.f16.f32
- F16C (between AVX and AVX2) version using vcvtph2ps/vcvtps2ph
The last two are fastest but need runtime detection.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1686543003
Review URL: https://codereview.chromium.org/1686543003
|
|
|
|
|
|
|
|
|
|
| |
refactoring.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1679053002
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
Review URL: https://codereview.chromium.org/1679053002
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Swizzle Bench Runtime
Nexus 6P 0.14x
Dell Venue 8 0.12x
CMYK Jpeg Decode Runtime
Nexus 6P 0.81x
Dell Venue 8 0.85x
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1676773003
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
Review URL: https://codereview.chromium.org/1676773003
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
PNG Decode Time Nexus 6P (for a test set of GrayAlpha encoded PNGs)
Regular Unpremul 0.91x
Zero Init Unpremul 0.92x
Regular Premul 0.84x
Zero Init Premul 0.86x
BUG=skia:4767
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1663623002
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
Review URL: https://codereview.chromium.org/1663623002
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Swizzle Bench Runtime
Nexus 6P 0.32x
Nexus 9 0.89x
PNG Decode Time (for test set of gray encoded PNGs)
Nexus 6P 0.88x
Nexus 9 0.91x
BUG=skia:4767
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1656383002
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
Review URL: https://codereview.chromium.org/1656383002
|
|
|
|
|
|
|
|
|
|
|
|
| |
This is the first of many little baby steps to have us stop runtime-detecting NEON.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1616013003
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
Committed: https://skia.googlesource.com/skia/+/efcc125acd2d71eb077caf6db65fdd6b9eb1dc0d
Review URL: https://codereview.chromium.org/1616013003
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
https://codereview.chromium.org/1532613002/ )
Reason for revert:
Bot failures
Original issue's description:
> AVX 2 SrcOver blits: color32, blitmask.
>
> As a follow up to the SSE 4.1 CL, this should look pretty familiar.
>
> I've made some organizational changes around how we load, store, pack, and unpack data that I think makes things clearer and more orthogonal, and it'll make it easier to try out a pmaddubsw lerp. I have backported these changes to the SSE 4.1 code, and I hope that I can actually get a lot of this code templated for sharing between the two later.
>
> Perf changes (relative to SSE 4.1):
> Xfermode_SrcOver: 1650 -> 1180 (0.71x) // large opaque blit
> Xfermode_SrcOver_aa: 1794 -> 1653 (0.92x) // large opaque + small transparent
> text_16_AA_{FF,BK,WT}: 1.72 -> 1.59 (0.92x) // small opaque blit
> text_16_AA_88: 1.83 -> 1.77 (0.97x) // small transparent blit
>
> This should be a big throughout win, and a small latency win.
> This should all be pixel-exact to the previous SSE 4.1 code.
>
>
> GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1532613002
> CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot;client.skia.compile:Build-Ubuntu-GCC-x86_64-Release-CMake-Trybot,Build-Mac10.9-Clang-x86_64-Release-CMake-Trybot
>
> Committed: https://skia.googlesource.com/skia/+/5d2117015eb271e09faf4a7ddd89093c9d618a36
TBR=herb@google.com,mtklein@google.com,mtklein@chromium.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
Review URL: https://codereview.chromium.org/1632713002
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
As a follow up to the SSE 4.1 CL, this should look pretty familiar.
I've made some organizational changes around how we load, store, pack, and unpack data that I think makes things clearer and more orthogonal, and it'll make it easier to try out a pmaddubsw lerp. I have backported these changes to the SSE 4.1 code, and I hope that I can actually get a lot of this code templated for sharing between the two later.
Perf changes (relative to SSE 4.1):
Xfermode_SrcOver: 1650 -> 1180 (0.71x) // large opaque blit
Xfermode_SrcOver_aa: 1794 -> 1653 (0.92x) // large opaque + small transparent
text_16_AA_{FF,BK,WT}: 1.72 -> 1.59 (0.92x) // small opaque blit
text_16_AA_88: 1.83 -> 1.77 (0.97x) // small transparent blit
This should be a big throughout win, and a small latency win.
This should all be pixel-exact to the previous SSE 4.1 code.
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1532613002
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot;client.skia.compile:Build-Ubuntu-GCC-x86_64-Release-CMake-Trybot,Build-Mac10.9-Clang-x86_64-Release-CMake-Trybot
Review URL: https://codereview.chromium.org/1532613002
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
https://codereview.chromium.org/1616013003/ )
Reason for revert:
This is somehow blocking the Google3 roll in ways neither Ben nor I understand. Precautionary revert... will try again Monday.
Original issue's description:
> de-proc sk_float_rsqrt
>
> This is the first of many little baby steps to have us stop runtime-detecting NEON.
>
> BUG=skia:
> GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1616013003
> CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
>
> Committed: https://skia.googlesource.com/skia/+/efcc125acd2d71eb077caf6db65fdd6b9eb1dc0d
TBR=reed@google.com,mtklein@chromium.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=skia:
Review URL: https://codereview.chromium.org/1629503002
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Swizzle Bench Runtime Nexus 6P
xxx_xxxa 0.32x
xxx_swaprb_xxxa 0.31x
Swizzle Bench Runtime Nexus 9
xxx_xxxa 1.11x
xxx_swaprb_xxxa 1.14x
(This is a slow down.)
Swizzle Bench Runtime Nexus 5
xxx_xxxa 0.12x
xxx_swaprb 0.12x
RGB PNG Decode Runtime
Nexus 6P 0.94x
Nexus 9 0.98x
I don't know how to explain the fact that the Swizzle Bench was
slower on Nexus 9, but the decode times got faster.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1618003002
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
Review URL: https://codereview.chromium.org/1618003002
|