| Commit message (Collapse) | Author | Age |
|
|
|
|
|
|
|
|
|
| |
Must have been we needed them to be weird (mutable, const setter) before.
It doesn't look like that's necessary now... we can just pass it to the
constructor.
BUG=skia:
Review URL: https://codereview.chromium.org/1112833003
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
WIP
- Can accumulate (write) data in one thread, and share snapshots of it in other threads
... e.g. network accumulates image data, and periodically we want to decode/draw it
- If this sort of thing sticks, should we promote SkData to have the same generality as
SkRBuffer?
BUG=skia:
TBR=
Review URL: https://codereview.chromium.org/1106113002
|
|
|
|
|
|
| |
parentheses [-Wparentheses-equality]
Review URL: https://codereview.chromium.org/1115433004
|
|
|
|
|
|
|
|
|
|
|
|
| |
We're currently decrementing early in checkForDeferredSave(), but that
means we break the getSaveCount() invariant during willSave() dispatch.
Instead, we can defer to doSave(), until after willSave() is called.
This allows overriders to call getSaveCount().
R=reed@google.com,robertphillips@google.com
Review URL: https://codereview.chromium.org/1113603002
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Also strips SK_SUPPORT_LEGACY_COLOR32_MATH,
which is no longer needed.
Seems handy to have SkTypes include the relevant intrinsics when
we know we've got them, but I'm not married to it.
Locally this looks like a pointlessly small perf win, but I'm mostly
keen to get all the code together.
BUG=skia:
Committed: https://skia.googlesource.com/skia/+/376e9bc206b69d9190f38dfebb132a8769bbd72b
Committed: https://skia.googlesource.com/skia/+/d65dc0cedd5b50dd407b6ff8fdc39123f11511cc
CQ_EXTRA_TRYBOTS=client.skia.compile:Build-Ubuntu-GCC-Mips-Debug-Android-Trybot
Review URL: https://codereview.chromium.org/1104183004
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
https://codereview.chromium.org/1104183004/)
Reason for revert:
duh
Original issue's description:
> De-proc Color32
>
> Also strips SK_SUPPORT_LEGACY_COLOR32_MATH,
> which is no longer needed.
>
> Seems handy to have SkTypes include the relevant intrinsics when
> we know we've got them, but I'm not married to it.
>
> Locally this looks like a pointlessly small perf win, but I'm mostly
> keen to get all the code together.
>
> BUG=skia:
>
> Committed: https://skia.googlesource.com/skia/+/376e9bc206b69d9190f38dfebb132a8769bbd72b
>
> Committed: https://skia.googlesource.com/skia/+/d65dc0cedd5b50dd407b6ff8fdc39123f11511cc
TBR=reed@google.com,mtklein@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=skia:
Review URL: https://codereview.chromium.org/1102363006
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Also strips SK_SUPPORT_LEGACY_COLOR32_MATH,
which is no longer needed.
Seems handy to have SkTypes include the relevant intrinsics when
we know we've got them, but I'm not married to it.
Locally this looks like a pointlessly small perf win, but I'm mostly
keen to get all the code together.
BUG=skia:
Committed: https://skia.googlesource.com/skia/+/376e9bc206b69d9190f38dfebb132a8769bbd72b
Review URL: https://codereview.chromium.org/1104183004
|
|
|
|
|
|
|
|
|
|
|
|
| |
This is a logical no-op. Everything was using the equivalent of rsqrt1() before, and is now after.
BUG=skia:
Committed: https://skia.googlesource.com/skia/+/9de16283fdc8cc0d31a84f503578d0ecea4e8297
CQ_EXTRA_TRYBOTS=client.skia.compile:Build-Ubuntu-GCC-Arm64-Debug-Android-Trybot
Review URL: https://codereview.chromium.org/1109913002
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
https://codereview.chromium.org/1104183004/)
Reason for revert:
MIPS
Original issue's description:
> De-proc Color32
>
> Also strips SK_SUPPORT_LEGACY_COLOR32_MATH,
> which is no longer needed.
>
> Seems handy to have SkTypes include the relevant intrinsics when
> we know we've got them, but I'm not married to it.
>
> Locally this looks like a pointlessly small perf win, but I'm mostly
> keen to get all the code together.
>
> BUG=skia:
>
> Committed: https://skia.googlesource.com/skia/+/376e9bc206b69d9190f38dfebb132a8769bbd72b
TBR=reed@google.com,mtklein@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=skia:
Review URL: https://codereview.chromium.org/1108163002
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Also strips SK_SUPPORT_LEGACY_COLOR32_MATH,
which is no longer needed.
Seems handy to have SkTypes include the relevant intrinsics when
we know we've got them, but I'm not married to it.
Locally this looks like a pointlessly small perf win, but I'm mostly
keen to get all the code together.
BUG=skia:
Review URL: https://codereview.chromium.org/1104183004
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
on ARM (patchset #2 id:20001 of https://codereview.chromium.org/1109913002/)
Reason for revert:
arm64 typos
Original issue's description:
> Split rsqrt into rsqrt{0,1,2}, with increasing cost and precision on ARM
>
> This is a logical no-op. Everything was using the equivalent of rsqrt1() before, and is now after.
>
> BUG=skia:
>
> Committed: https://skia.googlesource.com/skia/+/9de16283fdc8cc0d31a84f503578d0ecea4e8297
TBR=reed@google.com,mtklein@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=skia:
Review URL: https://codereview.chromium.org/1105233003
|
|
|
|
|
|
|
|
| |
This is a logical no-op. Everything was using the equivalent of rsqrt1() before, and is now after.
BUG=skia:
Review URL: https://codereview.chromium.org/1109913002
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
patch from issue 1072303005 at patchset 40001 (http://crrev.com/1072303005#ps40001)
This looks quite launchable. radial_gradient3, min of 100 samples:
N5: 985µs -> 946µs
MBP: 395µs -> 279µs
On my MBP, most of the meat looks like it's now in reading the cache and writing to dst one color at a time. Is that something we could do in float math rather than with a lookup table?
BUG=skia:
CQ_EXTRA_TRYBOTS=client.skia.compile:Build-Mac10.8-Clang-Arm7-Debug-Android-Trybot,Build-Ubuntu-GCC-Arm7-Release-Android_NoNeon-Trybot
Committed: https://skia.googlesource.com/skia/+/abf6c5cf95e921fae59efb487480e5b5081cf0ec
Review URL: https://codereview.chromium.org/1109643002
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
id:120001 of https://codereview.chromium.org/1109643002/)
Reason for revert:
compile failures.
Original issue's description:
> Mike's radial gradient CL with better float -> int.
>
> patch from issue 1072303005 at patchset 40001 (http://crrev.com/1072303005#ps40001)
>
> This looks quite launchable. radial_gradient3, min of 100 samples:
> N5: 985µs -> 946µs
> MBP: 395µs -> 279µs
>
> On my MBP, most of the meat looks like it's now in reading the cache and writing to dst one color at a time. Is that something we could do in float math rather than with a lookup table?
>
> BUG=skia:
>
> CQ_EXTRA_TRYBOTS=client.skia.android:Test-Android-GCC-Nexus5-CPU-NEON-Arm7-Debug-Trybot,Test-Android-GCC-Nexus9-CPU-Denver-Arm64-Debug-Trybot
>
> Committed: https://skia.googlesource.com/skia/+/abf6c5cf95e921fae59efb487480e5b5081cf0ec
TBR=reed@google.com,robertphillips@google.com,mtklein@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=skia:
Review URL: https://codereview.chromium.org/1109883003
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
patch from issue 1072303005 at patchset 40001 (http://crrev.com/1072303005#ps40001)
This looks quite launchable. radial_gradient3, min of 100 samples:
N5: 985µs -> 946µs
MBP: 395µs -> 279µs
On my MBP, most of the meat looks like it's now in reading the cache and writing to dst one color at a time. Is that something we could do in float math rather than with a lookup table?
BUG=skia:
CQ_EXTRA_TRYBOTS=client.skia.android:Test-Android-GCC-Nexus5-CPU-NEON-Arm7-Debug-Trybot,Test-Android-GCC-Nexus9-CPU-Denver-Arm64-Debug-Trybot
Review URL: https://codereview.chromium.org/1109643002
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This may be a small help to slimming paint:
picture_overhead_draw 1.25us -> 1.22us 0.98x
picture_overhead_nodraw 318ns -> 276ns 0.87x
It certainly cannot hurt performance.
BUG=chromium:470553
TBR=reed@google.com
No public API changes.
Review URL: https://codereview.chromium.org/1098183003
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This saves "up to" 22 bytes per SkPicture.
Due to alignment, this reduces sizeof(SkPicture):
- from 88 to 72 bytes on Linux x64
- from 68 to 48 bytes on Android
(with Chrome's build settings)
It also somewhat simplifies the GPU veto logic.
BUG=skia:
[mtklein] No public API changes.
TBR=reed@google.com
Review URL: https://codereview.chromium.org/1060863004
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The entries were found by the following command line:
$ find . -regex ".*\.[cChH]\(pp\)?" | xargs git grep -e ';;' --and --not
-e 'for *(.*;;'
Which is a combination of http://stackoverflow.com/a/3858879 and
http://gitster.livejournal.com/27674.html
BUG=None
R=mtklein@google.com
Review URL: https://codereview.chromium.org/1088763005
|
|
|
|
| |
Review URL: https://codereview.chromium.org/1101293003
|
|
|
|
|
|
|
| |
R=egdaniel@google.com
BUG=skia:
Review URL: https://codereview.chromium.org/1098753004
|
|
|
|
|
|
|
|
|
|
| |
Chrome wants to call this more often, and it's quite slow today.
Seems like this could be clearer if SkPictureUtils::ApproxBytesUsed() were SkPicture::approxBytesUsed().
BUG=chromium:471873
Review URL: https://codereview.chromium.org/1090943004
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
https://codereview.chromium.org/1098913002/)
Reason for revert:
Xfermode_SrcOver not looking encouraging. Up to 50% regressions.
https://perf.skia.org/#3242
Original issue's description:
> Convert Color32 code to perfect blend.
>
> Before we commit to blend_256_round_alt, let's make sure blend_perfect is
> really slower in practice (i.e. regresses on perf.skia.org).
>
> blend_perfect is really the most desirable algorithm if we can afford it. Not
> only is it correct, but it's easy to think about and break into correct pieces:
> for instance, its div255() doesn't require any coordination with the multiply.
>
> This looks like a 30% hit according to microbenches. That said, microbenches
> said my previous change would be a 20-25% perf improvement, but it didn't end
> up showing a significant effect at a high level.
>
> As for correctness, I see a bunch of off-by-1 compared to blend_256_round_alt
> (exactly what we'd expect), and one off-by-3 in a GM that looks like it has a
> bunch of overdraw.
>
> BUG=skia:
>
> Committed: https://skia.googlesource.com/skia/+/61221e7f87a99765b0e034020e06bb018e2a08c2
TBR=reed@google.com,fmalita@chromium.org,mtklein@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=skia:
Review URL: https://codereview.chromium.org/1083923006
|
|
|
|
|
|
| |
BUG=skia:
Review URL: https://codereview.chromium.org/1057043003
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Before we commit to blend_256_round_alt, let's make sure blend_perfect is
really slower in practice (i.e. regresses on perf.skia.org).
blend_perfect is really the most desirable algorithm if we can afford it. Not
only is it correct, but it's easy to think about and break into correct pieces:
for instance, its div255() doesn't require any coordination with the multiply.
This looks like a 30% hit according to microbenches. That said, microbenches
said my previous change would be a 20-25% perf improvement, but it didn't end
up showing a significant effect at a high level.
As for correctness, I see a bunch of off-by-1 compared to blend_256_round_alt
(exactly what we'd expect), and one off-by-3 in a GM that looks like it has a
bunch of overdraw.
BUG=skia:
Review URL: https://codereview.chromium.org/1098913002
|
|
|
|
|
|
|
|
|
|
|
| |
combination. There are still a class of cubic tests that fail and a handful of undiagnosed failures from skps and fuzz tests, but things are much better overall.
Extended tests (150M+) run to completion in release in about 6 minutes; the standard test suite exceeds 100K and finishes in a few seconds on desktops.
TBR=reed
BUG=skia:3588
Review URL: https://codereview.chromium.org/1037953004
|
|
|
|
|
|
| |
BUG=skia:3741
Review URL: https://codereview.chromium.org/1091173003
|
|
|
|
|
|
|
|
|
| |
no image changes expected in skia or chrome/blink
BUG=skia:
TBR=
Review URL: https://codereview.chromium.org/1050953003
|
|
|
|
| |
Review URL: https://codereview.chromium.org/1090103002
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This algorithm changes the blend math, guarded by SK_LEGACY_COLOR32_MATH. The new math is more correct: it's never off by more than 1, and correct in all the interesting 0x00 and 0xFF edge cases, where the old math was never off by more than 2, and not always correct on the edges.
If you look at tests/BlendTest.cpp, the old code was using the `blend_256_plus1_trunc` algorithm, while the new code uses `blend_256_round_alt`. Neither uses `blend_perfect`, which is about ~35% slower than `blend_256_round_alt`.
This will require an unfathomable number of rebaselines, first to Skia, then to Blink when I remove the guard.
I plan to follow up with some integer SIMD abstractions that can unify these two implementations into a single algorithm. This was originally what I was working on here, but the correctness gains seem to be quite compelling. The only places these two algorithms really differ greatly now is the kernel function, and even there they can really both be expressed abstractly as:
- multiply 8-bits and 8-bits producing 16-bits
- add 16-bits to 16-bits, returning the top 8 bits.
All the constants are the same, except SSE is a little faster to keep 8 16-bit inverse alphas, NEON's a little faster to keep 8 8-bit inverse alphas. I may need to take this small speed win back to unify the two.
We should expect a ~25% speedup on Intel (mostly from unrolling to 8 pixels) and a ~20% speedup on ARM (mostly from using vaddhn to add `color`, round, and narrow back down to 8-bit all into one instruction.
(I am probably missing several more related bugs here.)
BUG=skia:3738,skia:420,chromium:111470
Review URL: https://codereview.chromium.org/1092433002
|
|
|
|
|
|
|
|
|
|
| |
No one actually uses it, so remove it. Keep the string id for backward
compatibility.
TBR=reed@google.com
This doesn't change any public API.
Review URL: https://codereview.chromium.org/1057413005
|
|
|
|
|
|
| |
BUG=skia:3735
Review URL: https://codereview.chromium.org/1097433002
|
|
|
|
|
|
|
| |
BUG=skia:
TBR=
Review URL: https://codereview.chromium.org/1084623004
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
https://codereview.chromium.org/1078413003/)"
ah ha! Check for the define *after* we pull in SkUserConfig.h (indirectly)
This reverts commit 639a82855b94b93c4fa45560e67df8ec4a8bbb3a.
BUG=skia:
TBR=
Review URL: https://codereview.chromium.org/1084283003
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
id:1 of https://codereview.chromium.org/1048323005/)
Reason for revert:
end of test
Original issue's description:
> make blitAntiH2 virtual, to prep for larger change
>
> testing...
>
> BUG=skia:
> TBR=
>
> Committed: https://skia.googlesource.com/skia/+/33afdaa5213cf118b3ca290bb343253c8593756d
TBR=
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=skia:
Review URL: https://codereview.chromium.org/1086293002
|
|
|
|
|
|
|
|
|
| |
testing...
BUG=skia:
TBR=
Review URL: https://codereview.chromium.org/1048323005
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
of https://codereview.chromium.org/1078413003/)
Reason for revert:
Crazy failings in chrome tests
http://build.chromium.org/p/client.skia/builders/Linux%20Tests/builds/1862
Original issue's description:
> Speeup hairline curves (quads and cubics)
>
> /skia/trunk> cat ../old.txt
> maxrss loops min median mean max stddev samples config bench
> 9M 1 4.28ms 4.32ms 4.36ms 4.67ms 3% ▄▁▁▃▂▂▁▁▂█ 8888 path_hairline_small_AA_cubic
> 9M 1 743µs 767µs 770µs 825µs 4% ▃▃▇▃▁▁▅▁█▁ 8888 path_hairline_small_AA_conic
> 9M 1 533µs 606µs 598µs 680µs 9% ▁▂▂█▆▇▇▄▂▂ 8888 path_hairline_small_AA_quad
> 9M 1 451µs 452µs 456µs 495µs 3% ▁▁▁▁█▁▁▁▁▁ 8888 path_hairline_small_AA_line
>
> /skia/trunk> cat ../new.txt
> maxrss loops min median mean max stddev samples config bench
> 9M 1 827µs 827µs 831µs 869µs 2% ▁▁▁▁▁▁▁█▁▁ 8888 path_hairline_small_AA_cubic
> 9M 1 515µs 517µs 517µs 518µs 0% ▇█▆▅▃▃▁▁▁▅ 8888 path_hairline_small_AA_conic
> 9M 1 310µs 311µs 315µs 332µs 2% ▂▁█▆▁▁▁▁▁▁ 8888 path_hairline_small_AA_quad
> 9M 1 254µs 254µs 258µs 276µs 3% ▁▁▁▁▁▁▁█▇▂ 8888 path_hairline_small_AA_line
>
> Edited revert of https://codereview.chromium.org/1085013003
>
> TBR=
>
> Expectations:
> - no effect on chrome, as it is guarded
> - lots of tiny rebaselines for skia
TBR=
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
Review URL: https://codereview.chromium.org/1092483003
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
/skia/trunk> cat ../old.txt
maxrss loops min median mean max stddev samples config bench
9M 1 4.28ms 4.32ms 4.36ms 4.67ms 3% ▄▁▁▃▂▂▁▁▂█ 8888 path_hairline_small_AA_cubic
9M 1 743µs 767µs 770µs 825µs 4% ▃▃▇▃▁▁▅▁█▁ 8888 path_hairline_small_AA_conic
9M 1 533µs 606µs 598µs 680µs 9% ▁▂▂█▆▇▇▄▂▂ 8888 path_hairline_small_AA_quad
9M 1 451µs 452µs 456µs 495µs 3% ▁▁▁▁█▁▁▁▁▁ 8888 path_hairline_small_AA_line
/skia/trunk> cat ../new.txt
maxrss loops min median mean max stddev samples config bench
9M 1 827µs 827µs 831µs 869µs 2% ▁▁▁▁▁▁▁█▁▁ 8888 path_hairline_small_AA_cubic
9M 1 515µs 517µs 517µs 518µs 0% ▇█▆▅▃▃▁▁▁▅ 8888 path_hairline_small_AA_conic
9M 1 310µs 311µs 315µs 332µs 2% ▂▁█▆▁▁▁▁▁▁ 8888 path_hairline_small_AA_quad
9M 1 254µs 254µs 258µs 276µs 3% ▁▁▁▁▁▁▁█▇▂ 8888 path_hairline_small_AA_line
Edited revert of https://codereview.chromium.org/1085013003
TBR=
Review URL: https://codereview.chromium.org/1078413003
|
|
|
|
|
|
|
|
|
|
| |
Broken in https://chromium.googlesource.com/skia/+/76033be
I have confirmed that this fixes 470083.
BUG=470083
Review URL: https://codereview.chromium.org/1080923004
|
|
|
|
|
|
|
| |
BUG=skia:
TBR=
Review URL: https://codereview.chromium.org/1088143002
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
flag SK_SUPPORT_LEGACY_BLITANTIH2V2 (patchset #2 id:20001 of https://codereview.chromium.org/1060153003/)
Reason for revert:
DEPS roll is failing, one such example:
http://build.chromium.org/p/tryserver.chromium.win/builders/win8_chromium_rel/builds/73847/steps/ash_unittests%20%28with%20patch%29/logs/stdio
More here:
https://codereview.chromium.org/1091483003
This CL was the only CL in the roll when things started to fail.
Original issue's description:
> add new blit2 methods in support of antialiased hairlines
>
> before:
> 9M 1 528µs 530µs 539µs 607µs 5% ▁▁▁▁▁▁▁▁█▂ 8888 path_hairline_small_AA_quad
>
> after:
> 9M 1 355µs 356µs 358µs 375µs 2% ▂▁▁▁▁▁▁▁▁█ 8888 path_hairline_small_AA_quad
>
> BUG=skia:
>
> does require new baselines (bug chrome is guarded)
>
> Committed: https://skia.googlesource.com/skia/+/dd83031b98db4c6d3d0de2353bf115152a7d1464
TBR=caryclark@google.com,reed@chromium.org,reed@google.com
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=skia:
Review URL: https://codereview.chromium.org/1085013003
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
before:
9M 1 528µs 530µs 539µs 607µs 5% ▁▁▁▁▁▁▁▁█▂ 8888 path_hairline_small_AA_quad
after:
9M 1 355µs 356µs 358µs 375µs 2% ▂▁▁▁▁▁▁▁▁█ 8888 path_hairline_small_AA_quad
BUG=skia:
does require new baselines (bug chrome is guarded)
Review URL: https://codereview.chromium.org/1060153003
|
|
|
|
|
|
|
|
|
|
| |
These will underly the SkPMFloat-like class for uint16_t components.
Sk4h will back a single-pixel version, and Sk8h any larger number than that.
BUG=skia:
Review URL: https://codereview.chromium.org/1088883005
|
|
|
|
|
|
| |
BUG=skia:
Review URL: https://codereview.chromium.org/1076593002
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
As used today, SkNi is used in bool-y contexts. This keeps that, but under a
new name, SkNb. This makes room for a new SkNi that's focused on integer-y
things like loads, stores, arithmetic, etc.
The main reason to split these is that we want different specializations for
each use case: for bools, it's important for us to specialize 32- and 64-bit to
support efficient float- and double- comparisons, but for integer work we're
more likely to be looking at 8- and 16- bit lanes. Keeping these use cases
siloed helps me manage the compexity of the backend NEON and SSE code.
BUG=skia:
Review URL: https://codereview.chromium.org/1083123002
|
|
|
|
|
|
| |
BUG=skia:
Review URL: https://codereview.chromium.org/1085883002
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
When an SkRecord has more than kInlineRecords ops in it (today, 5 or more), the
current logic undercounts the bytes used by the SkRecord by sizeof(Record) *
kInlineRecords, i.e. 32 bytes. This isn't a huge deal... by the time you've
recorded 5 ops, we're typically up around 1KB anyway, and it's only ever off by
that constant 32 bytes, so somewhere between 3% to 0% error as the picture grows.
But now that I've noticed, we might as well fix it. Basically, this is a
reminder that the inline space used to store those first kInlineRecords ops
goes to waste once we pass that threshold. In contrast, fInlineAlloc, the
space we preallocate for the SkVarAlloc, never goes to waste. It always holds
the first few ops' data even when we grow past it.
BUG=skia:
Review URL: https://codereview.chromium.org/1081433002
|
|
|
|
|
|
|
|
|
|
| |
10% speedup for clipped haircubics
BUG=skia:
TBR=
NOTRY=True
Review URL: https://codereview.chromium.org/1079343002
|
|
|
|
|
|
|
| |
BUG=skia:
TBR=
Review URL: https://codereview.chromium.org/1077353002
|
|
|
|
|
|
|
|
|
| |
precursor for https://codereview.chromium.org/1074313002/
BUG=skia:
TBR=
Review URL: https://codereview.chromium.org/1084433002
|
|
|
|
|
|
| |
Missed this in https://codereview.chromium.org/1072603002/ (Add GM to repro crbug.com/472795)
Review URL: https://codereview.chromium.org/1078113002
|