| Commit message (Collapse) | Author | Age |
|
|
|
|
|
|
|
|
| |
This just cleans up an unused half->float entry point.
Change-Id: I7b869d3fd049d807453745c3cca4eab1162848d8
Reviewed-on: https://skia-review.googlesource.com/19451
Reviewed-by: Mike Klein <mtklein@chromium.org>
Commit-Queue: Mike Klein <mtklein@chromium.org>
|
|
|
|
|
|
|
|
|
|
|
|
| |
This refactors from_half() and to_half() a bit, totally
reimplementing the non-hardware cases to be more clearly correct.
CQ_INCLUDE_TRYBOTS=skia.primary:Test-Android-Clang-PixelC-CPU-TegraX1-arm64-Release-Android,Test-Android-Clang-Ci20-CPU-IngenicJZ4780-mipsel-Release-Android,Test-Android-Clang-Nexus10-CPU-Exynos5250-arm-Release-Android,Test-Mac-Clang-MacMini6.2-CPU-AVX-x86_64-Release,Test-Ubuntu-GCC-GCE-CPU-AVX2-x86-Debug,Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Debug
Change-Id: I439463cf90935c5e8fe2369cbcf45e07f3af62c7
Reviewed-on: https://skia-review.googlesource.com/13921
Commit-Queue: Mike Klein <mtklein@chromium.org>
Reviewed-by: Matt Sarett <msarett@google.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
We can't realistically use AVX and SkNx together because of ODR
problems, so remove the code that may tempt us to try.
Remaining code paths using AVX:
- one intrinsics-only routine in SkOpts_hsw.cpp
- SkJumper
Change-Id: I0d2d03b47ea4a0eec27f2de2b28a4c3d1ff8376f
Reviewed-on: https://skia-review.googlesource.com/13121
Reviewed-by: Herb Derby <herb@google.com>
Commit-Queue: Mike Klein <mtklein@chromium.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
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>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The ARMv8 asm path actually does it right... that should be okay.
My Nexus 5x fails `dm -m _finite_ftz` before this and passes after it.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2276533002
TBR=msarett@google.com
Review-Url: https://codereview.chromium.org/2276533002
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
I think we convinced ourselves that denorms, while a good chunk of half floats,
cover a rather small fraction of the representable range, which is always
close enough to zero to flush.
This makes both paths of the conversion to or from float considerably simpler.
These functions now work for zero-or-normal half floats (excluding infinite, NaN).
I'm not aware of a term for this class so I've called them "ordinary".
A handful of GMs and SKPs draw differently in --config f16, but all imperceptibly.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2256023002
Review-Url: https://codereview.chromium.org/2256023002
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Should feel very similar to Sk4h_store4:
NEON uses its native instruction, SSE unpacks manually.
Since we'll have our F16s in 4 Sk4h by the time we're done here,
this also extracts an Sk4h->Sk4f routine from the old uint64_t->Sk4f one.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2184753002
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/2184753002
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This should give us a good baseline to explore using SkRasterPipeline.
A particular colorxform to half float drops from 425us to 282us on my desktop.
Color Xform to Half Float (HP z620)
Original 425us
Trans16 (not 32) 355us
Vector Trans16 378us
Trans16 + Keep Halfs in Vector 335us
Vector Trans16 + Keep Halfs in Vector 282us
Final 282us
Color Xform to Half Float (Nexus 5X)
Original 556us
Final 472us
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2159993003
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/2159993003
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
It's become clear we need to sometimes deal with values <0 or >1.
I'm not yet convinced we care about NaN or +-inf.
We had some fairly clever tricks and optimizations here for NEON
and SSE. I've thrown them out in favor of a single implementation.
If we find the specializations mattered, we can certainly figure out
how to extend them to this new range/domain.
This happens to add a vectorized float -> half for ARMv7, which was
missing from the _01 version. (The SSE strategy was not portable to
platforms that flush denorm floats to zero.)
I've tested the full float range for FloatToHalf on my desktop and a 5x.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2145663003
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,Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-Fast-Trybot
Committed: https://skia.googlesource.com/skia/+/3296bee70d074bb8094b3229dbe12fa016657e90
Review-Url: https://codereview.chromium.org/2145663003
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
(patchset #7 id:120001 of https://codereview.chromium.org/2145663003/ )
Reason for revert:
Unit tests fail on Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-Fast
Original issue's description:
> Expand _01 half<->float limitation to _finite. Simplify.
>
> It's become clear we need to sometimes deal with values <0 or >1.
> I'm not yet convinced we care about NaN or +-inf.
>
> We had some fairly clever tricks and optimizations here for NEON
> and SSE. I've thrown them out in favor of a single implementation.
> If we find the specializations mattered, we can certainly figure out
> how to extend them to this new range/domain.
>
> This happens to add a vectorized float -> half for ARMv7, which was
> missing from the _01 version. (The SSE strategy was not portable to
> platforms that flush denorm floats to zero.)
>
> I've tested the full float range for FloatToHalf on my desktop and a 5x.
>
> BUG=skia:
> GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2145663003
> 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,Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-Fast-Trybot
>
> Committed: https://skia.googlesource.com/skia/+/3296bee70d074bb8094b3229dbe12fa016657e90
TBR=msarett@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/2151023003
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
It's become clear we need to sometimes deal with values <0 or >1.
I'm not yet convinced we care about NaN or +-inf.
We had some fairly clever tricks and optimizations here for NEON
and SSE. I've thrown them out in favor of a single implementation.
If we find the specializations mattered, we can certainly figure out
how to extend them to this new range/domain.
This happens to add a vectorized float -> half for ARMv7, which was
missing from the _01 version. (The SSE strategy was not portable to
platforms that flush denorm floats to zero.)
I've tested the full float range for FloatToHalf on my desktop and a 5x.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2145663003
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/2145663003
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This removes the code paths where we make SkCpu::Supports() calls
from within a tight loop. It keeps code paths using SkCpu::Supports()
to choose entire routines from src/opts/.
We can't rely on these hyper-local checks to be hoisted up reliably enough.
It worked pretty well with the first couple platforms we tried (e.g. Clang
on Linux/Mac) but we can't gaurantee it works everywhere.
Further, I'm not able to actually do anything fancy with those tests
outside of x86... I've not found a way to get, say, NEON+F16 conversion
code embedded into ordinary NEON code outside writing then entire function
in external assembly.
This whole idea becomes less important now that we've got a way to chain
separate function calls together efficiently. We can now, e.g., use an
AVX+F16C method to load some pixels, then chain that into an ordinary AVX
method to color filter them.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2138073002
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
Review-Url: https://codereview.chromium.org/2138073002
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
- floor with roundps is about 4.5x faster when available
- f16 srcover_n is similar to but a little faster than the version in https://codereview.chromium.org/1884683002. This new one fuses the dst load/stores into the f16<->f32 conversions:
+0x180 movups (%r15), %xmm1
+0x184 vcvtph2ps (%rbx), %xmm2
+0x189 movaps %xmm1, %xmm3
+0x18c shufps $255, %xmm3, %xmm3
+0x190 movaps %xmm0, %xmm4
+0x193 subps %xmm3, %xmm4
+0x196 mulps %xmm2, %xmm4
+0x199 addps %xmm1, %xmm4
+0x19c vcvtps2ph $0, %xmm4, (%rbx)
+0x1a2 addq $16, %r15
+0x1a6 addq $8, %rbx
+0x1aa decl %r14d
+0x1ad jne +0x180
If we decide to land this it'd be a good idea to convert most or all users of SkFloatToHalf_01 and SkHalfToFloat_01 over to the pointer-based versions.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1891513002
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
Committed: https://skia.googlesource.com/skia/+/cbe3c1af987d622ea67ef560d855b41bb14a0ce9
Committed: https://skia.googlesource.com/skia/+/3faf74b8364491ca806f523fbb1d8a97be592663
Review URL: https://codereview.chromium.org/1891513002
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
https://codereview.chromium.org/1891513002/ )
Reason for revert:
this depends on a CL I want to revert
Original issue's description:
> skcpu: sse4.1 floor, f16c f16<->f32
>
> - floor with roundps is about 4.5x faster when available
> - f16 srcover_n is similar to but a little faster than the version in https://codereview.chromium.org/1884683002. This new one fuses the dst load/stores into the f16<->f32 conversions:
>
> +0x180 movups (%r15), %xmm1
> +0x184 vcvtph2ps (%rbx), %xmm2
> +0x189 movaps %xmm1, %xmm3
> +0x18c shufps $255, %xmm3, %xmm3
> +0x190 movaps %xmm0, %xmm4
> +0x193 subps %xmm3, %xmm4
> +0x196 mulps %xmm2, %xmm4
> +0x199 addps %xmm1, %xmm4
> +0x19c vcvtps2ph $0, %xmm4, (%rbx)
> +0x1a2 addq $16, %r15
> +0x1a6 addq $8, %rbx
> +0x1aa decl %r14d
> +0x1ad jne +0x180
>
> If we decide to land this it'd be a good idea to convert most or all users of SkFloatToHalf_01 and SkHalfToFloat_01 over to the pointer-based versions.
>
> BUG=skia:
> GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1891513002
> CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
>
> Committed: https://skia.googlesource.com/skia/+/cbe3c1af987d622ea67ef560d855b41bb14a0ce9
>
> Committed: https://skia.googlesource.com/skia/+/3faf74b8364491ca806f523fbb1d8a97be592663
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/1897433002
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
- floor with roundps is about 4.5x faster when available
- f16 srcover_n is similar to but a little faster than the version in https://codereview.chromium.org/1884683002. This new one fuses the dst load/stores into the f16<->f32 conversions:
+0x180 movups (%r15), %xmm1
+0x184 vcvtph2ps (%rbx), %xmm2
+0x189 movaps %xmm1, %xmm3
+0x18c shufps $255, %xmm3, %xmm3
+0x190 movaps %xmm0, %xmm4
+0x193 subps %xmm3, %xmm4
+0x196 mulps %xmm2, %xmm4
+0x199 addps %xmm1, %xmm4
+0x19c vcvtps2ph $0, %xmm4, (%rbx)
+0x1a2 addq $16, %r15
+0x1a6 addq $8, %rbx
+0x1aa decl %r14d
+0x1ad jne +0x180
If we decide to land this it'd be a good idea to convert most or all users of SkFloatToHalf_01 and SkHalfToFloat_01 over to the pointer-based versions.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1891513002
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
Committed: https://skia.googlesource.com/skia/+/cbe3c1af987d622ea67ef560d855b41bb14a0ce9
Review URL: https://codereview.chromium.org/1891513002
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
https://codereview.chromium.org/1891513002/ )
Reason for revert:
Need to change around my #if guards so that clang-cl is treated like GCC and Clang, rather than MSVC.
Original issue's description:
> skcpu: sse4.1 floor, f16c f16<->f32
>
> - floor with roundps is about 4.5x faster when available
> - f16 srcover_n is similar to but a little faster than the version in https://codereview.chromium.org/1884683002. This new one fuses the dst load/stores into the f16<->f32 conversions:
>
> +0x180 movups (%r15), %xmm1
> +0x184 vcvtph2ps (%rbx), %xmm2
> +0x189 movaps %xmm1, %xmm3
> +0x18c shufps $255, %xmm3, %xmm3
> +0x190 movaps %xmm0, %xmm4
> +0x193 subps %xmm3, %xmm4
> +0x196 mulps %xmm2, %xmm4
> +0x199 addps %xmm1, %xmm4
> +0x19c vcvtps2ph $0, %xmm4, (%rbx)
> +0x1a2 addq $16, %r15
> +0x1a6 addq $8, %rbx
> +0x1aa decl %r14d
> +0x1ad jne +0x180
>
> If we decide to land this it'd be a good idea to convert most or all users of SkFloatToHalf_01 and SkHalfToFloat_01 over to the pointer-based versions.
>
> BUG=skia:
> GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1891513002
> CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
>
> Committed: https://skia.googlesource.com/skia/+/cbe3c1af987d622ea67ef560d855b41bb14a0ce9
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/1891993002
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
- floor with roundps is about 4.5x faster when available
- f16 srcover_n is similar to but a little faster than the version in https://codereview.chromium.org/1884683002. This new one fuses the dst load/stores into the f16<->f32 conversions:
+0x180 movups (%r15), %xmm1
+0x184 vcvtph2ps (%rbx), %xmm2
+0x189 movaps %xmm1, %xmm3
+0x18c shufps $255, %xmm3, %xmm3
+0x190 movaps %xmm0, %xmm4
+0x193 subps %xmm3, %xmm4
+0x196 mulps %xmm2, %xmm4
+0x199 addps %xmm1, %xmm4
+0x19c vcvtps2ph $0, %xmm4, (%rbx)
+0x1a2 addq $16, %r15
+0x1a6 addq $8, %rbx
+0x1aa decl %r14d
+0x1ad jne +0x180
If we decide to land this it'd be a good idea to convert most or all users of SkFloatToHalf_01 and SkHalfToFloat_01 over to the pointer-based versions.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1891513002
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
Review URL: https://codereview.chromium.org/1891513002
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Adds f32 <-> f16 ARMv7 and ARMv8 NEON code.
Also adds NEON f32 <-> u16 code to make the comparison fair.
The NDK GCC does not support the ARMv8 NEON intrinsics needed to go fastest, so we use a tiny amount of inline assembly.
The ARMv7 half -> float is different enough from the SSE version that it does not make sense to use SkNx.
Still TODO:
ARMv7 float -> half. Naively translating the SSE version results in 0x0000 where we'd expect a denormal output.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1700473003
CQ_EXTRA_TRYBOTS=client.skia.android:Test-Android-GCC-Nexus5-CPU-NEON-Arm7-Release-Trybot,Test-Android-GCC-Nexus9-CPU-Denver-Arm64-Release-Trybot;client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
Review URL: https://codereview.chromium.org/1700473003
|
|
|
|
|
|
|
|
|
| |
This is a little faster than the previous version, and much better explained.
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1688233002
Review URL: https://codereview.chromium.org/1688233002
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
These are basically inlined, 4-at-a-time versions of our existing functions,
but cut down to avoid any work that's only necessary outside [0,1].
Both f16 and f32 denorms should work fine modulo the usual ARMv7 NEON denorm==zero caveat.
In exchange for a little speed, f32->f16 does not round properly.
Instead it truncates, so it's never off by more than 1 bit.
Support for finite values >1 or <0 is straightforward to add back.
>1 might already work as-is.
Getting close to _u16 performance:
micros bench
261.13 xferu64_bw_1_opaque_u16
1833.51 xferu64_bw_1_alpha_u16
2762.32 ? xferu64_aa_1_opaque_u16
3334.29 xferu64_aa_1_alpha_u16
249.78 xferu64_bw_1_opaque_f16
3383.18 xferu64_bw_1_alpha_f16
4214.72 xferu64_aa_1_opaque_f16
4701.19 xferu64_aa_1_alpha_f16
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1685133005
Committed: https://skia.googlesource.com/skia/+/9ea11a4235b3e3521cc8bf914a27c2d0dc062db9
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
Review URL: https://codereview.chromium.org/1685133005
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
https://codereview.chromium.org/1685133005/ )
Reason for revert:
Gotta fix Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD
Original issue's description:
> SkHalfToFloat_01 / SkFloatToHalf_01
>
> These are basically inlined, 4-at-a-time versions of our existing functions,
> but cut down to avoid any work that's only necessary outside [0,1].
>
> Both f16 and f32 denorms should work fine modulo the usual ARMv7 NEON denorm==zero caveat.
>
> In exchange for a little speed, f32->f16 does not round properly.
> Instead it truncates, so it's never off by more than 1 bit.
>
> Support for finite values >1 or <0 is straightforward to add back.
> >1 might already work as-is.
>
> Getting close to _u16 performance:
> micros bench
> 261.13 xferu64_bw_1_opaque_u16
> 1833.51 xferu64_bw_1_alpha_u16
> 2762.32 ? xferu64_aa_1_opaque_u16
> 3334.29 xferu64_aa_1_alpha_u16
> 249.78 xferu64_bw_1_opaque_f16
> 3383.18 xferu64_bw_1_alpha_f16
> 4214.72 xferu64_aa_1_opaque_f16
> 4701.19 xferu64_aa_1_alpha_f16
>
>
> BUG=skia:
> GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1685133005
>
> Committed: https://skia.googlesource.com/skia/+/9ea11a4235b3e3521cc8bf914a27c2d0dc062db9
TBR=jvanverth@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/1693443003
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
These are basically inlined, 4-at-a-time versions of our existing functions,
but cut down to avoid any work that's only necessary outside [0,1].
Both f16 and f32 denorms should work fine modulo the usual ARMv7 NEON denorm==zero caveat.
In exchange for a little speed, f32->f16 does not round properly.
Instead it truncates, so it's never off by more than 1 bit.
Support for finite values >1 or <0 is straightforward to add back.
>1 might already work as-is.
Getting close to _u16 performance:
micros bench
261.13 xferu64_bw_1_opaque_u16
1833.51 xferu64_bw_1_alpha_u16
2762.32 ? xferu64_aa_1_opaque_u16
3334.29 xferu64_aa_1_alpha_u16
249.78 xferu64_bw_1_opaque_f16
3383.18 xferu64_bw_1_alpha_f16
4214.72 xferu64_aa_1_opaque_f16
4701.19 xferu64_aa_1_alpha_f16
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1685133005
Review URL: https://codereview.chromium.org/1685133005
|
|
|
|
|
|
|
|
|
| |
This allows us to create distance field textures with better precision,
which may help text quality.
BUG=skia:3103
Review URL: https://codereview.chromium.org/762923003
|
|
Based on code by Fabian Giesen at
https://fgiesen.wordpress.com/2012/03/28/half-to-float-done-quic/.
These will be needed for creating binary16 textures from floating point data.
BUG=skia:3103
Review URL: https://codereview.chromium.org/760753003
|