aboutsummaryrefslogtreecommitdiffhomepage
path: root/src
Commit message (Collapse)AuthorAge
* Remove outdated optimization to not flush for new textures in GrSWMaskHelperGravatar bsalomon2016-02-16
| | | | | | | BUG=skia: GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1704593002 Review URL: https://codereview.chromium.org/1704593002
* Workaround for bug in Tegra 3 when uploading to a render targetGravatar bsalomon2016-02-16
| | | | | | | BUG=skia: GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1703603002 Review URL: https://codereview.chromium.org/1703603002
* Delete GrTextContextGravatar joshualitt2016-02-16
| | | | | | | | TBR=bsalomon@google.com BUG=skia: GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1699073004 Review URL: https://codereview.chromium.org/1699073004
* Create GrPathRenderingDrawContextGravatar joshualitt2016-02-16
| | | | | | | | TBR=bsalomon@google.com BUG=skia: GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1701013002 Review URL: https://codereview.chromium.org/1701013002
* Change SkMatrix::fixedStepInX to return SkVector (of SkScalar) rather than ↵Gravatar benjaminwagner2016-02-16
| | | | | | | | | | | | | SkFixed. All users were immediately converting to SkScalar or SkFixed3232. This method is not used in Chromium, Android, or Google3. BUG=skia:4632 GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1692013002 Review URL: https://codereview.chromium.org/1692013002
* Image filters: change applyCropRect() to take a src rect.Gravatar senorblanco2016-02-16
| | | | | | | | | | Instead of taking the source bitmap and offset, we simply take the source rect bounds, and make the caller responsible for computing it. BUG=skia: GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1702683002 Review URL: https://codereview.chromium.org/1702683002
* SkOSFile_stdio: less ::Gravatar mtklein2016-02-16
| | | | | | | | | | | | It's not clear why we wrote this file in this manner. On OpenBSD, fileno is a macro, so ::fileno makes no sense. Also remove some end-of-line whitespace. BUG=skia:4953 GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1695333003 Review URL: https://codereview.chromium.org/1695333003
* Make png filter functions compatible with libpngGravatar msarett2016-02-16
| | | | | | | | | | | | We hope to upstream these functions to libpng (a C library). Let's make the code that Skia tests on look like the code that we are submitting. BUG=skia: GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1699953002 Review URL: https://codereview.chromium.org/1699953002
* Modify interface to GrPathUtils::convertCubicToQuadsGravatar bsalomon2016-02-16
| | | | | | | BUG=skia: GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1694403003 Review URL: https://codereview.chromium.org/1694403003
* Make copySurface work for texture dsts, return a bool, & add unit test.Gravatar bsalomon2016-02-16
| | | | | | | | GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1684313002 Committed: https://skia.googlesource.com/skia/+/7ea5e28065e5eb797e95f5d81c1a65cf3209d741 Review URL: https://codereview.chromium.org/1684313002
* Make SkRWBuffer destruct safelyGravatar scroggo2016-02-16
| | | | | | | | | | | Check for NULL before calling unref. Add a test. BUG=483369 GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1691383002 Review URL: https://codereview.chromium.org/1691383002
* Remove last bit of privacy violation for GrAtlasTextBlobGravatar joshualitt2016-02-16
| | | | | | | BUG=skia: GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1698503002 Review URL: https://codereview.chromium.org/1698503002
* starter kit for colorspacesGravatar reed2016-02-15
| | | | | | | | | BUG=skia: GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1692893003 TBR=jvanverth Review URL: https://codereview.chromium.org/1695353002
* Fix colorType/alphaType checks in SkCodecGravatar msarett2016-02-15
| | | | | | | | | | | | | | | | Make getPixels() and startScanlineDecode() behave consistently. Require that kGray8 decodes are opaque. Assert that creating the swizzler succeeds. BUG=skia:4203 GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1695473002 Committed: https://skia.googlesource.com/skia/+/c7578b6cdd03b61f076ffc7956efd952d6c301c0 Review URL: https://codereview.chromium.org/1695473002
* make SkComposeShader.h privateGravatar reed2016-02-15
| | | | | | | BUG=skia: GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1697523003 Review URL: https://codereview.chromium.org/1697523003
* blitters for sRGB and float16Gravatar reed2016-02-15
| | | | | | | BUG=skia: GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1697863002 Review URL: https://codereview.chromium.org/1697863002
* Roll back ellipse logic for now. This was (probably) still wrong.Gravatar brianosman2016-02-14
| | | | | | | BUG=skia: GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1692303003 Review URL: https://codereview.chromium.org/1692303003
* Revert of Add infastructure for gl_SampleMask (patchset #6 id:100001 of ↵Gravatar benjaminwagner2016-02-12
| | | | | | | | | | | | | | | | | | | | | | | | https://codereview.chromium.org/1690963003/ ) Reason for revert: Several bots failing with error 'invalid profile "es" for version 320'. Original issue's description: > Add infastructure for gl_SampleMask > > BUG=skia: > GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1690963003 > > Committed: https://skia.googlesource.com/skia/+/533cefe5b9c7cec2592fc7ca00ee4cf69a26c094 TBR=egdaniel@google.com,bsalomon@google.com,cdalton@nvidia.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/1693113002
* Revert of Fix colorType/alphaType checks in SkCodec (patchset #5 id:80001 of ↵Gravatar msarett2016-02-12
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | https://codereview.chromium.org/1695473002/ ) Reason for revert: Really bad images in Gold. Original issue's description: > Fix colorType/alphaType checks in SkCodec > > Make getPixels() and startScanlineDecode() behave > consistently. > > Require that kGray8 decodes are opaque. > > Assert that creating the swizzler succeeds. > > BUG=skia:4203 > GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1695473002 > > Committed: https://skia.googlesource.com/skia/+/c7578b6cdd03b61f076ffc7956efd952d6c301c0 TBR=scroggo@google.com # Skipping CQ checks because original CL landed less than 1 days ago. NOPRESUBMIT=true NOTREECHECKS=true NOTRY=true BUG=skia:4203 Review URL: https://codereview.chromium.org/1694023002
* Revert of SkNx refactoring (patchset #4 id:60001 of ↵Gravatar mtklein2016-02-12
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | https://codereview.chromium.org/1690633003/ ) Reason for revert: Precautionary revert for chromium:586487 Original issue's description: > SkNx refactoring > > - add back Sk4i typedef > - define SSE casts in terms of Sk4i > * uint8 <-> float becomes uint8 <-> int <-> float > * uint16 <-> float becomes uint16 <-> int <-> float > > This has the nice side effect of specializing uint8 <-> int > and uint16 <-> int, which are useful in their own right. > > There are many cast specializations now, some of which call each other. > I have tried to arrange them in some sort of sensible order, subject to > the constraint that those called must precede those who call. > > BUG=skia: > GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1690633003 > CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot > > Committed: https://skia.googlesource.com/skia/+/c1eb311f4e98934476f1b2ad5d6de772cf140d60 TBR=herb@google.com,mtklein@chromium.org # Not skipping CQ checks because original CL landed more than 1 days ago. BUG=chromium:586487 Review URL: https://codereview.chromium.org/1696903002
* Try fixing unused-variable warnings another way.Gravatar mtklein2016-02-12
| | | | | | | | | | | | | | If we mark these as extern, the compiler won't complain they're unused. This is an unsatisfying hack, but last time we tried to condition these on non-NEON only, Chromium's GN linux ARM bot failed. They must not be configured to use the NEON procs? BUG=skia:4798 GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1699473002 CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot Review URL: https://codereview.chromium.org/1699473002
* Add infastructure for gl_SampleMaskGravatar cdalton2016-02-12
| | | | | | | BUG=skia: GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1690963003 Review URL: https://codereview.chromium.org/1690963003
* Fix colorType/alphaType checks in SkCodecGravatar msarett2016-02-12
| | | | | | | | | | | | | | Make getPixels() and startScanlineDecode() behave consistently. Require that kGray8 decodes are opaque. Assert that creating the swizzler succeeds. BUG=skia:4203 GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1695473002 Review URL: https://codereview.chromium.org/1695473002
* Add support for caching font files in the Android SkFontMgr.Gravatar khushalsagar2016-02-12
| | | | | | | | | | | | | SkFontMgr_Android lazily access font files from disk, which is not possible when it is used in the renderer sandbox on Linux. Add a flag to SkFontMgr_Android_CustomFonts for caching readonly FILE streams when creating the font mgr. Since the font mgr is created before the sandbox is initialized, it can access these files on initialization, and use these cached streams safely for its lifetime. Review URL: https://codereview.chromium.org/1673373003
* Make SkJpegCodec compatible with libjpegGravatar msarett2016-02-12
| | | | | | | | BUG=skia:4470 BUG=skia:4520 GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1687863003 Review URL: https://codereview.chromium.org/1687863003
* Change ScalarTo256 to more efficient implementation.Gravatar benjaminwagner2016-02-12
| | | | | | | | | The previous implementation was likely more efficient when SkScalar was SkFixed. BUG=skia:4632 GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1693683002 Review URL: https://codereview.chromium.org/1693683002
* Use noperspective interpolation for 2D drawsGravatar cdalton2016-02-12
| | | | | | | | | | | | | | | | | | | | | | | | | | Adds a mechanism to notify GrGLSLVaryingHandler that a shader will not emit geometry in perspective. This gives it a chance to use the noperspective keyword if it is supported. Updates the existing processors to notify the varying handler when there is no perspective. Begins using the noperspective keyword in GrGLGpu internal shaders. Web scenes with observable benefit (Pixel C, gpu config): tabl_nofolo.skp 4.62 -> 3.33 ms 28% desk_tigersvg.skp 26.5 -> 24 ms 9% desk_pokemonwiki.skp 16.1 -> 14.9 ms 7% tabl_deviantart.skp 3.97 -> 3.7 ms 7% desk_gws.skp 3.85 -> 3.65 ms 5% Also adds new methods for creating varyings with flat interpolation. BUG=skia: GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1673093002 Review URL: https://codereview.chromium.org/1673093002
* Make copySurface work in more situations.Gravatar bsalomon2016-02-12
| | | | | | | BUG=skia: GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1693923002 Review URL: https://codereview.chromium.org/1693923002
* Put back the 0.5 bias in the shader. Only adjust the radii for the geometry, ↵Gravatar brianosman2016-02-12
| | | | | | | | | | | not the radii constants. After working through this with Jim, this is more mathematically correct, and also accounts for the inner radius. BUG=skia: GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1692683003 Review URL: https://codereview.chromium.org/1692683003
* Remove the 0.5 bias from the ellipse edge function. This makes coverage drop ↵Gravatar brianosman2016-02-12
| | | | | | | | | to zero before we get clipped by the rasterizer (i.e. ensures the generated ellipse is inscribed in the geometry). BUG=skia: GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1683843006 Review URL: https://codereview.chromium.org/1683843006
* Delete SkDecodingImageGeneratorGravatar msarett2016-02-11
| | | | | | | BUG=skia: GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1692053002 Review URL: https://codereview.chromium.org/1692053002
* Switch to vertex colors rather than uniforms for color in ↵Gravatar brianosman2016-02-11
| | | | | | | | | | | | | | Ellipse/Circle/RRect/DRRect [Notes:] - Performance delta on desktop looked like noise. I'm curious what the results will be on low-end mobile GPUs. I'm still trying to figure out how I determine that with perf, etc... - There appeared to be a single one-pixel image diff while running dm, if I'm understanding the results of skdiff correctly, so that's probably good news? - Should I include anyone else on the review? BUG=skia: GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1675833002 Review URL: https://codereview.chromium.org/1675833002
* Add #define directives to GrGLSLShaderBuilderGravatar cdalton2016-02-11
| | | | | | | BUG=skia: GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1660813002 Review URL: https://codereview.chromium.org/1660813002
* Implement onSkipScanlines() for bmp and wbmpGravatar msarett2016-02-11
| | | | | | | | TBR=reed@google.com BUG=skia:4270 GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1691083002 Review URL: https://codereview.chromium.org/1691083002
* Add GrShaderFlags enumGravatar cdalton2016-02-11
| | | | | | | | | Replaces GrGLSLUniformHandler::ShaderVisibility with GrShaderFlags. BUG=skia: GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1684063006 Review URL: https://codereview.chromium.org/1684063006
* new version of SkHalfToFloat_01Gravatar mtklein2016-02-11
| | | | | | | | | 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
* start to chip away at friending of GrAtlasTextBatch/GrAtlasTextBlobGravatar joshualitt2016-02-11
| | | | | | | BUG=skia: GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1684253002 Review URL: https://codereview.chromium.org/1684253002
* Revert of Make copySurface work for texture dsts, return a bool, & add unit ↵Gravatar kjlubick2016-02-11
| | | | | | | | | | | | | | | | | | | | | | | test. (patchset #6 id:100001 of https://codereview.chromium.org/1684313002/ ) Reason for revert: Copy surface tests are not happy for Windows, Android and probably others: https://uberchromegw.corp.google.com/i/client.skia.android/builders/Test-Android-GCC-Nexus7-GPU-Tegra3-Arm7-Release/builds/4161 https://uberchromegw.corp.google.com/i/client.skia/builders/Test-Win8-MSVC-ShuttleA-GPU-GTX960-x86_64-Release/builds/3694 Original issue's description: > Make copySurface work for texture dsts, return a bool, & add unit test. > GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1684313002 > > Committed: https://skia.googlesource.com/skia/+/7ea5e28065e5eb797e95f5d81c1a65cf3209d741 TBR=robertphillips@google.com,bsalomon@google.com # Skipping CQ checks because original CL landed less than 1 days ago. NOPRESUBMIT=true NOTREECHECKS=true NOTRY=true Review URL: https://codereview.chromium.org/1690053002
* Remove GrContext* from GrTextContextGravatar joshualitt2016-02-11
| | | | | | | BUG=skia: GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1686143003 Review URL: https://codereview.chromium.org/1686143003
* Make copySurface work for texture dsts, return a bool, & add unit test.Gravatar bsalomon2016-02-11
| | | | | | GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1684313002 Review URL: https://codereview.chromium.org/1684313002
* Support more color types for ICOsGravatar msarett2016-02-11
| | | | | | | BUG=skia:4620 GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1694513002 Review URL: https://codereview.chromium.org/1694513002
* Avoid initializing memory twice on incomplete RLE bmpsGravatar msarett2016-02-11
| | | | | | | BUG=skia:4582 GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1689983002 Review URL: https://codereview.chromium.org/1689983002
* Clarify bmp comment about when we support the alpha maskGravatar msarett2016-02-11
| | | | | | | BUG=skia: GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1689183002 Review URL: https://codereview.chromium.org/1689183002
* Fix scanline decoding of rare RLE bmpsGravatar msarett2016-02-11
| | | | | | | | | | | | | | | This also exposed a bug in rewinding RLE bmps, which I have also fixed in this CL. This should fix testcase7.bmp on Gold. The image that I am adding to resources is in the public domain. BUG=skia:4730 GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1689953002 Review URL: https://codereview.chromium.org/1689953002
* SkNx refactoringGravatar mtklein2016-02-11
| | | | | | | | | | | | | | | | | | | | - add back Sk4i typedef - define SSE casts in terms of Sk4i * uint8 <-> float becomes uint8 <-> int <-> float * uint16 <-> float becomes uint16 <-> int <-> float This has the nice side effect of specializing uint8 <-> int and uint16 <-> int, which are useful in their own right. There are many cast specializations now, some of which call each other. I have tried to arrange them in some sort of sensible order, subject to the constraint that those called must precede those who call. BUG=skia: GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1690633003 CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot Review URL: https://codereview.chromium.org/1690633003
* Remove SK_SUPPORT_LEGACY_BITMAP_FILTER guardGravatar fmalita2016-02-11
| | | | | | | TBR=reed@google.com GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1695443002 Review URL: https://codereview.chromium.org/1695443002
* SkDocument: remove use of SkTArray (part 1/3).Gravatar halcanary2016-02-11
| | | | | | GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1689683002 Review URL: https://codereview.chromium.org/1689683002
* Pass a GrContext pointer in GrDrawContext constructorGravatar joshualitt2016-02-11
| | | | | | | | TBR=bsalomon@google.com BUG=skia: GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1691503002 Review URL: https://codereview.chromium.org/1691503002
* SkHalfToFloat_01 / SkFloatToHalf_01Gravatar mtklein2016-02-11
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | 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
* Revert of SkHalfToFloat_01 / SkFloatToHalf_01 (patchset #11 id:200001 of ↵Gravatar mtklein2016-02-11
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | 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
generated by cgit v1.2.3 (git 2.39.1) at 2024-06-17 18:51:20 +0000