SkBlurImageFilter, sets it on an SkPaint, passes that paint to saveLayer(),
draws the primitives which are to be blurred, then calls restore(), which
applies the blur. The blurs have separate sizes in the horizontal and vertical
direction. This feature is GPU-only for now.
NB: Due to the clipping change, there are slight pixel differences on the
blurs_gpu and shadows_gpu tests, so those will require rebaselining on all
platforms, as will some of the WebKit layout tests (TBD).
Review URL: http://codereview.appspot.com/5322068/
git-svn-id: http://skia.googlecode.com/svn/trunk@2643 2bbb7eff-a529-9590-31e7-b0007b416f81
we can, because an entire row is conservatively either outside or inside the
gradient.) Change the gradient benchmark to capture both cases, and add new
gm to detect errors in these optimized paths.
git-svn-id: http://skia.googlecode.com/svn/trunk@2327 2bbb7eff-a529-9590-31e7-b0007b416f81
I have manually examined all of these diffs and restored a few files that
seem to require manual adjustment.
The following files still need to be modified manually, in a separate CL:
android_sample/SampleApp/AndroidManifest.xml
android_sample/SampleApp/res/layout/layout.xml
android_sample/SampleApp/res/menu/sample.xml
android_sample/SampleApp/res/values/strings.xml
android_sample/SampleApp/src/com/skia/sampleapp/SampleApp.java
android_sample/SampleApp/src/com/skia/sampleapp/SampleView.java
experimental/CiCarbonSampleMain.c
experimental/CocoaDebugger/main.m
experimental/FileReaderApp/main.m
experimental/SimpleCocoaApp/main.m
experimental/iOSSampleApp/Shared/SkAlertPrompt.h
experimental/iOSSampleApp/Shared/SkAlertPrompt.m
experimental/iOSSampleApp/SkiOSSampleApp-Base.xcconfig
experimental/iOSSampleApp/SkiOSSampleApp-Debug.xcconfig
experimental/iOSSampleApp/SkiOSSampleApp-Release.xcconfig
gpu/src/android/GrGLDefaultInterface_android.cpp
gyp/common.gypi
gyp_skia
include/ports/SkHarfBuzzFont.h
include/views/SkOSWindow_wxwidgets.h
make.bat
make.py
src/opts/memset.arm.S
src/opts/memset16_neon.S
src/opts/memset32_neon.S
src/opts/opts_check_arm.cpp
src/ports/SkDebug_brew.cpp
src/ports/SkMemory_brew.cpp
src/ports/SkOSFile_brew.cpp
src/ports/SkXMLParser_empty.cpp
src/utils/ios/SkImageDecoder_iOS.mm
src/utils/ios/SkOSFile_iOS.mm
src/utils/ios/SkStream_NSData.mm
tests/FillPathTest.cpp
Review URL: http://codereview.appspot.com/4816058
git-svn-id: http://skia.googlecode.com/svn/trunk@1982 2bbb7eff-a529-9590-31e7-b0007b416f81
We could speed-up again if we...
- respected kDither and only built 1/2 of the table for non-dither requests
- output simple params to the gpu rather than always a texture
- detected that we have no alpha, and then can skip premul per-entry
git-svn-id: http://skia.googlecode.com/svn/trunk@1772 2bbb7eff-a529-9590-31e7-b0007b416f81
SkGPipe needs this, since it cannot (unlike SkPicture) see all of the factories
before it hands its data to the reader.
In this mode, the writer embedds the factory name the first time it sees it,
and then after that writes an index (referencing the fFactorySet). The reader
installs an empty array, and as it encounters names, appends them to that array
so that subsequent indices can be used to retrieve the previously named factory.
Some of the existing patheffects did not register their factory names, so those
changes are also part of this CL. Annoyingly, to register your factory using the
current scheme, it has to be in the public section of the class definition.
git-svn-id: http://skia.googlecode.com/svn/trunk@1663 2bbb7eff-a529-9590-31e7-b0007b416f81
calculation for Sk2DPathEffect currently evaluates all the bytes in the
embedded SkMatrix. This includes the type mask, which contains some
uninitialized padding. Changing it to use SkMatrix::flatten() and
SkMatrix::unflatten() (as SkGroupShape was doing) avoids the uninitialized
data errors.
Review URL: http://codereview.appspot.com/4529074/
git-svn-id: http://skia.googlecode.com/svn/trunk@1395 2bbb7eff-a529-9590-31e7-b0007b416f81
Both shader and gradient_shader write matrices to the flatten stream. However, they were
just calling write(&matrix, sizeof(SkMatrix)) and the matrix can contain lazily-computed
function ptrs as part of its internal cache. Thus two matrices that are logically the
same may write different bytes.
This is a problem because picture relies on flattening objects and then using the
flatten stream as a key into its cache. This matrix-write bug effectively kills the
effectiveness of the cache for shaders.
The fix is to write proper read/write functions for matrix (and region btw). These
call through to the existing low-level flatten routines (which just write into a
memory ptr).
git-svn-id: http://skia.googlecode.com/svn/trunk@1290 2bbb7eff-a529-9590-31e7-b0007b416f81
note: gradient caller doesn't so we can still draw wrong when the caller
converts its initial fx from float->fixed. Perhaps SkClampRange should offer
a float interface as well.
git-svn-id: http://skia.googlecode.com/svn/trunk@1149 2bbb7eff-a529-9590-31e7-b0007b416f81
promise that const SkPaint& stay const (so we don't have bugs if a const paint
is referenced from two threads in pictures)
git-svn-id: http://skia.googlecode.com/svn/trunk@1074 2bbb7eff-a529-9590-31e7-b0007b416f81
for another iter that runs back-to-front (B2FIter).
add unittest for Deque
git-svn-id: http://skia.googlecode.com/svn/trunk@821 2bbb7eff-a529-9590-31e7-b0007b416f81
Fixed rendering of shadows under bitmaps.
Added an option to SkBlurMask to produce higher quality blurs.
The HQ blur option is propagated through SkBlurDrawLooper so that it can be invoked by WebKit for drawing shadows.
Added a shadow rendering test to gm.
Bugs fixed: 146, 150
Related Chromium bugs: 11153, 42654
Code review: http://codereview.appspot.com/4174049
git-svn-id: http://skia.googlecode.com/svn/trunk@818 2bbb7eff-a529-9590-31e7-b0007b416f81
SkSafeRef() and SkSafeUnref().
This is basically a bug waiting to happen. An optimizing compiler can remove
checks for null on "this" if it chooses. However, SkRefCnt::safeRef() relies on
precisely this check...
void SkRefCnt::safeRef() {
if (this) {
this->ref();
}
}
Since a compiler might skip the if-clause, it breaks the intention of this
method, hence its removal.
static inline void SkSafeRef(SkRefCnt* obj) {
if (obj) {
obj->ref();
}
}
This form is not ignored by an optimizing compile, so we use it instead.
git-svn-id: http://skia.googlecode.com/svn/trunk@762 2bbb7eff-a529-9590-31e7-b0007b416f81
The regression was due to the fact that we were calling PlatformColorProc() for
every span (which in turns makes CPUID, a fairly expensive call). Since we draw
a lot of rects, and rects have 1-pixel wide spans for the vertical segments,
that's a lot of CPUID.
Fixed by cacheing the result of PlatformColorProc(), as is done for the other
platform-specific blitters.
Review URL: http://codereview.appspot.com/3669042/
git-svn-id: http://skia.googlecode.com/svn/trunk@636 2bbb7eff-a529-9590-31e7-b0007b416f81
obvious artifacts. We need a more comprehensive fix for all gradients (32bit
and 16bit) when we have a large number of colors (or they are spaced out very
non-linearly), to avoid banding.
git-svn-id: http://skia.googlecode.com/svn/trunk@458 2bbb7eff-a529-9590-31e7-b0007b416f81
call these new procs in (nearly) all the places we had inlined loops before.
In once instance (blitter_argb32::blitAntiH) we get different results by a
tiny bit. The new code is more accurate, and exactly inline with all of the
other like-minded blits, so I think the change is good going forward.
git-svn-id: http://skia.googlecode.com/svn/trunk@366 2bbb7eff-a529-9590-31e7-b0007b416f81
effectively chopping off the top row of pixels when we computed a blur.
git-svn-id: http://skia.googlecode.com/svn/trunk@358 2bbb7eff-a529-9590-31e7-b0007b416f81
(like gradients) to support predithering. If they do, then they would suppress
kConstInY16, since they no longer are const.
The blitters now check for each flag separately, so we don't have to give up
const-in-Y in the 32bit case, since in that mode we don't care about dithering.
git-svn-id: http://skia.googlecode.com/svn/trunk@339 2bbb7eff-a529-9590-31e7-b0007b416f81
has been computed, othewise we can't distinguish 0x00000000 from 0x00FF0000
Add fast case for index blit where we read 4 src pixels at a time
git-svn-id: http://skia.googlecode.com/svn/trunk@248 2bbb7eff-a529-9590-31e7-b0007b416f81
define constant for max value flatten/unflatten can return (so clients can put
the buffer on the stack)
git-svn-id: http://skia.googlecode.com/svn/trunk@243 2bbb7eff-a529-9590-31e7-b0007b416f81
this signals blitters that the shader will return the same results for a given
x value, independent of y. Useful inside blitRect(), where it can cache the
first call to shadeSpan() and reuse it on all subsequent scans. Works with
(non-rotated) linear-gradients, and Nx1 bitmaps.
git-svn-id: http://skia.googlecode.com/svn/trunk@214 2bbb7eff-a529-9590-31e7-b0007b416f81
add SkShape baseclass, in the hopes of having SkPicture inherit from that, and
also using shapes as the extension mechanism for things like animated-gif
git-svn-id: http://skia.googlecode.com/svn/trunk@174 2bbb7eff-a529-9590-31e7-b0007b416f81
Should not hurt features (we hope), but is an easy way to survive malicious
values on a small-memory machine like a handset.
git-svn-id: http://skia.googlecode.com/svn/trunk@123 2bbb7eff-a529-9590-31e7-b0007b416f81
