modify threaded path ops tests to check
Background: this CL came out of a conversation with Eric where I learned that 10s of machines host 100s of bots. Since the bot hosting tests may be shared with many other tasks, it seems unwise for path ops to launch multiple test threads.
The change here is to make launching multiple threads "opt-in" and by default, bots can run path ops in a single thread.
Review URL: https://codereview.chromium.org/14002007
git-svn-id: http://skia.googlecode.com/svn/trunk@8750 2bbb7eff-a529-9590-31e7-b0007b416f81
fix bugs in tests on 32 bit release
Most changes revolve around pinning computed t values
very close to zero and one.
git-svn-id: http://skia.googlecode.com/svn/trunk@8745 2bbb7eff-a529-9590-31e7-b0007b416f81
Try to fix the 32 bit build by making some math
decisions more robust.
Rewrite the cubic intersection special case that
detects if only end points are shared.
Rewrite the angle sort setup that computes whether
a cubic bends to the left or right.
git-svn-id: http://skia.googlecode.com/svn/trunk@8726 2bbb7eff-a529-9590-31e7-b0007b416f81
This fixes the following clang error:
../../tests/XfermodeTest.cpp:43:44: error: comparison of constant -1 with expression of type 'SkXfermode::Mode' is always false [-Werror,-Wtautological-constant-out-of-range-compare]
REPORTER_ASSERT(reporter, reportedMode == -1);
~~~~~~~~~~~~ ^ ~~
R=vandebo@chromium.org,reed@google.com
Review URL: https://codereview.chromium.org/14180004
git-svn-id: http://skia.googlecode.com/svn/trunk@8692 2bbb7eff-a529-9590-31e7-b0007b416f81
standardize tests
use SK_ARRAY_COUNT everywhere
debug why x87 differs from SIMD 64
various platform specific fixes
git-svn-id: http://skia.googlecode.com/svn/trunk@8689 2bbb7eff-a529-9590-31e7-b0007b416f81
The fImmediateDevice member of DeferredDevice (SkDeferredCanvas.cpp) was becoming invalid after a fork of the backingstore in SkSurface_Gpu cause the device to be substituted.
New unit test code was to exercise SkSurface copy on write with draws that are deferred in GrInOrderDrawBuffer. The bad pointer was causing the test to crash.
TEST=skia unit test DeferredCanvas, subtest TestDeferredCanvasSurface
Review URL: https://codereview.chromium.org/14263015
git-svn-id: http://skia.googlecode.com/svn/trunk@8686 2bbb7eff-a529-9590-31e7-b0007b416f81
This temporarily disables SK_ENABLE_INST_COUNT
( skbug.com/1219 )
This fixes a linktime error on VS2012 in
PathTest.cpp; -SK_ScalarInfinity should be
SK_ScalarNegativeInfinity instead.
This adds pathops and pathops unit tests to the
main unit tests.
Should this change destabilize anything, it should
be sufficient to comment out the pathops gypi
includes. at test.gyp:18,21.
Review URL: https://codereview.chromium.org/14137010
git-svn-id: http://skia.googlecode.com/svn/trunk@8644 2bbb7eff-a529-9590-31e7-b0007b416f81
I don't understand why this change is necessary. On Android,
SkCommandLineFlags.h is not found, but only in this project.
Other projects depend on flags and include the file without
using the full path. Likewise, this works on other platforms.
Removing for now until I figure out the correct fix.
Review URL: https://codereview.chromium.org/13910008
git-svn-id: http://skia.googlecode.com/svn/trunk@8621 2bbb7eff-a529-9590-31e7-b0007b416f81
- fix rand for Android
- build unit test on linux
- use atomic inc in test count
- add casting for Android
git-svn-id: http://skia.googlecode.com/svn/trunk@8610 2bbb7eff-a529-9590-31e7-b0007b416f81
subDivide limit. This caused problems with degenate paths (too much recursion).
The fix was two parts:
1. decrement the subDivide limit as we recurse
2. up the limit for cubics to 7, to match our current quality
added unittest that replicated the too-much-recursion bug.
Review URL: https://codereview.chromium.org/14086002
git-svn-id: http://skia.googlecode.com/svn/trunk@8599 2bbb7eff-a529-9590-31e7-b0007b416f81
This CL depends on
https://codereview.chromium.org/12880016/
"Add intersections for path ops"
Given a path, iterate through its contour, and
construct an array of segments containing its curves.
Intersect each curve with every other curve, and for
cubics, with itself.
Given the set of intersections, find one with the
smallest y and sort the curves eminating from the
intersection. Assign each curve a winding value.
Operate on the curves, keeping and discarding them
according to the current operation and the sum of
the winding values.
Assemble the kept curves into an output path.
Review URL: https://codereview.chromium.org/13094010
git-svn-id: http://skia.googlecode.com/svn/trunk@8553 2bbb7eff-a529-9590-31e7-b0007b416f81
This CL depends on
https://codereview.chromium.org/12827020/
"Add base types for path ops"
The intersection of a line, quadratic, or cubic
with another curve (or with itself) is found by
solving the implicit equation for the curve pair.
The curves are first reduced to find the simplest
form that will describe the original, and to detect
degenerate or special-case data like horizontal and
vertical lines.
For cubic self-intersection, and for a pair of cubics,
the intersection is found by recursively
approximating the cubic with a series of quadratics.
The implicit solutions depend on the root finding
contained in the DCubic and DQuad structs, and
the quartic root finder included here.
Review URL: https://codereview.chromium.org/12880016
git-svn-id: http://skia.googlecode.com/svn/trunk@8552 2bbb7eff-a529-9590-31e7-b0007b416f81
Paths contain lines, quads, and cubics, which are
collectively curves.
To work with path intersections, intermediary curves
are constructed. For now, those intermediates use
doubles to guarantee sufficient precision.
The DVector, DPoint, DLine, DQuad, and DCubic
structs encapsulate these intermediate curves.
The DRect and DTriangle structs are created to
describe intersectable areas of interest.
The Bounds struct inherits from SkRect to create
a SkScalar-based rectangle that intersects shared
edges.
This also includes common math equalities and
debugging that the remainder of path ops builds on,
as well as a temporary top-level interface in
include/pathops/SkPathOps.h.
Review URL: https://codereview.chromium.org/12827020
git-svn-id: http://skia.googlecode.com/svn/trunk@8551 2bbb7eff-a529-9590-31e7-b0007b416f81
added capability to collect minidump and callstack if buildbot fails with heap
coruption in windows, and a NPE bug was fixed in SkPDFDocument, when document was destroyed without ever beeing used and a field was NULL + a few minor conflicts have been resolved)
git-svn-id: http://skia.googlecode.com/svn/trunk@8487 2bbb7eff-a529-9590-31e7-b0007b416f81
Replace the old attribute binding and index interface with one where we include the binding as part of the attribute array. Also removed the fixed attribute indices for constant color and coverage attributes, and replaced with dynamic ones based on current attribute set. Removed binding of color and coverage attributes unless they're actually set.
Original author: bsalomon@google.com
Author: jvanverth@google.com
Reviewed By: bsalomon@google.com,robertphillips@google.com
Review URL: https://chromiumcodereview.appspot.com/13296005
git-svn-id: http://skia.googlecode.com/svn/trunk@8468 2bbb7eff-a529-9590-31e7-b0007b416f81
Replace the old attribute binding and index interface with one where we include the binding as part of the attribute array. Also removed the fixed attribute indices for constant color and coverage attributes, and replaced with dynamic ones based on current attribute set. Removed binding of color and coverage attributes unless they're actually set.
Original author: bsalomon@google.com
Author: jvanverth@google.com
Reviewed By: bsalomon@google.com,robertphillips@google.com
Review URL: https://chromiumcodereview.appspot.com/13296005
git-svn-id: http://skia.googlecode.com/svn/trunk@8466 2bbb7eff-a529-9590-31e7-b0007b416f81
This strips out last of the edge types and the fixed function edge attribute and replaces them with using GrEdgeEffect. Also fixes a minor bug when checking attribute counts -- it was using kAttribIndexCount instead of kVertexAttribCnt.
Original Author: jvanverth@google.com
Review URL: https://codereview.chromium.org/13069003
git-svn-id: http://skia.googlecode.com/svn/trunk@8392 2bbb7eff-a529-9590-31e7-b0007b416f81
Effects can ask the builder for local coords which may or may not be distinct from positions.
GrEffectStage tracks changes to relationship between pos and local coords.
GrGLEffectMatrix and GrSingleTextureEffect can use either pos or textures as intput coords
GrSimpleTextureEffect now allows for an explicit texture coords attribute.
Review URL: https://codereview.chromium.org/12531015
git-svn-id: http://skia.googlecode.com/svn/trunk@8264 2bbb7eff-a529-9590-31e7-b0007b416f81
Adds some optimizations to the circle and ellipse shaders, static effect
instances for their GrEffects, and some minor changes to GrDrawState::setEffect
to make GrEffect setup faster.
git-svn-id: http://skia.googlecode.com/svn/trunk@8238 2bbb7eff-a529-9590-31e7-b0007b416f81
SkPurgeableImageCache:
New image cache that uses virtual memory to store the pixels. Combines
features of SkAshmemImageCache (which has been removed) with SkPurgeableMemoryBlock, which has android and Mac versions.
SkImageCache:
Modified the API. pinCache now returns a status out parameter which
states whether the pinned memory retained the old data. This allows
allocAndPinCache to only be used for allocations.
Add a new debug only interface to purge unpinned data.
Updates to documentation, clarifying behavior.
Changed CachedStatus to MemoryStatus
SkLruImageCache:
Implement the new function purgeAllUnpinnedCaches and change implementation
of pinCache for the new behavior.
SkLazyPixelRef:
Rewrite onLockPixels to account for the new behavior of pinCache.
BitmapFactoryTest:
Test the new SkPurgeableImageCache.
Write tests which directly test the SkImageCaches.
Create a larger bitmap, since some of the SkImageCaches are designed
to handle large bitmaps.
bench_ and render_pictures:
Consolidate lazy_decode_bitmap into one function.
Allow using a flag to specify using the purgeable image cache.
Clean up some #includes.
Review URL: https://codereview.chromium.org/12433020
git-svn-id: http://skia.googlecode.com/svn/trunk@8207 2bbb7eff-a529-9590-31e7-b0007b416f81
unused headers and fixes a couple of bugs exposed by changing the random
number generator:
First, the function SkMatrix::getMaxStretch() had an error where it was testing
the square of a number against near-zero. This led to it occasionally taking a
cheaper but imprecise path for computing the eigenvalues of the matrix. It's
been replaced with a check against the square of SK_ScalarNearlyZero.
The second case was a failure in ClipStackTest, where it hit the rare case of
a practically empty clip stack (it has a single Union) and we set a tight
bounds. The bounds rect doesn't get set by GrReducedClip::ReduceClipStack() in
this case, so when it clips the reduced stack it's clipping against garbage,
and the resulting regions don't match. The solution is to initialize the
tightBounds rect.
git-svn-id: http://skia.googlecode.com/svn/trunk@7952 2bbb7eff-a529-9590-31e7-b0007b416f81
This changes the old method of setting vertex layout to a new one where we
specify vertex attribute data separately from attribute bindings (i.e. program
functionality). Attribute data is now set up via an array of generic attribute
types and offsets, and this is mapped to the old program functionality by
setting specific attribute indices. This allows us to create more general
inputs to shaders.
git-svn-id: http://skia.googlecode.com/svn/trunk@7899 2bbb7eff-a529-9590-31e7-b0007b416f81
I removed unused parameters in the tests wherever it was trivial to do so. I'm trying to get the easy ones out of the way before we get into more involved discussions around this.
Review URL: https://codereview.appspot.com/7394055
git-svn-id: http://skia.googlecode.com/svn/trunk@7891 2bbb7eff-a529-9590-31e7-b0007b416f81
This patch changes the semantics of tileWidth/Height to include the border region, and
uses an offset to take into account the fact that there is no outer border for outer
tiles. This patch also fixes a previous bug where the right column and bottom row were
considered to be included in bounds that are expressed as an SkIRect.
Companion Chromium CL required for roll: https://codereview.chromium.org/12221077/
TEST=TileGrid unit test
Review URL: https://codereview.appspot.com/7350050
git-svn-id: http://skia.googlecode.com/svn/trunk@7885 2bbb7eff-a529-9590-31e7-b0007b416f81
The problem fixed - http://code.google.com/p/skia/issues/detail?id=940 - is that getResources will recursively obtain all child resource recursively without checking for duplicates.
If we have lots of duplicates, then we try to build a very large vector (exponential with the number of nodes usually) and sooner or later we end up using too much memory and crash.
A possible solution could have been to make sure resources do not have duplicates, but that requirement is impractical, and it this leaves the solution fragile, if there is any issue in the tree, we crash.
When we emit the pdf, the large number of duplicates is not an issue, because SkPDFCatalog::addObject will deal with duplicates.
I have run the gm with --config pdf, and the images are 100% same bits, while the pdfs have the same size but some very small changes, the order of some objects.
Review URL: https://codereview.appspot.com/6744050
git-svn-id: http://skia.googlecode.com/svn/trunk@7883 2bbb7eff-a529-9590-31e7-b0007b416f81
The new pixel ref behaves similarly to SkImageRef, with some key differences:
It does not depend on the images project.
It requires an SkImageCache, which handles allocation and caching of the pixel
memory.
It takes a function signature for decoding which decodes into already allocated
pixel memory rather than into an SkBitmap.
Add two implementations of SkImageCache: SkLruImageCache and SkAshmemImageCache.
Replace SkSerializationHelpers::DecodeBitmap with SkPicture::InstallPixelRefProc,
and update sites that referenced it.
SkBitmapFactory now sets the pixel ref to a new object of the new
class SkLazyPixelRef, provided it has an SkImageCache for caching.
Provide an option to do lazy decodes in render_pictures and bench_pictures.
SkPicture:
Eliminate the default parameters in the constructor.
If a proc for decoding bitmaps is installed, use it to decode any encoded
data in subpictures.
When parsing deserializing subpictures, check for success.
When serializing subpictures, pass the picture's bitmap encoder to the
subpicture's call to serialize.
Update BitmapFactoryTest to test its new behavior.
BUG=https://code.google.com/p/skia/issues/detail?id=1008
BUG=https://code.google.com/p/skia/issues/detail?id=1009
Review URL: https://codereview.appspot.com/7060052
git-svn-id: http://skia.googlecode.com/svn/trunk@7835 2bbb7eff-a529-9590-31e7-b0007b416f81
This new changelist also introduces a new image filter called SkRectShaderImageFilter which is make to simply apply a shader on a region without using any inputs.
TEST=Added ShaderImageFilter test
Review URL: https://codereview.appspot.com/7300046
git-svn-id: http://skia.googlecode.com/svn/trunk@7808 2bbb7eff-a529-9590-31e7-b0007b416f81
Currently we support 5 texture stages, each with 5 possible texture coordinate attributes.
However, we only ever use one explicit texture coordinate. This change removes all but one
(now named just "aTexCoord") of the possible explicit texture coordinates.
Review URL: https://codereview.appspot.com/7308094/
git-svn-id: http://skia.googlecode.com/svn/trunk@7737 2bbb7eff-a529-9590-31e7-b0007b416f81
My implementation of the Gorilla random number test had a bug in the code used
to track the random strings -- it was masking 6 bits instead of 5, which was
throwing off the counts. No idea how this worked on every platform except
Android.
git-svn-id: http://skia.googlecode.com/svn/trunk@7731 2bbb7eff-a529-9590-31e7-b0007b416f81
Upcoming changes will require GrDrawState to know things about the set of installed effects. Thus all setting of effects must go through a GrDrawState function (setEffect()). This change accomplishes that.
Review URL: https://codereview.appspot.com/7214045
git-svn-id: http://skia.googlecode.com/svn/trunk@7411 2bbb7eff-a529-9590-31e7-b0007b416f81
- Add drawOval base function to SkDevice, and override in SkGpuDevice
- Move isSimilarityMatrix to SkMatrix (renamed to isSimilarity) and fixed up unit test
- Since both SkGpuDevice::drawOval() and GrContext::drawPath() can try to draw ovals, added GrContext::canDrawOval() and GrContext::internalDrawOval() to avoid duplicate code
- Hooked in axis-aligned oval fill shader
- Enabled GPU stroked circles
- Added stroked circle bench test
Review URL: https://codereview.appspot.com/7137050
git-svn-id: http://skia.googlecode.com/svn/trunk@7304 2bbb7eff-a529-9590-31e7-b0007b416f81
This fixes crbug 170666. Before the fix, were were finishing the "clip" step
with a curve that was still partly negative, and so as assert would fire.
added unittest to confirm that the assert doesn't fire.
git-svn-id: http://skia.googlecode.com/svn/trunk@7278 2bbb7eff-a529-9590-31e7-b0007b416f81
of the cliprect, we (correctly) clamp the Y coordinate of the control-point
right next to the on-curve point that was chopped (this ensures we don't go
slightly outside of the clip-rect due to imperfect T value calculation). However,
the code was also clamping the other control-point as well, resulting in warping
the cubic, which could sometimes force it outside of the clip.
The fix is to just remove the line of code that clampped the 2nd control-point.
unittest added to reproduce a test cubic that triggered an assert, due to the
cubic being outside of the cliprect. The test (w/o the fix) will assert in
a SK_DEBUG build.
Review URL: https://codereview.appspot.com/7100056
git-svn-id: http://skia.googlecode.com/svn/trunk@7184 2bbb7eff-a529-9590-31e7-b0007b416f81
SkMemoryStream attempts to dereference fData in multiple places.
Instead of allowing it to be NULL, resulting in a crash, set it
to SkData::NewEmpty().
Add a test for SkStream that will crash when its SkData is set to NULL.
Review URL: https://codereview.appspot.com/7061059
git-svn-id: http://skia.googlecode.com/svn/trunk@7111 2bbb7eff-a529-9590-31e7-b0007b416f81
Most of these issues were due to functions whose definitions appear in header files; I changed those functions to be 'static inline' instead of just 'static' or 'inline', which kills the warning for such functions.
Other functions that were static or anonymous-namespaced but were unused in cpp files were probably called at some point but are no longer; someone who knows more than I do should probably scrub all the functions I either deleted or #if 0'ed out and make sure that the right thing is happening here.
Lots of unused variables removed, and one nasty const issue handled.
There remains a single warning in thirdparty/externals/cityhash/src/city.cc on line 146 related to a signed/unsigned mismatch. I don't know if we have control over this library so I didn't fix this one, but perhaps someone could do something about that one.
BUG=
Review URL: https://codereview.appspot.com/7067044
git-svn-id: http://skia.googlecode.com/svn/trunk@7051 2bbb7eff-a529-9590-31e7-b0007b416f81
Fixing an inverted test condition. The tested feature is not broken, but the test was passing the inverted test because code just above was missing a restore call, which was affecting the test.
Review URL: https://codereview.appspot.com/6921044
git-svn-id: http://skia.googlecode.com/svn/trunk@6745 2bbb7eff-a529-9590-31e7-b0007b416f81
method. Preliminary results using Chromium's cc_perftests
show that we can avoid almost half of the default constructors from
redundantly/unnecessarily initializing the matrix to identity.
Review URL: https://codereview.appspot.com/6872056
git-svn-id: http://skia.googlecode.com/svn/trunk@6742 2bbb7eff-a529-9590-31e7-b0007b416f81
If an SkBitmap has encoded data, write that during serialization
rather than reencoding it.
Add a test to ensure that this does not modify the output stream,
so the reader need not know the difference.
Review URL: https://codereview.appspot.com/6884054
git-svn-id: http://skia.googlecode.com/svn/trunk@6732 2bbb7eff-a529-9590-31e7-b0007b416f81
In SkBitmap::extractSubset, perform a deepCopy, if the pixelRef supports it.
Fixes a bug in the 'extractbitmap' gm, which attempts to draw a subset of a texture backed bitmap (if the canvas is really an SkGpuCanvas).
Also fix some bugs that happen when there is a pixel offset. These fixes get bypassed by the deepCopy, but a user can still set a pixel offset manually.
When copying GPU backed bitmap with a pixel offset, copy the offset.
If the new config is the same as the old, copy fRowBytes as well.
Add a function to SkBitmap.cpp (getUpperLeftFromOffset) to find the x,y coordinate to use when copying to a new config.
Fix a bug where readPixels copied to the correct desired config and we were setting the generation ID to match even though the desired config was not the same as the original config (caught by my new tests!).
Add some tests to verify the correct behavior.
Review URL: https://codereview.appspot.com/6839043
git-svn-id: http://skia.googlecode.com/svn/trunk@6710 2bbb7eff-a529-9590-31e7-b0007b416f81
If SkBitmapHeap::insert() returns INVALID_SLOT, assert at picture
record time so we can debug, but allow it to continue in release
mode, so that we can still capture a picture. At playback time,
print a message so we know that there was an error.
Review URL: https://codereview.appspot.com/6873050
git-svn-id: http://skia.googlecode.com/svn/trunk@6664 2bbb7eff-a529-9590-31e7-b0007b416f81
add bench
optimize operator== by performing 4 compares in a row before checking
optimize setconcat by noting when we can write the answer directly into this
At least on this macbook, I had to mark helpers like isIdentity() as inline to get them inlined.
Review URL: https://codereview.appspot.com/6863053
git-svn-id: http://skia.googlecode.com/svn/trunk@6655 2bbb7eff-a529-9590-31e7-b0007b416f81
Add a test and a GM for the factory, and a PNG file for it to decode.
The PNG file is copyright-free, obtained from
http://openclipart.org/detail/29213/paper-plane-by-ddoo
In cmykjpeg, do not attempt to decode in the constructor, since it
currently crashes on Mac (if you provide the correct resource path).
Even when we fix this crash there is no need to do it in the
constructor, since we create all of the gms in order to
get their names (to determine whether to run them).
Review URL: https://codereview.appspot.com/6847122
git-svn-id: http://skia.googlecode.com/svn/trunk@6618 2bbb7eff-a529-9590-31e7-b0007b416f81
The problem arises on devices like the Nexus 10 where we allow the
destruction of resources using the destructor of a static variable.
However, we have no guarentee that the GPU driver has not already
cleaned up it's resources prior to our static destructor.
Review URL: https://codereview.appspot.com/6851124
git-svn-id: http://skia.googlecode.com/svn/trunk@6599 2bbb7eff-a529-9590-31e7-b0007b416f81
Allow a rect to start in the middle of a span, and
wrap all the way around.
Initialize variable to suppress warning.
Add tests to detect rects constructed from a stroked
path.
Review URL: https://codereview.appspot.com/6847082
git-svn-id: http://skia.googlecode.com/svn/trunk@6522 2bbb7eff-a529-9590-31e7-b0007b416f81
Tease apart existing one rect path detector so
that a new variant can detect two nested rects as well.
Add tests to verify that both one and two rect
detectors both work and return the correct results.
Suppress other warnings in PathTest.
Review URL: https://codereview.appspot.com/6850059
git-svn-id: http://skia.googlecode.com/svn/trunk@6475 2bbb7eff-a529-9590-31e7-b0007b416f81
