- 5492424 Cleanup code related to SkFlattenable::flatten() by djsollen@google.com · 12 years ago
- 5370cd9 Consolidate PixelRef flattables with the standard impl by djsollen@google.com · 12 years ago
- a728e35 Fix a bunch of Coverity defects - class members not initialized in the constructor. by vandebo@chromium.org · 12 years ago
- 42aea28 handle null-device correctly in setDevice by reed@google.com · 12 years ago
- 19224c3 Fix a memory leak in SkAAClip::translate(). by tomhudson@google.com · 12 years ago
- 9fe287b add debuggin matrix-proc to validate its output before calling the samplers by reed@google.com · 12 years ago
- ba28d03 Create macro for registering classes for deserialization by djsollen@google.com · 12 years ago
- d43b0a8 Make use of the functions in the flattenable reader/writer by djsollen@google.com · 12 years ago
- a2ca41e Cleanup Macros arround static initializers for SkFlattenable. by djsollen@google.com · 12 years ago
- 8926b16 apply imagefilter to all draw calls by reed@google.com · 12 years ago
- 6db75fc Initialize margin, which was being used uninitialized in compute_bounds() by tomhudson@google.com · 12 years ago
- e63793a Upstream changes from Android repository. by djsollen@google.com · 12 years ago
- 5994451 Move SkClampRange into src/effects, where it isn't being advertised as by tomhudson@google.com · 12 years ago
- 166e653 Fix scaling issue when drawing text on a path. by djsollen@google.com · 12 years ago
- 0b53d59 add optional updateMatrixClip param to getTopDevice(). Chrome should set this by reed@google.com · 12 years ago
- 57e080c Another fix to r3423, reinstate blitter code that was commented out for by tomhudson@google.com · 12 years ago
- 8dd90a9 (SSE2) acceleration for rectangular opaque erases. by tomhudson@google.com · 12 years ago
- 80e1459 add IRect variant of getBounds() by reed@google.com · 12 years ago
- 86b2e43 add getBounds() helper by reed@google.com · 12 years ago
- 0724f43 don't clamp the stopScale to 0 unless it originally exceeded the clip. This is by reed@google.com · 12 years ago
- 6b2f7ea tweak gm bounds to better fit the content by reed@google.com · 12 years ago
- 68bc6f7 add kUTF32_TextEncoding by reed@google.com · 12 years ago
- 610f716 Fix four memory leaks uncovered by valgrinding gm tests. by vandebo@chromium.org · 12 years ago
- 77613a5 enable fractionalint (64bit) math in bitmap procs by reed@google.com · 12 years ago
- 9f79ece add build option to revert hairline-clipping fix from 3366 by reed@google.com · 12 years ago
- adb2e24 if we are extending past our clip in antihairlines, don't draw that last by reed@google.com · 12 years ago
- f9f258c detect all masks, and don't use masksuperblitter with those by reed@google.com · 12 years ago
- f444e8c don't flip the fractional bits (low 4) for bilerp when in kMirror_TileMode by reed@google.com · 12 years ago
- 4c43372 If we try to clip against a path and the path only contains a rect, treat it as a clip against a rect. by tomhudson@google.com · 12 years ago
- c5039dd disable 64bit opt for now, as I debug skew failure in DRT by reed@google.com · 12 years ago
- 411215a apply FractionalInt improvement to rotated by reed@google.com · 12 years ago
- 4bc0a9d use higher precision (64bit) fixed-point for matrix proc by reed@google.com · 12 years ago
- ae29b88 SSSE3 code for drawing rotated bitmaps; produces 25-30% speedup. by tomhudson@google.com · 12 years ago
- 05054f1 Erode and dilate image filter effects, CPU and GPU implementations. by senorblanco@chromium.org · 12 years ago
- 10d2d4d move the post cleanup of the rec until after we wack it for maskfilter by reed@google.com · 12 years ago
- 48a4cbc add experimental files to encapsulate device-level font settings by reed@google.com · 12 years ago
- 087d5aa fix edgecase in chopcubic where we computed duplicate tvalues by reed@google.com · 12 years ago
- 9b0da23 fix bug and add test for aaclip when blitV is called by reed@google.com · 12 years ago
- aefdd06 move SKPaint's defaults (that might be configurable) into a header by reed@google.com · 12 years ago
- d76dafb handle checking for large bounds correctly when SkScalar is fixed by reed@google.com · 12 years ago
- 5efaf26 SSE2 version of ClampX_ClampY_{no}filter_affine, courtesy of Jin Yang. by tomhudson@google.com · 12 years ago
- 07db861 since SkScalarCeil/Floor do not explicitly detect overflows from float->int, by reed@google.com · 12 years ago
- a4e9238 Disable check for large bounds in antipath by reed@google.com · 12 years ago
- 92d2a29 reapply r3259 (that was reverted) with fix. by djsollen@google.com · 12 years ago
- a907ac3 Modify SkDeferredCanvas so that it uses its inherited SkCanvas to track matrix and clipping state by junov@chromium.org · 12 years ago
- 3cc0d8a reverting r3259 because of gm/test failure. by djsollen@google.com · 12 years ago
- 42bbef7 Enable quickRejectY in SkCanvas to be inlined. by djsollen@google.com · 12 years ago
- 0b5b0ce reverting r3257 because of gm failure. by junov@chromium.org · 12 years ago
- 421bcc3 Modify SkDeferredCanvas so that it uses its inherited SkCanvas to track matrix and clipping state by junov@chromium.org · 12 years ago
- f68c5e2 check for null-device in iterator by reed@google.com · 12 years ago
- ffc9248 Fix math in coverage_to_alpha() (renamed coverage_to_partial_alpha()), behind by tomhudson@google.com · 12 years ago
- dceecc7 always allocate (only) the width of the clip for the AA row, since it will by reed@google.com · 12 years ago
- 8796ee6 remove bogus assert (leftover from some long-forgotten debugging session) by reed@google.com · 12 years ago
- a91e923 GPU device preserves pixel values across read/write/read of unpremul pixel values by bsalomon@google.com · 12 years ago
- fb0d741 Use nested template expansions to implement Config8888 conversions. by bsalomon@google.com · 12 years ago
- 1cc8f6f Adding more unit testing for SkCanvas and derived classes. by junov@chromium.org · 12 years ago
- 06a7313 SSE2 version of ClampX_ClampY_{no}filter_scale; yields 10-20% speedup in by tomhudson@google.com · 12 years ago
- 543c719 Revert r3036. This change breaks loading images as by caryclark@google.com · 12 years ago
- 6c5bf8d Added BSD license language where missing. by caryclark@google.com · 12 years ago
- f0ed80a Make cross_prod used in SkPath::cheapComputeDirection fallback to double computation when result is 0. Verbal LGTM from reed. by bsalomon@google.com · 12 years ago
- c2c80c4 Remove two static initializers from debug builds by digit@google.com · 12 years ago
- 302b861 be sure to mark fLastY with the bottom of the rect in blitRect by reed@google.com · 12 years ago
- 80cdb9a aaclip needs to handle paths with holes by reed@google.com · 12 years ago
- 897fc41 add a todo by reed@google.com · 12 years ago
- ee51d1a Fix TypefaceCache races. http://codereview.appspot.com/5656066/ by bungeman@google.com · 12 years ago
- 60abb07 Merge changes from the android repo upstream to Skia by djsollen@google.com · 12 years ago
- dd43df9 suport A8 grayscale text with COLOR_LUMINANCE build option by reed@google.com · 12 years ago
- d6770e6 SSE2 version of blit_lcd16, courtesy of Jin Yang. by tomhudson@google.com · 12 years ago
- 8f9ecbd Adding API and unit testing for deferred canvas clearing/purging by junov@chromium.org · 12 years ago
- 813d38b if SK_USE_COLOR_LUMINANCE is defined, then we store 2 bits of each component by reed@google.com · 12 years ago
- ce6dbb6 enable color-luminance in font cache (disabled by default) by reed@google.com · 12 years ago
- 1f9767c Fixing backing store access in SkDeferredCanvas. by junov@chromium.org · 12 years ago
- 0a9cc44 use macro SkFixedRoundToInt where necessary by reed@google.com · 12 years ago
- f3bf18b Don't change convexity status of a path after a moveTo by bsalomon@google.com · 12 years ago
- fb6f0f6 When adding a convex shape to a path, mark path as convex if path is initially degenerate. by bsalomon@google.com · 12 years ago
- 3e44e4d fall through to cross-product if the y-max-baseline returns a degenerate line by reed@google.com · 12 years ago
- c1ea60a handle multiple points all at the y-max when computing direction by reed@google.com · 12 years ago
- ac8543f use contour with global ymax to determine direction by reed@google.com · 12 years ago
- fa4d5bd add bandage to pin out-of-range values when converting our integral clip bounds to by reed@google.com · 12 years ago
- 5546ef2 inject a 32767 bounded cliprect before using SuperSampler blitter, to avoid by reed@google.com · 12 years ago
- 9efd9a0 extend fastbounds impls to include maskfilters and drawloopers. This allows by reed@google.com · 12 years ago
- bf6c1e4 Adding a flush method to SkCanvas by junov@chromium.org · 12 years ago
- 74b4619 Formatting cleanup: remove extra whitespace by vandebo@chromium.org · 12 years ago
- 1771cbf posix: Avoid static initializers in static/global mutexes by digit@google.com · 12 years ago
- 6d428d3 Make SkRegion::operator== a member function, rather than a privately-declared by reed@google.com · 13 years ago
- 8887be0 disable SPEW_PURGE_STATUS by reed@google.com · 13 years ago
- fde6b41 Cleanup of SkPath code to remove SK_OLD_EMPTY_PATH_BEHAVIOR and other by schenney@chromium.org · 13 years ago
- 13e812c Interpolate vertical linear gradients for improved quality. by tomhudson@google.com · 13 years ago
- dd1be60 Pull xfer mode test out of generic draw-as-hairline test. Use coverage rather than alpha to draw hairlines < 1pix wide in GPU. by bsalomon@google.com · 13 years ago
- a6d04d9 Fixing the behavior of SkPathMeasure to reflect changes in SkPath::Iter. by schenney@chromium.org · 13 years ago
- 8351aab add GenA8FromLCD as a hack to force GDI to create the A8 mask from the LCD by reed@google.com · 13 years ago
- 4370aed Adding class SkDeferredCanvas for deferred rendering. by junov@google.com · 13 years ago
- 188bfcf if the y-max has neighbors (prev and next) with the same Y values, then the by reed@google.com · 13 years ago
- 0eb7576 Modify SkBitmap::extractSubset() to respect volatility by skyostil@google.com · 13 years ago
- 2211b62 Subpixel text 3/8 of a pixel too far to the right. by bungeman@google.com · 13 years ago
- ce7adb5 Modify SkBitmap::extractSubset() to respect opaqueness by skyostil@google.com · 13 years ago
- d335d1d automatically inject a moveTo if we see a close followed by a line/quad/cubic by reed@google.com · 13 years ago
- cabaf1d reenable direction test, fix handling of degenerate segments in the non-convex case by reed@google.com · 13 years ago
- 3d19c38 fix cheapComputeDirection() in the non-convex case to pivot on the y-max point by reed@google.com · 13 years ago
- 6da3d17 improve clamping logic when handling imprecise chopping of cubics by reed@google.com · 13 years ago