Commit Graph

9 Commits

Author SHA1 Message Date
rmistry@google.com
d6176b0dca Result of running tools/sanitize_source_files.py (which was added in https://codereview.appspot.com/6465078/)
This CL is part II of IV (I broke down the 1280 files into 4 CLs).
Review URL: https://codereview.appspot.com/6474054

git-svn-id: http://skia.googlecode.com/svn/trunk@5263 2bbb7eff-a529-9590-31e7-b0007b416f81
2012-08-23 18:14:13 +00:00
reed@google.com
e2faf17bcc Even when the pts are restricted to 32K values, we can still overflow computing
a fixed-point coefficient for quadratics. To avoid this, we bias these
coefficients, storing 1/2 of their actual value, and then apply the 2x unbias
in updateQuadratic().

Fixes http://code.google.com/p/chromium/issues/detail?id=140803
Review URL: https://codereview.appspot.com/6450099

git-svn-id: http://skia.googlecode.com/svn/trunk@4960 2bbb7eff-a529-9590-31e7-b0007b416f81
2012-08-06 19:01:34 +00:00
reed@google.com
d9ee348720 If we lose precision computing sum of the dash intervals, then we can have the
same error when we subtract-in-a-loop with the phase. The result is that we can
read past the end of the array.

To fix this, we just pin the loop counter, and if we exhaust our intervals, we
just treat the phase as 0. Not precisely the exact answer, but we aren't going
to draw this dash correctly anyway, since it contains massive interval values
that will be imprecise given our current float implementation.

Fixes http://code.google.com/p/chromium/issues/detail?id=140642
Review URL: https://codereview.appspot.com/6458088

git-svn-id: http://skia.googlecode.com/svn/trunk@4959 2bbb7eff-a529-9590-31e7-b0007b416f81
2012-08-06 14:58:35 +00:00
reed@google.com
b59ed51564 add test for http://code.google.com/p/chromium/issues/detail?id=131181
git-svn-id: http://skia.googlecode.com/svn/trunk@4266 2bbb7eff-a529-9590-31e7-b0007b416f81
2012-06-15 18:26:04 +00:00
reed@google.com
9d5f76a250 fix bug (and add test) for drawing an inverse-path whose bounds do intersect
the clip, but whose edges do not (e.g. a curve). We used to overdraw a section
(and assert).



git-svn-id: http://skia.googlecode.com/svn/trunk@3809 2bbb7eff-a529-9590-31e7-b0007b416f81
2012-05-01 14:49:28 +00:00
reed@google.com
1df888bc3a check for numerical imprecision on phase
fixes http://code.google.com/p/chromium/issues/detail?id=124652



git-svn-id: http://skia.googlecode.com/svn/trunk@3761 2bbb7eff-a529-9590-31e7-b0007b416f81
2012-04-24 22:47:21 +00:00
reed@google.com
a90aa53498 fix clamping when we chop a cubic
Review URL: https://codereview.appspot.com/6039048

git-svn-id: http://skia.googlecode.com/svn/trunk@3691 2bbb7eff-a529-9590-31e7-b0007b416f81
2012-04-16 16:27:09 +00:00
mike@reedtribe.org
6093e65829 dd DrawPathTest to gyp
add regression tests for bug 533



git-svn-id: http://skia.googlecode.com/svn/trunk@3678 2bbb7eff-a529-9590-31e7-b0007b416f81
2012-04-14 12:55:17 +00:00
reed@google.com
dceecc70a8 always allocate (only) the width of the clip for the AA row, since it will
already be clamped to 32K (our impl limit at the moment.)

add unittest to confirm this fix



git-svn-id: http://skia.googlecode.com/svn/trunk@3242 2bbb7eff-a529-9590-31e7-b0007b416f81
2012-02-23 19:20:19 +00:00