Partially in preparation for building libpng on Windows.
Also, this makes us consistent across platforms for PDF.
Uses the version of zlib checked into the Chromium tree.
Remove miniz, which is replaced by zlib.
Review URL: https://codereview.chromium.org/966963002
Also, Revert "Revert of Flate: fix valgrind miniz
Conditional-jump-or-move-depends... error"
This reverts commit 6dc4ef01a6.
The alloc functions really do have different types depending
on zlib implementation.
CQ_INCLUDE_TRYBOTS=client.skia:Linux Builder-Trybot
Review URL: https://codereview.chromium.org/963853002
Add SkPDFCanon::reset function to unref all objects.
No longer possible to remove object from canon
Motivation: this doesn't change these object's lifetime, (they will
still be fully unrefed when SkDocument::close() is called, but we no
longer have to remove them from the array when their destructor is
called.
Review URL: https://codereview.chromium.org/966863002
Also fix a clip transform issue: clips are tracked in device space,
but if applied directly to leaf elements they are also subject to local
transforms ("transform='...'"). Instead, apply via <g> wrapper elements.
R=reed@google.com,mtklein@google.com
Review URL: https://codereview.chromium.org/959883003
/analyze recently reported two new warnings about reading from
potentially uninitialized local variables:
src\third_party\skia\src\gpu\grclipmaskmanager.cpp(290) :
warning C6001: Using uninitialized memory 'requiresAA'.
src\third_party\skia\src\gpu\grclipmaskmanager.cpp(336) :
warning C6001: Using uninitialized memory 'genID'.
It is not clear whether the uninitialized reads can actually happen,
but the guarantees seem sufficiently non-obvious that the prudent thing
to do to guarantee future stability is to initialized them -- it's
cheap. I also initialized initialState because it seemed to fall in
the same class, despite there being no warning for it.
BUG=427616
Review URL: https://codereview.chromium.org/957133002
- Adds miniz.c v115_r4 (latest release) to third_party.
- Merges SkDeflateWStream into SkFlate so including "miniz.c" links
without duplicating symbols.
The only interesting code change I've made is to remove the line
fImpl->fZStream.data_type = Z_BINARY;
from SkDeflateWStream::SkDeflateWStream(). miniz doesn't have Z_BINARY
defined, and as far as I can tell, both zlib and miniz ignore data_type.
We should be able to swap skflate.gyp's dependency between zlib.gyp:zlib and
zlib.gyp:miniz at will (except of course on Windows) if we're interested in
zlib itself. I've left android framework on its own zlib. I think this all
means we can stop defining SK_NO_FLATE on Windows.
I'll leave the possible cleanup of SK_NO_FLATE itself for another time. Might
be we always want to keep this dependency optional.
CQ_EXTRA_TRYBOTS=client.skia:Test-Win8-ShuttleA-HD7770-x86-Debug-Trybot
BUG=skia:
Review URL: https://codereview.chromium.org/957323003
This doesn't add them to the second-stringer Sk4i. It's unclear we should be
doing that often, and we don't have efficient ways to do it except via floats.
BUG=skia:
Review URL: https://codereview.chromium.org/964603002
This path renderer converts paths to linear contours, resolves intersections via Bentley-Ottman, implements a trapezoidal decomposition a la Fournier and Montuno to produce triangles, and renders those with a single draw call. It does not currently do antialiasing, so it must be used in conjunction with multisampling.
A fair amount of the code is to handle floating point edge cases in intersections. Rather than perform exact computations (which would require arbitrary precision arithmetic), we reconnect the mesh to reflect the intersection points. For example, intersections can occur above the current vertex, and force edges to be merged into the current vertex, requiring a restart of the intersections. Splitting edges for intersections can also force them to merge with formerly-distinct edges in the same polygon, or to violate the ordering of the active edge list, or the active edge state of split edges.
BUG=skia:
Review URL: https://codereview.chromium.org/855513004
Reason for revert:
TSAN troubles.
Original issue's description:
> Use an image generator to back SkPictureShader tiles.
>
> To avoid lifetime issues for tiles backed by discardable memory, use an
> image generator to re-generate them on the fly.
>
> With this CL, we are now caching bitmap shaders wrapping discardable
> pixel ref bitmaps backed by picture image generators.
>
> (the CL also includes some minor/unrelated SkPictureShader cleanup)
>
> BUG=skia:3220
> R=reed@google.com,halcanary@google.com
>
> Committed: https://skia.googlesource.com/skia/+/4739955e98ba86900a5bbac0e3661601427a54d4TBR=reed@google.com,halcanary@google.com,mtklein@google.com,reed@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=skia:3220
Review URL: https://codereview.chromium.org/956143002
We've always been using the portable ColorRect32, so we don't need the
ColorRectProc plumbing.
Furthermore, ColorRect32 doesn't seem to be very important (we're only using
it in the opaque case, which our row-by-row procs already specialize for).
Remove that too.
If we find we want specialization for really narrow rects again, let's put it in
blitRect() directly. It's pretty unlikely we're going to get platform-specific
speedup for blits to non-contiguous memory.
My local SKP comparison is +- 3%... most neutral I've ever seen.
BUG=skia:
Review URL: https://codereview.chromium.org/959873002
There's a small window where two threads can both see the gen ID is zero,
both go get new distinct genIDs, but race to write them.
This now uses compare_exchange to read and leave the winner's ID in place
when there's a race.
SkAtomics isn't public.
TBR=reed@google.com
BUG=skia:
Review URL: https://codereview.chromium.org/956013002
