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
Share common code between bench_ and render_ to set up the PictureRenderer.
Fix an include error in SkPictureRenderer.h.
Simplified parameter passing in render_pictures_main.
Switch to using an SkAutoTUnref for the PictureRenderer.
I also changed the input format somewhat, so the buildbots need to be updated as well: https://codereview.appspot.com/7441044/
Fixed a bug in PictureBenchmark where calling setTimeIndividualTiles(false) sets the member variable to true.
Removed setDeviceType from PictureBenchmark, since only the PictureRenderer needs to know which device type to use.
Some changes to the input format:
'--logPerIter' no longer takes a 1 or 0. Instead, '--logPerIter'
turns it on and '--nologPerIter' turns it off (with off as the
default). (Note that this is for bench_pictures; bench still uses the old format)
Change '--device' to '--config' and 'bitmap' to '8888' to be
the same as gm.
Requires '--r' before inputs (to match gm), though there can be multiple inputs following it.
Changed --enable-deferred-image-decoding (which no one uses but me yet anyway) to --deferImageDecoding, since the former is incompatible with the flag parser.
Changes to behavior:
Show a short error message on failure (rather than the explanation of all
flags).
BUG=https://code.google.com/p/skia/issues/detail?id=1094
Review URL: https://codereview.appspot.com/7230053
git-svn-id: http://skia.googlecode.com/svn/trunk@7961 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
When timing individual tiles in bench_pictures, keep a timer running
across all repeats, and then take the average. The former method of
timing each iteration separately runs into precision errors on some
platforms.
Running on my Mac Pro with OSX 10.8, the cmsecs for the new method
and the old method are roughly the same when checking the CPU time.
When checking the wall time, the old method often gives me 0ms,
while the new method gives me a larger value. I don't think this
can be entirely attributed to rounding though, since on occasion I
see the old method showing a short time period (.05 - .15ms) while
the new method shows .15ms higher (which is in range for the
difference I'm seeing for other tiles where the old method reports
0ms).
Some other changes:
PictureRenderer::resetState now takes a boolean parameter. If called
with false, it will only do a flush, while if called with true, it
will also call finish.
resetState is now called with true everywhere except in between
iterations of drawing the same tile (when timing individual tiles).
render_pictures_main no longer calls resetState directly, since it
already calls end, which calls resetState.
BUG=http://code.google.com/p/skia/issues/detail?id=1066
Review URL: https://codereview.appspot.com/7101060
git-svn-id: http://skia.googlecode.com/svn/trunk@7424 2bbb7eff-a529-9590-31e7-b0007b416f81
now we will default to max diff of 256, which means that for now we
report all pixels that are not as expected and we do not error out.
Ideally we will decrease the value of max diff to something that does
not have visual impact, e.g. 10, then we will report small changes with
the intensity under 10, but we will error out for anything larger.
Review URL: https://codereview.appspot.com/7137046
git-svn-id: http://skia.googlecode.com/svn/trunk@7232 2bbb7eff-a529-9590-31e7-b0007b416f81
means the tool should return failure or not. For now it is not defined, which
means any failed pixels are still reported to stdout, but the tool does not
return an error, allowing the bots to go green (until we can fix these failures).
Review URL: https://codereview.appspot.com/7105056
git-svn-id: http://skia.googlecode.com/svn/trunk@7211 2bbb7eff-a529-9590-31e7-b0007b416f81
1) flag to verify PNGs produced by render_picture that produce the same pixels
as simple renderer.
2) flag to write one single image (in tile we write individual tiles) - this
will help running skdiff also
3) flag to clone the picture before rendering
git-svn-id: http://skia.googlecode.com/svn/trunk@6890 2bbb7eff-a529-9590-31e7-b0007b416f81
In TiledPictureRenderer and CopyTilesRenderer, do a postTranslate so
the translate is not affected by the scale factor.
Likewise, use clipRegion for tiled renderer so it will be
unaffected by the scale factor.
Respect the viewport for record, playbackCreation, and pipe renderers.
Review URL: https://codereview.appspot.com/6947072
git-svn-id: http://skia.googlecode.com/svn/trunk@6853 2bbb7eff-a529-9590-31e7-b0007b416f81
Builds on https://codereview.appspot.com/6718046/ by mtklein.
Previously, each iteration of drawing a picture started new threads to draw the picture. Since each thread is using thread local storage for the font cache, this means that each iteration had to start with an empty font cache.
The newly added MultiCorePictureRenderer, separated from TiledPictureRenderer, now starts the drawing threads at the beginning of the test using an SkThreadPool, and keeps them alive through all iterations, so the font cache can be reused.
For now, I have removed the pipe version of the threaded renderer.
Updated bench_pictures_main and render_pictures_main to use the new
renderer, and to unref a renderer before early exit.
Review URL: https://codereview.appspot.com/6777063
git-svn-id: http://skia.googlecode.com/svn/trunk@6285 2bbb7eff-a529-9590-31e7-b0007b416f81
and deserialization for encoding and decoding bitmaps.
Remove kForceFlattenBitmapPixels_Flag, which is no longer used.
When an SkOrderedReadBuffer needs to read a bitmap, if it does not
have an image decoder, use a dummy bitmap.
In GM, add a tolerance option for color differences, used when
testing picture serialization, so it can assume two images are the
same even though PNG encoding/decoding may have resulted in small
differences.
Create dummy implementations for SkImageDecoder and SkImageEncoder
functions in SkImageDecoder_empty so that a project that does not
want to include the images project it can still build.
Allow ports to build without images project.
In Mac's image encoder, copy 4444 to 8888 before encoding.
Add SkWriter32::reservePad, to provide a pointer to write non 4 byte
aligned data, padded with zeroes.
In bench_ and render_ pictures, pass decode function to SkPicture
creation from a stream.
BUG=https://code.google.com/p/skia/issues/detail?id=842
Review URL: https://codereview.appspot.com/6551071
git-svn-id: http://skia.googlecode.com/svn/trunk@5818 2bbb7eff-a529-9590-31e7-b0007b416f81
When doing tiled rendering in multiple threads, clone fPicture,
rather than the blank fPictureClones. Also fix a precedence
problem so we get the correct rectangle.
In render_pictures_main, call PictureRenderer::setup().
Review URL: https://codereview.appspot.com/6585055
git-svn-id: http://skia.googlecode.com/svn/trunk@5758 2bbb7eff-a529-9590-31e7-b0007b416f81
Since the passed in picture may represent an image which is too large
to be represented on the GPU, never create such a large canvas. Instead,
after drawing to each tile, create a file showing just that tile.
Review URL: https://codereview.appspot.com/6532056
git-svn-id: http://skia.googlecode.com/svn/trunk@5603 2bbb7eff-a529-9590-31e7-b0007b416f81
bench_pictures with "--device gpu" is failing because we're trying to allocate
too much GPU memory. Move the recently-added scaling code into picture_utils
and share it between render_pictures and bench_pictures.
Review URL: https://codereview.appspot.com/6495125
git-svn-id: http://skia.googlecode.com/svn/trunk@5543 2bbb7eff-a529-9590-31e7-b0007b416f81
Move SkBenchLogger into separate files and make bench_pictures use it.
Remove sk_tools::print_msg, since SkBenchLogger is now used instead.
Combine picture_benchmark with bench_pictures, since that is the
only project that uses it.
Refactor the aggregator for bench timer data into its own class and
make bench_pictures use it.
Consolidate the various virtual PictureBenchmark::run functions
into one for reuse.
BUG=https://code.google.com/p/skia/issues/detail?id=822
Review URL: https://codereview.appspot.com/6488086
git-svn-id: http://skia.googlecode.com/svn/trunk@5432 2bbb7eff-a529-9590-31e7-b0007b416f81
Further, the picture is also stored in the PictureRenderer.
The main gain of all of this is that we will not have to change how the device in more than one place when we end up adding different devices.
Review URL: https://codereview.appspot.com/6458074
git-svn-id: http://skia.googlecode.com/svn/trunk@4990 2bbb7eff-a529-9590-31e7-b0007b416f81
