GLBenches do not expect gl state to change between onPerCanvasPreDraw and *PostDraw, but we do a clear and sometimes we clear as draw. This causes us to bind vertex objects / programs / etc.
This change creates two new virtual methods which are called right before and immediately after timing.
BUG=skia:
Review URL: https://codereview.chromium.org/1379853003
This fixes every case where virtual and SK_OVERRIDE were on the same line,
which should be the bulk of cases. We'll have to manually clean up the rest
over time unless I level up in regexes.
for f in (find . -type f); perl -p -i -e 's/virtual (.*)SK_OVERRIDE/\1SK_OVERRIDE/g' $f; end
BUG=skia:
Review URL: https://codereview.chromium.org/806653007
First bench: Simply render from a compressed ETC1 bitmap. This is roughly
equal between compressed and uncompressed data.
Second bench: Notify the bitmap that the pixels have changed (that's right,
like a liar) to force a re-upload to the GPU. On this bench, decompressed
textures seem to inexplicably do an order of magnitude better than
compressed textures when it should be the other way around, but that
investigation is reserved for a future CL.
R=robertphillips@google.com
Author: krajcevski@google.com
Review URL: https://codereview.chromium.org/317023002
