This means you should not need to set any special compiler flags; which should
mean gcc will generate non-neon specific asm (unless you tell it to). This
means it is possible to build an armv6 binary with neon suppor (as we always
meant to to do).
SVN revision: 55307
* Remove vim modelines:
find . -name '*.[chx]' -exec sed -i '/\/\*$/ {N;N;/ \* vim:ts/d}' \{\} \;
find . -name '*.[chx]' -exec sed -i '/\/[\*\/] *vim:/d' \{\} \;
* Remove leading blank lines:
find . -name '*.[cxh]' -exec sed -i '/./,$!d'
If you use vim, use this in your .vimrc:
set ts=8 sw=3 sts=8 expandtab cino=>5n-3f0^-2{2(0W1st0
SVN revision: 50816
Samsung Electronics just allowed me to release the first set of ARM neon
patches under the Evas licence. They are silently helping improve EFL and
this another product of their help.
These patches have been tested on a Cortex A8 and show consistent improvement
across the board. For expedite some tests up to 100% improvements, and
practical real world examples show that rendering limited applications show
similar improvements. For instance in one application from 17fps->30fps or
for another 40->63fps.
The patches are pure neon code (intrinsics tend to generate worse code). To
build under GCC you will need a recent GCC and the following C flags:
-mfloat-abi=softfp -mfpu=neon
I also recommend -O2 and -ffast-math.
If you have any problems please let me know.
SVN revision: 48733
enabled. the blending is not working/complete. the neon for fills and copies
isnt actually faster though currently :(
2. scalecache infra - disabled for now. working on it.
SVN revision: 39723