Unprotected reads -> relaxed reads.
Unprotected write -> relaxed write.
The only unprotected write we had was in SkTraceEvent, which it looks like we nabbed from Chrome at some point and changed only to silence TSAN. Chrome's version uses AtomicWord / NoBarrier_Load / NoBarrier_Store, which boils down to the same as here, intptr_t / relaxed load / relaxed store.
This leaves one place where we're lying a bit to TSAN, in include/core/SkLazyPtr.h where we're doing a data-dependent consume load. We're telling TSAN it's consume, but telling any other compiler to compile it as relaxed, given how they all upgrade consume to acquire. This eliminates a barrier for us on ARM. How do you guys deal with this? Just use a consume memory order, take the hit, and hope compilers get smarter one day?
BUG=chromium:465721
No public API changes.
TBR=reed@google.com
Review URL: https://codereview.chromium.org/996763002
1 Remove atExit feature: clients can do it just as well as SkOnce can.
2 Remove support for functors: no one but the unit test did that.
3 Remove support for unused non-static SkOnceFlag (no SK_ONCE_INIT).
4 Add SkOnce variants for no-arg functions so we're not forced to pass dummy values all the time.
5 Merge SkSpinlock and SkOnceFlag, making all members private.
6 More notes about memory barriers, adding an acquire load after acquiring the spinlock.
BUG=skia:
R=bungeman@google.com, mtklein@google.com, reed@google.com
Author: mtklein@chromium.org
Review URL: https://codereview.chromium.org/302083003
Aarch64 support
This change contains the necessary modifications to have Skia build and
run properly on an ARMv8 processor in aarch64 execution state.
Here's a list of the changes:
- add an arm64 target to the build system + SK_CPU_ARM64 flag
- MatrixTest was failing when built in Release mode. Fused MAC
instructions were generated which made some intermediate results
more accurate. As the test relies on result comparison, the more
precise results when compared to others led to a gap bigger than
what was tolerated. As I don't know if some actual skia code relies
on results being comparable, I've disabled fused MAC instruction
with -ffp-contract=off for arm64.
- Modify include/core/SkOnce.h to have barriers work.
- SK_CPU_ARM64 implies SK_ARM_NEON_MODE_ALWAYS.
- use existing Xfermode optimisations with modifications that can be
removed in the future when toolchains are ready. Also save a few
instructions is two Xfermodes (will apply to ARM too).
- use existing SkBoxBlur and SkMorphology optimisations.
- use existing SkBlitMask optimisations
- use existing BitmapProcState and Convolution optimisations.
Future changes will include:
- Blitters (only partialy merged upstream)
- SkUtils (there's little value in sending asm optimisations without
having them benchmarked on real hardware).
Signed-off-by: Kevin PETIT <kevin.petit@arm.com>
BUG=skia:
Committed: http://code.google.com/p/skia/source/detail?r=13980R=djsollen@google.com, reed@google.com, mtklein@google.com, halcanary@google.com
Author: kevin.petit@arm.com
Review URL: https://codereview.chromium.org/143423004
git-svn-id: http://skia.googlecode.com/svn/trunk@14025 2bbb7eff-a529-9590-31e7-b0007b416f81
Reason for revert:
GYP's failing on most (all?) bots.
Original issue's description:
> ARM Skia NEON patches - 35 - First AArch64 support
>
> Aarch64 support
>
> This change contains the necessary modifications to have Skia build and
> run properly on an ARMv8 processor in aarch64 execution state.
>
> Here's a list of the changes:
>
> - add an arm64 target to the build system + SK_CPU_ARM64 flag
>
> - MatrixTest was failing when built in Release mode. Fused MAC
> instructions were generated which made some intermediate results
> more accurate. As the test relies on result comparison, the more
> precise results when compared to others led to a gap bigger than
> what was tolerated. As I don't know if some actual skia code relies
> on results being comparable, I've disabled fused MAC instruction
> with -ffp-contract=off for arm64.
>
> - Modify include/core/SkOnce.h to have barriers work.
>
> - SK_CPU_ARM64 implies SK_ARM_NEON_MODE_ALWAYS.
>
> - use existing Xfermode optimisations with modifications that can be
> removed in the future when toolchains are ready. Also save a few
> instructions is two Xfermodes (will apply to ARM too).
>
> - use existing SkBoxBlur and SkMorphology optimisations.
>
> - use existing SkBlitMask optimisations
>
> - use existing BitmapProcState and Convolution optimisations.
>
> Future changes will include:
>
> - Blitters (only partialy merged upstream)
>
> - SkUtils (there's little value in sending asm optimisations without
> having them benchmarked on real hardware).
>
> Signed-off-by: Kevin PETIT <kevin.petit@arm.com>
>
> BUG=skia:
>
> Committed: http://code.google.com/p/skia/source/detail?r=13980R=djsollen@google.com, reed@google.com, halcanary@google.com, kevin.petit@arm.comTBR=djsollen@google.com, halcanary@google.com, kevin.petit@arm.com, reed@google.com
NOTREECHECKS=true
NOTRY=true
BUG=skia:
Author: mtklein@google.com
Review URL: https://codereview.chromium.org/216113005
git-svn-id: http://skia.googlecode.com/svn/trunk@13983 2bbb7eff-a529-9590-31e7-b0007b416f81
Aarch64 support
This change contains the necessary modifications to have Skia build and
run properly on an ARMv8 processor in aarch64 execution state.
Here's a list of the changes:
- add an arm64 target to the build system + SK_CPU_ARM64 flag
- MatrixTest was failing when built in Release mode. Fused MAC
instructions were generated which made some intermediate results
more accurate. As the test relies on result comparison, the more
precise results when compared to others led to a gap bigger than
what was tolerated. As I don't know if some actual skia code relies
on results being comparable, I've disabled fused MAC instruction
with -ffp-contract=off for arm64.
- Modify include/core/SkOnce.h to have barriers work.
- SK_CPU_ARM64 implies SK_ARM_NEON_MODE_ALWAYS.
- use existing Xfermode optimisations with modifications that can be
removed in the future when toolchains are ready. Also save a few
instructions is two Xfermodes (will apply to ARM too).
- use existing SkBoxBlur and SkMorphology optimisations.
- use existing SkBlitMask optimisations
- use existing BitmapProcState and Convolution optimisations.
Future changes will include:
- Blitters (only partialy merged upstream)
- SkUtils (there's little value in sending asm optimisations without
having them benchmarked on real hardware).
Signed-off-by: Kevin PETIT <kevin.petit@arm.com>
BUG=skia:
R=djsollen@google.com, reed@google.com, mtklein@google.com, halcanary@google.com
Author: kevin.petit@arm.com
Review URL: https://codereview.chromium.org/143423004
git-svn-id: http://skia.googlecode.com/svn/trunk@13980 2bbb7eff-a529-9590-31e7-b0007b416f81
Make leak counters implemented with SK_DECLARE_INST_COUNT thread-safe.
Enable the leak counting for Debug builds when Skia is built as a
static library. Having SK_DECLARE_INST_COUNT without
SK_DEFINE_INST_COUNT relies on static variables in member functions
declared in the header files. These might be duplicated in the clients
of the library when Skia is built as a dynamic library, producing
incorrect operation.
Protect the instance counter initialization step (initStep) by
using SkOnce.
Makes SkOnce.h part of the public API, since SkInstCnt is public.
Protect the per-class child list shared variable with a per-class mutex.
Changes the behavior in the way that if the child list has been
"cleaned up", it will still try to create subsequent child lists.
BUG=skia:1219
R=robertphillips@google.com, mtklein@google.com, bsalomon@google.com, bungeman@google.com, djsollen@google.com
Author: kkinnunen@nvidia.com
Review URL: https://codereview.chromium.org/99483003
git-svn-id: http://skia.googlecode.com/svn/trunk@13120 2bbb7eff-a529-9590-31e7-b0007b416f81
