Allow opt-out of implicit bounds-checking

[vchuravy]

KernelAbstractions currently creates kernels that look like:

if __validindex(ctx)
   # Body
end

This is problematic due to the convergence requirement on
@synchronize.

[vchuravy]

Warning

This pull request is not mergeable via GitHub because a downstack PR is open. Once all requirements are satisfied, merge this PR as a stack on Graphite.
Learn more

• Allow opt-out of implicit bounds-checking #563 👈 (View in Graphite)
• Implement a CPU backend using POCL #556 : 2 other dependent PRs (#558 , #562 )
• main

This stack of pull requests is managed by Graphite. Learn more about stacking.

[codecov]

Codecov Report

Attention: Patch coverage is 0% with 10 lines in your changes missing coverage. Please review.

Project coverage is 0.00%. Comparing base (f038d8c) to head (4dd0acc).

Files with missing lines	Patch %	Lines
src/macros.jl	0.00%	10 Missing ⚠️

Additional details and impacted files

@@           Coverage Diff           @@
##           vc/pocl    #563   +/-   ##
=======================================
  Coverage     0.00%   0.00%           
=======================================
  Files           21      21           
  Lines         1509    1513    +4     
=======================================
- Misses        1509    1513    +4     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[github-actions]

Benchmark Results

	main	4dd0acc...	main/4dd0accda3764a...
saxpy/default/Float16/1024	0.731 ± 0.006 μs	0.0553 ± 0.026 ms	0.0132
saxpy/default/Float16/1048576	0.174 ± 0.0017 ms	0.887 ± 0.023 ms	0.196
saxpy/default/Float16/16384	3.34 ± 0.048 μs	0.0615 ± 0.027 ms	0.0544
saxpy/default/Float16/2048	0.903 ± 0.0094 μs	0.0514 ± 0.023 ms	0.0176
saxpy/default/Float16/256	0.587 ± 0.0044 μs	0.0579 ± 0.027 ms	0.0101
saxpy/default/Float16/262144	0.0442 ± 0.00058 ms	0.27 ± 0.026 ms	0.164
saxpy/default/Float16/32768	6.02 ± 0.092 μs	0.0741 ± 0.027 ms	0.0812
saxpy/default/Float16/4096	1.31 ± 0.023 μs	0.0576 ± 0.025 ms	0.0227
saxpy/default/Float16/512	0.647 ± 0.0052 μs	0.0559 ± 0.026 ms	0.0116
saxpy/default/Float16/64	0.556 ± 0.0041 μs	0.0621 ± 0.026 ms	0.00896
saxpy/default/Float16/65536	11.7 ± 0.14 μs	0.103 ± 0.027 ms	0.114
saxpy/default/Float32/1024	0.628 ± 0.008 μs	0.0565 ± 0.026 ms	0.0111
saxpy/default/Float32/1048576	0.23 ± 0.0099 ms	0.469 ± 0.038 ms	0.491
saxpy/default/Float32/16384	2.96 ± 0.46 μs	0.0543 ± 0.024 ms	0.0545
saxpy/default/Float32/2048	0.752 ± 0.028 μs	0.049 ± 0.023 ms	0.0153
saxpy/default/Float32/256	0.567 ± 0.0048 μs	0.0561 ± 0.027 ms	0.0101
saxpy/default/Float32/262144	0.0577 ± 0.0025 ms	0.158 ± 0.035 ms	0.364
saxpy/default/Float32/32768	5.64 ± 0.72 μs	0.0598 ± 0.026 ms	0.0944
saxpy/default/Float32/4096	1.14 ± 0.12 μs	0.0529 ± 0.025 ms	0.0215
saxpy/default/Float32/512	0.6 ± 0.0063 μs	0.0573 ± 0.027 ms	0.0105
saxpy/default/Float32/64	0.554 ± 0.0049 μs	0.0566 ± 0.027 ms	0.00979
saxpy/default/Float32/65536	12.7 ± 0.93 μs	0.0744 ± 0.029 ms	0.17
saxpy/default/Float64/1024	0.747 ± 0.018 μs	0.0559 ± 0.026 ms	0.0134
saxpy/default/Float64/1048576	0.484 ± 0.017 ms	0.5 ± 0.047 ms	0.969
saxpy/default/Float64/16384	5.41 ± 0.51 μs	0.0558 ± 0.025 ms	0.0971
saxpy/default/Float64/2048	1.16 ± 0.14 μs	0.0497 ± 0.024 ms	0.0234
saxpy/default/Float64/256	0.586 ± 0.0058 μs	0.0592 ± 0.026 ms	0.00989
saxpy/default/Float64/262144	0.115 ± 0.0065 ms	0.171 ± 0.03 ms	0.669
saxpy/default/Float64/32768	12.5 ± 0.67 μs	0.0633 ± 0.025 ms	0.198
saxpy/default/Float64/4096	1.78 ± 0.26 μs	0.0546 ± 0.025 ms	0.0326
saxpy/default/Float64/512	0.635 ± 0.0079 μs	0.0523 ± 0.027 ms	0.0121
saxpy/default/Float64/64	0.564 ± 0.0043 μs	0.0596 ± 0.027 ms	0.00947
saxpy/default/Float64/65536	28.6 ± 1.4 μs	0.0835 ± 0.026 ms	0.342
saxpy/static workgroup=(1024,)/Float16/1024	2.18 ± 0.028 μs	0.0479 ± 0.026 ms	0.0456
saxpy/static workgroup=(1024,)/Float16/1048576	0.159 ± 0.0031 ms	0.898 ± 0.028 ms	0.177
saxpy/static workgroup=(1024,)/Float16/16384	4.43 ± 0.081 μs	0.0584 ± 0.025 ms	0.0758
saxpy/static workgroup=(1024,)/Float16/2048	2.36 ± 0.027 μs	0.0567 ± 0.023 ms	0.0416
saxpy/static workgroup=(1024,)/Float16/256	2.81 ± 0.037 μs	0.047 ± 0.026 ms	0.0599
saxpy/static workgroup=(1024,)/Float16/262144	0.0423 ± 0.00084 ms	0.268 ± 0.028 ms	0.158
saxpy/static workgroup=(1024,)/Float16/32768	6.87 ± 0.19 μs	0.0719 ± 0.025 ms	0.0956
saxpy/static workgroup=(1024,)/Float16/4096	2.68 ± 0.04 μs	0.0499 ± 0.026 ms	0.0538
saxpy/static workgroup=(1024,)/Float16/512	3.25 ± 0.033 μs	0.0489 ± 0.026 ms	0.0666
saxpy/static workgroup=(1024,)/Float16/64	2.51 ± 0.21 μs	0.05 ± 0.027 ms	0.0502
saxpy/static workgroup=(1024,)/Float16/65536	12.7 ± 0.33 μs	0.0994 ± 0.025 ms	0.127
saxpy/static workgroup=(1024,)/Float32/1024	2.23 ± 0.028 μs	0.0534 ± 0.026 ms	0.0418
saxpy/static workgroup=(1024,)/Float32/1048576	0.234 ± 0.02 ms	0.457 ± 0.027 ms	0.511
saxpy/static workgroup=(1024,)/Float32/16384	4.39 ± 0.26 μs	0.0518 ± 0.024 ms	0.0848
saxpy/static workgroup=(1024,)/Float32/2048	2.4 ± 0.06 μs	0.0467 ± 0.023 ms	0.0513
saxpy/static workgroup=(1024,)/Float32/256	2.69 ± 0.043 μs	0.0559 ± 0.026 ms	0.0481
saxpy/static workgroup=(1024,)/Float32/262144	0.0606 ± 0.0031 ms	0.153 ± 0.034 ms	0.395
saxpy/static workgroup=(1024,)/Float32/32768	7.59 ± 0.4 μs	0.0573 ± 0.025 ms	0.133
saxpy/static workgroup=(1024,)/Float32/4096	2.68 ± 0.082 μs	0.0486 ± 0.026 ms	0.055
saxpy/static workgroup=(1024,)/Float32/512	2.71 ± 0.028 μs	0.0564 ± 0.026 ms	0.048
saxpy/static workgroup=(1024,)/Float32/64	2.72 ± 5.2 μs	0.0557 ± 0.026 ms	0.0487
saxpy/static workgroup=(1024,)/Float32/65536	15.5 ± 0.85 μs	0.0709 ± 0.028 ms	0.219
saxpy/static workgroup=(1024,)/Float64/1024	2.32 ± 0.07 μs	0.0529 ± 0.026 ms	0.0439
saxpy/static workgroup=(1024,)/Float64/1048576	0.501 ± 0.021 ms	0.496 ± 0.05 ms	1.01
saxpy/static workgroup=(1024,)/Float64/16384	7.31 ± 0.32 μs	0.0527 ± 0.025 ms	0.139
saxpy/static workgroup=(1024,)/Float64/2048	2.61 ± 0.076 μs	0.0437 ± 0.022 ms	0.0596
saxpy/static workgroup=(1024,)/Float64/256	2.64 ± 0.055 μs	0.0557 ± 0.026 ms	0.0473
saxpy/static workgroup=(1024,)/Float64/262144	0.117 ± 0.0054 ms	0.166 ± 0.031 ms	0.706
saxpy/static workgroup=(1024,)/Float64/32768	15.2 ± 0.86 μs	0.0607 ± 0.025 ms	0.251
saxpy/static workgroup=(1024,)/Float64/4096	3.16 ± 0.19 μs	0.0504 ± 0.026 ms	0.0628
saxpy/static workgroup=(1024,)/Float64/512	2.64 ± 0.06 μs	0.0418 ± 0.027 ms	0.0632
saxpy/static workgroup=(1024,)/Float64/64	2.6 ± 0.066 μs	0.0434 ± 0.027 ms	0.0598
saxpy/static workgroup=(1024,)/Float64/65536	31.3 ± 1.1 μs	0.08 ± 0.027 ms	0.391
time_to_load	0.315 ± 0.00096 s	1.1 ± 0.0043 s	0.286

Benchmark Plots

A plot of the benchmark results have been uploaded as an artifact to the workflow run for this PR.
Go to "Actions"->"Benchmark a pull request"->[the most recent run]->"Artifacts" (at the bottom).