Backports for KernelAbstractions 0.9.35

[vchuravy]

• prevent get_backend from overflowing the stack #602
• Test correct backend in examples test #597

[github-actions]

Benchmark Results

	main	69198a1...	main / 69198a1...
saxpy/default/Float32/1024	0.0565 ± 0.029 ms	0.648 ± 0.011 μs	87.2 ± 46
saxpy/default/Float32/1048576	0.455 ± 0.021 ms	0.262 ± 0.027 ms	1.74 ± 0.19
saxpy/default/Float32/16384	0.0582 ± 0.029 ms	3.07 ± 0.92 μs	19 ± 11
saxpy/default/Float32/2048	0.0453 ± 0.027 ms	0.76 ± 0.069 μs	59.5 ± 37
saxpy/default/Float32/256	0.0623 ± 0.029 ms	0.574 ± 0.0063 μs	109 ± 50
saxpy/default/Float32/262144	0.157 ± 0.026 ms	0.0572 ± 0.0045 ms	2.75 ± 0.5
saxpy/default/Float32/32768	0.0636 ± 0.028 ms	6.1 ± 1.5 μs	10.4 ± 5.2
saxpy/default/Float32/4096	0.0481 ± 0.029 ms	1.12 ± 0.071 μs	43.2 ± 26
saxpy/default/Float32/512	0.0649 ± 0.028 ms	0.61 ± 0.0095 μs	106 ± 46
saxpy/default/Float32/64	0.0491 ± 0.029 ms	0.563 ± 0.0064 μs	87.2 ± 52
saxpy/default/Float32/65536	0.0754 ± 0.028 ms	13.4 ± 1.4 μs	5.63 ± 2.2
saxpy/default/Float64/1024	0.0516 ± 0.029 ms	0.761 ± 0.035 μs	67.8 ± 39
saxpy/default/Float64/1048576	0.591 ± 0.082 ms	0.535 ± 0.068 ms	1.11 ± 0.21
saxpy/default/Float64/16384	0.0673 ± 0.028 ms	5.67 ± 1.2 μs	11.9 ± 5.5
saxpy/default/Float64/2048	0.0438 ± 0.026 ms	1.12 ± 0.082 μs	39 ± 23
saxpy/default/Float64/256	0.062 ± 0.029 ms	0.59 ± 0.0063 μs	105 ± 49
saxpy/default/Float64/262144	0.169 ± 0.026 ms	0.117 ± 0.012 ms	1.45 ± 0.27
saxpy/default/Float64/32768	0.0683 ± 0.027 ms	13 ± 1.3 μs	5.25 ± 2.1
saxpy/default/Float64/4096	0.0492 ± 0.027 ms	1.74 ± 0.25 μs	28.3 ± 16
saxpy/default/Float64/512	0.0603 ± 0.029 ms	0.643 ± 0.011 μs	93.8 ± 46
saxpy/default/Float64/64	0.0551 ± 0.03 ms	0.564 ± 0.0053 μs	97.7 ± 54
saxpy/default/Float64/65536	0.0867 ± 0.027 ms	28.8 ± 2.9 μs	3.01 ± 0.98
saxpy/static workgroup=(1024,)/Float32/1024	0.0433 ± 0.029 ms	2.12 ± 0.03 μs	20.5 ± 14
saxpy/static workgroup=(1024,)/Float32/1048576	0.45 ± 0.02 ms	0.248 ± 0.024 ms	1.81 ± 0.19
saxpy/static workgroup=(1024,)/Float32/16384	0.0536 ± 0.027 ms	4.62 ± 0.75 μs	11.6 ± 6.1
saxpy/static workgroup=(1024,)/Float32/2048	0.0429 ± 0.026 ms	2.26 ± 0.047 μs	19 ± 12
saxpy/static workgroup=(1024,)/Float32/256	0.0558 ± 0.027 ms	2.61 ± 0.041 μs	21.4 ± 10
saxpy/static workgroup=(1024,)/Float32/262144	0.158 ± 0.027 ms	0.0602 ± 0.0053 ms	2.62 ± 0.5
saxpy/static workgroup=(1024,)/Float32/32768	0.0589 ± 0.026 ms	7.86 ± 1 μs	7.5 ± 3.5
saxpy/static workgroup=(1024,)/Float32/4096	0.0482 ± 0.028 ms	2.54 ± 0.067 μs	18.9 ± 11
saxpy/static workgroup=(1024,)/Float32/512	0.06 ± 0.028 ms	2.63 ± 0.037 μs	22.8 ± 11
saxpy/static workgroup=(1024,)/Float32/64	0.0539 ± 0.027 ms	2.63 ± 5.1 μs	20.5 ± 41
saxpy/static workgroup=(1024,)/Float32/65536	0.073 ± 0.026 ms	15.6 ± 1.4 μs	4.67 ± 1.7
saxpy/static workgroup=(1024,)/Float64/1024	0.0441 ± 0.03 ms	2.25 ± 0.053 μs	19.6 ± 13
saxpy/static workgroup=(1024,)/Float64/1048576	0.572 ± 0.064 ms	0.56 ± 0.042 ms	1.02 ± 0.14
saxpy/static workgroup=(1024,)/Float64/16384	0.063 ± 0.026 ms	7.69 ± 0.96 μs	8.19 ± 3.6
saxpy/static workgroup=(1024,)/Float64/2048	0.0427 ± 0.026 ms	2.52 ± 0.066 μs	16.9 ± 10
saxpy/static workgroup=(1024,)/Float64/256	0.0622 ± 0.027 ms	2.6 ± 0.056 μs	24 ± 10
saxpy/static workgroup=(1024,)/Float64/262144	0.169 ± 0.026 ms	0.102 ± 0.014 ms	1.67 ± 0.35
saxpy/static workgroup=(1024,)/Float64/32768	0.0668 ± 0.026 ms	15.7 ± 1.4 μs	4.25 ± 1.7
saxpy/static workgroup=(1024,)/Float64/4096	0.048 ± 0.026 ms	3.19 ± 0.22 μs	15.1 ± 8.3
saxpy/static workgroup=(1024,)/Float64/512	0.0585 ± 0.029 ms	2.6 ± 0.064 μs	22.5 ± 11
saxpy/static workgroup=(1024,)/Float64/64	0.0581 ± 0.027 ms	2.56 ± 0.067 μs	22.7 ± 11
saxpy/static workgroup=(1024,)/Float64/65536	0.0885 ± 0.027 ms	29.1 ± 5.9 μs	3.04 ± 1.1
time_to_load	1.37 ± 0.0047 s	0.319 ± 0.004 s	4.31 ± 0.056

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).

[christiangnrd]

LGTM