Add LinuxPerf support, add text-only logs summarizer, add more metrics

[jpsamaroo]

Written by Claude

Todo:

• Add docs for LinuxPerf metrics and :summary visualizer
• Basic test for :summary text output

[github-actions]

Dagger benchmarks: dirty vs master

	master	dirty	master / dirty
array/dagger/N=1024 (block 512)/add (X + X)	10.4 ± 0.59 ms	10.7 ± 0.02 ms	0.978 ± 0.055
array/dagger/N=1024 (block 512)/alloc (rand)	4.41 ± 0.43 ms	4.44 ± 1.8 ms	0.993 ± 0.41
array/dagger/N=1024 (block 512)/broadcast (X .+ 1)	6.51 ± 0.99 ms	7.44 ± 0.2 ms	0.875 ± 0.13
array/dagger/N=1024 (block 512)/map (sin.(X))	10.8 ± 2.7 ms	12.5 ± 1.6 ms	0.864 ± 0.24
array/dagger/N=1024 (block 512)/norm	3.88 ± 0.19 ms	4.59 ± 1 ms	0.845 ± 0.2
array/dagger/N=1024 (block 512)/reduce (sum)	7.17 ± 1.3 ms	7.83 ± 1.8 ms	0.916 ± 0.27
array/dagger/N=1024 (block 512)/transpose (permutedims)	10.4 ± 0.25 ms	10.9 ± 1.3 ms	0.951 ± 0.12
array/dagger/N=256 (block 256)/add (X + X)	3.73 ± 0.15 ms	3.31 ± 0.34 ms	1.13 ± 0.13
array/dagger/N=256 (block 256)/alloc (rand)	1.73 ± 0.51 ms	1.54 ± 0.036 ms	1.12 ± 0.33
array/dagger/N=256 (block 256)/broadcast (X .+ 1)	3.4 ± 0.33 ms	3.16 ± 0.073 ms	1.08 ± 0.11
array/dagger/N=256 (block 256)/map (sin.(X))	3.99 ± 0.22 ms	3.66 ± 0.078 ms	1.09 ± 0.065
array/dagger/N=256 (block 256)/norm	2.12 ± 0.17 ms	2.29 ± 0.17 ms	0.925 ± 0.1
array/dagger/N=256 (block 256)/reduce (sum)	4.06 ± 0.35 ms	4.51 ± 0.74 ms	0.9 ± 0.17
array/dagger/N=256 (block 256)/transpose (permutedims)	2.58 ± 0.048 ms	3.3 ± 0.12 ms	0.78 ± 0.033
linalg/dagger/N=1024 (block 512)/cholesky	0.0341 ± 0.0033 s	31.1 ± 1.8 ms	1.1 ± 0.12
linalg/dagger/N=1024 (block 512)/lu	7.73 s	7.69 s	1.01
linalg/dagger/N=1024 (block 512)/matmul (A*A)	0.0577 ± 0.0055 s	0.0605 ± 0.0051 s	0.954 ± 0.12
linalg/dagger/N=1024 (block 512)/matvec (A*x)	9.69 ± 1.2 ms	9 ± 0.79 ms	1.08 ± 0.16
linalg/dagger/N=1024 (block 512)/qr	0.399 ± 0.0031 s	0.398 ± 0.008 s	1 ± 0.022
linalg/dagger/N=1024 (block 512)/solve (A\b via lu)	11.1 s	11 s	1.01
linalg/dagger/N=1024 (block 512)/syrk (A'*A)	0.0447 ± 0.0031 s	0.0464 ± 0.005 s	0.964 ± 0.12
linalg/dagger/N=256 (block 256)/cholesky	6.91 ± 0.59 ms	6.81 ± 0.19 ms	1.01 ± 0.091
linalg/dagger/N=256 (block 256)/lu	1.35 ± 0.012 s	1.34 ± 0.021 s	1.01 ± 0.018
linalg/dagger/N=256 (block 256)/matmul (A*A)	4.28 ± 0.52 ms	4.51 ± 0.39 ms	0.951 ± 0.14
linalg/dagger/N=256 (block 256)/matvec (A*x)	3.55 ± 0.57 ms	3.53 ± 0.42 ms	1.01 ± 0.2
linalg/dagger/N=256 (block 256)/qr	10.8 ± 0.4 ms	11 ± 0.082 ms	0.981 ± 0.037
linalg/dagger/N=256 (block 256)/solve (A\b via lu)	2.23 ± 0.012 s	2.21 ± 0.0071 s	1.01 ± 0.0063
linalg/dagger/N=256 (block 256)/syrk (A'*A)	6.51 ± 0.86 ms	6.16 ± 0.26 ms	1.06 ± 0.15
stencil/dagger/N=1024 (block 512)/alloc (neighbors Wrap)	0.112 ± 0.013 s	0.0606 ± 0.0021 s	1.85 ± 0.23
stencil/dagger/N=1024 (block 512)/assign (const)	4.36 ± 0.28 ms	4.24 ± 0.38 ms	1.03 ± 0.11
stencil/dagger/N=1024 (block 512)/multi-expr	9.73 ± 0.67 ms	9.8 ± 0.59 ms	0.993 ± 0.091
stencil/dagger/N=1024 (block 512)/neighbors (Clamp)	0.0572 ± 0.0027 s	0.0514 ± 0.0016 s	1.11 ± 0.064
stencil/dagger/N=1024 (block 512)/neighbors (Pad)	0.054 ± 0.006 s	0.0583 ± 0.0037 s	0.927 ± 0.12
stencil/dagger/N=1024 (block 512)/neighbors (Reflect)	0.0533 ± 0.0059 s	0.0542 ± 0.0045 s	0.985 ± 0.14
stencil/dagger/N=1024 (block 512)/neighbors (Wrap)	0.0591 ± 0.0048 s	0.0562 ± 0.003 s	1.05 ± 0.1
stencil/dagger/N=1024 (block 512)/update (+)	6.33 ± 0.27 ms	6.26 ± 0.2 ms	1.01 ± 0.054
stencil/dagger/N=256 (block 256)/alloc (neighbors Wrap)	17.7 ± 0.79 ms	14.9 ± 0.27 ms	1.19 ± 0.057
stencil/dagger/N=256 (block 256)/assign (const)	2.68 ± 0.032 ms	2.54 ± 0.047 ms	1.06 ± 0.023
stencil/dagger/N=256 (block 256)/multi-expr	5.09 ± 0.26 ms	4.47 ± 0.26 ms	1.14 ± 0.088
stencil/dagger/N=256 (block 256)/neighbors (Clamp)	13.9 ± 0.4 ms	14.1 ± 0.28 ms	0.979 ± 0.035
stencil/dagger/N=256 (block 256)/neighbors (Pad)	13.9 ± 0.67 ms	13.7 ± 0.93 ms	1.02 ± 0.085
stencil/dagger/N=256 (block 256)/neighbors (Reflect)	13.9 ± 0.15 ms	13.7 ± 0.61 ms	1.02 ± 0.047
stencil/dagger/N=256 (block 256)/neighbors (Wrap)	13.6 ± 0.12 ms	14 ± 0.13 ms	0.971 ± 0.012
stencil/dagger/N=256 (block 256)/update (+)	3 ± 0.72 ms	3.19 ± 0.68 ms	0.942 ± 0.3
time_to_load	1.06 ± 0.027 s	1.05 ± 0.016 s	1.01 ± 0.03

Plots

⚠️ Regressions (> 25.0%)

• array/dagger/N=256 (block 256)/transpose (permutedims): +28.2%

Improvements (> 25.0% faster)

• stencil/dagger/N=1024 (block 512)/alloc (neighbors Wrap): -46.0%

Full results and plots (download the benchmark-results artifact).