No precomp

[chriselrod]

Precompilation kills runtime performance because of Julia bugs.
Master:

julia> @benchmark matmul!($Cf64, $Af64, $Bf64)
BenchmarkTools.Trial: 3727 samples with 1 evaluation.
 Range (min … max):  258.678 μs … 279.208 μs  ┊ GC (min … max): 0.00% … 0.00%
 Time  (median):     265.012 μs               ┊ GC (median):    0.00%
 Time  (mean ± σ):   265.171 μs ±   1.905 μs  ┊ GC (mean ± σ):  0.00% ± 0.00%

                     ▁▂▄▄▆█▆▆▇▇▇█▇▇▃▄▃▁▁                         
  ▂▁▁▁▁▂▂▁▂▃▂▃▃▃▄▄▅▅▇████████████████████▆▆▅▄▅▄▄▄▃▃▂▂▂▂▂▂▂▂▂▂▂▂ ▅
  259 μs           Histogram: frequency by time          272 μs <

 Memory estimate: 272 bytes, allocs estimate: 17.

PR:

julia> @benchmark matmul!($Cf64, $Af64, $Bf64)
BenchmarkTools.Trial: 10000 samples with 1 evaluation.
 Range (min … max):  23.756 μs …  43.892 μs  ┊ GC (min … max): 0.00% … 0.00%
 Time  (median):     24.462 μs               ┊ GC (median):    0.00%
 Time  (mean ± σ):   24.547 μs ± 586.526 ns  ┊ GC (mean ± σ):  0.00% ± 0.00%

           ▁▃▅▆██▆▄▃▁                                           
  ▂▂▂▂▂▃▃▅▆██████████▇▅▄▃▃▃▂▂▂▂▂▂▂▂▂▂▂▂▁▁▂▁▁▁▁▁▁▂▂▂▂▂▂▂▂▂▂▂▂▂▂ ▃
  23.8 μs         Histogram: frequency by time         26.7 μs <

 Memory estimate: 0 bytes, allocs estimate: 0.

@ranocha have you seen this?

This was the benchmark script I used:

julia> A = rand(-1_000:1_000, 200, 200);

julia> B = rand(-1_000:1_000, 200, 200);

julia> C = similar(A);

julia> @benchmark mul!($C, $A, $B)
BenchmarkTools.Trial: 411 samples with 1 evaluation.
 Range (min … max):  2.283 ms …  12.646 ms  ┊ GC (min … max): 0.00% … 77.02%
 Time  (median):     2.372 ms               ┊ GC (median):    0.00%
 Time  (mean ± σ):   2.430 ms ± 515.259 μs  ┊ GC (mean ± σ):  0.98% ±  3.80%

         ██        ▁                                           
  ▅▂▁▁▁▂▃██▄▄▅▅▅▄▄▇█▅▃▄▆▄▃▂▃▁▂▁▂▁▁▂▁▂▃▁▁▁▃▁▂▂▂▃▃▃▃▂▁▂▂▁▄▅▃▁▃▂ ▃
  2.28 ms         Histogram: frequency by time        2.67 ms <

 Memory estimate: 30.77 KiB, allocs estimate: 4.

julia> @benchmark matmul!($C, $A, $B)
BenchmarkTools.Trial: 10000 samples with 1 evaluation.
 Range (min … max):  90.987 μs … 118.231 μs  ┊ GC (min … max): 0.00% … 0.00%
 Time  (median):     91.633 μs               ┊ GC (median):    0.00%
 Time  (mean ± σ):   92.750 μs ±   2.644 μs  ┊ GC (mean ± σ):  0.00% ± 0.00%

  ▃▇█▅    ▂▄▃▁ ▃▆▅▁     ▁▁                                     ▂
  ████▇▃▄▆████▇████▁▃▁▆███▇▆▅▅▁▁▃▁▁▁▁▃▁▁▁▁▃▁▁▁▁▁▁▁▅▆▆▅▄▆▅▆▆▇▆▄ █
  91 μs         Histogram: log(frequency) by time       106 μs <

 Memory estimate: 0 bytes, allocs estimate: 0.

julia> 2.369e-3 / 91.302e-6
25.946857681102276

julia> Af64 = Float64.(A); Bf64 = Float64.(B); Cf64 = similar(Af64);

julia> @benchmark mul!($Cf64, $Af64, $Bf64)
BenchmarkTools.Trial: 10000 samples with 1 evaluation.
 Range (min … max):  14.808 μs … 34.542 μs  ┊ GC (min … max): 0.00% … 0.00%
 Time  (median):     15.224 μs              ┊ GC (median):    0.00%
 Time  (mean ± σ):   15.367 μs ±  1.002 μs  ┊ GC (mean ± σ):  0.00% ± 0.00%

     ▄█▆▂                                                      
  ▂▃▆████▆▄▃▂▂▂▁▁▂▂▂▂▂▁▂▁▁▁▂▂▂▂▂▂▂▂▂▁▂▂▂▂▂▂▂▂▂▂▂▂▂▂▁▂▂▂▂▁▂▂▂▂ ▃
  14.8 μs         Histogram: frequency by time        19.7 μs <

 Memory estimate: 0 bytes, allocs estimate: 0.

julia> @benchmark matmul!($Cf64, $Af64, $Bf64)
BenchmarkTools.Trial: 10000 samples with 1 evaluation.
 Range (min … max):  23.756 μs …  43.892 μs  ┊ GC (min … max): 0.00% … 0.00%
 Time  (median):     24.462 μs               ┊ GC (median):    0.00%
 Time  (mean ± σ):   24.547 μs ± 586.526 ns  ┊ GC (mean ± σ):  0.00% ± 0.00%

           ▁▃▅▆██▆▄▃▁                                           
  ▂▂▂▂▂▃▃▅▆██████████▇▅▄▃▃▃▂▂▂▂▂▂▂▂▂▂▂▂▁▁▂▁▁▁▁▁▁▂▂▂▂▂▂▂▂▂▂▂▂▂▂ ▃
  23.8 μs         Histogram: frequency by time         26.7 μs <

 Memory estimate: 0 bytes, allocs estimate: 0.

julia> Cf64 == C
true

[codecov]

Codecov Report

Patch coverage has no change and project coverage change: -0.47 ⚠️

Comparison is base (c63b5b9) 90.40% compared to head (9e0551e) 89.93%.

Additional details and impacted files

@@            Coverage Diff             @@
##           master     #181      +/-   ##
==========================================
- Coverage   90.40%   89.93%   -0.47%     
==========================================
  Files          14       14              
  Lines        1146     1133      -13     
==========================================
- Hits         1036     1019      -17     
- Misses        110      114       +4     

Impacted Files	Coverage Δ
src/Octavian.jl	50.00% <ø> (-50.00%)	⬇️

... and 2 files with indirect coverage changes

☔ View full report in Codecov by Sentry.
📢 Do you have feedback about the report comment? Let us know in this issue.

[ranocha]

@ranocha have you seen this?

No - but I have to admit that I haven't been running a lot of matmul!-based code recently, focused on other parts of our codebase instead.

Thanks a lot for fixing this! Do you have an idea which Julia bugs are involved here?

[chriselrod]

Thanks a lot for fixing this! Do you have an idea which Julia bugs are involved here?

No, but bugs such as this have been known for a long time.
TriangularSolve.jl had to disable precompilation as well for this reason.