perf(sort): Benchmark with a random 95MB file slow

[HackingRepo]

Hi, uutils mainteners

i confirmed that the uu sort extremly a slow with 95MB file, as we did in the benchmark

relunsec@relunsec:~/software/coreutils/target/release$ cat /dev/urandom > random_data
^C
relunsec@relunsec:~/software/coreutils/target/release$ ls -lh random_data 
-rw-rw-r-- 1 relunsec relunsec 95M Jun 29 08:37 random_data
relunsec@relunsec:~/software/coreutils/target/release$ hyperfine --shell=none --warmup 3 "./sort random_data" "gnusort random_data"
Benchmark 1: ./sort random_data
  Time (mean ± σ):      5.487 s ±  0.040 s    [User: 6.048 s, System: 0.211 s]
  Range (min … max):    5.447 s …  5.566 s    10 runs
 
Benchmark 2: gnusort random_data
  Time (mean ± σ):      1.620 s ±  0.030 s    [User: 4.360 s, System: 0.138 s]
  Range (min … max):    1.581 s …  1.676 s    10 runs
 
Summary
  gnusort random_data ran
    3.39 ± 0.07 times faster than ./sort random_data
relunsec@relunsec:~/software/coreutils/target/release$ 

[sylvestre]

as i said, could you please stop filling perf issues ? We know how to find them and we don't need plenty of issues of this
thanks

[HackingRepo]

ok, sorry

[sylvestre]

well, we should keep it
i was asking to pause the performance bug report. currently, i do need help with decreasing the backlog

[oech3]

We have label feature to filter-out uninterested issues. How about allowing to label when issues were opened?

[sylvestre]

i am sorry but we have almost 350 PR and 828 issues, we just don't need more issues that we can identify in a few seconds :)