Shrinkwrap: Taming Dynamic Shared Objects(fzakaria.com)

12 pointsby setheron4 tahun yang lalu4 comments

Shrinkwrap: Taming Dynamic Shared Objects

https://fzakaria.com/2022/03/14/shrinkwrap-taming-dynamic-shared-objects.html

4 comments

vup4 tahun yang lalu

As mentioned in the blog post, guix already does something similar, by patching glibc to support a per-application loader cache [1]. shrinkwrap seems to increase the binary size a lot (6.6 MB without shrinkwrap, 13 MB after running shrinkwrap for emacs).

Also when comparing the performance, the guix approach seems to be a bit faster:

  hyperfine -m 100 -w 5 -n shrinkwrap './emacs --version'  -n guix 'emacs --version'
  Benchmark #1: shrinkwrap
    Time (mean ± σ):      68.8 ms ±  13.7 ms    [User: 45.8 ms, System: 22.9 ms]
    Range (min … max):    52.8 ms … 106.3 ms    100 runs
  
  Benchmark #2: guix
    Time (mean ± σ):      56.7 ms ±  12.3 ms    [User: 34.9 ms, System: 21.7 ms]
    Range (min … max):    42.7 ms …  84.6 ms    100 runs
  
  Summary
    'guix' ran
      1.21 ± 0.36 times faster than 'shrinkwrap'

[1] https://guix.gnu.org/en/blog/2021/taming-the-stat-storm-with...

setheron4 tahun yang lalu

That's interesting. I'll have to think on why (or profile) the ld.so.conf.cache way is faster.

Thanks for the profile.

Animats4 tahun yang lalu

Didn't we do this last week?

setheron4 tahun yang lalu

I got an email asking for second change pool so I clicked it. Sorry. Didn't know someone else submitted.