Re: lintian: Performance log graphed

[Niels Thykier <niels@thykier.net>]

On 2015-07-03 15:40, Bastien ROUCARIES wrote:
> [...]
> 

Hi Bastien,

> 
> cat lintian-perf-20150702.log | sort -g -t, -k 3 | grep -v unpack,
> |grep -v ,total
> 

NB: you probably wanted "unpacked," - AFAIR, there is nothing that ends
on "unpack".

> give another view.
> 
> 
> [...]
> binary:manpages-dev/3.74-1/all,check/manpages,90.116526
> binary:linux-image-4.0.0-2-686-pae-dbg/4.0.5-1/i386,coll/file-info,96.109758
> binary:linux-image-4.0.0-2-686-pae-dbg/4.0.5-1/i386,coll/md5sums,96.468523
> source:linux/4.0.5-1,check/cruft,347.694435
> binary:linux-manual-4.0/4.0.5-1/all,check/manpages,733.472202
>

To be honest, I am not sure how to read your message.  Did you mean to
confirm or dispute the original graph?  AFAICT, the data you presented
seems to confirm it.


In other news:

 * I got about 2/3 to 3/4 off the runtime from objdump-info-helper by
   commenting out the UPX check (that doesn't work anyway - see #790932)

 * I got 1/2 off the runtime from strings-helper by replacing
   L::Command's spawn + reap with a manual pipe+fork+exec+wait.  Pushed
   to master.

~Niels