lists.openwall.net   lists  /  announce  owl-users  owl-dev  john-users  john-dev  passwdqc-users  yescrypt  popa3d-users  /  oss-security  kernel-hardening  musl  sabotage  tlsify  passwords  /  crypt-dev  xvendor  /  Bugtraq  Full-Disclosure  linux-kernel  linux-netdev  linux-ext4  linux-hardening  PHC 
Open Source and information security mailing list archives
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [day] [month] [year] [list]
Date:   Thu, 13 Aug 2020 12:57:52 +0200 (CEST)
From:   Julia Lawall <julia.lawall@...ia.fr>
To:     Sumera Priyadarsini <sylphrenadin@...il.com>
cc:     michal.lkml@...kovi.net, Gilles.Muller@...6.fr,
        gregkh@...uxfoundation.org, nicolas.palix@...g.fr,
        linux-kernel@...r.kernel.org, cocci@...teme.lip6.fr
Subject: Re: [Cocci] [PATCH] scripts: coccicheck: Change default value for
 parallelism



On Thu, 13 Aug 2020, Sumera Priyadarsini wrote:

>
>
> On Thu, 13 Aug, 2020, 1:45 AM Julia Lawall, <julia.lawall@...ia.fr> wrote:
>
>
>       On Wed, 12 Aug 2020, Sumera Priyadarsini wrote:
>
>       > By default, coccicheck utilizes all available threads to implement
>       > parallelisation. However, when hyperthreading is enabled, this leads
>       > to all threads per core being occupied resulting in longer wall-clock
>       > times and higher power consumption.
>
>       I have the feeling that the above sentence is not quite optimal.
>       Actually, using all of the available hardware threads would not be a bad
>       thing, if it was giving a benefit.  The point is that it doesn't.  It
>       makes the performance worse instead.
>
>
> How does this sound? 
>  
> However, when all available threads are used, a decrease in performance was noted. The elapsed time was minimum when at most one thread per core was used. 

This looks better, thanks.

julia

>
>
>
>
>       > Hence, to improve performance,
>       > modify coccicheck to use only one thread per core atmost.
>
>       "atmost" is not a word.  It would be clearer to say "to use at most one
>       thread per core".
>
>       > In the cases where the total number of threads is more than 8 and
>       > hyperthreading is enabled, it was observed that optimum performance
>       > is achieved around one-fourth of the total number of cores.
>       > Modify the script further to accommodate this use case.
>
>       It would be nice to give some performance numbers and some information
>       about the machine used.
>
>
> Alright, will add this. 
>
>
>       thanks,
>       julia
>
>       >
>       > Signed-off-by: Sumera Priyadarsini <sylphrenadin@...il.com>
>       > ---
>       >  scripts/coccicheck | 9 +++++++++
>       >  1 file changed, 9 insertions(+)
>       >
>       > diff --git a/scripts/coccicheck b/scripts/coccicheck
>       > index e04d328210ac..dd228dcc915e 100755
>       > --- a/scripts/coccicheck
>       > +++ b/scripts/coccicheck
>       > @@ -75,8 +75,17 @@ else
>       >          OPTIONS="--dir $KBUILD_EXTMOD $COCCIINCLUDE"
>       >      fi
>       >
>       > +    # Use only one thread per core by default if hyperthreading is enabled
>       > +    THREADS_PER_CORE=$(lscpu | grep "Thread(s) per core: " | tr -cd [:digit:])
>       >      if [ -z "$J" ]; then
>       >          NPROC=$(getconf _NPROCESSORS_ONLN)
>       > +     if [ $THREADS_PER_CORE -gt 1 -a $NPROC -gt 2 ] ; then
>       > +             if [ $NPROC -gt 8 ] ; then
>       > +                     NPROC=$((NPROC/4))
>       > +             else
>       > +                     NPROC=$((NPROC/2))
>       > +             fi
>       > +     fi
>       >      else
>       >          NPROC="$J"
>       >      fi
>       > --
>       > 2.17.1
>       >
>       > _______________________________________________
>       > Cocci mailing list
>       > Cocci@...teme.lip6.fr
>       > https://systeme.lip6.fr/mailman/listinfo/cocci
>       >
>
>
>

Powered by blists - more mailing lists