This is the mail archive of the
libc-alpha@sourceware.org
mailing list for the glibc project.
Re: [PATCH] Simple malloc benchtest.
- From: OndÅej BÃlka <neleai at seznam dot cz>
- To: Will Newton <will dot newton at linaro dot org>
- Cc: libc-alpha <libc-alpha at sourceware dot org>
- Date: Fri, 27 Dec 2013 18:36:21 +0100
- Subject: Re: [PATCH] Simple malloc benchtest.
- Authentication-results: sourceware.org; auth=none
- References: <20131221153303 dot GA8420 at domone dot podge> <CANu=DmiknJbnr0donGEEyG__o1Unpd7iEjeQ1LSEpv_vJNO1TA at mail dot gmail dot com> <20131221214545 dot GA5621 at domone> <CANu=DmiVik=zR+_XWkytaAMFu8iUDEbRBe+rF6O=Zk4Vo59SyA at mail dot gmail dot com>
On Sun, Dec 22, 2013 at 02:28:25PM +0000, Will Newton wrote:
> >
> >> Also I would recommend tetsing some larger allocations as some
> >> allocators are very fast for small allocations but perform very poorly
> >> for larger allocations. Switching to e.g. a power law distribution of
> >> allocation sizes would allow this to be done without using huge
> >> amounts of memory or runtime.
> >>
> > Preprocessing time does not matter here, it is not measured.
>
> I'm not sure I understand.
>
You talked that switching to power law distribution of allocation
sizes would allow this to be done without using huge amounts of memory or runtime.
But in general we are not that constrained by memory and runtime requirements,
we could spend much more resources in generating a distribution that
will be used.
In current case trying to be clever would not help anything, see
following pattern why:
http://www.cs.northwestern.edu/~pdinda/icsclass/doc/dsa.pdf