Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 17 Dec 2012 22:55:55 -0800
From:      Adrian Chadd <adrian@freebsd.org>
To:        Oleksandr Tymoshenko <gonzo@bluezbox.com>
Cc:        freebsd-mips@freebsd.org
Subject:   Re: Help: Reg Out of swap space
Message-ID:  <CAJ-Vmokb-LhJz10jCwUYpgzaM3nhOiW8yfPDgERjxz1eR9qG9A@mail.gmail.com>
In-Reply-To: <129337B8-C3C2-4365-8C8D-6F5B6B45DDB1@bluezbox.com>
References:  <CAK3PU_Am17vANaZxGNobSEgU3nVA-MTWw1cY9FFju8xLj%2B81nQ@mail.gmail.com> <CAJ-VmomnbEXN1YJSupw4UAXiPqdO=Bj9c_D=KGi8hzbLGAzE_w@mail.gmail.com> <CAK3PU_BDdLKYsQQCkXKrk5Mh2bFFgMbroxuVDJ9amH-uBuVC8A@mail.gmail.com> <50CFE88E.3020303@bluezbox.com> <266E3514-39BC-4E3C-A5BA-50B4D7DA635E@bsdimp.com> <129337B8-C3C2-4365-8C8D-6F5B6B45DDB1@bluezbox.com>

next in thread | previous in thread | raw e-mail | index | archive | help
i386 would suffer if it ran on say, 64 or 128mb ram.

I don't think it's a tier X question, I think it's a "hey, everything
has ${LOTS} of real RAM, let's make the allocator assume that!" design
from Jason.

Which is fine for the systems Jason is running on, but not for other situations.

Honestly, I think we should write some patches to jemalloc to
implement a better default set of behaviours (including run-time
configuration tweaks to set default parameters like arena sizes, etc)
and hand that off to Jason to push into his next install of jemalloc.

Thanks,



Adrian


On 17 December 2012 21:57, Oleksandr Tymoshenko <gonzo@bluezbox.com> wrote:
>
> On 2012-12-17, at 8:06 PM, Warner Losh <imp@bsdimp.com> wrote:
>
>>
>> On Dec 17, 2012, at 8:52 PM, Oleksandr Tymoshenko wrote:
>>
>>> On 12/17/2012 2:32 AM, Mukunda Haveri wrote:
>>>> I have done that already. Hope it's correct..
>>>> Please have a look at the attached make.mips.conf. and my build script (csh
>>>> script) sets,
>>>> setenv __MAKE_CONF /root/mips/make.conf.mips
>>>> setenv SRCCONF /root/mips/src.conf.mips
>>>>
>>>> I also have a new problem after updating the source to head [from 24k+
>>>> version]. Something related to <audisdistd>. -:)
>>>>
>>>>
>>>>
>>>> On Mon, Dec 17, 2012 at 2:14 PM, Adrian Chadd <adrian@freebsd.org> wrote:
>>>>
>>>>> You need to define MALLOC_PRODUCTION in your build or the default
>>>>> jemalloc options in -HEAD cause it to run out of RAM on embedded
>>>>> platforms.
>>>>>
>>>>>
>>>
>>> As far as I can see it's MALLOC_DISTRIBUTION in make.conf. Should be MALLOC_PRODUCTION
>>
>> It should be WITHOUT_MALLOC_DEBUG
>>
>
> BTW, why not make it a default for platforms with small footprint or non-tier1? As far as I understand
> the idea behind having debug enabled is to make getting information from running systems
> easier. I do not think this approach is applicable to ARM or MIPS targets.
>
> _______________________________________________
> freebsd-mips@freebsd.org mailing list
> http://lists.freebsd.org/mailman/listinfo/freebsd-mips
> To unsubscribe, send any mail to "freebsd-mips-unsubscribe@freebsd.org"



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CAJ-Vmokb-LhJz10jCwUYpgzaM3nhOiW8yfPDgERjxz1eR9qG9A>