Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 15 Feb 2006 11:11:49 +0800
From:      David Xu <davidxu@freebsd.org>
To:        Dmitry Pryanishnikov <dmitry@atlantis.dp.ua>
Cc:        freebsd-current@freebsd.org
Subject:   Re: Virtual memory consumption (both user and kernel) in modern CURRENT
Message-ID:  <43F29BF5.4060300@freebsd.org>
In-Reply-To: <20060215024339.N22450@atlantis.atlantis.dp.ua>
References:  <20060215024339.N22450@atlantis.atlantis.dp.ua>

next in thread | previous in thread | raw e-mail | index | archive | help
Dmitry Pryanishnikov wrote:
> 
> 
> Hello!
> 
>  I've got a relatively fresh (12-Feb-2006 14:20 UTC) CURRENT on my ASUS
> M5A notebook (CPU Pentium M 1.86GHz, 256 Mb memory), kernel config is:
> 
>     ftp://external.atlantis.dp.ua/FreeBSD/CURRENT/NOTEBOOK
> 
> I have several questions regarding virtual memory use, both user-mode 
> and kernel.
> 
> 1) Is it normal that virtual memory size for almost every non-kernel 
> process
>    is close to 50Mb now:
> 
>     ftp://external.atlantis.dp.ua/FreeBSD/CURRENT/top.txt
> 
>    Is it miscalculation or real growth of virtual address space?
> 
> 
I believe this is the new malloc code in libc, I am seeing this on my
Athlon64 machine, now it likes swap memory, in the old days, it seldom
touched it.

> 2) I can _trivially_ crash my box by extracting and deleting Openoffice.org
>    distribution:
> 
>     cd /usr/ports/editors/openoffice.org-2.0
>     NOCLEANDEPENDS=yes make extract clean
> 
>    However, I can't obtain crash dump in most cases, see the picture:
> 
>     ftp://external.atlantis.dp.ua/FreeBSD/CURRENT/nodump.jpg
> 
>    Is it due to new ATA DMA dump code, or due to some other reason?
>    I've never seen such dump failures before.
> 
I can not produce core on Intel 945 chipset too, south bridge is ICH7.

> 
> 3) Once I was lucky enought to obtain a valid crash dump in this situation.
>    Here is the backtrace:
> 
>     ftp://external.atlantis.dp.ua/FreeBSD/CURRENT/bt
> 
>    What tunables can I use to prevent this panic? Alas tuning(7) doesn't
>    say anything about "mem_map too small". Is it tuning issue at all
>    or a bug?
> 
> 
> Sincerely, Dmitry




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?43F29BF5.4060300>