lkml.org 
[lkml]   [1997]   [Nov]   [17]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
Subjectmemory management things to hack...
To the person who requested mem management things to hack:
(I don't have your address I think because I joined the list too late)

I have an SMP machine (2 PPro cpus) with 512M (well, I'd like to say it
has 512M). I can confirm for anyone who wants to know that 2.0.30 works
fine with 256M of ram. However, it crashes almost immediately (during
boot) when the machine has 512M of ram installed. It doesn't even work
if I use "mem=256M" ( i don't know why). Now, I have managed to get the
machine usable with 512M by using kernel 2.1.62.

Questions: why does 2.1.62 work? (Machine has been runing 4 days now -
but not without some odd quirks, not sure if its because its 2.1.X or if
not all is well with 512M still, or both)

can whatever be incorporated into next 2.0.X kernel?

some odd quirks observed:
-our model ("CORSIM" - http://misic.soc.cornell.edu) crashes at a
weird spot now with "error closing file." This may be related to new C
libraries however.
-users complain of FTPed file corruptions. (they might at fault)
-System crashed while FTPing a large file.
-mouse skips all around the screen at random times (and clicks on
things at random) as if someone was writing into /dev/mouse.
-telnet from or to any machine which is not also linux is bizarre
(suspect termcap library need upgrading or something, w/ respect to new
C library maybe)

Provided one of our staff figures out quirk#1, the rest seems minor. I
am curious to know other's experiences with 2.1.62 as a "production"
kernel rather than test kernel. Please email me
(koni@misic.soc.cornell.edu) any other oddities that have been observed.

Our "big" machine is usually under heavy use but occasionally over a
weekend there is some idle time that I could use to test SMP kernels and
memory management of 256M or 512M. Let me know if this is convienent for
anyone.

Unfortunately, due to PPro caching and our motherboard's interleaving (I
think), memory sizes >64M that aren't powers of 2 run orders of
magnitude slower so testing/using memory sizes of 192M or 384M is mostly
a waste of time for us. (30 second compiles take upwards of 35 minutes)

More info about our "memory problems" can be found by searching
comp.os.linux.hardware and/or comp.os.linux.setup for postings from me
in the last two months or so.

Thanks,

-Koni

From: Harald Koenig <koenig@tat.physik.uni-tuebingen.de>
Date: Mon, 17 Nov 1997 08:48:30 +0100
Subject: Re: Ideas for memory management hackers (2)

On Nov 16, Alan Cox wrote:

> > > - VM areas are limited to 128MB. This leads to many swap-files. (I
am
> > > not quite shure about further limitation. Does it mean that
bigger
> > > mmaps are not possible? The swapfile limitation is definitely in
the
> > > kernel.)
> >
> > no more than 128MB shared memory AFAIK?! this is bad for some (big)
databases...
>
> The kernel is happy with > 128Mbytes, the only limit is on physical
memory
> (1Gig) and virtual (3Gig). The limit he refers to is swap
file/partition size
> per partition and is becoming an irritiation in itself for osme


On Mon, 6 Oct 1997, greg@netset.com wrote in linux-kernel :

> > In the kernel include file include/asm/shmparam.h SHMMAX is set to
> > 16MB (0x1000000) and I was wondering if it would be a bad thing to
> > set it to 64MB (0x4000000) or 128MB (0x8000000) provided the machine

> > in question had 64MB+ or 128MB+ memory.
>
> In theory, increasing SHMMAX to 128MB should be possible... Beyond
that,
> not without difficulty - the mm subsystem uses 128MB swap areas, of
which
> shared memory is one.

so it's not true that SHMMAX is limited to 128MB right now ?
I don't have that much DRAM so I can't test it, sorry;) ?


Harald
- --
All SCSI disks will from now on ___ _____
be required to send an email notice 0--,| /OOOOOOO\
24 hours prior to complete hardware failure! <_/ / /OOOOOOOOOOO\
\
\/OOOOOOOOOOOOOOO\
\
OOOOOOOOOOOOOOOOO|//
Harald Koenig,
\/\/\/\/\/\/\/\/\/
Inst.f.Theoret.Astrophysik // / \\ \
koenig@tat.physik.uni-tuebingen.de ^^^^^ ^^^^^





\
 
 \ /
  Last update: 2005-03-22 13:40    [W:0.091 / U:0.280 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site