Re: With Daniel Phillips Patch (was: aic7xxx with 2.4.9 on 7899P)

[Stephan von Krawczynski <skraw@ithnet.com>]

On Tue, 21 Aug 2001 19:26:58 +0200
Daniel Phillips <phillips@bonn-fries.net> wrote:

> On August 21, 2001 06:48 pm, Sven Heinicke wrote:
> > Yes, highmem was on, the stystem got 4G of memory.  I turned off
> > highmem and got no messages apart from one:
> > 
> > Aug 21 07:29:19 ps1 kernel: (scsi0:A:0:0): Locking max tag count at 64
> > 
> > which I was getting before.
> >
> > Disk access is faster then before but still slower then the IDE
> > drive.  Any ideas?
> 
> Two separate problems, I think.  I don't know anything about the aic7xxx 
> driver but I can take a look at the highmem problem.  First, can you try
> it with highmem enabled, on a recent -ac kernel, say 2.4.8-ac7.

Ok, Daniel, here are the results of the german jury :-) (EU insider joke)

Aug 21 19:46:40 admin kernel: __alloc_pages: 2-order allocation failed (gfp=0x20/0).
Aug 21 19:46:40 admin kernel: __alloc_pages: 3-order allocation failed (gfp=0x20/0).
Aug 21 19:46:40 admin kernel: __alloc_pages: 3-order allocation failed (gfp=0x20/0).
Aug 21 19:46:40 admin kernel: __alloc_pages: 2-order allocation failed (gfp=0x20/0).
Aug 21 19:46:40 admin kernel: __alloc_pages: 3-order allocation failed (gfp=0x20/0).
Aug 21 19:46:40 admin kernel: __alloc_pages: 3-order allocation failed (gfp=0x20/0).
Aug 21 19:46:40 admin kernel: __alloc_pages: 2-order allocation failed (gfp=0x20/0).
Aug 21 19:46:40 admin kernel: __alloc_pages: 3-order allocation failed (gfp=0x20/0).
Aug 21 19:46:40 admin kernel: __alloc_pages: 3-order allocation failed (gfp=0x20/0).
Aug 21 19:46:40 admin kernel: __alloc_pages: 2-order allocation failed (gfp=0x20/0).

And what may be of big interest for Justin: I am using the _old_ AIC7xxx driver. 

The problem can quite easily be produced on my side. All you need is a (problem) host with NFS-server running and a client system. Then simply copy a lot of big files to the server. If you now go and read CD on the server you are in big trouble:
cpu load is shot through the ceiling and you cannot even type chars in a shell after about 3 minutes. Remember I am sitting in front of a dual P-III 1GHz with 1 GB of RAM and U160 SCSI, I simply cannot believe this. I have never seen such a thing under 2.2.

Regards,
Stephan

PS: I try to disable HighMem next.