Date: Fri, 11 Aug 2006 11:36:02 -0700
From: Andrew Morton <akpm@osdl.org>
To: Mark Haverkamp <markh@osdl.org>
Cc: linux-kernel@vger.kernel.org
Subject: Re: 2.6.18-rc3-mm2
Message-Id: <20060811113602.04867f46.akpm@osdl.org>
In-Reply-To: <1155319901.17493.9.camel@markh3.pdx.osdl.net>
References: <20060806030809.2cfb0b1e.akpm@osdl.org>
	<1155319901.17493.9.camel@markh3.pdx.osdl.net>
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1187
Lines: 33

On Fri, 11 Aug 2006 11:11:40 -0700
Mark Haverkamp <markh@osdl.org> wrote:

> On Sun, 2006-08-06 at 03:08 -0700, Andrew Morton wrote:
> > ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.18-rc3/2.6.18-rc3-mm2/
> 
> I am seeing problem loading modules at boot time.  My initrd tries to
> load scsi_mod and percpu_modalloc prints this;
> 
> Could not allocate 16 bytes percpu data
> 
> This is a 2 processor x86_64 machine.  I have attached the output from
> the serial console and the config file.
> 
> It is related to the mm patches.  I can boot OK from the main kernel
> tree and the scsi trees.

Yeah, sorry - this is almost certainly due to the increase in NR_IRQS.  It
made this, in include/linux/kernel_stat.h

	DECLARE_PER_CPU(struct kernel_stat, kstat);

really big and we consume all the per-cpu memory.


NR_IRQS is (sometimes) calculated from NR_CPUS via complex means.  Reducing
your NR_CPUS should fix things up.

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/