Re: ACPI warning from alloc_pages_nodemask on boot (2.6.33 regression)

[Zhang Rui <rui.zhang@intel.com>]

On Thu, 2009-12-31 at 02:05 +0800, Stephen Hemminger wrote:
> On Wed, 30 Dec 2009 16:35:44 +0100
> Peter Zijlstra <peterz@infradead.org> wrote:
> 
> > On Wed, 2009-12-30 at 15:21 +0900, KOSAKI Motohiro wrote:
> > > >> [    1.630020] ------------[ cut here ]------------
> > > >> [    1.630026] WARNING: at mm/page_alloc.c:1812 __alloc_pages_nodemask+0x617/0x730()
> > > >
> > > >        if (order >= MAX_ORDER) {
> > > >                WARN_ON_ONCE(!(gfp_mask & __GFP_NOWARN));
> > > >                return NULL;
> > > >        }
> > > >
> > > > I don't know what the mm alloc code is complaining about here.
> > 
> > > >> [    1.630028] Hardware name: System Product Name
> > > >> [    1.630029] Modules linked in:
> > > >> [    1.630032] Pid: 1, comm: swapper Not tainted 2.6.33-rc2 #4
> > > >> [    1.630034] Call Trace:
> > 
> > > >> [    1.630064]  [<ffffffff812cae3e>] acpi_os_allocate+0x25/0x27 
> > 
> > Right, so ACPI is trying to allocate something larger than 2^MAX_ORDER
> > pages, which on x86 computes to 4K * 2^11 = 8M.
> > 
> > That's not going to work.
> > 
> > Did this machine properly boot before? I seem to remember people working
> > on moving away from bootmem and getting th page/slab stuff up and
> > running sooner, it might be fallout from that...
> > 
> 
> Yes, and it still boots now.
> 
we have rootcaused the problem.
please refer to http://bugzilla.kernel.org/show_bug.cgi?id=14954

thanks,
rui

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>