[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Xen-devel] devel/next-2.6.39 SLUB: Unable to allocate memory on node -1 (gfp=0x20)



Tuesday, May 24, 2011, 3:48:47 PM, you wrote:

> On Mon, May 23, 2011 at 09:12:08AM +0200, Sander Eikelenboom wrote:
>> Hi Konrad,
>> 
>> It seems i have another problem with the devel/next-2.6.39 as dom0, don't 
>> know if it's xen related or not.
>> It doesn't crash the machine, also netwerk still seems to work, but these 
>> warns keep coming say one till 3 a minute.
>> I haven't changed the rest of the parameters/setup from the stable 
>> configuration i was running before with a 2.6.32 from jeremy.

> This is quite curious. There is this nasty issue that the folks at Virtual 
> Computer found with
> the rtl8169 where it would scribble over the DMA pages. It would only happen 
> with large amounts of data.

> But this looks to be memory related.

Hmm i think it's the same nasty issue then, at that moment i was indeed 
downloading a large file at high speed.
But i never saw it with the 2.6.32 series, so it must be introduced later.

>> The panic i reported earlier hasn't hit me again yet (after attaching serial 
>> console).

> I think that was due to kernel mismatch. The 'per_cpu' failure you saw is 
> similar to not having
> a specific patch. And that specific patch got removed from the #master branch 
> as a similar
> patch for it is being submitted upstream by tglrx.

> Specific patch being:     xen: do not implement irq_mask and irq_unmask in 
> xen_percpu_chip
> (48d036199309516f4e08f3cf256bab76c078864c)

Ah ok, i will lookup that patch and apply to my devel/next-2.6.39 clone !
But i haven't been able to trigger this one again so far ..

Thx

--

Sander



>> --
>> Sander
>> 
>> May 23 09:01:58 serveerstertje kernel: [128112.912479] swapper: page 
>> allocation failure. order:0, mode:0x4020
>> May 23 09:01:58 serveerstertje kernel: [128112.916453] Pid: 0, comm: swapper 
>> Not tainted 2.6.39-xen-next-2.6.39-konrad+ #5
>> May 23 09:01:58 serveerstertje kernel: [128112.916453] Call Trace:
>> May 23 09:01:58 serveerstertje kernel: [128112.916453]  <IRQ>  
>> [<ffffffff8112213c>] __alloc_pages_nodemask+0x64c/0x7e0
>> May 23 09:01:58 serveerstertje kernel: [128112.916453]  [<ffffffff8132bdf6>] 
>> ? security_sock_rcv_skb+0x16/0x20
>> May 23 09:01:58 serveerstertje kernel: [128112.916453]  [<ffffffff819899b8>] 
>> ? tcp_v4_rcv+0x828/0x930
>> May 23 09:01:58 serveerstertje kernel: [128112.916453]  [<ffffffff811588c3>] 
>> new_slab+0x233/0x250
>> May 23 09:01:58 serveerstertje kernel: [128112.916453]  [<ffffffff81158f34>] 
>> T.1026+0x194/0x440
>> May 23 09:01:58 serveerstertje kernel: [128112.916453]  [<ffffffff8190a014>] 
>> ? __netdev_alloc_skb+0x24/0x50
>> May 23 09:01:58 serveerstertje kernel: [128112.916453]  [<ffffffff8115afcc>] 
>> __kmalloc_track_caller+0x15c/0x1e0
>> May 23 09:01:58 serveerstertje kernel: [128112.916453]  [<ffffffff8190a014>] 
>> ? __netdev_alloc_skb+0x24/0x50
>> May 23 09:01:58 serveerstertje kernel: [128112.916453]  [<ffffffff81908d67>] 
>> __alloc_skb+0x77/0x160
>> May 23 09:01:58 serveerstertje kernel: [128112.916453]  [<ffffffff8190a014>] 
>> __netdev_alloc_skb+0x24/0x50
>> May 23 09:01:58 serveerstertje kernel: [128112.916453]  [<ffffffff815ea2fb>] 
>> rtl8169_rx_interrupt+0x20b/0x420
>> May 23 09:01:58 serveerstertje kernel: [128112.916453]  [<ffffffff815eb51e>] 
>> ? rtl8169_interrupt+0xfe/0x3e0
>> May 23 09:01:58 serveerstertje kernel: [128112.916453]  [<ffffffff815ed06c>] 
>> rtl8169_poll+0x3c/0x1c0
>> May 23 09:01:58 serveerstertje kernel: [128112.916453]  [<ffffffff8104c21d>] 
>> ? xen_force_evtchn_callback+0xd/0x10
>> May 23 09:01:58 serveerstertje kernel: [128112.916453]  [<ffffffff8191313e>] 
>> net_rx_action+0x14e/0x2d0
>> May 23 09:01:58 serveerstertje kernel: [128112.916453]  [<ffffffff8109ec61>] 
>> __do_softirq+0xd1/0x230
>> May 23 09:01:58 serveerstertje kernel: [128112.916453]  [<ffffffff810e7b2d>] 
>> ? handle_edge_irq+0x6d/0x120

> Hmm, edge. So this card is doing MSI, right?




-- 
Best regards,
 Sander                            mailto:linux@xxxxxxxxxxxxxx


_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-devel


 


Rackspace

Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.