Hello all,

we recently tried to stress several netblk configurations and got this kernel
log (amongst others). If additional information is required feel free to ask. We can try patches as well the config is not in production yet.

Jan 12 15:46:54 fs2 kernel: swapper: page allocation failure. order:0, mode:0x20
Jan 12 15:46:54 fs2 kernel:  [<c014c711>] __alloc_pages+0x283/0x294
Jan 12 15:46:54 fs2 kernel:  [<c016079e>] cache_alloc_refill+0x286/0x45f
Jan 12 15:46:54 fs2 kernel:  [<c01604dd>] kmem_cache_alloc+0x37/0x72
Jan 12 15:46:54 fs2 kernel:  [<c0279c36>] __alloc_skb+0x29/0x104
Jan 12 15:46:54 fs2 kernel:  [<c027a6bd>] __netdev_alloc_skb+0x1c/0x35
Jan 12 15:46:54 fs2 kernel:  [<f897b8e3>] tg3_alloc_rx_skb+0xb2/0x141 [tg3]
Jan 12 15:46:54 fs2 kernel:  [<c027dac9>] netif_receive_skb+0x257/0x271
Jan 12 15:46:54 fs2 kernel:  [<f8981d2a>] tg3_poll+0x370/0x7fd [tg3]
Jan 12 15:46:54 fs2 kernel:  [<f8981c17>] tg3_poll+0x25d/0x7fd [tg3]
Jan 12 15:46:54 fs2 kernel:  [<f8981c17>] tg3_poll+0x25d/0x7fd [tg3]
Jan 12 15:46:54 fs2 kernel:  [<c027f9cc>] net_rx_action+0x64/0xf4
Jan 12 15:46:54 fs2 kernel:  [<c0125d99>] __do_softirq+0x6c/0xcf
Jan 12 15:46:54 fs2 kernel:  [<c0125e2e>] do_softirq+0x32/0x36
Jan 12 15:46:54 fs2 kernel:  [<c0106355>] do_IRQ+0x5c/0x73
Jan 12 15:46:54 fs2 kernel:  [<c01149ab>] smp_apic_timer_interrupt+0x74/0x80
Jan 12 15:46:54 fs2 kernel:  [<c0102c56>] default_idle+0x0/0x3e
Jan 12 15:46:54 fs2 kernel:  [<c010479f>] common_interrupt+0x23/0x28
Jan 12 15:46:54 fs2 kernel:  [<c0102c56>] default_idle+0x0/0x3e 
Jan 12 15:46:54 fs2 kernel:  [<c0102c82>] default_idle+0x2c/0x3e
Jan 12 15:46:54 fs2 kernel:  [<c0102389>] cpu_idle+0x91/0xaa
Jan 12 15:46:54 fs2 kernel:  [<c038d8c0>] start_kernel+0x2b8/0x2c0
Jan 12 15:46:54 fs2 kernel:  [<c038d0e0>] unknown_bootoption+0x0/0x196
Jan 12 15:46:54 fs2 kernel:  =======================
Jan 12 15:46:54 fs2 kernel: DMA per-cpu:
Jan 12 15:46:54 fs2 kernel: CPU    0: Hot: hi:    0, btch:   1 usd:   0   Cold: hi:    0, btch:   1 usd:   0
Jan ...
[ message continues ]

On Sun, 13 Jan 2008 15:04:07 +0100

Does anybody have an idea what my original posting is all about?

Regards,
Stephan


--

[ message continues ]

I might have some idea.  From your lspci, it looks like a board I had some 
issues with about one year ago.  It was a Tyan server board that I had put a 
3ware card into a pci-X 133MHz slot.  Well, for some reason, the Broadcom 
chip would have trouble, bad data, kernel problems like yours.  I called 
Tyan and they had me short a jumper that forced the slot to 100MHz (the bus 
is shared between that slot and the Broadcom chip) and that fixed it for me.

I hope that might be some help to you.


Good luck,
Charles
--

[ message continues ]