2.6.10-rc3 b44 resume problem

2.6.10-rc3 b44 resume problem

[Robin Rosenberg]

Hi,

I can't get my wired network running after resume from suspend to ram:

Dec  7 18:48:36 xine kernel: ifconfig: page allocation failure. order:8, 
mode:0x21
After stopping the network, and netplugd and rmmodding b44 and mii I still get
this when trying to start my network (when b44 is activated).

Dec  7 18:51:18 xine kernel:  [dump_stack+30/32] dump_stack+0x1e/0x20
Dec  7 18:51:18 xine kernel:  [__alloc_pages+569/887] 
__alloc_pages+0x239/0x377
Dec  7 18:51:18 xine kernel:  [__get_free_pages+27/54] 
__get_free_pages+0x1b/0x36
Dec  7 18:51:18 xine kernel:  [dma_alloc_coherent+188/247] 
dma_alloc_coherent+0xbc/0xf7
Dec  7 18:51:18 xine kernel:  [pg0+546065382/1069609984] 
b44_alloc_consistent+0x99/0x152 [b44]
Dec  7 18:51:18 xine kernel:  [pg0+546066627/1069609984] b44_open+0x20/0xef 
[b44]
Dec  7 18:51:18 xine kernel:  [dev_open+116/131] dev_open+0x74/0x83
Dec  7 18:51:18 xine kernel:  [dev_change_flags+86/295] 
dev_change_flags+0x56/0x127
Dec  7 18:51:18 xine kernel:  [devinet_ioctl+1317/1501] 
devinet_ioctl+0x525/0x5dd
Dec  7 18:51:18 xine kernel:  [inet_ioctl+187/197] inet_ioctl+0xbb/0xc5
Dec  7 18:51:18 xine kernel:  [sock_ioctl+399/572] sock_ioctl+0x18f/0x23c
Dec  7 18:51:18 xine kernel:  [sys_ioctl+353/596] sys_ioctl+0x161/0x254
Dec  7 18:51:18 xine kernel:  [sysenter_past_esp+82/113] 
sysenter_past_esp+0x52/0x71

-- robin

Re: 2.6.10-rc3 b44 resume problem

[Pekka Pietikainen]

On Tue, Dec 07, 2004 at 10:25:31AM +0100, Robin Rosenberg wrote:
> Hi,
> 
> I can't get my wired network running after resume from suspend to ram:
> 
> Dec  7 18:48:36 xine kernel: ifconfig: page allocation failure. order:8, 
> mode:0x21
> After stopping the network, and netplugd and rmmodding b44 and mii I still get
> this when trying to start my network (when b44 is activate
Hi

For a quick "fix", you can set B44_DMA_MASK to 0xffffffff. But this 
is not really a fix (and will make your machine crash and burn if you have
more than a gig of memory and use something other than a 3:1 memory layout,
but it is essentially the pre-2.6.10 situation so if the driver worked
for you before it will do so again)

https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=118165, 

"See last paragraph of comment #66. The problem is that the driver
needs about 750k of memory that has to be located under 1GB physically
to not trigger the hardware bug that causes crashes and other fun. The
driver tries to allocate that kind of memory
(pci_set_consistent_dma_mask(pdev, 0x3fffffff) ). There should be
plenty, right?

Unfortunately the way it's implemented right now in the generic x86
pci code is that if you ask for some memory with a dma mask of < 4GB,
it falls back to giving you memory from the first 16MB. Now that's a
pretty limited resource :-(. There seems to be 3 drivers that need
similar workarounds (wanxl, aacraid and b44)."