FS#17551 - [kernel26] Can't mount USB device with DMA
Attached to Project:
Arch Linux
Opened by Dan Vratil (progdan) - Saturday, 19 December 2009, 20:41 GMT
Last edited by Tobias Powalowski (tpowa) - Friday, 08 January 2010, 08:09 GMT
Opened by Dan Vratil (progdan) - Saturday, 19 December 2009, 20:41 GMT
Last edited by Tobias Powalowski (tpowa) - Friday, 08 January 2010, 08:09 GMT
|
Details
Description:
When USB device with MDA support (mp3 player for example) is plugged in, dmesg outputs: usb 5-1: new full speed USB device using uhci_hcd and address 4 usb 5-1: configuration #1 chosen from 1 choice usb_storage: Unknown symbol scsi_add_host_with_dma Function scsi_add_host_with_dma was introduced in 2.6.32.1. I don't know if it is a bug in upstream or in our package. But better to start down here. Maybe something must be changed in kernel config? Steps to reproduce: Take some device with DMA, plug it in and look into dmesg. |
This task depends upon
http://git.kernel.org/?p=linux/kernel/git/stable/linux-2.6.32.y.git;a=commit;h=d888b1a2d5c7a9fbdc01e7395ea6a8d75cd729f5 changed the scsi module a bit, so 2.6.32.1 isn't compatible with 2.6.32 modules. If you happen to have 2.6.32 loaded and try to load 2.6.32.1 scsi modules, you'll get unresolved symbols.
Can this be reassigned? It's nobody's right now.
A bit more digging revealed that it was the particular "early kms" workaround that I was using that was the problem i.e. the one with the custom radeon hook, as described here: http://bbs.archlinux.org/viewtopic.php?id=86211. Changing to tpowa's workaround (
FS#17461) for early kms also made the usb problem go away.I have no idea why the radeon kms start method would cause an Unknown symbol error in the usb-storage module, but I now have both early radeon kms AND usb storage working again, so I'm happy. :)
As I asked for this to be reopened, I guess I should now say that it can be closed (again) - unless, of course, anyone wants to track down the exact cause of the problem.