Bug 43244 - firewire_ohci prevents boot
firewire_ohci prevents boot
Status: NEW
Product: Drivers
Classification: Unclassified
Component: IEEE1394
All Linux
: P1 high
Assigned To: drivers_ieee1394
:
Depends on:
Blocks:
  Show dependency treegraph
 
Reported: 2012-05-13 19:28 UTC by Rob Allen
Modified: 2012-06-19 21:07 UTC (History)
1 user (show)

See Also:
Kernel Version: 3.1.0-rc3
Tree: Mainline
Regression: Yes


Attachments
lspci -vvv (37.90 KB, text/plain)
2012-05-13 19:28 UTC, Rob Allen
Details
dmesg (filtered) from 3.0 series kernel (419 bytes, text/plain)
2012-05-13 19:30 UTC, Rob Allen
Details
dmesg from v3.4 with CONFIG_PCI_DEBUG (75.36 KB, text/plain)
2012-05-24 17:30 UTC, Rob Allen
Details
dmesg from 3.0 series kernel (firewire_ohci works) (71.16 KB, application/octet-stream)
2012-06-19 21:05 UTC, Rob Allen
Details
lspci -vvxxx -s 0a:09.4 (3.0 series, working) (1.60 KB, application/octet-stream)
2012-06-19 21:07 UTC, Rob Allen
Details
lspci -vvxxx -s 0a:09.4 (3.2 series, broken) (1.59 KB, application/octet-stream)
2012-06-19 21:07 UTC, Rob Allen
Details

Description Rob Allen 2012-05-13 19:28:20 UTC
Created attachment 73282 [details]
lspci -vvv

After upgrading to a 3.2 series kernel (from 3.0 series), my laptop failed to boot. Retrying with nomodeset, the final message is:

firewire_ohci: Failed to set Link Power Status

Afterwards, the machine is completely dead (ignoring Magic SysRq), requiring holding the power button to get anywhere.

Blacklisting firewire_ohci allows the kernel to boot okay.

Having experimented with Ubuntu's mainline kernel builds, it seem to have been introduced between 3.1.0-rc2 (working) and 3.1.0-rc3.
Comment 1 Rob Allen 2012-05-13 19:30:19 UTC
Created attachment 73283 [details]
dmesg (filtered) from 3.0 series kernel

This is the dmesg output that firewire_ohci used to produce (under 3.0 series).
Comment 2 Stefan Richter 2012-05-13 21:15:10 UTC
Discussion:  http://marc.info/?l=linux1394-devel&m=133694022804796
Comment 3 Rob Allen 2012-05-24 17:30:12 UTC
Created attachment 73381 [details]
dmesg from v3.4 with CONFIG_PCI_DEBUG
Comment 4 Rob Allen 2012-06-04 10:54:36 UTC
Bisected. The change that introduced the failure is:

commit 0a2daa1cf35004f5adbf4138555cc5669abf3a3e
Author: Ram Pai <linuxram@us.ibm.com>
Date:   Mon Jul 25 13:08:41 2011 -0700

    PCI: make cardbus-bridge resources optional
    
    Allocate resources to cardbus bridge only after all other genuine
    resources requests are satisfied. Dont retry if resource allocation
    for cardbus-bridges fail.
    
    Signed-off-by: Ram Pai <linuxram@us.ibm.com>
    Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
Comment 5 Rob Allen 2012-06-19 21:05:58 UTC
Created attachment 73861 [details]
dmesg from 3.0 series kernel (firewire_ohci works)
Comment 6 Rob Allen 2012-06-19 21:07:32 UTC
Created attachment 73871 [details]
lspci -vvxxx -s 0a:09.4 (3.0 series, working)
Comment 7 Rob Allen 2012-06-19 21:07:58 UTC
Created attachment 73881 [details]
lspci -vvxxx -s 0a:09.4 (3.2 series, broken)

Note You need to log in before you can comment on or make changes to this bug.