http://bugzilla.suse.com/show_bug.cgi?id=970249
http://bugzilla.suse.com/show_bug.cgi?id=970249#c7
Michal Kubeček changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |hare@suse.com
--- Comment #7 from Michal Kubeček ---
The screenshot from comment 3 provides most information. I checked the
addresses and the crash occurs in aac_src_intr_message() on line 73:
bellbits = src_readl(dev, MUnit.ODR_R);
At the moment, RAX should contains the value of dev->regs.src.bar0 and
it is null so that an attempt to access members of a structure pointed to
by it results in a page fault.
As the driver is still marked as loading, my hypothesis is that an IRQ
was received and processed before all data structures of the device were
fully initialized (perhaps because the interrupts were enabled too early).
I talked to Hannes about this and he thinks he has recently seen a bug
like this even if he is not completely sure it was about aacraid.
--
You are receiving this mail because:
You are on the CC list for the bug.