kern/122126: [ed] [panic] panic when using WD-8013 with ed(4)

[Volker Werth]

The following reply was made to PR kern/122126; it has been noted by GNATS.

From: Volker Werth <vwe at freebsd.org>
To: Peter Much <pmc at citylink.dinoex.sub.org>
Cc: bug-followup at FreeBSD.org
Subject: Re: kern/122126: [ed] [panic] panic when using WD-8013 with ed(4)
Date: Tue, 01 Apr 2008 23:38:46 +0200

 Dear Peter,
 
 thanks for all the requested infos.
 
 As running currently running a bit low on time, I'm unable to fully
 diagnose this issue really deep but I do have some first shots:
 
 I'm pretty sure, it's an interrupt handling problem
 
 My best hint for you for now is to update your BIOS. Please check your
 vendor site for an updated BIOS. If you really can't find one, we need
 to manually create an AML bugfix for your system as your AML code seems
 to be broken:
 > acpi_bus_number: root bus has no _BBN, assuming 0
 
 As you wrote, you're able to panic your box even w/o ACPI enabled, this
 may not relate to your problem but my experience shows, when running
 with a good ACPI implementation, the box will not show much strange
 behaviour.
 
 These lines:
 pcib0: slot 4 INTD routed to irq 9 via \\_SB_.LNKD
 pcib0: slot 9 INTA routed to irq 9 via \\_SB_.LNKD
 pcib0: slot 10 INTA routed to irq 9 via \\_SB_.LNKC
 pcib0: slot 10 INTB routed to irq 9 via \\_SB_.LNKD
 seem to indicate, the PCI slots aren't routed to IRQs nicely. Please try
 if you can fix this in the BIOS settings.
 
 Now, a bit for the real issue:
 
 Your box is receiving a packet from the network (ed0). ed0 fires an IRQ
 which is handled by the kernel. While the kernel tries to dispatch the
 IRQ to the right handler (the ed0 driver in this case), the box panics -
 for whatever reason but the "fault virtual address" is looking
 suspicious. With everything I've now checked on your issue, I'm always
 coming to the point, there's something wrong with the ithread dispatcher.
 
 Volker