Opened 9 years ago

Closed 9 years ago

Last modified 7 years ago

#12377 closed bug (fixed)

PANIC: calculated irq routing doesn't match bios for PCI 24:0:0

Reported by: vidrep Owned by: tqh
Priority: normal Milestone: R1/beta1
Component: System/Kernel Version: R1/Development
Keywords: Cc: tqh
Blocked By: Blocking: #12378
Platform: All

Description (last modified by diver)

Update from hrev49636 to hrev49638 x86_gcc2 Upon reboot KDL (photo attached)

Attachments (3)

DSC00034.JPG (1.0 MB ) - added by vidrep 9 years ago.
syslog (489.4 KB ) - added by vidrep 9 years ago.
09142015.txt (35.7 KB ) - added by vidrep 9 years ago.

Download all attachments as: .zip

Change History (21)

by vidrep, 9 years ago

Attachment: DSC00034.JPG added

comment:1 by vidrep, 9 years ago

I booted into x86_64 and grabbed the syslog from the non-booting x86_gcc2 partition (attached). Hopefully it will be of some help.

by vidrep, 9 years ago

Attachment: syslog added

comment:2 by diver, 9 years ago

Cc: tqh added
Component: - GeneralSystem/Kernel
Milestone: UnscheduledR1/beta1
Owner: changed from nobody to mmlr
Status: newassigned
Summary: hrev49638 update causes KDL on bootPANIC: calculated irq routing doesn't match bios for PCI 24:0:0
Last edited 9 years ago by diver (previous) (diff)

comment:3 by diver, 9 years ago

Blocking: 12378 added

comment:4 by diver, 9 years ago

Description: modified (diff)

comment:5 by mmlr, 9 years ago

Owner: changed from mmlr to tqh

Obviously something changed in the info we get from ACPICA. I don't think that it is generally broken, as that should throw much earlier (i.e. not finding a mapping at all, not finding one that doesn't match the BIOS). In any case more info is needed. The attached syslog isn't really helpful since the syslog from a boot is only flushed to disk once the syslog_daemon runs and it therefore doesn't contain any info from the failed boot. Instead please get a syslog via the bootloader menu (reboot using the reboot command in KDL and get the syslog as described in ReportingBugs).

comment:6 by vidrep, 9 years ago

I have this PC set up for serial debugging. Would capturing the output during boot help?

comment:7 by diver, 9 years ago

It would.

comment:8 by tqh, 9 years ago

I can reproduce the issue in a 32 bit build. If the errors just before is a hint the passed in buffer to ACPI is not big enough. I'll see what I can do.

Unfortunatly the boot from USB didn't preserve the syslog, and the syslog in this bug don't seem to be from a failed boot.

comment:9 by tqh, 9 years ago

It would help to provide a image of syslog: Type 'syslog' in KDL and scroll to end, then take image of that.

comment:10 by tqh, 9 years ago

There has been a change in address size in ACPI, 'temporary' fix coming once I've verified.

comment:11 by vidrep, 9 years ago

I tried to type commands after it went into KDL, but there are no keyboard functions. All I could do was reboot. I'll try to get a serial debug log tonight after work, if you haven't already fixed the problem by then.

comment:12 by tqh, 9 years ago

Should be fixed. Try hrev49642.

comment:13 by vidrep, 9 years ago

Attached is a serial debug log of the boot.

by vidrep, 9 years ago

Attachment: 09142015.txt added

comment:14 by vidrep, 9 years ago

The hrev49462 build now boots correctly.

comment:15 by tqh, 9 years ago

Resolution: fixed
Status: assignedclosed

Thanks for reporting! Closing this, will do some more work related to this later though.

comment:16 by vidrep, 9 years ago

Maybe keep the ticket open so the later work isn't forgotten, as sometimes happens ;) Thanks for the fix

comment:17 by tqh, 9 years ago

That would only lead to this bug being forgotten. Then you have two problems :)

comment:18 by tqh, 7 years ago

Pushed a real fix today, worked on my machine..

Note: See TracTickets for help on using tickets.