Bug 845

Summary:	VIA ACPI/APIC interrupts problems
Product:	ACPI	Reporter:	Nicolas Mailhot (Nicolas.Mailhot)
Component:	Config-Interrupts	Assignee:	Len Brown (lenb)
Status:	CLOSED PATCH_ALREADY_AVAILABLE
Severity:	blocking	CC:	acpi-bugzilla
Priority:	P2
Hardware:	i386
OS:	Linux
Kernel Version:	2.5.46 to 2.6.0-test4-bk2	Subsystem:
Regression:	---	Bisected commit-id:
Bug Depends on:	10, 71, 678, 1115, 1120
Bug Blocks:
Attachments:	Output of VIA C3 which won't work unless acpi=off is specified on 2.6.0-test4+ VIA IRQ routing fix

Description Nicolas Mailhot 2003-06-23 04:50:58 UTC

This is an umbrella bug to help trace all the VIA ACPI/APIC bug people have been
getting lately.

Most of the failures hit USB and/or ethernet subsystems and occur with varying
ACPI/LOCAL-APIC/IO-ACPI permutations on a wide range of motherboards

[ simply put via acpi is broken right now only the actual symptoms differ ]

People who hit this class of bug should open a new bug and mark it as blocker
for this one so it can be closed once all repported via acpi problems are
solved. It is very possible a single quirk will fix all via systems at once.

Comment 1 Nicolas Mailhot 2003-08-25 13:36:05 UTC

2.6.0-test4-bk2 : bug #10 still stuck, bug #71 almost solved

Comment 2 Nicolas Mailhot 2003-08-30 02:09:46 UTC

Bug #71 boot is solved except usb mouse is dog-slow

Bug #10 has a working fix for one system. Please everyone test it on their systems

Comment 3 Christian Guggenberger 2003-08-31 13:47:29 UTC

The working fix for Bug 10 does fix the IO-APIC problems here with my EPOX 8k5a3+.
That's the first time I got net, usb and sound working in APIC mode. thanks!

Comment 4 Danny ter Haar 2003-09-03 08:07:12 UTC

Created attachment 805 [details]
Output of VIA C3 which won't work unless acpi=off is specified on 2.6.0-test4+

the file "out" contains output of several of the suggested programs for ACPI
output

Comment 5 Nicolas Mailhot 2003-09-03 08:39:36 UTC

Please create a bug per system and make them block bug 845.

There are too many via systems suffering acpi/apic bugs to untangle them if they
are all put in a single report (plus some of the failures may actually be caused
by different bugs, so this way we can close a report each time a bug is squashed).

Comment 6 Andrew de Quincey 2003-09-07 14:13:10 UTC

Created attachment 834 [details]
VIA IRQ routing fix

Comment 7 Andrew de Quincey 2003-09-07 14:13:36 UTC

Please try the attached patch. It should fix VIA IRQ routing problems.

Comment 8 Len Brown 2003-12-04 13:30:07 UTC

The bugs this one depends on have all been resolved -- closing.