Re: kern/47648: NetBSD 6.1_RC1 ACPI interupt routing problem + other ACPI lossage

[Chuck Silvers <chuq%chuq.com@localhost>]

The following reply was made to PR kern/47648; it has been noted by GNATS.

From: Chuck Silvers <chuq%chuq.com@localhost>
To: gnats-bugs%NetBSD.org@localhost
Cc: kern-bug-people%netbsd.org@localhost, gnats-admin%netbsd.org@localhost,
        netbsd-bugs%netbsd.org@localhost
Subject: Re: kern/47648: NetBSD 6.1_RC1 ACPI interupt routing problem + other
 ACPI lossage
Date: Fri, 15 Mar 2013 08:52:11 -0700

 On Thu, Mar 14, 2013 at 02:55:00PM +0000, reinoud%NetBSD.org@localhost wrote:
 > Updating a P4 machine from NetBSD 6.0_BETA2 to NetBSD 6.1_RC1 fails due to
 > various ACPI lossage. It results in a machine having hundreds of thousands of
 > interrupts/second that are not handled properly and in broken devices, most
 > notably auixp(9).
 
 what is the specific problem with auixp?  what other devices have trouble
 and what's the specific problem with those?
 
 does this patch make any difference?
 http://mail-index.netbsd.org/port-amd64/2013/02/19/msg001860.html
 
 I made several changes to the ACPI code in the 6.x branch, in these files:
 
 sys/arch/x86/acpi/acpi_machdep.c
 sys/arch/x86/include/acpi_machdep.h
 sys/arch/x86/x86/mpacpi.c
 sys/dev/acpi/acpi.c
 
 if you revert all of those files to the netbsd-6-0-RELEASE tag,
 does that make the problem go away?  if so, could you turn on
 MPVERBOSE and INTRDEBUG and send the output from that with ACPI enabled?
 (with the current netbsd-6 versions of those files, not the older versions)
 
 also, could you provide the output of "acpidump -dt"?
 
 -Chuck