Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 9 Aug 2004 19:20:29 GMT
From:      John Baldwin <jhb@FreeBSD.org>
To:        freebsd-bugs@FreeBSD.org
Subject:   Re: kern/68351: bge0 watchdog timeout on 5.2.1 and -current, 5.1 is ok
Message-ID:  <200408091920.i79JKTd0045320@freefall.freebsd.org>

next in thread | raw e-mail | index | archive | help
The following reply was made to PR kern/68351; it has been noted by GNATS.

From: John Baldwin <jhb@FreeBSD.org>
To: Vadim Mikhailov <freebsd-bugs@mikhailov.org>
Cc: freebsd-gnats-submit@FreeBSD.org
Subject: Re: kern/68351: bge0 watchdog timeout on 5.2.1 and -current, 5.1 is ok
Date: Mon, 9 Aug 2004 15:06:32 -0400

 On Friday 06 August 2004 04:22 pm, Vadim Mikhailov wrote:
 > John Baldwin wrote:
 > > On Friday 06 August 2004 03:10 pm, Vadim Mikhailov wrote:
 > >>John Baldwin wrote:
 > >>>On Friday 06 August 2004 02:02 pm, Vadim Mikhailov wrote:
 > >>>>John Baldwin wrote:
 > >>>>>Have you tried disabling ACPI to see if it works better?
 > >>>>
 > >>>>  Can not try it now - this is production server under load,
 > >>>>and it is working fine since I installed Intel Pro/1000 XT card to
 > >>>>overcome bge problems.
 > >>>>  But, as stated in my original report, I did try it disabling ACPI
 > >>>>about month ago with current - didn't make any difference.
 > >>>>
 > >>>>>Also, can you provide mptable output?
 > >>>>
 > >>>>  Here it is:
 > >>>
 > >>>Thanks, can you also provide a full dmesg output?  The two bge's are
 > >>>routed to pins 0 and 1 of the second I/O APIC which should be IRQs 16
 > >>> and 17 if the first I/O APIC has 16 interrupt pins (dmesg will tell me
 > >>> that).
 > >>
 > >>   This is /var/run/dmesg.boot from verbose boot. Oddly enough, it does
 > >> not start with standard FreeBSD copyrights, probably because of dmesg
 > >> buffer overflow, however I didn't think this limit applies to
 > >> /var/run/dmesg.boot file. Hopefully it still has enough information
 > >> about APIC and interrupts.
 > >
 > > Thanks.  It looks like the interrupts are routed just as both ACPI and
 > > the MP Table say they should be.  Have you tried using cvs to change the
 > > sys/dev/bge source back to 5.1 sources with the rest of the kernel being
 > > 5.2.1 to see if it works ok?
 >
 >    Yes I did. Unfortunately difference between 5.1 and 5.2.1 is just too
 > big:
 > http://www.freebsd.org/cgi/cvsweb.cgi/src/sys/dev/bge/if_bge.c.diff?f=h&r1=
 >1.41&r2=1.60 With 5.2.1-p9 sources when I replace if_bge.c rev. 1.60
 > (5_2_1_RELEASE) with rev. 1.41 (5_1_0_RELEASE) kernel would not compile at
 > all:
 >
 > /usr/src/sys/dev/bge/if_bge.c:107:24: pci/pcireg.h: No such file or
 > directory /usr/src/sys/dev/bge/if_bge.c:108:24: pci/pcivar.h: No such file
 > or directory mkdep: compile failed
 > *** Error code 1
 
 Just change the files to say '#include <dev/pci/pcifoo.h>', that is, add the 
 'dev/' to the start of the filename and that should fix that problem.  If you 
 could try with those changes I'd appreciate it.
 
 -- 
 John Baldwin <jhb@FreeBSD.org>  <><  http://www.FreeBSD.org/~jhb/
 "Power Users Use the Power to Serve"  =  http://www.FreeBSD.org



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200408091920.i79JKTd0045320>