Re: syncing failed on shutdown: giving up on 1 buffer
:I installed DFBSD 1.0 a few days ago and I immediatly upgraded to STABLE
:with cvsup and enabled SMP. I am having a few problems installing the
:XFree86 ports from dfports, but this is not the matter here.
:
:In the process, I installed the dri port from the package at gobsd.com. It
:spat the following warnings:
:
:dragonfly# pkg_add -r dri
:Fetching http://gobsd.com/packages//Latest/dri.tgz... Done.
:mtree: lib/modules/dri/gamma_dri.so: No such file or directory
:mtree: lib/modules/dri/i810_dri.so: No such file or directory
This is a namecache bug that I had hoped was fixed in Stable. Stable
still has a bunch of compatibility code but most of it has been ripped
out in HEAD and the problem definitely not occur in HEAD.
I intend to sync Stable to Head the moment I get nullfs and unionfs
working again in HEAD.
:Please note that tab completion still works! It completed gamma_dri.so.
:
:...
:After this I get a bit suspicious and decide to go down to single user, unmount
:the /usr volume and fsck it. But everything is fine, which amazes me.
Yah, there's nothing wrong with the filesystem, it's the namecache
that was causing the confusion.
:After which, I /sbin/reboot:
:
:# /sbin/reboot
:boot() called on cpu#1
:Waiting (max 60 seconds) for system thread vnlru to stop...stopped
:Waiting [...] bufdaemon to stop...stopped
:Waiting [...] syncer to stop...stopped
:
:syncing disks... 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
:giving up on 1 buffers
:Debugger("busy buffer problem")
:...
:regards
:
:Raphael Marmier
We get reports of this occuring occassionally on SMP system and we
haven't quite figured out the cause yet. If this is repeatable on
your system and you can get a kernel core out of it that would be
helpful (make sure you are compiling up debug kernels, which we have
a default in GENERIC so you should be unless you removed the makeoptions
line for it).
-Matt
Matthew Dillon
<dillon@backplane.com>
討論串 (同標題文章)
完整討論串 (本文為第 2 之 3 篇)：