Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 30 Nov 2004 12:13:46 -0800
From:      Nate Lawson <nate@root.org>
To:        John Baldwin <jhb@FreeBSD.org>
Cc:        cvs-all@FreeBSD.org
Subject:   Re: cvs commit: src/sys/dev/acpica acpi_pci_link.c
Message-ID:  <41ACD47A.3020002@root.org>
In-Reply-To: <200411301341.01331.jhb@FreeBSD.org>
References:  <200411301748.iAUHmAiK010161@repoman.freebsd.org> <200411301341.01331.jhb@FreeBSD.org>

next in thread | previous in thread | raw e-mail | index | archive | help
John Baldwin wrote:
> On Tuesday 30 November 2004 12:48 pm, Nate Lawson wrote:
> 
>>njl         2004-11-30 17:48:10 UTC
>>
>>  FreeBSD src repository
>>
>>  Modified files:
>>    sys/dev/acpica       acpi_pci_link.c
>>  Log:
>>  Don't bother locking in attach().  At boot time, we're single-threaded
>>  anyway and for some reason, witness seems confused about what's already
>>  locked and triggers a false panic.
> 
> 
> Witness doesn't get confused, there's some other bug.  I'm not sure that all 
> device attachments are always single-threaded though.  Imagine hotplugging a 
> docking station, that's going to add a new PCI bus.  It probably won't have 
> an internal programming router (probably not physically possible), but I 
> don't think we should assume for now that attach is only a boot-time 
> scenario.

Some other guy had a problem before that was similar and Warner seemed 
to think that the xlock was being confused with some other lock since 
there was no way the lock could have already been held elsewhere.

I'm happy to let you troubleshoot this and sorry for any interference.

-Nate



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?41ACD47A.3020002>