LOR route vr0

[John Baldwin]

On Sunday 28 August 2005 11:31 am, Bjoern A. Zeeb wrote:
> On Sun, 28 Aug 2005, M. Warner Losh wrote:
> > In message:
> > <Pine.BSF.4.53.0508281123210.37621 at e0-0.zab2.int.zabbadoz.net>
> >
> >             "Bjoern A. Zeeb" <bzeeb-lists at lists.zabbadoz.net> writes:
> > : On Sat, 27 Aug 2005, M. Warner Losh wrote:
> > :
> > : for the archives...
> > :
> > : > ock order reversal
> > : >  1st 0xc17490e4 rtentry (rtentry) @ sys/netinet/if_ether.c:445
> > : >  2nd 0xc15c94b0 rl1 (network driver) @ sys/pci/if_rl.c:1451
> > :
> > : LOR http://sources.zabbadoz.net/freebsd/lor.html#144
> >
> > I cut and pasted this one from your LOR site...
>
> *gnaa* my grep-o-logic failed. I should really really really split
> this page...
>
> I removed the duplicate.
> The original one was http://sources.zabbadoz.net/freebsd/lor.html#139
>
> > : > and am seeing the following in my newly locked ed driver:
> > : >
> > : > lock order reversal
> > : >  1st 0xc1cb3588 rtentry (rtentry) @ net/route.c:1269
> > : >  2nd 0xc1fd3420 ed1 (network driver) @
> > : > /dell/imp/p4/newcard/src/sys/modules/ed/../../dev/ed/if_ed.c:697
> > :
> > : I added this one two though this LOR may not (yet) be seen in
> > : the wild.
> > :
> > : LOR http://sources.zabbadoz.net/freebsd/lor.html#145
> >
> > heh.  I'm expecting to figure out what the real LOR is here and then
> > about 20-25 of the LORs in your list can go away...
>
> great. let me know the IDs and the patch that will fix them then.

Note that all network driver locks are considered the same in witness (they 
have a lock "type" of MTX_NETWORK_LOCK).  You probably want to look at the 
lock "type" for LORs and don't log duplicate LORs for the same type.  The 
only widespread case of locks that use a type other than their name are 
network device drivers.  I imagine that other device drivers might do this in 
the future as well.

-- 
John Baldwin <jhb at FreeBSD.org>  <><  http://www.FreeBSD.org/~jhb/
"Power Users Use the Power to Serve"  =  http://www.FreeBSD.org