Re: 5.1-CURRENT hangs on disk i/o? sysctl_old_user() non-sleepable locks

From: Stefan Eßer <se_at_freebsd.org>
Date: Fri, 20 Jun 2003 14:04:11 +0200
On 2003-06-19 08:13 -0700, Don Lewis <truckman_at_FreeBSD.org> wrote:
> > In PR kern/46652 I reported, that DEBUG_VFS_LOCKS does never
> > check the **vpp parameters. A patch is included in the PR and
> > it does generate the missing tests.
> > 
> > I asked for feedback on the hackers mail list (IIRC), but did
> > not get any replies. 
> > 
> > Any objections against me committing the patch now ?
> > 
> > (A different fix is mentioned in the PR, the patch I suggested
> > was the minimal change to the code which made it work, the
> > alternative seems cleaner to me ...) Please read PR kern/46652 !
> 
> I think the alternative fix should be committed.  That would do the
> correct thing if another pointer to a pointer to a vnode argument is
> ever added.  I think this is better than adding magic to vpp.

Well, the alternative fix is much more work than I thought ...
I spent an hour on it, but the parameter names are assumed to only
consist of [a-z]* in a number of places and fixing this would add
complexity to the AWK script and make it harder to maintain.

Instead, I'm going to commit the trivial fix and add a comment about
double indirection as in **vpp requiring special code in the AWK script
to vnode_if.src ...

This will fix the debug code that is generated and who ever cares for
a better solution may back-out my two line fix and implement the clean
solution.

> Any idea if this change turns up more problems?

Sorry, no. I found this while looking for some other problem and just
opened the PR and sent a note to the hackers mail list ...

Regards, STefan
Received on Fri Jun 20 2003 - 03:04:17 UTC

This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:37:12 UTC