cvs commit: src/sys/sys malloc.h src/sys/kern kern_malloc.c src/sys/vm vm_contig.c vm_page.h

[John Baldwin]

On Monday 19 July 2004 06:39 am, Brian Fundakowski Feldman wrote:
> On Mon, Jul 19, 2004 at 02:11:47AM -0500, Alan Cox wrote:
> > On Mon, Jul 19, 2004 at 01:00:27AM -0600, Scott Long wrote:
> > > This is a pretty darn big change to something that is so vital to so
> > > many drivers (not including busdma, which means that this touches
> > > nearly every driver).  Who reviewed this?  Was it discussed on a public
> > > mailing list?  What kind of testing was done for it?  If you can't
> > > answer these questions then I'm not terribly thrilled.
> >
> > I reviewed an earlier version of this and gave Brian some feedback.
> >
> > It is important to note that he added the new implementation side-by-side
> > with the old. By default, the old implementation is still used.  So, I
> > wouldn't be too concerned.
>
> Aye, I didn't feel totally comfortable saying "Reviewed by:" since that
> was a while ago, but between alc and hmp who has been overhauling the
> contigmalloc() implementation in DFBSD there wasn't any concern that I
> know I hadn't addressed (and Dillon ignored my solicitation).

For future reference: In this case I usually add a line along the lines of:

Reviewed by:	bde (an earlier version)

Which is both truthful and avoids the ensuing flamewar due to the appearance 
of an unreviewed driveby. :)

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