The following reply was made to PR kernel/6348; it has been noted by GNATS.
From: Emilio Perea <[email protected]> To: "J.C. Roberts" <[email protected]> Cc: [email protected] Subject: Re: kernel/6348: -current i386 kernel crashes with parity errors Date: Thu, 22 Apr 2010 17:07:42 -0500 On Sat, Apr 03, 2010 at 07:24:57AM -0700, J.C. Roberts wrote: > On Fri, 2 Apr 2010 11:15:02 -0500 Emilio Perea <[email protected]> > wrote: > > > The problem came up for the first time on kernel > > built on the evening of 26 March. One built on 23 March works fine, > > as does the one in the current snapshot (dmesg picked up by sendbug). > > Now that cvs has been unlocked after the 4.7 release, and tagged > 4.7-current, this is when the heavy development occurs and large > changes get committed. > > After the March 24th snapshot, a number of important changes went into > the tree (pmem, isa, dma, ...). Since the problem has now disappeared > in the April 1st snapshot, you probably updated your local cvs checkout > at *just* the wrong time on the 26th, and only got some of the needed > commits. As of noon (US/Central) today, the -current cvs checkout GENERIC kernel still crashed. However, I'm happy to say that the 1600 checkout does not show the problem. AFAIK, none of the snapshot builds ever had this problem, so I didn't bother reporting that the problem continued until now. The fix was obviously in the pipeline somewhere, and I only wondered what had not been committed yet. Anyhow, this problem appears to be fixed. Thanks again to all involved in this project!
