From mboxrd@z Thu Jan 1 00:00:00 1970 From: Greg Kroah-Hartman Subject: Re: linux-next: Tree for Jun 18 (kmsg problems) Date: Mon, 18 Jun 2012 15:45:56 -0700 Message-ID: <20120618224556.GA2750@kroah.com> References: <20120618165336.48345227c6b64c314326ae57@canb.auug.org.au> <4FDF61A7.3050601@xenotime.net> <20120619080109.b28055c1b32fb81027a9562b@canb.auug.org.au> <20120618221535.GA13037@kroah.com> <20120619083720.996b2d743cff619a58e1de97@canb.auug.org.au> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Received: from mail-pz0-f46.google.com ([209.85.210.46]:55922 "EHLO mail-pz0-f46.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751348Ab2FRWqA (ORCPT ); Mon, 18 Jun 2012 18:46:00 -0400 Received: by dady13 with SMTP id y13so7547772dad.19 for ; Mon, 18 Jun 2012 15:46:00 -0700 (PDT) Content-Disposition: inline In-Reply-To: <20120619083720.996b2d743cff619a58e1de97@canb.auug.org.au> Sender: linux-next-owner@vger.kernel.org List-ID: To: Stephen Rothwell Cc: Randy Dunlap , linux-next@vger.kernel.org, LKML , Kay Sievers On Tue, Jun 19, 2012 at 08:37:20AM +1000, Stephen Rothwell wrote: > Hi Greg, > > On Mon, 18 Jun 2012 15:15:35 -0700 Greg Kroah-Hartman wrote: > > > > > Caused by commit e2ae715d66bf ("kmsg - kmsg_dump() use iterator to > > > receive log buffer content") from the driver-core.current tree. > > > > I'm building my tree right now, and I can't seem to hit this. > > You need CONFIG_PRINTK turned off ... Ugh, no wonder I couldn't reproduce it, who runs systems like that becides the CONFIG_TINY people? :) > > > This was reported a few days ago. Why is it not fixed yet? > > > > It was reported Friday evening, on a holiday weekend for some of us, we > > are trying the best we can :) > > Ah ha, another of those pesky holiday things :-) > > > > Especially since it is in a branch that is supposedly "bug fixes"? > > > > It fixes up a reported problem, and we had 3 people test it and said it > > resolved the issue for them, no build problems reported. > > My usual mantra is that if your patch is modifying code that clearly > depends on a CONFIG option, you must test with that option both on and > off. I see build problems in linux-next all the time where this has not > been done. :-( Thanks, as this was a kmsg thing, I didn't catch that CONFIG_PRINTK was involved, sorry about that. I'll go queue up Kay's patch right now, greg k-h