From mboxrd@z Thu Jan 1 00:00:00 1970 From: Roderick Colenbrander Subject: Re: Questions about GPLPV stability tests Date: Fri, 16 Dec 2011 02:25:25 +0000 Message-ID: References: <20111208234624.GD32474@konrad-lan> <20111211125247.GQ12984@reaktio.net> <20111213020229.GB2730@konrad-lan> <20111213212031.GA8410@konrad-lan> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <20111213212031.GA8410@konrad-lan> List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Sender: xen-devel-bounces@lists.xensource.com Errors-To: xen-devel-bounces@lists.xensource.com To: Konrad Rzeszutek Wilk Cc: xen-devel@lists.xensource.com, Andreas Kinzler List-Id: xen-devel@lists.xenproject.org Thanks. On Tuesday I set up your monitoring tool on the already running tests. Some of the systems were on the verge of showing results, but we just had a power outage and I have to restart the tests. It will probably take until the Holidays before there any results :( Roderick On Tue, Dec 13, 2011 at 9:20 PM, Konrad Rzeszutek Wilk wrote: >> > Does this feel like blktap issues? Or is it more likely that something >> > else happened which causes blktap and the other things to fail? >> >> It looks like the interrupts stopped. Perhaps you can run the C code >> (attached) on the serial console and see _what_ (or perhaps _when_) >> the interrupts stops (and also verify that the timeouts and 'frozen' >> happen due to no interrupts being received). > > The C code is http://darnok.org/xen/read_interrupts.c > >> >> There are a couple of bugs that were discovered in the interrupt code >> (that had been there since 2.6.27!) that went to the stable tree - so >> they should been backported to 2.6.32.x tree - but I honestly don't >> recall the names. I can look them up tomorrow. > > xen: x86_32: do not enable iterrupts when returning from exception in > interrupt context (git commit d198d499148a0c64a41b3aba9e7dd43772832b91) >