From mboxrd@z Thu Jan 1 00:00:00 1970 From: Paul Gortmaker Subject: Re: [ANNOUNCE] 4.8-rc5-rt1 beta Date: Sun, 2 Oct 2016 19:35:15 -0400 Message-ID: <20161002233515.GA18997@windriver.com> References: <20160906182508.GB9950@windriver.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Cc: LKML , Thomas Gleixner , Sebastian Andrzej Siewior , Steven Rostedt To: linux-rt-users Return-path: Content-Disposition: inline In-Reply-To: <20160906182508.GB9950@windriver.com> Sender: linux-kernel-owner@vger.kernel.org List-Id: linux-rt-users.vger.kernel.org [[ANNOUNCE] 4.8-rc5-rt1 beta] On 06/09/2016 (Tue 14:25) Paul Gortmaker wrote: > This is a carry forward of the -rt patches off v4.6.7-rt11 through > mainline commits of v4.7 and up to the current v4.8-rc5. > > https://git.kernel.org/cgit/linux/kernel/git/paulg/4.8-rt-patches.git > [...] > > Patch conflicts/issues of interest 4.7 --> 4.8 > ---------------------------------------------- > -mainline changed the qdisc from a simple bit to a seqcount; this causes a splat from migrate_enable, but I didn't initially see that since the sanity check relies on SCHED_DEBUG=y. This causes the -rt patch that faked out the lock status to splat; further details are at kernel.org: https://goo.gl/0fdRKR No point in anyone else having to redo the root cause research on this a second time; I've updated the above repo for this and for some other mm fallout from 4.8-rc7 in memcontrol.c so that the patches will apply for what inevitably becomes v4.8 final in a couple hours. That said, my original comments still apply -- this remains as a resource for developers who track the -rt progress, and how mainline will impact it. The release of 4.8 in mainline does not somehow change this repo into a stable/supported -rt release by osmosis. If you don't regularly manually apply and test patches, including resolving sleeping while atomic kernel backtrace dumps, you probably should not be using this. Paul.