From mboxrd@z Thu Jan 1 00:00:00 1970 From: axboe@kernel.dk (Jens Axboe) Date: Fri, 16 Mar 2018 09:26:24 -0700 Subject: [GIT PULL] nvme fix for 4.16-rc6 In-Reply-To: <20180316161655.GA13293@lst.de> References: <20180316160110.GC720@localhost.localdomain> <20180316161655.GA13293@lst.de> Message-ID: <8734416c-cd14-8bfb-8e2c-6eab84e0ce5d@kernel.dk> On 3/16/18 9:16 AM, Christoph Hellwig wrote: > On Fri, Mar 16, 2018@09:14:34AM -0700, Jens Axboe wrote: >> On 3/16/18 9:01 AM, Keith Busch wrote: >>> Hi Jens, >>> >>> One last fix from Jianchao to address a troubling scenario Ming identified >>> with offlining CPUs that could leave nvme controllers unusable. We can >>> not allow that to occur, so we wish to include a new IRQ spread that >>> prevents that situation. >> >> Since this isn't really a regression in this release, how about >> we push this one to 4.17 instead? > > It is a regression in 4.15. So I think we should fix this ASAP. It's not that I dislike the patch (in fact it makes the code easier to read), but it's pretty late for something that isn't a regression in this series. I can queue it up for some testing, but it's then -rc7 time. I guess we can see how it goes and push the decision until start next week. -- Jens Axboe