From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jens Axboe Subject: Re: [PULL] bcache fixes for v4.5-rc6 Date: Wed, 2 Mar 2016 20:26:43 -0700 Message-ID: <56D7AEF3.3040508@kernel.dk> References: <56D5B8E3.9050000@kernel.dk> Mime-Version: 1.0 Content-Type: text/plain; charset=windows-1252; format=flowed Content-Transfer-Encoding: 7bit Return-path: Received: from mail-pf0-f181.google.com ([209.85.192.181]:35368 "EHLO mail-pf0-f181.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752425AbcCCD0s (ORCPT ); Wed, 2 Mar 2016 22:26:48 -0500 Received: by mail-pf0-f181.google.com with SMTP id w128so6368453pfb.2 for ; Wed, 02 Mar 2016 19:26:47 -0800 (PST) In-Reply-To: Sender: linux-bcache-owner@vger.kernel.org List-Id: linux-bcache@vger.kernel.org To: Eric Wheeler Cc: Marc MERLIN , linux-bcache@vger.kernel.org, Zhu Yanhai , Kent Overstreet , Christoph Nelles On 03/02/2016 07:09 PM, Eric Wheeler wrote: > On Tue, 1 Mar 2016, Jens Axboe wrote: > >> On 02/29/2016 06:08 PM, Eric Wheeler wrote: >>> Hi Jens, >>> >>> Can we still get these in for 4.5? One is a race, the other is just error >>> handling. >>> >>> Please pull: >>> git pull https://bitbucket.org/ewheelerinc/linux.git >>> v4.5-rc6-bcache-fixes >>> >>> bcache: fix race of writeback thread starting before complete initialization >>> bcache: cleaned up error handling around register_cache() >> >> They look simple enough, though I generally don't like collapsing things like >> this: >> >> if ((ret = cache_alloc(sb, ca)) != 0) >> >> it's easy to screw up, doing >> >> ret = cache_alloc(sb, ca); >> if (ret != 0) >> ... > > Done. Rebased, fixed, and push --force'ed. Cool, I will pull it in. > BTW, Is there a better (preferred) way vs. push --force? Some say push > --force can break things. If you rebased, there's really no way around that, other than spinning a new branch with the new changes. The rebase is what breaks things for people, the --force is just a side effect since you have to do that to push it to the same branch again. It's only really a concern if people are continually pulling from you and developing on top of your branch, for something like this, it's not a huge problem. -- Jens Axboe