From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Message-ID: <1536163037.11534.6.camel@acm.org> Subject: Re: [PATCH 2/3] blk-core: introduce queue close feature From: Bart Van Assche To: Jianchao Wang , axboe@kernel.dk, ming.lei@redhat.com, bart.vanassche@wdc.com, sagi@grimberg.me, keith.busch@intel.com, jthumshirn@suse.de, jsmart2021@gmail.com Cc: linux-block@vger.kernel.org, linux-kernel@vger.kernel.org, linux-nvme@lists.infradead.org Date: Wed, 05 Sep 2018 08:57:17 -0700 In-Reply-To: <1536120586-3378-3-git-send-email-jianchao.w.wang@oracle.com> References: <1536120586-3378-1-git-send-email-jianchao.w.wang@oracle.com> <1536120586-3378-3-git-send-email-jianchao.w.wang@oracle.com> Content-Type: text/plain; charset="UTF-7" Mime-Version: 1.0 List-ID: On Wed, 2018-09-05 at 12:09 +-0800, Jianchao Wang wrote: +AD4 blk queue freeze is often used to prevent new IO from entering +AD4 request queue. However, becuase we kill the percpu-ref +AD4 q+AF8-usage+AF8-counter when freeze queue, we have to drain the request +AD4 queue when unfreeze. This is unnecessary for just preventing new +AD4 IO. In addition, If there is IO timeout or other issue when unfreeze +AD4 the queue, the scenario could be very tricky. +AD4 +AD4 So we introduce BLK+AF8-QUEUE+AF8-GATE+AF8-CLOSED to implement a light-weight +AD4 queue close feature base on the queue+AF8-gate to prevent new IO from +AD4 comming in queue which will not need to drain the queue any more. Does the +ACI-queue gate close+ACI feature cause blk+AF8-get+AF8-request() / blk+AF8-mq+AF8-get+AF8-request() to block until blk+AF8-clear+AF8-queue+AF8-closed() is called? If so, I think we need a better name for this feature. How about calling these two operations suspend and resume? Thanks, Bart.