diff for duplicates of <20181204212117.GC16751@localhost.localdomain> diff --git a/a/1.txt b/N1/1.txt index 02d8348..11d52b6 100644 --- a/a/1.txt +++ b/N1/1.txt @@ -1,8 +1,8 @@ -On Tue, Dec 04, 2018 at 11:33:33AM -0800, James Smart wrote: +On Tue, Dec 04, 2018@11:33:33AM -0800, James Smart wrote: > > > On 12/4/2018 9:48 AM, Keith Busch wrote: -> > On Tue, Dec 04, 2018 at 09:38:29AM -0800, Sagi Grimberg wrote: +> > On Tue, Dec 04, 2018@09:38:29AM -0800, Sagi Grimberg wrote: > > > > > > Yes, I'm very much in favour of this, too. > > > > > > We always have this IMO slightly weird notion of stopping the queue, set > > > > > > some error flags in the driver, then _restarting_ the queue, just so @@ -28,7 +28,7 @@ On Tue, Dec 04, 2018 at 11:33:33AM -0800, James Smart wrote: > > > > If you're doing a failover, you'd replace the freeze with a current path > > update in order to prevent new requests from entering. -> and if you're not multipath ? I assume you want the io queues to be +> and if you're not multipath ???? I assume you want the io queues to be > frozen so they queue there - which can block threads such as ns > verification. It's good to have them live, as todays checks bounce the io, > letting the thread terminate as its in a reset/reconnect state, which allows @@ -41,13 +41,13 @@ On Tue, Dec 04, 2018 at 11:33:33AM -0800, James Smart wrote: > > In either case, you don't need checks in queue_rq. The queue_rq check > > is redundant with the quiesce state that blk-mq already provides. > -> I disagree. The cases I've run into are on the admin queue - where we are +> I disagree.? The cases I've run into are on the admin queue - where we are > sending io to initialize the controller when another error/reset occurs, and > the checks are required to identify/reject the "old" initialization > commands, with another state check allowing them to proceed on the "new" -> initialization commands. And there are also cases for ioctls and other +> initialization commands.? And there are also cases for ioctls and other > things that occur during the middle of those initialization steps that need -> to be weeded out. The Admin queue has to be kept live to allow the +> to be weeded out.?? The Admin queue has to be kept live to allow the > initialization commands on the new controller. > > state checks are also needed for those namespace validation cases.... diff --git a/a/content_digest b/N1/content_digest index 2ed1d71..dd30fda 100644 --- a/a/content_digest +++ b/N1/content_digest @@ -7,23 +7,16 @@ "ref\011e02aa5-b0cb-b4cf-0af4-da3db8db9e36@grimberg.me\0" "ref\020181204174806.GA16751@localhost.localdomain\0" "ref\0b5b306cd-7341-7ce3-f2dc-fe98a01327fe@broadcom.com\0" - "From\0Keith Busch <keith.busch@intel.com>\0" - "Subject\0Re: [PATCH 1/2] blk-mq: Export iterating all tagged requests\0" + "From\0keith.busch@intel.com (Keith Busch)\0" + "Subject\0[PATCH 1/2] blk-mq: Export iterating all tagged requests\0" "Date\0Tue, 4 Dec 2018 14:21:17 -0700\0" - "To\0James Smart <james.smart@broadcom.com>\0" - "Cc\0Sagi Grimberg <sagi@grimberg.me>" - Jens Axboe <axboe@kernel.dk> - linux-block@vger.kernel.org <linux-block@vger.kernel.org> - Hannes Reinecke <hare@suse.de> - linux-nvme@lists.infradead.org <linux-nvme@lists.infradead.org> - " Christoph Hellwig <hch@lst.de>\0" "\00:1\0" "b\0" - "On Tue, Dec 04, 2018 at 11:33:33AM -0800, James Smart wrote:\n" + "On Tue, Dec 04, 2018@11:33:33AM -0800, James Smart wrote:\n" "> \n" "> \n" "> On 12/4/2018 9:48 AM, Keith Busch wrote:\n" - "> > On Tue, Dec 04, 2018 at 09:38:29AM -0800, Sagi Grimberg wrote:\n" + "> > On Tue, Dec 04, 2018@09:38:29AM -0800, Sagi Grimberg wrote:\n" "> > > > > > Yes, I'm very much in favour of this, too.\n" "> > > > > > We always have this IMO slightly weird notion of stopping the queue, set\n" "> > > > > > some error flags in the driver, then _restarting_ the queue, just so\n" @@ -49,7 +42,7 @@ "> > \n" "> > If you're doing a failover, you'd replace the freeze with a current path\n" "> > update in order to prevent new requests from entering.\n" - "> and if you're not multipath ?\302\240\302\240\302\240 I assume you want the io queues to be\n" + "> and if you're not multipath ???? I assume you want the io queues to be\n" "> frozen so they queue there - which can block threads such as ns\n" "> verification. It's good to have them live, as todays checks bounce the io,\n" "> letting the thread terminate as its in a reset/reconnect state, which allows\n" @@ -62,13 +55,13 @@ "> > In either case, you don't need checks in queue_rq. The queue_rq check\n" "> > is redundant with the quiesce state that blk-mq already provides.\n" "> \n" - "> I disagree.\302\240 The cases I've run into are on the admin queue - where we are\n" + "> I disagree.? The cases I've run into are on the admin queue - where we are\n" "> sending io to initialize the controller when another error/reset occurs, and\n" "> the checks are required to identify/reject the \"old\" initialization\n" "> commands, with another state check allowing them to proceed on the \"new\"\n" - "> initialization commands.\302\240 And there are also cases for ioctls and other\n" + "> initialization commands.? And there are also cases for ioctls and other\n" "> things that occur during the middle of those initialization steps that need\n" - "> to be weeded out.\302\240\302\240 The Admin queue has to be kept live to allow the\n" + "> to be weeded out.?? The Admin queue has to be kept live to allow the\n" "> initialization commands on the new controller.\n" "> \n" "> state checks are also needed for those namespace validation cases....\n" @@ -94,4 +87,4 @@ "have nvme-fc control when to dispatch those queued requests by utilizing\n" the blk-mq quiesce on/off states. Is there a reason that wouldn't work? -4d67c568181d67ab3865feaa3aa606b6d18e5c575b87e773c5d3d557bae35df1 +d7e4484f590535402dd1f2c0e07e2c89cfb9227e062c73a7f0eef73b709a9c9b
This is an external index of several public inboxes, see mirroring instructions on how to clone and mirror all data and code used by this external index.