Re: [PATCH 1/2] block: generic request_queue reference counting

On Tue, Oct 6, 2015 at 3:46 AM, Ming Lei wrote: > On Tue, Oct 6, 2015 at 7:23 AM, Dan Williams wrote: >> On Sun, Oct 4, 2015 at 12:52 AM, Ming Lei wrote: >>> On Wed, Sep 30, 2015 at 8:41 AM, Dan Williams >>> wrote: Allow pmem, and other synchronous/bio-based block drivers, to fallback

Re: [PATCH 1/2] block: generic request_queue reference counting

On Tue, Oct 6, 2015 at 7:23 AM, Dan Williams wrote: > On Sun, Oct 4, 2015 at 12:52 AM, Ming Lei wrote: >> On Wed, Sep 30, 2015 at 8:41 AM, Dan Williams >> wrote: >>> Allow pmem, and other synchronous/bio-based block drivers, to fallback >> >> Just a bit curious, why not extend it for all(both

Re: [PATCH 1/2] block: generic request_queue reference counting

On Tue, Oct 6, 2015 at 7:23 AM, Dan Williams wrote: > On Sun, Oct 4, 2015 at 12:52 AM, Ming Lei wrote: >> On Wed, Sep 30, 2015 at 8:41 AM, Dan Williams >> wrote: >>> Allow pmem, and other synchronous/bio-based block

Re: [PATCH 1/2] block: generic request_queue reference counting

On Tue, Oct 6, 2015 at 3:46 AM, Ming Lei wrote: > On Tue, Oct 6, 2015 at 7:23 AM, Dan Williams wrote: >> On Sun, Oct 4, 2015 at 12:52 AM, Ming Lei wrote: >>> On Wed, Sep 30, 2015 at 8:41 AM, Dan Williams

Re: [PATCH 1/2] block: generic request_queue reference counting

On Sat, Oct 3, 2015 at 11:40 PM, Christoph Hellwig wrote: > On Tue, Sep 29, 2015 at 08:41:31PM -0400, Dan Williams wrote: >> Allow pmem, and other synchronous/bio-based block drivers, to fallback >> on a per-cpu reference count managed by the core for tracking queue >> live/dead state. >> >> The

Re: [PATCH 1/2] block: generic request_queue reference counting

On Sun, Oct 4, 2015 at 12:52 AM, Ming Lei wrote: > On Wed, Sep 30, 2015 at 8:41 AM, Dan Williams > wrote: >> Allow pmem, and other synchronous/bio-based block drivers, to fallback > > Just a bit curious, why not extend it for all(both synchronous and > asynchrounous) bio-based drivers? As you

Re: [PATCH 1/2] block: generic request_queue reference counting

On Sat, Oct 3, 2015 at 11:40 PM, Christoph Hellwig wrote: > On Tue, Sep 29, 2015 at 08:41:31PM -0400, Dan Williams wrote: >> Allow pmem, and other synchronous/bio-based block drivers, to fallback >> on a per-cpu reference count managed by the core for tracking queue >> live/dead

Re: [PATCH 1/2] block: generic request_queue reference counting

On Sun, Oct 4, 2015 at 12:52 AM, Ming Lei wrote: > On Wed, Sep 30, 2015 at 8:41 AM, Dan Williams > wrote: >> Allow pmem, and other synchronous/bio-based block drivers, to fallback > > Just a bit curious, why not extend it for all(both synchronous

Re: [PATCH 1/2] block: generic request_queue reference counting

On Wed, Sep 30, 2015 at 8:41 AM, Dan Williams wrote: > Allow pmem, and other synchronous/bio-based block drivers, to fallback Just a bit curious, why not extend it for all(both synchronous and asynchrounous) bio-based drivers? As you mentioned in introductory message, all bio based drivers may

Re: [PATCH 1/2] block: generic request_queue reference counting

On Tue, Sep 29, 2015 at 08:41:31PM -0400, Dan Williams wrote: > Allow pmem, and other synchronous/bio-based block drivers, to fallback > on a per-cpu reference count managed by the core for tracking queue > live/dead state. > > The existing per-cpu reference count for the blk_mq case is promoted

Re: [PATCH 1/2] block: generic request_queue reference counting

On Wed, Sep 30, 2015 at 8:41 AM, Dan Williams wrote: > Allow pmem, and other synchronous/bio-based block drivers, to fallback Just a bit curious, why not extend it for all(both synchronous and asynchrounous) bio-based drivers? As you mentioned in introductory message,

Re: [PATCH 1/2] block: generic request_queue reference counting

On Tue, Sep 29, 2015 at 08:41:31PM -0400, Dan Williams wrote: > Allow pmem, and other synchronous/bio-based block drivers, to fallback > on a per-cpu reference count managed by the core for tracking queue > live/dead state. > > The existing per-cpu reference count for the blk_mq case is promoted

[PATCH 1/2] block: generic request_queue reference counting

Allow pmem, and other synchronous/bio-based block drivers, to fallback on a per-cpu reference count managed by the core for tracking queue live/dead state. The existing per-cpu reference count for the blk_mq case is promoted to be used in all block i/o scenarios. This involves initializing it by

[PATCH 1/2] block: generic request_queue reference counting

Allow pmem, and other synchronous/bio-based block drivers, to fallback on a per-cpu reference count managed by the core for tracking queue live/dead state. The existing per-cpu reference count for the blk_mq case is promoted to be used in all block i/o scenarios. This involves initializing it by