Re: [PATCH 12/12] io_uring: support true async buffered reads, if file provides it

2020-05-26 Thread Jens Axboe
On 5/26/20 1:44 AM, Pavel Begunkov wrote: > On 25/05/2020 22:59, Jens Axboe wrote: >> On 5/25/20 1:29 AM, Pavel Begunkov wrote: >>> On 23/05/2020 21:57, Jens Axboe wrote: If the file is flagged with FMODE_BUF_RASYNC, then we don't have to punt the buffered read to an io-wq worker.

Re: [PATCH 12/12] io_uring: support true async buffered reads, if file provides it

2020-05-26 Thread Jens Axboe
On 5/26/20 1:38 AM, Pavel Begunkov wrote: > On 23/05/2020 21:57, Jens Axboe wrote: >> If the file is flagged with FMODE_BUF_RASYNC, then we don't have to punt >> the buffered read to an io-wq worker. Instead we can rely on page >> unlocking callbacks to support retry based async IO. This is a lot

Re: [PATCH 12/12] io_uring: support true async buffered reads, if file provides it

2020-05-26 Thread Pavel Begunkov
On 25/05/2020 22:59, Jens Axboe wrote: > On 5/25/20 1:29 AM, Pavel Begunkov wrote: >> On 23/05/2020 21:57, Jens Axboe wrote: >>> If the file is flagged with FMODE_BUF_RASYNC, then we don't have to punt >>> the buffered read to an io-wq worker. Instead we can rely on page >>> unlocking callbacks to

Re: [PATCH 12/12] io_uring: support true async buffered reads, if file provides it

2020-05-26 Thread Pavel Begunkov
On 23/05/2020 21:57, Jens Axboe wrote: > If the file is flagged with FMODE_BUF_RASYNC, then we don't have to punt > the buffered read to an io-wq worker. Instead we can rely on page > unlocking callbacks to support retry based async IO. This is a lot more > efficient than doing async thread

Re: [PATCH 12/12] io_uring: support true async buffered reads, if file provides it

2020-05-25 Thread Jens Axboe
On 5/25/20 1:29 AM, Pavel Begunkov wrote: > On 23/05/2020 21:57, Jens Axboe wrote: >> If the file is flagged with FMODE_BUF_RASYNC, then we don't have to punt >> the buffered read to an io-wq worker. Instead we can rely on page >> unlocking callbacks to support retry based async IO. This is a lot

Re: [PATCH 12/12] io_uring: support true async buffered reads, if file provides it

2020-05-25 Thread Pavel Begunkov
On 23/05/2020 21:57, Jens Axboe wrote: > If the file is flagged with FMODE_BUF_RASYNC, then we don't have to punt > the buffered read to an io-wq worker. Instead we can rely on page > unlocking callbacks to support retry based async IO. This is a lot more > efficient than doing async thread