Re: io_pgetevents & aio fsync V4

2018-05-10 Thread Al Viro
On Wed, May 02, 2018 at 11:14:41PM +0200, Christoph Hellwig wrote: > Hi all, > > this patch adds workqueue based fsync offload. Version of this > patch have been floating around for a couple years, but we now > have a user with seastar used by ScyllaDB (who sponsored this > work) that really want

Re: io_pgetevents & aio fsync V2

2018-04-06 Thread Jeff Moyer
Christoph Hellwig writes: > On Fri, Apr 06, 2018 at 04:16:30AM +0100, Al Viro wrote: >> BTW, this is only tangentially related, but... does *anything* call >> io_submit() for huge amounts of iocb? I don't know. If an application did that, as many I/Os as could fit into the ring buffer would be

Re: io_pgetevents & aio fsync V2

2018-04-05 Thread Christoph Hellwig
On Fri, Apr 06, 2018 at 04:16:30AM +0100, Al Viro wrote: > BTW, this is only tangentially related, but... does *anything* call > io_submit() for huge amounts of iocb? Check in do_io_submit() is > insane - "no more than MAX_LONG total of _pointers_". Compat variant > goes for "no more than a page

Re: io_pgetevents & aio fsync V2

2018-04-05 Thread Al Viro
On Wed, Mar 28, 2018 at 09:26:33AM +0200, Christoph Hellwig wrote: > Hi all, > > this patch adds workqueue based fsync offload. Version of this > patch have been floating around for a couple years, but we now > have a user with seastar used by ScyllaDB (who sponsored this > work) that really want

Re: io_pgetevents & aio fsync

2018-03-22 Thread Al Viro
On Thu, Mar 22, 2018 at 04:36:05PM +, Al Viro wrote: > On Wed, Mar 21, 2018 at 08:32:23AM +0100, Christoph Hellwig wrote: > > Hi all, > > > > this patch adds workqueue based fsync offload. Version of this > > patch have been floating around for a couple years, but we now > > have a user with

Re: io_pgetevents & aio fsync

2018-03-22 Thread Al Viro
On Wed, Mar 21, 2018 at 08:32:23AM +0100, Christoph Hellwig wrote: > Hi all, > > this patch adds workqueue based fsync offload. Version of this > patch have been floating around for a couple years, but we now > have a user with seastar used by ScyllaDB (who sponsored this > work) that really want