Re: Odd locking pattern introduced as part of "nowait aio support"

2019-09-11 Thread Goldwyn Rodrigues
On 2:39 11/09, Andres Freund wrote: > > Ok. Goldwyn, do you want to write a patch, or do you want me to write > one up? I'll post one shortly. Thanks for bringing this up. -- Goldwyn

Re: Odd locking pattern introduced as part of "nowait aio support"

2019-09-11 Thread Ritesh Harjani
Hi, On 9/11/19 3:09 PM, Andres Freund wrote: Hi, On 2019-09-11 14:04:20 +1000, Dave Chinner wrote: On Tue, Sep 10, 2019 at 03:33:27PM -0700, Andres Freund wrote: Hi, Especially with buffered io it's fairly easy to hit contention on the inode lock, during writes. With something like io_uring,

Re: Odd locking pattern introduced as part of "nowait aio support"

2019-09-11 Thread Christoph Hellwig
On Wed, Sep 11, 2019 at 02:39:26AM -0700, Andres Freund wrote: > I do really wish buffered NOWAIT was supported... Send a patch..

Re: Odd locking pattern introduced as part of "nowait aio support"

2019-09-11 Thread Andres Freund
Hi, On 2019-09-11 14:04:20 +1000, Dave Chinner wrote: > On Tue, Sep 10, 2019 at 03:33:27PM -0700, Andres Freund wrote: > > Hi, > > > > Especially with buffered io it's fairly easy to hit contention on the > > inode lock, during writes. With something like io_uring, it's even > > easier, because i