Re: [PATCH v1 00/54] block: support multipage bvec

2017-01-17 Thread Ming Lei
On Tue, Jan 17, 2017 at 3:50 PM, Christoph Hellwig wrote: > On Tue, Jan 17, 2017 at 10:40:36AM +0800, Ming Lei wrote: >> IMO, the only one left is raid(1/5/10) which can be dealt with by the >> "NO_MP" flag. > > No, they can't in the long run. They need to handle the bvecs

Re: [PATCH v1 00/54] block: support multipage bvec

2017-01-17 Thread Ming Lei
On Tue, Jan 17, 2017 at 3:50 PM, Christoph Hellwig wrote: > On Tue, Jan 17, 2017 at 10:40:36AM +0800, Ming Lei wrote: >> IMO, the only one left is raid(1/5/10) which can be dealt with by the >> "NO_MP" flag. > > No, they can't in the long run. They need to handle the bvecs just > like everyone

Re: [PATCH v1 00/54] block: support multipage bvec

2017-01-16 Thread Christoph Hellwig
On Tue, Jan 17, 2017 at 10:40:36AM +0800, Ming Lei wrote: > IMO, the only one left is raid(1/5/10) which can be dealt with by the > "NO_MP" flag. No, they can't in the long run. They need to handle the bvecs just like everyone else.

Re: [PATCH v1 00/54] block: support multipage bvec

2017-01-16 Thread Christoph Hellwig
On Tue, Jan 17, 2017 at 10:40:36AM +0800, Ming Lei wrote: > IMO, the only one left is raid(1/5/10) which can be dealt with by the > "NO_MP" flag. No, they can't in the long run. They need to handle the bvecs just like everyone else.

Re: [PATCH v1 00/54] block: support multipage bvec

2017-01-16 Thread Ming Lei
On Mon, Jan 16, 2017 at 11:18 PM, Christoph Hellwig wrote: > On Mon, Jan 16, 2017 at 11:19:19AM +0800, Ming Lei wrote: >> Any comments on this version? > > We'll need to make sure all drivers can handle multi-page bvecs > before continuing any other work. Without that the

Re: [PATCH v1 00/54] block: support multipage bvec

2017-01-16 Thread Ming Lei
On Mon, Jan 16, 2017 at 11:18 PM, Christoph Hellwig wrote: > On Mon, Jan 16, 2017 at 11:19:19AM +0800, Ming Lei wrote: >> Any comments on this version? > > We'll need to make sure all drivers can handle multi-page bvecs > before continuing any other work. Without that the series is a no-go. >

Re: [PATCH v1 00/54] block: support multipage bvec

2017-01-16 Thread Christoph Hellwig
On Mon, Jan 16, 2017 at 11:19:19AM +0800, Ming Lei wrote: > Any comments on this version? We'll need to make sure all drivers can handle multi-page bvecs before continuing any other work. Without that the series is a no-go. Note that in general making a drivers capable of handling multipage

Re: [PATCH v1 00/54] block: support multipage bvec

2017-01-16 Thread Christoph Hellwig
On Mon, Jan 16, 2017 at 11:19:19AM +0800, Ming Lei wrote: > Any comments on this version? We'll need to make sure all drivers can handle multi-page bvecs before continuing any other work. Without that the series is a no-go. Note that in general making a drivers capable of handling multipage

Re: [PATCH v1 00/54] block: support multipage bvec

2017-01-15 Thread Ming Lei
Hi Guys, On Tue, Dec 27, 2016 at 11:55 PM, Ming Lei wrote: > Hi, > > This patchset brings multipage bvec into block layer. Basic > xfstests(-a auto) over virtio-blk/virtio-scsi have been run > and no regression is found, so it should be good enough > to show the approach

Re: [PATCH v1 00/54] block: support multipage bvec

2017-01-15 Thread Ming Lei
Hi Guys, On Tue, Dec 27, 2016 at 11:55 PM, Ming Lei wrote: > Hi, > > This patchset brings multipage bvec into block layer. Basic > xfstests(-a auto) over virtio-blk/virtio-scsi have been run > and no regression is found, so it should be good enough > to show the approach now, and any comments