Re: [PATCHv6 0/9] ext2: DIO to use iomap

["Darrick J. Wong" <djwong@kernel.org>]

On Fri, Apr 21, 2023 at 05:35:47PM +0530, Ritesh Harjani wrote:
> Jan Kara <jack@suse.cz> writes:
> 
> > Hello Ritesh,
> >
> > On Fri 21-04-23 15:16:10, Ritesh Harjani (IBM) wrote:
> >> Hello All,
> >>
> >> Please find the series which rewrites ext2 direct-io path to use modern
> >> iomap interface.
> >
> > The patches now all look good to me. I'd like to discuss a bit how to merge
> 
> Thanks Jan,
> 
> 
> > them. The series has an ext4 cleanup (patch 3) and three iomap patches
> 
> Also Patch-3 is on top of ext4 journalled data patch series of yours,
> otheriwse we might see a minor merge conflict.
> 
> https://lore.kernel.org/all/20230329154950.19720-6-jack@suse.cz/
> 
> > (patches 6, 8 and 9). Darrick, do you want to take the iomap patches through
> > your tree?

Hmm.  I could do that for 6.4 since the first one should be trivially
verifiable and so far Linus hasn't objected to patches that add
tracepoints being thrown into for-next right at the start of the merge
window.

--D

> > The only dependency is that patch 7 for ext2 is dependent on definitions
> > from patch 6
> 
> That's right. Patch 6 defines TRACE_IOCB_STRINGS definition which both
> ext2 and iomap tracepoints depend upon.
> 
> > so I'd have to pull your branch into my tree. Or I can take
> > all the iomap patches through my tree but for that it would be nice to have
> > Darrick's acks.
> >
> > I can take the ext4 patch through my tree unless Ted objects.
> 
> Sure, we might have to merge with Ted's ext4 tree as well to avoid the
> merge conflict I mentioned above.
> 
> >
> > I guess I won't rush this for the coming merge window (unless Linus decides
> > to do rc8) but once we settle on the merge strategy I'll push out some
> 
> Ok.
> 
> > branch on which we can base further ext2 iomap conversion work.
> >
> 
> Sure, will this branch also gets reflected in linux-next for wider testing?
> 
> -ritesh