[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20230421154058.GH360881@frogsfrogsfrogs>
Date: Fri, 21 Apr 2023 08:40:58 -0700
From: "Darrick J. Wong" <djwong@...nel.org>
To: Ritesh Harjani <ritesh.list@...il.com>
Cc: Jan Kara <jack@...e.cz>, linux-fsdevel@...r.kernel.org,
linux-ext4@...r.kernel.org, Christoph Hellwig <hch@...radead.org>,
Ojaswin Mujoo <ojaswin@...ux.ibm.com>,
Disha Goel <disgoel@...ux.ibm.com>, Ted Tso <tytso@....edu>
Subject: Re: [PATCHv6 0/9] ext2: DIO to use iomap
On Fri, Apr 21, 2023 at 05:35:47PM +0530, Ritesh Harjani wrote:
> Jan Kara <jack@...e.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
Powered by blists - more mailing lists