[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <Z9KOItsOJykGzI-F@infradead.org>
Date: Thu, 13 Mar 2025 00:49:54 -0700
From: Christoph Hellwig <hch@...radead.org>
To: John Garry <john.g.garry@...cle.com>
Cc: Christoph Hellwig <hch@...radead.org>,
Dave Chinner <david@...morbit.com>, brauner@...nel.org,
djwong@...nel.org, cem@...nel.org, linux-xfs@...r.kernel.org,
linux-fsdevel@...r.kernel.org, linux-kernel@...r.kernel.org,
ojaswin@...ux.ibm.com, ritesh.list@...il.com,
martin.petersen@...cle.com
Subject: Re: [PATCH RFC v5 10/10] iomap: Rename ATOMIC flags again
On Thu, Mar 13, 2025 at 07:41:11AM +0000, John Garry wrote:
> So how about this (I would re-add the write through comment):
This looks roughly sane. You'd probably want to turn the
iomap_dio_bio_opflags removal into a prep path, though.
> - blk_opf_t opflags = REQ_SYNC | REQ_IDLE;
This good lost and should move to the bio_opf declaration now.
> + (bdev_fua(iomap->bdev) || !bdev_write_cache(iomap->bdev))) {
> + bio_opf |= REQ_FUA; //reads as well?
REQ_FUA is not defined for reads in Linux Some of the storage standards
define it for reads, but the semantics are pretty nonsensical.
Powered by blists - more mailing lists