From: Mike Snitzer <snitzer@redhat.com>
To: Bart Van Assche <bart.vanassche@sandisk.com>
Cc: device-mapper development <dm-devel@redhat.com>
Subject: Re: [PATCH 7/7] dm-mpath: Fix a race condition in __multipath_map()
Date: Tue, 15 Nov 2016 20:50:21 -0500 [thread overview]
Message-ID: <20161116015021.GA19469@redhat.com> (raw)
In-Reply-To: <4411f9ca-da76-d222-a2ee-3edc59c82266@sandisk.com>
On Tue, Nov 15 2016 at 8:08pm -0500,
Bart Van Assche <bart.vanassche@sandisk.com> wrote:
> On 11/15/2016 05:01 PM, Mike Snitzer wrote:
> >On Tue, Nov 15 2016 at 7:40pm -0500,
> >Bart Van Assche <bart.vanassche@sandisk.com> wrote:
> >>Are you aware that dm_table_determine_type() sets "all_blk_mq" to
> >>false if there are no paths, even if the dm device is in blk-mq
> >>mode?
> >
> >That shouldn't matter. Once the type is established, it is used to
> >initialize the DM device's request_queue, the type cannot change across
> >different table loads.
>
> For a single queue dm device, what prevents a user from removing all
> single queue paths and to add one or more blk-mq paths? I think this
> will cause dm_table_determine_type() to change the table type.
A new table is created for every table load (any time a multipath device
is loaded into the kernel). The DM core disallows a table with a
different type to load. It will be rejected within an error.
See dm-ioctl.c:table_load()
next prev parent reply other threads:[~2016-11-16 1:50 UTC|newest]
Thread overview: 39+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-11-15 23:31 [PATCH 0/7] dm-mpath: Fix a race condition in the blk-mq path Bart Van Assche
2016-11-15 23:32 ` [PATCH 1/7] dm: Fix a (theoretical?) race condition in rq_completed() Bart Van Assche
2016-11-16 0:46 ` Mike Snitzer
2016-11-15 23:33 ` [PATCH 2/7] dm: Simplify dm_table_determine_type() Bart Van Assche
2016-11-16 14:54 ` Mike Snitzer
2016-11-16 20:14 ` Bart Van Assche
2016-11-16 21:11 ` Mike Snitzer
2016-11-16 21:53 ` Bart Van Assche
2016-11-16 23:09 ` Mike Snitzer
2016-11-15 23:33 ` [PATCH 3/7] dm-mpath: Document a locking assumption Bart Van Assche
2016-11-18 0:07 ` Mike Snitzer
2016-11-15 23:34 ` [PATCH 4/7] dm-mpath: Change return type of pg_init_all_paths() from int into void Bart Van Assche
2016-11-15 23:34 ` [PATCH 5/7] dm-mpath: Do not touch *__clone if request allocation fails Bart Van Assche
2016-11-15 23:34 ` [PATCH 6/7] dm-mpath: Avoid code duplication in __multipath_map() Bart Van Assche
2016-11-16 0:39 ` Mike Snitzer
2016-11-15 23:35 ` [PATCH 7/7] dm-mpath: Fix a race condition " Bart Van Assche
2016-11-16 0:37 ` Mike Snitzer
2016-11-16 0:40 ` Bart Van Assche
2016-11-16 1:01 ` Mike Snitzer
2016-11-16 1:08 ` Bart Van Assche
2016-11-16 1:50 ` Mike Snitzer [this message]
2016-11-21 21:44 ` Bart Van Assche
2016-11-21 23:43 ` Mike Snitzer
2016-11-21 23:57 ` Bart Van Assche
2016-11-22 0:34 ` Mike Snitzer
2016-11-22 23:47 ` Bart Van Assche
2016-11-23 0:48 ` Mike Snitzer
2016-11-23 3:16 ` Mike Snitzer
2016-11-23 18:28 ` Bart Van Assche
2016-11-23 18:50 ` Mike Snitzer
2016-11-16 0:47 ` [PATCH 0/7] dm-mpath: Fix a race condition in the blk-mq path Mike Snitzer
2016-11-16 0:57 ` Bart Van Assche
2016-11-16 1:08 ` Mike Snitzer
2016-11-16 1:10 ` Bart Van Assche
2016-11-16 1:53 ` Mike Snitzer
2016-11-16 7:39 ` Hannes Reinecke
2016-11-16 14:56 ` Mike Snitzer
2016-11-16 18:22 ` Bart Van Assche
2016-11-16 19:32 ` Mike Snitzer
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20161116015021.GA19469@redhat.com \
--to=snitzer@redhat.com \
--cc=bart.vanassche@sandisk.com \
--cc=dm-devel@redhat.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).