All of lore.kernel.org
 help / color / mirror / Atom feed
From: Benjamin Marzinski <bmarzins@redhat.com>
To: Martin Wilck <martin.wilck@suse.com>
Cc: Christophe Varoqui <christophe.varoqui@opensvc.com>,
	dm-devel@lists.linux.dev, Martin Wilck <mwilck@suse.com>
Subject: Re: [PATCH v2 0/8] multipath-tools fixes
Date: Tue, 3 Dec 2024 15:11:57 -0500	[thread overview]
Message-ID: <Z09mDdvigD5QqIDj@redhat.com> (raw)
In-Reply-To: <20241127230430.139639-1-mwilck@suse.com>

On Thu, Nov 28, 2024 at 12:04:22AM +0100, Martin Wilck wrote:

For all but patch 5/8:
Reviewed-by: Benjamin Marzinski <bmarzins@redhat.com>

> The first patch is an attempt to fix
> https://github.com/opensvc/multipath-tools/issues/105.
> 
> Patch 2-5 provide a re-implementation of pgcmp() without relying on
> pgp->id, as discussed during the review of v1 of this series.
> 
> Patch 6 and 7 change the way multipathd triggers uevents when
> it detects configuration changes, motivated by
> https://github.com/opensvc/multipath-tools/issues/105.
> 
> Up to now, multipathd had done this basically only if the "failed" state of a
> map had changed. This would miss cases where e.g. with find_multipaths strict,
> a path had been added to the WWIDs file and "multipathd reconfigure" had been
> run. In such cases, the paths would be added to the map, but udev would still
> not see them.  The same holds for changes of the blacklist.
> 
> With this set, when a map is created, or when map creation fails, multipathd
> checks whether udev already sees the correct properties for the device, and
> triggers an uevent otherwise.
> 
> When maps are to be removed because of configuration changes in
> coalesce_maps() (for example if a WWID has been removed from the WWIDs file),
> multipathd now also triggers uevents for the respective path devices. This
> is only done in coalesce_maps(), where current configuration changes are
> reflected. We can't do it after a "remove map" CLI command, for example,
> because that is by definition a temporary change. If we triggered an uevent
> for that, multipathd would recreate the map it just removed as soone as it
> receives the uevents.
> 
> Patch 8 is a minor log message change.
> 
> Martin Wilck (8):
>   libmultipath: fix handling of pp->pgindex
>   libmpathutil: change STATIC_BITFIELD to BITFIELD
>   libmpathutil: add cleanup_bitfield()
>   libmultipath: move pathcmp() to configure.c
>   libmultipath: re-implement pgcmp without the pathgroup id
>   libmultipath: trigger uevents upon map creation in domap()
>   multipathd: trigger uevents upon map removal in coalesce_maps()
>   multipathd: improve a log message in coalesce_maps()
> 
>  libmpathutil/libmpathutil.version |   1 +
>  libmpathutil/util.c               |   5 ++
>  libmpathutil/util.h               |  10 +--
>  libmultipath/configure.c          | 124 ++++++++++++++++++++----------
>  libmultipath/devmapper.c          |   9 +--
>  libmultipath/discovery.c          |   2 +-
>  libmultipath/dmparser.c           |   1 -
>  libmultipath/pgpolicies.c         |   6 ++
>  libmultipath/structs.c            |  31 +++-----
>  libmultipath/structs.h            |   3 -
>  libmultipath/structs_vec.c        |  15 ++++
>  multipathd/main.c                 |   6 +-
>  12 files changed, 135 insertions(+), 78 deletions(-)
> 
> -- 
> 2.47.0


      parent reply	other threads:[~2024-12-03 20:12 UTC|newest]

Thread overview: 13+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-11-27 23:04 [PATCH v2 0/8] multipath-tools fixes Martin Wilck
2024-11-27 23:04 ` [PATCH v2 1/8] libmultipath: fix handling of pp->pgindex Martin Wilck
2024-11-27 23:04 ` [PATCH v2 2/8] libmpathutil: change STATIC_BITFIELD to BITFIELD Martin Wilck
2024-11-27 23:04 ` [PATCH v2 3/8] libmpathutil: add cleanup_bitfield() Martin Wilck
2024-11-27 23:04 ` [PATCH v2 4/8] libmultipath: move pathcmp() to configure.c Martin Wilck
2024-11-27 23:04 ` [PATCH v2 5/8] libmultipath: re-implement pgcmp without the pathgroup id Martin Wilck
2024-12-03  0:17   ` Benjamin Marzinski
2024-12-03 12:40     ` Martin Wilck
2024-12-03 13:25       ` Martin Wilck
2024-11-27 23:04 ` [PATCH v2 6/8] libmultipath: trigger uevents upon map creation in domap() Martin Wilck
2024-11-27 23:04 ` [PATCH v2 7/8] multipathd: trigger uevents upon map removal in coalesce_maps() Martin Wilck
2024-11-27 23:04 ` [PATCH v2 8/8] multipathd: improve a log message " Martin Wilck
2024-12-03 20:11 ` Benjamin Marzinski [this message]

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=Z09mDdvigD5QqIDj@redhat.com \
    --to=bmarzins@redhat.com \
    --cc=christophe.varoqui@opensvc.com \
    --cc=dm-devel@lists.linux.dev \
    --cc=martin.wilck@suse.com \
    --cc=mwilck@suse.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.