Linux Container Development
 help / color / mirror / Atom feed
From: Nadia.Derbey-6ktuUTfB/bM@public.gmane.org
To: containers-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA@public.gmane.org
Subject: [RFC PATCH 0/4] IPC/sem - allow saving/restoring a process semundo_list
Date: Fri, 20 Jun 2008 13:48:38 +0200	[thread overview]
Message-ID: <20080620114838.779146000@bull.net> (raw)

Hi,

A couple of months ago, Pierre Peiffer has submitted a patch series to enable
checkpointing / restarting ipcs.
4 of these patches were related to semaphores:

https://lists.linux-foundation.org/pipermail/containers/2008-January/thread.html#9756
https://lists.linux-foundation.org/pipermail/containers/2008-January/thread.html#9757
https://lists.linux-foundation.org/pipermail/containers/2008-January/thread.html#9758
https://lists.linux-foundation.org/pipermail/containers/2008-January/thread.html#9759

They introduced a new procfs file: /proc/<pid>/semundo to read and write the
semaphores undo values for a given process.

These patches are widely used in the -lxc development tree cryo code is based
upon.
(more information about lxc can be found at http://lxc.sourceforge.net/ and
the development tree can be found there too - under the patches link).

Manfred Spraul, on his side, has rewritten an important part of the
semaphores code. See:

1. http://lkml.org/lkml/2008/5/24/92
2. http://lkml.org/lkml/2008/5/24/93
3. http://lkml.org/lkml/2008/5/24/90
4. http://lkml.org/lkml/2008/5/24/91

Mainly patches 1 and 4 made Pierre's patches unappliable.
 . patch 1 changed the semundo_list proc_list field into a linked list.
 . patch 2
    . reversed the locking order of the sem_undo_list lock and the semaphore
      lock.
    . converted the sem_undo structure to use rcu.

Since 2.6.26-rc5-mm3 is now the new target for the -lxc development tree, I've
ported them and taking this opportunity to resubmit them to the containers
list.

I have in mind a simpler solution, which I think I'll propose next week:
instead of writing into the /proc/<pid>/semundo of a third party process,
only allow the write operation to be done for <current>.

Kathy, can you please pull in these patches?


Regards,
Nadia


--

             reply	other threads:[~2008-06-20 11:48 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2008-06-20 11:48 Nadia.Derbey-6ktuUTfB/bM [this message]
2008-06-20 11:48 ` [RFC PATCH 1/4] IPC/sem: use RCU to free the sem_undo_list Nadia.Derbey-6ktuUTfB/bM
2008-06-20 11:48 ` [RFC PATCH 2/4] IPC/sem: per <pid> semundo file in procfs Nadia.Derbey-6ktuUTfB/bM
2008-06-20 11:48 ` [RFC PATCH 3/4] IPC/sem: prepare semundo code to work on a third party task Nadia.Derbey-6ktuUTfB/bM
2008-06-20 11:48 ` [RFC PATCH 4/4] IPC/sem: add the write() operation to the semundo file in procfs Nadia.Derbey-6ktuUTfB/bM

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=20080620114838.779146000@bull.net \
    --to=nadia.derbey-6ktuutfb/bm@public.gmane.org \
    --cc=containers-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA@public.gmane.org \
    /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