From: zkabelac@sourceware.org <zkabelac@sourceware.org>
To: lvm-devel@redhat.com
Subject: LVM2 ./WHATS_NEW daemons/clvmd/clvmd.c
Date: 11 Oct 2011 09:26:04 -0000 [thread overview]
Message-ID: <20111011092604.10366.qmail@sourceware.org> (raw)
CVSROOT: /cvs/lvm2
Module name: LVM2
Changes by: zkabelac at sourceware.org 2011-10-11 09:26:04
Modified files:
. : WHATS_NEW
daemons/clvmd : clvmd.c
Log message:
Use barrier instead of mutex
Barrier is supposed to be used in situation like this
and replace tricky mutex usage, where mutex has been unlocked
by a different thread than the locking thread.
Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/LVM2/WHATS_NEW.diff?cvsroot=lvm2&r1=1.2158&r2=1.2159
http://sourceware.org/cgi-bin/cvsweb.cgi/LVM2/daemons/clvmd/clvmd.c.diff?cvsroot=lvm2&r1=1.112&r2=1.113
--- LVM2/WHATS_NEW 2011/10/11 09:20:17 1.2158
+++ LVM2/WHATS_NEW 2011/10/11 09:26:04 1.2159
@@ -1,5 +1,6 @@
Version 2.02.89 -
==================================
+ Use pthread_barrier to synchronize clvmd threads at startup.
Limit clvmd's thread size to 128KiB.
Reduce default preallocated stack size to 64KiB.
Add check for access through NULL pointer when refresh_filter() fails.
--- LVM2/daemons/clvmd/clvmd.c 2011/10/11 09:23:48 1.112
+++ LVM2/daemons/clvmd/clvmd.c 2011/10/11 09:26:04 1.113
@@ -89,7 +89,7 @@
static pthread_attr_t stack_attr;
static pthread_mutex_t lvm_thread_mutex;
static pthread_cond_t lvm_thread_cond;
-static pthread_mutex_t lvm_start_mutex;
+static pthread_barrier_t lvm_start_barrier;
static struct dm_list lvm_cmd_head;
static volatile sig_atomic_t quit = 0;
static volatile sig_atomic_t reread_config = 0;
@@ -505,7 +505,7 @@
}
pthread_mutex_init(&lvm_thread_mutex, NULL);
pthread_cond_init(&lvm_thread_cond, NULL);
- pthread_mutex_init(&lvm_start_mutex, NULL);
+ pthread_barrier_init(&lvm_start_barrier, NULL, 2);
init_lvhash();
/* Start the cluster interface */
@@ -584,9 +584,11 @@
DEBUGLOG("starting LVM thread\n");
/* Don't let anyone else to do work until we are started */
- pthread_mutex_lock(&lvm_start_mutex);
pthread_create(&lvm_thread, &stack_attr, lvm_thread_fn, &lvm_params);
+ /* Don't start until the LVM thread is ready */
+ pthread_barrier_wait(&lvm_start_barrier);
+
/* Tell the rest of the cluster our version number */
if (clops->cluster_init_completed)
clops->cluster_init_completed();
@@ -1633,11 +1635,6 @@
DEBUGLOG("in sub thread: client = %p\n", client);
pthread_mutex_lock(&client->bits.localsock.mutex);
- /* Don't start until the LVM thread is ready */
- pthread_mutex_lock(&lvm_start_mutex);
- pthread_mutex_unlock(&lvm_start_mutex);
- DEBUGLOG("Sub thread ready for work.\n");
-
/* Ignore SIGUSR1 (handled by master process) but enable
SIGUSR2 (kills subthreads) */
sigemptyset(&ss);
@@ -2004,7 +2001,8 @@
init_clvm(lvm_params->excl_uuid);
/* Allow others to get moving */
- pthread_mutex_unlock(&lvm_start_mutex);
+ pthread_barrier_wait(&lvm_start_barrier);
+ DEBUGLOG("Sub thread ready for work.\n");
/* Now wait for some actual work */
while (!quit) {
next reply other threads:[~2011-10-11 9:26 UTC|newest]
Thread overview: 32+ messages / expand[flat|nested] mbox.gz Atom feed top
2011-10-11 9:26 zkabelac [this message]
-- strict thread matches above, loose matches on Subject: below --
2012-02-28 11:06 LVM2 ./WHATS_NEW daemons/clvmd/clvmd.c zkabelac
2012-02-28 9:58 zkabelac
2012-02-28 9:53 zkabelac
2012-02-27 11:26 zkabelac
2012-02-27 9:58 zkabelac
2011-10-11 9:54 zkabelac
2011-09-16 14:40 mbroz
2011-08-11 12:57 mbroz
2011-06-28 13:42 zkabelac
2011-03-30 12:36 zkabelac
2011-03-08 22:48 zkabelac
2011-03-08 13:27 mbroz
2010-12-13 10:49 prajnoha
2010-12-01 12:41 zkabelac
2010-12-01 10:46 zkabelac
2010-04-06 15:29 ccaulfield
2010-02-02 8:54 ccaulfield
2009-10-12 8:33 ccaulfield
2009-09-01 9:48 ccaulfield
2009-08-13 10:39 ccaulfield
2009-03-24 11:49 ccaulfield
2008-11-21 13:48 ccaulfield
2008-06-13 7:44 ccaulfield
2008-05-09 9:59 ccaulfield
2008-03-28 12:58 ccaulfield
2008-03-17 9:38 ccaulfield
2008-02-15 14:12 meyering
2007-11-15 10:16 pcaulfield
2007-03-29 13:59 pcaulfield
2006-12-11 13:48 pcaulfield
2006-11-30 9:44 pcaulfield
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=20111011092604.10366.qmail@sourceware.org \
--to=zkabelac@sourceware.org \
--cc=lvm-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 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.