Hi Steven, > The problem is localized to using pthread_spin_lock() on single cpu > system? Sounds like a priority inversion.
All right. > Which spin lock did you change to a mutex (or just change the whole > codebase)? I changed it in mutex in the next procedure. Step1) ./autogen.sh and ./configure. Step2) edit an include file and make next define line comment. /* Define to 1 if you have the `pthread_spin_lock' function. */ /* #define HAVE_PTHREAD_SPIN_LOCK 1 */ /* Define to 1 if you have the `pthread_spin_unlock' function. */ /* #define HAVE_PTHREAD_SPIN_UNLOCK 1 */ Step3) make install. If there is an mistake in my procedure, please support a right procedure. > I'll look into it this week. Thanks. Best Regards, Hideo Yamauchi. --- Steven Dake <[email protected]> wrote: > On 06/11/2010 12:14 AM, [email protected] wrote: > > Hi Steven, > > > > I report a result. > > When I started in corosync -p, I succeed in a stop. > > > > When I do not start with a p option, VM succeed in a stop when corosync > > operate it with a > mutex lock. > > > > GThe problem is localized to using pthread_spin_lock() on single cpu > system? Sounds like a priority inversion. > > Which spin lock did you change to a mutex (or just change the whole > codebase)? > > I'll look into it this week. > > Thanks > -steve > > > > > Best Regards, > > Hideo Yamauchi. > > > > --- [email protected] wrote: > > > >> Hi Steven, > >> > >> Thank you for comment. > >> > >> I try corosync -p. > >> > >> Best Regards, > >> Hideo Yamauchi. > >> > >> --- Steven Dake<[email protected]> wrote: > >> > >>> On 06/10/2010 01:58 AM, [email protected] wrote: > >>>> Hi All, > >>>> > >>>> I tried 1.2.4. > >>>> I do not use Pacemaker. > >>>> > >>>> * 4 node VM.(RHEL5.4 x86) > >>>> * VM on Esxi. > >>>> * debug : on > >>>> * rrp_mode : none > >>>> > >>>> However, at the time of debug:on, corosync fails in a stop. > >>>> Some VM fails in a stop and are hung up. > >>>> And it is used 100% for the CPU. > >>>> > >>>> We collect a little more detailed information and are going to report it. > >>>> > >>>> Best Regards, > >>>> Hideo Yamauchi. > >>>> > >>>> > >>> > >>> Thanks for the bug report. May be vm related or 1 cpu issue that isn't > >>> caught by our automatic tests. > >>> > >>> give it a run with corosync -p. > >>> > >>> Regards > >>> -steve > >>> > >>>> --- Steven Dake<[email protected]> wrote: > >>>> > >>>>> I am pleased to announce the release of corosync 1.2.4 for immediate > >>>>> download from our website. > >>>>> > >>>>> This version has the following changes: > >>>>> * Fixes defects in logsys which are crashing pacemaker installations. > >>>>> * Adds man pages for all binaries installed with the "make install" > >>>>> command. > >>>>> * Fixes several defects found in high packet loss field environments. > >>>>> > >>>>> The software is available via our website at: > >>>>> http://www.corosync.org > >>>>> _______________________________________________ > >>>>> Openais mailing list > >>>>> [email protected] > >>>>> https://lists.linux-foundation.org/mailman/listinfo/openais > >>>>> > >>>> > >>>> _______________________________________________ > >>>> Openais mailing list > >>>> [email protected] > >>>> https://lists.linux-foundation.org/mailman/listinfo/openais > >>> > >>> > >> > >> _______________________________________________ > >> Openais mailing list > >> [email protected] > >> https://lists.linux-foundation.org/mailman/listinfo/openais > >> > > > > _______________________________________________ Openais mailing list [email protected] https://lists.linux-foundation.org/mailman/listinfo/openais
