[Search for users] [Overall Top Noters] [List of all Conferences] [Download this site]

Conference smurf::buildhelp

Title:USG buildhelp questions/answers
Moderator:SMURF::FILTER
Created:Mon Apr 26 1993
Last Modified:Mon Jan 20 1997
Last Successful Update:Fri Jun 06 1997
Number of topics:2763
Total number of notes:5802

177.0. "Help - locking collision Maint vs Minor pools" by SMURF::FILTER (Automatic Posting Software - mail to flume::puck) Fri Jun 25 1993 12:36

Date Of Receipt: 	25-JUN-1993 11:18:53.91
From: 	LOCORE::"[email protected]"
To: 	"[email protected]"@SYSV.ENET.dec.com
CC: 	[email protected]
Subj: 	Help - locking collision Maint vs Minor pools

Help,

	Rich Larsen, the ODE giant for UNX, is out today.  A file,
src/kernel/procfs/procfs_vnops.c, has a lock from the Maint pool 1.1.26.6
stuck, and its preventing me from submitting a fix to the Sterling pool,
rev 1.1.31.?.


	Please call Dave Scoda at DTN 462-6084.

	Oh yeah, I've already tried bcs -o -u to unlock it, but it won't.

Thanks,
Dave

T.RTitleUserPersonal
Name
DateLines
177.1Re: Help - locking collision Maint vs Minor poolsSMURF::FILTERAutomatic Posting Software - mail to flume::puckFri Jun 25 1993 13:4173
Date Of Receipt: 	25-JUN-1993 11:44:27.12
	1993/06/17 17:42:54;  author: David_Scoda;  state: Exp;  lines: +18 -12
From: 	FLUME::"[email protected]" "Madeline Barcia-Asmus USG"
To: 	[email protected]
CC: 	[email protected], [email protected]
Subj: 	Re: Help - locking collision Maint vs Minor pools
I found the following: 	
locks: 	strict
David_Scoda: 	1.1.26.6
symbolic names: 	
AGOSMINOR: 	1.1.26
revision 1.1.26.6       locked by: 	David_Scoda;

        fix locking for PIOCRUN with PRCSIG set; add signal bounds check
        [1993/06/17  17:43:17  David_Scoda]


# grep procfs_vnops.c SNAPSHOT
./kernel/procfs/procfs_vnops.c,v        1.1.26.6


# more ../src/kernel/procfs/procfs_vnops.c
/*
 * @DEC_COPYRIGHT@
 */
/*
 * HISTORY
 * $Log: procfs_vnops.c,v $
 * Revision 1.1.26.6  1993/06/17  17:42:54  David_Scoda
 *      fix locking for PIOCRUN with PRCSIG set; add signal bounds check
 *      [1993/06/17  17:43:17  David_Scoda]
 *

Looks like  somehow bsubmit might have left the public branch locked .
Not sure. But since you hold the lock you should be able to  
       bcs -u1.1.26.6  procfs_vnops.c


and that will unlock the AGOSMINOR  public branch.

I don't think you will be submitting  1.1.31 since that one is owned by Michael_Fairbrother.
Yours is  1.1.30.3  .

----------------------------
revision 1.1.31.2
date: 1993/06/25 14:43:11;  author: Michael_Fairbrother;  state: Exp;  lines: +47
16 -0
*** Initial Branch Revision ***
----------------------------
revision 1.1.30.3
date: 1993/06/25 14:40:55;  author: David_Scoda;  state: Exp;  lines: +4 -1
        bci after bmerge
----------------------------

Let me know if you still have difficulties. 

-Madeline

 >>Help,
 >>
 >>	Rich Larsen, the ODE giant for UNX, is out today.  A file,
 >>src/kernel/procfs/procfs_vnops.c, has a lock from the Maint pool 1.1.26.6
 >>stuck, and its preventing me from submitting a fix to the Sterling pool,
 >>rev 1.1.31.?.
 >>
 >>
 >>	Please call Dave Scoda at DTN 462-6084.
 >>
 >>	Oh yeah, I've already tried bcs -o -u to unlock it, but it won't.
 >>
 >>Thanks,
 >>Dave

177.2Re: Help - locking collision Maint vs Minor poolsSMURF::FILTERAutomatic Posting Software - mail to flume::puckFri Jun 25 1993 13:4524
Date Of Receipt: 	25-JUN-1993 12:21:48.09
From: 	WASTED::jmf "Joshua M. Friedman ULTRIX SDE  25-Jun-1993 1222"
To: 	""Madeline T. Asmus"" <mad@wasted:zko.dec>
CC: 	unxa::scoda, odehelp@wasted:zko.dec, reng@wasted:zko.dec,
	jmf@wasted:zko.dec
Subj: 	Re: Help - locking collision Maint vs Minor pools

Dave's problem has been resolved.  The bsubmit happened to collide at
the rcs level with another user doing a bco, so the rcs file was busy
when bsubmit needed it.  

Bsubmit just stopped and issued a message to do a resub.  Dave wasn't
sure what was going on, and, finding the submit branch locked, unlocked
it.  

I had Dave just re-lock the submit branch in his name the way it had
been, (bcs -lAGOSMINOR) and do the bsubmit -resub, and it worked fine.

I also suggested to Dave he use gen-ptt to report the bug/suggestion
that bsubmit be able to do some waiting and retrying on an rcs lock
condition before just giving up.

-josh