This commit was manufactured by cvs2svn to create tag

'slurm-0-3-14-1'.
chaos · Jun 10, 2005 · 5a49bcb · 5a49bcb
1 parent 8e60bcf
commit 5a49bcb
Show file tree

Hide file tree

Showing 40 changed files with 522 additions and 248 deletions.
diff --git a/META b/META
@@ -9,8 +9,8 @@
   Name:		slurm
   Major:	0
   Minor:	3
-  Micro:        7
+  Micro:        14
-  Version:	0.3.7
+  Version:	0.3.14
   Release:	1
   API_CURRENT:	5	
   API_AGE:	4

diff --git a/NEWS b/NEWS
@@ -1,6 +1,53 @@
 This file describes changes in recent versions of SLURM. It primarily
 documents those changes that are of interest to users and admins. 
 
+* Changes in SLURM 0.3.14
+=========================
+ -- Fix bug in bitfmt2int() which can go off allocated memory.
+
+* Changes in SLURM 0.3.13
+=========================
+ -- "sinfo -i" no longer aborts when nodes have associated features
+
+* Changes in SLURM 0.3.12
+=========================
+ -- Always report job scheduling details (so prolog/epilog can find 
+    constraints).
+
+* Changes in SLURM 0.3.11
+=========================
+ -- Fixes for reported problems:
+   - slurm/538: user tasks block writing to stdio
+ -- Added srun option --disable-status,-X to disable srun status feature
+    and instead forward SIGINT immediately to job upon receipt of Ctrl-C.
+ -- Fix for bogus slurmd error message "Unable to put task N into pgrp..."
+ -- Fix case where slurmd may erroneously detect shared memory entry
+    as "stale" and delete entry for unkillable or slow-to-exit job.
+ -- (qsnet) Fix for running slurmd on node without and elan3 adapter.
+
+* Changes in SLURM 0.3.10
+=========================
+ -- Move startup script from "/etc/rc.d/init.d/slurm" to "/etc/init.d/slurm"
+ -- Avoid creating orphaned process group when starting job processes.
+ -- Remove calls in auth/munge plugin deprecated by munge-0.4.
+
+* Changes in SLURM 0.3.9
+========================
+ -- Fixes for reported problems:
+   - slurm/512: Let job steps run on DRAINING nodes
+   - slurm/513: Gracefully deal with UIDs missing from passwd file
+ -- Create new allocation as needed for debugger in case old allocation
+    has been purged
+ -- Fix bug in scheduling jobs when a processor count is specified 
+    and FastSchedule=0 and the cluster is heterogeneous.
+ -- Fix srun bug when --input, --output and --error are all "none"
+ -- Allow single task id to be selected with --input, --output, and --error.
+ -- Create shared memory segment for Elan statistics when using the 
+    switch/elan plugin.
+
+* Changes in SLURM 0.3.8
+========================
+ -- More fixes necessary for TotalView.
 
 * Changes in SLURM 0.3.7
 ========================

diff --git a/doc/man/man1/scontrol.1 b/doc/man/man1/scontrol.1
@@ -171,6 +171,7 @@ Possible values are"YES" and "NO".
 Set the job's required features on nodes specified value. Multiple values 
 may be comma separated if all features are required (AND operation) or 
 separated by "|" if any of the specified features are required (OR operation).
+Value may be cleared with blank data value, "Features=".
 .TP
 \fIJobId\fP=<id>
 Identify the job to be updated. This specification is required.
@@ -196,6 +197,7 @@ Set the job's priority to the specified value.
 \fIReqNodeList\fP=<nodes>
 Set the job's list of required node. Multiple node names may be specified using 
 simple node range expressions (e.g. "lx[10-20]"). 
+Value may be cleared with blank data value, "ReqNodeList=".
 .TP
 \fIReqNodes\fP=<count>
 Set the job's count of required nodes to the specified value.
@@ -232,8 +234,9 @@ changing its underlying state.
 \fBSPECIFICATIONS FOR UPDATE AND DELETE COMMANDS, PARTITIONS\fR
 .TP
 \fIAllowGroups\fP=<name>
-Identify the user groups which may use this partition. Multiple groups 
+Identify the user groups which may use this partition. 
-may be specified in a comma separated list. 
+Multiple groups may be specified in a comma separated list.
+To permit all groups to use the partition specify "AllowGroups=ALL" 
 .TP
 \fIDefault\fP=<yes|no>
 Specify if this partition is to be used by jobs which do not explicitly 
@@ -249,6 +252,7 @@ Possible values are"YES" and "NO".
 Identify the node(s) to be associated with this partition. Multiple node names 
 may be specified using simple node range expressions (e.g. "lx[10-20]"). 
 Note that jobs may only be associated with one partition at any time.
+Specify a blank data value to remove all nodes from a partition: "Nodes=".
 .TP
 \fIPartitionName\fP=<name>
 Identify the partition to be updated. This specification is required.

diff --git a/doc/man/man1/srun.1 b/doc/man/man1/srun.1
@@ -223,15 +223,24 @@ the job. By default only errors are displayed.
 .TP
 \fB\-W\fR, \fB\-\-wait\fR=\fIseconds\fR
 Specify how long to wait after the first task terminates before terminating 
-all remaining tasks. The default value is unlimited. This can be useful to 
+all remaining tasks. A value of 0 indicates an unlimited wait (a warning will
-insure that a job is terminated in a timely fashion in the event that one 
+be issued after 60 seconds). The default value is set by the WaitTime 
-or more tasks terminate prematurely.
+parameter in the slurm configuration file (see \fBslurm.conf(5)\fR). This 
+option can be useful to insure that a job is terminated in a timely fashion 
+in the event that one or more tasks terminate prematurely.
 .TP
 \fB\-q\fR, \fB\-\-quit-on-interrupt\fR
 Quit immediately on single SIGINT (Ctrl-C). Use of this option
 disables the status feature normally available when \fBsrun\fR receives 
 a single Ctrl-C and causes \fBsrun\fR to instead immediately terminate the
-running job.
+running job. 
+.TP
+\fB\-X\fR, \fB\-\-disable-status\fR
+Disable the display of task status when srun receives a single SIGINT
+(Ctrl-C). Instead immediately forward the SIGINT to the running job.
+A second Ctrl-C in one second will forcibly terminate the job and
+\fBsrun\fR will immediately exit. May also be set via the environment
+variable SLURM_DISABLE_STATUS.
 .TP
 \fB\-Q\fR, \fB\-\-quiet\fR
 Quiet operation. Suppress informational messages. Errors will still
@@ -604,6 +613,9 @@ SLURM_TIMELIMIT
 .TP
 SLURM_WAIT
 \fB\-W, \-\-wait\fR=\fIseconds\fR
+.TP
+SLURM_DISABLE_STATUS
+\fB\-X, \-\-disable-status\fR
 .PP
 Additionally,
 .B srun

diff --git a/etc/init.d.slurm b/etc/init.d.slurm
@@ -16,6 +16,11 @@
 
 # $Id$
 
+BINDIR=/usr/bin
+CONFDIR=/etc/slurm
+LIBDIR=/usr/lib
+SBINDIR=/usr/sbin
+
 # Source function library.
 [ -f /etc/rc.d/init.d/functions ] || exit 0
 . /etc/rc.d/init.d/functions
@@ -28,14 +33,17 @@ else
     SLURMD_OPTIONS=""
 fi
 
-[ -f /etc/slurm/slurm.conf ] || exit 1
+[ -f $CONFDIR/slurm.conf ] || exit 1
+
+# setup library paths for slurm and munge support
+export LD_LIBRARY_PATH="$LIBDIR:$LD_LIBRARY_PATH"
 
 RETVAL=0
 
 start() {
     echo -n "starting $1: " 
     unset HOME MAIL USER USERNAME 
-    daemon /usr/sbin/$1 $2
+    daemon $SBINDIR/$1 $2
     RETVAL=$?
     echo
     touch /var/lock/subsys/slurm
@@ -52,7 +60,7 @@ stop() {
 }
 
 startall() {
-    for prog in `scontrol show daemons`; do 
+    for prog in `$BINDIR/scontrol show daemons`; do 
         optvar=`echo ${prog}_OPTIONS | tr "a-z" "A-Z"`
         start $prog ${!optvar} 
     done
@@ -68,7 +76,7 @@ slurmstatus() {
     local rpid
     local pidfile
 
-    pidfile=`grep -i ${base}pid /etc/slurm/slurm.conf | grep -v '^ *#'`
+    pidfile=`grep -i ${base}pid $CONFDIR/slurm.conf | grep -v '^ *#'`
     if [ $? = 0 ]; then
         pidfile=${pidfile##*=}
         pidfile=${pidfile%#*}
@@ -123,33 +131,34 @@ case "$1" in
         startall
         ;;
     stop)
-	for prog in `scontrol show daemons`; do
+	for prog in `$BINDIR/scontrol show daemons`; do
 	    stop $prog
 	done
         ;;
     status)
-	for prog in `scontrol show daemons`; do
+	for prog in `$BINDIR/scontrol show daemons`; do
 	   slurmstatus $prog
 	done
         ;;
     restart)
-        $0 stop && $0 start
+        $0 stop
+	$0 start
         ;;
     condrestart)
         if [ -f /var/lock/subsys/slurm ]; then
-            for prog in `scontrol show daemons`; do
+            for prog in `$BINDIR/scontrol show daemons`; do
                  stop $prog
                  start $prog
             done
         fi
         ;;
     reconfig)
-	for prog in `scontrol show daemons`; do
+	for prog in `$BINDIR/scontrol show daemons`; do
 	    killproc $prog -HUP
 	done
 	;;
     test)
-	for prog in `scontrol show daemons`; do   
+	for prog in `$BINDIR/scontrol show daemons`; do   
 	    echo "$prog runs here"
 	done
 	;;

diff --git a/slurm.spec.in b/slurm.spec.in
@@ -24,8 +24,12 @@ Requires: openssl >= 0.9.6
 #
 %if %{?_with_debug:1}%{!?_with_debug:0}
   %define _enable_debug --enable-debug
-  %define __os_install_post /usr/lib/rpm/brp-compress
 %endif
+#
+# Never allow rpm to strip binaries as this will break
+#  parallel debugging capability
+#
+%define __os_install_post /usr/lib/rpm/brp-compress
 
 
 %package devel
@@ -108,16 +112,16 @@ rm -rf "$RPM_BUILD_ROOT"
 mkdir -p "$RPM_BUILD_ROOT"
 DESTDIR="$RPM_BUILD_ROOT" make install
 
-install -D -m755 etc/init.d.slurm $RPM_BUILD_ROOT/etc/rc.d/init.d/slurm
+install -D -m755 etc/init.d.slurm $RPM_BUILD_ROOT/etc/init.d/slurm
 install -D -m644 etc/slurm.conf.example $RPM_BUILD_ROOT/etc/slurm/slurm.conf
 # Delete unpackaged files:
-rm -f $RPM_BUILD_ROOT/usr/lib/slurm/*.{a,la}
+rm -f $RPM_BUILD_ROOT/%{_libdir}/slurm/*.{a,la}
 
 # Build file lists for optional plugin packages
 for plugin in auth_munge auth_authd switch_elan; do
    LIST=./${plugin}.files
    touch $LIST
-   test -f $RPM_BUILD_ROOT/usr/lib/slurm/${plugin}.so &&
+   test -f $RPM_BUILD_ROOT/%{_libdir}/slurm/${plugin}.so &&
      echo %{_libdir}/slurm/${plugin}.so > $LIST
 done
 
@@ -154,7 +158,7 @@ rm -rf $RPM_BUILD_ROOT
 %{_libdir}/slurm/sched_builtin.so
 %{_libdir}/slurm/switch_none.so
 %dir %{_libdir}/slurm/src
-%config(noreplace) /etc/rc.d/init.d/slurm
+%config(noreplace) /etc/init.d/slurm
 %config(noreplace) /etc/slurm/slurm.conf
 #############################################################################
 
@@ -189,29 +193,24 @@ rm -rf $RPM_BUILD_ROOT
 #############################################################################
 
 %pre
-#if [ -x /etc/rc.d/init.d/slurm ]; then
+#if [ -x /etc/init.d/slurm ]; then
-#    if /etc/rc.d/init.d/slurm status | grep -q running; then
+#    if /etc/init.d/slurm status | grep -q running; then
-#        /etc/rc.d/init.d/slurm stop
+#        /etc/init.d/slurm stop
 #    fi
 #fi
 
 %post
 /sbin/ldconfig %{_libdir}
-if [ -x /etc/rc.d/init.d/slurm ]; then
+if [ $1 = 1 ]; then
-    [ -x /sbin/chkconfig ] && /sbin/chkconfig --del slurm
     [ -x /sbin/chkconfig ] && /sbin/chkconfig --add slurm
-#    if ! /etc/rc.d/init.d/slurm status | grep -q running \
-#      || ! /etc/rc.d/init.d/slurm test | grep -q  slurmctld; then
-#        /etc/rc.d/init.d/slurm start
-#    fi 
 fi
 
 %preun
-if [ "$1" = 0 ]; then
+if [ $1 = 0 ]; then
-    if [ -x /etc/rc.d/init.d/slurm ]; then
+    if [ -x /etc/init.d/slurm ]; then
         [ -x /sbin/chkconfig ] && /sbin/chkconfig --del slurm
-        if /etc/rc.d/init.d/slurm status | grep -q running; then
+        if /etc/init.d/slurm status | grep -q running; then
-            /etc/rc.d/init.d/slurm stop
+            /etc/init.d/slurm stop
         fi
     fi
 fi
@@ -224,6 +223,11 @@ fi
 
 
 %changelog
+* Fri Oct 01 2004 Mark Grondona <mgrondona@llnl.gov>
+- don't delete and add service in %post
+* Wed Aug 04 2004 Mark Grondona <mgrondona@llnl.gov>
+- don't allow rpm to strip binaries since this breaks interface to parallel
+  debuggers.
 * Thu Jul 29 2004 Morris Jette <jette1@llnl.gov>
 - added checkpoint_none.so and jobcomp_script.so plugins
 * Fri Mar 07 2004 Mark Grondona <mgrondona@llnl.gov>

diff --git a/src/common/bitstring.c b/src/common/bitstring.c
@@ -624,7 +624,8 @@ bitfmt2int (char *bit_str_ptr)
 	if (bit_str_ptr == NULL) 
 		return NULL;
 	size = strlen (bit_str_ptr) + 1;
-	bit_int_ptr = xmalloc ( sizeof (int *) * size);
+	bit_int_ptr = xmalloc ( sizeof (int *) * 
+			(size * 2 + 1));	/* more than enough space */
 	if (bit_int_ptr == NULL)
 		return NULL;
 
@@ -653,6 +654,7 @@ bitfmt2int (char *bit_str_ptr)
 			sum = 0;
 		}
 	}
+	assert(bit_inx < (size*2+1));
 	bit_int_ptr[bit_inx] = -1;
 	return bit_int_ptr;
 }