Merge remote-tracking branch 'origin/slurm-2.4'

NEWS

@@ -50,6 +50,9 @@ documents those changes that are of interest to users and admins.
  -- BLUEGENE - If a large block (> 1 midplane) is in error and underlying
     hardware is marked bad remove the larger block and create a block over
     just the bad hardware making the other hardware available to run on.
+ -- BLUEGENE - Handle job completion correctly if an admin removes a block
+    where other blocks on an overlapping midplane are running jobs.
+ -- BLUEGENE - correctly remove running jobs when freeing a block.
 
 * Changes in SLURM 2.4.1
 ========================

slurm.spec

@@ -10,7 +10,7 @@
 # --with auth_none   %_with_auth_none   1    build auth-none RPM
 # --with blcr        %_with_blcr        1    require blcr support
 # --with bluegene    %_with_bluegene    1    build bluegene RPM
-# --with cray_xt     %_with_cray_xt     1    build for Cray XT system
+# --with cray        %_with_cray        1    build for a Cray system
 # --with debug       %_with_debug       1    enable extra debugging within SLURM
 # --with lua         %_with_lua         1    build SLURM lua bindings (proctrack only for now)
 # --without munge    %_without_munge    1    don't build auth-munge RPM
@@ -145,6 +145,11 @@ BuildRequires: mysql-devel >= 5.0.0
 BuildRequires: postgresql-devel >= 8.0.0
 %endif
 
+%if %{slurm_with cray}
+BuildRequires: cray-MySQL-devel-enterprise
+Requires: cray-MySQL-devel-enterprise
+%endif
+
 %ifnos aix5.3
 # FIXME: AIX can't seem to find this even though this is in existance there.
 # We should probably figure out a better way of doing this, but for now we
@@ -383,7 +388,6 @@ Gives the ability for SLURM to use Berkeley Lab Checkpoint/Restart
 
 %build
 %configure --program-prefix=%{?_program_prefix:%{_program_prefix}} \
-	%{?slurm_with_cray_xt:--enable-cray-xt} \
 	%{?slurm_with_debug:--enable-debug} \
 	%{?slurm_with_partial_attach:--enable-partial-attach} \
 	%{?slurm_with_sun_const:--enable-sun-const} \

src/plugins/select/bluegene/bg_core.c

@@ -534,7 +534,7 @@ extern int free_block_list(uint32_t job_id, List track_list,
 			itr = list_iterator_create(bg_record->job_list);
 			while ((job_ptr = list_next(itr))) {
 				if ((job_ptr->magic != JOB_MAGIC)
-				    || !IS_JOB_FINISHED(job_ptr))
+				    || IS_JOB_FINISHED(job_ptr))
 					continue;
 				freeit = xmalloc(sizeof(kill_job_struct_t));
 				freeit->jobid = job_ptr->job_id;

src/plugins/select/bluegene/select_bluegene.c

@@ -2528,14 +2528,15 @@ extern int select_p_update_block(update_block_msg_t *block_desc_ptr)
 					info("Pending job %u on block %s "
 					     "will try to be requeued "
 					     "because overlapping block %s "
-					     "is in an error state.",
+					     "is being removed.",
 					     found_record->job_running,
 					     found_record->bg_block_id,
 					     bg_record->bg_block_id);
 				else
-					info("Failing job %u on block %s "
+					info("Running job %u on block %s "
+					     "will try to be requeued "
 					     "because overlapping block %s "
-					     "is in an error state.",
+					     "is being removed.",
 					     found_record->job_running,
 					     found_record->bg_block_id,
 					     bg_record->bg_block_id);
@@ -2548,7 +2549,7 @@ extern int select_p_update_block(update_block_msg_t *block_desc_ptr)
 				struct job_record *job_ptr = NULL;
 				ListIterator job_itr = list_iterator_create(
 					found_record->job_list);
-				while ((job_ptr = list_next(itr))) {
+				while ((job_ptr = list_next(job_itr))) {
 					if (job_ptr->magic != JOB_MAGIC) {
 						error("select_p_update_block: "
 						      "bad magic found when "
@@ -2562,18 +2563,17 @@ extern int select_p_update_block(update_block_msg_t *block_desc_ptr)
 						info("Pending job %u on "
 						     "block %s "
 						     "will try to be requeued "
-						     "because overlapping "
-						     "block %s "
+						     "because related block %s "
 						     "is in an error state.",
 						     job_ptr->job_id,
 						     found_record->bg_block_id,
 						     bg_record->bg_block_id);
 					else
-						info("Failing job %u on "
-						     "block %s "
-						     "because overlapping "
+						info("Running job %u on "
 						     "block %s "
-						     "is in an error state.",
+						     "will try to be requeued "
+						     "because related block %s "
+						     "is being removed.",
 						     job_ptr->job_id,
 						     found_record->bg_block_id,
 						     bg_record->bg_block_id);