- 11 Jan, 2017 6 commits
-
-
Danny Auble authored
and erroneous error later on.
-
Morris Jette authored
The old logic would result in test16.4 failing some of the time. The failure was caused by the sattach command attaching to a job step before the original srun command received a RESPONSE_LAUNCH_TASKS message. That messsage would then be sent to the salloc command. Since srun never got the message, it would hang. This change does not mark the job step as RUNNING until after the original srun gets sent the RESPONSE_LAUNCH_TASKS message and sattach requests are blocked until that time.
-
Morris Jette authored
Identify the function where an error is generated.
-
Morris Jette authored
-
Morris Jette authored
-
Morris Jette authored
This bug was introduced in commit 155a65a4 A bad array index value was used.
-
- 10 Jan, 2017 18 commits
-
-
Morris Jette authored
This isn't a new bug, but logic in this area changed and Coverity reported it as a new problem (duplicate logic). The logic also incorrectly used a strstr() rather than strcmp() to find maching process names.
-
Morris Jette authored
Fixing a resource leak reported by Coverity in the SPANK logic resulted in unmounting the spank object and losing locally stored information. Regression test7.11 broke with the new code. We'll need to live with the error reported by Coverity and keep SPANK running. This is a partial reversion of commit 6fb3fea3
-
Morris Jette authored
Regression tests 1.63 was sometimes failing due to a race condition. Specifically, the srun command was signalled prior to completion of the task launch process, so the "tasks_started" bitmap did not yet have it's bits set, resulting in the signal send request being silently ignored.
-
Morris Jette authored
This sets a memory pointer to NULL after releasing its memory to avoid a duplicate xfree call. Bug introduced in commit 07ae3e72
-
Morris Jette authored
-
Morris Jette authored
This is a fix to commit d470d0a9
-
Tim Wickberg authored
-
Tim Wickberg authored
No longer needed (as noted by comment) since 14.11 protocol support is already removed.
-
Tim Wickberg authored
Missed in previous AIX removal.
-
Tim Wickberg authored
-
Tim Wickberg authored
-
Tim Wickberg authored
-
Tim Wickberg authored
All unused within their scope; no functional change.
-
Morris Jette authored
3 places where a string pointer referenced an out-of-scope memory location 1 memory leak and 1 unterminated string
-
Morris Jette authored
-
Morris Jette authored
Accidentally commited to Slurm version 16.05 instead of master
-
Morris Jette authored
-
Morris Jette authored
-
- 09 Jan, 2017 16 commits
-
-
Morris Jette authored
3 places where a string pointer referenced an out-of-scope memory location 1 memory leak and 1 unterminated string
-
Morris Jette authored
4 possible memory leaks 2 failures to check for errors returned by functions and 4 possibly unterminated strings
-
Morris Jette authored
-
Morris Jette authored
backfill scheduler: Stop trying to determine expected start time for a job after 2 seconds of wall time. This can happen if there are many running jobs and a pending job can not be started soon. byg 3373
-
Danny Auble authored
mobile devices didn't behave as expected.
-
Tim Wickberg authored
This reverts commit 17549a03.
-
Tim Wickberg authored
This reverts commit 7495090d.
-
Danny Auble authored
# Conflicts: # src/sbcast/opts.c
-
Tim Shaw authored
-
Danny Auble authored
-
Danny Auble authored
some of the tables (prolog_epilog) were hard to read.
-
Danny Auble authored
-
Dominik Bartkiewicz authored
Bug 3364.
-
Tim Shaw authored
Configuring slurm with munge manually installed in /usr/local, with the library in /usr/local/lib but an empty /usr/local/lib64 directory will cause the munge plugins to look for libmunge.so in the wrong place. The munge.spec file has historically provided libmunge.so as part of munge-devel, which Slurm depends on already.
-
Morris Jette authored
Add SchedulerParameters configuration parameter of "default_gbytes", which treats numeric only (no suffix) value for memory and tmp disk space as being in units of Gigabytes. Mostly for compatability with LSF.
-
Morris Jette authored
Move BatchScript to end of each job's information when using "scontrol -dd show job" to make it more readable.
-