- 29 Sep, 2016 5 commits
-
-
Tim Wickberg authored
-
Tim Wickberg authored
Switch to list_for_each, and check if access list actually changed after each update before updating last_prat_update. This prevents the backfill scheduler from resetting mid-cycle unnecessarily. Bug 3123.
-
Morris Jette authored
-
Josko Plazonic authored
Buffer was being allocated based upon the size of the wrong structure. The buffer would have been larger than required, so this bug should not have caused any failures. bug 3127
-
Morris Jette authored
Add protocol version to slurmd startup communications for slurmstepd to permit changes in the protocol.
-
- 28 Sep, 2016 12 commits
-
-
Morris Jette authored
-
Morris Jette authored
Added memory limits to the job and step. Without these, only one step may be able to run at a time and break the test
-
Morris Jette authored
Add "flag" field to launch_tasks_request_msg. Remove the following fields (moved into flags): multi_prog, task_flags, user_managed_io, pty, buffered_stdio, and labelio. More flags to be added later.
-
Brian Christiansen authored
-
Brian Christiansen authored
*** CID 149289: Null pointer dereferences (REVERSE_INULL) /slurm/src/slurmctld/step_mgr.c: 1993 in _step_dealloc_lps() 1987 xassert(job_resrcs_ptr->cpus); 1988 xassert(job_resrcs_ptr->cpus_used); 1989 1990 if (step_ptr->step_layout == NULL) /* batch step */ 1991 return; 1992 >>> CID 149289: Null pointer dereferences (REVERSE_INULL) >>> Null-checking "job_resrcs_ptr" suggests that it may be null, but it has already been dereferenced on all paths leading to the check. 1993 if (job_resrcs_ptr == NULL) 1994 return; 1995 i_first = bit_ffs(job_resrcs_ptr->node_bitmap); 1996 i_last = bit_fls(job_resrcs_ptr->node_bitmap); 1997 if (i_first == -1) /* empty bitmap */ 1998 return; ** CID 149288: Resource leaks (RESOURCE_LEAK) /slurm/src/api/reconfigure.c: 171 in _send_message_controller() ________________________________________________________________________________________________________ *** CID 149288: Resource leaks (RESOURCE_LEAK) /slurm/src/api/reconfigure.c: 171 in _send_message_controller() 165 } 166 resp_msg = xmalloc(sizeof(slurm_msg_t)); 167 slurm_msg_t_init(resp_msg); 168 169 if ((rc = slurm_receive_msg(fd, resp_msg, 0)) != 0) { 170 slurm_shutdown_msg_conn(fd); >>> CID 149288: Resource leaks (RESOURCE_LEAK) >>> Variable "resp_msg" going out of scope leaks the storage it points to. 171 return SLURMCTLD_COMMUNICATIONS_RECEIVE_ERROR; 172 } 173 174 if (slurm_shutdown_msg_conn(fd) != SLURM_SUCCESS) 175 rc = SLURMCTLD_COMMUNICATIONS_SHUTDOWN_ERROR; 176 else if (resp_msg->msg_type != RESPONSE_SLURM_RC) ** CID 149287: Integer handling issues (NEGATIVE_RETURNS) /slurm/src/slurmdbd/backup.c: 65 in run_dbd_backup() ________________________________________________________________________________________________________ *** CID 149287: Integer handling issues (NEGATIVE_RETURNS) /slurm/src/slurmdbd/backup.c: 65 in run_dbd_backup() 59 primary_resumed = false; 60 61 memset(&slurmdbd_conn, 0, sizeof(slurm_persist_conn_t)); 62 slurmdbd_conn.rem_host = slurmdbd_conf->dbd_addr; 63 slurmdbd_conn.rem_port = slurmdbd_conf->dbd_port; 64 slurmdbd_conn.cluster_name = "backup_slurmdbd"; >>> CID 149287: Integer handling issues (NEGATIVE_RETURNS) >>> Assigning: "slurmdbd_conn.fd" = a negative value. 65 slurmdbd_conn.fd = -1; 66 slurmdbd_conn.shutdown = &shutdown_time; 67 68 slurm_persist_conn_open_without_init(&slurmdbd_conn); 69 70 /* repeatedly ping Primary */ ** CID 149286: Integer handling issues (INCOMPATIBLE_CAST) ________________________________________________________________________________________________________ *** CID 149286: Integer handling issues (INCOMPATIBLE_CAST) /slurm/src/common/slurmdbd_defs.c: 3096 in slurmdbd_unpack_init_msg() 3090 3091 *msg = msg_ptr; 3092 3093 /* We don't use rollback going forward and version was packed after it 3094 * unfortunately */ 3095 if (rpc_version < SLURM_17_02_PROTOCOL_VERSION) >>> CID 149286: Integer handling issues (INCOMPATIBLE_CAST) >>> Pointer "&tmp32" points to an object whose effective type is "unsigned int" (32 bits, unsigned) but is dereferenced as a narrower "unsigned short" (16 bits, unsigned). This may lead to unexpected results depending on machine endianness. 3096 safe_unpack16((uint16_t *)&tmp32, buffer); 3097 safe_unpack16(&msg_ptr->version, buffer); 3098 safe_unpackstr_xmalloc(&msg_ptr->cluster_name, &tmp32, buffer); 3099 3100 /* We find out the version of the caller right here so use 3101 that as the rpc_version. */ ** CID 149285: Null pointer dereferences (FORWARD_NULL) /slurm/src/common/slurm_persist_conn.c: 588 in slurm_persist_conn_open() ________________________________________________________________________________________________________ *** CID 149285: Null pointer dereferences (FORWARD_NULL) /slurm/src/common/slurm_persist_conn.c: 588 in slurm_persist_conn_open() 582 error("%s: Failed to unpack persistent connection init resp message from %s:%d", 583 __func__, 584 persist_conn->rem_host, 585 persist_conn->rem_port); 586 _close_fd(&persist_conn->fd); 587 } else >>> CID 149285: Null pointer dereferences (FORWARD_NULL) >>> Dereferencing null pointer "resp". 588 persist_conn->version = resp->ret_info; 589 } 590 591 end_it: 592 593 slurm_persist_free_rc_msg(resp); ** CID 149284: Error handling issues (CHECKED_RETURN) /slurm/src/common/slurmdbd_defs.c: 1665 in _send_fini_msg() ________________________________________________________________________________________________________ *** CID 149284: Error handling issues (CHECKED_RETURN) /slurm/src/common/slurmdbd_defs.c: 1665 in _send_fini_msg() 1659 buffer = init_buf(1024); 1660 pack16((uint16_t) DBD_FINI, buffer); 1661 req.commit = 0; 1662 req.close_conn = 1; 1663 slurmdbd_pack_fini_msg(&req, SLURM_PROTOCOL_VERSION, buffer); 1664 >>> CID 149284: Error handling issues (CHECKED_RETURN) >>> Calling "slurm_persist_send_msg" without checking return value (as is done elsewhere 5 out of 6 times). 1665 slurm_persist_send_msg(slurmdbd_conn, buffer); 1666 free_buf(buffer); 1667 1668 return SLURM_SUCCESS; 1669 } 1670
-
Morris Jette authored
-
Tim Wickberg authored
Leave macros defined for now to avoid bit reuse.
-
Tim Wickberg authored
Leave the select/bluegene directory alone, as it will be removed entirely once BG/Q support is removed, and the cleanup isn't worth the risk as that is code is rarely changed.
-
Tim Wickberg authored
-
Tim Wickberg authored
Remove from build system, and delete L/P specific files. Run autogen.sh as well.
-
Tim Wickberg authored
Add in missing brackets. Caught by GCC 6.1 -Wall.
-
Morris Jette authored
Add "sbatch_wait_nodes" to SchedulerParameters to control default sbatch behaviour with respect to waiting for all allocated nodes to be ready for use. Job can override the configuration option using the --wait-all-nodes=# option. bug 3120
-
- 27 Sep, 2016 5 commits
-
-
Morris Jette authored
Prior logic would treat execute line like this: $ sbatch --wait-all-nodes -N3 tmp with "-N3" as being the argument to the "--wait-all-nodes" option. See bug 3120
-
Morris Jette authored
-
Morris Jette authored
Add salloc/sbatch/srun option --use-min-nodes to prefer smaller node counts when a range of node counts is specified (e.g. "-N 2-4"). bug 2996
-
Morris Jette authored
-
Tim Wickberg authored
Switch a few SLURM mentions for Slurm as well.
-
- 26 Sep, 2016 7 commits
-
-
Morris Jette authored
-
Morris Jette authored
It was out of alphabetic order before (e.g. after --power).
-
Morris Jette authored
Add salloc/sbatch/srun --priority option of "TOP" to set job priority to the highest possible value. This option is only available to Slurm operators and administrators. bug 3115
-
Morris Jette authored
The problem reported is just a configuration warning and not an error. Also change the test from ">=" to ">". bug 3086
-
Morris Jette authored
-
Morris Jette authored
-
Morris Jette authored
This patch finally resolves absolute/relative CPU mapping for nodes where the NUMA (or sockets) have different core counts (e.g. KNL SNC4).
-
- 25 Sep, 2016 2 commits
-
-
Morris Jette authored
-
Morris Jette authored
-
- 24 Sep, 2016 6 commits
-
-
Morris Jette authored
-
Morris Jette authored
-
Morris Jette authored
-
Morris Jette authored
bug 3090
-
Morris Jette authored
Cray was having problems with this for some reason
-
Morris Jette authored
Make sure no attempt is made to schedule a requeued job until all steps are cleaned (Node Health Check completes for all steps on a Cray). bug 3082
-
- 23 Sep, 2016 3 commits
-
-
Tim Wickberg authored
-
Brian Christiansen authored
-
Brian Christiansen authored
-