diff --git a/NEWS b/NEWS
index 1ce16b53954ac71ba036dd3a4a51b1135c30a7c6..6d6901f7cdb5d0c33831eb6646980ae562552ad4 100644
--- a/NEWS
+++ b/NEWS
@@ -3,8 +3,6 @@ documents those changes that are of interest to users and admins.
* Changes in SLURM 0.5.0-pre1
=============================
--- Fixes for reported problems:
- - slurm/538: user tasks block writing to stdio
-- Add network option to srun and job creation API for specification
of communication protocol over IBM Federation switch.
-- Add new slurm.conf parameter ProctrackType (process tracking) and
@@ -15,6 +13,8 @@ documents those changes that are of interest to users and admins.
-- Add federation switch plugin.
-- Add new configuration keyword, SchedulerRootFilter, to control
external scheduler control of RoolOnly partition (Chris Holmes, HP).
+ -- Modify logic to set process group ID for spawned processes (last
+ patch from slurm v0.3.11).
* Changes in SLURM 0.4.7
========================
@@ -84,6 +84,7 @@ documents those changes that are of interest to users and admins.
- 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.
+ - Fix for reported problem: slurm/538: user tasks block writing to stdio
* Changes in SLURM 0.4.0
========================
diff --git a/src/slurmd/smgr.c b/src/slurmd/smgr.c
index cc025b5982ab6e61d1190a35f3604bef6b98e2f3..2fa64d2fac5752cb28d047ea047fa61f38d5b3c5 100644
--- a/src/slurmd/smgr.c
+++ b/src/slurmd/smgr.c
@@ -77,7 +77,7 @@ static int smgr_sigarray[] = {
*/
static void _session_mgr(slurmd_job_t *job);
static int _exec_all_tasks(slurmd_job_t *job);
-static void _exec_task(slurmd_job_t *job, int i, int fd);
+static void _exec_task(slurmd_job_t *job, int i);
static int _become_user(slurmd_job_t *job);
static void _make_tmpdir(slurmd_job_t *job);
static int _child_exited(void);
@@ -172,7 +172,7 @@ _session_mgr(slurmd_job_t *job)
}
if (chdir(job->cwd) < 0) {
- error("couldn't chdir to '%s': %m: going to /tmp instead",
+ error("couldn't chdir to `%s': %m: going to /tmp instead",
job->cwd);
if (chdir("/tmp") < 0) {
error("couldn't chdir to /tmp either. dying.");
@@ -275,10 +275,9 @@ _become_user(slurmd_job_t *job)
static int
_exec_all_tasks(slurmd_job_t *job)
{
+ char c;
int i;
int fd = job->fdpair[1];
- int cpipe[2];
- uint8_t tid;
xassert(job != NULL);
xassert(fd >= 0);
@@ -291,26 +290,33 @@ _exec_all_tasks(slurmd_job_t *job)
if (xsignal_block(smgr_sigarray) < 0)
return error ("Unable to block signals");
- if (pipe (cpipe) < 0)
- return error ("Unable to open child pipe: %m");
-
for (i = 0; i < job->ntasks; i++) {
- pid_t pid = fork();
+ int fdpair[2];
+ pid_t pid;
+
+ if (pipe (fdpair) < 0)
+ error ("exec_all_tasks: pipe: %m");
- if (pid < 0) {
+ if ((pid = fork ()) < 0) {
error("fork: %m");
return SLURM_ERROR;
- } else if (pid == 0) { /* child */
- _exec_task(job, i, cpipe[1]);
- /* Should never return */
- exit(0);
+ } else if (pid == 0) { /* child */
+ /*
+ * Stall exec until pgid is set by parent
+ */
+ if (read (fdpair[0], &c, sizeof (c)) != 1)
+ error ("pgrp child read failed: %m");
+ close (fdpair[0]);
+ close (fdpair[1]);
+
+ _exec_task(job, i);
}
/* Parent continues:
*/
verbose ("task %lu (%lu) started %M",
(unsigned long) job->task[i]->gid,
- (unsigned long) pid);
+ (unsigned long) pid);
/*
* Send pid to job manager
@@ -323,10 +329,20 @@ _exec_all_tasks(slurmd_job_t *job)
job->task[i]->pid = pid;
/*
- * For task 0, wait until it has created a new pgrp.
+ * Set this child's pgid to pid of first task
+ */
+ if (setpgid (pid, job->task[0]->pid) < 0)
+ error ("Unable to put task %d (pid %ld) into pgrp %ld",
+ i, pid, job->task[0]->pid);
+
+ /*
+ * Now it's ok to unblock this child, so it may call exec
*/
- if (i == 0)
- read (cpipe[0], &tid, sizeof (tid));
+ if (write (fdpair[1], &c, sizeof (c)) != 1)
+ error ("write to unblock task %d failed", i);
+
+ close (fdpair[0]);
+ close (fdpair[1]);
/*
* Prepare process for attach by parallel debugger
@@ -335,50 +351,18 @@ _exec_all_tasks(slurmd_job_t *job)
_pdebug_trace_process(job, pid);
}
- /*
- * Wait for all tasks to finish joining new process group
- */
- for (i = 1; i < job->ntasks; i++)
- read (cpipe[0], &tid, sizeof (tid));
-
- close (cpipe[0]);
- close (cpipe[1]);
-
return SLURM_SUCCESS;
}
static void
-_exec_task(slurmd_job_t *job, int i, int fd)
+_exec_task(slurmd_job_t *job, int i)
{
- /*
- * Fit taskid into single byte for writing back to
- * initiating slurmd. Doesn't matter if the values
- * are truncated, we just need the number of bytes to
- * be right.
- */
- uint8_t tid = i;
-
if (xsignal_unblock(smgr_sigarray) < 0) {
error("unable to unblock signals");
exit(1);
}
- /*
- * Move this process into new pgrp within this session.
- */
- if (setpgid (0, i ? job->task[0]->pid : 0) < 0)
- error ("Unable to put task %d into pgrp %ld: %m",
- i, job->task[0]->pid);
-
- /*
- * Notify slurmd that pgid has been set for this task
- */
- if (write (fd, &tid, sizeof (tid)) < 0)
- error ("Unable to notify slurmd that task %d started\n", i);
-
- close (fd);
-
if (!job->batch) {
if (interconnect_attach(job->switch_job, &job->env,
job->nodeid, (uint32_t) i, job->nnodes,
@@ -609,11 +593,11 @@ _setup_env(slurmd_job_t *job, int taskid)
return -1;
if (setenvpf(&job->env, "SLURM_PROCID", "%d", t->gid ) < 0)
return -1;
-
if (getenvp(job->env, "SLURM_GMPI")) {
if (setenvpf(&job->env, "GMPI_ID", "%d", t->gid) < 0)
return -1;
}
+
return SLURM_SUCCESS;
}
@@ -648,7 +632,7 @@ _pdebug_trace_process(slurmd_job_t *job, pid_t pid)
if (job->task_flags & TASK_PARALLEL_DEBUG) {
int status;
waitpid(pid, &status, WUNTRACED);
- if (kill(pid, SIGSTOP) < 0)
+ if ((pid > (pid_t) 0) && (kill(pid, SIGSTOP) < 0))
error("kill(%lu): %m", (unsigned long) pid);
if (_PTRACE(PTRACE_DETACH, pid, NULL, 0))
error("ptrace(%lu): %m", (unsigned long) pid);