Assign all cpu's on node to the daemon
Signed-off-by: Ralph Castain <rhc@pmix.org>
(cherry picked from commit 7bac7eed6ef423e47fe980b4c32eae36b8e1d4cb)
Fixes#8195. This PR doesn't fix all the warnings from #8195, but
fixes many of them (e.g., I didn't get the "string might be truncated"
warnings on my Mac).
This is an adaptation of 14aa5fae3c42f14a1c6a259dede93d5ca7ecb82c from
master; it drops some things that aren't relevant here on the v4.1.x
branch and adds a few more warnings fixes that are relevant here on
v4.1.x that aren't relevant on master.
Signed-off-by: Jeff Squyres <jsquyres@cisco.com>
(cherry-picked from 14aa5fae3c42f14a1c6a259dede93d5ca7ecb82c)
Mark the node as "unusable" so it does not get included when computing
number of procs for the case where the user does not specify -np.
Signed-off-by: Ralph Castain <rhc@pmix.org>
Ensure we always pass the cpuset as well as the locality string for each
proc. Correct the mtl/ofi component's computation of relative locality
as the function being called expects to be given the locality string of
each proc, not the cpuset. If the locality string of the current proc
isn't available, then use the cpuset if available and compute the
locality before trying to compute relative localities of our peers.
Signed-off-by: Ralph Castain <rhc@pmix.org>
* The `--cpu-list` argument restricts the hwloc topology. When that topology is
sent from the remote daemon back to the HNP it is packed as XML. This packing
results in the loss of the applied cpu-list. Mapping will then be using
the full topology instead of the restricted topology when mapping the
processes. When the launch command reaches the remote node it will not
be congruent with the remote node's view of the topology leading to a
launch failure especially if the HNP and remote node have differing
topologies.
- The solution is to make sure that the HNP re-applies the cpu-list to
the incomming topology from the remote node. This way the HNP and the
remote node are using the same restricted topology.
Signed-off-by: Joshua Hursey <jhursey@us.ibm.com>
This patch fix a race condition, which caused the main thread
to sometimes write to a closed pipe.
The following are details:
Currently, during shut down, the main thread will do the following:
1. set listen_thread_action to false.
2. write to stop_thread pipe to tell the listener thread to exit.
The listener thread do the following:
1. call select() to listen to a set of file descriptors with
a maximum wait time.
2. check listen_thread_action. If it is false, close the stop_thread
pipe.
The main thread will write to closed pipe, when
1. listener's call to select() finished because maximum wait time reached.
2. main thread set listen_thread_action to false
3. listener thread check listen_thread_action and closed the pipe
4. main thread write to the closed pipe.
This patch address the issue by having the main thread close the pipe
after the listener thread has been joined. This way, main thread
will both write and close the thread, so there is no conflict.
Note This patch was opened directly against v4.1.x branch because
the orte/mca/oob/tcp directory has been removed from master branch.
Signed-off-by: Wei Zhang <wzam@amazon.com>
* Adds the `schizo/jsm` component that detects if the process was
direct launched with IBM's Job Step Manager (JSM). JSM is a PMIx
enhanced runtime environment so flag it as such.
Signed-off-by: Joshua Hursey <jhursey@us.ibm.com>
(cherry picked from commit 4f1de51371048085b86ee64e05849ad929c9f35c)
Every mapper is required to set the locale, which is why it is an error
if the locale attribute isn't found. Likewise, it is an error for any
mapper to set a NULL locale as it makes no sense. However, I can see
that maybe some compiler or static code checker might want to see
concrete evidence we checked it - so check it in the right place.
Backport the equivalent code from PRRTE as we know that works - more
confidence than trying to add another patch to this old code.
Signed-off-by: Ralph Castain <rhc@pmix.org>
* Additionally, fixes the `NULL` option to `OMPI_MCA_plm_rsh_agent`
would would also lead to a segv. Now it operates as intended by
disqualifying the `rsh` component and falling back onto the `isolated`
component.
Signed-off-by: Joshua Hursey <jhursey@us.ibm.com>
(cherry picked from commit 62d0058738e8a111cd099199bc5f1886f13aa8ec)
This change fixes the misleading error message. I added a conditional to
determine whether the error is due to a missing file or a bad interpreter.
If it is the latter, a new, more precise error message will be displayed.
Fixes#4528
Signed-off-by: Maxwell Coil <mcoil@nd.edu>
(cherry picked from commit 9b73f6ac83b1efd6db60d71a6e0a4c83b52af0aa)
in schizo/ompi, sets the new OMPI_MCA_mpi_oversubscribe environment
variable according to the node oversubscription state.
This MCA parameter is used to set the default value of the
mpi_yield_when_idle parameter.
This two steps tango is needed so the mpi_yield_when_idle setting
is always honored when set in a config file.
Refs. open-mpi/ompi#6433
Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>
(cherry-picked from cc97c0f6116ae62feef)
Signed-off-by: Scott Miller <scott.miller1@ibm.com>
(cherry picked from commit c1b8599528feaac3c1d9ad44f0113f65eb7cc97f)
Conflicts:
orte/mca/plm/rsh/plm_rsh_module.c
* The user can set `-mca odls_base_sigkill_timeout 30` to have ORTE wait
30 seconds before sending SIGTERM then another 30 seconds before sending
SIGKILL to remaining processes. This usually happens on an abnormal
termination. Sometimes the user wants to delay the cleanup to give the
system time to write out corefile or run other diagnostics.
* The problem is that child processes may be completing while ORTE is
in this loop. The SIGCHLD will interrupt the `sleep` system call.
Without the loop the sleep could effectively be ignored in this case.
- Sleep returns the amount of time remaining to sleep. If it was
interrupted by a signal then it is a positive number less than or
equal to the parameter passed to it. If it slept the whole time
then it returns 0.
Signed-off-by: Joshua Hursey <jhursey@us.ibm.com>
(cherry picked from commit 0e8a97c598d841d472047ea1025931813c3ef8a9)
Remove code for multiple OOB progress threads as it is an optimization
nobody uses. Also turns out to have a race condition that can cause
segfault on finalize, so maybe good that nobody is using it.
Signed-off-by: Ralph Castain <rhc@pmix.org>
(cherry picked from commit 41eb41c3f224abcacec78f4d77c138197c36f172)
(cherry picked from commit a2f35c1834ab2fcb216285621d177a179e33dfe7)
This commit fixes an compilation error when configured
with `--enable-timing`.
Procedures in the function `orte_ess_base_app_setup`
in `orte/mca/ess/base/ess_base_std_app.c` are moved
to `orte/mca/ess/pmi/ess_pmi_module.c`
and `orte/mca/ess/singleton/ess_singleton_module.c`
in the recent commit 57f6b94fa5.
In `ess_pmi_module.c`, the first argument of the
`OPAL_TIMING_ENV_NEXT` macro should have been adapted
to the destination function but was not.
In `ess_singleton_module.c`, `OPAL_TIMING_ENV_INIT`
was not used in the destination function originally.
So `OPAL_TIMING_ENV_NEXT` cannot be used in the function.
Signed-off-by: KAWASHIMA Takahiro <t-kawashima@jp.fujitsu.com>
(cherry picked from commit 8e7d874e14a5485dceff836419e36b6b24a66f48)
If both types of interfaces are enabled, don't error out if one of them
isn't able to open listener sockets. Only one interface family may be
available on some machines, but someone might want to build the code to
run more generally.
Refs https://github.com/pmix/prrte/pull/249
Signed-off-by: Ralph Castain <rhc@pmix.org>
(cherry picked from commit 06d188ebf3646760f50d4513361b50642af9cec4)
* Fix#6618
- See comments on Issue #6618 for finer details.
* The `plm/rsh` component uses the highest priority `routed` component
to construct the launch tree. The remote orted's will activate all
available `routed` components when updating routes. This allows the
opportunity for the parent vpid on the remote `orted` to not match
that which was expected in the tree launch. The result is that the
remote orted tries to contact their parent with the wrong contact
information and orted wireup will fail.
* This fix forces the orteds to use the same `routed` component as
the HNP used when contructing the tree, if tree launch is enabled.
Signed-off-by: Joshua Hursey <jhursey@us.ibm.com>
Provide a missing header and paren
Thanks to @zerothi for the assistance
Signed-off-by: Ralph Castain <rhc@pmix.org>
(cherry picked from commit bd5a1765eea200651babc5bfd9f45a9f3cedefbc)
Override the defaults when provided. Ignore LSF binding file if user
overrides by specifying a policy.
Fixes#6631
Signed-off-by: Ralph Castain <rhc@pmix.org>
(cherry picked from commit ea0dfc321809db50f78e742da1d22f9ef59650a3)
use strtol() instead of atoi() in order to handle hostnames
containing a large number.
This is a one-off commit for the release branches since
the regx framework has already been removed from master.
Refs. open-mpi/ompi#6729
Signed-off-by: perrynzhou <perrynzhou@gmail.com>
Remove the debruijn component as it changes the daemon's parent
process ID, thus breaking the other routed components
Signed-off-by: Ralph Castain <rhc@pmix.org>
SLURM 19 discontinued the use of --cpu_bind (and changed it to
--cpu-bind). There's no easy way to test at run time which one is
accepted, so set the environment variable SLURM_CPU_BIND to "none",
which should do the same thing as the srun CLI parameter.
Signed-off-by: Jordan Hayes <jhayes@ucr.edu>
Signed-off-by: Jeff Squyres <jsquyres@cisco.com>
(cherry picked from commit 7dad74032e30259506da7fa582dd8c4351e6e0a1)
This is so when a debugger attaches using MPIR, it can step out of this stack back into main.
This cannot be done with certain aggressive optimisations and missing debug information.
Signed-off-by: James Clark <james.clark@arm.com>
Signed-off-by: Jeff Squyres <jsquyres@cisco.com>
Co-authored-by: Jeff Squyres <jsquyres@cisco.com>
(cherry-picked from 20f5840)
Ensure we publish all the info required to be returned to the other
mpirun when executing this operation. We need to know the daemon (and
its URI) that is hosting each of the other procs so we can do a direct
modex operation and retrieve their connection info.
Signed-off-by: Ralph Castain <rhc@pmix.org>
(cherry picked from commit 60961ceb41cde6fbdc84694748696e90ddcfb345)
Since we intend to provide cross version compatibility
between versions with the same major and minor, use
MAJOR.MINOR.0 instead of orte_version_string
(e.g. MAJOR.MINOR.RELEASEGREEK).
Open MPI 4.0.0 has already been released, so in order to make
it compatible with future 4.0.x releases, we have to use 4.0.0
as the version string, that is why we use MAJOR.MINOR.0 instead
of MAJOR.MINOR
Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>
