The vader btl kept a per-peer registration cache to keep track of
attachments. This is not really a problem with small numbers of local
ranks but can be a problem with large SMP machines. To reduce the
footprint there is now one registration cache for all xpmem
attachments. This will probably increase the lookup time for large
transfers but is a worthwhile trade-off.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
OMPI header cannot be included in OPAL source code, hence removed it.
Fixes: (740b636db) btl/openib: Disqualify rdmacm CPC if
MPI_THREAD_MULTIPLE.
Signed-off-by: Potnuri Bharat Teja <bharat@chelsio.com>
We only support running with libfabric v1.3 or greater. So it's safe
to remove the legacy/adaptive cq_readerr() behavior.
Signed-off-by: Jeff Squyres <jsquyres@cisco.com>
There are critical usnic libfabric AV insert bugs before v1.3, so
don't allow any version prior to v1.3 at run time (still allow
*compiling* with earlier versions, though, since the ABI guarantees
allow us to compile with an earlier libfabric and run with a later
libfabric).
Switch to using fi_version() to check the version (instead of calling
fi_getinfo()) as a potentially lighter-weight / simpler solution.
This allows us to only call fi_getinfo() once.
Signed-off-by: Jeff Squyres <jsquyres@cisco.com>
The (effective) "+42" computation was, in fact, the incorrect answer
in this case (gasp!).
We should just take the max_msg_size from the command (which came from
the libfabric endpoint max_msg_size attribute in the client) and
subtract off the max header size: 68 (which is explained in the
comment). This will result in a "large" message size which is likely
slightly smaller than the MTU, but still right up near the MTU, and
therefore good enough.
Note: the old computation (i.e., -(68-42)) worked fine when we asked
for Libfabric API v1.1 because the usnic provider would return a
max_msg_size that was already less than the MTU due to FI_PREFIX
behavior shenanigans. Once we started asking for Libfabric API v1.4,
the usnic Libfabric provider started returning (MTU + prefix_size),
and the -(68-42) computation started giving a value that was over the
MTU. This caused sendto() on the connectivity checker UDP socket
to fail.
This commit also removes an old/misleading comment.
Signed-off-by: Jeff Squyres <jsquyres@cisco.com>
* Add a configure time option to rename libmpi(_FOO).*
- `--with-libmpi-name=STRING`
* This commit only impacts the installed libraries.
Internal, temporary libraries have not been renamed to limit the
scope of the patch to only what is needed.
For example:
```shell
shell$ ./configure --with-libmpi-name=wookie
...
shell$ find . -name "libmpi*"
shell$ find . -name "libwookie*"
./lib/libwookie.so.0.0.0
./lib/libwookie.so.0
./lib/libwookie.so
./lib/libwookie.la
./lib/libwookie_mpifh.so.0.0.0
./lib/libwookie_mpifh.so.0
./lib/libwookie_mpifh.so
./lib/libwookie_mpifh.la
./lib/libwookie_usempi.so.0.0.0
./lib/libwookie_usempi.so.0
./lib/libwookie_usempi.so
./lib/libwookie_usempi.la
shell$
```
The rdmacm CPC in the openib BTL is not thread safe. The rdmacm CPC
should disqualify itself (instead of failing in random ways) if
MPI_THREAD_MULTIPLE is the thread level.
Signed-off-by: Potnuri Bharat Teja <bharat@chelsio.com>
This patch is based on the "RFC: Reenabling the TCP BTL over local
interfaces (when specifically requested)". It removes the hardcoded
exception for the local devices that has been enforced by the
TCP BTL. Instead, we exclude the local interface only via the
exclude MCA (both IPv4 and IPv6 local addresses are already in the
default if_exclude), which is also the behavior currently described in
our README file.
This commit fixes an abort during finalize because pending events were
removed from the list twice.
References #2030
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
We commonly see messages on the users list where a peer has hung up
because it has crashed. Instead of having just a BTL_ERROR message,
make this a real opal_show_help() message that tells the user that the
peer unexpectedly hung up, and they should look into *why* that peer
hung up.
Signed-off-by: Jeff Squyres <jsquyres@cisco.com>
It looks like one help message was accidentally pasted in the middle
of another. Disentangle the two messages from each other, and
slightly tweak the one message to say that the job may also crash (in
addition to hanging).
Signed-off-by: Jeff Squyres <jsquyres@cisco.com>
This commit prevents the connection code from trying to connect an
endpoint if the directed datagram has been posted but not received.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
With libfabric v1.4, the usnic provider changed the values of its
fabric and domain name strings (compared to libfabric <v1.4). Update
the Open MPI usNIC BTL to handle both pre-v1.4 and v1.4 fabric/domain
names.
Signed-off-by: Jeff Squyres <jsquyres@cisco.com>
This commit fixes a race that can occur when two threads are in the
ugni progress function at the same time. This race occurs when one
thread calls GNI_PostDataProbeById then goes to sleep then another
thread calls GNI_PostDataProbeById then GNI_EpPostDataWaitById before
the other thread wakes up. If this happens the first thread will print
a warning on GNI_EpPostDataWaitById about no matching post.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
This commit fixes two issues that can occur during a connection:
- Re-entry to connection progress from modex lookup. Added an
additional endpoint state that will keep the code from re-entering
the common endpoint create.
- Fixed a race between a process posting a directed datagram through
a send and a connection being progressed through opal_progress().
The progress code was not obtaining the endpoint lock before
attempting to update the endpoint. To limit the amount of code
changed for 2.0.1 this commit makes the endpoint lock recursive. In
a future update this may be changed.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
The max inline send size on a queue pair is not available until after
the endpoint is connected. Before this commit the send flags
(including the inline flag) were set before this value was
initialized. This commit moves setting the send_flags down to
mca_btl_openib_put_internal which is only called after the endpoint is
connected. This fixes a bug when using osc/rdma.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
This commit fixes a long standing bug in rdmacm. It is required that
the thread that calls mca_btl_openib_endpoint_cpc_complete holds the
endpoint lock. This was not the case for rdmacm. This causes debug
builds to abort. This change also required changing
mca_btl_openib_endpoint_send_cts to require the endpoint lock to be
held when calling.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
This commit is an attempt to fix a hang in finalize of rdmacm. This fixes
a path where no rdmacm client is found for an endpoint.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
This commit fixes a segmentation fault that occurs if a device can be
initialized but not used. In this case the devices_count is not equal
to the number of usable devices in the devices pointer array.
Thanks to @artpol84 for tracking this down.
Fixesopen-mpi/ompi#1823
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
The rdma_disconnect function specifies that both the server and client
should call rdma_disconnect. The code was not calling rdma_disconnect
on an endpoint if the event came before the endpoint finalization.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
Before dynamic add_procs the openib_btl_size_queues was called exactly
once for non-dynamic jobs. Now the function is called on each new
connection so the calculation was wrong. Re-wrote the function to
correctly calculate the CQ size and only attempt to adjust the CQ if
the requested size has changed. This fixes a bug when using the openib
btl on psm2 hardware that is caused by the time needed to resize a
CQ. The overhead was causing udcm to timeout and fail.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
This commit fixes a programming error when using an aries nic. The
documentation of ugni shows that only the local alignment restriction
for get was lifted on aries. There is still a remote address alignment
restriction.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
This commit fixes a programming error when using an aries nic. The
documentation of ugni shows that only the local alignment restriction
for get was lifted on aries. There is still a remote address alignment
restriction.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
This commit adds support for Cray Aries atomic operations. This
includes 32-bit and floating point support.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
This commit add support for more atomic operations and type. The
operations added are logical and, logical or, logical xor, swap, min,
and max. New types are 32-bit int by using the
MCA_BTL_ATOMIC_FLAG_32BIT flag, 64-bit float by using the
MCA_BTL_ATOMIC_FLAG_FLOAT flag, and 32-bit float by using both
flags. Floating point numbers are supported by packing the number in
as an int64_t or int32_t. We will update the btl interface in the
future to make this less confusing.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
Before dynamic add_procs support was committed to master we called
add_procs with every proc in the job. The XRC code in the openib btl
was taking advantage of this and setting the number of work queue
entries (WQE) based on all the procs on a remote node. Since that is
no longer the case we can not simply increment the sd_wqe field on the
queue pair. To fix the issue a new field has been added to the xrc
queue pair structure to keep track of how many wqes there are total on
the queue pair. If a new endpoint is added that increases the number
of wqes and the xrc queue pair is already connected the code will
attempt to modify the number of wqes on the queue pair. A failure is
ignored because all that will happen is the number of active send work
requests on an XRC queue pair will be more limited.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
The original VADER_MAX_ADDRESS was tunned for x86_64 platforms only.
For non x86_64 platforms we can use XPMEM_MAXADDR_SIZE.
Signed-off-by: Pavel Shamis (Pasha) <pasharesearch@gmail.com>
This commit reduces the overhead of calling the ugni progress
function. It does the following:
- Check for new connections once every eight calls.
- Do not call remote smsg progress unless we are connected to at
least one remote peer.
- Do not call rdma progress unless at least one rdma fragment is
outstanding.
- Check endpoint wait list size before obtaining a lock.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
This commit reduces the default exclusivity so that btl/scif is not
used for send/recv over other shared memory transports.
Fixesopen-mpi/ompi#1712
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
The OPAL_CMA_NEED_SYSCALL_DEFS is always defined/set to 0 or 1. Therefore
instead of checking if the macro is defined, we have to look at the value
itself.
Signed-off-by: Pavel Shamis (Pasha) <pasharesearch@gmail.com>
This commit fixes several bugs in the registration cache code:
- Fix a programming error in the grdma invalidation function that can
cause an infinite loop if more than 100 registrations are
associated with a munmapped region. This happens because the
mca_rcache_base_vma_find_all function returns the same 100
registrations on each call. This has been fixed by adding an
iterate function to the vma tree interface.
- Always obtain the vma lock when needed. This is required because
there may be other threads in the system even if
opal_using_threads() is false. Additionally, since it is safe to do
so (the vma lock is recursive) the vma interface has been made
thread safe.
- Avoid calling free() while holding a lock. This avoids race
conditions with locks held outside the Open MPI code.
Fixesopen-mpi/ompi#1654.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
This commit fixes a compile/link issue caused by vader. The vader btl
was using OPAL_THREAD_ADD64 to increment a counter which may not be
available on 32-bit systems. Changed to use OPAL_THREAD_ADD_SIZE_T
which will be 64-bit or 32-bit depending on the system.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
This commit fixes a compilation issue with some versions of exp
verbs. In some cases struct ibv_exp_device_attr does not have either
the exp_atom or exp_atomic_cap fields. It is fine to drop one check
and fall back to the non-exp attribute check on the other.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
In some rare cases when a process receives the connect ack while
locally updating the peer endpoint structure, we could drop the
incomming connect ack due to the fact that the send handler is
protected with a try lock (on the endpoint) and our initial send
event was not persistent. Making the send event persistent solves
all issues.
This commit fixes a race between a thread calling the tcp btl's
add_procs and a thread processing an incomming connection. The race
occured because the add_procs thread adds a newly created proc object
to the hash table *before* the object is fully initialized. The
connection thread then attempts to use the object before the endpoints
array on the object has beeen allocation. The fix is to only add the
proc to the hash table after it has been completely initialized.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
The math for checking the number of QPs and CQs per usNIC/VF was
incorrect, allowing you to run MPI processes even when usNICs (i.e.,
VIC VFs) had fewer QPs and CQs than were necessary. This led to a
confusing error later when fi_enable(3) failed (because we lazily
create QPs). Fixing the math here ensure that we actually print a
helpful error message telling the user specifically what is wrong.
Signed-off-by: Jeff Squyres <jsquyres@cisco.com>
These macros should really be named OPAL_SUMMARY_*; they're used in
all projects, and therefore should be in the lowest later project (OPAL).
Signed-off-by: Jeff Squyres <jsquyres@cisco.com>
Because of the removal of the linux memory component it is no longer
necessary to initialize the memory component in opal_init(). This
commit moves the initialization to the creation of the first rcache
component.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
This comit adds code that sets the fragment reg_cnt to 1 when sending
the completion message for an eager get. Without this the btl will
either hang or abort.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
Added mca parameter to turn progress thread on/off
Add a flag to check if we have btl progress thread.
Added macro for ob1 matching lock.
Update the AUTHORS file.
All BTL-only operations (basically all data movements
with the exception of the matching operation) can now
be handled for the TCP BTL by a progress thread.
This reverts commit open-mpi/ompi@f7257a8310.
Ensure that we properly cleanup the session directory tree. Prior code had issues with symlinks, especially if the file that the link points to was already removed as we traverse the tree. Also found that the dirent checks for directory type weren't fully portable, and so fall back to the stat-based approach which is known to be portable.
Fix singularity singletons by detecting we are in a container and properly setting the pmix selection to pick the isolated component. Remove a stale restriction blocking use of the sm btl
This commit brings the scif btl up to date with changes made on master
to rework the mpool and rcache frameworks.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
This commit uses the new flag "enumerator" to support comma-delimited
lists of flags for both the btl and btl atomic flags. After this
commit is is valid to specify something like -mca btl_foo_flags
self,put,get,in-place. All non-deprecated flags are supported by the
enumerator.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
This commit adds a new type of enumerator meant to support flag
values. The enumerator parses comma-delimited strings and matches
each string or value to a list of valid flags. Additionally, the
enumerator does some basic checks to see if 1) a flag is valid in the
enumerator, and 2) if any conflicting flags are specified.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
This commit adds the data necessesary for supporting dynamic add_procs
to the rdma message (opal_process_name_t). The endpoint lookup
function has been updated to match the code in udcm.
Closesopen-mpi/ompi#1468.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
Fix CID 1345825 (1 of 1): Dereference before null check (REVERSE_INULL):
ib_proc should not be NULL in this case. Removed the check and added a
check for NULL after OBJ_NEW.
CID 1269821 (1 of 1): Dereference null return value (NULL_RETURNS):
I labeled this one as a false positive (which it is) but the code in
question could stand be be cleaned up.
Fix CID 1356424 (1 of 1): Argument cannot be negative (NEGATIVE_RETURNS):
While trying to silence another Coverity issue another was
flagged. Protect the close of fd with if (fd >= 0).
CID 70772 (1 of 1): Dereference null return value (NULL_RETURNS):
CID 70773 (1 of 1): Dereference null return value (NULL_RETURNS):
CID 70774 (1 of 1): Dereference null return value (NULL_RETURNS):
None of these are errors and are intentional but now that we have a
list release function use that to make these go away. The cleanup is
similar to CID 1269821.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
+ Added an mca parameter to allow connecting processes from different
subnets. Its current default value is 'false' - don't allow, to keep the
current flow the way it is now.
+ rmdacm: when calling ibv_query_gid, use the gid index from
btl_openib_gid_index.
This commit rewrites both the mpool and rcache frameworks. Summary of
changes:
- Before this change a significant portion of the rcache
functionality lived in mpool components. This meant that it was
impossible to add a new memory pool to use with rdma networks
(ugni, openib, etc) without duplicating the functionality of an
existing mpool component. All the registration functionality has
been removed from the mpool and placed in the rcache framework.
- All registration cache mpools components (udreg, grdma, gpusm,
rgpusm) have been changed to rcache components. rcaches are
allocated and released in the same way mpool components were.
- It is now valid to pass NULL as the resources argument when
creating an rcache. At this time the gpusm and rgpusm components
support this. All other rcache components require non-NULL
resources.
- A new mpool component has been added: hugepage. This component
supports huge page allocations on linux.
- Memory pools are now allocated using "hints". Each mpool component
is queried with the hints and returns a priority. The current hints
supported are NULL (uses posix_memalign/malloc), page_size=x (huge
page mpool), and mpool=x.
- The sm mpool has been moved to common/sm. This reflects that the sm
mpool is specialized and not meant for any general
allocations. This mpool may be moved back into the mpool framework
if there is any objection.
- The opal_free_list_init arguments have been updated. The unused0
argument is not used to pass in the registration cache module. The
mpool registration flags are now rcache registration flags.
- All components have been updated to make use of the new framework
interfaces.
As this commit makes significant changes to both the mpool and rcache
frameworks both versions have been bumped to 3.0.0.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
Add primitive magic number and version checking in the connectivity
checker protocol. These checks doesn't *guarantee* to we won't get
false PINGs and ACKs, but they do significantly reduce the possibility
of interpretating random incoming fragments as PINGs or ACKs.
The btl_recv.h:lookup_sender() function uses the hashed ORTE proc name
to determine the sender of the packet. With add_procs_cutoff>0, the
usnic BTL may not have knowledge of all the senders.
Until the usNIC BTL can be adjusted to do something like the
openib/ugni BTLs (i.e., use opal_proc_for_name() to lookup unknown
sender proc names), set MCA_BTL_FLAGS_SINGLE_ADD_PROCS, which means
that ob1 will only all add_procs() once -- with all the procs in it.
Also in this commit, adapt the connectivity checker to not rely on
knowing all the senders (which is a bit easier than adapting the main
BTL send path): the receiving connectivity agent will simply echo back
the same PING message (which contains the sender's IP address+UDP
port) back to the sender without checking that it knows who the sender
is. If the sender receives the echoed PING back on the expexted
interface, it will find a match in the pending pings list. If the
sender receives the echoed PING back an unexpected interface, a match
will not be found, and the incoming PING message will be dropped.
Fixesopen-mpi/ompi#1440
This commit fixes an inconsistency between btl_openib_receive_queues,
btl_openib_max_send_size and btl_openib_eager_limit. Before this
commit if the ini file specified a set of default receive queues that
happen to not contain one large enough for the default max_send_size
of eager_limit users would see an error like:
WARNING: The largest queue pair buffer size specified in the
btl_openib_receive_queues MCA parameter is smaller than the maximum
send size (i.e., the btl_openib_max_send_size MCA parameter), meaning
that no queue is large enough to receive the largest possible incoming
message fragment. The OpenFabrics (openib) BTL will therefore be
deactivated for this run.
Local host: somehost
Largest buffer size: 65536
Maximum send fragment size: 131072
This commit adds code that detects the source of the max_send_size and
eager_limit values and sets either or both of them to the size
supported by the largest queue pair if both 1) the value is larger
than the largest queue pair size, and 2) the value was not set by the
user or a MCA configuration file.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
This commit adds two m4 macros: OPAL_SUMMARY_ADD, OPAL_SUMMARY_PRINT.
OPAL_SUMMARY_ADD adds an item to a section in the summary. For example
OPAL_SUMMARY_ADD([[Transports]],[[Foo]],...,[yes]) will add the
following to the summary:
Transports
-----------------------
Foo: yes
With this commit two sections are added: Transports, Resource Managers.
The OPAL_SUMMARY_PRINT macro is called after AC_OUTPUT and prints out
some information about the build (version, projects, etc) and then
the summarys sections. It will additionally print a warning if
internal debugging is enabled.
Example output:
Open MPI configuration:
-----------------------
Version: 3.0.0 a1
Build Open Platform Abstration project: yes
Build Open Runtime project: yes
Build Open MPI project: yes
Build Open SHMEM project: no
MPI C++ bindings (deprecated): no
MPI Fortran bindings: mpif.h, use mpi, use mpi_f08
Debug build: yes
Transports
-----------------------
Cray uGNI (Gemini/Aries): no
Intel Omnipath (PSM2): no
KNEM Shared Memory: no
Linux CMA IPC: no
Mellanox MXM: no
Open UCX: no
OpenFabrics libfabric: no
OpenFabrics Verbs: no
portals4: no
QLogic Infinipath (PSM): no
tcp: yes
XPMEM Shared Memory: no
Resource Managers
-----------------------
Cray Alps: no
Grid Engine: no
LSF: no
Slurm: yes
Torque: yes
INTERNAL DEBUGGING IS ENABLED. DO NOT USE THIS BUILD FOR PERFORMANCE MEASUREMENTS!
Signed-off-by: Nathan Hjelm <hjelmn@me.com>
This commit fixes several threading bugs:
- Add an additional lock to the btl_base_endpoint_t structure to lock
the list of pending frags. This allows the progress function to
attempt to send pending frags without needing to drop/reaquire the
lock. This should provide a small improvement in performance and
fixes a potential race between adding an removing items from the
pending list.
- Ensure fast boxes are only set up once by updating the send count
using atomics when needed and do not set the fast box buffer
pointer until the fast box is set up.
Closesopen-mpi/ompi#1408
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
This commit fixes a bug that can occur when communicating via XRC to
peers on the same node. UDCM was not saving the SRQ numbers on the
loopback endpoint (which shares its ib_addr info with all local peers)
so any messages to local peers use an invalid SRQ number.
Fixesopen-mpi/ompi#1383
Signed-off-by: Nathan Hjelm <hjelmn@me.com>
This bug fixes two issue with the ib_addr lock:
- The ib_addr lock must always be obtained regardless of
opal_using_threads() as the CPC is run in a seperate thread.
- The ib_addr lock is held in mca_btl_openib_endpoint_connected when
calling back into the CPC start_connect on any pending
connections. This will attempt to obtain the ib_addr lock
again. Since this is not a performance-critical part of the code
the lock has been changed to be recursive.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
This commit fixes a bug that occurs when attempting a get or put
operation on an endpoint that is not already connected. In this case
the remote_srqn may be set to an invalid value as the rem_srqs array
on the endpoint is not populated. This commit moves the usage of the
rem_srqs array to the internal put/get functions where it is
guaranteed this array is populated.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
This commit ensures ib_addr->remote_xrc_rcv_qp_num value is set when
creating the loopback queue pair. This is needed when communicating
with any other local peer.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
This commit fixes two bugs in XRC support
- When dynamic add_procs support was added to master the remote
process name was added to the non-XRC request structure. The same
value was not added to the XRC xconnect structure. This error was
not caught because the send/recv code was incorrectly using the
wrong structure member. This commmit adds the member and ensure the
xconnect code uses the correct structure.
- XRC loopback QP support has been fixed. It was 1) not setting the
correct fields on the endpoint structure, 2) calling
udcm_xrc_recv_qp_connect, and 3) was not initializing the endpoint
data.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
mca_btl_openib_put incorrectly checks the qp inline max before
allowing an inline put. This check will always fail for an endpoint
that has not been connected. The commit changes the check to use the
btl_put_local_registration_threshold instead.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
Commit open-mpi/ompi@400af6c52d
introduced a regression in XRC support. The commit reversed the
ordering of shared receive queue (SRQ) and completion queue (CQ)
completion. CQ creation must always preceed SRQ creation when using
XRC as the CQs are needed to create the SRQs. This commit fixes the
ordering so that CQs are always created before SRQs.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
Cleanup the configury so we properly check for Singularity under the various typical use-cases
Bring the Singularity support online. We have to turn "off" the sm BTL as it segfaults from inside the container - root cause remains unclear. Also turned "off" the various OPAL shmem components in case they are involved and someone else tries to use them. Happily, the vader BTL works just fine!
These changes fix issue https://github.com/open-mpi/ompi/issues/1336
- improve abstractions: opal/memory/linux component should be single place that opeartes with
Memory Allocation Hooks.
- avoid collisions in case dynamic component open/close: it is safe because it is linked statically.
- does not change original behaivour.
The send code in the ugni btl has an optimization that enables it to
return 1 (fragment gone) in some cases. This optimization involved
removing the btl ownership and callback flags to ensure the fragment
stuck around long enough for its completion flag to be checked. This
works fine for the single-threaded case but not in the multi-threaded
case. It is possible that a fragment will be completed by another
thread while a thread is in mca_btl_ugni_send. This competition can
lead to a leaked fragment, missed callback, or both. To fix the issue
without removing the optimization a reference count has been added to
the fragment. Callbacks and fragment release will not be made until
the fragment reference count has reach 0. The count is incremented
before sending the frag and decremented after the completion flag has
been checked. The fix has been verified to work using a multi-threaded
RMA benchmark with the osc/pt2pt component.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
This commit fixes a race condition that can cause an endpoint to be
added to the wait list multiple times. To fix the issue an additional
check has been added to ensure the endpoint is not on the wait list
after the wait list lock is held. The wait list processing code has
also been updated to keep the wait list lock until all wait listed
endpoints have been handled. This reduces the chance that an endpoint
that is being processed by the wait list code is not re-added to the
list by a competing send.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
Three minor updates from the code review of
https://github.com/open-mpi/ompi-release/pull/933:
* Remove an extra blank line a show_help message
* We no longer allow -1 for the MCA param btl_usnic_av_eq_num, so
change the flag to REGINT_GE_ONE
* Change "num_blocks" definition to be in terms of block_len (not
eq_size)
A bunch of empirical testing has shown that increasing the retranmit
timeout from 1ms to 5ms doesn't adversely affect performance, yet
decreases the number of gratuitious retransmissions.
Add endpoints in a blocked manner so that we don't overrun the
fi_av_insert() event queue. Also make the AV EQ length an MCA param,
and report it in mca_btl_base_verbose >=5 output.
Sequence numbers will wrap around; it is not sufficient to check for
(seq-1) -- must use the SEQ_DIFF macro to properly handle the
wraparound.
This bug wasn't serious; it just meant we might retransmit one or two
extra times when retransmits were triggerd and the sequence numbers
wrapped around their sliding windows.
`cm_message_event_active == 1` but main thread has already stopped
processing messages and thus we will have the situation where one
message was left unhandled leading to a hang.
Per open-mpi/ompi#1230, add a comment explaining why we write to a
temporary file and then rename(2) the file, just so that future code
maintainers don't wonder why we do this seemingly-useless step.
The problem was in mca_btl_openib_proc_create. This function may be called
from several places simultaneously:
* from the main thread when somebody wants to do `MPI_Send()` (for example) for
the first time;
* from udcm if the counterpart peer is trying to connect and `mca_btl_openib_get_ep()`
is called.
In this case one of the threads may add an uninitialized proc structure
to the `mca_btl_openib_component.ib_procs` and the other will read it and
treat as initialized.
This commit turns ib_proc initialization into a single atomic operation.
It was decided some time ago that there is no benefit to using any
per-peer receive queues on infiniband. At the time we decided not to
change the default but that objection has been dropped. This commit
changes the 128 message queue to use SRQ instead of PP. This has no
impact on iWarp which sets the default in a different way.
Closesopen-mpi/ompi#1156
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
When PtlMDBind was removed, the result check was left in which
causes intermittent failures depending on the junk value found in
the 'ret' variable. The commit removes the result check.
Mofed 2.2 does not have the IBV_EXP_QP_INIT_ATTR_ATOMICS_ARG attribute
flag. Add a check to fix compilation for mofed 2.2. This commit only
fixes complilation with the older mofed. It will not allow an Open MPI
compiled with mofed 2.3 or newer to work on a machine with mofed 2.2.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
If btl-portals4 is configured to use logical mapping of ranks to
physical nodes, then the endpoint must have the rank field set.
This commit fixes a bug that caused the endpoint to have the
nid/pid instead of the rank if the endpoint already exists.
This update adds an additional check (if supported) to see if 8-byte
atomics are supported by the hardware. If 8-byte atomics are not
supported the atomics support is disabled.
This commit also includes some cleanup.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
This commit adds support for fetch-and-add and compare-and-swap when
using the mlx5 driver. The support is only enabled if the expanded
verbs interface is detected. This is required because mlx5 HCAs return
the atomic result in network byte order. This support may need to be
tweaked if Mellanox commits their changes into upstream verbs.
Closesopen-mpi/ompi#1077
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
1. Fix: old v1.6-era code reset the stats-emitting event to fire twice
for each time period.
1. Add the usNIC device name to the output for differentiating the
output in multi-rail scenarios.
This commit fixes the following bugs:
- On send failure release newly allocated message.
- In the destructor for udcm_message_sent_t always remove the send
timeout event from the event base. Failure to do this can lead to
memory corruption since the destructor may be called from an event
callback.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
This was fixed on my btl 3.0 branch but the changeset got lost in a
rebase. Fixes issues with lock ups when using osc/rdma.
References open-mpi/ompi#1010
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
The handling of RDMA get alignment in ugni BTL for Aries
(cray xc) was wrong, resulting in very poor bandwidth
for ugni BTL on aries.
Verified using osu_bw now gives sensible bandwidth on
Aries.
Fixes#1005
Signed-off-by: Howard Pritchard <howardp@lanl.gov>
This commit removes the service and async event threads from the
openib btl. Both threads are replaced by opal progress thread
support. The run_in_main function is now supported by allocating an
event and adding it to the sync event base. This ensures that the
requested function is called as part of opal_progress.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
This commit adds a access_flags argument to the mpool registration
function. This flag indicates what kind of access is being requested:
local write, remote read, remote write, and remote atomic. The values
of the registration access flags in the btl are tied to the new flags
in the mpool. All mpools have been updated to include the new argument
but only the grdma and udreg mpools have been updated to make use of
the access flags. In both mpools existing registrations are checked
for sufficient access before being returned. If a registration does
not contain sufficient access it is marked as invalid and a new
registration is generated.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
In the default mode of operation, the Portals4 components support
dynamic add_procs().
The Portals4 components have two alternate modes (flow control and
logical-to-physical) that require knowledge of all procs at startup.
In these modes, mtl-portals4 sets the MCA_MTL_BASE_FLAG_REQUIRE_WORLD
flag and btl-portals4 sets the MCA_BTL_FLAGS_SINGLE_ADD_PROCS flag
to tell the PML that we need all the procs in one add_procs() call.
After PMIx integration, the thrid parameter to OPAL_MODEX_RECV() is
opal_process_name_t instead of opal_proc_t. This commit replaces
proc with &proc->proc_name.
Fix CID 1312120: Uninitialized scalar variable
The response type will always be set unless a message of another type is passed to this function. To make sure that error is caught I am adding an assert.
Fix CID 1312116: Dereference after null check
This is a potential bug. If there is no endpoint data for an incoming connection a rejection should be sent. In this case we would just SEGV.
Fix CID 1312115: Dereference after null check
Clear error in the error message. Use the queue pair number that was passed in.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
This commit adds support to the pml, mtl, and btl frameworks for
components to indicate at runtime that they do not support the new
dynamic add_procs behavior. At the high end the lack of dynamic
add_procs support is signalled by the pml using the new pml_flags
member to the pml module structure. If the
MCA_PML_BASE_FLAG_REQUIRE_WORLD flag is set MPI_Init will generate the
ompi_proc_t array passed to add_proc from ompi_proc_world () instead
of ompi_proc_get_allocated ().
Both cm and ob1 have been updated to detect if the underlying mtl and
btl components support dynamic add_procs.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
In the v1.10 opal_progress_thread emulation, ensure to create a
blocking libevent timer far in the future. Without that,
opal_event_loop() will return immediately (and therefore the progress
thread spins hard, stealing CPU cycles).
We try to keep the source code the same between master and v1.10. So
put the #if's back for OPAL_HAVE_HWLOC (and just hard-code it to 1 on
master) so that this code is also compilable in v1.10.
This commit makes two changes to the tcp btl:
- If a tcp proc does not exist when handling a new connection create
a new proc and use it. The current implementation uses the
opal_proc_by_name() function to get the opal_proc_t then calls
add_procs on all btl modules. It may be sufficient to just call
add_procs until an endpoint is created so this may change somewhat.
- In add_procs add a check for an existing endpoint before creating
one.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
This commit adds an opal hash table to keep track of mapping between
process identifiers and ompi_proc_t's. This hash table is used by the
ompi_proc_by_name() function to lookup (in O(1) time) a given
process. This can be used by a BTL or other component to get a
ompi_proc_t when handling an incoming message from an as yet unknown
peer.
Additionally, this commit adds a new MCA variable to control the new
add_procs behavior: mpi_add_procs_cutoff. If the number of ranks in
the process falls below the threshold a ompi_proc_t is created for
every process. If the number of ranks is above the threshold then a
ompi_proc_t is only created for the local rank. The code needed to
generate additional ompi_proc_t's for a communicator is not yet
complete.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
This commit updates each non-compliant btl to send the
MCA_BTL_FLAGS_SEND flag in the btl_flags field if send is
supported. This fixes a problem identified after the latest bml/r2
update which excplicitly checks for the send flag.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
Bring Slurm PMI-1 component online
Bring the s2 component online
Little cleanup - let the various PMIx modules set the process name during init, and then just raise it up to the ORTE level. Required as the different PMI environments all pass the jobid in different ways.
Bring the OMPI pubsub/pmi component online
Get comm_spawn working again
Ensure we always provide a cpuset, even if it is NULL
pmix/cray: adjust cray pmix component for pmix
Make changes so cray pmix can work within the integrated
ompi/pmix framework.
Bring singletons back online. Implement the comm_spawn operation using pmix - not tested yet
Cleanup comm_spawn - procs now starting, error in connect_accept
Complete integration
There were several issues preventing the openib btl from running in
thread multiple mode:
- Missing locks in UDCM when generating a loopback endpoint. Fixed in
open-mpi/ompi@8205d79819.
- Incorrect sequence numbers generated in debug mode. This did not
prevent the openib btl from running but instead produced incorrect
error messages in debug builds.
- Recursive locking of the rcache lock caused by the malloc
hooks. This is fixed by open-mpi/ompi#827
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
When using eager RDMA in debug builds the openib btl generates a
sequence number for each send. The code independently updated the head
index and the sequence number for the eager rdma transaction. If
multiple threads enter this code at the same time and run in the
following order:
thread 1: update sequence (0 -> 1)
thread 2: update sequence (1 -> 2)
thread 2: update head (0 -> 1)
thread 1: update head (1 -> 2)
the sequence number for head[0] gets 1 and the sequence number for
head[1] gets 0. The fix is to generate the sequence number from the
head index.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
this test seems broken :
- some false positive were reported
- it fails to detect some OFED version mismatch
this commit simply removes this test, which means the application
will likely fail if XRC is used ad OFED version is different
between compile time and runtime
This code really had no purpose; just assign FI_VERSION(1, 1). This
fixes CID 1315274.
Also clarify the commet about why we still retain libfabric v1.0.0
compatibility code, even though configure.m4 requires libfabric >= v1.1.0.
In short applications, it's possible that the agent (i.e., local rank
0) will finalize after non-local rank 0 procs detect the connectivity
checker named socket, but before they complete a connect() on it. As
such, their connect() gets ECONNREFUSED.
This commit adds a simple counter in the agent that won't let it quit
before it accept()'s from all local procs, or 10 seconds goes by
(whichever occurs first). This is similar to the timeout for the
clients: they'll exit if they don't see the expected named socket
within 10 seconds.
There's no longer any need for the usnic BTL to have its own progress
thread: it can use the opal_progress_thread() infrastructure. This
commit removes the code to startup/shutdown the usnic-BTL-specific
progress thread and instead, just adds its events to the OPAL-wide
progress thread.
This necessitated a small change in the finalization step.
Previously, we would stop the progress thread and then tear down the
events. We can no longer stop the progress thread, and if we start
tearing down events, this will cause shutdown/hangups to be sent
across sockets, potentially firing some of the still-remaining events
while some (but not all) of the data structures have been torn down.
Chaos ensues.
Instead, queue up an event to tear down all the pending events. Since
the progress thread will only fire one event at a time, having a
teardown event means that it can tear down all the pending events
"atomically" and not have to worry that one of those events will get
fired in the middle of the teardown process.
Ensure that we have non-NULL on all levels of pointers, which will
save us if there are exitable errors very early during component /
module initialization.
since ibv_create_xrc_rcv_qp is now deprecated, and in order to
be "future-proof", we have to consider the case in which only XRC Domains are supported.
also, correctly handle distro that ship broken ibverbs devel headers
Thanks Paul Hargrove for the detailled report.
In libfabric v1.0.0 (i.e., API v1.0), the usnic provider handled
FI_MSG_PREFIX inconsistently between sends and receives. This has
been fixed in libfabric v1.1.0 (i.e., API v1.1): FI_MSG_PREFIX is
handled consistently for both sends and receives.
Run-time detect which libfabric we are running with and adapt behavior
appropriately.
Handle the differences between libfabric v1.0.0 and v1.1.0 in the
return value of fi_cq_readerr().
Also consolidate CRC and truncation errors into the same handling
block, since truncation errors are typically another symptom of CRC
errors. This ensures that buffers get reposted properly.
Instead of silently determining that the usnic BTL can't be built,
announce that usnic is checking for libfabric support, and then
AC_MSG_RESULT the result of that check.
@ggouaillardet is likely offline for the weekend, but master is broken
on RHEL 6.5 systems that do not have MOFED installed. So I'm taking
the liberty of revering this commit; I'm guessing Gilles will fixup
and re-commit next week.
This reverts commit 77f8282d51.
since ibv_create_xrc_rcv_qp is now deprecated, and in order to
be "future-proof", we have to consider the case in which only XRC Domains are supported.
Thanks Paul Hargrove for the detailled report.
The usnic BTL configure.m4 no longer needs to OPAL_CHECK_LIBFABRIC; it
just uses the results from opal/mca/common/libfabric's configure.m4.
We also now don't need to link against libfabric -- they just link
against the opal_common_libfabric library.
This commit does two things. It removes checks for C99 required
headers (stdlib.h, string.h, signal.h, etc). Additionally it removes
definitions for required C99 types (intptr_t, int64_t, int32_t, etc).
Signed-off-by: Nathan Hjelm <hjelmn@me.com>
Minor comment/whitespace fixes. Also some minor logic changes that
are mainly for defensive programming purposes (i.e., ensure to always
set malloc_hook_set to true or false, and then check it before we try
to actually invoke it).
Instead of unconditionally setting the memory hook, only set it when
the memory hooks are both available and have been enabled (e.g.,
opal/mca/memory/linux has decided that it *can* be enabled, and when
the mpi_leave_pinned MCA param is set to 1, or is set to -1 and some
component requested the memory hooks be enabled).
If we set the memory hook when memory hooks are not enabled,
__malloc_hook will be NULL, which will cause problems when
btl_openib_malloc_hook() tries to invoke it.
Fixesopen-mpi/ompi#638.
Also add another (superflous but symmetric) continue statement.
This missing "continue" statement allows IPv4 "private network"
matches to fall through and allow IPv6 matches to be made -- thereby
overriding the IPv4 match that was already made.
Fixes#585 (although several of the other issues identified on #585
still exist, the primary / initial bug that was reported there is now
fixed).
CID 1269821 Dereference null return value (NULL_RETURNS)
This is another false positive that can be silenced by looping on
opal_list_remove_first instead of using both opal_list_is_empty and
opal_list_remove_first.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
CID 1301390 Dereference before null check (REVERSE_INULL)
endpoint can not be NULL here. Remove NULL check.
CID 1269836 Unintentional integer overflow (OVERFLOW_BEFORE_WIDEN)
CID 1301388 Bad bit shift operation (BAD_SHIFT)
Add ull to integer constants to ensure the math is done in 64-bits not
32.
CID 715749 Explicit null dereferenced (FORWARD_NULL)
As far as I can tell this parser function does not accept a line that
does match key = value. If that is the case then value should never be
NULL. If it is it is a parse error. Updated the code to reflect
this. Also modified the intify function to do something more sane
(strtol vs atoi with hex detection).
CID 1269820 Dereference null return value (NULL_RETURNS)
This is a false positive as strchr will never return NULL here. It
makes sense, though, to quiet the warning by changing the do {} while
() loop to a while () loop.
CID 1269780 Dereference after null check (FORWARD_NULL)
Just return an error if the endpoint's cpc data is NULL.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
CID 1269674 Ignoring number of bytes read (CHECKED_RETURN)
Check that we read enough bytes to get a complete async command.
CID 1269793 Missing break in switch (MISSING_BREAK)
Added comment to indicate fall through was intentional.
CID 1269702: Constant variable guards dead code (DEADCODE)
Remove an unused argument to opal_show_help. This will quiet the
coverity issue.
CID 1269675 Ignoring number of bytes read (CHECKED_RETURN)
Check that at least sizeof(int) bytes are read. If this is not the
case then it is an error.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
CID 1269931 Uninitialized scalar variable (UNINIT)
Initialize complete async message. This was not a bug but the fix
contributes to valgrind cleanness (uninitialed write).
CID 1269915 Unintended sign extension (SIGN_EXTENSION)
Should never happen. Quieting this by explicitly casting to uint64_t.
CID 1269824 Dereference null return value (NULL_RETURNS)
It is impossible for opal_list_remove_first to return NULL if
opal_list_is_empty returns false. I refactored the code in question to
not use opal_list_is_empty but loop until NULL is returned by
opal_list_remove_first. That will quiet the issue.
CID 1269913 Dereference before null check (REVERSE_INULL)
The storage parameter should never be NULL. The check intended to
check if *storage was NULL not storage.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
CID 1269933 Uninitialized scalar variable (UNINIT)
This CID isn't really an error but it is best for both valgrind and
coverity cleanness to not write uninitialized data. Added an
initializer for async_command in btl_openib_component_close.
CID 1269930 Uninitialized scalar variable (UNINIT)
Same as above. Best not to write uninitialized data. Added an
initializer for async_command.
CID 1269701 Logically dead code (DEADCODE)
Coverity is correct. The smallest_pp_qp will always be 0. Changed the
initial value so that the smallest_pp_qp is set as intended. If no
per-per queue pair exists then use the last shared queue pair. This
queue pair should have the smallest message size. This will reduce
buffer waste.
CID 1269713 Logically dead code (DEADCODE)
False positive but easy to silence. The two check are meaningless if
HAVE_XRC is 0 so protect them with #if HAVE_XRC.
CID 1269726 Division or modulo by zero (DIVIDE_BY_ZERO)
Indeed an issue. If we get an invalid value for rd_win then this will
cause a divide-by-zero exception. Added a check to ensure rd_win is >
0. Also updated the help message to reflect this requirement.
CID 1269672 Ignoring number of bytes read (CHECKED_RETURN)
This error was somewhat intentional. Linux parameter files are
probably not empty but it is safer to check the return code of read to
make sure we got something. If 0 bytes are read this code could SEGV
whe running strtoull.
CID 1269836 Unintentional integer overflow (OVERFLOW_BEFORE_WIDEN)
Add a range check to read_module_param to ensure we do not
overflow. In the future it might be worthwhile to report an error
because these parameters should never cause overflow in this
calculation.
CID 1269692 Calling risky function (DC.WEAK_CRYPTO)
??? This call was added in 2006 but I see no calls to the rest of the
rand48 family of functions. Anyway, we SHOULD NEVER be calling seed48,
srand, etc because it messes with user code. Removed the call to
seed48.
CID 1269823 Dereference null return value (NULL_RETURNS)
This is likely a false positive. The endpoint lock is being held so no
other thread should be able to remove fragments from the list. Also,
mca_btl_openib_endpoint_post_send should not be removing items from
the list. If a NULL fragment is ever returned it will likely be a
coding error on the part of an Open MPI developer. Added an assert()
to catch this and quiet the coverity error.
CID 1269671 Unchecked return value (CHECKED_RETURN)
Added a check for the return code of mca_btl_openib_endpoint_post_send
to quiet the coverity error. It is unlikely this error path will be
traversed.
CID 1270229 Missing break in switch (MISSING_BREAK)
Add a comment to indicate that the fall-through is intentional.
CID 1269735 Dereference after null check (FORWARD_NULL)
There should always be an endpoint when handling a work
completion. The endpoint is either stored on the fragment or can be
looked up using the immediate data. Move the immediate data code up
and add an assert for a NULL endpoint.
CID 1269740 Dereference after null check (FORWARD_NULL)
CID 1269741 Explicit null dereferenced (FORWARD_NULL)
Similar to CID 1269735 fix.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
Followup to open-mpi/ompi@65b66ab: if we're not debugging, then #if
out an entire block so that the compiler doesn't warn about variables
that are assigned and not used.
When using an external libfabric (or really any libfabric newer than
libfabric commit 607e863), we must use fi_getname to determine the local
port of our endpoint. Without this fix, OMPI will hang endlessly
while retransmitting packets to port 0 on the remote host.
