Merge pull request #13467 from janjust/remove-hcoll

coll: Remove hcoll component

Author: bosilca

.ci/mellanox/README.md

@@ -6,7 +6,7 @@
 CI is managed by [Azure Pipelines](https://docs.microsoft.com/en-us/azure/devops/pipelines/?view=azure-devops) service.
 
 Mellanox Open MPI CI includes:
-* Open MPI building with internal stable engineering versions of UCX and HCOLL. The building is run in Docker-based environment.
+* Open MPI building with internal stable engineering versions of UCX. The building is run in Docker-based environment.
 * Sanity functional testing.
 ### How to Run CI
 Mellanox Open MPI CI is triggered upon the following events:

config/ompi_check_libhcoll.m4

@@ -63,7 +63,7 @@ btl=^usnic
 # The following frameworks/components are UNTESTED, and probably won't work.
 # They should run without faults, and will probably crash/deadlock after a fault.
 # You may try at your own risk.
-#   coll    hcoll, portals4
+#   coll    portals4
 #   topo    (all)
 #   osc     (all)
 #   io      (all)
@@ -72,7 +72,7 @@ btl=^usnic
 # We will disable only the components for which good components are known to exist.
 # Other untested components are selectable but will issue a runtime warning at
 # initiation if FT is enabled.
-coll=^hcoll,portals4
+coll=^portals4
 
 #
 # The following frameworks/components are NOT WORKING. Do not enable these with FT.

contrib/amca-param-sets/ft-mpi

@@ -13,7 +13,7 @@ enable_ipv6=no
 enable_man_pages=no
 enable_mpi_fortran=no
 enable_memchecker=no
-enable_mca_no_build=memchecker,coll-adapt,coll-cuda,coll-demo,coll-ftagree,coll-han,coll-hcoll,coll-inter,coll-libnbc,coll-monitoring,coll-portals4,coll-tuned,common-monitoring,common-ompio,fbtl,fcoll,fs,io,mtl,osc,pml-cm,pml-monitoring,pml-ucx,pml-v,sharedfp,topo,vprotocol,btl-ofi,btl-portals4,btl-smcuda,btl-uct,btl-ugni,btl-usnic,common-cuda,common-ofi,common-ucx
+enable_mca_no_build=memchecker,coll-adapt,coll-cuda,coll-demo,coll-ftagree,coll-han,coll-inter,coll-libnbc,coll-monitoring,coll-portals4,coll-tuned,common-monitoring,common-ompio,fbtl,fcoll,fs,io,mtl,osc,pml-cm,pml-monitoring,pml-ucx,pml-v,sharedfp,topo,vprotocol,btl-ofi,btl-portals4,btl-smcuda,btl-uct,btl-ugni,btl-usnic,common-cuda,common-ofi,common-ucx
 enable_contrib_no_build=libompitrace
 with_memory_manager=no
 with_devel_headers=yes

contrib/platform/intel/bend/linux

@@ -43,7 +43,6 @@ created.
       (change S to X; make sure numbers match those for the same entry in
       contrib/platform/lanl/toss/optimized-mlx.conf)
     - addition: pml = ob1 (disable MXM)
-    - addition: coll = ^hcoll (disable MXM)
 - toss3-hfi-optimized
   - copy of toss2-qib-optimized
 - toss3-hfi-optimized.conf

contrib/platform/lanl/toss/README.md

@@ -106,4 +106,3 @@ ras_base_launch_orted_on_hn = true
 
 ## Disable MXM
 pml = ob1
-coll = ^hcoll

contrib/platform/lanl/toss/toss2-mlx-optimized.conf

@@ -22,11 +22,6 @@ if [ "$mellanox_autodetect" == "yes" ]; then
         with_ucx=$ucx_dir
     fi
 
-    hcoll_dir=${hcoll_dir:="$(pkg-config --variable=prefix hcoll)"}
-    if [ -d $hcoll_dir ]; then
-        with_hcoll=$hcoll_dir
-    fi
-
     slurm_dir=${slurm_dir:="/usr"}
     if [ -f $slurm_dir/include/slurm/slurm.h ]; then
         with_slurm=$slurm_dir

contrib/platform/mellanox/optimized

@@ -333,7 +333,7 @@ correctly after a failure.
   * ``cuda``, ``inter``, ``sync``, ``sm``: **untested** (they have not
     been modified to handle faults, but we expect correct post-fault
     behavior)
-  * ``hcoll``, ``portals4`` **disabled** (they have not been modified
+  * ``portals4`` **disabled** (it has not been modified
     to handle faults, and we expect unspecified post-fault behavior)
 
 * ``osc``: MPI one-sided communications

docs/features/ulfm.rst

@@ -14,15 +14,6 @@ can be used with ``configure``:
 
   FCA is the support library for Mellanox switches and HCAs.
 
-* ``--with-hcoll=DIR``:
-  Specify the directory where the Mellanox hcoll library and header
-  files are located.  This option is generally only necessary if the
-  hcoll headers and libraries are not in default compiler/linker
-  search paths.
-
-  hcoll is the support library for MPI collective operation offload on
-  Mellanox ConnectX-3 HCAs (and later).
-
 * ``--with-knem=DIR``:
   Specify the directory where the knem libraries and header files are
   located.  This option is generally only necessary if the knem headers

docs/installing-open-mpi/configure-cli-options/networking.rst

@@ -3,7 +3,7 @@ Tuning Collectives
 
 Open MPI's ``coll`` framework provides a number of components implementing
 collective communication, including: ``han``, ``libnbc``, ``self``, ``ucc`` ``base``,
-``hcoll``, ``sync``, ``xhc``, ``accelerator``, ``basic``, ``ftagree``, ``inter``, ``portals4``,
+``sync``, ``xhc``, ``accelerator``, ``basic``, ``ftagree``, ``inter``, ``portals4``,
 and ``tuned``.  Some of these components may not be available depending on how
 Open MPI was compiled and what hardware is available on the system.  A run-time
 decision based on each component's self reported priority, selects which