Merge pull request #7594 from raffenet/fixes-for-rc2

[4.3.x] cherry-picked fixes for 4.3.2rc2

Approved-by: Hui Zhou <[email protected]>

Author: raffenet

CHANGES

@@ -25,6 +25,8 @@
 
 # Fix crash with GPU-aware build when running on systems with no GPUs
 
+# Fix HIP device query
+
 # Fix singleton init with Hydra
 
 # Fix thread safety for Level Zero memcpy functions

autogen.sh

@@ -570,6 +570,7 @@ autoreconf_amdir() {
         _patch_successful=no
 
         _patch_libtool $_dir/confdb/ltmain.sh  intel-compiler.patch
+        _patch_libtool $_dir/confdb/ltmain.sh  nagfor.patch
         _patch_libtool $_dir/confdb/libtool.m4 sys_lib_dlsearch_path_spec.patch
         _patch_libtool $_dir/confdb/libtool.m4 big-sur.patch
         _patch_libtool $_dir/confdb/libtool.m4 hpc-sdk.patch

maint/patches/optional/confdb/nagfor.patch

@@ -0,0 +1,15 @@
+--- confdb/ltmain.sh	2025-09-23 13:54:31.782898806 -0500
++++ confdb/ltmain.sh.new	2025-09-23 17:06:28.926502645 -0500
+@@ -7991,7 +7991,11 @@
+ 	      func_fatal_error "cannot find name of link library for '$lib'"
+ 	    fi
+ 	    # It is a libtool convenience library, so add in its objects.
+-	    func_append convenience " $ladir/$objdir/$old_library"
++            if test "$wl" = "-Wl,-Wl,," ; then
++                func_append convenience " -Wl,$ladir/$objdir/$old_library"
++            else
++                func_append convenience " $ladir/$objdir/$old_library"
++            fi
+ 	    func_append old_convenience " $ladir/$objdir/$old_library"
+ 	    tmp_libs=
+ 	    for deplib in $dependency_libs; do

maint/version.m4

@@ -14,7 +14,7 @@
 # changing this by playing with diversions, but then we would probably be
 # playing with autotools-fire.
 
-m4_define([MPICH_VERSION_m4],[4.3.2rc1])dnl
+m4_define([MPICH_VERSION_m4],[4.3.2rc2])dnl
 m4_define([MPICH_RELEASE_DATE_m4],[unreleased development copy])dnl
 
 # For libtool ABI versioning rules see:

src/binding/abi/mpi_abi.h

@@ -1001,9 +1001,9 @@ int MPI_Remove_error_code(int errorcode);
 int MPI_Remove_error_string(int errorcode);
 int MPI_Request_free(MPI_Request *request);
 int MPI_Request_get_status(MPI_Request request, int *flag, MPI_Status *status);
-int MPI_Request_get_status_all(int count, MPI_Request array_of_requests[], int *flag, MPI_Status *array_of_statuses);
-int MPI_Request_get_status_any(int count, MPI_Request array_of_requests[], int *indx, int *flag, MPI_Status *status);
-int MPI_Request_get_status_some(int incount, MPI_Request array_of_requests[], int *outcount, int array_of_indices[], MPI_Status *array_of_statuses);
+int MPI_Request_get_status_all(int count, const MPI_Request array_of_requests[], int *flag, MPI_Status *array_of_statuses);
+int MPI_Request_get_status_any(int count, const MPI_Request array_of_requests[], int *indx, int *flag, MPI_Status *status);
+int MPI_Request_get_status_some(int incount, const MPI_Request array_of_requests[], int *outcount, int array_of_indices[], MPI_Status *array_of_statuses);
 int MPI_Rget(void *origin_addr, int origin_count, MPI_Datatype origin_datatype, int target_rank, MPI_Aint target_disp, int target_count, MPI_Datatype target_datatype, MPI_Win win, MPI_Request *request);
 int MPI_Rget_c(void *origin_addr, MPI_Count origin_count, MPI_Datatype origin_datatype, int target_rank, MPI_Aint target_disp, MPI_Count target_count, MPI_Datatype target_datatype, MPI_Win win, MPI_Request *request);
 int MPI_Rget_accumulate(const void *origin_addr, int origin_count, MPI_Datatype origin_datatype, void *result_addr, int result_count, MPI_Datatype result_datatype, int target_rank, MPI_Aint target_disp, int target_count, MPI_Datatype target_datatype, MPI_Op op, MPI_Win win, MPI_Request *request);
@@ -1057,9 +1057,9 @@ int MPI_Ssend_init(const void *buf, int count, MPI_Datatype datatype, int dest,
 int MPI_Ssend_init_c(const void *buf, MPI_Count count, MPI_Datatype datatype, int dest, int tag, MPI_Comm comm, MPI_Request *request);
 int MPI_Start(MPI_Request *request);
 int MPI_Startall(int count, MPI_Request array_of_requests[]);
-int MPI_Status_get_error(MPI_Status *status, int *error);
-int MPI_Status_get_source(MPI_Status *status, int *source);
-int MPI_Status_get_tag(MPI_Status *status, int *tag);
+int MPI_Status_get_error(const MPI_Status *status, int *error);
+int MPI_Status_get_source(const MPI_Status *status, int *source);
+int MPI_Status_get_tag(const MPI_Status *status, int *tag);
 int MPI_Status_set_cancelled(MPI_Status *status, int flag);
 int MPI_Status_set_elements(MPI_Status *status, MPI_Datatype datatype, int count);
 int MPI_Status_set_elements_c(MPI_Status *status, MPI_Datatype datatype, MPI_Count count);
@@ -1673,9 +1673,9 @@ int PMPI_Remove_error_code(int errorcode);
 int PMPI_Remove_error_string(int errorcode);
 int PMPI_Request_free(MPI_Request *request);
 int PMPI_Request_get_status(MPI_Request request, int *flag, MPI_Status *status);
-int PMPI_Request_get_status_all(int count, MPI_Request array_of_requests[], int *flag, MPI_Status *array_of_statuses);
-int PMPI_Request_get_status_any(int count, MPI_Request array_of_requests[], int *indx, int *flag, MPI_Status *status);
-int PMPI_Request_get_status_some(int incount, MPI_Request array_of_requests[], int *outcount, int array_of_indices[], MPI_Status *array_of_statuses);
+int PMPI_Request_get_status_all(int count, const MPI_Request array_of_requests[], int *flag, MPI_Status *array_of_statuses);
+int PMPI_Request_get_status_any(int count, const MPI_Request array_of_requests[], int *indx, int *flag, MPI_Status *status);
+int PMPI_Request_get_status_some(int incount, const MPI_Request array_of_requests[], int *outcount, int array_of_indices[], MPI_Status *array_of_statuses);
 int PMPI_Rget(void *origin_addr, int origin_count, MPI_Datatype origin_datatype, int target_rank, MPI_Aint target_disp, int target_count, MPI_Datatype target_datatype, MPI_Win win, MPI_Request *request);
 int PMPI_Rget_c(void *origin_addr, MPI_Count origin_count, MPI_Datatype origin_datatype, int target_rank, MPI_Aint target_disp, MPI_Count target_count, MPI_Datatype target_datatype, MPI_Win win, MPI_Request *request);
 int PMPI_Rget_accumulate(const void *origin_addr, int origin_count, MPI_Datatype origin_datatype, void *result_addr, int result_count, MPI_Datatype result_datatype, int target_rank, MPI_Aint target_disp, int target_count, MPI_Datatype target_datatype, MPI_Op op, MPI_Win win, MPI_Request *request);
@@ -1729,9 +1729,9 @@ int PMPI_Ssend_init(const void *buf, int count, MPI_Datatype datatype, int dest,
 int PMPI_Ssend_init_c(const void *buf, MPI_Count count, MPI_Datatype datatype, int dest, int tag, MPI_Comm comm, MPI_Request *request);
 int PMPI_Start(MPI_Request *request);
 int PMPI_Startall(int count, MPI_Request array_of_requests[]);
-int PMPI_Status_get_error(MPI_Status *status, int *error);
-int PMPI_Status_get_source(MPI_Status *status, int *source);
-int PMPI_Status_get_tag(MPI_Status *status, int *tag);
+int PMPI_Status_get_error(const MPI_Status *status, int *error);
+int PMPI_Status_get_source(const MPI_Status *status, int *source);
+int PMPI_Status_get_tag(const MPI_Status *status, int *tag);
 int PMPI_Status_set_cancelled(MPI_Status *status, int flag);
 int PMPI_Status_set_elements(MPI_Status *status, MPI_Datatype datatype, int count);
 int PMPI_Status_set_elements_c(MPI_Status *status, MPI_Datatype datatype, MPI_Count count);

src/mpl/src/gpu/mpl_gpu_hip.c

@@ -50,6 +50,7 @@ int MPL_gpu_get_dev_list(int *dev_count, char ***dev_list, bool is_subdev)
         snprintf(device_list[i], str_len + 1, "%d", i);
     }
 
+    *dev_count = device_count;
     *dev_list = device_list;
     return ret;
 }