Accepting request 986553 from science:HPC

OBS-URL: https://build.opensuse.org/request/show/986553
OBS-URL: https://build.opensuse.org/package/show/openSUSE:Factory/mvapich2?expand=0&rev=31

_constraints

@@ -0,0 +1,11 @@
+<constraints>
+  <hardware>
+    <processors>8</processors>
+    <disk>
+      <size unit="G">10</size>
+    </disk>
+    <physicalmemory>
+      <size unit="G">16</size>
+    </physicalmemory>
+  </hardware>
+</constraints>

mvapich2-2.3.6.tar.gz

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:b3a62f2a05407191b856485f99da05f5e769d6381cd63e2fcb83ee98fc46a249
-size 21016727

mvapich2.changes

@@ -1,3 +1,49 @@
+-------------------------------------------------------------------
+Wed Jun 29 12:54:33 UTC 2022 - Klaus Kämpf <kkaempf@suse.com>
+
+- add pass-correct-size-to-snprintf.patch to fix potential buffer
+  overflows (required to make 'sundials' testsuite pass)
+
+- Update to mvapich2 2.3.7
+  * Features and Enhancements (since 2.3.6):
+    - Added support for systems with Rockport's switchless networks
+      * Added automatic architecture detection
+      * Optimized performance for point-to-point operations
+    - Added support for the Cray Slingshot 10 interconnect
+    - Enhanced support for blocking collective offload using
+      Mellanox SHARP
+        * Scatter and Scatterv
+    - Enhanced support for non-blocking collective offload using
+      Mellanox SHARP
+        * Iallreduce, Ibarrier, Ibcast, and Ireduce
+
+  * Bug Fixes (since 2.3.6):
+    - Removed several deprectated functions
+      - Thanks to Honggang Li @RedHat for the report
+    - Fixed a bug where tools like CMake FindMPI would not
+      detect MVAPICH  when compiled without Hydra mpiexec
+      - Thanks to Chris Chambreau and Adam Moody @LLNL for the report
+    - Fixed compilation error when building with mpirun and without hydra
+      - Thanks to James Long @University of Illinois for the report
+    - Fixed issue with setting RoCE mode correctly without RDMA_CM.
+      - Thanks to Nicolas Gagnon @Rockport Networks for the report
+    - Fixed an issue on heterogeneous clusters where QP attributes were
+      set incorrectly
+      - Thanks to X-ScaleSolutions for the report and fix
+    - Fixed a memory leak in improbe on the PSM channel
+      - Thanks to Gregory Lee @LLNL Beichuan Yan @University of Colorado
+        for the report
+    - Added retry logic for PSM connection establishment
+      - Thanks to Gregory Lee @LLNL for the report and X-ScaleSolutions
+        for the patch
+    - Fixed an initialization error when using PSM and gcc's -pg option
+      - Thanks to Gregory Lee @LLNL for the report and X-ScaleSolutions for 
+        the patch
+    - Fixed a potential integer overflow when transfering large arrays
+      - Thanks to Alexander Melnikov for the report and patch  
+
+- Fix Url: link
+
 -------------------------------------------------------------------
 Wed Feb 16 11:07:12 UTC 2022 - Nicolas Morey-Chaisemartin <nmoreychaisemartin@suse.com>
 

mvapich2.spec

@@ -19,8 +19,9 @@
 %global flavor @BUILD_FLAVOR@%{nil}
 
 %define pname mvapich2
-%define vers  2.3.6
+%define vers  2.3.7
-%define _vers 2_3_6
+%define _vers 2_3_7
+%define rc_ver -1
 
 %if "%{flavor}" == ""
 ExclusiveArch:  do_not_build
@@ -241,7 +242,7 @@ License:        BSD-3-Clause
 Group:          Development/Libraries/Parallel
 Version:        %{vers}
 Release:        0
-Source0:        http://mvapich.cse.ohio-state.edu/download/mvapich/mv2/mvapich2-%{version}.tar.gz
+Source0:        http://mvapich.cse.ohio-state.edu/download/mvapich/mv2/mvapich2-%{version}%{?rc_ver}.tar.gz
 Source1:        mpivars.sh
 Source2:        mpivars.csh
 Source3:        macros.hpc-mvapich2
@@ -252,13 +253,14 @@ Patch2:         mvapich2-arm-support.patch
 # It's been merged upstream, should be removed with the next release
 Patch3:         0001-Drop-GCC-check.patch
 Patch4:         reproducible.patch
+Patch5:         pass-correct-size-to-snprintf.patch
 
 ## Armv7 specific patches
 # PATCH-FIX-UPSTREAM 0001-Drop-real128.patch (https://github.com/pmodels/mpich/issues/4005)
 Patch50:        0001-Drop-real128.patch
 Patch51:        0001-Drop-Real-16.patch
 
-URL:            http://mvapich.cse.ohio-state.edu/overview/mvapich2/
+URL:            http://mvapich.cse.ohio-state.edu
 BuildRoot:      %{_tmppath}/%{name}-%{version}-build
 
 %if %{without skip_hpc_build}
@@ -389,6 +391,7 @@ is based on MPICH2 and MVICH. This package contains the static libraries
 %patch2
 %patch3
 %patch4
+%patch5 -p1
 
 # Only apply these patches on Armv7
 %ifarch armv7hl

pass-correct-size-to-snprintf.patch

@@ -0,0 +1,103 @@
+diff -wruN -x '*~' -x '*.o' -x '*.a' -x '*.so' -x '*.so.[0-9]' -x autom4te.cache -x .deps -x .libs ../orig-mvapich2-2.3.7-1/src/mpid/ch3/channels/common/src/affinity/hwloc_bind.c ./src/mpid/ch3/channels/common/src/affinity/hwloc_bind.c
+--- ../orig-mvapich2-2.3.7-1/src/mpid/ch3/channels/common/src/affinity/hwloc_bind.c	2022-05-16 18:58:22.000000000 +0200
++++ ./src/mpid/ch3/channels/common/src/affinity/hwloc_bind.c	2022-06-29 15:07:17.700058168 +0200
+@@ -2107,7 +2107,7 @@
+         for (i = 0; i < g_smpi.num_local_nodes; ++i) {
+             hwloc_bitmap_clr(*free_sock_cpuset, local_core_ids[i]);
+         }
+-        hwloc_bitmap_snprintf(cpu_str, 128, *free_sock_cpuset);
++        hwloc_bitmap_snprintf(cpu_str, sizeof(cpu_str), *free_sock_cpuset);
+         PRINT_DEBUG(DEBUG_INIT_verbose, "Free sock_cpuset = %s\n", cpu_str);
+     }
+ 
+@@ -3190,11 +3190,11 @@
+         for (i = 0; i < local_procs; i++) {
+             curr = count;
+             for (k = 0; k < num_app_threads; k++) {
+-                j += snprintf (mapping+j, _POSIX2_LINE_MAX, "%d,", mv2_core_map[curr]);
++                j += snprintf (mapping+j, sizeof(mapping)-j, "%d,", mv2_core_map[curr]);
+                 curr = (curr + 1) % num_pu;
+             }
+             mapping [--j] = '\0'; 
+-            j += snprintf (mapping+j, _POSIX2_LINE_MAX, ":");
++            j += snprintf (mapping+j, sizeof(mapping)-j, ":");
+             count = (count + hw_threads_per_core) % num_pu;
+         }
+     } else if (mv2_hybrid_binding_policy == HYBRID_LINEAR) {
+@@ -3203,14 +3203,14 @@
+          * resources  */
+         for (i = 0; i < local_procs; i++) {
+             for (k = 0; k < num_app_threads; k++) {
+-                j += snprintf (mapping+j, _POSIX2_LINE_MAX, "%d,", mv2_core_map[curr]);
++                j += snprintf (mapping+j, sizeof(mapping)-j, "%d,", mv2_core_map[curr]);
+ 
+                 curr = ((curr + hw_threads_per_core) >= num_pu) ?
+                             ((curr + hw_threads_per_core+ ++step) % num_pu) :
+                             (curr + hw_threads_per_core) % num_pu;
+             }
+             mapping [--j] = '\0';
+-            j += snprintf (mapping+j, _POSIX2_LINE_MAX, ":");
++            j += snprintf (mapping+j, sizeof(mapping)-j, ":");
+         }    
+     } else if (mv2_hybrid_binding_policy == HYBRID_SPREAD) {
+ #if defined(CHANNEL_MRAIL)
+@@ -3232,12 +3232,12 @@
+             for (i = 0; i < local_procs; i++) {
+                  for (k = curr; k < curr+chunk; k++) {
+                      for (l = 0; l < hw_threads_per_core; l++) {
+-                        j += snprintf (mapping+j, _POSIX2_LINE_MAX, "%d,", 
++                        j += snprintf (mapping+j, sizeof(mapping)-j, "%d,", 
+                                 mv2_core_map[k * hw_threads_per_core + l]);
+                      }
+                  }
+                  mapping [--j] = '\0';
+-                 j += snprintf (mapping+j, _POSIX2_LINE_MAX, ":");
++                 j += snprintf (mapping+j, sizeof(mapping)-j, ":");
+                  curr = (curr + chunk) % size;
+             } 
+         } else {
+@@ -3252,11 +3252,11 @@
+             for (i = 0; i < num_sockets; i++) {
+                 for (k = curr; k < curr+ranks_per_sock; k++) {
+                     for (l = 0; l < hw_threads_per_core; l++) {
+-                        j += snprintf (mapping+j, _POSIX2_LINE_MAX, "%d,",
++                        j += snprintf (mapping+j, sizeof(mapping)-j, "%d,",
+                                 mv2_core_map[k * hw_threads_per_core + l]);
+                     }
+                     mapping [--j] = '\0';
+-                    j += snprintf (mapping+j, _POSIX2_LINE_MAX, ":");
++                    j += snprintf (mapping+j, sizeof(mapping)-j, ":");
+                 }
+                 curr = (curr + ((num_pu_per_socket/hw_threads_per_core)  * chunk)) % size;
+             }
+@@ -3265,7 +3265,7 @@
+         /* Bunch mapping: Bind each MPI rank to a single phyical core of first
+          * socket followed by second secket */
+         for (i = 0; i < local_procs; i++) {
+-            j += snprintf (mapping+j, _POSIX2_LINE_MAX, "%d:", mv2_core_map[k]);
++            j += snprintf (mapping+j, sizeof(mapping)-j, "%d:", mv2_core_map[k]);
+             k = (k + hw_threads_per_core) % size;
+         } 
+     } else if (mv2_hybrid_binding_policy == HYBRID_SCATTER) {
+@@ -3283,7 +3283,7 @@
+             return MPI_ERR_OTHER;
+         }
+         for (i = 0; i < local_procs; i++) {
+-            j += snprintf (mapping+j, _POSIX2_LINE_MAX, "%d:", mv2_core_map[k]);
++            j += snprintf (mapping+j, sizeof(mapping)-j, "%d:", mv2_core_map[k]);
+             k = (i % num_sockets == 0) ?
+                     (k + num_pu_per_socket) % size :
+                     (k + num_pu_per_socket + hw_threads_per_core) % size;
+@@ -3315,10 +3315,10 @@
+         /* NUMA mapping: Bind consecutive MPI ranks to different NUMA domains in
+          * round-robin fashion. */
+         for (i = 0; i < local_procs; i++) {
+-            j += snprintf (mapping+j, _POSIX2_LINE_MAX, "%d,", 
++            j += snprintf (mapping+j, sizeof(mapping)-j, "%d,", 
+                                mv2_core_map_per_numa[node_base_pu+node_offset]);
+             mapping [--j] = '\0';
+-            j += snprintf (mapping+j, _POSIX2_LINE_MAX, ":");
++            j += snprintf (mapping+j, sizeof(mapping)-j, ":");
+             node_base_pu = (node_base_pu + num_pu_per_numanode) % size;
+             node_offset = (node_base_pu == 0) ? 
+                             (node_offset + ((hw_threads_per_core > 0) ? hw_threads_per_core : 1)) :