Several small improvements

* Makesure distclean can cleanup everything in mpi-proxy-split
* Check error returns
* Make some functions static when applicable
* Free pointers in `g_async_calls`
* Remove unsued function & variable definitions

Author: jungan

contrib/Makefile.in

@@ -26,7 +26,7 @@ LDFLAGS += -m32 -Wl,-m32 -Wl,-melf_i386 -Wa,--32
 endif
 
 plugins =
-plugins_distclean = ckptfile infiniband
+plugins_distclean = ckptfile infiniband mpi-proxy-split
 
 ifeq ($(INFINIBAND_SUPPORT),yes)
   plugins += infiniband  # ib2tcp

contrib/mpi-proxy-split/mpi-wrappers/.gitignore

@@ -0,0 +1,2 @@
+mana_p2p_update_logs
+p2p-deterministic.h

contrib/mpi-proxy-split/p2p_drain_send_recv.cpp

@@ -97,8 +97,11 @@ recvMsgIntoInternalBuffer(MPI_Status status, MPI_Comm comm)
   MPI_Type_size(MPI_BYTE, &size);
   JASSERT(size == 1);
   void *buf = JALLOC_HELPER_MALLOC(count);
-  MPI_Recv(buf, count, MPI_BYTE, status.MPI_SOURCE, status.MPI_TAG,
+  int rc = MPI_Recv(buf, count, MPI_BYTE, status.MPI_SOURCE, status.MPI_TAG,
            comm, MPI_STATUS_IGNORE);
+  if (rc != MPI_SUCCESS) {
+    fprintf(stderr, "MPI_Recv failed\n");
+  }
 
   mpi_message_t *message = (mpi_message_t *)JALLOC_HELPER_MALLOC(sizeof(mpi_message_t));
   message->buf        = buf;
@@ -116,7 +119,7 @@ recvMsgIntoInternalBuffer(MPI_Status status, MPI_Comm comm)
 
 // Go through each MPI_Irecv in the g_async_calls map and try to complete
 // the MPI_Irecv before checkpointing.
-int
+static int
 completePendingIrecvs()
 {
   int bytesReceived = 0;
@@ -135,6 +138,7 @@ completePendingIrecvs()
                                               call->comm);
         g_recvBytesByRank[worldRank] += call->count * size;
         bytesReceived += call->count * size;
+        JALLOC_HELPER_FREE(call);
         it = g_async_calls.erase(it);
       } else {
         /*  We update the iterator even if the MPI_Test fails.
@@ -162,7 +166,7 @@ completePendingIrecvs()
   return bytesReceived;
 }
 
-int
+static int
 recvFromAllComms()
 {
   int bytesReceived = 0;
@@ -190,17 +194,16 @@ recvFromAllComms()
   return bytesReceived;
 }
 
-void
+static void
 removePendingSendRequests()
 {
   dmtcp::map<MPI_Request, mpi_async_call_t*>::iterator it;
   for (it = g_async_calls.begin(); it != g_async_calls.end();) {
     MPI_Request request = it->first;
     mpi_async_call_t *call = it->second;
-    int flag = 0;
     if (call->type == ISEND_REQUEST) {
       UPDATE_REQUEST_MAP(request, MPI_REQUEST_NULL);
-      // FIXME: We should free `call' to avoid memory leak
+      JALLOC_HELPER_FREE(call);
       it = g_async_calls.erase(it);
     } else {
       it++;

contrib/mpi-proxy-split/p2p_drain_send_recv.h

@@ -37,14 +37,12 @@ extern dmtcp::vector<mpi_message_t*> g_message_queue;
 void initialize_drain_send_recv();
 void registerLocalSendsAndRecvs();
 void drainSendRecv();
-int recvFromAllComms(int source);
 int recvMsgIntoInternalBuffer(MPI_Status status);
 bool isBufferedPacket(int source, int tag, MPI_Comm comm, int *flag,
                       MPI_Status *status);
 int consumeBufferedPacket(void *buf, int count, MPI_Datatype datatype,
                           int source, int tag, MPI_Comm comm,
                           MPI_Status *mpi_status, int size);
-void removePendingSendRequests();
 void resetDrainCounters();
 int localRankToGlobalRank(int localRank, MPI_Comm localComm);
 #endif

contrib/mpi-proxy-split/p2p_log_replay.cpp

@@ -128,7 +128,6 @@ logRequestInfo(MPI_Request request, mpi_req_t req_type)
 request_info_t*
 lookupRequestInfo(MPI_Request request)
 {
-  request_info_t *req_info;
   std::unordered_map<MPI_Request, request_info_t*>::iterator it;
   it = request_log.find(request);
   if (it != request_log.end()) {

contrib/mpi-proxy-split/split_process.cpp

@@ -247,6 +247,7 @@ read_lh_proxy_bits(pid_t childpid)
   // NOTE:  This requires same privilege as ptrace_attach (valid for child).
   //        Anecdotally, in containers, we've seen a case where this errors out
   //        with ESRCH (no such proc.); it may need CAP_SYS_PTRACE privilege??
+  //TODO: why use loop since process_vm_readv can support passing array?
   for (int i = 0; i < IOV_SZ; i++) {
     JTRACE("Reading segment from lh_proxy")
           (remote_iov[i].iov_base)(remote_iov[i].iov_len);