NDT-DEV email list created

Revision: 348
Author: rcarlson501
Date: Wed Apr 21 17:45:18 2010
Log: Improve error handling when communicating with child.  Use select()  
with
a timer to prevent child processes from hanging in an accept() state.
the time will now expire and the child will return a error flag back to the  
parent
process.

Also handle write errors and terminate if a non EINTR error is encountered
while processing a write() call.

bumped version number to 3.6.3

RAC 4/21/10

http://code.google.com/p/ndt/source/detail?r=348

Modified:
 /trunk/Applet/Tcpbw100.java
 /trunk/configure
 /trunk/configure.ac
 /trunk/src/testoptions.c
 /trunk/src/web100srv.c

=======================================
--- /trunk/Applet/Tcpbw100.java Fri Apr  9 09:49:47 2010
+++ /trunk/Applet/Tcpbw100.java Wed Apr 21 17:45:18 2010
@@ -98,7 +98,7 @@

 public class Tcpbw100 extends JApplet implements ActionListener
 {
-  private static final String VERSION = "3.6.2b";
+  private static final String VERSION = "3.6.3";
   private static final byte TEST_MID = (1 << 0);
   private static final byte TEST_C2S = (1 << 1);
   private static final byte TEST_S2C = (1 << 2);
=======================================
--- /trunk/configure    Fri Apr  9 10:34:54 2010
+++ /trunk/configure    Wed Apr 21 17:45:18 2010
@@ -1,6 +1,6 @@
 #! /bin/sh
 # Guess values for system-dependent variables and create Makefiles.
-# Generated by GNU Autoconf 2.63 for NDT 3.6.2b.
+# Generated by GNU Autoconf 2.63 for NDT 3.6.3.
 #
 # Report bugs to 
<>.
 #
@@ -596,8 +596,8 @@
 # Identity of this package.
 PACKAGE_NAME='NDT'
 PACKAGE_TARNAME='ndt'
-PACKAGE_VERSION='3.6.2b'
-PACKAGE_STRING='NDT 3.6.2b'
+PACKAGE_VERSION='3.6.3'
+PACKAGE_STRING='NDT 3.6.3'
 
'

 ac_unique_file="src/analyze.c"
@@ -1331,7 +1331,7 @@
   # Omit some internal or obsolete options to make the list less imposing.
   # This message is too long to be a string in the A/UX 3.1 sh.
   cat <<_ACEOF
-\`configure' configures NDT 3.6.2b to adapt to many kinds of systems.
+\`configure' configures NDT 3.6.3 to adapt to many kinds of systems.

 Usage: $0 [OPTION]... [VAR=VALUE]...

@@ -1397,7 +1397,7 @@

 if test -n "$ac_init_help"; then
   case $ac_init_help in
-     short | recursive ) echo "Configuration of NDT 3.6.2b:";;
+     short | recursive ) echo "Configuration of NDT 3.6.3:";;
    esac
   cat <<\_ACEOF

@@ -1495,7 +1495,7 @@
 test -n "$ac_init_help" && exit $ac_status
 if $ac_init_version; then
   cat <<\_ACEOF
-NDT configure 3.6.2b
+NDT configure 3.6.3
 generated by GNU Autoconf 2.63

 Copyright (C) 1992, 1993, 1994, 1995, 1996, 1998, 1999, 2000, 2001,
@@ -1509,7 +1509,7 @@
 This file contains any messages produced by compilers while
 running configure, to aid debugging if configure makes a mistake.

-It was created by NDT $as_me 3.6.2b, which was
+It was created by NDT $as_me 3.6.3, which was
 generated by GNU Autoconf 2.63.  Invocation command line was

   $ $0 $@
@@ -2359,7 +2359,7 @@

 # Define the identity of the package.
  PACKAGE='ndt'
- VERSION='3.6.2b'
+ VERSION='3.6.3'


 cat >>confdefs.h <<_ACEOF
@@ -8645,7 +8645,7 @@
 # report actual input values of CONFIG_FILES etc. instead of their
 # values after options handling.
 ac_log="
-This file was extended by NDT $as_me 3.6.2b, which was
+This file was extended by NDT $as_me 3.6.3, which was
 generated by GNU Autoconf 2.63.  Invocation command line was

   CONFIG_FILES    = $CONFIG_FILES
@@ -8708,7 +8708,7 @@
 _ACEOF
 cat >>$CONFIG_STATUS <<_ACEOF || ac_write_fail=1
 ac_cs_version="\\
-NDT config.status 3.6.2b
+NDT config.status 3.6.3
 configured by $0, generated by GNU Autoconf 2.63,
   with options \\"`$as_echo "$ac_configure_args" | sed 's/^ //;  
s/[\\""\`\$]/\\\\&/g'`\\"

=======================================
--- /trunk/configure.ac Fri Apr  9 10:34:54 2010
+++ /trunk/configure.ac Wed Apr 21 17:45:18 2010
@@ -29,7 +29,7 @@
 # Process this file with autoconf to produce a configure script.
 #
 # AC_PREREQ(2.57)
-AC_INIT([NDT],[3.6.2b],[])
+AC_INIT([NDT],[3.6.3],[])
 AC_CONFIG_AUX_DIR(config)
 # AM_INIT_AUTOMAKE(NDT, v1.0, [no-define])
 AM_INIT_AUTOMAKE
=======================================
--- /trunk/src/testoptions.c    Fri Apr  9 09:23:03 2010
+++ /trunk/src/testoptions.c    Wed Apr 21 17:45:18 2010
@@ -263,7 +263,7 @@
   int maxseg=1456;
   /* int maxseg=1456, largewin=16*1024*1024; */
   /* int seg_size, win_size; */
-  int midfd, j, ret;
+  int midsfd, j, ret;
   struct sockaddr_storage cli_addr;
   /* socklen_t optlen, clilen; */
   socklen_t clilen;
@@ -274,6 +274,8 @@
   int msgLen;
   web100_connection* conn;
   char tmpstr[256];
+  struct timeval sel_tv;
+  fd_set rfd;

   assert(ctlsockfd != -1);
   assert(agent);
@@ -359,33 +361,47 @@
      * and does NAT detection.  More analysis functions (window scale)
      * will be done in the future.
      */
-    j = 0;
     clilen = sizeof(cli_addr);
-    for (;;) {
-      if ((midfd = accept(options->midsockfd, (struct sockaddr *)  
&cli_addr, &clilen)) > 0)
+    FD_ZERO(&rfd);
+    FD_SET(options->midsockfd, &rfd);
+    sel_tv.tv_sec = 5;
+    sel_tv.tv_usec = 0;
+    for (j=0; j<5; j++) {
+      ret = select((options->midsockfd)+1, &rfd, NULL, NULL, &sel_tv);
+      if ((ret == -1) && (errno == EINTR))
+       continue;
+      if (ret == 0)
+       return -100;  /* timeout */
+      if (ret < 0)
+       return -errno;
+      if (j == 4)
+       return -101;
+midfd:
+      if ((midsfd = accept(options->midsockfd, (struct sockaddr *)  
&cli_addr, &clilen)) > 0)
        break;

-      if ((midfd == -1) && (errno == EINTR))
-       continue;
+      if ((midsfd == -1) && (errno == EINTR))
+       goto midfd;

       sprintf(tmpstr, "-------     middlebox connection setup returned  
because (%d)", errno);
       if (get_debuglvl() > 1)
         perror(tmpstr);
-      if (++j == 4)
-        /* break; */
-       return -2;
+      if (midsfd < 0)
+       return -errno;
+      if (j == 4)
+       return -102;
     }
     memcpy(&meta.c_addr, &cli_addr, clilen);
     /* meta.c_addr = cli_addr; */
     meta.family = ((struct sockaddr *) &cli_addr)->sa_family;

     buff[0] = '\0';
-    if ((conn = web100_connection_from_socket(agent, midfd)) == NULL) {
+    if ((conn = web100_connection_from_socket(agent, midsfd)) == NULL) {
         log_println(0, "!!!!!!!!!!!  test_mid() failed to get web100  
connection data, rc=%d", errno);
         /* exit(-1); */
        return -3;
     }
-    web100_middlebox(midfd, agent, conn, buff);
+    web100_middlebox(midsfd, agent, conn, buff);
     send_msg(ctlsockfd, TEST_MSG, buff, strlen(buff));
     msgLen = sizeof(buff);
     if (recv_msg(ctlsockfd, &msgType, buff, &msgLen)) {
@@ -409,8 +425,8 @@
     *s2c2spd = atof(buff);
     log_println(4, "CWND limited throughput = %0.0f kbps (%s)", *s2c2spd,  
buff);

-    shutdown(midfd, SHUT_WR);
-    close(midfd);
+    shutdown(midsfd, SHUT_WR);
+    close(midsfd);
     close(options->midsockfd);
     send_msg(ctlsockfd, TEST_FINALIZE, "", 0);
     log_println(1, " <--------- %d ----------->", options->child0);
@@ -535,9 +551,22 @@
        return ret;

     clilen = sizeof(cli_addr);
-    /* j = 0; */
     log_println(6, "child %d - sent c2s prepare to client",  
testOptions->child0);
+    FD_ZERO(&rfd);
+    FD_SET(testOptions->c2ssockfd, &rfd);
+    sel_tv.tv_sec = 5;
+    sel_tv.tv_usec = 0;
     for (j=0; j<5; j++) {
+      ret = select((testOptions->c2ssockfd)+1, &rfd, NULL, NULL, &sel_tv);
+      if ((ret == -1) && (errno == EINTR))
+       continue;
+      if (ret == 0)
+       return -100;  /* timeout */
+      if (ret < 0)
+       return -errno;
+      if (j == 4)
+       return -101;
+recfd:
       recvsfd = accept(testOptions->c2ssockfd, (struct sockaddr *)  
&cli_addr, &clilen);
       if (recvsfd > 0) {
        log_println(6, "accept() for %d completed", testOptions->child0);
@@ -546,13 +575,15 @@
       if ((recvsfd == -1) && (errno == EINTR)) {
 	log_println(6, "Child %d interrupted while waiting for accept() to  
complete",
                testOptions->child0);
-       continue;
+       goto recfd;
       }
       log_println(6, "-------     C2S connection setup for %d returned  
because (%d)",
                testOptions->child0, errno);
-      if (++j == 4) {
+      if (recvsfd < 0)
+       return -errno;
+      if (j == 4) {
 	log_println(6, "c2s child %d, uable to open connection, return from  
test", testOptions->child0);
-        return -2;
+        return -102;
       }
     }
     log_println(6, "child %d - c2s ready for test with fd=%d",  
testOptions->child0, recvsfd);
@@ -967,24 +998,42 @@
      * This is the second throughput test, with data streaming from
      * the server back to the client.  Again stream data for 10 seconds.
      */
-    log_println(1, "waiting for data on testOptions->s2csockfd");
-
-    j = 0;
+    log_println(1, "%d waiting for data on testOptions->s2csockfd",  
testOptions->child0);
+
     clilen = sizeof(cli_addr);
-    for (;;) {
+    FD_ZERO(&rfd);
+    FD_SET(testOptions->c2ssockfd, &rfd);
+    sel_tv.tv_sec = 5;
+    sel_tv.tv_usec = 0;
+    for (j=0; j<5; j++) {
+      ret = select((testOptions->s2csockfd)+1, &rfd, NULL, NULL, &sel_tv);
+      if ((ret == -1) && (errno == EINTR))
+       continue;
+      if (ret == 0)
+       return -100;  /* timeout */
+      if (ret < 0)
+       return -errno;
+      if (j == 4)
+       return -101;
+ximfd:
       xmitsfd = accept(testOptions->s2csockfd, (struct sockaddr *)  
&cli_addr, &clilen);
       if (xmitsfd > 0) {
-	log_println(6, "S2C %d, has sfd=%d, read to stream data",  
testOptions->child0, xmitsfd);
-        break;
-      }
-      if ((xmitsfd == -1) && (errno == EINTR))
-       continue;
-
-      sprintf(tmpstr, "-------     S2C connection setup returned because  
(%d)", errno);
-      if (get_debuglvl() > 1)
-        perror(tmpstr);
-      if (++j == 4)
-        return -2;
+       log_println(6, "accept() for %d completed", testOptions->child0);
+       break;
+      }
+      if ((xmitsfd == -1) && (errno == EINTR)) {
+	log_println(6, "Child %d interrupted while waiting for accept() to  
complete",
+               testOptions->child0);
+       goto ximfd;
+      }
+      log_println(6, "-------     S2C connection setup for %d returned  
because (%d)",
+               testOptions->child0, errno);
+      if (xmitsfd < 0)
+       return -errno;
+      if (++j == 4) {
+	log_println(6, "s2c child %d, uable to open connection, return from  
test", testOptions->child0);
+        return -102;
+      }
     }
     src_addr = I2AddrByLocalSockFD(get_errhandle(), xmitsfd, 0);
     conn = web100_connection_from_socket(agent, xmitsfd);
=======================================
--- /trunk/src/web100srv.c      Fri Apr  9 09:49:47 2010
+++ /trunk/src/web100srv.c      Wed Apr 21 17:45:18 2010
@@ -842,7 +842,7 @@
     return NULL;
 }

-void
+int
 run_test(web100_agent* agent, int ctlsockfd, TestOptions* testopt, char  
*test_suite)
 {

@@ -931,6 +931,7 @@
        log_println(6, "Middlebox test failed with rc=%d", ret);
       log_println(0, "Middlebox test FAILED!, rc=%d", ret);
       testopt->midopt = TOPT_DISABLED;
+      return ret;
   }

 /*  alarm(20); */
@@ -938,8 +939,6 @@
   if ((ret = test_sfw_srv(ctlsockfd, agent, &*testopt, conn_options)) !=  
0) {
       if (ret < 0)
        log_println(6, "SFW test failed with rc=%d", ret);
-      log_println(0, "Simple firewall test FAILED!, rc=%d", ret);
-      testopt->sfwopt = TOPT_DISABLED;
   }

 /*  alarm(25); */
@@ -950,6 +949,7 @@
        log_println(6, "C2S test failed with rc=%d", ret);
       log_println(0, "C2S throughput test FAILED!, rc=%d", ret);
       testopt->c2sopt = TOPT_DISABLED;
+      return ret;
   }

 /*  alarm(25); */
@@ -960,6 +960,7 @@
        log_println(6, "S2C test failed with rc=%d", ret);
       log_println(0, "S2C throughput test FAILED!, rc=%d", ret);
       testopt->s2copt = TOPT_DISABLED;
+      return ret;
   }

   log_println(4, "Finished testing C2S = %0.2f Mbps, S2C = %0.2f Mbps",  
c2sspd/1000, s2cspd/1000);
@@ -1280,6 +1281,7 @@
   }
   shutdown(ctlsockfd, SHUT_WR);
   /* shutdown(ctlsockfd, SHUT_RDWR); */
+  return (0);
 }

 int
@@ -1695,6 +1697,7 @@

   for(;;){

+mainloop:
     if (head_ptr == NULL)
       log_println(3, "nothing in queue");
     else
@@ -1856,52 +1859,6 @@
        goto sel_11;
       tt = time(0);

-/*
-      if (head_ptr != NULL) {
-        log_println(3, "now = %ld Process started at %ld, run time = %ld",
-            tt, head_ptr->stime, (tt - head_ptr->stime));
-        if ((tt - head_ptr->stime) > 60) {
-          /-* process is stuck at the front of the queue. *-/
-          fp = fopen(get_logfile(),"a");
-          if (fp != NULL) {
-            fprintf(fp, "%d children waiting in queue: Killing off stuck  
process %d at %15.15s\n",
-                waiting, head_ptr->pid, ctime(&tt)+4);
-            fclose(fp);
-          }
-          log_println(6, "%d children waiting in queue: Killing off stuck  
process %d at %15.15s\n",
-                waiting, head_ptr->pid, ctime(&tt)+4);
-          -* kill(tmp_ptr->pid, SIGTERM); *-
-          -* kill(head_ptr->pid, SIGCHLD); *-
-         -* clean up more and inform the client that the test is ending
-          * rac 2/27/10
-          *-
-          log_println(6, "pid=%d, client='%s', stime=%ld, qtime=%ld  
now=%ld", head_ptr->pid, head_ptr->addr,
-                        head_ptr->stime, head_ptr->qtime, time(0));
-	  log_println(6, "pipe-fd=%d, running=%d, ctlsockfd=%d, client-type=%d,  
tests='%s'",
-                       head_ptr->pipe, head_ptr->running, 
head_ptr->ctlsockfd,
-                       head_ptr->oldclient, head_ptr->tests);
-          send_msg(head_ptr->ctlsockfd, SRV_QUEUE, "9555", 4);
-         shutdown(head_ptr->ctlsockfd, SHUT_WR);
-         close(head_ptr->ctlsockfd);
-          tpid = head_ptr->pid;
-         child_sig(-1);
-          kill(tpid, SIGTERM);
-         child_sig(tpid);
-
-         if (((multiple == 0) && (waiting == 1)) ||
-               ((multiple == 1) && (mclients == 0)))
-            testing = 0;
-	  -* should not decrement waiting here, it was decrementd in the  
child_sig() routine
-          * RAC 2/27/09
-          *-
-          -* if (waiting > 0)
-           *  waiting--;
-           *-
-         if (waiting == 0)
-           mclients = 0;
-        }
-      }
- */
     }
     else {
 	/* Nothing is in the queue, so wait forever until a new connection  
request arrives */
@@ -1976,7 +1933,26 @@
          continue;
        if (rc == 13)
          break;
+       if (rc == -1) {
+         log_println(1, "Initial contact with client failed errno=%d", 
errno);
+          close(chld_pipe[0]);
+          close(chld_pipe[1]);
+         shutdown(ctlsockfd, SHUT_WR);
+          close(ctlsockfd);
+          goto mainloop;
+        }
+
+       log_println(6, "xxx, calling initialize_tests()");
+       t_opts = initialize_tests(ctlsockfd, &testopt, test_suite);
+       if (t_opts < 1) {
+	    log_println(3, "Invalid test suite string '%s' received, terminate  
child", test_suite);
+            close(chld_pipe[0]);
+            close(chld_pipe[1]);
+           shutdown(ctlsockfd, SHUT_WR);
+            close(ctlsockfd);
+
        /* todo: handle other error contitions */
+        }
       }
       new_child = (struct ndtchild *) malloc(sizeof(struct ndtchild));
       memset(new_child, 0, sizeof(struct ndtchild));
@@ -2196,7 +2172,7 @@
          }
         }

-multi_client:
+/* multi_client: */
        if ((multiple == 1) && (mclients < max_clients)) {
          if (mwaiting == 0)
            continue;
@@ -2244,6 +2220,13 @@
 	    log_println(6, "Failed to write 'GO' message to client %d, reason=%d,  
errno=%d",
                        mchild->pid, rc, errno);
            /* TODO: handle other error conditions */
+           if (rc == -1) {
+	      log_println(1, "Dispatch multi-client failed because '%s'",  
strerror(errno));
+             shutdown(mchild->ctlsockfd, SHUT_WR);
+              close(mchild->ctlsockfd);
+             kill(chld_pid, SIGTERM);
+             goto mainloop;
+           }
          }
           close(mchild->pipe);
           close(mchild->ctlsockfd);
@@ -2261,6 +2244,13 @@
            if (rc == strlen(tmpstr))
              break;
            /* TODO: handle other error conditions */
+           if (rc == -1) {
+	      log_println(1, "Dispatch multi-client failed because '%s'",  
strerror(errno));
+             shutdown(head_ptr->ctlsockfd, SHUT_WR);
+              close(head_ptr->ctlsockfd);
+             kill(chld_pid, SIGTERM);
+             goto mainloop;
+           }
          }
           close(head_ptr->pipe);
           close(head_ptr->ctlsockfd);
@@ -2384,10 +2374,15 @@

        if (strncmp(test_suite, "Invalid", 7) != 0) {
 	  log_println(3, "Valid test sequence requested, run test for client=%d",  
getpid());
-          run_test(agent, ctlsockfd, &testopt, test_suite);
+          rc = run_test(agent, ctlsockfd, &testopt, test_suite);
        }

-        log_println(3, "Successfully returned from run_test() routine");
+       if (rc == 0)
+          log_println(3, "Successfully returned from run_test() routine");
+       else {
+          log_println(3, "Child %d returned non-zero (%d) from run_test()  
results some test failed!", getpid(), rc);
+         child_sig(0);
+       }
         close(ctlsockfd);
         web100_detach(agent);
        log_free();