Merge remote-tracking branch 'origin/mmap-dev'

Merging mmap feature development branch back into master. Darshan
can now mmap its data structures to temporary log files at runtime to
protect against applications that do not call MPI_Finalize. Feature
can be enabled at configure time using '--enable-mmap-logs'.

darshan-bgq-log-format.h

@@ -11,7 +11,7 @@
 #define DARSHAN_BGQ_VER 1
 
 #define BGQ_COUNTERS \
-    /* control system jobid*/\
+    /* control system jobid */\
     X(BGQ_CSJOBID) \
     /* number of BGQ compute nodes */\
     X(BGQ_NNODES) \
@@ -60,16 +60,13 @@ enum darshan_bgq_f_indices
  * which would actually be logged to file by Darshan for the "BGQ" example
  * module. This example implementation logs the following data for each
  * record:
- *      - a corresponding Darshan record identifier
- *      - the rank of the process responsible for the record
+ *      - a darshan_base_record structure, which contains the record id & rank
  *      - integer I/O counters (operation counts, I/O sizes, etc.)
  *      - floating point I/O counters (timestamps, cumulative timers, etc.)
  */
 struct darshan_bgq_record
 {
-    darshan_record_id f_id;
-    int64_t rank;
-    int alignment;
+    struct darshan_base_record base_rec;
     int64_t counters[BGQ_NUM_INDICES];
     double fcounters[BGQ_F_NUM_INDICES];
 };

darshan-hdf5-log-format.h

@@ -41,15 +41,13 @@ enum darshan_hdf5_f_indices
 /* file record structure for HDF5 files. a record is created and stored for
  * every HDF5 file opened by the original application. For the HDF5 module,
  * the record includes:
- *      - a corresponding record identifier (created by hashing the file path)
- *      - the rank of the process which opened the file (-1 for shared files)
+ *      - a darshan_base_record structure, which contains the record id & rank
  *      - integer file I/O statistics (open, read/write counts, etc)
  *      - floating point I/O statistics (timestamps, cumulative timers, etc.)
  */
 struct darshan_hdf5_file
 {
-    darshan_record_id f_id;
-    int64_t rank;
+    struct darshan_base_record base_rec;
     int64_t counters[HDF5_NUM_INDICES];
     double fcounters[HDF5_F_NUM_INDICES];
 };

darshan-log-format.h

@@ -20,7 +20,7 @@
 #endif
 
 /* update this on file format changes */
-#define DARSHAN_LOG_VERSION "3.00"
+#define DARSHAN_LOG_VERSION "3.01"
 
 /* magic number for validating output files and checking byte order */
 #define DARSHAN_MAGIC_NR 6567223
@@ -44,6 +44,7 @@ enum darshan_comp_type
 {
     DARSHAN_ZLIB_COMP,
     DARSHAN_BZIP2_COMP,
+    DARSHAN_NO_COMP, 
 };
 
 typedef uint64_t darshan_record_id;
@@ -68,7 +69,7 @@ struct darshan_header
     int64_t magic_nr;
     unsigned char comp_type;
     uint32_t partial_flag;
-    struct darshan_log_map rec_map;
+    struct darshan_log_map name_map;
     struct darshan_log_map mod_map[DARSHAN_MAX_MODS];
     uint32_t mod_ver[DARSHAN_MAX_MODS];
 };
@@ -85,11 +86,18 @@ struct darshan_job
     char metadata[DARSHAN_JOB_METADATA_LEN];
 };
 
-/* minimal record stored for each file/object accessed by Darshan */
-struct darshan_record
+/* record to store name->darshan_id mapping for each registered record */
+struct darshan_name_record
 {
-    char* name;
     darshan_record_id id;
+    char name[1];
+};
+
+/* base record definition that can be used by modules */
+struct darshan_base_record
+{
+    darshan_record_id id;
+    int64_t rank;
 };
 
 

darshan-mpiio-log-format.h

@@ -140,15 +140,13 @@ enum darshan_mpiio_f_indices
 /* file record structure for MPI-IO files. a record is created and stored for
  * every MPI-IO file opened by the original application. For the MPI-IO module,
  * the record includes:
- *      - a corresponding record identifier (created by hashing the file path)
- *      - the rank of the process which opened the file (-1 for shared files)
+ *      - a darshan_base_record structure, which contains the record id & rank
  *      - integer file I/O statistics (open, read/write counts, etc)
  *      - floating point I/O statistics (timestamps, cumulative timers, etc.)
  */
 struct darshan_mpiio_file
 {
-    darshan_record_id f_id;
-    int64_t rank;
+    struct darshan_base_record base_rec;
     int64_t counters[MPIIO_NUM_INDICES];
     double fcounters[MPIIO_F_NUM_INDICES];
 };

darshan-null-log-format.h

@@ -44,15 +44,13 @@ enum darshan_null_f_indices
  * which would actually be logged to file by Darshan for the "NULL" example
  * module. This example implementation logs the following data for each
  * record:
- *      - a corresponding Darshan record identifier
- *      - the rank of the process responsible for the record
+ *      - a darshan_base_record structure, which contains the record id & rank
  *      - integer I/O counters (operation counts, I/O sizes, etc.)
  *      - floating point I/O counters (timestamps, cumulative timers, etc.)
  */
 struct darshan_null_record
 {
-    darshan_record_id f_id;
-    int64_t rank;
+    struct darshan_base_record base_rec;
     int64_t counters[NULL_NUM_INDICES];
     double fcounters[NULL_F_NUM_INDICES];
 };

darshan-pnetcdf-log-format.h

@@ -43,15 +43,13 @@ enum darshan_pnetcdf_f_indices
 /* file record structure for PNETCDF files. a record is created and stored for
  * every PNETCDF file opened by the original application. For the PNETCDF module,
  * the record includes:
- *      - a corresponding record identifier (created by hashing the file path)
- *      - the rank of the process which opened the file (-1 for shared files)
+ *      - a darshan_base_record structure, which contains the record id & rank
  *      - integer file I/O statistics (open, read/write counts, etc)
  *      - floating point I/O statistics (timestamps, cumulative timers, etc.)
  */
 struct darshan_pnetcdf_file
 {
-    darshan_record_id f_id;
-    int64_t rank;
+    struct darshan_base_record base_rec;
     int64_t counters[PNETCDF_NUM_INDICES];
     double fcounters[PNETCDF_F_NUM_INDICES];
 };

darshan-posix-log-format.h

@@ -164,15 +164,13 @@ enum darshan_posix_f_indices
 /* file record structure for POSIX files. a record is created and stored for
  * every POSIX file opened by the original application. For the POSIX module,
  * the record includes:
- *      - a corresponding record identifier (created by hashing the file path)
- *      - the rank of the process which opened the file (-1 for shared files)
+ *      - a darshan_base_record structure, which contains the record id & rank
  *      - integer file I/O statistics (open, read/write counts, etc)
  *      - floating point I/O statistics (timestamps, cumulative timers, etc.)
  */
 struct darshan_posix_file
 {
-    darshan_record_id f_id;
-    int64_t rank;
+    struct darshan_base_record base_rec;
     int64_t counters[POSIX_NUM_INDICES];
     double fcounters[POSIX_F_NUM_INDICES];
 };

darshan-runtime/Makefile.in

@@ -17,6 +17,7 @@ CC = @CC@
 LD = @LD@
 
 DISABLE_LDPRELOAD = @DISABLE_LDPRELOAD@
+ENABLE_MMAP_LOGS = @ENABLE_MMAP_LOGS@
 DARSHAN_USE_BGQ = @DARSHAN_USE_BGQ@
 DARSHAN_LOG_FORMAT = $(srcdir)/../darshan-log-format.h
 DARSHAN_VERSION = @DARSHAN_VERSION@
@@ -157,6 +158,9 @@ endif
 	install -m 644 $(srcdir)/share/ld-opts/darshan-posix-ld-opts $(datarootdir)/ld-opts/darshan-posix-ld-opts
 	install -m 644 $(srcdir)/share/ld-opts/darshan-hdf5-ld-opts $(datarootdir)/ld-opts/darshan-hdf5-ld-opts
 	install -m 644 $(srcdir)/share/ld-opts/darshan-pnetcdf-ld-opts $(datarootdir)/ld-opts/darshan-pnetcdf-ld-opts
+ifdef ENABLE_MMAP_LOGS
+	install -m 755 share/darshan-mmap-epilog.sh $(datarootdir)/darshan-mmap-epilog.sh
+endif
 	install -d $(libdir)/pkgconfig
 	install -m 644 lib/pkgconfig/darshan-runtime.pc $(libdir)/pkgconfig/darshan-runtime.pc
 
@@ -164,5 +168,5 @@ clean::
 	rm -f *.o *.a lib/*.o lib/*.po lib/*.a lib/*.so 
 
 distclean:: clean
-	rm -f darshan-runtime-config.h darshan-gen-cxx.pl darshan-gen-fortran.pl darshan-gen-cc.pl darshan-mk-log-dirs.pl darshan-config darshan-base-ld-opts aclocal.m4 autom4te.cache/* config.status config.log Makefile 
+	rm -f darshan-runtime-config.h darshan-gen-cxx.pl darshan-gen-fortran.pl darshan-gen-cc.pl darshan-mk-log-dirs.pl darshan-config lib/pkgconfig/darshan-runtime.pc share/craype-1.x/darshan-module share/craype-2.x/darshan-module share/darshan-mmap-epilog.sh share/ld-opts/darshan-base-ld-opts share/mpi-profile/darshan-bg-cc.conf share/mpi-profile/darshan-bg-cxx.conf share/mpi-profile/darshan-bg-f.conf share/mpi-profile/darshan-cc.conf share/mpi-profile/darshan-cxx.conf share/mpi-profile/darshan-f.conf aclocal.m4 autom4te.cache/* config.status config.log Makefile 
 	rm -rf autom4te.cache

darshan-runtime/configure

@@ -624,6 +624,7 @@ LIBOBJS
 DARSHAN_USE_BGQ
 MPICH_LIB_OLD
 DARSHAN_VERSION
+ENABLE_MMAP_LOGS
 DISABLE_LDPRELOAD
 __DARSHAN_LOG_PATH
 darshan_share_path
@@ -684,15 +685,16 @@ ac_subst_files=''
 ac_user_opts='
 enable_option_checking
 with_zlib
-enable_cuserid
 enable_ld_preload
+enable_cuserid
 enable_group_readable_logs
+enable_bgq_mod
+enable_mmap_logs
 with_mem_align
 with_log_path_by_env
 with_log_hints
 with_log_path
 with_jobid_env
-enable_bgq_mod
 with_mod_mem
 '
       ac_precious_vars='build_alias
@@ -1313,10 +1315,12 @@ Optional Features:
   --disable-option-checking  ignore unrecognized --enable/--with options
   --disable-FEATURE       do not include FEATURE (same as --enable-FEATURE=no)
   --enable-FEATURE[=ARG]  include FEATURE [ARG=yes]
-  --disable-cuserid       Disables attempted use of cuserid() at run time
   --disable-ld-preload    Disables support for LD_PRELOAD library
-  --enable-group-readable-logs Set log files to be group readable
+  --disable-cuserid       Disables attempted use of cuserid() at run time
+  --enable-group-readable-logs
+                          Set log files to be group readable
   --disable-bgq-mod       Disables compilation and use of BG/Q module (for BG/Q systems)
+  --enable-mmmap-logs     Enables ability to mmap I/O data to log file
 
 Optional Packages:
   --with-PACKAGE[=ARG]    use PACKAGE [ARG=yes]
@@ -1325,13 +1329,16 @@ Optional Packages:
                     /usr/local or /usr if not found in /usr/local
   --without-zlib to disable zlib usage completely
   --with-mem-align=<num>  Memory alignment in bytes
-  --with-log-path-by-env=<env var list> Comma separated list of environment variables to check for log path location before --with-log-path path
-  --with-log-hints=<hint1=x;hint2=y>  Semicolon-separated list of MPI-IO hints for log file write
+  --with-log-path-by-env=<env var list>
+                          Comma separated list of environment variables to check for
+                          log path location before --with-log-path path
+  --with-log-hints=<hint1=x;hint2=y>
+                          Semicolon-separated list of MPI-IO hints for log file write
   --with-log-path=<path>  Location to store log files at run time
   --with-jobid-env=<name> Name of environment variable that stores the jobid
-    (specify "NONE" if no appropriate environment variable is available:
-    Darshan will use rank 0's pid instead)
-  --with-mod-mem=<num>  Maximum amount of memory (in MiB) for each Darshan module
+                          (specify "NONE" if no appropriate environment variable
+                          is available: Darshan will use rank 0's pid instead)
+  --with-mod-mem=<num>    Maximum amount of memory (in MiB) for each Darshan module
 
 Some influential environment variables:
   CC          C compiler command
@@ -3528,31 +3535,54 @@ fi
 
 
 
+# Check whether --enable-ld-preload was given.
+if test "${enable_ld_preload+set}" = set; then :
+  enableval=$enable_ld_preload; if test "x$enableval" = "xno" ; then
+    DISABLE_LDPRELOAD="1"
+fi
+
+fi
+
+
 # Check whether --enable-cuserid was given.
 if test "${enable_cuserid+set}" = set; then :
   enableval=$enable_cuserid; if test "x$enableval" = "xno" ; then
 
-$as_echo "#define DARSHAN_DISABLE_CUSERID 1" >>confdefs.h
+$as_echo "#define __DARSHAN_DISABLE_CUSERID 1" >>confdefs.h
 
 fi
 
 fi
 
 
-# Check whether --enable-ld-preload was given.
-if test "${enable_ld_preload+set}" = set; then :
-  enableval=$enable_ld_preload; if test "x$enableval" = "xno" ; then
-    DISABLE_LDPRELOAD="1"
+# Check whether --enable-group-readable-logs was given.
+if test "${enable_group_readable_logs+set}" = set; then :
+  enableval=$enable_group_readable_logs; if test "x$enableval" = "xyes" ; then
+
+$as_echo "#define __DARSHAN_GROUP_READABLE_LOGS 1" >>confdefs.h
+
 fi
 
 fi
 
 
-# Check whether --enable-group-readable-logs was given.
-if test "${enable_group_readable_logs+set}" = set; then :
-  enableval=$enable_group_readable_logs; if test "x$enableval" = "xyes" ; then
+# Check whether --enable-bgq_mod was given.
+if test "${enable_bgq_mod+set}" = set; then :
+  enableval=$enable_bgq_mod;
+else
+  enable_bgq_mod=check
+
+fi
 
-$as_echo "#define __DARSHAN_GROUP_READABLE_LOGS 1" >>confdefs.h
+
+# Check whether --enable-mmap-logs was given.
+if test "${enable_mmap_logs+set}" = set; then :
+  enableval=$enable_mmap_logs; if test "x$enableval" = "xyes" ; then
+
+$as_echo "#define __DARSHAN_ENABLE_MMAP_LOGS 1" >>confdefs.h
+
+    ENABLE_MMAP_LOGS=1
+    ac_config_files="$ac_config_files share/darshan-mmap-epilog.sh"
 
 fi
 
@@ -3648,6 +3678,54 @@ _ACEOF
 fi
 
 
+
+# Check whether --with-mod-mem was given.
+if test "${with_mod_mem+set}" = set; then :
+  withval=$with_mod_mem; if test x$withval = xyes; then
+        as_fn_error $? "--with-mod-mem must be given a number" "$LINENO" 5
+    else
+
+cat >>confdefs.h <<_ACEOF
+#define __DARSHAN_MOD_MEM_MAX ${withval}
+_ACEOF
+
+    fi
+
+fi
+
+
+if test x$enable_bgq_mod != xno; then
+    { $as_echo "$as_me:${as_lineno-$LINENO}: checking for BG/Q environment" >&5
+$as_echo_n "checking for BG/Q environment... " >&6; }
+    cat confdefs.h - <<_ACEOF >conftest.$ac_ext
+/* end confdefs.h.  */
+
+        #ifndef __bgq__
+        #error __bgq__ not set
+        #endif
+
+int
+main ()
+{
+
+  ;
+  return 0;
+}
+_ACEOF
+if ac_fn_c_try_compile "$LINENO"; then :
+  { $as_echo "$as_me:${as_lineno-$LINENO}: result: yes" >&5
+$as_echo "yes" >&6; }
+        DARSHAN_USE_BGQ=1
+else
+  { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5
+$as_echo "no" >&6; }
+        if test "x$enable_bgq_mod" = xyes; then :
+  as_fn_error $? "BG/Q module enabled in non-BG/Q environment" "$LINENO" 5
+fi
+fi
+rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext
+fi
+
 if test x$GOT_ALIGNMENT != x1; then
     as_fn_error $? "must provide --with-mem-align=<num> argument to configure." "$LINENO" 5
 fi
@@ -4243,63 +4321,6 @@ else
   MPICH_LIB_OLD=0
 fi
 
-# check to see whether the bgq instrumentation module should be built
-# Check whether --enable-bgq_mod was given.
-if test "${enable_bgq_mod+set}" = set; then :
-  enableval=$enable_bgq_mod;
-else
-  enable_bgq_mod=check
-
-fi
-
-if test x$enable_bgq_mod != xno; then
-    { $as_echo "$as_me:${as_lineno-$LINENO}: checking for BG/Q environment" >&5
-$as_echo_n "checking for BG/Q environment... " >&6; }
-    cat confdefs.h - <<_ACEOF >conftest.$ac_ext
-/* end confdefs.h.  */
-
-        #ifndef __bgq__
-        #error __bgq__ not set
-        #endif
-
-int
-main ()
-{
-
-  ;
-  return 0;
-}
-_ACEOF
-if ac_fn_c_try_compile "$LINENO"; then :
-  { $as_echo "$as_me:${as_lineno-$LINENO}: result: yes" >&5
-$as_echo "yes" >&6; }
-        DARSHAN_USE_BGQ=1
-else
-  { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5
-$as_echo "no" >&6; }
-        if test "x$enable_bgq_mod" = xyes; then :
-  as_fn_error $? "BG/Q module enabled in non-BG/Q environment" "$LINENO" 5
-fi
-fi
-rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext
-fi
-
-
-# Check whether --with-mod-mem was given.
-if test "${with_mod_mem+set}" = set; then :
-  withval=$with_mod_mem; if test x$withval = xyes; then
-        as_fn_error $? "--with-mod-mem must be given a number" "$LINENO" 5
-    else
-
-cat >>confdefs.h <<_ACEOF
-#define __DARSHAN_MOD_MEM_MAX ${withval}
-_ACEOF
-
-    fi
-
-fi
-
-
 DARSHAN_VERSION="3.0.1"
 
 
@@ -4309,6 +4330,7 @@ DARSHAN_VERSION="3.0.1"
 
 
 
+
 ac_config_files="$ac_config_files Makefile darshan-mk-log-dirs.pl darshan-gen-cc.pl darshan-gen-cxx.pl darshan-gen-fortran.pl darshan-config share/craype-1.x/darshan-module share/craype-2.x/darshan-module lib/pkgconfig/darshan-runtime.pc share/mpi-profile/darshan-cc.conf share/mpi-profile/darshan-cxx.conf share/mpi-profile/darshan-f.conf share/mpi-profile/darshan-bg-cc.conf share/mpi-profile/darshan-bg-cxx.conf share/mpi-profile/darshan-bg-f.conf share/ld-opts/darshan-base-ld-opts"
 
 cat >confcache <<\_ACEOF
@@ -5003,6 +5025,7 @@ for ac_config_target in $ac_config_targets
 do
   case $ac_config_target in
     "darshan-runtime-config.h") CONFIG_HEADERS="$CONFIG_HEADERS darshan-runtime-config.h" ;;
+    "share/darshan-mmap-epilog.sh") CONFIG_FILES="$CONFIG_FILES share/darshan-mmap-epilog.sh" ;;
     "Makefile") CONFIG_FILES="$CONFIG_FILES Makefile" ;;
     "darshan-mk-log-dirs.pl") CONFIG_FILES="$CONFIG_FILES darshan-mk-log-dirs.pl" ;;
     "darshan-gen-cc.pl") CONFIG_FILES="$CONFIG_FILES darshan-gen-cc.pl" ;;

darshan-runtime/configure.in

@@ -24,27 +24,44 @@ AC_TRY_COMPILE([#include <mpi.h>], [int ret = MPI_Init(0, (void*)0)],
 dnl runtime libraries require zlib
 CHECK_ZLIB
 
-AC_ARG_ENABLE(cuserid, 
-[  --disable-cuserid       Disables attempted use of cuserid() at run time], 
+AC_ARG_ENABLE(ld-preload, 
+[  --disable-ld-preload    Disables support for LD_PRELOAD library], 
 [if test "x$enableval" = "xno" ; then
-    AC_DEFINE(DARSHAN_DISABLE_CUSERID, 1, Define if cuserid() should be disabled)
+    DISABLE_LDPRELOAD="1"
 fi]
 ,)
 
-AC_ARG_ENABLE(ld-preload, 
-[  --disable-ld-preload    Disables support for LD_PRELOAD library], 
+AC_ARG_ENABLE(cuserid, 
+[  --disable-cuserid       Disables attempted use of cuserid() at run time], 
 [if test "x$enableval" = "xno" ; then
-    DISABLE_LDPRELOAD="1"
+    AC_DEFINE(__DARSHAN_DISABLE_CUSERID, 1, Define if cuserid() should be disabled)
 fi]
 ,)
 
 AC_ARG_ENABLE(group-readable-logs, 
-[  --enable-group-readable-logs Set log files to be group readable], 
+[  --enable-group-readable-logs
+                          Set log files to be group readable], 
 [if test "x$enableval" = "xyes" ; then
     AC_DEFINE(__DARSHAN_GROUP_READABLE_LOGS, 1, Define if Darshan should set log files to be group readable)
 fi]
 ,)
 
+AC_ARG_ENABLE(
+    [bgq_mod],
+    [  --disable-bgq-mod       Disables compilation and use of BG/Q module (for BG/Q systems)],
+    [],
+    [enable_bgq_mod=check]
+)
+
+AC_ARG_ENABLE(mmap-logs,
+[  --enable-mmmap-logs     Enables ability to mmap I/O data to log file],
+[if test "x$enableval" = "xyes" ; then
+    AC_DEFINE(__DARSHAN_ENABLE_MMAP_LOGS, 1, Define if Darshan should mmap data structures to log file)
+    ENABLE_MMAP_LOGS=1
+    AC_CONFIG_FILES(share/darshan-mmap-epilog.sh)
+fi]
+,)
+
 AC_ARG_WITH(mem-align,
 [  --with-mem-align=<num>  Memory alignment in bytes],
     if test x$withval = xyes; then
@@ -56,7 +73,9 @@ AC_ARG_WITH(mem-align,
 )
 
 AC_ARG_WITH(log-path-by-env,
-[  --with-log-path-by-env=<env var list> Comma separated list of environment variables to check for log path location before --with-log-path path],
+[  --with-log-path-by-env=<env var list>
+                          Comma separated list of environment variables to check for
+                          log path location before --with-log-path path],
     if test x$withval = xyes; then
         AC_MSG_ERROR(--with-log-path-by-env must be given at least one variable name)
     else
@@ -67,7 +86,8 @@ AC_ARG_WITH(log-path-by-env,
 )
 
 AC_ARG_WITH(log-hints,
-[  --with-log-hints=<hint1=x;hint2=y>  Semicolon-separated list of MPI-IO hints for log file write],
+[  --with-log-hints=<hint1=x;hint2=y>
+                          Semicolon-separated list of MPI-IO hints for log file write],
     if test x$withval = xyes; then
         AC_MSG_ERROR(--with-log-hints must be given an argument)
     else
@@ -90,8 +110,8 @@ AC_ARG_WITH(log-path,
 
 AC_ARG_WITH(jobid-env,
 [  --with-jobid-env=<name> Name of environment variable that stores the jobid
-    (specify "NONE" if no appropriate environment variable is available: 
-    Darshan will use rank 0's pid instead)],
+                          (specify "NONE" if no appropriate environment variable
+                          is available: Darshan will use rank 0's pid instead)],
     if test x$withval = xyes; then
         AC_MSG_ERROR(--with-jobid-env must be given a name)
     else
@@ -101,6 +121,30 @@ AC_ARG_WITH(jobid-env,
     fi
 )
 
+AC_ARG_WITH(mod-mem,
+[  --with-mod-mem=<num>    Maximum amount of memory (in MiB) for each Darshan module],
+    if test x$withval = xyes; then
+        AC_MSG_ERROR(--with-mod-mem must be given a number)
+    else
+        AC_DEFINE_UNQUOTED(__DARSHAN_MOD_MEM_MAX, ${withval}, Maximum memory (in MiB) for each Darshan module)
+    fi
+)
+
+if test x$enable_bgq_mod != xno; then
+    AC_MSG_CHECKING(for BG/Q environment)
+    AC_TRY_COMPILE([
+        #ifndef __bgq__
+        #error __bgq__ not set
+        #endif
+        ], [],
+        AC_MSG_RESULT(yes)
+        DARSHAN_USE_BGQ=1,
+        AC_MSG_RESULT(no)
+        AS_IF([test "x$enable_bgq_mod" = xyes],
+            AC_MSG_ERROR(BG/Q module enabled in non-BG/Q environment),
+            []))
+fi
+
 if test x$GOT_ALIGNMENT != x1; then
     AC_MSG_ERROR(must provide --with-mem-align=<num> argument to configure.)
 fi
@@ -280,43 +324,13 @@ AS_IF([$CC -show foo.c -o foo >& /dev/null],
 	    MPICH_LIB_OLD=0),
 	MPICH_LIB_OLD=0)
 
-# check to see whether the bgq instrumentation module should be built
-AC_ARG_ENABLE(
-    [bgq_mod],
-    [  --disable-bgq-mod       Disables compilation and use of BG/Q module (for BG/Q systems)],
-    [],
-    [enable_bgq_mod=check]
-)
-if test x$enable_bgq_mod != xno; then
-    AC_MSG_CHECKING(for BG/Q environment)
-    AC_TRY_COMPILE([
-        #ifndef __bgq__
-        #error __bgq__ not set
-        #endif
-        ], [],
-        AC_MSG_RESULT(yes)
-        DARSHAN_USE_BGQ=1,
-        AC_MSG_RESULT(no)
-        AS_IF([test "x$enable_bgq_mod" = xyes],
-            AC_MSG_ERROR(BG/Q module enabled in non-BG/Q environment),
-            []))
-fi
-
-AC_ARG_WITH(mod-mem,
-[  --with-mod-mem=<num>  Maximum amount of memory (in MiB) for each Darshan module],
-    if test x$withval = xyes; then
-        AC_MSG_ERROR(--with-mod-mem must be given a number)
-    else
-        AC_DEFINE_UNQUOTED(__DARSHAN_MOD_MEM_MAX, ${withval}, Maximum memory (in MiB) for each Darshan module)
-    fi
-)
-
 DARSHAN_VERSION="AC_PACKAGE_VERSION"
 AC_SUBST(darshan_lib_path)
 AC_SUBST(darshan_share_path)
 AC_SUBST(LDFLAGS)
 AC_SUBST(__DARSHAN_LOG_PATH)
 AC_SUBST(DISABLE_LDPRELOAD)
+AC_SUBST(ENABLE_MMAP_LOGS)
 AC_SUBST(DARSHAN_VERSION)
 AC_SUBST(MPICH_LIB_OLD)
 AC_SUBST(DARSHAN_USE_BGQ)

darshan-runtime/darshan-common.h

@@ -66,44 +66,68 @@
 } while(0)
 
 /* potentially set or increment a common value counter, depending on the __count
- * for the given __value
+ * for the given __value. This macro ensures common values are stored first in
+ * decreasing order of their total count, and second by decreasing order of
+ * their value.
+
  *
  * NOTE: This macro is hardcoded to expect that Darshan will only track the 4
  * most common (i.e., frequently occuring) values. __val_p is a pointer to the
  * base of the value counters (i.e., the first of 4 contiguous common value
  * counters) and __cnt_p is a pointer to the base of the count counters (i.e.
  * the first of 4 contiguous common count counters). It is assumed your counters
- * are stored as int64_t types.
+ * are stored as int64_t types. __add_flag is set if the given count should be
+ * added to the common access counter, rather than just incrementing it.
  */
-#define DARSHAN_COMMON_VAL_COUNTER_INC(__val_p, __cnt_p, __value, __count) do {\
+#define DARSHAN_COMMON_VAL_COUNTER_INC(__val_p, __cnt_p, __value, __count, __add_flag) do {\
     int i; \
-    int set = 0; \
-    int64_t min = *(__cnt_p); \
-    int min_index = 0; \
+    int inc_count, total_count; \
+    int64_t tmp_val[4] = {0}; \
+    int64_t tmp_cnt[4] = {0}; \
+    int tmp_ndx = 0; \
     if(__value == 0) break; \
+    if(__add_flag) \
+        inc_count = 1; \
+    else \
+        inc_count = __count; \
     for(i=0; i<4; i++) { \
-        /* increment bucket if already exists */ \
         if(*(__val_p + i) == __value) { \
-            *(__cnt_p + i) += __count; \
-            set = 1; \
+            total_count = *(__cnt_p + i) + inc_count; \
             break; \
         } \
-        /* otherwise find the least frequently used bucket */ \
-        else if(*(__cnt_p + i) < min) { \
-            min = *(__cnt_p + i); \
-            min_index = i; \
+    } \
+    if(i == 4) total_count = __count; \
+    /* first, copy over any counters that should be sorted above this one \
+     * (counters with higher counts or equal counts and larger values) \
+     */ \
+    for(i=0;i < 4; i++) { \
+        if((*(__cnt_p + i) > total_count) || \
+           ((*(__cnt_p + i) == total_count) && (*(__val_p + i) > __value))) { \
+            tmp_val[tmp_ndx] = *(__val_p + i); \
+            tmp_cnt[tmp_ndx] = *(__cnt_p + i); \
+            tmp_ndx++; \
         } \
+        else break; \
     } \
-    if(!set && (__count > min)) { \
-        *(__cnt_p + min_index) = __count; \
-        *(__val_p + min_index) = __value; \
+    if(tmp_ndx == 4) break; /* all done, updated counter is not added */ \
+    /* next, add the updated counter */ \
+    tmp_val[tmp_ndx] = __value; \
+    tmp_cnt[tmp_ndx] = total_count; \
+    tmp_ndx++; \
+    /* last, copy over any remaining counters to make sure we have 4 sets total */ \
+    while(tmp_ndx != 4) { \
+        if(*(__val_p + i) != __value) { \
+            tmp_val[tmp_ndx] = *(__val_p + i); \
+            tmp_cnt[tmp_ndx] = *(__cnt_p + i); \
+            tmp_ndx++; \
+        } \
+        i++; \
     } \
+    memcpy(__val_p, tmp_val, 4*sizeof(int64_t)); \
+    memcpy(__cnt_p, tmp_cnt, 4*sizeof(int64_t)); \
 } while(0)
 
-/* maximum number of common values that darshan will track per file at
- * runtime; at shutdown time these will be reduced to the 4 most
- * frequently occuring ones
- */
+/* maximum number of common values that darshan will track per file at runtime */
 #define DARSHAN_COMMON_VAL_MAX_RUNTIME_COUNT 32
 struct darshan_common_val_counter
 {
@@ -130,6 +154,67 @@ struct darshan_variance_dt
 * darshan-common functions for darshan modules *
 ***********************************************/
 
+/* darshan_lookup_record_ref()
+ *
+ * Lookup a record reference pointer using the given 'handle'.
+ * 'handle_sz' is the size of the handle structure, and 'hash_head'
+ * is the pointer to the hash table to search.
+ * If the handle is found, the corresponding record reference pointer
+ * is returned, otherwise NULL is returned.
+ */
+void *darshan_lookup_record_ref(
+    void *hash_head,
+    void *handle,
+    size_t handle_sz);
+
+/* darshan_add_record_ref()
+ *
+ * Add the given record reference pointer, 'rec_ref_p' to the hash
+ * table whose address is stored in the 'hash_head_p' pointer. The
+ * hash is generated from the given 'handle', with size 'handle_sz'.
+ * If the record reference is successfully added, 1 is returned,
+ * otherwise, 0 is returned.
+ */
+int darshan_add_record_ref(
+    void **hash_head_p,
+    void *handle,
+    size_t handle_sz,
+    void *rec_ref_p);
+
+/* darshan_delete_record_ref()
+ *
+ * Delete the record reference for the given 'handle', with size