benchpar.c is compiled when configure --enable-build-all --enable-paralllel.
But it failed to compile. It is decided to retire or move it to example.
I removed it from the PARA_BUILD_ALL list so that --enable-build-all will
not fail. The program file still need to be moved.
Tested: Koala using --enable-build-all --enable-parallel.
--in parallel mode.This line, and those below, will be ignored--
M perform/Makefile.in
M perform/Makefile.am
distribute by the author. It was an example program showing how to measure
parallel IO speed. Since then, we have created h5perf to do I/O performanace
measurement. This example is not needed any more.
mpi-perf.c:
Removed.
Makefile.am:
Makefile.in:
Removed the entry of mpi-perf.c so that "configure --enable-build-all ..."
will not build it.
The change is trivial that I got permission to skip the code review step.
Tested: h5committest plus "--enable-parallel --enable-build-all" in koala.
Ran bin/reconfigure to update the Makefile.in in directories not part of the fortran directory check=in. Updates Makefile.in due to changes made in configure.in for the Fortran 2003 additions.
Tested on all platforms run under daily tests.
Bring Coverity changes back to trunk:
r20340:
Address coverity issue 248. Set freed pointer to NULL if realloc fails, then check for NULL before attempting to close array of handles stored in memory addressed by pointer.
r20341:
Added initialization udata.path = NULL; udata.visited = NULL; to prevent unintialized use after done: label.
Do C-style initialization AFTER all variable definitions.
r20342:
Fix valgrind error in H5DUMP-XML-tall.h5
r20343:
Addressed Coverity issue 591 - check the return value of H5S_SELECT_ITER_RELEASE.
r20344:
Added initialization udata.path = NULL; to prevent uninitialized used after done: label
r20347:
Purpose: Fix memory leak in overhead.c
Description: Fixed leak in test() in overhead.c where array "had" was never
freed.
r20348:
Addressed Coverity issue 705 - added prototype for resize_configs_are_equal to cache_common.h.
r20392:
Purpose: Fix coverity issues 1677 and 1678
Description: Added return value checking for 4 funciton calls in
H5tools_get_symlink_info().
r20394:
Purpose: Fix coverity issues 1674 and 1675
Description: Modified H5Epush2() and H5E_printf_stack() to keep track of whether
va_start() has been called, and to call va_end() at the end if so.
r20395:
Should be H5MM_calloc(count * sizeof(void*)). Fixed --gh
r20396:
Should be H5MM_calloc(count * sizeof(void*)). Fixed. --gh
Tested on:
Mac OSX/32 10.6.7 (amazon) w/debug & production
(h5committested on branch)
General shared library improvements for CYGWIN / AIX
Description:
Shared libraries are disabled on both CYGWIN and AIX due
to inability to build them correctly. Part of the problem
in both of these situations is the lack of the libtool
flag -no-undefined, which tells libtool that all needed
symbols are defined at link time (a requirement on these
systems) and that it's okay to build shared libraries.
Another problem are lack of dependencies between wrapper
libraries and core C HDF5 library.
This patch addresses both of these by fixing configure to
add in -no-undefined flag for libtool during linking and
adds automake dependencies in the Makefile.am files.
After testing, both CYGWIN and AIX now generate shared
libraries, but there are still some test failures in each.
(cache_api, dt_arith, and testerror.sh on CYGWIN, and
fortran tests on AIX).
Even though the shared libraries are not quite perfect,
this is a general improvement to what we had before, so
I'm applying the patch anyways. Note that default behavior
of shared libraries on these systems being disabled has
NOT been changed and requires the use of the
--enable-unsupported to attempt to build them.
We will need to address the test failures in each
architecture prior to formally supporting shared
libraries on each.
Tested:
h5committested & CYGWIN tested (on bangan)
(AIX tested by Albert on bp-login2)
Add "silent make" mode configure option.
Description:
Automake 1.11 has a new option available that allows for a
silent make mode. This functionality needs to be explicitly
enabled in configure.in via the use of the automake macro
AM_SILENT_RULES, which is what this commit is adding.
This introduces a new configure option:
--{en|dis}able-silent-rules
This option is on by default, and simplies compile and link
line outputs when building the library. Disabling this option
will print full "verbose" output (i.e., full compile and
linking lines for each target).
Tested:
This was tested on jam & h5committested
- Revise shared Fortran library disabling scenarios in configure
- Improve configure output summary
Description:
Shared Fortran libraries are not supported on Mac, but were being
disabled by configure in a way that also forced the C libraries
to be static-only. This has been fixed, so now only shared Fortran
is disabled while shared C can remain.
This prompted two additional changes:
1. While working on the check that addresses whether or not
shared Fortran libraries are allowed, removed old and no
longer needed check(s) that disable shared Fortran
libraries with HP, Intel 8, PGI, and Absoft compilers.
(Essentially, Mac is the only situation in which Fortran
shared are disabled by configure.)
2. Having two different states of libraries (i.e. shared C
library with static-only Fortran library) was not apparent
in the configure summary, which labeled all libraries as
either shared and/or static. I've added lines to both the
C++ and Fortran output sections to list shared/static-ness
of these libraries specifically.
Additionally, I've made sure that the new --enable-unsupported
configure option correctly overrides configure if it tries to
disable a shared library.
Tested:
jam, fred, & h5committest
Also removed includes from top of file - they are included with proper config checks in h5private.h.
Changed name of Windows only gettimeofday function to avoid a define loop according to VS2008
Tested: local linux, windows
Bring Coverity changes back to trunk:
r19733:
Fix memory leak in h5perf_serial. Global buffer "buffer2" was allocated at the
beginning of do_sio() but never freed. Added call to free() at end of do_sio().
r19734:
Fix memory leak in iopipe. Buffer "the_data" was allocated at the beginning of
main() but never freed. Added call to free() at end of main().
Tested on:
Coverity branch in daily tests & Mac
not like it as it may get into a racing condition when "gmake -j N" is used--
two "rm" commands were trying to remove the same file.
Solution:
Rearranged the Makefile.am so that h5perf or h5perf_serial are built only
once, that is through bin_PROGRAMS only.
Bug 2135: h5perf_serial is not installed in --enable-parallel build.
It was because in parallel build, h5perf_serial is not was not in the
bin_PROGRAMS list but it was still being built and tested.
Solution:
Rearranged the Makefile.am code so that h5perf_serial is installed too.
(It is easier to make cleaner code to install it than to build and test
h5perf_serial but not test it.)
Tested: Jam, both serial and parallel.
Did not run h5committest because the changes are in perform/ directory only
and I think Jam has given the changes in Makefile a complete test already.
