Purpose:
Use port hunting to test the Stream VFD
Description:
The stream driver is tested by streaming data
between two different processes on the local
machine on a given default port.
If this port is already is use, port hunting
should find the next available port to use.
The hostname/port information which is actually
used by the sender is written to a temporary
file which is then read by the receiver process
to connect to the sender's port.
For the purpose of testing I switched back
the default port to use from 10007 to 5678
which is at least already used by another
service on modi4.
Platforms tested:
x86 Linux, Irix 32/64 bit (modi4), Dec Alpha,
Unicos on T3E, Hitachi SR8000, AIX on SP2
Purpose:
Implemented port hunting
Really catch SIGPIPE signals
Description:
Port hunting allows to try and bind to a successive port number
if the port number given in the filename is already used.
Because of a stupid typo in the code, SIGPIPE wasn't catched
which might have caused the sending side to hang.
Solution:
The Stream VFD's file access property list was extended by two
elements:
- maxhunt: how many successive ports to try if the one
given in the filename is already in use
Default is not to hunt for additional ports.
- port: port number which is finally used to bind a socket
This might be different to the port number
as given in the filename if port hunting is enabled.
The H5Pget_fapl_stream() can be used to obtain
this port number.
Platforms tested:
x86 Linux, Irix 32/64 bit, Dec Alpha, Unicos on T3E, AIX on SP2
Hitachi SR8000
Code cleanup
Description:
Recent CodeWarrior patches have broken the Unix builds and moved code
around in non-portable ways.
Solution:
Patched things back up to try to accomodate CodeWarrior and still let the
Unix builds work correctly.
Platforms tested:
FreeBSD 4.3 (hawkwind)
code warrior support
and some clean up
the macros file_seek and file_offset_t that were repeated over sevral files were put only in
H5private.h
H5private .h was updated for win32
vthe
Description:
Solution:
Platforms tested:
Document Solaris 64-bit support.
Description:
Added paragraph to "alternate compilers" section which describes how to
enable 64-bit support for Solaris.
Purpose:
Bug Fix
Description:
Modules weren't being found correctly.
Solution:
The F9XMODFLAG was being set and used in the test{par}/ and examples/
directories, but it's also needed in the src/ directory.
Platforms tested:
Modi4...cross fingers...
Purpose:
Bug Fix
Description:
Changed the order of the flags checking for modules from -M -I to -I
-M. Elena already fixed it in the configure ifle. This puts it in the
acsite.m4 file so that it will be propagated to future configure
builds.
Purpose:
bug fix
Description:
update H5pubconf.h and zconf.h(used for dll version)
Solution:
see above
Platforms tested:
elena will test on her machine
Purpose:
bug fix for windows MVS
Description:
adding various windows macros to make it work on windows platform
Solution:
see above
Platforms tested:
test on windows 2000, confirmed on linux.
Purpose:
1. update windows H5pubconf.h and zconf.h
still more work to be done to make sure that windows catches up
with UNIX platforms.
2. 1) bug fixings due to the deletion of windows.h
For several .c files under src and test, windows.h still needs to be included.
2) also IO.h and direct.h are needed for several files.
3) mkdir is _mkdir on windows without the second argument.
4) winsock.h needed for iopipe.c, winsock2.h needed for H5detect.c
3. add giftoh5 and h5togif in all_withhdf4.zip; update project settings.
Description:
see the above
Solution:
see the above
Platforms tested:
windows 2000
Purpose:
Maintenance on IBM SP
Description:
I used unnecessary system specific compiler falgs
Solution:
Removed unnecessary flags
Platforms tested:
NERSC IBM SP (gseaborg)
Purpose:
Bug bypass
Description:
Fortran configure checks flags where to find F90 modules in the
order "M", "I" and "p". On IBM SP "M" is a reserved flag, configure
does not understand compiler's warning and uses it to specify
module directories. Then tests and examples compilation fails since
module directory is not specified correctly on the compilation line.
Solution:
Changed the order of flags to be "I", "M", "p", so configure on IBM SP
finds the correct flag first.
Platforms tested:
NERSC IBM SP (gseaborg)
Code clean/bug fix
Description:
H5FL (free-list manager) code currently is taking an hsize_t as the size
of a memory block to allocate. On many machines, the size of an hsize_t
is greater than the size of a size_t, potentially leading to incorrect
memory allocations in rare circumstances.
Solution:
Changed hsize_t parameters and variables to size_t.
Platforms tested:
FreeBSD 4.3 (hawkwind)
Purpose:
IBM SP F90 HDF5 Library port
Description:
Added necessary type definitions for the port
Platforms tested:
Tested only for 1.4. branch. 1.5 branch will be tested later.
Purpose:
DEC UNIX port
Description:
Windows port caused problems on DEC UNIX.
Names of C functions called from F90 subroutines were converted
to UPPER case while DEC UNIX expected them to be in lower case.
Solution:
Added special DEC compiler directives to avoid the conflict on two
systems
Platforms tested:
Windows98, DEC UNIX, Linux
Purpose:
Bug fixes
Description:
Fortran test batch file had a typo in it that caused fflush2 test
to fail.
gif converter projects had wrong setting for the
include directories that caused build to fail.
Solution:
Fixed the projects, and replaced the projects in the archive
Platforms tested:
Windows98
Multiple-group testing
Description:
Added multiple groups under root group and multiple subgroups of
certain levels. Also writes several datasets in each group in
parallel.
Solution:
Platforms tested:
MPICH(IRIX 6.5, IRIX64 6.5(64), IRIX64 6.5(N32), Linux, SunOS 5.6)
and SGI MPI(IRIX64 6.5).
Purpose:
Multiple-group testing
Description:
Added multiple groups under root group and multiple subgroups of
certain levels. Also writes several datasets in each group in
parallel.
Solution:
Platforms tested:
MPICH(IRIX 6.5, IRIX64 6.5(64), IRIX64 6.5(N32), Linux, SunOS 5.6)
and SGI MPI(IRIX64 6.5).
Code Cleanup
Description:
Changed POSIX function calls to the HD* way of doing things. Checked
with checkposix and tested on Linux.
Platforms tested:
Linux
Bug Fix
Description:
On some platforms, long long is printed with the "ll" modifier.
HDfprintf didn't handle this case.
Solution:
Added code which checks if there's an "ll" modifier on the template.
If so, then it uses that as the modifier.
Platforms tested:
Kelgia and Dangermouse
Bug fix
Description:
'll' type description wasn't being handled correctly in HDfprintf, causing
testing failures.
Solution:
Added tests for 'll' to HDfprintf's type description parsing.
Platforms tested:
Solaris 2.7 (arabica)
Code cleanup
Description:
Several system header files needed to be included on Linux machines to
get the prototypes for some of the system calls we are using.
Solution:
Added a new feature test macro for including the <features.h>
header on Linux machines, to set up compile macros for system header
files.
Platforms tested:
Linux (eirene), FreeBSD 4.3 (hawkwind)
Code cleanup
Description:
Get ready for adding C99 extensions sometime in the future.
Solution:
Added some comments to the gcc flags..
Platforms tested:
Linux (eirene), FreeBSD 4.3 (hawkwind)
Code cleanup
Description:
Several system functions don't have prototypes on Linux machines with our
current compile flags.
Solution:
Add _POSIX_SOURCE and _BSD_SOURCE macros to the compile line when compiling
on Linux machines, to pick up missing prototypes.
Platforms tested:
Linux (eirene), FreeBSD 4.3 (hawkwind)
Bug Fix
Description:
HDfprintf wasn't aware of 'L' type modifier and was passing along the
'%Ld' verbatim instead of printing a long long integer value.
Solution:
Added 'L' to the type modifiers checked in HDfprintf
Platforms tested:
Linux (eirene)
POSIX Removal
Description:
Removed calls to POSIX functions and replaced them with the HD*
equivalents.
Solution:
Did just that. Also had to create two new HD* macros, the
gettimeofday() and getrusage()
Platforms tested:
Linux
