| Commit message (Collapse) | Author | Age | Files | Lines |
|---|
| |
|
|
| |
h5test.o, which requires the definition to work...
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The "FILENAME" declared extern in h5test.h is not always used.
It was used in h5_cleanup to remove temporary files created
during tests. Not all tests codes have used this routine.
Indeed, quite a few of test programs do "#define FILENAME ".
Also, h5_cleanup needs to work in tandem with h5_fixname.
h5_fixname accepts an explicite base_name argument instead
of using the global variable FILENAME. That is cleaner.
Solution:
Added char *base_name[] as a new argument to h5_cleanup, in
the same style as h5_fixname. Removed "extern char *FILENAME..."
from use. Also, undo some unnecessary declaration of "char *FILENAME"
from some tests which don't use it at all (yet).
Platforms tested:
modi4-64(irix64), arabica(solari2.7), eirene(linux)
(arabica could not launch tests automatically. I had to hack
in LD_LIBRARY_PATH to make them run.)
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
with
zlib, and zlib isn't in our LD_LIBRARY_PATH or in a standard system
library place (/usr/lib or /lib), then the tests can't run. This fixes it
if the person configures the source with the flag:
--with-zlib=/usr/fnord/include,/usr/fnord/lib
This only works if you're pointing to the shared library version in the
above flag...
|
| |
|
|
|
|
|
| |
libh5test, but
libh5test wants FILENAME to be defined. I have no clue why this was
working before...
|
| | |
|
| |
|
|
|
|
| |
static.
Also I'm keeping the place the user indicated the zlib would be...
|
| |
|
|
|
| |
RM_H5Z.html
Corrected compression cross-reference.
|
| | |
|
| | |
|
| |
|
|
|
|
|
|
|
| |
__cplusplus
macro didn't work on all systems. This fixes it.
This also introduces the ability to configure for fortran and c++. This
is not to be implemented yet, though!
|
| | |
|
| | |
|
| | |
|
| |
|
|
| |
autoheader, it was useless to add them in here...
|
| |
|
|
|
|
|
| |
src/H5{pub}config.h
files. I append to them an #undef of the inline macro. Then, if it's a
C++ compiler, the inline keyword isn't redefined incorrectly...
|
| |
|
|
|
|
| |
in C++
and, when adding the header to a C++ program, breaks it.
|
| | |
|
| |
|
|
|
|
| |
and dynamicly
allocated memory.
|
| |
|
|
|
|
|
|
| |
variable instead of
multi-dereferenced pointer chains. This buys us another ~20% improvement in
the hyperslab I/O speed. (From ~30 seconds to ~25 seconds on the h5hypers
benchmark)
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Bug fixes
Description:
All tests were core=dumping in IRIX64. The bug is in Generic
property list creation in which malloc asked for 2*64-1 bytes
due to coding bug. The object creation failed but the return
code was not checked. Program eventually crashed.
Solution:
H5F.c:
Check the return code from new file object creation and flag
error accordingly.
H5FL.c:
H5FL_arr_free is a replacement for H5MM_xfree which accepts
null value as a legal argument value. H5FL_arr_free assert
on it. Since other parts of the code have been passing null
value to H5MM_xfree, H5FL_arr_free must accept it too until
all the calling routines are changed to not pass Null.
H5P.c:
some routine passes in 0 as the hashsize value which is uintn.
The expression (hashsize-1) underflows to the largest unsigned
int for some machines. Thus the calloc failed. Cast hashsize
to unsigned int first (this assumes hashsize stays within the
signed int data range.
H5Smpio.c:
Added the extra parameter because the H5FD_write has been redefined.
Platforms tested:
IRIX64 -64 and -n32
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
changes
applied for various platforms which weren't supported by libtools are now
standard for libtools it seems. There were only a few of the previous
fixes I rolled forward (the diffs follow):
config.sub:
750a760,763
> tflop*)
> basic_machine=i386-intel
> os=-osf1
> ;;
ltmain.sh:
2911c2911
< $shlibpath_var=\"$temp_rpath\$$shlibpath_var\"
---
> $shlibpath_var=\"$finalize_shlibpath$temp_rpath\$$shlibpath_var\"
|
| |
|
|
| |
error).
|
| |
|
|
| |
FAIL...Fixed.
|
| |
|
|
|
|
|
|
| |
arrays of
hyperslab boundaries after adding them all, instead of maintaining the sorted
order during each addition. This boosts performance for sub-sampled (i.e.
strided) hyperslabs by about a factor of 10! :-)
|
| |
|
|
|
|
| |
garbage collect and
not eat all the memory in the machine (by default :-)
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
structures. Also
added code to allow metadata to be allocated out of a more contiguous block
("metadata aggregation") and also code for "catching" small metadata write
calls and building a buffer of the small pieces of metadata for later writing
as one, larger, block ("metadata accumulation"). These features are enabled
on a per VFL driver basis with the new VFL 'query' call and both currently
enabled for the sec2, family and stdio drivers. The mpio VFL driver enables
only the "metadata aggregation" code, not the "metadata accumulation" code.
All the other drivers have these features turned off.
|
| |
|
|
|
|
| |
aggregation and
accumulation to be enabled and tracked. Also updated some prototypes.
|
| | |
|
| |
|
|
|
|
| |
parameter to write
call.
|
| |
|
|
|
|
| |
(it's in H5FD.c
now) and updated driver to add the new VFL 'query' call.
|
| |
|
|
|
|
| |
field to file
access property lists.
|
| | |
|
| |
|
|
| |
bytes.
|
| |
|
|
|
|
|
| |
set/query the
metadata allocation size for file access property lists. These are new API
functions and should be documented for the next release.
|
| |
|
|
|
|
| |
added the
metadata allocation size to the default file access property lists.
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| |
|
|
| |
R1.2 branch)
|
| |
|
|
| |
R1.2 branch)
|
| | |
|
| |
|
|
|
|
| |
reduce malloc
abuse.
|
| |
|
|
|
|
| |
of code, to
get the metadata allocated in the file more tightly together.
|
| |
|
|
|
|
| |
logging and added
test code for aggregating metadata into more localized locations in the file.
|
| |
|
|
| |
possible.
|
| | |
|
