We have confirmed there is a bug introduced in the Intel compiler, version 14,
which causes the tests to fail with the option --enable-fortran2003. You can
see the discussion at:
http://software.intel.com/en-us/forums/topic/498698
No time frame was given for this bug being fixed by Intel,
Thank you for the report, H5F_provisional was missing from HDF5mpi.f90, which
is used when --enable-parallel is specified. The fix will be in the next
release, but for now the patch is:
--- HDF5mpio.f90(revision 24616)
+++ HDF5mpio.f90(working copy)
@@ -28,6 +28,7 @@
MODULE
According to Intel, the bug fix will be in Update 2, due late January or early
February. Until then, --enable-fortran2003 will not work with Intel 14.
Scot
From: Scot Breitenfeld
Sent: Monday, January 06, 2014 12:16 PM
To: hdf-forum@lists.hdfgroup.org
Subject
Assuming you have some derived type and array:
TYPE CmpField_struct
INTEGER :: a
REAL :: b
REAL :: c
ENDTYPE CmpField_struct
TYPE(CmpField_struct), DIMENSION(1), TARGET :: cf
You can get the offset of a, b, or c of the derived type by:
offset =
I've attached a modified version of h5ex_t_vlstring_F03.f90 to use attributes
instead. This example assumes you have a F2003 compiler and you used
--enable-fortran2003 when you built hdf5. Hope this helps.
Scot
From: Hdf-forum
only for Fortran
2003? I couldn't deduce it from
http://www.hdfgroup.org/HDF5/doc/RM/RM_H5A.html#Annot-Write
Regards,
Pierre
On Wed, Feb 05, 2014 at 05:43:37PM +, Scot Breitenfeld wrote:
I've attached a modified version of h5ex_t_vlstring_F03.f90 to use attributes
instead. This example
I could reproduce the problem you reported and we are looking into it.
It works if you disable shared, --disable-shared, and build the static
libraries. It also works for --enable-shared if you remove --enable-cxx, but
I'm assuming you need C++ support.
Scot
We believe libtool is getting generated with several blank '-l''s in its
postdeps variable due to mvapich's mpif90 wrapper. If you compare the output
from mvapich's mpif90:
mvapich2-2.0b/mvapich2/bin/mpif90 -shared -v
it lists the gfortran library as -l gfortran, with a space.
However, if
Hi Sean,
This is on our to-do list (the fix is slated for the next release) as we have
not yet incorporated any of the intrinsics from the Fortran 2008 standard.
Scot
On May 8, 2014, at 3:10 PM, Sean Byland se...@cray.commailto:se...@cray.com
wrote:
When building HDF5/1.8.13 with
On Jul 10, 2014, at 6:08 PM, Angel de Vicente ang...@iac.es wrote:
Hi,
Angel de Vicente ang...@iac.es writes:
I'm having a hard time trying to figure out what could be causing the
slow I/O behaviour that I see: the same code (in Fortran) run in three
different clusters behaves pretty
/projects/darshan/
Scot
On Jul 22, 2014, at 1:31 AM, Angel de Vicente
ang...@iac.esmailto:ang...@iac.es wrote:
Hi Scot,
thanks for trying this out.
Scot Breitenfeld brtn...@hdfgroup.orgmailto:brtn...@hdfgroup.org writes:
Can you be more specific about the hardware and the software you are
using
My bug report to gcc.gnu deals with this issue:
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48404
In summary, there is a mismatch of the LD_LIBRARY_PATH and the gcc version used
to compile the code on your OS.
The issue with the intel compiler is new, what is the output from the attached
0 0
0 0 0 0
skolio:~ dlane$
--- Dave
On 28/08/2014 1:49 PM, Scot Breitenfeld wrote:
My bug report to gcc.gnu deals with this issue:
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48404
In summary, there is a mismatch of the LD_LIBRARY_PATH
Have you looked at the examples from,
http://www.hdfgroup.org/HDF5/examples/api18-fortran.html
namely,
Read / Write String Datatype (Dataset) h5ex_t_string_F03.f90
Scot
On Dec 24, 2014, at 9:46 AM, Mark Carter
mcar...@markcarter.me.ukmailto:mcar...@markcarter.me.uk wrote:
I'm new to HDF5
to declare the variable in the subroutine call but it doesn't work
either.
Is there a specific way of doing it?
Thank you
On 4/2/15 17:05, Scot Breitenfeld wrote:
h5tset_size is expecting an integer(size_t) in the second argument, but you are
passing and integer (constant 18), instead pass INT
This issue will be fixed in the next release of HDF5 (the fix is already in the
trunk), but for now you will have to use the Fortran compiler flag
-assume nostd_value
when compiling HDF5 using the Intel 15 compiler.
Scot
On Feb 4, 2015, at 12:18 PM, Thompson, Matt (GSFC-610.1)[SCIENCE
h5tset_size is expecting an integer(size_t) in the second argument, but you are
passing and integer (constant 18), instead pass INT(18,size_t).
Scot
On Feb 4, 2015, at 9:34 AM, Fabio Berzaghi f...@uw.edu wrote:
Dear All,
I am not able to find what throws this error when compiling this
likely is size_t was 4 bytes on the
machine you were using. Glad it works now.
Scot
On 4/2/15 20:15, Scot Breitenfeld wrote:
The argument needs to be a scalar, so pass test(1), test(2), etc… If you have
multiple calls to h5tset_size then you will have to do this in a loop, and you
There are a few examples at:
http://www.hdfgroup.org/HDF5/examples/api18-fortran.html
which you can start from. Take a look in the “Datatypes” section, particularly
at h5ex_t_floatatt_F03.f90
Scot
On Mar 28, 2015, at 4:10 AM, Marco Lucchesi m.lucchesi.1...@gmail.com wrote:
Dear all,
Could you send the config.log file? What system is this for?
Thanks,
Scot
On May 6, 2015, at 7:26 AM, Andrea Negri negri.an...@gmail.com wrote:
Hi all,
I have compiled the hdf5-1.8.14 library with ifort 15.0.1 20141023 and
gcc 4.1.2, with the flags --enable-fortran --enable-fortran2003.
Good point Matt, you will have to use FCFLAGS=-assume nostd_value when using
intel 15 with 1.8.14. This issue has since been fixed in 1.8.15, so you will no
longer need the flag.
Scot
On May 6, 2015, at 8:16 AM, Thompson, Matt (GSFC-610.1)[SCIENCE SYSTEMS AND
APPLICATIONS INC]
You can have a look at the program h5ex_t_string_F03.f90 from:
https://www.hdfgroup.org/HDF5/examples/api18-fortran.html
You should be able to remove the writing section of the example and just modify
the reading part in order to read your dataset.
Scot
On Jul 7, 2015, at 2:42 PM,
If you are reading a variable length string with H5LTget_attribute_string
(H5T_VARIABLE) then you don’t need to allocate the string, just pass in a
pointer and the library will handle the allocations. If you are reading a fixed
length string then you need to allocate a string that is “large
A few things:
(1) If the integers in the dataset cannot be represented by INTEGER*4 (which
I’m assuming is your native INTEGER size) then the values read from the HDF5
file will be wrong. You should be reading them
into an INTEGER*8.
(2) What is the most recent version of hdf you can use
Can you show me how you declared your arguments in h5aread_f. Are you using any
compile flags to promote any integers or reals?
Thanks,
Scot
> On Oct 18, 2015, at 7:51 PM, Richa Mathur wrote:
>
> Hi
> I am trying to read an attribute (N_Number_Of_Scans) that has the
__
> From: Hdf-forum [hdf-forum-boun...@lists.hdfgroup.org] on behalf of Scot
> Breitenfeld [brtn...@hdfgroup.org]
> Sent: Monday, October 19, 2015 3:43 PM
> To: HDF Users Discussion List
> Subject: Re: [Hdf-forum] h5aread_f
>
> Can you show me how you declar
If you are not using h5fc to compile your code then you need to include the
path to the file “hdf5.mod”. This will be found where you installed hdf5
(/opt/hdf5/include in your case). You can also look in bin/h5fc to see what
include dir. you have to include, for example
(0,0): 12
}
Is that an array of dim(1,1) ? or a scalar?
I use
Integer :: ndata_dims, aread_data,error
CALL h5aread_f(attr_id, H5T_NATIVE_INTEGER, aread_data,ndata_dims , error)
Rgds
Gompie
On Mon, Oct 19, 2015 at 10:38 AM, Scot Breitenfeld
<brtn...@hdfgroup.org<mailto:brtn...@h
Sorry for the late response,
> On Sep 29, 2015, at 5:14 AM, Pascal wrote:
>
> Hi,
>
> We are trying to set up a fully open source tool chain to build our software
> and are stuck on hdf5 at the moment.
> When using mingw for compilation I got the error:
> from
>
You can combine/append hyperslabs to select complex regions if that helps, see
1.4.1.1 https://www.hdfgroup.org/HDF5/doc/UG/12_Dataspaces.html, and just have
one read call. Otherwise, you can use point selection, see 1.4.1.2, using one
read call. In general, if the size of your hyperslabs are
Did you also compile the HDF5 library with FCFLAGS = -fdefault-real-8 ? The
problem is the native double in HDF5 does not match the native double in your
program because you promoted real to double precision in your program but not
in the HDF5 library. Try building the HDF5 library with
k you for your fast answer, but I'm not an expert how to build the hdf5
library with this kind of Flag,
i just made:
sudo apt-get install libhdf5-serial-dev
Am 03.09.2015 um 18:55 schrieb Scot Breitenfeld:
Did you also compile the HDF5 library with FCFLAGS = -fdefault-real-8 ? The
problem is the
I was also getting the same error with MOAB from ANL when we were benchmarking
small mesh reads with large number of processors. When I ran on 16384 processes
the job would terminate with:
Out of memory in file
sudo apt-get install libhdf5-serial-dev
Am 03.09.2015 um 18:55 schrieb Scot Breitenfeld:
Did you also compile the HDF5 library with FCFLAGS = -fdefault-real-8 ? The
problem is the native double in HDF5 does not match the native double in your
program because you promoted real to double preci
Does anyone have a good (and current)recipe for building parallel hdf5 on BGQ?
We are trying to encourage a community based repository of build scripts for
those especially *unique* builds at:
https://github.com/brtnfld/build_hdf5
You can try the “build_hdf5” script found there for the
You cannot build shared libraries when enabling Fortran for 1.8 or 1.10 on the
Mac using ifort or gfortran.
We reported the problem to Intel at
https://software.intel.com/en-us/forums/intel-fortran-compiler-for-linux-and-mac-os-x/topic/543831?language=de
but I’m not sure if they’ve fixed the
A minor bug fix release of the High-level Library for handling HDF5 object and
region references was released today (6/21/2016).
The library contains C and Fortran APIs to:
* Get information and read data pointed to by region references
* Create an array of region references using paths
Does the example h5ex_t_cmpd_F03.f90 at
https://www.hdfgroup.org/HDF5/examples/api18-fortran.html
help?
> On Jun 23, 2016, at 7:41 AM, Øyvind Sylta wrote:
>
> Is there a simple Fortran90 instruction on reading a hdf5 dataset that uses
> compound and where the dataset
compiler
Oyvind
-Opprinnelig melding-
Fra: Hdf-forum [mailto:hdf-forum-boun...@lists.hdfgroup.org] På vegne av Scot
Breitenfeld
Sendt: 23 June 2016 15:58
Til: HDF Users Discussion List
<hdf-forum@lists.hdfgroup.org<mailto:hdf-forum@lists.hdfgroup.org>>
Emne: Re: [Hdf-forum] read
hdfgroup.org] På vegne av Scot
Breitenfeld
Sendt: 23 June 2016 18:51
Til: HDF Users Discussion List
<hdf-forum@lists.hdfgroup.org<mailto:hdf-forum@lists.hdfgroup.org>>
Emne: Re: [Hdf-forum] reading compound hdf5 data
If you are using the 1.8 version of hdf5 then you have to enabl
get problems. This is what confuses
me.
Øyvind
Fra: Hdf-forum [mailto:hdf-forum-boun...@lists.hdfgroup.org] På vegne av Scot
Breitenfeld
Sendt: 24 June 2016 18:00
Til: HDF Users Discussion List
<hdf-forum@lists.hdfgroup.org<mailto:hdf-forum@lists.hdfgroup.org>>
Emne: Re: [Hdf-forum] re
On Jan 21, 2016, at 6:55 AM, Pierre de Buyl
> wrote:
Hi,
I am developing a Fortran code that uses HDF5. The Fortran API datatypes listed
at https://www.hdfgroup.org/HDF5/doc/RM/PredefDTypes.html indicate that the
Fortran
Is it possible for you to try the 1.8.16 release and report back if you still
have the same issue? You did call h5open_f, correct?
Scot
On Apr 4, 2016, at 6:41 PM, Bhamidipati, Vikram
> wrote:
Hello,
I am static linking into
This was fixed in the trunk, but it appears it did not make it into 1.10.0. It
should be fixed in the 1.10.0 patch.
Thanks,
Scot
> On May 12, 2016, at 7:13 PM, Matthew Clay wrote:
>
> Hi,
>
> I am transitioning to HDF5 1.10.0 to use some of the new features in the
>
df5 Makefile MANIFEST test
bin CMakeLists.txt config.status CTestConfig.cmake hl
Makefile.am README.txt testpar
Fra: Hdf-forum [mailto:hdf-forum-boun...@lists.hdfgroup.org] På vegne av Scot
Breitenfeld
Sendt: 21 July 2016 19:59
Til: HDF Users Discussion
On Jul 27, 2016, at 1:18 PM, Blankenship, Clay B. (MSFC-ZP11)[USRA]
> wrote:
Greetings...I am trying to read an HDF5 data field in Fortran which is a 16-bit
integer. I have the analogous subroutines working for 4-byte reals and this
ould be fine. It is usually best to use the same compiler vendor
for C and Fortran (export CC=icc, export FC=ifort) if possible.
Øyvind
Fra: Hdf-forum [mailto:hdf-forum-boun...@lists.hdfgroup.org] På vegne av Scot
Breitenfeld
Sendt: 20 July 2016 21:15
Til: HDF Users Discussion List
You may have to include the path/lib to szip or zlib, and you need to link in
the C HDF5 library (-lhdf5), and you may need to link in the high level HDF5
library, -lhdf5hl_fortran -lhdf5_hl
Scot
> On Aug 12, 2016, at 9:59 AM, Akshay Subramaniam wrote:
>
> Hi,
>
> I'm
Hi,
I’ve fixed the issue in 1.8 and the fix will be in the next release of 1.8, or
you can try checking out 1.8 from svn,
https://svn.hdfgroup.uiuc.edu/hdf5/branches/hdf5_1_8
Also, we don’t have access to a NAG compiler, so if you can compile 1.8, run
the fortran tests (i.e. in the fortran
Also, the PT/TB APIs are high-level HDF5 calls, meaning they use a series of
low-level HDF5 APIs to create the PT/TB. If you are curious to see what
low-level HDF5 APIs are needed, for example, to create a PT, then you can look
in the HDF5 source, hl/src/H5PT.c —> H5PTcreate.
Scot
On Feb
Yes it is still the case that you cannot enable C++ or Fortran (or the High
Level APIs) when threadsafe is enabled. —enable-unsupported can override this
behavior.
Scot
> On Sep 22, 2016, at 12:36 PM, Elvis Stansvik
> wrote:
>
> 2016-09-22 19:23 GMT+02:00
> On Aug 23, 2016, at 10:42 AM, Schluenzen, Frank <frank.schluen...@desy.de>
> wrote:
>
> Hi,
>
> - Original Message -
>> From: "Scot Breitenfeld" <brtn...@hdfgroup.org>
>> To: "HDF Users Discussion List" <hdf-forum@lists.hdfgr
How far in the program does it get before it segfaults? Can you post the output
from the program?
I tried a sample program on Windows 10, Intel 17, and HDF5 1.8.17, but I’m
having issues with the HDF5 Fortran parameters not getting set when building
the program inside Visual Studio. Can you
We have noticed this behavior in the past.
I tried on my macOS 10.12.x, with 1_8, 1.8.18 and develop and did not have any
problems. I tried gfortran 6.3.0 with both gcc 6.3 and the system's gcc, but I
have cmake 3.8.0. From your previous email, it looks like cmake failed in
detecting the
Hi,
Take a look at the examples in the “Datatypes” section at:
https://support.hdfgroup.org/HDF5/examples/api-fortran.html
namely (for fixed sized string), h5ex_t_stringCatt_F03.f90.
Scot
> On May 16, 2017, at 9:55 AM, Greenberg, Naomi
> wrote:
>
> I am
denes
>> On 05/20/2017 09:12 PM, Quincey Koziol wrote:
>>>
>>>> On May 19, 2017, at 12:32 PM, Håkon Strandenes <haa...@hakostra.net> wrote:
>>>>
>>>> Yes, the issue is still there.
>>>>
>>>> I will try to make a dummy
Can you try it with 1.10.1 and see if you still have an issue.
Scot
> On May 19, 2017, at 1:11 PM, Quincey Koziol wrote:
>
> Hi Håkon,
>
>> On May 19, 2017, at 10:01 AM, Håkon Strandenes wrote:
>>
>> (sorry, forgot to cc mailing list in prev. mail)
>>
Hi Ricardo,
I just did a quick glance, and I noticed a few things.
(1) Have you considered passing the C address of the array arguments instead to
your functions? That would simplify your interfaces and remove a lot of code
duplication. Also, using the F2003 HDF5 functions instead would also
Did you patch the Fortran source for Intel 18?
https://software.intel.com/en-us/forums/intel-fortran-compiler-for-linux-and-mac-os-x/topic/747951
Scot
On Nov 30, 2017, at 8:13 AM, 江捷
> wrote:
Hello Barbara,
Thank you for your reply.
Thanks for the report Robert,
We’ve reproduced the issue, and it's slated to be fixed in the next release of
HDF5 1.10.
Scot
> On Dec 9, 2017, at 3:38 AM, Robert Sawko wrote:
>
> Dear HDFers,
>
> I am building a parallel version of HDF5 sub-version 1.10.1. I went with the
lists.hdfgroup.org>
When replying, please edit your Subject line so it is more specific
than "Re: Contents of Hdf-forum digest..."
Today's Topics:
1. Memory allocation/deallocation (Andreas Derler)
2. Re: no specif
target 'check-am' failed
make[1]: *** [check-am] Error 2
make[1]: Leaving directory '/home/ustc/下载/hdf5-1.10.1/testpar'
Makefile:650: recipe for target 'check-recursive' failed
make: *** [check-recursive] Error 1"
I have attached the configure.out, make.out, check.out and config.log.
Bests Reg
Can you send the config.log file, it should contain the details of the error.
Scot
On Jan 26, 2018, at 3:57 PM, Paul Arias
> wrote:
Hello,
I am trying to compile HDF5 (1.8.20) with openmpi, fortran (gnu).
This is the error message that I
If you are looking for an FEA “standard” I/O format, it does not exist. CGNS is
as close as it gets, but that is for CFD data. There has been some discussion
in the past to extend CGNS to include Solid Mechanics, but it never gained
traction. For FEA data formats which are geared towards solid
64 matches
Mail list logo