We re-opened it, though. :) I saw Nathan fix it in the c bindings; I'm not
sure if he fixed it in fortran yet. It's noted on the pull request, though.
Yes, if you'd like to file directly on Github, that would be great.
Sent from my phone. No type good.
> On Feb 11, 2016, at 2:49 PM, Lisandr
On 11 February 2016 at 14:41, Jeff Squyres (jsquyres)
wrote:
> Nope, this is not on purpose. I filed
> https://github.com/open-mpi/ompi/issues/1355 to track the issue.
>
Oh! I was not aware you are now tracking issues in GitHub. I think you
closed the issue too quickly :-) I added some addition
That is kind of odd. I believe that I put the tailing slash “/“ at the end of
the redirecting URL.
Anyway, I just restarted the apache daemon and it seems to be working fine now.
Maybe I did not restart the daemon after adding the tailing slash?
Regards,
--
- DongInn
> On Feb 11, 2016, at 11
DongInn --
When you enabled the https redirects for mtt.open-mpi.org, it looks like there
is a / missing in the redirect.
> On Feb 11, 2016, at 11:49 AM, Howard Pritchard wrote:
>
> Hi Folks
>
> When I go to
>
> https://mtt.open-mpi.org/
>
> and then click the summary button I get some kin
Hi Folks
When I go to
https://mtt.open-mpi.org/
and then click the summary button I get some kind of DNS lookup error.
Howard
Despite working for communicators and windows, setting a NULL
attribute value in datatypes fails with MPI_ERR_ARG. Run the attached
test case to reproduce.
--
Lisandro Dalcin
Research Scientist
Computer, Electrical and Mathematical Sciences & Engineering (CEMSE)
Numerical Porous Med
Nope, this is not on purpose. I filed
https://github.com/open-mpi/ompi/issues/1355 to track the issue.
Thanks!
> On Feb 11, 2016, at 3:15 AM, Lisandro Dalcin wrote:
>
> After writing some tests, I discovered Open MPI's MPI_Get_address()
> fails if fed with MPI_BOTTOM. Is this on purpose of j
I can’t speak to the packing question, but I can say that we have indeed
confirmed the lack of maintenance on OMPI for Debian/Ubuntu and are working to
resolve the problem.
> On Feb 11, 2016, at 1:16 AM, Gilles Gouaillardet
> wrote:
>
> Michael,
>
> MPI_Pack_external must convert data to big
Michael,
MPI_Pack_external must convert data to big endian, so it can be dumped into
a file, and be read correctly on big and little endianness arch, and with
any MPI flavor.
if you use only one MPI library on one arch, or if data is never
read/written from/to a file, then it is more efficient to
Hi Gilles,
I enhanced my simple test program to dump the contents of the buffer:
If I am not mistaken, it appears that the unpack is not doing the endian
conversion.
kindest regards
Mike
Good:
send data 04d2 162e
MPI_Pack_external: 0
buffer size: 8
Buffer contents
d2, 04, 00, 00, 2e, 16,
Hi Gilles,
thanks for thinking about this in more detail.
I understand what you are saying, but your comments raise some questions in my
mind:
If one is in a homogeneous cluster, is it important that, in the case of
little-endian, that the data be
converted to extern32 format (big-endian), only
Michael,
I think it is worst than that ...
without --enable-heterogeneous, it seems the data is not correctly packed
(e.g. it is not converted to big endian), at least on a x86_64 arch.
unpack looks broken too, but pack followed by unpack does work.
that means if you are reading data correctly wr
After writing some tests, I discovered Open MPI's MPI_Get_address()
fails if fed with MPI_BOTTOM. Is this on purpose of just an error
checking oversight?
$ cat get_address.c
#include
int main(int argc, char *argv[])
{
MPI_Aint addr;
MPI_Init(&argc, &argv);
MPI_Get_address(MPI_BOTTOM, &addr
Hi Ralph,
you are indeed correct. However, many of our users
have workstations such as me, with OpenMPI provided by installing a package.
So we don't know what has been configured.
Then we have failures, since, for instance, Ubuntu 14.04 by default appears
to have been built
with heterogeneous sup
14 matches
Mail list logo
