Hi Rayson
We take snapshots from time to time. We debated whether or not to update
again for the 1.7 release, but ultimately decided not to do so - IIRC, none
of our developers had the time.
If you are interested and willing to do the update, and perhaps look at
removing the limit, that is fine w
Hmm, not sure why I didn't see an error when I tested the change. It looks like
in this case yyterminate should have been defined as
orte_rmaps_rank_file_lex_destroy(). Looked a little deeper and it looks like
the default action for yyterminate is to call the *lex_destroy function so we
don't n
Ok, looks like the default yyterminate does not clean up the lex state. The
definition in rmaps/rankfile/rmaps_rank_file_lex.l should be
#define yyterminate() return orte_rmaps_rank_file_lex_destroy()
I can fix it if you want.
-Nathan
On Wed, Nov 07, 2012 at 08:34:59AM -0700, Nathan Hjelm wrot
Problem is that the orte function no longer seems to exist, so build fails
Sent from my iPhone
On Nov 7, 2012, at 7:55 AM, Nathan Hjelm wrote:
> Ok, looks like the default yyterminate does not clean up the lex state. The
> definition in rmaps/rankfile/rmaps_rank_file_lex.l should be
>
> #defi
Nathan,
Although this typo fix is correct, AFAIK, it is unnecessary to check
for NULL before calling free().
-- Tim "the perpetual OMPI lurker" Mattox
P.S. - I have a coding problem... I'm still watching commits 3+ years
after giving up gatekeeper duties!
On Tue, Nov 6, 2012 at 4:25 PM, wrote:
>
There appears to have been a change in the behaviour of -npersocket from
1.4.3 to 1.6.x (tested with 1.6.2). Below is what I see on a pair of dual
quad-core socket Nehalem nodes running under PBS. Is this expected?
Thanks
David
[dbs900@v482 ~/MPI]$ mpirun -V
mpirun (Open MPI) 1.4.3
...
[dbs90
