From: Joachim Hein (joachim_at_epcc.ed.ac.uk)
Date: Mon Nov 21 2005 - 04:12:05 CST
Hi,
Just a quick reply:
We are observing exactly the same on a p575 (one should not call this an
SP4). We also have problems with namd2.5 using AIX 5.2 on a p690+. We
are investigating this for some time already.
I have meet Jim and Gengbin on a conference last week. They made a number
of proposals to rectify the situation, which I now need to try. We will
keep you posted on the progress.
Best wishes
Joachim
On Mon, 21 Nov 2005, Sascha Tayefeh wrote:
>
> Hi!
>
> I have encountered a problem running namd2.6b1 on IBM-SP4 (Type IBM p575)
> with AIX53 installed. (uname -a: AIX nodeA 3 5) I thought this could be a
> vital information for the developers an the community, so I'd like to share
> this with you.
>
> There seems to be a problem concerning the memory allocation. This problem
> occurs only with these architecture/os-version and namd2.6b1 (both, with the
> binaries you provided and with our own binaries compiled from your
> sourcecode). Namd2.5 runs perfectly with this architecture/os-version. On the
> other hand, namd2.6b1 encounters no problems running on IBM-SP4 (Type p655+)
> with AIX5*2* installed (uname -a: AIX nodeB 2 5), neither does namd2.5.
>
> Here is the machine's errorlog when running namd2.6b1:
>
> <BEGIN>
> ATTENTION: 0031-408 8 tasks allocated by LoadLeveler, continuing...
> ------------- Processor 2 Exiting: Called CmiAbort ------------
> Reason: CthResume: swapcontext failed.
>
> ------------- Processor 4 Exiting: Called CmiAbort ------------
> Reason: CthResume: swapcontext failed.
>
> ------------- Processor 5 Exiting: Called CmiAbort ------------
> Reason: CthResume: swapcontext failed.
>
> ------------- Processor 6 Exiting: Called CmiAbort ------------
> Reason: CthResume: swapcontext failed.
>
> ------------- Processor 7 Exiting: Called CmiAbort ------------
> Reason: CthResume: swapcontext failed.
>
> ------------- Processor 0 Exiting: Called CmiAbort ------------
> Reason: CthResume: swapcontext failed.
>
> ERROR: 0031-250 task 2: Terminated
> ERROR: 0031-250 task 1: Terminated
> ERROR: 0031-250 task 3: Terminated
> ERROR: 0031-250 task 5: Terminated
> ERROR: 0031-250 task 6: Terminated
> ERROR: 0031-250 task 0: Terminated
> ERROR: 0031-250 task 4: Terminated
> ERROR: 0031-250 task 7: Terminated
> <END>
>
> And here's what namd2.6b1 (custom compile) complains (the binary you provide
> just stops without complaining):
>
> <BEGIN>
> ....
> FATAL ERROR: Memory allocation failed on processor 3
> FATAL ERROR: Memory allocation failed on processor 2
> FATAL ERROR: Memory allocation failed on processor 4
> FATAL ERROR: Memory allocation failed on processor 6
> FATAL ERROR: Memory allocation failed on processor 7
> FATAL ERROR: Memory allocation failed on processor 1
> FATAL ERROR: Memory allocation failed on processor 8
> <END>
>
>
> Sincerly
>
> Sascha Tayefeh
>
*******************************************************************************
dr joachim hein
epcc
the university of edinburgh
rm 3406 jcmb
mayfield road
edinburgh eh9 3jz tel: 0131 651 3390
scotland, uk fax: 0131 650 6555
*******************************************************************************
This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:40:10 CST