>>Any ideas what happened? It seems to be a small block of
>> text in the header of the mdout file and then some (significant)
>> differences in the last frame...
OK, on it...
Should be easy to trace the diffs.
Josh
On 12/02/2014, Jason Swails <jason.swails.gmail.com> wrote:
> Hi Josh,
>
> I think a lot of the pmemd TI tests are failing as a result of your
> recent commit. If I back my tree up to
> e1d5f28d18a0ab40a332cf58e0c1db83848a0753 (right before your commits)
> then all of the tests pass. If I use the current HEAD, I get the
> following diffs (attached). They seem to be both numerical and
> formatting. Any ideas what happened? It seems to be a small block of
> text in the header of the mdout file and then some (significant)
> differences in the last frame...
>
> For reference, I'm using the Intel compilers version 13.1.3 on an AMD
> machine.
>
> Thanks!
> Jason
>
> P.S. The attached file contains all of the diffs from both sander and
> pmemd-non-TI, but the TI ones are in there closer to the end.
>
> On Wed, 2014-02-12 at 08:40 +0100, Josh Berryman wrote:
>
>> Erm, I've got a big-ish emil commit that integrates Joe Kaus's emil_sc
>> work as well as tidying up the i/o (hence the docs and all emil test
>> cases are altered) and fixing a bug.
>>
>> cruisecontrol is green at the moment, so I will probably have pushed
>> by the time the sun rises in California.
>>
>>
>>
>>
>> On 11/02/2014, Ross Walker <ross.rosswalker.co.uk> wrote:
>> > Yeah this is actively being worked on right now (by several people all
>> > in
>> > the same room) - hence the mutual problem with us all trying to do
>> > things
>> > last minute. :-(
>> >
>> >
>> > On 2/11/14, 2:19 PM, "Jason Swails" <jason.swails.gmail.com> wrote:
>> >
>> >>
>> >>
>> >>> On Feb 11, 2014, at 5:05 PM, Ross Walker <ross.rosswalker.co.uk>
>> >>> wrote:
>> >>>
>> >>> Test Cases???
>> >>>
>> >>> And does this work with TI? What about IPS? - Don't see any
>> >>> protections
>> >>> for that.
>> >>>
>> >>> get_nb_energy.i, pairs_ips_calc.i and pairs_calc.i are about to get a
>> >>> massive overhaul so without any test cases there is no guarantee this
>> >>> addition will be working afterwards.
>> >>
>> >>It's all coming. It was about to get merged (test cases, GPU checks,
>> >>everything) and then you ripped out a ton of code and that got buggered
>> >>(sound emil-iar?). Pengfei is adding it back piece at a time and I'll
>> >>move over the test cases tonight.
>> >>
>> >>_______________________________________________
>> >>AMBER-Developers mailing list
>> >>AMBER-Developers.ambermd.org
>> >>http://lists.ambermd.org/mailman/listinfo/amber-developers
>> >
>> >
>> >
>> > _______________________________________________
>> > AMBER-Developers mailing list
>> > AMBER-Developers.ambermd.org
>> > http://lists.ambermd.org/mailman/listinfo/amber-developers
>> >
>>
>> _______________________________________________
>> AMBER-Developers mailing list
>> AMBER-Developers.ambermd.org
>> http://lists.ambermd.org/mailman/listinfo/amber-developers
>
> --
> Jason M. Swails
> BioMaPS,
> Rutgers University
> Postdoctoral Researcher
>
_______________________________________________
AMBER-Developers mailing list
AMBER-Developers.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber-developers
Received on Thu Feb 13 2014 - 02:00:02 PST
