[cam-users] segmentation fault in CAM2.0.1

[Jim Rosinski]

On Wed, 5 Feb 2003, Schwebach, Todd D. wrote:

> I am able to successfully compile CAM2.0.1 on a dual processor Linux machine
> with SMP enabled using the PGI 4.0 compiler. The model runs for a while and
> then outputs:
>
> ************************************************************
>
>  SSTINI: Read sst data for dates           816        43200  and
> 916
>             0
>  ICEINI: Read ice data for dates           816        43200  and
> 916
>             0
> t_start thread 0: timer tphysbc was already on: not restarting.
> t_start thread 0: timer zm_convr was already on: not restarting.
> t_stop: timer zm_convr was already off.
> t_start thread 0: timer pcond was already on: not restarting.
> t_stop: timer pcond was already off.

Don't have enough info to comment on the seg fault, but the timer complaints
are almost always a result of building the timer codes with threading
disabled but then running the model with threading enabled.  The timer codes
are in subdirectory cam1/models/utils/timing after untarring the
distribution.  Possibly this inconsistency (threading disabled in one part of
the code and enabled in another) is the cause of the seg fault?

Jim Rosinski