exit(3) and the RTEMS Shell

alex levy alevy at planetiq.com
Thu Jan 30 17:08:37 UTC 2020

Thanks for the quick response, Joel.

We have BSP_PRESS_KEY_FOR_RESET set to 0.  Also, note that we have
installed a user fatal error extension, but the keypress must occur before
the extension will run.

And, just to reiterate what Jonathan said, this only occurs some of the
time.  Most of the time our extension does run, and the system goes down as
designed without any keypress.


On Thu, Jan 30, 2020 at 9:46 AM Joel Sherrill <joel at rtems.org> wrote:

> On Thu, Jan 30, 2020 at 10:09 AM Jonathan Brandmeyer <
> jbrandmeyer at planetiq.com> wrote:
>> We're seeing some unexpected behavior.  It appears as though sometimes
>> the exit() function will wait until a key is pressed on the shell before
>> actually shutting down RTEMS.  This is independent of whether or not the
>> exit()ing thread is the shell thread or a different one.
>> We're using the zynq BSP, anchored to master as of a few months ago.
> This is a common configuration setting available to all BSPs.  The file
> c/src/aclocal/bsp-bspcleanup-options.m4 has the arguments that control
> the behavior but a specific BSP can provide defaults. The defaults may
> change
> over time and for different BSPs which might explain what you are seeing.
> RTEMS_BSP_CLEANUP_OPTIONS and the other macros are used from
> the BSP's configure.ac.
> You should be able to change the behavior by adding something like
> BSP_PRESS_KEY_FOR_RESET=0 to the RTEMS configure command
> like this
> ../rtems/configure BSP_PRESS_KEY_FOR_RESET=0 ...
> --joel
>> Thanks,
>> --
>> Jonathan Brandmeyer
>> PlanetiQ
>> _______________________________________________
>> users mailing list
>> users at rtems.org
>> http://lists.rtems.org/mailman/listinfo/users
> _______________________________________________
> users mailing list
> users at rtems.org
> http://lists.rtems.org/mailman/listinfo/users
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.rtems.org/pipermail/users/attachments/20200130/d04dc3a5/attachment.html>

More information about the users mailing list