Chris Johns commented on a discussion: 
https://gitlab.rtems.org/rtems/pkg/rtems-libbsd/-/merge_requests/168#note_160954


If the interface is present in other BSPs then I suggest the way it is in 
LibBSD for qoriq is broken and upside down. The decl and hook pointer should be 
in the BSP and ant call fails with an error unless hooked. If a BSP like qoriq 
defers it implementation to LibBSD it would register its function when LibBSD 
initializes.

I suggest:

* Add a call `rtems_clear_vmebridge_errors` to the shared part of the BSP 
source tree. Include any similar functions.
* Add a register call `rtems_register_clear_vmebridge_errors` 
* Have the BSPs in `rtems.git` register their `_BSP_clear_vmebridge_errors` 
calls in the BSP startup.
* Change all references to call the new function

_Note:_ `7-freebsd-14` has brought across more of FreeBSD's PCIe support. I am 
not sure how this effects a BSP like qoriq but it will need to be maintained 
and brought up to date if there are clashes or issues.

-- 
View it on GitLab: 
https://gitlab.rtems.org/rtems/pkg/rtems-libbsd/-/merge_requests/168#note_160954
You're receiving this email because of your account on gitlab.rtems.org. 
Unsubscribe from this thread: 
https://gitlab.rtems.org/-/namespace/20/sent_notifications/6-7zymwx1vak8ggmvuve21amhhb-k/unsubscribe
 | Manage all notifications: https://gitlab.rtems.org/-/profile/notifications | 
Help: https://gitlab.rtems.org/help


_______________________________________________
bugs mailing list
[email protected]
http://lists.rtems.org/mailman/listinfo/bugs

Reply via email to