Ayah Kashkoul commented: 
https://gitlab.rtems.org/rtems/rtos/rtems/-/issues/3791#note_143924


Hi @aptacc2421, @joel ,@gedare !

I’ve been following the discussion around #3791 and @aptacc2421 ’s proposed 
per-instance `mq_ispriority_np` approach. The direction makes a lot of sense, 
especially preserving FIFO as the default for backward compatibility.

I’m also planning to work on the Message Queue Improvements and Fixes project 
for GSoC, and I’d like to contribute by focusing on independent validation and 
regression testing for this change rather than re-implementing the feature 
itself.

My plan is to check out @aptacc2421 ’s !1084, verify behavior on RTEMS 7 
(erc32-sis), and develop a psxtest that:

* Confirms FIFO remains the default behavior
* Confirms `mq_ispriority_np` opt-in wakes the highest-priority waiting thread
* Helps guard against future regressions in wake-up discipline

I’ll report my findings here and open an MR for the tests once ready. Please 
let me know if this sounds aligned with the intended direction!

-- 
View it on GitLab: 
https://gitlab.rtems.org/rtems/rtos/rtems/-/issues/3791#note_143924
You're receiving this email because of your account on gitlab.rtems.org.


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

Reply via email to