yangsong8-a1 commented on PR #21588:
URL: https://github.com/apache/nuttx/pull/21588#issuecomment-6761728097

   > [@yangsong8-a1](https://github.com/yangsong8-a1) please test on real board 
to confirm the I2C and I3C still working correctly. If a new device get 
hot-plugged like QWIIC (for I2C) and QWIIIC (for I3C, it will exist someday), 
what will happen? Did you test this usage scenario?
   
   Hi @acassis, thanks for the review.
   
   This only affects controllers with a negative i3c_bus_id, which was not
   supported before (i3c_register() asserts on it). Such a controller is now
   used as an I2C-only controller. Controllers with i3c_bus_id >= 0 are
   unchanged.
   
   Hot-plug:
   - I2C (QWIIC): a new device is reachable as soon as it is addressed. This
     path is not touched.
   - I3C (QWIIIC): the I3C core has no generic Hot-Join support. The
     controller driver has to call i3c_master_do_daa(), and that path is
     unchanged.
   
   Tested on real hardware (I2C-only bus), plus a sim comparison test with a
   fake I3C controller. The sim test also checks the software side of
   hot-plug: a newly added I2C device is reachable without re-registration,
   and on an i3c_bus_id >= 0 bus a new I3C device is picked up when the
   driver calls i3c_master_do_daa(). Electrical hot-plug and real Hot-Join
   signaling are not covered. Details are in the PR description. The sim
   test also found a cleanup leak on unregister, fixed in the second commit.
   


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]

Reply via email to