[BUG] JACCL RDMA consecutive send/recv point-to-point with different shape produces wrong data or hangs.

Describe the bug
Consecutive send/recv point-to-point calls between two nodes with different shapes fail on the JACCL backend. Either I get wrong data sent/received or hang my system. It seems that adding an all_sum barrier before the send/recv fixes it. Ring backend always works fine with the same complex sequences of communications.

To Reproduce
I put in a repo the detailed setup instructions on my system, using conda and the mlx config setup. I added all the hostfiles generated by the setup and two test scripts (test_fail.py / test_pass.py):
https://github.com/dev-tb5-tester/rdma_jaccl

Expected behavior
Complex series of send/recv on jaccl backend are not currently working as expected. Prepending a barrier before each couple of point-to-point communication should not be required.

Desktop (please complete the following information):

• OS Version: MacOS 26.3
• Environment: M4 Max + M4 Pro, Thunderbolt 5, RDMA enabled.
• MLX version: 0.30.6

Additional context
Please review reproduction details and results on README

Start with the external reproduction repository's README and run test_fail.py and test_pass.py to confirm the JACCL behavior. Then trace the JACCL point-to-point send/recv implementation and its handling of consecutive calls with different shapes. Done means the reproduced sequences transfer correct data and do not hang without requiring an all_sum barrier.