Hi Robin,
Apologies for late the delay, I'll reply to the original email
On 10/10/2025 9:55 AM, Robin Jarry wrote:
Robin Jarry, Jun 10, 2025 at 16:53:
Currently, for the TRIE algorithm (actually, it should be called
DIR-24-8-8-8-8-8-8-8-8-8-8-8-8), the tbl8 group is freed even though the
readers might be using the tbl8 group entries. The freed tbl8 group can
be reallocated quickly. As a result, lookup may be performed
incorrectly.
To address that, RCU QSBR is integrated for safe tbl8 group reclamation.
Cc: Vladimir Medvedkin <[email protected]>
Signed-off-by: Robin Jarry <[email protected]>
---
Notes:
This is a semi-copy-paste of the FIB4 implementation.
I couldn't understand the implementation of trie_modify with regard to
depth_diff handling.
The unit tests fail because depth_diff is always 0 when deleting a route
which causes any subsequent add to fail with a -ENOSPC error.
Vladimir, could you give some more insights on the matter?
Gentle bump. Hi Vladimir, could you have a look? Thanks!
--
Regards,
Vladimir
