This avoids some atomics by using counter_u64 for TX and relying on
existing single-threading (single ithread per rxq) for RX.
Details
Details
Diff Detail
Diff Detail
- Repository
- rG FreeBSD src repository
- Lint
Lint Not Applicable - Unit
Tests Not Applicable
Event Timeline
sys/dev/cxgbe/t4_sge.c | ||
---|---|---|
1252 | A question here is if we want to have alloc_ofld_txq and free_ofld_txq helper functions in t4_sge.c as we currently have for offload RX queues. If so, I would probably go back and do that in the previous commit that introduced struct sge_ofld_txq |