[FS#174] Significant MT76 degradation after recent fq reordering changes

LEDE Bugs lede-bugs at lists.infradead.org
Thu Nov 3 07:08:34 PDT 2016


The following task has a new comment added:

FS#174 - Significant MT76 degradation after recent fq reordering changes
User who did this - Andrei Măceș (airend)

----------
I tested [[https://git.lede-project.org/?p=source.git;a=tree;h=12bd0f282005c57b9ec144c531390e2fa8cd1590;hb=12bd0f282005c57b9ec144c531390e2fa8cd1590| 12bd0f282005c57b9ec144c531390e2fa8cd1590]], and basic operation is improved; 11n/40MHz clients have more consistent download (upload seemed always fine/like before fq), but the issue remains with the 4-addr/bridged setup, where it maxes out around 20Mbps (all 5GHz, of course).

P.S. I also tried [[https://github.com/openwrt/mt76/commit/adf1bbd03cc7e1117d44b3da4e584f6dd590c4d0|latest mt76]], which didn't work well at all on current stack, fq or not.

----------

More information can be found at the following URL:
https://bugs.lede-project.org/index.php?do=details&task_id=174#comment752



More information about the lede-bugs mailing list