On 5/20/2025 2:12 PM, Johannes Berg wrote:
On Sun, 2025-05-18 at 22:04 +0530, Sarika Sharma wrote:
I reviewed the code again to check for pertid and realized that it is
applicable at the MLO level as well. The rx, tx fields are reported per
link (which can be accumulated for the MLO level), while the TXQ_STATS
related fields are applicable for station level meaning MLO level.
Therefore, pertid is applicable at MLO level and also at link level for
rx/tx fields.
Hm OK? Not sure I completely follow. Still seems mac80211 should only
fill that up with accumulated-from-removed-links stats?
And now it gets pretty tricky since we have some stats that are
accumulated from removed links, and some that aren't?
Then it seems I need to maintain all accumulated fields data to be
filled from accumulated-from-removed-links stats first, that will
automatically reset if some deflink values are there and fill with
removed links stats.
johannes
