[PATCH RFC/RFT] ath10k: use xmit encapsulation offloading

[Edward Matijevic]

Tested on a TP-Link Archer C2600 (IPQ8064 + QCA9980), using official
22.03-rc1 and your patch on top of 22.03
Load average peaks were still the same but with the offloading 1min
avg wasn't changing as sharply as without offload
My download speed was a bit more stable with the offload, tested with
a game download from Steam(20+ GB) maintaining over 40MB/s when it'd
only sometimes peak over 40.
Bufferbloat(from Waveform) latency testing wasn't noticeably
different(maybe a slight improvement to minimums), without SQM on
both.

No obvious regressions but haven't run the patch long enough yet to see issues