[PATCH 2/2] ath10k: do not use coherent memory for tx buffers

Kalle Valo kvalo at qca.qualcomm.com
Wed Nov 25 02:37:08 PST 2015

Felix Fietkau <nbd at openwrt.org> writes:

>>>   However, on the device that I'm testing on
>>> (IPQ806x based), this patch makes the difference between working and
>>> non-working wifi, fixing the regression introduced by your pre-allocated
>>> coherent memory patch.
>> Thank you for the catch up and fix.
>> Btw, the regression can be fixed by using GFP_KERNEL, instead of 
>> GFP_DMA, right?
> I just did some timing measurements, and it seems that the DMA coherent
> variant is roughly 200 nanoseconds faster. Maybe the extra latency is
> caused by the CPU filling the cacheline from RAM first.
> Kalle, please only merge the first one and drop this patch.
> I will send a replacement for it.

Ok, patch 2 dropped.

Kalle Valo

More information about the ath10k mailing list