On 25/04/2019 23:49, Jouni Malinen wrote:
On Thu, Apr 25, 2019 at 04:57:48PM +0200, John Crispin wrote:
The current code will always use the size required when all optional
elements are present. This will cause the linux kernel to consider the
field to be malformed if the elements are not actually flagged as being
present.
Thanks, applied. I left patch 2/2 into my queue for some more review and
testing.
By the way, do you have means for testing these properly with
mac80211_hwsim? I added a minimal test case to check that the HE
elements show up in Beacon frames, but it did not look like the issue
mentioned in this patch 1/2 was showing up there and the station side
does not show any sign of trying to use HE either.
Hi Jouni,
the kernel will silently fallback to VHT and not have the he_supported
flag set after the assoc succeeded.
-->
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/net/mac80211/mlme.c?h=v5.1-rc6#n4779
I have been doing testing using hwsim but moved on to using 2 hawkeye
boards as I am doing the ath11k support in parallel.
John
_______________________________________________
Hostap mailing list
Hostap@xxxxxxxxxxxxxxxxxxx
http://lists.infradead.org/mailman/listinfo/hostap
