Altough the patch is already accepted, I have a question regarding the current semantics of the whole function airtime_link_metric_get: My Impression was, that rate = DIV_ROUND_UP(sta_get_expected_throughput(sta), 100); will get the expected throughput from the rate control, so e.g. for ath9k it will receive the already averaged rate from minstrel(_ht). I don't know which hardware you use, but we have only mesh hardware in our lab, which use minstrel, so this code wouldn't be triggered. Unfortunately, we are on a quite old kernel in the lab, where this code is not present, so I'm not able to proof my assumption. Nevertheless, my former comment is still valid, as the averaged rate could be calculated over several kinds of frames (management/multicast) which have a inherently lower rate. This could have the inverse effect to slowly change the path metric, while this patch should stabilize it, isn't it ? -- M.Sc. Benjamin Beichler Universität Rostock, Fakultät für Informatik und Elektrotechnik Institut für Angewandte Mikroelektronik und Datentechnik University of Rostock, Department of CS and EE Institute of Applied Microelectronics and CE Richard-Wagner-Straße 31 18119 Rostock Deutschland/Germany phone: +49 (0) 381 498 - 7278 email: Benjamin.Beichler@xxxxxxxxxxxxxx www: http://www.imd.uni-rostock.de/
