On 10/04/14 19:14, Johannes Berg wrote: > On Thu, 2014-04-10 at 17:53 +0200, Antonio Quartulli wrote: >> >> On 08/04/14 12:04, Johannes Berg wrote: >>> On Sun, 2014-03-30 at 22:35 +0200, Antonio Quartulli wrote: >>> >>>> + u32 (*get_expected_throughput)(void *priv, void *priv_sta, >>>> + struct ieee80211_supported_band *sband); >>> >>> why would that need an sband argument? >>> >> >> I needed this because I saw it is required by some RC algo (only >> minstrel) to extract the bitrate starting from the index. >> >> You can check patch 4/6 for this: >> >> + bitrate = sband->bitrates[mi->r[idx].rix].bitrate; >> >> maybe I can extract it from another structure? I couldn't find any way.. > > Bit strange that the algorithm doesn't have the band, if it has an > index? but I guess I didn't check all the patches... I just rechecked, and it looked to me that all the other functions in Minstreal (non-HT) take the sband as argument. It is not stored anywhere and therefore it has to be passed as argument to this new API as well.. A general beautification might be applied to rate_control_ops (and not only)...but it is out of the scope of this patchset :) Cheers, -- Antonio Quartulli
Attachment:
signature.asc
Description: OpenPGP digital signature
