On Wed, 2020-03-25 at 11:14 -0700, Brian Norris wrote: > On Wed, Mar 25, 2020 at 8:52 AM Ben Greear <greearb@xxxxxxxxxxxxxxx> wrote: > > On 03/24/2020 10:16 PM, Brian Norris wrote: > > > Sure, but if you mask out all but 1 bitrate...voila! A fixed rate! > > > > So, see this thread from a while back. Has anyone even *tried* to use > > this API you are proposing? > > Yes, in fact, I have! Which is why I noted: > > > > Now, there are other problems, like the others that Ben mentioned: the > > > rest of the mac80211 framework doesn't like it too much if you really > > > disable all but 1 rate (arguably a mac80211 bug -- but not a nl80211 > > > bug) > > http://lists.infradead.org/pipermail/ath10k/2017-October/010291.html > > I hadn't seen that thread. So it sounds like maybe Johannes isn't > quite on the same page as Johannes ;) Hah. Happens all the time, not sure what that other Johannes is thinking ;-) More seriously though, I'm a bit lost now (and a big part of that is probably because I'm replying to a 2 months old thread, sorry). > If we're going to be particular about matching the AP's basic rates, > then this API is indeed probably not useful for the "single fixed rate > [for debugging/testing]" use case. > > > mac80211: Revert some of e8e4f5, fixes setting single rate in ath10k. > > Commit e8e4f5 was an unfortunate consequence of the stuff I mentioned > earlier about how Chrome OS used to use SET_TX_BITRATE_MAX -- we > weren't nuanced about it at all, so we might configure a set of > bitrates that doesn't intersect at all with the AP's BasicRates. That > does make it hard for the driver/framework to decide what to do: do we > listen to the user, or to the AP? Incidentally, that's also one reason > why Chrome OS no longer uses the API; it was too big of a hammer for > what we want (initial-connection reliability), and required us to be > more delicate about {Supported,Basic}Rates than we really wanted to. Right. But I'm not sure why Ben needs to just do a pr_err()? Shouldn't that just not happen? Hmm. johannes