On Tuesday 20 January 2009 00:17:13 Artur Skawina wrote: > Artur Skawina wrote: > > Christian Lamparter wrote: > >> On Monday 19 January 2009 22:53:03 Artur Skawina wrote: > >>> We can move the FREE_AFTER_TX(skb) check from the completion to the submission > >>> path, right? Then find a way to do the pull _before_ giving away the skbs. > >>> I can't shutdown the machine where i can reproduce this today, so it will have > >>> to wait until at least tomorrow. > >>> > >>> artur > >> Like this? > >> > > yes, that's what i was thinking. Will FREE_AFTER_TX(skb) be false for all skbs > > that are dropped in p54_rx_frame_sent()? Not exactly, but it's the right direction. FREE_AFTER_TX is true for _setting_ control frames... e.g frequency/filter/dcf. > > Also most of the push/pull business in completions can probably go, we just have to set > > up the pointers right on submission, ie transfer the rx header into the headroom, then > > no skb manipulations in completions should be needed, and these kind of races should > > be harmless. > > I only looked at the FREE_AFTER_TX(skb) part and completely missed that you had already > removed the pull, sorry. Yeah, this will make life easier for the "stuck package" detection... Although, I'll miss the skb_push & skb_pull checks. But that's a small price, if your device finally works. (Well "works" is maybe a bit too much, as you'll probably run into other bugs, e.g truesize bug or the "shot rc", or more problems with powersaving station that only have apsd). Regards, Chr -- To unsubscribe from this list: send the line "unsubscribe linux-wireless" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html