在 2021-08-19星期四的 00:13 +0800，Kyle Tso写道： > On Wed, Aug 18, 2021 at 10:40 PM Greg Kroah-Hartman > <gregkh@xxxxxxxxxxxxxxxxxxx> wrote: > > > > On Wed, Aug 18, 2021 at 10:02:24PM +0800, Icenowy Zheng wrote: > > > > > > > > > 于 2021年8月18日 GMT+08:00 下午9:56:38, Greg Kroah-Hartman < > > > gregkh@xxxxxxxxxxxxxxxxxxx> 写到: > > > > On Tue, Aug 17, 2021 at 02:00:33PM +0300, Heikki Krogerus > > > > wrote: > > > > > > > Why is it necessary to do discovery with data role swap > > > > > > > in general? > > > > > > > > > > > > I think it could be possible for devices to expose > > > > > > different altmode > > > > > > with different role. > > > > > > > > > > OK. FWIW: > > > > > > > > > > Acked-by: Heikki Krogerus <heikki.krogerus@xxxxxxxxxxxxxxx> > > > > > > > > Is this conflicting with > > > > https://lore.kernel.org/r/20210816075449.2236547-1-kyletso@xxxxxxxxxx > > > > ? > > > > > > > > Which of these two should I take, both? Neither? > > > > > > Don't take this, it's against spec. > > > > Ok, now dropped. What about the linked patch above? Does that > > work for > > you instead? > > > > thanks, > > > > greg k-h > > Hi Icenowy, > > Could you revisit the patch for the tcpm_unregister_altmodes part? > I think that is still a problem. Well I think it's okay, but I know none about PD3, and I have no PD3 device either. So I prefer to continue on fix on dropping the discover sent on snk_attach instead of working on this, because I know nothing about this, and in my limited knowledge the code looks fine here. > > thanks, > Kyle
