Occasional fail to initiate resume by incoming call, easy workaround proposed

Paul Fertser fercerpav at gmail.com
Mon Feb 9 21:03:45 CET 2009


Andy Green <andy at openmoko.com> writes:
> |> I was going to say then we stink up normal UART suspend function which
> |> doesn't know about GSM, but actually we can just do it in neo1973-pm-gsm
> |> suspend, it's neat and tidy.
> |
> | For some reasoning about why user-space needs to apply/release this
> | flow-control instead of the kernel please consider looking at a) and
> | b) at http://trac.freesmartphone.org/ticket/31#comment:3
>
> I don't think we're talking about setting flowcontrol state in a static
> way, just toggling it.  Because a message can come out of the GSM side
> at any time, there is no illegal violation of Mike's scheme if we
> briefly make sure to toggle CTS before setting it to the requested
> static drive source.

I might be misunderstanding you again, but in the first quote you
mentioned doing something extra in "neo1973-pm-gsm suspend" that i
interpreted like "what that driver hooks to kernel suspend", not what
it does on userspace access of "flowcontrolled" sysfs node. Otherwise,
i agree that it's possible to toggle CTS briefly (set 1->0->1) on
write 1 to "flowcontrolled" node, but it looks a bit excessive to me,
considering it needs to be done only once (and only with a broken
firmware). Otherwise, i see no problem with it.

-- 
Be free, use free (http://www.gnu.org/philosophy/free-sw.html) software!
mailto:fercerpav at gmail.com




More information about the openmoko-kernel mailing list