On Wed, May 05, 2010 at 03:46:10AM +0200, ext Ming Lei wrote:
2010/5/5 Sergei Shtylyov <sshtylyov@xxxxxxxxxx>:
Hello.
tom.leiming@xxxxxxxxx wrote:
From: Ming Lei <tom.leiming@xxxxxxxxx>
The plat->mode passed from platform_data always is MUSB_OTG,
so the init failures below will happen if we configure musb as
HOST or PERIPHERAL mode in Kconfig:
"incompatible Kconfig role setting"
The patch fixes the issue by degrading otg mode to
host or peripheral mode in such case.
Why not fix your platform data instead to pass the needed role like the
DaVinci code does (see arch/arm/mach-davinci/usb.c)?
IMO, this fix is better than the fix in board code:
-do not need to fix all board code which may support musb
-OTG mode means the hardware has the host-only or peripheral-only
capability
generally we want to keep SDPs or EVMs as OTG, so we can test all
possible scenarios, but real products are either peripheral only, otg or
host-only, and that's a real HW characteristic, not something we can
change easily. Take n900 for example, if you build it with otg support
you're doing nothing but bloating the kernel with useless code since
it'll never support the host role unless you route the id pin to the
correct place by hand on the PCB.
still, i think this patch is ok for mainline, that would allow
developers to play with other stuff but I'd like to see a big fat
warning when that happens. Can you update the patch ?
--
balbi
DefectiveByDesign.org
--
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
