* Andreas Fenkart <afenkart@xxxxxxxxx> [140311 02:45]: > Hi, > > 2014-03-05 17:33 GMT+01:00 Tony Lindgren <tony@xxxxxxxxxxx>: > > * Andreas Fenkart <afenkart@xxxxxxxxx> [140305 00:30]: > >> Hi, > >> > >> 2014-02-27 22:33 GMT+01:00 Tony Lindgren <tony@xxxxxxxxxxx>: > >> > > >> > The wake-irq is needed for omaps with wake-up path and also > >> > when doing GPIO remuxing. So the wake-up handling is pretty > >> > much the same for both cases. > >> I added this comment to the patch, since I was puzzled that you need > >> a wake_irq even whithout remuxing. > > > > Yeah we need it because omap_hsmmc is already doing runtime PM, > > so there's nothing stopping from shutting it down. And there's > > really no need to block runtime PM for it as it's working. > > > Also added this comment, since it really clarifies the issue. OK thanks. > FYI, just tried the patch without pin remuxing on am335x and > it actually works. Read: I'm always using the sdio pinmux never > reconfiguring the pin as a GPIO, but still get GPIO interrupts. > Yes, it fails if I comment out the pm_runtime_resume. > > So it's not that the am335x suddenly has a swakeup line, > but looks like an implementation detail of the am335x GPIO > block. Oh OK, that's good to know. I think 2420 behaved that way too for the serial wake-up. What also may affect things is if that GPIO line is in the first GPIO bank, for the other banks remuxing may still be needed for deeper idle states. Regards, Tony -- To unsubscribe from this list: send the line "unsubscribe linux-mmc" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
