On Sat, 17 Jan 2009, Benjamin Herrenschmidt wrote: > > Ben, we need to check with you on this also. These patches we have > > been throwing around move the code to disable and enable the PPC_PMAC > > ASIC clocks for USB. Can that code execute safely at any time while > > the host controller isn't running? Is it okay to execute the code > > with interrupts disabled? > > Wow, that's digging old stuff I haven't touched for some time :-) > > So I think running the pmac hooks with IRQs off might be ok. They have > delays but they use mdelay() not msleep(). > > As to when it is safe to run that code, it isn't a trivial answer. I'd > say any time after it's been suspended and before it's resumed sounds > about right, but my experience with those earlier Apple built-in OHCIs > is that they are -very- sensitive to random environmental conditions > when suspending/resuming and if -anything- goes wrong, they start > happily scribbling all over system memory when suspended :-( > > So I think it's going to boil down to testing. Might be useful to find > somebody with one of the really old powerbooks, I see if I can find that > around the lab here, mine is dead. > > Do you have a git tree I can pull to test your stuff ? Not exactly. Just try this patch: http://marc.info/?l=linux-kernel&m=123184641714487&w=2 with vanilla 2.6.29-rc1 or rc2. Alan Stern -- 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
