On Thu, 17 Nov 2011, Geoff Levand wrote: > Hi, > > On Tue, 2011-11-15 at 12:12 -0500, Alan Stern wrote: > > The HC gets reset during initialization, and you can reload the > > registers then. It also gets reset when the driver unbinds or the HC > > dies; it doesn't matter what happens to the registers at those times. > > > > The only other place where ehci_reset() gets called is at the start of > > ehci_run() -- not where your 4/5 patch added a hook. As far as I can > > tell, that call isn't necessary at all. We just have to audit the > > various platform drivers to make sure they all call ehci_reset() at > > some appropriate point during their probe routine. > > > > Therefore the 4/5 patch should merely remove the ehci_reset() call near > > the start of ehci_run(), and this patch should merge the > > ps3_ehci_post_reset code at the end of ps3_ehci_hc_reset() instead of > > installing a hook. > > I've made the two preliminary (untested on non-ps3) patches following > that I think do what you suggest. > > As I'm not familiar with it, I'm not sure what to about devices with TDI > (Transaction Translator) support. Will they just work like this, or do > we need to add a conditional on the ehci_reset()? I think that part of your patch was correct; just remove that section of code from ehci-hcd.c entirely. In fact, the comment there will now apply to non-TDI controllers as well -- they will all do the ehci_reset either in their probe or their reset routine. Therefore ehci_run doesn't need to call ehci_reset at all. As for whether the return value from ehci_reset needs to be checked... Practically nobody seems to do so now. I guess we don't need to worry about it. 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
