On Mon, May 13, 2002, David Brownell <[EMAIL PROTECTED]> wrote: > Johannes, in one note you said > > > I don't think anyone was arguing how the proper way to their write their > > host controller driver is, but rather what the proper coding convention > > is in the kernel. > > but then you later contradicted yourself by responding to Greg:
Huh? > > > So can the > > > allocate() and deallocate() calls for usb-ohci.c and hcd.c be removed > > > also? If so, can't they be dropped from the core all together? > > > > I don't think they can be completely dropped. > > Absolutely not. > > > It appears that OHCI has some device specific data that the HCD needs to > > track. I think it's relatively simple, but it would require the > > allocate() and deallocate() calls to remain. > > > > However, usb-ohci uses deallocate() to free the aforementioned data (and > > some minor related cleanup it appears) in addition to removing any URB's > > still around. > > > > It's that second part that's not needed. > > Doesn't seem correct, but then I'm focussing on "ohci-hcd" instead. I'm looking at usb-ohci.c, which Greg originally referred to. > And of course, EHCI uses those to ensure that every QH associated > with that device is deallocated along with the device. It looks similar to what's in usb-ohci.c. I'm not familiar enough with EHCI to know what is different and why it's required. Some of it looks superfluous. Like QH's being unlinked. That should be guaranteed to be unlinked before ehci_free_config is called. JE _______________________________________________________________ Have big pipes? SourceForge.net is looking for download mirrors. We supply the hardware. You get the recognition. Email Us: [EMAIL PROTECTED] _______________________________________________ [EMAIL PROTECTED] To unsubscribe, use the last form field at: https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
