Le 2002.10.14 18:53, David Brownell a �crit : >> Now it turns out I didn't do such a piss-poor of configuring my 2.5 >> kernel, since the only option I could find that make a difference >> was CONFIG_USB_DEBUG. When I accept flooding my system logs with >> obscure usb incantations the system boots:(. ... >> >> Can an helpful soul help me bring my system some relief ? I'd >> really like not to boot in debug mode. > > That's a new failure mode! Can you help narrow this down?
Of course ! > You're using the OHCI driver, so you can just tweak the lines at the > top of drivers/usb/host/ohci-hcd.c that can #define DEBUG. If you > comment out that #define, and leave CONFIG_USB_DEBUG on (and then > rebuild and re-init with the new OHCI driver), does that work or not? I'll try this ASAP, unfortunately compared to more modern setup it no longuer feels lighting-fast compiling kernels:( > If that works, it'd be time to see which OHCI printk()s morph init (?) > timing enough to matter to your K7 box. Looked to me like they were > all either before or after the timing-critical bits (chip init), so > disabling just the OHCI messages "should" not change your failure > mode. Well, they did and they do. Unfortunately I've found out today not even CONFIG_USB_DEBUG was sufficient, since I had a boot hand (cold boot) with my debug kernel today (warm boot afterwards was ok, though). Still CONFIG_USB_DEBUG is better than no CONFIG_USB_DEBUG, the first one boots almost always, the second hangs (maybe almost always, after 4-5 hangs I stopped testing it). I'll go on experimenting now. Regards, -- Nicolas Mailhot> ------------------------------------------------------- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf _______________________________________________ [EMAIL PROTECTED] To unsubscribe, use the last form field at: https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
