I am openint this as new since the base problem (15518) was not beta-only
(it was from the open beta).
This is also related to closed problems bug 15464 and bug 13605.
1. USB support in intiscripts is still broken.
2. I am still getting the error messages reported in bug 13605.
3. Using NOUSB appended to the kernel parameters still has no effect (bug
4. It is now worse -- if I append NOUSB, I get a "FAILED" durring the boot
process for the usb startup.
This defect is considered MUST-FIX for Winston Gold-release
Did you do an upgrade from previous betas? kudzu should not add new controllers
without IRQs, but won't remove them either.
Nope ... fresh install. This is highly repeatable.
In what way is it broken (how does it fail)?
nousb works fine here; did you actually specify
it in all caps?
(allcaps 'nousb' won't work in the current package, but has
been fixed in CVS.)
Specified in all lower case.
I have no idea how it failed. The "message" during bootup is "FAILED" after
saying Starting USB and the silly insmod error messages.
What initscripts do you have installed?
I have initscripts-5.44-1 from rc1.
Bill -- important point: I have some of the usb stuff disabled in the BIOS so
that I could "free up" an IRQ. This may be "causing" the error messages. On
other systems with USB such as the one I am on now, they also have USB but no
error messages and no problems.
BTW, if you have a newer version of initscripts you want me to test, give a
How are you doing 'nousb'?
Also, if you remove the usb-controller alias from
/etc/modules.conf, and then reboot (without 'nousb')
does it proceed normally?
Removing "alias usc-controller ..." from /etc/modules works -- no error
More to the point, I am marking this bug resolved. My brain must not have been
engaged -- I had NOUSB rather than nousb specified in a lilo append. That,
coupled with the fact that I have a "partially" bios-disable usb resulted in the
Thanks for the help and sorry for time lost.
On other systems I have with usb fully enabled, everything seems to work fine.