[PlanetCCRMA] Re: 2.6.7 ccrma kernel and usb / nvidia support

[Fernando Lopez-Lezcano]

On Fri, 2004-11-19 at 12:36, Rick B wrote:
> Shayne O'Connor wrote:
> >hi fernando
> >
> >i haven't been able to notice whether it fails - these are basically the
> >only messages i see when booting ...
> >
> >>Nov 18 17:17:36 localhost syslog: syslogd startup succeeded
> >>Nov 18 17:17:36 localhost syslog: klogd startup succeeded
> >>Nov 18 17:17:36 localhost irqbalance: irqbalance startup succeeded
> >>Nov 18 17:17:36 localhost portmap: portmap startup succeeded
> >>Nov 18 17:17:36 localhost nfslock: rpc.statd startup succeeded
> >>Nov 18 17:17:36 localhost ifup:  done. 
> >>Nov 18 17:17:36 localhost network: Bringing up interface eth0:  succeeded 
> >>Nov 18 17:17:41 localhost rpcidmapd: rpc.idmapd startup succeeded
> >>Nov 18 17:17:41 localhost random: Initializing random number generator:  succeeded
> >>Nov 18 17:17:41 localhost rc: Starting pcmcia:  succeeded
> >>Nov 18 17:17:41 localhost netfs: Mounting other filesystems:  succeeded
> >>Nov 18 17:17:41 localhost autofs: automount startup succeeded
> >>Nov 18 17:17:41 localhost smartd: smartd startup succeeded
> >>Nov 18 17:17:41 localhost acpid: acpid startup succeeded
> >>Nov 18 17:17:44 localhost cups: cupsd startup succeeded
> >>Nov 18 17:17:44 localhost sshd:  succeeded
> >>Nov 18 17:17:44 localhost xinetd: xinetd startup succeeded
> >>Nov 18 17:17:44 localhost sendmail: sendmail startup succeeded
> >>Nov 18 17:17:44 localhost sendmail: sm-client startup succeeded
> >>Nov 18 17:17:45 localhost gpm: gpm startup succeeded
> >>Nov 18 17:17:45 localhost crond: crond startup succeeded
> >>Nov 18 17:17:45 localhost xfs: xfs startup succeeded
> >>Nov 18 17:17:45 localhost anacron: anacron startup succeeded
> >>Nov 18 17:17:45 localhost atd: atd startup succeeded
> >>Nov 18 17:17:46 localhost readahead: Starting background readahead: 
> >>Nov 18 17:17:46 localhost rc: Starting readahead:  succeeded
> >>Nov 18 17:17:46 localhost messagebus: messagebus startup succeeded
> >>Nov 18 17:17:46 localhost wine: Registering binary handler for Windows applications
> >>Nov 18 17:17:46 localhost rc: Starting wine:  succeeded
> >>Nov 18 17:17:46 localhost mdmonitor: mdadm succeeded
> >>Nov 18 17:17:46 localhost mdmpd: mdmpd failed
> >
> >(what is that mdmpd module?)
> >
> >this is what's in my /proc/bus/usb folder:
> >
> >>[mrmachine@localhost usb]$ ls -l
> >>total 0
> >>dr-xr-xr-x  2 root root 0 Nov 18 21:02 001
> >>dr-xr-xr-x  2 root root 0 Nov 18 21:02 002
> >>dr-xr-xr-x  2 root root 0 Nov 18 21:02 003
> >>-r--r--r--  1 root root 0 Nov 18 21:02 devices
> >
> >if i run 
> >
> >>mount -t usbdevfs usbdevfs /proc/bus/usb
> >
> >from a command line, here's what i get:
> 
> >>[root@localhost usb]# mount -t usbdevfs usbdevfs /proc/bus/usb
> >>mount: usbdevfs already mounted or /proc/bus/usb busy
> >>mount: according to mtab, usbdevfs is already mounted on /proc/bus/usb

Ok, so that's not a problem in your case. Your /proc/bus/usb is mounted
and has stuff in it (and of course you will fail to mount it again, it
is already mounted). 

> >do you think it might help to add usbdevfs to fstab?

Nope, I think it is working fine in your case. 
One thing you could try is to see what is the software calling in terms
of system calls:

strace program_name

This will print LOTS of stuff (normally) and what the program was doing
before the crash may yield more information...

>      I was waiting for Fernando to give his expertise on this but since 
> he's no doubt very busy

Good guess!

> I'll go ahead and add my 2 cents worth. On 
> /proc/bus/usb/devices you only have read permission and this *might* be 
> a problem, but I don't have any expierience troubleshooting this kind of 
> problem, so I can't say for sure. Have you tried doing this as root? If 
> it works as root then it is a permissions problem. By the way, you have 
> a lot of stuff starting at boot that you really don't need 
> (sendmail,portmap,netfs,etc...).

strace could help finding out what is breaking the program. 
Not a lot of expertise in this...

-- Fernando