Re: still-extant potato boot problems (kmod/runany modprobe loop)
Martin Schulze <joey@finlandia.Infodrom.North.DE> writes:
> This is the proper list, I'm forwarding.
> > Are there any fixes (or even diagnoses) yet for the following problem?
> > Has anyone else even seen these symptoms?
> > Earlier I reported a problem with installing potato (from scratch)
> > and getting errors at boot time like:
> > kmod: runaway modprobe loop assumed and stopped
I've never seen that.
> > and many occurrences of:
> > modprobe: modprobe: Can't open dependencies files /lib/modules/2.2.17/modules.dep (No such file or directory)
Joey seem to indicate he had already made a fix in CVS for this -- is
Is this problem preventing installation or just annoying?
You can probably work around it by running 'depmod -a' on the cmdline.
> > (Note that those errors were from booting the installation boot
> > diskettes (rescue.bin/root.bit). Specifically, it was from
> > before I had a chance to make any choices or affect anything.)
Oh -- so they are just pretty much cosmetically annoying.
I don't know why, in rootdisk.sh, we couldn't run 'depmod -a' chrooted
in the new file system prior to making the drivers floppy or floppies.
I am loath to touch this though since it often breaks.
> > I thought I'd try getting to potato by upgrading (a clone of) my
> > slink system. The upgrade went pretty well (well, I accidentally
> > upgraded the original instead of the clone, but the clone seems
> > to be working fine) until I installed the 2.2.17 kernel and rebooted.
> > Those boot-time errors returned.
Ah, hum. The loop bug you mentioned looks like a kernel bug. Can you
file a bug against kernel-image-2.2.16[-flavor] where flavor is
(optionally) the kernel flavor you are using?
> > I also started getting a bunch of other errors, such as:
> > - "modprobe: Can't locate module xxx"
> > (at bootup; for roughly half of the modules)
Use depmod (see the man page) to check dependancies and locatable
If you are having a problem with the modules themselves, file bugs
against kernel-image. If the modules were simply placed in the wrong
place by the potato boot-floppies, file a bug here.
> > - "start ppp0: SIOCSIFMETRIC: Operation not supported"
> > (asynchonously, after something had triggered diald)
> > - "Use of setserial/setrocket to set SPD_* flags is deprecated"
> > (asynchonously)
> > - something like "xmit when interface down"
> > (asynchonously)
> > - (I can't even ping the other end of my diald's PPP connection
> > (I used the numeric IP address logged by pppd when it
> > connected).)
You are having serial / ppp problems and we cannot help you here.
Are you on i386? If so, try firstname.lastname@example.org. Otherwise,
.....Adam Di Carlo....adam@onShore.com.....<URL:http://www.onShore.com/>