"WARNING: cannot set battery backed clock...". Should I be concerned?Not really, but it means that you are using a rather old kernel. Newer kernels give the error message below .
"NetBSD/mac68k does not trust itself to update the RTC on shutdown." Should I be concerned?No. This is simply an indication that NetBSD/mac68k would rather not set your Real Time Clock since NetBSD loses more time the longer it runs (this would lead to your clock being off in Mac OS as well). The code to update the RTC on most machines does exist, though, and you can turn it on in the kernel if you truly wish to do so.
init: not found
panic: no init
Pausing forever.
/sbin/init was not installed. Use the Installer to "ls /sbin/init"
and make sure it is there. Make sure that you have installed a base
distribution set such as the NetBSD 1.3 release, or a -current snapshot:
audio at mainbus0 not configured
floppy at mainbus0 not configured
Audio support has been available since the prior to the 1.2 release. You should now see a line like:
asc0 at obio0: Apple Sound Chip
However, floppy support has not been implemented yet. It is in the works, though.
If you were at the bottom of the screen when this occurred, then this is the infamous "scrolling while ADB input is being received" problem. There are a few ways to avoid it:
changing root device to sd1a.
Swapping 409 and 401.
Swapdev = 409, dumpdev=ffffffff.
You need to edit the /etc/fstab file to mount the new root
partition. You can also simply use the Installer utility to delete the
current /etc/fstab and then do a "Build Devices" to create a new
one which the correct SCSI ID's for your drives.
Your symptoms probably look like:
If I boot with extra-debugger info on, using multi-user, I get:Changing root device to sd1a. PRAM: 0x30d482dc, macos_boottime: 0x30d482d1. vm_fault(10c000, 8bb3000, 1, 0) -> 1vm_fault(10c000, 8bb3000, 1, 0) -> 1vm_fault(10c000, 8bb3000, 1, 0) ->then it hangs.
Unfortunately, the 1.1 distribution kernel has quite a few bugs in it and does not run very well on many machines, especially the IIci. Try the 1.5 distribution
netstat -r I get: netstat: kvm_read kvm_read: Bad address. What's wrong?This is nothing to worry about. Basically, the file /netbsd is
not your current kernel. A number of programs (such as ps, who,
systat etc.) and libkvm, access /netbsd to learn what's going
on in the kernel. So you can simply rename your current kernel to
/netbsd to make this kind of error go away. Make sure that you
are not overwriting a working kernel when you do this, though, unless that
is exactly what you intend to do.
netstat -r I get an endless stream of question marks. What's wrong?That's a mis-match between libkvm/netstat and /netbsd. Or it's that
you're running a kernel that's not named /netbsd. Other symptoms of this
problem are likely to be that who, ps, ifconfig, and
systat will not work either. If you update your binaries and your
kernel at the same time, then you should be OK.
w, ps, and netstat, among others, don't work. What happened?One of two things. Either your currently booted kernel isn't named
/netbsd or else you have a mismatch between your kernel and the
binaries you are using. In the first case, simply making a link from your
currently booted kernel to /netbsd will solve the problem.
In the second case, dynamically linked binaries can usually be fixed by upgrading libkvm to match your new kernel. Statically linked binaries need to be replaced with more recent versions. Since they are statically linked, if you are going to recompile them yourself, you need to rebuild libkvm.a before you rebuild the program in question.
Thanks to John Wittkowski (jpw@netscape.com), here is a list of most of the
programs (besides /bin/ps) that depend on libkvm (all of these are
in /usr/bin):
fstatgdbipcsnetstatnfsstatsystatuptimevmstatwLike the previous three questions on this subject, the answer is most likely that your libkvm is out of sync with your kernel or binaries. To solve this problem, you can either get a binary distribution which matches your kernel, or you can build you own by following the instructions below:
If you get the "proc size mismatch" error and you determine
that you need to update your libs, here's what to do:
1. Get all the source code. If you're not willing to do this
and recompile things than you'll have to find someone who
will do it for you and you can try installing everything
by hand.
2. Make sure that your include files are up to date. Do
this by:
cd /usr/src
make includes
This will take a while. I had some trouble with this
because some of the Makefiles didn't have the INSTALL
variable defined. Whenever the "make includes" failed,
I went to the last directory listed and added the
following line to the Makefile:
INSTALL=/usr/bin/install
I had to do this several times before the make finished
without any errors.
(If you make sure that /usr/bin/make and all the files in
/usr/share/mk are up to date first, the above difficulties
can probably be avoided --Colin)
3. Rebuild the libkvm and install it:
cd /usr/src/lib/libkvm
make
make install
Note that in order to get the libkvm to compile on
my system I had to add the following link:
cd /usr/include/machine
ln -s ../m68k/kcore.h kcore.h
This may have been a quirk of my system so try
compiling without it first.
4. Then rebuild the binaries that are STATICALLY linked to
libkvm. The only statically linked program that I am
aware of is "/bin/ps". To rebuild ps, simply:
cd /usr/src/bin/ps
make
make install
5. You may or may not need to rebuild the binaries that
are dynamically linked to libkvm. This is because (I
think) if the major version number of the lib changes
then the old binaries will expect the old version
number and not work with the newer version of the lib.
For example, my old libkvm was libkvm.so.4.0. The new
one was libkvm.so.5.0. Without recompiling the
dynamically linked binaries, it would still complain
about "proc size mismatch" (if the 4.0 lib was still
there) or some lib missing error (if the 4.0 lib
had been removed from /usr/lib). If the version minor
number changes (4.0 to 4.1, for example), I think it
will run with a warning and so you may not need to
recompile all of these things.
The dynamically linked binaries that I am aware of
will give the "proc size mismatch" error (if the 4.0
lib was still there) or some lib missing error (if the
4.0 lib had been removed from /usr/lib). If the version
minor number changes (4.0 to 4.1, for example), I think
it will run with a warning, and you may not need to
recompile all of these things.
The dynamically linked binaries that I am aware of
are:
/usr/bin/fstat
/usr/bin/gdb
/usr/bin/ipcs
/usr/bin/netstat
/usr/bin/nfsstat
/usr/bin/systat
/usr/bin/uptime (linked to /usr/bin/w)
/usr/bin/vmstat
/usr/bin/w
Note that /usr/bin/uptime is a link to /usr/bin/w and
will be set up properly when you do the "make install"
for w.
To recompile these, do the following:
cd /usr/src/usr.bin/<cmd>
make
make install
For example, to recompile /usr/bin/vmstat:
cd /usr/src/usr.bin/vmstat
make
make install
Much thanks to John Wittkowski (jpw@netscape.com) for providing such a detailed answer for this one.
This error was due to a bug in the serial drivers. A fix has been made, but it did not make it into the 1.1 distribution kernel. Kernels since that time should have the fix incorporated.
(NOTE: If you are currently seeing this error, please read the paragraph at the bottom of this answer)
Here is the latest from Bill Studenmund (wrstuden@loki.stanford.edu):
I'm pleased to announce that we seem to have fixed the serial port
problems. The current source on ftp.NetBSD.org should be correct.
The problem turns out that we misunderstood how to set up the interrupt
levels (spltty specifically). We were turning off interrupts to the
chip when we should only have been turning off the passing of data to
the kernel. We were supposed to stop characters coming out of a
temporary buffer (the ring buffer) when in fact we were stopping characters
entering it. Thus it never did much, and we got lots of errors. :-(
I've gotten ppp to work very solidly. I've transferred a couple of meg
over the line w/o a single fifo error. :-)
The changed file is sys/arch/mac68k/include/param.h. Note that changing
it will mean a lot of the kernel needs re-compiling.
Noud de Brouwer, who is running a neighborhood ISP has used it, and gotten
some very impressive numbers. Like over 6k bytes/sec. :-)
WARNING about high speeds: If, while transferring data at over say 28800
(non-scientifically-picked number), you want to do something else, like
compile, log in elsewhere, or use ethernet, you might have problems.
You might not. Every time a byte gets received, the CPU has to drop
what it's doing, save what it's doing, get the byte, store it, and
resume what you're doing. It's life w/o DMA. :-( It works, but takes
a bit of CPU. Just be aware.
Recently, a change in the interrupt handling has caused this error to start cropping up again. Although a fix has been committed, newer serial drivers should permanently fix the problem. When the new serial drivers are committed, a note will be made of it.
Bootstrapping the pmap system.
Failure in BSD boot. nextpa=0x106000, high[0]=0x100000.
panic: You're hosed!
You forgot to put your Mac into 32-bit mode in the Memory control panel. Do this, and, barring other problems, it should boot fine.
The above is true unless you have a 24-bit video board installed. For some reason, the 1.2 kernels seem to have trouble detecting that the machine is in 32-bit mode when one of these boards (e.g. an Apple 8.24 video card or a Radius Preciesion Color Pro 24XP board) is installed.
This problem was not present in the 1.1 kernels, so hopefully a getting a more recent kernel will fix the problem (assuming the problem has been fixed by now), unless you are willing to simply remove the board for now.
Mar 2 13:03:04 myname init: kernel security level changed from 0 to 1
mrg: no trace functionality enabled
panic: kernel jump to zero
This is normal if you're using the serial console boot and have not fixed
your /etc/ttys to disable the getty on /dev/ttye0 and
enable the getty on /dev/tty00 like so:
Change /etc/ttys from this:
# Define console that we actually run getty on
ttye0 "/usr/libexec/getty Pc" vt220 on secure
.
.
.
#Hardwired lines are marked off ...
tty00 "/usr/libexec/getty std.9600" unknown off secure
to this:
# Define console that we actually run getty on
ttye0 "/usr/libexec/getty Pc" vt220 off secure
.
.
.
#Hardwired lines are marked off ...
tty00 "/usr/libexec/getty std.9600" unknown on secure
Thanks to Brian Wimberly (brianw@scripps.edu) for the above fix.
/netbsd: Panic switch: PC is 0x7f734.
/netbsd: ae0: warning - receiver ring buffer overrun
From Steve Weiss (srw@izzy.net):
This bug is known and fixed.
A little over a month ago I found a bug in dev/if_ae.c, the apple ethernet card driver, in which it failed to lower the interrupt that the card raises when its tally count registers overflow. This results in a perpetual interrupt hang. The mean time to fail is a function of network traffic.
Allen fixed this and checked the edits back into the source tree on February 2. There are several kernels in his outgoing directory on puma that have been built with this problem corrected. The earliest being netbsd.please.test (4feb), and the latest being netbsd.GENERIC-5 (7mar). Check the README in
ftp://ftp.macbsd.com/pub/outgoing/briggs/
My IIci used to hang every 3-5 minutes before this fix (we have very high traffic) and now it has been up for about 2 weeks running GENERIC-3.
/netbsd: ae0: warning - receiver ring buffer overrun
/netbsd: ae0: device timeout, recovered
If these messages continue after boot, you probably have a problem. However, if you only seem the during boot, they are most likely caused by the card having already been initialized by the Mac OS before you booted NetBSD.
Thanks to Henry B. Hotz (h.b.hotz@jpl.nasa.gov) for answering this one.
From Scott Reynolds (scottr@og.org):
This is a pretty well-known problem, and stems from the fact that Sun's telnet (and HP's, at least up to HP-UX 9.x) is based on very old, and broken, code. There are two workarounds for this: (a) get new telnet source from ftp.cray.com, compile, and install it on the Sun, or (b) add "-k" to the entry in inetd.conf for telnetd. The former solution is preferred; the latter is a kludge that will help the situation, but you will probably find other strangeness coming from "good" telnets.
Presumably, the Mac version of telnet suffers from a similar problem. Using rlogin instead usually works for me, but I'd recommend using ssh if it's available.
Unimplemented Trap system error?From Bill Studenmund (wrstuden@loki.stanford.edu):
I bet you have a "minimal" installed Mac OS, no? There is a call that was part of an experiment for supporting booting w/ VM on. The project wasn't completed, and this call was left in. It is not implimented on minimal systems, but is on full installs.
Upgrading to Booter 1.9.4 or later should fix the problem.
savecore: can't find device 0/0 when I go multi-user?From Allen Briggs (briggs@puma.macbsd.com):
That's because the file/netbsdand the current kernel are not the same, andsavecoreneeds to get the default dump device from the kernel, but if/netbsdand the current running kernel don't match, it gets confused. So do a few other programs.
If it hangs right after recognizing either the video card or the ethernet card, then you have probably been bitten by the interrupt conflict bug. For some time now, it seems that certain combinations of video and ethernet cards seem to have an interrupt conflict, usually causing the machine to hang on boot.
Although removing one of the cards will solve the problem, you might want to try a newer kernel as this kind of bug has been solved in several instances.
Allen Briggs (briggs@puma.macbsd.com) has provided a HOWTO on locating the information necessary to help fix this kind of problem: http://www.macbsd.com/macbsd/howto/video.html
ae0: NIC memory corrupt - invalid packet length 65280 when I boot?The full text of the error message is probably something more like:
ae0: length does not match next packet pointer
ae0: len 0000 nlen ff00 start 0c first 00 curr 20 next 00 stop 40
ae0: NIC memory corrupt - invalid packet length 65280
This is most likely because you had Mac OS networking enabled at the time you booted (most likely EtherTalk or MacTCP). Try booting with extensions off (well, at least those extensions) and it should go away. Occasionally, the boot process will hang at this point if you don't turn them off.
Thanks to Isaac Salpeter (isaac@ticalc.org) for the above answer.
According to Allen Briggs (briggs@puma.macbsd.com), these messages are fairly harmless, and you can eliminate them by compiling a kernel without the DIAGNOSTIC option enabled in its configuration file.
/dev/sd0a on /: specified device does not match mounted device?This indicates that your /etc/fstab doesn't match the actual
layout of the partitions on your disk. A likely cause it that versions 1.1a
through 1.1c of the Installer utility create a possibly incorrect version
of /etc/fstab putting all of your partitions on sd0. If you're
not using sd0, there are a couple of solutions:
/etc/fstab, edit it to reflect
reality in a text editor, and then use the Installer to copy it back in.Another possibility is that NetBSD 1.3 and later ignore Apple Driver type
partitions when filling in the disklabel, so some of your partitions might
move around due to this. Doing a disklabel sdX where "X" is the
device number of the drive in question should give you a listing of the
current layout of your partitions so you can fix your /etc/fstab.
panic: ResHndls table too small!
From Bob Nestor (rnestor@metronet.com):
The error comes from running out of ResHandles while making calls for ROM Resources. This is part of the MRG code solution for the ADB, so it probably won't be a problem at all with the Hardware Direct ADB solution. I'm not sure why the ResHandles Table gets exhausted with certain ADB devices, but it does. I recall working with someone on this some time ago, and we increased the table size to a value larger than the total number of Resources in the ROM and still couldn't eliminate the error. It's probably a case where there's an error in the use of the ResHandles Table *and* a loop of some kind in the ADB ROM routines trying to handle these devices. There are some ROM patches for the ADB routines in Mac OS that we're not currently using in the MRG code in NetBSD, and they may be related this this problem.
So, unless someone wants to dig back into the MRG code to solve this, real solutions are to either wait for the Hardware Direct code or run without the offending ADB device.
Warning: Battery clock earlier than filesystem date. What's wrong?From Nico van Eikema Hommes (hommes@ccc.uni-erlangen.de):
I think I have figured out what happens: it is a fug (that is a bug that would, in a Microsoft program, be called a "feature" :-> ) in the Installer. The timestamp is set to the local Mac OS time (a cpin followed by ls shows this), but it should be set to GMT. When the filesystem is mounted during the boot process, the timestamp is interpreted as GMT, what it should be, and the offset, which for MET is positive, is added to it. That of course gives a later time than that of the battery clock. The irregularity of course came from the time elapsed between installing and booting. The negative offset for US-based systems probably caused this to remain unnoticed.
The latest version of the installer (i.e. 1.1e) fixes this problem by using GMT for reads and writes instead of the local timezone (which is what the Mac OS uses).
"kernel is not in a format the booter can understand". What's wrong?You are probably trying to boot from a Mac OS partition and have forgotten to extract the raw kernel file from archive that it is in (many kernel files are uploaded as tar archives, gzipped archives, or both). You can get a copy of the enhanced Stuffit Expander or the MacGzip and Tar utilities to extract the raw kernel file. Make sure that you do not do any sort of end-of-line (i.e. text) translation in the process, or you will corrupt the kernel binary.
"sn0: receive descriptors exhausted"?From Denny Gentry (denny1@home.com):
The sonic DMAs a received packet into a data buffer. It then needs a descriptor to write the address of the buffer and some information about the packet (like its length) to hand it to the driver. If it runs out, it interrupts.
Usually this means the sn driver didn't get to run for a long time to clean up its descriptors. This can happen if there is a lot of SCSI activity, since SCSI interrupts will block network interrupts. Early in the morning the /etc/daily (etc) scripts run, causing lots of disk activity. If someone tried to reach your machine while the disk grinding was going on, it could easily cause it to run out of descriptors.
A few weeks ago I submitted changes to increase the number of receive descriptors by using a separate 4K page just for RX descs. Scott checked those changes in, so make sure you have a "recent" kernel. I note from another message you sent that you're running 1.2D, a 1.2G kernel will have more descriptors.
Since the Sonic requires its descriptors to be physically contiguous, giving it even more descriptors would require having a way to allocate physically contiguous memory.
"warning: no /dev/console" What's wrong?This and the similar message:
warning: lookup /dev/console: error #
Thanks to Bill Studenmund (wrstuden@loki.stanford.edu) for the above information.
You booted with booter version 1.8.
Booter version 1.11 is necessary to fully support this kernel.
From Scott Reynolds (scottr@NetBSD.org):
Harmless message, unless you are using a miniroot. (Well, if you're using a miniroot, the manner in which you are currently doing it is deprecated, anyway.) You can safely use new kernels with booter versions 1.9.5 or later, regardless of the warning message.
The booter version passed in has been out of date for quite some time. That's one of the reasons the next version of the Booter is 1.11, not 1.10.3; I wanted to circumvent any potential trouble with a Booter that actually had the version number fixed.
This is a problem that seems to most often plague IIcx users, but it has been seen in other older II-series Mac's as well. It appears to be a result of some strange low-memory interaction that depends on the size of the kernel and the location of certain data structures within it.
One workaround for this problem is to compile a custom kernel with unnecessary device drivers removed. This will often fix the problem (but not always). Another possible workaround is to try to reboot the machine and shutdown from Mac OS instead. If the reboot hangs as well (as is often the case), the best you can do is shutdown to single-user mode, sync the disks, and then manually power off or reboot the machine.
Table of contents of this chapter, General table of contents