On Sun, 8 Nov 1998, Mike Galbraith wrote:
> On Sun, 8 Nov 1998, John Summerfield wrote:
>
> > How do I discover what this means?
> > Nov 8 05:01:40 emu kernel: Unable to load interpreter
> > Nov 8 05:02:08 emu last message repeated 247 times
> > Nov 8 06:00:56 emu kernel: Unable to load interpreter
> > Nov 8 06:00:56 emu kernel: Unable to load interpreter
> >
> > It's coming from linux/fs/binfmt_elf.c.
> >
> > I am running Linux emu.os2.ami.com.au 2.0.35 #16 Mon Aug 10 10:26:24 WST 1998 i586 unknown
> >
> >
> > There have been no more messages since the last above - it's about 12 hours
> > later and the system seems normal except for this:
> > Nov 8 16:52:34 emu kernel: Sound error: Couldn't allocate DMA buffer
>
> Do you have bridging enabled? If so, there's a memory leak there which
> would account for this behavior. It's fixed in pre36.
No bridging
>
> It means exactly what it says and I don't know what your problem with the
> message is! The kernel is unable to load an interpreter. In this case
> it appears to be either shell (/bin/sh) or the elf loader itself!
I see you don't know either.
>
> Are you really trying to say "how do I discover what is causing this
> message to appear?". You're out of memory or another resource, such as
> process table entries. You discover that by looking.
>
Sure? Where?
> > There have been no more messages since the last above - it's about 12 hours
> > later and the system seems normal except for this:
>
> Seems normal to you, yes, but then you said that you couldn't understand the
> error message, so that implies to you that we should not rely on your
> powers of interpretation. Instead we need the raw data.
I gave you all the data I have. There is absolutely no call to be rude. I
am a user, not a kernel hacker.
>
> > Nov 8 16:52:34 emu kernel: Sound error: Couldn't allocate DMA buffer
>
> Not surprising. Obviously out of memory, Why don't you type "free"?
At the time it happened, I was in bed pushing out the zeds. When I got
those "Unable to load interpreter" messages it might have been running a
fairly big Perl job and if Perl's particularly wasteful of RAM, 96 Mb might
not have been enough. At 16:52 it was lightly loaded and the lack of sound
caused me to go looking when I found the other problem.
Had I seen the message when it happened, I'd not have thought to type
"free" - there's nothing in the message to suggest to me it's short on
memory. It looks more likely to me that the message is synonymous with
'internal processing error" or "the kernel is badly broken."
>
Paul Wouters and Brandon S. Allbery managed to tell me the probably cause
without being patronising or rude: thanks chaps. I'll be adding swapper. It
happened again this morning though I only got one message.
I will make this comment about the message: it's singularly unhelpful. It
doesn't tell anyone what the kernel was trying to do. It gives no clue as
to what program it was trying to run. The only interpreter I could think of
it might want to load is the Java bytecode interpreter: however this seemed
unlikely because I have no production java stuff that I run unattended and
I don't use the kernel built-in java support anyway (though it IS compiled
in).
Finally, it should NOT be necessary for users to grep the Linux source tree
to find the source of an error message or its meaning. Not that the source
code was helpful in discovering the meaning once I got there.
Cheers
John Summerfield
http://os2.ami.com.au/os2/ for OS/2 support.
Configuration, networking, combined IBM ftpsites index.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/