Re: Ugh in 2.1.60

Linus Torvalds (torvalds@transmeta.com)
30 Oct 1997 23:47:27 GMT

In article <199710302120.WAA08091@jebril.tornado.be>,
Michel Eyckmans (MCE) <mce@tornado.be> wrote:
>
>> Why are you using a SMP-enabled kernel on a uniprocessor system? In my
>> view, that would be the source of the message.
>
>No. I'm on an SMP machine over here, and have been seeing
>occasional ughs in do_page_fault for months now, starting
>(IIRC) somewhere in the late 2.1.3x series (I skipped the
>lower half of the thirties, so the problem may even be
>older than that). I'm now at 2.1.59.
>
>In fact, I even reported this to the list at the time, but
>got no reply. And since it doesn't seem to do any harm, I
>didn't bother after that.

Note that any "ugh" is really a bug, even if it happens on UP. The SMP
code _should_ work fine on UP modulo bugs (and some hardware
incompatibilities: the SMP code may expect a Pentium or better, for
example, and it also expects the BIOS to report correct SMP status
information even if the machine only has one CPU).

Getting a "ugh" in do_page_fault indicates that somebody does a copy
to/from user space with interrupts disabled. That is a _bug_.

I suspect it is some specific subsystem that does this, because I've
never seen it myself. It would be good if people who see this try to
notice what they tend to be doing when it happens, so that we might be
able to find out exactly _what_ it is that tries to take a page fault
with interrupts disabled..

Linus