> At 03:28 PM 12/20/99 +0000, you wrote:
> > > OK, I'll compile a 2.2.13/ikd kernel as soon as it is copied to that
> > > machine (an hour or so). What options do you want me to switch on?
> >
> >If you are seeing lockups probably all the lockup detection code. That
> >might trigger the occasional false positive.
>
> I have been experimenting with ikd and some of the memleak scripts while I
> wait for other things to happen. What are reasonable values to see from
> memleak/dosum ? It looks like dcache.c:492 continues to grow, while
> everything else flucuates. Below is some info from my PPro 200 that has
> 192M ram.
Kill as much cache as you can (run a memory hog), and it should drop
sharply. If it doesn't, it's likely a leak. There are no 'reasonable'
values to expect. Memleak shows the number of active allocations which
originated from a particular spot. In some cases, the numbers look higher
than expected because the allocation 'cause' is passed down to lower level
allocators.. let's you see what kind off memory is being lost in event
of a leak. As a result, the originator may be credited with lower level
allocations which were caused by servicing a higher level allocator's
request. (adjustable via MEMLEAK_PASS_ALLOCATION and MEMLEAK_UNWRAP_XXX)
-Mike
-
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/