Re: [PATCH] Fix kallsyms/insmod/rmmod race

[David Howells]

Rusty Russell <rusty@xxxxxxxxxxxxxxx> wrote:

> 	The more I looked at this, the more I warmed to it.  I've known for a
> while that people are using kallsyms not for OOPS (eg. /proc/$$/wchan),
> so we should provide a "grabs locks" version, but this solution gets
> around that nicely, while making life more certain for the oops case,
> too.


Hmmm... though it works on i386 SMP, it doesn't, however, seem to work on
ppc64 SMP:-/

My pSeries box seems to think that it can't find any symbols from previously
loaded modules, and my Power5 box is quite happy to load modules that depend
on other modules but panics because it can't mount its root fs.

This is very odd, because the patch is simple enough. Is there anything
obvious I've missed that you can see? Or maybe I'm just misunderstanding how
stop_machine_run() works... maybe it can't be called during initialisation.

David
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/