This is the mail archive of the systemtap@sourceware.org mailing list for the systemtap project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [RFC -mm][PATCH 5/6] prepare kprobes code for x86 unification


On Wed, Dec 12, 2007 at 08:21:01PM -0500, Masami Hiramatsu wrote:
> Masami Hiramatsu wrote:
> > Finally, I reproduced it by executing a command below:
> > % stap -e 'probe kernel.function("migration_init"){}'
> > 
> > And also, I found that oops occurred without my patch,
> > and even more, it occurred with 2.6.24-rc4 on x86-64,
> > but did not occurred with 2.6.23.8-34.fc7.
> >
> > We might find an unknown bug.
> 
> I compared both .config files and found that this bug
> did not occur with CONFIG_RELOCATABLE=y.
> I checked this with vanilla-2.6.23 on x86-64.
> 
> I also read the contents of the page which included migration_init().
> It was filled by "0xcc"(POISON_FREE_INITMEM).
> I think those freed pages are marked as freed and are not purged
> from the pagetable when CONFIG_RELOCATABLE=y.
> 
> Anyway, probing those "__init" functions after the kernel freed it
> is meaningless and even dangerous . I think kprobes should check it
> and reject it when registering a probe.

That does seem like the right thing to do. This also needs to live in
common code.

Srini,
Want to take a stab at it?

Ananth


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]