Re: [DynInst_API:] Snippets that need libs that the mutatee does not reference

[Bill Williams <bill@xxxxxxxxxxx>]

On 10/18/2013 02:59 PM, Wayne Motycka wrote:
> Bill Williams <bill <at> cs.wisc.edu> writes:
> ...
> > There are a couple of ways you can go here, with varying levels of
> > programmer and execution overhead. You can use BPatch_malloc to create
> > instrumentation variables of complex data types and manipulate their
> > fields with snippets, or you can inject an instrumentation library. The
> > standard technique when using an instrumentation library would be to
> > insert a call to some function that updates your assorted in-memory data
> > structures. In either case, you'd then register an exit callback that
> > will serialize the results to disk. The instrumentation library tends to
> > be easier to implement, and as your data structures grow more complex,
> > it will tend to be more efficient as well.
> Ah, instrumentation library sounds like the way to go then.  If I understand
> this it means that I'd need to define the in-memory data structure that
> the instrumentation library would manipulate/update and the exit callback
> would access this data structure to write it to disk, right?
Yep. The data structure can be a static global (and initialized when the 
library is loaded), or you can initialize it with a OneTimeCode before 
you start the process back up.
> Also, the library should be a shared lib (.so) and not an archive (.a), right?
Assuming you're working with a live process and not doing binary 
rewriting, a .so is the right tool. If you're doing binary rewriting, 
you can easily add either a shared lib or an archive, depending on how 
the original binary was linked.
> -Wayne
>
>
> _______________________________________________
> Dyninst-api mailing list
> Dyninst-api@xxxxxxxxxxx
> https://lists.cs.wisc.edu/mailman/listinfo/dyninst-api

--
--bw

Bill Williams
Paradyn Project
bill@xxxxxxxxxxx