On Sun, Mar 30, 2014 at 1:00 PM, gatesphere <[email protected]> wrote:

> Thought  [pyringe] might be of interest given Edward's recent work on 
> g.SherlockTracer...

Thanks for the link.

The interesting part appears to be pyringe/payload/gdb_service.py,
which (unless I am mistaken) patches the execution stack frame.  It's
an interesting hack.

This would have relevance for pep 3107: copyable generators.   This is
a very serious problem when tracing: there is *no* way to copy a
generator on the stack frame: itertools.tee does create two copies of
a generator, but the original becomes unusable, so tee can not be used
when the original is in a Python stack frame.

The pyringe way might work: create a proxy for the actual frame, then
put one of the results of tee in the frame.  Otoh, pyringe has, by its
own admission, some limitations.  It seems dubious to base a debugging
tool on a limited technique: when things go wrong we have to wonder:
is it a problem with the program under test or the testing tool
itself.

Edward

-- 
You received this message because you are subscribed to the Google Groups 
"leo-editor" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To post to this group, send email to [email protected].
Visit this group at http://groups.google.com/group/leo-editor.
For more options, visit https://groups.google.com/d/optout.

Reply via email to