Re: [csmith-dev] CSmith and tailcall optimisations

[Eric Eide <eeide@cs.utah.edu>]

Maxime Dénès <mail@maximedenes.fr> writes:

	Maxime> What do you think of this? Would you expect CSmith to be able
	Maxime> to find it?

You mean, should Csmith be able to generate a program that would allow some
other tool to observe the bug?

John has already talked answering this question in general, for arbitrary
compiler defects.

But for answering the question about the particular defect you introduced, I
took a quick look at Csmith's code.

I believe that most/all "return" statements generated current Csmith are of the
form "return var", which means that it will probably never generate a test
program of the style you're looking for.

(Why does it generate "return var", not "return expr"?  I don't recall.
Answering that is a harder question!)

Eric.

-- 
-------------------------------------------------------------------------------
Eric Eide <eeide@cs.utah.edu>  .         University of Utah School of Computing
http://www.cs.utah.edu/~eeide/ . +1 (801) 585-5512 voice, +1 (801) 581-5843 FAX