Re: Reference source for C-compiler validation?

Martin von Loewis <loewis@informatik.hu-berlin.de> wrote:
>jbredno@yahoo.de (J. Bredno) writes:
>
>> Simple question, but I'm afraid the answer won't be so simple:
>
>> Are there any widely accepted C and C++ source files that can be
>> used to validate a compiler? E.g., is there some test for ANSI
>> compliance or something similar?
>
>There are certainly widely-accepted C and C++ test suites. I don't
>think any of them is widely available. Common test suites are
>
>Plum Hall: http://www.plumhall.com/suites.html
>Perennial: http://www.peren.com/pages/cvsa_set.htm


There are, but there are some serious problems with them: the state of
the C standard. I have not been following C++.


The syntax of C is moderately well-defined, though there are some
areas where there is considerable dissent over what is specified.
Look at comp.std.c for more evidence than you want to see :-(


The semantics, however, isn't. In most places, the intent is clear,
but it is rarely agreed whether the intent is a requirement or merely
a guideline to recommended practice. In some other (often important)
places, the intent isn't clear, is known only from external sources or
even isn't agreed.


Any validation suite has to decide what language to test against, with
the obvious problems. I have listened to a debate between the author
of a validation suite and someone needed to prove conformance for a
contract on this very matter.


Regards,
Nick Maclaren,
University of Cambridge Computing Service,
New Museums Site, Pembroke Street, Cambridge CB2 3QG, England.
Email: nmm1@cam.ac.uk
Tel.: +44 1223 334761 Fax: +44 1223 334679

Return to the comp.compilers page.
Search the comp.compilers archives again.