Re: Patch for endianness issues in 4.3a

[email protected]
Newsgroups gmane.lisp.scheme.bigloo
Organization Inria
Message-ID <201792164153.59c34311@arches>
Hi Jerry,

> Greetings,
> 
> I tried to update the Fedora bigloo package to 4.3a once hop was
> updated to use the newer version, but the bigloo build failed on
> s390x.  We had already blacklisted the big endian ppc64 architecture
> for bigloo because of build failures there.  I decided to see if I
> could find the problem.  The attached patch gets the build to succeed
> on both s390x and big endian ppc64.
Wonderful. I have apply the patch. Thanks a million.

> The problem is mismatched declarations and definitions, as you can see
> by examining the patch.  I don't mean to be critical, but extern
> declarations in C files are a terrible idea, for precisely this
> reason.  If the declarations and definitions get out of sync, the
> compiler can't tell you about it, because it has no way of knowing.
> All of those declarations should be in header files that are included
> by both the defining code and the consuming code.
Yes. I agree. You are right. This is generally by laziness that we do not
include a legal C header file and we run in the very trouble you are
mentioning.

> Note that the patched code is wrong even for little endian
> architectures such as x86.  It just happens to "work" by accident.
What do you mean? What do you think is wrong?

Best,

-- 
Manuel
lmpx.com only provides a reader for public news (NNTP) servers. It is not affiliated with the servers or forums shown here and is not responsible for the content of articles, which is written by their respective authors.