Re: SunOS patch for memcmp()

Bruce Momjian writes:

> > What should we do? The only remaining issue is a non-8-bit-clean
> > memcmp, which seems pretty easy to fix it.
>
> OK, here is a patch to allow SunOS to pass the regression tests.

At a glance, this patch looks okay. However, since this issue does not
represent a regression from 7.1 I'm not exactly in favour of installing it
now. We want to get a release out, so I think we need to get stricter in
those matters. And this patch is not exactly trivial.

> One interesting item is that I had to compile with
> backend/utils/adt/varbit.c with -fno-builtin because my gcc 2.X manual
> says:

Yeah, if you use GCC with optimization on SunOS 4 then the issue should be
moot because the GCC version is used. However, I don't think that's the
setup in question.

This is actually a good situation to show that configuring with one kind
of compiler flag and building with another is not reliable, even if it's
only the optimization level.

--
Peter Eisentraut peter_e(at)gmx(dot)net