Re: headerscheck warnings with late-model gcc

Peter Eisentraut <peter(at)eisentraut(dot)org> writes:
> Attached are three patches to fix some unrelated problems with
> headerscheck in my environment.

0001 seems fine; it's an oversight that I'd not noticed because
ICU_CFLAGS is empty in my usage.

Don't like 0002 as-is. I'd be okay with skipping that header if
--with-llvm isn't given. However, we already tell users that
they'd better configure --with-perl and --with-python, so maybe
just add --with-llvm to that list?

0003: +1, I noticed that too yesterday.

0004: I prefer the solution I exhibited yesterday, ie add
externs to those headers.

> kwlist_d.h doesn't show up in my run, probably because I'm using a
> separate build directory, which headerscheck doesn't handle? Another
> thing to fix.

Yeah, as it stands headerscheck is really only meant for in-tree
builds. We will definitely have to do something about that to make
it handle built headers in meson builds, and I guess it'd be nice if
VPATH works that way as well. One idea could be to redefine it as
searching the installation include-file tree instead of the source
tree, so that the build process washes out of the matter.

regards, tom lane