This is the mail archive of the
libc-alpha@sourceware.org
mailing list for the glibc project.
Re: [PATCH 6/6] Installed header hygiene (BZ#20366): Test of installed headers.
On Fri, Sep 23, 2016 at 8:35 AM, Joseph Myers <joseph@codesourcery.com> wrote:
> On Fri, 23 Sep 2016, Zack Weinberg wrote:
>
>> On 09/22/2016 04:21 PM, Joseph Myers wrote:
>> > This patch is OK. Does it test headers installed from the toplevel
>> > Makefile as well as those installed by subdirectories?
>>
>> I wasn't aware there were any such headers! Indeed they were not being
>> tested.
>
> There's an argument for reducing the amount of stuff done by the toplevel
> Makefile - for example, by making include/ into a normal subdirectory the
> build recurses into, and installing the miscellaneous headers from there.
Or those headers could be moved to misc/, which might be less
confusing (and less work). include/ would then contain only wrapper
and internal-use-only headers.
> begin-end-check currently involves a hardcoded list of header paths
> (installed-headers in the toplevel Makefile). It would be better for it
> to run in each directory and work off the headers list from that
> directory, like your test does, but doing so without duplicating rules
> would also imply moving toplevel header installation to a subdirectory.
I saw that and I was thinking about doing a follow-on patch, but I've
got a lot on my plate for $dayjob right now...
zw