Am 17.10.2017 um 02:19 schrieb Yann Ylavic:
On Tue, Oct 17, 2017 at 1:23 AM, Yann Ylavic <[email protected]> wrote:
On Tue, Oct 17, 2017 at 12:48 AM, William A Rowe Jr <[email protected]> wrote:
Rainer,
https://ci.apache.org/builders/httpd-trunk/builds/1203
would you please re-kick this build from a clean svn checkout? I think we have
various mistakes in our exports.c preprocessor that become tangled in any
rebuild scenario.
Hmm, not sure, looks like an issue with the "inline" keyword used in
some system (systemd) header file, and c89.
According to github, systemd's master branch uses __inline__ there
instead (which should be fine with gcc's c89), so dunno if we can do
something like this (when -stdc=c89):
#undef inline
#define inline __inline__ /* or APR_INLINE? */
#include <sd-journal.h>
Nor if it can even work to #undef "inline"...
Looks fixed now: https://ci.apache.org/builders/httpd-trunk/builds/1205
Thanks Yann for fixing.
For those who didn't follow that closely: we now fixed the maintainer
build (which eg. is used by buildbot) to actually use the gcc flags that
we expected. All of the flags are tested for gcc supporting them before
adding them and those tests partially failed not due to gcc not
supporting them but due to the tests generating warnings and due to
-Werror actually failing.
Now that we have fixed maintainer mode, builds can fail due to new build
errors, that previously only showed up as warnings. The above is one
such case and Joe, Yann and me fixed at least three such cases in trunk.
Currently it looks like the build is fine, but depending on your
dependency libraries and the set of modules you build there's still a
little risk, that you might observe new failures.
Regards,
Rainer
