Re: coverage analysis improvements

On Tue, Sep 26, 2017 at 3:45 PM, Michael Paquier
<michael(dot)paquier(at)gmail(dot)com> wrote:
> Except for the plperl patch, I don't have more comments to offer about
> this patch set. It would be nice to make configure a bit smarter for
> lcov and gcov detection by not hard-failing if gcov can be found but
> not lcov. It is after all possible to run coverage without lcov, like
> on ArchLinux. Still that's a different request than what this patch
> set is doing, so this is not a blocker for this patch set.

-SPI.c: SPI.xs plperl_helpers.h
+%.c: %.xs
@if [ x"$(perl_privlibexp)" = x"" ]; then echo "configure switch
--with-perl was not specified."; exit 1; fi
- $(PERL) $(XSUBPPDIR)/ExtUtils/xsubpp -typemap
$(perl_privlibexp)/ExtUtils/typemap $< >$@
-
-Util.c: Util.xs plperl_helpers.h
- @if [ x"$(perl_privlibexp)" = x"" ]; then echo "configure switch
--with-perl was not specified."; exit 1; fi
- $(PERL) $(XSUBPPDIR)/ExtUtils/xsubpp -typemap
$(perl_privlibexp)/ExtUtils/typemap $< >$@
+ $(PERL) $(XSUBPPDIR)/ExtUtils/xsubpp -typemap
$(perl_privlibexp)/ExtUtils/typemap -output $@ $<
I just looked at the plperl portion of this patch, and I think that
01d83ffd has done some unnecessary things here. Contrary to
perlchunks.h and plperl_opmask.h which are generated during the build,
plperl_helpers.h is part of the source code so there is no meaning in
having a dependency with it. The .c files do not need this header
anyway, but their .o files would, still there is no need for that
either as plperl_helpers.h will not go away.

I am marking the full set of patches as ready for committer.
--
Michael