| Summary: | Bug in bsd.port.mk when USE_GCC not defined | ||||||
|---|---|---|---|---|---|---|---|
| Product: | Ports & Packages | Reporter: | David E. O'Brien <obrien> | ||||
| Component: | Individual Port(s) | Assignee: | Port Management Team <portmgr> | ||||
| Status: | Closed FIXED | ||||||
| Severity: | Affects Only Me | ||||||
| Priority: | Normal | ||||||
| Version: | Latest | ||||||
| Hardware: | Any | ||||||
| OS: | Any | ||||||
| Attachments: |
|
||||||
|
Description
David E. O'Brien
2004-04-20 01:50:18 UTC
Responsible Changed From-To: freebsd-ports-bugs->portmgr for the testing queue. On Mon, Apr 19, 2004 at 05:43:17PM -0700, David O'Brien wrote:
> A port no longer is told to use the compiler defined by 'CC' in the
> normal case of "USE_GCC" not being set.
>
> I think a ".else" was left out of the "USE_GCC" logic.
Actually I think my previous change in this area was too overzealous;
I moved the MAKE_ENV setting under the .ifdef USE_GCC, forgetting that
it's also needed in other cases. The fix would be to put it back
where it was.
Kris
On Mon, Apr 19, 2004 at 05:43:17PM -0700, David O'Brien wrote:
> >Description:
>
> A port no longer is told to use the compiler defined by 'CC' in the
> normal case of "USE_GCC" not being set.
>
> I think a ".else" was left out of the "USE_GCC" logic.
>
> >How-To-Repeat:
> $ make -V MAKE_ENV # shows no CC=
>
> >Fix:
>
> Index: bsd.port.mk
> ===================================================================
> RCS file: /home/ncvs/ports/Mk/bsd.port.mk,v
> retrieving revision 1.487
> diff -u -r1.487 bsd.port.mk
> --- bsd.port.mk 19 Apr 2004 01:37:11 -0000 1.487
> +++ bsd.port.mk 20 Apr 2004 00:39:37 -0000
> @@ -1320,6 +1321,7 @@
> BUILD_DEPENDS+= gcc34:${PORTSDIR}/lang/gcc34
> GCCVERSION= 030400
> .endif
> +.else
> MAKE_ENV+= CC="${CC}" CXX="${CXX}"
> .endif
Actually, I can't find evidence that the behaviour you describe was
previously the case. All I did in the recent USE_GCC commit was move
a bunch of separate MAKE_ENV settings into the outer .ifdef USE_GCC.
Kris
On Mon, Apr 19, 2004 at 11:36:23PM -0700, Kris Kennaway wrote: > > >Description: > > > > A port no longer is told to use the compiler defined by 'CC' in the > > normal case of "USE_GCC" not being set. > > > > I think a ".else" was left out of the "USE_GCC" logic. > > > > >How-To-Repeat: > > $ make -V MAKE_ENV # shows no CC= > > > > >Fix: > > +.else > > MAKE_ENV+= CC="${CC}" CXX="${CXX}" > > .endif > > Actually, I can't find evidence that the behaviour you describe was > previously the case. All I did in the recent USE_GCC commit was move > a bunch of separate MAKE_ENV settings into the outer .ifdef USE_GCC. Far enought. Do you think the ".else" should be added as a feature request then? -- -- David (obrien@FreeBSD.org) On Tue, Apr 20, 2004 at 06:32:27AM -0700, David O'Brien wrote:
> On Mon, Apr 19, 2004 at 11:36:23PM -0700, Kris Kennaway wrote:
> > > >Description:
> > >
> > > A port no longer is told to use the compiler defined by 'CC' in the
> > > normal case of "USE_GCC" not being set.
> > >
> > > I think a ".else" was left out of the "USE_GCC" logic.
> > >
> > > >How-To-Repeat:
> > > $ make -V MAKE_ENV # shows no CC=
> > >
> > > >Fix:
> > > +.else
> > > MAKE_ENV+= CC="${CC}" CXX="${CXX}"
> > > .endif
> >
> > Actually, I can't find evidence that the behaviour you describe was
> > previously the case. All I did in the recent USE_GCC commit was move
> > a bunch of separate MAKE_ENV settings into the outer .ifdef USE_GCC.
>
> Far enought. Do you think the ".else" should be added as a feature
> request then?
Well, that would still be wrong because it would avoid adding MAKE_ENV
when USE_GCC is set. Doing what you mean, I'm not sure..how many
ports would this affect?
Kris
On Tue, Apr 20, 2004 at 03:03:27PM -0700, Kris Kennaway wrote: > On Tue, Apr 20, 2004 at 06:32:27AM -0700, David O'Brien wrote: > > On Mon, Apr 19, 2004 at 11:36:23PM -0700, Kris Kennaway wrote: > > > > >Description: > > > > > > > > A port no longer is told to use the compiler defined by 'CC' in the > > > > normal case of "USE_GCC" not being set. > > > > > > > > I think a ".else" was left out of the "USE_GCC" logic. > > > > > > > > >How-To-Repeat: > > > > $ make -V MAKE_ENV # shows no CC= > > > > > > > > >Fix: > > > > +.else > > > > MAKE_ENV+= CC="${CC}" CXX="${CXX}" > > > > .endif > > > > > > Actually, I can't find evidence that the behaviour you describe was > > > previously the case. All I did in the recent USE_GCC commit was move > > > a bunch of separate MAKE_ENV settings into the outer .ifdef USE_GCC. > > > > Far enought. Do you think the ".else" should be added as a feature > > request then? > > Well, that would still be wrong because it would avoid adding MAKE_ENV > when USE_GCC is set. That's quite true. I wonder if the MAKE_ENV+= CC="${CC}" CXX="${CXX}" should be down in all cases. I thought it was in the distant past, but maybe I'm remembering wrong. -- -- David (obrien@FreeBSD.org) State Changed From-To: open->closed This wasn't a bug. |