RFA: Don't use -Wuninitialized if CFLAGS omits -O

RFA: Don't use -Wuninitialized if CFLAGS omits -O

[Jim Blandy]

I've committed this.

> This allows gdb_mbuild.sh to work with CFLAGS=-g, which improves
> compilation speed and helps with debugging.
> 
> 2005-02-28  Jim Blandy  <jimb@redhat.com>
> 
> 	Avoid warnings due to the use of -Wuninitialized without -O.
> 	* configure.ac (build_warnings): Include -Wuninitialized only if
> 	CFLAGS is unset at configure time, or CFLAGS is set and includes
> 	some -O option other than -O0.
> 	* configure: Regenerated.

Re: RFA: Don't use -Wuninitialized if CFLAGS omits -O

[Joel Brobecker]

Jim,

> 2005-02-28  Jim Blandy  <jimb@redhat.com>
> 
> 	Avoid warnings due to the use of -Wuninitialized without -O.
> 	* configure.ac (build_warnings): Include -Wuninitialized only if
> 	CFLAGS is unset at configure time, or CFLAGS is set and includes
> 	some -O option other than -O0.
> 	* configure: Regenerated.

I think there is a slight problem that broke the build with non-GNU
bourne shells. We have a build failure on our solaris 2.8 machine,
for instance. Here is what I think the problem is:

> + # GCC supports -Wuninitialized only with -O or -On, n != 0.
> + if test x${CFLAGS+set} == xset; then
                           ^^
I think the "==" should be "=", no? That's what "man test" documents.
But I've never been a shell guru...
                          
-- 
Joel

Re: RFA: Don't use -Wuninitialized if CFLAGS omits -O

[Andreas Schwab]

Joel Brobecker <brobecker@adacore.com> writes:

>> + # GCC supports -Wuninitialized only with -O or -On, n != 0.
>> + if test x${CFLAGS+set} == xset; then
>                            ^^
> I think the "==" should be "=", no? That's what "man test" documents.

Yes, "=" is correct, "==" is implemented in some shells (like bash and
ksh) as extension.

Andreas.

-- 
Andreas Schwab, SuSE Labs, schwab@suse.de
SuSE Linux Products GmbH, Maxfeldstraße 5, 90409 Nürnberg, Germany
Key fingerprint = 58CA 54C7 6D53 942B 1756  01D3 44D5 214B 8276 4ED5
"And now for something completely different."

Re: RFA: Don't use -Wuninitialized if CFLAGS omits -O

[Daniel Jacobowitz]

On Sat, Mar 12, 2005 at 01:05:06AM -0800, Joel Brobecker wrote:
> Jim,
> 
> > 2005-02-28  Jim Blandy  <jimb@redhat.com>
> > 
> > 	Avoid warnings due to the use of -Wuninitialized without -O.
> > 	* configure.ac (build_warnings): Include -Wuninitialized only if
> > 	CFLAGS is unset at configure time, or CFLAGS is set and includes
> > 	some -O option other than -O0.
> > 	* configure: Regenerated.
> 
> I think there is a slight problem that broke the build with non-GNU
> bourne shells. We have a build failure on our solaris 2.8 machine,
> for instance. Here is what I think the problem is:
> 
> > + # GCC supports -Wuninitialized only with -O or -On, n != 0.
> > + if test x${CFLAGS+set} == xset; then
>                            ^^
> I think the "==" should be "=", no? That's what "man test" documents.
> But I've never been a shell guru...

Yes, that's right.  Would you mind fixing this?

-- 
Daniel Jacobowitz
CodeSourcery, LLC

Re: RFA: Don't use -Wuninitialized if CFLAGS omits -O

[Joel Brobecker]

> > I think there is a slight problem that broke the build with non-GNU
> > bourne shells. We have a build failure on our solaris 2.8 machine,
> > for instance. Here is what I think the problem is:
> > 
> > > + # GCC supports -Wuninitialized only with -O or -On, n != 0.
> > > + if test x${CFLAGS+set} == xset; then
> >                            ^^
> > I think the "==" should be "=", no? That's what "man test" documents.
> > But I've never been a shell guru...
> 
> Yes, that's right.  Would you mind fixing this?

Sure, wildo later today.

-- 
Joel

Re: RFA: Don't use -Wuninitialized if CFLAGS omits -O

[Jim Blandy]

Joel Brobecker <brobecker@adacore.com> writes:

> > > I think there is a slight problem that broke the build with non-GNU
> > > bourne shells. We have a build failure on our solaris 2.8 machine,
> > > for instance. Here is what I think the problem is:
> > > 
> > > > + # GCC supports -Wuninitialized only with -O or -On, n != 0.
> > > > + if test x${CFLAGS+set} == xset; then
> > >                            ^^
> > > I think the "==" should be "=", no? That's what "man test" documents.
> > > But I've never been a shell guru...
> > 
> > Yes, that's right.  Would you mind fixing this?
> 
> Sure, wildo later today.

Beat you to it.  Sorry about that.