Re: use --enable-pic [libtool--gary--1.0--patch-32]

[Gary V. Vaughan]

Hallo Ralf,

Thanks for the review.  So much for my thinking this was a quick 
5-minute patch, eh? ;-)

Ralf Wildenhues wrote:
> * Gary V. Vaughan wrote on Tue, Aug 30, 2005 at 02:42:00PM CEST:
> > Ralf Wildenhues wrote:
> > > * Gary V. Vaughan wrote on Tue, Aug 30, 2005 at 01:30:35PM CEST:
> > > > This patch continues to support --with{,out}-pic[={yes,no,both}, but
> > > > only advertises the more compliant --enable-pic in the calling
> > > > configure --help text.
>
> Since the naming was unfortunate, and only documented in the test suite
> documentation, I think it's OK to change from --with to --enable.
> However, I'd like to test that it works before I approve.  :)

Okay.

> > http://tkd.kicks-ass.net/arch/address@hidden/libtool--gary--1.0--patch-32?cmd=view_patches
>
> First bug present in HEAD: the old macro was apparently named
> AC_LIBTOOL_PICMODE, not AC_LIBTOOL_PIC_MODE.  Please also fix that
> (2 instances) although it's orthogonal, thank you.

Fix applied to HEAD already.

> Then, NEWS should get an entry (since with-pic was mentioned there,
> too).  Also, tests/demo-{pic,nopic}.test need updates.

Agreed.

> Third, isn't the logic the other way round: instead of
> |  @item pic-only
> |  Change the default behaviour for @command{libtool} to try to use only
> |  @sc{pic} objects.  The user may still override this default by specifying
> | address@hidden to @command{configure}.
> | address@hidden to @command{configure}.
>
> shouldn't it be
> |  @item pic-only
> |  Change the default behaviour for @command{libtool} to try to use only
> |  @sc{pic} objects.  The user may still override this default by specifying
> | address@hidden to @command{configure}.
> | address@hidden to @command{configure}.

Yep.  That probably means the manual for 1.5 is wrong too... maybe this 
patch (libtool--gary--1.0--patch-32..33) should be backported infact?

> Fourth, the entry no-pic (a few lines below that) would need a similar
> update, too, I believe.  :)

Agreed.

> Fifth, this can only be tested comprehensively (including the old macro
> naming) if the ltoptions bug I just described is fixed.  ;->

Proposed fix for that bug posted separately.

I've also addressed all of your concerns against -patch-32, as -patch-33
(posted separately) -- I'll merge both to HEAD and commit as one 
changeset though.

Cheers,
        Gary.
--
Gary V. Vaughan      ())_.  address@hidden,gnu.org}
Research Scientist   ( '/   http://tkd.kicks-ass.net
GNU Hacker           / )=   http://www.gnu.org/software/libtool
Technical Author   `(_~)_   http://sources.redhat.com/autobook

signature.asc
Description: OpenPGP digital signature