Re: 329-gary-allow-RTLD_GLOBAL

[Gary V. Vaughan]

Morgen Ralf,

On 21 May 2007, at 18:20, Gary V. Vaughan wrote:
> > * Gary V. Vaughan wrote on Tue, May 08, 2007 at 04:42:09PM CEST:
> > > That would defeat building moddepend with unresolved symbols at  
> > > link time,
> > > so that we can test that it correctly calls the matching symbols  
> > > from
> > > modglobal after load time symbol resolution :-(  Known issue, but  
> > > I don't
> > > know how to fix it.  I guess we'll need to skip this test on  
> > > Windows and
> > > AIX due to linker brain damage?
> >
> > If we linked `moddepend'
> > against the other modules it depends upon at mode=link time, that
> > would work on w32, except now libtool still outputs the warning
> > | *** Warning: Linking the executable m against the loadable module
> > | *** liba.so is not portable!
> >
> > I guess a proper test would use a separate, real shared library
> > which moddepend would then depend upon.  WDYT?
>
> Yes that is doable, and is in fact how m4 modules are built on  
> HEAD, which
> works on linux, darwin and w32 (among others).  I'll make a start at
> implementing these changes while I wait for the 330 review.

Looking at this again in the fresh light of day, I realised that the  
entire
point of the moddepend test is to check that its unresolved symbols  
can be
satisfied from the program that lt_dlopen()'s it when using the  
global hint.
It makes no sense to resolve the symbols at link time when trying to  
test
that the hint is correctly causing them to resolve at load time.

The only way I see forward is to skip this test on hosts that require  
all
symbols to be resolved at link time... Should I also emit a 'not  
portable'
warning when using this feature on the hosts that do support it?

Cheers,
        Gary
--
  ())_.              Email me: address@hidden
  ( '/           Read my blog: http://blog.azazil.net
  / )=         ...and my book: http://sources.redhat.com/autobook
`(_~)_ Join my AGLOCO Network: http://www.agloco.com/r/BBBS7912

PGP.sig
Description: This is a digitally signed message part