Re: Testsuite de-Linuxification

[Thomas Schwinge <thomas at codesourcery dot com>]

Hi!

On Mon, 7 Jan 2013 16:19:47 +0000, "Maciej W. Rozycki" <macro@codesourcery.com> wrote:
> > diff --git ld/testsuite/ld-i386/export-class.exp ld/testsuite/ld-i386/export-class.exp
> > index cd96b4b..9dedd90 100644
> > --- ld/testsuite/ld-i386/export-class.exp
> > +++ ld/testsuite/ld-i386/export-class.exp
> > @@ -29,9 +29,8 @@ if { ![is_elf_format] } {
> >      return
> >  }
> >  
> > -# Exclude non-Linux targets; feel free to include your favorite one
> > -# if you like.
> > -if { ![istarget i?86-*-linux*] } {
> > +if { !([istarget i?86-*-linux*]
> > +       || [istarget i?86-*-gnu*]) } {
> >      return
> >  }
> >  
> 
>  I disagree with the removal of the comment (although a rewording would be 
> due of course), there are other x86 targets that use ELF shared libraries, 
> such as SCO OpenServer or Unixware, and that are likely to support export 
> classes.  Someone who cares about such targets may be still interested in 
> adding them to this test case.

You're right; I meant to do the same rewording as I had done in commit
16462cec50da2d070bd7d41fc4092a6abb469953 (2012-02-20).  I have now
committed the following:

ld/testsuite/
	* ld-i386/export-class.exp: Restore (and reword) comment about
	excluded targets.

Index: ld/testsuite/ld-i386/export-class.exp
===================================================================
RCS file: /cvs/src/src/ld/testsuite/ld-i386/export-class.exp,v
retrieving revision 1.3
diff -u -p -r1.3 export-class.exp
--- ld/testsuite/ld-i386/export-class.exp	17 Dec 2012 19:32:52 -0000	1.3
+++ ld/testsuite/ld-i386/export-class.exp	8 Jan 2013 10:23:56 -0000
@@ -29,6 +29,8 @@ if { ![is_elf_format] } {
     return
 }
 
+# Exclude some more targets; feel free to include your favorite one
+# if you like.
 if { !([istarget i?86-*-linux*]
        || [istarget i?86-*-nacl*]
        || [istarget i?86-*-gnu*]) } {


>  That said, I have since discovered the existence of the 
> check_shared_lib_support procedure -- perhaps we should simply use that 
> instead to relieve the maintainers (if any) of the more exotic platforms 
> from the burden of chasing such subtleties across the test suite.  

Fine my me.


GrÃÃe,
 Thomas

Attachment: pgp00000.pgp
Description: PGP signature