Re: [v3] Minimally exercise the other alias_decls in type_traits

2014-10-16 Thread Paolo Carlini
... the below completes the work, no -std=gnu++0x anymore. Paolo. / 2014-10-16 Paolo Carlini paolo.carl...@oracle.com * testsuite/lib/libstdc++.exp: Prefer -std=gnu++11. * testsuite/20_util/add_lvalue_reference/requirements/

[v3] Minimally exercise the other alias_decls in type_traits

2014-10-15 Thread Paolo Carlini
Hi, I committed the below. Thanks, Paolo. / 2014-10-15 Paolo Carlini paolo.carl...@oracle.com * testsuite/20_util/add_const/requirements/alias_decl.cc: New. * testsuite/20_util/add_cv/requirements/alias_decl.cc: Likewise. *

Re: [v3] Minimally exercise the other alias_decls in type_traits

2014-10-15 Thread Jonathan Wakely
On 15/10/14 11:27 +0200, Paolo Carlini wrote: * testsuite/20_util/add_const/requirements/alias_decl.cc: New. * testsuite/20_util/add_cv/requirements/alias_decl.cc: Likewise. * testsuite/20_util/add_pointer/requirements/alias_decl.cc: Likewise. *

Re: [v3] Minimally exercise the other alias_decls in type_traits

2014-10-15 Thread Jonathan Wakely
On 15/10/14 10:44 +0100, Jonathan Wakely wrote: We might also want to consider a global s/gnu++0x/gnu++11/ some day. One reason for doing that replacement is that I keep copying existing tests and then having to change it, and I'm lazy!

Re: [v3] Minimally exercise the other alias_decls in type_traits

2014-10-15 Thread Paolo Carlini
Hi, On 10/15/2014 11:44 AM, Jonathan Wakely wrote: Index: testsuite/20_util/add_const/requirements/alias_decl.cc === --- testsuite/20_util/add_const/requirements/alias_decl.cc (revision 0) +++

Re: [v3] Minimally exercise the other alias_decls in type_traits

2014-10-15 Thread Paolo Carlini
Hi again, On 10/15/2014 11:50 AM, Paolo Carlini wrote: We might also want to consider a global s/gnu++0x/gnu++11/ some day. And s/gnu++1y/gnu++14 of course ;) I'm going to commit the below. Thanks, Paolo. / 2014-10-15 Paolo Carlini paolo.carl...@oracle.com *

Re: [v3] Minimally exercise the other alias_decls in type_traits

2014-10-15 Thread Mike Stump
On Oct 15, 2014, at 8:41 AM, Paolo Carlini paolo.carl...@oracle.com wrote: ... the below replaces -std=c++0x and -std=c++11 with -std=gnu++11. Seems odd to go from c++11 - gnu++11. Why?

Re: [v3] Minimally exercise the other alias_decls in type_traits

2014-10-15 Thread Mike Stump
On Oct 15, 2014, at 2:45 AM, Jonathan Wakely jwak...@redhat.com wrote: On 15/10/14 10:44 +0100, Jonathan Wakely wrote: We might also want to consider a global s/gnu++0x/gnu++11/ some day. One reason for doing that replacement is that I keep copying existing tests and then having to change

Re: [v3] Minimally exercise the other alias_decls in type_traits

2014-10-15 Thread Paolo Carlini
On 10/15/2014 07:35 PM, Mike Stump wrote: On Oct 15, 2014, at 8:41 AM, Paolo Carlini paolo.carl...@oracle.com wrote: ... the below replaces -std=c++0x and -std=c++11 with -std=gnu++11. Seems odd to go from c++11 - gnu++11. Why? Because all the new testcases we are committing use

Re: [v3] Minimally exercise the other alias_decls in type_traits

2014-10-15 Thread Paolo Carlini
Hi, On 10/15/2014 11:44 AM, Jonathan Wakely wrote: We might also want to consider a global s/gnu++0x/gnu++11/ some day. The below takes care of 17-19, the rest over the next days... Paolo. // 2014-10-15 Paolo Carlini paolo.carl...@oracle.com *