[libcxx] r220724 - Add special case handling of linux target triples that do not contain `-gnu`.

[Eric Fiselier]

Woops. That was not the example I intended to give. Ill fix it. Thanks for
pointing this out.

/Eric
On Oct 28, 2014 8:51 AM, "Hal Finkel" <hfinkel at anl.gov> wrote:

> ----- Original Message -----
> > From: "Eric Fiselier" <eric at efcs.ca>
> > To: cfe-commits at cs.uiuc.edu
> > Sent: Monday, October 27, 2014 5:14:25 PM
> > Subject: [libcxx] r220724 - Add special case handling of linux target
> triples that do not contain `-gnu`.
> >
> > Author: ericwf
> > Date: Mon Oct 27 17:14:25 2014
> > New Revision: 220724
> >
> > URL: http://llvm.org/viewvc/llvm-project?rev=220724&view=rev
> > Log:
> > Add special case handling of linux target triples that do not contain
> > `-gnu`.
> >
> > For targets that end it `redhat-linux` and `suse-linux` manually add
> > the `-gnu`
> > section of the target since `linux-gnu` is needed in the testsuite.
> >
> > This patch also moves the removal of minor and patchlevel numbers
> > from OSX
> > triples to be handled when deducing the triple instead of when adding
> > available
> > features.
> >
> > Modified:
> >     libcxx/trunk/test/lit.cfg
> >
> > Modified: libcxx/trunk/test/lit.cfg
> > URL:
> >
> http://llvm.org/viewvc/llvm-project/libcxx/trunk/test/lit.cfg?rev=220724&r1=220723&r2=220724&view=diff
> >
> ==============================================================================
> > --- libcxx/trunk/test/lit.cfg (original)
> > +++ libcxx/trunk/test/lit.cfg Mon Oct 27 17:14:25 2014
> > @@ -349,13 +349,8 @@ class Configuration(object):
> >          # markers for tests that are known to fail with versions of
> >          libc++ as
> >          # were shipped with a particular triple.
> >          if self.use_system_lib:
> > -            # Drop sub-major version components from the triple,
> > because the
> > -            # current XFAIL handling expects exact matches for
> > feature checks.
> > -            sanitized_triple = re.sub(
> > -                r"([^-]+)-([^-]+)-([^-.]+).*", r"\1-\2-\3",
> > -                self.config.target_triple)
> >              self.config.available_features.add(
> > -                'with_system_lib=%s' % sanitized_triple)
> > +                'with_system_lib=%s' % self.config.target_triple)
> >
> >          if 'libcpp-has-no-threads' in
> >          self.config.available_features:
> >              self.compile_flags += ['-D_LIBCPP_HAS_NO_THREADS']
> > @@ -456,8 +451,22 @@ class Configuration(object):
> >          # If no target triple was given, try to infer it from the
> >          compiler
> >          # under test.
> >          if not self.config.target_triple:
> > -            self.config.target_triple = lit.util.capture(
> > +            target_triple = lit.util.capture(
> >                  [self.cxx, '-dumpmachine']).strip()
> > +            # Drop sub-major version components from the triple,
> > because the
> > +            # current XFAIL handling expects exact matches for
> > feature checks.
> > +            # Example: x86_64-unknown-linux-gnu ->
> > x86_64-unknown-linux-gnu.
>
> I don't understand this example, both look the same to me.
>
>  -Hal
>
> > +            # The 5th group handles triples greater than 3 parts
> > +            # (ex x86_64-pc-linux-gnu).
> > +            target_triple =
> > re.sub(r'([^-]+)-([^-]+)-([^.]+)([^-]*)(.*)',
> > +                                   r'\1-\2-\3\5', target_triple)
> > +            # linux-gnu is needed in the triple to properly identify
> > linuxes
> > +            # that use GLIBC. Handle redhat and opensuse triples as
> > special
> > +            # cases and append the missing `-gnu` portion.
> > +            if target_triple.endswith('redhat-linux') or \
> > +               target_triple.endswith('suse-linux'):
> > +                target_triple += '-gnu'
> > +            self.config.target_triple = target_triple
> >              self.lit_config.note(
> >                  "inferred target_triple as: %r" %
> >                  self.config.target_triple)
> >
> >
> >
> > _______________________________________________
> > cfe-commits mailing list
> > cfe-commits at cs.uiuc.edu
> > http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits
> >
>
> --
> Hal Finkel
> Assistant Computational Scientist
> Leadership Computing Facility
> Argonne National Laboratory
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/cfe-commits/attachments/20141028/de27f683/attachment.html>