#14676: testcc.sh identifies clang as GCC
-----------------------------------------------------+----------------------
Reporter: leif | Owner:
GeorgSWeber
Type: defect | Status: new
Priority: minor | Milestone:
sage-5.11
Component: build | Resolution:
Keywords: compiler detection testcxx CC CXX | Work issues:
Report Upstream: N/A | Reviewers:
Authors: | Merged in:
Dependencies: | Stopgaps:
-----------------------------------------------------+----------------------
Description changed by leif:
Old description:
> The problem is that `clang` defines `__GNUC__` as well (as may Intel's
> `icc`).
>
> So we should use
> {{{
> #!c
> #if defined(__GNUC__) && !defined(__clang__) &&
> !defined(__INTEL_COMPILER)
> GCC
> ...
> }}}
> or rather something like
> {{{
> #!c
> #ifdef __GNUC__
> # ifdef __clang__
> CLANG
> # elif defined(__INTEL_COMPILER)
> ICC
> # else
> GCC
> # endif
> #define KNOWN_CC
> #endif
> }}}
> instead.
>
> Similar applies to `testcxx.sh`. (`clang++`, i.e., `clang` in C++ mode
> defines both `__GNUC__` and `__GNUG__` as well.)
>
> ----
>
> But another problem is that a couple of scripts (e.g. `spkg-install`s)
> testing for GCC might be just happy with `clang` as well (because it is
> to some extent "GCC-compatible"), so these would have to get changed
> accordingly (to accept both `"GCC"` and `"CLANG"`), while others may
> indeed not work with `clang`.
>
> Other scripts may not (yet) use `test{cc,cxx}.sh` but some similar logic;
> those probably have to get fixed as well, be it to (currently) just bail
> out earlier (e.g. if `CXX=clang`; in the long run, Sage will most
> probably build with `clang` as well).
New description:
The problem is that `clang` defines `__GNUC__` as well (as may Intel's
`icc`).
So we should use
{{{
#!c
#if defined(__GNUC__) && !defined(__clang__) && !defined(__INTEL_COMPILER)
GCC
...
}}}
or rather something like
{{{
#!c
#ifdef __GNUC__
# ifdef __clang__
CLANG
# elif defined(__INTEL_COMPILER)
ICC
# else
GCC
# endif
#define KNOWN_CC
#endif
}}}
instead.
Similar applies to `testcxx.sh`. (`clang++`, i.e., `clang` in C++ mode
defines both `__GNUC__` and `__GNUG__` as well.)
----
But another problem is that a couple of scripts (e.g. `spkg-install`s)
testing for GCC might be just happy with `clang` as well (because it is to
some extent "GCC-compatible"), so these would have to get changed
accordingly (to accept both `"GCC"` and `"CLANG"`), while others may
indeed not work with `clang`.
Other scripts may not (yet) use `test{cc,cxx}.sh` but some similar logic;
those probably have to get fixed as well, be it to (currently) just bail
out earlier (e.g. if `CXX=clang`; in the long run, Sage will most probably
build with `clang` as well).
----
Related: #12820, #12426
--
--
Ticket URL: <http://trac.sagemath.org/sage_trac/ticket/14676#comment:4>
Sage <http://www.sagemath.org>
Sage: Creating a Viable Open Source Alternative to Magma, Maple, Mathematica,
and MATLAB
--
You received this message because you are subscribed to the Google Groups
"sage-trac" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To post to this group, send email to [email protected].
Visit this group at http://groups.google.com/group/sage-trac?hl=en.
For more options, visit https://groups.google.com/groups/opt_out.
