[Bug libstdc++/25913] Client's isnormal function is broken by cmath
--- Comment #7 from paolo at gcc dot gnu dot org 2007-11-26 18:25 --- Subject: Bug 25913 Author: paolo Date: Mon Nov 26 18:24:54 2007 New Revision: 130443 URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=130443 Log: 2007-11-26 Paolo Carlini [EMAIL PROTECTED] PR libstdc++/25913 * include/c_std/cmath (std::fpclassify, isfinite, isinf, isnan, isnormal, signbit, isgreater, isgreaterequal, isless, islessequal, islessgreater, isunordered): Guard with __enable_if and forward with __promote. * include/c_global/cmath: Likewise. * testsuite/26_numerics/headers/cmath/25913.cc: New. * include/c_std/cmath (__gnu_cxx::__capture_isfinite, __capture_isinf, __capture_isnan, __capture_isnormal, __capture_signbit, __capture_isgreater, __capture_isgreaterequal, __capture_isless, __capture_islessequal, __capture_islessgreater, __capture_isunordered): Remove. (std::isfinite, isinf, isnan, isnormal, signbit, isgreater, isgreaterequal, isless, islessequal, islessgreater, isunordered): Forward to the corresponding builtin. * include/c_global/cmath: Likewise. * include/c_global/cmath (std::atan2, pow): Guard with __enable_if. Added: trunk/libstdc++-v3/testsuite/26_numerics/headers/cmath/25913.cc Modified: trunk/libstdc++-v3/ChangeLog trunk/libstdc++-v3/include/c_global/cmath trunk/libstdc++-v3/include/c_std/cmath -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25913
[Bug libstdc++/25913] Client's isnormal function is broken by cmath
--- Comment #8 from pcarlini at suse dot de 2007-11-26 18:26 --- Fixed for 4.3.0. -- pcarlini at suse dot de changed: What|Removed |Added Status|ASSIGNED|RESOLVED Resolution||FIXED Target Milestone|--- |4.3.0 http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25913
[Bug libstdc++/25913] Client's isnormal function is broken by cmath
-- pcarlini at suse dot de changed: What|Removed |Added AssignedTo|unassigned at gcc dot gnu |pcarlini at suse dot de |dot org | Status|NEW |ASSIGNED http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25913
[Bug libstdc++/25913] Client's isnormal function is broken by cmath
--- Comment #6 from pinskia at gcc dot gnu dot org 2006-01-23 14:50 --- Confirmed. -- pinskia at gcc dot gnu dot org changed: What|Removed |Added Status|UNCONFIRMED |NEW Ever Confirmed|0 |1 Last reconfirmed|-00-00 00:00:00 |2006-01-23 14:50:10 date|| http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25913
[Bug libstdc++/25913] Client's isnormal function is broken by cmath
--- Comment #1 from gdr at cs dot tamu dot edu 2006-01-22 20:10 --- Subject: Re: New: Client's isnormal function is broken by cmath hhinnant at apple dot com [EMAIL PROTECTED] writes: | I believe the program below should compile: Does the problem exist if you configure with --disable-c99? -- Gaby -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25913
[Bug libstdc++/25913] Client's isnormal function is broken by cmath
--- Comment #2 from hhinnant at apple dot com 2006-01-22 20:25 --- (In reply to comment #1) Does the problem exist if you configure with --disable-c99? I haven't tried that configuration. The docs say that setting may change libstdc++'s ABI. So this is expected behavior for the default configuration? -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25913
[Bug libstdc++/25913] Client's isnormal function is broken by cmath
--- Comment #3 from gdr at cs dot tamu dot edu 2006-01-22 20:42 --- Subject: Re: Client's isnormal function is broken by cmath hhinnant at apple dot com [EMAIL PROTECTED] writes: | (In reply to comment #1) | | Does the problem exist if you configure with --disable-c99? | | I haven't tried that configuration. The docs say that setting may change | libstdc++'s ABI. Indeed. The question was to what extent the issue is completely fixable. | So this is expected behavior for the default configuration? if you keep ADL and template deduction, sort of :-) This is the error I get: /home/gdr/lib/gcc/i686-pc-linux-gnu/4.2.0/../../../../include/c++/4.2.0/cmath: In function 'int __gnu_cxx::__capture_isnormal(_Tp) [with _Tp = manager]': /home/gdr/lib/gcc/i686-pc-linux-gnu/4.2.0/../../../../include/c++/4.2.0/cmath:537: instantiated from 'int std::isnormal(_Tp) [with _Tp = manager]' t.C:22: instantiated from here /home/gdr/lib/gcc/i686-pc-linux-gnu/4.2.0/../../../../include/c++/4.2.0/cmath:466: error: cannot convert 'manager' to 'long double' for argument '1' to 'int __fpclassifyl(long double)' /home/gdr/lib/gcc/i686-pc-linux-gnu/4.2.0/../../../../include/c++/4.2.0/cmath:466: error: cannot convert 'manager' to 'double' for argument '1' to 'int __fpclassify(double)' /home/gdr/lib/gcc/i686-pc-linux-gnu/4.2.0/../../../../include/c++/4.2.0/cmath:466: error: cannot convert 'manager' to 'float' for argument '1' to 'int __fpclassifyf(float)' which shows that std::isnormal is being used -- because it provides a better match. Two issues here: (1) the std::isnormal template is way too liberal -- probably needs an enable-if decoration. (2) std::isnormal is there only if C99 is enabled with a compbination of C99 macro dynamic. See the contrived logic in std_cmath.h. (3) even when isnormal is enable-if hacked, you still potentially run into the same problem. -- Gaby -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25913
[Bug libstdc++/25913] Client's isnormal function is broken by cmath
--- Comment #4 from hhinnant at apple dot com 2006-01-22 20:49 --- (In reply to comment #3) (3) even when isnormal is enable-if hacked, you still potentially run into the same problem. For example? -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25913
[Bug libstdc++/25913] Client's isnormal function is broken by cmath
--- Comment #5 from gdr at cs dot tamu dot edu 2006-01-22 21:17 --- Subject: Re: Client's isnormal function is broken by cmath hhinnant at apple dot com [EMAIL PROTECTED] writes: | (In reply to comment #3) | |(3) even when isnormal is enable-if hacked, you still potentially |run into the same problem. | | For example? Plain #include math.h -- Gaby -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25913
