This is the mail archive of the mailing list for the GCC project.

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]

Re: PATCH: XFAIL c99-float-1.c on Solaris

On Mon, 14 May 2001, Mark Mitchell wrote:

> Solaris doesn't define FLT_EVAL_METHOD or DECIMAL_DIG.  This test
> failed with GCC 2.95.x as well, so this is not a regression.
> Tested on sparc-sun-solaris2.8, applied on the branch.

I think some documentation of the problem of systems which don't use GCC's
float.h is needed.  I've installed the following in both copies of

For a proper fix, such systems need either to get their own versions of
float.h in GCC or to have their system float.h fixincluded.

Index: c99status.html
RCS file: /cvs/gcc/wwwdocs/htdocs/c99status.html,v
retrieving revision 1.24
diff -u -r1.24 c99status.html
--- c99status.html	2001/02/17 10:22:10	1.24
+++ c99status.html	2001/05/14 22:14:30
@@ -309,7 +309,9 @@
 <code>FLT_EVAL_METHOD</code> and <code>DECIMAL_DIG</code>) are
 included in most versions of <code>&lt;float.h&gt;</code>, but not yet
 <code>float-c4x.h</code>, <code>float-m68k.h</code> or
+<code>float-vax.h</code>.  Also, some targets such as Solaris use the
+system <code>&lt;float.h&gt;</code> but do not fix it to add these
 <li>Compiler support is needed for <code>math_errhandling</code>; see
 messages <a

Joseph S. Myers

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]