This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: new real.c implementation
- From: Paul Koning <pkoning at equallogic dot com>
- To: rth at redhat dot com
- Cc: gcc-patches at gcc dot gnu dot org
- Date: Sat, 14 Sep 2002 17:10:44 -0400
- Subject: Re: new real.c implementation
- References: <20020914001909.GA32475@redhat.com>
Richard> For VAX, I370, and C4X, I compared the output of
Richard> float f[] = { 0.0f, -0.0f, 1.0f, -1.0f, 2.0f, 3.0f,
Richard> 1.0f/3.0f, __FLT_MIN__, __FLT_MAX__, __FLT_DENORM_MIN__ };
Richard> double d[] = { 0.0, -0.0, 1.0, -1.0, 2.0, 3.0, 1.0/3.0,
Richard> __DBL_MIN__, __DBL_MAX__, __DBL_DENORM_MIN__ };
Richard> long double l[] = { 0.0L, -0.0L, 1.0L, -1.0L, 2.0L, 3.0L,
Richard> 1.0L/3.0L, __LDBL_MIN__, __LDBL_MAX__, __LDBL_DENORM_MIN__
Richard> };
I'm not sure what xxx_DENORM_MIN means for VAX float format, which
doesn't have denormals.
You seem to have dropped the VAX_HALFWORD_ORDER stuff. I put that in
for the pdp11 backend (the current one doesn't use it, since it sends
float values to the assembler for processing, but I intended to use
the target format output support in real.c, and then this flag is
needed). Not a big deal for now... (If you do leave it out then you
might want to remove it from tm.texi.)
> #ifndef TARGET_G_FORAMT
...................^^^^^^ typo...
> #define TARGET_G_FORMAT 0
> #endif
paul