PowerPC: *printf in non-float-tasks

[Kulpinski, Jay]

On Friday 28 March 2003 03:59 am, Sergei Organov wrote:
>
> Joel,
>
> AFAIR, there was a patch for newlib's vfprintf to don't use floating point
> when no floating-point arguments are passed. Hasn't it been incorporated
> into the newlib sources? If not then the OP may wish to apply the patch to
> his newlib.

I had this problem a _long_ time ago.  At the time, this patch worked.
It's a bit of an ugly hack, but it was the simplest fix at the time.
The code was pretty convoluted, so making a more local declaration 
wasn't practical.  I don't know how well it applies against a newer newlib.



diff -uNr 
/usr1/rtems/work-tools/original/newlib-1.8.2/newlib/libc/stdio/vfprint
f.c newlib-1.8.2/newlib/libc/stdio/vfprintf.c
--- /usr1/rtems/work-tools/original/newlib-1.8.2/newlib/libc/stdio/vfprintf.c
Mon Jul  5 13:51:05 1999
+++ newlib-1.8.2/newlib/libc/stdio/vfprintf.c   Mon Jan 31 13:37:02 2000
@@ -300,7 +300,21 @@
 #ifdef FLOATING_POINT
        char *decimal_point = localeconv()->decimal_point;
        char softsign;          /* temporary negative sign for floats */
+       /*
+        * Although it is natural to declare this double here, the
+        * declaration causes gcc to save FP registers even when not
+        * printing an FP number.  This results in surprising use
+        * of FP registers to print integers or strings on at least the
+        * PowerPC.  A more proper solution would be to move FP printing
+        * to another file, but this does seem to work.
+        */
+#if 0
        double _double;         /* double precision arguments %[eEfgG] */
+#else
+                               /* double precision arguments %[eEfgG] */
+       union { int i; double d; } _double_ = {0};
+#define _double (_double_.d)
+#endif
        int expt;               /* integer value of exponent */
        int expsize;            /* character count for expstr */
        int ndig;               /* actual number of digits returned by cvt */