André Braga wrote:
On 10/16/06, Tom Marn wrote:
cast test 1 - float: 85.745110 [0x42ab7d7f] -> integer: 85 [0x0055]
cast test 2 - float: 85.745110 [0x42ab7d7f] -> integer: 57005
[0xdead] <-- 85 is correct
Am I the only one who found this result EXTREMELY amusing? :D
Tha
integer: 57005 [0xdead] <--
85 is correct
Tom Marn
#include
int cast_ftoi()
{
int a = 0xBEEF;
int b = 0xDEAD;
int i1,i2;
float f1,f2;
__asm__ __volatile__ ("nop");
f1 = (100.0 * a / b);
i1 = f1;
__asm__ __volatile__ ("nop");
i2 = (100.0 * a / b);
_
Hi
Resending fixed patch, mirror fix in glue(stfi, MEMSUFFIX) function; bitwise typo:
&& instead of &.
Tom Marn
Patch which appends optional "stfiwx" PowerPC instruction into QEMU.
Mirror fix of patch: 2006-10-11 : bitwise typo && instead &
Tom Marn
--- targ
guess,
because the qemu micro instruction couldn't be found (gen_op_load_fpr_FT1_fpr32) something flushed the
fpr registers to the correct values.
Tom Marn
Patch which appends optional "stfiwx" PowerPC instruction into QEMU.
Tom Marn
--- qemu/target-ppc/translate.c.orig 2006-10-1
Tom Marn wrote:
Here is my modified translate.c (still broken code)
Re-sending previous malformed patch (I hope this one will be formated ok).
--- target-ppc/translate.c.orig 2006-09-29 09:29:28.0 +0200
+++ target-ppc/translate.c 2006-09-29 09:39:17.0 +0200
@@ -1715,21
Hi
I'm using QEMU to emulate PowerPC 603e for testing my cross compiled root filesystem on i686.
When executing binary which is compiled (without -msoft-float) to use full floating point instruction set,
kernel returns "Illegal instruction" at runtime. After core dump debugging my
binary I fou
