[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[Qemu-devel] PPC floating point fixes
From: |
J. Mayer |
Subject: |
[Qemu-devel] PPC floating point fixes |
Date: |
Sat, 19 Feb 2005 13:20:16 +0100 |
I did some systematic tests on PPC FPU emulation.
Then, I found some bugs:
- single-precision operations are to be done in double precision, then
rounded to single precision.
- Make fctiw & fctiwz return the values a real G3 does
- Sign was false for fabs/fnabs whith +/- 0.0 operand
- fsel operands were interverted
- fres major opcode is 0x3B, not 0x3F
It also appears that most FPU flags are wrong, even when flag updates
are explicitely wanted (using '.' opcode) so more fixes will come soon.
--
J. Mayer <address@hidden>
Never organized
ppc_float_fixes.diff
Description: Text Data
[Prev in Thread] |
Current Thread |
[Next in Thread] |
- [Qemu-devel] PPC floating point fixes,
J. Mayer <=