For the alpha I see many compiler flags that control how floating point errors
are handled. Examples are
-mfp-trap-mode=TRAP MODE
-mfp-rounding-mode=ROUNDING MODE
-mtrap-precision=TRAP PRECISION
How are these functions implemented for the i386? I would like to be able to
ignore or trap to an exception handler things like floating underflow, division
by zero etc.
The alpha info-page says that n "only traps that are enabled are the ones that
cannot be disabled in software". How does one enable or disable the traps in
software?
Sorry, for I know this must be a FAQ.
don
--
-------------------------------------------------------------------
| Don Fisher [EMAIL PROTECTED] |
| MMT Observatory VOICE: (520)621-7647 |
| University of Arizona FAX: (520)670-5740 |
| Tucson, AZ 85721 |
-------------------------------------------------------------------
