On 6/22/06, Carl Lowenstein <[EMAIL PROTECTED]> wrote:
It's pretty well known that floating point on the Alpha is not quite
compliant with IEEE754 with regard to exception handling. I have
learned that there is a .configure option in with the Python sources:
"--with-fpectl" meaning floating-point exception control.
"./configure --with-fpectl" doesn't seem to help.
I have learned more about Python than I thought I would have to. If I
look at the test program that creates FP exceptions "test_long.py" and
work on it for a while, I can
do either of two things with it:
1) make it not produce FPE in which case the test gives the message
"TestFailed: expected OverflowError from float(huge)" but continues to
run.
2) make it produce FPE and then the whole test sequence bombs out
rather than handling the exception gracefully.
If I comment out all the FPE testing, the whole of "make test" runs to
completion with a few diagnostic messages but nothing major. So I
think I have a working version of Python 2.4.3 so long as I don't do
anything exceptional with floating point.
Interesting that I can't run the Python2.4.3 test set against the
existing Python2.1.3 because there have been some new features added,
like "//" truncating division, so the old program rejects the new test
set because it contains illegal operations.
carl
--
carl lowenstein marine physical lab u.c. san diego
[EMAIL PROTECTED]
--
[email protected]
http://www.kernel-panic.org/cgi-bin/mailman/listinfo/kplug-list
