[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Negative double weirdness
From: |
Marcus G. Daniels |
Subject: |
Re: Negative double weirdness |
Date: |
30 Jun 1998 22:39:48 -0700 |
>>>>> "SE" == Steve Emsley <address@hidden> writes:
SE> Basically, as long as I printf("%e\n",transmission) I get the
SE> correct answer.
It almost sounds like an optimizer bug. Have you tried building Swarm
with no optimization?
SE> I'm using RedHat5.0 but have upgraded to gcc2.8.0.
SE> What is a toolchain?
The toolchain consists of all the things that make it possible to
convert an Objective C program into an executable: as, cpp, cc1, ld,
the C library, Objective C support library, etc.
Is the gcc 2.8 you are using from Redhat or did you build it yourself?
Maybe there is some incompatibility there. If there is a egcs or gcc 2.8.1
RPM, that might be worth a try too.
Have you tried building Swarm with --disable-shared? One
significant difference between 1.1 and 1.2 is that on GNU/Linux you'll
get shared libraries for Swarm things by default.
Basically, I can't think of an explanation, although I suspect a code
generation or compiler configuration problem w.r.t. the C library dynamic
linker.
==================================
Swarm-Support is for discussion of the technical details of the day
to day usage of Swarm. For list administration needs (esp.
[un]subscribing), please send a message to <address@hidden>
with "help" in the body of the message.
==================================
- Re: Negative double weirdness,
Marcus G. Daniels <=