GCC 2.95 and floating point precision

[Toon Moene]

Dara Hazeghi wrote:

> The author is working on fixing it, but this really isn't a great way to
> promote gcc2.95. I know all you people put an incredible amount of work into
> gcc2.95 and for most things it excels, but unless I'm much mistaken it could
> have been left in with a warning that it might generate bad code.
> -fno-strict-aliasing was for the purpose of backward compatibility and it is
> essential. Something of the same style for clobber would have really
> simplified issues.

See the explanation at: http://egcs.cygnus.com/faq.html#asmclobber

It's not possible to have a `backwards compatibility switch' here,
because the asm's are really wrong.  The FAQ entry even explains that
the confusion was partly due to the fact that the documentation on asms
was ambiguous.  These constructs simply do not have a well-defined
meaning (as I wrote before) and therefore, strictly speaking, cannot be
compiled "correctly".

I'm sorry, I do not know enough of this side of gcc to explain this
better.

> On another aside, are there any changes with the configure scripts in gcc.
> When I do ./configure --target=alpha-linux I get the usage for the configure
> script. How is one supposed to configure to build a cross compiler in this
> day and age.

See:  http://egcs.cygnus.com/faq.html#cross

Hope this helps,

-- 
Toon Moene (toon@moene.indiv.nluug.nl)
Saturnushof 14, 3738 XG  Maartensdijk, The Netherlands
Phone: +31 346 214290; Fax: +31 346 214286
GNU Fortran: http://world.std.com/~burley/g77.html