Bug#204805: Intel have reproduced this 'bug'

[Duraid Madina <duraid@octopus.com.au>]

Hi Masanori,

	#204805 is fixed now. #256770 is a new bug. But the problem may not be 
with the Intel compiler. Why are symbols multiply defined in libc.a?? It 
seems a mistake.

	The Intel compiler works on RedHat and SuSE.

	Also, you can download the compiler for free, from this website:

http://developer.intel.com/software/products/compilers/clin/noncom.htm

	(but you must register).

	Good luck!

	Duraid

GOTO Masanori wrote:
> At Tue, 12 Aug 2003 06:40:11 +1000,
> Duraid Madina wrote:
>
> > And they say that 'this problem is due to our compiler not supporting 
> > glibc 2.3 yet.'
> >
> > A fix is in the works, apparently.
>
>
> #204805 is reported in 2.3.2, but now we have 2.3.2.ds1.  Is this bug
> still alive?
>
> BTW, there's two bugs about intel C/C++: #204805, #256770.  Both bugs
> are reported by you.  I guess the newer version fix #204805, but it's
> still broken (#256770).  I wonder why debian glibc obstruct
> compilation with intel C/C++.  Is it designed only for RedHat's glibc?
>
> Commercial applications are difficult to check because we cannot get
> easily.  I know the speed difference between gcc and intel cc.  It's
> currently different significantly in the high performance numerical
> computation.  So it's valuable to investigate on Debian IA-64.
> Duraid, please check this bug more.
>
> Regards,
> -- gotom