This is the mail archive of the
gcc-bugs@gcc.gnu.org
mailing list for the GCC project.
Re: bug report: Internal compiler error 378.
- To: "Martin v. Loewis" <martin at mira dot isdn dot cs dot tu-berlin dot de>
- Subject: Re: bug report: Internal compiler error 378.
- From: Ilan Finci <ifinci at mobileye dot com>
- Date: Fri, 24 Sep 1999 06:21:04 +0000
- CC: gcc-bugs at gcc dot gnu dot org
- Organization: MobilEye Vision Technologies LTD.
- References: <37EA1E7A.9A2F9E67@mobileye.com> <199909231526.RAA06298@mira.isdn.cs.tu-berlin.de>
"Martin v. Loewis" wrote:
>
> > I'm trying to compile a code I have (that compiles fine egcs-1.1.2)
> > here is the output of the compilation:
>
> Thanks for your bug report. gcc-2.95.1 still breaks; the mainline
> compiler says
>
By "mainline compiler", do you mean the current development version of
the compiler ?
> vec.h:76: warning: friend declaration `class MEmath::Vec<3> MEmath::operator /(const MEmath::Vec<3> &, const MEmath::Vec<3> &)'
> vec.h:76: warning: declares a non-template function
> vec.h:76: warning: (if this is not what you intended, make sure
> vec.h:76: warning: the function template has already been declared,
> vec.h:76: warning: and add <> after the function name here)
> vec.h:76: warning: -Wno-non-template-friend disables this warning.
> vec.cpp: In function `class MEmath::Vec<3> operator /(const MEmath::Vec<3> &, const MEmath::Vec<3> &)':
> vec.h:84: `double MEmath::Vec<3>::_el[3]' is private
> vec.cpp:18: within this context
> [message repeated a number of times]
>
> so it appears that the bug has been fixed.
This seems strange, since the operator/ is declered friend in the vec.h
file, so it should have access to the private variables of Vec<3>.
>
> Regards,
> Martin
Thanks, Ilan
--
MobilEye Vision Technologies LTD
24 Mishol Hadkalim st. Jerusalem, 97278, Israel
Tel: 972-2-5866989 Fax: 972-2-5867720 Personal Fax: 1-801-912-3251
Email: mailto:ifinci@mobileye.com http://www.mobileye.com