yOn Tue, 13 Nov 2001, Matthias Klose wrote:
> Adam Heath writes:
> > Package: libgcj2
> > Severity: serious
> >
> > Library packages should not include binaries. Please see policy section
> > 11.3.
> >
> > ...
> >
> >If your package has some run-time support programs which use the shared
>
Adam Heath writes:
> Package: libgcj2
> Severity: serious
>
> Library packages should not include binaries. Please see policy section 11.3.
>
> ...
>
>If your package has some run-time support programs which use the shared
> library you must not put them in the shared library package. If y
James Troup writes:
> Package: gcc-3.0
> Version: 1:3.0.2-3
>
> All the -3.0 packages don't depend, recommend or even suggest the
> packages making it possible (especially on hppa where -3.0 is
> the default compiler) to install just gcc-3.0 and not have a gcc
> symlink which is probably not a go
Phil Edwards writes:
>
> > All the -3.0 packages don't depend, recommend or even suggest the
> > packages making it possible (especially on hppa where -3.0 is
> > the default compiler) to install just gcc-3.0 and not have a gcc
> > symlink which is probably not a good thing?
>
> The 'gcc' packag
Package: g++
Version: 2:2.95.4-8
Severity: normal
In a member function such as:
int Partition::in_which_part()
{
int i;
}
A return statement with type of int is required. However, the compiler
fails to give any error/warning when the corresponding return statement
is omitted, when invoke
Processing commands for [EMAIL PROTECTED]:
> retitle 119384 provide java-virtual-machine
Bug#119384: New on the list - java on debian?
Changed Bug title.
> thanks
Stopping processing here.
Please contact me if you need assistance.
Darren Benham
(administrator, Debian Bugs database)
retitle 119384 provide java-virtual-machine
thanks
> Please have libgcj2 provide java-virtual-machine.
... only on the architectures where gij builds, of course.
--
Robbe
On Tue, Nov 13, 2001 at 08:44:48AM +0100, Matthias Klose wrote:
> Phil Edwards writes:
> > The 'gcc' package is hardcoded to gcc-2.95 right now. You can have all
> > three packages installed, and the symlink doesn't change:
Just to clarify: I wasn't complaining when I wrote that, merely making
i
Thanks for your bug report. This is not a bug in the compiler, but it
is documented behaviour. The code you write is perfectly correct; a
C++ program is not required to have a return statement in every
function. In fact, in some cases, it is completely reasonable not to
have any, e.g. when the las
Processing commands for [EMAIL PROTECTED]:
> severity 119440 wishlist
Bug#119440: g++: Compiler does not give any errors when a function fails to
return required value
Severity set to `wishlist'.
> thanks
Stopping processing here.
Please contact me if you need assistance.
Darren Benham
(admini
On Tue, Nov 13, 2001 at 17:54:14 +0200, Eray Ozkural (exa) wrote:
> In the current case, the compiler does not even give a warning message
> by default. Giving an error message might break valid C++ code (which
> I'm not sure of) but it seems that when you specify -Wall
-Wreturn-type suffices.
>
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Are there any warnings that are enabled by default?
Thanks,
On Tuesday 13 November 2001 23:40, Martin v. Loewis wrote:
> *If* control flow ever falls through the end of a function that needs
> to return a value, the behaviour is undefined. GCC offers
Your message dated Wed, 14 Nov 2001 00:28:30 +0200
with message-id <[EMAIL PROTECTED]>
and subject line closing this one
has caused the attached Bug report to be marked as done.
This means that you claim that the problem has been dealt with.
If this is not the case it is now your responsibility to
> Are there any warnings that are enabled by default?
Yes. For example, compiling
void x(){
short y = 0x;
}
gives the warning
warning: overflow in implicit constant conversion
However, this is different from the -Wreturn-type warning: The
compiler has reliable algorithm that gu
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
On Wednesday 14 November 2001 01:28, Martin v. Loewis wrote:
>
> A similar algorithm does not exist for reliably determining that a
> function falls off its end.
What if there is no return statement in a function with a non-void return
type?
Thanks,
Phil Edwards writes:
> On Tue, Nov 13, 2001 at 08:44:48AM +0100, Matthias Klose wrote:
> > > Perhaps gcc-3.0's postinst can do something like "if package 'gcc' is
> > > installed, then update the symlink"? I'm guessing out my butt here.
> >
> > we removed the gcc alternatives a year ago, which ac
> > A similar algorithm does not exist for reliably determining that a
> > function falls off its end.
>
> What if there is no return statement in a function with a non-void return
> type?
Like this?
int foo(){
exit(3);
}
Has no return statement, but still doesn't invoke undefined behavior.
17 matches
Mail list logo
