avr-gcc-list
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [avr-gcc-list] Re: Removal of unused functions


From: Joerg Wunsch
Subject: Re: [avr-gcc-list] Re: Removal of unused functions
Date: Mon, 21 Aug 2006 07:19:47 +0200 (MET DST)

Björn Haase <address@hidden> wrote:

> Joerg's statement used to be correct but no longer is. In fact there
> was a problem for older versions of binutils. The problem that you
> have encountered is fixed in the most recent cvs version of the
> binutils. If you don't know how to use cvs, you might use some
> recent snapshot.

I think avr-libc would at least have to add __attribute__((used)) to
the interrupt vector declarators, wouldn't it?  Without that, IMHO the
linker would assume the vectors to be dead code I'm afraid.

The question is whether any other of the library functions would need
this declaration as well (in particular stuff from gcrt1.S).  So far,
nobody in avr-libc land ever seriously took a stab on that -- that's
(together with the earlier problems you mention as being fixed) what
caused my statement of this feature being effectively not maintained
for the AVR.

We could continue discussing the avr-libc related part of that on the
libc mailing list if you want, so the remaining issues could be
cleared up before the next release (which might be quite soon), and
only then I'd like to officially announce the feature in the
accompanying documentation.  For that latter part (documentation
update), a volunteer would be highly welcome.  It's basically the FAQ
entry Anatoly mentioned, but as we also document some GCC and binutils
options that are interesting for the AVR, that needs to be updated,
too.

-- 
cheers, J"org               .-.-.   --... ...--   -.. .  DL8DTL

http://www.sax.de/~joerg/                        NIC: JW11-RIPE
Never trust an operating system you don't have sources for. ;-)





reply via email to

[Prev in Thread] Current Thread [Next in Thread]