This is the mail archive of the gcc@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: 3.3 problem of -fzero-initialized-in-bss w/-ffreestanding


 > From: Eric Botcazou <ebotcazou@libertysurf.fr>
 > 
 > > How does it break Emacs - and what's the workaround?
 > 
 > Emacs uses a PCH-like mechanism that makes assumptions about the layout of 
 > sections in the image of the executable in memory. This doesn't work anymore 
 > with GCC 3.3 on at least Solaris 7, 8 and 9.
 > 
 > The workaround is to pass CFLAGS="-g -O2 -fno-zero-initialized-in-bss" to the 
 > configure script (tested on Solaris 8 and 9).
 > 
 > See http://gcc.gnu.org/PR11386 for the whole story and a pointer to the Emacs 
 > patch.

Thanks for the pointer.  I read through the thread and the conclusion
apparently reached by the emacs folks is that this was a valid
optimization and that the problem was an emacs issue.  It only
affected the undumped emacs, and they fixed it.

http://mail.gnu.org/archive/html/emacs-devel/2003-07/msg00207.html
http://mail.gnu.org/archive/html/emacs-devel/2003-07/msg00257.html

OTOH, I'm not unbiased WRT this optmization, and I can certainly see
that it is important for GCC to continue to be able to compile older
releases of emacs.

So I'll leave it up to others to decide what to do and I'll be ok with
either outcome.

		--Kaveh
--
Kaveh R. Ghazi			ghazi@caip.rutgers.edu


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]