This is the mail archive of the gcc-patches@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: Patch to convert ASM_GLOBALIZE_LABEL to a target hook


 > From: David Edelsohn <dje@watson.ibm.com>
 > 
 > 	Congratulations!  You broke AIX bootstrap.

David - I understand its frustrating, Solaris2 bootstraps regularly
have been broken for a week or more over this Summer.  But note, I
didn't break AIX intentionally.  So there's no need for sarcasm.


 > 	I thought that your patch was a one for one translation.  How is
 > the before and after equivalent in xcoff.h?
 > 
 > -#define ASM_GLOBALIZE_LABEL(FILE,NAME) \
 > -  do { fputs ("\t.globl ", FILE);      \
 > -       RS6000_OUTPUT_BASENAME (FILE, NAME); putc ('\n', FILE);} while (0)
 > +/* Globalizing directive for a label.  */
 > +#define GLOBAL_ASM_OP "\t.globl "
 > 
 > Notice the RS6000_OUTPUT_BASENAME!  Just because you see ".globl" doesn't
 > mean the rest of the code is equivalent.
 > David


When I went through the code, RS6000_OUTPUT_BASENAME just does
 > assemble_name ((FILE), (*targetm.strip_name_encoding) (NAME))

and I thought assemble_name itself did a strip_name_encoding.  So I
thought RS6000_OUTPUT_BASENAME was redundant.  Is that not the case?
Can you please explain to me what went wrong so I don't repeat my
mistake?

		Thanks,
		--Kaveh
--
Kaveh R. Ghazi			Director of Systems Architecture
ghazi@caip.rutgers.edu		Qwest Solutions


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