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: solaris 2.9 sun assembler warnings


> > Jan Hubicka wrote:
> > 
> > >Hmm, so we are getting close.  Does these warnings stil reproduce on
> > >testcase I have?  Or can I get new one?
> > 
> > The one in errors.o is gone. But the one Kaveh sent is getting these 
> > errors here as well. Though bootstrap finished and I take it from stage2.
> > 
> > Attached the i and the s files
> > 
> > Andreas
> > 
> > 
> >  stage2/cc1 -fpreprocessed c-lang.i -quiet -dumpbase c-lang.c -mcpu=v7 
> > -auxbase-strip c-lang.o -g -O2 -W -Wall -Wwrite-strings 
> > -Wstrict-prototypes -Wmissing-prototypes -pedantic -Wno-long-long 
> > -Werror -version -fno-common -o c-lang.s
> > GNU C version 3.4 20030612 (experimental) (sparc-sun-solaris2.9)
> >         compiled by GNU C version 3.4 20030612 (experimental).
> > GGC heuristics: --param ggc-min-expand=91 --param ggc-min-heapsize=114688
> >  /usr/ccs/bin/as -V -Qy -s -xarch=v8 -o c-lang.o c-lang.s
> > /usr/ccs/bin/as: Sun WorkShop 6 update 2 Compiler Common 6.2 
> > Solaris_9_CBE 2001/04/02
> > /usr/ccs/bin/as: "c-lang.s", line 1184: warning: label in delay slot 
> > (follows CTI)
> > /usr/ccs/bin/as: "c-lang.s", line 1197: warning: label in delay slot 
> > (follows CTI)
> > /usr/ccs/bin/as: "c-lang.s", line 1210: warning: label in delay slot 
> > (follows CTI)
> > /usr/ccs/bin/as: "c-lang.s", line 1236: warning: label in delay slot 
> > (follows CTI)
> > /usr/ccs/bin/as: "c-lang.s", line 1249: warning: label in delay slot 
> > (follows CTI)
> > /usr/ccs/bin/as: "c-lang.s", line 1262: warning: label in delay slot 
> > (follows CTI)
> 
> 
> Hi,
> does the attached patch fix it?  It recuresc old code that prevented
> first line number information from being output.  I didn't understand
> why it is there and it seemed wrong.  Still I guess there should be
> better way to do that but I don't quite understand how the string
> prologues/epilogues are output.

This is actually cleaner fix.  Please let me know whether it bootstrap
and I will install it as obivous in case it does pass

Honza

Index: reorg.c
===================================================================
RCS file: /cvs/gcc/gcc/gcc/reorg.c,v
retrieving revision 1.84
diff -c -3 -p -r1.84 reorg.c
*** reorg.c	8 Apr 2003 20:16:54 -0000	1.84
--- reorg.c	12 Jun 2003 18:53:16 -0000
*************** emit_delay_sequence (insn, list, length)
*** 497,502 ****
--- 497,508 ----
        PREV_INSN (tem) = XVECEXP (seq, 0, i - 1);
        NEXT_INSN (XVECEXP (seq, 0, i - 1)) = tem;
  
+       /* SPARC assembler, for instance, emit warning when debug info is output
+          into the delay slot.  */
+       if (INSN_LOCATOR (tem) && !INSN_LOCATOR (seq_insn))
+ 	INSN_LOCATOR (seq_insn) = INSN_LOCATOR (tem);
+       INSN_LOCATOR (tem) = 0;
+ 
        for (note = REG_NOTES (tem); note; note = next)
  	{
  	  next = XEXP (note, 1);
*************** dbr_schedule (first, file)
*** 3841,3845 ****
--- 3847,3863 ----
      }
    free_resource_info ();
    free (uid_to_ruid);
+ #ifdef DELAY_SLOTS_FOR_EPILOGUE
+   /* SPARC assembler, for instance, emit warning when debug info is output
+      into the delay slot.  */
+   {
+     rtx link;
+ 
+     for (link = current_function_epilogue_delay_list;
+          link;
+          link = XEXP (link, 1))
+       INSN_LOCATOR (XEXP (link, 0)) = NULL;
+   }
+ #endif
  }
  #endif /* DELAY_SLOTS */


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