[rfa] Change line to long

Andrew Cagney ac131313@cygnus.com
Thu Mar 1 10:08:00 GMT 2001


Andrew Cagney wrote:
> 
> Hello,
> 
> (more sparc64 -Werror fallout)
> 
> This was triggered by some realy nasty code in os9kread.c.  It was
> casting a pointer down to an integer (sizeof(long) == sizeof(void*) ==
> 32 and sizeof(int) == 16 right? :-).
> 
> The patch changes the type of symtab's line to long (from int) and then
> propogates that through the most obvious places.  The os9k read then
> casts that pointer across to slightly less non-portable long.  Given the
> os9k is probably on its last legs I can't see a good reason to try and
> clean up that chunk code.
> 
> The alternative quick hack is to cast (void*) -> (long) -> (int) and
> change the assert to check that sizeof(void*) <= sizeof(int) ...
> 
> Look ok? Comments?
> 
>         Andrew
> 
>   ------------------------------------------------------------------------
> Thu Feb  1 00:29:42 2001  Andrew Cagney  <cagney@redhat.com>
> 
>         * symtab.h (struct linetable_entry): Change ``line'' to a long.
>         * buildsym.h (record_line): Change ``line'' to a long.
>         * buildsym.c (record_line): Ditto
>         * os9kread.c: Include "gdb_assert.h".
>         (os9k_process_one_symbol): Cast the pointer ``name'' to a long.

Just FYI,

I'm dropping this patch.

Instead, I'm going to commit the attatched.  It does to things:

os9kread is changed so that when sizeof(int)<sizeof(char*) ``they
loose'' with an assert failure rather than a possibly corrupt symbol
table.

os9kread is taken of the list of files that must compile with -Werror.

I think the addition of the assertion check goes under the obvious fix
rule.

	Andrew
2001-03-01  Andrew Cagney  <ac131313@redhat.com>
  
	* Makefile.in (os9kread.o): Do not compile with WERROR_CFLAGS.
	* os9kread.c (os9k_process_one_symbol): Add assert to detect
	``loses if sizeof (char *) > sizeof (int)''.

Index: Makefile.in
===================================================================
RCS file: /cvs/src/src/gdb/Makefile.in,v
retrieving revision 1.65
diff -p -r1.65 Makefile.in
*** Makefile.in	2001/03/01 16:13:18	1.65
--- Makefile.in	2001/03/01 17:55:12
*************** mdebugread.o: mdebugread.c buildsym.h co
*** 1635,1644 ****
  
  mipsm3-nat.o: mipsm3-nat.c $(defs_h) $(inferior_h) $(regcache_h)
  
  os9kread.o: os9kread.c buildsym.h complaints.h $(bfd_h) $(defs_h) \
!         $(expression_h) gdb-stabs.h $(gdbcore_h) $(gdbtypes_h) language.h \
!         objfiles.h stabsread.h symfile.h $(symtab_h) \
! 	target.h gdb_string.h
  
  mem-break.o: mem-break.c $(defs_h) 
  
--- 1635,1647 ----
  
  mipsm3-nat.o: mipsm3-nat.c $(defs_h) $(inferior_h) $(regcache_h)
  
+ # os9kread assumes that sizeof(char*) <= sizeof(int).  This looses on
+ # 64 bit targets where often, sizeof(int)=4 but sizeof(char*)=9.
  os9kread.o: os9kread.c buildsym.h complaints.h $(bfd_h) $(defs_h) \
! 	        $(expression_h) gdb-stabs.h $(gdbcore_h) $(gdbtypes_h) \
! 		language.h objfiles.h stabsread.h symfile.h $(symtab_h) \
! 		target.h gdb_string.h
! 	$(CC) -c $(INTERNAL_WARN_CFLAGS) $(NO_WERROR_CFLAGS) $<
  
  mem-break.o: mem-break.c $(defs_h) 
  
Index: os9kread.c
===================================================================
RCS file: /cvs/src/src/gdb/os9kread.c,v
retrieving revision 1.6
diff -p -r1.6 os9kread.c
*** os9kread.c	2000/09/27 17:43:07	1.6
--- os9kread.c	2001/03/01 17:55:13
***************
*** 35,40 ****
--- 35,41 ----
  
  #include "defs.h"
  #include "gdb_string.h"
+ #include "gdb_assert.h"
  #include <stdio.h>
  
  #if defined(USG) || defined(__CYGNUSCLIB__)
*************** os9k_process_one_symbol (int type, int d
*** 1509,1514 ****
--- 1510,1516 ----
        /* Relocate for dynamic loading and for ELF acc fn-relative syms.  */
        valu += ANOFFSET (section_offsets, SECT_OFF_TEXT (objfile));
        /* FIXME: loses if sizeof (char *) > sizeof (int) */
+       gdb_assert (sizeof (name) <= sizeof (int));
        record_line (current_subfile, (int) name, valu);
        break;
  


More information about the Gdb-patches mailing list