[PATCH] disassembly fixes

[PATCH] disassembly fixes

[Elena Zannoni]

A few fixes for the new disassembler. I'll commit in a day or so.

2003-04-04  Elena Zannoni  <ezannoni@redhat.com>

	* disasm.c (dump_insns):  Move variables inside loop, or they will
        be freed more than once, causing wild memory corruptions.
	(gdb_disassembly): Look for the substring "-thread",
        instead of "-threads" in the target name, to make sure to find
        the 'multi-thread' target.  Also, make sure we do the right thing 
        with the "core" target.

Index: disasm.c
===================================================================
RCS file: /cvs/uberbaum/gdb/disasm.c,v
retrieving revision 1.5
diff -u -p -r1.5 disasm.c
--- disasm.c	17 Mar 2003 19:57:12 -0000	1.5
+++ disasm.c	4 Apr 2003 22:56:46 -0000
@@ -93,14 +93,15 @@ dump_insns (struct ui_out *uiout, disass
 
   /* parts of the symbolic representation of the address */
   int unmapped;
-  char *filename = NULL;
-  char *name = NULL;
   int offset;
   int line;
   struct cleanup *ui_out_chain;
 
   for (pc = low; pc < high;)
     {
+      char *filename = NULL;
+      char *name = NULL;
+
       QUIT;
       if (how_many >= 0)
 	{
@@ -358,7 +359,8 @@ gdb_disassembly (struct ui_out *uiout,
       if (strcmp (target_shortname, "child") == 0
 	  || strcmp (target_shortname, "procfs") == 0
 	  || strcmp (target_shortname, "vxprocess") == 0
-	  || strstr (target_shortname, "-threads") != NULL)
+          || strcmp (target_shortname, "core") == 0
+	  || strstr (target_shortname, "-thread") != NULL)
 	gdb_disassemble_from_exec = 0;	/* It's a child process, read inferior mem */
       else
 	gdb_disassemble_from_exec = 1;	/* It's remote, read the exec file */

Re: [PATCH] disassembly fixes

[Daniel Jacobowitz]

On Fri, Apr 04, 2003 at 06:04:48PM -0500, Elena Zannoni wrote:
> 
> A few fixes for the new disassembler. I'll commit in a day or so.
> 
> 2003-04-04  Elena Zannoni  <ezannoni@redhat.com>
> 
> 	* disasm.c (dump_insns):  Move variables inside loop, or they will
>         be freed more than once, causing wild memory corruptions.
> 	(gdb_disassembly): Look for the substring "-thread",
>         instead of "-threads" in the target name, to make sure to find
>         the 'multi-thread' target.  Also, make sure we do the right thing 
>         with the "core" target.

While you're in there, any comments on:
  http://sources.redhat.com/ml/gdb/2003-02/msg00153.html
?

-- 
Daniel Jacobowitz
MontaVista Software                         Debian GNU/Linux Developer

Re: [PATCH] disassembly fixes

[Elena Zannoni]

Daniel Jacobowitz writes:
 > On Fri, Apr 04, 2003 at 06:04:48PM -0500, Elena Zannoni wrote:
 > > 
 > > A few fixes for the new disassembler. I'll commit in a day or so.
 > > 
 > > 2003-04-04  Elena Zannoni  <ezannoni@redhat.com>
 > > 
 > > 	* disasm.c (dump_insns):  Move variables inside loop, or they will
 > >         be freed more than once, causing wild memory corruptions.
 > > 	(gdb_disassembly): Look for the substring "-thread",
 > >         instead of "-threads" in the target name, to make sure to find
 > >         the 'multi-thread' target.  Also, make sure we do the right thing 
 > >         with the "core" target.
 > 
 > While you're in there, any comments on:
 >   http://sources.redhat.com/ml/gdb/2003-02/msg00153.html
 > ?
 > 

Hmm, yes, I think that logic is a bit screwed up (it came from gdbtk,
and maybe there it was more appropriate), I have no objections if you
want to revisit the heuristics.  n hte meantime I would apply these
changes anyway because disassembly with core files and threaded
programs just doesn't work at all.

elena


 > -- 
 > Daniel Jacobowitz
 > MontaVista Software                         Debian GNU/Linux Developer

Re: [PATCH] disassembly fixes

[Daniel Jacobowitz]

On Mon, Apr 07, 2003 at 09:31:08AM -0400, Elena Zannoni wrote:
> Daniel Jacobowitz writes:
>  > On Fri, Apr 04, 2003 at 06:04:48PM -0500, Elena Zannoni wrote:
>  > > 
>  > > A few fixes for the new disassembler. I'll commit in a day or so.
>  > > 
>  > > 2003-04-04  Elena Zannoni  <ezannoni@redhat.com>
>  > > 
>  > > 	* disasm.c (dump_insns):  Move variables inside loop, or they will
>  > >         be freed more than once, causing wild memory corruptions.
>  > > 	(gdb_disassembly): Look for the substring "-thread",
>  > >         instead of "-threads" in the target name, to make sure to find
>  > >         the 'multi-thread' target.  Also, make sure we do the right thing 
>  > >         with the "core" target.
>  > 
>  > While you're in there, any comments on:
>  >   http://sources.redhat.com/ml/gdb/2003-02/msg00153.html
>  > ?
>  > 
> 
> Hmm, yes, I think that logic is a bit screwed up (it came from gdbtk,
> and maybe there it was more appropriate), I have no objections if you
> want to revisit the heuristics.  n hte meantime I would apply these
> changes anyway because disassembly with core files and threaded
> programs just doesn't work at all.

Sounds good to me!

-- 
Daniel Jacobowitz
MontaVista Software                         Debian GNU/Linux Developer

Re: [PATCH] disassembly fixes

[Elena Zannoni]

Daniel Jacobowitz writes:
 > On Mon, Apr 07, 2003 at 09:31:08AM -0400, Elena Zannoni wrote:
 > > Daniel Jacobowitz writes:
 > >  > On Fri, Apr 04, 2003 at 06:04:48PM -0500, Elena Zannoni wrote:
 > >  > > 
 > >  > > A few fixes for the new disassembler. I'll commit in a day or so.
 > >  > > 
 > >  > > 2003-04-04  Elena Zannoni  <ezannoni@redhat.com>
 > >  > > 
 > >  > > 	* disasm.c (dump_insns):  Move variables inside loop, or they will
 > >  > >         be freed more than once, causing wild memory corruptions.
 > >  > > 	(gdb_disassembly): Look for the substring "-thread",
 > >  > >         instead of "-threads" in the target name, to make sure to find
 > >  > >         the 'multi-thread' target.  Also, make sure we do the right thing 
 > >  > >         with the "core" target.
 > >  > 
 > >  > While you're in there, any comments on:
 > >  >   http://sources.redhat.com/ml/gdb/2003-02/msg00153.html
 > >  > ?
 > >  > 
 > > 
 > > Hmm, yes, I think that logic is a bit screwed up (it came from gdbtk,
 > > and maybe there it was more appropriate), I have no objections if you
 > > want to revisit the heuristics.  n hte meantime I would apply these
 > > changes anyway because disassembly with core files and threaded
 > > programs just doesn't work at all.
 > 
 > Sounds good to me!

ok, applied.

elena


 > 
 > -- 
 > Daniel Jacobowitz
 > MontaVista Software                         Debian GNU/Linux Developer