[RFC] Remove exec_set_section_offsets

[RFC] Remove exec_set_section_offsets

[Thiago Jung Bauermann]

Hi,

I found out that exec_set_section_offsets is not referenced anywhere,
and the last ChangeLog entry related to it was in 1999, so this patch
removes it. Tested by compiling on Linux/ppc32. Is this ok?

Perhaps it would qualify as obvious, but I'm not used to removing code
from GDB so better ask about it...

2008-04-18  Thiago Jung Bauermann  <bauerman@br.ibm.com>

	* defs.h (exec_set_section_offsets): Remove prototype.
	* exec.c (exec_set_section_offsets): Remove function.

diff -r f3cbd26c0c0d gdb/defs.h
--- a/gdb/defs.h	Fri Apr 18 03:15:14 2008 -0700
+++ b/gdb/defs.h	Fri Apr 18 16:54:45 2008 -0700
@@ -602,10 +602,6 @@ extern void init_last_source_visited (vo
 extern void init_last_source_visited (void);
 
 /* From exec.c */
-
-extern void exec_set_section_offsets (bfd_signed_vma text_off,
-				      bfd_signed_vma data_off,
-				      bfd_signed_vma bss_off);
 
 /* Take over the 'find_mapped_memory' vector from exec.c. */
 extern void exec_set_find_memory_regions (int (*) (int (*) (CORE_ADDR, 
diff -r f3cbd26c0c0d gdb/exec.c
--- a/gdb/exec.c	Fri Apr 18 03:15:14 2008 -0700
+++ b/gdb/exec.c	Fri Apr 18 16:54:45 2008 -0700
@@ -598,42 +598,6 @@ exec_files_info (struct target_ops *t)
     }
 }
 
-/* msnyder 5/21/99:
-   exec_set_section_offsets sets the offsets of all the sections
-   in the exec objfile.  */
-
-void
-exec_set_section_offsets (bfd_signed_vma text_off, bfd_signed_vma data_off,
-			  bfd_signed_vma bss_off)
-{
-  struct section_table *sect;
-
-  for (sect = exec_ops.to_sections;
-       sect < exec_ops.to_sections_end;
-       sect++)
-    {
-      flagword flags;
-
-      flags = bfd_get_section_flags (exec_bfd, sect->the_bfd_section);
-
-      if (flags & SEC_CODE)
-	{
-	  sect->addr += text_off;
-	  sect->endaddr += text_off;
-	}
-      else if (flags & (SEC_DATA | SEC_LOAD))
-	{
-	  sect->addr += data_off;
-	  sect->endaddr += data_off;
-	}
-      else if (flags & SEC_ALLOC)
-	{
-	  sect->addr += bss_off;
-	  sect->endaddr += bss_off;
-	}
-    }
-}
-
 static void
 set_section_command (char *args, int from_tty)
 {

-- 
[]'s
Thiago Jung Bauermann
Software Engineer
IBM Linux Technology Center

Re: [RFC] Remove exec_set_section_offsets

[Joel Brobecker]

> Perhaps it would qualify as obvious, but I'm not used to removing code
> from GDB so better ask about it...

We just have to remember that insight builds on top of GDB. So, just
to be absolutely sure, I checked out the insight sources, and verified
that it wasn't using that function. So far, the functions I've seen
being used by insight started with a gdb_ prefix, but that might not
be a rule.

> 2008-04-18  Thiago Jung Bauermann  <bauerman@br.ibm.com>
> 
> 	* defs.h (exec_set_section_offsets): Remove prototype.
> 	* exec.c (exec_set_section_offsets): Remove function.

That's OK. Thank you!

-- 
Joel

Re: [RFC] Remove exec_set_section_offsets

[Thiago Jung Bauermann]

On Fri, 2008-04-18 at 18:58 -0700, Joel Brobecker wrote:
> > Perhaps it would qualify as obvious, but I'm not used to removing code
> > from GDB so better ask about it...
> 
> We just have to remember that insight builds on top of GDB. So, just
> to be absolutely sure, I checked out the insight sources, and verified
> that it wasn't using that function. So far, the functions I've seen
> being used by insight started with a gdb_ prefix, but that might not
> be a rule.

Nice to know, thanks for the tip. So if I observe that, this kind of
patch fits the obvious rule?

> > 2008-04-18  Thiago Jung Bauermann  <bauerman@br.ibm.com>
> > 
> > 	* defs.h (exec_set_section_offsets): Remove prototype.
> > 	* exec.c (exec_set_section_offsets): Remove function.
> 
> That's OK. Thank you!

Committed, thanks!
-- 
[]'s
Thiago Jung Bauermann
Software Engineer
IBM Linux Technology Center

Re: [RFC] Remove exec_set_section_offsets

[Joel Brobecker]

> Nice to know, thanks for the tip. So if I observe that, this kind of
> patch fits the obvious rule?

Yes :).

-- 
Joel