[patch] .gdb_index: Ignore higher versions

[patch] .gdb_index: Ignore higher versions

[Jan Kratochvil]

Hi,

currently introducing new .gdb_index version (latest is v3, so >=4) breaks all
existing GDBs as they try too interpret the index format the old way.

I have a reproducer here with some other patchset but I find it obvious.

Sending for a comment as we could also introduce some versioning features for
compatible/incompatible extensions.  I do not think it is worth it.


Thanks,
Jan


gdb/
2010-11-21  Jan Kratochvil  <jan.kratochvil@redhat.com>

	* dwarf2read.c (dwarf2_read_index): Ignore higher .gdb_index versions.

--- a/gdb/dwarf2read.c
+++ b/gdb/dwarf2read.c
@@ -1941,6 +1941,10 @@ dwarf2_read_index (struct objfile *objfile)
      it seems better to just ignore such indices.  */
   if (version < 3)
     return 0;
+  /* Indexes with higher version than the one supported by GDB may be no
+     longer backward compatible.  */
+  if (version > 3)
+    return 0;
 
   map = OBSTACK_ZALLOC (&objfile->objfile_obstack, struct mapped_index);
   map->total_size = dwarf2_per_objfile->gdb_index.size;

Re: [patch] .gdb_index: Ignore higher versions

[Tom Tromey]

>>>>> "Jan" == Jan Kratochvil <jan.kratochvil@redhat.com> writes:

Jan> currently introducing new .gdb_index version (latest is v3, so >=4)
Jan> breaks all existing GDBs as they try too interpret the index format
Jan> the old way.

Jan> I have a reproducer here with some other patchset but I find it obvious.

Me too.

Jan> Sending for a comment as we could also introduce some versioning
Jan> features for compatible/incompatible extensions.  I do not think it
Jan> is worth it.

I agree.

Tom

Re: [patch] .gdb_index: Ignore higher versions

[Jan Kratochvil]

On Mon, 22 Nov 2010 21:29:19 +0100, Tom Tromey wrote:
> >>>>> "Jan" == Jan Kratochvil <jan.kratochvil@redhat.com> writes:
> Jan> Sending for a comment as we could also introduce some versioning
> Jan> features for compatible/incompatible extensions.  I do not think it
> Jan> is worth it.
> 
> I agree.

Checked in:
	http://sourceware.org/ml/gdb-cvs/2010-11/msg00105.html


Thanks,
Jan