the exiting makedumpfile is almost there... :)

[Ken'ichi Ohmichi]

Hi Hedi, Jay,

Hedi Berriche wrote:
> In addition to what other folks have mentioned about giving the latest crash
> version a try, I'd like to point out that makedumpfile did spit a couple of
> warnings while creating the vmcore
> 
> 
> | Can't distinguish the pgtable.
> | The kernel version is not supported.
> | The created dumpfile may be incomplete.
> 
> these warnings added to the fact that later on crash choked with
> 
> | NOTE: page_hash_table does not exist in this kernel
> | crash: page excluded: kernel virtual address: e000006003108e00  type:
> 
> seem to suggest that the makedumpfile warnings could be relevant to the
> end result.

Oh, good point.
If makedumpfile cannot distinguish the pgtable, it guesses PGTABLE_3 and
creates a dumpfile. If Jay's kernel .config file has CONFIG_PGTABLE_4=y,
makedumpfile misunderstands the pgtable. If it has CONFIG_PGTABLE_4=y,
could you please try the attached patch ? This patch is only for debugging,
and I'll investigate the cause.


Thanks
Ken'ichi Ohmichi

---
diff -puN backup/makedumpfile-1.2.9/ia64.c makedumpfile/ia64.c
--- backup/makedumpfile-1.2.9/ia64.c	2008-09-04 16:31:58.000000000 +0900
+++ makedumpfile/ia64.c	2008-09-12 11:27:38.000000000 +0900
@@ -79,6 +79,7 @@ get_machdep_info_ia64(void)
 		vt.mem_flags |= MEMORY_PAGETABLE_3L;
 		DEBUG_MSG("PAGETABLE_3L : ON\n");
 	} else {
+		vt.mem_flags |= MEMORY_PAGETABLE_4L;
 		MSG("Can't distinguish the pgtable.\n");
 	}