[Crash-utility] x86 remap allocator in kernel 3.0

[Dave Anderson]

----- Original Message -----
> Dne St 11. ledna 2012 00:37:50 Petr Tesarik napsal(a):
> > [...]
> > I can see now that this is unnecessarily complicated, because the
> > node_remap_* variables are static arrays of MAX_NUMNODES elements, so I
> > can get their size from the debuginfo at POST_GDB init and initialize a
> > machine-specific data type with it. I'll post another patch tomorrow.
> 
> And here we go. Tested on my system and seems to work just fine.
> 
> Petr Tesarik
> SUSE Linux
 
Hi Petr,

This looks pretty good to me.  However, just to clarify the 
chicken-and-egg situation here...

When remap_init() does these 3 readmem() calls, they will pass
through x86_kvtop_remap() -- which I guess would fail because
the arrays would still be at least partially uninitialized?:

+	readmem(start_vaddr, KVADDR, ms->remap_start_vaddr,
+		ms->max_numnodes * sizeof(ulong), "node_remap_start_vaddr",
+		FAULT_ON_ERROR);
+	readmem(end_vaddr, KVADDR, ms->remap_end_vaddr,
+		ms->max_numnodes * sizeof(ulong), "node_remap_end_vaddr",
+		FAULT_ON_ERROR);
+	readmem(start_pfn, KVADDR, ms->remap_start_pfn,
+		ms->max_numnodes * sizeof(ulong), "node_remap_end_vaddr",
+		FAULT_ON_ERROR);

To fortify this (and put my mind at ease), can we also set a "remap_state"
flag in the machdep structure that describes the remap data structures in 
maybe two states:

  UNINITIALIZED - set at SETUP_ENV time, then cleared in x86_init() before 
                  remap_init() is called
  NOT_USED      - if remap_init() fails

Then have x86_kvtop_remap() -- which will get called on every readmem() or kvtop()
for the rest of time, check for (UNINITIALIZED|NOT_USED), and return FALSE immediately
if either are set?

What do you think?

Dave