[Crash-utility] [PATCH v3 0/3] vmalloc translation support for PPC

[Dave Anderson]

----- Original Message -----
> On 02/17/2012 07:25 PM, Dave Anderson wrote:
...
> >> Why doesn't it get called in x86_64 case even when the flags indicate
> >> page 'PRESENT' ?
> >
> > I don't know -- that's what I was asking!
> 
> The only suspect is :
> 
> 	if (page_exists) {
> >>>             if ((pc->flags & DEVMEM) && (paddr >= VTOP(vt->high_memory))) <<
>                          return;

I'm not sure whether you are testing ppc on a dumpfile or a live
system using /dev/mem, and I don't know what your paddr value and
vs your vt->high_memory are.

But I'm testing x86_64 on a live system with the crash memory
driver, so DEVMEM is not set.  And as it turns out, it doesn't
print the header because it returns here:

               if (!(section = valid_section_nr(section_nr))) {
#ifdef NOTDEF
                        break;    /* On a real sparsemem system we need to check
                                   * every section as gaps may exist.  But this
                                   * can be slow.  If we know we don't have gaps
                                   * just stop validating sections when we
                                   * get to the end of the valid ones.
                                   * In the future find a way to short circuit
                                   * this loop.
                                   */
#endif
                        if (mi->flags & ADDRESS_SPECIFIED) 
                                break;
                        continue;
                }

                if (print_hdr) {
...

Dave