[RFC v3 1/3] mm: add tracepoint for scanning pages

[Ebru Akagunduz <ebru.akagunduz@gmail.com>]

On Tue, Jul 14, 2015 at 12:06:46AM +0300, Kirill A. Shutemov wrote:
> On Mon, Jul 13, 2015 at 11:28:02PM +0300, Ebru Akagunduz wrote:
> > Using static tracepoints, data of functions is recorded.
> > It is good to automatize debugging without doing a lot
> > of changes in the source code.
> > 
> > This patch adds tracepoint for khugepaged_scan_pmd,
> > collapse_huge_page and __collapse_huge_page_isolate.
> > 
> > Signed-off-by: Ebru Akagunduz <ebru.akagunduz@gmail.com>
> > Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
> > Acked-by: Rik van Riel <riel@redhat.com>
> > ---
> > Changes in v2:
> >  - Nothing changed
> > 
> > Changes in v3:
> >  - Print page address instead of vm_start (Vlastimil Babka)
> >  - Define constants to specify exact tracepoint result (Vlastimil Babka)
> >  
> > 
> >  include/linux/mm.h                 |  18 ++++++
> >  include/trace/events/huge_memory.h | 100 ++++++++++++++++++++++++++++++++
> >  mm/huge_memory.c                   | 114 +++++++++++++++++++++++++++----------
> >  3 files changed, 203 insertions(+), 29 deletions(-)
> >  create mode 100644 include/trace/events/huge_memory.h
> > 
> > diff --git a/include/linux/mm.h b/include/linux/mm.h
> > index 7f47178..bf341c0 100644
> > --- a/include/linux/mm.h
> > +++ b/include/linux/mm.h
> > @@ -21,6 +21,24 @@
> >  #include <linux/resource.h>
> >  #include <linux/page_ext.h>
> >  
> > +#define MM_PMD_NULL		0
> > +#define MM_EXCEED_NONE_PTE	3
> > +#define MM_PTE_NON_PRESENT	4
> > +#define MM_PAGE_NULL		5
> > +#define MM_SCAN_ABORT		6
> > +#define MM_PAGE_COUNT		7
> > +#define MM_PAGE_LRU		8
> > +#define MM_ANY_PROCESS		0
> > +#define MM_VMA_NULL		2
> > +#define MM_VMA_CHECK		3
> > +#define MM_ADDRESS_RANGE	4
> > +#define MM_PAGE_LOCK		2
> > +#define MM_SWAP_CACHE_PAGE	6
> > +#define MM_ISOLATE_LRU_PAGE	7
> > +#define MM_ALLOC_HUGE_PAGE_FAIL	6
> > +#define MM_CGROUP_CHARGE_FAIL	7
> > +#define MM_COLLAPSE_ISOLATE_FAIL 5
> > +
> 
> These magic numbers looks very random. What's logic behind?
> 
I defined them to specify reason of all success and failure cases
of the functions with tracepoint. Only 1 means success case.

All other values mean failure, I give consecutive numbers as
far as possible, and tried to avoid conflicts of different functions
those can be fail for same reason.

kind regards,
Ebru