RE: [PATCHv3 2/3] mm: vmalloc: add VM_DMA flag to indicate areas used by dma-mapping framework

[Marek Szyprowski <m.szyprowski@samsung.com>]

Hi,

On Thursday, June 21, 2012 10:36 AM Minchan Kim wrote:

> On 06/13/2012 08:01 PM, Marek Szyprowski wrote:
> 
> > Add new type of vm_area intented to be used for mappings created by
> > dma-mapping framework.
> >
> > Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
> > Reviewed-by: Kyungmin Park <kyungmin.park@samsung.com>
> > ---
> >  include/linux/vmalloc.h |    1 +
> >  mm/vmalloc.c            |    3 +++
> >  2 files changed, 4 insertions(+), 0 deletions(-)
> >
> > diff --git a/include/linux/vmalloc.h b/include/linux/vmalloc.h
> > index 2e28f4d..e725b7b 100644
> > --- a/include/linux/vmalloc.h
> > +++ b/include/linux/vmalloc.h
> > @@ -14,6 +14,7 @@ struct vm_area_struct;		/* vma defining user mapping in
> mm_types.h */
> >  #define VM_USERMAP	0x00000008	/* suitable for remap_vmalloc_range */
> >  #define VM_VPAGES	0x00000010	/* buffer for pages was vmalloc'ed */
> >  #define VM_UNLIST	0x00000020	/* vm_struct is not listed in vmlist */
> > +#define VM_DMA		0x00000040	/* used by dma-mapping framework */
> >  /* bits [20..32] reserved for arch specific ioremap internals */
> >
> >  /*
> > diff --git a/mm/vmalloc.c b/mm/vmalloc.c
> > index 11308f0..e04d59b 100644
> > --- a/mm/vmalloc.c
> > +++ b/mm/vmalloc.c
> > @@ -2575,6 +2575,9 @@ static int s_show(struct seq_file *m, void *p)
> >  	if (v->flags & VM_IOREMAP)
> >  		seq_printf(m, " ioremap");
> >
> > +	if (v->flags & VM_DMA)
> > +		seq_printf(m, " dma");
> > +
> >  	if (v->flags & VM_ALLOC)
> >  		seq_printf(m, " vmalloc");
> >
> 
> 
> I still don't make sure that we should add new type for only ARM arch.
> I remember you said "It would be used for other architectures once we add" and
> Paul said he has a plan for SH. So at least, you should add such comment in changelog
> for persuading grumpy maintainers. :)
> 
> Frankly speaking, I could add my Reviewed-by but I think it wouldn't carry much weight
> because code is very tiny so you need Acked-by rather than Reviewed-by.
> IMHO, This problem is the thing only maintainer should decide.
> So I will toss the decision to akpm. Ccing akpm.(Ccing KOSAKI because he had a concern about
> this).
> 
> If anyone have a question to me, I'm Acked-by iff other architecture will use it.

Ok, if this flag is so controversial, I can change it to arch specific VM flag and utilize
'bits [20..32] reserved for arch specific ioremap internals'. Later, when other architectures 
start using similar flag, one can unify them and add such generic flag.

Best regards
-- 
Marek Szyprowski
Samsung Poland R&D Center