[PATCH 05/16] page-flags: define behavior of FS/IO-related flags on compound pages

2015-09-24 Thread Kirill A. Shutemov
It seems we don't have compound page on FS/IO path currently. Use PF_NO_COMPOUND to catch if we have. The odd exception is PG_dirty: sound uses compound pages and maps them with PTEs. PF_NO_COMPOUND triggers VM_BUG_ON() in set_page_dirty() on handling shared fault. Let's use PF_HEAD for PG_dirty.

Re: [PATCH 05/16] page-flags: define behavior of FS/IO-related flags on compound pages

2015-03-25 Thread Hugh Dickins
On Wed, 25 Mar 2015, Kirill A. Shutemov wrote: > > We only need tail refcounting for THP, so I think this should fix the issue: > > diff --git a/include/linux/mm.h b/include/linux/mm.h > index 4a3a38522ab4..9ab432660adb 100644 > --- a/include/linux/mm.h > +++ b/include/linux/mm.h > @@ -456,7 +456

Re: [PATCH 05/16] page-flags: define behavior of FS/IO-related flags on compound pages

2015-03-25 Thread Kirill A. Shutemov
On Tue, Mar 24, 2015 at 03:54:00PM -0700, Hugh Dickins wrote: > On Mon, 23 Mar 2015, Kirill A. Shutemov wrote: > > Should we avoid dirtying them in the first place? > > I don't think so: to do so would add more branches in hot paths, > just to avoid a rare case which works fine without them; and >

Re: [PATCH 05/16] page-flags: define behavior of FS/IO-related flags on compound pages

2015-03-24 Thread Hugh Dickins
On Mon, 23 Mar 2015, Kirill A. Shutemov wrote: > On Sun, Mar 22, 2015 at 05:02:58PM -0700, Hugh Dickins wrote: > > On Thu, 19 Mar 2015, Kirill A. Shutemov wrote: > > > On Thu, Mar 19, 2015 at 11:29:52AM -0700, Dave Hansen wrote: > > > > On 03/19/2015 10:08 AM, Kirill A. Shutemov wrote: > > > > > Th

Re: [PATCH 05/16] page-flags: define behavior of FS/IO-related flags on compound pages

2015-03-23 Thread Kirill A. Shutemov
On Sun, Mar 22, 2015 at 05:02:58PM -0700, Hugh Dickins wrote: > On Thu, 19 Mar 2015, Kirill A. Shutemov wrote: > > On Thu, Mar 19, 2015 at 11:29:52AM -0700, Dave Hansen wrote: > > > On 03/19/2015 10:08 AM, Kirill A. Shutemov wrote: > > > > The odd exception is PG_dirty: sound uses compound pages an

Re: [PATCH 05/16] page-flags: define behavior of FS/IO-related flags on compound pages

2015-03-22 Thread Hugh Dickins
On Thu, 19 Mar 2015, Kirill A. Shutemov wrote: > On Thu, Mar 19, 2015 at 11:29:52AM -0700, Dave Hansen wrote: > > On 03/19/2015 10:08 AM, Kirill A. Shutemov wrote: > > > The odd exception is PG_dirty: sound uses compound pages and maps them > > > with PTEs. NO_COMPOUND triggers VM_BUG_ON() in set_p

Re: [PATCH 05/16] page-flags: define behavior of FS/IO-related flags on compound pages

2015-03-19 Thread Kirill A. Shutemov
On Thu, Mar 19, 2015 at 11:29:52AM -0700, Dave Hansen wrote: > On 03/19/2015 10:08 AM, Kirill A. Shutemov wrote: > > The odd exception is PG_dirty: sound uses compound pages and maps them > > with PTEs. NO_COMPOUND triggers VM_BUG_ON() in set_page_dirty() on > > handling shared fault. Let's use HEA

Re: [PATCH 05/16] page-flags: define behavior of FS/IO-related flags on compound pages

2015-03-19 Thread Dave Hansen
On 03/19/2015 10:08 AM, Kirill A. Shutemov wrote: > The odd expection is PG_dirty: sound uses compound pages and maps them > with PTEs. NO_COMPOUND triggers VM_BUG_ON() in set_page_dirty() on > handling shared fault. Let's use HEAD for PG_dirty. Can we get the sound guys to look at this, btw? It

[PATCH 05/16] page-flags: define behavior of FS/IO-related flags on compound pages

2015-03-19 Thread Kirill A. Shutemov
It seems we don't have compound page on FS/IO path currently. Use NO_COMPOUND to catch if we have. The odd expection is PG_dirty: sound uses compound pages and maps them with PTEs. NO_COMPOUND triggers VM_BUG_ON() in set_page_dirty() on handling shared fault. Let's use HEAD for PG_dirty. Signed-o