Re: [RFC][PATCH 3/9] mm: page->pfmemalloc only used by slab/skb

2014-01-15 Thread Kirill A. Shutemov
On Tue, Jan 14, 2014 at 10:00:51AM -0800, Dave Hansen wrote: > > From: Dave Hansen > > page->pfmemalloc does not deserve a spot in 'struct page'. It is > only used transiently _just_ after a page leaves the buddy > allocator. > > Instead of declaring a union, we move its functionality behind

Re: [RFC][PATCH 3/9] mm: page-pfmemalloc only used by slab/skb

2014-01-15 Thread Kirill A. Shutemov
On Tue, Jan 14, 2014 at 10:00:51AM -0800, Dave Hansen wrote: From: Dave Hansen dave.han...@linux.intel.com page-pfmemalloc does not deserve a spot in 'struct page'. It is only used transiently _just_ after a page leaves the buddy allocator. Instead of declaring a union, we move its

Re: [RFC][PATCH 3/9] mm: page->pfmemalloc only used by slab/skb

2014-01-14 Thread David Rientjes
On Tue, 14 Jan 2014, Dave Hansen wrote: > >> page->pfmemalloc does not deserve a spot in 'struct page'. It is > >> only used transiently _just_ after a page leaves the buddy > >> allocator. > > > > Why would we need to do this if we are removing the cmpxchg_double? > > Why do we need the

Re: [RFC][PATCH 3/9] mm: page->pfmemalloc only used by slab/skb

2014-01-14 Thread Dave Hansen
On 01/14/2014 11:49 AM, Christoph Lameter wrote: > On Tue, 14 Jan 2014, Dave Hansen wrote: >> page->pfmemalloc does not deserve a spot in 'struct page'. It is >> only used transiently _just_ after a page leaves the buddy >> allocator. > > Why would we need to do this if we are removing the

Re: [RFC][PATCH 3/9] mm: page->pfmemalloc only used by slab/skb

2014-01-14 Thread Christoph Lameter
On Tue, 14 Jan 2014, Dave Hansen wrote: > page->pfmemalloc does not deserve a spot in 'struct page'. It is > only used transiently _just_ after a page leaves the buddy > allocator. Why would we need to do this if we are removing the cmpxchg_double? -- To unsubscribe from this list: send the

[RFC][PATCH 3/9] mm: page->pfmemalloc only used by slab/skb

2014-01-14 Thread Dave Hansen
From: Dave Hansen page->pfmemalloc does not deserve a spot in 'struct page'. It is only used transiently _just_ after a page leaves the buddy allocator. Instead of declaring a union, we move its functionality behind a few quick accessor functions. This way we could also much more easily

[RFC][PATCH 3/9] mm: page-pfmemalloc only used by slab/skb

2014-01-14 Thread Dave Hansen
From: Dave Hansen dave.han...@linux.intel.com page-pfmemalloc does not deserve a spot in 'struct page'. It is only used transiently _just_ after a page leaves the buddy allocator. Instead of declaring a union, we move its functionality behind a few quick accessor functions. This way we could

Re: [RFC][PATCH 3/9] mm: page-pfmemalloc only used by slab/skb

2014-01-14 Thread Christoph Lameter
On Tue, 14 Jan 2014, Dave Hansen wrote: page-pfmemalloc does not deserve a spot in 'struct page'. It is only used transiently _just_ after a page leaves the buddy allocator. Why would we need to do this if we are removing the cmpxchg_double? -- To unsubscribe from this list: send the line

Re: [RFC][PATCH 3/9] mm: page-pfmemalloc only used by slab/skb

2014-01-14 Thread Dave Hansen
On 01/14/2014 11:49 AM, Christoph Lameter wrote: On Tue, 14 Jan 2014, Dave Hansen wrote: page-pfmemalloc does not deserve a spot in 'struct page'. It is only used transiently _just_ after a page leaves the buddy allocator. Why would we need to do this if we are removing the cmpxchg_double?

Re: [RFC][PATCH 3/9] mm: page-pfmemalloc only used by slab/skb

2014-01-14 Thread David Rientjes
On Tue, 14 Jan 2014, Dave Hansen wrote: page-pfmemalloc does not deserve a spot in 'struct page'. It is only used transiently _just_ after a page leaves the buddy allocator. Why would we need to do this if we are removing the cmpxchg_double? Why do we need the patch? 'struct