similar to: [PATCH] add generic callbacks into compaction

Hello, This series try to enable migration of non-LRU pages, such as driver's page. My ARM-based platform occured severe fragmentation problem after long-term (several days) test. Sometimes even order-3 page allocation failed. It has memory size 512MB ~ 1024MB. 30% ~ 40% memory is consumed for graphic processing and 20~30 memory is reserved for zram. I found that many pages of GPU driver

Hello, This series try to enable migration of non-LRU pages, such as driver's page. My ARM-based platform occured severe fragmentation problem after long-term (several days) test. Sometimes even order-3 page allocation failed. It has memory size 512MB ~ 1024MB. 30% ~ 40% memory is consumed for graphic processing and 20~30 memory is reserved for zram. I found that many pages of GPU driver

Hello, This series try to enable migration of non-LRU pages, such as driver's page. My ARM-based platform occured severe fragmentation problem after long-term (several days) test. Sometimes even order-3 page allocation failed. It has memory size 512MB ~ 1024MB. 30% ~ 40% memory is consumed for graphic processing and 20~30 memory is reserved for zram. I found that many pages of GPU driver

Hello, This series try to enable migration of non-LRU pages, such as driver's page. My ARM-based platform occured severe fragmentation problem after long-term (several days) test. Sometimes even order-3 page allocation failed. It has memory size 512MB ~ 1024MB. 30% ~ 40% memory is consumed for graphic processing and 20~30 memory is reserved for zram. I found that many pages of GPU driver

From: Gioh Kim <gurugio at hanmail.net> Hello, This series try to enable migration of non-LRU pages, such as driver's page. My ARM-based platform occured severe fragmentation problem after long-term (several days) test. Sometimes even order-3 page allocation failed. It has memory size 512MB ~ 1024MB. 30% ~ 40% memory is consumed for graphic processing and 20~30 memory is reserved for

From: Gioh Kim <gurugio at hanmail.net> Hello, This series try to enable migration of non-LRU pages, such as driver's page. My ARM-based platform occured severe fragmentation problem after long-term (several days) test. Sometimes even order-3 page allocation failed. It has memory size 512MB ~ 1024MB. 30% ~ 40% memory is consumed for graphic processing and 20~30 memory is reserved for

Hello Gioh, I wanted to have such feature for zsmalloc. Thanks for the work. On Wed, Apr 01, 2015 at 08:11:30AM +0900, Gioh Kim wrote: > I sent a patch about page allocation for less fragmentation. > http://permalink.gmane.org/gmane.linux.kernel.mm/130599 > > It proposes a page allocator allocates pages in the same pageblock > for the drivers to move their unmovable pages. Some

On Tue, Jun 2, 2015 at 10:27 AM, Gioh Kim <gioh.kim at lge.com> wrote: > Apply driver page migration into balloon driver. > > Signed-off-by: Gioh Kim <gioh.kim at lge.com> > --- > drivers/virtio/virtio_balloon.c | 2 ++ > fs/proc/page.c | 4 +-- > include/linux/balloon_compaction.h | 42 ++++++++++++++++------- >

On Tue, Jun 2, 2015 at 10:27 AM, Gioh Kim <gioh.kim at lge.com> wrote: > Apply driver page migration into balloon driver. > > Signed-off-by: Gioh Kim <gioh.kim at lge.com> > --- > drivers/virtio/virtio_balloon.c | 2 ++ > fs/proc/page.c | 4 +-- > include/linux/balloon_compaction.h | 42 ++++++++++++++++------- >

From: Gioh Kim <gurugio at hanmail.net> Hello, This series try to enable migration of non-LRU pages, such as driver's page. My ARM-based platform occured severe fragmentation problem after long-term (several days) test. Sometimes even order-3 page allocation failed. It has memory size 512MB ~ 1024MB. 30% ~ 40% memory is consumed for graphic processing and 20~30 memory is reserved for

From: Gioh Kim <gurugio at hanmail.net> Hello, This series try to enable migration of non-LRU pages, such as driver's page. My ARM-based platform occured severe fragmentation problem after long-term (several days) test. Sometimes even order-3 page allocation failed. It has memory size 512MB ~ 1024MB. 30% ~ 40% memory is consumed for graphic processing and 20~30 memory is reserved for

On Fri, Jun 26, 2015 at 12:58 PM, Gioh Kim <gioh.kim at lge.com> wrote: > Compaction calls interfaces of driver page migration > instead of calling balloon migration directly. > > Signed-off-by: Gioh Kim <gioh.kim at lge.com> > --- > drivers/virtio/virtio_balloon.c | 1 + > mm/compaction.c | 9 +++++---- > mm/migrate.c | 21

On Fri, Jun 26, 2015 at 12:58 PM, Gioh Kim <gioh.kim at lge.com> wrote: > Compaction calls interfaces of driver page migration > instead of calling balloon migration directly. > > Signed-off-by: Gioh Kim <gioh.kim at lge.com> > --- > drivers/virtio/virtio_balloon.c | 1 + > mm/compaction.c | 9 +++++---- > mm/migrate.c | 21

Memory fragmentation introduced by ballooning might reduce significantly the number of 2MB contiguous memory blocks that can be used within a guest, thus imposing performance penalties associated with the reduced number of transparent huge pages that could be used by the guest workload. This patch-set follows the main idea discussed at 2012 LSFMMS session: "Ballooning for transparent huge

Memory fragmentation introduced by ballooning might reduce significantly the number of 2MB contiguous memory blocks that can be used within a guest, thus imposing performance penalties associated with the reduced number of transparent huge pages that could be used by the guest workload. This patch-set follows the main idea discussed at 2012 LSFMMS session: "Ballooning for transparent huge

Memory fragmentation introduced by ballooning might reduce significantly the number of 2MB contiguous memory blocks that can be used within a guest, thus imposing performance penalties associated with the reduced number of transparent huge pages that could be used by the guest workload. This patch-set follows the main idea discussed at 2012 LSFMMS session: "Ballooning for transparent huge

Memory fragmentation introduced by ballooning might reduce significantly the number of 2MB contiguous memory blocks that can be used within a guest, thus imposing performance penalties associated with the reduced number of transparent huge pages that could be used by the guest workload. This patch-set follows the main idea discussed at 2012 LSFMMS session: "Ballooning for transparent huge

Recently, I got many reports about perfermance degradation in embedded system(Android mobile phone, webOS TV and so on) and easy fork fail. The problem was fragmentation caused by zram and GPU driver mainly. With memory pressure, their pages were spread out all of pageblock and it cannot be migrated with current compaction algorithm which supports only LRU pages. In the end, compaction cannot

Recently, I got many reports about perfermance degradation in embedded system(Android mobile phone, webOS TV and so on) and easy fork fail. The problem was fragmentation caused by zram and GPU driver mainly. With memory pressure, their pages were spread out all of pageblock and it cannot be migrated with current compaction algorithm which supports only LRU pages. In the end, compaction cannot

Apply driver page migration into balloon driver. Signed-off-by: Gioh Kim <gioh.kim at lge.com> --- drivers/virtio/virtio_balloon.c | 2 ++ fs/proc/page.c | 4 +-- include/linux/balloon_compaction.h | 42 ++++++++++++++++------- include/linux/mm.h | 19 ----------- include/uapi/linux/kernel-page-flags.h | 2 +-