Re: [PATCH] z3fold: fix wrong handling of headless pages

Hi Jongseok, Den fre 6 juli 2018 kl 07:11 skrev Jongseok Kim : > > During the processing of headless pages in z3fold_reclaim_page(), > there was a problem that the zhdr pointed to another page > or a page was already released in z3fold_free(). So, the wrong page > is encoded in headless, or

Re: [PATCH] z3fold: fix wrong handling of headless pages

Hi Jongseok, Den fre 6 juli 2018 kl 07:11 skrev Jongseok Kim : > > During the processing of headless pages in z3fold_reclaim_page(), > there was a problem that the zhdr pointed to another page > or a page was already released in z3fold_free(). So, the wrong page > is encoded in headless, or

[PATCH] z3fold: fix possible reclaim races

unction. This patch supersedes "[PATCH] z3fold: fix reclaim lock-ups". Signed-off-by: Vitaly Wool Reviewed-by: Snild Dolkow --- mm/z3fold.c | 101 1 file changed, 62 insertions(+), 39 deletions(-) diff --git a/mm/z3fold.c b/

[PATCH] z3fold: fix possible reclaim races

unction. This patch supersedes "[PATCH] z3fold: fix reclaim lock-ups". Signed-off-by: Vitaly Wool Reviewed-by: Snild Dolkow --- mm/z3fold.c | 101 1 file changed, 62 insertions(+), 39 deletions(-) diff --git a/mm/z3fold.c b/

[PATCH] z3fold: encode object length in the handle

object we couldn't correctly map before. Signed-off-by: Vitaly Wool --- mm/z3fold.c | 48 +++- 1 file changed, 35 insertions(+), 13 deletions(-) diff --git a/mm/z3fold.c b/mm/z3fold.c index 4b366d181f35..86359b565d45 100644 --- a/mm/z3fold.c +++ b/m

[PATCH] z3fold: encode object length in the handle

object we couldn't correctly map before. Signed-off-by: Vitaly Wool --- mm/z3fold.c | 48 +++- 1 file changed, 35 insertions(+), 13 deletions(-) diff --git a/mm/z3fold.c b/mm/z3fold.c index 4b366d181f35..86359b565d45 100644 --- a/mm/z3fold.c +++ b/m

Re: [PATCH v2] z3fold: fix reclaim lock-ups

Hi Jongseok, Den tors 3 maj 2018 kl 08:36 skrev Jongseok Kim : > In the processing of headless pages, there was a problem that the > zhdr pointed to another page or a page was alread released in > z3fold_free(). So, the wrong page is encoded in headless, or test_bit > does not

Re: [PATCH v2] z3fold: fix reclaim lock-ups

Hi Jongseok, Den tors 3 maj 2018 kl 08:36 skrev Jongseok Kim : > In the processing of headless pages, there was a problem that the > zhdr pointed to another page or a page was alread released in > z3fold_free(). So, the wrong page is encoded in headless, or test_bit > does not work properly in

[PATCH] z3fold: fix reclaim lock-ups

Do not try to optimize in-page object layout while the page is under reclaim. This fixes lock-ups on reclaim and improves reclaim performance at the same time. Reported-by: Guenter Roeck <li...@roeck-us.net> Signed-off-by: Vitaly Wool <vitaly@sony.com> --- mm/z

[PATCH] z3fold: fix reclaim lock-ups

Do not try to optimize in-page object layout while the page is under reclaim. This fixes lock-ups on reclaim and improves reclaim performance at the same time. Reported-by: Guenter Roeck Signed-off-by: Vitaly Wool --- mm/z3fold.c | 42 ++ 1 file changed

Re: Crashes/hung tasks with z3pool under memory pressure

Den tis 17 apr. 2018 kl 18:35 skrev Guenter Roeck : > Getting better; the log is much less noisy. Unfortunately, there are still > locking problems, resulting in a hung task. I copied the log message to [1]. > This is with [2] applied on top of v4.17-rc1. Now this version

Re: Crashes/hung tasks with z3pool under memory pressure

Den tis 17 apr. 2018 kl 18:35 skrev Guenter Roeck : > Getting better; the log is much less noisy. Unfortunately, there are still > locking problems, resulting in a hung task. I copied the log message to [1]. > This is with [2] applied on top of v4.17-rc1. Now this version (this is a full patch

Re: Crashes/hung tasks with z3pool under memory pressure

Hi Guenter, > [ ... ] > > Ugh. Could you please keep that patch and apply this on top: > > > > diff --git a/mm/z3fold.c b/mm/z3fold.c > > index c0bca6153b95..e8a80d044d9e 100644 > > --- a/mm/z3fold.c > > +++ b/mm/z3fold.c > > @@ -840,6 +840,7 @@ static int z3fold_reclaim_page(struct z3fold_pool

Re: Crashes/hung tasks with z3pool under memory pressure

Hi Guenter, > [ ... ] > > Ugh. Could you please keep that patch and apply this on top: > > > > diff --git a/mm/z3fold.c b/mm/z3fold.c > > index c0bca6153b95..e8a80d044d9e 100644 > > --- a/mm/z3fold.c > > +++ b/mm/z3fold.c > > @@ -840,6 +840,7 @@ static int z3fold_reclaim_page(struct z3fold_pool

Re: Crashes/hung tasks with z3pool under memory pressure

On 4/16/18 5:58 PM, Guenter Roeck wrote: On Mon, Apr 16, 2018 at 02:43:01PM +0200, Vitaly Wool wrote: Hey Guenter, On 04/13/2018 07:56 PM, Guenter Roeck wrote: On Fri, Apr 13, 2018 at 05:40:18PM +, Vitaly Wool wrote: On Fri, Apr 13, 2018, 7:35 PM Guenter Roeck <li...@roeck-us.

Re: Crashes/hung tasks with z3pool under memory pressure

On 4/16/18 5:58 PM, Guenter Roeck wrote: On Mon, Apr 16, 2018 at 02:43:01PM +0200, Vitaly Wool wrote: Hey Guenter, On 04/13/2018 07:56 PM, Guenter Roeck wrote: On Fri, Apr 13, 2018 at 05:40:18PM +, Vitaly Wool wrote: On Fri, Apr 13, 2018, 7:35 PM Guenter Roeck wrote: On Fri, Apr 13

Re: Crashes/hung tasks with z3pool under memory pressure

Hey Guenter, On 04/13/2018 07:56 PM, Guenter Roeck wrote: On Fri, Apr 13, 2018 at 05:40:18PM +, Vitaly Wool wrote: On Fri, Apr 13, 2018, 7:35 PM Guenter Roeck <li...@roeck-us.net> wrote: On Fri, Apr 13, 2018 at 05:21:02AM +, Vitaly Wool wrote: Hi Guenter, Den fre 13 apr. 2

Re: Crashes/hung tasks with z3pool under memory pressure

Hey Guenter, On 04/13/2018 07:56 PM, Guenter Roeck wrote: On Fri, Apr 13, 2018 at 05:40:18PM +, Vitaly Wool wrote: On Fri, Apr 13, 2018, 7:35 PM Guenter Roeck wrote: On Fri, Apr 13, 2018 at 05:21:02AM +, Vitaly Wool wrote: Hi Guenter, Den fre 13 apr. 2018 kl 00:01 skrev Guenter

Re: Crashes/hung tasks with z3pool under memory pressure

Hi Guenter, Den fre 13 apr. 2018 kl 00:01 skrev Guenter Roeck : > Hi all, > we are observing crashes with z3pool under memory pressure. The kernel version > used to reproduce the problem is v4.16-11827-g5d1365940a68, but the problem was > also seen with v4.14 based kernels.

Re: Crashes/hung tasks with z3pool under memory pressure

Hi Guenter, Den fre 13 apr. 2018 kl 00:01 skrev Guenter Roeck : > Hi all, > we are observing crashes with z3pool under memory pressure. The kernel version > used to reproduce the problem is v4.16-11827-g5d1365940a68, but the problem was > also seen with v4.14 based kernels. just before I dig

[PATCH] z3fold: limit use of stale list for allocation

Currently if z3fold couldn't find an unbuddied page it would first try to pull a page off the stale list. The problem with this approach is that we can't 100% guarantee that the page is not processed by the workqueue thread at the same time unless we run cancel_work_sync() on it, which we can't

[PATCH] z3fold: limit use of stale list for allocation

Currently if z3fold couldn't find an unbuddied page it would first try to pull a page off the stale list. The problem with this approach is that we can't 100% guarantee that the page is not processed by the workqueue thread at the same time unless we run cancel_work_sync() on it, which we can't

[PATCH] binder: use lockless list for deferred_work

Binder uses hlist for deferred list, which isn't a good match. It's slow and requires mutual exclusion mechanism to protect its operations. Moreover, having schedule_work() called under a mutex may cause significant delays and creates noticeable adverse effect on Binder performance. Deferred list

[PATCH] binder: use lockless list for deferred_work

Binder uses hlist for deferred list, which isn't a good match. It's slow and requires mutual exclusion mechanism to protect its operations. Moreover, having schedule_work() called under a mutex may cause significant delays and creates noticeable adverse effect on Binder performance. Deferred list

[PATCH v2] llist: add llist_[add|del_first]_exclusive

It sometimes is necessary to be able to be able to use llist in the following manner: > if (node_unlisted(node)) > llst_add(node, list); i. e. only add a node to the list if it's not already on a list. This is not possible without taking locks because otherwise there's an

[PATCH v2] llist: add llist_[add|del_first]_exclusive

It sometimes is necessary to be able to be able to use llist in the following manner: > if (node_unlisted(node)) > llst_add(node, list); i. e. only add a node to the list if it's not already on a list. This is not possible without taking locks because otherwise there's an

[PATCH/RFC] llist: add llist_[add|del_first]_exclusive

will delete the first node off the list and mark it as not being on any list. Signed-off-by: Vitaly Wool <vitaly@sony.com> --- include/linux/llist.h | 25 + lib/llist.c | 29 + 2 files changed, 54 insertions(+) diff --git a/i

[PATCH/RFC] llist: add llist_[add|del_first]_exclusive

will delete the first node off the list and mark it as not being on any list. Signed-off-by: Vitaly Wool --- include/linux/llist.h | 25 + lib/llist.c | 29 + 2 files changed, 54 insertions(+) diff --git a/include/linux/llist.h b

[PATCH] z3fold: use kref to prevent page free / compact race

compaction is scheduled. It then becomes compaction function's responsibility to decrease the counter and quit immediately if the page was actually freed. Signed-off-by: Vitaly Wool <vitaly.w...@sonymobile.com> Cc: stable <sta...@vger.kernel.org> --- mm/z3fold.c | 10 -- 1 file

[PATCH] z3fold: use kref to prevent page free / compact race

compaction is scheduled. It then becomes compaction function's responsibility to decrease the counter and quit immediately if the page was actually freed. Signed-off-by: Vitaly Wool Cc: stable --- mm/z3fold.c | 10 -- 1 file changed, 8 insertions(+), 2 deletions(-) diff --git a/mm/z3f

Re: [PATCH] z3fold: fix stale list handling

Hi Andrew, 2017-09-14 23:15 GMT+02:00 Andrew Morton <a...@linux-foundation.org>: > On Thu, 14 Sep 2017 15:59:36 +0200 Vitaly Wool <vitalyw...@gmail.com> wrote: > >> Fix the situation when clear_bit() is called for page->private before >> the page pointer is actua

Re: [PATCH] z3fold: fix stale list handling

Hi Andrew, 2017-09-14 23:15 GMT+02:00 Andrew Morton : > On Thu, 14 Sep 2017 15:59:36 +0200 Vitaly Wool wrote: > >> Fix the situation when clear_bit() is called for page->private before >> the page pointer is actually assigned. While at it, remove work_busy() >>

[PATCH] z3fold: fix stale list handling

Fix the situation when clear_bit() is called for page->private before the page pointer is actually assigned. While at it, remove work_busy() check because it is costly and does not give 100% guarantee anyway. Signed-of-by: Vitaly Wool <vitalyw...@gmail.com> --- mm/z3fold.c | 6 ++---

[PATCH] z3fold: fix stale list handling

Fix the situation when clear_bit() is called for page->private before the page pointer is actually assigned. While at it, remove work_busy() check because it is costly and does not give 100% guarantee anyway. Signed-of-by: Vitaly Wool --- mm/z3fold.c | 6 ++ 1 file changed, 2 inserti

z3fold: fix potential race in z3fold_reclaim_page

a bug. To avoid that, spin_lock() has to be taken earlier, before the kref_put() call mentioned earlier. Signed-off-by: Vitaly Wool <vitalyw...@gmail.com> --- mm/z3fold.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/mm/z3fold.c b/mm/z3fold.c index 486550

z3fold: fix potential race in z3fold_reclaim_page

a bug. To avoid that, spin_lock() has to be taken earlier, before the kref_put() call mentioned earlier. Signed-off-by: Vitaly Wool --- mm/z3fold.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/mm/z3fold.c b/mm/z3fold.c index 486550df32be..b04fa3ba1bf2 100644 --- a/mm

Re: [PATCH/RFC] ion: add movability support for page pools

2017-09-06 2:19 GMT+02:00 Laura Abbott <labb...@redhat.com>: > On 09/05/2017 05:55 AM, Vitaly Wool wrote: >> ion page pool may become quite large and scattered all around >> the kernel memory area. These pages are actually not used so >> moving them around to reduce f

Re: [PATCH/RFC] ion: add movability support for page pools

2017-09-06 2:19 GMT+02:00 Laura Abbott : > On 09/05/2017 05:55 AM, Vitaly Wool wrote: >> ion page pool may become quite large and scattered all around >> the kernel memory area. These pages are actually not used so >> moving them around to reduce fragmentation is quite che

[PATCH/RFC] ion: add movability support for page pools

-off-by: Vitaly Wool <vitalyw...@gmail.com> --- drivers/staging/android/ion/ion.h | 2 + drivers/staging/android/ion/ion_page_pool.c | 165 +++- 2 files changed, 163 insertions(+), 4 deletions(-) diff --git a/drivers/staging/android/ion/ion.h b/drivers/s

[PATCH/RFC] ion: add movability support for page pools

-off-by: Vitaly Wool --- drivers/staging/android/ion/ion.h | 2 + drivers/staging/android/ion/ion_page_pool.c | 165 +++- 2 files changed, 163 insertions(+), 4 deletions(-) diff --git a/drivers/staging/android/ion/ion.h b/drivers/staging/android/ion/ion.h index

[PATCH v2] z3fold: use per-cpu unbuddied lists

in for almost 6x performance increase. Signed-off-by: Vitaly Wool <vitalyw...@gmail.com> --- mm/z3fold.c | 479 +++- 1 file changed, 344 insertions(+), 135 deletions(-) diff --git a/mm/z3fold.c b/mm/z3fold.c index 54f63c4a809a..b44ce5

[PATCH v2] z3fold: use per-cpu unbuddied lists

in for almost 6x performance increase. Signed-off-by: Vitaly Wool --- mm/z3fold.c | 479 +++- 1 file changed, 344 insertions(+), 135 deletions(-) diff --git a/mm/z3fold.c b/mm/z3fold.c index 54f63c4a809a..b44ce5059442 100644 --- a/mm/z3fold.c

[PATCH] z3fold: use per-cpu unbuddied lists

will go up. This patch also introduces two worker threads which: one for async in-page object layout optimization and one for releasing freed pages. Signed-off-by: Vitaly Wool <vitalyw...@gmail.com> --- mm/z3fold.c | 479 +++-

[PATCH] z3fold: use per-cpu unbuddied lists

will go up. This patch also introduces two worker threads which: one for async in-page object layout optimization and one for releasing freed pages. Signed-off-by: Vitaly Wool --- mm/z3fold.c | 479 +++- 1 file changed, 344 insertions

[PATCH] z3fold: fix page locking in z3fold_alloc()

Stress testing of the current z3fold implementation on a 8-core system revealed it was possible that a z3fold page deleted from its unbuddied list in z3fold_alloc() would be put on another unbuddied list by z3fold_free() while z3fold_alloc() is still processing it. This has been introduced with

[PATCH] z3fold: fix page locking in z3fold_alloc()

Stress testing of the current z3fold implementation on a 8-core system revealed it was possible that a z3fold page deleted from its unbuddied list in z3fold_alloc() would be put on another unbuddied list by z3fold_free() while z3fold_alloc() is still processing it. This has been introduced with

[PATCH] z3fold: fix spinlock unlocking in page reclaim

The patch "z3fold: add kref refcounting" introduced a bug in z3fold_reclaim_page() with function exit that may leave pool->lock spinlock held. Here comes the trivial fix. Reported-by: Alexey Khoroshilov <khoroshi...@ispras.ru> Signed-off-by: Vitaly Wool <vitalyw...@gmail.com

[PATCH] z3fold: fix spinlock unlocking in page reclaim

The patch "z3fold: add kref refcounting" introduced a bug in z3fold_reclaim_page() with function exit that may leave pool->lock spinlock held. Here comes the trivial fix. Reported-by: Alexey Khoroshilov Signed-off-by: Vitaly Wool --- mm/z3fold.c | 1 + 1 file changed, 1 inser

[PATCH/RESEND v3 5/5] z3fold: add kref refcounting

With both coming and already present locking optimizations, introducing kref to reference-count z3fold objects is the right thing to do. Moreover, it makes buddied list no longer necessary, and allows for a simpler handling of headless pages. Signed-off-by: Vitaly Wool <vitalyw...@gmail.

[PATCH/RESEND v3 5/5] z3fold: add kref refcounting

With both coming and already present locking optimizations, introducing kref to reference-count z3fold objects is the right thing to do. Moreover, it makes buddied list no longer necessary, and allows for a simpler handling of headless pages. Signed-off-by: Vitaly Wool --- mm/z3fold.c | 151

[PATCH/RESEND v3 3/5] z3fold: extend compaction function

, using BIG_CHUNK_GAP define as a threshold for middle chunk to be worth moving. Signed-off-by: Vitaly Wool <vitalyw...@gmail.com> --- mm/z3fold.c | 26 +- 1 file changed, 25 insertions(+), 1 deletion(-) diff --git a/mm/z3fold.c b/mm/z3fold.c index 98ab01f..be8b56e

[PATCH/RESEND v3 3/5] z3fold: extend compaction function

, using BIG_CHUNK_GAP define as a threshold for middle chunk to be worth moving. Signed-off-by: Vitaly Wool --- mm/z3fold.c | 26 +- 1 file changed, 25 insertions(+), 1 deletion(-) diff --git a/mm/z3fold.c b/mm/z3fold.c index 98ab01f..be8b56e 100644 --- a/mm/z3fold.c +++ b

[PATCH/RESEND v3 4/5] z3fold: use per-page spinlock

implements spinlock-based per-page locking mechanism which is lightweight enough to normally fit ok into the z3fold header. Signed-off-by: Vitaly Wool <vitalyw...@gmail.com> --- mm/z3fold.c | 148 +++- 1 file changed, 106 insertions(

[PATCH/RESEND v3 4/5] z3fold: use per-page spinlock

implements spinlock-based per-page locking mechanism which is lightweight enough to normally fit ok into the z3fold header. Signed-off-by: Vitaly Wool --- mm/z3fold.c | 148 +++- 1 file changed, 106 insertions(+), 42 deletions(-) diff --git a/mm

[PATCH/RESEND v3 2/5] z3fold: fix header size related issues

of num_free_chunks() and the address to move the middle chunk to in case of in-page compaction in z3fold_compact_page(). Signed-off-by: Vitaly Wool <vitalyw...@gmail.com> --- mm/z3fold.c | 114 ++-- 1 file changed, 64 insertions(+), 50 del

[PATCH/RESEND v3 2/5] z3fold: fix header size related issues

of num_free_chunks() and the address to move the middle chunk to in case of in-page compaction in z3fold_compact_page(). Signed-off-by: Vitaly Wool --- mm/z3fold.c | 114 ++-- 1 file changed, 64 insertions(+), 50 deletions(-) diff --git a/mm/z3fold.c

[PATCH/RESEND v3 1/5] z3fold: make pages_nr atomic

This patch converts pages_nr per-pool counter to atomic64_t. Signed-off-by: Vitaly Wool <vitalyw...@gmail.com> --- mm/z3fold.c | 20 +--- 1 file changed, 9 insertions(+), 11 deletions(-) diff --git a/mm/z3fold.c b/mm/z3fold.c index 207e5dd..2273789 100644 --- a/mm/z3fold.c

[PATCH/RESEND v3 1/5] z3fold: make pages_nr atomic

This patch converts pages_nr per-pool counter to atomic64_t. Signed-off-by: Vitaly Wool --- mm/z3fold.c | 20 +--- 1 file changed, 9 insertions(+), 11 deletions(-) diff --git a/mm/z3fold.c b/mm/z3fold.c index 207e5dd..2273789 100644 --- a/mm/z3fold.c +++ b/mm/z3fold.c @@ -80,7

[PATCH/RESEND v3 0/5] z3fold: optimizations and fixes

This is a new take on z3fold optimizations/fixes consolidation, revised after comments from Dan ([1] - [6]). The coming patches are to be applied on top of the following commit: Author: zhong jiang Date: Tue Dec 20 11:53:40 2016 +1100 mm/z3fold.c: limit first_num

[PATCH/RESEND v3 0/5] z3fold: optimizations and fixes

This is a new take on z3fold optimizations/fixes consolidation, revised after comments from Dan ([1] - [6]). The coming patches are to be applied on top of the following commit: Author: zhong jiang Date: Tue Dec 20 11:53:40 2016 +1100 mm/z3fold.c: limit first_num to the actual range of

Re: [PATCH/RESEND v2 5/5] z3fold: add kref refcounting

t but I can do that :) > > the header's already rounded up to chunk size, so if there's room then > it won't take any extra memory. but it works either way. So let's have it like this then: With both coming and already present locking optimizations, introducing kref to reference-count z

Re: [PATCH/RESEND v2 5/5] z3fold: add kref refcounting

t; > the header's already rounded up to chunk size, so if there's room then > it won't take any extra memory. but it works either way. So let's have it like this then: With both coming and already present locking optimizations, introducing kref to referen

Re: [PATCH/RESEND v2 3/5] z3fold: extend compaction function

On Wed, 11 Jan 2017 17:43:13 +0100 Vitaly Wool <vitalyw...@gmail.com> wrote: > On Wed, Jan 11, 2017 at 5:28 PM, Dan Streetman <ddstr...@ieee.org> wrote: > > On Wed, Jan 11, 2017 at 10:06 AM, Vitaly Wool <vitalyw...@gmail.com> wrote: > >> z3fold_compact_page()

Re: [PATCH/RESEND v2 3/5] z3fold: extend compaction function

On Wed, 11 Jan 2017 17:43:13 +0100 Vitaly Wool wrote: > On Wed, Jan 11, 2017 at 5:28 PM, Dan Streetman wrote: > > On Wed, Jan 11, 2017 at 10:06 AM, Vitaly Wool wrote: > >> z3fold_compact_page() currently only handles the situation when > >> there's a single middle c

Re: [PATCH/RESEND v2 5/5] z3fold: add kref refcounting

On Wed, Jan 11, 2017 at 6:39 PM, Dan Streetman <ddstr...@ieee.org> wrote: > On Wed, Jan 11, 2017 at 12:27 PM, Vitaly Wool <vitalyw...@gmail.com> wrote: >> On Wed, Jan 11, 2017 at 6:08 PM, Dan Streetman <ddstr...@ieee.org> wrote: >>> On Wed, Jan 11, 20

Re: [PATCH/RESEND v2 5/5] z3fold: add kref refcounting

On Wed, Jan 11, 2017 at 6:39 PM, Dan Streetman wrote: > On Wed, Jan 11, 2017 at 12:27 PM, Vitaly Wool wrote: >> On Wed, Jan 11, 2017 at 6:08 PM, Dan Streetman wrote: >>> On Wed, Jan 11, 2017 at 10:06 AM, Vitaly Wool wrote: >>>> With both coming and already

Re: [PATCH/RESEND v2 5/5] z3fold: add kref refcounting

On Wed, Jan 11, 2017 at 6:08 PM, Dan Streetman <ddstr...@ieee.org> wrote: > On Wed, Jan 11, 2017 at 10:06 AM, Vitaly Wool <vitalyw...@gmail.com> wrote: >> With both coming and already present locking optimizations, >> introducing kref to reference-count z3fold objects

Re: [PATCH/RESEND v2 5/5] z3fold: add kref refcounting

On Wed, Jan 11, 2017 at 6:08 PM, Dan Streetman wrote: > On Wed, Jan 11, 2017 at 10:06 AM, Vitaly Wool wrote: >> With both coming and already present locking optimizations, >> introducing kref to reference-count z3fold objects is the right >> thing to do. Moreover, it makes b

Re: [PATCH/RESEND 5/5] z3fold: add kref refcounting

On Wed, Jan 11, 2017 at 5:58 PM, Dan Streetman <ddstr...@ieee.org> wrote: > On Wed, Jan 11, 2017 at 5:52 AM, Vitaly Wool <vitalyw...@gmail.com> wrote: >> On Wed, Jan 4, 2017 at 7:42 PM, Dan Streetman <ddstr...@ieee.org> wrote: >>> On Sun, Dec 25, 2016 at 7:40 P

Re: [PATCH/RESEND 5/5] z3fold: add kref refcounting

On Wed, Jan 11, 2017 at 5:58 PM, Dan Streetman wrote: > On Wed, Jan 11, 2017 at 5:52 AM, Vitaly Wool wrote: >> On Wed, Jan 4, 2017 at 7:42 PM, Dan Streetman wrote: >>> On Sun, Dec 25, 2016 at 7:40 PM, Vitaly Wool wrote: >>>> With both coming and already

Re: [PATCH/RESEND v2 3/5] z3fold: extend compaction function

On Wed, Jan 11, 2017 at 5:28 PM, Dan Streetman <ddstr...@ieee.org> wrote: > On Wed, Jan 11, 2017 at 10:06 AM, Vitaly Wool <vitalyw...@gmail.com> wrote: >> z3fold_compact_page() currently only handles the situation when >> there's a single middle chunk within the z3f

Re: [PATCH/RESEND v2 3/5] z3fold: extend compaction function

On Wed, Jan 11, 2017 at 5:28 PM, Dan Streetman wrote: > On Wed, Jan 11, 2017 at 10:06 AM, Vitaly Wool wrote: >> z3fold_compact_page() currently only handles the situation when >> there's a single middle chunk within the z3fold page. However it >> may be worth it to mov

[PATCH/RESEND v2 4/5] z3fold: use per-page spinlock

implements spinlock-based per-page locking mechanism which is lightweight enough to normally fit ok into the z3fold header. Signed-off-by: Vitaly Wool <vitalyw...@gmail.com> --- mm/z3fold.c | 148 +++- 1 file changed, 106 insertions(

[PATCH/RESEND v2 4/5] z3fold: use per-page spinlock

implements spinlock-based per-page locking mechanism which is lightweight enough to normally fit ok into the z3fold header. Signed-off-by: Vitaly Wool --- mm/z3fold.c | 148 +++- 1 file changed, 106 insertions(+), 42 deletions(-) diff --git a/mm

[PATCH/RESEND v2 3/5] z3fold: extend compaction function

, using BIG_CHUNK_GAP define as a threshold for middle chunk to be worth moving. Signed-off-by: Vitaly Wool <vitalyw...@gmail.com> --- mm/z3fold.c | 26 +- 1 file changed, 25 insertions(+), 1 deletion(-) diff --git a/mm/z3fold.c b/mm/z3fold.c index 98ab01f..fca3310

[PATCH/RESEND v2 1/5] z3fold: make pages_nr atomic

This patch converts pages_nr per-pool counter to atomic64_t. Signed-off-by: Vitaly Wool <vitalyw...@gmail.com> --- mm/z3fold.c | 20 +--- 1 file changed, 9 insertions(+), 11 deletions(-) diff --git a/mm/z3fold.c b/mm/z3fold.c index 207e5dd..2273789 100644 --- a/mm/z3fold.c

[PATCH/RESEND v2 3/5] z3fold: extend compaction function

, using BIG_CHUNK_GAP define as a threshold for middle chunk to be worth moving. Signed-off-by: Vitaly Wool --- mm/z3fold.c | 26 +- 1 file changed, 25 insertions(+), 1 deletion(-) diff --git a/mm/z3fold.c b/mm/z3fold.c index 98ab01f..fca3310 100644 --- a/mm/z3fold.c +++ b

[PATCH/RESEND v2 1/5] z3fold: make pages_nr atomic

This patch converts pages_nr per-pool counter to atomic64_t. Signed-off-by: Vitaly Wool --- mm/z3fold.c | 20 +--- 1 file changed, 9 insertions(+), 11 deletions(-) diff --git a/mm/z3fold.c b/mm/z3fold.c index 207e5dd..2273789 100644 --- a/mm/z3fold.c +++ b/mm/z3fold.c @@ -80,7

[PATCH/RESEND v2 5/5] z3fold: add kref refcounting

With both coming and already present locking optimizations, introducing kref to reference-count z3fold objects is the right thing to do. Moreover, it makes buddied list no longer necessary, and allows for a simpler handling of headless pages. Signed-off-by: Vitaly Wool <vitalyw...@gmail.

[PATCH/RESEND v2 5/5] z3fold: add kref refcounting

With both coming and already present locking optimizations, introducing kref to reference-count z3fold objects is the right thing to do. Moreover, it makes buddied list no longer necessary, and allows for a simpler handling of headless pages. Signed-off-by: Vitaly Wool --- mm/z3fold.c | 145

[PATCH/RESEND v2 2/5] z3fold: fix header size related issues

of num_free_chunks() and the address to move the middle chunk to in case of in-page compaction in z3fold_compact_page(). Signed-off-by: Vitaly Wool <vitalyw...@gmail.com> --- mm/z3fold.c | 114 ++-- 1 file changed, 64 insertions(+), 50 del

[PATCH/RESEND v2 2/5] z3fold: fix header size related issues

of num_free_chunks() and the address to move the middle chunk to in case of in-page compaction in z3fold_compact_page(). Signed-off-by: Vitaly Wool --- mm/z3fold.c | 114 ++-- 1 file changed, 64 insertions(+), 50 deletions(-) diff --git a/mm/z3fold.c

[PATCH/RESEND v2 0/5] z3fold optimizations and fixes

This is a consolidation of z3fold optimizations and fixes done so far, revised after comments from Dan ([1], [2], [3], [4]). The coming patches are to be applied on top of the following commit: Author: zhong jiang Date: Tue Dec 20 11:53:40 2016 +1100 mm/z3fold.c:

[PATCH/RESEND v2 0/5] z3fold optimizations and fixes

This is a consolidation of z3fold optimizations and fixes done so far, revised after comments from Dan ([1], [2], [3], [4]). The coming patches are to be applied on top of the following commit: Author: zhong jiang Date: Tue Dec 20 11:53:40 2016 +1100 mm/z3fold.c: limit first_num to the

Re: [PATCH/RESEND 5/5] z3fold: add kref refcounting

On Wed, Jan 4, 2017 at 7:42 PM, Dan Streetman <ddstr...@ieee.org> wrote: > On Sun, Dec 25, 2016 at 7:40 PM, Vitaly Wool <vitalyw...@gmail.com> wrote: >> With both coming and already present locking optimizations, >> introducing kref to reference-count z3fold objects

Re: [PATCH/RESEND 5/5] z3fold: add kref refcounting

On Wed, Jan 4, 2017 at 7:42 PM, Dan Streetman wrote: > On Sun, Dec 25, 2016 at 7:40 PM, Vitaly Wool wrote: >> With both coming and already present locking optimizations, >> introducing kref to reference-count z3fold objects is the right >> thing to do. Moreover, it makes b

Re: [PATCH/RESEND 2/5] mm/z3fold.c: extend compaction function

On Wed, Jan 4, 2017 at 4:43 PM, Dan Streetman wrote: >> static int z3fold_compact_page(struct z3fold_header *zhdr) >> { >> struct page *page = virt_to_page(zhdr); >> - void *beg = zhdr; >> + int ret = 0; > > I still don't understand why you're adding ret

Re: [PATCH/RESEND 2/5] mm/z3fold.c: extend compaction function

On Wed, Jan 4, 2017 at 4:43 PM, Dan Streetman wrote: >> static int z3fold_compact_page(struct z3fold_header *zhdr) >> { >> struct page *page = virt_to_page(zhdr); >> - void *beg = zhdr; >> + int ret = 0; > > I still don't understand why you're adding ret and using goto.

[PATCH/RESEND 5/5] z3fold: add kref refcounting

With both coming and already present locking optimizations, introducing kref to reference-count z3fold objects is the right thing to do. Moreover, it makes buddied list no longer necessary, and allows for a simpler handling of headless pages. Signed-off-by: Vitaly Wool <vitalyw...@gmail.

[PATCH/RESEND 5/5] z3fold: add kref refcounting

With both coming and already present locking optimizations, introducing kref to reference-count z3fold objects is the right thing to do. Moreover, it makes buddied list no longer necessary, and allows for a simpler handling of headless pages. Signed-off-by: Vitaly Wool --- mm/z3fold.c | 137

[PATCH/RESEND 4/5] z3fold: fix header size related issues

of num_free_chunks() and the address to move the middle chunk to in case of in-page compaction in z3fold_compact_page(). Signed-off-by: Vitaly Wool <vitalyw...@gmail.com> --- mm/z3fold.c | 161 1 file changed, 87 insertions(+), 74 del

[PATCH/RESEND 4/5] z3fold: fix header size related issues

of num_free_chunks() and the address to move the middle chunk to in case of in-page compaction in z3fold_compact_page(). Signed-off-by: Vitaly Wool --- mm/z3fold.c | 161 1 file changed, 87 insertions(+), 74 deletions(-) diff --git a/mm/z3fold.c

[PATCH/RESEND 3/5] z3fold: use per-page spinlock

implements raw spinlock-based per-page locking mechanism which is lightweight enough to normally fit ok into the z3fold header. Signed-off-by: Vitaly Wool <vitalyw...@gmail.com> --- mm/z3fold.c | 148 +++- 1 file changed, 106 insertions(

[PATCH/RESEND 3/5] z3fold: use per-page spinlock

implements raw spinlock-based per-page locking mechanism which is lightweight enough to normally fit ok into the z3fold header. Signed-off-by: Vitaly Wool --- mm/z3fold.c | 148 +++- 1 file changed, 106 insertions(+), 42 deletions(-) diff --git

[PATCH/RESEND 2/5] mm/z3fold.c: extend compaction function

to less actual page allocations on hot path due to denser in-page allocation). This patch adds the relevant code, using BIG_CHUNK_GAP define as a threshold for middle chunk to be worth moving. Signed-off-by: Vitaly Wool <vitalyw...@gmail.com> --- mm/z3fold.

[PATCH/RESEND 2/5] mm/z3fold.c: extend compaction function

to less actual page allocations on hot path due to denser in-page allocation). This patch adds the relevant code, using BIG_CHUNK_GAP define as a threshold for middle chunk to be worth moving. Signed-off-by: Vitaly Wool --- mm/z3fold.c | 60

[PATCH/RESEND 2/5] mm/z3fold.c: extend compaction function

to less actual page allocations on hot path due to denser in-page allocation). This patch adds the relevant code, using BIG_CHUNK_GAP define as a threshold for middle chunk to be worth moving. Signed-off-by: Vitaly Wool <vitalyw...@gmail.com> --- mm/z3fold.

[PATCH/RESEND 2/5] mm/z3fold.c: extend compaction function

to less actual page allocations on hot path due to denser in-page allocation). This patch adds the relevant code, using BIG_CHUNK_GAP define as a threshold for middle chunk to be worth moving. Signed-off-by: Vitaly Wool --- mm/z3fold.c | 60

[PATCH/RESEND 1/5] mm/z3fold.c: make pages_nr atomic

Convert pages_nr per-pool counter to atomic64_t so that we won't have to care about locking for reading/updating it. Signed-off-by: Vitaly Wool <vitalyw...@gmail.com> --- mm/z3fold.c | 20 +--- 1 file changed, 9 insertions(+), 11 deletions(-) diff --git a/mm/z3fold.

[PATCH/RESEND 1/5] mm/z3fold.c: make pages_nr atomic

Convert pages_nr per-pool counter to atomic64_t so that we won't have to care about locking for reading/updating it. Signed-off-by: Vitaly Wool --- mm/z3fold.c | 20 +--- 1 file changed, 9 insertions(+), 11 deletions(-) diff --git a/mm/z3fold.c b/mm/z3fold.c index 207e5dd

<    1   2   3   4   5   >