Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]
Groups > linux.kernel > #1280444 > unrolled thread
| Started by | Jaegeuk Kim <jaegeuk@kernel.org> |
|---|---|
| First post | 2015-12-01 02:00 +0100 |
| Last post | 2015-12-01 03:20 +0100 |
| Articles | 5 — 3 participants |
Back to article view | Back to linux.kernel
[PATCH] f2fs: avoid deadlock in f2fs_shrink_extent_tree Jaegeuk Kim <jaegeuk@kernel.org> - 2015-12-01 02:00 +0100
Re: [PATCH] f2fs: avoid deadlock in f2fs_shrink_extent_tree kbuild test robot <lkp@intel.com> - 2015-12-01 02:40 +0100
Re: [PATCH v2] f2fs: avoid deadlock in f2fs_shrink_extent_tree Jaegeuk Kim <jaegeuk@kernel.org> - 2015-12-01 02:50 +0100
Re: [f2fs-dev] [PATCH v2] f2fs: avoid deadlock in f2fs_shrink_extent_tree Chao Yu <chao@kernel.org> - 2015-12-01 15:50 +0100
Re: [PATCH] f2fs: avoid deadlock in f2fs_shrink_extent_tree kbuild test robot <lkp@intel.com> - 2015-12-01 03:20 +0100
| From | Jaegeuk Kim <jaegeuk@kernel.org> |
|---|---|
| Date | 2015-12-01 02:00 +0100 |
| Subject | [PATCH] f2fs: avoid deadlock in f2fs_shrink_extent_tree |
| Message-ID | <qAHsu-2uI-5@gated-at.bofh.it> |
While handling extent trees, we can enter into a reclaiming path anytime.
If it tries to release some extent nodes in the same extent tree,
write_lock(&et->lock) would be hanged.
In order to avoid the deadlock, we can just skip it.
Note that, if it is an unreferenced tree, we should get write_lock(&et->lock)
successfully and release all of therein nodes.
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
---
fs/f2fs/extent_cache.c | 7 ++++---
1 file changed, 4 insertions(+), 3 deletions(-)
diff --git a/fs/f2fs/extent_cache.c b/fs/f2fs/extent_cache.c
index 7ddba81..21b1b8e 100644
--- a/fs/f2fs/extent_cache.c
+++ b/fs/f2fs/extent_cache.c
@@ -615,9 +615,10 @@ unsigned int f2fs_shrink_extent_tree(struct f2fs_sb_info *sbi, int nr_shrink)
for (i = 0; i < found; i++) {
struct extent_tree *et = treevec[i];
- write_lock(&et->lock);
- node_cnt += __free_extent_tree(sbi, et, false);
- write_unlock(&et->lock);
+ if (down_write_trylock(&et->lock)) {
+ node_cnt += __free_extent_tree(sbi, et, false);
+ write_unlock(&et->lock);
+ }
if (node_cnt + tree_cnt >= nr_shrink)
goto unlock_out;
--
2.4.9 (Apple Git-60)
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
[toc] | [next] | [standalone]
| From | kbuild test robot <lkp@intel.com> |
|---|---|
| Date | 2015-12-01 02:40 +0100 |
| Message-ID | <qAI5b-2YM-1@gated-at.bofh.it> |
| In reply to | #1280444 |
[Multipart message — attachments visible in raw view] — view raw
Hi Jaegeuk,
[auto build test WARNING on: f2fs/dev]
[also build test WARNING on: v4.4-rc3 next-20151127]
url: https://github.com/0day-ci/linux/commits/Jaegeuk-Kim/f2fs-avoid-deadlock-in-f2fs_shrink_extent_tree/20151201-085538
base: https://git.kernel.org/pub/scm/linux/kernel/git/jaegeuk/f2fs dev
config: i386-randconfig-i1-201548 (attached as .config)
reproduce:
# save the attached .config to linux build tree
make ARCH=i386
All warnings (new ones prefixed by >>):
fs/f2fs/extent_cache.c: In function 'f2fs_shrink_extent_tree':
>> fs/f2fs/extent_cache.c:618:27: warning: passing argument 1 of 'down_write_trylock' from incompatible pointer type [-Wincompatible-pointer-types]
if (down_write_trylock(&et->lock)) {
^
In file included from include/linux/fs.h:20:0,
from fs/f2fs/extent_cache.c:14:
include/linux/rwsem.h:123:12: note: expected 'struct rw_semaphore *' but argument is of type 'rwlock_t * {aka struct <anonymous> *}'
extern int down_write_trylock(struct rw_semaphore *sem);
^
vim +/down_write_trylock +618 fs/f2fs/extent_cache.c
602 }
603 spin_unlock(&sbi->extent_lock);
604
605 /*
606 * reset ino for searching victims from beginning of global extent tree.
607 */
608 ino = F2FS_ROOT_INO(sbi);
609
610 while ((found = radix_tree_gang_lookup(root,
611 (void **)treevec, ino, EXT_TREE_VEC_SIZE))) {
612 unsigned i;
613
614 ino = treevec[found - 1]->ino + 1;
615 for (i = 0; i < found; i++) {
616 struct extent_tree *et = treevec[i];
617
> 618 if (down_write_trylock(&et->lock)) {
619 node_cnt += __free_extent_tree(sbi, et, false);
620 write_unlock(&et->lock);
621 }
622
623 if (node_cnt + tree_cnt >= nr_shrink)
624 goto unlock_out;
625 }
626 }
---
0-DAY kernel test infrastructure Open Source Technology Center
https://lists.01.org/pipermail/kbuild-all Intel Corporation
[toc] | [prev] | [next] | [standalone]
| From | Jaegeuk Kim <jaegeuk@kernel.org> |
|---|---|
| Date | 2015-12-01 02:50 +0100 |
| Subject | Re: [PATCH v2] f2fs: avoid deadlock in f2fs_shrink_extent_tree |
| Message-ID | <qAIeR-32d-13@gated-at.bofh.it> |
| In reply to | #1280444 |
Change log from v1:
o fix bug
From 8cd63b5e91daeeccc1f158425c74de0f462c96af Mon Sep 17 00:00:00 2001
From: Jaegeuk Kim <jaegeuk@kernel.org>
Date: Mon, 30 Nov 2015 16:26:44 -0800
Subject: [PATCH] f2fs: avoid deadlock in f2fs_shrink_extent_tree
While handling extent trees, we can enter into a reclaiming path anytime.
If it tries to release some extent nodes in the same extent tree,
write_lock(&et->lock) would be hanged.
In order to avoid the deadlock, we can just skip it.
Note that, if it is an unreferenced tree, we should get write_lock(&et->lock)
successfully and release all of therein nodes.
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
---
fs/f2fs/extent_cache.c | 7 ++++---
1 file changed, 4 insertions(+), 3 deletions(-)
diff --git a/fs/f2fs/extent_cache.c b/fs/f2fs/extent_cache.c
index 7ddba81..de063f2 100644
--- a/fs/f2fs/extent_cache.c
+++ b/fs/f2fs/extent_cache.c
@@ -615,9 +615,10 @@ unsigned int f2fs_shrink_extent_tree(struct f2fs_sb_info *sbi, int nr_shrink)
for (i = 0; i < found; i++) {
struct extent_tree *et = treevec[i];
- write_lock(&et->lock);
- node_cnt += __free_extent_tree(sbi, et, false);
- write_unlock(&et->lock);
+ if (write_trylock(&et->lock)) {
+ node_cnt += __free_extent_tree(sbi, et, false);
+ write_unlock(&et->lock);
+ }
if (node_cnt + tree_cnt >= nr_shrink)
goto unlock_out;
--
2.4.9 (Apple Git-60)
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
[toc] | [prev] | [next] | [standalone]
| From | Chao Yu <chao@kernel.org> |
|---|---|
| Date | 2015-12-01 15:50 +0100 |
| Subject | Re: [f2fs-dev] [PATCH v2] f2fs: avoid deadlock in f2fs_shrink_extent_tree |
| Message-ID | <qAUpH-2rO-13@gated-at.bofh.it> |
| In reply to | #1280480 |
On 12/1/15 9:46 AM, Jaegeuk Kim wrote: > Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org> Reviewed-by: Chao Yu <chao2.yu@samsung.com> -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
[toc] | [prev] | [next] | [standalone]
| From | kbuild test robot <lkp@intel.com> |
|---|---|
| Date | 2015-12-01 03:20 +0100 |
| Message-ID | <qAIHT-3rX-3@gated-at.bofh.it> |
| In reply to | #1280444 |
[Multipart message — attachments visible in raw view] — view raw
Hi Jaegeuk,
[auto build test WARNING on: f2fs/dev]
[also build test WARNING on: v4.4-rc3 next-20151127]
url: https://github.com/0day-ci/linux/commits/Jaegeuk-Kim/f2fs-avoid-deadlock-in-f2fs_shrink_extent_tree/20151201-085538
base: https://git.kernel.org/pub/scm/linux/kernel/git/jaegeuk/f2fs dev
config: x86_64-randconfig-s2-12010958 (attached as .config)
reproduce:
# save the attached .config to linux build tree
make ARCH=x86_64
All warnings (new ones prefixed by >>):
In file included from include/linux/linkage.h:4:0,
from include/linux/fs.h:4,
from fs/f2fs/extent_cache.c:14:
fs/f2fs/extent_cache.c: In function 'f2fs_shrink_extent_tree':
fs/f2fs/extent_cache.c:618:27: warning: passing argument 1 of 'down_write_trylock' from incompatible pointer type [-Wincompatible-pointer-types]
if (down_write_trylock(&et->lock)) {
^
include/linux/compiler.h:147:28: note: in definition of macro '__trace_if'
if (__builtin_constant_p((cond)) ? !!(cond) : \
^
>> fs/f2fs/extent_cache.c:618:4: note: in expansion of macro 'if'
if (down_write_trylock(&et->lock)) {
^
In file included from include/linux/fs.h:20:0,
from fs/f2fs/extent_cache.c:14:
include/linux/rwsem.h:123:12: note: expected 'struct rw_semaphore *' but argument is of type 'rwlock_t * {aka struct <anonymous> *}'
extern int down_write_trylock(struct rw_semaphore *sem);
^
In file included from include/linux/linkage.h:4:0,
from include/linux/fs.h:4,
from fs/f2fs/extent_cache.c:14:
fs/f2fs/extent_cache.c:618:27: warning: passing argument 1 of 'down_write_trylock' from incompatible pointer type [-Wincompatible-pointer-types]
if (down_write_trylock(&et->lock)) {
^
include/linux/compiler.h:147:40: note: in definition of macro '__trace_if'
if (__builtin_constant_p((cond)) ? !!(cond) : \
^
>> fs/f2fs/extent_cache.c:618:4: note: in expansion of macro 'if'
if (down_write_trylock(&et->lock)) {
^
In file included from include/linux/fs.h:20:0,
from fs/f2fs/extent_cache.c:14:
include/linux/rwsem.h:123:12: note: expected 'struct rw_semaphore *' but argument is of type 'rwlock_t * {aka struct <anonymous> *}'
extern int down_write_trylock(struct rw_semaphore *sem);
^
In file included from include/linux/linkage.h:4:0,
from include/linux/fs.h:4,
from fs/f2fs/extent_cache.c:14:
fs/f2fs/extent_cache.c:618:27: warning: passing argument 1 of 'down_write_trylock' from incompatible pointer type [-Wincompatible-pointer-types]
if (down_write_trylock(&et->lock)) {
^
include/linux/compiler.h:158:16: note: in definition of macro '__trace_if'
______r = !!(cond); \
^
>> fs/f2fs/extent_cache.c:618:4: note: in expansion of macro 'if'
if (down_write_trylock(&et->lock)) {
^
In file included from include/linux/fs.h:20:0,
from fs/f2fs/extent_cache.c:14:
include/linux/rwsem.h:123:12: note: expected 'struct rw_semaphore *' but argument is of type 'rwlock_t * {aka struct <anonymous> *}'
extern int down_write_trylock(struct rw_semaphore *sem);
^
vim +/if +618 fs/f2fs/extent_cache.c
602 }
603 spin_unlock(&sbi->extent_lock);
604
605 /*
606 * reset ino for searching victims from beginning of global extent tree.
607 */
608 ino = F2FS_ROOT_INO(sbi);
609
610 while ((found = radix_tree_gang_lookup(root,
611 (void **)treevec, ino, EXT_TREE_VEC_SIZE))) {
612 unsigned i;
613
614 ino = treevec[found - 1]->ino + 1;
615 for (i = 0; i < found; i++) {
616 struct extent_tree *et = treevec[i];
617
> 618 if (down_write_trylock(&et->lock)) {
619 node_cnt += __free_extent_tree(sbi, et, false);
620 write_unlock(&et->lock);
621 }
622
623 if (node_cnt + tree_cnt >= nr_shrink)
624 goto unlock_out;
625 }
626 }
---
0-DAY kernel test infrastructure Open Source Technology Center
https://lists.01.org/pipermail/kbuild-all Intel Corporation
[toc] | [prev] | [standalone]
Back to top | Article view | linux.kernel
csiph-web