- Kernel - mailweb.openeuler.org

[PATCH] nvme: Assign subsys instance from first ctrl upstream 733e4b69
by Chen Taotao 06 Jun '22

06 Jun '22

The namespace disk names must be unique for the lifetime of the subsystem. This was accomplished by using their parent subsystems' instances which were allocated independently from the controllers connected to that subsystem. This allowed name prefixes assigned to namespaces to match a controller from an unrelated subsystem, and has created confusion among users examining device nodes. Ensure a namespace's subsystem instance never clashes with a controller instance of another subsystem by transferring the instance ownership to the parent subsystem from the first controller discovered in that subsystem. Reviewed-by: Logan Gunthorpe <logang(a)deltatee.com> Signed-off-by: Keith Busch <kbusch(a)kernel.org> Reviewed-by: Christoph Hellwig <hch(a)lst.de> Reviewed-by: Minwoo Im <minwoo.im(a)samsung.com> Reviewed-by: Hannes Reinecke <hare(a)suse.com> Reviewed-off-by: Sagi Grimberg <sagi(a)grimberg.me> Signed-off-by: Sagi Grimberg <sagi(a)grimberg.me> --- drivers/nvme/host/core.c | 21 ++++++++++----------- 1 file changed, 10 insertions(+), 11 deletions(-) diff --git a/drivers/nvme/host/core.c b/drivers/nvme/host/core.c index 8c760493354f..576da5820a57 100644 --- a/drivers/nvme/host/core.c +++ b/drivers/nvme/host/core.c @@ -88,7 +88,6 @@ EXPORT_SYMBOL_GPL(nvme_reset_wq); struct workqueue_struct *nvme_delete_wq; EXPORT_SYMBOL_GPL(nvme_delete_wq); -static DEFINE_IDA(nvme_subsystems_ida); static LIST_HEAD(nvme_subsystems); static DEFINE_MUTEX(nvme_subsystems_lock); @@ -2308,7 +2307,8 @@ static void nvme_release_subsystem(struct device *dev) struct nvme_subsystem *subsys = container_of(dev, struct nvme_subsystem, dev); - ida_simple_remove(&nvme_subsystems_ida, subsys->instance); + if (subsys->instance >= 0) + ida_simple_remove(&nvme_instance_ida, subsys->instance); kfree(subsys); } @@ -2419,12 +2419,8 @@ static int nvme_init_subsystem(struct nvme_ctrl *ctrl, struct nvme_id_ctrl *id) subsys = kzalloc(sizeof(*subsys), GFP_KERNEL); if (!subsys) return -ENOMEM; - ret = ida_simple_get(&nvme_subsystems_ida, 0, 0, GFP_KERNEL); - if (ret < 0) { - kfree(subsys); - return ret; - } - subsys->instance = ret; + + subsys->instance = -1; mutex_init(&subsys->lock); kref_init(&subsys->ref); INIT_LIST_HEAD(&subsys->ctrls); @@ -2439,7 +2435,7 @@ static int nvme_init_subsystem(struct nvme_ctrl *ctrl, struct nvme_id_ctrl *id) subsys->dev.class = nvme_subsys_class; subsys->dev.release = nvme_release_subsystem; subsys->dev.groups = nvme_subsys_attrs_groups; - dev_set_name(&subsys->dev, "nvme-subsys%d", subsys->instance); + dev_set_name(&subsys->dev, "nvme-subsys%d", ctrl->instance); device_initialize(&subsys->dev); mutex_lock(&nvme_subsystems_lock); @@ -2472,6 +2468,8 @@ static int nvme_init_subsystem(struct nvme_ctrl *ctrl, struct nvme_id_ctrl *id) list_add_tail(&subsys->entry, &nvme_subsystems); } + if (!found) + subsys->instance = ctrl->instance; ctrl->subsys = subsys; mutex_unlock(&nvme_subsystems_lock); @@ -3749,7 +3747,9 @@ static void nvme_free_ctrl(struct device *dev) container_of(dev, struct nvme_ctrl, ctrl_device); struct nvme_subsystem *subsys = ctrl->subsys; - ida_simple_remove(&nvme_instance_ida, ctrl->instance); + if (subsys && ctrl->instance != subsys->instance) + ida_simple_remove(&nvme_instance_ida, ctrl->instance); + kfree(ctrl->effects); nvme_mpath_uninit(ctrl); __free_page(ctrl->discard_page); @@ -4009,7 +4009,6 @@ int __init nvme_core_init(void) void nvme_core_exit(void) { - ida_destroy(&nvme_subsystems_ida); class_destroy(nvme_subsys_class); class_destroy(nvme_class); unregister_chrdev_region(nvme_chr_devt, NVME_MINORS); -- 2.27.0

1 0

Re: [PATCH OLK-5.10 1/4] eulerfs: add socket_num to eufs_sb_info
by Yu Kuai 06 Jun '22

06 Jun '22

在 2022/06/02 19:24, Gou Hao 写道: > > 在 2022/6/2 17:10, Yu Kuai 写道: >> 在 2022/06/02 10:50, Gou Hao 写道: >>> uniontech inclusion >>> category: bugfix >>> bugzilla: NA >>> CVE: NA >>> >>> ---------------------- >>> >>> We should not get the socket num in init_eufs_fs, becuse >>> system usually supports cpu hotplug,so we should get socket >>> num in dep_int. >> Hi, >> >> What if cpu is offline during mount, and online afterwards? >> > Well, in this case, we have to add a monitoring mechanism to > > dynamically adjust the number of threads according to the > > online status of the CPU. > > Or there are other better ways ... I prefer to keep the current implementation for now. Thanks, Kuai >> Thanks, >> Kuai >>> >>> Socket num is not necessarily the same when file system >>> mount,so it shoule be eufs_sb_info's member. >>> >>> Signed-off-by: Gou Hao <gouhao(a)uniontech.com> >>> --- >>> fs/eulerfs/dep.c | 28 ++++++++++++++++++++++------ >>> fs/eulerfs/euler.h | 2 -- >>> fs/eulerfs/euler_def.h | 1 + >>> fs/eulerfs/super.c | 15 +-------------- >>> 4 files changed, 24 insertions(+), 22 deletions(-) >>

2 1

[openEuler-22.09 00/34] bcache backport for openEuler-22.09 (up to Linux v5.19-rc1)
by Coly Li 04 Jun '22

04 Jun '22

This series is bcache backport up to Linux v5.19. The referenced bug is, https://gitee.com/openeuler/kernel/issues/I59A5L#note_10515416 All the patches are tested on Intel x86-64 platform, but not tested on ARM64 plantform yet. Coly Li --- Arnd Bergmann (1): md: bcache: avoid -Wempty-body warnings Bhaskar Chowdhury (1): md: bcache: Trivial typo fixes in the file journal.c Chao Yu (1): bcache: fix error info in register_bcache() Christoph Hellwig (1): bcache: remove PTR_CACHE Coly Li (10): bcache: fix a regression of code compiling failure in debug.c bcache: remove bcache device self-defined readahead bcache: avoid oversized read request in cache missing code path bcache: move uapi header bcache.h to bcache code directory bcache: improve multithreaded bch_btree_check() bcache: improve multithreaded bch_sectors_dirty_init() bcache: remove incremental dirty sector counting for bch_sectors_dirty_init() bcache: avoid journal no-space deadlock by reserving 1 journal bucket bcache: memset on stack variables in bch_btree_check() and bch_sectors_dirty_init() bcache: avoid unnecessary soft lockup in kworker update_writeback_rate() Ding Senjie (1): md: bcache: Fix spelling of 'acquire' Dongsheng Yang (1): bcache: fix race between setting bdev state to none and new write request direct to backing Greg Kroah-Hartman (1): bcache: use default_groups in kobj_type Gustavo A. R. Silva (1): bcache: Use 64-bit arithmetic instead of 32-bit Jia-Ju Bai (1): md: bcache: check the return value of kzalloc() in detached_dev_do_request() Joe Perches (1): bcache: Avoid comma separated statements Kai Krakow (1): bcache: Fix register_device_aync typo Lin Feng (2): bcache: move calc_cached_dev_sectors to proper place on backing device detach bcache: fix NULL pointer reference in cached_dev_detach_finish Ming Lei (1): bcache: don't pass BIOSET_NEED_BVECS for the 'bio_set' embedded in 'cache_set' Mingzhe Zou (2): bcache: fixup bcache_dev_sectors_dirty_add() multithreaded CPU false sharing bcache: fixup multiple threads crash Qing Wang (1): bcache: replace snprintf in show functions with sysfs_emit Yang Li (1): bcache: use NULL instead of using plain integer as pointer Yi Li (2): bcache:remove a superfluous check in register_bcache bcache: set pdev_set_uuid before scond loop iteration YueHaibing (1): lib: crc64: fix kernel-doc warning Zheng Yongjun (1): md/bcache: convert comma to semicolon Zhiqiang Liu (1): bcache: reduce redundant code in bch_cached_dev_run() dongdong tao (1): bcache: consider the fragmentation when update the writeback rate drivers/md/bcache/alloc.c | 5 +- drivers/md/bcache/bcache.h | 25 ++- .../md/bcache/bcache_ondisk.h | 0 drivers/md/bcache/bset.c | 12 +- drivers/md/bcache/bset.h | 2 +- drivers/md/bcache/btree.c | 67 +++--- drivers/md/bcache/btree.h | 2 +- drivers/md/bcache/debug.c | 2 +- drivers/md/bcache/extents.c | 4 +- drivers/md/bcache/features.c | 4 +- drivers/md/bcache/features.h | 3 +- drivers/md/bcache/io.c | 4 +- drivers/md/bcache/journal.c | 37 +++- drivers/md/bcache/journal.h | 2 + drivers/md/bcache/request.c | 25 +-- drivers/md/bcache/stats.c | 17 +- drivers/md/bcache/stats.h | 1 - drivers/md/bcache/super.c | 63 +++--- drivers/md/bcache/sysfs.c | 50 +++-- drivers/md/bcache/sysfs.h | 20 +- drivers/md/bcache/util.h | 19 +- drivers/md/bcache/writeback.c | 198 +++++++++++------- drivers/md/bcache/writeback.h | 6 +- lib/crc64.c | 2 +- 24 files changed, 324 insertions(+), 246 deletions(-) rename include/uapi/linux/bcache.h => drivers/md/bcache/bcache_ondisk.h (100%) -- 2.35.3

1 34

[openEuler-22.03-LTS 00/34] bcache backport for openEuler-22.03-LTS (up to Linux v5.19-rc1)
by Coly Li 04 Jun '22

04 Jun '22

This series is bcache backport up to Linux v5.19. The referenced bug is, https://gitee.com/openeuler/kernel/issues/I59A5L#note_10515416 All the patches are tested on Intel x86-64 platform, but not tested on ARM64 plantform yet. Coly Li --- Arnd Bergmann (1): md: bcache: avoid -Wempty-body warnings Bhaskar Chowdhury (1): md: bcache: Trivial typo fixes in the file journal.c Chao Yu (1): bcache: fix error info in register_bcache() Christoph Hellwig (1): bcache: remove PTR_CACHE Coly Li (10): bcache: fix a regression of code compiling failure in debug.c bcache: remove bcache device self-defined readahead bcache: avoid oversized read request in cache missing code path bcache: move uapi header bcache.h to bcache code directory bcache: improve multithreaded bch_btree_check() bcache: improve multithreaded bch_sectors_dirty_init() bcache: remove incremental dirty sector counting for bch_sectors_dirty_init() bcache: avoid journal no-space deadlock by reserving 1 journal bucket bcache: memset on stack variables in bch_btree_check() and bch_sectors_dirty_init() bcache: avoid unnecessary soft lockup in kworker update_writeback_rate() Ding Senjie (1): md: bcache: Fix spelling of 'acquire' Dongsheng Yang (1): bcache: fix race between setting bdev state to none and new write request direct to backing Greg Kroah-Hartman (1): bcache: use default_groups in kobj_type Gustavo A. R. Silva (1): bcache: Use 64-bit arithmetic instead of 32-bit Jia-Ju Bai (1): md: bcache: check the return value of kzalloc() in detached_dev_do_request() Joe Perches (1): bcache: Avoid comma separated statements Kai Krakow (1): bcache: Fix register_device_aync typo Lin Feng (2): bcache: move calc_cached_dev_sectors to proper place on backing device detach bcache: fix NULL pointer reference in cached_dev_detach_finish Ming Lei (1): bcache: don't pass BIOSET_NEED_BVECS for the 'bio_set' embedded in 'cache_set' Mingzhe Zou (2): bcache: fixup bcache_dev_sectors_dirty_add() multithreaded CPU false sharing bcache: fixup multiple threads crash Qing Wang (1): bcache: replace snprintf in show functions with sysfs_emit Yang Li (1): bcache: use NULL instead of using plain integer as pointer Yi Li (2): bcache:remove a superfluous check in register_bcache bcache: set pdev_set_uuid before scond loop iteration YueHaibing (1): lib: crc64: fix kernel-doc warning Zheng Yongjun (1): md/bcache: convert comma to semicolon Zhiqiang Liu (1): bcache: reduce redundant code in bch_cached_dev_run() dongdong tao (1): bcache: consider the fragmentation when update the writeback rate drivers/md/bcache/alloc.c | 5 +- drivers/md/bcache/bcache.h | 25 ++- .../md/bcache/bcache_ondisk.h | 0 drivers/md/bcache/bset.c | 12 +- drivers/md/bcache/bset.h | 2 +- drivers/md/bcache/btree.c | 67 +++--- drivers/md/bcache/btree.h | 2 +- drivers/md/bcache/debug.c | 2 +- drivers/md/bcache/extents.c | 4 +- drivers/md/bcache/features.c | 4 +- drivers/md/bcache/features.h | 3 +- drivers/md/bcache/io.c | 4 +- drivers/md/bcache/journal.c | 37 +++- drivers/md/bcache/journal.h | 2 + drivers/md/bcache/request.c | 25 +-- drivers/md/bcache/stats.c | 17 +- drivers/md/bcache/stats.h | 1 - drivers/md/bcache/super.c | 63 +++--- drivers/md/bcache/sysfs.c | 50 +++-- drivers/md/bcache/sysfs.h | 20 +- drivers/md/bcache/util.h | 19 +- drivers/md/bcache/writeback.c | 198 +++++++++++------- drivers/md/bcache/writeback.h | 6 +- lib/crc64.c | 2 +- 24 files changed, 324 insertions(+), 246 deletions(-) rename include/uapi/linux/bcache.h => drivers/md/bcache/bcache_ondisk.h (100%) -- 2.35.3

1 34

[OLK-5.10 PATCH 00/34] bcache backport for OLK-5.10 (up to Linux v5.19-rc1)
by Coly Li 04 Jun '22

04 Jun '22

This series is bcache backport up to Linux v5.19. The referenced bug is, https://gitee.com/openeuler/kernel/issues/I59A5L#note_10515416 All the patches are tested on Intel x86-64 platform, but not tested on ARM64 plantform yet. Coly Li --- Arnd Bergmann (1): md: bcache: avoid -Wempty-body warnings Bhaskar Chowdhury (1): md: bcache: Trivial typo fixes in the file journal.c Chao Yu (1): bcache: fix error info in register_bcache() Christoph Hellwig (1): bcache: remove PTR_CACHE Coly Li (10): bcache: fix a regression of code compiling failure in debug.c bcache: remove bcache device self-defined readahead bcache: avoid oversized read request in cache missing code path bcache: move uapi header bcache.h to bcache code directory bcache: improve multithreaded bch_btree_check() bcache: improve multithreaded bch_sectors_dirty_init() bcache: remove incremental dirty sector counting for bch_sectors_dirty_init() bcache: avoid journal no-space deadlock by reserving 1 journal bucket bcache: memset on stack variables in bch_btree_check() and bch_sectors_dirty_init() bcache: avoid unnecessary soft lockup in kworker update_writeback_rate() Ding Senjie (1): md: bcache: Fix spelling of 'acquire' Dongsheng Yang (1): bcache: fix race between setting bdev state to none and new write request direct to backing Greg Kroah-Hartman (1): bcache: use default_groups in kobj_type Gustavo A. R. Silva (1): bcache: Use 64-bit arithmetic instead of 32-bit Jia-Ju Bai (1): md: bcache: check the return value of kzalloc() in detached_dev_do_request() Joe Perches (1): bcache: Avoid comma separated statements Kai Krakow (1): bcache: Fix register_device_aync typo Lin Feng (2): bcache: move calc_cached_dev_sectors to proper place on backing device detach bcache: fix NULL pointer reference in cached_dev_detach_finish Ming Lei (1): bcache: don't pass BIOSET_NEED_BVECS for the 'bio_set' embedded in 'cache_set' Mingzhe Zou (2): bcache: fixup bcache_dev_sectors_dirty_add() multithreaded CPU false sharing bcache: fixup multiple threads crash Qing Wang (1): bcache: replace snprintf in show functions with sysfs_emit Yang Li (1): bcache: use NULL instead of using plain integer as pointer Yi Li (2): bcache:remove a superfluous check in register_bcache bcache: set pdev_set_uuid before scond loop iteration YueHaibing (1): lib: crc64: fix kernel-doc warning Zheng Yongjun (1): md/bcache: convert comma to semicolon Zhiqiang Liu (1): bcache: reduce redundant code in bch_cached_dev_run() dongdong tao (1): bcache: consider the fragmentation when update the writeback rate drivers/md/bcache/alloc.c | 5 +- drivers/md/bcache/bcache.h | 25 ++- .../md/bcache/bcache_ondisk.h | 0 drivers/md/bcache/bset.c | 12 +- drivers/md/bcache/bset.h | 2 +- drivers/md/bcache/btree.c | 67 +++--- drivers/md/bcache/btree.h | 2 +- drivers/md/bcache/debug.c | 2 +- drivers/md/bcache/extents.c | 4 +- drivers/md/bcache/features.c | 4 +- drivers/md/bcache/features.h | 3 +- drivers/md/bcache/io.c | 4 +- drivers/md/bcache/journal.c | 37 +++- drivers/md/bcache/journal.h | 2 + drivers/md/bcache/request.c | 25 +-- drivers/md/bcache/stats.c | 17 +- drivers/md/bcache/stats.h | 1 - drivers/md/bcache/super.c | 63 +++--- drivers/md/bcache/sysfs.c | 50 +++-- drivers/md/bcache/sysfs.h | 20 +- drivers/md/bcache/util.h | 19 +- drivers/md/bcache/writeback.c | 198 +++++++++++------- drivers/md/bcache/writeback.h | 6 +- lib/crc64.c | 2 +- 24 files changed, 324 insertions(+), 246 deletions(-) rename include/uapi/linux/bcache.h => drivers/md/bcache/bcache_ondisk.h (100%) -- 2.35.3

1 34

[PATCH openEuler-1.0-LTS] xfs: show the proper user quota options
by tangbin 02 Jun '22

02 Jun '22

From: Kaixu Xia <kaixuxia(a)tencent.com> mainline inclusion from mainline-v5.11-rc6 commit 237d7887ae723af7d978e8b9a385fdff416f357b category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I5AG18 CVE: NA -------------------------------- The quota option 'usrquota' should be shown if both the XFS_UQUOTA_ACCT and XFS_UQUOTA_ENFD flags are set. The option 'uqnoenforce' should be shown when only the XFS_UQUOTA_ACCT flag is set. The current code logic seems wrong, Fix it and show proper options. Signed-off-by: Kaixu Xia <kaixuxia(a)tencent.com> Reviewed-by: Darrick J. Wong <darrick.wong(a)oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong(a)oracle.com> Signed-off-by: tangbin <tangbin(a)cmss.chinamobile.com> --- fs/xfs/xfs_super.c | 10 ++++++---- 1 file changed, 6 insertions(+), 4 deletions(-) diff --git a/fs/xfs/xfs_super.c b/fs/xfs/xfs_super.c index e126bc6cdeb8..5d8e3a4d2671 100644 --- a/fs/xfs/xfs_super.c +++ b/fs/xfs/xfs_super.c @@ -505,10 +505,12 @@ xfs_showargs( seq_printf(m, ",swidth=%d", (int)XFS_FSB_TO_BB(mp, mp->m_swidth)); - if (mp->m_qflags & (XFS_UQUOTA_ACCT|XFS_UQUOTA_ENFD)) - seq_puts(m, ",usrquota"); - else if (mp->m_qflags & XFS_UQUOTA_ACCT) - seq_puts(m, ",uqnoenforce"); + if (mp->m_qflags & XFS_UQUOTA_ACCT) { + if (mp->m_qflags & XFS_UQUOTA_ENFD) + seq_puts(m, ",usrquota"); + else + seq_puts(m, ",uqnoenforce"); + } if (mp->m_qflags & XFS_PQUOTA_ACCT) { if (mp->m_qflags & XFS_PQUOTA_ENFD) -- 2.18.4

1 0

[PATCH] xfs: show the proper user quota options
by tangbin 02 Jun '22

02 Jun '22

From: Kaixu Xia <kaixuxia(a)tencent.com> mainline inclusion from mainline-v5.11-rc6 commit 237d7887ae723af7d978e8b9a385fdff416f357b category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I5AG18 CVE: NA -------------------------------- The quota option 'usrquota' should be shown if both the XFS_UQUOTA_ACCT and XFS_UQUOTA_ENFD flags are set. The option 'uqnoenforce' should be shown when only the XFS_UQUOTA_ACCT flag is set. The current code logic seems wrong, Fix it and show proper options. Signed-off-by: Kaixu Xia <kaixuxia(a)tencent.com> Reviewed-by: Darrick J. Wong <darrick.wong(a)oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong(a)oracle.com> Signed-off-by: tangbin <tangbin(a)cmss.chinamobile.com> --- fs/xfs/xfs_super.c | 10 ++++++---- 1 file changed, 6 insertions(+), 4 deletions(-) diff --git a/fs/xfs/xfs_super.c b/fs/xfs/xfs_super.c index e126bc6cdeb8..5d8e3a4d2671 100644 --- a/fs/xfs/xfs_super.c +++ b/fs/xfs/xfs_super.c @@ -505,10 +505,12 @@ xfs_showargs( seq_printf(m, ",swidth=%d", (int)XFS_FSB_TO_BB(mp, mp->m_swidth)); - if (mp->m_qflags & (XFS_UQUOTA_ACCT|XFS_UQUOTA_ENFD)) - seq_puts(m, ",usrquota"); - else if (mp->m_qflags & XFS_UQUOTA_ACCT) - seq_puts(m, ",uqnoenforce"); + if (mp->m_qflags & XFS_UQUOTA_ACCT) { + if (mp->m_qflags & XFS_UQUOTA_ENFD) + seq_puts(m, ",usrquota"); + else + seq_puts(m, ",uqnoenforce"); + } if (mp->m_qflags & XFS_PQUOTA_ACCT) { if (mp->m_qflags & XFS_PQUOTA_ENFD) -- 2.18.4

1 0

Re: [PATCH OLK-5.10 1/4] eulerfs: add socket_num to eufs_sb_info
by Gou Hao 02 Jun '22

02 Jun '22

在 2022/6/2 17:10, Yu Kuai 写道: > 在 2022/06/02 10:50, Gou Hao 写道: >> uniontech inclusion >> category: bugfix >> bugzilla: NA >> CVE: NA >> >> ---------------------- >> >> We should not get the socket num in init_eufs_fs, becuse >> system usually supports cpu hotplug,so we should get socket >> num in dep_int. > Hi, > > What if cpu is offline during mount, and online afterwards? > Well, in this case, we have to add a monitoring mechanism to dynamically adjust the number of threads according to the online status of the CPU. Or there are other better ways ... > Thanks, > Kuai >> >> Socket num is not necessarily the same when file system >> mount,so it shoule be eufs_sb_info's member. >> >> Signed-off-by: Gou Hao <gouhao(a)uniontech.com> >> --- >> fs/eulerfs/dep.c | 28 ++++++++++++++++++++++------ >> fs/eulerfs/euler.h | 2 -- >> fs/eulerfs/euler_def.h | 1 + >> fs/eulerfs/super.c | 15 +-------------- >> 4 files changed, 24 insertions(+), 22 deletions(-) > -- thanks, Gou Hao

1 0

Re: [PATCH OLK-5.10 4/4] eulerfs: fix potential sbi->persisters free error
by Gou Hao 02 Jun '22

02 Jun '22

在 2022/6/2 17:15, Yu Kuai 写道: > 在 2022/06/02 10:51, Gou Hao 写道: >> uniontech inclusion >> category: bugfix >> bugzilla: NA >> CVE: NA >> >> ------------------- >> >> After alloc the sbi->persisters memory, dep_init >> will call dep_fini when error happened.Because >> sbi->persisters is not set to 0, kthread_stop >> may happened error. > > Looks good to me, just one language fixup: > > sbi->persisters is not set to 0 > -> dep_fini() can be called with sbi->persisters[] uninitialized, thus > kthread_stop() can be called with random value. > ok, thank you : ) > Thanks, > Kuai >> >> Signed-off-by: Gou Hao <gouhao(a)uniontech.com> >> --- >> fs/eulerfs/dep.c | 4 ++-- >> 1 file changed, 2 insertions(+), 2 deletions(-) >> >> diff --git a/fs/eulerfs/dep.c b/fs/eulerfs/dep.c >> index da826a18fa31..62fd1376f980 100644 >> --- a/fs/eulerfs/dep.c >> +++ b/fs/eulerfs/dep.c >> @@ -728,8 +728,8 @@ int dep_init(struct super_block *sb) >> node_num++; >> sbi->persister_num = node_num * persisters_per_socket; >> - >> - sbi->persisters = kmalloc(sizeof(struct task_struct *) * >> + >> + sbi->persisters = kzalloc(sizeof(struct task_struct *) * >> sbi->persister_num, GFP_KERNEL); >> if (!sbi->persisters) { >> err = -ENOMEM; >> > -- thanks, Gou Hao

1 0

[PATCH OLK-5.10 0/4] eulerfs: improve the creation of persister
by Gou Hao 02 Jun '22

02 Jun '22

Patch series "improve the creation of persister". The main modification is to create a persistent thread only when a NUMA node has an online CPU at least, instead of on all NUMA nodes. In the current code, the number of NUMA nodes is obtained when the module is loaded, and the same number of persistent threads are created when the file system is mounted. However, we should ignore NUMA nodes that do not have an online CPU. So we should decide how many threads to create according to the online CPU condition when initializing persist. Patch4 is a bugfix about dep_init(). Gou Hao (4): eulerfs: add socket_num to eufs_sb_info eulerfs: create persister on node when it has online cpu eulerfs: record persister num eulerfs: fix potential sbi->persisters free error fs/eulerfs/dep.c | 48 +++++++++++++++++++++++++++--------------- fs/eulerfs/euler.h | 2 -- fs/eulerfs/euler_def.h | 1 + fs/eulerfs/super.c | 21 ++++-------------- 4 files changed, 36 insertions(+), 36 deletions(-) -- 2.20.1

1 4