- Kernel - mailweb.openeuler.org

[PATCH 42/75] bcache: avoid flushing btree node in cache_set_flush() if io disabled
by Coly Li 08 Jan '20

08 Jan '20

mainline inclusion from mainline-5.3-rc1 commit e6dcbd3e6c91b7828cb305ec324eb7fd9bdea8a0 category: backport When cache_set_flush() is called for too many I/O errors detected on cache device and the cache set is retiring, inside the function it doesn't make sense to flushing cached btree nodes from c->btree_cache because CACHE_SET_IO_DISABLE is set on c->flags already and all I/Os onto cache device will be rejected. This patch checks in cache_set_flush() that whether CACHE_SET_IO_DISABLE is set. If yes, then avoids to flush the cached btree nodes to reduce more time and make cache set retiring more faster. Signed-off-by: Coly Li <colyli(a)suse.de> Signed-off-by: Jens Axboe <axboe(a)kernel.dk> --- drivers/md/bcache/super.c | 18 +++++++++++------- 1 file changed, 11 insertions(+), 7 deletions(-) diff --git a/drivers/md/bcache/super.c b/drivers/md/bcache/super.c index e5b29aa7e18a..4b7c5567576b 100644 --- a/drivers/md/bcache/super.c +++ b/drivers/md/bcache/super.c @@ -1551,13 +1551,17 @@ static void cache_set_flush(struct closure *cl) if (!IS_ERR_OR_NULL(c->root)) list_add(&c->root->list, &c->btree_cache); - /* Should skip this if we're unregistering because of an error */ - list_for_each_entry(b, &c->btree_cache, list) { - mutex_lock(&b->write_lock); - if (btree_node_dirty(b)) - __bch_btree_node_write(b, NULL); - mutex_unlock(&b->write_lock); - } + /* + * Avoid flushing cached nodes if cache set is retiring + * due to too many I/O errors detected. + */ + if (!test_bit(CACHE_SET_IO_DISABLE, &c->flags)) + list_for_each_entry(b, &c->btree_cache, list) { + mutex_lock(&b->write_lock); + if (btree_node_dirty(b)) + __bch_btree_node_write(b, NULL); + mutex_unlock(&b->write_lock); + } for_each_cache(ca, c, i) if (ca->alloc_thread) -- 2.16.4

1 0

[RESEND PATCH 00/75] bcache backport for openEuler-1.0 upto 20200104
by Coly Li 08 Jan '20

08 Jan '20

Hi openEler kernel maintainer, This is the bcache backport since Linux v4.19 to v5.5 for openEuler-10 kernel. All the patches are based on openEuler-1.0 commit cf975ca6a64a ("Document: add guideline to submitting patches to openEuler"). This patch set survives for 2+ hours smoking test on a 48x2 cores Taishan machine. Pleasse review and consider to pick them. Thanks in advance. Coly Li --- Andrea Righi (1): bcache: fix deadlock in bcache_allocator Christoph Hellwig (2): bcache: remove the extra cflags for request.o bcache: don't export symbols Colin Ian King (1): bcache: fix indentation issue, remove tabs on a hunk of code Coly Li (50): bcache: fix typo in code comments of closure_return_with_destructor() bcache: introduce force_wake_up_gc() bcache: option to automatically run gc thread after writeback bcache: add MODULE_DESCRIPTION information bcache: make cutoff_writeback and cutoff_writeback_sync tunable bcache: set writeback_percent in a flexible range bcache: not use hard coded memset size in bch_cache_accounting_clear() bcache: export backing_dev_name via sysfs bcache: export backing_dev_uuid via sysfs bcache: fix input integer overflow of congested threshold bcache: add sysfs_strtoul_bool() for setting bit-field variables bcache: use sysfs_strtoul_bool() to set bit-field variables bcache: fix input overflow to writeback_delay bcache: fix input overflow to journal_delay_ms bcache: fix input overflow to cache set io_error_limit bcache: move definition of 'int ret' out of macro read_bucket() bcache: add comments for kobj release callback routine bcache: add error check for calling register_bdev() bcache: Add comments for blkdev_put() in registration code path bcache: add comments for closure_fn to be called in closure_queue() bcache: improve bcache_reboot() bcache: don't set max writeback rate if gc is running bcache: fix return value error in bch_journal_read() bcache: avoid flushing btree node in cache_set_flush() if io disabled bcache: add io error counting in write_bdev_super_endio() bcache: remove unnecessary prefetch() in bset_search_tree() bcache: add return value check to bch_cached_dev_run() bcache: remove unncessary code in bch_btree_keys_init() bcache: more detailed error message to bcache_device_link() bcache: add more error message in bch_cached_dev_attach() bcache: improve error message in bch_cached_dev_run() bcache: remove "XXX:" comment line from run_cache_set() bcache: make bset_search_tree() be more understandable bcache: add pendings_cleanup to stop pending bcache device bcache: stop writeback kthread and kworker when bch_cached_dev_run() failed bcache: avoid a deadlock in bcache_reboot() bcache: acquire bch_register_lock later in cached_dev_detach_finish() bcache: add code comments for journal_read_bucket() bcache: set largest seq to ja->seq[bucket_index] in journal_read_bucket() bcache: shrink btree node cache after bch_btree_check() bcache: remove retry_flush_write from struct cache_set bcache: performance improvement for btree_flush_write() bcache: add reclaimed_journal_buckets to struct cache_set bcache: fix static checker warning in bcache_device_free() bcache: add more accurate error messages in read_super() bcache: deleted code comments for dead code in bch_data_insert_keys() bcache: add code comment bch_keylist_pop() and bch_keylist_pop_front() bcache: add code comments in bch_btree_leaf_dirty() bcache: add idle_max_writeback_rate sysfs interface bcache: at least try to shrink 1 node in bch_mca_scan() Dan Carpenter (1): bcache: Fix an error code in bch_dump_read() Dongbo Cao (3): bcache: remove useless parameter of bch_debug_init() bcache: split combined if-condition code into separate ones bcache: panic fix for making cache device Geliang Tang (1): bcache: use kmemdup_nul for CACHED_LABEL buffer George Spelvin (1): bcache: Clean up bch_get_congested() Guoju Fang (4): bcache: print number of keys in trace_bcache_journal_write bcache: fix crashes stopping bcache device before read miss done bcache: fix inaccurate result of unused buckets bcache: fix a lost wake-up problem caused by mca_cannibalize_lock Jens Axboe (1): bcache: make is_discard_enabled() static Ming Lei (1): bcache: avoid to use bio_for_each_segment_all() in bch_bio_alloc_pages() Shenghui Wang (7): bcache: remove unused bch_passthrough_cache bcache: use MAX_CACHES_PER_SET instead of magic number 8 in __bch_bucket_alloc_set bcache: add comment for cache_set->fill_iter bcache: update comment for bch_data_insert bcache: update comment in sysfs.c bcache: cannot set writeback_running via sysfs if no writeback kthread created bcache: fix wrong usage use-after-freed on keylist in out_nocoalesce branch of btree_gc_coalesce Shile Zhang (1): bcache: add cond_resched() in __bch_cache_cmp() Wei Yongjun (1): bcache: fix possible memory leak in bch_cached_dev_run() drivers/md/bcache/Makefile | 2 - drivers/md/bcache/alloc.c | 7 +- drivers/md/bcache/bcache.h | 30 ++- drivers/md/bcache/bset.c | 78 ++----- drivers/md/bcache/btree.c | 30 ++- drivers/md/bcache/btree.h | 18 ++ drivers/md/bcache/closure.c | 7 - drivers/md/bcache/closure.h | 3 +- drivers/md/bcache/debug.c | 7 +- drivers/md/bcache/journal.c | 121 ++++++++--- drivers/md/bcache/journal.h | 4 + drivers/md/bcache/request.c | 59 ++--- drivers/md/bcache/request.h | 4 +- drivers/md/bcache/stats.c | 2 +- drivers/md/bcache/super.c | 486 +++++++++++++++++++++++++++++++++++------- drivers/md/bcache/sysfs.c | 166 +++++++++++---- drivers/md/bcache/sysfs.h | 10 + drivers/md/bcache/util.c | 6 +- drivers/md/bcache/util.h | 26 ++- drivers/md/bcache/writeback.c | 34 +++ drivers/md/bcache/writeback.h | 12 +- include/trace/events/bcache.h | 27 ++- 22 files changed, 861 insertions(+), 278 deletions(-) -- 2.16.4

1 74

[RESEND PATCH 00/75] bcache backport for openEuler-1.0 upto 20200104
by Coly Li 08 Jan '20

08 Jan '20

Hi openEler kernel maintainer, This is the bcache backport since Linux v4.19 to v5.5 for openEuler-10 kernel. All the patches are based on openEuler-1.0 commit cf975ca6a64a ("Document: add guideline to submitting patches to openEuler"). This patch set survives for 2+ hours smoking test on a 48x2 cores Taishan machine. Pleasse review and consider to pick them. Thanks in advance. Coly Li --- Andrea Righi (1): bcache: fix deadlock in bcache_allocator Christoph Hellwig (2): bcache: remove the extra cflags for request.o bcache: don't export symbols Colin Ian King (1): bcache: fix indentation issue, remove tabs on a hunk of code Coly Li (50): bcache: fix typo in code comments of closure_return_with_destructor() bcache: introduce force_wake_up_gc() bcache: option to automatically run gc thread after writeback bcache: add MODULE_DESCRIPTION information bcache: make cutoff_writeback and cutoff_writeback_sync tunable bcache: set writeback_percent in a flexible range bcache: not use hard coded memset size in bch_cache_accounting_clear() bcache: export backing_dev_name via sysfs bcache: export backing_dev_uuid via sysfs bcache: fix input integer overflow of congested threshold bcache: add sysfs_strtoul_bool() for setting bit-field variables bcache: use sysfs_strtoul_bool() to set bit-field variables bcache: fix input overflow to writeback_delay bcache: fix input overflow to journal_delay_ms bcache: fix input overflow to cache set io_error_limit bcache: move definition of 'int ret' out of macro read_bucket() bcache: add comments for kobj release callback routine bcache: add error check for calling register_bdev() bcache: Add comments for blkdev_put() in registration code path bcache: add comments for closure_fn to be called in closure_queue() bcache: improve bcache_reboot() bcache: don't set max writeback rate if gc is running bcache: fix return value error in bch_journal_read() bcache: avoid flushing btree node in cache_set_flush() if io disabled bcache: add io error counting in write_bdev_super_endio() bcache: remove unnecessary prefetch() in bset_search_tree() bcache: add return value check to bch_cached_dev_run() bcache: remove unncessary code in bch_btree_keys_init() bcache: more detailed error message to bcache_device_link() bcache: add more error message in bch_cached_dev_attach() bcache: improve error message in bch_cached_dev_run() bcache: remove "XXX:" comment line from run_cache_set() bcache: make bset_search_tree() be more understandable bcache: add pendings_cleanup to stop pending bcache device bcache: stop writeback kthread and kworker when bch_cached_dev_run() failed bcache: avoid a deadlock in bcache_reboot() bcache: acquire bch_register_lock later in cached_dev_detach_finish() bcache: add code comments for journal_read_bucket() bcache: set largest seq to ja->seq[bucket_index] in journal_read_bucket() bcache: shrink btree node cache after bch_btree_check() bcache: remove retry_flush_write from struct cache_set bcache: performance improvement for btree_flush_write() bcache: add reclaimed_journal_buckets to struct cache_set bcache: fix static checker warning in bcache_device_free() bcache: add more accurate error messages in read_super() bcache: deleted code comments for dead code in bch_data_insert_keys() bcache: add code comment bch_keylist_pop() and bch_keylist_pop_front() bcache: add code comments in bch_btree_leaf_dirty() bcache: add idle_max_writeback_rate sysfs interface bcache: at least try to shrink 1 node in bch_mca_scan() Dan Carpenter (1): bcache: Fix an error code in bch_dump_read() Dongbo Cao (3): bcache: remove useless parameter of bch_debug_init() bcache: split combined if-condition code into separate ones bcache: panic fix for making cache device Geliang Tang (1): bcache: use kmemdup_nul for CACHED_LABEL buffer George Spelvin (1): bcache: Clean up bch_get_congested() Guoju Fang (4): bcache: print number of keys in trace_bcache_journal_write bcache: fix crashes stopping bcache device before read miss done bcache: fix inaccurate result of unused buckets bcache: fix a lost wake-up problem caused by mca_cannibalize_lock Jens Axboe (1): bcache: make is_discard_enabled() static Ming Lei (1): bcache: avoid to use bio_for_each_segment_all() in bch_bio_alloc_pages() Shenghui Wang (7): bcache: remove unused bch_passthrough_cache bcache: use MAX_CACHES_PER_SET instead of magic number 8 in __bch_bucket_alloc_set bcache: add comment for cache_set->fill_iter bcache: update comment for bch_data_insert bcache: update comment in sysfs.c bcache: cannot set writeback_running via sysfs if no writeback kthread created bcache: fix wrong usage use-after-freed on keylist in out_nocoalesce branch of btree_gc_coalesce Shile Zhang (1): bcache: add cond_resched() in __bch_cache_cmp() Wei Yongjun (1): bcache: fix possible memory leak in bch_cached_dev_run() drivers/md/bcache/Makefile | 2 - drivers/md/bcache/alloc.c | 7 +- drivers/md/bcache/bcache.h | 30 ++- drivers/md/bcache/bset.c | 78 ++----- drivers/md/bcache/btree.c | 30 ++- drivers/md/bcache/btree.h | 18 ++ drivers/md/bcache/closure.c | 7 - drivers/md/bcache/closure.h | 3 +- drivers/md/bcache/debug.c | 7 +- drivers/md/bcache/journal.c | 121 ++++++++--- drivers/md/bcache/journal.h | 4 + drivers/md/bcache/request.c | 59 ++--- drivers/md/bcache/request.h | 4 +- drivers/md/bcache/stats.c | 2 +- drivers/md/bcache/super.c | 486 +++++++++++++++++++++++++++++++++++------- drivers/md/bcache/sysfs.c | 166 +++++++++++---- drivers/md/bcache/sysfs.h | 10 + drivers/md/bcache/util.c | 6 +- drivers/md/bcache/util.h | 26 ++- drivers/md/bcache/writeback.c | 34 +++ drivers/md/bcache/writeback.h | 12 +- include/trace/events/bcache.h | 27 ++- 22 files changed, 861 insertions(+), 278 deletions(-) -- 2.16.4

1 1

[Discussion] How to manage new projects in openEuler community
by Liyongle (Fred) 06 Jan '20

06 Jan '20

Hi all, If you are interested in managing new projects, or you are willing to create new projects in our community, you may be interested in the discussion [1]. Join to show your opinion. You opinion matters. [1] https://gitee.com/openeuler/community/issues/I17WWY?from=project-issue Fred 李永乐

1 0

Proposed to imporve your repos/projects readability
by Liyongle (Fred) 06 Jan '20

06 Jan '20

Hi all, At this early stage, many developers are coming to see the projects. To make visitors easier to understand the project, it is proposed to improve the following: - README It is better to provide both English[1] and Chinese[2]. - Meeting schedule Please provide meeting schedule and add the information to your readme. The example is [3]. - Contact Mail and IRC. - Team members Please be sure to including Maintainers, committers and contributors. This is a way to respect every contributor, and let new contributors to understand who to talk. - How to contribute Different projects have their own way to contribute. Some projects (like community) are mainly documents, some are codes and some are buildings. Please provide the contribution guidance for your own projects. - Other useful information Your advice and good practice are valuable. [1] https://gitee.com/openeuler/community/blob/master/README.md [2] https://gitee.com/openeuler/community/blob/master/README_cn.md [3] https://gitee.com/openeuler/community/blob/master/README.md#meeting Fred 李永乐

1 0

RE: [Dev] 如何向openEuler-1.0内核提交后向移植的补丁？
by Liyongle (Fred) 05 Jan '20

05 Jan '20

If you are committing patches for kernel, please send email to kernel(a)openeuler.org like in Kernel community. If for others, please find the project, or if you couldn't find it, follow [1] to create a new SIG for the project. [1] https://gitee.com/openeuler/community/blob/master/zh/technical-committee/go… Fred 李永乐 -----Original Message----- From: Coly Li via Dev [mailto:dev@openeuler.org] Sent: Saturday, January 4, 2020 22:44 To: dev(a)openeuler.org Subject: [Dev] 如何向openEuler-1.0内核提交后向移植的补丁？嗨，我今天按照openEuler-1.0内核里的指引，将bcache从Linux v5.5到v4.19里所有 openEuler内核缺失的补丁都做了后向移植，并在泰山服务器上做了简单的冒烟测试。请问这些补丁如何提交呢？是向dev(a)openeuler.xn--org(-3h5fs2ba998ajsdk5fk10bgbn7g3dgcvcw6zaga 内核开发那样），还是在gitee网站上发pull request（像github那样）。多谢！ -- Coly Li _______________________________________________ Dev mailing list -- dev(a)openeuler.org To unsubscribe send an email to dev-leave(a)openeuler.org

1 0

[PATCH 00/20] Fixes some bugs in drivers and kernel
by Yang Yingliang 03 Jan '20

03 Jan '20

This patchset fixes several bugs in drivers and kernel. Also backport serveral CVE fixes from mainline. Eric Dumazet (1): net: silence KCSAN warnings around sk_add_backlog() calls Ganapathi Bhat (1): mwifiex: fix possible heap overflow in mwifiex_process_country_ie() Gao Chuan (1): hisi_sas: Solve the bug of hisi sas tried to access other's interrupt. Hongbo Yao (1): Revert "vfio: relieve mmap_sem reader cacheline bouncing by holding it longer" Junxin Chen (2): net: hns3: fix bug when config of HNS3 is y net: hns3: fix bugs found by codereview Liu Yanshi (1): pcie: hisi pcie add saftly check Mike Christie (1): nbd: fix shutdown and recv work deadlock v2 Shaozhengchao (1): net/hinic: rectify issue Wei Li (1): sched/debug: Reset watchdog on all CPUs while processing sysrq-t Xiongfeng Wang (1): PCI: fix the wrong class type for HiSilicon NP 5896 device YueHaibing (1): dccp: Fix memleak in __feat_register_sp Zhang Wei (1): acc: fixup security problem and modify reset process fengsheng (2): drivers : localbus fixup some parameter dont check the legitimacy drivers : sysctl fixup some parameter dont check the legitimacy wanglin (2): RDMA/hns: Optimize roce code by review advice RDMA/hns: modify roce dfx code by review advice yangerkun (1): ext4: stop IO for page without buffer_head Zhang Wei (1): fixup sec2 free c_key but not cleanup it zhangwei (1): hpre: fixup security check review problem drivers/block/nbd.c | 6 +- drivers/crypto/hisilicon/hpre/hpre_crypto.c | 22 +- drivers/crypto/hisilicon/hpre/hpre_main.c | 157 ++++++----- drivers/crypto/hisilicon/qm.c | 205 +++++++++++++-- drivers/crypto/hisilicon/qm.h | 23 +- drivers/crypto/hisilicon/rde/rde.h | 25 +- drivers/crypto/hisilicon/rde/rde_api.c | 244 +++++++++--------- drivers/crypto/hisilicon/rde/rde_api.h | 2 +- drivers/crypto/hisilicon/rde/rde_data.c | 28 +- drivers/crypto/hisilicon/rde/rde_main.c | 183 +++++++------ drivers/crypto/hisilicon/sec/sec_algs.c | 13 +- drivers/crypto/hisilicon/sec2/sec.h | 10 +- drivers/crypto/hisilicon/sec2/sec_crypto.c | 43 +-- drivers/crypto/hisilicon/sec2/sec_crypto.h | 10 +- drivers/crypto/hisilicon/sec2/sec_main.c | 244 ++++++++++-------- drivers/crypto/hisilicon/sec2/sec_usr_if.h | 10 +- drivers/crypto/hisilicon/sgl.c | 6 +- drivers/crypto/hisilicon/zip/zip_crypto.c | 4 +- drivers/crypto/hisilicon/zip/zip_main.c | 190 +++++++++----- .../infiniband/hw/hns/hns_roce_hw_sysfs_v2.c | 6 +- .../hw/hns/roce-customer/rdfx_entry.c | 24 +- .../hw/hns/roce-customer/rdfx_hw_v2.c | 58 +---- .../hw/hns/roce-customer/rdfx_main.c | 2 + .../hw/hns/roce-customer/rdfx_sysfs.c | 96 +------ drivers/net/ethernet/hisilicon/hns3/hnae3.h | 2 +- .../hns3/hns-customer/hns3_enet_it.c | 2 +- .../hns3/hns-customer/hns3pf/hclge_ext.c | 12 +- .../hisilicon/hns3/hns3_cae/hns3_cae_cmd.c | 2 +- .../hisilicon/hns3/hns3_cae/hns3_cae_fd.c | 2 +- .../hisilicon/hns3/hns3_cae/hns3_cae_init.c | 13 +- .../hisilicon/hns3/hns3_cae/hns3_cae_pkt.c | 39 ++- .../hisilicon/hns3/hns3_cae/hns3_cae_pkt.h | 3 + .../hisilicon/hns3/hns3_cae/hns3_cae_qinfo.c | 2 +- .../hns3/hns3_cae/hns3_cae_version.c | 4 +- .../hns3/hns3_cae/hns3_cae_version.h | 2 +- .../net/ethernet/hisilicon/hns3/hns3_enet.h | 2 +- .../hisilicon/hns3/hns3pf/hclge_main.h | 2 +- .../hisilicon/hns3/hns3vf/hclgevf_main.h | 2 +- drivers/net/ethernet/huawei/hinic/hinic_tx.c | 2 +- .../net/wireless/marvell/mwifiex/sta_ioctl.c | 13 +- .../hisi-pcie-customer/hisi_pcie_cae.c | 108 +++++++- drivers/pci/quirks.c | 2 +- drivers/scsi/hisi_sas/hisi_sas_v3_hw.c | 2 +- drivers/soc/hisilicon/lbc/hs_lbc_pltfm.c | 12 +- drivers/soc/hisilicon/sysctl/sysctl_drv.c | 4 +- drivers/soc/hisilicon/sysctl/sysctl_pmbus.c | 47 +++- drivers/vfio/vfio_iommu_type1.c | 28 +- fs/ext4/inode.c | 26 ++ kernel/sched/debug.c | 11 +- net/core/sock.c | 2 +- net/dccp/feat.c | 7 +- net/ipv4/tcp_ipv4.c | 2 +- net/llc/llc_conn.c | 2 +- net/sctp/input.c | 4 +- net/tipc/socket.c | 6 +- net/x25/x25_dev.c | 2 +- 56 files changed, 1152 insertions(+), 828 deletions(-) -- 2.20.1

1 20

[infrastructure]What kind of blog system do you need in openEuler community?
by Liyongle (Fred) 03 Jan '20

03 Jan '20

Hi all, We are considering to build a blog in openEuler. Please leave your opinion in [1]. Your opinion matters. [1] https://gitee.com/openeuler/community-issue/issues/I17NN3 Fred 李永乐

1 0

[discussion] What code of conduct should we follow?
by Liyongle (Fred) 03 Jan '20

03 Jan '20

Hi all, Is openEuler a serious open source project? A very good question! Let’s try to figure out the code of conduct for our projects. You are welcome to provide your suggestion in [1]. [1] https://gitee.com/openeuler/community-issue/issues/I17R49?from=project-issue Fred 李永乐

1 0

openEuler is launched now
by Liyongle (Fred) 01 Jan '20

01 Jan '20

Dear openEuler members, I am excited to announce that openEuler is open now and you can access, read and play now. Thanks to all the OS developers, infrastructure team and friends who cares about openEuler, contributes in openEuler. It is just a start, and let’s continue collaborating to make openEuler better. The following links may help you to get a glance to openEuler community. Website: https://openeuler.org/ You can find two organizations on gitee.com Code sources: https://gitee.com/openeuler/ Package sources: https://gitee.com/src-openeuler/ To download ISO, visit https://openeuler.obs.cn-south-1.myhuaweicloud.com/release/openEuler-1.0-aa… To know how to start contributing, visit https://openeuler.org/en/developer.html Find all the SIGs: https://openeuler.org/en/sig.html To apply for a new SIG: https://gitee.com/openeuler/community/blob/master/en/technical-committee/go… or https://gitee.com/openeuler/community/blob/master/zh/technical-committee/go… in Chinese. To know technical Committee, visit https://gitee.com/openeuler/community/tree/master/zh/technical-committee. Happy New Year and wish you all have a happy 2020! Fred 李永乐

2 1