- Kernel - mailweb.openeuler.org

[PATCH openEuler-1.0-LTS v2] ext4: avoid deadlock in fs reclaim with page writeback
by Baokun Li 11 Jun '24

11 Jun '24

From: Jan Kara <jack(a)suse.cz> mainline inclusion from mainline-v6.4-rc2 commit 00d873c17e29cc32d90ca852b82685f1673acaa5 category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I9SYGK CVE: NA Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- Ext4 has a filesystem wide lock protecting ext4_writepages() calls to avoid races with switching of journalled data flag or inode format. This lock can however cause a deadlock like: CPU0 CPU1 ext4_writepages() percpu_down_read(sbi->s_writepages_rwsem); ext4_change_inode_journal_flag() percpu_down_write(sbi->s_writepages_rwsem); - blocks, all readers block from now on ext4_do_writepages() ext4_init_io_end() kmem_cache_zalloc(io_end_cachep, GFP_KERNEL) fs_reclaim frees dentry... dentry_unlink_inode() iput() - last ref => iput_final() - inode dirty => write_inode_now()... ext4_writepages() tries to acquire sbi->s_writepages_rwsem and blocks forever Make sure we cannot recurse into filesystem reclaim from writeback code to avoid the deadlock. Reported-by: syzbot+6898da502aef574c5f8a(a)syzkaller.appspotmail.com Link: https://lore.kernel.org/all/0000000000004c66b405fa108e27@google.com Fixes: c8585c6fcaf2 ("ext4: fix races between changing inode journal mode and ext4_writepages") CC: stable(a)vger.kernel.org Signed-off-by: Jan Kara <jack(a)suse.cz> Link: https://lore.kernel.org/r/20230504124723.20205-1-jack@suse.cz Signed-off-by: Theodore Ts'o <tytso(a)mit.edu> Conflicts: fs/ext4/ext4.h fs/ext4/super.c fs/ext4/inode.c fs/ext4/migrate.c [Because we have merged in 41d10c74ff04 ("ext4: fix race between writepages and remount") and we don't have ext4_do_writepages() yet and i_mmap_sem hasn't switched to invalidate_lock yet.] Signed-off-by: Baokun Li <libaokun1(a)huawei.com> --- V1->V2: Correct fs/ext4/migrate.c fs/ext4/ext4.h | 25 +++++++++++++++++++++++++ fs/ext4/inode.c | 18 ++++++++++-------- fs/ext4/migrate.c | 11 ++++++----- fs/ext4/super.c | 11 ++++++----- 4 files changed, 47 insertions(+), 18 deletions(-) diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index 6df919b154b4..1794f5ce1b46 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -37,6 +37,7 @@ #include <crypto/hash.h> #include <linux/falloc.h> #include <linux/percpu-rwsem.h> +#include <linux/sched/mm.h> #ifdef __KERNEL__ #include <linux/compat.h> #endif @@ -1571,6 +1572,30 @@ static inline struct ext4_inode_info *EXT4_I(struct inode *inode) return container_of(inode, struct ext4_inode_info, vfs_inode); } +static inline int ext4_writepages_down_read(struct super_block *sb) +{ + percpu_down_read(&EXT4_SB(sb)->s_writepages_rwsem); + return memalloc_nofs_save(); +} + +static inline void ext4_writepages_up_read(struct super_block *sb, int ctx) +{ + memalloc_nofs_restore(ctx); + percpu_up_read(&EXT4_SB(sb)->s_writepages_rwsem); +} + +static inline int ext4_writepages_down_write(struct super_block *sb) +{ + percpu_down_write(&EXT4_SB(sb)->s_writepages_rwsem); + return memalloc_nofs_save(); +} + +static inline void ext4_writepages_up_write(struct super_block *sb, int ctx) +{ + memalloc_nofs_restore(ctx); + percpu_up_write(&EXT4_SB(sb)->s_writepages_rwsem); +} + static inline int ext4_valid_inum(struct super_block *sb, unsigned long ino) { return ino == EXT4_ROOT_INO || diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 35033bb43149..7110fda36c14 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -2744,13 +2744,14 @@ static int ext4_writepages(struct address_space *mapping, struct blk_plug plug; bool give_up_on_write = false; unsigned long retry_warn_ddl = 0; + int alloc_ctx; #define RETRY_WARN_TIMEOUT (30 * HZ) if (unlikely(ext4_forced_shutdown(EXT4_SB(inode->i_sb)))) return -EIO; - percpu_down_read(&sbi->s_writepages_rwsem); + alloc_ctx = ext4_writepages_down_read(inode->i_sb); trace_ext4_writepages(inode, wbc); /* @@ -2981,7 +2982,7 @@ static int ext4_writepages(struct address_space *mapping, out_writepages: trace_ext4_writepages_result(inode, wbc, ret, nr_to_write - wbc->nr_to_write); - percpu_up_read(&sbi->s_writepages_rwsem); + ext4_writepages_up_read(inode->i_sb, alloc_ctx); return ret; } @@ -2992,17 +2993,18 @@ static int ext4_dax_writepages(struct address_space *mapping, long nr_to_write = wbc->nr_to_write; struct inode *inode = mapping->host; struct ext4_sb_info *sbi = EXT4_SB(mapping->host->i_sb); + int alloc_ctx; if (unlikely(ext4_forced_shutdown(EXT4_SB(inode->i_sb)))) return -EIO; - percpu_down_read(&sbi->s_writepages_rwsem); + alloc_ctx = ext4_writepages_down_read(inode->i_sb); trace_ext4_writepages(inode, wbc); ret = dax_writeback_mapping_range(mapping, inode->i_sb->s_bdev, wbc); trace_ext4_writepages_result(inode, wbc, ret, nr_to_write - wbc->nr_to_write); - percpu_up_read(&sbi->s_writepages_rwsem); + ext4_writepages_up_read(inode->i_sb, alloc_ctx); return ret; } @@ -6245,7 +6247,7 @@ int ext4_change_inode_journal_flag(struct inode *inode, int val) journal_t *journal; handle_t *handle; int err; - struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); + int alloc_ctx; /* * We have to be very careful here: changing a data block's @@ -6285,7 +6287,7 @@ int ext4_change_inode_journal_flag(struct inode *inode, int val) } } - percpu_down_write(&sbi->s_writepages_rwsem); + alloc_ctx = ext4_writepages_down_write(inode->i_sb); jbd2_journal_lock_updates(journal); /* @@ -6302,7 +6304,7 @@ int ext4_change_inode_journal_flag(struct inode *inode, int val) err = jbd2_journal_flush(journal); if (err < 0) { jbd2_journal_unlock_updates(journal); - percpu_up_write(&sbi->s_writepages_rwsem); + ext4_writepages_up_write(inode->i_sb, alloc_ctx); ext4_inode_resume_unlocked_dio(inode); return err; } @@ -6311,7 +6313,7 @@ int ext4_change_inode_journal_flag(struct inode *inode, int val) ext4_set_aops(inode); jbd2_journal_unlock_updates(journal); - percpu_up_write(&sbi->s_writepages_rwsem); + ext4_writepages_up_write(inode->i_sb, alloc_ctx); if (val) up_write(&EXT4_I(inode)->i_mmap_sem); diff --git a/fs/ext4/migrate.c b/fs/ext4/migrate.c index ed9e7816efbb..443a09087dac 100644 --- a/fs/ext4/migrate.c +++ b/fs/ext4/migrate.c @@ -407,7 +407,6 @@ static int free_ext_block(handle_t *handle, struct inode *inode) int ext4_ext_migrate(struct inode *inode) { - struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); handle_t *handle; int retval = 0, i; __le32 *i_data; @@ -417,6 +416,7 @@ int ext4_ext_migrate(struct inode *inode) unsigned long max_entries; __u32 goal, tmp_csum_seed; uid_t owner[2]; + int alloc_ctx; /* * If the filesystem does not support extents, or the inode @@ -432,7 +432,7 @@ int ext4_ext_migrate(struct inode *inode) */ return retval; - percpu_down_write(&sbi->s_writepages_rwsem); + alloc_ctx = ext4_writepages_down_write(inode->i_sb); /* * Worst case we can touch the allocation bitmaps and a block @@ -584,7 +584,7 @@ int ext4_ext_migrate(struct inode *inode) unlock_new_inode(tmp_inode); iput(tmp_inode); out_unlock: - percpu_up_write(&sbi->s_writepages_rwsem); + ext4_writepages_up_write(inode->i_sb, alloc_ctx); return retval; } @@ -603,6 +603,7 @@ int ext4_ind_migrate(struct inode *inode) ext4_fsblk_t blk; handle_t *handle; int ret; + int alloc_ctx; if (!ext4_has_feature_extents(inode->i_sb) || (!ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))) @@ -619,7 +620,7 @@ int ext4_ind_migrate(struct inode *inode) if (test_opt(inode->i_sb, DELALLOC)) ext4_alloc_da_blocks(inode); - percpu_down_write(&sbi->s_writepages_rwsem); + alloc_ctx = ext4_writepages_down_write(inode->i_sb); handle = ext4_journal_start(inode, EXT4_HT_MIGRATE, 1); if (IS_ERR(handle)) { @@ -661,6 +662,6 @@ int ext4_ind_migrate(struct inode *inode) ext4_journal_stop(handle); up_write(&EXT4_I(inode)->i_data_sem); out_unlock: - percpu_up_write(&sbi->s_writepages_rwsem); + ext4_writepages_up_write(inode->i_sb, alloc_ctx); return ret; } diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 1cfec3ac7c9e..8761fe5037a6 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -5559,6 +5559,7 @@ static int ext4_remount(struct super_block *sb, int *flags, char *data) ext4_group_t g; unsigned int journal_ioprio = DEFAULT_JOURNAL_IOPRIO; int err = 0; + int alloc_ctx; #ifdef CONFIG_QUOTA int i, j; char *to_free[EXT4_MAXQUOTAS]; @@ -5611,13 +5612,13 @@ static int ext4_remount(struct super_block *sb, int *flags, char *data) * s_writepages_rwsem to avoid race between writepages ops and * remount. */ - percpu_down_write(&sbi->s_writepages_rwsem); + alloc_ctx = ext4_writepages_down_write(sb); if (!parse_options(data, sb, NULL, &journal_ioprio, 1)) { err = -EINVAL; - percpu_up_write(&sbi->s_writepages_rwsem); + ext4_writepages_up_write(sb, alloc_ctx); goto restore_opts; } - percpu_up_write(&sbi->s_writepages_rwsem); + ext4_writepages_up_write(sb, alloc_ctx); if ((old_opts.s_mount_opt & EXT4_MOUNT_JOURNAL_CHECKSUM) ^ test_opt(sb, JOURNAL_CHECKSUM)) { @@ -5850,7 +5851,7 @@ static int ext4_remount(struct super_block *sb, int *flags, char *data) sb_any_quota_suspended(sb)) dquot_resume(sb, -1); - percpu_down_write(&sbi->s_writepages_rwsem); + alloc_ctx = ext4_writepages_down_write(sb); sb->s_flags = old_sb_flags; sbi->s_mount_opt = old_opts.s_mount_opt; sbi->s_mount_opt2 = old_opts.s_mount_opt2; @@ -5859,7 +5860,7 @@ static int ext4_remount(struct super_block *sb, int *flags, char *data) sbi->s_commit_interval = old_opts.s_commit_interval; sbi->s_min_batch_time = old_opts.s_min_batch_time; sbi->s_max_batch_time = old_opts.s_max_batch_time; - percpu_up_write(&sbi->s_writepages_rwsem); + ext4_writepages_up_write(sb, alloc_ctx); if (!test_opt(sb, BLOCK_VALIDITY) && sbi->system_blks) ext4_release_system_zone(sb); -- 2.31.1

2 1

[PATCH OLK-6.6] scsi: lpfc: Release hbalock before calling lpfc_worker_wake_up()
by Ye Bin 11 Jun '24

11 Jun '24

From: Justin Tee <justin.tee(a)broadcom.com> stable inclusion from stable-v6.6.31 commit e8bf2c05e8ad68e90f9d5889a9e4ef3f6fe00683 category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I9U211 CVE: CVE-2024-36924 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… -------------------------------- [ Upstream commit ded20192dff31c91cef2a04f7e20e60e9bb887d3 ] lpfc_worker_wake_up() calls the lpfc_work_done() routine, which takes the hbalock. Thus, lpfc_worker_wake_up() should not be called while holding the hbalock to avoid potential deadlock. Signed-off-by: Justin Tee <justin.tee(a)broadcom.com> Link: https://lore.kernel.org/r/20240305200503.57317-7-justintee8345@gmail.com Signed-off-by: Martin K. Petersen <martin.petersen(a)oracle.com> Signed-off-by: Sasha Levin <sashal(a)kernel.org> Signed-off-by: Ye Bin <yebin10(a)huawei.com> --- drivers/scsi/lpfc/lpfc_els.c | 20 ++++++++++---------- drivers/scsi/lpfc/lpfc_hbadisc.c | 5 ++--- drivers/scsi/lpfc/lpfc_sli.c | 14 +++++++------- 3 files changed, 19 insertions(+), 20 deletions(-) diff --git a/drivers/scsi/lpfc/lpfc_els.c b/drivers/scsi/lpfc/lpfc_els.c index 18b8325fd419..44d3ada9fbbc 100644 --- a/drivers/scsi/lpfc/lpfc_els.c +++ b/drivers/scsi/lpfc/lpfc_els.c @@ -4432,23 +4432,23 @@ lpfc_els_retry_delay(struct timer_list *t) unsigned long flags; struct lpfc_work_evt *evtp = &ndlp->els_retry_evt; + /* Hold a node reference for outstanding queued work */ + if (!lpfc_nlp_get(ndlp)) + return; + spin_lock_irqsave(&phba->hbalock, flags); if (!list_empty(&evtp->evt_listp)) { spin_unlock_irqrestore(&phba->hbalock, flags); + lpfc_nlp_put(ndlp); return; } - /* We need to hold the node by incrementing the reference - * count until the queued work is done - */ - evtp->evt_arg1 = lpfc_nlp_get(ndlp); - if (evtp->evt_arg1) { - evtp->evt = LPFC_EVT_ELS_RETRY; - list_add_tail(&evtp->evt_listp, &phba->work_list); - lpfc_worker_wake_up(phba); - } + evtp->evt_arg1 = ndlp; + evtp->evt = LPFC_EVT_ELS_RETRY; + list_add_tail(&evtp->evt_listp, &phba->work_list); spin_unlock_irqrestore(&phba->hbalock, flags); - return; + + lpfc_worker_wake_up(phba); } /** diff --git a/drivers/scsi/lpfc/lpfc_hbadisc.c b/drivers/scsi/lpfc/lpfc_hbadisc.c index 5154eeaee0ec..93703ab6ce03 100644 --- a/drivers/scsi/lpfc/lpfc_hbadisc.c +++ b/drivers/scsi/lpfc/lpfc_hbadisc.c @@ -257,7 +257,9 @@ lpfc_dev_loss_tmo_callbk(struct fc_rport *rport) if (evtp->evt_arg1) { evtp->evt = LPFC_EVT_DEV_LOSS; list_add_tail(&evtp->evt_listp, &phba->work_list); + spin_unlock_irqrestore(&phba->hbalock, iflags); lpfc_worker_wake_up(phba); + return; } spin_unlock_irqrestore(&phba->hbalock, iflags); } else { @@ -275,10 +277,7 @@ lpfc_dev_loss_tmo_callbk(struct fc_rport *rport) lpfc_disc_state_machine(vport, ndlp, NULL, NLP_EVT_DEVICE_RM); } - } - - return; } /** diff --git a/drivers/scsi/lpfc/lpfc_sli.c b/drivers/scsi/lpfc/lpfc_sli.c index 4dfadf254a72..9dab33686a93 100644 --- a/drivers/scsi/lpfc/lpfc_sli.c +++ b/drivers/scsi/lpfc/lpfc_sli.c @@ -1217,9 +1217,9 @@ lpfc_set_rrq_active(struct lpfc_hba *phba, struct lpfc_nodelist *ndlp, empty = list_empty(&phba->active_rrq_list); list_add_tail(&rrq->list, &phba->active_rrq_list); phba->hba_flag |= HBA_RRQ_ACTIVE; + spin_unlock_irqrestore(&phba->hbalock, iflags); if (empty) lpfc_worker_wake_up(phba); - spin_unlock_irqrestore(&phba->hbalock, iflags); return 0; out: spin_unlock_irqrestore(&phba->hbalock, iflags); @@ -11369,18 +11369,18 @@ lpfc_sli_post_recovery_event(struct lpfc_hba *phba, unsigned long iflags; struct lpfc_work_evt *evtp = &ndlp->recovery_evt; + /* Hold a node reference for outstanding queued work */ + if (!lpfc_nlp_get(ndlp)) + return; + spin_lock_irqsave(&phba->hbalock, iflags); if (!list_empty(&evtp->evt_listp)) { spin_unlock_irqrestore(&phba->hbalock, iflags); + lpfc_nlp_put(ndlp); return; } - /* Incrementing the reference count until the queued work is done. */ - evtp->evt_arg1 = lpfc_nlp_get(ndlp); - if (!evtp->evt_arg1) { - spin_unlock_irqrestore(&phba->hbalock, iflags); - return; - } + evtp->evt_arg1 = ndlp; evtp->evt = LPFC_EVT_RECOVER_PORT; list_add_tail(&evtp->evt_listp, &phba->work_list); spin_unlock_irqrestore(&phba->hbalock, iflags); -- 2.31.1

2 1

[PATCH openEuler-1.0-LTS] ext4: avoid deadlock in fs reclaim with page writeback
by Baokun Li 11 Jun '24

11 Jun '24

From: Jan Kara <jack(a)suse.cz> mainline inclusion from mainline-v6.4-rc2 commit 00d873c17e29cc32d90ca852b82685f1673acaa5 category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I9SYGK CVE: NA Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- Ext4 has a filesystem wide lock protecting ext4_writepages() calls to avoid races with switching of journalled data flag or inode format. This lock can however cause a deadlock like: CPU0 CPU1 ext4_writepages() percpu_down_read(sbi->s_writepages_rwsem); ext4_change_inode_journal_flag() percpu_down_write(sbi->s_writepages_rwsem); - blocks, all readers block from now on ext4_do_writepages() ext4_init_io_end() kmem_cache_zalloc(io_end_cachep, GFP_KERNEL) fs_reclaim frees dentry... dentry_unlink_inode() iput() - last ref => iput_final() - inode dirty => write_inode_now()... ext4_writepages() tries to acquire sbi->s_writepages_rwsem and blocks forever Make sure we cannot recurse into filesystem reclaim from writeback code to avoid the deadlock. Reported-by: syzbot+6898da502aef574c5f8a(a)syzkaller.appspotmail.com Link: https://lore.kernel.org/all/0000000000004c66b405fa108e27@google.com Fixes: c8585c6fcaf2 ("ext4: fix races between changing inode journal mode and ext4_writepages") CC: stable(a)vger.kernel.org Signed-off-by: Jan Kara <jack(a)suse.cz> Link: https://lore.kernel.org/r/20230504124723.20205-1-jack@suse.cz Signed-off-by: Theodore Ts'o <tytso(a)mit.edu> Conflicts: fs/ext4/ext4.h fs/ext4/super.c fs/ext4/inode.c fs/ext4/migrate.c [Because we have merged in 41d10c74ff04 ("ext4: fix race between writepages and remount") and we don't have ext4_do_writepages() yet and i_mmap_sem hasn't switched to invalidate_lock yet.] Signed-off-by: Baokun Li <libaokun1(a)huawei.com> --- fs/ext4/ext4.h | 25 +++++++++++++++++++++++++ fs/ext4/inode.c | 18 ++++++++++-------- fs/ext4/super.c | 11 ++++++----- 3 files changed, 41 insertions(+), 13 deletions(-) diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index 6df919b154b4..1794f5ce1b46 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -37,6 +37,7 @@ #include <crypto/hash.h> #include <linux/falloc.h> #include <linux/percpu-rwsem.h> +#include <linux/sched/mm.h> #ifdef __KERNEL__ #include <linux/compat.h> #endif @@ -1571,6 +1572,30 @@ static inline struct ext4_inode_info *EXT4_I(struct inode *inode) return container_of(inode, struct ext4_inode_info, vfs_inode); } +static inline int ext4_writepages_down_read(struct super_block *sb) +{ + percpu_down_read(&EXT4_SB(sb)->s_writepages_rwsem); + return memalloc_nofs_save(); +} + +static inline void ext4_writepages_up_read(struct super_block *sb, int ctx) +{ + memalloc_nofs_restore(ctx); + percpu_up_read(&EXT4_SB(sb)->s_writepages_rwsem); +} + +static inline int ext4_writepages_down_write(struct super_block *sb) +{ + percpu_down_write(&EXT4_SB(sb)->s_writepages_rwsem); + return memalloc_nofs_save(); +} + +static inline void ext4_writepages_up_write(struct super_block *sb, int ctx) +{ + memalloc_nofs_restore(ctx); + percpu_up_write(&EXT4_SB(sb)->s_writepages_rwsem); +} + static inline int ext4_valid_inum(struct super_block *sb, unsigned long ino) { return ino == EXT4_ROOT_INO || diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 35033bb43149..7110fda36c14 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -2744,13 +2744,14 @@ static int ext4_writepages(struct address_space *mapping, struct blk_plug plug; bool give_up_on_write = false; unsigned long retry_warn_ddl = 0; + int alloc_ctx; #define RETRY_WARN_TIMEOUT (30 * HZ) if (unlikely(ext4_forced_shutdown(EXT4_SB(inode->i_sb)))) return -EIO; - percpu_down_read(&sbi->s_writepages_rwsem); + alloc_ctx = ext4_writepages_down_read(inode->i_sb); trace_ext4_writepages(inode, wbc); /* @@ -2981,7 +2982,7 @@ static int ext4_writepages(struct address_space *mapping, out_writepages: trace_ext4_writepages_result(inode, wbc, ret, nr_to_write - wbc->nr_to_write); - percpu_up_read(&sbi->s_writepages_rwsem); + ext4_writepages_up_read(inode->i_sb, alloc_ctx); return ret; } @@ -2992,17 +2993,18 @@ static int ext4_dax_writepages(struct address_space *mapping, long nr_to_write = wbc->nr_to_write; struct inode *inode = mapping->host; struct ext4_sb_info *sbi = EXT4_SB(mapping->host->i_sb); + int alloc_ctx; if (unlikely(ext4_forced_shutdown(EXT4_SB(inode->i_sb)))) return -EIO; - percpu_down_read(&sbi->s_writepages_rwsem); + alloc_ctx = ext4_writepages_down_read(inode->i_sb); trace_ext4_writepages(inode, wbc); ret = dax_writeback_mapping_range(mapping, inode->i_sb->s_bdev, wbc); trace_ext4_writepages_result(inode, wbc, ret, nr_to_write - wbc->nr_to_write); - percpu_up_read(&sbi->s_writepages_rwsem); + ext4_writepages_up_read(inode->i_sb, alloc_ctx); return ret; } @@ -6245,7 +6247,7 @@ int ext4_change_inode_journal_flag(struct inode *inode, int val) journal_t *journal; handle_t *handle; int err; - struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); + int alloc_ctx; /* * We have to be very careful here: changing a data block's @@ -6285,7 +6287,7 @@ int ext4_change_inode_journal_flag(struct inode *inode, int val) } } - percpu_down_write(&sbi->s_writepages_rwsem); + alloc_ctx = ext4_writepages_down_write(inode->i_sb); jbd2_journal_lock_updates(journal); /* @@ -6302,7 +6304,7 @@ int ext4_change_inode_journal_flag(struct inode *inode, int val) err = jbd2_journal_flush(journal); if (err < 0) { jbd2_journal_unlock_updates(journal); - percpu_up_write(&sbi->s_writepages_rwsem); + ext4_writepages_up_write(inode->i_sb, alloc_ctx); ext4_inode_resume_unlocked_dio(inode); return err; } @@ -6311,7 +6313,7 @@ int ext4_change_inode_journal_flag(struct inode *inode, int val) ext4_set_aops(inode); jbd2_journal_unlock_updates(journal); - percpu_up_write(&sbi->s_writepages_rwsem); + ext4_writepages_up_write(inode->i_sb, alloc_ctx); if (val) up_write(&EXT4_I(inode)->i_mmap_sem); diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 1cfec3ac7c9e..8761fe5037a6 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -5559,6 +5559,7 @@ static int ext4_remount(struct super_block *sb, int *flags, char *data) ext4_group_t g; unsigned int journal_ioprio = DEFAULT_JOURNAL_IOPRIO; int err = 0; + int alloc_ctx; #ifdef CONFIG_QUOTA int i, j; char *to_free[EXT4_MAXQUOTAS]; @@ -5611,13 +5612,13 @@ static int ext4_remount(struct super_block *sb, int *flags, char *data) * s_writepages_rwsem to avoid race between writepages ops and * remount. */ - percpu_down_write(&sbi->s_writepages_rwsem); + alloc_ctx = ext4_writepages_down_write(sb); if (!parse_options(data, sb, NULL, &journal_ioprio, 1)) { err = -EINVAL; - percpu_up_write(&sbi->s_writepages_rwsem); + ext4_writepages_up_write(sb, alloc_ctx); goto restore_opts; } - percpu_up_write(&sbi->s_writepages_rwsem); + ext4_writepages_up_write(sb, alloc_ctx); if ((old_opts.s_mount_opt & EXT4_MOUNT_JOURNAL_CHECKSUM) ^ test_opt(sb, JOURNAL_CHECKSUM)) { @@ -5850,7 +5851,7 @@ static int ext4_remount(struct super_block *sb, int *flags, char *data) sb_any_quota_suspended(sb)) dquot_resume(sb, -1); - percpu_down_write(&sbi->s_writepages_rwsem); + alloc_ctx = ext4_writepages_down_write(sb); sb->s_flags = old_sb_flags; sbi->s_mount_opt = old_opts.s_mount_opt; sbi->s_mount_opt2 = old_opts.s_mount_opt2; @@ -5859,7 +5860,7 @@ static int ext4_remount(struct super_block *sb, int *flags, char *data) sbi->s_commit_interval = old_opts.s_commit_interval; sbi->s_min_batch_time = old_opts.s_min_batch_time; sbi->s_max_batch_time = old_opts.s_max_batch_time; - percpu_up_write(&sbi->s_writepages_rwsem); + ext4_writepages_up_write(sb, alloc_ctx); if (!test_opt(sb, BLOCK_VALIDITY) && sbi->system_blks) ext4_release_system_zone(sb); -- 2.31.1

2 1

[PATCH OLK-6.6] kabi: reserve kabi for struct device
by Peng Zhang 11 Jun '24

11 Jun '24

From: ZhangPeng <zhangpeng362(a)huawei.com> hulk inclusion category: cleanup bugzilla: https://gitee.com/openeuler/kernel/issues/I8OWRC -------------------------------- Just testing if the kabi check is working. Please ignore this patch. Signed-off-by: ZhangPeng <zhangpeng362(a)huawei.com> --- include/linux/device.h | 1 + 1 file changed, 1 insertion(+) diff --git a/include/linux/device.h b/include/linux/device.h index 17427fc8e035..0128883e652d 100644 --- a/include/linux/device.h +++ b/include/linux/device.h @@ -814,6 +814,7 @@ struct device { KABI_RESERVE(6) KABI_RESERVE(7) KABI_RESERVE(8) + KABI_RESERVE(9) }; /** -- 2.25.1

2 1

[PATCH] kabi: reserve kabi for struct device
by Peng Zhang 11 Jun '24

11 Jun '24

From: ZhangPeng <zhangpeng362(a)huawei.com> hulk inclusion category: cleanup bugzilla: https://gitee.com/openeuler/kernel/issues/I8OWRC -------------------------------- Just testing if the kabi check is working. Please ignore this patch. Signed-off-by: ZhangPeng <zhangpeng362(a)huawei.com> --- include/linux/device.h | 1 + 1 file changed, 1 insertion(+) diff --git a/include/linux/device.h b/include/linux/device.h index 17427fc8e035..0128883e652d 100644 --- a/include/linux/device.h +++ b/include/linux/device.h @@ -814,6 +814,7 @@ struct device { KABI_RESERVE(6) KABI_RESERVE(7) KABI_RESERVE(8) + KABI_RESERVE(9) }; /** -- 2.25.1

1 0

[PATCH openEuler-22.03-LTS-SP1] ext4: avoid deadlock in fs reclaim with page writeback
by Baokun Li 11 Jun '24

11 Jun '24

From: Jan Kara <jack(a)suse.cz> mainline inclusion from mainline-v6.4-rc2 commit 00d873c17e29cc32d90ca852b82685f1673acaa5 category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I9SYGK Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- Ext4 has a filesystem wide lock protecting ext4_writepages() calls to avoid races with switching of journalled data flag or inode format. This lock can however cause a deadlock like: CPU0 CPU1 ext4_writepages() percpu_down_read(sbi->s_writepages_rwsem); ext4_change_inode_journal_flag() percpu_down_write(sbi->s_writepages_rwsem); - blocks, all readers block from now on ext4_do_writepages() ext4_init_io_end() kmem_cache_zalloc(io_end_cachep, GFP_KERNEL) fs_reclaim frees dentry... dentry_unlink_inode() iput() - last ref => iput_final() - inode dirty => write_inode_now()... ext4_writepages() tries to acquire sbi->s_writepages_rwsem and blocks forever Make sure we cannot recurse into filesystem reclaim from writeback code to avoid the deadlock. Reported-by: syzbot+6898da502aef574c5f8a(a)syzkaller.appspotmail.com Link: https://lore.kernel.org/all/0000000000004c66b405fa108e27@google.com Fixes: c8585c6fcaf2 ("ext4: fix races between changing inode journal mode and ext4_writepages") CC: stable(a)vger.kernel.org Signed-off-by: Jan Kara <jack(a)suse.cz> Link: https://lore.kernel.org/r/20230504124723.20205-1-jack@suse.cz Signed-off-by: Theodore Ts'o <tytso(a)mit.edu> Conflicts: fs/ext4/ext4.h fs/ext4/super.c fs/ext4/inode.c [Because we have merged in 583b2fb551d0 ("ext4: fix race between writepages and remount") and we don't have ext4_do_writepages() yet and i_mmap_sem hasn't switched to invalidate_lock yet.] Signed-off-by: Baokun Li <libaokun1(a)huawei.com> --- fs/ext4/ext4.h | 25 +++++++++++++++++++++++++ fs/ext4/inode.c | 18 ++++++++++-------- fs/ext4/migrate.c | 11 ++++++----- fs/ext4/super.c | 11 ++++++----- 4 files changed, 47 insertions(+), 18 deletions(-) diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index 5d5ae6f44510..3cf1fe773f4f 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -37,6 +37,7 @@ #include <linux/falloc.h> #include <linux/percpu-rwsem.h> #include <linux/fiemap.h> +#include <linux/sched/mm.h> #ifdef __KERNEL__ #include <linux/compat.h> #endif @@ -1689,6 +1690,30 @@ static inline struct ext4_inode_info *EXT4_I(struct inode *inode) return container_of(inode, struct ext4_inode_info, vfs_inode); } +static inline int ext4_writepages_down_read(struct super_block *sb) +{ + percpu_down_read(&EXT4_SB(sb)->s_writepages_rwsem); + return memalloc_nofs_save(); +} + +static inline void ext4_writepages_up_read(struct super_block *sb, int ctx) +{ + memalloc_nofs_restore(ctx); + percpu_up_read(&EXT4_SB(sb)->s_writepages_rwsem); +} + +static inline int ext4_writepages_down_write(struct super_block *sb) +{ + percpu_down_write(&EXT4_SB(sb)->s_writepages_rwsem); + return memalloc_nofs_save(); +} + +static inline void ext4_writepages_up_write(struct super_block *sb, int ctx) +{ + memalloc_nofs_restore(ctx); + percpu_up_write(&EXT4_SB(sb)->s_writepages_rwsem); +} + static inline int ext4_valid_inum(struct super_block *sb, unsigned long ino) { return ino == EXT4_ROOT_INO || diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index f26d9b88d48d..9c438215cf56 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -2705,13 +2705,14 @@ static int ext4_writepages(struct address_space *mapping, struct blk_plug plug; bool give_up_on_write = false; unsigned long retry_warn_ddl = 0; + int alloc_ctx; #define RETRY_WARN_TIMEOUT (30 * HZ) if (unlikely(ext4_forced_shutdown(EXT4_SB(inode->i_sb)))) return -EIO; - percpu_down_read(&sbi->s_writepages_rwsem); + alloc_ctx = ext4_writepages_down_read(inode->i_sb); trace_ext4_writepages(inode, wbc); /* @@ -2932,7 +2933,7 @@ static int ext4_writepages(struct address_space *mapping, out_writepages: trace_ext4_writepages_result(inode, wbc, ret, nr_to_write - wbc->nr_to_write); - percpu_up_read(&sbi->s_writepages_rwsem); + ext4_writepages_up_read(inode->i_sb, alloc_ctx); return ret; } @@ -2943,17 +2944,18 @@ static int ext4_dax_writepages(struct address_space *mapping, long nr_to_write = wbc->nr_to_write; struct inode *inode = mapping->host; struct ext4_sb_info *sbi = EXT4_SB(mapping->host->i_sb); + int alloc_ctx; if (unlikely(ext4_forced_shutdown(EXT4_SB(inode->i_sb)))) return -EIO; - percpu_down_read(&sbi->s_writepages_rwsem); + alloc_ctx = ext4_writepages_down_read(inode->i_sb); trace_ext4_writepages(inode, wbc); ret = dax_writeback_mapping_range(mapping, sbi->s_daxdev, wbc); trace_ext4_writepages_result(inode, wbc, ret, nr_to_write - wbc->nr_to_write); - percpu_up_read(&sbi->s_writepages_rwsem); + ext4_writepages_up_read(inode->i_sb, alloc_ctx); return ret; } @@ -6065,7 +6067,7 @@ int ext4_change_inode_journal_flag(struct inode *inode, int val) journal_t *journal; handle_t *handle; int err; - struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); + int alloc_ctx; /* * We have to be very careful here: changing a data block's @@ -6103,7 +6105,7 @@ int ext4_change_inode_journal_flag(struct inode *inode, int val) } } - percpu_down_write(&sbi->s_writepages_rwsem); + alloc_ctx = ext4_writepages_down_write(inode->i_sb); jbd2_journal_lock_updates(journal); /* @@ -6120,7 +6122,7 @@ int ext4_change_inode_journal_flag(struct inode *inode, int val) err = jbd2_journal_flush(journal); if (err < 0) { jbd2_journal_unlock_updates(journal); - percpu_up_write(&sbi->s_writepages_rwsem); + ext4_writepages_up_write(inode->i_sb, alloc_ctx); return err; } ext4_clear_inode_flag(inode, EXT4_INODE_JOURNAL_DATA); @@ -6128,7 +6130,7 @@ int ext4_change_inode_journal_flag(struct inode *inode, int val) ext4_set_aops(inode); jbd2_journal_unlock_updates(journal); - percpu_up_write(&sbi->s_writepages_rwsem); + ext4_writepages_up_write(inode->i_sb, alloc_ctx); if (val) up_write(&EXT4_I(inode)->i_mmap_sem); diff --git a/fs/ext4/migrate.c b/fs/ext4/migrate.c index 04320715d61f..274f7ab37fff 100644 --- a/fs/ext4/migrate.c +++ b/fs/ext4/migrate.c @@ -409,7 +409,6 @@ static int free_ext_block(handle_t *handle, struct inode *inode) int ext4_ext_migrate(struct inode *inode) { - struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); handle_t *handle; int retval = 0, i; __le32 *i_data; @@ -419,6 +418,7 @@ int ext4_ext_migrate(struct inode *inode) unsigned long max_entries; __u32 goal, tmp_csum_seed; uid_t owner[2]; + int alloc_ctx; /* * If the filesystem does not support extents, or the inode @@ -434,7 +434,7 @@ int ext4_ext_migrate(struct inode *inode) */ return retval; - percpu_down_write(&sbi->s_writepages_rwsem); + alloc_ctx = ext4_writepages_down_write(inode->i_sb); /* * Worst case we can touch the allocation bitmaps and a block @@ -586,7 +586,7 @@ int ext4_ext_migrate(struct inode *inode) unlock_new_inode(tmp_inode); iput(tmp_inode); out_unlock: - percpu_up_write(&sbi->s_writepages_rwsem); + ext4_writepages_up_write(inode->i_sb, alloc_ctx); return retval; } @@ -605,6 +605,7 @@ int ext4_ind_migrate(struct inode *inode) ext4_fsblk_t blk; handle_t *handle; int ret, ret2 = 0; + int alloc_ctx; if (!ext4_has_feature_extents(inode->i_sb) || (!ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))) @@ -621,7 +622,7 @@ int ext4_ind_migrate(struct inode *inode) if (test_opt(inode->i_sb, DELALLOC)) ext4_alloc_da_blocks(inode); - percpu_down_write(&sbi->s_writepages_rwsem); + alloc_ctx = ext4_writepages_down_write(inode->i_sb); handle = ext4_journal_start(inode, EXT4_HT_MIGRATE, 1); if (IS_ERR(handle)) { @@ -665,6 +666,6 @@ int ext4_ind_migrate(struct inode *inode) ext4_journal_stop(handle); up_write(&EXT4_I(inode)->i_data_sem); out_unlock: - percpu_up_write(&sbi->s_writepages_rwsem); + ext4_writepages_up_write(inode->i_sb, alloc_ctx); return ret; } diff --git a/fs/ext4/super.c b/fs/ext4/super.c index ac032f517a60..2bdb16df5c23 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -5939,6 +5939,7 @@ static int ext4_remount(struct super_block *sb, int *flags, char *data) ext4_group_t g; unsigned int journal_ioprio = DEFAULT_JOURNAL_IOPRIO; int err = 0; + int alloc_ctx; #ifdef CONFIG_QUOTA int i, j; char *to_free[EXT4_MAXQUOTAS]; @@ -5991,13 +5992,13 @@ static int ext4_remount(struct super_block *sb, int *flags, char *data) * here s_writepages_rwsem to avoid race between writepages ops and * remount. */ - percpu_down_write(&sbi->s_writepages_rwsem); + alloc_ctx = ext4_writepages_down_write(sb); if (!parse_options(data, sb, NULL, &journal_ioprio, 1)) { err = -EINVAL; - percpu_up_write(&sbi->s_writepages_rwsem); + ext4_writepages_up_write(sb, alloc_ctx); goto restore_opts; } - percpu_up_write(&sbi->s_writepages_rwsem); + ext4_writepages_up_write(sb, alloc_ctx); if ((old_opts.s_mount_opt & EXT4_MOUNT_JOURNAL_CHECKSUM) ^ test_opt(sb, JOURNAL_CHECKSUM)) { @@ -6215,7 +6216,7 @@ static int ext4_remount(struct super_block *sb, int *flags, char *data) return 0; restore_opts: - percpu_down_write(&sbi->s_writepages_rwsem); + alloc_ctx = ext4_writepages_down_write(sb); sb->s_flags = old_sb_flags; sbi->s_mount_opt = old_opts.s_mount_opt; sbi->s_mount_opt2 = old_opts.s_mount_opt2; @@ -6224,7 +6225,7 @@ static int ext4_remount(struct super_block *sb, int *flags, char *data) sbi->s_commit_interval = old_opts.s_commit_interval; sbi->s_min_batch_time = old_opts.s_min_batch_time; sbi->s_max_batch_time = old_opts.s_max_batch_time; - percpu_up_write(&sbi->s_writepages_rwsem); + ext4_writepages_up_write(sb, alloc_ctx); if (!test_opt(sb, BLOCK_VALIDITY) && sbi->s_system_blks) ext4_release_system_zone(sb); -- 2.31.1

2 1

[PATCH OLK-5.10] ext4: avoid deadlock in fs reclaim with page writeback
by Baokun Li 11 Jun '24

11 Jun '24

From: Jan Kara <jack(a)suse.cz> mainline inclusion from mainline-v6.4-rc2 commit 00d873c17e29cc32d90ca852b82685f1673acaa5 category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I9SYGK Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- Ext4 has a filesystem wide lock protecting ext4_writepages() calls to avoid races with switching of journalled data flag or inode format. This lock can however cause a deadlock like: CPU0 CPU1 ext4_writepages() percpu_down_read(sbi->s_writepages_rwsem); ext4_change_inode_journal_flag() percpu_down_write(sbi->s_writepages_rwsem); - blocks, all readers block from now on ext4_do_writepages() ext4_init_io_end() kmem_cache_zalloc(io_end_cachep, GFP_KERNEL) fs_reclaim frees dentry... dentry_unlink_inode() iput() - last ref => iput_final() - inode dirty => write_inode_now()... ext4_writepages() tries to acquire sbi->s_writepages_rwsem and blocks forever Make sure we cannot recurse into filesystem reclaim from writeback code to avoid the deadlock. Reported-by: syzbot+6898da502aef574c5f8a(a)syzkaller.appspotmail.com Link: https://lore.kernel.org/all/0000000000004c66b405fa108e27@google.com Fixes: c8585c6fcaf2 ("ext4: fix races between changing inode journal mode and ext4_writepages") CC: stable(a)vger.kernel.org Signed-off-by: Jan Kara <jack(a)suse.cz> Link: https://lore.kernel.org/r/20230504124723.20205-1-jack@suse.cz Signed-off-by: Theodore Ts'o <tytso(a)mit.edu> Conflicts: fs/ext4/ext4.h fs/ext4/super.c fs/ext4/inode.c [Because we have merged in b66c23ec7fcf ("ext4: fix race between writepages and remount") and we don't have ext4_do_writepages() yet and i_mmap_sem hasn't switched to invalidate_lock yet.] Signed-off-by: Baokun Li <libaokun1(a)huawei.com> --- fs/ext4/ext4.h | 25 +++++++++++++++++++++++++ fs/ext4/inode.c | 18 ++++++++++-------- fs/ext4/migrate.c | 11 ++++++----- fs/ext4/super.c | 12 +++++++----- 4 files changed, 48 insertions(+), 18 deletions(-) diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index b701f7ac7f66..a9530484147d 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -37,6 +37,7 @@ #include <linux/falloc.h> #include <linux/percpu-rwsem.h> #include <linux/fiemap.h> +#include <linux/sched/mm.h> #ifdef __KERNEL__ #include <linux/compat.h> #endif @@ -1701,6 +1702,30 @@ static inline struct ext4_inode_info *EXT4_I(struct inode *inode) return container_of(inode, struct ext4_inode_info, vfs_inode); } +static inline int ext4_writepages_down_read(struct super_block *sb) +{ + percpu_down_read(&EXT4_SB(sb)->s_writepages_rwsem); + return memalloc_nofs_save(); +} + +static inline void ext4_writepages_up_read(struct super_block *sb, int ctx) +{ + memalloc_nofs_restore(ctx); + percpu_up_read(&EXT4_SB(sb)->s_writepages_rwsem); +} + +static inline int ext4_writepages_down_write(struct super_block *sb) +{ + percpu_down_write(&EXT4_SB(sb)->s_writepages_rwsem); + return memalloc_nofs_save(); +} + +static inline void ext4_writepages_up_write(struct super_block *sb, int ctx) +{ + memalloc_nofs_restore(ctx); + percpu_up_write(&EXT4_SB(sb)->s_writepages_rwsem); +} + static inline int ext4_valid_inum(struct super_block *sb, unsigned long ino) { return ino == EXT4_ROOT_INO || diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 412af3c5fde4..0f8d23aa2838 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -2709,13 +2709,14 @@ static int ext4_writepages(struct address_space *mapping, struct blk_plug plug; bool give_up_on_write = false; unsigned long retry_warn_ddl = 0; + int alloc_ctx; #define RETRY_WARN_TIMEOUT (30 * HZ) if (unlikely(ext4_forced_shutdown(EXT4_SB(inode->i_sb)))) return -EIO; - percpu_down_read(&sbi->s_writepages_rwsem); + alloc_ctx = ext4_writepages_down_read(inode->i_sb); trace_ext4_writepages(inode, wbc); /* @@ -2936,7 +2937,7 @@ static int ext4_writepages(struct address_space *mapping, out_writepages: trace_ext4_writepages_result(inode, wbc, ret, nr_to_write - wbc->nr_to_write); - percpu_up_read(&sbi->s_writepages_rwsem); + ext4_writepages_up_read(inode->i_sb, alloc_ctx); return ret; } @@ -2947,17 +2948,18 @@ static int ext4_dax_writepages(struct address_space *mapping, long nr_to_write = wbc->nr_to_write; struct inode *inode = mapping->host; struct ext4_sb_info *sbi = EXT4_SB(mapping->host->i_sb); + int alloc_ctx; if (unlikely(ext4_forced_shutdown(EXT4_SB(inode->i_sb)))) return -EIO; - percpu_down_read(&sbi->s_writepages_rwsem); + alloc_ctx = ext4_writepages_down_read(inode->i_sb); trace_ext4_writepages(inode, wbc); ret = dax_writeback_mapping_range(mapping, sbi->s_daxdev, wbc); trace_ext4_writepages_result(inode, wbc, ret, nr_to_write - wbc->nr_to_write); - percpu_up_read(&sbi->s_writepages_rwsem); + ext4_writepages_up_read(inode->i_sb, alloc_ctx); return ret; } @@ -6107,7 +6109,7 @@ int ext4_change_inode_journal_flag(struct inode *inode, int val) journal_t *journal; handle_t *handle; int err; - struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); + int alloc_ctx; /* * We have to be very careful here: changing a data block's @@ -6145,7 +6147,7 @@ int ext4_change_inode_journal_flag(struct inode *inode, int val) } } - percpu_down_write(&sbi->s_writepages_rwsem); + alloc_ctx = ext4_writepages_down_write(inode->i_sb); jbd2_journal_lock_updates(journal); /* @@ -6162,7 +6164,7 @@ int ext4_change_inode_journal_flag(struct inode *inode, int val) err = jbd2_journal_flush(journal); if (err < 0) { jbd2_journal_unlock_updates(journal); - percpu_up_write(&sbi->s_writepages_rwsem); + ext4_writepages_up_write(inode->i_sb, alloc_ctx); return err; } ext4_clear_inode_flag(inode, EXT4_INODE_JOURNAL_DATA); @@ -6170,7 +6172,7 @@ int ext4_change_inode_journal_flag(struct inode *inode, int val) ext4_set_aops(inode); jbd2_journal_unlock_updates(journal); - percpu_up_write(&sbi->s_writepages_rwsem); + ext4_writepages_up_write(inode->i_sb, alloc_ctx); if (val) up_write(&EXT4_I(inode)->i_mmap_sem); diff --git a/fs/ext4/migrate.c b/fs/ext4/migrate.c index b0ea646454ac..50e449ec62bf 100644 --- a/fs/ext4/migrate.c +++ b/fs/ext4/migrate.c @@ -409,7 +409,6 @@ static int free_ext_block(handle_t *handle, struct inode *inode) int ext4_ext_migrate(struct inode *inode) { - struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); handle_t *handle; int retval = 0, i; __le32 *i_data; @@ -419,6 +418,7 @@ int ext4_ext_migrate(struct inode *inode) unsigned long max_entries; __u32 goal, tmp_csum_seed; uid_t owner[2]; + int alloc_ctx; /* * If the filesystem does not support extents, or the inode @@ -435,7 +435,7 @@ int ext4_ext_migrate(struct inode *inode) */ return retval; - percpu_down_write(&sbi->s_writepages_rwsem); + alloc_ctx = ext4_writepages_down_write(inode->i_sb); /* * Worst case we can touch the allocation bitmaps and a block @@ -587,7 +587,7 @@ int ext4_ext_migrate(struct inode *inode) unlock_new_inode(tmp_inode); iput(tmp_inode); out_unlock: - percpu_up_write(&sbi->s_writepages_rwsem); + ext4_writepages_up_write(inode->i_sb, alloc_ctx); return retval; } @@ -606,6 +606,7 @@ int ext4_ind_migrate(struct inode *inode) ext4_fsblk_t blk; handle_t *handle; int ret, ret2 = 0; + int alloc_ctx; if (!ext4_has_feature_extents(inode->i_sb) || (!ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))) @@ -622,7 +623,7 @@ int ext4_ind_migrate(struct inode *inode) if (test_opt(inode->i_sb, DELALLOC)) ext4_alloc_da_blocks(inode); - percpu_down_write(&sbi->s_writepages_rwsem); + alloc_ctx = ext4_writepages_down_write(inode->i_sb); handle = ext4_journal_start(inode, EXT4_HT_MIGRATE, 1); if (IS_ERR(handle)) { @@ -666,6 +667,6 @@ int ext4_ind_migrate(struct inode *inode) ext4_journal_stop(handle); up_write(&EXT4_I(inode)->i_data_sem); out_unlock: - percpu_up_write(&sbi->s_writepages_rwsem); + ext4_writepages_up_write(inode->i_sb, alloc_ctx); return ret; } diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 90f886184fc6..bfa592bfd298 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -5961,6 +5961,7 @@ static int ext4_remount(struct super_block *sb, int *flags, char *data) ext4_group_t g; unsigned int journal_ioprio = DEFAULT_JOURNAL_IOPRIO; int err = 0; + int alloc_ctx; #ifdef CONFIG_QUOTA int enable_quota = 0; int i, j; @@ -6014,13 +6015,13 @@ static int ext4_remount(struct super_block *sb, int *flags, char *data) * here s_writepages_rwsem to avoid race between writepages ops and * remount. */ - percpu_down_write(&sbi->s_writepages_rwsem); + alloc_ctx = ext4_writepages_down_write(sb); if (!parse_options(data, sb, NULL, &journal_ioprio, 1)) { err = -EINVAL; - percpu_up_write(&sbi->s_writepages_rwsem); + ext4_writepages_up_write(sb, alloc_ctx); goto restore_opts; } - percpu_up_write(&sbi->s_writepages_rwsem); + ext4_writepages_up_write(sb, alloc_ctx); if ((old_opts.s_mount_opt & EXT4_MOUNT_JOURNAL_CHECKSUM) ^ test_opt(sb, JOURNAL_CHECKSUM)) { @@ -6247,7 +6248,8 @@ static int ext4_remount(struct super_block *sb, int *flags, char *data) if ((sb->s_flags & SB_RDONLY) && !(old_sb_flags & SB_RDONLY) && sb_any_quota_suspended(sb)) dquot_resume(sb, -1); - percpu_down_write(&sbi->s_writepages_rwsem); + + alloc_ctx = ext4_writepages_down_write(sb); sb->s_flags = old_sb_flags; sbi->s_mount_opt = old_opts.s_mount_opt; sbi->s_mount_opt2 = old_opts.s_mount_opt2; @@ -6256,7 +6258,7 @@ static int ext4_remount(struct super_block *sb, int *flags, char *data) sbi->s_commit_interval = old_opts.s_commit_interval; sbi->s_min_batch_time = old_opts.s_min_batch_time; sbi->s_max_batch_time = old_opts.s_max_batch_time; - percpu_up_write(&sbi->s_writepages_rwsem); + ext4_writepages_up_write(sb, alloc_ctx); if (!test_opt(sb, BLOCK_VALIDITY) && sbi->s_system_blks) ext4_release_system_zone(sb); -- 2.31.1

2 1

[PATCH OLK-5.10] phonet: fix rtm_phonet_notify() skb allocation
by Wang Wensheng 11 Jun '24

11 Jun '24

From: Eric Dumazet <edumazet(a)google.com> mainline inclusion from mainline-v6.9 commit d8cac8568618dcb8a51af3db1103e8d4cc4aeea7 category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/I9U4L2 CVE: CVE-2024-36946 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- fill_route() stores three components in the skb: - struct rtmsg - RTA_DST (u8) - RTA_OIF (u32) Therefore, rtm_phonet_notify() should use NLMSG_ALIGN(sizeof(struct rtmsg)) + nla_total_size(1) + nla_total_size(4) Fixes: f062f41d0657 ("Phonet: routing table Netlink interface") Signed-off-by: Eric Dumazet <edumazet(a)google.com> Acked-by: Rémi Denis-Courmont <courmisch(a)gmail.com> Link: https://lore.kernel.org/r/20240502161700.1804476-1-edumazet@google.com Signed-off-by: Jakub Kicinski <kuba(a)kernel.org> Signed-off-by: Wang Wensheng <wangwensheng4(a)huawei.com> --- net/phonet/pn_netlink.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/phonet/pn_netlink.c b/net/phonet/pn_netlink.c index 59aebe296890..dd4c7e9a634f 100644 --- a/net/phonet/pn_netlink.c +++ b/net/phonet/pn_netlink.c @@ -193,7 +193,7 @@ void rtm_phonet_notify(int event, struct net_device *dev, u8 dst) struct sk_buff *skb; int err = -ENOBUFS; - skb = nlmsg_new(NLMSG_ALIGN(sizeof(struct ifaddrmsg)) + + skb = nlmsg_new(NLMSG_ALIGN(sizeof(struct rtmsg)) + nla_total_size(1) + nla_total_size(4), GFP_KERNEL); if (skb == NULL) goto errout; -- 2.17.1

2 1

[PATCH openEuler-1.0-LTS] i40e: Fix NULL pointer dereference in i40e_dbg_dump_desc
by Li Huafei 11 Jun '24

11 Jun '24

From: Norbert Zulinski <norbertx.zulinski(a)intel.com> stable inclusion from stable-v5.10.85 commit e5b7fb2198abc50058f1a29c395b004f76ab1c83 category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/I9S22G CVE: CVE-2021-47501 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… -------------------------------- commit 23ec111bf3549aae37140330c31a16abfc172421 upstream. When trying to dump VFs VSI RX/TX descriptors using debugfs there was a crash due to NULL pointer dereference in i40e_dbg_dump_desc. Added a check to i40e_dbg_dump_desc that checks if VSI type is correct for dumping RX/TX descriptors. Fixes: 02e9c290814c ("i40e: debugfs interface") Signed-off-by: Sylwester Dziedziuch <sylwesterx.dziedziuch(a)intel.com> Signed-off-by: Norbert Zulinski <norbertx.zulinski(a)intel.com> Signed-off-by: Mateusz Palczewski <mateusz.palczewski(a)intel.com> Tested-by: Gurucharan G <gurucharanx.g(a)intel.com> Signed-off-by: Tony Nguyen <anthony.l.nguyen(a)intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Conflicts: drivers/net/ethernet/intel/i40e/i40e_debugfs.c [ Conflicts caused by not merging 44ea803e2fa7 ("i40e: introduce new dump desc XDP command"). ] Signed-off-by: Li Huafei <lihuafei1(a)huawei.com> --- drivers/net/ethernet/intel/i40e/i40e_debugfs.c | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/drivers/net/ethernet/intel/i40e/i40e_debugfs.c b/drivers/net/ethernet/intel/i40e/i40e_debugfs.c index 56b911a5dd8be..3e6c6585012f9 100644 --- a/drivers/net/ethernet/intel/i40e/i40e_debugfs.c +++ b/drivers/net/ethernet/intel/i40e/i40e_debugfs.c @@ -506,6 +506,14 @@ static void i40e_dbg_dump_desc(int cnt, int vsi_seid, int ring_id, int desc_n, dev_info(&pf->pdev->dev, "vsi %d not found\n", vsi_seid); return; } + if (vsi->type != I40E_VSI_MAIN && + vsi->type != I40E_VSI_FDIR && + vsi->type != I40E_VSI_VMDQ2) { + dev_info(&pf->pdev->dev, + "vsi %d type %d descriptor rings not available\n", + vsi_seid, vsi->type); + return; + } if (ring_id >= vsi->num_queue_pairs || ring_id < 0) { dev_info(&pf->pdev->dev, "ring %d not found\n", ring_id); return; -- 2.25.1

2 1

[PATCH openEuler-22.03-LTS-SP1] hugetlbfs: fix hugetlbfs_statfs() locking
by Yifan Qiao 11 Jun '24

11 Jun '24

From: Mina Almasry <almasrymina(a)google.com> mainline inclusion from mainline-v5.19-rc1 commit 4b25f030ae69ba710eff587cabb4c57cb7e7a8a1 category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I9SY02 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… ---------------------------------------------------------------------- After commit db71ef79b59b ("hugetlb: make free_huge_page irq safe"), the subpool lock should be locked with spin_lock_irq() and all call sites was modified as such, except for the ones in hugetlbfs_statfs(). Link: https://lkml.kernel.org/r/20220429202207.3045-1-almasrymina@google.com Fixes: db71ef79b59b ("hugetlb: make free_huge_page irq safe") Signed-off-by: Mina Almasry <almasrymina(a)google.com> Reviewed-by: Mike Kravetz <mike.kravetz(a)oracle.com> Signed-off-by: Andrew Morton <akpm(a)linux-foundation.org> Signed-off-by: Yifan Qiao <qiaoyifan4(a)huawei.com> --- fs/hugetlbfs/inode.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/fs/hugetlbfs/inode.c b/fs/hugetlbfs/inode.c index 8274f096e89f..73a008dfe322 100644 --- a/fs/hugetlbfs/inode.c +++ b/fs/hugetlbfs/inode.c @@ -1122,12 +1122,12 @@ static int hugetlbfs_statfs(struct dentry *dentry, struct kstatfs *buf) if (sbinfo->spool) { long free_pages; - spin_lock(&sbinfo->spool->lock); + spin_lock_irq(&sbinfo->spool->lock); buf->f_blocks = sbinfo->spool->max_hpages; free_pages = sbinfo->spool->max_hpages - sbinfo->spool->used_hpages; buf->f_bavail = buf->f_bfree = free_pages; - spin_unlock(&sbinfo->spool->lock); + spin_unlock_irq(&sbinfo->spool->lock); buf->f_files = sbinfo->max_inodes; buf->f_ffree = sbinfo->free_inodes; } -- 2.39.2

2 1