X-Git-Url: http://git.pld-linux.org/?a=blobdiff_plain;f=kernel-aufs4.patch;h=6627f25989a16e09366f8bdfcc2bbc03a13af418;hb=b62f1e654d5b42c316353d1d563a5989d99219ee;hp=3d47d24a6bbc372ac19be99df3ae4bbe8d86384b;hpb=521ced186fdfe27c112e57dac8af155cddf13974;p=packages%2Fkernel.git diff --git a/kernel-aufs4.patch b/kernel-aufs4.patch index 3d47d24a..6627f259 100644 --- a/kernel-aufs4.patch +++ b/kernel-aufs4.patch @@ -1,19 +1,18 @@ +SPDX-License-Identifier: GPL-2.0 aufs4.x-rcN kbuild patch -diff --git a/fs/Kconfig b/fs/Kconfig -index 83eab52..31f16c4 100644 ---- a/fs/Kconfig -+++ b/fs/Kconfig -@@ -248,6 +248,7 @@ source "fs/pstore/Kconfig" +--- linux-5.2/fs/Kconfig~ 2019-07-08 00:41:56.000000000 +0200 ++++ linux-5.2/fs/Kconfig 2019-07-08 09:33:44.244881357 +0200 +@@ -261,6 +261,7 @@ source "fs/romfs/Kconfig" + source "fs/pstore/Kconfig" source "fs/sysv/Kconfig" source "fs/ufs/Kconfig" - source "fs/exofs/Kconfig" +source "fs/aufs/Kconfig" endif # MISC_FILESYSTEMS diff --git a/fs/Makefile b/fs/Makefile -index 7bbaca9..a026491 100644 +index 293733f61594..12d19d0de07a 100644 --- a/fs/Makefile +++ b/fs/Makefile @@ -128,3 +128,4 @@ obj-y += exofs/ # Multiple modules @@ -21,32 +20,21 @@ index 7bbaca9..a026491 100644 obj-$(CONFIG_PSTORE) += pstore/ obj-$(CONFIG_EFIVAR_FS) += efivarfs/ +obj-$(CONFIG_AUFS_FS) += aufs/ -diff --git a/include/uapi/linux/Kbuild b/include/uapi/linux/Kbuild -index dd9820b..d0435d9 100644 ---- a/include/uapi/linux/Kbuild -+++ b/include/uapi/linux/Kbuild -@@ -59,6 +59,7 @@ header-y += atmsvc.h - header-y += atm_tcp.h - header-y += atm_zatm.h - header-y += audit.h -+header-y += aufs_type.h - header-y += auto_fs4.h - header-y += auto_fs.h - header-y += auxvec.h +SPDX-License-Identifier: GPL-2.0 aufs4.x-rcN base patch diff --git a/MAINTAINERS b/MAINTAINERS -index 1b0a87f..76fd7cc 100644 +index dce5c099f43c..d10f05c88860 100644 --- a/MAINTAINERS +++ b/MAINTAINERS -@@ -2319,6 +2319,19 @@ F: include/linux/audit.h +@@ -2635,6 +2635,19 @@ F: include/linux/audit.h F: include/uapi/linux/audit.h F: kernel/audit* +AUFS (advanced multi layered unification filesystem) FILESYSTEM +M: "J. R. Okajima" -+L: linux-unionfs@vger.kernel.org +L: aufs-users@lists.sourceforge.net (members only) ++L: linux-unionfs@vger.kernel.org +W: http://aufs.sourceforge.net +T: git://github.com/sfjro/aufs4-linux.git +S: Supported @@ -58,13 +46,13 @@ index 1b0a87f..76fd7cc 100644 + AUXILIARY DISPLAY DRIVERS M: Miguel Ojeda Sandonis - W: http://miguelojeda.es/auxdisplay.htm + S: Maintained diff --git a/drivers/block/loop.c b/drivers/block/loop.c -index 0ecb646..cf39138 100644 +index cf5538942834..a271e17e4b10 100644 --- a/drivers/block/loop.c +++ b/drivers/block/loop.c -@@ -701,6 +701,24 @@ static inline int is_loop_device(struct file *file) - return i && S_ISBLK(i->i_mode) && MAJOR(i->i_rdev) == LOOP_MAJOR; +@@ -738,6 +738,24 @@ static int loop_change_fd(struct loop_device *lo, struct block_device *bdev, + return error; } +/* @@ -89,23 +77,23 @@ index 0ecb646..cf39138 100644 static ssize_t loop_attr_show(struct device *dev, char *page, diff --git a/fs/dcache.c b/fs/dcache.c -index 95d71ed..8ca5f09 100644 +index aac41adf4743..fc18967c2522 100644 --- a/fs/dcache.c +++ b/fs/dcache.c -@@ -1164,7 +1164,7 @@ enum d_walk_ret { +@@ -1247,7 +1247,7 @@ enum d_walk_ret { * - * The @enter() and @finish() callbacks are called with d_lock held. + * The @enter() callbacks are called with d_lock held. */ -static void d_walk(struct dentry *parent, void *data, +void d_walk(struct dentry *parent, void *data, - enum d_walk_ret (*enter)(void *, struct dentry *), - void (*finish)(void *)) + enum d_walk_ret (*enter)(void *, struct dentry *)) { + struct dentry *this_parent; diff --git a/fs/fcntl.c b/fs/fcntl.c -index be8fbe2..f5f4f94 100644 +index 083185174c6d..78234ee16784 100644 --- a/fs/fcntl.c +++ b/fs/fcntl.c -@@ -30,7 +30,7 @@ +@@ -32,7 +32,7 @@ #define SETFL_MASK (O_APPEND | O_NONBLOCK | O_NDELAY | O_DIRECT | O_NOATIME) @@ -114,7 +102,7 @@ index be8fbe2..f5f4f94 100644 { struct inode * inode = file_inode(filp); int error = 0; -@@ -61,6 +61,8 @@ static int setfl(int fd, struct file * filp, unsigned long arg) +@@ -63,6 +63,8 @@ static int setfl(int fd, struct file * filp, unsigned long arg) if (filp->f_op->check_flags) error = filp->f_op->check_flags(arg); @@ -124,25 +112,42 @@ index be8fbe2..f5f4f94 100644 return error; diff --git a/fs/inode.c b/fs/inode.c -index 88110fd..9a9ba3a 100644 +index 73432e64f874..4a1756b8b4bd 100644 --- a/fs/inode.c +++ b/fs/inode.c -@@ -1642,7 +1642,7 @@ EXPORT_SYMBOL(generic_update_time); +@@ -1657,7 +1657,7 @@ EXPORT_SYMBOL(generic_update_time); * This does the actual work of updating an inodes time or version. Must have * had called mnt_want_write() before calling this. */ --static int update_time(struct inode *inode, struct timespec *time, int flags) -+int update_time(struct inode *inode, struct timespec *time, int flags) +-static int update_time(struct inode *inode, struct timespec64 *time, int flags) ++int update_time(struct inode *inode, struct timespec64 *time, int flags) { - int (*update_time)(struct inode *, struct timespec *, int); + int (*update_time)(struct inode *, struct timespec64 *, int); +diff --git a/fs/namespace.c b/fs/namespace.c +index 678ef175d63a..9b6a3d0f87a1 100644 +--- a/fs/namespace.c ++++ b/fs/namespace.c +@@ -768,6 +768,12 @@ static inline int check_mnt(struct mount *mnt) + return mnt->mnt_ns == current->nsproxy->mnt_ns; + } + ++/* for aufs, CONFIG_AUFS_BR_FUSE */ ++int is_current_mnt_ns(struct vfsmount *mnt) ++{ ++ return check_mnt(real_mount(mnt)); ++} ++ + /* + * vfsmount lock must be held for write + */ diff --git a/fs/read_write.c b/fs/read_write.c -index c4f88af..ac71b96 100644 +index ff3c5e6f87cf..5c18a9e51f37 100644 --- a/fs/read_write.c +++ b/fs/read_write.c -@@ -513,6 +513,28 @@ ssize_t __vfs_write(struct file *file, const char __user *p, size_t count, +@@ -489,6 +489,28 @@ ssize_t __vfs_write(struct file *file, const char __user *p, size_t count, + return -EINVAL; } - EXPORT_SYMBOL(__vfs_write); +vfs_readf_t vfs_readf(struct file *file) +{ @@ -166,14 +171,14 @@ index c4f88af..ac71b96 100644 + return ERR_PTR(-ENOSYS); +} + - ssize_t __kernel_write(struct file *file, const char *buf, size_t count, loff_t *pos) + ssize_t __kernel_write(struct file *file, const void *buf, size_t count, loff_t *pos) { mm_segment_t old_fs; diff --git a/fs/splice.c b/fs/splice.c -index 006ba50..46c87af 100644 +index de2ede048473..5dcf77b8e1b2 100644 --- a/fs/splice.c +++ b/fs/splice.c -@@ -859,8 +859,8 @@ EXPORT_SYMBOL(generic_splice_sendpage); +@@ -837,8 +837,8 @@ EXPORT_SYMBOL(generic_splice_sendpage); /* * Attempt to initiate a splice from pipe to file. */ @@ -184,7 +189,7 @@ index 006ba50..46c87af 100644 { ssize_t (*splice_write)(struct pipe_inode_info *, struct file *, loff_t *, size_t, unsigned int); -@@ -876,9 +876,9 @@ static long do_splice_from(struct pipe_inode_info *pipe, struct file *out, +@@ -854,9 +854,9 @@ static long do_splice_from(struct pipe_inode_info *pipe, struct file *out, /* * Attempt to initiate a splice from a file to a pipe. */ @@ -198,10 +203,10 @@ index 006ba50..46c87af 100644 ssize_t (*splice_read)(struct file *, loff_t *, struct pipe_inode_info *, size_t, unsigned int); diff --git a/fs/sync.c b/fs/sync.c -index 11ba023..abf6a5d 100644 +index b54e0541ad89..28607828e96f 100644 --- a/fs/sync.c +++ b/fs/sync.c -@@ -27,7 +27,7 @@ +@@ -28,7 +28,7 @@ * wait == 1 case since in that case write_inode() functions do * sync_dirty_buffer() and thus effectively write one block at a time. */ @@ -210,31 +215,19 @@ index 11ba023..abf6a5d 100644 { if (wait) sync_inodes_sb(sb); -diff --git a/include/linux/file.h b/include/linux/file.h -index 61eb82c..e700888 100644 ---- a/include/linux/file.h -+++ b/include/linux/file.h -@@ -19,6 +19,7 @@ struct dentry; - struct path; - extern struct file *alloc_file(const struct path *, fmode_t mode, - const struct file_operations *fop); -+extern struct file *get_empty_filp(void); - - static inline void fput_light(struct file *file, int fput_needed) - { diff --git a/include/linux/fs.h b/include/linux/fs.h -index 7251f7b..3668627 100644 +index 29d8e2cfed0e..5db265e7d35a 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h -@@ -1246,6 +1246,7 @@ extern void fasync_free(struct fasync_struct *); +@@ -1310,6 +1310,7 @@ extern void fasync_free(struct fasync_struct *); /* can be called from interrupts */ extern void kill_fasync(struct fasync_struct **, int, int); +extern int setfl(int fd, struct file * filp, unsigned long arg); extern void __f_setown(struct file *filp, struct pid *, enum pid_type, int force); - extern void f_setown(struct file *filp, unsigned long arg, int force); + extern int f_setown(struct file *filp, unsigned long arg, int force); extern void f_delown(struct file *filp); -@@ -1673,6 +1674,7 @@ struct file_operations { +@@ -1803,6 +1804,7 @@ struct file_operations { ssize_t (*sendpage) (struct file *, struct page *, int, size_t, loff_t *, int); unsigned long (*get_unmapped_area)(struct file *, unsigned long, unsigned long, unsigned long, unsigned long); int (*check_flags)(int); @@ -242,7 +235,7 @@ index 7251f7b..3668627 100644 int (*flock) (struct file *, int, struct file_lock *); ssize_t (*splice_write)(struct pipe_inode_info *, struct file *, loff_t *, size_t, unsigned int); ssize_t (*splice_read)(struct file *, loff_t *, struct pipe_inode_info *, size_t, unsigned int); -@@ -1749,6 +1751,12 @@ ssize_t rw_copy_check_uvector(int type, const struct iovec __user * uvector, +@@ -1873,6 +1875,12 @@ ssize_t rw_copy_check_uvector(int type, const struct iovec __user * uvector, struct iovec *fast_pointer, struct iovec **ret_pointer); @@ -253,17 +246,17 @@ index 7251f7b..3668627 100644 +vfs_writef_t vfs_writef(struct file *file); + extern ssize_t __vfs_read(struct file *, char __user *, size_t, loff_t *); - extern ssize_t __vfs_write(struct file *, const char __user *, size_t, loff_t *); extern ssize_t vfs_read(struct file *, char __user *, size_t, loff_t *); -@@ -2127,6 +2135,7 @@ extern int current_umask(void); + extern ssize_t vfs_write(struct file *, const char __user *, size_t, loff_t *); +@@ -2298,6 +2306,7 @@ extern int current_umask(void); extern void ihold(struct inode * inode); extern void iput(struct inode *); - extern int generic_update_time(struct inode *, struct timespec *, int); -+extern int update_time(struct inode *, struct timespec *, int); + extern int generic_update_time(struct inode *, struct timespec64 *, int); ++extern int update_time(struct inode *, struct timespec64 *, int); /* /sys/fs */ extern struct kobject *fs_kobj; -@@ -2407,6 +2416,7 @@ static inline bool sb_is_blkdev_sb(struct super_block *sb) +@@ -2585,6 +2594,7 @@ static inline bool sb_is_blkdev_sb(struct super_block *sb) return false; } #endif @@ -271,11 +264,51 @@ index 7251f7b..3668627 100644 extern int sync_filesystem(struct super_block *); extern const struct file_operations def_blk_fops; extern const struct file_operations def_chr_fops; +diff --git a/include/linux/lockdep.h b/include/linux/lockdep.h +index c5335df2372f..4e48a5059536 100644 +--- a/include/linux/lockdep.h ++++ b/include/linux/lockdep.h +@@ -306,6 +306,8 @@ static inline int lockdep_match_key(struct lockdep_map *lock, + return lock->key == key; + } + ++struct lock_class *lockdep_hlock_class(struct held_lock *hlock); ++ + /* + * Acquire a lock. + * +@@ -432,6 +434,7 @@ struct lockdep_map { }; + + #define lockdep_depth(tsk) (0) + ++#define lockdep_is_held(lock) (1) + #define lockdep_is_held_type(l, r) (1) + + #define lockdep_assert_held(l) do { (void)(l); } while (0) +diff --git a/include/linux/mnt_namespace.h b/include/linux/mnt_namespace.h +index 35942084cd40..24f5fd1a789d 100644 +--- a/include/linux/mnt_namespace.h ++++ b/include/linux/mnt_namespace.h +@@ -6,11 +6,14 @@ + struct mnt_namespace; + struct fs_struct; + struct user_namespace; ++struct vfsmount; + + extern struct mnt_namespace *copy_mnt_ns(unsigned long, struct mnt_namespace *, + struct user_namespace *, struct fs_struct *); + extern void put_mnt_ns(struct mnt_namespace *ns); + ++extern int is_current_mnt_ns(struct vfsmount *mnt); ++ + extern const struct file_operations proc_mounts_operations; + extern const struct file_operations proc_mountinfo_operations; + extern const struct file_operations proc_mountstats_operations; diff --git a/include/linux/splice.h b/include/linux/splice.h -index 00a2116..1f0a4a2 100644 +index 74b4911ac16d..19789fbea567 100644 --- a/include/linux/splice.h +++ b/include/linux/splice.h -@@ -86,4 +86,10 @@ extern void spd_release_page(struct splice_pipe_desc *, unsigned int); +@@ -87,4 +87,10 @@ extern void splice_shrink_spd(struct splice_pipe_desc *); extern const struct pipe_buf_operations page_cache_pipe_buf_ops; extern const struct pipe_buf_operations default_pipe_buf_ops; @@ -286,13 +319,35 @@ index 00a2116..1f0a4a2 100644 + struct pipe_inode_info *pipe, size_t len, + unsigned int flags); #endif +diff --git a/kernel/locking/lockdep.c b/kernel/locking/lockdep.c +index 95932333a48b..516ab7da26fb 100644 +--- a/kernel/locking/lockdep.c ++++ b/kernel/locking/lockdep.c +@@ -143,7 +143,7 @@ static + #endif + struct lock_class lock_classes[MAX_LOCKDEP_KEYS]; + +-static inline struct lock_class *hlock_class(struct held_lock *hlock) ++inline struct lock_class *lockdep_hlock_class(struct held_lock *hlock) + { + if (!hlock->class_idx) { + /* +@@ -154,6 +154,7 @@ static inline struct lock_class *hlock_class(struct held_lock *hlock) + } + return lock_classes + hlock->class_idx - 1; + } ++#define hlock_class(hlock) lockdep_hlock_class(hlock) + + #ifdef CONFIG_LOCK_STAT + static DEFINE_PER_CPU(struct lock_class_stats[MAX_LOCKDEP_KEYS], cpu_lock_stats); +SPDX-License-Identifier: GPL-2.0 aufs4.x-rcN mmap patch diff --git a/fs/proc/base.c b/fs/proc/base.c -index c87b6b9..6b71643 100644 +index f5ed9512d193..e8f8e59d93bb 100644 --- a/fs/proc/base.c +++ b/fs/proc/base.c -@@ -1946,7 +1946,7 @@ static int map_files_get_link(struct dentry *dentry, struct path *path) +@@ -2014,7 +2014,7 @@ static int map_files_get_link(struct dentry *dentry, struct path *path) down_read(&mm->mmap_sem); vma = find_exact_vma(mm, vm_start, vm_end); if (vma && vma->vm_file) { @@ -302,7 +357,7 @@ index c87b6b9..6b71643 100644 rc = 0; } diff --git a/fs/proc/nommu.c b/fs/proc/nommu.c -index 7563437..7c0dc0f 100644 +index 3b63be64e436..fb9913bf3d10 100644 --- a/fs/proc/nommu.c +++ b/fs/proc/nommu.c @@ -45,7 +45,10 @@ static int nommu_region_show(struct seq_file *m, struct vm_region *region) @@ -318,10 +373,10 @@ index 7563437..7c0dc0f 100644 ino = inode->i_ino; } diff --git a/fs/proc/task_mmu.c b/fs/proc/task_mmu.c -index f08bd31..e1b9a37 100644 +index 85b0ef890b28..0f7fb8138160 100644 --- a/fs/proc/task_mmu.c +++ b/fs/proc/task_mmu.c -@@ -292,7 +292,10 @@ show_map_vma(struct seq_file *m, struct vm_area_struct *vma, int is_pid) +@@ -305,7 +305,10 @@ show_map_vma(struct seq_file *m, struct vm_area_struct *vma) const char *name = NULL; if (file) { @@ -333,7 +388,7 @@ index f08bd31..e1b9a37 100644 dev = inode->i_sb->s_dev; ino = inode->i_ino; pgoff = ((loff_t)vma->vm_pgoff) << PAGE_SHIFT; -@@ -1629,7 +1632,7 @@ static int show_numa_map(struct seq_file *m, void *v, int is_pid) +@@ -1740,7 +1743,7 @@ static int show_numa_map(struct seq_file *m, void *v) struct proc_maps_private *proc_priv = &numa_priv->proc_maps; struct vm_area_struct *vma = v; struct numa_maps *md = &numa_priv->md; @@ -343,10 +398,10 @@ index f08bd31..e1b9a37 100644 struct mm_walk walk = { .hugetlb_entry = gather_hugetlb_stats, diff --git a/fs/proc/task_nommu.c b/fs/proc/task_nommu.c -index 23266694..58e59b6 100644 +index 0b63d68dedb2..400d1c594ceb 100644 --- a/fs/proc/task_nommu.c +++ b/fs/proc/task_nommu.c -@@ -157,7 +157,10 @@ static int nommu_vma_show(struct seq_file *m, struct vm_area_struct *vma, +@@ -155,7 +155,10 @@ static int nommu_vma_show(struct seq_file *m, struct vm_area_struct *vma) file = vma->vm_file; if (file) { @@ -359,12 +414,12 @@ index 23266694..58e59b6 100644 ino = inode->i_ino; pgoff = (loff_t)vma->vm_pgoff << PAGE_SHIFT; diff --git a/include/linux/mm.h b/include/linux/mm.h -index 00a8fa7..a90742b 100644 +index 80bb6408fe73..59a1ea611ba7 100644 --- a/include/linux/mm.h +++ b/include/linux/mm.h -@@ -1294,6 +1294,28 @@ static inline int fixup_user_fault(struct task_struct *tsk, +@@ -1518,6 +1518,28 @@ static inline void unmap_shared_mapping_range(struct address_space *mapping, + unmap_mapping_range(mapping, holebegin, holelen, 0); } - #endif +extern void vma_do_file_update_time(struct vm_area_struct *, const char[], int); +extern struct file *vma_do_pr_or_file(struct vm_area_struct *, const char[], @@ -388,14 +443,14 @@ index 00a8fa7..a90742b 100644 +#define vmr_fput(region) vmr_do_fput(region, __func__, __LINE__) +#endif /* !CONFIG_MMU */ + - extern int access_process_vm(struct task_struct *tsk, unsigned long addr, void *buf, int len, - unsigned int gup_flags); + extern int access_process_vm(struct task_struct *tsk, unsigned long addr, + void *buf, int len, unsigned int gup_flags); extern int access_remote_vm(struct mm_struct *mm, unsigned long addr, diff --git a/include/linux/mm_types.h b/include/linux/mm_types.h -index f60f45f..38c5f76 100644 +index 2c471a2c43fa..c6e7aa6996ff 100644 --- a/include/linux/mm_types.h +++ b/include/linux/mm_types.h -@@ -259,6 +259,7 @@ struct vm_region { +@@ -244,6 +244,7 @@ struct vm_region { unsigned long vm_top; /* region allocated to here */ unsigned long vm_pgoff; /* the offset in vm_file corresponding to vm_start */ struct file *vm_file; /* the backing file or NULL */ @@ -403,19 +458,19 @@ index f60f45f..38c5f76 100644 int vm_usage; /* region usage count (access under nommu_region_sem) */ bool vm_icache_flushed : 1; /* true if the icache has been flushed for -@@ -333,6 +334,7 @@ struct vm_area_struct { +@@ -318,6 +319,7 @@ struct vm_area_struct { unsigned long vm_pgoff; /* Offset (within vm_file) in PAGE_SIZE units */ struct file * vm_file; /* File we map to (can be NULL). */ + struct file *vm_prfile; /* shadow of vm_file */ void * vm_private_data; /* was vm_pte (shared mem) */ - #ifndef CONFIG_MMU + atomic_long_t swap_readahead_info; diff --git a/kernel/fork.c b/kernel/fork.c -index 6c463c80..24ae37c 100644 +index b69248e6f0e0..099b89978dd3 100644 --- a/kernel/fork.c +++ b/kernel/fork.c -@@ -641,7 +641,7 @@ static __latent_entropy int dup_mmap(struct mm_struct *mm, +@@ -547,7 +547,7 @@ static __latent_entropy int dup_mmap(struct mm_struct *mm, struct inode *inode = file_inode(file); struct address_space *mapping = file->f_mapping; @@ -425,24 +480,24 @@ index 6c463c80..24ae37c 100644 atomic_dec(&inode->i_writecount); i_mmap_lock_write(mapping); diff --git a/mm/Makefile b/mm/Makefile -index 026f6a8..723da17 100644 +index d210cc9d6f80..e77e80ce7298 100644 --- a/mm/Makefile +++ b/mm/Makefile -@@ -39,7 +39,7 @@ obj-y := filemap.o mempool.o oom_kill.o \ +@@ -39,7 +39,7 @@ obj-y := filemap.o mempool.o oom_kill.o fadvise.o \ mm_init.o mmu_context.o percpu.o slab_common.o \ - compaction.o vmacache.o swap_slots.o \ + compaction.o vmacache.o \ interval_tree.o list_lru.o workingset.o \ - debug.o $(mmu-y) + prfile.o debug.o $(mmu-y) obj-y += init-mm.o - + obj-y += memblock.o diff --git a/mm/filemap.c b/mm/filemap.c -index 1694623..68cd758 100644 +index 9f5e323e883e..90db684a7cf4 100644 --- a/mm/filemap.c +++ b/mm/filemap.c -@@ -2406,7 +2406,7 @@ int filemap_page_mkwrite(struct vm_fault *vmf) - int ret = VM_FAULT_LOCKED; +@@ -2677,7 +2677,7 @@ vm_fault_t filemap_page_mkwrite(struct vm_fault *vmf) + vm_fault_t ret = VM_FAULT_LOCKED; sb_start_pagefault(inode->i_sb); - file_update_time(vmf->vma->vm_file); @@ -451,19 +506,19 @@ index 1694623..68cd758 100644 if (page->mapping != inode->i_mapping) { unlock_page(page); diff --git a/mm/mmap.c b/mm/mmap.c -index bfbe885..8bd32f9 100644 +index fc1809b1bed6..8f62e47a60ea 100644 --- a/mm/mmap.c +++ b/mm/mmap.c -@@ -170,7 +170,7 @@ static struct vm_area_struct *remove_vma(struct vm_area_struct *vma) +@@ -180,7 +180,7 @@ static struct vm_area_struct *remove_vma(struct vm_area_struct *vma) if (vma->vm_ops && vma->vm_ops->close) vma->vm_ops->close(vma); if (vma->vm_file) - fput(vma->vm_file); + vma_fput(vma); mpol_put(vma_policy(vma)); - kmem_cache_free(vm_area_cachep, vma); + vm_area_free(vma); return next; -@@ -881,7 +881,7 @@ int __vma_adjust(struct vm_area_struct *vma, unsigned long start, +@@ -929,7 +929,7 @@ int __vma_adjust(struct vm_area_struct *vma, unsigned long start, if (remove_next) { if (file) { uprobe_munmap(next, next->vm_start, next->vm_end); @@ -472,7 +527,7 @@ index bfbe885..8bd32f9 100644 } if (next->anon_vma) anon_vma_merge(vma, next); -@@ -1731,8 +1731,8 @@ unsigned long mmap_region(struct file *file, unsigned long addr, +@@ -1845,8 +1845,8 @@ unsigned long mmap_region(struct file *file, unsigned long addr, return addr; unmap_and_free_vma: @@ -482,7 +537,7 @@ index bfbe885..8bd32f9 100644 /* Undo any partial mapping done by a device driver. */ unmap_region(mm, vma, prev, vma->vm_start, vma->vm_end); -@@ -2537,7 +2537,7 @@ int __split_vma(struct mm_struct *mm, struct vm_area_struct *vma, +@@ -2675,7 +2675,7 @@ int __split_vma(struct mm_struct *mm, struct vm_area_struct *vma, goto out_free_mpol; if (new->vm_file) @@ -491,7 +546,7 @@ index bfbe885..8bd32f9 100644 if (new->vm_ops && new->vm_ops->open) new->vm_ops->open(new); -@@ -2556,7 +2556,7 @@ int __split_vma(struct mm_struct *mm, struct vm_area_struct *vma, +@@ -2694,7 +2694,7 @@ int __split_vma(struct mm_struct *mm, struct vm_area_struct *vma, if (new->vm_ops && new->vm_ops->close) new->vm_ops->close(new); if (new->vm_file) @@ -500,16 +555,16 @@ index bfbe885..8bd32f9 100644 unlink_anon_vmas(new); out_free_mpol: mpol_put(vma_policy(new)); -@@ -2710,7 +2710,7 @@ SYSCALL_DEFINE5(remap_file_pages, unsigned long, start, unsigned long, size, +@@ -2884,7 +2884,7 @@ SYSCALL_DEFINE5(remap_file_pages, unsigned long, start, unsigned long, size, struct vm_area_struct *vma; unsigned long populate = 0; unsigned long ret = -EINVAL; - struct file *file; + struct file *file, *prfile; - pr_warn_once("%s (%d) uses deprecated remap_file_pages() syscall. See Documentation/vm/remap_file_pages.txt.\n", + pr_warn_once("%s (%d) uses deprecated remap_file_pages() syscall. See Documentation/vm/remap_file_pages.rst.\n", current->comm, current->pid); -@@ -2785,10 +2785,27 @@ SYSCALL_DEFINE5(remap_file_pages, unsigned long, start, unsigned long, size, +@@ -2959,10 +2959,27 @@ SYSCALL_DEFINE5(remap_file_pages, unsigned long, start, unsigned long, size, } } @@ -538,7 +593,7 @@ index bfbe885..8bd32f9 100644 out: up_write(&mm->mmap_sem); if (populate) -@@ -3079,7 +3096,7 @@ struct vm_area_struct *copy_vma(struct vm_area_struct **vmap, +@@ -3252,7 +3269,7 @@ struct vm_area_struct *copy_vma(struct vm_area_struct **vmap, if (anon_vma_clone(new_vma, vma)) goto out_free_mempol; if (new_vma->vm_file) @@ -548,10 +603,10 @@ index bfbe885..8bd32f9 100644 new_vma->vm_ops->open(new_vma); vma_link(mm, new_vma, prev, rb_link, rb_parent); diff --git a/mm/nommu.c b/mm/nommu.c -index 2d131b9..5dc9a19 100644 +index 749276beb109..c7957f06fbc4 100644 --- a/mm/nommu.c +++ b/mm/nommu.c -@@ -637,7 +637,7 @@ static void __put_nommu_region(struct vm_region *region) +@@ -625,7 +625,7 @@ static void __put_nommu_region(struct vm_region *region) up_write(&nommu_region_sem); if (region->vm_file) @@ -560,16 +615,16 @@ index 2d131b9..5dc9a19 100644 /* IO memory and memory shared directly out of the pagecache * from ramfs/tmpfs mustn't be released here */ -@@ -795,7 +795,7 @@ static void delete_vma(struct mm_struct *mm, struct vm_area_struct *vma) +@@ -763,7 +763,7 @@ static void delete_vma(struct mm_struct *mm, struct vm_area_struct *vma) if (vma->vm_ops && vma->vm_ops->close) vma->vm_ops->close(vma); if (vma->vm_file) - fput(vma->vm_file); + vma_fput(vma); put_nommu_region(vma->vm_region); - kmem_cache_free(vm_area_cachep, vma); + vm_area_free(vma); } -@@ -1322,7 +1322,7 @@ unsigned long do_mmap(struct file *file, +@@ -1286,7 +1286,7 @@ unsigned long do_mmap(struct file *file, goto error_just_free; } } @@ -578,7 +633,7 @@ index 2d131b9..5dc9a19 100644 kmem_cache_free(vm_region_jar, region); region = pregion; result = start; -@@ -1397,10 +1397,10 @@ unsigned long do_mmap(struct file *file, +@@ -1361,10 +1361,10 @@ unsigned long do_mmap(struct file *file, up_write(&nommu_region_sem); error: if (region->vm_file) @@ -588,23 +643,24 @@ index 2d131b9..5dc9a19 100644 if (vma->vm_file) - fput(vma->vm_file); + vma_fput(vma); - kmem_cache_free(vm_area_cachep, vma); + vm_area_free(vma); return ret; diff --git a/mm/prfile.c b/mm/prfile.c new file mode 100644 -index 0000000..b323b8a +index 000000000000..024cdcfae1b1 --- /dev/null +++ b/mm/prfile.c @@ -0,0 +1,86 @@ ++// SPDX-License-Identifier: GPL-2.0 +/* -+ * Mainly for aufs which mmap(2) diffrent file and wants to print different path -+ * in /proc/PID/maps. ++ * Mainly for aufs which mmap(2) different file and wants to print different ++ * path in /proc/PID/maps. + * Call these functions via macros defined in linux/mm.h. + * + * See Documentation/filesystems/aufs/design/06mmap.txt + * -+ * Copyright (c) 2014 Junjro R. Okajima ++ * Copyright (c) 2014-2019 Junjro R. Okajima + * Copyright (c) 2014 Ian Campbell + */ + @@ -618,8 +674,7 @@ index 0000000..b323b8a +{ +#ifdef PRFILE_TRACE + if (pr) -+ pr_info("%s:%d: %s, %s\n", func, line, func2, -+ f ? (char *)f->f_path.dentry->d_name.name : "(null)"); ++ pr_info("%s:%d: %s, %pD2\n", func, line, func2, f); +#endif +} + @@ -683,13 +738,14 @@ index 0000000..b323b8a + fput(pr); +} +#endif /* !CONFIG_MMU */ +SPDX-License-Identifier: GPL-2.0 aufs4.x-rcN standalone patch diff --git a/fs/dcache.c b/fs/dcache.c -index 8ca5f09..b1ff5be 100644 +index fc18967c2522..50eb587c1253 100644 --- a/fs/dcache.c +++ b/fs/dcache.c -@@ -1272,6 +1272,7 @@ void d_walk(struct dentry *parent, void *data, +@@ -1352,6 +1352,7 @@ void d_walk(struct dentry *parent, void *data, seq = 1; goto again; } @@ -697,7 +753,7 @@ index 8ca5f09..b1ff5be 100644 struct check_mount { struct vfsmount *mnt; -@@ -2864,6 +2865,7 @@ void d_exchange(struct dentry *dentry1, struct dentry *dentry2) +@@ -2845,6 +2846,7 @@ void d_exchange(struct dentry *dentry1, struct dentry *dentry2) write_sequnlock(&rename_lock); } @@ -706,7 +762,7 @@ index 8ca5f09..b1ff5be 100644 /** * d_ancestor - search for an ancestor diff --git a/fs/exec.c b/fs/exec.c -index 65145a3..8d35776 100644 +index fb72d36f7823..7ce68004cbf8 100644 --- a/fs/exec.c +++ b/fs/exec.c @@ -109,6 +109,7 @@ bool path_noexec(const struct path *path) @@ -718,10 +774,10 @@ index 65145a3..8d35776 100644 #ifdef CONFIG_USELIB /* diff --git a/fs/fcntl.c b/fs/fcntl.c -index f5f4f94..c671660 100644 +index 78234ee16784..2072f690b121 100644 --- a/fs/fcntl.c +++ b/fs/fcntl.c -@@ -83,6 +83,7 @@ int setfl(int fd, struct file * filp, unsigned long arg) +@@ -85,6 +85,7 @@ int setfl(int fd, struct file * filp, unsigned long arg) out: return error; } @@ -730,18 +786,18 @@ index f5f4f94..c671660 100644 static void f_modown(struct file *filp, struct pid *pid, enum pid_type type, int force) diff --git a/fs/file_table.c b/fs/file_table.c -index 954d510..4fb5b10 100644 +index 5679e7fcb6b0..961eec3df1eb 100644 --- a/fs/file_table.c +++ b/fs/file_table.c -@@ -148,6 +148,7 @@ struct file *get_empty_filp(void) +@@ -161,6 +161,7 @@ struct file *alloc_empty_file(int flags, const struct cred *cred) } return ERR_PTR(-ENFILE); } -+EXPORT_SYMBOL_GPL(get_empty_filp); ++EXPORT_SYMBOL_GPL(alloc_empty_file); - /** - * alloc_file - allocate and initialize a 'struct file' -@@ -259,6 +260,7 @@ void flush_delayed_fput(void) + /* + * Variant of alloc_empty_file() that doesn't check and modify nr_files. +@@ -323,6 +324,7 @@ void flush_delayed_fput(void) { delayed_fput(NULL); } @@ -749,27 +805,19 @@ index 954d510..4fb5b10 100644 static DECLARE_DELAYED_WORK(delayed_fput_work, delayed_fput); -@@ -301,6 +303,7 @@ void __fput_sync(struct file *file) +@@ -365,6 +367,7 @@ void __fput_sync(struct file *file) } EXPORT_SYMBOL(fput); +EXPORT_SYMBOL_GPL(__fput_sync); - void put_filp(struct file *file) - { -@@ -309,6 +312,7 @@ void put_filp(struct file *file) - file_free(file); - } - } -+EXPORT_SYMBOL_GPL(put_filp); - void __init files_init(void) - { + { diff --git a/fs/inode.c b/fs/inode.c -index 9a9ba3a..a3a18d8 100644 +index 4a1756b8b4bd..6e51d5f9b9f3 100644 --- a/fs/inode.c +++ b/fs/inode.c -@@ -1651,6 +1651,7 @@ int update_time(struct inode *inode, struct timespec *time, int flags) +@@ -1666,6 +1666,7 @@ int update_time(struct inode *inode, struct timespec64 *time, int flags) return update_time(inode, time, flags); } @@ -778,10 +826,10 @@ index 9a9ba3a..a3a18d8 100644 /** * touch_atime - update the access time diff --git a/fs/namespace.c b/fs/namespace.c -index cc1375ef..9b4c67c 100644 +index 9b6a3d0f87a1..8c4d3d56f2df 100644 --- a/fs/namespace.c +++ b/fs/namespace.c -@@ -465,6 +465,7 @@ void __mnt_drop_write(struct vfsmount *mnt) +@@ -434,6 +434,7 @@ void __mnt_drop_write(struct vfsmount *mnt) mnt_dec_writers(real_mount(mnt)); preempt_enable(); } @@ -789,7 +837,15 @@ index cc1375ef..9b4c67c 100644 /** * mnt_drop_write - give up write access to a mount -@@ -1884,6 +1885,7 @@ int iterate_mounts(int (*f)(struct vfsmount *, void *), void *arg, +@@ -773,6 +774,7 @@ int is_current_mnt_ns(struct vfsmount *mnt) + { + return check_mnt(real_mount(mnt)); + } ++EXPORT_SYMBOL_GPL(is_current_mnt_ns); + + /* + * vfsmount lock must be held for write +@@ -1842,6 +1844,7 @@ int iterate_mounts(int (*f)(struct vfsmount *, void *), void *arg, } return 0; } @@ -798,34 +854,26 @@ index cc1375ef..9b4c67c 100644 static void cleanup_group_ids(struct mount *mnt, struct mount *end) { diff --git a/fs/notify/group.c b/fs/notify/group.c -index fbe3cbe..bdfc61e 100644 +index c03b83662876..94d210ca384a 100644 --- a/fs/notify/group.c +++ b/fs/notify/group.c -@@ -22,6 +22,7 @@ - #include - #include - #include -+#include - - #include - #include "fsnotify.h" -@@ -100,6 +101,7 @@ void fsnotify_get_group(struct fsnotify_group *group) +@@ -112,6 +112,7 @@ void fsnotify_get_group(struct fsnotify_group *group) { - atomic_inc(&group->refcnt); + refcount_inc(&group->refcnt); } +EXPORT_SYMBOL_GPL(fsnotify_get_group); /* * Drop a reference to a group. Free it if it's through. -@@ -109,6 +111,7 @@ void fsnotify_put_group(struct fsnotify_group *group) - if (atomic_dec_and_test(&group->refcnt)) +@@ -121,6 +122,7 @@ void fsnotify_put_group(struct fsnotify_group *group) + if (refcount_dec_and_test(&group->refcnt)) fsnotify_final_destroy_group(group); } +EXPORT_SYMBOL_GPL(fsnotify_put_group); /* * Create a new fsnotify_group and hold a reference for the group returned. -@@ -137,6 +140,7 @@ struct fsnotify_group *fsnotify_alloc_group(const struct fsnotify_ops *ops) +@@ -150,6 +152,7 @@ struct fsnotify_group *fsnotify_alloc_group(const struct fsnotify_ops *ops) return group; } @@ -834,43 +882,43 @@ index fbe3cbe..bdfc61e 100644 int fsnotify_fasync(int fd, struct file *file, int on) { diff --git a/fs/notify/mark.c b/fs/notify/mark.c -index 6043306..fdb50e4 100644 +index d2dd16cb5989..cf709b7d611a 100644 --- a/fs/notify/mark.c +++ b/fs/notify/mark.c -@@ -113,6 +113,7 @@ void fsnotify_put_mark(struct fsnotify_mark *mark) - mark->free_mark(mark); - } +@@ -289,6 +289,7 @@ void fsnotify_put_mark(struct fsnotify_mark *mark) + queue_delayed_work(system_unbound_wq, &reaper_work, + FSNOTIFY_REAPER_DELAY); } +EXPORT_SYMBOL_GPL(fsnotify_put_mark); - /* Calculate mask of events for a list of marks */ - u32 fsnotify_recalc_mask(struct hlist_head *head) -@@ -230,6 +231,7 @@ void fsnotify_destroy_mark(struct fsnotify_mark *mark, + /* + * Get mark reference when we found the mark via lockless traversal of object +@@ -443,6 +444,7 @@ void fsnotify_destroy_mark(struct fsnotify_mark *mark, mutex_unlock(&group->mark_mutex); fsnotify_free_mark(mark); } +EXPORT_SYMBOL_GPL(fsnotify_destroy_mark); - void fsnotify_destroy_marks(struct hlist_head *head, spinlock_t *lock) - { -@@ -415,6 +417,7 @@ int fsnotify_add_mark_locked(struct fsnotify_mark *mark, - + /* + * Sorting function for lists of fsnotify marks. +@@ -658,6 +660,7 @@ int fsnotify_add_mark(struct fsnotify_mark *mark, fsnotify_connp_t *connp, + mutex_unlock(&group->mark_mutex); return ret; } +EXPORT_SYMBOL_GPL(fsnotify_add_mark); - int fsnotify_add_mark(struct fsnotify_mark *mark, struct fsnotify_group *group, - struct inode *inode, struct vfsmount *mnt, int allow_dups) -@@ -521,6 +524,7 @@ void fsnotify_init_mark(struct fsnotify_mark *mark, - atomic_set(&mark->refcnt, 1); - mark->free_mark = free_mark; + /* + * Given a list of marks, find the mark associated with given group. If found +@@ -781,6 +784,7 @@ void fsnotify_init_mark(struct fsnotify_mark *mark, + fsnotify_get_group(group); + mark->group = group; } +EXPORT_SYMBOL_GPL(fsnotify_init_mark); /* * Destroy all marks in destroy_list, waits for SRCU period to finish before diff --git a/fs/open.c b/fs/open.c -index 949cef2..9a892fb8 100644 +index 0285ce7dbd51..cb81623a8b09 100644 --- a/fs/open.c +++ b/fs/open.c @@ -64,6 +64,7 @@ int do_truncate(struct dentry *dentry, loff_t length, unsigned int time_attrs, @@ -881,19 +929,19 @@ index 949cef2..9a892fb8 100644 long vfs_truncate(const struct path *path, loff_t length) { -@@ -693,6 +694,7 @@ int open_check_o_direct(struct file *f) - } - return 0; - } -+EXPORT_SYMBOL_GPL(open_check_o_direct); - - static int do_dentry_open(struct file *f, - struct inode *inode, diff --git a/fs/read_write.c b/fs/read_write.c -index ac71b96..b69e687 100644 +index 5c18a9e51f37..542e59cd8d27 100644 --- a/fs/read_write.c +++ b/fs/read_write.c -@@ -523,6 +523,7 @@ vfs_readf_t vfs_readf(struct file *file) +@@ -459,6 +459,7 @@ ssize_t vfs_read(struct file *file, char __user *buf, size_t count, loff_t *pos) + + return ret; + } ++EXPORT_SYMBOL_GPL(vfs_read); + + static ssize_t new_sync_write(struct file *filp, const char __user *buf, size_t len, loff_t *ppos) + { +@@ -499,6 +500,7 @@ vfs_readf_t vfs_readf(struct file *file) return new_sync_read; return ERR_PTR(-ENOSYS); } @@ -901,19 +949,27 @@ index ac71b96..b69e687 100644 vfs_writef_t vfs_writef(struct file *file) { -@@ -534,6 +535,7 @@ vfs_writef_t vfs_writef(struct file *file) +@@ -510,6 +512,7 @@ vfs_writef_t vfs_writef(struct file *file) return new_sync_write; return ERR_PTR(-ENOSYS); } +EXPORT_SYMBOL_GPL(vfs_writef); - ssize_t __kernel_write(struct file *file, const char *buf, size_t count, loff_t *pos) + ssize_t __kernel_write(struct file *file, const void *buf, size_t count, loff_t *pos) + { +@@ -579,6 +582,7 @@ ssize_t vfs_write(struct file *file, const char __user *buf, size_t count, loff_ + + return ret; + } ++EXPORT_SYMBOL_GPL(vfs_write); + + static inline loff_t file_pos_read(struct file *file) { diff --git a/fs/splice.c b/fs/splice.c -index 46c87af..0efa652 100644 +index 5dcf77b8e1b2..63fe2652c67d 100644 --- a/fs/splice.c +++ b/fs/splice.c -@@ -872,6 +872,7 @@ long do_splice_from(struct pipe_inode_info *pipe, struct file *out, +@@ -850,6 +850,7 @@ long do_splice_from(struct pipe_inode_info *pipe, struct file *out, return splice_write(pipe, out, ppos, len, flags); } @@ -921,7 +977,7 @@ index 46c87af..0efa652 100644 /* * Attempt to initiate a splice from a file to a pipe. -@@ -901,6 +902,7 @@ long do_splice_to(struct file *in, loff_t *ppos, +@@ -879,6 +880,7 @@ long do_splice_to(struct file *in, loff_t *ppos, return splice_read(in, ppos, pipe, len, flags); } @@ -930,10 +986,10 @@ index 46c87af..0efa652 100644 /** * splice_direct_to_actor - splices data directly between two non-pipes diff --git a/fs/sync.c b/fs/sync.c -index abf6a5d..c86fe9c 100644 +index 28607828e96f..ffd7ea43831e 100644 --- a/fs/sync.c +++ b/fs/sync.c -@@ -38,6 +38,7 @@ int __sync_filesystem(struct super_block *sb, int wait) +@@ -39,6 +39,7 @@ int __sync_filesystem(struct super_block *sb, int wait) sb->s_op->sync_fs(sb, wait); return __sync_blockdev(sb->s_bdev, wait); } @@ -942,10 +998,10 @@ index abf6a5d..c86fe9c 100644 /* * Write out and wait upon all dirty data associated with this diff --git a/fs/xattr.c b/fs/xattr.c -index 7e3317c..88910fe 100644 +index 0d6a6a4af861..7ce4701b7289 100644 --- a/fs/xattr.c +++ b/fs/xattr.c -@@ -296,6 +296,7 @@ vfs_getxattr_alloc(struct dentry *dentry, const char *name, char **xattr_value, +@@ -295,6 +295,7 @@ vfs_getxattr_alloc(struct dentry *dentry, const char *name, char **xattr_value, *xattr_value = value; return error; } @@ -953,59 +1009,41 @@ index 7e3317c..88910fe 100644 ssize_t __vfs_getxattr(struct dentry *dentry, struct inode *inode, const char *name, +diff --git a/kernel/locking/lockdep.c b/kernel/locking/lockdep.c +index 516ab7da26fb..c4f72e461d28 100644 +--- a/kernel/locking/lockdep.c ++++ b/kernel/locking/lockdep.c +@@ -154,6 +154,7 @@ inline struct lock_class *lockdep_hlock_class(struct held_lock *hlock) + } + return lock_classes + hlock->class_idx - 1; + } ++EXPORT_SYMBOL_GPL(lockdep_hlock_class); + #define hlock_class(hlock) lockdep_hlock_class(hlock) + + #ifdef CONFIG_LOCK_STAT diff --git a/kernel/task_work.c b/kernel/task_work.c -index d513051..e056d54 100644 +index 0fef395662a6..83fb1ecfc33d 100644 --- a/kernel/task_work.c +++ b/kernel/task_work.c -@@ -119,3 +119,4 @@ void task_work_run(void) +@@ -116,3 +116,4 @@ void task_work_run(void) } while (work); } } +EXPORT_SYMBOL_GPL(task_work_run); -diff --git a/security/commoncap.c b/security/commoncap.c -index 78b3783..c8b3e88 100644 ---- a/security/commoncap.c -+++ b/security/commoncap.c -@@ -1062,12 +1062,14 @@ int cap_mmap_addr(unsigned long addr) - } - return ret; - } -+EXPORT_SYMBOL_GPL(cap_mmap_addr); - - int cap_mmap_file(struct file *file, unsigned long reqprot, - unsigned long prot, unsigned long flags) - { - return 0; - } -+EXPORT_SYMBOL_GPL(cap_mmap_file); - - #ifdef CONFIG_SECURITY - diff --git a/security/device_cgroup.c b/security/device_cgroup.c -index 03c1652..f88c84b 100644 +index cd97929fac66..20c59b2e3b26 100644 --- a/security/device_cgroup.c +++ b/security/device_cgroup.c -@@ -7,6 +7,7 @@ - #include - #include - #include -+#include - #include - #include - #include -@@ -849,6 +850,7 @@ int __devcgroup_inode_permission(struct inode *inode, int mask) - return __devcgroup_check_permission(type, imajor(inode), iminor(inode), - access); - } -+EXPORT_SYMBOL_GPL(__devcgroup_inode_permission); +@@ -824,3 +824,4 @@ int __devcgroup_check_permission(short type, u32 major, u32 minor, - int devcgroup_inode_mknod(int mode, dev_t dev) - { + return 0; + } ++EXPORT_SYMBOL_GPL(__devcgroup_check_permission); diff --git a/security/security.c b/security/security.c -index d0e07f2..5e323b0 100644 +index 55bc49027ba9..b46c4eef1960 100644 --- a/security/security.c +++ b/security/security.c -@@ -481,6 +481,7 @@ int security_path_rmdir(const struct path *dir, struct dentry *dentry) +@@ -566,6 +566,7 @@ int security_path_rmdir(const struct path *dir, struct dentry *dentry) return 0; return call_int_hook(path_rmdir, 0, dir, dentry); } @@ -1013,7 +1051,7 @@ index d0e07f2..5e323b0 100644 int security_path_unlink(const struct path *dir, struct dentry *dentry) { -@@ -497,6 +498,7 @@ int security_path_symlink(const struct path *dir, struct dentry *dentry, +@@ -582,6 +583,7 @@ int security_path_symlink(const struct path *dir, struct dentry *dentry, return 0; return call_int_hook(path_symlink, 0, dir, dentry, old_name); } @@ -1021,7 +1059,7 @@ index d0e07f2..5e323b0 100644 int security_path_link(struct dentry *old_dentry, const struct path *new_dir, struct dentry *new_dentry) -@@ -505,6 +507,7 @@ int security_path_link(struct dentry *old_dentry, const struct path *new_dir, +@@ -590,6 +592,7 @@ int security_path_link(struct dentry *old_dentry, const struct path *new_dir, return 0; return call_int_hook(path_link, 0, old_dentry, new_dir, new_dentry); } @@ -1029,7 +1067,7 @@ index d0e07f2..5e323b0 100644 int security_path_rename(const struct path *old_dir, struct dentry *old_dentry, const struct path *new_dir, struct dentry *new_dentry, -@@ -532,6 +535,7 @@ int security_path_truncate(const struct path *path) +@@ -617,6 +620,7 @@ int security_path_truncate(const struct path *path) return 0; return call_int_hook(path_truncate, 0, path); } @@ -1037,7 +1075,7 @@ index d0e07f2..5e323b0 100644 int security_path_chmod(const struct path *path, umode_t mode) { -@@ -539,6 +543,7 @@ int security_path_chmod(const struct path *path, umode_t mode) +@@ -624,6 +628,7 @@ int security_path_chmod(const struct path *path, umode_t mode) return 0; return call_int_hook(path_chmod, 0, path, mode); } @@ -1045,7 +1083,7 @@ index d0e07f2..5e323b0 100644 int security_path_chown(const struct path *path, kuid_t uid, kgid_t gid) { -@@ -546,6 +551,7 @@ int security_path_chown(const struct path *path, kuid_t uid, kgid_t gid) +@@ -631,6 +636,7 @@ int security_path_chown(const struct path *path, kuid_t uid, kgid_t gid) return 0; return call_int_hook(path_chown, 0, path, uid, gid); } @@ -1053,15 +1091,7 @@ index d0e07f2..5e323b0 100644 int security_path_chroot(const struct path *path) { -@@ -631,6 +637,7 @@ int security_inode_readlink(struct dentry *dentry) - return 0; - return call_int_hook(inode_readlink, 0, dentry); - } -+EXPORT_SYMBOL_GPL(security_inode_readlink); - - int security_inode_follow_link(struct dentry *dentry, struct inode *inode, - bool rcu) -@@ -646,6 +653,7 @@ int security_inode_permission(struct inode *inode, int mask) +@@ -731,6 +737,7 @@ int security_inode_permission(struct inode *inode, int mask) return 0; return call_int_hook(inode_permission, 0, inode, mask); } @@ -1069,7 +1099,7 @@ index d0e07f2..5e323b0 100644 int security_inode_setattr(struct dentry *dentry, struct iattr *attr) { -@@ -817,6 +825,7 @@ int security_file_permission(struct file *file, int mask) +@@ -902,6 +909,7 @@ int security_file_permission(struct file *file, int mask) return fsnotify_perm(file, mask); } @@ -1077,7 +1107,7 @@ index d0e07f2..5e323b0 100644 int security_file_alloc(struct file *file) { -@@ -876,6 +885,7 @@ int security_mmap_file(struct file *file, unsigned long prot, +@@ -961,6 +969,7 @@ int security_mmap_file(struct file *file, unsigned long prot, return ret; return ima_file_mmap(file, prot); } @@ -1087,8 +1117,8 @@ index d0e07f2..5e323b0 100644 { diff -urN /usr/share/empty/Documentation/ABI/testing/debugfs-aufs linux/Documentation/ABI/testing/debugfs-aufs --- /usr/share/empty/Documentation/ABI/testing/debugfs-aufs 1970-01-01 01:00:00.000000000 +0100 -+++ linux/Documentation/ABI/testing/debugfs-aufs 2017-05-06 22:16:52.774886674 +0200 -@@ -0,0 +1,50 @@ ++++ linux/Documentation/ABI/testing/debugfs-aufs 2018-08-12 23:43:05.450124426 +0200 +@@ -0,0 +1,55 @@ +What: /debug/aufs/si_/ +Date: March 2009 +Contact: J. R. Okajima @@ -1119,15 +1149,20 @@ diff -urN /usr/share/empty/Documentation/ABI/testing/debugfs-aufs linux/Document + When the aufs mount option 'noxino' is specified, it + will be empty. About XINO files, see the aufs manual. + -+What: /debug/aufs/si_/xino0, xino1 ... xinoN ++What: /debug/aufs/si_/xi0, xi1 ... xiN and xiN-N +Date: March 2009 +Contact: J. R. Okajima +Description: + It shows the consumed blocks by xino (External Inode Number + Translation Table), its link count, block size and file + size. -+ When the aufs mount option 'noxino' is specified, it -+ will be empty. About XINO files, see the aufs manual. ++ Due to the file size limit, there may exist multiple ++ xino files per branch. In this case, "-N" is added to ++ the filename and it corresponds to the index of the ++ internal xino array. "-0" is omitted. ++ When the aufs mount option 'noxino' is specified, Those ++ entries won't exist. About XINO files, see the aufs ++ manual. + +What: /debug/aufs/si_/xigen +Date: March 2009 @@ -1141,7 +1176,7 @@ diff -urN /usr/share/empty/Documentation/ABI/testing/debugfs-aufs linux/Document + will be empty. About XINO files, see the aufs manual. diff -urN /usr/share/empty/Documentation/ABI/testing/sysfs-aufs linux/Documentation/ABI/testing/sysfs-aufs --- /usr/share/empty/Documentation/ABI/testing/sysfs-aufs 1970-01-01 01:00:00.000000000 +0100 -+++ linux/Documentation/ABI/testing/sysfs-aufs 2017-05-06 22:16:52.774886674 +0200 ++++ linux/Documentation/ABI/testing/sysfs-aufs 2017-07-29 12:14:25.893041746 +0200 @@ -0,0 +1,31 @@ +What: /sys/fs/aufs/si_/ +Date: March 2009 @@ -1176,10 +1211,10 @@ diff -urN /usr/share/empty/Documentation/ABI/testing/sysfs-aufs linux/Documentat + will be empty. About XINO files, see the aufs manual. diff -urN /usr/share/empty/Documentation/filesystems/aufs/design/01intro.txt linux/Documentation/filesystems/aufs/design/01intro.txt --- /usr/share/empty/Documentation/filesystems/aufs/design/01intro.txt 1970-01-01 01:00:00.000000000 +0100 -+++ linux/Documentation/filesystems/aufs/design/01intro.txt 2017-05-06 22:16:52.774886674 +0200 -@@ -0,0 +1,170 @@ ++++ linux/Documentation/filesystems/aufs/design/01intro.txt 2019-03-05 12:13:00.132557473 +0100 +@@ -0,0 +1,171 @@ + -+# Copyright (C) 2005-2017 Junjiro R. Okajima ++# Copyright (C) 2005-2019 Junjiro R. Okajima +# +# This program is free software; you can redistribute it and/or modify +# it under the terms of the GNU General Public License as published by @@ -1197,12 +1232,13 @@ diff -urN /usr/share/empty/Documentation/filesystems/aufs/design/01intro.txt lin +Introduction +---------------------------------------- + -+aufs [ei ju: ef es] | [a u f s] ++aufs [ei ju: ef es] | /ey-yoo-ef-es/ | [a u f s] +1. abbrev. for "advanced multi-layered unification filesystem". +2. abbrev. for "another unionfs". +3. abbrev. for "auf das" in German which means "on the" in English. + Ex. "Butter aufs Brot"(G) means "butter onto bread"(E). + But "Filesystem aufs Filesystem" is hard to understand. ++4. abbrev. for "African Urban Fashion Show". + +AUFS is a filesystem with features: +- multi layered stackable unification filesystem, the member directory @@ -1350,10 +1386,10 @@ diff -urN /usr/share/empty/Documentation/filesystems/aufs/design/01intro.txt lin +about it. But currently I have implemented it in kernel space. diff -urN /usr/share/empty/Documentation/filesystems/aufs/design/02struct.txt linux/Documentation/filesystems/aufs/design/02struct.txt --- /usr/share/empty/Documentation/filesystems/aufs/design/02struct.txt 1970-01-01 01:00:00.000000000 +0100 -+++ linux/Documentation/filesystems/aufs/design/02struct.txt 2017-05-06 22:16:52.774886674 +0200 ++++ linux/Documentation/filesystems/aufs/design/02struct.txt 2019-03-05 12:13:00.132557473 +0100 @@ -0,0 +1,258 @@ + -+# Copyright (C) 2005-2017 Junjiro R. Okajima ++# Copyright (C) 2005-2019 Junjiro R. Okajima +# +# This program is free software; you can redistribute it and/or modify +# it under the terms of the GNU General Public License as published by @@ -1612,10 +1648,10 @@ diff -urN /usr/share/empty/Documentation/filesystems/aufs/design/02struct.txt li +For this purpose, use "aumvdown" command in aufs-util.git. diff -urN /usr/share/empty/Documentation/filesystems/aufs/design/03atomic_open.txt linux/Documentation/filesystems/aufs/design/03atomic_open.txt --- /usr/share/empty/Documentation/filesystems/aufs/design/03atomic_open.txt 1970-01-01 01:00:00.000000000 +0100 -+++ linux/Documentation/filesystems/aufs/design/03atomic_open.txt 2017-05-06 22:16:52.774886674 +0200 ++++ linux/Documentation/filesystems/aufs/design/03atomic_open.txt 2019-03-05 12:13:00.132557473 +0100 @@ -0,0 +1,85 @@ + -+# Copyright (C) 2015-2017 Junjiro R. Okajima ++# Copyright (C) 2015-2019 Junjiro R. Okajima +# +# This program is free software; you can redistribute it and/or modify +# it under the terms of the GNU General Public License as published by @@ -1701,10 +1737,10 @@ diff -urN /usr/share/empty/Documentation/filesystems/aufs/design/03atomic_open.t + be implemented in aufs, but not all I am afraid. diff -urN /usr/share/empty/Documentation/filesystems/aufs/design/03lookup.txt linux/Documentation/filesystems/aufs/design/03lookup.txt --- /usr/share/empty/Documentation/filesystems/aufs/design/03lookup.txt 1970-01-01 01:00:00.000000000 +0100 -+++ linux/Documentation/filesystems/aufs/design/03lookup.txt 2017-05-06 22:16:52.774886674 +0200 ++++ linux/Documentation/filesystems/aufs/design/03lookup.txt 2019-03-05 12:13:00.132557473 +0100 @@ -0,0 +1,113 @@ + -+# Copyright (C) 2005-2017 Junjiro R. Okajima ++# Copyright (C) 2005-2019 Junjiro R. Okajima +# +# This program is free software; you can redistribute it and/or modify +# it under the terms of the GNU General Public License as published by @@ -1818,10 +1854,10 @@ diff -urN /usr/share/empty/Documentation/filesystems/aufs/design/03lookup.txt li + by over-mounting something (or another method). diff -urN /usr/share/empty/Documentation/filesystems/aufs/design/04branch.txt linux/Documentation/filesystems/aufs/design/04branch.txt --- /usr/share/empty/Documentation/filesystems/aufs/design/04branch.txt 1970-01-01 01:00:00.000000000 +0100 -+++ linux/Documentation/filesystems/aufs/design/04branch.txt 2017-05-06 22:16:52.774886674 +0200 ++++ linux/Documentation/filesystems/aufs/design/04branch.txt 2019-03-05 12:13:00.132557473 +0100 @@ -0,0 +1,74 @@ + -+# Copyright (C) 2005-2017 Junjiro R. Okajima ++# Copyright (C) 2005-2019 Junjiro R. Okajima +# +# This program is free software; you can redistribute it and/or modify +# it under the terms of the GNU General Public License as published by @@ -1896,10 +1932,10 @@ diff -urN /usr/share/empty/Documentation/filesystems/aufs/design/04branch.txt li + same named entry on the upper branch. diff -urN /usr/share/empty/Documentation/filesystems/aufs/design/05wbr_policy.txt linux/Documentation/filesystems/aufs/design/05wbr_policy.txt --- /usr/share/empty/Documentation/filesystems/aufs/design/05wbr_policy.txt 1970-01-01 01:00:00.000000000 +0100 -+++ linux/Documentation/filesystems/aufs/design/05wbr_policy.txt 2017-05-06 22:16:52.774886674 +0200 ++++ linux/Documentation/filesystems/aufs/design/05wbr_policy.txt 2019-03-05 12:13:00.132557473 +0100 @@ -0,0 +1,64 @@ + -+# Copyright (C) 2005-2017 Junjiro R. Okajima ++# Copyright (C) 2005-2019 Junjiro R. Okajima +# +# This program is free software; you can redistribute it and/or modify +# it under the terms of the GNU General Public License as published by @@ -1962,12 +1998,153 @@ diff -urN /usr/share/empty/Documentation/filesystems/aufs/design/05wbr_policy.tx + where the source and the target exists and selects the higher + one. If the selected branch is readonly, then aufs follows the + copyup policy. +diff -urN /usr/share/empty/Documentation/filesystems/aufs/design/06dirren.dot linux/Documentation/filesystems/aufs/design/06dirren.dot +--- /usr/share/empty/Documentation/filesystems/aufs/design/06dirren.dot 1970-01-01 01:00:00.000000000 +0100 ++++ linux/Documentation/filesystems/aufs/design/06dirren.dot 2018-04-15 08:49:13.394483860 +0200 +@@ -0,0 +1,31 @@ ++ ++// to view this graph, run dot(1) command in GRAPHVIZ. ++ ++digraph G { ++node [shape=box]; ++whinfo [label="detailed info file\n(lower_brid_root-hinum, h_inum, namelen, old name)"]; ++ ++node [shape=oval]; ++ ++aufs_rename -> whinfo [label="store/remove"]; ++ ++node [shape=oval]; ++inode_list [label="h_inum list in branch\ncache"]; ++ ++node [shape=box]; ++whinode [label="h_inum list file"]; ++ ++node [shape=oval]; ++brmgmt [label="br_add/del/mod/umount"]; ++ ++brmgmt -> inode_list [label="create/remove"]; ++brmgmt -> whinode [label="load/store"]; ++ ++inode_list -> whinode [style=dashed,dir=both]; ++ ++aufs_rename -> inode_list [label="add/del"]; ++ ++aufs_lookup -> inode_list [label="search"]; ++ ++aufs_lookup -> whinfo [label="load/remove"]; ++} +diff -urN /usr/share/empty/Documentation/filesystems/aufs/design/06dirren.txt linux/Documentation/filesystems/aufs/design/06dirren.txt +--- /usr/share/empty/Documentation/filesystems/aufs/design/06dirren.txt 1970-01-01 01:00:00.000000000 +0100 ++++ linux/Documentation/filesystems/aufs/design/06dirren.txt 2019-03-05 12:13:00.132557473 +0100 +@@ -0,0 +1,102 @@ ++ ++# Copyright (C) 2017-2019 Junjiro R. Okajima ++# ++# This program is free software; you can redistribute it and/or modify ++# it under the terms of the GNU General Public License as published by ++# the Free Software Foundation; either version 2 of the License, or ++# (at your option) any later version. ++# ++# This program is distributed in the hope that it will be useful, ++# but WITHOUT ANY WARRANTY; without even the implied warranty of ++# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++# GNU General Public License for more details. ++# ++# You should have received a copy of the GNU General Public License ++# along with this program. If not, see . ++ ++Special handling for renaming a directory (DIRREN) ++---------------------------------------------------------------------- ++First, let's assume we have a simple usecase. ++ ++- /u = /rw + /ro ++- /rw/dirA exists ++- /ro/dirA and /ro/dirA/file exist too ++- there is no dirB on both branches ++- a user issues rename("dirA", "dirB") ++ ++Now, what should aufs behave against this rename(2)? ++There are a few possible cases. ++ ++A. returns EROFS. ++ since dirA exists on a readonly branch which cannot be renamed. ++B. returns EXDEV. ++ it is possible to copy-up dirA (only the dir itself), but the child ++ entries ("file" in this case) should not be. it must be a bad ++ approach to copy-up recursively. ++C. returns a success. ++ even the branch /ro is readonly, aufs tries renaming it. Obviously it ++ is a violation of aufs' policy. ++D. construct an extra information which indicates that /ro/dirA should ++ be handled as the name of dirB. ++ overlayfs has a similar feature called REDIRECT. ++ ++Until now, aufs implements the case B only which returns EXDEV, and ++expects the userspace application behaves like mv(1) which tries ++issueing rename(2) recursively. ++ ++A new aufs feature called DIRREN is introduced which implements the case ++D. There are several "extra information" added. ++ ++1. detailed info per renamed directory ++ path: /rw/dirB/$AUFS_WH_DR_INFO_PFX. ++2. the inode-number list of directories on a branch ++ path: /rw/dirB/$AUFS_WH_DR_BRHINO ++ ++The filename of "detailed info per directory" represents the lower ++branch, and its format is ++- a type of the branch id ++ one of these. ++ + uuid (not implemented yet) ++ + fsid ++ + dev ++- the inode-number of the branch root dir ++ ++And it contains these info in a single regular file. ++- magic number ++- branch's inode-number of the logically renamed dir ++- the name of the before-renamed dir ++ ++The "detailed info per directory" file is created in aufs rename(2), and ++loaded in any lookup. ++The info is considered in lookup for the matching case only. Here ++"matching" means that the root of branch (in the info filename) is same ++to the current looking-up branch. After looking-up the before-renamed ++name, the inode-number is compared. And the matched dentry is used. ++ ++The "inode-number list of directories" is a regular file which contains ++simply the inode-numbers on the branch. The file is created or updated ++in removing the branch, and loaded in adding the branch. Its lifetime is ++equal to the branch. ++The list is refered in lookup, and when the current target inode is ++found in the list, the aufs tries loading the "detailed info per ++directory" and get the changed and valid name of the dir. ++ ++Theoretically these "extra informaiton" may be able to be put into XATTR ++in the dir inode. But aufs doesn't choose this way because ++1. XATTR may not be supported by the branch (or its configuration) ++2. XATTR may have its size limit. ++3. XATTR may be less easy to convert than a regular file, when the ++ format of the info is changed in the future. ++At the same time, I agree that the regular file approach is much slower ++than XATTR approach. So, in the future, aufs may take the XATTR or other ++better approach. ++ ++This DIRREN feature is enabled by aufs configuration, and is activated ++by a new mount option. ++ ++For the more complicated case, there is a work with UDBA option, which ++is to dected the direct access to the branches (by-passing aufs) and to ++maintain the cashes in aufs. Since a single cached aufs dentry may ++contains two names, before- and after-rename, the name comparision in ++UDBA handler may not work correctly. In this case, the behaviour will be ++equivalen to udba=reval case. diff -urN /usr/share/empty/Documentation/filesystems/aufs/design/06fhsm.txt linux/Documentation/filesystems/aufs/design/06fhsm.txt --- /usr/share/empty/Documentation/filesystems/aufs/design/06fhsm.txt 1970-01-01 01:00:00.000000000 +0100 -+++ linux/Documentation/filesystems/aufs/design/06fhsm.txt 2017-05-06 22:16:52.774886674 +0200 ++++ linux/Documentation/filesystems/aufs/design/06fhsm.txt 2019-03-05 12:13:00.132557473 +0100 @@ -0,0 +1,120 @@ + -+# Copyright (C) 2011-2017 Junjiro R. Okajima ++# Copyright (C) 2011-2019 Junjiro R. Okajima +# +# This program is free software; you can redistribute it and/or modify +# it under the terms of the GNU General Public License as published by @@ -2088,10 +2265,10 @@ diff -urN /usr/share/empty/Documentation/filesystems/aufs/design/06fhsm.txt linu +should restore the original file state after an error happens. diff -urN /usr/share/empty/Documentation/filesystems/aufs/design/06mmap.txt linux/Documentation/filesystems/aufs/design/06mmap.txt --- /usr/share/empty/Documentation/filesystems/aufs/design/06mmap.txt 1970-01-01 01:00:00.000000000 +0100 -+++ linux/Documentation/filesystems/aufs/design/06mmap.txt 2017-05-06 22:16:52.774886674 +0200 ++++ linux/Documentation/filesystems/aufs/design/06mmap.txt 2019-03-05 12:13:00.132557473 +0100 @@ -0,0 +1,72 @@ + -+# Copyright (C) 2005-2017 Junjiro R. Okajima ++# Copyright (C) 2005-2019 Junjiro R. Okajima +# +# This program is free software; you can redistribute it and/or modify +# it under the terms of the GNU General Public License as published by @@ -2164,10 +2341,10 @@ diff -urN /usr/share/empty/Documentation/filesystems/aufs/design/06mmap.txt linu +I have to give up this "looks-smater" approach. diff -urN /usr/share/empty/Documentation/filesystems/aufs/design/06xattr.txt linux/Documentation/filesystems/aufs/design/06xattr.txt --- /usr/share/empty/Documentation/filesystems/aufs/design/06xattr.txt 1970-01-01 01:00:00.000000000 +0100 -+++ linux/Documentation/filesystems/aufs/design/06xattr.txt 2017-05-06 22:16:52.774886674 +0200 ++++ linux/Documentation/filesystems/aufs/design/06xattr.txt 2019-03-05 12:13:00.132557473 +0100 @@ -0,0 +1,96 @@ + -+# Copyright (C) 2014-2017 Junjiro R. Okajima ++# Copyright (C) 2014-2019 Junjiro R. Okajima +# +# This program is free software; you can redistribute it and/or modify +# it under the terms of the GNU General Public License as published by @@ -2264,10 +2441,10 @@ diff -urN /usr/share/empty/Documentation/filesystems/aufs/design/06xattr.txt lin +now, aufs implements the branch attributes to ignore the error. diff -urN /usr/share/empty/Documentation/filesystems/aufs/design/07export.txt linux/Documentation/filesystems/aufs/design/07export.txt --- /usr/share/empty/Documentation/filesystems/aufs/design/07export.txt 1970-01-01 01:00:00.000000000 +0100 -+++ linux/Documentation/filesystems/aufs/design/07export.txt 2017-05-06 22:16:52.774886674 +0200 ++++ linux/Documentation/filesystems/aufs/design/07export.txt 2019-03-05 12:13:00.132557473 +0100 @@ -0,0 +1,58 @@ + -+# Copyright (C) 2005-2017 Junjiro R. Okajima ++# Copyright (C) 2005-2019 Junjiro R. Okajima +# +# This program is free software; you can redistribute it and/or modify +# it under the terms of the GNU General Public License as published by @@ -2326,10 +2503,10 @@ diff -urN /usr/share/empty/Documentation/filesystems/aufs/design/07export.txt li + lookup_one_len(), vfs_getattr(), encode_fh() and others. diff -urN /usr/share/empty/Documentation/filesystems/aufs/design/08shwh.txt linux/Documentation/filesystems/aufs/design/08shwh.txt --- /usr/share/empty/Documentation/filesystems/aufs/design/08shwh.txt 1970-01-01 01:00:00.000000000 +0100 -+++ linux/Documentation/filesystems/aufs/design/08shwh.txt 2017-05-06 22:16:52.774886674 +0200 ++++ linux/Documentation/filesystems/aufs/design/08shwh.txt 2019-03-05 12:13:00.135890907 +0100 @@ -0,0 +1,52 @@ + -+# Copyright (C) 2005-2017 Junjiro R. Okajima ++# Copyright (C) 2005-2019 Junjiro R. Okajima +# +# This program is free software; you can redistribute it and/or modify +# it under the terms of the GNU General Public License as published by @@ -2382,10 +2559,10 @@ diff -urN /usr/share/empty/Documentation/filesystems/aufs/design/08shwh.txt linu +initramfs will use it to replace the old one at the next boot. diff -urN /usr/share/empty/Documentation/filesystems/aufs/design/10dynop.txt linux/Documentation/filesystems/aufs/design/10dynop.txt --- /usr/share/empty/Documentation/filesystems/aufs/design/10dynop.txt 1970-01-01 01:00:00.000000000 +0100 -+++ linux/Documentation/filesystems/aufs/design/10dynop.txt 2017-05-06 22:16:52.774886674 +0200 ++++ linux/Documentation/filesystems/aufs/design/10dynop.txt 2019-03-05 12:13:00.135890907 +0100 @@ -0,0 +1,47 @@ + -+# Copyright (C) 2010-2017 Junjiro R. Okajima ++# Copyright (C) 2010-2019 Junjiro R. Okajima +# +# This program is free software; you can redistribute it and/or modify +# it under the terms of the GNU General Public License as published by @@ -2433,8 +2610,8 @@ diff -urN /usr/share/empty/Documentation/filesystems/aufs/design/10dynop.txt lin +regular files only. diff -urN /usr/share/empty/Documentation/filesystems/aufs/README linux/Documentation/filesystems/aufs/README --- /usr/share/empty/Documentation/filesystems/aufs/README 1970-01-01 01:00:00.000000000 +0100 -+++ linux/Documentation/filesystems/aufs/README 2017-05-06 22:16:52.774886674 +0200 -@@ -0,0 +1,393 @@ ++++ linux/Documentation/filesystems/aufs/README 2019-07-08 09:31:38.324384703 +0200 +@@ -0,0 +1,402 @@ + +Aufs4 -- advanced multi layered unification filesystem version 4.x +http://aufs.sf.net @@ -2689,6 +2866,13 @@ diff -urN /usr/share/empty/Documentation/filesystems/aufs/README linux/Documenta + Supports a nested loopback mount in a branch-fs. This patch is + unnecessary until aufs produces a message like "you may want to try + another patch for loopback file". ++- proc_mounts.patch ++ When there are many mountpoints and many mount(2)/umount(2) are ++ running, then /proc/mounts may not show the all mountpoints. This ++ patch makes /proc/mounts always show the full mountpoints list. ++ If you don't want to apply this patch and meet such problem, then you ++ need to increase the value of 'ProcMounts_Times' make-variable in ++ aufs-util.git as a second best solution. +- vfs-ino.patch + Modifies a system global kernel internal function get_next_ino() in + order to stop assigning 0 for an inode-number. Not directly related to @@ -2808,9 +2992,11 @@ diff -urN /usr/share/empty/Documentation/filesystems/aufs/README linux/Documenta +James B made a donation (2014/7 and 2015/7). +Stefano Di Biase made a donation (2014/8). +Daniel Epellei made a donation (2015/1). -+OmegaPhil made a donation (2016/1). ++OmegaPhil made a donation (2016/1, 2018/4). +Tomasz Szewczyk made a donation (2016/4). +James Burry made a donation (2016/12). ++Carsten Rose made a donation (2018/9). ++Porteus Kiosk made a donation (2018/10). + +Thank you very much. +Donations are always, including future donations, very important and @@ -2830,10 +3016,11 @@ diff -urN /usr/share/empty/Documentation/filesystems/aufs/README linux/Documenta +# End: ; diff -urN /usr/share/empty/fs/aufs/aufs.h linux/fs/aufs/aufs.h --- /usr/share/empty/fs/aufs/aufs.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/aufs.h 2017-05-06 22:16:52.774886674 +0200 -@@ -0,0 +1,59 @@ ++++ linux/fs/aufs/aufs.h 2019-03-05 12:13:00.135890907 +0100 +@@ -0,0 +1,62 @@ ++/* SPDX-License-Identifier: GPL-2.0 */ +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -2874,15 +3061,17 @@ diff -urN /usr/share/empty/fs/aufs/aufs.h linux/fs/aufs/aufs.h +#include "dbgaufs.h" +#include "dentry.h" +#include "dir.h" ++#include "dirren.h" +#include "dynop.h" +#include "file.h" +#include "fstype.h" ++#include "hbl.h" +#include "inode.h" ++#include "lcnt.h" +#include "loop.h" +#include "module.h" +#include "opts.h" +#include "rwsem.h" -+#include "spl.h" +#include "super.h" +#include "sysaufs.h" +#include "vfsub.h" @@ -2893,10 +3082,11 @@ diff -urN /usr/share/empty/fs/aufs/aufs.h linux/fs/aufs/aufs.h +#endif /* __AUFS_H__ */ diff -urN /usr/share/empty/fs/aufs/branch.c linux/fs/aufs/branch.c --- /usr/share/empty/fs/aufs/branch.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/branch.c 2017-05-06 22:16:52.774886674 +0200 ++++ linux/fs/aufs/branch.c 2019-03-05 12:13:00.135890907 +0100 @@ -0,0 +1,1423 @@ ++// SPDX-License-Identifier: GPL-2.0 +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -2930,15 +3120,14 @@ diff -urN /usr/share/empty/fs/aufs/branch.c linux/fs/aufs/branch.c + struct au_dykey **key; + + au_hnotify_fin_br(br); ++ /* always, regardless the mount option */ ++ au_dr_hino_free(&br->br_dirren); ++ au_xino_put(br); + -+ if (br->br_xino.xi_file) -+ fput(br->br_xino.xi_file); -+ for (i = br->br_xino.xi_nondir.total - 1; i >= 0; i--) -+ AuDebugOn(br->br_xino.xi_nondir.array[i]); -+ au_delayed_kfree(br->br_xino.xi_nondir.array); -+ -+ AuDebugOn(au_br_count(br)); -+ au_br_count_fin(br); ++ AuLCntZero(au_lcnt_read(&br->br_nfiles, /*do_rev*/0)); ++ au_lcnt_fin(&br->br_nfiles, /*do_sync*/0); ++ AuLCntZero(au_lcnt_read(&br->br_count, /*do_rev*/0)); ++ au_lcnt_fin(&br->br_count, /*do_sync*/0); + + wbr = br->br_wbr; + if (wbr) { @@ -2950,7 +3139,7 @@ diff -urN /usr/share/empty/fs/aufs/branch.c linux/fs/aufs/branch.c + + if (br->br_fhsm) { + au_br_fhsm_fin(br->br_fhsm); -+ au_delayed_kfree(br->br_fhsm); ++ au_kfree_try_rcu(br->br_fhsm); + } + + key = br->br_dykey; @@ -2961,12 +3150,16 @@ diff -urN /usr/share/empty/fs/aufs/branch.c linux/fs/aufs/branch.c + break; + + /* recursive lock, s_umount of branch's */ ++ /* synchronize_rcu(); */ /* why? */ + lockdep_off(); + path_put(&br->br_path); + lockdep_on(); -+ if (wbr) -+ au_delayed_kfree(wbr); -+ au_delayed_kfree(br); ++ au_kfree_rcu(wbr); ++ au_lcnt_wait_for_fin(&br->br_nfiles); ++ au_lcnt_wait_for_fin(&br->br_count); ++ /* I don't know why, but percpu_refcount requires this */ ++ /* synchronize_rcu(); */ ++ au_kfree_rcu(br); +} + +/* @@ -3033,16 +3226,12 @@ diff -urN /usr/share/empty/fs/aufs/branch.c linux/fs/aufs/branch.c + add_branch = kzalloc(sizeof(*add_branch), GFP_NOFS); + if (unlikely(!add_branch)) + goto out; -+ add_branch->br_xino.xi_nondir.total = 8; /* initial size */ -+ add_branch->br_xino.xi_nondir.array -+ = kzalloc(sizeof(ino_t) * add_branch->br_xino.xi_nondir.total, -+ GFP_NOFS); -+ if (unlikely(!add_branch->br_xino.xi_nondir.array)) ++ add_branch->br_xino = au_xino_alloc(/*nfile*/1); ++ if (unlikely(!add_branch->br_xino)) + goto out_br; -+ + err = au_hnotify_init_br(add_branch, perm); + if (unlikely(err)) -+ goto out_xinondir; ++ goto out_xino; + + if (au_br_writable(perm)) { + /* may be freed separately at changing the branch permission */ @@ -3064,20 +3253,20 @@ diff -urN /usr/share/empty/fs/aufs/branch.c linux/fs/aufs/branch.c + err = au_di_realloc(au_di(root), new_nbranch, /*may_shrink*/0); + if (!err) { + inode = d_inode(root); -+ err = au_hinode_realloc(au_ii(inode), new_nbranch, /*may_shrink*/0); ++ err = au_hinode_realloc(au_ii(inode), new_nbranch, ++ /*may_shrink*/0); + } + if (!err) + return add_branch; /* success */ + +out_wbr: -+ if (add_branch->br_wbr) -+ au_delayed_kfree(add_branch->br_wbr); ++ au_kfree_rcu(add_branch->br_wbr); +out_hnotify: + au_hnotify_fin_br(add_branch); -+out_xinondir: -+ au_delayed_kfree(add_branch->br_xino.xi_nondir.array); ++out_xino: ++ au_xino_put(add_branch); +out_br: -+ au_delayed_kfree(add_branch); ++ au_kfree_rcu(add_branch); +out: + return ERR_PTR(err); +} @@ -3243,7 +3432,7 @@ diff -urN /usr/share/empty/fs/aufs/branch.c linux/fs/aufs/branch.c + br->br_perm = old_perm; + + if (!err && wbr && !au_br_writable(new_perm)) { -+ au_delayed_kfree(wbr); ++ au_kfree_rcu(wbr); + br->br_wbr = NULL; + } + @@ -3286,19 +3475,24 @@ diff -urN /usr/share/empty/fs/aufs/branch.c linux/fs/aufs/branch.c + struct au_opt_add *add) +{ + int err; ++ struct au_branch *brbase; ++ struct file *xf; + struct inode *h_inode; + + err = 0; -+ spin_lock_init(&br->br_xino.xi_nondir.spin); -+ init_waitqueue_head(&br->br_xino.xi_nondir.wqh); + br->br_perm = add->perm; + br->br_path = add->path; /* set first, path_get() later */ + spin_lock_init(&br->br_dykey_lock); -+ au_br_count_init(br); -+ atomic_set(&br->br_xino_running, 0); ++ au_lcnt_init(&br->br_nfiles, /*release*/NULL); ++ au_lcnt_init(&br->br_count, /*release*/NULL); + br->br_id = au_new_br_id(sb); + AuDebugOn(br->br_id < 0); + ++ /* always, regardless the given option */ ++ err = au_dr_br_init(sb, br, &add->path); ++ if (unlikely(err)) ++ goto out_err; ++ + if (au_br_writable(add->perm)) { + err = au_wbr_init(br, sb, add->perm); + if (unlikely(err)) @@ -3306,11 +3500,13 @@ diff -urN /usr/share/empty/fs/aufs/branch.c linux/fs/aufs/branch.c + } + + if (au_opt_test(au_mntflags(sb), XINO)) { ++ brbase = au_sbr(sb, 0); ++ xf = au_xino_file(brbase->br_xino, /*idx*/-1); ++ AuDebugOn(!xf); + h_inode = d_inode(add->path.dentry); -+ err = au_xino_br(sb, br, h_inode->i_ino, -+ au_sbr(sb, 0)->br_xino.xi_file, /*do_test*/1); ++ err = au_xino_init_br(sb, br, h_inode->i_ino, &xf->f_path); + if (unlikely(err)) { -+ AuDebugOn(br->br_xino.xi_file); ++ AuDebugOn(au_xino_file(br->br_xino, /*idx*/-1)); + goto out_err; + } + } @@ -3426,13 +3622,10 @@ diff -urN /usr/share/empty/fs/aufs/branch.c linux/fs/aufs/branch.c + } + + add_bindex = add->bindex; -+ if (!remount) -+ au_br_do_add(sb, add_branch, add_bindex); -+ else { -+ sysaufs_brs_del(sb, add_bindex); -+ au_br_do_add(sb, add_branch, add_bindex); -+ sysaufs_brs_add(sb, add_bindex); -+ } ++ sysaufs_brs_del(sb, add_bindex); /* remove successors */ ++ au_br_do_add(sb, add_branch, add_bindex); ++ sysaufs_brs_add(sb, add_bindex); /* append successors */ ++ dbgaufs_brs_add(sb, add_bindex, /*topdown*/0); /* rename successors */ + + h_dentry = add->path.dentry; + if (!add_bindex) { @@ -3441,18 +3634,6 @@ diff -urN /usr/share/empty/fs/aufs/branch.c linux/fs/aufs/branch.c + } else + au_add_nlink(root_inode, d_inode(h_dentry)); + -+ /* -+ * this test/set prevents aufs from handling unnecesary notify events -+ * of xino files, in case of re-adding a writable branch which was -+ * once detached from aufs. -+ */ -+ if (au_xino_brid(sb) < 0 -+ && au_br_writable(add_branch->br_perm) -+ && !au_test_fs_bad_xino(h_dentry->d_sb) -+ && add_branch->br_xino.xi_file -+ && add_branch->br_xino.xi_file->f_path.dentry->d_parent == h_dentry) -+ au_xino_brid_set(sb, add_branch->br_id); -+ +out: + return err; +} @@ -3465,14 +3646,15 @@ diff -urN /usr/share/empty/fs/aufs/branch.c linux/fs/aufs/branch.c +{ + unsigned long long n; + struct file **p, *f; -+ struct au_sphlhead *files; ++ struct hlist_bl_head *files; ++ struct hlist_bl_node *pos; + struct au_finfo *finfo; + + n = 0; + p = a; + files = &au_sbi(sb)->si_files; -+ spin_lock(&files->spin); -+ hlist_for_each_entry(finfo, &files->head, fi_hlist) { ++ hlist_bl_lock(files); ++ hlist_bl_for_each_entry(finfo, pos, files, fi_hlist) { + f = finfo->fi_file; + if (file_count(f) + && !special_file(file_inode(f)->i_mode)) { @@ -3482,7 +3664,7 @@ diff -urN /usr/share/empty/fs/aufs/branch.c linux/fs/aufs/branch.c + AuDebugOn(n > max); + } + } -+ spin_unlock(&files->spin); ++ hlist_bl_unlock(files); + + return n; +} @@ -3490,7 +3672,10 @@ diff -urN /usr/share/empty/fs/aufs/branch.c linux/fs/aufs/branch.c +static struct file **au_farray_alloc(struct super_block *sb, + unsigned long long *max) +{ -+ *max = au_nfiles(sb); ++ struct au_sbinfo *sbi; ++ ++ sbi = au_sbi(sb); ++ *max = au_lcnt_read(&sbi->si_nfiles, /*do_rev*/1); + return au_array_alloc(max, au_farray_cb, sb, /*arg*/NULL); +} + @@ -3753,7 +3938,7 @@ diff -urN /usr/share/empty/fs/aufs/branch.c linux/fs/aufs/branch.c +} + +static void br_del_file(struct file **to_free, unsigned long long opened, -+ aufs_bindex_t br_id) ++ aufs_bindex_t br_id) +{ + unsigned long long ull; + aufs_bindex_t bindex, btop, bbot, bfound; @@ -3885,6 +4070,9 @@ diff -urN /usr/share/empty/fs/aufs/branch.c linux/fs/aufs/branch.c + au_br_do_del_hip(au_ii(inode), bindex, bbot); + au_sbilist_unlock(); + ++ /* ignore an error */ ++ au_dr_br_fin(sb, br); /* always, regardless the mount option */ ++ + dput(h_root); + iput(h_inode); + au_br_do_free(br); @@ -3930,11 +4118,16 @@ diff -urN /usr/share/empty/fs/aufs/branch.c linux/fs/aufs/branch.c + AuVerbose(verbose, "no more branches left\n"); + goto out; + } ++ + br = au_sbr(sb, bindex); + AuDebugOn(!path_equal(&br->br_path, &del->h_path)); ++ if (unlikely(au_lcnt_read(&br->br_count, /*do_rev*/1))) { ++ AuVerbose(verbose, "br %pd2 is busy now\n", del->h_path.dentry); ++ goto out; ++ } + + br_id = br->br_id; -+ opened = au_br_count(br); ++ opened = au_lcnt_read(&br->br_nfiles, /*do_rev*/1); + if (unlikely(opened)) { + to_free = au_array_alloc(&opened, empty_cb, sb, NULL); + err = PTR_ERR(to_free); @@ -3977,13 +4170,11 @@ diff -urN /usr/share/empty/fs/aufs/branch.c linux/fs/aufs/branch.c + di_write_lock_child(root); + } + -+ if (!remount) -+ au_br_do_del(sb, bindex, br); -+ else { -+ sysaufs_brs_del(sb, bindex); -+ au_br_do_del(sb, bindex, br); -+ sysaufs_brs_add(sb, bindex); -+ } ++ sysaufs_brs_del(sb, bindex); /* remove successors */ ++ dbgaufs_xino_del(br); /* remove one */ ++ au_br_do_del(sb, bindex, br); ++ sysaufs_brs_add(sb, bindex); /* append successors */ ++ dbgaufs_brs_add(sb, bindex, /*topdown*/1); /* rename successors */ + + if (!bindex) { + au_cpup_attr_all(d_inode(root), /*force*/1); @@ -3993,8 +4184,6 @@ diff -urN /usr/share/empty/fs/aufs/branch.c linux/fs/aufs/branch.c + if (au_opt_test(mnt_flags, PLINK)) + au_plink_half_refresh(sb, br_id); + -+ if (au_xino_brid(sb) == br_id) -+ au_xino_brid_set(sb, -1); + goto out; /* success */ + +out_wh: @@ -4024,7 +4213,8 @@ diff -urN /usr/share/empty/fs/aufs/branch.c linux/fs/aufs/branch.c + + err = copy_from_user(&ibusy, arg, sizeof(ibusy)); + if (!err) -+ err = !access_ok(VERIFY_WRITE, &arg->h_ino, sizeof(arg->h_ino)); ++ /* VERIFY_WRITE */ ++ err = !access_ok(&arg->h_ino, sizeof(arg->h_ino)); + if (unlikely(err)) { + err = -EFAULT; + AuTraceErr(err); @@ -4270,7 +4460,7 @@ diff -urN /usr/share/empty/fs/aufs/branch.c linux/fs/aufs/branch.c + if (br->br_wbr) { + err = au_wbr_init(br, sb, mod->perm); + if (unlikely(err)) { -+ au_delayed_kfree(br->br_wbr); ++ au_kfree_rcu(br->br_wbr); + br->br_wbr = NULL; + } + } @@ -4282,7 +4472,7 @@ diff -urN /usr/share/empty/fs/aufs/branch.c linux/fs/aufs/branch.c + if (!au_br_fhsm(mod->perm)) { + /* fhsm --> non-fhsm */ + au_br_fhsm_fin(br->br_fhsm); -+ au_delayed_kfree(br->br_fhsm); ++ au_kfree_rcu(br->br_fhsm); + br->br_fhsm = NULL; + } + } else if (au_br_fhsm(mod->perm)) @@ -4294,8 +4484,7 @@ diff -urN /usr/share/empty/fs/aufs/branch.c linux/fs/aufs/branch.c + goto out; /* success */ + +out_bf: -+ if (bf) -+ au_delayed_kfree(bf); ++ au_kfree_try_rcu(bf); +out: + AuTraceErr(err); + return err; @@ -4320,10 +4509,11 @@ diff -urN /usr/share/empty/fs/aufs/branch.c linux/fs/aufs/branch.c +} diff -urN /usr/share/empty/fs/aufs/branch.h linux/fs/aufs/branch.h --- /usr/share/empty/fs/aufs/branch.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/branch.h 2017-05-06 22:16:52.774886674 +0200 -@@ -0,0 +1,321 @@ ++++ linux/fs/aufs/branch.h 2019-07-08 09:31:38.324384703 +0200 +@@ -0,0 +1,366 @@ ++/* SPDX-License-Identifier: GPL-2.0 */ +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -4349,15 +4539,19 @@ diff -urN /usr/share/empty/fs/aufs/branch.h linux/fs/aufs/branch.h +#ifdef __KERNEL__ + +#include ++#include "dirren.h" +#include "dynop.h" ++#include "lcnt.h" +#include "rwsem.h" +#include "super.h" + +/* ---------------------------------------------------------------------- */ + +/* a xino file */ -+struct au_xino_file { -+ struct file *xi_file; ++struct au_xino { ++ struct file **xi_file; ++ unsigned int xi_nfile; ++ + struct { + spinlock_t spin; + ino_t *array; @@ -4367,11 +4561,12 @@ diff -urN /usr/share/empty/fs/aufs/branch.h linux/fs/aufs/branch.h + wait_queue_head_t wqh; + } xi_nondir; + -+ /* todo: make xino files an array to support huge inode number */ ++ struct mutex xi_mtx; /* protects xi_file array */ ++ struct hlist_bl_head xi_writing; + -+#ifdef CONFIG_DEBUG_FS -+ struct dentry *xi_dbgaufs; -+#endif ++ atomic_t xi_truncating; ++ ++ struct kref xi_kref; +}; + +/* File-based Hierarchical Storage Management */ @@ -4422,7 +4617,7 @@ diff -urN /usr/share/empty/fs/aufs/branch.h linux/fs/aufs/branch.h + +/* protected by superblock rwsem */ +struct au_branch { -+ struct au_xino_file br_xino; ++ struct au_xino *br_xino; + + aufs_bindex_t br_id; + @@ -4430,14 +4625,12 @@ diff -urN /usr/share/empty/fs/aufs/branch.h linux/fs/aufs/branch.h + struct path br_path; + spinlock_t br_dykey_lock; + struct au_dykey *br_dykey[AuBrDynOp]; -+ struct percpu_counter br_count; ++ au_lcnt_t br_nfiles; /* opened files */ ++ au_lcnt_t br_count; /* in-use for other */ + + struct au_wbr *br_wbr; + struct au_br_fhsm *br_fhsm; + -+ /* xino truncation */ -+ atomic_t br_xino_running; -+ +#ifdef CONFIG_AUFS_HFSNOTIFY + struct au_br_hfsnotify *br_hfsn; +#endif @@ -4446,6 +4639,12 @@ diff -urN /usr/share/empty/fs/aufs/branch.h linux/fs/aufs/branch.h + /* entries under sysfs per mount-point */ + struct au_brsysfs br_sysfs[AuBrSysfs_Last]; +#endif ++ ++#ifdef CONFIG_DEBUG_FS ++ struct dentry *br_dbgaufs; /* xino */ ++#endif ++ ++ struct au_dr_br br_dirren; +}; + +/* ---------------------------------------------------------------------- */ @@ -4465,34 +4664,9 @@ diff -urN /usr/share/empty/fs/aufs/branch.h linux/fs/aufs/branch.h + return au_br_mnt(br)->mnt_sb; +} + -+static inline void au_br_get(struct au_branch *br) -+{ -+ percpu_counter_inc(&br->br_count); -+} -+ -+static inline void au_br_put(struct au_branch *br) -+{ -+ percpu_counter_dec(&br->br_count); -+} -+ -+static inline s64 au_br_count(struct au_branch *br) -+{ -+ return percpu_counter_sum(&br->br_count); -+} -+ -+static inline void au_br_count_init(struct au_branch *br) -+{ -+ percpu_counter_init(&br->br_count, 0, GFP_NOFS); -+} -+ -+static inline void au_br_count_fin(struct au_branch *br) -+{ -+ percpu_counter_destroy(&br->br_count); -+} -+ +static inline int au_br_rdonly(struct au_branch *br) +{ -+ return ((au_br_sb(br)->s_flags & MS_RDONLY) ++ return (sb_rdonly(au_br_sb(br)) + || !au_br_writable(br->br_perm)) + ? -EROFS : 0; +} @@ -4518,6 +4692,28 @@ diff -urN /usr/share/empty/fs/aufs/branch.h linux/fs/aufs/branch.h + return err; +} + ++static inline void au_xino_get(struct au_branch *br) ++{ ++ struct au_xino *xi; ++ ++ xi = br->br_xino; ++ if (xi) ++ kref_get(&xi->xi_kref); ++} ++ ++static inline int au_xino_count(struct au_branch *br) ++{ ++ int v; ++ struct au_xino *xi; ++ ++ v = 0; ++ xi = br->br_xino; ++ if (xi) ++ v = kref_read(&xi->xi_kref); ++ ++ return v; ++} ++ +/* ---------------------------------------------------------------------- */ + +/* branch.c */ @@ -4541,34 +4737,73 @@ diff -urN /usr/share/empty/fs/aufs/branch.h linux/fs/aufs/branch.h +/* xino.c */ +static const loff_t au_loff_max = LLONG_MAX; + -+int au_xib_trunc(struct super_block *sb); ++aufs_bindex_t au_xi_root(struct super_block *sb, struct dentry *dentry); ++struct file *au_xino_create(struct super_block *sb, char *fpath, int silent, ++ int wbrtop); ++struct file *au_xino_create2(struct super_block *sb, struct path *base, ++ struct file *copy_src); ++struct au_xi_new { ++ struct au_xino *xi; /* switch between xino and xigen */ ++ int idx; ++ struct path *base; ++ struct file *copy_src; ++}; ++struct file *au_xi_new(struct super_block *sb, struct au_xi_new *xinew); ++ ++int au_xino_read(struct super_block *sb, aufs_bindex_t bindex, ino_t h_ino, ++ ino_t *ino); ++int au_xino_write(struct super_block *sb, aufs_bindex_t bindex, ino_t h_ino, ++ ino_t ino); +ssize_t xino_fread(vfs_readf_t func, struct file *file, void *buf, size_t size, + loff_t *pos); +ssize_t xino_fwrite(vfs_writef_t func, struct file *file, void *buf, + size_t size, loff_t *pos); -+struct file *au_xino_create2(struct file *base_file, struct file *copy_src); -+struct file *au_xino_create(struct super_block *sb, char *fname, int silent); -+ino_t au_xino_new_ino(struct super_block *sb); -+void au_xino_delete_inode(struct inode *inode, const int unlinked); -+int au_xino_write(struct super_block *sb, aufs_bindex_t bindex, ino_t h_ino, -+ ino_t ino); -+int au_xino_read(struct super_block *sb, aufs_bindex_t bindex, ino_t h_ino, -+ ino_t *ino); -+int au_xino_br(struct super_block *sb, struct au_branch *br, ino_t hino, -+ struct file *base_file, int do_test); -+int au_xino_trunc(struct super_block *sb, aufs_bindex_t bindex); ++ ++int au_xib_trunc(struct super_block *sb); ++int au_xino_trunc(struct super_block *sb, aufs_bindex_t bindex, int idx_begin); ++ ++struct au_xino *au_xino_alloc(unsigned int nfile); ++int au_xino_put(struct au_branch *br); ++struct file *au_xino_file1(struct au_xino *xi); + +struct au_opt_xino; -+int au_xino_set(struct super_block *sb, struct au_opt_xino *xino, int remount); +void au_xino_clr(struct super_block *sb); ++int au_xino_set(struct super_block *sb, struct au_opt_xino *xiopt, int remount); +struct file *au_xino_def(struct super_block *sb); -+int au_xino_path(struct seq_file *seq, struct file *file); ++int au_xino_init_br(struct super_block *sb, struct au_branch *br, ino_t hino, ++ struct path *base); ++ ++ino_t au_xino_new_ino(struct super_block *sb); ++void au_xino_delete_inode(struct inode *inode, const int unlinked); + +void au_xinondir_leave(struct super_block *sb, aufs_bindex_t bindex, + ino_t h_ino, int idx); +int au_xinondir_enter(struct super_block *sb, aufs_bindex_t bindex, ino_t h_ino, + int *idx); + ++int au_xino_path(struct seq_file *seq, struct file *file); ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* @idx is signed to accept -1 meaning the first file */ ++static inline struct file *au_xino_file(struct au_xino *xi, int idx) ++{ ++ struct file *file; ++ ++ file = NULL; ++ if (!xi) ++ goto out; ++ ++ if (idx >= 0) { ++ if (idx < xi->xi_nfile) ++ file = xi->xi_file[idx]; ++ } else ++ file = au_xino_file1(xi); ++ ++out: ++ return file; ++} ++ +/* ---------------------------------------------------------------------- */ + +/* Superblock to branch */ @@ -4590,16 +4825,6 @@ diff -urN /usr/share/empty/fs/aufs/branch.h linux/fs/aufs/branch.h + return au_br_sb(au_sbr(sb, bindex)); +} + -+static inline void au_sbr_get(struct super_block *sb, aufs_bindex_t bindex) -+{ -+ au_br_get(au_sbr(sb, bindex)); -+} -+ -+static inline void au_sbr_put(struct super_block *sb, aufs_bindex_t bindex) -+{ -+ au_br_put(au_sbr(sb, bindex)); -+} -+ +static inline int au_sbr_perm(struct super_block *sb, aufs_bindex_t bindex) +{ + return au_sbr(sb, bindex)->br_perm; @@ -4612,15 +4837,24 @@ diff -urN /usr/share/empty/fs/aufs/branch.h linux/fs/aufs/branch.h + +/* ---------------------------------------------------------------------- */ + ++#define wbr_wh_read_lock(wbr) au_rw_read_lock(&(wbr)->wbr_wh_rwsem) ++#define wbr_wh_write_lock(wbr) au_rw_write_lock(&(wbr)->wbr_wh_rwsem) ++#define wbr_wh_read_trylock(wbr) au_rw_read_trylock(&(wbr)->wbr_wh_rwsem) ++#define wbr_wh_write_trylock(wbr) au_rw_write_trylock(&(wbr)->wbr_wh_rwsem) +/* -+ * wbr_wh_read_lock, wbr_wh_write_lock -+ * wbr_wh_read_unlock, wbr_wh_write_unlock, wbr_wh_downgrade_lock -+ */ -+AuSimpleRwsemFuncs(wbr_wh, struct au_wbr *wbr, &wbr->wbr_wh_rwsem); ++#define wbr_wh_read_trylock_nested(wbr) \ ++ au_rw_read_trylock_nested(&(wbr)->wbr_wh_rwsem) ++#define wbr_wh_write_trylock_nested(wbr) \ ++ au_rw_write_trylock_nested(&(wbr)->wbr_wh_rwsem) ++*/ ++ ++#define wbr_wh_read_unlock(wbr) au_rw_read_unlock(&(wbr)->wbr_wh_rwsem) ++#define wbr_wh_write_unlock(wbr) au_rw_write_unlock(&(wbr)->wbr_wh_rwsem) ++#define wbr_wh_downgrade_lock(wbr) au_rw_dgrade_lock(&(wbr)->wbr_wh_rwsem) + -+#define WbrWhMustNoWaiters(wbr) AuRwMustNoWaiters(&wbr->wbr_wh_rwsem) -+#define WbrWhMustAnyLock(wbr) AuRwMustAnyLock(&wbr->wbr_wh_rwsem) -+#define WbrWhMustWriteLock(wbr) AuRwMustWriteLock(&wbr->wbr_wh_rwsem) ++#define WbrWhMustNoWaiters(wbr) AuRwMustNoWaiters(&(wbr)->wbr_wh_rwsem) ++#define WbrWhMustAnyLock(wbr) AuRwMustAnyLock(&(wbr)->wbr_wh_rwsem) ++#define WbrWhMustWriteLock(wbr) AuRwMustWriteLock(&(wbr)->wbr_wh_rwsem) + +/* ---------------------------------------------------------------------- */ + @@ -4645,8 +4879,9 @@ diff -urN /usr/share/empty/fs/aufs/branch.h linux/fs/aufs/branch.h +#endif /* __AUFS_BRANCH_H__ */ diff -urN /usr/share/empty/fs/aufs/conf.mk linux/fs/aufs/conf.mk --- /usr/share/empty/fs/aufs/conf.mk 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/conf.mk 2017-05-06 22:16:52.774886674 +0200 -@@ -0,0 +1,38 @@ ++++ linux/fs/aufs/conf.mk 2018-06-04 09:08:09.181412645 +0200 +@@ -0,0 +1,40 @@ ++# SPDX-License-Identifier: GPL-2.0 + +AuConfStr = CONFIG_AUFS_FS=${CONFIG_AUFS_FS} + @@ -4663,6 +4898,7 @@ diff -urN /usr/share/empty/fs/aufs/conf.mk linux/fs/aufs/conf.mk + XATTR \ + FHSM \ + RDU \ ++ DIRREN \ + SHWH \ + BR_RAMFS \ + BR_FUSE POLL \ @@ -4687,10 +4923,11 @@ diff -urN /usr/share/empty/fs/aufs/conf.mk linux/fs/aufs/conf.mk +-include ${srctree}/${src}/conf_priv.mk diff -urN /usr/share/empty/fs/aufs/cpup.c linux/fs/aufs/cpup.c --- /usr/share/empty/fs/aufs/cpup.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/cpup.c 2017-05-06 22:16:52.774886674 +0200 -@@ -0,0 +1,1407 @@ ++++ linux/fs/aufs/cpup.c 2019-03-05 12:13:00.135890907 +0100 +@@ -0,0 +1,1458 @@ ++// SPDX-License-Identifier: GPL-2.0 +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -5027,9 +5264,11 @@ diff -urN /usr/share/empty/fs/aufs/cpup.c linux/fs/aufs/cpup.c + unsigned long blksize; + unsigned char do_kfree; + char *buf; ++ struct super_block *h_sb; + + err = -ENOMEM; -+ blksize = dst->f_path.dentry->d_sb->s_blocksize; ++ h_sb = file_inode(dst)->i_sb; ++ blksize = h_sb->s_blocksize; + if (!blksize || PAGE_SIZE < blksize) + blksize = PAGE_SIZE; + AuDbg("blksize %lu\n", blksize); @@ -5047,15 +5286,79 @@ diff -urN /usr/share/empty/fs/aufs/cpup.c linux/fs/aufs/cpup.c + src->f_pos = 0; + dst->f_pos = 0; + err = au_do_copy_file(dst, src, len, buf, blksize); -+ if (do_kfree) -+ au_delayed_kfree(buf); -+ else -+ au_delayed_free_page((unsigned long)buf); ++ if (do_kfree) { ++ AuDebugOn(!au_kfree_do_sz_test(blksize)); ++ au_kfree_do_rcu(buf); ++ } else ++ free_page((unsigned long)buf); + +out: + return err; +} + ++static int au_do_copy(struct file *dst, struct file *src, loff_t len) ++{ ++ int err; ++ struct super_block *h_src_sb; ++ struct inode *h_src_inode; ++ ++ h_src_inode = file_inode(src); ++ h_src_sb = h_src_inode->i_sb; ++ ++ /* XFS acquires inode_lock */ ++ if (!au_test_xfs(h_src_sb)) ++ err = au_copy_file(dst, src, len); ++ else { ++ inode_unlock_shared(h_src_inode); ++ err = au_copy_file(dst, src, len); ++ inode_lock_shared_nested(h_src_inode, AuLsc_I_CHILD); ++ } ++ ++ return err; ++} ++ ++static int au_clone_or_copy(struct file *dst, struct file *src, loff_t len) ++{ ++ int err; ++ loff_t lo; ++ struct super_block *h_src_sb; ++ struct inode *h_src_inode; ++ ++ h_src_inode = file_inode(src); ++ h_src_sb = h_src_inode->i_sb; ++ if (h_src_sb != file_inode(dst)->i_sb ++ || !dst->f_op->remap_file_range) { ++ err = au_do_copy(dst, src, len); ++ goto out; ++ } ++ ++ if (!au_test_nfs(h_src_sb)) { ++ inode_unlock_shared(h_src_inode); ++ lo = vfsub_clone_file_range(src, dst, len); ++ inode_lock_shared_nested(h_src_inode, AuLsc_I_CHILD); ++ } else ++ lo = vfsub_clone_file_range(src, dst, len); ++ if (lo == len) { ++ err = 0; ++ goto out; /* success */ ++ } else if (lo >= 0) ++ /* todo: possible? */ ++ /* paritially succeeded */ ++ AuDbg("lo %lld, len %lld. Retrying.\n", lo, len); ++ else if (lo != -EOPNOTSUPP) { ++ /* older XFS has a condition in cloning */ ++ err = lo; ++ goto out; ++ } ++ ++ /* the backend fs on NFS may not support cloning */ ++ err = au_do_copy(dst, src, len); ++ ++out: ++ AuTraceErr(err); ++ return err; ++} ++ +/* + * to support a sparse file which is opened with O_APPEND, + * we need to close the file. @@ -5070,20 +5373,18 @@ diff -urN /usr/share/empty/fs/aufs/cpup.c linux/fs/aufs/cpup.c + struct dentry *dentry; + int force_wr; + struct file *file; -+ void *label; + } *f, file[] = { + { + .bindex = cpg->bsrc, + .flags = O_RDONLY | O_NOATIME | O_LARGEFILE, -+ .label = &&out + }, + { + .bindex = cpg->bdst, + .flags = O_WRONLY | O_NOATIME | O_LARGEFILE, + .force_wr = !!au_ftest_cpup(cpg->flags, RWDST), -+ .label = &&out_src + } + }; ++ struct au_branch *br; + struct super_block *sb, *h_src_sb; + struct inode *h_src_inode; + struct task_struct *tsk = current; @@ -5095,9 +5396,13 @@ diff -urN /usr/share/empty/fs/aufs/cpup.c linux/fs/aufs/cpup.c + f->dentry = au_h_dptr(cpg->dentry, f->bindex); + f->file = au_h_open(cpg->dentry, f->bindex, f->flags, + /*file*/NULL, f->force_wr); -+ err = PTR_ERR(f->file); -+ if (IS_ERR(f->file)) -+ goto *f->label; ++ if (IS_ERR(f->file)) { ++ err = PTR_ERR(f->file); ++ if (i == SRC) ++ goto out; ++ else ++ goto out_src; ++ } + } + + /* try stopping to update while we copyup */ @@ -5105,32 +5410,13 @@ diff -urN /usr/share/empty/fs/aufs/cpup.c linux/fs/aufs/cpup.c + h_src_sb = h_src_inode->i_sb; + if (!au_test_nfs(h_src_sb)) + IMustLock(h_src_inode); -+ -+ if (h_src_sb != file_inode(file[DST].file)->i_sb -+ || !file[DST].file->f_op->clone_file_range) -+ err = au_copy_file(file[DST].file, file[SRC].file, cpg->len); -+ else { -+ if (!au_test_nfs(h_src_sb)) { -+ inode_unlock(h_src_inode); -+ err = vfsub_clone_file_range(file[SRC].file, -+ file[DST].file, cpg->len); -+ inode_lock(h_src_inode); -+ } else -+ err = vfsub_clone_file_range(file[SRC].file, -+ file[DST].file, cpg->len); -+ if (unlikely(err == -EOPNOTSUPP && au_test_nfs(h_src_sb))) -+ /* the backend fs on NFS may not support cloning */ -+ err = au_copy_file(file[DST].file, file[SRC].file, -+ cpg->len); -+ AuTraceErr(err); -+ } ++ err = au_clone_or_copy(file[DST].file, file[SRC].file, cpg->len); + + /* i wonder if we had O_NO_DELAY_FPUT flag */ + if (tsk->flags & PF_KTHREAD) + __fput_sync(file[DST].file); + else { -+ WARN(1, "%pD\nPlease report this warning to aufs-users ML", -+ file[DST].file); ++ /* it happened actually */ + fput(file[DST].file); + /* + * too bad. @@ -5140,11 +5426,13 @@ diff -urN /usr/share/empty/fs/aufs/cpup.c linux/fs/aufs/cpup.c + task_work_run(); + flush_delayed_fput(); + } -+ au_sbr_put(sb, file[DST].bindex); ++ br = au_sbr(sb, file[DST].bindex); ++ au_lcnt_dec(&br->br_nfiles); + +out_src: + fput(file[SRC].file); -+ au_sbr_put(sb, file[SRC].bindex); ++ br = au_sbr(sb, file[SRC].bindex); ++ au_lcnt_dec(&br->br_nfiles); +out: + return err; +} @@ -5164,7 +5452,7 @@ diff -urN /usr/share/empty/fs/aufs/cpup.c linux/fs/aufs/cpup.c + cpg->len = l; + if (cpg->len) { + /* try stopping to update while we are referencing */ -+ inode_lock_nested(h_src_inode, AuLsc_I_CHILD); ++ inode_lock_shared_nested(h_src_inode, AuLsc_I_CHILD); + au_pin_hdir_unlock(cpg->pin); + + h_path.dentry = au_h_dptr(cpg->dentry, cpg->bsrc); @@ -5173,20 +5461,20 @@ diff -urN /usr/share/empty/fs/aufs/cpup.c linux/fs/aufs/cpup.c + if (!au_test_nfs(h_src_inode->i_sb)) + err = vfsub_getattr(&h_path, &h_src_attr->st); + else { -+ inode_unlock(h_src_inode); ++ inode_unlock_shared(h_src_inode); + err = vfsub_getattr(&h_path, &h_src_attr->st); -+ inode_lock_nested(h_src_inode, AuLsc_I_CHILD); ++ inode_lock_shared_nested(h_src_inode, AuLsc_I_CHILD); + } + if (unlikely(err)) { -+ inode_unlock(h_src_inode); ++ inode_unlock_shared(h_src_inode); + goto out; + } + h_src_attr->valid = 1; + if (!au_test_nfs(h_src_inode->i_sb)) { + err = au_cp_regular(cpg); -+ inode_unlock(h_src_inode); ++ inode_unlock_shared(h_src_inode); + } else { -+ inode_unlock(h_src_inode); ++ inode_unlock_shared(h_src_inode); + err = au_cp_regular(cpg); + } + rerr = au_pin_hdir_relock(cpg->pin); @@ -5231,7 +5519,7 @@ diff -urN /usr/share/empty/fs/aufs/cpup.c linux/fs/aufs/cpup.c + sym.k[symlen] = 0; + err = vfsub_symlink(h_dir, h_path, sym.k); + } -+ au_delayed_free_page((unsigned long)sym.k); ++ free_page((unsigned long)sym.k); + +out: + return err; @@ -5276,7 +5564,7 @@ diff -urN /usr/share/empty/fs/aufs/cpup.c linux/fs/aufs/cpup.c + + /* + * strange behaviour from the users view, -+ * particularry setattr case ++ * particularly setattr case + */ + dir = d_inode(dst_parent); + if (au_ibtop(dir) == cpg->bdst) @@ -5332,8 +5620,7 @@ diff -urN /usr/share/empty/fs/aufs/cpup.c linux/fs/aufs/cpup.c + switch (mode & S_IFMT) { + case S_IFREG: + isreg = 1; -+ err = vfsub_create(h_dir, &h_path, S_IRUSR | S_IWUSR, -+ /*want_excl*/true); ++ err = vfsub_create(h_dir, &h_path, 0600, /*want_excl*/true); + if (!err) + err = au_do_cpup_regular(cpg, h_src_attr); + break; @@ -5438,7 +5725,7 @@ diff -urN /usr/share/empty/fs/aufs/cpup.c linux/fs/aufs/cpup.c + * in link/rename cases, @dst_parent may be different from the real one. + * basic->bsrc can be larger than basic->bdst. + * aufs doesn't touch the credential so -+ * security_inode_copy_up{,_xattr}() are unnecrssary. ++ * security_inode_copy_up{,_xattr}() are unnecessary. + */ +static int au_cpup_single(struct au_cp_generic *cpg, struct dentry *dst_parent) +{ @@ -5449,7 +5736,7 @@ diff -urN /usr/share/empty/fs/aufs/cpup.c linux/fs/aufs/cpup.c + struct inode *dst_inode, *h_dir, *inode, *delegated, *src_inode; + struct super_block *sb; + struct au_branch *br; -+ /* to reuduce stack size */ ++ /* to reduce stack size */ + struct { + struct au_dtime dt; + struct path h_path; @@ -5605,7 +5892,7 @@ diff -urN /usr/share/empty/fs/aufs/cpup.c linux/fs/aufs/cpup.c + } +out_parent: + dput(dst_parent); -+ au_delayed_kfree(a); ++ au_kfree_rcu(a); +out: + return err; +} @@ -6098,10 +6385,11 @@ diff -urN /usr/share/empty/fs/aufs/cpup.c linux/fs/aufs/cpup.c +} diff -urN /usr/share/empty/fs/aufs/cpup.h linux/fs/aufs/cpup.h --- /usr/share/empty/fs/aufs/cpup.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/cpup.h 2017-05-06 22:16:52.774886674 +0200 -@@ -0,0 +1,94 @@ ++++ linux/fs/aufs/cpup.h 2019-03-05 12:13:00.135890907 +0100 +@@ -0,0 +1,100 @@ ++/* SPDX-License-Identifier: GPL-2.0 */ +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -6161,6 +6449,11 @@ diff -urN /usr/share/empty/fs/aufs/cpup.h linux/fs/aufs/cpup.h +#define AuCpup_RWDST (1 << 5) /* force write target even if + the branch is marked as RO */ + ++#ifndef CONFIG_AUFS_BR_HFSPLUS ++#undef AuCpup_HOPEN ++#define AuCpup_HOPEN 0 ++#endif ++ +#define au_ftest_cpup(flags, name) ((flags) & AuCpup_##name) +#define au_fset_cpup(flags, name) \ + do { (flags) |= AuCpup_##name; } while (0) @@ -6186,7 +6479,7 @@ diff -urN /usr/share/empty/fs/aufs/cpup.h linux/fs/aufs/cpup.h +struct au_dtime { + struct dentry *dt_dentry; + struct path dt_h_path; -+ struct timespec dt_atime, dt_mtime; ++ struct timespec64 dt_atime, dt_mtime; +}; +void au_dtime_store(struct au_dtime *dt, struct dentry *dentry, + struct path *h_path); @@ -6196,10 +6489,11 @@ diff -urN /usr/share/empty/fs/aufs/cpup.h linux/fs/aufs/cpup.h +#endif /* __AUFS_CPUP_H__ */ diff -urN /usr/share/empty/fs/aufs/dbgaufs.c linux/fs/aufs/dbgaufs.c --- /usr/share/empty/fs/aufs/dbgaufs.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/dbgaufs.c 2017-05-06 22:16:52.774886674 +0200 -@@ -0,0 +1,438 @@ ++++ linux/fs/aufs/dbgaufs.c 2019-03-05 12:13:00.135890907 +0100 +@@ -0,0 +1,526 @@ ++// SPDX-License-Identifier: GPL-2.0 +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -6227,7 +6521,7 @@ diff -urN /usr/share/empty/fs/aufs/dbgaufs.c linux/fs/aufs/dbgaufs.c +#endif + +static struct dentry *dbgaufs; -+static const mode_t dbgaufs_mode = S_IRUSR | S_IRGRP | S_IROTH; ++static const mode_t dbgaufs_mode = 0444; + +/* 20 is max digits length of ulong 64 */ +struct dbgaufs_arg { @@ -6241,11 +6535,19 @@ diff -urN /usr/share/empty/fs/aufs/dbgaufs.c linux/fs/aufs/dbgaufs.c +static int dbgaufs_xi_release(struct inode *inode __maybe_unused, + struct file *file) +{ -+ au_delayed_kfree(file->private_data); ++ void *p; ++ ++ p = file->private_data; ++ if (p) { ++ /* this is struct dbgaufs_arg */ ++ AuDebugOn(!au_kfree_sz_test(p)); ++ au_kfree_do_rcu(p); ++ } + return 0; +} + -+static int dbgaufs_xi_open(struct file *xf, struct file *file, int do_fcnt) ++static int dbgaufs_xi_open(struct file *xf, struct file *file, int do_fcnt, ++ int cnt) +{ + int err; + struct kstat st; @@ -6266,8 +6568,8 @@ diff -urN /usr/share/empty/fs/aufs/dbgaufs.c linux/fs/aufs/dbgaufs.c + if (!err) { + if (do_fcnt) + p->n = snprintf -+ (p->a, sizeof(p->a), "%ld, %llux%u %lld\n", -+ (long)file_count(xf), st.blocks, st.blksize, ++ (p->a, sizeof(p->a), "%d, %llux%u %lld\n", ++ cnt, st.blocks, st.blksize, + (long long)st.size); + else + p->n = snprintf(p->a, sizeof(p->a), "%llux%u %lld\n", @@ -6281,7 +6583,6 @@ diff -urN /usr/share/empty/fs/aufs/dbgaufs.c linux/fs/aufs/dbgaufs.c + +out: + return err; -+ +} + +static ssize_t dbgaufs_xi_read(struct file *file, char __user *buf, @@ -6303,7 +6604,7 @@ diff -urN /usr/share/empty/fs/aufs/dbgaufs.c linux/fs/aufs/dbgaufs.c +static int dbgaufs_plink_release(struct inode *inode __maybe_unused, + struct file *file) +{ -+ au_delayed_free_page((unsigned long)file->private_data); ++ free_page((unsigned long)file->private_data); + return 0; +} + @@ -6314,7 +6615,7 @@ diff -urN /usr/share/empty/fs/aufs/dbgaufs.c linux/fs/aufs/dbgaufs.c + struct dbgaufs_plink_arg *p; + struct au_sbinfo *sbinfo; + struct super_block *sb; -+ struct au_sphlhead *sphl; ++ struct hlist_bl_head *hbl; + + err = -ENOMEM; + p = (void *)get_zeroed_page(GFP_NOFS); @@ -6334,10 +6635,9 @@ diff -urN /usr/share/empty/fs/aufs/dbgaufs.c linux/fs/aufs/dbgaufs.c + limit -= n; + + sum = 0; -+ for (i = 0, sphl = sbinfo->si_plink; -+ i < AuPlink_NHASH; -+ i++, sphl++) { -+ n = au_sphl_count(sphl); ++ for (i = 0, hbl = sbinfo->si_plink; i < AuPlink_NHASH; ++ i++, hbl++) { ++ n = au_hbl_count(hbl); + sum += n; + + n = snprintf(p->a + p->n, limit, "%lu ", n); @@ -6367,7 +6667,7 @@ diff -urN /usr/share/empty/fs/aufs/dbgaufs.c linux/fs/aufs/dbgaufs.c + goto out; /* success */ + +out_free: -+ au_delayed_free_page((unsigned long)p); ++ free_page((unsigned long)p); +out: + return err; +} @@ -6399,7 +6699,7 @@ diff -urN /usr/share/empty/fs/aufs/dbgaufs.c linux/fs/aufs/dbgaufs.c + sbinfo = inode->i_private; + sb = sbinfo->si_sb; + si_noflush_read_lock(sb); -+ err = dbgaufs_xi_open(sbinfo->si_xib, file, /*do_fcnt*/0); ++ err = dbgaufs_xi_open(sbinfo->si_xib, file, /*do_fcnt*/0, /*cnt*/0); + si_read_unlock(sb); + return err; +} @@ -6417,35 +6717,61 @@ diff -urN /usr/share/empty/fs/aufs/dbgaufs.c linux/fs/aufs/dbgaufs.c + +static int dbgaufs_xino_open(struct inode *inode, struct file *file) +{ -+ int err; ++ int err, idx; + long l; ++ aufs_bindex_t bindex; ++ char *p, a[sizeof(DbgaufsXi_PREFIX) + 8]; + struct au_sbinfo *sbinfo; + struct super_block *sb; ++ struct au_xino *xi; + struct file *xf; + struct qstr *name; ++ struct au_branch *br; + + err = -ENOENT; -+ xf = NULL; + name = &file->f_path.dentry->d_name; + if (unlikely(name->len < sizeof(DbgaufsXi_PREFIX) + || memcmp(name->name, DbgaufsXi_PREFIX, + sizeof(DbgaufsXi_PREFIX) - 1))) + goto out; -+ err = kstrtol(name->name + sizeof(DbgaufsXi_PREFIX) - 1, 10, &l); ++ ++ AuDebugOn(name->len >= sizeof(a)); ++ memcpy(a, name->name, name->len); ++ a[name->len] = '\0'; ++ p = strchr(a, '-'); ++ if (p) ++ *p = '\0'; ++ err = kstrtol(a + sizeof(DbgaufsXi_PREFIX) - 1, 10, &l); + if (unlikely(err)) + goto out; ++ bindex = l; ++ idx = 0; ++ if (p) { ++ err = kstrtol(p + 1, 10, &l); ++ if (unlikely(err)) ++ goto out; ++ idx = l; ++ } + ++ err = -ENOENT; + sbinfo = inode->i_private; + sb = sbinfo->si_sb; + si_noflush_read_lock(sb); -+ if (l <= au_sbbot(sb)) { -+ xf = au_sbr(sb, (aufs_bindex_t)l)->br_xino.xi_file; -+ err = dbgaufs_xi_open(xf, file, /*do_fcnt*/1); -+ } else -+ err = -ENOENT; -+ si_read_unlock(sb); ++ if (unlikely(bindex < 0 || bindex > au_sbbot(sb))) ++ goto out_si; ++ br = au_sbr(sb, bindex); ++ xi = br->br_xino; ++ if (unlikely(idx >= xi->xi_nfile)) ++ goto out_si; ++ xf = au_xino_file(xi, idx); ++ if (xf) ++ err = dbgaufs_xi_open(xf, file, /*do_fcnt*/1, ++ au_xino_count(br)); + ++out_si: ++ si_read_unlock(sb); +out: ++ AuTraceErr(err); + return err; +} + @@ -6456,11 +6782,25 @@ diff -urN /usr/share/empty/fs/aufs/dbgaufs.c linux/fs/aufs/dbgaufs.c + .read = dbgaufs_xi_read +}; + ++void dbgaufs_xino_del(struct au_branch *br) ++{ ++ struct dentry *dbgaufs; ++ ++ dbgaufs = br->br_dbgaufs; ++ if (!dbgaufs) ++ return; ++ ++ br->br_dbgaufs = NULL; ++ /* debugfs acquires the parent i_mutex */ ++ lockdep_off(); ++ debugfs_remove(dbgaufs); ++ lockdep_on(); ++} ++ +void dbgaufs_brs_del(struct super_block *sb, aufs_bindex_t bindex) +{ + aufs_bindex_t bbot; + struct au_branch *br; -+ struct au_xino_file *xi; + + if (!au_sbi(sb)->si_dbgaufs) + return; @@ -6468,23 +6808,70 @@ diff -urN /usr/share/empty/fs/aufs/dbgaufs.c linux/fs/aufs/dbgaufs.c + bbot = au_sbbot(sb); + for (; bindex <= bbot; bindex++) { + br = au_sbr(sb, bindex); -+ xi = &br->br_xino; -+ /* debugfs acquires the parent i_mutex */ ++ dbgaufs_xino_del(br); ++ } ++} ++ ++static void dbgaufs_br_do_add(struct super_block *sb, aufs_bindex_t bindex, ++ unsigned int idx, struct dentry *parent, ++ struct au_sbinfo *sbinfo) ++{ ++ struct au_branch *br; ++ struct dentry *d; ++ /* "xi" bindex(5) "-" idx(2) NULL */ ++ char name[sizeof(DbgaufsXi_PREFIX) + 8]; ++ ++ if (!idx) ++ snprintf(name, sizeof(name), DbgaufsXi_PREFIX "%d", bindex); ++ else ++ snprintf(name, sizeof(name), DbgaufsXi_PREFIX "%d-%u", ++ bindex, idx); ++ br = au_sbr(sb, bindex); ++ if (br->br_dbgaufs) { ++ struct qstr qstr = QSTR_INIT(name, strlen(name)); ++ ++ if (!au_qstreq(&br->br_dbgaufs->d_name, &qstr)) { ++ /* debugfs acquires the parent i_mutex */ ++ lockdep_off(); ++ d = debugfs_rename(parent, br->br_dbgaufs, parent, ++ name); ++ lockdep_on(); ++ if (unlikely(!d)) ++ pr_warn("failed renaming %pd/%s, ignored.\n", ++ parent, name); ++ } ++ } else { + lockdep_off(); -+ debugfs_remove(xi->xi_dbgaufs); ++ br->br_dbgaufs = debugfs_create_file(name, dbgaufs_mode, parent, ++ sbinfo, &dbgaufs_xino_fop); + lockdep_on(); -+ xi->xi_dbgaufs = NULL; ++ if (unlikely(!br->br_dbgaufs)) ++ pr_warn("failed creating %pd/%s, ignored.\n", ++ parent, name); + } +} + -+void dbgaufs_brs_add(struct super_block *sb, aufs_bindex_t bindex) ++static void dbgaufs_br_add(struct super_block *sb, aufs_bindex_t bindex, ++ struct dentry *parent, struct au_sbinfo *sbinfo) ++{ ++ struct au_branch *br; ++ struct au_xino *xi; ++ unsigned int u; ++ ++ br = au_sbr(sb, bindex); ++ xi = br->br_xino; ++ for (u = 0; u < xi->xi_nfile; u++) ++ dbgaufs_br_do_add(sb, bindex, u, parent, sbinfo); ++} ++ ++void dbgaufs_brs_add(struct super_block *sb, aufs_bindex_t bindex, int topdown) +{ + struct au_sbinfo *sbinfo; + struct dentry *parent; -+ struct au_branch *br; -+ struct au_xino_file *xi; + aufs_bindex_t bbot; -+ char name[sizeof(DbgaufsXi_PREFIX) + 5]; /* "xi" bindex NULL */ ++ ++ if (!au_opt_test(au_mntflags(sb), XINO)) ++ return; + + sbinfo = au_sbi(sb); + parent = sbinfo->si_dbgaufs; @@ -6492,20 +6879,12 @@ diff -urN /usr/share/empty/fs/aufs/dbgaufs.c linux/fs/aufs/dbgaufs.c + return; + + bbot = au_sbbot(sb); -+ for (; bindex <= bbot; bindex++) { -+ snprintf(name, sizeof(name), DbgaufsXi_PREFIX "%d", bindex); -+ br = au_sbr(sb, bindex); -+ xi = &br->br_xino; -+ AuDebugOn(xi->xi_dbgaufs); -+ /* debugfs acquires the parent i_mutex */ -+ lockdep_off(); -+ xi->xi_dbgaufs = debugfs_create_file(name, dbgaufs_mode, parent, -+ sbinfo, &dbgaufs_xino_fop); -+ lockdep_on(); -+ /* ignore an error */ -+ if (unlikely(!xi->xi_dbgaufs)) -+ AuWarn1("failed %s under debugfs\n", name); -+ } ++ if (topdown) ++ for (; bindex <= bbot; bindex++) ++ dbgaufs_br_add(sb, bindex, parent, sbinfo); ++ else ++ for (; bbot >= bindex; bbot--) ++ dbgaufs_br_add(sb, bbot, parent, sbinfo); +} + +/* ---------------------------------------------------------------------- */ @@ -6520,7 +6899,7 @@ diff -urN /usr/share/empty/fs/aufs/dbgaufs.c linux/fs/aufs/dbgaufs.c + sbinfo = inode->i_private; + sb = sbinfo->si_sb; + si_noflush_read_lock(sb); -+ err = dbgaufs_xi_open(sbinfo->si_xigen, file, /*do_fcnt*/0); ++ err = dbgaufs_xi_open(sbinfo->si_xigen, file, /*do_fcnt*/0, /*cnt*/0); + si_read_unlock(sb); + return err; +} @@ -6570,7 +6949,6 @@ diff -urN /usr/share/empty/fs/aufs/dbgaufs.c linux/fs/aufs/dbgaufs.c + + debugfs_remove_recursive(sbinfo->si_dbgaufs); + sbinfo->si_dbgaufs = NULL; -+ kobject_put(&sbinfo->si_kobj); +} + +int dbgaufs_si_init(struct au_sbinfo *sbinfo) @@ -6595,20 +6973,21 @@ diff -urN /usr/share/empty/fs/aufs/dbgaufs.c linux/fs/aufs/dbgaufs.c + sbinfo->si_dbgaufs = debugfs_create_dir(name, dbgaufs); + if (unlikely(!sbinfo->si_dbgaufs)) + goto out; -+ kobject_get(&sbinfo->si_kobj); -+ -+ sbinfo->si_dbgaufs_xib = debugfs_create_file -+ ("xib", dbgaufs_mode, sbinfo->si_dbgaufs, sbinfo, -+ &dbgaufs_xib_fop); -+ if (unlikely(!sbinfo->si_dbgaufs_xib)) -+ goto out_dir; + ++ /* regardless plink/noplink option */ + sbinfo->si_dbgaufs_plink = debugfs_create_file + ("plink", dbgaufs_mode, sbinfo->si_dbgaufs, sbinfo, + &dbgaufs_plink_fop); + if (unlikely(!sbinfo->si_dbgaufs_plink)) + goto out_dir; + ++ /* regardless xino/noxino option */ ++ sbinfo->si_dbgaufs_xib = debugfs_create_file ++ ("xib", dbgaufs_mode, sbinfo->si_dbgaufs, sbinfo, ++ &dbgaufs_xib_fop); ++ if (unlikely(!sbinfo->si_dbgaufs_xib)) ++ goto out_dir; ++ + err = dbgaufs_xigen_init(sbinfo); + if (!err) + goto out; /* success */ @@ -6616,6 +6995,8 @@ diff -urN /usr/share/empty/fs/aufs/dbgaufs.c linux/fs/aufs/dbgaufs.c +out_dir: + dbgaufs_si_fin(sbinfo); +out: ++ if (unlikely(err)) ++ pr_err("debugfs/aufs failed\n"); + return err; +} + @@ -6638,10 +7019,11 @@ diff -urN /usr/share/empty/fs/aufs/dbgaufs.c linux/fs/aufs/dbgaufs.c +} diff -urN /usr/share/empty/fs/aufs/dbgaufs.h linux/fs/aufs/dbgaufs.h --- /usr/share/empty/fs/aufs/dbgaufs.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/dbgaufs.h 2017-05-06 22:16:52.774886674 +0200 -@@ -0,0 +1,48 @@ ++++ linux/fs/aufs/dbgaufs.h 2019-03-05 12:13:00.135890907 +0100 +@@ -0,0 +1,53 @@ ++/* SPDX-License-Identifier: GPL-2.0 */ +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -6668,18 +7050,22 @@ diff -urN /usr/share/empty/fs/aufs/dbgaufs.h linux/fs/aufs/dbgaufs.h + +struct super_block; +struct au_sbinfo; ++struct au_branch; + +#ifdef CONFIG_DEBUG_FS +/* dbgaufs.c */ ++void dbgaufs_xino_del(struct au_branch *br); +void dbgaufs_brs_del(struct super_block *sb, aufs_bindex_t bindex); -+void dbgaufs_brs_add(struct super_block *sb, aufs_bindex_t bindex); ++void dbgaufs_brs_add(struct super_block *sb, aufs_bindex_t bindex, int topdown); +void dbgaufs_si_fin(struct au_sbinfo *sbinfo); +int dbgaufs_si_init(struct au_sbinfo *sbinfo); +void dbgaufs_fin(void); +int __init dbgaufs_init(void); +#else ++AuStubVoid(dbgaufs_xino_del, struct au_branch *br) +AuStubVoid(dbgaufs_brs_del, struct super_block *sb, aufs_bindex_t bindex) -+AuStubVoid(dbgaufs_brs_add, struct super_block *sb, aufs_bindex_t bindex) ++AuStubVoid(dbgaufs_brs_add, struct super_block *sb, aufs_bindex_t bindex, ++ int topdown) +AuStubVoid(dbgaufs_si_fin, struct au_sbinfo *sbinfo) +AuStubInt0(dbgaufs_si_init, struct au_sbinfo *sbinfo) +AuStubVoid(dbgaufs_fin, void) @@ -6690,10 +7076,11 @@ diff -urN /usr/share/empty/fs/aufs/dbgaufs.h linux/fs/aufs/dbgaufs.h +#endif /* __DBGAUFS_H__ */ diff -urN /usr/share/empty/fs/aufs/dcsub.c linux/fs/aufs/dcsub.c --- /usr/share/empty/fs/aufs/dcsub.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/dcsub.c 2017-05-06 22:16:52.774886674 +0200 ++++ linux/fs/aufs/dcsub.c 2019-03-05 12:13:00.135890907 +0100 @@ -0,0 +1,225 @@ ++// SPDX-License-Identifier: GPL-2.0 +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -6723,7 +7110,7 @@ diff -urN /usr/share/empty/fs/aufs/dcsub.c linux/fs/aufs/dcsub.c + p = dpage->dentries; + for (i = 0; i < dpage->ndentry; i++) + dput(*p++); -+ au_delayed_free_page((unsigned long)dpage->dentries); ++ free_page((unsigned long)dpage->dentries); +} + +int au_dpages_init(struct au_dcsub_pages *dpages, gfp_t gfp) @@ -6746,7 +7133,7 @@ diff -urN /usr/share/empty/fs/aufs/dcsub.c linux/fs/aufs/dcsub.c + return 0; /* success */ + +out_dpages: -+ au_delayed_kfree(dpages->dpages); ++ au_kfree_try_rcu(dpages->dpages); +out: + return err; +} @@ -6759,7 +7146,7 @@ diff -urN /usr/share/empty/fs/aufs/dcsub.c linux/fs/aufs/dcsub.c + p = dpages->dpages; + for (i = 0; i < dpages->ndpage; i++) + au_dpage_free(p++); -+ au_delayed_kfree(dpages->dpages); ++ au_kfree_try_rcu(dpages->dpages); +} + +static int au_dpages_append(struct au_dcsub_pages *dpages, @@ -6810,8 +7197,7 @@ diff -urN /usr/share/empty/fs/aufs/dcsub.c linux/fs/aufs/dcsub.c +}; + +extern void d_walk(struct dentry *parent, void *data, -+ enum d_walk_ret (*enter)(void *, struct dentry *), -+ void (*finish)(void *)); ++ enum d_walk_ret (*enter)(void *, struct dentry *)); + +struct ac_dpages_arg { + int err; @@ -6851,7 +7237,7 @@ diff -urN /usr/share/empty/fs/aufs/dcsub.c linux/fs/aufs/dcsub.c + .arg = arg + }; + -+ d_walk(root, &args, au_call_dpages_append, NULL); ++ d_walk(root, &args, au_call_dpages_append); + + return args.err; +} @@ -6919,10 +7305,11 @@ diff -urN /usr/share/empty/fs/aufs/dcsub.c linux/fs/aufs/dcsub.c +} diff -urN /usr/share/empty/fs/aufs/dcsub.h linux/fs/aufs/dcsub.h --- /usr/share/empty/fs/aufs/dcsub.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/dcsub.h 2017-05-06 22:16:52.778220093 +0200 -@@ -0,0 +1,136 @@ ++++ linux/fs/aufs/dcsub.h 2019-03-05 12:13:00.135890907 +0100 +@@ -0,0 +1,137 @@ ++/* SPDX-License-Identifier: GPL-2.0 */ +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -7059,10 +7446,11 @@ diff -urN /usr/share/empty/fs/aufs/dcsub.h linux/fs/aufs/dcsub.h +#endif /* __AUFS_DCSUB_H__ */ diff -urN /usr/share/empty/fs/aufs/debug.c linux/fs/aufs/debug.c --- /usr/share/empty/fs/aufs/debug.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/debug.c 2017-05-06 22:16:52.778220093 +0200 -@@ -0,0 +1,440 @@ ++++ linux/fs/aufs/debug.c 2019-05-06 09:03:04.816810402 +0200 +@@ -0,0 +1,441 @@ ++// SPDX-License-Identifier: GPL-2.0 +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -7082,6 +7470,7 @@ diff -urN /usr/share/empty/fs/aufs/debug.c linux/fs/aufs/debug.c + * debug print functions + */ + ++#include +#include "aufs.h" + +/* Returns 0, or -errno. arg is in kp->arg. */ @@ -7116,7 +7505,7 @@ diff -urN /usr/share/empty/fs/aufs/debug.c linux/fs/aufs/debug.c + +atomic_t aufs_debug = ATOMIC_INIT(0); +MODULE_PARM_DESC(debug, "debug print"); -+module_param_named(debug, aufs_debug, atomic_t, S_IRUGO | S_IWUSR | S_IWGRP); ++module_param_named(debug, aufs_debug, atomic_t, 0664); + +DEFINE_MUTEX(au_dbg_mtx); /* just to serialize the dbg msgs */ +char *au_plevel = KERN_DEBUG; @@ -7158,7 +7547,7 @@ diff -urN /usr/share/empty/fs/aufs/debug.c linux/fs/aufs/debug.c + return; + } + -+ dpri("deblk %u, nblk %lu, deblk %p, last{%lu, %p}, ver %lu\n", ++ dpri("deblk %u, nblk %lu, deblk %p, last{%lu, %p}, ver %llu\n", + vdir->vd_deblk_sz, vdir->vd_nblk, vdir->vd_deblk, + vdir->vd_last.ul, vdir->vd_last.p.deblk, vdir->vd_version); + for (ul = 0; ul < vdir->vd_nblk; ul++) { @@ -7193,9 +7582,9 @@ diff -urN /usr/share/empty/fs/aufs/debug.c linux/fs/aufs/debug.c + inode->i_ino, inode->i_sb ? au_sbtype(inode->i_sb) : "??", + atomic_read(&inode->i_count), inode->i_nlink, inode->i_mode, + i_size_read(inode), (unsigned long long)inode->i_blocks, -+ hn, (long long)timespec_to_ns(&inode->i_ctime) & 0x0ffff, ++ hn, (long long)timespec64_to_ns(&inode->i_ctime) & 0x0ffff, + inode->i_mapping ? inode->i_mapping->nrpages : 0, -+ inode->i_state, inode->i_flags, inode->i_version, ++ inode->i_state, inode->i_flags, inode_peek_iversion(inode), + inode->i_generation, + l ? ", wh " : "", l, n); + return 0; @@ -7362,13 +7751,14 @@ diff -urN /usr/share/empty/fs/aufs/debug.c linux/fs/aufs/debug.c + if (!sb || IS_ERR(sb)) + goto out; + -+ dpri("s%d: {perm 0x%x, id %d, cnt %lld, wbr %p}, " ++ dpri("s%d: {perm 0x%x, id %d, wbr %p}, " + "%s, dev 0x%02x%02x, flags 0x%lx, cnt %d, active %d, " + "xino %d\n", -+ bindex, br->br_perm, br->br_id, au_br_count(br), -+ br->br_wbr, au_sbtype(sb), MAJOR(sb->s_dev), MINOR(sb->s_dev), ++ bindex, br->br_perm, br->br_id, br->br_wbr, ++ au_sbtype(sb), MAJOR(sb->s_dev), MINOR(sb->s_dev), + sb->s_flags, sb->s_count, -+ atomic_read(&sb->s_active), !!br->br_xino.xi_file); ++ atomic_read(&sb->s_active), ++ !!au_xino_file(br->br_xino, /*idx*/-1)); + return 0; + +out: @@ -7381,7 +7771,7 @@ diff -urN /usr/share/empty/fs/aufs/debug.c linux/fs/aufs/debug.c + struct au_sbinfo *sbinfo; + aufs_bindex_t bindex; + int err; -+ /* to reuduce stack size */ ++ /* to reduce stack size */ + struct { + struct vfsmount mnt; + struct au_branch fake; @@ -7396,10 +7786,8 @@ diff -urN /usr/share/empty/fs/aufs/debug.c linux/fs/aufs/debug.c + + a->mnt.mnt_sb = sb; + a->fake.br_path.mnt = &a->mnt; -+ au_br_count_init(&a->fake); + err = do_pri_br(-1, &a->fake); -+ au_br_count_fin(&a->fake); -+ au_delayed_kfree(a); ++ au_kfree_rcu(a); + dpri("dev 0x%x\n", sb->s_dev); + if (err || !au_test_aufs(sb)) + return; @@ -7503,10 +7891,11 @@ diff -urN /usr/share/empty/fs/aufs/debug.c linux/fs/aufs/debug.c +} diff -urN /usr/share/empty/fs/aufs/debug.h linux/fs/aufs/debug.h --- /usr/share/empty/fs/aufs/debug.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/debug.h 2017-05-06 22:16:52.778220093 +0200 -@@ -0,0 +1,225 @@ ++++ linux/fs/aufs/debug.h 2019-03-05 12:13:00.135890907 +0100 +@@ -0,0 +1,226 @@ ++/* SPDX-License-Identifier: GPL-2.0 */ +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -7732,10 +8121,11 @@ diff -urN /usr/share/empty/fs/aufs/debug.h linux/fs/aufs/debug.h +#endif /* __AUFS_DEBUG_H__ */ diff -urN /usr/share/empty/fs/aufs/dentry.c linux/fs/aufs/dentry.c --- /usr/share/empty/fs/aufs/dentry.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/dentry.c 2017-05-06 22:16:52.778220093 +0200 -@@ -0,0 +1,1130 @@ ++++ linux/fs/aufs/dentry.c 2019-05-06 09:03:04.816810402 +0200 +@@ -0,0 +1,1154 @@ ++// SPDX-License-Identifier: GPL-2.0 +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -7755,22 +8145,17 @@ diff -urN /usr/share/empty/fs/aufs/dentry.c linux/fs/aufs/dentry.c + * lookup and dentry operations + */ + ++#include +#include +#include "aufs.h" + -+struct au_do_lookup_args { -+ unsigned int flags; -+ mode_t type; -+}; -+ +/* + * returns positive/negative dentry, NULL or an error. + * NULL means whiteout-ed or not-found. + */ +static struct dentry* +au_do_lookup(struct dentry *h_parent, struct dentry *dentry, -+ aufs_bindex_t bindex, struct qstr *wh_name, -+ struct au_do_lookup_args *args) ++ aufs_bindex_t bindex, struct au_do_lookup_args *args) +{ + struct dentry *h_dentry; + struct inode *h_inode; @@ -7785,7 +8170,7 @@ diff -urN /usr/share/empty/fs/aufs/dentry.c linux/fs/aufs/dentry.c + br = au_sbr(dentry->d_sb, bindex); + wh_able = !!au_br_whable(br->br_perm); + if (wh_able) -+ wh_found = au_wh_test(h_parent, wh_name, ignore_perm); ++ wh_found = au_wh_test(h_parent, &args->whname, ignore_perm); + h_dentry = ERR_PTR(wh_found); + if (!wh_found) + goto real_lookup; @@ -7800,9 +8185,9 @@ diff -urN /usr/share/empty/fs/aufs/dentry.c linux/fs/aufs/dentry.c + +real_lookup: + if (!ignore_perm) -+ h_dentry = vfsub_lkup_one(&dentry->d_name, h_parent); ++ h_dentry = vfsub_lkup_one(args->name, h_parent); + else -+ h_dentry = au_sio_lkup_one(&dentry->d_name, h_parent); ++ h_dentry = au_sio_lkup_one(args->name, h_parent); + if (IS_ERR(h_dentry)) { + if (PTR_ERR(h_dentry) == -ENAMETOOLONG + && !allow_neg) @@ -7817,6 +8202,13 @@ diff -urN /usr/share/empty/fs/aufs/dentry.c linux/fs/aufs/dentry.c + } else if (wh_found + || (args->type && args->type != (h_inode->i_mode & S_IFMT))) + goto out_neg; ++ else if (au_ftest_lkup(args->flags, DIRREN) ++ /* && h_inode */ ++ && !au_dr_lkup_h_ino(args, bindex, h_inode->i_ino)) { ++ AuDbg("b%d %pd ignored hi%llu\n", bindex, h_dentry, ++ (unsigned long long)h_inode->i_ino); ++ goto out_neg; ++ } + + if (au_dbbot(dentry) <= bindex) + au_set_dbbot(dentry, bindex); @@ -7829,9 +8221,9 @@ diff -urN /usr/share/empty/fs/aufs/dentry.c linux/fs/aufs/dentry.c + || (d_really_is_positive(dentry) && !d_is_dir(dentry))) + goto out; /* success */ + -+ inode_lock_nested(h_inode, AuLsc_I_CHILD); ++ inode_lock_shared_nested(h_inode, AuLsc_I_CHILD); + opq = au_diropq_test(h_dentry); -+ inode_unlock(h_inode); ++ inode_unlock_shared(h_inode); + if (opq > 0) + au_set_dbdiropq(dentry, bindex); + else if (unlikely(opq < 0)) { @@ -7865,26 +8257,28 @@ diff -urN /usr/share/empty/fs/aufs/dentry.c linux/fs/aufs/dentry.c +{ + int npositive, err; + aufs_bindex_t bindex, btail, bdiropq; -+ unsigned char isdir, dirperm1; -+ struct qstr whname; ++ unsigned char isdir, dirperm1, dirren; + struct au_do_lookup_args args = { -+ .flags = flags ++ .flags = flags, ++ .name = &dentry->d_name + }; -+ const struct qstr *name = &dentry->d_name; + struct dentry *parent; + struct super_block *sb; + + sb = dentry->d_sb; -+ err = au_test_shwh(sb, name); ++ err = au_test_shwh(sb, args.name); + if (unlikely(err)) + goto out; + -+ err = au_wh_name_alloc(&whname, name); ++ err = au_wh_name_alloc(&args.whname, args.name); + if (unlikely(err)) + goto out; + + isdir = !!d_is_dir(dentry); + dirperm1 = !!au_opt_test(au_mntflags(sb), DIRPERM1); ++ dirren = !!au_opt_test(au_mntflags(sb), DIRREN); ++ if (dirren) ++ au_fset_lkup(args.flags, DIRREN); + + npositive = 0; + parent = dget_parent(dentry); @@ -7892,6 +8286,7 @@ diff -urN /usr/share/empty/fs/aufs/dentry.c linux/fs/aufs/dentry.c + for (bindex = btop; bindex <= btail; bindex++) { + struct dentry *h_parent, *h_dentry; + struct inode *h_inode, *h_dir; ++ struct au_branch *br; + + h_dentry = au_h_dptr(dentry, bindex); + if (h_dentry) { @@ -7903,11 +8298,17 @@ diff -urN /usr/share/empty/fs/aufs/dentry.c linux/fs/aufs/dentry.c + if (!h_parent || !d_is_dir(h_parent)) + continue; + ++ if (dirren) { ++ /* if the inum matches, then use the prepared name */ ++ err = au_dr_lkup_name(&args, bindex); ++ if (unlikely(err)) ++ goto out_parent; ++ } ++ + h_dir = d_inode(h_parent); -+ inode_lock_nested(h_dir, AuLsc_I_PARENT); -+ h_dentry = au_do_lookup(h_parent, dentry, bindex, &whname, -+ &args); -+ inode_unlock(h_dir); ++ inode_lock_shared_nested(h_dir, AuLsc_I_PARENT); ++ h_dentry = au_do_lookup(h_parent, dentry, bindex, &args); ++ inode_unlock_shared(h_dir); + err = PTR_ERR(h_dentry); + if (IS_ERR(h_dentry)) + goto out_parent; @@ -7934,6 +8335,15 @@ diff -urN /usr/share/empty/fs/aufs/dentry.c linux/fs/aufs/dentry.c + if (bdiropq >= 0 && bdiropq <= bindex) + break; + } ++ br = au_sbr(sb, bindex); ++ if (dirren ++ && au_dr_hino_test_add(&br->br_dirren, h_inode->i_ino, ++ /*add_ent*/NULL)) { ++ /* prepare next name to lookup */ ++ err = au_dr_lkup(&args, dentry, bindex); ++ if (unlikely(err)) ++ goto out_parent; ++ } + } + + if (npositive) { @@ -7950,7 +8360,9 @@ diff -urN /usr/share/empty/fs/aufs/dentry.c linux/fs/aufs/dentry.c + +out_parent: + dput(parent); -+ au_delayed_kfree(whname.name); ++ au_kfree_try_rcu(args.whname.name); ++ if (dirren) ++ au_dr_lkup_fin(&args); +out: + return err; +} @@ -8037,7 +8449,7 @@ diff -urN /usr/share/empty/fs/aufs/dentry.c linux/fs/aufs/dentry.c + /* ia->i_nlink = h_inode->i_nlink; */ + ia->i_uid = h_inode->i_uid; + ia->i_gid = h_inode->i_gid; -+ ia->i_version = h_inode->i_version; ++ ia->i_version = inode_query_iversion(h_inode); +/* + ia->i_size = h_inode->i_size; + ia->i_blocks = h_inode->i_blocks; @@ -8051,7 +8463,7 @@ diff -urN /usr/share/empty/fs/aufs/dentry.c linux/fs/aufs/dentry.c + /* || ia->i_nlink != h_inode->i_nlink */ + || !uid_eq(ia->i_uid, h_inode->i_uid) + || !gid_eq(ia->i_gid, h_inode->i_gid) -+ || ia->i_version != h_inode->i_version ++ || !inode_eq_iversion(h_inode, ia->i_version) +/* + || ia->i_size != h_inode->i_size + || ia->i_blocks != h_inode->i_blocks @@ -8322,7 +8734,7 @@ diff -urN /usr/share/empty/fs/aufs/dentry.c linux/fs/aufs/dentry.c + if (d_really_is_positive(dentry)) + inode = d_inode(dentry); + if (!orig_h.inode) { -+ AuDbg("nagative originally\n"); ++ AuDbg("negative originally\n"); + if (inode) { + au_hide(dentry); + goto out; @@ -8559,7 +8971,7 @@ diff -urN /usr/share/empty/fs/aufs/dentry.c linux/fs/aufs/dentry.c + +/* todo: remove this */ +static int h_d_revalidate(struct dentry *dentry, struct inode *inode, -+ unsigned int flags, int do_udba) ++ unsigned int flags, int do_udba, int dirren) +{ + int err; + umode_t mode, h_mode; @@ -8610,7 +9022,7 @@ diff -urN /usr/share/empty/fs/aufs/dentry.c linux/fs/aufs/dentry.c + && !is_root + && ((!h_nfs + && (unhashed != !!d_unhashed(h_dentry) -+ || (!tmpfile ++ || (!tmpfile && !dirren + && !au_qstreq(name, h_name)) + )) + || (h_nfs @@ -8751,7 +9163,7 @@ diff -urN /usr/share/empty/fs/aufs/dentry.c linux/fs/aufs/dentry.c +{ + int valid, err; + unsigned int sigen; -+ unsigned char do_udba; ++ unsigned char do_udba, dirren; + struct super_block *sb; + struct inode *inode; + @@ -8824,7 +9236,8 @@ diff -urN /usr/share/empty/fs/aufs/dentry.c linux/fs/aufs/dentry.c + } + } + -+ err = h_d_revalidate(dentry, inode, flags, do_udba); ++ dirren = !!au_opt_test(au_mntflags(sb), DIRREN); ++ err = h_d_revalidate(dentry, inode, flags, do_udba, dirren); + if (unlikely(!err && do_udba && au_dbtop(dentry) < 0)) { + err = -EIO; + AuDbg("both of real entry and whiteout found, %p, err %d\n", @@ -8866,10 +9279,11 @@ diff -urN /usr/share/empty/fs/aufs/dentry.c linux/fs/aufs/dentry.c +}; diff -urN /usr/share/empty/fs/aufs/dentry.h linux/fs/aufs/dentry.h --- /usr/share/empty/fs/aufs/dentry.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/dentry.h 2017-05-06 22:16:52.778220093 +0200 -@@ -0,0 +1,255 @@ ++++ linux/fs/aufs/dentry.h 2019-03-05 12:13:00.135890907 +0100 +@@ -0,0 +1,268 @@ ++/* SPDX-License-Identifier: GPL-2.0 */ +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -8895,6 +9309,7 @@ diff -urN /usr/share/empty/fs/aufs/dentry.h linux/fs/aufs/dentry.h +#ifdef __KERNEL__ + +#include ++#include "dirren.h" +#include "rwsem.h" + +struct au_hdentry { @@ -8908,10 +9323,8 @@ diff -urN /usr/share/empty/fs/aufs/dentry.h linux/fs/aufs/dentry.h + struct au_rwsem di_rwsem; + aufs_bindex_t di_btop, di_bbot, di_bwh, di_bdiropq; + unsigned char di_tmpfile; /* to allow the different name */ -+ union { -+ struct au_hdentry *di_hdentry; -+ struct llist_node di_lnode; /* delayed free */ -+ }; ++ struct au_hdentry *di_hdentry; ++ struct rcu_head rcu; +} ____cacheline_aligned_in_smp; + +/* ---------------------------------------------------------------------- */ @@ -8919,12 +9332,25 @@ diff -urN /usr/share/empty/fs/aufs/dentry.h linux/fs/aufs/dentry.h +/* flags for au_lkup_dentry() */ +#define AuLkup_ALLOW_NEG 1 +#define AuLkup_IGNORE_PERM (1 << 1) ++#define AuLkup_DIRREN (1 << 2) +#define au_ftest_lkup(flags, name) ((flags) & AuLkup_##name) +#define au_fset_lkup(flags, name) \ + do { (flags) |= AuLkup_##name; } while (0) +#define au_fclr_lkup(flags, name) \ + do { (flags) &= ~AuLkup_##name; } while (0) + ++#ifndef CONFIG_AUFS_DIRREN ++#undef AuLkup_DIRREN ++#define AuLkup_DIRREN 0 ++#endif ++ ++struct au_do_lookup_args { ++ unsigned int flags; ++ mode_t type; ++ struct qstr whname, *name; ++ struct au_dr_lookup dirren; ++}; ++ +/* ---------------------------------------------------------------------- */ + +/* dentry.c */ @@ -9125,10 +9551,11 @@ diff -urN /usr/share/empty/fs/aufs/dentry.h linux/fs/aufs/dentry.h +#endif /* __AUFS_DENTRY_H__ */ diff -urN /usr/share/empty/fs/aufs/dinfo.c linux/fs/aufs/dinfo.c --- /usr/share/empty/fs/aufs/dinfo.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/dinfo.c 2017-05-06 22:16:52.778220093 +0200 -@@ -0,0 +1,553 @@ ++++ linux/fs/aufs/dinfo.c 2019-03-05 12:13:00.135890907 +0100 +@@ -0,0 +1,554 @@ ++// SPDX-License-Identifier: GPL-2.0 +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -9182,7 +9609,7 @@ diff -urN /usr/share/empty/fs/aufs/dinfo.c linux/fs/aufs/dinfo.c + goto out; + } + -+ au_cache_dfree_dinfo(dinfo); ++ au_cache_free_dinfo(dinfo); + dinfo = NULL; + +out: @@ -9202,8 +9629,8 @@ diff -urN /usr/share/empty/fs/aufs/dinfo.c linux/fs/aufs/dinfo.c + while (bindex++ <= bbot) + au_hdput(p++); + } -+ au_delayed_kfree(dinfo->di_hdentry); -+ au_cache_dfree_dinfo(dinfo); ++ au_kfree_try_rcu(dinfo->di_hdentry); ++ au_cache_free_dinfo(dinfo); +} + +void au_di_swap(struct au_dinfo *a, struct au_dinfo *b) @@ -9682,10 +10109,11 @@ diff -urN /usr/share/empty/fs/aufs/dinfo.c linux/fs/aufs/dinfo.c +} diff -urN /usr/share/empty/fs/aufs/dir.c linux/fs/aufs/dir.c --- /usr/share/empty/fs/aufs/dir.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/dir.c 2017-05-06 22:16:52.778220093 +0200 -@@ -0,0 +1,762 @@ ++++ linux/fs/aufs/dir.c 2019-05-06 09:03:04.816810402 +0200 +@@ -0,0 +1,763 @@ ++// SPDX-License-Identifier: GPL-2.0 +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -9706,6 +10134,7 @@ diff -urN /usr/share/empty/fs/aufs/dir.c linux/fs/aufs/dir.c + */ + +#include ++#include +#include "aufs.h" + +void au_add_nlink(struct inode *dir, struct inode *h_dir) @@ -9829,7 +10258,7 @@ diff -urN /usr/share/empty/fs/aufs/dir.c linux/fs/aufs/dir.c + au_hn_inode_lock_nested(hdir, AuLsc_I_PARENT); + h_dir = au_h_iptr(dir, btop); + if (h_dir->i_nlink -+ && timespec_compare(&h_dir->i_mtime, &dt.dt_mtime) < 0) { ++ && timespec64_compare(&h_dir->i_mtime, &dt.dt_mtime) < 0) { + dt.dt_h_path = h_path; + au_dtime_revert(&dt); + } @@ -9842,7 +10271,7 @@ diff -urN /usr/share/empty/fs/aufs/dir.c linux/fs/aufs/dir.c +out: + dput(a->dentry); + au_nwt_done(&au_sbi(sb)->si_nowait); -+ au_delayed_kfree(arg); ++ au_kfree_try_rcu(arg); +} + +void au_dir_ts(struct inode *dir, aufs_bindex_t bindex) @@ -9878,7 +10307,7 @@ diff -urN /usr/share/empty/fs/aufs/dir.c linux/fs/aufs/dir.c + if (unlikely(wkq_err)) { + pr_err("wkq %d\n", wkq_err); + dput(dentry); -+ au_delayed_kfree(arg); ++ au_kfree_try_rcu(arg); + } + +out: @@ -9944,7 +10373,7 @@ diff -urN /usr/share/empty/fs/aufs/dir.c linux/fs/aufs/dir.c + err = 0; + mnt = file->f_path.mnt; + dentry = file->f_path.dentry; -+ file->f_version = d_inode(dentry)->i_version; ++ file->f_version = inode_query_iversion(d_inode(dentry)); + bindex = au_dbtop(dentry); + au_set_fbtop(file, bindex); + btail = au_dbtaildir(dentry); @@ -9997,7 +10426,7 @@ diff -urN /usr/share/empty/fs/aufs/dir.c linux/fs/aufs/dir.c + }; + err = au_do_open(file, &args); + if (unlikely(err)) -+ au_delayed_kfree(fidir); ++ au_kfree_rcu(fidir); + } + si_read_unlock(sb); + return err; @@ -10011,21 +10440,18 @@ diff -urN /usr/share/empty/fs/aufs/dir.c linux/fs/aufs/dir.c + struct au_fidir *fidir; + struct au_hfile *hf; + aufs_bindex_t bindex, bbot; -+ int execed, delayed; + -+ delayed = (current->flags & PF_KTHREAD) || in_interrupt(); + finfo = au_fi(file); + fidir = finfo->fi_hdir; + if (fidir) { -+ au_sphl_del(&finfo->fi_hlist, -+ &au_sbi(file->f_path.dentry->d_sb)->si_files); ++ au_hbl_del(&finfo->fi_hlist, ++ &au_sbi(file->f_path.dentry->d_sb)->si_files); + vdir_cache = fidir->fd_vdir_cache; /* lock-free */ + if (vdir_cache) -+ au_vdir_free(vdir_cache, delayed); ++ au_vdir_free(vdir_cache); + + bindex = finfo->fi_btop; + if (bindex >= 0) { -+ execed = vfsub_file_execed(file); + hf = fidir->fd_hfile + bindex; + /* + * calls fput() instead of filp_close(), @@ -10034,12 +10460,12 @@ diff -urN /usr/share/empty/fs/aufs/dir.c linux/fs/aufs/dir.c + bbot = fidir->fd_bbot; + for (; bindex <= bbot; bindex++, hf++) + if (hf->hf_file) -+ au_hfput(hf, execed); ++ au_hfput(hf, /*execed*/0); + } -+ au_delayed_kfree(fidir); ++ au_kfree_rcu(fidir); + finfo->fi_hdir = NULL; + } -+ au_finfo_fin(file, delayed); ++ au_finfo_fin(file); + return 0; +} + @@ -10165,7 +10591,7 @@ diff -urN /usr/share/empty/fs/aufs/dir.c linux/fs/aufs/dir.c + struct inode *inode, *h_inode; + struct super_block *sb; + -+ AuDbg("%pD, ctx{%pf, %llu}\n", file, ctx->actor, ctx->pos); ++ AuDbg("%pD, ctx{%ps, %llu}\n", file, ctx->actor, ctx->pos); + + dentry = file->f_path.dentry; + inode = d_inode(dentry); @@ -10276,6 +10702,7 @@ diff -urN /usr/share/empty/fs/aufs/dir.c linux/fs/aufs/dir.c +{ + int err; + struct file *h_file; ++ struct au_branch *br; + + h_file = au_h_open(dentry, arg->bindex, + O_RDONLY | O_NONBLOCK | O_DIRECTORY | O_LARGEFILE, @@ -10300,7 +10727,8 @@ diff -urN /usr/share/empty/fs/aufs/dir.c linux/fs/aufs/dir.c + +out_put: + fput(h_file); -+ au_sbr_put(dentry->d_sb, arg->bindex); ++ br = au_sbr(dentry->d_sb, arg->bindex); ++ au_lcnt_dec(&br->br_nfiles); +out: + return err; +} @@ -10326,9 +10754,9 @@ diff -urN /usr/share/empty/fs/aufs/dir.c linux/fs/aufs/dir.c + h_dentry = au_h_dptr(dentry, arg->bindex); + h_inode = d_inode(h_dentry); + /* todo: i_mode changes anytime? */ -+ inode_lock_nested(h_inode, AuLsc_I_CHILD); ++ inode_lock_shared_nested(h_inode, AuLsc_I_CHILD); + err = au_test_h_perm_sio(h_inode, MAY_EXEC | MAY_READ); -+ inode_unlock(h_inode); ++ inode_unlock_shared(h_inode); + if (!err) + err = do_test_empty(dentry, arg); + else { @@ -10448,10 +10876,11 @@ diff -urN /usr/share/empty/fs/aufs/dir.c linux/fs/aufs/dir.c +}; diff -urN /usr/share/empty/fs/aufs/dir.h linux/fs/aufs/dir.h --- /usr/share/empty/fs/aufs/dir.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/dir.h 2017-05-06 22:16:52.778220093 +0200 -@@ -0,0 +1,137 @@ ++++ linux/fs/aufs/dir.h 2019-03-05 12:13:00.139224339 +0100 +@@ -0,0 +1,134 @@ ++/* SPDX-License-Identifier: GPL-2.0 */ +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -10494,10 +10923,8 @@ diff -urN /usr/share/empty/fs/aufs/dir.h linux/fs/aufs/dir.h + +struct au_vdir_dehstr { + struct hlist_node hash; -+ union { -+ struct au_vdir_destr *str; -+ struct llist_node lnode; /* delayed free */ -+ }; ++ struct au_vdir_destr *str; ++ struct rcu_head rcu; +} ____cacheline_aligned_in_smp; + +struct au_vdir_de { @@ -10533,12 +10960,10 @@ diff -urN /usr/share/empty/fs/aufs/dir.h linux/fs/aufs/dir.h + union au_vdir_deblk_p p; + } vd_last; + -+ unsigned long vd_version; ++ u64 vd_version; + unsigned int vd_deblk_sz; -+ union { -+ unsigned long vd_jiffy; -+ struct llist_node vd_lnode; /* delayed free */ -+ }; ++ unsigned long vd_jiffy; ++ struct rcu_head rcu; +} ____cacheline_aligned_in_smp; + +/* ---------------------------------------------------------------------- */ @@ -10562,7 +10987,7 @@ diff -urN /usr/share/empty/fs/aufs/dir.h linux/fs/aufs/dir.h +int au_nhash_append_wh(struct au_nhash *whlist, char *name, int nlen, ino_t ino, + unsigned int d_type, aufs_bindex_t bindex, + unsigned char shwh); -+void au_vdir_free(struct au_vdir *vdir, int atonce); ++void au_vdir_free(struct au_vdir *vdir); +int au_vdir_init(struct file *file); +int au_vdir_fill_de(struct file *file, struct dir_context *ctx); + @@ -10587,12 +11012,13 @@ diff -urN /usr/share/empty/fs/aufs/dir.h linux/fs/aufs/dir.h + +#endif /* __KERNEL__ */ +#endif /* __AUFS_DIR_H__ */ -diff -urN /usr/share/empty/fs/aufs/dynop.c linux/fs/aufs/dynop.c ---- /usr/share/empty/fs/aufs/dynop.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/dynop.c 2017-05-06 22:16:52.778220093 +0200 -@@ -0,0 +1,371 @@ +diff -urN /usr/share/empty/fs/aufs/dirren.c linux/fs/aufs/dirren.c +--- /usr/share/empty/fs/aufs/dirren.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux/fs/aufs/dirren.c 2019-03-05 12:13:00.139224339 +0100 +@@ -0,0 +1,1316 @@ ++// SPDX-License-Identifier: GPL-2.0 +/* -+ * Copyright (C) 2010-2017 Junjiro R. Okajima ++ * Copyright (C) 2017-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -10609,392 +11035,1855 @@ diff -urN /usr/share/empty/fs/aufs/dynop.c linux/fs/aufs/dynop.c + */ + +/* -+ * dynamically customizable operations for regular files ++ * special handling in renaming a directory ++ * in order to support looking-up the before-renamed name on the lower readonly ++ * branches + */ + ++#include +#include "aufs.h" + -+#define DyPrSym(key) AuDbgSym(key->dk_op.dy_hop) ++static void au_dr_hino_del(struct au_dr_br *dr, struct au_dr_hino *ent) ++{ ++ int idx; + -+/* -+ * How large will these lists be? -+ * Usually just a few elements, 20-30 at most for each, I guess. -+ */ -+static struct au_sphlhead dynop[AuDyLast]; ++ idx = au_dr_ihash(ent->dr_h_ino); ++ au_hbl_del(&ent->dr_hnode, dr->dr_h_ino + idx); ++} + -+static struct au_dykey *dy_gfind_get(struct au_sphlhead *sphl, const void *h_op) ++static int au_dr_hino_test_empty(struct au_dr_br *dr) +{ -+ struct au_dykey *key, *tmp; -+ struct hlist_head *head; ++ int ret, i; ++ struct hlist_bl_head *hbl; + -+ key = NULL; -+ head = &sphl->head; -+ rcu_read_lock(); -+ hlist_for_each_entry_rcu(tmp, head, dk_hnode) -+ if (tmp->dk_op.dy_hop == h_op) { -+ key = tmp; -+ kref_get(&key->dk_kref); -+ break; -+ } -+ rcu_read_unlock(); ++ ret = 1; ++ for (i = 0; ret && i < AuDirren_NHASH; i++) { ++ hbl = dr->dr_h_ino + i; ++ hlist_bl_lock(hbl); ++ ret &= hlist_bl_empty(hbl); ++ hlist_bl_unlock(hbl); ++ } + -+ return key; ++ return ret; +} + -+static struct au_dykey *dy_bradd(struct au_branch *br, struct au_dykey *key) ++static struct au_dr_hino *au_dr_hino_find(struct au_dr_br *dr, ino_t ino) +{ -+ struct au_dykey **k, *found; -+ const void *h_op = key->dk_op.dy_hop; -+ int i; ++ struct au_dr_hino *found, *ent; ++ struct hlist_bl_head *hbl; ++ struct hlist_bl_node *pos; ++ int idx; + + found = NULL; -+ k = br->br_dykey; -+ for (i = 0; i < AuBrDynOp; i++) -+ if (k[i]) { -+ if (k[i]->dk_op.dy_hop == h_op) { -+ found = k[i]; -+ break; -+ } -+ } else ++ idx = au_dr_ihash(ino); ++ hbl = dr->dr_h_ino + idx; ++ hlist_bl_lock(hbl); ++ hlist_bl_for_each_entry(ent, pos, hbl, dr_hnode) ++ if (ent->dr_h_ino == ino) { ++ found = ent; + break; -+ if (!found) { -+ spin_lock(&br->br_dykey_lock); -+ for (; i < AuBrDynOp; i++) -+ if (k[i]) { -+ if (k[i]->dk_op.dy_hop == h_op) { -+ found = k[i]; -+ break; -+ } -+ } else { -+ k[i] = key; -+ break; -+ } -+ spin_unlock(&br->br_dykey_lock); -+ BUG_ON(i == AuBrDynOp); /* expand the array */ -+ } ++ } ++ hlist_bl_unlock(hbl); + + return found; +} + -+/* kref_get() if @key is already added */ -+static struct au_dykey *dy_gadd(struct au_sphlhead *sphl, struct au_dykey *key) ++int au_dr_hino_test_add(struct au_dr_br *dr, ino_t ino, ++ struct au_dr_hino *add_ent) +{ -+ struct au_dykey *tmp, *found; -+ struct hlist_head *head; -+ const void *h_op = key->dk_op.dy_hop; ++ int found, idx; ++ struct hlist_bl_head *hbl; ++ struct hlist_bl_node *pos; ++ struct au_dr_hino *ent; + -+ found = NULL; -+ head = &sphl->head; -+ spin_lock(&sphl->spin); -+ hlist_for_each_entry(tmp, head, dk_hnode) -+ if (tmp->dk_op.dy_hop == h_op) { -+ kref_get(&tmp->dk_kref); -+ found = tmp; ++ found = 0; ++ idx = au_dr_ihash(ino); ++ hbl = dr->dr_h_ino + idx; ++#if 0 ++ { ++ struct hlist_bl_node *tmp; ++ ++ hlist_bl_for_each_entry_safe(ent, pos, tmp, hbl, dr_hnode) ++ AuDbg("hi%llu\n", (unsigned long long)ent->dr_h_ino); ++ } ++#endif ++ hlist_bl_lock(hbl); ++ hlist_bl_for_each_entry(ent, pos, hbl, dr_hnode) ++ if (ent->dr_h_ino == ino) { ++ found = 1; + break; + } -+ if (!found) -+ hlist_add_head_rcu(&key->dk_hnode, head); -+ spin_unlock(&sphl->spin); ++ if (!found && add_ent) ++ hlist_bl_add_head(&add_ent->dr_hnode, hbl); ++ hlist_bl_unlock(hbl); ++ ++ if (!found && add_ent) ++ AuDbg("i%llu added\n", (unsigned long long)add_ent->dr_h_ino); + -+ if (!found) -+ DyPrSym(key); + return found; +} + -+static void dy_free_rcu(struct rcu_head *rcu) ++void au_dr_hino_free(struct au_dr_br *dr) +{ -+ struct au_dykey *key; -+ -+ key = container_of(rcu, struct au_dykey, dk_rcu); -+ DyPrSym(key); -+ kfree(key); /* not delayed */ -+} ++ int i; ++ struct hlist_bl_head *hbl; ++ struct hlist_bl_node *pos, *tmp; ++ struct au_dr_hino *ent; + -+static void dy_free(struct kref *kref) -+{ -+ struct au_dykey *key; -+ struct au_sphlhead *sphl; ++ /* SiMustWriteLock(sb); */ + -+ key = container_of(kref, struct au_dykey, dk_kref); -+ sphl = dynop + key->dk_op.dy_type; -+ au_sphl_del_rcu(&key->dk_hnode, sphl); -+ call_rcu(&key->dk_rcu, dy_free_rcu); ++ for (i = 0; i < AuDirren_NHASH; i++) { ++ hbl = dr->dr_h_ino + i; ++ /* no spinlock since sbinfo must be write-locked */ ++ hlist_bl_for_each_entry_safe(ent, pos, tmp, hbl, dr_hnode) ++ au_kfree_rcu(ent); ++ INIT_HLIST_BL_HEAD(hbl); ++ } +} + -+void au_dy_put(struct au_dykey *key) ++/* returns the number of inodes or an error */ ++static int au_dr_hino_store(struct super_block *sb, struct au_branch *br, ++ struct file *hinofile) +{ -+ kref_put(&key->dk_kref, dy_free); -+} -+ -+/* ---------------------------------------------------------------------- */ ++ int err, i; ++ ssize_t ssz; ++ loff_t pos, oldsize; ++ __be64 u64; ++ struct inode *hinoinode; ++ struct hlist_bl_head *hbl; ++ struct hlist_bl_node *n1, *n2; ++ struct au_dr_hino *ent; + -+#define DyDbgSize(cnt, op) AuDebugOn(cnt != sizeof(op)/sizeof(void *)) ++ SiMustWriteLock(sb); ++ AuDebugOn(!au_br_writable(br->br_perm)); + -+#ifdef CONFIG_AUFS_DEBUG -+#define DyDbgDeclare(cnt) unsigned int cnt = 0 -+#define DyDbgInc(cnt) do { cnt++; } while (0) -+#else -+#define DyDbgDeclare(cnt) do {} while (0) -+#define DyDbgInc(cnt) do {} while (0) -+#endif ++ hinoinode = file_inode(hinofile); ++ oldsize = i_size_read(hinoinode); + -+#define DySet(func, dst, src, h_op, h_sb) do { \ -+ DyDbgInc(cnt); \ -+ if (h_op->func) { \ -+ if (src.func) \ -+ dst.func = src.func; \ -+ else \ -+ AuDbg("%s %s\n", au_sbtype(h_sb), #func); \ -+ } \ -+} while (0) ++ err = 0; ++ pos = 0; ++ hbl = br->br_dirren.dr_h_ino; ++ for (i = 0; !err && i < AuDirren_NHASH; i++, hbl++) { ++ /* no bit-lock since sbinfo must be write-locked */ ++ hlist_bl_for_each_entry_safe(ent, n1, n2, hbl, dr_hnode) { ++ AuDbg("hi%llu, %pD2\n", ++ (unsigned long long)ent->dr_h_ino, hinofile); ++ u64 = cpu_to_be64(ent->dr_h_ino); ++ ssz = vfsub_write_k(hinofile, &u64, sizeof(u64), &pos); ++ if (ssz == sizeof(u64)) ++ continue; + -+#define DySetForce(func, dst, src) do { \ -+ AuDebugOn(!src.func); \ -+ DyDbgInc(cnt); \ -+ dst.func = src.func; \ -+} while (0) ++ /* write error */ ++ pr_err("ssz %zd, %pD2\n", ssz, hinofile); ++ err = -ENOSPC; ++ if (ssz < 0) ++ err = ssz; ++ break; ++ } ++ } ++ /* regardless the error */ ++ if (pos < oldsize) { ++ err = vfsub_trunc(&hinofile->f_path, pos, /*attr*/0, hinofile); ++ AuTraceErr(err); ++ } + -+#define DySetAop(func) \ -+ DySet(func, dyaop->da_op, aufs_aop, h_aop, h_sb) -+#define DySetAopForce(func) \ -+ DySetForce(func, dyaop->da_op, aufs_aop) ++ AuTraceErr(err); ++ return err; ++} + -+static void dy_aop(struct au_dykey *key, const void *h_op, -+ struct super_block *h_sb __maybe_unused) ++static int au_dr_hino_load(struct au_dr_br *dr, struct file *hinofile) +{ -+ struct au_dyaop *dyaop = (void *)key; -+ const struct address_space_operations *h_aop = h_op; -+ DyDbgDeclare(cnt); ++ int err, hidx; ++ ssize_t ssz; ++ size_t sz, n; ++ loff_t pos; ++ uint64_t u64; ++ struct au_dr_hino *ent; ++ struct inode *hinoinode; ++ struct hlist_bl_head *hbl; + -+ AuDbg("%s\n", au_sbtype(h_sb)); ++ err = 0; ++ pos = 0; ++ hbl = dr->dr_h_ino; ++ hinoinode = file_inode(hinofile); ++ sz = i_size_read(hinoinode); ++ AuDebugOn(sz % sizeof(u64)); ++ n = sz / sizeof(u64); ++ while (n--) { ++ ssz = vfsub_read_k(hinofile, &u64, sizeof(u64), &pos); ++ if (unlikely(ssz != sizeof(u64))) { ++ pr_err("ssz %zd, %pD2\n", ssz, hinofile); ++ err = -EINVAL; ++ if (ssz < 0) ++ err = ssz; ++ goto out_free; ++ } + -+ DySetAop(writepage); -+ DySetAopForce(readpage); /* force */ -+ DySetAop(writepages); -+ DySetAop(set_page_dirty); -+ DySetAop(readpages); -+ DySetAop(write_begin); -+ DySetAop(write_end); -+ DySetAop(bmap); -+ DySetAop(invalidatepage); -+ DySetAop(releasepage); -+ DySetAop(freepage); -+ /* this one will be changed according to an aufs mount option */ -+ DySetAop(direct_IO); -+ DySetAop(migratepage); -+ DySetAop(isolate_page); -+ DySetAop(putback_page); -+ DySetAop(launder_page); -+ DySetAop(is_partially_uptodate); -+ DySetAop(is_dirty_writeback); -+ DySetAop(error_remove_page); -+ DySetAop(swap_activate); -+ DySetAop(swap_deactivate); ++ ent = kmalloc(sizeof(*ent), GFP_NOFS); ++ if (!ent) { ++ err = -ENOMEM; ++ AuTraceErr(err); ++ goto out_free; ++ } ++ ent->dr_h_ino = be64_to_cpu((__force __be64)u64); ++ AuDbg("hi%llu, %pD2\n", ++ (unsigned long long)ent->dr_h_ino, hinofile); ++ hidx = au_dr_ihash(ent->dr_h_ino); ++ au_hbl_add(&ent->dr_hnode, hbl + hidx); ++ } ++ goto out; /* success */ + -+ DyDbgSize(cnt, *h_aop); ++out_free: ++ au_dr_hino_free(dr); ++out: ++ AuTraceErr(err); ++ return err; +} + -+/* ---------------------------------------------------------------------- */ -+ -+static void dy_bug(struct kref *kref) ++/* ++ * @bindex/@br is a switch to distinguish whether suspending hnotify or not. ++ * @path is a switch to distinguish load and store. ++ */ ++static int au_dr_hino(struct super_block *sb, aufs_bindex_t bindex, ++ struct au_branch *br, const struct path *path) +{ -+ BUG(); -+} ++ int err, flags; ++ unsigned char load, suspend; ++ struct file *hinofile; ++ struct au_hinode *hdir; ++ struct inode *dir, *delegated; ++ struct path hinopath; ++ struct qstr hinoname = QSTR_INIT(AUFS_WH_DR_BRHINO, ++ sizeof(AUFS_WH_DR_BRHINO) - 1); + -+static struct au_dykey *dy_get(struct au_dynop *op, struct au_branch *br) -+{ -+ struct au_dykey *key, *old; -+ struct au_sphlhead *sphl; -+ struct op { -+ unsigned int sz; -+ void (*set)(struct au_dykey *key, const void *h_op, -+ struct super_block *h_sb __maybe_unused); -+ }; -+ static const struct op a[] = { -+ [AuDy_AOP] = { -+ .sz = sizeof(struct au_dyaop), -+ .set = dy_aop -+ } -+ }; -+ const struct op *p; ++ AuDebugOn(bindex < 0 && !br); ++ AuDebugOn(bindex >= 0 && br); + -+ sphl = dynop + op->dy_type; -+ key = dy_gfind_get(sphl, op->dy_hop); -+ if (key) -+ goto out_add; /* success */ ++ err = -EINVAL; ++ suspend = !br; ++ if (suspend) ++ br = au_sbr(sb, bindex); ++ load = !!path; ++ if (!load) { ++ path = &br->br_path; ++ AuDebugOn(!au_br_writable(br->br_perm)); ++ if (unlikely(!au_br_writable(br->br_perm))) ++ goto out; ++ } + -+ p = a + op->dy_type; -+ key = kzalloc(p->sz, GFP_NOFS); -+ if (unlikely(!key)) { -+ key = ERR_PTR(-ENOMEM); -+ goto out; ++ hdir = NULL; ++ if (suspend) { ++ dir = d_inode(sb->s_root); ++ hdir = au_hinode(au_ii(dir), bindex); ++ dir = hdir->hi_inode; ++ au_hn_inode_lock_nested(hdir, AuLsc_I_CHILD); ++ } else { ++ dir = d_inode(path->dentry); ++ inode_lock_nested(dir, AuLsc_I_CHILD); + } ++ hinopath.dentry = vfsub_lkup_one(&hinoname, path->dentry); ++ err = PTR_ERR(hinopath.dentry); ++ if (IS_ERR(hinopath.dentry)) ++ goto out_unlock; + -+ key->dk_op.dy_hop = op->dy_hop; -+ kref_init(&key->dk_kref); -+ p->set(key, op->dy_hop, au_br_sb(br)); -+ old = dy_gadd(sphl, key); -+ if (old) { -+ au_delayed_kfree(key); -+ key = old; ++ err = 0; ++ flags = O_RDONLY; ++ if (load) { ++ if (d_is_negative(hinopath.dentry)) ++ goto out_dput; /* success */ ++ } else { ++ if (au_dr_hino_test_empty(&br->br_dirren)) { ++ if (d_is_positive(hinopath.dentry)) { ++ delegated = NULL; ++ err = vfsub_unlink(dir, &hinopath, &delegated, ++ /*force*/0); ++ AuTraceErr(err); ++ if (unlikely(err)) ++ pr_err("ignored err %d, %pd2\n", ++ err, hinopath.dentry); ++ if (unlikely(err == -EWOULDBLOCK)) ++ iput(delegated); ++ err = 0; ++ } ++ goto out_dput; ++ } else if (!d_is_positive(hinopath.dentry)) { ++ err = vfsub_create(dir, &hinopath, 0600, ++ /*want_excl*/false); ++ AuTraceErr(err); ++ if (unlikely(err)) ++ goto out_dput; ++ } ++ flags = O_WRONLY; ++ } ++ hinopath.mnt = path->mnt; ++ hinofile = vfsub_dentry_open(&hinopath, flags); ++ if (suspend) ++ au_hn_inode_unlock(hdir); ++ else ++ inode_unlock(dir); ++ dput(hinopath.dentry); ++ AuTraceErrPtr(hinofile); ++ if (IS_ERR(hinofile)) { ++ err = PTR_ERR(hinofile); ++ goto out; + } + -+out_add: -+ old = dy_bradd(br, key); -+ if (old) -+ /* its ref-count should never be zero here */ -+ kref_put(&key->dk_kref, dy_bug); ++ if (load) ++ err = au_dr_hino_load(&br->br_dirren, hinofile); ++ else ++ err = au_dr_hino_store(sb, br, hinofile); ++ fput(hinofile); ++ goto out; ++ ++out_dput: ++ dput(hinopath.dentry); ++out_unlock: ++ if (suspend) ++ au_hn_inode_unlock(hdir); ++ else ++ inode_unlock(dir); +out: -+ return key; ++ AuTraceErr(err); ++ return err; +} + +/* ---------------------------------------------------------------------- */ -+/* -+ * Aufs prohibits O_DIRECT by defaut even if the branch supports it. -+ * This behaviour is necessary to return an error from open(O_DIRECT) instead -+ * of the succeeding I/O. The dio mount option enables O_DIRECT and makes -+ * open(O_DIRECT) always succeed, but the succeeding I/O may return an error. -+ * See the aufs manual in detail. -+ */ -+static void dy_adx(struct au_dyaop *dyaop, int do_dx) -+{ -+ if (!do_dx) -+ dyaop->da_op.direct_IO = NULL; -+ else -+ dyaop->da_op.direct_IO = aufs_aop.direct_IO; -+} + -+static struct au_dyaop *dy_aget(struct au_branch *br, -+ const struct address_space_operations *h_aop, -+ int do_dx) ++static int au_dr_brid_init(struct au_dr_brid *brid, const struct path *path) +{ -+ struct au_dyaop *dyaop; -+ struct au_dynop op; ++ int err; ++ struct kstatfs kstfs; ++ dev_t dev; ++ struct dentry *dentry; ++ struct super_block *sb; + -+ op.dy_type = AuDy_AOP; -+ op.dy_haop = h_aop; -+ dyaop = (void *)dy_get(&op, br); -+ if (IS_ERR(dyaop)) ++ err = vfs_statfs((void *)path, &kstfs); ++ AuTraceErr(err); ++ if (unlikely(err)) + goto out; -+ dy_adx(dyaop, do_dx); ++ ++ /* todo: support for UUID */ ++ ++ if (kstfs.f_fsid.val[0] || kstfs.f_fsid.val[1]) { ++ brid->type = AuBrid_FSID; ++ brid->fsid = kstfs.f_fsid; ++ } else { ++ dentry = path->dentry; ++ sb = dentry->d_sb; ++ dev = sb->s_dev; ++ if (dev) { ++ brid->type = AuBrid_DEV; ++ brid->dev = dev; ++ } ++ } + +out: -+ return dyaop; ++ return err; +} + -+int au_dy_iaop(struct inode *inode, aufs_bindex_t bindex, -+ struct inode *h_inode) ++int au_dr_br_init(struct super_block *sb, struct au_branch *br, ++ const struct path *path) +{ -+ int err, do_dx; -+ struct super_block *sb; -+ struct au_branch *br; -+ struct au_dyaop *dyaop; ++ int err, i; ++ struct au_dr_br *dr; ++ struct hlist_bl_head *hbl; + -+ AuDebugOn(!S_ISREG(h_inode->i_mode)); -+ IiMustWriteLock(inode); ++ dr = &br->br_dirren; ++ hbl = dr->dr_h_ino; ++ for (i = 0; i < AuDirren_NHASH; i++, hbl++) ++ INIT_HLIST_BL_HEAD(hbl); + -+ sb = inode->i_sb; -+ br = au_sbr(sb, bindex); -+ do_dx = !!au_opt_test(au_mntflags(sb), DIO); -+ dyaop = dy_aget(br, h_inode->i_mapping->a_ops, do_dx); -+ err = PTR_ERR(dyaop); -+ if (IS_ERR(dyaop)) -+ /* unnecessary to call dy_fput() */ ++ err = au_dr_brid_init(&dr->dr_brid, path); ++ if (unlikely(err)) + goto out; + -+ err = 0; -+ inode->i_mapping->a_ops = &dyaop->da_op; ++ if (au_opt_test(au_mntflags(sb), DIRREN)) ++ err = au_dr_hino(sb, /*bindex*/-1, br, path); + +out: ++ AuTraceErr(err); + return err; +} + -+/* -+ * Is it safe to replace a_ops during the inode/file is in operation? -+ * Yes, I hope so. -+ */ -+int au_dy_irefresh(struct inode *inode) ++int au_dr_br_fin(struct super_block *sb, struct au_branch *br) +{ + int err; -+ aufs_bindex_t btop; -+ struct inode *h_inode; + + err = 0; -+ if (S_ISREG(inode->i_mode)) { -+ btop = au_ibtop(inode); -+ h_inode = au_h_iptr(inode, btop); -+ err = au_dy_iaop(inode, btop, h_inode); -+ } -+ return err; -+} -+ -+void au_dy_arefresh(int do_dx) -+{ -+ struct au_sphlhead *sphl; -+ struct hlist_head *head; -+ struct au_dykey *key; ++ if (au_br_writable(br->br_perm)) ++ err = au_dr_hino(sb, /*bindex*/-1, br, /*path*/NULL); ++ if (!err) ++ au_dr_hino_free(&br->br_dirren); + -+ sphl = dynop + AuDy_AOP; -+ head = &sphl->head; -+ spin_lock(&sphl->spin); -+ hlist_for_each_entry(key, head, dk_hnode) -+ dy_adx((void *)key, do_dx); -+ spin_unlock(&sphl->spin); ++ return err; +} + +/* ---------------------------------------------------------------------- */ + -+void __init au_dy_init(void) ++static int au_brid_str(struct au_dr_brid *brid, struct inode *h_inode, ++ char *buf, size_t sz) +{ -+ int i; ++ int err; ++ unsigned int major, minor; ++ char *p; + -+ /* make sure that 'struct au_dykey *' can be any type */ -+ BUILD_BUG_ON(offsetof(struct au_dyaop, da_key)); ++ p = buf; ++ err = snprintf(p, sz, "%d_", brid->type); ++ AuDebugOn(err > sz); ++ p += err; ++ sz -= err; ++ switch (brid->type) { ++ case AuBrid_Unset: ++ return -EINVAL; ++ case AuBrid_UUID: ++ err = snprintf(p, sz, "%pU", brid->uuid.b); ++ break; ++ case AuBrid_FSID: ++ err = snprintf(p, sz, "%08x-%08x", ++ brid->fsid.val[0], brid->fsid.val[1]); ++ break; ++ case AuBrid_DEV: ++ major = MAJOR(brid->dev); ++ minor = MINOR(brid->dev); ++ if (major <= 0xff && minor <= 0xff) ++ err = snprintf(p, sz, "%02x%02x", major, minor); ++ else ++ err = snprintf(p, sz, "%03x:%05x", major, minor); ++ break; ++ } ++ AuDebugOn(err > sz); ++ p += err; ++ sz -= err; ++ err = snprintf(p, sz, "_%llu", (unsigned long long)h_inode->i_ino); ++ AuDebugOn(err > sz); ++ p += err; ++ sz -= err; + -+ for (i = 0; i < AuDyLast; i++) -+ au_sphl_init(dynop + i); ++ return p - buf; +} + -+void au_dy_fin(void) ++static int au_drinfo_name(struct au_branch *br, char *name, int len) +{ -+ int i; ++ int rlen; ++ struct dentry *br_dentry; ++ struct inode *br_inode; + -+ for (i = 0; i < AuDyLast; i++) -+ WARN_ON(!hlist_empty(&dynop[i].head)); ++ br_dentry = au_br_dentry(br); ++ br_inode = d_inode(br_dentry); ++ rlen = au_brid_str(&br->br_dirren.dr_brid, br_inode, name, len); ++ AuDebugOn(rlen >= AUFS_DIRREN_ENV_VAL_SZ); ++ AuDebugOn(rlen > len); ++ ++ return rlen; +} -diff -urN /usr/share/empty/fs/aufs/dynop.h linux/fs/aufs/dynop.h ---- /usr/share/empty/fs/aufs/dynop.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/dynop.h 2017-05-06 22:16:52.778220093 +0200 -@@ -0,0 +1,74 @@ -+/* -+ * Copyright (C) 2010-2017 Junjiro R. Okajima -+ * -+ * This program, aufs is free software; you can redistribute it and/or modify -+ * it under the terms of the GNU General Public License as published by -+ * the Free Software Foundation; either version 2 of the License, or -+ * (at your option) any later version. -+ * -+ * This program is distributed in the hope that it will be useful, -+ * but WITHOUT ANY WARRANTY; without even the implied warranty of -+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the -+ * GNU General Public License for more details. -+ * -+ * You should have received a copy of the GNU General Public License -+ * along with this program. If not, see . -+ */ ++ ++/* ---------------------------------------------------------------------- */ + +/* -+ * dynamically customizable operations (for regular files only) ++ * from the given @h_dentry, construct drinfo at @*fdata. ++ * when the size of @*fdata is not enough, reallocate and return new @fdata and ++ * @allocated. + */ ++static int au_drinfo_construct(struct au_drinfo_fdata **fdata, ++ struct dentry *h_dentry, ++ unsigned char *allocated) ++{ ++ int err, v; ++ struct au_drinfo_fdata *f, *p; ++ struct au_drinfo *drinfo; ++ struct inode *h_inode; ++ struct qstr *qname; ++ ++ err = 0; ++ f = *fdata; ++ h_inode = d_inode(h_dentry); ++ qname = &h_dentry->d_name; ++ drinfo = &f->drinfo; ++ drinfo->ino = (__force uint64_t)cpu_to_be64(h_inode->i_ino); ++ drinfo->oldnamelen = qname->len; ++ if (*allocated < sizeof(*f) + qname->len) { ++ v = roundup_pow_of_two(*allocated + qname->len); ++ p = au_krealloc(f, v, GFP_NOFS, /*may_shrink*/0); ++ if (unlikely(!p)) { ++ err = -ENOMEM; ++ AuTraceErr(err); ++ goto out; ++ } ++ f = p; ++ *fdata = f; ++ *allocated = v; ++ drinfo = &f->drinfo; ++ } ++ memcpy(drinfo->oldname, qname->name, qname->len); ++ AuDbg("i%llu, %.*s\n", ++ be64_to_cpu((__force __be64)drinfo->ino), drinfo->oldnamelen, ++ drinfo->oldname); ++ ++out: ++ AuTraceErr(err); ++ return err; ++} ++ ++/* callers have to free the return value */ ++static struct au_drinfo *au_drinfo_read_k(struct file *file, ino_t h_ino) ++{ ++ struct au_drinfo *ret, *drinfo; ++ struct au_drinfo_fdata fdata; ++ int len; ++ loff_t pos; ++ ssize_t ssz; ++ ++ ret = ERR_PTR(-EIO); ++ pos = 0; ++ ssz = vfsub_read_k(file, &fdata, sizeof(fdata), &pos); ++ if (unlikely(ssz != sizeof(fdata))) { ++ AuIOErr("ssz %zd, %u, %pD2\n", ++ ssz, (unsigned int)sizeof(fdata), file); ++ goto out; ++ } ++ ++ fdata.magic = ntohl((__force __be32)fdata.magic); ++ switch (fdata.magic) { ++ case AUFS_DRINFO_MAGIC_V1: ++ break; ++ default: ++ AuIOErr("magic-num 0x%x, 0x%x, %pD2\n", ++ fdata.magic, AUFS_DRINFO_MAGIC_V1, file); ++ goto out; ++ } ++ ++ drinfo = &fdata.drinfo; ++ len = drinfo->oldnamelen; ++ if (!len) { ++ AuIOErr("broken drinfo %pD2\n", file); ++ goto out; ++ } ++ ++ ret = NULL; ++ drinfo->ino = be64_to_cpu((__force __be64)drinfo->ino); ++ if (unlikely(h_ino && drinfo->ino != h_ino)) { ++ AuDbg("ignored i%llu, i%llu, %pD2\n", ++ (unsigned long long)drinfo->ino, ++ (unsigned long long)h_ino, file); ++ goto out; /* success */ ++ } ++ ++ ret = kmalloc(sizeof(*ret) + len, GFP_NOFS); ++ if (unlikely(!ret)) { ++ ret = ERR_PTR(-ENOMEM); ++ AuTraceErrPtr(ret); ++ goto out; ++ } ++ ++ *ret = *drinfo; ++ ssz = vfsub_read_k(file, (void *)ret->oldname, len, &pos); ++ if (unlikely(ssz != len)) { ++ au_kfree_rcu(ret); ++ ret = ERR_PTR(-EIO); ++ AuIOErr("ssz %zd, %u, %pD2\n", ssz, len, file); ++ goto out; ++ } ++ ++ AuDbg("oldname %.*s\n", ret->oldnamelen, ret->oldname); ++ ++out: ++ return ret; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* in order to be revertible */ ++struct au_drinfo_rev_elm { ++ int created; ++ struct dentry *info_dentry; ++ struct au_drinfo *info_last; ++}; ++ ++struct au_drinfo_rev { ++ unsigned char already; ++ aufs_bindex_t nelm; ++ struct au_drinfo_rev_elm elm[0]; ++}; ++ ++/* todo: isn't it too large? */ ++struct au_drinfo_store { ++ struct path h_ppath; ++ struct dentry *h_dentry; ++ struct au_drinfo_fdata *fdata; ++ char *infoname; /* inside of whname, just after PFX */ ++ char whname[sizeof(AUFS_WH_DR_INFO_PFX) + AUFS_DIRREN_ENV_VAL_SZ]; ++ aufs_bindex_t btgt, btail; ++ unsigned char no_sio, ++ allocated, /* current size of *fdata */ ++ infonamelen, /* room size for p */ ++ whnamelen, /* length of the generated name */ ++ renameback; /* renamed back */ ++}; ++ ++/* on rename(2) error, the caller should revert it using @elm */ ++static int au_drinfo_do_store(struct au_drinfo_store *w, ++ struct au_drinfo_rev_elm *elm) ++{ ++ int err, len; ++ ssize_t ssz; ++ loff_t pos; ++ struct path infopath = { ++ .mnt = w->h_ppath.mnt ++ }; ++ struct inode *h_dir, *h_inode, *delegated; ++ struct file *infofile; ++ struct qstr *qname; ++ ++ AuDebugOn(elm ++ && memcmp(elm, page_address(ZERO_PAGE(0)), sizeof(*elm))); ++ ++ infopath.dentry = vfsub_lookup_one_len(w->whname, w->h_ppath.dentry, ++ w->whnamelen); ++ AuTraceErrPtr(infopath.dentry); ++ if (IS_ERR(infopath.dentry)) { ++ err = PTR_ERR(infopath.dentry); ++ goto out; ++ } ++ ++ err = 0; ++ h_dir = d_inode(w->h_ppath.dentry); ++ if (elm && d_is_negative(infopath.dentry)) { ++ err = vfsub_create(h_dir, &infopath, 0600, /*want_excl*/true); ++ AuTraceErr(err); ++ if (unlikely(err)) ++ goto out_dput; ++ elm->created = 1; ++ elm->info_dentry = dget(infopath.dentry); ++ } ++ ++ infofile = vfsub_dentry_open(&infopath, O_RDWR); ++ AuTraceErrPtr(infofile); ++ if (IS_ERR(infofile)) { ++ err = PTR_ERR(infofile); ++ goto out_dput; ++ } ++ ++ h_inode = d_inode(infopath.dentry); ++ if (elm && i_size_read(h_inode)) { ++ h_inode = d_inode(w->h_dentry); ++ elm->info_last = au_drinfo_read_k(infofile, h_inode->i_ino); ++ AuTraceErrPtr(elm->info_last); ++ if (IS_ERR(elm->info_last)) { ++ err = PTR_ERR(elm->info_last); ++ elm->info_last = NULL; ++ AuDebugOn(elm->info_dentry); ++ goto out_fput; ++ } ++ } ++ ++ if (elm && w->renameback) { ++ delegated = NULL; ++ err = vfsub_unlink(h_dir, &infopath, &delegated, /*force*/0); ++ AuTraceErr(err); ++ if (unlikely(err == -EWOULDBLOCK)) ++ iput(delegated); ++ goto out_fput; ++ } ++ ++ pos = 0; ++ qname = &w->h_dentry->d_name; ++ len = sizeof(*w->fdata) + qname->len; ++ if (!elm) ++ len = sizeof(*w->fdata) + w->fdata->drinfo.oldnamelen; ++ ssz = vfsub_write_k(infofile, w->fdata, len, &pos); ++ if (ssz == len) { ++ AuDbg("hi%llu, %.*s\n", w->fdata->drinfo.ino, ++ w->fdata->drinfo.oldnamelen, w->fdata->drinfo.oldname); ++ goto out_fput; /* success */ ++ } else { ++ err = -EIO; ++ if (ssz < 0) ++ err = ssz; ++ /* the caller should revert it using @elm */ ++ } ++ ++out_fput: ++ fput(infofile); ++out_dput: ++ dput(infopath.dentry); ++out: ++ AuTraceErr(err); ++ return err; ++} ++ ++struct au_call_drinfo_do_store_args { ++ int *errp; ++ struct au_drinfo_store *w; ++ struct au_drinfo_rev_elm *elm; ++}; ++ ++static void au_call_drinfo_do_store(void *args) ++{ ++ struct au_call_drinfo_do_store_args *a = args; ++ ++ *a->errp = au_drinfo_do_store(a->w, a->elm); ++} ++ ++static int au_drinfo_store_sio(struct au_drinfo_store *w, ++ struct au_drinfo_rev_elm *elm) ++{ ++ int err, wkq_err; ++ ++ if (w->no_sio) ++ err = au_drinfo_do_store(w, elm); ++ else { ++ struct au_call_drinfo_do_store_args a = { ++ .errp = &err, ++ .w = w, ++ .elm = elm ++ }; ++ wkq_err = au_wkq_wait(au_call_drinfo_do_store, &a); ++ if (unlikely(wkq_err)) ++ err = wkq_err; ++ } ++ AuTraceErr(err); ++ ++ return err; ++} ++ ++static int au_drinfo_store_work_init(struct au_drinfo_store *w, ++ aufs_bindex_t btgt) ++{ ++ int err; ++ ++ memset(w, 0, sizeof(*w)); ++ w->allocated = roundup_pow_of_two(sizeof(*w->fdata) + 40); ++ strcpy(w->whname, AUFS_WH_DR_INFO_PFX); ++ w->infoname = w->whname + sizeof(AUFS_WH_DR_INFO_PFX) - 1; ++ w->infonamelen = sizeof(w->whname) - sizeof(AUFS_WH_DR_INFO_PFX); ++ w->btgt = btgt; ++ w->no_sio = !!uid_eq(current_fsuid(), GLOBAL_ROOT_UID); ++ ++ err = -ENOMEM; ++ w->fdata = kcalloc(1, w->allocated, GFP_NOFS); ++ if (unlikely(!w->fdata)) { ++ AuTraceErr(err); ++ goto out; ++ } ++ w->fdata->magic = (__force uint32_t)htonl(AUFS_DRINFO_MAGIC_V1); ++ err = 0; ++ ++out: ++ return err; ++} ++ ++static void au_drinfo_store_work_fin(struct au_drinfo_store *w) ++{ ++ au_kfree_rcu(w->fdata); ++} ++ ++static void au_drinfo_store_rev(struct au_drinfo_rev *rev, ++ struct au_drinfo_store *w) ++{ ++ struct au_drinfo_rev_elm *elm; ++ struct inode *h_dir, *delegated; ++ int err, nelm; ++ struct path infopath = { ++ .mnt = w->h_ppath.mnt ++ }; ++ ++ h_dir = d_inode(w->h_ppath.dentry); ++ IMustLock(h_dir); ++ ++ err = 0; ++ elm = rev->elm; ++ for (nelm = rev->nelm; nelm > 0; nelm--, elm++) { ++ AuDebugOn(elm->created && elm->info_last); ++ if (elm->created) { ++ AuDbg("here\n"); ++ delegated = NULL; ++ infopath.dentry = elm->info_dentry; ++ err = vfsub_unlink(h_dir, &infopath, &delegated, ++ !w->no_sio); ++ AuTraceErr(err); ++ if (unlikely(err == -EWOULDBLOCK)) ++ iput(delegated); ++ dput(elm->info_dentry); ++ } else if (elm->info_last) { ++ AuDbg("here\n"); ++ w->fdata->drinfo = *elm->info_last; ++ memcpy(w->fdata->drinfo.oldname, ++ elm->info_last->oldname, ++ elm->info_last->oldnamelen); ++ err = au_drinfo_store_sio(w, /*elm*/NULL); ++ au_kfree_rcu(elm->info_last); ++ } ++ if (unlikely(err)) ++ AuIOErr("%d, %s\n", err, w->whname); ++ /* go on even if err */ ++ } ++} ++ ++/* caller has to call au_dr_rename_fin() later */ ++static int au_drinfo_store(struct dentry *dentry, aufs_bindex_t btgt, ++ struct qstr *dst_name, void *_rev) ++{ ++ int err, sz, nelm; ++ aufs_bindex_t bindex, btail; ++ struct au_drinfo_store work; ++ struct au_drinfo_rev *rev, **p; ++ struct au_drinfo_rev_elm *elm; ++ struct super_block *sb; ++ struct au_branch *br; ++ struct au_hinode *hdir; ++ ++ err = au_drinfo_store_work_init(&work, btgt); ++ AuTraceErr(err); ++ if (unlikely(err)) ++ goto out; ++ ++ err = -ENOMEM; ++ btail = au_dbtaildir(dentry); ++ nelm = btail - btgt; ++ sz = sizeof(*rev) + sizeof(*elm) * nelm; ++ rev = kcalloc(1, sz, GFP_NOFS); ++ if (unlikely(!rev)) { ++ AuTraceErr(err); ++ goto out_args; ++ } ++ rev->nelm = nelm; ++ elm = rev->elm; ++ p = _rev; ++ *p = rev; ++ ++ err = 0; ++ sb = dentry->d_sb; ++ work.h_ppath.dentry = au_h_dptr(dentry, btgt); ++ work.h_ppath.mnt = au_sbr_mnt(sb, btgt); ++ hdir = au_hi(d_inode(dentry), btgt); ++ au_hn_inode_lock_nested(hdir, AuLsc_I_CHILD); ++ for (bindex = btgt + 1; bindex <= btail; bindex++, elm++) { ++ work.h_dentry = au_h_dptr(dentry, bindex); ++ if (!work.h_dentry) ++ continue; ++ ++ err = au_drinfo_construct(&work.fdata, work.h_dentry, ++ &work.allocated); ++ AuTraceErr(err); ++ if (unlikely(err)) ++ break; ++ ++ work.renameback = au_qstreq(&work.h_dentry->d_name, dst_name); ++ br = au_sbr(sb, bindex); ++ work.whnamelen = sizeof(AUFS_WH_DR_INFO_PFX) - 1; ++ work.whnamelen += au_drinfo_name(br, work.infoname, ++ work.infonamelen); ++ AuDbg("whname %.*s, i%llu, %.*s\n", ++ work.whnamelen, work.whname, ++ be64_to_cpu((__force __be64)work.fdata->drinfo.ino), ++ work.fdata->drinfo.oldnamelen, ++ work.fdata->drinfo.oldname); ++ ++ err = au_drinfo_store_sio(&work, elm); ++ AuTraceErr(err); ++ if (unlikely(err)) ++ break; ++ } ++ if (unlikely(err)) { ++ /* revert all drinfo */ ++ au_drinfo_store_rev(rev, &work); ++ au_kfree_try_rcu(rev); ++ *p = NULL; ++ } ++ au_hn_inode_unlock(hdir); ++ ++out_args: ++ au_drinfo_store_work_fin(&work); ++out: ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++int au_dr_rename(struct dentry *src, aufs_bindex_t bindex, ++ struct qstr *dst_name, void *_rev) ++{ ++ int err, already; ++ ino_t ino; ++ struct super_block *sb; ++ struct au_branch *br; ++ struct au_dr_br *dr; ++ struct dentry *h_dentry; ++ struct inode *h_inode; ++ struct au_dr_hino *ent; ++ struct au_drinfo_rev *rev, **p; ++ ++ AuDbg("bindex %d\n", bindex); ++ ++ err = -ENOMEM; ++ ent = kmalloc(sizeof(*ent), GFP_NOFS); ++ if (unlikely(!ent)) ++ goto out; ++ ++ sb = src->d_sb; ++ br = au_sbr(sb, bindex); ++ dr = &br->br_dirren; ++ h_dentry = au_h_dptr(src, bindex); ++ h_inode = d_inode(h_dentry); ++ ino = h_inode->i_ino; ++ ent->dr_h_ino = ino; ++ already = au_dr_hino_test_add(dr, ino, ent); ++ AuDbg("b%d, hi%llu, already %d\n", ++ bindex, (unsigned long long)ino, already); ++ ++ err = au_drinfo_store(src, bindex, dst_name, _rev); ++ AuTraceErr(err); ++ if (!err) { ++ p = _rev; ++ rev = *p; ++ rev->already = already; ++ goto out; /* success */ ++ } ++ ++ /* revert */ ++ if (!already) ++ au_dr_hino_del(dr, ent); ++ au_kfree_rcu(ent); ++ ++out: ++ AuTraceErr(err); ++ return err; ++} ++ ++void au_dr_rename_fin(struct dentry *src, aufs_bindex_t btgt, void *_rev) ++{ ++ struct au_drinfo_rev *rev; ++ struct au_drinfo_rev_elm *elm; ++ int nelm; ++ ++ rev = _rev; ++ elm = rev->elm; ++ for (nelm = rev->nelm; nelm > 0; nelm--, elm++) { ++ dput(elm->info_dentry); ++ au_kfree_rcu(elm->info_last); ++ } ++ au_kfree_try_rcu(rev); ++} ++ ++void au_dr_rename_rev(struct dentry *src, aufs_bindex_t btgt, void *_rev) ++{ ++ int err; ++ struct au_drinfo_store work; ++ struct au_drinfo_rev *rev = _rev; ++ struct super_block *sb; ++ struct au_branch *br; ++ struct inode *h_inode; ++ struct au_dr_br *dr; ++ struct au_dr_hino *ent; ++ ++ err = au_drinfo_store_work_init(&work, btgt); ++ if (unlikely(err)) ++ goto out; ++ ++ sb = src->d_sb; ++ br = au_sbr(sb, btgt); ++ work.h_ppath.dentry = au_h_dptr(src, btgt); ++ work.h_ppath.mnt = au_br_mnt(br); ++ au_drinfo_store_rev(rev, &work); ++ au_drinfo_store_work_fin(&work); ++ if (rev->already) ++ goto out; ++ ++ dr = &br->br_dirren; ++ h_inode = d_inode(work.h_ppath.dentry); ++ ent = au_dr_hino_find(dr, h_inode->i_ino); ++ BUG_ON(!ent); ++ au_dr_hino_del(dr, ent); ++ au_kfree_rcu(ent); ++ ++out: ++ au_kfree_try_rcu(rev); ++ if (unlikely(err)) ++ pr_err("failed to remove dirren info\n"); ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++static struct au_drinfo *au_drinfo_do_load(struct path *h_ppath, ++ char *whname, int whnamelen, ++ struct dentry **info_dentry) ++{ ++ struct au_drinfo *drinfo; ++ struct file *f; ++ struct inode *h_dir; ++ struct path infopath; ++ int unlocked; ++ ++ AuDbg("%pd/%.*s\n", h_ppath->dentry, whnamelen, whname); ++ ++ *info_dentry = NULL; ++ drinfo = NULL; ++ unlocked = 0; ++ h_dir = d_inode(h_ppath->dentry); ++ inode_lock_shared_nested(h_dir, AuLsc_I_PARENT); ++ infopath.dentry = vfsub_lookup_one_len(whname, h_ppath->dentry, ++ whnamelen); ++ if (IS_ERR(infopath.dentry)) { ++ drinfo = (void *)infopath.dentry; ++ goto out; ++ } ++ ++ if (d_is_negative(infopath.dentry)) ++ goto out_dput; /* success */ ++ ++ infopath.mnt = h_ppath->mnt; ++ f = vfsub_dentry_open(&infopath, O_RDONLY); ++ inode_unlock_shared(h_dir); ++ unlocked = 1; ++ if (IS_ERR(f)) { ++ drinfo = (void *)f; ++ goto out_dput; ++ } ++ ++ drinfo = au_drinfo_read_k(f, /*h_ino*/0); ++ if (IS_ERR_OR_NULL(drinfo)) ++ goto out_fput; ++ ++ AuDbg("oldname %.*s\n", drinfo->oldnamelen, drinfo->oldname); ++ *info_dentry = dget(infopath.dentry); /* keep it alive */ ++ ++out_fput: ++ fput(f); ++out_dput: ++ dput(infopath.dentry); ++out: ++ if (!unlocked) ++ inode_unlock_shared(h_dir); ++ AuTraceErrPtr(drinfo); ++ return drinfo; ++} ++ ++struct au_drinfo_do_load_args { ++ struct au_drinfo **drinfop; ++ struct path *h_ppath; ++ char *whname; ++ int whnamelen; ++ struct dentry **info_dentry; ++}; ++ ++static void au_call_drinfo_do_load(void *args) ++{ ++ struct au_drinfo_do_load_args *a = args; ++ ++ *a->drinfop = au_drinfo_do_load(a->h_ppath, a->whname, a->whnamelen, ++ a->info_dentry); ++} ++ ++struct au_drinfo_load { ++ struct path h_ppath; ++ struct qstr *qname; ++ unsigned char no_sio; ++ ++ aufs_bindex_t ninfo; ++ struct au_drinfo **drinfo; ++}; ++ ++static int au_drinfo_load(struct au_drinfo_load *w, aufs_bindex_t bindex, ++ struct au_branch *br) ++{ ++ int err, wkq_err, whnamelen, e; ++ char whname[sizeof(AUFS_WH_DR_INFO_PFX) + AUFS_DIRREN_ENV_VAL_SZ] ++ = AUFS_WH_DR_INFO_PFX; ++ struct au_drinfo *drinfo; ++ struct qstr oldname; ++ struct inode *h_dir, *delegated; ++ struct dentry *info_dentry; ++ struct path infopath; ++ ++ whnamelen = sizeof(AUFS_WH_DR_INFO_PFX) - 1; ++ whnamelen += au_drinfo_name(br, whname + whnamelen, ++ sizeof(whname) - whnamelen); ++ if (w->no_sio) ++ drinfo = au_drinfo_do_load(&w->h_ppath, whname, whnamelen, ++ &info_dentry); ++ else { ++ struct au_drinfo_do_load_args args = { ++ .drinfop = &drinfo, ++ .h_ppath = &w->h_ppath, ++ .whname = whname, ++ .whnamelen = whnamelen, ++ .info_dentry = &info_dentry ++ }; ++ wkq_err = au_wkq_wait(au_call_drinfo_do_load, &args); ++ if (unlikely(wkq_err)) ++ drinfo = ERR_PTR(wkq_err); ++ } ++ err = PTR_ERR(drinfo); ++ if (IS_ERR_OR_NULL(drinfo)) ++ goto out; ++ ++ err = 0; ++ oldname.len = drinfo->oldnamelen; ++ oldname.name = drinfo->oldname; ++ if (au_qstreq(w->qname, &oldname)) { ++ /* the name is renamed back */ ++ au_kfree_rcu(drinfo); ++ drinfo = NULL; ++ ++ infopath.dentry = info_dentry; ++ infopath.mnt = w->h_ppath.mnt; ++ h_dir = d_inode(w->h_ppath.dentry); ++ delegated = NULL; ++ inode_lock_nested(h_dir, AuLsc_I_PARENT); ++ e = vfsub_unlink(h_dir, &infopath, &delegated, !w->no_sio); ++ inode_unlock(h_dir); ++ if (unlikely(e)) ++ AuIOErr("ignored %d, %pd2\n", e, &infopath.dentry); ++ if (unlikely(e == -EWOULDBLOCK)) ++ iput(delegated); ++ } ++ au_kfree_rcu(w->drinfo[bindex]); ++ w->drinfo[bindex] = drinfo; ++ dput(info_dentry); ++ ++out: ++ AuTraceErr(err); ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++static void au_dr_lkup_free(struct au_drinfo **drinfo, int n) ++{ ++ struct au_drinfo **p = drinfo; ++ ++ while (n-- > 0) ++ au_kfree_rcu(*drinfo++); ++ au_kfree_try_rcu(p); ++} ++ ++int au_dr_lkup(struct au_do_lookup_args *lkup, struct dentry *dentry, ++ aufs_bindex_t btgt) ++{ ++ int err, ninfo; ++ struct au_drinfo_load w; ++ aufs_bindex_t bindex, bbot; ++ struct au_branch *br; ++ struct inode *h_dir; ++ struct au_dr_hino *ent; ++ struct super_block *sb; ++ ++ AuDbg("%.*s, name %.*s, whname %.*s, b%d\n", ++ AuLNPair(&dentry->d_name), AuLNPair(&lkup->dirren.dr_name), ++ AuLNPair(&lkup->whname), btgt); ++ ++ sb = dentry->d_sb; ++ bbot = au_sbbot(sb); ++ w.ninfo = bbot + 1; ++ if (!lkup->dirren.drinfo) { ++ lkup->dirren.drinfo = kcalloc(w.ninfo, ++ sizeof(*lkup->dirren.drinfo), ++ GFP_NOFS); ++ if (unlikely(!lkup->dirren.drinfo)) { ++ err = -ENOMEM; ++ goto out; ++ } ++ lkup->dirren.ninfo = w.ninfo; ++ } ++ w.drinfo = lkup->dirren.drinfo; ++ w.no_sio = !!uid_eq(current_fsuid(), GLOBAL_ROOT_UID); ++ w.h_ppath.dentry = au_h_dptr(dentry, btgt); ++ AuDebugOn(!w.h_ppath.dentry); ++ w.h_ppath.mnt = au_sbr_mnt(sb, btgt); ++ w.qname = &dentry->d_name; ++ ++ ninfo = 0; ++ for (bindex = btgt + 1; bindex <= bbot; bindex++) { ++ br = au_sbr(sb, bindex); ++ err = au_drinfo_load(&w, bindex, br); ++ if (unlikely(err)) ++ goto out_free; ++ if (w.drinfo[bindex]) ++ ninfo++; ++ } ++ if (!ninfo) { ++ br = au_sbr(sb, btgt); ++ h_dir = d_inode(w.h_ppath.dentry); ++ ent = au_dr_hino_find(&br->br_dirren, h_dir->i_ino); ++ AuDebugOn(!ent); ++ au_dr_hino_del(&br->br_dirren, ent); ++ au_kfree_rcu(ent); ++ } ++ goto out; /* success */ ++ ++out_free: ++ au_dr_lkup_free(lkup->dirren.drinfo, lkup->dirren.ninfo); ++ lkup->dirren.ninfo = 0; ++ lkup->dirren.drinfo = NULL; ++out: ++ AuTraceErr(err); ++ return err; ++} ++ ++void au_dr_lkup_fin(struct au_do_lookup_args *lkup) ++{ ++ au_dr_lkup_free(lkup->dirren.drinfo, lkup->dirren.ninfo); ++} ++ ++int au_dr_lkup_name(struct au_do_lookup_args *lkup, aufs_bindex_t btgt) ++{ ++ int err; ++ struct au_drinfo *drinfo; ++ ++ err = 0; ++ if (!lkup->dirren.drinfo) ++ goto out; ++ AuDebugOn(lkup->dirren.ninfo < btgt + 1); ++ drinfo = lkup->dirren.drinfo[btgt + 1]; ++ if (!drinfo) ++ goto out; ++ ++ au_kfree_try_rcu(lkup->whname.name); ++ lkup->whname.name = NULL; ++ lkup->dirren.dr_name.len = drinfo->oldnamelen; ++ lkup->dirren.dr_name.name = drinfo->oldname; ++ lkup->name = &lkup->dirren.dr_name; ++ err = au_wh_name_alloc(&lkup->whname, lkup->name); ++ if (!err) ++ AuDbg("name %.*s, whname %.*s, b%d\n", ++ AuLNPair(lkup->name), AuLNPair(&lkup->whname), ++ btgt); ++ ++out: ++ AuTraceErr(err); ++ return err; ++} ++ ++int au_dr_lkup_h_ino(struct au_do_lookup_args *lkup, aufs_bindex_t bindex, ++ ino_t h_ino) ++{ ++ int match; ++ struct au_drinfo *drinfo; ++ ++ match = 1; ++ if (!lkup->dirren.drinfo) ++ goto out; ++ AuDebugOn(lkup->dirren.ninfo < bindex + 1); ++ drinfo = lkup->dirren.drinfo[bindex + 1]; ++ if (!drinfo) ++ goto out; ++ ++ match = (drinfo->ino == h_ino); ++ AuDbg("match %d\n", match); ++ ++out: ++ return match; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++int au_dr_opt_set(struct super_block *sb) ++{ ++ int err; ++ aufs_bindex_t bindex, bbot; ++ struct au_branch *br; ++ ++ err = 0; ++ bbot = au_sbbot(sb); ++ for (bindex = 0; !err && bindex <= bbot; bindex++) { ++ br = au_sbr(sb, bindex); ++ err = au_dr_hino(sb, bindex, /*br*/NULL, &br->br_path); ++ } ++ ++ return err; ++} ++ ++int au_dr_opt_flush(struct super_block *sb) ++{ ++ int err; ++ aufs_bindex_t bindex, bbot; ++ struct au_branch *br; ++ ++ err = 0; ++ bbot = au_sbbot(sb); ++ for (bindex = 0; !err && bindex <= bbot; bindex++) { ++ br = au_sbr(sb, bindex); ++ if (au_br_writable(br->br_perm)) ++ err = au_dr_hino(sb, bindex, /*br*/NULL, /*path*/NULL); ++ } ++ ++ return err; ++} ++ ++int au_dr_opt_clr(struct super_block *sb, int no_flush) ++{ ++ int err; ++ aufs_bindex_t bindex, bbot; ++ struct au_branch *br; ++ ++ err = 0; ++ if (!no_flush) { ++ err = au_dr_opt_flush(sb); ++ if (unlikely(err)) ++ goto out; ++ } ++ ++ bbot = au_sbbot(sb); ++ for (bindex = 0; bindex <= bbot; bindex++) { ++ br = au_sbr(sb, bindex); ++ au_dr_hino_free(&br->br_dirren); ++ } ++ ++out: ++ return err; ++} +diff -urN /usr/share/empty/fs/aufs/dirren.h linux/fs/aufs/dirren.h +--- /usr/share/empty/fs/aufs/dirren.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux/fs/aufs/dirren.h 2019-03-05 12:13:00.139224339 +0100 +@@ -0,0 +1,140 @@ ++/* SPDX-License-Identifier: GPL-2.0 */ ++/* ++ * Copyright (C) 2017-2019 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see . ++ */ ++ ++/* ++ * renamed dir info ++ */ ++ ++#ifndef __AUFS_DIRREN_H__ ++#define __AUFS_DIRREN_H__ ++ ++#ifdef __KERNEL__ ++ ++#include ++#include ++#include ++#include "hbl.h" ++ ++#define AuDirren_NHASH 100 ++ ++#ifdef CONFIG_AUFS_DIRREN ++enum au_brid_type { ++ AuBrid_Unset, ++ AuBrid_UUID, ++ AuBrid_FSID, ++ AuBrid_DEV ++}; ++ ++struct au_dr_brid { ++ enum au_brid_type type; ++ union { ++ uuid_t uuid; /* unimplemented yet */ ++ fsid_t fsid; ++ dev_t dev; ++ }; ++}; ++ ++/* 20 is the max digits length of ulong 64 */ ++/* brid-type "_" uuid "_" inum */ ++#define AUFS_DIRREN_FNAME_SZ (1 + 1 + UUID_STRING_LEN + 20) ++#define AUFS_DIRREN_ENV_VAL_SZ (AUFS_DIRREN_FNAME_SZ + 1 + 20) ++ ++struct au_dr_hino { ++ struct hlist_bl_node dr_hnode; ++ ino_t dr_h_ino; ++}; ++ ++struct au_dr_br { ++ struct hlist_bl_head dr_h_ino[AuDirren_NHASH]; ++ struct au_dr_brid dr_brid; ++}; ++ ++struct au_dr_lookup { ++ /* dr_name is pointed by struct au_do_lookup_args.name */ ++ struct qstr dr_name; /* subset of dr_info */ ++ aufs_bindex_t ninfo; ++ struct au_drinfo **drinfo; ++}; ++#else ++struct au_dr_hino; ++/* empty */ ++struct au_dr_br { }; ++struct au_dr_lookup { }; ++#endif ++ ++/* ---------------------------------------------------------------------- */ ++ ++struct au_branch; ++struct au_do_lookup_args; ++struct au_hinode; ++#ifdef CONFIG_AUFS_DIRREN ++int au_dr_hino_test_add(struct au_dr_br *dr, ino_t h_ino, ++ struct au_dr_hino *add_ent); ++void au_dr_hino_free(struct au_dr_br *dr); ++int au_dr_br_init(struct super_block *sb, struct au_branch *br, ++ const struct path *path); ++int au_dr_br_fin(struct super_block *sb, struct au_branch *br); ++int au_dr_rename(struct dentry *src, aufs_bindex_t bindex, ++ struct qstr *dst_name, void *_rev); ++void au_dr_rename_fin(struct dentry *src, aufs_bindex_t btgt, void *rev); ++void au_dr_rename_rev(struct dentry *src, aufs_bindex_t bindex, void *rev); ++int au_dr_lkup(struct au_do_lookup_args *lkup, struct dentry *dentry, ++ aufs_bindex_t bindex); ++int au_dr_lkup_name(struct au_do_lookup_args *lkup, aufs_bindex_t btgt); ++int au_dr_lkup_h_ino(struct au_do_lookup_args *lkup, aufs_bindex_t bindex, ++ ino_t h_ino); ++void au_dr_lkup_fin(struct au_do_lookup_args *lkup); ++int au_dr_opt_set(struct super_block *sb); ++int au_dr_opt_flush(struct super_block *sb); ++int au_dr_opt_clr(struct super_block *sb, int no_flush); ++#else ++AuStubInt0(au_dr_hino_test_add, struct au_dr_br *dr, ino_t h_ino, ++ struct au_dr_hino *add_ent); ++AuStubVoid(au_dr_hino_free, struct au_dr_br *dr); ++AuStubInt0(au_dr_br_init, struct super_block *sb, struct au_branch *br, ++ const struct path *path); ++AuStubInt0(au_dr_br_fin, struct super_block *sb, struct au_branch *br); ++AuStubInt0(au_dr_rename, struct dentry *src, aufs_bindex_t bindex, ++ struct qstr *dst_name, void *_rev); ++AuStubVoid(au_dr_rename_fin, struct dentry *src, aufs_bindex_t btgt, void *rev); ++AuStubVoid(au_dr_rename_rev, struct dentry *src, aufs_bindex_t bindex, ++ void *rev); ++AuStubInt0(au_dr_lkup, struct au_do_lookup_args *lkup, struct dentry *dentry, ++ aufs_bindex_t bindex); ++AuStubInt0(au_dr_lkup_name, struct au_do_lookup_args *lkup, aufs_bindex_t btgt); ++AuStubInt0(au_dr_lkup_h_ino, struct au_do_lookup_args *lkup, ++ aufs_bindex_t bindex, ino_t h_ino); ++AuStubVoid(au_dr_lkup_fin, struct au_do_lookup_args *lkup); ++AuStubInt0(au_dr_opt_set, struct super_block *sb); ++AuStubInt0(au_dr_opt_flush, struct super_block *sb); ++AuStubInt0(au_dr_opt_clr, struct super_block *sb, int no_flush); ++#endif ++ ++/* ---------------------------------------------------------------------- */ ++ ++#ifdef CONFIG_AUFS_DIRREN ++static inline int au_dr_ihash(ino_t h_ino) ++{ ++ return h_ino % AuDirren_NHASH; ++} ++#else ++AuStubInt0(au_dr_ihash, ino_t h_ino); ++#endif ++ ++#endif /* __KERNEL__ */ ++#endif /* __AUFS_DIRREN_H__ */ +diff -urN /usr/share/empty/fs/aufs/dynop.c linux/fs/aufs/dynop.c +--- /usr/share/empty/fs/aufs/dynop.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux/fs/aufs/dynop.c 2019-07-08 09:31:38.324384703 +0200 +@@ -0,0 +1,370 @@ ++// SPDX-License-Identifier: GPL-2.0 ++/* ++ * Copyright (C) 2010-2019 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see . ++ */ ++ ++/* ++ * dynamically customizable operations for regular files ++ */ ++ ++#include "aufs.h" ++ ++#define DyPrSym(key) AuDbgSym(key->dk_op.dy_hop) ++ ++/* ++ * How large will these lists be? ++ * Usually just a few elements, 20-30 at most for each, I guess. ++ */ ++static struct hlist_bl_head dynop[AuDyLast]; ++ ++static struct au_dykey *dy_gfind_get(struct hlist_bl_head *hbl, ++ const void *h_op) ++{ ++ struct au_dykey *key, *tmp; ++ struct hlist_bl_node *pos; ++ ++ key = NULL; ++ hlist_bl_lock(hbl); ++ hlist_bl_for_each_entry(tmp, pos, hbl, dk_hnode) ++ if (tmp->dk_op.dy_hop == h_op) { ++ if (kref_get_unless_zero(&tmp->dk_kref)) ++ key = tmp; ++ break; ++ } ++ hlist_bl_unlock(hbl); ++ ++ return key; ++} ++ ++static struct au_dykey *dy_bradd(struct au_branch *br, struct au_dykey *key) ++{ ++ struct au_dykey **k, *found; ++ const void *h_op = key->dk_op.dy_hop; ++ int i; ++ ++ found = NULL; ++ k = br->br_dykey; ++ for (i = 0; i < AuBrDynOp; i++) ++ if (k[i]) { ++ if (k[i]->dk_op.dy_hop == h_op) { ++ found = k[i]; ++ break; ++ } ++ } else ++ break; ++ if (!found) { ++ spin_lock(&br->br_dykey_lock); ++ for (; i < AuBrDynOp; i++) ++ if (k[i]) { ++ if (k[i]->dk_op.dy_hop == h_op) { ++ found = k[i]; ++ break; ++ } ++ } else { ++ k[i] = key; ++ break; ++ } ++ spin_unlock(&br->br_dykey_lock); ++ BUG_ON(i == AuBrDynOp); /* expand the array */ ++ } ++ ++ return found; ++} ++ ++/* kref_get() if @key is already added */ ++static struct au_dykey *dy_gadd(struct hlist_bl_head *hbl, struct au_dykey *key) ++{ ++ struct au_dykey *tmp, *found; ++ struct hlist_bl_node *pos; ++ const void *h_op = key->dk_op.dy_hop; ++ ++ found = NULL; ++ hlist_bl_lock(hbl); ++ hlist_bl_for_each_entry(tmp, pos, hbl, dk_hnode) ++ if (tmp->dk_op.dy_hop == h_op) { ++ if (kref_get_unless_zero(&tmp->dk_kref)) ++ found = tmp; ++ break; ++ } ++ if (!found) ++ hlist_bl_add_head(&key->dk_hnode, hbl); ++ hlist_bl_unlock(hbl); ++ ++ if (!found) ++ DyPrSym(key); ++ return found; ++} ++ ++static void dy_free_rcu(struct rcu_head *rcu) ++{ ++ struct au_dykey *key; ++ ++ key = container_of(rcu, struct au_dykey, dk_rcu); ++ DyPrSym(key); ++ kfree(key); ++} ++ ++static void dy_free(struct kref *kref) ++{ ++ struct au_dykey *key; ++ struct hlist_bl_head *hbl; ++ ++ key = container_of(kref, struct au_dykey, dk_kref); ++ hbl = dynop + key->dk_op.dy_type; ++ au_hbl_del(&key->dk_hnode, hbl); ++ call_rcu(&key->dk_rcu, dy_free_rcu); ++} ++ ++void au_dy_put(struct au_dykey *key) ++{ ++ kref_put(&key->dk_kref, dy_free); ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++#define DyDbgSize(cnt, op) AuDebugOn(cnt != sizeof(op)/sizeof(void *)) ++ ++#ifdef CONFIG_AUFS_DEBUG ++#define DyDbgDeclare(cnt) unsigned int cnt = 0 ++#define DyDbgInc(cnt) do { cnt++; } while (0) ++#else ++#define DyDbgDeclare(cnt) do {} while (0) ++#define DyDbgInc(cnt) do {} while (0) ++#endif ++ ++#define DySet(func, dst, src, h_op, h_sb) do { \ ++ DyDbgInc(cnt); \ ++ if (h_op->func) { \ ++ if (src.func) \ ++ dst.func = src.func; \ ++ else \ ++ AuDbg("%s %s\n", au_sbtype(h_sb), #func); \ ++ } \ ++} while (0) ++ ++#define DySetForce(func, dst, src) do { \ ++ AuDebugOn(!src.func); \ ++ DyDbgInc(cnt); \ ++ dst.func = src.func; \ ++} while (0) ++ ++#define DySetAop(func) \ ++ DySet(func, dyaop->da_op, aufs_aop, h_aop, h_sb) ++#define DySetAopForce(func) \ ++ DySetForce(func, dyaop->da_op, aufs_aop) ++ ++static void dy_aop(struct au_dykey *key, const void *h_op, ++ struct super_block *h_sb __maybe_unused) ++{ ++ struct au_dyaop *dyaop = (void *)key; ++ const struct address_space_operations *h_aop = h_op; ++ DyDbgDeclare(cnt); ++ ++ AuDbg("%s\n", au_sbtype(h_sb)); ++ ++ DySetAop(writepage); ++ DySetAopForce(readpage); /* force */ ++ DySetAop(writepages); ++ DySetAop(set_page_dirty); ++ DySetAop(readpages); ++ DySetAop(write_begin); ++ DySetAop(write_end); ++ DySetAop(bmap); ++ DySetAop(invalidatepage); ++ DySetAop(releasepage); ++ DySetAop(freepage); ++ /* this one will be changed according to an aufs mount option */ ++ DySetAop(direct_IO); ++ DySetAop(migratepage); ++ DySetAop(isolate_page); ++ DySetAop(putback_page); ++ DySetAop(launder_page); ++ DySetAop(is_partially_uptodate); ++ DySetAop(is_dirty_writeback); ++ DySetAop(error_remove_page); ++ DySetAop(swap_activate); ++ DySetAop(swap_deactivate); ++ ++ DyDbgSize(cnt, *h_aop); ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++static void dy_bug(struct kref *kref) ++{ ++ BUG(); ++} ++ ++static struct au_dykey *dy_get(struct au_dynop *op, struct au_branch *br) ++{ ++ struct au_dykey *key, *old; ++ struct hlist_bl_head *hbl; ++ struct op { ++ unsigned int sz; ++ void (*set)(struct au_dykey *key, const void *h_op, ++ struct super_block *h_sb __maybe_unused); ++ }; ++ static const struct op a[] = { ++ [AuDy_AOP] = { ++ .sz = sizeof(struct au_dyaop), ++ .set = dy_aop ++ } ++ }; ++ const struct op *p; ++ ++ hbl = dynop + op->dy_type; ++ key = dy_gfind_get(hbl, op->dy_hop); ++ if (key) ++ goto out_add; /* success */ ++ ++ p = a + op->dy_type; ++ key = kzalloc(p->sz, GFP_NOFS); ++ if (unlikely(!key)) { ++ key = ERR_PTR(-ENOMEM); ++ goto out; ++ } ++ ++ key->dk_op.dy_hop = op->dy_hop; ++ kref_init(&key->dk_kref); ++ p->set(key, op->dy_hop, au_br_sb(br)); ++ old = dy_gadd(hbl, key); ++ if (old) { ++ au_kfree_rcu(key); ++ key = old; ++ } ++ ++out_add: ++ old = dy_bradd(br, key); ++ if (old) ++ /* its ref-count should never be zero here */ ++ kref_put(&key->dk_kref, dy_bug); ++out: ++ return key; ++} ++ ++/* ---------------------------------------------------------------------- */ ++/* ++ * Aufs prohibits O_DIRECT by default even if the branch supports it. ++ * This behaviour is necessary to return an error from open(O_DIRECT) instead ++ * of the succeeding I/O. The dio mount option enables O_DIRECT and makes ++ * open(O_DIRECT) always succeed, but the succeeding I/O may return an error. ++ * See the aufs manual in detail. ++ */ ++static void dy_adx(struct au_dyaop *dyaop, int do_dx) ++{ ++ if (!do_dx) ++ dyaop->da_op.direct_IO = NULL; ++ else ++ dyaop->da_op.direct_IO = aufs_aop.direct_IO; ++} ++ ++static struct au_dyaop *dy_aget(struct au_branch *br, ++ const struct address_space_operations *h_aop, ++ int do_dx) ++{ ++ struct au_dyaop *dyaop; ++ struct au_dynop op; ++ ++ op.dy_type = AuDy_AOP; ++ op.dy_haop = h_aop; ++ dyaop = (void *)dy_get(&op, br); ++ if (IS_ERR(dyaop)) ++ goto out; ++ dy_adx(dyaop, do_dx); ++ ++out: ++ return dyaop; ++} ++ ++int au_dy_iaop(struct inode *inode, aufs_bindex_t bindex, ++ struct inode *h_inode) ++{ ++ int err, do_dx; ++ struct super_block *sb; ++ struct au_branch *br; ++ struct au_dyaop *dyaop; ++ ++ AuDebugOn(!S_ISREG(h_inode->i_mode)); ++ IiMustWriteLock(inode); ++ ++ sb = inode->i_sb; ++ br = au_sbr(sb, bindex); ++ do_dx = !!au_opt_test(au_mntflags(sb), DIO); ++ dyaop = dy_aget(br, h_inode->i_mapping->a_ops, do_dx); ++ err = PTR_ERR(dyaop); ++ if (IS_ERR(dyaop)) ++ /* unnecessary to call dy_fput() */ ++ goto out; ++ ++ err = 0; ++ inode->i_mapping->a_ops = &dyaop->da_op; ++ ++out: ++ return err; ++} ++ ++/* ++ * Is it safe to replace a_ops during the inode/file is in operation? ++ * Yes, I hope so. ++ */ ++int au_dy_irefresh(struct inode *inode) ++{ ++ int err; ++ aufs_bindex_t btop; ++ struct inode *h_inode; ++ ++ err = 0; ++ if (S_ISREG(inode->i_mode)) { ++ btop = au_ibtop(inode); ++ h_inode = au_h_iptr(inode, btop); ++ err = au_dy_iaop(inode, btop, h_inode); ++ } ++ return err; ++} ++ ++void au_dy_arefresh(int do_dx) ++{ ++ struct hlist_bl_head *hbl; ++ struct hlist_bl_node *pos; ++ struct au_dykey *key; ++ ++ hbl = dynop + AuDy_AOP; ++ hlist_bl_lock(hbl); ++ hlist_bl_for_each_entry(key, pos, hbl, dk_hnode) ++ dy_adx((void *)key, do_dx); ++ hlist_bl_unlock(hbl); ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++void __init au_dy_init(void) ++{ ++ int i; ++ ++ /* make sure that 'struct au_dykey *' can be any type */ ++ BUILD_BUG_ON(offsetof(struct au_dyaop, da_key)); ++ ++ for (i = 0; i < AuDyLast; i++) ++ INIT_HLIST_BL_HEAD(dynop + i); ++} ++ ++void au_dy_fin(void) ++{ ++ int i; ++ ++ for (i = 0; i < AuDyLast; i++) ++ WARN_ON(!hlist_bl_empty(dynop + i)); ++} +diff -urN /usr/share/empty/fs/aufs/dynop.h linux/fs/aufs/dynop.h +--- /usr/share/empty/fs/aufs/dynop.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux/fs/aufs/dynop.h 2019-03-05 12:13:00.139224339 +0100 +@@ -0,0 +1,75 @@ ++/* SPDX-License-Identifier: GPL-2.0 */ ++/* ++ * Copyright (C) 2010-2019 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see . ++ */ ++ ++/* ++ * dynamically customizable operations (for regular files only) ++ */ ++ ++#ifndef __AUFS_DYNOP_H__ ++#define __AUFS_DYNOP_H__ ++ ++#ifdef __KERNEL__ ++ ++#include ++#include + -+#ifndef __AUFS_DYNOP_H__ -+#define __AUFS_DYNOP_H__ -+ -+#ifdef __KERNEL__ -+ -+#include -+#include -+ +enum {AuDy_AOP, AuDyLast}; + +struct au_dynop { @@ -11007,7 +12896,7 @@ diff -urN /usr/share/empty/fs/aufs/dynop.h linux/fs/aufs/dynop.h + +struct au_dykey { + union { -+ struct hlist_node dk_hnode; ++ struct hlist_bl_node dk_hnode; + struct rcu_head dk_rcu; + }; + struct au_dynop dk_op; @@ -11042,10 +12931,11 @@ diff -urN /usr/share/empty/fs/aufs/dynop.h linux/fs/aufs/dynop.h +#endif /* __AUFS_DYNOP_H__ */ diff -urN /usr/share/empty/fs/aufs/export.c linux/fs/aufs/export.c --- /usr/share/empty/fs/aufs/export.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/export.c 2017-05-06 22:16:52.778220093 +0200 -@@ -0,0 +1,836 @@ ++++ linux/fs/aufs/export.c 2019-03-05 12:13:00.139224339 +0100 +@@ -0,0 +1,838 @@ ++// SPDX-License-Identifier: GPL-2.0 +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -11227,7 +13117,7 @@ diff -urN /usr/share/empty/fs/aufs/export.c linux/fs/aufs/export.c + return err; +} + -+int au_xigen_set(struct super_block *sb, struct file *base) ++int au_xigen_set(struct super_block *sb, struct path *path) +{ + int err; + struct au_sbinfo *sbinfo; @@ -11236,7 +13126,7 @@ diff -urN /usr/share/empty/fs/aufs/export.c linux/fs/aufs/export.c + SiMustWriteLock(sb); + + sbinfo = au_sbi(sb); -+ file = au_xino_create2(base, sbinfo->si_xigen); ++ file = au_xino_create2(sb, path, sbinfo->si_xigen); + err = PTR_ERR(file); + if (IS_ERR(file)) + goto out; @@ -11246,6 +13136,7 @@ diff -urN /usr/share/empty/fs/aufs/export.c linux/fs/aufs/export.c + sbinfo->si_xigen = file; + +out: ++ AuTraceErr(err); + return err; +} + @@ -11462,7 +13353,7 @@ diff -urN /usr/share/empty/fs/aufs/export.c linux/fs/aufs/export.c + } + +out_name: -+ au_delayed_free_page((unsigned long)arg.name); ++ free_page((unsigned long)arg.name); +out_file: + fput(file); +out: @@ -11616,7 +13507,7 @@ diff -urN /usr/share/empty/fs/aufs/export.c linux/fs/aufs/export.c + dentry = ERR_PTR(-ESTALE); + } +out_pathname: -+ au_delayed_free_page((unsigned long)pathname); ++ free_page((unsigned long)pathname); +out_h_parent: + dput(h_parent); +out: @@ -11666,7 +13557,7 @@ diff -urN /usr/share/empty/fs/aufs/export.c linux/fs/aufs/export.c + + /* is the parent dir cached? */ + br = au_sbr(sb, nsi_lock.bindex); -+ au_br_get(br); ++ au_lcnt_inc(&br->br_nfiles); + dentry = decode_by_dir_ino(sb, ino, dir_ino, &nsi_lock); + if (IS_ERR(dentry)) + goto out_unlock; @@ -11690,7 +13581,7 @@ diff -urN /usr/share/empty/fs/aufs/export.c linux/fs/aufs/export.c + dentry = ERR_PTR(-ESTALE); +out_unlock: + if (br) -+ au_br_put(br); ++ au_lcnt_dec(&br->br_nfiles); + si_read_unlock(sb); +out: + AuTraceErrPtr(dentry); @@ -11882,10 +13773,11 @@ diff -urN /usr/share/empty/fs/aufs/export.c linux/fs/aufs/export.c +} diff -urN /usr/share/empty/fs/aufs/fhsm.c linux/fs/aufs/fhsm.c --- /usr/share/empty/fs/aufs/fhsm.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/fhsm.c 2017-05-06 22:16:52.778220093 +0200 -@@ -0,0 +1,426 @@ ++++ linux/fs/aufs/fhsm.c 2019-03-05 12:13:00.139224339 +0100 +@@ -0,0 +1,427 @@ ++// SPDX-License-Identifier: GPL-2.0 +/* -+ * Copyright (C) 2011-2017 Junjiro R. Okajima ++ * Copyright (C) 2011-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -12055,10 +13947,9 @@ diff -urN /usr/share/empty/fs/aufs/fhsm.c linux/fs/aufs/fhsm.c + +/* ---------------------------------------------------------------------- */ + -+static unsigned int au_fhsm_poll(struct file *file, -+ struct poll_table_struct *wait) ++static __poll_t au_fhsm_poll(struct file *file, struct poll_table_struct *wait) +{ -+ unsigned int mask; ++ __poll_t mask; + struct au_sbinfo *sbinfo; + struct au_fhsm *fhsm; + @@ -12067,9 +13958,10 @@ diff -urN /usr/share/empty/fs/aufs/fhsm.c linux/fs/aufs/fhsm.c + fhsm = &sbinfo->si_fhsm; + poll_wait(file, &fhsm->fhsm_wqh, wait); + if (atomic_read(&fhsm->fhsm_readable)) -+ mask = POLLIN /* | POLLRDNORM */; ++ mask = EPOLLIN /* | EPOLLRDNORM */; + -+ AuTraceErr((int)mask); ++ if (!mask) ++ AuDbg("mask 0x%x\n", mask); + return mask; +} + @@ -12240,7 +14132,7 @@ diff -urN /usr/share/empty/fs/aufs/fhsm.c linux/fs/aufs/fhsm.c + if (unlikely(fd < 0)) + goto out_pid; + -+ /* succeed reglardless 'fhsm' status */ ++ /* succeed regardless 'fhsm' status */ + kobject_get(&sbinfo->si_kobj); + si_noflush_read_lock(sb); + if (au_ftest_si(sbinfo, FHSM)) @@ -12312,10 +14204,11 @@ diff -urN /usr/share/empty/fs/aufs/fhsm.c linux/fs/aufs/fhsm.c +} diff -urN /usr/share/empty/fs/aufs/file.c linux/fs/aufs/file.c --- /usr/share/empty/fs/aufs/file.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/file.c 2017-05-06 22:16:52.778220093 +0200 -@@ -0,0 +1,860 @@ ++++ linux/fs/aufs/file.c 2019-03-05 12:13:00.139224339 +0100 +@@ -0,0 +1,863 @@ ++// SPDX-License-Identifier: GPL-2.0 +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -12398,7 +14291,7 @@ diff -urN /usr/share/empty/fs/aufs/file.c linux/fs/aufs/file.c + } + } + flags &= ~O_CREAT; -+ au_br_get(br); ++ au_lcnt_inc(&br->br_nfiles); + h_path.dentry = h_dentry; + h_path.mnt = au_br_mnt(br); + h_file = vfsub_dentry_open(&h_path, flags); @@ -12417,14 +14310,14 @@ diff -urN /usr/share/empty/fs/aufs/file.c linux/fs/aufs/file.c + goto out; /* success */ + +out_br: -+ au_br_put(br); ++ au_lcnt_dec(&br->br_nfiles); +out: + return h_file; +} + +static int au_cmoo(struct dentry *dentry) +{ -+ int err, cmoo; ++ int err, cmoo, matched; + unsigned int udba; + struct path h_path; + struct au_pin pin; @@ -12459,9 +14352,12 @@ diff -urN /usr/share/empty/fs/aufs/file.c linux/fs/aufs/file.c + sbinfo = au_sbi(sb); + fhsm = &sbinfo->si_fhsm; + pid = au_fhsm_pid(fhsm); -+ if (pid -+ && (current->pid == pid -+ || current->real_parent->pid == pid)) ++ rcu_read_lock(); ++ matched = (pid ++ && (current->pid == pid ++ || rcu_dereference(current->real_parent)->pid == pid)); ++ rcu_read_unlock(); ++ if (matched) + goto out; + + br = au_sbr(sb, cpg.bsrc); @@ -12538,11 +14434,11 @@ diff -urN /usr/share/empty/fs/aufs/file.c linux/fs/aufs/file.c + +int au_do_open(struct file *file, struct au_do_open_args *args) +{ -+ int err, no_lock = args->no_lock; ++ int err, aopen = args->aopen; + struct dentry *dentry; + struct au_finfo *finfo; + -+ if (!no_lock) ++ if (!aopen) + err = au_finfo_init(file, args->fidir); + else { + lockdep_off(); @@ -12554,33 +14450,28 @@ diff -urN /usr/share/empty/fs/aufs/file.c linux/fs/aufs/file.c + + dentry = file->f_path.dentry; + AuDebugOn(IS_ERR_OR_NULL(dentry)); -+ if (!no_lock) { -+ di_write_lock_child(dentry); -+ err = au_cmoo(dentry); -+ di_downgrade_lock(dentry, AuLock_IR); -+ if (!err) ++ di_write_lock_child(dentry); ++ err = au_cmoo(dentry); ++ di_downgrade_lock(dentry, AuLock_IR); ++ if (!err) { ++ if (!aopen) + err = args->open(file, vfsub_file_flags(file), NULL); -+ di_read_unlock(dentry, AuLock_IR); -+ } else { -+ err = au_cmoo(dentry); -+ if (!err) ++ else { ++ lockdep_off(); + err = args->open(file, vfsub_file_flags(file), + args->h_file); -+ if (!err && au_fbtop(file) != au_dbtop(dentry)) -+ /* -+ * cmoo happens after h_file was opened. -+ * need to refresh file later. -+ */ -+ atomic_dec(&au_fi(file)->fi_generation); ++ lockdep_on(); ++ } + } ++ di_read_unlock(dentry, AuLock_IR); + + finfo = au_fi(file); + if (!err) { + finfo->fi_file = file; -+ au_sphl_add(&finfo->fi_hlist, -+ &au_sbi(file->f_path.dentry->d_sb)->si_files); ++ au_hbl_add(&finfo->fi_hlist, ++ &au_sbi(file->f_path.dentry->d_sb)->si_files); + } -+ if (!no_lock) ++ if (!aopen) + fi_write_unlock(file); + else { + lockdep_off(); @@ -12589,10 +14480,11 @@ diff -urN /usr/share/empty/fs/aufs/file.c linux/fs/aufs/file.c + } + if (unlikely(err)) { + finfo->fi_hdir = NULL; -+ au_finfo_fin(file, /*atonce*/0); ++ au_finfo_fin(file); + } + +out: ++ AuTraceErr(err); + return err; +} + @@ -12601,10 +14493,12 @@ diff -urN /usr/share/empty/fs/aufs/file.c linux/fs/aufs/file.c + int err; + aufs_bindex_t btop; + struct dentry *dentry; ++ struct au_branch *br; + struct file *h_file, *h_file_tmp; + + dentry = file->f_path.dentry; + btop = au_dbtop(dentry); ++ br = au_sbr(dentry->d_sb, btop); + h_file_tmp = NULL; + if (au_fbtop(file) == btop) { + h_file = au_hf_top(file); @@ -12612,6 +14506,7 @@ diff -urN /usr/share/empty/fs/aufs/file.c linux/fs/aufs/file.c + return 0; /* success */ + h_file_tmp = h_file; + get_file(h_file_tmp); ++ au_lcnt_inc(&br->br_nfiles); + au_set_h_fptr(file, btop, NULL); + } + AuDebugOn(au_fi(file)->fi_hdir); @@ -12632,7 +14527,7 @@ diff -urN /usr/share/empty/fs/aufs/file.c linux/fs/aufs/file.c + err = PTR_ERR(h_file); + if (IS_ERR(h_file)) { + if (h_file_tmp) { -+ au_sbr_get(dentry->d_sb, btop); ++ /* revert */ + au_set_h_fptr(file, btop, h_file_tmp); + h_file_tmp = NULL; + } @@ -12647,8 +14542,10 @@ diff -urN /usr/share/empty/fs/aufs/file.c linux/fs/aufs/file.c + /* file->f_ra = h_file->f_ra; */ + +out: -+ if (h_file_tmp) ++ if (h_file_tmp) { + fput(h_file_tmp); ++ au_lcnt_dec(&br->br_nfiles); ++ } + return err; +} + @@ -12909,7 +14806,6 @@ diff -urN /usr/share/empty/fs/aufs/file.c linux/fs/aufs/file.c + +static void au_do_refresh_dir(struct file *file) +{ -+ int execed; + aufs_bindex_t bindex, bbot, new_bindex, brid; + struct au_hfile *p, tmp, *q; + struct au_finfo *finfo; @@ -12948,7 +14844,6 @@ diff -urN /usr/share/empty/fs/aufs/file.c linux/fs/aufs/file.c + } + } + -+ execed = vfsub_file_execed(file); + p = fidir->fd_hfile; + if (!au_test_mmapped(file) && !d_unlinked(file->f_path.dentry)) { + bbot = au_sbbot(sb); @@ -12957,14 +14852,14 @@ diff -urN /usr/share/empty/fs/aufs/file.c linux/fs/aufs/file.c + if (p->hf_file) { + if (file_inode(p->hf_file)) + break; -+ au_hfput(p, execed); ++ au_hfput(p, /*execed*/0); + } + } else { + bbot = au_br_index(sb, brid); + for (finfo->fi_btop = 0; finfo->fi_btop < bbot; + finfo->fi_btop++, p++) + if (p->hf_file) -+ au_hfput(p, execed); ++ au_hfput(p, /*execed*/0); + bbot = au_sbbot(sb); + } + @@ -12974,7 +14869,7 @@ diff -urN /usr/share/empty/fs/aufs/file.c linux/fs/aufs/file.c + if (p->hf_file) { + if (file_inode(p->hf_file)) + break; -+ au_hfput(p, execed); ++ au_hfput(p, /*execed*/0); + } + AuDebugOn(fidir->fd_bbot < finfo->fi_btop); +} @@ -13176,10 +15071,11 @@ diff -urN /usr/share/empty/fs/aufs/file.c linux/fs/aufs/file.c +}; diff -urN /usr/share/empty/fs/aufs/file.h linux/fs/aufs/file.h --- /usr/share/empty/fs/aufs/file.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/file.h 2017-05-06 22:16:52.778220093 +0200 -@@ -0,0 +1,333 @@ ++++ linux/fs/aufs/file.h 2019-03-05 12:13:00.139224339 +0100 +@@ -0,0 +1,342 @@ ++/* SPDX-License-Identifier: GPL-2.0 */ +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -13206,6 +15102,7 @@ diff -urN /usr/share/empty/fs/aufs/file.h linux/fs/aufs/file.h + +#include +#include ++#include +#include +#include "rwsem.h" + @@ -13242,11 +15139,9 @@ diff -urN /usr/share/empty/fs/aufs/file.h linux/fs/aufs/file.h + }; + struct au_fidir *fi_hdir; /* for dir only */ + -+ struct hlist_node fi_hlist; -+ union { -+ struct file *fi_file; /* very ugly */ -+ struct llist_node fi_lnode; /* delayed free */ -+ }; ++ struct hlist_bl_node fi_hlist; ++ struct file *fi_file; /* very ugly */ ++ struct rcu_head rcu; +} ____cacheline_aligned_in_smp; + +/* ---------------------------------------------------------------------- */ @@ -13257,7 +15152,7 @@ diff -urN /usr/share/empty/fs/aufs/file.h linux/fs/aufs/file.h +struct file *au_h_open(struct dentry *dentry, aufs_bindex_t bindex, int flags, + struct file *file, int force_wr); +struct au_do_open_args { -+ int no_lock; ++ int aopen; + int (*open)(struct file *file, int flags, + struct file *h_file); + struct au_fidir *fidir; @@ -13274,7 +15169,7 @@ diff -urN /usr/share/empty/fs/aufs/file.h linux/fs/aufs/file.h + +/* poll.c */ +#ifdef CONFIG_AUFS_POLL -+unsigned int aufs_poll(struct file *file, poll_table *wait); ++__poll_t aufs_poll(struct file *file, struct poll_table_struct *pt); +#endif + +#ifdef CONFIG_AUFS_BR_HFSPLUS @@ -13306,7 +15201,7 @@ diff -urN /usr/share/empty/fs/aufs/file.h linux/fs/aufs/file.h +int au_fidir_realloc(struct au_finfo *finfo, int nbr, int may_shrink); + +void au_fi_init_once(void *_fi); -+void au_finfo_fin(struct file *file, int atonce); ++void au_finfo_fin(struct file *file); +int au_finfo_init(struct file *file, struct au_fidir *fidir); + +/* ioctl.c */ @@ -13327,11 +15222,20 @@ diff -urN /usr/share/empty/fs/aufs/file.h linux/fs/aufs/file.h + +/* ---------------------------------------------------------------------- */ + ++#define fi_read_lock(f) au_rw_read_lock(&au_fi(f)->fi_rwsem) ++#define fi_write_lock(f) au_rw_write_lock(&au_fi(f)->fi_rwsem) ++#define fi_read_trylock(f) au_rw_read_trylock(&au_fi(f)->fi_rwsem) ++#define fi_write_trylock(f) au_rw_write_trylock(&au_fi(f)->fi_rwsem) +/* -+ * fi_read_lock, fi_write_lock, -+ * fi_read_unlock, fi_write_unlock, fi_downgrade_lock -+ */ -+AuSimpleRwsemFuncs(fi, struct file *f, &au_fi(f)->fi_rwsem); ++#define fi_read_trylock_nested(f) \ ++ au_rw_read_trylock_nested(&au_fi(f)->fi_rwsem) ++#define fi_write_trylock_nested(f) \ ++ au_rw_write_trylock_nested(&au_fi(f)->fi_rwsem) ++*/ ++ ++#define fi_read_unlock(f) au_rw_read_unlock(&au_fi(f)->fi_rwsem) ++#define fi_write_unlock(f) au_rw_write_unlock(&au_fi(f)->fi_rwsem) ++#define fi_downgrade_lock(f) au_rw_dgrade_lock(&au_fi(f)->fi_rwsem) + +/* lock subclass for finfo */ +enum { @@ -13513,10 +15417,11 @@ diff -urN /usr/share/empty/fs/aufs/file.h linux/fs/aufs/file.h +#endif /* __AUFS_FILE_H__ */ diff -urN /usr/share/empty/fs/aufs/finfo.c linux/fs/aufs/finfo.c --- /usr/share/empty/fs/aufs/finfo.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/finfo.c 2017-05-06 22:16:52.778220093 +0200 -@@ -0,0 +1,151 @@ ++++ linux/fs/aufs/finfo.c 2019-03-05 12:13:00.139224339 +0100 +@@ -0,0 +1,149 @@ ++// SPDX-License-Identifier: GPL-2.0 +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -13544,7 +15449,7 @@ diff -urN /usr/share/empty/fs/aufs/finfo.c linux/fs/aufs/finfo.c + allow_write_access(hf->hf_file); + fput(hf->hf_file); + hf->hf_file = NULL; -+ au_br_put(hf->hf_br); ++ au_lcnt_dec(&hf->hf_br->br_nfiles); + hf->hf_br = NULL; +} + @@ -13619,19 +15524,16 @@ diff -urN /usr/share/empty/fs/aufs/finfo.c linux/fs/aufs/finfo.c + +/* ---------------------------------------------------------------------- */ + -+void au_finfo_fin(struct file *file, int atonce) ++void au_finfo_fin(struct file *file) +{ + struct au_finfo *finfo; + -+ au_nfiles_dec(file->f_path.dentry->d_sb); ++ au_lcnt_dec(&au_sbi(file->f_path.dentry->d_sb)->si_nfiles); + + finfo = au_fi(file); + AuDebugOn(finfo->fi_hdir); + AuRwDestroy(&finfo->fi_rwsem); -+ if (!atonce) -+ au_cache_dfree_finfo(finfo); -+ else -+ au_cache_free_finfo(finfo); ++ au_cache_free_finfo(finfo); +} + +void au_fi_init_once(void *_finfo) @@ -13654,7 +15556,7 @@ diff -urN /usr/share/empty/fs/aufs/finfo.c linux/fs/aufs/finfo.c + goto out; + + err = 0; -+ au_nfiles_inc(dentry->d_sb); ++ au_lcnt_inc(&au_sbi(dentry->d_sb)->si_nfiles); + au_rw_write_lock(&finfo->fi_rwsem); + finfo->fi_btop = -1; + finfo->fi_hdir = fidir; @@ -13668,10 +15570,11 @@ diff -urN /usr/share/empty/fs/aufs/finfo.c linux/fs/aufs/finfo.c +} diff -urN /usr/share/empty/fs/aufs/f_op.c linux/fs/aufs/f_op.c --- /usr/share/empty/fs/aufs/f_op.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/f_op.c 2017-05-06 22:16:52.778220093 +0200 ++++ linux/fs/aufs/f_op.c 2019-03-05 12:13:00.139224339 +0100 @@ -0,0 +1,819 @@ ++// SPDX-License-Identifier: GPL-2.0 +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -13720,29 +15623,30 @@ diff -urN /usr/share/empty/fs/aufs/f_op.c linux/fs/aufs/f_op.c + if (unlikely(err)) + goto out; + h_file = au_h_open(dentry, bindex, flags, file, /*force_wr*/0); ++ if (IS_ERR(h_file)) { ++ err = PTR_ERR(h_file); ++ goto out; ++ } + } else { + h_dentry = h_file->f_path.dentry; + err = vfsub_test_mntns(file->f_path.mnt, h_dentry->d_sb); + if (unlikely(err)) + goto out; -+ get_file(h_file); ++ /* br ref is already inc-ed */ + } -+ if (IS_ERR(h_file)) -+ err = PTR_ERR(h_file); -+ else { -+ if ((flags & __O_TMPFILE) -+ && !(flags & O_EXCL)) { -+ h_inode = file_inode(h_file); -+ spin_lock(&h_inode->i_lock); -+ h_inode->i_state |= I_LINKABLE; -+ spin_unlock(&h_inode->i_lock); -+ } -+ au_set_fbtop(file, bindex); -+ au_set_h_fptr(file, bindex, h_file); -+ au_update_figen(file); -+ /* todo: necessary? */ -+ /* file->f_ra = h_file->f_ra; */ ++ ++ if ((flags & __O_TMPFILE) ++ && !(flags & O_EXCL)) { ++ h_inode = file_inode(h_file); ++ spin_lock(&h_inode->i_lock); ++ h_inode->i_state |= I_LINKABLE; ++ spin_unlock(&h_inode->i_lock); + } ++ au_set_fbtop(file, bindex); ++ au_set_h_fptr(file, bindex, h_file); ++ au_update_figen(file); ++ /* todo: necessary? */ ++ /* file->f_ra = h_file->f_ra; */ + +out: + return err; @@ -13771,17 +15675,15 @@ diff -urN /usr/share/empty/fs/aufs/f_op.c linux/fs/aufs/f_op.c +{ + struct au_finfo *finfo; + aufs_bindex_t bindex; -+ int delayed; + + finfo = au_fi(file); -+ au_sphl_del(&finfo->fi_hlist, -+ &au_sbi(file->f_path.dentry->d_sb)->si_files); ++ au_hbl_del(&finfo->fi_hlist, ++ &au_sbi(file->f_path.dentry->d_sb)->si_files); + bindex = finfo->fi_btop; + if (bindex >= 0) + au_set_h_fptr(file, bindex, NULL); + -+ delayed = (current->flags & PF_KTHREAD) || in_interrupt(); -+ au_finfo_fin(file, delayed); ++ au_finfo_fin(file); + return 0; +} + @@ -13808,7 +15710,7 @@ diff -urN /usr/share/empty/fs/aufs/f_op.c linux/fs/aufs/f_op.c +/* + * read and write functions acquire [fdi]_rwsem once, but release before + * mmap_sem. This is because to stop a race condition between mmap(2). -+ * Releasing these aufs-rwsem should be safe, no branch-mamagement (by keeping ++ * Releasing these aufs-rwsem should be safe, no branch-management (by keeping + * si_rwsem), no harmful copy-up should happen. Actually copy-up may happen in + * read functions after [fdi]_rwsem are released, but it should be harmless. + */ @@ -13907,12 +15809,11 @@ diff -urN /usr/share/empty/fs/aufs/f_op.c linux/fs/aufs/f_op.c + h_inode = file_inode(h_file); + inode->i_mode = h_inode->i_mode; + ii_write_unlock(inode); -+ fput(h_file); -+ + /* AuDbg("blks %llu, %llu\n", (u64)blks, (u64)h_inode->i_blocks); */ + if (written > 0) + au_fhsm_wrote(inode->i_sb, wpre->btop, + /*force*/h_inode->i_blocks > wpre->blks); ++ fput(h_file); +} + +static ssize_t aufs_read(struct file *file, char __user *buf, size_t count, @@ -14106,7 +16007,7 @@ diff -urN /usr/share/empty/fs/aufs/f_op.c linux/fs/aufs/f_op.c + goto out; + + err = vfsub_splice_to(h_file, ppos, pipe, len, flags); -+ /* todo: necessasry? */ ++ /* todo: necessary? */ + /* file->f_ra = h_file->f_ra; */ + au_read_post(inode, h_file); + @@ -14196,7 +16097,7 @@ diff -urN /usr/share/empty/fs/aufs/f_op.c linux/fs/aufs/f_op.c + au_mtx_and_read_lock(a_dst.inode); + /* + * in order to match the order in di_write_lock2_{child,parent}(), -+ * use f_path.dentry for this comparision. ++ * use f_path.dentry for this comparison. + */ + if (src->f_path.dentry < dst->f_path.dentry) { + a_src.h_file = au_read_pre(src, /*keep_fi*/1, AuLsc_FI_1); @@ -14262,9 +16163,9 @@ diff -urN /usr/share/empty/fs/aufs/f_op.c linux/fs/aufs/f_op.c + * (Note that [fdi]i_rwsem are released before mmap_sem). + * - in mmap case + * mmap(2) -- mmap_sem -- aufs_mmap() -- si_rwsem for read -- [fdi]i_rwsem -+ * This AB-BA order is definitly bad, but is not a problem since "si_rwsem for -+ * read" allows muliple processes to acquire it and [fdi]i_rwsem are not held in -+ * file I/O. Aufs needs to stop lockdep in aufs_mmap() though. ++ * This AB-BA order is definitely bad, but is not a problem since "si_rwsem for ++ * read" allows multiple processes to acquire it and [fdi]i_rwsem are not held ++ * in file I/O. Aufs needs to stop lockdep in aufs_mmap() though. + * It means that when aufs acquires si_rwsem for write, the process should never + * acquire mmap_sem. + * @@ -14434,7 +16335,8 @@ diff -urN /usr/share/empty/fs/aufs/f_op.c linux/fs/aufs/f_op.c + if (IS_ERR(h_file)) + goto out; + -+ arg |= vfsub_file_flags(file) & FASYNC; /* stop calling h_file->fasync */ ++ /* stop calling h_file->fasync */ ++ arg |= vfsub_file_flags(file) & FASYNC; + err = setfl(/*unused fd*/-1, h_file, arg); + fput(h_file); /* instead of au_read_post() */ + @@ -14491,10 +16393,11 @@ diff -urN /usr/share/empty/fs/aufs/f_op.c linux/fs/aufs/f_op.c +}; diff -urN /usr/share/empty/fs/aufs/fstype.h linux/fs/aufs/fstype.h --- /usr/share/empty/fs/aufs/fstype.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/fstype.h 2017-05-06 22:16:52.778220093 +0200 -@@ -0,0 +1,400 @@ ++++ linux/fs/aufs/fstype.h 2019-03-05 12:13:00.139224339 +0100 +@@ -0,0 +1,401 @@ ++/* SPDX-License-Identifier: GPL-2.0 */ +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -14882,7 +16785,7 @@ diff -urN /usr/share/empty/fs/aufs/fstype.h linux/fs/aufs/fstype.h + +/* + * test if the @inode is nfs with 'noacl' option -+ * NFS always sets MS_POSIXACL regardless its mount option 'noacl.' ++ * NFS always sets SB_POSIXACL regardless its mount option 'noacl.' + */ +static inline int au_test_nfs_noacl(struct inode *inode) +{ @@ -14893,12 +16796,82 @@ diff -urN /usr/share/empty/fs/aufs/fstype.h linux/fs/aufs/fstype.h + +#endif /* __KERNEL__ */ +#endif /* __AUFS_FSTYPE_H__ */ +diff -urN /usr/share/empty/fs/aufs/hbl.h linux/fs/aufs/hbl.h +--- /usr/share/empty/fs/aufs/hbl.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux/fs/aufs/hbl.h 2019-03-05 12:13:00.139224339 +0100 +@@ -0,0 +1,65 @@ ++/* SPDX-License-Identifier: GPL-2.0 */ ++/* ++ * Copyright (C) 2017-2019 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see . ++ */ ++ ++/* ++ * helpers for hlist_bl.h ++ */ ++ ++#ifndef __AUFS_HBL_H__ ++#define __AUFS_HBL_H__ ++ ++#ifdef __KERNEL__ ++ ++#include ++ ++static inline void au_hbl_add(struct hlist_bl_node *node, ++ struct hlist_bl_head *hbl) ++{ ++ hlist_bl_lock(hbl); ++ hlist_bl_add_head(node, hbl); ++ hlist_bl_unlock(hbl); ++} ++ ++static inline void au_hbl_del(struct hlist_bl_node *node, ++ struct hlist_bl_head *hbl) ++{ ++ hlist_bl_lock(hbl); ++ hlist_bl_del(node); ++ hlist_bl_unlock(hbl); ++} ++ ++#define au_hbl_for_each(pos, head) \ ++ for (pos = hlist_bl_first(head); \ ++ pos; \ ++ pos = pos->next) ++ ++static inline unsigned long au_hbl_count(struct hlist_bl_head *hbl) ++{ ++ unsigned long cnt; ++ struct hlist_bl_node *pos; ++ ++ cnt = 0; ++ hlist_bl_lock(hbl); ++ au_hbl_for_each(pos, hbl) ++ cnt++; ++ hlist_bl_unlock(hbl); ++ return cnt; ++} ++ ++#endif /* __KERNEL__ */ ++#endif /* __AUFS_HBL_H__ */ diff -urN /usr/share/empty/fs/aufs/hfsnotify.c linux/fs/aufs/hfsnotify.c --- /usr/share/empty/fs/aufs/hfsnotify.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/hfsnotify.c 2017-05-06 22:16:52.778220093 +0200 -@@ -0,0 +1,287 @@ ++++ linux/fs/aufs/hfsnotify.c 2019-03-05 12:13:00.139224339 +0100 +@@ -0,0 +1,289 @@ ++// SPDX-License-Identifier: GPL-2.0 +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -14931,8 +16904,8 @@ diff -urN /usr/share/empty/fs/aufs/hfsnotify.c linux/fs/aufs/hfsnotify.c + struct au_hnotify *hn = container_of(mark, struct au_hnotify, + hn_mark); + /* AuDbg("here\n"); */ -+ au_cache_dfree_hnotify(hn); -+ smp_mb__before_atomic(); ++ au_cache_free_hnotify(hn); ++ smp_mb__before_atomic(); /* for atomic64_dec */ + if (atomic64_dec_and_test(&au_hfsn_ifree)) + wake_up(&au_hfsn_wq); +} @@ -14953,15 +16926,14 @@ diff -urN /usr/share/empty/fs/aufs/hfsnotify.c linux/fs/aufs/hfsnotify.c + AuDebugOn(!br->br_hfsn); + + mark = &hn->hn_mark; -+ fsnotify_init_mark(mark, au_hfsn_free_mark); ++ fsnotify_init_mark(mark, br->br_hfsn->hfsn_group); + mark->mask = AuHfsnMask; + /* + * by udba rename or rmdir, aufs assign a new inode to the known + * h_inode, so specify 1 to allow dups. + */ + lockdep_off(); -+ err = fsnotify_add_mark(mark, br->br_hfsn->hfsn_group, hinode->hi_inode, -+ /*mnt*/NULL, /*allow_dups*/1); ++ err = fsnotify_add_inode_mark(mark, hinode->hi_inode, /*allow_dups*/1); + lockdep_on(); + + return err; @@ -15055,20 +17027,20 @@ diff -urN /usr/share/empty/fs/aufs/hfsnotify.c linux/fs/aufs/hfsnotify.c + struct au_br_hfsnotify *hfsn = group->private; + + /* AuDbg("here\n"); */ -+ au_delayed_kfree(hfsn); ++ au_kfree_try_rcu(hfsn); +} + +static int au_hfsn_handle_event(struct fsnotify_group *group, + struct inode *inode, -+ struct fsnotify_mark *inode_mark, -+ struct fsnotify_mark *vfsmount_mark, + u32 mask, const void *data, int data_type, -+ const unsigned char *file_name, u32 cookie) ++ const struct qstr *file_name, u32 cookie, ++ struct fsnotify_iter_info *iter_info) +{ + int err; + struct au_hnotify *hnotify; + struct inode *h_dir, *h_inode; -+ struct qstr h_child_qstr = QSTR_INIT(file_name, strlen(file_name)); ++ ++ struct fsnotify_mark *inode_mark; + + AuDebugOn(data_type != FSNOTIFY_EVENT_INODE); + @@ -15082,19 +17054,20 @@ diff -urN /usr/share/empty/fs/aufs/hfsnotify.c linux/fs/aufs/hfsnotify.c + h_inode = NULL; +#ifdef AuDbgHnotify + au_debug_on(); -+ if (1 || h_child_qstr.len != sizeof(AUFS_XINO_FNAME) - 1 -+ || strncmp(h_child_qstr.name, AUFS_XINO_FNAME, h_child_qstr.len)) { ++ if (1 || file_name.len != sizeof(AUFS_XINO_FNAME) - 1 ++ || strncmp(file_name.name, AUFS_XINO_FNAME, file_name.len)) { + AuDbg("i%lu, mask 0x%x %s, hcname %.*s, hi%lu\n", + h_dir->i_ino, mask, au_hfsn_name(mask), -+ AuLNPair(&h_child_qstr), h_inode ? h_inode->i_ino : 0); ++ AuLNPair(file_name), h_inode ? h_inode->i_ino : 0); + /* WARN_ON(1); */ + } + au_debug_off(); +#endif + ++ inode_mark = fsnotify_iter_inode_mark(iter_info); + AuDebugOn(!inode_mark); + hnotify = container_of(inode_mark, struct au_hnotify, hn_mark); -+ err = au_hnotify(h_dir, hnotify, mask, &h_child_qstr, h_inode); ++ err = au_hnotify(h_dir, hnotify, mask, file_name, h_inode); + +out: + return err; @@ -15102,7 +17075,8 @@ diff -urN /usr/share/empty/fs/aufs/hfsnotify.c linux/fs/aufs/hfsnotify.c + +static struct fsnotify_ops au_hfsn_ops = { + .handle_event = au_hfsn_handle_event, -+ .free_group_priv = au_hfsn_free_group ++ .free_group_priv = au_hfsn_free_group, ++ .free_mark = au_hfsn_free_mark +}; + +/* ---------------------------------------------------------------------- */ @@ -15149,7 +17123,7 @@ diff -urN /usr/share/empty/fs/aufs/hfsnotify.c linux/fs/aufs/hfsnotify.c + goto out; /* success */ + +out_hfsn: -+ au_delayed_kfree(hfsn); ++ au_kfree_try_rcu(hfsn); +out: + return err; +} @@ -15186,10 +17160,11 @@ diff -urN /usr/share/empty/fs/aufs/hfsnotify.c linux/fs/aufs/hfsnotify.c +}; diff -urN /usr/share/empty/fs/aufs/hfsplus.c linux/fs/aufs/hfsplus.c --- /usr/share/empty/fs/aufs/hfsplus.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/hfsplus.c 2017-05-06 22:16:52.778220093 +0200 -@@ -0,0 +1,56 @@ ++++ linux/fs/aufs/hfsplus.c 2019-03-05 12:13:00.139224339 +0100 +@@ -0,0 +1,60 @@ ++// SPDX-License-Identifier: GPL-2.0 +/* -+ * Copyright (C) 2010-2017 Junjiro R. Okajima ++ * Copyright (C) 2010-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -15206,11 +17181,11 @@ diff -urN /usr/share/empty/fs/aufs/hfsplus.c linux/fs/aufs/hfsplus.c + */ + +/* -+ * special support for filesystems which aqucires an inode mutex ++ * special support for filesystems which acquires an inode mutex + * at final closing a file, eg, hfsplus. + * + * This trick is very simple and stupid, just to open the file before really -+ * neceeary open to tell hfsplus that this is not the final closing. ++ * necessary open to tell hfsplus that this is not the final closing. + * The caller should call au_h_open_pre() after acquiring the inode mutex, + * and au_h_open_post() after releasing it. + */ @@ -15239,17 +17214,21 @@ diff -urN /usr/share/empty/fs/aufs/hfsplus.c linux/fs/aufs/hfsplus.c +void au_h_open_post(struct dentry *dentry, aufs_bindex_t bindex, + struct file *h_file) +{ ++ struct au_branch *br; ++ + if (h_file) { + fput(h_file); -+ au_sbr_put(dentry->d_sb, bindex); ++ br = au_sbr(dentry->d_sb, bindex); ++ au_lcnt_dec(&br->br_nfiles); + } +} diff -urN /usr/share/empty/fs/aufs/hnotify.c linux/fs/aufs/hnotify.c --- /usr/share/empty/fs/aufs/hnotify.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/hnotify.c 2017-05-06 22:16:52.778220093 +0200 -@@ -0,0 +1,723 @@ ++++ linux/fs/aufs/hnotify.c 2019-05-06 09:03:04.816810402 +0200 +@@ -0,0 +1,721 @@ ++// SPDX-License-Identifier: GPL-2.0 +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -15269,6 +17248,7 @@ diff -urN /usr/share/empty/fs/aufs/hnotify.c linux/fs/aufs/hnotify.c + * abstraction to notify the direct changes on lower directories + */ + ++/* #include */ +#include "aufs.h" + +int au_hn_alloc(struct au_hinode *hinode, struct inode *inode) @@ -15285,12 +17265,12 @@ diff -urN /usr/share/empty/fs/aufs/hnotify.c linux/fs/aufs/hnotify.c + AuTraceErr(err); + if (unlikely(err)) { + hinode->hi_notify = NULL; -+ au_cache_dfree_hnotify(hn); ++ au_cache_free_hnotify(hn); + /* + * The upper dir was removed by udba, but the same named -+ * dir left. In this case, aufs assignes a new inode ++ * dir left. In this case, aufs assigns a new inode + * number and set the monitor again. -+ * For the lower dir, the old monitnor is still left. ++ * For the lower dir, the old monitor is still left. + */ + if (err == -EEXIST) + err = 0; @@ -15309,7 +17289,7 @@ diff -urN /usr/share/empty/fs/aufs/hnotify.c linux/fs/aufs/hnotify.c + if (hn) { + hinode->hi_notify = NULL; + if (au_hnotify_op.free(hinode, hn)) -+ au_cache_dfree_hnotify(hn); ++ au_cache_free_hnotify(hn); + } +} + @@ -15572,11 +17552,11 @@ diff -urN /usr/share/empty/fs/aufs/hnotify.c linux/fs/aufs/hnotify.c + if (au_ftest_hnjob(a->flags, TRYXINO0) + && a->inode + && a->h_inode) { -+ inode_lock_nested(a->h_inode, AuLsc_I_CHILD); ++ inode_lock_shared_nested(a->h_inode, AuLsc_I_CHILD); + if (!a->h_inode->i_nlink + && !(a->h_inode->i_state & I_LINKABLE)) + hn_xino(a->inode, a->h_inode); /* ignore this error */ -+ inode_unlock(a->h_inode); ++ inode_unlock_shared(a->h_inode); + } + + /* make the generation obsolete */ @@ -15598,7 +17578,7 @@ diff -urN /usr/share/empty/fs/aufs/hnotify.c linux/fs/aufs/hnotify.c + if (vdir) + vdir->vd_jiffy = 0; + /* IMustLock(a->inode); */ -+ /* a->inode->i_version++; */ ++ /* inode_inc_iversion(a->inode); */ + } + + /* can do nothing but warn */ @@ -15712,6 +17692,14 @@ diff -urN /usr/share/empty/fs/aufs/hnotify.c linux/fs/aufs/hnotify.c + AuDebugOn(!sbinfo); + si_write_lock(sb, AuLock_NOPLMW); + ++ if (au_opt_test(sbinfo->si_mntflags, DIRREN)) ++ switch (a->mask & FS_EVENTS_POSS_ON_CHILD) { ++ case FS_MOVED_FROM: ++ case FS_MOVED_TO: ++ AuWarn1("DIRREN with UDBA may not work correctly " ++ "for the direct rename(2)\n"); ++ } ++ + ii_read_lock_parent(a->dir); + bfound = -1; + bbot = au_ibbot(a->dir); @@ -15782,13 +17770,13 @@ diff -urN /usr/share/empty/fs/aufs/hnotify.c linux/fs/aufs/hnotify.c + iput(a->dir); + si_write_unlock(sb); + au_nwt_done(&sbinfo->si_nowait); -+ au_delayed_kfree(a); ++ au_kfree_rcu(a); +} + +/* ---------------------------------------------------------------------- */ + +int au_hnotify(struct inode *h_dir, struct au_hnotify *hnotify, u32 mask, -+ struct qstr *h_child_qstr, struct inode *h_child_inode) ++ const struct qstr *h_child_qstr, struct inode *h_child_inode) +{ + int err, len; + unsigned int flags[AuHnLast], f; @@ -15888,7 +17876,7 @@ diff -urN /usr/share/empty/fs/aufs/hnotify.c linux/fs/aufs/hnotify.c + iput(args->h_child_inode); + iput(args->h_dir); + iput(args->dir); -+ au_delayed_kfree(args); ++ au_kfree_rcu(args); + } + +out: @@ -15929,26 +17917,17 @@ diff -urN /usr/share/empty/fs/aufs/hnotify.c linux/fs/aufs/hnotify.c + +static void au_hn_destroy_cache(void) +{ -+ struct au_cache *cp; -+ -+ flush_delayed_work(&au_dfree.dwork); -+ cp = au_dfree.cache + AuCache_HNOTIFY; -+ AuDebugOn(!llist_empty(&cp->llist)); -+ kmem_cache_destroy(cp->cache); -+ cp->cache = NULL; ++ kmem_cache_destroy(au_cache[AuCache_HNOTIFY]); ++ au_cache[AuCache_HNOTIFY] = NULL; +} + -+AU_CACHE_DFREE_FUNC(hnotify, HNOTIFY, hn_lnode); -+ +int __init au_hnotify_init(void) +{ + int err; -+ struct au_cache *cp; + + err = -ENOMEM; -+ cp = au_dfree.cache + AuCache_HNOTIFY; -+ cp->cache = AuCache(au_hnotify); -+ if (cp->cache) { ++ au_cache[AuCache_HNOTIFY] = AuCache(au_hnotify); ++ if (au_cache[AuCache_HNOTIFY]) { + err = 0; + if (au_hnotify_op.init) + err = au_hnotify_op.init(); @@ -15961,22 +17940,20 @@ diff -urN /usr/share/empty/fs/aufs/hnotify.c linux/fs/aufs/hnotify.c + +void au_hnotify_fin(void) +{ -+ struct au_cache *cp; -+ + if (au_hnotify_op.fin) + au_hnotify_op.fin(); + + /* cf. au_cache_fin() */ -+ cp = au_dfree.cache + AuCache_HNOTIFY; -+ if (cp->cache) ++ if (au_cache[AuCache_HNOTIFY]) + au_hn_destroy_cache(); +} diff -urN /usr/share/empty/fs/aufs/iinfo.c linux/fs/aufs/iinfo.c --- /usr/share/empty/fs/aufs/iinfo.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/iinfo.c 2017-05-06 22:16:52.778220093 +0200 -@@ -0,0 +1,285 @@ ++++ linux/fs/aufs/iinfo.c 2019-03-05 12:13:00.139224339 +0100 +@@ -0,0 +1,286 @@ ++// SPDX-License-Identifier: GPL-2.0 +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -16180,7 +18157,7 @@ diff -urN /usr/share/empty/fs/aufs/iinfo.c linux/fs/aufs/iinfo.c + nbr = 1; + hi = kmalloc_array(nbr, sizeof(*iinfo->ii_hinode), GFP_NOFS); + if (hi) { -+ au_ninodes_inc(sb); ++ au_lcnt_inc(&au_sbi(sb)->si_ninodes); + + iinfo->ii_hinode = hi; + for (i = 0; i < nbr; i++, hi++) @@ -16228,7 +18205,7 @@ diff -urN /usr/share/empty/fs/aufs/iinfo.c linux/fs/aufs/iinfo.c + AuDebugOn(au_is_bad_inode(inode)); + + sb = inode->i_sb; -+ au_ninodes_dec(sb); ++ au_lcnt_dec(&au_sbi(sb)->si_ninodes); + if (si_pid_test(sb)) + au_xino_delete_inode(inode, unlinked); + else { @@ -16245,7 +18222,7 @@ diff -urN /usr/share/empty/fs/aufs/iinfo.c linux/fs/aufs/iinfo.c + + iinfo = au_ii(inode); + if (iinfo->ii_vdir) -+ au_vdir_free(iinfo->ii_vdir, /*atonce*/0); ++ au_vdir_free(iinfo->ii_vdir); + + bindex = iinfo->ii_btop; + if (bindex >= 0) { @@ -16257,15 +18234,16 @@ diff -urN /usr/share/empty/fs/aufs/iinfo.c linux/fs/aufs/iinfo.c + hi++; + } + } -+ au_delayed_kfree(iinfo->ii_hinode); ++ au_kfree_rcu(iinfo->ii_hinode); + AuRwDestroy(&iinfo->ii_rwsem); +} diff -urN /usr/share/empty/fs/aufs/inode.c linux/fs/aufs/inode.c --- /usr/share/empty/fs/aufs/inode.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/inode.c 2017-05-06 22:16:52.781553511 +0200 -@@ -0,0 +1,527 @@ ++++ linux/fs/aufs/inode.c 2019-05-06 09:03:04.820143837 +0200 +@@ -0,0 +1,529 @@ ++// SPDX-License-Identifier: GPL-2.0 +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -16285,6 +18263,7 @@ diff -urN /usr/share/empty/fs/aufs/inode.c linux/fs/aufs/inode.c + * inode functions + */ + ++#include +#include "aufs.h" + +struct inode *au_igrab(struct inode *inode) @@ -16301,7 +18280,7 @@ diff -urN /usr/share/empty/fs/aufs/inode.c linux/fs/aufs/inode.c + au_cpup_attr_all(inode, /*force*/0); + au_update_iigen(inode, /*half*/1); + if (do_version) -+ inode->i_version++; ++ inode_inc_iversion(inode); +} + +static int au_ii_refresh(struct inode *inode, int *update) @@ -16793,10 +18772,11 @@ diff -urN /usr/share/empty/fs/aufs/inode.c linux/fs/aufs/inode.c +} diff -urN /usr/share/empty/fs/aufs/inode.h linux/fs/aufs/inode.h --- /usr/share/empty/fs/aufs/inode.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/inode.h 2017-05-06 22:16:52.781553511 +0200 -@@ -0,0 +1,692 @@ ++++ linux/fs/aufs/inode.h 2019-03-05 12:13:00.142557771 +0100 +@@ -0,0 +1,698 @@ ++/* SPDX-License-Identifier: GPL-2.0 */ +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -16832,10 +18812,8 @@ diff -urN /usr/share/empty/fs/aufs/inode.h linux/fs/aufs/inode.h + /* never use fsnotify_add_vfsmount_mark() */ + struct fsnotify_mark hn_mark; +#endif -+ union { -+ struct inode *hn_aufs_inode; /* no get/put */ -+ struct llist_node hn_lnode; /* delayed free */ -+ }; ++ struct inode *hn_aufs_inode; /* no get/put */ ++ struct rcu_head rcu; +#endif +} ____cacheline_aligned_in_smp; + @@ -16876,12 +18854,10 @@ diff -urN /usr/share/empty/fs/aufs/inode.h linux/fs/aufs/inode.h +}; + +struct au_icntnr { -+ struct au_iinfo iinfo; -+ struct inode vfs_inode; -+ union { -+ struct hlist_node plink; -+ struct llist_node lnode; /* delayed free */ -+ }; ++ struct au_iinfo iinfo; ++ struct inode vfs_inode; ++ struct hlist_bl_node plink; ++ struct rcu_head rcu; +} ____cacheline_aligned_in_smp; + +/* au_pin flags */ @@ -17197,10 +19173,9 @@ diff -urN /usr/share/empty/fs/aufs/inode.h linux/fs/aufs/inode.h +#undef AuWriteLockFunc +#undef AuRWLockFuncs + -+/* -+ * ii_read_unlock, ii_write_unlock, ii_downgrade_lock -+ */ -+AuSimpleUnlockRwsemFuncs(ii, struct inode *i, &au_ii(i)->ii_rwsem); ++#define ii_read_unlock(i) au_rw_read_unlock(&au_ii(i)->ii_rwsem) ++#define ii_write_unlock(i) au_rw_write_unlock(&au_ii(i)->ii_rwsem) ++#define ii_downgrade_lock(i) au_rw_dgrade_lock(&au_ii(i)->ii_rwsem) + +#define IiMustNoWaiters(i) AuRwMustNoWaiters(&au_ii(i)->ii_rwsem) +#define IiMustAnyLock(i) AuRwMustAnyLock(&au_ii(i)->ii_rwsem) @@ -17414,7 +19389,7 @@ diff -urN /usr/share/empty/fs/aufs/inode.h linux/fs/aufs/inode.h +void au_hn_ctl(struct au_hinode *hinode, int do_set); +void au_hn_reset(struct inode *inode, unsigned int flags); +int au_hnotify(struct inode *h_dir, struct au_hnotify *hnotify, u32 mask, -+ struct qstr *h_child_qstr, struct inode *h_child_inode); ++ const struct qstr *file_name, struct inode *h_child_inode); +int au_hnotify_reset_br(unsigned int udba, struct au_branch *br, int perm); +int au_hnotify_init_br(struct au_branch *br, int perm); +void au_hnotify_fin_br(struct au_branch *br); @@ -17479,6 +19454,16 @@ diff -urN /usr/share/empty/fs/aufs/inode.h linux/fs/aufs/inode.h + au_hn_suspend(hdir); +} + ++#if 0 /* unused */ ++#include "vfsub.h" ++static inline void au_hn_inode_lock_shared_nested(struct au_hinode *hdir, ++ unsigned int sc) ++{ ++ inode_lock_shared_nested(hdir->hi_inode, sc); ++ au_hn_suspend(hdir); ++} ++#endif ++ +static inline void au_hn_inode_unlock(struct au_hinode *hdir) +{ + au_hn_resume(hdir); @@ -17489,10 +19474,11 @@ diff -urN /usr/share/empty/fs/aufs/inode.h linux/fs/aufs/inode.h +#endif /* __AUFS_INODE_H__ */ diff -urN /usr/share/empty/fs/aufs/ioctl.c linux/fs/aufs/ioctl.c --- /usr/share/empty/fs/aufs/ioctl.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/ioctl.c 2017-05-06 22:16:52.781553511 +0200 -@@ -0,0 +1,219 @@ ++++ linux/fs/aufs/ioctl.c 2019-03-05 12:13:00.142557771 +0100 +@@ -0,0 +1,220 @@ ++// SPDX-License-Identifier: GPL-2.0 +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -17598,7 +19584,7 @@ diff -urN /usr/share/empty/fs/aufs/ioctl.c linux/fs/aufs/ioctl.c + if (IS_ERR(h_file)) + goto out_fd; + -+ au_br_put(br); /* cf. au_h_open() */ ++ au_lcnt_dec(&br->br_nfiles); /* cf. au_h_open() */ + fd_install(fd, h_file); + err = fd; + goto out; /* success */ @@ -17712,10 +19698,11 @@ diff -urN /usr/share/empty/fs/aufs/ioctl.c linux/fs/aufs/ioctl.c +#endif diff -urN /usr/share/empty/fs/aufs/i_op_add.c linux/fs/aufs/i_op_add.c --- /usr/share/empty/fs/aufs/i_op_add.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/i_op_add.c 2017-05-06 22:16:52.778220093 +0200 -@@ -0,0 +1,920 @@ ++++ linux/fs/aufs/i_op_add.c 2019-05-06 09:03:04.816810402 +0200 +@@ -0,0 +1,936 @@ ++// SPDX-License-Identifier: GPL-2.0 +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -17735,6 +19722,7 @@ diff -urN /usr/share/empty/fs/aufs/i_op_add.c linux/fs/aufs/i_op_add.c + * inode operations (add entry) + */ + ++#include +#include "aufs.h" + +/* @@ -17774,7 +19762,7 @@ diff -urN /usr/share/empty/fs/aufs/i_op_add.c linux/fs/aufs/i_op_add.c + dir = d_inode(dentry->d_parent); /* dir inode is locked */ + IMustLock(dir); + au_dir_ts(dir, bindex); -+ dir->i_version++; ++ inode_inc_iversion(dir); + au_fhsm_wrote(sb, bindex, /*force*/0); + return 0; /* success */ + } @@ -17962,11 +19950,12 @@ diff -urN /usr/share/empty/fs/aufs/i_op_add.c linux/fs/aufs/i_op_add.c + unsigned char created; + const unsigned char try_aopen + = (arg->type == Creat && arg->u.c.try_aopen); ++ struct vfsub_aopen_args *aopen = arg->u.c.aopen; + struct dentry *wh_dentry, *parent; + struct inode *h_dir; + struct super_block *sb; + struct au_branch *br; -+ /* to reuduce stack size */ ++ /* to reduce stack size */ + struct { + struct au_dtime dt; + struct au_pin pin; @@ -18009,30 +19998,42 @@ diff -urN /usr/share/empty/fs/aufs/i_op_add.c linux/fs/aufs/i_op_add.c + h_dir = au_pinned_h_dir(&a->pin); + switch (arg->type) { + case Creat: -+ err = 0; -+ if (!try_aopen || !h_dir->i_op->atomic_open) ++ if (!try_aopen || !h_dir->i_op->atomic_open) { + err = vfsub_create(h_dir, &a->h_path, arg->u.c.mode, + arg->u.c.want_excl); -+ else -+ err = vfsub_atomic_open(h_dir, a->h_path.dentry, -+ arg->u.c.aopen, br); ++ created = !err; ++ if (!err && try_aopen) ++ aopen->file->f_mode |= FMODE_CREATED; ++ } else { ++ aopen->br = br; ++ err = vfsub_atomic_open(h_dir, a->h_path.dentry, aopen); ++ AuDbg("err %d\n", err); ++ AuDbgFile(aopen->file); ++ created = err >= 0 ++ && !!(aopen->file->f_mode & FMODE_CREATED); ++ } + break; + case Symlink: + err = vfsub_symlink(h_dir, &a->h_path, arg->u.s.symname); ++ created = !err; + break; + case Mknod: + err = vfsub_mknod(h_dir, &a->h_path, arg->u.m.mode, + arg->u.m.dev); ++ created = !err; + break; + default: + BUG(); + } -+ created = !err; ++ if (unlikely(err < 0)) ++ goto out_unpin; ++ ++ err = epilog(dir, btop, wh_dentry, dentry); + if (!err) -+ err = epilog(dir, btop, wh_dentry, dentry); ++ goto out_unpin; /* success */ + + /* revert */ -+ if (unlikely(created && err && d_is_positive(a->h_path.dentry))) { ++ if (created /* && d_is_positive(a->h_path.dentry) */) { + /* no delegation since it is just created */ + rerr = vfsub_unlink(h_dir, &a->h_path, /*delegated*/NULL, + /*force*/0); @@ -18043,13 +20044,14 @@ diff -urN /usr/share/empty/fs/aufs/i_op_add.c linux/fs/aufs/i_op_add.c + } + au_dtime_revert(&a->dt); + } ++ if (try_aopen && h_dir->i_op->atomic_open ++ && (aopen->file->f_mode & FMODE_OPENED)) ++ /* aopen->file is still opened */ ++ au_lcnt_dec(&aopen->br->br_nfiles); + -+ if (!err && try_aopen && !h_dir->i_op->atomic_open) -+ *arg->u.c.aopen->opened |= FILE_CREATED; -+ ++out_unpin: + au_unpin(&a->pin); + dput(wh_dentry); -+ +out_parent: + if (!try_aopen) + di_write_unlock(parent); @@ -18061,7 +20063,7 @@ diff -urN /usr/share/empty/fs/aufs/i_op_add.c linux/fs/aufs/i_op_add.c + if (!try_aopen) + aufs_read_unlock(dentry, AuLock_DW); +out_free: -+ au_delayed_kfree(a); ++ au_kfree_rcu(a); +out: + return err; +} @@ -18422,7 +20424,7 @@ diff -urN /usr/share/empty/fs/aufs/i_op_add.c linux/fs/aufs/i_op_add.c + + /* + * aufs doesn't touch the credential so -+ * security_dentry_create_files_as() is unnecrssary. ++ * security_dentry_create_files_as() is unnecessary. + */ + if (au_opt_test(au_mntflags(sb), PLINK)) { + if (a->bdst < a->bsrc @@ -18487,7 +20489,7 @@ diff -urN /usr/share/empty/fs/aufs/i_op_add.c linux/fs/aufs/i_op_add.c + } + + au_dir_ts(dir, a->bdst); -+ dir->i_version++; ++ inode_inc_iversion(dir); + inc_nlink(inode); + inode->i_ctime = dir->i_ctime; + d_instantiate(dentry, au_igrab(inode)); @@ -18521,7 +20523,7 @@ diff -urN /usr/share/empty/fs/aufs/i_op_add.c linux/fs/aufs/i_op_add.c + } + aufs_read_and_write_unlock2(dentry, src_dentry); +out_kfree: -+ au_delayed_kfree(a); ++ au_kfree_rcu(a); +out: + AuTraceErr(err); + return err; @@ -18630,16 +20632,17 @@ diff -urN /usr/share/empty/fs/aufs/i_op_add.c linux/fs/aufs/i_op_add.c + } + aufs_read_unlock(dentry, AuLock_DW); +out_free: -+ au_delayed_kfree(a); ++ au_kfree_rcu(a); +out: + return err; +} diff -urN /usr/share/empty/fs/aufs/i_op.c linux/fs/aufs/i_op.c --- /usr/share/empty/fs/aufs/i_op.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/i_op.c 2017-05-06 22:16:52.778220093 +0200 -@@ -0,0 +1,1452 @@ ++++ linux/fs/aufs/i_op.c 2019-05-06 09:03:04.816810402 +0200 +@@ -0,0 +1,1507 @@ ++// SPDX-License-Identifier: GPL-2.0 +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -18661,6 +20664,7 @@ diff -urN /usr/share/empty/fs/aufs/i_op.c linux/fs/aufs/i_op.c + +#include +#include ++#include +#include +#include +#include "aufs.h" @@ -18679,14 +20683,14 @@ diff -urN /usr/share/empty/fs/aufs/i_op.c linux/fs/aufs/i_op.c + if (((mask & MAY_EXEC) + && S_ISREG(h_inode->i_mode) + && (path_noexec(h_path) -+ || !(h_inode->i_mode & S_IXUGO)))) ++ || !(h_inode->i_mode & 0111)))) + goto out; + + /* + * - skip the lower fs test in the case of write to ro branch. + * - nfs dir permission write check is optimized, but a policy for + * link/rename requires a real check. -+ * - nfs always sets MS_POSIXACL regardless its mount option 'noacl.' ++ * - nfs always sets SB_POSIXACL regardless its mount option 'noacl.' + * in this case, generic_permission() returns -EOPNOTSUPP. + */ + if ((write_mask && !au_br_writable(brperm)) @@ -18694,7 +20698,7 @@ diff -urN /usr/share/empty/fs/aufs/i_op.c linux/fs/aufs/i_op.c + && write_mask && !(mask & MAY_READ)) + || !h_inode->i_op->permission) { + /* AuLabel(generic_permission); */ -+ /* AuDbg("get_acl %pf\n", h_inode->i_op->get_acl); */ ++ /* AuDbg("get_acl %ps\n", h_inode->i_op->get_acl); */ + err = generic_permission(h_inode, mask); + if (err == -EOPNOTSUPP && au_test_nfs_noacl(h_inode)) + err = h_inode->i_op->permission(h_inode, mask); @@ -18891,45 +20895,100 @@ diff -urN /usr/share/empty/fs/aufs/i_op.c linux/fs/aufs/i_op.c + +/* ---------------------------------------------------------------------- */ + ++/* ++ * very dirty and complicated aufs ->atomic_open(). ++ * aufs_atomic_open() ++ * + au_aopen_or_create() ++ * + add_simple() ++ * + vfsub_atomic_open() ++ * + branch fs ->atomic_open() ++ * may call the actual 'open' for h_file ++ * + inc br_nfiles only if opened ++ * + au_aopen_no_open() or au_aopen_do_open() ++ * ++ * au_aopen_do_open() ++ * + finish_open() ++ * + au_do_aopen() ++ * + au_do_open() the body of all 'open' ++ * + au_do_open_nondir() ++ * set the passed h_file ++ * ++ * au_aopen_no_open() ++ * + finish_no_open() ++ */ ++ +struct aopen_node { -+ struct hlist_node hlist; ++ struct hlist_bl_node hblist; + struct file *file, *h_file; +}; + +static int au_do_aopen(struct inode *inode, struct file *file) +{ -+ struct au_sphlhead *aopen; ++ struct hlist_bl_head *aopen; ++ struct hlist_bl_node *pos; + struct aopen_node *node; + struct au_do_open_args args = { -+ .no_lock = 1, -+ .open = au_do_open_nondir ++ .aopen = 1, ++ .open = au_do_open_nondir + }; + + aopen = &au_sbi(inode->i_sb)->si_aopen; -+ spin_lock(&aopen->spin); -+ hlist_for_each_entry(node, &aopen->head, hlist) ++ hlist_bl_lock(aopen); ++ hlist_bl_for_each_entry(node, pos, aopen, hblist) + if (node->file == file) { + args.h_file = node->h_file; + break; + } -+ spin_unlock(&aopen->spin); ++ hlist_bl_unlock(aopen); + /* AuDebugOn(!args.h_file); */ + + return au_do_open(file, &args); +} + ++static int au_aopen_do_open(struct file *file, struct dentry *dentry, ++ struct aopen_node *aopen_node) ++{ ++ int err; ++ struct hlist_bl_head *aopen; ++ ++ AuLabel(here); ++ aopen = &au_sbi(dentry->d_sb)->si_aopen; ++ au_hbl_add(&aopen_node->hblist, aopen); ++ err = finish_open(file, dentry, au_do_aopen); ++ au_hbl_del(&aopen_node->hblist, aopen); ++ /* AuDbgFile(file); */ ++ AuDbg("%pd%s%s\n", dentry, ++ (file->f_mode & FMODE_CREATED) ? " created" : "", ++ (file->f_mode & FMODE_OPENED) ? " opened" : ""); ++ ++ AuTraceErr(err); ++ return err; ++} ++ ++static int au_aopen_no_open(struct file *file, struct dentry *dentry) ++{ ++ int err; ++ ++ AuLabel(here); ++ dget(dentry); ++ err = finish_no_open(file, dentry); ++ ++ AuTraceErr(err); ++ return err; ++} ++ +static int aufs_atomic_open(struct inode *dir, struct dentry *dentry, + struct file *file, unsigned int open_flag, -+ umode_t create_mode, int *opened) ++ umode_t create_mode) +{ -+ int err, h_opened = *opened; ++ int err, did_open; + unsigned int lkup_flags; ++ aufs_bindex_t bindex; ++ struct super_block *sb; + struct dentry *parent, *d; -+ struct au_sphlhead *aopen; + struct vfsub_aopen_args args = { + .open_flag = open_flag, -+ .create_mode = create_mode, -+ .opened = &h_opened ++ .create_mode = create_mode + }; + struct aopen_node aopen_node = { + .file = file @@ -18964,74 +21023,73 @@ diff -urN /usr/share/empty/fs/aufs/i_op.c linux/fs/aufs/i_op.c + if (d_is_positive(dentry) + || d_unhashed(dentry) + || d_unlinked(dentry) -+ || !(open_flag & O_CREAT)) -+ goto out_no_open; ++ || !(open_flag & O_CREAT)) { ++ err = au_aopen_no_open(file, dentry); ++ goto out; /* success */ ++ } + + err = aufs_read_lock(dentry, AuLock_DW | AuLock_FLUSH | AuLock_GEN); + if (unlikely(err)) + goto out; + ++ sb = dentry->d_sb; + parent = dentry->d_parent; /* dir is locked */ + di_write_lock_parent(parent); + err = au_lkup_dentry(dentry, /*btop*/0, AuLkup_ALLOW_NEG); -+ if (unlikely(err)) -+ goto out_unlock; ++ if (unlikely(err < 0)) ++ goto out_parent; + + AuDbgDentry(dentry); -+ if (d_is_positive(dentry)) -+ goto out_unlock; ++ if (d_is_positive(dentry)) { ++ err = au_aopen_no_open(file, dentry); ++ goto out_parent; /* success */ ++ } + -+ args.file = get_empty_filp(); ++ args.file = alloc_empty_file(file->f_flags, current_cred()); + err = PTR_ERR(args.file); + if (IS_ERR(args.file)) -+ goto out_unlock; ++ goto out_parent; + -+ args.file->f_flags = file->f_flags; ++ bindex = au_dbtop(dentry); + err = au_aopen_or_create(dir, dentry, &args); + AuTraceErr(err); + AuDbgFile(args.file); ++ file->f_mode = args.file->f_mode & ~FMODE_OPENED; ++ did_open = !!(args.file->f_mode & FMODE_OPENED); ++ if (!did_open) { ++ fput(args.file); ++ args.file = NULL; ++ } ++ di_write_unlock(parent); ++ di_write_unlock(dentry); + if (unlikely(err < 0)) { -+ if (h_opened & FILE_OPENED) ++ if (args.file) + fput(args.file); -+ else -+ put_filp(args.file); -+ goto out_unlock; ++ goto out_sb; + } + -+ /* some filesystems don't set FILE_CREATED while succeeded? */ -+ *opened |= FILE_CREATED; -+ if (h_opened & FILE_OPENED) -+ aopen_node.h_file = args.file; ++ if (!did_open) ++ err = au_aopen_no_open(file, dentry); + else { -+ put_filp(args.file); -+ args.file = NULL; ++ aopen_node.h_file = args.file; ++ err = au_aopen_do_open(file, dentry, &aopen_node); + } -+ aopen = &au_sbi(dir->i_sb)->si_aopen; -+ au_sphl_add(&aopen_node.hlist, aopen); -+ err = finish_open(file, dentry, au_do_aopen, opened); -+ au_sphl_del(&aopen_node.hlist, aopen); -+ AuTraceErr(err); -+ AuDbgFile(file); -+ if (aopen_node.h_file) -+ fput(aopen_node.h_file); ++ if (unlikely(err < 0)) { ++ if (args.file) ++ fput(args.file); ++ if (did_open) ++ au_lcnt_dec(&args.br->br_nfiles); ++ } ++ goto out_sb; /* success */ + -+out_unlock: ++out_parent: + di_write_unlock(parent); -+ aufs_read_unlock(dentry, AuLock_DW); -+ AuDbgDentry(dentry); -+ if (unlikely(err < 0)) -+ goto out; -+out_no_open: -+ if (err >= 0 && !(*opened & FILE_CREATED)) { -+ AuLabel(out_no_open); -+ dget(dentry); -+ err = finish_no_open(file, dentry); -+ } ++ di_write_unlock(dentry); ++out_sb: ++ si_read_unlock(sb); +out: -+ AuDbg("%pd%s%s\n", dentry, -+ (*opened & FILE_CREATED) ? " created" : "", -+ (*opened & FILE_OPENED) ? " opened" : ""); + AuTraceErr(err); ++ AuDbgFile(file); + return err; +} + @@ -19063,10 +21121,10 @@ diff -urN /usr/share/empty/fs/aufs/i_op.c linux/fs/aufs/i_op.c + if (!err && add_entry && !au_ftest_wrdir(add_entry, TMPFILE)) { + h_parent = au_h_dptr(parent, bcpup); + h_dir = d_inode(h_parent); -+ inode_lock_nested(h_dir, AuLsc_I_PARENT); ++ inode_lock_shared_nested(h_dir, AuLsc_I_PARENT); + err = au_lkup_neg(dentry, bcpup, /*wh*/0); + /* todo: no unlock here */ -+ inode_unlock(h_dir); ++ inode_unlock_shared(h_dir); + + AuDbg("bcpup %d\n", bcpup); + if (!err) { @@ -19450,10 +21508,10 @@ diff -urN /usr/share/empty/fs/aufs/i_op.c linux/fs/aufs/i_op.c + a->h_path.dentry = au_h_dptr(dentry, btop); + a->h_inode = d_inode(a->h_path.dentry); + if (ia && (ia->ia_valid & ATTR_SIZE)) { -+ inode_lock_nested(a->h_inode, AuLsc_I_CHILD); ++ inode_lock_shared_nested(a->h_inode, AuLsc_I_CHILD); + if (ia->ia_size < i_size_read(a->h_inode)) + sz = ia->ia_size; -+ inode_unlock(a->h_inode); ++ inode_unlock_shared(a->h_inode); + } + + hi_wh = NULL; @@ -19641,7 +21699,7 @@ diff -urN /usr/share/empty/fs/aufs/i_op.c linux/fs/aufs/i_op.c +out_si: + si_read_unlock(sb); +out_kfree: -+ au_delayed_kfree(a); ++ au_kfree_rcu(a); +out: + AuTraceErr(err); + return err; @@ -19732,7 +21790,7 @@ diff -urN /usr/share/empty/fs/aufs/i_op.c linux/fs/aufs/i_op.c + di_write_unlock(dentry); + si_read_unlock(sb); +out_kfree: -+ au_delayed_kfree(a); ++ au_kfree_rcu(a); +out: + AuTraceErr(err); + return err; @@ -19942,7 +22000,7 @@ diff -urN /usr/share/empty/fs/aufs/i_op.c linux/fs/aufs/i_op.c + goto out_unlock; + + err = 0; -+ AuDbg("%pf\n", h_inode->i_op->get_link); ++ AuDbg("%ps\n", h_inode->i_op->get_link); + AuDbgDentry(h_dentry); + ret = vfs_get_link(h_dentry, done); + dput(h_dentry); @@ -19965,7 +22023,8 @@ diff -urN /usr/share/empty/fs/aufs/i_op.c linux/fs/aufs/i_op.c + return (inode->i_mode & (S_IFBLK | S_IFCHR | S_IFIFO | S_IFSOCK)); +} + -+static int aufs_update_time(struct inode *inode, struct timespec *ts, int flags) ++static int aufs_update_time(struct inode *inode, struct timespec64 *ts, ++ int flags) +{ + int err; + aufs_bindex_t bindex; @@ -19981,7 +22040,6 @@ diff -urN /usr/share/empty/fs/aufs/i_op.c linux/fs/aufs/i_op.c + lockdep_off(); + si_read_lock(sb, AuLock_FLUSH); + ii_write_lock_child(inode); -+ lockdep_on(); + + err = 0; + bindex = au_ibtop(inode); @@ -20009,7 +22067,6 @@ diff -urN /usr/share/empty/fs/aufs/i_op.c linux/fs/aufs/i_op.c + AuDebugOn(1); + } + -+ lockdep_off(); + if (!err) + au_cpup_attr_timesizes(inode); + ii_write_unlock(inode); @@ -20092,10 +22149,11 @@ diff -urN /usr/share/empty/fs/aufs/i_op.c linux/fs/aufs/i_op.c +}; diff -urN /usr/share/empty/fs/aufs/i_op_del.c linux/fs/aufs/i_op_del.c --- /usr/share/empty/fs/aufs/i_op_del.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/i_op_del.c 2017-05-06 22:16:52.778220093 +0200 -@@ -0,0 +1,511 @@ ++++ linux/fs/aufs/i_op_del.c 2019-05-06 09:03:04.816810402 +0200 +@@ -0,0 +1,513 @@ ++// SPDX-License-Identifier: GPL-2.0 +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -20115,6 +22173,7 @@ diff -urN /usr/share/empty/fs/aufs/i_op_del.c linux/fs/aufs/i_op_del.c + * inode operations (del entry) + */ + ++#include +#include "aufs.h" + +/* @@ -20244,7 +22303,7 @@ diff -urN /usr/share/empty/fs/aufs/i_op_del.c linux/fs/aufs/i_op_del.c + +/* + * decide the branch where we operate for @dentry. the branch index will be set -+ * @rbcpup. after diciding it, 'pin' it and store the timestamps of the parent ++ * @rbcpup. after deciding it, 'pin' it and store the timestamps of the parent + * dir for reverting. + * when a new whiteout is necessary, create it. + */ @@ -20365,7 +22424,7 @@ diff -urN /usr/share/empty/fs/aufs/i_op_del.c linux/fs/aufs/i_op_del.c + inode->i_ctime = dir->i_ctime; + + au_dir_ts(dir, bindex); -+ dir->i_version++; ++ inode_inc_iversion(dir); +} + +/* @@ -20400,7 +22459,7 @@ diff -urN /usr/share/empty/fs/aufs/i_op_del.c linux/fs/aufs/i_op_del.c + aufs_bindex_t bwh, bindex, btop; + struct inode *inode, *h_dir, *delegated; + struct dentry *parent, *wh_dentry; -+ /* to reuduce stack size */ ++ /* to reduce stack size */ + struct { + struct au_dtime dt; + struct au_pin pin; @@ -20490,7 +22549,7 @@ diff -urN /usr/share/empty/fs/aufs/i_op_del.c linux/fs/aufs/i_op_del.c +out_unlock: + aufs_read_unlock(dentry, AuLock_DW); +out_free: -+ au_delayed_kfree(a); ++ au_kfree_rcu(a); +out: + return err; +} @@ -20502,7 +22561,7 @@ diff -urN /usr/share/empty/fs/aufs/i_op_del.c linux/fs/aufs/i_op_del.c + struct inode *inode; + struct dentry *parent, *wh_dentry, *h_dentry; + struct au_whtmp_rmdir *args; -+ /* to reuduce stack size */ ++ /* to reduce stack size */ + struct { + struct au_dtime dt; + struct au_pin pin; @@ -20600,17 +22659,18 @@ diff -urN /usr/share/empty/fs/aufs/i_op_del.c linux/fs/aufs/i_op_del.c +out_unlock: + aufs_read_unlock(dentry, AuLock_DW); +out_free: -+ au_delayed_kfree(a); ++ au_kfree_rcu(a); +out: + AuTraceErr(err); + return err; +} diff -urN /usr/share/empty/fs/aufs/i_op_ren.c linux/fs/aufs/i_op_ren.c --- /usr/share/empty/fs/aufs/i_op_ren.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/i_op_ren.c 2017-05-06 22:16:52.778220093 +0200 -@@ -0,0 +1,1165 @@ ++++ linux/fs/aufs/i_op_ren.c 2019-05-06 09:03:04.820143837 +0200 +@@ -0,0 +1,1250 @@ ++// SPDX-License-Identifier: GPL-2.0 +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -20631,6 +22691,7 @@ diff -urN /usr/share/empty/fs/aufs/i_op_ren.c linux/fs/aufs/i_op_ren.c + * todo: this is crazy monster + */ + ++#include +#include "aufs.h" + +enum { AuSRC, AuDST, AuSrcDst }; @@ -20645,12 +22706,20 @@ diff -urN /usr/share/empty/fs/aufs/i_op_ren.c linux/fs/aufs/i_op_ren.c +#define AuRen_DT_DSTDIR (1 << 6) +#define AuRen_DIROPQ_SRC (1 << 7) +#define AuRen_DIROPQ_DST (1 << 8) ++#define AuRen_DIRREN (1 << 9) ++#define AuRen_DROPPED_SRC (1 << 10) ++#define AuRen_DROPPED_DST (1 << 11) +#define au_ftest_ren(flags, name) ((flags) & AuRen_##name) +#define au_fset_ren(flags, name) \ + do { (flags) |= AuRen_##name; } while (0) +#define au_fclr_ren(flags, name) \ + do { (flags) &= ~AuRen_##name; } while (0) + ++#ifndef CONFIG_AUFS_DIRREN ++#undef AuRen_DIRREN ++#define AuRen_DIRREN 0 ++#endif ++ +struct au_ren_args { + struct { + struct dentry *dentry, *h_dentry, *parent, *h_parent, @@ -20703,6 +22772,7 @@ diff -urN /usr/share/empty/fs/aufs/i_op_ren.c linux/fs/aufs/i_op_ren.c + + struct au_whtmp_rmdir *thargs; + struct dentry *h_dst; ++ struct au_hinode *h_root; +}; + +/* ---------------------------------------------------------------------- */ @@ -20917,6 +22987,7 @@ diff -urN /usr/share/empty/fs/aufs/i_op_ren.c linux/fs/aufs/i_op_ren.c + d = a->dst_dentry; /* already renamed on the branch */ + always = !!au_opt_test(au_mntflags(d->d_sb), ALWAYS_DIROPQ); + if (au_ftest_ren(a->auren_flags, ISDIR_SRC) ++ && !au_ftest_ren(a->auren_flags, DIRREN) + && a->btgt != au_dbdiropq(a->src_dentry) + && (a->dst_wh_dentry + || a->btgt <= au_dbdiropq(d) @@ -20964,6 +23035,7 @@ diff -urN /usr/share/empty/fs/aufs/i_op_ren.c linux/fs/aufs/i_op_ren.c + /* prepare workqueue args for asynchronous rmdir */ + h_d = a->dst_h_dentry; + if (au_ftest_ren(a->auren_flags, ISDIR_DST) ++ /* && !au_ftest_ren(a->auren_flags, DIRREN) */ + && d_is_positive(h_d)) { + err = -ENOMEM; + a->thargs = au_whtmp_rmdir_alloc(a->src_dentry->d_sb, @@ -21013,6 +23085,11 @@ diff -urN /usr/share/empty/fs/aufs/i_op_ren.c linux/fs/aufs/i_op_ren.c + } + + BUG_ON(d_is_positive(a->dst_h_dentry) && a->src_btop != a->btgt); ++#if 0 ++ BUG_ON(!au_ftest_ren(a->auren_flags, DIRREN) ++ && d_is_positive(a->dst_h_dentry) ++ && a->src_btop != a->btgt); ++#endif + + /* rename by vfs_rename or cpup */ + err = au_ren_or_cpup(a); @@ -21095,25 +23172,35 @@ diff -urN /usr/share/empty/fs/aufs/i_op_ren.c linux/fs/aufs/i_op_ren.c +} + +/* -+ * test if @dentry dir can be rename source or not. -+ * if it can, return 0 and @children is filled. ++ * test if @a->src_dentry dir can be rename source or not. ++ * if it can, return 0. + * success means, + * - it is a logically empty dir. + * - or, it exists on writable branch and has no children including whiteouts -+ * on the lower branch. ++ * on the lower branch unless DIRREN is on. + */ -+static int may_rename_srcdir(struct dentry *dentry, aufs_bindex_t btgt) ++static int may_rename_srcdir(struct au_ren_args *a) +{ + int err; + unsigned int rdhash; -+ aufs_bindex_t btop; ++ aufs_bindex_t btop, btgt; ++ struct dentry *dentry; ++ struct super_block *sb; ++ struct au_sbinfo *sbinfo; ++ ++ dentry = a->src_dentry; ++ sb = dentry->d_sb; ++ sbinfo = au_sbi(sb); ++ if (au_opt_test(sbinfo->si_mntflags, DIRREN)) ++ au_fset_ren(a->auren_flags, DIRREN); + ++ btgt = a->btgt; + btop = au_dbtop(dentry); + if (btop != btgt) { + struct au_nhash whlist; + -+ SiMustAnyLock(dentry->d_sb); -+ rdhash = au_sbi(dentry->d_sb)->si_rdhash; ++ SiMustAnyLock(sb); ++ rdhash = sbinfo->si_rdhash; + if (!rdhash) + rdhash = au_rdhash_est(au_dir_size(/*file*/NULL, + dentry)); @@ -21132,9 +23219,13 @@ diff -urN /usr/share/empty/fs/aufs/i_op_ren.c linux/fs/aufs/i_op_ren.c + +out: + if (err == -ENOTEMPTY) { -+ AuWarn1("renaming dir who has child(ren) on multiple branches," -+ " is not supported\n"); -+ err = -EXDEV; ++ if (au_ftest_ren(a->auren_flags, DIRREN)) { ++ err = 0; ++ } else { ++ AuWarn1("renaming dir who has child(ren) on multiple " ++ "branches, is not supported\n"); ++ err = -EXDEV; ++ } + } + return err; +} @@ -21163,7 +23254,7 @@ diff -urN /usr/share/empty/fs/aufs/i_op_ren.c linux/fs/aufs/i_op_ren.c + err = may_rename_dstdir(d, &a->whlist); + au_set_dbtop(d, a->btgt); + } else -+ err = may_rename_srcdir(d, a->btgt); ++ err = may_rename_srcdir(a); + } + a->dst_h_dentry = au_h_dptr(d, au_dbtop(d)); + if (unlikely(err)) @@ -21172,7 +23263,7 @@ diff -urN /usr/share/empty/fs/aufs/i_op_ren.c linux/fs/aufs/i_op_ren.c + d = a->src_dentry; + a->src_h_dentry = au_h_dptr(d, au_dbtop(d)); + if (au_ftest_ren(a->auren_flags, ISDIR_SRC)) { -+ err = may_rename_srcdir(d, a->btgt); ++ err = may_rename_srcdir(a); + if (unlikely(err)) { + au_nhash_wh_free(&a->whlist); + a->whlist.nh_num = 0; @@ -21239,7 +23330,7 @@ diff -urN /usr/share/empty/fs/aufs/i_op_ren.c linux/fs/aufs/i_op_ren.c + * locking order + * (VFS) + * - src_dir and dir by lock_rename() -+ * - inode if exitsts ++ * - inode if exists + * (aufs) + * - lock all + * + src_dentry and dentry by aufs_read_and_write_lock2() which calls, @@ -21262,6 +23353,9 @@ diff -urN /usr/share/empty/fs/aufs/i_op_ren.c linux/fs/aufs/i_op_ren.c +{ + vfsub_unlock_rename(a->src_h_parent, a->src_hdir, + a->dst_h_parent, a->dst_hdir); ++ if (au_ftest_ren(a->auren_flags, DIRREN) ++ && a->h_root) ++ au_hn_inode_unlock(a->h_root); + if (au_ftest_ren(a->auren_flags, MNT_WRITE)) + vfsub_mnt_drop_write(au_br_mnt(a->br)); +} @@ -21281,6 +23375,23 @@ diff -urN /usr/share/empty/fs/aufs/i_op_ren.c linux/fs/aufs/i_op_ren.c + if (unlikely(err)) + goto out; + au_fset_ren(a->auren_flags, MNT_WRITE); ++ if (au_ftest_ren(a->auren_flags, DIRREN)) { ++ struct dentry *root; ++ struct inode *dir; ++ ++ /* ++ * sbinfo is already locked, so this ii_read_lock is ++ * unnecessary. but our debugging feature checks it. ++ */ ++ root = a->src_inode->i_sb->s_root; ++ if (root != a->src_parent && root != a->dst_parent) { ++ dir = d_inode(root); ++ ii_read_lock_parent3(dir); ++ a->h_root = au_hi(dir, a->btgt); ++ ii_read_unlock(dir); ++ au_hn_inode_lock_nested(a->h_root, AuLsc_I_PARENT3); ++ } ++ } + a->h_trap = vfsub_lock_rename(a->src_h_parent, a->src_hdir, + a->dst_h_parent, a->dst_hdir); + udba = au_opt_udba(a->src_dentry->d_sb); @@ -21312,7 +23423,7 @@ diff -urN /usr/share/empty/fs/aufs/i_op_ren.c linux/fs/aufs/i_op_ren.c + struct inode *dir; + + dir = a->dst_dir; -+ dir->i_version++; ++ inode_inc_iversion(dir); + if (au_ftest_ren(a->auren_flags, ISDIR_SRC)) { + /* is this updating defined in POSIX? */ + au_cpup_attr_timesizes(a->src_inode); @@ -21322,7 +23433,7 @@ diff -urN /usr/share/empty/fs/aufs/i_op_ren.c linux/fs/aufs/i_op_ren.c + + if (a->exchange) { + dir = a->src_dir; -+ dir->i_version++; ++ inode_inc_iversion(dir); + if (au_ftest_ren(a->auren_flags, ISDIR_DST)) { + /* is this updating defined in POSIX? */ + au_cpup_attr_timesizes(a->dst_inode); @@ -21335,7 +23446,7 @@ diff -urN /usr/share/empty/fs/aufs/i_op_ren.c linux/fs/aufs/i_op_ren.c + return; + + dir = a->src_dir; -+ dir->i_version++; ++ inode_inc_iversion(dir); + if (au_ftest_ren(a->auren_flags, ISDIR_SRC)) + au_cpup_attr_nlink(dir, /*force*/1); + au_dir_ts(dir, a->btgt); @@ -21376,34 +23487,39 @@ diff -urN /usr/share/empty/fs/aufs/i_op_ren.c linux/fs/aufs/i_op_ren.c + au_update_dbrange(d, /*do_put_zero*/0); + } + ++ if (a->exchange ++ || au_ftest_ren(a->auren_flags, DIRREN)) { ++ d_drop(a->src_dentry); ++ if (au_ftest_ren(a->auren_flags, DIRREN)) ++ au_set_dbwh(a->src_dentry, -1); ++ return; ++ } ++ + d = a->src_dentry; -+ if (!a->exchange) { -+ au_set_dbwh(d, -1); -+ bbot = au_dbbot(d); -+ for (bindex = a->btgt + 1; bindex <= bbot; bindex++) { -+ h_d = au_h_dptr(d, bindex); -+ if (h_d) -+ au_set_h_dptr(d, bindex, NULL); -+ } -+ au_set_dbbot(d, a->btgt); ++ au_set_dbwh(d, -1); ++ bbot = au_dbbot(d); ++ for (bindex = a->btgt + 1; bindex <= bbot; bindex++) { ++ h_d = au_h_dptr(d, bindex); ++ if (h_d) ++ au_set_h_dptr(d, bindex, NULL); ++ } ++ au_set_dbbot(d, a->btgt); + -+ sb = d->d_sb; -+ i = a->src_inode; -+ if (au_opt_test(au_mntflags(sb), PLINK) && au_plink_test(i)) -+ return; /* success */ ++ sb = d->d_sb; ++ i = a->src_inode; ++ if (au_opt_test(au_mntflags(sb), PLINK) && au_plink_test(i)) ++ return; /* success */ + -+ bbot = au_ibbot(i); -+ for (bindex = a->btgt + 1; bindex <= bbot; bindex++) { -+ h_i = au_h_iptr(i, bindex); -+ if (h_i) { -+ au_xino_write(sb, bindex, h_i->i_ino, /*ino*/0); -+ /* ignore this error */ -+ au_set_h_iptr(i, bindex, NULL, 0); -+ } ++ bbot = au_ibbot(i); ++ for (bindex = a->btgt + 1; bindex <= bbot; bindex++) { ++ h_i = au_h_iptr(i, bindex); ++ if (h_i) { ++ au_xino_write(sb, bindex, h_i->i_ino, /*ino*/0); ++ /* ignore this error */ ++ au_set_h_iptr(i, bindex, NULL, 0); + } -+ au_set_ibbot(i, a->btgt); + } -+ d_drop(a->src_dentry); ++ au_set_ibbot(i, a->btgt); +} + +/* ---------------------------------------------------------------------- */ @@ -21512,6 +23628,7 @@ diff -urN /usr/share/empty/fs/aufs/i_op_ren.c linux/fs/aufs/i_op_ren.c + unsigned int _flags) +{ + int err, lock_flags; ++ void *rev; + /* reduce stack space */ + struct au_ren_args *a; + struct au_pin pin; @@ -21531,6 +23648,8 @@ diff -urN /usr/share/empty/fs/aufs/i_op_ren.c linux/fs/aufs/i_op_ren.c + goto out; + + a->flags = _flags; ++ BUILD_BUG_ON(sizeof(a->exchange) == sizeof(u8) ++ && RENAME_EXCHANGE > U8_MAX); + a->exchange = _flags & RENAME_EXCHANGE; + a->src_dir = _src_dir; + a->src_dentry = _src_dentry; @@ -21571,7 +23690,8 @@ diff -urN /usr/share/empty/fs/aufs/i_op_ren.c linux/fs/aufs/i_op_ren.c + goto out_free; + lock_flags |= AuLock_DIRS; + } -+ err = aufs_read_and_write_lock2(a->dst_dentry, a->src_dentry, lock_flags); ++ err = aufs_read_and_write_lock2(a->dst_dentry, a->src_dentry, ++ lock_flags); + if (unlikely(err)) + goto out_free; + @@ -21724,10 +23844,22 @@ diff -urN /usr/share/empty/fs/aufs/i_op_ren.c linux/fs/aufs/i_op_ren.c + /* store timestamps to be revertible */ + au_ren_dt(a); + ++ /* store dirren info */ ++ if (au_ftest_ren(a->auren_flags, DIRREN)) { ++ err = au_dr_rename(a->src_dentry, a->btgt, ++ &a->dst_dentry->d_name, &rev); ++ AuTraceErr(err); ++ if (unlikely(err)) ++ goto out_dt; ++ } ++ + /* here we go */ + err = do_rename(a); + if (unlikely(err)) -+ goto out_dt; ++ goto out_dirren; ++ ++ if (au_ftest_ren(a->auren_flags, DIRREN)) ++ au_dr_rename_fin(a->src_dentry, a->btgt, rev); + + /* update dir attributes */ + au_ren_refresh_dir(a); @@ -21737,6 +23869,9 @@ diff -urN /usr/share/empty/fs/aufs/i_op_ren.c linux/fs/aufs/i_op_ren.c + + goto out_hdir; /* success */ + ++out_dirren: ++ if (au_ftest_ren(a->auren_flags, DIRREN)) ++ au_dr_rename_rev(a->src_dentry, a->btgt, rev); +out_dt: + au_ren_rev_dt(err, a); +out_hdir: @@ -21750,10 +23885,19 @@ diff -urN /usr/share/empty/fs/aufs/i_op_ren.c linux/fs/aufs/i_op_ren.c + } +out_parent: + if (!err) { ++ if (d_unhashed(a->src_dentry)) ++ au_fset_ren(a->auren_flags, DROPPED_SRC); ++ if (d_unhashed(a->dst_dentry)) ++ au_fset_ren(a->auren_flags, DROPPED_DST); + if (!a->exchange) + d_move(a->src_dentry, a->dst_dentry); -+ else ++ else { + d_exchange(a->src_dentry, a->dst_dentry); ++ if (au_ftest_ren(a->auren_flags, DROPPED_DST)) ++ d_drop(a->dst_dentry); ++ } ++ if (au_ftest_ren(a->auren_flags, DROPPED_SRC)) ++ d_drop(a->src_dentry); + } else { + au_update_dbtop(a->dst_dentry); + if (!a->dst_inode) @@ -21769,15 +23913,16 @@ diff -urN /usr/share/empty/fs/aufs/i_op_ren.c linux/fs/aufs/i_op_ren.c + iput(a->dst_inode); + if (a->thargs) + au_whtmp_rmdir_free(a->thargs); -+ au_delayed_kfree(a); ++ au_kfree_rcu(a); +out: + AuTraceErr(err); + return err; +} diff -urN /usr/share/empty/fs/aufs/Kconfig linux/fs/aufs/Kconfig --- /usr/share/empty/fs/aufs/Kconfig 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/Kconfig 2017-05-06 22:16:52.774886674 +0200 -@@ -0,0 +1,185 @@ ++++ linux/fs/aufs/Kconfig 2018-06-04 09:08:09.181412645 +0200 +@@ -0,0 +1,199 @@ ++# SPDX-License-Identifier: GPL-2.0 +config AUFS_FS + tristate "Aufs (Advanced multi layered unification filesystem) support" + help @@ -21896,6 +24041,19 @@ diff -urN /usr/share/empty/fs/aufs/Kconfig linux/fs/aufs/Kconfig + shows better performance in most cases. + See detail in aufs.5. + ++config AUFS_DIRREN ++ bool "Workaround for rename(2)-ing a directory" ++ help ++ By default, aufs returns EXDEV error in renameing a dir who has ++ his child on the lower branch, since it is a bad idea to issue ++ rename(2) internally for every lower branch. But user may not ++ accept this behaviour. So here is a workaround to allow such ++ rename(2) and store some extra infromation on the writable ++ branch. Obviously this costs high (and I don't like it). ++ To use this feature, you need to enable this configuration AND ++ to specify the mount option `dirren.' ++ See details in aufs.5 and the design documents. ++ +config AUFS_SHWH + bool "Show whiteouts" + help @@ -21963,12 +24121,203 @@ diff -urN /usr/share/empty/fs/aufs/Kconfig linux/fs/aufs/Kconfig + Automatic configuration for internal use. + When aufs supports Magic SysRq, enabled automatically. +endif +diff -urN /usr/share/empty/fs/aufs/lcnt.h linux/fs/aufs/lcnt.h +--- /usr/share/empty/fs/aufs/lcnt.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux/fs/aufs/lcnt.h 2019-03-05 12:13:00.142557771 +0100 +@@ -0,0 +1,186 @@ ++/* SPDX-License-Identifier: GPL-2.0 */ ++/* ++ * Copyright (C) 2018-2019 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see . ++ */ ++ ++/* ++ * simple long counter wrapper ++ */ ++ ++#ifndef __AUFS_LCNT_H__ ++#define __AUFS_LCNT_H__ ++ ++#ifdef __KERNEL__ ++ ++#include "debug.h" ++ ++#define AuLCntATOMIC 1 ++#define AuLCntPCPUCNT 2 ++/* ++ * why does percpu_refcount require extra synchronize_rcu()s in ++ * au_br_do_free() ++ */ ++#define AuLCntPCPUREF 3 ++ ++/* #define AuLCntChosen AuLCntATOMIC */ ++#define AuLCntChosen AuLCntPCPUCNT ++/* #define AuLCntChosen AuLCntPCPUREF */ ++ ++#if AuLCntChosen == AuLCntATOMIC ++#include ++ ++typedef atomic_long_t au_lcnt_t; ++ ++static inline int au_lcnt_init(au_lcnt_t *cnt, void *release __maybe_unused) ++{ ++ atomic_long_set(cnt, 0); ++ return 0; ++} ++ ++static inline void au_lcnt_wait_for_fin(au_lcnt_t *cnt __maybe_unused) ++{ ++ /* empty */ ++} ++ ++static inline void au_lcnt_fin(au_lcnt_t *cnt __maybe_unused, ++ int do_sync __maybe_unused) ++{ ++ /* empty */ ++} ++ ++static inline void au_lcnt_inc(au_lcnt_t *cnt) ++{ ++ atomic_long_inc(cnt); ++} ++ ++static inline void au_lcnt_dec(au_lcnt_t *cnt) ++{ ++ atomic_long_dec(cnt); ++} ++ ++static inline long au_lcnt_read(au_lcnt_t *cnt, int do_rev __maybe_unused) ++{ ++ return atomic_long_read(cnt); ++} ++#endif ++ ++#if AuLCntChosen == AuLCntPCPUCNT ++#include ++ ++typedef struct percpu_counter au_lcnt_t; ++ ++static inline int au_lcnt_init(au_lcnt_t *cnt, void *release __maybe_unused) ++{ ++ return percpu_counter_init(cnt, 0, GFP_NOFS); ++} ++ ++static inline void au_lcnt_wait_for_fin(au_lcnt_t *cnt __maybe_unused) ++{ ++ /* empty */ ++} ++ ++static inline void au_lcnt_fin(au_lcnt_t *cnt, int do_sync __maybe_unused) ++{ ++ percpu_counter_destroy(cnt); ++} ++ ++static inline void au_lcnt_inc(au_lcnt_t *cnt) ++{ ++ percpu_counter_inc(cnt); ++} ++ ++static inline void au_lcnt_dec(au_lcnt_t *cnt) ++{ ++ percpu_counter_dec(cnt); ++} ++ ++static inline long au_lcnt_read(au_lcnt_t *cnt, int do_rev __maybe_unused) ++{ ++ s64 n; ++ ++ n = percpu_counter_sum(cnt); ++ BUG_ON(n < 0); ++ if (LONG_MAX != LLONG_MAX ++ && n > LONG_MAX) ++ AuWarn1("%s\n", "wrap-around"); ++ ++ return n; ++} ++#endif ++ ++#if AuLCntChosen == AuLCntPCPUREF ++#include ++ ++typedef struct percpu_ref au_lcnt_t; ++ ++static inline int au_lcnt_init(au_lcnt_t *cnt, percpu_ref_func_t *release) ++{ ++ if (!release) ++ release = percpu_ref_exit; ++ return percpu_ref_init(cnt, release, /*percpu mode*/0, GFP_NOFS); ++} ++ ++static inline void au_lcnt_wait_for_fin(au_lcnt_t *cnt __maybe_unused) ++{ ++ synchronize_rcu(); ++} ++ ++static inline void au_lcnt_fin(au_lcnt_t *cnt, int do_sync) ++{ ++ percpu_ref_kill(cnt); ++ if (do_sync) ++ au_lcnt_wait_for_fin(cnt); ++} ++ ++static inline void au_lcnt_inc(au_lcnt_t *cnt) ++{ ++ percpu_ref_get(cnt); ++} ++ ++static inline void au_lcnt_dec(au_lcnt_t *cnt) ++{ ++ percpu_ref_put(cnt); ++} ++ ++/* ++ * avoid calling this func as possible. ++ */ ++static inline long au_lcnt_read(au_lcnt_t *cnt, int do_rev) ++{ ++ long l; ++ ++ percpu_ref_switch_to_atomic_sync(cnt); ++ l = atomic_long_read(&cnt->count); ++ if (do_rev) ++ percpu_ref_switch_to_percpu(cnt); ++ ++ /* percpu_ref is initialized by 1 instead of 0 */ ++ return l - 1; ++} ++#endif ++ ++#ifdef CONFIG_AUFS_DEBUG ++#define AuLCntZero(val) do { \ ++ long l = val; \ ++ if (l) \ ++ AuDbg("%s = %ld\n", #val, l); \ ++} while (0) ++#else ++#define AuLCntZero(val) do {} while (0) ++#endif ++ ++#endif /* __KERNEL__ */ ++#endif /* __AUFS_LCNT_H__ */ diff -urN /usr/share/empty/fs/aufs/loop.c linux/fs/aufs/loop.c --- /usr/share/empty/fs/aufs/loop.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/loop.c 2017-05-06 22:16:52.781553511 +0200 -@@ -0,0 +1,147 @@ ++++ linux/fs/aufs/loop.c 2019-03-05 12:13:00.142557771 +0100 +@@ -0,0 +1,148 @@ ++// SPDX-License-Identifier: GPL-2.0 +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -22112,14 +24461,15 @@ diff -urN /usr/share/empty/fs/aufs/loop.c linux/fs/aufs/loop.c +{ + if (backing_file_func) + symbol_put(loop_backing_file); -+ au_delayed_kfree(au_warn_loopback_array); ++ au_kfree_try_rcu(au_warn_loopback_array); +} diff -urN /usr/share/empty/fs/aufs/loop.h linux/fs/aufs/loop.h --- /usr/share/empty/fs/aufs/loop.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/loop.h 2017-05-06 22:16:52.781553511 +0200 -@@ -0,0 +1,52 @@ ++++ linux/fs/aufs/loop.h 2019-07-08 09:31:38.324384703 +0200 +@@ -0,0 +1,55 @@ ++/* SPDX-License-Identifier: GPL-2.0 */ +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -22159,6 +24509,8 @@ diff -urN /usr/share/empty/fs/aufs/loop.h linux/fs/aufs/loop.h +int au_loopback_init(void); +void au_loopback_fin(void); +#else ++AuStub(struct file *, loop_backing_file, return NULL, struct super_block *sb) ++ +AuStubInt0(au_test_loopback_overlap, struct super_block *sb, + struct dentry *h_adding) +AuStubInt0(au_test_loopback_kthread, void) @@ -22172,8 +24524,9 @@ diff -urN /usr/share/empty/fs/aufs/loop.h linux/fs/aufs/loop.h +#endif /* __AUFS_LOOP_H__ */ diff -urN /usr/share/empty/fs/aufs/magic.mk linux/fs/aufs/magic.mk --- /usr/share/empty/fs/aufs/magic.mk 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/magic.mk 2017-05-06 22:16:52.781553511 +0200 -@@ -0,0 +1,30 @@ ++++ linux/fs/aufs/magic.mk 2018-06-04 09:08:09.188079511 +0200 +@@ -0,0 +1,31 @@ ++# SPDX-License-Identifier: GPL-2.0 + +# defined in ${srctree}/fs/fuse/inode.c +# tristate @@ -22206,8 +24559,9 @@ diff -urN /usr/share/empty/fs/aufs/magic.mk linux/fs/aufs/magic.mk +endif diff -urN /usr/share/empty/fs/aufs/Makefile linux/fs/aufs/Makefile --- /usr/share/empty/fs/aufs/Makefile 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/Makefile 2017-05-06 22:16:52.774886674 +0200 -@@ -0,0 +1,44 @@ ++++ linux/fs/aufs/Makefile 2018-06-04 09:08:09.181412645 +0200 +@@ -0,0 +1,46 @@ ++# SPDX-License-Identifier: GPL-2.0 + +include ${src}/magic.mk +ifeq (${CONFIG_AUFS_FS},m) @@ -22246,6 +24600,7 @@ diff -urN /usr/share/empty/fs/aufs/Makefile linux/fs/aufs/Makefile +aufs-$(CONFIG_AUFS_EXPORT) += export.o +aufs-$(CONFIG_AUFS_XATTR) += xattr.o +aufs-$(CONFIG_FS_POSIX_ACL) += posix_acl.o ++aufs-$(CONFIG_AUFS_DIRREN) += dirren.o +aufs-$(CONFIG_AUFS_FHSM) += fhsm.o +aufs-$(CONFIG_AUFS_POLL) += poll.o +aufs-$(CONFIG_AUFS_RDU) += rdu.o @@ -22254,10 +24609,11 @@ diff -urN /usr/share/empty/fs/aufs/Makefile linux/fs/aufs/Makefile +aufs-$(CONFIG_AUFS_MAGIC_SYSRQ) += sysrq.o diff -urN /usr/share/empty/fs/aufs/module.c linux/fs/aufs/module.c --- /usr/share/empty/fs/aufs/module.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/module.c 2017-05-06 22:16:52.781553511 +0200 -@@ -0,0 +1,333 @@ ++++ linux/fs/aufs/module.c 2019-03-05 12:13:00.142557771 +0100 +@@ -0,0 +1,273 @@ ++// SPDX-License-Identifier: GPL-2.0 +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -22292,7 +24648,7 @@ diff -urN /usr/share/empty/fs/aufs/module.c linux/fs/aufs/module.c + if (p) { +#if 0 /* unused */ + if (!new_sz) { -+ au_delayed_kfree(p); ++ au_kfree_rcu(p); + p = NULL; + goto out; + } @@ -22316,7 +24672,7 @@ diff -urN /usr/share/empty/fs/aufs/module.c linux/fs/aufs/module.c + if (q) { + if (p) { + memcpy(q, p, new_sz); -+ au_delayed_kfree(p); ++ au_kfree_try_rcu(p); + } + p = q; + } else @@ -22340,61 +24696,11 @@ diff -urN /usr/share/empty/fs/aufs/module.c linux/fs/aufs/module.c +/* + * aufs caches + */ -+ -+struct au_dfree au_dfree; -+ -+/* delayed free */ -+static void au_do_dfree(struct work_struct *work __maybe_unused) -+{ -+ struct llist_head *head; -+ struct llist_node *node, *next; -+ -+#define AU_CACHE_DFREE_DO_BODY(name, idx, lnode) do { \ -+ head = &au_dfree.cache[AuCache_##idx].llist; \ -+ node = llist_del_all(head); \ -+ for (; node; node = next) { \ -+ struct au_##name *p \ -+ = llist_entry(node, struct au_##name, \ -+ lnode); \ -+ next = llist_next(node); \ -+ au_cache_free_##name(p); \ -+ } \ -+ } while (0) -+ -+ AU_CACHE_DFREE_DO_BODY(dinfo, DINFO, di_lnode); -+ AU_CACHE_DFREE_DO_BODY(icntnr, ICNTNR, lnode); -+ AU_CACHE_DFREE_DO_BODY(finfo, FINFO, fi_lnode); -+ AU_CACHE_DFREE_DO_BODY(vdir, VDIR, vd_lnode); -+ AU_CACHE_DFREE_DO_BODY(vdir_dehstr, DEHSTR, lnode); -+#ifdef CONFIG_AUFS_HNOTIFY -+ AU_CACHE_DFREE_DO_BODY(hnotify, HNOTIFY, hn_lnode); -+#endif -+ -+#define AU_DFREE_DO_BODY(llist, func) do { \ -+ node = llist_del_all(llist); \ -+ for (; node; node = next) { \ -+ next = llist_next(node); \ -+ func(node); \ -+ } \ -+ } while (0) -+ -+ AU_DFREE_DO_BODY(au_dfree.llist + AU_DFREE_KFREE, kfree); -+ AU_DFREE_DO_BODY(au_dfree.llist + AU_DFREE_FREE_PAGE, au_free_page); -+ -+#undef AU_CACHE_DFREE_DO_BODY -+#undef AU_DFREE_DO_BODY -+} -+ -+AU_CACHE_DFREE_FUNC(dinfo, DINFO, di_lnode); -+AU_CACHE_DFREE_FUNC(icntnr, ICNTNR, lnode); -+AU_CACHE_DFREE_FUNC(finfo, FINFO, fi_lnode); -+AU_CACHE_DFREE_FUNC(vdir, VDIR, vd_lnode); -+AU_CACHE_DFREE_FUNC(vdir_dehstr, DEHSTR, lnode); ++struct kmem_cache *au_cache[AuCache_Last]; + +static void au_cache_fin(void) +{ + int i; -+ struct au_cache *cp; + + /* + * Make sure all delayed rcu free inodes are flushed before we @@ -22404,33 +24710,27 @@ diff -urN /usr/share/empty/fs/aufs/module.c linux/fs/aufs/module.c + + /* excluding AuCache_HNOTIFY */ + BUILD_BUG_ON(AuCache_HNOTIFY + 1 != AuCache_Last); -+ flush_delayed_work(&au_dfree.dwork); + for (i = 0; i < AuCache_HNOTIFY; i++) { -+ cp = au_dfree.cache + i; -+ AuDebugOn(!llist_empty(&cp->llist)); -+ kmem_cache_destroy(cp->cache); -+ cp->cache = NULL; ++ kmem_cache_destroy(au_cache[i]); ++ au_cache[i] = NULL; + } +} + +static int __init au_cache_init(void) +{ -+ struct au_cache *cp; -+ -+ cp = au_dfree.cache; -+ cp[AuCache_DINFO].cache = AuCacheCtor(au_dinfo, au_di_init_once); -+ if (cp[AuCache_DINFO].cache) ++ au_cache[AuCache_DINFO] = AuCacheCtor(au_dinfo, au_di_init_once); ++ if (au_cache[AuCache_DINFO]) + /* SLAB_DESTROY_BY_RCU */ -+ cp[AuCache_ICNTNR].cache = AuCacheCtor(au_icntnr, ++ au_cache[AuCache_ICNTNR] = AuCacheCtor(au_icntnr, + au_icntnr_init_once); -+ if (cp[AuCache_ICNTNR].cache) -+ cp[AuCache_FINFO].cache = AuCacheCtor(au_finfo, ++ if (au_cache[AuCache_ICNTNR]) ++ au_cache[AuCache_FINFO] = AuCacheCtor(au_finfo, + au_fi_init_once); -+ if (cp[AuCache_FINFO].cache) -+ cp[AuCache_VDIR].cache = AuCache(au_vdir); -+ if (cp[AuCache_VDIR].cache) -+ cp[AuCache_DEHSTR].cache = AuCache(au_vdir_dehstr); -+ if (cp[AuCache_DEHSTR].cache) ++ if (au_cache[AuCache_FINFO]) ++ au_cache[AuCache_VDIR] = AuCache(au_vdir); ++ if (au_cache[AuCache_VDIR]) ++ au_cache[AuCache_DEHSTR] = AuCache(au_vdir_dehstr); ++ if (au_cache[AuCache_DEHSTR]) + return 0; + + au_cache_fin(); @@ -22446,7 +24746,7 @@ diff -urN /usr/share/empty/fs/aufs/module.c linux/fs/aufs/module.c + * iterate_supers_type() doesn't protect us from + * remounting (branch management) + */ -+struct au_sphlhead au_sbilist; ++struct hlist_bl_head au_sbilist; +#endif + +/* @@ -22463,12 +24763,12 @@ diff -urN /usr/share/empty/fs/aufs/module.c linux/fs/aufs/module.c +/* this module parameter has no meaning when SYSFS is disabled */ +int sysaufs_brs = 1; +MODULE_PARM_DESC(brs, "use /fs/aufs/si_*/brN"); -+module_param_named(brs, sysaufs_brs, int, S_IRUGO); ++module_param_named(brs, sysaufs_brs, int, 0444); + +/* this module parameter has no meaning when USER_NS is disabled */ +bool au_userns; +MODULE_PARM_DESC(allow_userns, "allow unprivileged to mount under userns"); -+module_param_named(allow_userns, au_userns, bool, S_IRUGO); ++module_param_named(allow_userns, au_userns, bool, 0444); + +/* ---------------------------------------------------------------------- */ + @@ -22479,9 +24779,9 @@ diff -urN /usr/share/empty/fs/aufs/module.c linux/fs/aufs/module.c + int err; + + err = seq_path(seq, path, au_esc_chars); -+ if (err > 0) ++ if (err >= 0) + err = 0; -+ else if (err < 0) ++ else + err = -ENOMEM; + + return err; @@ -22493,7 +24793,6 @@ diff -urN /usr/share/empty/fs/aufs/module.c linux/fs/aufs/module.c +{ + int err, i; + char *p; -+ struct au_cache *cp; + + p = au_esc_chars; + for (i = 1; i <= ' '; i++) @@ -22508,15 +24807,7 @@ diff -urN /usr/share/empty/fs/aufs/module.c linux/fs/aufs/module.c + for (i = 0; i < AuIop_Last; i++) + aufs_iop_nogetattr[i].getattr = NULL; + -+ /* First, initialize au_dfree */ -+ for (i = 0; i < AuCache_Last; i++) { /* including hnotify */ -+ cp = au_dfree.cache + i; -+ cp->cache = NULL; -+ init_llist_head(&cp->llist); -+ } -+ for (i = 0; i < AU_DFREE_Last; i++) -+ init_llist_head(au_dfree.llist + i); -+ INIT_DELAYED_WORK(&au_dfree.dwork, au_do_dfree); ++ memset(au_cache, 0, sizeof(au_cache)); /* including hnotify */ + + au_sbilist_init(); + sysaufs_brs_init(); @@ -22525,9 +24816,12 @@ diff -urN /usr/share/empty/fs/aufs/module.c linux/fs/aufs/module.c + err = sysaufs_init(); + if (unlikely(err)) + goto out; -+ err = au_procfs_init(); ++ err = dbgaufs_init(); + if (unlikely(err)) + goto out_sysaufs; ++ err = au_procfs_init(); ++ if (unlikely(err)) ++ goto out_dbgaufs; + err = au_wkq_init(); + if (unlikely(err)) + goto out_procfs; @@ -22565,10 +24859,11 @@ diff -urN /usr/share/empty/fs/aufs/module.c linux/fs/aufs/module.c + au_wkq_fin(); +out_procfs: + au_procfs_fin(); ++out_dbgaufs: ++ dbgaufs_fin(); +out_sysaufs: + sysaufs_fin(); + au_dy_fin(); -+ flush_delayed_work(&au_dfree.dwork); +out: + return err; +} @@ -22582,19 +24877,20 @@ diff -urN /usr/share/empty/fs/aufs/module.c linux/fs/aufs/module.c + au_loopback_fin(); + au_wkq_fin(); + au_procfs_fin(); ++ dbgaufs_fin(); + sysaufs_fin(); + au_dy_fin(); -+ flush_delayed_work(&au_dfree.dwork); +} + +module_init(aufs_init); +module_exit(aufs_exit); diff -urN /usr/share/empty/fs/aufs/module.h linux/fs/aufs/module.h --- /usr/share/empty/fs/aufs/module.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/module.h 2017-05-06 22:16:52.781553511 +0200 -@@ -0,0 +1,156 @@ ++++ linux/fs/aufs/module.h 2019-03-05 12:13:00.142557771 +0100 +@@ -0,0 +1,166 @@ ++/* SPDX-License-Identifier: GPL-2.0 */ +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -22621,6 +24917,10 @@ diff -urN /usr/share/empty/fs/aufs/module.h linux/fs/aufs/module.h + +#include +#include "debug.h" ++#include "dentry.h" ++#include "dir.h" ++#include "file.h" ++#include "inode.h" + +struct path; +struct seq_file; @@ -22637,6 +24937,53 @@ diff -urN /usr/share/empty/fs/aufs/module.h linux/fs/aufs/module.h +void *au_kzrealloc(void *p, unsigned int nused, unsigned int new_sz, gfp_t gfp, + int may_shrink); + ++/* ++ * Comparing the size of the object with sizeof(struct rcu_head) ++ * case 1: object is always larger ++ * --> au_kfree_rcu() or au_kfree_do_rcu() ++ * case 2: object is always smaller ++ * --> au_kfree_small() ++ * case 3: object can be any size ++ * --> au_kfree_try_rcu() ++ */ ++ ++static inline void au_kfree_do_rcu(const void *p) ++{ ++ struct { ++ struct rcu_head rcu; ++ } *a = (void *)p; ++ ++ kfree_rcu(a, rcu); ++} ++ ++#define au_kfree_rcu(_p) do { \ ++ typeof(_p) p = (_p); \ ++ BUILD_BUG_ON(sizeof(*p) < sizeof(struct rcu_head)); \ ++ if (p) \ ++ au_kfree_do_rcu(p); \ ++ } while (0) ++ ++#define au_kfree_do_sz_test(sz) (sz >= sizeof(struct rcu_head)) ++#define au_kfree_sz_test(p) (p && au_kfree_do_sz_test(ksize(p))) ++ ++static inline void au_kfree_try_rcu(const void *p) ++{ ++ if (!p) ++ return; ++ if (au_kfree_sz_test(p)) ++ au_kfree_do_rcu(p); ++ else ++ kfree(p); ++} ++ ++static inline void au_kfree_small(const void *p) ++{ ++ if (!p) ++ return; ++ AuDebugOn(au_kfree_sz_test(p)); ++ kfree(p); ++} ++ +static inline int au_kmidx_sub(size_t sz, size_t new_sz) +{ +#ifndef CONFIG_SLOB @@ -22659,7 +25006,7 @@ diff -urN /usr/share/empty/fs/aufs/module.h linux/fs/aufs/module.h + +/* ---------------------------------------------------------------------- */ + -+/* kmem cache and delayed free */ ++/* kmem cache */ +enum { + AuCache_DINFO, + AuCache_ICNTNR, @@ -22670,28 +25017,7 @@ diff -urN /usr/share/empty/fs/aufs/module.h linux/fs/aufs/module.h + AuCache_Last +}; + -+enum { -+ AU_DFREE_KFREE, -+ AU_DFREE_FREE_PAGE, -+ AU_DFREE_Last -+}; -+ -+struct au_cache { -+ struct kmem_cache *cache; -+ struct llist_head llist; /* delayed free */ -+}; -+ -+/* -+ * in order to reduce the cost of the internal timer, consolidate all the -+ * delayed free works into a single delayed_work. -+ */ -+struct au_dfree { -+ struct au_cache cache[AuCache_Last]; -+ struct llist_head llist[AU_DFREE_Last]; -+ struct delayed_work dwork; -+}; -+ -+extern struct au_dfree au_dfree; ++extern struct kmem_cache *au_cache[AuCache_Last]; + +#define AuCacheFlags (SLAB_RECLAIM_ACCOUNT | SLAB_MEM_SPREAD) +#define AuCache(type) KMEM_CACHE(type, AuCacheFlags) @@ -22699,25 +25025,23 @@ diff -urN /usr/share/empty/fs/aufs/module.h linux/fs/aufs/module.h + kmem_cache_create(#type, sizeof(struct type), \ + __alignof__(struct type), AuCacheFlags, ctor) + -+#define AU_DFREE_DELAY msecs_to_jiffies(10) -+#define AU_DFREE_BODY(lnode, llist) do { \ -+ if (llist_add(lnode, llist)) \ -+ schedule_delayed_work(&au_dfree.dwork, \ -+ AU_DFREE_DELAY); \ -+ } while (0) -+#define AU_CACHE_DFREE_FUNC(name, idx, lnode) \ -+ void au_cache_dfree_##name(struct au_##name *p) \ -+ { \ -+ struct au_cache *cp = au_dfree.cache + AuCache_##idx; \ -+ AU_DFREE_BODY(&p->lnode, &cp->llist); \ -+ } -+ -+#define AuCacheFuncs(name, index) \ -+static inline struct au_##name *au_cache_alloc_##name(void) \ -+{ return kmem_cache_alloc(au_dfree.cache[AuCache_##index].cache, GFP_NOFS); } \ -+static inline void au_cache_free_##name(struct au_##name *p) \ -+{ kmem_cache_free(au_dfree.cache[AuCache_##index].cache, p); } \ -+void au_cache_dfree_##name(struct au_##name *p) ++#define AuCacheFuncs(name, index) \ ++ static inline struct au_##name *au_cache_alloc_##name(void) \ ++ { return kmem_cache_alloc(au_cache[AuCache_##index], GFP_NOFS); } \ ++ static inline void au_cache_free_##name##_norcu(struct au_##name *p) \ ++ { kmem_cache_free(au_cache[AuCache_##index], p); } \ ++ \ ++ static inline void au_cache_free_##name##_rcu_cb(struct rcu_head *rcu) \ ++ { void *p = rcu; \ ++ p -= offsetof(struct au_##name, rcu); \ ++ kmem_cache_free(au_cache[AuCache_##index], p); } \ ++ static inline void au_cache_free_##name##_rcu(struct au_##name *p) \ ++ { BUILD_BUG_ON(sizeof(struct au_##name) < sizeof(struct rcu_head)); \ ++ call_rcu(&p->rcu, au_cache_free_##name##_rcu_cb); } \ ++ \ ++ static inline void au_cache_free_##name(struct au_##name *p) \ ++ { /* au_cache_free_##name##_norcu(p); */ \ ++ au_cache_free_##name##_rcu(p); } + +AuCacheFuncs(dinfo, DINFO); +AuCacheFuncs(icntnr, ICNTNR); @@ -22728,33 +25052,15 @@ diff -urN /usr/share/empty/fs/aufs/module.h linux/fs/aufs/module.h +AuCacheFuncs(hnotify, HNOTIFY); +#endif + -+static inline void au_delayed_kfree(const void *p) -+{ -+ AuDebugOn(!p); -+ AuDebugOn(ksize(p) < sizeof(struct llist_node)); -+ -+ AU_DFREE_BODY((void *)p, au_dfree.llist + AU_DFREE_KFREE); -+} -+ -+/* cast only */ -+static inline void au_free_page(void *p) -+{ -+ free_page((unsigned long)p); -+} -+ -+static inline void au_delayed_free_page(unsigned long addr) -+{ -+ AU_DFREE_BODY((void *)addr, au_dfree.llist + AU_DFREE_FREE_PAGE); -+} -+ +#endif /* __KERNEL__ */ +#endif /* __AUFS_MODULE_H__ */ diff -urN /usr/share/empty/fs/aufs/mvdown.c linux/fs/aufs/mvdown.c --- /usr/share/empty/fs/aufs/mvdown.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/mvdown.c 2017-05-06 22:16:52.781553511 +0200 -@@ -0,0 +1,704 @@ ++++ linux/fs/aufs/mvdown.c 2019-03-05 12:13:00.142557771 +0100 +@@ -0,0 +1,706 @@ ++// SPDX-License-Identifier: GPL-2.0 +/* -+ * Copyright (C) 2011-2017 Junjiro R. Okajima ++ * Copyright (C) 2011-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -22831,7 +25137,7 @@ diff -urN /usr/share/empty/fs/aufs/mvdown.c linux/fs/aufs/mvdown.c + for (bindex++; bindex <= bbot; bindex++) { + br = au_sbr(sb, bindex); + if (au_br_fhsm(br->br_perm) -+ && (!(au_br_sb(br)->s_flags & MS_RDONLY))) ++ && !sb_rdonly(au_br_sb(br))) + return bindex; + } + else if (!(a->mvdown.flags & AUFS_MVDOWN_ROLOWER)) @@ -22843,7 +25149,7 @@ diff -urN /usr/share/empty/fs/aufs/mvdown.c linux/fs/aufs/mvdown.c + else + for (bindex++; bindex <= bbot; bindex++) { + br = au_sbr(sb, bindex); -+ if (!(au_br_sb(br)->s_flags & MS_RDONLY)) { ++ if (!sb_rdonly(au_br_sb(br))) { + if (au_br_rdonly(br)) + a->mvdown.flags + |= AUFS_MVDOWN_ROLOWER_R; @@ -23397,7 +25703,8 @@ diff -urN /usr/share/empty/fs/aufs/mvdown.c linux/fs/aufs/mvdown.c + + err = copy_from_user(&args->mvdown, uarg, sizeof(args->mvdown)); + if (!err) -+ err = !access_ok(VERIFY_WRITE, uarg, sizeof(*uarg)); ++ /* VERIFY_WRITE */ ++ err = !access_ok(uarg, sizeof(*uarg)); + if (unlikely(err)) { + err = -EFAULT; + AuTraceErr(err); @@ -23452,17 +25759,18 @@ diff -urN /usr/share/empty/fs/aufs/mvdown.c linux/fs/aufs/mvdown.c + e = copy_to_user(uarg, &args->mvdown, sizeof(args->mvdown)); + if (unlikely(e)) + err = -EFAULT; -+ au_delayed_kfree(args); ++ au_kfree_rcu(args); +out: + AuTraceErr(err); + return err; +} diff -urN /usr/share/empty/fs/aufs/opts.c linux/fs/aufs/opts.c --- /usr/share/empty/fs/aufs/opts.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/opts.c 2017-05-06 22:16:52.781553511 +0200 -@@ -0,0 +1,1848 @@ ++++ linux/fs/aufs/opts.c 2019-07-08 09:31:38.324384703 +0200 +@@ -0,0 +1,1877 @@ ++// SPDX-License-Identifier: GPL-2.0 +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -23510,6 +25818,7 @@ diff -urN /usr/share/empty/fs/aufs/opts.c linux/fs/aufs/opts.c + Opt_verbose, Opt_noverbose, + Opt_sum, Opt_nosum, Opt_wsum, + Opt_dirperm1, Opt_nodirperm1, ++ Opt_dirren, Opt_nodirren, + Opt_acl, Opt_noacl, + Opt_tail, Opt_ignore, Opt_ignore_silent, Opt_err +}; @@ -23564,10 +25873,18 @@ diff -urN /usr/share/empty/fs/aufs/opts.c linux/fs/aufs/opts.c + {Opt_dio, "dio"}, + {Opt_nodio, "nodio"}, + ++#ifdef CONFIG_AUFS_DIRREN ++ {Opt_dirren, "dirren"}, ++ {Opt_nodirren, "nodirren"}, ++#else ++ {Opt_ignore, "dirren"}, ++ {Opt_ignore_silent, "nodirren"}, ++#endif ++ +#ifdef CONFIG_AUFS_FHSM + {Opt_fhsm_sec, "fhsm_sec=%d"}, +#else -+ {Opt_ignore_silent, "fhsm_sec=%d"}, ++ {Opt_ignore, "fhsm_sec=%d"}, +#endif + + {Opt_diropq_a, "diropq=always"}, @@ -23580,7 +25897,7 @@ diff -urN /usr/share/empty/fs/aufs/opts.c linux/fs/aufs/opts.c + + /* keep them temporary */ + {Opt_ignore_silent, "nodlgt"}, -+ {Opt_ignore_silent, "clean_plink"}, ++ {Opt_ignore, "clean_plink"}, + +#ifdef CONFIG_AUFS_SHWH + {Opt_shwh, "shwh"}, @@ -23618,7 +25935,7 @@ diff -urN /usr/share/empty/fs/aufs/opts.c linux/fs/aufs/opts.c + {Opt_acl, "acl"}, + {Opt_noacl, "noacl"}, +#else -+ {Opt_ignore_silent, "acl"}, ++ {Opt_ignore, "acl"}, + {Opt_ignore_silent, "noacl"}, +#endif + @@ -24184,6 +26501,12 @@ diff -urN /usr/share/empty/fs/aufs/opts.c linux/fs/aufs/opts.c + case Opt_fhsm_sec: + AuDbg("fhsm_sec %u\n", opt->fhsm_second); + break; ++ case Opt_dirren: ++ AuLabel(dirren); ++ break; ++ case Opt_nodirren: ++ AuLabel(nodirren); ++ break; + case Opt_acl: + AuLabel(acl); + break; @@ -24249,7 +26572,7 @@ diff -urN /usr/share/empty/fs/aufs/opts.c linux/fs/aufs/opts.c + add->perm = AuBrPerm_RO; + if (au_test_fs_rr(add->path.dentry->d_sb)) + add->perm = AuBrPerm_RR; -+ else if (!bindex && !(sb_flags & MS_RDONLY)) ++ else if (!bindex && !(sb_flags & SB_RDONLY)) + add->perm = AuBrPerm_RW; + } + opt->type = Opt_add; @@ -24312,7 +26635,7 @@ diff -urN /usr/share/empty/fs/aufs/opts.c linux/fs/aufs/opts.c + mod->path = args[0].from; + p = strchr(mod->path, '='); + if (unlikely(!p)) { -+ pr_err("no permssion %s\n", args[0].from); ++ pr_err("no permission %s\n", args[0].from); + goto out; + } + @@ -24365,7 +26688,7 @@ diff -urN /usr/share/empty/fs/aufs/opts.c linux/fs/aufs/opts.c + int err; + struct file *file; + -+ file = au_xino_create(sb, args[0].from, /*silent*/0); ++ file = au_xino_create(sb, args[0].from, /*silent*/0, /*wbrtop*/0); + err = PTR_ERR(file); + if (IS_ERR(file)) + goto out; @@ -24634,6 +26957,8 @@ diff -urN /usr/share/empty/fs/aufs/opts.c linux/fs/aufs/opts.c + case Opt_wsum: + case Opt_rdblk_def: + case Opt_rdhash_def: ++ case Opt_dirren: ++ case Opt_nodirren: + case Opt_acl: + case Opt_noacl: + err = 0; @@ -24705,7 +27030,7 @@ diff -urN /usr/share/empty/fs/aufs/opts.c linux/fs/aufs/opts.c + } + } + -+ au_delayed_kfree(a); ++ au_kfree_rcu(a); + dump_opts(opts); + if (unlikely(err)) + au_opts_free(opts); @@ -24888,7 +27213,8 @@ diff -urN /usr/share/empty/fs/aufs/opts.c linux/fs/aufs/opts.c + + case Opt_trunc_xino_path: + case Opt_itrunc_xino: -+ err = au_xino_trunc(sb, opt->xino_itrunc.bindex); ++ err = au_xino_trunc(sb, opt->xino_itrunc.bindex, ++ /*idx_begin*/0); + if (!err) + err = 1; + break; @@ -24900,11 +27226,33 @@ diff -urN /usr/share/empty/fs/aufs/opts.c linux/fs/aufs/opts.c + au_fclr_opts(opts->flags, TRUNC_XIB); + break; + ++ case Opt_dirren: ++ err = 1; ++ if (!au_opt_test(sbinfo->si_mntflags, DIRREN)) { ++ err = au_dr_opt_set(sb); ++ if (!err) ++ err = 1; ++ } ++ if (err == 1) ++ au_opt_set(sbinfo->si_mntflags, DIRREN); ++ break; ++ case Opt_nodirren: ++ err = 1; ++ if (au_opt_test(sbinfo->si_mntflags, DIRREN)) { ++ err = au_dr_opt_clr(sb, au_ftest_opts(opts->flags, ++ DR_FLUSHED)); ++ if (!err) ++ err = 1; ++ } ++ if (err == 1) ++ au_opt_clr(sbinfo->si_mntflags, DIRREN); ++ break; ++ + case Opt_acl: -+ sb->s_flags |= MS_POSIXACL; ++ sb->s_flags |= SB_POSIXACL; + break; + case Opt_noacl: -+ sb->s_flags &= ~MS_POSIXACL; ++ sb->s_flags &= ~SB_POSIXACL; + break; + + default: @@ -24968,7 +27316,6 @@ diff -urN /usr/share/empty/fs/aufs/opts.c linux/fs/aufs/opts.c + } + break; + } -+ + return err; +} + @@ -24977,8 +27324,6 @@ diff -urN /usr/share/empty/fs/aufs/opts.c linux/fs/aufs/opts.c + struct au_opts *opts) +{ + int err; -+ aufs_bindex_t bbot, bindex; -+ struct dentry *root, *parent, *h_root; + + err = 0; + switch (opt->type) { @@ -24989,24 +27334,10 @@ diff -urN /usr/share/empty/fs/aufs/opts.c linux/fs/aufs/opts.c + break; + + *opt_xino = &opt->xino; -+ au_xino_brid_set(sb, -1); -+ -+ /* safe d_parent access */ -+ parent = opt->xino.file->f_path.dentry->d_parent; -+ root = sb->s_root; -+ bbot = au_sbbot(sb); -+ for (bindex = 0; bindex <= bbot; bindex++) { -+ h_root = au_h_dptr(root, bindex); -+ if (h_root == parent) { -+ au_xino_brid_set(sb, au_sbr_id(sb, bindex)); -+ break; -+ } -+ } + break; + + case Opt_noxino: + au_xino_clr(sb); -+ au_xino_brid_set(sb, -1); + *opt_xino = (void *)-1; + break; + } @@ -25032,7 +27363,7 @@ diff -urN /usr/share/empty/fs/aufs/opts.c linux/fs/aufs/opts.c + sbinfo = au_sbi(sb); + AuDebugOn(!(sbinfo->si_mntflags & AuOptMask_UDBA)); + -+ if (!(sb_flags & MS_RDONLY)) { ++ if (!(sb_flags & SB_RDONLY)) { + if (unlikely(!au_br_writable(au_sbr_perm(sb, 0)))) + pr_warn("first branch should be rw\n"); + if (unlikely(au_opt_test(sbinfo->si_mntflags, SHWH))) @@ -25065,7 +27396,7 @@ diff -urN /usr/share/empty/fs/aufs/opts.c linux/fs/aufs/opts.c + br->br_perm &= ~AuBrAttr_ICEX; +#if 0 + if ((br->br_perm & AuBrAttr_ICEX_SEC) -+ && (au_br_sb(br)->s_flags & MS_NOSEC)) ++ && (au_br_sb(br)->s_flags & SB_NOSEC)) + br->br_perm &= ~AuBrAttr_ICEX_SEC; +#endif + @@ -25129,8 +27460,7 @@ diff -urN /usr/share/empty/fs/aufs/opts.c linux/fs/aufs/opts.c + au_hn_inode_unlock(hdir); + + if (!err && do_free) { -+ if (wbr) -+ au_delayed_kfree(wbr); ++ au_kfree_rcu(wbr); + br->br_wbr = NULL; + } + } @@ -25259,7 +27589,11 @@ diff -urN /usr/share/empty/fs/aufs/opts.c linux/fs/aufs/opts.c + + SiMustWriteLock(sb); + -+ err = 0; ++ err = au_dr_opt_flush(sb); ++ if (unlikely(err)) ++ goto out; ++ au_fset_opts(opts->flags, DR_FLUSHED); ++ + dir = d_inode(sb->s_root); + sbinfo = au_sbi(sb); + opt_xino = NULL; @@ -25300,6 +27634,8 @@ diff -urN /usr/share/empty/fs/aufs/opts.c linux/fs/aufs/opts.c + au_fset_opts(opts->flags, REFRESH); + + AuDbg("status 0x%x\n", opts->flags); ++ ++out: + return err; +} + @@ -25311,10 +27647,11 @@ diff -urN /usr/share/empty/fs/aufs/opts.c linux/fs/aufs/opts.c +} diff -urN /usr/share/empty/fs/aufs/opts.h linux/fs/aufs/opts.h --- /usr/share/empty/fs/aufs/opts.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/opts.h 2017-05-06 22:16:52.781553511 +0200 -@@ -0,0 +1,213 @@ ++++ linux/fs/aufs/opts.h 2019-05-06 09:03:04.820143837 +0200 +@@ -0,0 +1,225 @@ ++/* SPDX-License-Identifier: GPL-2.0 */ +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -25342,7 +27679,6 @@ diff -urN /usr/share/empty/fs/aufs/opts.h linux/fs/aufs/opts.h +#include + +struct file; -+struct super_block; + +/* ---------------------------------------------------------------------- */ + @@ -25361,13 +27697,18 @@ diff -urN /usr/share/empty/fs/aufs/opts.h linux/fs/aufs/opts.h +#define AuOpt_SUM (1 << 10) /* summation for statfs(2) */ +#define AuOpt_SUM_W (1 << 11) /* unimplemented */ +#define AuOpt_WARN_PERM (1 << 12) /* warn when add-branch */ -+#define AuOpt_VERBOSE (1 << 13) /* busy inode when del-branch */ ++#define AuOpt_VERBOSE (1 << 13) /* print the cause of error */ +#define AuOpt_DIO (1 << 14) /* direct io */ ++#define AuOpt_DIRREN (1 << 15) /* directory rename */ + +#ifndef CONFIG_AUFS_HNOTIFY +#undef AuOpt_UDBA_HNOTIFY +#define AuOpt_UDBA_HNOTIFY 0 +#endif ++#ifndef CONFIG_AUFS_DIRREN ++#undef AuOpt_DIRREN ++#define AuOpt_DIRREN 0 ++#endif +#ifndef CONFIG_AUFS_SHWH +#undef AuOpt_SHWH +#define AuOpt_SHWH 0 @@ -25492,12 +27833,18 @@ diff -urN /usr/share/empty/fs/aufs/opts.h linux/fs/aufs/opts.h +#define AuOpts_TRUNC_XIB (1 << 2) +#define AuOpts_REFRESH_DYAOP (1 << 3) +#define AuOpts_REFRESH_IDOP (1 << 4) ++#define AuOpts_DR_FLUSHED (1 << 5) +#define au_ftest_opts(flags, name) ((flags) & AuOpts_##name) +#define au_fset_opts(flags, name) \ + do { (flags) |= AuOpts_##name; } while (0) +#define au_fclr_opts(flags, name) \ + do { (flags) &= ~AuOpts_##name; } while (0) + ++#ifndef CONFIG_AUFS_DIRREN ++#undef AuOpts_DR_FLUSHED ++#define AuOpts_DR_FLUSHED 0 ++#endif ++ +struct au_opts { + struct au_opt *opt; + int max_opt; @@ -25516,6 +27863,7 @@ diff -urN /usr/share/empty/fs/aufs/opts.h linux/fs/aufs/opts.h +const char *au_optstr_wbr_create(int wbr_create); + +void au_opts_free(struct au_opts *opts); ++struct super_block; +int au_opts_parse(struct super_block *sb, char *str, struct au_opts *opts); +int au_opts_verify(struct super_block *sb, unsigned long sb_flags, + unsigned int pending); @@ -25528,10 +27876,11 @@ diff -urN /usr/share/empty/fs/aufs/opts.h linux/fs/aufs/opts.h +#endif /* __AUFS_OPTS_H__ */ diff -urN /usr/share/empty/fs/aufs/plink.c linux/fs/aufs/plink.c --- /usr/share/empty/fs/aufs/plink.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/plink.c 2017-05-06 22:16:52.781553511 +0200 -@@ -0,0 +1,514 @@ ++++ linux/fs/aufs/plink.c 2019-03-05 12:13:00.142557771 +0100 +@@ -0,0 +1,516 @@ ++// SPDX-License-Identifier: GPL-2.0 +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -25667,7 +28016,8 @@ diff -urN /usr/share/empty/fs/aufs/plink.c linux/fs/aufs/plink.c +{ + int i; + struct au_sbinfo *sbinfo; -+ struct hlist_head *plink_hlist; ++ struct hlist_bl_head *hbl; ++ struct hlist_bl_node *pos; + struct au_icntnr *icntnr; + + SiMustAnyLock(sb); @@ -25677,11 +28027,11 @@ diff -urN /usr/share/empty/fs/aufs/plink.c linux/fs/aufs/plink.c + AuDebugOn(au_plink_maint(sb, AuLock_NOPLM)); + + for (i = 0; i < AuPlink_NHASH; i++) { -+ plink_hlist = &sbinfo->si_plink[i].head; -+ rcu_read_lock(); -+ hlist_for_each_entry_rcu(icntnr, plink_hlist, plink) ++ hbl = sbinfo->si_plink + i; ++ hlist_bl_lock(hbl); ++ hlist_bl_for_each_entry(icntnr, pos, hbl, plink) + AuDbg("%lu\n", icntnr->vfs_inode.i_ino); -+ rcu_read_unlock(); ++ hlist_bl_unlock(hbl); + } +} +#endif @@ -25691,7 +28041,8 @@ diff -urN /usr/share/empty/fs/aufs/plink.c linux/fs/aufs/plink.c +{ + int found, i; + struct au_sbinfo *sbinfo; -+ struct hlist_head *plink_hlist; ++ struct hlist_bl_head *hbl; ++ struct hlist_bl_node *pos; + struct au_icntnr *icntnr; + + sbinfo = au_sbi(inode->i_sb); @@ -25701,14 +28052,14 @@ diff -urN /usr/share/empty/fs/aufs/plink.c linux/fs/aufs/plink.c + + found = 0; + i = au_plink_hash(inode->i_ino); -+ plink_hlist = &sbinfo->si_plink[i].head; -+ rcu_read_lock(); -+ hlist_for_each_entry_rcu(icntnr, plink_hlist, plink) ++ hbl = sbinfo->si_plink + i; ++ hlist_bl_lock(hbl); ++ hlist_bl_for_each_entry(icntnr, pos, hbl, plink) + if (&icntnr->vfs_inode == inode) { + found = 1; + break; + } -+ rcu_read_unlock(); ++ hlist_bl_unlock(hbl); + return found; +} + @@ -25747,9 +28098,9 @@ diff -urN /usr/share/empty/fs/aufs/plink.c linux/fs/aufs/plink.c + struct inode *h_inode; + + h_inode = d_inode(h_parent); -+ inode_lock_nested(h_inode, AuLsc_I_CHILD2); ++ inode_lock_shared_nested(h_inode, AuLsc_I_CHILD2); + h_dentry = vfsub_lkup_one(tgtname, h_parent); -+ inode_unlock(h_inode); ++ inode_unlock_shared(h_inode); + return h_dentry; +} + @@ -25896,9 +28247,9 @@ diff -urN /usr/share/empty/fs/aufs/plink.c linux/fs/aufs/plink.c +{ + struct super_block *sb; + struct au_sbinfo *sbinfo; -+ struct hlist_head *plink_hlist; ++ struct hlist_bl_head *hbl; ++ struct hlist_bl_node *pos; + struct au_icntnr *icntnr; -+ struct au_sphlhead *sphl; + int found, err, cnt, i; + + sb = inode->i_sb; @@ -25911,12 +28262,11 @@ diff -urN /usr/share/empty/fs/aufs/plink.c linux/fs/aufs/plink.c + return; + + i = au_plink_hash(inode->i_ino); -+ sphl = sbinfo->si_plink + i; -+ plink_hlist = &sphl->head; ++ hbl = sbinfo->si_plink + i; + au_igrab(inode); + -+ spin_lock(&sphl->spin); -+ hlist_for_each_entry(icntnr, plink_hlist, plink) { ++ hlist_bl_lock(hbl); ++ hlist_bl_for_each_entry(icntnr, pos, hbl, plink) { + if (&icntnr->vfs_inode == inode) { + found = 1; + break; @@ -25924,19 +28274,19 @@ diff -urN /usr/share/empty/fs/aufs/plink.c linux/fs/aufs/plink.c + } + if (!found) { + icntnr = container_of(inode, struct au_icntnr, vfs_inode); -+ hlist_add_head_rcu(&icntnr->plink, plink_hlist); ++ hlist_bl_add_head(&icntnr->plink, hbl); + } -+ spin_unlock(&sphl->spin); ++ hlist_bl_unlock(hbl); + if (!found) { -+ cnt = au_sphl_count(sphl); -+#define msg "unexpectedly unblanced or too many pseudo-links" ++ cnt = au_hbl_count(hbl); ++#define msg "unexpectedly unbalanced or too many pseudo-links" + if (cnt > AUFS_PLINK_WARN) + AuWarn1(msg ", %d\n", cnt); +#undef msg + err = whplink(h_dentry, inode, bindex, au_sbr(sb, bindex)); + if (unlikely(err)) { + pr_warn("err %d, damaged pseudo link.\n", err); -+ au_sphl_del_rcu(&icntnr->plink, sphl); ++ au_hbl_del(&icntnr->plink, hbl); + iput(&icntnr->vfs_inode); + } + } else @@ -25948,8 +28298,8 @@ diff -urN /usr/share/empty/fs/aufs/plink.c linux/fs/aufs/plink.c +{ + int i, warned; + struct au_sbinfo *sbinfo; -+ struct hlist_head *plink_hlist; -+ struct hlist_node *tmp; ++ struct hlist_bl_head *hbl; ++ struct hlist_bl_node *pos, *tmp; + struct au_icntnr *icntnr; + + SiMustWriteLock(sb); @@ -25961,14 +28311,14 @@ diff -urN /usr/share/empty/fs/aufs/plink.c linux/fs/aufs/plink.c + /* no spin_lock since sbinfo is write-locked */ + warned = 0; + for (i = 0; i < AuPlink_NHASH; i++) { -+ plink_hlist = &sbinfo->si_plink[i].head; -+ if (!warned && verbose && !hlist_empty(plink_hlist)) { ++ hbl = sbinfo->si_plink + i; ++ if (!warned && verbose && !hlist_bl_empty(hbl)) { + pr_warn("pseudo-link is not flushed"); + warned = 1; + } -+ hlist_for_each_entry_safe(icntnr, tmp, plink_hlist, plink) ++ hlist_bl_for_each_entry_safe(icntnr, pos, tmp, hbl, plink) + iput(&icntnr->vfs_inode); -+ INIT_HLIST_HEAD(plink_hlist); ++ INIT_HLIST_BL_HEAD(hbl); + } +} + @@ -26016,8 +28366,8 @@ diff -urN /usr/share/empty/fs/aufs/plink.c linux/fs/aufs/plink.c +void au_plink_half_refresh(struct super_block *sb, aufs_bindex_t br_id) +{ + struct au_sbinfo *sbinfo; -+ struct hlist_head *plink_hlist; -+ struct hlist_node *tmp; ++ struct hlist_bl_head *hbl; ++ struct hlist_bl_node *pos, *tmp; + struct au_icntnr *icntnr; + struct inode *inode; + int i, do_put; @@ -26028,15 +28378,15 @@ diff -urN /usr/share/empty/fs/aufs/plink.c linux/fs/aufs/plink.c + AuDebugOn(!au_opt_test(au_mntflags(sb), PLINK)); + AuDebugOn(au_plink_maint(sb, AuLock_NOPLM)); + -+ /* no spin_lock since sbinfo is write-locked */ ++ /* no bit_lock since sbinfo is write-locked */ + for (i = 0; i < AuPlink_NHASH; i++) { -+ plink_hlist = &sbinfo->si_plink[i].head; -+ hlist_for_each_entry_safe(icntnr, tmp, plink_hlist, plink) { ++ hbl = sbinfo->si_plink + i; ++ hlist_bl_for_each_entry_safe(icntnr, pos, tmp, hbl, plink) { + inode = au_igrab(&icntnr->vfs_inode); + ii_write_lock_child(inode); + do_put = au_plink_do_half_refresh(inode, br_id); + if (do_put) { -+ hlist_del(&icntnr->plink); ++ hlist_bl_del(&icntnr->plink); + iput(inode); + } + ii_write_unlock(inode); @@ -26046,10 +28396,11 @@ diff -urN /usr/share/empty/fs/aufs/plink.c linux/fs/aufs/plink.c +} diff -urN /usr/share/empty/fs/aufs/poll.c linux/fs/aufs/poll.c --- /usr/share/empty/fs/aufs/poll.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/poll.c 2017-05-06 22:16:52.781553511 +0200 -@@ -0,0 +1,52 @@ ++++ linux/fs/aufs/poll.c 2019-03-05 12:13:00.142557771 +0100 +@@ -0,0 +1,51 @@ ++// SPDX-License-Identifier: GPL-2.0 +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -26072,40 +28423,39 @@ diff -urN /usr/share/empty/fs/aufs/poll.c linux/fs/aufs/poll.c + +#include "aufs.h" + -+unsigned int aufs_poll(struct file *file, poll_table *wait) ++__poll_t aufs_poll(struct file *file, struct poll_table_struct *pt) +{ -+ unsigned int mask; -+ int err; ++ __poll_t mask; + struct file *h_file; + struct super_block *sb; + + /* We should pretend an error happened. */ -+ mask = POLLERR /* | POLLIN | POLLOUT */; ++ mask = EPOLLERR /* | EPOLLIN | EPOLLOUT */; + sb = file->f_path.dentry->d_sb; + si_read_lock(sb, AuLock_FLUSH | AuLock_NOPLMW); + + h_file = au_read_pre(file, /*keep_fi*/0, /*lsc*/0); -+ err = PTR_ERR(h_file); -+ if (IS_ERR(h_file)) ++ if (IS_ERR(h_file)) { ++ AuDbg("h_file %ld\n", PTR_ERR(h_file)); + goto out; ++ } + -+ /* it is not an error if h_file has no operation */ -+ mask = DEFAULT_POLLMASK; -+ if (h_file->f_op->poll) -+ mask = h_file->f_op->poll(h_file, wait); ++ mask = vfs_poll(h_file, pt); + fput(h_file); /* instead of au_read_post() */ + +out: + si_read_unlock(sb); -+ AuTraceErr((int)mask); ++ if (mask & EPOLLERR) ++ AuDbg("mask 0x%x\n", mask); + return mask; +} diff -urN /usr/share/empty/fs/aufs/posix_acl.c linux/fs/aufs/posix_acl.c --- /usr/share/empty/fs/aufs/posix_acl.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/posix_acl.c 2017-05-06 22:16:52.781553511 +0200 -@@ -0,0 +1,102 @@ ++++ linux/fs/aufs/posix_acl.c 2019-03-05 12:13:00.142557771 +0100 +@@ -0,0 +1,103 @@ ++// SPDX-License-Identifier: GPL-2.0 +/* -+ * Copyright (C) 2014-2017 Junjiro R. Okajima ++ * Copyright (C) 2014-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -26140,7 +28490,7 @@ diff -urN /usr/share/empty/fs/aufs/posix_acl.c linux/fs/aufs/posix_acl.c + sb = inode->i_sb; + si_read_lock(sb, AuLock_FLUSH); + ii_read_lock_child(inode); -+ if (!(sb->s_flags & MS_POSIXACL)) ++ if (!(sb->s_flags & SB_POSIXACL)) + goto out; + + bindex = au_ibtop(inode); @@ -26208,10 +28558,11 @@ diff -urN /usr/share/empty/fs/aufs/posix_acl.c linux/fs/aufs/posix_acl.c +} diff -urN /usr/share/empty/fs/aufs/procfs.c linux/fs/aufs/procfs.c --- /usr/share/empty/fs/aufs/procfs.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/procfs.c 2017-05-06 22:16:52.781553511 +0200 -@@ -0,0 +1,169 @@ ++++ linux/fs/aufs/procfs.c 2019-07-08 09:31:38.324384703 +0200 +@@ -0,0 +1,171 @@ ++// SPDX-License-Identifier: GPL-2.0 +/* -+ * Copyright (C) 2010-2017 Junjiro R. Okajima ++ * Copyright (C) 2010-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -26261,6 +28612,7 @@ diff -urN /usr/share/empty/fs/aufs/procfs.c linux/fs/aufs/procfs.c + int err; + struct super_block *sb; + struct au_sbinfo *sbinfo; ++ struct hlist_bl_node *pos; + + err = -EBUSY; + if (unlikely(file->private_data)) @@ -26268,14 +28620,14 @@ diff -urN /usr/share/empty/fs/aufs/procfs.c linux/fs/aufs/procfs.c + + sb = NULL; + /* don't use au_sbilist_lock() here */ -+ spin_lock(&au_sbilist.spin); -+ hlist_for_each_entry(sbinfo, &au_sbilist.head, si_list) ++ hlist_bl_lock(&au_sbilist); ++ hlist_bl_for_each_entry(sbinfo, pos, &au_sbilist, si_list) + if (id == sysaufs_si_id(sbinfo)) { -+ kobject_get(&sbinfo->si_kobj); -+ sb = sbinfo->si_sb; ++ if (kobject_get_unless_zero(&sbinfo->si_kobj)) ++ sb = sbinfo->si_sb; + break; + } -+ spin_unlock(&au_sbilist.spin); ++ hlist_bl_unlock(&au_sbilist); + + err = -EINVAL; + if (unlikely(!sb)) @@ -26365,7 +28717,7 @@ diff -urN /usr/share/empty/fs/aufs/procfs.c linux/fs/aufs/procfs.c + if (unlikely(!au_procfs_dir)) + goto out; + -+ entry = proc_create(AUFS_PLINK_MAINT_NAME, S_IFREG | S_IWUSR, ++ entry = proc_create(AUFS_PLINK_MAINT_NAME, S_IFREG | 0200, + au_procfs_dir, &au_procfs_plm_fop); + if (unlikely(!entry)) + goto out_dir; @@ -26381,10 +28733,11 @@ diff -urN /usr/share/empty/fs/aufs/procfs.c linux/fs/aufs/procfs.c +} diff -urN /usr/share/empty/fs/aufs/rdu.c linux/fs/aufs/rdu.c --- /usr/share/empty/fs/aufs/rdu.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/rdu.c 2017-05-06 22:16:52.781553511 +0200 -@@ -0,0 +1,381 @@ ++++ linux/fs/aufs/rdu.c 2019-03-05 12:13:00.142557771 +0100 +@@ -0,0 +1,384 @@ ++// SPDX-License-Identifier: GPL-2.0 +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -26518,7 +28871,8 @@ diff -urN /usr/share/empty/fs/aufs/rdu.c linux/fs/aufs/rdu.c + struct file *h_file; + struct au_rdu_cookie *cookie = &rdu->cookie; + -+ err = !access_ok(VERIFY_WRITE, rdu->ent.e, rdu->sz); ++ /* VERIFY_WRITE */ ++ err = !access_ok(rdu->ent.e, rdu->sz); + if (unlikely(err)) { + err = -EFAULT; + AuTraceErr(err); @@ -26624,7 +28978,8 @@ diff -urN /usr/share/empty/fs/aufs/rdu.c linux/fs/aufs/rdu.c + /* unnecessary to support mmap_sem since this is a dir */ + err = copy_from_user(&ent, u->e, sizeof(ent)); + if (!err) -+ err = !access_ok(VERIFY_WRITE, &u->e->ino, sizeof(ino)); ++ /* VERIFY_WRITE */ ++ err = !access_ok(&u->e->ino, sizeof(ino)); + if (unlikely(err)) { + err = -EFAULT; + AuTraceErr(err); @@ -26766,10 +29121,11 @@ diff -urN /usr/share/empty/fs/aufs/rdu.c linux/fs/aufs/rdu.c +#endif diff -urN /usr/share/empty/fs/aufs/rwsem.h linux/fs/aufs/rwsem.h --- /usr/share/empty/fs/aufs/rwsem.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/rwsem.h 2017-05-06 22:16:52.781553511 +0200 -@@ -0,0 +1,198 @@ ++++ linux/fs/aufs/rwsem.h 2019-03-05 12:13:00.142557771 +0100 +@@ -0,0 +1,73 @@ ++/* SPDX-License-Identifier: GPL-2.0 */ +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -26796,182 +29152,57 @@ diff -urN /usr/share/empty/fs/aufs/rwsem.h linux/fs/aufs/rwsem.h + +#include "debug.h" + -+struct au_rwsem { -+ struct rw_semaphore rwsem; -+#ifdef CONFIG_AUFS_DEBUG -+ /* just for debugging, not almighty counter */ -+ atomic_t rcnt, wcnt; -+#endif -+}; -+ -+#ifdef CONFIG_LOCKDEP -+#define au_lockdep_set_name(rw) \ -+ lockdep_set_class_and_name(&(rw)->rwsem, \ -+ /*original key*/(rw)->rwsem.dep_map.key, \ -+ /*name*/#rw) -+#else -+#define au_lockdep_set_name(rw) do {} while (0) -+#endif -+ -+#ifdef CONFIG_AUFS_DEBUG -+#define AuDbgCntInit(rw) do { \ -+ atomic_set(&(rw)->rcnt, 0); \ -+ atomic_set(&(rw)->wcnt, 0); \ -+ smp_mb(); /* atomic set */ \ -+} while (0) -+ -+#define AuDbgCnt(rw, cnt) atomic_read(&(rw)->cnt) -+#define AuDbgCntInc(rw, cnt) atomic_inc(&(rw)->cnt) -+#define AuDbgCntDec(rw, cnt) WARN_ON(atomic_dec_return(&(rw)->cnt) < 0) -+#define AuDbgRcntInc(rw) AuDbgCntInc(rw, rcnt) -+#define AuDbgRcntDec(rw) AuDbgCntDec(rw, rcnt) -+#define AuDbgWcntInc(rw) AuDbgCntInc(rw, wcnt) -+#define AuDbgWcntDec(rw) AuDbgCntDec(rw, wcnt) -+#else -+#define AuDbgCnt(rw, cnt) 0 -+#define AuDbgCntInit(rw) do {} while (0) -+#define AuDbgRcntInc(rw) do {} while (0) -+#define AuDbgRcntDec(rw) do {} while (0) -+#define AuDbgWcntInc(rw) do {} while (0) -+#define AuDbgWcntDec(rw) do {} while (0) -+#endif /* CONFIG_AUFS_DEBUG */ ++/* in the future, the name 'au_rwsem' will be totally gone */ ++#define au_rwsem rw_semaphore + +/* to debug easier, do not make them inlined functions */ -+#define AuRwMustNoWaiters(rw) AuDebugOn(rwsem_is_contended(&(rw)->rwsem)) ++#define AuRwMustNoWaiters(rw) AuDebugOn(rwsem_is_contended(rw)) +/* rwsem_is_locked() is unusable */ -+#define AuRwMustReadLock(rw) AuDebugOn(AuDbgCnt(rw, rcnt) <= 0) -+#define AuRwMustWriteLock(rw) AuDebugOn(AuDbgCnt(rw, wcnt) <= 0) -+#define AuRwMustAnyLock(rw) AuDebugOn(AuDbgCnt(rw, rcnt) <= 0 \ -+ && AuDbgCnt(rw, wcnt) <= 0) -+#define AuRwDestroy(rw) AuDebugOn(AuDbgCnt(rw, rcnt) \ -+ || AuDbgCnt(rw, wcnt)) -+ -+#define au_rw_init(rw) do { \ -+ AuDbgCntInit(rw); \ -+ init_rwsem(&(rw)->rwsem); \ -+ au_lockdep_set_name(rw); \ -+ } while (0) ++#define AuRwMustReadLock(rw) AuDebugOn(!lockdep_recursing(current) \ ++ && debug_locks \ ++ && !lockdep_is_held_type(rw, 1)) ++#define AuRwMustWriteLock(rw) AuDebugOn(!lockdep_recursing(current) \ ++ && debug_locks \ ++ && !lockdep_is_held_type(rw, 0)) ++#define AuRwMustAnyLock(rw) AuDebugOn(!lockdep_recursing(current) \ ++ && debug_locks \ ++ && !lockdep_is_held(rw)) ++#define AuRwDestroy(rw) AuDebugOn(!lockdep_recursing(current) \ ++ && debug_locks \ ++ && lockdep_is_held(rw)) ++ ++#define au_rw_init(rw) init_rwsem(rw) + +#define au_rw_init_wlock(rw) do { \ + au_rw_init(rw); \ -+ down_write(&(rw)->rwsem); \ -+ AuDbgWcntInc(rw); \ ++ down_write(rw); \ + } while (0) + -+#define au_rw_init_wlock_nested(rw, lsc) do { \ -+ au_rw_init(rw); \ -+ down_write_nested(&(rw)->rwsem, lsc); \ -+ AuDbgWcntInc(rw); \ ++#define au_rw_init_wlock_nested(rw, lsc) do { \ ++ au_rw_init(rw); \ ++ down_write_nested(rw, lsc); \ + } while (0) + -+static inline void au_rw_read_lock(struct au_rwsem *rw) -+{ -+ down_read(&rw->rwsem); -+ AuDbgRcntInc(rw); -+} -+ -+static inline void au_rw_read_lock_nested(struct au_rwsem *rw, unsigned int lsc) -+{ -+ down_read_nested(&rw->rwsem, lsc); -+ AuDbgRcntInc(rw); -+} -+ -+static inline void au_rw_read_unlock(struct au_rwsem *rw) -+{ -+ AuRwMustReadLock(rw); -+ AuDbgRcntDec(rw); -+ up_read(&rw->rwsem); -+} -+ -+static inline void au_rw_dgrade_lock(struct au_rwsem *rw) -+{ -+ AuRwMustWriteLock(rw); -+ AuDbgRcntInc(rw); -+ AuDbgWcntDec(rw); -+ downgrade_write(&rw->rwsem); -+} -+ -+static inline void au_rw_write_lock(struct au_rwsem *rw) -+{ -+ down_write(&rw->rwsem); -+ AuDbgWcntInc(rw); -+} -+ -+static inline void au_rw_write_lock_nested(struct au_rwsem *rw, -+ unsigned int lsc) -+{ -+ down_write_nested(&rw->rwsem, lsc); -+ AuDbgWcntInc(rw); -+} -+ -+static inline void au_rw_write_unlock(struct au_rwsem *rw) -+{ -+ AuRwMustWriteLock(rw); -+ AuDbgWcntDec(rw); -+ up_write(&rw->rwsem); -+} -+ -+/* why is not _nested version defined */ -+static inline int au_rw_read_trylock(struct au_rwsem *rw) -+{ -+ int ret; -+ -+ ret = down_read_trylock(&rw->rwsem); -+ if (ret) -+ AuDbgRcntInc(rw); -+ return ret; -+} -+ -+static inline int au_rw_write_trylock(struct au_rwsem *rw) -+{ -+ int ret; -+ -+ ret = down_write_trylock(&rw->rwsem); -+ if (ret) -+ AuDbgWcntInc(rw); -+ return ret; -+} -+ -+#undef AuDbgCntDec -+#undef AuDbgRcntInc -+#undef AuDbgRcntDec -+#undef AuDbgWcntDec -+ -+#define AuSimpleLockRwsemFuncs(prefix, param, rwsem) \ -+static inline void prefix##_read_lock(param) \ -+{ au_rw_read_lock(rwsem); } \ -+static inline void prefix##_write_lock(param) \ -+{ au_rw_write_lock(rwsem); } \ -+static inline int prefix##_read_trylock(param) \ -+{ return au_rw_read_trylock(rwsem); } \ -+static inline int prefix##_write_trylock(param) \ -+{ return au_rw_write_trylock(rwsem); } -+/* why is not _nested version defined */ -+/* static inline void prefix##_read_trylock_nested(param, lsc) -+{ au_rw_read_trylock_nested(rwsem, lsc)); } -+static inline void prefix##_write_trylock_nestd(param, lsc) -+{ au_rw_write_trylock_nested(rwsem, lsc); } */ -+ -+#define AuSimpleUnlockRwsemFuncs(prefix, param, rwsem) \ -+static inline void prefix##_read_unlock(param) \ -+{ au_rw_read_unlock(rwsem); } \ -+static inline void prefix##_write_unlock(param) \ -+{ au_rw_write_unlock(rwsem); } \ -+static inline void prefix##_downgrade_lock(param) \ -+{ au_rw_dgrade_lock(rwsem); } -+ -+#define AuSimpleRwsemFuncs(prefix, param, rwsem) \ -+ AuSimpleLockRwsemFuncs(prefix, param, rwsem) \ -+ AuSimpleUnlockRwsemFuncs(prefix, param, rwsem) ++#define au_rw_read_lock(rw) down_read(rw) ++#define au_rw_read_lock_nested(rw, lsc) down_read_nested(rw, lsc) ++#define au_rw_read_unlock(rw) up_read(rw) ++#define au_rw_dgrade_lock(rw) downgrade_write(rw) ++#define au_rw_write_lock(rw) down_write(rw) ++#define au_rw_write_lock_nested(rw, lsc) down_write_nested(rw, lsc) ++#define au_rw_write_unlock(rw) up_write(rw) ++/* why is not _nested version defined? */ ++#define au_rw_read_trylock(rw) down_read_trylock(rw) ++#define au_rw_write_trylock(rw) down_write_trylock(rw) + +#endif /* __KERNEL__ */ +#endif /* __AUFS_RWSEM_H__ */ diff -urN /usr/share/empty/fs/aufs/sbinfo.c linux/fs/aufs/sbinfo.c --- /usr/share/empty/fs/aufs/sbinfo.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/sbinfo.c 2017-05-06 22:16:52.781553511 +0200 -@@ -0,0 +1,304 @@ ++++ linux/fs/aufs/sbinfo.c 2019-05-06 09:03:04.820143837 +0200 +@@ -0,0 +1,314 @@ ++// SPDX-License-Identifier: GPL-2.0 +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -26991,6 +29222,7 @@ diff -urN /usr/share/empty/fs/aufs/sbinfo.c linux/fs/aufs/sbinfo.c + * superblock private data + */ + ++#include +#include "aufs.h" + +/* @@ -27004,23 +29236,26 @@ diff -urN /usr/share/empty/fs/aufs/sbinfo.c linux/fs/aufs/sbinfo.c + + sbinfo = container_of(kobj, struct au_sbinfo, si_kobj); + for (i = 0; i < AuPlink_NHASH; i++) -+ AuDebugOn(!hlist_empty(&sbinfo->si_plink[i].head)); ++ AuDebugOn(!hlist_bl_empty(sbinfo->si_plink + i)); + AuDebugOn(atomic_read(&sbinfo->si_nowait.nw_len)); + -+ AuDebugOn(percpu_counter_sum(&sbinfo->si_ninodes)); -+ percpu_counter_destroy(&sbinfo->si_ninodes); -+ AuDebugOn(percpu_counter_sum(&sbinfo->si_nfiles)); -+ percpu_counter_destroy(&sbinfo->si_nfiles); ++ AuLCntZero(au_lcnt_read(&sbinfo->si_ninodes, /*do_rev*/0)); ++ au_lcnt_fin(&sbinfo->si_ninodes, /*do_sync*/0); ++ AuLCntZero(au_lcnt_read(&sbinfo->si_nfiles, /*do_rev*/0)); ++ au_lcnt_fin(&sbinfo->si_nfiles, /*do_sync*/0); + ++ dbgaufs_si_fin(sbinfo); + au_rw_write_lock(&sbinfo->si_rwsem); + au_br_free(sbinfo); + au_rw_write_unlock(&sbinfo->si_rwsem); + -+ au_delayed_kfree(sbinfo->si_branch); ++ au_kfree_try_rcu(sbinfo->si_branch); + mutex_destroy(&sbinfo->si_xib_mtx); + AuRwDestroy(&sbinfo->si_rwsem); + -+ au_delayed_kfree(sbinfo); ++ au_lcnt_wait_for_fin(&sbinfo->si_ninodes); ++ /* si_nfiles is waited too */ ++ au_kfree_rcu(sbinfo); +} + +int au_si_alloc(struct super_block *sb) @@ -27039,14 +29274,20 @@ diff -urN /usr/share/empty/fs/aufs/sbinfo.c linux/fs/aufs/sbinfo.c + goto out_sbinfo; + + err = sysaufs_si_init(sbinfo); ++ if (!err) { ++ dbgaufs_si_null(sbinfo); ++ err = dbgaufs_si_init(sbinfo); ++ if (unlikely(err)) ++ kobject_put(&sbinfo->si_kobj); ++ } + if (unlikely(err)) + goto out_br; + + au_nwt_init(&sbinfo->si_nowait); + au_rw_init_wlock(&sbinfo->si_rwsem); + -+ percpu_counter_init(&sbinfo->si_ninodes, 0, GFP_NOFS); -+ percpu_counter_init(&sbinfo->si_nfiles, 0, GFP_NOFS); ++ au_lcnt_init(&sbinfo->si_ninodes, /*release*/NULL); ++ au_lcnt_init(&sbinfo->si_nfiles, /*release*/NULL); + + sbinfo->si_bbot = -1; + sbinfo->si_last_br_id = AUFS_BRANCH_MAX / 2; @@ -27064,10 +29305,9 @@ diff -urN /usr/share/empty/fs/aufs/sbinfo.c linux/fs/aufs/sbinfo.c + sbinfo->si_xino_expire + = msecs_to_jiffies(AUFS_XINO_DEF_SEC * MSEC_PER_SEC); + mutex_init(&sbinfo->si_xib_mtx); -+ sbinfo->si_xino_brid = -1; + /* leave si_xib_last_pindex and si_xib_next_bit */ + -+ au_sphl_init(&sbinfo->si_aopen); ++ INIT_HLIST_BL_HEAD(&sbinfo->si_aopen); + + sbinfo->si_rdcache = msecs_to_jiffies(AUFS_RDCACHE_DEF * MSEC_PER_SEC); + sbinfo->si_rdblk = AUFS_RDBLK_DEF; @@ -27075,11 +29315,11 @@ diff -urN /usr/share/empty/fs/aufs/sbinfo.c linux/fs/aufs/sbinfo.c + sbinfo->si_dirwh = AUFS_DIRWH_DEF; + + for (i = 0; i < AuPlink_NHASH; i++) -+ au_sphl_init(sbinfo->si_plink + i); ++ INIT_HLIST_BL_HEAD(sbinfo->si_plink + i); + init_waitqueue_head(&sbinfo->si_plink_wq); + spin_lock_init(&sbinfo->si_plink_maint_lock); + -+ au_sphl_init(&sbinfo->si_files); ++ INIT_HLIST_BL_HEAD(&sbinfo->si_files); + + /* with getattr by default */ + sbinfo->si_iop_array = aufs_iop; @@ -27091,9 +29331,9 @@ diff -urN /usr/share/empty/fs/aufs/sbinfo.c linux/fs/aufs/sbinfo.c + return 0; /* success */ + +out_br: -+ au_delayed_kfree(sbinfo->si_branch); ++ au_kfree_try_rcu(sbinfo->si_branch); +out_sbinfo: -+ au_delayed_kfree(sbinfo); ++ au_kfree_rcu(sbinfo); +out: + return err; +} @@ -27132,7 +29372,7 @@ diff -urN /usr/share/empty/fs/aufs/sbinfo.c linux/fs/aufs/sbinfo.c + au_update_digen(sb->s_root); + inode = d_inode(sb->s_root); + au_update_iigen(inode, /*half*/0); -+ inode->i_version++; ++ inode_inc_iversion(inode); + return gen; +} + @@ -27274,129 +29514,13 @@ diff -urN /usr/share/empty/fs/aufs/sbinfo.c linux/fs/aufs/sbinfo.c + di_write_unlock2(d1, d2); + si_read_unlock(d1->d_sb); +} -diff -urN /usr/share/empty/fs/aufs/spl.h linux/fs/aufs/spl.h ---- /usr/share/empty/fs/aufs/spl.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/spl.h 2017-05-06 22:16:52.781553511 +0200 -@@ -0,0 +1,113 @@ -+/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima -+ * -+ * This program, aufs is free software; you can redistribute it and/or modify -+ * it under the terms of the GNU General Public License as published by -+ * the Free Software Foundation; either version 2 of the License, or -+ * (at your option) any later version. -+ * -+ * This program is distributed in the hope that it will be useful, -+ * but WITHOUT ANY WARRANTY; without even the implied warranty of -+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the -+ * GNU General Public License for more details. -+ * -+ * You should have received a copy of the GNU General Public License -+ * along with this program. If not, see . -+ */ -+ -+/* -+ * simple list protected by a spinlock -+ */ -+ -+#ifndef __AUFS_SPL_H__ -+#define __AUFS_SPL_H__ -+ -+#ifdef __KERNEL__ -+ -+#if 0 -+struct au_splhead { -+ spinlock_t spin; -+ struct list_head head; -+}; -+ -+static inline void au_spl_init(struct au_splhead *spl) -+{ -+ spin_lock_init(&spl->spin); -+ INIT_LIST_HEAD(&spl->head); -+} -+ -+static inline void au_spl_add(struct list_head *list, struct au_splhead *spl) -+{ -+ spin_lock(&spl->spin); -+ list_add(list, &spl->head); -+ spin_unlock(&spl->spin); -+} -+ -+static inline void au_spl_del(struct list_head *list, struct au_splhead *spl) -+{ -+ spin_lock(&spl->spin); -+ list_del(list); -+ spin_unlock(&spl->spin); -+} -+ -+static inline void au_spl_del_rcu(struct list_head *list, -+ struct au_splhead *spl) -+{ -+ spin_lock(&spl->spin); -+ list_del_rcu(list); -+ spin_unlock(&spl->spin); -+} -+#endif -+ -+/* ---------------------------------------------------------------------- */ -+ -+struct au_sphlhead { -+ spinlock_t spin; -+ struct hlist_head head; -+}; -+ -+static inline void au_sphl_init(struct au_sphlhead *sphl) -+{ -+ spin_lock_init(&sphl->spin); -+ INIT_HLIST_HEAD(&sphl->head); -+} -+ -+static inline void au_sphl_add(struct hlist_node *hlist, -+ struct au_sphlhead *sphl) -+{ -+ spin_lock(&sphl->spin); -+ hlist_add_head(hlist, &sphl->head); -+ spin_unlock(&sphl->spin); -+} -+ -+static inline void au_sphl_del(struct hlist_node *hlist, -+ struct au_sphlhead *sphl) -+{ -+ spin_lock(&sphl->spin); -+ hlist_del(hlist); -+ spin_unlock(&sphl->spin); -+} -+ -+static inline void au_sphl_del_rcu(struct hlist_node *hlist, -+ struct au_sphlhead *sphl) -+{ -+ spin_lock(&sphl->spin); -+ hlist_del_rcu(hlist); -+ spin_unlock(&sphl->spin); -+} -+ -+static inline unsigned long au_sphl_count(struct au_sphlhead *sphl) -+{ -+ unsigned long cnt; -+ struct hlist_node *pos; -+ -+ cnt = 0; -+ spin_lock(&sphl->spin); -+ hlist_for_each(pos, &sphl->head) -+ cnt++; -+ spin_unlock(&sphl->spin); -+ return cnt; -+} -+ -+#endif /* __KERNEL__ */ -+#endif /* __AUFS_SPL_H__ */ diff -urN /usr/share/empty/fs/aufs/super.c linux/fs/aufs/super.c --- /usr/share/empty/fs/aufs/super.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/super.c 2017-05-06 22:16:52.781553511 +0200 -@@ -0,0 +1,1044 @@ ++++ linux/fs/aufs/super.c 2019-05-06 09:03:04.820143837 +0200 +@@ -0,0 +1,1049 @@ ++// SPDX-License-Identifier: GPL-2.0 +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -27416,6 +29540,7 @@ diff -urN /usr/share/empty/fs/aufs/super.c linux/fs/aufs/super.c + * mount and super_block operations + */ + ++#include +#include +#include +#include @@ -27432,7 +29557,7 @@ diff -urN /usr/share/empty/fs/aufs/super.c linux/fs/aufs/super.c + c = au_cache_alloc_icntnr(); + if (c) { + au_icntnr_init(c); -+ c->vfs_inode.i_version = 1; /* sigen(sb); */ ++ inode_set_iversion(&c->vfs_inode, 1); /* sigen(sb); */ + c->iinfo.ii_hinode = NULL; + return &c->vfs_inode; + } @@ -27443,7 +29568,7 @@ diff -urN /usr/share/empty/fs/aufs/super.c linux/fs/aufs/super.c +{ + struct inode *inode = container_of(head, struct inode, i_rcu); + -+ au_cache_dfree_icntnr(container_of(inode, struct au_icntnr, vfs_inode)); ++ au_cache_free_icntnr(container_of(inode, struct au_icntnr, vfs_inode)); +} + +static void aufs_destroy_inode(struct inode *inode) @@ -27470,7 +29595,7 @@ diff -urN /usr/share/empty/fs/aufs/super.c linux/fs/aufs/super.c + if (!err) + err = au_iinfo_init(inode); + if (!err) -+ inode->i_version++; ++ inode_inc_iversion(inode); + else { + iget_failed(inode); + inode = ERR_PTR(err); @@ -27583,6 +29708,7 @@ diff -urN /usr/share/empty/fs/aufs/super.c linux/fs/aufs/super.c + struct qstr *name; + struct file *f; + struct dentry *d, *h_root; ++ struct au_branch *br; + + AuRwMustAnyLock(&sbinfo->si_rwsem); + @@ -27593,11 +29719,12 @@ diff -urN /usr/share/empty/fs/aufs/super.c linux/fs/aufs/super.c + + /* stop printing the default xino path on the first writable branch */ + h_root = NULL; -+ brid = au_xino_brid(sb); -+ if (brid >= 0) { -+ bindex = au_br_index(sb, brid); -+ h_root = au_hdentry(au_di(sb->s_root), bindex)->hd_dentry; ++ bindex = au_xi_root(sb, f->f_path.dentry); ++ if (bindex >= 0) { ++ br = au_sbr_sb(sb, bindex); ++ h_root = au_br_dentry(br); + } ++ + d = f->f_path.dentry; + name = &d->d_name; + /* safe ->d_parent because the file is unlinked */ @@ -27640,8 +29767,12 @@ diff -urN /usr/share/empty/fs/aufs/super.c linux/fs/aufs/super.c +} while (0) + + sb = dentry->d_sb; -+ if (sb->s_flags & MS_POSIXACL) ++ if (sb->s_flags & SB_POSIXACL) + seq_puts(m, ",acl"); ++#if 0 ++ if (sb->s_flags & SB_I_VERSION) ++ seq_puts(m, ",i_version"); ++#endif + + /* lock free root dinfo */ + si_noflush_read_lock(sb); @@ -27685,6 +29816,7 @@ diff -urN /usr/share/empty/fs/aufs/super.c linux/fs/aufs/super.c + + au_fhsm_show(m, sbinfo); + ++ AuBool(DIRREN, dirren); + AuBool(SUM, sum); + /* AuBool(SUM_W, wsum); */ + AuBool(WARN_PERM, warn_perm); @@ -27862,11 +29994,8 @@ diff -urN /usr/share/empty/fs/aufs/super.c linux/fs/aufs/super.c + struct au_sbinfo *sbinfo; + + sbinfo = au_sbi(sb); -+ if (!sbinfo) -+ return; -+ -+ dbgaufs_si_fin(sbinfo); -+ kobject_put(&sbinfo->si_kobj); ++ if (sbinfo) ++ kobject_put(&sbinfo->si_kobj); +} + +/* ---------------------------------------------------------------------- */ @@ -27937,7 +30066,10 @@ diff -urN /usr/share/empty/fs/aufs/super.c linux/fs/aufs/super.c + +struct inode **au_iarray_alloc(struct super_block *sb, unsigned long long *max) +{ -+ *max = au_ninodes(sb); ++ struct au_sbinfo *sbi; ++ ++ sbi = au_sbi(sb); ++ *max = au_lcnt_read(&sbi->si_ninodes, /*do_rev*/1); + return au_array_alloc(max, au_iarray_cb, sb, &sb->s_inodes); +} + @@ -28135,7 +30267,7 @@ diff -urN /usr/share/empty/fs/aufs/super.c linux/fs/aufs/super.c + AuDebugOn(sbi->si_iop_array == aufs_iop); + sbi->si_iop_array = aufs_iop; + } -+ pr_info("reset to %pf and %pf\n", ++ pr_info("reset to %ps and %ps\n", + sb->s_d_op, sbi->si_iop_array); + } + @@ -28231,7 +30363,7 @@ diff -urN /usr/share/empty/fs/aufs/super.c linux/fs/aufs/super.c +out_mtx: + inode_unlock(inode); +out_opts: -+ au_delayed_free_page((unsigned long)opts.opt); ++ free_page((unsigned long)opts.opt); +out: + err = cvt_err(err); + AuTraceErr(err); @@ -28319,7 +30451,8 @@ diff -urN /usr/share/empty/fs/aufs/super.c linux/fs/aufs/super.c + sbinfo = au_sbi(sb); + + /* all timestamps always follow the ones on the branch */ -+ sb->s_flags |= MS_NOATIME | MS_NODIRATIME; ++ sb->s_flags |= SB_NOATIME | SB_NODIRATIME; ++ sb->s_flags |= SB_I_VERSION; /* do we really need this? */ + sb->s_op = &aufs_sop; + sb->s_d_op = &aufs_dop; + sb->s_magic = AUFS_SUPER_MAGIC; @@ -28354,7 +30487,7 @@ diff -urN /usr/share/empty/fs/aufs/super.c linux/fs/aufs/super.c + au_opts_free(&opts); + if (!err && au_ftest_si(sbinfo, NO_DREVAL)) { + sb->s_d_op = &aufs_dop_noreval; -+ pr_info("%pf\n", sb->s_d_op); ++ pr_info("%ps\n", sb->s_d_op); + au_refresh_dop(root, /*force_reval*/0); + sbinfo->si_iop_array = aufs_iop_nogetattr; + au_refresh_iop(inode, /*force_getattr*/0); @@ -28368,11 +30501,10 @@ diff -urN /usr/share/empty/fs/aufs/super.c linux/fs/aufs/super.c + dput(root); + sb->s_root = NULL; +out_info: -+ dbgaufs_si_fin(sbinfo); + kobject_put(&sbinfo->si_kobj); + sb->s_fs_info = NULL; +out_opts: -+ au_delayed_free_page((unsigned long)opts.opt); ++ free_page((unsigned long)opts.opt); +out: + AuTraceErr(err); + err = cvt_err(err); @@ -28387,7 +30519,6 @@ diff -urN /usr/share/empty/fs/aufs/super.c linux/fs/aufs/super.c + void *raw_data) +{ + struct dentry *root; -+ struct super_block *sb; + + /* all timestamps always follow the ones on the branch */ + /* mnt->mnt_flags |= MNT_NOATIME | MNT_NODIRATIME; */ @@ -28395,11 +30526,7 @@ diff -urN /usr/share/empty/fs/aufs/super.c linux/fs/aufs/super.c + if (IS_ERR(root)) + goto out; + -+ sb = root->d_sb; -+ si_write_lock(sb, !AuLock_FLUSH); -+ sysaufs_brs_add(sb, 0); -+ si_write_unlock(sb); -+ au_sbilist_add(sb); ++ au_sbilist_add(root->d_sb); + +out: + return root; @@ -28423,6 +30550,7 @@ diff -urN /usr/share/empty/fs/aufs/super.c linux/fs/aufs/super.c + if (au_opt_test(sbinfo->si_mntflags, PLINK)) + au_plink_put(sb, /*verbose*/1); + au_xino_clr(sb); ++ au_dr_opt_flush(sb); + sbinfo->si_sb = NULL; + aufs_write_unlock(sb->s_root); + au_nwt_flush(&sbinfo->si_nowait); @@ -28441,10 +30569,11 @@ diff -urN /usr/share/empty/fs/aufs/super.c linux/fs/aufs/super.c +}; diff -urN /usr/share/empty/fs/aufs/super.h linux/fs/aufs/super.h --- /usr/share/empty/fs/aufs/super.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/super.h 2017-05-06 22:16:52.781553511 +0200 -@@ -0,0 +1,617 @@ ++++ linux/fs/aufs/super.h 2019-03-05 12:13:00.142557771 +0100 +@@ -0,0 +1,589 @@ ++/* SPDX-License-Identifier: GPL-2.0 */ +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -28471,8 +30600,9 @@ diff -urN /usr/share/empty/fs/aufs/super.h linux/fs/aufs/super.h + +#include +#include ++#include "hbl.h" ++#include "lcnt.h" +#include "rwsem.h" -+#include "spl.h" +#include "wkq.h" + +/* policies to select one among multiple writable branches */ @@ -28530,7 +30660,7 @@ diff -urN /usr/share/empty/fs/aufs/super.h linux/fs/aufs/super.h + struct au_nowait_tasks si_nowait; + + /* -+ * tried sb->s_umount, but failed due to the dependecy between i_mutex. ++ * tried sb->s_umount, but failed due to the dependency between i_mutex. + * rwsem for au_sbinfo is necessary. + */ + struct au_rwsem si_rwsem; @@ -28539,7 +30669,7 @@ diff -urN /usr/share/empty/fs/aufs/super.h linux/fs/aufs/super.h + * dirty approach to protect sb->sb_inodes and ->s_files (gone) from + * remount. + */ -+ struct percpu_counter si_ninodes, si_nfiles; ++ au_lcnt_t si_ninodes, si_nfiles; + + /* branch management */ + unsigned int si_generation; @@ -28576,12 +30706,14 @@ diff -urN /usr/share/empty/fs/aufs/super.h linux/fs/aufs/super.h + /* external inode number (bitmap and translation table) */ + vfs_readf_t si_xread; + vfs_writef_t si_xwrite; ++ loff_t si_ximaxent; /* max entries in a xino */ ++ + struct file *si_xib; + struct mutex si_xib_mtx; /* protect xib members */ + unsigned long *si_xib_buf; + unsigned long si_xib_last_pindex; + int si_xib_next_bit; -+ aufs_bindex_t si_xino_brid; ++ + unsigned long si_xino_jiffy; + unsigned long si_xino_expire; + /* reserved for future use */ @@ -28589,12 +30721,13 @@ diff -urN /usr/share/empty/fs/aufs/super.h linux/fs/aufs/super.h + +#ifdef CONFIG_AUFS_EXPORT + /* i_generation */ ++ /* todo: make xigen file an array to support many inode numbers */ + struct file *si_xigen; + atomic_t si_xigen_next; +#endif + -+ /* dirty trick to suppoer atomic_open */ -+ struct au_sphlhead si_aopen; ++ /* dirty trick to support atomic_open */ ++ struct hlist_bl_head si_aopen; + + /* vdir parameters */ + unsigned long si_rdcache; /* max cache time in jiffies */ @@ -28610,13 +30743,13 @@ diff -urN /usr/share/empty/fs/aufs/super.h linux/fs/aufs/super.h + unsigned int si_dirwh; + + /* pseudo_link list */ -+ struct au_sphlhead si_plink[AuPlink_NHASH]; ++ struct hlist_bl_head si_plink[AuPlink_NHASH]; + wait_queue_head_t si_plink_wq; + spinlock_t si_plink_maint_lock; + pid_t si_plink_maint_pid; + + /* file list */ -+ struct au_sphlhead si_files; ++ struct hlist_bl_head si_files; + + /* with/without getattr, brother of sb->s_d_op */ + struct inode_operations *si_iop_array; @@ -28624,7 +30757,7 @@ diff -urN /usr/share/empty/fs/aufs/super.h linux/fs/aufs/super.h + /* + * sysfs and lifetime management. + * this is not a small structure and it may be a waste of memory in case -+ * of sysfs is disabled, particulary when many aufs-es are mounted. ++ * of sysfs is disabled, particularly when many aufs-es are mounted. + * but using sysfs is majority. + */ + struct kobject si_kobj; @@ -28638,7 +30771,7 @@ diff -urN /usr/share/empty/fs/aufs/super.h linux/fs/aufs/super.h +#endif + +#ifdef CONFIG_AUFS_SBILIST -+ struct hlist_node si_list; ++ struct hlist_bl_node si_list; +#endif + + /* dirty, necessary for unmounting, sysfs and sysrq */ @@ -28649,7 +30782,7 @@ diff -urN /usr/share/empty/fs/aufs/super.h linux/fs/aufs/super.h +/* + * set true when refresh_dirs() failed at remount time. + * then try refreshing dirs at access time again. -+ * if it is false, refreshing dirs at access time is unnecesary ++ * if it is false, refreshing dirs at access time is unnecessary + */ +#define AuSi_FAILED_REFRESH_DIR 1 +#define AuSi_FHSM (1 << 1) /* fhsm is active now */ @@ -28790,7 +30923,7 @@ diff -urN /usr/share/empty/fs/aufs/super.h linux/fs/aufs/super.h +void au_export_init(struct super_block *sb); +void au_xigen_inc(struct inode *inode); +int au_xigen_new(struct inode *inode); -+int au_xigen_set(struct super_block *sb, struct file *base); ++int au_xigen_set(struct super_block *sb, struct path *path); +void au_xigen_clr(struct super_block *sb); + +static inline int au_busy_or_stale(void) @@ -28804,7 +30937,7 @@ diff -urN /usr/share/empty/fs/aufs/super.h linux/fs/aufs/super.h +AuStubVoid(au_export_init, struct super_block *sb) +AuStubVoid(au_xigen_inc, struct inode *inode) +AuStubInt0(au_xigen_new, struct inode *inode) -+AuStubInt0(au_xigen_set, struct super_block *sb, struct file *base) ++AuStubInt0(au_xigen_set, struct super_block *sb, struct path *path) +AuStubVoid(au_xigen_clr, struct super_block *sb) +AuStub(int, au_busy_or_stale, return -EBUSY, void) +#endif /* CONFIG_AUFS_EXPORT */ @@ -28813,32 +30946,32 @@ diff -urN /usr/share/empty/fs/aufs/super.h linux/fs/aufs/super.h + +#ifdef CONFIG_AUFS_SBILIST +/* module.c */ -+extern struct au_sphlhead au_sbilist; ++extern struct hlist_bl_head au_sbilist; + +static inline void au_sbilist_init(void) +{ -+ au_sphl_init(&au_sbilist); ++ INIT_HLIST_BL_HEAD(&au_sbilist); +} + +static inline void au_sbilist_add(struct super_block *sb) +{ -+ au_sphl_add(&au_sbi(sb)->si_list, &au_sbilist); ++ au_hbl_add(&au_sbi(sb)->si_list, &au_sbilist); +} + +static inline void au_sbilist_del(struct super_block *sb) +{ -+ au_sphl_del(&au_sbi(sb)->si_list, &au_sbilist); ++ au_hbl_del(&au_sbi(sb)->si_list, &au_sbilist); +} + +#ifdef CONFIG_AUFS_MAGIC_SYSRQ +static inline void au_sbilist_lock(void) +{ -+ spin_lock(&au_sbilist.spin); ++ hlist_bl_lock(&au_sbilist); +} + +static inline void au_sbilist_unlock(void) +{ -+ spin_unlock(&au_sbilist.spin); ++ hlist_bl_unlock(&au_sbilist); +} +#define AuGFP_SBILIST GFP_ATOMIC +#else @@ -28904,11 +31037,20 @@ diff -urN /usr/share/empty/fs/aufs/super.h linux/fs/aufs/super.h +/* ---------------------------------------------------------------------- */ + +/* lock superblock. mainly for entry point functions */ -+/* -+ * __si_read_lock, __si_write_lock, -+ * __si_read_unlock, __si_write_unlock, __si_downgrade_lock -+ */ -+AuSimpleRwsemFuncs(__si, struct super_block *sb, &au_sbi(sb)->si_rwsem); ++#define __si_read_lock(sb) au_rw_read_lock(&au_sbi(sb)->si_rwsem) ++#define __si_write_lock(sb) au_rw_write_lock(&au_sbi(sb)->si_rwsem) ++#define __si_read_trylock(sb) au_rw_read_trylock(&au_sbi(sb)->si_rwsem) ++#define __si_write_trylock(sb) au_rw_write_trylock(&au_sbi(sb)->si_rwsem) ++/* ++#define __si_read_trylock_nested(sb) \ ++ au_rw_read_trylock_nested(&au_sbi(sb)->si_rwsem) ++#define __si_write_trylock_nested(sb) \ ++ au_rw_write_trylock_nested(&au_sbi(sb)->si_rwsem) ++*/ ++ ++#define __si_read_unlock(sb) au_rw_read_unlock(&au_sbi(sb)->si_rwsem) ++#define __si_write_unlock(sb) au_rw_write_unlock(&au_sbi(sb)->si_rwsem) ++#define __si_downgrade_lock(sb) au_rw_dgrade_lock(&au_sbi(sb)->si_rwsem) + +#define SiMustNoWaiters(sb) AuRwMustNoWaiters(&au_sbi(sb)->si_rwsem) +#define SiMustAnyLock(sb) AuRwMustAnyLock(&au_sbi(sb)->si_rwsem) @@ -29003,42 +31145,6 @@ diff -urN /usr/share/empty/fs/aufs/super.h linux/fs/aufs/super.h + return au_sbi(sb)->si_generation; +} + -+static inline unsigned long long au_ninodes(struct super_block *sb) -+{ -+ s64 n = percpu_counter_sum(&au_sbi(sb)->si_ninodes); -+ -+ BUG_ON(n < 0); -+ return n; -+} -+ -+static inline void au_ninodes_inc(struct super_block *sb) -+{ -+ percpu_counter_inc(&au_sbi(sb)->si_ninodes); -+} -+ -+static inline void au_ninodes_dec(struct super_block *sb) -+{ -+ percpu_counter_dec(&au_sbi(sb)->si_ninodes); -+} -+ -+static inline unsigned long long au_nfiles(struct super_block *sb) -+{ -+ s64 n = percpu_counter_sum(&au_sbi(sb)->si_nfiles); -+ -+ BUG_ON(n < 0); -+ return n; -+} -+ -+static inline void au_nfiles_inc(struct super_block *sb) -+{ -+ percpu_counter_inc(&au_sbi(sb)->si_nfiles); -+} -+ -+static inline void au_nfiles_dec(struct super_block *sb) -+{ -+ percpu_counter_dec(&au_sbi(sb)->si_nfiles); -+} -+ +static inline struct au_branch *au_sbr(struct super_block *sb, + aufs_bindex_t bindex) +{ @@ -29046,26 +31152,21 @@ diff -urN /usr/share/empty/fs/aufs/super.h linux/fs/aufs/super.h + return au_sbi(sb)->si_branch[0 + bindex]; +} + -+static inline void au_xino_brid_set(struct super_block *sb, aufs_bindex_t brid) -+{ -+ SiMustWriteLock(sb); -+ au_sbi(sb)->si_xino_brid = brid; -+} -+ -+static inline aufs_bindex_t au_xino_brid(struct super_block *sb) ++static inline loff_t au_xi_maxent(struct super_block *sb) +{ + SiMustAnyLock(sb); -+ return au_sbi(sb)->si_xino_brid; ++ return au_sbi(sb)->si_ximaxent; +} + +#endif /* __KERNEL__ */ +#endif /* __AUFS_SUPER_H__ */ diff -urN /usr/share/empty/fs/aufs/sysaufs.c linux/fs/aufs/sysaufs.c --- /usr/share/empty/fs/aufs/sysaufs.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/sysaufs.c 2017-05-06 22:16:52.781553511 +0200 -@@ -0,0 +1,104 @@ ++++ linux/fs/aufs/sysaufs.c 2019-03-05 12:13:00.142557771 +0100 +@@ -0,0 +1,93 @@ ++// SPDX-License-Identifier: GPL-2.0 +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -29125,18 +31226,11 @@ diff -urN /usr/share/empty/fs/aufs/sysaufs.c linux/fs/aufs/sysaufs.c + (&sbinfo->si_kobj, &au_sbi_ktype, /*&sysaufs_kset->kobj*/NULL, + SysaufsSiNamePrefix "%lx", sysaufs_si_id(sbinfo)); + -+ dbgaufs_si_null(sbinfo); -+ if (!err) { -+ err = dbgaufs_si_init(sbinfo); -+ if (unlikely(err)) -+ kobject_put(&sbinfo->si_kobj); -+ } + return err; +} + +void sysaufs_fin(void) +{ -+ dbgaufs_fin(); + sysfs_remove_group(&sysaufs_kset->kobj, sysaufs_attr_group); + kset_unregister(sysaufs_kset); +} @@ -29157,23 +31251,19 @@ diff -urN /usr/share/empty/fs/aufs/sysaufs.c linux/fs/aufs/sysaufs.c + if (IS_ERR(sysaufs_kset)) + goto out; + err = sysfs_create_group(&sysaufs_kset->kobj, sysaufs_attr_group); -+ if (unlikely(err)) { ++ if (unlikely(err)) + kset_unregister(sysaufs_kset); -+ goto out; -+ } + -+ err = dbgaufs_init(); -+ if (unlikely(err)) -+ sysaufs_fin(); +out: + return err; +} diff -urN /usr/share/empty/fs/aufs/sysaufs.h linux/fs/aufs/sysaufs.h --- /usr/share/empty/fs/aufs/sysaufs.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/sysaufs.h 2017-05-06 22:16:52.781553511 +0200 -@@ -0,0 +1,101 @@ ++++ linux/fs/aufs/sysaufs.h 2019-03-05 12:13:00.142557771 +0100 +@@ -0,0 +1,102 @@ ++/* SPDX-License-Identifier: GPL-2.0 */ +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -29275,10 +31365,11 @@ diff -urN /usr/share/empty/fs/aufs/sysaufs.h linux/fs/aufs/sysaufs.h +#endif /* __SYSAUFS_H__ */ diff -urN /usr/share/empty/fs/aufs/sysfs.c linux/fs/aufs/sysfs.c --- /usr/share/empty/fs/aufs/sysfs.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/sysfs.c 2017-05-06 22:16:52.781553511 +0200 -@@ -0,0 +1,376 @@ ++++ linux/fs/aufs/sysfs.c 2019-03-05 12:13:00.142557771 +0100 +@@ -0,0 +1,374 @@ ++// SPDX-License-Identifier: GPL-2.0 +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -29488,7 +31579,7 @@ diff -urN /usr/share/empty/fs/aufs/sysfs.c linux/fs/aufs/sysfs.c + if (unlikely(err == PAGE_SIZE)) + err = -EFBIG; + } -+ au_delayed_kfree(seq); ++ au_kfree_rcu(seq); +out_unlock: + si_read_unlock(sb); +out: @@ -29525,7 +31616,8 @@ diff -urN /usr/share/empty/fs/aufs/sysfs.c linux/fs/aufs/sysfs.c + + sz = sizeof(*arg) - offsetof(union aufs_brinfo, path); + for (bindex = 0; bindex <= bbot; bindex++, arg++) { -+ err = !access_ok(VERIFY_WRITE, arg, sizeof(*arg)); ++ /* VERIFY_WRITE */ ++ err = !access_ok(arg, sizeof(*arg)); + if (unlikely(err)) + break; + @@ -29559,9 +31651,9 @@ diff -urN /usr/share/empty/fs/aufs/sysfs.c linux/fs/aufs/sysfs.c + err = -EFAULT; + +out_seq: -+ au_delayed_kfree(seq); ++ au_kfree_rcu(seq); +out_buf: -+ au_delayed_free_page((unsigned long)buf); ++ free_page((unsigned long)buf); +out: + si_read_unlock(sb); + return err; @@ -29592,7 +31684,7 @@ diff -urN /usr/share/empty/fs/aufs/sysfs.c linux/fs/aufs/sysfs.c + attr = &br_sysfs->attr; + sysfs_attr_init(attr); + attr->name = br_sysfs->name; -+ attr->mode = S_IRUGO; ++ attr->mode = 0444; + br_sysfs++; + } +} @@ -29605,8 +31697,6 @@ diff -urN /usr/share/empty/fs/aufs/sysfs.c linux/fs/aufs/sysfs.c + int i; + aufs_bindex_t bbot; + -+ dbgaufs_brs_del(sb, bindex); -+ + if (!sysaufs_brs) + return; + @@ -29630,8 +31720,6 @@ diff -urN /usr/share/empty/fs/aufs/sysfs.c linux/fs/aufs/sysfs.c + struct au_branch *br; + struct au_brsysfs *br_sysfs; + -+ dbgaufs_brs_add(sb, bindex); -+ + if (!sysaufs_brs) + return; + @@ -29655,10 +31743,11 @@ diff -urN /usr/share/empty/fs/aufs/sysfs.c linux/fs/aufs/sysfs.c +} diff -urN /usr/share/empty/fs/aufs/sysrq.c linux/fs/aufs/sysrq.c --- /usr/share/empty/fs/aufs/sysrq.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/sysrq.c 2017-05-06 22:16:52.781553511 +0200 -@@ -0,0 +1,157 @@ ++++ linux/fs/aufs/sysrq.c 2019-03-05 12:13:00.142557771 +0100 +@@ -0,0 +1,160 @@ ++// SPDX-License-Identifier: GPL-2.0 +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -29675,7 +31764,7 @@ diff -urN /usr/share/empty/fs/aufs/sysrq.c linux/fs/aufs/sysrq.c + */ + +/* -+ * magic sysrq hanlder ++ * magic sysrq handler + */ + +/* #include */ @@ -29689,7 +31778,8 @@ diff -urN /usr/share/empty/fs/aufs/sysrq.c linux/fs/aufs/sysrq.c + char *plevel; + struct au_sbinfo *sbinfo; + struct file *file; -+ struct au_sphlhead *files; ++ struct hlist_bl_head *files; ++ struct hlist_bl_node *pos; + struct au_finfo *finfo; + + plevel = au_plevel; @@ -29748,8 +31838,8 @@ diff -urN /usr/share/empty/fs/aufs/sysrq.c linux/fs/aufs/sysrq.c +#endif + pr("files\n"); + files = &au_sbi(sb)->si_files; -+ spin_lock(&files->spin); -+ hlist_for_each_entry(finfo, &files->head, fi_hlist) { ++ hlist_bl_lock(files); ++ hlist_bl_for_each_entry(finfo, pos, files, fi_hlist) { + umode_t mode; + + file = finfo->fi_file; @@ -29757,7 +31847,7 @@ diff -urN /usr/share/empty/fs/aufs/sysrq.c linux/fs/aufs/sysrq.c + if (!special_file(mode)) + au_dpri_file(file); + } -+ spin_unlock(&files->spin); ++ hlist_bl_unlock(files); + pr("done\n"); + +#undef pr @@ -29768,16 +31858,17 @@ diff -urN /usr/share/empty/fs/aufs/sysrq.c linux/fs/aufs/sysrq.c + +/* module parameter */ +static char *aufs_sysrq_key = "a"; -+module_param_named(sysrq, aufs_sysrq_key, charp, S_IRUGO); ++module_param_named(sysrq, aufs_sysrq_key, charp, 0444); +MODULE_PARM_DESC(sysrq, "MagicSysRq key for " AUFS_NAME); + +static void au_sysrq(int key __maybe_unused) +{ + struct au_sbinfo *sbinfo; ++ struct hlist_bl_node *pos; + + lockdep_off(); + au_sbilist_lock(); -+ hlist_for_each_entry(sbinfo, &au_sbilist.head, si_list) ++ hlist_bl_for_each_entry(sbinfo, pos, &au_sbilist, si_list) + sysrq_sb(sbinfo->si_sb); + au_sbilist_unlock(); + lockdep_on(); @@ -29816,10 +31907,11 @@ diff -urN /usr/share/empty/fs/aufs/sysrq.c linux/fs/aufs/sysrq.c +} diff -urN /usr/share/empty/fs/aufs/vdir.c linux/fs/aufs/vdir.c --- /usr/share/empty/fs/aufs/vdir.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/vdir.c 2017-05-06 22:16:52.781553511 +0200 -@@ -0,0 +1,900 @@ ++++ linux/fs/aufs/vdir.c 2019-05-06 09:03:04.820143837 +0200 +@@ -0,0 +1,896 @@ ++// SPDX-License-Identifier: GPL-2.0 +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -29839,6 +31931,7 @@ diff -urN /usr/share/empty/fs/aufs/vdir.c linux/fs/aufs/vdir.c + * virtual or vertical directory + */ + ++#include +#include "aufs.h" + +static unsigned int calc_size(int nlen) @@ -29930,7 +32023,7 @@ diff -urN /usr/share/empty/fs/aufs/vdir.c linux/fs/aufs/vdir.c + struct hlist_node *node; + + hlist_for_each_entry_safe(pos, node, head, wh_hash) -+ au_delayed_kfree(pos); ++ au_kfree_rcu(pos); +} + +static void au_nhash_de_do_free(struct hlist_head *head) @@ -29939,7 +32032,7 @@ diff -urN /usr/share/empty/fs/aufs/vdir.c linux/fs/aufs/vdir.c + struct hlist_node *node; + + hlist_for_each_entry_safe(pos, node, head, hash) -+ au_cache_dfree_vdir_dehstr(pos); ++ au_cache_free_vdir_dehstr(pos); +} + +static void au_nhash_do_free(struct au_nhash *nhash, @@ -29957,7 +32050,7 @@ diff -urN /usr/share/empty/fs/aufs/vdir.c linux/fs/aufs/vdir.c + nhash_count(head); + free(head++); + } -+ au_delayed_kfree(nhash->nh_head); ++ au_kfree_try_rcu(nhash->nh_head); +} + +void au_nhash_wh_free(struct au_nhash *whlist) @@ -30170,23 +32263,15 @@ diff -urN /usr/share/empty/fs/aufs/vdir.c linux/fs/aufs/vdir.c + +/* ---------------------------------------------------------------------- */ + -+void au_vdir_free(struct au_vdir *vdir, int atonce) ++void au_vdir_free(struct au_vdir *vdir) +{ + unsigned char **deblk; + + deblk = vdir->vd_deblk; -+ if (!atonce) { -+ while (vdir->vd_nblk--) -+ au_delayed_kfree(*deblk++); -+ au_delayed_kfree(vdir->vd_deblk); -+ au_cache_dfree_vdir(vdir); -+ } else { -+ /* not delayed */ -+ while (vdir->vd_nblk--) -+ kfree(*deblk++); -+ kfree(vdir->vd_deblk); -+ au_cache_free_vdir(vdir); -+ } ++ while (vdir->vd_nblk--) ++ au_kfree_try_rcu(*deblk++); ++ au_kfree_try_rcu(vdir->vd_deblk); ++ au_cache_free_vdir(vdir); +} + +static struct au_vdir *alloc_vdir(struct file *file) @@ -30220,10 +32305,10 @@ diff -urN /usr/share/empty/fs/aufs/vdir.c linux/fs/aufs/vdir.c + if (!err) + return vdir; /* success */ + -+ au_delayed_kfree(vdir->vd_deblk); ++ au_kfree_try_rcu(vdir->vd_deblk); + +out_free: -+ au_cache_dfree_vdir(vdir); ++ au_cache_free_vdir(vdir); +out: + vdir = ERR_PTR(err); + return vdir; @@ -30235,7 +32320,7 @@ diff -urN /usr/share/empty/fs/aufs/vdir.c linux/fs/aufs/vdir.c + union au_vdir_deblk_p p, deblk_end; + + while (vdir->vd_nblk > 1) { -+ au_delayed_kfree(vdir->vd_deblk[vdir->vd_nblk - 1]); ++ au_kfree_try_rcu(vdir->vd_deblk[vdir->vd_nblk - 1]); + /* vdir->vd_deblk[vdir->vd_nblk - 1] = NULL; */ + vdir->vd_nblk--; + } @@ -30311,6 +32396,7 @@ diff -urN /usr/share/empty/fs/aufs/vdir.c linux/fs/aufs/vdir.c + if (au_nhash_test_known_wh(&arg->whlist, name, nlen)) + goto out; /* already whiteouted */ + ++ ino = 0; /* just to suppress a warning */ + if (shwh) + arg->err = au_wh_ino(sb, arg->bindex, h_ino, d_type, + &ino); @@ -30366,7 +32452,7 @@ diff -urN /usr/share/empty/fs/aufs/vdir.c linux/fs/aufs/vdir.c + } + } + -+ au_delayed_free_page((unsigned long)o); ++ free_page((unsigned long)o); + +out: + AuTraceErr(err); @@ -30483,7 +32569,7 @@ diff -urN /usr/share/empty/fs/aufs/vdir.c linux/fs/aufs/vdir.c + err = 0; + allocated = vdir; + } else if (may_read -+ && (inode->i_version != vdir->vd_version ++ && (!inode_eq_iversion(inode, vdir->vd_version) + || time_after(jiffies, vdir->vd_jiffy + expire))) { + do_read = 1; + err = reinit_vdir(vdir); @@ -30499,13 +32585,13 @@ diff -urN /usr/share/empty/fs/aufs/vdir.c linux/fs/aufs/vdir.c + err = au_do_read_vdir(&arg); + if (!err) { + /* file->f_pos = 0; */ /* todo: ctx->pos? */ -+ vdir->vd_version = inode->i_version; ++ vdir->vd_version = inode_query_iversion(inode); + vdir->vd_last.ul = 0; + vdir->vd_last.p.deblk = vdir->vd_deblk[0]; + if (allocated) + au_set_ivdir(inode, allocated); + } else if (allocated) -+ au_vdir_free(allocated, /*atonce*/0); ++ au_vdir_free(allocated); + +out: + return err; @@ -30596,11 +32682,11 @@ diff -urN /usr/share/empty/fs/aufs/vdir.c linux/fs/aufs/vdir.c + inode = file_inode(file); + err = copy_vdir(vdir_cache, au_ivdir(inode)); + if (!err) { -+ file->f_version = inode->i_version; ++ file->f_version = inode_query_iversion(inode); + if (allocated) + au_set_fvdir_cache(file, allocated); + } else if (allocated) -+ au_vdir_free(allocated, /*atonce*/0); ++ au_vdir_free(allocated); + +out: + return err; @@ -30669,7 +32755,8 @@ diff -urN /usr/share/empty/fs/aufs/vdir.c linux/fs/aufs/vdir.c + +out: + /* smp_mb(); */ -+ AuTraceErr(!valid); ++ if (!valid) ++ AuDbg("valid %d\n", !valid); + return valid; +} + @@ -30680,10 +32767,10 @@ diff -urN /usr/share/empty/fs/aufs/vdir.c linux/fs/aufs/vdir.c + struct au_vdir *vdir_cache; + struct au_vdir_de *de; + -+ vdir_cache = au_fvdir_cache(file); + if (!seek_vdir(file, ctx)) + return 0; + ++ vdir_cache = au_fvdir_cache(file); + deblk_sz = vdir_cache->vd_deblk_sz; + while (1) { + deblk_end.deblk = vdir_cache->vd_deblk[vdir_cache->vd_last.ul]; @@ -30720,10 +32807,11 @@ diff -urN /usr/share/empty/fs/aufs/vdir.c linux/fs/aufs/vdir.c +} diff -urN /usr/share/empty/fs/aufs/vfsub.c linux/fs/aufs/vfsub.c --- /usr/share/empty/fs/aufs/vfsub.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/vfsub.c 2017-05-06 22:16:52.781553511 +0200 -@@ -0,0 +1,899 @@ ++++ linux/fs/aufs/vfsub.c 2019-03-05 12:13:00.145891204 +0100 +@@ -0,0 +1,902 @@ ++// SPDX-License-Identifier: GPL-2.0 +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -30743,26 +32831,20 @@ diff -urN /usr/share/empty/fs/aufs/vfsub.c linux/fs/aufs/vfsub.c + * sub-routines for VFS + */ + ++#include +#include +#include +#include +#include -+#ifdef CONFIG_AUFS_BR_FUSE -+#include "../fs/mount.h" -+#endif +#include "aufs.h" + +#ifdef CONFIG_AUFS_BR_FUSE +int vfsub_test_mntns(struct vfsmount *mnt, struct super_block *h_sb) +{ -+ struct nsproxy *ns; -+ + if (!au_test_fuse(h_sb) || !au_userns) + return 0; + -+ ns = current->nsproxy; -+ /* no {get,put}_nsproxy(ns) */ -+ return real_mount(mnt)->mnt_ns == ns->mnt_ns ? 0 : -EACCES; ++ return is_current_mnt_ns(mnt) ? 0 : -EACCES; +} +#endif + @@ -30840,9 +32922,10 @@ diff -urN /usr/share/empty/fs/aufs/vfsub.c linux/fs/aufs/vfsub.c + * cf. linux/fs/namei.c:do_last(), lookup_open() and atomic_open(). + */ +int vfsub_atomic_open(struct inode *dir, struct dentry *dentry, -+ struct vfsub_aopen_args *args, struct au_branch *br) ++ struct vfsub_aopen_args *args) +{ + int err; ++ struct au_branch *br = args->br; + struct file *file = args->file; + /* copied from linux/fs/namei.c:atomic_open() */ + struct dentry *const DENTRY_NOT_SET = (void *)-1UL; @@ -30854,31 +32937,37 @@ diff -urN /usr/share/empty/fs/aufs/vfsub.c linux/fs/aufs/vfsub.c + if (unlikely(err)) + goto out; + -+ args->file->f_path.dentry = DENTRY_NOT_SET; -+ args->file->f_path.mnt = au_br_mnt(br); ++ au_lcnt_inc(&br->br_nfiles); ++ file->f_path.dentry = DENTRY_NOT_SET; ++ file->f_path.mnt = au_br_mnt(br); ++ AuDbg("%ps\n", dir->i_op->atomic_open); + err = dir->i_op->atomic_open(dir, dentry, file, args->open_flag, -+ args->create_mode, args->opened); -+ if (err >= 0) { -+ /* some filesystems don't set FILE_CREATED while succeeded? */ -+ if (*args->opened & FILE_CREATED) -+ fsnotify_create(dir, dentry); -+ } else ++ args->create_mode); ++ if (unlikely(err < 0)) { ++ au_lcnt_dec(&br->br_nfiles); + goto out; ++ } + ++ /* temporary workaround for nfsv4 branch */ ++ if (au_test_nfs(dir->i_sb)) ++ nfs_mark_for_revalidate(dir); + -+ if (!err) { -+ /* todo: call VFS:may_open() here */ -+ err = open_check_o_direct(file); -+ /* todo: ima_file_check() too? */ -+ if (!err && (args->open_flag & __FMODE_EXEC)) -+ err = deny_write_access(file); -+ if (unlikely(err)) -+ /* note that the file is created and still opened */ -+ goto out; ++ if (file->f_mode & FMODE_CREATED) ++ fsnotify_create(dir, dentry); ++ if (!(file->f_mode & FMODE_OPENED)) { ++ au_lcnt_dec(&br->br_nfiles); ++ goto out; + } + -+ au_br_get(br); -+ fsnotify_open(file); ++ /* todo: call VFS:may_open() here */ ++ /* todo: ima_file_check() too? */ ++ if (!err && (args->open_flag & __FMODE_EXEC)) ++ err = deny_write_access(file); ++ if (!err) ++ fsnotify_open(file); ++ else ++ au_lcnt_dec(&br->br_nfiles); ++ /* note that the file is created and still opened */ + +out: + return err; @@ -31315,13 +33404,14 @@ diff -urN /usr/share/empty/fs/aufs/vfsub.c linux/fs/aufs/vfsub.c +{ + int err; + -+ AuDbg("%pD, ctx{%pf, %llu}\n", file, ctx->actor, ctx->pos); ++ AuDbg("%pD, ctx{%ps, %llu}\n", file, ctx->actor, ctx->pos); + + lockdep_off(); + err = iterate_dir(file, ctx); + lockdep_on(); + if (err >= 0) + vfsub_update_h_iattr(&file->f_path, /*did*/NULL); /*ignore*/ ++ + return err; +} + @@ -31623,10 +33713,11 @@ diff -urN /usr/share/empty/fs/aufs/vfsub.c linux/fs/aufs/vfsub.c +} diff -urN /usr/share/empty/fs/aufs/vfsub.h linux/fs/aufs/vfsub.h --- /usr/share/empty/fs/aufs/vfsub.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/vfsub.h 2017-05-06 22:16:52.781553511 +0200 -@@ -0,0 +1,353 @@ ++++ linux/fs/aufs/vfsub.h 2019-05-06 09:03:04.820143837 +0200 +@@ -0,0 +1,354 @@ ++/* SPDX-License-Identifier: GPL-2.0 */ +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -31660,7 +33751,7 @@ diff -urN /usr/share/empty/fs/aufs/vfsub.h linux/fs/aufs/vfsub.h +/* copied from linux/fs/internal.h */ +/* todo: BAD approach!! */ +extern void __mnt_drop_write(struct vfsmount *); -+extern int open_check_o_direct(struct file *f); ++extern struct file *alloc_empty_file(int, const struct cred *); + +/* ---------------------------------------------------------------------- */ + @@ -31698,7 +33789,7 @@ diff -urN /usr/share/empty/fs/aufs/vfsub.h linux/fs/aufs/vfsub.h + +static inline int vfsub_native_ro(struct inode *inode) +{ -+ return (inode->i_sb->s_flags & MS_RDONLY) ++ return sb_rdonly(inode->i_sb) + || IS_RDONLY(inode) + /* || IS_APPEND(inode) */ + || IS_IMMUTABLE(inode); @@ -31717,15 +33808,15 @@ diff -urN /usr/share/empty/fs/aufs/vfsub.h linux/fs/aufs/vfsub.h +int vfsub_update_h_iattr(struct path *h_path, int *did); +struct file *vfsub_dentry_open(struct path *path, int flags); +struct file *vfsub_filp_open(const char *path, int oflags, int mode); ++struct au_branch; +struct vfsub_aopen_args { -+ struct file *file; -+ unsigned int open_flag; -+ umode_t create_mode; -+ int *opened; ++ struct file *file; ++ unsigned int open_flag; ++ umode_t create_mode; ++ struct au_branch *br; +}; -+struct au_branch; +int vfsub_atomic_open(struct inode *dir, struct dentry *dentry, -+ struct vfsub_aopen_args *args, struct au_branch *br); ++ struct vfsub_aopen_args *args); +int vfsub_kern_path(const char *name, unsigned int flags, struct path *path); + +struct dentry *vfsub_lookup_one_len_unlocked(const char *name, @@ -31852,8 +33943,8 @@ diff -urN /usr/share/empty/fs/aufs/vfsub.h linux/fs/aufs/vfsub.h +} +#endif + -+static inline int vfsub_update_time(struct inode *h_inode, struct timespec *ts, -+ int flags) ++static inline int vfsub_update_time(struct inode *h_inode, ++ struct timespec64 *ts, int flags) +{ + return update_time(h_inode, ts, flags); + /* no vfsub_update_h_iattr() since we don't have struct path */ @@ -31897,13 +33988,13 @@ diff -urN /usr/share/empty/fs/aufs/vfsub.h linux/fs/aufs/vfsub.h + * re-use branch fs's ioctl(FICLONE) while aufs itself doesn't support such + * ioctl. + */ -+static inline int vfsub_clone_file_range(struct file *src, struct file *dst, -+ u64 len) ++static inline loff_t vfsub_clone_file_range(struct file *src, struct file *dst, ++ loff_t len) +{ -+ int err; ++ loff_t err; + + lockdep_off(); -+ err = vfs_clone_file_range(src, 0, dst, 0, len); ++ err = vfs_clone_file_range(src, 0, dst, 0, len, /*remap_flags*/0); + lockdep_on(); + + return err; @@ -31980,10 +34071,11 @@ diff -urN /usr/share/empty/fs/aufs/vfsub.h linux/fs/aufs/vfsub.h +#endif /* __AUFS_VFSUB_H__ */ diff -urN /usr/share/empty/fs/aufs/wbr_policy.c linux/fs/aufs/wbr_policy.c --- /usr/share/empty/fs/aufs/wbr_policy.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/wbr_policy.c 2017-05-06 22:16:52.784886929 +0200 ++++ linux/fs/aufs/wbr_policy.c 2019-03-05 12:13:00.145891204 +0100 @@ -0,0 +1,830 @@ ++// SPDX-License-Identifier: GPL-2.0 +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -32110,8 +34202,7 @@ diff -urN /usr/share/empty/fs/aufs/wbr_policy.c linux/fs/aufs/wbr_policy.c + goto out; + h_path.dentry = au_h_dptr(dentry, bdst); + h_path.mnt = au_sbr_mnt(dentry->d_sb, bdst); -+ err = vfsub_sio_mkdir(au_h_iptr(dir, bdst), &h_path, -+ S_IRWXU | S_IRUGO | S_IXUGO); ++ err = vfsub_sio_mkdir(au_h_iptr(dir, bdst), &h_path, 0755); + if (unlikely(err)) + goto out_put; + au_fset_cpdown(*flags, MADE_DIR); @@ -32445,7 +34536,7 @@ diff -urN /usr/share/empty/fs/aufs/wbr_policy.c linux/fs/aufs/wbr_policy.c + + mfs->mfsrr_bytes = bavail; + AuDbg("b%d\n", mfs->mfs_bindex); -+ au_delayed_kfree(st); ++ au_kfree_rcu(st); +} + +static int au_wbr_create_mfs(struct dentry *dentry, unsigned int flags) @@ -32814,10 +34905,11 @@ diff -urN /usr/share/empty/fs/aufs/wbr_policy.c linux/fs/aufs/wbr_policy.c +}; diff -urN /usr/share/empty/fs/aufs/whout.c linux/fs/aufs/whout.c --- /usr/share/empty/fs/aufs/whout.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/whout.c 2017-05-06 22:16:52.784886929 +0200 -@@ -0,0 +1,1061 @@ ++++ linux/fs/aufs/whout.c 2019-03-05 12:13:00.145891204 +0100 +@@ -0,0 +1,1062 @@ ++// SPDX-License-Identifier: GPL-2.0 +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -32839,7 +34931,7 @@ diff -urN /usr/share/empty/fs/aufs/whout.c linux/fs/aufs/whout.c + +#include "aufs.h" + -+#define WH_MASK S_IRUGO ++#define WH_MASK 0444 + +/* + * If a directory contains this file, then it is opaque. We start with the @@ -32980,7 +35072,7 @@ diff -urN /usr/share/empty/fs/aufs/whout.c linux/fs/aufs/whout.c + +out_name: + if (name != defname) -+ au_delayed_kfree(name); ++ au_kfree_try_rcu(name); +out: + AuTraceErrPtr(dentry); + return dentry; @@ -33132,10 +35224,10 @@ diff -urN /usr/share/empty/fs/aufs/whout.c linux/fs/aufs/whout.c + + err = -EEXIST; + if (d_is_negative(path->dentry)) { -+ int mode = S_IRWXU; ++ int mode = 0700; + + if (au_test_nfs(path->dentry->d_sb)) -+ mode |= S_IXUGO; ++ mode |= 0111; + err = vfsub_mkdir(h_dir, path, mode); + } else if (d_is_dir(path->dentry)) + err = 0; @@ -33164,7 +35256,7 @@ diff -urN /usr/share/empty/fs/aufs/whout.c linux/fs/aufs/whout.c +/* + * returns tri-state, + * minus: error, caller should print the message -+ * zero: succuess ++ * zero: success + * plus: error, caller should NOT print the message + */ +static int au_wh_init_rw_nolink(struct dentry *h_root, struct au_wbr *wbr, @@ -33417,10 +35509,10 @@ diff -urN /usr/share/empty/fs/aufs/whout.c linux/fs/aufs/whout.c +out: + if (wbr) + atomic_dec(&wbr->wbr_wh_running); -+ au_br_put(a->br); ++ au_lcnt_dec(&a->br->br_count); + si_write_unlock(a->sb); + au_nwt_done(&au_sbi(a->sb)->si_nowait); -+ au_delayed_kfree(arg); ++ au_kfree_rcu(a); + if (unlikely(err)) + AuIOErr("err %d\n", err); +} @@ -33443,12 +35535,12 @@ diff -urN /usr/share/empty/fs/aufs/whout.c linux/fs/aufs/whout.c + */ + arg->sb = sb; + arg->br = br; -+ au_br_get(br); ++ au_lcnt_inc(&br->br_count); + wkq_err = au_wkq_nowait(reinit_br_wh, arg, sb, /*flags*/0); + if (unlikely(wkq_err)) { + atomic_dec(&br->br_wbr->wbr_wh_running); -+ au_br_put(br); -+ au_delayed_kfree(arg); ++ au_lcnt_dec(&br->br_count); ++ au_kfree_rcu(arg); + } + do_dec = 0; + } @@ -33607,7 +35699,7 @@ diff -urN /usr/share/empty/fs/aufs/whout.c linux/fs/aufs/whout.c + wh_dentry = ERR_PTR(err); + if (!err) { + wh_dentry = vfsub_lkup_one(&wh_name, h_parent); -+ au_delayed_kfree(wh_name.name); ++ au_kfree_try_rcu(wh_name.name); + } + return wh_dentry; +} @@ -33683,7 +35775,7 @@ diff -urN /usr/share/empty/fs/aufs/whout.c linux/fs/aufs/whout.c + break; + } + } -+ au_delayed_free_page((unsigned long)wh_name.name); ++ free_page((unsigned long)wh_name.name); + +out: + return err; @@ -33725,7 +35817,7 @@ diff -urN /usr/share/empty/fs/aufs/whout.c linux/fs/aufs/whout.c + rdhash = AUFS_RDHASH_DEF; + err = au_nhash_alloc(&whtmp->whlist, rdhash, gfp); + if (unlikely(err)) { -+ au_delayed_kfree(whtmp); ++ au_kfree_rcu(whtmp); + whtmp = ERR_PTR(err); + } + @@ -33736,11 +35828,11 @@ diff -urN /usr/share/empty/fs/aufs/whout.c linux/fs/aufs/whout.c +void au_whtmp_rmdir_free(struct au_whtmp_rmdir *whtmp) +{ + if (whtmp->br) -+ au_br_put(whtmp->br); ++ au_lcnt_dec(&whtmp->br->br_count); + dput(whtmp->wh_dentry); + iput(whtmp->dir); + au_nhash_wh_free(&whtmp->whlist); -+ au_delayed_kfree(whtmp); ++ au_kfree_rcu(whtmp); +} + +/* @@ -33869,7 +35961,7 @@ diff -urN /usr/share/empty/fs/aufs/whout.c linux/fs/aufs/whout.c + sb = dir->i_sb; + args->dir = au_igrab(dir); + args->br = au_sbr(sb, bindex); -+ au_br_get(args->br); ++ au_lcnt_inc(&args->br->br_count); + args->wh_dentry = dget(wh_dentry); + wkq_err = au_wkq_nowait(call_rmdir_whtmp, args, sb, /*flags*/0); + if (unlikely(wkq_err)) { @@ -33879,10 +35971,11 @@ diff -urN /usr/share/empty/fs/aufs/whout.c linux/fs/aufs/whout.c +} diff -urN /usr/share/empty/fs/aufs/whout.h linux/fs/aufs/whout.h --- /usr/share/empty/fs/aufs/whout.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/whout.h 2017-05-06 22:16:52.784886929 +0200 -@@ -0,0 +1,85 @@ ++++ linux/fs/aufs/whout.h 2019-03-05 12:13:00.145891204 +0100 +@@ -0,0 +1,86 @@ ++/* SPDX-License-Identifier: GPL-2.0 */ +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -33968,10 +36061,11 @@ diff -urN /usr/share/empty/fs/aufs/whout.h linux/fs/aufs/whout.h +#endif /* __AUFS_WHOUT_H__ */ diff -urN /usr/share/empty/fs/aufs/wkq.c linux/fs/aufs/wkq.c --- /usr/share/empty/fs/aufs/wkq.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/wkq.c 2017-05-06 22:16:52.784886929 +0200 -@@ -0,0 +1,213 @@ ++++ linux/fs/aufs/wkq.c 2019-03-05 12:13:00.145891204 +0100 +@@ -0,0 +1,392 @@ ++// SPDX-License-Identifier: GPL-2.0 +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -33989,7 +36083,7 @@ diff -urN /usr/share/empty/fs/aufs/wkq.c linux/fs/aufs/wkq.c + +/* + * workqueue for asynchronous/super-io operations -+ * todo: try new dredential scheme ++ * todo: try new credential scheme + */ + +#include @@ -33999,19 +36093,186 @@ diff -urN /usr/share/empty/fs/aufs/wkq.c linux/fs/aufs/wkq.c + +static struct workqueue_struct *au_wkq; + -+struct au_wkinfo { -+ struct work_struct wk; -+ struct kobject *kobj; ++struct au_wkinfo { ++ struct work_struct wk; ++ struct kobject *kobj; ++ ++ unsigned int flags; /* see wkq.h */ ++ ++ au_wkq_func_t func; ++ void *args; ++ ++#ifdef CONFIG_LOCKDEP ++ int dont_check; ++ struct held_lock **hlock; ++#endif ++ ++ struct completion *comp; ++}; ++ ++/* ---------------------------------------------------------------------- */ ++/* ++ * Aufs passes some operations to the workqueue such as the internal copyup. ++ * This scheme looks rather unnatural for LOCKDEP debugging feature, since the ++ * job run by workqueue depends upon the locks acquired in the other task. ++ * Delegating a small operation to the workqueue, aufs passes its lockdep ++ * information too. And the job in the workqueue restores the info in order to ++ * pretend as if it acquired those locks. This is just to make LOCKDEP work ++ * correctly and expectedly. ++ */ ++ ++#ifndef CONFIG_LOCKDEP ++AuStubInt0(au_wkq_lockdep_alloc, struct au_wkinfo *wkinfo); ++AuStubVoid(au_wkq_lockdep_free, struct au_wkinfo *wkinfo); ++AuStubVoid(au_wkq_lockdep_pre, struct au_wkinfo *wkinfo); ++AuStubVoid(au_wkq_lockdep_post, struct au_wkinfo *wkinfo); ++AuStubVoid(au_wkq_lockdep_init, struct au_wkinfo *wkinfo); ++#else ++static void au_wkq_lockdep_init(struct au_wkinfo *wkinfo) ++{ ++ wkinfo->hlock = NULL; ++ wkinfo->dont_check = 0; ++} ++ ++/* ++ * 1: matched ++ * 0: unmatched ++ */ ++static int au_wkq_lockdep_test(struct lock_class_key *key, const char *name) ++{ ++ static DEFINE_SPINLOCK(spin); ++ static struct { ++ char *name; ++ struct lock_class_key *key; ++ } a[] = { ++ { .name = "&sbinfo->si_rwsem" }, ++ { .name = "&finfo->fi_rwsem" }, ++ { .name = "&dinfo->di_rwsem" }, ++ { .name = "&iinfo->ii_rwsem" } ++ }; ++ static int set; ++ int i; ++ ++ /* lockless read from 'set.' see below */ ++ if (set == ARRAY_SIZE(a)) { ++ for (i = 0; i < ARRAY_SIZE(a); i++) ++ if (a[i].key == key) ++ goto match; ++ goto unmatch; ++ } ++ ++ spin_lock(&spin); ++ if (set) ++ for (i = 0; i < ARRAY_SIZE(a); i++) ++ if (a[i].key == key) { ++ spin_unlock(&spin); ++ goto match; ++ } ++ for (i = 0; i < ARRAY_SIZE(a); i++) { ++ if (a[i].key) { ++ if (unlikely(a[i].key == key)) { /* rare but possible */ ++ spin_unlock(&spin); ++ goto match; ++ } else ++ continue; ++ } ++ if (strstr(a[i].name, name)) { ++ /* ++ * the order of these three lines is important for the ++ * lockless read above. ++ */ ++ a[i].key = key; ++ spin_unlock(&spin); ++ set++; ++ /* AuDbg("%d, %s\n", set, name); */ ++ goto match; ++ } ++ } ++ spin_unlock(&spin); ++ goto unmatch; ++ ++match: ++ return 1; ++unmatch: ++ return 0; ++} ++ ++static int au_wkq_lockdep_alloc(struct au_wkinfo *wkinfo) ++{ ++ int err, n; ++ struct task_struct *curr; ++ struct held_lock **hl, *held_locks, *p; ++ ++ err = 0; ++ curr = current; ++ wkinfo->dont_check = lockdep_recursing(curr); ++ if (wkinfo->dont_check) ++ goto out; ++ n = curr->lockdep_depth; ++ if (!n) ++ goto out; ++ ++ err = -ENOMEM; ++ wkinfo->hlock = kmalloc_array(n + 1, sizeof(*wkinfo->hlock), GFP_NOFS); ++ if (unlikely(!wkinfo->hlock)) ++ goto out; ++ ++ err = 0; ++#if 0 ++ if (0 && au_debug_test()) /* left for debugging */ ++ lockdep_print_held_locks(curr); ++#endif ++ held_locks = curr->held_locks; ++ hl = wkinfo->hlock; ++ while (n--) { ++ p = held_locks++; ++ if (au_wkq_lockdep_test(p->instance->key, p->instance->name)) ++ *hl++ = p; ++ } ++ *hl = NULL; ++ ++out: ++ return err; ++} ++ ++static void au_wkq_lockdep_free(struct au_wkinfo *wkinfo) ++{ ++ au_kfree_try_rcu(wkinfo->hlock); ++} + -+ unsigned int flags; /* see wkq.h */ ++static void au_wkq_lockdep_pre(struct au_wkinfo *wkinfo) ++{ ++ struct held_lock *p, **hl = wkinfo->hlock; ++ int subclass; + -+ au_wkq_func_t func; -+ void *args; ++ if (wkinfo->dont_check) ++ lockdep_off(); ++ if (!hl) ++ return; ++ while ((p = *hl++)) { /* assignment */ ++ subclass = lockdep_hlock_class(p)->subclass; ++ /* AuDbg("%s, %d\n", p->instance->name, subclass); */ ++ if (p->read) ++ rwsem_acquire_read(p->instance, subclass, 0, ++ /*p->acquire_ip*/_RET_IP_); ++ else ++ rwsem_acquire(p->instance, subclass, 0, ++ /*p->acquire_ip*/_RET_IP_); ++ } ++} + -+ struct completion *comp; -+}; ++static void au_wkq_lockdep_post(struct au_wkinfo *wkinfo) ++{ ++ struct held_lock *p, **hl = wkinfo->hlock; + -+/* ---------------------------------------------------------------------- */ ++ if (wkinfo->dont_check) ++ lockdep_on(); ++ if (!hl) ++ return; ++ while ((p = *hl++)) /* assignment */ ++ rwsem_release(p->instance, 0, /*p->acquire_ip*/_RET_IP_); ++} ++#endif + +static void wkq_func(struct work_struct *wk) +{ @@ -34020,13 +36281,15 @@ diff -urN /usr/share/empty/fs/aufs/wkq.c linux/fs/aufs/wkq.c + AuDebugOn(!uid_eq(current_fsuid(), GLOBAL_ROOT_UID)); + AuDebugOn(rlimit(RLIMIT_FSIZE) != RLIM_INFINITY); + ++ au_wkq_lockdep_pre(wkinfo); + wkinfo->func(wkinfo->args); ++ au_wkq_lockdep_post(wkinfo); + if (au_ftest_wkq(wkinfo->flags, WAIT)) + complete(wkinfo->comp); + else { + kobject_put(wkinfo->kobj); + module_put(THIS_MODULE); /* todo: ?? */ -+ au_delayed_kfree(wkinfo); ++ au_kfree_rcu(wkinfo); + } +} + @@ -34049,7 +36312,7 @@ diff -urN /usr/share/empty/fs/aufs/wkq.c linux/fs/aufs/wkq.c + +static void au_wkq_comp_free(struct completion *comp) +{ -+ au_delayed_kfree(comp); ++ au_kfree_rcu(comp); +} + +#else @@ -34076,7 +36339,8 @@ diff -urN /usr/share/empty/fs/aufs/wkq.c linux/fs/aufs/wkq.c + if (au_ftest_wkq(wkinfo->flags, NEST)) { + if (au_wkq_test()) { + AuWarn1("wkq from wkq, unless silly-rename on NFS," -+ " due to a dead dir by UDBA?\n"); ++ " due to a dead dir by UDBA," ++ " or async xino write?\n"); + AuDebugOn(au_ftest_wkq(wkinfo->flags, WAIT)); + } + } else @@ -34108,16 +36372,23 @@ diff -urN /usr/share/empty/fs/aufs/wkq.c linux/fs/aufs/wkq.c + }; + + err = au_wkq_comp_alloc(&wkinfo, &comp); ++ if (unlikely(err)) ++ goto out; ++ err = au_wkq_lockdep_alloc(&wkinfo); ++ if (unlikely(err)) ++ goto out_comp; + if (!err) { + au_wkq_run(&wkinfo); + /* no timeout, no interrupt */ + wait_for_completion(wkinfo.comp); -+ au_wkq_comp_free(comp); -+ destroy_work_on_stack(&wkinfo.wk); + } ++ au_wkq_lockdep_free(&wkinfo); + ++out_comp: ++ au_wkq_comp_free(comp); ++out: ++ destroy_work_on_stack(&wkinfo.wk); + return err; -+ +} + +/* @@ -34144,6 +36415,7 @@ diff -urN /usr/share/empty/fs/aufs/wkq.c linux/fs/aufs/wkq.c + wkinfo->func = func; + wkinfo->args = args; + wkinfo->comp = NULL; ++ au_wkq_lockdep_init(wkinfo); + kobject_get(wkinfo->kobj); + __module_get(THIS_MODULE); /* todo: ?? */ + @@ -34185,10 +36457,11 @@ diff -urN /usr/share/empty/fs/aufs/wkq.c linux/fs/aufs/wkq.c +} diff -urN /usr/share/empty/fs/aufs/wkq.h linux/fs/aufs/wkq.h --- /usr/share/empty/fs/aufs/wkq.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/wkq.h 2017-05-06 22:16:52.784886929 +0200 -@@ -0,0 +1,93 @@ ++++ linux/fs/aufs/wkq.h 2019-03-05 12:13:00.145891204 +0100 +@@ -0,0 +1,89 @@ ++/* SPDX-License-Identifier: GPL-2.0 */ +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -34214,7 +36487,7 @@ diff -urN /usr/share/empty/fs/aufs/wkq.h linux/fs/aufs/wkq.h + +#ifdef __KERNEL__ + -+#include ++#include + +struct super_block; + @@ -34241,11 +36514,6 @@ diff -urN /usr/share/empty/fs/aufs/wkq.h linux/fs/aufs/wkq.h +#define au_fclr_wkq(flags, name) \ + do { (flags) &= ~AuWkq_##name; } while (0) + -+#ifndef CONFIG_AUFS_HNOTIFY -+#undef AuWkq_NEST -+#define AuWkq_NEST 0 -+#endif -+ +/* wkq.c */ +int au_wkq_do_wait(unsigned int flags, au_wkq_func_t func, void *args); +int au_wkq_nowait(au_wkq_func_t func, void *args, struct super_block *sb, @@ -34282,10 +36550,11 @@ diff -urN /usr/share/empty/fs/aufs/wkq.h linux/fs/aufs/wkq.h +#endif /* __AUFS_WKQ_H__ */ diff -urN /usr/share/empty/fs/aufs/xattr.c linux/fs/aufs/xattr.c --- /usr/share/empty/fs/aufs/xattr.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/xattr.c 2017-05-06 22:16:52.784886929 +0200 -@@ -0,0 +1,357 @@ ++++ linux/fs/aufs/xattr.c 2019-03-05 12:13:00.145891204 +0100 +@@ -0,0 +1,356 @@ ++// SPDX-License-Identifier: GPL-2.0 +/* -+ * Copyright (C) 2014-2017 Junjiro R. Okajima ++ * Copyright (C) 2014-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -34402,7 +36671,7 @@ diff -urN /usr/share/empty/fs/aufs/xattr.c linux/fs/aufs/xattr.c + h_isrc = d_inode(h_src); + h_idst = d_inode(h_dst); + inode_unlock(h_idst); -+ inode_lock_nested(h_isrc, AuLsc_I_CHILD); ++ inode_lock_shared_nested(h_isrc, AuLsc_I_CHILD); + inode_lock_nested(h_idst, AuLsc_I_CHILD2); + unlocked = 0; + @@ -34428,7 +36697,7 @@ diff -urN /usr/share/empty/fs/aufs/xattr.c linux/fs/aufs/xattr.c + goto out; + err = vfs_listxattr(h_src, p, ssz); + } -+ inode_unlock(h_isrc); ++ inode_unlock_shared(h_isrc); + unlocked = 1; + AuDbg("err %d, ssz %zd\n", err, ssz); + if (unlikely(err < 0)) @@ -34464,15 +36733,13 @@ diff -urN /usr/share/empty/fs/aufs/xattr.c linux/fs/aufs/xattr.c + AuTraceErr(err); + } + -+ if (value) -+ au_delayed_kfree(value); ++ au_kfree_try_rcu(value); + +out_free: -+ if (o) -+ au_delayed_kfree(o); ++ au_kfree_try_rcu(o); +out: + if (!unlocked) -+ inode_unlock(h_isrc); ++ inode_unlock_shared(h_isrc); + AuTraceErr(err); + return err; +} @@ -34643,10 +36910,11 @@ diff -urN /usr/share/empty/fs/aufs/xattr.c linux/fs/aufs/xattr.c +} diff -urN /usr/share/empty/fs/aufs/xino.c linux/fs/aufs/xino.c --- /usr/share/empty/fs/aufs/xino.c 1970-01-01 01:00:00.000000000 +0100 -+++ linux/fs/aufs/xino.c 2017-05-06 22:16:52.784886929 +0200 -@@ -0,0 +1,1415 @@ ++++ linux/fs/aufs/xino.c 2019-07-08 09:31:38.327718137 +0200 +@@ -0,0 +1,1965 @@ ++// SPDX-License-Identifier: GPL-2.0 +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -34664,194 +36932,273 @@ diff -urN /usr/share/empty/fs/aufs/xino.c linux/fs/aufs/xino.c + +/* + * external inode number translation table and bitmap ++ * ++ * things to consider ++ * - the lifetime ++ * + au_xino object ++ * + XINO files (xino, xib, xigen) ++ * + dynamic debugfs entries (xiN) ++ * + static debugfs entries (xib, xigen) ++ * + static sysfs entry (xi_path) ++ * - several entry points to handle them. ++ * + mount(2) without xino option (default) ++ * + mount(2) with xino option ++ * + mount(2) with noxino option ++ * + umount(2) ++ * + remount with add/del branches ++ * + remount with xino/noxino options + */ + +#include +#include +#include "aufs.h" + -+/* todo: unnecessary to support mmap_sem since kernel-space? */ -+ssize_t xino_fread(vfs_readf_t func, struct file *file, void *kbuf, size_t size, -+ loff_t *pos) ++static aufs_bindex_t sbr_find_shared(struct super_block *sb, aufs_bindex_t btop, ++ aufs_bindex_t bbot, ++ struct super_block *h_sb) +{ -+ ssize_t err; -+ mm_segment_t oldfs; -+ union { -+ void *k; -+ char __user *u; -+ } buf; ++ /* todo: try binary-search if the branches are many */ ++ for (; btop <= bbot; btop++) ++ if (h_sb == au_sbr_sb(sb, btop)) ++ return btop; ++ return -1; ++} + -+ buf.k = kbuf; -+ oldfs = get_fs(); -+ set_fs(KERNEL_DS); -+ do { -+ /* todo: signal_pending? */ -+ err = func(file, buf.u, size, pos); -+ } while (err == -EAGAIN || err == -EINTR); -+ set_fs(oldfs); ++/* ++ * find another branch who is on the same filesystem of the specified ++ * branch{@btgt}. search until @bbot. ++ */ ++static aufs_bindex_t is_sb_shared(struct super_block *sb, aufs_bindex_t btgt, ++ aufs_bindex_t bbot) ++{ ++ aufs_bindex_t bindex; ++ struct super_block *tgt_sb; + -+#if 0 /* reserved for future use */ -+ if (err > 0) -+ fsnotify_access(file->f_path.dentry); -+#endif ++ tgt_sb = au_sbr_sb(sb, btgt); ++ bindex = sbr_find_shared(sb, /*btop*/0, btgt - 1, tgt_sb); ++ if (bindex < 0) ++ bindex = sbr_find_shared(sb, btgt + 1, bbot, tgt_sb); + -+ return err; ++ return bindex; +} + +/* ---------------------------------------------------------------------- */ + -+static ssize_t xino_fwrite_wkq(vfs_writef_t func, struct file *file, void *buf, -+ size_t size, loff_t *pos); ++/* ++ * stop unnecessary notify events at creating xino files ++ */ + -+static ssize_t do_xino_fwrite(vfs_writef_t func, struct file *file, void *kbuf, -+ size_t size, loff_t *pos) ++aufs_bindex_t au_xi_root(struct super_block *sb, struct dentry *dentry) +{ -+ ssize_t err; -+ mm_segment_t oldfs; -+ union { -+ void *k; -+ const char __user *u; -+ } buf; -+ int i; -+ const int prevent_endless = 10; ++ aufs_bindex_t bfound, bindex, bbot; ++ struct dentry *parent; ++ struct au_branch *br; + -+ i = 0; -+ buf.k = kbuf; -+ oldfs = get_fs(); -+ set_fs(KERNEL_DS); -+ do { -+ err = func(file, buf.u, size, pos); -+ if (err == -EINTR -+ && !au_wkq_test() -+ && fatal_signal_pending(current)) { -+ set_fs(oldfs); -+ err = xino_fwrite_wkq(func, file, kbuf, size, pos); -+ BUG_ON(err == -EINTR); -+ oldfs = get_fs(); -+ set_fs(KERNEL_DS); ++ bfound = -1; ++ parent = dentry->d_parent; /* safe d_parent access */ ++ bbot = au_sbbot(sb); ++ for (bindex = 0; bindex <= bbot; bindex++) { ++ br = au_sbr(sb, bindex); ++ if (au_br_dentry(br) == parent) { ++ bfound = bindex; ++ break; + } -+ } while (i++ < prevent_endless -+ && (err == -EAGAIN || err == -EINTR)); -+ set_fs(oldfs); -+ -+#if 0 /* reserved for future use */ -+ if (err > 0) -+ fsnotify_modify(file->f_path.dentry); -+#endif ++ } + -+ return err; ++ AuDbg("bfound b%d\n", bfound); ++ return bfound; +} + -+struct do_xino_fwrite_args { -+ ssize_t *errp; -+ vfs_writef_t func; -+ struct file *file; -+ void *buf; -+ size_t size; -+ loff_t *pos; ++struct au_xino_lock_dir { ++ struct au_hinode *hdir; ++ struct dentry *parent; ++ struct inode *dir; +}; + -+static void call_do_xino_fwrite(void *args) ++static struct dentry *au_dget_parent_lock(struct dentry *dentry, ++ unsigned int lsc) +{ -+ struct do_xino_fwrite_args *a = args; -+ *a->errp = do_xino_fwrite(a->func, a->file, a->buf, a->size, a->pos); ++ struct dentry *parent; ++ struct inode *dir; ++ ++ parent = dget_parent(dentry); ++ dir = d_inode(parent); ++ inode_lock_nested(dir, lsc); ++#if 0 /* it should not happen */ ++ spin_lock(&dentry->d_lock); ++ if (unlikely(dentry->d_parent != parent)) { ++ spin_unlock(&dentry->d_lock); ++ inode_unlock(dir); ++ dput(parent); ++ parent = NULL; ++ goto out; ++ } ++ spin_unlock(&dentry->d_lock); ++ ++out: ++#endif ++ return parent; +} + -+static ssize_t xino_fwrite_wkq(vfs_writef_t func, struct file *file, void *buf, -+ size_t size, loff_t *pos) ++static void au_xino_lock_dir(struct super_block *sb, struct path *xipath, ++ struct au_xino_lock_dir *ldir) +{ -+ ssize_t err; -+ int wkq_err; -+ struct do_xino_fwrite_args args = { -+ .errp = &err, -+ .func = func, -+ .file = file, -+ .buf = buf, -+ .size = size, -+ .pos = pos -+ }; ++ aufs_bindex_t bindex; + -+ /* -+ * it breaks RLIMIT_FSIZE and normal user's limit, -+ * users should care about quota and real 'filesystem full.' -+ */ -+ wkq_err = au_wkq_wait(call_do_xino_fwrite, &args); -+ if (unlikely(wkq_err)) -+ err = wkq_err; ++ ldir->hdir = NULL; ++ bindex = au_xi_root(sb, xipath->dentry); ++ if (bindex >= 0) { ++ /* rw branch root */ ++ ldir->hdir = au_hi(d_inode(sb->s_root), bindex); ++ au_hn_inode_lock_nested(ldir->hdir, AuLsc_I_PARENT); ++ } else { ++ /* other */ ++ ldir->parent = au_dget_parent_lock(xipath->dentry, ++ AuLsc_I_PARENT); ++ ldir->dir = d_inode(ldir->parent); ++ } ++} + -+ return err; ++static void au_xino_unlock_dir(struct au_xino_lock_dir *ldir) ++{ ++ if (ldir->hdir) ++ au_hn_inode_unlock(ldir->hdir); ++ else { ++ inode_unlock(ldir->dir); ++ dput(ldir->parent); ++ } +} + -+ssize_t xino_fwrite(vfs_writef_t func, struct file *file, void *buf, -+ size_t size, loff_t *pos) ++/* ---------------------------------------------------------------------- */ ++ ++/* ++ * create and set a new xino file ++ */ ++struct file *au_xino_create(struct super_block *sb, char *fpath, int silent, ++ int wbrtop) +{ -+ ssize_t err; ++ struct file *file; ++ struct dentry *h_parent, *d; ++ struct inode *h_dir, *inode; ++ int err; ++ static DEFINE_MUTEX(mtx); + -+ if (rlimit(RLIMIT_FSIZE) == RLIM_INFINITY) { -+ lockdep_off(); -+ err = do_xino_fwrite(func, file, buf, size, pos); -+ lockdep_on(); -+ } else -+ err = xino_fwrite_wkq(func, file, buf, size, pos); ++ /* ++ * at mount-time, and the xino file is the default path, ++ * hnotify is disabled so we have no notify events to ignore. ++ * when a user specified the xino, we cannot get au_hdir to be ignored. ++ */ ++ if (!wbrtop) ++ mutex_lock(&mtx); ++ file = vfsub_filp_open(fpath, O_RDWR | O_CREAT | O_EXCL | O_LARGEFILE ++ /* | __FMODE_NONOTIFY */, ++ 0666); ++ if (IS_ERR(file)) { ++ if (!wbrtop) ++ mutex_unlock(&mtx); ++ if (!silent) ++ pr_err("open %s(%ld)\n", fpath, PTR_ERR(file)); ++ return file; ++ } + -+ return err; -+} ++ /* keep file count */ ++ err = 0; ++ d = file->f_path.dentry; ++ h_parent = au_dget_parent_lock(d, AuLsc_I_PARENT); ++ if (!wbrtop) ++ mutex_unlock(&mtx); ++ /* mnt_want_write() is unnecessary here */ ++ h_dir = d_inode(h_parent); ++ inode = file_inode(file); ++ /* no delegation since it is just created */ ++ if (inode->i_nlink) ++ err = vfsub_unlink(h_dir, &file->f_path, /*delegated*/NULL, ++ /*force*/0); ++ inode_unlock(h_dir); ++ dput(h_parent); ++ if (unlikely(err)) { ++ if (!silent) ++ pr_err("unlink %s(%d)\n", fpath, err); ++ goto out; ++ } + -+/* ---------------------------------------------------------------------- */ ++ err = -EINVAL; ++ if (unlikely(sb == d->d_sb)) { ++ if (!silent) ++ pr_err("%s must be outside\n", fpath); ++ goto out; ++ } ++ if (unlikely(au_test_fs_bad_xino(d->d_sb))) { ++ if (!silent) ++ pr_err("xino doesn't support %s(%s)\n", ++ fpath, au_sbtype(d->d_sb)); ++ goto out; ++ } ++ return file; /* success */ ++ ++out: ++ fput(file); ++ file = ERR_PTR(err); ++ return file; ++} + +/* -+ * create a new xinofile at the same place/path as @base_file. ++ * create a new xinofile at the same place/path as @base. + */ -+struct file *au_xino_create2(struct file *base_file, struct file *copy_src) ++struct file *au_xino_create2(struct super_block *sb, struct path *base, ++ struct file *copy_src) +{ + struct file *file; -+ struct dentry *base, *parent; ++ struct dentry *dentry, *parent; + struct inode *dir, *delegated; + struct qstr *name; + struct path path; -+ int err; ++ int err, do_unlock; ++ struct au_xino_lock_dir ldir; + -+ base = base_file->f_path.dentry; -+ parent = base->d_parent; /* dir inode is locked */ ++ do_unlock = 1; ++ au_xino_lock_dir(sb, base, &ldir); ++ dentry = base->dentry; ++ parent = dentry->d_parent; /* dir inode is locked */ + dir = d_inode(parent); + IMustLock(dir); + -+ file = ERR_PTR(-EINVAL); -+ name = &base->d_name; ++ name = &dentry->d_name; + path.dentry = vfsub_lookup_one_len(name->name, parent, name->len); + if (IS_ERR(path.dentry)) { + file = (void *)path.dentry; -+ pr_err("%pd lookup err %ld\n", -+ base, PTR_ERR(path.dentry)); ++ pr_err("%pd lookup err %ld\n", dentry, PTR_ERR(path.dentry)); + goto out; + } + + /* no need to mnt_want_write() since we call dentry_open() later */ -+ err = vfs_create(dir, path.dentry, S_IRUGO | S_IWUGO, NULL); ++ err = vfs_create(dir, path.dentry, 0666, NULL); + if (unlikely(err)) { + file = ERR_PTR(err); -+ pr_err("%pd create err %d\n", base, err); ++ pr_err("%pd create err %d\n", dentry, err); + goto out_dput; + } + -+ path.mnt = base_file->f_path.mnt; ++ path.mnt = base->mnt; + file = vfsub_dentry_open(&path, + O_RDWR | O_CREAT | O_EXCL | O_LARGEFILE + /* | __FMODE_NONOTIFY */); + if (IS_ERR(file)) { -+ pr_err("%pd open err %ld\n", base, PTR_ERR(file)); ++ pr_err("%pd open err %ld\n", dentry, PTR_ERR(file)); + goto out_dput; + } + + delegated = NULL; + err = vfsub_unlink(dir, &file->f_path, &delegated, /*force*/0); ++ au_xino_unlock_dir(&ldir); ++ do_unlock = 0; + if (unlikely(err == -EWOULDBLOCK)) { + pr_warn("cannot retry for NFSv4 delegation" + " for an internal unlink\n"); + iput(delegated); + } + if (unlikely(err)) { -+ pr_err("%pd unlink err %d\n", base, err); ++ pr_err("%pd unlink err %d\n", dentry, err); + goto out_fput; + } + @@ -34859,7 +37206,7 @@ diff -urN /usr/share/empty/fs/aufs/xino.c linux/fs/aufs/xino.c + /* no one can touch copy_src xino */ + err = au_copy_file(file, copy_src, vfsub_f_size_read(copy_src)); + if (unlikely(err)) { -+ pr_err("%pd copy err %d\n", base, err); ++ pr_err("%pd copy err %d\n", dentry, err); + goto out_fput; + } + } @@ -34871,60 +37218,161 @@ diff -urN /usr/share/empty/fs/aufs/xino.c linux/fs/aufs/xino.c +out_dput: + dput(path.dentry); +out: ++ if (do_unlock) ++ au_xino_unlock_dir(&ldir); + return file; +} + -+struct au_xino_lock_dir { -+ struct au_hinode *hdir; -+ struct dentry *parent; -+ struct inode *dir; -+}; ++struct file *au_xino_file1(struct au_xino *xi) ++{ ++ struct file *file; ++ unsigned int u, nfile; + -+static void au_xino_lock_dir(struct super_block *sb, struct file *xino, -+ struct au_xino_lock_dir *ldir) ++ file = NULL; ++ nfile = xi->xi_nfile; ++ for (u = 0; u < nfile; u++) { ++ file = xi->xi_file[u]; ++ if (file) ++ break; ++ } ++ ++ return file; ++} ++ ++static int au_xino_file_set(struct au_xino *xi, int idx, struct file *file) +{ -+ aufs_bindex_t brid, bindex; ++ int err; ++ struct file *f; ++ void *p; + -+ ldir->hdir = NULL; -+ bindex = -1; -+ brid = au_xino_brid(sb); -+ if (brid >= 0) -+ bindex = au_br_index(sb, brid); -+ if (bindex >= 0) { -+ ldir->hdir = au_hi(d_inode(sb->s_root), bindex); -+ au_hn_inode_lock_nested(ldir->hdir, AuLsc_I_PARENT); ++ if (file) ++ get_file(file); ++ ++ err = 0; ++ f = NULL; ++ if (idx < xi->xi_nfile) { ++ f = xi->xi_file[idx]; ++ if (f) ++ fput(f); + } else { -+ ldir->parent = dget_parent(xino->f_path.dentry); -+ ldir->dir = d_inode(ldir->parent); -+ inode_lock_nested(ldir->dir, AuLsc_I_PARENT); ++ p = au_kzrealloc(xi->xi_file, ++ sizeof(*xi->xi_file) * xi->xi_nfile, ++ sizeof(*xi->xi_file) * (idx + 1), ++ GFP_NOFS, /*may_shrink*/0); ++ if (p) { ++ MtxMustLock(&xi->xi_mtx); ++ xi->xi_file = p; ++ xi->xi_nfile = idx + 1; ++ } else { ++ err = -ENOMEM; ++ if (file) ++ fput(file); ++ goto out; ++ } + } ++ xi->xi_file[idx] = file; ++ ++out: ++ return err; +} + -+static void au_xino_unlock_dir(struct au_xino_lock_dir *ldir) ++/* ++ * if @xinew->xi is not set, then create new xigen file. ++ */ ++struct file *au_xi_new(struct super_block *sb, struct au_xi_new *xinew) +{ -+ if (ldir->hdir) -+ au_hn_inode_unlock(ldir->hdir); ++ struct file *file; ++ int err; ++ ++ SiMustAnyLock(sb); ++ ++ file = au_xino_create2(sb, xinew->base, xinew->copy_src); ++ if (IS_ERR(file)) { ++ err = PTR_ERR(file); ++ pr_err("%s[%d], err %d\n", ++ xinew->xi ? "xino" : "xigen", ++ xinew->idx, err); ++ goto out; ++ } ++ ++ if (xinew->xi) ++ err = au_xino_file_set(xinew->xi, xinew->idx, file); + else { -+ inode_unlock(ldir->dir); -+ dput(ldir->parent); ++ BUG(); ++ /* todo: make xigen file an array */ ++ /* err = au_xigen_file_set(sb, xinew->idx, file); */ + } ++ fput(file); ++ if (unlikely(err)) ++ file = ERR_PTR(err); ++ ++out: ++ return file; +} + +/* ---------------------------------------------------------------------- */ + -+/* trucate xino files asynchronously */ -+ -+int au_xino_trunc(struct super_block *sb, aufs_bindex_t bindex) ++/* ++ * truncate xino files ++ */ ++static int au_xino_do_trunc(struct super_block *sb, aufs_bindex_t bindex, ++ int idx, struct kstatfs *st) +{ + int err; -+ unsigned long jiffy; + blkcnt_t blocks; -+ aufs_bindex_t bi, bbot; ++ struct file *file, *new_xino; ++ struct au_xi_new xinew = { ++ .idx = idx ++ }; ++ ++ err = 0; ++ xinew.xi = au_sbr(sb, bindex)->br_xino; ++ file = au_xino_file(xinew.xi, idx); ++ if (!file) ++ goto out; ++ ++ xinew.base = &file->f_path; ++ err = vfs_statfs(xinew.base, st); ++ if (unlikely(err)) { ++ AuErr1("statfs err %d, ignored\n", err); ++ err = 0; ++ goto out; ++ } ++ ++ blocks = file_inode(file)->i_blocks; ++ pr_info("begin truncating xino(b%d-%d), ib%llu, %llu/%llu free blks\n", ++ bindex, idx, (u64)blocks, st->f_bfree, st->f_blocks); ++ ++ xinew.copy_src = file; ++ new_xino = au_xi_new(sb, &xinew); ++ if (IS_ERR(new_xino)) { ++ err = PTR_ERR(new_xino); ++ pr_err("xino(b%d-%d), err %d, ignored\n", bindex, idx, err); ++ goto out; ++ } ++ ++ err = vfs_statfs(&new_xino->f_path, st); ++ if (!err) ++ pr_info("end truncating xino(b%d-%d), ib%llu, %llu/%llu free blks\n", ++ bindex, idx, (u64)file_inode(new_xino)->i_blocks, ++ st->f_bfree, st->f_blocks); ++ else { ++ AuErr1("statfs err %d, ignored\n", err); ++ err = 0; ++ } ++ ++out: ++ return err; ++} ++ ++int au_xino_trunc(struct super_block *sb, aufs_bindex_t bindex, int idx_begin) ++{ ++ int err, i; ++ unsigned long jiffy; ++ aufs_bindex_t bbot; + struct kstatfs *st; + struct au_branch *br; -+ struct file *new_xino, *file; -+ struct super_block *h_sb; -+ struct au_xino_lock_dir ldir; ++ struct au_xino *xi; + + err = -ENOMEM; + st = kmalloc(sizeof(*st), GFP_NOFS); @@ -34935,57 +37383,18 @@ diff -urN /usr/share/empty/fs/aufs/xino.c linux/fs/aufs/xino.c + bbot = au_sbbot(sb); + if (unlikely(bindex < 0 || bbot < bindex)) + goto out_st; -+ br = au_sbr(sb, bindex); -+ file = br->br_xino.xi_file; -+ if (!file) -+ goto out_st; -+ -+ err = vfs_statfs(&file->f_path, st); -+ if (unlikely(err)) -+ AuErr1("statfs err %d, ignored\n", err); -+ jiffy = jiffies; -+ blocks = file_inode(file)->i_blocks; -+ pr_info("begin truncating xino(b%d), ib%llu, %llu/%llu free blks\n", -+ bindex, (u64)blocks, st->f_bfree, st->f_blocks); + -+ au_xino_lock_dir(sb, file, &ldir); -+ /* mnt_want_write() is unnecessary here */ -+ new_xino = au_xino_create2(file, file); -+ au_xino_unlock_dir(&ldir); -+ err = PTR_ERR(new_xino); -+ if (IS_ERR(new_xino)) { -+ pr_err("err %d, ignored\n", err); -+ goto out_st; -+ } + err = 0; -+ fput(file); -+ br->br_xino.xi_file = new_xino; -+ -+ h_sb = au_br_sb(br); -+ for (bi = 0; bi <= bbot; bi++) { -+ if (unlikely(bi == bindex)) -+ continue; -+ br = au_sbr(sb, bi); -+ if (au_br_sb(br) != h_sb) -+ continue; -+ -+ fput(br->br_xino.xi_file); -+ br->br_xino.xi_file = new_xino; -+ get_file(new_xino); -+ } -+ -+ err = vfs_statfs(&new_xino->f_path, st); -+ if (!err) { -+ pr_info("end truncating xino(b%d), ib%llu, %llu/%llu free blks\n", -+ bindex, (u64)file_inode(new_xino)->i_blocks, -+ st->f_bfree, st->f_blocks); -+ if (file_inode(new_xino)->i_blocks < blocks) -+ au_sbi(sb)->si_xino_jiffy = jiffy; -+ } else -+ AuErr1("statfs err %d, ignored\n", err); ++ jiffy = jiffies; ++ br = au_sbr(sb, bindex); ++ xi = br->br_xino; ++ for (i = idx_begin; !err && i < xi->xi_nfile; i++) ++ err = au_xino_do_trunc(sb, bindex, i, st); ++ if (!err) ++ au_sbi(sb)->si_xino_jiffy = jiffy; + +out_st: -+ au_delayed_kfree(st); ++ au_kfree_rcu(st); +out: + return err; +} @@ -34993,6 +37402,7 @@ diff -urN /usr/share/empty/fs/aufs/xino.c linux/fs/aufs/xino.c +struct xino_do_trunc_args { + struct super_block *sb; + struct au_branch *br; ++ int idx; +}; + +static void xino_do_trunc(void *_args) @@ -35001,61 +37411,79 @@ diff -urN /usr/share/empty/fs/aufs/xino.c linux/fs/aufs/xino.c + struct super_block *sb; + struct au_branch *br; + struct inode *dir; -+ int err; ++ int err, idx; + aufs_bindex_t bindex; + + err = 0; + sb = args->sb; + dir = d_inode(sb->s_root); + br = args->br; ++ idx = args->idx; + + si_noflush_write_lock(sb); + ii_read_lock_parent(dir); + bindex = au_br_index(sb, br->br_id); -+ err = au_xino_trunc(sb, bindex); ++ err = au_xino_trunc(sb, bindex, idx); + ii_read_unlock(dir); + if (unlikely(err)) + pr_warn("err b%d, (%d)\n", bindex, err); -+ atomic_dec(&br->br_xino_running); -+ au_br_put(br); ++ atomic_dec(&br->br_xino->xi_truncating); ++ au_lcnt_dec(&br->br_count); + si_write_unlock(sb); + au_nwt_done(&au_sbi(sb)->si_nowait); -+ au_delayed_kfree(args); ++ au_kfree_rcu(args); +} + ++/* ++ * returns the index in the xi_file array whose corresponding file is necessary ++ * to truncate, or -1 which means no need to truncate. ++ */ +static int xino_trunc_test(struct super_block *sb, struct au_branch *br) +{ + int err; ++ unsigned int u; + struct kstatfs st; + struct au_sbinfo *sbinfo; ++ struct au_xino *xi; ++ struct file *file; + + /* todo: si_xino_expire and the ratio should be customizable */ + sbinfo = au_sbi(sb); + if (time_before(jiffies, + sbinfo->si_xino_jiffy + sbinfo->si_xino_expire)) -+ return 0; ++ return -1; + + /* truncation border */ -+ err = vfs_statfs(&br->br_xino.xi_file->f_path, &st); -+ if (unlikely(err)) { -+ AuErr1("statfs err %d, ignored\n", err); -+ return 0; ++ xi = br->br_xino; ++ for (u = 0; u < xi->xi_nfile; u++) { ++ file = au_xino_file(xi, u); ++ if (!file) ++ continue; ++ ++ err = vfs_statfs(&file->f_path, &st); ++ if (unlikely(err)) { ++ AuErr1("statfs err %d, ignored\n", err); ++ return -1; ++ } ++ if (div64_u64(st.f_bfree * 100, st.f_blocks) ++ >= AUFS_XINO_DEF_TRUNC) ++ return u; + } -+ if (div64_u64(st.f_bfree * 100, st.f_blocks) >= AUFS_XINO_DEF_TRUNC) -+ return 0; + -+ return 1; ++ return -1; +} + +static void xino_try_trunc(struct super_block *sb, struct au_branch *br) +{ ++ int idx; + struct xino_do_trunc_args *args; + int wkq_err; + -+ if (!xino_trunc_test(sb, br)) ++ idx = xino_trunc_test(sb, br); ++ if (idx < 0) + return; + -+ if (atomic_inc_return(&br->br_xino_running) > 1) ++ if (atomic_inc_return(&br->br_xino->xi_truncating) > 1) + goto out; + + /* lock and kfree() will be called in trunc_xino() */ @@ -35065,36 +37493,236 @@ diff -urN /usr/share/empty/fs/aufs/xino.c linux/fs/aufs/xino.c + goto out; + } + -+ au_br_get(br); -+ args->sb = sb; -+ args->br = br; -+ wkq_err = au_wkq_nowait(xino_do_trunc, args, sb, /*flags*/0); -+ if (!wkq_err) -+ return; /* success */ ++ au_lcnt_inc(&br->br_count); ++ args->sb = sb; ++ args->br = br; ++ args->idx = idx; ++ wkq_err = au_wkq_nowait(xino_do_trunc, args, sb, /*flags*/0); ++ if (!wkq_err) ++ return; /* success */ ++ ++ pr_err("wkq %d\n", wkq_err); ++ au_lcnt_dec(&br->br_count); ++ au_kfree_rcu(args); ++ ++out: ++ atomic_dec(&br->br_xino->xi_truncating); ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++struct au_xi_calc { ++ int idx; ++ loff_t pos; ++}; ++ ++static void au_xi_calc(struct super_block *sb, ino_t h_ino, ++ struct au_xi_calc *calc) ++{ ++ loff_t maxent; ++ ++ maxent = au_xi_maxent(sb); ++ calc->idx = div64_u64_rem(h_ino, maxent, &calc->pos); ++ calc->pos *= sizeof(ino_t); ++} ++ ++static int au_xino_do_new_async(struct super_block *sb, struct au_branch *br, ++ struct au_xi_calc *calc) ++{ ++ int err; ++ struct file *file; ++ struct au_xino *xi = br->br_xino; ++ struct au_xi_new xinew = { ++ .xi = xi ++ }; ++ ++ SiMustAnyLock(sb); ++ ++ err = 0; ++ if (!xi) ++ goto out; ++ ++ mutex_lock(&xi->xi_mtx); ++ file = au_xino_file(xi, calc->idx); ++ if (file) ++ goto out_mtx; ++ ++ file = au_xino_file(xi, /*idx*/-1); ++ AuDebugOn(!file); ++ xinew.idx = calc->idx; ++ xinew.base = &file->f_path; ++ /* xinew.copy_src = NULL; */ ++ file = au_xi_new(sb, &xinew); ++ if (IS_ERR(file)) ++ err = PTR_ERR(file); ++ ++out_mtx: ++ mutex_unlock(&xi->xi_mtx); ++out: ++ return err; ++} ++ ++struct au_xino_do_new_async_args { ++ struct super_block *sb; ++ struct au_branch *br; ++ struct au_xi_calc calc; ++ ino_t ino; ++}; ++ ++struct au_xi_writing { ++ struct hlist_bl_node node; ++ ino_t h_ino, ino; ++}; ++ ++static int au_xino_do_write(vfs_writef_t write, struct file *file, ++ struct au_xi_calc *calc, ino_t ino); ++ ++static void au_xino_call_do_new_async(void *args) ++{ ++ struct au_xino_do_new_async_args *a = args; ++ struct au_branch *br; ++ struct super_block *sb; ++ struct au_sbinfo *sbi; ++ struct inode *root; ++ struct file *file; ++ struct au_xi_writing *del, *p; ++ struct hlist_bl_head *hbl; ++ struct hlist_bl_node *pos; ++ int err; ++ ++ br = a->br; ++ sb = a->sb; ++ sbi = au_sbi(sb); ++ si_noflush_read_lock(sb); ++ root = d_inode(sb->s_root); ++ ii_read_lock_child(root); ++ err = au_xino_do_new_async(sb, br, &a->calc); ++ if (unlikely(err)) { ++ AuIOErr("err %d\n", err); ++ goto out; ++ } ++ ++ file = au_xino_file(br->br_xino, a->calc.idx); ++ AuDebugOn(!file); ++ err = au_xino_do_write(sbi->si_xwrite, file, &a->calc, a->ino); ++ if (unlikely(err)) { ++ AuIOErr("err %d\n", err); ++ goto out; ++ } ++ ++ del = NULL; ++ hbl = &br->br_xino->xi_writing; ++ hlist_bl_lock(hbl); ++ au_hbl_for_each(pos, hbl) { ++ p = container_of(pos, struct au_xi_writing, node); ++ if (p->ino == a->ino) { ++ del = p; ++ hlist_bl_del(&p->node); ++ break; ++ } ++ } ++ hlist_bl_unlock(hbl); ++ au_kfree_rcu(del); ++ ++out: ++ au_lcnt_dec(&br->br_count); ++ ii_read_unlock(root); ++ si_read_unlock(sb); ++ au_nwt_done(&sbi->si_nowait); ++ au_kfree_rcu(a); ++} ++ ++/* ++ * create a new xino file asynchronously ++ */ ++static int au_xino_new_async(struct super_block *sb, struct au_branch *br, ++ struct au_xi_calc *calc, ino_t ino) ++{ ++ int err; ++ struct au_xino_do_new_async_args *arg; ++ ++ err = -ENOMEM; ++ arg = kmalloc(sizeof(*arg), GFP_NOFS); ++ if (unlikely(!arg)) ++ goto out; ++ ++ arg->sb = sb; ++ arg->br = br; ++ arg->calc = *calc; ++ arg->ino = ino; ++ au_lcnt_inc(&br->br_count); ++ err = au_wkq_nowait(au_xino_call_do_new_async, arg, sb, AuWkq_NEST); ++ if (unlikely(err)) { ++ pr_err("wkq %d\n", err); ++ au_lcnt_dec(&br->br_count); ++ au_kfree_rcu(arg); ++ } ++ ++out: ++ return err; ++} ++ ++/* ++ * read @ino from xinofile for the specified branch{@sb, @bindex} ++ * at the position of @h_ino. ++ */ ++int au_xino_read(struct super_block *sb, aufs_bindex_t bindex, ino_t h_ino, ++ ino_t *ino) ++{ ++ int err; ++ ssize_t sz; ++ struct au_xi_calc calc; ++ struct au_sbinfo *sbinfo; ++ struct file *file; ++ struct au_xino *xi; ++ struct hlist_bl_head *hbl; ++ struct hlist_bl_node *pos; ++ struct au_xi_writing *p; ++ ++ *ino = 0; ++ if (!au_opt_test(au_mntflags(sb), XINO)) ++ return 0; /* no xino */ ++ ++ err = 0; ++ au_xi_calc(sb, h_ino, &calc); ++ xi = au_sbr(sb, bindex)->br_xino; ++ file = au_xino_file(xi, calc.idx); ++ if (!file) { ++ hbl = &xi->xi_writing; ++ hlist_bl_lock(hbl); ++ au_hbl_for_each(pos, hbl) { ++ p = container_of(pos, struct au_xi_writing, node); ++ if (p->h_ino == h_ino) { ++ AuDbg("hi%llu, i%llu, found\n", ++ (u64)p->h_ino, (u64)p->ino); ++ *ino = p->ino; ++ break; ++ } ++ } ++ hlist_bl_unlock(hbl); ++ return 0; ++ } else if (vfsub_f_size_read(file) < calc.pos + sizeof(*ino)) ++ return 0; /* no xino */ + -+ pr_err("wkq %d\n", wkq_err); -+ au_br_put(br); -+ au_delayed_kfree(args); ++ sbinfo = au_sbi(sb); ++ sz = xino_fread(sbinfo->si_xread, file, ino, sizeof(*ino), &calc.pos); ++ if (sz == sizeof(*ino)) ++ return 0; /* success */ + -+out: -+ atomic_dec(&br->br_xino_running); ++ err = sz; ++ if (unlikely(sz >= 0)) { ++ err = -EIO; ++ AuIOErr("xino read error (%zd)\n", sz); ++ } ++ return err; +} + -+/* ---------------------------------------------------------------------- */ -+ +static int au_xino_do_write(vfs_writef_t write, struct file *file, -+ ino_t h_ino, ino_t ino) ++ struct au_xi_calc *calc, ino_t ino) +{ -+ loff_t pos; + ssize_t sz; + -+ pos = h_ino; -+ if (unlikely(au_loff_max / sizeof(ino) - 1 < pos)) { -+ AuIOErr1("too large hi%lu\n", (unsigned long)h_ino); -+ return -EFBIG; -+ } -+ pos *= sizeof(ino); -+ sz = xino_fwrite(write, file, &ino, sizeof(ino), &pos); ++ sz = xino_fwrite(write, file, &ino, sizeof(ino), &calc->pos); + if (sz == sizeof(ino)) + return 0; /* success */ + @@ -35114,34 +37742,234 @@ diff -urN /usr/share/empty/fs/aufs/xino.c linux/fs/aufs/xino.c +{ + int err; + unsigned int mnt_flags; ++ struct au_xi_calc calc; ++ struct file *file; + struct au_branch *br; ++ struct au_xino *xi; ++ struct au_xi_writing *p; + -+ BUILD_BUG_ON(sizeof(long long) != sizeof(au_loff_max) -+ || ((loff_t)-1) > 0); + SiMustAnyLock(sb); + + mnt_flags = au_mntflags(sb); + if (!au_opt_test(mnt_flags, XINO)) + return 0; + ++ au_xi_calc(sb, h_ino, &calc); + br = au_sbr(sb, bindex); -+ err = au_xino_do_write(au_sbi(sb)->si_xwrite, br->br_xino.xi_file, -+ h_ino, ino); ++ xi = br->br_xino; ++ file = au_xino_file(xi, calc.idx); ++ if (!file) { ++ /* store the inum pair into the list */ ++ p = kmalloc(sizeof(*p), GFP_NOFS | __GFP_NOFAIL); ++ p->h_ino = h_ino; ++ p->ino = ino; ++ au_hbl_add(&p->node, &xi->xi_writing); ++ ++ /* create and write a new xino file asynchronously */ ++ err = au_xino_new_async(sb, br, &calc, ino); ++ if (!err) ++ return 0; /* success */ ++ goto out; ++ } ++ ++ err = au_xino_do_write(au_sbi(sb)->si_xwrite, file, &calc, ino); + if (!err) { ++ br = au_sbr(sb, bindex); + if (au_opt_test(mnt_flags, TRUNC_XINO) + && au_test_fs_trunc_xino(au_br_sb(br))) + xino_try_trunc(sb, br); + return 0; /* success */ + } + ++out: + AuIOErr("write failed (%d)\n", err); + return -EIO; +} + -+/* ---------------------------------------------------------------------- */ ++static ssize_t xino_fread_wkq(vfs_readf_t func, struct file *file, void *buf, ++ size_t size, loff_t *pos); ++ ++/* todo: unnecessary to support mmap_sem since kernel-space? */ ++ssize_t xino_fread(vfs_readf_t func, struct file *file, void *kbuf, size_t size, ++ loff_t *pos) ++{ ++ ssize_t err; ++ mm_segment_t oldfs; ++ union { ++ void *k; ++ char __user *u; ++ } buf; ++ int i; ++ const int prevent_endless = 10; ++ ++ i = 0; ++ buf.k = kbuf; ++ oldfs = get_fs(); ++ set_fs(KERNEL_DS); ++ do { ++ err = func(file, buf.u, size, pos); ++ if (err == -EINTR ++ && !au_wkq_test() ++ && fatal_signal_pending(current)) { ++ set_fs(oldfs); ++ err = xino_fread_wkq(func, file, kbuf, size, pos); ++ BUG_ON(err == -EINTR); ++ oldfs = get_fs(); ++ set_fs(KERNEL_DS); ++ } ++ } while (i++ < prevent_endless ++ && (err == -EAGAIN || err == -EINTR)); ++ set_fs(oldfs); ++ ++#if 0 /* reserved for future use */ ++ if (err > 0) ++ fsnotify_access(file->f_path.dentry); ++#endif ++ ++ return err; ++} ++ ++struct xino_fread_args { ++ ssize_t *errp; ++ vfs_readf_t func; ++ struct file *file; ++ void *buf; ++ size_t size; ++ loff_t *pos; ++}; ++ ++static void call_xino_fread(void *args) ++{ ++ struct xino_fread_args *a = args; ++ *a->errp = xino_fread(a->func, a->file, a->buf, a->size, a->pos); ++} ++ ++static ssize_t xino_fread_wkq(vfs_readf_t func, struct file *file, void *buf, ++ size_t size, loff_t *pos) ++{ ++ ssize_t err; ++ int wkq_err; ++ struct xino_fread_args args = { ++ .errp = &err, ++ .func = func, ++ .file = file, ++ .buf = buf, ++ .size = size, ++ .pos = pos ++ }; ++ ++ wkq_err = au_wkq_wait(call_xino_fread, &args); ++ if (unlikely(wkq_err)) ++ err = wkq_err; ++ ++ return err; ++} ++ ++static ssize_t xino_fwrite_wkq(vfs_writef_t func, struct file *file, void *buf, ++ size_t size, loff_t *pos); ++ ++static ssize_t do_xino_fwrite(vfs_writef_t func, struct file *file, void *kbuf, ++ size_t size, loff_t *pos) ++{ ++ ssize_t err; ++ mm_segment_t oldfs; ++ union { ++ void *k; ++ const char __user *u; ++ } buf; ++ int i; ++ const int prevent_endless = 10; ++ ++ i = 0; ++ buf.k = kbuf; ++ oldfs = get_fs(); ++ set_fs(KERNEL_DS); ++ do { ++ err = func(file, buf.u, size, pos); ++ if (err == -EINTR ++ && !au_wkq_test() ++ && fatal_signal_pending(current)) { ++ set_fs(oldfs); ++ err = xino_fwrite_wkq(func, file, kbuf, size, pos); ++ BUG_ON(err == -EINTR); ++ oldfs = get_fs(); ++ set_fs(KERNEL_DS); ++ } ++ } while (i++ < prevent_endless ++ && (err == -EAGAIN || err == -EINTR)); ++ set_fs(oldfs); ++ ++#if 0 /* reserved for future use */ ++ if (err > 0) ++ fsnotify_modify(file->f_path.dentry); ++#endif ++ ++ return err; ++} ++ ++struct do_xino_fwrite_args { ++ ssize_t *errp; ++ vfs_writef_t func; ++ struct file *file; ++ void *buf; ++ size_t size; ++ loff_t *pos; ++}; ++ ++static void call_do_xino_fwrite(void *args) ++{ ++ struct do_xino_fwrite_args *a = args; ++ *a->errp = do_xino_fwrite(a->func, a->file, a->buf, a->size, a->pos); ++} ++ ++static ssize_t xino_fwrite_wkq(vfs_writef_t func, struct file *file, void *buf, ++ size_t size, loff_t *pos) ++{ ++ ssize_t err; ++ int wkq_err; ++ struct do_xino_fwrite_args args = { ++ .errp = &err, ++ .func = func, ++ .file = file, ++ .buf = buf, ++ .size = size, ++ .pos = pos ++ }; ++ ++ /* ++ * it breaks RLIMIT_FSIZE and normal user's limit, ++ * users should care about quota and real 'filesystem full.' ++ */ ++ wkq_err = au_wkq_wait(call_do_xino_fwrite, &args); ++ if (unlikely(wkq_err)) ++ err = wkq_err; ++ ++ return err; ++} ++ ++ssize_t xino_fwrite(vfs_writef_t func, struct file *file, void *buf, ++ size_t size, loff_t *pos) ++{ ++ ssize_t err; + -+/* aufs inode number bitmap */ ++ if (rlimit(RLIMIT_FSIZE) == RLIM_INFINITY) { ++ lockdep_off(); ++ err = do_xino_fwrite(func, file, buf, size, pos); ++ lockdep_on(); ++ } else { ++ lockdep_off(); ++ err = xino_fwrite_wkq(func, file, buf, size, pos); ++ lockdep_on(); ++ } ++ ++ return err; ++} + ++/* ---------------------------------------------------------------------- */ ++ ++/* ++ * inode number bitmap ++ */ +static const int page_bits = (int)PAGE_SIZE * BITS_PER_BYTE; +static ino_t xib_calc_ino(unsigned long pindex, int bit) +{ @@ -35206,8 +38034,6 @@ diff -urN /usr/share/empty/fs/aufs/xino.c linux/fs/aufs/xino.c + return err; +} + -+/* ---------------------------------------------------------------------- */ -+ +static void au_xib_clear_bit(struct inode *inode) +{ + int err, bit; @@ -35230,310 +38056,11 @@ diff -urN /usr/share/empty/fs/aufs/xino.c linux/fs/aufs/xino.c + mutex_unlock(&sbinfo->si_xib_mtx); +} + -+/* for s_op->delete_inode() */ -+void au_xino_delete_inode(struct inode *inode, const int unlinked) -+{ -+ int err; -+ unsigned int mnt_flags; -+ aufs_bindex_t bindex, bbot, bi; -+ unsigned char try_trunc; -+ struct au_iinfo *iinfo; -+ struct super_block *sb; -+ struct au_hinode *hi; -+ struct inode *h_inode; -+ struct au_branch *br; -+ vfs_writef_t xwrite; -+ -+ AuDebugOn(au_is_bad_inode(inode)); -+ -+ sb = inode->i_sb; -+ mnt_flags = au_mntflags(sb); -+ if (!au_opt_test(mnt_flags, XINO) -+ || inode->i_ino == AUFS_ROOT_INO) -+ return; -+ -+ if (unlinked) { -+ au_xigen_inc(inode); -+ au_xib_clear_bit(inode); -+ } -+ -+ iinfo = au_ii(inode); -+ bindex = iinfo->ii_btop; -+ if (bindex < 0) -+ return; -+ -+ xwrite = au_sbi(sb)->si_xwrite; -+ try_trunc = !!au_opt_test(mnt_flags, TRUNC_XINO); -+ hi = au_hinode(iinfo, bindex); -+ bbot = iinfo->ii_bbot; -+ for (; bindex <= bbot; bindex++, hi++) { -+ h_inode = hi->hi_inode; -+ if (!h_inode -+ || (!unlinked && h_inode->i_nlink)) -+ continue; -+ -+ /* inode may not be revalidated */ -+ bi = au_br_index(sb, hi->hi_id); -+ if (bi < 0) -+ continue; -+ -+ br = au_sbr(sb, bi); -+ err = au_xino_do_write(xwrite, br->br_xino.xi_file, -+ h_inode->i_ino, /*ino*/0); -+ if (!err && try_trunc -+ && au_test_fs_trunc_xino(au_br_sb(br))) -+ xino_try_trunc(sb, br); -+ } -+} -+ -+/* get an unused inode number from bitmap */ -+ino_t au_xino_new_ino(struct super_block *sb) -+{ -+ ino_t ino; -+ unsigned long *p, pindex, ul, pend; -+ struct au_sbinfo *sbinfo; -+ struct file *file; -+ int free_bit, err; -+ -+ if (!au_opt_test(au_mntflags(sb), XINO)) -+ return iunique(sb, AUFS_FIRST_INO); -+ -+ sbinfo = au_sbi(sb); -+ mutex_lock(&sbinfo->si_xib_mtx); -+ p = sbinfo->si_xib_buf; -+ free_bit = sbinfo->si_xib_next_bit; -+ if (free_bit < page_bits && !test_bit(free_bit, p)) -+ goto out; /* success */ -+ free_bit = find_first_zero_bit(p, page_bits); -+ if (free_bit < page_bits) -+ goto out; /* success */ -+ -+ pindex = sbinfo->si_xib_last_pindex; -+ for (ul = pindex - 1; ul < ULONG_MAX; ul--) { -+ err = xib_pindex(sb, ul); -+ if (unlikely(err)) -+ goto out_err; -+ free_bit = find_first_zero_bit(p, page_bits); -+ if (free_bit < page_bits) -+ goto out; /* success */ -+ } -+ -+ file = sbinfo->si_xib; -+ pend = vfsub_f_size_read(file) / PAGE_SIZE; -+ for (ul = pindex + 1; ul <= pend; ul++) { -+ err = xib_pindex(sb, ul); -+ if (unlikely(err)) -+ goto out_err; -+ free_bit = find_first_zero_bit(p, page_bits); -+ if (free_bit < page_bits) -+ goto out; /* success */ -+ } -+ BUG(); -+ -+out: -+ set_bit(free_bit, p); -+ sbinfo->si_xib_next_bit = free_bit + 1; -+ pindex = sbinfo->si_xib_last_pindex; -+ mutex_unlock(&sbinfo->si_xib_mtx); -+ ino = xib_calc_ino(pindex, free_bit); -+ AuDbg("i%lu\n", (unsigned long)ino); -+ return ino; -+out_err: -+ mutex_unlock(&sbinfo->si_xib_mtx); -+ AuDbg("i0\n"); -+ return 0; -+} -+ -+/* -+ * read @ino from xinofile for the specified branch{@sb, @bindex} -+ * at the position of @h_ino. -+ * if @ino does not exist and @do_new is true, get new one. -+ */ -+int au_xino_read(struct super_block *sb, aufs_bindex_t bindex, ino_t h_ino, -+ ino_t *ino) -+{ -+ int err; -+ ssize_t sz; -+ loff_t pos; -+ struct file *file; -+ struct au_sbinfo *sbinfo; -+ -+ *ino = 0; -+ if (!au_opt_test(au_mntflags(sb), XINO)) -+ return 0; /* no xino */ -+ -+ err = 0; -+ sbinfo = au_sbi(sb); -+ pos = h_ino; -+ if (unlikely(au_loff_max / sizeof(*ino) - 1 < pos)) { -+ AuIOErr1("too large hi%lu\n", (unsigned long)h_ino); -+ return -EFBIG; -+ } -+ pos *= sizeof(*ino); -+ -+ file = au_sbr(sb, bindex)->br_xino.xi_file; -+ if (vfsub_f_size_read(file) < pos + sizeof(*ino)) -+ return 0; /* no ino */ -+ -+ sz = xino_fread(sbinfo->si_xread, file, ino, sizeof(*ino), &pos); -+ if (sz == sizeof(*ino)) -+ return 0; /* success */ -+ -+ err = sz; -+ if (unlikely(sz >= 0)) { -+ err = -EIO; -+ AuIOErr("xino read error (%zd)\n", sz); -+ } -+ -+ return err; -+} -+ -+/* ---------------------------------------------------------------------- */ -+ -+/* create and set a new xino file */ -+ -+struct file *au_xino_create(struct super_block *sb, char *fname, int silent) -+{ -+ struct file *file; -+ struct dentry *h_parent, *d; -+ struct inode *h_dir, *inode; -+ int err; -+ -+ /* -+ * at mount-time, and the xino file is the default path, -+ * hnotify is disabled so we have no notify events to ignore. -+ * when a user specified the xino, we cannot get au_hdir to be ignored. -+ */ -+ file = vfsub_filp_open(fname, O_RDWR | O_CREAT | O_EXCL | O_LARGEFILE -+ /* | __FMODE_NONOTIFY */, -+ S_IRUGO | S_IWUGO); -+ if (IS_ERR(file)) { -+ if (!silent) -+ pr_err("open %s(%ld)\n", fname, PTR_ERR(file)); -+ return file; -+ } -+ -+ /* keep file count */ -+ err = 0; -+ inode = file_inode(file); -+ h_parent = dget_parent(file->f_path.dentry); -+ h_dir = d_inode(h_parent); -+ inode_lock_nested(h_dir, AuLsc_I_PARENT); -+ /* mnt_want_write() is unnecessary here */ -+ /* no delegation since it is just created */ -+ if (inode->i_nlink) -+ err = vfsub_unlink(h_dir, &file->f_path, /*delegated*/NULL, -+ /*force*/0); -+ inode_unlock(h_dir); -+ dput(h_parent); -+ if (unlikely(err)) { -+ if (!silent) -+ pr_err("unlink %s(%d)\n", fname, err); -+ goto out; -+ } -+ -+ err = -EINVAL; -+ d = file->f_path.dentry; -+ if (unlikely(sb == d->d_sb)) { -+ if (!silent) -+ pr_err("%s must be outside\n", fname); -+ goto out; -+ } -+ if (unlikely(au_test_fs_bad_xino(d->d_sb))) { -+ if (!silent) -+ pr_err("xino doesn't support %s(%s)\n", -+ fname, au_sbtype(d->d_sb)); -+ goto out; -+ } -+ return file; /* success */ -+ -+out: -+ fput(file); -+ file = ERR_PTR(err); -+ return file; -+} -+ -+/* -+ * find another branch who is on the same filesystem of the specified -+ * branch{@btgt}. search until @bbot. -+ */ -+static int is_sb_shared(struct super_block *sb, aufs_bindex_t btgt, -+ aufs_bindex_t bbot) -+{ -+ aufs_bindex_t bindex; -+ struct super_block *tgt_sb = au_sbr_sb(sb, btgt); -+ -+ for (bindex = 0; bindex < btgt; bindex++) -+ if (unlikely(tgt_sb == au_sbr_sb(sb, bindex))) -+ return bindex; -+ for (bindex++; bindex <= bbot; bindex++) -+ if (unlikely(tgt_sb == au_sbr_sb(sb, bindex))) -+ return bindex; -+ return -1; -+} -+ +/* ---------------------------------------------------------------------- */ + +/* -+ * initialize the xinofile for the specified branch @br -+ * at the place/path where @base_file indicates. -+ * test whether another branch is on the same filesystem or not, -+ * if @do_test is true. ++ * truncate a xino bitmap file + */ -+int au_xino_br(struct super_block *sb, struct au_branch *br, ino_t h_ino, -+ struct file *base_file, int do_test) -+{ -+ int err; -+ ino_t ino; -+ aufs_bindex_t bbot, bindex; -+ struct au_branch *shared_br, *b; -+ struct file *file; -+ struct super_block *tgt_sb; -+ -+ shared_br = NULL; -+ bbot = au_sbbot(sb); -+ if (do_test) { -+ tgt_sb = au_br_sb(br); -+ for (bindex = 0; bindex <= bbot; bindex++) { -+ b = au_sbr(sb, bindex); -+ if (tgt_sb == au_br_sb(b)) { -+ shared_br = b; -+ break; -+ } -+ } -+ } -+ -+ if (!shared_br || !shared_br->br_xino.xi_file) { -+ struct au_xino_lock_dir ldir; -+ -+ au_xino_lock_dir(sb, base_file, &ldir); -+ /* mnt_want_write() is unnecessary here */ -+ file = au_xino_create2(base_file, NULL); -+ au_xino_unlock_dir(&ldir); -+ err = PTR_ERR(file); -+ if (IS_ERR(file)) -+ goto out; -+ br->br_xino.xi_file = file; -+ } else { -+ br->br_xino.xi_file = shared_br->br_xino.xi_file; -+ get_file(br->br_xino.xi_file); -+ } -+ -+ ino = AUFS_ROOT_INO; -+ err = au_xino_do_write(au_sbi(sb)->si_xwrite, br->br_xino.xi_file, -+ h_ino, ino); -+ if (unlikely(err)) { -+ fput(br->br_xino.xi_file); -+ br->br_xino.xi_file = NULL; -+ } -+ -+out: -+ return err; -+} -+ -+/* ---------------------------------------------------------------------- */ -+ -+/* trucate a xino bitmap file */ + +/* todo: slow */ +static int do_xib_restore(struct super_block *sb, struct file *file, void *page) @@ -35581,9 +38108,13 @@ diff -urN /usr/share/empty/fs/aufs/xino.c linux/fs/aufs/xino.c + +static int xib_restore(struct super_block *sb) +{ -+ int err; ++ int err, i; ++ unsigned int nfile; + aufs_bindex_t bindex, bbot; + void *page; ++ struct au_branch *br; ++ struct au_xino *xi; ++ struct file *file; + + err = -ENOMEM; + page = (void *)__get_free_page(GFP_NOFS); @@ -35593,12 +38124,18 @@ diff -urN /usr/share/empty/fs/aufs/xino.c linux/fs/aufs/xino.c + err = 0; + bbot = au_sbbot(sb); + for (bindex = 0; !err && bindex <= bbot; bindex++) -+ if (!bindex || is_sb_shared(sb, bindex, bindex - 1) < 0) -+ err = do_xib_restore -+ (sb, au_sbr(sb, bindex)->br_xino.xi_file, page); -+ else -+ AuDbg("b%d\n", bindex); -+ au_delayed_free_page((unsigned long)page); ++ if (!bindex || is_sb_shared(sb, bindex, bindex - 1) < 0) { ++ br = au_sbr(sb, bindex); ++ xi = br->br_xino; ++ nfile = xi->xi_nfile; ++ for (i = 0; i < nfile; i++) { ++ file = au_xino_file(xi, i); ++ if (file) ++ err = do_xib_restore(sb, file, page); ++ } ++ } else ++ AuDbg("skip shared b%d\n", bindex); ++ free_page((unsigned long)page); + +out: + return err; @@ -35609,7 +38146,6 @@ diff -urN /usr/share/empty/fs/aufs/xino.c linux/fs/aufs/xino.c + int err; + ssize_t sz; + loff_t pos; -+ struct au_xino_lock_dir ldir; + struct au_sbinfo *sbinfo; + unsigned long *p; + struct file *file; @@ -35625,10 +38161,7 @@ diff -urN /usr/share/empty/fs/aufs/xino.c linux/fs/aufs/xino.c + if (vfsub_f_size_read(file) <= PAGE_SIZE) + goto out; + -+ au_xino_lock_dir(sb, file, &ldir); -+ /* mnt_want_write() is unnecessary here */ -+ file = au_xino_create2(sbinfo->si_xib, NULL); -+ au_xino_unlock_dir(&ldir); ++ file = au_xino_create2(sb, &sbinfo->si_xib->f_path, NULL); + err = PTR_ERR(file); + if (IS_ERR(file)) + goto out; @@ -35658,6 +38191,112 @@ diff -urN /usr/share/empty/fs/aufs/xino.c linux/fs/aufs/xino.c + +/* ---------------------------------------------------------------------- */ + ++struct au_xino *au_xino_alloc(unsigned int nfile) ++{ ++ struct au_xino *xi; ++ ++ xi = kzalloc(sizeof(*xi), GFP_NOFS); ++ if (unlikely(!xi)) ++ goto out; ++ xi->xi_nfile = nfile; ++ xi->xi_file = kcalloc(nfile, sizeof(*xi->xi_file), GFP_NOFS); ++ if (unlikely(!xi->xi_file)) ++ goto out_free; ++ ++ xi->xi_nondir.total = 8; /* initial size */ ++ xi->xi_nondir.array = kcalloc(xi->xi_nondir.total, sizeof(ino_t), ++ GFP_NOFS); ++ if (unlikely(!xi->xi_nondir.array)) ++ goto out_file; ++ ++ spin_lock_init(&xi->xi_nondir.spin); ++ init_waitqueue_head(&xi->xi_nondir.wqh); ++ mutex_init(&xi->xi_mtx); ++ INIT_HLIST_BL_HEAD(&xi->xi_writing); ++ atomic_set(&xi->xi_truncating, 0); ++ kref_init(&xi->xi_kref); ++ goto out; /* success */ ++ ++out_file: ++ au_kfree_try_rcu(xi->xi_file); ++out_free: ++ au_kfree_rcu(xi); ++ xi = NULL; ++out: ++ return xi; ++} ++ ++static int au_xino_init(struct au_branch *br, int idx, struct file *file) ++{ ++ int err; ++ struct au_xino *xi; ++ ++ err = 0; ++ xi = au_xino_alloc(idx + 1); ++ if (unlikely(!xi)) { ++ err = -ENOMEM; ++ goto out; ++ } ++ ++ if (file) ++ get_file(file); ++ xi->xi_file[idx] = file; ++ AuDebugOn(br->br_xino); ++ br->br_xino = xi; ++ ++out: ++ return err; ++} ++ ++static void au_xino_release(struct kref *kref) ++{ ++ struct au_xino *xi; ++ int i; ++ unsigned long ul; ++ struct hlist_bl_head *hbl; ++ struct hlist_bl_node *pos, *n; ++ struct au_xi_writing *p; ++ ++ xi = container_of(kref, struct au_xino, xi_kref); ++ for (i = 0; i < xi->xi_nfile; i++) ++ if (xi->xi_file[i]) ++ fput(xi->xi_file[i]); ++ for (i = xi->xi_nondir.total - 1; i >= 0; i--) ++ AuDebugOn(xi->xi_nondir.array[i]); ++ mutex_destroy(&xi->xi_mtx); ++ hbl = &xi->xi_writing; ++ ul = au_hbl_count(hbl); ++ if (unlikely(ul)) { ++ pr_warn("xi_writing %lu\n", ul); ++ hlist_bl_lock(hbl); ++ hlist_bl_for_each_entry_safe (p, pos, n, hbl, node) { ++ hlist_bl_del(&p->node); ++ au_kfree_rcu(p); ++ } ++ hlist_bl_unlock(hbl); ++ } ++ au_kfree_try_rcu(xi->xi_file); ++ au_kfree_try_rcu(xi->xi_nondir.array); ++ au_kfree_rcu(xi); ++} ++ ++int au_xino_put(struct au_branch *br) ++{ ++ int ret; ++ struct au_xino *xi; ++ ++ ret = 0; ++ xi = br->br_xino; ++ if (xi) { ++ br->br_xino = NULL; ++ ret = kref_put(&xi->xi_kref, au_xino_release); ++ } ++ ++ return ret; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ +/* + * xino mount option handlers + */ @@ -35670,27 +38309,27 @@ diff -urN /usr/share/empty/fs/aufs/xino.c linux/fs/aufs/xino.c + SiMustWriteLock(sb); + + sbinfo = au_sbi(sb); -+ sbinfo->si_xread = NULL; -+ sbinfo->si_xwrite = NULL; ++ /* unnecessary to clear sbinfo->si_xread and ->si_xwrite */ + if (sbinfo->si_xib) + fput(sbinfo->si_xib); + sbinfo->si_xib = NULL; + if (sbinfo->si_xib_buf) -+ au_delayed_free_page((unsigned long)sbinfo->si_xib_buf); ++ free_page((unsigned long)sbinfo->si_xib_buf); + sbinfo->si_xib_buf = NULL; +} + -+static int au_xino_set_xib(struct super_block *sb, struct file *base) ++static int au_xino_set_xib(struct super_block *sb, struct path *path) +{ + int err; + loff_t pos; + struct au_sbinfo *sbinfo; + struct file *file; ++ struct super_block *xi_sb; + + SiMustWriteLock(sb); + + sbinfo = au_sbi(sb); -+ file = au_xino_create2(base, sbinfo->si_xib); ++ file = au_xino_create2(sb, path, sbinfo->si_xib); + err = PTR_ERR(file); + if (IS_ERR(file)) + goto out; @@ -35699,6 +38338,15 @@ diff -urN /usr/share/empty/fs/aufs/xino.c linux/fs/aufs/xino.c + sbinfo->si_xib = file; + sbinfo->si_xread = vfs_readf(file); + sbinfo->si_xwrite = vfs_writef(file); ++ xi_sb = file_inode(file)->i_sb; ++ sbinfo->si_ximaxent = xi_sb->s_maxbytes; ++ if (unlikely(sbinfo->si_ximaxent < PAGE_SIZE)) { ++ err = -EIO; ++ pr_err("s_maxbytes(%llu) on %s is too small\n", ++ (u64)sbinfo->si_ximaxent, au_sbtype(xi_sb)); ++ goto out_unset; ++ } ++ sbinfo->si_ximaxent /= sizeof(ino_t); + + err = -ENOMEM; + if (!sbinfo->si_xib_buf) @@ -35720,16 +38368,15 @@ diff -urN /usr/share/empty/fs/aufs/xino.c linux/fs/aufs/xino.c + +out_free: + if (sbinfo->si_xib_buf) -+ au_delayed_free_page((unsigned long)sbinfo->si_xib_buf); ++ free_page((unsigned long)sbinfo->si_xib_buf); + sbinfo->si_xib_buf = NULL; + if (err >= 0) + err = -EIO; +out_unset: + fput(sbinfo->si_xib); + sbinfo->si_xib = NULL; -+ sbinfo->si_xread = NULL; -+ sbinfo->si_xwrite = NULL; +out: ++ AuTraceErr(err); + return err; +} + @@ -35742,79 +38389,100 @@ diff -urN /usr/share/empty/fs/aufs/xino.c linux/fs/aufs/xino.c + bbot = au_sbbot(sb); + for (bindex = 0; bindex <= bbot; bindex++) { + br = au_sbr(sb, bindex); -+ if (!br || !br->br_xino.xi_file) -+ continue; ++ AuDebugOn(!br); ++ au_xino_put(br); ++ } ++} + -+ fput(br->br_xino.xi_file); -+ br->br_xino.xi_file = NULL; ++static void au_xino_set_br_shared(struct super_block *sb, struct au_branch *br, ++ aufs_bindex_t bshared) ++{ ++ struct au_branch *brshared; ++ ++ brshared = au_sbr(sb, bshared); ++ AuDebugOn(!brshared->br_xino); ++ AuDebugOn(!brshared->br_xino->xi_file); ++ if (br->br_xino != brshared->br_xino) { ++ au_xino_get(brshared); ++ au_xino_put(br); ++ br->br_xino = brshared->br_xino; + } +} + -+static int au_xino_set_br(struct super_block *sb, struct file *base) ++struct au_xino_do_set_br { ++ vfs_writef_t writef; ++ struct au_branch *br; ++ ino_t h_ino; ++ aufs_bindex_t bshared; ++}; ++ ++static int au_xino_do_set_br(struct super_block *sb, struct path *path, ++ struct au_xino_do_set_br *args) +{ + int err; -+ ino_t ino; -+ aufs_bindex_t bindex, bbot, bshared; -+ struct { -+ struct file *old, *new; -+ } *fpair, *p; ++ struct au_xi_calc calc; ++ struct file *file; + struct au_branch *br; ++ struct au_xi_new xinew = { ++ .base = path ++ }; ++ ++ br = args->br; ++ xinew.xi = br->br_xino; ++ au_xi_calc(sb, args->h_ino, &calc); ++ xinew.copy_src = au_xino_file(xinew.xi, calc.idx); ++ if (args->bshared >= 0) ++ /* shared xino */ ++ au_xino_set_br_shared(sb, br, args->bshared); ++ else if (!xinew.xi) { ++ /* new xino */ ++ err = au_xino_init(br, calc.idx, xinew.copy_src); ++ if (unlikely(err)) ++ goto out; ++ } ++ ++ /* force re-creating */ ++ xinew.xi = br->br_xino; ++ xinew.idx = calc.idx; ++ mutex_lock(&xinew.xi->xi_mtx); ++ file = au_xi_new(sb, &xinew); ++ mutex_unlock(&xinew.xi->xi_mtx); ++ err = PTR_ERR(file); ++ if (IS_ERR(file)) ++ goto out; ++ AuDebugOn(!file); ++ ++ err = au_xino_do_write(args->writef, file, &calc, AUFS_ROOT_INO); ++ if (unlikely(err)) ++ au_xino_put(br); ++ ++out: ++ AuTraceErr(err); ++ return err; ++} ++ ++static int au_xino_set_br(struct super_block *sb, struct path *path) ++{ ++ int err; ++ aufs_bindex_t bindex, bbot; ++ struct au_xino_do_set_br args; + struct inode *inode; -+ vfs_writef_t writef; + + SiMustWriteLock(sb); + -+ err = -ENOMEM; + bbot = au_sbbot(sb); -+ fpair = kcalloc(bbot + 1, sizeof(*fpair), GFP_NOFS); -+ if (unlikely(!fpair)) -+ goto out; -+ + inode = d_inode(sb->s_root); -+ ino = AUFS_ROOT_INO; -+ writef = au_sbi(sb)->si_xwrite; -+ for (bindex = 0, p = fpair; bindex <= bbot; bindex++, p++) { -+ bshared = is_sb_shared(sb, bindex, bindex - 1); -+ if (bshared >= 0) { -+ /* shared xino */ -+ *p = fpair[bshared]; -+ get_file(p->new); -+ } -+ -+ if (!p->new) { -+ /* new xino */ -+ br = au_sbr(sb, bindex); -+ p->old = br->br_xino.xi_file; -+ p->new = au_xino_create2(base, br->br_xino.xi_file); -+ err = PTR_ERR(p->new); -+ if (IS_ERR(p->new)) { -+ p->new = NULL; -+ goto out_pair; -+ } -+ } -+ -+ err = au_xino_do_write(writef, p->new, -+ au_h_iptr(inode, bindex)->i_ino, ino); ++ args.writef = au_sbi(sb)->si_xwrite; ++ for (bindex = 0; bindex <= bbot; bindex++) { ++ args.h_ino = au_h_iptr(inode, bindex)->i_ino; ++ args.br = au_sbr(sb, bindex); ++ args.bshared = is_sb_shared(sb, bindex, bindex - 1); ++ err = au_xino_do_set_br(sb, path, &args); + if (unlikely(err)) -+ goto out_pair; -+ } -+ -+ for (bindex = 0, p = fpair; bindex <= bbot; bindex++, p++) { -+ br = au_sbr(sb, bindex); -+ if (br->br_xino.xi_file) -+ fput(br->br_xino.xi_file); -+ get_file(p->new); -+ br->br_xino.xi_file = p->new; ++ break; + } + -+out_pair: -+ for (bindex = 0, p = fpair; bindex <= bbot; bindex++, p++) -+ if (p->new) -+ fput(p->new); -+ else -+ break; -+ au_delayed_kfree(fpair); -+out: ++ AuTraceErr(err); + return err; +} + @@ -35825,32 +38493,37 @@ diff -urN /usr/share/empty/fs/aufs/xino.c linux/fs/aufs/xino.c + au_xigen_clr(sb); + xino_clear_xib(sb); + xino_clear_br(sb); ++ dbgaufs_brs_del(sb, 0); + sbinfo = au_sbi(sb); + /* lvalue, do not call au_mntflags() */ + au_opt_clr(sbinfo->si_mntflags, XINO); +} + -+int au_xino_set(struct super_block *sb, struct au_opt_xino *xino, int remount) ++int au_xino_set(struct super_block *sb, struct au_opt_xino *xiopt, int remount) +{ + int err, skip; -+ struct dentry *parent, *cur_parent; ++ struct dentry *dentry, *parent, *cur_dentry, *cur_parent; + struct qstr *dname, *cur_name; + struct file *cur_xino; -+ struct inode *dir; + struct au_sbinfo *sbinfo; ++ struct path *path, *cur_path; + + SiMustWriteLock(sb); + + err = 0; + sbinfo = au_sbi(sb); -+ parent = dget_parent(xino->file->f_path.dentry); ++ path = &xiopt->file->f_path; ++ dentry = path->dentry; ++ parent = dget_parent(dentry); + if (remount) { + skip = 0; -+ dname = &xino->file->f_path.dentry->d_name; + cur_xino = sbinfo->si_xib; + if (cur_xino) { -+ cur_parent = dget_parent(cur_xino->f_path.dentry); -+ cur_name = &cur_xino->f_path.dentry->d_name; ++ cur_path = &cur_xino->f_path; ++ cur_dentry = cur_path->dentry; ++ cur_parent = dget_parent(cur_dentry); ++ cur_name = &cur_dentry->d_name; ++ dname = &dentry->d_name; + skip = (cur_parent == parent + && au_qstreq(dname, cur_name)); + dput(cur_parent); @@ -35860,30 +38533,26 @@ diff -urN /usr/share/empty/fs/aufs/xino.c linux/fs/aufs/xino.c + } + + au_opt_set(sbinfo->si_mntflags, XINO); -+ dir = d_inode(parent); -+ inode_lock_nested(dir, AuLsc_I_PARENT); -+ /* mnt_want_write() is unnecessary here */ -+ err = au_xino_set_xib(sb, xino->file); ++ err = au_xino_set_xib(sb, path); ++ /* si_x{read,write} are set */ + if (!err) -+ err = au_xigen_set(sb, xino->file); -+ if (!err) -+ err = au_xino_set_br(sb, xino->file); -+ inode_unlock(dir); ++ err = au_xigen_set(sb, path); + if (!err) ++ err = au_xino_set_br(sb, path); ++ if (!err) { ++ dbgaufs_brs_add(sb, 0, /*topdown*/1); + goto out; /* success */ ++ } + + /* reset all */ -+ AuIOErr("failed creating xino(%d).\n", err); -+ au_xigen_clr(sb); -+ xino_clear_xib(sb); ++ AuIOErr("failed setting xino(%d).\n", err); ++ au_xino_clr(sb); + +out: + dput(parent); + return err; +} + -+/* ---------------------------------------------------------------------- */ -+ +/* + * create a xinofile at the default place/path. + */ @@ -35920,13 +38589,12 @@ diff -urN /usr/share/empty/fs/aufs/xino.c linux/fs/aufs/xino.c + if (!IS_ERR(p)) { + strcat(p, "/" AUFS_XINO_FNAME); + AuDbg("%s\n", p); -+ file = au_xino_create(sb, p, /*silent*/0); -+ if (!IS_ERR(file)) -+ au_xino_brid_set(sb, br->br_id); ++ file = au_xino_create(sb, p, /*silent*/0, /*wbrtop*/1); + } -+ au_delayed_free_page((unsigned long)page); ++ free_page((unsigned long)page); + } else { -+ file = au_xino_create(sb, AUFS_XINO_DEFPATH, /*silent*/0); ++ file = au_xino_create(sb, AUFS_XINO_DEFPATH, /*silent*/0, ++ /*wbrtop*/0); + if (IS_ERR(file)) + goto out; + h_sb = file->f_path.dentry->d_sb; @@ -35936,8 +38604,6 @@ diff -urN /usr/share/empty/fs/aufs/xino.c linux/fs/aufs/xino.c + fput(file); + file = ERR_PTR(-EINVAL); + } -+ if (!IS_ERR(file)) -+ au_xino_brid_set(sb, -1); + } + +out: @@ -35946,50 +38612,165 @@ diff -urN /usr/share/empty/fs/aufs/xino.c linux/fs/aufs/xino.c + +/* ---------------------------------------------------------------------- */ + -+int au_xino_path(struct seq_file *seq, struct file *file) ++/* ++ * initialize the xinofile for the specified branch @br ++ * at the place/path where @base_file indicates. ++ * test whether another branch is on the same filesystem or not, ++ * if found then share the xinofile with another branch. ++ */ ++int au_xino_init_br(struct super_block *sb, struct au_branch *br, ino_t h_ino, ++ struct path *base) +{ + int err; ++ struct au_xino_do_set_br args = { ++ .h_ino = h_ino, ++ .br = br ++ }; + -+ err = au_seq_path(seq, &file->f_path); ++ args.writef = au_sbi(sb)->si_xwrite; ++ args.bshared = sbr_find_shared(sb, /*btop*/0, au_sbbot(sb), ++ au_br_sb(br)); ++ err = au_xino_do_set_br(sb, base, &args); + if (unlikely(err)) -+ goto out; -+ -+#define Deleted "\\040(deleted)" -+ seq->count -= sizeof(Deleted) - 1; -+ AuDebugOn(memcmp(seq->buf + seq->count, Deleted, -+ sizeof(Deleted) - 1)); -+#undef Deleted ++ au_xino_put(br); + -+out: + return err; +} + +/* ---------------------------------------------------------------------- */ + -+void au_xinondir_leave(struct super_block *sb, aufs_bindex_t bindex, -+ ino_t h_ino, int idx) ++/* ++ * get an unused inode number from bitmap ++ */ ++ino_t au_xino_new_ino(struct super_block *sb) +{ -+ struct au_xino_file *xino; ++ ino_t ino; ++ unsigned long *p, pindex, ul, pend; ++ struct au_sbinfo *sbinfo; ++ struct file *file; ++ int free_bit, err; + -+ AuDebugOn(!au_opt_test(au_mntflags(sb), XINO)); -+ xino = &au_sbr(sb, bindex)->br_xino; -+ AuDebugOn(idx < 0 || xino->xi_nondir.total <= idx); ++ if (!au_opt_test(au_mntflags(sb), XINO)) ++ return iunique(sb, AUFS_FIRST_INO); ++ ++ sbinfo = au_sbi(sb); ++ mutex_lock(&sbinfo->si_xib_mtx); ++ p = sbinfo->si_xib_buf; ++ free_bit = sbinfo->si_xib_next_bit; ++ if (free_bit < page_bits && !test_bit(free_bit, p)) ++ goto out; /* success */ ++ free_bit = find_first_zero_bit(p, page_bits); ++ if (free_bit < page_bits) ++ goto out; /* success */ ++ ++ pindex = sbinfo->si_xib_last_pindex; ++ for (ul = pindex - 1; ul < ULONG_MAX; ul--) { ++ err = xib_pindex(sb, ul); ++ if (unlikely(err)) ++ goto out_err; ++ free_bit = find_first_zero_bit(p, page_bits); ++ if (free_bit < page_bits) ++ goto out; /* success */ ++ } ++ ++ file = sbinfo->si_xib; ++ pend = vfsub_f_size_read(file) / PAGE_SIZE; ++ for (ul = pindex + 1; ul <= pend; ul++) { ++ err = xib_pindex(sb, ul); ++ if (unlikely(err)) ++ goto out_err; ++ free_bit = find_first_zero_bit(p, page_bits); ++ if (free_bit < page_bits) ++ goto out; /* success */ ++ } ++ BUG(); ++ ++out: ++ set_bit(free_bit, p); ++ sbinfo->si_xib_next_bit = free_bit + 1; ++ pindex = sbinfo->si_xib_last_pindex; ++ mutex_unlock(&sbinfo->si_xib_mtx); ++ ino = xib_calc_ino(pindex, free_bit); ++ AuDbg("i%lu\n", (unsigned long)ino); ++ return ino; ++out_err: ++ mutex_unlock(&sbinfo->si_xib_mtx); ++ AuDbg("i0\n"); ++ return 0; ++} ++ ++/* for s_op->delete_inode() */ ++void au_xino_delete_inode(struct inode *inode, const int unlinked) ++{ ++ int err; ++ unsigned int mnt_flags; ++ aufs_bindex_t bindex, bbot, bi; ++ unsigned char try_trunc; ++ struct au_iinfo *iinfo; ++ struct super_block *sb; ++ struct au_hinode *hi; ++ struct inode *h_inode; ++ struct au_branch *br; ++ vfs_writef_t xwrite; ++ struct au_xi_calc calc; ++ struct file *file; ++ ++ AuDebugOn(au_is_bad_inode(inode)); ++ ++ sb = inode->i_sb; ++ mnt_flags = au_mntflags(sb); ++ if (!au_opt_test(mnt_flags, XINO) ++ || inode->i_ino == AUFS_ROOT_INO) ++ return; ++ ++ if (unlinked) { ++ au_xigen_inc(inode); ++ au_xib_clear_bit(inode); ++ } ++ ++ iinfo = au_ii(inode); ++ bindex = iinfo->ii_btop; ++ if (bindex < 0) ++ return; ++ ++ xwrite = au_sbi(sb)->si_xwrite; ++ try_trunc = !!au_opt_test(mnt_flags, TRUNC_XINO); ++ hi = au_hinode(iinfo, bindex); ++ bbot = iinfo->ii_bbot; ++ for (; bindex <= bbot; bindex++, hi++) { ++ h_inode = hi->hi_inode; ++ if (!h_inode ++ || (!unlinked && h_inode->i_nlink)) ++ continue; + -+ spin_lock(&xino->xi_nondir.spin); -+ AuDebugOn(xino->xi_nondir.array[idx] != h_ino); -+ xino->xi_nondir.array[idx] = 0; -+ spin_unlock(&xino->xi_nondir.spin); -+ wake_up_all(&xino->xi_nondir.wqh); ++ /* inode may not be revalidated */ ++ bi = au_br_index(sb, hi->hi_id); ++ if (bi < 0) ++ continue; ++ ++ br = au_sbr(sb, bi); ++ au_xi_calc(sb, h_inode->i_ino, &calc); ++ file = au_xino_file(br->br_xino, calc.idx); ++ if (IS_ERR_OR_NULL(file)) ++ continue; ++ ++ err = au_xino_do_write(xwrite, file, &calc, /*ino*/0); ++ if (!err && try_trunc ++ && au_test_fs_trunc_xino(au_br_sb(br))) ++ xino_try_trunc(sb, br); ++ } +} + -+static int au_xinondir_find(struct au_xino_file *xino, ino_t h_ino) ++/* ---------------------------------------------------------------------- */ ++ ++static int au_xinondir_find(struct au_xino *xi, ino_t h_ino) +{ + int found, total, i; + + found = -1; -+ total = xino->xi_nondir.total; ++ total = xi->xi_nondir.total; + for (i = 0; i < total; i++) { -+ if (xino->xi_nondir.array[i] != h_ino) ++ if (xi->xi_nondir.array[i] != h_ino) + continue; + found = i; + break; @@ -35998,7 +38779,7 @@ diff -urN /usr/share/empty/fs/aufs/xino.c linux/fs/aufs/xino.c + return found; +} + -+static int au_xinondir_expand(struct au_xino_file *xino) ++static int au_xinondir_expand(struct au_xino *xi) +{ + int err, sz; + ino_t *p; @@ -36006,15 +38787,15 @@ diff -urN /usr/share/empty/fs/aufs/xino.c linux/fs/aufs/xino.c + BUILD_BUG_ON(KMALLOC_MAX_SIZE > INT_MAX); + + err = -ENOMEM; -+ sz = xino->xi_nondir.total * sizeof(ino_t); ++ sz = xi->xi_nondir.total * sizeof(ino_t); + if (unlikely(sz > KMALLOC_MAX_SIZE / 2)) + goto out; -+ p = au_kzrealloc(xino->xi_nondir.array, sz, sz << 1, GFP_ATOMIC, ++ p = au_kzrealloc(xi->xi_nondir.array, sz, sz << 1, GFP_ATOMIC, + /*may_shrink*/0); + if (p) { -+ xino->xi_nondir.array = p; -+ xino->xi_nondir.total <<= 1; -+ AuDbg("xi_nondir.total %d\n", xino->xi_nondir.total); ++ xi->xi_nondir.array = p; ++ xi->xi_nondir.total <<= 1; ++ AuDbg("xi_nondir.total %d\n", xi->xi_nondir.total); + err = 0; + } + @@ -36022,50 +38803,87 @@ diff -urN /usr/share/empty/fs/aufs/xino.c linux/fs/aufs/xino.c + return err; +} + ++void au_xinondir_leave(struct super_block *sb, aufs_bindex_t bindex, ++ ino_t h_ino, int idx) ++{ ++ struct au_xino *xi; ++ ++ AuDebugOn(!au_opt_test(au_mntflags(sb), XINO)); ++ xi = au_sbr(sb, bindex)->br_xino; ++ AuDebugOn(idx < 0 || xi->xi_nondir.total <= idx); ++ ++ spin_lock(&xi->xi_nondir.spin); ++ AuDebugOn(xi->xi_nondir.array[idx] != h_ino); ++ xi->xi_nondir.array[idx] = 0; ++ spin_unlock(&xi->xi_nondir.spin); ++ wake_up_all(&xi->xi_nondir.wqh); ++} ++ +int au_xinondir_enter(struct super_block *sb, aufs_bindex_t bindex, ino_t h_ino, + int *idx) +{ + int err, found, empty; -+ struct au_xino_file *xino; ++ struct au_xino *xi; + + err = 0; + *idx = -1; + if (!au_opt_test(au_mntflags(sb), XINO)) + goto out; /* no xino */ + -+ xino = &au_sbr(sb, bindex)->br_xino; ++ xi = au_sbr(sb, bindex)->br_xino; + +again: -+ spin_lock(&xino->xi_nondir.spin); -+ found = au_xinondir_find(xino, h_ino); ++ spin_lock(&xi->xi_nondir.spin); ++ found = au_xinondir_find(xi, h_ino); + if (found == -1) { -+ empty = au_xinondir_find(xino, /*h_ino*/0); ++ empty = au_xinondir_find(xi, /*h_ino*/0); + if (empty == -1) { -+ empty = xino->xi_nondir.total; -+ err = au_xinondir_expand(xino); ++ empty = xi->xi_nondir.total; ++ err = au_xinondir_expand(xi); + if (unlikely(err)) + goto out_unlock; + } -+ xino->xi_nondir.array[empty] = h_ino; ++ xi->xi_nondir.array[empty] = h_ino; + *idx = empty; + } else { -+ spin_unlock(&xino->xi_nondir.spin); -+ wait_event(xino->xi_nondir.wqh, -+ xino->xi_nondir.array[found] != h_ino); ++ spin_unlock(&xi->xi_nondir.spin); ++ wait_event(xi->xi_nondir.wqh, ++ xi->xi_nondir.array[found] != h_ino); + goto again; + } + +out_unlock: -+ spin_unlock(&xino->xi_nondir.spin); ++ spin_unlock(&xi->xi_nondir.spin); ++out: ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++int au_xino_path(struct seq_file *seq, struct file *file) ++{ ++ int err; ++ ++ err = au_seq_path(seq, &file->f_path); ++ if (unlikely(err)) ++ goto out; ++ ++#define Deleted "\\040(deleted)" ++ seq->count -= sizeof(Deleted) - 1; ++ AuDebugOn(memcmp(seq->buf + seq->count, Deleted, ++ sizeof(Deleted) - 1)); ++#undef Deleted ++ +out: + return err; +} diff -urN /usr/share/empty/include/uapi/linux/aufs_type.h linux/include/uapi/linux/aufs_type.h --- /usr/share/empty/include/uapi/linux/aufs_type.h 1970-01-01 01:00:00.000000000 +0100 -+++ linux/include/uapi/linux/aufs_type.h 2017-05-06 22:16:52.784886929 +0200 -@@ -0,0 +1,419 @@ ++++ linux/include/uapi/linux/aufs_type.h 2019-07-08 09:31:38.327718137 +0200 +@@ -0,0 +1,448 @@ ++/* SPDX-License-Identifier: GPL-2.0 */ +/* -+ * Copyright (C) 2005-2017 Junjiro R. Okajima ++ * Copyright (C) 2005-2019 Junjiro R. Okajima + * + * This program, aufs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by @@ -36105,7 +38923,7 @@ diff -urN /usr/share/empty/include/uapi/linux/aufs_type.h linux/include/uapi/lin + +#include + -+#define AUFS_VERSION "4.x-rcN-20170410" ++#define AUFS_VERSION "4.x-rcN-20190610" + +/* todo? move this to linux-2.6.19/include/magic.h */ +#define AUFS_SUPER_MAGIC ('a' << 24 | 'u' << 16 | 'f' << 8 | 's') @@ -36167,6 +38985,13 @@ diff -urN /usr/share/empty/include/uapi/linux/aufs_type.h linux/include/uapi/lin +#define AUFS_PLINK_MAINT_DIR "fs/" AUFS_NAME +#define AUFS_PLINK_MAINT_PATH AUFS_PLINK_MAINT_DIR "/" AUFS_PLINK_MAINT_NAME + ++/* dirren, renamed dir */ ++#define AUFS_DR_INFO_PFX AUFS_WH_PFX ".dr." ++#define AUFS_DR_BRHINO_NAME AUFS_WH_PFX "hino" ++/* whiteouted doubly */ ++#define AUFS_WH_DR_INFO_PFX AUFS_WH_PFX AUFS_DR_INFO_PFX ++#define AUFS_WH_DR_BRHINO AUFS_WH_PFX AUFS_DR_BRHINO_NAME ++ +#define AUFS_DIROPQ_NAME AUFS_WH_PFX ".opq" /* whiteouted doubly */ +#define AUFS_WH_DIROPQ AUFS_WH_PFX AUFS_DIROPQ_NAME + @@ -36385,6 +39210,27 @@ diff -urN /usr/share/empty/include/uapi/linux/aufs_type.h linux/include/uapi/lin + +/* ---------------------------------------------------------------------- */ + ++/* dirren. the branch is identified by the filename who contains this */ ++struct au_drinfo { ++ uint64_t ino; ++ union { ++ uint8_t oldnamelen; ++ uint64_t _padding; ++ }; ++ uint8_t oldname[0]; ++} __aligned(8); ++ ++struct au_drinfo_fdata { ++ uint32_t magic; ++ struct au_drinfo drinfo; ++} __aligned(8); ++ ++#define AUFS_DRINFO_MAGIC_V1 ('a' << 24 | 'd' << 16 | 'r' << 8 | 0x01) ++/* future */ ++#define AUFS_DRINFO_MAGIC_V2 ('a' << 24 | 'd' << 16 | 'r' << 8 | 0x02) ++ ++/* ---------------------------------------------------------------------- */ ++ +struct aufs_wbr_fd { + uint32_t oflags; + int16_t brid; @@ -36483,52 +39329,17 @@ diff -urN /usr/share/empty/include/uapi/linux/aufs_type.h linux/include/uapi/lin +#define AUFS_CTL_FHSM_FD _IOW(AuCtlType, AuCtl_FHSM_FD, int) + +#endif /* __AUFS_TYPE_H__ */ +SPDX-License-Identifier: GPL-2.0 aufs4.x-rcN loopback patch diff --git a/drivers/block/loop.c b/drivers/block/loop.c -index cf39138..29b33b2 100644 +index a271e17e4b10..ee17c7577224 100644 --- a/drivers/block/loop.c +++ b/drivers/block/loop.c -@@ -551,7 +551,7 @@ static int do_req_filebacked(struct loop_device *lo, struct request *rq) +@@ -625,6 +625,15 @@ static inline void loop_update_dio(struct loop_device *lo) + lo->use_dio); } - struct switch_request { -- struct file *file; -+ struct file *file, *virt_file; - struct completion wait; - }; - -@@ -577,6 +577,7 @@ static void do_loop_switch(struct loop_device *lo, struct switch_request *p) - mapping = file->f_mapping; - mapping_set_gfp_mask(old_file->f_mapping, lo->old_gfp_mask); - lo->lo_backing_file = file; -+ lo->lo_backing_virt_file = p->virt_file; - lo->lo_blocksize = S_ISBLK(mapping->host->i_mode) ? - mapping->host->i_bdev->bd_block_size : PAGE_SIZE; - lo->old_gfp_mask = mapping_gfp_mask(mapping); -@@ -589,11 +590,13 @@ static void do_loop_switch(struct loop_device *lo, struct switch_request *p) - * First it needs to flush existing IO, it does this by sending a magic - * BIO down the pipe. The completion of this BIO does the actual switch. - */ --static int loop_switch(struct loop_device *lo, struct file *file) -+static int loop_switch(struct loop_device *lo, struct file *file, -+ struct file *virt_file) - { - struct switch_request w; - - w.file = file; -+ w.virt_file = virt_file; - - /* freeze queue and wait for completion of scheduled requests */ - blk_mq_freeze_queue(lo->lo_queue); -@@ -612,7 +615,16 @@ static int loop_switch(struct loop_device *lo, struct file *file) - */ - static int loop_flush(struct loop_device *lo) - { -- return loop_switch(lo, NULL); -+ return loop_switch(lo, NULL, NULL); -+} -+ +static struct file *loop_real_file(struct file *file) +{ + struct file *f = NULL; @@ -36536,21 +39347,23 @@ index cf39138..29b33b2 100644 + if (file->f_path.dentry->d_sb->s_op->real_loop) + f = file->f_path.dentry->d_sb->s_op->real_loop(file); + return f; - } - ++} ++ static void loop_reread_partitions(struct loop_device *lo, -@@ -649,6 +661,7 @@ static int loop_change_fd(struct loop_device *lo, struct block_device *bdev, + struct block_device *bdev) + { +@@ -678,6 +687,7 @@ static int loop_change_fd(struct loop_device *lo, struct block_device *bdev, unsigned int arg) { - struct file *file, *old_file; + struct file *file = NULL, *old_file; + struct file *f, *virt_file = NULL, *old_virt_file; - struct inode *inode; int error; + bool partscan; -@@ -665,9 +678,16 @@ static int loop_change_fd(struct loop_device *lo, struct block_device *bdev, +@@ -697,12 +707,19 @@ static int loop_change_fd(struct loop_device *lo, struct block_device *bdev, file = fget(arg); if (!file) - goto out; + goto out_err; + f = loop_real_file(file); + if (f) { + virt_file = file; @@ -36558,45 +39371,51 @@ index cf39138..29b33b2 100644 + get_file(file); + } - inode = file->f_mapping->host; + error = loop_validate_file(file, bdev); + if (error) + goto out_err; + old_file = lo->lo_backing_file; + old_virt_file = lo->lo_backing_virt_file; error = -EINVAL; -@@ -679,17 +699,21 @@ static int loop_change_fd(struct loop_device *lo, struct block_device *bdev, - goto out_putf; - - /* and ... switch */ -- error = loop_switch(lo, file); -+ error = loop_switch(lo, file, virt_file); - if (error) - goto out_putf; - +@@ -714,6 +731,7 @@ static int loop_change_fd(struct loop_device *lo, struct block_device *bdev, + blk_mq_freeze_queue(lo->lo_queue); + mapping_set_gfp_mask(old_file->f_mapping, lo->old_gfp_mask); + lo->lo_backing_file = file; ++ lo->lo_backing_virt_file = virt_file; + lo->old_gfp_mask = mapping_gfp_mask(file->f_mapping); + mapping_set_gfp_mask(file->f_mapping, + lo->old_gfp_mask & ~(__GFP_IO|__GFP_FS)); +@@ -727,6 +745,8 @@ static int loop_change_fd(struct loop_device *lo, struct block_device *bdev, + * dependency. + */ fput(old_file); + if (old_virt_file) + fput(old_virt_file); - if (lo->lo_flags & LO_FLAGS_PARTSCAN) + if (partscan) loop_reread_partitions(lo, bdev); return 0; - - out_putf: - fput(file); +@@ -735,6 +755,8 @@ static int loop_change_fd(struct loop_device *lo, struct block_device *bdev, + mutex_unlock(&loop_ctl_mutex); + if (file) + fput(file); + if (virt_file) + fput(virt_file); - out: return error; } -@@ -876,7 +900,7 @@ static int loop_prepare_queue(struct loop_device *lo) + +@@ -921,7 +943,7 @@ static int loop_prepare_queue(struct loop_device *lo) static int loop_set_fd(struct loop_device *lo, fmode_t mode, struct block_device *bdev, unsigned int arg) { -- struct file *file, *f; +- struct file *file; + struct file *file, *f, *virt_file = NULL; struct inode *inode; struct address_space *mapping; - unsigned lo_blocksize; -@@ -891,6 +915,12 @@ static int loop_set_fd(struct loop_device *lo, fmode_t mode, + int lo_flags = 0; +@@ -936,6 +958,12 @@ static int loop_set_fd(struct loop_device *lo, fmode_t mode, file = fget(arg); if (!file) goto out; @@ -36607,9 +39426,9 @@ index cf39138..29b33b2 100644 + get_file(file); + } - error = -EBUSY; - if (lo->lo_state != Lo_unbound) -@@ -943,6 +973,7 @@ static int loop_set_fd(struct loop_device *lo, fmode_t mode, + error = mutex_lock_killable(&loop_ctl_mutex); + if (error) +@@ -972,6 +1000,7 @@ static int loop_set_fd(struct loop_device *lo, fmode_t mode, lo->lo_device = bdev; lo->lo_flags = lo_flags; lo->lo_backing_file = file; @@ -36617,42 +39436,42 @@ index cf39138..29b33b2 100644 lo->transfer = NULL; lo->ioctl = NULL; lo->lo_sizelimit = 0; -@@ -975,6 +1006,8 @@ static int loop_set_fd(struct loop_device *lo, fmode_t mode, - - out_putf: +@@ -1009,6 +1038,8 @@ static int loop_set_fd(struct loop_device *lo, fmode_t mode, + mutex_unlock(&loop_ctl_mutex); + out_putf: fput(file); + if (virt_file) + fput(virt_file); - out: + out: /* This is safe: open() is still holding a reference. */ module_put(THIS_MODULE); -@@ -1021,6 +1054,7 @@ loop_init_xfer(struct loop_device *lo, struct loop_func_table *xfer, - static int loop_clr_fd(struct loop_device *lo) +@@ -1055,6 +1086,7 @@ loop_init_xfer(struct loop_device *lo, struct loop_func_table *xfer, + static int __loop_clr_fd(struct loop_device *lo, bool release) { - struct file *filp = lo->lo_backing_file; + struct file *filp = NULL; + struct file *virt_filp = lo->lo_backing_virt_file; gfp_t gfp = lo->old_gfp_mask; struct block_device *bdev = lo->lo_device; + int err = 0; +@@ -1078,6 +1110,7 @@ static int __loop_clr_fd(struct loop_device *lo, bool release) -@@ -1052,6 +1086,7 @@ static int loop_clr_fd(struct loop_device *lo) spin_lock_irq(&lo->lo_lock); - lo->lo_state = Lo_rundown; lo->lo_backing_file = NULL; + lo->lo_backing_virt_file = NULL; spin_unlock_irq(&lo->lo_lock); loop_release_xfer(lo); -@@ -1096,6 +1131,8 @@ static int loop_clr_fd(struct loop_device *lo) - * bd_mutex which is usually taken before lo_ctl_mutex. +@@ -1146,6 +1179,8 @@ static int __loop_clr_fd(struct loop_device *lo, bool release) */ - fput(filp); + if (filp) + fput(filp); + if (virt_filp) + fput(virt_filp); - return 0; + return err; } diff --git a/drivers/block/loop.h b/drivers/block/loop.h -index fb2237c..c3888c5 100644 +index af75a5ee4094..1e6ee5a4f623 100644 --- a/drivers/block/loop.h +++ b/drivers/block/loop.h @@ -46,7 +46,7 @@ struct loop_device { @@ -36662,13 +39481,13 @@ index fb2237c..c3888c5 100644 - struct file * lo_backing_file; + struct file * lo_backing_file, *lo_backing_virt_file; struct block_device *lo_device; - unsigned lo_blocksize; void *key_data; + diff --git a/fs/aufs/f_op.c b/fs/aufs/f_op.c -index 7e03507..f31af1d 100644 +index 0309f0d502ff..19feb4f3cb5f 100644 --- a/fs/aufs/f_op.c +++ b/fs/aufs/f_op.c -@@ -360,7 +360,7 @@ static ssize_t aufs_read_iter(struct kiocb *kio, struct iov_iter *iov_iter) +@@ -359,7 +359,7 @@ static ssize_t aufs_read_iter(struct kiocb *kio, struct iov_iter *iov_iter) if (IS_ERR(h_file)) goto out; @@ -36678,12 +39497,12 @@ index 7e03507..f31af1d 100644 if (file->f_mapping != h_file->f_mapping) { file->f_mapping = h_file->f_mapping; diff --git a/fs/aufs/loop.c b/fs/aufs/loop.c -index 61e9197..9910bea 100644 +index f05887d612fd..6f0505723cb4 100644 --- a/fs/aufs/loop.c +++ b/fs/aufs/loop.c -@@ -132,3 +132,19 @@ void au_loopback_fin(void) +@@ -133,3 +133,19 @@ void au_loopback_fin(void) symbol_put(loop_backing_file); - au_delayed_kfree(au_warn_loopback_array); + au_kfree_try_rcu(au_warn_loopback_array); } + +/* ---------------------------------------------------------------------- */ @@ -36702,22 +39521,19 @@ index 61e9197..9910bea 100644 + return f; +} diff --git a/fs/aufs/loop.h b/fs/aufs/loop.h -index e2df495..36e5052 100644 +index f31e40aff267..e13fb1a0717a 100644 --- a/fs/aufs/loop.h +++ b/fs/aufs/loop.h -@@ -25,7 +25,11 @@ void au_warn_loopback(struct super_block *h_sb); +@@ -26,6 +26,8 @@ void au_warn_loopback(struct super_block *h_sb); int au_loopback_init(void); void au_loopback_fin(void); + +struct file *aufs_real_loop(struct file *file); #else -+AuStub(struct file *, loop_backing_file, return NULL) -+ - AuStubInt0(au_test_loopback_overlap, struct super_block *sb, - struct dentry *h_adding) - AuStubInt0(au_test_loopback_kthread, void) -@@ -33,6 +37,8 @@ AuStubVoid(au_warn_loopback, struct super_block *h_sb) + AuStub(struct file *, loop_backing_file, return NULL, struct super_block *sb) + +@@ -36,6 +38,8 @@ AuStubVoid(au_warn_loopback, struct super_block *h_sb) AuStubInt0(au_loopback_init, void) AuStubVoid(au_loopback_fin, void) @@ -36727,10 +39543,10 @@ index e2df495..36e5052 100644 #endif /* __KERNEL__ */ diff --git a/fs/aufs/super.c b/fs/aufs/super.c -index 0e9bbc3..b90d52f 100644 +index 73255f63b80c..f04e211cbf89 100644 --- a/fs/aufs/super.c +++ b/fs/aufs/super.c -@@ -837,7 +837,10 @@ static const struct super_operations aufs_sop = { +@@ -846,7 +846,10 @@ static const struct super_operations aufs_sop = { .statfs = aufs_statfs, .put_super = aufs_put_super, .sync_fs = aufs_sync_fs, @@ -36743,10 +39559,10 @@ index 0e9bbc3..b90d52f 100644 /* ---------------------------------------------------------------------- */ diff --git a/include/linux/fs.h b/include/linux/fs.h -index 3668627..2cbc8fe 100644 +index 5db265e7d35a..8895cd2343e3 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h -@@ -1810,6 +1810,10 @@ struct super_operations { +@@ -1937,6 +1937,10 @@ struct super_operations { struct shrink_control *); long (*free_cached_objects)(struct super_block *, struct shrink_control *);