From: Jeff Mahoney Subject: vfs: add super_operations->get_inode_dev References: bsc#927455 bsc#1176983 Patch-mainline: Never, Upstream submissions have been met with a request to use per-subvolume superblocks, which is unworkable in practice There are many places where a dev_t:ino_t pair are passed to userspace to uniquely describe an inode. Some file systems, like btrfs, have multiple inode namespace internally and use a separate dev_t to make the distinction between them. This patch adds a super_operations->get_inode_dev operation to allow the file system to export those dev_ts to callers within the kernel for further export to userspace. Without this patch, things like audit and some perf and trace events will not distinguish between subvolumes within a single btrfs filesystem. Signed-off-by: Jeff Mahoney --- arch/arc/kernel/troubleshoot.c | 2 +- drivers/staging/lustre/lustre/llite/dir.c | 2 +- drivers/staging/lustre/lustre/llite/file.c | 2 +- fs/locks.c | 4 ++-- fs/nfsd/nfs3xdr.c | 2 +- fs/nfsd/vfs.c | 6 +++--- fs/notify/fdinfo.c | 4 ++-- fs/proc/nommu.c | 2 +- fs/proc/task_mmu.c | 2 +- fs/proc/task_nommu.c | 2 +- fs/proc_namespace.c | 3 ++- fs/stat.c | 2 +- include/linux/fs.h | 9 +++++++++ include/trace/events/filelock.h | 8 ++++---- include/trace/events/filemap.h | 2 +- include/trace/events/writeback.h | 2 +- kernel/audit.c | 2 +- kernel/audit_fsnotify.c | 2 +- kernel/audit_watch.c | 6 +++--- kernel/auditsc.c | 4 ++-- kernel/events/core.c | 2 +- mm/memory-failure.c | 2 +- net/unix/diag.c | 2 +- security/tomoyo/condition.c | 2 +- 24 files changed, 43 insertions(+), 33 deletions(-) --- a/arch/arc/kernel/troubleshoot.c +++ b/arch/arc/kernel/troubleshoot.c @@ -104,7 +104,7 @@ static void show_faulting_vma(unsigned l if (file) { nm = file_path(file, buf, PAGE_SIZE - 1); inode = file_inode(vma->vm_file); - dev = inode->i_sb->s_dev; + dev = inode_get_dev(inode); ino = inode->i_ino; } pr_info(" @off 0x%lx in [%s]\n" --- a/drivers/staging/lustre/lustre/llite/dir.c +++ b/drivers/staging/lustre/lustre/llite/dir.c @@ -1360,7 +1360,7 @@ skip_lmm: struct lov_user_mds_data __user *lmdp; lstat_t st = { 0 }; - st.st_dev = inode->i_sb->s_dev; + st.st_dev = inode_get_dev(inode); st.st_mode = body->mbo_mode; st.st_nlink = body->mbo_nlink; st.st_uid = body->mbo_uid; --- a/drivers/staging/lustre/lustre/llite/file.c +++ b/drivers/staging/lustre/lustre/llite/file.c @@ -2964,7 +2964,7 @@ int ll_getattr(const struct path *path, OBD_FAIL_TIMEOUT(OBD_FAIL_GETATTR_DELAY, 30); - stat->dev = inode->i_sb->s_dev; + stat->dev = inode_get_dev(inode); if (ll_need_32bit_api(sbi)) stat->ino = cl_fid_build_ino(&lli->lli_fid, 1); else --- a/fs/locks.c +++ b/fs/locks.c @@ -2682,8 +2682,8 @@ static void lock_get_status(struct seq_f if (inode) { /* userspace relies on this representation of dev_t */ seq_printf(f, "%d %02x:%02x:%lu ", fl_pid, - MAJOR(inode->i_sb->s_dev), - MINOR(inode->i_sb->s_dev), inode->i_ino); + MAJOR(inode_get_dev(inode)), + MINOR(inode_get_dev(inode)), inode->i_ino); } else { seq_printf(f, "%d :0 ", fl_pid); } --- a/fs/nfsd/nfs3xdr.c +++ b/fs/nfsd/nfs3xdr.c @@ -147,7 +147,7 @@ static __be32 *encode_fsid(__be32 *p, st default: case FSIDSOURCE_DEV: p = xdr_encode_hyper(p, (u64)huge_encode_dev - (fhp->fh_dentry->d_sb->s_dev)); + (inode_get_dev(d_inode(fhp->fh_dentry)))); break; case FSIDSOURCE_FSID: p = xdr_encode_hyper(p, (u64) fhp->fh_export->ex_fsid); --- a/fs/nfsd/vfs.c +++ b/fs/nfsd/vfs.c @@ -795,7 +795,7 @@ struct raparms * nfsd_init_raparms(struct file *file) { struct inode *inode = file_inode(file); - dev_t dev = inode->i_sb->s_dev; + dev_t dev = inode_get_dev(inode); ino_t ino = inode->i_ino; struct raparms *ra, **rap, **frap = NULL; int depth = 0; @@ -963,7 +963,7 @@ static int wait_for_concurrent_writes(st int err = 0; if (atomic_read(&inode->i_writecount) > 1 - || (last_ino == inode->i_ino && last_dev == inode->i_sb->s_dev)) { + || (last_ino == inode->i_ino && last_dev == inode_get_dev(inode))) { dprintk("nfsd: write defer %d\n", task_pid_nr(current)); msleep(10); dprintk("nfsd: write resume %d\n", task_pid_nr(current)); @@ -974,7 +974,7 @@ static int wait_for_concurrent_writes(st err = vfs_fsync(file, 0); } last_ino = inode->i_ino; - last_dev = inode->i_sb->s_dev; + last_dev = inode_get_dev(inode); return err; } --- a/fs/notify/fdinfo.c +++ b/fs/notify/fdinfo.c @@ -90,7 +90,7 @@ static void inotify_fdinfo(struct seq_fi */ u32 mask = mark->mask & IN_ALL_EVENTS; seq_printf(m, "inotify wd:%x ino:%lx sdev:%x mask:%x ignored_mask:%x ", - inode_mark->wd, inode->i_ino, inode->i_sb->s_dev, + inode_mark->wd, inode->i_ino, inode_get_dev(inode), mask, mark->ignored_mask); show_mark_fhandle(m, inode); seq_putc(m, '\n'); @@ -120,7 +120,7 @@ static void fanotify_fdinfo(struct seq_f if (!inode) return; seq_printf(m, "fanotify ino:%lx sdev:%x mflags:%x mask:%x ignored_mask:%x ", - inode->i_ino, inode->i_sb->s_dev, + inode->i_ino, inode_get_dev(inode), mflags, mark->mask, mark->ignored_mask); show_mark_fhandle(m, inode); seq_putc(m, '\n'); --- a/fs/proc/nommu.c +++ b/fs/proc/nommu.c @@ -46,7 +46,7 @@ static int nommu_region_show(struct seq_ if (file) { struct inode *inode = file_inode(region->vm_file); - dev = inode->i_sb->s_dev; + dev = inode_get_dev(inode); ino = inode->i_ino; } --- a/fs/proc/task_mmu.c +++ b/fs/proc/task_mmu.c @@ -295,7 +295,7 @@ show_map_vma(struct seq_file *m, struct if (file) { struct inode *inode = file_inode(vma->vm_file); - dev = inode->i_sb->s_dev; + dev = inode_get_dev(inode); ino = inode->i_ino; pgoff = ((loff_t)vma->vm_pgoff) << PAGE_SHIFT; } --- a/fs/proc/task_nommu.c +++ b/fs/proc/task_nommu.c @@ -158,7 +158,7 @@ static int nommu_vma_show(struct seq_fil if (file) { struct inode *inode = file_inode(vma->vm_file); - dev = inode->i_sb->s_dev; + dev = inode_get_dev(inode); ino = inode->i_ino; pgoff = (loff_t)vma->vm_pgoff << PAGE_SHIFT; } --- a/fs/proc_namespace.c +++ b/fs/proc_namespace.c @@ -133,10 +133,11 @@ static int show_mountinfo(struct seq_fil struct mount *r = real_mount(mnt); struct super_block *sb = mnt->mnt_sb; struct path mnt_path = { .dentry = mnt->mnt_root, .mnt = mnt }; + dev_t dev = inode_get_dev(mnt->mnt_root->d_inode); int err; seq_printf(m, "%i %i %u:%u ", r->mnt_id, r->mnt_parent->mnt_id, - MAJOR(sb->s_dev), MINOR(sb->s_dev)); + MAJOR(dev), MINOR(dev)); if (sb->s_op->show_path) { err = sb->s_op->show_path(m, mnt->mnt_root); if (err) --- a/fs/stat.c +++ b/fs/stat.c @@ -31,7 +31,7 @@ */ void generic_fillattr(struct inode *inode, struct kstat *stat) { - stat->dev = inode->i_sb->s_dev; + stat->dev = inode_get_dev(inode); stat->ino = inode->i_ino; stat->mode = inode->i_mode; stat->nlink = inode->i_nlink; --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -1834,6 +1834,7 @@ struct super_operations { struct shrink_control *); long (*free_cached_objects)(struct super_block *, struct shrink_control *); + dev_t (*get_inode_dev)(const struct inode *); }; /* @@ -3363,4 +3364,12 @@ static inline bool dir_relax_shared(stru extern bool path_noexec(const struct path *path); extern void inode_nohighmem(struct inode *inode); +static inline dev_t inode_get_dev(const struct inode *inode) +{ + if (inode->i_sb->s_op->get_inode_dev) + return inode->i_sb->s_op->get_inode_dev(inode); + + return inode->i_sb->s_dev; +} + #endif /* _LINUX_FS_H */ --- a/include/trace/events/filelock.h +++ b/include/trace/events/filelock.h @@ -47,7 +47,7 @@ TRACE_EVENT(locks_get_lock_context, ), TP_fast_assign( - __entry->s_dev = inode->i_sb->s_dev; + __entry->s_dev = inode_get_dev(inode); __entry->i_ino = inode->i_ino; __entry->type = type; __entry->ctx = ctx; @@ -79,7 +79,7 @@ DECLARE_EVENT_CLASS(filelock_lock, TP_fast_assign( __entry->fl = fl ? fl : NULL; - __entry->s_dev = inode->i_sb->s_dev; + __entry->s_dev = inode_get_dev(inode); __entry->i_ino = inode->i_ino; __entry->fl_next = fl ? fl->fl_next : NULL; __entry->fl_owner = fl ? fl->fl_owner : NULL; @@ -131,7 +131,7 @@ DECLARE_EVENT_CLASS(filelock_lease, TP_fast_assign( __entry->fl = fl ? fl : NULL; - __entry->s_dev = inode->i_sb->s_dev; + __entry->s_dev = inode_get_dev(inode); __entry->i_ino = inode->i_ino; __entry->fl_next = fl ? fl->fl_next : NULL; __entry->fl_owner = fl ? fl->fl_owner : NULL; @@ -181,7 +181,7 @@ TRACE_EVENT(generic_add_lease, ), TP_fast_assign( - __entry->s_dev = inode->i_sb->s_dev; + __entry->s_dev = inode_get_dev(inode); __entry->i_ino = inode->i_ino; __entry->wcount = atomic_read(&inode->i_writecount); __entry->dcount = d_count(fl->fl_file->f_path.dentry); --- a/include/trace/events/filemap.h +++ b/include/trace/events/filemap.h @@ -30,7 +30,7 @@ DECLARE_EVENT_CLASS(mm_filemap_op_page_c __entry->i_ino = page->mapping->host->i_ino; __entry->index = page->index; if (page->mapping->host->i_sb) - __entry->s_dev = page->mapping->host->i_sb->s_dev; + __entry->s_dev = inode_get_dev(page->mapping->host); else __entry->s_dev = page->mapping->host->i_rdev; ), --- a/include/trace/events/writeback.h +++ b/include/trace/events/writeback.h @@ -708,7 +708,7 @@ DECLARE_EVENT_CLASS(writeback_inode_temp ), TP_fast_assign( - __entry->dev = inode->i_sb->s_dev; + __entry->dev = inode_get_dev(inode); __entry->ino = inode->i_ino; __entry->state = inode->i_state; __entry->mode = inode->i_mode; --- a/kernel/audit.c +++ b/kernel/audit.c @@ -2050,7 +2050,7 @@ void audit_copy_inode(struct audit_names struct inode *inode) { name->ino = inode->i_ino; - name->dev = inode->i_sb->s_dev; + name->dev = inode_get_dev(inode); name->mode = inode->i_mode; name->uid = inode->i_uid; name->gid = inode->i_gid; --- a/kernel/audit_fsnotify.c +++ b/kernel/audit_fsnotify.c @@ -76,7 +76,7 @@ int audit_mark_compare(struct audit_fsno static void audit_update_mark(struct audit_fsnotify_mark *audit_mark, const struct inode *inode) { - audit_mark->dev = inode ? inode->i_sb->s_dev : AUDIT_DEV_UNSET; + audit_mark->dev = inode ? inode_get_dev(inode) : AUDIT_DEV_UNSET; audit_mark->ino = inode ? inode->i_ino : AUDIT_INO_UNSET; } --- a/kernel/audit_watch.c +++ b/kernel/audit_watch.c @@ -368,7 +368,7 @@ static int audit_get_nd(struct audit_wat inode_unlock(d_backing_inode(parent->dentry)); if (d_is_positive(d)) { /* update watch filter fields */ - watch->dev = d->d_sb->s_dev; + watch->dev = inode_get_dev(d_backing_inode(d)); watch->ino = d_backing_inode(d)->i_ino; } dput(d); @@ -509,7 +509,7 @@ static int audit_watch_handle_event(stru } if (mask & (FS_CREATE|FS_MOVED_TO) && inode) - audit_update_watch(parent, dname, inode->i_sb->s_dev, inode->i_ino, 0); + audit_update_watch(parent, dname, inode_get_dev(inode), inode->i_ino, 0); else if (mask & (FS_DELETE|FS_MOVED_FROM)) audit_update_watch(parent, dname, AUDIT_DEV_UNSET, AUDIT_INO_UNSET, 1); else if (mask & (FS_DELETE_SELF|FS_UNMOUNT|FS_MOVE_SELF)) @@ -563,7 +563,7 @@ int audit_exe_compare(struct task_struct if (!exe_file) return 0; ino = file_inode(exe_file)->i_ino; - dev = file_inode(exe_file)->i_sb->s_dev; + dev = inode_get_dev(file_inode(exe_file)); fput(exe_file); return audit_mark_compare(mark, ino, dev); } --- a/kernel/auditsc.c +++ b/kernel/auditsc.c @@ -1808,7 +1808,7 @@ void __audit_inode(struct filename *name if (n->ino) { /* valid inode number, use that for the comparison */ if (n->ino != inode->i_ino || - n->dev != inode->i_sb->s_dev) + n->dev != inode_get_dev(inode)) continue; } else if (n->name) { /* inode number has not been set, check the name */ @@ -1894,7 +1894,7 @@ void __audit_inode_child(struct inode *p n->type != AUDIT_TYPE_UNKNOWN)) continue; - if (n->ino == parent->i_ino && n->dev == parent->i_sb->s_dev && + if (n->ino == parent->i_ino && n->dev == inode_get_dev(parent) && !audit_compare_dname_path(dname, n->name->name, n->name_len)) { if (n->type == AUDIT_TYPE_UNKNOWN) --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -7169,7 +7169,7 @@ static void perf_event_mmap_event(struct goto cpy_name; } inode = file_inode(vma->vm_file); - dev = inode->i_sb->s_dev; + dev = inode_get_dev(inode); ino = inode->i_ino; gen = inode->i_generation; maj = MAJOR(dev); --- a/mm/memory-failure.c +++ b/mm/memory-failure.c @@ -100,7 +100,7 @@ static int hwpoison_filter_dev(struct pa if (mapping == NULL || mapping->host == NULL) return -EINVAL; - dev = mapping->host->i_sb->s_dev; + dev = inode_get_dev(mapping->host); if (hwpoison_filter_dev_major != ~0U && hwpoison_filter_dev_major != MAJOR(dev)) return -EINVAL; --- a/net/unix/diag.c +++ b/net/unix/diag.c @@ -27,7 +27,7 @@ static int sk_diag_dump_vfs(struct sock if (dentry) { struct unix_diag_vfs uv = { .udiag_vfs_ino = d_backing_inode(dentry)->i_ino, - .udiag_vfs_dev = dentry->d_sb->s_dev, + .udiag_vfs_dev = inode_get_dev(d_backing_inode(dentry)), }; return nla_put(nlskb, UNIX_DIAG_VFS, sizeof(uv), &uv); --- a/security/tomoyo/condition.c +++ b/security/tomoyo/condition.c @@ -721,7 +721,7 @@ void tomoyo_get_attributes(struct tomoyo stat->gid = inode->i_gid; stat->ino = inode->i_ino; stat->mode = inode->i_mode; - stat->dev = inode->i_sb->s_dev; + stat->dev = inode_get_dev(inode); stat->rdev = inode->i_rdev; obj->stat_valid[i] = true; }