diff options
Diffstat (limited to 'fs/f2fs/file.c')
| -rw-r--r-- | fs/f2fs/file.c | 1073 |
1 files changed, 717 insertions, 356 deletions
diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c index 92ec2699bc85..82cda1258227 100644 --- a/fs/f2fs/file.c +++ b/fs/f2fs/file.c @@ -24,6 +24,7 @@ #include <linux/sched/signal.h> #include <linux/fileattr.h> #include <linux/fadvise.h> +#include <linux/iomap.h> #include "f2fs.h" #include "node.h" @@ -42,8 +43,8 @@ static vm_fault_t f2fs_filemap_fault(struct vm_fault *vmf) ret = filemap_fault(vmf); if (!ret) - f2fs_update_iostat(F2FS_I_SB(inode), APP_MAPPED_READ_IO, - F2FS_BLKSIZE); + f2fs_update_iostat(F2FS_I_SB(inode), inode, + APP_MAPPED_READ_IO, F2FS_BLKSIZE); trace_f2fs_filemap_fault(inode, vmf->pgoff, (unsigned long)ret); @@ -153,7 +154,7 @@ static vm_fault_t f2fs_vm_page_mkwrite(struct vm_fault *vmf) if (!PageUptodate(page)) SetPageUptodate(page); - f2fs_update_iostat(sbi, APP_MAPPED_IO, F2FS_BLKSIZE); + f2fs_update_iostat(sbi, inode, APP_MAPPED_IO, F2FS_BLKSIZE); f2fs_update_time(sbi, REQ_TIME); trace_f2fs_vm_page_mkwrite(page, DATA); @@ -236,13 +237,13 @@ static void try_to_fix_pino(struct inode *inode) struct f2fs_inode_info *fi = F2FS_I(inode); nid_t pino; - down_write(&fi->i_sem); + f2fs_down_write(&fi->i_sem); if (file_wrong_pino(inode) && inode->i_nlink == 1 && get_parent_ino(inode, &pino)) { f2fs_i_pino_write(inode, pino); file_got_pino(inode); } - up_write(&fi->i_sem); + f2fs_up_write(&fi->i_sem); } static int f2fs_do_sync_file(struct file *file, loff_t start, loff_t end, @@ -317,9 +318,9 @@ go_write: * Both of fdatasync() and fsync() are able to be recovered from * sudden-power-off. */ - down_read(&F2FS_I(inode)->i_sem); + f2fs_down_read(&F2FS_I(inode)->i_sem); cp_reason = need_do_checkpoint(inode); - up_read(&F2FS_I(inode)->i_sem); + f2fs_up_read(&F2FS_I(inode)->i_sem); if (cp_reason) { /* all the dirty node pages should be flushed for POR */ @@ -371,7 +372,8 @@ sync_nodes: f2fs_remove_ino_entry(sbi, ino, APPEND_INO); clear_inode_flag(inode, FI_APPEND_WRITE); flush_out: - if (!atomic && F2FS_OPTION(sbi).fsync_mode != FSYNC_MODE_NOBARRIER) + if ((!atomic && F2FS_OPTION(sbi).fsync_mode != FSYNC_MODE_NOBARRIER) || + (atomic && !test_opt(sbi, NOBARRIER) && f2fs_sb_has_blkzoned(sbi))) ret = f2fs_issue_flush(sbi, inode->i_ino); if (!ret) { f2fs_remove_ino_entry(sbi, ino, UPDATE_INO); @@ -806,12 +808,40 @@ int f2fs_truncate(struct inode *inode) return 0; } +static bool f2fs_force_buffered_io(struct inode *inode, int rw) +{ + struct f2fs_sb_info *sbi = F2FS_I_SB(inode); + + if (!fscrypt_dio_supported(inode)) + return true; + if (fsverity_active(inode)) + return true; + if (f2fs_compressed_file(inode)) + return true; + + /* disallow direct IO if any of devices has unaligned blksize */ + if (f2fs_is_multi_device(sbi) && !sbi->aligned_blksize) + return true; + /* + * for blkzoned device, fallback direct IO to buffered IO, so + * all IOs can be serialized by log-structured write. + */ + if (f2fs_sb_has_blkzoned(sbi) && (rw == WRITE)) + return true; + if (f2fs_lfs_mode(sbi) && rw == WRITE && F2FS_IO_ALIGNED(sbi)) + return true; + if (is_sbi_flag_set(sbi, SBI_CP_DISABLED)) + return true; + + return false; +} + int f2fs_getattr(struct user_namespace *mnt_userns, const struct path *path, struct kstat *stat, u32 request_mask, unsigned int query_flags) { struct inode *inode = d_inode(path->dentry); struct f2fs_inode_info *fi = F2FS_I(inode); - struct f2fs_inode *ri; + struct f2fs_inode *ri = NULL; unsigned int flags; if (f2fs_has_extra_attr(inode) && @@ -822,6 +852,24 @@ int f2fs_getattr(struct user_namespace *mnt_userns, const struct path *path, stat->btime.tv_nsec = fi->i_crtime.tv_nsec; } + /* + * Return the DIO alignment restrictions if requested. We only return + * this information when requested, since on encrypted files it might + * take a fair bit of work to get if the file wasn't opened recently. + * + * f2fs sometimes supports DIO reads but not DIO writes. STATX_DIOALIGN + * cannot represent that, so in that case we report no DIO support. + */ + if ((request_mask & STATX_DIOALIGN) && S_ISREG(inode->i_mode)) { + unsigned int bsize = i_blocksize(inode); + + stat->result_mask |= STATX_DIOALIGN; + if (!f2fs_force_buffered_io(inode, WRITE)) { + stat->dio_mem_align = bsize; + stat->dio_offset_align = bsize; + } + } + flags = fi->i_flags; if (flags & F2FS_COMPR_FL) stat->attributes |= STATX_ATTR_COMPRESSED; @@ -843,7 +891,7 @@ int f2fs_getattr(struct user_namespace *mnt_userns, const struct path *path, STATX_ATTR_NODUMP | STATX_ATTR_VERITY); - generic_fillattr(&init_user_ns, inode, stat); + generic_fillattr(mnt_userns, inode, stat); /* we need to show initial sectors used for inline_data/dentries */ if ((S_ISREG(inode->i_mode) && f2fs_has_inline_data(inode)) || @@ -859,10 +907,8 @@ static void __setattr_copy(struct user_namespace *mnt_userns, { unsigned int ia_valid = attr->ia_valid; - if (ia_valid & ATTR_UID) - inode->i_uid = attr->ia_uid; - if (ia_valid & ATTR_GID) - inode->i_gid = attr->ia_gid; + i_uid_update(mnt_userns, attr, inode); + i_gid_update(mnt_userns, attr, inode); if (ia_valid & ATTR_ATIME) inode->i_atime = attr->ia_atime; if (ia_valid & ATTR_MTIME) @@ -871,9 +917,10 @@ static void __setattr_copy(struct user_namespace *mnt_userns, inode->i_ctime = attr->ia_ctime; if (ia_valid & ATTR_MODE) { umode_t mode = attr->ia_mode; - kgid_t kgid = i_gid_into_mnt(mnt_userns, inode); + vfsgid_t vfsgid = i_gid_into_vfsgid(mnt_userns, inode); - if (!in_group_p(kgid) && !capable_wrt_inode_uidgid(mnt_userns, inode, CAP_FSETID)) + if (!vfsgid_in_group_p(vfsgid) && + !capable_wrt_inode_uidgid(mnt_userns, inode, CAP_FSETID)) mode &= ~S_ISGID; set_acl_inode(inode, mode); } @@ -903,7 +950,7 @@ int f2fs_setattr(struct user_namespace *mnt_userns, struct dentry *dentry, !f2fs_is_compress_backend_ready(inode)) return -EOPNOTSUPP; - err = setattr_prepare(&init_user_ns, dentry, attr); + err = setattr_prepare(mnt_userns, dentry, attr); if (err) return err; @@ -915,17 +962,15 @@ int f2fs_setattr(struct user_namespace *mnt_userns, struct dentry *dentry, if (err) return err; - if (is_quota_modification(inode, attr)) { + if (is_quota_modification(mnt_userns, inode, attr)) { err = f2fs_dquot_initialize(inode); if (err) return err; } - if ((attr->ia_valid & ATTR_UID && - !uid_eq(attr->ia_uid, inode->i_uid)) || - (attr->ia_valid & ATTR_GID && - !gid_eq(attr->ia_gid, inode->i_gid))) { + if (i_uid_needs_update(mnt_userns, attr, inode) || + i_gid_needs_update(mnt_userns, attr, inode)) { f2fs_lock_op(F2FS_I_SB(inode)); - err = dquot_transfer(inode, attr); + err = dquot_transfer(mnt_userns, inode, attr); if (err) { set_sbi_flag(F2FS_I_SB(inode), SBI_QUOTA_NEED_REPAIR); @@ -936,10 +981,8 @@ int f2fs_setattr(struct user_namespace *mnt_userns, struct dentry *dentry, * update uid/gid under lock_op(), so that dquot and inode can * be updated atomically. */ - if (attr->ia_valid & ATTR_UID) - inode->i_uid = attr->ia_uid; - if (attr->ia_valid & ATTR_GID) - inode->i_gid = attr->ia_gid; + i_uid_update(mnt_userns, attr, inode); + i_gid_update(mnt_userns, attr, inode); f2fs_mark_inode_dirty_sync(inode, true); f2fs_unlock_op(F2FS_I_SB(inode)); } @@ -957,7 +1000,7 @@ int f2fs_setattr(struct user_namespace *mnt_userns, struct dentry *dentry, return err; } - down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]); + f2fs_down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]); filemap_invalidate_lock(inode->i_mapping); truncate_setsize(inode, attr->ia_size); @@ -969,7 +1012,7 @@ int f2fs_setattr(struct user_namespace *mnt_userns, struct dentry *dentry, * larger than i_size. */ filemap_invalidate_unlock(inode->i_mapping); - up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]); + f2fs_up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]); if (err) return err; @@ -979,10 +1022,10 @@ int f2fs_setattr(struct user_namespace *mnt_userns, struct dentry *dentry, spin_unlock(&F2FS_I(inode)->i_size_lock); } - __setattr_copy(&init_user_ns, inode, attr); + __setattr_copy(mnt_userns, inode, attr); if (attr->ia_valid & ATTR_MODE) { - err = posix_acl_chmod(&init_user_ns, inode, f2fs_get_inode_mode(inode)); + err = posix_acl_chmod(mnt_userns, inode, f2fs_get_inode_mode(inode)); if (is_inode_flag_set(inode, FI_ACL_MODE)) { if (!err) @@ -1111,7 +1154,7 @@ static int punch_hole(struct inode *inode, loff_t offset, loff_t len) blk_start = (loff_t)pg_start << PAGE_SHIFT; blk_end = (loff_t)pg_end << PAGE_SHIFT; - down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]); + f2fs_down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]); filemap_invalidate_lock(inode->i_mapping); truncate_pagecache_range(inode, blk_start, blk_end - 1); @@ -1121,7 +1164,7 @@ static int punch_hole(struct inode *inode, loff_t offset, loff_t len) f2fs_unlock_op(sbi); filemap_invalidate_unlock(inode->i_mapping); - up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]); + f2fs_up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]); } } @@ -1159,6 +1202,7 @@ next_dnode: !f2fs_is_valid_blkaddr(sbi, *blkaddr, DATA_GENERIC_ENHANCE)) { f2fs_put_dnode(&dn); + f2fs_handle_error(sbi, ERROR_INVALID_BLKADDR); return -EFSCORRUPTED; } @@ -1232,7 +1276,7 @@ static int __clone_blkaddrs(struct inode *src_inode, struct inode *dst_inode, if (ret) return ret; - ret = f2fs_get_node_info(sbi, dn.nid, &ni); + ret = f2fs_get_node_info(sbi, dn.nid, &ni, false); if (ret) { f2fs_put_dnode(&dn); return ret; @@ -1276,7 +1320,7 @@ static int __clone_blkaddrs(struct inode *src_inode, struct inode *dst_inode, f2fs_put_page(psrc, 1); return PTR_ERR(pdst); } - f2fs_copy_page(psrc, pdst); + memcpy_page(pdst, 0, psrc, 0, PAGE_SIZE); set_page_dirty(pdst); f2fs_put_page(pdst, 1); f2fs_put_page(psrc, 1); @@ -1354,7 +1398,7 @@ static int f2fs_do_collapse(struct inode *inode, loff_t offset, loff_t len) f2fs_balance_fs(sbi, true); /* avoid gc operation during block exchange */ - down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]); + f2fs_down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]); filemap_invalidate_lock(inode->i_mapping); f2fs_lock_op(sbi); @@ -1364,7 +1408,7 @@ static int f2fs_do_collapse(struct inode *inode, loff_t offset, loff_t len) f2fs_unlock_op(sbi); filemap_invalidate_unlock(inode->i_mapping); - up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]); + f2fs_up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]); return ret; } @@ -1436,11 +1480,20 @@ static int f2fs_do_zero_range(struct dnode_of_data *dn, pgoff_t start, ret = -ENOSPC; break; } - if (dn->data_blkaddr != NEW_ADDR) { - f2fs_invalidate_blocks(sbi, dn->data_blkaddr); - dn->data_blkaddr = NEW_ADDR; - f2fs_set_data_blkaddr(dn); + + if (dn->data_blkaddr == NEW_ADDR) + continue; + + if (!f2fs_is_valid_blkaddr(sbi, dn->data_blkaddr, + DATA_GENERIC_ENHANCE)) { + ret = -EFSCORRUPTED; + f2fs_handle_error(sbi, ERROR_INVALID_BLKADDR); + break; } + + f2fs_invalidate_blocks(sbi, dn->data_blkaddr); + dn->data_blkaddr = NEW_ADDR; + f2fs_set_data_blkaddr(dn); } f2fs_update_extent_cache_range(dn, start, 0, index - start); @@ -1499,7 +1552,7 @@ static int f2fs_zero_range(struct inode *inode, loff_t offset, loff_t len, unsigned int end_offset; pgoff_t end; - down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]); + f2fs_down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]); filemap_invalidate_lock(mapping); truncate_pagecache_range(inode, @@ -1513,7 +1566,7 @@ static int f2fs_zero_range(struct inode *inode, loff_t offset, loff_t len, if (ret) { f2fs_unlock_op(sbi); filemap_invalidate_unlock(mapping); - up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]); + f2fs_up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]); goto out; } @@ -1525,7 +1578,7 @@ static int f2fs_zero_range(struct inode *inode, loff_t offset, loff_t len, f2fs_unlock_op(sbi); filemap_invalidate_unlock(mapping); - up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]); + f2fs_up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]); f2fs_balance_fs(sbi, dn.node_changed); @@ -1599,7 +1652,7 @@ static int f2fs_insert_range(struct inode *inode, loff_t offset, loff_t len) idx = DIV_ROUND_UP(i_size_read(inode), PAGE_SIZE); /* avoid gc operation during block exchange */ - down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]); + f2fs_down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]); filemap_invalidate_lock(mapping); truncate_pagecache(inode, offset); @@ -1617,7 +1670,7 @@ static int f2fs_insert_range(struct inode *inode, loff_t offset, loff_t len) f2fs_unlock_op(sbi); } filemap_invalidate_unlock(mapping); - up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]); + f2fs_up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]); /* write out all moved pages, if possible */ filemap_invalidate_lock(mapping); @@ -1637,6 +1690,11 @@ static int expand_inode_data(struct inode *inode, loff_t offset, struct f2fs_map_blocks map = { .m_next_pgofs = NULL, .m_next_extent = NULL, .m_seg_type = NO_CHECK_TYPE, .m_may_create = true }; + struct f2fs_gc_control gc_control = { .victim_segno = NULL_SEGNO, + .init_gc_type = FG_GC, + .should_migrate_blocks = false, + .err_gc_skipped = true, + .nr_free_secs = 0 }; pgoff_t pg_start, pg_end; loff_t new_size = i_size_read(inode); loff_t off_end; @@ -1666,20 +1724,20 @@ static int expand_inode_data(struct inode *inode, loff_t offset, return 0; if (f2fs_is_pinned_file(inode)) { - block_t sec_blks = BLKS_PER_SEC(sbi); + block_t sec_blks = CAP_BLKS_PER_SEC(sbi); block_t sec_len = roundup(map.m_len, sec_blks); map.m_len = sec_blks; next_alloc: if (has_not_enough_free_secs(sbi, 0, GET_SEC_FROM_SEG(sbi, overprovision_segments(sbi)))) { - down_write(&sbi->gc_lock); - err = f2fs_gc(sbi, true, false, false, NULL_SEGNO); - if (err && err != -ENODATA && err != -EAGAIN) + f2fs_down_write(&sbi->gc_lock); + err = f2fs_gc(sbi, &gc_control); + if (err && err != -ENODATA) goto out_err; } - down_write(&sbi->pin_sem); + f2fs_down_write(&sbi->pin_sem); f2fs_lock_op(sbi); f2fs_allocate_new_section(sbi, CURSEG_COLD_DATA_PINNED, false); @@ -1687,8 +1745,9 @@ next_alloc: map.m_seg_type = CURSEG_COLD_DATA_PINNED; err = f2fs_map_blocks(inode, &map, 1, F2FS_GET_BLOCK_PRE_DIO); + file_dont_truncate(inode); - up_write(&sbi->pin_sem); + f2fs_up_write(&sbi->pin_sem); expanded += map.m_len; sec_len -= map.m_len; @@ -1748,7 +1807,11 @@ static long f2fs_fallocate(struct file *file, int mode, (mode & (FALLOC_FL_COLLAPSE_RANGE | FALLOC_FL_INSERT_RANGE))) return -EOPNOTSUPP; - if (f2fs_compressed_file(inode) && + /* + * Pinned file should not support partial trucation since the block + * can be used by applications. + */ + if ((f2fs_compressed_file(inode) || f2fs_is_pinned_file(inode)) && (mode & (FALLOC_FL_PUNCH_HOLE | FALLOC_FL_COLLAPSE_RANGE | FALLOC_FL_ZERO_RANGE | FALLOC_FL_INSERT_RANGE))) return -EOPNOTSUPP; @@ -1760,6 +1823,10 @@ static long f2fs_fallocate(struct file *file, int mode, inode_lock(inode); + ret = file_modified(file); + if (ret) + goto out; + if (mode & FALLOC_FL_PUNCH_HOLE) { if (offset >= inode->i_size) goto out; @@ -1798,16 +1865,7 @@ static int f2fs_release_file(struct inode *inode, struct file *filp) atomic_read(&inode->i_writecount) != 1) return 0; - /* some remained atomic pages should discarded */ - if (f2fs_is_atomic_file(inode)) - f2fs_drop_inmem_pages(inode); - if (f2fs_is_volatile_file(inode)) { - set_inode_flag(inode, FI_DROP_CACHE); - filemap_fdatawrite(inode->i_mapping); - clear_inode_flag(inode, FI_DROP_CACHE); - clear_inode_flag(inode, FI_VOLATILE_FILE); - stat_dec_volatile_write(inode); - } + f2fs_abort_atomic_write(inode, true); return 0; } @@ -1821,9 +1879,8 @@ static int f2fs_file_flush(struct file *file, fl_owner_t id) * until all the writers close its file. Since this should be done * before dropping file lock, it needs to do in ->flush. */ - if (f2fs_is_atomic_file(inode) && - F2FS_I(inode)->inmem_task == current) - f2fs_drop_inmem_pages(inode); + if (F2FS_I(inode)->atomic_write_task == current) + f2fs_abort_atomic_write(inode, true); return 0; } @@ -1857,22 +1914,15 @@ static int f2fs_setflags_common(struct inode *inode, u32 iflags, u32 mask) if (masked_flags & F2FS_COMPR_FL) { if (!f2fs_disable_compressed_file(inode)) return -EINVAL; - } - if (iflags & F2FS_NOCOMP_FL) - return -EINVAL; - if (iflags & F2FS_COMPR_FL) { + } else { if (!f2fs_may_compress(inode)) return -EINVAL; - if (S_ISREG(inode->i_mode) && inode->i_size) + if (S_ISREG(inode->i_mode) && F2FS_HAS_BLOCKS(inode)) return -EINVAL; - - set_compress_context(inode); + if (set_compress_context(inode)) + return -EOPNOTSUPP; } } - if ((iflags ^ masked_flags) & F2FS_NOCOMP_FL) { - if (masked_flags & F2FS_COMPR_FL) - return -EINVAL; - } fi->i_flags = iflags | (fi->i_flags & ~mask); f2fs_bug_on(F2FS_I_SB(inode), (fi->i_flags & F2FS_COMPR_FL) && @@ -1983,11 +2033,13 @@ static int f2fs_ioc_getversion(struct file *filp, unsigned long arg) static int f2fs_ioc_start_atomic_write(struct file *filp) { struct inode *inode = file_inode(filp); + struct user_namespace *mnt_userns = file_mnt_user_ns(filp); struct f2fs_inode_info *fi = F2FS_I(inode); struct f2fs_sb_info *sbi = F2FS_I_SB(inode); + struct inode *pinode; int ret; - if (!inode_owner_or_capable(&init_user_ns, inode)) + if (!inode_owner_or_capable(mnt_userns, inode)) return -EACCES; if (!S_ISREG(inode->i_mode)) @@ -2002,47 +2054,60 @@ static int f2fs_ioc_start_atomic_write(struct file *filp) inode_lock(inode); - f2fs_disable_compressed_file(inode); - - if (f2fs_is_atomic_file(inode)) { - if (is_inode_flag_set(inode, FI_ATOMIC_REVOKE_REQUEST)) - ret = -EINVAL; + if (!f2fs_disable_compressed_file(inode)) { + ret = -EINVAL; goto out; } + if (f2fs_is_atomic_file(inode)) + goto out; + ret = f2fs_convert_inline_inode(inode); if (ret) goto out; - down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]); + f2fs_down_write(&fi->i_gc_rwsem[WRITE]); /* * Should wait end_io to count F2FS_WB_CP_DATA correctly by * f2fs_is_atomic_file. */ if (get_dirty_pages(inode)) - f2fs_warn(F2FS_I_SB(inode), "Unexpected flush for atomic writes: ino=%lu, npages=%u", + f2fs_warn(sbi, "Unexpected flush for atomic writes: ino=%lu, npages=%u", inode->i_ino, get_dirty_pages(inode)); ret = filemap_write_and_wait_range(inode->i_mapping, 0, LLONG_MAX); if (ret) { - up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]); + f2fs_up_write(&fi->i_gc_rwsem[WRITE]); + goto out; + } + + /* Create a COW inode for atomic write */ + pinode = f2fs_iget(inode->i_sb, fi->i_pino); + if (IS_ERR(pinode)) { + f2fs_up_write(&fi->i_gc_rwsem[WRITE]); + ret = PTR_ERR(pinode); + goto out; + } + + ret = f2fs_get_tmpfile(mnt_userns, pinode, &fi->cow_inode); + iput(pinode); + if (ret) { + f2fs_up_write(&fi->i_gc_rwsem[WRITE]); goto out; } + f2fs_i_size_write(fi->cow_inode, i_size_read(inode)); - spin_lock(&sbi->inode_lock[ATOMIC_FILE]); - if (list_empty(&fi->inmem_ilist)) - list_add_tail(&fi->inmem_ilist, &sbi->inode_list[ATOMIC_FILE]); - sbi->atomic_files++; - spin_unlock(&sbi->inode_lock[ATOMIC_FILE]); + stat_inc_atomic_inode(inode); - /* add inode in inmem_list first and set atomic_file */ set_inode_flag(inode, FI_ATOMIC_FILE); - clear_inode_flag(inode, FI_ATOMIC_REVOKE_REQUEST); - up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]); + set_inode_flag(fi->cow_inode, FI_COW_FILE); + clear_inode_flag(fi->cow_inode, FI_INLINE_DATA); + f2fs_up_write(&fi->i_gc_rwsem[WRITE]); - f2fs_update_time(F2FS_I_SB(inode), REQ_TIME); - F2FS_I(inode)->inmem_task = current; + f2fs_update_time(sbi, REQ_TIME); + fi->atomic_write_task = current; stat_update_max_atomic_write(inode); + fi->atomic_write_cnt = 0; out: inode_unlock(inode); mnt_drop_write_file(filp); @@ -2052,9 +2117,10 @@ out: static int f2fs_ioc_commit_atomic_write(struct file *filp) { struct inode *inode = file_inode(filp); + struct user_namespace *mnt_userns = file_mnt_user_ns(filp); int ret; - if (!inode_owner_or_capable(&init_user_ns, inode)) + if (!inode_owner_or_capable(mnt_userns, inode)) return -EACCES; ret = mnt_want_write_file(filp); @@ -2065,119 +2131,39 @@ static int f2fs_ioc_commit_atomic_write(struct file *filp) inode_lock(inode); - if (f2fs_is_volatile_file(inode)) { - ret = -EINVAL; - goto err_out; - } - if (f2fs_is_atomic_file(inode)) { - ret = f2fs_commit_inmem_pages(inode); + ret = f2fs_commit_atomic_write(inode); if (ret) - goto err_out; + goto unlock_out; ret = f2fs_do_sync_file(filp, 0, LLONG_MAX, 0, true); if (!ret) - f2fs_drop_inmem_pages(inode); + f2fs_abort_atomic_write(inode, false); } else { ret = f2fs_do_sync_file(filp, 0, LLONG_MAX, 1, false); } -err_out: - if (is_inode_flag_set(inode, FI_ATOMIC_REVOKE_REQUEST)) { - clear_inode_flag(inode, FI_ATOMIC_REVOKE_REQUEST); - ret = -EINVAL; - } +unlock_out: inode_unlock(inode); mnt_drop_write_file(filp); return ret; } -static int f2fs_ioc_start_volatile_write(struct file *filp) +static int f2fs_ioc_abort_atomic_write(struct file *filp) { struct inode *inode = file_inode(filp); + struct user_namespace *mnt_userns = file_mnt_user_ns(filp); int ret; - if (!inode_owner_or_capable(&init_user_ns, inode)) + if (!inode_owner_or_capable(mnt_userns, inode)) return -EACCES; - if (!S_ISREG(inode->i_mode)) - return -EINVAL; - ret = mnt_want_write_file(filp); if (ret) return ret; inode_lock(inode); - if (f2fs_is_volatile_file(inode)) - goto out; - - ret = f2fs_convert_inline_inode(inode); - if (ret) - goto out; - - stat_inc_volatile_write(inode); - stat_update_max_volatile_write(inode); - - set_inode_flag(inode, FI_VOLATILE_FILE); - f2fs_update_time(F2FS_I_SB(inode), REQ_TIME); -out: - inode_unlock(inode); - mnt_drop_write_file(filp); - return ret; -} - -static int f2fs_ioc_release_volatile_write(struct file *filp) -{ - struct inode *inode = file_inode(filp); - int ret; - - if (!inode_owner_or_capable(&init_user_ns, inode)) - return -EACCES; - - ret = mnt_want_write_file(filp); - if (ret) - return ret; - - inode_lock(inode); - - if (!f2fs_is_volatile_file(inode)) - goto out; - - if (!f2fs_is_first_block_written(inode)) { - ret = truncate_partial_data_page(inode, 0, true); - goto out; - } - - ret = punch_hole(inode, 0, F2FS_BLKSIZE); -out: - inode_unlock(inode); - mnt_drop_write_file(filp); - return ret; -} - -static int f2fs_ioc_abort_volatile_write(struct file *filp) -{ - struct inode *inode = file_inode(filp); - int ret; - - if (!inode_owner_or_capable(&init_user_ns, inode)) - return -EACCES; - - ret = mnt_want_write_file(filp); - if (ret) - return ret; - - inode_lock(inode); - - if (f2fs_is_atomic_file(inode)) - f2fs_drop_inmem_pages(inode); - if (f2fs_is_volatile_file(inode)) { - clear_inode_flag(inode, FI_VOLATILE_FILE); - stat_dec_volatile_write(inode); - ret = f2fs_do_sync_file(filp, 0, LLONG_MAX, 0, true); - } - - clear_inode_flag(inode, FI_ATOMIC_REVOKE_REQUEST); + f2fs_abort_atomic_write(inode, true); inode_unlock(inode); @@ -2205,7 +2191,8 @@ static int f2fs_ioc_shutdown(struct file *filp, unsigned long arg) if (ret) { if (ret == -EROFS) { ret = 0; - f2fs_stop_checkpoint(sbi, false); + f2fs_stop_checkpoint(sbi, false, + STOP_CP_REASON_SHUTDOWN); set_sbi_flag(sbi, SBI_IS_SHUTDOWN); trace_f2fs_shutdown(sbi, in, ret); } @@ -2218,7 +2205,7 @@ static int f2fs_ioc_shutdown(struct file *filp, unsigned long arg) ret = freeze_bdev(sb->s_bdev); if (ret) goto out; - f2fs_stop_checkpoint(sbi, false); + f2fs_stop_checkpoint(sbi, false, STOP_CP_REASON_SHUTDOWN); set_sbi_flag(sbi, SBI_IS_SHUTDOWN); thaw_bdev(sb->s_bdev); break; @@ -2227,16 +2214,16 @@ static int f2fs_ioc_shutdown(struct file *filp, unsigned long arg) ret = f2fs_sync_fs(sb, 1); if (ret) goto out; - f2fs_stop_checkpoint(sbi, false); + f2fs_stop_checkpoint(sbi, false, STOP_CP_REASON_SHUTDOWN); set_sbi_flag(sbi, SBI_IS_SHUTDOWN); break; case F2FS_GOING_DOWN_NOSYNC: - f2fs_stop_checkpoint(sbi, false); + f2fs_stop_checkpoint(sbi, false, STOP_CP_REASON_SHUTDOWN); set_sbi_flag(sbi, SBI_IS_SHUTDOWN); break; case F2FS_GOING_DOWN_METAFLUSH: f2fs_sync_meta_pages(sbi, META, LONG_MAX, FS_META_IO); - f2fs_stop_checkpoint(sbi, false); + f2fs_stop_checkpoint(sbi, false, STOP_CP_REASON_SHUTDOWN); set_sbi_flag(sbi, SBI_IS_SHUTDOWN); break; case F2FS_GOING_DOWN_NEED_FSCK: @@ -2271,7 +2258,6 @@ static int f2fs_ioc_fitrim(struct file *filp, unsigned long arg) { struct inode *inode = file_inode(filp); struct super_block *sb = inode->i_sb; - struct request_queue *q = bdev_get_queue(sb->s_bdev); struct fstrim_range range; int ret; @@ -2290,7 +2276,7 @@ static int f2fs_ioc_fitrim(struct file *filp, unsigned long arg) return ret; range.minlen = max((unsigned int)range.minlen, - q->limits.discard_granularity); + bdev_discard_granularity(sb->s_bdev)); ret = f2fs_trim_fs(F2FS_SB(sb), &range); mnt_drop_write_file(filp); if (ret < 0) @@ -2345,7 +2331,7 @@ static int f2fs_ioc_get_encryption_pwsalt(struct file *filp, unsigned long arg) if (err) return err; - down_write(&sbi->sb_lock); + f2fs_down_write(&sbi->sb_lock); if (uuid_is_nonzero(sbi->raw_super->encrypt_pw_salt)) goto got_it; @@ -2364,7 +2350,7 @@ got_it: 16)) err = -EFAULT; out_err: - up_write(&sbi->sb_lock); + f2fs_up_write(&sbi->sb_lock); mnt_drop_write_file(filp); return err; } @@ -2424,6 +2410,10 @@ static int f2fs_ioc_gc(struct file *filp, unsigned long arg) { struct inode *inode = file_inode(filp); struct f2fs_sb_info *sbi = F2FS_I_SB(inode); + struct f2fs_gc_control gc_control = { .victim_segno = NULL_SEGNO, + .no_bg_gc = false, + .should_migrate_blocks = false, + .nr_free_secs = 0 }; __u32 sync; int ret; @@ -2441,15 +2431,17 @@ static int f2fs_ioc_gc(struct file *filp, unsigned long arg) return ret; if (!sync) { - if (!down_write_trylock(&sbi->gc_lock)) { + if (!f2fs_down_write_trylock(&sbi->gc_lock)) { ret = -EBUSY; goto out; } } else { - down_write(&sbi->gc_lock); + f2fs_down_write(&sbi->gc_lock); } - ret = f2fs_gc(sbi, sync, true, false, NULL_SEGNO); + gc_control.init_gc_type = sync ? FG_GC : BG_GC; + gc_control.err_gc_skipped = sync; + ret = f2fs_gc(sbi, &gc_control); out: mnt_drop_write_file(filp); return ret; @@ -2458,6 +2450,12 @@ out: static int __f2fs_ioc_gc_range(struct file *filp, struct f2fs_gc_range *range) { struct f2fs_sb_info *sbi = F2FS_I_SB(file_inode(filp)); + struct f2fs_gc_control gc_control = { + .init_gc_type = range->sync ? FG_GC : BG_GC, + .no_bg_gc = false, + .should_migrate_blocks = false, + .err_gc_skipped = range->sync, + .nr_free_secs = 0 }; u64 end; int ret; @@ -2477,22 +2475,22 @@ static int __f2fs_ioc_gc_range(struct file *filp, struct f2fs_gc_range *range) do_more: if (!range->sync) { - if (!down_write_trylock(&sbi->gc_lock)) { + if (!f2fs_down_write_trylock(&sbi->gc_lock)) { ret = -EBUSY; goto out; } } else { - down_write(&sbi->gc_lock); + f2fs_down_write(&sbi->gc_lock); } - ret = f2fs_gc(sbi, range->sync, true, false, - GET_SEGNO(sbi, range->start)); + gc_control.victim_segno = GET_SEGNO(sbi, range->start); + ret = f2fs_gc(sbi, &gc_control); if (ret) { if (ret == -EBUSY) ret = -EAGAIN; goto out; } - range->start += BLKS_PER_SEC(sbi); + range->start += CAP_BLKS_PER_SEC(sbi); if (range->start <= end) goto do_more; out: @@ -2553,10 +2551,6 @@ static int f2fs_defragment_range(struct f2fs_sb_info *sbi, bool fragmented = false; int err; - /* if in-place-update policy is enabled, don't waste time here */ - if (f2fs_should_update_inplace(inode, NULL)) - return -EINVAL; - pg_start = range->start >> PAGE_SHIFT; pg_end = (range->start + range->len) >> PAGE_SHIFT; @@ -2564,6 +2558,13 @@ static int f2fs_defragment_range(struct f2fs_sb_info *sbi, inode_lock(inode); + /* if in-place-update policy is enabled, don't waste time here */ + set_inode_flag(inode, FI_OPU_WRITE); + if (f2fs_should_update_inplace(inode, NULL)) { + err = -EINVAL; + goto out; + } + /* writeback all dirty pages in the range */ err = filemap_write_and_wait_range(inode->i_mapping, range->start, range->start + range->len - 1); @@ -2614,7 +2615,7 @@ static int f2fs_defragment_range(struct f2fs_sb_info *sbi, goto out; } - sec_num = DIV_ROUND_UP(total, BLKS_PER_SEC(sbi)); + sec_num = DIV_ROUND_UP(total, CAP_BLKS_PER_SEC(sbi)); /* * make sure there are enough free section for LFS allocation, this can @@ -2645,7 +2646,7 @@ do_map: goto check; } - set_inode_flag(inode, FI_DO_DEFRAG); + set_inode_flag(inode, FI_SKIP_WRITES); idx = map.m_lblk; while (idx < map.m_lblk + map.m_len && cnt < blk_per_seg) { @@ -2658,6 +2659,7 @@ do_map: } set_page_dirty(page); + set_page_private_gcing(page); f2fs_put_page(page, 1); idx++; @@ -2670,15 +2672,16 @@ check: if (map.m_lblk < pg_end && cnt < blk_per_seg) goto do_map; - clear_inode_flag(inode, FI_DO_DEFRAG); + clear_inode_flag(inode, FI_SKIP_WRITES); err = filemap_fdatawrite(inode->i_mapping); if (err) goto out; } clear_out: - clear_inode_flag(inode, FI_DO_DEFRAG); + clear_inode_flag(inode, FI_SKIP_WRITES); out: + clear_inode_flag(inode, FI_OPU_WRITE); inode_unlock(inode); if (!err) range->len = (u64)total << PAGE_SHIFT; @@ -2814,10 +2817,10 @@ static int f2fs_move_file_range(struct file *file_in, loff_t pos_in, f2fs_balance_fs(sbi, true); - down_write(&F2FS_I(src)->i_gc_rwsem[WRITE]); + f2fs_down_write(&F2FS_I(src)->i_gc_rwsem[WRITE]); if (src != dst) { ret = -EBUSY; - if (!down_write_trylock(&F2FS_I(dst)->i_gc_rwsem[WRITE])) + if (!f2fs_down_write_trylock(&F2FS_I(dst)->i_gc_rwsem[WRITE])) goto out_src; } @@ -2835,9 +2838,9 @@ static int f2fs_move_file_range(struct file *file_in, loff_t pos_in, f2fs_unlock_op(sbi); if (src != dst) - up_write(&F2FS_I(dst)->i_gc_rwsem[WRITE]); + f2fs_up_write(&F2FS_I(dst)->i_gc_rwsem[WRITE]); out_src: - up_write(&F2FS_I(src)->i_gc_rwsem[WRITE]); + f2fs_up_write(&F2FS_I(src)->i_gc_rwsem[WRITE]); out_unlock: if (src != dst) inode_unlock(dst); @@ -2896,6 +2899,11 @@ static int f2fs_ioc_flush_device(struct file *filp, unsigned long arg) unsigned int start_segno = 0, end_segno = 0; unsigned int dev_start_segno = 0, dev_end_segno = 0; struct f2fs_flush_device range; + struct f2fs_gc_control gc_control = { + .init_gc_type = FG_GC, + .should_migrate_blocks = true, + .err_gc_skipped = true, + .nr_free_secs = 0 }; int ret; if (!capable(CAP_SYS_ADMIN)) @@ -2932,14 +2940,16 @@ static int f2fs_ioc_flush_device(struct file *filp, unsigned long arg) end_segno = min(start_segno + range.segments, dev_end_segno); while (start_segno < end_segno) { - if (!down_write_trylock(&sbi->gc_lock)) { + if (!f2fs_down_write_trylock(&sbi->gc_lock)) { ret = -EBUSY; goto out; } sm->last_victim[GC_CB] = end_segno + 1; sm->last_victim[GC_GREEDY] = end_segno + 1; sm->last_victim[ALLOC_NEXT] = end_segno + 1; - ret = f2fs_gc(sbi, true, true, true, start_segno); + + gc_control.victim_segno = start_segno; + ret = f2fs_gc(sbi, &gc_control); if (ret == -EAGAIN) ret = 0; else if (ret < 0) @@ -2984,7 +2994,7 @@ static int f2fs_ioc_setproject(struct inode *inode, __u32 projid) { struct f2fs_inode_info *fi = F2FS_I(inode); struct f2fs_sb_info *sbi = F2FS_I_SB(inode); - struct page *ipage; + struct f2fs_inode *ri = NULL; kprojid_t kprojid; int err; @@ -3000,7 +3010,7 @@ static int f2fs_ioc_setproject(struct inode *inode, __u32 projid) kprojid = make_kprojid(&init_user_ns, (projid_t)projid); - if (projid_eq(kprojid, F2FS_I(inode)->i_projid)) + if (projid_eq(kprojid, fi->i_projid)) return 0; err = -EPERM; @@ -3008,17 +3018,8 @@ static int f2fs_ioc_setproject(struct inode *inode, __u32 projid) if (IS_NOQUOTA(inode)) return err; - ipage = f2fs_get_node_page(sbi, inode->i_ino); - if (IS_ERR(ipage)) - return PTR_ERR(ipage); - - if (!F2FS_FITS_IN_INODE(F2FS_INODE(ipage), fi->i_extra_isize, - i_projid)) { - err = -EOVERFLOW; - f2fs_put_page(ipage, 1); - return err; - } - f2fs_put_page(ipage, 1); + if (!F2FS_FITS_IN_INODE(ri, fi->i_extra_isize, i_projid)) + return -EOVERFLOW; err = f2fs_dquot_initialize(inode); if (err) @@ -3029,7 +3030,7 @@ static int f2fs_ioc_setproject(struct inode *inode, __u32 projid) if (err) goto out_unlock; - F2FS_I(inode)->i_projid = kprojid; + fi->i_projid = kprojid; inode->i_ctime = current_time(inode); f2fs_mark_inode_dirty_sync(inode, true); out_unlock: @@ -3143,17 +3144,17 @@ static int f2fs_ioc_set_pin_file(struct file *filp, unsigned long arg) inode_lock(inode); - if (f2fs_should_update_outplace(inode, NULL)) { - ret = -EINVAL; - goto out; - } - if (!pin) { clear_inode_flag(inode, FI_PIN_FILE); f2fs_i_gc_failures_write(inode, 0); goto done; } + if (f2fs_should_update_outplace(inode, NULL)) { + ret = -EINVAL; + goto out; + } + if (f2fs_pin_file_control(inode, false)) { ret = -EAGAIN; goto out; @@ -3209,9 +3210,9 @@ int f2fs_precache_extents(struct inode *inode) while (map.m_lblk < end) { map.m_len = end - map.m_lblk; - down_write(&fi->i_gc_rwsem[WRITE]); + f2fs_down_write(&fi->i_gc_rwsem[WRITE]); err = f2fs_map_blocks(inode, &map, 0, F2FS_GET_BLOCK_PRECACHE); - up_write(&fi->i_gc_rwsem[WRITE]); + f2fs_up_write(&fi->i_gc_rwsem[WRITE]); if (err) return err; @@ -3288,11 +3289,11 @@ static int f2fs_ioc_getfslabel(struct file *filp, unsigned long arg) if (!vbuf) return -ENOMEM; - down_read(&sbi->sb_lock); + f2fs_down_read(&sbi->sb_lock); count = utf16s_to_utf8s(sbi->raw_super->volume_name, ARRAY_SIZE(sbi->raw_super->volume_name), UTF16_LITTLE_ENDIAN, vbuf, MAX_VOLUME_NAME); - up_read(&sbi->sb_lock); + f2fs_up_read(&sbi->sb_lock); if (copy_to_user((char __user *)arg, vbuf, min(FSLABEL_MAX, count))) @@ -3320,7 +3321,7 @@ static int f2fs_ioc_setfslabel(struct file *filp, unsigned long arg) if (err) goto out; - down_write(&sbi->sb_lock); + f2fs_down_write(&sbi->sb_lock); memset(sbi->raw_super->volume_name, 0, sizeof(sbi->raw_super->volume_name)); @@ -3330,7 +3331,7 @@ static int f2fs_ioc_setfslabel(struct file *filp, unsigned long arg) err = f2fs_commit_super(sbi, false); - up_write(&sbi->sb_lock); + f2fs_up_write(&sbi->sb_lock); mnt_drop_write_file(filp); out: @@ -3368,8 +3369,10 @@ static int release_compress_blocks(struct dnode_of_data *dn, pgoff_t count) if (!__is_valid_data_blkaddr(blkaddr)) continue; if (unlikely(!f2fs_is_valid_blkaddr(sbi, blkaddr, - DATA_GENERIC_ENHANCE))) + DATA_GENERIC_ENHANCE))) { + f2fs_handle_error(sbi, ERROR_INVALID_BLKADDR); return -EFSCORRUPTED; + } } while (count) { @@ -3456,7 +3459,7 @@ static int f2fs_release_compress_blocks(struct file *filp, unsigned long arg) if (!atomic_read(&F2FS_I(inode)->i_compr_blocks)) goto out; - down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]); + f2fs_down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]); filemap_invalidate_lock(inode->i_mapping); last_idx = DIV_ROUND_UP(i_size_read(inode), PAGE_SIZE); @@ -3493,7 +3496,7 @@ static int f2fs_release_compress_blocks(struct file *filp, unsigned long arg) } filemap_invalidate_unlock(inode->i_mapping); - up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]); + f2fs_up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]); out: inode_unlock(inode); @@ -3530,8 +3533,10 @@ static int reserve_compress_blocks(struct dnode_of_data *dn, pgoff_t count) if (!__is_valid_data_blkaddr(blkaddr)) continue; if (unlikely(!f2fs_is_valid_blkaddr(sbi, blkaddr, - DATA_GENERIC_ENHANCE))) + DATA_GENERIC_ENHANCE))) { + f2fs_handle_error(sbi, ERROR_INVALID_BLKADDR); return -EFSCORRUPTED; + } } while (count) { @@ -3609,7 +3614,7 @@ static int f2fs_reserve_compress_blocks(struct file *filp, unsigned long arg) goto unlock_inode; } - down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]); + f2fs_down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]); filemap_invalidate_lock(inode->i_mapping); last_idx = DIV_ROUND_UP(i_size_read(inode), PAGE_SIZE); @@ -3646,7 +3651,7 @@ static int f2fs_reserve_compress_blocks(struct file *filp, unsigned long arg) } filemap_invalidate_unlock(inode->i_mapping); - up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]); + f2fs_up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]); if (ret >= 0) { clear_inode_flag(inode, FI_COMPRESS_RELEASED); @@ -3677,18 +3682,18 @@ out: static int f2fs_secure_erase(struct block_device *bdev, struct inode *inode, pgoff_t off, block_t block, block_t len, u32 flags) { - struct request_queue *q = bdev_get_queue(bdev); sector_t sector = SECTOR_FROM_BLOCK(block); sector_t nr_sects = SECTOR_FROM_BLOCK(len); int ret = 0; - if (!q) - return -ENXIO; - - if (flags & F2FS_TRIM_FILE_DISCARD) - ret = blkdev_issue_discard(bdev, sector, nr_sects, GFP_NOFS, - blk_queue_secure_erase(q) ? - BLKDEV_DISCARD_SECURE : 0); + if (flags & F2FS_TRIM_FILE_DISCARD) { + if (bdev_max_secure_erase_sectors(bdev)) + ret = blkdev_issue_secure_erase(bdev, sector, nr_sects, + GFP_NOFS); + else + ret = blkdev_issue_discard(bdev, sector, nr_sects, + GFP_NOFS); + } if (!ret && (flags & F2FS_TRIM_FILE_ZEROOUT)) { if (IS_ENCRYPTED(inode)) @@ -3764,7 +3769,7 @@ static int f2fs_sec_trim_file(struct file *filp, unsigned long arg) if (ret) goto err; - down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]); + f2fs_down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]); filemap_invalidate_lock(mapping); ret = filemap_write_and_wait_range(mapping, range.start, @@ -3803,6 +3808,8 @@ static int f2fs_sec_trim_file(struct file *filp, unsigned long arg) DATA_GENERIC_ENHANCE)) { ret = -EFSCORRUPTED; f2fs_put_dnode(&dn); + f2fs_handle_error(sbi, + ERROR_INVALID_BLKADDR); goto out; } @@ -3853,7 +3860,7 @@ static int f2fs_sec_trim_file(struct file *filp, unsigned long arg) prev_block, len, range.flags); out: filemap_invalidate_unlock(mapping); - up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]); + f2fs_up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]); err: inode_unlock(inode); file_end_write(filp); @@ -3960,10 +3967,10 @@ static int redirty_blocks(struct inode *inode, pgoff_t page_idx, int len) for (i = 0; i < page_len; i++, redirty_idx++) { page = find_lock_page(mapping, redirty_idx); - if (!page) { - ret = -ENOMEM; - break; - } + + /* It will never fail, when page has pinned above */ + f2fs_bug_on(F2FS_I_SB(inode), !page); + set_page_dirty(page); f2fs_put_page(page, 1); f2fs_put_page(page, 0); @@ -3979,7 +3986,7 @@ static int f2fs_ioc_decompress_file(struct file *filp, unsigned long arg) struct f2fs_inode_info *fi = F2FS_I(inode); pgoff_t page_idx = 0, last_idx; unsigned int blk_per_seg = sbi->blocks_per_seg; - int cluster_size = F2FS_I(inode)->i_cluster_size; + int cluster_size = fi->i_cluster_size; int count, ret; if (!f2fs_sb_has_compression(sbi) || @@ -4002,8 +4009,8 @@ static int f2fs_ioc_decompress_file(struct file *filp, unsigned long arg) goto out; } - if (f2fs_is_mmap_file(inode)) { - ret = -EBUSY; + if (is_inode_flag_set(inode, FI_COMPRESS_RELEASED)) { + ret = -EINVAL; goto out; } @@ -4074,8 +4081,8 @@ static int f2fs_ioc_compress_file(struct file *filp, unsigned long arg) goto out; } - if (f2fs_is_mmap_file(inode)) { - ret = -EBUSY; + if (is_inode_flag_set(inode, FI_COMPRESS_RELEASED)) { + ret = -EINVAL; goto out; } @@ -4127,12 +4134,11 @@ static long __f2fs_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) return f2fs_ioc_start_atomic_write(filp); case F2FS_IOC_COMMIT_ATOMIC_WRITE: return f2fs_ioc_commit_atomic_write(filp); + case F2FS_IOC_ABORT_ATOMIC_WRITE: + return f2fs_ioc_abort_atomic_write(filp); case F2FS_IOC_START_VOLATILE_WRITE: - return f2fs_ioc_start_volatile_write(filp); case F2FS_IOC_RELEASE_VOLATILE_WRITE: - return f2fs_ioc_release_volatile_write(filp); - case F2FS_IOC_ABORT_VOLATILE_WRITE: - return f2fs_ioc_abort_volatile_write(filp); + return -EOPNOTSUPP; case F2FS_IOC_SHUTDOWN: return f2fs_ioc_shutdown(filp, arg); case FITRIM: @@ -4218,27 +4224,403 @@ long f2fs_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) return __f2fs_ioctl(filp, cmd, arg); } -static ssize_t f2fs_file_read_iter(struct kiocb *iocb, struct iov_iter *iter) +/* + * Return %true if the given read or write request should use direct I/O, or + * %false if it should use buffered I/O. + */ +static bool f2fs_should_use_dio(struct inode *inode, struct kiocb *iocb, + struct iov_iter *iter) +{ + unsigned int align; + + if (!(iocb->ki_flags & IOCB_DIRECT)) + return false; + + if (f2fs_force_buffered_io(inode, iov_iter_rw(iter))) + return false; + + /* + * Direct I/O not aligned to the disk's logical_block_size will be + * attempted, but will fail with -EINVAL. + * + * f2fs additionally requires that direct I/O be aligned to the + * filesystem block size, which is often a stricter requirement. + * However, f2fs traditionally falls back to buffered I/O on requests + * that are logical_block_size-aligned but not fs-block aligned. + * + * The below logic implements this behavior. + */ + align = iocb->ki_pos | iov_iter_alignment(iter); + if (!IS_ALIGNED(align, i_blocksize(inode)) && + IS_ALIGNED(align, bdev_logical_block_size(inode->i_sb->s_bdev))) + return false; + + return true; +} + +static int f2fs_dio_read_end_io(struct kiocb *iocb, ssize_t size, int error, + unsigned int flags) +{ + struct f2fs_sb_info *sbi = F2FS_I_SB(file_inode(iocb->ki_filp)); + + dec_page_count(sbi, F2FS_DIO_READ); + if (error) + return error; + f2fs_update_iostat(sbi, NULL, APP_DIRECT_READ_IO, size); + return 0; +} + +static const struct iomap_dio_ops f2fs_iomap_dio_read_ops = { + .end_io = f2fs_dio_read_end_io, +}; + +static ssize_t f2fs_dio_read_iter(struct kiocb *iocb, struct iov_iter *to) { struct file *file = iocb->ki_filp; struct inode *inode = file_inode(file); - int ret; + struct f2fs_sb_info *sbi = F2FS_I_SB(inode); + struct f2fs_inode_info *fi = F2FS_I(inode); + const loff_t pos = iocb->ki_pos; + const size_t count = iov_iter_count(to); + struct iomap_dio *dio; + ssize_t ret; + + if (count == 0) + return 0; /* skip atime update */ + + trace_f2fs_direct_IO_enter(inode, iocb, count, READ); + + if (iocb->ki_flags & IOCB_NOWAIT) { + if (!f2fs_down_read_trylock(&fi->i_gc_rwsem[READ])) { + ret = -EAGAIN; + goto out; + } + } else { + f2fs_down_read(&fi->i_gc_rwsem[READ]); + } + + /* + * We have to use __iomap_dio_rw() and iomap_dio_complete() instead of + * the higher-level function iomap_dio_rw() in order to ensure that the + * F2FS_DIO_READ counter will be decremented correctly in all cases. + */ + inc_page_count(sbi, F2FS_DIO_READ); + dio = __iomap_dio_rw(iocb, to, &f2fs_iomap_ops, + &f2fs_iomap_dio_read_ops, 0, NULL, 0); + if (IS_ERR_OR_NULL(dio)) { + ret = PTR_ERR_OR_ZERO(dio); + if (ret != -EIOCBQUEUED) + dec_page_count(sbi, F2FS_DIO_READ); + } else { + ret = iomap_dio_complete(dio); + } + + f2fs_up_read(&fi->i_gc_rwsem[READ]); + + file_accessed(file); +out: + trace_f2fs_direct_IO_exit(inode, pos, count, READ, ret); + return ret; +} + +static ssize_t f2fs_file_read_iter(struct kiocb *iocb, struct iov_iter *to) +{ + struct inode *inode = file_inode(iocb->ki_filp); + const loff_t pos = iocb->ki_pos; + ssize_t ret; if (!f2fs_is_compress_backend_ready(inode)) return -EOPNOTSUPP; - ret = generic_file_read_iter(iocb, iter); + if (trace_f2fs_dataread_start_enabled()) { + char *p = f2fs_kmalloc(F2FS_I_SB(inode), PATH_MAX, GFP_KERNEL); + char *path; + + if (!p) + goto skip_read_trace; - if (ret > 0) - f2fs_update_iostat(F2FS_I_SB(inode), APP_READ_IO, ret); + path = dentry_path_raw(file_dentry(iocb->ki_filp), p, PATH_MAX); + if (IS_ERR(path)) { + kfree(p); + goto skip_read_trace; + } + trace_f2fs_dataread_start(inode, pos, iov_iter_count(to), + current->pid, path, current->comm); + kfree(p); + } +skip_read_trace: + if (f2fs_should_use_dio(inode, iocb, to)) { + ret = f2fs_dio_read_iter(iocb, to); + } else { + ret = filemap_read(iocb, to, 0); + if (ret > 0) + f2fs_update_iostat(F2FS_I_SB(inode), inode, + APP_BUFFERED_READ_IO, ret); + } + if (trace_f2fs_dataread_end_enabled()) + trace_f2fs_dataread_end(inode, pos, ret); return ret; } -static ssize_t f2fs_file_write_iter(struct kiocb *iocb, struct iov_iter *from) +static ssize_t f2fs_write_checks(struct kiocb *iocb, struct iov_iter *from) +{ + struct file *file = iocb->ki_filp; + struct inode *inode = file_inode(file); + ssize_t count; + int err; + + if (IS_IMMUTABLE(inode)) + return -EPERM; + + if (is_inode_flag_set(inode, FI_COMPRESS_RELEASED)) + return -EPERM; + + count = generic_write_checks(iocb, from); + if (count <= 0) + return count; + + err = file_modified(file); + if (err) + return err; + return count; +} + +/* + * Preallocate blocks for a write request, if it is possible and helpful to do + * so. Returns a positive number if blocks may have been preallocated, 0 if no + * blocks were preallocated, or a negative errno value if something went + * seriously wrong. Also sets FI_PREALLOCATED_ALL on the inode if *all* the + * requested blocks (not just some of them) have been allocated. + */ +static int f2fs_preallocate_blocks(struct kiocb *iocb, struct iov_iter *iter, + bool dio) +{ + struct inode *inode = file_inode(iocb->ki_filp); + struct f2fs_sb_info *sbi = F2FS_I_SB(inode); + const loff_t pos = iocb->ki_pos; + const size_t count = iov_iter_count(iter); + struct f2fs_map_blocks map = {}; + int flag; + int ret; + + /* If it will be an out-of-place direct write, don't bother. */ + if (dio && f2fs_lfs_mode(sbi)) + return 0; + /* + * Don't preallocate holes aligned to DIO_SKIP_HOLES which turns into + * buffered IO, if DIO meets any holes. + */ + if (dio && i_size_read(inode) && + (F2FS_BYTES_TO_BLK(pos) < F2FS_BLK_ALIGN(i_size_read(inode)))) + return 0; + + /* No-wait I/O can't allocate blocks. */ + if (iocb->ki_flags & IOCB_NOWAIT) + return 0; + + /* If it will be a short write, don't bother. */ + if (fault_in_iov_iter_readable(iter, count)) + return 0; + + if (f2fs_has_inline_data(inode)) { + /* If the data will fit inline, don't bother. */ + if (pos + count <= MAX_INLINE_DATA(inode)) + return 0; + ret = f2fs_convert_inline_inode(inode); + if (ret) + return ret; + } + + /* Do not preallocate blocks that will be written partially in 4KB. */ + map.m_lblk = F2FS_BLK_ALIGN(pos); + map.m_len = F2FS_BYTES_TO_BLK(pos + count); + if (map.m_len > map.m_lblk) + map.m_len -= map.m_lblk; + else + map.m_len = 0; + map.m_may_create = true; + if (dio) { + map.m_seg_type = f2fs_rw_hint_to_seg_type(inode->i_write_hint); + flag = F2FS_GET_BLOCK_PRE_DIO; + } else { + map.m_seg_type = NO_CHECK_TYPE; + flag = F2FS_GET_BLOCK_PRE_AIO; + } + + ret = f2fs_map_blocks(inode, &map, 1, flag); + /* -ENOSPC|-EDQUOT are fine to report the number of allocated blocks. */ + if (ret < 0 && !((ret == -ENOSPC || ret == -EDQUOT) && map.m_len > 0)) + return ret; + if (ret == 0) + set_inode_flag(inode, FI_PREALLOCATED_ALL); + return map.m_len; +} + +static ssize_t f2fs_buffered_write_iter(struct kiocb *iocb, + struct iov_iter *from) +{ + struct file *file = iocb->ki_filp; + struct inode *inode = file_inode(file); + ssize_t ret; + + if (iocb->ki_flags & IOCB_NOWAIT) + return -EOPNOTSUPP; + + current->backing_dev_info = inode_to_bdi(inode); + ret = generic_perform_write(iocb, from); + current->backing_dev_info = NULL; + + if (ret > 0) { + iocb->ki_pos += ret; + f2fs_update_iostat(F2FS_I_SB(inode), inode, + APP_BUFFERED_IO, ret); + } + return ret; +} + +static int f2fs_dio_write_end_io(struct kiocb *iocb, ssize_t size, int error, + unsigned int flags) +{ + struct f2fs_sb_info *sbi = F2FS_I_SB(file_inode(iocb->ki_filp)); + + dec_page_count(sbi, F2FS_DIO_WRITE); + if (error) + return error; + f2fs_update_iostat(sbi, NULL, APP_DIRECT_IO, size); + return 0; +} + +static const struct iomap_dio_ops f2fs_iomap_dio_write_ops = { + .end_io = f2fs_dio_write_end_io, +}; + +static ssize_t f2fs_dio_write_iter(struct kiocb *iocb, struct iov_iter *from, + bool *may_need_sync) { struct file *file = iocb->ki_filp; struct inode *inode = file_inode(file); + struct f2fs_inode_info *fi = F2FS_I(inode); + struct f2fs_sb_info *sbi = F2FS_I_SB(inode); + const bool do_opu = f2fs_lfs_mode(sbi); + const loff_t pos = iocb->ki_pos; + const ssize_t count = iov_iter_count(from); + unsigned int dio_flags; + struct iomap_dio *dio; + ssize_t ret; + + trace_f2fs_direct_IO_enter(inode, iocb, count, WRITE); + + if (iocb->ki_flags & IOCB_NOWAIT) { + /* f2fs_convert_inline_inode() and block allocation can block */ + if (f2fs_has_inline_data(inode) || + !f2fs_overwrite_io(inode, pos, count)) { + ret = -EAGAIN; + goto out; + } + + if (!f2fs_down_read_trylock(&fi->i_gc_rwsem[WRITE])) { + ret = -EAGAIN; + goto out; + } + if (do_opu && !f2fs_down_read_trylock(&fi->i_gc_rwsem[READ])) { + f2fs_up_read(&fi->i_gc_rwsem[WRITE]); + ret = -EAGAIN; + goto out; + } + } else { + ret = f2fs_convert_inline_inode(inode); + if (ret) + goto out; + + f2fs_down_read(&fi->i_gc_rwsem[WRITE]); + if (do_opu) + f2fs_down_read(&fi->i_gc_rwsem[READ]); + } + + /* + * We have to use __iomap_dio_rw() and iomap_dio_complete() instead of + * the higher-level function iomap_dio_rw() in order to ensure that the + * F2FS_DIO_WRITE counter will be decremented correctly in all cases. + */ + inc_page_count(sbi, F2FS_DIO_WRITE); + dio_flags = 0; + if (pos + count > inode->i_size) + dio_flags |= IOMAP_DIO_FORCE_WAIT; + dio = __iomap_dio_rw(iocb, from, &f2fs_iomap_ops, + &f2fs_iomap_dio_write_ops, dio_flags, NULL, 0); + if (IS_ERR_OR_NULL(dio)) { + ret = PTR_ERR_OR_ZERO(dio); + if (ret == -ENOTBLK) + ret = 0; + if (ret != -EIOCBQUEUED) + dec_page_count(sbi, F2FS_DIO_WRITE); + } else { + ret = iomap_dio_complete(dio); + } + + if (do_opu) + f2fs_up_read(&fi->i_gc_rwsem[READ]); + f2fs_up_read(&fi->i_gc_rwsem[WRITE]); + + if (ret < 0) + goto out; + if (pos + ret > inode->i_size) + f2fs_i_size_write(inode, pos + ret); + if (!do_opu) + set_inode_flag(inode, FI_UPDATE_WRITE); + + if (iov_iter_count(from)) { + ssize_t ret2; + loff_t bufio_start_pos = iocb->ki_pos; + + /* + * The direct write was partial, so we need to fall back to a + * buffered write for the remainder. + */ + + ret2 = f2fs_buffered_write_iter(iocb, from); + if (iov_iter_count(from)) + f2fs_write_failed(inode, iocb->ki_pos); + if (ret2 < 0) + goto out; + + /* + * Ensure that the pagecache pages are written to disk and + * invalidated to preserve the expected O_DIRECT semantics. + */ + if (ret2 > 0) { + loff_t bufio_end_pos = bufio_start_pos + ret2 - 1; + + ret += ret2; + + ret2 = filemap_write_and_wait_range(file->f_mapping, + bufio_start_pos, + bufio_end_pos); + if (ret2 < 0) + goto out; + invalidate_mapping_pages(file->f_mapping, + bufio_start_pos >> PAGE_SHIFT, + bufio_end_pos >> PAGE_SHIFT); + } + } else { + /* iomap_dio_rw() already handled the generic_write_sync(). */ + *may_need_sync = false; + } +out: + trace_f2fs_direct_IO_exit(inode, pos, count, WRITE, ret); + return ret; +} + +static ssize_t f2fs_file_write_iter(struct kiocb *iocb, struct iov_iter *from) +{ + struct inode *inode = file_inode(iocb->ki_filp); + const loff_t orig_pos = iocb->ki_pos; + const size_t orig_count = iov_iter_count(from); + loff_t target_size; + bool dio; + bool may_need_sync = true; + int preallocated; ssize_t ret; if (unlikely(f2fs_cp_error(F2FS_I_SB(inode)))) { @@ -4260,91 +4642,64 @@ static ssize_t f2fs_file_write_iter(struct kiocb *iocb, struct iov_iter *from) inode_lock(inode); } - if (unlikely(IS_IMMUTABLE(inode))) { - ret = -EPERM; - goto unlock; - } + ret = f2fs_write_checks(iocb, from); + if (ret <= 0) + goto out_unlock; - if (is_inode_flag_set(inode, FI_COMPRESS_RELEASED)) { - ret = -EPERM; - goto unlock; - } + /* Determine whether we will do a direct write or a buffered write. */ + dio = f2fs_should_use_dio(inode, iocb, from); - ret = generic_write_checks(iocb, from); - if (ret > 0) { - bool preallocated = false; - size_t target_size = 0; - int err; - - if (fault_in_iov_iter_readable(from, iov_iter_count(from))) - set_inode_flag(inode, FI_NO_PREALLOC); - - if ((iocb->ki_flags & IOCB_NOWAIT)) { - if (!f2fs_overwrite_io(inode, iocb->ki_pos, - iov_iter_count(from)) || - f2fs_has_inline_data(inode) || - f2fs_force_buffered_io(inode, iocb, from)) { - clear_inode_flag(inode, FI_NO_PREALLOC); - inode_unlock(inode); - ret = -EAGAIN; - goto out; + /* Possibly preallocate the blocks for the write. */ + target_size = iocb->ki_pos + iov_iter_count(from); + preallocated = f2fs_preallocate_blocks(iocb, from, dio); + if (preallocated < 0) { + ret = preallocated; + } else { + if (trace_f2fs_datawrite_start_enabled()) { + char *p = f2fs_kmalloc(F2FS_I_SB(inode), + PATH_MAX, GFP_KERNEL); + char *path; + + if (!p) + goto skip_write_trace; + path = dentry_path_raw(file_dentry(iocb->ki_filp), + p, PATH_MAX); + if (IS_ERR(path)) { + kfree(p); + goto skip_write_trace; } - goto write; + trace_f2fs_datawrite_start(inode, orig_pos, orig_count, + current->pid, path, current->comm); + kfree(p); } +skip_write_trace: + /* Do the actual write. */ + ret = dio ? + f2fs_dio_write_iter(iocb, from, &may_need_sync) : + f2fs_buffered_write_iter(iocb, from); - if (is_inode_flag_set(inode, FI_NO_PREALLOC)) - goto write; - - if (iocb->ki_flags & IOCB_DIRECT) { - /* - * Convert inline data for Direct I/O before entering - * f2fs_direct_IO(). - */ - err = f2fs_convert_inline_inode(inode); - if (err) - goto out_err; - /* - * If force_buffere_io() is true, we have to allocate - * blocks all the time, since f2fs_direct_IO will fall - * back to buffered IO. - */ - if (!f2fs_force_buffered_io(inode, iocb, from) && - f2fs_lfs_mode(F2FS_I_SB(inode))) - goto write; - } - preallocated = true; - target_size = iocb->ki_pos + iov_iter_count(from); - - err = f2fs_preallocate_blocks(iocb, from); - if (err) { -out_err: - clear_inode_flag(inode, FI_NO_PREALLOC); - inode_unlock(inode); - ret = err; - goto out; - } -write: - ret = __generic_file_write_iter(iocb, from); - clear_inode_flag(inode, FI_NO_PREALLOC); - - /* if we couldn't write data, we should deallocate blocks. */ - if (preallocated && i_size_read(inode) < target_size) { - down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]); - filemap_invalidate_lock(inode->i_mapping); - f2fs_truncate(inode); - filemap_invalidate_unlock(inode->i_mapping); - up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]); - } + if (trace_f2fs_datawrite_end_enabled()) + trace_f2fs_datawrite_end(inode, orig_pos, ret); + } - if (ret > 0) - f2fs_update_iostat(F2FS_I_SB(inode), APP_WRITE_IO, ret); + /* Don't leave any preallocated blocks around past i_size. */ + if (preallocated && i_size_read(inode) < target_size) { + f2fs_down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]); + filemap_invalidate_lock(inode->i_mapping); + if (!f2fs_truncate(inode)) + file_dont_truncate(inode); + filemap_invalidate_unlock(inode->i_mapping); + f2fs_up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]); + } else { + file_dont_truncate(inode); } -unlock: + + clear_inode_flag(inode, FI_PREALLOCATED_ALL); +out_unlock: inode_unlock(inode); out: - trace_f2fs_file_write_iter(inode, iocb->ki_pos, - iov_iter_count(from), ret); - if (ret > 0) + trace_f2fs_file_write_iter(inode, orig_pos, orig_count, ret); + if (ret > 0 && may_need_sync) ret = generic_write_sync(iocb, ret); return ret; } @@ -4352,12 +4707,12 @@ out: static int f2fs_file_fadvise(struct file *filp, loff_t offset, loff_t len, int advice) { - struct inode *inode; struct address_space *mapping; struct backing_dev_info *bdi; + struct inode *inode = file_inode(filp); + int err; if (advice == POSIX_FADV_SEQUENTIAL) { - inode = file_inode(filp); if (S_ISFIFO(inode->i_mode)) return -ESPIPE; @@ -4374,7 +4729,13 @@ static int f2fs_file_fadvise(struct file *filp, loff_t offset, loff_t len, return 0; } - return generic_fadvise(filp, offset, len, advice); + err = generic_fadvise(filp, offset, len, advice); + if (!err && advice == POSIX_FADV_DONTNEED && + test_opt(F2FS_I_SB(inode), COMPRESS_CACHE) && + f2fs_compressed_file(inode)) + f2fs_invalidate_compress_pages(F2FS_I_SB(inode), inode->i_ino); + + return err; } #ifdef CONFIG_COMPAT @@ -4447,7 +4808,7 @@ long f2fs_compat_ioctl(struct file *file, unsigned int cmd, unsigned long arg) case F2FS_IOC_COMMIT_ATOMIC_WRITE: case F2FS_IOC_START_VOLATILE_WRITE: case F2FS_IOC_RELEASE_VOLATILE_WRITE: - case F2FS_IOC_ABORT_VOLATILE_WRITE: + case F2FS_IOC_ABORT_ATOMIC_WRITE: case F2FS_IOC_SHUTDOWN: case FITRIM: case FS_IOC_SET_ENCRYPTION_POLICY: |