From ccfb30001f37ace4690a74c27b4812cf054e123a Mon Sep 17 00:00:00 2001 From: Jaegeuk Kim Date: Fri, 13 Jun 2014 13:02:11 +0900 Subject: f2fs: fix to report newly allocate region as extent Previous get_block in f2fs didn't report the newly allocated region which has NEW_ADDR. For reader, it should not report, but fiemap needs this. So, this patch introduces two get_block sharing core function. Signed-off-by: Jaegeuk Kim diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c index 0924521..f8cf619 100644 --- a/fs/f2fs/data.c +++ b/fs/f2fs/data.c @@ -608,8 +608,8 @@ static int __allocate_data_block(struct dnode_of_data *dn) * b. do not use extent cache for better performance * c. give the block addresses to blockdev */ -static int get_data_block(struct inode *inode, sector_t iblock, - struct buffer_head *bh_result, int create) +static int __get_data_block(struct inode *inode, sector_t iblock, + struct buffer_head *bh_result, int create, bool fiemap) { struct f2fs_sb_info *sbi = F2FS_SB(inode->i_sb); unsigned int blkbits = inode->i_sb->s_blocksize_bits; @@ -637,7 +637,7 @@ static int get_data_block(struct inode *inode, sector_t iblock, err = 0; goto unlock_out; } - if (dn.data_blkaddr == NEW_ADDR) + if (dn.data_blkaddr == NEW_ADDR && !fiemap) goto put_out; if (dn.data_blkaddr != NULL_ADDR) { @@ -671,7 +671,7 @@ get_next: err = 0; goto unlock_out; } - if (dn.data_blkaddr == NEW_ADDR) + if (dn.data_blkaddr == NEW_ADDR && !fiemap) goto put_out; end_offset = ADDRS_PER_PAGE(dn.node_page, F2FS_I(inode)); @@ -708,10 +708,23 @@ out: return err; } +static int get_data_block(struct inode *inode, sector_t iblock, + struct buffer_head *bh_result, int create) +{ + return __get_data_block(inode, iblock, bh_result, create, false); +} + +static int get_data_block_fiemap(struct inode *inode, sector_t iblock, + struct buffer_head *bh_result, int create) +{ + return __get_data_block(inode, iblock, bh_result, create, true); +} + int f2fs_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo, u64 start, u64 len) { - return generic_block_fiemap(inode, fieinfo, start, len, get_data_block); + return generic_block_fiemap(inode, fieinfo, + start, len, get_data_block_fiemap); } static int f2fs_read_data_page(struct file *file, struct page *page) -- cgit v0.10.2 From ead432756ab2c76b1f1de742a1c8a06992cb98eb Mon Sep 17 00:00:00 2001 From: Jaegeuk Kim Date: Fri, 13 Jun 2014 13:05:55 +0900 Subject: f2fs: recover fallocated data and its i_size together This patch arranges the f2fs_locks to cover the fallocated data and its i_size. Signed-off-by: Jaegeuk Kim diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c index c58e330..f6c4bda 100644 --- a/fs/f2fs/file.c +++ b/fs/f2fs/file.c @@ -659,13 +659,13 @@ static int expand_inode_data(struct inode *inode, loff_t offset, off_start = offset & (PAGE_CACHE_SIZE - 1); off_end = (offset + len) & (PAGE_CACHE_SIZE - 1); + f2fs_lock_op(sbi); + for (index = pg_start; index <= pg_end; index++) { struct dnode_of_data dn; - f2fs_lock_op(sbi); set_new_dnode(&dn, inode, NULL, NULL, 0); ret = f2fs_reserve_block(&dn, index); - f2fs_unlock_op(sbi); if (ret) break; @@ -683,8 +683,9 @@ static int expand_inode_data(struct inode *inode, loff_t offset, i_size_read(inode) < new_size) { i_size_write(inode, new_size); mark_inode_dirty(inode); - f2fs_write_inode(inode, NULL); + update_inode_page(inode); } + f2fs_unlock_op(sbi); return ret; } -- cgit v0.10.2 From 98397ff3cddcfdedd2ba1701bd30a73c1d733769 Mon Sep 17 00:00:00 2001 From: Jaegeuk Kim Date: Fri, 13 Jun 2014 13:07:31 +0900 Subject: f2fs: fix not to allocate unnecessary blocks during fallocate This patch fixes the fallocate bug like below. (See xfstests/255) In fallocate(fd, 0, 20480), expand_inode_data processes for (index = pg_start; index <= pg_end; index++) { f2fs_reserve_block(); ... } So, even though fallocate requests 20480, 5 blocks, f2fs allocates 6 blocks including pg_end. So, this patch adds one condition to avoid block allocation. Signed-off-by: Jaegeuk Kim diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c index f6c4bda..7d8b962 100644 --- a/fs/f2fs/file.c +++ b/fs/f2fs/file.c @@ -664,11 +664,14 @@ static int expand_inode_data(struct inode *inode, loff_t offset, for (index = pg_start; index <= pg_end; index++) { struct dnode_of_data dn; + if (index == pg_end && !off_end) + goto noalloc; + set_new_dnode(&dn, inode, NULL, NULL, 0); ret = f2fs_reserve_block(&dn, index); if (ret) break; - +noalloc: if (pg_start == pg_end) new_size = offset + len; else if (index == pg_start && off_start) -- cgit v0.10.2 From 8bc6f60e3f7f31c4ce370b4b27b8f4b355b7f07e Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Wed, 11 Jun 2014 18:32:23 +0800 Subject: f2fs: remove unused variables in f2fs_sm_info Remove unused variables in struct f2fs_sm_info. Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index e51c732..7ef7acd 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -342,9 +342,6 @@ struct f2fs_sm_info { struct dirty_seglist_info *dirty_info; /* dirty segment information */ struct curseg_info *curseg_array; /* active segment information */ - struct list_head wblist_head; /* list of under-writeback pages */ - spinlock_t wblist_lock; /* lock for checkpoint */ - block_t seg0_blkaddr; /* block address of 0'th segment */ block_t main_blkaddr; /* start block address of main area */ block_t ssa_blkaddr; /* start block address of SSA area */ diff --git a/fs/f2fs/segment.c b/fs/f2fs/segment.c index f25f0e0..b22d5a0 100644 --- a/fs/f2fs/segment.c +++ b/fs/f2fs/segment.c @@ -1885,8 +1885,6 @@ int build_segment_manager(struct f2fs_sb_info *sbi) /* init sm info */ sbi->sm_info = sm_info; - INIT_LIST_HEAD(&sm_info->wblist_head); - spin_lock_init(&sm_info->wblist_lock); sm_info->seg0_blkaddr = le32_to_cpu(raw_super->segment0_blkaddr); sm_info->main_blkaddr = le32_to_cpu(raw_super->main_blkaddr); sm_info->segment_count = le32_to_cpu(raw_super->segment_count); -- cgit v0.10.2 From d6b7d4b31dfd5a454a71c445b8086bc098237334 Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Thu, 12 Jun 2014 13:23:41 +0800 Subject: f2fs: check lower bound nid value in check_nid_range This patch add lower bound verification for nid in check_nid_range, so nids reserved like 0, node, meta passed by caller could be checked there. And then check_nid_range could be used in f2fs_nfs_get_inode for simplifying code. Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index 7ef7acd..58df97e 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -641,7 +641,8 @@ static inline void f2fs_unlock_all(struct f2fs_sb_info *sbi) */ static inline int check_nid_range(struct f2fs_sb_info *sbi, nid_t nid) { - WARN_ON((nid >= NM_I(sbi)->max_nid)); + if (unlikely(nid < F2FS_ROOT_INO(sbi))) + return -EINVAL; if (unlikely(nid >= NM_I(sbi)->max_nid)) return -EINVAL; return 0; diff --git a/fs/f2fs/inode.c b/fs/f2fs/inode.c index adc622c..2cf6962 100644 --- a/fs/f2fs/inode.c +++ b/fs/f2fs/inode.c @@ -78,6 +78,7 @@ static int do_read_inode(struct inode *inode) if (check_nid_range(sbi, inode->i_ino)) { f2fs_msg(inode->i_sb, KERN_ERR, "bad inode number: %lu", (unsigned long) inode->i_ino); + WARN_ON(1); return -EINVAL; } diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c index b2b1863..8f96d93 100644 --- a/fs/f2fs/super.c +++ b/fs/f2fs/super.c @@ -689,9 +689,7 @@ static struct inode *f2fs_nfs_get_inode(struct super_block *sb, struct f2fs_sb_info *sbi = F2FS_SB(sb); struct inode *inode; - if (unlikely(ino < F2FS_ROOT_INO(sbi))) - return ERR_PTR(-ESTALE); - if (unlikely(ino >= NM_I(sbi)->max_nid)) + if (check_nid_range(sbi, ino)) return ERR_PTR(-ESTALE); /* -- cgit v0.10.2 From 90d72459ccb47335a4348947506fd091e63f7cf8 Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Tue, 24 Jun 2014 10:34:00 +0800 Subject: f2fs: fix error path in init_inode_metadata If we fail in this path: ->init_inode_metadata ->make_empty_dir ->get_new_data_page ->grab_cache_page return -ENOMEM We will bug on in error path of init_inode_metadata when call remove_inode_page because i_block = 2 (one inode block will be released later & one dentry block). We should release the dentry block in init_inode_metadata to avoid this BUG_ON, and avoid leak of dentry block resource, because we never have second chance to release that block in ->evict_inode as in upper error path we make this inode 'bad'. Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim diff --git a/fs/f2fs/dir.c b/fs/f2fs/dir.c index 966acb0..a4addd7 100644 --- a/fs/f2fs/dir.c +++ b/fs/f2fs/dir.c @@ -376,11 +376,11 @@ static struct page *init_inode_metadata(struct inode *inode, put_error: f2fs_put_page(page, 1); +error: /* once the failed inode becomes a bad inode, i_mode is S_IFREG */ truncate_inode_pages(&inode->i_data, 0); truncate_blocks(inode, 0); remove_dirty_dir_inode(inode); -error: remove_inode_page(inode); return ERR_PTR(err); } -- cgit v0.10.2 From dd4d961fe7d1cb1cba6e9d8d132475d4917ba864 Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Tue, 24 Jun 2014 14:13:13 +0800 Subject: f2fs: release new entry page correctly in error path of f2fs_rename This patch correct releasing code of new_page to avoid BUG_ON in error patch of f2fs_rename. Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim diff --git a/fs/f2fs/namei.c b/fs/f2fs/namei.c index 9138c32..aa3ddb2 100644 --- a/fs/f2fs/namei.c +++ b/fs/f2fs/namei.c @@ -474,7 +474,8 @@ static int f2fs_rename(struct inode *old_dir, struct dentry *old_dentry, return 0; put_out_dir: - f2fs_put_page(new_page, 1); + kunmap(new_page); + f2fs_put_page(new_page, 0); out_dir: if (old_dir_entry) { kunmap(old_dir_page); -- cgit v0.10.2 From b2c0829912493df596706a1a036c67beb1bd6ff5 Mon Sep 17 00:00:00 2001 From: Jaegeuk Kim Date: Mon, 30 Jun 2014 18:09:55 +0900 Subject: f2fs: do checkpoint for the renamed inode If an inode is renamed, it should be registered as file_lost_pino to conduct checkpoint at f2fs_sync_file. Otherwise, the inode cannot be recovered due to no dent_mark in the following scenario. Note that, this scenario is from xfstests/322. 1. create "a" 2. fsync "a" 3. rename "a" to "b" 4. fsync "b" 5. Sudden power-cut After recovery is done, "b" should be seen. However, the result shows "a", since the recovery procedure does not enter recover_dentry due to no dent_mark. The reason is like below. - The nid of "a" is checkpointed during #2, f2fs_sync_file. - The inode page for "b" produced by #3 is written without dent_mark by sync_node_pages. So, this patch fixes this bug by assinging file_lost_pino to the "a"'s inode. If the pino is lost, f2fs_sync_file conducts checkpoint, and then recovers the latest pino and its dentry information for further recovery. Signed-off-by: Jaegeuk Kim diff --git a/fs/f2fs/namei.c b/fs/f2fs/namei.c index aa3ddb2..a6bdddc 100644 --- a/fs/f2fs/namei.c +++ b/fs/f2fs/namei.c @@ -417,9 +417,6 @@ static int f2fs_rename(struct inode *old_dir, struct dentry *old_dentry, } f2fs_set_link(new_dir, new_entry, new_page, old_inode); - down_write(&F2FS_I(old_inode)->i_sem); - F2FS_I(old_inode)->i_pino = new_dir->i_ino; - up_write(&F2FS_I(old_inode)->i_sem); new_inode->i_ctime = CURRENT_TIME; down_write(&F2FS_I(new_inode)->i_sem); @@ -448,6 +445,10 @@ static int f2fs_rename(struct inode *old_dir, struct dentry *old_dentry, } } + down_write(&F2FS_I(old_inode)->i_sem); + file_lost_pino(old_inode); + up_write(&F2FS_I(old_inode)->i_sem); + old_inode->i_ctime = CURRENT_TIME; mark_inode_dirty(old_inode); @@ -457,9 +458,6 @@ static int f2fs_rename(struct inode *old_dir, struct dentry *old_dentry, if (old_dir != new_dir) { f2fs_set_link(old_inode, old_dir_entry, old_dir_page, new_dir); - down_write(&F2FS_I(old_inode)->i_sem); - F2FS_I(old_inode)->i_pino = new_dir->i_ino; - up_write(&F2FS_I(old_inode)->i_sem); update_inode_page(old_inode); } else { kunmap(old_dir_page); -- cgit v0.10.2 From 2743f865543c0c4a5e12fc13edb2bf89a6e9687c Mon Sep 17 00:00:00 2001 From: Jaegeuk Kim Date: Sat, 28 Jun 2014 01:00:41 +0900 Subject: f2fs: check bdi->dirty_exceeded when trying to skip data writes If we don't check the current backing device status, balance_dirty_pages can fall into infinite pausing routine. This can be occurred when a lot of directories make a small number of dirty dentry pages including files. Reported-by: Brian Chadwick Signed-off-by: Jaegeuk Kim diff --git a/fs/f2fs/node.c b/fs/f2fs/node.c index 9dfb9a0..4b697cc 100644 --- a/fs/f2fs/node.c +++ b/fs/f2fs/node.c @@ -42,6 +42,8 @@ bool available_free_memory(struct f2fs_sb_info *sbi, int type) mem_size = (nm_i->nat_cnt * sizeof(struct nat_entry)) >> 12; res = mem_size < ((val.totalram * nm_i->ram_thresh / 100) >> 2); } else if (type == DIRTY_DENTS) { + if (sbi->sb->s_bdi->dirty_exceeded) + return false; mem_size = get_pages(sbi, F2FS_DIRTY_DENTS); res = mem_size < ((val.totalram * nm_i->ram_thresh / 100) >> 1); } -- cgit v0.10.2 From 50e1f8d22199b557337b3d1ec8520e4c5aa5c76e Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Mon, 7 Jul 2014 09:39:32 +0800 Subject: f2fs: avoid to access NULL pointer in issue_flush_thread Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=75861 Denis 2014-05-10 11:28:59 UTC reported: "F2FS-fs (mmcblk0p28): mounting.. Unable to handle kernel NULL pointer dereference at virtual address 00000018 ... [] (_raw_spin_lock+0x3c/0x70) from [] (issue_flush_thread+0x50/0x17c) [] (issue_flush_thread+0x50/0x17c) from [] (kthread+0x98/0xa4) [] (kthread+0x98/0xa4) from [] (kernel_thread_exit+0x0/0x8)" This patch assign cmd_control_info in sm_info before issue_flush_thread is being created, so this make sure that issue flush thread will have no chance to access invalid info in fcc. Signed-off-by: Chao Yu Reviewed-by: Gu Zheng Signed-off-by: Jaegeuk Kim diff --git a/fs/f2fs/segment.c b/fs/f2fs/segment.c index b22d5a0..d04613d 100644 --- a/fs/f2fs/segment.c +++ b/fs/f2fs/segment.c @@ -272,14 +272,15 @@ int create_flush_cmd_control(struct f2fs_sb_info *sbi) return -ENOMEM; spin_lock_init(&fcc->issue_lock); init_waitqueue_head(&fcc->flush_wait_queue); + sbi->sm_info->cmd_control_info = fcc; fcc->f2fs_issue_flush = kthread_run(issue_flush_thread, sbi, "f2fs_flush-%u:%u", MAJOR(dev), MINOR(dev)); if (IS_ERR(fcc->f2fs_issue_flush)) { err = PTR_ERR(fcc->f2fs_issue_flush); kfree(fcc); + sbi->sm_info->cmd_control_info = NULL; return err; } - sbi->sm_info->cmd_control_info = fcc; return err; } -- cgit v0.10.2