From e4d7f2d3593e703a4d58b813f332a5c2aef5106f Mon Sep 17 00:00:00 2001 From: Keyur Patel Date: Wed, 10 Jun 2020 23:19:46 -0400 Subject: ext4: fix spelling mistakes in extents.c Fix spelling issues over the comments in the code. requsted ==> requested deterimined ==> determined insde ==> inside neet ==> need somthing ==> something Signed-off-by: Keyur Patel Reviewed-by: Jan Kara Link: https://lore.kernel.org/r/20200611031947.165079-1-iamkeyur96@gmail.com Signed-off-by: Theodore Ts'o --- fs/ext4/extents.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c index 221f240eae60..1106ee17aa21 100644 --- a/fs/ext4/extents.c +++ b/fs/ext4/extents.c @@ -1915,7 +1915,7 @@ out: /* * ext4_ext_insert_extent: - * tries to merge requsted extent into the existing extent or + * tries to merge requested extent into the existing extent or * inserts requested extent as new one into the tree, * creating new leaf in the no-space case. */ @@ -3125,7 +3125,7 @@ static int ext4_ext_zeroout(struct inode *inode, struct ext4_extent *ex) * * * Splits extent [a, b] into two extents [a, @split) and [@split, b], states - * of which are deterimined by split_flag. + * of which are determined by split_flag. * * There are two cases: * a> the extent are splitted into two extent. @@ -3650,7 +3650,7 @@ static int ext4_split_convert_extents(handle_t *handle, eof_block = map->m_lblk + map->m_len; /* * It is safe to convert extent to initialized via explicit - * zeroout only if extent is fully insde i_size or new_size. + * zeroout only if extent is fully inside i_size or new_size. */ depth = ext_depth(inode); ex = path[depth].p_ext; @@ -4495,7 +4495,7 @@ static long ext4_zero_range(struct file *file, loff_t offset, } /* - * Round up offset. This is not fallocate, we neet to zero out + * Round up offset. This is not fallocate, we need to zero out * blocks, so convert interior block aligned part of the range to * unwritten and possibly manually zero out unaligned parts of the * range. @@ -5579,7 +5579,7 @@ ext4_swap_extents(handle_t *handle, struct inode *inode1, } ex1 = path1[path1->p_depth].p_ext; ex2 = path2[path2->p_depth].p_ext; - /* Do we have somthing to swap ? */ + /* Do we have something to swap ? */ if (unlikely(!ex2 || !ex1)) goto finish; -- cgit v1.2.3 From e0f49d270d9d04beec113aa7cfacac6bddcb5765 Mon Sep 17 00:00:00 2001 From: Markus Elfring Date: Sat, 13 Jun 2020 19:12:24 +0200 Subject: ext4: delete unnecessary checks before brelse() The brelse() function tests whether its argument is NULL and then returns immediately. Thus remove the tests which are not needed around the shown calls. This issue was detected by using the Coccinelle software. Signed-off-by: Markus Elfring Reviewed-by: Ritesh Harjani Link: https://lore.kernel.org/r/0d713702-072f-a89c-20ec-ca70aa83a432@web.de Signed-off-by: Theodore Ts'o --- fs/ext4/extents.c | 6 ++---- fs/ext4/xattr.c | 3 +-- 2 files changed, 3 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c index 1106ee17aa21..4af3f36c8351 100644 --- a/fs/ext4/extents.c +++ b/fs/ext4/extents.c @@ -693,10 +693,8 @@ void ext4_ext_drop_refs(struct ext4_ext_path *path) return; depth = path->p_depth; for (i = 0; i <= depth; i++, path++) { - if (path->p_bh) { - brelse(path->p_bh); - path->p_bh = NULL; - } + brelse(path->p_bh); + path->p_bh = NULL; } } diff --git a/fs/ext4/xattr.c b/fs/ext4/xattr.c index 7d2f6576d954..cba4b877c606 100644 --- a/fs/ext4/xattr.c +++ b/fs/ext4/xattr.c @@ -1356,8 +1356,7 @@ retry: block = 0; while (wsize < bufsize) { - if (bh != NULL) - brelse(bh); + brelse(bh); csize = (bufsize - wsize) > blocksize ? blocksize : bufsize - wsize; bh = ext4_getblk(handle, ea_inode, block, 0); -- cgit v1.2.3 From e030a28810daf48b43ef6e92fa502ae410b12c8c Mon Sep 17 00:00:00 2001 From: Dio Putra Date: Sun, 14 Jun 2020 11:45:44 +0700 Subject: ext4: fix coding style in file.c Fixed a few coding style issues in file.c Signed-off-by: Dio Putra Link: https://lore.kernel.org/r/239fcd8f-d33f-8621-9e82-0416dd3f9c94@gmail.com Signed-off-by: Theodore Ts'o --- fs/ext4/file.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/ext4/file.c b/fs/ext4/file.c index 2a01e31a032c..453cafecf5cc 100644 --- a/fs/ext4/file.c +++ b/fs/ext4/file.c @@ -145,8 +145,7 @@ static int ext4_release_file(struct inode *inode, struct file *filp) /* if we are the last writer on the inode, drop the block reservation */ if ((filp->f_mode & FMODE_WRITE) && (atomic_read(&inode->i_writecount) == 1) && - !EXT4_I(inode)->i_reserved_data_blocks) - { + !EXT4_I(inode)->i_reserved_data_blocks) { down_write(&EXT4_I(inode)->i_data_sem); ext4_discard_preallocations(inode); up_write(&EXT4_I(inode)->i_data_sem); @@ -810,7 +809,7 @@ out: return err; } -static int ext4_file_open(struct inode * inode, struct file * filp) +static int ext4_file_open(struct inode *inode, struct file *filp) { int ret; -- cgit v1.2.3 From 24dc9864914eb5813173cfa53313fcd02e4aea7d Mon Sep 17 00:00:00 2001 From: Lukas Czerner Date: Wed, 17 Jun 2020 11:25:49 +0200 Subject: jbd2: make sure jh have b_transaction set in refile/unfile_buffer Callers of __jbd2_journal_unfile_buffer() and __jbd2_journal_refile_buffer() assume that the b_transaction is set. In fact if it's not, we can end up with journal_head refcounting errors leading to crash much later that might be very hard to track down. Add asserts to make sure that is the case. We also make sure that b_next_transaction is NULL in __jbd2_journal_unfile_buffer() since the callers expect that as well and we should not get into that stage in this state anyway, leading to problems later on if we do. Tested with fstests. Signed-off-by: Lukas Czerner Reviewed-by: Jan Kara Link: https://lore.kernel.org/r/20200617092549.6712-1-lczerner@redhat.com Signed-off-by: Theodore Ts'o --- fs/jbd2/transaction.c | 10 ++++++++++ 1 file changed, 10 insertions(+) (limited to 'fs') diff --git a/fs/jbd2/transaction.c b/fs/jbd2/transaction.c index e91aad3637a2..e65e0aca2826 100644 --- a/fs/jbd2/transaction.c +++ b/fs/jbd2/transaction.c @@ -2026,6 +2026,9 @@ static void __jbd2_journal_temp_unlink_buffer(struct journal_head *jh) */ static void __jbd2_journal_unfile_buffer(struct journal_head *jh) { + J_ASSERT_JH(jh, jh->b_transaction != NULL); + J_ASSERT_JH(jh, jh->b_next_transaction == NULL); + __jbd2_journal_temp_unlink_buffer(jh); jh->b_transaction = NULL; } @@ -2572,6 +2575,13 @@ bool __jbd2_journal_refile_buffer(struct journal_head *jh) was_dirty = test_clear_buffer_jbddirty(bh); __jbd2_journal_temp_unlink_buffer(jh); + + /* + * b_transaction must be set, otherwise the new b_transaction won't + * be holding jh reference + */ + J_ASSERT_JH(jh, jh->b_transaction != NULL); + /* * We set b_transaction here because b_next_transaction will inherit * our jh reference and thus __jbd2_journal_file_buffer() must not -- cgit v1.2.3 From 5872331b3d91820e14716632ebb56b1399b34fe1 Mon Sep 17 00:00:00 2001 From: Eric Sandeen Date: Wed, 17 Jun 2020 14:19:04 -0500 Subject: ext4: fix potential negative array index in do_split() If for any reason a directory passed to do_split() does not have enough active entries to exceed half the size of the block, we can end up iterating over all "count" entries without finding a split point. In this case, count == move, and split will be zero, and we will attempt a negative index into map[]. Guard against this by detecting this case, and falling back to split-to-half-of-count instead; in this case we will still have plenty of space (> half blocksize) in each split block. Fixes: ef2b02d3e617 ("ext34: ensure do_split leaves enough free space in both blocks") Signed-off-by: Eric Sandeen Reviewed-by: Andreas Dilger Reviewed-by: Jan Kara Link: https://lore.kernel.org/r/f53e246b-647c-64bb-16ec-135383c70ad7@redhat.com Signed-off-by: Theodore Ts'o --- fs/ext4/namei.c | 16 +++++++++++++--- 1 file changed, 13 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c index 56738b538ddf..ef606301a106 100644 --- a/fs/ext4/namei.c +++ b/fs/ext4/namei.c @@ -1858,7 +1858,7 @@ static struct ext4_dir_entry_2 *do_split(handle_t *handle, struct inode *dir, blocksize, hinfo, map); map -= count; dx_sort_map(map, count); - /* Split the existing block in the middle, size-wise */ + /* Ensure that neither split block is over half full */ size = 0; move = 0; for (i = count-1; i >= 0; i--) { @@ -1868,8 +1868,18 @@ static struct ext4_dir_entry_2 *do_split(handle_t *handle, struct inode *dir, size += map[i].size; move++; } - /* map index at which we will split */ - split = count - move; + /* + * map index at which we will split + * + * If the sum of active entries didn't exceed half the block size, just + * split it in half by count; each resulting block will have at least + * half the space free. + */ + if (i > 0) + split = count - move; + else + split = count/2; + hash2 = map[split].hash; continued = hash2 == map[split - 1].hash; dxtrace(printk(KERN_INFO "Split block %lu at %x, %i/%i\n", -- cgit v1.2.3 From ef3f5830b859604eda8723c26d90ab23edc027a4 Mon Sep 17 00:00:00 2001 From: "zhangyi (F)" Date: Sat, 20 Jun 2020 14:19:48 +0800 Subject: jbd2: add the missing unlock_buffer() in the error path of jbd2_write_superblock() jbd2_write_superblock() is under the buffer lock of journal superblock before ending that superblock write, so add a missing unlock_buffer() in in the error path before submitting buffer. Fixes: 742b06b5628f ("jbd2: check superblock mapped prior to committing") Signed-off-by: zhangyi (F) Reviewed-by: Ritesh Harjani Cc: stable@kernel.org Link: https://lore.kernel.org/r/20200620061948.2049579-1-yi.zhang@huawei.com Signed-off-by: Theodore Ts'o --- fs/jbd2/journal.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c index e4944436e733..5493a0da23dd 100644 --- a/fs/jbd2/journal.c +++ b/fs/jbd2/journal.c @@ -1367,8 +1367,10 @@ static int jbd2_write_superblock(journal_t *journal, int write_flags) int ret; /* Buffer got discarded which means block device got invalidated */ - if (!buffer_mapped(bh)) + if (!buffer_mapped(bh)) { + unlock_buffer(bh); return -EIO; + } trace_jbd2_write_superblock(journal, write_flags); if (!(journal->j_flags & JBD2_BARRIER)) -- cgit v1.2.3 From 9a5d265fed014115f35e598022c956e5d2fb863e Mon Sep 17 00:00:00 2001 From: zhengliang Date: Wed, 1 Jul 2020 16:30:27 +0800 Subject: ext4: lost matching-pair of trace in ext4_truncate It should call trace exit in all return path for ext4_truncate. Signed-off-by: zhengliang Reviewed-by: Andreas Dilger Reviewed-by: Ritesh Harjani Link: https://lore.kernel.org/r/20200701083027.45996-1-zhengliang6@huawei.com Signed-off-by: Theodore Ts'o --- fs/ext4/inode.c | 17 +++++++++-------- 1 file changed, 9 insertions(+), 8 deletions(-) (limited to 'fs') diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 10dd470876b3..6187c8880c02 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -4163,7 +4163,7 @@ int ext4_truncate(struct inode *inode) trace_ext4_truncate_enter(inode); if (!ext4_can_truncate(inode)) - return 0; + goto out_trace; if (inode->i_size == 0 && !test_opt(inode->i_sb, NO_AUTO_DA_ALLOC)) ext4_set_inode_state(inode, EXT4_STATE_DA_ALLOC_CLOSE); @@ -4172,16 +4172,14 @@ int ext4_truncate(struct inode *inode) int has_inline = 1; err = ext4_inline_data_truncate(inode, &has_inline); - if (err) - return err; - if (has_inline) - return 0; + if (err || has_inline) + goto out_trace; } /* If we zero-out tail of the page, we have to create jinode for jbd2 */ if (inode->i_size & (inode->i_sb->s_blocksize - 1)) { if (ext4_inode_attach_jinode(inode) < 0) - return 0; + goto out_trace; } if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) @@ -4190,8 +4188,10 @@ int ext4_truncate(struct inode *inode) credits = ext4_blocks_for_truncate(inode); handle = ext4_journal_start(inode, EXT4_HT_TRUNCATE, credits); - if (IS_ERR(handle)) - return PTR_ERR(handle); + if (IS_ERR(handle)) { + err = PTR_ERR(handle); + goto out_trace; + } if (inode->i_size & (inode->i_sb->s_blocksize - 1)) ext4_block_truncate_page(handle, mapping, inode->i_size); @@ -4242,6 +4242,7 @@ out_stop: err = err2; ext4_journal_stop(handle); +out_trace: trace_ext4_truncate_exit(inode); return err; } -- cgit v1.2.3 From e5f78159d62f79eb7923561275290b7c95e7f03f Mon Sep 17 00:00:00 2001 From: Yi Zhuang Date: Mon, 29 Jun 2020 20:26:21 +0800 Subject: ext4: lost matching-pair of trace in ext4_unlink If dquot_initialize() return non-zero and trace of ext4_unlink_enter/exit enabled then the matching-pair of trace_exit will lost in log. Signed-off-by: Yi Zhuang Reviewed-by: Andreas Dilger Reviewed-by: Ritesh Harjani Link: https://lore.kernel.org/r/20200629122621.129953-1-zhuangyi1@huawei.com Signed-off-by: Theodore Ts'o --- fs/ext4/namei.c | 38 +++++++++++++++++++++----------------- 1 file changed, 21 insertions(+), 17 deletions(-) (limited to 'fs') diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c index ef606301a106..cb2eb1967e73 100644 --- a/fs/ext4/namei.c +++ b/fs/ext4/namei.c @@ -3203,30 +3203,33 @@ static int ext4_unlink(struct inode *dir, struct dentry *dentry) * in separate transaction */ retval = dquot_initialize(dir); if (retval) - return retval; + goto out_trace; retval = dquot_initialize(d_inode(dentry)); if (retval) - return retval; + goto out_trace; - retval = -ENOENT; bh = ext4_find_entry(dir, &dentry->d_name, &de, NULL); - if (IS_ERR(bh)) - return PTR_ERR(bh); - if (!bh) - goto end_unlink; + if (IS_ERR(bh)) { + retval = PTR_ERR(bh); + goto out_trace; + } + if (!bh) { + retval = -ENOENT; + goto out_trace; + } inode = d_inode(dentry); - retval = -EFSCORRUPTED; - if (le32_to_cpu(de->inode) != inode->i_ino) - goto end_unlink; + if (le32_to_cpu(de->inode) != inode->i_ino) { + retval = -EFSCORRUPTED; + goto out_bh; + } handle = ext4_journal_start(dir, EXT4_HT_DIR, EXT4_DATA_TRANS_BLOCKS(dir->i_sb)); if (IS_ERR(handle)) { retval = PTR_ERR(handle); - handle = NULL; - goto end_unlink; + goto out_bh; } if (IS_DIRSYNC(dir)) @@ -3234,12 +3237,12 @@ static int ext4_unlink(struct inode *dir, struct dentry *dentry) retval = ext4_delete_entry(handle, dir, de, bh); if (retval) - goto end_unlink; + goto out_handle; dir->i_ctime = dir->i_mtime = current_time(dir); ext4_update_dx_flag(dir); retval = ext4_mark_inode_dirty(handle, dir); if (retval) - goto end_unlink; + goto out_handle; if (inode->i_nlink == 0) ext4_warning_inode(inode, "Deleting file '%.*s' with no links", dentry->d_name.len, dentry->d_name.name); @@ -3261,10 +3264,11 @@ static int ext4_unlink(struct inode *dir, struct dentry *dentry) d_invalidate(dentry); #endif -end_unlink: +out_handle: + ext4_journal_stop(handle); +out_bh: brelse(bh); - if (handle) - ext4_journal_stop(handle); +out_trace: trace_ext4_unlink_exit(dentry, retval); return retval; } -- cgit v1.2.3 From e65bf6e468c39a60c1dfae439198ad875bb8a870 Mon Sep 17 00:00:00 2001 From: "Alexander A. Klimov" Date: Mon, 6 Jul 2020 21:03:39 +0200 Subject: ext4: replace HTTP links with HTTPS ones Rationale: Reduces attack surface on kernel devs opening the links for MITM as HTTPS traffic is much harder to manipulate. Deterministic algorithm: For each file: If not .svg: For each line: If doesn't contain `\bxmlns\b`: For each link, `\bhttp://[^# \t\r\n]*(?:\w|/)`: If both the HTTP and HTTPS versions return 200 OK and serve the same content: Replace HTTP with HTTPS. Signed-off-by: Alexander A. Klimov Link: https://lore.kernel.org/r/20200706190339.20709-1-grandmaster@al2klimov.de Signed-off-by: Theodore Ts'o --- Documentation/filesystems/ext4/about.rst | 2 +- fs/ext4/Kconfig | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/Documentation/filesystems/ext4/about.rst b/Documentation/filesystems/ext4/about.rst index 0aadba052264..cc76b577d2f4 100644 --- a/Documentation/filesystems/ext4/about.rst +++ b/Documentation/filesystems/ext4/about.rst @@ -39,6 +39,6 @@ entry. Other References ---------------- -Also see http://www.nongnu.org/ext2-doc/ for quite a collection of +Also see https://www.nongnu.org/ext2-doc/ for quite a collection of information about ext2/3. Here's another old reference: http://wiki.osdev.org/Ext2 diff --git a/fs/ext4/Kconfig b/fs/ext4/Kconfig index 1afa5a4bcb5f..619dd35ddd48 100644 --- a/fs/ext4/Kconfig +++ b/fs/ext4/Kconfig @@ -110,7 +110,7 @@ config EXT4_KUNIT_TESTS This builds the ext4 KUnit tests. KUnit tests run during boot and output the results to the debug log - in TAP format (http://testanything.org/). Only useful for kernel devs + in TAP format (https://testanything.org/). Only useful for kernel devs running KUnit test harness and are not for inclusion into a production build. -- cgit v1.2.3 From 0b3171b6d195637f84ddf8b59bae818ea20bc8ac Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Wed, 8 Jul 2020 17:35:16 +0200 Subject: ext4: do not block RWF_NOWAIT dio write on unallocated space Since commit 378f32bab371 ("ext4: introduce direct I/O write using iomap infrastructure") we don't properly bail out of RWF_NOWAIT direct IO write if underlying blocks are not allocated. Also ext4_dio_write_checks() does not honor RWF_NOWAIT when re-acquiring i_rwsem. Fix both issues. Fixes: 378f32bab371 ("ext4: introduce direct I/O write using iomap infrastructure") Cc: stable@kernel.org Reported-by: Filipe Manana Signed-off-by: Jan Kara Reviewed-by: Ritesh Harjani Link: https://lore.kernel.org/r/20200708153516.9507-1-jack@suse.cz Signed-off-by: Theodore Ts'o --- fs/ext4/file.c | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'fs') diff --git a/fs/ext4/file.c b/fs/ext4/file.c index 453cafecf5cc..7a2720517bbb 100644 --- a/fs/ext4/file.c +++ b/fs/ext4/file.c @@ -427,6 +427,10 @@ restart: */ if (*ilock_shared && (!IS_NOSEC(inode) || *extend || !ext4_overwrite_io(inode, offset, count))) { + if (iocb->ki_flags & IOCB_NOWAIT) { + ret = -EAGAIN; + goto out; + } inode_unlock_shared(inode); *ilock_shared = false; inode_lock(inode); -- cgit v1.2.3 From 11215630aada28307ba555a43138db6ac54fa825 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Fri, 10 Jul 2020 16:07:59 +0200 Subject: ext4: don't BUG on inconsistent journal feature A customer has reported a BUG_ON in ext4_clear_journal_err() hitting during an LTP testing. Either this has been caused by a test setup issue where the filesystem was being overwritten while LTP was mounting it or the journal replay has overwritten the superblock with invalid data. In either case it is preferable we don't take the machine down with a BUG_ON. So handle the situation of unexpectedly missing has_journal feature more gracefully. We issue warning and fail the mount in the cases where the race window is narrow and the failed check is most likely a programming error. In cases where fs corruption is more likely, we do full ext4_error() handling before failing mount / remount. Reviewed-by: Lukas Czerner Signed-off-by: Jan Kara Link: https://lore.kernel.org/r/20200710140759.18031-1-jack@suse.cz Signed-off-by: Theodore Ts'o --- fs/ext4/super.c | 68 +++++++++++++++++++++++++++++++++++++++------------------ 1 file changed, 47 insertions(+), 21 deletions(-) (limited to 'fs') diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 330957ed1f05..9fdad843b30e 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -66,10 +66,10 @@ static int ext4_load_journal(struct super_block *, struct ext4_super_block *, unsigned long journal_devnum); static int ext4_show_options(struct seq_file *seq, struct dentry *root); static int ext4_commit_super(struct super_block *sb, int sync); -static void ext4_mark_recovery_complete(struct super_block *sb, +static int ext4_mark_recovery_complete(struct super_block *sb, struct ext4_super_block *es); -static void ext4_clear_journal_err(struct super_block *sb, - struct ext4_super_block *es); +static int ext4_clear_journal_err(struct super_block *sb, + struct ext4_super_block *es); static int ext4_sync_fs(struct super_block *sb, int wait); static int ext4_remount(struct super_block *sb, int *flags, char *data); static int ext4_statfs(struct dentry *dentry, struct kstatfs *buf); @@ -4770,7 +4770,9 @@ no_journal: EXT4_SB(sb)->s_mount_state &= ~EXT4_ORPHAN_FS; if (needs_recovery) { ext4_msg(sb, KERN_INFO, "recovery complete"); - ext4_mark_recovery_complete(sb, es); + err = ext4_mark_recovery_complete(sb, es); + if (err) + goto failed_mount8; } if (EXT4_SB(sb)->s_journal) { if (test_opt(sb, DATA_FLAGS) == EXT4_MOUNT_JOURNAL_DATA) @@ -4813,10 +4815,8 @@ cantfind_ext4: ext4_msg(sb, KERN_ERR, "VFS: Can't find ext4 filesystem"); goto failed_mount; -#ifdef CONFIG_QUOTA failed_mount8: ext4_unregister_sysfs(sb); -#endif failed_mount7: ext4_unregister_li_request(sb); failed_mount6: @@ -4956,7 +4956,8 @@ static journal_t *ext4_get_journal(struct super_block *sb, struct inode *journal_inode; journal_t *journal; - BUG_ON(!ext4_has_feature_journal(sb)); + if (WARN_ON_ONCE(!ext4_has_feature_journal(sb))) + return NULL; journal_inode = ext4_get_journal_inode(sb, journal_inum); if (!journal_inode) @@ -4986,7 +4987,8 @@ static journal_t *ext4_get_dev_journal(struct super_block *sb, struct ext4_super_block *es; struct block_device *bdev; - BUG_ON(!ext4_has_feature_journal(sb)); + if (WARN_ON_ONCE(!ext4_has_feature_journal(sb))) + return NULL; bdev = ext4_blkdev_get(j_dev, sb); if (bdev == NULL) @@ -5078,7 +5080,8 @@ static int ext4_load_journal(struct super_block *sb, int err = 0; int really_read_only; - BUG_ON(!ext4_has_feature_journal(sb)); + if (WARN_ON_ONCE(!ext4_has_feature_journal(sb))) + return -EFSCORRUPTED; if (journal_devnum && journal_devnum != le32_to_cpu(es->s_journal_dev)) { @@ -5148,7 +5151,12 @@ static int ext4_load_journal(struct super_block *sb, } EXT4_SB(sb)->s_journal = journal; - ext4_clear_journal_err(sb, es); + err = ext4_clear_journal_err(sb, es); + if (err) { + EXT4_SB(sb)->s_journal = NULL; + jbd2_journal_destroy(journal); + return err; + } if (!really_read_only && journal_devnum && journal_devnum != le32_to_cpu(es->s_journal_dev)) { @@ -5244,26 +5252,32 @@ static int ext4_commit_super(struct super_block *sb, int sync) * remounting) the filesystem readonly, then we will end up with a * consistent fs on disk. Record that fact. */ -static void ext4_mark_recovery_complete(struct super_block *sb, - struct ext4_super_block *es) +static int ext4_mark_recovery_complete(struct super_block *sb, + struct ext4_super_block *es) { + int err; journal_t *journal = EXT4_SB(sb)->s_journal; if (!ext4_has_feature_journal(sb)) { - BUG_ON(journal != NULL); - return; + if (journal != NULL) { + ext4_error(sb, "Journal got removed while the fs was " + "mounted!"); + return -EFSCORRUPTED; + } + return 0; } jbd2_journal_lock_updates(journal); - if (jbd2_journal_flush(journal) < 0) + err = jbd2_journal_flush(journal); + if (err < 0) goto out; if (ext4_has_feature_journal_needs_recovery(sb) && sb_rdonly(sb)) { ext4_clear_feature_journal_needs_recovery(sb); ext4_commit_super(sb, 1); } - out: jbd2_journal_unlock_updates(journal); + return err; } /* @@ -5271,14 +5285,17 @@ out: * has recorded an error from a previous lifetime, move that error to the * main filesystem now. */ -static void ext4_clear_journal_err(struct super_block *sb, +static int ext4_clear_journal_err(struct super_block *sb, struct ext4_super_block *es) { journal_t *journal; int j_errno; const char *errstr; - BUG_ON(!ext4_has_feature_journal(sb)); + if (!ext4_has_feature_journal(sb)) { + ext4_error(sb, "Journal got removed while the fs was mounted!"); + return -EFSCORRUPTED; + } journal = EXT4_SB(sb)->s_journal; @@ -5303,6 +5320,7 @@ static void ext4_clear_journal_err(struct super_block *sb, jbd2_journal_clear_err(journal); jbd2_journal_update_sb_errno(journal); } + return 0; } /* @@ -5573,8 +5591,13 @@ static int ext4_remount(struct super_block *sb, int *flags, char *data) (sbi->s_mount_state & EXT4_VALID_FS)) es->s_state = cpu_to_le16(sbi->s_mount_state); - if (sbi->s_journal) + if (sbi->s_journal) { + /* + * We let remount-ro finish even if marking fs + * as clean failed... + */ ext4_mark_recovery_complete(sb, es); + } if (sbi->s_mmp_tsk) kthread_stop(sbi->s_mmp_tsk); } else { @@ -5622,8 +5645,11 @@ static int ext4_remount(struct super_block *sb, int *flags, char *data) * been changed by e2fsck since we originally mounted * the partition.) */ - if (sbi->s_journal) - ext4_clear_journal_err(sb, es); + if (sbi->s_journal) { + err = ext4_clear_journal_err(sb, es); + if (err) + goto restore_opts; + } sbi->s_mount_state = le16_to_cpu(es->s_state); err = ext4_setup_super(sb, es, 0); -- cgit v1.2.3 From 2a12e147da38807f8aab03c92de99a70ab983a96 Mon Sep 17 00:00:00 2001 From: Eric Biggers Date: Sun, 12 Jul 2020 20:10:12 -0700 Subject: ext4: don't hardcode bit values in EXT4_FL_USER_* Define the EXT4_FL_USER_* constants by OR-ing together the appropriate flags, rather than hard-coding a numeric value. This makes it much easier to see which flags are listed. No change in the actual values. Signed-off-by: Eric Biggers Reviewed-by: Jan Kara Link: https://lore.kernel.org/r/20200713031012.192440-1-ebiggers@kernel.org Signed-off-by: Theodore Ts'o --- fs/ext4/ext4.h | 30 ++++++++++++++++++++++++++++-- 1 file changed, 28 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index 42f5060f3cdf..b603a28a3696 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -434,8 +434,34 @@ struct flex_groups { #define EXT4_CASEFOLD_FL 0x40000000 /* Casefolded directory */ #define EXT4_RESERVED_FL 0x80000000 /* reserved for ext4 lib */ -#define EXT4_FL_USER_VISIBLE 0x725BDFFF /* User visible flags */ -#define EXT4_FL_USER_MODIFIABLE 0x624BC0FF /* User modifiable flags */ +/* User modifiable flags */ +#define EXT4_FL_USER_MODIFIABLE (EXT4_SECRM_FL | \ + EXT4_UNRM_FL | \ + EXT4_COMPR_FL | \ + EXT4_SYNC_FL | \ + EXT4_IMMUTABLE_FL | \ + EXT4_APPEND_FL | \ + EXT4_NODUMP_FL | \ + EXT4_NOATIME_FL | \ + EXT4_JOURNAL_DATA_FL | \ + EXT4_NOTAIL_FL | \ + EXT4_DIRSYNC_FL | \ + EXT4_TOPDIR_FL | \ + EXT4_EXTENTS_FL | \ + 0x00400000 /* EXT4_EOFBLOCKS_FL */ | \ + EXT4_DAX_FL | \ + EXT4_PROJINHERIT_FL | \ + EXT4_CASEFOLD_FL) + +/* User visible flags */ +#define EXT4_FL_USER_VISIBLE (EXT4_FL_USER_MODIFIABLE | \ + EXT4_DIRTY_FL | \ + EXT4_COMPRBLK_FL | \ + EXT4_NOCOMPR_FL | \ + EXT4_ENCRYPT_FL | \ + EXT4_INDEX_FL | \ + EXT4_VERITY_FL | \ + EXT4_INLINE_DATA_FL) /* Flags we can manipulate with through EXT4_IOC_FSSETXATTR */ #define EXT4_FL_XFLAG_VISIBLE (EXT4_SYNC_FL | \ -- cgit v1.2.3 From cb29a02d3a9d51c749f556b3bbf6551fbc0454eb Mon Sep 17 00:00:00 2001 From: Eric Biggers Date: Tue, 14 Jul 2020 16:09:09 -0700 Subject: ext4: use generic names for generic ioctls Don't define EXT4_IOC_* aliases to ioctls that already have a generic FS_IOC_* name. These aliases are unnecessary, and they make it unclear which ioctls are ext4-specific and which are generic. Exception: leave EXT4_IOC_GETVERSION_OLD and EXT4_IOC_SETVERSION_OLD as-is for now, since renaming them to FS_IOC_GETVERSION and FS_IOC_SETVERSION would probably make them more likely to be confused with EXT4_IOC_GETVERSION and EXT4_IOC_SETVERSION which also exist. Signed-off-by: Eric Biggers Reviewed-by: Jan Kara Link: https://lore.kernel.org/r/20200714230909.56349-1-ebiggers@kernel.org Signed-off-by: Theodore Ts'o --- Documentation/admin-guide/ext4.rst | 20 ++++++++++---------- fs/ext4/ext4.h | 12 +----------- fs/ext4/ioctl.c | 32 ++++++++++++++++---------------- 3 files changed, 27 insertions(+), 37 deletions(-) (limited to 'fs') diff --git a/Documentation/admin-guide/ext4.rst b/Documentation/admin-guide/ext4.rst index 9443fcef1876..7fc6a72920c9 100644 --- a/Documentation/admin-guide/ext4.rst +++ b/Documentation/admin-guide/ext4.rst @@ -522,21 +522,21 @@ Files in /sys/fs/ext4/: Ioctls ====== -There is some Ext4 specific functionality which can be accessed by applications -through the system call interfaces. The list of all Ext4 specific ioctls are -shown in the table below. +Ext4 implements various ioctls which can be used by applications to access +ext4-specific functionality. An incomplete list of these ioctls is shown in the +table below. This list includes truly ext4-specific ioctls (``EXT4_IOC_*``) as +well as ioctls that may have been ext4-specific originally but are now supported +by some other filesystem(s) too (``FS_IOC_*``). -Table of Ext4 specific ioctls +Table of Ext4 ioctls - EXT4_IOC_GETFLAGS + FS_IOC_GETFLAGS Get additional attributes associated with inode. The ioctl argument is - an integer bitfield, with bit values described in ext4.h. This ioctl is - an alias for FS_IOC_GETFLAGS. + an integer bitfield, with bit values described in ext4.h. - EXT4_IOC_SETFLAGS + FS_IOC_SETFLAGS Set additional attributes associated with inode. The ioctl argument is - an integer bitfield, with bit values described in ext4.h. This ioctl is - an alias for FS_IOC_SETFLAGS. + an integer bitfield, with bit values described in ext4.h. EXT4_IOC_GETVERSION, EXT4_IOC_GETVERSION_OLD Get the inode i_generation number stored for each inode. The diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index b603a28a3696..dd5e6c645442 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -463,7 +463,7 @@ struct flex_groups { EXT4_VERITY_FL | \ EXT4_INLINE_DATA_FL) -/* Flags we can manipulate with through EXT4_IOC_FSSETXATTR */ +/* Flags we can manipulate with through FS_IOC_FSSETXATTR */ #define EXT4_FL_XFLAG_VISIBLE (EXT4_SYNC_FL | \ EXT4_IMMUTABLE_FL | \ EXT4_APPEND_FL | \ @@ -695,8 +695,6 @@ enum { /* * ioctl commands */ -#define EXT4_IOC_GETFLAGS FS_IOC_GETFLAGS -#define EXT4_IOC_SETFLAGS FS_IOC_SETFLAGS #define EXT4_IOC_GETVERSION _IOR('f', 3, long) #define EXT4_IOC_SETVERSION _IOW('f', 4, long) #define EXT4_IOC_GETVERSION_OLD FS_IOC_GETVERSION @@ -713,17 +711,11 @@ enum { #define EXT4_IOC_RESIZE_FS _IOW('f', 16, __u64) #define EXT4_IOC_SWAP_BOOT _IO('f', 17) #define EXT4_IOC_PRECACHE_EXTENTS _IO('f', 18) -#define EXT4_IOC_SET_ENCRYPTION_POLICY FS_IOC_SET_ENCRYPTION_POLICY -#define EXT4_IOC_GET_ENCRYPTION_PWSALT FS_IOC_GET_ENCRYPTION_PWSALT -#define EXT4_IOC_GET_ENCRYPTION_POLICY FS_IOC_GET_ENCRYPTION_POLICY /* ioctl codes 19--39 are reserved for fscrypt */ #define EXT4_IOC_CLEAR_ES_CACHE _IO('f', 40) #define EXT4_IOC_GETSTATE _IOW('f', 41, __u32) #define EXT4_IOC_GET_ES_CACHE _IOWR('f', 42, struct fiemap) -#define EXT4_IOC_FSGETXATTR FS_IOC_FSGETXATTR -#define EXT4_IOC_FSSETXATTR FS_IOC_FSSETXATTR - #define EXT4_IOC_SHUTDOWN _IOR ('X', 125, __u32) /* @@ -748,8 +740,6 @@ enum { /* * ioctl commands in 32 bit emulation */ -#define EXT4_IOC32_GETFLAGS FS_IOC32_GETFLAGS -#define EXT4_IOC32_SETFLAGS FS_IOC32_SETFLAGS #define EXT4_IOC32_GETVERSION _IOR('f', 3, int) #define EXT4_IOC32_SETVERSION _IOW('f', 4, int) #define EXT4_IOC32_GETRSVSZ _IOR('f', 5, int) diff --git a/fs/ext4/ioctl.c b/fs/ext4/ioctl.c index 999cf6add39c..6e70a63dcca7 100644 --- a/fs/ext4/ioctl.c +++ b/fs/ext4/ioctl.c @@ -819,12 +819,12 @@ long ext4_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) switch (cmd) { case FS_IOC_GETFSMAP: return ext4_ioc_getfsmap(sb, (void __user *)arg); - case EXT4_IOC_GETFLAGS: + case FS_IOC_GETFLAGS: flags = ei->i_flags & EXT4_FL_USER_VISIBLE; if (S_ISREG(inode->i_mode)) flags &= ~EXT4_PROJINHERIT_FL; return put_user(flags, (int __user *) arg); - case EXT4_IOC_SETFLAGS: { + case FS_IOC_SETFLAGS: { int err; if (!inode_owner_or_capable(inode)) @@ -1129,12 +1129,12 @@ resizefs_out: case EXT4_IOC_PRECACHE_EXTENTS: return ext4_ext_precache(inode); - case EXT4_IOC_SET_ENCRYPTION_POLICY: + case FS_IOC_SET_ENCRYPTION_POLICY: if (!ext4_has_feature_encrypt(sb)) return -EOPNOTSUPP; return fscrypt_ioctl_set_policy(filp, (const void __user *)arg); - case EXT4_IOC_GET_ENCRYPTION_PWSALT: { + case FS_IOC_GET_ENCRYPTION_PWSALT: { #ifdef CONFIG_FS_ENCRYPTION int err, err2; struct ext4_sb_info *sbi = EXT4_SB(sb); @@ -1174,7 +1174,7 @@ resizefs_out: return -EOPNOTSUPP; #endif } - case EXT4_IOC_GET_ENCRYPTION_POLICY: + case FS_IOC_GET_ENCRYPTION_POLICY: if (!ext4_has_feature_encrypt(sb)) return -EOPNOTSUPP; return fscrypt_ioctl_get_policy(filp, (void __user *)arg); @@ -1236,7 +1236,7 @@ resizefs_out: case EXT4_IOC_GET_ES_CACHE: return ext4_ioctl_get_es_cache(filp, arg); - case EXT4_IOC_FSGETXATTR: + case FS_IOC_FSGETXATTR: { struct fsxattr fa; @@ -1247,7 +1247,7 @@ resizefs_out: return -EFAULT; return 0; } - case EXT4_IOC_FSSETXATTR: + case FS_IOC_FSSETXATTR: { struct fsxattr fa, old_fa; int err; @@ -1313,11 +1313,11 @@ long ext4_compat_ioctl(struct file *file, unsigned int cmd, unsigned long arg) { /* These are just misnamed, they actually get/put from/to user an int */ switch (cmd) { - case EXT4_IOC32_GETFLAGS: - cmd = EXT4_IOC_GETFLAGS; + case FS_IOC32_GETFLAGS: + cmd = FS_IOC_GETFLAGS; break; - case EXT4_IOC32_SETFLAGS: - cmd = EXT4_IOC_SETFLAGS; + case FS_IOC32_SETFLAGS: + cmd = FS_IOC_SETFLAGS; break; case EXT4_IOC32_GETVERSION: cmd = EXT4_IOC_GETVERSION; @@ -1361,9 +1361,9 @@ long ext4_compat_ioctl(struct file *file, unsigned int cmd, unsigned long arg) case EXT4_IOC_RESIZE_FS: case FITRIM: case EXT4_IOC_PRECACHE_EXTENTS: - case EXT4_IOC_SET_ENCRYPTION_POLICY: - case EXT4_IOC_GET_ENCRYPTION_PWSALT: - case EXT4_IOC_GET_ENCRYPTION_POLICY: + case FS_IOC_SET_ENCRYPTION_POLICY: + case FS_IOC_GET_ENCRYPTION_PWSALT: + case FS_IOC_GET_ENCRYPTION_POLICY: case FS_IOC_GET_ENCRYPTION_POLICY_EX: case FS_IOC_ADD_ENCRYPTION_KEY: case FS_IOC_REMOVE_ENCRYPTION_KEY: @@ -1377,8 +1377,8 @@ long ext4_compat_ioctl(struct file *file, unsigned int cmd, unsigned long arg) case EXT4_IOC_CLEAR_ES_CACHE: case EXT4_IOC_GETSTATE: case EXT4_IOC_GET_ES_CACHE: - case EXT4_IOC_FSGETXATTR: - case EXT4_IOC_FSSETXATTR: + case FS_IOC_FSGETXATTR: + case FS_IOC_FSSETXATTR: break; default: return -ENOIOCTLCMD; -- cgit v1.2.3 From 3cb77bd241a48c54bca7bd8510516939264944c5 Mon Sep 17 00:00:00 2001 From: brookxu Date: Wed, 15 Jul 2020 11:00:44 +0800 Subject: ext4: fix spelling typos in ext4_mb_initialize_context Fix spelling typos in ext4_mb_initialize_context. Signed-off-by: Chunguang Xu Reviewed-by: Jan Kara Link: https://lore.kernel.org/r/883b523c-58ec-7f38-0bb8-cd2ea4393684@gmail.com Signed-off-by: Theodore Ts'o --- fs/ext4/mballoc.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index c0a331e2feb0..6dc2c6c535ef 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -4399,7 +4399,7 @@ ext4_mb_initialize_context(struct ext4_allocation_context *ac, ac->ac_g_ex = ac->ac_o_ex; ac->ac_flags = ar->flags; - /* we have to define context: we'll we work with a file or + /* we have to define context: we'll work with a file or * locality group. this is a policy, actually */ ext4_mb_group_or_file(ac); -- cgit v1.2.3 From 273108fa5015eeffc4bacfa5ce272af3434b96e4 Mon Sep 17 00:00:00 2001 From: Lukas Czerner Date: Fri, 17 Jul 2020 11:06:05 +0200 Subject: ext4: handle read only external journal device Ext4 uses blkdev_get_by_dev() to get the block_device for journal device which does check to see if the read-only block device was opened read-only. As a result ext4 will hapily proceed mounting the file system with external journal on read-only device. This is bad as we would not be able to use the journal leading to errors later on. Instead of simply failing to mount file system in this case, treat it in a similar way we treat internal journal on read-only device. Allow to mount with -o noload in read-only mode. This can be reproduced easily like this: mke2fs -F -O journal_dev $JOURNAL_DEV 100M mkfs.$FSTYPE -F -J device=$JOURNAL_DEV $FS_DEV blockdev --setro $JOURNAL_DEV mount $FS_DEV $MNT touch $MNT/file umount $MNT leading to error like this [ 1307.318713] ------------[ cut here ]------------ [ 1307.323362] generic_make_request: Trying to write to read-only block-device dm-2 (partno 0) [ 1307.331741] WARNING: CPU: 36 PID: 3224 at block/blk-core.c:855 generic_make_request_checks+0x2c3/0x580 [ 1307.341041] Modules linked in: ext4 mbcache jbd2 rfkill intel_rapl_msr intel_rapl_common isst_if_commd [ 1307.419445] CPU: 36 PID: 3224 Comm: jbd2/dm-2 Tainted: G W I 5.8.0-rc5 #2 [ 1307.427359] Hardware name: Dell Inc. PowerEdge R740/01KPX8, BIOS 2.3.10 08/15/2019 [ 1307.434932] RIP: 0010:generic_make_request_checks+0x2c3/0x580 [ 1307.440676] Code: 94 03 00 00 48 89 df 48 8d 74 24 08 c6 05 cf 2b 18 01 01 e8 7f a4 ff ff 48 c7 c7 50e [ 1307.459420] RSP: 0018:ffffc0d70eb5fb48 EFLAGS: 00010286 [ 1307.464646] RAX: 0000000000000000 RBX: ffff9b33b2978300 RCX: 0000000000000000 [ 1307.471780] RDX: ffff9b33e12a81e0 RSI: ffff9b33e1298000 RDI: ffff9b33e1298000 [ 1307.478913] RBP: ffff9b7b9679e0c0 R08: 0000000000000837 R09: 0000000000000024 [ 1307.486044] R10: 0000000000000000 R11: ffffc0d70eb5f9f0 R12: 0000000000000400 [ 1307.493177] R13: 0000000000000000 R14: 0000000000000001 R15: 0000000000000000 [ 1307.500308] FS: 0000000000000000(0000) GS:ffff9b33e1280000(0000) knlGS:0000000000000000 [ 1307.508396] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 1307.514142] CR2: 000055eaf4109000 CR3: 0000003dee40a006 CR4: 00000000007606e0 [ 1307.521273] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [ 1307.528407] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 [ 1307.535538] PKRU: 55555554 [ 1307.538250] Call Trace: [ 1307.540708] generic_make_request+0x30/0x340 [ 1307.544985] submit_bio+0x43/0x190 [ 1307.548393] ? bio_add_page+0x62/0x90 [ 1307.552068] submit_bh_wbc+0x16a/0x190 [ 1307.555833] jbd2_write_superblock+0xec/0x200 [jbd2] [ 1307.560803] jbd2_journal_update_sb_log_tail+0x65/0xc0 [jbd2] [ 1307.566557] jbd2_journal_commit_transaction+0x2ae/0x1860 [jbd2] [ 1307.572566] ? check_preempt_curr+0x7a/0x90 [ 1307.576756] ? update_curr+0xe1/0x1d0 [ 1307.580421] ? account_entity_dequeue+0x7b/0xb0 [ 1307.584955] ? newidle_balance+0x231/0x3d0 [ 1307.589056] ? __switch_to_asm+0x42/0x70 [ 1307.592986] ? __switch_to_asm+0x36/0x70 [ 1307.596918] ? lock_timer_base+0x67/0x80 [ 1307.600851] kjournald2+0xbd/0x270 [jbd2] [ 1307.604873] ? finish_wait+0x80/0x80 [ 1307.608460] ? commit_timeout+0x10/0x10 [jbd2] [ 1307.612915] kthread+0x114/0x130 [ 1307.616152] ? kthread_park+0x80/0x80 [ 1307.619816] ret_from_fork+0x22/0x30 [ 1307.623400] ---[ end trace 27490236265b1630 ]--- Signed-off-by: Lukas Czerner Reviewed-by: Andreas Dilger Link: https://lore.kernel.org/r/20200717090605.2612-1-lczerner@redhat.com Signed-off-by: Theodore Ts'o --- fs/ext4/super.c | 51 +++++++++++++++++++++++++++++++++------------------ 1 file changed, 33 insertions(+), 18 deletions(-) (limited to 'fs') diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 9fdad843b30e..dda967efcbc2 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -5079,6 +5079,7 @@ static int ext4_load_journal(struct super_block *sb, dev_t journal_dev; int err = 0; int really_read_only; + int journal_dev_ro; if (WARN_ON_ONCE(!ext4_has_feature_journal(sb))) return -EFSCORRUPTED; @@ -5091,7 +5092,31 @@ static int ext4_load_journal(struct super_block *sb, } else journal_dev = new_decode_dev(le32_to_cpu(es->s_journal_dev)); - really_read_only = bdev_read_only(sb->s_bdev); + if (journal_inum && journal_dev) { + ext4_msg(sb, KERN_ERR, + "filesystem has both journal inode and journal device!"); + return -EINVAL; + } + + if (journal_inum) { + journal = ext4_get_journal(sb, journal_inum); + if (!journal) + return -EINVAL; + } else { + journal = ext4_get_dev_journal(sb, journal_dev); + if (!journal) + return -EINVAL; + } + + journal_dev_ro = bdev_read_only(journal->j_dev); + really_read_only = bdev_read_only(sb->s_bdev) | journal_dev_ro; + + if (journal_dev_ro && !sb_rdonly(sb)) { + ext4_msg(sb, KERN_ERR, + "journal device read-only, try mounting with '-o ro'"); + err = -EROFS; + goto err_out; + } /* * Are we loading a blank journal or performing recovery after a @@ -5106,27 +5131,14 @@ static int ext4_load_journal(struct super_block *sb, ext4_msg(sb, KERN_ERR, "write access " "unavailable, cannot proceed " "(try mounting with noload)"); - return -EROFS; + err = -EROFS; + goto err_out; } ext4_msg(sb, KERN_INFO, "write access will " "be enabled during recovery"); } } - if (journal_inum && journal_dev) { - ext4_msg(sb, KERN_ERR, "filesystem has both journal " - "and inode journals!"); - return -EINVAL; - } - - if (journal_inum) { - if (!(journal = ext4_get_journal(sb, journal_inum))) - return -EINVAL; - } else { - if (!(journal = ext4_get_dev_journal(sb, journal_dev))) - return -EINVAL; - } - if (!(journal->j_flags & JBD2_BARRIER)) ext4_msg(sb, KERN_INFO, "barriers disabled"); @@ -5146,8 +5158,7 @@ static int ext4_load_journal(struct super_block *sb, if (err) { ext4_msg(sb, KERN_ERR, "error loading journal"); - jbd2_journal_destroy(journal); - return err; + goto err_out; } EXT4_SB(sb)->s_journal = journal; @@ -5167,6 +5178,10 @@ static int ext4_load_journal(struct super_block *sb, } return 0; + +err_out: + jbd2_journal_destroy(journal); + return err; } static int ext4_commit_super(struct super_block *sb, int sync) -- cgit v1.2.3 From cfd73237722135807967f389bcbda558a60a30d6 Mon Sep 17 00:00:00 2001 From: Alex Zhuravlev Date: Tue, 21 Apr 2020 10:54:07 +0300 Subject: ext4: add prefetching for block allocation bitmaps This should significantly improve bitmap loading, especially for flex groups as it tries to load all bitmaps within a flex.group instead of one by one synchronously. Prefetching is done in 8 * flex_bg groups, so it should be 8 read-ahead reads for a single allocating thread. At the end of allocation the thread waits for read-ahead completion and initializes buddy information so that read-aheads are not lost in case of memory pressure. At cr=0 the number of prefetching IOs is limited per allocation context to prevent a situation when mballoc loads thousands of bitmaps looking for a perfect group and ignoring groups with good chunks. Together with the patch "ext4: limit scanning of uninitialized groups" the mount time (which includes few tiny allocations) of a 1PB filesystem is reduced significantly: 0% full 50%-full unpatched patched mount time 33s 9279s 563s [ Restructured by tytso; removed the state flags in the allocation context, so it can be used to lazily prefetch the allocation bitmaps immediately after the file system is mounted. Skip prefetching block groups which are uninitialized. Finally pass in the REQ_RAHEAD flag to the block layer while prefetching. ] Signed-off-by: Alex Zhuravlev Reviewed-by: Andreas Dilger Signed-off-by: Theodore Ts'o --- fs/ext4/balloc.c | 14 ++++-- fs/ext4/ext4.h | 8 +++- fs/ext4/mballoc.c | 133 +++++++++++++++++++++++++++++++++++++++++++++++++++++- fs/ext4/sysfs.c | 4 ++ 4 files changed, 153 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/ext4/balloc.c b/fs/ext4/balloc.c index 1ba46d87cdf1..1e2b1b4093aa 100644 --- a/fs/ext4/balloc.c +++ b/fs/ext4/balloc.c @@ -413,7 +413,8 @@ verified: * Return buffer_head on success or an ERR_PTR in case of failure. */ struct buffer_head * -ext4_read_block_bitmap_nowait(struct super_block *sb, ext4_group_t block_group) +ext4_read_block_bitmap_nowait(struct super_block *sb, ext4_group_t block_group, + bool ignore_locked) { struct ext4_group_desc *desc; struct ext4_sb_info *sbi = EXT4_SB(sb); @@ -441,6 +442,12 @@ ext4_read_block_bitmap_nowait(struct super_block *sb, ext4_group_t block_group) return ERR_PTR(-ENOMEM); } + if (ignore_locked && buffer_locked(bh)) { + /* buffer under IO already, return if called for prefetching */ + put_bh(bh); + return NULL; + } + if (bitmap_uptodate(bh)) goto verify; @@ -490,7 +497,8 @@ ext4_read_block_bitmap_nowait(struct super_block *sb, ext4_group_t block_group) trace_ext4_read_block_bitmap_load(sb, block_group); bh->b_end_io = ext4_end_bitmap_read; get_bh(bh); - submit_bh(REQ_OP_READ, REQ_META | REQ_PRIO, bh); + submit_bh(REQ_OP_READ, REQ_META | REQ_PRIO | + (ignore_locked ? REQ_RAHEAD : 0), bh); return bh; verify: err = ext4_validate_block_bitmap(sb, desc, block_group, bh); @@ -534,7 +542,7 @@ ext4_read_block_bitmap(struct super_block *sb, ext4_group_t block_group) struct buffer_head *bh; int err; - bh = ext4_read_block_bitmap_nowait(sb, block_group); + bh = ext4_read_block_bitmap_nowait(sb, block_group, false); if (IS_ERR(bh)) return bh; err = ext4_wait_block_bitmap(sb, block_group, bh); diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index dd5e6c645442..4fba138b8722 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -1521,6 +1521,8 @@ struct ext4_sb_info { /* where last allocation was done - for stream allocation */ unsigned long s_mb_last_group; unsigned long s_mb_last_start; + unsigned int s_mb_prefetch; + unsigned int s_mb_prefetch_limit; /* stats for buddy allocator */ atomic_t s_bal_reqs; /* number of reqs with len > 1 */ @@ -2462,7 +2464,8 @@ extern struct ext4_group_desc * ext4_get_group_desc(struct super_block * sb, extern int ext4_should_retry_alloc(struct super_block *sb, int *retries); extern struct buffer_head *ext4_read_block_bitmap_nowait(struct super_block *sb, - ext4_group_t block_group); + ext4_group_t block_group, + bool ignore_locked); extern int ext4_wait_block_bitmap(struct super_block *sb, ext4_group_t block_group, struct buffer_head *bh); @@ -3161,6 +3164,7 @@ struct ext4_group_info { (1 << EXT4_GROUP_INFO_BBITMAP_CORRUPT_BIT) #define EXT4_GROUP_INFO_IBITMAP_CORRUPT \ (1 << EXT4_GROUP_INFO_IBITMAP_CORRUPT_BIT) +#define EXT4_GROUP_INFO_BBITMAP_READ_BIT 4 #define EXT4_MB_GRP_NEED_INIT(grp) \ (test_bit(EXT4_GROUP_INFO_NEED_INIT_BIT, &((grp)->bb_state))) @@ -3175,6 +3179,8 @@ struct ext4_group_info { (set_bit(EXT4_GROUP_INFO_WAS_TRIMMED_BIT, &((grp)->bb_state))) #define EXT4_MB_GRP_CLEAR_TRIMMED(grp) \ (clear_bit(EXT4_GROUP_INFO_WAS_TRIMMED_BIT, &((grp)->bb_state))) +#define EXT4_MB_GRP_TEST_AND_SET_READ(grp) \ + (test_and_set_bit(EXT4_GROUP_INFO_BBITMAP_READ_BIT, &((grp)->bb_state))) #define EXT4_MAX_CONTENTION 8 #define EXT4_CONTENTION_THRESHOLD 2 diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index 6dc2c6c535ef..c350ae9cb0d9 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -922,7 +922,7 @@ static int ext4_mb_init_cache(struct page *page, char *incore, gfp_t gfp) bh[i] = NULL; continue; } - bh[i] = ext4_read_block_bitmap_nowait(sb, group); + bh[i] = ext4_read_block_bitmap_nowait(sb, group, false); if (IS_ERR(bh[i])) { err = PTR_ERR(bh[i]); bh[i] = NULL; @@ -2209,12 +2209,93 @@ out: return ret; } +/* + * Start prefetching @nr block bitmaps starting at @group. + * Return the next group which needs to be prefetched. + */ +static ext4_group_t +ext4_mb_prefetch(struct super_block *sb, ext4_group_t group, + unsigned int nr, int *cnt) +{ + ext4_group_t ngroups = ext4_get_groups_count(sb); + struct buffer_head *bh; + struct blk_plug plug; + + blk_start_plug(&plug); + while (nr-- > 0) { + struct ext4_group_desc *gdp = ext4_get_group_desc(sb, group, + NULL); + struct ext4_group_info *grp = ext4_get_group_info(sb, group); + + /* + * Prefetch block groups with free blocks; but don't + * bother if it is marked uninitialized on disk, since + * it won't require I/O to read. Also only try to + * prefetch once, so we avoid getblk() call, which can + * be expensive. + */ + if (!EXT4_MB_GRP_TEST_AND_SET_READ(grp) && + EXT4_MB_GRP_NEED_INIT(grp) && + ext4_free_group_clusters(sb, gdp) > 0 && + !(ext4_has_group_desc_csum(sb) && + (gdp->bg_flags & cpu_to_le16(EXT4_BG_BLOCK_UNINIT)))) { + bh = ext4_read_block_bitmap_nowait(sb, group, true); + if (bh && !IS_ERR(bh)) { + if (!buffer_uptodate(bh) && cnt) + (*cnt)++; + brelse(bh); + } + } + if (++group >= ngroups) + group = 0; + } + blk_finish_plug(&plug); + return group; +} + +/* + * Prefetching reads the block bitmap into the buffer cache; but we + * need to make sure that the buddy bitmap in the page cache has been + * initialized. Note that ext4_mb_init_group() will block if the I/O + * is not yet completed, or indeed if it was not initiated by + * ext4_mb_prefetch did not start the I/O. + * + * TODO: We should actually kick off the buddy bitmap setup in a work + * queue when the buffer I/O is completed, so that we don't block + * waiting for the block allocation bitmap read to finish when + * ext4_mb_prefetch_fini is called from ext4_mb_regular_allocator(). + */ +static void +ext4_mb_prefetch_fini(struct super_block *sb, ext4_group_t group, + unsigned int nr) +{ + while (nr-- > 0) { + struct ext4_group_desc *gdp = ext4_get_group_desc(sb, group, + NULL); + struct ext4_group_info *grp = ext4_get_group_info(sb, group); + + if (!group) + group = ext4_get_groups_count(sb); + group--; + grp = ext4_get_group_info(sb, group); + + if (EXT4_MB_GRP_NEED_INIT(grp) && + ext4_free_group_clusters(sb, gdp) > 0 && + !(ext4_has_group_desc_csum(sb) && + (gdp->bg_flags & cpu_to_le16(EXT4_BG_BLOCK_UNINIT)))) { + if (ext4_mb_init_group(sb, group, GFP_NOFS)) + break; + } + } +} + static noinline_for_stack int ext4_mb_regular_allocator(struct ext4_allocation_context *ac) { - ext4_group_t ngroups, group, i; + ext4_group_t prefetch_grp = 0, ngroups, group, i; int cr = -1; int err = 0, first_err = 0; + unsigned int nr = 0, prefetch_ios = 0; struct ext4_sb_info *sbi; struct super_block *sb; struct ext4_buddy e4b; @@ -2282,6 +2363,7 @@ repeat: * from the goal value specified */ group = ac->ac_g_ex.fe_group; + prefetch_grp = group; for (i = 0; i < ngroups; group++, i++) { int ret = 0; @@ -2293,6 +2375,29 @@ repeat: if (group >= ngroups) group = 0; + /* + * Batch reads of the block allocation bitmaps + * to get multiple READs in flight; limit + * prefetching at cr=0/1, otherwise mballoc can + * spend a lot of time loading imperfect groups + */ + if ((prefetch_grp == group) && + (cr > 1 || + prefetch_ios < sbi->s_mb_prefetch_limit)) { + unsigned int curr_ios = prefetch_ios; + + nr = sbi->s_mb_prefetch; + if (ext4_has_feature_flex_bg(sb)) { + nr = (group / sbi->s_mb_prefetch) * + sbi->s_mb_prefetch; + nr = nr + sbi->s_mb_prefetch - group; + } + prefetch_grp = ext4_mb_prefetch(sb, group, + nr, &prefetch_ios); + if (prefetch_ios == curr_ios) + nr = 0; + } + /* This now checks without needing the buddy page */ ret = ext4_mb_good_group_nolock(ac, group, cr); if (ret <= 0) { @@ -2367,6 +2472,10 @@ out: mb_debug(sb, "Best len %d, origin len %d, ac_status %u, ac_flags 0x%x, cr %d ret %d\n", ac->ac_b_ex.fe_len, ac->ac_o_ex.fe_len, ac->ac_status, ac->ac_flags, cr, err); + + if (nr) + ext4_mb_prefetch_fini(sb, prefetch_grp, nr); + return err; } @@ -2613,6 +2722,26 @@ static int ext4_mb_init_backend(struct super_block *sb) goto err_freebuddy; } + if (ext4_has_feature_flex_bg(sb)) { + /* a single flex group is supposed to be read by a single IO */ + sbi->s_mb_prefetch = 1 << sbi->s_es->s_log_groups_per_flex; + sbi->s_mb_prefetch *= 8; /* 8 prefetch IOs in flight at most */ + } else { + sbi->s_mb_prefetch = 32; + } + if (sbi->s_mb_prefetch > ext4_get_groups_count(sb)) + sbi->s_mb_prefetch = ext4_get_groups_count(sb); + /* now many real IOs to prefetch within a single allocation at cr=0 + * given cr=0 is an CPU-related optimization we shouldn't try to + * load too many groups, at some point we should start to use what + * we've got in memory. + * with an average random access time 5ms, it'd take a second to get + * 200 groups (* N with flex_bg), so let's make this limit 4 + */ + sbi->s_mb_prefetch_limit = sbi->s_mb_prefetch * 4; + if (sbi->s_mb_prefetch_limit > ext4_get_groups_count(sb)) + sbi->s_mb_prefetch_limit = ext4_get_groups_count(sb); + return 0; err_freebuddy: diff --git a/fs/ext4/sysfs.c b/fs/ext4/sysfs.c index 6c9fc9e21c13..31e0db726d21 100644 --- a/fs/ext4/sysfs.c +++ b/fs/ext4/sysfs.c @@ -240,6 +240,8 @@ EXT4_RO_ATTR_ES_STRING(last_error_func, s_last_error_func, 32); EXT4_ATTR(first_error_time, 0444, first_error_time); EXT4_ATTR(last_error_time, 0444, last_error_time); EXT4_ATTR(journal_task, 0444, journal_task); +EXT4_RW_ATTR_SBI_UI(mb_prefetch, s_mb_prefetch); +EXT4_RW_ATTR_SBI_UI(mb_prefetch_limit, s_mb_prefetch_limit); static unsigned int old_bump_val = 128; EXT4_ATTR_PTR(max_writeback_mb_bump, 0444, pointer_ui, &old_bump_val); @@ -283,6 +285,8 @@ static struct attribute *ext4_attrs[] = { #ifdef CONFIG_EXT4_DEBUG ATTR_LIST(simulate_fail), #endif + ATTR_LIST(mb_prefetch), + ATTR_LIST(mb_prefetch_limit), NULL, }; ATTRIBUTE_GROUPS(ext4); -- cgit v1.2.3 From c1d2c7d47e15482bb23cda83a5021e60f624a09c Mon Sep 17 00:00:00 2001 From: Alex Zhuravlev Date: Fri, 19 Jun 2020 22:08:56 -0400 Subject: ext4: skip non-loaded groups at cr=0/1 when scanning for good groups cr=0 is supposed to be an optimization to save CPU cycles, but if buddy data (in memory) is not initialized then all this makes no sense as we have to do sync IO taking a lot of cycles. Also, at cr=0 mballoc doesn't choose any available chunk. cr=1 also skips groups using heuristic based on avg. fragment size. It's more useful to skip such groups and switch to cr=2 where groups will be scanned for available chunks. However, we always read the first block group in a flex_bg so metadata blocks will get read into the first flex_bg if possible. Using sparse image and dm-slow virtual device of 120TB was simulated, then the image was formatted and filled using debugfs to mark ~85% of available space as busy. mount process w/o the patch couldn't complete in half an hour (according to vmstat it would take ~10-11 hours). With the patch applied mount took ~20 seconds. Lustre-bug-id: https://jira.whamcloud.com/browse/LU-12988 Signed-off-by: Alex Zhuravlev Reviewed-by: Andreas Dilger Reviewed-by: Artem Blagodarenko --- fs/ext4/mballoc.c | 21 ++++++++++++++++++++- 1 file changed, 20 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index c350ae9cb0d9..9a07da53ab7b 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -2177,6 +2177,7 @@ static int ext4_mb_good_group_nolock(struct ext4_allocation_context *ac, { struct ext4_group_info *grp = ext4_get_group_info(ac->ac_sb, group); struct super_block *sb = ac->ac_sb; + struct ext4_sb_info *sbi = EXT4_SB(sb); bool should_lock = ac->ac_flags & EXT4_MB_STRICT_CHECK; ext4_grpblk_t free; int ret = 0; @@ -2195,7 +2196,25 @@ static int ext4_mb_good_group_nolock(struct ext4_allocation_context *ac, /* We only do this if the grp has never been initialized */ if (unlikely(EXT4_MB_GRP_NEED_INIT(grp))) { - ret = ext4_mb_init_group(ac->ac_sb, group, GFP_NOFS); + struct ext4_group_desc *gdp = + ext4_get_group_desc(sb, group, NULL); + int ret; + + /* cr=0/1 is a very optimistic search to find large + * good chunks almost for free. If buddy data is not + * ready, then this optimization makes no sense. But + * we never skip the first block group in a flex_bg, + * since this gets used for metadata block allocation, + * and we want to make sure we locate metadata blocks + * in the first block group in the flex_bg if possible. + */ + if (cr < 2 && + (!sbi->s_log_groups_per_flex || + ((group & ((1 << sbi->s_log_groups_per_flex) - 1)) != 0)) && + !(ext4_has_group_desc_csum(sb) && + (gdp->bg_flags & cpu_to_le16(EXT4_BG_BLOCK_UNINIT)))) + return 0; + ret = ext4_mb_init_group(sb, group, GFP_NOFS); if (ret) return ret; } -- cgit v1.2.3 From bc71726c725767205757821df364acff87f92ac5 Mon Sep 17 00:00:00 2001 From: "zhangyi (F)" Date: Sat, 20 Jun 2020 10:54:23 +0800 Subject: ext4: abort the filesystem if failed to async write metadata buffer There is a risk of filesystem inconsistency if we failed to async write back metadata buffer in the background. Because of current buffer's end io procedure is handled by end_buffer_async_write() in the block layer, and it only clear the buffer's uptodate flag and mark the write_io_error flag, so ext4 cannot detect such failure immediately. In most cases of getting metadata buffer (e.g. ext4_read_inode_bitmap()), although the buffer's data is actually uptodate, it may still read data from disk because the buffer's uptodate flag has been cleared. Finally, it may lead to on-disk filesystem inconsistency if reading old data from the disk successfully and write them out again. This patch detect bdev mapping->wb_err when getting journal's write access and mark the filesystem error if bdev's mapping->wb_err was increased, this could prevent further writing and potential inconsistency. Signed-off-by: zhangyi (F) Suggested-by: Jan Kara Link: https://lore.kernel.org/r/20200620025427.1756360-2-yi.zhang@huawei.com Signed-off-by: Theodore Ts'o --- fs/ext4/ext4.h | 3 +++ fs/ext4/ext4_jbd2.c | 25 +++++++++++++++++++++++++ fs/ext4/super.c | 17 +++++++++++++++++ 3 files changed, 45 insertions(+) (limited to 'fs') diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index 4fba138b8722..26ae31a994a2 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -1603,6 +1603,9 @@ struct ext4_sb_info { #ifdef CONFIG_EXT4_DEBUG unsigned long s_simulate_fail; #endif + /* Record the errseq of the backing block device */ + errseq_t s_bdev_wb_err; + spinlock_t s_bdev_wb_lock; }; static inline struct ext4_sb_info *EXT4_SB(struct super_block *sb) diff --git a/fs/ext4/ext4_jbd2.c b/fs/ext4/ext4_jbd2.c index 0c76cdd44d90..760b9ee49dc0 100644 --- a/fs/ext4/ext4_jbd2.c +++ b/fs/ext4/ext4_jbd2.c @@ -195,6 +195,28 @@ static void ext4_journal_abort_handle(const char *caller, unsigned int line, jbd2_journal_abort_handle(handle); } +static void ext4_check_bdev_write_error(struct super_block *sb) +{ + struct address_space *mapping = sb->s_bdev->bd_inode->i_mapping; + struct ext4_sb_info *sbi = EXT4_SB(sb); + int err; + + /* + * If the block device has write error flag, it may have failed to + * async write out metadata buffers in the background. In this case, + * we could read old data from disk and write it out again, which + * may lead to on-disk filesystem inconsistency. + */ + if (errseq_check(&mapping->wb_err, READ_ONCE(sbi->s_bdev_wb_err))) { + spin_lock(&sbi->s_bdev_wb_lock); + err = errseq_check_and_advance(&mapping->wb_err, &sbi->s_bdev_wb_err); + spin_unlock(&sbi->s_bdev_wb_lock); + if (err) + ext4_error_err(sb, -err, + "Error while async write back metadata"); + } +} + int __ext4_journal_get_write_access(const char *where, unsigned int line, handle_t *handle, struct buffer_head *bh) { @@ -202,6 +224,9 @@ int __ext4_journal_get_write_access(const char *where, unsigned int line, might_sleep(); + if (bh->b_bdev->bd_super) + ext4_check_bdev_write_error(bh->b_bdev->bd_super); + if (ext4_handle_valid(handle)) { err = jbd2_journal_get_write_access(handle, bh); if (err) diff --git a/fs/ext4/super.c b/fs/ext4/super.c index dda967efcbc2..c77b10257b36 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -4765,6 +4765,15 @@ no_journal: } #endif /* CONFIG_QUOTA */ + /* + * Save the original bdev mapping's wb_err value which could be + * used to detect the metadata async write error. + */ + spin_lock_init(&sbi->s_bdev_wb_lock); + if (!sb_rdonly(sb)) + errseq_check_and_advance(&sb->s_bdev->bd_inode->i_mapping->wb_err, + &sbi->s_bdev_wb_err); + sb->s_bdev->bd_super = sb; EXT4_SB(sb)->s_mount_state |= EXT4_ORPHAN_FS; ext4_orphan_cleanup(sb, es); EXT4_SB(sb)->s_mount_state &= ~EXT4_ORPHAN_FS; @@ -5654,6 +5663,14 @@ static int ext4_remount(struct super_block *sb, int *flags, char *data) goto restore_opts; } + /* + * Update the original bdev mapping's wb_err value + * which could be used to detect the metadata async + * write error. + */ + errseq_check_and_advance(&sb->s_bdev->bd_inode->i_mapping->wb_err, + &sbi->s_bdev_wb_err); + /* * Mounting a RDONLY partition read-write, so reread * and store the current valid flag. (It may have -- cgit v1.2.3 From c044f3d8360d2ecf831ba2cc9f08cf9fb2c699fb Mon Sep 17 00:00:00 2001 From: "zhangyi (F)" Date: Sat, 20 Jun 2020 10:54:26 +0800 Subject: jbd2: abort journal if free a async write error metadata buffer If we free a metadata buffer which has been failed to async write out in the background, the jbd2 checkpoint procedure will not detect this failure in jbd2_log_do_checkpoint(), so it may lead to filesystem inconsistency after cleanup journal tail. This patch abort the journal if free a buffer has write_io_error flag to prevent potential further inconsistency. Signed-off-by: zhangyi (F) Link: https://lore.kernel.org/r/20200620025427.1756360-5-yi.zhang@huawei.com Signed-off-by: Theodore Ts'o --- fs/jbd2/transaction.c | 16 ++++++++++++++++ 1 file changed, 16 insertions(+) (limited to 'fs') diff --git a/fs/jbd2/transaction.c b/fs/jbd2/transaction.c index e65e0aca2826..6250c9faa4cb 100644 --- a/fs/jbd2/transaction.c +++ b/fs/jbd2/transaction.c @@ -2120,6 +2120,7 @@ int jbd2_journal_try_to_free_buffers(journal_t *journal, { struct buffer_head *head; struct buffer_head *bh; + bool has_write_io_error = false; int ret = 0; J_ASSERT(PageLocked(page)); @@ -2144,11 +2145,26 @@ int jbd2_journal_try_to_free_buffers(journal_t *journal, jbd2_journal_put_journal_head(jh); if (buffer_jbd(bh)) goto busy; + + /* + * If we free a metadata buffer which has been failed to + * write out, the jbd2 checkpoint procedure will not detect + * this failure and may lead to filesystem inconsistency + * after cleanup journal tail. + */ + if (buffer_write_io_error(bh)) { + pr_err("JBD2: Error while async write back metadata bh %llu.", + (unsigned long long)bh->b_blocknr); + has_write_io_error = true; + } } while ((bh = bh->b_this_page) != head); ret = try_to_free_buffers(page); busy: + if (has_write_io_error) + jbd2_journal_abort(journal, -EIO); + return ret; } -- cgit v1.2.3 From 529a781ee07aaa58be8164d75ba5998eb7dd216c Mon Sep 17 00:00:00 2001 From: "zhangyi (F)" Date: Sat, 20 Jun 2020 10:54:27 +0800 Subject: jbd2: remove unused parameter in jbd2_journal_try_to_free_buffers() Parameter gfp_mask in jbd2_journal_try_to_free_buffers() is no longer used after commit <536fc240e7147> ("jbd2: clean up jbd2_journal_try_to_free_buffers()"), so just remove it. Signed-off-by: zhangyi (F) Link: https://lore.kernel.org/r/20200620025427.1756360-6-yi.zhang@huawei.com Signed-off-by: Theodore Ts'o --- fs/ext4/inode.c | 2 +- fs/ext4/super.c | 4 ++-- fs/jbd2/transaction.c | 7 +------ include/linux/jbd2.h | 2 +- 4 files changed, 5 insertions(+), 10 deletions(-) (limited to 'fs') diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 6187c8880c02..551a1056870d 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -3288,7 +3288,7 @@ static int ext4_releasepage(struct page *page, gfp_t wait) if (PageChecked(page)) return 0; if (journal) - return jbd2_journal_try_to_free_buffers(journal, page, wait); + return jbd2_journal_try_to_free_buffers(journal, page); else return try_to_free_buffers(page); } diff --git a/fs/ext4/super.c b/fs/ext4/super.c index c77b10257b36..8c00f0a09f4d 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -1288,8 +1288,8 @@ static int bdev_try_to_free_page(struct super_block *sb, struct page *page, if (!page_has_buffers(page)) return 0; if (journal) - return jbd2_journal_try_to_free_buffers(journal, page, - wait & ~__GFP_DIRECT_RECLAIM); + return jbd2_journal_try_to_free_buffers(journal, page); + return try_to_free_buffers(page); } diff --git a/fs/jbd2/transaction.c b/fs/jbd2/transaction.c index 6250c9faa4cb..43985738aa86 100644 --- a/fs/jbd2/transaction.c +++ b/fs/jbd2/transaction.c @@ -2081,10 +2081,6 @@ out: * int jbd2_journal_try_to_free_buffers() - try to free page buffers. * @journal: journal for operation * @page: to try and free - * @gfp_mask: we use the mask to detect how hard should we try to release - * buffers. If __GFP_DIRECT_RECLAIM and __GFP_FS is set, we wait for commit - * code to release the buffers. - * * * For all the buffers on this page, * if they are fully written out ordered data, move them onto BUF_CLEAN @@ -2115,8 +2111,7 @@ out: * * Return 0 on failure, 1 on success */ -int jbd2_journal_try_to_free_buffers(journal_t *journal, - struct page *page, gfp_t gfp_mask) +int jbd2_journal_try_to_free_buffers(journal_t *journal, struct page *page) { struct buffer_head *head; struct buffer_head *bh; diff --git a/include/linux/jbd2.h b/include/linux/jbd2.h index d56128df2aff..a756a4cdf939 100644 --- a/include/linux/jbd2.h +++ b/include/linux/jbd2.h @@ -1380,7 +1380,7 @@ extern int jbd2_journal_dirty_metadata (handle_t *, struct buffer_head *); extern int jbd2_journal_forget (handle_t *, struct buffer_head *); extern int jbd2_journal_invalidatepage(journal_t *, struct page *, unsigned int, unsigned int); -extern int jbd2_journal_try_to_free_buffers(journal_t *, struct page *, gfp_t); +extern int jbd2_journal_try_to_free_buffers(journal_t *journal, struct page *page); extern int jbd2_journal_stop(handle_t *); extern int jbd2_journal_flush (journal_t *); extern void jbd2_journal_lock_updates (journal_t *); -- cgit v1.2.3 From ab74c7b23f3770935016e3eb3ecdf1e42b73efaa Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Wed, 15 Jul 2020 11:48:55 -0400 Subject: ext4: indicate via a block bitmap read is prefetched via a tracepoint Modify the ext4_read_block_bitmap_load tracepoint so that it tells us whether a block bitmap is being prefetched. Signed-off-by: Theodore Ts'o Reviewed-by: Artem Blagodarenko --- fs/ext4/balloc.c | 2 +- include/trace/events/ext4.h | 24 ++++++++++++++++++++---- 2 files changed, 21 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/ext4/balloc.c b/fs/ext4/balloc.c index 1e2b1b4093aa..48c3df47748d 100644 --- a/fs/ext4/balloc.c +++ b/fs/ext4/balloc.c @@ -494,7 +494,7 @@ ext4_read_block_bitmap_nowait(struct super_block *sb, ext4_group_t block_group, * submit the buffer_head for reading */ set_buffer_new(bh); - trace_ext4_read_block_bitmap_load(sb, block_group); + trace_ext4_read_block_bitmap_load(sb, block_group, ignore_locked); bh->b_end_io = ext4_end_bitmap_read; get_bh(bh); submit_bh(REQ_OP_READ, REQ_META | REQ_PRIO | diff --git a/include/trace/events/ext4.h b/include/trace/events/ext4.h index cc41d692ae8e..cbcd2e1a608d 100644 --- a/include/trace/events/ext4.h +++ b/include/trace/events/ext4.h @@ -1312,18 +1312,34 @@ DEFINE_EVENT(ext4__bitmap_load, ext4_mb_buddy_bitmap_load, TP_ARGS(sb, group) ); -DEFINE_EVENT(ext4__bitmap_load, ext4_read_block_bitmap_load, +DEFINE_EVENT(ext4__bitmap_load, ext4_load_inode_bitmap, TP_PROTO(struct super_block *sb, unsigned long group), TP_ARGS(sb, group) ); -DEFINE_EVENT(ext4__bitmap_load, ext4_load_inode_bitmap, +TRACE_EVENT(ext4_read_block_bitmap_load, + TP_PROTO(struct super_block *sb, unsigned long group, bool prefetch), - TP_PROTO(struct super_block *sb, unsigned long group), + TP_ARGS(sb, group, prefetch), - TP_ARGS(sb, group) + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( __u32, group ) + __field( bool, prefetch ) + + ), + + TP_fast_assign( + __entry->dev = sb->s_dev; + __entry->group = group; + __entry->prefetch = prefetch; + ), + + TP_printk("dev %d,%d group %u prefetch %d", + MAJOR(__entry->dev), MINOR(__entry->dev), + __entry->group, __entry->prefetch) ); TRACE_EVENT(ext4_direct_IO_enter, -- cgit v1.2.3 From 3d392b2676bf3199863a1e5efb2c087ad9d442a4 Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Fri, 17 Jul 2020 00:14:40 -0400 Subject: ext4: add prefetch_block_bitmaps mount option For file systems where we can afford to keep the buddy bitmaps cached, we can speed up initial writes to large file systems by starting to load the block allocation bitmaps as soon as the file system is mounted. This won't work well for _super_ large file systems, or memory constrained systems, so we only enable this when it is requested via a mount option. Addresses-Google-Bug: 159488342 Signed-off-by: Theodore Ts'o Reviewed-by: Andreas Dilger --- fs/ext4/ext4.h | 15 +++++++++++- fs/ext4/mballoc.c | 10 ++++---- fs/ext4/super.c | 57 ++++++++++++++++++++++++++++++++------------- include/trace/events/ext4.h | 44 ++++++++++++++++++++++++++++++++++ 4 files changed, 103 insertions(+), 23 deletions(-) (limited to 'fs') diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index 26ae31a994a2..e5f0c66a6156 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -1188,6 +1188,7 @@ struct ext4_inode_info { #define EXT4_MOUNT_JOURNAL_CHECKSUM 0x800000 /* Journal checksums */ #define EXT4_MOUNT_JOURNAL_ASYNC_COMMIT 0x1000000 /* Journal Async Commit */ #define EXT4_MOUNT_WARN_ON_ERROR 0x2000000 /* Trigger WARN_ON on error */ +#define EXT4_MOUNT_PREFETCH_BLOCK_BITMAPS 0x4000000 #define EXT4_MOUNT_DELALLOC 0x8000000 /* Delalloc support */ #define EXT4_MOUNT_DATA_ERR_ABORT 0x10000000 /* Abort on file data write */ #define EXT4_MOUNT_BLOCK_VALIDITY 0x20000000 /* Block validity checking */ @@ -2334,9 +2335,15 @@ struct ext4_lazy_init { struct mutex li_list_mtx; }; +enum ext4_li_mode { + EXT4_LI_MODE_PREFETCH_BBITMAP, + EXT4_LI_MODE_ITABLE, +}; + struct ext4_li_request { struct super_block *lr_super; - struct ext4_sb_info *lr_sbi; + enum ext4_li_mode lr_mode; + ext4_group_t lr_first_not_zeroed; ext4_group_t lr_next_group; struct list_head lr_request; unsigned long lr_next_sched; @@ -2676,6 +2683,12 @@ extern int ext4_mb_reserve_blocks(struct super_block *, int); extern void ext4_discard_preallocations(struct inode *); extern int __init ext4_init_mballoc(void); extern void ext4_exit_mballoc(void); +extern ext4_group_t ext4_mb_prefetch(struct super_block *sb, + ext4_group_t group, + unsigned int nr, int *cnt); +extern void ext4_mb_prefetch_fini(struct super_block *sb, ext4_group_t group, + unsigned int nr); + extern void ext4_free_blocks(handle_t *handle, struct inode *inode, struct buffer_head *bh, ext4_fsblk_t block, unsigned long count, int flags); diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index 9a07da53ab7b..8ecd49dd3906 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -2232,9 +2232,8 @@ out: * Start prefetching @nr block bitmaps starting at @group. * Return the next group which needs to be prefetched. */ -static ext4_group_t -ext4_mb_prefetch(struct super_block *sb, ext4_group_t group, - unsigned int nr, int *cnt) +ext4_group_t ext4_mb_prefetch(struct super_block *sb, ext4_group_t group, + unsigned int nr, int *cnt) { ext4_group_t ngroups = ext4_get_groups_count(sb); struct buffer_head *bh; @@ -2284,9 +2283,8 @@ ext4_mb_prefetch(struct super_block *sb, ext4_group_t group, * waiting for the block allocation bitmap read to finish when * ext4_mb_prefetch_fini is called from ext4_mb_regular_allocator(). */ -static void -ext4_mb_prefetch_fini(struct super_block *sb, ext4_group_t group, - unsigned int nr) +void ext4_mb_prefetch_fini(struct super_block *sb, ext4_group_t group, + unsigned int nr) { while (nr-- > 0) { struct ext4_group_desc *gdp = ext4_get_group_desc(sb, group, diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 8c00f0a09f4d..1cbe3f248faf 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -1521,6 +1521,7 @@ enum { Opt_dioread_nolock, Opt_dioread_lock, Opt_discard, Opt_nodiscard, Opt_init_itable, Opt_noinit_itable, Opt_max_dir_size_kb, Opt_nojournal_checksum, Opt_nombcache, + Opt_prefetch_block_bitmaps, }; static const match_table_t tokens = { @@ -1612,6 +1613,7 @@ static const match_table_t tokens = { {Opt_test_dummy_encryption, "test_dummy_encryption"}, {Opt_nombcache, "nombcache"}, {Opt_nombcache, "no_mbcache"}, /* for backward compatibility */ + {Opt_prefetch_block_bitmaps, "prefetch_block_bitmaps"}, {Opt_removed, "check=none"}, /* mount option from ext2/3 */ {Opt_removed, "nocheck"}, /* mount option from ext2/3 */ {Opt_removed, "reservation"}, /* mount option from ext2/3 */ @@ -1829,6 +1831,8 @@ static const struct mount_opts { {Opt_max_dir_size_kb, 0, MOPT_GTE0}, {Opt_test_dummy_encryption, 0, MOPT_STRING}, {Opt_nombcache, EXT4_MOUNT_NO_MBCACHE, MOPT_SET}, + {Opt_prefetch_block_bitmaps, EXT4_MOUNT_PREFETCH_BLOCK_BITMAPS, + MOPT_SET}, {Opt_err, 0, 0} }; @@ -3201,15 +3205,34 @@ static void print_daily_error_info(struct timer_list *t) static int ext4_run_li_request(struct ext4_li_request *elr) { struct ext4_group_desc *gdp = NULL; - ext4_group_t group, ngroups; - struct super_block *sb; + struct super_block *sb = elr->lr_super; + ext4_group_t ngroups = EXT4_SB(sb)->s_groups_count; + ext4_group_t group = elr->lr_next_group; unsigned long timeout = 0; + unsigned int prefetch_ios = 0; int ret = 0; - sb = elr->lr_super; - ngroups = EXT4_SB(sb)->s_groups_count; + if (elr->lr_mode == EXT4_LI_MODE_PREFETCH_BBITMAP) { + elr->lr_next_group = ext4_mb_prefetch(sb, group, + EXT4_SB(sb)->s_mb_prefetch, &prefetch_ios); + if (prefetch_ios) + ext4_mb_prefetch_fini(sb, elr->lr_next_group, + prefetch_ios); + trace_ext4_prefetch_bitmaps(sb, group, elr->lr_next_group, + prefetch_ios); + if (group >= elr->lr_next_group) { + ret = 1; + if (elr->lr_first_not_zeroed != ngroups && + !sb_rdonly(sb) && test_opt(sb, INIT_INODE_TABLE)) { + elr->lr_next_group = elr->lr_first_not_zeroed; + elr->lr_mode = EXT4_LI_MODE_ITABLE; + ret = 0; + } + } + return ret; + } - for (group = elr->lr_next_group; group < ngroups; group++) { + for (; group < ngroups; group++) { gdp = ext4_get_group_desc(sb, group, NULL); if (!gdp) { ret = 1; @@ -3227,9 +3250,10 @@ static int ext4_run_li_request(struct ext4_li_request *elr) timeout = jiffies; ret = ext4_init_inode_table(sb, group, elr->lr_timeout ? 0 : 1); + trace_ext4_lazy_itable_init(sb, group); if (elr->lr_timeout == 0) { timeout = (jiffies - timeout) * - elr->lr_sbi->s_li_wait_mult; + EXT4_SB(elr->lr_super)->s_li_wait_mult; elr->lr_timeout = timeout; } elr->lr_next_sched = jiffies + elr->lr_timeout; @@ -3244,15 +3268,11 @@ static int ext4_run_li_request(struct ext4_li_request *elr) */ static void ext4_remove_li_request(struct ext4_li_request *elr) { - struct ext4_sb_info *sbi; - if (!elr) return; - sbi = elr->lr_sbi; - list_del(&elr->lr_request); - sbi->s_li_request = NULL; + EXT4_SB(elr->lr_super)->s_li_request = NULL; kfree(elr); } @@ -3461,7 +3481,6 @@ static int ext4_li_info_new(void) static struct ext4_li_request *ext4_li_request_new(struct super_block *sb, ext4_group_t start) { - struct ext4_sb_info *sbi = EXT4_SB(sb); struct ext4_li_request *elr; elr = kzalloc(sizeof(*elr), GFP_KERNEL); @@ -3469,8 +3488,13 @@ static struct ext4_li_request *ext4_li_request_new(struct super_block *sb, return NULL; elr->lr_super = sb; - elr->lr_sbi = sbi; - elr->lr_next_group = start; + elr->lr_first_not_zeroed = start; + if (test_opt(sb, PREFETCH_BLOCK_BITMAPS)) + elr->lr_mode = EXT4_LI_MODE_PREFETCH_BBITMAP; + else { + elr->lr_mode = EXT4_LI_MODE_ITABLE; + elr->lr_next_group = start; + } /* * Randomize first schedule time of the request to @@ -3500,8 +3524,9 @@ int ext4_register_li_request(struct super_block *sb, goto out; } - if (first_not_zeroed == ngroups || sb_rdonly(sb) || - !test_opt(sb, INIT_INODE_TABLE)) + if (!test_opt(sb, PREFETCH_BLOCK_BITMAPS) && + (first_not_zeroed == ngroups || sb_rdonly(sb) || + !test_opt(sb, INIT_INODE_TABLE))) goto out; elr = ext4_li_request_new(sb, first_not_zeroed); diff --git a/include/trace/events/ext4.h b/include/trace/events/ext4.h index cbcd2e1a608d..8008d2e116b9 100644 --- a/include/trace/events/ext4.h +++ b/include/trace/events/ext4.h @@ -2742,6 +2742,50 @@ TRACE_EVENT(ext4_error, __entry->function, __entry->line) ); +TRACE_EVENT(ext4_prefetch_bitmaps, + TP_PROTO(struct super_block *sb, ext4_group_t group, + ext4_group_t next, unsigned int prefetch_ios), + + TP_ARGS(sb, group, next, prefetch_ios), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( __u32, group ) + __field( __u32, next ) + __field( __u32, ios ) + ), + + TP_fast_assign( + __entry->dev = sb->s_dev; + __entry->group = group; + __entry->next = next; + __entry->ios = prefetch_ios; + ), + + TP_printk("dev %d,%d group %u next %u ios %u", + MAJOR(__entry->dev), MINOR(__entry->dev), + __entry->group, __entry->next, __entry->ios) +); + +TRACE_EVENT(ext4_lazy_itable_init, + TP_PROTO(struct super_block *sb, ext4_group_t group), + + TP_ARGS(sb, group), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( __u32, group ) + ), + + TP_fast_assign( + __entry->dev = sb->s_dev; + __entry->group = group; + ), + + TP_printk("dev %d,%d group %u", + MAJOR(__entry->dev), MINOR(__entry->dev), __entry->group) +); + #endif /* _TRACE_EXT4_H */ /* This part must be outside protection */ -- cgit v1.2.3 From 60ed633f51d0c675150a117d96a45e78c3613f91 Mon Sep 17 00:00:00 2001 From: Xianting Tian Date: Sat, 18 Jul 2020 08:57:37 -0400 Subject: jbd2: fix incorrect code style Remove unnecessary blank. Signed-off-by: Xianting Tian Reviewed-by: Jan Kara Link: https://lore.kernel.org/r/1595077057-8048-1-git-send-email-xianting_tian@126.com Signed-off-by: Theodore Ts'o --- fs/jbd2/journal.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c index 5493a0da23dd..17fdc482f554 100644 --- a/fs/jbd2/journal.c +++ b/fs/jbd2/journal.c @@ -1285,7 +1285,7 @@ journal_t *jbd2_journal_init_inode(struct inode *inode) * superblock as being NULL to prevent the journal destroy from writing * back a bogus superblock. */ -static void journal_fail_superblock (journal_t *journal) +static void journal_fail_superblock(journal_t *journal) { struct buffer_head *bh = journal->j_sb_buffer; brelse(bh); @@ -1817,7 +1817,7 @@ int jbd2_journal_destroy(journal_t *journal) /** - *int jbd2_journal_check_used_features () - Check if features specified are used. + *int jbd2_journal_check_used_features() - Check if features specified are used. * @journal: Journal to check. * @compat: bitmask of compatible features * @ro: bitmask of features that force read-only mount @@ -1827,7 +1827,7 @@ int jbd2_journal_destroy(journal_t *journal) * features. Return true (non-zero) if it does. **/ -int jbd2_journal_check_used_features (journal_t *journal, unsigned long compat, +int jbd2_journal_check_used_features(journal_t *journal, unsigned long compat, unsigned long ro, unsigned long incompat) { journal_superblock_t *sb; @@ -1862,7 +1862,7 @@ int jbd2_journal_check_used_features (journal_t *journal, unsigned long compat, * all of a given set of features on this journal. Return true * (non-zero) if it can. */ -int jbd2_journal_check_available_features (journal_t *journal, unsigned long compat, +int jbd2_journal_check_available_features(journal_t *journal, unsigned long compat, unsigned long ro, unsigned long incompat) { if (!compat && !ro && !incompat) @@ -1884,7 +1884,7 @@ int jbd2_journal_check_available_features (journal_t *journal, unsigned long com } /** - * int jbd2_journal_set_features () - Mark a given journal feature in the superblock + * int jbd2_journal_set_features() - Mark a given journal feature in the superblock * @journal: Journal to act on. * @compat: bitmask of compatible features * @ro: bitmask of features that force read-only mount @@ -1895,7 +1895,7 @@ int jbd2_journal_check_available_features (journal_t *journal, unsigned long com * */ -int jbd2_journal_set_features (journal_t *journal, unsigned long compat, +int jbd2_journal_set_features(journal_t *journal, unsigned long compat, unsigned long ro, unsigned long incompat) { #define INCOMPAT_FEATURE_ON(f) \ -- cgit v1.2.3 From f25391ebb475d3ffb3aa61bb90e3594c841749ef Mon Sep 17 00:00:00 2001 From: Lukas Czerner Date: Thu, 23 Jul 2020 17:05:26 +0200 Subject: ext4: handle option set by mount flags correctly Currently there is a problem with mount options that can be both set by vfs using mount flags or by a string parsing in ext4. i_version/iversion options gets lost after remount, for example $ mount -o i_version /dev/pmem0 /mnt $ grep pmem0 /proc/self/mountinfo | grep i_version 310 95 259:0 / /mnt rw,relatime shared:163 - ext4 /dev/pmem0 rw,seclabel,i_version $ mount -o remount,ro /mnt $ grep pmem0 /proc/self/mountinfo | grep i_version nolazytime gets ignored by ext4 on remount, for example $ mount -o lazytime /dev/pmem0 /mnt $ grep pmem0 /proc/self/mountinfo | grep lazytime 310 95 259:0 / /mnt rw,relatime shared:163 - ext4 /dev/pmem0 rw,lazytime,seclabel $ mount -o remount,nolazytime /mnt $ grep pmem0 /proc/self/mountinfo | grep lazytime 310 95 259:0 / /mnt rw,relatime shared:163 - ext4 /dev/pmem0 rw,lazytime,seclabel Fix it by applying the SB_LAZYTIME and SB_I_VERSION flags from *flags to s_flags before we parse the option and use the resulting state of the same flags in *flags at the end of successful remount. Signed-off-by: Lukas Czerner Reviewed-by: Ritesh Harjani Link: https://lore.kernel.org/r/20200723150526.19931-1-lczerner@redhat.com Signed-off-by: Theodore Ts'o --- fs/ext4/super.c | 21 ++++++++++++++++----- 1 file changed, 16 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 1cbe3f248faf..a6de88fc7060 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -5512,7 +5512,7 @@ static int ext4_remount(struct super_block *sb, int *flags, char *data) { struct ext4_super_block *es; struct ext4_sb_info *sbi = EXT4_SB(sb); - unsigned long old_sb_flags; + unsigned long old_sb_flags, vfs_flags; struct ext4_mount_options old_opts; int enable_quota = 0; ext4_group_t g; @@ -5555,6 +5555,14 @@ static int ext4_remount(struct super_block *sb, int *flags, char *data) if (sbi->s_journal && sbi->s_journal->j_task->io_context) journal_ioprio = sbi->s_journal->j_task->io_context->ioprio; + /* + * Some options can be enabled by ext4 and/or by VFS mount flag + * either way we need to make sure it matches in both *flags and + * s_flags. Copy those selected flags from *flags to s_flags + */ + vfs_flags = SB_LAZYTIME | SB_I_VERSION; + sb->s_flags = (sb->s_flags & ~vfs_flags) | (*flags & vfs_flags); + if (!parse_options(data, sb, NULL, &journal_ioprio, 1)) { err = -EINVAL; goto restore_opts; @@ -5608,9 +5616,6 @@ static int ext4_remount(struct super_block *sb, int *flags, char *data) set_task_ioprio(sbi->s_journal->j_task, journal_ioprio); } - if (*flags & SB_LAZYTIME) - sb->s_flags |= SB_LAZYTIME; - if ((bool)(*flags & SB_RDONLY) != sb_rdonly(sb)) { if (sbi->s_mount_flags & EXT4_MF_FS_ABORTED) { err = -EROFS; @@ -5758,7 +5763,13 @@ static int ext4_remount(struct super_block *sb, int *flags, char *data) } #endif - *flags = (*flags & ~SB_LAZYTIME) | (sb->s_flags & SB_LAZYTIME); + /* + * Some options can be enabled by ext4 and/or by VFS mount flag + * either way we need to make sure it matches in both *flags and + * s_flags. Copy those selected flags from s_flags to *flags + */ + *flags = (*flags & ~vfs_flags) | (sb->s_flags & vfs_flags); + ext4_msg(sb, KERN_INFO, "re-mounted. Opts: %s", orig_data); kfree(orig_data); return 0; -- cgit v1.2.3 From 6dbd300129089ad57de20be89ab7c122b36a4c26 Mon Sep 17 00:00:00 2001 From: Shijie Luo Date: Thu, 23 Jul 2020 23:29:54 -0400 Subject: ext4: remove some redundant function declarations ext4 update feature functions do not exist now, remove these useless function declarations. Signed-off-by: Shijie Luo Reviewed-by: Ritesh Harjani Link: https://lore.kernel.org/r/20200724032954.22097-1-luoshijie1@huawei.com Signed-off-by: Theodore Ts'o --- fs/ext4/ext4.h | 6 ------ 1 file changed, 6 deletions(-) (limited to 'fs') diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index e5f0c66a6156..32260ceb9cb5 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -2959,12 +2959,6 @@ do { \ #endif -extern int ext4_update_compat_feature(handle_t *handle, struct super_block *sb, - __u32 compat); -extern int ext4_update_rocompat_feature(handle_t *handle, - struct super_block *sb, __u32 rocompat); -extern int ext4_update_incompat_feature(handle_t *handle, - struct super_block *sb, __u32 incompat); extern ext4_fsblk_t ext4_block_bitmap(struct super_block *sb, struct ext4_group_desc *bg); extern ext4_fsblk_t ext4_inode_bitmap(struct super_block *sb, -- cgit v1.2.3 From 1cf006ed19a887c22e085189c8b4a3cbf60d2246 Mon Sep 17 00:00:00 2001 From: Dmitry Monakhov Date: Sat, 25 Jul 2020 12:33:13 +0000 Subject: ext4: export msg_count and warning_count via sysfs This numbers can be analized by system automation similar to errors_count. In ideal world it would be nice to have separate counters for different log-levels, but this makes this patch too intrusive. Signed-off-by: Dmitry Monakhov Link: https://lore.kernel.org/r/20200725123313.4467-1-dmtrmonakhov@yandex-team.ru Signed-off-by: Theodore Ts'o --- fs/ext4/ext4.h | 2 ++ fs/ext4/super.c | 12 +++++++++--- fs/ext4/sysfs.c | 7 +++++++ 3 files changed, 18 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index 32260ceb9cb5..c73b9a1959a9 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -1591,6 +1591,8 @@ struct ext4_sb_info { struct ratelimit_state s_err_ratelimit_state; struct ratelimit_state s_warning_ratelimit_state; struct ratelimit_state s_msg_ratelimit_state; + atomic_t s_warning_count; + atomic_t s_msg_count; /* Encryption context for '-o test_dummy_encryption' */ struct fscrypt_dummy_context s_dummy_enc_ctx; diff --git a/fs/ext4/super.c b/fs/ext4/super.c index a6de88fc7060..7deb2b75fcd0 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -744,6 +744,7 @@ void __ext4_msg(struct super_block *sb, struct va_format vaf; va_list args; + atomic_inc(&EXT4_SB(sb)->s_msg_count); if (!___ratelimit(&(EXT4_SB(sb)->s_msg_ratelimit_state), "EXT4-fs")) return; @@ -754,9 +755,12 @@ void __ext4_msg(struct super_block *sb, va_end(args); } -#define ext4_warning_ratelimit(sb) \ - ___ratelimit(&(EXT4_SB(sb)->s_warning_ratelimit_state), \ - "EXT4-fs warning") +static int ext4_warning_ratelimit(struct super_block *sb) +{ + atomic_inc(&EXT4_SB(sb)->s_warning_count); + return ___ratelimit(&(EXT4_SB(sb)->s_warning_ratelimit_state), + "EXT4-fs warning"); +} void __ext4_warning(struct super_block *sb, const char *function, unsigned int line, const char *fmt, ...) @@ -4840,6 +4844,8 @@ no_journal: ratelimit_state_init(&sbi->s_err_ratelimit_state, 5 * HZ, 10); ratelimit_state_init(&sbi->s_warning_ratelimit_state, 5 * HZ, 10); ratelimit_state_init(&sbi->s_msg_ratelimit_state, 5 * HZ, 10); + atomic_set(&sbi->s_warning_count, 0); + atomic_set(&sbi->s_msg_count, 0); kfree(orig_data); return 0; diff --git a/fs/ext4/sysfs.c b/fs/ext4/sysfs.c index 31e0db726d21..7fee11cc30e7 100644 --- a/fs/ext4/sysfs.c +++ b/fs/ext4/sysfs.c @@ -189,6 +189,9 @@ static struct ext4_attr ext4_attr_##_name = { \ #define EXT4_RW_ATTR_SBI_UL(_name,_elname) \ EXT4_ATTR_OFFSET(_name, 0644, pointer_ul, ext4_sb_info, _elname) +#define EXT4_RO_ATTR_SBI_ATOMIC(_name,_elname) \ + EXT4_ATTR_OFFSET(_name, 0444, pointer_atomic, ext4_sb_info, _elname) + #define EXT4_ATTR_PTR(_name,_mode,_id,_ptr) \ static struct ext4_attr ext4_attr_##_name = { \ .attr = {.name = __stringify(_name), .mode = _mode }, \ @@ -226,6 +229,8 @@ EXT4_RW_ATTR_SBI_UI(msg_ratelimit_burst, s_msg_ratelimit_state.burst); #ifdef CONFIG_EXT4_DEBUG EXT4_RW_ATTR_SBI_UL(simulate_fail, s_simulate_fail); #endif +EXT4_RO_ATTR_SBI_ATOMIC(warning_count, s_warning_count); +EXT4_RO_ATTR_SBI_ATOMIC(msg_count, s_msg_count); EXT4_RO_ATTR_ES_UI(errors_count, s_error_count); EXT4_RO_ATTR_ES_U8(first_error_errcode, s_first_error_errcode); EXT4_RO_ATTR_ES_U8(last_error_errcode, s_last_error_errcode); @@ -269,6 +274,8 @@ static struct attribute *ext4_attrs[] = { ATTR_LIST(msg_ratelimit_interval_ms), ATTR_LIST(msg_ratelimit_burst), ATTR_LIST(errors_count), + ATTR_LIST(warning_count), + ATTR_LIST(msg_count), ATTR_LIST(first_error_ino), ATTR_LIST(last_error_ino), ATTR_LIST(first_error_block), -- cgit v1.2.3 From 9375ac770cda6bdff829f43868d0ebcaba41f675 Mon Sep 17 00:00:00 2001 From: brookxu Date: Mon, 27 Jul 2020 09:54:14 +0800 Subject: ext4: delete the invalid BUGON in ext4_mb_load_buddy_gfp() Delete the invalid BUGON in ext4_mb_load_buddy_gfp(), the previous code has already judged whether page is NULL. Signed-off-by: Chunguang Xu Reviewed-by: Ritesh Harjani Link: https://lore.kernel.org/r/ad68e8a2-5ec3-5beb-537f-f3e53f55367a@gmail.com Signed-off-by: Theodore Ts'o --- fs/ext4/mballoc.c | 3 --- 1 file changed, 3 deletions(-) (limited to 'fs') diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index 8ecd49dd3906..19101941dadd 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -1279,9 +1279,6 @@ ext4_mb_load_buddy_gfp(struct super_block *sb, ext4_group_t group, e4b->bd_buddy_page = page; e4b->bd_buddy = page_address(page) + (poff * sb->s_blocksize); - BUG_ON(e4b->bd_bitmap_page == NULL); - BUG_ON(e4b->bd_buddy_page == NULL); - return 0; err: -- cgit v1.2.3 From d176b1f62f242ab259ff665a26fbac69db1aecba Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Tue, 28 Jul 2020 15:04:32 +0200 Subject: ext4: handle error of ext4_setup_system_zone() on remount ext4_setup_system_zone() can fail. Handle the failure in ext4_remount(). Reviewed-by: Lukas Czerner Signed-off-by: Jan Kara Link: https://lore.kernel.org/r/20200728130437.7804-2-jack@suse.cz Signed-off-by: Theodore Ts'o --- fs/ext4/super.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 7deb2b75fcd0..6507e30e58b5 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -5747,7 +5747,10 @@ static int ext4_remount(struct super_block *sb, int *flags, char *data) ext4_register_li_request(sb, first_not_zeroed); } - ext4_setup_system_zone(sb); + err = ext4_setup_system_zone(sb); + if (err) + goto restore_opts; + if (sbi->s_journal == NULL && !(old_sb_flags & SB_RDONLY)) { err = ext4_commit_super(sb, 1); if (err) -- cgit v1.2.3 From bf9a379d0980e7413d94cb18dac73db2bfc5f470 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Tue, 28 Jul 2020 15:04:33 +0200 Subject: ext4: don't allow overlapping system zones Currently, add_system_zone() just silently merges two added system zones that overlap. However the overlap should not happen and it generally suggests that some unrelated metadata overlap which indicates the fs is corrupted. We should have caught such problems earlier (e.g. in ext4_check_descriptors()) but add this check as another line of defense. In later patch we also use this for stricter checking of journal inode extent tree. Reviewed-by: Lukas Czerner Signed-off-by: Jan Kara Link: https://lore.kernel.org/r/20200728130437.7804-3-jack@suse.cz Signed-off-by: Theodore Ts'o --- fs/ext4/block_validity.c | 36 +++++++++++++----------------------- 1 file changed, 13 insertions(+), 23 deletions(-) (limited to 'fs') diff --git a/fs/ext4/block_validity.c b/fs/ext4/block_validity.c index 16e9b2fda03a..b394a50ebbe3 100644 --- a/fs/ext4/block_validity.c +++ b/fs/ext4/block_validity.c @@ -68,7 +68,7 @@ static int add_system_zone(struct ext4_system_blocks *system_blks, ext4_fsblk_t start_blk, unsigned int count) { - struct ext4_system_zone *new_entry = NULL, *entry; + struct ext4_system_zone *new_entry, *entry; struct rb_node **n = &system_blks->root.rb_node, *node; struct rb_node *parent = NULL, *new_node = NULL; @@ -79,30 +79,20 @@ static int add_system_zone(struct ext4_system_blocks *system_blks, n = &(*n)->rb_left; else if (start_blk >= (entry->start_blk + entry->count)) n = &(*n)->rb_right; - else { - if (start_blk + count > (entry->start_blk + - entry->count)) - entry->count = (start_blk + count - - entry->start_blk); - new_node = *n; - new_entry = rb_entry(new_node, struct ext4_system_zone, - node); - break; - } + else /* Unexpected overlap of system zones. */ + return -EFSCORRUPTED; } - if (!new_entry) { - new_entry = kmem_cache_alloc(ext4_system_zone_cachep, - GFP_KERNEL); - if (!new_entry) - return -ENOMEM; - new_entry->start_blk = start_blk; - new_entry->count = count; - new_node = &new_entry->node; - - rb_link_node(new_node, parent, n); - rb_insert_color(new_node, &system_blks->root); - } + new_entry = kmem_cache_alloc(ext4_system_zone_cachep, + GFP_KERNEL); + if (!new_entry) + return -ENOMEM; + new_entry->start_blk = start_blk; + new_entry->count = count; + new_node = &new_entry->node; + + rb_link_node(new_node, parent, n); + rb_insert_color(new_node, &system_blks->root); /* Can we merge to the left? */ node = rb_prev(new_node); -- cgit v1.2.3 From ce9f24cccdc019229b70a5c15e2b09ad9c0ab5d1 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Tue, 28 Jul 2020 15:04:34 +0200 Subject: ext4: check journal inode extents more carefully Currently, system zones just track ranges of block, that are "important" fs metadata (bitmaps, group descriptors, journal blocks, etc.). This however complicates how extent tree (or indirect blocks) can be checked for inodes that actually track such metadata - currently the journal inode but arguably we should be treating quota files or resize inode similarly. We cannot run __ext4_ext_check() on such metadata inodes when loading their extents as that would immediately trigger the validity checks and so we just hack around that and special-case the journal inode. This however leads to a situation that a journal inode which has extent tree of depth at least one can have invalid extent tree that gets unnoticed until ext4_cache_extents() crashes. To overcome this limitation, track inode number each system zone belongs to (0 is used for zones not belonging to any inode). We can then verify inode number matches the expected one when verifying extent tree and thus avoid the false errors. With this there's no need to to special-case journal inode during extent tree checking anymore so remove it. Fixes: 0a944e8a6c66 ("ext4: don't perform block validity checks on the journal inode") Reported-by: Wolfgang Frisch Reviewed-by: Lukas Czerner Signed-off-by: Jan Kara Link: https://lore.kernel.org/r/20200728130437.7804-4-jack@suse.cz Signed-off-by: Theodore Ts'o --- fs/ext4/block_validity.c | 51 ++++++++++++++++++++++++------------------------ fs/ext4/ext4.h | 6 +++--- fs/ext4/extents.c | 16 ++++++--------- fs/ext4/indirect.c | 6 ++---- fs/ext4/inode.c | 5 ++--- fs/ext4/mballoc.c | 4 ++-- 6 files changed, 41 insertions(+), 47 deletions(-) (limited to 'fs') diff --git a/fs/ext4/block_validity.c b/fs/ext4/block_validity.c index b394a50ebbe3..e830a9d4e10d 100644 --- a/fs/ext4/block_validity.c +++ b/fs/ext4/block_validity.c @@ -24,6 +24,7 @@ struct ext4_system_zone { struct rb_node node; ext4_fsblk_t start_blk; unsigned int count; + u32 ino; }; static struct kmem_cache *ext4_system_zone_cachep; @@ -45,7 +46,8 @@ void ext4_exit_system_zone(void) static inline int can_merge(struct ext4_system_zone *entry1, struct ext4_system_zone *entry2) { - if ((entry1->start_blk + entry1->count) == entry2->start_blk) + if ((entry1->start_blk + entry1->count) == entry2->start_blk && + entry1->ino == entry2->ino) return 1; return 0; } @@ -66,7 +68,7 @@ static void release_system_zone(struct ext4_system_blocks *system_blks) */ static int add_system_zone(struct ext4_system_blocks *system_blks, ext4_fsblk_t start_blk, - unsigned int count) + unsigned int count, u32 ino) { struct ext4_system_zone *new_entry, *entry; struct rb_node **n = &system_blks->root.rb_node, *node; @@ -89,6 +91,7 @@ static int add_system_zone(struct ext4_system_blocks *system_blks, return -ENOMEM; new_entry->start_blk = start_blk; new_entry->count = count; + new_entry->ino = ino; new_node = &new_entry->node; rb_link_node(new_node, parent, n); @@ -149,7 +152,7 @@ static void debug_print_tree(struct ext4_sb_info *sbi) static int ext4_data_block_valid_rcu(struct ext4_sb_info *sbi, struct ext4_system_blocks *system_blks, ext4_fsblk_t start_blk, - unsigned int count) + unsigned int count, ino_t ino) { struct ext4_system_zone *entry; struct rb_node *n; @@ -170,7 +173,7 @@ static int ext4_data_block_valid_rcu(struct ext4_sb_info *sbi, else if (start_blk >= (entry->start_blk + entry->count)) n = n->rb_right; else - return 0; + return entry->ino == ino; } return 1; } @@ -204,19 +207,18 @@ static int ext4_protect_reserved_inode(struct super_block *sb, if (n == 0) { i++; } else { - if (!ext4_data_block_valid_rcu(sbi, system_blks, - map.m_pblk, n)) { - err = -EFSCORRUPTED; - __ext4_error(sb, __func__, __LINE__, -err, - map.m_pblk, "blocks %llu-%llu " - "from inode %u overlap system zone", - map.m_pblk, - map.m_pblk + map.m_len - 1, ino); + err = add_system_zone(system_blks, map.m_pblk, n, ino); + if (err < 0) { + if (err == -EFSCORRUPTED) { + __ext4_error(sb, __func__, __LINE__, + -err, map.m_pblk, + "blocks %llu-%llu from inode %u overlap system zone", + map.m_pblk, + map.m_pblk + map.m_len - 1, + ino); + } break; } - err = add_system_zone(system_blks, map.m_pblk, n); - if (err < 0) - break; i += n; } } @@ -270,19 +272,19 @@ int ext4_setup_system_zone(struct super_block *sb) ((i < 5) || ((i % flex_size) == 0))) add_system_zone(system_blks, ext4_group_first_block_no(sb, i), - ext4_bg_num_gdb(sb, i) + 1); + ext4_bg_num_gdb(sb, i) + 1, 0); gdp = ext4_get_group_desc(sb, i, NULL); ret = add_system_zone(system_blks, - ext4_block_bitmap(sb, gdp), 1); + ext4_block_bitmap(sb, gdp), 1, 0); if (ret) goto err; ret = add_system_zone(system_blks, - ext4_inode_bitmap(sb, gdp), 1); + ext4_inode_bitmap(sb, gdp), 1, 0); if (ret) goto err; ret = add_system_zone(system_blks, ext4_inode_table(sb, gdp), - sbi->s_itb_per_group); + sbi->s_itb_per_group, 0); if (ret) goto err; } @@ -331,7 +333,7 @@ void ext4_release_system_zone(struct super_block *sb) call_rcu(&system_blks->rcu, ext4_destroy_system_zone); } -int ext4_data_block_valid(struct ext4_sb_info *sbi, ext4_fsblk_t start_blk, +int ext4_inode_block_valid(struct inode *inode, ext4_fsblk_t start_blk, unsigned int count) { struct ext4_system_blocks *system_blks; @@ -343,9 +345,9 @@ int ext4_data_block_valid(struct ext4_sb_info *sbi, ext4_fsblk_t start_blk, * mount option. */ rcu_read_lock(); - system_blks = rcu_dereference(sbi->system_blks); - ret = ext4_data_block_valid_rcu(sbi, system_blks, start_blk, - count); + system_blks = rcu_dereference(EXT4_SB(inode->i_sb)->system_blks); + ret = ext4_data_block_valid_rcu(EXT4_SB(inode->i_sb), system_blks, + start_blk, count, inode->i_ino); rcu_read_unlock(); return ret; } @@ -364,8 +366,7 @@ int ext4_check_blockref(const char *function, unsigned int line, while (bref < p+max) { blk = le32_to_cpu(*bref++); if (blk && - unlikely(!ext4_data_block_valid(EXT4_SB(inode->i_sb), - blk, 1))) { + unlikely(!ext4_inode_block_valid(inode, blk, 1))) { ext4_error_inode(inode, function, line, blk, "invalid block"); return -EFSCORRUPTED; diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index c73b9a1959a9..3994f012a9de 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -3397,9 +3397,9 @@ extern void ext4_release_system_zone(struct super_block *sb); extern int ext4_setup_system_zone(struct super_block *sb); extern int __init ext4_init_system_zone(void); extern void ext4_exit_system_zone(void); -extern int ext4_data_block_valid(struct ext4_sb_info *sbi, - ext4_fsblk_t start_blk, - unsigned int count); +extern int ext4_inode_block_valid(struct inode *inode, + ext4_fsblk_t start_blk, + unsigned int count); extern int ext4_check_blockref(const char *, unsigned int, struct inode *, __le32 *, unsigned int); diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c index 4af3f36c8351..0eea09aa0f26 100644 --- a/fs/ext4/extents.c +++ b/fs/ext4/extents.c @@ -340,7 +340,7 @@ static int ext4_valid_extent(struct inode *inode, struct ext4_extent *ext) */ if (lblock + len <= lblock) return 0; - return ext4_data_block_valid(EXT4_SB(inode->i_sb), block, len); + return ext4_inode_block_valid(inode, block, len); } static int ext4_valid_extent_idx(struct inode *inode, @@ -348,7 +348,7 @@ static int ext4_valid_extent_idx(struct inode *inode, { ext4_fsblk_t block = ext4_idx_pblock(ext_idx); - return ext4_data_block_valid(EXT4_SB(inode->i_sb), block, 1); + return ext4_inode_block_valid(inode, block, 1); } static int ext4_valid_extent_entries(struct inode *inode, @@ -507,14 +507,10 @@ __read_extent_tree_block(const char *function, unsigned int line, } if (buffer_verified(bh) && !(flags & EXT4_EX_FORCE_CACHE)) return bh; - if (!ext4_has_feature_journal(inode->i_sb) || - (inode->i_ino != - le32_to_cpu(EXT4_SB(inode->i_sb)->s_es->s_journal_inum))) { - err = __ext4_ext_check(function, line, inode, - ext_block_hdr(bh), depth, pblk); - if (err) - goto errout; - } + err = __ext4_ext_check(function, line, inode, + ext_block_hdr(bh), depth, pblk); + if (err) + goto errout; set_buffer_verified(bh); /* * If this is a leaf block, cache all of its entries diff --git a/fs/ext4/indirect.c b/fs/ext4/indirect.c index be2b66eb65f7..402641825712 100644 --- a/fs/ext4/indirect.c +++ b/fs/ext4/indirect.c @@ -858,8 +858,7 @@ static int ext4_clear_blocks(handle_t *handle, struct inode *inode, else if (ext4_should_journal_data(inode)) flags |= EXT4_FREE_BLOCKS_FORGET; - if (!ext4_data_block_valid(EXT4_SB(inode->i_sb), block_to_free, - count)) { + if (!ext4_inode_block_valid(inode, block_to_free, count)) { EXT4_ERROR_INODE(inode, "attempt to clear invalid " "blocks %llu len %lu", (unsigned long long) block_to_free, count); @@ -1004,8 +1003,7 @@ static void ext4_free_branches(handle_t *handle, struct inode *inode, if (!nr) continue; /* A hole */ - if (!ext4_data_block_valid(EXT4_SB(inode->i_sb), - nr, 1)) { + if (!ext4_inode_block_valid(inode, nr, 1)) { EXT4_ERROR_INODE(inode, "invalid indirect mapped " "block %lu (level %d)", diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 551a1056870d..0b07576af3bf 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -394,8 +394,7 @@ static int __check_block_validity(struct inode *inode, const char *func, (inode->i_ino == le32_to_cpu(EXT4_SB(inode->i_sb)->s_es->s_journal_inum))) return 0; - if (!ext4_data_block_valid(EXT4_SB(inode->i_sb), map->m_pblk, - map->m_len)) { + if (!ext4_inode_block_valid(inode, map->m_pblk, map->m_len)) { ext4_error_inode(inode, func, line, map->m_pblk, "lblock %lu mapped to illegal pblock %llu " "(length %d)", (unsigned long) map->m_lblk, @@ -4761,7 +4760,7 @@ struct inode *__ext4_iget(struct super_block *sb, unsigned long ino, ret = 0; if (ei->i_file_acl && - !ext4_data_block_valid(EXT4_SB(sb), ei->i_file_acl, 1)) { + !ext4_inode_block_valid(inode, ei->i_file_acl, 1)) { ext4_error_inode(inode, function, line, 0, "iget: bad extended attribute block %llu", ei->i_file_acl); diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index 19101941dadd..8b8c84db243a 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -3233,7 +3233,7 @@ ext4_mb_mark_diskspace_used(struct ext4_allocation_context *ac, block = ext4_grp_offs_to_block(sb, &ac->ac_b_ex); len = EXT4_C2B(sbi, ac->ac_b_ex.fe_len); - if (!ext4_data_block_valid(sbi, block, len)) { + if (!ext4_inode_block_valid(ac->ac_inode, block, len)) { ext4_error(sb, "Allocating blocks %llu-%llu which overlap " "fs metadata", block, block+len); /* File system mounted not to panic on error @@ -5058,7 +5058,7 @@ void ext4_free_blocks(handle_t *handle, struct inode *inode, sbi = EXT4_SB(sb); if (!(flags & EXT4_FREE_BLOCKS_VALIDATED) && - !ext4_data_block_valid(sbi, block, count)) { + !ext4_inode_block_valid(inode, block, count)) { ext4_error(sb, "Freeing blocks not in datazone - " "block = %llu, count = %lu", block, count); goto error_return; -- cgit v1.2.3 From 3f67e7cffa6ee2d49ccb697c569c6d30e6024160 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Tue, 28 Jul 2020 15:04:35 +0200 Subject: ext4: fold ext4_data_block_valid_rcu() into the caller After the previous patch, ext4_data_block_valid_rcu() has a single caller. Fold it into it. Reviewed-by: Lukas Czerner Signed-off-by: Jan Kara Link: https://lore.kernel.org/r/20200728130437.7804-5-jack@suse.cz Signed-off-by: Theodore Ts'o --- fs/ext4/block_validity.c | 69 ++++++++++++++++++++++-------------------------- 1 file changed, 31 insertions(+), 38 deletions(-) (limited to 'fs') diff --git a/fs/ext4/block_validity.c b/fs/ext4/block_validity.c index e830a9d4e10d..9c40214f31f9 100644 --- a/fs/ext4/block_validity.c +++ b/fs/ext4/block_validity.c @@ -144,40 +144,6 @@ static void debug_print_tree(struct ext4_sb_info *sbi) printk(KERN_CONT "\n"); } -/* - * Returns 1 if the passed-in block region (start_blk, - * start_blk+count) is valid; 0 if some part of the block region - * overlaps with filesystem metadata blocks. - */ -static int ext4_data_block_valid_rcu(struct ext4_sb_info *sbi, - struct ext4_system_blocks *system_blks, - ext4_fsblk_t start_blk, - unsigned int count, ino_t ino) -{ - struct ext4_system_zone *entry; - struct rb_node *n; - - if ((start_blk <= le32_to_cpu(sbi->s_es->s_first_data_block)) || - (start_blk + count < start_blk) || - (start_blk + count > ext4_blocks_count(sbi->s_es))) - return 0; - - if (system_blks == NULL) - return 1; - - n = system_blks->root.rb_node; - while (n) { - entry = rb_entry(n, struct ext4_system_zone, node); - if (start_blk + count - 1 < entry->start_blk) - n = n->rb_left; - else if (start_blk >= (entry->start_blk + entry->count)) - n = n->rb_right; - else - return entry->ino == ino; - } - return 1; -} - static int ext4_protect_reserved_inode(struct super_block *sb, struct ext4_system_blocks *system_blks, u32 ino) @@ -333,11 +299,24 @@ void ext4_release_system_zone(struct super_block *sb) call_rcu(&system_blks->rcu, ext4_destroy_system_zone); } +/* + * Returns 1 if the passed-in block region (start_blk, + * start_blk+count) is valid; 0 if some part of the block region + * overlaps with some other filesystem metadata blocks. + */ int ext4_inode_block_valid(struct inode *inode, ext4_fsblk_t start_blk, unsigned int count) { + struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); struct ext4_system_blocks *system_blks; - int ret; + struct ext4_system_zone *entry; + struct rb_node *n; + int ret = 1; + + if ((start_blk <= le32_to_cpu(sbi->s_es->s_first_data_block)) || + (start_blk + count < start_blk) || + (start_blk + count > ext4_blocks_count(sbi->s_es))) + return 0; /* * Lock the system zone to prevent it being released concurrently @@ -345,9 +324,23 @@ int ext4_inode_block_valid(struct inode *inode, ext4_fsblk_t start_blk, * mount option. */ rcu_read_lock(); - system_blks = rcu_dereference(EXT4_SB(inode->i_sb)->system_blks); - ret = ext4_data_block_valid_rcu(EXT4_SB(inode->i_sb), system_blks, - start_blk, count, inode->i_ino); + system_blks = rcu_dereference(sbi->system_blks); + if (system_blks == NULL) + goto out_rcu; + + n = system_blks->root.rb_node; + while (n) { + entry = rb_entry(n, struct ext4_system_zone, node); + if (start_blk + count - 1 < entry->start_blk) + n = n->rb_left; + else if (start_blk >= (entry->start_blk + entry->count)) + n = n->rb_right; + else { + ret = (entry->ino == inode->i_ino); + break; + } + } +out_rcu: rcu_read_unlock(); return ret; } -- cgit v1.2.3 From e7bfb5c9bb3d63cb2abb3ceaf1a429d9f02f942d Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Tue, 28 Jul 2020 15:04:36 +0200 Subject: ext4: handle add_system_zone() failure in ext4_setup_system_zone() There's one place that fails to handle error from add_system_zone() call and thus we can fail to protect superblock and group-descriptor blocks properly in case of ENOMEM. Fix it. Reported-by: Lukas Czerner Reviewed-by: Lukas Czerner Signed-off-by: Jan Kara Link: https://lore.kernel.org/r/20200728130437.7804-6-jack@suse.cz Signed-off-by: Theodore Ts'o --- fs/ext4/block_validity.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/ext4/block_validity.c b/fs/ext4/block_validity.c index 9c40214f31f9..2d008c1b58f2 100644 --- a/fs/ext4/block_validity.c +++ b/fs/ext4/block_validity.c @@ -235,10 +235,13 @@ int ext4_setup_system_zone(struct super_block *sb) for (i=0; i < ngroups; i++) { cond_resched(); if (ext4_bg_has_super(sb, i) && - ((i < 5) || ((i % flex_size) == 0))) - add_system_zone(system_blks, + ((i < 5) || ((i % flex_size) == 0))) { + ret = add_system_zone(system_blks, ext4_group_first_block_no(sb, i), ext4_bg_num_gdb(sb, i) + 1, 0); + if (ret) + goto err; + } gdp = ext4_get_group_desc(sb, i, NULL); ret = add_system_zone(system_blks, ext4_block_bitmap(sb, gdp), 1, 0); -- cgit v1.2.3 From 0f5bde1db174f6c471f0bd27198575719dabe3e5 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Tue, 28 Jul 2020 15:04:37 +0200 Subject: ext4: correctly restore system zone info when remount fails When remounting filesystem fails late during remount handling and block_validity mount option is also changed during the remount, we fail to restore system zone information to a state matching the mount option. This is mostly harmless, just the block validity checking will not match the situation described by the mount option. Make sure these two are always consistent. Reported-by: Lukas Czerner Reviewed-by: Lukas Czerner Signed-off-by: Jan Kara Link: https://lore.kernel.org/r/20200728130437.7804-7-jack@suse.cz Signed-off-by: Theodore Ts'o --- fs/ext4/block_validity.c | 8 -------- fs/ext4/super.c | 29 +++++++++++++++++++++-------- 2 files changed, 21 insertions(+), 16 deletions(-) (limited to 'fs') diff --git a/fs/ext4/block_validity.c b/fs/ext4/block_validity.c index 2d008c1b58f2..c54ba52f2dd4 100644 --- a/fs/ext4/block_validity.c +++ b/fs/ext4/block_validity.c @@ -220,14 +220,6 @@ int ext4_setup_system_zone(struct super_block *sb) int flex_size = ext4_flex_bg_size(sbi); int ret; - if (!test_opt(sb, BLOCK_VALIDITY)) { - if (sbi->system_blks) - ext4_release_system_zone(sb); - return 0; - } - if (sbi->system_blks) - return 0; - system_blks = kzalloc(sizeof(*system_blks), GFP_KERNEL); if (!system_blks) return -ENOMEM; diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 6507e30e58b5..0892d71a52bd 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -4727,11 +4727,13 @@ no_journal: ext4_set_resv_clusters(sb); - err = ext4_setup_system_zone(sb); - if (err) { - ext4_msg(sb, KERN_ERR, "failed to initialize system " - "zone (%d)", err); - goto failed_mount4a; + if (test_opt(sb, BLOCK_VALIDITY)) { + err = ext4_setup_system_zone(sb); + if (err) { + ext4_msg(sb, KERN_ERR, "failed to initialize system " + "zone (%d)", err); + goto failed_mount4a; + } } ext4_ext_init(sb); @@ -5747,9 +5749,16 @@ static int ext4_remount(struct super_block *sb, int *flags, char *data) ext4_register_li_request(sb, first_not_zeroed); } - err = ext4_setup_system_zone(sb); - if (err) - goto restore_opts; + /* + * Handle creation of system zone data early because it can fail. + * Releasing of existing data is done when we are sure remount will + * succeed. + */ + if (test_opt(sb, BLOCK_VALIDITY) && !sbi->system_blks) { + err = ext4_setup_system_zone(sb); + if (err) + goto restore_opts; + } if (sbi->s_journal == NULL && !(old_sb_flags & SB_RDONLY)) { err = ext4_commit_super(sb, 1); @@ -5771,6 +5780,8 @@ static int ext4_remount(struct super_block *sb, int *flags, char *data) } } #endif + if (!test_opt(sb, BLOCK_VALIDITY) && sbi->system_blks) + ext4_release_system_zone(sb); /* * Some options can be enabled by ext4 and/or by VFS mount flag @@ -5792,6 +5803,8 @@ restore_opts: sbi->s_commit_interval = old_opts.s_commit_interval; sbi->s_min_batch_time = old_opts.s_min_batch_time; sbi->s_max_batch_time = old_opts.s_max_batch_time; + if (!test_opt(sb, BLOCK_VALIDITY) && sbi->system_blks) + ext4_release_system_zone(sb); #ifdef CONFIG_QUOTA sbi->s_jquota_fmt = old_opts.s_jquota_fmt; for (i = 0; i < EXT4_MAXQUOTAS; i++) { -- cgit v1.2.3 From 377254b2cd2252c7c3151b113cbdf93a7736c2e9 Mon Sep 17 00:00:00 2001 From: Xianting Tian Date: Fri, 31 Jul 2020 12:10:25 -0400 Subject: fs: prevent BUG_ON in submit_bh_wbc() If a device is hot-removed --- for example, when a physical device is unplugged from pcie slot or a nbd device's network is shutdown --- this can result in a BUG_ON() crash in submit_bh_wbc(). This is because the when the block device dies, the buffer heads will have their Buffer_Mapped flag get cleared, leading to the crash in submit_bh_wbc. We had attempted to work around this problem in commit a17712c8 ("ext4: check superblock mapped prior to committing"). Unfortunately, it's still possible to hit the BUG_ON(!buffer_mapped(bh)) if the device dies between when the work-around check in ext4_commit_super() and when submit_bh_wbh() is finally called: Code path: ext4_commit_super judge if 'buffer_mapped(sbh)' is false, return <== commit a17712c8 lock_buffer(sbh) ... unlock_buffer(sbh) __sync_dirty_buffer(sbh,... lock_buffer(sbh) judge if 'buffer_mapped(sbh))' is false, return <== added by this patch submit_bh(...,sbh) submit_bh_wbc(...,sbh,...) [100722.966497] kernel BUG at fs/buffer.c:3095! <== BUG_ON(!buffer_mapped(bh))' in submit_bh_wbc() [100722.966503] invalid opcode: 0000 [#1] SMP [100722.966566] task: ffff8817e15a9e40 task.stack: ffffc90024744000 [100722.966574] RIP: 0010:submit_bh_wbc+0x180/0x190 [100722.966575] RSP: 0018:ffffc90024747a90 EFLAGS: 00010246 [100722.966576] RAX: 0000000000620005 RBX: ffff8818a80603a8 RCX: 0000000000000000 [100722.966576] RDX: ffff8818a80603a8 RSI: 0000000000020800 RDI: 0000000000000001 [100722.966577] RBP: ffffc90024747ac0 R08: 0000000000000000 R09: ffff88207f94170d [100722.966578] R10: 00000000000437c8 R11: 0000000000000001 R12: 0000000000020800 [100722.966578] R13: 0000000000000001 R14: 000000000bf9a438 R15: ffff88195f333000 [100722.966580] FS: 00007fa2eee27700(0000) GS:ffff88203d840000(0000) knlGS:0000000000000000 [100722.966580] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [100722.966581] CR2: 0000000000f0b008 CR3: 000000201a622003 CR4: 00000000007606e0 [100722.966582] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [100722.966583] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 [100722.966583] PKRU: 55555554 [100722.966583] Call Trace: [100722.966588] __sync_dirty_buffer+0x6e/0xd0 [100722.966614] ext4_commit_super+0x1d8/0x290 [ext4] [100722.966626] __ext4_std_error+0x78/0x100 [ext4] [100722.966635] ? __ext4_journal_get_write_access+0xca/0x120 [ext4] [100722.966646] ext4_reserve_inode_write+0x58/0xb0 [ext4] [100722.966655] ? ext4_dirty_inode+0x48/0x70 [ext4] [100722.966663] ext4_mark_inode_dirty+0x53/0x1e0 [ext4] [100722.966671] ? __ext4_journal_start_sb+0x6d/0xf0 [ext4] [100722.966679] ext4_dirty_inode+0x48/0x70 [ext4] [100722.966682] __mark_inode_dirty+0x17f/0x350 [100722.966686] generic_update_time+0x87/0xd0 [100722.966687] touch_atime+0xa9/0xd0 [100722.966690] generic_file_read_iter+0xa09/0xcd0 [100722.966694] ? page_cache_tree_insert+0xb0/0xb0 [100722.966704] ext4_file_read_iter+0x4a/0x100 [ext4] [100722.966707] ? __inode_security_revalidate+0x4f/0x60 [100722.966709] __vfs_read+0xec/0x160 [100722.966711] vfs_read+0x8c/0x130 [100722.966712] SyS_pread64+0x87/0xb0 [100722.966716] do_syscall_64+0x67/0x1b0 [100722.966719] entry_SYSCALL64_slow_path+0x25/0x25 To address this, add the check of 'buffer_mapped(bh)' to __sync_dirty_buffer(). This also has the benefit of fixing this for other file systems. With this addition, we can drop the workaround in ext4_commit_supper(). [ Commit description rewritten by tytso. ] Signed-off-by: Xianting Tian Link: https://lore.kernel.org/r/1596211825-8750-1-git-send-email-xianting_tian@126.com Signed-off-by: Theodore Ts'o --- fs/buffer.c | 9 +++++++++ fs/ext4/super.c | 7 ------- 2 files changed, 9 insertions(+), 7 deletions(-) (limited to 'fs') diff --git a/fs/buffer.c b/fs/buffer.c index 64fe82ec65ff..75a8849abb5d 100644 --- a/fs/buffer.c +++ b/fs/buffer.c @@ -3160,6 +3160,15 @@ int __sync_dirty_buffer(struct buffer_head *bh, int op_flags) WARN_ON(atomic_read(&bh->b_count) < 1); lock_buffer(bh); if (test_clear_buffer_dirty(bh)) { + /* + * The bh should be mapped, but it might not be if the + * device was hot-removed. Not much we can do but fail the I/O. + */ + if (!buffer_mapped(bh)) { + unlock_buffer(bh); + return -EIO; + } + get_bh(bh); bh->b_end_io = end_buffer_write_sync; ret = submit_bh(REQ_OP_WRITE, op_flags, bh); diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 0892d71a52bd..daa94c7f7271 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -5235,13 +5235,6 @@ static int ext4_commit_super(struct super_block *sb, int sync) if (!sbh || block_device_ejected(sb)) return error; - /* - * The superblock bh should be mapped, but it might not be if the - * device was hot-removed. Not much we can do but fail the I/O. - */ - if (!buffer_mapped(sbh)) - return error; - /* * If the file system is mounted read-only, don't update the * superblock write time. This avoids updating the superblock -- cgit v1.2.3 From 7303cb5bfe845f7d43cd9b2dbd37dbb266efda9b Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Fri, 31 Jul 2020 18:21:35 +0200 Subject: ext4: fix checking of directory entry validity for inline directories ext4_search_dir() and ext4_generic_delete_entry() can be called both for standard director blocks and for inline directories stored inside inode or inline xattr space. For the second case we didn't call ext4_check_dir_entry() with proper constraints that could result in accepting corrupted directory entry as well as false positive filesystem errors like: EXT4-fs error (device dm-0): ext4_search_dir:1395: inode #28320400: block 113246792: comm dockerd: bad entry in directory: directory entry too close to block end - offset=0, inode=28320403, rec_len=32, name_len=8, size=4096 Fix the arguments passed to ext4_check_dir_entry(). Fixes: 109ba779d6cc ("ext4: check for directory entries too close to block end") CC: stable@vger.kernel.org Signed-off-by: Jan Kara Link: https://lore.kernel.org/r/20200731162135.8080-1-jack@suse.cz Signed-off-by: Theodore Ts'o --- fs/ext4/namei.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c index cb2eb1967e73..b92571beab72 100644 --- a/fs/ext4/namei.c +++ b/fs/ext4/namei.c @@ -1396,8 +1396,8 @@ int ext4_search_dir(struct buffer_head *bh, char *search_buf, int buf_size, ext4_match(dir, fname, de)) { /* found a match - just to be sure, do * a full check */ - if (ext4_check_dir_entry(dir, NULL, de, bh, bh->b_data, - bh->b_size, offset)) + if (ext4_check_dir_entry(dir, NULL, de, bh, search_buf, + buf_size, offset)) return -1; *res_dir = de; return 1; @@ -2482,7 +2482,7 @@ int ext4_generic_delete_entry(handle_t *handle, de = (struct ext4_dir_entry_2 *)entry_buf; while (i < buf_size - csum_size) { if (ext4_check_dir_entry(dir, NULL, de, bh, - bh->b_data, bh->b_size, i)) + entry_buf, buf_size, i)) return -EFSCORRUPTED; if (de == de_del) { if (pde) -- cgit v1.2.3 From e9a3cd48d65307b1a489300d7e3e0b28ae94dbfd Mon Sep 17 00:00:00 2001 From: brookxu Date: Fri, 7 Aug 2020 22:01:23 +0800 Subject: ext4: fix typos in ext4_mb_regular_allocator() comment Fix typos in ext4_mb_regular_allocator() comment Signed-off-by: Chunguang Xu Reviewed-by: Ritesh Harjani Reviewed-by: Andreas Dilger Link: https://lore.kernel.org/r/d6514145-73b3-808b-ec5a-a8be27c51f9c@gmail.com Signed-off-by: Theodore Ts'o --- fs/ext4/mballoc.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index 8b8c84db243a..5be24244426a 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -2332,8 +2332,8 @@ ext4_mb_regular_allocator(struct ext4_allocation_context *ac) goto out; /* - * ac->ac2_order is set only if the fe_len is a power of 2 - * if ac2_order is set we also set criteria to 0 so that we + * ac->ac_2order is set only if the fe_len is a power of 2 + * if ac->ac_2order is set we also set criteria to 0 so that we * try exact allocation using buddy. */ i = fls(ac->ac_g_ex.fe_len); -- cgit v1.2.3 From 051e2ce8cb90592c8acdc7056ffed966310d91b4 Mon Sep 17 00:00:00 2001 From: brookxu Date: Fri, 7 Aug 2020 22:01:33 +0800 Subject: ext4: delete invalid comments near ext4_mb_check_limits() These comments do not seem to be related to ext4_mb_check_limits(), it may be invalid. Signed-off-by: Chunguang Xu Reviewed-by: Ritesh Harjani Link: https://lore.kernel.org/r/c49faf0c-d5d5-9c51-6911-9e0ff57c6bfa@gmail.com Signed-off-by: Theodore Ts'o --- fs/ext4/mballoc.c | 4 ---- 1 file changed, 4 deletions(-) (limited to 'fs') diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index 5be24244426a..63ffd0057471 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -1740,10 +1740,6 @@ static void ext4_mb_use_best_found(struct ext4_allocation_context *ac, } -/* - * regular allocator, for general purposes allocation - */ - static void ext4_mb_check_limits(struct ext4_allocation_context *ac, struct ext4_buddy *e4b, int finish_group) -- cgit v1.2.3 From dddcd2f9ebdeca9fbd36526e950bbcd0f7c1765f Mon Sep 17 00:00:00 2001 From: brookxu Date: Fri, 7 Aug 2020 22:01:39 +0800 Subject: ext4: optimize the implementation of ext4_mb_good_group() It might be better to adjust the code in two places: 1. Determine whether grp is currupt or not should be placed first. 2. (cr<=2 && free ac_g_ex.fe_len)should may belong to the crx strategy, and it may be more appropriate to put it in the subsequent switch statement block. For cr1, cr2, the conditions in switch potentially realize the above judgment. For cr0, we should add (free ac_g_ex.fe_len) judgment, and then delete (free / fragments) >= ac->ac_g_ex.fe_len), because cr0 returns true by default. Signed-off-by: Chunguang Xu Reviewed-by: Andreas Dilger Reviewed-by: Ritesh Harjani Link: https://lore.kernel.org/r/e20b2d8f-1154-adb7-3831-a9e11ba842e9@gmail.com Signed-off-by: Theodore Ts'o --- fs/ext4/mballoc.c | 14 +++++++------- 1 file changed, 7 insertions(+), 7 deletions(-) (limited to 'fs') diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index 63ffd0057471..f68730eec205 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -2112,13 +2112,11 @@ static bool ext4_mb_good_group(struct ext4_allocation_context *ac, BUG_ON(cr < 0 || cr >= 4); - free = grp->bb_free; - if (free == 0) - return false; - if (cr <= 2 && free < ac->ac_g_ex.fe_len) + if (unlikely(EXT4_MB_GRP_BBITMAP_CORRUPT(grp))) return false; - if (unlikely(EXT4_MB_GRP_BBITMAP_CORRUPT(grp))) + free = grp->bb_free; + if (free == 0) return false; fragments = grp->bb_fragments; @@ -2135,8 +2133,10 @@ static bool ext4_mb_good_group(struct ext4_allocation_context *ac, ((group % flex_size) == 0)) return false; - if ((ac->ac_2order > ac->ac_sb->s_blocksize_bits+1) || - (free / fragments) >= ac->ac_g_ex.fe_len) + if (free < ac->ac_g_ex.fe_len) + return false; + + if (ac->ac_2order > ac->ac_sb->s_blocksize_bits+1) return true; if (grp->bb_largest_free_order < ac->ac_2order) -- cgit v1.2.3 From e0d438c72abe0cb000d67c0795b0341f9e763836 Mon Sep 17 00:00:00 2001 From: Xu Wang Date: Mon, 10 Aug 2020 02:21:58 +0000 Subject: mballoc: replace seq_printf with seq_puts seq_puts is a lot cheaper than seq_printf, so use that to print literal strings. Signed-off-by: Xu Wang Reviewed-by: Ritesh Harjani Link: https://lore.kernel.org/r/20200810022158.9167-1-vulab@iscas.ac.cn Signed-off-by: Theodore Ts'o --- fs/ext4/mballoc.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index f68730eec205..51e2fe6094be 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -2558,7 +2558,7 @@ static int ext4_mb_seq_groups_show(struct seq_file *seq, void *v) for (i = 0; i <= 13; i++) seq_printf(seq, " %-5u", i <= blocksize_bits + 1 ? sg.info.bb_counters[i] : 0); - seq_printf(seq, " ]\n"); + seq_puts(seq, " ]\n"); return 0; } -- cgit v1.2.3 From 2fe34d2938181b26f86bceb7b87fbf7370bd92c4 Mon Sep 17 00:00:00 2001 From: Kyoungho Koo Date: Mon, 10 Aug 2020 17:07:05 +0900 Subject: ext4: remove unused parameter of ext4_generic_delete_entry function The ext4_generic_delete_entry function does not use the parameter handle, so it can be removed. Signed-off-by: Kyoungho Koo Reviewed-by: Ritesh Harjani Link: https://lore.kernel.org/r/20200810080701.GA14160@koo-Z370-HD3 Signed-off-by: Theodore Ts'o --- fs/ext4/ext4.h | 3 +-- fs/ext4/inline.c | 2 +- fs/ext4/namei.c | 6 ++---- 3 files changed, 4 insertions(+), 7 deletions(-) (limited to 'fs') diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index 3994f012a9de..71b4370a3f91 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -2802,8 +2802,7 @@ extern int ext4_search_dir(struct buffer_head *bh, struct ext4_filename *fname, unsigned int offset, struct ext4_dir_entry_2 **res_dir); -extern int ext4_generic_delete_entry(handle_t *handle, - struct inode *dir, +extern int ext4_generic_delete_entry(struct inode *dir, struct ext4_dir_entry_2 *de_del, struct buffer_head *bh, void *entry_buf, diff --git a/fs/ext4/inline.c b/fs/ext4/inline.c index c3a1ad2db122..13054653a06a 100644 --- a/fs/ext4/inline.c +++ b/fs/ext4/inline.c @@ -1706,7 +1706,7 @@ int ext4_delete_inline_entry(handle_t *handle, if (err) goto out; - err = ext4_generic_delete_entry(handle, dir, de_del, bh, + err = ext4_generic_delete_entry(dir, de_del, bh, inline_start, inline_size, 0); if (err) goto out; diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c index b92571beab72..153a9fbe1dd0 100644 --- a/fs/ext4/namei.c +++ b/fs/ext4/namei.c @@ -2465,8 +2465,7 @@ cleanup: * ext4_generic_delete_entry deletes a directory entry by merging it * with the previous entry */ -int ext4_generic_delete_entry(handle_t *handle, - struct inode *dir, +int ext4_generic_delete_entry(struct inode *dir, struct ext4_dir_entry_2 *de_del, struct buffer_head *bh, void *entry_buf, @@ -2527,8 +2526,7 @@ static int ext4_delete_entry(handle_t *handle, if (unlikely(err)) goto out; - err = ext4_generic_delete_entry(handle, dir, de_del, - bh, bh->b_data, + err = ext4_generic_delete_entry(dir, de_del, bh, bh->b_data, dir->i_sb->s_blocksize, csum_size); if (err) goto out; -- cgit v1.2.3 From 70d7ced2ed4dc3f7f21b4f4fec0e5be5c2a5a55c Mon Sep 17 00:00:00 2001 From: Shijie Luo Date: Mon, 10 Aug 2020 07:44:35 -0400 Subject: ext4: change to use fallthrough macro Change to use fallthrough macro in switch case. Signed-off-by: Shijie Luo Reviewed-by: Ritesh Harjani Link: https://lore.kernel.org/r/20200810114435.24182-1-luoshijie1@huawei.com Signed-off-by: Theodore Ts'o --- fs/ext4/hash.c | 4 ++-- fs/ext4/indirect.c | 12 ++++++------ fs/ext4/readpage.c | 4 ++-- 3 files changed, 10 insertions(+), 10 deletions(-) (limited to 'fs') diff --git a/fs/ext4/hash.c b/fs/ext4/hash.c index 3e133793a5a3..2924261226e0 100644 --- a/fs/ext4/hash.c +++ b/fs/ext4/hash.c @@ -233,7 +233,7 @@ static int __ext4fs_dirhash(const char *name, int len, break; case DX_HASH_HALF_MD4_UNSIGNED: str2hashbuf = str2hashbuf_unsigned; - /* fall through */ + fallthrough; case DX_HASH_HALF_MD4: p = name; while (len > 0) { @@ -247,7 +247,7 @@ static int __ext4fs_dirhash(const char *name, int len, break; case DX_HASH_TEA_UNSIGNED: str2hashbuf = str2hashbuf_unsigned; - /* fall through */ + fallthrough; case DX_HASH_TEA: p = name; while (len > 0) { diff --git a/fs/ext4/indirect.c b/fs/ext4/indirect.c index 402641825712..433ca8415c5a 100644 --- a/fs/ext4/indirect.c +++ b/fs/ext4/indirect.c @@ -1180,21 +1180,21 @@ do_indirects: ext4_free_branches(handle, inode, NULL, &nr, &nr+1, 1); i_data[EXT4_IND_BLOCK] = 0; } - /* fall through */ + fallthrough; case EXT4_IND_BLOCK: nr = i_data[EXT4_DIND_BLOCK]; if (nr) { ext4_free_branches(handle, inode, NULL, &nr, &nr+1, 2); i_data[EXT4_DIND_BLOCK] = 0; } - /* fall through */ + fallthrough; case EXT4_DIND_BLOCK: nr = i_data[EXT4_TIND_BLOCK]; if (nr) { ext4_free_branches(handle, inode, NULL, &nr, &nr+1, 3); i_data[EXT4_TIND_BLOCK] = 0; } - /* fall through */ + fallthrough; case EXT4_TIND_BLOCK: ; } @@ -1434,7 +1434,7 @@ do_indirects: ext4_free_branches(handle, inode, NULL, &nr, &nr+1, 1); i_data[EXT4_IND_BLOCK] = 0; } - /* fall through */ + fallthrough; case EXT4_IND_BLOCK: if (++n >= n2) break; @@ -1443,7 +1443,7 @@ do_indirects: ext4_free_branches(handle, inode, NULL, &nr, &nr+1, 2); i_data[EXT4_DIND_BLOCK] = 0; } - /* fall through */ + fallthrough; case EXT4_DIND_BLOCK: if (++n >= n2) break; @@ -1452,7 +1452,7 @@ do_indirects: ext4_free_branches(handle, inode, NULL, &nr, &nr+1, 3); i_data[EXT4_TIND_BLOCK] = 0; } - /* fall through */ + fallthrough; case EXT4_TIND_BLOCK: ; } diff --git a/fs/ext4/readpage.c b/fs/ext4/readpage.c index 5761e9961682..153a838e9e3c 100644 --- a/fs/ext4/readpage.c +++ b/fs/ext4/readpage.c @@ -140,7 +140,7 @@ static void bio_post_read_processing(struct bio_post_read_ctx *ctx) return; } ctx->cur_step++; - /* fall-through */ + fallthrough; case STEP_VERITY: if (ctx->enabled_steps & (1 << STEP_VERITY)) { INIT_WORK(&ctx->work, verity_work); @@ -148,7 +148,7 @@ static void bio_post_read_processing(struct bio_post_read_ctx *ctx) return; } ctx->cur_step++; - /* fall-through */ + fallthrough; default: __read_end_io(ctx->bio); } -- cgit v1.2.3 From 00a3fff0712cd9cc4112ecf6da0916f8503e2a86 Mon Sep 17 00:00:00 2001 From: Shijie Luo Date: Mon, 10 Aug 2020 22:21:28 -0400 Subject: jbd2: clean up checksum verification in do_one_pass() Remove the unnecessary chksum_err and checksum_seen variables as well as some redundant code to make the function easier to understand. [ With changes suggested by jack@ and tytso@ ] Signed-off-by: Shijie Luo Signed-off-by: Theodore Ts'o Reviewed-by: Jan Kara Link: https://lore.kernel.org/r/20200819122955.33526-1-luoshijie1@huawei.com Signed-off-by: Theodore Ts'o --- fs/jbd2/recovery.c | 46 ++++++++++++---------------------------------- 1 file changed, 12 insertions(+), 34 deletions(-) (limited to 'fs') diff --git a/fs/jbd2/recovery.c b/fs/jbd2/recovery.c index 2ed278f0dced..faa97d748474 100644 --- a/fs/jbd2/recovery.c +++ b/fs/jbd2/recovery.c @@ -690,14 +690,11 @@ static int do_one_pass(journal_t *journal, * number. */ if (pass == PASS_SCAN && jbd2_has_feature_checksum(journal)) { - int chksum_err, chksum_seen; struct commit_header *cbh = (struct commit_header *)bh->b_data; unsigned found_chksum = be32_to_cpu(cbh->h_chksum[0]); - chksum_err = chksum_seen = 0; - if (info->end_transaction) { journal->j_failed_commit = info->end_transaction; @@ -705,42 +702,23 @@ static int do_one_pass(journal_t *journal, break; } - if (crc32_sum == found_chksum && - cbh->h_chksum_type == JBD2_CRC32_CHKSUM && - cbh->h_chksum_size == - JBD2_CRC32_CHKSUM_SIZE) - chksum_seen = 1; - else if (!(cbh->h_chksum_type == 0 && - cbh->h_chksum_size == 0 && - found_chksum == 0 && - !chksum_seen)) - /* - * If fs is mounted using an old kernel and then - * kernel with journal_chksum is used then we - * get a situation where the journal flag has - * checksum flag set but checksums are not - * present i.e chksum = 0, in the individual - * commit blocks. - * Hence to avoid checksum failures, in this - * situation, this extra check is added. - */ - chksum_err = 1; - - if (chksum_err) { - info->end_transaction = next_commit_ID; - - if (!jbd2_has_feature_async_commit(journal)) { - journal->j_failed_commit = - next_commit_ID; - brelse(bh); - break; - } - } + /* Neither checksum match nor unused? */ + if (!((crc32_sum == found_chksum && + cbh->h_chksum_type == + JBD2_CRC32_CHKSUM && + cbh->h_chksum_size == + JBD2_CRC32_CHKSUM_SIZE) || + (cbh->h_chksum_type == 0 && + cbh->h_chksum_size == 0 && + found_chksum == 0))) + goto chksum_error; + crc32_sum = ~0; } if (pass == PASS_SCAN && !jbd2_commit_block_csum_verify(journal, bh->b_data)) { + chksum_error: info->end_transaction = next_commit_ID; if (!jbd2_has_feature_async_commit(journal)) { -- cgit v1.2.3 From 7ca4fcba92edf82b05533a6e4f6b19a5fbec6f54 Mon Sep 17 00:00:00 2001 From: kyoungho koo Date: Sat, 25 Apr 2020 02:16:24 +0900 Subject: ext4: Fix comment typo "the the". I have found double typed comments "the the". So i modified it to one "the" Signed-off-by: kyoungho koo Link: https://lore.kernel.org/r/20200424171620.GA11943@koo-Z370-HD3 Signed-off-by: Theodore Ts'o --- fs/ext4/inline.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ext4/inline.c b/fs/ext4/inline.c index 13054653a06a..75c97bca0815 100644 --- a/fs/ext4/inline.c +++ b/fs/ext4/inline.c @@ -276,7 +276,7 @@ static int ext4_create_inline_data(handle_t *handle, len = 0; } - /* Insert the the xttr entry. */ + /* Insert the xttr entry. */ i.value = value; i.value_len = len; -- cgit v1.2.3 From c55ee7d202857b000531c2acfe8ce7cba5b77b5c Mon Sep 17 00:00:00 2001 From: brookxu Date: Sat, 15 Aug 2020 08:10:44 +0800 Subject: ext4: add mb_debug logging when there are lost chunks Lost chunks are when some other process raced with the current thread to grab a particular block allocation. Add mb_debug log for developers who wants to see how often this is happening for a particular workload. Signed-off-by: Chunguang Xu Link: https://lore.kernel.org/r/0a165ac0-1912-aebd-8a0d-b42e7cd1aea1@gmail.com Signed-off-by: Theodore Ts'o --- fs/ext4/mballoc.c | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index 51e2fe6094be..0993044cfe46 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -2309,6 +2309,7 @@ ext4_mb_regular_allocator(struct ext4_allocation_context *ac) struct ext4_sb_info *sbi; struct super_block *sb; struct ext4_buddy e4b; + unsigned int lost; sb = ac->ac_sb; sbi = EXT4_SB(sb); @@ -2456,22 +2457,24 @@ repeat: * We've been searching too long. Let's try to allocate * the best chunk we've found so far */ - ext4_mb_try_best_found(ac, &e4b); if (ac->ac_status != AC_STATUS_FOUND) { /* * Someone more lucky has already allocated it. * The only thing we can do is just take first * found block(s) - printk(KERN_DEBUG "EXT4-fs: someone won our chunk\n"); */ + lost = (unsigned int)atomic_inc_return(&sbi->s_mb_lost_chunks); + mb_debug(sb, "lost chunk, group: %u, start: %d, len: %d, lost: %u\n", + ac->ac_b_ex.fe_group, ac->ac_b_ex.fe_start, + ac->ac_b_ex.fe_len, lost); + ac->ac_b_ex.fe_group = 0; ac->ac_b_ex.fe_start = 0; ac->ac_b_ex.fe_len = 0; ac->ac_status = AC_STATUS_CONTINUE; ac->ac_flags |= EXT4_MB_HINT_FIRST; cr = 3; - atomic_inc(&sbi->s_mb_lost_chunks); goto repeat; } } -- cgit v1.2.3 From 66d5e0277e225cdc5d272fc22b1aa90a9b0d21ac Mon Sep 17 00:00:00 2001 From: brookxu Date: Mon, 17 Aug 2020 15:36:06 +0800 Subject: ext4: reorganize if statement of ext4_mb_release_context() Reorganize the if statement of ext4_mb_release_context(), make it easier to read. Signed-off-by: Chunguang Xu Link: https://lore.kernel.org/r/5439ac6f-db79-ad68-76c1-a4dda9aa0cc3@gmail.com Reviewed-by: Ritesh Harjani Signed-off-by: Theodore Ts'o --- fs/ext4/mballoc.c | 33 ++++++++++++++++----------------- 1 file changed, 16 insertions(+), 17 deletions(-) (limited to 'fs') diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index 0993044cfe46..45ac6088b4ac 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -2309,7 +2309,7 @@ ext4_mb_regular_allocator(struct ext4_allocation_context *ac) struct ext4_sb_info *sbi; struct super_block *sb; struct ext4_buddy e4b; - unsigned int lost; + int lost; sb = ac->ac_sb; sbi = EXT4_SB(sb); @@ -2464,8 +2464,8 @@ repeat: * The only thing we can do is just take first * found block(s) */ - lost = (unsigned int)atomic_inc_return(&sbi->s_mb_lost_chunks); - mb_debug(sb, "lost chunk, group: %u, start: %d, len: %d, lost: %u\n", + lost = atomic_inc_return(&sbi->s_mb_lost_chunks); + mb_debug(sb, "lost chunk, group: %u, start: %d, len: %d, lost: %d\n", ac->ac_b_ex.fe_group, ac->ac_b_ex.fe_start, ac->ac_b_ex.fe_len, lost); @@ -4706,20 +4706,19 @@ static int ext4_mb_release_context(struct ext4_allocation_context *ac) pa->pa_free -= ac->ac_b_ex.fe_len; pa->pa_len -= ac->ac_b_ex.fe_len; spin_unlock(&pa->pa_lock); - } - } - if (pa) { - /* - * We want to add the pa to the right bucket. - * Remove it from the list and while adding - * make sure the list to which we are adding - * doesn't grow big. - */ - if ((pa->pa_type == MB_GROUP_PA) && likely(pa->pa_free)) { - spin_lock(pa->pa_obj_lock); - list_del_rcu(&pa->pa_inode_list); - spin_unlock(pa->pa_obj_lock); - ext4_mb_add_n_trim(ac); + + /* + * We want to add the pa to the right bucket. + * Remove it from the list and while adding + * make sure the list to which we are adding + * doesn't grow big. + */ + if (likely(pa->pa_free)) { + spin_lock(pa->pa_obj_lock); + list_del_rcu(&pa->pa_inode_list); + spin_unlock(pa->pa_obj_lock); + ext4_mb_add_n_trim(ac); + } } ext4_mb_put_pa(ac, ac->ac_sb, pa); } -- cgit v1.2.3 From 27bc446e2def38db3244a6eb4bb1d6312936610a Mon Sep 17 00:00:00 2001 From: brookxu Date: Mon, 17 Aug 2020 15:36:15 +0800 Subject: ext4: limit the length of per-inode prealloc list MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit In the scenario of writing sparse files, the per-inode prealloc list may be very long, resulting in high overhead for ext4_mb_use_preallocated(). To circumvent this problem, we limit the maximum length of per-inode prealloc list to 512 and allow users to modify it. After patching, we observed that the sys ratio of cpu has dropped, and the system throughput has increased significantly. We created a process to write the sparse file, and the running time of the process on the fixed kernel was significantly reduced, as follows: Running time on unfixed kernel： [root@TENCENT64 ~]# time taskset 0x01 ./sparse /data1/sparce.dat real 0m2.051s user 0m0.008s sys 0m2.026s Running time on fixed kernel： [root@TENCENT64 ~]# time taskset 0x01 ./sparse /data1/sparce.dat real 0m0.471s user 0m0.004s sys 0m0.395s Signed-off-by: Chunguang Xu Link: https://lore.kernel.org/r/d7a98178-056b-6db5-6bce-4ead23f4a257@gmail.com Signed-off-by: Theodore Ts'o --- Documentation/admin-guide/ext4.rst | 3 ++ fs/ext4/ext4.h | 4 ++- fs/ext4/extents.c | 10 +++--- fs/ext4/file.c | 2 +- fs/ext4/indirect.c | 2 +- fs/ext4/inode.c | 6 ++-- fs/ext4/ioctl.c | 2 +- fs/ext4/mballoc.c | 74 +++++++++++++++++++++++++++++++++----- fs/ext4/mballoc.h | 4 +++ fs/ext4/move_extent.c | 4 +-- fs/ext4/super.c | 3 +- fs/ext4/sysfs.c | 2 ++ include/trace/events/ext4.h | 17 +++++---- 13 files changed, 104 insertions(+), 29 deletions(-) (limited to 'fs') diff --git a/Documentation/admin-guide/ext4.rst b/Documentation/admin-guide/ext4.rst index 7fc6a72920c9..59bcc4a92602 100644 --- a/Documentation/admin-guide/ext4.rst +++ b/Documentation/admin-guide/ext4.rst @@ -482,6 +482,9 @@ Files in /sys/fs/ext4/: multiple of this tuning parameter if the stripe size is not set in the ext4 superblock + mb_max_inode_prealloc + The maximum length of per-inode ext4_prealloc_space list. + mb_max_to_scan The maximum number of extents the multiblock allocator will search to find the best extent. diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index 71b4370a3f91..523e00d7b392 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -1070,6 +1070,7 @@ struct ext4_inode_info { struct timespec64 i_crtime; /* mballoc */ + atomic_t i_prealloc_active; struct list_head i_prealloc_list; spinlock_t i_prealloc_lock; @@ -1518,6 +1519,7 @@ struct ext4_sb_info { unsigned int s_mb_stats; unsigned int s_mb_order2_reqs; unsigned int s_mb_group_prealloc; + unsigned int s_mb_max_inode_prealloc; unsigned int s_max_dir_size_kb; /* where last allocation was done - for stream allocation */ unsigned long s_mb_last_group; @@ -2682,7 +2684,7 @@ extern int ext4_mb_release(struct super_block *); extern ext4_fsblk_t ext4_mb_new_blocks(handle_t *, struct ext4_allocation_request *, int *); extern int ext4_mb_reserve_blocks(struct super_block *, int); -extern void ext4_discard_preallocations(struct inode *); +extern void ext4_discard_preallocations(struct inode *, unsigned int); extern int __init ext4_init_mballoc(void); extern void ext4_exit_mballoc(void); extern ext4_group_t ext4_mb_prefetch(struct super_block *sb, diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c index 0eea09aa0f26..a0481582187a 100644 --- a/fs/ext4/extents.c +++ b/fs/ext4/extents.c @@ -100,7 +100,7 @@ static int ext4_ext_trunc_restart_fn(struct inode *inode, int *dropped) * i_mutex. So we can safely drop the i_data_sem here. */ BUG_ON(EXT4_JOURNAL(inode) == NULL); - ext4_discard_preallocations(inode); + ext4_discard_preallocations(inode, 0); up_write(&EXT4_I(inode)->i_data_sem); *dropped = 1; return 0; @@ -4266,7 +4266,7 @@ got_allocated_blocks: * not a good idea to call discard here directly, * but otherwise we'd need to call it every free(). */ - ext4_discard_preallocations(inode); + ext4_discard_preallocations(inode, 0); if (flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE) fb_flags = EXT4_FREE_BLOCKS_NO_QUOT_UPDATE; ext4_free_blocks(handle, inode, NULL, newblock, @@ -5293,7 +5293,7 @@ static int ext4_collapse_range(struct inode *inode, loff_t offset, loff_t len) } down_write(&EXT4_I(inode)->i_data_sem); - ext4_discard_preallocations(inode); + ext4_discard_preallocations(inode, 0); ret = ext4_es_remove_extent(inode, punch_start, EXT_MAX_BLOCKS - punch_start); @@ -5307,7 +5307,7 @@ static int ext4_collapse_range(struct inode *inode, loff_t offset, loff_t len) up_write(&EXT4_I(inode)->i_data_sem); goto out_stop; } - ext4_discard_preallocations(inode); + ext4_discard_preallocations(inode, 0); ret = ext4_ext_shift_extents(inode, handle, punch_stop, punch_stop - punch_start, SHIFT_LEFT); @@ -5439,7 +5439,7 @@ static int ext4_insert_range(struct inode *inode, loff_t offset, loff_t len) goto out_stop; down_write(&EXT4_I(inode)->i_data_sem); - ext4_discard_preallocations(inode); + ext4_discard_preallocations(inode, 0); path = ext4_find_extent(inode, offset_lblk, NULL, 0); if (IS_ERR(path)) { diff --git a/fs/ext4/file.c b/fs/ext4/file.c index 7a2720517bbb..e608ce3fb535 100644 --- a/fs/ext4/file.c +++ b/fs/ext4/file.c @@ -147,7 +147,7 @@ static int ext4_release_file(struct inode *inode, struct file *filp) (atomic_read(&inode->i_writecount) == 1) && !EXT4_I(inode)->i_reserved_data_blocks) { down_write(&EXT4_I(inode)->i_data_sem); - ext4_discard_preallocations(inode); + ext4_discard_preallocations(inode, 0); up_write(&EXT4_I(inode)->i_data_sem); } if (is_dx(inode) && filp->private_data) diff --git a/fs/ext4/indirect.c b/fs/ext4/indirect.c index 433ca8415c5a..80c9f33800be 100644 --- a/fs/ext4/indirect.c +++ b/fs/ext4/indirect.c @@ -696,7 +696,7 @@ static int ext4_ind_trunc_restart_fn(handle_t *handle, struct inode *inode, * i_mutex. So we can safely drop the i_data_sem here. */ BUG_ON(EXT4_JOURNAL(inode) == NULL); - ext4_discard_preallocations(inode); + ext4_discard_preallocations(inode, 0); up_write(&EXT4_I(inode)->i_data_sem); *dropped = 1; return 0; diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 0b07576af3bf..77543f988258 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -383,7 +383,7 @@ void ext4_da_update_reserve_space(struct inode *inode, */ if ((ei->i_reserved_data_blocks == 0) && !inode_is_open_for_write(inode)) - ext4_discard_preallocations(inode); + ext4_discard_preallocations(inode, 0); } static int __check_block_validity(struct inode *inode, const char *func, @@ -4055,7 +4055,7 @@ int ext4_punch_hole(struct inode *inode, loff_t offset, loff_t length) if (stop_block > first_block) { down_write(&EXT4_I(inode)->i_data_sem); - ext4_discard_preallocations(inode); + ext4_discard_preallocations(inode, 0); ret = ext4_es_remove_extent(inode, first_block, stop_block - first_block); @@ -4210,7 +4210,7 @@ int ext4_truncate(struct inode *inode) down_write(&EXT4_I(inode)->i_data_sem); - ext4_discard_preallocations(inode); + ext4_discard_preallocations(inode, 0); if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) err = ext4_ext_truncate(handle, inode); diff --git a/fs/ext4/ioctl.c b/fs/ext4/ioctl.c index 6e70a63dcca7..36eca3bc036a 100644 --- a/fs/ext4/ioctl.c +++ b/fs/ext4/ioctl.c @@ -202,7 +202,7 @@ static long swap_inode_boot_loader(struct super_block *sb, reset_inode_seed(inode); reset_inode_seed(inode_bl); - ext4_discard_preallocations(inode); + ext4_discard_preallocations(inode, 0); err = ext4_mark_inode_dirty(handle, inode); if (err < 0) { diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index 45ac6088b4ac..132c118d12e1 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -2878,6 +2878,7 @@ int ext4_mb_init(struct super_block *sb) sbi->s_mb_stats = MB_DEFAULT_STATS; sbi->s_mb_stream_request = MB_DEFAULT_STREAM_THRESHOLD; sbi->s_mb_order2_reqs = MB_DEFAULT_ORDER2_REQS; + sbi->s_mb_max_inode_prealloc = MB_DEFAULT_MAX_INODE_PREALLOC; /* * The default group preallocation is 512, which for 4k block * sizes translates to 2 megabytes. However for bigalloc file @@ -3816,6 +3817,26 @@ void ext4_mb_generate_from_pa(struct super_block *sb, void *bitmap, mb_debug(sb, "preallocated %d for group %u\n", preallocated, group); } +static void ext4_mb_mark_pa_deleted(struct super_block *sb, + struct ext4_prealloc_space *pa) +{ + struct ext4_inode_info *ei; + + if (pa->pa_deleted) { + ext4_warning(sb, "deleted pa, type:%d, pblk:%llu, lblk:%u, len:%d\n", + pa->pa_type, pa->pa_pstart, pa->pa_lstart, + pa->pa_len); + return; + } + + pa->pa_deleted = 1; + + if (pa->pa_type == MB_INODE_PA) { + ei = EXT4_I(pa->pa_inode); + atomic_dec(&ei->i_prealloc_active); + } +} + static void ext4_mb_pa_callback(struct rcu_head *head) { struct ext4_prealloc_space *pa; @@ -3848,7 +3869,7 @@ static void ext4_mb_put_pa(struct ext4_allocation_context *ac, return; } - pa->pa_deleted = 1; + ext4_mb_mark_pa_deleted(sb, pa); spin_unlock(&pa->pa_lock); grp_blk = pa->pa_pstart; @@ -3972,6 +3993,7 @@ ext4_mb_new_inode_pa(struct ext4_allocation_context *ac) spin_lock(pa->pa_obj_lock); list_add_rcu(&pa->pa_inode_list, &ei->i_prealloc_list); spin_unlock(pa->pa_obj_lock); + atomic_inc(&ei->i_prealloc_active); } /* @@ -4182,7 +4204,7 @@ repeat: } /* seems this one can be freed ... */ - pa->pa_deleted = 1; + ext4_mb_mark_pa_deleted(sb, pa); /* we can trust pa_free ... */ free += pa->pa_free; @@ -4245,7 +4267,7 @@ out_dbg: * * FIXME!! Make sure it is valid at all the call sites */ -void ext4_discard_preallocations(struct inode *inode) +void ext4_discard_preallocations(struct inode *inode, unsigned int needed) { struct ext4_inode_info *ei = EXT4_I(inode); struct super_block *sb = inode->i_sb; @@ -4263,15 +4285,19 @@ void ext4_discard_preallocations(struct inode *inode) mb_debug(sb, "discard preallocation for inode %lu\n", inode->i_ino); - trace_ext4_discard_preallocations(inode); + trace_ext4_discard_preallocations(inode, + atomic_read(&ei->i_prealloc_active), needed); INIT_LIST_HEAD(&list); + if (needed == 0) + needed = UINT_MAX; + repeat: /* first, collect all pa's in the inode */ spin_lock(&ei->i_prealloc_lock); - while (!list_empty(&ei->i_prealloc_list)) { - pa = list_entry(ei->i_prealloc_list.next, + while (!list_empty(&ei->i_prealloc_list) && needed) { + pa = list_entry(ei->i_prealloc_list.prev, struct ext4_prealloc_space, pa_inode_list); BUG_ON(pa->pa_obj_lock != &ei->i_prealloc_lock); spin_lock(&pa->pa_lock); @@ -4288,10 +4314,11 @@ repeat: } if (pa->pa_deleted == 0) { - pa->pa_deleted = 1; + ext4_mb_mark_pa_deleted(sb, pa); spin_unlock(&pa->pa_lock); list_del_rcu(&pa->pa_inode_list); list_add(&pa->u.pa_tmp_list, &list); + needed--; continue; } @@ -4592,7 +4619,7 @@ ext4_mb_discard_lg_preallocations(struct super_block *sb, BUG_ON(pa->pa_type != MB_GROUP_PA); /* seems this one can be freed ... */ - pa->pa_deleted = 1; + ext4_mb_mark_pa_deleted(sb, pa); spin_unlock(&pa->pa_lock); list_del_rcu(&pa->pa_inode_list); @@ -4690,11 +4717,30 @@ static void ext4_mb_add_n_trim(struct ext4_allocation_context *ac) return ; } +/* + * if per-inode prealloc list is too long, trim some PA + */ +static void ext4_mb_trim_inode_pa(struct inode *inode) +{ + struct ext4_inode_info *ei = EXT4_I(inode); + struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); + int count, delta; + + count = atomic_read(&ei->i_prealloc_active); + delta = (sbi->s_mb_max_inode_prealloc >> 2) + 1; + if (count > sbi->s_mb_max_inode_prealloc + delta) { + count -= sbi->s_mb_max_inode_prealloc; + ext4_discard_preallocations(inode, count); + } +} + /* * release all resource we used in allocation */ static int ext4_mb_release_context(struct ext4_allocation_context *ac) { + struct inode *inode = ac->ac_inode; + struct ext4_inode_info *ei = EXT4_I(inode); struct ext4_sb_info *sbi = EXT4_SB(ac->ac_sb); struct ext4_prealloc_space *pa = ac->ac_pa; if (pa) { @@ -4720,6 +4766,17 @@ static int ext4_mb_release_context(struct ext4_allocation_context *ac) ext4_mb_add_n_trim(ac); } } + + if (pa->pa_type == MB_INODE_PA) { + /* + * treat per-inode prealloc list as a lru list, then try + * to trim the least recently used PA. + */ + spin_lock(pa->pa_obj_lock); + list_move(&pa->pa_inode_list, &ei->i_prealloc_list); + spin_unlock(pa->pa_obj_lock); + } + ext4_mb_put_pa(ac, ac->ac_sb, pa); } if (ac->ac_bitmap_page) @@ -4729,6 +4786,7 @@ static int ext4_mb_release_context(struct ext4_allocation_context *ac) if (ac->ac_flags & EXT4_MB_HINT_GROUP_ALLOC) mutex_unlock(&ac->ac_lg->lg_mutex); ext4_mb_collect_stats(ac); + ext4_mb_trim_inode_pa(inode); return 0; } diff --git a/fs/ext4/mballoc.h b/fs/ext4/mballoc.h index 6b4d17c2935d..e75b4749aa1c 100644 --- a/fs/ext4/mballoc.h +++ b/fs/ext4/mballoc.h @@ -73,6 +73,10 @@ */ #define MB_DEFAULT_GROUP_PREALLOC 512 +/* + * maximum length of inode prealloc list + */ +#define MB_DEFAULT_MAX_INODE_PREALLOC 512 struct ext4_free_data { /* this links the free block information from sb_info */ diff --git a/fs/ext4/move_extent.c b/fs/ext4/move_extent.c index 1ed86fb6c302..0d601b822875 100644 --- a/fs/ext4/move_extent.c +++ b/fs/ext4/move_extent.c @@ -686,8 +686,8 @@ ext4_move_extents(struct file *o_filp, struct file *d_filp, __u64 orig_blk, out: if (*moved_len) { - ext4_discard_preallocations(orig_inode); - ext4_discard_preallocations(donor_inode); + ext4_discard_preallocations(orig_inode, 0); + ext4_discard_preallocations(donor_inode, 0); } ext4_ext_drop_refs(path); diff --git a/fs/ext4/super.c b/fs/ext4/super.c index daa94c7f7271..13bdddc081e0 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -1127,6 +1127,7 @@ static struct inode *ext4_alloc_inode(struct super_block *sb) inode_set_iversion(&ei->vfs_inode, 1); spin_lock_init(&ei->i_raw_lock); INIT_LIST_HEAD(&ei->i_prealloc_list); + atomic_set(&ei->i_prealloc_active, 0); spin_lock_init(&ei->i_prealloc_lock); ext4_es_init_tree(&ei->i_es_tree); rwlock_init(&ei->i_es_lock); @@ -1220,7 +1221,7 @@ void ext4_clear_inode(struct inode *inode) { invalidate_inode_buffers(inode); clear_inode(inode); - ext4_discard_preallocations(inode); + ext4_discard_preallocations(inode, 0); ext4_es_remove_extent(inode, 0, EXT_MAX_BLOCKS); dquot_drop(inode); if (EXT4_I(inode)->jinode) { diff --git a/fs/ext4/sysfs.c b/fs/ext4/sysfs.c index 7fee11cc30e7..bfabb799fa45 100644 --- a/fs/ext4/sysfs.c +++ b/fs/ext4/sysfs.c @@ -218,6 +218,7 @@ EXT4_RW_ATTR_SBI_UI(mb_min_to_scan, s_mb_min_to_scan); EXT4_RW_ATTR_SBI_UI(mb_order2_req, s_mb_order2_reqs); EXT4_RW_ATTR_SBI_UI(mb_stream_req, s_mb_stream_request); EXT4_RW_ATTR_SBI_UI(mb_group_prealloc, s_mb_group_prealloc); +EXT4_RW_ATTR_SBI_UI(mb_max_inode_prealloc, s_mb_max_inode_prealloc); EXT4_RW_ATTR_SBI_UI(extent_max_zeroout_kb, s_extent_max_zeroout_kb); EXT4_ATTR(trigger_fs_error, 0200, trigger_test_error); EXT4_RW_ATTR_SBI_UI(err_ratelimit_interval_ms, s_err_ratelimit_state.interval); @@ -264,6 +265,7 @@ static struct attribute *ext4_attrs[] = { ATTR_LIST(mb_order2_req), ATTR_LIST(mb_stream_req), ATTR_LIST(mb_group_prealloc), + ATTR_LIST(mb_max_inode_prealloc), ATTR_LIST(max_writeback_mb_bump), ATTR_LIST(extent_max_zeroout_kb), ATTR_LIST(trigger_fs_error), diff --git a/include/trace/events/ext4.h b/include/trace/events/ext4.h index 8008d2e116b9..4c8b99ec8606 100644 --- a/include/trace/events/ext4.h +++ b/include/trace/events/ext4.h @@ -746,24 +746,29 @@ TRACE_EVENT(ext4_mb_release_group_pa, ); TRACE_EVENT(ext4_discard_preallocations, - TP_PROTO(struct inode *inode), + TP_PROTO(struct inode *inode, unsigned int len, unsigned int needed), - TP_ARGS(inode), + TP_ARGS(inode, len, needed), TP_STRUCT__entry( - __field( dev_t, dev ) - __field( ino_t, ino ) + __field( dev_t, dev ) + __field( ino_t, ino ) + __field( unsigned int, len ) + __field( unsigned int, needed ) ), TP_fast_assign( __entry->dev = inode->i_sb->s_dev; __entry->ino = inode->i_ino; + __entry->len = len; + __entry->needed = needed; ), - TP_printk("dev %d,%d ino %lu", + TP_printk("dev %d,%d ino %lu len: %u needed %u", MAJOR(__entry->dev), MINOR(__entry->dev), - (unsigned long) __entry->ino) + (unsigned long) __entry->ino, __entry->len, + __entry->needed) ); TRACE_EVENT(ext4_mb_discard_preallocations, -- cgit v1.2.3