From 7d21cde893a05f22a145bf5c2902c84035a048bc Mon Sep 17 00:00:00 2001 From: Eric Biggers Date: Thu, 5 Mar 2020 00:41:38 -0800 Subject: [PATCH 01/86] fscrypt: don't evict dirty inodes after removing key After FS_IOC_REMOVE_ENCRYPTION_KEY removes a key, it syncs the filesystem and tries to get and put all inodes that were unlocked by the key so that unused inodes get evicted via fscrypt_drop_inode(). Normally, the inodes are all clean due to the sync. However, after the filesystem is sync'ed, userspace can modify and close one of the files. (Userspace is *supposed* to close the files before removing the key. But it doesn't always happen, and the kernel can't assume it.) This causes the inode to be dirtied and have i_count == 0. Then, fscrypt_drop_inode() failed to consider this case and indicated that the inode can be dropped, causing the write to be lost. On f2fs, other problems such as a filesystem freeze could occur due to the inode being freed while still on f2fs's dirty inode list. Fix this bug by making fscrypt_drop_inode() only drop clean inodes. I've written an xfstest which detects this bug on ext4, f2fs, and ubifs. Fixes: b1c0ec3599f4 ("fscrypt: add FS_IOC_REMOVE_ENCRYPTION_KEY ioctl") Cc: # v5.4+ Link: https://lore.kernel.org/r/20200305084138.653498-1-ebiggers@kernel.org Signed-off-by: Eric Biggers --- fs/crypto/keysetup.c | 9 +++++++++ 1 file changed, 9 insertions(+) diff --git a/fs/crypto/keysetup.c b/fs/crypto/keysetup.c index 65cb09fa6ead..08c9f216a54d 100644 --- a/fs/crypto/keysetup.c +++ b/fs/crypto/keysetup.c @@ -538,6 +538,15 @@ int fscrypt_drop_inode(struct inode *inode) return 0; mk = ci->ci_master_key->payload.data[0]; + /* + * With proper, non-racy use of FS_IOC_REMOVE_ENCRYPTION_KEY, all inodes + * protected by the key were cleaned by sync_filesystem(). But if + * userspace is still using the files, inodes can be dirtied between + * then and now. We mustn't lose any writes, so skip dirty inodes here. + */ + if (inode->i_state & I_DIRTY_ALL) + return 0; + /* * Note: since we aren't holding ->mk_secret_sem, the result here can * immediately become outdated. But there's no correctness problem with From 30ccfb38e9dfe46be4c34e226cf3ff5f1380e598 Mon Sep 17 00:00:00 2001 From: Naohiro Aota Date: Sat, 28 Mar 2020 00:28:08 -0700 Subject: [PATCH 02/86] mm/swapfile.c: move inode_lock out of claim_swapfile claim_swapfile() currently keeps the inode locked when it is successful, or the file is already swapfile (with -ebusy). and, on the other error cases, it does not lock the inode. this inconsistency of the lock state and return value is quite confusing and actually causing a bad unlock balance as below in the "bad_swap" section of __do_sys_swapon(). this commit fixes this issue by moving the inode_lock() and is_swapfile check out of claim_swapfile(). the inode is unlocked in "bad_swap_unlock_inode" section, so that the inode is ensured to be unlocked at "bad_swap". thus, error handling codes after the locking now jumps to "bad_swap_unlock_inode" instead of "bad_swap". ===================================== warning: bad unlock balance detected! 5.5.0-rc7+ #176 not tainted ------------------------------------- swapon/4294 is trying to release lock (&sb->s_type->i_mutex_key) at: [] __do_sys_swapon+0x94b/0x3550 but there are no more locks to release! other info that might help us debug this: no locks held by swapon/4294. stack backtrace: cpu: 5 pid: 4294 comm: swapon not tainted 5.5.0-rc7-btrfs-zns+ #176 hardware name: asus all series/h87-pro, bios 2102 07/29/2014 call trace: dump_stack+0xa1/0xea ? __do_sys_swapon+0x94b/0x3550 print_unlock_imbalance_bug.cold+0x114/0x123 ? __do_sys_swapon+0x94b/0x3550 lock_release+0x562/0xed0 ? kvfree+0x31/0x40 ? lock_downgrade+0x770/0x770 ? kvfree+0x31/0x40 ? rcu_read_lock_sched_held+0xa1/0xd0 ? rcu_read_lock_bh_held+0xb0/0xb0 up_write+0x2d/0x490 ? kfree+0x293/0x2f0 __do_sys_swapon+0x94b/0x3550 ? putname+0xb0/0xf0 ? kmem_cache_free+0x2e7/0x370 ? do_sys_open+0x184/0x3e0 ? generic_max_swapfile_size+0x40/0x40 ? do_syscall_64+0x27/0x4b0 ? entry_syscall_64_after_hwframe+0x49/0xbe ? lockdep_hardirqs_on+0x38c/0x590 __x64_sys_swapon+0x54/0x80 do_syscall_64+0xa4/0x4b0 entry_syscall_64_after_hwframe+0x49/0xbe rip: 0033:0x7f15da0a0dc7 link: http://lkml.kernel.org/r/20200206090132.154869-1-naohiro.aota@wdc.com fixes: 1638045c3677 ("mm: set s_swapfile on blockdev swap devices") signed-off-by: naohiro aota reviewed-by: andrew morton reviewed-by: darrick j. wong tested-by: qais youef cc: christoph hellwig cc: --- mm/swapfile.c | 40 +++++++++++++++++++--------------------- 1 file changed, 19 insertions(+), 21 deletions(-) diff --git a/mm/swapfile.c b/mm/swapfile.c index dab43523afdd..a7de92334b44 100644 --- a/mm/swapfile.c +++ b/mm/swapfile.c @@ -2892,10 +2892,6 @@ static int claim_swapfile(struct swap_info_struct *p, struct inode *inode) p->bdev = inode->i_sb->s_bdev; } - inode_lock(inode); - if (IS_SWAPFILE(inode)) - return -EBUSY; - return 0; } @@ -3150,36 +3146,40 @@ SYSCALL_DEFINE2(swapon, const char __user *, specialfile, int, swap_flags) mapping = swap_file->f_mapping; inode = mapping->host; - /* If S_ISREG(inode->i_mode) will do inode_lock(inode); */ error = claim_swapfile(p, inode); if (unlikely(error)) goto bad_swap; + inode_lock(inode); + if (IS_SWAPFILE(inode)) { + error = -EBUSY; + goto bad_swap_unlock_inode; + } /* * Read the swap header. */ if (!mapping->a_ops->readpage) { error = -EINVAL; - goto bad_swap; + goto bad_swap_unlock_inode; } page = read_mapping_page(mapping, 0, swap_file); if (IS_ERR(page)) { error = PTR_ERR(page); - goto bad_swap; + goto bad_swap_unlock_inode; } swap_header = kmap(page); maxpages = read_swap_header(p, swap_header, inode); if (unlikely(!maxpages)) { error = -EINVAL; - goto bad_swap; + goto bad_swap_unlock_inode; } /* OK, set up the swap map and apply the bad block list */ swap_map = vzalloc(maxpages); if (!swap_map) { error = -ENOMEM; - goto bad_swap; + goto bad_swap_unlock_inode; } if (bdi_cap_stable_pages_required(inode_to_bdi(inode))) @@ -3204,7 +3204,7 @@ SYSCALL_DEFINE2(swapon, const char __user *, specialfile, int, swap_flags) GFP_KERNEL); if (!cluster_info) { error = -ENOMEM; - goto bad_swap; + goto bad_swap_unlock_inode; } for (ci = 0; ci < nr_cluster; ci++) @@ -3213,7 +3213,7 @@ SYSCALL_DEFINE2(swapon, const char __user *, specialfile, int, swap_flags) p->percpu_cluster = alloc_percpu(struct percpu_cluster); if (!p->percpu_cluster) { error = -ENOMEM; - goto bad_swap; + goto bad_swap_unlock_inode; } for_each_possible_cpu(cpu) { struct percpu_cluster *cluster; @@ -3227,13 +3227,13 @@ SYSCALL_DEFINE2(swapon, const char __user *, specialfile, int, swap_flags) error = swap_cgroup_swapon(p->type, maxpages); if (error) - goto bad_swap; + goto bad_swap_unlock_inode; nr_extents = setup_swap_map_and_extents(p, swap_header, swap_map, cluster_info, maxpages, &span); if (unlikely(nr_extents < 0)) { error = nr_extents; - goto bad_swap; + goto bad_swap_unlock_inode; } /* frontswap enabled? set up bit-per-page map for frontswap */ if (IS_ENABLED(CONFIG_FRONTSWAP)) @@ -3273,7 +3273,7 @@ SYSCALL_DEFINE2(swapon, const char __user *, specialfile, int, swap_flags) error = init_swap_address_space(p->type, maxpages); if (error) - goto bad_swap; + goto bad_swap_unlock_inode; /* * Flush any pending IO and dirty mappings before we start using this @@ -3283,7 +3283,7 @@ SYSCALL_DEFINE2(swapon, const char __user *, specialfile, int, swap_flags) error = inode_drain_writes(inode); if (error) { inode->i_flags &= ~S_SWAPFILE; - goto bad_swap; + goto bad_swap_unlock_inode; } mutex_lock(&swapon_mutex); @@ -3308,6 +3308,8 @@ SYSCALL_DEFINE2(swapon, const char __user *, specialfile, int, swap_flags) error = 0; goto out; +bad_swap_unlock_inode: + inode_unlock(inode); bad_swap: free_percpu(p->percpu_cluster); p->percpu_cluster = NULL; @@ -3315,6 +3317,7 @@ bad_swap: set_blocksize(p->bdev, p->old_block_size); blkdev_put(p->bdev, FMODE_READ | FMODE_WRITE | FMODE_EXCL); } + inode = NULL; destroy_swap_extents(p); swap_cgroup_swapoff(p->type); spin_lock(&swap_lock); @@ -3326,13 +3329,8 @@ bad_swap: kvfree(frontswap_map); if (inced_nr_rotate_swap) atomic_dec(&nr_rotate_swap); - if (swap_file) { - if (inode) { - inode_unlock(inode); - inode = NULL; - } + if (swap_file) filp_close(swap_file, NULL); - } out: if (page && !IS_ERR(page)) { kunmap(page); From 7e79ae46f49857778a203fbd0d572a5cbc3a407b Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Fri, 14 Feb 2020 17:45:12 +0800 Subject: [PATCH 03/86] f2fs: fix to wait all node page writeback There is a race condition that we may miss to wait for all node pages writeback, fix it. - fsync() - shrink - f2fs_do_sync_file - __write_node_page - set_page_writeback(page#0) : remove DIRTY/TOWRITE flag - f2fs_fsync_node_pages : won't find page #0 as TOWRITE flag was removeD - f2fs_wait_on_node_pages_writeback : wont' wait page #0 writeback as it was not in fsync_node_list list. - f2fs_add_fsync_node_entry Fixes: 50fa53eccf9f ("f2fs: fix to avoid broken of dnode block list") Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/node.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/fs/f2fs/node.c b/fs/f2fs/node.c index 3314a0f3405e..905e8e39c9bf 100644 --- a/fs/f2fs/node.c +++ b/fs/f2fs/node.c @@ -1562,15 +1562,16 @@ static int __write_node_page(struct page *page, bool atomic, bool *submitted, if (atomic && !test_opt(sbi, NOBARRIER)) fio.op_flags |= REQ_PREFLUSH | REQ_FUA; - set_page_writeback(page); - ClearPageError(page); - + /* should add to global list before clearing PAGECACHE status */ if (f2fs_in_warm_node_list(sbi, page)) { seq = f2fs_add_fsync_node_entry(sbi, page); if (seq_id) *seq_id = seq; } + set_page_writeback(page); + ClearPageError(page); + fio.old_blkaddr = ni.blk_addr; f2fs_do_write_node_page(nid, &fio); set_node_addr(sbi, &ni, fio.new_blkaddr, is_fsync_dnode(page)); From 49996449d3b78639978879e595e5abb4e4cc46be Mon Sep 17 00:00:00 2001 From: Sahitya Tummala Date: Tue, 18 Feb 2020 09:19:07 +0530 Subject: [PATCH 04/86] f2fs: fix the panic in do_checkpoint() There could be a scenario where f2fs_sync_meta_pages() will not ensure that all F2FS_DIRTY_META pages are submitted for IO. Thus, resulting in the below panic in do_checkpoint() - f2fs_bug_on(sbi, get_pages(sbi, F2FS_DIRTY_META) && !f2fs_cp_error(sbi)); This can happen in a low-memory condition, where shrinker could also be doing the writepage operation (stack shown below) at the same time when checkpoint is running on another core. schedule down_write f2fs_submit_page_write -> by this time, this page in page cache is tagged as PAGECACHE_TAG_WRITEBACK and PAGECACHE_TAG_DIRTY is cleared, due to which f2fs_sync_meta_pages() cannot sync this page in do_checkpoint() path. f2fs_do_write_meta_page __f2fs_write_meta_page f2fs_write_meta_page shrink_page_list shrink_inactive_list shrink_node_memcg shrink_node kswapd Signed-off-by: Sahitya Tummala Reviewed-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/checkpoint.c | 16 +++++++--------- fs/f2fs/f2fs.h | 2 +- fs/f2fs/super.c | 2 +- 3 files changed, 9 insertions(+), 11 deletions(-) diff --git a/fs/f2fs/checkpoint.c b/fs/f2fs/checkpoint.c index 44e84ac5c941..9c88fb3d255a 100644 --- a/fs/f2fs/checkpoint.c +++ b/fs/f2fs/checkpoint.c @@ -1250,20 +1250,20 @@ static void unblock_operations(struct f2fs_sb_info *sbi) f2fs_unlock_all(sbi); } -void f2fs_wait_on_all_pages_writeback(struct f2fs_sb_info *sbi) +void f2fs_wait_on_all_pages(struct f2fs_sb_info *sbi, int type) { DEFINE_WAIT(wait); for (;;) { prepare_to_wait(&sbi->cp_wait, &wait, TASK_UNINTERRUPTIBLE); - if (!get_pages(sbi, F2FS_WB_CP_DATA)) + if (!get_pages(sbi, type)) break; if (unlikely(f2fs_cp_error(sbi))) break; - io_schedule_timeout(5*HZ); + io_schedule_timeout(HZ/50); } finish_wait(&sbi->cp_wait, &wait); } @@ -1384,8 +1384,6 @@ static int do_checkpoint(struct f2fs_sb_info *sbi, struct cp_control *cpc) /* Flush all the NAT/SIT pages */ f2fs_sync_meta_pages(sbi, META, LONG_MAX, FS_CP_META_IO); - f2fs_bug_on(sbi, get_pages(sbi, F2FS_DIRTY_META) && - !f2fs_cp_error(sbi)); /* * modify checkpoint @@ -1493,11 +1491,11 @@ static int do_checkpoint(struct f2fs_sb_info *sbi, struct cp_control *cpc) /* Here, we have one bio having CP pack except cp pack 2 page */ f2fs_sync_meta_pages(sbi, META, LONG_MAX, FS_CP_META_IO); - f2fs_bug_on(sbi, get_pages(sbi, F2FS_DIRTY_META) && - !f2fs_cp_error(sbi)); + /* Wait for all dirty meta pages to be submitted for IO */ + f2fs_wait_on_all_pages(sbi, F2FS_DIRTY_META); /* wait for previous submitted meta pages writeback */ - f2fs_wait_on_all_pages_writeback(sbi); + f2fs_wait_on_all_pages(sbi, F2FS_WB_CP_DATA); /* flush all device cache */ err = f2fs_flush_device_cache(sbi); @@ -1506,7 +1504,7 @@ static int do_checkpoint(struct f2fs_sb_info *sbi, struct cp_control *cpc) /* barrier and flush checkpoint cp pack 2 page if it can */ commit_checkpoint(sbi, ckpt, start_blk); - f2fs_wait_on_all_pages_writeback(sbi); + f2fs_wait_on_all_pages(sbi, F2FS_WB_CP_DATA); /* * invalidate intermediate page cache borrowed from meta inode which are diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index 5355be6b6755..d39f5de11420 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -3308,7 +3308,7 @@ int f2fs_get_valid_checkpoint(struct f2fs_sb_info *sbi); void f2fs_update_dirty_page(struct inode *inode, struct page *page); void f2fs_remove_dirty_inode(struct inode *inode); int f2fs_sync_dirty_inodes(struct f2fs_sb_info *sbi, enum inode_type type); -void f2fs_wait_on_all_pages_writeback(struct f2fs_sb_info *sbi); +void f2fs_wait_on_all_pages(struct f2fs_sb_info *sbi, int type); int f2fs_write_checkpoint(struct f2fs_sb_info *sbi, struct cp_control *cpc); void f2fs_init_ino_entry_info(struct f2fs_sb_info *sbi); int __init f2fs_create_checkpoint_caches(void); diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c index 83df68509703..490eafe8645e 100644 --- a/fs/f2fs/super.c +++ b/fs/f2fs/super.c @@ -1172,7 +1172,7 @@ static void f2fs_put_super(struct super_block *sb) /* our cp_error case, we can wait for any writeback page */ f2fs_flush_merged_writes(sbi); - f2fs_wait_on_all_pages_writeback(sbi); + f2fs_wait_on_all_pages(sbi, F2FS_WB_CP_DATA); f2fs_bug_on(sbi, sbi->fsync_node_num); From 4bd833dbcf8b9418c2962779a798c669ac5f979b Mon Sep 17 00:00:00 2001 From: Eric Biggers Date: Thu, 20 Feb 2020 20:50:37 -0800 Subject: [PATCH 05/86] f2fs: fix leaking uninitialized memory in compressed clusters When the compressed data of a cluster doesn't end on a page boundary, the remainder of the last page must be zeroed in order to avoid leaking uninitialized memory to disk. Fixes: 4c8ff7095bef ("f2fs: support data compression") Signed-off-by: Eric Biggers Reviewed-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/compress.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/fs/f2fs/compress.c b/fs/f2fs/compress.c index d8a64be90a50..ef7dd04312fe 100644 --- a/fs/f2fs/compress.c +++ b/fs/f2fs/compress.c @@ -385,11 +385,15 @@ static int f2fs_compress_pages(struct compress_ctx *cc) for (i = 0; i < COMPRESS_DATA_RESERVED_SIZE; i++) cc->cbuf->reserved[i] = cpu_to_le32(0); + nr_cpages = DIV_ROUND_UP(cc->clen + COMPRESS_HEADER_SIZE, PAGE_SIZE); + + /* zero out any unused part of the last page */ + memset(&cc->cbuf->cdata[cc->clen], 0, + (nr_cpages * PAGE_SIZE) - (cc->clen + COMPRESS_HEADER_SIZE)); + vunmap(cc->cbuf); vunmap(cc->rbuf); - nr_cpages = DIV_ROUND_UP(cc->clen + COMPRESS_HEADER_SIZE, PAGE_SIZE); - for (i = nr_cpages; i < cc->nr_cpages; i++) { f2fs_put_compressed_page(cc->cpages[i]); cc->cpages[i] = NULL; From 56ec790bb7445cbe79e41bd8ec136df295f589e5 Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Mon, 24 Feb 2020 19:20:17 +0800 Subject: [PATCH 06/86] f2fs: fix to avoid NULL pointer dereference Unable to handle kernel NULL pointer dereference at virtual address 00000000 PC is at f2fs_free_dic+0x60/0x2c8 LR is at f2fs_decompress_pages+0x3c4/0x3e8 f2fs_free_dic+0x60/0x2c8 f2fs_decompress_pages+0x3c4/0x3e8 __read_end_io+0x78/0x19c f2fs_post_read_work+0x6c/0x94 process_one_work+0x210/0x48c worker_thread+0x2e8/0x44c kthread+0x110/0x120 ret_from_fork+0x10/0x18 In f2fs_free_dic(), we can not use f2fs_put_page(,1) to release dic->tpages[i], as the page's mapping is NULL. Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/compress.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/fs/f2fs/compress.c b/fs/f2fs/compress.c index ef7dd04312fe..dde8ca08e94d 100644 --- a/fs/f2fs/compress.c +++ b/fs/f2fs/compress.c @@ -1137,7 +1137,8 @@ void f2fs_free_dic(struct decompress_io_ctx *dic) for (i = 0; i < dic->cluster_size; i++) { if (dic->rpages[i]) continue; - f2fs_put_page(dic->tpages[i], 1); + unlock_page(dic->tpages[i]); + put_page(dic->tpages[i]); } kfree(dic->tpages); } From 225a27719d9574f9ec71e70488770d0611758e56 Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Mon, 24 Feb 2020 19:20:18 +0800 Subject: [PATCH 07/86] f2fs: recycle unused compress_data.chksum feild In Struct compress_data, chksum field was never used, remove it. Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/compress.c | 1 - fs/f2fs/f2fs.h | 3 +-- 2 files changed, 1 insertion(+), 3 deletions(-) diff --git a/fs/f2fs/compress.c b/fs/f2fs/compress.c index dde8ca08e94d..5ca2ea0e1e45 100644 --- a/fs/f2fs/compress.c +++ b/fs/f2fs/compress.c @@ -380,7 +380,6 @@ static int f2fs_compress_pages(struct compress_ctx *cc) } cc->cbuf->clen = cpu_to_le32(cc->clen); - cc->cbuf->chksum = cpu_to_le32(0); for (i = 0; i < COMPRESS_DATA_RESERVED_SIZE; i++) cc->cbuf->reserved[i] = cpu_to_le32(0); diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index d39f5de11420..16ac00f01586 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -1214,10 +1214,9 @@ enum compress_algorithm_type { COMPRESS_MAX, }; -#define COMPRESS_DATA_RESERVED_SIZE 4 +#define COMPRESS_DATA_RESERVED_SIZE 5 struct compress_data { __le32 clen; /* compressed data size */ - __le32 chksum; /* checksum of compressed data */ __le32 reserved[COMPRESS_DATA_RESERVED_SIZE]; /* reserved */ u8 cdata[]; /* compressed data */ }; From fff4770044d41efe322d7814c22e21a00e1b5175 Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Mon, 24 Feb 2020 19:20:19 +0800 Subject: [PATCH 08/86] f2fs: add missing function name in kernel message Otherwise, we can not distinguish the exact location of messages, when there are more than one places printing same message. Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/f2fs.h | 2 +- fs/f2fs/node.c | 5 +++-- 2 files changed, 4 insertions(+), 3 deletions(-) diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index 16ac00f01586..a5656dd00bc2 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -2209,7 +2209,7 @@ static inline void dec_valid_node_count(struct f2fs_sb_info *sbi, dquot_free_inode(inode); } else { if (unlikely(inode->i_blocks == 0)) { - f2fs_warn(sbi, "Inconsistent i_blocks, ino:%lu, iblocks:%llu", + f2fs_warn(sbi, "dec_valid_node_count: inconsistent i_blocks, ino:%lu, iblocks:%llu", inode->i_ino, (unsigned long long)inode->i_blocks); set_sbi_flag(sbi, SBI_NEED_FSCK); diff --git a/fs/f2fs/node.c b/fs/f2fs/node.c index 905e8e39c9bf..75a3b2c4dc55 100644 --- a/fs/f2fs/node.c +++ b/fs/f2fs/node.c @@ -1188,8 +1188,9 @@ int f2fs_remove_inode_page(struct inode *inode) } if (unlikely(inode->i_blocks != 0 && inode->i_blocks != 8)) { - f2fs_warn(F2FS_I_SB(inode), "Inconsistent i_blocks, ino:%lu, iblocks:%llu", - inode->i_ino, (unsigned long long)inode->i_blocks); + f2fs_warn(F2FS_I_SB(inode), + "f2fs_remove_inode_page: inconsistent i_blocks, ino:%lu, iblocks:%llu", + inode->i_ino, (unsigned long long)inode->i_blocks); set_sbi_flag(F2FS_I_SB(inode), SBI_NEED_FSCK); } From c014fe3982aaf4781f71a12d57999365c0886984 Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Mon, 24 Feb 2020 19:20:16 +0800 Subject: [PATCH 09/86] f2fs: fix to avoid potential deadlock Using f2fs_trylock_op() in f2fs_write_compressed_pages() to avoid potential deadlock like we did in f2fs_write_single_data_page(). Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/compress.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/fs/f2fs/compress.c b/fs/f2fs/compress.c index 5ca2ea0e1e45..6da33ba2b2db 100644 --- a/fs/f2fs/compress.c +++ b/fs/f2fs/compress.c @@ -775,7 +775,6 @@ static int f2fs_write_compressed_pages(struct compress_ctx *cc, .encrypted_page = NULL, .compressed_page = NULL, .submitted = false, - .need_lock = LOCK_RETRY, .io_type = io_type, .io_wbc = wbc, .encrypted = f2fs_encrypted_file(cc->inode), @@ -788,9 +787,10 @@ static int f2fs_write_compressed_pages(struct compress_ctx *cc, loff_t psize; int i, err; - set_new_dnode(&dn, cc->inode, NULL, NULL, 0); + if (!f2fs_trylock_op(sbi)) + return -EAGAIN; - f2fs_lock_op(sbi); + set_new_dnode(&dn, cc->inode, NULL, NULL, 0); err = f2fs_get_dnode_of_data(&dn, start_idx, LOOKUP_NODE); if (err) From 10ed1e2397d335da3d52d0944a7e0fe3d0154ac9 Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Tue, 25 Feb 2020 18:26:46 +0800 Subject: [PATCH 10/86] f2fs: fix to check i_compr_blocks correctly inode.i_blocks counts based on 512byte sector, we need to convert to 4kb sized block count before comparing to i_compr_blocks. In addition, add to print message when sanity check on inode compression configs failed. Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/inode.c | 23 ++++++++++++++++++++--- 1 file changed, 20 insertions(+), 3 deletions(-) diff --git a/fs/f2fs/inode.c b/fs/f2fs/inode.c index 78c3f1d70f1d..9b49832296cf 100644 --- a/fs/f2fs/inode.c +++ b/fs/f2fs/inode.c @@ -291,13 +291,30 @@ static bool sanity_check_inode(struct inode *inode, struct page *node_page) fi->i_flags & F2FS_COMPR_FL && F2FS_FITS_IN_INODE(ri, fi->i_extra_isize, i_log_cluster_size)) { - if (ri->i_compress_algorithm >= COMPRESS_MAX) + if (ri->i_compress_algorithm >= COMPRESS_MAX) { + f2fs_warn(sbi, "%s: inode (ino=%lx) has unsupported " + "compress algorithm: %u, run fsck to fix", + __func__, inode->i_ino, + ri->i_compress_algorithm); return false; - if (le64_to_cpu(ri->i_compr_blocks) > inode->i_blocks) + } + if (le64_to_cpu(ri->i_compr_blocks) > + SECTOR_TO_BLOCK(inode->i_blocks)) { + f2fs_warn(sbi, "%s: inode (ino=%lx) has inconsistent " + "i_compr_blocks:%llu, i_blocks:%llu, run fsck to fix", + __func__, inode->i_ino, + le64_to_cpu(ri->i_compr_blocks), + SECTOR_TO_BLOCK(inode->i_blocks)); return false; + } if (ri->i_log_cluster_size < MIN_COMPRESS_LOG_SIZE || - ri->i_log_cluster_size > MAX_COMPRESS_LOG_SIZE) + ri->i_log_cluster_size > MAX_COMPRESS_LOG_SIZE) { + f2fs_warn(sbi, "%s: inode (ino=%lx) has unsupported " + "log cluster size: %u, run fsck to fix", + __func__, inode->i_ino, + ri->i_log_cluster_size); return false; + } } return true; From 67276e0286e0507369233c8122fca19d3b74a795 Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Thu, 27 Feb 2020 19:30:03 +0800 Subject: [PATCH 11/86] f2fs: cover last_disk_size update with spinlock MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This change solves below hangtask issue: INFO: task kworker/u16:1:58 blocked for more than 122 seconds. Not tainted 5.6.0-rc2-00590-g9983bdae4974e #11 "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. kworker/u16:1 D 0 58 2 0x00000000 Workqueue: writeback wb_workfn (flush-179:0) Backtrace: (__schedule) from [] (schedule+0x78/0xf4) (schedule) from [] (rwsem_down_write_slowpath+0x24c/0x4c0) (rwsem_down_write_slowpath) from [] (down_write+0x6c/0x70) (down_write) from [] (f2fs_write_single_data_page+0x608/0x7ac) (f2fs_write_single_data_page) from [] (f2fs_write_cache_pages+0x2b4/0x7c4) (f2fs_write_cache_pages) from [] (f2fs_write_data_pages+0x344/0x35c) (f2fs_write_data_pages) from [] (do_writepages+0x3c/0xd4) (do_writepages) from [] (__writeback_single_inode+0x44/0x454) (__writeback_single_inode) from [] (writeback_sb_inodes+0x204/0x4b0) (writeback_sb_inodes) from [] (__writeback_inodes_wb+0x50/0xe4) (__writeback_inodes_wb) from [] (wb_writeback+0x294/0x338) (wb_writeback) from [] (wb_workfn+0x35c/0x54c) (wb_workfn) from [] (process_one_work+0x214/0x544) (process_one_work) from [] (worker_thread+0x4c/0x574) (worker_thread) from [] (kthread+0x144/0x170) (kthread) from [] (ret_from_fork+0x14/0x2c) Reported-and-tested-by: Ondřej Jirman Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/compress.c | 4 ++-- fs/f2fs/data.c | 4 ++-- fs/f2fs/f2fs.h | 5 +++-- fs/f2fs/file.c | 4 ++-- fs/f2fs/super.c | 1 + 5 files changed, 10 insertions(+), 8 deletions(-) diff --git a/fs/f2fs/compress.c b/fs/f2fs/compress.c index 6da33ba2b2db..25bc6154c31c 100644 --- a/fs/f2fs/compress.c +++ b/fs/f2fs/compress.c @@ -898,10 +898,10 @@ unlock_continue: f2fs_put_dnode(&dn); f2fs_unlock_op(sbi); - down_write(&fi->i_sem); + spin_lock(&fi->i_size_lock); if (fi->last_disk_size < psize) fi->last_disk_size = psize; - up_write(&fi->i_sem); + spin_unlock(&fi->i_size_lock); f2fs_put_rpages(cc); f2fs_destroy_compress_ctx(cc); diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c index 8bd9afa81c54..5f4948bc40e9 100644 --- a/fs/f2fs/data.c +++ b/fs/f2fs/data.c @@ -2647,10 +2647,10 @@ write: if (err) { file_set_keep_isize(inode); } else { - down_write(&F2FS_I(inode)->i_sem); + spin_lock(&F2FS_I(inode)->i_size_lock); if (F2FS_I(inode)->last_disk_size < psize) F2FS_I(inode)->last_disk_size = psize; - up_write(&F2FS_I(inode)->i_sem); + spin_unlock(&F2FS_I(inode)->i_size_lock); } done: diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index a5656dd00bc2..43900c1e34fc 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -696,6 +696,7 @@ struct f2fs_inode_info { struct task_struct *cp_task; /* separate cp/wb IO stats*/ nid_t i_xattr_nid; /* node id that contains xattrs */ loff_t last_disk_size; /* lastly written file size */ + spinlock_t i_size_lock; /* protect last_disk_size */ #ifdef CONFIG_QUOTA struct dquot *i_dquot[MAXQUOTAS]; @@ -2855,9 +2856,9 @@ static inline bool f2fs_skip_inode_update(struct inode *inode, int dsync) if (!f2fs_is_time_consistent(inode)) return false; - down_read(&F2FS_I(inode)->i_sem); + spin_lock(&F2FS_I(inode)->i_size_lock); ret = F2FS_I(inode)->last_disk_size == i_size_read(inode); - up_read(&F2FS_I(inode)->i_sem); + spin_unlock(&F2FS_I(inode)->i_size_lock); return ret; } diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c index 2362f147efc3..5adda28a73a0 100644 --- a/fs/f2fs/file.c +++ b/fs/f2fs/file.c @@ -935,10 +935,10 @@ int f2fs_setattr(struct dentry *dentry, struct iattr *attr) if (err) return err; - down_write(&F2FS_I(inode)->i_sem); + spin_lock(&F2FS_I(inode)->i_size_lock); inode->i_mtime = inode->i_ctime = current_time(inode); F2FS_I(inode)->last_disk_size = i_size_read(inode); - up_write(&F2FS_I(inode)->i_sem); + spin_unlock(&F2FS_I(inode)->i_size_lock); } __setattr_copy(inode, attr); diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c index 490eafe8645e..456e6a1d703c 100644 --- a/fs/f2fs/super.c +++ b/fs/f2fs/super.c @@ -960,6 +960,7 @@ static struct inode *f2fs_alloc_inode(struct super_block *sb) /* Initialize f2fs-specific inode info */ atomic_set(&fi->dirty_pages, 0); init_rwsem(&fi->i_sem); + spin_lock_init(&fi->i_size_lock); INIT_LIST_HEAD(&fi->dirty_list); INIT_LIST_HEAD(&fi->gdirty_list); INIT_LIST_HEAD(&fi->inmem_ilist); From e89b741a74f7b53b74415e724e69577ef87e6fc7 Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Thu, 27 Feb 2020 19:30:04 +0800 Subject: [PATCH 12/86] f2fs: remove i_sem lock coverage in f2fs_setxattr() f2fs_inode.xattr_ver field was gone after commit d260081ccf37 ("f2fs: change recovery policy of xattr node block"), remove i_sem lock coverage in f2fs_setxattr() Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/xattr.c | 3 --- 1 file changed, 3 deletions(-) diff --git a/fs/f2fs/xattr.c b/fs/f2fs/xattr.c index 296b3189448a..6794a13064be 100644 --- a/fs/f2fs/xattr.c +++ b/fs/f2fs/xattr.c @@ -758,12 +758,9 @@ int f2fs_setxattr(struct inode *inode, int index, const char *name, f2fs_balance_fs(sbi, true); f2fs_lock_op(sbi); - /* protect xattr_ver */ - down_write(&F2FS_I(inode)->i_sem); down_write(&F2FS_I(inode)->i_xattr_sem); err = __f2fs_setxattr(inode, index, name, value, size, ipage, flags); up_write(&F2FS_I(inode)->i_xattr_sem); - up_write(&F2FS_I(inode)->i_sem); f2fs_unlock_op(sbi); f2fs_update_time(sbi, REQ_TIME); From 271ca0e6d05d8b77270e915caaf52cd798fb16c6 Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Thu, 27 Feb 2020 19:30:05 +0800 Subject: [PATCH 13/86] f2fs: fix inconsistent comments Lack of maintenance on comments may mislead developers, fix them. Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/checkpoint.c | 18 ++++-------------- fs/f2fs/data.c | 19 ++++++------------- fs/f2fs/f2fs.h | 2 +- fs/f2fs/file.c | 1 - fs/f2fs/gc.c | 5 ++++- fs/f2fs/inode.c | 2 +- fs/f2fs/namei.c | 2 +- fs/f2fs/node.c | 6 +----- fs/f2fs/shrinker.c | 2 +- fs/f2fs/super.c | 4 ++-- 10 files changed, 21 insertions(+), 40 deletions(-) diff --git a/fs/f2fs/checkpoint.c b/fs/f2fs/checkpoint.c index 9c88fb3d255a..0dea31f1ddad 100644 --- a/fs/f2fs/checkpoint.c +++ b/fs/f2fs/checkpoint.c @@ -50,9 +50,6 @@ repeat: return page; } -/* - * We guarantee no failure on the returned page. - */ static struct page *__get_meta_page(struct f2fs_sb_info *sbi, pgoff_t index, bool is_meta) { @@ -206,7 +203,7 @@ bool f2fs_is_valid_blkaddr(struct f2fs_sb_info *sbi, } /* - * Readahead CP/NAT/SIT/SSA pages + * Readahead CP/NAT/SIT/SSA/POR pages */ int f2fs_ra_meta_pages(struct f2fs_sb_info *sbi, block_t start, int nrpages, int type, bool sync) @@ -898,7 +895,7 @@ int f2fs_get_valid_checkpoint(struct f2fs_sb_info *sbi) return -ENOMEM; /* * Finding out valid cp block involves read both - * sets( cp pack1 and cp pack 2) + * sets( cp pack 1 and cp pack 2) */ cp_start_blk_no = le32_to_cpu(fsb->cp_blkaddr); cp1 = validate_checkpoint(sbi, cp_start_blk_no, &cp1_version); @@ -1385,10 +1382,7 @@ static int do_checkpoint(struct f2fs_sb_info *sbi, struct cp_control *cpc) /* Flush all the NAT/SIT pages */ f2fs_sync_meta_pages(sbi, META, LONG_MAX, FS_CP_META_IO); - /* - * modify checkpoint - * version number is already updated - */ + /* start to update checkpoint, cp ver is already updated previously */ ckpt->elapsed_time = cpu_to_le64(get_mtime(sbi, true)); ckpt->free_segment_count = cpu_to_le32(free_segments(sbi)); for (i = 0; i < NR_CURSEG_NODE_TYPE; i++) { @@ -1541,9 +1535,6 @@ static int do_checkpoint(struct f2fs_sb_info *sbi, struct cp_control *cpc) return unlikely(f2fs_cp_error(sbi)) ? -EIO : 0; } -/* - * We guarantee that this checkpoint procedure will not fail. - */ int f2fs_write_checkpoint(struct f2fs_sb_info *sbi, struct cp_control *cpc) { struct f2fs_checkpoint *ckpt = F2FS_CKPT(sbi); @@ -1611,7 +1602,6 @@ int f2fs_write_checkpoint(struct f2fs_sb_info *sbi, struct cp_control *cpc) f2fs_flush_sit_entries(sbi, cpc); - /* unlock all the fs_lock[] in do_checkpoint() */ err = do_checkpoint(sbi, cpc); if (err) f2fs_release_discard_addrs(sbi); @@ -1624,7 +1614,7 @@ stop: if (cpc->reason & CP_RECOVERY) f2fs_notice(sbi, "checkpoint: version = %llx", ckpt_ver); - /* do checkpoint periodically */ + /* update CP_TIME to trigger checkpoint periodically */ f2fs_update_time(sbi, CP_TIME); trace_f2fs_write_checkpoint(sbi->sb, cpc->reason, "finish checkpoint"); out: diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c index 5f4948bc40e9..ecf5ea5f7f07 100644 --- a/fs/f2fs/data.c +++ b/fs/f2fs/data.c @@ -364,9 +364,6 @@ static void f2fs_write_end_io(struct bio *bio) bio_put(bio); } -/* - * Return true, if pre_bio's bdev is same as its target device. - */ struct block_device *f2fs_target_device(struct f2fs_sb_info *sbi, block_t blk_addr, struct bio *bio) { @@ -403,6 +400,9 @@ int f2fs_target_device_index(struct f2fs_sb_info *sbi, block_t blkaddr) return 0; } +/* + * Return true, if pre_bio's bdev is same as its target device. + */ static bool __same_bdev(struct f2fs_sb_info *sbi, block_t blk_addr, struct bio *bio) { @@ -410,9 +410,6 @@ static bool __same_bdev(struct f2fs_sb_info *sbi, return bio->bi_disk == b->bd_disk && bio->bi_partno == b->bd_partno; } -/* - * Low-level block read/write IO operations. - */ static struct bio *__bio_alloc(struct f2fs_io_info *fio, int npages) { struct f2fs_sb_info *sbi = fio->sbi; @@ -1388,13 +1385,9 @@ void __do_map_lock(struct f2fs_sb_info *sbi, int flag, bool lock) } /* - * f2fs_map_blocks() now supported readahead/bmap/rw direct_IO with - * f2fs_map_blocks structure. - * If original data blocks are allocated, then give them to blockdev. - * Otherwise, - * a. preallocate requested block addresses - * b. do not use extent cache for better performance - * c. give the block addresses to blockdev + * f2fs_map_blocks() tries to find or build mapping relationship which + * maps continuous logical blocks to physical blocks, and return such + * info via f2fs_map_blocks structure. */ int f2fs_map_blocks(struct inode *inode, struct f2fs_map_blocks *map, int create, int flag) diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index 43900c1e34fc..760a05f98ac7 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -2388,9 +2388,9 @@ static inline block_t datablock_addr(struct inode *inode, raw_node = F2FS_NODE(node_page); - /* from GC path only */ if (is_inode) { if (!inode) + /* from GC path only */ base = offset_in_addr(&raw_node->i); else if (f2fs_has_extra_attr(inode)) base = get_extra_isize(inode); diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c index 5adda28a73a0..e16be6f99a1a 100644 --- a/fs/f2fs/file.c +++ b/fs/f2fs/file.c @@ -112,7 +112,6 @@ static vm_fault_t f2fs_vm_page_mkwrite(struct vm_fault *vmf) } } - /* fill the page */ f2fs_wait_on_page_writeback(page, DATA, false, true); /* wait for GCed page writeback via META_MAPPING */ diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c index db8725d473b5..608d04950f05 100644 --- a/fs/f2fs/gc.c +++ b/fs/f2fs/gc.c @@ -192,7 +192,10 @@ static void select_policy(struct f2fs_sb_info *sbi, int gc_type, p->ofs_unit = sbi->segs_per_sec; } - /* we need to check every dirty segments in the FG_GC case */ + /* + * adjust candidates range, should select all dirty segments for + * foreground GC and urgent GC cases. + */ if (gc_type != FG_GC && (sbi->gc_mode != GC_URGENT) && p->max_search > sbi->max_victim_search) diff --git a/fs/f2fs/inode.c b/fs/f2fs/inode.c index 9b49832296cf..45aff90af5c5 100644 --- a/fs/f2fs/inode.c +++ b/fs/f2fs/inode.c @@ -776,7 +776,7 @@ no_delete: else f2fs_inode_synced(inode); - /* ino == 0, if f2fs_new_inode() was failed t*/ + /* for the case f2fs_new_inode() was failed, .i_ino is zero, skip it */ if (inode->i_ino) invalidate_mapping_pages(NODE_MAPPING(sbi), inode->i_ino, inode->i_ino); diff --git a/fs/f2fs/namei.c b/fs/f2fs/namei.c index 2aa035422c0f..b75c70813f9e 100644 --- a/fs/f2fs/namei.c +++ b/fs/f2fs/namei.c @@ -177,7 +177,7 @@ static inline int is_extension_exist(const unsigned char *s, const char *sub) } /* - * Set multimedia files as cold files for hot/cold data separation + * Set file's temperature for hot/cold data separation */ static inline void set_file_temperature(struct f2fs_sb_info *sbi, struct inode *inode, const unsigned char *name) diff --git a/fs/f2fs/node.c b/fs/f2fs/node.c index 75a3b2c4dc55..f9b39c8675b5 100644 --- a/fs/f2fs/node.c +++ b/fs/f2fs/node.c @@ -510,9 +510,6 @@ int f2fs_try_to_free_nats(struct f2fs_sb_info *sbi, int nr_shrink) return nr - nr_shrink; } -/* - * This function always returns success - */ int f2fs_get_node_info(struct f2fs_sb_info *sbi, nid_t nid, struct node_info *ni) { @@ -716,8 +713,7 @@ got: /* * Caller should call f2fs_put_dnode(dn). * Also, it should grab and release a rwsem by calling f2fs_lock_op() and - * f2fs_unlock_op() only if ro is not set RDONLY_NODE. - * In the case of RDONLY_NODE, we don't need to care about mutex. + * f2fs_unlock_op() only if mode is set with ALLOC_NODE. */ int f2fs_get_dnode_of_data(struct dnode_of_data *dn, pgoff_t index, int mode) { diff --git a/fs/f2fs/shrinker.c b/fs/f2fs/shrinker.c index a467aca29cfe..d66de5999a26 100644 --- a/fs/f2fs/shrinker.c +++ b/fs/f2fs/shrinker.c @@ -58,7 +58,7 @@ unsigned long f2fs_shrink_count(struct shrinker *shrink, /* count extent cache entries */ count += __count_extent_cache(sbi); - /* shrink clean nat cache entries */ + /* count clean nat cache entries */ count += __count_nat_entries(sbi); /* count free nids cache entries */ diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c index 456e6a1d703c..758fc6006831 100644 --- a/fs/f2fs/super.c +++ b/fs/f2fs/super.c @@ -1658,7 +1658,7 @@ static int f2fs_disable_checkpoint(struct f2fs_sb_info *sbi) out_unlock: up_write(&sbi->gc_lock); restore_flag: - sbi->sb->s_flags = s_flags; /* Restore MS_RDONLY status */ + sbi->sb->s_flags = s_flags; /* Restore SB_RDONLY status */ return err; } @@ -3590,7 +3590,7 @@ try_onemore: f2fs_err(sbi, "Cannot turn on quotas: error %d", err); } #endif - /* if there are nt orphan nodes free them */ + /* if there are any orphan inodes, free them */ err = f2fs_recover_orphan_inodes(sbi); if (err) goto free_meta; From 31bc1dc12f4d777349985ea2fac38d275df55e06 Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Mon, 2 Mar 2020 17:34:27 +0800 Subject: [PATCH 14/86] f2fs: fix to avoid using uninitialized variable In f2fs_vm_page_mkwrite(), if inode is compress one, and current mmapped page locates in compressed cluster, we have to call f2fs_get_dnode_of_data() to get its physical block address before f2fs_wait_on_block_writeback(). Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/file.c | 16 ++++++++++++---- 1 file changed, 12 insertions(+), 4 deletions(-) diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c index e16be6f99a1a..353887b29ab1 100644 --- a/fs/f2fs/file.c +++ b/fs/f2fs/file.c @@ -106,10 +106,18 @@ static vm_fault_t f2fs_vm_page_mkwrite(struct vm_fault *vmf) err = f2fs_get_block(&dn, page->index); f2fs_put_dnode(&dn); __do_map_lock(sbi, F2FS_GET_BLOCK_PRE_AIO, false); - if (err) { - unlock_page(page); - goto out_sem; - } + } + +#ifdef CONFIG_F2FS_FS_COMPRESSION + if (!need_alloc) { + set_new_dnode(&dn, inode, NULL, NULL, 0); + err = f2fs_get_dnode_of_data(&dn, page->index, LOOKUP_NODE); + f2fs_put_dnode(&dn); + } +#endif + if (err) { + unlock_page(page); + goto out_sem; } f2fs_wait_on_page_writeback(page, DATA, false, true); From e1be20951a2a0ba3e4c8538cf78b0d4214578e27 Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Fri, 28 Feb 2020 18:08:46 +0800 Subject: [PATCH 15/86] f2fs: fix to avoid use-after-free in f2fs_write_multi_pages() In compress cluster, if physical block number is less than logic page number, race condition will cause use-after-free issue as described below: - f2fs_write_compressed_pages - fio.page = cic->rpages[0]; - f2fs_outplace_write_data - f2fs_compress_write_end_io - kfree(cic->rpages); - kfree(cic); - fio.page = cic->rpages[1]; f2fs_write_multi_pages+0xfd0/0x1a98 f2fs_write_data_pages+0x74c/0xb5c do_writepages+0x64/0x108 __writeback_single_inode+0xdc/0x4b8 writeback_sb_inodes+0x4d0/0xa68 __writeback_inodes_wb+0x88/0x178 wb_writeback+0x1f0/0x424 wb_workfn+0x2f4/0x574 process_one_work+0x210/0x48c worker_thread+0x2e8/0x44c kthread+0x110/0x120 ret_from_fork+0x10/0x18 Fixes: 4c8ff7095bef ("f2fs: support data compression") Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/compress.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/f2fs/compress.c b/fs/f2fs/compress.c index 25bc6154c31c..411d1a5679ac 100644 --- a/fs/f2fs/compress.c +++ b/fs/f2fs/compress.c @@ -848,7 +848,7 @@ static int f2fs_write_compressed_pages(struct compress_ctx *cc, blkaddr = datablock_addr(dn.inode, dn.node_page, dn.ofs_in_node); - fio.page = cic->rpages[i]; + fio.page = cc->rpages[i]; fio.old_blkaddr = blkaddr; /* cluster header */ From 872315f532769118ee10ac5efd6f4d7358f71422 Mon Sep 17 00:00:00 2001 From: Jaegeuk Kim Date: Thu, 5 Mar 2020 15:20:26 -0800 Subject: [PATCH 16/86] f2fs: fix wrong check on F2FS_IOC_FSSETXATTR This fixes the incorrect failure when enabling project quota on casefold-enabled file. Cc: Daniel Rosenberg Cc: kernel-team@android.com Reviewed-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/file.c | 13 ++++++++----- 1 file changed, 8 insertions(+), 5 deletions(-) diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c index 353887b29ab1..530bb3f014dc 100644 --- a/fs/f2fs/file.c +++ b/fs/f2fs/file.c @@ -1800,12 +1800,15 @@ static int f2fs_file_flush(struct file *file, fl_owner_t id) static int f2fs_setflags_common(struct inode *inode, u32 iflags, u32 mask) { struct f2fs_inode_info *fi = F2FS_I(inode); + u32 masked_flags = fi->i_flags & mask; + + f2fs_bug_on(F2FS_I_SB(inode), (iflags & ~mask)); /* Is it quota file? Do not allow user to mess with it */ if (IS_NOQUOTA(inode)) return -EPERM; - if ((iflags ^ fi->i_flags) & F2FS_CASEFOLD_FL) { + if ((iflags ^ masked_flags) & F2FS_CASEFOLD_FL) { if (!f2fs_sb_has_casefold(F2FS_I_SB(inode))) return -EOPNOTSUPP; if (!f2fs_empty_dir(inode)) @@ -1819,9 +1822,9 @@ static int f2fs_setflags_common(struct inode *inode, u32 iflags, u32 mask) return -EINVAL; } - if ((iflags ^ fi->i_flags) & F2FS_COMPR_FL) { + if ((iflags ^ masked_flags) & F2FS_COMPR_FL) { if (S_ISREG(inode->i_mode) && - (fi->i_flags & F2FS_COMPR_FL || i_size_read(inode) || + (masked_flags & F2FS_COMPR_FL || i_size_read(inode) || F2FS_HAS_BLOCKS(inode))) return -EINVAL; if (iflags & F2FS_NOCOMP_FL) @@ -1838,8 +1841,8 @@ static int f2fs_setflags_common(struct inode *inode, u32 iflags, u32 mask) set_compress_context(inode); } } - if ((iflags ^ fi->i_flags) & F2FS_NOCOMP_FL) { - if (fi->i_flags & F2FS_COMPR_FL) + if ((iflags ^ masked_flags) & F2FS_NOCOMP_FL) { + if (masked_flags & F2FS_COMPR_FL) return -EINVAL; } From 5d060716dfec2702ea2e5111b8a43f2af3064512 Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Tue, 10 Mar 2020 20:50:07 +0800 Subject: [PATCH 17/86] f2fs: fix to account compressed inode correctly stat_inc_compr_inode() needs to check FI_COMPRESSED_FILE flag, so in f2fs_disable_compressed_file(), we should call stat_dec_compr_inode() before clearing FI_COMPRESSED_FILE flag. Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/f2fs.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index 760a05f98ac7..23013d3b59ba 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -3824,8 +3824,8 @@ static inline u64 f2fs_disable_compressed_file(struct inode *inode) return fi->i_compr_blocks; fi->i_flags &= ~F2FS_COMPR_FL; - clear_inode_flag(inode, FI_COMPRESSED_FILE); stat_dec_compr_inode(inode); + clear_inode_flag(inode, FI_COMPRESSED_FILE); return 0; } From d37e8ddb57155acd79357ebc4068a7310fc56510 Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Tue, 10 Mar 2020 20:50:08 +0800 Subject: [PATCH 18/86] f2fs: fix to check dirty pages during compressed inode conversion Compressed cluster can be generated during dirty data writeback, if there is dirty pages on compressed inode, it needs to disable converting compressed inode to non-compressed one. Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/f2fs.h | 2 ++ 1 file changed, 2 insertions(+) diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index 23013d3b59ba..dbca4c6a7aaf 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -3820,6 +3820,8 @@ static inline u64 f2fs_disable_compressed_file(struct inode *inode) if (!f2fs_compressed_file(inode)) return 0; + if (get_dirty_pages(inode)) + return 1; if (fi->i_compr_blocks) return fi->i_compr_blocks; From 47e56012f4e8b1d579886de639b5b842c320f13d Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Tue, 10 Mar 2020 20:50:09 +0800 Subject: [PATCH 19/86] f2fs: allow to clear F2FS_COMPR_FL flag If regular inode has no compressed cluster, allow using 'chattr -c' to remove its compress flag, recovering it to a non-compressed file. Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/file.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c index 530bb3f014dc..605a726b8bee 100644 --- a/fs/f2fs/file.c +++ b/fs/f2fs/file.c @@ -1823,10 +1823,10 @@ static int f2fs_setflags_common(struct inode *inode, u32 iflags, u32 mask) } if ((iflags ^ masked_flags) & F2FS_COMPR_FL) { - if (S_ISREG(inode->i_mode) && - (masked_flags & F2FS_COMPR_FL || i_size_read(inode) || - F2FS_HAS_BLOCKS(inode))) - return -EINVAL; + if (S_ISREG(inode->i_mode) && (masked_flags & F2FS_COMPR_FL)) { + if (f2fs_disable_compressed_file(inode)) + return -EINVAL; + } if (iflags & F2FS_NOCOMP_FL) return -EINVAL; if (iflags & F2FS_COMPR_FL) { From 8c753bfc20406f4333f7ca06ca00649ad3e296bb Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Wed, 11 Mar 2020 10:33:53 +0100 Subject: [PATCH 20/86] f2fs: Use scnprintf() for avoiding potential buffer overflow Since snprintf() returns the would-be-output size instead of the actual output size, the succeeding calls may go beyond the given buffer limit. Fix it by replacing with scnprintf(). Signed-off-by: Takashi Iwai Reviewed-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/sysfs.c | 38 +++++++++++++++++++------------------- 1 file changed, 19 insertions(+), 19 deletions(-) diff --git a/fs/f2fs/sysfs.c b/fs/f2fs/sysfs.c index 91d649790b1b..5cc58a1fa645 100644 --- a/fs/f2fs/sysfs.c +++ b/fs/f2fs/sysfs.c @@ -109,47 +109,47 @@ static ssize_t features_show(struct f2fs_attr *a, return sprintf(buf, "0\n"); if (f2fs_sb_has_encrypt(sbi)) - len += snprintf(buf, PAGE_SIZE - len, "%s", + len += scnprintf(buf, PAGE_SIZE - len, "%s", "encryption"); if (f2fs_sb_has_blkzoned(sbi)) - len += snprintf(buf + len, PAGE_SIZE - len, "%s%s", + len += scnprintf(buf + len, PAGE_SIZE - len, "%s%s", len ? ", " : "", "blkzoned"); if (f2fs_sb_has_extra_attr(sbi)) - len += snprintf(buf + len, PAGE_SIZE - len, "%s%s", + len += scnprintf(buf + len, PAGE_SIZE - len, "%s%s", len ? ", " : "", "extra_attr"); if (f2fs_sb_has_project_quota(sbi)) - len += snprintf(buf + len, PAGE_SIZE - len, "%s%s", + len += scnprintf(buf + len, PAGE_SIZE - len, "%s%s", len ? ", " : "", "projquota"); if (f2fs_sb_has_inode_chksum(sbi)) - len += snprintf(buf + len, PAGE_SIZE - len, "%s%s", + len += scnprintf(buf + len, PAGE_SIZE - len, "%s%s", len ? ", " : "", "inode_checksum"); if (f2fs_sb_has_flexible_inline_xattr(sbi)) - len += snprintf(buf + len, PAGE_SIZE - len, "%s%s", + len += scnprintf(buf + len, PAGE_SIZE - len, "%s%s", len ? ", " : "", "flexible_inline_xattr"); if (f2fs_sb_has_quota_ino(sbi)) - len += snprintf(buf + len, PAGE_SIZE - len, "%s%s", + len += scnprintf(buf + len, PAGE_SIZE - len, "%s%s", len ? ", " : "", "quota_ino"); if (f2fs_sb_has_inode_crtime(sbi)) - len += snprintf(buf + len, PAGE_SIZE - len, "%s%s", + len += scnprintf(buf + len, PAGE_SIZE - len, "%s%s", len ? ", " : "", "inode_crtime"); if (f2fs_sb_has_lost_found(sbi)) - len += snprintf(buf + len, PAGE_SIZE - len, "%s%s", + len += scnprintf(buf + len, PAGE_SIZE - len, "%s%s", len ? ", " : "", "lost_found"); if (f2fs_sb_has_verity(sbi)) - len += snprintf(buf + len, PAGE_SIZE - len, "%s%s", + len += scnprintf(buf + len, PAGE_SIZE - len, "%s%s", len ? ", " : "", "verity"); if (f2fs_sb_has_sb_chksum(sbi)) - len += snprintf(buf + len, PAGE_SIZE - len, "%s%s", + len += scnprintf(buf + len, PAGE_SIZE - len, "%s%s", len ? ", " : "", "sb_checksum"); if (f2fs_sb_has_casefold(sbi)) - len += snprintf(buf + len, PAGE_SIZE - len, "%s%s", + len += scnprintf(buf + len, PAGE_SIZE - len, "%s%s", len ? ", " : "", "casefold"); if (f2fs_sb_has_compression(sbi)) - len += snprintf(buf + len, PAGE_SIZE - len, "%s%s", + len += scnprintf(buf + len, PAGE_SIZE - len, "%s%s", len ? ", " : "", "compression"); - len += snprintf(buf + len, PAGE_SIZE - len, "%s%s", + len += scnprintf(buf + len, PAGE_SIZE - len, "%s%s", len ? ", " : "", "pin_file"); - len += snprintf(buf + len, PAGE_SIZE - len, "\n"); + len += scnprintf(buf + len, PAGE_SIZE - len, "\n"); return len; } @@ -233,16 +233,16 @@ static ssize_t f2fs_sbi_show(struct f2fs_attr *a, int hot_count = sbi->raw_super->hot_ext_count; int len = 0, i; - len += snprintf(buf + len, PAGE_SIZE - len, + len += scnprintf(buf + len, PAGE_SIZE - len, "cold file extension:\n"); for (i = 0; i < cold_count; i++) - len += snprintf(buf + len, PAGE_SIZE - len, "%s\n", + len += scnprintf(buf + len, PAGE_SIZE - len, "%s\n", extlist[i]); - len += snprintf(buf + len, PAGE_SIZE - len, + len += scnprintf(buf + len, PAGE_SIZE - len, "hot file extension:\n"); for (i = cold_count; i < cold_count + hot_count; i++) - len += snprintf(buf + len, PAGE_SIZE - len, "%s\n", + len += scnprintf(buf + len, PAGE_SIZE - len, "%s\n", extlist[i]); return len; } From 2b6f655ecffbbf17257476c32af5cec341c69ba1 Mon Sep 17 00:00:00 2001 From: Jaegeuk Kim Date: Tue, 25 Feb 2020 19:08:16 -0800 Subject: [PATCH 21/86] f2fs: show mounted time Let's show mounted time. Reviewed-by: Chao Yu Signed-off-by: Jaegeuk Kim --- Documentation/ABI/testing/sysfs-fs-f2fs | 5 +++++ fs/f2fs/debug.c | 3 +++ fs/f2fs/segment.c | 2 +- fs/f2fs/segment.h | 2 +- fs/f2fs/sysfs.c | 8 ++++++++ 5 files changed, 18 insertions(+), 2 deletions(-) diff --git a/Documentation/ABI/testing/sysfs-fs-f2fs b/Documentation/ABI/testing/sysfs-fs-f2fs index 1a6cd5397129..bd8a0d19abe6 100644 --- a/Documentation/ABI/testing/sysfs-fs-f2fs +++ b/Documentation/ABI/testing/sysfs-fs-f2fs @@ -318,3 +318,8 @@ Date: September 2019 Contact: "Hridya Valsaraju" Description: Average number of valid blocks. Available when CONFIG_F2FS_STAT_FS=y. + +What: /sys/fs/f2fs//mounted_time_sec +Date: February 2020 +Contact: "Jaegeuk Kim" +Description: Show the mounted time in secs of this partition. diff --git a/fs/f2fs/debug.c b/fs/f2fs/debug.c index 6b89eae5e4ca..0dbcb0f9c019 100644 --- a/fs/f2fs/debug.c +++ b/fs/f2fs/debug.c @@ -301,6 +301,9 @@ static int stat_show(struct seq_file *s, void *v) si->ssa_area_segs, si->main_area_segs); seq_printf(s, "(OverProv:%d Resv:%d)]\n\n", si->overp_segs, si->rsvd_segs); + seq_printf(s, "Current Time Sec: %llu / Mounted Time Sec: %llu\n\n", + ktime_get_boottime_seconds(), + SIT_I(si->sbi)->mounted_time); if (test_opt(si->sbi, DISCARD)) seq_printf(s, "Utilization: %u%% (%u valid blocks, %u discard blocks)\n", si->utilization, si->valid_count, si->discard_blks); diff --git a/fs/f2fs/segment.c b/fs/f2fs/segment.c index cf0eb002cfd4..c00cf5c27e95 100644 --- a/fs/f2fs/segment.c +++ b/fs/f2fs/segment.c @@ -4071,7 +4071,7 @@ static int build_sit_info(struct f2fs_sb_info *sbi) sit_i->dirty_sentries = 0; sit_i->sents_per_block = SIT_ENTRY_PER_BLOCK; sit_i->elapsed_time = le64_to_cpu(sbi->ckpt->elapsed_time); - sit_i->mounted_time = ktime_get_real_seconds(); + sit_i->mounted_time = ktime_get_boottime_seconds(); init_rwsem(&sit_i->sentry_lock); return 0; } diff --git a/fs/f2fs/segment.h b/fs/f2fs/segment.h index 459dc3901a57..7a83bd530812 100644 --- a/fs/f2fs/segment.h +++ b/fs/f2fs/segment.h @@ -756,7 +756,7 @@ static inline unsigned long long get_mtime(struct f2fs_sb_info *sbi, bool base_time) { struct sit_info *sit_i = SIT_I(sbi); - time64_t diff, now = ktime_get_real_seconds(); + time64_t diff, now = ktime_get_boottime_seconds(); if (now >= sit_i->mounted_time) return sit_i->elapsed_time + now - sit_i->mounted_time; diff --git a/fs/f2fs/sysfs.c b/fs/f2fs/sysfs.c index 5cc58a1fa645..5f4fec83a6c6 100644 --- a/fs/f2fs/sysfs.c +++ b/fs/f2fs/sysfs.c @@ -185,6 +185,12 @@ static ssize_t encoding_show(struct f2fs_attr *a, return sprintf(buf, "(none)"); } +static ssize_t mounted_time_sec_show(struct f2fs_attr *a, + struct f2fs_sb_info *sbi, char *buf) +{ + return sprintf(buf, "%llu", SIT_I(sbi)->mounted_time); +} + #ifdef CONFIG_F2FS_STAT_FS static ssize_t moved_blocks_foreground_show(struct f2fs_attr *a, struct f2fs_sb_info *sbi, char *buf) @@ -544,6 +550,7 @@ F2FS_GENERAL_RO_ATTR(features); F2FS_GENERAL_RO_ATTR(current_reserved_blocks); F2FS_GENERAL_RO_ATTR(unusable); F2FS_GENERAL_RO_ATTR(encoding); +F2FS_GENERAL_RO_ATTR(mounted_time_sec); #ifdef CONFIG_F2FS_STAT_FS F2FS_STAT_ATTR(STAT_INFO, f2fs_stat_info, cp_foreground_calls, cp_count); F2FS_STAT_ATTR(STAT_INFO, f2fs_stat_info, cp_background_calls, bg_cp_count); @@ -621,6 +628,7 @@ static struct attribute *f2fs_attrs[] = { ATTR_LIST(reserved_blocks), ATTR_LIST(current_reserved_blocks), ATTR_LIST(encoding), + ATTR_LIST(mounted_time_sec), #ifdef CONFIG_F2FS_STAT_FS ATTR_LIST(cp_foreground_calls), ATTR_LIST(cp_background_calls), From e14abdaa9dd38c51c427a58d45bd13c88834831b Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Fri, 14 Feb 2020 17:44:10 +0800 Subject: [PATCH 22/86] f2fs: clean up codes with {f2fs_,}data_blkaddr() - rename datablock_addr() to data_blkaddr(). - wrap data_blkaddr() with f2fs_data_blkaddr() to clean up parameters. Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/compress.c | 7 +++---- fs/f2fs/data.c | 12 +++++------- fs/f2fs/f2fs.h | 7 ++++++- fs/f2fs/file.c | 15 +++++---------- fs/f2fs/gc.c | 2 +- fs/f2fs/node.c | 3 +-- fs/f2fs/recovery.c | 7 +++---- 7 files changed, 24 insertions(+), 29 deletions(-) diff --git a/fs/f2fs/compress.c b/fs/f2fs/compress.c index 411d1a5679ac..5ef7d7cfc0e7 100644 --- a/fs/f2fs/compress.c +++ b/fs/f2fs/compress.c @@ -557,7 +557,7 @@ static int f2fs_compressed_blocks(struct compress_ctx *cc) for (i = 1; i < cc->cluster_size; i++) { block_t blkaddr; - blkaddr = datablock_addr(dn.inode, + blkaddr = data_blkaddr(dn.inode, dn.node_page, dn.ofs_in_node + i); if (blkaddr != NULL_ADDR) ret++; @@ -797,7 +797,7 @@ static int f2fs_write_compressed_pages(struct compress_ctx *cc, goto out_unlock_op; for (i = 0; i < cc->cluster_size; i++) { - if (datablock_addr(dn.inode, dn.node_page, + if (data_blkaddr(dn.inode, dn.node_page, dn.ofs_in_node + i) == NULL_ADDR) goto out_put_dnode; } @@ -846,8 +846,7 @@ static int f2fs_write_compressed_pages(struct compress_ctx *cc, for (i = 0; i < cc->cluster_size; i++, dn.ofs_in_node++) { block_t blkaddr; - blkaddr = datablock_addr(dn.inode, dn.node_page, - dn.ofs_in_node); + blkaddr = f2fs_data_blkaddr(&dn); fio.page = cc->rpages[i]; fio.old_blkaddr = blkaddr; diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c index ecf5ea5f7f07..3b5bf4390014 100644 --- a/fs/f2fs/data.c +++ b/fs/f2fs/data.c @@ -1044,8 +1044,7 @@ int f2fs_reserve_new_blocks(struct dnode_of_data *dn, blkcnt_t count) f2fs_wait_on_page_writeback(dn->node_page, NODE, true, true); for (; count > 0; dn->ofs_in_node++) { - block_t blkaddr = datablock_addr(dn->inode, - dn->node_page, dn->ofs_in_node); + block_t blkaddr = f2fs_data_blkaddr(dn); if (blkaddr == NULL_ADDR) { dn->data_blkaddr = NEW_ADDR; __set_data_blkaddr(dn); @@ -1297,8 +1296,7 @@ static int __allocate_data_block(struct dnode_of_data *dn, int seg_type) if (err) return err; - dn->data_blkaddr = datablock_addr(dn->inode, - dn->node_page, dn->ofs_in_node); + dn->data_blkaddr = f2fs_data_blkaddr(dn); if (dn->data_blkaddr != NULL_ADDR) goto alloc; @@ -1460,7 +1458,7 @@ next_dnode: end_offset = ADDRS_PER_PAGE(dn.node_page, inode); next_block: - blkaddr = datablock_addr(dn.inode, dn.node_page, dn.ofs_in_node); + blkaddr = f2fs_data_blkaddr(&dn); if (__is_valid_data_blkaddr(blkaddr) && !f2fs_is_valid_blkaddr(sbi, blkaddr, DATA_GENERIC_ENHANCE)) { @@ -2060,7 +2058,7 @@ int f2fs_read_multi_pages(struct compress_ctx *cc, struct bio **bio_ret, for (i = 1; i < cc->cluster_size; i++) { block_t blkaddr; - blkaddr = datablock_addr(dn.inode, dn.node_page, + blkaddr = data_blkaddr(dn.inode, dn.node_page, dn.ofs_in_node + i); if (!__is_valid_data_blkaddr(blkaddr)) @@ -2089,7 +2087,7 @@ int f2fs_read_multi_pages(struct compress_ctx *cc, struct bio **bio_ret, struct page *page = dic->cpages[i]; block_t blkaddr; - blkaddr = datablock_addr(dn.inode, dn.node_page, + blkaddr = data_blkaddr(dn.inode, dn.node_page, dn.ofs_in_node + i + 1); if (bio && !page_is_mergeable(sbi, bio, diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index dbca4c6a7aaf..59d6ade50e22 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -2378,7 +2378,7 @@ static inline __le32 *blkaddr_in_node(struct f2fs_node *node) } static inline int f2fs_has_extra_attr(struct inode *inode); -static inline block_t datablock_addr(struct inode *inode, +static inline block_t data_blkaddr(struct inode *inode, struct page *node_page, unsigned int offset) { struct f2fs_node *raw_node; @@ -2400,6 +2400,11 @@ static inline block_t datablock_addr(struct inode *inode, return le32_to_cpu(addr_array[base + offset]); } +static inline block_t f2fs_data_blkaddr(struct dnode_of_data *dn) +{ + return data_blkaddr(dn->inode, dn->node_page, dn->ofs_in_node); +} + static inline int f2fs_test_bit(unsigned int nr, char *addr) { int mask; diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c index 605a726b8bee..5a82e867b072 100644 --- a/fs/f2fs/file.c +++ b/fs/f2fs/file.c @@ -455,8 +455,7 @@ static loff_t f2fs_seek_block(struct file *file, loff_t offset, int whence) data_ofs = (loff_t)pgofs << PAGE_SHIFT) { block_t blkaddr; - blkaddr = datablock_addr(dn.inode, - dn.node_page, dn.ofs_in_node); + blkaddr = f2fs_data_blkaddr(&dn); if (__is_valid_data_blkaddr(blkaddr) && !f2fs_is_valid_blkaddr(F2FS_I_SB(inode), @@ -1122,8 +1121,7 @@ next_dnode: done = min((pgoff_t)ADDRS_PER_PAGE(dn.node_page, inode) - dn.ofs_in_node, len); for (i = 0; i < done; i++, blkaddr++, do_replace++, dn.ofs_in_node++) { - *blkaddr = datablock_addr(dn.inode, - dn.node_page, dn.ofs_in_node); + *blkaddr = f2fs_data_blkaddr(&dn); if (__is_valid_data_blkaddr(*blkaddr) && !f2fs_is_valid_blkaddr(sbi, *blkaddr, @@ -1212,8 +1210,7 @@ static int __clone_blkaddrs(struct inode *src_inode, struct inode *dst_inode, ADDRS_PER_PAGE(dn.node_page, dst_inode) - dn.ofs_in_node, len - i); do { - dn.data_blkaddr = datablock_addr(dn.inode, - dn.node_page, dn.ofs_in_node); + dn.data_blkaddr = f2fs_data_blkaddr(&dn); f2fs_truncate_data_blocks_range(&dn, 1); if (do_replace[i]) { @@ -1389,8 +1386,7 @@ static int f2fs_do_zero_range(struct dnode_of_data *dn, pgoff_t start, int ret; for (; index < end; index++, dn->ofs_in_node++) { - if (datablock_addr(dn->inode, dn->node_page, - dn->ofs_in_node) == NULL_ADDR) + if (f2fs_data_blkaddr(dn) == NULL_ADDR) count++; } @@ -1401,8 +1397,7 @@ static int f2fs_do_zero_range(struct dnode_of_data *dn, pgoff_t start, dn->ofs_in_node = ofs_in_node; for (index = start; index < end; index++, dn->ofs_in_node++) { - dn->data_blkaddr = datablock_addr(dn->inode, - dn->node_page, dn->ofs_in_node); + dn->data_blkaddr = f2fs_data_blkaddr(dn); /* * f2fs_reserve_new_blocks will not guarantee entire block * allocation. diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c index 608d04950f05..3fdc3d6a874c 100644 --- a/fs/f2fs/gc.c +++ b/fs/f2fs/gc.c @@ -637,7 +637,7 @@ static bool is_alive(struct f2fs_sb_info *sbi, struct f2fs_summary *sum, } *nofs = ofs_of_node(node_page); - source_blkaddr = datablock_addr(NULL, node_page, ofs_in_node); + source_blkaddr = data_blkaddr(NULL, node_page, ofs_in_node); f2fs_put_page(node_page, 1); if (source_blkaddr != blkaddr) { diff --git a/fs/f2fs/node.c b/fs/f2fs/node.c index f9b39c8675b5..bdcab8aeb542 100644 --- a/fs/f2fs/node.c +++ b/fs/f2fs/node.c @@ -805,8 +805,7 @@ int f2fs_get_dnode_of_data(struct dnode_of_data *dn, pgoff_t index, int mode) dn->nid = nids[level]; dn->ofs_in_node = offset[level]; dn->node_page = npage[level]; - dn->data_blkaddr = datablock_addr(dn->inode, - dn->node_page, dn->ofs_in_node); + dn->data_blkaddr = f2fs_data_blkaddr(dn); return 0; release_pages: diff --git a/fs/f2fs/recovery.c b/fs/f2fs/recovery.c index 763d5c0951d1..348e8d463b3e 100644 --- a/fs/f2fs/recovery.c +++ b/fs/f2fs/recovery.c @@ -496,8 +496,7 @@ out: return 0; truncate_out: - if (datablock_addr(tdn.inode, tdn.node_page, - tdn.ofs_in_node) == blkaddr) + if (f2fs_data_blkaddr(&tdn) == blkaddr) f2fs_truncate_data_blocks_range(&tdn, 1); if (dn->inode->i_ino == nid && !dn->inode_page_locked) unlock_page(dn->inode_page); @@ -560,8 +559,8 @@ retry_dn: for (; start < end; start++, dn.ofs_in_node++) { block_t src, dest; - src = datablock_addr(dn.inode, dn.node_page, dn.ofs_in_node); - dest = datablock_addr(dn.inode, page, dn.ofs_in_node); + src = f2fs_data_blkaddr(&dn); + dest = data_blkaddr(dn.inode, page, dn.ofs_in_node); if (__is_valid_data_blkaddr(src) && !f2fs_is_valid_blkaddr(sbi, src, META_POR)) { From 76fb65bef352eeb188c04fac4994e4bac5efd27d Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Fri, 14 Feb 2020 17:44:11 +0800 Subject: [PATCH 23/86] f2fs: clean up parameter of macro XATTR_SIZE() Just cleanup, no logic change. Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/xattr.c | 10 ++++------ fs/f2fs/xattr.h | 3 ++- 2 files changed, 6 insertions(+), 7 deletions(-) diff --git a/fs/f2fs/xattr.c b/fs/f2fs/xattr.c index 6794a13064be..d4800df6302d 100644 --- a/fs/f2fs/xattr.c +++ b/fs/f2fs/xattr.c @@ -312,12 +312,12 @@ static int lookup_all_xattrs(struct inode *inode, struct page *ipage, if (!xnid && !inline_size) return -ENODATA; - *base_size = XATTR_SIZE(xnid, inode) + XATTR_PADDING_SIZE; + *base_size = XATTR_SIZE(inode) + XATTR_PADDING_SIZE; txattr_addr = f2fs_kzalloc(F2FS_I_SB(inode), *base_size, GFP_NOFS); if (!txattr_addr) return -ENOMEM; - last_txattr_addr = (void *)txattr_addr + XATTR_SIZE(xnid, inode); + last_txattr_addr = (void *)txattr_addr + XATTR_SIZE(inode); /* read from inline xattr */ if (inline_size) { @@ -539,7 +539,6 @@ out: ssize_t f2fs_listxattr(struct dentry *dentry, char *buffer, size_t buffer_size) { struct inode *inode = d_inode(dentry); - nid_t xnid = F2FS_I(inode)->i_xattr_nid; struct f2fs_xattr_entry *entry; void *base_addr, *last_base_addr; int error = 0; @@ -551,7 +550,7 @@ ssize_t f2fs_listxattr(struct dentry *dentry, char *buffer, size_t buffer_size) if (error) return error; - last_base_addr = (void *)base_addr + XATTR_SIZE(xnid, inode); + last_base_addr = (void *)base_addr + XATTR_SIZE(inode); list_for_each_xattr(entry, base_addr) { const struct xattr_handler *handler = @@ -609,7 +608,6 @@ static int __f2fs_setxattr(struct inode *inode, int index, { struct f2fs_xattr_entry *here, *last; void *base_addr, *last_base_addr; - nid_t xnid = F2FS_I(inode)->i_xattr_nid; int found, newsize; size_t len; __u32 new_hsize; @@ -633,7 +631,7 @@ static int __f2fs_setxattr(struct inode *inode, int index, if (error) return error; - last_base_addr = (void *)base_addr + XATTR_SIZE(xnid, inode); + last_base_addr = (void *)base_addr + XATTR_SIZE(inode); /* find entry with wanted name. */ here = __find_xattr(base_addr, last_base_addr, index, len, name); diff --git a/fs/f2fs/xattr.h b/fs/f2fs/xattr.h index de0c600b9cab..574beea46494 100644 --- a/fs/f2fs/xattr.h +++ b/fs/f2fs/xattr.h @@ -73,7 +73,8 @@ struct f2fs_xattr_entry { entry = XATTR_NEXT_ENTRY(entry)) #define VALID_XATTR_BLOCK_SIZE (PAGE_SIZE - sizeof(struct node_footer)) #define XATTR_PADDING_SIZE (sizeof(__u32)) -#define XATTR_SIZE(x,i) (((x) ? VALID_XATTR_BLOCK_SIZE : 0) + \ +#define XATTR_SIZE(i) ((F2FS_I(i)->i_xattr_nid ? \ + VALID_XATTR_BLOCK_SIZE : 0) + \ (inline_xattr_size(i))) #define MIN_OFFSET(i) XATTR_ALIGN(inline_xattr_size(i) + \ VALID_XATTR_BLOCK_SIZE) From 545adf83e38a690f1ee43c102ba82a078fef0f75 Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Fri, 14 Feb 2020 17:45:11 +0800 Subject: [PATCH 24/86] f2fs: fix to show norecovery mount option Previously, 'norecovery' mount option will be shown as 'disable_roll_forward', fix to show original option name correctly. Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/f2fs.h | 1 + fs/f2fs/super.c | 7 +++++-- 2 files changed, 6 insertions(+), 2 deletions(-) diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index 59d6ade50e22..221d035729f6 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -100,6 +100,7 @@ extern const char *f2fs_fault_name[FAULT_MAX]; #define F2FS_MOUNT_INLINE_XATTR_SIZE 0x00800000 #define F2FS_MOUNT_RESERVE_ROOT 0x01000000 #define F2FS_MOUNT_DISABLE_CHECKPOINT 0x02000000 +#define F2FS_MOUNT_NORECOVERY 0x04000000 #define F2FS_OPTION(sbi) ((sbi)->mount_opt) #define clear_opt(sbi, option) (F2FS_OPTION(sbi).opt &= ~F2FS_MOUNT_##option) diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c index 758fc6006831..035118662cfb 100644 --- a/fs/f2fs/super.c +++ b/fs/f2fs/super.c @@ -446,7 +446,7 @@ static int parse_options(struct super_block *sb, char *options) break; case Opt_norecovery: /* this option mounts f2fs with ro */ - set_opt(sbi, DISABLE_ROLL_FORWARD); + set_opt(sbi, NORECOVERY); if (!f2fs_readonly(sb)) return -EINVAL; break; @@ -1447,6 +1447,8 @@ static int f2fs_show_options(struct seq_file *seq, struct dentry *root) } if (test_opt(sbi, DISABLE_ROLL_FORWARD)) seq_puts(seq, ",disable_roll_forward"); + if (test_opt(sbi, NORECOVERY)) + seq_puts(seq, ",norecovery"); if (test_opt(sbi, DISCARD)) seq_puts(seq, ",discard"); else @@ -3599,7 +3601,8 @@ try_onemore: goto reset_checkpoint; /* recover fsynced data */ - if (!test_opt(sbi, DISABLE_ROLL_FORWARD)) { + if (!test_opt(sbi, DISABLE_ROLL_FORWARD) && + !test_opt(sbi, NORECOVERY)) { /* * mount should be failed, when device has readonly mode, and * previous checkpoint was not done by clean system shutdown. From d0fef6f0716275ad3accb92c4938269690deac0d Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Fri, 14 Feb 2020 17:44:12 +0800 Subject: [PATCH 25/86] f2fs: clean up lfs/adaptive mount option This patch removes F2FS_MOUNT_ADAPTIVE and F2FS_MOUNT_LFS mount options, and add F2FS_OPTION.fs_mode with below two status to indicate filesystem mode. enum { FS_MODE_ADAPTIVE, /* use both lfs/ssr allocation */ FS_MODE_LFS, /* use lfs allocation only */ }; It can enhance code readability and fs mode's scalability. Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/data.c | 8 ++++---- fs/f2fs/f2fs.h | 27 ++++++++++----------------- fs/f2fs/file.c | 2 +- fs/f2fs/gc.c | 2 +- fs/f2fs/segment.c | 12 ++++++------ fs/f2fs/super.c | 16 ++++++++-------- 6 files changed, 30 insertions(+), 37 deletions(-) diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c index 3b5bf4390014..04cf514ca36d 100644 --- a/fs/f2fs/data.c +++ b/fs/f2fs/data.c @@ -442,7 +442,7 @@ static inline void __submit_bio(struct f2fs_sb_info *sbi, if (type != DATA && type != NODE) goto submit_io; - if (test_opt(sbi, LFS) && current->plug) + if (f2fs_lfs_mode(sbi) && current->plug) blk_finish_plug(current->plug); if (F2FS_IO_ALIGNED(sbi)) @@ -1413,7 +1413,7 @@ int f2fs_map_blocks(struct inode *inode, struct f2fs_map_blocks *map, end = pgofs + maxblocks; if (!create && f2fs_lookup_extent_cache(inode, pgofs, &ei)) { - if (test_opt(sbi, LFS) && flag == F2FS_GET_BLOCK_DIO && + if (f2fs_lfs_mode(sbi) && flag == F2FS_GET_BLOCK_DIO && map->m_may_create) goto next_dnode; @@ -1468,7 +1468,7 @@ next_block: if (__is_valid_data_blkaddr(blkaddr)) { /* use out-place-update for driect IO under LFS mode */ - if (test_opt(sbi, LFS) && flag == F2FS_GET_BLOCK_DIO && + if (f2fs_lfs_mode(sbi) && flag == F2FS_GET_BLOCK_DIO && map->m_may_create) { err = __allocate_data_block(&dn, map->m_seg_type); if (err) @@ -2388,7 +2388,7 @@ bool f2fs_should_update_outplace(struct inode *inode, struct f2fs_io_info *fio) { struct f2fs_sb_info *sbi = F2FS_I_SB(inode); - if (test_opt(sbi, LFS)) + if (f2fs_lfs_mode(sbi)) return true; if (S_ISDIR(inode->i_mode)) return true; diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index 221d035729f6..e27d30d6a387 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -91,8 +91,6 @@ extern const char *f2fs_fault_name[FAULT_MAX]; #define F2FS_MOUNT_FORCE_FG_GC 0x00004000 #define F2FS_MOUNT_DATA_FLUSH 0x00008000 #define F2FS_MOUNT_FAULT_INJECTION 0x00010000 -#define F2FS_MOUNT_ADAPTIVE 0x00020000 -#define F2FS_MOUNT_LFS 0x00040000 #define F2FS_MOUNT_USRQUOTA 0x00080000 #define F2FS_MOUNT_GRPQUOTA 0x00100000 #define F2FS_MOUNT_PRJQUOTA 0x00200000 @@ -139,6 +137,7 @@ struct f2fs_mount_info { int whint_mode; int alloc_mode; /* segment allocation policy */ int fsync_mode; /* fsync policy */ + int fs_mode; /* fs mode: LFS or ADAPTIVE */ bool test_dummy_encryption; /* test dummy encryption */ block_t unusable_cap; /* Amount of space allowed to be * unusable when disabling checkpoint @@ -1173,6 +1172,11 @@ enum { GC_URGENT, }; +enum { + FS_MODE_ADAPTIVE, /* use both lfs/ssr allocation */ + FS_MODE_LFS, /* use lfs allocation only */ +}; + enum { WHINT_MODE_OFF, /* not pass down write hints */ WHINT_MODE_USER, /* try to pass down hints given by users */ @@ -3910,20 +3914,9 @@ static inline bool f2fs_hw_is_readonly(struct f2fs_sb_info *sbi) return false; } - -static inline void set_opt_mode(struct f2fs_sb_info *sbi, unsigned int mt) +static inline bool f2fs_lfs_mode(struct f2fs_sb_info *sbi) { - clear_opt(sbi, ADAPTIVE); - clear_opt(sbi, LFS); - - switch (mt) { - case F2FS_MOUNT_ADAPTIVE: - set_opt(sbi, ADAPTIVE); - break; - case F2FS_MOUNT_LFS: - set_opt(sbi, LFS); - break; - } + return F2FS_OPTION(sbi).fs_mode == FS_MODE_LFS; } static inline bool f2fs_may_encrypt(struct inode *inode) @@ -3978,7 +3971,7 @@ static inline int allow_outplace_dio(struct inode *inode, struct f2fs_sb_info *sbi = F2FS_I_SB(inode); int rw = iov_iter_rw(iter); - return (test_opt(sbi, LFS) && (rw == WRITE) && + return (f2fs_lfs_mode(sbi) && (rw == WRITE) && !block_unaligned_IO(inode, iocb, iter)); } @@ -4000,7 +3993,7 @@ static inline bool f2fs_force_buffered_io(struct inode *inode, */ if (f2fs_sb_has_blkzoned(sbi)) return true; - if (test_opt(sbi, LFS) && (rw == WRITE)) { + if (f2fs_lfs_mode(sbi) && (rw == WRITE)) { if (block_unaligned_IO(inode, iocb, iter)) return true; if (F2FS_IO_ALIGNED(sbi)) diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c index 5a82e867b072..2e8276d54378 100644 --- a/fs/f2fs/file.c +++ b/fs/f2fs/file.c @@ -1132,7 +1132,7 @@ next_dnode: if (!f2fs_is_checkpointed_data(sbi, *blkaddr)) { - if (test_opt(sbi, LFS)) { + if (f2fs_lfs_mode(sbi)) { f2fs_put_dnode(&dn); return -EOPNOTSUPP; } diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c index 3fdc3d6a874c..d883943f7f46 100644 --- a/fs/f2fs/gc.c +++ b/fs/f2fs/gc.c @@ -765,7 +765,7 @@ static int move_data_block(struct inode *inode, block_t bidx, struct page *page, *mpage; block_t newaddr; int err = 0; - bool lfs_mode = test_opt(fio.sbi, LFS); + bool lfs_mode = f2fs_lfs_mode(fio.sbi); /* do not read out */ page = f2fs_grab_cache_page(inode->i_mapping, bidx, false); diff --git a/fs/f2fs/segment.c b/fs/f2fs/segment.c index c00cf5c27e95..04156df83e7a 100644 --- a/fs/f2fs/segment.c +++ b/fs/f2fs/segment.c @@ -172,7 +172,7 @@ bool f2fs_need_SSR(struct f2fs_sb_info *sbi) int dent_secs = get_blocktype_secs(sbi, F2FS_DIRTY_DENTS); int imeta_secs = get_blocktype_secs(sbi, F2FS_DIRTY_IMETA); - if (test_opt(sbi, LFS)) + if (f2fs_lfs_mode(sbi)) return false; if (sbi->gc_mode == GC_URGENT) return true; @@ -1940,7 +1940,7 @@ void f2fs_clear_prefree_segments(struct f2fs_sb_info *sbi, unsigned int start = 0, end = -1; unsigned int secno, start_segno; bool force = (cpc->reason & CP_DISCARD); - bool need_align = test_opt(sbi, LFS) && __is_large_section(sbi); + bool need_align = f2fs_lfs_mode(sbi) && __is_large_section(sbi); mutex_lock(&dirty_i->seglist_lock); @@ -1972,7 +1972,7 @@ void f2fs_clear_prefree_segments(struct f2fs_sb_info *sbi, (end - 1) <= cpc->trim_end) continue; - if (!test_opt(sbi, LFS) || !__is_large_section(sbi)) { + if (!f2fs_lfs_mode(sbi) || !__is_large_section(sbi)) { f2fs_issue_discard(sbi, START_BLOCK(sbi, start), (end - start) << sbi->log_blocks_per_seg); continue; @@ -2830,7 +2830,7 @@ int f2fs_trim_fs(struct f2fs_sb_info *sbi, struct fstrim_range *range) struct discard_policy dpolicy; unsigned long long trimmed = 0; int err = 0; - bool need_align = test_opt(sbi, LFS) && __is_large_section(sbi); + bool need_align = f2fs_lfs_mode(sbi) && __is_large_section(sbi); if (start >= MAX_BLKADDR(sbi) || range->len < sbi->blocksize) return -EINVAL; @@ -3193,7 +3193,7 @@ static void update_device_state(struct f2fs_io_info *fio) static void do_write_page(struct f2fs_summary *sum, struct f2fs_io_info *fio) { int type = __get_segment_type(fio); - bool keep_order = (test_opt(fio->sbi, LFS) && type == CURSEG_COLD_DATA); + bool keep_order = (f2fs_lfs_mode(fio->sbi) && type == CURSEG_COLD_DATA); if (keep_order) down_read(&fio->sbi->io_order_lock); @@ -4678,7 +4678,7 @@ int f2fs_build_segment_manager(struct f2fs_sb_info *sbi) if (sm_info->rec_prefree_segments > DEF_MAX_RECLAIM_PREFREE_SEGMENTS) sm_info->rec_prefree_segments = DEF_MAX_RECLAIM_PREFREE_SEGMENTS; - if (!test_opt(sbi, LFS)) + if (!f2fs_lfs_mode(sbi)) sm_info->ipu_policy = 1 << F2FS_IPU_FSYNC; sm_info->min_ipu_util = DEF_MIN_IPU_UTIL; sm_info->min_fsync_blocks = DEF_MIN_FSYNC_BLOCKS; diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c index 035118662cfb..8006644ee0e3 100644 --- a/fs/f2fs/super.c +++ b/fs/f2fs/super.c @@ -600,10 +600,10 @@ static int parse_options(struct super_block *sb, char *options) kvfree(name); return -EINVAL; } - set_opt_mode(sbi, F2FS_MOUNT_ADAPTIVE); + F2FS_OPTION(sbi).fs_mode = FS_MODE_ADAPTIVE; } else if (strlen(name) == 3 && !strncmp(name, "lfs", 3)) { - set_opt_mode(sbi, F2FS_MOUNT_LFS); + F2FS_OPTION(sbi).fs_mode = FS_MODE_LFS; } else { kvfree(name); return -EINVAL; @@ -904,7 +904,7 @@ static int parse_options(struct super_block *sb, char *options) } #endif - if (F2FS_IO_SIZE_BITS(sbi) && !test_opt(sbi, LFS)) { + if (F2FS_IO_SIZE_BITS(sbi) && !f2fs_lfs_mode(sbi)) { f2fs_err(sbi, "Should set mode=lfs with %uKB-sized IO", F2FS_IO_SIZE_KB(sbi)); return -EINVAL; @@ -934,7 +934,7 @@ static int parse_options(struct super_block *sb, char *options) } } - if (test_opt(sbi, DISABLE_CHECKPOINT) && test_opt(sbi, LFS)) { + if (test_opt(sbi, DISABLE_CHECKPOINT) && f2fs_lfs_mode(sbi)) { f2fs_err(sbi, "LFS not compatible with checkpoint=disable\n"); return -EINVAL; } @@ -1500,9 +1500,9 @@ static int f2fs_show_options(struct seq_file *seq, struct dentry *root) seq_puts(seq, ",data_flush"); seq_puts(seq, ",mode="); - if (test_opt(sbi, ADAPTIVE)) + if (F2FS_OPTION(sbi).fs_mode == FS_MODE_ADAPTIVE) seq_puts(seq, "adaptive"); - else if (test_opt(sbi, LFS)) + else if (F2FS_OPTION(sbi).fs_mode == FS_MODE_LFS) seq_puts(seq, "lfs"); seq_printf(seq, ",active_logs=%u", F2FS_OPTION(sbi).active_logs); if (test_opt(sbi, RESERVE_ROOT)) @@ -1589,9 +1589,9 @@ static void default_options(struct f2fs_sb_info *sbi) set_opt(sbi, FLUSH_MERGE); set_opt(sbi, DISCARD); if (f2fs_sb_has_blkzoned(sbi)) - set_opt_mode(sbi, F2FS_MOUNT_LFS); + F2FS_OPTION(sbi).fs_mode = FS_MODE_LFS; else - set_opt_mode(sbi, F2FS_MOUNT_ADAPTIVE); + F2FS_OPTION(sbi).fs_mode = FS_MODE_ADAPTIVE; #ifdef CONFIG_F2FS_FS_XATTR set_opt(sbi, XATTR_USER); From 846c645aa5969e2514457b347946da0f0570987d Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Fri, 14 Feb 2020 17:44:13 +0800 Subject: [PATCH 26/86] f2fs: clean up bggc mount option There are three status for background gc: on, off and sync, it's a little bit confused to use test_opt(BG_GC) and test_opt(FORCE_FG_GC) combinations to indicate status of background gc. So let's remove F2FS_MOUNT_BG_GC and F2FS_MOUNT_FORCE_FG_GC mount options, and add F2FS_OPTION().bggc_mode with below three status to clean up codes and enhance bggc mode's scalability. enum { BGGC_MODE_ON, /* background gc is on */ BGGC_MODE_OFF, /* background gc is off */ BGGC_MODE_SYNC, /* * background gc is on, migrating blocks * like foreground gc */ }; Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/f2fs.h | 12 ++++++++++-- fs/f2fs/gc.c | 6 +++++- fs/f2fs/super.c | 29 +++++++++++++---------------- 3 files changed, 28 insertions(+), 19 deletions(-) diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index e27d30d6a387..ef3ec6548629 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -74,7 +74,6 @@ extern const char *f2fs_fault_name[FAULT_MAX]; /* * For mount options */ -#define F2FS_MOUNT_BG_GC 0x00000001 #define F2FS_MOUNT_DISABLE_ROLL_FORWARD 0x00000002 #define F2FS_MOUNT_DISCARD 0x00000004 #define F2FS_MOUNT_NOHEAP 0x00000008 @@ -88,7 +87,6 @@ extern const char *f2fs_fault_name[FAULT_MAX]; #define F2FS_MOUNT_NOBARRIER 0x00000800 #define F2FS_MOUNT_FASTBOOT 0x00001000 #define F2FS_MOUNT_EXTENT_CACHE 0x00002000 -#define F2FS_MOUNT_FORCE_FG_GC 0x00004000 #define F2FS_MOUNT_DATA_FLUSH 0x00008000 #define F2FS_MOUNT_FAULT_INJECTION 0x00010000 #define F2FS_MOUNT_USRQUOTA 0x00080000 @@ -138,6 +136,7 @@ struct f2fs_mount_info { int alloc_mode; /* segment allocation policy */ int fsync_mode; /* fsync policy */ int fs_mode; /* fs mode: LFS or ADAPTIVE */ + int bggc_mode; /* bggc mode: off, on or sync */ bool test_dummy_encryption; /* test dummy encryption */ block_t unusable_cap; /* Amount of space allowed to be * unusable when disabling checkpoint @@ -1172,6 +1171,15 @@ enum { GC_URGENT, }; +enum { + BGGC_MODE_ON, /* background gc is on */ + BGGC_MODE_OFF, /* background gc is off */ + BGGC_MODE_SYNC, /* + * background gc is on, migrating blocks + * like foreground gc + */ +}; + enum { FS_MODE_ADAPTIVE, /* use both lfs/ssr allocation */ FS_MODE_LFS, /* use lfs allocation only */ diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c index d883943f7f46..a19966b0e018 100644 --- a/fs/f2fs/gc.c +++ b/fs/f2fs/gc.c @@ -31,6 +31,8 @@ static int gc_thread_func(void *data) set_freezable(); do { + bool sync_mode; + wait_event_interruptible_timeout(*wq, kthread_should_stop() || freezing(current) || gc_th->gc_wake, @@ -101,8 +103,10 @@ static int gc_thread_func(void *data) do_gc: stat_inc_bggc_count(sbi->stat_info); + sync_mode = F2FS_OPTION(sbi).bggc_mode == BGGC_MODE_SYNC; + /* if return value is not zero, no victim was selected */ - if (f2fs_gc(sbi, test_opt(sbi, FORCE_FG_GC), true, NULL_SEGNO)) + if (f2fs_gc(sbi, sync_mode, true, NULL_SEGNO)) wait_ms = gc_th->no_gc_sleep_time; trace_f2fs_background_gc(sbi->sb, wait_ms, diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c index 8006644ee0e3..40d0a6190be1 100644 --- a/fs/f2fs/super.c +++ b/fs/f2fs/super.c @@ -427,14 +427,11 @@ static int parse_options(struct super_block *sb, char *options) if (!name) return -ENOMEM; if (strlen(name) == 2 && !strncmp(name, "on", 2)) { - set_opt(sbi, BG_GC); - clear_opt(sbi, FORCE_FG_GC); + F2FS_OPTION(sbi).bggc_mode = BGGC_MODE_ON; } else if (strlen(name) == 3 && !strncmp(name, "off", 3)) { - clear_opt(sbi, BG_GC); - clear_opt(sbi, FORCE_FG_GC); + F2FS_OPTION(sbi).bggc_mode = BGGC_MODE_OFF; } else if (strlen(name) == 4 && !strncmp(name, "sync", 4)) { - set_opt(sbi, BG_GC); - set_opt(sbi, FORCE_FG_GC); + F2FS_OPTION(sbi).bggc_mode = BGGC_MODE_SYNC; } else { kvfree(name); return -EINVAL; @@ -1437,14 +1434,13 @@ static int f2fs_show_options(struct seq_file *seq, struct dentry *root) { struct f2fs_sb_info *sbi = F2FS_SB(root->d_sb); - if (!f2fs_readonly(sbi->sb) && test_opt(sbi, BG_GC)) { - if (test_opt(sbi, FORCE_FG_GC)) - seq_printf(seq, ",background_gc=%s", "sync"); - else - seq_printf(seq, ",background_gc=%s", "on"); - } else { + if (F2FS_OPTION(sbi).bggc_mode == BGGC_MODE_SYNC) + seq_printf(seq, ",background_gc=%s", "sync"); + else if (F2FS_OPTION(sbi).bggc_mode == BGGC_MODE_ON) + seq_printf(seq, ",background_gc=%s", "on"); + else if (F2FS_OPTION(sbi).bggc_mode == BGGC_MODE_OFF) seq_printf(seq, ",background_gc=%s", "off"); - } + if (test_opt(sbi, DISABLE_ROLL_FORWARD)) seq_puts(seq, ",disable_roll_forward"); if (test_opt(sbi, NORECOVERY)) @@ -1576,8 +1572,8 @@ static void default_options(struct f2fs_sb_info *sbi) F2FS_OPTION(sbi).compress_algorithm = COMPRESS_LZO; F2FS_OPTION(sbi).compress_log_size = MIN_COMPRESS_LOG_SIZE; F2FS_OPTION(sbi).compress_ext_cnt = 0; + F2FS_OPTION(sbi).bggc_mode = BGGC_MODE_ON; - set_opt(sbi, BG_GC); set_opt(sbi, INLINE_XATTR); set_opt(sbi, INLINE_DATA); set_opt(sbi, INLINE_DENTRY); @@ -1783,7 +1779,8 @@ static int f2fs_remount(struct super_block *sb, int *flags, char *data) * or if background_gc = off is passed in mount * option. Also sync the filesystem. */ - if ((*flags & SB_RDONLY) || !test_opt(sbi, BG_GC)) { + if ((*flags & SB_RDONLY) || + F2FS_OPTION(sbi).bggc_mode == BGGC_MODE_OFF) { if (sbi->gc_thread) { f2fs_stop_gc_thread(sbi); need_restart_gc = true; @@ -3668,7 +3665,7 @@ reset_checkpoint: * If filesystem is not mounted as read-only then * do start the gc_thread. */ - if (test_opt(sbi, BG_GC) && !f2fs_readonly(sb)) { + if (F2FS_OPTION(sbi).bggc_mode != BGGC_MODE_OFF && !f2fs_readonly(sb)) { /* After POR, we can run background GC thread.*/ err = f2fs_start_gc_thread(sbi); if (err) From 391d9ab1f1c4e6a051c1aa7c508279afb61ec796 Mon Sep 17 00:00:00 2001 From: Jaegeuk Kim Date: Sun, 9 Feb 2020 13:27:09 -0800 Subject: [PATCH 27/86] f2fs: add migration count iff migration happens If first segment is empty and migration_granularity is 1, we can't move this at all. Reviewed-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/gc.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c index a19966b0e018..bb0b6435dcf7 100644 --- a/fs/f2fs/gc.c +++ b/fs/f2fs/gc.c @@ -1240,12 +1240,12 @@ static int do_garbage_collect(struct f2fs_sb_info *sbi, segno, gc_type); stat_inc_seg_count(sbi, type, gc_type); + migrated++; freed: if (gc_type == FG_GC && get_valid_blocks(sbi, segno, false) == 0) seg_freed++; - migrated++; if (__is_large_section(sbi) && segno + 1 < end_segno) sbi->next_victim_seg[gc_type] = segno + 1; From a8398719f7bc4cec7680308e4960d93d357e6034 Mon Sep 17 00:00:00 2001 From: Jaegeuk Kim Date: Sun, 9 Feb 2020 13:23:28 -0800 Subject: [PATCH 28/86] f2fs: skip GC when section is full This fixes skipping GC when segment is full in large section. Signed-off-by: Jaegeuk Kim --- fs/f2fs/gc.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c index bb0b6435dcf7..0dfdec4926e4 100644 --- a/fs/f2fs/gc.c +++ b/fs/f2fs/gc.c @@ -1025,8 +1025,8 @@ next_step: * race condition along with SSR block allocation. */ if ((gc_type == BG_GC && has_not_enough_free_secs(sbi, 0, 0)) || - get_valid_blocks(sbi, segno, false) == - sbi->blocks_per_seg) + get_valid_blocks(sbi, segno, true) == + BLKS_PER_SEC(sbi)) return submitted; if (check_valid_map(sbi, segno, off) == 0) From 81c36db3205d299cf7772d0b07d9e7123720f7c2 Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Mon, 17 Feb 2020 17:45:44 +0800 Subject: [PATCH 29/86] f2fs: introduce DEFAULT_IO_TIMEOUT As Geert Uytterhoeven reported: for parameter HZ/50 in congestion_wait(BLK_RW_ASYNC, HZ/50); On some platforms, HZ can be less than 50, then unexpected 0 timeout jiffies will be set in congestion_wait(). This patch introduces a macro DEFAULT_IO_TIMEOUT to wrap a determinate value with msecs_to_jiffies(20) to instead HZ/50 to avoid such issue. Quoted from Geert Uytterhoeven: "A timeout of HZ means 1 second. HZ/50 means 20 ms, but has the risk of being zero, if HZ < 50. If you want to use a timeout of 20 ms, you best use msecs_to_jiffies(20), as that takes care of the special cases, and never returns 0." Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/checkpoint.c | 2 +- fs/f2fs/compress.c | 3 ++- fs/f2fs/data.c | 4 ++-- fs/f2fs/f2fs.h | 3 +++ fs/f2fs/gc.c | 3 ++- fs/f2fs/inode.c | 2 +- fs/f2fs/node.c | 2 +- fs/f2fs/recovery.c | 5 +++-- fs/f2fs/segment.c | 10 ++++++---- fs/f2fs/super.c | 6 ++++-- 10 files changed, 25 insertions(+), 15 deletions(-) diff --git a/fs/f2fs/checkpoint.c b/fs/f2fs/checkpoint.c index 0dea31f1ddad..46fc9c1542fe 100644 --- a/fs/f2fs/checkpoint.c +++ b/fs/f2fs/checkpoint.c @@ -1260,7 +1260,7 @@ void f2fs_wait_on_all_pages(struct f2fs_sb_info *sbi, int type) if (unlikely(f2fs_cp_error(sbi))) break; - io_schedule_timeout(HZ/50); + io_schedule_timeout(DEFAULT_IO_TIMEOUT); } finish_wait(&sbi->cp_wait, &wait); } diff --git a/fs/f2fs/compress.c b/fs/f2fs/compress.c index 5ef7d7cfc0e7..748933813cd1 100644 --- a/fs/f2fs/compress.c +++ b/fs/f2fs/compress.c @@ -988,7 +988,8 @@ retry_write: } else if (ret == -EAGAIN) { ret = 0; cond_resched(); - congestion_wait(BLK_RW_ASYNC, HZ/50); + congestion_wait(BLK_RW_ASYNC, + DEFAULT_IO_TIMEOUT); lock_page(cc->rpages[i]); clear_page_dirty_for_io(cc->rpages[i]); goto retry_write; diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c index 04cf514ca36d..921e1fe6aecd 100644 --- a/fs/f2fs/data.c +++ b/fs/f2fs/data.c @@ -2317,7 +2317,7 @@ retry_encrypt: /* flush pending IOs and wait for a while in the ENOMEM case */ if (PTR_ERR(fio->encrypted_page) == -ENOMEM) { f2fs_flush_merged_writes(fio->sbi); - congestion_wait(BLK_RW_ASYNC, HZ/50); + congestion_wait(BLK_RW_ASYNC, DEFAULT_IO_TIMEOUT); gfp_flags |= __GFP_NOFAIL; goto retry_encrypt; } @@ -2908,7 +2908,7 @@ result: if (wbc->sync_mode == WB_SYNC_ALL) { cond_resched(); congestion_wait(BLK_RW_ASYNC, - HZ/50); + DEFAULT_IO_TIMEOUT); goto retry_write; } goto next; diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index ef3ec6548629..241a87d0d71c 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -558,6 +558,9 @@ enum { #define DEFAULT_RETRY_IO_COUNT 8 /* maximum retry read IO count */ +/* congestion wait timeout value, default: 20ms */ +#define DEFAULT_IO_TIMEOUT (msecs_to_jiffies(20)) + /* maximum retry quota flush count */ #define DEFAULT_RETRY_QUOTA_FLUSH_COUNT 8 diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c index 0dfdec4926e4..f6958ae8c157 100644 --- a/fs/f2fs/gc.c +++ b/fs/f2fs/gc.c @@ -977,7 +977,8 @@ retry: if (err) { clear_cold_data(page); if (err == -ENOMEM) { - congestion_wait(BLK_RW_ASYNC, HZ/50); + congestion_wait(BLK_RW_ASYNC, + DEFAULT_IO_TIMEOUT); goto retry; } if (is_dirty) diff --git a/fs/f2fs/inode.c b/fs/f2fs/inode.c index 45aff90af5c5..44e08bf2e2b4 100644 --- a/fs/f2fs/inode.c +++ b/fs/f2fs/inode.c @@ -535,7 +535,7 @@ retry: inode = f2fs_iget(sb, ino); if (IS_ERR(inode)) { if (PTR_ERR(inode) == -ENOMEM) { - congestion_wait(BLK_RW_ASYNC, HZ/50); + congestion_wait(BLK_RW_ASYNC, DEFAULT_IO_TIMEOUT); goto retry; } } diff --git a/fs/f2fs/node.c b/fs/f2fs/node.c index bdcab8aeb542..9f827d851de9 100644 --- a/fs/f2fs/node.c +++ b/fs/f2fs/node.c @@ -2599,7 +2599,7 @@ int f2fs_recover_inode_page(struct f2fs_sb_info *sbi, struct page *page) retry: ipage = f2fs_grab_cache_page(NODE_MAPPING(sbi), ino, false); if (!ipage) { - congestion_wait(BLK_RW_ASYNC, HZ/50); + congestion_wait(BLK_RW_ASYNC, DEFAULT_IO_TIMEOUT); goto retry; } diff --git a/fs/f2fs/recovery.c b/fs/f2fs/recovery.c index 348e8d463b3e..dd804c07eeb0 100644 --- a/fs/f2fs/recovery.c +++ b/fs/f2fs/recovery.c @@ -534,7 +534,7 @@ retry_dn: err = f2fs_get_dnode_of_data(&dn, start, ALLOC_NODE); if (err) { if (err == -ENOMEM) { - congestion_wait(BLK_RW_ASYNC, HZ/50); + congestion_wait(BLK_RW_ASYNC, DEFAULT_IO_TIMEOUT); goto retry_dn; } goto out; @@ -617,7 +617,8 @@ retry_prev: err = check_index_in_prev_nodes(sbi, dest, &dn); if (err) { if (err == -ENOMEM) { - congestion_wait(BLK_RW_ASYNC, HZ/50); + congestion_wait(BLK_RW_ASYNC, + DEFAULT_IO_TIMEOUT); goto retry_prev; } goto err; diff --git a/fs/f2fs/segment.c b/fs/f2fs/segment.c index 04156df83e7a..a2beb0e81440 100644 --- a/fs/f2fs/segment.c +++ b/fs/f2fs/segment.c @@ -245,7 +245,8 @@ retry: LOOKUP_NODE); if (err) { if (err == -ENOMEM) { - congestion_wait(BLK_RW_ASYNC, HZ/50); + congestion_wait(BLK_RW_ASYNC, + DEFAULT_IO_TIMEOUT); cond_resched(); goto retry; } @@ -312,7 +313,7 @@ next: skip: iput(inode); } - congestion_wait(BLK_RW_ASYNC, HZ/50); + congestion_wait(BLK_RW_ASYNC, DEFAULT_IO_TIMEOUT); cond_resched(); if (gc_failure) { if (++looped >= count) @@ -415,7 +416,8 @@ retry: err = f2fs_do_write_data_page(&fio); if (err) { if (err == -ENOMEM) { - congestion_wait(BLK_RW_ASYNC, HZ/50); + congestion_wait(BLK_RW_ASYNC, + DEFAULT_IO_TIMEOUT); cond_resched(); goto retry; } @@ -2801,7 +2803,7 @@ next: blk_finish_plug(&plug); mutex_unlock(&dcc->cmd_lock); trimmed += __wait_all_discard_cmd(sbi, NULL); - congestion_wait(BLK_RW_ASYNC, HZ/50); + congestion_wait(BLK_RW_ASYNC, DEFAULT_IO_TIMEOUT); goto next; } skip: diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c index 40d0a6190be1..350ec725bc3b 100644 --- a/fs/f2fs/super.c +++ b/fs/f2fs/super.c @@ -1885,7 +1885,8 @@ repeat: page = read_cache_page_gfp(mapping, blkidx, GFP_NOFS); if (IS_ERR(page)) { if (PTR_ERR(page) == -ENOMEM) { - congestion_wait(BLK_RW_ASYNC, HZ/50); + congestion_wait(BLK_RW_ASYNC, + DEFAULT_IO_TIMEOUT); goto repeat; } set_sbi_flag(F2FS_SB(sb), SBI_QUOTA_NEED_REPAIR); @@ -1939,7 +1940,8 @@ retry: &page, NULL); if (unlikely(err)) { if (err == -ENOMEM) { - congestion_wait(BLK_RW_ASYNC, HZ/50); + congestion_wait(BLK_RW_ASYNC, + DEFAULT_IO_TIMEOUT); goto retry; } set_sbi_flag(F2FS_SB(sb), SBI_QUOTA_NEED_REPAIR); From 2f2dd104f45de20692af3788f10acf43451780ff Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Mon, 17 Feb 2020 17:46:20 +0800 Subject: [PATCH 30/86] f2fs: add prefix for f2fs slab cache name In order to avoid polluting global slab cache namespace. Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/data.c | 2 +- fs/f2fs/node.c | 8 ++++---- fs/f2fs/segment.c | 8 ++++---- 3 files changed, 9 insertions(+), 9 deletions(-) diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c index 921e1fe6aecd..4e8e70555688 100644 --- a/fs/f2fs/data.c +++ b/fs/f2fs/data.c @@ -3846,7 +3846,7 @@ void f2fs_destroy_post_read_wq(struct f2fs_sb_info *sbi) int __init f2fs_init_bio_entry_cache(void) { - bio_entry_slab = f2fs_kmem_cache_create("bio_entry_slab", + bio_entry_slab = f2fs_kmem_cache_create("f2fs_bio_entry_slab", sizeof(struct bio_entry)); if (!bio_entry_slab) return -ENOMEM; diff --git a/fs/f2fs/node.c b/fs/f2fs/node.c index 9f827d851de9..07214e0dab26 100644 --- a/fs/f2fs/node.c +++ b/fs/f2fs/node.c @@ -3190,22 +3190,22 @@ void f2fs_destroy_node_manager(struct f2fs_sb_info *sbi) int __init f2fs_create_node_manager_caches(void) { - nat_entry_slab = f2fs_kmem_cache_create("nat_entry", + nat_entry_slab = f2fs_kmem_cache_create("f2fs_nat_entry", sizeof(struct nat_entry)); if (!nat_entry_slab) goto fail; - free_nid_slab = f2fs_kmem_cache_create("free_nid", + free_nid_slab = f2fs_kmem_cache_create("f2fs_free_nid", sizeof(struct free_nid)); if (!free_nid_slab) goto destroy_nat_entry; - nat_entry_set_slab = f2fs_kmem_cache_create("nat_entry_set", + nat_entry_set_slab = f2fs_kmem_cache_create("f2fs_nat_entry_set", sizeof(struct nat_entry_set)); if (!nat_entry_set_slab) goto destroy_free_nid; - fsync_node_entry_slab = f2fs_kmem_cache_create("fsync_node_entry", + fsync_node_entry_slab = f2fs_kmem_cache_create("f2fs_fsync_node_entry", sizeof(struct fsync_node_entry)); if (!fsync_node_entry_slab) goto destroy_nat_entry_set; diff --git a/fs/f2fs/segment.c b/fs/f2fs/segment.c index a2beb0e81440..601d67e72c50 100644 --- a/fs/f2fs/segment.c +++ b/fs/f2fs/segment.c @@ -4832,22 +4832,22 @@ void f2fs_destroy_segment_manager(struct f2fs_sb_info *sbi) int __init f2fs_create_segment_manager_caches(void) { - discard_entry_slab = f2fs_kmem_cache_create("discard_entry", + discard_entry_slab = f2fs_kmem_cache_create("f2fs_discard_entry", sizeof(struct discard_entry)); if (!discard_entry_slab) goto fail; - discard_cmd_slab = f2fs_kmem_cache_create("discard_cmd", + discard_cmd_slab = f2fs_kmem_cache_create("f2fs_discard_cmd", sizeof(struct discard_cmd)); if (!discard_cmd_slab) goto destroy_discard_entry; - sit_entry_set_slab = f2fs_kmem_cache_create("sit_entry_set", + sit_entry_set_slab = f2fs_kmem_cache_create("f2fs_sit_entry_set", sizeof(struct sit_entry_set)); if (!sit_entry_set_slab) goto destroy_discard_cmd; - inmem_entry_slab = f2fs_kmem_cache_create("inmem_page_entry", + inmem_entry_slab = f2fs_kmem_cache_create("f2fs_inmem_page_entry", sizeof(struct inmem_pages)); if (!inmem_entry_slab) goto destroy_sit_entry_set; From c628779520f1fdad2025c9a64157af792b8b85b7 Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Tue, 18 Feb 2020 18:21:35 +0800 Subject: [PATCH 31/86] f2fs: fix to avoid triggering IO in write path If we are in write IO path, we need to avoid using GFP_KERNEL. Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/compress.c | 2 +- fs/f2fs/data.c | 24 +++++++++++++----------- fs/f2fs/f2fs.h | 2 +- 3 files changed, 15 insertions(+), 13 deletions(-) diff --git a/fs/f2fs/compress.c b/fs/f2fs/compress.c index 748933813cd1..bd3ea01db448 100644 --- a/fs/f2fs/compress.c +++ b/fs/f2fs/compress.c @@ -656,7 +656,7 @@ retry: struct bio *bio = NULL; ret = f2fs_read_multi_pages(cc, &bio, cc->cluster_size, - &last_block_in_bio, false); + &last_block_in_bio, false, true); f2fs_destroy_compress_ctx(cc); if (ret) goto release_pages; diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c index 4e8e70555688..896364b825a0 100644 --- a/fs/f2fs/data.c +++ b/fs/f2fs/data.c @@ -925,14 +925,15 @@ static inline bool f2fs_need_verity(const struct inode *inode, pgoff_t idx) static struct bio *f2fs_grab_read_bio(struct inode *inode, block_t blkaddr, unsigned nr_pages, unsigned op_flag, - pgoff_t first_idx) + pgoff_t first_idx, bool for_write) { struct f2fs_sb_info *sbi = F2FS_I_SB(inode); struct bio *bio; struct bio_post_read_ctx *ctx; unsigned int post_read_steps = 0; - bio = f2fs_bio_alloc(sbi, min_t(int, nr_pages, BIO_MAX_PAGES), false); + bio = f2fs_bio_alloc(sbi, min_t(int, nr_pages, BIO_MAX_PAGES), + for_write); if (!bio) return ERR_PTR(-ENOMEM); f2fs_target_device(sbi, blkaddr, bio); @@ -967,12 +968,12 @@ static void f2fs_release_read_bio(struct bio *bio) /* This can handle encryption stuffs */ static int f2fs_submit_page_read(struct inode *inode, struct page *page, - block_t blkaddr) + block_t blkaddr, bool for_write) { struct f2fs_sb_info *sbi = F2FS_I_SB(inode); struct bio *bio; - bio = f2fs_grab_read_bio(inode, blkaddr, 1, 0, page->index); + bio = f2fs_grab_read_bio(inode, blkaddr, 1, 0, page->index, for_write); if (IS_ERR(bio)) return PTR_ERR(bio); @@ -1158,7 +1159,7 @@ got_it: return page; } - err = f2fs_submit_page_read(inode, page, dn.data_blkaddr); + err = f2fs_submit_page_read(inode, page, dn.data_blkaddr, for_write); if (err) goto put_err; return page; @@ -1971,7 +1972,8 @@ submit_and_realloc: } if (bio == NULL) { bio = f2fs_grab_read_bio(inode, block_nr, nr_pages, - is_readahead ? REQ_RAHEAD : 0, page->index); + is_readahead ? REQ_RAHEAD : 0, page->index, + false); if (IS_ERR(bio)) { ret = PTR_ERR(bio); bio = NULL; @@ -2006,7 +2008,7 @@ out: #ifdef CONFIG_F2FS_FS_COMPRESSION int f2fs_read_multi_pages(struct compress_ctx *cc, struct bio **bio_ret, unsigned nr_pages, sector_t *last_block_in_bio, - bool is_readahead) + bool is_readahead, bool for_write) { struct dnode_of_data dn; struct inode *inode = cc->inode; @@ -2100,7 +2102,7 @@ submit_and_realloc: if (!bio) { bio = f2fs_grab_read_bio(inode, blkaddr, nr_pages, is_readahead ? REQ_RAHEAD : 0, - page->index); + page->index, for_write); if (IS_ERR(bio)) { ret = PTR_ERR(bio); bio = NULL; @@ -2201,7 +2203,7 @@ int f2fs_mpage_readpages(struct address_space *mapping, ret = f2fs_read_multi_pages(&cc, &bio, max_nr_pages, &last_block_in_bio, - is_readahead); + is_readahead, false); f2fs_destroy_compress_ctx(&cc); if (ret) goto set_error_page; @@ -2244,7 +2246,7 @@ next_page: ret = f2fs_read_multi_pages(&cc, &bio, max_nr_pages, &last_block_in_bio, - is_readahead); + is_readahead, false); f2fs_destroy_compress_ctx(&cc); } } @@ -3274,7 +3276,7 @@ repeat: err = -EFSCORRUPTED; goto fail; } - err = f2fs_submit_page_read(inode, page, blkaddr); + err = f2fs_submit_page_read(inode, page, blkaddr, true); if (err) goto fail; diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index 241a87d0d71c..14f7af37ff4b 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -3796,7 +3796,7 @@ int f2fs_write_multi_pages(struct compress_ctx *cc, int f2fs_is_compressed_cluster(struct inode *inode, pgoff_t index); int f2fs_read_multi_pages(struct compress_ctx *cc, struct bio **bio_ret, unsigned nr_pages, sector_t *last_block_in_bio, - bool is_readahead); + bool is_readahead, bool for_write); struct decompress_io_ctx *f2fs_alloc_dic(struct compress_ctx *cc); void f2fs_free_dic(struct decompress_io_ctx *dic); void f2fs_decompress_end_io(struct page **rpages, From 50893b23d238a7074dd3e2a30eadd1301787c691 Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Fri, 21 Feb 2020 18:09:21 +0800 Subject: [PATCH 32/86] f2fs: introduce F2FS_IOC_GET_COMPRESS_BLOCKS With this newly introduced interface, user can get block number compression saved in target inode. Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/f2fs.h | 1 + fs/f2fs/file.c | 18 ++++++++++++++++++ 2 files changed, 19 insertions(+) diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index 14f7af37ff4b..60e9da1c1c0a 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -426,6 +426,7 @@ static inline bool __has_cursum_space(struct f2fs_journal *journal, #define F2FS_IOC_GET_PIN_FILE _IOR(F2FS_IOCTL_MAGIC, 14, __u32) #define F2FS_IOC_PRECACHE_EXTENTS _IO(F2FS_IOCTL_MAGIC, 15) #define F2FS_IOC_RESIZE_FS _IOW(F2FS_IOCTL_MAGIC, 16, __u64) +#define F2FS_IOC_GET_COMPRESS_BLOCKS _IOR(F2FS_IOCTL_MAGIC, 17, __u64) #define F2FS_IOC_GET_VOLUME_NAME FS_IOC_GETFSLABEL #define F2FS_IOC_SET_VOLUME_NAME FS_IOC_SETFSLABEL diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c index 2e8276d54378..178052490a0e 100644 --- a/fs/f2fs/file.c +++ b/fs/f2fs/file.c @@ -3404,6 +3404,21 @@ out: return err; } +static int f2fs_get_compress_blocks(struct file *filp, unsigned long arg) +{ + struct inode *inode = file_inode(filp); + __u64 blocks; + + if (!f2fs_sb_has_compression(F2FS_I_SB(inode))) + return -EOPNOTSUPP; + + if (!f2fs_compressed_file(inode)) + return -EINVAL; + + blocks = F2FS_I(inode)->i_compr_blocks; + return put_user(blocks, (u64 __user *)arg); +} + long f2fs_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) { if (unlikely(f2fs_cp_error(F2FS_I_SB(file_inode(filp))))) @@ -3482,6 +3497,8 @@ long f2fs_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) return f2fs_get_volume_name(filp, arg); case F2FS_IOC_SET_VOLUME_NAME: return f2fs_set_volume_name(filp, arg); + case F2FS_IOC_GET_COMPRESS_BLOCKS: + return f2fs_get_compress_blocks(filp, arg); default: return -ENOTTY; } @@ -3638,6 +3655,7 @@ long f2fs_compat_ioctl(struct file *file, unsigned int cmd, unsigned long arg) case FS_IOC_MEASURE_VERITY: case F2FS_IOC_GET_VOLUME_NAME: case F2FS_IOC_SET_VOLUME_NAME: + case F2FS_IOC_GET_COMPRESS_BLOCKS: break; default: return -ENOIOCTLCMD; From 8ae3cefc7ee9ca62bdae97fee7bf0e988532cdc7 Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Tue, 18 Feb 2020 18:21:34 +0800 Subject: [PATCH 33/86] f2fs: avoid __GFP_NOFAIL in f2fs_bio_alloc __f2fs_bio_alloc() won't fail due to memory pool backend, remove unneeded __GFP_NOFAIL flag in __f2fs_bio_alloc(). Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/data.c | 12 ++++-------- fs/f2fs/f2fs.h | 2 +- 2 files changed, 5 insertions(+), 9 deletions(-) diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c index 896364b825a0..aa29b33f1537 100644 --- a/fs/f2fs/data.c +++ b/fs/f2fs/data.c @@ -54,17 +54,13 @@ static inline struct bio *__f2fs_bio_alloc(gfp_t gfp_mask, return bio_alloc_bioset(gfp_mask, nr_iovecs, &f2fs_bioset); } -struct bio *f2fs_bio_alloc(struct f2fs_sb_info *sbi, int npages, bool no_fail) +struct bio *f2fs_bio_alloc(struct f2fs_sb_info *sbi, int npages, bool noio) { - struct bio *bio; - - if (no_fail) { + if (noio) { /* No failure on bio allocation */ - bio = __f2fs_bio_alloc(GFP_NOIO, npages); - if (!bio) - bio = __f2fs_bio_alloc(GFP_NOIO | __GFP_NOFAIL, npages); - return bio; + return __f2fs_bio_alloc(GFP_NOIO, npages); } + if (time_to_inject(sbi, FAULT_ALLOC_BIO)) { f2fs_show_injection_info(sbi, FAULT_ALLOC_BIO); return NULL; diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index 60e9da1c1c0a..baaf7f0fd02a 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -3341,7 +3341,7 @@ void f2fs_destroy_checkpoint_caches(void); */ int __init f2fs_init_bioset(void); void f2fs_destroy_bioset(void); -struct bio *f2fs_bio_alloc(struct f2fs_sb_info *sbi, int npages, bool no_fail); +struct bio *f2fs_bio_alloc(struct f2fs_sb_info *sbi, int npages, bool noio); int f2fs_init_bio_entry_cache(void); void f2fs_destroy_bio_entry_cache(void); void f2fs_submit_bio(struct f2fs_sb_info *sbi, From 586d27acc8aea3e8790e7f30bd1221ada207031d Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Mon, 24 Feb 2020 19:20:15 +0800 Subject: [PATCH 34/86] f2fs: fix to show tracepoint correctly Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/file.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c index 178052490a0e..f0152181f3e8 100644 --- a/fs/f2fs/file.c +++ b/fs/f2fs/file.c @@ -3526,8 +3526,10 @@ static ssize_t f2fs_file_write_iter(struct kiocb *iocb, struct iov_iter *from) goto out; } - if (!f2fs_is_compress_backend_ready(inode)) - return -EOPNOTSUPP; + if (!f2fs_is_compress_backend_ready(inode)) { + ret = -EOPNOTSUPP; + goto out; + } if (iocb->ki_flags & IOCB_NOWAIT) { if (!inode_trylock(inode)) { From 394959e71f85aa01321194a7251f8949d36b2ef3 Mon Sep 17 00:00:00 2001 From: Jaegeuk Kim Date: Sun, 9 Feb 2020 13:28:45 -0800 Subject: [PATCH 35/86] f2fs: skip migration only when BG_GC is called FG_GC needs to move entire section more quickly. Reviewed-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/gc.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c index f6958ae8c157..9ead93fcf78a 100644 --- a/fs/f2fs/gc.c +++ b/fs/f2fs/gc.c @@ -1211,7 +1211,7 @@ static int do_garbage_collect(struct f2fs_sb_info *sbi, if (get_valid_blocks(sbi, segno, false) == 0) goto freed; - if (__is_large_section(sbi) && + if (gc_type == BG_GC && __is_large_section(sbi) && migrated >= sbi->migration_granularity) goto skip; if (!PageUptodate(sum_page) || unlikely(f2fs_cp_error(sbi))) From 41257e7726a51ed0088c84b8683390403b8a9071 Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Tue, 25 Feb 2020 18:17:10 +0800 Subject: [PATCH 36/86] f2fs: use kmem_cache pool during inline xattr lookups It's been observed that kzalloc() on lookup_all_xattrs() are called millions of times on Android, quickly becoming the top abuser of slub memory allocator. Use a dedicated kmem cache pool for xattr lookups to mitigate this. Signed-off-by: Park Ju Hyung Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/f2fs.h | 3 +++ fs/f2fs/super.c | 10 ++++++++- fs/f2fs/xattr.c | 54 ++++++++++++++++++++++++++++++++++++++++++++----- fs/f2fs/xattr.h | 4 ++++ 4 files changed, 65 insertions(+), 6 deletions(-) diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index baaf7f0fd02a..3e727a6af08e 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -1487,6 +1487,9 @@ struct f2fs_sb_info { __u32 s_chksum_seed; struct workqueue_struct *post_read_wq; /* post read workqueue */ + + struct kmem_cache *inline_xattr_slab; /* inline xattr entry */ + unsigned int inline_xattr_slab_size; /* default inline xattr slab size */ }; struct f2fs_private_dio { diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c index 350ec725bc3b..c1f2c03a2e5a 100644 --- a/fs/f2fs/super.c +++ b/fs/f2fs/super.c @@ -1202,6 +1202,7 @@ static void f2fs_put_super(struct super_block *sb) kvfree(sbi->raw_super); destroy_device_list(sbi); + f2fs_destroy_xattr_caches(sbi); mempool_destroy(sbi->write_io_dummy); #ifdef CONFIG_QUOTA for (i = 0; i < MAXQUOTAS; i++) @@ -3458,12 +3459,17 @@ try_onemore: } } + /* init per sbi slab cache */ + err = f2fs_init_xattr_caches(sbi); + if (err) + goto free_io_dummy; + /* get an inode for meta space */ sbi->meta_inode = f2fs_iget(sb, F2FS_META_INO(sbi)); if (IS_ERR(sbi->meta_inode)) { f2fs_err(sbi, "Failed to read F2FS meta data inode"); err = PTR_ERR(sbi->meta_inode); - goto free_io_dummy; + goto free_xattr_cache; } err = f2fs_get_valid_checkpoint(sbi); @@ -3736,6 +3742,8 @@ free_meta_inode: make_bad_inode(sbi->meta_inode); iput(sbi->meta_inode); sbi->meta_inode = NULL; +free_xattr_cache: + f2fs_destroy_xattr_caches(sbi); free_io_dummy: mempool_destroy(sbi->write_io_dummy); free_percpu: diff --git a/fs/f2fs/xattr.c b/fs/f2fs/xattr.c index d4800df6302d..4f6582ef7ee3 100644 --- a/fs/f2fs/xattr.c +++ b/fs/f2fs/xattr.c @@ -23,6 +23,25 @@ #include "xattr.h" #include "segment.h" +static void *xattr_alloc(struct f2fs_sb_info *sbi, int size, bool *is_inline) +{ + if (likely(size == sbi->inline_xattr_slab_size)) { + *is_inline = true; + return kmem_cache_zalloc(sbi->inline_xattr_slab, GFP_NOFS); + } + *is_inline = false; + return f2fs_kzalloc(sbi, size, GFP_NOFS); +} + +static void xattr_free(struct f2fs_sb_info *sbi, void *xattr_addr, + bool is_inline) +{ + if (is_inline) + kmem_cache_free(sbi->inline_xattr_slab, xattr_addr); + else + kvfree(xattr_addr); +} + static int f2fs_xattr_generic_get(const struct xattr_handler *handler, struct dentry *unused, struct inode *inode, const char *name, void *buffer, size_t size) @@ -301,7 +320,8 @@ static int read_xattr_block(struct inode *inode, void *txattr_addr) static int lookup_all_xattrs(struct inode *inode, struct page *ipage, unsigned int index, unsigned int len, const char *name, struct f2fs_xattr_entry **xe, - void **base_addr, int *base_size) + void **base_addr, int *base_size, + bool *is_inline) { void *cur_addr, *txattr_addr, *last_txattr_addr; void *last_addr = NULL; @@ -313,7 +333,7 @@ static int lookup_all_xattrs(struct inode *inode, struct page *ipage, return -ENODATA; *base_size = XATTR_SIZE(inode) + XATTR_PADDING_SIZE; - txattr_addr = f2fs_kzalloc(F2FS_I_SB(inode), *base_size, GFP_NOFS); + txattr_addr = xattr_alloc(F2FS_I_SB(inode), *base_size, is_inline); if (!txattr_addr) return -ENOMEM; @@ -362,7 +382,7 @@ check: *base_addr = txattr_addr; return 0; out: - kvfree(txattr_addr); + xattr_free(F2FS_I_SB(inode), txattr_addr, *is_inline); return err; } @@ -499,6 +519,7 @@ int f2fs_getxattr(struct inode *inode, int index, const char *name, unsigned int size, len; void *base_addr = NULL; int base_size; + bool is_inline; if (name == NULL) return -EINVAL; @@ -509,7 +530,7 @@ int f2fs_getxattr(struct inode *inode, int index, const char *name, down_read(&F2FS_I(inode)->i_xattr_sem); error = lookup_all_xattrs(inode, ipage, index, len, name, - &entry, &base_addr, &base_size); + &entry, &base_addr, &base_size, &is_inline); up_read(&F2FS_I(inode)->i_xattr_sem); if (error) return error; @@ -532,7 +553,7 @@ int f2fs_getxattr(struct inode *inode, int index, const char *name, } error = size; out: - kvfree(base_addr); + xattr_free(F2FS_I_SB(inode), base_addr, is_inline); return error; } @@ -764,3 +785,26 @@ int f2fs_setxattr(struct inode *inode, int index, const char *name, f2fs_update_time(sbi, REQ_TIME); return err; } + +int f2fs_init_xattr_caches(struct f2fs_sb_info *sbi) +{ + dev_t dev = sbi->sb->s_bdev->bd_dev; + char slab_name[32]; + + sprintf(slab_name, "f2fs_xattr_entry-%u:%u", MAJOR(dev), MINOR(dev)); + + sbi->inline_xattr_slab_size = F2FS_OPTION(sbi).inline_xattr_size * + sizeof(__le32) + XATTR_PADDING_SIZE; + + sbi->inline_xattr_slab = f2fs_kmem_cache_create(slab_name, + sbi->inline_xattr_slab_size); + if (!sbi->inline_xattr_slab) + return -ENOMEM; + + return 0; +} + +void f2fs_destroy_xattr_caches(struct f2fs_sb_info *sbi) +{ + kmem_cache_destroy(sbi->inline_xattr_slab); +} diff --git a/fs/f2fs/xattr.h b/fs/f2fs/xattr.h index 574beea46494..0153b4c9ef21 100644 --- a/fs/f2fs/xattr.h +++ b/fs/f2fs/xattr.h @@ -131,6 +131,8 @@ extern int f2fs_setxattr(struct inode *, int, const char *, extern int f2fs_getxattr(struct inode *, int, const char *, void *, size_t, struct page *); extern ssize_t f2fs_listxattr(struct dentry *, char *, size_t); +extern int f2fs_init_xattr_caches(struct f2fs_sb_info *); +extern void f2fs_destroy_xattr_caches(struct f2fs_sb_info *); #else #define f2fs_xattr_handlers NULL @@ -151,6 +153,8 @@ static inline ssize_t f2fs_listxattr(struct dentry *dentry, char *buffer, { return -EOPNOTSUPP; } +static int f2fs_init_xattr_caches(struct f2fs_sb_info *sbi) { return 0; } +static void f2fs_destroy_xattr_caches(struct f2fs_sb_info *sbi) { } #endif #ifdef CONFIG_F2FS_FS_SECURITY From 1440f98dc558785aac1caaee60ef1d2e98e8868b Mon Sep 17 00:00:00 2001 From: Sahitya Tummala Date: Tue, 3 Mar 2020 19:59:25 +0530 Subject: [PATCH 37/86] f2fs: Fix mount failure due to SPO after a successful online resize FS Even though online resize is successfully done, a SPO immediately after resize, still causes below error in the next mount. [ 11.294650] F2FS-fs (sda8): Wrong user_block_count: 2233856 [ 11.300272] F2FS-fs (sda8): Failed to get valid F2FS checkpoint This is because after FS metadata is updated in update_fs_metadata() if the SBI_IS_DIRTY is not dirty, then CP will not be done to reflect the new user_block_count. Signed-off-by: Sahitya Tummala Reviewed-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/gc.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c index 9ead93fcf78a..bc203e24a039 100644 --- a/fs/f2fs/gc.c +++ b/fs/f2fs/gc.c @@ -1578,11 +1578,17 @@ int f2fs_resize_fs(struct f2fs_sb_info *sbi, __u64 block_count) goto out; } + mutex_lock(&sbi->cp_mutex); update_fs_metadata(sbi, -secs); clear_sbi_flag(sbi, SBI_IS_RESIZEFS); + set_sbi_flag(sbi, SBI_IS_DIRTY); + mutex_unlock(&sbi->cp_mutex); + err = f2fs_sync_fs(sbi->sb, 1); if (err) { + mutex_lock(&sbi->cp_mutex); update_fs_metadata(sbi, secs); + mutex_unlock(&sbi->cp_mutex); update_sb_metadata(sbi, secs); f2fs_commit_super(sbi, false); } From 8449e28786ff3181b5e9b13ba749c7830bc5b816 Mon Sep 17 00:00:00 2001 From: Sahitya Tummala Date: Tue, 3 Mar 2020 19:59:26 +0530 Subject: [PATCH 38/86] f2fs: Add a new CP flag to help fsck fix resize SPO issues Add and set a new CP flag CP_RESIZEFS_FLAG during online resize FS to help fsck fix the metadata mismatch that may happen due to SPO during resize, where SB got updated but CP data couldn't be written yet. fsck errors - Info: CKPT version = 6ed7bccb Wrong user_block_count(2233856) [f2fs_do_mount:3365] Checkpoint is polluted Signed-off-by: Sahitya Tummala Reviewed-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/checkpoint.c | 8 ++++++-- include/linux/f2fs_fs.h | 1 + 2 files changed, 7 insertions(+), 2 deletions(-) diff --git a/fs/f2fs/checkpoint.c b/fs/f2fs/checkpoint.c index 46fc9c1542fe..852890b72d6a 100644 --- a/fs/f2fs/checkpoint.c +++ b/fs/f2fs/checkpoint.c @@ -1298,10 +1298,14 @@ static void update_ckpt_flags(struct f2fs_sb_info *sbi, struct cp_control *cpc) else __clear_ckpt_flags(ckpt, CP_ORPHAN_PRESENT_FLAG); - if (is_sbi_flag_set(sbi, SBI_NEED_FSCK) || - is_sbi_flag_set(sbi, SBI_IS_RESIZEFS)) + if (is_sbi_flag_set(sbi, SBI_NEED_FSCK)) __set_ckpt_flags(ckpt, CP_FSCK_FLAG); + if (is_sbi_flag_set(sbi, SBI_IS_RESIZEFS)) + __set_ckpt_flags(ckpt, CP_RESIZEFS_FLAG); + else + __clear_ckpt_flags(ckpt, CP_RESIZEFS_FLAG); + if (is_sbi_flag_set(sbi, SBI_CP_DISABLED)) __set_ckpt_flags(ckpt, CP_DISABLED_FLAG); else diff --git a/include/linux/f2fs_fs.h b/include/linux/f2fs_fs.h index ac3f4888b3df..3c383ddd92dd 100644 --- a/include/linux/f2fs_fs.h +++ b/include/linux/f2fs_fs.h @@ -125,6 +125,7 @@ struct f2fs_super_block { /* * For checkpoint */ +#define CP_RESIZEFS_FLAG 0x00004000 #define CP_DISABLED_QUICK_FLAG 0x00002000 #define CP_DISABLED_FLAG 0x00001000 #define CP_QUOTA_NEED_FSCK_FLAG 0x00000800 From 76f3725f75db4c4ae3f7c4d28a6e197c7d704640 Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Tue, 3 Mar 2020 20:09:25 +0800 Subject: [PATCH 39/86] f2fs: fix to update f2fs_super_block fields under sb_lock Fields in struct f2fs_super_block should be updated under coverage of sb_lock, fix to adjust update_sb_metadata() for that rule. Fixes: 04f0b2eaa3b3 ("f2fs: ioctl for removing a range from F2FS") Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/gc.c | 17 +++++++++++++---- 1 file changed, 13 insertions(+), 4 deletions(-) diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c index bc203e24a039..f122fe3dbba3 100644 --- a/fs/f2fs/gc.c +++ b/fs/f2fs/gc.c @@ -1442,12 +1442,19 @@ static int free_segment_range(struct f2fs_sb_info *sbi, unsigned int start, static void update_sb_metadata(struct f2fs_sb_info *sbi, int secs) { struct f2fs_super_block *raw_sb = F2FS_RAW_SUPER(sbi); - int section_count = le32_to_cpu(raw_sb->section_count); - int segment_count = le32_to_cpu(raw_sb->segment_count); - int segment_count_main = le32_to_cpu(raw_sb->segment_count_main); - long long block_count = le64_to_cpu(raw_sb->block_count); + int section_count; + int segment_count; + int segment_count_main; + long long block_count; int segs = secs * sbi->segs_per_sec; + down_write(&sbi->sb_lock); + + section_count = le32_to_cpu(raw_sb->section_count); + segment_count = le32_to_cpu(raw_sb->segment_count); + segment_count_main = le32_to_cpu(raw_sb->segment_count_main); + block_count = le64_to_cpu(raw_sb->block_count); + raw_sb->section_count = cpu_to_le32(section_count + secs); raw_sb->segment_count = cpu_to_le32(segment_count + segs); raw_sb->segment_count_main = cpu_to_le32(segment_count_main + segs); @@ -1461,6 +1468,8 @@ static void update_sb_metadata(struct f2fs_sb_info *sbi, int secs) raw_sb->devs[last_dev].total_segments = cpu_to_le32(dev_segs + segs); } + + up_write(&sbi->sb_lock); } static void update_fs_metadata(struct f2fs_sb_info *sbi, int secs) From 0887a35951fee4baa7ef0881981dfb99a35a429d Mon Sep 17 00:00:00 2001 From: "Gustavo A. R. Silva" Date: Mon, 9 Mar 2020 13:10:08 -0500 Subject: [PATCH 40/86] f2fs: xattr.h: Replace zero-length array with flexible-array member The current codebase makes use of the zero-length array language extension to the C90 standard, but the preferred mechanism to declare variable-length types such as these ones is a flexible array member[1][2], introduced in C99: struct foo { int stuff; struct boo array[]; }; By making use of the mechanism above, we will get a compiler warning in case the flexible array does not occur last in the structure, which will help us prevent some kind of undefined behavior bugs from being inadvertently introduced[3] to the codebase from now on. Also, notice that, dynamic memory allocations won't be affected by this change: "Flexible array members have incomplete type, and so the sizeof operator may not be applied. As a quirk of the original implementation of zero-length arrays, sizeof evaluates to zero."[1] This issue was found with the help of Coccinelle. [1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html [2] https://github.com/KSPP/linux/issues/21 [3] commit 76497732932f ("cxgb3/l2t: Fix undefined behaviour") Signed-off-by: Gustavo A. R. Silva Reviewed-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/xattr.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/f2fs/xattr.h b/fs/f2fs/xattr.h index 0153b4c9ef21..e471be77f8f0 100644 --- a/fs/f2fs/xattr.h +++ b/fs/f2fs/xattr.h @@ -49,7 +49,7 @@ struct f2fs_xattr_entry { __u8 e_name_index; __u8 e_name_len; __le16 e_value_size; /* size of attribute value */ - char e_name[0]; /* attribute name */ + char e_name[]; /* attribute name */ }; #define XATTR_HDR(ptr) ((struct f2fs_xattr_header *)(ptr)) From 276321288ece8c982f3df24047b737afd541a9fc Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Thu, 12 Mar 2020 10:45:29 +0800 Subject: [PATCH 41/86] f2fs: fix to account compressed blocks in f2fs_compressed_blocks() por_fsstress reports inconsistent status in orphan inode, the root cause of this is in f2fs_write_raw_pages() we decrease i_compr_blocks incorrectly due to wrong calculation in f2fs_compressed_blocks(). So this patch exposes below two functions based on __f2fs_cluster_blocks: - f2fs_compressed_blocks: get count of compressed blocks in compressed cluster - f2fs_cluster_blocks: get count of valid blocks (including reserved blocks) in compressed cluster. Then use f2fs_compress_blocks() to get correct compressed blocks count in f2fs_write_raw_pages(). sanity_check_inode: inode (ino=ad80) hash inconsistent i_compr_blocks:2, i_blocks:1, run fsck to fix Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/compress.c | 28 ++++++++++++++++++++++------ 1 file changed, 22 insertions(+), 6 deletions(-) diff --git a/fs/f2fs/compress.c b/fs/f2fs/compress.c index bd3ea01db448..f9cf75fe4a12 100644 --- a/fs/f2fs/compress.c +++ b/fs/f2fs/compress.c @@ -535,8 +535,7 @@ static bool __cluster_may_compress(struct compress_ctx *cc) return true; } -/* return # of compressed block addresses */ -static int f2fs_compressed_blocks(struct compress_ctx *cc) +static int __f2fs_cluster_blocks(struct compress_ctx *cc, bool compr) { struct dnode_of_data dn; int ret; @@ -559,8 +558,13 @@ static int f2fs_compressed_blocks(struct compress_ctx *cc) blkaddr = data_blkaddr(dn.inode, dn.node_page, dn.ofs_in_node + i); - if (blkaddr != NULL_ADDR) - ret++; + if (compr) { + if (__is_valid_data_blkaddr(blkaddr)) + ret++; + } else { + if (blkaddr != NULL_ADDR) + ret++; + } } } fail: @@ -568,6 +572,18 @@ fail: return ret; } +/* return # of compressed blocks in compressed cluster */ +static int f2fs_compressed_blocks(struct compress_ctx *cc) +{ + return __f2fs_cluster_blocks(cc, true); +} + +/* return # of valid blocks in compressed cluster */ +static int f2fs_cluster_blocks(struct compress_ctx *cc, bool compr) +{ + return __f2fs_cluster_blocks(cc, false); +} + int f2fs_is_compressed_cluster(struct inode *inode, pgoff_t index) { struct compress_ctx cc = { @@ -577,7 +593,7 @@ int f2fs_is_compressed_cluster(struct inode *inode, pgoff_t index) .cluster_idx = index >> F2FS_I(inode)->i_log_cluster_size, }; - return f2fs_compressed_blocks(&cc); + return f2fs_cluster_blocks(&cc, false); } static bool cluster_may_compress(struct compress_ctx *cc) @@ -626,7 +642,7 @@ static int prepare_compress_overwrite(struct compress_ctx *cc, bool prealloc; retry: - ret = f2fs_compressed_blocks(cc); + ret = f2fs_cluster_blocks(cc, false); if (ret <= 0) return ret; From 7c6716c9a418378723244ddbeec25e6274bbfc0a Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Wed, 18 Mar 2020 19:40:45 +0800 Subject: [PATCH 42/86] f2fs: don't mark compressed inode dirty during f2fs_iget() - f2fs_iget - do_read_inode - set_inode_flag(, FI_COMPRESSED_FILE) - __mark_inode_dirty_flag(, true) It's unnecessary, so let's just mark compressed inode dirty while compressed inode conversion. Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/f2fs.h | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index 3e727a6af08e..e720054c9062 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -2573,7 +2573,6 @@ static inline void __mark_inode_dirty_flag(struct inode *inode, case FI_DATA_EXIST: case FI_INLINE_DOTS: case FI_PIN_FILE: - case FI_COMPRESSED_FILE: f2fs_mark_inode_dirty_sync(inode, true); } } @@ -3837,6 +3836,7 @@ static inline void set_compress_context(struct inode *inode) F2FS_I(inode)->i_flags |= F2FS_COMPR_FL; set_inode_flag(inode, FI_COMPRESSED_FILE); stat_inc_compr_inode(inode); + f2fs_mark_inode_dirty_sync(inode, true); } static inline u64 f2fs_disable_compressed_file(struct inode *inode) @@ -3853,6 +3853,7 @@ static inline u64 f2fs_disable_compressed_file(struct inode *inode) fi->i_flags &= ~F2FS_COMPR_FL; stat_dec_compr_inode(inode); clear_inode_flag(inode, FI_COMPRESSED_FILE); + f2fs_mark_inode_dirty_sync(inode, true); return 0; } From 6c4b72f0ffc2d7875e4d4b9490fad7d1c67857b5 Mon Sep 17 00:00:00 2001 From: DongDongJu Date: Fri, 20 Mar 2020 15:01:32 +0900 Subject: [PATCH 43/86] f2fs: delete DIO read lock This lock can be a contention with multi 4k random read IO with single inode. example) fio --output=test --name=test --numjobs=60 --filename=/media/samsung960pro/file_test --rw=randread --bs=4k --direct=1 --time_based --runtime=7 --ioengine=libaio --iodepth=256 --group_reporting --size=10G With this commit, it remove that possible lock contention. Signed-off-by: Dongjoo Seo Signed-off-by: Jaegeuk Kim --- fs/f2fs/data.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c index aa29b33f1537..4baf3a33f251 100644 --- a/fs/f2fs/data.c +++ b/fs/f2fs/data.c @@ -3453,7 +3453,8 @@ static ssize_t f2fs_direct_IO(struct kiocb *iocb, struct iov_iter *iter) err = __blockdev_direct_IO(iocb, inode, inode->i_sb->s_bdev, iter, rw == WRITE ? get_data_block_dio_write : get_data_block_dio, NULL, f2fs_dio_submit_bio, - DIO_LOCKING | DIO_SKIP_HOLES); + rw == WRITE ? DIO_LOCKING | DIO_SKIP_HOLES : + DIO_SKIP_HOLES); if (do_opu) up_read(&fi->i_gc_rwsem[READ]); From 1f6b30ae7fb4e99a121ccf786e5504fdc69f5511 Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Fri, 20 Mar 2020 18:14:31 +0800 Subject: [PATCH 44/86] f2fs: fix potential deadlock on compressed quota file generic/232 reports below deadlock: fsstress D 0 96980 96969 0x00084000 Call Trace: schedule+0x4a/0xb0 io_schedule+0x12/0x40 __lock_page+0x127/0x1d0 pagecache_get_page+0x1d8/0x250 prepare_compress_overwrite+0xe0/0x490 [f2fs] f2fs_prepare_compress_overwrite+0x5d/0x80 [f2fs] f2fs_write_begin+0x833/0xb90 [f2fs] f2fs_quota_write+0x145/0x1e0 [f2fs] write_blk+0x36/0x80 [quota_tree] do_insert_tree+0x2ac/0x4a0 [quota_tree] do_insert_tree+0x26e/0x4a0 [quota_tree] qtree_write_dquot+0x70/0x190 [quota_tree] v2_write_dquot+0x43/0x90 [quota_v2] dquot_acquire+0x77/0x100 f2fs_dquot_acquire+0x2f/0x60 [f2fs] dqget+0x310/0x450 dquot_transfer+0xb2/0x120 f2fs_setattr+0x11a/0x4a0 [f2fs] notify_change+0x349/0x480 chown_common+0x168/0x1c0 do_fchownat+0xbc/0xf0 __x64_sys_lchown+0x21/0x30 do_syscall_64+0x5f/0x220 entry_SYSCALL_64_after_hwframe+0x44/0xa9 task PC stack pid father kworker/u256:0 D 0 103444 2 0x80084000 Workqueue: writeback wb_workfn (flush-251:1) Call Trace: schedule+0x4a/0xb0 schedule_timeout+0x15e/0x2f0 io_schedule_timeout+0x19/0x40 congestion_wait+0x7e/0x120 f2fs_write_multi_pages+0x12a/0x840 [f2fs] f2fs_write_cache_pages+0x48f/0x790 [f2fs] f2fs_write_data_pages+0x2db/0x330 [f2fs] do_writepages+0x1a/0x60 __writeback_single_inode+0x3d/0x340 writeback_sb_inodes+0x225/0x4a0 wb_writeback+0xf7/0x320 wb_workfn+0xba/0x470 process_one_work+0x16c/0x3f0 worker_thread+0x4c/0x440 kthread+0xf8/0x130 ret_from_fork+0x35/0x40 fsstress D 0 5277 5266 0x00084000 Call Trace: schedule+0x4a/0xb0 rwsem_down_write_slowpath+0x29d/0x540 block_operations+0x105/0x360 [f2fs] f2fs_write_checkpoint+0x101/0x1010 [f2fs] f2fs_sync_fs+0xa8/0x130 [f2fs] f2fs_do_sync_file+0x1ad/0x890 [f2fs] do_fsync+0x38/0x60 __x64_sys_fdatasync+0x13/0x20 do_syscall_64+0x5f/0x220 entry_SYSCALL_64_after_hwframe+0x44/0xa9 The root cause is there is potential deadlock between quota data update and writeback. Kworker Thread B Thread C - f2fs_write_cache_pages - lock whole cluster --- A - f2fs_write_multi_pages - f2fs_write_raw_pages - f2fs_write_single_data_page - f2fs_do_write_data_page - f2fs_setattr - f2fs_lock_op --- B - f2fs_write_checkpoint - block_operations - f2fs_lock_all --- B - dquot_transfer - f2fs_quota_write - f2fs_prepare_compress_overwrite - pagecache_get_page --- A - f2fs_trylock_op failed --- B - congestion_wait - goto rewrite To fix this issue, during quota file writeback, just redirty all pages left in cluster rather holding pages' lock in cluster and looping retrying lock cp_rwsem. Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/compress.c | 15 ++++++++++----- 1 file changed, 10 insertions(+), 5 deletions(-) diff --git a/fs/f2fs/compress.c b/fs/f2fs/compress.c index f9cf75fe4a12..00fd6f904139 100644 --- a/fs/f2fs/compress.c +++ b/fs/f2fs/compress.c @@ -1002,6 +1002,15 @@ retry_write: unlock_page(cc->rpages[i]); ret = 0; } else if (ret == -EAGAIN) { + /* + * for quota file, just redirty left pages to + * avoid deadlock caused by cluster update race + * from foreground operation. + */ + if (IS_NOQUOTA(cc->inode)) { + err = 0; + goto out_err; + } ret = 0; cond_resched(); congestion_wait(BLK_RW_ASYNC, @@ -1011,16 +1020,12 @@ retry_write: goto retry_write; } err = ret; - goto out_fail; + goto out_err; } *submitted += _submitted; } return 0; - -out_fail: - /* TODO: revoke partially updated block addresses */ - BUG_ON(compr_blocks); out_err: for (++i; i < cc->cluster_size; i++) { if (!cc->rpages[i]) From 4e626be63ee1d439348b051a9fe6f15e526c9052 Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Fri, 20 Mar 2020 18:17:54 +0800 Subject: [PATCH 45/86] f2fs: don't change inode status under page lock In order to shrink page lock coverage. Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/dir.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/fs/f2fs/dir.c b/fs/f2fs/dir.c index 27d0dd7a16d6..0971ccc4664a 100644 --- a/fs/f2fs/dir.c +++ b/fs/f2fs/dir.c @@ -850,12 +850,6 @@ void f2fs_delete_entry(struct f2fs_dir_entry *dentry, struct page *page, 0); set_page_dirty(page); - dir->i_ctime = dir->i_mtime = current_time(dir); - f2fs_mark_inode_dirty_sync(dir, false); - - if (inode) - f2fs_drop_nlink(dir, inode); - if (bit_pos == NR_DENTRY_IN_BLOCK && !f2fs_truncate_hole(dir, page->index, page->index + 1)) { f2fs_clear_page_cache_dirty_tag(page); @@ -867,6 +861,12 @@ void f2fs_delete_entry(struct f2fs_dir_entry *dentry, struct page *page, f2fs_remove_dirty_inode(dir); } f2fs_put_page(page, 1); + + dir->i_ctime = dir->i_mtime = current_time(dir); + f2fs_mark_inode_dirty_sync(dir, false); + + if (inode) + f2fs_drop_nlink(dir, inode); } bool f2fs_empty_dir(struct inode *dir) From 42c5bf882252bb81cb90ee6ff73f4079b0fc7bab Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Thu, 19 Mar 2020 19:57:57 +0800 Subject: [PATCH 46/86] f2fs: fix to avoid potential deadlock We should always check F2FS_I(inode)->cp_task condition in prior to other conditions in __should_serialize_io() to avoid deadloop described in commit 040d2bb318d1 ("f2fs: fix to avoid deadloop if data_flush is on"), however we break this rule when we support compression, fix it. Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/data.c | 14 ++++++++------ 1 file changed, 8 insertions(+), 6 deletions(-) diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c index 4baf3a33f251..6a2f2427793a 100644 --- a/fs/f2fs/data.c +++ b/fs/f2fs/data.c @@ -2962,15 +2962,17 @@ next: static inline bool __should_serialize_io(struct inode *inode, struct writeback_control *wbc) { - if (!S_ISREG(inode->i_mode)) - return false; - if (f2fs_compressed_file(inode)) - return true; - if (IS_NOQUOTA(inode)) - return false; /* to avoid deadlock in path of data flush */ if (F2FS_I(inode)->cp_task) return false; + + if (!S_ISREG(inode->i_mode)) + return false; + if (IS_NOQUOTA(inode)) + return false; + + if (f2fs_compressed_file(inode)) + return true; if (wbc->sync_mode != WB_SYNC_ALL) return true; if (get_dirty_pages(inode) >= SM_I(F2FS_I_SB(inode))->min_seq_blocks) From d17c4e8fe7c83701152535832bbf7f4901fe43ae Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Sat, 21 Mar 2020 20:19:33 +0800 Subject: [PATCH 47/86] f2fs: clean up f2fs_may_encrypt() Merge below two conditions into f2fs_may_encrypt() for cleanup - IS_ENCRYPTED() - DUMMY_ENCRYPTION_ENABLED() Check IS_ENCRYPTED(inode) condition in f2fs_init_inode_metadata() is enough since we have already set encrypt flag in f2fs_new_inode(). Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/dir.c | 4 +--- fs/f2fs/f2fs.h | 13 +++++++++---- fs/f2fs/namei.c | 4 +--- 3 files changed, 11 insertions(+), 10 deletions(-) diff --git a/fs/f2fs/dir.c b/fs/f2fs/dir.c index 0971ccc4664a..44bfc464df78 100644 --- a/fs/f2fs/dir.c +++ b/fs/f2fs/dir.c @@ -471,7 +471,6 @@ struct page *f2fs_init_inode_metadata(struct inode *inode, struct inode *dir, struct page *dpage) { struct page *page; - int dummy_encrypt = DUMMY_ENCRYPTION_ENABLED(F2FS_I_SB(dir)); int err; if (is_inode_flag_set(inode, FI_NEW_INODE)) { @@ -498,8 +497,7 @@ struct page *f2fs_init_inode_metadata(struct inode *inode, struct inode *dir, if (err) goto put_error; - if ((IS_ENCRYPTED(dir) || dummy_encrypt) && - f2fs_may_encrypt(inode)) { + if (IS_ENCRYPTED(inode)) { err = fscrypt_inherit_context(dir, inode, page, false); if (err) goto put_error; diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index e720054c9062..47c8c763a9b5 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -3935,15 +3935,20 @@ static inline bool f2fs_lfs_mode(struct f2fs_sb_info *sbi) return F2FS_OPTION(sbi).fs_mode == FS_MODE_LFS; } -static inline bool f2fs_may_encrypt(struct inode *inode) +static inline bool f2fs_may_encrypt(struct inode *dir, struct inode *inode) { #ifdef CONFIG_FS_ENCRYPTION + struct f2fs_sb_info *sbi = F2FS_I_SB(dir); umode_t mode = inode->i_mode; - return (S_ISREG(mode) || S_ISDIR(mode) || S_ISLNK(mode)); -#else - return false; + /* + * If the directory encrypted or dummy encryption enabled, + * then we should encrypt the inode. + */ + if (IS_ENCRYPTED(dir) || DUMMY_ENCRYPTION_ENABLED(sbi)) + return (S_ISREG(mode) || S_ISDIR(mode) || S_ISLNK(mode)); #endif + return false; } static inline bool f2fs_may_compress(struct inode *inode) diff --git a/fs/f2fs/namei.c b/fs/f2fs/namei.c index b75c70813f9e..95cfbce062e8 100644 --- a/fs/f2fs/namei.c +++ b/fs/f2fs/namei.c @@ -75,9 +75,7 @@ static struct inode *f2fs_new_inode(struct inode *dir, umode_t mode) set_inode_flag(inode, FI_NEW_INODE); - /* If the directory encrypted, then we should encrypt the inode. */ - if ((IS_ENCRYPTED(dir) || DUMMY_ENCRYPTION_ENABLED(sbi)) && - f2fs_may_encrypt(inode)) + if (f2fs_may_encrypt(dir, inode)) f2fs_set_encrypted_inode(inode); if (f2fs_sb_has_extra_attr(sbi)) { From 73cfe4df2452a47032a3496e23c58023f67213f8 Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Thu, 19 Mar 2020 19:58:00 +0800 Subject: [PATCH 48/86] f2fs: fix NULL pointer dereference in f2fs_write_begin() BUG: kernel NULL pointer dereference, address: 0000000000000000 RIP: 0010:f2fs_write_begin+0x823/0xb90 [f2fs] Call Trace: f2fs_quota_write+0x139/0x1d0 [f2fs] write_blk+0x36/0x80 [quota_tree] get_free_dqblk+0x42/0xa0 [quota_tree] do_insert_tree+0x235/0x4a0 [quota_tree] do_insert_tree+0x26e/0x4a0 [quota_tree] do_insert_tree+0x26e/0x4a0 [quota_tree] do_insert_tree+0x26e/0x4a0 [quota_tree] qtree_write_dquot+0x70/0x190 [quota_tree] v2_write_dquot+0x43/0x90 [quota_v2] dquot_acquire+0x77/0x100 f2fs_dquot_acquire+0x2f/0x60 [f2fs] dqget+0x310/0x450 dquot_transfer+0x7e/0x120 f2fs_setattr+0x11a/0x4a0 [f2fs] notify_change+0x349/0x480 chown_common+0x168/0x1c0 do_fchownat+0xbc/0xf0 __x64_sys_fchownat+0x20/0x30 do_syscall_64+0x5f/0x220 entry_SYSCALL_64_after_hwframe+0x44/0xa9 Passing fsdata parameter to .write_{begin,end} in f2fs_quota_write(), so that if quota file is compressed one, we can avoid above NULL pointer dereference when updating quota content. Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/super.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c index c1f2c03a2e5a..204388b4cd55 100644 --- a/fs/f2fs/super.c +++ b/fs/f2fs/super.c @@ -1929,6 +1929,7 @@ static ssize_t f2fs_quota_write(struct super_block *sb, int type, int offset = off & (sb->s_blocksize - 1); size_t towrite = len; struct page *page; + void *fsdata = NULL; char *kaddr; int err = 0; int tocopy; @@ -1938,7 +1939,7 @@ static ssize_t f2fs_quota_write(struct super_block *sb, int type, towrite); retry: err = a_ops->write_begin(NULL, mapping, off, tocopy, 0, - &page, NULL); + &page, &fsdata); if (unlikely(err)) { if (err == -ENOMEM) { congestion_wait(BLK_RW_ASYNC, @@ -1955,7 +1956,7 @@ retry: flush_dcache_page(page); a_ops->write_end(NULL, mapping, off, tocopy, tocopy, - page, NULL); + page, fsdata); offset = 0; towrite -= tocopy; off += tocopy; From 4fc7fa5c4388166b7b87c3c2b8d27aab5561e19a Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Thu, 19 Mar 2020 19:57:58 +0800 Subject: [PATCH 49/86] f2fs: don't trigger data flush in foreground operation Data flush can generate heavy IO and cause long latency during flush, so it's not appropriate to trigger it in foreground operation. And also, we may face below potential deadlock during data flush: - f2fs_write_multi_pages - f2fs_write_raw_pages - f2fs_write_single_data_page - f2fs_balance_fs - f2fs_balance_fs_bg - f2fs_sync_dirty_inodes - filemap_fdatawrite -- stuck on flush same cluster Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/f2fs.h | 2 +- fs/f2fs/gc.c | 2 +- fs/f2fs/node.c | 2 +- fs/f2fs/segment.c | 6 +++--- 4 files changed, 6 insertions(+), 6 deletions(-) diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index 47c8c763a9b5..d84c53782c06 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -3236,7 +3236,7 @@ void f2fs_drop_inmem_pages(struct inode *inode); void f2fs_drop_inmem_page(struct inode *inode, struct page *page); int f2fs_commit_inmem_pages(struct inode *inode); void f2fs_balance_fs(struct f2fs_sb_info *sbi, bool need); -void f2fs_balance_fs_bg(struct f2fs_sb_info *sbi); +void f2fs_balance_fs_bg(struct f2fs_sb_info *sbi, bool from_bg); int f2fs_issue_flush(struct f2fs_sb_info *sbi, nid_t ino); int f2fs_create_flush_cmd_control(struct f2fs_sb_info *sbi); int f2fs_flush_device_cache(struct f2fs_sb_info *sbi); diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c index f122fe3dbba3..26248c8936db 100644 --- a/fs/f2fs/gc.c +++ b/fs/f2fs/gc.c @@ -113,7 +113,7 @@ do_gc: prefree_segments(sbi), free_segments(sbi)); /* balancing f2fs's metadata periodically */ - f2fs_balance_fs_bg(sbi); + f2fs_balance_fs_bg(sbi, true); next: sb_end_write(sbi->sb); diff --git a/fs/f2fs/node.c b/fs/f2fs/node.c index 07214e0dab26..c74b6da2ef05 100644 --- a/fs/f2fs/node.c +++ b/fs/f2fs/node.c @@ -1976,7 +1976,7 @@ static int f2fs_write_node_pages(struct address_space *mapping, goto skip_write; /* balancing f2fs's metadata in background */ - f2fs_balance_fs_bg(sbi); + f2fs_balance_fs_bg(sbi, true); /* collect a number of dirty node pages and write together */ if (wbc->sync_mode != WB_SYNC_ALL && diff --git a/fs/f2fs/segment.c b/fs/f2fs/segment.c index 601d67e72c50..aece09a184fa 100644 --- a/fs/f2fs/segment.c +++ b/fs/f2fs/segment.c @@ -496,7 +496,7 @@ void f2fs_balance_fs(struct f2fs_sb_info *sbi, bool need) /* balance_fs_bg is able to be pending */ if (need && excess_cached_nats(sbi)) - f2fs_balance_fs_bg(sbi); + f2fs_balance_fs_bg(sbi, false); if (!f2fs_is_checkpoint_ready(sbi)) return; @@ -511,7 +511,7 @@ void f2fs_balance_fs(struct f2fs_sb_info *sbi, bool need) } } -void f2fs_balance_fs_bg(struct f2fs_sb_info *sbi) +void f2fs_balance_fs_bg(struct f2fs_sb_info *sbi, bool from_bg) { if (unlikely(is_sbi_flag_set(sbi, SBI_POR_DOING))) return; @@ -540,7 +540,7 @@ void f2fs_balance_fs_bg(struct f2fs_sb_info *sbi) excess_dirty_nats(sbi) || excess_dirty_nodes(sbi) || f2fs_time_over(sbi, CP_TIME)) { - if (test_opt(sbi, DATA_FLUSH)) { + if (test_opt(sbi, DATA_FLUSH) && from_bg) { struct blk_plug plug; mutex_lock(&sbi->flush_lock); From 580c87b050f5a1946c6b633f1540d0dc57a0ffc4 Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Sat, 21 Mar 2020 20:23:27 +0800 Subject: [PATCH 50/86] f2fs: don't call fscrypt_get_encryption_info() explicitly in f2fs_tmpfile() In f2fs_tmpfile(), parent inode's encryption info is only used when inheriting encryption context to its child inode, however, we have already called fscrypt_get_encryption_info() in fscrypt_inherit_context() to get the encryption info, so just removing unneeded one in f2fs_tmpfile(). Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/namei.c | 6 ------ 1 file changed, 6 deletions(-) diff --git a/fs/f2fs/namei.c b/fs/f2fs/namei.c index 95cfbce062e8..f54119da2217 100644 --- a/fs/f2fs/namei.c +++ b/fs/f2fs/namei.c @@ -874,12 +874,6 @@ static int f2fs_tmpfile(struct inode *dir, struct dentry *dentry, umode_t mode) if (!f2fs_is_checkpoint_ready(sbi)) return -ENOSPC; - if (IS_ENCRYPTED(dir) || DUMMY_ENCRYPTION_ENABLED(sbi)) { - int err = fscrypt_get_encryption_info(dir); - if (err) - return err; - } - return __f2fs_tmpfile(dir, dentry, mode, NULL); } From c31bdd2c890684b5a8fab492bf6bf240730be161 Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Sat, 21 Mar 2020 20:24:11 +0800 Subject: [PATCH 51/86] f2fs: fix to clear PG_error if fsverity failed In f2fs_decompress_end_io(), we should clear PG_error flag before page unlock, otherwise reread will fail due to the flag as described in commit fb7d70db305a ("f2fs: clear PageError on the read path"). Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/compress.c | 18 ++++++++++-------- 1 file changed, 10 insertions(+), 8 deletions(-) diff --git a/fs/f2fs/compress.c b/fs/f2fs/compress.c index 00fd6f904139..d66d372d104d 100644 --- a/fs/f2fs/compress.c +++ b/fs/f2fs/compress.c @@ -1187,15 +1187,17 @@ void f2fs_decompress_end_io(struct page **rpages, if (!rpage) continue; - if (err || PageError(rpage)) { - ClearPageUptodate(rpage); - ClearPageError(rpage); - } else { - if (!verity || fsverity_verify_page(rpage)) - SetPageUptodate(rpage); - else - SetPageError(rpage); + if (err || PageError(rpage)) + goto clear_uptodate; + + if (!verity || fsverity_verify_page(rpage)) { + SetPageUptodate(rpage); + goto unlock; } +clear_uptodate: + ClearPageUptodate(rpage); + ClearPageError(rpage); +unlock: unlock_page(rpage); } } From dce539dcff5f450b07688992e51148b6da1b780d Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Mon, 23 Mar 2020 17:43:04 +0800 Subject: [PATCH 52/86] f2fs: fix NULL pointer dereference in f2fs_verity_work() If both compression and fsverity feature is on, generic/572 will report below NULL pointer dereference bug. BUG: kernel NULL pointer dereference, address: 0000000000000018 RIP: 0010:f2fs_verity_work+0x60/0x90 [f2fs] #PF: supervisor read access in kernel mode Workqueue: fsverity_read_queue f2fs_verity_work [f2fs] RIP: 0010:f2fs_verity_work+0x60/0x90 [f2fs] Call Trace: process_one_work+0x16c/0x3f0 worker_thread+0x4c/0x440 ? rescuer_thread+0x350/0x350 kthread+0xf8/0x130 ? kthread_unpark+0x70/0x70 ret_from_fork+0x35/0x40 There are two issue in f2fs_verity_work(): - it needs to traverse and verify all pages in bio. - if pages in bio belong to non-compressed cluster, accessing decompress IO context stored in page private will cause NULL pointer dereference. Fix them. Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/compress.c | 2 ++ fs/f2fs/data.c | 35 ++++++++++++++++++++++++++++++----- 2 files changed, 32 insertions(+), 5 deletions(-) diff --git a/fs/f2fs/compress.c b/fs/f2fs/compress.c index d66d372d104d..adac6fcb2830 100644 --- a/fs/f2fs/compress.c +++ b/fs/f2fs/compress.c @@ -477,6 +477,8 @@ out_vunmap_cbuf: out_vunmap_rbuf: vunmap(dic->rbuf); out_free_dic: + if (verity) + refcount_add(dic->nr_cpages - 1, &dic->ref); if (!verity) f2fs_decompress_end_io(dic->rpages, dic->cluster_size, ret, false); diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c index 6a2f2427793a..453cfee85cf6 100644 --- a/fs/f2fs/data.c +++ b/fs/f2fs/data.c @@ -187,12 +187,37 @@ static void f2fs_verify_pages(struct page **rpages, unsigned int cluster_size) static void f2fs_verify_bio(struct bio *bio) { - struct page *page = bio_first_page_all(bio); - struct decompress_io_ctx *dic = - (struct decompress_io_ctx *)page_private(page); + struct bio_vec *bv; + struct bvec_iter_all iter_all; - f2fs_verify_pages(dic->rpages, dic->cluster_size); - f2fs_free_dic(dic); + bio_for_each_segment_all(bv, bio, iter_all) { + struct page *page = bv->bv_page; + struct decompress_io_ctx *dic; + + dic = (struct decompress_io_ctx *)page_private(page); + + if (dic) { + if (refcount_dec_not_one(&dic->ref)) + continue; + f2fs_verify_pages(dic->rpages, + dic->cluster_size); + f2fs_free_dic(dic); + continue; + } + + if (bio->bi_status || PageError(page)) + goto clear_uptodate; + + if (fsverity_verify_page(page)) { + SetPageUptodate(page); + goto unlock; + } +clear_uptodate: + ClearPageUptodate(page); + ClearPageError(page); +unlock: + unlock_page(page); + } } #endif From 57cfbf77f25e6119b749cc528a056beb4cdb26ab Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Mon, 23 Mar 2020 11:18:07 +0800 Subject: [PATCH 53/86] f2fs: fix potential .flags overflow on 32bit architecture f2fs_inode_info.flags is unsigned long variable, it has 32 bits in 32bit architecture, since we introduced FI_MMAP_FILE flag when we support data compression, we may access memory cross the border of .flags field, corrupting .i_sem field, result in below deadlock. To fix this issue, let's expand .flags as an array to grab enough space to store new flags. Call Trace: __schedule+0x8d0/0x13fc ? mark_held_locks+0xac/0x100 schedule+0xcc/0x260 rwsem_down_write_slowpath+0x3ab/0x65d down_write+0xc7/0xe0 f2fs_drop_nlink+0x3d/0x600 [f2fs] f2fs_delete_inline_entry+0x300/0x440 [f2fs] f2fs_delete_entry+0x3a1/0x7f0 [f2fs] f2fs_unlink+0x500/0x790 [f2fs] vfs_unlink+0x211/0x490 do_unlinkat+0x483/0x520 sys_unlink+0x4a/0x70 do_fast_syscall_32+0x12b/0x683 entry_SYSENTER_32+0xaa/0x102 Fixes: 4c8ff7095bef ("f2fs: support data compression") Tested-by: Ondrej Jirman Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/f2fs.h | 99 ++++++++++++++++++++++++------------------------- fs/f2fs/inode.c | 2 +- 2 files changed, 50 insertions(+), 51 deletions(-) diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index d84c53782c06..dea45bb57862 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -678,6 +678,44 @@ enum { MAX_GC_FAILURE }; +/* used for f2fs_inode_info->flags */ +enum { + FI_NEW_INODE, /* indicate newly allocated inode */ + FI_DIRTY_INODE, /* indicate inode is dirty or not */ + FI_AUTO_RECOVER, /* indicate inode is recoverable */ + FI_DIRTY_DIR, /* indicate directory has dirty pages */ + FI_INC_LINK, /* need to increment i_nlink */ + FI_ACL_MODE, /* indicate acl mode */ + FI_NO_ALLOC, /* should not allocate any blocks */ + FI_FREE_NID, /* free allocated nide */ + FI_NO_EXTENT, /* not to use the extent cache */ + FI_INLINE_XATTR, /* used for inline xattr */ + FI_INLINE_DATA, /* used for inline data*/ + FI_INLINE_DENTRY, /* used for inline dentry */ + FI_APPEND_WRITE, /* inode has appended data */ + FI_UPDATE_WRITE, /* inode has in-place-update data */ + FI_NEED_IPU, /* used for ipu per file */ + FI_ATOMIC_FILE, /* indicate atomic file */ + FI_ATOMIC_COMMIT, /* indicate the state of atomical committing */ + FI_VOLATILE_FILE, /* indicate volatile file */ + FI_FIRST_BLOCK_WRITTEN, /* indicate #0 data block was written */ + FI_DROP_CACHE, /* drop dirty page cache */ + FI_DATA_EXIST, /* indicate data exists */ + FI_INLINE_DOTS, /* indicate inline dot dentries */ + FI_DO_DEFRAG, /* indicate defragment is running */ + FI_DIRTY_FILE, /* indicate regular/symlink has dirty pages */ + FI_NO_PREALLOC, /* indicate skipped preallocated blocks */ + FI_HOT_DATA, /* indicate file is hot */ + FI_EXTRA_ATTR, /* indicate file has extra attribute */ + FI_PROJ_INHERIT, /* indicate file inherits projectid */ + FI_PIN_FILE, /* indicate file should not be gced */ + FI_ATOMIC_REVOKE_REQUEST, /* request to drop atomic data */ + FI_VERITY_IN_PROGRESS, /* building fs-verity Merkle tree */ + FI_COMPRESSED_FILE, /* indicate file's data can be compressed */ + FI_MMAP_FILE, /* indicate file was mmapped */ + FI_MAX, /* max flag, never be used */ +}; + struct f2fs_inode_info { struct inode vfs_inode; /* serve a vfs inode */ unsigned long i_flags; /* keep an inode flags for ioctl */ @@ -690,7 +728,7 @@ struct f2fs_inode_info { umode_t i_acl_mode; /* keep file acl mode temporarily */ /* Use below internally in f2fs*/ - unsigned long flags; /* use to pass per-file flags */ + unsigned long flags[BITS_TO_LONGS(FI_MAX)]; /* use to pass per-file flags */ struct rw_semaphore i_sem; /* protect fi info */ atomic_t dirty_pages; /* # of dirty pages */ f2fs_hash_t chash; /* hash value of given file name */ @@ -2522,43 +2560,6 @@ static inline __u32 f2fs_mask_flags(umode_t mode, __u32 flags) return flags & F2FS_OTHER_FLMASK; } -/* used for f2fs_inode_info->flags */ -enum { - FI_NEW_INODE, /* indicate newly allocated inode */ - FI_DIRTY_INODE, /* indicate inode is dirty or not */ - FI_AUTO_RECOVER, /* indicate inode is recoverable */ - FI_DIRTY_DIR, /* indicate directory has dirty pages */ - FI_INC_LINK, /* need to increment i_nlink */ - FI_ACL_MODE, /* indicate acl mode */ - FI_NO_ALLOC, /* should not allocate any blocks */ - FI_FREE_NID, /* free allocated nide */ - FI_NO_EXTENT, /* not to use the extent cache */ - FI_INLINE_XATTR, /* used for inline xattr */ - FI_INLINE_DATA, /* used for inline data*/ - FI_INLINE_DENTRY, /* used for inline dentry */ - FI_APPEND_WRITE, /* inode has appended data */ - FI_UPDATE_WRITE, /* inode has in-place-update data */ - FI_NEED_IPU, /* used for ipu per file */ - FI_ATOMIC_FILE, /* indicate atomic file */ - FI_ATOMIC_COMMIT, /* indicate the state of atomical committing */ - FI_VOLATILE_FILE, /* indicate volatile file */ - FI_FIRST_BLOCK_WRITTEN, /* indicate #0 data block was written */ - FI_DROP_CACHE, /* drop dirty page cache */ - FI_DATA_EXIST, /* indicate data exists */ - FI_INLINE_DOTS, /* indicate inline dot dentries */ - FI_DO_DEFRAG, /* indicate defragment is running */ - FI_DIRTY_FILE, /* indicate regular/symlink has dirty pages */ - FI_NO_PREALLOC, /* indicate skipped preallocated blocks */ - FI_HOT_DATA, /* indicate file is hot */ - FI_EXTRA_ATTR, /* indicate file has extra attribute */ - FI_PROJ_INHERIT, /* indicate file inherits projectid */ - FI_PIN_FILE, /* indicate file should not be gced */ - FI_ATOMIC_REVOKE_REQUEST, /* request to drop atomic data */ - FI_VERITY_IN_PROGRESS, /* building fs-verity Merkle tree */ - FI_COMPRESSED_FILE, /* indicate file's data can be compressed */ - FI_MMAP_FILE, /* indicate file was mmapped */ -}; - static inline void __mark_inode_dirty_flag(struct inode *inode, int flag, bool set) { @@ -2579,20 +2580,18 @@ static inline void __mark_inode_dirty_flag(struct inode *inode, static inline void set_inode_flag(struct inode *inode, int flag) { - if (!test_bit(flag, &F2FS_I(inode)->flags)) - set_bit(flag, &F2FS_I(inode)->flags); + test_and_set_bit(flag, F2FS_I(inode)->flags); __mark_inode_dirty_flag(inode, flag, true); } static inline int is_inode_flag_set(struct inode *inode, int flag) { - return test_bit(flag, &F2FS_I(inode)->flags); + return test_bit(flag, F2FS_I(inode)->flags); } static inline void clear_inode_flag(struct inode *inode, int flag) { - if (test_bit(flag, &F2FS_I(inode)->flags)) - clear_bit(flag, &F2FS_I(inode)->flags); + test_and_clear_bit(flag, F2FS_I(inode)->flags); __mark_inode_dirty_flag(inode, flag, false); } @@ -2683,19 +2682,19 @@ static inline void get_inline_info(struct inode *inode, struct f2fs_inode *ri) struct f2fs_inode_info *fi = F2FS_I(inode); if (ri->i_inline & F2FS_INLINE_XATTR) - set_bit(FI_INLINE_XATTR, &fi->flags); + set_bit(FI_INLINE_XATTR, fi->flags); if (ri->i_inline & F2FS_INLINE_DATA) - set_bit(FI_INLINE_DATA, &fi->flags); + set_bit(FI_INLINE_DATA, fi->flags); if (ri->i_inline & F2FS_INLINE_DENTRY) - set_bit(FI_INLINE_DENTRY, &fi->flags); + set_bit(FI_INLINE_DENTRY, fi->flags); if (ri->i_inline & F2FS_DATA_EXIST) - set_bit(FI_DATA_EXIST, &fi->flags); + set_bit(FI_DATA_EXIST, fi->flags); if (ri->i_inline & F2FS_INLINE_DOTS) - set_bit(FI_INLINE_DOTS, &fi->flags); + set_bit(FI_INLINE_DOTS, fi->flags); if (ri->i_inline & F2FS_EXTRA_ATTR) - set_bit(FI_EXTRA_ATTR, &fi->flags); + set_bit(FI_EXTRA_ATTR, fi->flags); if (ri->i_inline & F2FS_PIN_FILE) - set_bit(FI_PIN_FILE, &fi->flags); + set_bit(FI_PIN_FILE, fi->flags); } static inline void set_raw_inline(struct inode *inode, struct f2fs_inode *ri) diff --git a/fs/f2fs/inode.c b/fs/f2fs/inode.c index 44e08bf2e2b4..44582a4db513 100644 --- a/fs/f2fs/inode.c +++ b/fs/f2fs/inode.c @@ -362,7 +362,7 @@ static int do_read_inode(struct inode *inode) fi->i_flags = le32_to_cpu(ri->i_flags); if (S_ISREG(inode->i_mode)) fi->i_flags &= ~F2FS_PROJINHERIT_FL; - fi->flags = 0; + bitmap_zero(fi->flags, FI_MAX); fi->i_advise = ri->i_advise; fi->i_pino = le32_to_cpu(ri->i_pino); fi->i_dir_level = ri->i_dir_level; From a1869cd66f554ca498ccbf23378e410f08ef2701 Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Tue, 24 Mar 2020 14:20:57 +0800 Subject: [PATCH 54/86] f2fs: fix to avoid double unlock On image that has verity and compression feature, if compressed pages and non-compressed pages are mixed in one bio, we may double unlock non-compressed page in below flow: - f2fs_post_read_work - f2fs_decompress_work - f2fs_decompress_bio - __read_end_io - unlock_page - fsverity_enqueue_verify_work - f2fs_verity_work - f2fs_verify_bio - unlock_page So it should skip handling non-compressed page in f2fs_decompress_work() if verity is on. Besides, add missing dec_page_count() in f2fs_verify_bio(). Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/data.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c index 453cfee85cf6..0f88caee8a31 100644 --- a/fs/f2fs/data.c +++ b/fs/f2fs/data.c @@ -139,6 +139,8 @@ static void __read_end_io(struct bio *bio, bool compr, bool verity) f2fs_decompress_pages(bio, page, verity); continue; } + if (verity) + continue; #endif /* PG_error was set if any post_read step failed */ @@ -216,6 +218,7 @@ clear_uptodate: ClearPageUptodate(page); ClearPageError(page); unlock: + dec_page_count(F2FS_P_SB(page), __read_io_type(page)); unlock_page(page); } } From 285d3d447d36009eb0e7eb9beb19ba4b36329796 Mon Sep 17 00:00:00 2001 From: YueHaibing Date: Sat, 28 Mar 2020 19:27:36 +0800 Subject: [PATCH 55/86] f2fs: xattr.h: Make stub helpers inline Fix gcc warnings: In file included from fs/f2fs/dir.c:15:0: fs/f2fs/xattr.h:157:13: warning: 'f2fs_destroy_xattr_caches' defined but not used [-Wunused-function] static void f2fs_destroy_xattr_caches(struct f2fs_sb_info *sbi) { } ^~~~~~~~~~~~~~~~~~~~~~~~~ fs/f2fs/xattr.h:156:12: warning: 'f2fs_init_xattr_caches' defined but not used [-Wunused-function] static int f2fs_init_xattr_caches(struct f2fs_sb_info *sbi) { return 0; } Reported-by: Hulk Robot Fixes: a999150f4fe3 ("f2fs: use kmem_cache pool during inline xattr lookups") Signed-off-by: YueHaibing Reviewed-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/xattr.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/fs/f2fs/xattr.h b/fs/f2fs/xattr.h index e471be77f8f0..938fcd20565d 100644 --- a/fs/f2fs/xattr.h +++ b/fs/f2fs/xattr.h @@ -153,8 +153,8 @@ static inline ssize_t f2fs_listxattr(struct dentry *dentry, char *buffer, { return -EOPNOTSUPP; } -static int f2fs_init_xattr_caches(struct f2fs_sb_info *sbi) { return 0; } -static void f2fs_destroy_xattr_caches(struct f2fs_sb_info *sbi) { } +static inline int f2fs_init_xattr_caches(struct f2fs_sb_info *sbi) { return 0; } +static inline void f2fs_destroy_xattr_caches(struct f2fs_sb_info *sbi) { } #endif #ifdef CONFIG_F2FS_FS_SECURITY From b4ae0bfb535e721d4a2459a8030aed5187df43d0 Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Mon, 30 Mar 2020 18:03:15 +0800 Subject: [PATCH 56/86] f2fs: fix to use f2fs_readpage_limit() in f2fs_read_multi_pages() Multipage read flow should consider fsverity, so it needs to use f2fs_readpage_limit() instead of i_size_read() to check EOF condition. Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/data.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c index 0f88caee8a31..3355239b317e 100644 --- a/fs/f2fs/data.c +++ b/fs/f2fs/data.c @@ -2048,7 +2048,8 @@ int f2fs_read_multi_pages(struct compress_ctx *cc, struct bio **bio_ret, f2fs_bug_on(sbi, f2fs_cluster_is_empty(cc)); - last_block_in_file = (i_size_read(inode) + blocksize - 1) >> blkbits; + last_block_in_file = (f2fs_readpage_limit(inode) + + blocksize - 1) >> blkbits; /* get rid of pages beyond EOF */ for (i = 0; i < cc->cluster_size; i++) { From 8182a99781cb7b62578cec6d745f63a15b0e041a Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Sat, 28 Mar 2020 17:33:23 +0800 Subject: [PATCH 57/86] f2fs: clean up {cic,dic}.ref handling {cic,dic}.ref should be initialized to number of compressed pages, let's initialize it directly rather than doing w/ f2fs_set_compressed_page(). Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/compress.c | 16 ++++++---------- 1 file changed, 6 insertions(+), 10 deletions(-) diff --git a/fs/f2fs/compress.c b/fs/f2fs/compress.c index adac6fcb2830..fde4af20c523 100644 --- a/fs/f2fs/compress.c +++ b/fs/f2fs/compress.c @@ -52,7 +52,7 @@ bool f2fs_is_compressed_page(struct page *page) } static void f2fs_set_compressed_page(struct page *page, - struct inode *inode, pgoff_t index, void *data, refcount_t *r) + struct inode *inode, pgoff_t index, void *data) { SetPagePrivate(page); set_page_private(page, (unsigned long)data); @@ -60,8 +60,6 @@ static void f2fs_set_compressed_page(struct page *page, /* i_crypto_info and iv index */ page->index = index; page->mapping = inode->i_mapping; - if (r) - refcount_inc(r); } static void f2fs_put_compressed_page(struct page *page) @@ -478,7 +476,7 @@ out_vunmap_rbuf: vunmap(dic->rbuf); out_free_dic: if (verity) - refcount_add(dic->nr_cpages - 1, &dic->ref); + refcount_set(&dic->ref, dic->nr_cpages); if (!verity) f2fs_decompress_end_io(dic->rpages, dic->cluster_size, ret, false); @@ -834,7 +832,7 @@ static int f2fs_write_compressed_pages(struct compress_ctx *cc, cic->magic = F2FS_COMPRESSED_PAGE_MAGIC; cic->inode = inode; - refcount_set(&cic->ref, 1); + refcount_set(&cic->ref, cc->nr_cpages); cic->rpages = f2fs_kzalloc(sbi, sizeof(struct page *) << cc->log_cluster_size, GFP_NOFS); if (!cic->rpages) @@ -844,8 +842,7 @@ static int f2fs_write_compressed_pages(struct compress_ctx *cc, for (i = 0; i < cc->nr_cpages; i++) { f2fs_set_compressed_page(cc->cpages[i], inode, - cc->rpages[i + 1]->index, - cic, i ? &cic->ref : NULL); + cc->rpages[i + 1]->index, cic); fio.compressed_page = cc->cpages[i]; if (fio.encrypted) { fio.page = cc->rpages[i + 1]; @@ -1095,7 +1092,7 @@ struct decompress_io_ctx *f2fs_alloc_dic(struct compress_ctx *cc) dic->magic = F2FS_COMPRESSED_PAGE_MAGIC; dic->inode = cc->inode; - refcount_set(&dic->ref, 1); + refcount_set(&dic->ref, cc->nr_cpages); dic->cluster_idx = cc->cluster_idx; dic->cluster_size = cc->cluster_size; dic->log_cluster_size = cc->log_cluster_size; @@ -1119,8 +1116,7 @@ struct decompress_io_ctx *f2fs_alloc_dic(struct compress_ctx *cc) goto out_free; f2fs_set_compressed_page(page, cc->inode, - start_idx + i + 1, - dic, i ? &dic->ref : NULL); + start_idx + i + 1, dic); dic->cpages[i] = page; } From fed59ca9bd8c4a497519eda4dbfe7abe13a59ab5 Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Tue, 10 Mar 2020 20:50:05 +0800 Subject: [PATCH 58/86] f2fs: change default compression algorithm Use LZ4 as default compression algorithm, as compared to LZO, it shows almost the same compression ratio and much better decompression speed. Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/super.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c index 204388b4cd55..8a3883c92163 100644 --- a/fs/f2fs/super.c +++ b/fs/f2fs/super.c @@ -1570,7 +1570,7 @@ static void default_options(struct f2fs_sb_info *sbi) F2FS_OPTION(sbi).test_dummy_encryption = false; F2FS_OPTION(sbi).s_resuid = make_kuid(&init_user_ns, F2FS_DEF_RESUID); F2FS_OPTION(sbi).s_resgid = make_kgid(&init_user_ns, F2FS_DEF_RESGID); - F2FS_OPTION(sbi).compress_algorithm = COMPRESS_LZO; + F2FS_OPTION(sbi).compress_algorithm = COMPRESS_LZ4; F2FS_OPTION(sbi).compress_log_size = MIN_COMPRESS_LOG_SIZE; F2FS_OPTION(sbi).compress_ext_cnt = 0; F2FS_OPTION(sbi).bggc_mode = BGGC_MODE_ON; From a39def5af858966b2cd3ed3f0b08ba4f7e1cea0e Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Tue, 3 Mar 2020 16:57:06 +0800 Subject: [PATCH 59/86] f2fs: compress: fix to call missing destroy_compress_ctx() Otherwise, it will cause memory leak. Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/compress.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/fs/f2fs/compress.c b/fs/f2fs/compress.c index fde4af20c523..1e5bbb9c1d3e 100644 --- a/fs/f2fs/compress.c +++ b/fs/f2fs/compress.c @@ -396,6 +396,8 @@ static int f2fs_compress_pages(struct compress_ctx *cc) cc->cpages[i] = NULL; } + cops->destroy_compress_ctx(cc); + cc->nr_cpages = nr_cpages; trace_f2fs_compress_pages_end(cc->inode, cc->cluster_idx, From 6fcaefbc94343b3bf65aff69528545ba167ed4fa Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Tue, 3 Mar 2020 16:57:07 +0800 Subject: [PATCH 60/86] f2fs: compress: add .{init,destroy}_decompress_ctx callback Add below two callback interfaces in struct f2fs_compress_ops: int (*init_decompress_ctx)(struct decompress_io_ctx *dic); void (*destroy_decompress_ctx)(struct decompress_io_ctx *dic); Which will be used by zstd compress algorithm later. In addition, this patch adds callback function pointer check, so that specified algorithm can avoid defining unneeded functions. Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/compress.c | 27 +++++++++++++++++++++------ 1 file changed, 21 insertions(+), 6 deletions(-) diff --git a/fs/f2fs/compress.c b/fs/f2fs/compress.c index 1e5bbb9c1d3e..13666735f567 100644 --- a/fs/f2fs/compress.c +++ b/fs/f2fs/compress.c @@ -20,6 +20,8 @@ struct f2fs_compress_ops { int (*init_compress_ctx)(struct compress_ctx *cc); void (*destroy_compress_ctx)(struct compress_ctx *cc); int (*compress_pages)(struct compress_ctx *cc); + int (*init_decompress_ctx)(struct decompress_io_ctx *dic); + void (*destroy_decompress_ctx)(struct decompress_io_ctx *dic); int (*decompress_pages)(struct decompress_io_ctx *dic); }; @@ -332,9 +334,11 @@ static int f2fs_compress_pages(struct compress_ctx *cc) trace_f2fs_compress_pages_start(cc->inode, cc->cluster_idx, cc->cluster_size, fi->i_compress_algorithm); - ret = cops->init_compress_ctx(cc); - if (ret) - goto out; + if (cops->init_compress_ctx) { + ret = cops->init_compress_ctx(cc); + if (ret) + goto out; + } max_len = COMPRESS_HEADER_SIZE + cc->clen; cc->nr_cpages = DIV_ROUND_UP(max_len, PAGE_SIZE); @@ -396,7 +400,8 @@ static int f2fs_compress_pages(struct compress_ctx *cc) cc->cpages[i] = NULL; } - cops->destroy_compress_ctx(cc); + if (cops->destroy_compress_ctx) + cops->destroy_compress_ctx(cc); cc->nr_cpages = nr_cpages; @@ -416,7 +421,8 @@ out_free_cpages: kfree(cc->cpages); cc->cpages = NULL; destroy_compress_ctx: - cops->destroy_compress_ctx(cc); + if (cops->destroy_compress_ctx) + cops->destroy_compress_ctx(cc); out: trace_f2fs_compress_pages_end(cc->inode, cc->cluster_idx, cc->clen, ret); @@ -450,10 +456,16 @@ void f2fs_decompress_pages(struct bio *bio, struct page *page, bool verity) goto out_free_dic; } + if (cops->init_decompress_ctx) { + ret = cops->init_decompress_ctx(dic); + if (ret) + goto out_free_dic; + } + dic->rbuf = vmap(dic->tpages, dic->cluster_size, VM_MAP, PAGE_KERNEL); if (!dic->rbuf) { ret = -ENOMEM; - goto out_free_dic; + goto destroy_decompress_ctx; } dic->cbuf = vmap(dic->cpages, dic->nr_cpages, VM_MAP, PAGE_KERNEL_RO); @@ -476,6 +488,9 @@ out_vunmap_cbuf: vunmap(dic->cbuf); out_vunmap_rbuf: vunmap(dic->rbuf); +destroy_decompress_ctx: + if (cops->destroy_decompress_ctx) + cops->destroy_decompress_ctx(dic); out_free_dic: if (verity) refcount_set(&dic->ref, dic->nr_cpages); From 6a881d4790ecb0b526712a358073a364ca67d95b Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Tue, 3 Mar 2020 17:46:02 +0800 Subject: [PATCH 61/86] f2fs: compress: support zstd compress algorithm Add zstd compress algorithm support, use "compress_algorithm=zstd" mountoption to enable it. Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- Documentation/filesystems/f2fs.txt | 4 +- fs/f2fs/Kconfig | 9 ++ fs/f2fs/compress.c | 165 +++++++++++++++++++++++++++++ fs/f2fs/f2fs.h | 5 + fs/f2fs/super.c | 7 ++ include/trace/events/f2fs.h | 3 +- 6 files changed, 190 insertions(+), 3 deletions(-) diff --git a/Documentation/filesystems/f2fs.txt b/Documentation/filesystems/f2fs.txt index 4eb3e2ddd00e..b1a66cf0e967 100644 --- a/Documentation/filesystems/f2fs.txt +++ b/Documentation/filesystems/f2fs.txt @@ -235,8 +235,8 @@ checkpoint=%s[:%u[%]] Set to "disable" to turn off checkpointing. Set to "en hide up to all remaining free space. The actual space that would be unusable can be viewed at /sys/fs/f2fs//unusable This space is reclaimed once checkpoint=enable. -compress_algorithm=%s Control compress algorithm, currently f2fs supports "lzo" - and "lz4" algorithm. +compress_algorithm=%s Control compress algorithm, currently f2fs supports "lzo", + "lz4" and "zstd" algorithm. compress_log_size=%u Support configuring compress cluster size, the size will be 4KB * (1 << %u), 16KB is minimum size, also it's default size. diff --git a/fs/f2fs/Kconfig b/fs/f2fs/Kconfig index f0faada30f30..bb68d21e1f8c 100644 --- a/fs/f2fs/Kconfig +++ b/fs/f2fs/Kconfig @@ -118,3 +118,12 @@ config F2FS_FS_LZ4 default y help Support LZ4 compress algorithm, if unsure, say Y. + +config F2FS_FS_ZSTD + bool "ZSTD compression support" + depends on F2FS_FS_COMPRESSION + select ZSTD_COMPRESS + select ZSTD_DECOMPRESS + default y + help + Support ZSTD compress algorithm, if unsure, say Y. diff --git a/fs/f2fs/compress.c b/fs/f2fs/compress.c index 13666735f567..c7284dd7d52f 100644 --- a/fs/f2fs/compress.c +++ b/fs/f2fs/compress.c @@ -11,6 +11,7 @@ #include #include #include +#include #include "f2fs.h" #include "node.h" @@ -291,6 +292,165 @@ static const struct f2fs_compress_ops f2fs_lz4_ops = { }; #endif +#ifdef CONFIG_F2FS_FS_ZSTD +#define F2FS_ZSTD_DEFAULT_CLEVEL 1 + +static int zstd_init_compress_ctx(struct compress_ctx *cc) +{ + ZSTD_parameters params; + ZSTD_CStream *stream; + void *workspace; + unsigned int workspace_size; + + params = ZSTD_getParams(F2FS_ZSTD_DEFAULT_CLEVEL, cc->rlen, 0); + workspace_size = ZSTD_CStreamWorkspaceBound(params.cParams); + + workspace = f2fs_kvmalloc(F2FS_I_SB(cc->inode), + workspace_size, GFP_NOFS); + if (!workspace) + return -ENOMEM; + + stream = ZSTD_initCStream(params, 0, workspace, workspace_size); + if (!stream) { + printk_ratelimited("%sF2FS-fs (%s): %s ZSTD_initCStream failed\n", + KERN_ERR, F2FS_I_SB(cc->inode)->sb->s_id, + __func__); + kvfree(workspace); + return -EIO; + } + + cc->private = workspace; + cc->private2 = stream; + + cc->clen = cc->rlen - PAGE_SIZE - COMPRESS_HEADER_SIZE; + return 0; +} + +static void zstd_destroy_compress_ctx(struct compress_ctx *cc) +{ + kvfree(cc->private); + cc->private = NULL; + cc->private2 = NULL; +} + +static int zstd_compress_pages(struct compress_ctx *cc) +{ + ZSTD_CStream *stream = cc->private2; + ZSTD_inBuffer inbuf; + ZSTD_outBuffer outbuf; + int src_size = cc->rlen; + int dst_size = src_size - PAGE_SIZE - COMPRESS_HEADER_SIZE; + int ret; + + inbuf.pos = 0; + inbuf.src = cc->rbuf; + inbuf.size = src_size; + + outbuf.pos = 0; + outbuf.dst = cc->cbuf->cdata; + outbuf.size = dst_size; + + ret = ZSTD_compressStream(stream, &outbuf, &inbuf); + if (ZSTD_isError(ret)) { + printk_ratelimited("%sF2FS-fs (%s): %s ZSTD_compressStream failed, ret: %d\n", + KERN_ERR, F2FS_I_SB(cc->inode)->sb->s_id, + __func__, ZSTD_getErrorCode(ret)); + return -EIO; + } + + ret = ZSTD_endStream(stream, &outbuf); + if (ZSTD_isError(ret)) { + printk_ratelimited("%sF2FS-fs (%s): %s ZSTD_endStream returned %d\n", + KERN_ERR, F2FS_I_SB(cc->inode)->sb->s_id, + __func__, ZSTD_getErrorCode(ret)); + return -EIO; + } + + cc->clen = outbuf.pos; + return 0; +} + +static int zstd_init_decompress_ctx(struct decompress_io_ctx *dic) +{ + ZSTD_DStream *stream; + void *workspace; + unsigned int workspace_size; + + workspace_size = ZSTD_DStreamWorkspaceBound(MAX_COMPRESS_WINDOW_SIZE); + + workspace = f2fs_kvmalloc(F2FS_I_SB(dic->inode), + workspace_size, GFP_NOFS); + if (!workspace) + return -ENOMEM; + + stream = ZSTD_initDStream(MAX_COMPRESS_WINDOW_SIZE, + workspace, workspace_size); + if (!stream) { + printk_ratelimited("%sF2FS-fs (%s): %s ZSTD_initDStream failed\n", + KERN_ERR, F2FS_I_SB(dic->inode)->sb->s_id, + __func__); + kvfree(workspace); + return -EIO; + } + + dic->private = workspace; + dic->private2 = stream; + + return 0; +} + +static void zstd_destroy_decompress_ctx(struct decompress_io_ctx *dic) +{ + kvfree(dic->private); + dic->private = NULL; + dic->private2 = NULL; +} + +static int zstd_decompress_pages(struct decompress_io_ctx *dic) +{ + ZSTD_DStream *stream = dic->private2; + ZSTD_inBuffer inbuf; + ZSTD_outBuffer outbuf; + int ret; + + inbuf.pos = 0; + inbuf.src = dic->cbuf->cdata; + inbuf.size = dic->clen; + + outbuf.pos = 0; + outbuf.dst = dic->rbuf; + outbuf.size = dic->rlen; + + ret = ZSTD_decompressStream(stream, &outbuf, &inbuf); + if (ZSTD_isError(ret)) { + printk_ratelimited("%sF2FS-fs (%s): %s ZSTD_compressStream failed, ret: %d\n", + KERN_ERR, F2FS_I_SB(dic->inode)->sb->s_id, + __func__, ZSTD_getErrorCode(ret)); + return -EIO; + } + + if (dic->rlen != outbuf.pos) { + printk_ratelimited("%sF2FS-fs (%s): %s ZSTD invalid rlen:%zu, " + "expected:%lu\n", KERN_ERR, + F2FS_I_SB(dic->inode)->sb->s_id, + __func__, dic->rlen, + PAGE_SIZE << dic->log_cluster_size); + return -EIO; + } + + return 0; +} + +static const struct f2fs_compress_ops f2fs_zstd_ops = { + .init_compress_ctx = zstd_init_compress_ctx, + .destroy_compress_ctx = zstd_destroy_compress_ctx, + .compress_pages = zstd_compress_pages, + .init_decompress_ctx = zstd_init_decompress_ctx, + .destroy_decompress_ctx = zstd_destroy_decompress_ctx, + .decompress_pages = zstd_decompress_pages, +}; +#endif + static const struct f2fs_compress_ops *f2fs_cops[COMPRESS_MAX] = { #ifdef CONFIG_F2FS_FS_LZO &f2fs_lzo_ops, @@ -302,6 +462,11 @@ static const struct f2fs_compress_ops *f2fs_cops[COMPRESS_MAX] = { #else NULL, #endif +#ifdef CONFIG_F2FS_FS_ZSTD + &f2fs_zstd_ops, +#else + NULL, +#endif }; bool f2fs_is_compress_backend_ready(struct inode *inode) diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index dea45bb57862..8d206dcc33ab 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -1267,6 +1267,7 @@ enum fsync_mode { enum compress_algorithm_type { COMPRESS_LZO, COMPRESS_LZ4, + COMPRESS_ZSTD, COMPRESS_MAX, }; @@ -1296,6 +1297,7 @@ struct compress_ctx { size_t rlen; /* valid data length in rbuf */ size_t clen; /* valid data length in cbuf */ void *private; /* payload buffer for specified compression algorithm */ + void *private2; /* extra payload buffer */ }; /* compress context for write IO path */ @@ -1325,11 +1327,14 @@ struct decompress_io_ctx { size_t clen; /* valid data length in cbuf */ refcount_t ref; /* referrence count of compressed page */ bool failed; /* indicate IO error during decompression */ + void *private; /* payload buffer for specified decompression algorithm */ + void *private2; /* extra payload buffer */ }; #define NULL_CLUSTER ((unsigned int)(~0)) #define MIN_COMPRESS_LOG_SIZE 2 #define MAX_COMPRESS_LOG_SIZE 8 +#define MAX_COMPRESS_WINDOW_SIZE ((PAGE_SIZE) << MAX_COMPRESS_LOG_SIZE) struct f2fs_sb_info { struct super_block *sb; /* pointer to VFS super block */ diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c index 8a3883c92163..dcf506e3cc32 100644 --- a/fs/f2fs/super.c +++ b/fs/f2fs/super.c @@ -829,6 +829,10 @@ static int parse_options(struct super_block *sb, char *options) !strcmp(name, "lz4")) { F2FS_OPTION(sbi).compress_algorithm = COMPRESS_LZ4; + } else if (strlen(name) == 4 && + !strcmp(name, "zstd")) { + F2FS_OPTION(sbi).compress_algorithm = + COMPRESS_ZSTD; } else { kfree(name); return -EINVAL; @@ -1419,6 +1423,9 @@ static inline void f2fs_show_compress_options(struct seq_file *seq, case COMPRESS_LZ4: algtype = "lz4"; break; + case COMPRESS_ZSTD: + algtype = "zstd"; + break; } seq_printf(seq, ",compress_algorithm=%s", algtype); diff --git a/include/trace/events/f2fs.h b/include/trace/events/f2fs.h index e3feba13e437..d4647dc49275 100644 --- a/include/trace/events/f2fs.h +++ b/include/trace/events/f2fs.h @@ -153,7 +153,8 @@ TRACE_DEFINE_ENUM(CP_PAUSE); #define show_compress_algorithm(type) \ __print_symbolic(type, \ { COMPRESS_LZO, "LZO" }, \ - { COMPRESS_LZ4, "LZ4" }) + { COMPRESS_LZ4, "LZ4" }, \ + { COMPRESS_ZSTD, "ZSTD" }) struct f2fs_sb_info; struct f2fs_io_info; From cd3a848cd102e08f262bd60f1d4b47b245805791 Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Wed, 25 Mar 2020 17:25:07 +0800 Subject: [PATCH 62/86] f2fs: clean up dic->tpages assignment Just cleanup, no logic change. Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/compress.c | 10 +++------- 1 file changed, 3 insertions(+), 7 deletions(-) diff --git a/fs/f2fs/compress.c b/fs/f2fs/compress.c index c7284dd7d52f..ca4f54fbbd2e 100644 --- a/fs/f2fs/compress.c +++ b/fs/f2fs/compress.c @@ -1308,20 +1308,16 @@ struct decompress_io_ctx *f2fs_alloc_dic(struct compress_ctx *cc) goto out_free; for (i = 0; i < dic->cluster_size; i++) { - if (cc->rpages[i]) + if (cc->rpages[i]) { + dic->tpages[i] = cc->rpages[i]; continue; + } dic->tpages[i] = f2fs_grab_page(); if (!dic->tpages[i]) goto out_free; } - for (i = 0; i < dic->cluster_size; i++) { - if (dic->tpages[i]) - continue; - dic->tpages[i] = cc->rpages[i]; - } - return dic; out_free: From a73ede190590f14168fd5cf3f35affb485beba3b Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Wed, 25 Mar 2020 10:22:09 +0800 Subject: [PATCH 63/86] f2fs: show compression in statx fstest reports below message when compression is on: generic/424 1s ... - output mismatch --- tests/generic/424.out +++ results/generic/424.out.bad @@ -1,2 +1,26 @@ QA output created by 424 +[!] Attribute compressed should be set +Failed +stat_test failed +[!] Attribute compressed should be set +Failed +stat_test failed We missed to set STATX_ATTR_COMPRESSED on compressed inode in getattr(), fix it. Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/file.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c index f0152181f3e8..adda8b8d7095 100644 --- a/fs/f2fs/file.c +++ b/fs/f2fs/file.c @@ -799,6 +799,8 @@ int f2fs_getattr(const struct path *path, struct kstat *stat, } flags = fi->i_flags; + if (flags & F2FS_COMPR_FL) + stat->attributes |= STATX_ATTR_COMPRESSED; if (flags & F2FS_APPEND_FL) stat->attributes |= STATX_ATTR_APPEND; if (IS_ENCRYPTED(inode)) @@ -810,7 +812,8 @@ int f2fs_getattr(const struct path *path, struct kstat *stat, if (IS_VERITY(inode)) stat->attributes |= STATX_ATTR_VERITY; - stat->attributes_mask |= (STATX_ATTR_APPEND | + stat->attributes_mask |= (STATX_ATTR_COMPRESSED | + STATX_ATTR_APPEND | STATX_ATTR_ENCRYPTED | STATX_ATTR_IMMUTABLE | STATX_ATTR_NODUMP | From c27eb0091f157abeb0e63bffcb9dd610a62f2207 Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Thu, 26 Mar 2020 17:42:26 +0800 Subject: [PATCH 64/86] f2fs: fix to verify tpage before releasing in f2fs_free_dic() In below error path, tpages[i] could be NULL, fix to check it before releasing it. - f2fs_read_multi_pages - f2fs_alloc_dic - f2fs_free_dic Fixes: 61fbae2b2b12 ("f2fs: fix to avoid NULL pointer dereference") Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/compress.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/fs/f2fs/compress.c b/fs/f2fs/compress.c index ca4f54fbbd2e..df7b2d15eacd 100644 --- a/fs/f2fs/compress.c +++ b/fs/f2fs/compress.c @@ -1333,6 +1333,8 @@ void f2fs_free_dic(struct decompress_io_ctx *dic) for (i = 0; i < dic->cluster_size; i++) { if (dic->rpages[i]) continue; + if (!dic->tpages[i]) + continue; unlock_page(dic->tpages[i]); put_page(dic->tpages[i]); } From 6df9eeba23ab975786baadb354cb3c736f93b0a0 Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Thu, 26 Mar 2020 17:43:56 +0800 Subject: [PATCH 65/86] f2fs: switch discard_policy.timeout to bool type While checking discard timeout, we use specified type UMOUNT_DISCARD_TIMEOUT, so just replace doplicy.timeout with it, and switch doplicy.timeout to bool type. Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/f2fs.h | 2 +- fs/f2fs/segment.c | 16 ++++++++-------- 2 files changed, 9 insertions(+), 9 deletions(-) diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index 8d206dcc33ab..c44afc598638 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -330,8 +330,8 @@ struct discard_policy { bool io_aware; /* issue discard in idle time */ bool sync; /* submit discard with REQ_SYNC flag */ bool ordered; /* issue discard by lba order */ + bool timeout; /* discard timeout for put_super */ unsigned int granularity; /* discard granularity */ - int timeout; /* discard timeout for put_super */ }; struct discard_cmd_control { diff --git a/fs/f2fs/segment.c b/fs/f2fs/segment.c index aece09a184fa..b7a9421472a7 100644 --- a/fs/f2fs/segment.c +++ b/fs/f2fs/segment.c @@ -1080,7 +1080,7 @@ static void __init_discard_policy(struct f2fs_sb_info *sbi, dpolicy->max_requests = DEF_MAX_DISCARD_REQUEST; dpolicy->io_aware_gran = MAX_PLIST_NUM; - dpolicy->timeout = 0; + dpolicy->timeout = false; if (discard_type == DPOLICY_BG) { dpolicy->min_interval = DEF_MIN_DISCARD_ISSUE_TIME; @@ -1105,6 +1105,7 @@ static void __init_discard_policy(struct f2fs_sb_info *sbi, dpolicy->io_aware = false; /* we need to issue all to keep CP_TRIMMED_FLAG */ dpolicy->granularity = 1; + dpolicy->timeout = true; } } @@ -1473,12 +1474,12 @@ static int __issue_discard_cmd(struct f2fs_sb_info *sbi, int i, issued = 0; bool io_interrupted = false; - if (dpolicy->timeout != 0) - f2fs_update_time(sbi, dpolicy->timeout); + if (dpolicy->timeout) + f2fs_update_time(sbi, UMOUNT_DISCARD_TIMEOUT); for (i = MAX_PLIST_NUM - 1; i >= 0; i--) { - if (dpolicy->timeout != 0 && - f2fs_time_over(sbi, dpolicy->timeout)) + if (dpolicy->timeout && + f2fs_time_over(sbi, UMOUNT_DISCARD_TIMEOUT)) break; if (i + 1 < dpolicy->granularity) @@ -1499,8 +1500,8 @@ static int __issue_discard_cmd(struct f2fs_sb_info *sbi, list_for_each_entry_safe(dc, tmp, pend_list, list) { f2fs_bug_on(sbi, dc->state != D_PREP); - if (dpolicy->timeout != 0 && - f2fs_time_over(sbi, dpolicy->timeout)) + if (dpolicy->timeout && + f2fs_time_over(sbi, UMOUNT_DISCARD_TIMEOUT)) break; if (dpolicy->io_aware && i < dpolicy->io_aware_gran && @@ -1679,7 +1680,6 @@ bool f2fs_issue_discard_timeout(struct f2fs_sb_info *sbi) __init_discard_policy(sbi, &dpolicy, DPOLICY_UMOUNT, dcc->discard_granularity); - dpolicy.timeout = UMOUNT_DISCARD_TIMEOUT; __issue_discard_cmd(sbi, &dpolicy); dropped = __drop_discard_cmd(sbi); From 61ab7d87815eee9ebb79534dac61c2e12e162809 Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Fri, 27 Mar 2020 18:29:00 +0800 Subject: [PATCH 66/86] f2fs: add missing CONFIG_F2FS_FS_COMPRESSION Compression sysfs node should not be shown if f2fs module disables compression feature. Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/sysfs.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/fs/f2fs/sysfs.c b/fs/f2fs/sysfs.c index 5f4fec83a6c6..e3bbbef9b4f0 100644 --- a/fs/f2fs/sysfs.c +++ b/fs/f2fs/sysfs.c @@ -580,7 +580,9 @@ F2FS_FEATURE_RO_ATTR(verity, FEAT_VERITY); #endif F2FS_FEATURE_RO_ATTR(sb_checksum, FEAT_SB_CHECKSUM); F2FS_FEATURE_RO_ATTR(casefold, FEAT_CASEFOLD); +#ifdef CONFIG_F2FS_FS_COMPRESSION F2FS_FEATURE_RO_ATTR(compression, FEAT_COMPRESSION); +#endif #define ATTR_LIST(name) (&f2fs_attr_##name.attr) static struct attribute *f2fs_attrs[] = { @@ -662,7 +664,9 @@ static struct attribute *f2fs_feat_attrs[] = { #endif ATTR_LIST(sb_checksum), ATTR_LIST(casefold), +#ifdef CONFIG_F2FS_FS_COMPRESSION ATTR_LIST(compression), +#endif NULL, }; ATTRIBUTE_GROUPS(f2fs_feat); From 196176207e0a7fb87a45cc9fc2dbd43583c5c48f Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Fri, 27 Mar 2020 18:29:51 +0800 Subject: [PATCH 67/86] f2fs: fix to disable compression on directory It needs to call f2fs_disable_compressed_file() to disable compression on directory. Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/f2fs.h | 10 ++++++---- fs/f2fs/file.c | 2 +- 2 files changed, 7 insertions(+), 5 deletions(-) diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index c44afc598638..c84442eefc56 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -3849,10 +3849,12 @@ static inline u64 f2fs_disable_compressed_file(struct inode *inode) if (!f2fs_compressed_file(inode)) return 0; - if (get_dirty_pages(inode)) - return 1; - if (fi->i_compr_blocks) - return fi->i_compr_blocks; + if (S_ISREG(inode->i_mode)) { + if (get_dirty_pages(inode)) + return 1; + if (fi->i_compr_blocks) + return fi->i_compr_blocks; + } fi->i_flags &= ~F2FS_COMPR_FL; stat_dec_compr_inode(inode); diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c index adda8b8d7095..7ee14452bc3b 100644 --- a/fs/f2fs/file.c +++ b/fs/f2fs/file.c @@ -1821,7 +1821,7 @@ static int f2fs_setflags_common(struct inode *inode, u32 iflags, u32 mask) } if ((iflags ^ masked_flags) & F2FS_COMPR_FL) { - if (S_ISREG(inode->i_mode) && (masked_flags & F2FS_COMPR_FL)) { + if (masked_flags & F2FS_COMPR_FL) { if (f2fs_disable_compressed_file(inode)) return -EINVAL; } From e8a1b83895aa9fde8c709fba4dfe07c3f3ace759 Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Fri, 27 Mar 2020 18:29:52 +0800 Subject: [PATCH 68/86] f2fs: keep inline_data when compression conversion We can keep compressed inode's data inline before inline conversion. Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/file.c | 5 ----- 1 file changed, 5 deletions(-) diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c index 7ee14452bc3b..eb869bbe5cd5 100644 --- a/fs/f2fs/file.c +++ b/fs/f2fs/file.c @@ -1828,11 +1828,6 @@ static int f2fs_setflags_common(struct inode *inode, u32 iflags, u32 mask) if (iflags & F2FS_NOCOMP_FL) return -EINVAL; if (iflags & F2FS_COMPR_FL) { - int err = f2fs_convert_inline_inode(inode); - - if (err) - return err; - if (!f2fs_may_compress(inode)) return -EINVAL; From cc53008ea6dbbd3ba062687d9e464321594e7b52 Mon Sep 17 00:00:00 2001 From: Eric Biggers Date: Sat, 14 Mar 2020 13:50:49 -0700 Subject: [PATCH 69/86] fscrypt: add FS_IOC_GET_ENCRYPTION_NONCE ioctl Add an ioctl FS_IOC_GET_ENCRYPTION_NONCE which retrieves the nonce from an encrypted file or directory. The nonce is the 16-byte random value stored in the inode's encryption xattr. It is normally used together with the master key to derive the inode's actual encryption key. The nonces are needed by automated tests that verify the correctness of the ciphertext on-disk. Except for the IV_INO_LBLK_64 case, there's no way to replicate a file's ciphertext without knowing that file's nonce. The nonces aren't secret, and the existing ciphertext verification tests in xfstests retrieve them from disk using debugfs or dump.f2fs. But in environments that lack these debugging tools, getting the nonces by manually parsing the filesystem structure would be very hard. To make this important type of testing much easier, let's just add an ioctl that retrieves the nonce. Link: https://lore.kernel.org/r/20200314205052.93294-2-ebiggers@kernel.org Reviewed-by: Theodore Ts'o Signed-off-by: Eric Biggers --- Documentation/filesystems/fscrypt.rst | 11 +++++++++++ fs/crypto/fscrypt_private.h | 20 ++++++++++++++++++++ fs/crypto/keysetup.c | 16 ++-------------- fs/crypto/policy.c | 21 ++++++++++++++++++++- include/linux/fscrypt.h | 6 ++++++ include/uapi/linux/fscrypt.h | 1 + 6 files changed, 60 insertions(+), 15 deletions(-) diff --git a/Documentation/filesystems/fscrypt.rst b/Documentation/filesystems/fscrypt.rst index fbcd185d15da..dc444b8d3704 100644 --- a/Documentation/filesystems/fscrypt.rst +++ b/Documentation/filesystems/fscrypt.rst @@ -633,6 +633,17 @@ from a passphrase or other low-entropy user credential. FS_IOC_GET_ENCRYPTION_PWSALT is deprecated. Instead, prefer to generate and manage any needed salt(s) in userspace. +Getting a file's encryption nonce +--------------------------------- + +Since Linux v5.7, the ioctl FS_IOC_GET_ENCRYPTION_NONCE is supported. +On encrypted files and directories it gets the inode's 16-byte nonce. +On unencrypted files and directories, it fails with ENODATA. + +This ioctl can be useful for automated tests which verify that the +encryption is being done correctly. It is not needed for normal use +of fscrypt. + Adding keys ----------- diff --git a/fs/crypto/fscrypt_private.h b/fs/crypto/fscrypt_private.h index 9aae851409e5..dbced2937ec8 100644 --- a/fs/crypto/fscrypt_private.h +++ b/fs/crypto/fscrypt_private.h @@ -76,6 +76,26 @@ static inline int fscrypt_context_size(const union fscrypt_context *ctx) return 0; } +/* Check whether an fscrypt_context has a recognized version number and size */ +static inline bool fscrypt_context_is_valid(const union fscrypt_context *ctx, + int ctx_size) +{ + return ctx_size >= 1 && ctx_size == fscrypt_context_size(ctx); +} + +/* Retrieve the context's nonce, assuming the context was already validated */ +static inline const u8 *fscrypt_context_nonce(const union fscrypt_context *ctx) +{ + switch (ctx->version) { + case FSCRYPT_CONTEXT_V1: + return ctx->v1.nonce; + case FSCRYPT_CONTEXT_V2: + return ctx->v2.nonce; + } + WARN_ON(1); + return NULL; +} + #undef fscrypt_policy union fscrypt_policy { u8 version; diff --git a/fs/crypto/keysetup.c b/fs/crypto/keysetup.c index 08c9f216a54d..302375e9f719 100644 --- a/fs/crypto/keysetup.c +++ b/fs/crypto/keysetup.c @@ -425,20 +425,8 @@ int fscrypt_get_encryption_info(struct inode *inode) goto out; } - switch (ctx.version) { - case FSCRYPT_CONTEXT_V1: - memcpy(crypt_info->ci_nonce, ctx.v1.nonce, - FS_KEY_DERIVATION_NONCE_SIZE); - break; - case FSCRYPT_CONTEXT_V2: - memcpy(crypt_info->ci_nonce, ctx.v2.nonce, - FS_KEY_DERIVATION_NONCE_SIZE); - break; - default: - WARN_ON(1); - res = -EINVAL; - goto out; - } + memcpy(crypt_info->ci_nonce, fscrypt_context_nonce(&ctx), + FS_KEY_DERIVATION_NONCE_SIZE); if (!fscrypt_supported_policy(&crypt_info->ci_policy, inode)) { res = -EINVAL; diff --git a/fs/crypto/policy.c b/fs/crypto/policy.c index cf2a9d26ef7d..10ccf945020c 100644 --- a/fs/crypto/policy.c +++ b/fs/crypto/policy.c @@ -258,7 +258,7 @@ int fscrypt_policy_from_context(union fscrypt_policy *policy_u, { memset(policy_u, 0, sizeof(*policy_u)); - if (ctx_size <= 0 || ctx_size != fscrypt_context_size(ctx_u)) + if (!fscrypt_context_is_valid(ctx_u, ctx_size)) return -EINVAL; switch (ctx_u->version) { @@ -481,6 +481,25 @@ int fscrypt_ioctl_get_policy_ex(struct file *filp, void __user *uarg) } EXPORT_SYMBOL_GPL(fscrypt_ioctl_get_policy_ex); +/* FS_IOC_GET_ENCRYPTION_NONCE: retrieve file's encryption nonce for testing */ +int fscrypt_ioctl_get_nonce(struct file *filp, void __user *arg) +{ + struct inode *inode = file_inode(filp); + union fscrypt_context ctx; + int ret; + + ret = inode->i_sb->s_cop->get_context(inode, &ctx, sizeof(ctx)); + if (ret < 0) + return ret; + if (!fscrypt_context_is_valid(&ctx, ret)) + return -EINVAL; + if (copy_to_user(arg, fscrypt_context_nonce(&ctx), + FS_KEY_DERIVATION_NONCE_SIZE)) + return -EFAULT; + return 0; +} +EXPORT_SYMBOL_GPL(fscrypt_ioctl_get_nonce); + /** * fscrypt_has_permitted_context() - is a file's encryption policy permitted * within its directory? diff --git a/include/linux/fscrypt.h b/include/linux/fscrypt.h index 556f4adf5dc5..e3c2d2a15525 100644 --- a/include/linux/fscrypt.h +++ b/include/linux/fscrypt.h @@ -139,6 +139,7 @@ extern void fscrypt_free_bounce_page(struct page *bounce_page); extern int fscrypt_ioctl_set_policy(struct file *, const void __user *); extern int fscrypt_ioctl_get_policy(struct file *, void __user *); extern int fscrypt_ioctl_get_policy_ex(struct file *, void __user *); +extern int fscrypt_ioctl_get_nonce(struct file *filp, void __user *arg); extern int fscrypt_has_permitted_context(struct inode *, struct inode *); extern int fscrypt_inherit_context(struct inode *, struct inode *, void *, bool); @@ -300,6 +301,11 @@ static inline int fscrypt_ioctl_get_policy_ex(struct file *filp, return -EOPNOTSUPP; } +static inline int fscrypt_ioctl_get_nonce(struct file *filp, void __user *arg) +{ + return -EOPNOTSUPP; +} + static inline int fscrypt_has_permitted_context(struct inode *parent, struct inode *child) { diff --git a/include/uapi/linux/fscrypt.h b/include/uapi/linux/fscrypt.h index 0d8a6f47711c..a10e3cdc2839 100644 --- a/include/uapi/linux/fscrypt.h +++ b/include/uapi/linux/fscrypt.h @@ -163,6 +163,7 @@ struct fscrypt_get_key_status_arg { #define FS_IOC_REMOVE_ENCRYPTION_KEY _IOWR('f', 24, struct fscrypt_remove_key_arg) #define FS_IOC_REMOVE_ENCRYPTION_KEY_ALL_USERS _IOWR('f', 25, struct fscrypt_remove_key_arg) #define FS_IOC_GET_ENCRYPTION_KEY_STATUS _IOWR('f', 26, struct fscrypt_get_key_status_arg) +#define FS_IOC_GET_ENCRYPTION_NONCE _IOR('f', 27, __u8[16]) /**********************************************************************/ From 93bc1cef7bd405bb8f84782ac6f46723de6945f6 Mon Sep 17 00:00:00 2001 From: Eric Biggers Date: Sat, 14 Mar 2020 13:50:50 -0700 Subject: [PATCH 70/86] ext4: wire up FS_IOC_GET_ENCRYPTION_NONCE This new ioctl retrieves a file's encryption nonce, which is useful for testing. See the corresponding fs/crypto/ patch for more details. Link: https://lore.kernel.org/r/20200314205052.93294-3-ebiggers@kernel.org Reviewed-by: Theodore Ts'o Signed-off-by: Eric Biggers --- fs/ext4/ioctl.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/fs/ext4/ioctl.c b/fs/ext4/ioctl.c index 0b7f316fd30f..1bf6811a263d 100644 --- a/fs/ext4/ioctl.c +++ b/fs/ext4/ioctl.c @@ -1210,6 +1210,11 @@ resizefs_out: return -EOPNOTSUPP; return fscrypt_ioctl_get_key_status(filp, (void __user *)arg); + case FS_IOC_GET_ENCRYPTION_NONCE: + if (!ext4_has_feature_encrypt(sb)) + return -EOPNOTSUPP; + return fscrypt_ioctl_get_nonce(filp, (void __user *)arg); + case EXT4_IOC_CLEAR_ES_CACHE: { if (!inode_owner_or_capable(inode)) @@ -1369,6 +1374,7 @@ long ext4_compat_ioctl(struct file *file, unsigned int cmd, unsigned long arg) case FS_IOC_REMOVE_ENCRYPTION_KEY: case FS_IOC_REMOVE_ENCRYPTION_KEY_ALL_USERS: case FS_IOC_GET_ENCRYPTION_KEY_STATUS: + case FS_IOC_GET_ENCRYPTION_NONCE: case EXT4_IOC_SHUTDOWN: case FS_IOC_GETFSMAP: case FS_IOC_ENABLE_VERITY: From 48d09b926aa6fc88144cc39ea3dccc6c4f09a2a8 Mon Sep 17 00:00:00 2001 From: Eric Biggers Date: Sat, 14 Mar 2020 13:50:51 -0700 Subject: [PATCH 71/86] f2fs: wire up FS_IOC_GET_ENCRYPTION_NONCE This new ioctl retrieves a file's encryption nonce, which is useful for testing. See the corresponding fs/crypto/ patch for more details. Link: https://lore.kernel.org/r/20200314205052.93294-4-ebiggers@kernel.org Signed-off-by: Eric Biggers --- fs/f2fs/file.c | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c index eb869bbe5cd5..c15a83c58b60 100644 --- a/fs/f2fs/file.c +++ b/fs/f2fs/file.c @@ -2432,6 +2432,14 @@ static int f2fs_ioc_get_encryption_key_status(struct file *filp, return fscrypt_ioctl_get_key_status(filp, (void __user *)arg); } +static int f2fs_ioc_get_encryption_nonce(struct file *filp, unsigned long arg) +{ + if (!f2fs_sb_has_encrypt(F2FS_I_SB(file_inode(filp)))) + return -EOPNOTSUPP; + + return fscrypt_ioctl_get_nonce(filp, (void __user *)arg); +} + static int f2fs_ioc_gc(struct file *filp, unsigned long arg) { struct inode *inode = file_inode(filp); @@ -3461,6 +3469,8 @@ long f2fs_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) return f2fs_ioc_remove_encryption_key_all_users(filp, arg); case FS_IOC_GET_ENCRYPTION_KEY_STATUS: return f2fs_ioc_get_encryption_key_status(filp, arg); + case FS_IOC_GET_ENCRYPTION_NONCE: + return f2fs_ioc_get_encryption_nonce(filp, arg); case F2FS_IOC_GARBAGE_COLLECT: return f2fs_ioc_gc(filp, arg); case F2FS_IOC_GARBAGE_COLLECT_RANGE: @@ -3638,6 +3648,7 @@ long f2fs_compat_ioctl(struct file *file, unsigned int cmd, unsigned long arg) case FS_IOC_REMOVE_ENCRYPTION_KEY: case FS_IOC_REMOVE_ENCRYPTION_KEY_ALL_USERS: case FS_IOC_GET_ENCRYPTION_KEY_STATUS: + case FS_IOC_GET_ENCRYPTION_NONCE: case F2FS_IOC_GARBAGE_COLLECT: case F2FS_IOC_GARBAGE_COLLECT_RANGE: case F2FS_IOC_WRITE_CHECKPOINT: From 32dff5fe0a82f13973a2089ba5dff6d73523fc3a Mon Sep 17 00:00:00 2001 From: Eric Biggers Date: Sat, 14 Mar 2020 13:50:52 -0700 Subject: [PATCH 72/86] ubifs: wire up FS_IOC_GET_ENCRYPTION_NONCE This new ioctl retrieves a file's encryption nonce, which is useful for testing. See the corresponding fs/crypto/ patch for more details. Link: https://lore.kernel.org/r/20200314205052.93294-5-ebiggers@kernel.org Signed-off-by: Eric Biggers --- fs/ubifs/ioctl.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/fs/ubifs/ioctl.c b/fs/ubifs/ioctl.c index 5dc5abca11c7..f69eb15f4719 100644 --- a/fs/ubifs/ioctl.c +++ b/fs/ubifs/ioctl.c @@ -200,6 +200,9 @@ long ubifs_ioctl(struct file *file, unsigned int cmd, unsigned long arg) case FS_IOC_GET_ENCRYPTION_KEY_STATUS: return fscrypt_ioctl_get_key_status(file, (void __user *)arg); + case FS_IOC_GET_ENCRYPTION_NONCE: + return fscrypt_ioctl_get_nonce(file, (void __user *)arg); + default: return -ENOTTY; } @@ -222,6 +225,7 @@ long ubifs_compat_ioctl(struct file *file, unsigned int cmd, unsigned long arg) case FS_IOC_REMOVE_ENCRYPTION_KEY: case FS_IOC_REMOVE_ENCRYPTION_KEY_ALL_USERS: case FS_IOC_GET_ENCRYPTION_KEY_STATUS: + case FS_IOC_GET_ENCRYPTION_NONCE: break; default: return -ENOIOCTLCMD; From bf8f3e073ca2279e53a855f79e67fe8927f7f791 Mon Sep 17 00:00:00 2001 From: Jaegeuk Kim Date: Thu, 2 Apr 2020 09:32:35 -0700 Subject: [PATCH 73/86] f2fs: introduce sysfs/data_io_flag to attach REQ_META/FUA This patch introduces a way to attach REQ_META/FUA explicitly to all the data writes given temperature. -> attach REQ_FUA to Hot Data writes -> attach REQ_FUA to Hot|Warm Data writes -> attach REQ_FUA to Hot|Warm|Cold Data writes -> attach REQ_FUA to Hot|Warm|Cold Data writes as well as REQ_META to Hot Data writes Reviewed-by: Chao Yu Signed-off-by: Jaegeuk Kim --- Documentation/ABI/testing/sysfs-fs-f2fs | 9 +++++++++ fs/f2fs/data.c | 23 +++++++++++++++++++++++ fs/f2fs/f2fs.h | 3 +++ fs/f2fs/sysfs.c | 3 ++- 4 files changed, 37 insertions(+), 1 deletion(-) diff --git a/Documentation/ABI/testing/sysfs-fs-f2fs b/Documentation/ABI/testing/sysfs-fs-f2fs index bd8a0d19abe6..c8620ea7022a 100644 --- a/Documentation/ABI/testing/sysfs-fs-f2fs +++ b/Documentation/ABI/testing/sysfs-fs-f2fs @@ -323,3 +323,12 @@ What: /sys/fs/f2fs//mounted_time_sec Date: February 2020 Contact: "Jaegeuk Kim" Description: Show the mounted time in secs of this partition. + +What: /sys/fs/f2fs//data_io_flag +Date: April 2020 +Contact: "Jaegeuk Kim" +Description: Give a way to attach REQ_META|FUA to data writes + given temperature-based bits. Now the bits indicate: + * REQ_META | REQ_FUA | + * 5 | 4 | 3 | 2 | 1 | 0 | + * Cold | Warm | Hot | Cold | Warm | Hot | diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c index 3355239b317e..cfe05dfa3fa8 100644 --- a/fs/f2fs/data.c +++ b/fs/f2fs/data.c @@ -513,6 +513,28 @@ void f2fs_submit_bio(struct f2fs_sb_info *sbi, __submit_bio(sbi, bio, type); } +static void __attach_data_io_flag(struct f2fs_io_info *fio) +{ + struct f2fs_sb_info *sbi = fio->sbi; + unsigned int temp_mask = (1 << NR_TEMP_TYPE) - 1; + unsigned int fua_flag = sbi->data_io_flag & temp_mask; + unsigned int meta_flag = (sbi->data_io_flag >> NR_TEMP_TYPE) & + temp_mask; + /* + * data io flag bits per temp: + * REQ_META | REQ_FUA | + * 5 | 4 | 3 | 2 | 1 | 0 | + * Cold | Warm | Hot | Cold | Warm | Hot | + */ + if (fio->type != DATA) + return; + + if ((1 << fio->temp) & meta_flag) + fio->op_flags |= REQ_META; + if ((1 << fio->temp) & fua_flag) + fio->op_flags |= REQ_FUA; +} + static void __submit_merged_bio(struct f2fs_bio_info *io) { struct f2fs_io_info *fio = &io->fio; @@ -520,6 +542,7 @@ static void __submit_merged_bio(struct f2fs_bio_info *io) if (!io->bio) return; + __attach_data_io_flag(fio); bio_set_op_attrs(io->bio, fio->op, fio->op_flags); if (is_read_io(fio->op)) diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index c84442eefc56..56f39f396b3b 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -1506,6 +1506,9 @@ struct f2fs_sb_info { unsigned long long write_iostat[NR_IO_TYPE]; bool iostat_enable; + /* to attach REQ_META|REQ_FUA flags */ + unsigned int data_io_flag; + /* For sysfs suppport */ struct kobject s_kobj; struct completion s_kobj_unregister; diff --git a/fs/f2fs/sysfs.c b/fs/f2fs/sysfs.c index e3bbbef9b4f0..aeebfb5024a2 100644 --- a/fs/f2fs/sysfs.c +++ b/fs/f2fs/sysfs.c @@ -372,7 +372,6 @@ out: return count; } - if (!strcmp(a->attr.name, "iostat_enable")) { sbi->iostat_enable = !!t; if (!sbi->iostat_enable) @@ -543,6 +542,7 @@ F2FS_RW_ATTR(F2FS_SBI, f2fs_super_block, extension_list, extension_list); F2FS_RW_ATTR(FAULT_INFO_RATE, f2fs_fault_info, inject_rate, inject_rate); F2FS_RW_ATTR(FAULT_INFO_TYPE, f2fs_fault_info, inject_type, inject_type); #endif +F2FS_RW_ATTR(F2FS_SBI, f2fs_sb_info, data_io_flag, data_io_flag); F2FS_GENERAL_RO_ATTR(dirty_segments); F2FS_GENERAL_RO_ATTR(free_segments); F2FS_GENERAL_RO_ATTR(lifetime_write_kbytes); @@ -622,6 +622,7 @@ static struct attribute *f2fs_attrs[] = { ATTR_LIST(inject_rate), ATTR_LIST(inject_type), #endif + ATTR_LIST(data_io_flag), ATTR_LIST(dirty_segments), ATTR_LIST(free_segments), ATTR_LIST(unusable), From a6b67c6e52f1d5b33ac194b83c0fd295c8dd9dda Mon Sep 17 00:00:00 2001 From: Daeho Jeong Date: Mon, 30 Mar 2020 03:30:59 +0000 Subject: [PATCH 74/86] f2fs: add tracepoint for f2fs iostat Added a tracepoint to see iostat of f2fs. Default period of that is 3 second. This tracepoint can be used to be monitoring I/O statistics periodically. Signed-off-by: Daeho Jeong Reviewed-by: Chao Yu Signed-off-by: Jaegeuk Kim --- Documentation/ABI/testing/sysfs-fs-f2fs | 6 +++ fs/f2fs/f2fs.h | 16 +++++++- fs/f2fs/super.c | 1 + fs/f2fs/sysfs.c | 39 +++++++++++++++++++ include/trace/events/f2fs.h | 52 +++++++++++++++++++++++++ 5 files changed, 113 insertions(+), 1 deletion(-) diff --git a/Documentation/ABI/testing/sysfs-fs-f2fs b/Documentation/ABI/testing/sysfs-fs-f2fs index c8620ea7022a..427f5b45c67f 100644 --- a/Documentation/ABI/testing/sysfs-fs-f2fs +++ b/Documentation/ABI/testing/sysfs-fs-f2fs @@ -332,3 +332,9 @@ Description: Give a way to attach REQ_META|FUA to data writes * REQ_META | REQ_FUA | * 5 | 4 | 3 | 2 | 1 | 0 | * Cold | Warm | Hot | Cold | Warm | Hot | + +What: /sys/fs/f2fs//iostat_period_ms +Date: April 2020 +Contact: "Daeho Jeong" +Description: Give a way to change iostat_period time. 3secs by default. + The new iostat trace gives stats gap given the period. diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index 56f39f396b3b..b41bfd7ab106 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -1504,7 +1504,10 @@ struct f2fs_sb_info { /* For app/fs IO statistics */ spinlock_t iostat_lock; unsigned long long write_iostat[NR_IO_TYPE]; + unsigned long long prev_write_iostat[NR_IO_TYPE]; bool iostat_enable; + unsigned long iostat_next_period; + unsigned int iostat_period_ms; /* to attach REQ_META|REQ_FUA flags */ unsigned int data_io_flag; @@ -2998,16 +3001,25 @@ static inline int get_inline_xattr_addrs(struct inode *inode) sizeof((f2fs_inode)->field)) \ <= (F2FS_OLD_ATTRIBUTE_SIZE + (extra_isize))) \ +#define DEFAULT_IOSTAT_PERIOD_MS 3000 +#define MIN_IOSTAT_PERIOD_MS 100 +/* maximum period of iostat tracing is 1 day */ +#define MAX_IOSTAT_PERIOD_MS 8640000 + static inline void f2fs_reset_iostat(struct f2fs_sb_info *sbi) { int i; spin_lock(&sbi->iostat_lock); - for (i = 0; i < NR_IO_TYPE; i++) + for (i = 0; i < NR_IO_TYPE; i++) { sbi->write_iostat[i] = 0; + sbi->prev_write_iostat[i] = 0; + } spin_unlock(&sbi->iostat_lock); } +extern void f2fs_record_iostat(struct f2fs_sb_info *sbi); + static inline void f2fs_update_iostat(struct f2fs_sb_info *sbi, enum iostat_type type, unsigned long long io_bytes) { @@ -3021,6 +3033,8 @@ static inline void f2fs_update_iostat(struct f2fs_sb_info *sbi, sbi->write_iostat[APP_WRITE_IO] - sbi->write_iostat[APP_DIRECT_IO]; spin_unlock(&sbi->iostat_lock); + + f2fs_record_iostat(sbi); } #define __is_large_section(sbi) ((sbi)->segs_per_sec > 1) diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c index dcf506e3cc32..2a29964729f0 100644 --- a/fs/f2fs/super.c +++ b/fs/f2fs/super.c @@ -3423,6 +3423,7 @@ try_onemore: /* init iostat info */ spin_lock_init(&sbi->iostat_lock); sbi->iostat_enable = false; + sbi->iostat_period_ms = DEFAULT_IOSTAT_PERIOD_MS; for (i = 0; i < NR_PAGE_TYPE; i++) { int n = (i == META) ? 1: NR_TEMP_TYPE; diff --git a/fs/f2fs/sysfs.c b/fs/f2fs/sysfs.c index aeebfb5024a2..d05cb68c2637 100644 --- a/fs/f2fs/sysfs.c +++ b/fs/f2fs/sysfs.c @@ -15,6 +15,7 @@ #include "f2fs.h" #include "segment.h" #include "gc.h" +#include static struct proc_dir_entry *f2fs_proc_root; @@ -379,6 +380,15 @@ out: return count; } + if (!strcmp(a->attr.name, "iostat_period_ms")) { + if (t < MIN_IOSTAT_PERIOD_MS || t > MAX_IOSTAT_PERIOD_MS) + return -EINVAL; + spin_lock(&sbi->iostat_lock); + sbi->iostat_period_ms = (unsigned int)t; + spin_unlock(&sbi->iostat_lock); + return count; + } + *ui = (unsigned int)t; return count; @@ -535,6 +545,7 @@ F2FS_RW_ATTR(F2FS_SBI, f2fs_sb_info, gc_idle_interval, interval_time[GC_TIME]); F2FS_RW_ATTR(F2FS_SBI, f2fs_sb_info, umount_discard_timeout, interval_time[UMOUNT_DISCARD_TIMEOUT]); F2FS_RW_ATTR(F2FS_SBI, f2fs_sb_info, iostat_enable, iostat_enable); +F2FS_RW_ATTR(F2FS_SBI, f2fs_sb_info, iostat_period_ms, iostat_period_ms); F2FS_RW_ATTR(F2FS_SBI, f2fs_sb_info, readdir_ra, readdir_ra); F2FS_RW_ATTR(F2FS_SBI, f2fs_sb_info, gc_pin_file_thresh, gc_pin_file_threshold); F2FS_RW_ATTR(F2FS_SBI, f2fs_super_block, extension_list, extension_list); @@ -615,6 +626,7 @@ static struct attribute *f2fs_attrs[] = { ATTR_LIST(gc_idle_interval), ATTR_LIST(umount_discard_timeout), ATTR_LIST(iostat_enable), + ATTR_LIST(iostat_period_ms), ATTR_LIST(readdir_ra), ATTR_LIST(gc_pin_file_thresh), ATTR_LIST(extension_list), @@ -751,6 +763,33 @@ static int __maybe_unused segment_bits_seq_show(struct seq_file *seq, return 0; } +void f2fs_record_iostat(struct f2fs_sb_info *sbi) +{ + unsigned long long iostat_diff[NR_IO_TYPE]; + int i; + + if (time_is_after_jiffies(sbi->iostat_next_period)) + return; + + /* Need double check under the lock */ + spin_lock(&sbi->iostat_lock); + if (time_is_after_jiffies(sbi->iostat_next_period)) { + spin_unlock(&sbi->iostat_lock); + return; + } + sbi->iostat_next_period = jiffies + + msecs_to_jiffies(sbi->iostat_period_ms); + + for (i = 0; i < NR_IO_TYPE; i++) { + iostat_diff[i] = sbi->write_iostat[i] - + sbi->prev_write_iostat[i]; + sbi->prev_write_iostat[i] = sbi->write_iostat[i]; + } + spin_unlock(&sbi->iostat_lock); + + trace_f2fs_iostat(sbi, iostat_diff); +} + static int __maybe_unused iostat_info_seq_show(struct seq_file *seq, void *offset) { diff --git a/include/trace/events/f2fs.h b/include/trace/events/f2fs.h index d4647dc49275..ded046fce8f9 100644 --- a/include/trace/events/f2fs.h +++ b/include/trace/events/f2fs.h @@ -1812,6 +1812,58 @@ DEFINE_EVENT(f2fs_zip_end, f2fs_decompress_pages_end, TP_ARGS(inode, cluster_idx, compressed_size, ret) ); +TRACE_EVENT(f2fs_iostat, + + TP_PROTO(struct f2fs_sb_info *sbi, unsigned long long *iostat), + + TP_ARGS(sbi, iostat), + + TP_STRUCT__entry( + __field(dev_t, dev) + __field(unsigned long long, app_dio) + __field(unsigned long long, app_bio) + __field(unsigned long long, app_wio) + __field(unsigned long long, app_mio) + __field(unsigned long long, fs_dio) + __field(unsigned long long, fs_nio) + __field(unsigned long long, fs_mio) + __field(unsigned long long, fs_gc_dio) + __field(unsigned long long, fs_gc_nio) + __field(unsigned long long, fs_cp_dio) + __field(unsigned long long, fs_cp_nio) + __field(unsigned long long, fs_cp_mio) + __field(unsigned long long, fs_discard) + ), + + TP_fast_assign( + __entry->dev = sbi->sb->s_dev; + __entry->app_dio = iostat[APP_DIRECT_IO]; + __entry->app_bio = iostat[APP_BUFFERED_IO]; + __entry->app_wio = iostat[APP_WRITE_IO]; + __entry->app_mio = iostat[APP_MAPPED_IO]; + __entry->fs_dio = iostat[FS_DATA_IO]; + __entry->fs_nio = iostat[FS_NODE_IO]; + __entry->fs_mio = iostat[FS_META_IO]; + __entry->fs_gc_dio = iostat[FS_GC_DATA_IO]; + __entry->fs_gc_nio = iostat[FS_GC_NODE_IO]; + __entry->fs_cp_dio = iostat[FS_CP_DATA_IO]; + __entry->fs_cp_nio = iostat[FS_CP_NODE_IO]; + __entry->fs_cp_mio = iostat[FS_CP_META_IO]; + __entry->fs_discard = iostat[FS_DISCARD]; + ), + + TP_printk("dev = (%d,%d), " + "app [write=%llu (direct=%llu, buffered=%llu), mapped=%llu], " + "fs [data=%llu, node=%llu, meta=%llu, discard=%llu], " + "gc [data=%llu, node=%llu], " + "cp [data=%llu, node=%llu, meta=%llu]", + show_dev(__entry->dev), __entry->app_wio, __entry->app_dio, + __entry->app_bio, __entry->app_mio, __entry->fs_dio, + __entry->fs_nio, __entry->fs_mio, __entry->fs_discard, + __entry->fs_gc_dio, __entry->fs_gc_nio, __entry->fs_cp_dio, + __entry->fs_cp_nio, __entry->fs_cp_mio) +); + #endif /* _TRACE_F2FS_H */ /* This part must be outside protection */ From 5bc3833f1667d534adae3e1f63e45e9b374ac1fe Mon Sep 17 00:00:00 2001 From: Sahitya Tummala Date: Wed, 15 Apr 2020 14:37:53 +0530 Subject: [PATCH 75/86] f2fs: fix long latency due to discard during umount F2FS already has a default timeout of 5 secs for discards that can be issued during umount, but it can take more than the 5 sec timeout if the underlying UFS device queue is already full and there are no more available free tags to be used. Fix this by submitting a small batch of discard requests so that it won't cause the device queue to be full at any time and thus doesn't incur its wait time in the umount context. Signed-off-by: Sahitya Tummala Reviewed-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/segment.c | 12 ++++++++++-- 1 file changed, 10 insertions(+), 2 deletions(-) diff --git a/fs/f2fs/segment.c b/fs/f2fs/segment.c index b7a9421472a7..90c75822fef6 100644 --- a/fs/f2fs/segment.c +++ b/fs/f2fs/segment.c @@ -1101,7 +1101,6 @@ static void __init_discard_policy(struct f2fs_sb_info *sbi, } else if (discard_type == DPOLICY_FSTRIM) { dpolicy->io_aware = false; } else if (discard_type == DPOLICY_UMOUNT) { - dpolicy->max_requests = UINT_MAX; dpolicy->io_aware = false; /* we need to issue all to keep CP_TRIMMED_FLAG */ dpolicy->granularity = 1; @@ -1463,6 +1462,8 @@ next: return issued; } +static unsigned int __wait_all_discard_cmd(struct f2fs_sb_info *sbi, + struct discard_policy *dpolicy); static int __issue_discard_cmd(struct f2fs_sb_info *sbi, struct discard_policy *dpolicy) @@ -1471,12 +1472,14 @@ static int __issue_discard_cmd(struct f2fs_sb_info *sbi, struct list_head *pend_list; struct discard_cmd *dc, *tmp; struct blk_plug plug; - int i, issued = 0; + int i, issued; bool io_interrupted = false; if (dpolicy->timeout) f2fs_update_time(sbi, UMOUNT_DISCARD_TIMEOUT); +retry: + issued = 0; for (i = MAX_PLIST_NUM - 1; i >= 0; i--) { if (dpolicy->timeout && f2fs_time_over(sbi, UMOUNT_DISCARD_TIMEOUT)) @@ -1523,6 +1526,11 @@ next: break; } + if (dpolicy->type == DPOLICY_UMOUNT && issued) { + __wait_all_discard_cmd(sbi, dpolicy); + goto retry; + } + if (!issued && io_interrupted) issued = -1; From 9a26fc81a1b913fafb3ffab8751ae5393b8009ca Mon Sep 17 00:00:00 2001 From: Sahitya Tummala Date: Wed, 15 Apr 2020 09:35:54 +0530 Subject: [PATCH 76/86] f2fs: report the discard cmd errors properly In case a discard_cmd is split into several bios, the dc->error must not be overwritten once an error is reported by a bio. Also, move it under dc->lock. Signed-off-by: Sahitya Tummala Reviewed-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/segment.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/fs/f2fs/segment.c b/fs/f2fs/segment.c index 90c75822fef6..728ff6e316ac 100644 --- a/fs/f2fs/segment.c +++ b/fs/f2fs/segment.c @@ -1029,9 +1029,9 @@ static void f2fs_submit_discard_endio(struct bio *bio) struct discard_cmd *dc = (struct discard_cmd *)bio->bi_private; unsigned long flags; - dc->error = blk_status_to_errno(bio->bi_status); - spin_lock_irqsave(&dc->lock, flags); + if (!dc->error) + dc->error = blk_status_to_errno(bio->bi_status); dc->bio_ref--; if (!dc->bio_ref && dc->state == D_SUBMIT) { dc->state = D_DONE; From ccd59287f7650809ee8100387bba985a50228de1 Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Fri, 10 Apr 2020 18:07:20 +0800 Subject: [PATCH 77/86] f2fs: fix to handle error path of f2fs_ra_meta_pages() In f2fs_ra_meta_pages(), if f2fs_submit_page_bio() failed, we need to unlock page, fix it. Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/checkpoint.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/fs/f2fs/checkpoint.c b/fs/f2fs/checkpoint.c index 852890b72d6a..6be357c8e002 100644 --- a/fs/f2fs/checkpoint.c +++ b/fs/f2fs/checkpoint.c @@ -220,6 +220,7 @@ int f2fs_ra_meta_pages(struct f2fs_sb_info *sbi, block_t start, int nrpages, .is_por = (type == META_POR), }; struct blk_plug plug; + int err; if (unlikely(type == META_POR)) fio.op_flags &= ~REQ_META; @@ -263,8 +264,8 @@ int f2fs_ra_meta_pages(struct f2fs_sb_info *sbi, block_t start, int nrpages, } fio.page = page; - f2fs_submit_page_bio(&fio); - f2fs_put_page(page, 0); + err = f2fs_submit_page_bio(&fio); + f2fs_put_page(page, err ? 1 : 0); } out: blk_finish_plug(&plug); From 90054756a839924ea91d3c33fb3b56050237c87f Mon Sep 17 00:00:00 2001 From: Sahitya Tummala Date: Thu, 16 Apr 2020 11:47:41 +0530 Subject: [PATCH 78/86] f2fs: Fix the accounting of dcc->undiscard_blks When a discard_cmd needs to be split due to dpolicy->max_requests, then for the remaining length it will be either merged into another cmd or a new discard_cmd will be created. In this case, there is double accounting of dcc->undiscard_blks for the remaining len, due to which it shows incorrect value in stats. Signed-off-by: Sahitya Tummala Reviewed-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/segment.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/fs/f2fs/segment.c b/fs/f2fs/segment.c index 728ff6e316ac..1c48ec866b8c 100644 --- a/fs/f2fs/segment.c +++ b/fs/f2fs/segment.c @@ -1214,8 +1214,10 @@ submit: len = total_len; } - if (!err && len) + if (!err && len) { + dcc->undiscard_blks -= len; __update_discard_tree_range(sbi, bdev, lstart, start, len); + } return err; } From d7bbecc6b83cfd68dbc53bfa24937d8985c3bd9e Mon Sep 17 00:00:00 2001 From: Chao Yu Date: Thu, 16 Apr 2020 18:16:56 +0800 Subject: [PATCH 79/86] f2fs: support read iostat Adds to support accounting read IOs from userspace/kernel. Signed-off-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/checkpoint.c | 5 ++++ fs/f2fs/data.c | 6 +++++ fs/f2fs/f2fs.h | 37 +++++++++++++++++++------- fs/f2fs/file.c | 12 ++++++++- fs/f2fs/gc.c | 6 +++++ fs/f2fs/node.c | 8 +++++- fs/f2fs/sysfs.c | 52 +++++++++++++++++++++++++------------ include/trace/events/f2fs.h | 23 ++++++++++++++-- 8 files changed, 118 insertions(+), 31 deletions(-) diff --git a/fs/f2fs/checkpoint.c b/fs/f2fs/checkpoint.c index 6be357c8e002..5ba649e17c72 100644 --- a/fs/f2fs/checkpoint.c +++ b/fs/f2fs/checkpoint.c @@ -86,6 +86,8 @@ repeat: return ERR_PTR(err); } + f2fs_update_iostat(sbi, FS_META_READ_IO, F2FS_BLKSIZE); + lock_page(page); if (unlikely(page->mapping != mapping)) { f2fs_put_page(page, 1); @@ -266,6 +268,9 @@ int f2fs_ra_meta_pages(struct f2fs_sb_info *sbi, block_t start, int nrpages, fio.page = page; err = f2fs_submit_page_bio(&fio); f2fs_put_page(page, err ? 1 : 0); + + if (!err) + f2fs_update_iostat(sbi, FS_META_READ_IO, F2FS_BLKSIZE); } out: blk_finish_plug(&plug); diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c index cfe05dfa3fa8..aacb0b5aa4b0 100644 --- a/fs/f2fs/data.c +++ b/fs/f2fs/data.c @@ -1033,6 +1033,7 @@ static int f2fs_submit_page_read(struct inode *inode, struct page *page, } ClearPageError(page); inc_page_count(sbi, F2FS_RD_DATA); + f2fs_update_iostat(sbi, FS_DATA_READ_IO, F2FS_BLKSIZE); __submit_bio(sbi, bio, DATA); return 0; } @@ -2038,6 +2039,7 @@ submit_and_realloc: goto submit_and_realloc; inc_page_count(F2FS_I_SB(inode), F2FS_RD_DATA); + f2fs_update_iostat(F2FS_I_SB(inode), FS_DATA_READ_IO, F2FS_BLKSIZE); ClearPageError(page); *last_block_in_bio = block_nr; goto out; @@ -2173,6 +2175,7 @@ submit_and_realloc: goto submit_and_realloc; inc_page_count(sbi, F2FS_RD_DATA); + f2fs_update_iostat(sbi, FS_DATA_READ_IO, F2FS_BLKSIZE); ClearPageError(page); *last_block_in_bio = blkaddr; } @@ -3526,6 +3529,9 @@ static ssize_t f2fs_direct_IO(struct kiocb *iocb, struct iov_iter *iter) } else if (err < 0) { f2fs_write_failed(mapping, offset + count); } + } else { + if (err > 0) + f2fs_update_iostat(sbi, APP_DIRECT_READ_IO, err); } out: diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index b41bfd7ab106..72016c29aa46 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -1087,8 +1087,9 @@ enum cp_reason_type { }; enum iostat_type { - APP_DIRECT_IO, /* app direct IOs */ - APP_BUFFERED_IO, /* app buffered IOs */ + /* WRITE IO */ + APP_DIRECT_IO, /* app direct write IOs */ + APP_BUFFERED_IO, /* app buffered write IOs */ APP_WRITE_IO, /* app write IOs */ APP_MAPPED_IO, /* app mapped IOs */ FS_DATA_IO, /* data IOs from kworker/fsync/reclaimer */ @@ -1099,6 +1100,17 @@ enum iostat_type { FS_CP_DATA_IO, /* data IOs from checkpoint */ FS_CP_NODE_IO, /* node IOs from checkpoint */ FS_CP_META_IO, /* meta IOs from checkpoint */ + + /* READ IO */ + APP_DIRECT_READ_IO, /* app direct read IOs */ + APP_BUFFERED_READ_IO, /* app buffered read IOs */ + APP_READ_IO, /* app read IOs */ + APP_MAPPED_READ_IO, /* app mapped read IOs */ + FS_DATA_READ_IO, /* data read IOs */ + FS_NODE_READ_IO, /* node read IOs */ + FS_META_READ_IO, /* meta read IOs */ + + /* other */ FS_DISCARD, /* discard */ NR_IO_TYPE, }; @@ -1503,8 +1515,8 @@ struct f2fs_sb_info { /* For app/fs IO statistics */ spinlock_t iostat_lock; - unsigned long long write_iostat[NR_IO_TYPE]; - unsigned long long prev_write_iostat[NR_IO_TYPE]; + unsigned long long rw_iostat[NR_IO_TYPE]; + unsigned long long prev_rw_iostat[NR_IO_TYPE]; bool iostat_enable; unsigned long iostat_next_period; unsigned int iostat_period_ms; @@ -3012,8 +3024,8 @@ static inline void f2fs_reset_iostat(struct f2fs_sb_info *sbi) spin_lock(&sbi->iostat_lock); for (i = 0; i < NR_IO_TYPE; i++) { - sbi->write_iostat[i] = 0; - sbi->prev_write_iostat[i] = 0; + sbi->rw_iostat[i] = 0; + sbi->prev_rw_iostat[i] = 0; } spin_unlock(&sbi->iostat_lock); } @@ -3026,12 +3038,17 @@ static inline void f2fs_update_iostat(struct f2fs_sb_info *sbi, if (!sbi->iostat_enable) return; spin_lock(&sbi->iostat_lock); - sbi->write_iostat[type] += io_bytes; + sbi->rw_iostat[type] += io_bytes; if (type == APP_WRITE_IO || type == APP_DIRECT_IO) - sbi->write_iostat[APP_BUFFERED_IO] = - sbi->write_iostat[APP_WRITE_IO] - - sbi->write_iostat[APP_DIRECT_IO]; + sbi->rw_iostat[APP_BUFFERED_IO] = + sbi->rw_iostat[APP_WRITE_IO] - + sbi->rw_iostat[APP_DIRECT_IO]; + + if (type == APP_READ_IO || type == APP_DIRECT_READ_IO) + sbi->rw_iostat[APP_BUFFERED_READ_IO] = + sbi->rw_iostat[APP_READ_IO] - + sbi->rw_iostat[APP_DIRECT_READ_IO]; spin_unlock(&sbi->iostat_lock); f2fs_record_iostat(sbi); diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c index c15a83c58b60..3d24f3e6dbe7 100644 --- a/fs/f2fs/file.c +++ b/fs/f2fs/file.c @@ -40,6 +40,10 @@ static vm_fault_t f2fs_filemap_fault(struct vm_fault *vmf) ret = filemap_fault(vmf); up_read(&F2FS_I(inode)->i_mmap_sem); + if (!ret) + f2fs_update_iostat(F2FS_I_SB(inode), APP_MAPPED_READ_IO, + F2FS_BLKSIZE); + trace_f2fs_filemap_fault(inode, vmf->pgoff, (unsigned long)ret); return ret; @@ -3516,11 +3520,17 @@ static ssize_t f2fs_file_read_iter(struct kiocb *iocb, struct iov_iter *iter) { struct file *file = iocb->ki_filp; struct inode *inode = file_inode(file); + int ret; if (!f2fs_is_compress_backend_ready(inode)) return -EOPNOTSUPP; - return generic_file_read_iter(iocb, iter); + ret = generic_file_read_iter(iocb, iter); + + if (ret > 0) + f2fs_update_iostat(F2FS_I_SB(inode), APP_READ_IO, ret); + + return ret; } static ssize_t f2fs_file_write_iter(struct kiocb *iocb, struct iov_iter *from) diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c index 26248c8936db..28a8c79c8bdc 100644 --- a/fs/f2fs/gc.c +++ b/fs/f2fs/gc.c @@ -737,6 +737,9 @@ got_it: goto put_encrypted_page; f2fs_put_page(fio.encrypted_page, 0); f2fs_put_page(page, 1); + + f2fs_update_iostat(sbi, FS_DATA_READ_IO, F2FS_BLKSIZE); + return 0; put_encrypted_page: f2fs_put_page(fio.encrypted_page, 1); @@ -840,6 +843,9 @@ static int move_data_block(struct inode *inode, block_t bidx, f2fs_put_page(mpage, 1); goto up_out; } + + f2fs_update_iostat(fio.sbi, FS_DATA_READ_IO, F2FS_BLKSIZE); + lock_page(mpage); if (unlikely(mpage->mapping != META_MAPPING(fio.sbi) || !PageUptodate(mpage))) { diff --git a/fs/f2fs/node.c b/fs/f2fs/node.c index c74b6da2ef05..8f78deeca538 100644 --- a/fs/f2fs/node.c +++ b/fs/f2fs/node.c @@ -1300,7 +1300,13 @@ static int read_node_page(struct page *page, int op_flags) } fio.new_blkaddr = fio.old_blkaddr = ni.blk_addr; - return f2fs_submit_page_bio(&fio); + + err = f2fs_submit_page_bio(&fio); + + if (!err) + f2fs_update_iostat(sbi, FS_NODE_READ_IO, F2FS_BLKSIZE); + + return err; } /* diff --git a/fs/f2fs/sysfs.c b/fs/f2fs/sysfs.c index d05cb68c2637..eaf8088548f0 100644 --- a/fs/f2fs/sysfs.c +++ b/fs/f2fs/sysfs.c @@ -781,9 +781,9 @@ void f2fs_record_iostat(struct f2fs_sb_info *sbi) msecs_to_jiffies(sbi->iostat_period_ms); for (i = 0; i < NR_IO_TYPE; i++) { - iostat_diff[i] = sbi->write_iostat[i] - - sbi->prev_write_iostat[i]; - sbi->prev_write_iostat[i] = sbi->write_iostat[i]; + iostat_diff[i] = sbi->rw_iostat[i] - + sbi->prev_rw_iostat[i]; + sbi->prev_rw_iostat[i] = sbi->rw_iostat[i]; } spin_unlock(&sbi->iostat_lock); @@ -802,33 +802,51 @@ static int __maybe_unused iostat_info_seq_show(struct seq_file *seq, seq_printf(seq, "time: %-16llu\n", now); - /* print app IOs */ + /* print app write IOs */ seq_printf(seq, "app buffered: %-16llu\n", - sbi->write_iostat[APP_BUFFERED_IO]); + sbi->rw_iostat[APP_BUFFERED_IO]); seq_printf(seq, "app direct: %-16llu\n", - sbi->write_iostat[APP_DIRECT_IO]); + sbi->rw_iostat[APP_DIRECT_IO]); seq_printf(seq, "app mapped: %-16llu\n", - sbi->write_iostat[APP_MAPPED_IO]); + sbi->rw_iostat[APP_MAPPED_IO]); - /* print fs IOs */ + /* print fs write IOs */ seq_printf(seq, "fs data: %-16llu\n", - sbi->write_iostat[FS_DATA_IO]); + sbi->rw_iostat[FS_DATA_IO]); seq_printf(seq, "fs node: %-16llu\n", - sbi->write_iostat[FS_NODE_IO]); + sbi->rw_iostat[FS_NODE_IO]); seq_printf(seq, "fs meta: %-16llu\n", - sbi->write_iostat[FS_META_IO]); + sbi->rw_iostat[FS_META_IO]); seq_printf(seq, "fs gc data: %-16llu\n", - sbi->write_iostat[FS_GC_DATA_IO]); + sbi->rw_iostat[FS_GC_DATA_IO]); seq_printf(seq, "fs gc node: %-16llu\n", - sbi->write_iostat[FS_GC_NODE_IO]); + sbi->rw_iostat[FS_GC_NODE_IO]); seq_printf(seq, "fs cp data: %-16llu\n", - sbi->write_iostat[FS_CP_DATA_IO]); + sbi->rw_iostat[FS_CP_DATA_IO]); seq_printf(seq, "fs cp node: %-16llu\n", - sbi->write_iostat[FS_CP_NODE_IO]); + sbi->rw_iostat[FS_CP_NODE_IO]); seq_printf(seq, "fs cp meta: %-16llu\n", - sbi->write_iostat[FS_CP_META_IO]); + sbi->rw_iostat[FS_CP_META_IO]); + + /* print app read IOs */ + seq_printf(seq, "app buffered: %-16llu\n", + sbi->rw_iostat[APP_BUFFERED_READ_IO]); + seq_printf(seq, "app direct: %-16llu\n", + sbi->rw_iostat[APP_DIRECT_READ_IO]); + seq_printf(seq, "app mapped: %-16llu\n", + sbi->rw_iostat[APP_MAPPED_READ_IO]); + + /* print fs read IOs */ + seq_printf(seq, "fs data: %-16llu\n", + sbi->rw_iostat[FS_DATA_READ_IO]); + seq_printf(seq, "fs node: %-16llu\n", + sbi->rw_iostat[FS_NODE_READ_IO]); + seq_printf(seq, "fs meta: %-16llu\n", + sbi->rw_iostat[FS_META_READ_IO]); + + /* print other IOs */ seq_printf(seq, "fs discard: %-16llu\n", - sbi->write_iostat[FS_DISCARD]); + sbi->rw_iostat[FS_DISCARD]); return 0; } diff --git a/include/trace/events/f2fs.h b/include/trace/events/f2fs.h index ded046fce8f9..542b02544162 100644 --- a/include/trace/events/f2fs.h +++ b/include/trace/events/f2fs.h @@ -1832,6 +1832,13 @@ TRACE_EVENT(f2fs_iostat, __field(unsigned long long, fs_cp_dio) __field(unsigned long long, fs_cp_nio) __field(unsigned long long, fs_cp_mio) + __field(unsigned long long, app_drio) + __field(unsigned long long, app_brio) + __field(unsigned long long, app_rio) + __field(unsigned long long, app_mrio) + __field(unsigned long long, fs_drio) + __field(unsigned long long, fs_nrio) + __field(unsigned long long, fs_mrio) __field(unsigned long long, fs_discard) ), @@ -1849,6 +1856,13 @@ TRACE_EVENT(f2fs_iostat, __entry->fs_cp_dio = iostat[FS_CP_DATA_IO]; __entry->fs_cp_nio = iostat[FS_CP_NODE_IO]; __entry->fs_cp_mio = iostat[FS_CP_META_IO]; + __entry->app_drio = iostat[APP_DIRECT_READ_IO]; + __entry->app_brio = iostat[APP_BUFFERED_READ_IO]; + __entry->app_rio = iostat[APP_READ_IO]; + __entry->app_mrio = iostat[APP_MAPPED_READ_IO]; + __entry->fs_drio = iostat[FS_DATA_READ_IO]; + __entry->fs_nrio = iostat[FS_NODE_READ_IO]; + __entry->fs_mrio = iostat[FS_META_READ_IO]; __entry->fs_discard = iostat[FS_DISCARD]; ), @@ -1856,12 +1870,17 @@ TRACE_EVENT(f2fs_iostat, "app [write=%llu (direct=%llu, buffered=%llu), mapped=%llu], " "fs [data=%llu, node=%llu, meta=%llu, discard=%llu], " "gc [data=%llu, node=%llu], " - "cp [data=%llu, node=%llu, meta=%llu]", + "cp [data=%llu, node=%llu, meta=%llu], " + "app [read=%llu (direct=%llu, buffered=%llu), mapped=%llu], " + "fs [data=%llu, node=%llu, meta=%llu]", show_dev(__entry->dev), __entry->app_wio, __entry->app_dio, __entry->app_bio, __entry->app_mio, __entry->fs_dio, __entry->fs_nio, __entry->fs_mio, __entry->fs_discard, __entry->fs_gc_dio, __entry->fs_gc_nio, __entry->fs_cp_dio, - __entry->fs_cp_nio, __entry->fs_cp_mio) + __entry->fs_cp_nio, __entry->fs_cp_mio, + __entry->app_rio, __entry->app_drio, __entry->app_brio, + __entry->app_mrio, __entry->fs_drio, __entry->fs_nrio, + __entry->fs_mrio) ); #endif /* _TRACE_F2FS_H */ From 8c77bd62d6e4d68a70ba4baecb296b73ab252d2a Mon Sep 17 00:00:00 2001 From: Jaegeuk Kim Date: Thu, 9 Apr 2020 10:25:21 -0700 Subject: [PATCH 80/86] f2fs: fix quota_sync failure due to f2fs_lock_op f2fs_quota_sync() uses f2fs_lock_op() before flushing dirty pages, but f2fs_write_data_page() returns EAGAIN. Likewise dentry blocks, we can just bypass getting the lock, since quota blocks are also maintained by checkpoint. Reviewed-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/compress.c | 2 +- fs/f2fs/data.c | 4 ++-- 2 files changed, 3 insertions(+), 3 deletions(-) diff --git a/fs/f2fs/compress.c b/fs/f2fs/compress.c index df7b2d15eacd..faaa35828901 100644 --- a/fs/f2fs/compress.c +++ b/fs/f2fs/compress.c @@ -985,7 +985,7 @@ static int f2fs_write_compressed_pages(struct compress_ctx *cc, loff_t psize; int i, err; - if (!f2fs_trylock_op(sbi)) + if (!IS_NOQUOTA(inode) && !f2fs_trylock_op(sbi)) return -EAGAIN; set_new_dnode(&dn, cc->inode, NULL, NULL, 0); diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c index aacb0b5aa4b0..71d17ad11c77 100644 --- a/fs/f2fs/data.c +++ b/fs/f2fs/data.c @@ -2659,8 +2659,8 @@ write: f2fs_available_free_memory(sbi, BASE_CHECK)))) goto redirty_out; - /* Dentry blocks are controlled by checkpoint */ - if (S_ISDIR(inode->i_mode)) { + /* Dentry/quota blocks are controlled by checkpoint */ + if (S_ISDIR(inode->i_mode) || IS_NOQUOTA(inode)) { fio.need_lock = LOCK_DONE; err = f2fs_do_write_data_page(&fio); goto done; From a51f842b5079f1f9c331c74b59784f13ea814cb1 Mon Sep 17 00:00:00 2001 From: Jaegeuk Kim Date: Thu, 23 Apr 2020 20:37:43 -0700 Subject: [PATCH 81/86] f2fs: fix missing check for f2fs_unlock_op This fixes what Chao pointed. Upstream patch should be fine. Fixes: 8c77bd62d6e4 ("f2fs: fix quota_sync failure due to f2fs_lock_op") Signed-off-by: Jaegeuk Kim --- fs/f2fs/compress.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/fs/f2fs/compress.c b/fs/f2fs/compress.c index faaa35828901..26b071afe48a 100644 --- a/fs/f2fs/compress.c +++ b/fs/f2fs/compress.c @@ -1092,7 +1092,8 @@ unlock_continue: set_inode_flag(inode, FI_FIRST_BLOCK_WRITTEN); f2fs_put_dnode(&dn); - f2fs_unlock_op(sbi); + if (!IS_NOQUOTA(inode)) + f2fs_unlock_op(sbi); spin_lock(&fi->i_size_lock); if (fi->last_disk_size < psize) @@ -1118,7 +1119,8 @@ out_put_cic: out_put_dnode: f2fs_put_dnode(&dn); out_unlock_op: - f2fs_unlock_op(sbi); + if (!IS_NOQUOTA(inode)) + f2fs_unlock_op(sbi); return -EAGAIN; } From 17662a0b4aa31a15bf83a679d1310866fdc7f04e Mon Sep 17 00:00:00 2001 From: Eric Biggers Date: Thu, 7 May 2020 00:59:02 -0700 Subject: [PATCH 82/86] f2fs: don't leak filename in f2fs_try_convert_inline_dir() We need to call fscrypt_free_filename() to free the memory allocated by fscrypt_setup_filename(). Fixes: b06af2aff28b ("f2fs: convert inline_dir early before starting rename") Cc: # v5.6+ Signed-off-by: Eric Biggers Reviewed-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/inline.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/fs/f2fs/inline.c b/fs/f2fs/inline.c index 4167e5408151..59a4b7ff11e1 100644 --- a/fs/f2fs/inline.c +++ b/fs/f2fs/inline.c @@ -559,12 +559,12 @@ int f2fs_try_convert_inline_dir(struct inode *dir, struct dentry *dentry) ipage = f2fs_get_node_page(sbi, dir->i_ino); if (IS_ERR(ipage)) { err = PTR_ERR(ipage); - goto out; + goto out_fname; } if (f2fs_has_enough_room(dir, ipage, &fname)) { f2fs_put_page(ipage, 1); - goto out; + goto out_fname; } inline_dentry = inline_data_addr(dir, ipage); @@ -572,6 +572,8 @@ int f2fs_try_convert_inline_dir(struct inode *dir, struct dentry *dentry) err = do_convert_inline_dir(dir, ipage, inline_dentry); if (!err) f2fs_put_page(ipage, 1); +out_fname: + fscrypt_free_filename(&fname); out: f2fs_unlock_op(sbi); return err; From 8950f8d6e47b14abeb3ef90fed193a2805a06b8b Mon Sep 17 00:00:00 2001 From: Eric Biggers Date: Thu, 7 May 2020 00:59:03 -0700 Subject: [PATCH 83/86] f2fs: split f2fs_d_compare() from f2fs_match_name() Sharing f2fs_ci_compare() between comparing cached dentries (f2fs_d_compare()) and comparing on-disk dentries (f2fs_match_name()) doesn't work as well as intended, as these actions fundamentally differ in several ways (e.g. whether the task may sleep, whether the directory is stable, whether the casefolded name was precomputed, whether the dentry will need to be decrypted once we allow casefold+encrypt, etc.) Just make f2fs_d_compare() implement what it needs directly, and rework f2fs_ci_compare() to be specialized for f2fs_match_name(). Signed-off-by: Eric Biggers Signed-off-by: Jaegeuk Kim --- fs/f2fs/dir.c | 66 +++++++++++++++++++++++++------------------------- fs/f2fs/f2fs.h | 5 ---- 2 files changed, 33 insertions(+), 38 deletions(-) diff --git a/fs/f2fs/dir.c b/fs/f2fs/dir.c index 44bfc464df78..44eb12a00cd0 100644 --- a/fs/f2fs/dir.c +++ b/fs/f2fs/dir.c @@ -107,36 +107,28 @@ static struct f2fs_dir_entry *find_in_block(struct inode *dir, /* * Test whether a case-insensitive directory entry matches the filename * being searched for. - * - * Returns: 0 if the directory entry matches, more than 0 if it - * doesn't match or less than zero on error. */ -int f2fs_ci_compare(const struct inode *parent, const struct qstr *name, - const struct qstr *entry, bool quick) +static bool f2fs_match_ci_name(const struct inode *dir, const struct qstr *name, + const struct qstr *entry, bool quick) { - const struct f2fs_sb_info *sbi = F2FS_SB(parent->i_sb); + const struct f2fs_sb_info *sbi = F2FS_SB(dir->i_sb); const struct unicode_map *um = sbi->s_encoding; - int ret; + int res; if (quick) - ret = utf8_strncasecmp_folded(um, name, entry); + res = utf8_strncasecmp_folded(um, name, entry); else - ret = utf8_strncasecmp(um, name, entry); - - if (ret < 0) { - /* Handle invalid character sequence as either an error - * or as an opaque byte sequence. + res = utf8_strncasecmp(um, name, entry); + if (res < 0) { + /* + * In strict mode, ignore invalid names. In non-strict mode, + * fall back to treating them as opaque byte sequences. */ - if (f2fs_has_strict_mode(sbi)) - return -EINVAL; - - if (name->len != entry->len) - return 1; - - return !!memcmp(name->name, entry->name, name->len); + if (f2fs_has_strict_mode(sbi) || name->len != entry->len) + return false; + return !memcmp(name->name, entry->name, name->len); } - - return ret; + return res == 0; } static void f2fs_fname_setup_ci_filename(struct inode *dir, @@ -188,10 +180,10 @@ static inline bool f2fs_match_name(struct f2fs_dentry_ptr *d, if (cf_str->name) { struct qstr cf = {.name = cf_str->name, .len = cf_str->len}; - return !f2fs_ci_compare(parent, &cf, &entry, true); + return f2fs_match_ci_name(parent, &cf, &entry, true); } - return !f2fs_ci_compare(parent, fname->usr_fname, &entry, - false); + return f2fs_match_ci_name(parent, fname->usr_fname, &entry, + false); } #endif if (fscrypt_match_name(fname, d->filename[bit_pos], @@ -1080,17 +1072,25 @@ const struct file_operations f2fs_dir_operations = { static int f2fs_d_compare(const struct dentry *dentry, unsigned int len, const char *str, const struct qstr *name) { - struct qstr qstr = {.name = str, .len = len }; const struct dentry *parent = READ_ONCE(dentry->d_parent); - const struct inode *inode = READ_ONCE(parent->d_inode); + const struct inode *dir = READ_ONCE(parent->d_inode); + const struct f2fs_sb_info *sbi = F2FS_SB(dentry->d_sb); + struct qstr entry = QSTR_INIT(str, len); + int res; - if (!inode || !IS_CASEFOLDED(inode)) { - if (len != name->len) - return -1; - return memcmp(str, name->name, len); - } + if (!dir || !IS_CASEFOLDED(dir)) + goto fallback; - return f2fs_ci_compare(inode, name, &qstr, false); + res = utf8_strncasecmp(sbi->s_encoding, name, &entry); + if (res >= 0) + return res; + + if (f2fs_has_strict_mode(sbi)) + return -EINVAL; +fallback: + if (len != name->len) + return 1; + return !!memcmp(str, name->name, len); } static int f2fs_d_hash(const struct dentry *dentry, struct qstr *str) diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index 72016c29aa46..f91af69c5e48 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -3139,11 +3139,6 @@ int f2fs_update_extension_list(struct f2fs_sb_info *sbi, const char *name, bool hot, bool set); struct dentry *f2fs_get_parent(struct dentry *child); -extern int f2fs_ci_compare(const struct inode *parent, - const struct qstr *name, - const struct qstr *entry, - bool quick); - /* * dir.c */ From 6ccc7b00639e0156eefea4428d528f194ec9191a Mon Sep 17 00:00:00 2001 From: Eric Biggers Date: Thu, 7 May 2020 00:59:04 -0700 Subject: [PATCH 84/86] f2fs: rework filename handling Rework f2fs's handling of filenames to use a new 'struct f2fs_filename'. Similar to 'struct ext4_filename', this stores the usr_fname, disk_name, dirhash, crypto_buf, and casefolded name. Some of these names can be NULL in some cases. 'struct f2fs_filename' differs from 'struct fscrypt_name' mainly in that the casefolded name is included. For user-initiated directory operations like lookup() and create(), initialize the f2fs_filename by translating the corresponding fscrypt_name, then computing the dirhash and casefolded name if needed. This makes the dirhash and casefolded name be cached for each syscall, so we don't have to recompute them repeatedly. (Previously, f2fs computed the dirhash once per directory level, and the casefolded name once per directory block.) This improves performance. This rework also makes it much easier to correctly handle all combinations of normal, encrypted, casefolded, and encrypted+casefolded directories. (The fourth isn't supported yet but is being worked on.) The only other cases where an f2fs_filename gets initialized are for two filesystem-internal operations: (1) when converting an inline directory to a regular one, we grab the needed disk_name and hash from an existing f2fs_dir_entry; and (2) when roll-forward recovering a new dentry, we grab the needed disk_name from f2fs_inode::i_name and compute the hash. Signed-off-by: Eric Biggers Signed-off-by: Jaegeuk Kim --- fs/f2fs/dir.c | 310 +++++++++++++++++++++++++-------------------- fs/f2fs/f2fs.h | 78 +++++++++--- fs/f2fs/hash.c | 76 +++++------ fs/f2fs/inline.c | 45 ++++--- fs/f2fs/namei.c | 6 +- fs/f2fs/recovery.c | 51 ++++++-- 6 files changed, 331 insertions(+), 235 deletions(-) diff --git a/fs/f2fs/dir.c b/fs/f2fs/dir.c index 44eb12a00cd0..29f70f2295cc 100644 --- a/fs/f2fs/dir.c +++ b/fs/f2fs/dir.c @@ -70,6 +70,111 @@ unsigned char f2fs_get_de_type(struct f2fs_dir_entry *de) return DT_UNKNOWN; } +/* If @dir is casefolded, initialize @fname->cf_name from @fname->usr_fname. */ +int f2fs_init_casefolded_name(const struct inode *dir, + struct f2fs_filename *fname) +{ +#ifdef CONFIG_UNICODE + struct f2fs_sb_info *sbi = F2FS_SB(dir->i_sb); + + if (IS_CASEFOLDED(dir)) { + fname->cf_name.name = f2fs_kmalloc(sbi, F2FS_NAME_LEN, + GFP_NOFS); + if (!fname->cf_name.name) + return -ENOMEM; + fname->cf_name.len = utf8_casefold(sbi->s_encoding, + fname->usr_fname, + fname->cf_name.name, + F2FS_NAME_LEN); + if ((int)fname->cf_name.len <= 0) { + kfree(fname->cf_name.name); + fname->cf_name.name = NULL; + if (f2fs_has_strict_mode(sbi)) + return -EINVAL; + /* fall back to treating name as opaque byte sequence */ + } + } +#endif + return 0; +} + +static int __f2fs_setup_filename(const struct inode *dir, + const struct fscrypt_name *crypt_name, + struct f2fs_filename *fname) +{ + int err; + + memset(fname, 0, sizeof(*fname)); + + fname->usr_fname = crypt_name->usr_fname; + fname->disk_name = crypt_name->disk_name; +#ifdef CONFIG_FS_ENCRYPTION + fname->crypto_buf = crypt_name->crypto_buf; +#endif + if (crypt_name->is_ciphertext_name) { + /* hash was decoded from the no-key name */ + fname->hash = cpu_to_le32(crypt_name->hash); + } else { + err = f2fs_init_casefolded_name(dir, fname); + if (err) { + f2fs_free_filename(fname); + return err; + } + f2fs_hash_filename(dir, fname); + } + return 0; +} + +/* + * Prepare to search for @iname in @dir. This is similar to + * fscrypt_setup_filename(), but this also handles computing the casefolded name + * and the f2fs dirhash if needed, then packing all the information about this + * filename up into a 'struct f2fs_filename'. + */ +int f2fs_setup_filename(struct inode *dir, const struct qstr *iname, + int lookup, struct f2fs_filename *fname) +{ + struct fscrypt_name crypt_name; + int err; + + err = fscrypt_setup_filename(dir, iname, lookup, &crypt_name); + if (err) + return err; + + return __f2fs_setup_filename(dir, &crypt_name, fname); +} + +/* + * Prepare to look up @dentry in @dir. This is similar to + * fscrypt_prepare_lookup(), but this also handles computing the casefolded name + * and the f2fs dirhash if needed, then packing all the information about this + * filename up into a 'struct f2fs_filename'. + */ +int f2fs_prepare_lookup(struct inode *dir, struct dentry *dentry, + struct f2fs_filename *fname) +{ + struct fscrypt_name crypt_name; + int err; + + err = fscrypt_prepare_lookup(dir, dentry, &crypt_name); + if (err) + return err; + + return __f2fs_setup_filename(dir, &crypt_name, fname); +} + +void f2fs_free_filename(struct f2fs_filename *fname) +{ +#ifdef CONFIG_FS_ENCRYPTION + kfree(fname->crypto_buf.name); + fname->crypto_buf.name = NULL; +#endif +#ifdef CONFIG_UNICODE + kfree(fname->cf_name.name); + fname->cf_name.name = NULL; +#endif +} + static unsigned long dir_block_index(unsigned int level, int dir_level, unsigned int idx) { @@ -84,8 +189,7 @@ static unsigned long dir_block_index(unsigned int level, static struct f2fs_dir_entry *find_in_block(struct inode *dir, struct page *dentry_page, - struct fscrypt_name *fname, - f2fs_hash_t namehash, + const struct f2fs_filename *fname, int *max_slots, struct page **res_page) { @@ -96,7 +200,7 @@ static struct f2fs_dir_entry *find_in_block(struct inode *dir, dentry_blk = (struct f2fs_dentry_block *)page_address(dentry_page); make_dentry_ptr_block(dir, &d, dentry_blk); - de = f2fs_find_target_dentry(fname, namehash, max_slots, &d); + de = f2fs_find_target_dentry(&d, fname, max_slots); if (de) *res_page = dentry_page; @@ -109,102 +213,55 @@ static struct f2fs_dir_entry *find_in_block(struct inode *dir, * being searched for. */ static bool f2fs_match_ci_name(const struct inode *dir, const struct qstr *name, - const struct qstr *entry, bool quick) + const u8 *de_name, u32 de_name_len) { const struct f2fs_sb_info *sbi = F2FS_SB(dir->i_sb); const struct unicode_map *um = sbi->s_encoding; + struct qstr entry = QSTR_INIT(de_name, de_name_len); int res; - if (quick) - res = utf8_strncasecmp_folded(um, name, entry); - else - res = utf8_strncasecmp(um, name, entry); + res = utf8_strncasecmp_folded(um, name, &entry); if (res < 0) { /* * In strict mode, ignore invalid names. In non-strict mode, * fall back to treating them as opaque byte sequences. */ - if (f2fs_has_strict_mode(sbi) || name->len != entry->len) + if (f2fs_has_strict_mode(sbi) || name->len != entry.len) return false; - return !memcmp(name->name, entry->name, name->len); + return !memcmp(name->name, entry.name, name->len); } return res == 0; } +#endif /* CONFIG_UNICODE */ -static void f2fs_fname_setup_ci_filename(struct inode *dir, - const struct qstr *iname, - struct fscrypt_str *cf_name) +static inline bool f2fs_match_name(const struct inode *dir, + const struct f2fs_filename *fname, + const u8 *de_name, u32 de_name_len) { - struct f2fs_sb_info *sbi = F2FS_I_SB(dir); - - if (!IS_CASEFOLDED(dir)) { - cf_name->name = NULL; - return; - } - - cf_name->name = f2fs_kmalloc(sbi, F2FS_NAME_LEN, GFP_NOFS); - if (!cf_name->name) - return; - - cf_name->len = utf8_casefold(sbi->s_encoding, - iname, cf_name->name, - F2FS_NAME_LEN); - if ((int)cf_name->len <= 0) { - kvfree(cf_name->name); - cf_name->name = NULL; - } -} -#endif - -static inline bool f2fs_match_name(struct f2fs_dentry_ptr *d, - struct f2fs_dir_entry *de, - struct fscrypt_name *fname, - struct fscrypt_str *cf_str, - unsigned long bit_pos, - f2fs_hash_t namehash) -{ -#ifdef CONFIG_UNICODE - struct inode *parent = d->inode; - struct f2fs_sb_info *sbi = F2FS_I_SB(parent); - struct qstr entry; -#endif - - if (de->hash_code != namehash) - return false; + struct fscrypt_name f; #ifdef CONFIG_UNICODE - entry.name = d->filename[bit_pos]; - entry.len = de->name_len; + if (fname->cf_name.name) { + struct qstr cf = FSTR_TO_QSTR(&fname->cf_name); - if (sbi->s_encoding && IS_CASEFOLDED(parent)) { - if (cf_str->name) { - struct qstr cf = {.name = cf_str->name, - .len = cf_str->len}; - return f2fs_match_ci_name(parent, &cf, &entry, true); - } - return f2fs_match_ci_name(parent, fname->usr_fname, &entry, - false); + return f2fs_match_ci_name(dir, &cf, de_name, de_name_len); } #endif - if (fscrypt_match_name(fname, d->filename[bit_pos], - le16_to_cpu(de->name_len))) - return true; - return false; + f.usr_fname = fname->usr_fname; + f.disk_name = fname->disk_name; +#ifdef CONFIG_FS_ENCRYPTION + f.crypto_buf = fname->crypto_buf; +#endif + return fscrypt_match_name(&f, de_name, de_name_len); } -struct f2fs_dir_entry *f2fs_find_target_dentry(struct fscrypt_name *fname, - f2fs_hash_t namehash, int *max_slots, - struct f2fs_dentry_ptr *d) +struct f2fs_dir_entry *f2fs_find_target_dentry(const struct f2fs_dentry_ptr *d, + const struct f2fs_filename *fname, int *max_slots) { struct f2fs_dir_entry *de; - struct fscrypt_str cf_str = { .name = NULL, .len = 0 }; unsigned long bit_pos = 0; int max_len = 0; -#ifdef CONFIG_UNICODE - f2fs_fname_setup_ci_filename(d->inode, fname->usr_fname, &cf_str); -#endif - if (max_slots) *max_slots = 0; while (bit_pos < d->max) { @@ -221,7 +278,9 @@ struct f2fs_dir_entry *f2fs_find_target_dentry(struct fscrypt_name *fname, continue; } - if (f2fs_match_name(d, de, fname, &cf_str, bit_pos, namehash)) + if (de->hash_code == fname->hash && + f2fs_match_name(d->inode, fname, d->filename[bit_pos], + le16_to_cpu(de->name_len))) goto found; if (max_slots && max_len > *max_slots) @@ -235,33 +294,27 @@ struct f2fs_dir_entry *f2fs_find_target_dentry(struct fscrypt_name *fname, found: if (max_slots && max_len > *max_slots) *max_slots = max_len; - -#ifdef CONFIG_UNICODE - kvfree(cf_str.name); -#endif return de; } static struct f2fs_dir_entry *find_in_level(struct inode *dir, unsigned int level, - struct fscrypt_name *fname, + const struct f2fs_filename *fname, struct page **res_page) { - struct qstr name = FSTR_TO_QSTR(&fname->disk_name); - int s = GET_DENTRY_SLOTS(name.len); + int s = GET_DENTRY_SLOTS(fname->disk_name.len); unsigned int nbucket, nblock; unsigned int bidx, end_block; struct page *dentry_page; struct f2fs_dir_entry *de = NULL; bool room = false; int max_slots; - f2fs_hash_t namehash = f2fs_dentry_hash(dir, &name, fname); nbucket = dir_buckets(level, F2FS_I(dir)->i_dir_level); nblock = bucket_blocks(level); bidx = dir_block_index(level, F2FS_I(dir)->i_dir_level, - le32_to_cpu(namehash) % nbucket); + le32_to_cpu(fname->hash) % nbucket); end_block = bidx + nblock; for (; bidx < end_block; bidx++) { @@ -277,8 +330,8 @@ static struct f2fs_dir_entry *find_in_level(struct inode *dir, } } - de = find_in_block(dir, dentry_page, fname, namehash, - &max_slots, res_page); + de = find_in_block(dir, dentry_page, fname, &max_slots, + res_page); if (de) break; @@ -287,8 +340,8 @@ static struct f2fs_dir_entry *find_in_level(struct inode *dir, f2fs_put_page(dentry_page, 0); } - if (!de && room && F2FS_I(dir)->chash != namehash) { - F2FS_I(dir)->chash = namehash; + if (!de && room && F2FS_I(dir)->chash != fname->hash) { + F2FS_I(dir)->chash = fname->hash; F2FS_I(dir)->clevel = level; } @@ -296,7 +349,8 @@ static struct f2fs_dir_entry *find_in_level(struct inode *dir, } struct f2fs_dir_entry *__f2fs_find_entry(struct inode *dir, - struct fscrypt_name *fname, struct page **res_page) + const struct f2fs_filename *fname, + struct page **res_page) { unsigned long npages = dir_blocks(dir); struct f2fs_dir_entry *de = NULL; @@ -345,18 +399,10 @@ struct f2fs_dir_entry *f2fs_find_entry(struct inode *dir, const struct qstr *child, struct page **res_page) { struct f2fs_dir_entry *de = NULL; - struct fscrypt_name fname; + struct f2fs_filename fname; int err; -#ifdef CONFIG_UNICODE - if (f2fs_has_strict_mode(F2FS_I_SB(dir)) && IS_CASEFOLDED(dir) && - utf8_validate(F2FS_I_SB(dir)->s_encoding, child)) { - *res_page = ERR_PTR(-EINVAL); - return NULL; - } -#endif - - err = fscrypt_setup_filename(dir, child, 1, &fname); + err = f2fs_setup_filename(dir, child, 1, &fname); if (err) { if (err == -ENOENT) *res_page = NULL; @@ -367,7 +413,7 @@ struct f2fs_dir_entry *f2fs_find_entry(struct inode *dir, de = __f2fs_find_entry(dir, &fname, res_page); - fscrypt_free_filename(&fname); + f2fs_free_filename(&fname); return de; } @@ -408,7 +454,8 @@ void f2fs_set_link(struct inode *dir, struct f2fs_dir_entry *de, f2fs_put_page(page, 1); } -static void init_dent_inode(const struct qstr *name, struct page *ipage) +static void init_dent_inode(const struct f2fs_filename *fname, + struct page *ipage) { struct f2fs_inode *ri; @@ -416,16 +463,16 @@ static void init_dent_inode(const struct qstr *name, struct page *ipage) /* copy name info. to this inode page */ ri = F2FS_INODE(ipage); - ri->i_namelen = cpu_to_le32(name->len); - memcpy(ri->i_name, name->name, name->len); + ri->i_namelen = cpu_to_le32(fname->disk_name.len); + memcpy(ri->i_name, fname->disk_name.name, fname->disk_name.len); set_page_dirty(ipage); } void f2fs_do_make_empty_dir(struct inode *inode, struct inode *parent, struct f2fs_dentry_ptr *d) { - struct qstr dot = QSTR_INIT(".", 1); - struct qstr dotdot = QSTR_INIT("..", 2); + struct fscrypt_str dot = FSTR_INIT(".", 1); + struct fscrypt_str dotdot = FSTR_INIT("..", 2); /* update dirent of "." */ f2fs_update_dentry(inode->i_ino, inode->i_mode, d, &dot, 0, 0); @@ -459,8 +506,7 @@ static int make_empty_dir(struct inode *inode, } struct page *f2fs_init_inode_metadata(struct inode *inode, struct inode *dir, - const struct qstr *new_name, const struct qstr *orig_name, - struct page *dpage) + const struct f2fs_filename *fname, struct page *dpage) { struct page *page; int err; @@ -485,7 +531,8 @@ struct page *f2fs_init_inode_metadata(struct inode *inode, struct inode *dir, if (err) goto put_error; - err = f2fs_init_security(inode, dir, orig_name, page); + err = f2fs_init_security(inode, dir, + fname ? fname->usr_fname : NULL, page); if (err) goto put_error; @@ -500,8 +547,8 @@ struct page *f2fs_init_inode_metadata(struct inode *inode, struct inode *dir, return page; } - if (new_name) { - init_dent_inode(new_name, page); + if (fname) { + init_dent_inode(fname, page); if (IS_ENCRYPTED(dir)) file_set_enc_name(inode); } @@ -569,11 +616,11 @@ next: } bool f2fs_has_enough_room(struct inode *dir, struct page *ipage, - struct fscrypt_name *fname) + const struct f2fs_filename *fname) { struct f2fs_dentry_ptr d; unsigned int bit_pos; - int slots = GET_DENTRY_SLOTS(fname_len(fname)); + int slots = GET_DENTRY_SLOTS(fname->disk_name.len); make_dentry_ptr_inline(dir, &d, inline_data_addr(dir, ipage)); @@ -583,8 +630,8 @@ bool f2fs_has_enough_room(struct inode *dir, struct page *ipage, } void f2fs_update_dentry(nid_t ino, umode_t mode, struct f2fs_dentry_ptr *d, - const struct qstr *name, f2fs_hash_t name_hash, - unsigned int bit_pos) + const struct fscrypt_str *name, f2fs_hash_t name_hash, + unsigned int bit_pos) { struct f2fs_dir_entry *de; int slots = GET_DENTRY_SLOTS(name->len); @@ -604,15 +651,13 @@ void f2fs_update_dentry(nid_t ino, umode_t mode, struct f2fs_dentry_ptr *d, } } -int f2fs_add_regular_entry(struct inode *dir, const struct qstr *new_name, - const struct qstr *orig_name, - struct inode *inode, nid_t ino, umode_t mode) +int f2fs_add_regular_entry(struct inode *dir, const struct f2fs_filename *fname, + struct inode *inode, nid_t ino, umode_t mode) { unsigned int bit_pos; unsigned int level; unsigned int current_depth; unsigned long bidx, block; - f2fs_hash_t dentry_hash; unsigned int nbucket, nblock; struct page *dentry_page = NULL; struct f2fs_dentry_block *dentry_blk = NULL; @@ -621,11 +666,10 @@ int f2fs_add_regular_entry(struct inode *dir, const struct qstr *new_name, int slots, err = 0; level = 0; - slots = GET_DENTRY_SLOTS(new_name->len); - dentry_hash = f2fs_dentry_hash(dir, new_name, NULL); + slots = GET_DENTRY_SLOTS(fname->disk_name.len); current_depth = F2FS_I(dir)->i_current_depth; - if (F2FS_I(dir)->chash == dentry_hash) { + if (F2FS_I(dir)->chash == fname->hash) { level = F2FS_I(dir)->clevel; F2FS_I(dir)->chash = 0; } @@ -647,7 +691,7 @@ start: nblock = bucket_blocks(level); bidx = dir_block_index(level, F2FS_I(dir)->i_dir_level, - (le32_to_cpu(dentry_hash) % nbucket)); + (le32_to_cpu(fname->hash) % nbucket)); for (block = bidx; block <= (bidx + nblock - 1); block++) { dentry_page = f2fs_get_new_data_page(dir, NULL, block, true); @@ -671,8 +715,7 @@ add_dentry: if (inode) { down_write(&F2FS_I(inode)->i_sem); - page = f2fs_init_inode_metadata(inode, dir, new_name, - orig_name, NULL); + page = f2fs_init_inode_metadata(inode, dir, fname, NULL); if (IS_ERR(page)) { err = PTR_ERR(page); goto fail; @@ -680,7 +723,8 @@ add_dentry: } make_dentry_ptr_block(NULL, &d, dentry_blk); - f2fs_update_dentry(ino, mode, &d, new_name, dentry_hash, bit_pos); + f2fs_update_dentry(ino, mode, &d, &fname->disk_name, fname->hash, + bit_pos); set_page_dirty(dentry_page); @@ -704,21 +748,15 @@ fail: return err; } -int f2fs_add_dentry(struct inode *dir, struct fscrypt_name *fname, - struct inode *inode, nid_t ino, umode_t mode) +int f2fs_add_dentry(struct inode *dir, const struct f2fs_filename *fname, + struct inode *inode, nid_t ino, umode_t mode) { - struct qstr new_name; int err = -EAGAIN; - new_name.name = fname_name(fname); - new_name.len = fname_len(fname); - if (f2fs_has_inline_dentry(dir)) - err = f2fs_add_inline_entry(dir, &new_name, fname->usr_fname, - inode, ino, mode); + err = f2fs_add_inline_entry(dir, fname, inode, ino, mode); if (err == -EAGAIN) - err = f2fs_add_regular_entry(dir, &new_name, fname->usr_fname, - inode, ino, mode); + err = f2fs_add_regular_entry(dir, fname, inode, ino, mode); f2fs_update_time(F2FS_I_SB(dir), REQ_TIME); return err; @@ -731,12 +769,12 @@ int f2fs_add_dentry(struct inode *dir, struct fscrypt_name *fname, int f2fs_do_add_link(struct inode *dir, const struct qstr *name, struct inode *inode, nid_t ino, umode_t mode) { - struct fscrypt_name fname; + struct f2fs_filename fname; struct page *page = NULL; struct f2fs_dir_entry *de = NULL; int err; - err = fscrypt_setup_filename(dir, name, 0, &fname); + err = f2fs_setup_filename(dir, name, 0, &fname); if (err) return err; @@ -759,7 +797,7 @@ int f2fs_do_add_link(struct inode *dir, const struct qstr *name, } else { err = f2fs_add_dentry(dir, &fname, inode, ino, mode); } - fscrypt_free_filename(&fname); + f2fs_free_filename(&fname); return err; } @@ -769,7 +807,7 @@ int f2fs_do_tmpfile(struct inode *inode, struct inode *dir) int err = 0; down_write(&F2FS_I(inode)->i_sem); - page = f2fs_init_inode_metadata(inode, dir, NULL, NULL, NULL); + page = f2fs_init_inode_metadata(inode, dir, NULL, NULL); if (IS_ERR(page)) { err = PTR_ERR(page); goto fail; diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index f91af69c5e48..afe9ba41ac22 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -505,6 +505,42 @@ static inline int get_inline_xattr_addrs(struct inode *inode); * For INODE and NODE manager */ /* for directory operations */ + +struct f2fs_filename { + /* + * The filename the user specified. This is NULL for some + * filesystem-internal operations, e.g. converting an inline directory + * to a non-inline one, or roll-forward recovering an encrypted dentry. + */ + const struct qstr *usr_fname; + + /* + * The on-disk filename. For encrypted directories, this is encrypted. + * This may be NULL for lookups in an encrypted dir without the key. + */ + struct fscrypt_str disk_name; + + /* The dirhash of this filename */ + f2fs_hash_t hash; + +#ifdef CONFIG_FS_ENCRYPTION + /* + * For lookups in encrypted directories: either the buffer backing + * disk_name, or a buffer that holds the decoded no-key name. + */ + struct fscrypt_str crypto_buf; +#endif +#ifdef CONFIG_UNICODE + /* + * For casefolded directories: the casefolded name, but it's left NULL + * if the original name is not valid Unicode or if the filesystem is + * doing an internal operation where usr_fname is also NULL. In these + * cases we fall back to treating the name as an opaque byte sequence. + */ + struct fscrypt_str cf_name; +#endif +}; + struct f2fs_dentry_ptr { struct inode *inode; void *bitmap; @@ -2919,12 +2955,12 @@ static inline bool f2fs_cp_error(struct f2fs_sb_info *sbi) return is_set_ckpt_flags(sbi, CP_ERROR_FLAG); } -static inline bool is_dot_dotdot(const struct qstr *str) +static inline bool is_dot_dotdot(const u8 *name, size_t len) { - if (str->len == 1 && str->name[0] == '.') + if (len == 1 && name[0] == '.') return true; - if (str->len == 2 && str->name[0] == '.' && str->name[1] == '.') + if (len == 2 && name[0] == '.' && name[1] == '.') return true; return false; @@ -3143,22 +3179,28 @@ struct dentry *f2fs_get_parent(struct dentry *child); * dir.c */ unsigned char f2fs_get_de_type(struct f2fs_dir_entry *de); -struct f2fs_dir_entry *f2fs_find_target_dentry(struct fscrypt_name *fname, - f2fs_hash_t namehash, int *max_slots, - struct f2fs_dentry_ptr *d); +int f2fs_init_casefolded_name(const struct inode *dir, + struct f2fs_filename *fname); +int f2fs_setup_filename(struct inode *dir, const struct qstr *iname, + int lookup, struct f2fs_filename *fname); +int f2fs_prepare_lookup(struct inode *dir, struct dentry *dentry, + struct f2fs_filename *fname); +void f2fs_free_filename(struct f2fs_filename *fname); +struct f2fs_dir_entry *f2fs_find_target_dentry(const struct f2fs_dentry_ptr *d, + const struct f2fs_filename *fname, int *max_slots); int f2fs_fill_dentries(struct dir_context *ctx, struct f2fs_dentry_ptr *d, unsigned int start_pos, struct fscrypt_str *fstr); void f2fs_do_make_empty_dir(struct inode *inode, struct inode *parent, struct f2fs_dentry_ptr *d); struct page *f2fs_init_inode_metadata(struct inode *inode, struct inode *dir, - const struct qstr *new_name, - const struct qstr *orig_name, struct page *dpage); + const struct f2fs_filename *fname, struct page *dpage); void f2fs_update_parent_metadata(struct inode *dir, struct inode *inode, unsigned int current_depth); int f2fs_room_for_filename(const void *bitmap, int slots, int max_slots); void f2fs_drop_nlink(struct inode *dir, struct inode *inode); struct f2fs_dir_entry *__f2fs_find_entry(struct inode *dir, - struct fscrypt_name *fname, struct page **res_page); + const struct f2fs_filename *fname, + struct page **res_page); struct f2fs_dir_entry *f2fs_find_entry(struct inode *dir, const struct qstr *child, struct page **res_page); struct f2fs_dir_entry *f2fs_parent_dir(struct inode *dir, struct page **p); @@ -3167,14 +3209,13 @@ ino_t f2fs_inode_by_name(struct inode *dir, const struct qstr *qstr, void f2fs_set_link(struct inode *dir, struct f2fs_dir_entry *de, struct page *page, struct inode *inode); bool f2fs_has_enough_room(struct inode *dir, struct page *ipage, - struct fscrypt_name *fname); + const struct f2fs_filename *fname); void f2fs_update_dentry(nid_t ino, umode_t mode, struct f2fs_dentry_ptr *d, - const struct qstr *name, f2fs_hash_t name_hash, + const struct fscrypt_str *name, f2fs_hash_t name_hash, unsigned int bit_pos); -int f2fs_add_regular_entry(struct inode *dir, const struct qstr *new_name, - const struct qstr *orig_name, +int f2fs_add_regular_entry(struct inode *dir, const struct f2fs_filename *fname, struct inode *inode, nid_t ino, umode_t mode); -int f2fs_add_dentry(struct inode *dir, struct fscrypt_name *fname, +int f2fs_add_dentry(struct inode *dir, const struct f2fs_filename *fname, struct inode *inode, nid_t ino, umode_t mode); int f2fs_do_add_link(struct inode *dir, const struct qstr *name, struct inode *inode, nid_t ino, umode_t mode); @@ -3204,8 +3245,7 @@ int f2fs_sanity_check_ckpt(struct f2fs_sb_info *sbi); /* * hash.c */ -f2fs_hash_t f2fs_dentry_hash(const struct inode *dir, - const struct qstr *name_info, struct fscrypt_name *fname); +void f2fs_hash_filename(const struct inode *dir, struct f2fs_filename *fname); /* * node.c @@ -3716,11 +3756,11 @@ int f2fs_try_convert_inline_dir(struct inode *dir, struct dentry *dentry); int f2fs_write_inline_data(struct inode *inode, struct page *page); bool f2fs_recover_inline_data(struct inode *inode, struct page *npage); struct f2fs_dir_entry *f2fs_find_in_inline_dir(struct inode *dir, - struct fscrypt_name *fname, struct page **res_page); + const struct f2fs_filename *fname, + struct page **res_page); int f2fs_make_empty_inline_dir(struct inode *inode, struct inode *parent, struct page *ipage); -int f2fs_add_inline_entry(struct inode *dir, const struct qstr *new_name, - const struct qstr *orig_name, +int f2fs_add_inline_entry(struct inode *dir, const struct f2fs_filename *fname, struct inode *inode, nid_t ino, umode_t mode); void f2fs_delete_inline_entry(struct f2fs_dir_entry *dentry, struct page *page, struct inode *dir, diff --git a/fs/f2fs/hash.c b/fs/f2fs/hash.c index 5bc4dcd8fc03..e5997919472d 100644 --- a/fs/f2fs/hash.c +++ b/fs/f2fs/hash.c @@ -68,22 +68,9 @@ static void str2hashbuf(const unsigned char *msg, size_t len, *buf++ = pad; } -static f2fs_hash_t __f2fs_dentry_hash(const struct qstr *name_info, - struct fscrypt_name *fname) +static u32 TEA_hash_name(const u8 *p, size_t len) { - __u32 hash; - f2fs_hash_t f2fs_hash; - const unsigned char *p; __u32 in[8], buf[4]; - const unsigned char *name = name_info->name; - size_t len = name_info->len; - - /* encrypted bigname case */ - if (fname && !fname->disk_name.name) - return cpu_to_le32(fname->hash); - - if (is_dot_dotdot(name_info)) - return 0; /* Initialize the default seed for the hash checksum functions */ buf[0] = 0x67452301; @@ -91,7 +78,6 @@ static f2fs_hash_t __f2fs_dentry_hash(const struct qstr *name_info, buf[2] = 0x98badcfe; buf[3] = 0x10325476; - p = name; while (1) { str2hashbuf(p, len, in, 4); TEA_transform(buf, in); @@ -100,41 +86,43 @@ static f2fs_hash_t __f2fs_dentry_hash(const struct qstr *name_info, break; len -= 16; } - hash = buf[0]; - f2fs_hash = cpu_to_le32(hash & ~F2FS_HASH_COL_BIT); - return f2fs_hash; + return buf[0] & ~F2FS_HASH_COL_BIT; } -f2fs_hash_t f2fs_dentry_hash(const struct inode *dir, - const struct qstr *name_info, struct fscrypt_name *fname) +/* + * Compute @fname->hash. For all directories, @fname->disk_name must be set. + * For casefolded directories, @fname->usr_fname must be set, and also + * @fname->cf_name if the filename is valid Unicode. + */ +void f2fs_hash_filename(const struct inode *dir, struct f2fs_filename *fname) { -#ifdef CONFIG_UNICODE - struct f2fs_sb_info *sbi = F2FS_SB(dir->i_sb); - const struct unicode_map *um = sbi->s_encoding; - int r, dlen; - unsigned char *buff; - struct qstr folded; + const u8 *name = fname->disk_name.name; + size_t len = fname->disk_name.len; - if (!name_info->len || !IS_CASEFOLDED(dir)) - goto opaque_seq; + WARN_ON_ONCE(!name); - buff = f2fs_kzalloc(sbi, sizeof(char) * PATH_MAX, GFP_KERNEL); - if (!buff) - return -ENOMEM; - - dlen = utf8_casefold(um, name_info, buff, PATH_MAX); - if (dlen < 0) { - kvfree(buff); - goto opaque_seq; + if (is_dot_dotdot(name, len)) { + fname->hash = 0; + return; } - folded.name = buff; - folded.len = dlen; - r = __f2fs_dentry_hash(&folded, fname); - kvfree(buff); - return r; - -opaque_seq: +#ifdef CONFIG_UNICODE + if (IS_CASEFOLDED(dir)) { + /* + * If the casefolded name is provided, hash it instead of the + * on-disk name. If the casefolded name is *not* provided, that + * should only be because the name wasn't valid Unicode, so fall + * back to treating the name as an opaque byte sequence. + */ + WARN_ON_ONCE(!fname->usr_fname->name); + if (fname->cf_name.name) { + name = fname->cf_name.name; + len = fname->cf_name.len; + } else { + name = fname->usr_fname->name; + len = fname->usr_fname->len; + } + } #endif - return __f2fs_dentry_hash(name_info, fname); + fname->hash = cpu_to_le32(TEA_hash_name(name, len)); } diff --git a/fs/f2fs/inline.c b/fs/f2fs/inline.c index 59a4b7ff11e1..751a012ff840 100644 --- a/fs/f2fs/inline.c +++ b/fs/f2fs/inline.c @@ -305,15 +305,14 @@ process_inline: } struct f2fs_dir_entry *f2fs_find_in_inline_dir(struct inode *dir, - struct fscrypt_name *fname, struct page **res_page) + const struct f2fs_filename *fname, + struct page **res_page) { struct f2fs_sb_info *sbi = F2FS_SB(dir->i_sb); - struct qstr name = FSTR_TO_QSTR(&fname->disk_name); struct f2fs_dir_entry *de; struct f2fs_dentry_ptr d; struct page *ipage; void *inline_dentry; - f2fs_hash_t namehash; ipage = f2fs_get_node_page(sbi, dir->i_ino); if (IS_ERR(ipage)) { @@ -321,12 +320,10 @@ struct f2fs_dir_entry *f2fs_find_in_inline_dir(struct inode *dir, return NULL; } - namehash = f2fs_dentry_hash(dir, &name, fname); - inline_dentry = inline_data_addr(dir, ipage); make_dentry_ptr_inline(dir, &d, inline_dentry); - de = f2fs_find_target_dentry(fname, namehash, NULL, &d); + de = f2fs_find_target_dentry(&d, fname, NULL); unlock_page(ipage); if (de) *res_page = ipage; @@ -443,7 +440,7 @@ static int f2fs_add_inline_entries(struct inode *dir, void *inline_dentry) while (bit_pos < d.max) { struct f2fs_dir_entry *de; - struct qstr new_name; + struct f2fs_filename fname; nid_t ino; umode_t fake_mode; @@ -459,14 +456,19 @@ static int f2fs_add_inline_entries(struct inode *dir, void *inline_dentry) continue; } - new_name.name = d.filename[bit_pos]; - new_name.len = le16_to_cpu(de->name_len); + /* + * We only need the disk_name and hash to move the dentry. + * We don't need the original or casefolded filenames. + */ + memset(&fname, 0, sizeof(fname)); + fname.disk_name.name = d.filename[bit_pos]; + fname.disk_name.len = le16_to_cpu(de->name_len); + fname.hash = de->hash_code; ino = le32_to_cpu(de->ino); fake_mode = f2fs_get_de_type(de) << S_SHIFT; - err = f2fs_add_regular_entry(dir, &new_name, NULL, NULL, - ino, fake_mode); + err = f2fs_add_regular_entry(dir, &fname, NULL, ino, fake_mode); if (err) goto punch_dentry_pages; @@ -543,7 +545,7 @@ int f2fs_try_convert_inline_dir(struct inode *dir, struct dentry *dentry) { struct f2fs_sb_info *sbi = F2FS_I_SB(dir); struct page *ipage; - struct fscrypt_name fname; + struct f2fs_filename fname; void *inline_dentry = NULL; int err = 0; @@ -552,7 +554,7 @@ int f2fs_try_convert_inline_dir(struct inode *dir, struct dentry *dentry) f2fs_lock_op(sbi); - err = fscrypt_setup_filename(dir, &dentry->d_name, 0, &fname); + err = f2fs_setup_filename(dir, &dentry->d_name, 0, &fname); if (err) goto out; @@ -573,23 +575,21 @@ int f2fs_try_convert_inline_dir(struct inode *dir, struct dentry *dentry) if (!err) f2fs_put_page(ipage, 1); out_fname: - fscrypt_free_filename(&fname); + f2fs_free_filename(&fname); out: f2fs_unlock_op(sbi); return err; } -int f2fs_add_inline_entry(struct inode *dir, const struct qstr *new_name, - const struct qstr *orig_name, - struct inode *inode, nid_t ino, umode_t mode) +int f2fs_add_inline_entry(struct inode *dir, const struct f2fs_filename *fname, + struct inode *inode, nid_t ino, umode_t mode) { struct f2fs_sb_info *sbi = F2FS_I_SB(dir); struct page *ipage; unsigned int bit_pos; - f2fs_hash_t name_hash; void *inline_dentry = NULL; struct f2fs_dentry_ptr d; - int slots = GET_DENTRY_SLOTS(new_name->len); + int slots = GET_DENTRY_SLOTS(fname->disk_name.len); struct page *page = NULL; int err = 0; @@ -611,8 +611,7 @@ int f2fs_add_inline_entry(struct inode *dir, const struct qstr *new_name, if (inode) { down_write(&F2FS_I(inode)->i_sem); - page = f2fs_init_inode_metadata(inode, dir, new_name, - orig_name, ipage); + page = f2fs_init_inode_metadata(inode, dir, fname, ipage); if (IS_ERR(page)) { err = PTR_ERR(page); goto fail; @@ -621,8 +620,8 @@ int f2fs_add_inline_entry(struct inode *dir, const struct qstr *new_name, f2fs_wait_on_page_writeback(ipage, NODE, true, true); - name_hash = f2fs_dentry_hash(dir, new_name, NULL); - f2fs_update_dentry(ino, mode, &d, new_name, name_hash, bit_pos); + f2fs_update_dentry(ino, mode, &d, &fname->disk_name, fname->hash, + bit_pos); set_page_dirty(ipage); diff --git a/fs/f2fs/namei.c b/fs/f2fs/namei.c index f54119da2217..58db1de7ca94 100644 --- a/fs/f2fs/namei.c +++ b/fs/f2fs/namei.c @@ -482,7 +482,7 @@ static struct dentry *f2fs_lookup(struct inode *dir, struct dentry *dentry, nid_t ino = -1; int err = 0; unsigned int root_ino = F2FS_ROOT_INO(F2FS_I_SB(dir)); - struct fscrypt_name fname; + struct f2fs_filename fname; trace_f2fs_lookup_start(dir, dentry, flags); @@ -491,13 +491,13 @@ static struct dentry *f2fs_lookup(struct inode *dir, struct dentry *dentry, goto out; } - err = fscrypt_prepare_lookup(dir, dentry, &fname); + err = f2fs_prepare_lookup(dir, dentry, &fname); if (err == -ENOENT) goto out_splice; if (err) goto out; de = __f2fs_find_entry(dir, &fname, &page); - fscrypt_free_filename(&fname); + f2fs_free_filename(&fname); if (!de) { if (IS_ERR(page)) { diff --git a/fs/f2fs/recovery.c b/fs/f2fs/recovery.c index dd804c07eeb0..ae5310f02e7f 100644 --- a/fs/f2fs/recovery.c +++ b/fs/f2fs/recovery.c @@ -107,13 +107,51 @@ static void del_fsync_inode(struct fsync_inode_entry *entry, int drop) kmem_cache_free(fsync_entry_slab, entry); } +static int init_recovered_filename(const struct inode *dir, + struct f2fs_inode *raw_inode, + struct f2fs_filename *fname, + struct qstr *usr_fname) +{ + int err; + + memset(fname, 0, sizeof(*fname)); + fname->disk_name.len = le32_to_cpu(raw_inode->i_namelen); + fname->disk_name.name = raw_inode->i_name; + + if (WARN_ON(fname->disk_name.len > F2FS_NAME_LEN)) + return -ENAMETOOLONG; + + if (!IS_ENCRYPTED(dir)) { + usr_fname->name = fname->disk_name.name; + usr_fname->len = fname->disk_name.len; + fname->usr_fname = usr_fname; + } + + /* Compute the hash of the filename */ + if (IS_CASEFOLDED(dir)) { + err = f2fs_init_casefolded_name(dir, fname); + if (err) + return err; + f2fs_hash_filename(dir, fname); +#ifdef CONFIG_UNICODE + /* Case-sensitive match is fine for recovery */ + kfree(fname->cf_name.name); + fname->cf_name.name = NULL; +#endif + } else { + f2fs_hash_filename(dir, fname); + } + return 0; +} + static int recover_dentry(struct inode *inode, struct page *ipage, struct list_head *dir_list) { struct f2fs_inode *raw_inode = F2FS_INODE(ipage); nid_t pino = le32_to_cpu(raw_inode->i_pino); struct f2fs_dir_entry *de; - struct fscrypt_name fname; + struct f2fs_filename fname; + struct qstr usr_fname; struct page *page; struct inode *dir, *einode; struct fsync_inode_entry *entry; @@ -132,16 +170,9 @@ static int recover_dentry(struct inode *inode, struct page *ipage, } dir = entry->inode; - - memset(&fname, 0, sizeof(struct fscrypt_name)); - fname.disk_name.len = le32_to_cpu(raw_inode->i_namelen); - fname.disk_name.name = raw_inode->i_name; - - if (unlikely(fname.disk_name.len > F2FS_NAME_LEN)) { - WARN_ON(1); - err = -ENAMETOOLONG; + err = init_recovered_filename(dir, raw_inode, &fname, &usr_fname); + if (err) goto out; - } retry: de = __f2fs_find_entry(dir, &fname, &page); if (de && inode->i_ino == le32_to_cpu(de->ino)) From d27708574208b5501946bc2a1b978399525063fd Mon Sep 17 00:00:00 2001 From: Jaegeuk Kim Date: Fri, 15 May 2020 17:20:50 -0700 Subject: [PATCH 85/86] f2fs: fix checkpoint=disable:%u%% When parsing the mount option, we don't have sbi->user_block_count. Should do it after getting it. Cc: Reviewed-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/f2fs.h | 1 + fs/f2fs/super.c | 25 +++++++++++++++++++------ 2 files changed, 20 insertions(+), 6 deletions(-) diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index afe9ba41ac22..0ee1fa56f048 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -138,6 +138,7 @@ struct f2fs_mount_info { int fs_mode; /* fs mode: LFS or ADAPTIVE */ int bggc_mode; /* bggc mode: off, on or sync */ bool test_dummy_encryption; /* test dummy encryption */ + block_t unusable_cap_perc; /* percentage for cap */ block_t unusable_cap; /* Amount of space allowed to be * unusable when disabling checkpoint */ diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c index 2a29964729f0..bfc55bf8bc93 100644 --- a/fs/f2fs/super.c +++ b/fs/f2fs/super.c @@ -283,6 +283,22 @@ static inline void limit_reserve_root(struct f2fs_sb_info *sbi) F2FS_OPTION(sbi).s_resgid)); } +static inline void adjust_unusable_cap_perc(struct f2fs_sb_info *sbi) +{ + if (!F2FS_OPTION(sbi).unusable_cap_perc) + return; + + if (F2FS_OPTION(sbi).unusable_cap_perc == 100) + F2FS_OPTION(sbi).unusable_cap = sbi->user_block_count; + else + F2FS_OPTION(sbi).unusable_cap = (sbi->user_block_count / 100) * + F2FS_OPTION(sbi).unusable_cap_perc; + + f2fs_info(sbi, "Adjust unusable cap for checkpoint=disable = %u / %u%%", + F2FS_OPTION(sbi).unusable_cap, + F2FS_OPTION(sbi).unusable_cap_perc); +} + static void init_once(void *foo) { struct f2fs_inode_info *fi = (struct f2fs_inode_info *) foo; @@ -794,12 +810,7 @@ static int parse_options(struct super_block *sb, char *options) return -EINVAL; if (arg < 0 || arg > 100) return -EINVAL; - if (arg == 100) - F2FS_OPTION(sbi).unusable_cap = - sbi->user_block_count; - else - F2FS_OPTION(sbi).unusable_cap = - (sbi->user_block_count / 100) * arg; + F2FS_OPTION(sbi).unusable_cap_perc = arg; set_opt(sbi, DISABLE_CHECKPOINT); break; case Opt_checkpoint_disable_cap: @@ -1844,6 +1855,7 @@ skip: (test_opt(sbi, POSIX_ACL) ? SB_POSIXACL : 0); limit_reserve_root(sbi); + adjust_unusable_cap_perc(sbi); *flags = (*flags & ~SB_LAZYTIME) | (sb->s_flags & SB_LAZYTIME); return 0; restore_gc: @@ -3521,6 +3533,7 @@ try_onemore: sbi->reserved_blocks = 0; sbi->current_reserved_blocks = 0; limit_reserve_root(sbi); + adjust_unusable_cap_perc(sbi); for (i = 0; i < NR_INODE_TYPE; i++) { INIT_LIST_HEAD(&sbi->inode_list[i]); From 2d9a99d8c320509f67452228756ef8e504fe96a9 Mon Sep 17 00:00:00 2001 From: Jaegeuk Kim Date: Wed, 13 May 2020 21:12:53 -0700 Subject: [PATCH 86/86] f2fs: flush dirty meta pages when flushing them Let's guarantee flusing dirty meta pages to avoid infinite loop. Signed-off-by: Jaegeuk Kim --- fs/f2fs/checkpoint.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/fs/f2fs/checkpoint.c b/fs/f2fs/checkpoint.c index 5ba649e17c72..e3604a3a43f5 100644 --- a/fs/f2fs/checkpoint.c +++ b/fs/f2fs/checkpoint.c @@ -1266,6 +1266,9 @@ void f2fs_wait_on_all_pages(struct f2fs_sb_info *sbi, int type) if (unlikely(f2fs_cp_error(sbi))) break; + if (type == F2FS_DIRTY_META) + f2fs_sync_meta_pages(sbi, META, LONG_MAX, + FS_CP_META_IO); io_schedule_timeout(DEFAULT_IO_TIMEOUT); } finish_wait(&sbi->cp_wait, &wait);