kernel/linux.git/fs/nilfs2, branch linux-6.0.y

nilfs2: fix shift-out-of-bounds due to too large exponent of block size

2022-12-31T12:26:41+00:00

[ Upstream commit ebeccaaef67a4895d2496ab8d9c2fb8d89201211 ] If field s_log_block_size of superblock data is corrupted and too large, init_nilfs() and load_nilfs() still can trigger a shift-out-of-bounds warning followed by a kernel panic (if panic_on_warn is set): shift exponent 38973 is too large for 32-bit type 'int' Call Trace: dump_stack_lvl+0xcd/0x134 ubsan_epilogue+0xb/0x50 __ubsan_handle_shift_out_of_bounds.cold.12+0x17b/0x1f5 init_nilfs.cold.11+0x18/0x1d [nilfs2] nilfs_mount+0x9b5/0x12b0 [nilfs2] ... This fixes the issue by adding and using a new helper function for getting block size with sanity check. Link: https://lkml.kernel.org/r/20221027044306.42774-3-konishi.ryusuke@gmail.com Signed-off-by: Ryusuke Konishi Tested-by: Ryusuke Konishi Signed-off-by: Andrew Morton Signed-off-by: Sasha Levin

nilfs2: fix shift-out-of-bounds/overflow in nilfs_sb2_bad_offset()

2022-12-31T12:26:41+00:00

[ Upstream commit 610a2a3d7d8be3537458a378ec69396a76c385b6 ] Patch series "nilfs2: fix UBSAN shift-out-of-bounds warnings on mount time". The first patch fixes a bug reported by syzbot, and the second one fixes the remaining bug of the same kind. Although they are triggered by the same super block data anomaly, I divided it into the above two because the details of the issues and how to fix it are different. Both are required to eliminate the shift-out-of-bounds issues at mount time. This patch (of 2): If the block size exponent information written in an on-disk superblock is corrupted, nilfs_sb2_bad_offset helper function can trigger shift-out-of-bounds warning followed by a kernel panic (if panic_on_warn is set): shift exponent 38983 is too large for 64-bit type 'unsigned long long' Call Trace: __dump_stack lib/dump_stack.c:88 [inline] dump_stack_lvl+0x1b1/0x28e lib/dump_stack.c:106 ubsan_epilogue lib/ubsan.c:151 [inline] __ubsan_handle_shift_out_of_bounds+0x33d/0x3b0 lib/ubsan.c:322 nilfs_sb2_bad_offset fs/nilfs2/the_nilfs.c:449 [inline] nilfs_load_super_block+0xdf5/0xe00 fs/nilfs2/the_nilfs.c:523 init_nilfs+0xb7/0x7d0 fs/nilfs2/the_nilfs.c:577 nilfs_fill_super+0xb1/0x5d0 fs/nilfs2/super.c:1047 nilfs_mount+0x613/0x9b0 fs/nilfs2/super.c:1317 ... In addition, since nilfs_sb2_bad_offset() performs multiplication without considering the upper bound, the computation may overflow if the disk layout parameters are not normal. This fixes these issues by inserting preliminary sanity checks for those parameters and by converting the comparison from one involving multiplication and left bit-shifting to one using division and right bit-shifting. Link: https://lkml.kernel.org/r/20221027044306.42774-1-konishi.ryusuke@gmail.com Link: https://lkml.kernel.org/r/20221027044306.42774-2-konishi.ryusuke@gmail.com Signed-off-by: Ryusuke Konishi Reported-by: syzbot+e91619dd4c11c4960706@syzkaller.appspotmail.com Tested-by: Ryusuke Konishi Signed-off-by: Andrew Morton Signed-off-by: Sasha Levin

nilfs2: fix NULL pointer dereference in nilfs_palloc_commit_free_entry()

2022-12-08T10:30:19+00:00

commit f0a0ccda18d6fd826d7c7e7ad48a6ed61c20f8b4 upstream. Syzbot reported a null-ptr-deref bug: NILFS (loop0): segctord starting. Construction interval = 5 seconds, CP frequency < 30 seconds general protection fault, probably for non-canonical address 0xdffffc0000000002: 0000 [#1] PREEMPT SMP KASAN KASAN: null-ptr-deref in range [0x0000000000000010-0x0000000000000017] CPU: 1 PID: 3603 Comm: segctord Not tainted 6.1.0-rc2-syzkaller-00105-gb229b6ca5abb #0 Hardware name: Google Compute Engine/Google Compute Engine, BIOS Google 10/11/2022 RIP: 0010:nilfs_palloc_commit_free_entry+0xe5/0x6b0 fs/nilfs2/alloc.c:608 Code: 00 00 00 00 fc ff df 80 3c 02 00 0f 85 cd 05 00 00 48 b8 00 00 00 00 00 fc ff df 4c 8b 73 08 49 8d 7e 10 48 89 fa 48 c1 ea 03 <80> 3c 02 00 0f 85 26 05 00 00 49 8b 46 10 be a6 00 00 00 48 c7 c7 RSP: 0018:ffffc90003dff830 EFLAGS: 00010212 RAX: dffffc0000000000 RBX: ffff88802594e218 RCX: 000000000000000d RDX: 0000000000000002 RSI: 0000000000002000 RDI: 0000000000000010 RBP: ffff888071880222 R08: 0000000000000005 R09: 000000000000003f R10: 000000000000000d R11: 0000000000000000 R12: ffff888071880158 R13: ffff88802594e220 R14: 0000000000000000 R15: 0000000000000004 FS: 0000000000000000(0000) GS:ffff8880b9b00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 00007fb1c08316a8 CR3: 0000000018560000 CR4: 0000000000350ee0 Call Trace: nilfs_dat_commit_free fs/nilfs2/dat.c:114 [inline] nilfs_dat_commit_end+0x464/0x5f0 fs/nilfs2/dat.c:193 nilfs_dat_commit_update+0x26/0x40 fs/nilfs2/dat.c:236 nilfs_btree_commit_update_v+0x87/0x4a0 fs/nilfs2/btree.c:1940 nilfs_btree_commit_propagate_v fs/nilfs2/btree.c:2016 [inline] nilfs_btree_propagate_v fs/nilfs2/btree.c:2046 [inline] nilfs_btree_propagate+0xa00/0xd60 fs/nilfs2/btree.c:2088 nilfs_bmap_propagate+0x73/0x170 fs/nilfs2/bmap.c:337 nilfs_collect_file_data+0x45/0xd0 fs/nilfs2/segment.c:568 nilfs_segctor_apply_buffers+0x14a/0x470 fs/nilfs2/segment.c:1018 nilfs_segctor_scan_file+0x3f4/0x6f0 fs/nilfs2/segment.c:1067 nilfs_segctor_collect_blocks fs/nilfs2/segment.c:1197 [inline] nilfs_segctor_collect fs/nilfs2/segment.c:1503 [inline] nilfs_segctor_do_construct+0x12fc/0x6af0 fs/nilfs2/segment.c:2045 nilfs_segctor_construct+0x8e3/0xb30 fs/nilfs2/segment.c:2379 nilfs_segctor_thread_construct fs/nilfs2/segment.c:2487 [inline] nilfs_segctor_thread+0x3c3/0xf30 fs/nilfs2/segment.c:2570 kthread+0x2e4/0x3a0 kernel/kthread.c:376 ret_from_fork+0x1f/0x30 arch/x86/entry/entry_64.S:306 ... If DAT metadata file is corrupted on disk, there is a case where req->pr_desc_bh is NULL and blocknr is 0 at nilfs_dat_commit_end() during a b-tree operation that cascadingly updates ancestor nodes of the b-tree, because nilfs_dat_commit_alloc() for a lower level block can initialize the blocknr on the same DAT entry between nilfs_dat_prepare_end() and nilfs_dat_commit_end(). If this happens, nilfs_dat_commit_end() calls nilfs_dat_commit_free() without valid buffer heads in req->pr_desc_bh and req->pr_bitmap_bh, and causes the NULL pointer dereference above in nilfs_palloc_commit_free_entry() function, which leads to a crash. Fix this by adding a NULL check on req->pr_desc_bh and req->pr_bitmap_bh before nilfs_palloc_commit_free_entry() in nilfs_dat_commit_free(). This also calls nilfs_error() in that case to notify that there is a fatal flaw in the filesystem metadata and prevent further operations. Link: https://lkml.kernel.org/r/00000000000097c20205ebaea3d6@google.com Link: https://lkml.kernel.org/r/20221114040441.1649940-1-zhangpeng362@huawei.com Link: https://lkml.kernel.org/r/20221119120542.17204-1-konishi.ryusuke@gmail.com Signed-off-by: ZhangPeng Signed-off-by: Ryusuke Konishi Reported-by: syzbot+ebe05ee8e98f755f61d0@syzkaller.appspotmail.com Tested-by: Ryusuke Konishi Cc: Signed-off-by: Andrew Morton Signed-off-by: Greg Kroah-Hartman

nilfs2: fix nilfs_sufile_mark_dirty() not set segment usage as dirty

2022-12-02T16:43:12+00:00

commit 512c5ca01a3610ab14ff6309db363de51f1c13a6 upstream. When extending segments, nilfs_sufile_alloc() is called to get an unassigned segment, then mark it as dirty to avoid accidentally allocating the same segment in the future. But for some special cases such as a corrupted image it can be unreliable. If such corruption of the dirty state of the segment occurs, nilfs2 may reallocate a segment that is in use and pick the same segment for writing twice at the same time. This will cause the problem reported by syzkaller: https://syzkaller.appspot.com/bug?id=c7c4748e11ffcc367cef04f76e02e931833cbd24 This case started with segbuf1.segnum = 3, nextnum = 4 when constructed. It supposed segment 4 has already been allocated and marked as dirty. However the dirty state was corrupted and segment 4 usage was not dirty. For the first time nilfs_segctor_extend_segments() segment 4 was allocated again, which made segbuf2 and next segbuf3 had same segment 4. sb_getblk() will get same bh for segbuf2 and segbuf3, and this bh is added to both buffer lists of two segbuf. It makes the lists broken which causes NULL pointer dereference. Fix the problem by setting usage as dirty every time in nilfs_sufile_mark_dirty(), which is called during constructing current segment to be written out and before allocating next segment. [chenzhongjin@huawei.com: add lock protection per Ryusuke] Link: https://lkml.kernel.org/r/20221121091141.214703-1-chenzhongjin@huawei.com Link: https://lkml.kernel.org/r/20221118063304.140187-1-chenzhongjin@huawei.com Fixes: 9ff05123e3bf ("nilfs2: segment constructor") Signed-off-by: Chen Zhongjin Reported-by: Reported-by: Liu Shixin Acked-by: Ryusuke Konishi Tested-by: Ryusuke Konishi Cc: Signed-off-by: Andrew Morton Signed-off-by: Greg Kroah-Hartman

nilfs2: fix use-after-free bug of ns_writer on remount

2022-11-16T09:04:07+00:00

commit 8cccf05fe857a18ee26e20d11a8455a73ffd4efd upstream. If a nilfs2 filesystem is downgraded to read-only due to metadata corruption on disk and is remounted read/write, or if emergency read-only remount is performed, detaching a log writer and synchronizing the filesystem can be done at the same time. In these cases, use-after-free of the log writer (hereinafter nilfs->ns_writer) can happen as shown in the scenario below: Task1 Task2 -------------------------------- ------------------------------ nilfs_construct_segment nilfs_segctor_sync init_wait init_waitqueue_entry add_wait_queue schedule nilfs_remount (R/W remount case) nilfs_attach_log_writer nilfs_detach_log_writer nilfs_segctor_destroy kfree finish_wait _raw_spin_lock_irqsave __raw_spin_lock_irqsave do_raw_spin_lock debug_spin_lock_before <-- use-after-free While Task1 is sleeping, nilfs->ns_writer is freed by Task2. After Task1 waked up, Task1 accesses nilfs->ns_writer which is already freed. This scenario diagram is based on the Shigeru Yoshida's post [1]. This patch fixes the issue by not detaching nilfs->ns_writer on remount so that this UAF race doesn't happen. Along with this change, this patch also inserts a few necessary read-only checks with superblock instance where only the ns_writer pointer was used to check if the filesystem is read-only. Link: https://syzkaller.appspot.com/bug?id=79a4c002e960419ca173d55e863bd09e8112df8b Link: https://lkml.kernel.org/r/20221103141759.1836312-1-syoshida@redhat.com [1] Link: https://lkml.kernel.org/r/20221104142959.28296-1-konishi.ryusuke@gmail.com Signed-off-by: Ryusuke Konishi Reported-by: syzbot+f816fa82f8783f7a02bb@syzkaller.appspotmail.com Reported-by: Shigeru Yoshida Tested-by: Ryusuke Konishi Cc: Signed-off-by: Andrew Morton Signed-off-by: Greg Kroah-Hartman

nilfs2: fix deadlock in nilfs_count_free_blocks()

2022-11-16T09:04:06+00:00

commit 8ac932a4921a96ca52f61935dbba64ea87bbd5dc upstream. A semaphore deadlock can occur if nilfs_get_block() detects metadata corruption while locating data blocks and a superblock writeback occurs at the same time: task 1 task 2 ------ ------ * A file operation * nilfs_truncate() nilfs_get_block() down_read(rwsem A) <-- nilfs_bmap_lookup_contig() ... generic_shutdown_super() nilfs_put_super() * Prepare to write superblock * down_write(rwsem B) <-- nilfs_cleanup_super() * Detect b-tree corruption * nilfs_set_log_cursor() nilfs_bmap_convert_error() nilfs_count_free_blocks() __nilfs_error() down_read(rwsem A) <-- nilfs_set_error() down_write(rwsem B) <-- *** DEADLOCK *** Here, nilfs_get_block() readlocks rwsem A (= NILFS_MDT(dat_inode)->mi_sem) and then calls nilfs_bmap_lookup_contig(), but if it fails due to metadata corruption, __nilfs_error() is called from nilfs_bmap_convert_error() inside the lock section. Since __nilfs_error() calls nilfs_set_error() unless the filesystem is read-only and nilfs_set_error() attempts to writelock rwsem B (= nilfs->ns_sem) to write back superblock exclusively, hierarchical lock acquisition occurs in the order rwsem A -> rwsem B. Now, if another task starts updating the superblock, it may writelock rwsem B during the lock sequence above, and can deadlock trying to readlock rwsem A in nilfs_count_free_blocks(). However, there is actually no need to take rwsem A in nilfs_count_free_blocks() because it, within the lock section, only reads a single integer data on a shared struct with nilfs_sufile_get_ncleansegs(). This has been the case after commit aa474a220180 ("nilfs2: add local variable to cache the number of clean segments"), that is, even before this bug was introduced. So, this resolves the deadlock problem by just not taking the semaphore in nilfs_count_free_blocks(). Link: https://lkml.kernel.org/r/20221029044912.9139-1-konishi.ryusuke@gmail.com Fixes: e828949e5b42 ("nilfs2: call nilfs_error inside bmap routines") Signed-off-by: Ryusuke Konishi Reported-by: syzbot+45d6ce7b7ad7ef455d03@syzkaller.appspotmail.com Tested-by: Ryusuke Konishi Cc: [2.6.38+ Signed-off-by: Andrew Morton Signed-off-by: Greg Kroah-Hartman

nilfs2: replace WARN_ONs by nilfs_error for checkpoint acquisition failure

2022-10-15T06:02:54+00:00

commit 723ac751208f6d6540191689cfbf6c77135a7a1b upstream. If creation or finalization of a checkpoint fails due to anomalies in the checkpoint metadata on disk, a kernel warning is generated. This patch replaces the WARN_ONs by nilfs_error, so that a kernel, booted with panic_on_warn, does not panic. A nilfs_error is appropriate here to handle the abnormal filesystem condition. This also replaces the detected error codes with an I/O error so that neither of the internal error codes is returned to callers. Link: https://lkml.kernel.org/r/20220929123330.19658-1-konishi.ryusuke@gmail.com Signed-off-by: Ryusuke Konishi Reported-by: syzbot+fbb3e0b24e8dae5a16ee@syzkaller.appspotmail.com Signed-off-by: Andrew Morton Signed-off-by: Greg Kroah-Hartman

nilfs2: fix leak of nilfs_root in case of writer thread creation failure

2022-10-15T06:02:54+00:00

commit d0d51a97063db4704a5ef6bc978dddab1636a306 upstream. If nilfs_attach_log_writer() failed to create a log writer thread, it frees a data structure of the log writer without any cleanup. After commit e912a5b66837 ("nilfs2: use root object to get ifile"), this causes a leak of struct nilfs_root, which started to leak an ifile metadata inode and a kobject on that struct. In addition, if the kernel is booted with panic_on_warn, the above ifile metadata inode leak will cause the following panic when the nilfs2 kernel module is removed: kmem_cache_destroy nilfs2_inode_cache: Slab cache still has objects when called from nilfs_destroy_cachep+0x16/0x3a [nilfs2] WARNING: CPU: 8 PID: 1464 at mm/slab_common.c:494 kmem_cache_destroy+0x138/0x140 ... RIP: 0010:kmem_cache_destroy+0x138/0x140 Code: 00 20 00 00 e8 a9 55 d8 ff e9 76 ff ff ff 48 8b 53 60 48 c7 c6 20 70 65 86 48 c7 c7 d8 69 9c 86 48 8b 4c 24 28 e8 ef 71 c7 00 <0f> 0b e9 53 ff ff ff c3 48 81 ff ff 0f 00 00 77 03 31 c0 c3 53 48 ... Call Trace: ? nilfs_palloc_freev.cold.24+0x58/0x58 [nilfs2] nilfs_destroy_cachep+0x16/0x3a [nilfs2] exit_nilfs_fs+0xa/0x1b [nilfs2] __x64_sys_delete_module+0x1d9/0x3a0 ? __sanitizer_cov_trace_pc+0x1a/0x50 ? syscall_trace_enter.isra.19+0x119/0x190 do_syscall_64+0x34/0x80 entry_SYSCALL_64_after_hwframe+0x63/0xcd ... Kernel panic - not syncing: panic_on_warn set ... This patch fixes these issues by calling nilfs_detach_log_writer() cleanup function if spawning the log writer thread fails. Link: https://lkml.kernel.org/r/20221007085226.57667-1-konishi.ryusuke@gmail.com Fixes: e912a5b66837 ("nilfs2: use root object to get ifile") Signed-off-by: Ryusuke Konishi Reported-by: syzbot+7381dc4ad60658ca4c05@syzkaller.appspotmail.com Tested-by: Ryusuke Konishi Cc: Signed-off-by: Andrew Morton Signed-off-by: Greg Kroah-Hartman

nilfs2: fix use-after-free bug of struct nilfs_root

2022-10-15T06:02:54+00:00

commit d325dc6eb763c10f591c239550b8c7e5466a5d09 upstream. If the beginning of the inode bitmap area is corrupted on disk, an inode with the same inode number as the root inode can be allocated and fail soon after. In this case, the subsequent call to nilfs_clear_inode() on that bogus root inode will wrongly decrement the reference counter of struct nilfs_root, and this will erroneously free struct nilfs_root, causing kernel oopses. This fixes the problem by changing nilfs_new_inode() to skip reserved inode numbers while repairing the inode bitmap. Link: https://lkml.kernel.org/r/20221003150519.39789-1-konishi.ryusuke@gmail.com Signed-off-by: Ryusuke Konishi Reported-by: syzbot+b8c672b0e22615c80fe0@syzkaller.appspotmail.com Reported-by: Khalid Masum Tested-by: Ryusuke Konishi Cc: Signed-off-by: Andrew Morton Signed-off-by: Greg Kroah-Hartman

nilfs2: fix NULL pointer dereference at nilfs_bmap_lookup_at_level()

2022-10-15T06:02:54+00:00

commit 21a87d88c2253350e115029f14fe2a10a7e6c856 upstream. If the i_mode field in inode of metadata files is corrupted on disk, it can cause the initialization of bmap structure, which should have been called from nilfs_read_inode_common(), not to be called. This causes a lockdep warning followed by a NULL pointer dereference at nilfs_bmap_lookup_at_level(). This patch fixes these issues by adding a missing sanitiy check for the i_mode field of metadata file's inode. Link: https://lkml.kernel.org/r/20221002030804.29978-1-konishi.ryusuke@gmail.com Signed-off-by: Ryusuke Konishi Reported-by: syzbot+2b32eb36c1a825b7a74c@syzkaller.appspotmail.com Reported-by: Tetsuo Handa Tested-by: Ryusuke Konishi Cc: Signed-off-by: Andrew Morton Signed-off-by: Greg Kroah-Hartman