From be3f1938d3e6ea8186f0de3dd95245dda4f22c1e Mon Sep 17 00:00:00 2001 From: Dave Chen Date: Tue, 15 Apr 2025 14:33:42 +0800 Subject: [PATCH 1/3] btrfs: fix COW handling in run_delalloc_nocow() In run_delalloc_nocow(), when the found btrfs_key's offset > cur_offset, it indicates a gap between the current processing region and the next file extent. The original code would directly jump to the "must_cow" label, which increments the slot and forces a fallback to COW. This behavior might skip an extent item and result in an overestimated COW fallback range. This patch modifies the logic so that when a gap is detected: - If no COW range is already being recorded (cow_start is unset), cow_start is set to cur_offset. - cur_offset is then advanced to the beginning of the next extent. - Instead of jumping to "must_cow", control flows directly to "next_slot" so that the same extent item can be reexamined properly. The change ensures that we accurately account for the extent gap and avoid accidentally extending the range that needs to fallback to COW. CC: stable@vger.kernel.org # 6.6+ Reviewed-by: Filipe Manana Signed-off-by: Dave Chen Signed-off-by: David Sterba --- fs/btrfs/inode.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 6eedfbfce1cb..312fa996a987 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -2129,12 +2129,13 @@ static noinline int run_delalloc_nocow(struct btrfs_inode *inode, /* * If the found extent starts after requested offset, then - * adjust extent_end to be right before this extent begins + * adjust cur_offset to be right before this extent begins. */ if (found_key.offset > cur_offset) { - extent_end = found_key.offset; - extent_type = 0; - goto must_cow; + if (cow_start == (u64)-1) + cow_start = cur_offset; + cur_offset = found_key.offset; + goto next_slot; } /* From 48c1d1bb525b1c44b8bdc8e7ec5629cb6c2b9fc4 Mon Sep 17 00:00:00 2001 From: Penglei Jiang Date: Mon, 21 Apr 2025 08:40:29 -0700 Subject: [PATCH 2/3] btrfs: fix the inode leak in btrfs_iget() [BUG] There is a bug report that a syzbot reproducer can lead to the following busy inode at unmount time: BTRFS info (device loop1): last unmount of filesystem 1680000e-3c1e-4c46-84b6-56bd3909af50 VFS: Busy inodes after unmount of loop1 (btrfs) ------------[ cut here ]------------ kernel BUG at fs/super.c:650! Oops: invalid opcode: 0000 [#1] SMP KASAN NOPTI CPU: 0 UID: 0 PID: 48168 Comm: syz-executor Not tainted 6.15.0-rc2-00471-g119009db2674 #2 PREEMPT(full) Hardware name: QEMU Ubuntu 24.04 PC (i440FX + PIIX, 1996), BIOS 1.16.3-debian-1.16.3-2 04/01/2014 RIP: 0010:generic_shutdown_super+0x2e9/0x390 fs/super.c:650 Call Trace: kill_anon_super+0x3a/0x60 fs/super.c:1237 btrfs_kill_super+0x3b/0x50 fs/btrfs/super.c:2099 deactivate_locked_super+0xbe/0x1a0 fs/super.c:473 deactivate_super fs/super.c:506 [inline] deactivate_super+0xe2/0x100 fs/super.c:502 cleanup_mnt+0x21f/0x440 fs/namespace.c:1435 task_work_run+0x14d/0x240 kernel/task_work.c:227 resume_user_mode_work include/linux/resume_user_mode.h:50 [inline] exit_to_user_mode_loop kernel/entry/common.c:114 [inline] exit_to_user_mode_prepare include/linux/entry-common.h:329 [inline] __syscall_exit_to_user_mode_work kernel/entry/common.c:207 [inline] syscall_exit_to_user_mode+0x269/0x290 kernel/entry/common.c:218 do_syscall_64+0xd4/0x250 arch/x86/entry/syscall_64.c:100 entry_SYSCALL_64_after_hwframe+0x77/0x7f [CAUSE] When btrfs_alloc_path() failed, btrfs_iget() directly returned without releasing the inode already allocated by btrfs_iget_locked(). This results the above busy inode and trigger the kernel BUG. [FIX] Fix it by calling iget_failed() if btrfs_alloc_path() failed. If we hit error inside btrfs_read_locked_inode(), it will properly call iget_failed(), so nothing to worry about. Although the iget_failed() cleanup inside btrfs_read_locked_inode() is a break of the normal error handling scheme, let's fix the obvious bug and backport first, then rework the error handling later. Reported-by: Penglei Jiang Link: https://lore.kernel.org/linux-btrfs/20250421102425.44431-1-superman.xpt@gmail.com/ Fixes: 7c855e16ab72 ("btrfs: remove conditional path allocation in btrfs_read_locked_inode()") CC: stable@vger.kernel.org # 6.13+ Reviewed-by: Qu Wenruo Signed-off-by: Penglei Jiang Signed-off-by: David Sterba --- fs/btrfs/inode.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 312fa996a987..d295a37fa049 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -5682,8 +5682,10 @@ struct btrfs_inode *btrfs_iget(u64 ino, struct btrfs_root *root) return inode; path = btrfs_alloc_path(); - if (!path) + if (!path) { + iget_failed(&inode->vfs_inode); return ERR_PTR(-ENOMEM); + } ret = btrfs_read_locked_inode(inode, path); btrfs_free_path(path); From e08e49d986f82c30f42ad0ed43ebbede1e1e3739 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Mon, 14 Apr 2025 14:51:58 -0400 Subject: [PATCH 3/3] btrfs: adjust subpage bit start based on sectorsize When running machines with 64k page size and a 16k nodesize we started seeing tree log corruption in production. This turned out to be because we were not writing out dirty blocks sometimes, so this in fact affects all metadata writes. When writing out a subpage EB we scan the subpage bitmap for a dirty range. If the range isn't dirty we do bit_start++; to move onto the next bit. The problem is the bitmap is based on the number of sectors that an EB has. So in this case, we have a 64k pagesize, 16k nodesize, but a 4k sectorsize. This means our bitmap is 4 bits for every node. With a 64k page size we end up with 4 nodes per page. To make this easier this is how everything looks [0 16k 32k 48k ] logical address [0 4 8 12 ] radix tree offset [ 64k page ] folio [ 16k eb ][ 16k eb ][ 16k eb ][ 16k eb ] extent buffers [ | | | | | | | | | | | | | | | | ] bitmap Now we use all of our addressing based on fs_info->sectorsize_bits, so as you can see the above our 16k eb->start turns into radix entry 4. When we find a dirty range for our eb, we correctly do bit_start += sectors_per_node, because if we start at bit 0, the next bit for the next eb is 4, to correspond to eb->start 16k. However if our range is clean, we will do bit_start++, which will now put us offset from our radix tree entries. In our case, assume that the first time we check the bitmap the block is not dirty, we increment bit_start so now it == 1, and then we loop around and check again. This time it is dirty, and we go to find that start using the following equation start = folio_start + bit_start * fs_info->sectorsize; so in the case above, eb->start 0 is now dirty, and we calculate start as 0 + 1 * fs_info->sectorsize = 4096 4096 >> 12 = 1 Now we're looking up the radix tree for 1, and we won't find an eb. What's worse is now we're using bit_start == 1, so we do bit_start += sectors_per_node, which is now 5. If that eb is dirty we will run into the same thing, we will look at an offset that is not populated in the radix tree, and now we're skipping the writeout of dirty extent buffers. The best fix for this is to not use sectorsize_bits to address nodes, but that's a larger change. Since this is a fs corruption problem fix it simply by always using sectors_per_node to increment the start bit. Fixes: c4aec299fa8f ("btrfs: introduce submit_eb_subpage() to submit a subpage metadata page") CC: stable@vger.kernel.org # 5.15+ Reviewed-by: Boris Burkov Reviewed-by: Qu Wenruo Signed-off-by: Josef Bacik Signed-off-by: David Sterba --- fs/btrfs/extent_io.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c index 197f5e51c474..8515c31f563b 100644 --- a/fs/btrfs/extent_io.c +++ b/fs/btrfs/extent_io.c @@ -2047,7 +2047,7 @@ static int submit_eb_subpage(struct folio *folio, struct writeback_control *wbc) subpage->bitmaps)) { spin_unlock_irqrestore(&subpage->lock, flags); spin_unlock(&folio->mapping->i_private_lock); - bit_start++; + bit_start += sectors_per_node; continue; }