Re: [PATCH] hugetlbfs: emit hugetlbfs_free_inode from ->destroy_inode()

[Muchun Song <muchun.song@linux.dev>]

This is the V2 version. you should specify it in your subject line prefix.

> On Nov 1, 2026, at 00:06, Joas Antonio dos Santos via B4 Relay <devnull+joasantonio108.gmail.com@kernel.org> wrote:
> 
> From: Joas Antonio dos Santos <joasantonio108@gmail.com>
> 
> hugetlbfs_free_inode() calls trace_hugetlbfs_free_inode(), and the
> hugetlbfs__inode event class records inode->i_sb->s_dev.  ->free_inode()
> runs from an RCU callback (destroy_inode() queues i_callback() with
> call_rcu()), and Documentation/filesystems/porting.rst says that this
> callback "is *NOT* ordered wrt superblock destruction; as the matter of
> fact, the superblock and all associated structures might be already
> gone".  put_super() frees the superblock with call_rcu() followed by a
> workqueue item (destroy_super_rcu() -> destroy_super_work()).  syzbot
> hit this with the event enabled:
> 
>  BUG: KASAN: slab-use-after-free in trace_event_raw_event_hugetlbfs__inode+0x45b/0x4c0 include/trace/events/hugetlbfs.h:36
>  Read of size 4 at addr ffff888038850010 by task syz-executor952/6134
>   hugetlbfs_free_inode+0x15e/0x1f0 fs/hugetlbfs/inode.c:1182
>   i_callback+0x46/0x80 fs/inode.c:317
>   rcu_do_batch kernel/rcu/tree.c:2650 [inline]
>   rcu_core+0x5f0/0x1220 kernel/rcu/tree.c:2919
>  Allocated by task 6134:
>   alloc_super fs/super.c:326 [inline]
>   sget_fc+0x1ba/0x1f60 fs/super.c:874
>   hugetlbfs_get_tree+0x1e7/0x2b0 fs/hugetlbfs/inode.c:1431
>  Freed by task 855:
>   kfree+0x22b/0x6d0 mm/slub.c:6923
>   process_one_work+0xac7/0x1b10 kernel/workqueue.c:3495
> 
> Emit the event from hugetlbfs_destroy_inode() instead.  destroy_inode()
> calls ->destroy_inode() synchronously, before it queues the RCU
> callback, while the superblock is still alive; hugetlbfs_destroy_inode()
> already dereferences inode->i_sb there.  ->free_inode() is left with
> only the kmem_cache_free().
> 
> Fixes: 014ad7c42a69 ("hugetlbfs: use tracepoints in hugetlbfs functions.")
> Reported-by: syzbot+91e1fc35f51774c479b2@syzkaller.appspotmail.com
> Closes: https://lore.kernel.org/all/6ae27525.5451848c.f7707.0001.GAE@google.com/
> Cc: stable@vger.kernel.org
> Signed-off-by: Joas Antonio dos Santos <joasantonio108@gmail.com>
> ---
> Resend: my mail provider blocked the original submission to the lists
> on 2026-10-29, so only the direct recipients got it. No change to the
> patch, except dropping a deprecated Assisted-by trailer.
> 
> David asked whether the tracepoint should be renamed now that it no
> longer fires from ->free_inode(). I'd keep this patch as is, since it
> is tagged for stable and the event name is visible in tracefs
> (events/hugetlbfs/hugetlbfs_free_inode); renaming it in stable kernels
> would break existing tracing setups just to fix the use-after-free. I
> can send a follow-up for mm-unstable that renames it to
> hugetlbfs_destroy_inode, next to hugetlbfs_evict_inode, or fold the
> rename into a v2 if preferred.

At least for me, I think we should rename it, and it belongs to the fix
itself. Otherwise, we will introduce another behavior change, someone
thinks the tracepoint is fired through free_inode callback, but it's
not, it's weird.

Thanks,
Muchun

> 
> Tested with the syzbot C reproducer on a KASAN kernel with the event
> enabled: the use-after-free reproduced in 3 of 3 boots without the
> patch and in none of 12 runs with it, and the event is still emitted.
> 
> Tool-use note: the analysis and testing were done with help from an
> LLM-based assistant; I reviewed the change and ran the tests myself.
> ---
> fs/hugetlbfs/inode.c | 6 +++++-
> 1 file changed, 5 insertions(+), 1 deletion(-)
> 
> diff --git a/fs/hugetlbfs/inode.c b/fs/hugetlbfs/inode.c
> index 01c4d6f44..5bd348288 100644
> --- a/fs/hugetlbfs/inode.c
> +++ b/fs/hugetlbfs/inode.c
> @@ -1179,12 +1179,16 @@ static struct inode *hugetlbfs_alloc_inode(struct super_block *sb)
> 
> static void hugetlbfs_free_inode(struct inode *inode)
> {
> -	trace_hugetlbfs_free_inode(inode);
> 	kmem_cache_free(hugetlbfs_inode_cachep, HUGETLBFS_I(inode));
> }
> 
> static void hugetlbfs_destroy_inode(struct inode *inode)
> {
> +	/*
> +	 * Not in ->free_inode(): that runs from an RCU callback, when the
> +	 * superblock the event reads may already be freed.
> +	 */
> +	trace_hugetlbfs_free_inode(inode);
> 	hugetlbfs_inc_free_inodes(HUGETLBFS_SB(inode->i_sb));
> }
> 
> 
> ---
> base-commit: c339f4010dea8b0d754f84ace70cd1b089878ef8
> change-id: 20261031-b4-hugetlbfs-trace-341b338ab57b
> 
> Best regards,
> -- 
> Joas Antonio dos Santos <joasantonio108@gmail.com>
> 
>