Re: [PATCH 5.4/5.10 1/1] mac80211: mesh: embedd mesh_paths and mpp_paths into ieee80211_if_mesh

From: Greg Kroah-Hartman
Date: Thu Feb 23 2023 - 04:11:57 EST


On Wed, Feb 22, 2023 at 11:03:01PM +0300, Fedor Pchelkin wrote:
> From: Pavel Skripkin <paskripkin@xxxxxxxxx>
>
> commit 8b5cb7e41d9d77ffca036b0239177de123394a55 upstream.
>
> Syzbot hit NULL deref in rhashtable_free_and_destroy(). The problem was
> in mesh_paths and mpp_paths being NULL.
>
> mesh_pathtbl_init() could fail in case of memory allocation failure, but
> nobody cared, since ieee80211_mesh_init_sdata() returns void. It led to
> leaving 2 pointers as NULL. Syzbot has found null deref on exit path,
> but it could happen anywhere else, because code assumes these pointers are
> valid.
>
> Since all ieee80211_*_setup_sdata functions are void and do not fail,
> let's embedd mesh_paths and mpp_paths into parent struct to avoid
> adding error handling on higher levels and follow the pattern of others
> setup_sdata functions
>
> Fixes: 60854fd94573 ("mac80211: mesh: convert path table to rhashtable")
> Reported-and-tested-by: syzbot+860268315ba86ea6b96b@xxxxxxxxxxxxxxxxxxxxxxxxx
> Signed-off-by: Pavel Skripkin <paskripkin@xxxxxxxxx>
> Link: https://lore.kernel.org/r/20211230195547.23977-1-paskripkin@xxxxxxxxx
> Signed-off-by: Johannes Berg <johannes.berg@xxxxxxxxx>
> [pchelkin@xxxxxxxxx: adapt a comment spell fixing issue]
> Signed-off-by: Fedor Pchelkin <pchelkin@xxxxxxxxx>
> ---

This also worked for 4.19.y, but not 4.14.y, care to also fix it there?

thanks,

greg k-h