Re: [PATCH] slab: fix a crash by reading /proc/slab_allocators

["Tobin C. Harding" <me@tobin.cc>]

On Sat, Apr 06, 2019 at 06:59:01PM -0400, Qian Cai wrote:
> The commit 510ded33e075 ("slab: implement slab_root_caches list")
> changes the name of the list node within "struct kmem_cache" from
> "list" to "root_caches_node", but leaks_show() still use the "list"
> which causes a crash when reading /proc/slab_allocators.
> 
> BUG: unable to handle kernel NULL pointer dereference at
> 00000000000000aa
> PGD 0 P4D 0
> Oops: 0000 [#1] SMP DEBUG_PAGEALLOC PTI
> CPU: 3 PID: 5925 Comm: ldd Not tainted 5.1.0-rc3-mm1+ #6
> RIP: 0010:__lock_acquire.isra.14+0x4b4/0xa50
> Call Trace:
>  <IRQ>
>  lock_acquire+0xa3/0x180
>  _raw_spin_lock+0x2f/0x40
>  do_drain+0x61/0xc0
>  flush_smp_call_function_queue+0x3a/0x110
>  generic_smp_call_function_single_interrupt+0x13/0x2b
>  smp_call_function_interrupt+0x66/0x1a0
>  call_function_interrupt+0xf/0x20
>  </IRQ>
> RIP: 0010:__tlb_remove_page_size+0x8c/0xe0
>  zap_pte_range+0x39f/0xc80
>  unmap_page_range+0x38a/0x550
>  unmap_single_vma+0x7d/0xe0
>  unmap_vmas+0xae/0xd0
>  exit_mmap+0xae/0x190
>  mmput+0x7a/0x150
>  do_exit+0x2d9/0xd40
>  do_group_exit+0x41/0xd0
>  __x64_sys_exit_group+0x18/0x20
>  do_syscall_64+0x68/0x381
>  entry_SYSCALL_64_after_hwframe+0x44/0xa9
> 
> Fixes: 510ded33e075 ("slab: implement slab_root_caches list")
> Signed-off-by: Qian Cai <cai@lca.pw>
> ---
>  mm/slab.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/mm/slab.c b/mm/slab.c
> index 46a6e084222b..9142ee992493 100644
> --- a/mm/slab.c
> +++ b/mm/slab.c
> @@ -4307,7 +4307,8 @@ static void show_symbol(struct seq_file *m, unsigned long address)
>  
>  static int leaks_show(struct seq_file *m, void *p)
>  {
> -	struct kmem_cache *cachep = list_entry(p, struct kmem_cache, list);
> +	struct kmem_cache *cachep = list_entry(p, struct kmem_cache,
> +					       root_caches_node);
>  	struct page *page;
>  	struct kmem_cache_node *n;
>  	const char *name;
> -- 
> 2.17.2 (Apple Git-113)
> 

For what its worth

Reviewed-by: Tobin C. Harding <tobin@kernel.org>

thanks,
Tobin.