Re: [PATCH] [v2] clk: fix slab-use-after-free when clk_core_populate_parent_map failed

["Zhou, Yun" <yun.zhou@windriver.com>]

Hi Brian,

Can you help me check if there are any issues with this new patch?

Thanks,
Yun

On 10/2/25 14:12, Yun Zhou wrote:
> If clk_core_populate_parent_map() fails, core->parents will be immediately
> released within clk_core_populate_parent_map(). Therefore it is can't be
> released in __clk_release() again.
>
> This fixes the following KASAN reported issue:
>
> ==================================================================
> BUG: KASAN: slab-use-after-free in __clk_release+0x80/0x160
> Read of size 8 at addr ffffff8043fd0980 by task kworker/u6:0/27
>
> CPU: 1 PID: 27 Comm: kworker/u6:0 Tainted: G        W          6.6.69-yocto-standard+ #7
> Hardware name: Raspberry Pi 4 Model B (DT)
> Workqueue: events_unbound deferred_probe_work_func
> Call trace:
>   dump_backtrace+0x98/0xf8
>   show_stack+0x20/0x38
>   dump_stack_lvl+0x48/0x60
>   print_report+0xf8/0x5d8
>   kasan_report+0xb4/0x100
>   __asan_load8+0x9c/0xc0
>   __clk_release+0x80/0x160
>   __clk_register+0x6dc/0xfb8
>   devm_clk_hw_register+0x70/0x108
>   bcm2835_register_clock+0x284/0x358
>   bcm2835_clk_probe+0x2c4/0x438
>   platform_probe+0x98/0x110
>   really_probe+0x1e4/0x3e8
>   __driver_probe_device+0xc0/0x1a0
>   driver_probe_device+0x110/0x1e8
>   __device_attach_driver+0xf0/0x1a8
>   bus_for_each_drv+0xf8/0x178
>   __device_attach+0x120/0x240
>   device_initial_probe+0x1c/0x30
>   bus_probe_device+0xdc/0xe8
>   deferred_probe_work_func+0xe8/0x130
>   process_one_work+0x2a4/0x698
>   worker_thread+0x53c/0x708
>   kthread+0x1b4/0x1c8
>   ret_from_fork+0x10/0x20
>
> Allocated by task 27:
>   kasan_save_stack+0x3c/0x68
>   kasan_set_track+0x2c/0x40
>   kasan_save_alloc_info+0x24/0x38
>   __kasan_kmalloc+0xd4/0xd8
>   __kmalloc+0x74/0x238
>   __clk_register+0x718/0xfb8
>   devm_clk_hw_register+0x70/0x108
>   bcm2835_register_clock+0x284/0x358
>   bcm2835_clk_probe+0x2c4/0x438
>   platform_probe+0x98/0x110
>   really_probe+0x1e4/0x3e8
>   __driver_probe_device+0xc0/0x1a0
>   driver_probe_device+0x110/0x1e8
>   __device_attach_driver+0xf0/0x1a8
>   bus_for_each_drv+0xf8/0x178
>   __device_attach+0x120/0x240
>   device_initial_probe+0x1c/0x30
>   bus_probe_device+0xdc/0xe8
>   deferred_probe_work_func+0xe8/0x130
>   process_one_work+0x2a4/0x698
>   worker_thread+0x53c/0x708
>   kthread+0x1b4/0x1c8
>   ret_from_fork+0x10/0x20
>
> Freed by task 27:
>   kasan_save_stack+0x3c/0x68
>   kasan_set_track+0x2c/0x40
>   kasan_save_free_info+0x38/0x60
>   __kasan_slab_free+0x100/0x170
>   slab_free_freelist_hook+0xcc/0x218
>   __kmem_cache_free+0x158/0x210
>   kfree+0x88/0x140
>   __clk_register+0x9d0/0xfb8
>   devm_clk_hw_register+0x70/0x108
>   bcm2835_register_clock+0x284/0x358
>   bcm2835_clk_probe+0x2c4/0x438
>   platform_probe+0x98/0x110
>   really_probe+0x1e4/0x3e8
>   __driver_probe_device+0xc0/0x1a0
>   driver_probe_device+0x110/0x1e8
>   __device_attach_driver+0xf0/0x1a8
>   bus_for_each_drv+0xf8/0x178
>   __device_attach+0x120/0x240
>   device_initial_probe+0x1c/0x30
>   bus_probe_device+0xdc/0xe8
>   deferred_probe_work_func+0xe8/0x130
>   process_one_work+0x2a4/0x698
>   worker_thread+0x53c/0x708
>   kthread+0x1b4/0x1c8
>   ret_from_fork+0x10/0x20
>
> The buggy address belongs to the object at ffffff8043fd0800
>   which belongs to the cache kmalloc-512 of size 512
> The buggy address is located 384 bytes inside of
>   freed 512-byte region [ffffff8043fd0800, ffffff8043fd0a00)
>
> The buggy address belongs to the physical page:
> page:fffffffe010ff400 refcount:1 mapcount:0 mapping:0000000000000000 index:0xffffff8043fd0e00 pfn:0x43fd0
> head:fffffffe010ff400 order:3 entire_mapcount:0 nr_pages_mapped:0 pincount:0
> flags: 0x4000000000000840(slab|head|zone=1)
> page_type: 0xffffffff()
> raw: 4000000000000840 ffffff8040002f40 ffffff8040000a50 ffffff8040000a50
> raw: ffffff8043fd0e00 0000000000150002 00000001ffffffff 0000000000000000
> page dumped because: kasan: bad access detected
>
> Memory state around the buggy address:
>   ffffff8043fd0880: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
>   ffffff8043fd0900: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
>> ffffff8043fd0980: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
>                     ^
>   ffffff8043fd0a00: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc
>   ffffff8043fd0a80: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc
> ==================================================================
>
> Fixes: fc0c209c147f ("clk: Allow parents to be specified without string names")
> Signed-off-by: Yun Zhou <yun.zhou@windriver.com>
> Reviewed-by: Brian Masney <bmasney@redhat.com>
> ---
>   drivers/clk/clk.c | 5 +++--
>   1 file changed, 3 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/clk/clk.c b/drivers/clk/clk.c
> index b821b2cdb155..b93f38de4cac 100644
> --- a/drivers/clk/clk.c
> +++ b/drivers/clk/clk.c
> @@ -4254,7 +4254,6 @@ static int clk_core_populate_parent_map(struct clk_core *core,
>   	 * having a cache of names/clk_hw pointers to clk_core pointers.
>   	 */
>   	parents = kcalloc(num_parents, sizeof(*parents), GFP_KERNEL);
> -	core->parents = parents;
>   	if (!parents)
>   		return -ENOMEM;
>   
> @@ -4295,6 +4294,8 @@ static int clk_core_populate_parent_map(struct clk_core *core,
>   		}
>   	}
>   
> +	core->parents = parents;
> +
>   	return 0;
>   }
>   
> @@ -4302,7 +4303,7 @@ static void clk_core_free_parent_map(struct clk_core *core)
>   {
>   	int i = core->num_parents;
>   
> -	if (!core->num_parents)
> +	if (!core->parents)
>   		return;
>   
>   	while (--i >= 0) {