Re: [PATCH] workqueue: Try to catch flush_work() without INIT_WORK().

[Daniel Jordan]

On Sat, Jan 19, 2019 at 11:41:22AM +0900, Tetsuo Handa wrote:
> On 2019/01/19 4:48, Daniel Jordan wrote:
> > On Sat, Jan 19, 2019 at 02:04:58AM +0900, Tetsuo Handa wrote:
> > __queue_work has a sanity check already for work, but using list_empty.  Seems
> > slightly better to be consistent?
> > 
> 
> list_empty() won't work, for "struct work_struct" is embedded into a struct
> which is allocated by kzalloc().

Please check list_empty's definition again, it compares the address of the node
to its next pointer, so it should work for a zeroed node.  I'll reiterate that
it seems slightly better to be consistent in "is work_struct initialized?"
checks, but it's not a big deal and I'm fine either way.

> From 1bbf8d9c7aaef78d7f483d2131261162485c6f72 Mon Sep 17 00:00:00 2001
> From: Tetsuo Handa <penguin-kernel@xxxxxxxxxxxxxxxxxxx>
> Date: Sat, 19 Jan 2019 01:38:53 +0900
> Subject: [PATCH] drm/vkms: Fix flush_work() without INIT_WORK().
> 
> syzbot is hitting a lockdep warning [1] because flush_work() is called
> without INIT_WORK() after kzalloc() at vkms_atomic_crtc_reset().
> 
> Commit 6c234fe37c57627a ("drm/vkms: Implement CRC debugfs API") added
> INIT_WORK() to only vkms_atomic_crtc_duplicate_state() side. Assuming
> that lifecycle of crc_work is appropriately managed, fix this problem
> by adding INIT_WORK() to vkms_atomic_crtc_reset() side.
> 
> [1] https://syzkaller.appspot.com/bug?id=a5954455fcfa51c29ca2ab55b203076337e1c770
> 
> Reported-and-tested-by: syzbot <syzbot+12f1b031b6da017e34f8@xxxxxxxxxxxxxxxxxxxxxxxxx>
> Signed-off-by: Tetsuo Handa <penguin-kernel@xxxxxxxxxxxxxxxxxxx>
> ---
>  drivers/gpu/drm/vkms/vkms_crtc.c | 1 +
>  1 file changed, 1 insertion(+)
> 
> diff --git a/drivers/gpu/drm/vkms/vkms_crtc.c b/drivers/gpu/drm/vkms/vkms_crtc.c
> index 177bbcb..3c37d8c 100644
> --- a/drivers/gpu/drm/vkms/vkms_crtc.c
> +++ b/drivers/gpu/drm/vkms/vkms_crtc.c
> @@ -104,6 +104,7 @@ static void vkms_atomic_crtc_reset(struct drm_crtc *crtc)
>  	vkms_state = kzalloc(sizeof(*vkms_state), GFP_KERNEL);
>  	if (!vkms_state)
>  		return;
> +	INIT_WORK(&vkms_state->crc_work, vkms_crc_work_handle);
>  
>  	crtc->state = &vkms_state->base;
>  	crtc->state->crtc = crtc;

Great, thanks, I think mentioning this path would improve the changelog.