dri-devel.lists.freedesktop.org archive mirror
 help / color / mirror / Atom feed
* [PATCH] drm/msm/kms: Make a lock_class_key for each crtc mutex
@ 2021-01-25 23:49 Stephen Boyd
  2021-01-28 16:39 ` Rob Clark
  2021-02-02 15:46 ` Daniel Vetter
  0 siblings, 2 replies; 10+ messages in thread
From: Stephen Boyd @ 2021-01-25 23:49 UTC (permalink / raw)
  To: Rob Clark
  Cc: linux-arm-msm, freedreno, linux-kernel, dri-devel, Krishna Manikandan

Lockdep complains about an AA deadlock when rebooting the device.

============================================
WARNING: possible recursive locking detected
5.4.91 #1 Not tainted
--------------------------------------------
reboot/5213 is trying to acquire lock:
ffffff80d13391b0 (&kms->commit_lock[i]){+.+.}, at: lock_crtcs+0x60/0xa4

but task is already holding lock:
ffffff80d1339110 (&kms->commit_lock[i]){+.+.}, at: lock_crtcs+0x60/0xa4

other info that might help us debug this:
Possible unsafe locking scenario:

CPU0
----
lock(&kms->commit_lock[i]);
lock(&kms->commit_lock[i]);

*** DEADLOCK ***

May be due to missing lock nesting notation

6 locks held by reboot/5213:
__arm64_sys_reboot+0x148/0x2a0
device_shutdown+0x10c/0x2c4
drm_atomic_helper_shutdown+0x48/0xfc
modeset_lock+0x120/0x24c
lock_crtcs+0x60/0xa4

stack backtrace:
CPU: 4 PID: 5213 Comm: reboot Not tainted 5.4.91 #1
Hardware name: Google Pompom (rev1) with LTE (DT)
Call trace:
dump_backtrace+0x0/0x1dc
show_stack+0x24/0x30
dump_stack+0xfc/0x1a8
__lock_acquire+0xcd0/0x22b8
lock_acquire+0x1ec/0x240
__mutex_lock_common+0xe0/0xc84
mutex_lock_nested+0x48/0x58
lock_crtcs+0x60/0xa4
msm_atomic_commit_tail+0x348/0x570
commit_tail+0xdc/0x178
drm_atomic_helper_commit+0x160/0x168
drm_atomic_commit+0x68/0x80

This is because lockdep thinks all the locks taken in lock_crtcs() are
the same lock, when they actually aren't. That's because we call
mutex_init() in msm_kms_init() and that assigns on static key for every
lock initialized in this loop. Let's allocate a dynamic number of
lock_class_keys and assign them to each lock so that lockdep can figure
out an AA deadlock isn't possible here.

Fixes: b3d91800d9ac ("drm/msm: Fix race condition in msm driver with async layer updates")
Cc: Krishna Manikandan <mkrishn@codeaurora.org>
Signed-off-by: Stephen Boyd <swboyd@chromium.org>
---
 drivers/gpu/drm/msm/msm_kms.h | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/msm/msm_kms.h b/drivers/gpu/drm/msm/msm_kms.h
index d8151a89e163..4735251a394d 100644
--- a/drivers/gpu/drm/msm/msm_kms.h
+++ b/drivers/gpu/drm/msm/msm_kms.h
@@ -157,6 +157,7 @@ struct msm_kms {
 	 * from the crtc's pending_timer close to end of the frame:
 	 */
 	struct mutex commit_lock[MAX_CRTCS];
+	struct lock_class_key commit_lock_keys[MAX_CRTCS];
 	unsigned pending_crtc_mask;
 	struct msm_pending_timer pending_timers[MAX_CRTCS];
 };
@@ -166,8 +167,11 @@ static inline int msm_kms_init(struct msm_kms *kms,
 {
 	unsigned i, ret;
 
-	for (i = 0; i < ARRAY_SIZE(kms->commit_lock); i++)
-		mutex_init(&kms->commit_lock[i]);
+	for (i = 0; i < ARRAY_SIZE(kms->commit_lock); i++) {
+		lockdep_register_key(&kms->commit_lock_keys[i]);
+		__mutex_init(&kms->commit_lock[i], "&kms->commit_lock[i]",
+			     &kms->commit_lock_keys[i]);
+	}
 
 	kms->funcs = funcs;
 

base-commit: 19c329f6808995b142b3966301f217c831e7cf31
-- 
https://chromeos.dev

_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

^ permalink raw reply related	[flat|nested] 10+ messages in thread

* Re: [PATCH] drm/msm/kms: Make a lock_class_key for each crtc mutex
  2021-01-25 23:49 [PATCH] drm/msm/kms: Make a lock_class_key for each crtc mutex Stephen Boyd
@ 2021-01-28 16:39 ` Rob Clark
  2021-02-02 15:46 ` Daniel Vetter
  1 sibling, 0 replies; 10+ messages in thread
From: Rob Clark @ 2021-01-28 16:39 UTC (permalink / raw)
  To: Stephen Boyd
  Cc: linux-arm-msm, freedreno, Linux Kernel Mailing List, dri-devel,
	Krishna Manikandan

On Mon, Jan 25, 2021 at 3:49 PM Stephen Boyd <swboyd@chromium.org> wrote:
>
> Lockdep complains about an AA deadlock when rebooting the device.
>
> ============================================
> WARNING: possible recursive locking detected
> 5.4.91 #1 Not tainted
> --------------------------------------------
> reboot/5213 is trying to acquire lock:
> ffffff80d13391b0 (&kms->commit_lock[i]){+.+.}, at: lock_crtcs+0x60/0xa4
>
> but task is already holding lock:
> ffffff80d1339110 (&kms->commit_lock[i]){+.+.}, at: lock_crtcs+0x60/0xa4
>
> other info that might help us debug this:
> Possible unsafe locking scenario:
>
> CPU0
> ----
> lock(&kms->commit_lock[i]);
> lock(&kms->commit_lock[i]);
>
> *** DEADLOCK ***
>
> May be due to missing lock nesting notation
>
> 6 locks held by reboot/5213:
> __arm64_sys_reboot+0x148/0x2a0
> device_shutdown+0x10c/0x2c4
> drm_atomic_helper_shutdown+0x48/0xfc
> modeset_lock+0x120/0x24c
> lock_crtcs+0x60/0xa4
>
> stack backtrace:
> CPU: 4 PID: 5213 Comm: reboot Not tainted 5.4.91 #1
> Hardware name: Google Pompom (rev1) with LTE (DT)
> Call trace:
> dump_backtrace+0x0/0x1dc
> show_stack+0x24/0x30
> dump_stack+0xfc/0x1a8
> __lock_acquire+0xcd0/0x22b8
> lock_acquire+0x1ec/0x240
> __mutex_lock_common+0xe0/0xc84
> mutex_lock_nested+0x48/0x58
> lock_crtcs+0x60/0xa4
> msm_atomic_commit_tail+0x348/0x570
> commit_tail+0xdc/0x178
> drm_atomic_helper_commit+0x160/0x168
> drm_atomic_commit+0x68/0x80
>
> This is because lockdep thinks all the locks taken in lock_crtcs() are
> the same lock, when they actually aren't. That's because we call
> mutex_init() in msm_kms_init() and that assigns on static key for every

nit, s/on/one/ ?

BR,
-R

> lock initialized in this loop. Let's allocate a dynamic number of
> lock_class_keys and assign them to each lock so that lockdep can figure
> out an AA deadlock isn't possible here.
>
> Fixes: b3d91800d9ac ("drm/msm: Fix race condition in msm driver with async layer updates")
> Cc: Krishna Manikandan <mkrishn@codeaurora.org>
> Signed-off-by: Stephen Boyd <swboyd@chromium.org>
> ---
>  drivers/gpu/drm/msm/msm_kms.h | 8 ++++++--
>  1 file changed, 6 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/gpu/drm/msm/msm_kms.h b/drivers/gpu/drm/msm/msm_kms.h
> index d8151a89e163..4735251a394d 100644
> --- a/drivers/gpu/drm/msm/msm_kms.h
> +++ b/drivers/gpu/drm/msm/msm_kms.h
> @@ -157,6 +157,7 @@ struct msm_kms {
>          * from the crtc's pending_timer close to end of the frame:
>          */
>         struct mutex commit_lock[MAX_CRTCS];
> +       struct lock_class_key commit_lock_keys[MAX_CRTCS];
>         unsigned pending_crtc_mask;
>         struct msm_pending_timer pending_timers[MAX_CRTCS];
>  };
> @@ -166,8 +167,11 @@ static inline int msm_kms_init(struct msm_kms *kms,
>  {
>         unsigned i, ret;
>
> -       for (i = 0; i < ARRAY_SIZE(kms->commit_lock); i++)
> -               mutex_init(&kms->commit_lock[i]);
> +       for (i = 0; i < ARRAY_SIZE(kms->commit_lock); i++) {
> +               lockdep_register_key(&kms->commit_lock_keys[i]);
> +               __mutex_init(&kms->commit_lock[i], "&kms->commit_lock[i]",
> +                            &kms->commit_lock_keys[i]);
> +       }
>
>         kms->funcs = funcs;
>
>
> base-commit: 19c329f6808995b142b3966301f217c831e7cf31
> --
> https://chromeos.dev
>
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

^ permalink raw reply	[flat|nested] 10+ messages in thread

* Re: [PATCH] drm/msm/kms: Make a lock_class_key for each crtc mutex
  2021-01-25 23:49 [PATCH] drm/msm/kms: Make a lock_class_key for each crtc mutex Stephen Boyd
  2021-01-28 16:39 ` Rob Clark
@ 2021-02-02 15:46 ` Daniel Vetter
  2021-02-02 16:51   ` Rob Clark
  1 sibling, 1 reply; 10+ messages in thread
From: Daniel Vetter @ 2021-02-02 15:46 UTC (permalink / raw)
  To: Stephen Boyd
  Cc: Krishna Manikandan, linux-arm-msm, linux-kernel, dri-devel, freedreno

On Mon, Jan 25, 2021 at 03:49:01PM -0800, Stephen Boyd wrote:
> Lockdep complains about an AA deadlock when rebooting the device.
> 
> ============================================
> WARNING: possible recursive locking detected
> 5.4.91 #1 Not tainted
> --------------------------------------------
> reboot/5213 is trying to acquire lock:
> ffffff80d13391b0 (&kms->commit_lock[i]){+.+.}, at: lock_crtcs+0x60/0xa4
> 
> but task is already holding lock:
> ffffff80d1339110 (&kms->commit_lock[i]){+.+.}, at: lock_crtcs+0x60/0xa4
> 
> other info that might help us debug this:
> Possible unsafe locking scenario:
> 
> CPU0
> ----
> lock(&kms->commit_lock[i]);
> lock(&kms->commit_lock[i]);
> 
> *** DEADLOCK ***
> 
> May be due to missing lock nesting notation
> 
> 6 locks held by reboot/5213:
> __arm64_sys_reboot+0x148/0x2a0
> device_shutdown+0x10c/0x2c4
> drm_atomic_helper_shutdown+0x48/0xfc
> modeset_lock+0x120/0x24c
> lock_crtcs+0x60/0xa4
> 
> stack backtrace:
> CPU: 4 PID: 5213 Comm: reboot Not tainted 5.4.91 #1
> Hardware name: Google Pompom (rev1) with LTE (DT)
> Call trace:
> dump_backtrace+0x0/0x1dc
> show_stack+0x24/0x30
> dump_stack+0xfc/0x1a8
> __lock_acquire+0xcd0/0x22b8
> lock_acquire+0x1ec/0x240
> __mutex_lock_common+0xe0/0xc84
> mutex_lock_nested+0x48/0x58
> lock_crtcs+0x60/0xa4
> msm_atomic_commit_tail+0x348/0x570
> commit_tail+0xdc/0x178
> drm_atomic_helper_commit+0x160/0x168
> drm_atomic_commit+0x68/0x80
> 
> This is because lockdep thinks all the locks taken in lock_crtcs() are
> the same lock, when they actually aren't. That's because we call
> mutex_init() in msm_kms_init() and that assigns on static key for every
> lock initialized in this loop. Let's allocate a dynamic number of
> lock_class_keys and assign them to each lock so that lockdep can figure
> out an AA deadlock isn't possible here.
> 
> Fixes: b3d91800d9ac ("drm/msm: Fix race condition in msm driver with async layer updates")
> Cc: Krishna Manikandan <mkrishn@codeaurora.org>
> Signed-off-by: Stephen Boyd <swboyd@chromium.org>

This smells like throwing more bad after initial bad code ...

First a rant: https://blog.ffwll.ch/2020/08/lockdep-false-positives.html

Yes I know the locking you're doing here is correct, but that goes to the
second issue: Why is this needed? atomic_async_update helpers are supposed
to take care of ordering fun like this, if they're not, we need to address
things there. The problem that

commit b3d91800d9ac35014e0349292273a6fa7938d402
Author: Krishna Manikandan <mkrishn@codeaurora.org>
Date:   Fri Oct 16 19:40:43 2020 +0530

    drm/msm: Fix race condition in msm driver with async layer updates

is _the_ reason we have drm_crtc_commit to track stuff, and Maxime has
recently rolled out a pile of changes to vc4 to use these things
correctly. Hacking some glorious hand-rolled locking for synchronization
of updates really should be the exception for kms drivers, not the rule.
And this one here doesn't look like an exception by far (the one legit I
know of is the locking issues amdgpu has between atomic_commit_tail and
gpu reset, and that one is really nasty, so not going to get fixed in
helpers, ever).

Cheers, Daniel

> ---
>  drivers/gpu/drm/msm/msm_kms.h | 8 ++++++--
>  1 file changed, 6 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/gpu/drm/msm/msm_kms.h b/drivers/gpu/drm/msm/msm_kms.h
> index d8151a89e163..4735251a394d 100644
> --- a/drivers/gpu/drm/msm/msm_kms.h
> +++ b/drivers/gpu/drm/msm/msm_kms.h
> @@ -157,6 +157,7 @@ struct msm_kms {
>  	 * from the crtc's pending_timer close to end of the frame:
>  	 */
>  	struct mutex commit_lock[MAX_CRTCS];
> +	struct lock_class_key commit_lock_keys[MAX_CRTCS];
>  	unsigned pending_crtc_mask;
>  	struct msm_pending_timer pending_timers[MAX_CRTCS];
>  };
> @@ -166,8 +167,11 @@ static inline int msm_kms_init(struct msm_kms *kms,
>  {
>  	unsigned i, ret;
>  
> -	for (i = 0; i < ARRAY_SIZE(kms->commit_lock); i++)
> -		mutex_init(&kms->commit_lock[i]);
> +	for (i = 0; i < ARRAY_SIZE(kms->commit_lock); i++) {
> +		lockdep_register_key(&kms->commit_lock_keys[i]);
> +		__mutex_init(&kms->commit_lock[i], "&kms->commit_lock[i]",
> +			     &kms->commit_lock_keys[i]);
> +	}
>  
>  	kms->funcs = funcs;
>  
> 
> base-commit: 19c329f6808995b142b3966301f217c831e7cf31
> -- 
> https://chromeos.dev
> 
> _______________________________________________
> dri-devel mailing list
> dri-devel@lists.freedesktop.org
> https://lists.freedesktop.org/mailman/listinfo/dri-devel

-- 
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

^ permalink raw reply	[flat|nested] 10+ messages in thread

* Re: [PATCH] drm/msm/kms: Make a lock_class_key for each crtc mutex
  2021-02-02 15:46 ` Daniel Vetter
@ 2021-02-02 16:51   ` Rob Clark
  2021-02-03 10:10     ` Daniel Vetter
  0 siblings, 1 reply; 10+ messages in thread
From: Rob Clark @ 2021-02-02 16:51 UTC (permalink / raw)
  To: Stephen Boyd, Rob Clark, linux-arm-msm, freedreno,
	Linux Kernel Mailing List, dri-devel, Krishna Manikandan

On Tue, Feb 2, 2021 at 7:46 AM Daniel Vetter <daniel@ffwll.ch> wrote:
>
> On Mon, Jan 25, 2021 at 03:49:01PM -0800, Stephen Boyd wrote:
> > Lockdep complains about an AA deadlock when rebooting the device.
> >
> > ============================================
> > WARNING: possible recursive locking detected
> > 5.4.91 #1 Not tainted
> > --------------------------------------------
> > reboot/5213 is trying to acquire lock:
> > ffffff80d13391b0 (&kms->commit_lock[i]){+.+.}, at: lock_crtcs+0x60/0xa4
> >
> > but task is already holding lock:
> > ffffff80d1339110 (&kms->commit_lock[i]){+.+.}, at: lock_crtcs+0x60/0xa4
> >
> > other info that might help us debug this:
> > Possible unsafe locking scenario:
> >
> > CPU0
> > ----
> > lock(&kms->commit_lock[i]);
> > lock(&kms->commit_lock[i]);
> >
> > *** DEADLOCK ***
> >
> > May be due to missing lock nesting notation
> >
> > 6 locks held by reboot/5213:
> > __arm64_sys_reboot+0x148/0x2a0
> > device_shutdown+0x10c/0x2c4
> > drm_atomic_helper_shutdown+0x48/0xfc
> > modeset_lock+0x120/0x24c
> > lock_crtcs+0x60/0xa4
> >
> > stack backtrace:
> > CPU: 4 PID: 5213 Comm: reboot Not tainted 5.4.91 #1
> > Hardware name: Google Pompom (rev1) with LTE (DT)
> > Call trace:
> > dump_backtrace+0x0/0x1dc
> > show_stack+0x24/0x30
> > dump_stack+0xfc/0x1a8
> > __lock_acquire+0xcd0/0x22b8
> > lock_acquire+0x1ec/0x240
> > __mutex_lock_common+0xe0/0xc84
> > mutex_lock_nested+0x48/0x58
> > lock_crtcs+0x60/0xa4
> > msm_atomic_commit_tail+0x348/0x570
> > commit_tail+0xdc/0x178
> > drm_atomic_helper_commit+0x160/0x168
> > drm_atomic_commit+0x68/0x80
> >
> > This is because lockdep thinks all the locks taken in lock_crtcs() are
> > the same lock, when they actually aren't. That's because we call
> > mutex_init() in msm_kms_init() and that assigns on static key for every
> > lock initialized in this loop. Let's allocate a dynamic number of
> > lock_class_keys and assign them to each lock so that lockdep can figure
> > out an AA deadlock isn't possible here.
> >
> > Fixes: b3d91800d9ac ("drm/msm: Fix race condition in msm driver with async layer updates")
> > Cc: Krishna Manikandan <mkrishn@codeaurora.org>
> > Signed-off-by: Stephen Boyd <swboyd@chromium.org>
>
> This smells like throwing more bad after initial bad code ...
>
> First a rant: https://blog.ffwll.ch/2020/08/lockdep-false-positives.html
>
> Yes I know the locking you're doing here is correct, but that goes to the
> second issue: Why is this needed? atomic_async_update helpers are supposed
> to take care of ordering fun like this, if they're not, we need to address
> things there. The problem that

Maybe a better solution would be helper awareness of hw that has
double-buffered state and flush bits.. ie. something that looks a bit
more like the internal kms fxn ptrs. Currently the locking is
protecting something that the atomic helpers are not aware of, ie.
we've already written previous cursor updates to hw and are just
waiting until close to vblank to write the flush bits

But, we've been over this before. I'd tried various approaches.. the
current scheme replaces seanpaul's earlier attempts to do it the
"helper" way.  The current implementation does the best job of
avoiding fps drops when the legacy cursor uapi is in play.  (And yes,
legacy cursor + atomic ioctls is maybe not the greatest, but it is
what it is.)

BR,
-R

>
> commit b3d91800d9ac35014e0349292273a6fa7938d402
> Author: Krishna Manikandan <mkrishn@codeaurora.org>
> Date:   Fri Oct 16 19:40:43 2020 +0530
>
>     drm/msm: Fix race condition in msm driver with async layer updates
>
> is _the_ reason we have drm_crtc_commit to track stuff, and Maxime has
> recently rolled out a pile of changes to vc4 to use these things
> correctly. Hacking some glorious hand-rolled locking for synchronization
> of updates really should be the exception for kms drivers, not the rule.
> And this one here doesn't look like an exception by far (the one legit I
> know of is the locking issues amdgpu has between atomic_commit_tail and
> gpu reset, and that one is really nasty, so not going to get fixed in
> helpers, ever).
>
> Cheers, Daniel
>
> > ---
> >  drivers/gpu/drm/msm/msm_kms.h | 8 ++++++--
> >  1 file changed, 6 insertions(+), 2 deletions(-)
> >
> > diff --git a/drivers/gpu/drm/msm/msm_kms.h b/drivers/gpu/drm/msm/msm_kms.h
> > index d8151a89e163..4735251a394d 100644
> > --- a/drivers/gpu/drm/msm/msm_kms.h
> > +++ b/drivers/gpu/drm/msm/msm_kms.h
> > @@ -157,6 +157,7 @@ struct msm_kms {
> >        * from the crtc's pending_timer close to end of the frame:
> >        */
> >       struct mutex commit_lock[MAX_CRTCS];
> > +     struct lock_class_key commit_lock_keys[MAX_CRTCS];
> >       unsigned pending_crtc_mask;
> >       struct msm_pending_timer pending_timers[MAX_CRTCS];
> >  };
> > @@ -166,8 +167,11 @@ static inline int msm_kms_init(struct msm_kms *kms,
> >  {
> >       unsigned i, ret;
> >
> > -     for (i = 0; i < ARRAY_SIZE(kms->commit_lock); i++)
> > -             mutex_init(&kms->commit_lock[i]);
> > +     for (i = 0; i < ARRAY_SIZE(kms->commit_lock); i++) {
> > +             lockdep_register_key(&kms->commit_lock_keys[i]);
> > +             __mutex_init(&kms->commit_lock[i], "&kms->commit_lock[i]",
> > +                          &kms->commit_lock_keys[i]);
> > +     }
> >
> >       kms->funcs = funcs;
> >
> >
> > base-commit: 19c329f6808995b142b3966301f217c831e7cf31
> > --
> > https://chromeos.dev
> >
> > _______________________________________________
> > dri-devel mailing list
> > dri-devel@lists.freedesktop.org
> > https://lists.freedesktop.org/mailman/listinfo/dri-devel
>
> --
> Daniel Vetter
> Software Engineer, Intel Corporation
> http://blog.ffwll.ch
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

^ permalink raw reply	[flat|nested] 10+ messages in thread

* Re: [PATCH] drm/msm/kms: Make a lock_class_key for each crtc mutex
  2021-02-02 16:51   ` Rob Clark
@ 2021-02-03 10:10     ` Daniel Vetter
  2021-02-03 17:29       ` Rob Clark
  0 siblings, 1 reply; 10+ messages in thread
From: Daniel Vetter @ 2021-02-03 10:10 UTC (permalink / raw)
  To: Rob Clark
  Cc: Krishna Manikandan, linux-arm-msm, Linux Kernel Mailing List,
	dri-devel, Stephen Boyd, freedreno

On Tue, Feb 02, 2021 at 08:51:25AM -0800, Rob Clark wrote:
> On Tue, Feb 2, 2021 at 7:46 AM Daniel Vetter <daniel@ffwll.ch> wrote:
> >
> > On Mon, Jan 25, 2021 at 03:49:01PM -0800, Stephen Boyd wrote:
> > > Lockdep complains about an AA deadlock when rebooting the device.
> > >
> > > ============================================
> > > WARNING: possible recursive locking detected
> > > 5.4.91 #1 Not tainted
> > > --------------------------------------------
> > > reboot/5213 is trying to acquire lock:
> > > ffffff80d13391b0 (&kms->commit_lock[i]){+.+.}, at: lock_crtcs+0x60/0xa4
> > >
> > > but task is already holding lock:
> > > ffffff80d1339110 (&kms->commit_lock[i]){+.+.}, at: lock_crtcs+0x60/0xa4
> > >
> > > other info that might help us debug this:
> > > Possible unsafe locking scenario:
> > >
> > > CPU0
> > > ----
> > > lock(&kms->commit_lock[i]);
> > > lock(&kms->commit_lock[i]);
> > >
> > > *** DEADLOCK ***
> > >
> > > May be due to missing lock nesting notation
> > >
> > > 6 locks held by reboot/5213:
> > > __arm64_sys_reboot+0x148/0x2a0
> > > device_shutdown+0x10c/0x2c4
> > > drm_atomic_helper_shutdown+0x48/0xfc
> > > modeset_lock+0x120/0x24c
> > > lock_crtcs+0x60/0xa4
> > >
> > > stack backtrace:
> > > CPU: 4 PID: 5213 Comm: reboot Not tainted 5.4.91 #1
> > > Hardware name: Google Pompom (rev1) with LTE (DT)
> > > Call trace:
> > > dump_backtrace+0x0/0x1dc
> > > show_stack+0x24/0x30
> > > dump_stack+0xfc/0x1a8
> > > __lock_acquire+0xcd0/0x22b8
> > > lock_acquire+0x1ec/0x240
> > > __mutex_lock_common+0xe0/0xc84
> > > mutex_lock_nested+0x48/0x58
> > > lock_crtcs+0x60/0xa4
> > > msm_atomic_commit_tail+0x348/0x570
> > > commit_tail+0xdc/0x178
> > > drm_atomic_helper_commit+0x160/0x168
> > > drm_atomic_commit+0x68/0x80
> > >
> > > This is because lockdep thinks all the locks taken in lock_crtcs() are
> > > the same lock, when they actually aren't. That's because we call
> > > mutex_init() in msm_kms_init() and that assigns on static key for every
> > > lock initialized in this loop. Let's allocate a dynamic number of
> > > lock_class_keys and assign them to each lock so that lockdep can figure
> > > out an AA deadlock isn't possible here.
> > >
> > > Fixes: b3d91800d9ac ("drm/msm: Fix race condition in msm driver with async layer updates")
> > > Cc: Krishna Manikandan <mkrishn@codeaurora.org>
> > > Signed-off-by: Stephen Boyd <swboyd@chromium.org>
> >
> > This smells like throwing more bad after initial bad code ...
> >
> > First a rant: https://blog.ffwll.ch/2020/08/lockdep-false-positives.html

Some technical on the patch itself: I think you want
mutex_lock_nested(crtc->lock, drm_crtc_index(crtc)), not your own locking
classes hand-rolled. It's defacto the same, but much more obviously
correct since self-documenting.

> > Yes I know the locking you're doing here is correct, but that goes to the
> > second issue: Why is this needed? atomic_async_update helpers are supposed
> > to take care of ordering fun like this, if they're not, we need to address
> > things there. The problem that
> 
> Maybe a better solution would be helper awareness of hw that has
> double-buffered state and flush bits.. ie. something that looks a bit
> more like the internal kms fxn ptrs. Currently the locking is
> protecting something that the atomic helpers are not aware of, ie.
> we've already written previous cursor updates to hw and are just
> waiting until close to vblank to write the flush bits
> 
> But, we've been over this before. I'd tried various approaches.. the
> current scheme replaces seanpaul's earlier attempts to do it the
> "helper" way.  The current implementation does the best job of
> avoiding fps drops when the legacy cursor uapi is in play.  (And yes,
> legacy cursor + atomic ioctls is maybe not the greatest, but it is
> what it is.)

I didn't read enough of the context and got confused, the flush handling
looks all reasonable and obviously needs some locks to avoid races with
updates.

It still looks a bit strange that you need multi-crtc locks for cursor
(generally this stuff is supposed to be solved with ordering) and why the
async helpers don't work since msm has something that's pretty close
itself. Atomic+cursor is a bit nasty, but if every driver hacks this
together themselves then there's not much chance of this ever really
working well across the board. And aside from the flush bit instead of
automatic double buffering (which you're just emulating) there's not
really anything special with msm afaics. So pretty sure that if this
doesn't work for msm, it doesn't work anywhere else.
-Daniel

> 
> BR,
> -R
> 
> >
> > commit b3d91800d9ac35014e0349292273a6fa7938d402
> > Author: Krishna Manikandan <mkrishn@codeaurora.org>
> > Date:   Fri Oct 16 19:40:43 2020 +0530
> >
> >     drm/msm: Fix race condition in msm driver with async layer updates
> >
> > is _the_ reason we have drm_crtc_commit to track stuff, and Maxime has
> > recently rolled out a pile of changes to vc4 to use these things
> > correctly. Hacking some glorious hand-rolled locking for synchronization
> > of updates really should be the exception for kms drivers, not the rule.
> > And this one here doesn't look like an exception by far (the one legit I
> > know of is the locking issues amdgpu has between atomic_commit_tail and
> > gpu reset, and that one is really nasty, so not going to get fixed in
> > helpers, ever).
> >
> > Cheers, Daniel
> >
> > > ---
> > >  drivers/gpu/drm/msm/msm_kms.h | 8 ++++++--
> > >  1 file changed, 6 insertions(+), 2 deletions(-)
> > >
> > > diff --git a/drivers/gpu/drm/msm/msm_kms.h b/drivers/gpu/drm/msm/msm_kms.h
> > > index d8151a89e163..4735251a394d 100644
> > > --- a/drivers/gpu/drm/msm/msm_kms.h
> > > +++ b/drivers/gpu/drm/msm/msm_kms.h
> > > @@ -157,6 +157,7 @@ struct msm_kms {
> > >        * from the crtc's pending_timer close to end of the frame:
> > >        */
> > >       struct mutex commit_lock[MAX_CRTCS];
> > > +     struct lock_class_key commit_lock_keys[MAX_CRTCS];
> > >       unsigned pending_crtc_mask;
> > >       struct msm_pending_timer pending_timers[MAX_CRTCS];
> > >  };
> > > @@ -166,8 +167,11 @@ static inline int msm_kms_init(struct msm_kms *kms,
> > >  {
> > >       unsigned i, ret;
> > >
> > > -     for (i = 0; i < ARRAY_SIZE(kms->commit_lock); i++)
> > > -             mutex_init(&kms->commit_lock[i]);
> > > +     for (i = 0; i < ARRAY_SIZE(kms->commit_lock); i++) {
> > > +             lockdep_register_key(&kms->commit_lock_keys[i]);
> > > +             __mutex_init(&kms->commit_lock[i], "&kms->commit_lock[i]",
> > > +                          &kms->commit_lock_keys[i]);
> > > +     }
> > >
> > >       kms->funcs = funcs;
> > >
> > >
> > > base-commit: 19c329f6808995b142b3966301f217c831e7cf31
> > > --
> > > https://chromeos.dev
> > >
> > > _______________________________________________
> > > dri-devel mailing list
> > > dri-devel@lists.freedesktop.org
> > > https://lists.freedesktop.org/mailman/listinfo/dri-devel
> >
> > --
> > Daniel Vetter
> > Software Engineer, Intel Corporation
> > http://blog.ffwll.ch
> _______________________________________________
> dri-devel mailing list
> dri-devel@lists.freedesktop.org
> https://lists.freedesktop.org/mailman/listinfo/dri-devel

-- 
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

^ permalink raw reply	[flat|nested] 10+ messages in thread

* Re: [PATCH] drm/msm/kms: Make a lock_class_key for each crtc mutex
  2021-02-03 10:10     ` Daniel Vetter
@ 2021-02-03 17:29       ` Rob Clark
  2021-02-03 21:58         ` Stephen Boyd
  0 siblings, 1 reply; 10+ messages in thread
From: Rob Clark @ 2021-02-03 17:29 UTC (permalink / raw)
  To: Rob Clark, Stephen Boyd, linux-arm-msm, freedreno,
	Linux Kernel Mailing List, dri-devel, Krishna Manikandan

On Wed, Feb 3, 2021 at 2:10 AM Daniel Vetter <daniel@ffwll.ch> wrote:
>
> On Tue, Feb 02, 2021 at 08:51:25AM -0800, Rob Clark wrote:
> > On Tue, Feb 2, 2021 at 7:46 AM Daniel Vetter <daniel@ffwll.ch> wrote:
> > >
> > > On Mon, Jan 25, 2021 at 03:49:01PM -0800, Stephen Boyd wrote:
> > > > Lockdep complains about an AA deadlock when rebooting the device.
> > > >
> > > > ============================================
> > > > WARNING: possible recursive locking detected
> > > > 5.4.91 #1 Not tainted
> > > > --------------------------------------------
> > > > reboot/5213 is trying to acquire lock:
> > > > ffffff80d13391b0 (&kms->commit_lock[i]){+.+.}, at: lock_crtcs+0x60/0xa4
> > > >
> > > > but task is already holding lock:
> > > > ffffff80d1339110 (&kms->commit_lock[i]){+.+.}, at: lock_crtcs+0x60/0xa4
> > > >
> > > > other info that might help us debug this:
> > > > Possible unsafe locking scenario:
> > > >
> > > > CPU0
> > > > ----
> > > > lock(&kms->commit_lock[i]);
> > > > lock(&kms->commit_lock[i]);
> > > >
> > > > *** DEADLOCK ***
> > > >
> > > > May be due to missing lock nesting notation
> > > >
> > > > 6 locks held by reboot/5213:
> > > > __arm64_sys_reboot+0x148/0x2a0
> > > > device_shutdown+0x10c/0x2c4
> > > > drm_atomic_helper_shutdown+0x48/0xfc
> > > > modeset_lock+0x120/0x24c
> > > > lock_crtcs+0x60/0xa4
> > > >
> > > > stack backtrace:
> > > > CPU: 4 PID: 5213 Comm: reboot Not tainted 5.4.91 #1
> > > > Hardware name: Google Pompom (rev1) with LTE (DT)
> > > > Call trace:
> > > > dump_backtrace+0x0/0x1dc
> > > > show_stack+0x24/0x30
> > > > dump_stack+0xfc/0x1a8
> > > > __lock_acquire+0xcd0/0x22b8
> > > > lock_acquire+0x1ec/0x240
> > > > __mutex_lock_common+0xe0/0xc84
> > > > mutex_lock_nested+0x48/0x58
> > > > lock_crtcs+0x60/0xa4
> > > > msm_atomic_commit_tail+0x348/0x570
> > > > commit_tail+0xdc/0x178
> > > > drm_atomic_helper_commit+0x160/0x168
> > > > drm_atomic_commit+0x68/0x80
> > > >
> > > > This is because lockdep thinks all the locks taken in lock_crtcs() are
> > > > the same lock, when they actually aren't. That's because we call
> > > > mutex_init() in msm_kms_init() and that assigns on static key for every
> > > > lock initialized in this loop. Let's allocate a dynamic number of
> > > > lock_class_keys and assign them to each lock so that lockdep can figure
> > > > out an AA deadlock isn't possible here.
> > > >
> > > > Fixes: b3d91800d9ac ("drm/msm: Fix race condition in msm driver with async layer updates")
> > > > Cc: Krishna Manikandan <mkrishn@codeaurora.org>
> > > > Signed-off-by: Stephen Boyd <swboyd@chromium.org>
> > >
> > > This smells like throwing more bad after initial bad code ...
> > >
> > > First a rant: https://blog.ffwll.ch/2020/08/lockdep-false-positives.html
>
> Some technical on the patch itself: I think you want
> mutex_lock_nested(crtc->lock, drm_crtc_index(crtc)), not your own locking
> classes hand-rolled. It's defacto the same, but much more obviously
> correct since self-documenting.

hmm, yeah, that is a bit cleaner.. but this patch is already on
msm-next, maybe I'll add a patch on top to change it

> > > Yes I know the locking you're doing here is correct, but that goes to the
> > > second issue: Why is this needed? atomic_async_update helpers are supposed
> > > to take care of ordering fun like this, if they're not, we need to address
> > > things there. The problem that
> >
> > Maybe a better solution would be helper awareness of hw that has
> > double-buffered state and flush bits.. ie. something that looks a bit
> > more like the internal kms fxn ptrs. Currently the locking is
> > protecting something that the atomic helpers are not aware of, ie.
> > we've already written previous cursor updates to hw and are just
> > waiting until close to vblank to write the flush bits
> >
> > But, we've been over this before. I'd tried various approaches.. the
> > current scheme replaces seanpaul's earlier attempts to do it the
> > "helper" way.  The current implementation does the best job of
> > avoiding fps drops when the legacy cursor uapi is in play.  (And yes,
> > legacy cursor + atomic ioctls is maybe not the greatest, but it is
> > what it is.)
>
> I didn't read enough of the context and got confused, the flush handling
> looks all reasonable and obviously needs some locks to avoid races with
> updates.
>
> It still looks a bit strange that you need multi-crtc locks for cursor
> (generally this stuff is supposed to be solved with ordering) and why the
> async helpers don't work since msm has something that's pretty close
> itself. Atomic+cursor is a bit nasty, but if every driver hacks this
> together themselves then there's not much chance of this ever really
> working well across the board. And aside from the flush bit instead of
> automatic double buffering (which you're just emulating) there's not
> really anything special with msm afaics. So pretty sure that if this
> doesn't work for msm, it doesn't work anywhere else.

I did dig out a hub/keyboard/mouse for the one mtk device I have to
see how it behaves with cursor updates.. the fps drops are pretty
bad.. roughly the same as they were with msm prior to the current
"defer the flush bits" scheme.  Maybe I'm just more picky than others.

Better helper support for this sort of hw would be useful, since I
think it is not uncommon.  (At least omap is similar, I didn't really
look at mtk/rockchip.)  If I had a clone to handle the display side of
things, my clone would work on that ;-)

BR,
-R

> -Daniel
>
> >
> > BR,
> > -R
> >
> > >
> > > commit b3d91800d9ac35014e0349292273a6fa7938d402
> > > Author: Krishna Manikandan <mkrishn@codeaurora.org>
> > > Date:   Fri Oct 16 19:40:43 2020 +0530
> > >
> > >     drm/msm: Fix race condition in msm driver with async layer updates
> > >
> > > is _the_ reason we have drm_crtc_commit to track stuff, and Maxime has
> > > recently rolled out a pile of changes to vc4 to use these things
> > > correctly. Hacking some glorious hand-rolled locking for synchronization
> > > of updates really should be the exception for kms drivers, not the rule.
> > > And this one here doesn't look like an exception by far (the one legit I
> > > know of is the locking issues amdgpu has between atomic_commit_tail and
> > > gpu reset, and that one is really nasty, so not going to get fixed in
> > > helpers, ever).
> > >
> > > Cheers, Daniel
> > >
> > > > ---
> > > >  drivers/gpu/drm/msm/msm_kms.h | 8 ++++++--
> > > >  1 file changed, 6 insertions(+), 2 deletions(-)
> > > >
> > > > diff --git a/drivers/gpu/drm/msm/msm_kms.h b/drivers/gpu/drm/msm/msm_kms.h
> > > > index d8151a89e163..4735251a394d 100644
> > > > --- a/drivers/gpu/drm/msm/msm_kms.h
> > > > +++ b/drivers/gpu/drm/msm/msm_kms.h
> > > > @@ -157,6 +157,7 @@ struct msm_kms {
> > > >        * from the crtc's pending_timer close to end of the frame:
> > > >        */
> > > >       struct mutex commit_lock[MAX_CRTCS];
> > > > +     struct lock_class_key commit_lock_keys[MAX_CRTCS];
> > > >       unsigned pending_crtc_mask;
> > > >       struct msm_pending_timer pending_timers[MAX_CRTCS];
> > > >  };
> > > > @@ -166,8 +167,11 @@ static inline int msm_kms_init(struct msm_kms *kms,
> > > >  {
> > > >       unsigned i, ret;
> > > >
> > > > -     for (i = 0; i < ARRAY_SIZE(kms->commit_lock); i++)
> > > > -             mutex_init(&kms->commit_lock[i]);
> > > > +     for (i = 0; i < ARRAY_SIZE(kms->commit_lock); i++) {
> > > > +             lockdep_register_key(&kms->commit_lock_keys[i]);
> > > > +             __mutex_init(&kms->commit_lock[i], "&kms->commit_lock[i]",
> > > > +                          &kms->commit_lock_keys[i]);
> > > > +     }
> > > >
> > > >       kms->funcs = funcs;
> > > >
> > > >
> > > > base-commit: 19c329f6808995b142b3966301f217c831e7cf31
> > > > --
> > > > https://chromeos.dev
> > > >
> > > > _______________________________________________
> > > > dri-devel mailing list
> > > > dri-devel@lists.freedesktop.org
> > > > https://lists.freedesktop.org/mailman/listinfo/dri-devel
> > >
> > > --
> > > Daniel Vetter
> > > Software Engineer, Intel Corporation
> > > http://blog.ffwll.ch
> > _______________________________________________
> > dri-devel mailing list
> > dri-devel@lists.freedesktop.org
> > https://lists.freedesktop.org/mailman/listinfo/dri-devel
>
> --
> Daniel Vetter
> Software Engineer, Intel Corporation
> http://blog.ffwll.ch
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

^ permalink raw reply	[flat|nested] 10+ messages in thread

* Re: [PATCH] drm/msm/kms: Make a lock_class_key for each crtc mutex
  2021-02-03 17:29       ` Rob Clark
@ 2021-02-03 21:58         ` Stephen Boyd
  2021-02-03 22:11           ` Rob Clark
  0 siblings, 1 reply; 10+ messages in thread
From: Stephen Boyd @ 2021-02-03 21:58 UTC (permalink / raw)
  To: Krishna Manikandan, Linux Kernel Mailing List, Rob Clark,
	dri-devel, freedreno, linux-arm-msm

Quoting Rob Clark (2021-02-03 09:29:09)
> On Wed, Feb 3, 2021 at 2:10 AM Daniel Vetter <daniel@ffwll.ch> wrote:
> >
> > On Tue, Feb 02, 2021 at 08:51:25AM -0800, Rob Clark wrote:
> > > On Tue, Feb 2, 2021 at 7:46 AM Daniel Vetter <daniel@ffwll.ch> wrote:
> > > >
> > > > On Mon, Jan 25, 2021 at 03:49:01PM -0800, Stephen Boyd wrote:
> > > > > This is because lockdep thinks all the locks taken in lock_crtcs() are
> > > > > the same lock, when they actually aren't. That's because we call
> > > > > mutex_init() in msm_kms_init() and that assigns on static key for every
> > > > > lock initialized in this loop. Let's allocate a dynamic number of
> > > > > lock_class_keys and assign them to each lock so that lockdep can figure
> > > > > out an AA deadlock isn't possible here.
> > > > >
> > > > > Fixes: b3d91800d9ac ("drm/msm: Fix race condition in msm driver with async layer updates")
> > > > > Cc: Krishna Manikandan <mkrishn@codeaurora.org>
> > > > > Signed-off-by: Stephen Boyd <swboyd@chromium.org>
> > > >
> > > > This smells like throwing more bad after initial bad code ...
> > > >
> > > > First a rant: https://blog.ffwll.ch/2020/08/lockdep-false-positives.html
> >
> > Some technical on the patch itself: I think you want
> > mutex_lock_nested(crtc->lock, drm_crtc_index(crtc)), not your own locking
> > classes hand-rolled. It's defacto the same, but much more obviously
> > correct since self-documenting.
> 
> hmm, yeah, that is a bit cleaner.. but this patch is already on
> msm-next, maybe I'll add a patch on top to change it

How many CRTCs are there? The subclass number tops out at 8, per
MAX_LOCKDEP_SUBCLASSES so if we have more than that many bits possible
then it will fail.
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

^ permalink raw reply	[flat|nested] 10+ messages in thread

* Re: [PATCH] drm/msm/kms: Make a lock_class_key for each crtc mutex
  2021-02-03 21:58         ` Stephen Boyd
@ 2021-02-03 22:11           ` Rob Clark
  2021-02-04 15:17             ` Daniel Vetter
  0 siblings, 1 reply; 10+ messages in thread
From: Rob Clark @ 2021-02-03 22:11 UTC (permalink / raw)
  To: Stephen Boyd
  Cc: Krishna Manikandan, freedreno, Linux Kernel Mailing List,
	dri-devel, linux-arm-msm

On Wed, Feb 3, 2021 at 1:58 PM Stephen Boyd <swboyd@chromium.org> wrote:
>
> Quoting Rob Clark (2021-02-03 09:29:09)
> > On Wed, Feb 3, 2021 at 2:10 AM Daniel Vetter <daniel@ffwll.ch> wrote:
> > >
> > > On Tue, Feb 02, 2021 at 08:51:25AM -0800, Rob Clark wrote:
> > > > On Tue, Feb 2, 2021 at 7:46 AM Daniel Vetter <daniel@ffwll.ch> wrote:
> > > > >
> > > > > On Mon, Jan 25, 2021 at 03:49:01PM -0800, Stephen Boyd wrote:
> > > > > > This is because lockdep thinks all the locks taken in lock_crtcs() are
> > > > > > the same lock, when they actually aren't. That's because we call
> > > > > > mutex_init() in msm_kms_init() and that assigns on static key for every
> > > > > > lock initialized in this loop. Let's allocate a dynamic number of
> > > > > > lock_class_keys and assign them to each lock so that lockdep can figure
> > > > > > out an AA deadlock isn't possible here.
> > > > > >
> > > > > > Fixes: b3d91800d9ac ("drm/msm: Fix race condition in msm driver with async layer updates")
> > > > > > Cc: Krishna Manikandan <mkrishn@codeaurora.org>
> > > > > > Signed-off-by: Stephen Boyd <swboyd@chromium.org>
> > > > >
> > > > > This smells like throwing more bad after initial bad code ...
> > > > >
> > > > > First a rant: https://blog.ffwll.ch/2020/08/lockdep-false-positives.html
> > >
> > > Some technical on the patch itself: I think you want
> > > mutex_lock_nested(crtc->lock, drm_crtc_index(crtc)), not your own locking
> > > classes hand-rolled. It's defacto the same, but much more obviously
> > > correct since self-documenting.
> >
> > hmm, yeah, that is a bit cleaner.. but this patch is already on
> > msm-next, maybe I'll add a patch on top to change it
>
> How many CRTCs are there? The subclass number tops out at 8, per
> MAX_LOCKDEP_SUBCLASSES so if we have more than that many bits possible
> then it will fail.

conveniently MAX_CRTCS is 8.. realistically I don't *think* you'd ever
see more than 2 or 3

BR,
-R
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

^ permalink raw reply	[flat|nested] 10+ messages in thread

* Re: [PATCH] drm/msm/kms: Make a lock_class_key for each crtc mutex
  2021-02-03 22:11           ` Rob Clark
@ 2021-02-04 15:17             ` Daniel Vetter
  0 siblings, 0 replies; 10+ messages in thread
From: Daniel Vetter @ 2021-02-04 15:17 UTC (permalink / raw)
  To: Rob Clark
  Cc: Krishna Manikandan, linux-arm-msm, Linux Kernel Mailing List,
	dri-devel, Stephen Boyd, freedreno

On Wed, Feb 03, 2021 at 02:11:09PM -0800, Rob Clark wrote:
> On Wed, Feb 3, 2021 at 1:58 PM Stephen Boyd <swboyd@chromium.org> wrote:
> >
> > Quoting Rob Clark (2021-02-03 09:29:09)
> > > On Wed, Feb 3, 2021 at 2:10 AM Daniel Vetter <daniel@ffwll.ch> wrote:
> > > >
> > > > On Tue, Feb 02, 2021 at 08:51:25AM -0800, Rob Clark wrote:
> > > > > On Tue, Feb 2, 2021 at 7:46 AM Daniel Vetter <daniel@ffwll.ch> wrote:
> > > > > >
> > > > > > On Mon, Jan 25, 2021 at 03:49:01PM -0800, Stephen Boyd wrote:
> > > > > > > This is because lockdep thinks all the locks taken in lock_crtcs() are
> > > > > > > the same lock, when they actually aren't. That's because we call
> > > > > > > mutex_init() in msm_kms_init() and that assigns on static key for every
> > > > > > > lock initialized in this loop. Let's allocate a dynamic number of
> > > > > > > lock_class_keys and assign them to each lock so that lockdep can figure
> > > > > > > out an AA deadlock isn't possible here.
> > > > > > >
> > > > > > > Fixes: b3d91800d9ac ("drm/msm: Fix race condition in msm driver with async layer updates")
> > > > > > > Cc: Krishna Manikandan <mkrishn@codeaurora.org>
> > > > > > > Signed-off-by: Stephen Boyd <swboyd@chromium.org>
> > > > > >
> > > > > > This smells like throwing more bad after initial bad code ...
> > > > > >
> > > > > > First a rant: https://blog.ffwll.ch/2020/08/lockdep-false-positives.html
> > > >
> > > > Some technical on the patch itself: I think you want
> > > > mutex_lock_nested(crtc->lock, drm_crtc_index(crtc)), not your own locking
> > > > classes hand-rolled. It's defacto the same, but much more obviously
> > > > correct since self-documenting.
> > >
> > > hmm, yeah, that is a bit cleaner.. but this patch is already on
> > > msm-next, maybe I'll add a patch on top to change it
> >
> > How many CRTCs are there? The subclass number tops out at 8, per
> > MAX_LOCKDEP_SUBCLASSES so if we have more than that many bits possible
> > then it will fail.

Hm good point, tbh the mutex_lock_nested annotations isn't super awesome
either, it would be kinda neat if we could put that annotation into
mutex_lock_init fairly statically (and at that point we could allos resize
the array fairly easily I think at runtime).

The nice thing with the nesting index is just that it makes it a bit more
obvious that there's a static nesting going on and why it's ok.
-Daniel

> conveniently MAX_CRTCS is 8.. realistically I don't *think* you'd ever
> see more than 2 or 3
> 
> BR,
> -R
> _______________________________________________
> dri-devel mailing list
> dri-devel@lists.freedesktop.org
> https://lists.freedesktop.org/mailman/listinfo/dri-devel

-- 
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

^ permalink raw reply	[flat|nested] 10+ messages in thread

* [PATCH] drm/msm/kms: Make a lock_class_key for each crtc mutex
@ 2021-01-26  2:01 Stephen Boyd
  0 siblings, 0 replies; 10+ messages in thread
From: Stephen Boyd @ 2021-01-26  2:01 UTC (permalink / raw)
  To: Rob Clark
  Cc: linux-arm-msm, freedreno, linux-kernel, dri-devel, Krishna Manikandan

Lockdep complains about an AA deadlock when rebooting the device.

============================================
WARNING: possible recursive locking detected
5.4.91 #1 Not tainted
--------------------------------------------
reboot/5213 is trying to acquire lock:
ffffff80d13391b0 (&kms->commit_lock[i]){+.+.}, at: lock_crtcs+0x60/0xa4

but task is already holding lock:
ffffff80d1339110 (&kms->commit_lock[i]){+.+.}, at: lock_crtcs+0x60/0xa4

other info that might help us debug this:
Possible unsafe locking scenario:

CPU0
----
lock(&kms->commit_lock[i]);
lock(&kms->commit_lock[i]);

*** DEADLOCK ***

May be due to missing lock nesting notation

6 locks held by reboot/5213:
__arm64_sys_reboot+0x148/0x2a0
device_shutdown+0x10c/0x2c4
drm_atomic_helper_shutdown+0x48/0xfc
modeset_lock+0x120/0x24c
lock_crtcs+0x60/0xa4

stack backtrace:
CPU: 4 PID: 5213 Comm: reboot Not tainted 5.4.91 #1
Hardware name: Google Pompom (rev1) with LTE (DT)
Call trace:
dump_backtrace+0x0/0x1dc
show_stack+0x24/0x30
dump_stack+0xfc/0x1a8
__lock_acquire+0xcd0/0x22b8
lock_acquire+0x1ec/0x240
__mutex_lock_common+0xe0/0xc84
mutex_lock_nested+0x48/0x58
lock_crtcs+0x60/0xa4
msm_atomic_commit_tail+0x348/0x570
commit_tail+0xdc/0x178
drm_atomic_helper_commit+0x160/0x168
drm_atomic_commit+0x68/0x80

This is because lockdep thinks all the locks taken in lock_crtcs() are
the same lock, when they actually aren't. That's because we call
mutex_init() in msm_kms_init() and that assigns on static key for every
lock initialized in this loop. Let's allocate a dynamic number of
lock_class_keys and assign them to each lock so that lockdep can figure
out an AA deadlock isn't possible here.

Fixes: b3d91800d9ac ("drm/msm: Fix race condition in msm driver with async layer updates")
Cc: Krishna Manikandan <mkrishn@codeaurora.org>
Signed-off-by: Stephen Boyd <swboyd@chromium.org>
---
 drivers/gpu/drm/msm/msm_kms.h | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/msm/msm_kms.h b/drivers/gpu/drm/msm/msm_kms.h
index d8151a89e163..4735251a394d 100644
--- a/drivers/gpu/drm/msm/msm_kms.h
+++ b/drivers/gpu/drm/msm/msm_kms.h
@@ -157,6 +157,7 @@ struct msm_kms {
 	 * from the crtc's pending_timer close to end of the frame:
 	 */
 	struct mutex commit_lock[MAX_CRTCS];
+	struct lock_class_key commit_lock_keys[MAX_CRTCS];
 	unsigned pending_crtc_mask;
 	struct msm_pending_timer pending_timers[MAX_CRTCS];
 };
@@ -166,8 +167,11 @@ static inline int msm_kms_init(struct msm_kms *kms,
 {
 	unsigned i, ret;
 
-	for (i = 0; i < ARRAY_SIZE(kms->commit_lock); i++)
-		mutex_init(&kms->commit_lock[i]);
+	for (i = 0; i < ARRAY_SIZE(kms->commit_lock); i++) {
+		lockdep_register_key(&kms->commit_lock_keys[i]);
+		__mutex_init(&kms->commit_lock[i], "&kms->commit_lock[i]",
+			     &kms->commit_lock_keys[i]);
+	}
 
 	kms->funcs = funcs;
 

base-commit: 19c329f6808995b142b3966301f217c831e7cf31
-- 
https://chromeos.dev

_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

^ permalink raw reply related	[flat|nested] 10+ messages in thread

end of thread, other threads:[~2021-02-04 15:18 UTC | newest]

Thread overview: 10+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2021-01-25 23:49 [PATCH] drm/msm/kms: Make a lock_class_key for each crtc mutex Stephen Boyd
2021-01-28 16:39 ` Rob Clark
2021-02-02 15:46 ` Daniel Vetter
2021-02-02 16:51   ` Rob Clark
2021-02-03 10:10     ` Daniel Vetter
2021-02-03 17:29       ` Rob Clark
2021-02-03 21:58         ` Stephen Boyd
2021-02-03 22:11           ` Rob Clark
2021-02-04 15:17             ` Daniel Vetter
2021-01-26  2:01 Stephen Boyd

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).