[PATCH] drm/irq: BUG_ON() -> WARN_ON()

[PATCH] drm/irq: BUG_ON() -> WARN_ON()

[Rob Clark]

Let's make things a bit easier to debug when things go bad (potentially
under console_lock).

Signed-off-by: Rob Clark <robdclark@gmail.com>
---
 drivers/gpu/drm/drm_irq.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/drm_irq.c b/drivers/gpu/drm/drm_irq.c
index 5ef03c2..c4edea9 100644
--- a/drivers/gpu/drm/drm_irq.c
+++ b/drivers/gpu/drm/drm_irq.c
@@ -1029,7 +1029,8 @@ void drm_vblank_put(struct drm_device *dev, int crtc)
 {
 	struct drm_vblank_crtc *vblank = &dev->vblank[crtc];
 
-	BUG_ON(atomic_read(&vblank->refcount) == 0);
+	if (WARN_ON(atomic_read(&vblank->refcount) == 0))
+		return;
 
 	if (WARN_ON(crtc >= dev->num_crtcs))
 		return;
-- 
1.9.3

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

Re: [PATCH] drm/irq: BUG_ON() -> WARN_ON()

[Michel Dänzer]

On 09.11.2014 00:16, Rob Clark wrote:
> Let's make things a bit easier to debug when things go bad (potentially
> under console_lock).
>
> Signed-off-by: Rob Clark <robdclark@gmail.com>
> ---
>   drivers/gpu/drm/drm_irq.c | 3 ++-
>   1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/gpu/drm/drm_irq.c b/drivers/gpu/drm/drm_irq.c
> index 5ef03c2..c4edea9 100644
> --- a/drivers/gpu/drm/drm_irq.c
> +++ b/drivers/gpu/drm/drm_irq.c
> @@ -1029,7 +1029,8 @@ void drm_vblank_put(struct drm_device *dev, int crtc)
>   {
>   	struct drm_vblank_crtc *vblank = &dev->vblank[crtc];
>
> -	BUG_ON(atomic_read(&vblank->refcount) == 0);
> +	if (WARN_ON(atomic_read(&vblank->refcount) == 0))
> +		return;
>
>   	if (WARN_ON(crtc >= dev->num_crtcs))
>   		return;
>

Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>


-- 
Earthling Michel Dänzer            |                  http://www.amd.com
Libre software enthusiast          |                Mesa and X developer
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/dri-devel

Re: [PATCH] drm/irq: BUG_ON() -> WARN_ON()

[Jani Nikula]

On Sat, 08 Nov 2014, Rob Clark <robdclark@gmail.com> wrote:
> Let's make things a bit easier to debug when things go bad (potentially
> under console_lock).
>
> Signed-off-by: Rob Clark <robdclark@gmail.com>
> ---
>  drivers/gpu/drm/drm_irq.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/gpu/drm/drm_irq.c b/drivers/gpu/drm/drm_irq.c
> index 5ef03c2..c4edea9 100644
> --- a/drivers/gpu/drm/drm_irq.c
> +++ b/drivers/gpu/drm/drm_irq.c
> @@ -1029,7 +1029,8 @@ void drm_vblank_put(struct drm_device *dev, int crtc)
>  {
>  	struct drm_vblank_crtc *vblank = &dev->vblank[crtc];
>  
> -	BUG_ON(atomic_read(&vblank->refcount) == 0);
> +	if (WARN_ON(atomic_read(&vblank->refcount) == 0))
> +		return;

While I approve of the change, I'd like to promote the use of WARN with
a message instead of WARN_ON. See [1]. In fact it lead to us redefining
WARN_ON in i915 [2].

Thanks,
Jani.


[1] http://mid.gmane.org/87y4sfkoy4.fsf@intel.com
[2] http://mid.gmane.org/1414510350-15713-1-git-send-email-miku@iki.fi



>  
>  	if (WARN_ON(crtc >= dev->num_crtcs))
>  		return;
> -- 
> 1.9.3
>
> _______________________________________________
> dri-devel mailing list
> dri-devel@lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/dri-devel

-- 
Jani Nikula, Intel Open Source Technology Center
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/dri-devel

Re: [PATCH] drm/irq: BUG_ON() -> WARN_ON()

[Maarten Lankhorst]

Hey,

Op 10-11-14 om 12:01 schreef Jani Nikula:
> On Sat, 08 Nov 2014, Rob Clark <robdclark@gmail.com> wrote:
>> Let's make things a bit easier to debug when things go bad (potentially
>> under console_lock).
>>
>> Signed-off-by: Rob Clark <robdclark@gmail.com>
>> ---
>>  drivers/gpu/drm/drm_irq.c | 3 ++-
>>  1 file changed, 2 insertions(+), 1 deletion(-)
>>
>> diff --git a/drivers/gpu/drm/drm_irq.c b/drivers/gpu/drm/drm_irq.c
>> index 5ef03c2..c4edea9 100644
>> --- a/drivers/gpu/drm/drm_irq.c
>> +++ b/drivers/gpu/drm/drm_irq.c
>> @@ -1029,7 +1029,8 @@ void drm_vblank_put(struct drm_device *dev, int crtc)
>>  {
>>  	struct drm_vblank_crtc *vblank = &dev->vblank[crtc];
>>  
>> -	BUG_ON(atomic_read(&vblank->refcount) == 0);
>> +	if (WARN_ON(atomic_read(&vblank->refcount) == 0))
>> +		return;
> While I approve of the change, I'd like to promote the use of WARN with
> a message instead of WARN_ON. See [1]. In fact it lead to us redefining
> WARN_ON in i915 [2].
I love the redefinition of WARN, but I think this should not be overridden locally, instead it should be a kernel debug option.

~Maarten

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

Re: [PATCH] drm/irq: BUG_ON() -> WARN_ON()

[Maarten Lankhorst]

Hey,

Op 10-11-14 om 12:01 schreef Jani Nikula:
> On Sat, 08 Nov 2014, Rob Clark <robdclark@gmail.com> wrote:
>> Let's make things a bit easier to debug when things go bad (potentially
>> under console_lock).
>>
>> Signed-off-by: Rob Clark <robdclark@gmail.com>
>> ---
>>  drivers/gpu/drm/drm_irq.c | 3 ++-
>>  1 file changed, 2 insertions(+), 1 deletion(-)
>>
>> diff --git a/drivers/gpu/drm/drm_irq.c b/drivers/gpu/drm/drm_irq.c
>> index 5ef03c2..c4edea9 100644
>> --- a/drivers/gpu/drm/drm_irq.c
>> +++ b/drivers/gpu/drm/drm_irq.c
>> @@ -1029,7 +1029,8 @@ void drm_vblank_put(struct drm_device *dev, int crtc)
>>  {
>>  	struct drm_vblank_crtc *vblank = &dev->vblank[crtc];
>>  
>> -	BUG_ON(atomic_read(&vblank->refcount) == 0);
>> +	if (WARN_ON(atomic_read(&vblank->refcount) == 0))
>> +		return;
> While I approve of the change, I'd like to promote the use of WARN with
> a message instead of WARN_ON. See [1]. In fact it lead to us redefining
> WARN_ON in i915 [2].
I love the redefinition of WARN, but I think this should not be overridden locally, instead it should be a kernel debug option.

~Maarten