[PATCH 4.1,4.2] drm/i915: Silence DDR DVFS errors on CHV

[PATCH 4.1,4.2] drm/i915: Silence DDR DVFS errors on CHV

[ville.syrjala]

From: Ville Syrjälä <ville.syrjala@linux.intel.com>

commit 58590c14d80defc94e900308a9d8fa55284de6f2 upstream.

58590c1 drm/i915: Don't try to use DDR DVFS on CHV when disabled in the BIOS
can't be backported as is because proper DDR DVFS support didn't even
exist before
6f9c784 drm/i915: Don't do PM5/DDR DVFS with multiple pipes

However we still want to get rid of the error messages. The simplest thing
is to just suppress them, so let's do that instead.

Cc: <stable@vger.kernel.org> # 4.1, 4.2
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
---
 drivers/gpu/drm/i915/intel_pm.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c
index eadc15c..4a904f2 100644
--- a/drivers/gpu/drm/i915/intel_pm.c
+++ b/drivers/gpu/drm/i915/intel_pm.c
@@ -303,7 +303,7 @@ static void chv_set_memory_dvfs(struct drm_i915_private *dev_priv, bool enable)
 
 	if (wait_for((vlv_punit_read(dev_priv, PUNIT_REG_DDR_SETUP2) &
 		      FORCE_DDR_FREQ_REQ_ACK) == 0, 3))
-		DRM_ERROR("timed out waiting for Punit DDR DVFS request\n");
+		DRM_DEBUG_DRIVER("timed out waiting for Punit DDR DVFS request\n");
 
 	mutex_unlock(&dev_priv->rps.hw_lock);
 }
-- 
2.4.6

Re: [PATCH 4.1,4.2] drm/i915: Silence DDR DVFS errors on CHV

[Greg KH]

On Mon, Sep 28, 2015 at 10:09:11PM +0300, ville.syrjala@linux.intel.com wrote:
> From: Ville Syrj�l� <ville.syrjala@linux.intel.com>
> 
> commit 58590c14d80defc94e900308a9d8fa55284de6f2 upstream.

This is not the commit id of the patch below at all, I can't take this,
please be more careful in the future.

thanks,

greg k-h

> 
> 58590c1 drm/i915: Don't try to use DDR DVFS on CHV when disabled in the BIOS
> can't be backported as is because proper DDR DVFS support didn't even
> exist before
> 6f9c784 drm/i915: Don't do PM5/DDR DVFS with multiple pipes
> 
> However we still want to get rid of the error messages. The simplest thing
> is to just suppress them, so let's do that instead.
> 
> Cc: <stable@vger.kernel.org> # 4.1, 4.2
> Signed-off-by: Ville Syrj�l� <ville.syrjala@linux.intel.com>
> ---
>  drivers/gpu/drm/i915/intel_pm.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c
> index eadc15c..4a904f2 100644
> --- a/drivers/gpu/drm/i915/intel_pm.c
> +++ b/drivers/gpu/drm/i915/intel_pm.c
> @@ -303,7 +303,7 @@ static void chv_set_memory_dvfs(struct drm_i915_private *dev_priv, bool enable)
>  
>  	if (wait_for((vlv_punit_read(dev_priv, PUNIT_REG_DDR_SETUP2) &
>  		      FORCE_DDR_FREQ_REQ_ACK) == 0, 3))
> -		DRM_ERROR("timed out waiting for Punit DDR DVFS request\n");
> +		DRM_DEBUG_DRIVER("timed out waiting for Punit DDR DVFS request\n");
>  
>  	mutex_unlock(&dev_priv->rps.hw_lock);
>  }
> -- 
> 2.4.6
> 
> --
> To unsubscribe from this list: send the line "unsubscribe stable" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [Intel-gfx] [PATCH 4.1, 4.2] drm/i915: Silence DDR DVFS errors on CHV

[Jani Nikula]

On Sat, 17 Oct 2015, Greg KH <gregkh@linuxfoundation.org> wrote:
> On Mon, Sep 28, 2015 at 10:09:11PM +0300, ville.syrjala@linux.intel.com wrote:
>> From: Ville Syrjälä <ville.syrjala@linux.intel.com>
>> 
>> commit 58590c14d80defc94e900308a9d8fa55284de6f2 upstream.
>
> This is not the commit id of the patch below at all, I can't take this,
> please be more careful in the future.

Greg, the commit message tries (and apparently fails) to explain that we
can't really backport all of the commits to fix this properly.

The referenced upstream commit looks totally different because it
prevents us from entering the failing path to begin with. Since we can't
do that in stable, Ville was proposing to just the tune down the error
message, referencing the commit that gets rid of the error message
upstream.

Was this just a communication failure, or do you think this is still
against the stable kernel rules?


BR,
Jani.

>
> thanks,
>
> greg k-h
>
>> 
>> 58590c1 drm/i915: Don't try to use DDR DVFS on CHV when disabled in the BIOS
>> can't be backported as is because proper DDR DVFS support didn't even
>> exist before
>> 6f9c784 drm/i915: Don't do PM5/DDR DVFS with multiple pipes
>> 
>> However we still want to get rid of the error messages. The simplest thing
>> is to just suppress them, so let's do that instead.
>> 
>> Cc: <stable@vger.kernel.org> # 4.1, 4.2
>> Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
>> ---
>>  drivers/gpu/drm/i915/intel_pm.c | 2 +-
>>  1 file changed, 1 insertion(+), 1 deletion(-)
>> 
>> diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c
>> index eadc15c..4a904f2 100644
>> --- a/drivers/gpu/drm/i915/intel_pm.c
>> +++ b/drivers/gpu/drm/i915/intel_pm.c
>> @@ -303,7 +303,7 @@ static void chv_set_memory_dvfs(struct drm_i915_private *dev_priv, bool enable)
>>  
>>  	if (wait_for((vlv_punit_read(dev_priv, PUNIT_REG_DDR_SETUP2) &
>>  		      FORCE_DDR_FREQ_REQ_ACK) == 0, 3))
>> -		DRM_ERROR("timed out waiting for Punit DDR DVFS request\n");
>> +		DRM_DEBUG_DRIVER("timed out waiting for Punit DDR DVFS request\n");
>>  
>>  	mutex_unlock(&dev_priv->rps.hw_lock);
>>  }
>> -- 
>> 2.4.6
>> 
>> --
>> To unsubscribe from this list: send the line "unsubscribe stable" in
>> the body of a message to majordomo@vger.kernel.org
>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> _______________________________________________
> Intel-gfx mailing list
> Intel-gfx@lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/intel-gfx

-- 
Jani Nikula, Intel Open Source Technology Center

Re: [Intel-gfx] [PATCH 4.1, 4.2] drm/i915: Silence DDR DVFS errors on CHV

[Greg KH]

On Mon, Oct 19, 2015 at 11:02:35AM +0300, Jani Nikula wrote:
> On Sat, 17 Oct 2015, Greg KH <gregkh@linuxfoundation.org> wrote:
> > On Mon, Sep 28, 2015 at 10:09:11PM +0300, ville.syrjala@linux.intel.com wrote:
> >> From: Ville Syrj�l� <ville.syrjala@linux.intel.com>
> >> 
> >> commit 58590c14d80defc94e900308a9d8fa55284de6f2 upstream.
> >
> > This is not the commit id of the patch below at all, I can't take this,
> > please be more careful in the future.
> 
> Greg, the commit message tries (and apparently fails) to explain that we
> can't really backport all of the commits to fix this properly.

Yeah, it failed at that, as this isn't the same patch, so please don't
say that in the first line :(

> The referenced upstream commit looks totally different because it
> prevents us from entering the failing path to begin with. Since we can't
> do that in stable, Ville was proposing to just the tune down the error
> message, referencing the commit that gets rid of the error message
> upstream.

Why can't we do that in the stable tree?  I _REALLY_ do not like taking
patches that are different from what is in Linus's tree.  It always
burns us in the end, no matter how hard we try to prevent it...

thanks,

greg k-h

Re: [Intel-gfx] [PATCH 4.1, 4.2] drm/i915: Silence DDR DVFS errors on CHV

[Daniel Vetter]

On Mon, Oct 19, 2015 at 08:13:05AM -0700, Greg KH wrote:
> On Mon, Oct 19, 2015 at 11:02:35AM +0300, Jani Nikula wrote:
> > On Sat, 17 Oct 2015, Greg KH <gregkh@linuxfoundation.org> wrote:
> > > On Mon, Sep 28, 2015 at 10:09:11PM +0300, ville.syrjala@linux.intel.com wrote:
> > >> From: Ville Syrj�l� <ville.syrjala@linux.intel.com>
> > >> 
> > >> commit 58590c14d80defc94e900308a9d8fa55284de6f2 upstream.
> > >
> > > This is not the commit id of the patch below at all, I can't take this,
> > > please be more careful in the future.
> > 
> > Greg, the commit message tries (and apparently fails) to explain that we
> > can't really backport all of the commits to fix this properly.
> 
> Yeah, it failed at that, as this isn't the same patch, so please don't
> say that in the first line :(
> 
> > The referenced upstream commit looks totally different because it
> > prevents us from entering the failing path to begin with. Since we can't
> > do that in stable, Ville was proposing to just the tune down the error
> > message, referencing the commit that gets rid of the error message
> > upstream.
> 
> Why can't we do that in the stable tree?  I _REALLY_ do not like taking
> patches that are different from what is in Linus's tree.  It always
> burns us in the end, no matter how hard we try to prevent it...

It shuts up a spurious error in dmesg about a feature which isn't even
implemented and took about 20+ patches to get right in Linus tree. Option
B is to waste reporters time if we leave that DRM_ERROR in there with
filing bugs that we'll close right away. There's no way we can backport
DVFS support without breaking the world.
-Daniel
-- 
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch

Re: [Intel-gfx] [PATCH 4.1, 4.2] drm/i915: Silence DDR DVFS errors on CHV

[Greg KH]

On Mon, Oct 19, 2015 at 06:10:39PM +0200, Daniel Vetter wrote:
> On Mon, Oct 19, 2015 at 08:13:05AM -0700, Greg KH wrote:
> > On Mon, Oct 19, 2015 at 11:02:35AM +0300, Jani Nikula wrote:
> > > On Sat, 17 Oct 2015, Greg KH <gregkh@linuxfoundation.org> wrote:
> > > > On Mon, Sep 28, 2015 at 10:09:11PM +0300, ville.syrjala@linux.intel.com wrote:
> > > >> From: Ville Syrj�l� <ville.syrjala@linux.intel.com>
> > > >> 
> > > >> commit 58590c14d80defc94e900308a9d8fa55284de6f2 upstream.
> > > >
> > > > This is not the commit id of the patch below at all, I can't take this,
> > > > please be more careful in the future.
> > > 
> > > Greg, the commit message tries (and apparently fails) to explain that we
> > > can't really backport all of the commits to fix this properly.
> > 
> > Yeah, it failed at that, as this isn't the same patch, so please don't
> > say that in the first line :(
> > 
> > > The referenced upstream commit looks totally different because it
> > > prevents us from entering the failing path to begin with. Since we can't
> > > do that in stable, Ville was proposing to just the tune down the error
> > > message, referencing the commit that gets rid of the error message
> > > upstream.
> > 
> > Why can't we do that in the stable tree?  I _REALLY_ do not like taking
> > patches that are different from what is in Linus's tree.  It always
> > burns us in the end, no matter how hard we try to prevent it...
> 
> It shuts up a spurious error in dmesg about a feature which isn't even
> implemented and took about 20+ patches to get right in Linus tree. Option
> B is to waste reporters time if we leave that DRM_ERROR in there with
> filing bugs that we'll close right away. There's no way we can backport
> DVFS support without breaking the world.

Ok, then document the heck out of this, as it is, the changelog text
didn't match what was going on at all.

greg k-h

Re: [Intel-gfx] [PATCH 4.1, 4.2] drm/i915: Silence DDR DVFS errors on CHV

[Ville Syrjälä]

On Mon, Oct 19, 2015 at 08:13:05AM -0700, Greg KH wrote:
> On Mon, Oct 19, 2015 at 11:02:35AM +0300, Jani Nikula wrote:
> > On Sat, 17 Oct 2015, Greg KH <gregkh@linuxfoundation.org> wrote:
> > > On Mon, Sep 28, 2015 at 10:09:11PM +0300, ville.syrjala@linux.intel.com wrote:
> > >> From: Ville Syrj�l� <ville.syrjala@linux.intel.com>
> > >> 
> > >> commit 58590c14d80defc94e900308a9d8fa55284de6f2 upstream.
> > >
> > > This is not the commit id of the patch below at all, I can't take this,
> > > please be more careful in the future.
> > 
> > Greg, the commit message tries (and apparently fails) to explain that we
> > can't really backport all of the commits to fix this properly.
> 
> Yeah, it failed at that, as this isn't the same patch, so please don't
> say that in the first line :(

I remember I had trouble figuring out what exactly I should put in the
commit message. The documentation said I should specify the upsteam
commit, and so that's what I did in the end. There wasn't really much
extra guidance for cases where you can't simply cherry-pick the upstream
commit as is.

> > The referenced upstream commit looks totally different because it
> > prevents us from entering the failing path to begin with. Since we can't
> > do that in stable, Ville was proposing to just the tune down the error
> > message, referencing the commit that gets rid of the error message
> > upstream.
> 
> Why can't we do that in the stable tree?

First we'd probably get to backport most of whatever atomic modeset
work that landed in the meantime (and somehow massage that into a shape
that doesn't break everything), then we'd get to backport at least one
total rewrite of the VLV/CHV watermark code, and finally we might be
able to cherry-pick the patch as is.

So that's a non-trivial amount of work, and the risk of breaking everything
modeset related is very real. Definitely not something I want to put
into stable.

> I _REALLY_ do not like taking
> patches that are different from what is in Linus's tree.  It always
> burns us in the end, no matter how hard we try to prevent it...
> 
> thanks,
> 
> greg k-h

-- 
Ville Syrj�l�
Intel OTC