diff mbox series

[v2] drm/i915: Try to guess PCH type even without ISA bridge

Message ID 20201218090531.23241-1-xiong.y.zhang@intel.com (mailing list archive)
State New, archived
Headers show
Series [v2] drm/i915: Try to guess PCH type even without ISA bridge | expand

Commit Message

Zhang, Xiong Y Dec. 18, 2020, 9:05 a.m. UTC
From: Zhenyu Wang <zhenyuw@linux.intel.com>

Some vmm like hyperv and crosvm don't supply any ISA bridge to their guest,
when igd passthrough is equipped on these vmm, guest i915 display may
couldn't work as guest i915 detects PCH_NONE pch type.

When i915 runs as guest, this patch guess pch type through gpu type even
without ISA bridge.

v2: Fix CI warning

Signed-off-by: Zhenyu Wang <zhenyuw@linux.intel.com>
---
 drivers/gpu/drm/i915/i915_drv.h  |  7 +++++-
 drivers/gpu/drm/i915/intel_pch.c | 38 ++++++++++++++++++++++----------
 2 files changed, 32 insertions(+), 13 deletions(-)

Comments

Zhenyu Wang Dec. 22, 2020, 5:15 a.m. UTC | #1
On 2020.12.18 17:05:31 +0800, Xiong Zhang wrote:
> From: Zhenyu Wang <zhenyuw@linux.intel.com>
> 
> Some vmm like hyperv and crosvm don't supply any ISA bridge to their guest,
> when igd passthrough is equipped on these vmm, guest i915 display may
> couldn't work as guest i915 detects PCH_NONE pch type.
> 
> When i915 runs as guest, this patch guess pch type through gpu type even
> without ISA bridge.
> 
> v2: Fix CI warning
> 
> Signed-off-by: Zhenyu Wang <zhenyuw@linux.intel.com>

Need to add yourself in sob.

> ---
>  drivers/gpu/drm/i915/i915_drv.h  |  7 +++++-
>  drivers/gpu/drm/i915/intel_pch.c | 38 ++++++++++++++++++++++----------
>  2 files changed, 32 insertions(+), 13 deletions(-)
> 
> diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
> index 5a7df5621aa3..df0b8f9268b2 100644
> --- a/drivers/gpu/drm/i915/i915_drv.h
> +++ b/drivers/gpu/drm/i915/i915_drv.h
> @@ -1758,6 +1758,11 @@ tgl_revids_get(struct drm_i915_private *dev_priv)
>  #define INTEL_DISPLAY_ENABLED(dev_priv) \
>  	(drm_WARN_ON(&(dev_priv)->drm, !HAS_DISPLAY(dev_priv)), !(dev_priv)->params.disable_display)
>  
> +static inline bool run_as_guest(void)
> +{
> +	return !hypervisor_is_type(X86_HYPER_NATIVE);
> +}
> +
>  static inline bool intel_vtd_active(void)
>  {
>  #ifdef CONFIG_INTEL_IOMMU
> @@ -1766,7 +1771,7 @@ static inline bool intel_vtd_active(void)
>  #endif
>  
>  	/* Running as a guest, we assume the host is enforcing VT'd */
> -	return !hypervisor_is_type(X86_HYPER_NATIVE);
> +	return run_as_guest();
>  }
>  
>  static inline bool intel_scanout_needs_vtd_wa(struct drm_i915_private *dev_priv)
> diff --git a/drivers/gpu/drm/i915/intel_pch.c b/drivers/gpu/drm/i915/intel_pch.c
> index f31c0dabd0cc..a73c60bf349e 100644
> --- a/drivers/gpu/drm/i915/intel_pch.c
> +++ b/drivers/gpu/drm/i915/intel_pch.c
> @@ -184,6 +184,23 @@ intel_virt_detect_pch(const struct drm_i915_private *dev_priv)
>  	return id;
>  }
>  
> +static void intel_detect_pch_virt(struct drm_i915_private *dev_priv)
> +{
> +	unsigned short id;
> +	enum intel_pch pch_type;
> +
> +	id = intel_virt_detect_pch(dev_priv);
> +	pch_type = intel_pch_type(dev_priv, id);
> +
> +	/* Sanity check virtual PCH id */
> +	if (drm_WARN_ON(&dev_priv->drm,
> +			id && pch_type == PCH_NONE))
> +		id = 0;
> +
> +	dev_priv->pch_type = pch_type;
> +	dev_priv->pch_id = id;
> +}
> +
>  void intel_detect_pch(struct drm_i915_private *dev_priv)
>  {
>  	struct pci_dev *pch = NULL;
> @@ -221,16 +238,7 @@ void intel_detect_pch(struct drm_i915_private *dev_priv)
>  			break;
>  		} else if (intel_is_virt_pch(id, pch->subsystem_vendor,
>  					     pch->subsystem_device)) {
> -			id = intel_virt_detect_pch(dev_priv);
> -			pch_type = intel_pch_type(dev_priv, id);
> -
> -			/* Sanity check virtual PCH id */
> -			if (drm_WARN_ON(&dev_priv->drm,
> -					id && pch_type == PCH_NONE))
> -				id = 0;
> -
> -			dev_priv->pch_type = pch_type;
> -			dev_priv->pch_id = id;
> +			intel_detect_pch_virt(dev_priv);
>  			break;
>  		}
>  	}
> @@ -246,8 +254,14 @@ void intel_detect_pch(struct drm_i915_private *dev_priv)
>  		dev_priv->pch_id = 0;
>  	}
>  
> -	if (!pch)
> -		drm_dbg_kms(&dev_priv->drm, "No PCH found.\n");
> +	if (!pch) {

Require HAS_DISPLAY() here?

> +		if (run_as_guest()) {
> +			drm_dbg_kms(&dev_priv->drm, "No PCH found in vm, try guess..\n");
> +			intel_detect_pch_virt(dev_priv);
> +		} else {
> +			drm_dbg_kms(&dev_priv->drm, "No PCH found.\n");
> +		}
> +	}
>  
>  	pci_dev_put(pch);
>  }
> -- 
> 2.17.1
>
Zhang, Xiong Y Dec. 23, 2020, 8:39 a.m. UTC | #2
> On 2020.12.18 17:05:31 +0800, Xiong Zhang wrote:
> > From: Zhenyu Wang <zhenyuw@linux.intel.com>
> >
> > Some vmm like hyperv and crosvm don't supply any ISA bridge to their
> > guest, when igd passthrough is equipped on these vmm, guest i915
> > display may couldn't work as guest i915 detects PCH_NONE pch type.
> >
> > When i915 runs as guest, this patch guess pch type through gpu type
> > even without ISA bridge.
> >
> > v2: Fix CI warning
> >
> > Signed-off-by: Zhenyu Wang <zhenyuw@linux.intel.com>
> 
> Need to add yourself in sob.
> 
> > ---
> >  drivers/gpu/drm/i915/i915_drv.h  |  7 +++++-
> > drivers/gpu/drm/i915/intel_pch.c | 38 ++++++++++++++++++++++----------
> >  2 files changed, 32 insertions(+), 13 deletions(-)
> >
> > diff --git a/drivers/gpu/drm/i915/i915_drv.h
> > b/drivers/gpu/drm/i915/i915_drv.h index 5a7df5621aa3..df0b8f9268b2
> > 100644
> > --- a/drivers/gpu/drm/i915/i915_drv.h
> > +++ b/drivers/gpu/drm/i915/i915_drv.h
> > @@ -1758,6 +1758,11 @@ tgl_revids_get(struct drm_i915_private
> > *dev_priv)  #define INTEL_DISPLAY_ENABLED(dev_priv) \
> >  	(drm_WARN_ON(&(dev_priv)->drm, !HAS_DISPLAY(dev_priv)),
> > !(dev_priv)->params.disable_display)
> >
> > +static inline bool run_as_guest(void) {
> > +	return !hypervisor_is_type(X86_HYPER_NATIVE);
> > +}
> > +
> >  static inline bool intel_vtd_active(void)  {  #ifdef
> > CONFIG_INTEL_IOMMU @@ -1766,7 +1771,7 @@ static inline bool
> > intel_vtd_active(void)  #endif
> >
> >  	/* Running as a guest, we assume the host is enforcing VT'd */
> > -	return !hypervisor_is_type(X86_HYPER_NATIVE);
> > +	return run_as_guest();
> >  }
> >
> >  static inline bool intel_scanout_needs_vtd_wa(struct drm_i915_private
> > *dev_priv) diff --git a/drivers/gpu/drm/i915/intel_pch.c
> > b/drivers/gpu/drm/i915/intel_pch.c
> > index f31c0dabd0cc..a73c60bf349e 100644
> > --- a/drivers/gpu/drm/i915/intel_pch.c
> > +++ b/drivers/gpu/drm/i915/intel_pch.c
> > @@ -184,6 +184,23 @@ intel_virt_detect_pch(const struct
> drm_i915_private *dev_priv)
> >  	return id;
> >  }
> >
> > +static void intel_detect_pch_virt(struct drm_i915_private *dev_priv)
> > +{
> > +	unsigned short id;
> > +	enum intel_pch pch_type;
> > +
> > +	id = intel_virt_detect_pch(dev_priv);
> > +	pch_type = intel_pch_type(dev_priv, id);
> > +
> > +	/* Sanity check virtual PCH id */
> > +	if (drm_WARN_ON(&dev_priv->drm,
> > +			id && pch_type == PCH_NONE))
> > +		id = 0;
> > +
> > +	dev_priv->pch_type = pch_type;
> > +	dev_priv->pch_id = id;
> > +}
> > +
> >  void intel_detect_pch(struct drm_i915_private *dev_priv)  {
> >  	struct pci_dev *pch = NULL;
> > @@ -221,16 +238,7 @@ void intel_detect_pch(struct drm_i915_private
> *dev_priv)
> >  			break;
> >  		} else if (intel_is_virt_pch(id, pch->subsystem_vendor,
> >  					     pch->subsystem_device)) {
> > -			id = intel_virt_detect_pch(dev_priv);
> > -			pch_type = intel_pch_type(dev_priv, id);
> > -
> > -			/* Sanity check virtual PCH id */
> > -			if (drm_WARN_ON(&dev_priv->drm,
> > -					id && pch_type == PCH_NONE))
> > -				id = 0;
> > -
> > -			dev_priv->pch_type = pch_type;
> > -			dev_priv->pch_id = id;
> > +			intel_detect_pch_virt(dev_priv);
> >  			break;
> >  		}
> >  	}
> > @@ -246,8 +254,14 @@ void intel_detect_pch(struct drm_i915_private
> *dev_priv)
> >  		dev_priv->pch_id = 0;
> >  	}
> >
> > -	if (!pch)
> > -		drm_dbg_kms(&dev_priv->drm, "No PCH found.\n");
> > +	if (!pch) {
> 
> Require HAS_DISPLAY() here?
[Zhang, Xiong Y] yes, require it.

thanks
> 
> > +		if (run_as_guest()) {
> > +			drm_dbg_kms(&dev_priv->drm, "No PCH found in
> vm, try guess..\n");
> > +			intel_detect_pch_virt(dev_priv);
> > +		} else {
> > +			drm_dbg_kms(&dev_priv->drm, "No PCH found.\n");
> > +		}
> > +	}
> >
> >  	pci_dev_put(pch);
> >  }
> > --
> > 2.17.1
> >
> 
> --
> 
> $gpg --keyserver wwwkeys.pgp.net --recv-keys 4D781827
Jani Nikula Dec. 23, 2020, 8:59 a.m. UTC | #3
On Fri, 18 Dec 2020, Xiong Zhang <xiong.y.zhang@intel.com> wrote:
> From: Zhenyu Wang <zhenyuw@linux.intel.com>
>
> Some vmm like hyperv and crosvm don't supply any ISA bridge to their guest,
> when igd passthrough is equipped on these vmm, guest i915 display may
> couldn't work as guest i915 detects PCH_NONE pch type.
>
> When i915 runs as guest, this patch guess pch type through gpu type even
> without ISA bridge.
>
> v2: Fix CI warning
>
> Signed-off-by: Zhenyu Wang <zhenyuw@linux.intel.com>
> ---
>  drivers/gpu/drm/i915/i915_drv.h  |  7 +++++-
>  drivers/gpu/drm/i915/intel_pch.c | 38 ++++++++++++++++++++++----------
>  2 files changed, 32 insertions(+), 13 deletions(-)
>
> diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
> index 5a7df5621aa3..df0b8f9268b2 100644
> --- a/drivers/gpu/drm/i915/i915_drv.h
> +++ b/drivers/gpu/drm/i915/i915_drv.h
> @@ -1758,6 +1758,11 @@ tgl_revids_get(struct drm_i915_private *dev_priv)
>  #define INTEL_DISPLAY_ENABLED(dev_priv) \
>  	(drm_WARN_ON(&(dev_priv)->drm, !HAS_DISPLAY(dev_priv)), !(dev_priv)->params.disable_display)
>  
> +static inline bool run_as_guest(void)
> +{
> +	return !hypervisor_is_type(X86_HYPER_NATIVE);
> +}
> +
>  static inline bool intel_vtd_active(void)
>  {
>  #ifdef CONFIG_INTEL_IOMMU
> @@ -1766,7 +1771,7 @@ static inline bool intel_vtd_active(void)
>  #endif
>  
>  	/* Running as a guest, we assume the host is enforcing VT'd */
> -	return !hypervisor_is_type(X86_HYPER_NATIVE);
> +	return run_as_guest();
>  }
>  
>  static inline bool intel_scanout_needs_vtd_wa(struct drm_i915_private *dev_priv)
> diff --git a/drivers/gpu/drm/i915/intel_pch.c b/drivers/gpu/drm/i915/intel_pch.c
> index f31c0dabd0cc..a73c60bf349e 100644
> --- a/drivers/gpu/drm/i915/intel_pch.c
> +++ b/drivers/gpu/drm/i915/intel_pch.c
> @@ -184,6 +184,23 @@ intel_virt_detect_pch(const struct drm_i915_private *dev_priv)
>  	return id;
>  }
>  
> +static void intel_detect_pch_virt(struct drm_i915_private *dev_priv)
> +{
> +	unsigned short id;
> +	enum intel_pch pch_type;
> +
> +	id = intel_virt_detect_pch(dev_priv);

intel_detect_pch_virt() calls intel_virt_detect_pch()... the naming
should be clarified to make some difference.

> +	pch_type = intel_pch_type(dev_priv, id);
> +
> +	/* Sanity check virtual PCH id */
> +	if (drm_WARN_ON(&dev_priv->drm,
> +			id && pch_type == PCH_NONE))
> +		id = 0;
> +
> +	dev_priv->pch_type = pch_type;
> +	dev_priv->pch_id = id;

Previously the pch type and id assignments were all done in
intel_detect_pch(), so moving this to a separate function in *some* but
not all cases reduces clarity too.

BR,
Jani.

> +}
> +
>  void intel_detect_pch(struct drm_i915_private *dev_priv)
>  {
>  	struct pci_dev *pch = NULL;
> @@ -221,16 +238,7 @@ void intel_detect_pch(struct drm_i915_private *dev_priv)
>  			break;
>  		} else if (intel_is_virt_pch(id, pch->subsystem_vendor,
>  					     pch->subsystem_device)) {
> -			id = intel_virt_detect_pch(dev_priv);
> -			pch_type = intel_pch_type(dev_priv, id);
> -
> -			/* Sanity check virtual PCH id */
> -			if (drm_WARN_ON(&dev_priv->drm,
> -					id && pch_type == PCH_NONE))
> -				id = 0;
> -
> -			dev_priv->pch_type = pch_type;
> -			dev_priv->pch_id = id;
> +			intel_detect_pch_virt(dev_priv);
>  			break;
>  		}
>  	}
> @@ -246,8 +254,14 @@ void intel_detect_pch(struct drm_i915_private *dev_priv)
>  		dev_priv->pch_id = 0;
>  	}
>  
> -	if (!pch)
> -		drm_dbg_kms(&dev_priv->drm, "No PCH found.\n");
> +	if (!pch) {
> +		if (run_as_guest()) {
> +			drm_dbg_kms(&dev_priv->drm, "No PCH found in vm, try guess..\n");
> +			intel_detect_pch_virt(dev_priv);
> +		} else {
> +			drm_dbg_kms(&dev_priv->drm, "No PCH found.\n");
> +		}
> +	}
>  
>  	pci_dev_put(pch);
>  }
Zhang, Xiong Y Dec. 24, 2020, 2:42 a.m. UTC | #4
> On Fri, 18 Dec 2020, Xiong Zhang <xiong.y.zhang@intel.com> wrote:
> > From: Zhenyu Wang <zhenyuw@linux.intel.com>
> >
> > Some vmm like hyperv and crosvm don't supply any ISA bridge to their
> > guest, when igd passthrough is equipped on these vmm, guest i915
> > display may couldn't work as guest i915 detects PCH_NONE pch type.
> >
> > When i915 runs as guest, this patch guess pch type through gpu type
> > even without ISA bridge.
> >
> > v2: Fix CI warning
> >
> > Signed-off-by: Zhenyu Wang <zhenyuw@linux.intel.com>
> > ---
> >  drivers/gpu/drm/i915/i915_drv.h  |  7 +++++-
> > drivers/gpu/drm/i915/intel_pch.c | 38 ++++++++++++++++++++++----------
> >  2 files changed, 32 insertions(+), 13 deletions(-)
> >
> > diff --git a/drivers/gpu/drm/i915/i915_drv.h
> > b/drivers/gpu/drm/i915/i915_drv.h index 5a7df5621aa3..df0b8f9268b2
> > 100644
> > --- a/drivers/gpu/drm/i915/i915_drv.h
> > +++ b/drivers/gpu/drm/i915/i915_drv.h
> > @@ -1758,6 +1758,11 @@ tgl_revids_get(struct drm_i915_private
> > *dev_priv)  #define INTEL_DISPLAY_ENABLED(dev_priv) \
> >  	(drm_WARN_ON(&(dev_priv)->drm, !HAS_DISPLAY(dev_priv)),
> > !(dev_priv)->params.disable_display)
> >
> > +static inline bool run_as_guest(void) {
> > +	return !hypervisor_is_type(X86_HYPER_NATIVE);
> > +}
> > +
> >  static inline bool intel_vtd_active(void)  {  #ifdef
> > CONFIG_INTEL_IOMMU @@ -1766,7 +1771,7 @@ static inline bool
> > intel_vtd_active(void)  #endif
> >
> >  	/* Running as a guest, we assume the host is enforcing VT'd */
> > -	return !hypervisor_is_type(X86_HYPER_NATIVE);
> > +	return run_as_guest();
> >  }
> >
> >  static inline bool intel_scanout_needs_vtd_wa(struct drm_i915_private
> > *dev_priv) diff --git a/drivers/gpu/drm/i915/intel_pch.c
> > b/drivers/gpu/drm/i915/intel_pch.c
> > index f31c0dabd0cc..a73c60bf349e 100644
> > --- a/drivers/gpu/drm/i915/intel_pch.c
> > +++ b/drivers/gpu/drm/i915/intel_pch.c
> > @@ -184,6 +184,23 @@ intel_virt_detect_pch(const struct
> drm_i915_private *dev_priv)
> >  	return id;
> >  }
> >
> > +static void intel_detect_pch_virt(struct drm_i915_private *dev_priv)
> > +{
> > +	unsigned short id;
> > +	enum intel_pch pch_type;
> > +
> > +	id = intel_virt_detect_pch(dev_priv);
> 
> intel_detect_pch_virt() calls intel_virt_detect_pch()... the naming should be
> clarified to make some difference.
[Zhang, Xiong Y] Indeed the naming is confusing. How about deleting this new added function, then move intel_pch_type() calling into original intel_virt
> 
> > +	pch_type = intel_pch_type(dev_priv, id);
> > +
> > +	/* Sanity check virtual PCH id */
> > +	if (drm_WARN_ON(&dev_priv->drm,
> > +			id && pch_type == PCH_NONE))
> > +		id = 0;
> > +
> > +	dev_priv->pch_type = pch_type;
> > +	dev_priv->pch_id = id;
> 
> Previously the pch type and id assignments were all done in
> intel_detect_pch(), so moving this to a separate function in *some* but not
> all cases reduces clarity too.
> 
> BR,
> Jani.
> 
> > +}
> > +
> >  void intel_detect_pch(struct drm_i915_private *dev_priv)  {
> >  	struct pci_dev *pch = NULL;
> > @@ -221,16 +238,7 @@ void intel_detect_pch(struct drm_i915_private
> *dev_priv)
> >  			break;
> >  		} else if (intel_is_virt_pch(id, pch->subsystem_vendor,
> >  					     pch->subsystem_device)) {
> > -			id = intel_virt_detect_pch(dev_priv);
> > -			pch_type = intel_pch_type(dev_priv, id);
> > -
> > -			/* Sanity check virtual PCH id */
> > -			if (drm_WARN_ON(&dev_priv->drm,
> > -					id && pch_type == PCH_NONE))
> > -				id = 0;
> > -
> > -			dev_priv->pch_type = pch_type;
> > -			dev_priv->pch_id = id;
> > +			intel_detect_pch_virt(dev_priv);
> >  			break;
> >  		}
> >  	}
> > @@ -246,8 +254,14 @@ void intel_detect_pch(struct drm_i915_private
> *dev_priv)
> >  		dev_priv->pch_id = 0;
> >  	}
> >
> > -	if (!pch)
> > -		drm_dbg_kms(&dev_priv->drm, "No PCH found.\n");
> > +	if (!pch) {
> > +		if (run_as_guest()) {
> > +			drm_dbg_kms(&dev_priv->drm, "No PCH found in
> vm, try guess..\n");
> > +			intel_detect_pch_virt(dev_priv);
> > +		} else {
> > +			drm_dbg_kms(&dev_priv->drm, "No PCH found.\n");
> > +		}
> > +	}
> >
> >  	pci_dev_put(pch);
> >  }
> 
> --
> Jani Nikula, Intel Open Source Graphics Center
Zhang, Xiong Y Dec. 24, 2020, 2:54 a.m. UTC | #5
> On Fri, 18 Dec 2020, Xiong Zhang <xiong.y.zhang@intel.com> wrote:
> > From: Zhenyu Wang <zhenyuw@linux.intel.com>
> >
> > Some vmm like hyperv and crosvm don't supply any ISA bridge to their
> > guest, when igd passthrough is equipped on these vmm, guest i915
> > display may couldn't work as guest i915 detects PCH_NONE pch type.
> >
> > When i915 runs as guest, this patch guess pch type through gpu type
> > even without ISA bridge.
> >
> > v2: Fix CI warning
> >
> > Signed-off-by: Zhenyu Wang <zhenyuw@linux.intel.com>
> > ---
> >  drivers/gpu/drm/i915/i915_drv.h  |  7 +++++-
> > drivers/gpu/drm/i915/intel_pch.c | 38 ++++++++++++++++++++++----------
> >  2 files changed, 32 insertions(+), 13 deletions(-)
> >
> > diff --git a/drivers/gpu/drm/i915/i915_drv.h
> > b/drivers/gpu/drm/i915/i915_drv.h index 5a7df5621aa3..df0b8f9268b2
> > 100644
> > --- a/drivers/gpu/drm/i915/i915_drv.h
> > +++ b/drivers/gpu/drm/i915/i915_drv.h
> > @@ -1758,6 +1758,11 @@ tgl_revids_get(struct drm_i915_private
> > *dev_priv)  #define INTEL_DISPLAY_ENABLED(dev_priv) \
> >  	(drm_WARN_ON(&(dev_priv)->drm, !HAS_DISPLAY(dev_priv)),
> > !(dev_priv)->params.disable_display)
> >
> > +static inline bool run_as_guest(void) {
> > +	return !hypervisor_is_type(X86_HYPER_NATIVE);
> > +}
> > +
> >  static inline bool intel_vtd_active(void)  {  #ifdef
> > CONFIG_INTEL_IOMMU @@ -1766,7 +1771,7 @@ static inline bool
> > intel_vtd_active(void)  #endif
> >
> >  	/* Running as a guest, we assume the host is enforcing VT'd */
> > -	return !hypervisor_is_type(X86_HYPER_NATIVE);
> > +	return run_as_guest();
> >  }
> >
> >  static inline bool intel_scanout_needs_vtd_wa(struct drm_i915_private
> > *dev_priv) diff --git a/drivers/gpu/drm/i915/intel_pch.c
> > b/drivers/gpu/drm/i915/intel_pch.c
> > index f31c0dabd0cc..a73c60bf349e 100644
> > --- a/drivers/gpu/drm/i915/intel_pch.c
> > +++ b/drivers/gpu/drm/i915/intel_pch.c
> > @@ -184,6 +184,23 @@ intel_virt_detect_pch(const struct
> drm_i915_private *dev_priv)
> >  	return id;
> >  }
> >
> > +static void intel_detect_pch_virt(struct drm_i915_private *dev_priv)
> > +{
> > +	unsigned short id;
> > +	enum intel_pch pch_type;
> > +
> > +	id = intel_virt_detect_pch(dev_priv);
> 
> intel_detect_pch_virt() calls intel_virt_detect_pch()... the naming should be
> clarified to make some difference.
 [Zhang, Xiong Y] Indeed the naming is confusing. How about deleting this new added function, then move intel_pch_type() calling into original intel_virt_detect_pch(), and let intel_virt_detect_pch() return id and pch_type,  finally assign id and pch_type  in intel_detect_pch().

thanks
> 
> > +	pch_type = intel_pch_type(dev_priv, id);
> > +
> > +	/* Sanity check virtual PCH id */
> > +	if (drm_WARN_ON(&dev_priv->drm,
> > +			id && pch_type == PCH_NONE))
> > +		id = 0;
> > +
> > +	dev_priv->pch_type = pch_type;
> > +	dev_priv->pch_id = id;
> 
> Previously the pch type and id assignments were all done in
> intel_detect_pch(), so moving this to a separate function in *some* but not
> all cases reduces clarity too.
> 
> BR,
> Jani.
> 
> > +}
> > +
> >  void intel_detect_pch(struct drm_i915_private *dev_priv)  {
> >  	struct pci_dev *pch = NULL;
> > @@ -221,16 +238,7 @@ void intel_detect_pch(struct drm_i915_private
> *dev_priv)
> >  			break;
> >  		} else if (intel_is_virt_pch(id, pch->subsystem_vendor,
> >  					     pch->subsystem_device)) {
> > -			id = intel_virt_detect_pch(dev_priv);
> > -			pch_type = intel_pch_type(dev_priv, id);
> > -
> > -			/* Sanity check virtual PCH id */
> > -			if (drm_WARN_ON(&dev_priv->drm,
> > -					id && pch_type == PCH_NONE))
> > -				id = 0;
> > -
> > -			dev_priv->pch_type = pch_type;
> > -			dev_priv->pch_id = id;
> > +			intel_detect_pch_virt(dev_priv);
> >  			break;
> >  		}
> >  	}
> > @@ -246,8 +254,14 @@ void intel_detect_pch(struct drm_i915_private
> *dev_priv)
> >  		dev_priv->pch_id = 0;
> >  	}
> >
> > -	if (!pch)
> > -		drm_dbg_kms(&dev_priv->drm, "No PCH found.\n");
> > +	if (!pch) {
> > +		if (run_as_guest()) {
> > +			drm_dbg_kms(&dev_priv->drm, "No PCH found in
> vm, try guess..\n");
> > +			intel_detect_pch_virt(dev_priv);
> > +		} else {
> > +			drm_dbg_kms(&dev_priv->drm, "No PCH found.\n");
> > +		}
> > +	}
> >
> >  	pci_dev_put(pch);
> >  }
> 
> --
> Jani Nikula, Intel Open Source Graphics Center
Zhenyu Wang Dec. 24, 2020, 4:49 a.m. UTC | #6
On 2020.12.24 02:54:18 +0000, Zhang, Xiong Y wrote:
> > On Fri, 18 Dec 2020, Xiong Zhang <xiong.y.zhang@intel.com> wrote:
> > > From: Zhenyu Wang <zhenyuw@linux.intel.com>
> > >
> > > Some vmm like hyperv and crosvm don't supply any ISA bridge to their
> > > guest, when igd passthrough is equipped on these vmm, guest i915
> > > display may couldn't work as guest i915 detects PCH_NONE pch type.
> > >
> > > When i915 runs as guest, this patch guess pch type through gpu type
> > > even without ISA bridge.
> > >
> > > v2: Fix CI warning
> > >
> > > Signed-off-by: Zhenyu Wang <zhenyuw@linux.intel.com>
> > > ---
> > >  drivers/gpu/drm/i915/i915_drv.h  |  7 +++++-
> > > drivers/gpu/drm/i915/intel_pch.c | 38 ++++++++++++++++++++++----------
> > >  2 files changed, 32 insertions(+), 13 deletions(-)
> > >
> > > diff --git a/drivers/gpu/drm/i915/i915_drv.h
> > > b/drivers/gpu/drm/i915/i915_drv.h index 5a7df5621aa3..df0b8f9268b2
> > > 100644
> > > --- a/drivers/gpu/drm/i915/i915_drv.h
> > > +++ b/drivers/gpu/drm/i915/i915_drv.h
> > > @@ -1758,6 +1758,11 @@ tgl_revids_get(struct drm_i915_private
> > > *dev_priv)  #define INTEL_DISPLAY_ENABLED(dev_priv) \
> > >  	(drm_WARN_ON(&(dev_priv)->drm, !HAS_DISPLAY(dev_priv)),
> > > !(dev_priv)->params.disable_display)
> > >
> > > +static inline bool run_as_guest(void) {
> > > +	return !hypervisor_is_type(X86_HYPER_NATIVE);
> > > +}
> > > +
> > >  static inline bool intel_vtd_active(void)  {  #ifdef
> > > CONFIG_INTEL_IOMMU @@ -1766,7 +1771,7 @@ static inline bool
> > > intel_vtd_active(void)  #endif
> > >
> > >  	/* Running as a guest, we assume the host is enforcing VT'd */
> > > -	return !hypervisor_is_type(X86_HYPER_NATIVE);
> > > +	return run_as_guest();
> > >  }
> > >
> > >  static inline bool intel_scanout_needs_vtd_wa(struct drm_i915_private
> > > *dev_priv) diff --git a/drivers/gpu/drm/i915/intel_pch.c
> > > b/drivers/gpu/drm/i915/intel_pch.c
> > > index f31c0dabd0cc..a73c60bf349e 100644
> > > --- a/drivers/gpu/drm/i915/intel_pch.c
> > > +++ b/drivers/gpu/drm/i915/intel_pch.c
> > > @@ -184,6 +184,23 @@ intel_virt_detect_pch(const struct
> > drm_i915_private *dev_priv)
> > >  	return id;
> > >  }
> > >
> > > +static void intel_detect_pch_virt(struct drm_i915_private *dev_priv)
> > > +{
> > > +	unsigned short id;
> > > +	enum intel_pch pch_type;
> > > +
> > > +	id = intel_virt_detect_pch(dev_priv);
> > 
> > intel_detect_pch_virt() calls intel_virt_detect_pch()... the naming should be
> > clarified to make some difference.
>  [Zhang, Xiong Y] Indeed the naming is confusing. How about deleting this new added function, then move intel_pch_type() calling into original intel_virt_detect_pch(), and let intel_virt_detect_pch() return id and pch_type,  finally assign id and pch_type  in intel_detect_pch().
> 

May just put those in intel_virt_detect_pch() for all virt cases.

> > 
> > > +	pch_type = intel_pch_type(dev_priv, id);
> > > +
> > > +	/* Sanity check virtual PCH id */
> > > +	if (drm_WARN_ON(&dev_priv->drm,
> > > +			id && pch_type == PCH_NONE))
> > > +		id = 0;
> > > +
> > > +	dev_priv->pch_type = pch_type;
> > > +	dev_priv->pch_id = id;
> > 
> > Previously the pch type and id assignments were all done in
> > intel_detect_pch(), so moving this to a separate function in *some* but not
> > all cases reduces clarity too.
> > 
> > BR,
> > Jani.
> > 
> > > +}
> > > +
> > >  void intel_detect_pch(struct drm_i915_private *dev_priv)  {
> > >  	struct pci_dev *pch = NULL;
> > > @@ -221,16 +238,7 @@ void intel_detect_pch(struct drm_i915_private
> > *dev_priv)
> > >  			break;
> > >  		} else if (intel_is_virt_pch(id, pch->subsystem_vendor,
> > >  					     pch->subsystem_device)) {
> > > -			id = intel_virt_detect_pch(dev_priv);
> > > -			pch_type = intel_pch_type(dev_priv, id);
> > > -
> > > -			/* Sanity check virtual PCH id */
> > > -			if (drm_WARN_ON(&dev_priv->drm,
> > > -					id && pch_type == PCH_NONE))
> > > -				id = 0;
> > > -
> > > -			dev_priv->pch_type = pch_type;
> > > -			dev_priv->pch_id = id;
> > > +			intel_detect_pch_virt(dev_priv);
> > >  			break;
> > >  		}
> > >  	}
> > > @@ -246,8 +254,14 @@ void intel_detect_pch(struct drm_i915_private
> > *dev_priv)
> > >  		dev_priv->pch_id = 0;
> > >  	}
> > >
> > > -	if (!pch)
> > > -		drm_dbg_kms(&dev_priv->drm, "No PCH found.\n");
> > > +	if (!pch) {
> > > +		if (run_as_guest()) {
> > > +			drm_dbg_kms(&dev_priv->drm, "No PCH found in
> > vm, try guess..\n");
> > > +			intel_detect_pch_virt(dev_priv);
> > > +		} else {
> > > +			drm_dbg_kms(&dev_priv->drm, "No PCH found.\n");
> > > +		}
> > > +	}
> > >
> > >  	pci_dev_put(pch);
> > >  }
> > 
> > --
> > Jani Nikula, Intel Open Source Graphics Center
diff mbox series

Patch

diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index 5a7df5621aa3..df0b8f9268b2 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -1758,6 +1758,11 @@  tgl_revids_get(struct drm_i915_private *dev_priv)
 #define INTEL_DISPLAY_ENABLED(dev_priv) \
 	(drm_WARN_ON(&(dev_priv)->drm, !HAS_DISPLAY(dev_priv)), !(dev_priv)->params.disable_display)
 
+static inline bool run_as_guest(void)
+{
+	return !hypervisor_is_type(X86_HYPER_NATIVE);
+}
+
 static inline bool intel_vtd_active(void)
 {
 #ifdef CONFIG_INTEL_IOMMU
@@ -1766,7 +1771,7 @@  static inline bool intel_vtd_active(void)
 #endif
 
 	/* Running as a guest, we assume the host is enforcing VT'd */
-	return !hypervisor_is_type(X86_HYPER_NATIVE);
+	return run_as_guest();
 }
 
 static inline bool intel_scanout_needs_vtd_wa(struct drm_i915_private *dev_priv)
diff --git a/drivers/gpu/drm/i915/intel_pch.c b/drivers/gpu/drm/i915/intel_pch.c
index f31c0dabd0cc..a73c60bf349e 100644
--- a/drivers/gpu/drm/i915/intel_pch.c
+++ b/drivers/gpu/drm/i915/intel_pch.c
@@ -184,6 +184,23 @@  intel_virt_detect_pch(const struct drm_i915_private *dev_priv)
 	return id;
 }
 
+static void intel_detect_pch_virt(struct drm_i915_private *dev_priv)
+{
+	unsigned short id;
+	enum intel_pch pch_type;
+
+	id = intel_virt_detect_pch(dev_priv);
+	pch_type = intel_pch_type(dev_priv, id);
+
+	/* Sanity check virtual PCH id */
+	if (drm_WARN_ON(&dev_priv->drm,
+			id && pch_type == PCH_NONE))
+		id = 0;
+
+	dev_priv->pch_type = pch_type;
+	dev_priv->pch_id = id;
+}
+
 void intel_detect_pch(struct drm_i915_private *dev_priv)
 {
 	struct pci_dev *pch = NULL;
@@ -221,16 +238,7 @@  void intel_detect_pch(struct drm_i915_private *dev_priv)
 			break;
 		} else if (intel_is_virt_pch(id, pch->subsystem_vendor,
 					     pch->subsystem_device)) {
-			id = intel_virt_detect_pch(dev_priv);
-			pch_type = intel_pch_type(dev_priv, id);
-
-			/* Sanity check virtual PCH id */
-			if (drm_WARN_ON(&dev_priv->drm,
-					id && pch_type == PCH_NONE))
-				id = 0;
-
-			dev_priv->pch_type = pch_type;
-			dev_priv->pch_id = id;
+			intel_detect_pch_virt(dev_priv);
 			break;
 		}
 	}
@@ -246,8 +254,14 @@  void intel_detect_pch(struct drm_i915_private *dev_priv)
 		dev_priv->pch_id = 0;
 	}
 
-	if (!pch)
-		drm_dbg_kms(&dev_priv->drm, "No PCH found.\n");
+	if (!pch) {
+		if (run_as_guest()) {
+			drm_dbg_kms(&dev_priv->drm, "No PCH found in vm, try guess..\n");
+			intel_detect_pch_virt(dev_priv);
+		} else {
+			drm_dbg_kms(&dev_priv->drm, "No PCH found.\n");
+		}
+	}
 
 	pci_dev_put(pch);
 }