Skip to contentSkip to conversation

drm/i915: add PCH_NONE to enum intel_pch

PATCH In review Paulo Zanoni 2 participants 10 messages Jul 3, 2012 – Jul 4, 2012

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

PATCH 1/3 drm/i915: add PCH_NONE to enum intel_pch

From: Paulo Zanoni <[email protected]>

And rely on the fact that it's 0 to assume that machines without a PCH will have PCH_NONE as dev_priv->pch_type.

Just today I finally realized that HAS_PCH_IBX is true for machines without a PCH. IMHO this is totally counter-intuitive and I don't think it's a good idea to assume that we're going to check for HAS_PCH_IBX only after we check for HAS_PCH_SPLIT.

I believe that in the future we'll have more PCH types and checks like:

    if (HAS_PCH_IBX(dev) || HAS_PCH_CPT(dev))

will become more and more common. There's a good chance that we may break non-PCH machines by adding these checks in code that runs on all machines. I also believe that the HAS_PCH_SPLIT check will become less common as we add more and more different PCH types.

Also: are we sure we don't already have any bugs triggered by checking for HAS_PCH_IBX on non-PCH machines?

---
 drivers/gpu/drm/i915/i915_drv.h |    1 +
 1 file changed, 1 insertion(+)

Another alternative would have been to change HAS_PCH_IBX to also check for HAS_PCH_SPLIT, but I'm not exactly in favor of adding more conditionals...

drivers/gpu/drm/i915/i915_drv.h +1 -0
index b7a1eaa..b12e79a 100644
@@ -333,6 +333,7 @@ enum no_fbc_reason {
 };
 
 enum intel_pch {
+	PCH_NONE = 0,	/* No PCH present */
 	PCH_IBX,	/* Ibexpeak PCH */
 	PCH_CPT,	/* Cougarpoint PCH */
 	PCH_LPT,	/* Lynxpoint PCH */
Signature · 1 line
1.7.10.2
#1 lore raw

PATCH 1/3 drm/i915: add PCH_NONE to enum intel_pch

From: Paulo Zanoni <[email protected]>

And rely on the fact that it's 0 to assume that machines without a PCH will have PCH_NONE as dev_priv->pch_type.

Just today I finally realized that HAS_PCH_IBX is true for machines without a PCH. IMHO this is totally counter-intuitive and I don't think it's a good idea to assume that we're going to check for HAS_PCH_IBX only after we check for HAS_PCH_SPLIT.

I believe that in the future we'll have more PCH types and checks like:

    if (HAS_PCH_IBX(dev) || HAS_PCH_CPT(dev))

will become more and more common. There's a good chance that we may break non-PCH machines by adding these checks in code that runs on all machines. I also believe that the HAS_PCH_SPLIT check will become less common as we add more and more different PCH types. We'll probably start replacing checks like:

    if (HAS_PCH_SPLIT(dev))
        foo();
    else
        bar();

with:

    if (HAS_PCH_NEW(dev))
        baz();
    else if (HAS_PCH_OLD(dev) || HAS_PCH_IBX(dev))
        foo();
    else
        bar();

and this may break gen 2/3/4.

As far as we have investigated, this patch will affect the behavior of intel_hdmi_dpms and intel_dp_link_down on gen 4. In both functions the code inside the HAS_PCH_IBX check is for IBX-specific workarounds, so we should be safe. If we start bisecting gen 2/3/4 bugs to this commit we should consider replacing the HAS_PCH_IBX checks with something else.

V2: Improve commit message, list possible side effects and solution.

---
 drivers/gpu/drm/i915/i915_drv.h |    1 +
 1 file changed, 1 insertion(+)

Another alternative would have been to change HAS_PCH_IBX to also check for HAS_PCH_SPLIT, but I'm not exactly in favor of adding more conditionals...

drivers/gpu/drm/i915/i915_drv.h +1 -0
index b7a1eaa..b12e79a 100644
@@ -333,6 +333,7 @@ enum no_fbc_reason {
 };
 
 enum intel_pch {
+	PCH_NONE = 0,	/* No PCH present */
 	PCH_IBX,	/* Ibexpeak PCH */
 	PCH_CPT,	/* Cougarpoint PCH */
 	PCH_LPT,	/* Lynxpoint PCH */
Signature · 1 line
1.7.10.2
#2 lore raw
Quoted 44 lines · Paulo Zanoni

From: Paulo Zanoni <[email protected]>

And rely on the fact that it's 0 to assume that machines without a PCH will have PCH_NONE as dev_priv->pch_type.

Just today I finally realized that HAS_PCH_IBX is true for machines without a PCH. IMHO this is totally counter-intuitive and I don't think it's a good idea to assume that we're going to check for HAS_PCH_IBX only after we check for HAS_PCH_SPLIT.

I believe that in the future we'll have more PCH types and checks like:

    if (HAS_PCH_IBX(dev) || HAS_PCH_CPT(dev))

will become more and more common. There's a good chance that we may break non-PCH machines by adding these checks in code that runs on all machines. I also believe that the HAS_PCH_SPLIT check will become less common as we add more and more different PCH types. We'll probably start replacing checks like:

    if (HAS_PCH_SPLIT(dev))
        foo();
    else
        bar();

with:

    if (HAS_PCH_NEW(dev))
        baz();
    else if (HAS_PCH_OLD(dev) || HAS_PCH_IBX(dev))
        foo();
    else
        bar();

and this may break gen 2/3/4.

As far as we have investigated, this patch will affect the behavior of intel_hdmi_dpms and intel_dp_link_down on gen 4. In both functions the code inside the HAS_PCH_IBX check is for IBX-specific workarounds, so we should be safe. If we start bisecting gen 2/3/4 bugs to this commit we should consider replacing the HAS_PCH_IBX checks with something else.

Quoted 3 lines

V2: Improve commit message, list possible side effects and solution.

All three patches queued for -next, thanks.
-Daniel

Signature · 3 lines
Daniel Vetter
Mail: [email protected]
Mobile: +41 (0)79 365 57 48
#3 lore raw

PATCH 2/3 drm/i915: get rid of dev_priv->info->has_pch_split

In reply to Paulo Zanoni

From: Paulo Zanoni <[email protected]>

Previously we had has_pch_split to tell us whether we had a PCH or not and we also had dev_priv->pch_type to tell us which kind of PCH it was, but it could only be used if we were 100% sure we did have a PCH. Now that PCH_NONE was added to dev_priv->pch_type we don't need has_pch_split anymore: we can just check for pch_type != PCH_NONE.

The HAS_PCH_{IBX,CPT,LPT} macros use dev_priv->pch_type, so they can only be called after intel_detect_pch. The HAS_PCH_SPLIT macro looks at dev_priv->info->has_pch_split, which is available earlier.

Since the goal is to implement HAS_PCH_SPLIT using dev_priv->pch_type instead of dev_priv->info->has_pch_split, we need to make sure that intel_detect_pch is called before any calls to HAS_PCH_SPLIT are made. So we moved the intel_detect_pch call to an earlier stage.

---
 drivers/gpu/drm/i915/i915_dma.c |    5 +++--
 drivers/gpu/drm/i915/i915_drv.c |    8 --------
 drivers/gpu/drm/i915/i915_drv.h |    3 +--
 3 files changed, 4 insertions(+), 12 deletions(-)

This patch does not solve any real problem: it's just a "suggestion" of something we could do after the previous patch. Some people may argue that looking at the has_pch_split variable might make it easier for us to find out which machines actually have a pch split without running the machine. So I really won't complain if we don't accept this patch: patch 01 is the important one.

drivers/gpu/drm/i915/i915_dma.c +3 -2
index 2166519..f8bc9ea 100644
@@ -1547,6 +1547,9 @@ int i915_driver_load(struct drm_device *dev, unsigned long flags)
 		goto out_mtrrfree;
 	}
 
+	/* This must be called before any calls to HAS_PCH_* */
+	intel_detect_pch(dev);
+
 	intel_irq_init(dev);
 
 	/* Try to make sure MCHBAR is enabled before poking at it */
@@ -1599,8 +1602,6 @@ int i915_driver_load(struct drm_device *dev, unsigned long flags)
 	/* Start out suspended */
 	dev_priv->mm.suspended = 1;
 
-	intel_detect_pch(dev);
-
 	if (drm_core_check_feature(dev, DRIVER_MODESET)) {
 		ret = i915_load_modeset_init(dev);
 		if (ret < 0) {
drivers/gpu/drm/i915/i915_drv.c +0 -8
index 7d0eb82..1794833 100644
@@ -225,7 +225,6 @@ static const struct intel_device_info intel_ironlake_d_info = {
 	.gen = 5,
 	.need_gfx_hws = 1, .has_hotplug = 1,
 	.has_bsd_ring = 1,
-	.has_pch_split = 1,
 };
 
 static const struct intel_device_info intel_ironlake_m_info = {
@@ -233,7 +232,6 @@ static const struct intel_device_info intel_ironlake_m_info = {
 	.need_gfx_hws = 1, .has_hotplug = 1,
 	.has_fbc = 1,
 	.has_bsd_ring = 1,
-	.has_pch_split = 1,
 };
 
 static const struct intel_device_info intel_sandybridge_d_info = {
@@ -242,7 +240,6 @@ static const struct intel_device_info intel_sandybridge_d_info = {
 	.has_bsd_ring = 1,
 	.has_blt_ring = 1,
 	.has_llc = 1,
-	.has_pch_split = 1,
 	.has_force_wake = 1,
 };
 
@@ -253,7 +250,6 @@ static const struct intel_device_info intel_sandybridge_m_info = {
 	.has_bsd_ring = 1,
 	.has_blt_ring = 1,
 	.has_llc = 1,
-	.has_pch_split = 1,
 	.has_force_wake = 1,
 };
 
@@ -263,7 +259,6 @@ static const struct intel_device_info intel_ivybridge_d_info = {
 	.has_bsd_ring = 1,
 	.has_blt_ring = 1,
 	.has_llc = 1,
-	.has_pch_split = 1,
 	.has_force_wake = 1,
 };
 
@@ -274,7 +269,6 @@ static const struct intel_device_info intel_ivybridge_m_info = {
 	.has_bsd_ring = 1,
 	.has_blt_ring = 1,
 	.has_llc = 1,
-	.has_pch_split = 1,
 	.has_force_wake = 1,
 };
 
@@ -302,7 +296,6 @@ static const struct intel_device_info intel_haswell_d_info = {
 	.has_bsd_ring = 1,
 	.has_blt_ring = 1,
 	.has_llc = 1,
-	.has_pch_split = 1,
 	.has_force_wake = 1,
 };
 
@@ -312,7 +305,6 @@ static const struct intel_device_info intel_haswell_m_info = {
 	.has_bsd_ring = 1,
 	.has_blt_ring = 1,
 	.has_llc = 1,
-	.has_pch_split = 1,
 	.has_force_wake = 1,
 };
 
drivers/gpu/drm/i915/i915_drv.h +1 -2
index b12e79a..89025ab 100644
@@ -285,7 +285,6 @@ struct intel_device_info {
 	u8 is_crestline:1;
 	u8 is_ivybridge:1;
 	u8 is_valleyview:1;
-	u8 has_pch_split:1;
 	u8 has_force_wake:1;
 	u8 is_haswell:1;
 	u8 has_fbc:1;
@@ -1113,13 +1112,13 @@ struct drm_i915_file_private {
 #define HAS_PIPE_CXSR(dev) (INTEL_INFO(dev)->has_pipe_cxsr)
 #define I915_HAS_FBC(dev) (INTEL_INFO(dev)->has_fbc)
 
-#define HAS_PCH_SPLIT(dev) (INTEL_INFO(dev)->has_pch_split)
 #define HAS_PIPE_CONTROL(dev) (INTEL_INFO(dev)->gen >= 5)
 
 #define INTEL_PCH_TYPE(dev) (((struct drm_i915_private *)(dev)->dev_private)->pch_type)
 #define HAS_PCH_LPT(dev) (INTEL_PCH_TYPE(dev) == PCH_LPT)
 #define HAS_PCH_CPT(dev) (INTEL_PCH_TYPE(dev) == PCH_CPT)
 #define HAS_PCH_IBX(dev) (INTEL_PCH_TYPE(dev) == PCH_IBX)
+#define HAS_PCH_SPLIT(dev) (INTEL_PCH_TYPE(dev) != PCH_NONE)
 
 #define HAS_FORCE_WAKE(dev) (INTEL_INFO(dev)->has_force_wake)
 
Signature · 1 line
1.7.10.2
#4 lore raw

Re: [PATCH 2/3] drm/i915: get rid of dev_priv->info->has_pch_split

Quoted 36 lines · Paulo Zanoni

From: Paulo Zanoni <[email protected]>

Previously we had has_pch_split to tell us whether we had a PCH or not and we also had dev_priv->pch_type to tell us which kind of PCH it was, but it could only be used if we were 100% sure we did have a PCH. Now that PCH_NONE was added to dev_priv->pch_type we don't need has_pch_split anymore: we can just check for pch_type != PCH_NONE.

The HAS_PCH_{IBX,CPT,LPT} macros use dev_priv->pch_type, so they can only be called after intel_detect_pch. The HAS_PCH_SPLIT macro looks at dev_priv->info->has_pch_split, which is available earlier.

Since the goal is to implement HAS_PCH_SPLIT using dev_priv->pch_type instead of dev_priv->info->has_pch_split, we need to make sure that intel_detect_pch is called before any calls to HAS_PCH_SPLIT are made. So we moved the intel_detect_pch call to an earlier stage.

---
 drivers/gpu/drm/i915/i915_dma.c |    5 +++--
 drivers/gpu/drm/i915/i915_drv.c |    8 --------
 drivers/gpu/drm/i915/i915_drv.h |    3 +--
 3 files changed, 4 insertions(+), 12 deletions(-)

This patch does not solve any real problem: it's just a "suggestion" of something we could do after the previous patch. Some people may argue that looking at the has_pch_split variable might make it easier for us to find out which machines actually have a pch split without running the machine. So I really won't complain if we don't accept this patch: patch 01 is the important one.

drivers/gpu/drm/i915/i915_dma.c +0 -0
index 2166519..f8bc9ea 100644

drivers/gpu/drm/i915/i915_dma.c int i915_driver_load(struct drm_device *dev, unsigned long flags) · line 1552

Quoted 6 lines
drivers/gpu/drm/i915/i915_dma.c +3 -0
@@ -1547,6 +1547,9 @@ int i915_driver_load(struct drm_device *dev, unsigned long flags)
 		goto out_mtrrfree;
 	}
 
+	/* This must be called before any calls to HAS_PCH_* */
+	intel_detect_pch(dev);
+

Hm, what about defining PCH_NONE as -1, PCH_RESERVED as 0 and then adding a WARN_ON(dev_priv->pch_type == PCH_RESERVED)? detect_pch is called unconditionally, and that way we would catch this. Might be overkill otoh, so if you think this is not worth it, np.
-Daniel

Quoted 114 lines
drivers/gpu/drm/i915/i915_dma.c +0 -2
 	intel_irq_init(dev);
 
 	/* Try to make sure MCHBAR is enabled before poking at it */
@@ -1599,8 +1602,6 @@ int i915_driver_load(struct drm_device *dev, unsigned long flags)
 	/* Start out suspended */
 	dev_priv->mm.suspended = 1;
 
-	intel_detect_pch(dev);
-
 	if (drm_core_check_feature(dev, DRIVER_MODESET)) {
 		ret = i915_load_modeset_init(dev);
 		if (ret < 0) {
drivers/gpu/drm/i915/i915_drv.c +0 -8
index 7d0eb82..1794833 100644
@@ -225,7 +225,6 @@ static const struct intel_device_info intel_ironlake_d_info = {
 	.gen = 5,
 	.need_gfx_hws = 1, .has_hotplug = 1,
 	.has_bsd_ring = 1,
-	.has_pch_split = 1,
 };
 
 static const struct intel_device_info intel_ironlake_m_info = {
@@ -233,7 +232,6 @@ static const struct intel_device_info intel_ironlake_m_info = {
 	.need_gfx_hws = 1, .has_hotplug = 1,
 	.has_fbc = 1,
 	.has_bsd_ring = 1,
-	.has_pch_split = 1,
 };
 
 static const struct intel_device_info intel_sandybridge_d_info = {
@@ -242,7 +240,6 @@ static const struct intel_device_info intel_sandybridge_d_info = {
 	.has_bsd_ring = 1,
 	.has_blt_ring = 1,
 	.has_llc = 1,
-	.has_pch_split = 1,
 	.has_force_wake = 1,
 };
 
@@ -253,7 +250,6 @@ static const struct intel_device_info intel_sandybridge_m_info = {
 	.has_bsd_ring = 1,
 	.has_blt_ring = 1,
 	.has_llc = 1,
-	.has_pch_split = 1,
 	.has_force_wake = 1,
 };
 
@@ -263,7 +259,6 @@ static const struct intel_device_info intel_ivybridge_d_info = {
 	.has_bsd_ring = 1,
 	.has_blt_ring = 1,
 	.has_llc = 1,
-	.has_pch_split = 1,
 	.has_force_wake = 1,
 };
 
@@ -274,7 +269,6 @@ static const struct intel_device_info intel_ivybridge_m_info = {
 	.has_bsd_ring = 1,
 	.has_blt_ring = 1,
 	.has_llc = 1,
-	.has_pch_split = 1,
 	.has_force_wake = 1,
 };
 
@@ -302,7 +296,6 @@ static const struct intel_device_info intel_haswell_d_info = {
 	.has_bsd_ring = 1,
 	.has_blt_ring = 1,
 	.has_llc = 1,
-	.has_pch_split = 1,
 	.has_force_wake = 1,
 };
 
@@ -312,7 +305,6 @@ static const struct intel_device_info intel_haswell_m_info = {
 	.has_bsd_ring = 1,
 	.has_blt_ring = 1,
 	.has_llc = 1,
-	.has_pch_split = 1,
 	.has_force_wake = 1,
 };
 
drivers/gpu/drm/i915/i915_drv.h +1 -2
index b12e79a..89025ab 100644
@@ -285,7 +285,6 @@ struct intel_device_info {
 	u8 is_crestline:1;
 	u8 is_ivybridge:1;
 	u8 is_valleyview:1;
-	u8 has_pch_split:1;
 	u8 has_force_wake:1;
 	u8 is_haswell:1;
 	u8 has_fbc:1;
@@ -1113,13 +1112,13 @@ struct drm_i915_file_private {
 #define HAS_PIPE_CXSR(dev) (INTEL_INFO(dev)->has_pipe_cxsr)
 #define I915_HAS_FBC(dev) (INTEL_INFO(dev)->has_fbc)
 
-#define HAS_PCH_SPLIT(dev) (INTEL_INFO(dev)->has_pch_split)
 #define HAS_PIPE_CONTROL(dev) (INTEL_INFO(dev)->gen >= 5)
 
 #define INTEL_PCH_TYPE(dev) (((struct drm_i915_private *)(dev)->dev_private)->pch_type)
 #define HAS_PCH_LPT(dev) (INTEL_PCH_TYPE(dev) == PCH_LPT)
 #define HAS_PCH_CPT(dev) (INTEL_PCH_TYPE(dev) == PCH_CPT)
 #define HAS_PCH_IBX(dev) (INTEL_PCH_TYPE(dev) == PCH_IBX)
+#define HAS_PCH_SPLIT(dev) (INTEL_PCH_TYPE(dev) != PCH_NONE)
 
 #define HAS_FORCE_WAKE(dev) (INTEL_INFO(dev)->has_force_wake)
 
Signature · 6 lines
1.7.10.2

_______________________________________________
Intel-gfx mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx
Signature · 3 lines
Daniel Vetter
Mail: [email protected]
Mobile: +41 (0)79 365 57 48
#5 lore raw

Re: [PATCH 2/3] drm/i915: get rid of dev_priv->info->has_pch_split

2012/7/3 Daniel Vetter <[email protected]>:

Quoted 4 lines

Hm, what about defining PCH_NONE as -1, PCH_RESERVED as 0 and then adding a WARN_ON(dev_priv->pch_type == PCH_RESERVED)? detect_pch is called unconditionally, and that way we would catch this. Might be overkill otoh, so if you think this is not worth it, np.

I actually thought about this idea before. But it would only make sense if we add these WARNs to the HAS_PCH_FOO macros. But these macros are supposed to be simple and cheap and fast... Do we really want to start adding assertions inside them? If we think the price is worth paying, then we might do as you suggested. Or maybe this could be inside some #ifdef DEBUG...

On my local machines, I changed the HAS_PCH_FOO macros to print some stuff so I could check whether any of them was called before intel_detect_pch. At least on these machines (SNB, HSW and a netbook), everything was fine.

Quoted 121 lines

-Daniel

Quoted 110 lines
      intel_irq_init(dev);

      /* Try to make sure MCHBAR is enabled before poking at it */
drivers/gpu/drm/i915/i915_dma.c +0 -2
@@ -1599,8 +1602,6 @@ int i915_driver_load(struct drm_device *dev, unsigned long flags)
      /* Start out suspended */
      dev_priv->mm.suspended = 1;
 
-     intel_detect_pch(dev);
-
      if (drm_core_check_feature(dev, DRIVER_MODESET)) {
              ret = i915_load_modeset_init(dev);
              if (ret < 0) {
drivers/gpu/drm/i915/i915_drv.c +0 -8
index 7d0eb82..1794833 100644
@@ -225,7 +225,6 @@ static const struct intel_device_info intel_ironlake_d_info = {
      .gen = 5,
      .need_gfx_hws = 1, .has_hotplug = 1,
      .has_bsd_ring = 1,
-     .has_pch_split = 1,
 };
 
 static const struct intel_device_info intel_ironlake_m_info = {
@@ -233,7 +232,6 @@ static const struct intel_device_info intel_ironlake_m_info = {
      .need_gfx_hws = 1, .has_hotplug = 1,
      .has_fbc = 1,
      .has_bsd_ring = 1,
-     .has_pch_split = 1,
 };
 
 static const struct intel_device_info intel_sandybridge_d_info = {
@@ -242,7 +240,6 @@ static const struct intel_device_info intel_sandybridge_d_info = {
      .has_bsd_ring = 1,
      .has_blt_ring = 1,
      .has_llc = 1,
-     .has_pch_split = 1,
      .has_force_wake = 1,
 };
 
@@ -253,7 +250,6 @@ static const struct intel_device_info intel_sandybridge_m_info = {
      .has_bsd_ring = 1,
      .has_blt_ring = 1,
      .has_llc = 1,
-     .has_pch_split = 1,
      .has_force_wake = 1,
 };
 
@@ -263,7 +259,6 @@ static const struct intel_device_info intel_ivybridge_d_info = {
      .has_bsd_ring = 1,
      .has_blt_ring = 1,
      .has_llc = 1,
-     .has_pch_split = 1,
      .has_force_wake = 1,
 };
 
@@ -274,7 +269,6 @@ static const struct intel_device_info intel_ivybridge_m_info = {
      .has_bsd_ring = 1,
      .has_blt_ring = 1,
      .has_llc = 1,
-     .has_pch_split = 1,
      .has_force_wake = 1,
 };
 
@@ -302,7 +296,6 @@ static const struct intel_device_info intel_haswell_d_info = {
      .has_bsd_ring = 1,
      .has_blt_ring = 1,
      .has_llc = 1,
-     .has_pch_split = 1,
      .has_force_wake = 1,
 };
 
@@ -312,7 +305,6 @@ static const struct intel_device_info intel_haswell_m_info = {
      .has_bsd_ring = 1,
      .has_blt_ring = 1,
      .has_llc = 1,
-     .has_pch_split = 1,
      .has_force_wake = 1,
 };
 
drivers/gpu/drm/i915/i915_drv.h +1 -3
index b12e79a..89025ab 100644
@@ -285,7 +285,6 @@ struct intel_device_info {
      u8 is_crestline:1;
      u8 is_ivybridge:1;
      u8 is_valleyview:1;
-     u8 has_pch_split:1;
      u8 has_force_wake:1;
      u8 is_haswell:1;
      u8 has_fbc:1;
@@ -1113,13 +1112,13 @@ struct drm_i915_file_private {
 #define HAS_PIPE_CXSR(dev) (INTEL_INFO(dev)->has_pipe_cxsr)
 #define I915_HAS_FBC(dev) (INTEL_INFO(dev)->has_fbc)
 
-#define HAS_PCH_SPLIT(dev) (INTEL_INFO(dev)->has_pch_split)
 #define HAS_PIPE_CONTROL(dev) (INTEL_INFO(dev)->gen >= 5)
 
 #define INTEL_PCH_TYPE(dev) (((struct drm_i915_private *)(dev)->dev_private)->pch_type)
 #define HAS_PCH_LPT(dev) (INTEL_PCH_TYPE(dev) == PCH_LPT)
 #define HAS_PCH_CPT(dev) (INTEL_PCH_TYPE(dev) == PCH_CPT)
 #define HAS_PCH_IBX(dev) (INTEL_PCH_TYPE(dev) == PCH_IBX)
+#define HAS_PCH_SPLIT(dev) (INTEL_PCH_TYPE(dev) != PCH_NONE)
 
 #define HAS_FORCE_WAKE(dev) (INTEL_INFO(dev)->has_force_wake)
 
--

1.7.10.2

Quoted 4 lines

_______________________________________________
Intel-gfx mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

--
Daniel Vetter
Mail: [email protected]
Mobile: +41 (0)79 365 57 48

Signature · 1 line
Paulo Zanoni
#6 lore raw
In reply to Paulo Zanoni
Quoted 15 lines · Paulo Zanoni

From: Paulo Zanoni <[email protected]>

And rely on the fact that it's 0 to assume that machines without a PCH will have PCH_NONE as dev_priv->pch_type.

Just today I finally realized that HAS_PCH_IBX is true for machines without a PCH. IMHO this is totally counter-intuitive and I don't think it's a good idea to assume that we're going to check for HAS_PCH_IBX only after we check for HAS_PCH_SPLIT.

I believe that in the future we'll have more PCH types and checks like:

    if (HAS_PCH_IBX(dev) || HAS_PCH_CPT(dev))
Quoted 7 lines

will become more and more common. There's a good chance that we may break non-PCH machines by adding these checks in code that runs on all machines. I also believe that the HAS_PCH_SPLIT check will become less common as we add more and more different PCH types.

Also: are we sure we don't already have any bugs triggered by checking for HAS_PCH_IBX on non-PCH machines?

I think most of the HAS_PCH_xxx are implicitly guarded because we've split up the pch modeset into it's own functions. I think there might only be a few issues in the encoder functions maybe. Have your checked all the HAS_PCH_IBX checks there? If you want, I can go through the code, too.

Otherwise I really like this.
-Daniel

Quoted 29 lines
---
 drivers/gpu/drm/i915/i915_drv.h |    1 +
 1 file changed, 1 insertion(+)

Another alternative would have been to change HAS_PCH_IBX to also check for HAS_PCH_SPLIT, but I'm not exactly in favor of adding more conditionals...

drivers/gpu/drm/i915/i915_drv.h +1 -0
index b7a1eaa..b12e79a 100644
@@ -333,6 +333,7 @@ enum no_fbc_reason {
 };
 
 enum intel_pch {
+	PCH_NONE = 0,	/* No PCH present */
 	PCH_IBX,	/* Ibexpeak PCH */
 	PCH_CPT,	/* Cougarpoint PCH */
 	PCH_LPT,	/* Lynxpoint PCH */
Signature · 6 lines
1.7.10.2

_______________________________________________
Intel-gfx mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx
Signature · 3 lines
Daniel Vetter
Mail: [email protected]
Mobile: +41 (0)79 365 57 48
#7 lore raw

2012/7/3 Daniel Vetter <[email protected]>:

Quoted 5 lines

I think most of the HAS_PCH_xxx are implicitly guarded because we've split up the pch modeset into it's own functions. I think there might only be a few issues in the encoder functions maybe. Have your checked all the HAS_PCH_IBX checks there? If you want, I can go through the code, too.

I did check. At the moment we have just a few HAS_PCH_IBX calls in our driver. The only possible issues may be inside intel_hdmi.c and intel_dp.c (and they need more investigation).

My biggest worry here is being "future-proof": are we sure whenever someone suggests adding HAS_PCH_IBX we'll remember that machines without a PCH return true for HAS_PCH_IBX? This is highly counter-intuitive. I really think that in future hardware enablement code we'll replace a lot of the "if (HAS_PCH_SPLIT) { foo(); } else { bar(); }" code for "if (HAS_PCH_NEW) { baz(); } else if (HAS_PCH_OLD) { foo(); } else { bar(); }".

Thanks,
Paulo

Quoted 36 lines

Otherwise I really like this.
-Daniel

Quoted 25 lines
---
 drivers/gpu/drm/i915/i915_drv.h |    1 +
 1 file changed, 1 insertion(+)

Another alternative would have been to change HAS_PCH_IBX to also check for HAS_PCH_SPLIT, but I'm not exactly in favor of adding more conditionals...

drivers/gpu/drm/i915/i915_drv.h +1 -1
index b7a1eaa..b12e79a 100644
@@ -333,6 +333,7 @@ enum no_fbc_reason {
 };
 
 enum intel_pch {
+     PCH_NONE = 0,   /* No PCH present */
      PCH_IBX,        /* Ibexpeak PCH */
      PCH_CPT,        /* Cougarpoint PCH */
      PCH_LPT,        /* Lynxpoint PCH */
--

1.7.10.2

Quoted 4 lines

_______________________________________________
Intel-gfx mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

--
Daniel Vetter
Mail: [email protected]
Mobile: +41 (0)79 365 57 48

Signature · 1 line
Paulo Zanoni
#8 lore raw
Quoted 11 lines · Paulo Zanoni

2012/7/3 Daniel Vetter <[email protected]>:

Quoted 5 lines

I think most of the HAS_PCH_xxx are implicitly guarded because we've split up the pch modeset into it's own functions. I think there might only be a few issues in the encoder functions maybe. Have your checked all the HAS_PCH_IBX checks there? If you want, I can go through the code, too.

I did check. At the moment we have just a few HAS_PCH_IBX calls in our driver. The only possible issues may be inside intel_hdmi.c and intel_dp.c (and they need more investigation).

Quoted 7 lines

My biggest worry here is being "future-proof": are we sure whenever someone suggests adding HAS_PCH_IBX we'll remember that machines without a PCH return true for HAS_PCH_IBX? This is highly counter-intuitive. I really think that in future hardware enablement code we'll replace a lot of the "if (HAS_PCH_SPLIT) { foo(); } else { bar(); }" code for "if (HAS_PCH_NEW) { baz(); } else if (HAS_PCH_OLD) { foo(); } else { bar(); }".

Ok, I've quickly checked them. The one in intel_dp.c isn't an issue, because DP is a gen5+ feature. So the only thing accidentally affected is vlv, which isn't such a big deal ;-)

The only other check that isn't guarded with a HAS_PCH_SPLIT check is in intel_hdmi.c for a ibx-only w/a. That one will also leak out into gm45 platforms (which support hdmi, too).

Otherwise I haven't found anything. Can you please amend the commit message detailing the effects on these two places? Just in case a bisect hits this patch and someone is totally confused what's going on here ...
-Daniel

Signature · 2 lines
Daniel Vetter
[email protected] - +41 (0) 79 364 57 48 - http://blog.ffwll.ch
#9 lore raw

PATCH 3/3 drm/i915: don't ironlake_init_pch_refclk() on LPT

In reply to Paulo Zanoni

From: Paulo Zanoni <[email protected]>

This function is used to set the PCH_DREF_CONTROL register, which does not exist on LPT anymore.

---
 drivers/gpu/drm/i915/i915_drv.c      |    2 +-
 drivers/gpu/drm/i915/intel_display.c |    2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

My crystal ball tells me patches like this will become more and more common in the future... That's why patch 01 was proposed.

This patch depends on patch 01, otherwise we'll start calling ironlake_init_pch_refclk on non-pch machines.

drivers/gpu/drm/i915/i915_drv.c +1 -1
index 1794833..0630471 100644
@@ -687,7 +687,7 @@ static int i915_drm_thaw(struct drm_device *dev)
 
 	/* KMS EnterVT equivalent */
 	if (drm_core_check_feature(dev, DRIVER_MODESET)) {
-		if (HAS_PCH_SPLIT(dev))
+		if (HAS_PCH_IBX(dev) || HAS_PCH_CPT(dev))
 			ironlake_init_pch_refclk(dev);
 
 		mutex_lock(&dev->struct_mutex);
drivers/gpu/drm/i915/intel_display.c +1 -1
index 8cd5032..da33911 100644
@@ -6855,7 +6855,7 @@ static void intel_setup_outputs(struct drm_device *dev)
 	/* disable all the possible outputs/crtcs before entering KMS mode */
 	drm_helper_disable_unused_functions(dev);
 
-	if (HAS_PCH_SPLIT(dev))
+	if (HAS_PCH_IBX(dev) || HAS_PCH_CPT(dev))
 		ironlake_init_pch_refclk(dev);
 }
 
Signature · 1 line
1.7.10.2