Skip to contentSkip to conversation

drm/i915: save/restore GMBUS freq across suspend/resume on gen4

PATCH Applied Jesse Barnes 5 participants 6 messages Dec 10, 2014 – Dec 15, 2014

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

PATCH 1/1 drm/i915: save/restore GMBUS freq across suspend/resume on gen4

Should probably just init this in the GMbus code all the time, based on the cdclk and HPLL like we do on newer platforms. Ville has code for that in a rework branch, but until then we can fix this bug fairly easily.

References: https://bugs.freedesktop.org/show_bug.cgi?id=76301

---
 drivers/gpu/drm/i915/i915_drv.h     | 1 +
 drivers/gpu/drm/i915/i915_reg.h     | 1 +
 drivers/gpu/drm/i915/i915_suspend.c | 8 ++++++++
 3 files changed, 10 insertions(+)
drivers/gpu/drm/i915/i915_drv.h +1 -0
index 2725243..f33102b 100644
@@ -925,6 +925,7 @@ struct i915_suspend_saved_registers {
 	u32 savePIPEB_LINK_N1;
 	u32 saveMCHBAR_RENDER_STANDBY;
 	u32 savePCH_PORT_HOTPLUG;
+	u16 saveGCDGMBUS;
 };
 
 struct vlv_s0ix_state {
drivers/gpu/drm/i915/i915_reg.h +1 -0
index dc03fac..f7c2856 100644
@@ -76,6 +76,7 @@
 #define   I915_GC_RENDER_CLOCK_166_MHZ	(0 << 0)
 #define   I915_GC_RENDER_CLOCK_200_MHZ	(1 << 0)
 #define   I915_GC_RENDER_CLOCK_333_MHZ	(4 << 0)
+#define GCDGMBUS 0xcc
 #define PCI_LBPC 0xf4 /* legacy/combination backlight modes, also called LBB */
 
 
drivers/gpu/drm/i915/i915_suspend.c +8 -0
index dfe6617..2636882 100644
@@ -303,6 +303,10 @@ int i915_save_state(struct drm_device *dev)
 		}
 	}
 
+	if (IS_GEN4(dev))
+		pci_read_config_word(dev->pdev, GCDGMBUS,
+				     &dev_priv->regfile.saveGCDGMBUS);
+
 	/* Cache mode state */
 	if (INTEL_INFO(dev)->gen < 7)
 		dev_priv->regfile.saveCACHE_MODE_0 = I915_READ(CACHE_MODE_0);
@@ -331,6 +335,10 @@ int i915_restore_state(struct drm_device *dev)
 	mutex_lock(&dev->struct_mutex);
 
 	i915_gem_restore_fences(dev);
+
+	if (IS_GEN4(dev))
+		pci_write_config_word(dev->pdev, GCDGMBUS,
+				      dev_priv->regfile.saveGCDGMBUS);
 	i915_restore_display(dev);
 
 	if (!drm_core_check_feature(dev, DRIVER_MODESET)) {
Signature · 6 lines
1.9.1

_______________________________________________
Intel-gfx mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx
Reviewed-by on v1 1/1
#1 lore raw
Quoted 7 lines · Jesse Barnes

Should probably just init this in the GMbus code all the time, based on the cdclk and HPLL like we do on newer platforms. Ville has code for that in a rework branch, but until then we can fix this bug fairly easily.

References: https://bugs.freedesktop.org/show_bug.cgi?id=76301

My cdclk extraction code doesn't seem to agree with this register for this particular bug reporter at least. So I think I need to go double check my code. The other options are that GMBUS clock isn't derived from cdclk on that platform, or that the HPLL/cdclk bits in configdb are simply not valid for this particular chipset.

In the meantime however, we can at least get some machines working with this patch. I'm not entirely sure which platforms have this register, but IS_GEN4() looks safe enough since my 946GZ has it and the reporter has a G41.

Quoted 63 lines
---
 drivers/gpu/drm/i915/i915_drv.h     | 1 +
 drivers/gpu/drm/i915/i915_reg.h     | 1 +
 drivers/gpu/drm/i915/i915_suspend.c | 8 ++++++++
 3 files changed, 10 insertions(+)
drivers/gpu/drm/i915/i915_drv.h +1 -0
index 2725243..f33102b 100644
@@ -925,6 +925,7 @@ struct i915_suspend_saved_registers {
 	u32 savePIPEB_LINK_N1;
 	u32 saveMCHBAR_RENDER_STANDBY;
 	u32 savePCH_PORT_HOTPLUG;
+	u16 saveGCDGMBUS;
 };
 
 struct vlv_s0ix_state {
drivers/gpu/drm/i915/i915_reg.h +1 -0
index dc03fac..f7c2856 100644
@@ -76,6 +76,7 @@
 #define   I915_GC_RENDER_CLOCK_166_MHZ	(0 << 0)
 #define   I915_GC_RENDER_CLOCK_200_MHZ	(1 << 0)
 #define   I915_GC_RENDER_CLOCK_333_MHZ	(4 << 0)
+#define GCDGMBUS 0xcc
 #define PCI_LBPC 0xf4 /* legacy/combination backlight modes, also called LBB */
 
 
drivers/gpu/drm/i915/i915_suspend.c +8 -0
index dfe6617..2636882 100644
@@ -303,6 +303,10 @@ int i915_save_state(struct drm_device *dev)
 		}
 	}
 
+	if (IS_GEN4(dev))
+		pci_read_config_word(dev->pdev, GCDGMBUS,
+				     &dev_priv->regfile.saveGCDGMBUS);
+
 	/* Cache mode state */
 	if (INTEL_INFO(dev)->gen < 7)
 		dev_priv->regfile.saveCACHE_MODE_0 = I915_READ(CACHE_MODE_0);
@@ -331,6 +335,10 @@ int i915_restore_state(struct drm_device *dev)
 	mutex_lock(&dev->struct_mutex);
 
 	i915_gem_restore_fences(dev);
+
+	if (IS_GEN4(dev))
+		pci_write_config_word(dev->pdev, GCDGMBUS,
+				      dev_priv->regfile.saveGCDGMBUS);
 	i915_restore_display(dev);
 
 	if (!drm_core_check_feature(dev, DRIVER_MODESET)) {
Signature · 6 lines
1.9.1

_______________________________________________
Intel-gfx mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx
Signature · 6 lines
Ville Syrjälä
Intel OTC
_______________________________________________
Intel-gfx mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx
#2 lore raw
Quoted 15 lines · On Wed, 10 Dec 2014 22:35:37 +0200 Ville Syrjälä…
Quoted 7 lines · Jesse Barnes

Should probably just init this in the GMbus code all the time, based on the cdclk and HPLL like we do on newer platforms. Ville has code for that in a rework branch, but until then we can fix this bug fairly easily.

References: https://bugs.freedesktop.org/show_bug.cgi?id=76301

My cdclk extraction code doesn't seem to agree with this register for this particular bug reporter at least. So I think I need to go double check my code. The other options are that GMBUS clock isn't derived from cdclk on that platform, or that the HPLL/cdclk bits in configdb are simply not valid for this particular chipset.

Quoted 6 lines

In the meantime however, we can at least get some machines working with this patch. I'm not entirely sure which platforms have this register, but IS_GEN4() looks safe enough since my 946GZ has it and the reporter has a G41.

Great, thanks. Jani, can you pick this up? I'll bounce the original over to stable@ too.

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

maintainer Applied
#3 lore raw
Quoted 23 lines · Jesse Barnes
Quoted 20 lines · On Wed, 10 Dec 2014 22:35:37 +0200 Ville Syrjälä…
Quoted 7 lines · Jesse Barnes

Should probably just init this in the GMbus code all the time, based on the cdclk and HPLL like we do on newer platforms. Ville has code for that in a rework branch, but until then we can fix this bug fairly easily.

References: https://bugs.freedesktop.org/show_bug.cgi?id=76301

My cdclk extraction code doesn't seem to agree with this register for this particular bug reporter at least. So I think I need to go double check my code. The other options are that GMBUS clock isn't derived from cdclk on that platform, or that the HPLL/cdclk bits in configdb are simply not valid for this particular chipset.

In the meantime however, we can at least get some machines working with this patch. I'm not entirely sure which platforms have this register, but IS_GEN4() looks safe enough since my 946GZ has it and the reporter has a G41.

Quoted 4 lines
Quoted 1 line

Great, thanks. Jani, can you pick this up? I'll bounce the original over to stable@ too.

Pushed to drm-intel-next-fixes, with cc: stable added. Thanks for the patch and review.

BR,
Jani

Quoted 7 lines

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

Signature · 5 lines
Jani Nikula, Intel Open Source Technology Center
_______________________________________________
Intel-gfx mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx
#4 lore raw
In reply to Ville Syrjälä
Quoted 15 lines · Ville Syrjälä
Quoted 7 lines · Jesse Barnes

Should probably just init this in the GMbus code all the time, based on the cdclk and HPLL like we do on newer platforms. Ville has code for that in a rework branch, but until then we can fix this bug fairly easily.

References: https://bugs.freedesktop.org/show_bug.cgi?id=76301

My cdclk extraction code doesn't seem to agree with this register for this particular bug reporter at least. So I think I need to go double check my code. The other options are that GMBUS clock isn't derived from cdclk on that platform, or that the HPLL/cdclk bits in configdb are simply not valid for this particular chipset.

Quoted 4 lines

In the meantime however, we can at least get some machines working with this patch. I'm not entirely sure which platforms have this register, but IS_GEN4() looks safe enough since my 946GZ has it and the reporter has a G41.

Yeah I really don't like the save/restore_state functions since they just bash registers without any regard to ordering constraints and stuff. I guess your cdclk series will address this? - At least we have the vlv_update_cdclk sprinkled all over already, probably better to call the intel_i2c_update_cdlk and move it to intel_i2c.c, too. -Daniel

Signature · 7 lines
Daniel Vetter
Software Engineer, Intel Corporation
+41 (0) 79 365 57 48 - http://blog.ffwll.ch
_______________________________________________
Intel-gfx mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx
Tested-by on v1 1/1
#5 lore raw
In reply to Jesse Barnes

-------------------------------------Summary-------------------------------------
Platform Delta drm-intel-nightly Series Applied
PNV 364/364 364/364
ILK +1 364/366 365/366

SNB                                  448/450              448/450
IVB                                  497/498              497/498
BYT                                  289/289              289/289
HSW                                  563/564              563/564
BDW                                  417/417              417/417

-------------------------------------Detailed-------------------------------------
Platform Test drm-intel-nightly Series Applied
ILK igt_kms_flip_wf_vblank-ts-check DMESG_WARN(6, M26)PASS(21, M26M37) PASS(1, M37)
Note: You need to pay more attention to line start with '*'
_______________________________________________
Intel-gfx mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx