Merge branch 'drm-fixes-intel' of git://people.freedesktop.org/~airlied/linux

Pull drm update from Dave Airlie:
 "This pull just contains a forward of the Intel fixes from Daniel.

  The only annoyance is the RC6 enable, which really should have made
  -next, but since Ubuntu are shipping it I reckon its getting a good
  testing now by the time 3.4 comes out.

  The pull from Daniel contains his pull message to me:

    "A few patches for 3.4, major part is 3 regression fixes:
     - ppgtt broke hibernate on snb/ivb.  Somehow our QA claims that it
       still works, which is why this has not been caught earlier.
     - ppgtt flails in combination with dmar.  I kinda expected this one :(
     - fence handling bugfix for gen2/3.  Iirc this one is about a year
       old, fix curtesy Chris Wilson.  I've created an shockingly simple
       i-g-t test to catch this in the future."

     Wrt regressions I've just got a report that gmbus (newly enabled
     again in 3.4) is a bit noisy.  I'm looking into this atm.

     Also included are the rc6 enable patches for snb from Eugeni.  I
     wanted to include these in the main 3.4 pull but screwed it up.
     Please hit me.  Imo these kind of patches really should go in
     before -rc1, but in thise case rc6 has brought us tons of press and
     guinea pigs^W^W testers and ubuntu is already running with it.  So
     I estimate a pretty small chance for this to blow up.

     And some smaller things:
      - two minor locking snafus
      - server gt2 ivb pciid
      - 2 patches to sanitize the register state left behind by the bios
        some more
      - 2 new quirk entries
      - cs readback trick against missed IRQs from ivb also enabled on snb
      - sprite fix from Jesse"

Let's see if the "enable RC6 on sandybridge" finally works and sticks.
I've been enabling it by hand (i915.i915_enable_rc6=1) for several
months on my Macbook Air, and it definitely makes a difference (and has
worked for me).  But every time we enabled it before it showed some odd
hw buglet for *somebody*.

This time it's all good, I'm sure.

* 'drm-fixes-intel' of git://people.freedesktop.org/~airlied/linux:
  drm/i915: treat src w & h as fixed point in sprite handling code
  drm/i915: no-lvds quirk on MSI DC500
  drm/i915: Add lock on drm_helper_resume_force_mode
  drm/i915: don't leak struct_mutex lock on ppgtt init failures
  drm/i915: disable ppgtt on snb when dmar is enabled
  drm/i915: add Ivy Bridge GT2 Server entries
  drm/i915: properly clear SSC1 bit in the pch refclock init code
  drm/i915: apply CS reg readback trick against missed IRQ on snb
  drm/i915: quirk away broken OpRegion VBT
  drm/i915: enable plain RC6 on Sandy Bridge by default
  drm/i915: allow to select rc6 modes via kernel parameter
  drm/i915: Mark untiled BLT commands as fenced on gen2/3
  drm/i915: properly restore the ppgtt page directory on resume
  drm/i915: Sanitize BIOS debugging bits from PIPECONF

drivers/char/agp/intel-agp.h

@@ -234,6 +234,7 @@
 #define PCI_DEVICE_ID_INTEL_IVYBRIDGE_M_GT2_IG		0x0166
 #define PCI_DEVICE_ID_INTEL_IVYBRIDGE_S_HB		0x0158  /* Server */
 #define PCI_DEVICE_ID_INTEL_IVYBRIDGE_S_GT1_IG		0x015A
+#define PCI_DEVICE_ID_INTEL_IVYBRIDGE_S_GT2_IG		0x016A
 
 int intel_gmch_probe(struct pci_dev *pdev,
 			       struct agp_bridge_data *bridge);

drivers/char/agp/intel-gtt.c

@@ -1190,7 +1190,6 @@ static inline int needs_idle_maps(void)
 {
 #ifdef CONFIG_INTEL_IOMMU
 	const unsigned short gpu_devid = intel_private.pcidev->device;
-	extern int intel_iommu_gfx_mapped;
 
 	/* Query intel_iommu to see if we need the workaround. Presumably that
 	 * was loaded first.
@@ -1459,6 +1458,8 @@ static const struct intel_gtt_driver_description {
 	    "Ivybridge", &sandybridge_gtt_driver },
 	{ PCI_DEVICE_ID_INTEL_IVYBRIDGE_S_GT1_IG,
 	    "Ivybridge", &sandybridge_gtt_driver },
+	{ PCI_DEVICE_ID_INTEL_IVYBRIDGE_S_GT2_IG,
+	    "Ivybridge", &sandybridge_gtt_driver },
 	{ 0, NULL, NULL }
 };
 

drivers/gpu/drm/i915/i915_dma.c

@@ -1183,6 +1183,21 @@ static bool i915_switcheroo_can_switch(struct pci_dev *pdev)
 	return can_switch;
 }
 
+static bool
+intel_enable_ppgtt(struct drm_device *dev)
+{
+	if (i915_enable_ppgtt >= 0)
+		return i915_enable_ppgtt;
+
+#ifdef CONFIG_INTEL_IOMMU
+	/* Disable ppgtt on SNB if VT-d is on. */
+	if (INTEL_INFO(dev)->gen == 6 && intel_iommu_gfx_mapped)
+		return false;
+#endif
+
+	return true;
+}
+
 static int i915_load_gem_init(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -1197,7 +1212,7 @@ static int i915_load_gem_init(struct drm_device *dev)
 	drm_mm_init(&dev_priv->mm.stolen, 0, prealloc_size);
 
 	mutex_lock(&dev->struct_mutex);
-	if (i915_enable_ppgtt && HAS_ALIASING_PPGTT(dev)) {
+	if (intel_enable_ppgtt(dev) && HAS_ALIASING_PPGTT(dev)) {
 		/* PPGTT pdes are stolen from global gtt ptes, so shrink the
 		 * aperture accordingly when using aliasing ppgtt. */
 		gtt_size -= I915_PPGTT_PD_ENTRIES*PAGE_SIZE;
@@ -1207,8 +1222,10 @@ static int i915_load_gem_init(struct drm_device *dev)
 		i915_gem_do_init(dev, 0, mappable_size, gtt_size);
 
 		ret = i915_gem_init_aliasing_ppgtt(dev);
-		if (ret)
+		if (ret) {
+			mutex_unlock(&dev->struct_mutex);
 			return ret;
+		}
 	} else {
 		/* Let GEM Manage all of the aperture.
 		 *

drivers/gpu/drm/i915/i915_drv.c

@@ -66,7 +66,11 @@ MODULE_PARM_DESC(semaphores,
 int i915_enable_rc6 __read_mostly = -1;
 module_param_named(i915_enable_rc6, i915_enable_rc6, int, 0600);
 MODULE_PARM_DESC(i915_enable_rc6,
-		"Enable power-saving render C-state 6 (default: -1 (use per-chip default)");
+		"Enable power-saving render C-state 6. "
+		"Different stages can be selected via bitmask values "
+		"(0 = disable; 1 = enable rc6; 2 = enable deep rc6; 4 = enable deepest rc6). "
+		"For example, 3 would enable rc6 and deep rc6, and 7 would enable everything. "
+		"default: -1 (use per-chip default)");
 
 int i915_enable_fbc __read_mostly = -1;
 module_param_named(i915_enable_fbc, i915_enable_fbc, int, 0600);
@@ -103,8 +107,8 @@ MODULE_PARM_DESC(enable_hangcheck,
 		"WARNING: Disabling this can cause system wide hangs. "
 		"(default: true)");
 
-bool i915_enable_ppgtt __read_mostly = 1;
-module_param_named(i915_enable_ppgtt, i915_enable_ppgtt, bool, 0600);
+int i915_enable_ppgtt __read_mostly = -1;
+module_param_named(i915_enable_ppgtt, i915_enable_ppgtt, int, 0600);
 MODULE_PARM_DESC(i915_enable_ppgtt,
 		"Enable PPGTT (default: true)");
 
@@ -292,6 +296,7 @@ static const struct pci_device_id pciidlist[] = {		/* aka */
 	INTEL_VGA_DEVICE(0x0152, &intel_ivybridge_d_info), /* GT1 desktop */
 	INTEL_VGA_DEVICE(0x0162, &intel_ivybridge_d_info), /* GT2 desktop */
 	INTEL_VGA_DEVICE(0x015a, &intel_ivybridge_d_info), /* GT1 server */
+	INTEL_VGA_DEVICE(0x016a, &intel_ivybridge_d_info), /* GT2 server */
 	{0, 0, 0}
 };
 
@@ -533,7 +538,9 @@ static int i915_drm_thaw(struct drm_device *dev)
 		drm_irq_install(dev);
 
 		/* Resume the modeset for every activated CRTC */
+		mutex_lock(&dev->mode_config.mutex);
 		drm_helper_resume_force_mode(dev);
+		mutex_unlock(&dev->mode_config.mutex);
 
 		if (IS_IRONLAKE_M(dev))
 			ironlake_enable_rc6(dev);

drivers/gpu/drm/i915/i915_drv.h

@@ -1053,6 +1053,27 @@ struct drm_i915_file_private {
 
 #include "i915_trace.h"
 
+/**
+ * RC6 is a special power stage which allows the GPU to enter an very
+ * low-voltage mode when idle, using down to 0V while at this stage.  This
+ * stage is entered automatically when the GPU is idle when RC6 support is
+ * enabled, and as soon as new workload arises GPU wakes up automatically as well.
+ *
+ * There are different RC6 modes available in Intel GPU, which differentiate
+ * among each other with the latency required to enter and leave RC6 and
+ * voltage consumed by the GPU in different states.
+ *
+ * The combination of the following flags define which states GPU is allowed
+ * to enter, while RC6 is the normal RC6 state, RC6p is the deep RC6, and
+ * RC6pp is deepest RC6. Their support by hardware varies according to the
+ * GPU, BIOS, chipset and platform. RC6 is usually the safest one and the one
+ * which brings the most power savings; deeper states save more power, but
+ * require higher latency to switch to and wake up.
+ */
+#define INTEL_RC6_ENABLE			(1<<0)
+#define INTEL_RC6p_ENABLE			(1<<1)
+#define INTEL_RC6pp_ENABLE			(1<<2)
+
 extern struct drm_ioctl_desc i915_ioctls[];
 extern int i915_max_ioctl;
 extern unsigned int i915_fbpercrtc __always_unused;
@@ -1065,7 +1086,7 @@ extern int i915_vbt_sdvo_panel_type __read_mostly;
 extern int i915_enable_rc6 __read_mostly;
 extern int i915_enable_fbc __read_mostly;
 extern bool i915_enable_hangcheck __read_mostly;
-extern bool i915_enable_ppgtt __read_mostly;
+extern int i915_enable_ppgtt __read_mostly;
 
 extern int i915_suspend(struct drm_device *dev, pm_message_t state);
 extern int i915_resume(struct drm_device *dev);

drivers/gpu/drm/i915/i915_gem.c

@@ -1472,16 +1472,19 @@ i915_gem_object_move_to_active(struct drm_i915_gem_object *obj,
 	list_move_tail(&obj->ring_list, &ring->active_list);
 
 	obj->last_rendering_seqno = seqno;
-	if (obj->fenced_gpu_access) {
-		struct drm_i915_fence_reg *reg;
-
-		BUG_ON(obj->fence_reg == I915_FENCE_REG_NONE);
 
+	if (obj->fenced_gpu_access) {
 		obj->last_fenced_seqno = seqno;
 		obj->last_fenced_ring = ring;
 
+		/* Bump MRU to take account of the delayed flush */
+		if (obj->fence_reg != I915_FENCE_REG_NONE) {
+			struct drm_i915_fence_reg *reg;
+
 			reg = &dev_priv->fence_regs[obj->fence_reg];
-		list_move_tail(&reg->lru_list, &dev_priv->mm.fence_list);
+			list_move_tail(&reg->lru_list,
+				       &dev_priv->mm.fence_list);
+		}
 	}
 }
 
@@ -3754,12 +3757,32 @@ void i915_gem_init_ppgtt(struct drm_device *dev)
 	drm_i915_private_t *dev_priv = dev->dev_private;
 	uint32_t pd_offset;
 	struct intel_ring_buffer *ring;
+	struct i915_hw_ppgtt *ppgtt = dev_priv->mm.aliasing_ppgtt;
+	uint32_t __iomem *pd_addr;
+	uint32_t pd_entry;
 	int i;
 
 	if (!dev_priv->mm.aliasing_ppgtt)
 		return;
 
-	pd_offset = dev_priv->mm.aliasing_ppgtt->pd_offset;
+
+	pd_addr = dev_priv->mm.gtt->gtt + ppgtt->pd_offset/sizeof(uint32_t);
+	for (i = 0; i < ppgtt->num_pd_entries; i++) {
+		dma_addr_t pt_addr;
+
+		if (dev_priv->mm.gtt->needs_dmar)
+			pt_addr = ppgtt->pt_dma_addr[i];
+		else
+			pt_addr = page_to_phys(ppgtt->pt_pages[i]);
+
+		pd_entry = GEN6_PDE_ADDR_ENCODE(pt_addr);
+		pd_entry |= GEN6_PDE_VALID;
+
+		writel(pd_entry, pd_addr + i);
+	}
+	readl(pd_addr);
+
+	pd_offset = ppgtt->pd_offset;
 	pd_offset /= 64; /* in cachelines, */
 	pd_offset <<= 16;
 

drivers/gpu/drm/i915/i915_gem_execbuffer.c

@@ -498,8 +498,8 @@ pin_and_fence_object(struct drm_i915_gem_object *obj,
 				if (ret)
 					goto err_unpin;
 			}
+			obj->pending_fenced_gpu_access = true;
 		}
-		obj->pending_fenced_gpu_access = need_fence;
 	}
 
 	entry->offset = obj->gtt_offset;

drivers/gpu/drm/i915/i915_gem_gtt.c

@@ -65,9 +65,7 @@ int i915_gem_init_aliasing_ppgtt(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct i915_hw_ppgtt *ppgtt;
-	uint32_t pd_entry;
 	unsigned first_pd_entry_in_global_pt;
-	uint32_t __iomem *pd_addr;
 	int i;
 	int ret = -ENOMEM;
 
@@ -100,7 +98,6 @@ int i915_gem_init_aliasing_ppgtt(struct drm_device *dev)
 			goto err_pt_alloc;
 	}
 
-	pd_addr = dev_priv->mm.gtt->gtt + first_pd_entry_in_global_pt;
 	for (i = 0; i < ppgtt->num_pd_entries; i++) {
 		dma_addr_t pt_addr;
 		if (dev_priv->mm.gtt->needs_dmar) {
@@ -117,13 +114,7 @@ int i915_gem_init_aliasing_ppgtt(struct drm_device *dev)
 			ppgtt->pt_dma_addr[i] = pt_addr;
 		} else
 			pt_addr = page_to_phys(ppgtt->pt_pages[i]);
-
-		pd_entry = GEN6_PDE_ADDR_ENCODE(pt_addr);
-		pd_entry |= GEN6_PDE_VALID;
-
-		writel(pd_entry, pd_addr + i);
 	}
-	readl(pd_addr);
 
 	ppgtt->scratch_page_dma_addr = dev_priv->mm.gtt->scratch_page_dma;
 

drivers/gpu/drm/i915/i915_reg.h

@@ -2385,6 +2385,7 @@
 #define   PIPECONF_DISABLE	0
 #define   PIPECONF_DOUBLE_WIDE	(1<<30)
 #define   I965_PIPECONF_ACTIVE	(1<<30)
+#define   PIPECONF_FRAME_START_DELAY_MASK (3<<27)
 #define   PIPECONF_SINGLE_WIDE	0
 #define   PIPECONF_PIPE_UNLOCKED 0
 #define   PIPECONF_PIPE_LOCKED	(1<<25)

drivers/gpu/drm/i915/intel_bios.c

@@ -24,6 +24,7 @@
  *    Eric Anholt <eric@anholt.net>
  *
  */
+#include <linux/dmi.h>
 #include <drm/drm_dp_helper.h>
 #include "drmP.h"
 #include "drm.h"
@@ -621,6 +622,26 @@ init_vbt_defaults(struct drm_i915_private *dev_priv)
 	dev_priv->edp.bpp = 18;
 }
 
+static int __init intel_no_opregion_vbt_callback(const struct dmi_system_id *id)
+{
+	DRM_DEBUG_KMS("Falling back to manually reading VBT from "
+		      "VBIOS ROM for %s\n",
+		      id->ident);
+	return 1;
+}
+
+static const struct dmi_system_id intel_no_opregion_vbt[] = {
+	{
+		.callback = intel_no_opregion_vbt_callback,
+		.ident = "ThinkCentre A57",
+		.matches = {
+			DMI_MATCH(DMI_SYS_VENDOR, "LENOVO"),
+			DMI_MATCH(DMI_PRODUCT_NAME, "97027RG"),
+		},
+	},
+	{ }
+};
+
 /**
  * intel_parse_bios - find VBT and initialize settings from the BIOS
  * @dev: DRM device
@@ -641,7 +662,7 @@ intel_parse_bios(struct drm_device *dev)
 	init_vbt_defaults(dev_priv);
 
 	/* XXX Should this validation be moved to intel_opregion.c? */
-	if (dev_priv->opregion.vbt) {
+	if (!dmi_check_system(intel_no_opregion_vbt) && dev_priv->opregion.vbt) {
 		struct vbt_header *vbt = dev_priv->opregion.vbt;
 		if (memcmp(vbt->signature, "$VBT", 4) == 0) {
 			DRM_DEBUG_KMS("Using VBT from OpRegion: %20s\n",

drivers/gpu/drm/i915/intel_display.c

@@ -5539,7 +5539,8 @@ void ironlake_init_pch_refclk(struct drm_device *dev)
 		if (intel_panel_use_ssc(dev_priv) && can_ssc) {
 			DRM_DEBUG_KMS("Using SSC on panel\n");
 			temp |= DREF_SSC1_ENABLE;
-		}
+		} else
+			temp &= ~DREF_SSC1_ENABLE;
 
 		/* Get SSC going before enabling the outputs */
 		I915_WRITE(PCH_DREF_CONTROL, temp);
@@ -7580,6 +7581,12 @@ static void intel_sanitize_modesetting(struct drm_device *dev,
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 reg, val;
 
+	/* Clear any frame start delays used for debugging left by the BIOS */
+	for_each_pipe(pipe) {
+		reg = PIPECONF(pipe);
+		I915_WRITE(reg, I915_READ(reg) & ~PIPECONF_FRAME_START_DELAY_MASK);
+	}
+
 	if (HAS_PCH_SPLIT(dev))
 		return;
 
@@ -8215,7 +8222,7 @@ void intel_init_emon(struct drm_device *dev)
 	dev_priv->corr = (lcfuse & LCFUSE_HIV_MASK);
 }
 
-static bool intel_enable_rc6(struct drm_device *dev)
+static int intel_enable_rc6(struct drm_device *dev)
 {
 	/*
 	 * Respect the kernel parameter if it is set
@@ -8233,11 +8240,11 @@ static bool intel_enable_rc6(struct drm_device *dev)
 	 * Disable rc6 on Sandybridge
 	 */
 	if (INTEL_INFO(dev)->gen == 6) {
-		DRM_DEBUG_DRIVER("Sandybridge: RC6 disabled\n");
-		return 0;
+		DRM_DEBUG_DRIVER("Sandybridge: deep RC6 disabled\n");
+		return INTEL_RC6_ENABLE;
 	}
-	DRM_DEBUG_DRIVER("RC6 enabled\n");
-	return 1;
+	DRM_DEBUG_DRIVER("RC6 and deep RC6 enabled\n");
+	return (INTEL_RC6_ENABLE | INTEL_RC6p_ENABLE);
 }
 
 void gen6_enable_rps(struct drm_i915_private *dev_priv)
@@ -8247,6 +8254,7 @@ void gen6_enable_rps(struct drm_i915_private *dev_priv)
 	u32 pcu_mbox, rc6_mask = 0;
 	u32 gtfifodbg;
 	int cur_freq, min_freq, max_freq;
+	int rc6_mode;
 	int i;
 
 	/* Here begins a magic sequence of register writes to enable
@@ -8284,9 +8292,20 @@ void gen6_enable_rps(struct drm_i915_private *dev_priv)
 	I915_WRITE(GEN6_RC6p_THRESHOLD, 100000);
 	I915_WRITE(GEN6_RC6pp_THRESHOLD, 64000); /* unused */
 
-	if (intel_enable_rc6(dev_priv->dev))
-		rc6_mask = GEN6_RC_CTL_RC6_ENABLE |
-			((IS_GEN7(dev_priv->dev)) ? GEN6_RC_CTL_RC6p_ENABLE : 0);
+	rc6_mode = intel_enable_rc6(dev_priv->dev);
+	if (rc6_mode & INTEL_RC6_ENABLE)
+		rc6_mask |= GEN6_RC_CTL_RC6_ENABLE;
+
+	if (rc6_mode & INTEL_RC6p_ENABLE)
+		rc6_mask |= GEN6_RC_CTL_RC6p_ENABLE;
+
+	if (rc6_mode & INTEL_RC6pp_ENABLE)
+		rc6_mask |= GEN6_RC_CTL_RC6pp_ENABLE;
+
+	DRM_INFO("Enabling RC6 states: RC6 %s, RC6p %s, RC6pp %s\n",
+			(rc6_mode & INTEL_RC6_ENABLE) ? "on" : "off",
+			(rc6_mode & INTEL_RC6p_ENABLE) ? "on" : "off",
+			(rc6_mode & INTEL_RC6pp_ENABLE) ? "on" : "off");
 
 	I915_WRITE(GEN6_RC_CONTROL,
 		   rc6_mask |

drivers/gpu/drm/i915/intel_lvds.c

@@ -755,6 +755,14 @@ static const struct dmi_system_id intel_no_lvds[] = {
 			DMI_MATCH(DMI_BOARD_NAME, "hp st5747"),
 		},
 	},
+	{
+		.callback = intel_no_lvds_dmi_callback,
+		.ident = "MSI Wind Box DC500",
+		.matches = {
+			DMI_MATCH(DMI_BOARD_VENDOR, "MICRO-STAR INTERNATIONAL CO., LTD"),
+			DMI_MATCH(DMI_BOARD_NAME, "MS-7469"),
+		},
+	},
 
 	{ }	/* terminating entry */
 };

drivers/gpu/drm/i915/intel_ringbuffer.c

@@ -626,7 +626,7 @@ gen6_ring_get_seqno(struct intel_ring_buffer *ring)
 	/* Workaround to force correct ordering between irq and seqno writes on
 	 * ivb (and maybe also on snb) by reading from a CS register (like
 	 * ACTHD) before reading the status page. */
-	if (IS_GEN7(dev))
+	if (IS_GEN6(dev) || IS_GEN7(dev))
 		intel_ring_get_active_head(ring);
 	return intel_read_status_page(ring, I915_GEM_HWS_INDEX);
 }

drivers/gpu/drm/i915/intel_sprite.c

@@ -411,6 +411,9 @@ intel_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 
 	old_obj = intel_plane->obj;
 
+	src_w = src_w >> 16;
+	src_h = src_h >> 16;
+
 	/* Pipe must be running... */
 	if (!(I915_READ(PIPECONF(pipe)) & PIPECONF_ENABLE))
 		return -EINVAL;

include/drm/intel-gtt.h

@@ -44,4 +44,8 @@ void intel_gtt_insert_pages(unsigned int first_entry, unsigned int num_entries,
 /* flag for GFDT type */
 #define AGP_USER_CACHED_MEMORY_GFDT (1 << 3)
 
+#ifdef CONFIG_INTEL_IOMMU
+extern int intel_iommu_gfx_mapped;
+#endif
+
 #endif