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

Pull drm fixes from Dave Airlie:
 "Xmas fixes pull:

  core:
      one atomic fix, revert the WARN_ON dumb buffers patch.

  agp:
      fixup Dave J.

  nouveau:
      fix 3.18 regression for old userspace

  tegra fixes:
      vblank and iommu fixes

  amdkfd:
      fix bugs shown by testing with userspace, init apertures once

  msm:
      hdmi fixes and cleanup

  i915:
      misc fixes

  There is also a link ordering fix that I've asked to be cc'ed to you,
  putting iommu before gpu, it fixes an issue with amdkfd when things
  are all in the kernel, but I didn't like sending it via my tree
  without discussion.

  I'll probably be a bit on/off for a few weeks with pulls now, due to
  holidays and LCA, so don't be surprised if stuff gets a bit backed up,
  and things end up a bit large due to lag"

* 'drm-fixes' of git://people.freedesktop.org/~airlied/linux: (28 commits)
  Revert "drm/gem: Warn on illegal use of the dumb buffer interface v2"
  agp: Fix up email address & attributions in AGP MODULE_AUTHOR tags
  nouveau: bring back legacy mmap handler
  drm/msm/hdmi: rework HDMI IRQ handler
  drm/msm/hdmi: enable regulators before clocks to avoid warnings
  drm/msm/mdp5: update irqs on crtc<->encoder link change
  drm/msm: block incoming update on pending updates
  drm/atomic: fix potential null ptr on plane enable
  drm/msm: Deletion of unnecessary checks before the function call "release_firmware"
  drm/msm: Deletion of unnecessary checks before two function calls
  drm/tegra: dc: Select root window for event dispatch
  drm/tegra: gem: Use the proper size for GEM objects
  drm/tegra: gem: Flush buffer objects upon allocation
  drm/tegra: dc: Fix a potential race on page-flip completion
  drm/tegra: dc: Consistently use the same pipe
  drm/irq: Add drm_crtc_vblank_count()
  drm/irq: Add drm_crtc_handle_vblank()
  drm/irq: Add drm_crtc_send_vblank_event()
  drm/i915: Disable PSMI sleep messages on all rings around context switches
  drm/i915: Force the CS stall for invalidate flushes
  ...

47 files changed, 450 insertions, 221 deletions

diff --git a/drivers/char/agp/ali-agp.c b/drivers/char/agp/ali-agp.c
index 19db03667650..dcbbb4ea3cc1 100644
--- a/drivers/char/agp/ali-agp.c
+++ b/drivers/char/agp/ali-agp.c
@@ -417,6 +417,6 @@ static void __exit agp_ali_cleanup(void)
 module_init(agp_ali_init);
 module_exit(agp_ali_cleanup);
 
-MODULE_AUTHOR("Dave Jones <davej@redhat.com>");
+MODULE_AUTHOR("Dave Jones");
 MODULE_LICENSE("GPL and additional rights");
 
diff --git a/drivers/char/agp/amd64-agp.c b/drivers/char/agp/amd64-agp.c
index 3b47ed0310e1..0ef350010766 100644
--- a/drivers/char/agp/amd64-agp.c
+++ b/drivers/char/agp/amd64-agp.c
@@ -813,6 +813,6 @@ static void __exit agp_amd64_cleanup(void)
 module_init(agp_amd64_mod_init);
 module_exit(agp_amd64_cleanup);
 
-MODULE_AUTHOR("Dave Jones <davej@redhat.com>, Andi Kleen");
+MODULE_AUTHOR("Dave Jones, Andi Kleen");
 module_param(agp_try_unsupported, bool, 0);
 MODULE_LICENSE("GPL");
diff --git a/drivers/char/agp/ati-agp.c b/drivers/char/agp/ati-agp.c
index 18a7a6baa304..75a9786a77e6 100644
--- a/drivers/char/agp/ati-agp.c
+++ b/drivers/char/agp/ati-agp.c
@@ -579,6 +579,6 @@ static void __exit agp_ati_cleanup(void)
 module_init(agp_ati_init);
 module_exit(agp_ati_cleanup);
 
-MODULE_AUTHOR("Dave Jones <davej@redhat.com>");
+MODULE_AUTHOR("Dave Jones");
 MODULE_LICENSE("GPL and additional rights");
 
diff --git a/drivers/char/agp/backend.c b/drivers/char/agp/backend.c
index 317c28ce8328..38ffb281df97 100644
--- a/drivers/char/agp/backend.c
+++ b/drivers/char/agp/backend.c
@@ -356,7 +356,7 @@ static __init int agp_setup(char *s)
 __setup("agp=", agp_setup);
 #endif
 
-MODULE_AUTHOR("Dave Jones <davej@redhat.com>");
+MODULE_AUTHOR("Dave Jones, Jeff Hartmann");
 MODULE_DESCRIPTION("AGP GART driver");
 MODULE_LICENSE("GPL and additional rights");
 MODULE_ALIAS_MISCDEV(AGPGART_MINOR);
diff --git a/drivers/char/agp/intel-agp.c b/drivers/char/agp/intel-agp.c
index f9b9ca5d31b7..0a21daed5b62 100644
--- a/drivers/char/agp/intel-agp.c
+++ b/drivers/char/agp/intel-agp.c
@@ -920,5 +920,5 @@ static void __exit agp_intel_cleanup(void)
 module_init(agp_intel_init);
 module_exit(agp_intel_cleanup);
 
-MODULE_AUTHOR("Dave Jones <davej@redhat.com>");
+MODULE_AUTHOR("Dave Jones, Various @Intel");
 MODULE_LICENSE("GPL and additional rights");
diff --git a/drivers/char/agp/intel-gtt.c b/drivers/char/agp/intel-gtt.c
index f3334829e55a..92aa43fa8d70 100644
--- a/drivers/char/agp/intel-gtt.c
+++ b/drivers/char/agp/intel-gtt.c
@@ -1438,5 +1438,5 @@ void intel_gmch_remove(void)
 }
 EXPORT_SYMBOL(intel_gmch_remove);
 
-MODULE_AUTHOR("Dave Jones <davej@redhat.com>");
+MODULE_AUTHOR("Dave Jones, Various @Intel");
 MODULE_LICENSE("GPL and additional rights");
diff --git a/drivers/char/agp/nvidia-agp.c b/drivers/char/agp/nvidia-agp.c
index a1861b75eb31..6c8d39cb566e 100644
--- a/drivers/char/agp/nvidia-agp.c
+++ b/drivers/char/agp/nvidia-agp.c
@@ -1,7 +1,7 @@
 /*
  * Nvidia AGPGART routines.
  * Based upon a 2.4 agpgart diff by the folks from NVIDIA, and hacked up
- * to work in 2.5 by Dave Jones <davej@redhat.com>
+ * to work in 2.5 by Dave Jones.
  */
 
 #include <linux/module.h>
diff --git a/drivers/char/agp/via-agp.c b/drivers/char/agp/via-agp.c
index 228f20cddc05..a4961d35e940 100644
--- a/drivers/char/agp/via-agp.c
+++ b/drivers/char/agp/via-agp.c
@@ -595,4 +595,4 @@ module_init(agp_via_init);
 module_exit(agp_via_cleanup);
 
 MODULE_LICENSE("GPL");
-MODULE_AUTHOR("Dave Jones <davej@redhat.com>");
+MODULE_AUTHOR("Dave Jones");
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c
index 4f7b275f2f7b..7d4974b83af7 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c
@@ -121,13 +121,9 @@ static int kfd_open(struct inode *inode, struct file *filep)
 	if (IS_ERR(process))
 		return PTR_ERR(process);
 
-	process->is_32bit_user_mode = is_32bit_user_mode;
-
 	dev_dbg(kfd_device, "process %d opened, compat mode (32 bit) - %d\n",
 		process->pasid, process->is_32bit_user_mode);
 
-	kfd_init_apertures(process);
-
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_flat_memory.c b/drivers/gpu/drm/amd/amdkfd/kfd_flat_memory.c
index 66df4da01c29..e64aa99e5e41 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_flat_memory.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_flat_memory.c
@@ -299,13 +299,13 @@ int kfd_init_apertures(struct kfd_process *process)
 	struct kfd_dev *dev;
 	struct kfd_process_device *pdd;
 
-	mutex_lock(&process->mutex);
-
 	/*Iterating over all devices*/
 	while ((dev = kfd_topology_enum_kfd_devices(id)) != NULL &&
 		id < NUM_OF_SUPPORTED_GPUS) {
 
 		pdd = kfd_get_process_device_data(dev, process, 1);
+		if (!pdd)
+			return -1;
 
 		/*
 		 * For 64 bit process aperture will be statically reserved in
@@ -348,8 +348,6 @@ int kfd_init_apertures(struct kfd_process *process)
 		id++;
 	}
 
-	mutex_unlock(&process->mutex);
-
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_process.c b/drivers/gpu/drm/amd/amdkfd/kfd_process.c
index b85eb0b830b4..3c76ef05cbcf 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_process.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_process.c
@@ -26,6 +26,8 @@
 #include <linux/slab.h>
 #include <linux/amd-iommu.h>
 #include <linux/notifier.h>
+#include <linux/compat.h>
+
 struct mm_struct;
 
 #include "kfd_priv.h"
@@ -285,8 +287,15 @@ static struct kfd_process *create_process(const struct task_struct *thread)
 	if (err != 0)
 		goto err_process_pqm_init;
 
+	/* init process apertures*/
+	process->is_32bit_user_mode = is_compat_task();
+	if (kfd_init_apertures(process) != 0)
+		goto err_init_apretures;
+
 	return process;
 
+err_init_apretures:
+	pqm_uninit(&process->pqm);
 err_process_pqm_init:
 	hash_del_rcu(&process->kfd_processes);
 	synchronize_rcu();
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_topology.c b/drivers/gpu/drm/amd/amdkfd/kfd_topology.c
index 5733e2859e8a..b11792d7e70e 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_topology.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_topology.c
@@ -700,8 +700,6 @@ static ssize_t node_show(struct kobject *kobj, struct attribute *attr,
 				dev->node_props.simd_per_cu);
 		sysfs_show_32bit_prop(buffer, "max_slots_scratch_cu",
 				dev->node_props.max_slots_scratch_cu);
-		sysfs_show_32bit_prop(buffer, "engine_id",
-				dev->node_props.engine_id);
 		sysfs_show_32bit_prop(buffer, "vendor_id",
 				dev->node_props.vendor_id);
 		sysfs_show_32bit_prop(buffer, "device_id",
@@ -715,6 +713,12 @@ static ssize_t node_show(struct kobject *kobj, struct attribute *attr,
 						dev->gpu->kgd));
 			sysfs_show_64bit_prop(buffer, "local_mem_size",
 					kfd2kgd->get_vmem_size(dev->gpu->kgd));
+
+			sysfs_show_32bit_prop(buffer, "fw_version",
+					kfd2kgd->get_fw_version(
+							dev->gpu->kgd,
+							KGD_ENGINE_MEC1));
+
 		}
 
 		ret = sysfs_show_32bit_prop(buffer, "max_engine_clk_ccompute",
diff --git a/drivers/gpu/drm/amd/include/kgd_kfd_interface.h b/drivers/gpu/drm/amd/include/kgd_kfd_interface.h
index 9c729dd8dd50..47b551970a14 100644
--- a/drivers/gpu/drm/amd/include/kgd_kfd_interface.h
+++ b/drivers/gpu/drm/amd/include/kgd_kfd_interface.h
@@ -45,6 +45,17 @@ enum kgd_memory_pool {
 	KGD_POOL_FRAMEBUFFER = 3,
 };
 
+enum kgd_engine_type {
+	KGD_ENGINE_PFP = 1,
+	KGD_ENGINE_ME,
+	KGD_ENGINE_CE,
+	KGD_ENGINE_MEC1,
+	KGD_ENGINE_MEC2,
+	KGD_ENGINE_RLC,
+	KGD_ENGINE_SDMA,
+	KGD_ENGINE_MAX
+};
+
 struct kgd2kfd_shared_resources {
 	/* Bit n == 1 means VMID n is available for KFD. */
 	unsigned int compute_vmid_bitmap;
@@ -137,6 +148,8 @@ struct kgd2kfd_calls {
  *
  * @hqd_destroy: Destructs and preempts the queue assigned to that hqd slot.
  *
+ * @get_fw_version: Returns FW versions from the header
+ *
  * This structure contains function pointers to services that the kgd driver
  * provides to amdkfd driver.
  *
@@ -176,6 +189,8 @@ struct kfd2kgd_calls {
 	int (*hqd_destroy)(struct kgd_dev *kgd, uint32_t reset_type,
 				unsigned int timeout, uint32_t pipe_id,
 				uint32_t queue_id);
+	uint16_t (*get_fw_version)(struct kgd_dev *kgd,
+				enum kgd_engine_type type);
 };
 
 bool kgd2kfd_init(unsigned interface_version,
diff --git a/drivers/gpu/drm/drm_atomic_helper.c b/drivers/gpu/drm/drm_atomic_helper.c
index 4a78a773151c..bbdbe4721573 100644
--- a/drivers/gpu/drm/drm_atomic_helper.c
+++ b/drivers/gpu/drm/drm_atomic_helper.c
@@ -61,7 +61,7 @@ drm_atomic_helper_plane_changed(struct drm_atomic_state *state,
 	struct drm_crtc_state *crtc_state;
 
 	if (plane->state->crtc) {
-		crtc_state = state->crtc_states[drm_crtc_index(plane->crtc)];
+		crtc_state = state->crtc_states[drm_crtc_index(plane->state->crtc)];
 
 		if (WARN_ON(!crtc_state))
 			return;
diff --git a/drivers/gpu/drm/drm_irq.c b/drivers/gpu/drm/drm_irq.c
index f5a5f18efa5b..4d79dad9d44f 100644
--- a/drivers/gpu/drm/drm_irq.c
+++ b/drivers/gpu/drm/drm_irq.c
@@ -830,6 +830,8 @@ drm_get_last_vbltimestamp(struct drm_device *dev, int crtc,
  * vblank events since the system was booted, including lost events due to
  * modesetting activity.
  *
+ * This is the legacy version of drm_crtc_vblank_count().
+ *
  * Returns:
  * The software vblank counter.
  */
@@ -844,6 +846,25 @@ u32 drm_vblank_count(struct drm_device *dev, int crtc)
 EXPORT_SYMBOL(drm_vblank_count);
 
 /**
+ * drm_crtc_vblank_count - retrieve "cooked" vblank counter value
+ * @crtc: which counter to retrieve
+ *
+ * Fetches the "cooked" vblank count value that represents the number of
+ * vblank events since the system was booted, including lost events due to
+ * modesetting activity.
+ *
+ * This is the native KMS version of drm_vblank_count().
+ *
+ * Returns:
+ * The software vblank counter.
+ */
+u32 drm_crtc_vblank_count(struct drm_crtc *crtc)
+{
+	return drm_vblank_count(crtc->dev, drm_crtc_index(crtc));
+}
+EXPORT_SYMBOL(drm_crtc_vblank_count);
+
+/**
  * drm_vblank_count_and_time - retrieve "cooked" vblank counter value
  * and the system timestamp corresponding to that vblank counter value.
  *
@@ -904,6 +925,8 @@ static void send_vblank_event(struct drm_device *dev,
  *
  * Updates sequence # and timestamp on event, and sends it to userspace.
  * Caller must hold event lock.
+ *
+ * This is the legacy version of drm_crtc_send_vblank_event().
  */
 void drm_send_vblank_event(struct drm_device *dev, int crtc,
 		struct drm_pending_vblank_event *e)
@@ -923,6 +946,23 @@ void drm_send_vblank_event(struct drm_device *dev, int crtc,
 EXPORT_SYMBOL(drm_send_vblank_event);
 
 /**
+ * drm_crtc_send_vblank_event - helper to send vblank event after pageflip
+ * @crtc: the source CRTC of the vblank event
+ * @e: the event to send
+ *
+ * Updates sequence # and timestamp on event, and sends it to userspace.
+ * Caller must hold event lock.
+ *
+ * This is the native KMS version of drm_send_vblank_event().
+ */
+void drm_crtc_send_vblank_event(struct drm_crtc *crtc,
+				struct drm_pending_vblank_event *e)
+{
+	drm_send_vblank_event(crtc->dev, drm_crtc_index(crtc), e);
+}
+EXPORT_SYMBOL(drm_crtc_send_vblank_event);
+
+/**
  * drm_vblank_enable - enable the vblank interrupt on a CRTC
  * @dev: DRM device
  * @crtc: CRTC in question
@@ -1594,6 +1634,8 @@ static void drm_handle_vblank_events(struct drm_device *dev, int crtc)
  *
  * Drivers should call this routine in their vblank interrupt handlers to
  * update the vblank counter and send any signals that may be pending.
+ *
+ * This is the legacy version of drm_crtc_handle_vblank().
  */
 bool drm_handle_vblank(struct drm_device *dev, int crtc)
 {
@@ -1670,3 +1712,21 @@ bool drm_handle_vblank(struct drm_device *dev, int crtc)
 	return true;
 }
 EXPORT_SYMBOL(drm_handle_vblank);
+
+/**
+ * drm_crtc_handle_vblank - handle a vblank event
+ * @crtc: where this event occurred
+ *
+ * Drivers should call this routine in their vblank interrupt handlers to
+ * update the vblank counter and send any signals that may be pending.
+ *
+ * This is the native KMS version of drm_handle_vblank().
+ *
+ * Returns:
+ * True if the event was successfully handled, false on failure.
+ */
+bool drm_crtc_handle_vblank(struct drm_crtc *crtc)
+{
+	return drm_handle_vblank(crtc->dev, drm_crtc_index(crtc));
+}
+EXPORT_SYMBOL(drm_crtc_handle_vblank);
diff --git a/drivers/gpu/drm/i915/i915_drv.c b/drivers/gpu/drm/i915/i915_drv.c
index f990ab4c3efb..574057cd1d09 100644
--- a/drivers/gpu/drm/i915/i915_drv.c
+++ b/drivers/gpu/drm/i915/i915_drv.c
@@ -811,6 +811,8 @@ int i915_reset(struct drm_device *dev)
 	if (!i915.reset)
 		return 0;
 
+	intel_reset_gt_powersave(dev);
+
 	mutex_lock(&dev->struct_mutex);
 
 	i915_gem_reset(dev);
@@ -880,7 +882,7 @@ int i915_reset(struct drm_device *dev)
 		 * of re-init after reset.
 		 */
 		if (INTEL_INFO(dev)->gen > 5)
-			intel_reset_gt_powersave(dev);
+			intel_enable_gt_powersave(dev);
 	} else {
 		mutex_unlock(&dev->struct_mutex);
 	}
@@ -1584,7 +1586,7 @@ static struct drm_driver driver = {
 	.gem_prime_import = i915_gem_prime_import,
 
 	.dumb_create = i915_gem_dumb_create,
-	.dumb_map_offset = i915_gem_dumb_map_offset,
+	.dumb_map_offset = i915_gem_mmap_gtt,
 	.dumb_destroy = drm_gem_dumb_destroy,
 	.ioctls = i915_ioctls,
 	.fops = &i915_driver_fops,
diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index 63bcda5541ec..70d0f0f06f1a 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -2501,9 +2501,8 @@ void i915_vma_move_to_active(struct i915_vma *vma,
 int i915_gem_dumb_create(struct drm_file *file_priv,
 			 struct drm_device *dev,
 			 struct drm_mode_create_dumb *args);
-int i915_gem_dumb_map_offset(struct drm_file *file_priv,
-			     struct drm_device *dev, uint32_t handle,
-			     uint64_t *offset);
+int i915_gem_mmap_gtt(struct drm_file *file_priv, struct drm_device *dev,
+		      uint32_t handle, uint64_t *offset);
 /**
  * Returns true if seq1 is later than seq2.
  */
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index 4a9faea626db..52adcb680be3 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -401,7 +401,6 @@ static int
 i915_gem_create(struct drm_file *file,
 		struct drm_device *dev,
 		uint64_t size,
-		bool dumb,
 		uint32_t *handle_p)
 {
 	struct drm_i915_gem_object *obj;
@@ -417,7 +416,6 @@ i915_gem_create(struct drm_file *file,
 	if (obj == NULL)
 		return -ENOMEM;
 
-	obj->base.dumb = dumb;
 	ret = drm_gem_handle_create(file, &obj->base, &handle);
 	/* drop reference from allocate - handle holds it now */
 	drm_gem_object_unreference_unlocked(&obj->base);
@@ -437,7 +435,7 @@ i915_gem_dumb_create(struct drm_file *file,
 	args->pitch = ALIGN(args->width * DIV_ROUND_UP(args->bpp, 8), 64);
 	args->size = args->pitch * args->height;
 	return i915_gem_create(file, dev,
-			       args->size, true, &args->handle);
+			       args->size, &args->handle);
 }
 
 /**
@@ -450,7 +448,7 @@ i915_gem_create_ioctl(struct drm_device *dev, void *data,
 	struct drm_i915_gem_create *args = data;
 
 	return i915_gem_create(file, dev,
-			       args->size, false, &args->handle);
+			       args->size, &args->handle);
 }
 
 static inline int
@@ -1840,10 +1838,10 @@ static void i915_gem_object_free_mmap_offset(struct drm_i915_gem_object *obj)
 	drm_gem_free_mmap_offset(&obj->base);
 }
 
-static int
+int
 i915_gem_mmap_gtt(struct drm_file *file,
 		  struct drm_device *dev,
-		  uint32_t handle, bool dumb,
+		  uint32_t handle,
 		  uint64_t *offset)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -1860,13 +1858,6 @@ i915_gem_mmap_gtt(struct drm_file *file,
 		goto unlock;
 	}
 
-	/*
-	 * We don't allow dumb mmaps on objects created using another
-	 * interface.
-	 */
-	WARN_ONCE(dumb && !(obj->base.dumb || obj->base.import_attach),
-		  "Illegal dumb map of accelerated buffer.\n");
-
 	if (obj->base.size > dev_priv->gtt.mappable_end) {
 		ret = -E2BIG;
 		goto out;
@@ -1891,15 +1882,6 @@ unlock:
 	return ret;
 }
 
-int
-i915_gem_dumb_map_offset(struct drm_file *file,
-			 struct drm_device *dev,
-			 uint32_t handle,
-			 uint64_t *offset)
-{
-	return i915_gem_mmap_gtt(file, dev, handle, true, offset);
-}
-
 /**
  * i915_gem_mmap_gtt_ioctl - prepare an object for GTT mmap'ing
  * @dev: DRM device
@@ -1921,7 +1903,7 @@ i915_gem_mmap_gtt_ioctl(struct drm_device *dev, void *data,
 {
 	struct drm_i915_gem_mmap_gtt *args = data;
 
-	return i915_gem_mmap_gtt(file, dev, args->handle, false, &args->offset);
+	return i915_gem_mmap_gtt(file, dev, args->handle, &args->offset);
 }
 
 static inline int
diff --git a/drivers/gpu/drm/i915/i915_gem_context.c b/drivers/gpu/drm/i915/i915_gem_context.c
index d17ff435f276..d011ec82ef1e 100644
--- a/drivers/gpu/drm/i915/i915_gem_context.c
+++ b/drivers/gpu/drm/i915/i915_gem_context.c
@@ -473,7 +473,12 @@ mi_set_context(struct intel_engine_cs *ring,
 	       u32 hw_flags)
 {
 	u32 flags = hw_flags | MI_MM_SPACE_GTT;
-	int ret;
+	const int num_rings =
+		/* Use an extended w/a on ivb+ if signalling from other rings */
+		i915_semaphore_is_enabled(ring->dev) ?
+		hweight32(INTEL_INFO(ring->dev)->ring_mask) - 1 :
+		0;
+	int len, i, ret;
 
 	/* w/a: If Flush TLB Invalidation Mode is enabled, driver must do a TLB
 	 * invalidation prior to MI_SET_CONTEXT. On GEN6 we don't set the value
@@ -490,15 +495,31 @@ mi_set_context(struct intel_engine_cs *ring,
 	if (!IS_HASWELL(ring->dev) && INTEL_INFO(ring->dev)->gen < 8)
 		flags |= (MI_SAVE_EXT_STATE_EN | MI_RESTORE_EXT_STATE_EN);
 
-	ret = intel_ring_begin(ring, 6);
+
+	len = 4;
+	if (INTEL_INFO(ring->dev)->gen >= 7)
+		len += 2 + (num_rings ? 4*num_rings + 2 : 0);
+
+	ret = intel_ring_begin(ring, len);
 	if (ret)
 		return ret;
 
 	/* WaProgramMiArbOnOffAroundMiSetContext:ivb,vlv,hsw,bdw,chv */
-	if (INTEL_INFO(ring->dev)->gen >= 7)
+	if (INTEL_INFO(ring->dev)->gen >= 7) {
 		intel_ring_emit(ring, MI_ARB_ON_OFF | MI_ARB_DISABLE);
-	else
-		intel_ring_emit(ring, MI_NOOP);
+		if (num_rings) {
+			struct intel_engine_cs *signaller;
+
+			intel_ring_emit(ring, MI_LOAD_REGISTER_IMM(num_rings));
+			for_each_ring(signaller, to_i915(ring->dev), i) {
+				if (signaller == ring)
+					continue;
+
+				intel_ring_emit(ring, RING_PSMI_CTL(signaller->mmio_base));
+				intel_ring_emit(ring, _MASKED_BIT_ENABLE(GEN6_PSMI_SLEEP_MSG_DISABLE));
+			}
+		}
+	}
 
 	intel_ring_emit(ring, MI_NOOP);
 	intel_ring_emit(ring, MI_SET_CONTEXT);
@@ -510,10 +531,21 @@ mi_set_context(struct intel_engine_cs *ring,
 	 */
 	intel_ring_emit(ring, MI_NOOP);
 
-	if (INTEL_INFO(ring->dev)->gen >= 7)
+	if (INTEL_INFO(ring->dev)->gen >= 7) {
+		if (num_rings) {
+			struct intel_engine_cs *sig