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

Pull drm fixes from Dave Airlie:
 - some small fixes for msm and exynos
 - a regression revert affecting nouveau users with old userspace
 - intel pageflip deadlock and gpu hang fixes, hsw modesetting hangs

* 'drm-fixes' of git://people.freedesktop.org/~airlied/linux: (22 commits)
  Revert "drm: mark context support as a legacy subsystem"
  drm/i915: Don't enable the cursor on a disable pipe
  drm/i915: do not update cursor in crtc mode set
  drm/exynos: fix return value check in lowlevel_buffer_allocate()
  drm/exynos: Fix address space warnings in exynos_drm_fbdev.c
  drm/exynos: Fix address space warning in exynos_drm_buf.c
  drm/exynos: Remove redundant OF dependency
  drm/msm: drop unnecessary set_need_resched()
  drm/i915: kill set_need_resched
  drm/msm: fix potential NULL pointer dereference
  drm/i915/dvo: set crtc timings again for panel fixed modes
  drm/i915/sdvo: Robustify the dtd<->drm_mode conversions
  drm/msm: workaround for missing irq
  drm/msm: return -EBUSY if bo still active
  drm/msm: fix return value check in ERR_PTR()
  drm/msm: fix cmdstream size check
  drm/msm: hangcheck harder
  drm/msm: handle read vs write fences
  drm/i915/sdvo: Fully translate sync flags in the dtd->mode conversion
  drm/i915: Use proper print format for debug prints
  ...

23 files changed, 276 insertions, 197 deletions

diff --git a/drivers/gpu/drm/drm_context.c b/drivers/gpu/drm/drm_context.c
index b4fb86d89850..224ff965bcf7 100644
--- a/drivers/gpu/drm/drm_context.c
+++ b/drivers/gpu/drm/drm_context.c
@@ -42,6 +42,10 @@
 
 #include <drm/drmP.h>
 
+/******************************************************************/
+/** \name Context bitmap support */
+/*@{*/
+
 /**
  * Free a handle from the context bitmap.
  *
@@ -52,48 +56,13 @@
  * in drm_device::ctx_idr, while holding the drm_device::struct_mutex
  * lock.
  */
-static void drm_ctxbitmap_free(struct drm_device * dev, int ctx_handle)
+void drm_ctxbitmap_free(struct drm_device * dev, int ctx_handle)
 {
-        if (drm_core_check_feature(dev, DRIVER_MODESET))
-                return;
-
         mutex_lock(&dev->struct_mutex);
         idr_remove(&dev->ctx_idr, ctx_handle);
         mutex_unlock(&dev->struct_mutex);
 }
 
-/******************************************************************/
-/** \name Context bitmap support */
-/*@{*/
-
-void drm_legacy_ctxbitmap_release(struct drm_device *dev,
-                                  struct drm_file *file_priv)
-{
-        if (drm_core_check_feature(dev, DRIVER_MODESET))
-                return;
-
-        mutex_lock(&dev->ctxlist_mutex);
-        if (!list_empty(&dev->ctxlist)) {
-                struct drm_ctx_list *pos, *n;
-
-                list_for_each_entry_safe(pos, n, &dev->ctxlist, head) {
-                        if (pos->tag == file_priv &&
-                            pos->handle != DRM_KERNEL_CONTEXT) {
-                                if (dev->driver->context_dtor)
-                                        dev->driver->context_dtor(dev,
-                                                                  pos->handle);
-
-                                drm_ctxbitmap_free(dev, pos->handle);
-
-                                list_del(&pos->head);
-                                kfree(pos);
-                                --dev->ctx_count;
-                        }
-                }
-        }
-        mutex_unlock(&dev->ctxlist_mutex);
-}
-
 /**
  * Context bitmap allocation.
  *
@@ -121,12 +90,10 @@ static int drm_ctxbitmap_next(struct drm_device * dev)
  *
  * Initialise the drm_device::ctx_idr
  */
-void drm_legacy_ctxbitmap_init(struct drm_device * dev)
+int drm_ctxbitmap_init(struct drm_device * dev)
 {
-        if (drm_core_check_feature(dev, DRIVER_MODESET))
-                return;
-
         idr_init(&dev->ctx_idr);
+        return 0;
 }
 
 /**
@@ -137,7 +104,7 @@ void drm_legacy_ctxbitmap_init(struct drm_device * dev)
  * Free all idr members using drm_ctx_sarea_free helper function
  * while holding the drm_device::struct_mutex lock.
  */
-void drm_legacy_ctxbitmap_cleanup(struct drm_device * dev)
+void drm_ctxbitmap_cleanup(struct drm_device * dev)
 {
         mutex_lock(&dev->struct_mutex);
         idr_destroy(&dev->ctx_idr);
@@ -169,9 +136,6 @@ int drm_getsareactx(struct drm_device *dev, void *data,
         struct drm_local_map *map;
         struct drm_map_list *_entry;
 
-        if (drm_core_check_feature(dev, DRIVER_MODESET))
-                return -EINVAL;
-
         mutex_lock(&dev->struct_mutex);
 
         map = idr_find(&dev->ctx_idr, request->ctx_id);
@@ -216,9 +180,6 @@ int drm_setsareactx(struct drm_device *dev, void *data,
         struct drm_local_map *map = NULL;
         struct drm_map_list *r_list = NULL;
 
-        if (drm_core_check_feature(dev, DRIVER_MODESET))
-                return -EINVAL;
-
         mutex_lock(&dev->struct_mutex);
         list_for_each_entry(r_list, &dev->maplist, head) {
                 if (r_list->map
@@ -319,9 +280,6 @@ int drm_resctx(struct drm_device *dev, void *data,
         struct drm_ctx ctx;
         int i;
 
-        if (drm_core_check_feature(dev, DRIVER_MODESET))
-                return -EINVAL;
-
         if (res->count >= DRM_RESERVED_CONTEXTS) {
                 memset(&ctx, 0, sizeof(ctx));
                 for (i = 0; i < DRM_RESERVED_CONTEXTS; i++) {
@@ -352,9 +310,6 @@ int drm_addctx(struct drm_device *dev, void *data,
         struct drm_ctx_list *ctx_entry;
         struct drm_ctx *ctx = data;
 
-        if (drm_core_check_feature(dev, DRIVER_MODESET))
-                return -EINVAL;
-
         ctx->handle = drm_ctxbitmap_next(dev);
         if (ctx->handle == DRM_KERNEL_CONTEXT) {
                 /* Skip kernel's context and get a new one. */
@@ -398,9 +353,6 @@ int drm_getctx(struct drm_device *dev, void *data, struct drm_file *file_priv)
 {
         struct drm_ctx *ctx = data;
 
-        if (drm_core_check_feature(dev, DRIVER_MODESET))
-                return -EINVAL;
-
         /* This is 0, because we don't handle any context flags */
         ctx->flags = 0;
 
@@ -423,9 +375,6 @@ int drm_switchctx(struct drm_device *dev, void *data,
 {
         struct drm_ctx *ctx = data;
 
-        if (drm_core_check_feature(dev, DRIVER_MODESET))
-                return -EINVAL;
-
         DRM_DEBUG("%d\n", ctx->handle);
         return drm_context_switch(dev, dev->last_context, ctx->handle);
 }
@@ -446,9 +395,6 @@ int drm_newctx(struct drm_device *dev, void *data,
 {
         struct drm_ctx *ctx = data;
 
-        if (drm_core_check_feature(dev, DRIVER_MODESET))
-                return -EINVAL;
-
         DRM_DEBUG("%d\n", ctx->handle);
         drm_context_switch_complete(dev, file_priv, ctx->handle);
 
@@ -471,9 +417,6 @@ int drm_rmctx(struct drm_device *dev, void *data,
 {
         struct drm_ctx *ctx = data;
 
-        if (drm_core_check_feature(dev, DRIVER_MODESET))
-                return -EINVAL;
-
         DRM_DEBUG("%d\n", ctx->handle);
         if (ctx->handle != DRM_KERNEL_CONTEXT) {
                 if (dev->driver->context_dtor)
diff --git a/drivers/gpu/drm/drm_fops.c b/drivers/gpu/drm/drm_fops.c
index 4be8e09a32ef..3f84277d7036 100644
--- a/drivers/gpu/drm/drm_fops.c
+++ b/drivers/gpu/drm/drm_fops.c
@@ -439,7 +439,26 @@ int drm_release(struct inode *inode, struct file *filp)
         if (dev->driver->driver_features & DRIVER_GEM)
                 drm_gem_release(dev, file_priv);
 
-        drm_legacy_ctxbitmap_release(dev, file_priv);
+        mutex_lock(&dev->ctxlist_mutex);
+        if (!list_empty(&dev->ctxlist)) {
+                struct drm_ctx_list *pos, *n;
+
+                list_for_each_entry_safe(pos, n, &dev->ctxlist, head) {
+                        if (pos->tag == file_priv &&
+                            pos->handle != DRM_KERNEL_CONTEXT) {
+                                if (dev->driver->context_dtor)
+                                        dev->driver->context_dtor(dev,
+                                                                  pos->handle);
+
+                                drm_ctxbitmap_free(dev, pos->handle);
+
+                                list_del(&pos->head);
+                                kfree(pos);
+                                --dev->ctx_count;
+                        }
+                }
+        }
+        mutex_unlock(&dev->ctxlist_mutex);
 
         mutex_lock(&dev->struct_mutex);
 
diff --git a/drivers/gpu/drm/drm_stub.c b/drivers/gpu/drm/drm_stub.c
index e7eb0276f7f1..39d864576be4 100644
--- a/drivers/gpu/drm/drm_stub.c
+++ b/drivers/gpu/drm/drm_stub.c
@@ -292,7 +292,13 @@ int drm_fill_in_dev(struct drm_device *dev,
                         goto error_out_unreg;
         }
 
-        drm_legacy_ctxbitmap_init(dev);
+
+
+        retcode = drm_ctxbitmap_init(dev);
+        if (retcode) {
+                DRM_ERROR("Cannot allocate memory for context bitmap.\n");
+                goto error_out_unreg;
+        }
 
         if (driver->driver_features & DRIVER_GEM) {
                 retcode = drm_gem_init(dev);
@@ -446,7 +452,7 @@ void drm_put_dev(struct drm_device *dev)
                 drm_rmmap(dev, r_list->map);
         drm_ht_remove(&dev->map_hash);
 
-        drm_legacy_ctxbitmap_cleanup(dev);
+        drm_ctxbitmap_cleanup(dev);
 
         if (drm_core_check_feature(dev, DRIVER_MODESET))
                 drm_put_minor(&dev->control);
diff --git a/drivers/gpu/drm/exynos/Kconfig b/drivers/gpu/drm/exynos/Kconfig
index 4752f223e5b2..45b6ef595965 100644
--- a/drivers/gpu/drm/exynos/Kconfig
+++ b/drivers/gpu/drm/exynos/Kconfig
@@ -56,7 +56,7 @@ config DRM_EXYNOS_IPP
 
 config DRM_EXYNOS_FIMC
         bool "Exynos DRM FIMC"
-        depends on DRM_EXYNOS_IPP && MFD_SYSCON && OF
+        depends on DRM_EXYNOS_IPP && MFD_SYSCON
         help
           Choose this option if you want to use Exynos FIMC for DRM.
 
diff --git a/drivers/gpu/drm/exynos/exynos_drm_buf.c b/drivers/gpu/drm/exynos/exynos_drm_buf.c
index 3445a0f3a6b2..9c8088462c26 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_buf.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_buf.c
@@ -63,7 +63,8 @@ static int lowlevel_buffer_allocate(struct drm_device *dev,
                         return -ENOMEM;
                 }
 
-                buf->kvaddr = dma_alloc_attrs(dev->dev, buf->size,
+                buf->kvaddr = (void __iomem *)dma_alloc_attrs(dev->dev,
+                                        buf->size,
                                         &buf->dma_addr, GFP_KERNEL,
                                         &buf->dma_attrs);
                 if (!buf->kvaddr) {
@@ -90,9 +91,9 @@ static int lowlevel_buffer_allocate(struct drm_device *dev,
         }
 
         buf->sgt = drm_prime_pages_to_sg(buf->pages, nr_pages);
-        if (!buf->sgt) {
+        if (IS_ERR(buf->sgt)) {
                 DRM_ERROR("failed to get sg table.\n");
-                ret = -ENOMEM;
+                ret = PTR_ERR(buf->sgt);
                 goto err_free_attrs;
         }
 
diff --git a/drivers/gpu/drm/exynos/exynos_drm_fbdev.c b/drivers/gpu/drm/exynos/exynos_drm_fbdev.c
index 78e868bcf1ec..e7c2f2d07f19 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_fbdev.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_fbdev.c
@@ -99,12 +99,13 @@ static int exynos_drm_fbdev_update(struct drm_fb_helper *helper,
                 if (is_drm_iommu_supported(dev)) {
                         unsigned int nr_pages = buffer->size >> PAGE_SHIFT;
 
-                        buffer->kvaddr = vmap(buffer->pages, nr_pages, VM_MAP,
+                        buffer->kvaddr = (void __iomem *) vmap(buffer->pages,
+                                        nr_pages, VM_MAP,
                                         pgprot_writecombine(PAGE_KERNEL));
                 } else {
                         phys_addr_t dma_addr = buffer->dma_addr;
                         if (dma_addr)
-                                buffer->kvaddr = phys_to_virt(dma_addr);
+                                buffer->kvaddr = (void __iomem *)phys_to_virt(dma_addr);
                         else
                                 buffer->kvaddr = (void __iomem *)NULL;
                 }
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index 8507c6d1e642..df9253d890ee 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -1392,14 +1392,11 @@ out:
                 if (i915_terminally_wedged(&dev_priv->gpu_error))
                         return VM_FAULT_SIGBUS;
         case -EAGAIN:
-                /* Give the error handler a chance to run and move the
-                 * objects off the GPU active list. Next time we service the
-                 * fault, we should be able to transition the page into the
-                 * GTT without touching the GPU (and so avoid further
-                 * EIO/EGAIN). If the GPU is wedged, then there is no issue
-                 * with coherency, just lost writes.
+                /*
+                 * EAGAIN means the gpu is hung and we'll wait for the error
+                 * handler to reset everything when re-faulting in
+                 * i915_mutex_lock_interruptible.
                  */
-                set_need_resched();
         case 0:
         case -ERESTARTSYS:
         case -EINTR:
diff --git a/drivers/gpu/drm/i915/i915_irq.c b/drivers/gpu/drm/i915/i915_irq.c
index 83cce0cdb769..4b91228fd9bd 100644
--- a/drivers/gpu/drm/i915/i915_irq.c
+++ b/drivers/gpu/drm/i915/i915_irq.c
@@ -1469,6 +1469,34 @@ static irqreturn_t ironlake_irq_handler(int irq, void *arg)
         return ret;
 }
 
+static void i915_error_wake_up(struct drm_i915_private *dev_priv,
+                               bool reset_completed)
+{
+        struct intel_ring_buffer *ring;
+        int i;
+
+        /*
+         * Notify all waiters for GPU completion events that reset state has
+         * been changed, and that they need to restart their wait after
+         * checking for potential errors (and bail out to drop locks if there is
+         * a gpu reset pending so that i915_error_work_func can acquire them).
+         */
+
+        /* Wake up __wait_seqno, potentially holding dev->struct_mutex. */
+        for_each_ring(ring, dev_priv, i)
+                wake_up_all(&ring->irq_queue);
+
+        /* Wake up intel_crtc_wait_for_pending_flips, holding crtc->mutex. */
+        wake_up_all(&dev_priv->pending_flip_queue);
+
+        /*
+         * Signal tasks blocked in i915_gem_wait_for_error that the pending
+         * reset state is cleared.
+         */
+        if (reset_completed)
+                wake_up_all(&dev_priv->gpu_error.reset_queue);
+}
+
 /**
  * i915_error_work_func - do process context error handling work
  * @work: work struct
@@ -1483,11 +1511,10 @@ static void i915_error_work_func(struct work_struct *work)
         drm_i915_private_t *dev_priv = container_of(error, drm_i915_private_t,
                                                     gpu_error);
         struct drm_device *dev = dev_priv->dev;
-        struct intel_ring_buffer *ring;
         char *error_event[] = { I915_ERROR_UEVENT "=1", NULL };
         char *reset_event[] = { I915_RESET_UEVENT "=1", NULL };
         char *reset_done_event[] = { I915_ERROR_UEVENT "=0", NULL };
-        int i, ret;
+        int ret;
 
         kobject_uevent_env(&dev->primary->kdev.kobj, KOBJ_CHANGE, error_event);
 
@@ -1506,8 +1533,16 @@ static void i915_error_work_func(struct work_struct *work)
                 kobject_uevent_env(&dev->primary->kdev.kobj, KOBJ_CHANGE,
                                    reset_event);
 
+                /*
+                 * All state reset _must_ be completed before we update the
+                 * reset counter, for otherwise waiters might miss the reset
+                 * pending state and not properly drop locks, resulting in
+                 * deadlocks with the reset work.
+                 */
                 ret = i915_reset(dev);
 
+                intel_display_handle_reset(dev);
+
                 if (ret == 0) {
                         /*
                          * After all the gem state is reset, increment the reset
@@ -1528,12 +1563,11 @@ static void i915_error_work_func(struct work_struct *work)
                         atomic_set(&error->reset_counter, I915_WEDGED);
                 }
 
-                for_each_ring(ring, dev_priv, i)
-                        wake_up_all(&ring->irq_queue);
-
-                intel_display_handle_reset(dev);
-
-                wake_up_all(&dev_priv->gpu_error.reset_queue);
+                /*
+                 * Note: The wake_up also serves as a memory barrier so that
+                 * waiters see the update value of the reset counter atomic_t.
+                 */
+                i915_error_wake_up(dev_priv, true);
         }
 }
 
@@ -1642,8 +1676,6 @@ static void i915_report_and_clear_eir(struct drm_device *dev)
 void i915_handle_error(struct drm_device *dev, bool wedged)
 {
         struct drm_i915_private *dev_priv = dev->dev_private;
-        struct intel_ring_buffer *ring;
-        int i;
 
         i915_capture_error_state(dev);
         i915_report_and_clear_eir(dev);
@@ -1653,11 +1685,19 @@ void i915_handle_error(struct drm_device *dev, bool wedged)
                                 &dev_priv->gpu_error.reset_counter);
 
                 /*
-                 * Wakeup waiting processes so that the reset work item
-                 * doesn't deadlock trying to grab various locks.
+                 * Wakeup waiting processes so that the reset work function
+                 * i915_error_work_func doesn't deadlock trying to grab various
+                 * locks. By bumping the reset counter first, the woken
+                 * processes will see a reset in progress and back off,
+                 * releasing their locks and then wait for the reset completion.
+                 * We must do this for _all_ gpu waiters that might hold locks
+                 * that the reset work needs to acquire.
+                 *
+                 * Note: The wake_up serves as the required memory barrier to
+                 * ensure that the waiters see the updated value of the reset
+                 * counter atomic_t.
                  */
-                for_each_ring(ring, dev_priv, i)
-                        wake_up_all(&ring->irq_queue);
+                i915_error_wake_up(dev_priv, false);
         }
 
         /*
diff --git a/drivers/gpu/drm/i915/intel_ddi.c b/drivers/gpu/drm/i915/intel_ddi.c
index 63aca49d11a8..63de2701b974 100644
--- a/drivers/gpu/drm/i915/intel_ddi.c
+++ b/drivers/gpu/drm/i915/intel_ddi.c
@@ -778,7 +778,7 @@ void intel_ddi_enable_transcoder_func(struct drm_crtc *crtc)
                         /* Can only use the always-on power well for eDP when
                          * not using the panel fitter, and when not using motion
                           * blur mitigation (which we don't support). */
-                        if (intel_crtc->config.pch_pfit.size)
+                        if (intel_crtc->config.pch_pfit.enabled)
                                 temp |= TRANS_DDI_EDP_INPUT_A_ONOFF;
                         else
                                 temp |= TRANS_DDI_EDP_INPUT_A_ON;
diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index 2489d0b4c7d2..d8a1d98693e7 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -2249,7 +2249,7 @@ intel_pipe_set_base(struct drm_crtc *crtc, int x, int y,
                 I915_WRITE(PIPESRC(intel_crtc->pipe),
                            ((crtc->mode.hdisplay - 1) << 16) |
                            (crtc->mode.vdisplay - 1));
-                if (!intel_crtc->config.pch_pfit.size &&
+                if (!intel_crtc->config.pch_pfit.enabled &&
                     (intel_pipe_has_type(crtc, INTEL_OUTPUT_LVDS) ||
                      intel_pipe_has_type(crtc, INTEL_OUTPUT_EDP))) {
                         I915_WRITE(PF_CTL(intel_crtc->pipe), 0);
@@ -3203,7 +3203,7 @@ static void ironlake_pfit_enable(struct intel_crtc *crtc)
         struct drm_i915_private *dev_priv = dev->dev_private;
         int pipe = crtc->pipe;
 
-        if (crtc->config.pch_pfit.size) {
+        if (crtc->config.pch_pfit.enabled) {
                 /* Force use of hard-coded filter coefficients
                  * as some pre-programmed values are broken,
                  * e.g. x201.
@@ -3428,7 +3428,7 @@ static void ironlake_pfit_disable(struct intel_crtc *crtc)
 
         /* To avoid upsetting the power well on haswell only disable the pfit if
          * it's in use. The hw state code will make sure we get this right. */
-        if (crtc->config.pch_pfit.size) {
+        if (crtc->config.pch_pfit.enabled) {
                 I915_WRITE(PF_CTL(pipe), 0);
                 I915_WRITE(PF_WIN_POS(pipe), 0);
                 I915_WRITE(PF_WIN_SZ(pipe), 0);
@@ -4877,9 +4877,6 @@ static int i9xx_crtc_mode_set(struct drm_crtc *crtc,
                 return -EINVAL;
         }
 
-        /* Ensure that the cursor is valid for the new mode before changing... */
-        intel_crtc_update_cursor(crtc, true);
-
         if (is_lvds && dev_priv->lvds_downclock_avail) {
                 /*
                  * Ensure we match the reduced clock's P to the target clock.
@@ -5768,9 +5765,6 @@ static int ironlake_crtc_mode_set(struct drm_crtc *crtc,
                 intel_crtc->config.dpll.p2 = clock.p2;
         }
 
-        /* Ensure that the cursor is valid for the new mode before changing... */
-        intel_crtc_update_cursor(crtc, true);
-
         /* CPU eDP is the only output that doesn't need a PCH PLL of its own. */
         if (intel_crtc->config.has_pch_encoder) {
                 fp = i9xx_dpll_compute_fp(&intel_crtc->config.dpll);
@@ -5859,6 +5853,7 @@ static void ironlake_get_pfit_config(struct intel_crtc *crtc,
         tmp = I915_READ(PF_CTL(crtc->pipe));
 
         if (tmp & PF_ENABLE) {
+                pipe_config->pch_pfit.enabled = true;
                 pipe_config->pch_pfit.pos = I915_READ(PF_WIN_POS(crtc->pipe));
                 pipe_config->pch_pfit.size = I915_READ(PF_WIN_SZ(crtc->pipe));
 
@@ -6236,7 +6231,7 @@ static void haswell_modeset_global_resources(struct drm_device *dev)
                 if (!crtc->base.enabled)
                         continue;
 
-                if (crtc->pipe != PIPE_A || crtc->config.pch_pfit.size ||
+                if (crtc->pipe != PIPE_A || crtc->config.pch_pfit.enabled ||
                     crtc->config.cpu_transcoder != TRANSCODER_EDP)
                         enable = true;
         }
@@ -6259,9 +6254,6 @@ static int haswell_crtc_mode_set(struct drm_crtc *crtc,
         if (!intel_ddi_pll_mode_set(crtc))
                 return -EINVAL;
 
-        /* Ensure that the cursor is valid for the new mode before changing... */
-        intel_crtc_update_cursor(crtc, true);
-
         if (intel_crtc->config.has_dp_encoder)
                 intel_dp_set_m_n(intel_crtc);
 
@@ -6494,15 +6486,15 @@ static void haswell_write_eld(struct drm_connector *connector,
 
         /* Set ELD valid state */
         tmp = I915_READ(aud_cntrl_st2);
-        DRM_DEBUG_DRIVER("HDMI audio: pin eld vld status=0x%8x\n", tmp);
+        DRM_DEBUG_DRIVER("HDMI audio: pin eld vld status=0x%08x\n", tmp);
         tmp |= (AUDIO_ELD_VALID_A << (pipe * 4));
         I915_WRITE(aud_cntrl_st2, tmp);
         tmp = I915_READ(aud_cntrl_st2);
-        DRM_DEBUG_DRIVER("HDMI audio: eld vld status=0x%8x\n", tmp);
+        DRM_DEBUG_DRIVER("HDMI audio: eld vld status=0x%08x\n", tmp);
 
         /* Enable HDMI mode */
         tmp = I915_READ(aud_config);
-        DRM_DEBUG_DRIVER("HDMI audio: audio conf: 0x%8x\n", tmp);
+        DRM_DEBUG_DRIVER("HDMI audio: audio conf: 0x%08x\n", tmp);
         /* clear N_programing_enable and N_value_index */
         tmp &= ~(AUD_CONFIG_N_VALUE_INDEX | AUD_CONFIG_N_PROG_ENABLE);
         I915_WRITE(aud_config, tmp);
@@ -6937,7 +6929,8 @@ static int intel_crtc_cursor_set(struct drm_crtc *crtc,
         intel_crtc->cursor_width = width;
         intel_crtc->cursor_height = height;
 
-        intel_crtc_update_cursor(crtc, intel_crtc->cursor_bo != NULL);
+        if (intel_crtc->active)
+                intel_crtc_update_cursor(crtc, intel_crtc->cursor_bo != NULL);
 
         return 0;
 fail_unpin:
@@ -6956,7 +6949,8 @@ static int intel_crtc_cursor_move(struct drm_crtc *crtc, int x, int y)
         intel_crtc->cursor_x = x;
         intel_crtc->cursor_y = y;
 
-        intel_crtc_update_cursor(crtc, intel_crtc->cursor_bo != NULL);
+        if (intel_crtc->active)
+                intel_crtc_update_cursor(crtc, intel_crtc->cursor_bo != NULL);
 
         return 0;
 }
@@ -8205,9 +8199,10 @@ static void intel_dump_pipe_config(struct intel_crtc *crtc,
                       pipe_config->gmch_pfit.control,
                       pipe_config->gmch_pfit.pgm_ratios,
                       pipe_config->gmch_pfit.lvds_border_bits);
-        DRM_DEBUG_KMS("pch pfit: pos: 0x%08x, size: 0x%08x\n",
+        DRM_DEBUG_KMS("pch pfit: pos: 0x%08x, size: 0x%08x, %s\n",
                       pipe_config->pch_pfit.pos,
-                      pipe_config->pch_pfit.size);
+                      pipe_config->pch_pfit.size,
+                      pipe_config->pch_pfit.enabled ? "enabled" : "disabled");
         DRM_DEBUG_KMS("ips: %i\n", pipe_config->ips_enabled);
 }
 
@@ -8603,8 +8598,11 @@ intel_pipe_config_compare(struct drm_device *dev,
         if (INTEL_INFO(dev)->gen < 4)
                 PIPE_CONF_CHECK_I(gmch_pfit.pgm_ratios);
         PIPE_CONF_CHECK_I(gmch_pfit.lvds_border_bits);
-        PIPE_CONF_CHECK_I(pch_pfit.pos);
-        PIPE_CONF_CHECK_I(pch_pfit.size);
+        PIPE_CONF_CHECK_I(pch_pfit.enabled);
+        if (current_config->pch_pfit.enabled) {
+                PIPE_CONF_CHECK_I(pch_pfit.pos);
+                PIPE_CONF_CHECK_I(pch_pfit.size);
+        }
 
         PIPE_CONF_CHECK_I(ips_enabled);
 
diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h
index a47799e832c6..28cae80495e2 100644
--- a/drivers/gpu/drm/i915/intel_drv.h
+++ b/drivers/gpu/drm/i915/intel_drv.h
@@ -280,6 +280,7 @@ struct intel_crtc_config {
         struct {
                 u32 pos;
                 u32 size;
+                bool enabled;
         } pch_pfit;
 
         /* FDI configuration, only valid if has_pch_encoder is set. */
diff --git a/drivers/gpu/drm/i915/intel_dvo.c b/drivers/gpu/drm/i915/intel_dvo.c
index 406303b509c1..7fa7df546c1e 100644
--- a/drivers/gpu/drm/i915/intel_dvo.c
+++ b/drivers/gpu/drm/i915/intel_dvo.c
@@ -263,6 +263,8 @@ static bool intel_dvo_compute_config(struct intel_encoder *encoder,
                 C(vtotal);
                 C(clock);
 #undef C
+
+                drm_mode_set_crtcinfo(adjusted_mode, 0);
         }
 
         if (intel_dvo->dev.dev_ops->mode_fixup)
diff --git a/drivers/gpu/drm/i915/intel_panel.c b/drivers/gpu/drm/i915/intel_panel.c
index 42114ecbae0e..293564a2896a 100644
--- a/drivers/gpu/drm/i915/intel_panel.c
+++ b/drivers/gpu/drm/i915/intel_panel.c
@@ -112,6 +112,7 @@ intel_pch_panel_fitting(struct intel_crtc *intel_crtc,
 done:
         pipe_config->pch_pfit.pos = (x << 16) | y;
         pipe_config->pch_pfit.size = (width << 16) | height;
+        pipe_config->pch_pfit.enabled = pipe_config->pch_pfit.size != 0;
 }
 
 static void
diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c
index 0c115cc4899f..dd176b7296c1 100644
--- a/drivers/gpu/drm/i915/intel_pm.c
+++ b/drivers/gpu/drm/i915/intel_pm.c
@@ -2096,16 +2096,16 @@ static uint32_t ilk_pipe_pixel_rate(struct drm_device *dev,
                                     struct drm_crtc *crtc)
 {
         struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-        uint32_t pixel_rate, pfit_size;
+        uint32_t pixel_rate;
 
         pixel_rate = intel_crtc->config.adjusted_mode.clock;
 
         /* We only use IF-ID interlacing. If we ever use PF-ID we'll need to
          * adjust the pixel_rate here. */
 
-        pfit_size = intel_crtc->config.pch_pfit.size;
-        if (pfit_size) {
+        if (intel_crtc->config.pch_pfit.enabled) {
                 uint64_t pipe_w, pipe_h, pfit_w, pfit_h;
+                uint32_t pfit_size = intel_crtc->config.pch_pfit.size;
 
                 pipe_w = intel_crtc->config.requested_mode.hdisplay;
                 pipe_h = intel_crtc->config.requested_mode.vdisplay;
diff --git a/drivers/gpu/drm/i915/intel_sdvo.c b/drivers/gpu/drm/i915/intel_sdvo.c
index 85037b9d4934..49482fd5b76c 100644
--- a/drivers/gpu/drm/i915/intel_sdvo.c
+++ b/drivers/gpu/drm/i915/intel_sdvo.c
@@ -788,6 +788,8 @@ static void intel_sdvo_get_dtd_from_mode(struct intel_sdvo_dtd *dtd,
         uint16_t h_sync_offset, v_sync_offset;
         int mode_clock;
 
+        memset(dtd, 0, sizeof(*dtd));
+
         width = mode->hdisplay;
         height = mode->vdisplay;
 
@@ -830,44 +832,51 @@ static void intel_sdvo_get_dtd_from_mode(struct intel_sdvo_dtd *dtd,
         if (mode->flags & DRM_MODE_FLAG_PVSYNC)
                 dtd->part2.dtd_flags |= DTD_FLAG_VSYNC_POSITIVE;
 
-        dtd->part2.sdvo_flags = 0;
         dtd->part2.v_sync_off_high = v_sync_offset & 0xc0;
-        dtd->part2.reserved = 0;
 }
 
-static void intel_sdvo_get_mode_from_dtd(struct drm_display_mode * mode,
+static void intel_sdvo_get_mode_from_dtd(struct drm_display_mode *pmode,
                                          const struct intel_sdvo_dtd *dtd)
 {
-        mode->hdisplay = dtd->part1.h_active;
-        mode->hdisplay += ((dtd->part1.h_high >> 4) & 0x0f) << 8;
-        mode->hsync_start = mode->hdisplay + dtd->part2.h_sync_off;
-        mode->hsync_start += (dtd->part2.sync_off_width_high & 0xc0) << 2;
-        mode->hsync_end = mode->hsync_start + dtd->part2.h_sync_width;
-        mode->hsync_end += (dtd->part2.sync_off_width_high & 0x30) << 4;
-        mode->htotal = mode->hdisplay + dtd->part1.h_blank;
-        mode->htotal += (dtd->part1.h_high & 0xf) << 8;
-
-        mode->vdisplay = dtd->part1.v_active;
-        mode->vdisplay += ((dtd->part1.v_high >> 4) & 0x0f) << 8;
-        mode->vsync_start = mode->vdisplay;
-        mode->vsync_start += (dtd->part2.v_sync_off_width >> 4) & 0xf;
-        mode->vsync_start += (dtd->part2.sync_off_width_high & 0x0c) << 2;
-        mode->vsync_start += dtd->part2.v_sync_off_high & 0xc0;
-        mode->vsync_end = mode->vsync_start +
+        struct drm_display_mode mode = {};
+
+        mode.hdisplay = dtd->part1.h_active;
+        mode.hdisplay += ((dtd->part1.h_high >> 4) & 0x0f) << 8;
+        mode.hsync_start = mode.hdisplay + dtd->part2.h_sync_off;
+        mode.hsync_start += (dtd->part2.sync_off_width_high & 0xc0) << 2;
+        mode.hsync_end = mode.hsync_start + dtd->part2.h_sync_width;
+        mode.hsync_end += (dtd->part2.sync_off_width_high & 0x30) << 4;
+        mode.htotal = mode.hdisplay + dtd->part1.h_blank;
+        mode.htotal += (dtd->part1.h_high & 0xf) << 8;
+
+        mode.vdisplay = dtd->part1.v_active;
+        mode.vdisplay += ((dtd->part1.v_high >> 4) & 0x0f) << 8;
+        mode.vsync_start = mode.vdisplay;
+        mode.vsync_start += (dtd->part2.v_sync_off_width >> 4) & 0xf;
+        mode.vsync_start += (dtd->part2.sync_off_width_high & 0x0c) << 2;
+        mode.vsync_start += dtd->part2.v_sync_off_high & 0xc0;
+        mode.vsync_end = mode.vsync_start +
                 (dtd->part2.v_sync_off_width & 0xf);
-        mode->vsync_end += (dtd->part2.sync_off_width_high & 0x3) << 4;
-        mode->vtotal = mode->vdisplay + dtd->part1.v_blank;
-        mode->vtotal += (dtd->part1.v_high & 0xf) << 8;
+        mode.vsync_end += (dtd->part2.sync_off_width_high & 0x3) << 4;
+        mode.vtotal = mode.vdisplay + dtd->part1.v_blank;
+        mode.vtotal += (dtd->part1.v_high & 0xf) << 8;
 
-        mode->clock = dtd->part1.clock * 10;
+        mode.clock = dtd->part1.clock * 10;
 
-        mode->flags &= ~(DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC);
         if (dtd->part2.dtd_flags & DTD_FLAG_INTERLACE)
-                mode->flags |= DRM_MODE_FLAG_INTERLACE;
+                mode.flags |= DRM_MODE_FLAG_INTERLACE;
         if (dtd->part2.dtd_flags & DTD_FLAG_HSYNC_POSITIVE)
-                mode->flags |= DRM_MODE_FLAG_PHSYNC;
+                mode.flags |= DRM_MODE_FLAG_PHSYNC;
+        else
+                mode.flags |= DRM_MODE_FLAG_NHSYNC;
         if (dtd->part2.dtd_flags & DTD_FLAG_VSYNC_POSITIVE)
-                mode->flags |= DRM_MODE_FLAG_PVSYNC;
+                mode.flags |= DRM_MODE_FLAG_PVSYNC;
+        else
+                mode.flags |= DRM_MODE_FLAG_NVSYNC;
+
+        drm_mode_set_crtcinfo(&mode, 0);
+
+        drm_mode_copy(pmode, &mode);
 }
 
 static bool intel_sdvo_check_supp_encode(struct intel_sdvo *intel_sdvo)
diff --git a/drivers/gpu/drm/msm/adreno/adreno_gpu.c b/drivers/gpu/drm/msm/adreno/adreno_gpu.c
index a60584763b61..a0b9d8a95b16 100644
--- a/drivers/gpu/drm/msm/adreno/adreno_gpu.c
+++ b/drivers/gpu/drm/msm/adreno/adreno_gpu.c
@@ -124,6 +124,8 @@ void adreno_recover(struct msm_gpu *gpu)
 
         /* reset completed fence seqno, just discard anything pending: */
         adreno_gpu->memptrs->fence = gpu->submitted_fence;
+        adreno_gpu->memptrs->rptr  = 0;
+        adreno_gpu->memptrs->wptr  = 0;
 
         gpu->funcs->pm_resume(gpu);
         ret = gpu->funcs->hw_init(gpu);
@@ -229,7 +231,7 @@ void adreno_idle(struct msm_gpu *gpu)
                         return;
         } while(time_before(jiffies, t));
 
-        DRM_ERROR("timeout waiting for %s to drain ringbuffer!\n", gpu->name);
+        DRM_ERROR("%s: timeout waiting to drain ringbuffer!\n", gpu->name);
 
         /* TODO maybe we need to reset GPU here to recover from hang? */
 }
@@ -256,11 +258,17 @@ void adreno_wait_ring(struct msm_gpu *gpu, uint32_t ndwords)
 {
         struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu);
         uint32_t freedwords;
+        unsigned long t = jiffies + ADRENO_IDLE_TIMEOUT;
         do {
                 uint32_t size = gpu->rb->size / 4;
                 uint32_t wptr = get_wptr(gpu->rb);
                 uint32_t rptr = adreno_gpu->memptrs->rptr;
                 freedwords = (rptr + (size - 1) - wptr) % size;
+
+                if (time_after(jiffies, t)) {
+                        DRM_ERROR("%s: timeout waiting for ringbuffer space\n", gpu->name);
+                        break;
+                }
         } while(freedwords < ndwords);
 }
 
diff --git a/drivers/gpu/drm/msm/msm_drv.c b/drivers/gpu/drm/msm/msm_drv.c
index 864c9773636b..008d772384c7 100644
--- a/drivers/gpu/drm/msm/msm_drv.c
+++ b/drivers/gpu/drm/msm/msm_drv.c
@@ -499,25 +499,41 @@ int msm_wait_fence_interruptable(struct drm_device *dev, uint32_t fence,
                 struct timespec *timeout)
 {
         struct msm_drm_private *priv = dev->dev_private;
-        unsigned long timeout_jiffies = timespec_to_jiffies(timeout);
-        unsigned long start_jiffies = jiffies;
-        unsigned long remaining_jiffies;
         int ret;
 
-        if (time_after(start_jiffies, timeout_jiffies))
-                remaining_jiffies = 0;
-        else
-                remaining_jiffies = timeout_jiffies - start_jiffies;
-
-        ret = wait_event_interruptible_timeout(priv->fence_event,
-                        priv->completed_fence >= fence,
-                        remaining_jiffies);
-        if (ret == 0) {
-                DBG("timeout waiting for fence: %u (completed: %u)",
-                                fence, priv->completed_fence);
-                ret = -ETIMEDOUT;
-        } else if (ret != -ERESTARTSYS) {
-                ret = 0;
+        if (!priv->gpu)
+                return 0;
+
+        if (fence > priv->gpu->submitted_fence) {
+                DRM_ERROR("waiting on invalid fence: %u (of %u)\n",
+                                fence, priv->gpu->submitted_fence);
+                return -EINVAL;
+        }
+
+        if (!timeout) {
+                /* no-wait: */
+                ret = fence_completed(dev, fence) ? 0 : -EBUSY;
+        } else {
+                unsigned long timeout_jiffies = timespec_to_jiffies(timeout);
+                unsigned long start_jiffies = jiffies;
+                unsigned long remaining_jiffies;
+
+                if (time_after(start_jiffies, timeout_jiffies))
+                        remaining_jiffies = 0;
+                else
+                        remaining_jiffies = timeout_jiffies - start_jiffies;
+
+                ret = wait_event_interruptible_timeout(priv->fence_event,
+                                fence_completed(dev, fence),
+                                remaining_jiffies);
+
+                if (ret == 0) {
+                        DBG("timeout waiting for fence: %u (completed: %u)",
+                                        fence, priv->completed_fence);
+                        ret = -ETIMEDOUT;
+                } else if (ret != -ERESTARTSYS) {
+                        ret = 0;
+                }
         }
 
         return ret;
diff --git a/drivers/gpu/drm/msm/msm_drv.h b/drivers/gpu/drm/msm/msm_drv.h
index 80d75094bf0a..df8f1d084bc1 100644
--- a/drivers/gpu/drm/msm/msm_drv.h
+++ b/drivers/gpu/drm/msm/msm_drv.h
@@ -153,7 +153,7 @@ void *msm_gem_vaddr(struct drm_gem_object *obj);
 int msm_gem_queue_inactive_work(struct drm_gem_object *obj,
                 struct work_struct *work);
 void msm_gem_move_to_active(struct drm_gem_object *obj,
-                struct msm_gpu *gpu, uint32_t fence);
+                struct msm_gpu *gpu, bool write, uint32_t fence);
 void msm_gem_move_to_inactive(struct drm_gem_object *obj);
 int msm_gem_cpu_prep(struct drm_gem_object *obj, uint32_t op,
                 struct timespec *timeout);
@@ -191,6 +191,12 @@ u32 msm_readl(const void __iomem *addr);
 #define DBG(fmt, ...) DRM_DEBUG(fmt"\n", ##__VA_ARGS__)
 #define VERB(fmt, ...) if (0) DRM_DEBUG(fmt"\n", ##__VA_ARGS__)
 
+static inline bool fence_completed(struct drm_device *dev, uint32_t fence)
+{
+        struct msm_drm_private *priv = dev->dev_private;
+        return priv->completed_fence >= fence;
+}
+
 static inline int align_pitch(int width, int bpp)
 {
         int bytespp = (bpp + 7) / 8;
diff --git a/drivers/gpu/drm/msm/msm_gem.c b/drivers/gpu/drm/msm/msm_gem.c
index 6b5a6c8c7658..29eacfa29cfb 100644
--- a/drivers/gpu/drm/msm/msm_gem.c
+++ b/drivers/gpu/drm/msm/msm_gem.c
@@ -40,9 +40,9 @@ static struct page **get_pages(struct drm_gem_object *obj)
                 }
 
                 msm_obj->sgt = drm_prime_pages_to_sg(p, npages);
-                if (!msm_obj->sgt) {
+                if (IS_ERR(msm_obj->sgt)) {
                         dev_err(dev->dev, "failed to allocate sgt\n");
-                        return ERR_PTR(-ENOMEM);
+                        return ERR_CAST(msm_obj->sgt);
                 }
 
                 msm_obj->pages = p;
@@ -159,7 +159,6 @@ out_unlock:
 out:
         switch (ret) {
         case -EAGAIN:
-                set_need_resched();
         case 0:
         case -ERESTARTSYS:
         case -EINTR:
@@ -393,11 +392,14 @@ int msm_gem_queue_inactive_work(struct drm_gem_object *obj,
 }
 
 void msm_gem_move_to_active(struct drm_gem_object *obj,
-                struct msm_gpu *gpu, uint32_t fence)
+                struct msm_gpu *gpu, bool write, uint32_t fence)
 {
         struct msm_gem_object *msm_obj = to_msm_bo(obj);
         msm_obj->gpu = gpu;
-        msm_obj->fence = fence;
+        if (write)
+                msm_obj->write_fence = fence;
+        else
+                msm_obj->read_fence = fence;
         list_del_init(&msm_obj->mm_list);
         list_add_tail(&msm_obj->mm_list, &gpu->active_list);
 }
@@ -411,7 +413,8 @@ void msm_gem_move_to_inactive(struct drm_gem_object *obj)
         WARN_ON(!mutex_is_locked(&dev->struct_mutex));
 
         msm_obj->gpu = NULL;
-        msm_obj->fence = 0;
+        msm_obj->read_fence = 0;
+        msm_obj->write_fence = 0;
         list_del_init(&msm_obj->mm_list);
         list_add_tail(&msm_obj->mm_list, &priv->inactive_list);
 
@@ -433,8 +436,18 @@ int msm_gem_cpu_prep(struct drm_gem_object *obj, uint32_t op,
         struct msm_gem_object *msm_obj = to_msm_bo(obj);
         int ret = 0;
 
-        if (is_active(msm_obj) && !(op & MSM_PREP_NOSYNC))
-                ret = msm_wait_fence_interruptable(dev, msm_obj->fence, timeout);
+        if (is_active(msm_obj)) {
+                uint32_t fence = 0;
+
+                if (op & MSM_PREP_READ)
+                        fence = msm_obj->write_fence;
+                if (op & MSM_PREP_WRITE)
+                        fence = max(fence, msm_obj->read_fence);
+                if (op & MSM_PREP_NOSYNC)
+                        timeout = NULL;
+
+                ret = msm_wait_fence_interruptable(dev, fence, timeout);
+        }
 
         /* TODO cache maintenance */
 
@@ -455,9 +468,10 @@ void msm_gem_describe(struct drm_gem_object *obj, struct seq_file *m)
         uint64_t off = drm_vma_node_start(&obj->vma_node);
 
         WARN_ON(!mutex_is_locked(&dev->struct_mutex));
-        seq_printf(m, "%08x: %c(%d) %2d (%2d) %08llx %p %d\n",
+        seq_printf(m, "%08x: %c(r=%u,w=%u) %2d (%2d) %08llx %p %d\n",
                         msm_obj->flags, is_active(msm_obj) ? 'A' : 'I',
-                        msm_obj->fence, obj->name, obj->refcount.refcount.counter,
+                        msm_obj->read_fence, msm_obj->write_fence,
+                        obj->name, obj->refcount.refcount.counter,
                         off, msm_obj->vaddr, obj->size);
 }
 
diff --git a/drivers/gpu/drm/msm/msm_gem.h b/drivers/gpu/drm/msm/msm_gem.h
index d746f13d283c..0676f32e2c6a 100644
--- a/drivers/gpu/drm/msm/msm_gem.h
+++ b/drivers/gpu/drm/msm/msm_gem.h
@@ -36,7 +36,7 @@ struct msm_gem_object {
          */
         struct list_head mm_list;
         struct msm_gpu *gpu;     /* non-null if active */
-        uint32_t fence;
+        uint32_t read_fence, write_fence;
 
         /* Transiently in the process of submit ioctl, objects associated
          * with the submit are on submit->bo_list.. this only lasts for
diff --git a/drivers/gpu/drm/msm/msm_gem_submit.c b/drivers/gpu/drm/msm/msm_gem_submit.c
index 3e1ef3a00f60..5281d4bc37f7 100644
--- a/drivers/gpu/drm/msm/msm_gem_submit.c
+++ b/drivers/gpu/drm/msm/msm_gem_submit.c
@@ -78,7 +78,7 @@ static int submit_lookup_objects(struct msm_gem_submit *submit,
                 }
 
                 if (submit_bo.flags & BO_INVALID_FLAGS) {
-                        DBG("invalid flags: %x", submit_bo.flags);
+                        DRM_ERROR("invalid flags: %x\n", submit_bo.flags);
                         ret = -EINVAL;
                         goto out_unlock;
                 }
@@ -92,7 +92,7 @@ static int submit_lookup_objects(struct msm_gem_submit *submit,
                  */
                 obj = idr_find(&file->object_idr, submit_bo.handle);
                 if (!obj) {
-                        DBG("invalid handle %u at index %u", submit_bo.handle, i);
+                        DRM_ERROR("invalid handle %u at index %u\n", submit_bo.handle, i);
                         ret = -EINVAL;
                         goto out_unlock;
                 }
@@ -100,7 +100,7 @@ static int submit_lookup_objects(struct msm_gem_submit *submit,
                 msm_obj = to_msm_bo(obj);
 
                 if (!list_empty(&msm_obj->submit_entry)) {
-                        DBG("handle %u at index %u already on submit list",
+                        DRM_ERROR("handle %u at index %u already on submit list\n",
                                         submit_bo.handle, i);
                         ret = -EINVAL;
                         goto out_unlock;
@@ -216,8 +216,9 @@ static int submit_bo(struct msm_gem_submit *submit, uint32_t idx,
                 struct msm_gem_object **obj, uint32_t *iova, bool *valid)
 {
         if (idx >= submit->nr_bos) {
-                DBG("invalid buffer index: %u (out of %u)", idx, submit->nr_bos);
-                return EINVAL;
+                DRM_ERROR("invalid buffer index: %u (out of %u)\n",
+                                idx, submit->nr_bos);
+                return -EINVAL;
         }
 
         if (obj)
@@ -239,7 +240,7 @@ static int submit_reloc(struct msm_gem_submit *submit, struct msm_gem_object *ob
         int ret;
 
         if (offset % 4) {
-                DBG("non-aligned cmdstream buffer: %u", offset);
+                DRM_ERROR("non-aligned cmdstream buffer: %u\n", offset);
                 return -EINVAL;
         }
 
@@ -266,7 +267,7 @@ static int submit_reloc(struct msm_gem_submit *submit, struct msm_gem_object *ob
                         return -EFAULT;
 
                 if (submit_reloc.submit_offset % 4) {
-                        DBG("non-aligned reloc offset: %u",
+                        DRM_ERROR("non-aligned reloc offset: %u\n",
                                         submit_reloc.submit_offset);
                         return -EINVAL;
                 }
@@ -276,7 +277,7 @@ static int submit_reloc(struct msm_gem_submit *submit, struct msm_gem_object *ob
 
                 if ((off >= (obj->base.size / 4)) ||
                                 (off < last_offset)) {
-                        DBG("invalid offset %u at reloc %u", off, i);
+                        DRM_ERROR("invalid offset %u at reloc %u\n", off, i);
                         return -EINVAL;
                 }
 
@@ -374,14 +375,15 @@ int msm_ioctl_gem_submit(struct drm_device *dev, void *data,
                         goto out;
 
                 if (submit_cmd.size % 4) {
-                        DBG("non-aligned cmdstream buffer size: %u",
+                        DRM_ERROR("non-aligned cmdstream buffer size: %u\n",
                                         submit_cmd.size);
                         ret = -EINVAL;
                         goto out;
                 }
 
-                if (submit_cmd.size >= msm_obj->base.size) {
-                        DBG("invalid cmdstream size: %u", submit_cmd.size);
+                if ((submit_cmd.size + submit_cmd.submit_offset) >=
+                                msm_obj->base.size) {
+                        DRM_ERROR("invalid cmdstream size: %u\n", submit_cmd.size);
                         ret = -EINVAL;
                         goto out;
                 }
diff --git a/drivers/gpu/drm/msm/msm_gpu.c b/drivers/gpu/drm/msm/msm_gpu.c
index e1e1ec9321ff..3bab937965d1 100644
--- a/drivers/gpu/drm/msm/msm_gpu.c
+++ b/drivers/gpu/drm/msm/msm_gpu.c
@@ -29,13 +29,14 @@
 static void bs_init(struct msm_gpu *gpu, struct platform_device *pdev)
 {
         struct drm_device *dev = gpu->dev;
-        struct kgsl_device_platform_data *pdata = pdev->dev.platform_data;
+        struct kgsl_device_platform_data *pdata;
 
         if (!pdev) {
                 dev_err(dev->dev, "could not find dtv pdata\n");
                 return;
         }
 
+        pdata = pdev->dev.platform_data;
         if (pdata->bus_scale_table) {
                 gpu->bsc = msm_bus_scale_register_client(pdata->bus_scale_table);
                 DBG("bus scale client: %08x", gpu->bsc);
@@ -230,6 +231,8 @@ static void hangcheck_timer_reset(struct msm_gpu *gpu)
 static void hangcheck_handler(unsigned long data)
 {
         struct msm_gpu *gpu = (struct msm_gpu *)data;
+        struct drm_device *dev = gpu->dev;
+        struct msm_drm_private *priv = dev->dev_private;
         uint32_t fence = gpu->funcs->last_fence(gpu);
 
         if (fence != gpu->hangcheck_fence) {
@@ -237,14 +240,22 @@ static void hangcheck_handler(unsigned long data)
                 gpu->hangcheck_fence = fence;
         } else if (fence < gpu->submitted_fence) {
                 /* no progress and not done.. hung! */
-                struct msm_drm_private *priv = gpu->dev->dev_private;
                 gpu->hangcheck_fence = fence;
+                dev_err(dev->dev, "%s: hangcheck detected gpu lockup!\n",
+                                gpu->name);
+                dev_err(dev->dev, "%s:     completed fence: %u\n",
+                                gpu->name, fence);
+                dev_err(dev->dev, "%s:     submitted fence: %u\n",
+                                gpu->name, gpu->submitted_fence);
                 queue_work(priv->wq, &gpu->recover_work);
         }
 
         /* if still more pending work, reset the hangcheck timer: */
         if (gpu->submitted_fence > gpu->hangcheck_fence)
                 hangcheck_timer_reset(gpu);
+
+        /* workaround for missing irq: */
+        queue_work(priv->wq, &gpu->retire_work);
 }
 
 /*
@@ -265,7 +276,8 @@ static void retire_worker(struct work_struct *work)
                 obj = list_first_entry(&gpu->active_list,
                                 struct msm_gem_object, mm_list);
 
-                if (obj->fence <= fence) {
+                if ((obj->read_fence <= fence) &&
+                                (obj->write_fence <= fence)) {
                         /* move to inactive: */
                         msm_gem_move_to_inactive(&obj->base);
                         msm_gem_put_iova(&obj->base, gpu->id);
@@ -321,7 +333,11 @@ int msm_gpu_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit,
                                         submit->gpu->id, &iova);
                 }
 
-                msm_gem_move_to_active(&msm_obj->base, gpu, submit->fence);
+                if (submit->bos[i].flags & MSM_SUBMIT_BO_READ)
+                        msm_gem_move_to_active(&msm_obj->base, gpu, false, submit->fence);
+
+                if (submit->bos[i].flags & MSM_SUBMIT_BO_WRITE)
+                        msm_gem_move_to_active(&msm_obj->base, gpu, true, submit->fence);
         }
         hangcheck_timer_reset(gpu);
         mutex_unlock(&dev->struct_mutex);
diff --git a/include/drm/drmP.h b/include/drm/drmP.h
index 290734191f72..b46fb45f2cca 100644
--- a/include/drm/drmP.h
+++ b/include/drm/drmP.h
@@ -1322,10 +1322,9 @@ extern int drm_newctx(struct drm_device *dev, void *data,
 extern int drm_rmctx(struct drm_device *dev, void *data,
                      struct drm_file *file_priv);
 
-extern void drm_legacy_ctxbitmap_init(struct drm_device *dev);
-extern void drm_legacy_ctxbitmap_cleanup(struct drm_device *dev);
-extern void drm_legacy_ctxbitmap_release(struct drm_device *dev,
-                                         struct drm_file *file_priv);
+extern int drm_ctxbitmap_init(struct drm_device *dev);
+extern void drm_ctxbitmap_cleanup(struct drm_device *dev);
+extern void drm_ctxbitmap_free(struct drm_device *dev, int ctx_handle);
 
 extern int drm_setsareactx(struct drm_device *dev, void *data,
                            struct drm_file *file_priv);