68 files changed, 1061 insertions, 581 deletions
diff --git a/drivers/gpu/drm/drm_fb_cma_helper.c b/drivers/gpu/drm/drm_fb_cma_helper.c
index 09e11a5d921a..fd9d0af4d536 100644
--- a/drivers/gpu/drm/drm_fb_cma_helper.c
+++ b/drivers/gpu/drm/drm_fb_cma_helper.c
@@ -206,7 +206,7 @@ static int drm_fbdev_cma_create(struct drm_fb_helper *helper,
        size_t size;
        int ret;
-        DRM_DEBUG_KMS("surface width(%d), height(%d) and bpp(%d\n",
+        DRM_DEBUG_KMS("surface width(%d), height(%d) and bpp(%d)\n",
                        sizes->surface_width, sizes->surface_height,
                        sizes->surface_bpp);
@@ -220,7 +220,7 @@ static int drm_fbdev_cma_create(struct drm_fb_helper *helper,
        size = mode_cmd.pitches[0] * mode_cmd.height;
        obj = drm_gem_cma_create(dev, size);
-        if (!obj)
+        if (IS_ERR(obj))
                return -ENOMEM;
        fbi = framebuffer_alloc(0, dev->dev);
diff --git a/drivers/gpu/drm/drm_info.c b/drivers/gpu/drm/drm_info.c
index cdf8b1e7602d..d4b20ceda3fb 100644
--- a/drivers/gpu/drm/drm_info.c
+++ b/drivers/gpu/drm/drm_info.c
@@ -205,8 +205,6 @@ static int drm_gem_one_name_info(int id, void *ptr, void *data)
        struct drm_gem_object *obj = ptr;
        struct seq_file *m = data;
-        seq_printf(m, "name %d size %zd\n", obj->name, obj->size);
        seq_printf(m, "%6d %8zd %7d %8d\n",
                   obj->name, obj->size,
                   atomic_read(&obj->handle_count),
@@ -239,7 +237,7 @@ int drm_vma_info(struct seq_file *m, void *data)
        mutex_lock(&dev->struct_mutex);
        seq_printf(m, "vma use count: %d, high_memory = %pK, 0x%pK\n",
                   atomic_read(&dev->vma_count),
-                   high_memory, (void *)virt_to_phys(high_memory));
+                   high_memory, (void *)(unsigned long)virt_to_phys(high_memory));
        list_for_each_entry(pt, &dev->vmalist, head) {
                vma = pt->vma;
diff --git a/drivers/gpu/drm/drm_platform.c b/drivers/gpu/drm/drm_platform.c
index aaeb6f8d69ce..b8a282ea8751 100644
--- a/drivers/gpu/drm/drm_platform.c
+++ b/drivers/gpu/drm/drm_platform.c
@@ -64,7 +64,6 @@ int drm_get_platform_dev(struct platform_device *platdev,
        }
        if (drm_core_check_feature(dev, DRIVER_MODESET)) {
-                dev_set_drvdata(&platdev->dev, dev);
                ret = drm_get_minor(dev, &dev->control, DRM_MINOR_CONTROL);
                if (ret)
                        goto err_g1;
diff --git a/drivers/gpu/drm/exynos/Kconfig b/drivers/gpu/drm/exynos/Kconfig
index 59a26e577b57..fc345d4ebb03 100644
--- a/drivers/gpu/drm/exynos/Kconfig
+++ b/drivers/gpu/drm/exynos/Kconfig
@@ -1,6 +1,6 @@
 config DRM_EXYNOS
        tristate "DRM Support for Samsung SoC EXYNOS Series"
-        depends on DRM && PLAT_SAMSUNG
+        depends on DRM && (PLAT_SAMSUNG || ARCH_MULTIPLATFORM)
        select DRM_KMS_HELPER
        select FB_CFB_FILLRECT
        select FB_CFB_COPYAREA
diff --git a/drivers/gpu/drm/exynos/exynos_drm_connector.c b/drivers/gpu/drm/exynos/exynos_drm_connector.c
index 18c271862ca8..0f68a2872673 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_connector.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_connector.c
@@ -374,6 +374,7 @@ struct drm_connector *exynos_drm_connector_create(struct drm_device *dev,
        exynos_connector->encoder_id = encoder->base.id;
        exynos_connector->manager = manager;
        exynos_connector->dpms = DRM_MODE_DPMS_OFF;
+        connector->dpms = DRM_MODE_DPMS_OFF;
        connector->encoder = encoder;
        err = drm_mode_connector_attach_encoder(connector, encoder);
diff --git a/drivers/gpu/drm/exynos/exynos_drm_encoder.c b/drivers/gpu/drm/exynos/exynos_drm_encoder.c
index e51503fbaf2b..241ad1eeec64 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_encoder.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_encoder.c
@@ -43,12 +43,14 @@
 * @manager: specific encoder has its own manager to control a hardware
 *      appropriately and we can access a hardware drawing on this manager.
 * @dpms: store the encoder dpms value.
+ * @updated: indicate whether overlay data updating is needed or not.
 */
 struct exynos_drm_encoder {
        struct drm_crtc                 *old_crtc;
        struct drm_encoder              drm_encoder;
        struct exynos_drm_manager       *manager;
-        int dpms;
+        int                             dpms;
+        bool                            updated;
 };
 static void exynos_drm_connector_power(struct drm_encoder *encoder, int mode)
@@ -85,7 +87,9 @@ static void exynos_drm_encoder_dpms(struct drm_encoder *encoder, int mode)
        switch (mode) {
        case DRM_MODE_DPMS_ON:
                if (manager_ops && manager_ops->apply)
-                        manager_ops->apply(manager->dev);
+                        if (!exynos_encoder->updated)
+                                manager_ops->apply(manager->dev);
                exynos_drm_connector_power(encoder, mode);
                exynos_encoder->dpms = mode;
                break;
@@ -94,6 +98,7 @@ static void exynos_drm_encoder_dpms(struct drm_encoder *encoder, int mode)
        case DRM_MODE_DPMS_OFF:
                exynos_drm_connector_power(encoder, mode);
                exynos_encoder->dpms = mode;
+                exynos_encoder->updated = false;
                break;
        default:
                DRM_ERROR("unspecified mode %d\n", mode);
@@ -205,13 +210,22 @@ static void exynos_drm_encoder_prepare(struct drm_encoder *encoder)
 static void exynos_drm_encoder_commit(struct drm_encoder *encoder)
 {
-        struct exynos_drm_manager *manager = exynos_drm_get_manager(encoder);
+        struct exynos_drm_encoder *exynos_encoder = to_exynos_encoder(encoder);
+        struct exynos_drm_manager *manager = exynos_encoder->manager;
        struct exynos_drm_manager_ops *manager_ops = manager->ops;
        DRM_DEBUG_KMS("%s\n", __FILE__);
        if (manager_ops && manager_ops->commit)
                manager_ops->commit(manager->dev);
+        /*
+         * this will avoid one issue that overlay data is updated to
+         * real hardware two times.
+         * And this variable will be used to check if the data was
+         * already updated or not by exynos_drm_encoder_dpms function.
+         */
+        exynos_encoder->updated = true;
 }
 static void exynos_drm_encoder_disable(struct drm_encoder *encoder)
@@ -401,19 +415,6 @@ void exynos_drm_encoder_crtc_dpms(struct drm_encoder *encoder, void *data)
                manager_ops->dpms(manager->dev, mode);
        /*
-         * set current mode to new one so that data aren't updated into
-         * registers by drm_helper_connector_dpms two times.
-         *
-         * in case that drm_crtc_helper_set_mode() is called,
-         * overlay_ops->commit() and manager_ops->commit() callbacks
-         * can be called two times, first at drm_crtc_helper_set_mode()
-         * and second at drm_helper_connector_dpms().
-         * so with this setting, when drm_helper_connector_dpms() is called
-         * encoder->funcs->dpms() will be ignored.
-         */
-        exynos_encoder->dpms = mode;
-        /*
         * if this condition is ok then it means that the crtc is already
         * detached from encoder and last function for detaching is properly
         * done, so clear pipe from manager to prevent repeated call.
diff --git a/drivers/gpu/drm/exynos/exynos_mixer.c b/drivers/gpu/drm/exynos/exynos_mixer.c
index 614b2e9ac462..e7fbb823fd8e 100644
--- a/drivers/gpu/drm/exynos/exynos_mixer.c
+++ b/drivers/gpu/drm/exynos/exynos_mixer.c
@@ -1142,7 +1142,7 @@ static int __devinit mixer_probe(struct platform_device *pdev)
                const struct of_device_id *match;
                match = of_match_node(of_match_ptr(mixer_match_types),
                                                          pdev->dev.of_node);
-                drv = match->data;
+                drv = (struct mixer_drv_data *)match->data;
        } else {
                drv = (struct mixer_drv_data *)
                        platform_get_device_id(pdev)->driver_data;
diff --git a/drivers/gpu/drm/i915/dvo_ch7xxx.c b/drivers/gpu/drm/i915/dvo_ch7xxx.c
index 38f3a6cb8c7d..3edd981e0770 100644
--- a/drivers/gpu/drm/i915/dvo_ch7xxx.c
+++ b/drivers/gpu/drm/i915/dvo_ch7xxx.c
@@ -303,10 +303,10 @@ static bool ch7xxx_get_hw_state(struct intel_dvo_device *dvo)
        ch7xxx_readb(dvo, CH7xxx_PM, &val);
-        if (val & CH7xxx_PM_FPD)
+        if (val & (CH7xxx_PM_DVIL | CH7xxx_PM_DVIP))
-                return false;
-        else
                return true;
+        else
+                return false;
 }
 static void ch7xxx_dump_regs(struct intel_dvo_device *dvo)
diff --git a/drivers/gpu/drm/i915/i915_dma.c b/drivers/gpu/drm/i915/i915_dma.c
index c9bfd83dde64..61ae104dca8c 100644
--- a/drivers/gpu/drm/i915/i915_dma.c
+++ b/drivers/gpu/drm/i915/i915_dma.c
@@ -1505,7 +1505,8 @@ int i915_driver_load(struct drm_device *dev, unsigned long flags)
                goto put_gmch;
        }
-        i915_kick_out_firmware_fb(dev_priv);
+        if (drm_core_check_feature(dev, DRIVER_MODESET))
+                i915_kick_out_firmware_fb(dev_priv);
        pci_set_master(dev->pdev);
diff --git a/drivers/gpu/drm/i915/i915_drv.c b/drivers/gpu/drm/i915/i915_drv.c
index aac4e5e1a5b9..6770ee6084b4 100644
--- a/drivers/gpu/drm/i915/i915_drv.c
+++ b/drivers/gpu/drm/i915/i915_drv.c
@@ -118,6 +118,13 @@ module_param_named(i915_enable_ppgtt, i915_enable_ppgtt, int, 0600);
 MODULE_PARM_DESC(i915_enable_ppgtt,
                "Enable PPGTT (default: true)");
+unsigned int i915_preliminary_hw_support __read_mostly = 0;
+module_param_named(preliminary_hw_support, i915_preliminary_hw_support, int, 0600);
+MODULE_PARM_DESC(preliminary_hw_support,
+                "Enable preliminary hardware support. "
+                "Enable Haswell and ValleyView Support. "
+                "(default: false)");
 static struct drm_driver driver;
 extern int intel_agp_enabled;
@@ -826,6 +833,12 @@ i915_pci_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
        struct intel_device_info *intel_info =
                (struct intel_device_info *) ent->driver_data;
+        if (intel_info->is_haswell || intel_info->is_valleyview)
+                if(!i915_preliminary_hw_support) {
+                        DRM_ERROR("Preliminary hardware support disabled\n");
+                        return -ENODEV;
+                }
        /* Only bind to function 0 of the device. Early generations
         * used function 1 as a placeholder for multi-head. This causes
         * us confusion instead, especially on the systems where both
diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index 4f2831aa5fed..f511fa2f4168 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -1217,6 +1217,7 @@ extern int i915_enable_rc6 __read_mostly;
 extern int i915_enable_fbc __read_mostly;
 extern bool i915_enable_hangcheck __read_mostly;
 extern int i915_enable_ppgtt __read_mostly;
+extern unsigned int i915_preliminary_hw_support __read_mostly;
 extern int i915_suspend(struct drm_device *dev, pm_message_t state);
 extern int i915_resume(struct drm_device *dev);
@@ -1341,9 +1342,14 @@ int __must_check i915_gem_object_get_pages(struct drm_i915_gem_object *obj);
 static inline struct page *i915_gem_object_get_page(struct drm_i915_gem_object *obj, int n)
 {
        struct scatterlist *sg = obj->pages->sgl;
-        while (n >= SG_MAX_SINGLE_ALLOC) {
+        int nents = obj->pages->nents;
+        while (nents > SG_MAX_SINGLE_ALLOC) {
+                if (n < SG_MAX_SINGLE_ALLOC - 1)
+                        break;
                sg = sg_chain_ptr(sg + SG_MAX_SINGLE_ALLOC - 1);
                n -= SG_MAX_SINGLE_ALLOC - 1;
+                nents -= SG_MAX_SINGLE_ALLOC - 1;
        }
        return sg_page(sg+n);
 }
@@ -1427,7 +1433,7 @@ int __must_check i915_gpu_idle(struct drm_device *dev);
 int __must_check i915_gem_idle(struct drm_device *dev);
 int i915_add_request(struct intel_ring_buffer *ring,
                     struct drm_file *file,
-                     struct drm_i915_gem_request *request);
+                     u32 *seqno);
 int __must_check i915_wait_seqno(struct intel_ring_buffer *ring,
                                 uint32_t seqno);
 int i915_gem_fault(struct vm_area_struct *vma, struct vm_fault *vmf);
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index 19dbdd7dd564..107f09befe92 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -1407,8 +1407,10 @@ out:
                return VM_FAULT_NOPAGE;
        case -ENOMEM:
                return VM_FAULT_OOM;
+        case -ENOSPC:
+                return VM_FAULT_SIGBUS;
        default:
-                WARN_ON_ONCE(ret);
+                WARN_ONCE(ret, "unhandled error in i915_gem_fault: %i\n", ret);
                return VM_FAULT_SIGBUS;
        }
 }
@@ -1822,10 +1824,11 @@ i915_gem_object_get_pages_gtt(struct drm_i915_gem_object *obj)
                sg_set_page(sg, page, PAGE_SIZE, 0);
        }
+        obj->pages = st;
        if (i915_gem_object_needs_bit17_swizzle(obj))
                i915_gem_object_do_bit_17_swizzle(obj);
-        obj->pages = st;
        return 0;
 err_pages:
@@ -1955,11 +1958,12 @@ i915_gem_next_request_seqno(struct intel_ring_buffer *ring)
 int
 i915_add_request(struct intel_ring_buffer *ring,
                 struct drm_file *file,
-                 struct drm_i915_gem_request *request)
+                 u32 *out_seqno)
 {
        drm_i915_private_t *dev_priv = ring->dev->dev_private;
-        uint32_t seqno;
+        struct drm_i915_gem_request *request;
        u32 request_ring_position;
+        u32 seqno;
        int was_empty;
        int ret;
@@ -1974,11 +1978,9 @@ i915_add_request(struct intel_ring_buffer *ring,
        if (ret)
                return ret;
-        if (request == NULL) {
+        request = kmalloc(sizeof(*request), GFP_KERNEL);
-                request = kmalloc(sizeof(*request), GFP_KERNEL);
+        if (request == NULL)
-                if (request == NULL)
+                return -ENOMEM;
-                        return -ENOMEM;
-        }
        seqno = i915_gem_next_request_seqno(ring);
@@ -2030,6 +2032,8 @@ i915_add_request(struct intel_ring_buffer *ring,
                }
        }
+        if (out_seqno)
+                *out_seqno = seqno;
        return 0;
 }
@@ -3959,6 +3963,9 @@ i915_gem_init_hw(struct drm_device *dev)
        if (!intel_enable_gtt())
                return -EIO;
+        if (IS_HASWELL(dev) && (I915_READ(0x120010) == 1))
+                I915_WRITE(0x9008, I915_READ(0x9008) | 0xf0000);
        i915_gem_l3_remap(dev);
        i915_gem_init_swizzling(dev);
@@ -4098,7 +4105,6 @@ i915_gem_entervt_ioctl(struct drm_device *dev, void *data,
        }
        BUG_ON(!list_empty(&dev_priv->mm.active_list));
-        BUG_ON(!list_empty(&dev_priv->mm.inactive_list));
        mutex_unlock(&dev->struct_mutex);
        ret = drm_irq_install(dev);
diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h
index 64c1be0a9cfd..a4162ddff6c5 100644
--- a/drivers/gpu/drm/i915/i915_reg.h
+++ b/drivers/gpu/drm/i915/i915_reg.h
@@ -521,7 +521,7 @@
 */
 # define _3D_CHICKEN2_WM_READ_PIPELINED                 (1 << 14)
 #define _3D_CHICKEN3    0x02090
-#define  _3D_CHICKEN_SF_DISABLE_FASTCLIP_CULL           (1 << 5)
+#define  _3D_CHICKEN3_SF_DISABLE_FASTCLIP_CULL          (1 << 5)
 #define MI_MODE         0x0209c
 # define VS_TIMER_DISPATCH                              (1 << 6)
diff --git a/drivers/gpu/drm/i915/intel_crt.c b/drivers/gpu/drm/i915/intel_crt.c
index 893f30164b7e..b726b478a4f5 100644
--- a/drivers/gpu/drm/i915/intel_crt.c
+++ b/drivers/gpu/drm/i915/intel_crt.c
@@ -219,20 +219,7 @@ static void intel_crt_mode_set(struct drm_encoder *encoder,
                intel_encoder_to_crt(to_intel_encoder(encoder));
        struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
        struct drm_i915_private *dev_priv = dev->dev_private;
-        int dpll_md_reg;
+        u32 adpa;
-        u32 adpa, dpll_md;
-        dpll_md_reg = DPLL_MD(intel_crtc->pipe);
-        /*
-         * Disable separate mode multiplier used when cloning SDVO to CRT
-         * XXX this needs to be adjusted when we really are cloning
-         */
-        if (INTEL_INFO(dev)->gen >= 4 && !HAS_PCH_SPLIT(dev)) {
-                dpll_md = I915_READ(dpll_md_reg);
-                I915_WRITE(dpll_md_reg,
-                           dpll_md & ~DPLL_MD_UDI_MULTIPLIER_MASK);
-        }
        adpa = ADPA_HOTPLUG_BITS;
        if (adjusted_mode->flags & DRM_MODE_FLAG_PHSYNC)
@@ -742,7 +729,7 @@ void intel_crt_init(struct drm_device *dev)
        crt->base.type = INTEL_OUTPUT_ANALOG;
        crt->base.cloneable = true;
-        if (IS_HASWELL(dev))
+        if (IS_HASWELL(dev) || IS_I830(dev))
                crt->base.crtc_mask = (1 << 0);
        else
                crt->base.crtc_mask = (1 << 0) | (1 << 1) | (1 << 2);
diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index 2b6ce9b2674a..461a637f1ef7 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -3253,6 +3253,16 @@ static void ironlake_crtc_enable(struct drm_crtc *crtc)
        if (HAS_PCH_CPT(dev))
                intel_cpt_verify_modeset(dev, intel_crtc->pipe);
+        /*
+         * There seems to be a race in PCH platform hw (at least on some
+         * outputs) where an enabled pipe still completes any pageflip right
+         * away (as if the pipe is off) instead of waiting for vblank. As soon
+         * as the first vblank happend, everything works as expected. Hence just
+         * wait for one vblank before returning to avoid strange things
+         * happening.
+         */
+        intel_wait_for_vblank(dev, intel_crtc->pipe);
 }
 static void ironlake_crtc_disable(struct drm_crtc *crtc)
@@ -7882,6 +7892,34 @@ struct intel_quirk {
        void (*hook)(struct drm_device *dev);
 };
+/* For systems that don't have a meaningful PCI subdevice/subvendor ID */
+struct intel_dmi_quirk {
+        void (*hook)(struct drm_device *dev);
+        const struct dmi_system_id (*dmi_id_list)[];
+};
+static int intel_dmi_reverse_brightness(const struct dmi_system_id *id)
+{
+        DRM_INFO("Backlight polarity reversed on %s\n", id->ident);
+        return 1;
+}
+static const struct intel_dmi_quirk intel_dmi_quirks[] = {
+        {
+                .dmi_id_list = &(const struct dmi_system_id[]) {
+                        {
+                                .callback = intel_dmi_reverse_brightness,
+                                .ident = "NCR Corporation",
+                                .matches = {DMI_MATCH(DMI_SYS_VENDOR, "NCR Corporation"),
+                                            DMI_MATCH(DMI_PRODUCT_NAME, ""),
+                                },
+                        },
+                        { }  /* terminating entry */
+                },
+                .hook = quirk_invert_brightness,
+        },
+};
 static struct intel_quirk intel_quirks[] = {
        /* HP Mini needs pipe A force quirk (LP: #322104) */
        { 0x27ae, 0x103c, 0x361a, quirk_pipea_force },
@@ -7892,8 +7930,7 @@ static struct intel_quirk intel_quirks[] = {
        /* ThinkPad T60 needs pipe A force quirk (bug #16494) */
        { 0x2782, 0x17aa, 0x201a, quirk_pipea_force },
-        /* 855 & before need to leave pipe A & dpll A up */
+        /* 830/845 need to leave pipe A & dpll A up */
-        { 0x3582, PCI_ANY_ID, PCI_ANY_ID, quirk_pipea_force },
        { 0x2562, PCI_ANY_ID, PCI_ANY_ID, quirk_pipea_force },
        { 0x3577, PCI_ANY_ID, PCI_ANY_ID, quirk_pipea_force },
@@ -7922,6 +7959,10 @@ static void intel_init_quirks(struct drm_device *dev)
                     q->subsystem_device == PCI_ANY_ID))
                        q->hook(dev);
        }
+        for (i = 0; i < ARRAY_SIZE(intel_dmi_quirks); i++) {
+                if (dmi_check_system(*intel_dmi_quirks[i].dmi_id_list) != 0)
+                        intel_dmi_quirks[i].hook(dev);
+        }
 }
 /* Disable the VGA plane that we never use */
@@ -8049,29 +8090,42 @@ static void intel_enable_pipe_a(struct drm_device *dev)
 }
+static bool
+intel_check_plane_mapping(struct intel_crtc *crtc)
+{
+        struct drm_i915_private *dev_priv = crtc->base.dev->dev_private;
+        u32 reg, val;
+        if (dev_priv->num_pipe == 1)
+                return true;
+        reg = DSPCNTR(!crtc->plane);
+        val = I915_READ(reg);
+        if ((val & DISPLAY_PLANE_ENABLE) &&
+            (!!(val & DISPPLANE_SEL_PIPE_MASK) == crtc->pipe))
+                return false;
+        return true;
+}
 static void intel_sanitize_crtc(struct intel_crtc *crtc)
 {
        struct drm_device *dev = crtc->base.dev;
        struct drm_i915_private *dev_priv = dev->dev_private;
-        u32 reg, val;
+        u32 reg;
        /* Clear any frame start delays used for debugging left by the BIOS */
        reg = PIPECONF(crtc->pipe);
        I915_WRITE(reg, I915_READ(reg) & ~PIPECONF_FRAME_START_DELAY_MASK);
        /* We need to sanitize the plane -> pipe mapping first because this will
-         * disable the crtc (and hence change the state) if it is wrong. */
+         * disable the crtc (and hence change the state) if it is wrong. Note
-        if (!HAS_PCH_SPLIT(dev)) {
+         * that gen4+ has a fixed plane -> pipe mapping.  */
+        if (INTEL_INFO(dev)->gen < 4 && !intel_check_plane_mapping(crtc)) {
                struct intel_connector *connector;
                bool plane;
-                reg = DSPCNTR(crtc->plane);
-                val = I915_READ(reg);
-                if ((val & DISPLAY_PLANE_ENABLE) == 0 &&
-                    (!!(val & DISPPLANE_SEL_PIPE_MASK) == crtc->pipe))
-                        goto ok;
                DRM_DEBUG_KMS("[CRTC:%d] wrong plane connection detected!\n",
                              crtc->base.base.id);
@@ -8095,7 +8149,6 @@ static void intel_sanitize_crtc(struct intel_crtc *crtc)
                WARN_ON(crtc->active);
                crtc->base.enabled = false;
        }
-ok:
        if (dev_priv->quirks & QUIRK_PIPEA_FORCE &&
            crtc->pipe == PIPE_A && !crtc->active) {
diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c
index d1e8ddb2d6c0..368ed8ef1600 100644
--- a/drivers/gpu/drm/i915/intel_dp.c
+++ b/drivers/gpu/drm/i915/intel_dp.c
@@ -1797,7 +1797,8 @@ intel_dp_start_link_train(struct intel_dp *intel_dp)
                        if ((intel_dp->train_set[i] & DP_TRAIN_MAX_SWING_REACHED) == 0)
                                break;
                if (i == intel_dp->lane_count && voltage_tries == 5) {
-                        if (++loop_tries == 5) {
+                        ++loop_tries;
+                        if (loop_tries == 5) {
                                DRM_DEBUG_KMS("too many full retries, give up\n");
                                break;
                        }
@@ -1807,11 +1808,15 @@ intel_dp_start_link_train(struct intel_dp *intel_dp)
                }
                /* Check to see if we've tried the same voltage 5 times */
-                if ((intel_dp->train_set[0] & DP_TRAIN_VOLTAGE_SWING_MASK) != voltage) {
+                if ((intel_dp->train_set[0] & DP_TRAIN_VOLTAGE_SWING_MASK) == voltage) {
-                        voltage = intel_dp->train_set[0] & DP_TRAIN_VOLTAGE_SWING_MASK;
-                        voltage_tries = 0;
-                } else
                        ++voltage_tries;
+                        if (voltage_tries == 5) {
+                                DRM_DEBUG_KMS("too many voltage retries, give up\n");
+                                break;
+                        }
+                } else
+                        voltage_tries = 0;
+                voltage = intel_dp->train_set[0] & DP_TRAIN_VOLTAGE_SWING_MASK;
                /* Compute new intel_dp->train_set as requested by target */
                intel_get_adjust_train(intel_dp, link_status);
@@ -2369,8 +2374,9 @@ static void
 intel_dp_destroy(struct drm_connector *connector)
 {
        struct drm_device *dev = connector->dev;
+        struct intel_dp *intel_dp = intel_attached_dp(connector);
-        if (intel_dpd_is_edp(dev))
+        if (is_edp(intel_dp))
                intel_panel_destroy_backlight(dev);
        drm_sysfs_connector_remove(connector);
diff --git a/drivers/gpu/drm/i915/intel_lvds.c b/drivers/gpu/drm/i915/intel_lvds.c
index e3166df55daa..edba93b3474b 100644
--- a/drivers/gpu/drm/i915/intel_lvds.c
+++ b/drivers/gpu/drm/i915/intel_lvds.c
@@ -777,6 +777,14 @@ static const struct dmi_system_id intel_no_lvds[] = {
                        DMI_MATCH(DMI_BOARD_NAME, "D525TUD"),
                },
        },
+        {
+                .callback = intel_no_lvds_dmi_callback,
+                .ident = "Supermicro X7SPA-H",
+                .matches = {
+                        DMI_MATCH(DMI_SYS_VENDOR, "Supermicro"),
+                        DMI_MATCH(DMI_PRODUCT_NAME, "X7SPA-H"),
+                },
+        },
        { }     /* terminating entry */
 };
diff --git a/drivers/gpu/drm/i915/intel_overlay.c b/drivers/gpu/drm/i915/intel_overlay.c
index ebff850a9ab6..d7bc817f51a0 100644
--- a/drivers/gpu/drm/i915/intel_overlay.c
+++ b/drivers/gpu/drm/i915/intel_overlay.c
@@ -209,7 +209,6 @@ static void intel_overlay_unmap_regs(struct intel_overlay *overlay,
 }
 static int intel_overlay_do_wait_request(struct intel_overlay *overlay,
-                                         struct drm_i915_gem_request *request,
                                         void (*tail)(struct intel_overlay *))
 {
        struct drm_device *dev = overlay->dev;
@@ -218,12 +217,10 @@ static int intel_overlay_do_wait_request(struct intel_overlay *overlay,
        int ret;
        BUG_ON(overlay->last_flip_req);
-        ret = i915_add_request(ring, NULL, request);
+        ret = i915_add_request(ring, NULL, &overlay->last_flip_req);
-        if (ret) {
+        if (ret)
-            kfree(request);
+                return ret;
-            return ret;
-        }
-        overlay->last_flip_req = request->seqno;
        overlay->flip_tail = tail;
        ret = i915_wait_seqno(ring, overlay->last_flip_req);
        if (ret)
@@ -240,7 +237,6 @@ static int intel_overlay_on(struct intel_overlay *overlay)
        struct drm_device *dev = overlay->dev;
        struct drm_i915_private *dev_priv = dev->dev_private;
        struct intel_ring_buffer *ring = &dev_priv->ring[RCS];
-        struct drm_i915_gem_request *request;
        int ret;
        BUG_ON(overlay->active);
@@ -248,17 +244,9 @@ static int intel_overlay_on(struct intel_overlay *overlay)
        WARN_ON(IS_I830(dev) && !(dev_priv->quirks & QUIRK_PIPEA_FORCE));
-        request = kzalloc(sizeof(*request), GFP_KERNEL);
-        if (request == NULL) {
-                ret = -ENOMEM;
-                goto out;
-        }
        ret = intel_ring_begin(ring, 4);
-        if (ret) {
+        if (ret)
-                kfree(request);
+                return ret;
-                goto out;
-        }
        intel_ring_emit(ring, MI_OVERLAY_FLIP | MI_OVERLAY_ON);
        intel_ring_emit(ring, overlay->flip_addr | OFC_UPDATE);
@@ -266,9 +254,7 @@ static int intel_overlay_on(struct intel_overlay *overlay)
        intel_ring_emit(ring, MI_NOOP);
        intel_ring_advance(ring);
-        ret = intel_overlay_do_wait_request(overlay, request, NULL);
+        return intel_overlay_do_wait_request(overlay, NULL);
-out:
-        return ret;
 }
 /* overlay needs to be enabled in OCMD reg */
@@ -278,17 +264,12 @@ static int intel_overlay_continue(struct intel_overlay *overlay,
        struct drm_device *dev = overlay->dev;
        drm_i915_private_t *dev_priv = dev->dev_private;
        struct intel_ring_buffer *ring = &dev_priv->ring[RCS];
-        struct drm_i915_gem_request *request;
        u32 flip_addr = overlay->flip_addr;
        u32 tmp;
        int ret;
        BUG_ON(!overlay->active);
-        request = kzalloc(sizeof(*request), GFP_KERNEL);
-        if (request == NULL)
-                return -ENOMEM;
        if (load_polyphase_filter)
                flip_addr |= OFC_UPDATE;
@@ -298,22 +279,14 @@ static int intel_overlay_continue(struct intel_overlay *overlay,
                DRM_DEBUG("overlay underrun, DOVSTA: %x\n", tmp);
        ret = intel_ring_begin(ring, 2);
-        if (ret) {
+        if (ret)
-                kfree(request);
                return ret;
-        }
        intel_ring_emit(ring, MI_OVERLAY_FLIP | MI_OVERLAY_CONTINUE);
        intel_ring_emit(ring, flip_addr);
        intel_ring_advance(ring);
-        ret = i915_add_request(ring, NULL, request);
+        return i915_add_request(ring, NULL, &overlay->last_flip_req);
-        if (ret) {
-                kfree(request);
-                return ret;
-        }
-        overlay->last_flip_req = request->seqno;
-        return 0;
 }
 static void intel_overlay_release_old_vid_tail(struct intel_overlay *overlay)
@@ -349,15 +322,10 @@ static int intel_overlay_off(struct intel_overlay *overlay)
        struct drm_i915_private *dev_priv = dev->dev_private;
        struct intel_ring_buffer *ring = &dev_priv->ring[RCS];
        u32 flip_addr = overlay->flip_addr;
-        struct drm_i915_gem_request *request;
        int ret;
        BUG_ON(!overlay->active);
-        request = kzalloc(sizeof(*request), GFP_KERNEL);
-        if (request == NULL)
-                return -ENOMEM;
        /* According to intel docs the overlay hw may hang (when switching
         * off) without loading the filter coeffs. It is however unclear whether
         * this applies to the disabling of the overlay or to the switching off
@@ -365,22 +333,28 @@ static int intel_overlay_off(struct intel_overlay *overlay)
        flip_addr |= OFC_UPDATE;
        ret = intel_ring_begin(ring, 6);
-        if (ret) {
+        if (ret)
-                kfree(request);
                return ret;
-        }
        /* wait for overlay to go idle */
        intel_ring_emit(ring, MI_OVERLAY_FLIP | MI_OVERLAY_CONTINUE);
        intel_ring_emit(ring, flip_addr);
        intel_ring_emit(ring, MI_WAIT_FOR_EVENT | MI_WAIT_FOR_OVERLAY_FLIP);
        /* turn overlay off */
-        intel_ring_emit(ring, MI_OVERLAY_FLIP | MI_OVERLAY_OFF);
+        if (IS_I830(dev)) {
-        intel_ring_emit(ring, flip_addr);
+                /* Workaround: Don't disable the overlay fully, since otherwise
-        intel_ring_emit(ring, MI_WAIT_FOR_EVENT | MI_WAIT_FOR_OVERLAY_FLIP);
+                 * it dies on the next OVERLAY_ON cmd. */
+                intel_ring_emit(ring, MI_NOOP);
+                intel_ring_emit(ring, MI_NOOP);
+                intel_ring_emit(ring, MI_NOOP);
+        } else {
+                intel_ring_emit(ring, MI_OVERLAY_FLIP | MI_OVERLAY_OFF);
+                intel_ring_emit(ring, flip_addr);
+                intel_ring_emit(ring, MI_WAIT_FOR_EVENT | MI_WAIT_FOR_OVERLAY_FLIP);
+        }
        intel_ring_advance(ring);
-        return intel_overlay_do_wait_request(overlay, request,
+        return intel_overlay_do_wait_request(overlay, intel_overlay_off_tail);
-                                             intel_overlay_off_tail);
 }
 /* recover from an interruption due to a signal
@@ -425,24 +399,16 @@ static int intel_overlay_release_old_vid(struct intel_overlay *overlay)
                return 0;
        if (I915_READ(ISR) & I915_OVERLAY_PLANE_FLIP_PENDING_INTERRUPT) {
-                struct drm_i915_gem_request *request;
                /* synchronous slowpath */
-                request = kzalloc(sizeof(*request), GFP_KERNEL);
-                if (request == NULL)
-                        return -ENOMEM;
                ret = intel_ring_begin(ring, 2);
-                if (ret) {
+                if (ret)
-                        kfree(request);
                        return ret;
-                }
                intel_ring_emit(ring, MI_WAIT_FOR_EVENT | MI_WAIT_FOR_OVERLAY_FLIP);
                intel_ring_emit(ring, MI_NOOP);
                intel_ring_advance(ring);
-                ret = intel_overlay_do_wait_request(overlay, request,
+                ret = intel_overlay_do_wait_request(overlay,
                                                    intel_overlay_release_old_vid_tail);
                if (ret)
                        return ret;
diff --git a/drivers/gpu/drm/i915/intel_panel.c b/drivers/gpu/drm/i915/intel_panel.c
index e019b2369861..e2aacd329545 100644
--- a/drivers/gpu/drm/i915/intel_panel.c
+++ b/drivers/gpu/drm/i915/intel_panel.c
@@ -435,7 +435,7 @@ int intel_panel_setup_backlight(struct drm_device *dev)
        props.type = BACKLIGHT_RAW;
        props.max_brightness = _intel_panel_get_max_backlight(dev);
        if (props.max_brightness == 0) {
-                DRM_ERROR("Failed to get maximum backlight value\n");
+                DRM_DEBUG_DRIVER("Failed to get maximum backlight value\n");
                return -ENODEV;
        }
        dev_priv->backlight =
diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c
index b3b4b6cea8b0..72f41aaa71ff 100644
--- a/drivers/gpu/drm/i915/intel_pm.c
+++ b/drivers/gpu/drm/i915/intel_pm.c
@@ -3442,8 +3442,8 @@ static void gen6_init_clock_gating(struct drm_device *dev)
                   GEN6_RCCUNIT_CLOCK_GATE_DISABLE);
        /* Bspec says we need to always set all mask bits. */
-        I915_WRITE(_3D_CHICKEN, (0xFFFF << 16) |
+        I915_WRITE(_3D_CHICKEN3, (0xFFFF << 16) |
-                   _3D_CHICKEN_SF_DISABLE_FASTCLIP_CULL);
+                   _3D_CHICKEN3_SF_DISABLE_FASTCLIP_CULL);
        /*
         * According to the spec the following bits should be
diff --git a/drivers/gpu/drm/i915/intel_sdvo.c b/drivers/gpu/drm/i915/intel_sdvo.c
index 0007a4d9bf6e..79d308da29ff 100644
--- a/drivers/gpu/drm/i915/intel_sdvo.c
+++ b/drivers/gpu/drm/i915/intel_sdvo.c
@@ -139,6 +139,11 @@ struct intel_sdvo {
        /* DDC bus used by this SDVO encoder */
        uint8_t ddc_bus;
+        /*
+         * the sdvo flag gets lost in round trip: dtd->adjusted_mode->dtd
+         */
+        uint8_t dtd_sdvo_flags;
 };
 struct intel_sdvo_connector {
@@ -889,6 +894,45 @@ static void intel_sdvo_dump_hdmi_buf(struct intel_sdvo *intel_sdvo)
 }
 #endif
+static bool intel_sdvo_write_infoframe(struct intel_sdvo *intel_sdvo,
+                                       unsigned if_index, uint8_t tx_rate,
+                                       uint8_t *data, unsigned length)
+{
+        uint8_t set_buf_index[2] = { if_index, 0 };
+        uint8_t hbuf_size, tmp[8];
+        int i;
+        if (!intel_sdvo_set_value(intel_sdvo,
+                                  SDVO_CMD_SET_HBUF_INDEX,
+                                  set_buf_index, 2))
+                return false;
+        if (!intel_sdvo_get_value(intel_sdvo, SDVO_CMD_GET_HBUF_INFO,
+                                  &hbuf_size, 1))
+                return false;
+        /* Buffer size is 0 based, hooray! */
+        hbuf_size++;
+        DRM_DEBUG_KMS("writing sdvo hbuf: %i, hbuf_size %i, hbuf_size: %i\n",
+                      if_index, length, hbuf_size);
+        for (i = 0; i < hbuf_size; i += 8) {
+                memset(tmp, 0, 8);
+                if (i < length)
+                        memcpy(tmp, data + i, min_t(unsigned, 8, length - i));
+                if (!intel_sdvo_set_value(intel_sdvo,
+                                          SDVO_CMD_SET_HBUF_DATA,
+                                          tmp, 8))
+                        return false;
+        }
+        return intel_sdvo_set_value(intel_sdvo,
+                                    SDVO_CMD_SET_HBUF_TXRATE,
+                                    &tx_rate, 1);
+}
 static bool intel_sdvo_set_avi_infoframe(struct intel_sdvo *intel_sdvo)
 {
        struct dip_infoframe avi_if = {
@@ -896,11 +940,7 @@ static bool intel_sdvo_set_avi_infoframe(struct intel_sdvo *intel_sdvo)
                .ver = DIP_VERSION_AVI,
                .len = DIP_LEN_AVI,
        };
-        uint8_t tx_rate = SDVO_HBUF_TX_VSYNC;
-        uint8_t set_buf_index[2] = { 1, 0 };
        uint8_t sdvo_data[4 + sizeof(avi_if.body.avi)];
-        uint64_t *data = (uint64_t *)sdvo_data;
-        unsigned i;
        intel_dip_infoframe_csum(&avi_if);
@@ -910,22 +950,9 @@ static bool intel_sdvo_set_avi_infoframe(struct intel_sdvo *intel_sdvo)
        sdvo_data[3] = avi_if.checksum;
        memcpy(&sdvo_data[4], &avi_if.body, sizeof(avi_if.body.avi));
-        if (!intel_sdvo_set_value(intel_sdvo,
+        return intel_sdvo_write_infoframe(intel_sdvo, SDVO_HBUF_INDEX_AVI_IF,
-                                  SDVO_CMD_SET_HBUF_INDEX,
+                                          SDVO_HBUF_TX_VSYNC,
-                                  set_buf_index, 2))
+                                          sdvo_data, sizeof(sdvo_data));
-                return false;
-        for (i = 0; i < sizeof(sdvo_data); i += 8) {
-                if (!intel_sdvo_set_value(intel_sdvo,
-                                          SDVO_CMD_SET_HBUF_DATA,
-                                          data, 8))
-                        return false;
-                data++;
-        }
-        return intel_sdvo_set_value(intel_sdvo,
-                                    SDVO_CMD_SET_HBUF_TXRATE,
-                                    &tx_rate, 1);
 }
 static bool intel_sdvo_set_tv_format(struct intel_sdvo *intel_sdvo)
@@ -984,6 +1011,7 @@ intel_sdvo_get_preferred_input_mode(struct intel_sdvo *intel_sdvo,
                return false;
        intel_sdvo_get_mode_from_dtd(adjusted_mode, &input_dtd);
+        intel_sdvo->dtd_sdvo_flags = input_dtd.part2.sdvo_flags;
        return true;
 }
@@ -1092,6 +1120,8 @@ static void intel_sdvo_mode_set(struct drm_encoder *encoder,
         * adjusted_mode.
         */
        intel_sdvo_get_dtd_from_mode(&input_dtd, adjusted_mode);
+        if (intel_sdvo->is_tv || intel_sdvo->is_lvds)
+                input_dtd.part2.sdvo_flags = intel_sdvo->dtd_sdvo_flags;
        if (!intel_sdvo_set_input_timing(intel_sdvo, &input_dtd))
                DRM_INFO("Setting input timings on %s failed\n",
                         SDVO_NAME(intel_sdvo));
@@ -2277,10 +2307,8 @@ intel_sdvo_lvds_init(struct intel_sdvo *intel_sdvo, int device)
                intel_sdvo_connector->output_flag = SDVO_OUTPUT_LVDS1;
        }
-        /* SDVO LVDS is cloneable because the SDVO encoder does the upscaling,
+        /* SDVO LVDS is not cloneable because the input mode gets adjusted by the encoder */
-         * as opposed to native LVDS, where we upscale with the panel-fitter
+        intel_sdvo->base.cloneable = false;
-         * (and hence only the native LVDS resolution could be cloned). */
-        intel_sdvo->base.cloneable = true;
        intel_sdvo_connector_init(intel_sdvo_connector, intel_sdvo);
        if (!intel_sdvo_create_enhance_property(intel_sdvo, intel_sdvo_connector))
diff --git a/drivers/gpu/drm/i915/intel_sdvo_regs.h b/drivers/gpu/drm/i915/intel_sdvo_regs.h
index 9d030142ee43..770bdd6ecd9f 100644
--- a/drivers/gpu/drm/i915/intel_sdvo_regs.h
+++ b/drivers/gpu/drm/i915/intel_sdvo_regs.h
@@ -708,6 +708,8 @@ struct intel_sdvo_enhancements_arg {
 #define SDVO_CMD_SET_AUDIO_STAT         0x91
 #define SDVO_CMD_GET_AUDIO_STAT         0x92
 #define SDVO_CMD_SET_HBUF_INDEX         0x93
+  #define SDVO_HBUF_INDEX_ELD           0
+  #define SDVO_HBUF_INDEX_AVI_IF        1
 #define SDVO_CMD_GET_HBUF_INDEX         0x94
 #define SDVO_CMD_GET_HBUF_INFO          0x95
 #define SDVO_CMD_SET_HBUF_AV_SPLIT      0x96
diff --git a/drivers/gpu/drm/nouveau/core/core/gpuobj.c b/drivers/gpu/drm/nouveau/core/core/gpuobj.c
index 1f34549aff18..70586fde69cf 100644
--- a/drivers/gpu/drm/nouveau/core/core/gpuobj.c
+++ b/drivers/gpu/drm/nouveau/core/core/gpuobj.c
@@ -39,6 +39,11 @@ nouveau_gpuobj_destroy(struct nouveau_gpuobj *gpuobj)
                        nv_wo32(gpuobj, i, 0x00000000);
        }
+        if (gpuobj->node) {
+                nouveau_mm_free(&nv_gpuobj(gpuobj->parent)->heap,
+                                &gpuobj->node);
+        }
        if (gpuobj->heap.block_size)
                nouveau_mm_fini(&gpuobj->heap);
diff --git a/drivers/gpu/drm/nouveau/core/core/mm.c b/drivers/gpu/drm/nouveau/core/core/mm.c
index bfddf87926dd..a6d3cd6490f7 100644
--- a/drivers/gpu/drm/nouveau/core/core/mm.c
+++ b/drivers/gpu/drm/nouveau/core/core/mm.c
@@ -218,13 +218,16 @@ nouveau_mm_init(struct nouveau_mm *mm, u32 offset, u32 length, u32 block)
        node = kzalloc(sizeof(*node), GFP_KERNEL);
        if (!node)
                return -ENOMEM;
-        node->offset = roundup(offset, mm->block_size);
-        node->length = rounddown(offset + length, mm->block_size) - node->offset;
+        if (length) {
+                node->offset  = roundup(offset, mm->block_size);
+                node->length  = rounddown(offset + length, mm->block_size);
+                node->length -= node->offset;
+        }
        list_add_tail(&node->nl_entry, &mm->nodes);
        list_add_tail(&node->fl_entry, &mm->free);
        mm->heap_nodes++;
-        mm->heap_size += length;
        return 0;
 }
@@ -236,7 +239,7 @@ nouveau_mm_fini(struct nouveau_mm *mm)
        int nodes = 0;
        list_for_each_entry(node, &mm->nodes, nl_entry) {
-                if (nodes++ == mm->heap_nodes)
+                if (WARN_ON(nodes++ == mm->heap_nodes))
                        return -EBUSY;
        }
diff --git a/drivers/gpu/drm/nouveau/core/include/core/mm.h b/drivers/gpu/drm/nouveau/core/include/core/mm.h
index 9ee9bf4028ca..975137ba34a6 100644
--- a/drivers/gpu/drm/nouveau/core/include/core/mm.h
+++ b/drivers/gpu/drm/nouveau/core/include/core/mm.h
@@ -19,7 +19,6 @@ struct nouveau_mm {
        u32 block_size;
        int heap_nodes;
-        u32 heap_size;
 };
 int  nouveau_mm_init(struct nouveau_mm *, u32 offset, u32 length, u32 block);
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bios/base.c b/drivers/gpu/drm/nouveau/core/subdev/bios/base.c
index dcb5c2befc92..70ca7d5a1aa1 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/bios/base.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/bios/base.c
@@ -72,7 +72,7 @@ nouveau_bios_shadow_of(struct nouveau_bios *bios)
        }
        data = of_get_property(dn, "NVDA,BMP", &size);
-        if (data) {
+        if (data && size) {
                bios->size = size;
                bios->data = kmalloc(bios->size, GFP_KERNEL);
                if (bios->data)
@@ -104,6 +104,9 @@ nouveau_bios_shadow_pramin(struct nouveau_bios *bios)
                goto out;
        bios->size = nv_rd08(bios, 0x700002) * 512;
+        if (!bios->size)
+                goto out;
        bios->data = kmalloc(bios->size, GFP_KERNEL);
        if (bios->data) {
                for (i = 0; i < bios->size; i++)
@@ -155,6 +158,9 @@ nouveau_bios_shadow_prom(struct nouveau_bios *bios)
        /* read entire bios image to system memory */
        bios->size = nv_rd08(bios, 0x300002) * 512;
+        if (!bios->size)
+                goto out;
        bios->data = kmalloc(bios->size, GFP_KERNEL);
        if (bios->data) {
                for (i = 0; i < bios->size; i++)
@@ -186,14 +192,22 @@ nouveau_bios_shadow_acpi(struct nouveau_bios *bios)
 {
        struct pci_dev *pdev = nv_device(bios)->pdev;
        int ret, cnt, i;
-        u8  data[3];
-        if (!nouveau_acpi_rom_supported(pdev))
+        if (!nouveau_acpi_rom_supported(pdev)) {
+                bios->data = NULL;
                return;
+        }
        bios->size = 0;
-        if (nouveau_acpi_get_bios_chunk(data, 0, 3) == 3)
+        bios->data = kmalloc(4096, GFP_KERNEL);
-                bios->size = data[2] * 512;
+        if (bios->data) {
+                if (nouveau_acpi_get_bios_chunk(bios->data, 0, 4096) == 4096)
+                        bios->size = bios->data[2] * 512;
+                kfree(bios->data);
+        }
+        if (!bios->size)
+                return;
        bios->data = kmalloc(bios->size, GFP_KERNEL);
        for (i = 0; bios->data && i < bios->size; i += cnt) {
@@ -229,12 +243,14 @@ nouveau_bios_shadow_pci(struct nouveau_bios *bios)
 static int
 nouveau_bios_score(struct nouveau_bios *bios, const bool writeable)
 {
-        if (!bios->data || bios->data[0] != 0x55 || bios->data[1] != 0xAA) {
+        if (bios->size < 3 || !bios->data || bios->data[0] != 0x55 ||
+                        bios->data[1] != 0xAA) {
                nv_info(bios, "... signature not found\n");
                return 0;
        }
-        if (nvbios_checksum(bios->data, bios->data[2] * 512)) {
+        if (nvbios_checksum(bios->data,
+                        min_t(u32, bios->data[2] * 512, bios->size))) {
                nv_info(bios, "... checksum invalid\n");
                /* if a ro image is somewhat bad, it's probably all rubbish */
                return writeable ? 2 : 1;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bios/dcb.c b/drivers/gpu/drm/nouveau/core/subdev/bios/dcb.c
index 9ed6e728a94c..7d750382a833 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/bios/dcb.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/bios/dcb.c
@@ -43,7 +43,7 @@ dcb_table(struct nouveau_bios *bios, u8 *ver, u8 *hdr, u8 *cnt, u8 *len)
        *ver = nv_ro08(bios, dcb);
        if (*ver >= 0x41) {
-                nv_warn(bios, "DCB *ver 0x%02x unknown\n", *ver);
+                nv_warn(bios, "DCB version 0x%02x unknown\n", *ver);
                return 0x0000;
        } else
        if (*ver >= 0x30) {
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bios/pll.c b/drivers/gpu/drm/nouveau/core/subdev/bios/pll.c
index 5e5f4cddae3c..f835501203e5 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/bios/pll.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/bios/pll.c
@@ -157,11 +157,10 @@ pll_map_reg(struct nouveau_bios *bios, u32 reg, u32 *type, u8 *ver, u8 *len)
        while (map->reg) {
                if (map->reg == reg && *ver >= 0x20) {
                        u16 addr = (data += hdr);
+                        *type = map->type;
                        while (cnt--) {
-                                if (nv_ro32(bios, data) == map->reg) {
+                                if (nv_ro32(bios, data) == map->reg)
-                                        *type = map->type;
                                        return data;
-                                }
                                data += *len;
                        }
                        return addr;
@@ -200,11 +199,10 @@ pll_map_type(struct nouveau_bios *bios, u8 type, u32 *reg, u8 *ver, u8 *len)
        while (map->reg) {
                if (map->type == type && *ver >= 0x20) {
                        u16 addr = (data += hdr);
+                        *reg = map->reg;
                        while (cnt--) {
-                                if (nv_ro32(bios, data) == map->reg) {
+                                if (nv_ro32(bios, data) == map->reg)
-                                        *reg = map->reg;
                                        return data;
-                                }
                                data += *len;
                        }
                        return addr;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/nv50.c b/drivers/gpu/drm/nouveau/core/subdev/fb/nv50.c
index 436e9efe7ef5..5f570806143a 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/nv50.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/nv50.c
@@ -219,13 +219,11 @@ nv50_fb_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
                             ((priv->base.ram.size & 0x000000ff) << 32);
        tags = nv_rd32(priv, 0x100320);
-        if (tags) {
+        ret = nouveau_mm_init(&priv->base.tags, 0, tags, 1);
-                ret = nouveau_mm_init(&priv->base.tags, 0, tags, 1);
+        if (ret)
-                if (ret)
+                return ret;
-                        return ret;
-                nv_debug(priv, "%d compression tags\n", tags);
+        nv_debug(priv, "%d compression tags\n", tags);
-        }
        size = (priv->base.ram.size >> 12) - rsvd_head - rsvd_tail;
        switch (device->chipset) {
@@ -237,6 +235,7 @@ nv50_fb_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
                        return ret;
                priv->base.ram.stolen = (u64)nv_rd32(priv, 0x100e10) << 12;
+                priv->base.ram.type = NV_MEM_TYPE_STOLEN;
                break;
        default:
                ret = nouveau_mm_init(&priv->base.vram, rsvd_head, size,
@@ -277,7 +276,6 @@ nv50_fb_dtor(struct nouveau_object *object)
                __free_page(priv->r100c08_page);
        }
-        nouveau_mm_fini(&priv->base.vram);
        nouveau_fb_destroy(&priv->base);
 }
diff --git a/drivers/gpu/drm/nouveau/core/subdev/i2c/base.c b/drivers/gpu/drm/nouveau/core/subdev/i2c/base.c
index 3d2c88310f98..dbfc2abf0cfe 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/i2c/base.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/i2c/base.c
@@ -292,7 +292,7 @@ nouveau_i2c_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
                case DCB_I2C_NVIO_BIT:
                        port->drive = info.drive & 0x0f;
                        if (device->card_type < NV_D0) {
-                                if (info.drive >= ARRAY_SIZE(nv50_i2c_port))
+                                if (port->drive >= ARRAY_SIZE(nv50_i2c_port))
                                        break;
                                port->drive = nv50_i2c_port[port->drive];
                                port->sense = port->drive;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/therm/fan.c b/drivers/gpu/drm/nouveau/core/subdev/therm/fan.c
index b29237970fa0..523178685180 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/therm/fan.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/therm/fan.c
@@ -134,7 +134,7 @@ nouveau_therm_fan_sense(struct nouveau_therm *therm)
        end = ptimer->read(ptimer);
        if (cycles == 5) {
-                tach = (u64)60000000000;
+                tach = (u64)60000000000ULL;
                do_div(tach, (end - start));
                return tach;
        } else
diff --git a/drivers/gpu/drm/nouveau/core/subdev/vm/nv41.c b/drivers/gpu/drm/nouveau/core/subdev/vm/nv41.c
index 0203e1e12caa..49050d991e75 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/vm/nv41.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/vm/nv41.c
@@ -92,7 +92,8 @@ nv41_vmmgr_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
        struct nv04_vmmgr_priv *priv;
        int ret;
-        if (!nouveau_boolopt(device->cfgopt, "NvPCIE", true)) {
+        if (pci_find_capability(device->pdev, PCI_CAP_ID_AGP) ||
+            !nouveau_boolopt(device->cfgopt, "NvPCIE", true)) {
                return nouveau_object_ctor(parent, engine, &nv04_vmmgr_oclass,
                                           data, size, pobject);
        }
diff --git a/drivers/gpu/drm/nouveau/core/subdev/vm/nv44.c b/drivers/gpu/drm/nouveau/core/subdev/vm/nv44.c
index 0ac18d05a146..aa8131436e3d 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/vm/nv44.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/vm/nv44.c
@@ -163,7 +163,8 @@ nv44_vmmgr_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
        struct nv04_vmmgr_priv *priv;
        int ret;
-        if (!nouveau_boolopt(device->cfgopt, "NvPCIE", true)) {
+        if (pci_find_capability(device->pdev, PCI_CAP_ID_AGP) ||
+            !nouveau_boolopt(device->cfgopt, "NvPCIE", true)) {
                return nouveau_object_ctor(parent, engine, &nv04_vmmgr_oclass,
                                           data, size, pobject);
        }
diff --git a/drivers/gpu/drm/nouveau/nouveau_bo.c b/drivers/gpu/drm/nouveau/nouveau_bo.c
index 259e5f1adf47..35ac57f0aab6 100644
--- a/drivers/gpu/drm/nouveau/nouveau_bo.c
+++ b/drivers/gpu/drm/nouveau/nouveau_bo.c
@@ -456,6 +456,7 @@ static struct ttm_tt *
 nouveau_ttm_tt_create(struct ttm_bo_device *bdev, unsigned long size,
                      uint32_t page_flags, struct page *dummy_read)
 {
+#if __OS_HAS_AGP
        struct nouveau_drm *drm = nouveau_bdev(bdev);
        struct drm_device *dev = drm->dev;
@@ -463,6 +464,7 @@ nouveau_ttm_tt_create(struct ttm_bo_device *bdev, unsigned long size,
                return ttm_agp_tt_create(bdev, dev->agp->bridge, size,
                                         page_flags, dummy_read);
        }
+#endif
        return nouveau_sgdma_create_ttm(bdev, size, page_flags, dummy_read);
 }
diff --git a/drivers/gpu/drm/nouveau/nouveau_display.c b/drivers/gpu/drm/nouveau/nouveau_display.c
index 8f98e5a8c488..86124b131f4f 100644
--- a/drivers/gpu/drm/nouveau/nouveau_display.c
+++ b/drivers/gpu/drm/nouveau/nouveau_display.c
@@ -290,6 +290,7 @@ nouveau_display_create(struct drm_device *dev)
        struct nouveau_drm *drm = nouveau_drm(dev);
        struct nouveau_disp *pdisp = nouveau_disp(drm->device);
        struct nouveau_display *disp;
+        u32 pclass = dev->pdev->class >> 8;
        int ret, gen;
        disp = drm->display = kzalloc(sizeof(*disp), GFP_KERNEL);
@@ -360,23 +361,27 @@ nouveau_display_create(struct drm_device *dev)
        drm_kms_helper_poll_init(dev);
        drm_kms_helper_poll_disable(dev);
-        if (nv_device(drm->device)->card_type < NV_50)
+        if (nouveau_modeset == 1 ||
-                ret = nv04_display_create(dev);
+            (nouveau_modeset < 0 && pclass == PCI_CLASS_DISPLAY_VGA)) {
-        else
+                if (nv_device(drm->device)->card_type < NV_50)
-        if (nv_device(drm->device)->card_type < NV_D0)
+                        ret = nv04_display_create(dev);
-                ret = nv50_display_create(dev);
+                else
-        else
+                if (nv_device(drm->device)->card_type < NV_D0)
-                ret = nvd0_display_create(dev);
+                        ret = nv50_display_create(dev);
-        if (ret)
+                else
-                goto disp_create_err;
+                        ret = nvd0_display_create(dev);
-        if (dev->mode_config.num_crtc) {
-                ret = drm_vblank_init(dev, dev->mode_config.num_crtc);
                if (ret)
-                        goto vblank_err;
+                        goto disp_create_err;
+                if (dev->mode_config.num_crtc) {
+                        ret = drm_vblank_init(dev, dev->mode_config.num_crtc);
+                        if (ret)
+                                goto vblank_err;
+                }
+                nouveau_backlight_init(dev);
        }
-        nouveau_backlight_init(dev);
        return 0;
 vblank_err:
@@ -395,7 +400,8 @@ nouveau_display_destroy(struct drm_device *dev)
        nouveau_backlight_exit(dev);
        drm_vblank_cleanup(dev);
-        disp->dtor(dev);
+        if (disp->dtor)
+                disp->dtor(dev);
        drm_kms_helper_poll_fini(dev);
        drm_mode_config_cleanup(dev);
@@ -530,9 +536,11 @@ nouveau_page_flip_reserve(struct nouveau_bo *old_bo,
        if (ret)
                goto fail;
-        ret = ttm_bo_reserve(&old_bo->bo, false, false, false, 0);
+        if (likely(old_bo != new_bo)) {
-        if (ret)
+                ret = ttm_bo_reserve(&old_bo->bo, false, false, false, 0);
-                goto fail_unreserve;
+                if (ret)
+                        goto fail_unreserve;
+        }
        return 0;
@@ -551,8 +559,10 @@ nouveau_page_flip_unreserve(struct nouveau_bo *old_bo,
        nouveau_bo_fence(new_bo, fence);
        ttm_bo_unreserve(&new_bo->bo);
-        nouveau_bo_fence(old_bo, fence);
+        if (likely(old_bo != new_bo)) {
-        ttm_bo_unreserve(&old_bo->bo);
+                nouveau_bo_fence(old_bo, fence);
+                ttm_bo_unreserve(&old_bo->bo);
+        }
        nouveau_bo_unpin(old_bo);
 }
diff --git a/drivers/gpu/drm/nouveau/nouveau_drm.c b/drivers/gpu/drm/nouveau/nouveau_drm.c
index ccae8c26ae2b..0910125cbbc3 100644
--- a/drivers/gpu/drm/nouveau/nouveau_drm.c
+++ b/drivers/gpu/drm/nouveau/nouveau_drm.c
@@ -63,8 +63,9 @@ MODULE_PARM_DESC(noaccel, "disable kernel/abi16 acceleration");
 static int nouveau_noaccel = 0;
 module_param_named(noaccel, nouveau_noaccel, int, 0400);
-MODULE_PARM_DESC(modeset, "enable driver");
+MODULE_PARM_DESC(modeset, "enable driver (default: auto, "
-static int nouveau_modeset = -1;
+                          "0 = disabled, 1 = enabled, 2 = headless)");
+int nouveau_modeset = -1;
 module_param_named(modeset, nouveau_modeset, int, 0400);
 static struct drm_driver driver;
@@ -363,7 +364,8 @@ nouveau_drm_unload(struct drm_device *dev)
        nouveau_pm_fini(dev);
-        nouveau_display_fini(dev);
+        if (dev->mode_config.num_crtc)
+                nouveau_display_fini(dev);
        nouveau_display_destroy(dev);
        nouveau_irq_fini(dev);
@@ -403,13 +405,15 @@ nouveau_drm_suspend(struct pci_dev *pdev, pm_message_t pm_state)
            pm_state.event == PM_EVENT_PRETHAW)
                return 0;
-        NV_INFO(drm, "suspending fbcon...\n");
+        if (dev->mode_config.num_crtc) {
-        nouveau_fbcon_set_suspend(dev, 1);
+                NV_INFO(drm, "suspending fbcon...\n");
+                nouveau_fbcon_set_suspend(dev, 1);
-        NV_INFO(drm, "suspending display...\n");
+                NV_INFO(drm, "suspending display...\n");
-        ret = nouveau_display_suspend(dev);
+                ret = nouveau_display_suspend(dev);
-        if (ret)
+                if (ret)
-                return ret;
+                        return ret;
+        }
        NV_INFO(drm, "evicting buffers...\n");
        ttm_bo_evict_mm(&drm->ttm.bdev, TTM_PL_VRAM);
@@ -445,8 +449,10 @@ fail_client:
                nouveau_client_init(&cli->base);
        }
-        NV_INFO(drm, "resuming display...\n");
+        if (dev->mode_config.num_crtc) {
-        nouveau_display_resume(dev);
+                NV_INFO(drm, "resuming display...\n");
+                nouveau_display_resume(dev);
+        }
        return ret;
 }
@@ -486,8 +492,10 @@ nouveau_drm_resume(struct pci_dev *pdev)
        nouveau_irq_postinstall(dev);
        nouveau_pm_resume(dev);
-        NV_INFO(drm, "resuming display...\n");
+        if (dev->mode_config.num_crtc) {
-        nouveau_display_resume(dev);
+                NV_INFO(drm, "resuming display...\n");
+                nouveau_display_resume(dev);
+        }
        return 0;
 }
@@ -662,9 +670,7 @@ nouveau_drm_init(void)
 #ifdef CONFIG_VGA_CONSOLE
                if (vgacon_text_force())
                        nouveau_modeset = 0;
-                else
 #endif
-                        nouveau_modeset = 1;
        }
        if (!nouveau_modeset)
diff --git a/drivers/gpu/drm/nouveau/nouveau_drm.h b/drivers/gpu/drm/nouveau/nouveau_drm.h
index 819471217546..a10169927086 100644
--- a/drivers/gpu/drm/nouveau/nouveau_drm.h
+++ b/drivers/gpu/drm/nouveau/nouveau_drm.h
@@ -141,4 +141,6 @@ int nouveau_drm_resume(struct pci_dev *);
                nv_info((cli), fmt, ##args);                                   \
 } while (0)
+extern int nouveau_modeset;
 #endif
diff --git a/drivers/gpu/drm/nouveau/nouveau_irq.c b/drivers/gpu/drm/nouveau/nouveau_irq.c
index 9ca8afdb5549..1d8cb506a28a 100644
--- a/drivers/gpu/drm/nouveau/nouveau_irq.c
+++ b/drivers/gpu/drm/nouveau/nouveau_irq.c
@@ -61,13 +61,15 @@ nouveau_irq_handler(DRM_IRQ_ARGS)
        nv_subdev(pmc)->intr(nv_subdev(pmc));
-        if (device->card_type >= NV_D0) {
+        if (dev->mode_config.num_crtc) {
-                if (nv_rd32(device, 0x000100) & 0x04000000)
+                if (device->card_type >= NV_D0) {
-                        nvd0_display_intr(dev);
+                        if (nv_rd32(device, 0x000100) & 0x04000000)
-        } else
+                                nvd0_display_intr(dev);
-        if (device->card_type >= NV_50) {
+                } else
-                if (nv_rd32(device, 0x000100) & 0x04000000)
+                if (device->card_type >= NV_50) {
-                        nv50_display_intr(dev);
+                        if (nv_rd32(device, 0x000100) & 0x04000000)
+                                nv50_display_intr(dev);
+                }
        }
        return IRQ_HANDLED;
diff --git a/drivers/gpu/drm/nouveau/nouveau_pm.c b/drivers/gpu/drm/nouveau/nouveau_pm.c
index 0bf64c90aa20..5566172774df 100644
--- a/drivers/gpu/drm/nouveau/nouveau_pm.c
+++ b/drivers/gpu/drm/nouveau/nouveau_pm.c
@@ -52,7 +52,7 @@ nouveau_pm_perflvl_aux(struct drm_device *dev, struct nouveau_pm_level *perflvl,
 {
        struct nouveau_drm *drm = nouveau_drm(dev);
        struct nouveau_pm *pm = nouveau_pm(dev);
-        struct nouveau_therm *therm = nouveau_therm(drm);
+        struct nouveau_therm *therm = nouveau_therm(drm->device);
        int ret;
        /*XXX: not on all boards, we should control based on temperature
@@ -64,7 +64,6 @@ nouveau_pm_perflvl_aux(struct drm_device *dev, struct nouveau_pm_level *perflvl,
                ret = therm->fan_set(therm, perflvl->fanspeed);
                if (ret && ret != -ENODEV) {
                        NV_ERROR(drm, "fanspeed set failed: %d\n", ret);
-                        return ret;
                }
        }
@@ -706,8 +705,7 @@ nouveau_hwmon_init(struct drm_device *dev)
        struct device *hwmon_dev;
        int ret = 0;
-        if (!therm || !therm->temp_get || !therm->attr_get ||
+        if (!therm || !therm->temp_get || !therm->attr_get || !therm->attr_set)
-                !therm->attr_set || therm->temp_get(therm) < 0)
                return -ENODEV;
        hwmon_dev = hwmon_device_register(&dev->pdev->dev);
diff --git a/drivers/gpu/drm/nouveau/nv04_dac.c b/drivers/gpu/drm/nouveau/nv04_dac.c
index 347a3bd78d04..64f7020fb605 100644
--- a/drivers/gpu/drm/nouveau/nv04_dac.c
+++ b/drivers/gpu/drm/nouveau/nv04_dac.c
@@ -220,7 +220,7 @@ out:
        NVWriteVgaCrtc(dev, 0, NV_CIO_CR_MODE_INDEX, saved_cr_mode);
        if (blue == 0x18) {
-                NV_INFO(drm, "Load detected on head A\n");
+                NV_DEBUG(drm, "Load detected on head A\n");
                return connector_status_connected;
        }
@@ -338,8 +338,8 @@ nv17_dac_detect(struct drm_encoder *encoder, struct drm_connector *connector)
        if (nv17_dac_sample_load(encoder) &
            NV_PRAMDAC_TEST_CONTROL_SENSEB_ALLHI) {
-                NV_INFO(drm, "Load detected on output %c\n",
+                NV_DEBUG(drm, "Load detected on output %c\n",
-                        '@' + ffs(dcb->or));
+                         '@' + ffs(dcb->or));
                return connector_status_connected;
        } else {
                return connector_status_disconnected;
@@ -413,9 +413,9 @@ static void nv04_dac_commit(struct drm_encoder *encoder)
        helper->dpms(encoder, DRM_MODE_DPMS_ON);
-        NV_INFO(drm, "Output %s is running on CRTC %d using output %c\n",
+        NV_DEBUG(drm, "Output %s is running on CRTC %d using output %c\n",
-                drm_get_connector_name(&nouveau_encoder_connector_get(nv_encoder)->base),
+                 drm_get_connector_name(&nouveau_encoder_connector_get(nv_encoder)->base),
-                nv_crtc->index, '@' + ffs(nv_encoder->dcb->or));
+                 nv_crtc->index, '@' + ffs(nv_encoder->dcb->or));
 }
 void nv04_dac_update_dacclk(struct drm_encoder *encoder, bool enable)
@@ -461,8 +461,8 @@ static void nv04_dac_dpms(struct drm_encoder *encoder, int mode)
                return;
        nv_encoder->last_dpms = mode;
-        NV_INFO(drm, "Setting dpms mode %d on vga encoder (output %d)\n",
+        NV_DEBUG(drm, "Setting dpms mode %d on vga encoder (output %d)\n",
-                     mode, nv_encoder->dcb->index);
+                 mode, nv_encoder->dcb->index);
        nv04_dac_update_dacclk(encoder, mode == DRM_MODE_DPMS_ON);
 }
diff --git a/drivers/gpu/drm/nouveau/nv04_dfp.c b/drivers/gpu/drm/nouveau/nv04_dfp.c
index da55d7642c8c..184cdf806761 100644
--- a/drivers/gpu/drm/nouveau/nv04_dfp.c
+++ b/drivers/gpu/drm/nouveau/nv04_dfp.c
@@ -476,9 +476,9 @@ static void nv04_dfp_commit(struct drm_encoder *encoder)
        helper->dpms(encoder, DRM_MODE_DPMS_ON);
-        NV_INFO(drm, "Output %s is running on CRTC %d using output %c\n",
+        NV_DEBUG(drm, "Output %s is running on CRTC %d using output %c\n",
-                drm_get_connector_name(&nouveau_encoder_connector_get(nv_encoder)->base),
+                 drm_get_connector_name(&nouveau_encoder_connector_get(nv_encoder)->base),
-                nv_crtc->index, '@' + ffs(nv_encoder->dcb->or));
+                 nv_crtc->index, '@' + ffs(nv_encoder->dcb->or));
 }
 static void nv04_dfp_update_backlight(struct drm_encoder *encoder, int mode)
@@ -520,8 +520,8 @@ static void nv04_lvds_dpms(struct drm_encoder *encoder, int mode)
                return;
        nv_encoder->last_dpms = mode;
-        NV_INFO(drm, "Setting dpms mode %d on lvds encoder (output %d)\n",
+        NV_DEBUG(drm, "Setting dpms mode %d on lvds encoder (output %d)\n",
-                     mode, nv_encoder->dcb->index);
+                 mode, nv_encoder->dcb->index);
        if (was_powersaving && is_powersaving_dpms(mode))
                return;
@@ -565,8 +565,8 @@ static void nv04_tmds_dpms(struct drm_encoder *encoder, int mode)
                return;
        nv_encoder->last_dpms = mode;
-        NV_INFO(drm, "Setting dpms mode %d on tmds encoder (output %d)\n",
+        NV_DEBUG(drm, "Setting dpms mode %d on tmds encoder (output %d)\n",
-                     mode, nv_encoder->dcb->index);
+                 mode, nv_encoder->dcb->index);
        nv04_dfp_update_backlight(encoder, mode);
        nv04_dfp_update_fp_control(encoder, mode);
diff --git a/drivers/gpu/drm/nouveau/nv04_tv.c b/drivers/gpu/drm/nouveau/nv04_tv.c
index 099fbeda6e2e..62e826a139b3 100644
--- a/drivers/gpu/drm/nouveau/nv04_tv.c
+++ b/drivers/gpu/drm/nouveau/nv04_tv.c
@@ -75,8 +75,8 @@ static void nv04_tv_dpms(struct drm_encoder *encoder, int mode)
        struct nv04_mode_state *state = &nv04_display(dev)->mode_reg;
        uint8_t crtc1A;
-        NV_INFO(drm, "Setting dpms mode %d on TV encoder (output %d)\n",
+        NV_DEBUG(drm, "Setting dpms mode %d on TV encoder (output %d)\n",
-                mode, nv_encoder->dcb->index);
+                 mode, nv_encoder->dcb->index);
        state->pllsel &= ~(PLLSEL_TV_CRTC1_MASK | PLLSEL_TV_CRTC2_MASK);
@@ -167,9 +167,8 @@ static void nv04_tv_commit(struct drm_encoder *encoder)
        helper->dpms(encoder, DRM_MODE_DPMS_ON);
-        NV_INFO(drm, "Output %s is running on CRTC %d using output %c\n",
+        NV_DEBUG(drm, "Output %s is running on CRTC %d using output %c\n",
-                      drm_get_connector_name(&nouveau_encoder_connector_get(nv_encoder)->base), nv_crtc->index,
+                 drm_get_connector_name(&nouveau_encoder_connector_get(nv_encoder)->base), nv_crtc->index, '@' + ffs(nv_encoder->dcb->or));
-                      '@' + ffs(nv_encoder->dcb->or));
 }
 static void nv04_tv_destroy(struct drm_encoder *encoder)
diff --git a/drivers/gpu/drm/radeon/atombios_crtc.c b/drivers/gpu/drm/radeon/atombios_crtc.c
index 96184d02c8d9..2e566e123e9e 100644
--- a/drivers/gpu/drm/radeon/atombios_crtc.c
+++ b/drivers/gpu/drm/radeon/atombios_crtc.c
@@ -1690,10 +1690,10 @@ static int radeon_atom_pick_pll(struct drm_crtc *crtc)
                }
                /* all other cases */
                pll_in_use = radeon_get_pll_use_mask(crtc);
-                if (!(pll_in_use & (1 << ATOM_PPLL2)))
-                        return ATOM_PPLL2;
                if (!(pll_in_use & (1 << ATOM_PPLL1)))
                        return ATOM_PPLL1;
+                if (!(pll_in_use & (1 << ATOM_PPLL2)))
+                        return ATOM_PPLL2;
                DRM_ERROR("unable to allocate a PPLL\n");
                return ATOM_PPLL_INVALID;
        } else {
@@ -1715,10 +1715,10 @@ static int radeon_atom_pick_pll(struct drm_crtc *crtc)
                        }
                        /* all other cases */
                        pll_in_use = radeon_get_pll_use_mask(crtc);
-                        if (!(pll_in_use & (1 << ATOM_PPLL2)))
-                                return ATOM_PPLL2;
                        if (!(pll_in_use & (1 << ATOM_PPLL1)))
                                return ATOM_PPLL1;
+                        if (!(pll_in_use & (1 << ATOM_PPLL2)))
+                                return ATOM_PPLL2;
                        DRM_ERROR("unable to allocate a PPLL\n");
                        return ATOM_PPLL_INVALID;
                } else {
diff --git a/drivers/gpu/drm/radeon/atombios_encoders.c b/drivers/gpu/drm/radeon/atombios_encoders.c
index 49cbb3795a10..ba498f8e47a2 100644
--- a/drivers/gpu/drm/radeon/atombios_encoders.c
+++ b/drivers/gpu/drm/radeon/atombios_encoders.c
@@ -184,6 +184,7 @@ void radeon_atom_backlight_init(struct radeon_encoder *radeon_encoder,
        struct radeon_backlight_privdata *pdata;
        struct radeon_encoder_atom_dig *dig;
        u8 backlight_level;
+        char bl_name[16];
        if (!radeon_encoder->enc_priv)
                return;
@@ -203,7 +204,9 @@ void radeon_atom_backlight_init(struct radeon_encoder *radeon_encoder,
        memset(&props, 0, sizeof(props));
        props.max_brightness = RADEON_MAX_BL_LEVEL;
        props.type = BACKLIGHT_RAW;
-        bd = backlight_device_register("radeon_bl", &drm_connector->kdev,
+        snprintf(bl_name, sizeof(bl_name),
+                 "radeon_bl%d", dev->primary->index);
+        bd = backlight_device_register(bl_name, &drm_connector->kdev,
                                       pdata, &radeon_atom_backlight_ops, &props);
        if (IS_ERR(bd)) {
                DRM_ERROR("Backlight registration failed\n");
diff --git a/drivers/gpu/drm/radeon/evergreen.c b/drivers/gpu/drm/radeon/evergreen.c
index a1f49c5fd74b..14313ad43b76 100644
--- a/drivers/gpu/drm/radeon/evergreen.c
+++ b/drivers/gpu/drm/radeon/evergreen.c
@@ -3431,9 +3431,14 @@ void evergreen_pcie_gen2_enable(struct radeon_device *rdev)
        if (!(mask & DRM_PCIE_SPEED_50))
                return;
+        speed_cntl = RREG32_PCIE_P(PCIE_LC_SPEED_CNTL);
+        if (speed_cntl & LC_CURRENT_DATA_RATE) {
+                DRM_INFO("PCIE gen 2 link speeds already enabled\n");
+                return;
+        }
        DRM_INFO("enabling PCIE gen 2 link speeds, disable with radeon.pcie_gen2=0\n");
-        speed_cntl = RREG32_PCIE_P(PCIE_LC_SPEED_CNTL);
        if ((speed_cntl & LC_OTHER_SIDE_EVER_SENT_GEN2) ||
            (speed_cntl & LC_OTHER_SIDE_SUPPORTS_GEN2)) {
diff --git a/drivers/gpu/drm/radeon/evergreen_cs.c b/drivers/gpu/drm/radeon/evergreen_cs.c
index 573ed1bc6cf7..95e6318b6268 100644
--- a/drivers/gpu/drm/radeon/evergreen_cs.c
+++ b/drivers/gpu/drm/radeon/evergreen_cs.c
@@ -264,7 +264,7 @@ static int evergreen_surface_check_2d(struct radeon_cs_parser *p,
        /* macro tile width & height */
        palign = (8 * surf->bankw * track->npipes) * surf->mtilea;
        halign = (8 * surf->bankh * surf->nbanks) / surf->mtilea;
-        mtileb = (palign / 8) * (halign / 8) * tileb;;
+        mtileb = (palign / 8) * (halign / 8) * tileb;
        mtile_pr = surf->nbx / palign;
        mtile_ps = (mtile_pr * surf->nby) / halign;
        surf->layer_size = mtile_ps * mtileb * slice_pt;
@@ -2829,6 +2829,7 @@ static bool evergreen_vm_reg_valid(u32 reg)
        case CAYMAN_SQ_EX_ALLOC_TABLE_SLOTS:
                return true;
        default:
+                DRM_ERROR("Invalid register 0x%x in CS\n", reg);
                return false;
        }
 }
diff --git a/drivers/gpu/drm/radeon/ni.c b/drivers/gpu/drm/radeon/ni.c
index 8bcb554ea0c5..81e6a568c29d 100644
--- a/drivers/gpu/drm/radeon/ni.c
+++ b/drivers/gpu/drm/radeon/ni.c
@@ -770,9 +770,13 @@ static int cayman_pcie_gart_enable(struct radeon_device *rdev)
        WREG32(0x15DC, 0);
        /* empty context1-7 */
+        /* Assign the pt base to something valid for now; the pts used for
+         * the VMs are determined by the application and setup and assigned
+         * on the fly in the vm part of radeon_gart.c
+         */
        for (i = 1; i < 8; i++) {
                WREG32(VM_CONTEXT0_PAGE_TABLE_START_ADDR + (i << 2), 0);
-                WREG32(VM_CONTEXT0_PAGE_TABLE_END_ADDR + (i << 2), 0);
+                WREG32(VM_CONTEXT0_PAGE_TABLE_END_ADDR + (i << 2), rdev->vm_manager.max_pfn);
                WREG32(VM_CONTEXT0_PAGE_TABLE_BASE_ADDR + (i << 2),
                        rdev->gart.table_addr >> 12);
        }
@@ -1534,26 +1538,31 @@ void cayman_vm_set_page(struct radeon_device *rdev, uint64_t pe,
 {
        struct radeon_ring *ring = &rdev->ring[rdev->asic->vm.pt_ring_index];
        uint32_t r600_flags = cayman_vm_page_flags(rdev, flags);
-        int i;
-        radeon_ring_write(ring, PACKET3(PACKET3_ME_WRITE, 1 + count * 2));
+        while (count) {
-        radeon_ring_write(ring, pe);
+                unsigned ndw = 1 + count * 2;
-        radeon_ring_write(ring, upper_32_bits(pe) & 0xff);
+                if (ndw > 0x3FFF)
-        for (i = 0; i < count; ++i) {
+                        ndw = 0x3FFF;
-                uint64_t value = 0;
-                if (flags & RADEON_VM_PAGE_SYSTEM) {
+                radeon_ring_write(ring, PACKET3(PACKET3_ME_WRITE, ndw));
-                        value = radeon_vm_map_gart(rdev, addr);
+                radeon_ring_write(ring, pe);
-                        value &= 0xFFFFFFFFFFFFF000ULL;
+                radeon_ring_write(ring, upper_32_bits(pe) & 0xff);
-                        addr += incr;
+                for (; ndw > 1; ndw -= 2, --count, pe += 8) {
+                        uint64_t value = 0;
-                } else if (flags & RADEON_VM_PAGE_VALID) {
+                        if (flags & RADEON_VM_PAGE_SYSTEM) {
-                        value = addr;
+                                value = radeon_vm_map_gart(rdev, addr);
-                        addr += incr;
+                                value &= 0xFFFFFFFFFFFFF000ULL;
-                }
+                                addr += incr;
+                        } else if (flags & RADEON_VM_PAGE_VALID) {
+                                value = addr;
+                                addr += incr;
+                        }
-                value |= r600_flags;
+                        value |= r600_flags;
-                radeon_ring_write(ring, value);
+                        radeon_ring_write(ring, value);
-                radeon_ring_write(ring, upper_32_bits(value));
+                        radeon_ring_write(ring, upper_32_bits(value));
+                }
        }
 }
@@ -1572,12 +1581,6 @@ void cayman_vm_flush(struct radeon_device *rdev, int ridx, struct radeon_vm *vm)
        if (vm == NULL)
                return;
-        radeon_ring_write(ring, PACKET0(VM_CONTEXT0_PAGE_TABLE_START_ADDR + (vm->id << 2), 0));
-        radeon_ring_write(ring, 0);
-        radeon_ring_write(ring, PACKET0(VM_CONTEXT0_PAGE_TABLE_END_ADDR + (vm->id << 2), 0));
-        radeon_ring_write(ring, vm->last_pfn);
        radeon_ring_write(ring, PACKET0(VM_CONTEXT0_PAGE_TABLE_BASE_ADDR + (vm->id << 2), 0));
        radeon_ring_write(ring, vm->pd_gpu_addr >> 12);
@@ -1588,4 +1591,8 @@ void cayman_vm_flush(struct radeon_device *rdev, int ridx, struct radeon_vm *vm)
        /* bits 0-7 are the VM contexts0-7 */
        radeon_ring_write(ring, PACKET0(VM_INVALIDATE_REQUEST, 0));
        radeon_ring_write(ring, 1 << vm->id);
+        /* sync PFP to ME, otherwise we might get invalid PFP reads */
+        radeon_ring_write(ring, PACKET3(PACKET3_PFP_SYNC_ME, 0));
+        radeon_ring_write(ring, 0x0);
 }
diff --git a/drivers/gpu/drm/radeon/nid.h b/drivers/gpu/drm/radeon/nid.h
index 2423d1b5d385..cbef6815907a 100644
--- a/drivers/gpu/drm/radeon/nid.h
+++ b/drivers/gpu/drm/radeon/nid.h
@@ -502,6 +502,7 @@
 #define PACKET3_MPEG_INDEX                              0x3A
 #define PACKET3_WAIT_REG_MEM                            0x3C
 #define PACKET3_MEM_WRITE                               0x3D
+#define PACKET3_PFP_SYNC_ME                             0x42
 #define PACKET3_SURFACE_SYNC                            0x43
 #              define PACKET3_CB0_DEST_BASE_ENA    (1 << 6)
 #              define PACKET3_CB1_DEST_BASE_ENA    (1 << 7)
diff --git a/drivers/gpu/drm/radeon/r600.c b/drivers/gpu/drm/radeon/r600.c
index 70c800ff6190..cda280d157da 100644
--- a/drivers/gpu/drm/radeon/r600.c
+++ b/drivers/gpu/drm/radeon/r600.c
@@ -3703,6 +3703,12 @@ static void r600_pcie_gen2_enable(struct radeon_device *rdev)
        if (!(mask & DRM_PCIE_SPEED_50))
                return;
+        speed_cntl = RREG32_PCIE_P(PCIE_LC_SPEED_CNTL);
+        if (speed_cntl & LC_CURRENT_DATA_RATE) {
+                DRM_INFO("PCIE gen 2 link speeds already enabled\n");
+                return;
+        }
        DRM_INFO("enabling PCIE gen 2 link speeds, disable with radeon.pcie_gen2=0\n");
        /* 55 nm r6xx asics */
diff --git a/drivers/gpu/drm/radeon/radeon.h b/drivers/gpu/drm/radeon/radeon.h
index b04c06444d8b..8c42d54c2e26 100644
--- a/drivers/gpu/drm/radeon/radeon.h
+++ b/drivers/gpu/drm/radeon/radeon.h
@@ -663,9 +663,14 @@ struct radeon_vm {
        struct list_head                list;
        struct list_head                va;
        unsigned                        id;
-        unsigned                        last_pfn;
-        u64                             pd_gpu_addr;
+        /* contains the page directory */
-        struct radeon_sa_bo             *sa_bo;
+        struct radeon_sa_bo             *page_directory;
+        uint64_t                        pd_gpu_addr;
+        /* array of page tables, one for each page directory entry */
+        struct radeon_sa_bo             **page_tables;
        struct mutex                    mutex;
        /* last fence for cs using this vm */
        struct radeon_fence             *fence;
@@ -1843,9 +1848,10 @@ extern void radeon_ttm_set_active_vram_size(struct radeon_device *rdev, u64 size
 */
 int radeon_vm_manager_init(struct radeon_device *rdev);
 void radeon_vm_manager_fini(struct radeon_device *rdev);
-int radeon_vm_init(struct radeon_device *rdev, struct radeon_vm *vm);
+void radeon_vm_init(struct radeon_device *rdev, struct radeon_vm *vm);
 void radeon_vm_fini(struct radeon_device *rdev, struct radeon_vm *vm);
 int radeon_vm_alloc_pt(struct radeon_device *rdev, struct radeon_vm *vm);
+void radeon_vm_add_to_lru(struct radeon_device *rdev, struct radeon_vm *vm);
 struct radeon_fence *radeon_vm_grab_id(struct radeon_device *rdev,
                                       struct radeon_vm *vm, int ring);
 void radeon_vm_fence(struct radeon_device *rdev,
diff --git a/drivers/gpu/drm/radeon/radeon_acpi.c b/drivers/gpu/drm/radeon/radeon_acpi.c
index b0a5688c67f8..196d28d99570 100644
--- a/drivers/gpu/drm/radeon/radeon_acpi.c
+++ b/drivers/gpu/drm/radeon/radeon_acpi.c
@@ -201,7 +201,7 @@ static int radeon_atif_verify_interface(acpi_handle handle,
        size = *(u16 *) info->buffer.pointer;
        if (size < 12) {
-                DRM_INFO("ATIF buffer is too small: %lu\n", size);
+                DRM_INFO("ATIF buffer is too small: %zu\n", size);
                err = -EINVAL;
                goto out;
        }
@@ -370,6 +370,7 @@ int radeon_atif_handler(struct radeon_device *rdev,
                        radeon_set_backlight_level(rdev, enc, req.backlight_level);
+#if defined(CONFIG_BACKLIGHT_CLASS_DEVICE) || defined(CONFIG_BACKLIGHT_CLASS_DEVICE_MODULE)
                        if (rdev->is_atom_bios) {
                                struct radeon_encoder_atom_dig *dig = enc->enc_priv;
                                backlight_force_update(dig->bl_dev,
@@ -379,6 +380,7 @@ int radeon_atif_handler(struct radeon_device *rdev,
                                backlight_force_update(dig->bl_dev,
                                                       BACKLIGHT_UPDATE_HOTKEY);
                        }
+#endif
                }
        }
        /* TODO: check other events */
@@ -485,7 +487,7 @@ static int radeon_atcs_verify_interface(acpi_handle handle,
        size = *(u16 *) info->buffer.pointer;
        if (size < 8) {
-                DRM_INFO("ATCS buffer is too small: %lu\n", size);
+                DRM_INFO("ATCS buffer is too small: %zu\n", size);
                err = -EINVAL;
                goto out;
        }
diff --git a/drivers/gpu/drm/radeon/radeon_atpx_handler.c b/drivers/gpu/drm/radeon/radeon_atpx_handler.c
index 582e99449c12..15f5ded65e0c 100644
--- a/drivers/gpu/drm/radeon/radeon_atpx_handler.c
+++ b/drivers/gpu/drm/radeon/radeon_atpx_handler.c
@@ -87,7 +87,7 @@ static union acpi_object *radeon_atpx_call(acpi_handle handle, int function,
                atpx_arg_elements[1].integer.value = 0;
        }
-        status = acpi_evaluate_object(handle, "ATPX", &atpx_arg, &buffer);
+        status = acpi_evaluate_object(handle, NULL, &atpx_arg, &buffer);
        /* Fail only if calling the method fails and ATPX is supported */
        if (ACPI_FAILURE(status) && status != AE_NOT_FOUND) {
@@ -148,7 +148,7 @@ static int radeon_atpx_verify_interface(struct radeon_atpx *atpx)
        size = *(u16 *) info->buffer.pointer;
        if (size < 8) {
-                printk("ATPX buffer is too small: %lu\n", size);
+                printk("ATPX buffer is too small: %zu\n", size);
                err = -EINVAL;
                goto out;
        }
@@ -352,9 +352,9 @@ static int radeon_atpx_switchto(enum vga_switcheroo_client_id id)
 }
 /**
- * radeon_atpx_switchto - switch to the requested GPU
+ * radeon_atpx_power_state - power down/up the requested GPU
 *
- * @id: GPU to switch to
+ * @id: GPU to power down/up
 * @state: requested power state (0 = off, 1 = on)
 *
 * Execute the necessary ATPX function to power down/up the discrete GPU
@@ -373,11 +373,11 @@ static int radeon_atpx_power_state(enum vga_switcheroo_client_id id,
 }
 /**
- * radeon_atpx_pci_probe_handle - look up the ATRM and ATPX handles
+ * radeon_atpx_pci_probe_handle - look up the ATPX handle
 *
 * @pdev: pci device
 *
- * Look up the ATPX and ATRM handles (all asics).
+ * Look up the ATPX handles (all asics).
 * Returns true if the handles are found, false if not.
 */
 static bool radeon_atpx_pci_probe_handle(struct pci_dev *pdev)
diff --git a/drivers/gpu/drm/radeon/radeon_connectors.c b/drivers/gpu/drm/radeon/radeon_connectors.c
index 67cfc1795ecd..b884c362a8c2 100644
--- a/drivers/gpu/drm/radeon/radeon_connectors.c
+++ b/drivers/gpu/drm/radeon/radeon_connectors.c
@@ -941,7 +941,7 @@ radeon_dvi_detect(struct drm_connector *connector, bool force)
        struct drm_mode_object *obj;
        int i;
        enum drm_connector_status ret = connector_status_disconnected;
-        bool dret = false;
+        bool dret = false, broken_edid = false;
        if (!force && radeon_check_hpd_status_unchanged(connector))
                return connector->status;
@@ -965,6 +965,9 @@ radeon_dvi_detect(struct drm_connector *connector, bool force)
                                ret = connector_status_disconnected;
                                DRM_ERROR("%s: detected RS690 floating bus bug, stopping ddc detect\n", drm_get_connector_name(connector));
                                radeon_connector->ddc_bus = NULL;
+                        } else {
+                                ret = connector_status_connected;
+                                broken_edid = true; /* defer use_digital to later */
                        }
                } else {
                        radeon_connector->use_digital = !!(radeon_connector->edid->input & DRM_EDID_INPUT_DIGITAL);
@@ -1047,13 +1050,24 @@ radeon_dvi_detect(struct drm_connector *connector, bool force)
                        encoder_funcs = encoder->helper_private;
                        if (encoder_funcs->detect) {
-                                if (ret != connector_status_connected) {
+                                if (!broken_edid) {
-                                        ret = encoder_funcs->detect(encoder, connector);
+                                        if (ret != connector_status_connected) {
-                                        if (ret == connector_status_connected) {
+                                                /* deal with analog monitors without DDC */
-                                                radeon_connector->use_digital = false;
+                                                ret = encoder_funcs->detect(encoder, connector);
+                                                if (ret == connector_status_connected) {
+                                                        radeon_connector->use_digital = false;
+                                                }
+                                                if (ret != connector_status_disconnected)
+                                                        radeon_connector->detected_by_load = true;
                                        }
-                                        if (ret != connector_status_disconnected)
+                                } else {
-                                                radeon_connector->detected_by_load = true;
+                                        enum drm_connector_status lret;
+                                        /* assume digital unless load detected otherwise */
+                                        radeon_connector->use_digital = true;
+                                        lret = encoder_funcs->detect(encoder, connector);
+                                        DRM_DEBUG_KMS("load_detect %x returned: %x\n",encoder->encoder_type,lret);
+                                        if (lret == connector_status_connected)
+                                                radeon_connector->use_digital = false;
                                }
                                break;
                        }
diff --git a/drivers/gpu/drm/radeon/radeon_cs.c b/drivers/gpu/drm/radeon/radeon_cs.c
index cb7b7c062fef..41672cc563fb 100644
--- a/drivers/gpu/drm/radeon/radeon_cs.c
+++ b/drivers/gpu/drm/radeon/radeon_cs.c
@@ -478,6 +478,7 @@ static int radeon_cs_ib_vm_chunk(struct radeon_device *rdev,
        }
 out:
+        radeon_vm_add_to_lru(rdev, vm);
        mutex_unlock(&vm->mutex);
        mutex_unlock(&rdev->vm_manager.lock);
        return r;
diff --git a/drivers/gpu/drm/radeon/radeon_device.c b/drivers/gpu/drm/radeon/radeon_device.c
index 64a42647f08a..e2f5f888c374 100644
--- a/drivers/gpu/drm/radeon/radeon_device.c
+++ b/drivers/gpu/drm/radeon/radeon_device.c
@@ -355,6 +355,8 @@ int radeon_wb_init(struct radeon_device *rdev)
 */
 void radeon_vram_location(struct radeon_device *rdev, struct radeon_mc *mc, u64 base)
 {
+        uint64_t limit = (uint64_t)radeon_vram_limit << 20;
        mc->vram_start = base;
        if (mc->mc_vram_size > (0xFFFFFFFF - base + 1)) {
                dev_warn(rdev->dev, "limiting VRAM to PCI aperture size\n");
@@ -368,8 +370,8 @@ void radeon_vram_location(struct radeon_device *rdev, struct radeon_mc *mc, u64
                mc->mc_vram_size = mc->aper_size;
        }
        mc->vram_end = mc->vram_start + mc->mc_vram_size - 1;
-        if (radeon_vram_limit && radeon_vram_limit < mc->real_vram_size)
+        if (limit && limit < mc->real_vram_size)
-                mc->real_vram_size = radeon_vram_limit;
+                mc->real_vram_size = limit;
        dev_info(rdev->dev, "VRAM: %lluM 0x%016llX - 0x%016llX (%lluM used)\n",
                        mc->mc_vram_size >> 20, mc->vram_start,
                        mc->vram_end, mc->real_vram_size >> 20);
@@ -835,6 +837,19 @@ static unsigned int radeon_vga_set_decode(void *cookie, bool state)
 }
 /**
+ * radeon_check_pot_argument - check that argument is a power of two
+ *
+ * @arg: value to check
+ *
+ * Validates that a certain argument is a power of two (all asics).
+ * Returns true if argument is valid.
+ */
+static bool radeon_check_pot_argument(int arg)
+{
+        return (arg & (arg - 1)) == 0;
+}
+/**
 * radeon_check_arguments - validate module params
 *
 * @rdev: radeon_device pointer
@@ -845,52 +860,25 @@ static unsigned int radeon_vga_set_decode(void *cookie, bool state)
 static void radeon_check_arguments(struct radeon_device *rdev)
 {
        /* vramlimit must be a power of two */
-        switch (radeon_vram_limit) {
+        if (!radeon_check_pot_argument(radeon_vram_limit)) {
-        case 0:
-        case 4:
-        case 8:
-        case 16:
-        case 32:
-        case 64:
-        case 128:
-        case 256:
-        case 512:
-        case 1024:
-        case 2048:
-        case 4096:
-                break;
-        default:
                dev_warn(rdev->dev, "vram limit (%d) must be a power of 2\n",
                                radeon_vram_limit);
                radeon_vram_limit = 0;
-                break;
        }
-        radeon_vram_limit = radeon_vram_limit << 20;
        /* gtt size must be power of two and greater or equal to 32M */
-        switch (radeon_gart_size) {
+        if (radeon_gart_size < 32) {
-        case 4:
-        case 8:
-        case 16:
                dev_warn(rdev->dev, "gart size (%d) too small forcing to 512M\n",
                                radeon_gart_size);
                radeon_gart_size = 512;
-                break;
-        case 32:
+        } else if (!radeon_check_pot_argument(radeon_gart_size)) {
-        case 64:
-        case 128:
-        case 256:
-        case 512:
-        case 1024:
-        case 2048:
-        case 4096:
-                break;
-        default:
                dev_warn(rdev->dev, "gart size (%d) must be a power of 2\n",
                                radeon_gart_size);
                radeon_gart_size = 512;
-                break;
        }
-        rdev->mc.gtt_size = radeon_gart_size * 1024 * 1024;
+        rdev->mc.gtt_size = (uint64_t)radeon_gart_size << 20;
        /* AGP mode can only be -1, 1, 2, 4, 8 */
        switch (radeon_agpmode) {
        case -1:
@@ -1018,6 +1006,10 @@ int radeon_device_init(struct radeon_device *rdev,
                return r;
        /* initialize vm here */
        mutex_init(&rdev->vm_manager.lock);
+        /* Adjust VM size here.
+         * Currently set to 4GB ((1 << 20) 4k pages).
+         * Max GPUVM size for cayman and SI is 40 bits.
+         */
        rdev->vm_manager.max_pfn = 1 << 20;
        INIT_LIST_HEAD(&rdev->vm_manager.lru_vm);
diff --git a/drivers/gpu/drm/radeon/radeon_gart.c b/drivers/gpu/drm/radeon/radeon_gart.c
index f0c06d196b75..4debd60e5aa6 100644
--- a/drivers/gpu/drm/radeon/radeon_gart.c
+++ b/drivers/gpu/drm/radeon/radeon_gart.c
@@ -355,14 +355,13 @@ int radeon_gart_init(struct radeon_device *rdev)
        DRM_INFO("GART: num cpu pages %u, num gpu pages %u\n",
                 rdev->gart.num_cpu_pages, rdev->gart.num_gpu_pages);
        /* Allocate pages table */
-        rdev->gart.pages = kzalloc(sizeof(void *) * rdev->gart.num_cpu_pages,
+        rdev->gart.pages = vzalloc(sizeof(void *) * rdev->gart.num_cpu_pages);
-                                   GFP_KERNEL);
        if (rdev->gart.pages == NULL) {
                radeon_gart_fini(rdev);
                return -ENOMEM;
        }
-        rdev->gart.pages_addr = kzalloc(sizeof(dma_addr_t) *
+        rdev->gart.pages_addr = vzalloc(sizeof(dma_addr_t) *
-                                        rdev->gart.num_cpu_pages, GFP_KERNEL);
+                                        rdev->gart.num_cpu_pages);
        if (rdev->gart.pages_addr == NULL) {
                radeon_gart_fini(rdev);
                return -ENOMEM;
@@ -388,8 +387,8 @@ void radeon_gart_fini(struct radeon_device *rdev)
                radeon_gart_unbind(rdev, 0, rdev->gart.num_cpu_pages);
        }
        rdev->gart.ready = false;
-        kfree(rdev->gart.pages);
+        vfree(rdev->gart.pages);
-        kfree(rdev->gart.pages_addr);
+        vfree(rdev->gart.pages_addr);
        rdev->gart.pages = NULL;
        rdev->gart.pages_addr = NULL;
@@ -423,6 +422,18 @@ void radeon_gart_fini(struct radeon_device *rdev)
 */
 /**
+ * radeon_vm_num_pde - return the number of page directory entries
+ *
+ * @rdev: radeon_device pointer
+ *
+ * Calculate the number of page directory entries (cayman+).
+ */
+static unsigned radeon_vm_num_pdes(struct radeon_device *rdev)
+{
+        return rdev->vm_manager.max_pfn >> RADEON_VM_BLOCK_SIZE;
+}
+/**
 * radeon_vm_directory_size - returns the size of the page directory in bytes
 *
 * @rdev: radeon_device pointer
@@ -431,7 +442,7 @@ void radeon_gart_fini(struct radeon_device *rdev)
 */
 static unsigned radeon_vm_directory_size(struct radeon_device *rdev)
 {
-        return (rdev->vm_manager.max_pfn >> RADEON_VM_BLOCK_SIZE) * 8;
+        return RADEON_GPU_PAGE_ALIGN(radeon_vm_num_pdes(rdev) * 8);
 }
 /**
@@ -451,11 +462,11 @@ int radeon_vm_manager_init(struct radeon_device *rdev)
        if (!rdev->vm_manager.enabled) {
                /* allocate enough for 2 full VM pts */
-                size = RADEON_GPU_PAGE_ALIGN(radeon_vm_directory_size(rdev));
+                size = radeon_vm_directory_size(rdev);
-                size += RADEON_GPU_PAGE_ALIGN(rdev->vm_manager.max_pfn * 8);
+                size += rdev->vm_manager.max_pfn * 8;
                size *= 2;
                r = radeon_sa_bo_manager_init(rdev, &rdev->vm_manager.sa_manager,
-                                              size,
+                                              RADEON_GPU_PAGE_ALIGN(size),
                                              RADEON_GEM_DOMAIN_VRAM);
                if (r) {
                        dev_err(rdev->dev, "failed to allocate vm bo (%dKB)\n",
@@ -476,7 +487,7 @@ int radeon_vm_manager_init(struct radeon_device *rdev)
        /* restore page table */
        list_for_each_entry(vm, &rdev->vm_manager.lru_vm, list) {
-                if (vm->sa_bo == NULL)
+                if (vm->page_directory == NULL)
                        continue;
                list_for_each_entry(bo_va, &vm->va, vm_list) {
@@ -500,16 +511,25 @@ static void radeon_vm_free_pt(struct radeon_device *rdev,
                                    struct radeon_vm *vm)
 {
        struct radeon_bo_va *bo_va;
+        int i;
-        if (!vm->sa_bo)
+        if (!vm->page_directory)
                return;
        list_del_init(&vm->list);
-        radeon_sa_bo_free(rdev, &vm->sa_bo, vm->fence);
+        radeon_sa_bo_free(rdev, &vm->page_directory, vm->fence);
        list_for_each_entry(bo_va, &vm->va, vm_list) {
                bo_va->valid = false;
        }
+        if (vm->page_tables == NULL)
+                return;
+        for (i = 0; i < radeon_vm_num_pdes(rdev); i++)
+                radeon_sa_bo_free(rdev, &vm->page_tables[i], vm->fence);
+        kfree(vm->page_tables);
 }
 /**
@@ -546,63 +566,106 @@ void radeon_vm_manager_fini(struct radeon_device *rdev)
 }
 /**
+ * radeon_vm_evict - evict page table to make room for new one
+ *
+ * @rdev: radeon_device pointer
+ * @vm: VM we want to allocate something for
+ *
+ * Evict a VM from the lru, making sure that it isn't @vm. (cayman+).
+ * Returns 0 for success, -ENOMEM for failure.
+ *
+ * Global and local mutex must be locked!
+ */
+static int radeon_vm_evict(struct radeon_device *rdev, struct radeon_vm *vm)
+{
+        struct radeon_vm *vm_evict;
+        if (list_empty(&rdev->vm_manager.lru_vm))
+                return -ENOMEM;
+        vm_evict = list_first_entry(&rdev->vm_manager.lru_vm,
+                                    struct radeon_vm, list);
+        if (vm_evict == vm)
+                return -ENOMEM;
+        mutex_lock(&vm_evict->mutex);
+        radeon_vm_free_pt(rdev, vm_evict);
+        mutex_unlock(&vm_evict->mutex);
+        return 0;
+}
+/**
 * radeon_vm_alloc_pt - allocates a page table for a VM
 *
 * @rdev: radeon_device pointer
 * @vm: vm to bind
 *
 * Allocate a page table for the requested vm (cayman+).
- * Also starts to populate the page table.
 * Returns 0 for success, error for failure.
 *
 * Global and local mutex must be locked!
 */
 int radeon_vm_alloc_pt(struct radeon_device *rdev, struct radeon_vm *vm)
 {
-        struct radeon_vm *vm_evict;
+        unsigned pd_size, pts_size;
-        int r;
        u64 *pd_addr;
-        int tables_size;
+        int r;
        if (vm == NULL) {
                return -EINVAL;
        }
-        /* allocate enough to cover the current VM size */
+        if (vm->page_directory != NULL) {
-        tables_size = RADEON_GPU_PAGE_ALIGN(radeon_vm_directory_size(rdev));
-        tables_size += RADEON_GPU_PAGE_ALIGN(vm->last_pfn * 8);
-        if (vm->sa_bo != NULL) {
-                /* update lru */
-                list_del_init(&vm->list);
-                list_add_tail(&vm->list, &rdev->vm_manager.lru_vm);
                return 0;
        }
 retry:
-        r = radeon_sa_bo_new(rdev, &rdev->vm_manager.sa_manager, &vm->sa_bo,
+        pd_size = RADEON_GPU_PAGE_ALIGN(radeon_vm_directory_size(rdev));
-                             tables_size, RADEON_GPU_PAGE_SIZE, false);
+        r = radeon_sa_bo_new(rdev, &rdev->vm_manager.sa_manager,
+                             &vm->page_directory, pd_size,
+                             RADEON_GPU_PAGE_SIZE, false);
        if (r == -ENOMEM) {
-                if (list_empty(&rdev->vm_manager.lru_vm)) {
+                r = radeon_vm_evict(rdev, vm);
+                if (r)
                        return r;
-                }
-                vm_evict = list_first_entry(&rdev->vm_manager.lru_vm, struct radeon_vm, list);
-                mutex_lock(&vm_evict->mutex);
-                radeon_vm_free_pt(rdev, vm_evict);
-                mutex_unlock(&vm_evict->mutex);
                goto retry;
        } else if (r) {
                return r;
        }
-        pd_addr = radeon_sa_bo_cpu_addr(vm->sa_bo);
+        vm->pd_gpu_addr = radeon_sa_bo_gpu_addr(vm->page_directory);
-        vm->pd_gpu_addr = radeon_sa_bo_gpu_addr(vm->sa_bo);
-        memset(pd_addr, 0, tables_size);
+        /* Initially clear the page directory */
+        pd_addr = radeon_sa_bo_cpu_addr(vm->page_directory);
+        memset(pd_addr, 0, pd_size);
+        pts_size = radeon_vm_num_pdes(rdev) * sizeof(struct radeon_sa_bo *);
+        vm->page_tables = kzalloc(pts_size, GFP_KERNEL);
+        if (vm->page_tables == NULL) {
+                DRM_ERROR("Cannot allocate memory for page table array\n");
+                radeon_sa_bo_free(rdev, &vm->page_directory, vm->fence);
+                return -ENOMEM;
+        }
+        return 0;
+}
+/**
+ * radeon_vm_add_to_lru - add VMs page table to LRU list
+ *
+ * @rdev: radeon_device pointer
+ * @vm: vm to add to LRU
+ *
+ * Add the allocated page table to the LRU list (cayman+).
+ *
+ * Global mutex must be locked!
+ */
+void radeon_vm_add_to_lru(struct radeon_device *rdev, struct radeon_vm *vm)
+{
+        list_del_init(&vm->list);
        list_add_tail(&vm->list, &rdev->vm_manager.lru_vm);
-        return radeon_vm_bo_update_pte(rdev, vm, rdev->ring_tmp_bo.bo,
-                                       &rdev->ring_tmp_bo.bo->tbo.mem);
 }
 /**
@@ -793,20 +856,6 @@ int radeon_vm_bo_set_addr(struct radeon_device *rdev,
        }
        mutex_lock(&vm->mutex);
-        if (last_pfn > vm->last_pfn) {
-                /* release mutex and lock in right order */
-                mutex_unlock(&vm->mutex);
-                mutex_lock(&rdev->vm_manager.lock);
-                mutex_lock(&vm->mutex);
-                /* and check again */
-                if (last_pfn > vm->last_pfn) {
-                        /* grow va space 32M by 32M */
-                        unsigned align = ((32 << 20) >> 12) - 1;
-                        radeon_vm_free_pt(rdev, vm);
-                        vm->last_pfn = (last_pfn + align) & ~align;
-                }
-                mutex_unlock(&rdev->vm_manager.lock);
-        }
        head = &vm->va;
        last_offset = 0;
        list_for_each_entry(tmp, &vm->va, vm_list) {
@@ -865,6 +914,154 @@ uint64_t radeon_vm_map_gart(struct radeon_device *rdev, uint64_t addr)
 }
 /**
+ * radeon_vm_update_pdes - make sure that page directory is valid
+ *
+ * @rdev: radeon_device pointer
+ * @vm: requested vm
+ * @start: start of GPU address range
+ * @end: end of GPU address range
+ *
+ * Allocates new page tables if necessary
+ * and updates the page directory (cayman+).
+ * Returns 0 for success, error for failure.
+ *
+ * Global and local mutex must be locked!
+ */
+static int radeon_vm_update_pdes(struct radeon_device *rdev,
+                                 struct radeon_vm *vm,
+                                 uint64_t start, uint64_t end)
+{
+        static const uint32_t incr = RADEON_VM_PTE_COUNT * 8;
+        uint64_t last_pde = ~0, last_pt = ~0;
+        unsigned count = 0;
+        uint64_t pt_idx;
+        int r;
+        start = (start / RADEON_GPU_PAGE_SIZE) >> RADEON_VM_BLOCK_SIZE;
+        end = (end / RADEON_GPU_PAGE_SIZE) >> RADEON_VM_BLOCK_SIZE;
+        /* walk over the address space and update the page directory */
+        for (pt_idx = start; pt_idx <= end; ++pt_idx) {
+                uint64_t pde, pt;
+                if (vm->page_tables[pt_idx])
+                        continue;
+retry:
+                r = radeon_sa_bo_new(rdev, &rdev->vm_manager.sa_manager,
+                                     &vm->page_tables[pt_idx],
+                                     RADEON_VM_PTE_COUNT * 8,
+                                     RADEON_GPU_PAGE_SIZE, false);
+                if (r == -ENOMEM) {
+                        r = radeon_vm_evict(rdev, vm);
+                        if (r)
+                                return r;
+                        goto retry;
+                } else if (r) {
+                        return r;
+                }
+                pde = vm->pd_gpu_addr + pt_idx * 8;
+                pt = radeon_sa_bo_gpu_addr(vm->page_tables[pt_idx]);
+                if (((last_pde + 8 * count) != pde) ||
+                    ((last_pt + incr * count) != pt)) {
+                        if (count) {
+                                radeon_asic_vm_set_page(rdev, last_pde,
+                                                        last_pt, count, incr,
+                                                        RADEON_VM_PAGE_VALID);
+                        }
+                        count = 1;
+                        last_pde = pde;
+                        last_pt = pt;
+                } else {
+                        ++count;
+                }
+        }
+        if (count) {
+                radeon_asic_vm_set_page(rdev, last_pde, last_pt, count,
+                                        incr, RADEON_VM_PAGE_VALID);
+        }
+        return 0;
+}
+/**
+ * radeon_vm_update_ptes - make sure that page tables are valid
+ *
+ * @rdev: radeon_device pointer
+ * @vm: requested vm
+ * @start: start of GPU address range
+ * @end: end of GPU address range
+ * @dst: destination address to map to
+ * @flags: mapping flags
+ *
+ * Update the page tables in the range @start - @end (cayman+).
+ *
+ * Global and local mutex must be locked!
+ */
+static void radeon_vm_update_ptes(struct radeon_device *rdev,
+                                  struct radeon_vm *vm,
+                                  uint64_t start, uint64_t end,
+                                  uint64_t dst, uint32_t flags)
+{
+        static const uint64_t mask = RADEON_VM_PTE_COUNT - 1;
+        uint64_t last_pte = ~0, last_dst = ~0;
+        unsigned count = 0;
+        uint64_t addr;
+        start = start / RADEON_GPU_PAGE_SIZE;
+        end = end / RADEON_GPU_PAGE_SIZE;
+        /* walk over the address space and update the page tables */
+        for (addr = start; addr < end; ) {
+                uint64_t pt_idx = addr >> RADEON_VM_BLOCK_SIZE;
+                unsigned nptes;
+                uint64_t pte;
+                if ((addr & ~mask) == (end & ~mask))
+                        nptes = end - addr;
+                else
+                        nptes = RADEON_VM_PTE_COUNT - (addr & mask);
+                pte = radeon_sa_bo_gpu_addr(vm->page_tables[pt_idx]);
+                pte += (addr & mask) * 8;
+                if ((last_pte + 8 * count) != pte) {
+                        if (count) {
+                                radeon_asic_vm_set_page(rdev, last_pte,
+                                                        last_dst, count,
+                                                        RADEON_GPU_PAGE_SIZE,
+                                                        flags);
+                        }
+                        count = nptes;
+                        last_pte = pte;
+                        last_dst = dst;
+                } else {
+                        count += nptes;
+                }
+                addr += nptes;
+                dst += nptes * RADEON_GPU_PAGE_SIZE;
+        }
+        if (count) {
+                radeon_asic_vm_set_page(rdev, last_pte, last_dst, count,
+                                        RADEON_GPU_PAGE_SIZE, flags);
+        }
+}
+/**
 * radeon_vm_bo_update_pte - map a bo into the vm page table
 *
 * @rdev: radeon_device pointer
@@ -887,12 +1084,11 @@ int radeon_vm_bo_update_pte(struct radeon_device *rdev,
        struct radeon_semaphore *sem = NULL;
        struct radeon_bo_va *bo_va;
        unsigned nptes, npdes, ndw;
-        uint64_t pe, addr;
+        uint64_t addr;
-        uint64_t pfn;
        int r;
        /* nothing to do if vm isn't bound */
-        if (vm->sa_bo == NULL)
+        if (vm->page_directory == NULL)
                return 0;
        bo_va = radeon_vm_bo_find(vm, bo);
@@ -939,25 +1135,29 @@ int radeon_vm_bo_update_pte(struct radeon_device *rdev,
                }
        }
-        /* estimate number of dw needed */
-        /* reserve space for 32-bit padding */
-        ndw = 32;
        nptes = radeon_bo_ngpu_pages(bo);
-        pfn = (bo_va->soffset / RADEON_GPU_PAGE_SIZE);
+        /* assume two extra pdes in case the mapping overlaps the borders */
+        npdes = (nptes >> RADEON_VM_BLOCK_SIZE) + 2;
-        /* handle cases where a bo spans several pdes  */
+        /* estimate number of dw needed */
-        npdes = (ALIGN(pfn + nptes, RADEON_VM_PTE_COUNT) -
+        /* semaphore, fence and padding */
-                 (pfn & ~(RADEON_VM_PTE_COUNT - 1))) >> RADEON_VM_BLOCK_SIZE;
+        ndw = 32;
+        if (RADEON_VM_BLOCK_SIZE > 11)
+                /* reserve space for one header for every 2k dwords */
+                ndw += (nptes >> 11) * 4;
+        else
+                /* reserve space for one header for
+                    every (1 << BLOCK_SIZE) entries */
+                ndw += (nptes >> RADEON_VM_BLOCK_SIZE) * 4;
-        /* reserve space for one header for every 2k dwords */
-        ndw += (nptes >> 11) * 3;
        /* reserve space for pte addresses */
        ndw += nptes * 2;
        /* reserve space for one header for every 2k dwords */
-        ndw += (npdes >> 11) * 3;
+        ndw += (npdes >> 11) * 4;
        /* reserve space for pde addresses */
        ndw += npdes * 2;
@@ -971,22 +1171,14 @@ int radeon_vm_bo_update_pte(struct radeon_device *rdev,
                radeon_fence_note_sync(vm->fence, ridx);
        }
-        /* update page table entries */
+        r = radeon_vm_update_pdes(rdev, vm, bo_va->soffset, bo_va->eoffset);
-        pe = vm->pd_gpu_addr;
+        if (r) {
-        pe += radeon_vm_directory_size(rdev);
+                radeon_ring_unlock_undo(rdev, ring);
-        pe += (bo_va->soffset / RADEON_GPU_PAGE_SIZE) * 8;
+                return r;
+        }
-        radeon_asic_vm_set_page(rdev, pe, addr, nptes,
-                                RADEON_GPU_PAGE_SIZE, bo_va->flags);
-        /* update page directory entries */
-        addr = pe;
-        pe = vm->pd_gpu_addr;
-        pe += ((bo_va->soffset / RADEON_GPU_PAGE_SIZE) >> RADEON_VM_BLOCK_SIZE) * 8;
-        radeon_asic_vm_set_page(rdev, pe, addr, npdes,
+        radeon_vm_update_ptes(rdev, vm, bo_va->soffset, bo_va->eoffset,
-                                RADEON_VM_PTE_COUNT * 8, RADEON_VM_PAGE_VALID);
+                              addr, bo_va->flags);
        radeon_fence_unref(&vm->fence);
        r = radeon_fence_emit(rdev, &vm->fence, ridx);
@@ -997,6 +1189,7 @@ int radeon_vm_bo_update_pte(struct radeon_device *rdev,
        radeon_ring_unlock_commit(rdev, ring);
        radeon_semaphore_free(rdev, &sem, vm->fence);
        radeon_fence_unref(&vm->last_flush);
        return 0;
 }
@@ -1056,31 +1249,15 @@ void radeon_vm_bo_invalidate(struct radeon_device *rdev,
 * @rdev: radeon_device pointer
 * @vm: requested vm
 *
- * Init @vm (cayman+).
+ * Init @vm fields (cayman+).
- * Map the IB pool and any other shared objects into the VM
- * by default as it's used by all VMs.
- * Returns 0 for success, error for failure.
 */
-int radeon_vm_init(struct radeon_device *rdev, struct radeon_vm *vm)
+void radeon_vm_init(struct radeon_device *rdev, struct radeon_vm *vm)
 {
-        struct radeon_bo_va *bo_va;
-        int r;
        vm->id = 0;
        vm->fence = NULL;
-        vm->last_pfn = 0;
        mutex_init(&vm->mutex);
        INIT_LIST_HEAD(&vm->list);
        INIT_LIST_HEAD(&vm->va);
-        /* map the ib pool buffer at 0 in virtual address space, set
-         * read only
-         */
-        bo_va = radeon_vm_bo_add(rdev, vm, rdev->ring_tmp_bo.bo);
-        r = radeon_vm_bo_set_addr(rdev, bo_va, RADEON_VA_IB_OFFSET,
-                                  RADEON_VM_PAGE_READABLE |
-                                  RADEON_VM_PAGE_SNOOPED);
-        return r;
 }
 /**
@@ -1102,17 +1279,6 @@ void radeon_vm_fini(struct radeon_device *rdev, struct radeon_vm *vm)
        radeon_vm_free_pt(rdev, vm);
        mutex_unlock(&rdev->vm_manager.lock);
-        /* remove all bo at this point non are busy any more because unbind
-         * waited for the last vm fence to signal
-         */
-        r = radeon_bo_reserve(rdev->ring_tmp_bo.bo, false);
-        if (!r) {
-                bo_va = radeon_vm_bo_find(vm, rdev->ring_tmp_bo.bo);
-                list_del_init(&bo_va->bo_list);
-                list_del_init(&bo_va->vm_list);
-                radeon_bo_unreserve(rdev->ring_tmp_bo.bo);
-                kfree(bo_va);
-        }
        if (!list_empty(&vm->va)) {
                dev_err(rdev->dev, "still active bo inside vm\n");
        }
diff --git a/drivers/gpu/drm/radeon/radeon_gem.c b/drivers/gpu/drm/radeon/radeon_gem.c
index f38fbcc46935..fe5c1f6b7957 100644
--- a/drivers/gpu/drm/radeon/radeon_gem.c
+++ b/drivers/gpu/drm/radeon/radeon_gem.c
@@ -53,6 +53,7 @@ int radeon_gem_object_create(struct radeon_device *rdev, int size,
                                struct drm_gem_object **obj)
 {
        struct radeon_bo *robj;
+        unsigned long max_size;
        int r;
        *obj = NULL;
@@ -60,11 +61,26 @@ int radeon_gem_object_create(struct radeon_device *rdev, int size,
        if (alignment < PAGE_SIZE) {
                alignment = PAGE_SIZE;
        }
+        /* maximun bo size is the minimun btw visible vram and gtt size */
+        max_size = min(rdev->mc.visible_vram_size, rdev->mc.gtt_size);
+        if (size > max_size) {
+                printk(KERN_WARNING "%s:%d alloc size %dMb bigger than %ldMb limit\n",
+                       __func__, __LINE__, size >> 20, max_size >> 20);
+                return -ENOMEM;
+        }
+retry:
        r = radeon_bo_create(rdev, size, alignment, kernel, initial_domain, NULL, &robj);
        if (r) {
-                if (r != -ERESTARTSYS)
+                if (r != -ERESTARTSYS) {
+                        if (initial_domain == RADEON_GEM_DOMAIN_VRAM) {
+                                initial_domain |= RADEON_GEM_DOMAIN_GTT;
+                                goto retry;
+                        }
                        DRM_ERROR("Failed to allocate GEM object (%d, %d, %u, %d)\n",
                                  size, initial_domain, alignment, r);
+                }
                return r;
        }
        *obj = &robj->gem_base;
diff --git a/drivers/gpu/drm/radeon/radeon_kms.c b/drivers/gpu/drm/radeon/radeon_kms.c
index 83b8d8aa71c0..dc781c49b96b 100644
--- a/drivers/gpu/drm/radeon/radeon_kms.c
+++ b/drivers/gpu/drm/radeon/radeon_kms.c
@@ -419,6 +419,7 @@ int radeon_driver_open_kms(struct drm_device *dev, struct drm_file *file_priv)
        /* new gpu have virtual address space support */
        if (rdev->family >= CHIP_CAYMAN) {
                struct radeon_fpriv *fpriv;
+                struct radeon_bo_va *bo_va;
                int r;
                fpriv = kzalloc(sizeof(*fpriv), GFP_KERNEL);
@@ -426,7 +427,15 @@ int radeon_driver_open_kms(struct drm_device *dev, struct drm_file *file_priv)
                        return -ENOMEM;
                }
-                r = radeon_vm_init(rdev, &fpriv->vm);
+                radeon_vm_init(rdev, &fpriv->vm);
+                /* map the ib pool buffer read only into
+                 * virtual address space */
+                bo_va = radeon_vm_bo_add(rdev, &fpriv->vm,
+                                         rdev->ring_tmp_bo.bo);
+                r = radeon_vm_bo_set_addr(rdev, bo_va, RADEON_VA_IB_OFFSET,
+                                          RADEON_VM_PAGE_READABLE |
+                                          RADEON_VM_PAGE_SNOOPED);
                if (r) {
                        radeon_vm_fini(rdev, &fpriv->vm);
                        kfree(fpriv);
@@ -454,6 +463,17 @@ void radeon_driver_postclose_kms(struct drm_device *dev,
        /* new gpu have virtual address space support */
        if (rdev->family >= CHIP_CAYMAN && file_priv->driver_priv) {
                struct radeon_fpriv *fpriv = file_priv->driver_priv;
+                struct radeon_bo_va *bo_va;
+                int r;
+                r = radeon_bo_reserve(rdev->ring_tmp_bo.bo, false);
+                if (!r) {
+                        bo_va = radeon_vm_bo_find(&fpriv->vm,
+                                                  rdev->ring_tmp_bo.bo);
+                        if (bo_va)
+                                radeon_vm_bo_rmv(rdev, bo_va);
+                        radeon_bo_unreserve(rdev->ring_tmp_bo.bo);
+                }
                radeon_vm_fini(rdev, &fpriv->vm);
                kfree(fpriv);
diff --git a/drivers/gpu/drm/radeon/radeon_legacy_crtc.c b/drivers/gpu/drm/radeon/radeon_legacy_crtc.c
index 5677a424b585..6857cb4efb76 100644
--- a/drivers/gpu/drm/radeon/radeon_legacy_crtc.c
+++ b/drivers/gpu/drm/radeon/radeon_legacy_crtc.c
@@ -295,6 +295,7 @@ static void radeon_crtc_dpms(struct drm_crtc *crtc, int mode)
        struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc);
        struct drm_device *dev = crtc->dev;
        struct radeon_device *rdev = dev->dev_private;
+        uint32_t crtc_ext_cntl = 0;
        uint32_t mask;
        if (radeon_crtc->crtc_id)
@@ -307,6 +308,16 @@ static void radeon_crtc_dpms(struct drm_crtc *crtc, int mode)
                        RADEON_CRTC_VSYNC_DIS |
                        RADEON_CRTC_HSYNC_DIS);
+        /*
+         * On all dual CRTC GPUs this bit controls the CRTC of the primary DAC.
+         * Therefore it is set in the DAC DMPS function.
+         * This is different for GPU's with a single CRTC but a primary and a
+         * TV DAC: here it controls the single CRTC no matter where it is
+         * routed. Therefore we set it here.
+         */
+        if (rdev->flags & RADEON_SINGLE_CRTC)
+                crtc_ext_cntl = RADEON_CRTC_CRT_ON;
+        
        switch (mode) {
        case DRM_MODE_DPMS_ON:
                radeon_crtc->enabled = true;
@@ -317,7 +328,7 @@ static void radeon_crtc_dpms(struct drm_crtc *crtc, int mode)
                else {
                        WREG32_P(RADEON_CRTC_GEN_CNTL, RADEON_CRTC_EN, ~(RADEON_CRTC_EN |
                                                                         RADEON_CRTC_DISP_REQ_EN_B));
-                        WREG32_P(RADEON_CRTC_EXT_CNTL, 0, ~mask);
+                        WREG32_P(RADEON_CRTC_EXT_CNTL, crtc_ext_cntl, ~(mask | crtc_ext_cntl));
                }
                drm_vblank_post_modeset(dev, radeon_crtc->crtc_id);
                radeon_crtc_load_lut(crtc);
@@ -331,7 +342,7 @@ static void radeon_crtc_dpms(struct drm_crtc *crtc, int mode)
                else {
                        WREG32_P(RADEON_CRTC_GEN_CNTL, RADEON_CRTC_DISP_REQ_EN_B, ~(RADEON_CRTC_EN |
                                                                                    RADEON_CRTC_DISP_REQ_EN_B));
-                        WREG32_P(RADEON_CRTC_EXT_CNTL, mask, ~mask);
+                        WREG32_P(RADEON_CRTC_EXT_CNTL, mask, ~(mask | crtc_ext_cntl));
                }
                radeon_crtc->enabled = false;
                /* adjust pm to dpms changes AFTER disabling crtcs */
diff --git a/drivers/gpu/drm/radeon/radeon_legacy_encoders.c b/drivers/gpu/drm/radeon/radeon_legacy_encoders.c
index 92487e614778..f5ba2241dacc 100644
--- a/drivers/gpu/drm/radeon/radeon_legacy_encoders.c
+++ b/drivers/gpu/drm/radeon/radeon_legacy_encoders.c
@@ -269,27 +269,6 @@ static const struct drm_encoder_helper_funcs radeon_legacy_lvds_helper_funcs = {
        .disable = radeon_legacy_encoder_disable,
 };
-#if defined(CONFIG_BACKLIGHT_CLASS_DEVICE) || defined(CONFIG_BACKLIGHT_CLASS_DEVICE_MODULE)
-static uint8_t radeon_legacy_lvds_level(struct backlight_device *bd)
-{
-        struct radeon_backlight_privdata *pdata = bl_get_data(bd);
-        uint8_t level;
-        /* Convert brightness to hardware level */
-        if (bd->props.brightness < 0)
-                level = 0;
-        else if (bd->props.brightness > RADEON_MAX_BL_LEVEL)
-                level = RADEON_MAX_BL_LEVEL;
-        else
-                level = bd->props.brightness;
-        if (pdata->negative)
-                level = RADEON_MAX_BL_LEVEL - level;
-        return level;
-}
 u8
 radeon_legacy_get_backlight_level(struct radeon_encoder *radeon_encoder)
 {
@@ -331,6 +310,27 @@ radeon_legacy_set_backlight_level(struct radeon_encoder *radeon_encoder, u8 leve
        radeon_legacy_lvds_update(&radeon_encoder->base, dpms_mode);
 }
+#if defined(CONFIG_BACKLIGHT_CLASS_DEVICE) || defined(CONFIG_BACKLIGHT_CLASS_DEVICE_MODULE)
+static uint8_t radeon_legacy_lvds_level(struct backlight_device *bd)
+{
+        struct radeon_backlight_privdata *pdata = bl_get_data(bd);
+        uint8_t level;
+        /* Convert brightness to hardware level */
+        if (bd->props.brightness < 0)
+                level = 0;
+        else if (bd->props.brightness > RADEON_MAX_BL_LEVEL)
+                level = RADEON_MAX_BL_LEVEL;
+        else
+                level = bd->props.brightness;
+        if (pdata->negative)
+                level = RADEON_MAX_BL_LEVEL - level;
+        return level;
+}
 static int radeon_legacy_backlight_update_status(struct backlight_device *bd)
 {
        struct radeon_backlight_privdata *pdata = bl_get_data(bd);
@@ -370,6 +370,7 @@ void radeon_legacy_backlight_init(struct radeon_encoder *radeon_encoder,
        struct backlight_properties props;
        struct radeon_backlight_privdata *pdata;
        uint8_t backlight_level;
+        char bl_name[16];
        if (!radeon_encoder->enc_priv)
                return;
@@ -389,7 +390,9 @@ void radeon_legacy_backlight_init(struct radeon_encoder *radeon_encoder,
        memset(&props, 0, sizeof(props));
        props.max_brightness = RADEON_MAX_BL_LEVEL;
        props.type = BACKLIGHT_RAW;
-        bd = backlight_device_register("radeon_bl", &drm_connector->kdev,
+        snprintf(bl_name, sizeof(bl_name),
+                 "radeon_bl%d", dev->primary->index);
+        bd = backlight_device_register(bl_name, &drm_connector->kdev,
                                       pdata, &radeon_backlight_ops, &props);
        if (IS_ERR(bd)) {
                DRM_ERROR("Backlight registration failed\n");
@@ -534,7 +537,9 @@ static void radeon_legacy_primary_dac_dpms(struct drm_encoder *encoder, int mode
                break;
        }
-        WREG32(RADEON_CRTC_EXT_CNTL, crtc_ext_cntl);
+        /* handled in radeon_crtc_dpms() */
+        if (!(rdev->flags & RADEON_SINGLE_CRTC))
+                WREG32(RADEON_CRTC_EXT_CNTL, crtc_ext_cntl);
        WREG32(RADEON_DAC_CNTL, dac_cntl);
        WREG32(RADEON_DAC_MACRO_CNTL, dac_macro_cntl);
@@ -659,6 +664,8 @@ static enum drm_connector_status radeon_legacy_primary_dac_detect(struct drm_enc
        if (ASIC_IS_R300(rdev))
                tmp |= (0x1b6 << RADEON_DAC_FORCE_DATA_SHIFT);
+        else if (ASIC_IS_RV100(rdev))
+                tmp |= (0x1ac << RADEON_DAC_FORCE_DATA_SHIFT);
        else
                tmp |= (0x180 << RADEON_DAC_FORCE_DATA_SHIFT);
@@ -668,6 +675,7 @@ static enum drm_connector_status radeon_legacy_primary_dac_detect(struct drm_enc
        tmp |= RADEON_DAC_RANGE_CNTL_PS2 | RADEON_DAC_CMP_EN;
        WREG32(RADEON_DAC_CNTL, tmp);
+        tmp = dac_macro_cntl;
        tmp &= ~(RADEON_DAC_PDWN_R |
                 RADEON_DAC_PDWN_G |
                 RADEON_DAC_PDWN_B);
@@ -991,11 +999,7 @@ static void radeon_legacy_tmds_ext_mode_set(struct drm_encoder *encoder,
 static void radeon_ext_tmds_enc_destroy(struct drm_encoder *encoder)
 {
        struct radeon_encoder *radeon_encoder = to_radeon_encoder(encoder);
-        struct radeon_encoder_ext_tmds *tmds = radeon_encoder->enc_priv;
+        /* don't destroy the i2c bus record here, this will be done in radeon_i2c_fini */
-        if (tmds) {
-                if (tmds->i2c_bus)
-                        radeon_i2c_destroy(tmds->i2c_bus);
-        }
        kfree(radeon_encoder->enc_priv);
        drm_encoder_cleanup(encoder);
        kfree(radeon_encoder);
@@ -1093,7 +1097,8 @@ static void radeon_legacy_tv_dac_dpms(struct drm_encoder *encoder, int mode)
        } else {
                if (is_tv)
                        WREG32(RADEON_TV_MASTER_CNTL, tv_master_cntl);
-                else
+                /* handled in radeon_crtc_dpms() */
+                else if (!(rdev->flags & RADEON_SINGLE_CRTC))
                        WREG32(RADEON_CRTC2_GEN_CNTL, crtc2_gen_cntl);
                WREG32(RADEON_TV_DAC_CNTL, tv_dac_cntl);
        }
@@ -1417,13 +1422,104 @@ static bool radeon_legacy_tv_detect(struct drm_encoder *encoder,
        return found;
 }
+static bool radeon_legacy_ext_dac_detect(struct drm_encoder *encoder,
+                                         struct drm_connector *connector)
+{
+        struct drm_device *dev = encoder->dev;
+        struct radeon_device *rdev = dev->dev_private;
+        uint32_t gpio_monid, fp2_gen_cntl, disp_output_cntl, crtc2_gen_cntl;
+        uint32_t disp_lin_trans_grph_a, disp_lin_trans_grph_b, disp_lin_trans_grph_c;
+        uint32_t disp_lin_trans_grph_d, disp_lin_trans_grph_e, disp_lin_trans_grph_f;
+        uint32_t tmp, crtc2_h_total_disp, crtc2_v_total_disp;
+        uint32_t crtc2_h_sync_strt_wid, crtc2_v_sync_strt_wid;
+        bool found = false;
+        int i;
+        /* save the regs we need */
+        gpio_monid = RREG32(RADEON_GPIO_MONID);
+        fp2_gen_cntl = RREG32(RADEON_FP2_GEN_CNTL);
+        disp_output_cntl = RREG32(RADEON_DISP_OUTPUT_CNTL);
+        crtc2_gen_cntl = RREG32(RADEON_CRTC2_GEN_CNTL);
+        disp_lin_trans_grph_a = RREG32(RADEON_DISP_LIN_TRANS_GRPH_A);
+        disp_lin_trans_grph_b = RREG32(RADEON_DISP_LIN_TRANS_GRPH_B);
+        disp_lin_trans_grph_c = RREG32(RADEON_DISP_LIN_TRANS_GRPH_C);
+        disp_lin_trans_grph_d = RREG32(RADEON_DISP_LIN_TRANS_GRPH_D);
+        disp_lin_trans_grph_e = RREG32(RADEON_DISP_LIN_TRANS_GRPH_E);
+        disp_lin_trans_grph_f = RREG32(RADEON_DISP_LIN_TRANS_GRPH_F);
+        crtc2_h_total_disp = RREG32(RADEON_CRTC2_H_TOTAL_DISP);
+        crtc2_v_total_disp = RREG32(RADEON_CRTC2_V_TOTAL_DISP);
+        crtc2_h_sync_strt_wid = RREG32(RADEON_CRTC2_H_SYNC_STRT_WID);
+        crtc2_v_sync_strt_wid = RREG32(RADEON_CRTC2_V_SYNC_STRT_WID);
+        tmp = RREG32(RADEON_GPIO_MONID);
+        tmp &= ~RADEON_GPIO_A_0;
+        WREG32(RADEON_GPIO_MONID, tmp);
+        WREG32(RADEON_FP2_GEN_CNTL, (RADEON_FP2_ON |
+                                     RADEON_FP2_PANEL_FORMAT |
+                                     R200_FP2_SOURCE_SEL_TRANS_UNIT |
+                                     RADEON_FP2_DVO_EN |
+                                     R200_FP2_DVO_RATE_SEL_SDR));
+        WREG32(RADEON_DISP_OUTPUT_CNTL, (RADEON_DISP_DAC_SOURCE_RMX |
+                                         RADEON_DISP_TRANS_MATRIX_GRAPHICS));
+        WREG32(RADEON_CRTC2_GEN_CNTL, (RADEON_CRTC2_EN |
+                                       RADEON_CRTC2_DISP_REQ_EN_B));
+        WREG32(RADEON_DISP_LIN_TRANS_GRPH_A, 0x00000000);
+        WREG32(RADEON_DISP_LIN_TRANS_GRPH_B, 0x000003f0);
+        WREG32(RADEON_DISP_LIN_TRANS_GRPH_C, 0x00000000);
+        WREG32(RADEON_DISP_LIN_TRANS_GRPH_D, 0x000003f0);
+        WREG32(RADEON_DISP_LIN_TRANS_GRPH_E, 0x00000000);
+        WREG32(RADEON_DISP_LIN_TRANS_GRPH_F, 0x000003f0);
+        WREG32(RADEON_CRTC2_H_TOTAL_DISP, 0x01000008);
+        WREG32(RADEON_CRTC2_H_SYNC_STRT_WID, 0x00000800);
+        WREG32(RADEON_CRTC2_V_TOTAL_DISP, 0x00080001);
+        WREG32(RADEON_CRTC2_V_SYNC_STRT_WID, 0x00000080);
+        for (i = 0; i < 200; i++) {
+                tmp = RREG32(RADEON_GPIO_MONID);
+                if (tmp & RADEON_GPIO_Y_0)
+                        found = true;
+                if (found)
+                        break;
+                if (!drm_can_sleep())
+                        mdelay(1);
+                else
+                        msleep(1);
+        }
+        /* restore the regs we used */
+        WREG32(RADEON_DISP_LIN_TRANS_GRPH_A, disp_lin_trans_grph_a);
+        WREG32(RADEON_DISP_LIN_TRANS_GRPH_B, disp_lin_trans_grph_b);
+        WREG32(RADEON_DISP_LIN_TRANS_GRPH_C, disp_lin_trans_grph_c);
+        WREG32(RADEON_DISP_LIN_TRANS_GRPH_D, disp_lin_trans_grph_d);
+        WREG32(RADEON_DISP_LIN_TRANS_GRPH_E, disp_lin_trans_grph_e);
+        WREG32(RADEON_DISP_LIN_TRANS_GRPH_F, disp_lin_trans_grph_f);
+        WREG32(RADEON_CRTC2_H_TOTAL_DISP, crtc2_h_total_disp);
+        WREG32(RADEON_CRTC2_V_TOTAL_DISP, crtc2_v_total_disp);
+        WREG32(RADEON_CRTC2_H_SYNC_STRT_WID, crtc2_h_sync_strt_wid);
+        WREG32(RADEON_CRTC2_V_SYNC_STRT_WID, crtc2_v_sync_strt_wid);
+        WREG32(RADEON_CRTC2_GEN_CNTL, crtc2_gen_cntl);
+        WREG32(RADEON_DISP_OUTPUT_CNTL, disp_output_cntl);
+        WREG32(RADEON_FP2_GEN_CNTL, fp2_gen_cntl);
+        WREG32(RADEON_GPIO_MONID, gpio_monid);
+        return found;
+}
 static enum drm_connector_status radeon_legacy_tv_dac_detect(struct drm_encoder *encoder,
                                                             struct drm_connector *connector)
 {
        struct drm_device *dev = encoder->dev;
        struct radeon_device *rdev = dev->dev_private;
-        uint32_t crtc2_gen_cntl, tv_dac_cntl, dac_cntl2, dac_ext_cntl;
+        uint32_t crtc2_gen_cntl = 0, tv_dac_cntl, dac_cntl2, dac_ext_cntl;
-        uint32_t disp_hw_debug, disp_output_cntl, gpiopad_a, pixclks_cntl, tmp;
+        uint32_t gpiopad_a = 0, pixclks_cntl, tmp;
+        uint32_t disp_output_cntl = 0, disp_hw_debug = 0, crtc_ext_cntl = 0;
        enum drm_connector_status found = connector_status_disconnected;
        struct radeon_encoder *radeon_encoder = to_radeon_encoder(encoder);
        struct radeon_encoder_tv_dac *tv_dac = radeon_encoder->enc_priv;
@@ -1460,12 +1556,27 @@ static enum drm_connector_status radeon_legacy_tv_dac_detect(struct drm_encoder
                return connector_status_disconnected;
        }
+        /* R200 uses an external DAC for secondary DAC */
+        if (rdev->family == CHIP_R200) {
+                if (radeon_legacy_ext_dac_detect(encoder, connector))
+                        found = connector_status_connected;
+                return found;
+        }
        /* save the regs we need */
        pixclks_cntl = RREG32_PLL(RADEON_PIXCLKS_CNTL);
-        gpiopad_a = ASIC_IS_R300(rdev) ? RREG32(RADEON_GPIOPAD_A) : 0;
-        disp_output_cntl = ASIC_IS_R300(rdev) ? RREG32(RADEON_DISP_OUTPUT_CNTL) : 0;
+        if (rdev->flags & RADEON_SINGLE_CRTC) {
-        disp_hw_debug = ASIC_IS_R300(rdev) ? 0 : RREG32(RADEON_DISP_HW_DEBUG);
+                crtc_ext_cntl = RREG32(RADEON_CRTC_EXT_CNTL);
-        crtc2_gen_cntl = RREG32(RADEON_CRTC2_GEN_CNTL);
+        } else {
+                if (ASIC_IS_R300(rdev)) {
+                        gpiopad_a = RREG32(RADEON_GPIOPAD_A);
+                        disp_output_cntl = RREG32(RADEON_DISP_OUTPUT_CNTL);
+                } else {
+                        disp_hw_debug = RREG32(RADEON_DISP_HW_DEBUG);
+                }
+                crtc2_gen_cntl = RREG32(RADEON_CRTC2_GEN_CNTL);
+        }
        tv_dac_cntl = RREG32(RADEON_TV_DAC_CNTL);
        dac_ext_cntl = RREG32(RADEON_DAC_EXT_CNTL);
        dac_cntl2 = RREG32(RADEON_DAC_CNTL2);
@@ -1474,22 +1585,24 @@ static enum drm_connector_status radeon_legacy_tv_dac_detect(struct drm_encoder
                               | RADEON_PIX2CLK_DAC_ALWAYS_ONb);
        WREG32_PLL(RADEON_PIXCLKS_CNTL, tmp);
-        if (ASIC_IS_R300(rdev))
+        if (rdev->flags & RADEON_SINGLE_CRTC) {
-                WREG32_P(RADEON_GPIOPAD_A, 1, ~1);
+                tmp = crtc_ext_cntl | RADEON_CRTC_CRT_ON;
+                WREG32(RADEON_CRTC_EXT_CNTL, tmp);
-        tmp = crtc2_gen_cntl & ~RADEON_CRTC2_PIX_WIDTH_MASK;
-        tmp |= RADEON_CRTC2_CRT2_ON |
-                (2 << RADEON_CRTC2_PIX_WIDTH_SHIFT);
-        WREG32(RADEON_CRTC2_GEN_CNTL, tmp);
-        if (ASIC_IS_R300(rdev)) {
-                tmp = disp_output_cntl & ~RADEON_DISP_TVDAC_SOURCE_MASK;
-                tmp |= RADEON_DISP_TVDAC_SOURCE_CRTC2;
-                WREG32(RADEON_DISP_OUTPUT_CNTL, tmp);
        } else {
-                tmp = disp_hw_debug & ~RADEON_CRT2_DISP1_SEL;
+                tmp = crtc2_gen_cntl & ~RADEON_CRTC2_PIX_WIDTH_MASK;
-                WREG32(RADEON_DISP_HW_DEBUG, tmp);
+                tmp |= RADEON_CRTC2_CRT2_ON |
+                        (2 << RADEON_CRTC2_PIX_WIDTH_SHIFT);
+                WREG32(RADEON_CRTC2_GEN_CNTL, tmp);
+                if (ASIC_IS_R300(rdev)) {
+                        WREG32_P(RADEON_GPIOPAD_A, 1, ~1);
+                        tmp = disp_output_cntl & ~RADEON_DISP_TVDAC_SOURCE_MASK;
+                        tmp |= RADEON_DISP_TVDAC_SOURCE_CRTC2;
+                        WREG32(RADEON_DISP_OUTPUT_CNTL, tmp);
+                } else {
+                        tmp = disp_hw_debug & ~RADEON_CRT2_DISP1_SEL;
+                        WREG32(RADEON_DISP_HW_DEBUG, tmp);
+                }
        }
        tmp = RADEON_TV_DAC_NBLANK |
@@ -1531,14 +1644,19 @@ static enum drm_connector_status radeon_legacy_tv_dac_detect(struct drm_encoder
        WREG32(RADEON_DAC_CNTL2, dac_cntl2);
        WREG32(RADEON_DAC_EXT_CNTL, dac_ext_cntl);
        WREG32(RADEON_TV_DAC_CNTL, tv_dac_cntl);
-        WREG32(RADEON_CRTC2_GEN_CNTL, crtc2_gen_cntl);
-        if (ASIC_IS_R300(rdev)) {
+        if (rdev->flags & RADEON_SINGLE_CRTC) {
-                WREG32(RADEON_DISP_OUTPUT_CNTL, disp_output_cntl);
+                WREG32(RADEON_CRTC_EXT_CNTL, crtc_ext_cntl);
-                WREG32_P(RADEON_GPIOPAD_A, gpiopad_a, ~1);
        } else {
-                WREG32(RADEON_DISP_HW_DEBUG, disp_hw_debug);
+                WREG32(RADEON_CRTC2_GEN_CNTL, crtc2_gen_cntl);
+                if (ASIC_IS_R300(rdev)) {
+                        WREG32(RADEON_DISP_OUTPUT_CNTL, disp_output_cntl);
+                        WREG32_P(RADEON_GPIOPAD_A, gpiopad_a, ~1);
+                } else {
+                        WREG32(RADEON_DISP_HW_DEBUG, disp_hw_debug);
+                }
        }
        WREG32_PLL(RADEON_PIXCLKS_CNTL, pixclks_cntl);
        return found;
diff --git a/drivers/gpu/drm/radeon/radeon_object.c b/drivers/gpu/drm/radeon/radeon_object.c
index 8b27dd6e3144..b91118ccef86 100644
--- a/drivers/gpu/drm/radeon/radeon_object.c
+++ b/drivers/gpu/drm/radeon/radeon_object.c
@@ -105,7 +105,6 @@ int radeon_bo_create(struct radeon_device *rdev,
        struct radeon_bo *bo;
        enum ttm_bo_type type;
        unsigned long page_align = roundup(byte_align, PAGE_SIZE) >> PAGE_SHIFT;
-        unsigned long max_size = 0;
        size_t acc_size;
        int r;
@@ -121,18 +120,9 @@ int radeon_bo_create(struct radeon_device *rdev,
        }
        *bo_ptr = NULL;
-        /* maximun bo size is the minimun btw visible vram and gtt size */
-        max_size = min(rdev->mc.visible_vram_size, rdev->mc.gtt_size);
-        if ((page_align << PAGE_SHIFT) >= max_size) {
-                printk(KERN_WARNING "%s:%d alloc size %ldM bigger than %ldMb limit\n",
-                        __func__, __LINE__, page_align  >> (20 - PAGE_SHIFT), max_size >> 20);
-                return -ENOMEM;
-        }
        acc_size = ttm_bo_dma_acc_size(&rdev->mman.bdev, size,
                                       sizeof(struct radeon_bo));
-retry:
        bo = kzalloc(sizeof(struct radeon_bo), GFP_KERNEL);
        if (bo == NULL)
                return -ENOMEM;
@@ -154,15 +144,6 @@ retry:
                        acc_size, sg, &radeon_ttm_bo_destroy);
        up_read(&rdev->pm.mclk_lock);
        if (unlikely(r != 0)) {
-                if (r != -ERESTARTSYS) {
-                        if (domain == RADEON_GEM_DOMAIN_VRAM) {
-                                domain |= RADEON_GEM_DOMAIN_GTT;
-                                goto retry;
-                        }
-                        dev_err(rdev->dev,
-                                "object_init failed for (%lu, 0x%08X)\n",
-                                size, domain);
-                }
                return r;
        }
        *bo_ptr = bo;
diff --git a/drivers/gpu/drm/radeon/radeon_ring.c b/drivers/gpu/drm/radeon/radeon_ring.c
index bba66902c83b..47634f27f2e5 100644
--- a/drivers/gpu/drm/radeon/radeon_ring.c
+++ b/drivers/gpu/drm/radeon/radeon_ring.c
@@ -305,7 +305,7 @@ void radeon_ring_write(struct radeon_ring *ring, uint32_t v)
 {
 #if DRM_DEBUG_CODE
        if (ring->count_dw <= 0) {
-                DRM_ERROR("radeon: writting more dword to ring than expected !\n");
+                DRM_ERROR("radeon: writing more dwords to the ring than expected!\n");
        }
 #endif
        ring->ring[ring->wptr++] = v;
diff --git a/drivers/gpu/drm/radeon/si.c b/drivers/gpu/drm/radeon/si.c
index f79633a036c3..b0db712060fb 100644
--- a/drivers/gpu/drm/radeon/si.c
+++ b/drivers/gpu/drm/radeon/si.c
@@ -2407,12 +2407,13 @@ static int si_pcie_gart_enable(struct radeon_device *rdev)
        WREG32(0x15DC, 0);
        /* empty context1-15 */
-        /* FIXME start with 4G, once using 2 level pt switch to full
-         * vm size space
-         */
        /* set vm size, must be a multiple of 4 */
        WREG32(VM_CONTEXT1_PAGE_TABLE_START_ADDR, 0);
        WREG32(VM_CONTEXT1_PAGE_TABLE_END_ADDR, rdev->vm_manager.max_pfn);
+        /* Assign the pt base to something valid for now; the pts used for
+         * the VMs are determined by the application and setup and assigned
+         * on the fly in the vm part of radeon_gart.c
+         */
        for (i = 1; i < 16; i++) {
                if (i < 8)
                        WREG32(VM_CONTEXT0_PAGE_TABLE_BASE_ADDR + (i << 2),
@@ -2807,26 +2808,31 @@ void si_vm_set_page(struct radeon_device *rdev, uint64_t pe,
 {
        struct radeon_ring *ring = &rdev->ring[rdev->asic->vm.pt_ring_index];
        uint32_t r600_flags = cayman_vm_page_flags(rdev, flags);
-        int i;
-        uint64_t value;
-        radeon_ring_write(ring, PACKET3(PACKET3_WRITE_DATA, 2 + count * 2));
+        while (count) {
-        radeon_ring_write(ring, (WRITE_DATA_ENGINE_SEL(0) |
+                unsigned ndw = 2 + count * 2;
-                                 WRITE_DATA_DST_SEL(1)));
+                if (ndw > 0x3FFE)
-        radeon_ring_write(ring, pe);
+                        ndw = 0x3FFE;
-        radeon_ring_write(ring, upper_32_bits(pe));
-        for (i = 0; i < count; ++i) {
+                radeon_ring_write(ring, PACKET3(PACKET3_WRITE_DATA, ndw));
-                if (flags & RADEON_VM_PAGE_SYSTEM) {
+                radeon_ring_write(ring, (WRITE_DATA_ENGINE_SEL(0) |
-                        value = radeon_vm_map_gart(rdev, addr);
+                                         WRITE_DATA_DST_SEL(1)));
-                        value &= 0xFFFFFFFFFFFFF000ULL;
+                radeon_ring_write(ring, pe);
-                } else if (flags & RADEON_VM_PAGE_VALID)
+                radeon_ring_write(ring, upper_32_bits(pe));
-                        value = addr;
+                for (; ndw > 2; ndw -= 2, --count, pe += 8) {
-                else
+                        uint64_t value;
-                        value = 0;
+                        if (flags & RADEON_VM_PAGE_SYSTEM) {
-                addr += incr;
+                                value = radeon_vm_map_gart(rdev, addr);
-                value |= r600_flags;
+                                value &= 0xFFFFFFFFFFFFF000ULL;
-                radeon_ring_write(ring, value);
+                        } else if (flags & RADEON_VM_PAGE_VALID)
-                radeon_ring_write(ring, upper_32_bits(value));
+                                value = addr;
+                        else
+                                value = 0;
+                        addr += incr;
+                        value |= r600_flags;
+                        radeon_ring_write(ring, value);
+                        radeon_ring_write(ring, upper_32_bits(value));
+                }
        }
 }
@@ -2867,6 +2873,10 @@ void si_vm_flush(struct radeon_device *rdev, int ridx, struct radeon_vm *vm)
        radeon_ring_write(ring, VM_INVALIDATE_REQUEST >> 2);
        radeon_ring_write(ring, 0);
        radeon_ring_write(ring, 1 << vm->id);
+        /* sync PFP to ME, otherwise we might get invalid PFP reads */
+        radeon_ring_write(ring, PACKET3(PACKET3_PFP_SYNC_ME, 0));
+        radeon_ring_write(ring, 0x0);
 }
 /*
diff --git a/drivers/gpu/drm/shmobile/shmob_drm_drv.c b/drivers/gpu/drm/shmobile/shmob_drm_drv.c
index c71d493fd0c5..1c350fc4e449 100644
--- a/drivers/gpu/drm/shmobile/shmob_drm_drv.c
+++ b/drivers/gpu/drm/shmobile/shmob_drm_drv.c
@@ -201,6 +201,8 @@ static int shmob_drm_load(struct drm_device *dev, unsigned long flags)
                goto done;
        }
+        platform_set_drvdata(pdev, sdev);
 done:
        if (ret)
                shmob_drm_unload(dev);
@@ -299,11 +301,9 @@ static struct drm_driver shmob_drm_driver = {
 #if CONFIG_PM_SLEEP
 static int shmob_drm_pm_suspend(struct device *dev)
 {
-        struct platform_device *pdev = to_platform_device(dev);
+        struct shmob_drm_device *sdev = dev_get_drvdata(dev);
-        struct drm_device *ddev = platform_get_drvdata(pdev);
-        struct shmob_drm_device *sdev = ddev->dev_private;
-        drm_kms_helper_poll_disable(ddev);
+        drm_kms_helper_poll_disable(sdev->ddev);
        shmob_drm_crtc_suspend(&sdev->crtc);
        return 0;
@@ -311,9 +311,7 @@ static int shmob_drm_pm_suspend(struct device *dev)
 static int shmob_drm_pm_resume(struct device *dev)
 {
-        struct platform_device *pdev = to_platform_device(dev);
+        struct shmob_drm_device *sdev = dev_get_drvdata(dev);
-        struct drm_device *ddev = platform_get_drvdata(pdev);
-        struct shmob_drm_device *sdev = ddev->dev_private;
        mutex_lock(&sdev->ddev->mode_config.mutex);
        shmob_drm_crtc_resume(&sdev->crtc);
diff --git a/drivers/gpu/drm/ttm/ttm_bo.c b/drivers/gpu/drm/ttm/ttm_bo.c
index 402ab69f9f99..bf6e4b5a73b5 100644
--- a/drivers/gpu/drm/ttm/ttm_bo.c
+++ b/drivers/gpu/drm/ttm/ttm_bo.c
@@ -580,6 +580,7 @@ retry:
        if (unlikely(ret != 0))
                return ret;
+retry_reserve:
        spin_lock(&glob->lru_lock);
        if (unlikely(list_empty(&bo->ddestroy))) {
@@ -587,14 +588,20 @@ retry:
                return 0;
        }
-        ret = ttm_bo_reserve_locked(bo, interruptible,
+        ret = ttm_bo_reserve_locked(bo, false, true, false, 0);
-                                    no_wait_reserve, false, 0);
-        if (unlikely(ret != 0)) {
+        if (unlikely(ret == -EBUSY)) {
                spin_unlock(&glob->lru_lock);
-                return ret;
+                if (likely(!no_wait_reserve))
+                        ret = ttm_bo_wait_unreserved(bo, interruptible);
+                if (unlikely(ret != 0))
+                        return ret;
+                goto retry_reserve;
        }
+        BUG_ON(ret != 0);
        /**
         * We can re-check for sync object without taking
         * the bo::lock since setting the sync object requires
@@ -811,17 +818,14 @@ retry:
                                          no_wait_reserve, no_wait_gpu);
                kref_put(&bo->list_kref, ttm_bo_release_list);
-                if (likely(ret == 0 || ret == -ERESTARTSYS))
+                return ret;
-                        return ret;
-                goto retry;
        }
-        ret = ttm_bo_reserve_locked(bo, false, no_wait_reserve, false, 0);
+        ret = ttm_bo_reserve_locked(bo, false, true, false, 0);
        if (unlikely(ret == -EBUSY)) {
                spin_unlock(&glob->lru_lock);
-                if (likely(!no_wait_gpu))
+                if (likely(!no_wait_reserve))
                        ret = ttm_bo_wait_unreserved(bo, interruptible);
                kref_put(&bo->list_kref, ttm_bo_release_list);
diff --git a/drivers/gpu/drm/udl/udl_drv.h b/drivers/gpu/drm/udl/udl_drv.h
index fccd361f7b50..87aa5f5d3c88 100644
--- a/drivers/gpu/drm/udl/udl_drv.h
+++ b/drivers/gpu/drm/udl/udl_drv.h
@@ -104,7 +104,7 @@ udl_fb_user_fb_create(struct drm_device *dev,
 int udl_render_hline(struct drm_device *dev, int bpp, struct urb **urb_ptr,
                     const char *front, char **urb_buf_ptr,
-                     u32 byte_offset, u32 byte_width,
+                     u32 byte_offset, u32 device_byte_offset, u32 byte_width,
                     int *ident_ptr, int *sent_ptr);
 int udl_dumb_create(struct drm_file *file_priv,
diff --git a/drivers/gpu/drm/udl/udl_fb.c b/drivers/gpu/drm/udl/udl_fb.c
index 69a2b16f42a6..d4ab3beaada0 100644
--- a/drivers/gpu/drm/udl/udl_fb.c
+++ b/drivers/gpu/drm/udl/udl_fb.c
@@ -114,9 +114,10 @@ static void udlfb_dpy_deferred_io(struct fb_info *info,
        list_for_each_entry(cur, &fbdefio->pagelist, lru) {
                if (udl_render_hline(dev, (ufbdev->ufb.base.bits_per_pixel / 8),
-                                  &urb, (char *) info->fix.smem_start,
+                                     &urb, (char *) info->fix.smem_start,
-                                  &cmd, cur->index << PAGE_SHIFT,
+                                     &cmd, cur->index << PAGE_SHIFT,
-                                  PAGE_SIZE, &bytes_identical, &bytes_sent))
+                                     cur->index << PAGE_SHIFT,
+                                     PAGE_SIZE, &bytes_identical, &bytes_sent))
                        goto error;
                bytes_rendered += PAGE_SIZE;
        }
@@ -187,10 +188,11 @@ int udl_handle_damage(struct udl_framebuffer *fb, int x, int y,
        for (i = y; i < y + height ; i++) {
                const int line_offset = fb->base.pitches[0] * i;
                const int byte_offset = line_offset + (x * bpp);
+                const int dev_byte_offset = (fb->base.width * bpp * i) + (x * bpp);
                if (udl_render_hline(dev, bpp, &urb,
                                     (char *) fb->obj->vmapping,
-                                     &cmd, byte_offset, width * bpp,
+                                     &cmd, byte_offset, dev_byte_offset,
+                                     width * bpp,
                                     &bytes_identical, &bytes_sent))
                        goto error;
        }
diff --git a/drivers/gpu/drm/udl/udl_transfer.c b/drivers/gpu/drm/udl/udl_transfer.c
index dc095526ffb7..142fee5f983f 100644
--- a/drivers/gpu/drm/udl/udl_transfer.c
+++ b/drivers/gpu/drm/udl/udl_transfer.c
@@ -213,11 +213,12 @@ static void udl_compress_hline16(
 */
 int udl_render_hline(struct drm_device *dev, int bpp, struct urb **urb_ptr,
                     const char *front, char **urb_buf_ptr,
-                     u32 byte_offset, u32 byte_width,
+                     u32 byte_offset, u32 device_byte_offset,
+                     u32 byte_width,
                     int *ident_ptr, int *sent_ptr)
 {
        const u8 *line_start, *line_end, *next_pixel;
-        u32 base16 = 0 + (byte_offset / bpp) * 2;
+        u32 base16 = 0 + (device_byte_offset / bpp) * 2;
        struct urb *urb = *urb_ptr;
        u8 *cmd = *urb_buf_ptr;
        u8 *cmd_end = (u8 *) urb->transfer_buffer + urb->transfer_buffer_length;