Merge remote-tracking branch 'pfdo/drm-next' into drm-next

Pull in drm-next for the object find API changes.

Fix the one place the API crashes.

Signed-off-by: Dave Airlie <airlied@redhat.com>

Documentation/devicetree/bindings/display/bridge/adi,adv7511.txt

@@ -68,6 +68,8 @@ Optional properties:
 - adi,disable-timing-generator: Only for ADV7533. Disables the internal timing
   generator. The chip will rely on the sync signals in the DSI data lanes,
   rather than generate its own timings for HDMI output.
+- clocks: from common clock binding: reference to the CEC clock.
+- clock-names: from common clock binding: must be "cec".
 
 Required nodes:
 
@@ -89,6 +91,8 @@ Example
 		reg = <39>;
 		interrupt-parent = <&gpio3>;
 		interrupts = <29 IRQ_TYPE_EDGE_FALLING>;
+		clocks = <&cec_clock>;
+		clock-names = "cec";
 
 		adi,input-depth = <8>;
 		adi,input-colorspace = "rgb";

Documentation/devicetree/bindings/display/bridge/sii9234.txt

@@ -0,0 +1,49 @@
+Silicon Image SiI9234 HDMI/MHL bridge bindings
+
+Required properties:
+	- compatible : "sil,sii9234".
+	- reg : I2C address for TPI interface, use 0x39
+	- avcc33-supply : MHL/USB Switch Supply Voltage (3.3V)
+	- iovcc18-supply : I/O Supply Voltage (1.8V)
+	- avcc12-supply : TMDS Analog Supply Voltage (1.2V)
+	- cvcc12-supply : Digital Core Supply Voltage (1.2V)
+	- interrupts, interrupt-parent: interrupt specifier of INT pin
+	- reset-gpios: gpio specifier of RESET pin (active low)
+	- video interfaces: Device node can contain two video interface port
+			    nodes for HDMI encoder and connector according to [1].
+			    - port@0 - MHL to HDMI
+			    - port@1 - MHL to connector
+
+[1]: Documentation/devicetree/bindings/media/video-interfaces.txt
+
+
+Example:
+	sii9234@39 {
+		compatible = "sil,sii9234";
+		reg = <0x39>;
+		avcc33-supply = <&vcc33mhl>;
+		iovcc18-supply = <&vcc18mhl>;
+		avcc12-supply = <&vsil12>;
+		cvcc12-supply = <&vsil12>;
+		reset-gpios = <&gpf3 4 GPIO_ACTIVE_LOW>;
+		interrupt-parent = <&gpf3>;
+		interrupts = <5 IRQ_TYPE_LEVEL_HIGH>;
+
+		ports {
+			#address-cells = <1>;
+			#size-cells = <0>;
+
+			port@0 {
+				reg = <0>;
+				mhl_to_hdmi: endpoint {
+					remote-endpoint = <&hdmi_to_mhl>;
+				};
+			};
+			port@1 {
+				reg = <1>;
+				mhl_to_connector: endpoint {
+					remote-endpoint = <&connector_to_mhl>;
+				};
+			};
+		};
+	};

Documentation/devicetree/bindings/display/panel/raspberrypi,7inch-touchscreen.txt

@@ -0,0 +1,49 @@
+This binding covers the official 7" (800x480) Raspberry Pi touchscreen
+panel.
+
+This DSI panel contains:
+
+- TC358762 DSI->DPI bridge
+- Atmel microcontroller on I2C for power sequencing the DSI bridge and
+  controlling backlight
+- Touchscreen controller on I2C for touch input
+
+and this binding covers the DSI display parts but not its touch input.
+
+Required properties:
+- compatible:	Must be "raspberrypi,7inch-touchscreen-panel"
+- reg:		Must be "45"
+- port:		See panel-common.txt
+
+Example:
+
+dsi1: dsi@7e700000 {
+	#address-cells = <1>;
+	#size-cells = <0>;
+	<...>
+
+	port {
+		dsi_out_port: endpoint {
+			remote-endpoint = <&panel_dsi_port>;
+		};
+	};
+};
+
+i2c_dsi: i2c {
+	compatible = "i2c-gpio";
+	#address-cells = <1>;
+	#size-cells = <0>;
+	gpios = <&gpio 28 0
+		 &gpio 29 0>;
+
+	lcd@45 {
+		compatible = "raspberrypi,7inch-touchscreen-panel";
+		reg = <0x45>;
+
+		port {
+			panel_dsi_port: endpoint {
+				remote-endpoint = <&dsi_out_port>;
+			};
+		};
+	};
+};

Documentation/devicetree/bindings/display/sunxi/sun4i-drm.txt

@@ -41,14 +41,17 @@ CEC. It is one end of the pipeline.
 Required properties:
   - compatible: value must be one of:
     * allwinner,sun5i-a10s-hdmi
+    * allwinner,sun6i-a31-hdmi
   - reg: base address and size of memory-mapped region
   - interrupts: interrupt associated to this IP
   - clocks: phandles to the clocks feeding the HDMI encoder
     * ahb: the HDMI interface clock
     * mod: the HDMI module clock
+    * ddc: the HDMI ddc clock (A31 only)
     * pll-0: the first video PLL
     * pll-1: the second video PLL
   - clock-names: the clock names mentioned above
+  - resets: phandle to the reset control for the HDMI encoder (A31 only)
   - dmas: phandles to the DMA channels used by the HDMI encoder
     * ddc-tx: The channel for DDC transmission
     * ddc-rx: The channel for DDC reception

drivers/dma-buf/reservation.c

@@ -266,8 +266,7 @@ EXPORT_SYMBOL(reservation_object_add_excl_fence);
 * @dst: the destination reservation object
 * @src: the source reservation object
 *
-* Copy all fences from src to dst. Both src->lock as well as dst-lock must be
-* held.
+* Copy all fences from src to dst. dst-lock must be held.
 */
 int reservation_object_copy_fences(struct reservation_object *dst,
 				   struct reservation_object *src)
@@ -277,33 +276,62 @@ int reservation_object_copy_fences(struct reservation_object *dst,
 	size_t size;
 	unsigned i;
 
-	src_list = reservation_object_get_list(src);
+	rcu_read_lock();
+	src_list = rcu_dereference(src->fence);
 
+retry:
 	if (src_list) {
-		size = offsetof(typeof(*src_list),
-				shared[src_list->shared_count]);
+		unsigned shared_count = src_list->shared_count;
+
+		size = offsetof(typeof(*src_list), shared[shared_count]);
+		rcu_read_unlock();
+
 		dst_list = kmalloc(size, GFP_KERNEL);
 		if (!dst_list)
 			return -ENOMEM;
 
-		dst_list->shared_count = src_list->shared_count;
-		dst_list->shared_max = src_list->shared_count;
-		for (i = 0; i < src_list->shared_count; ++i)
-			dst_list->shared[i] =
-				dma_fence_get(src_list->shared[i]);
+		rcu_read_lock();
+		src_list = rcu_dereference(src->fence);
+		if (!src_list || src_list->shared_count > shared_count) {
+			kfree(dst_list);
+			goto retry;
+		}
+
+		dst_list->shared_count = 0;
+		dst_list->shared_max = shared_count;
+		for (i = 0; i < src_list->shared_count; ++i) {
+			struct dma_fence *fence;
+
+			fence = rcu_dereference(src_list->shared[i]);
+			if (test_bit(DMA_FENCE_FLAG_SIGNALED_BIT,
+				     &fence->flags))
+				continue;
+
+			if (!dma_fence_get_rcu(fence)) {
+				kfree(dst_list);
+				src_list = rcu_dereference(src->fence);
+				goto retry;
+			}
+
+			if (dma_fence_is_signaled(fence)) {
+				dma_fence_put(fence);
+				continue;
+			}
+
+			dst_list->shared[dst_list->shared_count++] = fence;
+		}
 	} else {
 		dst_list = NULL;
 	}
 
+	new = dma_fence_get_rcu_safe(&src->fence_excl);
+	rcu_read_unlock();
+
 	kfree(dst->staged);
 	dst->staged = NULL;
 
 	src_list = reservation_object_get_list(dst);
-
 	old = reservation_object_get_excl(dst);
-	new = reservation_object_get_excl(src);
-
-	dma_fence_get(new);
 
 	preempt_disable();
 	write_seqcount_begin(&dst->seq);

drivers/gpu/drm/amd/amdgpu/amdgpu_connectors.c

@@ -231,7 +231,7 @@ amdgpu_connector_update_scratch_regs(struct drm_connector *connector,
 		if (connector->encoder_ids[i] == 0)
 			break;
 
-		encoder = drm_encoder_find(connector->dev,
+		encoder = drm_encoder_find(connector->dev, NULL,
 					connector->encoder_ids[i]);
 		if (!encoder)
 			continue;
@@ -256,7 +256,7 @@ amdgpu_connector_find_encoder(struct drm_connector *connector,
 	for (i = 0; i < DRM_CONNECTOR_MAX_ENCODER; i++) {
 		if (connector->encoder_ids[i] == 0)
 			break;
-		encoder = drm_encoder_find(connector->dev,
+		encoder = drm_encoder_find(connector->dev, NULL,
 					connector->encoder_ids[i]);
 		if (!encoder)
 			continue;
@@ -372,7 +372,7 @@ amdgpu_connector_best_single_encoder(struct drm_connector *connector)
 
 	/* pick the encoder ids */
 	if (enc_id)
-		return drm_encoder_find(connector->dev, enc_id);
+		return drm_encoder_find(connector->dev, NULL, enc_id);
 	return NULL;
 }
 
@@ -1077,7 +1077,7 @@ amdgpu_connector_dvi_detect(struct drm_connector *connector, bool force)
 			if (connector->encoder_ids[i] == 0)
 				break;
 
-			encoder = drm_encoder_find(connector->dev, connector->encoder_ids[i]);
+			encoder = drm_encoder_find(connector->dev, NULL, connector->encoder_ids[i]);
 			if (!encoder)
 				continue;
 
@@ -1134,7 +1134,7 @@ amdgpu_connector_dvi_encoder(struct drm_connector *connector)
 		if (connector->encoder_ids[i] == 0)
 			break;
 
-		encoder = drm_encoder_find(connector->dev, connector->encoder_ids[i]);
+		encoder = drm_encoder_find(connector->dev, NULL, connector->encoder_ids[i]);
 		if (!encoder)
 			continue;
 
@@ -1153,7 +1153,7 @@ amdgpu_connector_dvi_encoder(struct drm_connector *connector)
 	/* then check use digitial */
 	/* pick the first one */
 	if (enc_id)
-		return drm_encoder_find(connector->dev, enc_id);
+		return drm_encoder_find(connector->dev, NULL, enc_id);
 	return NULL;
 }
 
@@ -1294,7 +1294,7 @@ u16 amdgpu_connector_encoder_get_dp_bridge_encoder_id(struct drm_connector *conn
 		if (connector->encoder_ids[i] == 0)
 			break;
 
-		encoder = drm_encoder_find(connector->dev,
+		encoder = drm_encoder_find(connector->dev, NULL,
 					connector->encoder_ids[i]);
 		if (!encoder)
 			continue;
@@ -1323,7 +1323,7 @@ static bool amdgpu_connector_encoder_is_hbr2(struct drm_connector *connector)
 	for (i = 0; i < DRM_CONNECTOR_MAX_ENCODER; i++) {
 		if (connector->encoder_ids[i] == 0)
 			break;
-		encoder = drm_encoder_find(connector->dev,
+		encoder = drm_encoder_find(connector->dev, NULL,
 					connector->encoder_ids[i]);
 		if (!encoder)
 			continue;

drivers/gpu/drm/amd/amdgpu/dce_virtual.c

@@ -288,7 +288,7 @@ dce_virtual_encoder(struct drm_connector *connector)
 		if (connector->encoder_ids[i] == 0)
 			break;
 
-		encoder = drm_encoder_find(connector->dev, connector->encoder_ids[i]);
+		encoder = drm_encoder_find(connector->dev, NULL, connector->encoder_ids[i]);
 		if (!encoder)
 			continue;
 
@@ -298,7 +298,7 @@ dce_virtual_encoder(struct drm_connector *connector)
 
 	/* pick the first one */
 	if (enc_id)
-		return drm_encoder_find(connector->dev, enc_id);
+		return drm_encoder_find(connector->dev, NULL, enc_id);
 	return NULL;
 }
 

drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c

@@ -2638,7 +2638,7 @@ static struct drm_encoder *best_encoder(struct drm_connector *connector)
 
 	/* pick the encoder ids */
 	if (enc_id) {
-		obj = drm_mode_object_find(connector->dev, enc_id, DRM_MODE_OBJECT_ENCODER);
+		obj = drm_mode_object_find(connector->dev, NULL, enc_id, DRM_MODE_OBJECT_ENCODER);
 		if (!obj) {
 			DRM_ERROR("Couldn't find a matching encoder for our connector\n");
 			return NULL;

drivers/gpu/drm/ast/ast_mode.c

@@ -713,7 +713,7 @@ static struct drm_encoder *ast_best_single_encoder(struct drm_connector *connect
 	int enc_id = connector->encoder_ids[0];
 	/* pick the encoder ids */
 	if (enc_id)
-		return drm_encoder_find(connector->dev, enc_id);
+		return drm_encoder_find(connector->dev, NULL, enc_id);
 	return NULL;
 }
 

drivers/gpu/drm/bochs/bochs_kms.c

@@ -213,7 +213,7 @@ bochs_connector_best_encoder(struct drm_connector *connector)
 	int enc_id = connector->encoder_ids[0];
 	/* pick the encoder ids */
 	if (enc_id)
-		return drm_encoder_find(connector->dev, enc_id);
+		return drm_encoder_find(connector->dev, NULL, enc_id);
 	return NULL;
 }
 

drivers/gpu/drm/bridge/Kconfig

@@ -71,7 +71,7 @@ config DRM_PARADE_PS8622
 
 config DRM_SIL_SII8620
 	tristate "Silicon Image SII8620 HDMI/MHL bridge"
-	depends on OF
+	depends on OF && RC_CORE
 	select DRM_KMS_HELPER
 	help
 	  Silicon Image SII8620 HDMI/MHL bridge chip driver.
@@ -84,6 +84,14 @@ config DRM_SII902X
 	---help---
 	  Silicon Image sii902x bridge chip driver.
 
+config DRM_SII9234
+	tristate "Silicon Image SII9234 HDMI/MHL bridge"
+	depends on OF
+	---help---
+	  Say Y here if you want support for the MHL interface.
+	  It is an I2C driver, that detects connection of MHL bridge
+	  and starts encapsulation of HDMI signal.
+
 config DRM_TOSHIBA_TC358767
 	tristate "Toshiba TC358767 eDP bridge"
 	depends on OF

drivers/gpu/drm/bridge/Makefile

@@ -6,6 +6,7 @@ obj-$(CONFIG_DRM_NXP_PTN3460) += nxp-ptn3460.o
 obj-$(CONFIG_DRM_PARADE_PS8622) += parade-ps8622.o
 obj-$(CONFIG_DRM_SIL_SII8620) += sil-sii8620.o
 obj-$(CONFIG_DRM_SII902X) += sii902x.o
+obj-$(CONFIG_DRM_SII9234) += sii9234.o
 obj-$(CONFIG_DRM_TOSHIBA_TC358767) += tc358767.o
 obj-$(CONFIG_DRM_ANALOGIX_DP) += analogix/
 obj-$(CONFIG_DRM_I2C_ADV7511) += adv7511/

drivers/gpu/drm/bridge/adv7511/Kconfig

@@ -21,3 +21,11 @@ config DRM_I2C_ADV7533
 	default y
 	help
 	  Support for the Analog Devices ADV7533 DSI to HDMI encoder.
+
+config DRM_I2C_ADV7511_CEC
+	bool "ADV7511/33 HDMI CEC driver"
+	depends on DRM_I2C_ADV7511
+	select CEC_CORE
+	default y
+	help
+	  When selected the HDMI transmitter will support the CEC feature.

drivers/gpu/drm/bridge/adv7511/Makefile

@@ -1,4 +1,5 @@
 adv7511-y := adv7511_drv.o
 adv7511-$(CONFIG_DRM_I2C_ADV7511_AUDIO) += adv7511_audio.o
+adv7511-$(CONFIG_DRM_I2C_ADV7511_CEC) += adv7511_cec.o
 adv7511-$(CONFIG_DRM_I2C_ADV7533) += adv7533.o
 obj-$(CONFIG_DRM_I2C_ADV7511) += adv7511.o

drivers/gpu/drm/bridge/adv7511/adv7511.h

@@ -195,6 +195,25 @@
 #define ADV7511_PACKET_GM(x)	    ADV7511_PACKET(5, x)
 #define ADV7511_PACKET_SPARE(x)	    ADV7511_PACKET(6, x)
 
+#define ADV7511_REG_CEC_TX_FRAME_HDR	0x00
+#define ADV7511_REG_CEC_TX_FRAME_DATA0	0x01
+#define ADV7511_REG_CEC_TX_FRAME_LEN	0x10
+#define ADV7511_REG_CEC_TX_ENABLE	0x11
+#define ADV7511_REG_CEC_TX_RETRY	0x12
+#define ADV7511_REG_CEC_TX_LOW_DRV_CNT	0x14
+#define ADV7511_REG_CEC_RX_FRAME_HDR	0x15
+#define ADV7511_REG_CEC_RX_FRAME_DATA0	0x16
+#define ADV7511_REG_CEC_RX_FRAME_LEN	0x25
+#define ADV7511_REG_CEC_RX_ENABLE	0x26
+#define ADV7511_REG_CEC_RX_BUFFERS	0x4a
+#define ADV7511_REG_CEC_LOG_ADDR_MASK	0x4b
+#define ADV7511_REG_CEC_LOG_ADDR_0_1	0x4c
+#define ADV7511_REG_CEC_LOG_ADDR_2	0x4d
+#define ADV7511_REG_CEC_CLK_DIV		0x4e
+#define ADV7511_REG_CEC_SOFT_RESET	0x50
+
+#define ADV7533_REG_CEC_OFFSET		0x70
+
 enum adv7511_input_clock {
 	ADV7511_INPUT_CLOCK_1X,
 	ADV7511_INPUT_CLOCK_2X,
@@ -297,6 +316,8 @@ enum adv7511_type {
 	ADV7533,
 };
 
+#define ADV7511_MAX_ADDRS 3
+
 struct adv7511 {
 	struct i2c_client *i2c_main;
 	struct i2c_client *i2c_edid;
@@ -341,15 +362,27 @@ struct adv7511 {
 
 	enum adv7511_type type;
 	struct platform_device *audio_pdev;
+
+	struct cec_adapter *cec_adap;
+	u8   cec_addr[ADV7511_MAX_ADDRS];
+	u8   cec_valid_addrs;
+	bool cec_enabled_adap;
+	struct clk *cec_clk;
+	u32 cec_clk_freq;
 };
 
+#ifdef CONFIG_DRM_I2C_ADV7511_CEC
+int adv7511_cec_init(struct device *dev, struct adv7511 *adv7511,
+		     unsigned int offset);
+void adv7511_cec_irq_process(struct adv7511 *adv7511, unsigned int irq1);
+#endif
+
 #ifdef CONFIG_DRM_I2C_ADV7533
 void adv7533_dsi_power_on(struct adv7511 *adv);
 void adv7533_dsi_power_off(struct adv7511 *adv);
 void adv7533_mode_set(struct adv7511 *adv, struct drm_display_mode *mode);
 int adv7533_patch_registers(struct adv7511 *adv);
-void adv7533_uninit_cec(struct adv7511 *adv);
-int adv7533_init_cec(struct adv7511 *adv);
+int adv7533_patch_cec_registers(struct adv7511 *adv);
 int adv7533_attach_dsi(struct adv7511 *adv);
 void adv7533_detach_dsi(struct adv7511 *adv);
 int adv7533_parse_dt(struct device_node *np, struct adv7511 *adv);
@@ -372,11 +405,7 @@ static inline int adv7533_patch_registers(struct adv7511 *adv)
 	return -ENODEV;
 }
 
-static inline void adv7533_uninit_cec(struct adv7511 *adv)
-{
-}
-
-static inline int adv7533_init_cec(struct adv7511 *adv)
+static inline int adv7533_patch_cec_registers(struct adv7511 *adv)
 {
 	return -ENODEV;
 }

drivers/gpu/drm/bridge/adv7511/adv7511_cec.c

@@ -0,0 +1,337 @@
+/*
+ * adv7511_cec.c - Analog Devices ADV7511/33 cec driver
+ *
+ * Copyright 2017 Cisco Systems, Inc. and/or its affiliates. All rights reserved.
+ *
+ * This program is free software; you may redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; version 2 of the License.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ *
+ */
+
+#include <linux/device.h>
+#include <linux/module.h>
+#include <linux/of_device.h>
+#include <linux/slab.h>
+#include <linux/clk.h>
+
+#include <media/cec.h>
+
+#include "adv7511.h"
+
+#define ADV7511_INT1_CEC_MASK \
+	(ADV7511_INT1_CEC_TX_READY | ADV7511_INT1_CEC_TX_ARBIT_LOST | \
+	 ADV7511_INT1_CEC_TX_RETRY_TIMEOUT | ADV7511_INT1_CEC_RX_READY1)
+
+static void adv_cec_tx_raw_status(struct adv7511 *adv7511, u8 tx_raw_status)
+{
+	unsigned int offset = adv7511->type == ADV7533 ?
+					ADV7533_REG_CEC_OFFSET : 0;
+	unsigned int val;
+
+	if (regmap_read(adv7511->regmap_cec,
+			ADV7511_REG_CEC_TX_ENABLE + offset, &val))
+		return;
+
+	if ((val & 0x01) == 0)
+		return;
+
+	if (tx_raw_status & ADV7511_INT1_CEC_TX_ARBIT_LOST) {
+		cec_transmit_attempt_done(adv7511->cec_adap,
+					  CEC_TX_STATUS_ARB_LOST);
+		return;
+	}
+	if (tx_raw_status & ADV7511_INT1_CEC_TX_RETRY_TIMEOUT) {
+		u8 status;
+		u8 err_cnt = 0;
+		u8 nack_cnt = 0;
+		u8 low_drive_cnt = 0;
+		unsigned int cnt;
+
+		/*
+		 * We set this status bit since this hardware performs
+		 * retransmissions.
+		 */
+		status = CEC_TX_STATUS_MAX_RETRIES;
+		if (regmap_read(adv7511->regmap_cec,
+			    ADV7511_REG_CEC_TX_LOW_DRV_CNT + offset, &cnt)) {
+			err_cnt = 1;
+			status |= CEC_TX_STATUS_ERROR;
+		} else {
+			nack_cnt = cnt & 0xf;
+			if (nack_cnt)
+				status |= CEC_TX_STATUS_NACK;
+			low_drive_cnt = cnt >> 4;
+			if (low_drive_cnt)
+				status |= CEC_TX_STATUS_LOW_DRIVE;
+		}
+		cec_transmit_done(adv7511->cec_adap, status,
+				  0, nack_cnt, low_drive_cnt, err_cnt);
+		return;
+	}
+	if (tx_raw_status & ADV7511_INT1_CEC_TX_READY) {
+		cec_transmit_attempt_done(adv7511->cec_adap, CEC_TX_STATUS_OK);
+		return;
+	}
+}
+
+void adv7511_cec_irq_process(struct adv7511 *adv7511, unsigned int irq1)
+{
+	unsigned int offset = adv7511->type == ADV7533 ?
+					ADV7533_REG_CEC_OFFSET : 0;
+	const u32 irq_tx_mask = ADV7511_INT1_CEC_TX_READY |
+				ADV7511_INT1_CEC_TX_ARBIT_LOST |
+				ADV7511_INT1_CEC_TX_RETRY_TIMEOUT;
+	struct cec_msg msg = {};
+	unsigned int len;
+	unsigned int val;
+	u8 i;
+
+	if (irq1 & irq_tx_mask)
+		adv_cec_tx_raw_status(adv7511, irq1);
+
+	if (!(irq1 & ADV7511_INT1_CEC_RX_READY1))
+		return;
+
+	if (regmap_read(adv7511->regmap_cec,
+			ADV7511_REG_CEC_RX_FRAME_LEN + offset, &len))
+		return;
+
+	msg.len = len & 0x1f;
+
+	if (msg.len > 16)
+		msg.len = 16;
+
+	if (!msg.len)
+		return;
+
+	for (i = 0; i < msg.len; i++) {
+		regmap_read(adv7511->regmap_cec,
+			    i + ADV7511_REG_CEC_RX_FRAME_HDR + offset, &val);
+		msg.msg[i] = val;
+	}
+
+	/* toggle to re-enable rx 1 */
+	regmap_write(adv7511->regmap_cec,
+		     ADV7511_REG_CEC_RX_BUFFERS + offset, 1);
+	regmap_write(adv7511->regmap_cec,
+		     ADV7511_REG_CEC_RX_BUFFERS + offset, 0);
+	cec_received_msg(adv7511->cec_adap, &msg);
+}
+
+static int adv7511_cec_adap_enable(struct cec_adapter *adap, bool enable)
+{
+	struct adv7511 *adv7511 = cec_get_drvdata(adap);
+	unsigned int offset = adv7511->type == ADV7533 ?
+					ADV7533_REG_CEC_OFFSET : 0;
+
+	if (adv7511->i2c_cec == NULL)
+		return -EIO;
+
+	if (!adv7511->cec_enabled_adap && enable) {
+		/* power up cec section */
+		regmap_update_bits(adv7511->regmap_cec,
+				   ADV7511_REG_CEC_CLK_DIV + offset,
+				   0x03, 0x01);
+		/* legacy mode and clear all rx buffers */
+		regmap_write(adv7511->regmap_cec,
+			     ADV7511_REG_CEC_RX_BUFFERS + offset, 0x07);
+		regmap_write(adv7511->regmap_cec,
+			     ADV7511_REG_CEC_RX_BUFFERS + offset, 0);
+		/* initially disable tx */
+		regmap_update_bits(adv7511->regmap_cec,
+				   ADV7511_REG_CEC_TX_ENABLE + offset, 1, 0);
+		/* enabled irqs: */
+		/* tx: ready */
+		/* tx: arbitration lost */
+		/* tx: retry timeout */
+		/* rx: ready 1 */
+		regmap_update_bits(adv7511->regmap,
+				   ADV7511_REG_INT_ENABLE(1), 0x3f,
+				   ADV7511_INT1_CEC_MASK);
+	} else if (adv7511->cec_enabled_adap && !enable) {
+		regmap_update_bits(adv7511->regmap,
+				   ADV7511_REG_INT_ENABLE(1), 0x3f, 0);
+		/* disable address mask 1-3 */
+		regmap_update_bits(adv7511->regmap_cec,
+				   ADV7511_REG_CEC_LOG_ADDR_MASK + offset,
+				   0x70, 0x00);
+		/* power down cec section */
+		regmap_update_bits(adv7511->regmap_cec,
+				   ADV7511_REG_CEC_CLK_DIV + offset,
+				   0x03, 0x00);
+		adv7511->cec_valid_addrs = 0;
+	}
+	adv7511->cec_enabled_adap = enable;
+	return 0;
+}
+
+static int adv7511_cec_adap_log_addr(struct cec_adapter *adap, u8 addr)
+{
+	struct adv7511 *adv7511 = cec_get_drvdata(adap);
+	unsigned int offset = adv7511->type == ADV7533 ?
+					ADV7533_REG_CEC_OFFSET : 0;
+	unsigned int i, free_idx = ADV7511_MAX_ADDRS;
+
+	if (!adv7511->cec_enabled_adap)
+		return addr == CEC_LOG_ADDR_INVALID ? 0 : -EIO;
+
+	if (addr == CEC_LOG_ADDR_INVALID) {
+		regmap_update_bits(adv7511->regmap_cec,
+				   ADV7511_REG_CEC_LOG_ADDR_MASK + offset,
+				   0x70, 0);
+		adv7511->cec_valid_addrs = 0;
+		return 0;
+	}
+
+	for (i = 0; i < ADV7511_MAX_ADDRS; i++) {
+		bool is_valid = adv7511->cec_valid_addrs & (1 << i);
+
+		if (free_idx == ADV7511_MAX_ADDRS && !is_valid)
+			free_idx = i;
+		if (is_valid && adv7511->cec_addr[i] == addr)
+			return 0;
+	}
+	if (i == ADV7511_MAX_ADDRS) {
+		i = free_idx;
+		if (i == ADV7511_MAX_ADDRS)
+			return -ENXIO;
+	}
+	adv7511->cec_addr[i] = addr;
+	adv7511->cec_valid_addrs |= 1 << i;
+
+	switch (i) {
+	case 0:
+		/* enable address mask 0 */
+		regmap_update_bits(adv7511->regmap_cec,
+				   ADV7511_REG_CEC_LOG_ADDR_MASK + offset,
+				   0x10, 0x10);
+		/* set address for mask 0 */
+		regmap_update_bits(adv7511->regmap_cec,
+				   ADV7511_REG_CEC_LOG_ADDR_0_1 + offset,
+				   0x0f, addr);
+		break;
+	case 1:
+		/* enable address mask 1 */
+		regmap_update_bits(adv7511->regmap_cec,
+				   ADV7511_REG_CEC_LOG_ADDR_MASK + offset,
+				   0x20, 0x20);
+		/* set address for mask 1 */
+		regmap_update_bits(adv7511->regmap_cec,
+				   ADV7511_REG_CEC_LOG_ADDR_0_1 + offset,
+				   0xf0, addr << 4);
+		break;
+	case 2:
+		/* enable address mask 2 */
+		regmap_update_bits(adv7511->regmap_cec,
+				   ADV7511_REG_CEC_LOG_ADDR_MASK + offset,
+				   0x40, 0x40);
+		/* set address for mask 1 */
+		regmap_update_bits(adv7511->regmap_cec,
+				   ADV7511_REG_CEC_LOG_ADDR_2 + offset,
+				   0x0f, addr);
+		break;
+	}
+	return 0;
+}
+
+static int adv7511_cec_adap_transmit(struct cec_adapter *adap, u8 attempts,
+				     u32 signal_free_time, struct cec_msg *msg)
+{
+	struct adv7511 *adv7511 = cec_get_drvdata(adap);
+	unsigned int offset = adv7511->type == ADV7533 ?
+					ADV7533_REG_CEC_OFFSET : 0;
+	u8 len = msg->len;
+	unsigned int i;
+
+	/*
+	 * The number of retries is the number of attempts - 1, but retry
+	 * at least once. It's not clear if a value of 0 is allowed, so
+	 * let's do at least one retry.
+	 */
+	regmap_update_bits(adv7511->regmap_cec,
+			   ADV7511_REG_CEC_TX_RETRY + offset,
+			   0x70, max(1, attempts - 1) << 4);
+
+	/* blocking, clear cec tx irq status */
+	regmap_update_bits(adv7511->regmap, ADV7511_REG_INT(1), 0x38, 0x38);
+
+	/* write data */
+	for (i = 0; i < len; i++)
+		regmap_write(adv7511->regmap_cec,
+			     i + ADV7511_REG_CEC_TX_FRAME_HDR + offset,
+			     msg->msg[i]);
+
+	/* set length (data + header) */
+	regmap_write(adv7511->regmap_cec,
+		     ADV7511_REG_CEC_TX_FRAME_LEN + offset, len);
+	/* start transmit, enable tx */
+	regmap_write(adv7511->regmap_cec,
+		     ADV7511_REG_CEC_TX_ENABLE + offset, 0x01);
+	return 0;
+}
+
+static const struct cec_adap_ops adv7511_cec_adap_ops = {
+	.adap_enable = adv7511_cec_adap_enable,
+	.adap_log_addr = adv7511_cec_adap_log_addr,
+	.adap_transmit = adv7511_cec_adap_transmit,
+};
+
+static int adv7511_cec_parse_dt(struct device *dev, struct adv7511 *adv7511)
+{
+	adv7511->cec_clk = devm_clk_get(dev, "cec");
+	if (IS_ERR(adv7511->cec_clk)) {
+		int ret = PTR_ERR(adv7511->cec_clk);
+
+		adv7511->cec_clk = NULL;
+		return ret;
+	}
+	clk_prepare_enable(adv7511->cec_clk);
+	adv7511->cec_clk_freq = clk_get_rate(adv7511->cec_clk);
+	return 0;
+}
+
+int adv7511_cec_init(struct device *dev, struct adv7511 *adv7511,
+		     unsigned int offset)
+{
+	int ret = adv7511_cec_parse_dt(dev, adv7511);
+
+	if (ret)
+		return ret;
+
+	adv7511->cec_adap = cec_allocate_adapter(&adv7511_cec_adap_ops,
+		adv7511, dev_name(dev), CEC_CAP_DEFAULTS, ADV7511_MAX_ADDRS);
+	if (IS_ERR(adv7511->cec_adap))
+		return PTR_ERR(adv7511->cec_adap);
+
+	regmap_write(adv7511->regmap, ADV7511_REG_CEC_CTRL + offset, 0);
+	/* cec soft reset */
+	regmap_write(adv7511->regmap_cec,
+		     ADV7511_REG_CEC_SOFT_RESET + offset, 0x01);
+	regmap_write(adv7511->regmap_cec,
+		     ADV7511_REG_CEC_SOFT_RESET + offset, 0x00);
+
+	/* legacy mode */
+	regmap_write(adv7511->regmap_cec,
+		     ADV7511_REG_CEC_RX_BUFFERS + offset, 0x00);
+
+	regmap_write(adv7511->regmap_cec,
+		     ADV7511_REG_CEC_CLK_DIV + offset,
+		     ((adv7511->cec_clk_freq / 750000) - 1) << 2);
+
+	ret = cec_register_adapter(adv7511->cec_adap, dev);
+	if (ret) {
+		cec_delete_adapter(adv7511->cec_adap);
+		adv7511->cec_adap = NULL;
+	}
+	return ret;
+}

drivers/gpu/drm/bridge/adv7511/adv7511_drv.c

@@ -11,12 +11,15 @@
 #include <linux/module.h>
 #include <linux/of_device.h>
 #include <linux/slab.h>
+#include <linux/clk.h>
 
 #include <drm/drmP.h>
 #include <drm/drm_atomic.h>
 #include <drm/drm_atomic_helper.h>
 #include <drm/drm_edid.h>
 
+#include <media/cec.h>
+
 #include "adv7511.h"
 
 /* ADI recommended values for proper operation. */
@@ -336,8 +339,10 @@ static void __adv7511_power_on(struct adv7511 *adv7511)
 		 */
 		regmap_write(adv7511->regmap, ADV7511_REG_INT_ENABLE(0),
 			     ADV7511_INT0_EDID_READY | ADV7511_INT0_HPD);
-		regmap_write(adv7511->regmap, ADV7511_REG_INT_ENABLE(1),
-			     ADV7511_INT1_DDC_ERROR);
+		regmap_update_bits(adv7511->regmap,
+				   ADV7511_REG_INT_ENABLE(1),
+				   ADV7511_INT1_DDC_ERROR,
+				   ADV7511_INT1_DDC_ERROR);
 	}
 
 	/*
@@ -373,6 +378,9 @@ static void __adv7511_power_off(struct adv7511 *adv7511)
 	regmap_update_bits(adv7511->regmap, ADV7511_REG_POWER,
 			   ADV7511_POWER_POWER_DOWN,
 			   ADV7511_POWER_POWER_DOWN);
+	regmap_update_bits(adv7511->regmap,
+			   ADV7511_REG_INT_ENABLE(1),
+			   ADV7511_INT1_DDC_ERROR, 0);
 	regcache_mark_dirty(adv7511->regmap);
 }
 
@@ -423,6 +431,8 @@ static void adv7511_hpd_work(struct work_struct *work)
 
 	if (adv7511->connector.status != status) {
 		adv7511->connector.status = status;
+		if (status == connector_status_disconnected)
+			cec_phys_addr_invalidate(adv7511->cec_adap);
 		drm_kms_helper_hotplug_event(adv7511->connector.dev);
 	}
 }
@@ -453,6 +463,10 @@ static int adv7511_irq_process(struct adv7511 *adv7511, bool process_hpd)
 			wake_up_all(&adv7511->wq);
 	}
 
+#ifdef CONFIG_DRM_I2C_ADV7511_CEC
+	adv7511_cec_irq_process(adv7511, irq1);
+#endif
+
 	return 0;
 }
 
@@ -595,6 +609,8 @@ static int adv7511_get_modes(struct adv7511 *adv7511,
 
 	kfree(edid);
 
+	cec_s_phys_addr_from_edid(adv7511->cec_adap, edid);
+
 	return count;
 }
 
@@ -919,6 +935,65 @@ static void adv7511_uninit_regulators(struct adv7511 *adv)
 	regulator_bulk_disable(adv->num_supplies, adv->supplies);
 }
 
+static bool adv7511_cec_register_volatile(struct device *dev, unsigned int reg)
+{
+	struct i2c_client *i2c = to_i2c_client(dev);
+	struct adv7511 *adv7511 = i2c_get_clientdata(i2c);
+
+	if (adv7511->type == ADV7533)
+		reg -= ADV7533_REG_CEC_OFFSET;
+
+	switch (reg) {
+	case ADV7511_REG_CEC_RX_FRAME_HDR:
+	case ADV7511_REG_CEC_RX_FRAME_DATA0...
+		ADV7511_REG_CEC_RX_FRAME_DATA0 + 14:
+	case ADV7511_REG_CEC_RX_FRAME_LEN:
+	case ADV7511_REG_CEC_RX_BUFFERS:
+	case ADV7511_REG_CEC_TX_LOW_DRV_CNT:
+		return true;
+	}
+
+	return false;
+}
+
+static const struct regmap_config adv7511_cec_regmap_config = {
+	.reg_bits = 8,
+	.val_bits = 8,
+
+	.max_register = 0xff,
+	.cache_type = REGCACHE_RBTREE,
+	.volatile_reg = adv7511_cec_register_volatile,
+};
+
+static int adv7511_init_cec_regmap(struct adv7511 *adv)
+{
+	int ret;
+
+	adv->i2c_cec = i2c_new_dummy(adv->i2c_main->adapter,
+				     adv->i2c_main->addr - 1);
+	if (!adv->i2c_cec)
+		return -ENOMEM;
+	i2c_set_clientdata(adv->i2c_cec, adv);
+
+	adv->regmap_cec = devm_regmap_init_i2c(adv->i2c_cec,
+					&adv7511_cec_regmap_config);
+	if (IS_ERR(adv->regmap_cec)) {
+		ret = PTR_ERR(adv->regmap_cec);
+		goto err;
+	}
+
+	if (adv->type == ADV7533) {
+		ret = adv7533_patch_cec_registers(adv);
+		if (ret)
+			goto err;
+	}
+
+	return 0;
+err:
+	i2c_unregister_device(adv->i2c_cec);
+	return ret;
+}
+
 static int adv7511_parse_dt(struct device_node *np,
 			    struct adv7511_link_config *config)
 {
@@ -1009,6 +1084,7 @@ static int adv7511_probe(struct i2c_client *i2c, const struct i2c_device_id *id)
 	struct device *dev = &i2c->dev;
 	unsigned int main_i2c_addr = i2c->addr << 1;
 	unsigned int edid_i2c_addr = main_i2c_addr + 4;
+	unsigned int offset;
 	unsigned int val;
 	int ret;
 
@@ -1092,11 +1168,9 @@ static int adv7511_probe(struct i2c_client *i2c, const struct i2c_device_id *id)
 		goto uninit_regulators;
 	}
 
-	if (adv7511->type == ADV7533) {
-		ret = adv7533_init_cec(adv7511);
-		if (ret)
-			goto err_i2c_unregister_edid;
-	}
+	ret = adv7511_init_cec_regmap(adv7511);
+	if (ret)
+		goto err_i2c_unregister_edid;
 
 	INIT_WORK(&adv7511->hpd_work, adv7511_hpd_work);
 
@@ -1111,10 +1185,6 @@ static int adv7511_probe(struct i2c_client *i2c, const struct i2c_device_id *id)
 			goto err_unregister_cec;
 	}
 
-	/* CEC is unused for now */
-	regmap_write(adv7511->regmap, ADV7511_REG_CEC_CTRL,
-		     ADV7511_CEC_CTRL_POWER_DOWN);
-
 	adv7511_power_off(adv7511);
 
 	i2c_set_clientdata(i2c, adv7511);
@@ -1129,10 +1199,23 @@ static int adv7511_probe(struct i2c_client *i2c, const struct i2c_device_id *id)
 
 	adv7511_audio_init(dev, adv7511);
 
+	offset = adv7511->type == ADV7533 ? ADV7533_REG_CEC_OFFSET : 0;
+
+#ifdef CONFIG_DRM_I2C_ADV7511_CEC
+	ret = adv7511_cec_init(dev, adv7511, offset);
+	if (ret)
+		goto err_unregister_cec;
+#else
+	regmap_write(adv7511->regmap, ADV7511_REG_CEC_CTRL + offset,
+		     ADV7511_CEC_CTRL_POWER_DOWN);
+#endif
+
 	return 0;
 
 err_unregister_cec:
-	adv7533_uninit_cec(adv7511);
+	i2c_unregister_device(adv7511->i2c_cec);
+	if (adv7511->cec_clk)
+		clk_disable_unprepare(adv7511->cec_clk);
 err_i2c_unregister_edid:
 	i2c_unregister_device(adv7511->i2c_edid);
 uninit_regulators:
@@ -1145,10 +1228,11 @@ static int adv7511_remove(struct i2c_client *i2c)
 {
 	struct adv7511 *adv7511 = i2c_get_clientdata(i2c);
 
-	if (adv7511->type == ADV7533) {
+	if (adv7511->type == ADV7533)
 		adv7533_detach_dsi(adv7511);
-		adv7533_uninit_cec(adv7511);
-	}
+	i2c_unregister_device(adv7511->i2c_cec);
+	if (adv7511->cec_clk)
+		clk_disable_unprepare(adv7511->cec_clk);
 
 	adv7511_uninit_regulators(adv7511);
 
@@ -1156,6 +1240,8 @@ static int adv7511_remove(struct i2c_client *i2c)
 
 	adv7511_audio_exit(adv7511);
 
+	cec_unregister_adapter(adv7511->cec_adap);
+
 	i2c_unregister_device(adv7511->i2c_edid);
 
 	return 0;

drivers/gpu/drm/bridge/adv7511/adv7533.c

@@ -32,14 +32,6 @@ static const struct reg_sequence adv7533_cec_fixed_registers[] = {
 	{ 0x05, 0xc8 },
 };
 
-static const struct regmap_config adv7533_cec_regmap_config = {
-	.reg_bits = 8,
-	.val_bits = 8,
-
-	.max_register = 0xff,
-	.cache_type = REGCACHE_RBTREE,
-};
-
 static void adv7511_dsi_config_timing_gen(struct adv7511 *adv)
 {
 	struct mipi_dsi_device *dsi = adv->dsi;
@@ -145,37 +137,11 @@ int adv7533_patch_registers(struct adv7511 *adv)
 				     ARRAY_SIZE(adv7533_fixed_registers));
 }
 
-void adv7533_uninit_cec(struct adv7511 *adv)
-{
-	i2c_unregister_device(adv->i2c_cec);
-}
-
-int adv7533_init_cec(struct adv7511 *adv)
+int adv7533_patch_cec_registers(struct adv7511 *adv)
 {
-	int ret;
-
-	adv->i2c_cec = i2c_new_dummy(adv->i2c_main->adapter,
-				     adv->i2c_main->addr - 1);
-	if (!adv->i2c_cec)
-		return -ENOMEM;
-
-	adv->regmap_cec = devm_regmap_init_i2c(adv->i2c_cec,
-					&adv7533_cec_regmap_config);
-	if (IS_ERR(adv->regmap_cec)) {
-		ret = PTR_ERR(adv->regmap_cec);
-		goto err;
-	}
-
-	ret = regmap_register_patch(adv->regmap_cec,
+	return regmap_register_patch(adv->regmap_cec,
 				    adv7533_cec_fixed_registers,
 				    ARRAY_SIZE(adv7533_cec_fixed_registers));
-	if (ret)
-		goto err;
-
-	return 0;
-err:
-	adv7533_uninit_cec(adv);
-	return ret;
 }
 
 int adv7533_attach_dsi(struct adv7511 *adv)

drivers/gpu/drm/bridge/panel.c

@@ -188,7 +188,15 @@ EXPORT_SYMBOL(drm_panel_bridge_add);
  */
 void drm_panel_bridge_remove(struct drm_bridge *bridge)
 {
-	struct panel_bridge *panel_bridge = drm_bridge_to_panel_bridge(bridge);
+	struct panel_bridge *panel_bridge;
+
+	if (!bridge)
+		return;
+
+	if (bridge->funcs != &panel_bridge_bridge_funcs)
+		return;
+
+	panel_bridge = drm_bridge_to_panel_bridge(bridge);
 
 	drm_bridge_remove(bridge);
 	devm_kfree(panel_bridge->panel->dev, bridge);

drivers/gpu/drm/bridge/sii9234.c

@@ -0,0 +1,994 @@
+/*
+ * Copyright (C) 2017 Samsung Electronics
+ *
+ * Authors:
+ *    Tomasz Stanislawski <t.stanislaws@samsung.com>
+ *    Maciej Purski <m.purski@samsung.com>
+ *
+ * Based on sii9234 driver created by:
+ *    Adam Hampson <ahampson@sta.samsung.com>
+ *    Erik Gilling <konkers@android.com>
+ *    Shankar Bandal <shankar.b@samsung.com>
+ *    Dharam Kumar <dharam.kr@samsung.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program
+ *
+ */
+#include <drm/bridge/mhl.h>
+#include <drm/drm_crtc.h>
+#include <drm/drm_edid.h>
+
+#include <linux/delay.h>
+#include <linux/err.h>
+#include <linux/gpio/consumer.h>
+#include <linux/i2c.h>
+#include <linux/interrupt.h>
+#include <linux/irq.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/mutex.h>
+#include <linux/regulator/consumer.h>
+#include <linux/slab.h>
+
+#define CBUS_DEVCAP_OFFSET		0x80
+
+#define SII9234_MHL_VERSION		0x11
+#define SII9234_SCRATCHPAD_SIZE		0x10
+#define SII9234_INT_STAT_SIZE		0x33
+
+#define BIT_TMDS_CCTRL_TMDS_OE		BIT(4)
+#define MHL_HPD_OUT_OVR_EN		BIT(4)
+#define MHL_HPD_OUT_OVR_VAL		BIT(5)
+#define MHL_INIT_TIMEOUT		0x0C
+
+/* MHL Tx registers and bits */
+#define MHL_TX_SRST			0x05
+#define MHL_TX_SYSSTAT_REG		0x09
+#define MHL_TX_INTR1_REG		0x71
+#define MHL_TX_INTR4_REG		0x74
+#define MHL_TX_INTR1_ENABLE_REG		0x75
+#define MHL_TX_INTR4_ENABLE_REG		0x78
+#define MHL_TX_INT_CTRL_REG		0x79
+#define MHL_TX_TMDS_CCTRL		0x80
+#define MHL_TX_DISC_CTRL1_REG		0x90
+#define MHL_TX_DISC_CTRL2_REG		0x91
+#define MHL_TX_DISC_CTRL3_REG		0x92
+#define MHL_TX_DISC_CTRL4_REG		0x93
+#define MHL_TX_DISC_CTRL5_REG		0x94
+#define MHL_TX_DISC_CTRL6_REG		0x95
+#define MHL_TX_DISC_CTRL7_REG		0x96
+#define MHL_TX_DISC_CTRL8_REG		0x97
+#define MHL_TX_STAT2_REG		0x99
+#define MHL_TX_MHLTX_CTL1_REG		0xA0
+#define MHL_TX_MHLTX_CTL2_REG		0xA1
+#define MHL_TX_MHLTX_CTL4_REG		0xA3
+#define MHL_TX_MHLTX_CTL6_REG		0xA5
+#define MHL_TX_MHLTX_CTL7_REG		0xA6
+
+#define RSEN_STATUS			BIT(2)
+#define HPD_CHANGE_INT			BIT(6)
+#define RSEN_CHANGE_INT			BIT(5)
+#define RGND_READY_INT			BIT(6)
+#define VBUS_LOW_INT			BIT(5)
+#define CBUS_LKOUT_INT			BIT(4)
+#define MHL_DISC_FAIL_INT		BIT(3)
+#define MHL_EST_INT			BIT(2)
+#define HPD_CHANGE_INT_MASK		BIT(6)
+#define RSEN_CHANGE_INT_MASK		BIT(5)
+
+#define RGND_READY_MASK			BIT(6)
+#define CBUS_LKOUT_MASK			BIT(4)
+#define MHL_DISC_FAIL_MASK		BIT(3)
+#define MHL_EST_MASK			BIT(2)
+
+#define SKIP_GND			BIT(6)
+
+#define ATT_THRESH_SHIFT		0x04
+#define ATT_THRESH_MASK			(0x03 << ATT_THRESH_SHIFT)
+#define USB_D_OEN			BIT(3)
+#define DEGLITCH_TIME_MASK		0x07
+#define DEGLITCH_TIME_2MS		0
+#define DEGLITCH_TIME_4MS		1
+#define DEGLITCH_TIME_8MS		2
+#define DEGLITCH_TIME_16MS		3
+#define DEGLITCH_TIME_40MS		4
+#define DEGLITCH_TIME_50MS		5
+#define DEGLITCH_TIME_60MS		6
+#define DEGLITCH_TIME_128MS		7
+
+#define USB_D_OVR			BIT(7)
+#define USB_ID_OVR			BIT(6)
+#define DVRFLT_SEL			BIT(5)
+#define BLOCK_RGND_INT			BIT(4)
+#define SKIP_DEG			BIT(3)
+#define CI2CA_POL			BIT(2)
+#define CI2CA_WKUP			BIT(1)
+#define SINGLE_ATT			BIT(0)
+
+#define USB_D_ODN			BIT(5)
+#define VBUS_CHECK			BIT(2)
+#define RGND_INTP_MASK			0x03
+#define RGND_INTP_OPEN			0
+#define RGND_INTP_2K			1
+#define RGND_INTP_1K			2
+#define RGND_INTP_SHORT			3
+
+/* HDMI registers */
+#define HDMI_RX_TMDS0_CCTRL1_REG	0x10
+#define HDMI_RX_TMDS_CLK_EN_REG		0x11
+#define HDMI_RX_TMDS_CH_EN_REG		0x12
+#define HDMI_RX_PLL_CALREFSEL_REG	0x17
+#define HDMI_RX_PLL_VCOCAL_REG		0x1A
+#define HDMI_RX_EQ_DATA0_REG		0x22
+#define HDMI_RX_EQ_DATA1_REG		0x23
+#define HDMI_RX_EQ_DATA2_REG		0x24
+#define HDMI_RX_EQ_DATA3_REG		0x25
+#define HDMI_RX_EQ_DATA4_REG		0x26
+#define HDMI_RX_TMDS_ZONE_CTRL_REG	0x4C
+#define HDMI_RX_TMDS_MODE_CTRL_REG	0x4D
+
+/* CBUS registers */
+#define CBUS_INT_STATUS_1_REG		0x08
+#define CBUS_INTR1_ENABLE_REG		0x09
+#define CBUS_MSC_REQ_ABORT_REASON_REG	0x0D
+#define CBUS_INT_STATUS_2_REG		0x1E
+#define CBUS_INTR2_ENABLE_REG		0x1F
+#define CBUS_LINK_CONTROL_2_REG		0x31
+#define CBUS_MHL_STATUS_REG_0		0xB0
+#define CBUS_MHL_STATUS_REG_1		0xB1
+
+#define BIT_CBUS_RESET			BIT(3)
+#define SET_HPD_DOWNSTREAM		BIT(6)
+
+/* TPI registers */
+#define TPI_DPD_REG			0x3D
+
+/* Timeouts in msec */
+#define T_SRC_VBUS_CBUS_TO_STABLE	200
+#define T_SRC_CBUS_FLOAT		100
+#define T_SRC_CBUS_DEGLITCH		2
+#define T_SRC_RXSENSE_DEGLITCH		110
+
+#define MHL1_MAX_CLK			75000 /* in kHz */
+
+#define I2C_TPI_ADDR			0x3D
+#define I2C_HDMI_ADDR			0x49
+#define I2C_CBUS_ADDR			0x64
+
+enum sii9234_state {
+	ST_OFF,
+	ST_D3,
+	ST_RGND_INIT,
+	ST_RGND_1K,
+	ST_RSEN_HIGH,
+	ST_MHL_ESTABLISHED,
+	ST_FAILURE_DISCOVERY,
+	ST_FAILURE,
+};
+
+struct sii9234 {
+	struct i2c_client *client[4];
+	struct drm_bridge bridge;
+	struct device *dev;
+	struct gpio_desc *gpio_reset;
+	int i2c_error;
+	struct regulator_bulk_data supplies[4];
+
+	struct mutex lock; /* Protects fields below and device registers */
+	enum sii9234_state state;
+};
+
+enum sii9234_client_id {
+	I2C_MHL,
+	I2C_TPI,
+	I2C_HDMI,
+	I2C_CBUS,
+};
+
+static const char * const sii9234_client_name[] = {
+	[I2C_MHL] = "MHL",
+	[I2C_TPI] = "TPI",
+	[I2C_HDMI] = "HDMI",
+	[I2C_CBUS] = "CBUS",
+};
+
+static int sii9234_writeb(struct sii9234 *ctx, int id, int offset,
+			  int value)
+{
+	int ret;
+	struct i2c_client *client = ctx->client[id];
+
+	if (ctx->i2c_error)
+		return ctx->i2c_error;
+
+	ret = i2c_smbus_write_byte_data(client, offset, value);
+	if (ret < 0)
+		dev_err(ctx->dev, "writeb: %4s[0x%02x] <- 0x%02x\n",
+			sii9234_client_name[id], offset, value);
+	ctx->i2c_error = ret;
+
+	return ret;
+}
+
+static int sii9234_writebm(struct sii9234 *ctx, int id, int offset,
+			   int value, int mask)
+{
+	int ret;
+	struct i2c_client *client = ctx->client[id];
+
+	if (ctx->i2c_error)
+		return ctx->i2c_error;
+
+	ret = i2c_smbus_write_byte(client, offset);
+	if (ret < 0) {
+		dev_err(ctx->dev, "writebm: %4s[0x%02x] <- 0x%02x\n",
+			sii9234_client_name[id], offset, value);
+		ctx->i2c_error = ret;
+		return ret;
+	}
+
+	ret = i2c_smbus_read_byte(client);
+	if (ret < 0) {
+		dev_err(ctx->dev, "writebm: %4s[0x%02x] <- 0x%02x\n",
+			sii9234_client_name[id], offset, value);
+		ctx->i2c_error = ret;
+		return ret;
+	}
+
+	value = (value & mask) | (ret & ~mask);
+
+	ret = i2c_smbus_write_byte_data(client, offset, value);
+	if (ret < 0) {
+		dev_err(ctx->dev, "writebm: %4s[0x%02x] <- 0x%02x\n",
+			sii9234_client_name[id], offset, value);
+		ctx->i2c_error = ret;
+	}
+
+	return ret;
+}
+
+static int sii9234_readb(struct sii9234 *ctx, int id, int offset)
+{
+	int ret;
+	struct i2c_client *client = ctx->client[id];
+
+	if (ctx->i2c_error)
+		return ctx->i2c_error;
+
+	ret = i2c_smbus_write_byte(client, offset);
+	if (ret < 0) {
+		dev_err(ctx->dev, "readb: %4s[0x%02x]\n",
+			sii9234_client_name[id], offset);
+		ctx->i2c_error = ret;
+		return ret;
+	}
+
+	ret = i2c_smbus_read_byte(client);
+	if (ret < 0) {
+		dev_err(ctx->dev, "readb: %4s[0x%02x]\n",
+			sii9234_client_name[id], offset);
+		ctx->i2c_error = ret;
+	}
+
+	return ret;
+}
+
+static int sii9234_clear_error(struct sii9234 *ctx)
+{
+	int ret = ctx->i2c_error;
+
+	ctx->i2c_error = 0;
+
+	return ret;
+}
+
+#define mhl_tx_writeb(sii9234, offset, value) \
+	sii9234_writeb(sii9234, I2C_MHL, offset, value)
+#define mhl_tx_writebm(sii9234, offset, value, mask) \
+	sii9234_writebm(sii9234, I2C_MHL, offset, value, mask)
+#define mhl_tx_readb(sii9234, offset) \
+	sii9234_readb(sii9234, I2C_MHL, offset)
+#define cbus_writeb(sii9234, offset, value) \
+	sii9234_writeb(sii9234, I2C_CBUS, offset, value)
+#define cbus_writebm(sii9234, offset, value, mask) \
+	sii9234_writebm(sii9234, I2C_CBUS, offset, value, mask)
+#define cbus_readb(sii9234, offset) \
+	sii9234_readb(sii9234, I2C_CBUS, offset)
+#define hdmi_writeb(sii9234, offset, value) \
+	sii9234_writeb(sii9234, I2C_HDMI, offset, value)
+#define hdmi_writebm(sii9234, offset, value, mask) \
+	sii9234_writebm(sii9234, I2C_HDMI, offset, value, mask)
+#define hdmi_readb(sii9234, offset) \
+	sii9234_readb(sii9234, I2C_HDMI, offset)
+#define tpi_writeb(sii9234, offset, value) \
+	sii9234_writeb(sii9234, I2C_TPI, offset, value)
+#define tpi_writebm(sii9234, offset, value, mask) \
+	sii9234_writebm(sii9234, I2C_TPI, offset, value, mask)
+#define tpi_readb(sii9234, offset) \
+	sii9234_readb(sii9234, I2C_TPI, offset)
+
+static u8 sii9234_tmds_control(struct sii9234 *ctx, bool enable)
+{
+	mhl_tx_writebm(ctx, MHL_TX_TMDS_CCTRL, enable ? ~0 : 0,
+		       BIT_TMDS_CCTRL_TMDS_OE);
+	mhl_tx_writebm(ctx, MHL_TX_INT_CTRL_REG, enable ? ~0 : 0,
+		       MHL_HPD_OUT_OVR_EN | MHL_HPD_OUT_OVR_VAL);
+	return sii9234_clear_error(ctx);
+}
+
+static int sii9234_cbus_reset(struct sii9234 *ctx)
+{
+	int i;
+
+	mhl_tx_writebm(ctx, MHL_TX_SRST, ~0, BIT_CBUS_RESET);
+	msleep(T_SRC_CBUS_DEGLITCH);
+	mhl_tx_writebm(ctx, MHL_TX_SRST, 0, BIT_CBUS_RESET);
+
+	for (i = 0; i < 4; i++) {
+		/*
+		 * Enable WRITE_STAT interrupt for writes to all
+		 * 4 MSC Status registers.
+		 */
+		cbus_writeb(ctx, 0xE0 + i, 0xF2);
+		/*
+		 * Enable SET_INT interrupt for writes to all
+		 * 4 MSC Interrupt registers.
+		 */
+		cbus_writeb(ctx, 0xF0 + i, 0xF2);
+	}
+
+	return sii9234_clear_error(ctx);
+}
+
+/* Require to chek mhl imformation of samsung in cbus_init_register */
+static int sii9234_cbus_init(struct sii9234 *ctx)
+{
+	cbus_writeb(ctx, 0x07, 0xF2);
+	cbus_writeb(ctx, 0x40, 0x03);
+	cbus_writeb(ctx, 0x42, 0x06);
+	cbus_writeb(ctx, 0x36, 0x0C);
+	cbus_writeb(ctx, 0x3D, 0xFD);
+	cbus_writeb(ctx, 0x1C, 0x01);
+	cbus_writeb(ctx, 0x1D, 0x0F);
+	cbus_writeb(ctx, 0x44, 0x02);
+	/* Setup our devcap */
+	cbus_writeb(ctx, CBUS_DEVCAP_OFFSET + MHL_DCAP_DEV_STATE, 0x00);
+	cbus_writeb(ctx, CBUS_DEVCAP_OFFSET + MHL_DCAP_MHL_VERSION,
+		    SII9234_MHL_VERSION);
+	cbus_writeb(ctx, CBUS_DEVCAP_OFFSET + MHL_DCAP_CAT,
+		    MHL_DCAP_CAT_SOURCE);
+	cbus_writeb(ctx, CBUS_DEVCAP_OFFSET + MHL_DCAP_ADOPTER_ID_H, 0x01);
+	cbus_writeb(ctx, CBUS_DEVCAP_OFFSET + MHL_DCAP_ADOPTER_ID_L, 0x41);
+	cbus_writeb(ctx, CBUS_DEVCAP_OFFSET + MHL_DCAP_VID_LINK_MODE,
+		    MHL_DCAP_VID_LINK_RGB444 | MHL_DCAP_VID_LINK_YCBCR444);
+	cbus_writeb(ctx, CBUS_DEVCAP_OFFSET + MHL_DCAP_VIDEO_TYPE,
+		    MHL_DCAP_VT_GRAPHICS);
+	cbus_writeb(ctx, CBUS_DEVCAP_OFFSET + MHL_DCAP_LOG_DEV_MAP,
+		    MHL_DCAP_LD_GUI);
+	cbus_writeb(ctx, CBUS_DEVCAP_OFFSET + MHL_DCAP_BANDWIDTH, 0x0F);
+	cbus_writeb(ctx, CBUS_DEVCAP_OFFSET + MHL_DCAP_FEATURE_FLAG,
+		    MHL_DCAP_FEATURE_RCP_SUPPORT | MHL_DCAP_FEATURE_RAP_SUPPORT
+			| MHL_DCAP_FEATURE_SP_SUPPORT);
+	cbus_writeb(ctx, CBUS_DEVCAP_OFFSET + MHL_DCAP_DEVICE_ID_H, 0x0);
+	cbus_writeb(ctx, CBUS_DEVCAP_OFFSET + MHL_DCAP_DEVICE_ID_L, 0x0);
+	cbus_writeb(ctx, CBUS_DEVCAP_OFFSET + MHL_DCAP_SCRATCHPAD_SIZE,
+		    SII9234_SCRATCHPAD_SIZE);
+	cbus_writeb(ctx, CBUS_DEVCAP_OFFSET + MHL_DCAP_INT_STAT_SIZE,
+		    SII9234_INT_STAT_SIZE);
+	cbus_writeb(ctx, CBUS_DEVCAP_OFFSET + MHL_DCAP_RESERVED, 0);
+	cbus_writebm(ctx, 0x31, 0x0C, 0x0C);
+	cbus_writeb(ctx, 0x30, 0x01);
+	cbus_writebm(ctx, 0x3C, 0x30, 0x38);
+	cbus_writebm(ctx, 0x22, 0x0D, 0x0F);
+	cbus_writebm(ctx, 0x2E, 0x15, 0x15);
+	cbus_writeb(ctx, CBUS_INTR1_ENABLE_REG, 0);
+	cbus_writeb(ctx, CBUS_INTR2_ENABLE_REG, 0);
+
+	return sii9234_clear_error(ctx);
+}
+
+static void force_usb_id_switch_open(struct sii9234 *ctx)
+{
+	/* Disable CBUS discovery */
+	mhl_tx_writebm(ctx, MHL_TX_DISC_CTRL1_REG, 0, 0x01);
+	/* Force USB ID switch to open */
+	mhl_tx_writebm(ctx, MHL_TX_DISC_CTRL6_REG, ~0, USB_ID_OVR);
+	mhl_tx_writebm(ctx, MHL_TX_DISC_CTRL3_REG, ~0, 0x86);
+	/* Force upstream HPD to 0 when not in MHL mode. */
+	mhl_tx_writebm(ctx, MHL_TX_INT_CTRL_REG, 0, 0x30);
+}
+
+static void release_usb_id_switch_open(struct sii9234 *ctx)
+{
+	msleep(T_SRC_CBUS_FLOAT);
+	/* Clear USB ID switch to open */
+	mhl_tx_writebm(ctx, MHL_TX_DISC_CTRL6_REG, 0, USB_ID_OVR);
+	/* Enable CBUS discovery */
+	mhl_tx_writebm(ctx, MHL_TX_DISC_CTRL1_REG, ~0, 0x01);
+}
+
+static int sii9234_power_init(struct sii9234 *ctx)
+{
+	/* Force the SiI9234 into the D0 state. */
+	tpi_writeb(ctx, TPI_DPD_REG, 0x3F);
+	/* Enable TxPLL Clock */
+	hdmi_writeb(ctx, HDMI_RX_TMDS_CLK_EN_REG, 0x01);
+	/* Enable Tx Clock Path & Equalizer */
+	hdmi_writeb(ctx, HDMI_RX_TMDS_CH_EN_REG, 0x15);
+	/* Power Up TMDS */
+	mhl_tx_writeb(ctx, 0x08, 0x35);
+	return sii9234_clear_error(ctx);
+}
+
+static int sii9234_hdmi_init(struct sii9234 *ctx)
+{
+	hdmi_writeb(ctx, HDMI_RX_TMDS0_CCTRL1_REG, 0xC1);
+	hdmi_writeb(ctx, HDMI_RX_PLL_CALREFSEL_REG, 0x03);
+	hdmi_writeb(ctx, HDMI_RX_PLL_VCOCAL_REG, 0x20);
+	hdmi_writeb(ctx, HDMI_RX_EQ_DATA0_REG, 0x8A);
+	hdmi_writeb(ctx, HDMI_RX_EQ_DATA1_REG, 0x6A);
+	hdmi_writeb(ctx, HDMI_RX_EQ_DATA2_REG, 0xAA);
+	hdmi_writeb(ctx, HDMI_RX_EQ_DATA3_REG, 0xCA);
+	hdmi_writeb(ctx, HDMI_RX_EQ_DATA4_REG, 0xEA);
+	hdmi_writeb(ctx, HDMI_RX_TMDS_ZONE_CTRL_REG, 0xA0);
+	hdmi_writeb(ctx, HDMI_RX_TMDS_MODE_CTRL_REG, 0x00);
+	mhl_tx_writeb(ctx, MHL_TX_TMDS_CCTRL, 0x34);
+	hdmi_writeb(ctx, 0x45, 0x44);
+	hdmi_writeb(ctx, 0x31, 0x0A);
+	hdmi_writeb(ctx, HDMI_RX_TMDS0_CCTRL1_REG, 0xC1);
+
+	return sii9234_clear_error(ctx);
+}
+
+static int sii9234_mhl_tx_ctl_int(struct sii9234 *ctx)
+{
+	mhl_tx_writeb(ctx, MHL_TX_MHLTX_CTL1_REG, 0xD0);
+	mhl_tx_writeb(ctx, MHL_TX_MHLTX_CTL2_REG, 0xFC);
+	mhl_tx_writeb(ctx, MHL_TX_MHLTX_CTL4_REG, 0xEB);
+	mhl_tx_writeb(ctx, MHL_TX_MHLTX_CTL7_REG, 0x0C);
+
+	return sii9234_clear_error(ctx);
+}
+
+static int sii9234_reset(struct sii9234 *ctx)
+{
+	int ret;
+
+	sii9234_clear_error(ctx);
+
+	ret = sii9234_power_init(ctx);
+	if (ret < 0)
+		return ret;
+	ret = sii9234_cbus_reset(ctx);
+	if (ret < 0)
+		return ret;
+	ret = sii9234_hdmi_init(ctx);
+	if (ret < 0)
+		return ret;
+	ret = sii9234_mhl_tx_ctl_int(ctx);
+	if (ret < 0)
+		return ret;
+
+	/* Enable HDCP Compliance safety */
+	mhl_tx_writeb(ctx, 0x2B, 0x01);
+	/* CBUS discovery cycle time for each drive and float = 150us */
+	mhl_tx_writebm(ctx, MHL_TX_DISC_CTRL1_REG, 0x04, 0x06);
+	/* Clear bit 6 (reg_skip_rgnd) */
+	mhl_tx_writeb(ctx, MHL_TX_DISC_CTRL2_REG, (1 << 7) /* Reserved */
+		      | 2 << ATT_THRESH_SHIFT | DEGLITCH_TIME_50MS);
+	/*
+	 * Changed from 66 to 65 for 94[1:0] = 01 = 5k reg_cbusmhl_pup_sel
+	 * 1.8V CBUS VTH & GND threshold
+	 * to meet CTS 3.3.7.2 spec
+	 */
+	mhl_tx_writeb(ctx, MHL_TX_DISC_CTRL5_REG, 0x77);
+	cbus_writebm(ctx, CBUS_LINK_CONTROL_2_REG, ~0, MHL_INIT_TIMEOUT);
+	mhl_tx_writeb(ctx, MHL_TX_MHLTX_CTL6_REG, 0xA0);
+	/* RGND & single discovery attempt (RGND blocking) */
+	mhl_tx_writeb(ctx, MHL_TX_DISC_CTRL6_REG, BLOCK_RGND_INT |
+		      DVRFLT_SEL | SINGLE_ATT);
+	/* Use VBUS path of discovery state machine */
+	mhl_tx_writeb(ctx, MHL_TX_DISC_CTRL8_REG, 0);
+	/* 0x92[3] sets the CBUS / ID switch */
+	mhl_tx_writebm(ctx, MHL_TX_DISC_CTRL6_REG, ~0, USB_ID_OVR);
+	/*
+	 * To allow RGND engine to operate correctly.
+	 * When moving the chip from D2 to D0 (power up, init regs)
+	 * the values should be
+	 * 94[1:0] = 01  reg_cbusmhl_pup_sel[1:0] should be set for 5k
+	 * 93[7:6] = 10  reg_cbusdisc_pup_sel[1:0] should be
+	 * set for 10k (default)
+	 * 93[5:4] = 00  reg_cbusidle_pup_sel[1:0] = open (default)
+	 */
+	mhl_tx_writebm(ctx, MHL_TX_DISC_CTRL3_REG, ~0, 0x86);
+	/*
+	 * Change from CC to 8C to match 5K
+	 * to meet CTS 3.3.72 spec
+	 */
+	mhl_tx_writebm(ctx, MHL_TX_DISC_CTRL4_REG, ~0, 0x8C);
+	/* Configure the interrupt as active high */
+	mhl_tx_writebm(ctx, MHL_TX_INT_CTRL_REG, 0, 0x06);
+
+	msleep(25);
+
+	/* Release usb_id switch */
+	mhl_tx_writebm(ctx, MHL_TX_DISC_CTRL6_REG, 0,  USB_ID_OVR);
+	mhl_tx_writeb(ctx, MHL_TX_DISC_CTRL1_REG, 0x27);
+
+	ret = sii9234_clear_error(ctx);
+	if (ret < 0)
+		return ret;
+	ret = sii9234_cbus_init(ctx);
+	if (ret < 0)
+		return ret;
+
+	/* Enable Auto soft reset on SCDT = 0 */
+	mhl_tx_writeb(ctx, 0x05, 0x04);
+	/* HDMI Transcode mode enable */
+	mhl_tx_writeb(ctx, 0x0D, 0x1C);
+	mhl_tx_writeb(ctx, MHL_TX_INTR4_ENABLE_REG,
+		      RGND_READY_MASK | CBUS_LKOUT_MASK
+			| MHL_DISC_FAIL_MASK | MHL_EST_MASK);
+	mhl_tx_writeb(ctx, MHL_TX_INTR1_ENABLE_REG, 0x60);
+
+	/* This point is very important before measure RGND impedance */
+	force_usb_id_switch_open(ctx);
+	mhl_tx_writebm(ctx, MHL_TX_DISC_CTRL4_REG, 0, 0xF0);
+	mhl_tx_writebm(ctx, MHL_TX_DISC_CTRL5_REG, 0, 0x03);
+	release_usb_id_switch_open(ctx);
+
+	/* Force upstream HPD to 0 when not in MHL mode */
+	mhl_tx_writebm(ctx, MHL_TX_INT_CTRL_REG, 0, 1 << 5);
+	mhl_tx_writebm(ctx, MHL_TX_INT_CTRL_REG, ~0, 1 << 4);
+
+	return sii9234_clear_error(ctx);
+}
+
+static int sii9234_goto_d3(struct sii9234 *ctx)
+{
+	int ret;
+
+	dev_dbg(ctx->dev, "sii9234: detection started d3\n");
+
+	ret = sii9234_reset(ctx);
+	if (ret < 0)
+		goto exit;
+
+	hdmi_writeb(ctx, 0x01, 0x03);
+	tpi_writebm(ctx, TPI_DPD_REG, 0, 1);
+	/* I2C above is expected to fail because power goes down */
+	sii9234_clear_error(ctx);
+
+	ctx->state = ST_D3;
+
+	return 0;
+ exit:
+	dev_err(ctx->dev, "%s failed\n", __func__);
+	return -1;
+}
+
+static int sii9234_hw_on(struct sii9234 *ctx)
+{
+	return regulator_bulk_enable(ARRAY_SIZE(ctx->supplies), ctx->supplies);
+}
+
+static void sii9234_hw_off(struct sii9234 *ctx)
+{
+	gpiod_set_value(ctx->gpio_reset, 1);
+	msleep(20);
+	regulator_bulk_disable(ARRAY_SIZE(ctx->supplies), ctx->supplies);
+}
+
+static void sii9234_hw_reset(struct sii9234 *ctx)
+{
+	gpiod_set_value(ctx->gpio_reset, 1);
+	msleep(20);
+	gpiod_set_value(ctx->gpio_reset, 0);
+}
+
+static void sii9234_cable_in(struct sii9234 *ctx)
+{
+	int ret;
+
+	mutex_lock(&ctx->lock);
+	if (ctx->state != ST_OFF)
+		goto unlock;
+	ret = sii9234_hw_on(ctx);
+	if (ret < 0)
+		goto unlock;
+
+	sii9234_hw_reset(ctx);
+	sii9234_goto_d3(ctx);
+	/* To avoid irq storm, when hw is in meta state */
+	enable_irq(to_i2c_client(ctx->dev)->irq);
+
+unlock:
+	mutex_unlock(&ctx->lock);
+}
+
+static void sii9234_cable_out(struct sii9234 *ctx)
+{
+	mutex_lock(&ctx->lock);
+
+	if (ctx->state == ST_OFF)
+		goto unlock;
+
+	disable_irq(to_i2c_client(ctx->dev)->irq);
+	tpi_writeb(ctx, TPI_DPD_REG, 0);
+	/* Turn on&off hpd festure for only QCT HDMI */
+	sii9234_hw_off(ctx);
+
+	ctx->state = ST_OFF;
+
+unlock:
+	mutex_unlock(&ctx->lock);
+}
+
+static enum sii9234_state sii9234_rgnd_ready_irq(struct sii9234 *ctx)
+{
+	int value;
+
+	if (ctx->state == ST_D3) {
+		int ret;
+
+		dev_dbg(ctx->dev, "RGND_READY_INT\n");
+		sii9234_hw_reset(ctx);
+
+		ret = sii9234_reset(ctx);
+		if (ret < 0) {
+			dev_err(ctx->dev, "sii9234_reset() failed\n");
+			return ST_FAILURE;
+		}
+
+		return ST_RGND_INIT;
+	}
+
+	/* Got interrupt in inappropriate state */
+	if (ctx->state != ST_RGND_INIT)
+		return ST_FAILURE;
+
+	value = mhl_tx_readb(ctx, MHL_TX_STAT2_REG);
+	if (sii9234_clear_error(ctx))
+		return ST_FAILURE;
+
+	if ((value & RGND_INTP_MASK) != RGND_INTP_1K) {
+		dev_warn(ctx->dev, "RGND is not 1k\n");
+		return ST_RGND_INIT;
+	}
+	dev_dbg(ctx->dev, "RGND 1K!!\n");
+	mhl_tx_writebm(ctx, MHL_TX_DISC_CTRL4_REG, ~0, 0x8C);
+	mhl_tx_writeb(ctx, MHL_TX_DISC_CTRL5_REG, 0x77);
+	mhl_tx_writebm(ctx, MHL_TX_DISC_CTRL6_REG, ~0, 0x05);
+	if (sii9234_clear_error(ctx))
+		return ST_FAILURE;
+
+	msleep(T_SRC_VBUS_CBUS_TO_STABLE);
+	return ST_RGND_1K;
+}
+
+static enum sii9234_state sii9234_mhl_established(struct sii9234 *ctx)
+{
+	dev_dbg(ctx->dev, "mhl est interrupt\n");
+
+	/* Discovery override */
+	mhl_tx_writeb(ctx, MHL_TX_MHLTX_CTL1_REG, 0x10);
+	/* Increase DDC translation layer timer (byte mode) */
+	cbus_writeb(ctx, 0x07, 0x32);
+	cbus_writebm(ctx, 0x44, ~0, 1 << 1);
+	/* Keep the discovery enabled. Need RGND interrupt */
+	mhl_tx_writebm(ctx, MHL_TX_DISC_CTRL1_REG, ~0, 1);
+	mhl_tx_writeb(ctx, MHL_TX_INTR1_ENABLE_REG,
+		      RSEN_CHANGE_INT_MASK | HPD_CHANGE_INT_MASK);
+
+	if (sii9234_clear_error(ctx))
+		return ST_FAILURE;
+
+	return ST_MHL_ESTABLISHED;
+}
+
+static enum sii9234_state sii9234_hpd_change(struct sii9234 *ctx)
+{
+	int value;
+
+	value = cbus_readb(ctx, CBUS_MSC_REQ_ABORT_REASON_REG);
+	if (sii9234_clear_error(ctx))
+		return ST_FAILURE;
+
+	if (value & SET_HPD_DOWNSTREAM) {
+		/* Downstream HPD High, Enable TMDS */
+		sii9234_tmds_control(ctx, true);
+	} else {
+		/* Downstream HPD Low, Disable TMDS */
+		sii9234_tmds_control(ctx, false);
+	}
+
+	return ctx->state;
+}
+
+static enum sii9234_state sii9234_rsen_change(struct sii9234 *ctx)
+{
+	int value;
+
+	/* Work_around code to handle wrong interrupt */
+	if (ctx->state != ST_RGND_1K) {
+		dev_err(ctx->dev, "RSEN_HIGH without RGND_1K\n");
+		return ST_FAILURE;
+	}
+	value = mhl_tx_readb(ctx, MHL_TX_SYSSTAT_REG);
+	if (value < 0)
+		return ST_FAILURE;
+
+	if (value & RSEN_STATUS) {
+		dev_dbg(ctx->dev, "MHL cable connected.. RSEN High\n");
+		return ST_RSEN_HIGH;
+	}
+	dev_dbg(ctx->dev, "RSEN lost\n");
+	/*
+	 * Once RSEN loss is confirmed,we need to check
+	 * based on cable status and chip power status,whether
+	 * it is SINK Loss(HDMI cable not connected, TV Off)
+	 * or MHL cable disconnection
+	 * TODO: Define the below mhl_disconnection()
+	 */
+	msleep(T_SRC_RXSENSE_DEGLITCH);
+	value = mhl_tx_readb(ctx, MHL_TX_SYSSTAT_REG);
+	if (value < 0)
+		return ST_FAILURE;
+	dev_dbg(ctx->dev, "sys_stat: %x\n", value);
+
+	if (value & RSEN_STATUS) {
+		dev_dbg(ctx->dev, "RSEN recovery\n");
+		return ST_RSEN_HIGH;
+	}
+	dev_dbg(ctx->dev, "RSEN Really LOW\n");
+	/* To meet CTS 3.3.22.2 spec */
+	sii9234_tmds_control(ctx, false);
+	force_usb_id_switch_open(ctx);
+	release_usb_id_switch_open(ctx);
+
+	return ST_FAILURE;
+}
+
+static irqreturn_t sii9234_irq_thread(int irq, void *data)
+{
+	struct sii9234 *ctx = data;
+	int intr1, intr4;
+	int intr1_en, intr4_en;
+	int cbus_intr1, cbus_intr2;
+
+	dev_dbg(ctx->dev, "%s\n", __func__);
+
+	mutex_lock(&ctx->lock);
+
+	intr1 = mhl_tx_readb(ctx, MHL_TX_INTR1_REG);
+	intr4 = mhl_tx_readb(ctx, MHL_TX_INTR4_REG);
+	intr1_en = mhl_tx_readb(ctx, MHL_TX_INTR1_ENABLE_REG);
+	intr4_en = mhl_tx_readb(ctx, MHL_TX_INTR4_ENABLE_REG);
+	cbus_intr1 = cbus_readb(ctx, CBUS_INT_STATUS_1_REG);
+	cbus_intr2 = cbus_readb(ctx, CBUS_INT_STATUS_2_REG);
+
+	if (sii9234_clear_error(ctx))
+		goto done;
+
+	dev_dbg(ctx->dev, "irq %02x/%02x %02x/%02x %02x/%02x\n",
+		intr1, intr1_en, intr4, intr4_en, cbus_intr1, cbus_intr2);
+
+	if (intr4 & RGND_READY_INT)
+		ctx->state = sii9234_rgnd_ready_irq(ctx);
+	if (intr1 & RSEN_CHANGE_INT)
+		ctx->state = sii9234_rsen_change(ctx);
+	if (intr4 & MHL_EST_INT)
+		ctx->state = sii9234_mhl_established(ctx);
+	if (intr1 & HPD_CHANGE_INT)
+		ctx->state = sii9234_hpd_change(ctx);
+	if (intr4 & CBUS_LKOUT_INT)
+		ctx->state = ST_FAILURE;
+	if (intr4 & MHL_DISC_FAIL_INT)
+		ctx->state = ST_FAILURE_DISCOVERY;
+
+ done:
+	/* Clean interrupt status and pending flags */
+	mhl_tx_writeb(ctx, MHL_TX_INTR1_REG, intr1);
+	mhl_tx_writeb(ctx, MHL_TX_INTR4_REG, intr4);
+	cbus_writeb(ctx, CBUS_MHL_STATUS_REG_0, 0xFF);
+	cbus_writeb(ctx, CBUS_MHL_STATUS_REG_1, 0xFF);
+	cbus_writeb(ctx, CBUS_INT_STATUS_1_REG, cbus_intr1);
+	cbus_writeb(ctx, CBUS_INT_STATUS_2_REG, cbus_intr2);
+
+	sii9234_clear_error(ctx);
+
+	if (ctx->state == ST_FAILURE) {
+		dev_dbg(ctx->dev, "try to reset after failure\n");
+		sii9234_hw_reset(ctx);
+		sii9234_goto_d3(ctx);
+	}
+
+	if (ctx->state == ST_FAILURE_DISCOVERY) {
+		dev_err(ctx->dev, "discovery failed, no power for MHL?\n");
+		tpi_writebm(ctx, TPI_DPD_REG, 0, 1);
+		ctx->state = ST_D3;
+	}
+
+	mutex_unlock(&ctx->lock);
+
+	return IRQ_HANDLED;
+}
+
+static int sii9234_init_resources(struct sii9234 *ctx,
+				  struct i2c_client *client)
+{
+	struct i2c_adapter *adapter = to_i2c_adapter(client->dev.parent);
+	int ret;
+
+	if (!ctx->dev->of_node) {
+		dev_err(ctx->dev, "not DT device\n");
+		return -ENODEV;
+	}
+
+	ctx->gpio_reset = devm_gpiod_get(ctx->dev, "reset", GPIOD_OUT_LOW);
+	if (IS_ERR(ctx->gpio_reset)) {
+		dev_err(ctx->dev, "failed to get reset gpio from DT\n");
+		return PTR_ERR(ctx->gpio_reset);
+	}
+
+	ctx->supplies[0].supply = "avcc12";
+	ctx->supplies[1].supply = "avcc33";
+	ctx->supplies[2].supply = "iovcc18";
+	ctx->supplies[3].supply = "cvcc12";
+	ret = devm_regulator_bulk_get(ctx->dev, 4, ctx->supplies);
+	if (ret) {
+		dev_err(ctx->dev, "regulator_bulk failed\n");
+		return ret;
+	}
+
+	ctx->client[I2C_MHL] = client;
+
+	ctx->client[I2C_TPI] = i2c_new_dummy(adapter, I2C_TPI_ADDR);
+	if (!ctx->client[I2C_TPI]) {
+		dev_err(ctx->dev, "failed to create TPI client\n");
+		return -ENODEV;
+	}
+
+	ctx->client[I2C_HDMI] = i2c_new_dummy(adapter, I2C_HDMI_ADDR);
+	if (!ctx->client[I2C_HDMI]) {
+		dev_err(ctx->dev, "failed to create HDMI RX client\n");
+		goto fail_tpi;
+	}
+
+	ctx->client[I2C_CBUS] = i2c_new_dummy(adapter, I2C_CBUS_ADDR);
+	if (!ctx->client[I2C_CBUS]) {
+		dev_err(ctx->dev, "failed to create CBUS client\n");
+		goto fail_hdmi;
+	}
+
+	return 0;
+
+fail_hdmi:
+	i2c_unregister_device(ctx->client[I2C_HDMI]);
+fail_tpi:
+	i2c_unregister_device(ctx->client[I2C_TPI]);
+
+	return -ENODEV;
+}
+
+static void sii9234_deinit_resources(struct sii9234 *ctx)
+{
+	i2c_unregister_device(ctx->client[I2C_CBUS]);
+	i2c_unregister_device(ctx->client[I2C_HDMI]);
+	i2c_unregister_device(ctx->client[I2C_TPI]);
+}
+
+static inline struct sii9234 *bridge_to_sii9234(struct drm_bridge *bridge)
+{
+	return container_of(bridge, struct sii9234, bridge);
+}
+
+static enum drm_mode_status sii9234_mode_valid(struct drm_bridge *bridge,
+					 const struct drm_display_mode *mode)
+{
+	if (mode->clock > MHL1_MAX_CLK)
+		return MODE_CLOCK_HIGH;
+
+	return MODE_OK;
+}
+
+static const struct drm_bridge_funcs sii9234_bridge_funcs = {
+	.mode_valid = sii9234_mode_valid,
+};
+
+static int sii9234_probe(struct i2c_client *client,
+			 const struct i2c_device_id *id)
+{
+	struct i2c_adapter *adapter = to_i2c_adapter(client->dev.parent);
+	struct sii9234 *ctx;
+	struct device *dev = &client->dev;
+	int ret;
+
+	ctx = devm_kzalloc(dev, sizeof(*ctx), GFP_KERNEL);
+	if (!ctx)
+		return -ENOMEM;
+
+	ctx->dev = dev;
+	mutex_init(&ctx->lock);
+
+	if (!i2c_check_functionality(adapter, I2C_FUNC_SMBUS_BYTE_DATA)) {
+		dev_err(dev, "I2C adapter lacks SMBUS feature\n");
+		return -EIO;
+	}
+
+	if (!client->irq) {
+		dev_err(dev, "no irq provided\n");
+		return -EINVAL;
+	}
+
+	irq_set_status_flags(client->irq, IRQ_NOAUTOEN);
+	ret = devm_request_threaded_irq(dev, client->irq, NULL,
+					sii9234_irq_thread,
+					IRQF_TRIGGER_HIGH | IRQF_ONESHOT,
+					"sii9234", ctx);
+	if (ret < 0) {
+		dev_err(dev, "failed to install IRQ handler\n");
+		return ret;
+	}
+
+	ret = sii9234_init_resources(ctx, client);
+	if (ret < 0)
+		return ret;
+
+	i2c_set_clientdata(client, ctx);
+
+	ctx->bridge.funcs = &sii9234_bridge_funcs;
+	ctx->bridge.of_node = dev->of_node;
+	drm_bridge_add(&ctx->bridge);
+
+	sii9234_cable_in(ctx);
+
+	return 0;
+}
+
+static int sii9234_remove(struct i2c_client *client)
+{
+	struct sii9234 *ctx = i2c_get_clientdata(client);
+
+	sii9234_cable_out(ctx);
+	drm_bridge_remove(&ctx->bridge);
+	sii9234_deinit_resources(ctx);
+
+	return 0;
+}
+
+static const struct of_device_id sii9234_dt_match[] = {
+	{ .compatible = "sil,sii9234" },
+	{ },
+};
+MODULE_DEVICE_TABLE(of, sii9234_dt_match);
+
+static const struct i2c_device_id sii9234_id[] = {
+	{ "SII9234", 0 },
+	{ },
+};
+MODULE_DEVICE_TABLE(i2c, sii9234_id);
+
+static struct i2c_driver sii9234_driver = {
+	.driver = {
+		.name	= "sii9234",
+		.of_match_table = sii9234_dt_match,
+	},
+	.probe = sii9234_probe,
+	.remove = sii9234_remove,
+	.id_table = sii9234_id,
+};
+
+module_i2c_driver(sii9234_driver);
+MODULE_LICENSE("GPL");

drivers/gpu/drm/bridge/sil-sii8620.c

@@ -28,6 +28,8 @@
 #include <linux/regulator/consumer.h>
 #include <linux/slab.h>
 
+#include <media/rc-core.h>
+
 #include "sil-sii8620.h"
 
 #define SII8620_BURST_BUF_LEN 288
@@ -58,6 +60,7 @@ enum sii8620_mt_state {
 struct sii8620 {
 	struct drm_bridge bridge;
 	struct device *dev;
+	struct rc_dev *rc_dev;
 	struct clk *clk_xtal;
 	struct gpio_desc *gpio_reset;
 	struct gpio_desc *gpio_int;
@@ -431,6 +434,16 @@ static void sii8620_mt_rap(struct sii8620 *ctx, u8 code)
 	sii8620_mt_msc_msg(ctx, MHL_MSC_MSG_RAP, code);
 }
 
+static void sii8620_mt_rcpk(struct sii8620 *ctx, u8 code)
+{
+	sii8620_mt_msc_msg(ctx, MHL_MSC_MSG_RCPK, code);
+}
+
+static void sii8620_mt_rcpe(struct sii8620 *ctx, u8 code)
+{
+	sii8620_mt_msc_msg(ctx, MHL_MSC_MSG_RCPE, code);
+}
+
 static void sii8620_mt_read_devcap_send(struct sii8620 *ctx,
 					struct sii8620_mt_msg *msg)
 {
@@ -1753,6 +1766,25 @@ static void sii8620_send_features(struct sii8620 *ctx)
 	sii8620_write_buf(ctx, REG_MDT_XMIT_WRITE_PORT, buf, ARRAY_SIZE(buf));
 }
 
+static bool sii8620_rcp_consume(struct sii8620 *ctx, u8 scancode)
+{
+	bool pressed = !(scancode & MHL_RCP_KEY_RELEASED_MASK);
+
+	scancode &= MHL_RCP_KEY_ID_MASK;
+
+	if (!ctx->rc_dev) {
+		dev_dbg(ctx->dev, "RCP input device not initialized\n");
+		return false;
+	}
+
+	if (pressed)
+		rc_keydown(ctx->rc_dev, RC_PROTO_CEC, scancode, 0);
+	else
+		rc_keyup(ctx->rc_dev);
+
+	return true;
+}
+
 static void sii8620_msc_mr_set_int(struct sii8620 *ctx)
 {
 	u8 ints[MHL_INT_SIZE];
@@ -1804,19 +1836,25 @@ static void sii8620_msc_mt_done(struct sii8620 *ctx)
 
 static void sii8620_msc_mr_msc_msg(struct sii8620 *ctx)
 {
-	struct sii8620_mt_msg *msg = sii8620_msc_msg_first(ctx);
+	struct sii8620_mt_msg *msg;
 	u8 buf[2];
 
-	if (!msg)
-		return;
-
 	sii8620_read_buf(ctx, REG_MSC_MR_MSC_MSG_RCVD_1ST_DATA, buf, 2);
 
 	switch (buf[0]) {
 	case MHL_MSC_MSG_RAPK:
+		msg = sii8620_msc_msg_first(ctx);
+		if (!msg)
+			return;
 		msg->ret = buf[1];
 		ctx->mt_state = MT_STATE_DONE;
 		break;
+	case MHL_MSC_MSG_RCP:
+		if (!sii8620_rcp_consume(ctx, buf[1]))
+			sii8620_mt_rcpe(ctx,
+					MHL_RCPE_STATUS_INEFFECTIVE_KEY_CODE);
+		sii8620_mt_rcpk(ctx, buf[1]);
+		break;
 	default:
 		dev_err(ctx->dev, "%s message type %d,%d not supported",
 			__func__, buf[0], buf[1]);
@@ -2102,11 +2140,57 @@ static void sii8620_cable_in(struct sii8620 *ctx)
 	enable_irq(to_i2c_client(ctx->dev)->irq);
 }
 
+static void sii8620_init_rcp_input_dev(struct sii8620 *ctx)
+{
+	struct rc_dev *rc_dev;
+	int ret;
+
+	rc_dev = rc_allocate_device(RC_DRIVER_SCANCODE);
+	if (!rc_dev) {
+		dev_err(ctx->dev, "Failed to allocate RC device\n");
+		ctx->error = -ENOMEM;
+		return;
+	}
+
+	rc_dev->input_phys = "sii8620/input0";
+	rc_dev->input_id.bustype = BUS_VIRTUAL;
+	rc_dev->map_name = RC_MAP_CEC;
+	rc_dev->allowed_protocols = RC_PROTO_BIT_CEC;
+	rc_dev->driver_name = "sii8620";
+	rc_dev->device_name = "sii8620";
+
+	ret = rc_register_device(rc_dev);
+
+	if (ret) {
+		dev_err(ctx->dev, "Failed to register RC device\n");
+		ctx->error = ret;
+		rc_free_device(ctx->rc_dev);
+		return;
+	}
+	ctx->rc_dev = rc_dev;
+}
+
 static inline struct sii8620 *bridge_to_sii8620(struct drm_bridge *bridge)
 {
 	return container_of(bridge, struct sii8620, bridge);
 }
 
+static int sii8620_attach(struct drm_bridge *bridge)
+{
+	struct sii8620 *ctx = bridge_to_sii8620(bridge);
+
+	sii8620_init_rcp_input_dev(ctx);
+
+	return sii8620_clear_error(ctx);
+}
+
+static void sii8620_detach(struct drm_bridge *bridge)
+{
+	struct sii8620 *ctx = bridge_to_sii8620(bridge);
+
+	rc_unregister_device(ctx->rc_dev);
+}
+
 static bool sii8620_mode_fixup(struct drm_bridge *bridge,
 			       const struct drm_display_mode *mode,
 			       struct drm_display_mode *adjusted_mode)
@@ -2151,6 +2235,8 @@ end:
 }
 
 static const struct drm_bridge_funcs sii8620_bridge_funcs = {
+	.attach = sii8620_attach,
+	.detach = sii8620_detach,
 	.mode_fixup = sii8620_mode_fixup,
 };
 
@@ -2217,8 +2303,8 @@ static int sii8620_remove(struct i2c_client *client)
 	struct sii8620 *ctx = i2c_get_clientdata(client);
 
 	disable_irq(to_i2c_client(ctx->dev)->irq);
-	drm_bridge_remove(&ctx->bridge);
 	sii8620_hw_off(ctx);
+	drm_bridge_remove(&ctx->bridge);
 
 	return 0;
 }

drivers/gpu/drm/bridge/synopsys/dw-mipi-dsi.c

@@ -221,7 +221,6 @@ struct dw_mipi_dsi {
 	struct drm_bridge bridge;
 	struct mipi_dsi_host dsi_host;
 	struct drm_bridge *panel_bridge;
-	bool is_panel_bridge;
 	struct device *dev;
 	void __iomem *base;
 
@@ -297,7 +296,6 @@ static int dw_mipi_dsi_host_attach(struct mipi_dsi_host *host,
 		bridge = drm_panel_bridge_add(panel, DRM_MODE_CONNECTOR_DSI);
 		if (IS_ERR(bridge))
 			return PTR_ERR(bridge);
-		dsi->is_panel_bridge = true;
 	}
 
 	dsi->panel_bridge = bridge;
@@ -312,8 +310,7 @@ static int dw_mipi_dsi_host_detach(struct mipi_dsi_host *host,
 {
 	struct dw_mipi_dsi *dsi = host_to_dsi(host);
 
-	if (dsi->is_panel_bridge)
-		drm_panel_bridge_remove(dsi->panel_bridge);
+	drm_of_panel_bridge_remove(host->dev->of_node, 1, 0);
 
 	drm_bridge_remove(&dsi->bridge);
 

drivers/gpu/drm/cirrus/cirrus_mode.c

@@ -457,7 +457,7 @@ static struct drm_encoder *cirrus_connector_best_encoder(struct drm_connector
 	int enc_id = connector->encoder_ids[0];
 	/* pick the encoder ids */
 	if (enc_id)
-		return drm_encoder_find(connector->dev, enc_id);
+		return drm_encoder_find(connector->dev, NULL, enc_id);
 	return NULL;
 }
 

drivers/gpu/drm/drm_atomic.c

@@ -182,9 +182,6 @@ void drm_atomic_state_default_clear(struct drm_atomic_state *state)
 	for (i = 0; i < state->num_private_objs; i++) {
 		struct drm_private_obj *obj = state->private_objs[i].ptr;
 
-		if (!obj)
-			continue;
-
 		obj->funcs->atomic_destroy_state(obj,
 						 state->private_objs[i].state);
 		state->private_objs[i].ptr = NULL;
@@ -718,7 +715,7 @@ static int drm_atomic_plane_set_property(struct drm_plane *plane,
 	struct drm_mode_config *config = &dev->mode_config;
 
 	if (property == config->prop_fb_id) {
-		struct drm_framebuffer *fb = drm_framebuffer_lookup(dev, val);
+		struct drm_framebuffer *fb = drm_framebuffer_lookup(dev, NULL, val);
 		drm_atomic_set_fb_for_plane(state, fb);
 		if (fb)
 			drm_framebuffer_put(fb);
@@ -734,7 +731,7 @@ static int drm_atomic_plane_set_property(struct drm_plane *plane,
 			return -EINVAL;
 
 	} else if (property == config->prop_crtc_id) {
-		struct drm_crtc *crtc = drm_crtc_find(dev, val);
+		struct drm_crtc *crtc = drm_crtc_find(dev, NULL, val);
 		return drm_atomic_set_crtc_for_plane(state, crtc);
 	} else if (property == config->prop_crtc_x) {
 		state->crtc_x = U642I64(val);
@@ -1149,7 +1146,7 @@ static int drm_atomic_connector_set_property(struct drm_connector *connector,
 	struct drm_mode_config *config = &dev->mode_config;
 
 	if (property == config->prop_crtc_id) {
-		struct drm_crtc *crtc = drm_crtc_find(dev, val);
+		struct drm_crtc *crtc = drm_crtc_find(dev, NULL, val);
 		return drm_atomic_set_crtc_for_connector(state, crtc);
 	} else if (property == config->dpms_property) {
 		/* setting DPMS property requires special handling, which
@@ -2259,7 +2256,7 @@ retry:
 			goto out;
 		}
 
-		obj = drm_mode_object_find(dev, obj_id, DRM_MODE_OBJECT_ANY);
+		obj = drm_mode_object_find(dev, file_priv, obj_id, DRM_MODE_OBJECT_ANY);
 		if (!obj) {
 			ret = -ENOENT;
 			goto out;

drivers/gpu/drm/drm_atomic_helper.c

@@ -1704,7 +1704,7 @@ crtc_or_fake_commit(struct drm_atomic_state *state, struct drm_crtc *crtc)
  * drm_atomic_helper_commit_cleanup_done().
  *
  * This is all implemented by in drm_atomic_helper_commit(), giving drivers a
- * complete and esay-to-use default implementation of the atomic_commit() hook.
+ * complete and easy-to-use default implementation of the atomic_commit() hook.
  *
  * The tracking of asynchronously executed and still pending commits is done
  * using the core structure &drm_crtc_commit.
@@ -1819,7 +1819,7 @@ EXPORT_SYMBOL(drm_atomic_helper_setup_commit);
  * This function waits for all preceeding commits that touch the same CRTC as
  * @old_state to both be committed to the hardware (as signalled by
  * drm_atomic_helper_commit_hw_done) and executed by the hardware (as signalled
- * by calling drm_crtc_vblank_send_event() on the &drm_crtc_state.event).
+ * by calling drm_crtc_send_vblank_event() on the &drm_crtc_state.event).
  *
  * This is part of the atomic helper support for nonblocking commits, see
  * drm_atomic_helper_setup_commit() for an overview.
@@ -3052,6 +3052,7 @@ out:
 		drm_modeset_backoff(&ctx);
 	}
 
+	drm_atomic_state_put(state);
 	drm_modeset_drop_locks(&ctx);
 	drm_modeset_acquire_fini(&ctx);
 
@@ -3206,7 +3207,7 @@ struct drm_encoder *
 drm_atomic_helper_best_encoder(struct drm_connector *connector)
 {
 	WARN_ON(connector->encoder_ids[1]);
-	return drm_encoder_find(connector->dev, connector->encoder_ids[0]);
+	return drm_encoder_find(connector->dev, NULL, connector->encoder_ids[0]);
 }
 EXPORT_SYMBOL(drm_atomic_helper_best_encoder);
 

drivers/gpu/drm/drm_color_mgmt.c

@@ -230,7 +230,7 @@ int drm_mode_gamma_set_ioctl(struct drm_device *dev,
 	if (!drm_core_check_feature(dev, DRIVER_MODESET))
 		return -EINVAL;
 
-	crtc = drm_crtc_find(dev, crtc_lut->crtc_id);
+	crtc = drm_crtc_find(dev, file_priv, crtc_lut->crtc_id);
 	if (!crtc)
 		return -ENOENT;
 
@@ -308,7 +308,7 @@ int drm_mode_gamma_get_ioctl(struct drm_device *dev,
 	if (!drm_core_check_feature(dev, DRIVER_MODESET))
 		return -EINVAL;
 
-	crtc = drm_crtc_find(dev, crtc_lut->crtc_id);
+	crtc = drm_crtc_find(dev, file_priv, crtc_lut->crtc_id);
 	if (!crtc)
 		return -ENOENT;
 

drivers/gpu/drm/drm_connector.c

@@ -1310,7 +1310,7 @@ int drm_mode_getconnector(struct drm_device *dev, void *data,
 
 	memset(&u_mode, 0, sizeof(struct drm_mode_modeinfo));
 
-	connector = drm_connector_lookup(dev, out_resp->connector_id);
+	connector = drm_connector_lookup(dev, file_priv, out_resp->connector_id);
 	if (!connector)
 		return -ENOENT;
 

drivers/gpu/drm/drm_crtc.c

@@ -402,7 +402,7 @@ int drm_mode_getcrtc(struct drm_device *dev,
 	if (!drm_core_check_feature(dev, DRIVER_MODESET))
 		return -EINVAL;
 
-	crtc = drm_crtc_find(dev, crtc_resp->crtc_id);
+	crtc = drm_crtc_find(dev, file_priv, crtc_resp->crtc_id);
 	if (!crtc)
 		return -ENOENT;
 
@@ -569,7 +569,7 @@ int drm_mode_setcrtc(struct drm_device *dev, void *data,
 	if (crtc_req->x & 0xffff0000 || crtc_req->y & 0xffff0000)
 		return -ERANGE;
 
-	crtc = drm_crtc_find(dev, crtc_req->crtc_id);
+	crtc = drm_crtc_find(dev, file_priv, crtc_req->crtc_id);
 	if (!crtc) {
 		DRM_DEBUG_KMS("Unknown CRTC ID %d\n", crtc_req->crtc_id);
 		return -ENOENT;
@@ -595,7 +595,7 @@ retry:
 			/* Make refcounting symmetric with the lookup path. */
 			drm_framebuffer_get(fb);
 		} else {
-			fb = drm_framebuffer_lookup(dev, crtc_req->fb_id);
+			fb = drm_framebuffer_lookup(dev, file_priv, crtc_req->fb_id);
 			if (!fb) {
 				DRM_DEBUG_KMS("Unknown FB ID%d\n",
 						crtc_req->fb_id);
@@ -680,7 +680,7 @@ retry:
 				goto out;
 			}
 
-			connector = drm_connector_lookup(dev, out_id);
+			connector = drm_connector_lookup(dev, file_priv, out_id);
 			if (!connector) {
 				DRM_DEBUG_KMS("Connector id %d unknown\n",
 						out_id);

drivers/gpu/drm/drm_crtc_helper.c

@@ -562,12 +562,12 @@ int drm_crtc_helper_set_config(struct drm_mode_set *set,
 	 * Allocate space for the backup of all (non-pointer) encoder and
 	 * connector data.
 	 */
-	save_encoder_crtcs = kzalloc(dev->mode_config.num_encoder *
+	save_encoder_crtcs = kcalloc(dev->mode_config.num_encoder,
 				sizeof(struct drm_crtc *), GFP_KERNEL);
 	if (!save_encoder_crtcs)
 		return -ENOMEM;
 
-	save_connector_encoders = kzalloc(dev->mode_config.num_connector *
+	save_connector_encoders = kcalloc(dev->mode_config.num_connector,
 				sizeof(struct drm_encoder *), GFP_KERNEL);
 	if (!save_connector_encoders) {
 		kfree(save_encoder_crtcs);

drivers/gpu/drm/drm_crtc_internal.h

@@ -106,6 +106,7 @@ int drm_mode_object_add(struct drm_device *dev, struct drm_mode_object *obj,
 void drm_mode_object_register(struct drm_device *dev,
 			      struct drm_mode_object *obj);
 struct drm_mode_object *__drm_mode_object_find(struct drm_device *dev,
+					       struct drm_file *file_priv,
 					       uint32_t id, uint32_t type);
 void drm_mode_object_unregister(struct drm_device *dev,
 				struct drm_mode_object *object);

drivers/gpu/drm/drm_dp_helper.c

@@ -137,8 +137,10 @@ EXPORT_SYMBOL(drm_dp_link_train_channel_eq_delay);
 u8 drm_dp_link_rate_to_bw_code(int link_rate)
 {
 	switch (link_rate) {
-	case 162000:
 	default:
+		WARN(1, "unknown DP link rate %d, using %x\n", link_rate,
+		     DP_LINK_BW_1_62);
+	case 162000:
 		return DP_LINK_BW_1_62;
 	case 270000:
 		return DP_LINK_BW_2_7;
@@ -151,8 +153,9 @@ EXPORT_SYMBOL(drm_dp_link_rate_to_bw_code);
 int drm_dp_bw_code_to_link_rate(u8 link_bw)
 {
 	switch (link_bw) {
-	case DP_LINK_BW_1_62:
 	default:
+		WARN(1, "unknown DP link BW code %x, using 162000\n", link_bw);
+	case DP_LINK_BW_1_62:
 		return 162000;
 	case DP_LINK_BW_2_7:
 		return 270000;

drivers/gpu/drm/drm_encoder.c

@@ -220,7 +220,7 @@ int drm_mode_getencoder(struct drm_device *dev, void *data,
 	if (!drm_core_check_feature(dev, DRIVER_MODESET))
 		return -EINVAL;
 
-	encoder = drm_encoder_find(dev, enc_resp->encoder_id);
+	encoder = drm_encoder_find(dev, file_priv, enc_resp->encoder_id);
 	if (!encoder)
 		return -ENOENT;
 

drivers/gpu/drm/drm_fb_helper.c

@@ -2266,7 +2266,7 @@ static int drm_pick_crtcs(struct drm_fb_helper *fb_helper,
 	if (modes[n] == NULL)
 		return best_score;
 
-	crtcs = kzalloc(fb_helper->connector_count *
+	crtcs = kcalloc(fb_helper->connector_count,
 			sizeof(struct drm_fb_helper_crtc *), GFP_KERNEL);
 	if (!crtcs)
 		return best_score;

drivers/gpu/drm/drm_framebuffer.c

@@ -381,7 +381,7 @@ int drm_mode_rmfb(struct drm_device *dev,
 	if (!drm_core_check_feature(dev, DRIVER_MODESET))
 		return -EINVAL;
 
-	fb = drm_framebuffer_lookup(dev, *id);
+	fb = drm_framebuffer_lookup(dev, file_priv, *id);
 	if (!fb)
 		return -ENOENT;
 
@@ -450,7 +450,7 @@ int drm_mode_getfb(struct drm_device *dev,
 	if (!drm_core_check_feature(dev, DRIVER_MODESET))
 		return -EINVAL;
 
-	fb = drm_framebuffer_lookup(dev, r->fb_id);
+	fb = drm_framebuffer_lookup(dev, file_priv, r->fb_id);
 	if (!fb)
 		return -ENOENT;
 
@@ -515,7 +515,7 @@ int drm_mode_dirtyfb_ioctl(struct drm_device *dev,
 	if (!drm_core_check_feature(dev, DRIVER_MODESET))
 		return -EINVAL;
 
-	fb = drm_framebuffer_lookup(dev, r->fb_id);
+	fb = drm_framebuffer_lookup(dev, file_priv, r->fb_id);
 	if (!fb)
 		return -ENOENT;
 
@@ -688,12 +688,13 @@ EXPORT_SYMBOL(drm_framebuffer_init);
  * again, using drm_framebuffer_put().
  */
 struct drm_framebuffer *drm_framebuffer_lookup(struct drm_device *dev,
+					       struct drm_file *file_priv,
 					       uint32_t id)
 {
 	struct drm_mode_object *obj;
 	struct drm_framebuffer *fb = NULL;
 
-	obj = __drm_mode_object_find(dev, id, DRM_MODE_OBJECT_FB);
+	obj = __drm_mode_object_find(dev, file_priv, id, DRM_MODE_OBJECT_FB);
 	if (obj)
 		fb = obj_to_fb(obj);
 	return fb;

drivers/gpu/drm/drm_gem_framebuffer_helper.c

@@ -27,19 +27,24 @@
  * DOC: overview
  *
  * This library provides helpers for drivers that don't subclass
- * &drm_framebuffer and and use &drm_gem_object for their backing storage.
+ * &drm_framebuffer and use &drm_gem_object for their backing storage.
  *
  * Drivers without additional needs to validate framebuffers can simply use
- * drm_gem_fb_create() and everything is wired up automatically. But all
- * parts can be used individually.
+ * drm_gem_fb_create() and everything is wired up automatically. Other drivers
+ * can use all parts independently.
  */
 
 /**
- * drm_gem_fb_get_obj() - Get GEM object for framebuffer
- * @fb: The framebuffer
- * @plane: Which plane
+ * drm_gem_fb_get_obj() - Get GEM object backing the framebuffer
+ * @fb: Framebuffer
+ * @plane: Plane index
  *
- * Returns the GEM object for given framebuffer.
+ * No additional reference is taken beyond the one that the &drm_frambuffer
+ * already holds.
+ *
+ * Returns:
+ * Pointer to &drm_gem_object for the given framebuffer and plane index or NULL
+ * if it does not exist.
  */
 struct drm_gem_object *drm_gem_fb_get_obj(struct drm_framebuffer *fb,
 					  unsigned int plane)
@@ -82,7 +87,7 @@ drm_gem_fb_alloc(struct drm_device *dev,
 
 /**
  * drm_gem_fb_destroy - Free GEM backed framebuffer
- * @fb: DRM framebuffer
+ * @fb: Framebuffer
  *
  * Frees a GEM backed framebuffer with its backing buffer(s) and the structure
  * itself. Drivers can use this as their &drm_framebuffer_funcs->destroy
@@ -102,12 +107,13 @@ EXPORT_SYMBOL(drm_gem_fb_destroy);
 
 /**
  * drm_gem_fb_create_handle - Create handle for GEM backed framebuffer
- * @fb: DRM framebuffer
- * @file: drm file
- * @handle: handle created
+ * @fb: Framebuffer
+ * @file: DRM file to register the handle for
+ * @handle: Pointer to return the created handle
  *
+ * This function creates a handle for the GEM object backing the framebuffer.
  * Drivers can use this as their &drm_framebuffer_funcs->create_handle
- * callback.
+ * callback. The GETFB IOCTL calls into this callback.
  *
  * Returns:
  * 0 on success or a negative error code on failure.
@@ -120,18 +126,21 @@ int drm_gem_fb_create_handle(struct drm_framebuffer *fb, struct drm_file *file,
 EXPORT_SYMBOL(drm_gem_fb_create_handle);
 
 /**
- * drm_gem_fb_create_with_funcs() - helper function for the
+ * drm_gem_fb_create_with_funcs() - Helper function for the
  *                                  &drm_mode_config_funcs.fb_create
  *                                  callback
  * @dev: DRM device
- * @file: drm file for the ioctl call
- * @mode_cmd: metadata from the userspace fb creation request
+ * @file: DRM file that holds the GEM handle(s) backing the framebuffer
+ * @mode_cmd: Metadata from the userspace framebuffer creation request
  * @funcs: vtable to be used for the new framebuffer object
  *
  * This can be used to set &drm_framebuffer_funcs for drivers that need the
  * &drm_framebuffer_funcs.dirty callback. Use drm_gem_fb_create() if you don't
  * need to change &drm_framebuffer_funcs.
  * The function does buffer size validation.
+ *
+ * Returns:
+ * Pointer to a &drm_framebuffer on success or an error pointer on failure.
  */
 struct drm_framebuffer *
 drm_gem_fb_create_with_funcs(struct drm_device *dev, struct drm_file *file,
@@ -192,15 +201,26 @@ static const struct drm_framebuffer_funcs drm_gem_fb_funcs = {
 };
 
 /**
- * drm_gem_fb_create() - &drm_mode_config_funcs.fb_create callback function
+ * drm_gem_fb_create() - Helper function for the
+ *                       &drm_mode_config_funcs.fb_create callback
  * @dev: DRM device
- * @file: drm file for the ioctl call
- * @mode_cmd: metadata from the userspace fb creation request
+ * @file: DRM file that holds the GEM handle(s) backing the framebuffer
+ * @mode_cmd: Metadata from the userspace framebuffer creation request
+ *
+ * This function creates a new framebuffer object described by
+ * &drm_mode_fb_cmd2. This description includes handles for the buffer(s)
+ * backing the framebuffer.
  *
  * If your hardware has special alignment or pitch requirements these should be
  * checked before calling this function. The function does buffer size
  * validation. Use drm_gem_fb_create_with_funcs() if you need to set
  * &drm_framebuffer_funcs.dirty.
+ *
+ * Drivers can use this as their &drm_mode_config_funcs.fb_create callback.
+ * The ADDFB2 IOCTL calls into this callback.
+ *
+ * Returns:
+ * Pointer to a &drm_framebuffer on success or an error pointer on failure.
  */
 struct drm_framebuffer *
 drm_gem_fb_create(struct drm_device *dev, struct drm_file *file,
@@ -212,15 +232,15 @@ drm_gem_fb_create(struct drm_device *dev, struct drm_file *file,
 EXPORT_SYMBOL_GPL(drm_gem_fb_create);
 
 /**
- * drm_gem_fb_prepare_fb() - Prepare gem framebuffer
- * @plane: Which plane
- * @state: Plane state attach fence to
+ * drm_gem_fb_prepare_fb() - Prepare a GEM backed framebuffer
+ * @plane: Plane
+ * @state: Plane state the fence will be attached to
  *
- * This can be used as the &drm_plane_helper_funcs.prepare_fb hook.
- *
- * This function checks if the plane FB has an dma-buf attached, extracts
- * the exclusive fence and attaches it to plane state for the atomic helper
- * to wait on.
+ * This function prepares a GEM backed framebuffer for scanout by checking if
+ * the plane framebuffer has a DMA-BUF attached. If it does, it extracts the
+ * exclusive fence and attaches it to the plane state for the atomic helper to
+ * wait on. This function can be used as the &drm_plane_helper_funcs.prepare_fb
+ * callback.
  *
  * There is no need for &drm_plane_helper_funcs.cleanup_fb hook for simple
  * gem based framebuffer drivers which have their buffers always pinned in
@@ -246,17 +266,19 @@ int drm_gem_fb_prepare_fb(struct drm_plane *plane,
 EXPORT_SYMBOL_GPL(drm_gem_fb_prepare_fb);
 
 /**
- * drm_gem_fbdev_fb_create - Create a drm_framebuffer for fbdev emulation
+ * drm_gem_fbdev_fb_create - Create a GEM backed &drm_framebuffer for fbdev
+ *                           emulation
  * @dev: DRM device
  * @sizes: fbdev size description
- * @pitch_align: optional pitch alignment
+ * @pitch_align: Optional pitch alignment
  * @obj: GEM object backing the framebuffer
  * @funcs: vtable to be used for the new framebuffer object
  *
- * This function creates a framebuffer for use with fbdev emulation.
+ * This function creates a framebuffer from a &drm_fb_helper_surface_size
+ * description for use in the &drm_fb_helper_funcs.fb_probe callback.
  *
  * Returns:
- * Pointer to a drm_framebuffer on success or an error pointer on failure.
+ * Pointer to a &drm_framebuffer on success or an error pointer on failure.
  */
 struct drm_framebuffer *
 drm_gem_fbdev_fb_create(struct drm_device *dev,

drivers/gpu/drm/drm_internal.h

@@ -55,7 +55,6 @@ int drm_clients_info(struct seq_file *m, void* data);
 int drm_gem_name_info(struct seq_file *m, void *data);
 
 /* drm_vblank.c */
-extern unsigned int drm_timestamp_monotonic;
 void drm_vblank_disable_and_save(struct drm_device *dev, unsigned int pipe);
 void drm_vblank_cleanup(struct drm_device *dev);
 

drivers/gpu/drm/drm_ioctl.c

@@ -235,7 +235,7 @@ static int drm_getcap(struct drm_device *dev, void *data, struct drm_file *file_
 	/* Only some caps make sense with UMS/render-only drivers. */
 	switch (req->capability) {
 	case DRM_CAP_TIMESTAMP_MONOTONIC:
-		req->value = drm_timestamp_monotonic;
+		req->value = 1;
 		return 0;
 	case DRM_CAP_PRIME:
 		req->value |= dev->driver->prime_fd_to_handle ? DRM_PRIME_CAP_IMPORT : 0;

drivers/gpu/drm/drm_mode_object.c

@@ -105,6 +105,7 @@ void drm_mode_object_unregister(struct drm_device *dev,
 }
 
 struct drm_mode_object *__drm_mode_object_find(struct drm_device *dev,
+					       struct drm_file *file_priv,
 					       uint32_t id, uint32_t type)
 {
 	struct drm_mode_object *obj = NULL;
@@ -127,7 +128,7 @@ struct drm_mode_object *__drm_mode_object_find(struct drm_device *dev,
 
 /**
  * drm_mode_object_find - look up a drm object with static lifetime
- * @dev: drm device
+ * @file_priv: drm file
  * @id: id of the mode object
  * @type: type of the mode object
  *
@@ -136,11 +137,12 @@ struct drm_mode_object *__drm_mode_object_find(struct drm_device *dev,
  * by callind drm_mode_object_put().
  */
 struct drm_mode_object *drm_mode_object_find(struct drm_device *dev,
+		struct drm_file *file_priv,
 		uint32_t id, uint32_t type)
 {
 	struct drm_mode_object *obj = NULL;
 
-	obj = __drm_mode_object_find(dev, id, type);
+	obj = __drm_mode_object_find(dev, file_priv, id, type);
 	return obj;
 }
 EXPORT_SYMBOL(drm_mode_object_find);
@@ -359,7 +361,7 @@ int drm_mode_obj_get_properties_ioctl(struct drm_device *dev, void *data,
 
 	drm_modeset_lock_all(dev);
 
-	obj = drm_mode_object_find(dev, arg->obj_id, arg->obj_type);
+	obj = drm_mode_object_find(dev, file_priv, arg->obj_id, arg->obj_type);
 	if (!obj) {
 		ret = -ENOENT;
 		goto out;
@@ -481,7 +483,7 @@ int drm_mode_obj_set_property_ioctl(struct drm_device *dev, void *data,
 	if (!drm_core_check_feature(dev, DRIVER_MODESET))
 		return -EINVAL;
 
-	arg_obj = drm_mode_object_find(dev, arg->obj_id, arg->obj_type);
+	arg_obj = drm_mode_object_find(dev, file_priv, arg->obj_id, arg->obj_type);
 	if (!arg_obj)
 		return -ENOENT;
 

drivers/gpu/drm/drm_plane.c

@@ -513,7 +513,7 @@ int drm_mode_getplane(struct drm_device *dev, void *data,
 	if (!drm_core_check_feature(dev, DRIVER_MODESET))
 		return -EINVAL;
 
-	plane = drm_plane_find(dev, plane_resp->plane_id);
+	plane = drm_plane_find(dev, file_priv, plane_resp->plane_id);
 	if (!plane)
 		return -ENOENT;
 
@@ -703,7 +703,7 @@ int drm_mode_setplane(struct drm_device *dev, void *data,
 	 * First, find the plane, crtc, and fb objects.  If not available,
 	 * we don't bother to call the driver.
 	 */
-	plane = drm_plane_find(dev, plane_req->plane_id);
+	plane = drm_plane_find(dev, file_priv, plane_req->plane_id);
 	if (!plane) {
 		DRM_DEBUG_KMS("Unknown plane ID %d\n",
 			      plane_req->plane_id);
@@ -711,14 +711,14 @@ int drm_mode_setplane(struct drm_device *dev, void *data,
 	}
 
 	if (plane_req->fb_id) {
-		fb = drm_framebuffer_lookup(dev, plane_req->fb_id);
+		fb = drm_framebuffer_lookup(dev, file_priv, plane_req->fb_id);
 		if (!fb) {
 			DRM_DEBUG_KMS("Unknown framebuffer ID %d\n",
 				      plane_req->fb_id);
 			return -ENOENT;
 		}
 
-		crtc = drm_crtc_find(dev, plane_req->crtc_id);
+		crtc = drm_crtc_find(dev, file_priv, plane_req->crtc_id);
 		if (!crtc) {
 			drm_framebuffer_put(fb);
 			DRM_DEBUG_KMS("Unknown crtc ID %d\n",
@@ -829,7 +829,7 @@ static int drm_mode_cursor_common(struct drm_device *dev,
 	if (!req->flags || (~DRM_MODE_CURSOR_FLAGS & req->flags))
 		return -EINVAL;
 
-	crtc = drm_crtc_find(dev, req->crtc_id);
+	crtc = drm_crtc_find(dev, file_priv, req->crtc_id);
 	if (!crtc) {
 		DRM_DEBUG_KMS("Unknown CRTC ID %d\n", req->crtc_id);
 		return -ENOENT;
@@ -944,7 +944,7 @@ int drm_mode_page_flip_ioctl(struct drm_device *dev,
 	if ((page_flip->flags & DRM_MODE_PAGE_FLIP_ASYNC) && !dev->mode_config.async_page_flip)
 		return -EINVAL;
 
-	crtc = drm_crtc_find(dev, page_flip->crtc_id);
+	crtc = drm_crtc_find(dev, file_priv, page_flip->crtc_id);
 	if (!crtc)
 		return -ENOENT;
 
@@ -1005,7 +1005,7 @@ retry:
 		goto out;
 	}
 
-	fb = drm_framebuffer_lookup(dev, page_flip->fb_id);
+	fb = drm_framebuffer_lookup(dev, file_priv, page_flip->fb_id);
 	if (!fb) {
 		ret = -ENOENT;
 		goto out;

drivers/gpu/drm/drm_plane_helper.c

@@ -354,7 +354,7 @@ int drm_primary_helper_update(struct drm_plane *plane, struct drm_crtc *crtc,
 	/* Find current connectors for CRTC */
 	num_connectors = get_connectors_for_crtc(crtc, NULL, 0);
 	BUG_ON(num_connectors == 0);
-	connector_list = kzalloc(num_connectors * sizeof(*connector_list),
+	connector_list = kcalloc(num_connectors, sizeof(*connector_list),
 				 GFP_KERNEL);
 	if (!connector_list)
 		return -ENOMEM;

drivers/gpu/drm/drm_probe_helper.c

@@ -99,7 +99,7 @@ drm_mode_validate_pipeline(struct drm_display_mode *mode,
 
 	/* Step 2: Validate against encoders and crtcs */
 	for (i = 0; i < DRM_CONNECTOR_MAX_ENCODER; i++) {
-		struct drm_encoder *encoder = drm_encoder_find(dev, ids[i]);
+		struct drm_encoder *encoder = drm_encoder_find(dev, NULL, ids[i]);
 		struct drm_crtc *crtc;
 
 		if (!encoder)

drivers/gpu/drm/drm_property.c

@@ -450,7 +450,7 @@ int drm_mode_getproperty_ioctl(struct drm_device *dev,
 	if (!drm_core_check_feature(dev, DRIVER_MODESET))
 		return -EINVAL;
 
-	property = drm_property_find(dev, out_resp->prop_id);
+	property = drm_property_find(dev, file_priv, out_resp->prop_id);
 	if (!property)
 		return -ENOENT;
 
@@ -634,7 +634,7 @@ struct drm_property_blob *drm_property_lookup_blob(struct drm_device *dev,
 	struct drm_mode_object *obj;
 	struct drm_property_blob *blob = NULL;
 
-	obj = __drm_mode_object_find(dev, id, DRM_MODE_OBJECT_BLOB);
+	obj = __drm_mode_object_find(dev, NULL, id, DRM_MODE_OBJECT_BLOB);
 	if (obj)
 		blob = obj_to_blob(obj);
 	return blob;
@@ -897,7 +897,7 @@ bool drm_property_change_valid_get(struct drm_property *property,
 		if (value == 0)
 			return true;
 
-		*ref = __drm_mode_object_find(property->dev, value,
+		*ref = __drm_mode_object_find(property->dev, NULL, value,
 					      property->values[0]);
 		return *ref != NULL;
 	}

drivers/gpu/drm/drm_syncobj.c

@@ -845,7 +845,8 @@ static int drm_syncobj_array_wait(struct drm_device *dev,
 }
 
 static int drm_syncobj_array_find(struct drm_file *file_private,
-				  void *user_handles, uint32_t count_handles,
+				  void __user *user_handles,
+				  uint32_t count_handles,
 				  struct drm_syncobj ***syncobjs_out)
 {
 	uint32_t i, *handles;

drivers/gpu/drm/drm_vblank.c

@@ -78,28 +78,20 @@
 
 static bool
 drm_get_last_vbltimestamp(struct drm_device *dev, unsigned int pipe,
-			  struct timeval *tvblank, bool in_vblank_irq);
+			  ktime_t *tvblank, bool in_vblank_irq);
 
 static unsigned int drm_timestamp_precision = 20;  /* Default to 20 usecs. */
 
-/*
- * Default to use monotonic timestamps for wait-for-vblank and page-flip
- * complete events.
- */
-unsigned int drm_timestamp_monotonic = 1;
-
 static int drm_vblank_offdelay = 5000;    /* Default to 5000 msecs. */
 
 module_param_named(vblankoffdelay, drm_vblank_offdelay, int, 0600);
 module_param_named(timestamp_precision_usec, drm_timestamp_precision, int, 0600);
-module_param_named(timestamp_monotonic, drm_timestamp_monotonic, int, 0600);
 MODULE_PARM_DESC(vblankoffdelay, "Delay until vblank irq auto-disable [msecs] (0: never disable, <0: disable immediately)");
 MODULE_PARM_DESC(timestamp_precision_usec, "Max. error on timestamps [usecs]");
-MODULE_PARM_DESC(timestamp_monotonic, "Use monotonic timestamps");
 
 static void store_vblank(struct drm_device *dev, unsigned int pipe,
 			 u32 vblank_count_inc,
-			 struct timeval *t_vblank, u32 last)
+			 ktime_t t_vblank, u32 last)
 {
 	struct drm_vblank_crtc *vblank = &dev->vblank[pipe];
 
@@ -108,7 +100,7 @@ static void store_vblank(struct drm_device *dev, unsigned int pipe,
 	vblank->last = last;
 
 	write_seqlock(&vblank->seqlock);
-	vblank->time = *t_vblank;
+	vblank->time = t_vblank;
 	vblank->count += vblank_count_inc;
 	write_sequnlock(&vblank->seqlock);
 }
@@ -151,7 +143,7 @@ static void drm_reset_vblank_timestamp(struct drm_device *dev, unsigned int pipe
 {
 	u32 cur_vblank;
 	bool rc;
-	struct timeval t_vblank;
+	ktime_t t_vblank;
 	int count = DRM_TIMESTAMP_MAXRETRIES;
 
 	spin_lock(&dev->vblank_time_lock);
@@ -171,13 +163,13 @@ static void drm_reset_vblank_timestamp(struct drm_device *dev, unsigned int pipe
 	 * interrupt and assign 0 for now, to mark the vblanktimestamp as invalid.
 	 */
 	if (!rc)
-		t_vblank = (struct timeval) {0, 0};
+		t_vblank = 0;
 
 	/*
 	 * +1 to make sure user will never see the same
 	 * vblank counter value before and after a modeset
 	 */
-	store_vblank(dev, pipe, 1, &t_vblank, cur_vblank);
+	store_vblank(dev, pipe, 1, t_vblank, cur_vblank);
 
 	spin_unlock(&dev->vblank_time_lock);
 }
@@ -200,7 +192,7 @@ static void drm_update_vblank_count(struct drm_device *dev, unsigned int pipe,
 	struct drm_vblank_crtc *vblank = &dev->vblank[pipe];
 	u32 cur_vblank, diff;
 	bool rc;
-	struct timeval t_vblank;
+	ktime_t t_vblank;
 	int count = DRM_TIMESTAMP_MAXRETRIES;
 	int framedur_ns = vblank->framedur_ns;
 
@@ -225,11 +217,7 @@ static void drm_update_vblank_count(struct drm_device *dev, unsigned int pipe,
 		/* trust the hw counter when it's around */
 		diff = (cur_vblank - vblank->last) & dev->max_vblank_count;
 	} else if (rc && framedur_ns) {
-		const struct timeval *t_old;
-		u64 diff_ns;
-
-		t_old = &vblank->time;
-		diff_ns = timeval_to_ns(&t_vblank) - timeval_to_ns(t_old);
+		u64 diff_ns = ktime_to_ns(ktime_sub(t_vblank, vblank->time));
 
 		/*
 		 * Figure out how many vblanks we've missed based
@@ -278,9 +266,9 @@ static void drm_update_vblank_count(struct drm_device *dev, unsigned int pipe,
 	 * for now, to mark the vblanktimestamp as invalid.
 	 */
 	if (!rc && !in_vblank_irq)
-		t_vblank = (struct timeval) {0, 0};
+		t_vblank = 0;
 
-	store_vblank(dev, pipe, diff, &t_vblank, cur_vblank);
+	store_vblank(dev, pipe, diff, t_vblank, cur_vblank);
 }
 
 static u32 drm_vblank_count(struct drm_device *dev, unsigned int pipe)
@@ -556,7 +544,7 @@ EXPORT_SYMBOL(drm_calc_timestamping_constants);
  * @pipe: index of CRTC whose vblank timestamp to retrieve
  * @max_error: Desired maximum allowable error in timestamps (nanosecs)
  *             On return contains true maximum error of timestamp
- * @vblank_time: Pointer to struct timeval which should receive the timestamp
+ * @vblank_time: Pointer to time which should receive the timestamp
  * @in_vblank_irq:
  *     True when called from drm_crtc_handle_vblank().  Some drivers
  *     need to apply some workarounds for gpu-specific vblank irq quirks
@@ -584,10 +572,10 @@ EXPORT_SYMBOL(drm_calc_timestamping_constants);
 bool drm_calc_vbltimestamp_from_scanoutpos(struct drm_device *dev,
 					   unsigned int pipe,
 					   int *max_error,
-					   struct timeval *vblank_time,
+					   ktime_t *vblank_time,
 					   bool in_vblank_irq)
 {
-	struct timeval tv_etime;
+	struct timespec64 ts_etime, ts_vblank_time;
 	ktime_t stime, etime;
 	bool vbl_status;
 	struct drm_crtc *crtc;
@@ -676,41 +664,31 @@ bool drm_calc_vbltimestamp_from_scanoutpos(struct drm_device *dev,
 	delta_ns = div_s64(1000000LL * (vpos * mode->crtc_htotal + hpos),
 			   mode->crtc_clock);
 
-	if (!drm_timestamp_monotonic)
-		etime = ktime_mono_to_real(etime);
-
 	/* save this only for debugging purposes */
-	tv_etime = ktime_to_timeval(etime);
+	ts_etime = ktime_to_timespec64(etime);
+	ts_vblank_time = ktime_to_timespec64(*vblank_time);
 	/* Subtract time delta from raw timestamp to get final
 	 * vblank_time timestamp for end of vblank.
 	 */
 	etime = ktime_sub_ns(etime, delta_ns);
-	*vblank_time = ktime_to_timeval(etime);
+	*vblank_time = etime;
 
-	DRM_DEBUG_VBL("crtc %u : v p(%d,%d)@ %ld.%ld -> %ld.%ld [e %d us, %d rep]\n",
+	DRM_DEBUG_VBL("crtc %u : v p(%d,%d)@ %lld.%06ld -> %lld.%06ld [e %d us, %d rep]\n",
 		      pipe, hpos, vpos,
-		      (long)tv_etime.tv_sec, (long)tv_etime.tv_usec,
-		      (long)vblank_time->tv_sec, (long)vblank_time->tv_usec,
-		      duration_ns/1000, i);
+		      (u64)ts_etime.tv_sec, ts_etime.tv_nsec / 1000,
+		      (u64)ts_vblank_time.tv_sec, ts_vblank_time.tv_nsec / 1000,
+		      duration_ns / 1000, i);
 
 	return true;
 }
 EXPORT_SYMBOL(drm_calc_vbltimestamp_from_scanoutpos);
 
-static struct timeval get_drm_timestamp(void)
-{
-	ktime_t now;
-
-	now = drm_timestamp_monotonic ? ktime_get() : ktime_get_real();
-	return ktime_to_timeval(now);
-}
-
 /**
  * drm_get_last_vbltimestamp - retrieve raw timestamp for the most recent
  *                             vblank interval
  * @dev: DRM device
  * @pipe: index of CRTC whose vblank timestamp to retrieve
- * @tvblank: Pointer to target struct timeval which should receive the timestamp
+ * @tvblank: Pointer to target time which should receive the timestamp
  * @in_vblank_irq:
  *     True when called from drm_crtc_handle_vblank().  Some drivers
  *     need to apply some workarounds for gpu-specific vblank irq quirks
@@ -728,7 +706,7 @@ static struct timeval get_drm_timestamp(void)
  */
 static bool
 drm_get_last_vbltimestamp(struct drm_device *dev, unsigned int pipe,
-			  struct timeval *tvblank, bool in_vblank_irq)
+			  ktime_t *tvblank, bool in_vblank_irq)
 {
 	bool ret = false;
 
@@ -744,7 +722,7 @@ drm_get_last_vbltimestamp(struct drm_device *dev, unsigned int pipe,
 	 * Return current monotonic/gettimeofday timestamp as best estimate.
 	 */
 	if (!ret)
-		*tvblank = get_drm_timestamp();
+		*tvblank = ktime_get();
 
 	return ret;
 }
@@ -769,14 +747,14 @@ u32 drm_crtc_vblank_count(struct drm_crtc *crtc)
 EXPORT_SYMBOL(drm_crtc_vblank_count);
 
 static u32 drm_vblank_count_and_time(struct drm_device *dev, unsigned int pipe,
-				     struct timeval *vblanktime)
+				     ktime_t *vblanktime)
 {
 	struct drm_vblank_crtc *vblank = &dev->vblank[pipe];
 	u32 vblank_count;
 	unsigned int seq;
 
 	if (WARN_ON(pipe >= dev->num_crtcs)) {
-		*vblanktime = (struct timeval) { 0 };
+		*vblanktime = 0;
 		return 0;
 	}
 
@@ -793,7 +771,7 @@ static u32 drm_vblank_count_and_time(struct drm_device *dev, unsigned int pipe,
  * drm_crtc_vblank_count_and_time - retrieve "cooked" vblank counter value
  *     and the system timestamp corresponding to that vblank counter value
  * @crtc: which counter to retrieve
- * @vblanktime: Pointer to struct timeval to receive the vblank timestamp.
+ * @vblanktime: Pointer to time to receive the vblank timestamp.
  *
  * Fetches the "cooked" vblank count value that represents the number of
  * vblank events since the system was booted, including lost events due to
@@ -801,7 +779,7 @@ static u32 drm_vblank_count_and_time(struct drm_device *dev, unsigned int pipe,
  * of the vblank interval that corresponds to the current vblank counter value.
  */
 u32 drm_crtc_vblank_count_and_time(struct drm_crtc *crtc,
-				   struct timeval *vblanktime)
+				   ktime_t *vblanktime)
 {
 	return drm_vblank_count_and_time(crtc->dev, drm_crtc_index(crtc),
 					 vblanktime);
@@ -810,11 +788,18 @@ EXPORT_SYMBOL(drm_crtc_vblank_count_and_time);
 
 static void send_vblank_event(struct drm_device *dev,
 		struct drm_pending_vblank_event *e,
-		unsigned long seq, struct timeval *now)
+		unsigned long seq, ktime_t now)
 {
+	struct timespec64 tv = ktime_to_timespec64(now);
+
 	e->event.sequence = seq;
-	e->event.tv_sec = now->tv_sec;
-	e->event.tv_usec = now->tv_usec;
+	/*
+	 * e->event is a user space structure, with hardcoded unsigned
+	 * 32-bit seconds/microseconds. This is safe as we always use
+	 * monotonic timestamps since linux-4.15
+	 */
+	e->event.tv_sec = tv.tv_sec;
+	e->event.tv_usec = tv.tv_nsec / 1000;
 
 	trace_drm_vblank_event_delivered(e->base.file_priv, e->pipe,
 					 e->event.sequence);
@@ -869,7 +854,7 @@ void drm_crtc_arm_vblank_event(struct drm_crtc *crtc,
 	assert_spin_locked(&dev->event_lock);
 
 	e->pipe = pipe;
-	e->event.sequence = drm_vblank_count(dev, pipe);
+	e->event.sequence = drm_crtc_accurate_vblank_count(crtc) + 1;
 	e->event.crtc_id = crtc->base.id;
 	list_add_tail(&e->base.link, &dev->vblank_event_list);
 }
@@ -891,18 +876,18 @@ void drm_crtc_send_vblank_event(struct drm_crtc *crtc,
 {
 	struct drm_device *dev = crtc->dev;
 	unsigned int seq, pipe = drm_crtc_index(crtc);
-	struct timeval now;
+	ktime_t now;
 
 	if (dev->num_crtcs > 0) {
 		seq = drm_vblank_count_and_time(dev, pipe, &now);
 	} else {
 		seq = 0;
 
-		now = get_drm_timestamp();
+		now = ktime_get();
 	}
 	e->pipe = pipe;
 	e->event.crtc_id = crtc->base.id;
-	send_vblank_event(dev, e, seq, &now);
+	send_vblank_event(dev, e, seq, now);
 }
 EXPORT_SYMBOL(drm_crtc_send_vblank_event);
 
@@ -1100,7 +1085,8 @@ void drm_crtc_vblank_off(struct drm_crtc *crtc)
 	unsigned int pipe = drm_crtc_index(crtc);
 	struct drm_vblank_crtc *vblank = &dev->vblank[pipe];
 	struct drm_pending_vblank_event *e, *t;
-	struct timeval now;
+
+	ktime_t now;
 	unsigned long irqflags;
 	unsigned int seq;
 
@@ -1141,7 +1127,7 @@ void drm_crtc_vblank_off(struct drm_crtc *crtc)
 			  e->event.sequence, seq);
 		list_del(&e->base.link);
 		drm_vblank_put(dev, pipe);
-		send_vblank_event(dev, e, seq, &now);
+		send_vblank_event(dev, e, seq, now);
 	}
 	spin_unlock_irqrestore(&dev->event_lock, irqflags);
 
@@ -1321,7 +1307,7 @@ static int drm_queue_vblank_event(struct drm_device *dev, unsigned int pipe,
 {
 	struct drm_vblank_crtc *vblank = &dev->vblank[pipe];
 	struct drm_pending_vblank_event *e;
-	struct timeval now;
+	ktime_t now;
 	unsigned long flags;
 	unsigned int seq;
 	int ret;
@@ -1367,7 +1353,7 @@ static int drm_queue_vblank_event(struct drm_device *dev, unsigned int pipe,
 	e->event.sequence = vblwait->request.sequence;
 	if (vblank_passed(seq, vblwait->request.sequence)) {
 		drm_vblank_put(dev, pipe);
-		send_vblank_event(dev, e, seq, &now);
+		send_vblank_event(dev, e, seq, now);
 		vblwait->reply.sequence = seq;
 	} else {
 		/* drm_handle_vblank_events will call drm_vblank_put */
@@ -1398,6 +1384,23 @@ static bool drm_wait_vblank_is_query(union drm_wait_vblank *vblwait)
 					  _DRM_VBLANK_NEXTONMISS));
 }
 
+static void drm_wait_vblank_reply(struct drm_device *dev, unsigned int pipe,
+				  struct drm_wait_vblank_reply *reply)
+{
+	ktime_t now;
+	struct timespec64 ts;
+
+	/*
+	 * drm_wait_vblank_reply is a UAPI structure that uses 'long'
+	 * to store the seconds. This is safe as we always use monotonic
+	 * timestamps since linux-4.15.
+	 */
+	reply->sequence = drm_vblank_count_and_time(dev, pipe, &now);
+	ts = ktime_to_timespec64(now);
+	reply->tval_sec = (u32)ts.tv_sec;
+	reply->tval_usec = ts.tv_nsec / 1000;
+}
+
 int drm_wait_vblank_ioctl(struct drm_device *dev, void *data,
 			  struct drm_file *file_priv)
 {
@@ -1439,12 +1442,7 @@ int drm_wait_vblank_ioctl(struct drm_device *dev, void *data,
 	if (dev->vblank_disable_immediate &&
 	    drm_wait_vblank_is_query(vblwait) &&
 	    READ_ONCE(vblank->enabled)) {
-		struct timeval now;
-
-		vblwait->reply.sequence =
-			drm_vblank_count_and_time(dev, pipe, &now);
-		vblwait->reply.tval_sec = now.tv_sec;
-		vblwait->reply.tval_usec = now.tv_usec;
+		drm_wait_vblank_reply(dev, pipe, &vblwait->reply);
 		return 0;
 	}
 
@@ -1487,11 +1485,7 @@ int drm_wait_vblank_ioctl(struct drm_device *dev, void *data,
 	}
 
 	if (ret != -EINTR) {
-		struct timeval now;
-
-		vblwait->reply.sequence = drm_vblank_count_and_time(dev, pipe, &now);
-		vblwait->reply.tval_sec = now.tv_sec;
-		vblwait->reply.tval_usec = now.tv_usec;
+		drm_wait_vblank_reply(dev, pipe, &vblwait->reply);
 
 		DRM_DEBUG("crtc %d returning %u to client\n",
 			  pipe, vblwait->reply.sequence);
@@ -1507,7 +1501,7 @@ done:
 static void drm_handle_vblank_events(struct drm_device *dev, unsigned int pipe)
 {
 	struct drm_pending_vblank_event *e, *t;
-	struct timeval now;
+	ktime_t now;
 	unsigned int seq;
 
 	assert_spin_locked(&dev->event_lock);
@@ -1525,7 +1519,7 @@ static void drm_handle_vblank_events(struct drm_device *dev, unsigned int pipe)
 
 		list_del(&e->base.link);
 		drm_vblank_put(dev, pipe);
-		send_vblank_event(dev, e, seq, &now);
+		send_vblank_event(dev, e, seq, now);
 	}
 
 	trace_drm_vblank_event(pipe, seq);

drivers/gpu/drm/etnaviv/Kconfig

@@ -7,8 +7,6 @@ config DRM_ETNAVIV
 	select SHMEM
 	select SYNC_FILE
 	select TMPFS
-	select IOMMU_API
-	select IOMMU_SUPPORT
 	select WANT_DEV_COREDUMP
 	select CMA if HAVE_DMA_CONTIGUOUS
 	select DMA_CMA if HAVE_DMA_CONTIGUOUS

drivers/gpu/drm/etnaviv/Makefile

@@ -10,6 +10,7 @@ etnaviv-y := \
 	etnaviv_gpu.o \
 	etnaviv_iommu_v2.o \
 	etnaviv_iommu.o \
-	etnaviv_mmu.o
+	etnaviv_mmu.o \
+	etnaviv_perfmon.o
 
 obj-$(CONFIG_DRM_ETNAVIV)	+= etnaviv.o

drivers/gpu/drm/etnaviv/etnaviv_buffer.c

@@ -250,6 +250,42 @@ void etnaviv_buffer_end(struct etnaviv_gpu *gpu)
 	}
 }
 
+/* Append a 'sync point' to the ring buffer. */
+void etnaviv_sync_point_queue(struct etnaviv_gpu *gpu, unsigned int event)
+{
+	struct etnaviv_cmdbuf *buffer = gpu->buffer;
+	unsigned int waitlink_offset = buffer->user_size - 16;
+	u32 dwords, target;
+
+	/*
+	 * We need at most 3 dwords in the return target:
+	 * 1 event + 1 end + 1 wait + 1 link.
+	 */
+	dwords = 4;
+	target = etnaviv_buffer_reserve(gpu, buffer, dwords);
+
+	/* Signal sync point event */
+	CMD_LOAD_STATE(buffer, VIVS_GL_EVENT, VIVS_GL_EVENT_EVENT_ID(event) |
+		       VIVS_GL_EVENT_FROM_PE);
+
+	/* Stop the FE to 'pause' the GPU */
+	CMD_END(buffer);
+
+	/* Append waitlink */
+	CMD_WAIT(buffer);
+	CMD_LINK(buffer, 2, etnaviv_cmdbuf_get_va(buffer) +
+			    buffer->user_size - 4);
+
+	/*
+	 * Kick off the 'sync point' command by replacing the previous
+	 * WAIT with a link to the address in the ring buffer.
+	 */
+	etnaviv_buffer_replace_wait(buffer, waitlink_offset,
+				    VIV_FE_LINK_HEADER_OP_LINK |
+				    VIV_FE_LINK_HEADER_PREFETCH(dwords),
+				    target);
+}
+
 /* Append a command buffer to the ring buffer. */
 void etnaviv_buffer_queue(struct etnaviv_gpu *gpu, unsigned int event,
 	struct etnaviv_cmdbuf *cmdbuf)

drivers/gpu/drm/etnaviv/etnaviv_cmdbuf.c

@@ -19,6 +19,7 @@
 #include "etnaviv_cmdbuf.h"
 #include "etnaviv_gpu.h"
 #include "etnaviv_mmu.h"
+#include "etnaviv_perfmon.h"
 
 #define SUBALLOC_SIZE		SZ_256K
 #define SUBALLOC_GRANULE	SZ_4K
@@ -87,9 +88,10 @@ void etnaviv_cmdbuf_suballoc_destroy(struct etnaviv_cmdbuf_suballoc *suballoc)
 
 struct etnaviv_cmdbuf *
 etnaviv_cmdbuf_new(struct etnaviv_cmdbuf_suballoc *suballoc, u32 size,
-		   size_t nr_bos)
+		   size_t nr_bos, size_t nr_pmrs)
 {
 	struct etnaviv_cmdbuf *cmdbuf;
+	struct etnaviv_perfmon_request *pmrs;
 	size_t sz = size_vstruct(nr_bos, sizeof(cmdbuf->bo_map[0]),
 				 sizeof(*cmdbuf));
 	int granule_offs, order, ret;
@@ -98,6 +100,12 @@ etnaviv_cmdbuf_new(struct etnaviv_cmdbuf_suballoc *suballoc, u32 size,
 	if (!cmdbuf)
 		return NULL;
 
+	sz = sizeof(*pmrs) * nr_pmrs;
+	pmrs = kzalloc(sz, GFP_KERNEL);
+	if (!pmrs)
+		goto out_free_cmdbuf;
+
+	cmdbuf->pmrs = pmrs;
 	cmdbuf->suballoc = suballoc;
 	cmdbuf->size = size;
 
@@ -124,6 +132,10 @@ retry:
 	cmdbuf->vaddr = suballoc->vaddr + cmdbuf->suballoc_offset;
 
 	return cmdbuf;
+
+out_free_cmdbuf:
+	kfree(cmdbuf);
+	return NULL;
 }
 
 void etnaviv_cmdbuf_free(struct etnaviv_cmdbuf *cmdbuf)
@@ -139,6 +151,7 @@ void etnaviv_cmdbuf_free(struct etnaviv_cmdbuf *cmdbuf)
 	suballoc->free_space = 1;
 	mutex_unlock(&suballoc->lock);
 	wake_up_all(&suballoc->free_event);
+	kfree(cmdbuf->pmrs);
 	kfree(cmdbuf);
 }
 

drivers/gpu/drm/etnaviv/etnaviv_cmdbuf.h

@@ -21,6 +21,7 @@
 
 struct etnaviv_gpu;
 struct etnaviv_cmdbuf_suballoc;
+struct etnaviv_perfmon_request;
 
 struct etnaviv_cmdbuf {
 	/* suballocator this cmdbuf is allocated from */
@@ -38,6 +39,9 @@ struct etnaviv_cmdbuf {
 	u32 exec_state;
 	/* per GPU in-flight list */
 	struct list_head node;
+	/* perfmon requests */
+	unsigned int nr_pmrs;
+	struct etnaviv_perfmon_request *pmrs;
 	/* BOs attached to this command buffer */
 	unsigned int nr_bos;
 	struct etnaviv_vram_mapping *bo_map[0];
@@ -49,7 +53,7 @@ void etnaviv_cmdbuf_suballoc_destroy(struct etnaviv_cmdbuf_suballoc *suballoc);
 
 struct etnaviv_cmdbuf *
 etnaviv_cmdbuf_new(struct etnaviv_cmdbuf_suballoc *suballoc, u32 size,
-		   size_t nr_bos);
+		   size_t nr_bos, size_t nr_pmrs);
 void etnaviv_cmdbuf_free(struct etnaviv_cmdbuf *cmdbuf);
 
 u32 etnaviv_cmdbuf_get_va(struct etnaviv_cmdbuf *buf);

drivers/gpu/drm/etnaviv/etnaviv_drv.c

@@ -23,6 +23,7 @@
 #include "etnaviv_gpu.h"
 #include "etnaviv_gem.h"
 #include "etnaviv_mmu.h"
+#include "etnaviv_perfmon.h"
 
 #ifdef CONFIG_DRM_ETNAVIV_REGISTER_LOGGING
 static bool reglog;
@@ -451,6 +452,40 @@ static int etnaviv_ioctl_gem_wait(struct drm_device *dev, void *data,
 	return ret;
 }
 
+static int etnaviv_ioctl_pm_query_dom(struct drm_device *dev, void *data,
+	struct drm_file *file)
+{
+	struct etnaviv_drm_private *priv = dev->dev_private;
+	struct drm_etnaviv_pm_domain *args = data;
+	struct etnaviv_gpu *gpu;
+
+	if (args->pipe >= ETNA_MAX_PIPES)
+		return -EINVAL;
+
+	gpu = priv->gpu[args->pipe];
+	if (!gpu)
+		return -ENXIO;
+
+	return etnaviv_pm_query_dom(gpu, args);
+}
+
+static int etnaviv_ioctl_pm_query_sig(struct drm_device *dev, void *data,
+	struct drm_file *file)
+{
+	struct etnaviv_drm_private *priv = dev->dev_private;
+	struct drm_etnaviv_pm_signal *args = data;
+	struct etnaviv_gpu *gpu;
+
+	if (args->pipe >= ETNA_MAX_PIPES)
+		return -EINVAL;
+
+	gpu = priv->gpu[args->pipe];
+	if (!gpu)
+		return -ENXIO;
+
+	return etnaviv_pm_query_sig(gpu, args);
+}
+
 static const struct drm_ioctl_desc etnaviv_ioctls[] = {
 #define ETNA_IOCTL(n, func, flags) \
 	DRM_IOCTL_DEF_DRV(ETNAVIV_##n, etnaviv_ioctl_##func, flags)
@@ -463,6 +498,8 @@ static const struct drm_ioctl_desc etnaviv_ioctls[] = {
 	ETNA_IOCTL(WAIT_FENCE,   wait_fence,   DRM_AUTH|DRM_RENDER_ALLOW),
 	ETNA_IOCTL(GEM_USERPTR,  gem_userptr,  DRM_AUTH|DRM_RENDER_ALLOW),
 	ETNA_IOCTL(GEM_WAIT,     gem_wait,     DRM_AUTH|DRM_RENDER_ALLOW),
+	ETNA_IOCTL(PM_QUERY_DOM, pm_query_dom, DRM_AUTH|DRM_RENDER_ALLOW),
+	ETNA_IOCTL(PM_QUERY_SIG, pm_query_sig, DRM_AUTH|DRM_RENDER_ALLOW),
 };
 
 static const struct vm_operations_struct vm_ops = {
@@ -513,7 +550,7 @@ static struct drm_driver etnaviv_drm_driver = {
 	.desc               = "etnaviv DRM",
 	.date               = "20151214",
 	.major              = 1,
-	.minor              = 1,
+	.minor              = 2,
 };
 
 /*

drivers/gpu/drm/etnaviv/etnaviv_drv.h

@@ -26,7 +26,6 @@
 #include <linux/pm_runtime.h>
 #include <linux/slab.h>
 #include <linux/list.h>
-#include <linux/iommu.h>
 #include <linux/types.h>
 #include <linux/sizes.h>
 
@@ -92,15 +91,12 @@ int etnaviv_gem_cpu_fini(struct drm_gem_object *obj);
 void etnaviv_gem_free_object(struct drm_gem_object *obj);
 int etnaviv_gem_new_handle(struct drm_device *dev, struct drm_file *file,
 		u32 size, u32 flags, u32 *handle);
-struct drm_gem_object *etnaviv_gem_new_locked(struct drm_device *dev,
-		u32 size, u32 flags);
-struct drm_gem_object *etnaviv_gem_new(struct drm_device *dev,
-		u32 size, u32 flags);
 int etnaviv_gem_new_userptr(struct drm_device *dev, struct drm_file *file,
 	uintptr_t ptr, u32 size, u32 flags, u32 *handle);
 u16 etnaviv_buffer_init(struct etnaviv_gpu *gpu);
 u16 etnaviv_buffer_config_mmuv2(struct etnaviv_gpu *gpu, u32 mtlb_addr, u32 safe_addr);
 void etnaviv_buffer_end(struct etnaviv_gpu *gpu);
+void etnaviv_sync_point_queue(struct etnaviv_gpu *gpu, unsigned int event);
 void etnaviv_buffer_queue(struct etnaviv_gpu *gpu, unsigned int event,
 	struct etnaviv_cmdbuf *cmdbuf);
 void etnaviv_validate_init(void);

drivers/gpu/drm/etnaviv/etnaviv_gem.c

@@ -704,25 +704,6 @@ int etnaviv_gem_new_handle(struct drm_device *dev, struct drm_file *file,
 	return ret;
 }
 
-struct drm_gem_object *etnaviv_gem_new(struct drm_device *dev,
-		u32 size, u32 flags)
-{
-	struct drm_gem_object *obj;
-	int ret;
-
-	obj = __etnaviv_gem_new(dev, size, flags);
-	if (IS_ERR(obj))
-		return obj;
-
-	ret = etnaviv_gem_obj_add(dev, obj);
-	if (ret < 0) {
-		drm_gem_object_put_unlocked(obj);
-		return ERR_PTR(ret);
-	}
-
-	return obj;
-}
-
 int etnaviv_gem_new_private(struct drm_device *dev, size_t size, u32 flags,
 	struct reservation_object *robj, const struct etnaviv_gem_ops *ops,
 	struct etnaviv_gem_object **res)

drivers/gpu/drm/etnaviv/etnaviv_gem_submit.c

@@ -21,6 +21,7 @@
 #include "etnaviv_drv.h"
 #include "etnaviv_gpu.h"
 #include "etnaviv_gem.h"
+#include "etnaviv_perfmon.h"
 
 /*
  * Cmdstream submission:
@@ -283,6 +284,54 @@ static int submit_reloc(struct etnaviv_gem_submit *submit, void *stream,
 	return 0;
 }
 
+static int submit_perfmon_validate(struct etnaviv_gem_submit *submit,
+		struct etnaviv_cmdbuf *cmdbuf,
+		const struct drm_etnaviv_gem_submit_pmr *pmrs,
+		u32 nr_pms)
+{
+	u32 i;
+
+	for (i = 0; i < nr_pms; i++) {
+		const struct drm_etnaviv_gem_submit_pmr *r = pmrs + i;
+		struct etnaviv_gem_submit_bo *bo;
+		int ret;
+
+		ret = submit_bo(submit, r->read_idx, &bo);
+		if (ret)
+			return ret;
+
+		/* at offset 0 a sequence number gets stored used for userspace sync */
+		if (r->read_offset == 0) {
+			DRM_ERROR("perfmon request: offset is 0");
+			return -EINVAL;
+		}
+
+		if (r->read_offset >= bo->obj->base.size - sizeof(u32)) {
+			DRM_ERROR("perfmon request: offset %u outside object", i);
+			return -EINVAL;
+		}
+
+		if (r->flags & ~(ETNA_PM_PROCESS_PRE | ETNA_PM_PROCESS_POST)) {
+			DRM_ERROR("perfmon request: flags are not valid");
+			return -EINVAL;
+		}
+
+		if (etnaviv_pm_req_validate(r, cmdbuf->exec_state)) {
+			DRM_ERROR("perfmon request: domain or signal not valid");
+			return -EINVAL;
+		}
+
+		cmdbuf->pmrs[i].flags = r->flags;
+		cmdbuf->pmrs[i].domain = r->domain;
+		cmdbuf->pmrs[i].signal = r->signal;
+		cmdbuf->pmrs[i].sequence = r->sequence;
+		cmdbuf->pmrs[i].offset = r->read_offset;
+		cmdbuf->pmrs[i].bo_vma = etnaviv_gem_vmap(&bo->obj->base);
+	}
+
+	return 0;
+}
+
 static void submit_cleanup(struct etnaviv_gem_submit *submit)
 {
 	unsigned i;
@@ -306,6 +355,7 @@ int etnaviv_ioctl_gem_submit(struct drm_device *dev, void *data,
 	struct etnaviv_drm_private *priv = dev->dev_private;
 	struct drm_etnaviv_gem_submit *args = data;
 	struct drm_etnaviv_gem_submit_reloc *relocs;
+	struct drm_etnaviv_gem_submit_pmr *pmrs;
 	struct drm_etnaviv_gem_submit_bo *bos;
 	struct etnaviv_gem_submit *submit;
 	struct etnaviv_cmdbuf *cmdbuf;
@@ -347,11 +397,12 @@ int etnaviv_ioctl_gem_submit(struct drm_device *dev, void *data,
 	 */
 	bos = kvmalloc_array(args->nr_bos, sizeof(*bos), GFP_KERNEL);
 	relocs = kvmalloc_array(args->nr_relocs, sizeof(*relocs), GFP_KERNEL);
+	pmrs = kvmalloc_array(args->nr_pmrs, sizeof(*pmrs), GFP_KERNEL);
 	stream = kvmalloc_array(1, args->stream_size, GFP_KERNEL);
 	cmdbuf = etnaviv_cmdbuf_new(gpu->cmdbuf_suballoc,
 				    ALIGN(args->stream_size, 8) + 8,
-				    args->nr_bos);
-	if (!bos || !relocs || !stream || !cmdbuf) {
+				    args->nr_bos, args->nr_pmrs);
+	if (!bos || !relocs || !pmrs || !stream || !cmdbuf) {
 		ret = -ENOMEM;
 		goto err_submit_cmds;
 	}
@@ -373,6 +424,14 @@ int etnaviv_ioctl_gem_submit(struct drm_device *dev, void *data,
 		goto err_submit_cmds;
 	}
 
+	ret = copy_from_user(pmrs, u64_to_user_ptr(args->pmrs),
+			     args->nr_pmrs * sizeof(*pmrs));
+	if (ret) {
+		ret = -EFAULT;
+		goto err_submit_cmds;
+	}
+	cmdbuf->nr_pmrs = args->nr_pmrs;
+
 	ret = copy_from_user(stream, u64_to_user_ptr(args->stream),
 			     args->stream_size);
 	if (ret) {
@@ -441,6 +500,10 @@ int etnaviv_ioctl_gem_submit(struct drm_device *dev, void *data,
 	if (ret)
 		goto out;
 
+	ret = submit_perfmon_validate(submit, cmdbuf, pmrs, args->nr_pmrs);
+	if (ret)
+		goto out;
+
 	memcpy(cmdbuf->vaddr, stream, args->stream_size);
 	cmdbuf->user_size = ALIGN(args->stream_size, 8);
 
@@ -496,6 +559,8 @@ err_submit_cmds:
 		kvfree(bos);
 	if (relocs)
 		kvfree(relocs);
+	if (pmrs)
+		kvfree(pmrs);
 
 	return ret;
 }

drivers/gpu/drm/etnaviv/etnaviv_gpu.c

@@ -25,6 +25,7 @@
 #include "etnaviv_gpu.h"
 #include "etnaviv_gem.h"
 #include "etnaviv_mmu.h"
+#include "etnaviv_perfmon.h"
 #include "common.xml.h"
 #include "state.xml.h"
 #include "state_hi.xml.h"
@@ -420,9 +421,10 @@ static void etnaviv_gpu_update_clock(struct etnaviv_gpu *gpu)
 			     gpu->base_rate_shader >> gpu->freq_scale);
 	} else {
 		unsigned int fscale = 1 << (6 - gpu->freq_scale);
-		u32 clock = VIVS_HI_CLOCK_CONTROL_DISABLE_DEBUG_REGISTERS |
-			    VIVS_HI_CLOCK_CONTROL_FSCALE_VAL(fscale);
+		u32 clock = gpu_read(gpu, VIVS_HI_CLOCK_CONTROL);
 
+		clock &= ~VIVS_HI_CLOCK_CONTROL_FSCALE_VAL__MASK;
+		clock |= VIVS_HI_CLOCK_CONTROL_FSCALE_VAL(fscale);
 		etnaviv_gpu_load_clock(gpu, clock);
 	}
 }
@@ -433,24 +435,14 @@ static int etnaviv_hw_reset(struct etnaviv_gpu *gpu)
 	unsigned long timeout;
 	bool failed = true;
 
-	/* TODO
-	 *
-	 * - clock gating
-	 * - puls eater
-	 * - what about VG?
-	 */
-
 	/* We hope that the GPU resets in under one second */
 	timeout = jiffies + msecs_to_jiffies(1000);
 
 	while (time_is_after_jiffies(timeout)) {
 		/* enable clock */
-		etnaviv_gpu_update_clock(gpu);
-
-		control = gpu_read(gpu, VIVS_HI_CLOCK_CONTROL);
-
-		/* Wait for stable clock.  Vivante's code waited for 1ms */
-		usleep_range(1000, 10000);
+		unsigned int fscale = 1 << (6 - gpu->freq_scale);
+		control = VIVS_HI_CLOCK_CONTROL_FSCALE_VAL(fscale);
+		etnaviv_gpu_load_clock(gpu, control);
 
 		/* isolate the GPU. */
 		control |= VIVS_HI_CLOCK_CONTROL_ISOLATE_GPU;
@@ -461,7 +453,7 @@ static int etnaviv_hw_reset(struct etnaviv_gpu *gpu)
 		gpu_write(gpu, VIVS_HI_CLOCK_CONTROL, control);
 
 		/* wait for reset. */
-		msleep(1);
+		usleep_range(10, 20);
 
 		/* reset soft reset bit. */
 		control &= ~VIVS_HI_CLOCK_CONTROL_SOFT_RESET;
@@ -490,6 +482,10 @@ static int etnaviv_hw_reset(struct etnaviv_gpu *gpu)
 			continue;
 		}
 
+		/* disable debug registers, as they are not normally needed */
+		control |= VIVS_HI_CLOCK_CONTROL_DISABLE_DEBUG_REGISTERS;
+		gpu_write(gpu, VIVS_HI_CLOCK_CONTROL, control);
+
 		failed = false;
 		break;
 	}
@@ -721,7 +717,7 @@ int etnaviv_gpu_init(struct etnaviv_gpu *gpu)
 	}
 
 	/* Create buffer: */
-	gpu->buffer = etnaviv_cmdbuf_new(gpu->cmdbuf_suballoc, PAGE_SIZE, 0);
+	gpu->buffer = etnaviv_cmdbuf_new(gpu->cmdbuf_suballoc, PAGE_SIZE, 0, 0);
 	if (!gpu->buffer) {
 		ret = -ENOMEM;
 		dev_err(gpu->dev, "could not create command buffer\n");
@@ -739,10 +735,9 @@ int etnaviv_gpu_init(struct etnaviv_gpu *gpu)
 	/* Setup event management */
 	spin_lock_init(&gpu->event_spinlock);
 	init_completion(&gpu->event_free);
-	for (i = 0; i < ARRAY_SIZE(gpu->event); i++) {
-		gpu->event[i].used = false;
+	bitmap_zero(gpu->event_bitmap, ETNA_NR_EVENTS);
+	for (i = 0; i < ARRAY_SIZE(gpu->event); i++)
 		complete(&gpu->event_free);
-	}
 
 	/* Now program the hardware */
 	mutex_lock(&gpu->lock);
@@ -926,7 +921,7 @@ static void recover_worker(struct work_struct *work)
 	struct etnaviv_gpu *gpu = container_of(work, struct etnaviv_gpu,
 					       recover_work);
 	unsigned long flags;
-	unsigned int i;
+	unsigned int i = 0;
 
 	dev_err(gpu->dev, "hangcheck recover!\n");
 
@@ -945,14 +940,12 @@ static void recover_worker(struct work_struct *work)
 
 	/* complete all events, the GPU won't do it after the reset */
 	spin_lock_irqsave(&gpu->event_spinlock, flags);
-	for (i = 0; i < ARRAY_SIZE(gpu->event); i++) {
-		if (!gpu->event[i].used)
-			continue;
+	for_each_set_bit_from(i, gpu->event_bitmap, ETNA_NR_EVENTS) {
 		dma_fence_signal(gpu->event[i].fence);
 		gpu->event[i].fence = NULL;
-		gpu->event[i].used = false;
 		complete(&gpu->event_free);
 	}
+	bitmap_zero(gpu->event_bitmap, ETNA_NR_EVENTS);
 	spin_unlock_irqrestore(&gpu->event_spinlock, flags);
 	gpu->completed_fence = gpu->active_fence;
 
@@ -1140,30 +1133,45 @@ int etnaviv_gpu_fence_sync_obj(struct etnaviv_gem_object *etnaviv_obj,
  * event management:
  */
 
-static unsigned int event_alloc(struct etnaviv_gpu *gpu)
+static int event_alloc(struct etnaviv_gpu *gpu, unsigned nr_events,
+	unsigned int *events)
 {
-	unsigned long ret, flags;
-	unsigned int i, event = ~0U;
+	unsigned long flags, timeout = msecs_to_jiffies(10 * 10000);
+	unsigned i, acquired = 0;
 
-	ret = wait_for_completion_timeout(&gpu->event_free,
-					  msecs_to_jiffies(10 * 10000));
-	if (!ret)
-		dev_err(gpu->dev, "wait_for_completion_timeout failed");
+	for (i = 0; i < nr_events; i++) {
+		unsigned long ret;
 
-	spin_lock_irqsave(&gpu->event_spinlock, flags);
+		ret = wait_for_completion_timeout(&gpu->event_free, timeout);
 
-	/* find first free event */
-	for (i = 0; i < ARRAY_SIZE(gpu->event); i++) {
-		if (gpu->event[i].used == false) {
-			gpu->event[i].used = true;
-			event = i;
-			break;
+		if (!ret) {
+			dev_err(gpu->dev, "wait_for_completion_timeout failed");
+			goto out;
 		}
+
+		acquired++;
+		timeout = ret;
+	}
+
+	spin_lock_irqsave(&gpu->event_spinlock, flags);
+
+	for (i = 0; i < nr_events; i++) {
+		int event = find_first_zero_bit(gpu->event_bitmap, ETNA_NR_EVENTS);
+
+		events[i] = event;
+		memset(&gpu->event[event], 0, sizeof(struct etnaviv_event));
+		set_bit(event, gpu->event_bitmap);
 	}
 
 	spin_unlock_irqrestore(&gpu->event_spinlock, flags);
 
-	return event;
+	return 0;
+
+out:
+	for (i = 0; i < acquired; i++)
+		complete(&gpu->event_free);
+
+	return -EBUSY;
 }
 
 static void event_free(struct etnaviv_gpu *gpu, unsigned int event)
@@ -1172,12 +1180,12 @@ static void event_free(struct etnaviv_gpu *gpu, unsigned int event)
 
 	spin_lock_irqsave(&gpu->event_spinlock, flags);
 
-	if (gpu->event[event].used == false) {
+	if (!test_bit(event, gpu->event_bitmap)) {
 		dev_warn(gpu->dev, "event %u is already marked as free",
 			 event);
 		spin_unlock_irqrestore(&gpu->event_spinlock, flags);
 	} else {
-		gpu->event[event].used = false;
+		clear_bit(event, gpu->event_bitmap);
 		spin_unlock_irqrestore(&gpu->event_spinlock, flags);
 
 		complete(&gpu->event_free);
@@ -1311,12 +1319,71 @@ void etnaviv_gpu_pm_put(struct etnaviv_gpu *gpu)
 	pm_runtime_put_autosuspend(gpu->dev);
 }
 
+static void sync_point_perfmon_sample(struct etnaviv_gpu *gpu,
+	struct etnaviv_event *event, unsigned int flags)
+{
+	const struct etnaviv_cmdbuf *cmdbuf = event->cmdbuf;
+	unsigned int i;
+
+	for (i = 0; i < cmdbuf->nr_pmrs; i++) {
+		const struct etnaviv_perfmon_request *pmr = cmdbuf->pmrs + i;
+
+		if (pmr->flags == flags)
+			etnaviv_perfmon_process(gpu, pmr);
+	}
+}
+
+static void sync_point_perfmon_sample_pre(struct etnaviv_gpu *gpu,
+	struct etnaviv_event *event)
+{
+	u32 val;
+
+	/* disable clock gating */
+	val = gpu_read(gpu, VIVS_PM_POWER_CONTROLS);
+	val &= ~VIVS_PM_POWER_CONTROLS_ENABLE_MODULE_CLOCK_GATING;
+	gpu_write(gpu, VIVS_PM_POWER_CONTROLS, val);
+
+	/* enable debug register */
+	val = gpu_read(gpu, VIVS_HI_CLOCK_CONTROL);
+	val &= ~VIVS_HI_CLOCK_CONTROL_DISABLE_DEBUG_REGISTERS;
+	gpu_write(gpu, VIVS_HI_CLOCK_CONTROL, val);
+
+	sync_point_perfmon_sample(gpu, event, ETNA_PM_PROCESS_PRE);
+}
+
+static void sync_point_perfmon_sample_post(struct etnaviv_gpu *gpu,
+	struct etnaviv_event *event)
+{
+	const struct etnaviv_cmdbuf *cmdbuf = event->cmdbuf;
+	unsigned int i;
+	u32 val;
+
+	sync_point_perfmon_sample(gpu, event, ETNA_PM_PROCESS_POST);
+
+	for (i = 0; i < cmdbuf->nr_pmrs; i++) {
+		const struct etnaviv_perfmon_request *pmr = cmdbuf->pmrs + i;
+
+		*pmr->bo_vma = pmr->sequence;
+	}
+
+	/* disable debug register */
+	val = gpu_read(gpu, VIVS_HI_CLOCK_CONTROL);
+	val |= VIVS_HI_CLOCK_CONTROL_DISABLE_DEBUG_REGISTERS;
+	gpu_write(gpu, VIVS_HI_CLOCK_CONTROL, val);
+
+	/* enable clock gating */
+	val = gpu_read(gpu, VIVS_PM_POWER_CONTROLS);
+	val |= VIVS_PM_POWER_CONTROLS_ENABLE_MODULE_CLOCK_GATING;
+	gpu_write(gpu, VIVS_PM_POWER_CONTROLS, val);
+}
+
+
 /* add bo's to gpu's ring, and kick gpu: */
 int etnaviv_gpu_submit(struct etnaviv_gpu *gpu,
 	struct etnaviv_gem_submit *submit, struct etnaviv_cmdbuf *cmdbuf)
 {
 	struct dma_fence *fence;
-	unsigned int event, i;
+	unsigned int i, nr_events = 1, event[3];
 	int ret;
 
 	ret = etnaviv_gpu_pm_get_sync(gpu);
@@ -1332,10 +1399,19 @@ int etnaviv_gpu_submit(struct etnaviv_gpu *gpu,
 	 *
 	 */
 
-	event = event_alloc(gpu);
-	if (unlikely(event == ~0U)) {
-		DRM_ERROR("no free event\n");
-		ret = -EBUSY;
+	/*
+	 * if there are performance monitor requests we need to have
+	 * - a sync point to re-configure gpu and process ETNA_PM_PROCESS_PRE
+	 *   requests.
+	 * - a sync point to re-configure gpu, process ETNA_PM_PROCESS_POST requests
+	 *   and update the sequence number for userspace.
+	 */
+	if (cmdbuf->nr_pmrs)
+		nr_events = 3;
+
+	ret = event_alloc(gpu, nr_events, event);
+	if (ret) {
+		DRM_ERROR("no free events\n");
 		goto out_pm_put;
 	}
 
@@ -1343,12 +1419,14 @@ int etnaviv_gpu_submit(struct etnaviv_gpu *gpu,
 
 	fence = etnaviv_gpu_fence_alloc(gpu);
 	if (!fence) {
-		event_free(gpu, event);
+		for (i = 0; i < nr_events; i++)
+			event_free(gpu, event[i]);
+
 		ret = -ENOMEM;
 		goto out_unlock;
 	}
 
-	gpu->event[event].fence = fence;
+	gpu->event[event[0]].fence = fence;
 	submit->fence = dma_fence_get(fence);
 	gpu->active_fence = submit->fence->seqno;
 
@@ -1358,7 +1436,19 @@ int etnaviv_gpu_submit(struct etnaviv_gpu *gpu,
 		gpu->lastctx = cmdbuf->ctx;
 	}
 
-	etnaviv_buffer_queue(gpu, event, cmdbuf);
+	if (cmdbuf->nr_pmrs) {
+		gpu->event[event[1]].sync_point = &sync_point_perfmon_sample_pre;
+		gpu->event[event[1]].cmdbuf = cmdbuf;
+		etnaviv_sync_point_queue(gpu, event[1]);
+	}
+
+	etnaviv_buffer_queue(gpu, event[0], cmdbuf);
+
+	if (cmdbuf->nr_pmrs) {
+		gpu->event[event[2]].sync_point = &sync_point_perfmon_sample_post;
+		gpu->event[event[2]].cmdbuf = cmdbuf;
+		etnaviv_sync_point_queue(gpu, event[2]);
+	}
 
 	cmdbuf->fence = fence;
 	list_add_tail(&cmdbuf->node, &gpu->active_cmd_list);
@@ -1394,6 +1484,24 @@ out_pm_put:
 	return ret;
 }
 
+static void etnaviv_process_sync_point(struct etnaviv_gpu *gpu,
+	struct etnaviv_event *event)
+{
+	u32 addr = gpu_read(gpu, VIVS_FE_DMA_ADDRESS);
+
+	event->sync_point(gpu, event);
+	etnaviv_gpu_start_fe(gpu, addr + 2, 2);
+}
+
+static void sync_point_worker(struct work_struct *work)
+{
+	struct etnaviv_gpu *gpu = container_of(work, struct etnaviv_gpu,
+					       sync_point_work);
+
+	etnaviv_process_sync_point(gpu, &gpu->event[gpu->sync_point_event]);
+	event_free(gpu, gpu->sync_point_event);
+}
+
 /*
  * Init/Cleanup:
  */
@@ -1440,7 +1548,15 @@ static irqreturn_t irq_handler(int irq, void *data)
 
 			dev_dbg(gpu->dev, "event %u\n", event);
 
+			if (gpu->event[event].sync_point) {
+				gpu->sync_point_event = event;
+				etnaviv_queue_work(gpu->drm, &gpu->sync_point_work);
+			}
+
 			fence = gpu->event[event].fence;
+			if (!fence)
+				continue;
+
 			gpu->event[event].fence = NULL;
 			dma_fence_signal(fence);
 
@@ -1645,6 +1761,7 @@ static int etnaviv_gpu_bind(struct device *dev, struct device *master,
 
 	INIT_LIST_HEAD(&gpu->active_cmd_list);
 	INIT_WORK(&gpu->retire_work, retire_worker);
+	INIT_WORK(&gpu->sync_point_work, sync_point_worker);
 	INIT_WORK(&gpu->recover_work, recover_worker);
 	init_waitqueue_head(&gpu->fence_event);
 

drivers/gpu/drm/etnaviv/etnaviv_gpu.h

@@ -88,13 +88,17 @@ struct etnaviv_chip_identity {
 };
 
 struct etnaviv_event {
-	bool used;
 	struct dma_fence *fence;
+	struct etnaviv_cmdbuf *cmdbuf;
+
+	void (*sync_point)(struct etnaviv_gpu *gpu, struct etnaviv_event *event);
 };
 
 struct etnaviv_cmdbuf_suballoc;
 struct etnaviv_cmdbuf;
 
+#define ETNA_NR_EVENTS 30
+
 struct etnaviv_gpu {
 	struct drm_device *drm;
 	struct thermal_cooling_device *cooling;
@@ -112,7 +116,8 @@ struct etnaviv_gpu {
 	u32 memory_base;
 
 	/* event management: */
-	struct etnaviv_event event[30];
+	DECLARE_BITMAP(event_bitmap, ETNA_NR_EVENTS);
+	struct etnaviv_event event[ETNA_NR_EVENTS];
 	struct completion event_free;
 	spinlock_t event_spinlock;
 
@@ -133,6 +138,10 @@ struct etnaviv_gpu {
 	/* worker for handling active-list retiring: */
 	struct work_struct retire_work;
 
+	/* worker for handling 'sync' points: */
+	struct work_struct sync_point_work;
+	int sync_point_event;
+
 	void __iomem *mmio;
 	int irq;
 

drivers/gpu/drm/etnaviv/etnaviv_iommu.c

@@ -14,7 +14,6 @@
  * this program.  If not, see <http://www.gnu.org/licenses/>.
  */
 
-#include <linux/iommu.h>
 #include <linux/platform_device.h>
 #include <linux/sizes.h>
 #include <linux/slab.h>
@@ -31,174 +30,115 @@
 
 #define GPU_MEM_START	0x80000000
 
-struct etnaviv_iommu_domain_pgtable {
-	u32 *pgtable;
-	dma_addr_t paddr;
+struct etnaviv_iommuv1_domain {
+	struct etnaviv_iommu_domain base;
+	u32 *pgtable_cpu;
+	dma_addr_t pgtable_dma;
 };
 
-struct etnaviv_iommu_domain {
-	struct iommu_domain domain;
-	struct device *dev;
-	void *bad_page_cpu;
-	dma_addr_t bad_page_dma;
-	struct etnaviv_iommu_domain_pgtable pgtable;
-	spinlock_t map_lock;
-};
-
-static struct etnaviv_iommu_domain *to_etnaviv_domain(struct iommu_domain *domain)
-{
-	return container_of(domain, struct etnaviv_iommu_domain, domain);
-}
-
-static int pgtable_alloc(struct etnaviv_iommu_domain_pgtable *pgtable,
-			 size_t size)
-{
-	pgtable->pgtable = dma_alloc_coherent(NULL, size, &pgtable->paddr, GFP_KERNEL);
-	if (!pgtable->pgtable)
-		return -ENOMEM;
-
-	return 0;
-}
-
-static void pgtable_free(struct etnaviv_iommu_domain_pgtable *pgtable,
-			 size_t size)
+static struct etnaviv_iommuv1_domain *
+to_etnaviv_domain(struct etnaviv_iommu_domain *domain)
 {
-	dma_free_coherent(NULL, size, pgtable->pgtable, pgtable->paddr);
-}
-
-static u32 pgtable_read(struct etnaviv_iommu_domain_pgtable *pgtable,
-			   unsigned long iova)
-{
-	/* calcuate index into page table */
-	unsigned int index = (iova - GPU_MEM_START) / SZ_4K;
-	phys_addr_t paddr;
-
-	paddr = pgtable->pgtable[index];
-
-	return paddr;
+	return container_of(domain, struct etnaviv_iommuv1_domain, base);
 }
 
-static void pgtable_write(struct etnaviv_iommu_domain_pgtable *pgtable,
-			  unsigned long iova, phys_addr_t paddr)
-{
-	/* calcuate index into page table */
-	unsigned int index = (iova - GPU_MEM_START) / SZ_4K;
-
-	pgtable->pgtable[index] = paddr;
-}
-
-static int __etnaviv_iommu_init(struct etnaviv_iommu_domain *etnaviv_domain)
+static int __etnaviv_iommu_init(struct etnaviv_iommuv1_domain *etnaviv_domain)
 {
 	u32 *p;
-	int ret, i;
-
-	etnaviv_domain->bad_page_cpu = dma_alloc_coherent(etnaviv_domain->dev,
-						  SZ_4K,
-						  &etnaviv_domain->bad_page_dma,
-						  GFP_KERNEL);
-	if (!etnaviv_domain->bad_page_cpu)
+	int i;
+
+	etnaviv_domain->base.bad_page_cpu = dma_alloc_coherent(
+						etnaviv_domain->base.dev,
+						SZ_4K,
+						&etnaviv_domain->base.bad_page_dma,
+						GFP_KERNEL);
+	if (!etnaviv_domain->base.bad_page_cpu)
 		return -ENOMEM;
 
-	p = etnaviv_domain->bad_page_cpu;
+	p = etnaviv_domain->base.bad_page_cpu;
 	for (i = 0; i < SZ_4K / 4; i++)
 		*p++ = 0xdead55aa;
 
-	ret = pgtable_alloc(&etnaviv_domain->pgtable, PT_SIZE);
-	if (ret < 0) {
-		dma_free_coherent(etnaviv_domain->dev, SZ_4K,
-				  etnaviv_domain->bad_page_cpu,
-				  etnaviv_domain->bad_page_dma);
-		return ret;
+	etnaviv_domain->pgtable_cpu =
+			dma_alloc_coherent(etnaviv_domain->base.dev, PT_SIZE,
+					   &etnaviv_domain->pgtable_dma,
+					   GFP_KERNEL);
+	if (!etnaviv_domain->pgtable_cpu) {
+		dma_free_coherent(etnaviv_domain->base.dev, SZ_4K,
+				  etnaviv_domain->base.bad_page_cpu,
+				  etnaviv_domain->base.bad_page_dma);
+		return -ENOMEM;
 	}
 
 	for (i = 0; i < PT_ENTRIES; i++)
-		etnaviv_domain->pgtable.pgtable[i] =
-			etnaviv_domain->bad_page_dma;
-
-	spin_lock_init(&etnaviv_domain->map_lock);
+		etnaviv_domain->pgtable_cpu[i] =
+				etnaviv_domain->base.bad_page_dma;
 
 	return 0;
 }
 
-static void etnaviv_domain_free(struct iommu_domain *domain)
+static void etnaviv_iommuv1_domain_free(struct etnaviv_iommu_domain *domain)
 {
-	struct etnaviv_iommu_domain *etnaviv_domain = to_etnaviv_domain(domain);
+	struct etnaviv_iommuv1_domain *etnaviv_domain =
+			to_etnaviv_domain(domain);
 
-	pgtable_free(&etnaviv_domain->pgtable, PT_SIZE);
+	dma_free_coherent(etnaviv_domain->base.dev, PT_SIZE,
+			  etnaviv_domain->pgtable_cpu,
+			  etnaviv_domain->pgtable_dma);
 
-	dma_free_coherent(etnaviv_domain->dev, SZ_4K,
-			  etnaviv_domain->bad_page_cpu,
-			  etnaviv_domain->bad_page_dma);
+	dma_free_coherent(etnaviv_domain->base.dev, SZ_4K,
+			  etnaviv_domain->base.bad_page_cpu,
+			  etnaviv_domain->base.bad_page_dma);
 
 	kfree(etnaviv_domain);
 }
 
-static int etnaviv_iommuv1_map(struct iommu_domain *domain, unsigned long iova,
-	   phys_addr_t paddr, size_t size, int prot)
+static int etnaviv_iommuv1_map(struct etnaviv_iommu_domain *domain,
+			       unsigned long iova, phys_addr_t paddr,
+			       size_t size, int prot)
 {
-	struct etnaviv_iommu_domain *etnaviv_domain = to_etnaviv_domain(domain);
+	struct etnaviv_iommuv1_domain *etnaviv_domain = to_etnaviv_domain(domain);
+	unsigned int index = (iova - GPU_MEM_START) / SZ_4K;
 
 	if (size != SZ_4K)
 		return -EINVAL;
 
-	spin_lock(&etnaviv_domain->map_lock);
-	pgtable_write(&etnaviv_domain->pgtable, iova, paddr);
-	spin_unlock(&etnaviv_domain->map_lock);
+	etnaviv_domain->pgtable_cpu[index] = paddr;
 
 	return 0;
 }
 
-static size_t etnaviv_iommuv1_unmap(struct iommu_domain *domain,
+static size_t etnaviv_iommuv1_unmap(struct etnaviv_iommu_domain *domain,
 	unsigned long iova, size_t size)
 {
-	struct etnaviv_iommu_domain *etnaviv_domain = to_etnaviv_domain(domain);
+	struct etnaviv_iommuv1_domain *etnaviv_domain =
+			to_etnaviv_domain(domain);
+	unsigned int index = (iova - GPU_MEM_START) / SZ_4K;
 
 	if (size != SZ_4K)
 		return -EINVAL;
 
-	spin_lock(&etnaviv_domain->map_lock);
-	pgtable_write(&etnaviv_domain->pgtable, iova,
-		      etnaviv_domain->bad_page_dma);
-	spin_unlock(&etnaviv_domain->map_lock);
+	etnaviv_domain->pgtable_cpu[index] = etnaviv_domain->base.bad_page_dma;
 
 	return SZ_4K;
 }
 
-static phys_addr_t etnaviv_iommu_iova_to_phys(struct iommu_domain *domain,
-	dma_addr_t iova)
-{
-	struct etnaviv_iommu_domain *etnaviv_domain = to_etnaviv_domain(domain);
-
-	return pgtable_read(&etnaviv_domain->pgtable, iova);
-}
-
-static size_t etnaviv_iommuv1_dump_size(struct iommu_domain *domain)
+static size_t etnaviv_iommuv1_dump_size(struct etnaviv_iommu_domain *domain)
 {
 	return PT_SIZE;
 }
 
-static void etnaviv_iommuv1_dump(struct iommu_domain *domain, void *buf)
+static void etnaviv_iommuv1_dump(struct etnaviv_iommu_domain *domain, void *buf)
 {
-	struct etnaviv_iommu_domain *etnaviv_domain = to_etnaviv_domain(domain);
+	struct etnaviv_iommuv1_domain *etnaviv_domain =
+			to_etnaviv_domain(domain);
 
-	memcpy(buf, etnaviv_domain->pgtable.pgtable, PT_SIZE);
+	memcpy(buf, etnaviv_domain->pgtable_cpu, PT_SIZE);
 }
 
-static const struct etnaviv_iommu_ops etnaviv_iommu_ops = {
-	.ops = {
-		.domain_free = etnaviv_domain_free,
-		.map = etnaviv_iommuv1_map,
-		.unmap = etnaviv_iommuv1_unmap,
-		.iova_to_phys = etnaviv_iommu_iova_to_phys,
-		.pgsize_bitmap = SZ_4K,
-	},
-	.dump_size = etnaviv_iommuv1_dump_size,
-	.dump = etnaviv_iommuv1_dump,
-};
-
 void etnaviv_iommuv1_restore(struct etnaviv_gpu *gpu)
 {
-	struct etnaviv_iommu_domain *etnaviv_domain =
+	struct etnaviv_iommuv1_domain *etnaviv_domain =
 			to_etnaviv_domain(gpu->mmu->domain);
 	u32 pgtable;
 
@@ -210,7 +150,7 @@ void etnaviv_iommuv1_restore(struct etnaviv_gpu *gpu)
 	gpu_write(gpu, VIVS_MC_MEMORY_BASE_ADDR_PE, gpu->memory_base);
 
 	/* set page table address in MC */
-	pgtable = (u32)etnaviv_domain->pgtable.paddr;
+	pgtable = (u32)etnaviv_domain->pgtable_dma;
 
 	gpu_write(gpu, VIVS_MC_MMU_FE_PAGE_TABLE, pgtable);
 	gpu_write(gpu, VIVS_MC_MMU_TX_PAGE_TABLE, pgtable);
@@ -219,28 +159,37 @@ void etnaviv_iommuv1_restore(struct etnaviv_gpu *gpu)
 	gpu_write(gpu, VIVS_MC_MMU_RA_PAGE_TABLE, pgtable);
 }
 
-struct iommu_domain *etnaviv_iommuv1_domain_alloc(struct etnaviv_gpu *gpu)
+const struct etnaviv_iommu_domain_ops etnaviv_iommuv1_ops = {
+	.free = etnaviv_iommuv1_domain_free,
+	.map = etnaviv_iommuv1_map,
+	.unmap = etnaviv_iommuv1_unmap,
+	.dump_size = etnaviv_iommuv1_dump_size,
+	.dump = etnaviv_iommuv1_dump,
+};
+
+struct etnaviv_iommu_domain *
+etnaviv_iommuv1_domain_alloc(struct etnaviv_gpu *gpu)
 {
-	struct etnaviv_iommu_domain *etnaviv_domain;
+	struct etnaviv_iommuv1_domain *etnaviv_domain;
+	struct etnaviv_iommu_domain *domain;
 	int ret;
 
 	etnaviv_domain = kzalloc(sizeof(*etnaviv_domain), GFP_KERNEL);
 	if (!etnaviv_domain)
 		return NULL;
 
-	etnaviv_domain->dev = gpu->dev;
+	domain = &etnaviv_domain->base;
 
-	etnaviv_domain->domain.type = __IOMMU_DOMAIN_PAGING;
-	etnaviv_domain->domain.ops = &etnaviv_iommu_ops.ops;
-	etnaviv_domain->domain.pgsize_bitmap = SZ_4K;
-	etnaviv_domain->domain.geometry.aperture_start = GPU_MEM_START;
-	etnaviv_domain->domain.geometry.aperture_end = GPU_MEM_START + PT_ENTRIES * SZ_4K - 1;
+	domain->dev = gpu->dev;
+	domain->base = GPU_MEM_START;
+	domain->size = PT_ENTRIES * SZ_4K;
+	domain->ops = &etnaviv_iommuv1_ops;
 
 	ret = __etnaviv_iommu_init(etnaviv_domain);
 	if (ret)
 		goto out_free;
 
-	return &etnaviv_domain->domain;
+	return &etnaviv_domain->base;
 
 out_free:
 	kfree(etnaviv_domain);

drivers/gpu/drm/etnaviv/etnaviv_iommu.h

@@ -18,11 +18,14 @@
 #define __ETNAVIV_IOMMU_H__
 
 struct etnaviv_gpu;
+struct etnaviv_iommu_domain;
 
-struct iommu_domain *etnaviv_iommuv1_domain_alloc(struct etnaviv_gpu *gpu);
+struct etnaviv_iommu_domain *
+etnaviv_iommuv1_domain_alloc(struct etnaviv_gpu *gpu);
 void etnaviv_iommuv1_restore(struct etnaviv_gpu *gpu);
 
-struct iommu_domain *etnaviv_iommuv2_domain_alloc(struct etnaviv_gpu *gpu);
+struct etnaviv_iommu_domain *
+etnaviv_iommuv2_domain_alloc(struct etnaviv_gpu *gpu);
 void etnaviv_iommuv2_restore(struct etnaviv_gpu *gpu);
 
 #endif /* __ETNAVIV_IOMMU_H__ */

drivers/gpu/drm/etnaviv/etnaviv_iommu_v2.c

@@ -14,7 +14,6 @@
  * this program.  If not, see <http://www.gnu.org/licenses/>.
  */
 
-#include <linux/iommu.h>
 #include <linux/platform_device.h>
 #include <linux/sizes.h>
 #include <linux/slab.h>
@@ -40,10 +39,7 @@
 #define MMUv2_MAX_STLB_ENTRIES		1024
 
 struct etnaviv_iommuv2_domain {
-	struct iommu_domain domain;
-	struct device *dev;
-	void *bad_page_cpu;
-	dma_addr_t bad_page_dma;
+	struct etnaviv_iommu_domain base;
 	/* M(aster) TLB aka first level pagetable */
 	u32 *mtlb_cpu;
 	dma_addr_t mtlb_dma;
@@ -52,13 +48,15 @@ struct etnaviv_iommuv2_domain {
 	dma_addr_t stlb_dma[1024];
 };
 
-static struct etnaviv_iommuv2_domain *to_etnaviv_domain(struct iommu_domain *domain)
+static struct etnaviv_iommuv2_domain *
+to_etnaviv_domain(struct etnaviv_iommu_domain *domain)
 {
-	return container_of(domain, struct etnaviv_iommuv2_domain, domain);
+	return container_of(domain, struct etnaviv_iommuv2_domain, base);
 }
 
-static int etnaviv_iommuv2_map(struct iommu_domain *domain, unsigned long iova,
-	   phys_addr_t paddr, size_t size, int prot)
+static int etnaviv_iommuv2_map(struct etnaviv_iommu_domain *domain,
+			       unsigned long iova, phys_addr_t paddr,
+			       size_t size, int prot)
 {
 	struct etnaviv_iommuv2_domain *etnaviv_domain =
 			to_etnaviv_domain(domain);
@@ -68,7 +66,7 @@ static int etnaviv_iommuv2_map(struct iommu_domain *domain, unsigned long iova,
 	if (size != SZ_4K)
 		return -EINVAL;
 
-	if (prot & IOMMU_WRITE)
+	if (prot & ETNAVIV_PROT_WRITE)
 		entry |= MMUv2_PTE_WRITEABLE;
 
 	mtlb_entry = (iova & MMUv2_MTLB_MASK) >> MMUv2_MTLB_SHIFT;
@@ -79,8 +77,8 @@ static int etnaviv_iommuv2_map(struct iommu_domain *domain, unsigned long iova,
 	return 0;
 }
 
-static size_t etnaviv_iommuv2_unmap(struct iommu_domain *domain,
-	unsigned long iova, size_t size)
+static size_t etnaviv_iommuv2_unmap(struct etnaviv_iommu_domain *domain,
+				    unsigned long iova, size_t size)
 {
 	struct etnaviv_iommuv2_domain *etnaviv_domain =
 			to_etnaviv_domain(domain);
@@ -97,38 +95,26 @@ static size_t etnaviv_iommuv2_unmap(struct iommu_domain *domain,
 	return SZ_4K;
 }
 
-static phys_addr_t etnaviv_iommuv2_iova_to_phys(struct iommu_domain *domain,
-	dma_addr_t iova)
-{
-	struct etnaviv_iommuv2_domain *etnaviv_domain =
-			to_etnaviv_domain(domain);
-	int mtlb_entry, stlb_entry;
-
-	mtlb_entry = (iova & MMUv2_MTLB_MASK) >> MMUv2_MTLB_SHIFT;
-	stlb_entry = (iova & MMUv2_STLB_MASK) >> MMUv2_STLB_SHIFT;
-
-	return etnaviv_domain->stlb_cpu[mtlb_entry][stlb_entry] & ~(SZ_4K - 1);
-}
-
 static int etnaviv_iommuv2_init(struct etnaviv_iommuv2_domain *etnaviv_domain)
 {
 	u32 *p;
 	int ret, i, j;
 
 	/* allocate scratch page */
-	etnaviv_domain->bad_page_cpu = dma_alloc_coherent(etnaviv_domain->dev,
-						  SZ_4K,
-						  &etnaviv_domain->bad_page_dma,
-						  GFP_KERNEL);
-	if (!etnaviv_domain->bad_page_cpu) {
+	etnaviv_domain->base.bad_page_cpu = dma_alloc_coherent(
+						etnaviv_domain->base.dev,
+						SZ_4K,
+						&etnaviv_domain->base.bad_page_dma,
+						GFP_KERNEL);
+	if (!etnaviv_domain->base.bad_page_cpu) {
 		ret = -ENOMEM;
 		goto fail_mem;
 	}
-	p = etnaviv_domain->bad_page_cpu;
+	p = etnaviv_domain->base.bad_page_cpu;
 	for (i = 0; i < SZ_4K / 4; i++)
 		*p++ = 0xdead55aa;
 
-	etnaviv_domain->mtlb_cpu = dma_alloc_coherent(etnaviv_domain->dev,
+	etnaviv_domain->mtlb_cpu = dma_alloc_coherent(etnaviv_domain->base.dev,
 						  SZ_4K,
 						  &etnaviv_domain->mtlb_dma,
 						  GFP_KERNEL);
@@ -140,7 +126,7 @@ static int etnaviv_iommuv2_init(struct etnaviv_iommuv2_domain *etnaviv_domain)
 	/* pre-populate STLB pages (may want to switch to on-demand later) */
 	for (i = 0; i < MMUv2_MAX_STLB_ENTRIES; i++) {
 		etnaviv_domain->stlb_cpu[i] =
-				dma_alloc_coherent(etnaviv_domain->dev,
+				dma_alloc_coherent(etnaviv_domain->base.dev,
 						   SZ_4K,
 						   &etnaviv_domain->stlb_dma[i],
 						   GFP_KERNEL);
@@ -159,19 +145,19 @@ static int etnaviv_iommuv2_init(struct etnaviv_iommuv2_domain *etnaviv_domain)
 	return 0;
 
 fail_mem:
-	if (etnaviv_domain->bad_page_cpu)
-		dma_free_coherent(etnaviv_domain->dev, SZ_4K,
-				  etnaviv_domain->bad_page_cpu,
-				  etnaviv_domain->bad_page_dma);
+	if (etnaviv_domain->base.bad_page_cpu)
+		dma_free_coherent(etnaviv_domain->base.dev, SZ_4K,
+				  etnaviv_domain->base.bad_page_cpu,
+				  etnaviv_domain->base.bad_page_dma);
 
 	if (etnaviv_domain->mtlb_cpu)
-		dma_free_coherent(etnaviv_domain->dev, SZ_4K,
+		dma_free_coherent(etnaviv_domain->base.dev, SZ_4K,
 				  etnaviv_domain->mtlb_cpu,
 				  etnaviv_domain->mtlb_dma);
 
 	for (i = 0; i < MMUv2_MAX_STLB_ENTRIES; i++) {
 		if (etnaviv_domain->stlb_cpu[i])
-			dma_free_coherent(etnaviv_domain->dev, SZ_4K,
+			dma_free_coherent(etnaviv_domain->base.dev, SZ_4K,
 					  etnaviv_domain->stlb_cpu[i],
 					  etnaviv_domain->stlb_dma[i]);
 	}
@@ -179,23 +165,23 @@ fail_mem:
 	return ret;
 }
 
-static void etnaviv_iommuv2_domain_free(struct iommu_domain *domain)
+static void etnaviv_iommuv2_domain_free(struct etnaviv_iommu_domain *domain)
 {
 	struct etnaviv_iommuv2_domain *etnaviv_domain =
 			to_etnaviv_domain(domain);
 	int i;
 
-	dma_free_coherent(etnaviv_domain->dev, SZ_4K,
-			  etnaviv_domain->bad_page_cpu,
-			  etnaviv_domain->bad_page_dma);
+	dma_free_coherent(etnaviv_domain->base.dev, SZ_4K,
+			  etnaviv_domain->base.bad_page_cpu,
+			  etnaviv_domain->base.bad_page_dma);
 
-	dma_free_coherent(etnaviv_domain->dev, SZ_4K,
+	dma_free_coherent(etnaviv_domain->base.dev, SZ_4K,
 			  etnaviv_domain->mtlb_cpu,
 			  etnaviv_domain->mtlb_dma);
 
 	for (i = 0; i < MMUv2_MAX_STLB_ENTRIES; i++) {
 		if (etnaviv_domain->stlb_cpu[i])
-			dma_free_coherent(etnaviv_domain->dev, SZ_4K,
+			dma_free_coherent(etnaviv_domain->base.dev, SZ_4K,
 					  etnaviv_domain->stlb_cpu[i],
 					  etnaviv_domain->stlb_dma[i]);
 	}
@@ -203,7 +189,7 @@ static void etnaviv_iommuv2_domain_free(struct iommu_domain *domain)
 	vfree(etnaviv_domain);
 }
 
-static size_t etnaviv_iommuv2_dump_size(struct iommu_domain *domain)
+static size_t etnaviv_iommuv2_dump_size(struct etnaviv_iommu_domain *domain)
 {
 	struct etnaviv_iommuv2_domain *etnaviv_domain =
 			to_etnaviv_domain(domain);
@@ -217,7 +203,7 @@ static size_t etnaviv_iommuv2_dump_size(struct iommu_domain *domain)
 	return dump_size;
 }
 
-static void etnaviv_iommuv2_dump(struct iommu_domain *domain, void *buf)
+static void etnaviv_iommuv2_dump(struct etnaviv_iommu_domain *domain, void *buf)
 {
 	struct etnaviv_iommuv2_domain *etnaviv_domain =
 			to_etnaviv_domain(domain);
@@ -230,18 +216,6 @@ static void etnaviv_iommuv2_dump(struct iommu_domain *domain, void *buf)
 			memcpy(buf, etnaviv_domain->stlb_cpu[i], SZ_4K);
 }
 
-static const struct etnaviv_iommu_ops etnaviv_iommu_ops = {
-	.ops = {
-		.domain_free = etnaviv_iommuv2_domain_free,
-		.map = etnaviv_iommuv2_map,
-		.unmap = etnaviv_iommuv2_unmap,
-		.iova_to_phys = etnaviv_iommuv2_iova_to_phys,
-		.pgsize_bitmap = SZ_4K,
-	},
-	.dump_size = etnaviv_iommuv2_dump_size,
-	.dump = etnaviv_iommuv2_dump,
-};
-
 void etnaviv_iommuv2_restore(struct etnaviv_gpu *gpu)
 {
 	struct etnaviv_iommuv2_domain *etnaviv_domain =
@@ -254,35 +228,45 @@ void etnaviv_iommuv2_restore(struct etnaviv_gpu *gpu)
 
 	prefetch = etnaviv_buffer_config_mmuv2(gpu,
 				(u32)etnaviv_domain->mtlb_dma,
-				(u32)etnaviv_domain->bad_page_dma);
+				(u32)etnaviv_domain->base.bad_page_dma);
 	etnaviv_gpu_start_fe(gpu, (u32)etnaviv_cmdbuf_get_pa(gpu->buffer),
 			     prefetch);
 	etnaviv_gpu_wait_idle(gpu, 100);
 
 	gpu_write(gpu, VIVS_MMUv2_CONTROL, VIVS_MMUv2_CONTROL_ENABLE);
 }
-struct iommu_domain *etnaviv_iommuv2_domain_alloc(struct etnaviv_gpu *gpu)
+
+const struct etnaviv_iommu_domain_ops etnaviv_iommuv2_ops = {
+	.free = etnaviv_iommuv2_domain_free,
+	.map = etnaviv_iommuv2_map,
+	.unmap = etnaviv_iommuv2_unmap,
+	.dump_size = etnaviv_iommuv2_dump_size,
+	.dump = etnaviv_iommuv2_dump,
+};
+
+struct etnaviv_iommu_domain *
+etnaviv_iommuv2_domain_alloc(struct etnaviv_gpu *gpu)
 {
 	struct etnaviv_iommuv2_domain *etnaviv_domain;
+	struct etnaviv_iommu_domain *domain;
 	int ret;
 
 	etnaviv_domain = vzalloc(sizeof(*etnaviv_domain));
 	if (!etnaviv_domain)
 		return NULL;
 
-	etnaviv_domain->dev = gpu->dev;
+	domain = &etnaviv_domain->base;
 
-	etnaviv_domain->domain.type = __IOMMU_DOMAIN_PAGING;
-	etnaviv_domain->domain.ops = &etnaviv_iommu_ops.ops;
-	etnaviv_domain->domain.pgsize_bitmap = SZ_4K;
-	etnaviv_domain->domain.geometry.aperture_start = 0;
-	etnaviv_domain->domain.geometry.aperture_end = ~0UL & ~(SZ_4K - 1);
+	domain->dev = gpu->dev;
+	domain->base = 0;
+	domain->size = (u64)SZ_1G * 4;
+	domain->ops = &etnaviv_iommuv2_ops;
 
 	ret = etnaviv_iommuv2_init(etnaviv_domain);
 	if (ret)
 		goto out_free;
 
-	return &etnaviv_domain->domain;
+	return &etnaviv_domain->base;
 
 out_free:
 	vfree(etnaviv_domain);

drivers/gpu/drm/etnaviv/etnaviv_mmu.c

@@ -22,17 +22,64 @@
 #include "etnaviv_iommu.h"
 #include "etnaviv_mmu.h"
 
-static int etnaviv_fault_handler(struct iommu_domain *iommu, struct device *dev,
-		unsigned long iova, int flags, void *arg)
+static void etnaviv_domain_unmap(struct etnaviv_iommu_domain *domain,
+				 unsigned long iova, size_t size)
 {
-	DBG("*** fault: iova=%08lx, flags=%d", iova, flags);
-	return 0;
+	size_t unmapped_page, unmapped = 0;
+	size_t pgsize = SZ_4K;
+
+	if (!IS_ALIGNED(iova | size, pgsize)) {
+		pr_err("unaligned: iova 0x%lx size 0x%zx min_pagesz 0x%x\n",
+		       iova, size, pgsize);
+		return;
+	}
+
+	while (unmapped < size) {
+		unmapped_page = domain->ops->unmap(domain, iova, pgsize);
+		if (!unmapped_page)
+			break;
+
+		iova += unmapped_page;
+		unmapped += unmapped_page;
+	}
 }
 
-int etnaviv_iommu_map(struct etnaviv_iommu *iommu, u32 iova,
-		struct sg_table *sgt, unsigned len, int prot)
+static int etnaviv_domain_map(struct etnaviv_iommu_domain *domain,
+			      unsigned long iova, phys_addr_t paddr,
+			      size_t size, int prot)
 {
-	struct iommu_domain *domain = iommu->domain;
+	unsigned long orig_iova = iova;
+	size_t pgsize = SZ_4K;
+	size_t orig_size = size;
+	int ret = 0;
+
+	if (!IS_ALIGNED(iova | paddr | size, pgsize)) {
+		pr_err("unaligned: iova 0x%lx pa %pa size 0x%zx min_pagesz 0x%x\n",
+		       iova, &paddr, size, pgsize);
+		return -EINVAL;
+	}
+
+	while (size) {
+		ret = domain->ops->map(domain, iova, paddr, pgsize, prot);
+		if (ret)
+			break;
+
+		iova += pgsize;
+		paddr += pgsize;
+		size -= pgsize;
+	}
+
+	/* unroll mapping in case something went wrong */
+	if (ret)
+		etnaviv_domain_unmap(domain, orig_iova, orig_size - size);
+
+	return ret;
+}
+
+static int etnaviv_iommu_map(struct etnaviv_iommu *iommu, u32 iova,
+			     struct sg_table *sgt, unsigned len, int prot)
+{
+	struct etnaviv_iommu_domain *domain = iommu->domain;
 	struct scatterlist *sg;
 	unsigned int da = iova;
 	unsigned int i, j;
@@ -47,7 +94,7 @@ int etnaviv_iommu_map(struct etnaviv_iommu *iommu, u32 iova,
 
 		VERB("map[%d]: %08x %08x(%zx)", i, iova, pa, bytes);
 
-		ret = iommu_map(domain, da, pa, bytes, prot);
+		ret = etnaviv_domain_map(domain, da, pa, bytes, prot);
 		if (ret)
 			goto fail;
 
@@ -62,27 +109,24 @@ fail:
 	for_each_sg(sgt->sgl, sg, i, j) {
 		size_t bytes = sg_dma_len(sg) + sg->offset;
 
-		iommu_unmap(domain, da, bytes);
+		etnaviv_domain_unmap(domain, da, bytes);
 		da += bytes;
 	}
 	return ret;
 }
 
-int etnaviv_iommu_unmap(struct etnaviv_iommu *iommu, u32 iova,
-		struct sg_table *sgt, unsigned len)
+static void etnaviv_iommu_unmap(struct etnaviv_iommu *iommu, u32 iova,
+				struct sg_table *sgt, unsigned len)
 {
-	struct iommu_domain *domain = iommu->domain;
+	struct etnaviv_iommu_domain *domain = iommu->domain;
 	struct scatterlist *sg;
 	unsigned int da = iova;
 	int i;
 
 	for_each_sg(sgt->sgl, sg, sgt->nents, i) {
 		size_t bytes = sg_dma_len(sg) + sg->offset;
-		size_t unmapped;
 
-		unmapped = iommu_unmap(domain, da, bytes);
-		if (unmapped < bytes)
-			return unmapped;
+		etnaviv_domain_unmap(domain, da, bytes);
 
 		VERB("unmap[%d]: %08x(%zx)", i, iova, bytes);
 
@@ -90,8 +134,6 @@ int etnaviv_iommu_unmap(struct etnaviv_iommu *iommu, u32 iova,
 
 		da += bytes;
 	}
-
-	return 0;
 }
 
 static void etnaviv_iommu_remove_mapping(struct etnaviv_iommu *mmu,
@@ -237,7 +279,7 @@ int etnaviv_iommu_map_gem(struct etnaviv_iommu *mmu,
 	mmu->last_iova = node->start + etnaviv_obj->base.size;
 	mapping->iova = node->start;
 	ret = etnaviv_iommu_map(mmu, node->start, sgt, etnaviv_obj->base.size,
-				IOMMU_READ | IOMMU_WRITE);
+				ETNAVIV_PROT_READ | ETNAVIV_PROT_WRITE);
 
 	if (ret < 0) {
 		drm_mm_remove_node(node);
@@ -271,7 +313,7 @@ void etnaviv_iommu_unmap_gem(struct etnaviv_iommu *mmu,
 void etnaviv_iommu_destroy(struct etnaviv_iommu *mmu)
 {
 	drm_mm_takedown(&mmu->mm);
-	iommu_domain_free(mmu->domain);
+	mmu->domain->ops->free(mmu->domain);
 	kfree(mmu);
 }
 
@@ -303,11 +345,7 @@ struct etnaviv_iommu *etnaviv_iommu_new(struct etnaviv_gpu *gpu)
 	mutex_init(&mmu->lock);
 	INIT_LIST_HEAD(&mmu->mappings);
 
-	drm_mm_init(&mmu->mm, mmu->domain->geometry.aperture_start,
-		    mmu->domain->geometry.aperture_end -
-		    mmu->domain->geometry.aperture_start + 1);
-
-	iommu_set_fault_handler(mmu->domain, etnaviv_fault_handler, gpu->dev);
+	drm_mm_init(&mmu->mm, mmu->domain->base, mmu->domain->size);
 
 	return mmu;
 }
@@ -338,8 +376,8 @@ int etnaviv_iommu_get_suballoc_va(struct etnaviv_gpu *gpu, dma_addr_t paddr,
 			mutex_unlock(&mmu->lock);
 			return ret;
 		}
-		ret = iommu_map(mmu->domain, vram_node->start, paddr, size,
-				IOMMU_READ);
+		ret = etnaviv_domain_map(mmu->domain, vram_node->start, paddr,
+					 size, ETNAVIV_PROT_READ);
 		if (ret < 0) {
 			drm_mm_remove_node(vram_node);
 			mutex_unlock(&mmu->lock);
@@ -362,25 +400,17 @@ void etnaviv_iommu_put_suballoc_va(struct etnaviv_gpu *gpu,
 
 	if (mmu->version == ETNAVIV_IOMMU_V2) {
 		mutex_lock(&mmu->lock);
-		iommu_unmap(mmu->domain,iova, size);
+		etnaviv_domain_unmap(mmu->domain, iova, size);
 		drm_mm_remove_node(vram_node);
 		mutex_unlock(&mmu->lock);
 	}
 }
 size_t etnaviv_iommu_dump_size(struct etnaviv_iommu *iommu)
 {
-	struct etnaviv_iommu_ops *ops;
-
-	ops = container_of(iommu->domain->ops, struct etnaviv_iommu_ops, ops);
-
-	return ops->dump_size(iommu->domain);
+	return iommu->domain->ops->dump_size(iommu->domain);
 }
 
 void etnaviv_iommu_dump(struct etnaviv_iommu *iommu, void *buf)
 {
-	struct etnaviv_iommu_ops *ops;
-
-	ops = container_of(iommu->domain->ops, struct etnaviv_iommu_ops, ops);
-
-	ops->dump(iommu->domain, buf);
+	iommu->domain->ops->dump(iommu->domain, buf);
 }

drivers/gpu/drm/etnaviv/etnaviv_mmu.h

@@ -17,7 +17,8 @@
 #ifndef __ETNAVIV_MMU_H__
 #define __ETNAVIV_MMU_H__
 
-#include <linux/iommu.h>
+#define ETNAVIV_PROT_READ	(1 << 0)
+#define ETNAVIV_PROT_WRITE	(1 << 1)
 
 enum etnaviv_iommu_version {
 	ETNAVIV_IOMMU_V1 = 0,
@@ -26,16 +27,31 @@ enum etnaviv_iommu_version {
 
 struct etnaviv_gpu;
 struct etnaviv_vram_mapping;
+struct etnaviv_iommu_domain;
 
-struct etnaviv_iommu_ops {
-	struct iommu_ops ops;
-	size_t (*dump_size)(struct iommu_domain *);
-	void (*dump)(struct iommu_domain *, void *);
+struct etnaviv_iommu_domain_ops {
+	void (*free)(struct etnaviv_iommu_domain *);
+	int (*map)(struct etnaviv_iommu_domain *domain, unsigned long iova,
+		   phys_addr_t paddr, size_t size, int prot);
+	size_t (*unmap)(struct etnaviv_iommu_domain *domain, unsigned long iova,
+			size_t size);
+	size_t (*dump_size)(struct etnaviv_iommu_domain *);
+	void (*dump)(struct etnaviv_iommu_domain *, void *);
+};
+
+struct etnaviv_iommu_domain {
+	struct device *dev;
+	void *bad_page_cpu;
+	dma_addr_t bad_page_dma;
+	u64 base;
+	u64 size;
+
+	const struct etnaviv_iommu_domain_ops *ops;
 };
 
 struct etnaviv_iommu {
 	struct etnaviv_gpu *gpu;
-	struct iommu_domain *domain;
+	struct etnaviv_iommu_domain *domain;
 
 	enum etnaviv_iommu_version version;
 
@@ -49,18 +65,11 @@ struct etnaviv_iommu {
 
 struct etnaviv_gem_object;
 
-int etnaviv_iommu_attach(struct etnaviv_iommu *iommu, const char **names,
-	int cnt);
-int etnaviv_iommu_map(struct etnaviv_iommu *iommu, u32 iova,
-	struct sg_table *sgt, unsigned len, int prot);
-int etnaviv_iommu_unmap(struct etnaviv_iommu *iommu, u32 iova,
-	struct sg_table *sgt, unsigned len);
 int etnaviv_iommu_map_gem(struct etnaviv_iommu *mmu,
 	struct etnaviv_gem_object *etnaviv_obj, u32 memory_base,
 	struct etnaviv_vram_mapping *mapping);
 void etnaviv_iommu_unmap_gem(struct etnaviv_iommu *mmu,
 	struct etnaviv_vram_mapping *mapping);
-void etnaviv_iommu_destroy(struct etnaviv_iommu *iommu);
 
 int etnaviv_iommu_get_suballoc_va(struct etnaviv_gpu *gpu, dma_addr_t paddr,
 				  struct drm_mm_node *vram_node, size_t size,
@@ -73,6 +82,7 @@ size_t etnaviv_iommu_dump_size(struct etnaviv_iommu *iommu);
 void etnaviv_iommu_dump(struct etnaviv_iommu *iommu, void *buf);
 
 struct etnaviv_iommu *etnaviv_iommu_new(struct etnaviv_gpu *gpu);
+void etnaviv_iommu_destroy(struct etnaviv_iommu *iommu);
 void etnaviv_iommu_restore(struct etnaviv_gpu *gpu);
 
 #endif /* __ETNAVIV_MMU_H__ */

drivers/gpu/drm/etnaviv/etnaviv_perfmon.c

@@ -0,0 +1,495 @@
+/*
+ * Copyright (C) 2017 Etnaviv Project
+ * Copyright (C) 2017 Zodiac Inflight Innovations
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License version 2 as published by
+ * the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include "etnaviv_gpu.h"
+#include "etnaviv_perfmon.h"
+#include "state_hi.xml.h"
+
+struct etnaviv_pm_domain;
+
+struct etnaviv_pm_signal {
+	char name[64];
+	u32 data;
+
+	u32 (*sample)(struct etnaviv_gpu *gpu,
+	              const struct etnaviv_pm_domain *domain,
+	              const struct etnaviv_pm_signal *signal);
+};
+
+struct etnaviv_pm_domain {
+	char name[64];
+
+	/* profile register */
+	u32 profile_read;
+	u32 profile_config;
+
+	u8 nr_signals;
+	const struct etnaviv_pm_signal *signal;
+};
+
+struct etnaviv_pm_domain_meta {
+	const struct etnaviv_pm_domain *domains;
+	u32 nr_domains;
+};
+
+static u32 simple_reg_read(struct etnaviv_gpu *gpu,
+	const struct etnaviv_pm_domain *domain,
+	const struct etnaviv_pm_signal *signal)
+{
+	return gpu_read(gpu, signal->data);
+}
+
+static u32 perf_reg_read(struct etnaviv_gpu *gpu,
+	const struct etnaviv_pm_domain *domain,
+	const struct etnaviv_pm_signal *signal)
+{
+	gpu_write(gpu, domain->profile_config, signal->data);
+
+	return gpu_read(gpu, domain->profile_read);
+}
+
+static u32 pipe_reg_read(struct etnaviv_gpu *gpu,
+	const struct etnaviv_pm_domain *domain,
+	const struct etnaviv_pm_signal *signal)
+{
+	u32 clock = gpu_read(gpu, VIVS_HI_CLOCK_CONTROL);
+	u32 value = 0;
+	unsigned i;
+
+	for (i = 0; i < gpu->identity.pixel_pipes; i++) {
+		clock &= ~(VIVS_HI_CLOCK_CONTROL_DEBUG_PIXEL_PIPE__MASK);
+		clock |= VIVS_HI_CLOCK_CONTROL_DEBUG_PIXEL_PIPE(i);
+		gpu_write(gpu, VIVS_HI_CLOCK_CONTROL, clock);
+		gpu_write(gpu, domain->profile_config, signal->data);
+		value += gpu_read(gpu, domain->profile_read);
+	}
+
+	/* switch back to pixel pipe 0 to prevent GPU hang */
+	clock &= ~(VIVS_HI_CLOCK_CONTROL_DEBUG_PIXEL_PIPE__MASK);
+	clock |= VIVS_HI_CLOCK_CONTROL_DEBUG_PIXEL_PIPE(0);
+	gpu_write(gpu, VIVS_HI_CLOCK_CONTROL, clock);
+
+	return value;
+}
+
+static const struct etnaviv_pm_domain doms_3d[] = {
+	{
+		.name = "HI",
+		.profile_read = VIVS_MC_PROFILE_HI_READ,
+		.profile_config = VIVS_MC_PROFILE_CONFIG2,
+		.nr_signals = 5,
+		.signal = (const struct etnaviv_pm_signal[]) {
+			{
+				"TOTAL_CYCLES",
+				VIVS_HI_PROFILE_TOTAL_CYCLES,
+				&simple_reg_read
+			},
+			{
+				"IDLE_CYCLES",
+				VIVS_HI_PROFILE_IDLE_CYCLES,
+				&simple_reg_read
+			},
+			{
+				"AXI_CYCLES_READ_REQUEST_STALLED",
+				VIVS_MC_PROFILE_CONFIG2_HI_AXI_CYCLES_READ_REQUEST_STALLED,
+				&perf_reg_read
+			},
+			{
+				"AXI_CYCLES_WRITE_REQUEST_STALLED",
+				VIVS_MC_PROFILE_CONFIG2_HI_AXI_CYCLES_WRITE_REQUEST_STALLED,
+				&perf_reg_read
+			},
+			{
+				"AXI_CYCLES_WRITE_DATA_STALLED",
+				VIVS_MC_PROFILE_CONFIG2_HI_AXI_CYCLES_WRITE_DATA_STALLED,
+				&perf_reg_read
+			}
+		}
+	},
+	{
+		.name = "PE",
+		.profile_read = VIVS_MC_PROFILE_PE_READ,
+		.profile_config = VIVS_MC_PROFILE_CONFIG0,
+		.nr_signals = 5,
+		.signal = (const struct etnaviv_pm_signal[]) {
+			{
+				"PIXEL_COUNT_KILLED_BY_COLOR_PIPE",
+				VIVS_MC_PROFILE_CONFIG0_PE_PIXEL_COUNT_KILLED_BY_COLOR_PIPE,
+				&pipe_reg_read
+			},
+			{
+				"PIXEL_COUNT_KILLED_BY_DEPTH_PIPE",
+				VIVS_MC_PROFILE_CONFIG0_PE_PIXEL_COUNT_KILLED_BY_DEPTH_PIPE,
+				&pipe_reg_read
+			},
+			{
+				"PIXEL_COUNT_DRAWN_BY_COLOR_PIPE",
+				VIVS_MC_PROFILE_CONFIG0_PE_PIXEL_COUNT_DRAWN_BY_COLOR_PIPE,
+				&pipe_reg_read
+			},
+			{
+				"PIXEL_COUNT_DRAWN_BY_DEPTH_PIPE",
+				VIVS_MC_PROFILE_CONFIG0_PE_PIXEL_COUNT_DRAWN_BY_DEPTH_PIPE,
+				&pipe_reg_read
+			}
+		}
+	},
+	{
+		.name = "SH",
+		.profile_read = VIVS_MC_PROFILE_SH_READ,
+		.profile_config = VIVS_MC_PROFILE_CONFIG0,
+		.nr_signals = 9,
+		.signal = (const struct etnaviv_pm_signal[]) {
+			{
+				"SHADER_CYCLES",
+				VIVS_MC_PROFILE_CONFIG0_SH_SHADER_CYCLES,
+				&perf_reg_read
+			},
+			{
+				"PS_INST_COUNTER",
+				VIVS_MC_PROFILE_CONFIG0_SH_PS_INST_COUNTER,
+				&perf_reg_read
+			},
+			{
+				"RENDERED_PIXEL_COUNTER",
+				VIVS_MC_PROFILE_CONFIG0_SH_RENDERED_PIXEL_COUNTER,
+				&perf_reg_read
+			},
+			{
+				"VS_INST_COUNTER",
+				VIVS_MC_PROFILE_CONFIG0_SH_VS_INST_COUNTER,
+				&pipe_reg_read
+			},
+			{
+				"RENDERED_VERTICE_COUNTER",
+				VIVS_MC_PROFILE_CONFIG0_SH_RENDERED_VERTICE_COUNTER,
+				&pipe_reg_read
+			},
+			{
+				"VTX_BRANCH_INST_COUNTER",
+				VIVS_MC_PROFILE_CONFIG0_SH_VTX_BRANCH_INST_COUNTER,
+				&pipe_reg_read
+			},
+			{
+				"VTX_TEXLD_INST_COUNTER",
+				VIVS_MC_PROFILE_CONFIG0_SH_VTX_TEXLD_INST_COUNTER,
+				&pipe_reg_read
+			},
+			{
+				"PXL_BRANCH_INST_COUNTER",
+				VIVS_MC_PROFILE_CONFIG0_SH_PXL_BRANCH_INST_COUNTER,
+				&pipe_reg_read
+			},
+			{
+				"PXL_TEXLD_INST_COUNTER",
+				VIVS_MC_PROFILE_CONFIG0_SH_PXL_TEXLD_INST_COUNTER,
+				&pipe_reg_read
+			}
+		}
+	},
+	{
+		.name = "PA",
+		.profile_read = VIVS_MC_PROFILE_PA_READ,
+		.profile_config = VIVS_MC_PROFILE_CONFIG1,
+		.nr_signals = 6,
+		.signal = (const struct etnaviv_pm_signal[]) {
+			{
+				"INPUT_VTX_COUNTER",
+				VIVS_MC_PROFILE_CONFIG1_PA_INPUT_VTX_COUNTER,
+				&perf_reg_read
+			},
+			{
+				"INPUT_PRIM_COUNTER",
+				VIVS_MC_PROFILE_CONFIG1_PA_INPUT_PRIM_COUNTER,
+				&perf_reg_read
+			},
+			{
+				"OUTPUT_PRIM_COUNTER",
+				VIVS_MC_PROFILE_CONFIG1_PA_OUTPUT_PRIM_COUNTER,
+				&perf_reg_read
+			},
+			{
+				"DEPTH_CLIPPED_COUNTER",
+				VIVS_MC_PROFILE_CONFIG1_PA_DEPTH_CLIPPED_COUNTER,
+				&pipe_reg_read
+			},
+			{
+				"TRIVIAL_REJECTED_COUNTER",
+				VIVS_MC_PROFILE_CONFIG1_PA_TRIVIAL_REJECTED_COUNTER,
+				&pipe_reg_read
+			},
+			{
+				"CULLED_COUNTER",
+				VIVS_MC_PROFILE_CONFIG1_PA_CULLED_COUNTER,
+				&pipe_reg_read
+			}
+		}
+	},
+	{
+		.name = "SE",
+		.profile_read = VIVS_MC_PROFILE_SE_READ,
+		.profile_config = VIVS_MC_PROFILE_CONFIG1,
+		.nr_signals = 2,
+		.signal = (const struct etnaviv_pm_signal[]) {
+			{
+				"CULLED_TRIANGLE_COUNT",
+				VIVS_MC_PROFILE_CONFIG1_SE_CULLED_TRIANGLE_COUNT,
+				&perf_reg_read
+			},
+			{
+				"CULLED_LINES_COUNT",
+				VIVS_MC_PROFILE_CONFIG1_SE_CULLED_LINES_COUNT,
+				&perf_reg_read
+			}
+		}
+	},
+	{
+		.name = "RA",
+		.profile_read = VIVS_MC_PROFILE_RA_READ,
+		.profile_config = VIVS_MC_PROFILE_CONFIG1,
+		.nr_signals = 7,
+		.signal = (const struct etnaviv_pm_signal[]) {
+			{
+				"VALID_PIXEL_COUNT",
+				VIVS_MC_PROFILE_CONFIG1_RA_VALID_PIXEL_COUNT,
+				&perf_reg_read
+			},
+			{
+				"TOTAL_QUAD_COUNT",
+				VIVS_MC_PROFILE_CONFIG1_RA_TOTAL_QUAD_COUNT,
+				&perf_reg_read
+			},
+			{
+				"VALID_QUAD_COUNT_AFTER_EARLY_Z",
+				VIVS_MC_PROFILE_CONFIG1_RA_VALID_QUAD_COUNT_AFTER_EARLY_Z,
+				&perf_reg_read
+			},
+			{
+				"TOTAL_PRIMITIVE_COUNT",
+				VIVS_MC_PROFILE_CONFIG1_RA_TOTAL_PRIMITIVE_COUNT,
+				&perf_reg_read
+			},
+			{
+				"PIPE_CACHE_MISS_COUNTER",
+				VIVS_MC_PROFILE_CONFIG1_RA_PIPE_CACHE_MISS_COUNTER,
+				&perf_reg_read
+			},
+			{
+				"PREFETCH_CACHE_MISS_COUNTER",
+				VIVS_MC_PROFILE_CONFIG1_RA_PREFETCH_CACHE_MISS_COUNTER,
+				&perf_reg_read
+			},
+			{
+				"CULLED_QUAD_COUNT",
+				VIVS_MC_PROFILE_CONFIG1_RA_CULLED_QUAD_COUNT,
+				&perf_reg_read
+			}
+		}
+	},
+	{
+		.name = "TX",
+		.profile_read = VIVS_MC_PROFILE_TX_READ,
+		.profile_config = VIVS_MC_PROFILE_CONFIG1,
+		.nr_signals = 9,
+		.signal = (const struct etnaviv_pm_signal[]) {
+			{
+				"TOTAL_BILINEAR_REQUESTS",
+				VIVS_MC_PROFILE_CONFIG1_TX_TOTAL_BILINEAR_REQUESTS,
+				&perf_reg_read
+			},
+			{
+				"TOTAL_TRILINEAR_REQUESTS",
+				VIVS_MC_PROFILE_CONFIG1_TX_TOTAL_TRILINEAR_REQUESTS,
+				&perf_reg_read
+			},
+			{
+				"TOTAL_DISCARDED_TEXTURE_REQUESTS",
+				VIVS_MC_PROFILE_CONFIG1_TX_TOTAL_DISCARDED_TEXTURE_REQUESTS,
+				&perf_reg_read
+			},
+			{
+				"TOTAL_TEXTURE_REQUESTS",
+				VIVS_MC_PROFILE_CONFIG1_TX_TOTAL_TEXTURE_REQUESTS,
+				&perf_reg_read
+			},
+			{
+				"MEM_READ_COUNT",
+				VIVS_MC_PROFILE_CONFIG1_TX_MEM_READ_COUNT,
+				&perf_reg_read
+			},
+			{
+				"MEM_READ_IN_8B_COUNT",
+				VIVS_MC_PROFILE_CONFIG1_TX_MEM_READ_IN_8B_COUNT,
+				&perf_reg_read
+			},
+			{
+				"CACHE_MISS_COUNT",
+				VIVS_MC_PROFILE_CONFIG1_TX_CACHE_MISS_COUNT,
+				&perf_reg_read
+			},
+			{
+				"CACHE_HIT_TEXEL_COUNT",
+				VIVS_MC_PROFILE_CONFIG1_TX_CACHE_HIT_TEXEL_COUNT,
+				&perf_reg_read
+			},
+			{
+				"CACHE_MISS_TEXEL_COUNT",
+				VIVS_MC_PROFILE_CONFIG1_TX_CACHE_MISS_TEXEL_COUNT,
+				&perf_reg_read
+			}
+		}
+	},
+	{
+		.name = "MC",
+		.profile_read = VIVS_MC_PROFILE_MC_READ,
+		.profile_config = VIVS_MC_PROFILE_CONFIG2,
+		.nr_signals = 3,
+		.signal = (const struct etnaviv_pm_signal[]) {
+			{
+				"TOTAL_READ_REQ_8B_FROM_PIPELINE",
+				VIVS_MC_PROFILE_CONFIG2_MC_TOTAL_READ_REQ_8B_FROM_PIPELINE,
+				&perf_reg_read
+			},
+			{
+				"TOTAL_READ_REQ_8B_FROM_IP",
+				VIVS_MC_PROFILE_CONFIG2_MC_TOTAL_READ_REQ_8B_FROM_IP,
+				&perf_reg_read
+			},
+			{
+				"TOTAL_WRITE_REQ_8B_FROM_PIPELINE",
+				VIVS_MC_PROFILE_CONFIG2_MC_TOTAL_WRITE_REQ_8B_FROM_PIPELINE,
+				&perf_reg_read
+			}
+		}
+	}
+};
+
+static const struct etnaviv_pm_domain doms_2d[] = {
+	{
+		.name = "PE",
+		.profile_read = VIVS_MC_PROFILE_PE_READ,
+		.profile_config = VIVS_MC_PROFILE_CONFIG0,
+		.nr_signals = 1,
+		.signal = (const struct etnaviv_pm_signal[]) {
+			{
+				"PIXELS_RENDERED_2D",
+				VIVS_MC_PROFILE_CONFIG0_PE_PIXELS_RENDERED_2D,
+				&pipe_reg_read
+			}
+		}
+	}
+};
+
+static const struct etnaviv_pm_domain doms_vg[] = {
+};
+
+static const struct etnaviv_pm_domain_meta doms_meta[] = {
+	{
+		.nr_domains = ARRAY_SIZE(doms_3d),
+		.domains = &doms_3d[0]
+	},
+	{
+		.nr_domains = ARRAY_SIZE(doms_2d),
+		.domains = &doms_2d[0]
+	},
+	{
+		.nr_domains = ARRAY_SIZE(doms_vg),
+		.domains = &doms_vg[0]
+	}
+};
+
+int etnaviv_pm_query_dom(struct etnaviv_gpu *gpu,
+	struct drm_etnaviv_pm_domain *domain)
+{
+	const struct etnaviv_pm_domain_meta *meta = &doms_meta[domain->pipe];
+	const struct etnaviv_pm_domain *dom;
+
+	if (domain->iter >= meta->nr_domains)
+		return -EINVAL;
+
+	dom = meta->domains + domain->iter;
+
+	domain->id = domain->iter;
+	domain->nr_signals = dom->nr_signals;
+	strncpy(domain->name, dom->name, sizeof(domain->name));
+
+	domain->iter++;
+	if (domain->iter == meta->nr_domains)
+		domain->iter = 0xff;
+
+	return 0;
+}
+
+int etnaviv_pm_query_sig(struct etnaviv_gpu *gpu,
+	struct drm_etnaviv_pm_signal *signal)
+{
+	const struct etnaviv_pm_domain_meta *meta = &doms_meta[signal->pipe];
+	const struct etnaviv_pm_domain *dom;
+	const struct etnaviv_pm_signal *sig;
+
+	if (signal->domain >= meta->nr_domains)
+		return -EINVAL;
+
+	dom = meta->domains + signal->domain;
+
+	if (signal->iter > dom->nr_signals)
+		return -EINVAL;
+
+	sig = &dom->signal[signal->iter];
+
+	signal->id = signal->iter;
+	strncpy(signal->name, sig->name, sizeof(signal->name));
+
+	signal->iter++;
+	if (signal->iter == dom->nr_signals)
+		signal->iter = 0xffff;
+
+	return 0;
+}
+
+int etnaviv_pm_req_validate(const struct drm_etnaviv_gem_submit_pmr *r,
+	u32 exec_state)
+{
+	const struct etnaviv_pm_domain_meta *meta = &doms_meta[exec_state];
+	const struct etnaviv_pm_domain *dom;
+
+	if (r->domain >= meta->nr_domains)
+		return -EINVAL;
+
+	dom = meta->domains + r->domain;
+
+	if (r->signal > dom->nr_signals)
+		return -EINVAL;
+
+	return 0;
+}
+
+void etnaviv_perfmon_process(struct etnaviv_gpu *gpu,
+	const struct etnaviv_perfmon_request *pmr)
+{
+	const struct etnaviv_pm_domain_meta *meta = &doms_meta[gpu->exec_state];
+	const struct etnaviv_pm_domain *dom;
+	const struct etnaviv_pm_signal *sig;
+	u32 *bo = pmr->bo_vma;
+	u32 val;
+
+	dom = meta->domains + pmr->domain;
+	sig = &dom->signal[pmr->signal];
+	val = sig->sample(gpu, dom, sig);
+
+	*(bo + pmr->offset) = val;
+}

drivers/gpu/drm/etnaviv/etnaviv_perfmon.h

@@ -0,0 +1,49 @@
+/*
+ * Copyright (C) 2017 Etnaviv Project
+ * Copyright (C) 2017 Zodiac Inflight Innovations
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License version 2 as published by
+ * the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#ifndef __ETNAVIV_PERFMON_H__
+#define __ETNAVIV_PERFMON_H__
+
+struct etnaviv_gpu;
+struct drm_etnaviv_pm_domain;
+struct drm_etnaviv_pm_signal;
+
+struct etnaviv_perfmon_request
+{
+	u32 flags;
+	u8 domain;
+	u8 signal;
+	u32 sequence;
+
+	/* bo to store a value */
+	u32 *bo_vma;
+	u32 offset;
+};
+
+int etnaviv_pm_query_dom(struct etnaviv_gpu *gpu,
+	struct drm_etnaviv_pm_domain *domain);
+
+int etnaviv_pm_query_sig(struct etnaviv_gpu *gpu,
+	struct drm_etnaviv_pm_signal *signal);
+
+int etnaviv_pm_req_validate(const struct drm_etnaviv_gem_submit_pmr *r,
+	u32 exec_state);
+
+void etnaviv_perfmon_process(struct etnaviv_gpu *gpu,
+	const struct etnaviv_perfmon_request *pmr);
+
+#endif /* __ETNAVIV_PERFMON_H__ */

drivers/gpu/drm/gma500/mid_bios.c

@@ -237,7 +237,7 @@ static int mid_get_vbt_data_r10(struct drm_psb_private *dev_priv, u32 addr)
 
 	gct = kmalloc(sizeof(*gct) * vbt.panel_count, GFP_KERNEL);
 	if (!gct)
-		return -1;
+		return -ENOMEM;
 
 	gct_virtual = ioremap(addr + sizeof(vbt),
 			sizeof(*gct) * vbt.panel_count);

drivers/gpu/drm/gma500/psb_intel_sdvo.c

@@ -37,6 +37,7 @@
 #include "psb_drv.h"
 #include "psb_intel_sdvo_regs.h"
 #include "psb_intel_reg.h"
+#include <linux/kernel.h>
 
 #define SDVO_TMDS_MASK (SDVO_OUTPUT_TMDS0 | SDVO_OUTPUT_TMDS1)
 #define SDVO_RGB_MASK  (SDVO_OUTPUT_RGB0 | SDVO_OUTPUT_RGB1)
@@ -62,8 +63,6 @@ static const char *tv_format_names[] = {
 	"SECAM_60"
 };
 
-#define TV_FORMAT_NUM  (sizeof(tv_format_names) / sizeof(*tv_format_names))
-
 struct psb_intel_sdvo {
 	struct gma_encoder base;
 
@@ -148,7 +147,7 @@ struct psb_intel_sdvo_connector {
 	int force_audio;
 
 	/* This contains all current supported TV format */
-	u8 tv_format_supported[TV_FORMAT_NUM];
+	u8 tv_format_supported[ARRAY_SIZE(tv_format_names)];
 	int   format_supported_num;
 	struct drm_property *tv_format;
 
@@ -1709,7 +1708,7 @@ psb_intel_sdvo_set_property(struct drm_connector *connector,
 	}
 
 	if (property == psb_intel_sdvo_connector->tv_format) {
-		if (val >= TV_FORMAT_NUM)
+		if (val >= ARRAY_SIZE(tv_format_names))
 			return -EINVAL;
 
 		if (psb_intel_sdvo->tv_format_index ==
@@ -2269,7 +2268,7 @@ static bool psb_intel_sdvo_tv_create_property(struct psb_intel_sdvo *psb_intel_s
 		return false;
 
 	psb_intel_sdvo_connector->format_supported_num = 0;
-	for (i = 0 ; i < TV_FORMAT_NUM; i++)
+	for (i = 0 ; i < ARRAY_SIZE(tv_format_names); i++)
 		if (format_map & (1 << i))
 			psb_intel_sdvo_connector->tv_format_supported[psb_intel_sdvo_connector->format_supported_num++] = i;
 

drivers/gpu/drm/hisilicon/hibmc/hibmc_drm_vdac.c

@@ -36,7 +36,7 @@ static int hibmc_connector_mode_valid(struct drm_connector *connector,
 static struct drm_encoder *
 hibmc_connector_best_encoder(struct drm_connector *connector)
 {
-	return drm_encoder_find(connector->dev, connector->encoder_ids[0]);
+	return drm_encoder_find(connector->dev, NULL, connector->encoder_ids[0]);
 }
 
 static const struct drm_connector_helper_funcs

drivers/gpu/drm/hisilicon/kirin/kirin_drm_drv.c

@@ -237,8 +237,8 @@ static int kirin_drm_platform_probe(struct platform_device *pdev)
 	}
 
 	remote = of_graph_get_remote_node(np, 0, 0);
-	if (IS_ERR(remote))
-		return PTR_ERR(remote);
+	if (!remote)
+		return -ENODEV;
 
 	drm_of_component_match_add(dev, &match, compare_of, remote);
 	of_node_put(remote);

drivers/gpu/drm/i915/Kconfig

@@ -12,6 +12,7 @@ config DRM_I915
 	select DRM_PANEL
 	select DRM_MIPI_DSI
 	select RELAY
+	select IRQ_WORK
 	# i915 depends on ACPI_VIDEO when ACPI is enabled
 	# but for select to work, need to select ACPI_VIDEO's dependencies, ick
 	select BACKLIGHT_LCD_SUPPORT if ACPI

drivers/gpu/drm/i915/Makefile

@@ -139,7 +139,8 @@ i915-y += i915_perf.o \
 	  i915_oa_bxt.o \
 	  i915_oa_kblgt2.o \
 	  i915_oa_kblgt3.o \
-	  i915_oa_glk.o
+	  i915_oa_glk.o \
+	  i915_oa_cflgt2.o
 
 ifeq ($(CONFIG_DRM_I915_GVT),y)
 i915-y += intel_gvt.o

drivers/gpu/drm/i915/gvt/cfg_space.c

@@ -101,7 +101,7 @@ int intel_vgpu_emulate_cfg_read(struct intel_vgpu *vgpu, unsigned int offset,
 	if (WARN_ON(bytes > 4))
 		return -EINVAL;
 
-	if (WARN_ON(offset + bytes > INTEL_GVT_MAX_CFG_SPACE_SZ))
+	if (WARN_ON(offset + bytes > vgpu->gvt->device_info.cfg_space_size))
 		return -EINVAL;
 
 	memcpy(p_data, vgpu_cfg_space(vgpu) + offset, bytes);
@@ -110,13 +110,25 @@ int intel_vgpu_emulate_cfg_read(struct intel_vgpu *vgpu, unsigned int offset,
 
 static int map_aperture(struct intel_vgpu *vgpu, bool map)
 {
-	u64 first_gfn, first_mfn;
+	phys_addr_t aperture_pa = vgpu_aperture_pa_base(vgpu);
+	unsigned long aperture_sz = vgpu_aperture_sz(vgpu);
+	u64 first_gfn;
 	u64 val;
 	int ret;
 
 	if (map == vgpu->cfg_space.bar[INTEL_GVT_PCI_BAR_APERTURE].tracked)
 		return 0;
 
+	if (map) {
+		vgpu->gm.aperture_va = memremap(aperture_pa, aperture_sz,
+						MEMREMAP_WC);
+		if (!vgpu->gm.aperture_va)
+			return -ENOMEM;
+	} else {
+		memunmap(vgpu->gm.aperture_va);
+		vgpu->gm.aperture_va = NULL;
+	}
+
 	val = vgpu_cfg_space(vgpu)[PCI_BASE_ADDRESS_2];
 	if (val & PCI_BASE_ADDRESS_MEM_TYPE_64)
 		val = *(u64 *)(vgpu_cfg_space(vgpu) + PCI_BASE_ADDRESS_2);
@@ -124,14 +136,16 @@ static int map_aperture(struct intel_vgpu *vgpu, bool map)
 		val = *(u32 *)(vgpu_cfg_space(vgpu) + PCI_BASE_ADDRESS_2);
 
 	first_gfn = (val + vgpu_aperture_offset(vgpu)) >> PAGE_SHIFT;
-	first_mfn = vgpu_aperture_pa_base(vgpu) >> PAGE_SHIFT;
 
 	ret = intel_gvt_hypervisor_map_gfn_to_mfn(vgpu, first_gfn,
-						  first_mfn,
-						  vgpu_aperture_sz(vgpu) >>
-						  PAGE_SHIFT, map);
-	if (ret)
+						  aperture_pa >> PAGE_SHIFT,
+						  aperture_sz >> PAGE_SHIFT,
+						  map);
+	if (ret) {
+		memunmap(vgpu->gm.aperture_va);
+		vgpu->gm.aperture_va = NULL;
 		return ret;
+	}
 
 	vgpu->cfg_space.bar[INTEL_GVT_PCI_BAR_APERTURE].tracked = map;
 	return 0;
@@ -275,7 +289,7 @@ int intel_vgpu_emulate_cfg_write(struct intel_vgpu *vgpu, unsigned int offset,
 	if (WARN_ON(bytes > 4))
 		return -EINVAL;
 
-	if (WARN_ON(offset + bytes > INTEL_GVT_MAX_CFG_SPACE_SZ))
+	if (WARN_ON(offset + bytes > vgpu->gvt->device_info.cfg_space_size))
 		return -EINVAL;
 
 	/* First check if it's PCI_COMMAND */

drivers/gpu/drm/i915/gvt/cmd_parser.c

@@ -1576,11 +1576,11 @@ static int batch_buffer_needs_scan(struct parser_exec_state *s)
 	return 1;
 }
 
-static uint32_t find_bb_size(struct parser_exec_state *s)
+static int find_bb_size(struct parser_exec_state *s)
 {
 	unsigned long gma = 0;
 	struct cmd_info *info;
-	uint32_t bb_size = 0;
+	int bb_size = 0;
 	uint32_t cmd_len = 0;
 	bool met_bb_end = false;
 	struct intel_vgpu *vgpu = s->vgpu;
@@ -1637,6 +1637,8 @@ static int perform_bb_shadow(struct parser_exec_state *s)
 
 	/* get the size of the batch buffer */
 	bb_size = find_bb_size(s);
+	if (bb_size < 0)
+		return -EINVAL;
 
 	/* allocate shadow batch buffer */
 	entry_obj = kmalloc(sizeof(*entry_obj), GFP_KERNEL);
@@ -2603,7 +2605,8 @@ static int shadow_workload_ring_buffer(struct intel_vgpu_workload *workload)
 {
 	struct intel_vgpu *vgpu = workload->vgpu;
 	unsigned long gma_head, gma_tail, gma_top, guest_rb_size;
-	u32 *cs;
+	void *shadow_ring_buffer_va;
+	int ring_id = workload->ring_id;
 	int ret;
 
 	guest_rb_size = _RING_CTL_BUF_SIZE(workload->rb_ctl);
@@ -2616,34 +2619,42 @@ static int shadow_workload_ring_buffer(struct intel_vgpu_workload *workload)
 	gma_tail = workload->rb_start + workload->rb_tail;
 	gma_top = workload->rb_start + guest_rb_size;
 
-	/* allocate shadow ring buffer */
-	cs = intel_ring_begin(workload->req, workload->rb_len / sizeof(u32));
-	if (IS_ERR(cs))
-		return PTR_ERR(cs);
+	if (workload->rb_len > vgpu->reserve_ring_buffer_size[ring_id]) {
+		void *va = vgpu->reserve_ring_buffer_va[ring_id];
+		/* realloc the new ring buffer if needed */
+		vgpu->reserve_ring_buffer_va[ring_id] =
+			krealloc(va, workload->rb_len, GFP_KERNEL);
+		if (!vgpu->reserve_ring_buffer_va[ring_id]) {
+			gvt_vgpu_err("fail to alloc reserve ring buffer\n");
+			return -ENOMEM;
+		}
+		vgpu->reserve_ring_buffer_size[ring_id] = workload->rb_len;
+	}
+
+	shadow_ring_buffer_va = vgpu->reserve_ring_buffer_va[ring_id];
 
 	/* get shadow ring buffer va */
-	workload->shadow_ring_buffer_va = cs;
+	workload->shadow_ring_buffer_va = shadow_ring_buffer_va;
 
 	/* head > tail --> copy head <-> top */
 	if (gma_head > gma_tail) {
 		ret = copy_gma_to_hva(vgpu, vgpu->gtt.ggtt_mm,
-				      gma_head, gma_top, cs);
+				      gma_head, gma_top, shadow_ring_buffer_va);
 		if (ret < 0) {
 			gvt_vgpu_err("fail to copy guest ring buffer\n");
 			return ret;
 		}
-		cs += ret / sizeof(u32);
+		shadow_ring_buffer_va += ret;
 		gma_head = workload->rb_start;
 	}
 
 	/* copy head or start <-> tail */
-	ret = copy_gma_to_hva(vgpu, vgpu->gtt.ggtt_mm, gma_head, gma_tail, cs);
+	ret = copy_gma_to_hva(vgpu, vgpu->gtt.ggtt_mm, gma_head, gma_tail,
+				shadow_ring_buffer_va);
 	if (ret < 0) {
 		gvt_vgpu_err("fail to copy guest ring buffer\n");
 		return ret;
 	}
-	cs += ret / sizeof(u32);
-	intel_ring_advance(workload->req, cs);
 	return 0;
 }
 

drivers/gpu/drm/i915/gvt/execlist.c

@@ -368,7 +368,7 @@ static void free_workload(struct intel_vgpu_workload *workload)
 #define get_desc_from_elsp_dwords(ed, i) \
 	((struct execlist_ctx_descriptor_format *)&((ed)->data[i * 2]))
 
-static void prepare_shadow_batch_buffer(struct intel_vgpu_workload *workload)
+static int prepare_shadow_batch_buffer(struct intel_vgpu_workload *workload)
 {
 	const int gmadr_bytes = workload->vgpu->gvt->device_info.gmadr_bytes_in_cmd;
 	struct intel_shadow_bb_entry *entry_obj;
@@ -379,7 +379,7 @@ static void prepare_shadow_batch_buffer(struct intel_vgpu_workload *workload)
 
 		vma = i915_gem_object_ggtt_pin(entry_obj->obj, NULL, 0, 4, 0);
 		if (IS_ERR(vma)) {
-			return;
+			return PTR_ERR(vma);
 		}
 
 		/* FIXME: we are not tracking our pinned VMA leaving it
@@ -392,6 +392,7 @@ static void prepare_shadow_batch_buffer(struct intel_vgpu_workload *workload)
 		if (gmadr_bytes == 8)
 			entry_obj->bb_start_cmd_va[2] = 0;
 	}
+	return 0;
 }
 
 static int update_wa_ctx_2_shadow_ctx(struct intel_shadow_wa_ctx *wa_ctx)
@@ -420,7 +421,7 @@ static int update_wa_ctx_2_shadow_ctx(struct intel_shadow_wa_ctx *wa_ctx)
 	return 0;
 }
 
-static void prepare_shadow_wa_ctx(struct intel_shadow_wa_ctx *wa_ctx)
+static int prepare_shadow_wa_ctx(struct intel_shadow_wa_ctx *wa_ctx)
 {
 	struct i915_vma *vma;
 	unsigned char *per_ctx_va =
@@ -428,12 +429,12 @@ static void prepare_shadow_wa_ctx(struct intel_shadow_wa_ctx *wa_ctx)
 		wa_ctx->indirect_ctx.size;
 
 	if (wa_ctx->indirect_ctx.size == 0)
-		return;
+		return 0;
 
 	vma = i915_gem_object_ggtt_pin(wa_ctx->indirect_ctx.obj, NULL,
 				       0, CACHELINE_BYTES, 0);
 	if (IS_ERR(vma)) {
-		return;
+		return PTR_ERR(vma);
 	}
 
 	/* FIXME: we are not tracking our pinned VMA leaving it
@@ -447,26 +448,7 @@ static void prepare_shadow_wa_ctx(struct intel_shadow_wa_ctx *wa_ctx)
 	memset(per_ctx_va, 0, CACHELINE_BYTES);
 
 	update_wa_ctx_2_shadow_ctx(wa_ctx);
-}
-
-static int prepare_execlist_workload(struct intel_vgpu_workload *workload)
-{
-	struct intel_vgpu *vgpu = workload->vgpu;
-	struct execlist_ctx_descriptor_format ctx[2];
-	int ring_id = workload->ring_id;
-
-	intel_vgpu_pin_mm(workload->shadow_mm);
-	intel_vgpu_sync_oos_pages(workload->vgpu);
-	intel_vgpu_flush_post_shadow(workload->vgpu);
-	prepare_shadow_batch_buffer(workload);
-	prepare_shadow_wa_ctx(&workload->wa_ctx);
-	if (!workload->emulate_schedule_in)
-		return 0;
-
-	ctx[0] = *get_desc_from_elsp_dwords(&workload->elsp_dwords, 1);
-	ctx[1] = *get_desc_from_elsp_dwords(&workload->elsp_dwords, 0);
-
-	return emulate_execlist_schedule_in(&vgpu->execlist[ring_id], ctx);
+	return 0;
 }
 
 static void release_shadow_batch_buffer(struct intel_vgpu_workload *workload)
@@ -489,13 +471,62 @@ static void release_shadow_batch_buffer(struct intel_vgpu_workload *workload)
 	}
 }
 
-static void release_shadow_wa_ctx(struct intel_shadow_wa_ctx *wa_ctx)
+static int prepare_execlist_workload(struct intel_vgpu_workload *workload)
 {
-	if (!wa_ctx->indirect_ctx.obj)
-		return;
+	struct intel_vgpu *vgpu = workload->vgpu;
+	struct execlist_ctx_descriptor_format ctx[2];
+	int ring_id = workload->ring_id;
+	int ret;
+
+	ret = intel_vgpu_pin_mm(workload->shadow_mm);
+	if (ret) {
+		gvt_vgpu_err("fail to vgpu pin mm\n");
+		goto out;
+	}
+
+	ret = intel_vgpu_sync_oos_pages(workload->vgpu);
+	if (ret) {
+		gvt_vgpu_err("fail to vgpu sync oos pages\n");
+		goto err_unpin_mm;
+	}
 
-	i915_gem_object_unpin_map(wa_ctx->indirect_ctx.obj);
-	i915_gem_object_put(wa_ctx->indirect_ctx.obj);
+	ret = intel_vgpu_flush_post_shadow(workload->vgpu);
+	if (ret) {
+		gvt_vgpu_err("fail to flush post shadow\n");
+		goto err_unpin_mm;
+	}
+
+	ret = prepare_shadow_batch_buffer(workload);
+	if (ret) {
+		gvt_vgpu_err("fail to prepare_shadow_batch_buffer\n");
+		goto err_unpin_mm;
+	}
+
+	ret = prepare_shadow_wa_ctx(&workload->wa_ctx);
+	if (ret) {
+		gvt_vgpu_err("fail to prepare_shadow_wa_ctx\n");
+		goto err_shadow_batch;
+	}
+
+	if (!workload->emulate_schedule_in)
+		return 0;
+
+	ctx[0] = *get_desc_from_elsp_dwords(&workload->elsp_dwords, 1);
+	ctx[1] = *get_desc_from_elsp_dwords(&workload->elsp_dwords, 0);
+
+	ret = emulate_execlist_schedule_in(&vgpu->execlist[ring_id], ctx);
+	if (!ret)
+		goto out;
+	else
+		gvt_vgpu_err("fail to emulate execlist schedule in\n");
+
+	release_shadow_wa_ctx(&workload->wa_ctx);
+err_shadow_batch:
+	release_shadow_batch_buffer(workload);
+err_unpin_mm:
+	intel_vgpu_unpin_mm(workload->shadow_mm);
+out:
+	return ret;
 }
 
 static int complete_execlist_workload(struct intel_vgpu_workload *workload)
@@ -511,8 +542,10 @@ static int complete_execlist_workload(struct intel_vgpu_workload *workload)
 	gvt_dbg_el("complete workload %p status %d\n", workload,
 			workload->status);
 
-	release_shadow_batch_buffer(workload);
-	release_shadow_wa_ctx(&workload->wa_ctx);
+	if (!workload->status) {
+		release_shadow_batch_buffer(workload);
+		release_shadow_wa_ctx(&workload->wa_ctx);
+	}
 
 	if (workload->status || (vgpu->resetting_eng & ENGINE_MASK(ring_id))) {
 		/* if workload->status is not successful means HW GPU
@@ -820,10 +853,21 @@ static void clean_workloads(struct intel_vgpu *vgpu, unsigned long engine_mask)
 
 void intel_vgpu_clean_execlist(struct intel_vgpu *vgpu)
 {
+	enum intel_engine_id i;
+	struct intel_engine_cs *engine;
+
 	clean_workloads(vgpu, ALL_ENGINES);
 	kmem_cache_destroy(vgpu->workloads);
+
+	for_each_engine(engine, vgpu->gvt->dev_priv, i) {
+		kfree(vgpu->reserve_ring_buffer_va[i]);
+		vgpu->reserve_ring_buffer_va[i] = NULL;
+		vgpu->reserve_ring_buffer_size[i] = 0;
+	}
+
 }
 
+#define RESERVE_RING_BUFFER_SIZE		((1 * PAGE_SIZE)/8)
 int intel_vgpu_init_execlist(struct intel_vgpu *vgpu)
 {
 	enum intel_engine_id i;
@@ -843,7 +887,26 @@ int intel_vgpu_init_execlist(struct intel_vgpu *vgpu)
 	if (!vgpu->workloads)
 		return -ENOMEM;
 
+	/* each ring has a shadow ring buffer until vgpu destroyed */
+	for_each_engine(engine, vgpu->gvt->dev_priv, i) {
+		vgpu->reserve_ring_buffer_va[i] =
+			kmalloc(RESERVE_RING_BUFFER_SIZE, GFP_KERNEL);
+		if (!vgpu->reserve_ring_buffer_va[i]) {
+			gvt_vgpu_err("fail to alloc reserve ring buffer\n");
+			goto out;
+		}
+		vgpu->reserve_ring_buffer_size[i] = RESERVE_RING_BUFFER_SIZE;
+	}
 	return 0;
+out:
+	for_each_engine(engine, vgpu->gvt->dev_priv, i) {
+		if (vgpu->reserve_ring_buffer_size[i]) {
+			kfree(vgpu->reserve_ring_buffer_va[i]);
+			vgpu->reserve_ring_buffer_va[i] = NULL;
+			vgpu->reserve_ring_buffer_size[i] = 0;
+		}
+	}
+	return -ENOMEM;
 }
 
 void intel_vgpu_reset_execlist(struct intel_vgpu *vgpu,

drivers/gpu/drm/i915/gvt/gtt.c

@@ -1647,14 +1647,13 @@ int intel_vgpu_pin_mm(struct intel_vgpu_mm *mm)
 	if (WARN_ON(mm->type != INTEL_GVT_MM_PPGTT))
 		return 0;
 
-	atomic_inc(&mm->pincount);
-
 	if (!mm->shadowed) {
 		ret = shadow_mm(mm);
 		if (ret)
 			return ret;
 	}
 
+	atomic_inc(&mm->pincount);
 	list_del_init(&mm->lru_list);
 	list_add_tail(&mm->lru_list, &mm->vgpu->gvt->gtt.mm_lru_list_head);
 	return 0;
@@ -1972,7 +1971,7 @@ static int alloc_scratch_pages(struct intel_vgpu *vgpu,
 		 */
 		se.val64 |= _PAGE_PRESENT | _PAGE_RW;
 		if (type == GTT_TYPE_PPGTT_PDE_PT)
-			se.val64 |= PPAT_CACHED_INDEX;
+			se.val64 |= PPAT_CACHED;
 
 		for (i = 0; i < page_entry_num; i++)
 			ops->set_entry(scratch_pt, &se, i, false, 0, vgpu);

drivers/gpu/drm/i915/gvt/gvt.c

@@ -111,7 +111,7 @@ static void init_device_info(struct intel_gvt *gvt)
 	if (IS_BROADWELL(gvt->dev_priv) || IS_SKYLAKE(gvt->dev_priv)
 		|| IS_KABYLAKE(gvt->dev_priv)) {
 		info->max_support_vgpus = 8;
-		info->cfg_space_size = 256;
+		info->cfg_space_size = PCI_CFG_SPACE_EXP_SIZE;
 		info->mmio_size = 2 * 1024 * 1024;
 		info->mmio_bar = 0;
 		info->gtt_start_offset = 8 * 1024 * 1024;

drivers/gpu/drm/i915/gvt/gvt.h

@@ -80,6 +80,7 @@ struct intel_gvt_device_info {
 struct intel_vgpu_gm {
 	u64 aperture_sz;
 	u64 hidden_sz;
+	void *aperture_va;
 	struct drm_mm_node low_gm_node;
 	struct drm_mm_node high_gm_node;
 };
@@ -99,7 +100,6 @@ struct intel_vgpu_mmio {
 	bool disable_warn_untrack;
 };
 
-#define INTEL_GVT_MAX_CFG_SPACE_SZ 256
 #define INTEL_GVT_MAX_BAR_NUM 4
 
 struct intel_vgpu_pci_bar {
@@ -108,7 +108,7 @@ struct intel_vgpu_pci_bar {
 };
 
 struct intel_vgpu_cfg_space {
-	unsigned char virtual_cfg_space[INTEL_GVT_MAX_CFG_SPACE_SZ];
+	unsigned char virtual_cfg_space[PCI_CFG_SPACE_EXP_SIZE];
 	struct intel_vgpu_pci_bar bar[INTEL_GVT_MAX_BAR_NUM];
 };
 
@@ -165,6 +165,9 @@ struct intel_vgpu {
 	struct list_head workload_q_head[I915_NUM_ENGINES];
 	struct kmem_cache *workloads;
 	atomic_t running_workload_num;
+	/* 1/2K for each reserve ring buffer */
+	void *reserve_ring_buffer_va[I915_NUM_ENGINES];
+	int reserve_ring_buffer_size[I915_NUM_ENGINES];
 	DECLARE_BITMAP(tlb_handle_pending, I915_NUM_ENGINES);
 	struct i915_gem_context *shadow_ctx;
 	DECLARE_BITMAP(shadow_ctx_desc_updated, I915_NUM_ENGINES);
@@ -474,6 +477,13 @@ int intel_vgpu_emulate_cfg_read(struct intel_vgpu *vgpu, unsigned int offset,
 int intel_vgpu_emulate_cfg_write(struct intel_vgpu *vgpu, unsigned int offset,
 		void *p_data, unsigned int bytes);
 
+static inline u64 intel_vgpu_get_bar_gpa(struct intel_vgpu *vgpu, int bar)
+{
+	/* We are 64bit bar. */
+	return (*(u64 *)(vgpu->cfg_space.virtual_cfg_space + bar)) &
+			PCI_BASE_ADDRESS_MEM_MASK;
+}
+
 void intel_gvt_clean_opregion(struct intel_gvt *gvt);
 int intel_gvt_init_opregion(struct intel_gvt *gvt);
 

drivers/gpu/drm/i915/gvt/kvmgt.c

@@ -609,21 +609,20 @@ static void intel_vgpu_release_work(struct work_struct *work)
 	__intel_vgpu_release(vgpu);
 }
 
-static uint64_t intel_vgpu_get_bar0_addr(struct intel_vgpu *vgpu)
+static uint64_t intel_vgpu_get_bar_addr(struct intel_vgpu *vgpu, int bar)
 {
 	u32 start_lo, start_hi;
 	u32 mem_type;
-	int pos = PCI_BASE_ADDRESS_0;
 
-	start_lo = (*(u32 *)(vgpu->cfg_space.virtual_cfg_space + pos)) &
+	start_lo = (*(u32 *)(vgpu->cfg_space.virtual_cfg_space + bar)) &
 			PCI_BASE_ADDRESS_MEM_MASK;
-	mem_type = (*(u32 *)(vgpu->cfg_space.virtual_cfg_space + pos)) &
+	mem_type = (*(u32 *)(vgpu->cfg_space.virtual_cfg_space + bar)) &
 			PCI_BASE_ADDRESS_MEM_TYPE_MASK;
 
 	switch (mem_type) {
 	case PCI_BASE_ADDRESS_MEM_TYPE_64:
 		start_hi = (*(u32 *)(vgpu->cfg_space.virtual_cfg_space
-						+ pos + 4));
+						+ bar + 4));
 		break;
 	case PCI_BASE_ADDRESS_MEM_TYPE_32:
 	case PCI_BASE_ADDRESS_MEM_TYPE_1M:
@@ -637,6 +636,21 @@ static uint64_t intel_vgpu_get_bar0_addr(struct intel_vgpu *vgpu)
 	return ((u64)start_hi << 32) | start_lo;
 }
 
+static int intel_vgpu_bar_rw(struct intel_vgpu *vgpu, int bar, uint64_t off,
+			     void *buf, unsigned int count, bool is_write)
+{
+	uint64_t bar_start = intel_vgpu_get_bar_addr(vgpu, bar);
+	int ret;
+
+	if (is_write)
+		ret = intel_gvt_ops->emulate_mmio_write(vgpu,
+					bar_start + off, buf, count);
+	else
+		ret = intel_gvt_ops->emulate_mmio_read(vgpu,
+					bar_start + off, buf, count);
+	return ret;
+}
+
 static ssize_t intel_vgpu_rw(struct mdev_device *mdev, char *buf,
 			size_t count, loff_t *ppos, bool is_write)
 {
@@ -661,20 +675,14 @@ static ssize_t intel_vgpu_rw(struct mdev_device *mdev, char *buf,
 						buf, count);
 		break;
 	case VFIO_PCI_BAR0_REGION_INDEX:
-	case VFIO_PCI_BAR1_REGION_INDEX:
-		if (is_write) {
-			uint64_t bar0_start = intel_vgpu_get_bar0_addr(vgpu);
-
-			ret = intel_gvt_ops->emulate_mmio_write(vgpu,
-						bar0_start + pos, buf, count);
-		} else {
-			uint64_t bar0_start = intel_vgpu_get_bar0_addr(vgpu);
-
-			ret = intel_gvt_ops->emulate_mmio_read(vgpu,
-						bar0_start + pos, buf, count);
-		}
+		ret = intel_vgpu_bar_rw(vgpu, PCI_BASE_ADDRESS_0, pos,
+					buf, count, is_write);
 		break;
 	case VFIO_PCI_BAR2_REGION_INDEX:
+		ret = intel_vgpu_bar_rw(vgpu, PCI_BASE_ADDRESS_2, pos,
+					buf, count, is_write);
+		break;
+	case VFIO_PCI_BAR1_REGION_INDEX:
 	case VFIO_PCI_BAR3_REGION_INDEX:
 	case VFIO_PCI_BAR4_REGION_INDEX:
 	case VFIO_PCI_BAR5_REGION_INDEX:
@@ -970,7 +978,7 @@ static long intel_vgpu_ioctl(struct mdev_device *mdev, unsigned int cmd,
 		switch (info.index) {
 		case VFIO_PCI_CONFIG_REGION_INDEX:
 			info.offset = VFIO_PCI_INDEX_TO_OFFSET(info.index);
-			info.size = INTEL_GVT_MAX_CFG_SPACE_SZ;
+			info.size = vgpu->gvt->device_info.cfg_space_size;
 			info.flags = VFIO_REGION_INFO_FLAG_READ |
 				     VFIO_REGION_INFO_FLAG_WRITE;
 			break;

drivers/gpu/drm/i915/gvt/mmio.c

@@ -45,8 +45,7 @@
  */
 int intel_vgpu_gpa_to_mmio_offset(struct intel_vgpu *vgpu, u64 gpa)
 {
-	u64 gttmmio_gpa = *(u64 *)(vgpu_cfg_space(vgpu) + PCI_BASE_ADDRESS_0) &
-			  ~GENMASK(3, 0);
+	u64 gttmmio_gpa = intel_vgpu_get_bar_gpa(vgpu, PCI_BASE_ADDRESS_0);
 	return gpa - gttmmio_gpa;
 }
 
@@ -57,6 +56,38 @@ int intel_vgpu_gpa_to_mmio_offset(struct intel_vgpu *vgpu, u64 gpa)
 	(reg >= gvt->device_info.gtt_start_offset \
 	 && reg < gvt->device_info.gtt_start_offset + gvt_ggtt_sz(gvt))
 
+static bool vgpu_gpa_is_aperture(struct intel_vgpu *vgpu, uint64_t gpa)
+{
+	u64 aperture_gpa = intel_vgpu_get_bar_gpa(vgpu, PCI_BASE_ADDRESS_2);
+	u64 aperture_sz = vgpu_aperture_sz(vgpu);
+
+	return gpa >= aperture_gpa && gpa < aperture_gpa + aperture_sz;
+}
+
+static int vgpu_aperture_rw(struct intel_vgpu *vgpu, uint64_t gpa,
+			    void *pdata, unsigned int size, bool is_read)
+{
+	u64 aperture_gpa = intel_vgpu_get_bar_gpa(vgpu, PCI_BASE_ADDRESS_2);
+	u64 offset = gpa - aperture_gpa;
+
+	if (!vgpu_gpa_is_aperture(vgpu, gpa + size - 1)) {
+		gvt_vgpu_err("Aperture rw out of range, offset %llx, size %d\n",
+			     offset, size);
+		return -EINVAL;
+	}
+
+	if (!vgpu->gm.aperture_va) {
+		gvt_vgpu_err("BAR is not enabled\n");
+		return -ENXIO;
+	}
+
+	if (is_read)
+		memcpy(pdata, vgpu->gm.aperture_va + offset, size);
+	else
+		memcpy(vgpu->gm.aperture_va + offset, pdata, size);
+	return 0;
+}
+
 static void failsafe_emulate_mmio_rw(struct intel_vgpu *vgpu, uint64_t pa,
 		void *p_data, unsigned int bytes, bool read)
 {
@@ -133,6 +164,12 @@ int intel_vgpu_emulate_mmio_read(struct intel_vgpu *vgpu, uint64_t pa,
 	}
 	mutex_lock(&gvt->lock);
 
+	if (vgpu_gpa_is_aperture(vgpu, pa)) {
+		ret = vgpu_aperture_rw(vgpu, pa, p_data, bytes, true);
+		mutex_unlock(&gvt->lock);
+		return ret;
+	}
+
 	if (atomic_read(&vgpu->gtt.n_write_protected_guest_page)) {
 		struct intel_vgpu_guest_page *gp;
 
@@ -224,6 +261,12 @@ int intel_vgpu_emulate_mmio_write(struct intel_vgpu *vgpu, uint64_t pa,
 
 	mutex_lock(&gvt->lock);
 
+	if (vgpu_gpa_is_aperture(vgpu, pa)) {
+		ret = vgpu_aperture_rw(vgpu, pa, p_data, bytes, false);
+		mutex_unlock(&gvt->lock);
+		return ret;
+	}
+
 	if (atomic_read(&vgpu->gtt.n_write_protected_guest_page)) {
 		struct intel_vgpu_guest_page *gp;
 

drivers/gpu/drm/i915/gvt/render.c

@@ -293,7 +293,7 @@ static void switch_mmio_to_vgpu(struct intel_vgpu *vgpu, int ring_id)
 		 */
 		if (mmio->in_context &&
 				((ctx_ctrl & inhibit_mask) != inhibit_mask) &&
-				i915.enable_execlists)
+				i915_modparams.enable_execlists)
 			continue;
 
 		if (mmio->mask)

drivers/gpu/drm/i915/gvt/scheduler.c

@@ -87,7 +87,7 @@ static int populate_shadow_context(struct intel_vgpu_workload *workload)
 			return -EINVAL;
 		}
 
-		page = i915_gem_object_get_page(ctx_obj, LRC_PPHWSP_PN + i);
+		page = i915_gem_object_get_page(ctx_obj, LRC_HEADER_PAGES + i);
 		dst = kmap(page);
 		intel_gvt_hypervisor_read_gpa(vgpu, context_gpa, dst,
 				GTT_PAGE_SIZE);
@@ -201,6 +201,43 @@ static void shadow_context_descriptor_update(struct i915_gem_context *ctx,
 	ce->lrc_desc = desc;
 }
 
+static int copy_workload_to_ring_buffer(struct intel_vgpu_workload *workload)
+{
+	struct intel_vgpu *vgpu = workload->vgpu;
+	void *shadow_ring_buffer_va;
+	u32 *cs;
+
+	/* allocate shadow ring buffer */
+	cs = intel_ring_begin(workload->req, workload->rb_len / sizeof(u32));
+	if (IS_ERR(cs)) {
+		gvt_vgpu_err("fail to alloc size =%ld shadow  ring buffer\n",
+			workload->rb_len);
+		return PTR_ERR(cs);
+	}
+
+	shadow_ring_buffer_va = workload->shadow_ring_buffer_va;
+
+	/* get shadow ring buffer va */
+	workload->shadow_ring_buffer_va = cs;
+
+	memcpy(cs, shadow_ring_buffer_va,
+			workload->rb_len);
+
+	cs += workload->rb_len / sizeof(u32);
+	intel_ring_advance(workload->req, cs);
+
+	return 0;
+}
+
+void release_shadow_wa_ctx(struct intel_shadow_wa_ctx *wa_ctx)
+{
+	if (!wa_ctx->indirect_ctx.obj)
+		return;
+
+	i915_gem_object_unpin_map(wa_ctx->indirect_ctx.obj);
+	i915_gem_object_put(wa_ctx->indirect_ctx.obj);
+}
+
 /**
  * intel_gvt_scan_and_shadow_workload - audit the workload by scanning and
  * shadow it as well, include ringbuffer,wa_ctx and ctx.
@@ -214,8 +251,10 @@ int intel_gvt_scan_and_shadow_workload(struct intel_vgpu_workload *workload)
 	int ring_id = workload->ring_id;
 	struct i915_gem_context *shadow_ctx = workload->vgpu->shadow_ctx;
 	struct drm_i915_private *dev_priv = workload->vgpu->gvt->dev_priv;
+	struct intel_engine_cs *engine = dev_priv->engine[ring_id];
 	struct drm_i915_gem_request *rq;
 	struct intel_vgpu *vgpu = workload->vgpu;
+	struct intel_ring *ring;
 	int ret;
 
 	lockdep_assert_held(&dev_priv->drm.struct_mutex);
@@ -231,35 +270,56 @@ int intel_gvt_scan_and_shadow_workload(struct intel_vgpu_workload *workload)
 		shadow_context_descriptor_update(shadow_ctx,
 					dev_priv->engine[ring_id]);
 
-	rq = i915_gem_request_alloc(dev_priv->engine[ring_id], shadow_ctx);
-	if (IS_ERR(rq)) {
-		gvt_vgpu_err("fail to allocate gem request\n");
-		ret = PTR_ERR(rq);
-		goto out;
-	}
-
-	gvt_dbg_sched("ring id %d get i915 gem request %p\n", ring_id, rq);
-
-	workload->req = i915_gem_request_get(rq);
-
 	ret = intel_gvt_scan_and_shadow_ringbuffer(workload);
 	if (ret)
-		goto out;
+		goto err_scan;
 
 	if ((workload->ring_id == RCS) &&
 	    (workload->wa_ctx.indirect_ctx.size != 0)) {
 		ret = intel_gvt_scan_and_shadow_wa_ctx(&workload->wa_ctx);
 		if (ret)
-			goto out;
+			goto err_scan;
+	}
+
+	/* pin shadow context by gvt even the shadow context will be pinned
+	 * when i915 alloc request. That is because gvt will update the guest
+	 * context from shadow context when workload is completed, and at that
+	 * moment, i915 may already unpined the shadow context to make the
+	 * shadow_ctx pages invalid. So gvt need to pin itself. After update
+	 * the guest context, gvt can unpin the shadow_ctx safely.
+	 */
+	ring = engine->context_pin(engine, shadow_ctx);
+	if (IS_ERR(ring)) {
+		ret = PTR_ERR(ring);
+		gvt_vgpu_err("fail to pin shadow context\n");
+		goto err_shadow;
 	}
 
 	ret = populate_shadow_context(workload);
 	if (ret)
-		goto out;
+		goto err_unpin;
 
+	rq = i915_gem_request_alloc(dev_priv->engine[ring_id], shadow_ctx);
+	if (IS_ERR(rq)) {
+		gvt_vgpu_err("fail to allocate gem request\n");
+		ret = PTR_ERR(rq);
+		goto err_unpin;
+	}
+
+	gvt_dbg_sched("ring id %d get i915 gem request %p\n", ring_id, rq);
+
+	workload->req = i915_gem_request_get(rq);
+	ret = copy_workload_to_ring_buffer(workload);
+	if (ret)
+		goto err_unpin;
 	workload->shadowed = true;
+	return 0;
 
-out:
+err_unpin:
+	engine->context_unpin(engine, shadow_ctx);
+err_shadow:
+	release_shadow_wa_ctx(&workload->wa_ctx);
+err_scan:
 	return ret;
 }
 
@@ -269,8 +329,6 @@ static int dispatch_workload(struct intel_vgpu_workload *workload)
 	struct i915_gem_context *shadow_ctx = workload->vgpu->shadow_ctx;
 	struct drm_i915_private *dev_priv = workload->vgpu->gvt->dev_priv;
 	struct intel_engine_cs *engine = dev_priv->engine[ring_id];
-	struct intel_vgpu *vgpu = workload->vgpu;
-	struct intel_ring *ring;
 	int ret = 0;
 
 	gvt_dbg_sched("ring id %d prepare to dispatch workload %p\n",
@@ -284,22 +342,10 @@ static int dispatch_workload(struct intel_vgpu_workload *workload)
 
 	if (workload->prepare) {
 		ret = workload->prepare(workload);
-		if (ret)
+		if (ret) {
+			engine->context_unpin(engine, shadow_ctx);
 			goto out;
-	}
-
-	/* pin shadow context by gvt even the shadow context will be pinned
-	 * when i915 alloc request. That is because gvt will update the guest
-	 * context from shadow context when workload is completed, and at that
-	 * moment, i915 may already unpined the shadow context to make the
-	 * shadow_ctx pages invalid. So gvt need to pin itself. After update
-	 * the guest context, gvt can unpin the shadow_ctx safely.
-	 */
-	ring = engine->context_pin(engine, shadow_ctx);
-	if (IS_ERR(ring)) {
-		ret = PTR_ERR(ring);
-		gvt_vgpu_err("fail to pin shadow context\n");
-		goto out;
+		}
 	}
 
 out:
@@ -408,7 +454,7 @@ static void update_guest_context(struct intel_vgpu_workload *workload)
 			return;
 		}
 
-		page = i915_gem_object_get_page(ctx_obj, LRC_PPHWSP_PN + i);
+		page = i915_gem_object_get_page(ctx_obj, LRC_HEADER_PAGES + i);
 		src = kmap(page);
 		intel_gvt_hypervisor_write_gpa(vgpu, context_gpa, src,
 				GTT_PAGE_SIZE);

drivers/gpu/drm/i915/gvt/scheduler.h

@@ -140,4 +140,5 @@ int intel_vgpu_init_gvt_context(struct intel_vgpu *vgpu);
 
 void intel_vgpu_clean_gvt_context(struct intel_vgpu *vgpu);
 
+void release_shadow_wa_ctx(struct intel_shadow_wa_ctx *wa_ctx);
 #endif

drivers/gpu/drm/i915/i915_debugfs.c

@@ -67,7 +67,7 @@ static int i915_capabilities(struct seq_file *m, void *data)
 #undef PRINT_FLAG
 
 	kernel_param_lock(THIS_MODULE);
-#define PRINT_PARAM(T, x) seq_print_param(m, #x, #T, &i915.x);
+#define PRINT_PARAM(T, x, ...) seq_print_param(m, #x, #T, &i915_modparams.x);
 	I915_PARAMS_FOR_EACH(PRINT_PARAM);
 #undef PRINT_PARAM
 	kernel_param_unlock(THIS_MODULE);
@@ -1267,7 +1267,7 @@ static int i915_hangcheck_info(struct seq_file *m, void *unused)
 	if (waitqueue_active(&dev_priv->gpu_error.reset_queue))
 		seq_puts(m, "struct_mutex blocked for reset\n");
 
-	if (!i915.enable_hangcheck) {
+	if (!i915_modparams.enable_hangcheck) {
 		seq_puts(m, "Hangcheck disabled\n");
 		return 0;
 	}
@@ -1422,6 +1422,9 @@ static int i915_forcewake_domains(struct seq_file *m, void *data)
 	struct intel_uncore_forcewake_domain *fw_domain;
 	unsigned int tmp;
 
+	seq_printf(m, "user.bypass_count = %u\n",
+		   i915->uncore.user_forcewake.count);
+
 	for_each_fw_domain(fw_domain, i915, tmp)
 		seq_printf(m, "%s.wake_count = %u\n",
 			   intel_uncore_forcewake_domain_to_str(fw_domain->id),
@@ -1699,7 +1702,7 @@ static int i915_ips_status(struct seq_file *m, void *unused)
 	intel_runtime_pm_get(dev_priv);
 
 	seq_printf(m, "Enabled by kernel parameter: %s\n",
-		   yesno(i915.enable_ips));
+		   yesno(i915_modparams.enable_ips));
 
 	if (INTEL_GEN(dev_priv) >= 8) {
 		seq_puts(m, "Currently: unknown\n");
@@ -2014,7 +2017,7 @@ static int i915_dump_lrc(struct seq_file *m, void *unused)
 	enum intel_engine_id id;
 	int ret;
 
-	if (!i915.enable_execlists) {
+	if (!i915_modparams.enable_execlists) {
 		seq_printf(m, "Logical Ring Contexts are disabled\n");
 		return 0;
 	}
@@ -2443,12 +2446,8 @@ static void i915_guc_client_info(struct seq_file *m,
 
 	seq_printf(m, "\tPriority %d, GuC stage index: %u, PD offset 0x%x\n",
 		client->priority, client->stage_id, client->proc_desc_offset);
-	seq_printf(m, "\tDoorbell id %d, offset: 0x%lx, cookie 0x%x\n",
-		client->doorbell_id, client->doorbell_offset, client->doorbell_cookie);
-	seq_printf(m, "\tWQ size %d, offset: 0x%x, tail %d\n",
-		client->wq_size, client->wq_offset, client->wq_tail);
-
-	seq_printf(m, "\tWork queue full: %u\n", client->no_wq_space);
+	seq_printf(m, "\tDoorbell id %d, offset: 0x%lx\n",
+		client->doorbell_id, client->doorbell_offset);
 
 	for_each_engine(engine, dev_priv, id) {
 		u64 submissions = client->submissions[id];
@@ -2594,7 +2593,7 @@ static int i915_guc_log_control_get(void *data, u64 *val)
 	if (!dev_priv->guc.log.vma)
 		return -EINVAL;
 
-	*val = i915.guc_log_level;
+	*val = i915_modparams.guc_log_level;
 
 	return 0;
 }
@@ -3312,7 +3311,9 @@ static int i915_engine_info(struct seq_file *m, void *unused)
 		seq_printf(m, "\tBBADDR: 0x%08x_%08x\n",
 			   upper_32_bits(addr), lower_32_bits(addr));
 
-		if (i915.enable_execlists) {
+		if (i915_modparams.enable_execlists) {
+			const u32 *hws = &engine->status_page.page_addr[I915_HWS_CSB_BUF0_INDEX];
+			struct intel_engine_execlists * const execlists = &engine->execlists;
 			u32 ptr, read, write;
 			unsigned int idx;
 
@@ -3323,8 +3324,10 @@ static int i915_engine_info(struct seq_file *m, void *unused)
 			ptr = I915_READ(RING_CONTEXT_STATUS_PTR(engine));
 			read = GEN8_CSB_READ_PTR(ptr);
 			write = GEN8_CSB_WRITE_PTR(ptr);
-			seq_printf(m, "\tExeclist CSB read %d, write %d, interrupt posted? %s\n",
-				   read, write,
+			seq_printf(m, "\tExeclist CSB read %d [%d cached], write %d [%d from hws], interrupt posted? %s\n",
+				   read, execlists->csb_head,
+				   write,
+				   intel_read_status_page(engine, intel_hws_csb_write_index(engine->i915)),
 				   yesno(test_bit(ENGINE_IRQ_EXECLIST,
 						  &engine->irq_posted)));
 			if (read >= GEN8_CSB_ENTRIES)
@@ -3335,18 +3338,19 @@ static int i915_engine_info(struct seq_file *m, void *unused)
 				write += GEN8_CSB_ENTRIES;
 			while (read < write) {
 				idx = ++read % GEN8_CSB_ENTRIES;
-				seq_printf(m, "\tExeclist CSB[%d]: 0x%08x, context: %d\n",
+				seq_printf(m, "\tExeclist CSB[%d]: 0x%08x [0x%08x in hwsp], context: %d [%d in hwsp]\n",
 					   idx,
 					   I915_READ(RING_CONTEXT_STATUS_BUF_LO(engine, idx)),
-					   I915_READ(RING_CONTEXT_STATUS_BUF_HI(engine, idx)));
+					   hws[idx * 2],
+					   I915_READ(RING_CONTEXT_STATUS_BUF_HI(engine, idx)),
+					   hws[idx * 2 + 1]);
 			}
 
 			rcu_read_lock();
-			for (idx = 0; idx < ARRAY_SIZE(engine->execlist_port); idx++) {
+			for (idx = 0; idx < execlists_num_ports(execlists); idx++) {
 				unsigned int count;
 
-				rq = port_unpack(&engine->execlist_port[idx],
-						 &count);
+				rq = port_unpack(&execlists->port[idx], &count);
 				if (rq) {
 					seq_printf(m, "\t\tELSP[%d] count=%d, ",
 						   idx, count);
@@ -3359,7 +3363,7 @@ static int i915_engine_info(struct seq_file *m, void *unused)
 			rcu_read_unlock();
 
 			spin_lock_irq(&engine->timeline->lock);
-			for (rb = engine->execlist_first; rb; rb = rb_next(rb)){
+			for (rb = execlists->first; rb; rb = rb_next(rb)) {
 				struct i915_priolist *p =
 					rb_entry(rb, typeof(*p), node);
 
@@ -3403,7 +3407,7 @@ static int i915_semaphore_status(struct seq_file *m, void *unused)
 	enum intel_engine_id id;
 	int j, ret;
 
-	if (!i915.semaphores) {
+	if (!i915_modparams.semaphores) {
 		seq_puts(m, "Semaphores are disabled\n");
 		return 0;
 	}
@@ -3523,6 +3527,57 @@ static int i915_wa_registers(struct seq_file *m, void *unused)
 	return 0;
 }
 
+static int i915_ipc_status_show(struct seq_file *m, void *data)
+{
+	struct drm_i915_private *dev_priv = m->private;
+
+	seq_printf(m, "Isochronous Priority Control: %s\n",
+			yesno(dev_priv->ipc_enabled));
+	return 0;
+}
+
+static int i915_ipc_status_open(struct inode *inode, struct file *file)
+{
+	struct drm_i915_private *dev_priv = inode->i_private;
+
+	if (!HAS_IPC(dev_priv))
+		return -ENODEV;
+
+	return single_open(file, i915_ipc_status_show, dev_priv);
+}
+
+static ssize_t i915_ipc_status_write(struct file *file, const char __user *ubuf,
+				     size_t len, loff_t *offp)
+{
+	struct seq_file *m = file->private_data;
+	struct drm_i915_private *dev_priv = m->private;
+	int ret;
+	bool enable;
+
+	ret = kstrtobool_from_user(ubuf, len, &enable);
+	if (ret < 0)
+		return ret;
+
+	intel_runtime_pm_get(dev_priv);
+	if (!dev_priv->ipc_enabled && enable)
+		DRM_INFO("Enabling IPC: WM will be proper only after next commit\n");
+	dev_priv->wm.distrust_bios_wm = true;
+	dev_priv->ipc_enabled = enable;
+	intel_enable_ipc(dev_priv);
+	intel_runtime_pm_put(dev_priv);
+
+	return len;
+}
+
+static const struct file_operations i915_ipc_status_fops = {
+	.owner = THIS_MODULE,
+	.open = i915_ipc_status_open,
+	.read = seq_read,
+	.llseek = seq_lseek,
+	.release = single_release,
+	.write = i915_ipc_status_write
+};
+
 static int i915_ddb_info(struct seq_file *m, void *unused)
 {
 	struct drm_i915_private *dev_priv = node_to_i915(m->private);
@@ -4674,26 +4729,26 @@ static int i915_sseu_status(struct seq_file *m, void *unused)
 
 static int i915_forcewake_open(struct inode *inode, struct file *file)
 {
-	struct drm_i915_private *dev_priv = inode->i_private;
+	struct drm_i915_private *i915 = inode->i_private;
 
-	if (INTEL_GEN(dev_priv) < 6)
+	if (INTEL_GEN(i915) < 6)
 		return 0;
 
-	intel_runtime_pm_get(dev_priv);
-	intel_uncore_forcewake_get(dev_priv, FORCEWAKE_ALL);
+	intel_runtime_pm_get(i915);
+	intel_uncore_forcewake_user_get(i915);
 
 	return 0;
 }
 
 static int i915_forcewake_release(struct inode *inode, struct file *file)
 {
-	struct drm_i915_private *dev_priv = inode->i_private;
+	struct drm_i915_private *i915 = inode->i_private;
 
-	if (INTEL_GEN(dev_priv) < 6)
+	if (INTEL_GEN(i915) < 6)
 		return 0;
 
-	intel_uncore_forcewake_put(dev_priv, FORCEWAKE_ALL);
-	intel_runtime_pm_put(dev_priv);
+	intel_uncore_forcewake_user_put(i915);
+	intel_runtime_pm_put(i915);
 
 	return 0;
 }
@@ -4859,7 +4914,8 @@ static const struct i915_debugfs_files {
 	{"i915_dp_test_type", &i915_displayport_test_type_fops},
 	{"i915_dp_test_active", &i915_displayport_test_active_fops},
 	{"i915_guc_log_control", &i915_guc_log_control_fops},
-	{"i915_hpd_storm_ctl", &i915_hpd_storm_ctl_fops}
+	{"i915_hpd_storm_ctl", &i915_hpd_storm_ctl_fops},
+	{"i915_ipc_status", &i915_ipc_status_fops}
 };
 
 int i915_debugfs_register(struct drm_i915_private *dev_priv)

drivers/gpu/drm/i915/i915_drv.c

@@ -58,12 +58,12 @@ static unsigned int i915_load_fail_count;
 
 bool __i915_inject_load_failure(const char *func, int line)
 {
-	if (i915_load_fail_count >= i915.inject_load_failure)
+	if (i915_load_fail_count >= i915_modparams.inject_load_failure)
 		return false;
 
-	if (++i915_load_fail_count == i915.inject_load_failure) {
+	if (++i915_load_fail_count == i915_modparams.inject_load_failure) {
 		DRM_INFO("Injecting failure at checkpoint %u [%s:%d]\n",
-			 i915.inject_load_failure, func, line);
+			 i915_modparams.inject_load_failure, func, line);
 		return true;
 	}
 
@@ -106,8 +106,8 @@ __i915_printk(struct drm_i915_private *dev_priv, const char *level,
 
 static bool i915_error_injected(struct drm_i915_private *dev_priv)
 {
-	return i915.inject_load_failure &&
-	       i915_load_fail_count == i915.inject_load_failure;
+	return i915_modparams.inject_load_failure &&
+	       i915_load_fail_count == i915_modparams.inject_load_failure;
 }
 
 #define i915_load_error(dev_priv, fmt, ...)				     \
@@ -321,7 +321,7 @@ static int i915_getparam(struct drm_device *dev, void *data,
 		value = USES_PPGTT(dev_priv);
 		break;
 	case I915_PARAM_HAS_SEMAPHORES:
-		value = i915.semaphores;
+		value = i915_modparams.semaphores;
 		break;
 	case I915_PARAM_HAS_SECURE_BATCHES:
 		value = capable(CAP_SYS_ADMIN);
@@ -340,7 +340,8 @@ static int i915_getparam(struct drm_device *dev, void *data,
 			return -ENODEV;
 		break;
 	case I915_PARAM_HAS_GPU_RESET:
-		value = i915.enable_hangcheck && intel_has_gpu_reset(dev_priv);
+		value = i915_modparams.enable_hangcheck &&
+			intel_has_gpu_reset(dev_priv);
 		if (value && intel_has_reset_engine(dev_priv))
 			value = 2;
 		break;
@@ -869,6 +870,10 @@ static int i915_driver_init_early(struct drm_i915_private *dev_priv,
 	memcpy(device_info, match_info, sizeof(*device_info));
 	device_info->device_id = dev_priv->drm.pdev->device;
 
+	BUILD_BUG_ON(INTEL_MAX_PLATFORMS >
+		     sizeof(device_info->platform_mask) * BITS_PER_BYTE);
+	device_info->platform_mask = BIT(device_info->platform);
+
 	BUG_ON(device_info->gen > sizeof(device_info->gen_mask) * BITS_PER_BYTE);
 	device_info->gen_mask = BIT(device_info->gen - 1);
 
@@ -1031,9 +1036,9 @@ static void i915_driver_cleanup_mmio(struct drm_i915_private *dev_priv)
 
 static void intel_sanitize_options(struct drm_i915_private *dev_priv)
 {
-	i915.enable_execlists =
+	i915_modparams.enable_execlists =
 		intel_sanitize_enable_execlists(dev_priv,
-						i915.enable_execlists);
+						i915_modparams.enable_execlists);
 
 	/*
 	 * i915.enable_ppgtt is read-only, so do an early pass to validate the
@@ -1041,12 +1046,15 @@ static void intel_sanitize_options(struct drm_i915_private *dev_priv)
 	 * do this now so that we can print out any log messages once rather
 	 * than every time we check intel_enable_ppgtt().
 	 */
-	i915.enable_ppgtt =
-		intel_sanitize_enable_ppgtt(dev_priv, i915.enable_ppgtt);
-	DRM_DEBUG_DRIVER("ppgtt mode: %i\n", i915.enable_ppgtt);
+	i915_modparams.enable_ppgtt =
+		intel_sanitize_enable_ppgtt(dev_priv,
+					    i915_modparams.enable_ppgtt);
+	DRM_DEBUG_DRIVER("ppgtt mode: %i\n", i915_modparams.enable_ppgtt);
 
-	i915.semaphores = intel_sanitize_semaphores(dev_priv, i915.semaphores);
-	DRM_DEBUG_DRIVER("use GPU semaphores? %s\n", yesno(i915.semaphores));
+	i915_modparams.semaphores =
+		intel_sanitize_semaphores(dev_priv, i915_modparams.semaphores);
+	DRM_DEBUG_DRIVER("use GPU semaphores? %s\n",
+			 yesno(i915_modparams.semaphores));
 
 	intel_uc_sanitize_options(dev_priv);
 
@@ -1277,7 +1285,7 @@ int i915_driver_load(struct pci_dev *pdev, const struct pci_device_id *ent)
 	int ret;
 
 	/* Enable nuclear pageflip on ILK+ */
-	if (!i915.nuclear_pageflip && match_info->gen < 5)
+	if (!i915_modparams.nuclear_pageflip && match_info->gen < 5)
 		driver.driver_features &= ~DRIVER_ATOMIC;
 
 	ret = -ENOMEM;
@@ -1341,7 +1349,7 @@ int i915_driver_load(struct pci_dev *pdev, const struct pci_device_id *ent)
 
 	intel_runtime_pm_enable(dev_priv);
 
-	dev_priv->ipc_enabled = false;
+	intel_init_ipc(dev_priv);
 
 	if (IS_ENABLED(CONFIG_DRM_I915_DEBUG))
 		DRM_INFO("DRM_I915_DEBUG enabled\n");
@@ -2609,6 +2617,8 @@ static int intel_runtime_resume(struct device *kdev)
 	if (!IS_VALLEYVIEW(dev_priv) && !IS_CHERRYVIEW(dev_priv))
 		intel_hpd_init(dev_priv);
 
+	intel_enable_ipc(dev_priv);
+
 	enable_rpm_wakeref_asserts(dev_priv);
 
 	if (ret)

drivers/gpu/drm/i915/i915_drv.h

@@ -80,8 +80,8 @@
 
 #define DRIVER_NAME		"i915"
 #define DRIVER_DESC		"Intel Graphics"
-#define DRIVER_DATE		"20170907"
-#define DRIVER_TIMESTAMP	1504772900
+#define DRIVER_DATE		"20170929"
+#define DRIVER_TIMESTAMP	1506682238
 
 /* Use I915_STATE_WARN(x) and I915_STATE_WARN_ON() (rather than WARN() and
  * WARN_ON()) for hw state sanity checks to check for unexpected conditions
@@ -93,7 +93,7 @@
 #define I915_STATE_WARN(condition, format...) ({			\
 	int __ret_warn_on = !!(condition);				\
 	if (unlikely(__ret_warn_on))					\
-		if (!WARN(i915.verbose_state_checks, format))		\
+		if (!WARN(i915_modparams.verbose_state_checks, format))	\
 			DRM_ERROR(format);				\
 	unlikely(__ret_warn_on);					\
 })
@@ -126,7 +126,7 @@ static inline uint_fixed_16_16_t u32_to_fixed16(uint32_t val)
 {
 	uint_fixed_16_16_t fp;
 
-	WARN_ON(val >> 16);
+	WARN_ON(val > U16_MAX);
 
 	fp.val = val << 16;
 	return fp;
@@ -163,8 +163,8 @@ static inline uint_fixed_16_16_t max_fixed16(uint_fixed_16_16_t max1,
 static inline uint_fixed_16_16_t clamp_u64_to_fixed16(uint64_t val)
 {
 	uint_fixed_16_16_t fp;
-	WARN_ON(val >> 32);
-	fp.val = clamp_t(uint32_t, val, 0, ~0);
+	WARN_ON(val > U32_MAX);
+	fp.val = (uint32_t) val;
 	return fp;
 }
 
@@ -181,8 +181,8 @@ static inline uint32_t mul_round_up_u32_fixed16(uint32_t val,
 
 	intermediate_val = (uint64_t) val * mul.val;
 	intermediate_val = DIV_ROUND_UP_ULL(intermediate_val, 1 << 16);
-	WARN_ON(intermediate_val >> 32);
-	return clamp_t(uint32_t, intermediate_val, 0, ~0);
+	WARN_ON(intermediate_val > U32_MAX);
+	return (uint32_t) intermediate_val;
 }
 
 static inline uint_fixed_16_16_t mul_fixed16(uint_fixed_16_16_t val,
@@ -211,8 +211,8 @@ static inline uint32_t div_round_up_u32_fixed16(uint32_t val,
 
 	interm_val = (uint64_t)val << 16;
 	interm_val = DIV_ROUND_UP_ULL(interm_val, d.val);
-	WARN_ON(interm_val >> 32);
-	return clamp_t(uint32_t, interm_val, 0, ~0);
+	WARN_ON(interm_val > U32_MAX);
+	return (uint32_t) interm_val;
 }
 
 static inline uint_fixed_16_16_t mul_u32_fixed16(uint32_t val,
@@ -776,7 +776,6 @@ struct intel_csr {
 	func(has_fpga_dbg); \
 	func(has_full_ppgtt); \
 	func(has_full_48bit_ppgtt); \
-	func(has_gmbus_irq); \
 	func(has_gmch_display); \
 	func(has_guc); \
 	func(has_guc_ct); \
@@ -797,7 +796,8 @@ struct intel_csr {
 	func(cursor_needs_physical); \
 	func(hws_needs_physical); \
 	func(overlay_needs_physical); \
-	func(supports_tv);
+	func(supports_tv); \
+	func(has_ipc);
 
 struct sseu_dev_info {
 	u8 slice_mask;
@@ -851,21 +851,28 @@ enum intel_platform {
 };
 
 struct intel_device_info {
-	u32 display_mmio_offset;
 	u16 device_id;
-	u8 num_pipes;
-	u8 num_sprites[I915_MAX_PIPES];
-	u8 num_scalers[I915_MAX_PIPES];
-	u8 gen;
 	u16 gen_mask;
-	enum intel_platform platform;
+
+	u8 gen;
 	u8 gt; /* GT number, 0 if undefined */
-	u8 ring_mask; /* Rings supported by the HW */
 	u8 num_rings;
+	u8 ring_mask; /* Rings supported by the HW */
+
+	enum intel_platform platform;
+	u32 platform_mask;
+
+	u32 display_mmio_offset;
+
+	u8 num_pipes;
+	u8 num_sprites[I915_MAX_PIPES];
+	u8 num_scalers[I915_MAX_PIPES];
+
 #define DEFINE_FLAG(name) u8 name:1
 	DEV_INFO_FOR_EACH_FLAG(DEFINE_FLAG);
 #undef DEFINE_FLAG
 	u16 ddb_size; /* in blocks */
+
 	/* Register offsets for the various display pipes and transcoders */
 	int pipe_offsets[I915_MAX_TRANSCODERS];
 	int trans_offsets[I915_MAX_TRANSCODERS];
@@ -1000,7 +1007,8 @@ struct i915_gpu_state {
 			u32 seqno;
 			u32 head;
 			u32 tail;
-		} *requests, execlist[2];
+		} *requests, execlist[EXECLIST_MAX_PORTS];
+		unsigned int num_ports;
 
 		struct drm_i915_error_waiter {
 			char comm[TASK_COMM_LEN];
@@ -1178,6 +1186,14 @@ struct i915_psr {
 	bool y_cord_support;
 	bool colorimetry_support;
 	bool alpm;
+
+	void (*enable_source)(struct intel_dp *,
+			      const struct intel_crtc_state *);
+	void (*disable_source)(struct intel_dp *,
+			       const struct intel_crtc_state *);
+	void (*enable_sink)(struct intel_dp *);
+	void (*activate)(struct intel_dp *);
+	void (*setup_vsc)(struct intel_dp *, const struct intel_crtc_state *);
 };
 
 enum intel_pch {
@@ -1836,6 +1852,20 @@ struct skl_wm_level {
 	uint8_t plane_res_l;
 };
 
+/* Stores plane specific WM parameters */
+struct skl_wm_params {
+	bool x_tiled, y_tiled;
+	bool rc_surface;
+	uint32_t width;
+	uint8_t cpp;
+	uint32_t plane_pixel_rate;
+	uint32_t y_min_scanlines;
+	uint32_t plane_bytes_per_line;
+	uint_fixed_16_16_t plane_blocks_per_line;
+	uint_fixed_16_16_t y_tile_minimum;
+	uint32_t linetime_us;
+};
+
 /*
  * This struct helps tracking the state needed for runtime PM, which puts the
  * device in PCI D3 state. Notice that when this happens, nothing on the
@@ -2331,6 +2361,8 @@ struct drm_i915_private {
 	DECLARE_HASHTABLE(mm_structs, 7);
 	struct mutex mm_lock;
 
+	struct intel_ppat ppat;
+
 	/* Kernel Modesetting */
 
 	struct intel_crtc *plane_to_crtc_mapping[I915_MAX_PIPES];
@@ -2811,8 +2843,8 @@ static inline struct scatterlist *__sg_next(struct scatterlist *sg)
 #define for_each_sgt_dma(__dmap, __iter, __sgt)				\
 	for ((__iter) = __sgt_iter((__sgt)->sgl, true);			\
 	     ((__dmap) = (__iter).dma + (__iter).curr);			\
-	     (((__iter).curr += PAGE_SIZE) < (__iter).max) ||		\
-	     ((__iter) = __sgt_iter(__sg_next((__iter).sgp), true), 0))
+	     (((__iter).curr += PAGE_SIZE) >= (__iter).max) ?		\
+	     (__iter) = __sgt_iter(__sg_next((__iter).sgp), true), 0 : 0)
 
 /**
  * for_each_sgt_page - iterate over the pages of the given sg_table
@@ -2824,8 +2856,23 @@ static inline struct scatterlist *__sg_next(struct scatterlist *sg)
 	for ((__iter) = __sgt_iter((__sgt)->sgl, false);		\
 	     ((__pp) = (__iter).pfn == 0 ? NULL :			\
 	      pfn_to_page((__iter).pfn + ((__iter).curr >> PAGE_SHIFT))); \
-	     (((__iter).curr += PAGE_SIZE) < (__iter).max) ||		\
-	     ((__iter) = __sgt_iter(__sg_next((__iter).sgp), false), 0))
+	     (((__iter).curr += PAGE_SIZE) >= (__iter).max) ?		\
+	     (__iter) = __sgt_iter(__sg_next((__iter).sgp), false), 0 : 0)
+
+static inline unsigned int i915_sg_segment_size(void)
+{
+	unsigned int size = swiotlb_max_segment();
+
+	if (size == 0)
+		return SCATTERLIST_MAX_SEGMENT;
+
+	size = rounddown(size, PAGE_SIZE);
+	/* swiotlb_max_segment_size can return 1 byte when it means one page. */
+	if (size < PAGE_SIZE)
+		size = PAGE_SIZE;
+
+	return size;
+}
 
 static inline const struct intel_device_info *
 intel_info(const struct drm_i915_private *dev_priv)
@@ -2842,23 +2889,21 @@ intel_info(const struct drm_i915_private *dev_priv)
 #define INTEL_REVID(dev_priv)	((dev_priv)->drm.pdev->revision)
 
 #define GEN_FOREVER (0)
+
+#define INTEL_GEN_MASK(s, e) ( \
+	BUILD_BUG_ON_ZERO(!__builtin_constant_p(s)) + \
+	BUILD_BUG_ON_ZERO(!__builtin_constant_p(e)) + \
+	GENMASK((e) != GEN_FOREVER ? (e) - 1 : BITS_PER_LONG - 1, \
+		(s) != GEN_FOREVER ? (s) - 1 : 0) \
+)
+
 /*
  * Returns true if Gen is in inclusive range [Start, End].
  *
  * Use GEN_FOREVER for unbound start and or end.
  */
-#define IS_GEN(dev_priv, s, e) ({ \
-	unsigned int __s = (s), __e = (e); \
-	BUILD_BUG_ON(!__builtin_constant_p(s)); \
-	BUILD_BUG_ON(!__builtin_constant_p(e)); \
-	if ((__s) != GEN_FOREVER) \
-		__s = (s) - 1; \
-	if ((__e) == GEN_FOREVER) \
-		__e = BITS_PER_LONG - 1; \
-	else \
-		__e = (e) - 1; \
-	!!((dev_priv)->info.gen_mask & GENMASK((__e), (__s))); \
-})
+#define IS_GEN(dev_priv, s, e) \
+	(!!((dev_priv)->info.gen_mask & INTEL_GEN_MASK((s), (e))))
 
 /*
  * Return true if revision is in range [since,until] inclusive.
@@ -2868,37 +2913,39 @@ intel_info(const struct drm_i915_private *dev_priv)
 #define IS_REVID(p, since, until) \
 	(INTEL_REVID(p) >= (since) && INTEL_REVID(p) <= (until))
 
-#define IS_I830(dev_priv)	((dev_priv)->info.platform == INTEL_I830)
-#define IS_I845G(dev_priv)	((dev_priv)->info.platform == INTEL_I845G)
-#define IS_I85X(dev_priv)	((dev_priv)->info.platform == INTEL_I85X)
-#define IS_I865G(dev_priv)	((dev_priv)->info.platform == INTEL_I865G)
-#define IS_I915G(dev_priv)	((dev_priv)->info.platform == INTEL_I915G)
-#define IS_I915GM(dev_priv)	((dev_priv)->info.platform == INTEL_I915GM)
-#define IS_I945G(dev_priv)	((dev_priv)->info.platform == INTEL_I945G)
-#define IS_I945GM(dev_priv)	((dev_priv)->info.platform == INTEL_I945GM)
-#define IS_I965G(dev_priv)	((dev_priv)->info.platform == INTEL_I965G)
-#define IS_I965GM(dev_priv)	((dev_priv)->info.platform == INTEL_I965GM)
-#define IS_G45(dev_priv)	((dev_priv)->info.platform == INTEL_G45)
-#define IS_GM45(dev_priv)	((dev_priv)->info.platform == INTEL_GM45)
+#define IS_PLATFORM(dev_priv, p) ((dev_priv)->info.platform_mask & BIT(p))
+
+#define IS_I830(dev_priv)	IS_PLATFORM(dev_priv, INTEL_I830)
+#define IS_I845G(dev_priv)	IS_PLATFORM(dev_priv, INTEL_I845G)
+#define IS_I85X(dev_priv)	IS_PLATFORM(dev_priv, INTEL_I85X)
+#define IS_I865G(dev_priv)	IS_PLATFORM(dev_priv, INTEL_I865G)
+#define IS_I915G(dev_priv)	IS_PLATFORM(dev_priv, INTEL_I915G)
+#define IS_I915GM(dev_priv)	IS_PLATFORM(dev_priv, INTEL_I915GM)
+#define IS_I945G(dev_priv)	IS_PLATFORM(dev_priv, INTEL_I945G)
+#define IS_I945GM(dev_priv)	IS_PLATFORM(dev_priv, INTEL_I945GM)
+#define IS_I965G(dev_priv)	IS_PLATFORM(dev_priv, INTEL_I965G)
+#define IS_I965GM(dev_priv)	IS_PLATFORM(dev_priv, INTEL_I965GM)
+#define IS_G45(dev_priv)	IS_PLATFORM(dev_priv, INTEL_G45)
+#define IS_GM45(dev_priv)	IS_PLATFORM(dev_priv, INTEL_GM45)
 #define IS_G4X(dev_priv)	(IS_G45(dev_priv) || IS_GM45(dev_priv))
 #define IS_PINEVIEW_G(dev_priv)	(INTEL_DEVID(dev_priv) == 0xa001)
 #define IS_PINEVIEW_M(dev_priv)	(INTEL_DEVID(dev_priv) == 0xa011)
-#define IS_PINEVIEW(dev_priv)	((dev_priv)->info.platform == INTEL_PINEVIEW)
-#define IS_G33(dev_priv)	((dev_priv)->info.platform == INTEL_G33)
+#define IS_PINEVIEW(dev_priv)	IS_PLATFORM(dev_priv, INTEL_PINEVIEW)
+#define IS_G33(dev_priv)	IS_PLATFORM(dev_priv, INTEL_G33)
 #define IS_IRONLAKE_M(dev_priv)	(INTEL_DEVID(dev_priv) == 0x0046)
-#define IS_IVYBRIDGE(dev_priv)	((dev_priv)->info.platform == INTEL_IVYBRIDGE)
+#define IS_IVYBRIDGE(dev_priv)	IS_PLATFORM(dev_priv, INTEL_IVYBRIDGE)
 #define IS_IVB_GT1(dev_priv)	(IS_IVYBRIDGE(dev_priv) && \
 				 (dev_priv)->info.gt == 1)
-#define IS_VALLEYVIEW(dev_priv)	((dev_priv)->info.platform == INTEL_VALLEYVIEW)
-#define IS_CHERRYVIEW(dev_priv)	((dev_priv)->info.platform == INTEL_CHERRYVIEW)
-#define IS_HASWELL(dev_priv)	((dev_priv)->info.platform == INTEL_HASWELL)
-#define IS_BROADWELL(dev_priv)	((dev_priv)->info.platform == INTEL_BROADWELL)
-#define IS_SKYLAKE(dev_priv)	((dev_priv)->info.platform == INTEL_SKYLAKE)
-#define IS_BROXTON(dev_priv)	((dev_priv)->info.platform == INTEL_BROXTON)
-#define IS_KABYLAKE(dev_priv)	((dev_priv)->info.platform == INTEL_KABYLAKE)
-#define IS_GEMINILAKE(dev_priv)	((dev_priv)->info.platform == INTEL_GEMINILAKE)
-#define IS_COFFEELAKE(dev_priv)	((dev_priv)->info.platform == INTEL_COFFEELAKE)
-#define IS_CANNONLAKE(dev_priv)	((dev_priv)->info.platform == INTEL_CANNONLAKE)
+#define IS_VALLEYVIEW(dev_priv)	IS_PLATFORM(dev_priv, INTEL_VALLEYVIEW)
+#define IS_CHERRYVIEW(dev_priv)	IS_PLATFORM(dev_priv, INTEL_CHERRYVIEW)
+#define IS_HASWELL(dev_priv)	IS_PLATFORM(dev_priv, INTEL_HASWELL)
+#define IS_BROADWELL(dev_priv)	IS_PLATFORM(dev_priv, INTEL_BROADWELL)
+#define IS_SKYLAKE(dev_priv)	IS_PLATFORM(dev_priv, INTEL_SKYLAKE)
+#define IS_BROXTON(dev_priv)	IS_PLATFORM(dev_priv, INTEL_BROXTON)
+#define IS_KABYLAKE(dev_priv)	IS_PLATFORM(dev_priv, INTEL_KABYLAKE)
+#define IS_GEMINILAKE(dev_priv)	IS_PLATFORM(dev_priv, INTEL_GEMINILAKE)
+#define IS_COFFEELAKE(dev_priv)	IS_PLATFORM(dev_priv, INTEL_COFFEELAKE)
+#define IS_CANNONLAKE(dev_priv)	IS_PLATFORM(dev_priv, INTEL_CANNONLAKE)
 #define IS_MOBILE(dev_priv)	((dev_priv)->info.is_mobile)
 #define IS_HSW_EARLY_SDV(dev_priv) (IS_HASWELL(dev_priv) && \
 				    (INTEL_DEVID(dev_priv) & 0xFF00) == 0x0C00)
@@ -2946,6 +2993,8 @@ intel_info(const struct drm_i915_private *dev_priv)
 				 (dev_priv)->info.gt == 3)
 #define IS_CFL_ULT(dev_priv)	(IS_COFFEELAKE(dev_priv) && \
 				 (INTEL_DEVID(dev_priv) & 0x00F0) == 0x00A0)
+#define IS_CFL_GT2(dev_priv)	(IS_COFFEELAKE(dev_priv) && \
+				 (dev_priv)->info.gt == 2)
 
 #define IS_ALPHA_SUPPORT(intel_info) ((intel_info)->is_alpha_support)
 
@@ -3036,9 +3085,9 @@ intel_info(const struct drm_i915_private *dev_priv)
 
 #define HAS_LOGICAL_RING_CONTEXTS(dev_priv) \
 		((dev_priv)->info.has_logical_ring_contexts)
-#define USES_PPGTT(dev_priv)		(i915.enable_ppgtt)
-#define USES_FULL_PPGTT(dev_priv)	(i915.enable_ppgtt >= 2)
-#define USES_FULL_48BIT_PPGTT(dev_priv)	(i915.enable_ppgtt == 3)
+#define USES_PPGTT(dev_priv)		(i915_modparams.enable_ppgtt)
+#define USES_FULL_PPGTT(dev_priv)	(i915_modparams.enable_ppgtt >= 2)
+#define USES_FULL_48BIT_PPGTT(dev_priv)	(i915_modparams.enable_ppgtt == 3)
 
 #define HAS_OVERLAY(dev_priv)		 ((dev_priv)->info.has_overlay)
 #define OVERLAY_NEEDS_PHYSICAL(dev_priv) \
@@ -3056,9 +3105,12 @@ intel_info(const struct drm_i915_private *dev_priv)
  * even when in MSI mode. This results in spurious interrupt warnings if the
  * legacy irq no. is shared with another device. The kernel then disables that
  * interrupt source and so prevents the other device from working properly.
+ *
+ * Since we don't enable MSI anymore on gen4, we can always use GMBUS/AUX
+ * interrupts.
  */
-#define HAS_AUX_IRQ(dev_priv)   ((dev_priv)->info.gen >= 5)
-#define HAS_GMBUS_IRQ(dev_priv) ((dev_priv)->info.has_gmbus_irq)
+#define HAS_AUX_IRQ(dev_priv)   true
+#define HAS_GMBUS_IRQ(dev_priv) (INTEL_GEN(dev_priv) >= 4)
 
 /* With the 945 and later, Y tiling got adjusted so that it was 32 128-byte
  * rows, which changed the alignment requirements and fence programming.
@@ -3089,6 +3141,8 @@ intel_info(const struct drm_i915_private *dev_priv)
 #define HAS_RUNTIME_PM(dev_priv) ((dev_priv)->info.has_runtime_pm)
 #define HAS_64BIT_RELOC(dev_priv) ((dev_priv)->info.has_64bit_reloc)
 
+#define HAS_IPC(dev_priv)		 ((dev_priv)->info.has_ipc)
+
 /*
  * For now, anything with a GuC requires uCode loading, and then supports
  * command submission once loaded. But these are logically independent
@@ -3234,7 +3288,7 @@ static inline void i915_queue_hangcheck(struct drm_i915_private *dev_priv)
 {
 	unsigned long delay;
 
-	if (unlikely(!i915.enable_hangcheck))
+	if (unlikely(!i915_modparams.enable_hangcheck))
 		return;
 
 	/* Don't continually defer the hangcheck so that it is always run at
@@ -3267,6 +3321,8 @@ static inline bool intel_vgpu_active(struct drm_i915_private *dev_priv)
 	return dev_priv->vgpu.active;
 }
 
+u32 i915_pipestat_enable_mask(struct drm_i915_private *dev_priv,
+			      enum pipe pipe);
 void
 i915_enable_pipestat(struct drm_i915_private *dev_priv, enum pipe pipe,
 		     u32 status_mask);
@@ -4360,4 +4416,12 @@ int remap_io_mapping(struct vm_area_struct *vma,
 		     unsigned long addr, unsigned long pfn, unsigned long size,
 		     struct io_mapping *iomap);
 
+static inline int intel_hws_csb_write_index(struct drm_i915_private *i915)
+{
+	if (INTEL_GEN(i915) >= 10)
+		return CNL_HWS_CSB_WRITE_INDEX;
+	else
+		return I915_HWS_CSB_WRITE_INDEX;
+}
+
 #endif

drivers/gpu/drm/i915/i915_gem.c

@@ -179,7 +179,7 @@ i915_gem_object_get_pages_phys(struct drm_i915_gem_object *obj)
 	 * the alignment of the buddy allocation will naturally match.
 	 */
 	phys = drm_pci_alloc(obj->base.dev,
-			     obj->base.size,
+			     roundup_pow_of_two(obj->base.size),
 			     roundup_pow_of_two(obj->base.size));
 	if (!phys)
 		return ERR_PTR(-ENOMEM);
@@ -694,10 +694,10 @@ flush_write_domain(struct drm_i915_gem_object *obj, unsigned int flush_domains)
 
 	switch (obj->base.write_domain) {
 	case I915_GEM_DOMAIN_GTT:
-		if (INTEL_GEN(dev_priv) >= 6 && !HAS_LLC(dev_priv)) {
+		if (!HAS_LLC(dev_priv)) {
 			intel_runtime_pm_get(dev_priv);
 			spin_lock_irq(&dev_priv->uncore.lock);
-			POSTING_READ_FW(RING_ACTHD(dev_priv->engine[RCS]->mmio_base));
+			POSTING_READ_FW(RING_HEAD(dev_priv->engine[RCS]->mmio_base));
 			spin_unlock_irq(&dev_priv->uncore.lock);
 			intel_runtime_pm_put(dev_priv);
 		}
@@ -2303,7 +2303,7 @@ i915_gem_object_get_pages_gtt(struct drm_i915_gem_object *obj)
 	struct sgt_iter sgt_iter;
 	struct page *page;
 	unsigned long last_pfn = 0;	/* suppress gcc warning */
-	unsigned int max_segment;
+	unsigned int max_segment = i915_sg_segment_size();
 	gfp_t noreclaim;
 	int ret;
 
@@ -2314,10 +2314,6 @@ i915_gem_object_get_pages_gtt(struct drm_i915_gem_object *obj)
 	GEM_BUG_ON(obj->base.read_domains & I915_GEM_GPU_DOMAINS);
 	GEM_BUG_ON(obj->base.write_domain & I915_GEM_GPU_DOMAINS);
 
-	max_segment = swiotlb_max_segment();
-	if (!max_segment)
-		max_segment = rounddown(UINT_MAX, PAGE_SIZE);
-
 	st = kmalloc(sizeof(*st), GFP_KERNEL);
 	if (st == NULL)
 		return ERR_PTR(-ENOMEM);
@@ -2819,8 +2815,8 @@ i915_gem_reset_prepare_engine(struct intel_engine_cs *engine)
 	 * Turning off the engine->irq_tasklet until the reset is over
 	 * prevents the race.
 	 */
-	tasklet_kill(&engine->irq_tasklet);
-	tasklet_disable(&engine->irq_tasklet);
+	tasklet_kill(&engine->execlists.irq_tasklet);
+	tasklet_disable(&engine->execlists.irq_tasklet);
 
 	if (engine->irq_seqno_barrier)
 		engine->irq_seqno_barrier(engine);
@@ -2999,7 +2995,7 @@ void i915_gem_reset(struct drm_i915_private *dev_priv)
 
 void i915_gem_reset_finish_engine(struct intel_engine_cs *engine)
 {
-	tasklet_enable(&engine->irq_tasklet);
+	tasklet_enable(&engine->execlists.irq_tasklet);
 	kthread_unpark(engine->breadcrumbs.signaler);
 }
 
@@ -3026,9 +3022,6 @@ static void nop_submit_request(struct drm_i915_gem_request *request)
 
 static void engine_set_wedged(struct intel_engine_cs *engine)
 {
-	struct drm_i915_gem_request *request;
-	unsigned long flags;
-
 	/* We need to be sure that no thread is running the old callback as
 	 * we install the nop handler (otherwise we would submit a request
 	 * to hardware that will never complete). In order to prevent this
@@ -3038,40 +3031,7 @@ static void engine_set_wedged(struct intel_engine_cs *engine)
 	engine->submit_request = nop_submit_request;
 
 	/* Mark all executing requests as skipped */
-	spin_lock_irqsave(&engine->timeline->lock, flags);
-	list_for_each_entry(request, &engine->timeline->requests, link)
-		if (!i915_gem_request_completed(request))
-			dma_fence_set_error(&request->fence, -EIO);
-	spin_unlock_irqrestore(&engine->timeline->lock, flags);
-
-	/*
-	 * Clear the execlists queue up before freeing the requests, as those
-	 * are the ones that keep the context and ringbuffer backing objects
-	 * pinned in place.
-	 */
-
-	if (i915.enable_execlists) {
-		struct execlist_port *port = engine->execlist_port;
-		unsigned long flags;
-		unsigned int n;
-
-		spin_lock_irqsave(&engine->timeline->lock, flags);
-
-		for (n = 0; n < ARRAY_SIZE(engine->execlist_port); n++)
-			i915_gem_request_put(port_request(&port[n]));
-		memset(engine->execlist_port, 0, sizeof(engine->execlist_port));
-		engine->execlist_queue = RB_ROOT;
-		engine->execlist_first = NULL;
-
-		spin_unlock_irqrestore(&engine->timeline->lock, flags);
-
-		/* The port is checked prior to scheduling a tasklet, but
-		 * just in case we have suspended the tasklet to do the
-		 * wedging make sure that when it wakes, it decides there
-		 * is no work to do by clearing the irq_posted bit.
-		 */
-		clear_bit(ENGINE_IRQ_EXECLIST, &engine->irq_posted);
-	}
+	engine->cancel_requests(engine);
 
 	/* Mark all pending requests as complete so that any concurrent
 	 * (lockless) lookup doesn't try and wait upon the request as we
@@ -4778,7 +4738,7 @@ bool intel_sanitize_semaphores(struct drm_i915_private *dev_priv, int value)
 		return false;
 
 	/* TODO: make semaphores and Execlists play nicely together */
-	if (i915.enable_execlists)
+	if (i915_modparams.enable_execlists)
 		return false;
 
 	if (value >= 0)
@@ -4799,7 +4759,7 @@ int i915_gem_init(struct drm_i915_private *dev_priv)
 
 	dev_priv->mm.unordered_timeline = dma_fence_context_alloc(1);
 
-	if (!i915.enable_execlists) {
+	if (!i915_modparams.enable_execlists) {
 		dev_priv->gt.resume = intel_legacy_submission_resume;
 		dev_priv->gt.cleanup_engine = intel_engine_cleanup;
 	} else {

drivers/gpu/drm/i915/i915_gem_context.c

@@ -314,7 +314,7 @@ __create_hw_context(struct drm_i915_private *dev_priv,
 	 * present or not in use we still need a small bias as ring wraparound
 	 * at offset 0 sometimes hangs. No idea why.
 	 */
-	if (HAS_GUC(dev_priv) && i915.enable_guc_loading)
+	if (HAS_GUC(dev_priv) && i915_modparams.enable_guc_loading)
 		ctx->ggtt_offset_bias = GUC_WOPCM_TOP;
 	else
 		ctx->ggtt_offset_bias = I915_GTT_PAGE_SIZE;
@@ -407,7 +407,7 @@ i915_gem_context_create_gvt(struct drm_device *dev)
 	i915_gem_context_set_closed(ctx); /* not user accessible */
 	i915_gem_context_clear_bannable(ctx);
 	i915_gem_context_set_force_single_submission(ctx);
-	if (!i915.enable_guc_submission)
+	if (!i915_modparams.enable_guc_submission)
 		ctx->ring_size = 512 * PAGE_SIZE; /* Max ring buffer size */
 
 	GEM_BUG_ON(i915_gem_context_is_kernel(ctx));
@@ -431,7 +431,7 @@ int i915_gem_contexts_init(struct drm_i915_private *dev_priv)
 
 	if (intel_vgpu_active(dev_priv) &&
 	    HAS_LOGICAL_RING_CONTEXTS(dev_priv)) {
-		if (!i915.enable_execlists) {
+		if (!i915_modparams.enable_execlists) {
 			DRM_INFO("Only EXECLIST mode is supported in vgpu.\n");
 			return -EINVAL;
 		}
@@ -483,7 +483,7 @@ void i915_gem_contexts_lost(struct drm_i915_private *dev_priv)
 	}
 
 	/* Force the GPU state to be restored on enabling */
-	if (!i915.enable_execlists) {
+	if (!i915_modparams.enable_execlists) {
 		struct i915_gem_context *ctx;
 
 		list_for_each_entry(ctx, &dev_priv->contexts.list, link) {
@@ -568,7 +568,7 @@ mi_set_context(struct drm_i915_gem_request *req, u32 flags)
 	enum intel_engine_id id;
 	const int num_rings =
 		/* Use an extended w/a on gen7 if signalling from other rings */
-		(i915.semaphores && INTEL_GEN(dev_priv) == 7) ?
+		(i915_modparams.semaphores && INTEL_GEN(dev_priv) == 7) ?
 		INTEL_INFO(dev_priv)->num_rings - 1 :
 		0;
 	int len;
@@ -837,7 +837,7 @@ int i915_switch_context(struct drm_i915_gem_request *req)
 	struct intel_engine_cs *engine = req->engine;
 
 	lockdep_assert_held(&req->i915->drm.struct_mutex);
-	if (i915.enable_execlists)
+	if (i915_modparams.enable_execlists)
 		return 0;
 
 	if (!req->ctx->engine[engine->id].state) {

drivers/gpu/drm/i915/i915_gem_execbuffer.c

@@ -58,6 +58,7 @@ enum {
 
 #define __EXEC_HAS_RELOC	BIT(31)
 #define __EXEC_VALIDATED	BIT(30)
+#define __EXEC_INTERNAL_FLAGS	(~0u << 30)
 #define UPDATE			PIN_OFFSET_FIXED
 
 #define BATCH_OFFSET_BIAS (256*1024)
@@ -679,7 +680,7 @@ static int eb_select_context(struct i915_execbuffer *eb)
 static int eb_lookup_vmas(struct i915_execbuffer *eb)
 {
 	struct radix_tree_root *handles_vma = &eb->ctx->handles_vma;
-	struct drm_i915_gem_object *uninitialized_var(obj);
+	struct drm_i915_gem_object *obj;
 	unsigned int i;
 	int err;
 
@@ -725,19 +726,17 @@ static int eb_lookup_vmas(struct i915_execbuffer *eb)
 			goto err_obj;
 		}
 
+		/* transfer ref to ctx */
 		vma->open_count++;
 		list_add(&lut->obj_link, &obj->lut_list);
 		list_add(&lut->ctx_link, &eb->ctx->handles_list);
 		lut->ctx = eb->ctx;
 		lut->handle = handle;
 
-		/* transfer ref to ctx */
-		obj = NULL;
-
 add_vma:
 		err = eb_add_vma(eb, i, vma);
 		if (unlikely(err))
-			goto err_obj;
+			goto err_vma;
 
 		GEM_BUG_ON(vma != eb->vma[i]);
 		GEM_BUG_ON(vma->exec_flags != &eb->flags[i]);
@@ -766,8 +765,7 @@ add_vma:
 	return eb_reserve(eb);
 
 err_obj:
-	if (obj)
-		i915_gem_object_put(obj);
+	i915_gem_object_put(obj);
 err_vma:
 	eb->vma[i] = NULL;
 	return err;
@@ -1587,7 +1585,7 @@ static int eb_prefault_relocations(const struct i915_execbuffer *eb)
 	const unsigned int count = eb->buffer_count;
 	unsigned int i;
 
-	if (unlikely(i915.prefault_disable))
+	if (unlikely(i915_modparams.prefault_disable))
 		return 0;
 
 	for (i = 0; i < count; i++) {
@@ -2188,6 +2186,7 @@ i915_gem_do_execbuffer(struct drm_device *dev,
 	int out_fence_fd = -1;
 	int err;
 
+	BUILD_BUG_ON(__EXEC_INTERNAL_FLAGS & ~__I915_EXEC_ILLEGAL_FLAGS);
 	BUILD_BUG_ON(__EXEC_OBJECT_INTERNAL_FLAGS &
 		     ~__EXEC_OBJECT_UNKNOWN_FLAGS);
 

drivers/gpu/drm/i915/i915_gem_gtt.c

@@ -180,7 +180,7 @@ int intel_sanitize_enable_ppgtt(struct drm_i915_private *dev_priv,
 		return 0;
 	}
 
-	if (INTEL_GEN(dev_priv) >= 8 && i915.enable_execlists) {
+	if (INTEL_GEN(dev_priv) >= 8 && i915_modparams.enable_execlists) {
 		if (has_full_48bit_ppgtt)
 			return 3;
 
@@ -230,13 +230,13 @@ static gen8_pte_t gen8_pte_encode(dma_addr_t addr,
 
 	switch (level) {
 	case I915_CACHE_NONE:
-		pte |= PPAT_UNCACHED_INDEX;
+		pte |= PPAT_UNCACHED;
 		break;
 	case I915_CACHE_WT:
-		pte |= PPAT_DISPLAY_ELLC_INDEX;
+		pte |= PPAT_DISPLAY_ELLC;
 		break;
 	default:
-		pte |= PPAT_CACHED_INDEX;
+		pte |= PPAT_CACHED;
 		break;
 	}
 
@@ -249,9 +249,9 @@ static gen8_pde_t gen8_pde_encode(const dma_addr_t addr,
 	gen8_pde_t pde = _PAGE_PRESENT | _PAGE_RW;
 	pde |= addr;
 	if (level != I915_CACHE_NONE)
-		pde |= PPAT_CACHED_PDE_INDEX;
+		pde |= PPAT_CACHED_PDE;
 	else
-		pde |= PPAT_UNCACHED_INDEX;
+		pde |= PPAT_UNCACHED;
 	return pde;
 }
 
@@ -481,10 +481,8 @@ static void fill_page_dma(struct i915_address_space *vm,
 			  const u64 val)
 {
 	u64 * const vaddr = kmap_atomic(p->page);
-	int i;
 
-	for (i = 0; i < 512; i++)
-		vaddr[i] = val;
+	memset64(vaddr, val, PAGE_SIZE / sizeof(val));
 
 	kunmap_atomic(vaddr);
 }
@@ -1168,19 +1166,22 @@ static int gen8_ppgtt_alloc_pd(struct i915_address_space *vm,
 	unsigned int pde;
 
 	gen8_for_each_pde(pt, pd, start, length, pde) {
+		int count = gen8_pte_count(start, length);
+
 		if (pt == vm->scratch_pt) {
 			pt = alloc_pt(vm);
 			if (IS_ERR(pt))
 				goto unwind;
 
-			gen8_initialize_pt(vm, pt);
+			if (count < GEN8_PTES)
+				gen8_initialize_pt(vm, pt);
 
 			gen8_ppgtt_set_pde(vm, pd, pt, pde);
 			pd->used_pdes++;
 			GEM_BUG_ON(pd->used_pdes > I915_PDES);
 		}
 
-		pt->used_ptes += gen8_pte_count(start, length);
+		pt->used_ptes += count;
 	}
 	return 0;
 
@@ -1969,7 +1970,7 @@ int i915_ppgtt_init_hw(struct drm_i915_private *dev_priv)
 	/* In the case of execlists, PPGTT is enabled by the context descriptor
 	 * and the PDPs are contained within the context itself.  We don't
 	 * need to do anything here. */
-	if (i915.enable_execlists)
+	if (i915_modparams.enable_execlists)
 		return 0;
 
 	if (!USES_PPGTT(dev_priv))
@@ -2816,41 +2817,209 @@ static int ggtt_probe_common(struct i915_ggtt *ggtt, u64 size)
 	return 0;
 }
 
-static void cnl_setup_private_ppat(struct drm_i915_private *dev_priv)
+static struct intel_ppat_entry *
+__alloc_ppat_entry(struct intel_ppat *ppat, unsigned int index, u8 value)
+{
+	struct intel_ppat_entry *entry = &ppat->entries[index];
+
+	GEM_BUG_ON(index >= ppat->max_entries);
+	GEM_BUG_ON(test_bit(index, ppat->used));
+
+	entry->ppat = ppat;
+	entry->value = value;
+	kref_init(&entry->ref);
+	set_bit(index, ppat->used);
+	set_bit(index, ppat->dirty);
+
+	return entry;
+}
+
+static void __free_ppat_entry(struct intel_ppat_entry *entry)
+{
+	struct intel_ppat *ppat = entry->ppat;
+	unsigned int index = entry - ppat->entries;
+
+	GEM_BUG_ON(index >= ppat->max_entries);
+	GEM_BUG_ON(!test_bit(index, ppat->used));
+
+	entry->value = ppat->clear_value;
+	clear_bit(index, ppat->used);
+	set_bit(index, ppat->dirty);
+}
+
+/**
+ * intel_ppat_get - get a usable PPAT entry
+ * @i915: i915 device instance
+ * @value: the PPAT value required by the caller
+ *
+ * The function tries to search if there is an existing PPAT entry which
+ * matches with the required value. If perfectly matched, the existing PPAT
+ * entry will be used. If only partially matched, it will try to check if
+ * there is any available PPAT index. If yes, it will allocate a new PPAT
+ * index for the required entry and update the HW. If not, the partially
+ * matched entry will be used.
+ */
+const struct intel_ppat_entry *
+intel_ppat_get(struct drm_i915_private *i915, u8 value)
 {
+	struct intel_ppat *ppat = &i915->ppat;
+	struct intel_ppat_entry *entry;
+	unsigned int scanned, best_score;
+	int i;
+
+	GEM_BUG_ON(!ppat->max_entries);
+
+	scanned = best_score = 0;
+	for_each_set_bit(i, ppat->used, ppat->max_entries) {
+		unsigned int score;
+
+		score = ppat->match(ppat->entries[i].value, value);
+		if (score > best_score) {
+			entry = &ppat->entries[i];
+			if (score == INTEL_PPAT_PERFECT_MATCH) {
+				kref_get(&entry->ref);
+				return entry;
+			}
+			best_score = score;
+		}
+		scanned++;
+	}
+
+	if (scanned == ppat->max_entries) {
+		if (!best_score)
+			return ERR_PTR(-ENOSPC);
+
+		kref_get(&entry->ref);
+		return entry;
+	}
+
+	i = find_first_zero_bit(ppat->used, ppat->max_entries);
+	entry = __alloc_ppat_entry(ppat, i, value);
+	ppat->update_hw(i915);
+	return entry;
+}
+
+static void release_ppat(struct kref *kref)
+{
+	struct intel_ppat_entry *entry =
+		container_of(kref, struct intel_ppat_entry, ref);
+	struct drm_i915_private *i915 = entry->ppat->i915;
+
+	__free_ppat_entry(entry);
+	entry->ppat->update_hw(i915);
+}
+
+/**
+ * intel_ppat_put - put back the PPAT entry got from intel_ppat_get()
+ * @entry: an intel PPAT entry
+ *
+ * Put back the PPAT entry got from intel_ppat_get(). If the PPAT index of the
+ * entry is dynamically allocated, its reference count will be decreased. Once
+ * the reference count becomes into zero, the PPAT index becomes free again.
+ */
+void intel_ppat_put(const struct intel_ppat_entry *entry)
+{
+	struct intel_ppat *ppat = entry->ppat;
+	unsigned int index = entry - ppat->entries;
+
+	GEM_BUG_ON(!ppat->max_entries);
+
+	kref_put(&ppat->entries[index].ref, release_ppat);
+}
+
+static void cnl_private_pat_update_hw(struct drm_i915_private *dev_priv)
+{
+	struct intel_ppat *ppat = &dev_priv->ppat;
+	int i;
+
+	for_each_set_bit(i, ppat->dirty, ppat->max_entries) {
+		I915_WRITE(GEN10_PAT_INDEX(i), ppat->entries[i].value);
+		clear_bit(i, ppat->dirty);
+	}
+}
+
+static void bdw_private_pat_update_hw(struct drm_i915_private *dev_priv)
+{
+	struct intel_ppat *ppat = &dev_priv->ppat;
+	u64 pat = 0;
+	int i;
+
+	for (i = 0; i < ppat->max_entries; i++)
+		pat |= GEN8_PPAT(i, ppat->entries[i].value);
+
+	bitmap_clear(ppat->dirty, 0, ppat->max_entries);
+
+	I915_WRITE(GEN8_PRIVATE_PAT_LO, lower_32_bits(pat));
+	I915_WRITE(GEN8_PRIVATE_PAT_HI, upper_32_bits(pat));
+}
+
+static unsigned int bdw_private_pat_match(u8 src, u8 dst)
+{
+	unsigned int score = 0;
+	enum {
+		AGE_MATCH = BIT(0),
+		TC_MATCH = BIT(1),
+		CA_MATCH = BIT(2),
+	};
+
+	/* Cache attribute has to be matched. */
+	if (GEN8_PPAT_GET_CA(src) != GEN8_PPAT_GET_CA(dst))
+		return 0;
+
+	score |= CA_MATCH;
+
+	if (GEN8_PPAT_GET_TC(src) == GEN8_PPAT_GET_TC(dst))
+		score |= TC_MATCH;
+
+	if (GEN8_PPAT_GET_AGE(src) == GEN8_PPAT_GET_AGE(dst))
+		score |= AGE_MATCH;
+
+	if (score == (AGE_MATCH | TC_MATCH | CA_MATCH))
+		return INTEL_PPAT_PERFECT_MATCH;
+
+	return score;
+}
+
+static unsigned int chv_private_pat_match(u8 src, u8 dst)
+{
+	return (CHV_PPAT_GET_SNOOP(src) == CHV_PPAT_GET_SNOOP(dst)) ?
+		INTEL_PPAT_PERFECT_MATCH : 0;
+}
+
+static void cnl_setup_private_ppat(struct intel_ppat *ppat)
+{
+	ppat->max_entries = 8;
+	ppat->update_hw = cnl_private_pat_update_hw;
+	ppat->match = bdw_private_pat_match;
+	ppat->clear_value = GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(3);
+
 	/* XXX: spec is unclear if this is still needed for CNL+ */
-	if (!USES_PPGTT(dev_priv)) {
-		I915_WRITE(GEN10_PAT_INDEX(0), GEN8_PPAT_UC);
+	if (!USES_PPGTT(ppat->i915)) {
+		__alloc_ppat_entry(ppat, 0, GEN8_PPAT_UC);
 		return;
 	}
 
-	I915_WRITE(GEN10_PAT_INDEX(0), GEN8_PPAT_WB | GEN8_PPAT_LLC);
-	I915_WRITE(GEN10_PAT_INDEX(1), GEN8_PPAT_WC | GEN8_PPAT_LLCELLC);
-	I915_WRITE(GEN10_PAT_INDEX(2), GEN8_PPAT_WT | GEN8_PPAT_LLCELLC);
-	I915_WRITE(GEN10_PAT_INDEX(3), GEN8_PPAT_UC);
-	I915_WRITE(GEN10_PAT_INDEX(4), GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(0));
-	I915_WRITE(GEN10_PAT_INDEX(5), GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(1));
-	I915_WRITE(GEN10_PAT_INDEX(6), GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(2));
-	I915_WRITE(GEN10_PAT_INDEX(7), GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(3));
+	__alloc_ppat_entry(ppat, 0, GEN8_PPAT_WB | GEN8_PPAT_LLC);
+	__alloc_ppat_entry(ppat, 1, GEN8_PPAT_WC | GEN8_PPAT_LLCELLC);
+	__alloc_ppat_entry(ppat, 2, GEN8_PPAT_WT | GEN8_PPAT_LLCELLC);
+	__alloc_ppat_entry(ppat, 3, GEN8_PPAT_UC);
+	__alloc_ppat_entry(ppat, 4, GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(0));
+	__alloc_ppat_entry(ppat, 5, GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(1));
+	__alloc_ppat_entry(ppat, 6, GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(2));
+	__alloc_ppat_entry(ppat, 7, GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(3));
 }
 
 /* The GGTT and PPGTT need a private PPAT setup in order to handle cacheability
  * bits. When using advanced contexts each context stores its own PAT, but
  * writing this data shouldn't be harmful even in those cases. */
-static void bdw_setup_private_ppat(struct drm_i915_private *dev_priv)
+static void bdw_setup_private_ppat(struct intel_ppat *ppat)
 {
-	u64 pat;
+	ppat->max_entries = 8;
+	ppat->update_hw = bdw_private_pat_update_hw;
+	ppat->match = bdw_private_pat_match;
+	ppat->clear_value = GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(3);
 
-	pat = GEN8_PPAT(0, GEN8_PPAT_WB | GEN8_PPAT_LLC)     | /* for normal objects, no eLLC */
-	      GEN8_PPAT(1, GEN8_PPAT_WC | GEN8_PPAT_LLCELLC) | /* for something pointing to ptes? */
-	      GEN8_PPAT(2, GEN8_PPAT_WT | GEN8_PPAT_LLCELLC) | /* for scanout with eLLC */
-	      GEN8_PPAT(3, GEN8_PPAT_UC)                     | /* Uncached objects, mostly for scanout */
-	      GEN8_PPAT(4, GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(0)) |
-	      GEN8_PPAT(5, GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(1)) |
-	      GEN8_PPAT(6, GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(2)) |
-	      GEN8_PPAT(7, GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(3));
-
-	if (!USES_PPGTT(dev_priv))
+	if (!USES_PPGTT(ppat->i915)) {
 		/* Spec: "For GGTT, there is NO pat_sel[2:0] from the entry,
 		 * so RTL will always use the value corresponding to
 		 * pat_sel = 000".
@@ -2864,17 +3033,26 @@ static void bdw_setup_private_ppat(struct drm_i915_private *dev_priv)
 		 * So we can still hold onto all our assumptions wrt cpu
 		 * clflushing on LLC machines.
 		 */
-		pat = GEN8_PPAT(0, GEN8_PPAT_UC);
+		__alloc_ppat_entry(ppat, 0, GEN8_PPAT_UC);
+		return;
+	}
 
-	/* XXX: spec defines this as 2 distinct registers. It's unclear if a 64b
-	 * write would work. */
-	I915_WRITE(GEN8_PRIVATE_PAT_LO, pat);
-	I915_WRITE(GEN8_PRIVATE_PAT_HI, pat >> 32);
+	__alloc_ppat_entry(ppat, 0, GEN8_PPAT_WB | GEN8_PPAT_LLC);      /* for normal objects, no eLLC */
+	__alloc_ppat_entry(ppat, 1, GEN8_PPAT_WC | GEN8_PPAT_LLCELLC);  /* for something pointing to ptes? */
+	__alloc_ppat_entry(ppat, 2, GEN8_PPAT_WT | GEN8_PPAT_LLCELLC);  /* for scanout with eLLC */
+	__alloc_ppat_entry(ppat, 3, GEN8_PPAT_UC);                      /* Uncached objects, mostly for scanout */
+	__alloc_ppat_entry(ppat, 4, GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(0));
+	__alloc_ppat_entry(ppat, 5, GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(1));
+	__alloc_ppat_entry(ppat, 6, GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(2));
+	__alloc_ppat_entry(ppat, 7, GEN8_PPAT_WB | GEN8_PPAT_LLCELLC | GEN8_PPAT_AGE(3));
 }
 
-static void chv_setup_private_ppat(struct drm_i915_private *dev_priv)
+static void chv_setup_private_ppat(struct intel_ppat *ppat)
 {
-	u64 pat;
+	ppat->max_entries = 8;
+	ppat->update_hw = bdw_private_pat_update_hw;
+	ppat->match = chv_private_pat_match;
+	ppat->clear_value = CHV_PPAT_SNOOP;
 
 	/*
 	 * Map WB on BDW to snooped on CHV.
@@ -2894,17 +3072,15 @@ static void chv_setup_private_ppat(struct drm_i915_private *dev_priv)
 	 * Which means we must set the snoop bit in PAT entry 0
 	 * in order to keep the global status page working.
 	 */
-	pat = GEN8_PPAT(0, CHV_PPAT_SNOOP) |
-	      GEN8_PPAT(1, 0) |
-	      GEN8_PPAT(2, 0) |
-	      GEN8_PPAT(3, 0) |
-	      GEN8_PPAT(4, CHV_PPAT_SNOOP) |
-	      GEN8_PPAT(5, CHV_PPAT_SNOOP) |
-	      GEN8_PPAT(6, CHV_PPAT_SNOOP) |
-	      GEN8_PPAT(7, CHV_PPAT_SNOOP);
 
-	I915_WRITE(GEN8_PRIVATE_PAT_LO, pat);
-	I915_WRITE(GEN8_PRIVATE_PAT_HI, pat >> 32);
+	__alloc_ppat_entry(ppat, 0, CHV_PPAT_SNOOP);
+	__alloc_ppat_entry(ppat, 1, 0);
+	__alloc_ppat_entry(ppat, 2, 0);
+	__alloc_ppat_entry(ppat, 3, 0);
+	__alloc_ppat_entry(ppat, 4, CHV_PPAT_SNOOP);
+	__alloc_ppat_entry(ppat, 5, CHV_PPAT_SNOOP);
+	__alloc_ppat_entry(ppat, 6, CHV_PPAT_SNOOP);
+	__alloc_ppat_entry(ppat, 7, CHV_PPAT_SNOOP);
 }
 
 static void gen6_gmch_remove(struct i915_address_space *vm)
@@ -2915,6 +3091,31 @@ static void gen6_gmch_remove(struct i915_address_space *vm)
 	cleanup_scratch_page(vm);
 }
 
+static void setup_private_pat(struct drm_i915_private *dev_priv)
+{
+	struct intel_ppat *ppat = &dev_priv->ppat;
+	int i;
+
+	ppat->i915 = dev_priv;
+
+	if (INTEL_GEN(dev_priv) >= 10)
+		cnl_setup_private_ppat(ppat);
+	else if (IS_CHERRYVIEW(dev_priv) || IS_GEN9_LP(dev_priv))
+		chv_setup_private_ppat(ppat);
+	else
+		bdw_setup_private_ppat(ppat);
+
+	GEM_BUG_ON(ppat->max_entries > INTEL_MAX_PPAT_ENTRIES);
+
+	for_each_clear_bit(i, ppat->used, ppat->max_entries) {
+		ppat->entries[i].value = ppat->clear_value;
+		ppat->entries[i].ppat = ppat;
+		set_bit(i, ppat->dirty);
+	}
+
+	ppat->update_hw(dev_priv);
+}
+
 static int gen8_gmch_probe(struct i915_ggtt *ggtt)
 {
 	struct drm_i915_private *dev_priv = ggtt->base.i915;
@@ -2947,14 +3148,6 @@ static int gen8_gmch_probe(struct i915_ggtt *ggtt)
 	}
 
 	ggtt->base.total = (size / sizeof(gen8_pte_t)) << PAGE_SHIFT;
-
-	if (INTEL_GEN(dev_priv) >= 10)
-		cnl_setup_private_ppat(dev_priv);
-	else if (IS_CHERRYVIEW(dev_priv) || IS_GEN9_LP(dev_priv))
-		chv_setup_private_ppat(dev_priv);
-	else
-		bdw_setup_private_ppat(dev_priv);
-
 	ggtt->base.cleanup = gen6_gmch_remove;
 	ggtt->base.bind_vma = ggtt_bind_vma;
 	ggtt->base.unbind_vma = ggtt_unbind_vma;
@@ -2975,6 +3168,8 @@ static int gen8_gmch_probe(struct i915_ggtt *ggtt)
 
 	ggtt->invalidate = gen6_ggtt_invalidate;
 
+	setup_private_pat(dev_priv);
+
 	return ggtt_probe_common(ggtt, size);
 }
 
@@ -3095,7 +3290,7 @@ int i915_ggtt_probe_hw(struct drm_i915_private *dev_priv)
 	 * currently don't have any bits spare to pass in this upper
 	 * restriction!
 	 */
-	if (HAS_GUC(dev_priv) && i915.enable_guc_loading) {
+	if (HAS_GUC(dev_priv) && i915_modparams.enable_guc_loading) {
 		ggtt->base.total = min_t(u64, ggtt->base.total, GUC_GGTT_TOP);
 		ggtt->mappable_end = min(ggtt->mappable_end, ggtt->base.total);
 	}
@@ -3232,13 +3427,10 @@ void i915_gem_restore_gtt_mappings(struct drm_i915_private *dev_priv)
 	ggtt->base.closed = false;
 
 	if (INTEL_GEN(dev_priv) >= 8) {
-		if (INTEL_GEN(dev_priv) >= 10)
-			cnl_setup_private_ppat(dev_priv);
-		else if (IS_CHERRYVIEW(dev_priv) || IS_GEN9_LP(dev_priv))
-			chv_setup_private_ppat(dev_priv);
-		else
-			bdw_setup_private_ppat(dev_priv);
+		struct intel_ppat *ppat = &dev_priv->ppat;
 
+		bitmap_set(ppat->dirty, 0, ppat->max_entries);
+		dev_priv->ppat.update_hw(dev_priv);
 		return;
 	}
 

drivers/gpu/drm/i915/i915_gem_gtt.h

@@ -126,13 +126,13 @@ typedef u64 gen8_ppgtt_pml4e_t;
  * tables */
 #define GEN8_PDPE_MASK			0x1ff
 
-#define PPAT_UNCACHED_INDEX		(_PAGE_PWT | _PAGE_PCD)
-#define PPAT_CACHED_PDE_INDEX		0 /* WB LLC */
-#define PPAT_CACHED_INDEX		_PAGE_PAT /* WB LLCeLLC */
-#define PPAT_DISPLAY_ELLC_INDEX		_PAGE_PCD /* WT eLLC */
+#define PPAT_UNCACHED			(_PAGE_PWT | _PAGE_PCD)
+#define PPAT_CACHED_PDE			0 /* WB LLC */
+#define PPAT_CACHED			_PAGE_PAT /* WB LLCeLLC */
+#define PPAT_DISPLAY_ELLC		_PAGE_PCD /* WT eLLC */
 
 #define CHV_PPAT_SNOOP			(1<<6)
-#define GEN8_PPAT_AGE(x)		(x<<4)
+#define GEN8_PPAT_AGE(x)		((x)<<4)
 #define GEN8_PPAT_LLCeLLC		(3<<2)
 #define GEN8_PPAT_LLCELLC		(2<<2)
 #define GEN8_PPAT_LLC			(1<<2)
@@ -143,6 +143,11 @@ typedef u64 gen8_ppgtt_pml4e_t;
 #define GEN8_PPAT_ELLC_OVERRIDE		(0<<2)
 #define GEN8_PPAT(i, x)			((u64)(x) << ((i) * 8))
 
+#define GEN8_PPAT_GET_CA(x) ((x) & 3)
+#define GEN8_PPAT_GET_TC(x) ((x) & (3 << 2))
+#define GEN8_PPAT_GET_AGE(x) ((x) & (3 << 4))
+#define CHV_PPAT_GET_SNOOP(x) ((x) & (1 << 6))
+
 struct sg_table;
 
 struct intel_rotation_info {
@@ -536,6 +541,37 @@ i915_vm_to_ggtt(struct i915_address_space *vm)
 	return container_of(vm, struct i915_ggtt, base);
 }
 
+#define INTEL_MAX_PPAT_ENTRIES 8
+#define INTEL_PPAT_PERFECT_MATCH (~0U)
+
+struct intel_ppat;
+
+struct intel_ppat_entry {
+	struct intel_ppat *ppat;
+	struct kref ref;
+	u8 value;
+};
+
+struct intel_ppat {
+	struct intel_ppat_entry entries[INTEL_MAX_PPAT_ENTRIES];
+	DECLARE_BITMAP(used, INTEL_MAX_PPAT_ENTRIES);
+	DECLARE_BITMAP(dirty, INTEL_MAX_PPAT_ENTRIES);
+	unsigned int max_entries;
+	u8 clear_value;
+	/*
+	 * Return a score to show how two PPAT values match,
+	 * a INTEL_PPAT_PERFECT_MATCH indicates a perfect match
+	 */
+	unsigned int (*match)(u8 src, u8 dst);
+	void (*update_hw)(struct drm_i915_private *i915);
+
+	struct drm_i915_private *i915;
+};
+
+const struct intel_ppat_entry *
+intel_ppat_get(struct drm_i915_private *i915, u8 value);
+void intel_ppat_put(const struct intel_ppat_entry *entry);
+
 int i915_gem_init_aliasing_ppgtt(struct drm_i915_private *i915);
 void i915_gem_fini_aliasing_ppgtt(struct drm_i915_private *i915);
 

drivers/gpu/drm/i915/i915_gem_request.c

@@ -1021,12 +1021,28 @@ static bool busywait_stop(unsigned long timeout, unsigned int cpu)
 	return this_cpu != cpu;
 }
 
-bool __i915_spin_request(const struct drm_i915_gem_request *req,
-			 u32 seqno, int state, unsigned long timeout_us)
+static bool __i915_spin_request(const struct drm_i915_gem_request *req,
+				u32 seqno, int state, unsigned long timeout_us)
 {
 	struct intel_engine_cs *engine = req->engine;
 	unsigned int irq, cpu;
 
+	GEM_BUG_ON(!seqno);
+
+	/*
+	 * Only wait for the request if we know it is likely to complete.
+	 *
+	 * We don't track the timestamps around requests, nor the average
+	 * request length, so we do not have a good indicator that this
+	 * request will complete within the timeout. What we do know is the
+	 * order in which requests are executed by the engine and so we can
+	 * tell if the request has started. If the request hasn't started yet,
+	 * it is a fair assumption that it will not complete within our
+	 * relatively short timeout.
+	 */
+	if (!i915_seqno_passed(intel_engine_get_seqno(engine), seqno - 1))
+		return false;
+
 	/* When waiting for high frequency requests, e.g. during synchronous
 	 * rendering split between the CPU and GPU, the finite amount of time
 	 * required to set up the irq and wait upon it limits the response
@@ -1040,12 +1056,8 @@ bool __i915_spin_request(const struct drm_i915_gem_request *req,
 	irq = atomic_read(&engine->irq_count);
 	timeout_us += local_clock_us(&cpu);
 	do {
-		if (seqno != i915_gem_request_global_seqno(req))
-			break;
-
-		if (i915_seqno_passed(intel_engine_get_seqno(req->engine),
-				      seqno))
-			return true;
+		if (i915_seqno_passed(intel_engine_get_seqno(engine), seqno))
+			return seqno == i915_gem_request_global_seqno(req);
 
 		/* Seqno are meant to be ordered *before* the interrupt. If
 		 * we see an interrupt without a corresponding seqno advance,
@@ -1156,7 +1168,7 @@ restart:
 	GEM_BUG_ON(!i915_sw_fence_signaled(&req->submit));
 
 	/* Optimistic short spin before touching IRQs */
-	if (i915_spin_request(req, state, 5))
+	if (__i915_spin_request(req, wait.seqno, state, 5))
 		goto complete;
 
 	set_current_state(state);
@@ -1213,7 +1225,7 @@ wakeup:
 			continue;
 
 		/* Only spin if we know the GPU is processing this request */
-		if (i915_spin_request(req, state, 2))
+		if (__i915_spin_request(req, wait.seqno, state, 2))
 			break;
 
 		if (!intel_wait_check_request(&wait, req)) {

drivers/gpu/drm/i915/i915_gem_request.h

@@ -312,26 +312,6 @@ static inline bool i915_seqno_passed(u32 seq1, u32 seq2)
 	return (s32)(seq1 - seq2) >= 0;
 }
 
-static inline bool
-__i915_gem_request_started(const struct drm_i915_gem_request *req, u32 seqno)
-{
-	GEM_BUG_ON(!seqno);
-	return i915_seqno_passed(intel_engine_get_seqno(req->engine),
-				 seqno - 1);
-}
-
-static inline bool
-i915_gem_request_started(const struct drm_i915_gem_request *req)
-{
-	u32 seqno;
-
-	seqno = i915_gem_request_global_seqno(req);
-	if (!seqno)
-		return false;
-
-	return __i915_gem_request_started(req, seqno);
-}
-
 static inline bool
 __i915_gem_request_completed(const struct drm_i915_gem_request *req, u32 seqno)
 {
@@ -352,21 +332,6 @@ i915_gem_request_completed(const struct drm_i915_gem_request *req)
 	return __i915_gem_request_completed(req, seqno);
 }
 
-bool __i915_spin_request(const struct drm_i915_gem_request *request,
-			 u32 seqno, int state, unsigned long timeout_us);
-static inline bool i915_spin_request(const struct drm_i915_gem_request *request,
-				     int state, unsigned long timeout_us)
-{
-	u32 seqno;
-
-	seqno = i915_gem_request_global_seqno(request);
-	if (!seqno)
-		return 0;
-
-	return (__i915_gem_request_started(request, seqno) &&
-		__i915_spin_request(request, seqno, state, timeout_us));
-}
-
 /* We treat requests as fences. This is not be to confused with our
  * "fence registers" but pipeline synchronisation objects ala GL_ARB_sync.
  * We use the fences to synchronize access from the CPU with activity on the

drivers/gpu/drm/i915/i915_gem_userptr.c

@@ -399,64 +399,42 @@ struct get_pages_work {
 	struct task_struct *task;
 };
 
-#if IS_ENABLED(CONFIG_SWIOTLB)
-#define swiotlb_active() swiotlb_nr_tbl()
-#else
-#define swiotlb_active() 0
-#endif
-
-static int
-st_set_pages(struct sg_table **st, struct page **pvec, int num_pages)
-{
-	struct scatterlist *sg;
-	int ret, n;
-
-	*st = kmalloc(sizeof(**st), GFP_KERNEL);
-	if (*st == NULL)
-		return -ENOMEM;
-
-	if (swiotlb_active()) {
-		ret = sg_alloc_table(*st, num_pages, GFP_KERNEL);
-		if (ret)
-			goto err;
-
-		for_each_sg((*st)->sgl, sg, num_pages, n)
-			sg_set_page(sg, pvec[n], PAGE_SIZE, 0);
-	} else {
-		ret = sg_alloc_table_from_pages(*st, pvec, num_pages,
-						0, num_pages << PAGE_SHIFT,
-						GFP_KERNEL);
-		if (ret)
-			goto err;
-	}
-
-	return 0;
-
-err:
-	kfree(*st);
-	*st = NULL;
-	return ret;
-}
-
 static struct sg_table *
-__i915_gem_userptr_set_pages(struct drm_i915_gem_object *obj,
-			     struct page **pvec, int num_pages)
+__i915_gem_userptr_alloc_pages(struct drm_i915_gem_object *obj,
+			       struct page **pvec, int num_pages)
 {
-	struct sg_table *pages;
+	unsigned int max_segment = i915_sg_segment_size();
+	struct sg_table *st;
 	int ret;
 
-	ret = st_set_pages(&pages, pvec, num_pages);
-	if (ret)
+	st = kmalloc(sizeof(*st), GFP_KERNEL);
+	if (!st)
+		return ERR_PTR(-ENOMEM);
+
+alloc_table:
+	ret = __sg_alloc_table_from_pages(st, pvec, num_pages,
+					  0, num_pages << PAGE_SHIFT,
+					  max_segment,
+					  GFP_KERNEL);
+	if (ret) {
+		kfree(st);
 		return ERR_PTR(ret);
+	}
 
-	ret = i915_gem_gtt_prepare_pages(obj, pages);
+	ret = i915_gem_gtt_prepare_pages(obj, st);
 	if (ret) {
-		sg_free_table(pages);
-		kfree(pages);
+		sg_free_table(st);
+
+		if (max_segment > PAGE_SIZE) {
+			max_segment = PAGE_SIZE;
+			goto alloc_table;
+		}
+
+		kfree(st);
 		return ERR_PTR(ret);
 	}
 
-	return pages;
+	return st;
 }
 
 static int
@@ -540,7 +518,8 @@ __i915_gem_userptr_get_pages_worker(struct work_struct *_work)
 		struct sg_table *pages = ERR_PTR(ret);
 
 		if (pinned == npages) {
-			pages = __i915_gem_userptr_set_pages(obj, pvec, npages);
+			pages = __i915_gem_userptr_alloc_pages(obj, pvec,
+							       npages);
 			if (!IS_ERR(pages)) {
 				__i915_gem_object_set_pages(obj, pages);
 				pinned = 0;
@@ -661,7 +640,7 @@ i915_gem_userptr_get_pages(struct drm_i915_gem_object *obj)
 		pages = __i915_gem_userptr_get_pages_schedule(obj);
 		active = pages == ERR_PTR(-EAGAIN);
 	} else {
-		pages = __i915_gem_userptr_set_pages(obj, pvec, num_pages);
+		pages = __i915_gem_userptr_alloc_pages(obj, pvec, num_pages);
 		active = !IS_ERR(pages);
 	}
 	if (active)
@@ -834,7 +813,9 @@ int i915_gem_init_userptr(struct drm_i915_private *dev_priv)
 	hash_init(dev_priv->mm_structs);
 
 	dev_priv->mm.userptr_wq =
-		alloc_workqueue("i915-userptr-acquire", WQ_HIGHPRI, 0);
+		alloc_workqueue("i915-userptr-acquire",
+				WQ_HIGHPRI | WQ_MEM_RECLAIM,
+				0);
 	if (!dev_priv->mm.userptr_wq)
 		return -ENOMEM;
 

drivers/gpu/drm/i915/i915_gpu_error.c

@@ -396,6 +396,8 @@ static void error_print_context(struct drm_i915_error_state_buf *m,
 static void error_print_engine(struct drm_i915_error_state_buf *m,
 			       const struct drm_i915_error_engine *ee)
 {
+	int n;
+
 	err_printf(m, "%s command stream:\n", engine_str(ee->engine_id));
 	err_printf(m, "  START: 0x%08x\n", ee->start);
 	err_printf(m, "  HEAD:  0x%08x [0x%08x]\n", ee->head, ee->rq_head);
@@ -465,8 +467,11 @@ static void error_print_engine(struct drm_i915_error_state_buf *m,
 		   jiffies_to_msecs(jiffies - ee->hangcheck_timestamp));
 	err_printf(m, "  engine reset count: %u\n", ee->reset_count);
 
-	error_print_request(m, "  ELSP[0]: ", &ee->execlist[0]);
-	error_print_request(m, "  ELSP[1]: ", &ee->execlist[1]);
+	for (n = 0; n < ee->num_ports; n++) {
+		err_printf(m, "  ELSP[%d]:", n);
+		error_print_request(m, " ", &ee->execlist[n]);
+	}
+
 	error_print_context(m, "  Active context: ", &ee->context);
 }
 
@@ -567,7 +572,7 @@ static __always_inline void err_print_param(struct drm_i915_error_state_buf *m,
 static void err_print_params(struct drm_i915_error_state_buf *m,
 			     const struct i915_params *p)
 {
-#define PRINT(T, x) err_print_param(m, #x, #T, &p->x);
+#define PRINT(T, x, ...) err_print_param(m, #x, #T, &p->x);
 	I915_PARAMS_FOR_EACH(PRINT);
 #undef PRINT
 }
@@ -861,7 +866,7 @@ void __i915_gpu_state_free(struct kref *error_ref)
 	kfree(error->overlay);
 	kfree(error->display);
 
-#define FREE(T, x) free_param(#T, &error->params.x);
+#define FREE(T, x, ...) free_param(#T, &error->params.x);
 	I915_PARAMS_FOR_EACH(FREE);
 #undef FREE
 
@@ -1327,17 +1332,19 @@ static void engine_record_requests(struct intel_engine_cs *engine,
 static void error_record_engine_execlists(struct intel_engine_cs *engine,
 					  struct drm_i915_error_engine *ee)
 {
-	const struct execlist_port *port = engine->execlist_port;
+	const struct intel_engine_execlists * const execlists = &engine->execlists;
 	unsigned int n;
 
-	for (n = 0; n < ARRAY_SIZE(engine->execlist_port); n++) {
-		struct drm_i915_gem_request *rq = port_request(&port[n]);
+	for (n = 0; n < execlists_num_ports(execlists); n++) {
+		struct drm_i915_gem_request *rq = port_request(&execlists->port[n]);
 
 		if (!rq)
 			break;
 
 		record_request(rq, &ee->execlist[n]);
 	}
+
+	ee->num_ports = n;
 }
 
 static void record_context(struct drm_i915_error_context *e,
@@ -1554,7 +1561,7 @@ static void i915_gem_capture_guc_log_buffer(struct drm_i915_private *dev_priv,
 					    struct i915_gpu_state *error)
 {
 	/* Capturing log buf contents won't be useful if logging was disabled */
-	if (!dev_priv->guc.log.vma || (i915.guc_log_level < 0))
+	if (!dev_priv->guc.log.vma || (i915_modparams.guc_log_level < 0))
 		return;
 
 	error->guc_log = i915_error_object_create(dev_priv,
@@ -1696,8 +1703,8 @@ static int capture(void *data)
 		ktime_to_timeval(ktime_sub(ktime_get(),
 					   error->i915->gt.last_init_time));
 
-	error->params = i915;
-#define DUP(T, x) dup_param(#T, &error->params.x);
+	error->params = i915_modparams;
+#define DUP(T, x, ...) dup_param(#T, &error->params.x);
 	I915_PARAMS_FOR_EACH(DUP);
 #undef DUP
 
@@ -1751,7 +1758,7 @@ void i915_capture_error_state(struct drm_i915_private *dev_priv,
 	struct i915_gpu_state *error;
 	unsigned long flags;
 
-	if (!i915.error_capture)
+	if (!i915_modparams.error_capture)
 		return;
 
 	if (READ_ONCE(dev_priv->gpu_error.first_error))

drivers/gpu/drm/i915/i915_guc_submission.c

@@ -192,13 +192,12 @@ static int __create_doorbell(struct i915_guc_client *client)
 
 	doorbell = __get_doorbell(client);
 	doorbell->db_status = GUC_DOORBELL_ENABLED;
-	doorbell->cookie = client->doorbell_cookie;
+	doorbell->cookie = 0;
 
 	err = __guc_allocate_doorbell(client->guc, client->stage_id);
-	if (err) {
+	if (err)
 		doorbell->db_status = GUC_DOORBELL_DISABLED;
-		doorbell->cookie = 0;
-	}
+
 	return err;
 }
 
@@ -306,7 +305,7 @@ static void guc_proc_desc_init(struct intel_guc *guc,
 	desc->db_base_addr = 0;
 
 	desc->stage_id = client->stage_id;
-	desc->wq_size_bytes = client->wq_size;
+	desc->wq_size_bytes = GUC_WQ_SIZE;
 	desc->wq_status = WQ_STATUS_ACTIVE;
 	desc->priority = client->priority;
 }
@@ -391,8 +390,8 @@ static void guc_stage_desc_init(struct intel_guc *guc,
 	desc->db_trigger_cpu = (uintptr_t)__get_doorbell(client);
 	desc->db_trigger_uk = gfx_addr + client->doorbell_offset;
 	desc->process_desc = gfx_addr + client->proc_desc_offset;
-	desc->wq_addr = gfx_addr + client->wq_offset;
-	desc->wq_size = client->wq_size;
+	desc->wq_addr = gfx_addr + GUC_DB_SIZE;
+	desc->wq_size = GUC_WQ_SIZE;
 
 	desc->desc_private = (uintptr_t)client;
 }
@@ -406,82 +405,23 @@ static void guc_stage_desc_fini(struct intel_guc *guc,
 	memset(desc, 0, sizeof(*desc));
 }
 
-/**
- * i915_guc_wq_reserve() - reserve space in the GuC's workqueue
- * @request:	request associated with the commands
- *
- * Return:	0 if space is available
- *		-EAGAIN if space is not currently available
- *
- * This function must be called (and must return 0) before a request
- * is submitted to the GuC via i915_guc_submit() below. Once a result
- * of 0 has been returned, it must be balanced by a corresponding
- * call to submit().
- *
- * Reservation allows the caller to determine in advance that space
- * will be available for the next submission before committing resources
- * to it, and helps avoid late failures with complicated recovery paths.
- */
-int i915_guc_wq_reserve(struct drm_i915_gem_request *request)
-{
-	const size_t wqi_size = sizeof(struct guc_wq_item);
-	struct i915_guc_client *client = request->i915->guc.execbuf_client;
-	struct guc_process_desc *desc = __get_process_desc(client);
-	u32 freespace;
-	int ret;
-
-	spin_lock_irq(&client->wq_lock);
-	freespace = CIRC_SPACE(client->wq_tail, desc->head, client->wq_size);
-	freespace -= client->wq_rsvd;
-	if (likely(freespace >= wqi_size)) {
-		client->wq_rsvd += wqi_size;
-		ret = 0;
-	} else {
-		client->no_wq_space++;
-		ret = -EAGAIN;
-	}
-	spin_unlock_irq(&client->wq_lock);
-
-	return ret;
-}
-
-static void guc_client_update_wq_rsvd(struct i915_guc_client *client, int size)
-{
-	unsigned long flags;
-
-	spin_lock_irqsave(&client->wq_lock, flags);
-	client->wq_rsvd += size;
-	spin_unlock_irqrestore(&client->wq_lock, flags);
-}
-
-void i915_guc_wq_unreserve(struct drm_i915_gem_request *request)
-{
-	const int wqi_size = sizeof(struct guc_wq_item);
-	struct i915_guc_client *client = request->i915->guc.execbuf_client;
-
-	GEM_BUG_ON(READ_ONCE(client->wq_rsvd) < wqi_size);
-	guc_client_update_wq_rsvd(client, -wqi_size);
-}
-
 /* Construct a Work Item and append it to the GuC's Work Queue */
 static void guc_wq_item_append(struct i915_guc_client *client,
 			       struct drm_i915_gem_request *rq)
 {
 	/* wqi_len is in DWords, and does not include the one-word header */
 	const size_t wqi_size = sizeof(struct guc_wq_item);
-	const u32 wqi_len = wqi_size/sizeof(u32) - 1;
+	const u32 wqi_len = wqi_size / sizeof(u32) - 1;
 	struct intel_engine_cs *engine = rq->engine;
+	struct i915_gem_context *ctx = rq->ctx;
 	struct guc_process_desc *desc = __get_process_desc(client);
 	struct guc_wq_item *wqi;
-	u32 freespace, tail, wq_off;
+	u32 ring_tail, wq_off;
 
-	/* Free space is guaranteed, see i915_guc_wq_reserve() above */
-	freespace = CIRC_SPACE(client->wq_tail, desc->head, client->wq_size);
-	GEM_BUG_ON(freespace < wqi_size);
+	lockdep_assert_held(&client->wq_lock);
 
-	/* The GuC firmware wants the tail index in QWords, not bytes */
-	tail = intel_ring_set_tail(rq->ring, rq->tail) >> 3;
-	GEM_BUG_ON(tail > WQ_RING_TAIL_MAX);
+	ring_tail = intel_ring_set_tail(rq->ring, rq->tail) / sizeof(u64);
+	GEM_BUG_ON(ring_tail > WQ_RING_TAIL_MAX);
 
 	/* For now workqueue item is 4 DWs; workqueue buffer is 2 pages. So we
 	 * should not have the case where structure wqi is across page, neither
@@ -491,29 +431,29 @@ static void guc_wq_item_append(struct i915_guc_client *client,
 	 * workqueue buffer dw by dw.
 	 */
 	BUILD_BUG_ON(wqi_size != 16);
-	GEM_BUG_ON(client->wq_rsvd < wqi_size);
 
-	/* postincrement WQ tail for next time */
-	wq_off = client->wq_tail;
+	/* Free space is guaranteed. */
+	wq_off = READ_ONCE(desc->tail);
+	GEM_BUG_ON(CIRC_SPACE(wq_off, READ_ONCE(desc->head),
+			      GUC_WQ_SIZE) < wqi_size);
 	GEM_BUG_ON(wq_off & (wqi_size - 1));
-	client->wq_tail += wqi_size;
-	client->wq_tail &= client->wq_size - 1;
-	client->wq_rsvd -= wqi_size;
 
 	/* WQ starts from the page after doorbell / process_desc */
 	wqi = client->vaddr + wq_off + GUC_DB_SIZE;
 
 	/* Now fill in the 4-word work queue item */
 	wqi->header = WQ_TYPE_INORDER |
-			(wqi_len << WQ_LEN_SHIFT) |
-			(engine->guc_id << WQ_TARGET_SHIFT) |
-			WQ_NO_WCFLUSH_WAIT;
+		      (wqi_len << WQ_LEN_SHIFT) |
+		      (engine->guc_id << WQ_TARGET_SHIFT) |
+		      WQ_NO_WCFLUSH_WAIT;
 
-	/* The GuC wants only the low-order word of the context descriptor */
-	wqi->context_desc = (u32)intel_lr_context_descriptor(rq->ctx, engine);
+	wqi->context_desc = lower_32_bits(intel_lr_context_descriptor(ctx, engine));
 
-	wqi->submit_element_info = tail << WQ_RING_TAIL_SHIFT;
+	wqi->submit_element_info = ring_tail << WQ_RING_TAIL_SHIFT;
 	wqi->fence_id = rq->global_seqno;
+
+	/* Postincrement WQ tail for next time. */
+	WRITE_ONCE(desc->tail, (wq_off + wqi_size) & (GUC_WQ_SIZE - 1));
 }
 
 static void guc_reset_wq(struct i915_guc_client *client)
@@ -522,106 +462,64 @@ static void guc_reset_wq(struct i915_guc_client *client)
 
 	desc->head = 0;
 	desc->tail = 0;
-
-	client->wq_tail = 0;
 }
 
-static int guc_ring_doorbell(struct i915_guc_client *client)
+static void guc_ring_doorbell(struct i915_guc_client *client)
 {
-	struct guc_process_desc *desc = __get_process_desc(client);
-	union guc_doorbell_qw db_cmp, db_exc, db_ret;
-	union guc_doorbell_qw *db;
-	int attempt = 2, ret = -EAGAIN;
-
-	/* Update the tail so it is visible to GuC */
-	desc->tail = client->wq_tail;
-
-	/* current cookie */
-	db_cmp.db_status = GUC_DOORBELL_ENABLED;
-	db_cmp.cookie = client->doorbell_cookie;
+	struct guc_doorbell_info *db;
+	u32 cookie;
 
-	/* cookie to be updated */
-	db_exc.db_status = GUC_DOORBELL_ENABLED;
-	db_exc.cookie = client->doorbell_cookie + 1;
-	if (db_exc.cookie == 0)
-		db_exc.cookie = 1;
+	lockdep_assert_held(&client->wq_lock);
 
 	/* pointer of current doorbell cacheline */
-	db = (union guc_doorbell_qw *)__get_doorbell(client);
-
-	while (attempt--) {
-		/* lets ring the doorbell */
-		db_ret.value_qw = atomic64_cmpxchg((atomic64_t *)db,
-			db_cmp.value_qw, db_exc.value_qw);
-
-		/* if the exchange was successfully executed */
-		if (db_ret.value_qw == db_cmp.value_qw) {
-			/* db was successfully rung */
-			client->doorbell_cookie = db_exc.cookie;
-			ret = 0;
-			break;
-		}
-
-		/* XXX: doorbell was lost and need to acquire it again */
-		if (db_ret.db_status == GUC_DOORBELL_DISABLED)
-			break;
+	db = __get_doorbell(client);
 
-		DRM_WARN("Cookie mismatch. Expected %d, found %d\n",
-			 db_cmp.cookie, db_ret.cookie);
-
-		/* update the cookie to newly read cookie from GuC */
-		db_cmp.cookie = db_ret.cookie;
-		db_exc.cookie = db_ret.cookie + 1;
-		if (db_exc.cookie == 0)
-			db_exc.cookie = 1;
-	}
+	/* we're not expecting the doorbell cookie to change behind our back */
+	cookie = READ_ONCE(db->cookie);
+	WARN_ON_ONCE(xchg(&db->cookie, cookie + 1) != cookie);
 
-	return ret;
+	/* XXX: doorbell was lost and need to acquire it again */
+	GEM_BUG_ON(db->db_status != GUC_DOORBELL_ENABLED);
 }
 
 /**
- * __i915_guc_submit() - Submit commands through GuC
- * @rq:		request associated with the commands
- *
- * The caller must have already called i915_guc_wq_reserve() above with
- * a result of 0 (success), guaranteeing that there is space in the work
- * queue for the new request, so enqueuing the item cannot fail.
- *
- * Bad Things Will Happen if the caller violates this protocol e.g. calls
- * submit() when _reserve() says there's no space, or calls _submit()
- * a different number of times from (successful) calls to _reserve().
+ * i915_guc_submit() - Submit commands through GuC
+ * @engine: engine associated with the commands
  *
  * The only error here arises if the doorbell hardware isn't functioning
  * as expected, which really shouln't happen.
  */
-static void __i915_guc_submit(struct drm_i915_gem_request *rq)
+static void i915_guc_submit(struct intel_engine_cs *engine)
 {
-	struct drm_i915_private *dev_priv = rq->i915;
-	struct intel_engine_cs *engine = rq->engine;
-	unsigned int engine_id = engine->id;
-	struct intel_guc *guc = &rq->i915->guc;
+	struct drm_i915_private *dev_priv = engine->i915;
+	struct intel_guc *guc = &dev_priv->guc;
 	struct i915_guc_client *client = guc->execbuf_client;
-	unsigned long flags;
-	int b_ret;
+	struct intel_engine_execlists * const execlists = &engine->execlists;
+	struct execlist_port *port = execlists->port;
+	const unsigned int engine_id = engine->id;
+	unsigned int n;
 
-	/* WA to flush out the pending GMADR writes to ring buffer. */
-	if (i915_vma_is_map_and_fenceable(rq->ring->vma))
-		POSTING_READ_FW(GUC_STATUS);
+	for (n = 0; n < ARRAY_SIZE(execlists->port); n++) {
+		struct drm_i915_gem_request *rq;
+		unsigned int count;
 
-	spin_lock_irqsave(&client->wq_lock, flags);
+		rq = port_unpack(&port[n], &count);
+		if (rq && count == 0) {
+			port_set(&port[n], port_pack(rq, ++count));
 
-	guc_wq_item_append(client, rq);
-	b_ret = guc_ring_doorbell(client);
+			if (i915_vma_is_map_and_fenceable(rq->ring->vma))
+				POSTING_READ_FW(GUC_STATUS);
 
-	client->submissions[engine_id] += 1;
+			spin_lock(&client->wq_lock);
 
-	spin_unlock_irqrestore(&client->wq_lock, flags);
-}
+			guc_wq_item_append(client, rq);
+			guc_ring_doorbell(client);
 
-static void i915_guc_submit(struct drm_i915_gem_request *rq)
-{
-	__i915_gem_request_submit(rq);
-	__i915_guc_submit(rq);
+			client->submissions[engine_id] += 1;
+
+			spin_unlock(&client->wq_lock);
+		}
+	}
 }
 
 static void nested_enable_signaling(struct drm_i915_gem_request *rq)
@@ -655,27 +553,33 @@ static void port_assign(struct execlist_port *port,
 	if (port_isset(port))
 		i915_gem_request_put(port_request(port));
 
-	port_set(port, i915_gem_request_get(rq));
+	port_set(port, port_pack(i915_gem_request_get(rq), port_count(port)));
 	nested_enable_signaling(rq);
 }
 
-static bool i915_guc_dequeue(struct intel_engine_cs *engine)
+static void i915_guc_dequeue(struct intel_engine_cs *engine)
 {
-	struct execlist_port *port = engine->execlist_port;
-	struct drm_i915_gem_request *last = port_request(port);
-	struct rb_node *rb;
+	struct intel_engine_execlists * const execlists = &engine->execlists;
+	struct execlist_port *port = execlists->port;
+	struct drm_i915_gem_request *last = NULL;
+	const struct execlist_port * const last_port =
+		&execlists->port[execlists->port_mask];
 	bool submit = false;
+	struct rb_node *rb;
+
+	if (port_isset(port))
+		port++;
 
 	spin_lock_irq(&engine->timeline->lock);
-	rb = engine->execlist_first;
-	GEM_BUG_ON(rb_first(&engine->execlist_queue) != rb);
+	rb = execlists->first;
+	GEM_BUG_ON(rb_first(&execlists->queue) != rb);
 	while (rb) {
 		struct i915_priolist *p = rb_entry(rb, typeof(*p), node);
 		struct drm_i915_gem_request *rq, *rn;
 
 		list_for_each_entry_safe(rq, rn, &p->requests, priotree.link) {
 			if (last && rq->ctx != last->ctx) {
-				if (port != engine->execlist_port) {
+				if (port == last_port) {
 					__list_del_many(&p->requests,
 							&rq->priotree.link);
 					goto done;
@@ -689,50 +593,48 @@ static bool i915_guc_dequeue(struct intel_engine_cs *engine)
 			INIT_LIST_HEAD(&rq->priotree.link);
 			rq->priotree.priority = INT_MAX;
 
-			i915_guc_submit(rq);
-			trace_i915_gem_request_in(rq, port_index(port, engine));
+			__i915_gem_request_submit(rq);
+			trace_i915_gem_request_in(rq, port_index(port, execlists));
 			last = rq;
 			submit = true;
 		}
 
 		rb = rb_next(rb);
-		rb_erase(&p->node, &engine->execlist_queue);
+		rb_erase(&p->node, &execlists->queue);
 		INIT_LIST_HEAD(&p->requests);
 		if (p->priority != I915_PRIORITY_NORMAL)
 			kmem_cache_free(engine->i915->priorities, p);
 	}
 done:
-	engine->execlist_first = rb;
-	if (submit)
+	execlists->first = rb;
+	if (submit) {
 		port_assign(port, last);
+		i915_guc_submit(engine);
+	}
 	spin_unlock_irq(&engine->timeline->lock);
-
-	return submit;
 }
 
 static void i915_guc_irq_handler(unsigned long data)
 {
-	struct intel_engine_cs *engine = (struct intel_engine_cs *)data;
-	struct execlist_port *port = engine->execlist_port;
+	struct intel_engine_cs * const engine = (struct intel_engine_cs *)data;
+	struct intel_engine_execlists * const execlists = &engine->execlists;
+	struct execlist_port *port = execlists->port;
+	const struct execlist_port * const last_port =
+		&execlists->port[execlists->port_mask];
 	struct drm_i915_gem_request *rq;
-	bool submit;
 
-	do {
-		rq = port_request(&port[0]);
-		while (rq && i915_gem_request_completed(rq)) {
-			trace_i915_gem_request_out(rq);
-			i915_gem_request_put(rq);
+	rq = port_request(&port[0]);
+	while (rq && i915_gem_request_completed(rq)) {
+		trace_i915_gem_request_out(rq);
+		i915_gem_request_put(rq);
 
-			port[0] = port[1];
-			memset(&port[1], 0, sizeof(port[1]));
+		execlists_port_complete(execlists, port);
 
-			rq = port_request(&port[0]);
-		}
+		rq = port_request(&port[0]);
+	}
 
-		submit = false;
-		if (!port_count(&port[1]))
-			submit = i915_guc_dequeue(engine);
-	} while (submit);
+	if (!port_isset(last_port))
+		i915_guc_dequeue(engine);
 }
 
 /*
@@ -913,8 +815,6 @@ guc_client_alloc(struct drm_i915_private *dev_priv,
 	client->engines = engines;
 	client->priority = priority;
 	client->doorbell_id = GUC_DOORBELL_INVALID;
-	client->wq_offset = GUC_DB_SIZE;
-	client->wq_size = GUC_WQ_SIZE;
 	spin_lock_init(&client->wq_lock);
 
 	ret = ida_simple_get(&guc->stage_ids, 0, GUC_MAX_STAGE_DESCRIPTORS,
@@ -996,28 +896,39 @@ static void guc_client_free(struct i915_guc_client *client)
 	kfree(client);
 }
 
+static void guc_policy_init(struct guc_policy *policy)
+{
+	policy->execution_quantum = POLICY_DEFAULT_EXECUTION_QUANTUM_US;
+	policy->preemption_time = POLICY_DEFAULT_PREEMPTION_TIME_US;
+	policy->fault_time = POLICY_DEFAULT_FAULT_TIME_US;
+	policy->policy_flags = 0;
+}
+
 static void guc_policies_init(struct guc_policies *policies)
 {
 	struct guc_policy *policy;
 	u32 p, i;
 
-	policies->dpc_promote_time = 500000;
+	policies->dpc_promote_time = POLICY_DEFAULT_DPC_PROMOTE_TIME_US;
 	policies->max_num_work_items = POLICY_MAX_NUM_WI;
 
 	for (p = 0; p < GUC_CLIENT_PRIORITY_NUM; p++) {
 		for (i = GUC_RENDER_ENGINE; i < GUC_MAX_ENGINES_NUM; i++) {
 			policy = &policies->policy[p][i];
 
-			policy->execution_quantum = 1000000;
-			policy->preemption_time = 500000;
-			policy->fault_time = 250000;
-			policy->policy_flags = 0;
+			guc_policy_init(policy);
 		}
 	}
 
 	policies->is_valid = 1;
 }
 
+/*
+ * The first 80 dwords of the register state context, containing the
+ * execlists and ppgtt registers.
+ */
+#define LR_HW_CONTEXT_SIZE	(80 * sizeof(u32))
+
 static int guc_ads_create(struct intel_guc *guc)
 {
 	struct drm_i915_private *dev_priv = guc_to_i915(guc);
@@ -1032,6 +943,8 @@ static int guc_ads_create(struct intel_guc *guc)
 	} __packed *blob;
 	struct intel_engine_cs *engine;
 	enum intel_engine_id id;
+	const u32 skipped_offset = LRC_HEADER_PAGES * PAGE_SIZE;
+	const u32 skipped_size = LRC_PPHWSP_SZ * PAGE_SIZE + LR_HW_CONTEXT_SIZE;
 	u32 base;
 
 	GEM_BUG_ON(guc->ads_vma);
@@ -1062,13 +975,20 @@ static int guc_ads_create(struct intel_guc *guc)
 	 * engines after a reset. Here we use the Render ring default
 	 * context, which must already exist and be pinned in the GGTT,
 	 * so its address won't change after we've told the GuC where
-	 * to find it.
+	 * to find it. Note that we have to skip our header (1 page),
+	 * because our GuC shared data is there.
 	 */
 	blob->ads.golden_context_lrca =
-		dev_priv->engine[RCS]->status_page.ggtt_offset;
+		guc_ggtt_offset(dev_priv->kernel_context->engine[RCS].state) + skipped_offset;
 
+	/*
+	 * The GuC expects us to exclude the portion of the context image that
+	 * it skips from the size it is to read. It starts reading from after
+	 * the execlist context (so skipping the first page [PPHWSP] and 80
+	 * dwords). Weird guc is weird.
+	 */
 	for_each_engine(engine, dev_priv, id)
-		blob->ads.eng_state_size[engine->guc_id] = engine->context_size;
+		blob->ads.eng_state_size[engine->guc_id] = engine->context_size - skipped_size;
 
 	base = guc_ggtt_offset(vma);
 	blob->ads.scheduler_policies = base + ptr_offset(blob, policies);
@@ -1221,6 +1141,19 @@ int i915_guc_submission_enable(struct drm_i915_private *dev_priv)
 	enum intel_engine_id id;
 	int err;
 
+	/*
+	 * We're using GuC work items for submitting work through GuC. Since
+	 * we're coalescing multiple requests from a single context into a
+	 * single work item prior to assigning it to execlist_port, we can
+	 * never have more work items than the total number of ports (for all
+	 * engines). The GuC firmware is controlling the HEAD of work queue,
+	 * and it is guaranteed that it will remove the work item from the
+	 * queue before our request is completed.
+	 */
+	BUILD_BUG_ON(ARRAY_SIZE(engine->execlists.port) *
+		     sizeof(struct guc_wq_item) *
+		     I915_NUM_ENGINES > GUC_WQ_SIZE);
+
 	if (!client) {
 		client = guc_client_alloc(dev_priv,
 					  INTEL_INFO(dev_priv)->ring_mask,
@@ -1248,24 +1181,15 @@ int i915_guc_submission_enable(struct drm_i915_private *dev_priv)
 	guc_interrupts_capture(dev_priv);
 
 	for_each_engine(engine, dev_priv, id) {
-		const int wqi_size = sizeof(struct guc_wq_item);
-		struct drm_i915_gem_request *rq;
-
+		struct intel_engine_execlists * const execlists = &engine->execlists;
 		/* The tasklet was initialised by execlists, and may be in
 		 * a state of flux (across a reset) and so we just want to
 		 * take over the callback without changing any other state
 		 * in the tasklet.
 		 */
-		engine->irq_tasklet.func = i915_guc_irq_handler;
+		execlists->irq_tasklet.func = i915_guc_irq_handler;
 		clear_bit(ENGINE_IRQ_EXECLIST, &engine->irq_posted);
-
-		/* Replay the current set of previously submitted requests */
-		spin_lock_irq(&engine->timeline->lock);
-		list_for_each_entry(rq, &engine->timeline->requests, link) {
-			guc_client_update_wq_rsvd(client, wqi_size);
-			__i915_guc_submit(rq);
-		}
-		spin_unlock_irq(&engine->timeline->lock);
+		tasklet_schedule(&execlists->irq_tasklet);
 	}
 
 	return 0;
@@ -1310,7 +1234,7 @@ int intel_guc_suspend(struct drm_i915_private *dev_priv)
 	/* any value greater than GUC_POWER_D0 */
 	data[1] = GUC_POWER_D1;
 	/* first page is shared data with GuC */
-	data[2] = guc_ggtt_offset(ctx->engine[RCS].state);
+	data[2] = guc_ggtt_offset(ctx->engine[RCS].state) + LRC_GUCSHR_PN * PAGE_SIZE;
 
 	return intel_guc_send(guc, data, ARRAY_SIZE(data));
 }
@@ -1328,7 +1252,7 @@ int intel_guc_resume(struct drm_i915_private *dev_priv)
 	if (guc->fw.load_status != INTEL_UC_FIRMWARE_SUCCESS)
 		return 0;
 
-	if (i915.guc_log_level >= 0)
+	if (i915_modparams.guc_log_level >= 0)
 		gen9_enable_guc_interrupts(dev_priv);
 
 	ctx = dev_priv->kernel_context;
@@ -1336,7 +1260,7 @@ int intel_guc_resume(struct drm_i915_private *dev_priv)
 	data[0] = INTEL_GUC_ACTION_EXIT_S_STATE;
 	data[1] = GUC_POWER_D0;
 	/* first page is shared data with GuC */
-	data[2] = guc_ggtt_offset(ctx->engine[RCS].state);
+	data[2] = guc_ggtt_offset(ctx->engine[RCS].state) + LRC_GUCSHR_PN * PAGE_SIZE;
 
 	return intel_guc_send(guc, data, ARRAY_SIZE(data));
 }

drivers/gpu/drm/i915/i915_irq.c

@@ -126,7 +126,7 @@ static const u32 hpd_bxt[HPD_NUM_PINS] = {
 	POSTING_READ(GEN8_##type##_IIR(which)); \
 } while (0)
 
-#define GEN5_IRQ_RESET(type) do { \
+#define GEN3_IRQ_RESET(type) do { \
 	I915_WRITE(type##IMR, 0xffffffff); \
 	POSTING_READ(type##IMR); \
 	I915_WRITE(type##IER, 0); \
@@ -136,10 +136,20 @@ static const u32 hpd_bxt[HPD_NUM_PINS] = {
 	POSTING_READ(type##IIR); \
 } while (0)
 
+#define GEN2_IRQ_RESET(type) do { \
+	I915_WRITE16(type##IMR, 0xffff); \
+	POSTING_READ16(type##IMR); \
+	I915_WRITE16(type##IER, 0); \
+	I915_WRITE16(type##IIR, 0xffff); \
+	POSTING_READ16(type##IIR); \
+	I915_WRITE16(type##IIR, 0xffff); \
+	POSTING_READ16(type##IIR); \
+} while (0)
+
 /*
  * We should clear IMR at preinstall/uninstall, and just check at postinstall.
  */
-static void gen5_assert_iir_is_zero(struct drm_i915_private *dev_priv,
+static void gen3_assert_iir_is_zero(struct drm_i915_private *dev_priv,
 				    i915_reg_t reg)
 {
 	u32 val = I915_READ(reg);
@@ -155,20 +165,43 @@ static void gen5_assert_iir_is_zero(struct drm_i915_private *dev_priv,
 	POSTING_READ(reg);
 }
 
+static void gen2_assert_iir_is_zero(struct drm_i915_private *dev_priv,
+				    i915_reg_t reg)
+{
+	u16 val = I915_READ16(reg);
+
+	if (val == 0)
+		return;
+
+	WARN(1, "Interrupt register 0x%x is not zero: 0x%08x\n",
+	     i915_mmio_reg_offset(reg), val);
+	I915_WRITE16(reg, 0xffff);
+	POSTING_READ16(reg);
+	I915_WRITE16(reg, 0xffff);
+	POSTING_READ16(reg);
+}
+
 #define GEN8_IRQ_INIT_NDX(type, which, imr_val, ier_val) do { \
-	gen5_assert_iir_is_zero(dev_priv, GEN8_##type##_IIR(which)); \
+	gen3_assert_iir_is_zero(dev_priv, GEN8_##type##_IIR(which)); \
 	I915_WRITE(GEN8_##type##_IER(which), (ier_val)); \
 	I915_WRITE(GEN8_##type##_IMR(which), (imr_val)); \
 	POSTING_READ(GEN8_##type##_IMR(which)); \
 } while (0)
 
-#define GEN5_IRQ_INIT(type, imr_val, ier_val) do { \
-	gen5_assert_iir_is_zero(dev_priv, type##IIR); \
+#define GEN3_IRQ_INIT(type, imr_val, ier_val) do { \
+	gen3_assert_iir_is_zero(dev_priv, type##IIR); \
 	I915_WRITE(type##IER, (ier_val)); \
 	I915_WRITE(type##IMR, (imr_val)); \
 	POSTING_READ(type##IMR); \
 } while (0)
 
+#define GEN2_IRQ_INIT(type, imr_val, ier_val) do { \
+	gen2_assert_iir_is_zero(dev_priv, type##IIR); \
+	I915_WRITE16(type##IER, (ier_val)); \
+	I915_WRITE16(type##IMR, (imr_val)); \
+	POSTING_READ16(type##IMR); \
+} while (0)
+
 static void gen6_rps_irq_handler(struct drm_i915_private *dev_priv, u32 pm_iir);
 static void gen9_guc_irq_handler(struct drm_i915_private *dev_priv, u32 pm_iir);
 
@@ -534,62 +567,16 @@ void ibx_display_interrupt_update(struct drm_i915_private *dev_priv,
 	POSTING_READ(SDEIMR);
 }
 
-static void
-__i915_enable_pipestat(struct drm_i915_private *dev_priv, enum pipe pipe,
-		       u32 enable_mask, u32 status_mask)
+u32 i915_pipestat_enable_mask(struct drm_i915_private *dev_priv,
+			      enum pipe pipe)
 {
-	i915_reg_t reg = PIPESTAT(pipe);
-	u32 pipestat = I915_READ(reg) & PIPESTAT_INT_ENABLE_MASK;
-
-	lockdep_assert_held(&dev_priv->irq_lock);
-	WARN_ON(!intel_irqs_enabled(dev_priv));
-
-	if (WARN_ONCE(enable_mask & ~PIPESTAT_INT_ENABLE_MASK ||
-		      status_mask & ~PIPESTAT_INT_STATUS_MASK,
-		      "pipe %c: enable_mask=0x%x, status_mask=0x%x\n",
-		      pipe_name(pipe), enable_mask, status_mask))
-		return;
-
-	if ((pipestat & enable_mask) == enable_mask)
-		return;
-
-	dev_priv->pipestat_irq_mask[pipe] |= status_mask;
-
-	/* Enable the interrupt, clear any pending status */
-	pipestat |= enable_mask | status_mask;
-	I915_WRITE(reg, pipestat);
-	POSTING_READ(reg);
-}
-
-static void
-__i915_disable_pipestat(struct drm_i915_private *dev_priv, enum pipe pipe,
-		        u32 enable_mask, u32 status_mask)
-{
-	i915_reg_t reg = PIPESTAT(pipe);
-	u32 pipestat = I915_READ(reg) & PIPESTAT_INT_ENABLE_MASK;
+	u32 status_mask = dev_priv->pipestat_irq_mask[pipe];
+	u32 enable_mask = status_mask << 16;
 
 	lockdep_assert_held(&dev_priv->irq_lock);
-	WARN_ON(!intel_irqs_enabled(dev_priv));
-
-	if (WARN_ONCE(enable_mask & ~PIPESTAT_INT_ENABLE_MASK ||
-		      status_mask & ~PIPESTAT_INT_STATUS_MASK,
-		      "pipe %c: enable_mask=0x%x, status_mask=0x%x\n",
-		      pipe_name(pipe), enable_mask, status_mask))
-		return;
-
-	if ((pipestat & enable_mask) == 0)
-		return;
-
-	dev_priv->pipestat_irq_mask[pipe] &= ~status_mask;
 
-	pipestat &= ~enable_mask;
-	I915_WRITE(reg, pipestat);
-	POSTING_READ(reg);
-}
-
-static u32 vlv_get_pipestat_enable_mask(struct drm_device *dev, u32 status_mask)
-{
-	u32 enable_mask = status_mask << 16;
+	if (INTEL_GEN(dev_priv) < 5)
+		goto out;
 
 	/*
 	 * On pipe A we don't support the PSR interrupt yet,
@@ -612,35 +599,59 @@ static u32 vlv_get_pipestat_enable_mask(struct drm_device *dev, u32 status_mask)
 	if (status_mask & SPRITE1_FLIP_DONE_INT_STATUS_VLV)
 		enable_mask |= SPRITE1_FLIP_DONE_INT_EN_VLV;
 
+out:
+	WARN_ONCE(enable_mask & ~PIPESTAT_INT_ENABLE_MASK ||
+		  status_mask & ~PIPESTAT_INT_STATUS_MASK,
+		  "pipe %c: enable_mask=0x%x, status_mask=0x%x\n",
+		  pipe_name(pipe), enable_mask, status_mask);
+
 	return enable_mask;
 }
 
-void
-i915_enable_pipestat(struct drm_i915_private *dev_priv, enum pipe pipe,
-		     u32 status_mask)
+void i915_enable_pipestat(struct drm_i915_private *dev_priv,
+			  enum pipe pipe, u32 status_mask)
 {
+	i915_reg_t reg = PIPESTAT(pipe);
 	u32 enable_mask;
 
-	if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv))
-		enable_mask = vlv_get_pipestat_enable_mask(&dev_priv->drm,
-							   status_mask);
-	else
-		enable_mask = status_mask << 16;
-	__i915_enable_pipestat(dev_priv, pipe, enable_mask, status_mask);
+	WARN_ONCE(status_mask & ~PIPESTAT_INT_STATUS_MASK,
+		  "pipe %c: status_mask=0x%x\n",
+		  pipe_name(pipe), status_mask);
+
+	lockdep_assert_held(&dev_priv->irq_lock);
+	WARN_ON(!intel_irqs_enabled(dev_priv));
+
+	if ((dev_priv->pipestat_irq_mask[pipe] & status_mask) == status_mask)
+		return;
+
+	dev_priv->pipestat_irq_mask[pipe] |= status_mask;
+	enable_mask = i915_pipestat_enable_mask(dev_priv, pipe);
+
+	I915_WRITE(reg, enable_mask | status_mask);
+	POSTING_READ(reg);
 }
 
-void
-i915_disable_pipestat(struct drm_i915_private *dev_priv, enum pipe pipe,
-		      u32 status_mask)
+void i915_disable_pipestat(struct drm_i915_private *dev_priv,
+			   enum pipe pipe, u32 status_mask)
 {
+	i915_reg_t reg = PIPESTAT(pipe);
 	u32 enable_mask;
 
-	if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv))
-		enable_mask = vlv_get_pipestat_enable_mask(&dev_priv->drm,
-							   status_mask);
-	else
-		enable_mask = status_mask << 16;
-	__i915_disable_pipestat(dev_priv, pipe, enable_mask, status_mask);
+	WARN_ONCE(status_mask & ~PIPESTAT_INT_STATUS_MASK,
+		  "pipe %c: status_mask=0x%x\n",
+		  pipe_name(pipe), status_mask);
+
+	lockdep_assert_held(&dev_priv->irq_lock);
+	WARN_ON(!intel_irqs_enabled(dev_priv));
+
+	if ((dev_priv->pipestat_irq_mask[pipe] & status_mask) == 0)
+		return;
+
+	dev_priv->pipestat_irq_mask[pipe] &= ~status_mask;
+	enable_mask = i915_pipestat_enable_mask(dev_priv, pipe);
+
+	I915_WRITE(reg, enable_mask | status_mask);
+	POSTING_READ(reg);
 }
 
 /**
@@ -772,6 +783,57 @@ static u32 g4x_get_vblank_counter(struct drm_device *dev, unsigned int pipe)
 	return I915_READ(PIPE_FRMCOUNT_G4X(pipe));
 }
 
+/*
+ * On certain encoders on certain platforms, pipe
+ * scanline register will not work to get the scanline,
+ * since the timings are driven from the PORT or issues
+ * with scanline register updates.
+ * This function will use Framestamp and current
+ * timestamp registers to calculate the scanline.
+ */
+static u32 __intel_get_crtc_scanline_from_timestamp(struct intel_crtc *crtc)
+{
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
+	struct drm_vblank_crtc *vblank =
+		&crtc->base.dev->vblank[drm_crtc_index(&crtc->base)];
+	const struct drm_display_mode *mode = &vblank->hwmode;
+	u32 vblank_start = mode->crtc_vblank_start;
+	u32 vtotal = mode->crtc_vtotal;
+	u32 htotal = mode->crtc_htotal;
+	u32 clock = mode->crtc_clock;
+	u32 scanline, scan_prev_time, scan_curr_time, scan_post_time;
+
+	/*
+	 * To avoid the race condition where we might cross into the
+	 * next vblank just between the PIPE_FRMTMSTMP and TIMESTAMP_CTR
+	 * reads. We make sure we read PIPE_FRMTMSTMP and TIMESTAMP_CTR
+	 * during the same frame.
+	 */
+	do {
+		/*
+		 * This field provides read back of the display
+		 * pipe frame time stamp. The time stamp value
+		 * is sampled at every start of vertical blank.
+		 */
+		scan_prev_time = I915_READ_FW(PIPE_FRMTMSTMP(crtc->pipe));
+
+		/*
+		 * The TIMESTAMP_CTR register has the current
+		 * time stamp value.
+		 */
+		scan_curr_time = I915_READ_FW(IVB_TIMESTAMP_CTR);
+
+		scan_post_time = I915_READ_FW(PIPE_FRMTMSTMP(crtc->pipe));
+	} while (scan_post_time != scan_prev_time);
+
+	scanline = div_u64(mul_u32_u32(scan_curr_time - scan_prev_time,
+					clock), 1000 * htotal);
+	scanline = min(scanline, vtotal - 1);
+	scanline = (scanline + vblank_start) % vtotal;
+
+	return scanline;
+}
+
 /* I915_READ_FW, only for fast reads of display block, no need for forcewake etc. */
 static int __intel_get_crtc_scanline(struct intel_crtc *crtc)
 {
@@ -788,6 +850,9 @@ static int __intel_get_crtc_scanline(struct intel_crtc *crtc)
 	vblank = &crtc->base.dev->vblank[drm_crtc_index(&crtc->base)];
 	mode = &vblank->hwmode;
 
+	if (mode->private_flags & I915_MODE_FLAG_GET_SCANLINE_FROM_TIMESTAMP)
+		return __intel_get_crtc_scanline_from_timestamp(crtc);
+
 	vtotal = mode->crtc_vtotal;
 	if (mode->flags & DRM_MODE_FLAG_INTERLACE)
 		vtotal /= 2;
@@ -1005,6 +1070,8 @@ static void notify_ring(struct intel_engine_cs *engine)
 	spin_lock(&engine->breadcrumbs.irq_lock);
 	wait = engine->breadcrumbs.irq_wait;
 	if (wait) {
+		bool wakeup = engine->irq_seqno_barrier;
+
 		/* We use a callback from the dma-fence to submit
 		 * requests after waiting on our own requests. To
 		 * ensure minimum delay in queuing the next request to
@@ -1017,12 +1084,18 @@ static void notify_ring(struct intel_engine_cs *engine)
 		 * and many waiters.
 		 */
 		if (i915_seqno_passed(intel_engine_get_seqno(engine),
-				      wait->seqno) &&
-		    !test_bit(DMA_FENCE_FLAG_SIGNALED_BIT,
-			      &wait->request->fence.flags))
-			rq = i915_gem_request_get(wait->request);
+				      wait->seqno)) {
+			struct drm_i915_gem_request *waiter = wait->request;
+
+			wakeup = true;
+			if (!test_bit(DMA_FENCE_FLAG_SIGNALED_BIT,
+				      &waiter->fence.flags) &&
+			    intel_wait_check_request(wait, waiter))
+				rq = i915_gem_request_get(waiter);
+		}
 
-		wake_up_process(wait->tsk);
+		if (wakeup)
+			wake_up_process(wait->tsk);
 	} else {
 		__intel_engine_disarm_breadcrumbs(engine);
 	}
@@ -1305,10 +1378,11 @@ static void snb_gt_irq_handler(struct drm_i915_private *dev_priv,
 static void
 gen8_cs_irq_handler(struct intel_engine_cs *engine, u32 iir, int test_shift)
 {
+	struct intel_engine_execlists * const execlists = &engine->execlists;
 	bool tasklet = false;
 
 	if (iir & (GT_CONTEXT_SWITCH_INTERRUPT << test_shift)) {
-		if (port_count(&engine->execlist_port[0])) {
+		if (port_count(&execlists->port[0])) {
 			__set_bit(ENGINE_IRQ_EXECLIST, &engine->irq_posted);
 			tasklet = true;
 		}
@@ -1316,11 +1390,11 @@ gen8_cs_irq_handler(struct intel_engine_cs *engine, u32 iir, int test_shift)
 
 	if (iir & (GT_RENDER_USER_INTERRUPT << test_shift)) {
 		notify_ring(engine);
-		tasklet |= i915.enable_guc_submission;
+		tasklet |= i915_modparams.enable_guc_submission;
 	}
 
 	if (tasklet)
-		tasklet_hi_schedule(&engine->irq_tasklet);
+		tasklet_hi_schedule(&execlists->irq_tasklet);
 }
 
 static irqreturn_t gen8_gt_irq_ack(struct drm_i915_private *dev_priv,
@@ -1573,11 +1647,11 @@ static void display_pipe_crc_irq_handler(struct drm_i915_private *dev_priv,
 		 * bonkers. So let's just wait for the next vblank and read
 		 * out the buggy result.
 		 *
-		 * On CHV sometimes the second CRC is bonkers as well, so
+		 * On GEN8+ sometimes the second CRC is bonkers as well, so
 		 * don't trust that one either.
 		 */
 		if (pipe_crc->skipped == 0 ||
-		    (IS_CHERRYVIEW(dev_priv) && pipe_crc->skipped == 1)) {
+		    (INTEL_GEN(dev_priv) >= 8 && pipe_crc->skipped == 1)) {
 			pipe_crc->skipped++;
 			spin_unlock(&pipe_crc->lock);
 			return;
@@ -1706,8 +1780,21 @@ static void gen9_guc_irq_handler(struct drm_i915_private *dev_priv, u32 gt_iir)
 	}
 }
 
-static void valleyview_pipestat_irq_ack(struct drm_i915_private *dev_priv,
-					u32 iir, u32 pipe_stats[I915_MAX_PIPES])
+static void i9xx_pipestat_irq_reset(struct drm_i915_private *dev_priv)
+{
+	enum pipe pipe;
+
+	for_each_pipe(dev_priv, pipe) {
+		I915_WRITE(PIPESTAT(pipe),
+			   PIPESTAT_INT_STATUS_MASK |
+			   PIPE_FIFO_UNDERRUN_STATUS);
+
+		dev_priv->pipestat_irq_mask[pipe] = 0;
+	}
+}
+
+static void i9xx_pipestat_irq_ack(struct drm_i915_private *dev_priv,
+				  u32 iir, u32 pipe_stats[I915_MAX_PIPES])
 {
 	int pipe;
 
@@ -1720,7 +1807,7 @@ static void valleyview_pipestat_irq_ack(struct drm_i915_private *dev_priv,
 
 	for_each_pipe(dev_priv, pipe) {
 		i915_reg_t reg;
-		u32 mask, iir_bit = 0;
+		u32 status_mask, enable_mask, iir_bit = 0;
 
 		/*
 		 * PIPESTAT bits get signalled even when the interrupt is
@@ -1731,7 +1818,7 @@ static void valleyview_pipestat_irq_ack(struct drm_i915_private *dev_priv,
 		 */
 
 		/* fifo underruns are filterered in the underrun handler. */
-		mask = PIPE_FIFO_UNDERRUN_STATUS;
+		status_mask = PIPE_FIFO_UNDERRUN_STATUS;
 
 		switch (pipe) {
 		case PIPE_A:
@@ -1745,25 +1832,92 @@ static void valleyview_pipestat_irq_ack(struct drm_i915_private *dev_priv,
 			break;
 		}
 		if (iir & iir_bit)
-			mask |= dev_priv->pipestat_irq_mask[pipe];
+			status_mask |= dev_priv->pipestat_irq_mask[pipe];
 
-		if (!mask)
+		if (!status_mask)
 			continue;
 
 		reg = PIPESTAT(pipe);
-		mask |= PIPESTAT_INT_ENABLE_MASK;
-		pipe_stats[pipe] = I915_READ(reg) & mask;
+		pipe_stats[pipe] = I915_READ(reg) & status_mask;
+		enable_mask = i915_pipestat_enable_mask(dev_priv, pipe);
 
 		/*
 		 * Clear the PIPE*STAT regs before the IIR
 		 */
-		if (pipe_stats[pipe] & (PIPE_FIFO_UNDERRUN_STATUS |
-					PIPESTAT_INT_STATUS_MASK))
-			I915_WRITE(reg, pipe_stats[pipe]);
+		if (pipe_stats[pipe])
+			I915_WRITE(reg, enable_mask | pipe_stats[pipe]);
 	}
 	spin_unlock(&dev_priv->irq_lock);
 }
 
+static void i8xx_pipestat_irq_handler(struct drm_i915_private *dev_priv,
+				      u16 iir, u32 pipe_stats[I915_MAX_PIPES])
+{
+	enum pipe pipe;
+
+	for_each_pipe(dev_priv, pipe) {
+		if (pipe_stats[pipe] & PIPE_VBLANK_INTERRUPT_STATUS)
+			drm_handle_vblank(&dev_priv->drm, pipe);
+
+		if (pipe_stats[pipe] & PIPE_CRC_DONE_INTERRUPT_STATUS)
+			i9xx_pipe_crc_irq_handler(dev_priv, pipe);
+
+		if (pipe_stats[pipe] & PIPE_FIFO_UNDERRUN_STATUS)
+			intel_cpu_fifo_underrun_irq_handler(dev_priv, pipe);
+	}
+}
+
+static void i915_pipestat_irq_handler(struct drm_i915_private *dev_priv,
+				      u32 iir, u32 pipe_stats[I915_MAX_PIPES])
+{
+	bool blc_event = false;
+	enum pipe pipe;
+
+	for_each_pipe(dev_priv, pipe) {
+		if (pipe_stats[pipe] & PIPE_VBLANK_INTERRUPT_STATUS)
+			drm_handle_vblank(&dev_priv->drm, pipe);
+
+		if (pipe_stats[pipe] & PIPE_LEGACY_BLC_EVENT_STATUS)
+			blc_event = true;
+
+		if (pipe_stats[pipe] & PIPE_CRC_DONE_INTERRUPT_STATUS)
+			i9xx_pipe_crc_irq_handler(dev_priv, pipe);
+
+		if (pipe_stats[pipe] & PIPE_FIFO_UNDERRUN_STATUS)
+			intel_cpu_fifo_underrun_irq_handler(dev_priv, pipe);
+	}
+
+	if (blc_event || (iir & I915_ASLE_INTERRUPT))
+		intel_opregion_asle_intr(dev_priv);
+}
+
+static void i965_pipestat_irq_handler(struct drm_i915_private *dev_priv,
+				      u32 iir, u32 pipe_stats[I915_MAX_PIPES])
+{
+	bool blc_event = false;
+	enum pipe pipe;
+
+	for_each_pipe(dev_priv, pipe) {
+		if (pipe_stats[pipe] & PIPE_START_VBLANK_INTERRUPT_STATUS)
+			drm_handle_vblank(&dev_priv->drm, pipe);
+
+		if (pipe_stats[pipe] & PIPE_LEGACY_BLC_EVENT_STATUS)
+			blc_event = true;
+
+		if (pipe_stats[pipe] & PIPE_CRC_DONE_INTERRUPT_STATUS)
+			i9xx_pipe_crc_irq_handler(dev_priv, pipe);
+
+		if (pipe_stats[pipe] & PIPE_FIFO_UNDERRUN_STATUS)
+			intel_cpu_fifo_underrun_irq_handler(dev_priv, pipe);
+	}
+
+	if (blc_event || (iir & I915_ASLE_INTERRUPT))
+		intel_opregion_asle_intr(dev_priv);
+
+	if (pipe_stats[0] & PIPE_GMBUS_INTERRUPT_STATUS)
+		gmbus_irq_handler(dev_priv);
+}
+
 static void valleyview_pipestat_irq_handler(struct drm_i915_private *dev_priv,
 					    u32 pipe_stats[I915_MAX_PIPES])
 {
@@ -1879,7 +2033,7 @@ static irqreturn_t valleyview_irq_handler(int irq, void *arg)
 
 		/* Call regardless, as some status bits might not be
 		 * signalled in iir */
-		valleyview_pipestat_irq_ack(dev_priv, iir, pipe_stats);
+		i9xx_pipestat_irq_ack(dev_priv, iir, pipe_stats);
 
 		if (iir & (I915_LPE_PIPE_A_INTERRUPT |
 			   I915_LPE_PIPE_B_INTERRUPT))
@@ -1963,7 +2117,7 @@ static irqreturn_t cherryview_irq_handler(int irq, void *arg)
 
 		/* Call regardless, as some status bits might not be
 		 * signalled in iir */
-		valleyview_pipestat_irq_ack(dev_priv, iir, pipe_stats);
+		i9xx_pipestat_irq_ack(dev_priv, iir, pipe_stats);
 
 		if (iir & (I915_LPE_PIPE_A_INTERRUPT |
 			   I915_LPE_PIPE_B_INTERRUPT |
@@ -2860,7 +3014,7 @@ static void ibx_irq_reset(struct drm_i915_private *dev_priv)
 	if (HAS_PCH_NOP(dev_priv))
 		return;
 
-	GEN5_IRQ_RESET(SDE);
+	GEN3_IRQ_RESET(SDE);
 
 	if (HAS_PCH_CPT(dev_priv) || HAS_PCH_LPT(dev_priv))
 		I915_WRITE(SERR_INT, 0xffffffff);
@@ -2888,15 +3042,13 @@ static void ibx_irq_pre_postinstall(struct drm_device *dev)
 
 static void gen5_gt_irq_reset(struct drm_i915_private *dev_priv)
 {
-	GEN5_IRQ_RESET(GT);
+	GEN3_IRQ_RESET(GT);
 	if (INTEL_GEN(dev_priv) >= 6)
-		GEN5_IRQ_RESET(GEN6_PM);
+		GEN3_IRQ_RESET(GEN6_PM);
 }
 
 static void vlv_display_irq_reset(struct drm_i915_private *dev_priv)
 {
-	enum pipe pipe;
-
 	if (IS_CHERRYVIEW(dev_priv))
 		I915_WRITE(DPINVGTT, DPINVGTT_STATUS_MASK_CHV);
 	else
@@ -2905,14 +3057,9 @@ static void vlv_display_irq_reset(struct drm_i915_private *dev_priv)
 	i915_hotplug_interrupt_update_locked(dev_priv, 0xffffffff, 0);
 	I915_WRITE(PORT_HOTPLUG_STAT, I915_READ(PORT_HOTPLUG_STAT));
 
-	for_each_pipe(dev_priv, pipe) {
-		I915_WRITE(PIPESTAT(pipe),
-			   PIPE_FIFO_UNDERRUN_STATUS |
-			   PIPESTAT_INT_STATUS_MASK);
-		dev_priv->pipestat_irq_mask[pipe] = 0;
-	}
+	i9xx_pipestat_irq_reset(dev_priv);
 
-	GEN5_IRQ_RESET(VLV_);
+	GEN3_IRQ_RESET(VLV_);
 	dev_priv->irq_mask = ~0;
 }
 
@@ -2922,8 +3069,7 @@ static void vlv_display_irq_postinstall(struct drm_i915_private *dev_priv)
 	u32 enable_mask;
 	enum pipe pipe;
 
-	pipestat_mask = PLANE_FLIP_DONE_INT_STATUS_VLV |
-			PIPE_CRC_DONE_INTERRUPT_STATUS;
+	pipestat_mask = PIPE_CRC_DONE_INTERRUPT_STATUS;
 
 	i915_enable_pipestat(dev_priv, PIPE_A, PIPE_GMBUS_INTERRUPT_STATUS);
 	for_each_pipe(dev_priv, pipe)
@@ -2943,7 +3089,7 @@ static void vlv_display_irq_postinstall(struct drm_i915_private *dev_priv)
 
 	dev_priv->irq_mask = ~enable_mask;
 
-	GEN5_IRQ_INIT(VLV_, dev_priv->irq_mask, enable_mask);
+	GEN3_IRQ_INIT(VLV_, dev_priv->irq_mask, enable_mask);
 }
 
 /* drm_dma.h hooks
@@ -2952,9 +3098,10 @@ static void ironlake_irq_reset(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = to_i915(dev);
 
-	I915_WRITE(HWSTAM, 0xffffffff);
+	if (IS_GEN5(dev_priv))
+		I915_WRITE(HWSTAM, 0xffffffff);
 
-	GEN5_IRQ_RESET(DE);
+	GEN3_IRQ_RESET(DE);
 	if (IS_GEN7(dev_priv))
 		I915_WRITE(GEN7_ERR_INT, 0xffffffff);
 
@@ -2963,7 +3110,7 @@ static void ironlake_irq_reset(struct drm_device *dev)
 	ibx_irq_reset(dev_priv);
 }
 
-static void valleyview_irq_preinstall(struct drm_device *dev)
+static void valleyview_irq_reset(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = to_i915(dev);
 
@@ -3001,9 +3148,9 @@ static void gen8_irq_reset(struct drm_device *dev)
 						   POWER_DOMAIN_PIPE(pipe)))
 			GEN8_IRQ_RESET_NDX(DE_PIPE, pipe);
 
-	GEN5_IRQ_RESET(GEN8_DE_PORT_);
-	GEN5_IRQ_RESET(GEN8_DE_MISC_);
-	GEN5_IRQ_RESET(GEN8_PCU_);
+	GEN3_IRQ_RESET(GEN8_DE_PORT_);
+	GEN3_IRQ_RESET(GEN8_DE_MISC_);
+	GEN3_IRQ_RESET(GEN8_PCU_);
 
 	if (HAS_PCH_SPLIT(dev_priv))
 		ibx_irq_reset(dev_priv);
@@ -3037,7 +3184,7 @@ void gen8_irq_power_well_pre_disable(struct drm_i915_private *dev_priv,
 	synchronize_irq(dev_priv->drm.irq);
 }
 
-static void cherryview_irq_preinstall(struct drm_device *dev)
+static void cherryview_irq_reset(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = to_i915(dev);
 
@@ -3046,7 +3193,7 @@ static void cherryview_irq_preinstall(struct drm_device *dev)
 
 	gen8_gt_irq_reset(dev_priv);
 
-	GEN5_IRQ_RESET(GEN8_PCU_);
+	GEN3_IRQ_RESET(GEN8_PCU_);
 
 	spin_lock_irq(&dev_priv->irq_lock);
 	if (dev_priv->display_irqs_enabled)
@@ -3111,7 +3258,15 @@ static void ibx_hpd_irq_setup(struct drm_i915_private *dev_priv)
 
 static void spt_hpd_detection_setup(struct drm_i915_private *dev_priv)
 {
-	u32 hotplug;
+	u32 val, hotplug;
+
+	/* Display WA #1179 WaHardHangonHotPlug: cnp */
+	if (HAS_PCH_CNP(dev_priv)) {
+		val = I915_READ(SOUTH_CHICKEN1);
+		val &= ~CHASSIS_CLK_REQ_DURATION_MASK;
+		val |= CHASSIS_CLK_REQ_DURATION(0xf);
+		I915_WRITE(SOUTH_CHICKEN1, val);
+	}
 
 	/* Enable digital hotplug on the PCH */
 	hotplug = I915_READ(PCH_PORT_HOTPLUG);
@@ -3238,10 +3393,12 @@ static void ibx_irq_postinstall(struct drm_device *dev)
 
 	if (HAS_PCH_IBX(dev_priv))
 		mask = SDE_GMBUS | SDE_AUX_MASK | SDE_POISON;
-	else
+	else if (HAS_PCH_CPT(dev_priv) || HAS_PCH_LPT(dev_priv))
 		mask = SDE_GMBUS_CPT | SDE_AUX_MASK_CPT;
+	else
+		mask = SDE_GMBUS_CPT;
 
-	gen5_assert_iir_is_zero(dev_priv, SDEIIR);
+	gen3_assert_iir_is_zero(dev_priv, SDEIIR);
 	I915_WRITE(SDEIMR, ~mask);
 
 	if (HAS_PCH_IBX(dev_priv) || HAS_PCH_CPT(dev_priv) ||
@@ -3272,7 +3429,7 @@ static void gen5_gt_irq_postinstall(struct drm_device *dev)
 		gt_irqs |= GT_BLT_USER_INTERRUPT | GT_BSD_USER_INTERRUPT;
 	}
 
-	GEN5_IRQ_INIT(GT, dev_priv->gt_irq_mask, gt_irqs);
+	GEN3_IRQ_INIT(GT, dev_priv->gt_irq_mask, gt_irqs);
 
 	if (INTEL_GEN(dev_priv) >= 6) {
 		/*
@@ -3285,7 +3442,7 @@ static void gen5_gt_irq_postinstall(struct drm_device *dev)
 		}
 
 		dev_priv->pm_imr = 0xffffffff;
-		GEN5_IRQ_INIT(GEN6_PM, dev_priv->pm_imr, pm_irqs);
+		GEN3_IRQ_INIT(GEN6_PM, dev_priv->pm_imr, pm_irqs);
 	}
 }
 
@@ -3296,18 +3453,14 @@ static int ironlake_irq_postinstall(struct drm_device *dev)
 
 	if (INTEL_GEN(dev_priv) >= 7) {
 		display_mask = (DE_MASTER_IRQ_CONTROL | DE_GSE_IVB |
-				DE_PCH_EVENT_IVB | DE_PLANEC_FLIP_DONE_IVB |
-				DE_PLANEB_FLIP_DONE_IVB |
-				DE_PLANEA_FLIP_DONE_IVB | DE_AUX_CHANNEL_A_IVB);
+				DE_PCH_EVENT_IVB | DE_AUX_CHANNEL_A_IVB);
 		extra_mask = (DE_PIPEC_VBLANK_IVB | DE_PIPEB_VBLANK_IVB |
 			      DE_PIPEA_VBLANK_IVB | DE_ERR_INT_IVB |
 			      DE_DP_A_HOTPLUG_IVB);
 	} else {
 		display_mask = (DE_MASTER_IRQ_CONTROL | DE_GSE | DE_PCH_EVENT |
-				DE_PLANEA_FLIP_DONE | DE_PLANEB_FLIP_DONE |
-				DE_AUX_CHANNEL_A |
-				DE_PIPEB_CRC_DONE | DE_PIPEA_CRC_DONE |
-				DE_POISON);
+				DE_AUX_CHANNEL_A | DE_PIPEB_CRC_DONE |
+				DE_PIPEA_CRC_DONE | DE_POISON);
 		extra_mask = (DE_PIPEA_VBLANK | DE_PIPEB_VBLANK | DE_PCU_EVENT |
 			      DE_PIPEB_FIFO_UNDERRUN | DE_PIPEA_FIFO_UNDERRUN |
 			      DE_DP_A_HOTPLUG);
@@ -3315,11 +3468,9 @@ static int ironlake_irq_postinstall(struct drm_device *dev)
 
 	dev_priv->irq_mask = ~display_mask;
 
-	I915_WRITE(HWSTAM, 0xeffe);
-
 	ibx_irq_pre_postinstall(dev);
 
-	GEN5_IRQ_INIT(DE, dev_priv->irq_mask, display_mask | extra_mask);
+	GEN3_IRQ_INIT(DE, dev_priv->irq_mask, display_mask | extra_mask);
 
 	gen5_gt_irq_postinstall(dev);
 
@@ -3429,15 +3580,13 @@ static void gen8_de_irq_postinstall(struct drm_i915_private *dev_priv)
 	enum pipe pipe;
 
 	if (INTEL_GEN(dev_priv) >= 9) {
-		de_pipe_masked |= GEN9_PIPE_PLANE1_FLIP_DONE |
-				  GEN9_DE_PIPE_IRQ_FAULT_ERRORS;
+		de_pipe_masked |= GEN9_DE_PIPE_IRQ_FAULT_ERRORS;
 		de_port_masked |= GEN9_AUX_CHANNEL_B | GEN9_AUX_CHANNEL_C |
 				  GEN9_AUX_CHANNEL_D;
 		if (IS_GEN9_LP(dev_priv))
 			de_port_masked |= BXT_DE_PORT_GMBUS;
 	} else {
-		de_pipe_masked |= GEN8_PIPE_PRIMARY_FLIP_DONE |
-				  GEN8_DE_PIPE_IRQ_FAULT_ERRORS;
+		de_pipe_masked |= GEN8_DE_PIPE_IRQ_FAULT_ERRORS;
 	}
 
 	de_pipe_enables = de_pipe_masked | GEN8_PIPE_VBLANK |
@@ -3460,8 +3609,8 @@ static void gen8_de_irq_postinstall(struct drm_i915_private *dev_priv)
 					  dev_priv->de_irq_mask[pipe],
 					  de_pipe_enables);
 
-	GEN5_IRQ_INIT(GEN8_DE_PORT_, ~de_port_masked, de_port_enables);
-	GEN5_IRQ_INIT(GEN8_DE_MISC_, ~de_misc_masked, de_misc_masked);
+	GEN3_IRQ_INIT(GEN8_DE_PORT_, ~de_port_masked, de_port_enables);
+	GEN3_IRQ_INIT(GEN8_DE_MISC_, ~de_misc_masked, de_misc_masked);
 
 	if (IS_GEN9_LP(dev_priv))
 		bxt_hpd_detection_setup(dev_priv);
@@ -3505,98 +3654,36 @@ static int cherryview_irq_postinstall(struct drm_device *dev)
 	return 0;
 }
 
-static void gen8_irq_uninstall(struct drm_device *dev)
-{
-	struct drm_i915_private *dev_priv = to_i915(dev);
-
-	if (!dev_priv)
-		return;
-
-	gen8_irq_reset(dev);
-}
-
-static void valleyview_irq_uninstall(struct drm_device *dev)
-{
-	struct drm_i915_private *dev_priv = to_i915(dev);
-
-	if (!dev_priv)
-		return;
-
-	I915_WRITE(VLV_MASTER_IER, 0);
-	POSTING_READ(VLV_MASTER_IER);
-
-	gen5_gt_irq_reset(dev_priv);
-
-	I915_WRITE(HWSTAM, 0xffffffff);
-
-	spin_lock_irq(&dev_priv->irq_lock);
-	if (dev_priv->display_irqs_enabled)
-		vlv_display_irq_reset(dev_priv);
-	spin_unlock_irq(&dev_priv->irq_lock);
-}
-
-static void cherryview_irq_uninstall(struct drm_device *dev)
+static void i8xx_irq_reset(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = to_i915(dev);
 
-	if (!dev_priv)
-		return;
-
-	I915_WRITE(GEN8_MASTER_IRQ, 0);
-	POSTING_READ(GEN8_MASTER_IRQ);
-
-	gen8_gt_irq_reset(dev_priv);
-
-	GEN5_IRQ_RESET(GEN8_PCU_);
-
-	spin_lock_irq(&dev_priv->irq_lock);
-	if (dev_priv->display_irqs_enabled)
-		vlv_display_irq_reset(dev_priv);
-	spin_unlock_irq(&dev_priv->irq_lock);
-}
-
-static void ironlake_irq_uninstall(struct drm_device *dev)
-{
-	struct drm_i915_private *dev_priv = to_i915(dev);
+	i9xx_pipestat_irq_reset(dev_priv);
 
-	if (!dev_priv)
-		return;
-
-	ironlake_irq_reset(dev);
-}
-
-static void i8xx_irq_preinstall(struct drm_device * dev)
-{
-	struct drm_i915_private *dev_priv = to_i915(dev);
-	int pipe;
+	I915_WRITE16(HWSTAM, 0xffff);
 
-	for_each_pipe(dev_priv, pipe)
-		I915_WRITE(PIPESTAT(pipe), 0);
-	I915_WRITE16(IMR, 0xffff);
-	I915_WRITE16(IER, 0x0);
-	POSTING_READ16(IER);
+	GEN2_IRQ_RESET();
 }
 
 static int i8xx_irq_postinstall(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = to_i915(dev);
+	u16 enable_mask;
 
-	I915_WRITE16(EMR,
-		     ~(I915_ERROR_PAGE_TABLE | I915_ERROR_MEMORY_REFRESH));
+	I915_WRITE16(EMR, ~(I915_ERROR_PAGE_TABLE |
+			    I915_ERROR_MEMORY_REFRESH));
 
 	/* Unmask the interrupts that we always want on. */
 	dev_priv->irq_mask =
 		~(I915_DISPLAY_PIPE_A_EVENT_INTERRUPT |
-		  I915_DISPLAY_PIPE_B_EVENT_INTERRUPT |
-		  I915_DISPLAY_PLANE_A_FLIP_PENDING_INTERRUPT |
-		  I915_DISPLAY_PLANE_B_FLIP_PENDING_INTERRUPT);
-	I915_WRITE16(IMR, dev_priv->irq_mask);
+		  I915_DISPLAY_PIPE_B_EVENT_INTERRUPT);
+
+	enable_mask =
+		I915_DISPLAY_PIPE_A_EVENT_INTERRUPT |
+		I915_DISPLAY_PIPE_B_EVENT_INTERRUPT |
+		I915_USER_INTERRUPT;
 
-	I915_WRITE16(IER,
-		     I915_DISPLAY_PIPE_A_EVENT_INTERRUPT |
-		     I915_DISPLAY_PIPE_B_EVENT_INTERRUPT |
-		     I915_USER_INTERRUPT);
-	POSTING_READ16(IER);
+	GEN2_IRQ_INIT(, dev_priv->irq_mask, enable_mask);
 
 	/* Interrupt setup is already guaranteed to be single-threaded, this is
 	 * just to make the assert_spin_locked check happy. */
@@ -3608,17 +3695,11 @@ static int i8xx_irq_postinstall(struct drm_device *dev)
 	return 0;
 }
 
-/*
- * Returns true when a page flip has completed.
- */
 static irqreturn_t i8xx_irq_handler(int irq, void *arg)
 {
 	struct drm_device *dev = arg;
 	struct drm_i915_private *dev_priv = to_i915(dev);
-	u16 iir, new_iir;
-	u32 pipe_stats[2];
-	int pipe;
-	irqreturn_t ret;
+	irqreturn_t ret = IRQ_NONE;
 
 	if (!intel_irqs_enabled(dev_priv))
 		return IRQ_NONE;
@@ -3626,96 +3707,50 @@ static irqreturn_t i8xx_irq_handler(int irq, void *arg)
 	/* IRQs are synced during runtime_suspend, we don't require a wakeref */
 	disable_rpm_wakeref_asserts(dev_priv);
 
-	ret = IRQ_NONE;
-	iir = I915_READ16(IIR);
-	if (iir == 0)
-		goto out;
+	do {
+		u32 pipe_stats[I915_MAX_PIPES] = {};
+		u16 iir;
 
-	while (iir) {
-		/* Can't rely on pipestat interrupt bit in iir as it might
-		 * have been cleared after the pipestat interrupt was received.
-		 * It doesn't set the bit in iir again, but it still produces
-		 * interrupts (for non-MSI).
-		 */
-		spin_lock(&dev_priv->irq_lock);
-		if (iir & I915_RENDER_COMMAND_PARSER_ERROR_INTERRUPT)
-			DRM_DEBUG("Command parser error, iir 0x%08x\n", iir);
+		iir = I915_READ16(IIR);
+		if (iir == 0)
+			break;
 
-		for_each_pipe(dev_priv, pipe) {
-			i915_reg_t reg = PIPESTAT(pipe);
-			pipe_stats[pipe] = I915_READ(reg);
+		ret = IRQ_HANDLED;
 
-			/*
-			 * Clear the PIPE*STAT regs before the IIR
-			 */
-			if (pipe_stats[pipe] & 0x8000ffff)
-				I915_WRITE(reg, pipe_stats[pipe]);
-		}
-		spin_unlock(&dev_priv->irq_lock);
+		/* Call regardless, as some status bits might not be
+		 * signalled in iir */
+		i9xx_pipestat_irq_ack(dev_priv, iir, pipe_stats);
 
 		I915_WRITE16(IIR, iir);
-		new_iir = I915_READ16(IIR); /* Flush posted writes */
 
 		if (iir & I915_USER_INTERRUPT)
 			notify_ring(dev_priv->engine[RCS]);
 
-		for_each_pipe(dev_priv, pipe) {
-			int plane = pipe;
-			if (HAS_FBC(dev_priv))
-				plane = !plane;
-
-			if (pipe_stats[pipe] & PIPE_VBLANK_INTERRUPT_STATUS)
-				drm_handle_vblank(&dev_priv->drm, pipe);
-
-			if (pipe_stats[pipe] & PIPE_CRC_DONE_INTERRUPT_STATUS)
-				i9xx_pipe_crc_irq_handler(dev_priv, pipe);
-
-			if (pipe_stats[pipe] & PIPE_FIFO_UNDERRUN_STATUS)
-				intel_cpu_fifo_underrun_irq_handler(dev_priv,
-								    pipe);
-		}
+		if (iir & I915_RENDER_COMMAND_PARSER_ERROR_INTERRUPT)
+			DRM_DEBUG("Command parser error, iir 0x%08x\n", iir);
 
-		iir = new_iir;
-	}
-	ret = IRQ_HANDLED;
+		i8xx_pipestat_irq_handler(dev_priv, iir, pipe_stats);
+	} while (0);
 
-out:
 	enable_rpm_wakeref_asserts(dev_priv);
 
 	return ret;
 }
 
-static void i8xx_irq_uninstall(struct drm_device * dev)
-{
-	struct drm_i915_private *dev_priv = to_i915(dev);
-	int pipe;
-
-	for_each_pipe(dev_priv, pipe) {
-		/* Clear enable bits; then clear status bits */
-		I915_WRITE(PIPESTAT(pipe), 0);
-		I915_WRITE(PIPESTAT(pipe), I915_READ(PIPESTAT(pipe)));
-	}
-	I915_WRITE16(IMR, 0xffff);
-	I915_WRITE16(IER, 0x0);
-	I915_WRITE16(IIR, I915_READ16(IIR));
-}
-
-static void i915_irq_preinstall(struct drm_device * dev)
+static void i915_irq_reset(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = to_i915(dev);
-	int pipe;
 
 	if (I915_HAS_HOTPLUG(dev_priv)) {
 		i915_hotplug_interrupt_update(dev_priv, 0xffffffff, 0);
 		I915_WRITE(PORT_HOTPLUG_STAT, I915_READ(PORT_HOTPLUG_STAT));
 	}
 
-	I915_WRITE16(HWSTAM, 0xeffe);
-	for_each_pipe(dev_priv, pipe)
-		I915_WRITE(PIPESTAT(pipe), 0);
-	I915_WRITE(IMR, 0xffffffff);
-	I915_WRITE(IER, 0x0);
-	POSTING_READ(IER);
+	i9xx_pipestat_irq_reset(dev_priv);
+
+	I915_WRITE(HWSTAM, 0xffffffff);
+
+	GEN3_IRQ_RESET();
 }
 
 static int i915_irq_postinstall(struct drm_device *dev)
@@ -3723,15 +3758,14 @@ static int i915_irq_postinstall(struct drm_device *dev)
 	struct drm_i915_private *dev_priv = to_i915(dev);
 	u32 enable_mask;
 
-	I915_WRITE(EMR, ~(I915_ERROR_PAGE_TABLE | I915_ERROR_MEMORY_REFRESH));
+	I915_WRITE(EMR, ~(I915_ERROR_PAGE_TABLE |
+			  I915_ERROR_MEMORY_REFRESH));
 
 	/* Unmask the interrupts that we always want on. */
 	dev_priv->irq_mask =
 		~(I915_ASLE_INTERRUPT |
 		  I915_DISPLAY_PIPE_A_EVENT_INTERRUPT |
-		  I915_DISPLAY_PIPE_B_EVENT_INTERRUPT |
-		  I915_DISPLAY_PLANE_A_FLIP_PENDING_INTERRUPT |
-		  I915_DISPLAY_PLANE_B_FLIP_PENDING_INTERRUPT);
+		  I915_DISPLAY_PIPE_B_EVENT_INTERRUPT);
 
 	enable_mask =
 		I915_ASLE_INTERRUPT |
@@ -3740,20 +3774,13 @@ static int i915_irq_postinstall(struct drm_device *dev)
 		I915_USER_INTERRUPT;
 
 	if (I915_HAS_HOTPLUG(dev_priv)) {
-		i915_hotplug_interrupt_update(dev_priv, 0xffffffff, 0);
-		POSTING_READ(PORT_HOTPLUG_EN);
-
 		/* Enable in IER... */
 		enable_mask |= I915_DISPLAY_PORT_INTERRUPT;
 		/* and unmask in IMR */
 		dev_priv->irq_mask &= ~I915_DISPLAY_PORT_INTERRUPT;
 	}
 
-	I915_WRITE(IMR, dev_priv->irq_mask);
-	I915_WRITE(IER, enable_mask);
-	POSTING_READ(IER);
-
-	i915_enable_asle_pipestat(dev_priv);
+	GEN3_IRQ_INIT(, dev_priv->irq_mask, enable_mask);
 
 	/* Interrupt setup is already guaranteed to be single-threaded, this is
 	 * just to make the assert_spin_locked check happy. */
@@ -3762,6 +3789,8 @@ static int i915_irq_postinstall(struct drm_device *dev)
 	i915_enable_pipestat(dev_priv, PIPE_B, PIPE_CRC_DONE_INTERRUPT_STATUS);
 	spin_unlock_irq(&dev_priv->irq_lock);
 
+	i915_enable_asle_pipestat(dev_priv);
+
 	return 0;
 }
 
@@ -3769,8 +3798,7 @@ static irqreturn_t i915_irq_handler(int irq, void *arg)
 {
 	struct drm_device *dev = arg;
 	struct drm_i915_private *dev_priv = to_i915(dev);
-	u32 iir, new_iir, pipe_stats[I915_MAX_PIPES];
-	int pipe, ret = IRQ_NONE;
+	irqreturn_t ret = IRQ_NONE;
 
 	if (!intel_irqs_enabled(dev_priv))
 		return IRQ_NONE;
@@ -3778,131 +3806,56 @@ static irqreturn_t i915_irq_handler(int irq, void *arg)
 	/* IRQs are synced during runtime_suspend, we don't require a wakeref */
 	disable_rpm_wakeref_asserts(dev_priv);
 
-	iir = I915_READ(IIR);
 	do {
-		bool irq_received = (iir) != 0;
-		bool blc_event = false;
-
-		/* Can't rely on pipestat interrupt bit in iir as it might
-		 * have been cleared after the pipestat interrupt was received.
-		 * It doesn't set the bit in iir again, but it still produces
-		 * interrupts (for non-MSI).
-		 */
-		spin_lock(&dev_priv->irq_lock);
-		if (iir & I915_RENDER_COMMAND_PARSER_ERROR_INTERRUPT)
-			DRM_DEBUG("Command parser error, iir 0x%08x\n", iir);
-
-		for_each_pipe(dev_priv, pipe) {
-			i915_reg_t reg = PIPESTAT(pipe);
-			pipe_stats[pipe] = I915_READ(reg);
-
-			/* Clear the PIPE*STAT regs before the IIR */
-			if (pipe_stats[pipe] & 0x8000ffff) {
-				I915_WRITE(reg, pipe_stats[pipe]);
-				irq_received = true;
-			}
-		}
-		spin_unlock(&dev_priv->irq_lock);
+		u32 pipe_stats[I915_MAX_PIPES] = {};
+		u32 hotplug_status = 0;
+		u32 iir;
 
-		if (!irq_received)
+		iir = I915_READ(IIR);
+		if (iir == 0)
 			break;
 
-		/* Consume port.  Then clear IIR or we'll miss events */
+		ret = IRQ_HANDLED;
+
 		if (I915_HAS_HOTPLUG(dev_priv) &&
-		    iir & I915_DISPLAY_PORT_INTERRUPT) {
-			u32 hotplug_status = i9xx_hpd_irq_ack(dev_priv);
-			if (hotplug_status)
-				i9xx_hpd_irq_handler(dev_priv, hotplug_status);
-		}
+		    iir & I915_DISPLAY_PORT_INTERRUPT)
+			hotplug_status = i9xx_hpd_irq_ack(dev_priv);
+
+		/* Call regardless, as some status bits might not be
+		 * signalled in iir */
+		i9xx_pipestat_irq_ack(dev_priv, iir, pipe_stats);
 
 		I915_WRITE(IIR, iir);
-		new_iir = I915_READ(IIR); /* Flush posted writes */
 
 		if (iir & I915_USER_INTERRUPT)
 			notify_ring(dev_priv->engine[RCS]);
 
-		for_each_pipe(dev_priv, pipe) {
-			int plane = pipe;
-			if (HAS_FBC(dev_priv))
-				plane = !plane;
-
-			if (pipe_stats[pipe] & PIPE_VBLANK_INTERRUPT_STATUS)
-				drm_handle_vblank(&dev_priv->drm, pipe);
-
-			if (pipe_stats[pipe] & PIPE_LEGACY_BLC_EVENT_STATUS)
-				blc_event = true;
-
-			if (pipe_stats[pipe] & PIPE_CRC_DONE_INTERRUPT_STATUS)
-				i9xx_pipe_crc_irq_handler(dev_priv, pipe);
-
-			if (pipe_stats[pipe] & PIPE_FIFO_UNDERRUN_STATUS)
-				intel_cpu_fifo_underrun_irq_handler(dev_priv,
-								    pipe);
-		}
+		if (iir & I915_RENDER_COMMAND_PARSER_ERROR_INTERRUPT)
+			DRM_DEBUG("Command parser error, iir 0x%08x\n", iir);
 
-		if (blc_event || (iir & I915_ASLE_INTERRUPT))
-			intel_opregion_asle_intr(dev_priv);
+		if (hotplug_status)
+			i9xx_hpd_irq_handler(dev_priv, hotplug_status);
 
-		/* With MSI, interrupts are only generated when iir
-		 * transitions from zero to nonzero.  If another bit got
-		 * set while we were handling the existing iir bits, then
-		 * we would never get another interrupt.
-		 *
-		 * This is fine on non-MSI as well, as if we hit this path
-		 * we avoid exiting the interrupt handler only to generate
-		 * another one.
-		 *
-		 * Note that for MSI this could cause a stray interrupt report
-		 * if an interrupt landed in the time between writing IIR and
-		 * the posting read.  This should be rare enough to never
-		 * trigger the 99% of 100,000 interrupts test for disabling
-		 * stray interrupts.
-		 */
-		ret = IRQ_HANDLED;
-		iir = new_iir;
-	} while (iir);
+		i915_pipestat_irq_handler(dev_priv, iir, pipe_stats);
+	} while (0);
 
 	enable_rpm_wakeref_asserts(dev_priv);
 
 	return ret;
 }
 
-static void i915_irq_uninstall(struct drm_device * dev)
-{
-	struct drm_i915_private *dev_priv = to_i915(dev);
-	int pipe;
-
-	if (I915_HAS_HOTPLUG(dev_priv)) {
-		i915_hotplug_interrupt_update(dev_priv, 0xffffffff, 0);
-		I915_WRITE(PORT_HOTPLUG_STAT, I915_READ(PORT_HOTPLUG_STAT));
-	}
-
-	I915_WRITE16(HWSTAM, 0xffff);
-	for_each_pipe(dev_priv, pipe) {
-		/* Clear enable bits; then clear status bits */
-		I915_WRITE(PIPESTAT(pipe), 0);
-		I915_WRITE(PIPESTAT(pipe), I915_READ(PIPESTAT(pipe)));
-	}
-	I915_WRITE(IMR, 0xffffffff);
-	I915_WRITE(IER, 0x0);
-
-	I915_WRITE(IIR, I915_READ(IIR));
-}
-
-static void i965_irq_preinstall(struct drm_device * dev)
+static void i965_irq_reset(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = to_i915(dev);
-	int pipe;
 
 	i915_hotplug_interrupt_update(dev_priv, 0xffffffff, 0);
 	I915_WRITE(PORT_HOTPLUG_STAT, I915_READ(PORT_HOTPLUG_STAT));
 
-	I915_WRITE(HWSTAM, 0xeffe);
-	for_each_pipe(dev_priv, pipe)
-		I915_WRITE(PIPESTAT(pipe), 0);
-	I915_WRITE(IMR, 0xffffffff);
-	I915_WRITE(IER, 0x0);
-	POSTING_READ(IER);
+	i9xx_pipestat_irq_reset(dev_priv);
+
+	I915_WRITE(HWSTAM, 0xffffffff);
+
+	GEN3_IRQ_RESET();
 }
 
 static int i965_irq_postinstall(struct drm_device *dev)
@@ -3911,31 +3864,6 @@ static int i965_irq_postinstall(struct drm_device *dev)
 	u32 enable_mask;
 	u32 error_mask;
 
-	/* Unmask the interrupts that we always want on. */
-	dev_priv->irq_mask = ~(I915_ASLE_INTERRUPT |
-			       I915_DISPLAY_PORT_INTERRUPT |
-			       I915_DISPLAY_PIPE_A_EVENT_INTERRUPT |
-			       I915_DISPLAY_PIPE_B_EVENT_INTERRUPT |
-			       I915_DISPLAY_PLANE_A_FLIP_PENDING_INTERRUPT |
-			       I915_DISPLAY_PLANE_B_FLIP_PENDING_INTERRUPT |
-			       I915_RENDER_COMMAND_PARSER_ERROR_INTERRUPT);
-
-	enable_mask = ~dev_priv->irq_mask;
-	enable_mask &= ~(I915_DISPLAY_PLANE_A_FLIP_PENDING_INTERRUPT |
-			 I915_DISPLAY_PLANE_B_FLIP_PENDING_INTERRUPT);
-	enable_mask |= I915_USER_INTERRUPT;
-
-	if (IS_G4X(dev_priv))
-		enable_mask |= I915_BSD_USER_INTERRUPT;
-
-	/* Interrupt setup is already guaranteed to be single-threaded, this is
-	 * just to make the assert_spin_locked check happy. */
-	spin_lock_irq(&dev_priv->irq_lock);
-	i915_enable_pipestat(dev_priv, PIPE_A, PIPE_GMBUS_INTERRUPT_STATUS);
-	i915_enable_pipestat(dev_priv, PIPE_A, PIPE_CRC_DONE_INTERRUPT_STATUS);
-	i915_enable_pipestat(dev_priv, PIPE_B, PIPE_CRC_DONE_INTERRUPT_STATUS);
-	spin_unlock_irq(&dev_priv->irq_lock);
-
 	/*
 	 * Enable some error detection, note the instruction error mask
 	 * bit is reserved, so we leave it masked.
@@ -3951,12 +3879,34 @@ static int i965_irq_postinstall(struct drm_device *dev)
 	}
 	I915_WRITE(EMR, error_mask);
 
-	I915_WRITE(IMR, dev_priv->irq_mask);
-	I915_WRITE(IER, enable_mask);
-	POSTING_READ(IER);
+	/* Unmask the interrupts that we always want on. */
+	dev_priv->irq_mask =
+		~(I915_ASLE_INTERRUPT |
+		  I915_DISPLAY_PORT_INTERRUPT |
+		  I915_DISPLAY_PIPE_A_EVENT_INTERRUPT |
+		  I915_DISPLAY_PIPE_B_EVENT_INTERRUPT |
+		  I915_RENDER_COMMAND_PARSER_ERROR_INTERRUPT);
 
-	i915_hotplug_interrupt_update(dev_priv, 0xffffffff, 0);
-	POSTING_READ(PORT_HOTPLUG_EN);
+	enable_mask =
+		I915_ASLE_INTERRUPT |
+		I915_DISPLAY_PORT_INTERRUPT |
+		I915_DISPLAY_PIPE_A_EVENT_INTERRUPT |
+		I915_DISPLAY_PIPE_B_EVENT_INTERRUPT |
+		I915_RENDER_COMMAND_PARSER_ERROR_INTERRUPT |
+		I915_USER_INTERRUPT;
+
+	if (IS_G4X(dev_priv))
+		enable_mask |= I915_BSD_USER_INTERRUPT;
+
+	GEN3_IRQ_INIT(, dev_priv->irq_mask, enable_mask);
+
+	/* Interrupt setup is already guaranteed to be single-threaded, this is
+	 * just to make the assert_spin_locked check happy. */
+	spin_lock_irq(&dev_priv->irq_lock);
+	i915_enable_pipestat(dev_priv, PIPE_A, PIPE_GMBUS_INTERRUPT_STATUS);
+	i915_enable_pipestat(dev_priv, PIPE_A, PIPE_CRC_DONE_INTERRUPT_STATUS);
+	i915_enable_pipestat(dev_priv, PIPE_B, PIPE_CRC_DONE_INTERRUPT_STATUS);
+	spin_unlock_irq(&dev_priv->irq_lock);
 
 	i915_enable_asle_pipestat(dev_priv);
 
@@ -3992,9 +3942,7 @@ static irqreturn_t i965_irq_handler(int irq, void *arg)
 {
 	struct drm_device *dev = arg;
 	struct drm_i915_private *dev_priv = to_i915(dev);
-	u32 iir, new_iir;
-	u32 pipe_stats[I915_MAX_PIPES];
-	int ret = IRQ_NONE, pipe;
+	irqreturn_t ret = IRQ_NONE;
 
 	if (!intel_irqs_enabled(dev_priv))
 		return IRQ_NONE;
@@ -4002,121 +3950,46 @@ static irqreturn_t i965_irq_handler(int irq, void *arg)
 	/* IRQs are synced during runtime_suspend, we don't require a wakeref */
 	disable_rpm_wakeref_asserts(dev_priv);
 
-	iir = I915_READ(IIR);
-
-	for (;;) {
-		bool irq_received = (iir) != 0;
-		bool blc_event = false;
-
-		/* Can't rely on pipestat interrupt bit in iir as it might
-		 * have been cleared after the pipestat interrupt was received.
-		 * It doesn't set the bit in iir again, but it still produces
-		 * interrupts (for non-MSI).
-		 */
-		spin_lock(&dev_priv->irq_lock);
-		if (iir & I915_RENDER_COMMAND_PARSER_ERROR_INTERRUPT)
-			DRM_DEBUG("Command parser error, iir 0x%08x\n", iir);
-
-		for_each_pipe(dev_priv, pipe) {
-			i915_reg_t reg = PIPESTAT(pipe);
-			pipe_stats[pipe] = I915_READ(reg);
-
-			/*
-			 * Clear the PIPE*STAT regs before the IIR
-			 */
-			if (pipe_stats[pipe] & 0x8000ffff) {
-				I915_WRITE(reg, pipe_stats[pipe]);
-				irq_received = true;
-			}
-		}
-		spin_unlock(&dev_priv->irq_lock);
+	do {
+		u32 pipe_stats[I915_MAX_PIPES] = {};
+		u32 hotplug_status = 0;
+		u32 iir;
 
-		if (!irq_received)
+		iir = I915_READ(IIR);
+		if (iir == 0)
 			break;
 
 		ret = IRQ_HANDLED;
 
-		/* Consume port.  Then clear IIR or we'll miss events */
-		if (iir & I915_DISPLAY_PORT_INTERRUPT) {
-			u32 hotplug_status = i9xx_hpd_irq_ack(dev_priv);
-			if (hotplug_status)
-				i9xx_hpd_irq_handler(dev_priv, hotplug_status);
-		}
+		if (iir & I915_DISPLAY_PORT_INTERRUPT)
+			hotplug_status = i9xx_hpd_irq_ack(dev_priv);
+
+		/* Call regardless, as some status bits might not be
+		 * signalled in iir */
+		i9xx_pipestat_irq_ack(dev_priv, iir, pipe_stats);
 
 		I915_WRITE(IIR, iir);
-		new_iir = I915_READ(IIR); /* Flush posted writes */
 
 		if (iir & I915_USER_INTERRUPT)
 			notify_ring(dev_priv->engine[RCS]);
+
 		if (iir & I915_BSD_USER_INTERRUPT)
 			notify_ring(dev_priv->engine[VCS]);
 
-		for_each_pipe(dev_priv, pipe) {
-			if (pipe_stats[pipe] & PIPE_START_VBLANK_INTERRUPT_STATUS)
-				drm_handle_vblank(&dev_priv->drm, pipe);
-
-			if (pipe_stats[pipe] & PIPE_LEGACY_BLC_EVENT_STATUS)
-				blc_event = true;
-
-			if (pipe_stats[pipe] & PIPE_CRC_DONE_INTERRUPT_STATUS)
-				i9xx_pipe_crc_irq_handler(dev_priv, pipe);
-
-			if (pipe_stats[pipe] & PIPE_FIFO_UNDERRUN_STATUS)
-				intel_cpu_fifo_underrun_irq_handler(dev_priv, pipe);
-		}
-
-		if (blc_event || (iir & I915_ASLE_INTERRUPT))
-			intel_opregion_asle_intr(dev_priv);
+		if (iir & I915_RENDER_COMMAND_PARSER_ERROR_INTERRUPT)
+			DRM_DEBUG("Command parser error, iir 0x%08x\n", iir);
 
-		if (pipe_stats[0] & PIPE_GMBUS_INTERRUPT_STATUS)
-			gmbus_irq_handler(dev_priv);
+		if (hotplug_status)
+			i9xx_hpd_irq_handler(dev_priv, hotplug_status);
 
-		/* With MSI, interrupts are only generated when iir
-		 * transitions from zero to nonzero.  If another bit got
-		 * set while we were handling the existing iir bits, then
-		 * we would never get another interrupt.
-		 *
-		 * This is fine on non-MSI as well, as if we hit this path
-		 * we avoid exiting the interrupt handler only to generate
-		 * another one.
-		 *
-		 * Note that for MSI this could cause a stray interrupt report
-		 * if an interrupt landed in the time between writing IIR and
-		 * the posting read.  This should be rare enough to never
-		 * trigger the 99% of 100,000 interrupts test for disabling
-		 * stray interrupts.
-		 */
-		iir = new_iir;
-	}
+		i965_pipestat_irq_handler(dev_priv, iir, pipe_stats);
+	} while (0);
 
 	enable_rpm_wakeref_asserts(dev_priv);
 
 	return ret;
 }
 
-static void i965_irq_uninstall(struct drm_device * dev)
-{
-	struct drm_i915_private *dev_priv = to_i915(dev);
-	int pipe;
-
-	if (!dev_priv)
-		return;
-
-	i915_hotplug_interrupt_update(dev_priv, 0xffffffff, 0);
-	I915_WRITE(PORT_HOTPLUG_STAT, I915_READ(PORT_HOTPLUG_STAT));
-
-	I915_WRITE(HWSTAM, 0xffffffff);
-	for_each_pipe(dev_priv, pipe)
-		I915_WRITE(PIPESTAT(pipe), 0);
-	I915_WRITE(IMR, 0xffffffff);
-	I915_WRITE(IER, 0x0);
-
-	for_each_pipe(dev_priv, pipe)
-		I915_WRITE(PIPESTAT(pipe),
-			   I915_READ(PIPESTAT(pipe)) & 0x8000ffff);
-	I915_WRITE(IIR, I915_READ(IIR));
-}
-
 /**
  * intel_irq_init - initializes irq support
  * @dev_priv: i915 device instance
@@ -4197,17 +4070,17 @@ void intel_irq_init(struct drm_i915_private *dev_priv)
 
 	if (IS_CHERRYVIEW(dev_priv)) {
 		dev->driver->irq_handler = cherryview_irq_handler;
-		dev->driver->irq_preinstall = cherryview_irq_preinstall;
+		dev->driver->irq_preinstall = cherryview_irq_reset;
 		dev->driver->irq_postinstall = cherryview_irq_postinstall;
-		dev->driver->irq_uninstall = cherryview_irq_uninstall;
+		dev->driver->irq_uninstall = cherryview_irq_reset;
 		dev->driver->enable_vblank = i965_enable_vblank;
 		dev->driver->disable_vblank = i965_disable_vblank;
 		dev_priv->display.hpd_irq_setup = i915_hpd_irq_setup;
 	} else if (IS_VALLEYVIEW(dev_priv)) {
 		dev->driver->irq_handler = valleyview_irq_handler;
-		dev->driver->irq_preinstall = valleyview_irq_preinstall;
+		dev->driver->irq_preinstall = valleyview_irq_reset;
 		dev->driver->irq_postinstall = valleyview_irq_postinstall;
-		dev->driver->irq_uninstall = valleyview_irq_uninstall;
+		dev->driver->irq_uninstall = valleyview_irq_reset;
 		dev->driver->enable_vblank = i965_enable_vblank;
 		dev->driver->disable_vblank = i965_disable_vblank;
 		dev_priv->display.hpd_irq_setup = i915_hpd_irq_setup;
@@ -4215,7 +4088,7 @@ void intel_irq_init(struct drm_i915_private *dev_priv)
 		dev->driver->irq_handler = gen8_irq_handler;
 		dev->driver->irq_preinstall = gen8_irq_reset;
 		dev->driver->irq_postinstall = gen8_irq_postinstall;
-		dev->driver->irq_uninstall = gen8_irq_uninstall;
+		dev->driver->irq_uninstall = gen8_irq_reset;
 		dev->driver->enable_vblank = gen8_enable_vblank;
 		dev->driver->disable_vblank = gen8_disable_vblank;
 		if (IS_GEN9_LP(dev_priv))
@@ -4229,29 +4102,29 @@ void intel_irq_init(struct drm_i915_private *dev_priv)
 		dev->driver->irq_handler = ironlake_irq_handler;
 		dev->driver->irq_preinstall = ironlake_irq_reset;
 		dev->driver->irq_postinstall = ironlake_irq_postinstall;
-		dev->driver->irq_uninstall = ironlake_irq_uninstall;
+		dev->driver->irq_uninstall = ironlake_irq_reset;
 		dev->driver->enable_vblank = ironlake_enable_vblank;
 		dev->driver->disable_vblank = ironlake_disable_vblank;
 		dev_priv->display.hpd_irq_setup = ilk_hpd_irq_setup;
 	} else {
 		if (IS_GEN2(dev_priv)) {
-			dev->driver->irq_preinstall = i8xx_irq_preinstall;
+			dev->driver->irq_preinstall = i8xx_irq_reset;
 			dev->driver->irq_postinstall = i8xx_irq_postinstall;
 			dev->driver->irq_handler = i8xx_irq_handler;
-			dev->driver->irq_uninstall = i8xx_irq_uninstall;
+			dev->driver->irq_uninstall = i8xx_irq_reset;
 			dev->driver->enable_vblank = i8xx_enable_vblank;
 			dev->driver->disable_vblank = i8xx_disable_vblank;
 		} else if (IS_GEN3(dev_priv)) {
-			dev->driver->irq_preinstall = i915_irq_preinstall;
+			dev->driver->irq_preinstall = i915_irq_reset;
 			dev->driver->irq_postinstall = i915_irq_postinstall;
-			dev->driver->irq_uninstall = i915_irq_uninstall;
+			dev->driver->irq_uninstall = i915_irq_reset;
 			dev->driver->irq_handler = i915_irq_handler;
 			dev->driver->enable_vblank = i8xx_enable_vblank;
 			dev->driver->disable_vblank = i8xx_disable_vblank;
 		} else {
-			dev->driver->irq_preinstall = i965_irq_preinstall;
+			dev->driver->irq_preinstall = i965_irq_reset;
 			dev->driver->irq_postinstall = i965_irq_postinstall;
-			dev->driver->irq_uninstall = i965_irq_uninstall;
+			dev->driver->irq_uninstall = i965_irq_reset;
 			dev->driver->irq_handler = i965_irq_handler;
 			dev->driver->enable_vblank = i965_enable_vblank;
 			dev->driver->disable_vblank = i965_disable_vblank;

drivers/gpu/drm/i915/i915_oa_cflgt2.c

@@ -0,0 +1,109 @@
+/*
+ * Autogenerated file by GPU Top : https://github.com/rib/gputop
+ * DO NOT EDIT manually!
+ *
+ *
+ * Copyright (c) 2015 Intel Corporation
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the next
+ * paragraph) shall be included in all copies or substantial portions of the
+ * Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
+ * IN THE SOFTWARE.
+ *
+ */
+
+#include <linux/sysfs.h>
+
+#include "i915_drv.h"
+#include "i915_oa_cflgt2.h"
+
+static const struct i915_oa_reg b_counter_config_test_oa[] = {
+	{ _MMIO(0x2740), 0x00000000 },
+	{ _MMIO(0x2744), 0x00800000 },
+	{ _MMIO(0x2714), 0xf0800000 },
+	{ _MMIO(0x2710), 0x00000000 },
+	{ _MMIO(0x2724), 0xf0800000 },
+	{ _MMIO(0x2720), 0x00000000 },
+	{ _MMIO(0x2770), 0x00000004 },
+	{ _MMIO(0x2774), 0x00000000 },
+	{ _MMIO(0x2778), 0x00000003 },
+	{ _MMIO(0x277c), 0x00000000 },
+	{ _MMIO(0x2780), 0x00000007 },
+	{ _MMIO(0x2784), 0x00000000 },
+	{ _MMIO(0x2788), 0x00100002 },
+	{ _MMIO(0x278c), 0x0000fff7 },
+	{ _MMIO(0x2790), 0x00100002 },
+	{ _MMIO(0x2794), 0x0000ffcf },
+	{ _MMIO(0x2798), 0x00100082 },
+	{ _MMIO(0x279c), 0x0000ffef },
+	{ _MMIO(0x27a0), 0x001000c2 },
+	{ _MMIO(0x27a4), 0x0000ffe7 },
+	{ _MMIO(0x27a8), 0x00100001 },
+	{ _MMIO(0x27ac), 0x0000ffe7 },
+};
+
+static const struct i915_oa_reg flex_eu_config_test_oa[] = {
+};
+
+static const struct i915_oa_reg mux_config_test_oa[] = {
+	{ _MMIO(0x9840), 0x00000080 },
+	{ _MMIO(0x9888), 0x11810000 },
+	{ _MMIO(0x9888), 0x07810013 },
+	{ _MMIO(0x9888), 0x1f810000 },
+	{ _MMIO(0x9888), 0x1d810000 },
+	{ _MMIO(0x9888), 0x1b930040 },
+	{ _MMIO(0x9888), 0x07e54000 },
+	{ _MMIO(0x9888), 0x1f908000 },
+	{ _MMIO(0x9888), 0x11900000 },
+	{ _MMIO(0x9888), 0x37900000 },
+	{ _MMIO(0x9888), 0x53900000 },
+	{ _MMIO(0x9888), 0x45900000 },
+	{ _MMIO(0x9888), 0x33900000 },
+};
+
+static ssize_t
+show_test_oa_id(struct device *kdev, struct device_attribute *attr, char *buf)
+{
+	return sprintf(buf, "1\n");
+}
+
+void
+i915_perf_load_test_config_cflgt2(struct drm_i915_private *dev_priv)
+{
+	strncpy(dev_priv->perf.oa.test_config.uuid,
+		"74fb4902-d3d3-4237-9e90-cbdc68d0a446",
+		UUID_STRING_LEN);
+	dev_priv->perf.oa.test_config.id = 1;
+
+	dev_priv->perf.oa.test_config.mux_regs = mux_config_test_oa;
+	dev_priv->perf.oa.test_config.mux_regs_len = ARRAY_SIZE(mux_config_test_oa);
+
+	dev_priv->perf.oa.test_config.b_counter_regs = b_counter_config_test_oa;
+	dev_priv->perf.oa.test_config.b_counter_regs_len = ARRAY_SIZE(b_counter_config_test_oa);
+
+	dev_priv->perf.oa.test_config.flex_regs = flex_eu_config_test_oa;
+	dev_priv->perf.oa.test_config.flex_regs_len = ARRAY_SIZE(flex_eu_config_test_oa);
+
+	dev_priv->perf.oa.test_config.sysfs_metric.name = "74fb4902-d3d3-4237-9e90-cbdc68d0a446";
+	dev_priv->perf.oa.test_config.sysfs_metric.attrs = dev_priv->perf.oa.test_config.attrs;
+
+	dev_priv->perf.oa.test_config.attrs[0] = &dev_priv->perf.oa.test_config.sysfs_metric_id.attr;
+
+	dev_priv->perf.oa.test_config.sysfs_metric_id.attr.name = "id";
+	dev_priv->perf.oa.test_config.sysfs_metric_id.attr.mode = 0444;
+	dev_priv->perf.oa.test_config.sysfs_metric_id.show = show_test_oa_id;
+}

drivers/gpu/drm/i915/i915_oa_cflgt2.h

@@ -0,0 +1,34 @@
+/*
+ * Autogenerated file by GPU Top : https://github.com/rib/gputop
+ * DO NOT EDIT manually!
+ *
+ *
+ * Copyright (c) 2015 Intel Corporation
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the next
+ * paragraph) shall be included in all copies or substantial portions of the
+ * Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
+ * IN THE SOFTWARE.
+ *
+ */
+
+#ifndef __I915_OA_CFLGT2_H__
+#define __I915_OA_CFLGT2_H__
+
+extern void i915_perf_load_test_config_cflgt2(struct drm_i915_private *dev_priv);
+
+#endif

drivers/gpu/drm/i915/i915_params.c

@@ -25,235 +25,171 @@
 #include "i915_params.h"
 #include "i915_drv.h"
 
-struct i915_params i915 __read_mostly = {
-	.modeset = -1,
-	.panel_ignore_lid = 1,
-	.semaphores = -1,
-	.lvds_channel_mode = 0,
-	.panel_use_ssc = -1,
-	.vbt_sdvo_panel_type = -1,
-	.enable_rc6 = -1,
-	.enable_dc = -1,
-	.enable_fbc = -1,
-	.enable_execlists = -1,
-	.enable_hangcheck = true,
-	.enable_ppgtt = -1,
-	.enable_psr = -1,
-	.alpha_support = IS_ENABLED(CONFIG_DRM_I915_ALPHA_SUPPORT),
-	.disable_power_well = -1,
-	.enable_ips = 1,
-	.fastboot = 0,
-	.prefault_disable = 0,
-	.load_detect_test = 0,
-	.force_reset_modeset_test = 0,
-	.reset = 2,
-	.error_capture = true,
-	.invert_brightness = 0,
-	.disable_display = 0,
-	.enable_cmd_parser = true,
-	.use_mmio_flip = 0,
-	.mmio_debug = 0,
-	.verbose_state_checks = 1,
-	.nuclear_pageflip = 0,
-	.edp_vswing = 0,
-	.enable_guc_loading = 0,
-	.enable_guc_submission = 0,
-	.guc_log_level = -1,
-	.guc_firmware_path = NULL,
-	.huc_firmware_path = NULL,
-	.enable_dp_mst = true,
-	.inject_load_failure = 0,
-	.enable_dpcd_backlight = false,
-	.enable_gvt = false,
+#define i915_param_named(name, T, perm, desc) \
+	module_param_named(name, i915_modparams.name, T, perm); \
+	MODULE_PARM_DESC(name, desc)
+#define i915_param_named_unsafe(name, T, perm, desc) \
+	module_param_named_unsafe(name, i915_modparams.name, T, perm); \
+	MODULE_PARM_DESC(name, desc)
+
+struct i915_params i915_modparams __read_mostly = {
+#define MEMBER(T, member, value) .member = (value),
+	I915_PARAMS_FOR_EACH(MEMBER)
+#undef MEMBER
 };
 
-module_param_named(modeset, i915.modeset, int, 0400);
-MODULE_PARM_DESC(modeset,
+i915_param_named(modeset, int, 0400,
 	"Use kernel modesetting [KMS] (0=disable, "
 	"1=on, -1=force vga console preference [default])");
 
-module_param_named_unsafe(panel_ignore_lid, i915.panel_ignore_lid, int, 0600);
-MODULE_PARM_DESC(panel_ignore_lid,
+i915_param_named_unsafe(panel_ignore_lid, int, 0600,
 	"Override lid status (0=autodetect, 1=autodetect disabled [default], "
 	"-1=force lid closed, -2=force lid open)");
 
-module_param_named_unsafe(semaphores, i915.semaphores, int, 0400);
-MODULE_PARM_DESC(semaphores,
+i915_param_named_unsafe(semaphores, int, 0400,
 	"Use semaphores for inter-ring sync "
 	"(default: -1 (use per-chip defaults))");
 
-module_param_named_unsafe(enable_rc6, i915.enable_rc6, int, 0400);
-MODULE_PARM_DESC(enable_rc6,
+i915_param_named_unsafe(enable_rc6, int, 0400,
 	"Enable power-saving render C-state 6. "
 	"Different stages can be selected via bitmask values "
 	"(0 = disable; 1 = enable rc6; 2 = enable deep rc6; 4 = enable deepest rc6). "
 	"For example, 3 would enable rc6 and deep rc6, and 7 would enable everything. "
 	"default: -1 (use per-chip default)");
 
-module_param_named_unsafe(enable_dc, i915.enable_dc, int, 0400);
-MODULE_PARM_DESC(enable_dc,
+i915_param_named_unsafe(enable_dc, int, 0400,
 	"Enable power-saving display C-states. "
 	"(-1=auto [default]; 0=disable; 1=up to DC5; 2=up to DC6)");
 
-module_param_named_unsafe(enable_fbc, i915.enable_fbc, int, 0600);
-MODULE_PARM_DESC(enable_fbc,
+i915_param_named_unsafe(enable_fbc, int, 0600,
 	"Enable frame buffer compression for power savings "
 	"(default: -1 (use per-chip default))");
 
-module_param_named_unsafe(lvds_channel_mode, i915.lvds_channel_mode, int, 0400);
-MODULE_PARM_DESC(lvds_channel_mode,
+i915_param_named_unsafe(lvds_channel_mode, int, 0400,
 	 "Specify LVDS channel mode "
 	 "(0=probe BIOS [default], 1=single-channel, 2=dual-channel)");
 
-module_param_named_unsafe(lvds_use_ssc, i915.panel_use_ssc, int, 0600);
-MODULE_PARM_DESC(lvds_use_ssc,
+i915_param_named_unsafe(panel_use_ssc, int, 0600,
 	"Use Spread Spectrum Clock with panels [LVDS/eDP] "
 	"(default: auto from VBT)");
 
-module_param_named_unsafe(vbt_sdvo_panel_type, i915.vbt_sdvo_panel_type, int, 0400);
-MODULE_PARM_DESC(vbt_sdvo_panel_type,
+i915_param_named_unsafe(vbt_sdvo_panel_type, int, 0400,
 	"Override/Ignore selection of SDVO panel mode in the VBT "
 	"(-2=ignore, -1=auto [default], index in VBT BIOS table)");
 
-module_param_named_unsafe(reset, i915.reset, int, 0600);
-MODULE_PARM_DESC(reset, "Attempt GPU resets (0=disabled, 1=full gpu reset, 2=engine reset [default])");
+i915_param_named_unsafe(reset, int, 0600,
+	"Attempt GPU resets (0=disabled, 1=full gpu reset, 2=engine reset [default])");
 
-module_param_named_unsafe(vbt_firmware, i915.vbt_firmware, charp, 0400);
-MODULE_PARM_DESC(vbt_firmware,
-		 "Load VBT from specified file under /lib/firmware");
+i915_param_named_unsafe(vbt_firmware, charp, 0400,
+	"Load VBT from specified file under /lib/firmware");
 
 #if IS_ENABLED(CONFIG_DRM_I915_CAPTURE_ERROR)
-module_param_named(error_capture, i915.error_capture, bool, 0600);
-MODULE_PARM_DESC(error_capture,
+i915_param_named(error_capture, bool, 0600,
 	"Record the GPU state following a hang. "
 	"This information in /sys/class/drm/card<N>/error is vital for "
 	"triaging and debugging hangs.");
 #endif
 
-module_param_named_unsafe(enable_hangcheck, i915.enable_hangcheck, bool, 0644);
-MODULE_PARM_DESC(enable_hangcheck,
+i915_param_named_unsafe(enable_hangcheck, bool, 0644,
 	"Periodically check GPU activity for detecting hangs. "
 	"WARNING: Disabling this can cause system wide hangs. "
 	"(default: true)");
 
-module_param_named_unsafe(enable_ppgtt, i915.enable_ppgtt, int, 0400);
-MODULE_PARM_DESC(enable_ppgtt,
+i915_param_named_unsafe(enable_ppgtt, int, 0400,
 	"Override PPGTT usage. "
 	"(-1=auto [default], 0=disabled, 1=aliasing, 2=full, 3=full with extended address space)");
 
-module_param_named_unsafe(enable_execlists, i915.enable_execlists, int, 0400);
-MODULE_PARM_DESC(enable_execlists,
+i915_param_named_unsafe(enable_execlists, int, 0400,
 	"Override execlists usage. "
 	"(-1=auto [default], 0=disabled, 1=enabled)");
 
-module_param_named_unsafe(enable_psr, i915.enable_psr, int, 0600);
-MODULE_PARM_DESC(enable_psr, "Enable PSR "
-		 "(0=disabled, 1=enabled - link mode chosen per-platform, 2=force link-standby mode, 3=force link-off mode) "
-		 "Default: -1 (use per-chip default)");
+i915_param_named_unsafe(enable_psr, int, 0600,
+	"Enable PSR "
+	"(0=disabled, 1=enabled - link mode chosen per-platform, 2=force link-standby mode, 3=force link-off mode) "
+	"Default: -1 (use per-chip default)");
 
-module_param_named_unsafe(alpha_support, i915.alpha_support, bool, 0400);
-MODULE_PARM_DESC(alpha_support,
+i915_param_named_unsafe(alpha_support, bool, 0400,
 	"Enable alpha quality driver support for latest hardware. "
 	"See also CONFIG_DRM_I915_ALPHA_SUPPORT.");
 
-module_param_named_unsafe(disable_power_well, i915.disable_power_well, int, 0400);
-MODULE_PARM_DESC(disable_power_well,
+i915_param_named_unsafe(disable_power_well, int, 0400,
 	"Disable display power wells when possible "
 	"(-1=auto [default], 0=power wells always on, 1=power wells disabled when possible)");
 
-module_param_named_unsafe(enable_ips, i915.enable_ips, int, 0600);
-MODULE_PARM_DESC(enable_ips, "Enable IPS (default: true)");
+i915_param_named_unsafe(enable_ips, int, 0600, "Enable IPS (default: true)");
 
-module_param_named(fastboot, i915.fastboot, bool, 0600);
-MODULE_PARM_DESC(fastboot,
+i915_param_named(fastboot, bool, 0600,
 	"Try to skip unnecessary mode sets at boot time (default: false)");
 
-module_param_named_unsafe(prefault_disable, i915.prefault_disable, bool, 0600);
-MODULE_PARM_DESC(prefault_disable,
+i915_param_named_unsafe(prefault_disable, bool, 0600,
 	"Disable page prefaulting for pread/pwrite/reloc (default:false). "
 	"For developers only.");
 
-module_param_named_unsafe(load_detect_test, i915.load_detect_test, bool, 0600);
-MODULE_PARM_DESC(load_detect_test,
+i915_param_named_unsafe(load_detect_test, bool, 0600,
 	"Force-enable the VGA load detect code for testing (default:false). "
 	"For developers only.");
 
-module_param_named_unsafe(force_reset_modeset_test, i915.force_reset_modeset_test, bool, 0600);
-MODULE_PARM_DESC(force_reset_modeset_test,
+i915_param_named_unsafe(force_reset_modeset_test, bool, 0600,
 	"Force a modeset during gpu reset for testing (default:false). "
 	"For developers only.");
 
-module_param_named_unsafe(invert_brightness, i915.invert_brightness, int, 0600);
-MODULE_PARM_DESC(invert_brightness,
+i915_param_named_unsafe(invert_brightness, int, 0600,
 	"Invert backlight brightness "
 	"(-1 force normal, 0 machine defaults, 1 force inversion), please "
 	"report PCI device ID, subsystem vendor and subsystem device ID "
 	"to dri-devel@lists.freedesktop.org, if your machine needs it. "
 	"It will then be included in an upcoming module version.");
 
-module_param_named(disable_display, i915.disable_display, bool, 0400);
-MODULE_PARM_DESC(disable_display, "Disable display (default: false)");
+i915_param_named(disable_display, bool, 0400,
+	"Disable display (default: false)");
 
-module_param_named_unsafe(enable_cmd_parser, i915.enable_cmd_parser, bool, 0400);
-MODULE_PARM_DESC(enable_cmd_parser,
-		 "Enable command parsing (true=enabled [default], false=disabled)");
+i915_param_named_unsafe(enable_cmd_parser, bool, 0400,
+	"Enable command parsing (true=enabled [default], false=disabled)");
 
-module_param_named_unsafe(use_mmio_flip, i915.use_mmio_flip, int, 0600);
-MODULE_PARM_DESC(use_mmio_flip,
-		 "use MMIO flips (-1=never, 0=driver discretion [default], 1=always)");
+i915_param_named_unsafe(use_mmio_flip, int, 0600,
+	"use MMIO flips (-1=never, 0=driver discretion [default], 1=always)");
 
-module_param_named(mmio_debug, i915.mmio_debug, int, 0600);
-MODULE_PARM_DESC(mmio_debug,
+i915_param_named(mmio_debug, int, 0600,
 	"Enable the MMIO debug code for the first N failures (default: off). "
 	"This may negatively affect performance.");
 
-module_param_named(verbose_state_checks, i915.verbose_state_checks, bool, 0600);
-MODULE_PARM_DESC(verbose_state_checks,
+i915_param_named(verbose_state_checks, bool, 0600,
 	"Enable verbose logs (ie. WARN_ON()) in case of unexpected hw state conditions.");
 
-module_param_named_unsafe(nuclear_pageflip, i915.nuclear_pageflip, bool, 0400);
-MODULE_PARM_DESC(nuclear_pageflip,
-		 "Force enable atomic functionality on platforms that don't have full support yet.");
+i915_param_named_unsafe(nuclear_pageflip, bool, 0400,
+	"Force enable atomic functionality on platforms that don't have full support yet.");
 
 /* WA to get away with the default setting in VBT for early platforms.Will be removed */
-module_param_named_unsafe(edp_vswing, i915.edp_vswing, int, 0400);
-MODULE_PARM_DESC(edp_vswing,
-		 "Ignore/Override vswing pre-emph table selection from VBT "
-		 "(0=use value from vbt [default], 1=low power swing(200mV),"
-		 "2=default swing(400mV))");
-
-module_param_named_unsafe(enable_guc_loading, i915.enable_guc_loading, int, 0400);
-MODULE_PARM_DESC(enable_guc_loading,
-		"Enable GuC firmware loading "
-		"(-1=auto, 0=never [default], 1=if available, 2=required)");
-
-module_param_named_unsafe(enable_guc_submission, i915.enable_guc_submission, int, 0400);
-MODULE_PARM_DESC(enable_guc_submission,
-		"Enable GuC submission "
-		"(-1=auto, 0=never [default], 1=if available, 2=required)");
-
-module_param_named(guc_log_level, i915.guc_log_level, int, 0400);
-MODULE_PARM_DESC(guc_log_level,
+i915_param_named_unsafe(edp_vswing, int, 0400,
+	"Ignore/Override vswing pre-emph table selection from VBT "
+	"(0=use value from vbt [default], 1=low power swing(200mV),"
+	"2=default swing(400mV))");
+
+i915_param_named_unsafe(enable_guc_loading, int, 0400,
+	"Enable GuC firmware loading "
+	"(-1=auto, 0=never [default], 1=if available, 2=required)");
+
+i915_param_named_unsafe(enable_guc_submission, int, 0400,
+	"Enable GuC submission "
+	"(-1=auto, 0=never [default], 1=if available, 2=required)");
+
+i915_param_named(guc_log_level, int, 0400,
 	"GuC firmware logging level (-1:disabled (default), 0-3:enabled)");
 
-module_param_named_unsafe(guc_firmware_path, i915.guc_firmware_path, charp, 0400);
-MODULE_PARM_DESC(guc_firmware_path,
+i915_param_named_unsafe(guc_firmware_path, charp, 0400,
 	"GuC firmware path to use instead of the default one");
 
-module_param_named_unsafe(huc_firmware_path, i915.huc_firmware_path, charp, 0400);
-MODULE_PARM_DESC(huc_firmware_path,
+i915_param_named_unsafe(huc_firmware_path, charp, 0400,
 	"HuC firmware path to use instead of the default one");
 
-module_param_named_unsafe(enable_dp_mst, i915.enable_dp_mst, bool, 0600);
-MODULE_PARM_DESC(enable_dp_mst,
+i915_param_named_unsafe(enable_dp_mst, bool, 0600,
 	"Enable multi-stream transport (MST) for new DisplayPort sinks. (default: true)");
-module_param_named_unsafe(inject_load_failure, i915.inject_load_failure, uint, 0400);
-MODULE_PARM_DESC(inject_load_failure,
+
+i915_param_named_unsafe(inject_load_failure, uint, 0400,
 	"Force an error after a number of failure check points (0:disabled (default), N:force failure at the Nth failure check point)");
-module_param_named(enable_dpcd_backlight, i915.enable_dpcd_backlight, bool, 0600);
-MODULE_PARM_DESC(enable_dpcd_backlight,
+
+i915_param_named(enable_dpcd_backlight, bool, 0600,
 	"Enable support for DPCD backlight control (default:false)");
 
-module_param_named(enable_gvt, i915.enable_gvt, bool, 0400);
-MODULE_PARM_DESC(enable_gvt,
+i915_param_named(enable_gvt, bool, 0400,
 	"Enable support for Intel GVT-g graphics virtualization host support(default:false)");

drivers/gpu/drm/i915/i915_params.h

@@ -27,56 +27,56 @@
 
 #include <linux/cache.h> /* for __read_mostly */
 
-#define I915_PARAMS_FOR_EACH(func) \
-	func(char *, vbt_firmware); \
-	func(int, modeset); \
-	func(int, panel_ignore_lid); \
-	func(int, semaphores); \
-	func(int, lvds_channel_mode); \
-	func(int, panel_use_ssc); \
-	func(int, vbt_sdvo_panel_type); \
-	func(int, enable_rc6); \
-	func(int, enable_dc); \
-	func(int, enable_fbc); \
-	func(int, enable_ppgtt); \
-	func(int, enable_execlists); \
-	func(int, enable_psr); \
-	func(int, disable_power_well); \
-	func(int, enable_ips); \
-	func(int, invert_brightness); \
-	func(int, enable_guc_loading); \
-	func(int, enable_guc_submission); \
-	func(int, guc_log_level); \
-	func(char *, guc_firmware_path); \
-	func(char *, huc_firmware_path); \
-	func(int, use_mmio_flip); \
-	func(int, mmio_debug); \
-	func(int, edp_vswing); \
-	func(int, reset); \
-	func(unsigned int, inject_load_failure); \
+#define I915_PARAMS_FOR_EACH(param) \
+	param(char *, vbt_firmware, NULL) \
+	param(int, modeset, -1) \
+	param(int, panel_ignore_lid, 1) \
+	param(int, semaphores, -1) \
+	param(int, lvds_channel_mode, 0) \
+	param(int, panel_use_ssc, -1) \
+	param(int, vbt_sdvo_panel_type, -1) \
+	param(int, enable_rc6, -1) \
+	param(int, enable_dc, -1) \
+	param(int, enable_fbc, -1) \
+	param(int, enable_ppgtt, -1) \
+	param(int, enable_execlists, -1) \
+	param(int, enable_psr, -1) \
+	param(int, disable_power_well, -1) \
+	param(int, enable_ips, 1) \
+	param(int, invert_brightness, 0) \
+	param(int, enable_guc_loading, 0) \
+	param(int, enable_guc_submission, 0) \
+	param(int, guc_log_level, -1) \
+	param(char *, guc_firmware_path, NULL) \
+	param(char *, huc_firmware_path, NULL) \
+	param(int, use_mmio_flip, 0) \
+	param(int, mmio_debug, 0) \
+	param(int, edp_vswing, 0) \
+	param(int, reset, 2) \
+	param(unsigned int, inject_load_failure, 0) \
 	/* leave bools at the end to not create holes */ \
-	func(bool, alpha_support); \
-	func(bool, enable_cmd_parser); \
-	func(bool, enable_hangcheck); \
-	func(bool, fastboot); \
-	func(bool, prefault_disable); \
-	func(bool, load_detect_test); \
-	func(bool, force_reset_modeset_test); \
-	func(bool, error_capture); \
-	func(bool, disable_display); \
-	func(bool, verbose_state_checks); \
-	func(bool, nuclear_pageflip); \
-	func(bool, enable_dp_mst); \
-	func(bool, enable_dpcd_backlight); \
-	func(bool, enable_gvt)
+	param(bool, alpha_support, IS_ENABLED(CONFIG_DRM_I915_ALPHA_SUPPORT)) \
+	param(bool, enable_cmd_parser, true) \
+	param(bool, enable_hangcheck, true) \
+	param(bool, fastboot, false) \
+	param(bool, prefault_disable, false) \
+	param(bool, load_detect_test, false) \
+	param(bool, force_reset_modeset_test, false) \
+	param(bool, error_capture, true) \
+	param(bool, disable_display, false) \
+	param(bool, verbose_state_checks, true) \
+	param(bool, nuclear_pageflip, false) \
+	param(bool, enable_dp_mst, true) \
+	param(bool, enable_dpcd_backlight, false) \
+	param(bool, enable_gvt, false)
 
-#define MEMBER(T, member) T member
+#define MEMBER(T, member, ...) T member;
 struct i915_params {
 	I915_PARAMS_FOR_EACH(MEMBER);
 };
 #undef MEMBER
 
-extern struct i915_params i915 __read_mostly;
+extern struct i915_params i915_modparams __read_mostly;
 
 #endif
 

drivers/gpu/drm/i915/i915_pci.c

@@ -168,6 +168,7 @@ static const struct intel_device_info intel_i965g_info __initconst = {
 	.platform = INTEL_I965G,
 	.has_overlay = 1,
 	.hws_needs_physical = 1,
+	.has_snoop = false,
 };
 
 static const struct intel_device_info intel_i965gm_info __initconst = {
@@ -177,6 +178,7 @@ static const struct intel_device_info intel_i965gm_info __initconst = {
 	.has_overlay = 1,
 	.supports_tv = 1,
 	.hws_needs_physical = 1,
+	.has_snoop = false,
 };
 
 static const struct intel_device_info intel_g45_info __initconst = {
@@ -198,7 +200,6 @@ static const struct intel_device_info intel_gm45_info __initconst = {
 #define GEN5_FEATURES \
 	.gen = 5, .num_pipes = 2, \
 	.has_hotplug = 1, \
-	.has_gmbus_irq = 1, \
 	.ring_mask = RENDER_RING | BSD_RING, \
 	.has_snoop = true, \
 	GEN_DEFAULT_PIPEOFFSETS, \
@@ -223,7 +224,6 @@ static const struct intel_device_info intel_ironlake_m_info __initconst = {
 	.has_llc = 1, \
 	.has_rc6 = 1, \
 	.has_rc6p = 1, \
-	.has_gmbus_irq = 1, \
 	.has_aliasing_ppgtt = 1, \
 	GEN_DEFAULT_PIPEOFFSETS, \
 	CURSOR_OFFSETS
@@ -266,7 +266,6 @@ static const struct intel_device_info intel_sandybridge_m_gt2_info __initconst =
 	.has_llc = 1, \
 	.has_rc6 = 1, \
 	.has_rc6p = 1, \
-	.has_gmbus_irq = 1, \
 	.has_aliasing_ppgtt = 1, \
 	.has_full_ppgtt = 1, \
 	GEN_DEFAULT_PIPEOFFSETS, \
@@ -319,7 +318,6 @@ static const struct intel_device_info intel_valleyview_info __initconst = {
 	.has_psr = 1,
 	.has_runtime_pm = 1,
 	.has_rc6 = 1,
-	.has_gmbus_irq = 1,
 	.has_gmch_display = 1,
 	.has_hotplug = 1,
 	.has_aliasing_ppgtt = 1,
@@ -410,7 +408,6 @@ static const struct intel_device_info intel_cherryview_info __initconst = {
 	.has_runtime_pm = 1,
 	.has_resource_streamer = 1,
 	.has_rc6 = 1,
-	.has_gmbus_irq = 1,
 	.has_logical_ring_contexts = 1,
 	.has_gmch_display = 1,
 	.has_aliasing_ppgtt = 1,
@@ -472,7 +469,6 @@ static const struct intel_device_info intel_skylake_gt4_info __initconst = {
 	.has_resource_streamer = 1, \
 	.has_rc6 = 1, \
 	.has_dp_mst = 1, \
-	.has_gmbus_irq = 1, \
 	.has_logical_ring_contexts = 1, \
 	.has_guc = 1, \
 	.has_aliasing_ppgtt = 1, \
@@ -480,6 +476,7 @@ static const struct intel_device_info intel_skylake_gt4_info __initconst = {
 	.has_full_48bit_ppgtt = 1, \
 	.has_reset_engine = 1, \
 	.has_snoop = true, \
+	.has_ipc = 1, \
 	GEN_DEFAULT_PIPEOFFSETS, \
 	IVB_CURSOR_OFFSETS, \
 	BDW_COLORS
@@ -503,6 +500,7 @@ static const struct intel_device_info intel_geminilake_info __initconst = {
 	.platform = INTEL_KABYLAKE, \
 	.has_csr = 1, \
 	.has_guc = 1, \
+	.has_ipc = 1, \
 	.ddb_size = 896
 
 static const struct intel_device_info intel_kabylake_gt1_info __initconst = {
@@ -522,12 +520,12 @@ static const struct intel_device_info intel_kabylake_gt3_info __initconst = {
 };
 
 #define CFL_PLATFORM \
-	.is_alpha_support = 1, \
 	BDW_FEATURES, \
 	.gen = 9, \
 	.platform = INTEL_COFFEELAKE, \
 	.has_csr = 1, \
 	.has_guc = 1, \
+	.has_ipc = 1, \
 	.ddb_size = 896
 
 static const struct intel_device_info intel_coffeelake_gt1_info __initconst = {
@@ -554,6 +552,7 @@ static const struct intel_device_info intel_cannonlake_gt2_info __initconst = {
 	.gt = 2,
 	.ddb_size = 1024,
 	.has_csr = 1,
+	.has_ipc = 1,
 	.color = { .degamma_lut_size = 0, .gamma_lut_size = 1024 }
 };
 
@@ -632,7 +631,7 @@ static int i915_pci_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 		(struct intel_device_info *) ent->driver_data;
 	int err;
 
-	if (IS_ALPHA_SUPPORT(intel_info) && !i915.alpha_support) {
+	if (IS_ALPHA_SUPPORT(intel_info) && !i915_modparams.alpha_support) {
 		DRM_INFO("The driver support for your hardware in this kernel version is alpha quality\n"
 			 "See CONFIG_DRM_I915_ALPHA_SUPPORT or i915.alpha_support module parameter\n"
 			 "to enable support in this kernel version, or check for kernel updates.\n");
@@ -690,10 +689,10 @@ static int __init i915_init(void)
 	 * vga_text_mode_force boot option.
 	 */
 
-	if (i915.modeset == 0)
+	if (i915_modparams.modeset == 0)
 		use_kms = false;
 
-	if (vgacon_text_force() && i915.modeset == -1)
+	if (vgacon_text_force() && i915_modparams.modeset == -1)
 		use_kms = false;
 
 	if (!use_kms) {

drivers/gpu/drm/i915/i915_perf.c

@@ -206,6 +206,7 @@
 #include "i915_oa_kblgt2.h"
 #include "i915_oa_kblgt3.h"
 #include "i915_oa_glk.h"
+#include "i915_oa_cflgt2.h"
 
 /* HW requires this to be a power of two, between 128k and 16M, though driver
  * is currently generally designed assuming the largest 16M size is used such
@@ -1213,7 +1214,7 @@ static int oa_get_render_ctx_id(struct i915_perf_stream *stream)
 {
 	struct drm_i915_private *dev_priv = stream->dev_priv;
 
-	if (i915.enable_execlists)
+	if (i915_modparams.enable_execlists)
 		dev_priv->perf.oa.specific_ctx_id = stream->ctx->hw_id;
 	else {
 		struct intel_engine_cs *engine = dev_priv->engine[RCS];
@@ -1259,7 +1260,7 @@ static void oa_put_render_ctx_id(struct i915_perf_stream *stream)
 {
 	struct drm_i915_private *dev_priv = stream->dev_priv;
 
-	if (i915.enable_execlists) {
+	if (i915_modparams.enable_execlists) {
 		dev_priv->perf.oa.specific_ctx_id = INVALID_CTX_ID;
 	} else {
 		struct intel_engine_cs *engine = dev_priv->engine[RCS];
@@ -1850,8 +1851,7 @@ static int gen8_enable_metric_set(struct drm_i915_private *dev_priv,
 	 * be read back from automatically triggered reports, as part of the
 	 * RPT_ID field.
 	 */
-	if (IS_SKYLAKE(dev_priv) || IS_BROXTON(dev_priv) ||
-	    IS_KABYLAKE(dev_priv) || IS_GEMINILAKE(dev_priv)) {
+	if (IS_GEN9(dev_priv)) {
 		I915_WRITE(GEN8_OA_DEBUG,
 			   _MASKED_BIT_ENABLE(GEN9_OA_DEBUG_DISABLE_CLK_RATIO_REPORTS |
 					      GEN9_OA_DEBUG_INCLUDE_CLK_RATIO));
@@ -2927,6 +2927,9 @@ void i915_perf_register(struct drm_i915_private *dev_priv)
 			i915_perf_load_test_config_kblgt3(dev_priv);
 	} else if (IS_GEMINILAKE(dev_priv)) {
 		i915_perf_load_test_config_glk(dev_priv);
+	} else if (IS_COFFEELAKE(dev_priv)) {
+		if (IS_CFL_GT2(dev_priv))
+			i915_perf_load_test_config_cflgt2(dev_priv);
 	}
 
 	if (dev_priv->perf.oa.test_config.id == 0)
@@ -3405,7 +3408,7 @@ void i915_perf_init(struct drm_i915_private *dev_priv)
 		dev_priv->perf.oa.timestamp_frequency = 12500000;
 
 		dev_priv->perf.oa.oa_formats = hsw_oa_formats;
-	} else if (i915.enable_execlists) {
+	} else if (i915_modparams.enable_execlists) {
 		/* Note: that although we could theoretically also support the
 		 * legacy ringbuffer mode on BDW (and earlier iterations of
 		 * this driver, before upstreaming did this) it didn't seem
@@ -3453,6 +3456,7 @@ void i915_perf_init(struct drm_i915_private *dev_priv)
 				break;
 			case INTEL_SKYLAKE:
 			case INTEL_KABYLAKE:
+			case INTEL_COFFEELAKE:
 				dev_priv->perf.oa.timestamp_frequency = 12000000;
 				break;
 			default:

drivers/gpu/drm/i915/i915_reg.h

@@ -2336,7 +2336,7 @@ enum i915_power_well_id {
 #define DONE_REG		_MMIO(0x40b0)
 #define GEN8_PRIVATE_PAT_LO	_MMIO(0x40e0)
 #define GEN8_PRIVATE_PAT_HI	_MMIO(0x40e0 + 4)
-#define GEN10_PAT_INDEX(index)	_MMIO(0x40e0 + index*4)
+#define GEN10_PAT_INDEX(index)	_MMIO(0x40e0 + (index)*4)
 #define BSD_HWS_PGA_GEN7	_MMIO(0x04180)
 #define BLT_HWS_PGA_GEN7	_MMIO(0x04280)
 #define VEBOX_HWS_PGA_GEN7	_MMIO(0x04380)
@@ -2730,6 +2730,11 @@ enum i915_power_well_id {
 #define   GEN9_F2_SS_DIS_SHIFT		20
 #define   GEN9_F2_SS_DIS_MASK		(0xf << GEN9_F2_SS_DIS_SHIFT)
 
+#define   GEN10_F2_S_ENA_SHIFT		22
+#define   GEN10_F2_S_ENA_MASK		(0x3f << GEN10_F2_S_ENA_SHIFT)
+#define   GEN10_F2_SS_DIS_SHIFT		18
+#define   GEN10_F2_SS_DIS_MASK		(0xf << GEN10_F2_SS_DIS_SHIFT)
+
 #define GEN8_EU_DISABLE0		_MMIO(0x9134)
 #define   GEN8_EU_DIS0_S0_MASK		0xffffff
 #define   GEN8_EU_DIS0_S1_SHIFT		24
@@ -2745,6 +2750,9 @@ enum i915_power_well_id {
 
 #define GEN9_EU_DISABLE(slice)		_MMIO(0x9134 + (slice)*0x4)
 
+#define GEN10_EU_DISABLE3		_MMIO(0x9140)
+#define   GEN10_EU_DIS_SS_MASK		0xff
+
 #define GEN6_BSD_SLEEP_PSMI_CONTROL	_MMIO(0x12050)
 #define   GEN6_BSD_SLEEP_MSG_DISABLE	(1 << 0)
 #define   GEN6_BSD_SLEEP_FLUSH_DISABLE	(1 << 2)
@@ -4047,7 +4055,7 @@ enum {
 #define   EDP_PSR2_FRAME_BEFORE_SU_SHIFT 4
 #define   EDP_PSR2_FRAME_BEFORE_SU_MASK	(0xf<<4)
 #define   EDP_PSR2_IDLE_MASK		0xf
-#define   EDP_FRAMES_BEFORE_SU_ENTRY   (1<<4)
+#define   EDP_PSR2_FRAME_BEFORE_SU(a)	((a)<<4)
 
 #define EDP_PSR2_STATUS_CTL            _MMIO(0x6f940)
 #define EDP_PSR2_STATUS_STATE_MASK     (0xf<<28)
@@ -6913,7 +6921,7 @@ enum {
 # define CHICKEN3_DGMG_DONE_FIX_DISABLE		(1 << 2)
 
 #define CHICKEN_PAR1_1		_MMIO(0x42080)
-#define  SKL_RC_HASH_OUTSIDE	(1 << 15)
+#define  SKL_DE_COMPRESSED_HASH_MODE	(1 << 15)
 #define  DPA_MASK_VBLANK_SRD	(1 << 15)
 #define  FORCE_ARB_IDLE_PLANES	(1 << 14)
 #define  SKL_EDP_PSR_FIX_RDWRAP	(1 << 3)
@@ -6949,6 +6957,7 @@ enum {
 #define  DISP_FBC_WM_DIS		(1<<15)
 #define DISP_ARB_CTL2	_MMIO(0x45004)
 #define  DISP_DATA_PARTITION_5_6	(1<<6)
+#define  DISP_IPC_ENABLE		(1<<3)
 #define DBUF_CTL	_MMIO(0x45008)
 #define  DBUF_POWER_REQUEST		(1<<31)
 #define  DBUF_POWER_STATE		(1<<30)
@@ -6990,6 +6999,7 @@ enum {
 # define GEN7_CSC1_RHWO_OPT_DISABLE_IN_RCC	((1<<10) | (1<<26))
 # define GEN9_RHWO_OPTIMIZATION_DISABLE		(1<<14)
 #define COMMON_SLICE_CHICKEN2			_MMIO(0x7014)
+# define GEN9_PBE_COMPRESSED_HASH_SELECTION	(1<<13)
 # define GEN9_DISABLE_GATHER_AT_SET_SHADER_COMMON_SLICE (1<<12)
 # define GEN8_SBE_DISABLE_REPLAY_BUF_OPTIMIZATION (1<<8)
 # define GEN8_CSC2_SBE_VUE_CACHE_CONSERVATIVE	(1<<0)
@@ -7469,6 +7479,8 @@ enum {
 #define  FDI_PHASE_SYNC_OVR(pipe) (1<<(FDIA_PHASE_SYNC_SHIFT_OVR - ((pipe) * 2)))
 #define  FDI_PHASE_SYNC_EN(pipe) (1<<(FDIA_PHASE_SYNC_SHIFT_EN - ((pipe) * 2)))
 #define  FDI_BC_BIFURCATION_SELECT	(1 << 12)
+#define  CHASSIS_CLK_REQ_DURATION_MASK	(0xf << 8)
+#define  CHASSIS_CLK_REQ_DURATION(x)	((x) << 8)
 #define  SPT_PWM_GRANULARITY		(1<<0)
 #define SOUTH_CHICKEN2		_MMIO(0xc2004)
 #define  FDI_MPHY_IOSFSB_RESET_STATUS	(1<<13)
@@ -7953,8 +7965,8 @@ enum {
 #define     GEN7_PCODE_TIMEOUT			0x2
 #define     GEN7_PCODE_ILLEGAL_DATA		0x3
 #define     GEN7_PCODE_MIN_FREQ_TABLE_GT_RATIO_OUT_OF_RANGE 0x10
-#define	  GEN6_PCODE_WRITE_RC6VIDS		0x4
-#define	  GEN6_PCODE_READ_RC6VIDS		0x5
+#define   GEN6_PCODE_WRITE_RC6VIDS		0x4
+#define   GEN6_PCODE_READ_RC6VIDS		0x5
 #define     GEN6_ENCODE_RC6_VID(mv)		(((mv) - 245) / 5)
 #define     GEN6_DECODE_RC6_VID(vids)		(((vids) * 5) + 245)
 #define   BDW_PCODE_DISPLAY_FREQ_CHANGE_REQ	0x18
@@ -7973,7 +7985,9 @@ enum {
 #define   GEN6_PCODE_WRITE_D_COMP		0x11
 #define   HSW_PCODE_DE_WRITE_FREQ_REQ		0x17
 #define   DISPLAY_IPS_CONTROL			0x19
-#define	  HSW_PCODE_DYNAMIC_DUTY_CYCLE_CONTROL	0x1A
+            /* See also IPS_CTL */
+#define     IPS_PCODE_CONTROL			(1 << 30)
+#define   HSW_PCODE_DYNAMIC_DUTY_CYCLE_CONTROL	0x1A
 #define   GEN9_PCODE_SAGV_CONTROL		0x21
 #define     GEN9_SAGV_DISABLE			0x0
 #define     GEN9_SAGV_IS_DISABLED		0x1
@@ -8082,6 +8096,7 @@ enum {
 #define   GEN8_SAMPLER_POWER_BYPASS_DIS	(1<<1)
 
 #define GEN9_HALF_SLICE_CHICKEN7	_MMIO(0xe194)
+#define   GEN9_SAMPLER_HASH_COMPRESSED_READ_ADDR	(1<<8)
 #define   GEN9_ENABLE_YV12_BUGFIX	(1<<4)
 #define   GEN9_ENABLE_GPGPU_PREEMPTION	(1<<2)
 
@@ -8594,7 +8609,7 @@ enum skl_power_gate {
 #define  DPLL_CFGCR0_LINK_RATE_3240	(6 << 25)
 #define  DPLL_CFGCR0_LINK_RATE_4050	(7 << 25)
 #define  DPLL_CFGCR0_DCO_FRACTION_MASK	(0x7fff << 10)
-#define  DPLL_CFGCR0_DCO_FRAC_SHIFT	(10)
+#define  DPLL_CFGCR0_DCO_FRACTION_SHIFT	(10)
 #define  DPLL_CFGCR0_DCO_FRACTION(x)	((x) << 10)
 #define  DPLL_CFGCR0_DCO_INTEGER_MASK	(0x3ff)
 #define CNL_DPLL_CFGCR0(pll)		_MMIO_PLL(pll, _CNL_DPLL0_CFGCR0, _CNL_DPLL1_CFGCR0)
@@ -8801,6 +8816,15 @@ enum skl_power_gate {
 #define MIPIO_TXESC_CLK_DIV2			_MMIO(0x160008)
 #define  GLK_TX_ESC_CLK_DIV2_MASK			0x3FF
 
+/* Gen4+ Timestamp and Pipe Frame time stamp registers */
+#define GEN4_TIMESTAMP		_MMIO(0x2358)
+#define ILK_TIMESTAMP_HI	_MMIO(0x70070)
+#define IVB_TIMESTAMP_CTR	_MMIO(0x44070)
+
+#define _PIPE_FRMTMSTMP_A		0x70048
+#define PIPE_FRMTMSTMP(pipe)		\
+			_MMIO_PIPE2(pipe, _PIPE_FRMTMSTMP_A)
+
 /* BXT MIPI clock controls */
 #define BXT_MAX_VAR_OUTPUT_KHZ			39500
 
@@ -9382,4 +9406,8 @@ enum skl_power_gate {
 #define   GEN9_L3_LRA_1_GPGPU_DEFAULT_VALUE_SKL  0x67F1427F /*    "        " */
 #define   GEN9_L3_LRA_1_GPGPU_DEFAULT_VALUE_BXT  0x5FF101FF /*    "        " */
 
+#define MMCD_MISC_CTRL		_MMIO(0x4ddc) /* skl+ */
+#define  MMCD_PCLA		(1 << 31)
+#define  MMCD_HOTSPOT_EN	(1 << 27)
+
 #endif /* _I915_REG_H_ */