drm pull request for 5.2

-----BEGIN PGP SIGNATURE----- iQIcBAABAgAGBQJc04M6AAoJEAx081l5xIa+SJgP/0uIgIOM53vPpydgmr+2IEHF jbDqrd+mipgNriRVHjDsWdUHCUNtyhB7YEBCMrj3mY0rRFI7FlQQf4lOwYGoHiKP 4JZg4kwC37997lFXl1uabGj3DmJLtxKL2/D15zCH/uLe+2EDzWznP6NVdFT3WK0P YKZQCWT19PWSsLoBRPutWxkmop4AYvkqE0a6vXUlJlFYZK3Bbytx6/179uWKfiX5 ZkKEEtx1XiDAvcp5gBb6PISurycrBY0e/bkPBnK3ES5vawMbTU5IrmWOrQ4D8yOd z9qOVZawZ6+b2XBDgBWjQ9bM7I5R7Il1q/LglYEaFI9+wHUnlUdDSm6ft5/5BiCZ fqgkh5Bj2iEsajbSsacoljMOpxpYPqj63mqc+7fAGXF34V+B+9U1bpt8kCbMKowf 7Abb7IuiCR6vLDapjP6VqTMvdQ4O466OEAN83ULGFTdmMqYYH4AxaIwc+xcAk/aP RNq7/RHhh4FRynRAj9fCkGlF3ArnM88gLINwWuEQq4SClWGcvdw7eaHpwWo77c4g iccCnTLqSIg5pDVu07AQzzBlW6KulWxh5o72x+Xx+EXWdYUDHQ1SlNs11bSNUBV1 5MkrzY2GuD+NFEjsXJEDIPOr40mQOyJCXnxq8nXPsz/hD9kHeJPvWn3J3eVKyb5B Z6/knNqM0BDn3SaYR/rD =YFiQ -----END PGP SIGNATURE----- Merge tag 'drm-next-2019-05-09' of git://anongit.freedesktop.org/drm/drm Pull drm updates from Dave Airlie: "This has two exciting community drivers for ARM Mali accelerators. Since ARM has never been open source friendly on the GPU side of the house, the community has had to create open source drivers for the Mali GPUs. Lima covers the older t4xx and panfrost the newer 6xx/7xx series. Well done to all involved and hopefully this will help ARM head in the right direction. There is also now the ability if you don't have any of the legacy drivers enabled (pre-KMS) to remove all the pre-KMS support code from the core drm, this saves 10% or so in codesize on my machine. i915 also enable Icelake/Elkhart Lake Gen11 GPUs by default, vboxvideo moves out of staging. There are also some rcar-du patches which crossover with media tree but all should be acked by Mauro. Summary: uapi changes: - Colorspace connector property - fourcc - new YUV formts - timeline sync objects initially merged - expose FB_DAMAGE_CLIPS to atomic userspace new drivers: - vboxvideo: moved out of staging - aspeed: ASPEED SoC BMC chip display support - lima: ARM Mali4xx GPU acceleration driver support - panfrost: ARM Mali6xx/7xx Midgard/Bitfrost acceleration driver support core: - component helper docs - unplugging fixes - devm device init - MIPI/DSI rate control - shmem backed gem objects - connector, display_info, edid_quirks cleanups - dma_buf fence chain support - 64-bit dma-fence seqno comparison fixes - move initial fb config code to core - gem fence array helpers for Lima - ability to remove legacy support code if no drivers requires it (removes 10% of drm.ko size) - lease fixes ttm: - unified DRM_FILE_PAGE_OFFSET handling - Account for kernel allocations in kernel zone only panel: - OSD070T1718-19TS panel support - panel-tpo-td028ttec1 backlight support - Ronbo RB070D30 MIPI/DSI - Feiyang FY07024DI26A30-D MIPI-DSI panel - Rocktech jh057n00900 MIPI-DSI panel i915: - Comet Lake (Gen9) PCI IDs - Updated Icelake PCI IDs - Elkhartlake (Gen11) support - DP MST property addtions - plane and watermark fixes - Icelake port sync and VEBOX disable fixes - struct_mutex usage reduction - Icelake gamma fix - GuC reset fixes - make mmap more asynchronous - sound display power well race fixes - DDI/MIPI-DSI clocks for Icelake - Icelake RPS frequency changing support - Icelake workarounds amdgpu: - Use HMM for userptr - vega20 experimental smu11 support - RAS support for vega20 - BACO support for vega12 + fixes for vega20 - reworked IH interrupt handling - amdkfd RAS support - Freesync improvements - initial timeline sync object support - DC Z ordering fixes - NV12 planes support - colorspace properties for planes= - eDP opts if eDP already initialized nouveau: - misc fixes etnaviv: - misc fixes msm: - GPU zap shader support expansion - robustness ABI addition exynos: - Logging cleanups tegra: - Shared reset fix - CPU cache maintenance fix cirrus: - driver rewritten using simple helpers meson: - G12A support vmwgfx: - Resource dirtying management improvements - Userspace logging improvements virtio: - PRIME fixes rockchip: - rk3066 hdmi support sun4i: - DSI burst mode support vc4: - load tracker to detect underflow v3d: - v3d v4.2 support malidp: - initial Mali D71 support in komeda driver tfp410: - omap related improvement omapdrm: - drm bridge/panel support - drop some omap specific panels rcar-du: - Display writeback support" * tag 'drm-next-2019-05-09' of git://anongit.freedesktop.org/drm/drm: (1507 commits) drm/msm/a6xx: No zap shader is not an error drm/cma-helper: Fix drm_gem_cma_free_object() drm: Fix timestamp docs for variable refresh properties. drm/komeda: Mark the local functions as static drm/komeda: Fixed warning: Function parameter or member not described drm/komeda: Expose bus_width to Komeda-CORE drm/komeda: Add sysfs attribute: core_id and config_id drm: add non-desktop quirk for Valve HMDs drm/panfrost: Show stored feature registers drm/panfrost: Don't scream about deferred probe drm/panfrost: Disable PM on probe failure drm/panfrost: Set DMA masks earlier drm/panfrost: Add sanity checks to submit IOCTL drm/etnaviv: initialize idle mask before querying the HW db drm: introduce a capability flag for syncobj timeline support drm: report consistent errors when checking syncobj capibility drm/nouveau/nouveau: forward error generated while resuming objects tree drm/nouveau/fb/ramgk104: fix spelling mistake "sucessfully" -> "successfully" drm/nouveau/i2c: Disable i2c bus access after ->fini() drm/nouveau: Remove duplicate ACPI_VIDEO_NOTIFY_PROBE definition ...
2019-05-08 21:35:19 -07:00 · 2019-05-08 21:35:19 -07:00 · a2d635decb
parent 89c3b37af8 eb85d03e01
commit a2d635decb
1012 changed files with 63845 additions and 24075 deletions
--- a/Documentation/devicetree/bindings/display/amlogic,meson-dw-hdmi.txt
+++ b/Documentation/devicetree/bindings/display/amlogic,meson-dw-hdmi.txt
@ -37,6 +37,7 @@ Required properties:
 	- GXL (S905X, S905D) : "amlogic,meson-gxl-dw-hdmi"
 	- GXM (S912) : "amlogic,meson-gxm-dw-hdmi"
 	followed by the common "amlogic,meson-gx-dw-hdmi"
+	- G12A (S905X2, S905Y2, S905D2) : "amlogic,meson-g12a-dw-hdmi"
 - reg: Physical base address and length of the controller's registers.
 - interrupts: The HDMI interrupt number
 - clocks, clock-names : must have the phandles to the HDMI iahb and isfr clocks,
@ -66,6 +67,9 @@ corresponding to each HDMI output and input.
 S905X (GXL)	VENC Input	TMDS Output
 S905D (GXL)	VENC Input	TMDS Output
 S912 (GXM)	VENC Input	TMDS Output
+ S905X2 (G12A)	VENC Input	TMDS Output
+ S905Y2 (G12A)	VENC Input	TMDS Output
+ S905D2 (G12A)	VENC Input	TMDS Output

 Example:

--- a/Documentation/devicetree/bindings/display/amlogic,meson-vpu.txt
+++ b/Documentation/devicetree/bindings/display/amlogic,meson-vpu.txt
@ -57,18 +57,18 @@ Required properties:
 	- GXL (S905X, S905D) : "amlogic,meson-gxl-vpu"
 	- GXM (S912) : "amlogic,meson-gxm-vpu"
 	followed by the common "amlogic,meson-gx-vpu"
+	- G12A (S905X2, S905Y2, S905D2) : "amlogic,meson-g12a-vpu"
 - reg: base address and size of he following memory-mapped regions :
 	- vpu
 	- hhi
-	- dmc
 - reg-names: should contain the names of the previous memory regions
 - interrupts: should contain the VENC Vsync interrupt number
+- amlogic,canvas: phandle to canvas provider node as described in the file
+	../soc/amlogic/amlogic,canvas.txt

 Optional properties:
 - power-domains: Optional phandle to associated power domain as described in
 	the file ../power/power_domain.txt
- amlogic,canvas: phandle to canvas provider node as described in the file
-	../soc/amlogic/amlogic,canvas.txt

 Required nodes:

@ -84,6 +84,9 @@ corresponding to each VPU output.
 S905X (GXL)	CVBS VDAC	HDMI-TX
 S905D (GXL)	CVBS VDAC	HDMI-TX
 S912 (GXM)	CVBS VDAC	HDMI-TX
+ S905X2 (G12A)	CVBS VDAC	HDMI-TX
+ S905Y2 (G12A)	CVBS VDAC	HDMI-TX
+ S905D2 (G12A)	CVBS VDAC	HDMI-TX

 Example:

--- a/Documentation/devicetree/bindings/display/bridge/ti,tfp410.txt
+++ b/Documentation/devicetree/bindings/display/bridge/ti,tfp410.txt
@ -6,15 +6,32 @@ Required properties:

 Optional properties:
 - powerdown-gpios: power-down gpio
- reg: I2C address. If and only if present the device node
-       should be placed into the i2c controller node where the
-       tfp410 i2c is connected to.
+- reg: I2C address. If and only if present the device node should be placed
+  into the I2C controller node where the TFP410 I2C is connected to.
+- ti,deskew: data de-skew in 350ps increments, from -4 to +3, as configured
+  through th DK[3:1] pins. This property shall be present only if the TFP410
+  is not connected through I2C.

 Required nodes:
- Video port 0 for DPI input [1].
- Video port 1 for DVI output [1].

-[1]: Documentation/devicetree/bindings/media/video-interfaces.txt
+This device has two video ports. Their connections are modeled using the OF
+graph bindings specified in [1]. Each port node shall have a single endpoint.
+
+- Port 0 is the DPI input port. Its endpoint subnode shall contain a
+  pclk-sample and bus-width property and a remote-endpoint property as specified
+  in [1].
+  - If pclk-sample is not defined, pclk-sample = 0 should be assumed for
+    backward compatibility.
+  - If bus-width is not defined then bus-width = 24 should be assumed for
+    backward compatibility.
+    bus-width = 24: 24 data lines are connected and single-edge mode
+    bus-width = 12: 12 data lines are connected and dual-edge mode
+
+- Port 1 is the DVI output port. Its endpoint subnode shall contain a
+  remote-endpoint property is specified in [1].
+
+[1] Documentation/devicetree/bindings/media/video-interfaces.txt
+

 Example
 -------
@ -22,6 +39,7 @@ Example
 tfp410: encoder@0 {
 	compatible = "ti,tfp410";
 	powerdown-gpios = <&twl_gpio 2 GPIO_ACTIVE_LOW>;
+	ti,deskew = <4>;

 	ports {
 		#address-cells = <1>;
@ -31,6 +49,8 @@ tfp410: encoder@0 {
 			reg = <0>;

 			tfp410_in: endpoint@0 {
+				pclk-sample = <1>;
+				bus-width = <24>;
 				remote-endpoint = <&dpi_out>;
 			};
 		};
--- a/Documentation/devicetree/bindings/display/msm/gmu.txt
+++ b/Documentation/devicetree/bindings/display/msm/gmu.txt
@ -24,7 +24,10 @@ Required properties:
   * "cxo"
   * "axi"
   * "mnoc"
- power-domains: should be <&clock_gpucc GPU_CX_GDSC>
+- power-domains: should be:
+	<&clock_gpucc GPU_CX_GDSC>
+	<&clock_gpucc GPU_GX_GDSC>
+- power-domain-names: Matching names for the power domains
 - iommus: phandle to the adreno iommu
 - operating-points-v2: phandle to the OPP operating points

@ -51,7 +54,10 @@ Example:
 			<&gcc GCC_GPU_MEMNOC_GFX_CLK>;
 		clock-names = "gmu", "cxo", "axi", "memnoc";

-		power-domains = <&gpucc GPU_CX_GDSC>;
+		power-domains = <&gpucc GPU_CX_GDSC>,
+				<&gpucc GPU_GX_GDSC>;
+		power-domain-names = "cx", "gx";
+
 		iommus = <&adreno_smmu 5>;

 		operating-points-v2 = <&gmu_opp_table>;
--- a/Documentation/devicetree/bindings/display/msm/gpu.txt
+++ b/Documentation/devicetree/bindings/display/msm/gpu.txt
@ -22,9 +22,14 @@ Required properties:
   - qcom,adreno-630.2
 - iommus: optional phandle to an adreno iommu instance
 - operating-points-v2: optional phandle to the OPP operating points
+- interconnects: optional phandle to an interconnect provider.  See
+  ../interconnect/interconnect.txt for details.
 - qcom,gmu: For GMU attached devices a phandle to the GMU device that will
  control the power for the GPU. Applicable targets:
    - qcom,adreno-630.2
+- zap-shader: For a5xx and a6xx devices this node contains a memory-region that
+  points to reserved memory to store the zap shader that can be used to help
+  bring the GPU out of secure mode.

 Example 3xx/4xx/a5xx:

@ -70,6 +75,12 @@ Example a6xx (with GMU):

 		operating-points-v2 = <&gpu_opp_table>;

+		interconnects = <&rsc_hlos MASTER_GFX3D &rsc_hlos SLAVE_EBI1>;
+
 		qcom,gmu = <&gmu>;
+
+		zap-shader {
+			memory-region = <&zap_shader_region>;
+		};
 	};
 };
--- a/Documentation/devicetree/bindings/display/panel/feiyang,fy07024di26a30d.txt
+++ b/Documentation/devicetree/bindings/display/panel/feiyang,fy07024di26a30d.txt
@ -0,0 +1,20 @@
+Feiyang FY07024DI26A30-D 7" MIPI-DSI LCD Panel
+
+Required properties:
+- compatible: must be "feiyang,fy07024di26a30d"
+- reg: DSI virtual channel used by that screen
+- avdd-supply: analog regulator dc1 switch
+- dvdd-supply: 3v3 digital regulator
+- reset-gpios: a GPIO phandle for the reset pin
+
+Optional properties:
+- backlight: phandle for the backlight control.
+
+panel@0 {
+	compatible = "feiyang,fy07024di26a30d";
+	reg = <0>;
+	avdd-supply = <&reg_dc1sw>;
+	dvdd-supply = <&reg_dldo2>;
+	reset-gpios = <&pio 3 24 GPIO_ACTIVE_HIGH>; /* LCD-RST: PD24 */
+	backlight = <&backlight>;
+};
--- a/Documentation/devicetree/bindings/display/panel/innolux,p079zca.txt
+++ b/Documentation/devicetree/bindings/display/panel/innolux,p079zca.txt
@ -12,7 +12,7 @@ Optional properties:
 Example:

 	&mipi_dsi {
-		panel {
+		panel@0 {
 			compatible = "innolux,p079zca";
 			reg = <0>;
 			power-supply = <...>;
--- a/Documentation/devicetree/bindings/display/panel/innolux,p097pfg.txt
+++ b/Documentation/devicetree/bindings/display/panel/innolux,p097pfg.txt
@ -13,7 +13,7 @@ Optional properties:
 Example:

 	&mipi_dsi {
-		panel {
+		panel@0 {
 			compatible = "innolux,p079zca";
 			reg = <0>;
 			avdd-supply = <...>;
--- a/Documentation/devicetree/bindings/display/panel/kingdisplay,kd097d04.txt
+++ b/Documentation/devicetree/bindings/display/panel/kingdisplay,kd097d04.txt
@ -12,7 +12,7 @@ Optional properties:
 Example:

 	&mipi_dsi {
-		panel {
+		panel@0 {
 			compatible = "kingdisplay,kd097d04";
 			reg = <0>;
 			power-supply = <...>;
--- a/Documentation/devicetree/bindings/display/panel/lg,acx467akm-7.txt
+++ b/Documentation/devicetree/bindings/display/panel/lg,acx467akm-7.txt
@ -0,0 +1,7 @@
+LG ACX467AKM-7 4.95" 1080×1920 LCD Panel
+
+Required properties:
+- compatible: must be "lg,acx467akm-7"
+
+This binding is compatible with the simple-panel binding, which is specified
+in simple-panel.txt in this directory.
--- a/Documentation/devicetree/bindings/display/panel/osddisplays,osd070t1718-19ts.txt
+++ b/Documentation/devicetree/bindings/display/panel/osddisplays,osd070t1718-19ts.txt
@ -0,0 +1,12 @@
+OSD Displays OSD070T1718-19TS 7" WVGA TFT LCD panel
+
+Required properties:
+- compatible: shall be "osddisplays,osd070t1718-19ts"
+- power-supply: see simple-panel.txt
+
+Optional properties:
+- backlight: see simple-panel.txt
+
+This binding is compatible with the simple-panel binding, which is specified
+in simple-panel.txt in this directory. No other simple-panel properties than
+the ones specified herein are valid.
--- a/Documentation/devicetree/bindings/display/panel/rocktech,jh057n00900.txt
+++ b/Documentation/devicetree/bindings/display/panel/rocktech,jh057n00900.txt
@ -0,0 +1,18 @@
+Rocktech jh057n00900 5.5" 720x1440 TFT LCD panel
+
+Required properties:
+- compatible: should be "rocktech,jh057n00900"
+- reg: DSI virtual channel of the peripheral
+- reset-gpios: panel reset gpio
+- backlight: phandle of the backlight device attached to the panel
+
+Example:
+
+	&mipi_dsi {
+		panel@0 {
+			compatible = "rocktech,jh057n00900";
+			reg = <0>;
+			backlight = <&backlight>;
+			reset-gpios = <&gpio3 13 GPIO_ACTIVE_LOW>;
+		};
+	};
--- a/Documentation/devicetree/bindings/display/panel/ronbo,rb070d30.yaml
+++ b/Documentation/devicetree/bindings/display/panel/ronbo,rb070d30.yaml
@ -0,0 +1,51 @@
+# SPDX-License-Identifier: (GPL-2.0+ OR X11)
+%YAML 1.2
+---
+$id: http://devicetree.org/schemas/display/panel/ronbo,rb070d30.yaml#
+$schema: http://devicetree.org/meta-schemas/core.yaml#
+
+title: Ronbo RB070D30 DSI Display Panel
+
+maintainers:
+  - Maxime Ripard <maxime.ripard@bootlin.com>
+
+properties:
+  compatible:
+    const: ronbo,rb070d30
+
+  reg:
+    description: MIPI-DSI virtual channel
+
+  power-gpios:
+    description: GPIO used for the power pin
+    maxItems: 1
+
+  reset-gpios:
+    description: GPIO used for the reset pin
+    maxItems: 1
+
+  shlr-gpios:
+    description: GPIO used for the shlr pin (horizontal flip)
+    maxItems: 1
+
+  updn-gpios:
+    description: GPIO used for the updn pin (vertical flip)
+    maxItems: 1
+
+  vcc-lcd-supply:
+    description: Power regulator
+
+  backlight:
+    description: Backlight used by the panel
+    $ref: "/schemas/types.yaml#/definitions/phandle"
+
+required:
+  - compatible
+  - power-gpios
+  - reg
+  - reset-gpios
+  - shlr-gpios
+  - updn-gpios
+  - vcc-lcd-supply
+
+additionalProperties: false
--- a/Documentation/devicetree/bindings/display/panel/tpo,td028ttec1.txt
+++ b/Documentation/devicetree/bindings/display/panel/tpo,td028ttec1.txt
@ -6,6 +6,7 @@ Required properties:

 Optional properties:
 - label: a symbolic name for the panel
+- backlight: phandle of the backlight device

 Required nodes:
 - Video port for DPI input
@ -21,6 +22,7 @@ lcd-panel: td028ttec1@0 {
 	spi-cpha;

 	label = "lcd";
+	backlight = <&backlight>;
 	port {
 		lcd_in: endpoint {
 			remote-endpoint = <&dpi_out>;
--- a/Documentation/devicetree/bindings/display/rockchip/rockchip,rk3066-hdmi.txt
+++ b/Documentation/devicetree/bindings/display/rockchip/rockchip,rk3066-hdmi.txt
@ -0,0 +1,72 @@
+Rockchip specific extensions for rk3066 HDMI
+============================================
+
+Required properties:
+- compatible:
+	"rockchip,rk3066-hdmi";
+- reg:
+	Physical base address and length of the controller's registers.
+- clocks, clock-names:
+	Phandle to HDMI controller clock, name should be "hclk".
+- interrupts:
+	HDMI interrupt number.
+- power-domains:
+	Phandle to the RK3066_PD_VIO power domain.
+- rockchip,grf:
+	This soc uses GRF regs to switch the HDMI TX input between vop0 and vop1.
+- ports:
+	Contains one port node with two endpoints, numbered 0 and 1,
+	connected respectively to vop0 and vop1.
+	Contains one port node with one endpoint
+	connected to a hdmi-connector node.
+- pinctrl-0, pinctrl-name:
+	Switch the iomux for the HPD/I2C pins to HDMI function.
+
+Example:
+	hdmi: hdmi@10116000 {
+		compatible = "rockchip,rk3066-hdmi";
+		reg = <0x10116000 0x2000>;
+		interrupts = <GIC_SPI 64 IRQ_TYPE_LEVEL_HIGH>;
+		clocks = <&cru HCLK_HDMI>;
+		clock-names = "hclk";
+		power-domains = <&power RK3066_PD_VIO>;
+		rockchip,grf = <&grf>;
+		pinctrl-names = "default";
+		pinctrl-0 = <&hdmii2c_xfer>, <&hdmi_hpd>;
+
+		ports {
+			#address-cells = <1>;
+			#size-cells = <0>;
+			hdmi_in: port@0 {
+				reg = <0>;
+				#address-cells = <1>;
+				#size-cells = <0>;
+				hdmi_in_vop0: endpoint@0 {
+					reg = <0>;
+					remote-endpoint = <&vop0_out_hdmi>;
+				};
+				hdmi_in_vop1: endpoint@1 {
+					reg = <1>;
+					remote-endpoint = <&vop1_out_hdmi>;
+				};
+			};
+			hdmi_out: port@1 {
+				reg = <1>;
+				hdmi_out_con: endpoint {
+					remote-endpoint = <&hdmi_con_in>;
+				};
+			};
+		};
+	};
+
+&pinctrl {
+		hdmi {
+			hdmi_hpd: hdmi-hpd {
+				rockchip,pins = <0 RK_PA0 1 &pcfg_pull_default>;
+			};
+			hdmii2c_xfer: hdmii2c-xfer {
+				rockchip,pins = <0 RK_PA1 1 &pcfg_pull_none>,
+						<0 RK_PA2 1 &pcfg_pull_none>;
+			};
+		};
+};
--- a/Documentation/devicetree/bindings/display/ste,mcde.txt
+++ b/Documentation/devicetree/bindings/display/ste,mcde.txt
@ -0,0 +1,104 @@
+ST-Ericsson Multi Channel Display Engine MCDE
+
+The ST-Ericsson MCDE is a display controller with support for compositing
+and displaying several channels memory resident graphics data on DSI or
+LCD displays or bridges. It is used in the ST-Ericsson U8500 platform.
+
+Required properties:
+
+- compatible: must be:
+  "ste,mcde"
+- reg: register base for the main MCDE control registers, should be
+  0x1000 in size
+- interrupts: the interrupt line for the MCDE
+- epod-supply: a phandle to the EPOD regulator
+- vana-supply: a phandle to the analog voltage regulator
+- clocks: an array of the MCDE clocks in this strict order:
+  MCDECLK (main MCDE clock), LCDCLK (LCD clock), PLLDSI
+  (HDMI clock), DSI0ESCLK (DSI0 energy save clock),
+  DSI1ESCLK (DSI1 energy save clock), DSI2ESCLK (DSI2 energy
+  save clock)
+- clock-names: must be the following array:
+  "mcde", "lcd", "hdmi"
+  to match the required clock inputs above.
+- #address-cells: should be <1> (for the DSI hosts that will be children)
+- #size-cells: should be <1> (for the DSI hosts that will be children)
+- ranges: this should always be stated
+
+Required subnodes:
+
+The devicetree must specify subnodes for the DSI host adapters.
+These must have the following characteristics:
+
+- compatible: must be:
+  "ste,mcde-dsi"
+- reg: must specify the register range for the DSI host
+- vana-supply: phandle to the VANA voltage regulator
+- clocks: phandles to the high speed and low power (energy save) clocks
+  the high speed clock is not present on the third (dsi2) block, so it
+  should only have the "lp" clock
+- clock-names: "hs" for the high speed clock and "lp" for the low power
+  (energy save) clock
+- #address-cells: should be <1>
+- #size-cells: should be <0>
+
+Display panels and bridges will appear as children on the DSI hosts, and
+the displays are connected to the DSI hosts using the common binding
+for video transmitter interfaces; see
+Documentation/devicetree/bindings/media/video-interfaces.txt
+
+If a DSI host is unused (not connected) it will have no children defined.
+
+Example:
+
+mcde@a0350000 {
+	compatible = "ste,mcde";
+	reg = <0xa0350000 0x1000>;
+	interrupts = <GIC_SPI 48 IRQ_TYPE_LEVEL_HIGH>;
+	epod-supply = <&db8500_b2r2_mcde_reg>;
+	vana-supply = <&ab8500_ldo_ana_reg>;
+	clocks = <&prcmu_clk PRCMU_MCDECLK>, /* Main MCDE clock */
+		 <&prcmu_clk PRCMU_LCDCLK>, /* LCD clock */
+		 <&prcmu_clk PRCMU_PLLDSI>; /* HDMI clock */
+	clock-names = "mcde", "lcd", "hdmi";
+	#address-cells = <1>;
+	#size-cells = <1>;
+	ranges;
+
+	dsi0: dsi@a0351000 {
+		compatible = "ste,mcde-dsi";
+		reg = <0xa0351000 0x1000>;
+		vana-supply = <&ab8500_ldo_ana_reg>;
+		clocks = <&prcmu_clk PRCMU_DSI0CLK>, <&prcmu_clk PRCMU_DSI0ESCCLK>;
+		clock-names = "hs", "lp";
+		#address-cells = <1>;
+		#size-cells = <0>;
+
+		panel {
+			compatible = "samsung,s6d16d0";
+			reg = <0>;
+			vdd1-supply = <&ab8500_ldo_aux1_reg>;
+			reset-gpios = <&gpio2 1 GPIO_ACTIVE_LOW>;
+		};
+
+	};
+	dsi1: dsi@a0352000 {
+		compatible = "ste,mcde-dsi";
+		reg = <0xa0352000 0x1000>;
+		vana-supply = <&ab8500_ldo_ana_reg>;
+		clocks = <&prcmu_clk PRCMU_DSI1CLK>, <&prcmu_clk PRCMU_DSI1ESCCLK>;
+		clock-names = "hs", "lp";
+		#address-cells = <1>;
+		#size-cells = <0>;
+	};
+	dsi2: dsi@a0353000 {
+		compatible = "ste,mcde-dsi";
+		reg = <0xa0353000 0x1000>;
+		vana-supply = <&ab8500_ldo_ana_reg>;
+		/* This DSI port only has the Low Power / Energy Save clock */
+		clocks = <&prcmu_clk PRCMU_DSI2ESCCLK>;
+		clock-names = "lp";
+		#address-cells = <1>;
+		#size-cells = <0>;
+	};
+};
--- a/Documentation/devicetree/bindings/gpu/arm,mali-bifrost.txt
+++ b/Documentation/devicetree/bindings/gpu/arm,mali-bifrost.txt
@ -0,0 +1,92 @@
+ARM Mali Bifrost GPU
+====================
+
+Required properties:
+
+- compatible :
+  * Since Mali Bifrost GPU model/revision is fully discoverable by reading
+    some determined registers, must contain the following:
+    + "arm,mali-bifrost"
+  * which must be preceded by one of the following vendor specifics:
+    + "amlogic,meson-g12a-mali"
+
+- reg : Physical base address of the device and length of the register area.
+
+- interrupts : Contains the three IRQ lines required by Mali Bifrost devices,
+  in the following defined order.
+
+- interrupt-names : Contains the names of IRQ resources in this exact defined
+  order: "job", "mmu", "gpu".
+
+Optional properties:
+
+- clocks : Phandle to clock for the Mali Bifrost device.
+
+- mali-supply : Phandle to regulator for the Mali device. Refer to
+  Documentation/devicetree/bindings/regulator/regulator.txt for details.
+
+- operating-points-v2 : Refer to Documentation/devicetree/bindings/opp/opp.txt
+  for details.
+
+- resets : Phandle of the GPU reset line.
+
+Vendor-specific bindings
+------------------------
+
+The Mali GPU is integrated very differently from one SoC to
+another. In order to accommodate those differences, you have the option
+to specify one more vendor-specific compatible, among:
+
+- "amlogic,meson-g12a-mali"
+  Required properties:
+  - resets : Should contain phandles of :
+    + GPU reset line
+    + GPU APB glue reset line
+
+Example for a Mali-G31:
+
+gpu@ffa30000 {
+	compatible = "amlogic,meson-g12a-mali", "arm,mali-bifrost";
+	reg = <0xffe40000 0x10000>;
+	interrupts = <GIC_SPI 160 IRQ_TYPE_LEVEL_HIGH>,
+		     <GIC_SPI 161 IRQ_TYPE_LEVEL_HIGH>,
+		     <GIC_SPI 162 IRQ_TYPE_LEVEL_HIGH>;
+	interrupt-names = "job", "mmu", "gpu";
+	clocks = <&clk CLKID_MALI>;
+	mali-supply = <&vdd_gpu>;
+	operating-points-v2 = <&gpu_opp_table>;
+	resets = <&reset RESET_DVALIN_CAPB3>, <&reset RESET_DVALIN>;
+};
+
+gpu_opp_table: opp_table0 {
+	compatible = "operating-points-v2";
+
+	opp@533000000 {
+		opp-hz = /bits/ 64 <533000000>;
+		opp-microvolt = <1250000>;
+	};
+	opp@450000000 {
+		opp-hz = /bits/ 64 <450000000>;
+		opp-microvolt = <1150000>;
+	};
+	opp@400000000 {
+		opp-hz = /bits/ 64 <400000000>;
+		opp-microvolt = <1125000>;
+	};
+	opp@350000000 {
+		opp-hz = /bits/ 64 <350000000>;
+		opp-microvolt = <1075000>;
+	};
+	opp@266000000 {
+		opp-hz = /bits/ 64 <266000000>;
+		opp-microvolt = <1025000>;
+	};
+	opp@160000000 {
+		opp-hz = /bits/ 64 <160000000>;
+		opp-microvolt = <925000>;
+	};
+	opp@100000000 {
+		opp-hz = /bits/ 64 <100000000>;
+		opp-microvolt = <912500>;
+	};
+};
--- a/Documentation/devicetree/bindings/gpu/aspeed-gfx.txt
+++ b/Documentation/devicetree/bindings/gpu/aspeed-gfx.txt
@ -0,0 +1,41 @@
+Device tree configuration for the GFX display device on the ASPEED SoCs
+
+Required properties:
+  - compatible
+    * Must be one of the following:
+      + aspeed,ast2500-gfx
+      + aspeed,ast2400-gfx
+    * In addition, the ASPEED pinctrl bindings require the 'syscon' property to
+      be present
+
+  - reg: Physical base address and length of the GFX registers
+
+  - interrupts: interrupt number for the GFX device
+
+  - clocks: clock number used to generate the pixel clock
+
+  - resets: reset line that must be released to use the GFX device
+
+  - memory-region:
+    Phandle to a memory region to allocate from, as defined in
+    Documentation/devicetree/bindings/reserved-memory/reserved-memory.txt
+
+
+Example:
+
+gfx: display@1e6e6000 {
+	compatible = "aspeed,ast2500-gfx", "syscon";
+	reg = <0x1e6e6000 0x1000>;
+	reg-io-width = <4>;
+	clocks = <&syscon ASPEED_CLK_GATE_D1CLK>;
+	resets = <&syscon ASPEED_RESET_CRT1>;
+	interrupts = <0x19>;
+	memory-region = <&gfx_memory>;
+};
+
+gfx_memory: framebuffer {
+	size = <0x01000000>;
+	alignment = <0x01000000>;
+	compatible = "shared-dma-pool";
+	reusable;
+};
--- a/Documentation/devicetree/bindings/gpu/brcm,bcm-v3d.txt
+++ b/Documentation/devicetree/bindings/gpu/brcm,bcm-v3d.txt
@ -6,15 +6,20 @@ For V3D 2.x, see brcm,bcm-vc4.txt.
 Required properties:
 - compatible:	Should be "brcm,7268-v3d" or "brcm,7278-v3d"
 - reg:		Physical base addresses and lengths of the register areas
- reg-names:	Names for the register areas.  The "hub", "bridge", and "core0"
+- reg-names:	Names for the register areas.  The "hub" and "core0"
 		  register areas are always required.  The "gca" register area
-		  is required if the GCA cache controller is present.
+		  is required if the GCA cache controller is present.  The
+		  "bridge" register area is required if an external reset
+		  controller is not present.
 - interrupts:	The interrupt numbers.  The first interrupt is for the hub,
-		  while the following interrupts are for the cores.
+		  while the following interrupts are separate interrupt lines
+		  for the cores (if they don't share the hub's interrupt).
 		  See bindings/interrupt-controller/interrupts.txt

 Optional properties:
 - clocks:	The core clock the unit runs on
+- resets:	The reset line for v3d, if not using a mapping of the bridge
+		  See bindings/reset/reset.txt

 v3d {
 	compatible = "brcm,7268-v3d";
--- a/Documentation/devicetree/bindings/vendor-prefixes.txt
+++ b/Documentation/devicetree/bindings/vendor-prefixes.txt
@ -305,6 +305,7 @@ oranth	Shenzhen Oranth Technology Co., Ltd.
 ORCL	Oracle Corporation
 orisetech	Orise Technology
 ortustech	Ortus Technology Co., Ltd.
+osddisplays	OSD Displays
 ovti	OmniVision Technologies
 oxsemi	Oxford Semiconductor, Ltd.
 panasonic	Panasonic Corporation
@ -347,7 +348,9 @@ ricoh	Ricoh Co. Ltd.
 rikomagic	Rikomagic Tech Corp. Ltd
 riscv	RISC-V Foundation
 rockchip	Fuzhou Rockchip Electronics Co., Ltd
+rocktech	ROCKTECH DISPLAYS LIMITED
 rohm	ROHM Semiconductor Co., Ltd
+ronbo   Ronbo Electronics
 roofull	Shenzhen Roofull Technology Co, Ltd
 samsung	Samsung Semiconductor
 samtec	Samtec/Softing company
--- a/Documentation/driver-api/component.rst
+++ b/Documentation/driver-api/component.rst
@ -1,3 +1,5 @@
+.. _component:
+
 ======================================
 Component Helper for Aggregate Drivers
 ======================================
--- a/Documentation/driver-model/devres.txt
+++ b/Documentation/driver-model/devres.txt
@ -256,6 +256,9 @@ DMA
  dmam_pool_create()
  dmam_pool_destroy()

+DRM
+  devm_drm_dev_init()
+
 GPIO
  devm_gpiod_get()
  devm_gpiod_get_index()
--- a/Documentation/gpu/drm-internals.rst
+++ b/Documentation/gpu/drm-internals.rst
@ -93,6 +93,11 @@ Device Instance and Driver Handling
 Driver Load
 -----------

+Component Helper Usage
+~~~~~~~~~~~~~~~~~~~~~~
+
+.. kernel-doc:: drivers/gpu/drm/drm_drv.c
+   :doc: component helper usage recommendations

 IRQ Helper Library
 ~~~~~~~~~~~~~~~~~~
--- a/Documentation/gpu/drm-kms-helpers.rst
+++ b/Documentation/gpu/drm-kms-helpers.rst
@ -107,6 +107,12 @@ fbdev Helper Functions Reference
 .. kernel-doc:: drivers/gpu/drm/drm_fb_helper.c
   :export:

+format Helper Functions Reference
+=================================
+
+.. kernel-doc:: drivers/gpu/drm/drm_format_helper.c
+   :export:
+
 Framebuffer CMA Helper Functions Reference
 ==========================================

@ -369,3 +375,15 @@ Legacy CRTC/Modeset Helper Functions Reference

 .. kernel-doc:: drivers/gpu/drm/drm_crtc_helper.c
   :export:
+
+SHMEM GEM Helper Reference
+==========================
+
+.. kernel-doc:: drivers/gpu/drm/drm_gem_shmem_helper.c
+   :doc: overview
+
+.. kernel-doc:: include/drm/drm_gem_shmem_helper.h
+   :internal:
+
+.. kernel-doc:: drivers/gpu/drm/drm_gem_shmem_helper.c
+   :export:
--- a/Documentation/gpu/kms-properties.csv
+++ b/Documentation/gpu/kms-properties.csv
@ -17,7 +17,6 @@ Owner Module/Drivers,Group,Property Name,Type,Property Values,Object attached,De
 ,Virtual GPU,“suggested X”,RANGE,"Min=0, Max=0xffffffff",Connector,property to suggest an X offset for a connector
 ,,“suggested Y”,RANGE,"Min=0, Max=0xffffffff",Connector,property to suggest an Y offset for a connector
 ,Optional,"""aspect ratio""",ENUM,"{ ""None"", ""4:3"", ""16:9"" }",Connector,TDB
-,Optional,"""content type""",ENUM,"{ ""No Data"", ""Graphics"", ""Photo"", ""Cinema"", ""Game"" }",Connector,TBD
 i915,Generic,"""Broadcast RGB""",ENUM,"{ ""Automatic"", ""Full"", ""Limited 16:235"" }",Connector,"When this property is set to Limited 16:235 and CTM is set, the hardware will be programmed with the result of the multiplication of CTM by the limited range matrix to ensure the pixels normaly in the range 0..1.0 are remapped to the range 16/255..235/255."
 ,,“audio”,ENUM,"{ ""force-dvi"", ""off"", ""auto"", ""on"" }",Connector,TBD
 ,SDVO-TV,“mode”,ENUM,"{ ""NTSC_M"", ""NTSC_J"", ""NTSC_443"", ""PAL_B"" } etc.",Connector,TBD
--- a/Documentation/gpu/meson.rst
+++ b/Documentation/gpu/meson.rst
@ -42,12 +42,6 @@ Video Encoder
 .. kernel-doc:: drivers/gpu/drm/meson/meson_venc.c
   :doc: Video Encoder

-Video Canvas Management
-=======================
-
-.. kernel-doc:: drivers/gpu/drm/meson/meson_canvas.c
-   :doc: Canvas
-
 Video Clocks
 ============

--- a/Documentation/gpu/tinydrm.rst
+++ b/Documentation/gpu/tinydrm.rst
@ -1,27 +1,12 @@
-==========================
-drm/tinydrm Driver library
-==========================
+============================
+drm/tinydrm Tiny DRM drivers
+============================

-.. kernel-doc:: drivers/gpu/drm/tinydrm/core/tinydrm-core.c
-   :doc: overview
+tinydrm is a collection of DRM drivers that are so small they can fit in a
+single source file.

-Core functionality
-==================
-
-.. kernel-doc:: drivers/gpu/drm/tinydrm/core/tinydrm-core.c
-   :doc: core
-
-.. kernel-doc:: include/drm/tinydrm/tinydrm.h
-   :internal:
-
-.. kernel-doc:: drivers/gpu/drm/tinydrm/core/tinydrm-core.c
-   :export:
-
-.. kernel-doc:: drivers/gpu/drm/tinydrm/core/tinydrm-pipe.c
-   :export:
-
-Additional helpers
-==================
+Helpers
+=======

 .. kernel-doc:: include/drm/tinydrm/tinydrm-helpers.h
   :internal:
@ -29,6 +14,9 @@ Additional helpers
 .. kernel-doc:: drivers/gpu/drm/tinydrm/core/tinydrm-helpers.c
   :export:

+.. kernel-doc:: drivers/gpu/drm/tinydrm/core/tinydrm-pipe.c
+   :export:
+
 MIPI DBI Compatible Controllers
 ===============================

--- a/Documentation/gpu/todo.rst
+++ b/Documentation/gpu/todo.rst
@ -215,12 +215,12 @@ Might be good to also have some igt testcases for this.

 Contact: Daniel Vetter, Noralf Tronnes

-Put a reservation_object into drm_gem_object
+Remove the ->gem_prime_res_obj callback
 --------------------------------------------

-This would remove the need for the ->gem_prime_res_obj callback. It would also
-allow us to implement generic helpers for waiting for a bo, allowing for quite a
-bit of refactoring in the various wait ioctl implementations.
+The ->gem_prime_res_obj callback can be removed from drivers by using the
+reservation_object in the drm_gem_object. It may also be possible to use the
+generic drm_gem_reservation_object_wait helper for waiting for a bo.

 Contact: Daniel Vetter

@ -469,10 +469,6 @@ those drivers as simple as possible, so lots of room for refactoring:
  one of the ideas for having a shared dsi/dbi helper, abstracting away the
  transport details more.

- Quick aside: The unregister devm stuff is kinda getting the lifetimes of
-  a drm_device wrong. Doesn't matter, since everyone else gets it wrong
-  too :-)
-
 Contact: Noralf Trønnes, Daniel Vetter

 AMD DC Display Driver
--- a/47
+++ b/47
@ -1169,7 +1169,7 @@ S:	Supported
 T:	git git://linux-arm.org/linux-ld.git for-upstream/mali-dp
 F:	drivers/gpu/drm/arm/display/include/
 F:	drivers/gpu/drm/arm/display/komeda/
-F:	Documentation/devicetree/bindings/display/arm/arm,komeda.txt
+F:	Documentation/devicetree/bindings/display/arm,komeda.txt
 F:	Documentation/gpu/komeda-kms.rst

 ARM MALI-DP DRM DRIVER
@ -1182,6 +1182,15 @@ F:	drivers/gpu/drm/arm/
 F:	Documentation/devicetree/bindings/display/arm,malidp.txt
 F:	Documentation/gpu/afbc.rst

+ARM MALI PANFROST DRM DRIVER
+M:	Rob Herring <robh@kernel.org>
+M:	Tomeu Vizoso <tomeu.vizoso@collabora.com>
+L:	dri-devel@lists.freedesktop.org
+S:	Supported
+T:	git git://anongit.freedesktop.org/drm/drm-misc
+F:	drivers/gpu/drm/panfrost/
+F:	include/uapi/drm/panfrost_drm.h
+
 ARM MFM AND FLOPPY DRIVERS
 M:	Ian Molton <spyro@f2s.com>
 S:	Maintained
@ -4917,6 +4926,14 @@ M:	Dave Airlie <airlied@redhat.com>
 S:	Odd Fixes
 F:	drivers/gpu/drm/ast/

+DRM DRIVER FOR ASPEED BMC GFX
+M:	Joel Stanley <joel@jms.id.au>
+L:	linux-aspeed@lists.ozlabs.org
+T:	git git://anongit.freedesktop.org/drm/drm-misc
+S:	Supported
+F:	drivers/gpu/drm/aspeed/
+F:	Documentation/devicetree/bindings/gpu/aspeed-gfx.txt
+
 DRM DRIVER FOR BOCHS VIRTUAL GPU
 M:	Gerd Hoffmann <kraxel@redhat.com>
 L:	virtualization@lists.linux-foundation.org
@ -4930,6 +4947,12 @@ T:	git git://anongit.freedesktop.org/drm/drm-misc
 S:	Maintained
 F:	drivers/gpu/drm/tve200/

+DRM DRIVER FOR FEIYANG FY07024DI26A30-D MIPI-DSI LCD PANELS
+M:	Jagan Teki <jagan@amarulasolutions.com>
+S:	Maintained
+F:	drivers/gpu/drm/panel/panel-feiyang-fy07024di26a30d.c
+F:	Documentation/devicetree/bindings/display/panel/feiyang,fy07024di26a30d.txt
+
 DRM DRIVER FOR ILITEK ILI9225 PANELS
 M:	David Lechner <david@lechnology.com>
 S:	Maintained
@ -5021,6 +5044,12 @@ S:	Orphan / Obsolete
 F:	drivers/gpu/drm/r128/
 F:	include/uapi/drm/r128_drm.h

+DRM DRIVER FOR ROCKTECH JH057N00900 PANELS
+M:	Guido Günther <agx@sigxcpu.org>
+S:	Maintained
+F:	drivers/gpu/drm/panel/panel-rocktech-jh057n00900.c
+F:	Documentation/devicetree/bindings/display/panel/rocktech,jh057n00900.txt
+
 DRM DRIVER FOR SAVAGE VIDEO CARDS
 S:	Orphan / Obsolete
 F:	drivers/gpu/drm/savage/
@ -5068,6 +5097,13 @@ S:	Odd Fixes
 F:	drivers/gpu/drm/udl/
 T:	git git://anongit.freedesktop.org/drm/drm-misc

+DRM DRIVER FOR VIRTUALBOX VIRTUAL GPU
+M:	Hans de Goede <hdegoede@redhat.com>
+L:	dri-devel@lists.freedesktop.org
+S:	Maintained
+F:	drivers/gpu/drm/vboxvideo/
+T:	git git://anongit.freedesktop.org/drm/drm-misc
+
 DRM DRIVER FOR VIRTUAL KERNEL MODESETTING (VKMS)
 M:	Rodrigo Siqueira <rodrigosiqueiramelo@gmail.com>
 R:	Haneen Mohammed <hamohammed.sa@gmail.com>
@ -5202,6 +5238,15 @@ S:	Maintained
 F:	drivers/gpu/drm/hisilicon/
 F:	Documentation/devicetree/bindings/display/hisilicon/

+DRM DRIVERS FOR LIMA
+M:	Qiang Yu <yuq825@gmail.com>
+L:	dri-devel@lists.freedesktop.org
+L:	lima@lists.freedesktop.org (moderated for non-subscribers)
+S:	Maintained
+F:	drivers/gpu/drm/lima/
+F:	include/uapi/drm/lima_drm.h
+T:	git git://anongit.freedesktop.org/drm/drm-misc
+
 DRM DRIVERS FOR MEDIATEK
 M:	CK Hu <ck.hu@mediatek.com>
 M:	Philipp Zabel <p.zabel@pengutronix.de>
--- a/arch/x86/kernel/early-quirks.c
+++ b/arch/x86/kernel/early-quirks.c
@ -525,7 +525,8 @@ static const struct pci_device_id intel_early_ids[] __initconst = {
 	INTEL_I945G_IDS(&gen3_early_ops),
 	INTEL_I945GM_IDS(&gen3_early_ops),
 	INTEL_VLV_IDS(&gen6_early_ops),
-	INTEL_PINEVIEW_IDS(&gen3_early_ops),
+	INTEL_PINEVIEW_G_IDS(&gen3_early_ops),
+	INTEL_PINEVIEW_M_IDS(&gen3_early_ops),
 	INTEL_I965G_IDS(&gen3_early_ops),
 	INTEL_G33_IDS(&gen3_early_ops),
 	INTEL_I965GM_IDS(&gen3_early_ops),
@ -547,6 +548,7 @@ static const struct pci_device_id intel_early_ids[] __initconst = {
 	INTEL_GLK_IDS(&gen9_early_ops),
 	INTEL_CNL_IDS(&gen9_early_ops),
 	INTEL_ICL_11_IDS(&gen11_early_ops),
+	INTEL_EHL_IDS(&gen11_early_ops),
 };

 struct resource intel_graphics_stolen_res __ro_after_init = DEFINE_RES_MEM(0, 0);
--- a/drivers/dma-buf/Makefile
+++ b/drivers/dma-buf/Makefile
@ -1,4 +1,5 @@
-obj-y := dma-buf.o dma-fence.o dma-fence-array.o reservation.o seqno-fence.o
+obj-y := dma-buf.o dma-fence.o dma-fence-array.o dma-fence-chain.o \
+	 reservation.o seqno-fence.o
 obj-$(CONFIG_SYNC_FILE)		+= sync_file.o
 obj-$(CONFIG_SW_SYNC)		+= sw_sync.o sync_debug.o
 obj-$(CONFIG_UDMABUF)		+= udmabuf.o
--- a/drivers/dma-buf/dma-fence-chain.c
+++ b/drivers/dma-buf/dma-fence-chain.c
@ -0,0 +1,242 @@
+/*
+ * fence-chain: chain fences together in a timeline
+ *
+ * Copyright (C) 2018 Advanced Micro Devices, Inc.
+ * Authors:
+ *	Christian König <christian.koenig@amd.com>
+ *
+ * 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.
+ */
+
+#include <linux/dma-fence-chain.h>
+
+static bool dma_fence_chain_enable_signaling(struct dma_fence *fence);
+
+/**
+ * dma_fence_chain_get_prev - use RCU to get a reference to the previous fence
+ * @chain: chain node to get the previous node from
+ *
+ * Use dma_fence_get_rcu_safe to get a reference to the previous fence of the
+ * chain node.
+ */
+static struct dma_fence *dma_fence_chain_get_prev(struct dma_fence_chain *chain)
+{
+	struct dma_fence *prev;
+
+	rcu_read_lock();
+	prev = dma_fence_get_rcu_safe(&chain->prev);
+	rcu_read_unlock();
+	return prev;
+}
+
+/**
+ * dma_fence_chain_walk - chain walking function
+ * @fence: current chain node
+ *
+ * Walk the chain to the next node. Returns the next fence or NULL if we are at
+ * the end of the chain. Garbage collects chain nodes which are already
+ * signaled.
+ */
+struct dma_fence *dma_fence_chain_walk(struct dma_fence *fence)
+{
+	struct dma_fence_chain *chain, *prev_chain;
+	struct dma_fence *prev, *replacement, *tmp;
+
+	chain = to_dma_fence_chain(fence);
+	if (!chain) {
+		dma_fence_put(fence);
+		return NULL;
+	}
+
+	while ((prev = dma_fence_chain_get_prev(chain))) {
+
+		prev_chain = to_dma_fence_chain(prev);
+		if (prev_chain) {
+			if (!dma_fence_is_signaled(prev_chain->fence))
+				break;
+
+			replacement = dma_fence_chain_get_prev(prev_chain);
+		} else {
+			if (!dma_fence_is_signaled(prev))
+				break;
+
+			replacement = NULL;
+		}
+
+		tmp = cmpxchg((void **)&chain->prev, (void *)prev, (void *)replacement);
+		if (tmp == prev)
+			dma_fence_put(tmp);
+		else
+			dma_fence_put(replacement);
+		dma_fence_put(prev);
+	}
+
+	dma_fence_put(fence);
+	return prev;
+}
+EXPORT_SYMBOL(dma_fence_chain_walk);
+
+/**
+ * dma_fence_chain_find_seqno - find fence chain node by seqno
+ * @pfence: pointer to the chain node where to start
+ * @seqno: the sequence number to search for
+ *
+ * Advance the fence pointer to the chain node which will signal this sequence
+ * number. If no sequence number is provided then this is a no-op.
+ *
+ * Returns EINVAL if the fence is not a chain node or the sequence number has
+ * not yet advanced far enough.
+ */
+int dma_fence_chain_find_seqno(struct dma_fence **pfence, uint64_t seqno)
+{
+	struct dma_fence_chain *chain;
+
+	if (!seqno)
+		return 0;
+
+	chain = to_dma_fence_chain(*pfence);
+	if (!chain || chain->base.seqno < seqno)
+		return -EINVAL;
+
+	dma_fence_chain_for_each(*pfence, &chain->base) {
+		if ((*pfence)->context != chain->base.context ||
+		    to_dma_fence_chain(*pfence)->prev_seqno < seqno)
+			break;
+	}
+	dma_fence_put(&chain->base);
+
+	return 0;
+}
+EXPORT_SYMBOL(dma_fence_chain_find_seqno);
+
+static const char *dma_fence_chain_get_driver_name(struct dma_fence *fence)
+{
+        return "dma_fence_chain";
+}
+
+static const char *dma_fence_chain_get_timeline_name(struct dma_fence *fence)
+{
+        return "unbound";
+}
+
+static void dma_fence_chain_irq_work(struct irq_work *work)
+{
+	struct dma_fence_chain *chain;
+
+	chain = container_of(work, typeof(*chain), work);
+
+	/* Try to rearm the callback */
+	if (!dma_fence_chain_enable_signaling(&chain->base))
+		/* Ok, we are done. No more unsignaled fences left */
+		dma_fence_signal(&chain->base);
+	dma_fence_put(&chain->base);
+}
+
+static void dma_fence_chain_cb(struct dma_fence *f, struct dma_fence_cb *cb)
+{
+	struct dma_fence_chain *chain;
+
+	chain = container_of(cb, typeof(*chain), cb);
+	irq_work_queue(&chain->work);
+	dma_fence_put(f);
+}
+
+static bool dma_fence_chain_enable_signaling(struct dma_fence *fence)
+{
+	struct dma_fence_chain *head = to_dma_fence_chain(fence);
+
+	dma_fence_get(&head->base);
+	dma_fence_chain_for_each(fence, &head->base) {
+		struct dma_fence_chain *chain = to_dma_fence_chain(fence);
+		struct dma_fence *f = chain ? chain->fence : fence;
+
+		dma_fence_get(f);
+		if (!dma_fence_add_callback(f, &head->cb, dma_fence_chain_cb)) {
+			dma_fence_put(fence);
+			return true;
+		}
+		dma_fence_put(f);
+	}
+	dma_fence_put(&head->base);
+	return false;
+}
+
+static bool dma_fence_chain_signaled(struct dma_fence *fence)
+{
+	dma_fence_chain_for_each(fence, fence) {
+		struct dma_fence_chain *chain = to_dma_fence_chain(fence);
+		struct dma_fence *f = chain ? chain->fence : fence;
+
+		if (!dma_fence_is_signaled(f)) {
+			dma_fence_put(fence);
+			return false;
+		}
+	}
+
+	return true;
+}
+
+static void dma_fence_chain_release(struct dma_fence *fence)
+{
+	struct dma_fence_chain *chain = to_dma_fence_chain(fence);
+
+	dma_fence_put(rcu_dereference_protected(chain->prev, true));
+	dma_fence_put(chain->fence);
+	dma_fence_free(fence);
+}
+
+const struct dma_fence_ops dma_fence_chain_ops = {
+	.use_64bit_seqno = true,
+	.get_driver_name = dma_fence_chain_get_driver_name,
+	.get_timeline_name = dma_fence_chain_get_timeline_name,
+	.enable_signaling = dma_fence_chain_enable_signaling,
+	.signaled = dma_fence_chain_signaled,
+	.release = dma_fence_chain_release,
+};
+EXPORT_SYMBOL(dma_fence_chain_ops);
+
+/**
+ * dma_fence_chain_init - initialize a fence chain
+ * @chain: the chain node to initialize
+ * @prev: the previous fence
+ * @fence: the current fence
+ *
+ * Initialize a new chain node and either start a new chain or add the node to
+ * the existing chain of the previous fence.
+ */
+void dma_fence_chain_init(struct dma_fence_chain *chain,
+			  struct dma_fence *prev,
+			  struct dma_fence *fence,
+			  uint64_t seqno)
+{
+	struct dma_fence_chain *prev_chain = to_dma_fence_chain(prev);
+	uint64_t context;
+
+	spin_lock_init(&chain->lock);
+	rcu_assign_pointer(chain->prev, prev);
+	chain->fence = fence;
+	chain->prev_seqno = 0;
+	init_irq_work(&chain->work, dma_fence_chain_irq_work);
+
+	/* Try to reuse the context of the previous chain node. */
+	if (prev_chain && __dma_fence_is_later(seqno, prev->seqno, prev->ops)) {
+		context = prev->context;
+		chain->prev_seqno = prev->seqno;
+	} else {
+		context = dma_fence_context_alloc(1);
+		/* Make sure that we always have a valid sequence number. */
+		if (prev_chain)
+			seqno = max(prev->seqno, seqno);
+	}
+
+	dma_fence_init(&chain->base, &dma_fence_chain_ops,
+		       &chain->lock, context, seqno);
+}
+EXPORT_SYMBOL(dma_fence_chain_init);
--- a/drivers/dma-buf/reservation.c
+++ b/drivers/dma-buf/reservation.c
@ -73,6 +73,8 @@ int reservation_object_reserve_shared(struct reservation_object *obj,
 	struct reservation_object_list *old, *new;
 	unsigned int i, j, k, max;

+	reservation_object_assert_held(obj);
+
 	old = reservation_object_get_list(obj);

 	if (old && old->shared_max) {
@ -151,6 +153,8 @@ void reservation_object_add_shared_fence(struct reservation_object *obj,

 	dma_fence_get(fence);

+	reservation_object_assert_held(obj);
+
 	fobj = reservation_object_get_list(obj);
 	count = fobj->shared_count;

@ -196,6 +200,8 @@ void reservation_object_add_excl_fence(struct reservation_object *obj,
 	struct reservation_object_list *old;
 	u32 i = 0;

+	reservation_object_assert_held(obj);
+
 	old = reservation_object_get_list(obj);
 	if (old)
 		i = old->shared_count;
@ -236,6 +242,8 @@ int reservation_object_copy_fences(struct reservation_object *dst,
 	size_t size;
 	unsigned i;

+	reservation_object_assert_held(dst);
+
 	rcu_read_lock();
 	src_list = rcu_dereference(src->fence);

--- a/drivers/dma-buf/sw_sync.c
+++ b/drivers/dma-buf/sw_sync.c
@ -161,7 +161,7 @@ static bool timeline_fence_signaled(struct dma_fence *fence)
 {
 	struct sync_timeline *parent = dma_fence_parent(fence);

-	return !__dma_fence_is_later(fence->seqno, parent->value);
+	return !__dma_fence_is_later(fence->seqno, parent->value, fence->ops);
 }

 static bool timeline_fence_enable_signaling(struct dma_fence *fence)
--- a/drivers/dma-buf/sync_file.c
+++ b/drivers/dma-buf/sync_file.c
@ -258,7 +258,8 @@ static struct sync_file *sync_file_merge(const char *name, struct sync_file *a,

 			i_b++;
 		} else {
-			if (__dma_fence_is_later(pt_a->seqno, pt_b->seqno))
+			if (__dma_fence_is_later(pt_a->seqno, pt_b->seqno,
+						 pt_a->ops))
 				add_fence(fences, &i, pt_a);
 			else
 				add_fence(fences, &i, pt_b);
--- a/drivers/gpu/drm/Kconfig
+++ b/drivers/gpu/drm/Kconfig
@ -173,6 +173,12 @@ config DRM_KMS_CMA_HELPER
 	help
 	  Choose this if you need the KMS CMA helper functions

+config DRM_GEM_SHMEM_HELPER
+	bool
+	depends on DRM
+	help
+	  Choose this if you need the GEM shmem helper functions
+
 config DRM_VM
 	bool
 	depends on DRM && MMU
@ -225,8 +231,6 @@ config DRM_AMDGPU

 source "drivers/gpu/drm/amd/amdgpu/Kconfig"

-source "drivers/gpu/drm/amd/lib/Kconfig"
-
 source "drivers/gpu/drm/nouveau/Kconfig"

 source "drivers/gpu/drm/i915/Kconfig"
@ -251,6 +255,9 @@ config DRM_VKMS

 	  If M is selected the module will be called vkms.

+config DRM_ATI_PCIGART
+        bool
+
 source "drivers/gpu/drm/exynos/Kconfig"

 source "drivers/gpu/drm/rockchip/Kconfig"
@ -329,12 +336,21 @@ source "drivers/gpu/drm/tve200/Kconfig"

 source "drivers/gpu/drm/xen/Kconfig"

+source "drivers/gpu/drm/vboxvideo/Kconfig"
+
+source "drivers/gpu/drm/lima/Kconfig"
+
+source "drivers/gpu/drm/panfrost/Kconfig"
+
+source "drivers/gpu/drm/aspeed/Kconfig"
+
 # Keep legacy drivers last

 menuconfig DRM_LEGACY
 	bool "Enable legacy drivers (DANGEROUS)"
 	depends on DRM && MMU
 	select DRM_VM
+	select DRM_ATI_PCIGART if PCI
 	help
 	  Enable legacy DRI1 drivers. Those drivers expose unsafe and dangerous
 	  APIs to user-space, which can be used to circumvent access
--- a/drivers/gpu/drm/Makefile
+++ b/drivers/gpu/drm/Makefile
@ -3,11 +3,9 @@
 # Makefile for the drm device driver.  This driver provides support for the
 # Direct Rendering Infrastructure (DRI) in XFree86 4.1.0 and higher.

-drm-y       :=	drm_auth.o drm_bufs.o drm_cache.o \
-		drm_context.o drm_dma.o \
+drm-y       :=	drm_auth.o drm_cache.o \
 		drm_file.o drm_gem.o drm_ioctl.o drm_irq.o \
-		drm_lock.o drm_memory.o drm_drv.o \
-		drm_scatter.o drm_pci.o \
+		drm_memory.o drm_drv.o drm_pci.o \
 		drm_sysfs.o drm_hashtab.o drm_mm.o \
 		drm_crtc.o drm_fourcc.o drm_modes.o drm_edid.o \
 		drm_encoder_slave.o \
@ -21,11 +19,13 @@ drm-y       :=	drm_auth.o drm_bufs.o drm_cache.o \
 		drm_syncobj.o drm_lease.o drm_writeback.o drm_client.o \
 		drm_atomic_uapi.o

+drm-$(CONFIG_DRM_LEGACY) += drm_legacy_misc.o drm_bufs.o drm_context.o drm_dma.o drm_scatter.o drm_lock.o
 drm-$(CONFIG_DRM_LIB_RANDOM) += lib/drm_random.o
 drm-$(CONFIG_DRM_VM) += drm_vm.o
 drm-$(CONFIG_COMPAT) += drm_ioc32.o
 drm-$(CONFIG_DRM_GEM_CMA_HELPER) += drm_gem_cma_helper.o
-drm-$(CONFIG_PCI) += ati_pcigart.o
+drm-$(CONFIG_DRM_GEM_SHMEM_HELPER) += drm_gem_shmem_helper.o
+drm-$(CONFIG_DRM_ATI_PCIGART) += ati_pcigart.o
 drm-$(CONFIG_DRM_PANEL) += drm_panel.o
 drm-$(CONFIG_OF) += drm_of.o
 drm-$(CONFIG_AGP) += drm_agpsupport.o
@ -37,7 +37,8 @@ drm_kms_helper-y := drm_crtc_helper.o drm_dp_helper.o drm_dsc.o drm_probe_helper
 		drm_kms_helper_common.o drm_dp_dual_mode_helper.o \
 		drm_simple_kms_helper.o drm_modeset_helper.o \
 		drm_scdc_helper.o drm_gem_framebuffer_helper.o \
-		drm_atomic_state_helper.o drm_damage_helper.o
+		drm_atomic_state_helper.o drm_damage_helper.o \
+		drm_format_helper.o

 drm_kms_helper-$(CONFIG_DRM_PANEL_BRIDGE) += bridge/panel.o
 drm_kms_helper-$(CONFIG_DRM_FBDEV_EMULATION) += drm_fb_helper.o
@ -56,7 +57,6 @@ obj-$(CONFIG_DRM_TTM)	+= ttm/
 obj-$(CONFIG_DRM_SCHED)	+= scheduler/
 obj-$(CONFIG_DRM_TDFX)	+= tdfx/
 obj-$(CONFIG_DRM_R128)	+= r128/
-obj-y			+= amd/lib/
 obj-$(CONFIG_HSA_AMD) += amd/amdkfd/
 obj-$(CONFIG_DRM_RADEON)+= radeon/
 obj-$(CONFIG_DRM_AMDGPU)+= amd/amdgpu/
@ -109,3 +109,7 @@ obj-$(CONFIG_DRM_TINYDRM) += tinydrm/
 obj-$(CONFIG_DRM_PL111) += pl111/
 obj-$(CONFIG_DRM_TVE200) += tve200/
 obj-$(CONFIG_DRM_XEN) += xen/
+obj-$(CONFIG_DRM_VBOXVIDEO) += vboxvideo/
+obj-$(CONFIG_DRM_LIMA)  += lima/
+obj-$(CONFIG_DRM_PANFROST) += panfrost/
+obj-$(CONFIG_DRM_ASPEED_GFX) += aspeed/
--- a/drivers/gpu/drm/amd/amdgpu/Makefile
+++ b/drivers/gpu/drm/amd/amdgpu/Makefile
@ -23,7 +23,7 @@
 # Makefile for the drm device driver.  This driver provides support for the
 # Direct Rendering Infrastructure (DRI) in XFree86 4.1.0 and higher.

-FULL_AMD_PATH=$(src)/..
+FULL_AMD_PATH=$(srctree)/$(src)/..
 DISPLAY_FOLDER_NAME=display
 FULL_AMD_DISPLAY_PATH = $(FULL_AMD_PATH)/$(DISPLAY_FOLDER_NAME)

@ -53,7 +53,8 @@ amdgpu-y += amdgpu_device.o amdgpu_kms.o \
 	amdgpu_ucode.o amdgpu_bo_list.o amdgpu_ctx.o amdgpu_sync.o \
 	amdgpu_gtt_mgr.o amdgpu_vram_mgr.o amdgpu_virt.o amdgpu_atomfirmware.o \
 	amdgpu_vf_error.o amdgpu_sched.o amdgpu_debugfs.o amdgpu_ids.o \
-	amdgpu_gmc.o amdgpu_xgmi.o amdgpu_csa.o
+	amdgpu_gmc.o amdgpu_xgmi.o amdgpu_csa.o amdgpu_ras.o amdgpu_vm_cpu.o \
+	amdgpu_vm_sdma.o

 # add asic specific block
 amdgpu-$(CONFIG_DRM_AMDGPU_CIK)+= cik.o cik_ih.o kv_smc.o kv_dpm.o \
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
@ -83,6 +83,7 @@
 #include "amdgpu_gem.h"
 #include "amdgpu_doorbell.h"
 #include "amdgpu_amdkfd.h"
+#include "amdgpu_smu.h"

 #define MAX_GPU_INSTANCE		16

@ -156,6 +157,8 @@ extern int amdgpu_emu_mode;
 extern uint amdgpu_smu_memory_pool_size;
 extern uint amdgpu_dc_feature_mask;
 extern struct amdgpu_mgpu_info mgpu_info;
+extern int amdgpu_ras_enable;
+extern uint amdgpu_ras_mask;

 #ifdef CONFIG_DRM_AMDGPU_SI
 extern int amdgpu_si_support;
@ -433,6 +436,12 @@ struct amdgpu_cs_chunk {
 	void			*kdata;
 };

+struct amdgpu_cs_post_dep {
+	struct drm_syncobj *syncobj;
+	struct dma_fence_chain *chain;
+	u64 point;
+};
+
 struct amdgpu_cs_parser {
 	struct amdgpu_device	*adev;
 	struct drm_file		*filp;
@ -462,8 +471,8 @@ struct amdgpu_cs_parser {
 	/* user fence */
 	struct amdgpu_bo_list_entry	uf_entry;

-	unsigned num_post_dep_syncobjs;
-	struct drm_syncobj **post_dep_syncobjs;
+	unsigned			num_post_deps;
+	struct amdgpu_cs_post_dep	*post_deps;
 };

 static inline u32 amdgpu_get_ib_value(struct amdgpu_cs_parser *p,
@ -702,7 +711,6 @@ enum amd_hw_ip_block_type {
 struct amd_powerplay {
 	void *pp_handle;
 	const struct amd_pm_funcs *pp_funcs;
-	uint32_t pp_feature;
 };

 #define AMDGPU_RESET_MAGIC_NUM 64
@ -825,6 +833,7 @@ struct amdgpu_device {
 	/* For pre-DCE11. DCE11 and later are in "struct amdgpu_device->dm" */
 	struct work_struct		hotplug_work;
 	struct amdgpu_irq_src		crtc_irq;
+	struct amdgpu_irq_src		vupdate_irq;
 	struct amdgpu_irq_src		pageflip_irq;
 	struct amdgpu_irq_src		hpd_irq;

@ -842,6 +851,9 @@ struct amdgpu_device {
 	struct amd_powerplay		powerplay;
 	bool				pp_force_state_enabled;

+	/* smu */
+	struct smu_context		smu;
+
 	/* dpm */
 	struct amdgpu_pm		pm;
 	u32				cg_flags;
@ -922,6 +934,8 @@ struct amdgpu_device {

 	int asic_reset_res;
 	struct work_struct		xgmi_reset_work;
+
+	bool                            in_baco_reset;
 };

 static inline struct amdgpu_device *amdgpu_ttm_adev(struct ttm_bo_device *bdev)
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c
@ -335,6 +335,43 @@ void amdgpu_amdkfd_free_gtt_mem(struct kgd_dev *kgd, void *mem_obj)
 	amdgpu_bo_unref(&(bo));
 }

+uint32_t amdgpu_amdkfd_get_fw_version(struct kgd_dev *kgd,
+				      enum kgd_engine_type type)
+{
+	struct amdgpu_device *adev = (struct amdgpu_device *)kgd;
+
+	switch (type) {
+	case KGD_ENGINE_PFP:
+		return adev->gfx.pfp_fw_version;
+
+	case KGD_ENGINE_ME:
+		return adev->gfx.me_fw_version;
+
+	case KGD_ENGINE_CE:
+		return adev->gfx.ce_fw_version;
+
+	case KGD_ENGINE_MEC1:
+		return adev->gfx.mec_fw_version;
+
+	case KGD_ENGINE_MEC2:
+		return adev->gfx.mec2_fw_version;
+
+	case KGD_ENGINE_RLC:
+		return adev->gfx.rlc_fw_version;
+
+	case KGD_ENGINE_SDMA1:
+		return adev->sdma.instance[0].fw_version;
+
+	case KGD_ENGINE_SDMA2:
+		return adev->sdma.instance[1].fw_version;
+
+	default:
+		return 0;
+	}
+
+	return 0;
+}
+
 void amdgpu_amdkfd_get_local_mem_info(struct kgd_dev *kgd,
 				      struct kfd_local_mem_info *mem_info)
 {
@ -640,4 +677,8 @@ int kgd2kfd_post_reset(struct kfd_dev *kfd)
 void kgd2kfd_interrupt(struct kfd_dev *kfd, const void *ih_ring_entry)
 {
 }
+
+void kgd2kfd_set_sram_ecc_flag(struct kfd_dev *kfd)
+{
+}
 #endif
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h
@ -81,6 +81,18 @@ struct amdgpu_kfd_dev {
 	uint64_t vram_used;
 };

+enum kgd_engine_type {
+	KGD_ENGINE_PFP = 1,
+	KGD_ENGINE_ME,
+	KGD_ENGINE_CE,
+	KGD_ENGINE_MEC1,
+	KGD_ENGINE_MEC2,
+	KGD_ENGINE_RLC,
+	KGD_ENGINE_SDMA1,
+	KGD_ENGINE_SDMA2,
+	KGD_ENGINE_MAX
+};
+
 struct amdgpu_amdkfd_fence *amdgpu_amdkfd_fence_create(u64 context,
 						       struct mm_struct *mm);
 bool amdkfd_fence_check_mm(struct dma_fence *f, struct mm_struct *mm);
@ -142,6 +154,8 @@ int amdgpu_amdkfd_alloc_gtt_mem(struct kgd_dev *kgd, size_t size,
 				void **mem_obj, uint64_t *gpu_addr,
 				void **cpu_ptr, bool mqd_gfx9);
 void amdgpu_amdkfd_free_gtt_mem(struct kgd_dev *kgd, void *mem_obj);
+uint32_t amdgpu_amdkfd_get_fw_version(struct kgd_dev *kgd,
+				      enum kgd_engine_type type);
 void amdgpu_amdkfd_get_local_mem_info(struct kgd_dev *kgd,
 				      struct kfd_local_mem_info *mem_info);
 uint64_t amdgpu_amdkfd_get_gpu_clock_counter(struct kgd_dev *kgd);
@ -230,5 +244,6 @@ int kgd2kfd_quiesce_mm(struct mm_struct *mm);
 int kgd2kfd_resume_mm(struct mm_struct *mm);
 int kgd2kfd_schedule_evict_and_restore_process(struct mm_struct *mm,
 					       struct dma_fence *fence);
+void kgd2kfd_set_sram_ecc_flag(struct kfd_dev *kfd);

 #endif /* AMDGPU_AMDKFD_H_INCLUDED */
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v7.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v7.c
@ -22,14 +22,12 @@

 #include <linux/fdtable.h>
 #include <linux/uaccess.h>
-#include <linux/firmware.h>
 #include <linux/mmu_context.h>
 #include <drm/drmP.h>
 #include "amdgpu.h"
 #include "amdgpu_amdkfd.h"
 #include "cikd.h"
 #include "cik_sdma.h"
-#include "amdgpu_ucode.h"
 #include "gfx_v7_0.h"
 #include "gca/gfx_7_2_d.h"
 #include "gca/gfx_7_2_enum.h"
@ -139,7 +137,6 @@ static bool get_atc_vmid_pasid_mapping_valid(struct kgd_dev *kgd, uint8_t vmid);
 static uint16_t get_atc_vmid_pasid_mapping_pasid(struct kgd_dev *kgd,
 							uint8_t vmid);

-static uint16_t get_fw_version(struct kgd_dev *kgd, enum kgd_engine_type type);
 static void set_scratch_backing_va(struct kgd_dev *kgd,
 					uint64_t va, uint32_t vmid);
 static void set_vm_context_page_table_base(struct kgd_dev *kgd, uint32_t vmid,
@ -191,7 +188,6 @@ static const struct kfd2kgd_calls kfd2kgd = {
 	.address_watch_get_offset = kgd_address_watch_get_offset,
 	.get_atc_vmid_pasid_mapping_pasid = get_atc_vmid_pasid_mapping_pasid,
 	.get_atc_vmid_pasid_mapping_valid = get_atc_vmid_pasid_mapping_valid,
-	.get_fw_version = get_fw_version,
 	.set_scratch_backing_va = set_scratch_backing_va,
 	.get_tile_config = get_tile_config,
 	.set_vm_context_page_table_base = set_vm_context_page_table_base,
@ -792,63 +788,6 @@ static void set_scratch_backing_va(struct kgd_dev *kgd,
 	unlock_srbm(kgd);
 }

-static uint16_t get_fw_version(struct kgd_dev *kgd, enum kgd_engine_type type)
-{
-	struct amdgpu_device *adev = (struct amdgpu_device *) kgd;
-	const union amdgpu_firmware_header *hdr;
-
-	switch (type) {
-	case KGD_ENGINE_PFP:
-		hdr = (const union amdgpu_firmware_header *)
-						adev->gfx.pfp_fw->data;
-		break;
-
-	case KGD_ENGINE_ME:
-		hdr = (const union amdgpu_firmware_header *)
-						adev->gfx.me_fw->data;
-		break;
-
-	case KGD_ENGINE_CE:
-		hdr = (const union amdgpu_firmware_header *)
-						adev->gfx.ce_fw->data;
-		break;
-
-	case KGD_ENGINE_MEC1:
-		hdr = (const union amdgpu_firmware_header *)
-						adev->gfx.mec_fw->data;
-		break;
-
-	case KGD_ENGINE_MEC2:
-		hdr = (const union amdgpu_firmware_header *)
-						adev->gfx.mec2_fw->data;
-		break;
-
-	case KGD_ENGINE_RLC:
-		hdr = (const union amdgpu_firmware_header *)
-						adev->gfx.rlc_fw->data;
-		break;
-
-	case KGD_ENGINE_SDMA1:
-		hdr = (const union amdgpu_firmware_header *)
-						adev->sdma.instance[0].fw->data;
-		break;
-
-	case KGD_ENGINE_SDMA2:
-		hdr = (const union amdgpu_firmware_header *)
-						adev->sdma.instance[1].fw->data;
-		break;
-
-	default:
-		return 0;
-	}
-
-	if (hdr == NULL)
-		return 0;
-
-	/* Only 12 bit in use*/
-	return hdr->common.ucode_version;
-}
-
 static void set_vm_context_page_table_base(struct kgd_dev *kgd, uint32_t vmid,
 			uint64_t page_table_base)
 {
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v8.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v8.c
@ -23,12 +23,10 @@
 #include <linux/module.h>
 #include <linux/fdtable.h>
 #include <linux/uaccess.h>
-#include <linux/firmware.h>
 #include <linux/mmu_context.h>
 #include <drm/drmP.h>
 #include "amdgpu.h"
 #include "amdgpu_amdkfd.h"
-#include "amdgpu_ucode.h"
 #include "gfx_v8_0.h"
 #include "gca/gfx_8_0_sh_mask.h"
 #include "gca/gfx_8_0_d.h"
@ -95,7 +93,6 @@ static bool get_atc_vmid_pasid_mapping_valid(struct kgd_dev *kgd,
 		uint8_t vmid);
 static uint16_t get_atc_vmid_pasid_mapping_pasid(struct kgd_dev *kgd,
 		uint8_t vmid);
-static uint16_t get_fw_version(struct kgd_dev *kgd, enum kgd_engine_type type);
 static void set_scratch_backing_va(struct kgd_dev *kgd,
 					uint64_t va, uint32_t vmid);
 static void set_vm_context_page_table_base(struct kgd_dev *kgd, uint32_t vmid,
@ -148,7 +145,6 @@ static const struct kfd2kgd_calls kfd2kgd = {
 			get_atc_vmid_pasid_mapping_pasid,
 	.get_atc_vmid_pasid_mapping_valid =
 			get_atc_vmid_pasid_mapping_valid,
-	.get_fw_version = get_fw_version,
 	.set_scratch_backing_va = set_scratch_backing_va,
 	.get_tile_config = get_tile_config,
 	.set_vm_context_page_table_base = set_vm_context_page_table_base,
@ -751,63 +747,6 @@ static void set_scratch_backing_va(struct kgd_dev *kgd,
 	unlock_srbm(kgd);
 }

-static uint16_t get_fw_version(struct kgd_dev *kgd, enum kgd_engine_type type)
-{
-	struct amdgpu_device *adev = (struct amdgpu_device *) kgd;
-	const union amdgpu_firmware_header *hdr;
-
-	switch (type) {
-	case KGD_ENGINE_PFP:
-		hdr = (const union amdgpu_firmware_header *)
-						adev->gfx.pfp_fw->data;
-		break;
-
-	case KGD_ENGINE_ME:
-		hdr = (const union amdgpu_firmware_header *)
-						adev->gfx.me_fw->data;
-		break;
-
-	case KGD_ENGINE_CE:
-		hdr = (const union amdgpu_firmware_header *)
-						adev->gfx.ce_fw->data;
-		break;
-
-	case KGD_ENGINE_MEC1:
-		hdr = (const union amdgpu_firmware_header *)
-						adev->gfx.mec_fw->data;
-		break;
-
-	case KGD_ENGINE_MEC2:
-		hdr = (const union amdgpu_firmware_header *)
-						adev->gfx.mec2_fw->data;
-		break;
-
-	case KGD_ENGINE_RLC:
-		hdr = (const union amdgpu_firmware_header *)
-						adev->gfx.rlc_fw->data;
-		break;
-
-	case KGD_ENGINE_SDMA1:
-		hdr = (const union amdgpu_firmware_header *)
-						adev->sdma.instance[0].fw->data;
-		break;
-
-	case KGD_ENGINE_SDMA2:
-		hdr = (const union amdgpu_firmware_header *)
-						adev->sdma.instance[1].fw->data;
-		break;
-
-	default:
-		return 0;
-	}
-
-	if (hdr == NULL)
-		return 0;
-
-	/* Only 12 bit in use*/
-	return hdr->common.ucode_version;
-}
-
 static void set_vm_context_page_table_base(struct kgd_dev *kgd, uint32_t vmid,
 		uint64_t page_table_base)
 {
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v9.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v9.c
@ -25,12 +25,10 @@
 #include <linux/module.h>
 #include <linux/fdtable.h>
 #include <linux/uaccess.h>
-#include <linux/firmware.h>
 #include <linux/mmu_context.h>
 #include <drm/drmP.h>
 #include "amdgpu.h"
 #include "amdgpu_amdkfd.h"
-#include "amdgpu_ucode.h"
 #include "soc15_hw_ip.h"
 #include "gc/gc_9_0_offset.h"
 #include "gc/gc_9_0_sh_mask.h"
@ -111,7 +109,6 @@ static uint16_t get_atc_vmid_pasid_mapping_pasid(struct kgd_dev *kgd,
 		uint8_t vmid);
 static void set_vm_context_page_table_base(struct kgd_dev *kgd, uint32_t vmid,
 		uint64_t page_table_base);
-static uint16_t get_fw_version(struct kgd_dev *kgd, enum kgd_engine_type type);
 static void set_scratch_backing_va(struct kgd_dev *kgd,
 					uint64_t va, uint32_t vmid);
 static int invalidate_tlbs(struct kgd_dev *kgd, uint16_t pasid);
@ -158,7 +155,6 @@ static const struct kfd2kgd_calls kfd2kgd = {
 			get_atc_vmid_pasid_mapping_pasid,
 	.get_atc_vmid_pasid_mapping_valid =
 			get_atc_vmid_pasid_mapping_valid,
-	.get_fw_version = get_fw_version,
 	.set_scratch_backing_va = set_scratch_backing_va,
 	.get_tile_config = amdgpu_amdkfd_get_tile_config,
 	.set_vm_context_page_table_base = set_vm_context_page_table_base,
@ -874,56 +870,6 @@ static void set_scratch_backing_va(struct kgd_dev *kgd,
 	 */
 }

-/* FIXME: Does this need to be ASIC-specific code? */
-static uint16_t get_fw_version(struct kgd_dev *kgd, enum kgd_engine_type type)
-{
-	struct amdgpu_device *adev = (struct amdgpu_device *) kgd;
-	const union amdgpu_firmware_header *hdr;
-
-	switch (type) {
-	case KGD_ENGINE_PFP:
-		hdr = (const union amdgpu_firmware_header *)adev->gfx.pfp_fw->data;
-		break;
-
-	case KGD_ENGINE_ME:
-		hdr = (const union amdgpu_firmware_header *)adev->gfx.me_fw->data;
-		break;
-
-	case KGD_ENGINE_CE:
-		hdr = (const union amdgpu_firmware_header *)adev->gfx.ce_fw->data;
-		break;
-
-	case KGD_ENGINE_MEC1:
-		hdr = (const union amdgpu_firmware_header *)adev->gfx.mec_fw->data;
-		break;
-
-	case KGD_ENGINE_MEC2:
-		hdr = (const union amdgpu_firmware_header *)adev->gfx.mec2_fw->data;
-		break;
-
-	case KGD_ENGINE_RLC:
-		hdr = (const union amdgpu_firmware_header *)adev->gfx.rlc_fw->data;
-		break;
-
-	case KGD_ENGINE_SDMA1:
-		hdr = (const union amdgpu_firmware_header *)adev->sdma.instance[0].fw->data;
-		break;
-
-	case KGD_ENGINE_SDMA2:
-		hdr = (const union amdgpu_firmware_header *)adev->sdma.instance[1].fw->data;
-		break;
-
-	default:
-		return 0;
-	}
-
-	if (hdr == NULL)
-		return 0;
-
-	/* Only 12 bit in use*/
-	return hdr->common.ucode_version;
-}
-
 static void set_vm_context_page_table_base(struct kgd_dev *kgd, uint32_t vmid,
 		uint64_t page_table_base)
 {
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c
@ -410,15 +410,7 @@ static int add_bo_to_vm(struct amdgpu_device *adev, struct kgd_mem *mem,
 	if (p_bo_va_entry)
 		*p_bo_va_entry = bo_va_entry;

-	/* Allocate new page tables if needed and validate
-	 * them.
-	 */
-	ret = amdgpu_vm_alloc_pts(adev, vm, va, amdgpu_bo_size(bo));
-	if (ret) {
-		pr_err("Failed to allocate pts, err=%d\n", ret);
-		goto err_alloc_pts;
-	}
-
+	/* Allocate validate page tables if needed */
 	ret = vm_validate_pt_pd_bos(vm);
 	if (ret) {
 		pr_err("validate_pt_pd_bos() failed\n");
@ -741,13 +733,7 @@ static int update_gpuvm_pte(struct amdgpu_device *adev,
 		struct amdgpu_sync *sync)
 {
 	int ret;
-	struct amdgpu_vm *vm;
-	struct amdgpu_bo_va *bo_va;
-	struct amdgpu_bo *bo;
-
-	bo_va = entry->bo_va;
-	vm = bo_va->base.vm;
-	bo = bo_va->base.bo;
+	struct amdgpu_bo_va *bo_va = entry->bo_va;

 	/* Update the page tables  */
 	ret = amdgpu_vm_bo_update(adev, bo_va, false);
@ -906,7 +892,8 @@ static int init_kfd_vm(struct amdgpu_vm *vm, void **process_info,
 		pr_err("validate_pt_pd_bos() failed\n");
 		goto validate_pd_fail;
 	}
-	amdgpu_bo_sync_wait(vm->root.base.bo, AMDGPU_FENCE_OWNER_KFD, false);
+	ret = amdgpu_bo_sync_wait(vm->root.base.bo,
+				  AMDGPU_FENCE_OWNER_KFD, false);
 	if (ret)
 		goto wait_pd_fail;
 	amdgpu_bo_fence(vm->root.base.bo,
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.c
@ -28,8 +28,6 @@
 #include "atom.h"
 #include "atombios.h"

-#define get_index_into_master_table(master_table, table_name) (offsetof(struct master_table, table_name) / sizeof(uint16_t))
-
 bool amdgpu_atomfirmware_gpu_supports_virtualization(struct amdgpu_device *adev)
 {
 	int index = get_index_into_master_table(atom_master_list_of_data_tables_v2_1,
@ -238,10 +236,71 @@ int amdgpu_atomfirmware_get_vram_type(struct amdgpu_device *adev)
 	return 0;
 }

+/*
+ * Return true if vbios enabled ecc by default, if umc info table is available
+ * or false if ecc is not enabled or umc info table is not available
+ */
+bool amdgpu_atomfirmware_mem_ecc_supported(struct amdgpu_device *adev)
+{
+	struct amdgpu_mode_info *mode_info = &adev->mode_info;
+	int index;
+	u16 data_offset, size;
+	union umc_info *umc_info;
+	u8 frev, crev;
+	bool ecc_default_enabled = false;
+
+	index = get_index_into_master_table(atom_master_list_of_data_tables_v2_1,
+			umc_info);
+
+	if (amdgpu_atom_parse_data_header(mode_info->atom_context,
+				index, &size, &frev, &crev, &data_offset)) {
+		/* support umc_info 3.1+ */
+		if ((frev == 3 && crev >= 1) || (frev > 3)) {
+			umc_info = (union umc_info *)
+				(mode_info->atom_context->bios + data_offset);
+			ecc_default_enabled =
+				(le32_to_cpu(umc_info->v31.umc_config) &
+				 UMC_CONFIG__DEFAULT_MEM_ECC_ENABLE) ? true : false;
+		}
+	}
+
+	return ecc_default_enabled;
+}
+
 union firmware_info {
 	struct atom_firmware_info_v3_1 v31;
 };

+/*
+ * Return true if vbios supports sram ecc or false if not
+ */
+bool amdgpu_atomfirmware_sram_ecc_supported(struct amdgpu_device *adev)
+{
+	struct amdgpu_mode_info *mode_info = &adev->mode_info;
+	int index;
+	u16 data_offset, size;
+	union firmware_info *firmware_info;
+	u8 frev, crev;
+	bool sram_ecc_supported = false;
+
+	index = get_index_into_master_table(atom_master_list_of_data_tables_v2_1,
+			firmwareinfo);
+
+	if (amdgpu_atom_parse_data_header(adev->mode_info.atom_context,
+				index, &size, &frev, &crev, &data_offset)) {
+		/* support firmware_info 3.1 + */
+		if ((frev == 3 && crev >=1) || (frev > 3)) {
+			firmware_info = (union firmware_info *)
+				(mode_info->atom_context->bios + data_offset);
+			sram_ecc_supported =
+				(le32_to_cpu(firmware_info->v31.firmware_capability) &
+				 ATOM_FIRMWARE_CAP_SRAM_ECC) ? true : false;
+		}
+	}
+
+	return sram_ecc_supported;
+}
+
 union smu_info {
 	struct atom_smu_info_v3_1 v31;
 };
@ -346,11 +405,11 @@ int amdgpu_atomfirmware_get_gfx_info(struct amdgpu_device *adev)
 			(mode_info->atom_context->bios + data_offset);
 		switch (crev) {
 		case 4:
-			adev->gfx.config.max_shader_engines = gfx_info->v24.gc_num_se;
-			adev->gfx.config.max_cu_per_sh = gfx_info->v24.gc_num_cu_per_sh;
-			adev->gfx.config.max_sh_per_se = gfx_info->v24.gc_num_sh_per_se;
-			adev->gfx.config.max_backends_per_se = gfx_info->v24.gc_num_rb_per_se;
-			adev->gfx.config.max_texture_channel_caches = gfx_info->v24.gc_num_tccs;
+			adev->gfx.config.max_shader_engines = gfx_info->v24.max_shader_engines;
+			adev->gfx.config.max_cu_per_sh = gfx_info->v24.max_cu_per_sh;
+			adev->gfx.config.max_sh_per_se = gfx_info->v24.max_sh_per_se;
+			adev->gfx.config.max_backends_per_se = gfx_info->v24.max_backends_per_se;
+			adev->gfx.config.max_texture_channel_caches = gfx_info->v24.max_texture_channel_caches;
 			adev->gfx.config.max_gprs = le16_to_cpu(gfx_info->v24.gc_num_gprs);
 			adev->gfx.config.max_gs_threads = gfx_info->v24.gc_num_max_gs_thds;
 			adev->gfx.config.gs_vgt_table_depth = gfx_info->v24.gc_gs_table_depth;
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.h
@ -24,6 +24,8 @@
 #ifndef __AMDGPU_ATOMFIRMWARE_H__
 #define __AMDGPU_ATOMFIRMWARE_H__

+#define get_index_into_master_table(master_table, table_name) (offsetof(struct master_table, table_name) / sizeof(uint16_t))
+
 bool amdgpu_atomfirmware_gpu_supports_virtualization(struct amdgpu_device *adev);
 void amdgpu_atomfirmware_scratch_regs_init(struct amdgpu_device *adev);
 int amdgpu_atomfirmware_allocate_fb_scratch(struct amdgpu_device *adev);
@ -31,5 +33,7 @@ int amdgpu_atomfirmware_get_vram_width(struct amdgpu_device *adev);
 int amdgpu_atomfirmware_get_vram_type(struct amdgpu_device *adev);
 int amdgpu_atomfirmware_get_clock_info(struct amdgpu_device *adev);
 int amdgpu_atomfirmware_get_gfx_info(struct amdgpu_device *adev);
+bool amdgpu_atomfirmware_mem_ecc_supported(struct amdgpu_device *adev);
+bool amdgpu_atomfirmware_sram_ecc_supported(struct amdgpu_device *adev);

 #endif
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
@ -215,6 +215,8 @@ static int amdgpu_cs_parser_init(struct amdgpu_cs_parser *p, union drm_amdgpu_cs
 		case AMDGPU_CHUNK_ID_SYNCOBJ_IN:
 		case AMDGPU_CHUNK_ID_SYNCOBJ_OUT:
 		case AMDGPU_CHUNK_ID_SCHEDULED_DEPENDENCIES:
+		case AMDGPU_CHUNK_ID_SYNCOBJ_TIMELINE_WAIT:
+		case AMDGPU_CHUNK_ID_SYNCOBJ_TIMELINE_SIGNAL:
 			break;

 		default:
@ -804,9 +806,11 @@ static void amdgpu_cs_parser_fini(struct amdgpu_cs_parser *parser, int error,
 		ttm_eu_backoff_reservation(&parser->ticket,
 					   &parser->validated);

-	for (i = 0; i < parser->num_post_dep_syncobjs; i++)
-		drm_syncobj_put(parser->post_dep_syncobjs[i]);
-	kfree(parser->post_dep_syncobjs);
+	for (i = 0; i < parser->num_post_deps; i++) {
+		drm_syncobj_put(parser->post_deps[i].syncobj);
+		kfree(parser->post_deps[i].chain);
+	}
+	kfree(parser->post_deps);

 	dma_fence_put(parser->fence);

@ -1117,13 +1121,18 @@ static int amdgpu_cs_process_fence_dep(struct amdgpu_cs_parser *p,
 }

 static int amdgpu_syncobj_lookup_and_add_to_sync(struct amdgpu_cs_parser *p,
-						 uint32_t handle)
+						 uint32_t handle, u64 point,
+						 u64 flags)
 {
-	int r;
 	struct dma_fence *fence;
-	r = drm_syncobj_find_fence(p->filp, handle, 0, 0, &fence);
-	if (r)
+	int r;
+
+	r = drm_syncobj_find_fence(p->filp, handle, point, flags, &fence);
+	if (r) {
+		DRM_ERROR("syncobj %u failed to find fence @ %llu (%d)!\n",
+			  handle, point, r);
 		return r;
+	}

 	r = amdgpu_sync_fence(p->adev, &p->job->sync, fence, true);
 	dma_fence_put(fence);
@ -1134,46 +1143,118 @@ static int amdgpu_syncobj_lookup_and_add_to_sync(struct amdgpu_cs_parser *p,
 static int amdgpu_cs_process_syncobj_in_dep(struct amdgpu_cs_parser *p,
 					    struct amdgpu_cs_chunk *chunk)
 {
+	struct drm_amdgpu_cs_chunk_sem *deps;
 	unsigned num_deps;
 	int i, r;
-	struct drm_amdgpu_cs_chunk_sem *deps;

 	deps = (struct drm_amdgpu_cs_chunk_sem *)chunk->kdata;
 	num_deps = chunk->length_dw * 4 /
 		sizeof(struct drm_amdgpu_cs_chunk_sem);
-
 	for (i = 0; i < num_deps; ++i) {
-		r = amdgpu_syncobj_lookup_and_add_to_sync(p, deps[i].handle);
+		r = amdgpu_syncobj_lookup_and_add_to_sync(p, deps[i].handle,
+							  0, 0);
 		if (r)
 			return r;
 	}
+
+	return 0;
+}
+
+
+static int amdgpu_cs_process_syncobj_timeline_in_dep(struct amdgpu_cs_parser *p,
+						     struct amdgpu_cs_chunk *chunk)
+{
+	struct drm_amdgpu_cs_chunk_syncobj *syncobj_deps;
+	unsigned num_deps;
+	int i, r;
+
+	syncobj_deps = (struct drm_amdgpu_cs_chunk_syncobj *)chunk->kdata;
+	num_deps = chunk->length_dw * 4 /
+		sizeof(struct drm_amdgpu_cs_chunk_syncobj);
+	for (i = 0; i < num_deps; ++i) {
+		r = amdgpu_syncobj_lookup_and_add_to_sync(p,
+							  syncobj_deps[i].handle,
+							  syncobj_deps[i].point,
+							  syncobj_deps[i].flags);
+		if (r)
+			return r;
+	}
+
 	return 0;
 }

 static int amdgpu_cs_process_syncobj_out_dep(struct amdgpu_cs_parser *p,
 					     struct amdgpu_cs_chunk *chunk)
 {
+	struct drm_amdgpu_cs_chunk_sem *deps;
 	unsigned num_deps;
 	int i;
-	struct drm_amdgpu_cs_chunk_sem *deps;
+
 	deps = (struct drm_amdgpu_cs_chunk_sem *)chunk->kdata;
 	num_deps = chunk->length_dw * 4 /
 		sizeof(struct drm_amdgpu_cs_chunk_sem);

-	p->post_dep_syncobjs = kmalloc_array(num_deps,
-					     sizeof(struct drm_syncobj *),
-					     GFP_KERNEL);
-	p->num_post_dep_syncobjs = 0;
+	p->post_deps = kmalloc_array(num_deps, sizeof(*p->post_deps),
+				     GFP_KERNEL);
+	p->num_post_deps = 0;

-	if (!p->post_dep_syncobjs)
+	if (!p->post_deps)
+		return -ENOMEM;
+
+
+	for (i = 0; i < num_deps; ++i) {
+		p->post_deps[i].syncobj =
+			drm_syncobj_find(p->filp, deps[i].handle);
+		if (!p->post_deps[i].syncobj)
+			return -EINVAL;
+		p->post_deps[i].chain = NULL;
+		p->post_deps[i].point = 0;
+		p->num_post_deps++;
+	}
+
+	return 0;
+}
+
+
+static int amdgpu_cs_process_syncobj_timeline_out_dep(struct amdgpu_cs_parser *p,
+						      struct amdgpu_cs_chunk
+						      *chunk)
+{
+	struct drm_amdgpu_cs_chunk_syncobj *syncobj_deps;
+	unsigned num_deps;
+	int i;
+
+	syncobj_deps = (struct drm_amdgpu_cs_chunk_syncobj *)chunk->kdata;
+	num_deps = chunk->length_dw * 4 /
+		sizeof(struct drm_amdgpu_cs_chunk_syncobj);
+
+	p->post_deps = kmalloc_array(num_deps, sizeof(*p->post_deps),
+				     GFP_KERNEL);
+	p->num_post_deps = 0;
+
+	if (!p->post_deps)
 		return -ENOMEM;

 	for (i = 0; i < num_deps; ++i) {
-		p->post_dep_syncobjs[i] = drm_syncobj_find(p->filp, deps[i].handle);
-		if (!p->post_dep_syncobjs[i])
+		struct amdgpu_cs_post_dep *dep = &p->post_deps[i];
+
+		dep->chain = NULL;
+		if (syncobj_deps[i].point) {
+			dep->chain = kmalloc(sizeof(*dep->chain), GFP_KERNEL);
+			if (!dep->chain)
+				return -ENOMEM;
+		}
+
+		dep->syncobj = drm_syncobj_find(p->filp,
+						syncobj_deps[i].handle);
+		if (!dep->syncobj) {
+			kfree(dep->chain);
 			return -EINVAL;
-		p->num_post_dep_syncobjs++;
+		}
+		dep->point = syncobj_deps[i].point;
+		p->num_post_deps++;
 	}
+
 	return 0;
 }

@ -1187,19 +1268,33 @@ static int amdgpu_cs_dependencies(struct amdgpu_device *adev,

 		chunk = &p->chunks[i];

-		if (chunk->chunk_id == AMDGPU_CHUNK_ID_DEPENDENCIES ||
-		    chunk->chunk_id == AMDGPU_CHUNK_ID_SCHEDULED_DEPENDENCIES) {
+		switch (chunk->chunk_id) {
+		case AMDGPU_CHUNK_ID_DEPENDENCIES:
+		case AMDGPU_CHUNK_ID_SCHEDULED_DEPENDENCIES:
 			r = amdgpu_cs_process_fence_dep(p, chunk);
 			if (r)
 				return r;
-		} else if (chunk->chunk_id == AMDGPU_CHUNK_ID_SYNCOBJ_IN) {
+			break;
+		case AMDGPU_CHUNK_ID_SYNCOBJ_IN:
 			r = amdgpu_cs_process_syncobj_in_dep(p, chunk);
 			if (r)
 				return r;
-		} else if (chunk->chunk_id == AMDGPU_CHUNK_ID_SYNCOBJ_OUT) {
+			break;
+		case AMDGPU_CHUNK_ID_SYNCOBJ_OUT:
 			r = amdgpu_cs_process_syncobj_out_dep(p, chunk);
 			if (r)
 				return r;
+			break;
+		case AMDGPU_CHUNK_ID_SYNCOBJ_TIMELINE_WAIT:
+			r = amdgpu_cs_process_syncobj_timeline_in_dep(p, chunk);
+			if (r)
+				return r;
+			break;
+		case AMDGPU_CHUNK_ID_SYNCOBJ_TIMELINE_SIGNAL:
+			r = amdgpu_cs_process_syncobj_timeline_out_dep(p, chunk);
+			if (r)
+				return r;
+			break;
 		}
 	}

@ -1210,8 +1305,17 @@ static void amdgpu_cs_post_dependencies(struct amdgpu_cs_parser *p)
 {
 	int i;

-	for (i = 0; i < p->num_post_dep_syncobjs; ++i)
-		drm_syncobj_replace_fence(p->post_dep_syncobjs[i], p->fence);
+	for (i = 0; i < p->num_post_deps; ++i) {
+		if (p->post_deps[i].chain && p->post_deps[i].point) {
+			drm_syncobj_add_point(p->post_deps[i].syncobj,
+					      p->post_deps[i].chain,
+					      p->fence, p->post_deps[i].point);
+			p->post_deps[i].chain = NULL;
+		} else {
+			drm_syncobj_replace_fence(p->post_deps[i].syncobj,
+						  p->fence);
+		}
+	}
 }

 static int amdgpu_cs_submit(struct amdgpu_cs_parser *p,
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_csa.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_csa.c
@ -92,15 +92,6 @@ int amdgpu_map_static_csa(struct amdgpu_device *adev, struct amdgpu_vm *vm,
 		return -ENOMEM;
 	}

-	r = amdgpu_vm_alloc_pts(adev, (*bo_va)->base.vm, csa_addr,
-				size);
-	if (r) {
-		DRM_ERROR("failed to allocate pts for static CSA, err=%d\n", r);
-		amdgpu_vm_bo_rmv(adev, *bo_va);
-		ttm_eu_backoff_reservation(&ticket, &list);
-		return r;
-	}
-
 	r = amdgpu_vm_bo_map(adev, *bo_va, csa_addr, 0, size,
 			     AMDGPU_PTE_READABLE | AMDGPU_PTE_WRITEABLE |
 			     AMDGPU_PTE_EXECUTABLE);
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c
@ -26,6 +26,7 @@
 #include <drm/drm_auth.h>
 #include "amdgpu.h"
 #include "amdgpu_sched.h"
+#include "amdgpu_ras.h"

 #define to_amdgpu_ctx_entity(e)	\
 	container_of((e), struct amdgpu_ctx_entity, entity)
@ -344,6 +345,7 @@ static int amdgpu_ctx_query2(struct amdgpu_device *adev,
 {
 	struct amdgpu_ctx *ctx;
 	struct amdgpu_ctx_mgr *mgr;
+	uint32_t ras_counter;

 	if (!fpriv)
 		return -EINVAL;
@ -368,6 +370,21 @@ static int amdgpu_ctx_query2(struct amdgpu_device *adev,
 	if (atomic_read(&ctx->guilty))
 		out->state.flags |= AMDGPU_CTX_QUERY2_FLAGS_GUILTY;

+	/*query ue count*/
+	ras_counter = amdgpu_ras_query_error_count(adev, false);
+	/*ras counter is monotonic increasing*/
+	if (ras_counter != ctx->ras_counter_ue) {
+		out->state.flags |= AMDGPU_CTX_QUERY2_FLAGS_RAS_UE;
+		ctx->ras_counter_ue = ras_counter;
+	}
+
+	/*query ce count*/
+	ras_counter = amdgpu_ras_query_error_count(adev, true);
+	if (ras_counter != ctx->ras_counter_ce) {
+		out->state.flags |= AMDGPU_CTX_QUERY2_FLAGS_RAS_CE;
+		ctx->ras_counter_ce = ras_counter;
+	}
+
 	mutex_unlock(&mgr->lock);
 	return 0;
 }
@ -541,32 +558,26 @@ void amdgpu_ctx_mgr_init(struct amdgpu_ctx_mgr *mgr)
 	idr_init(&mgr->ctx_handles);
 }

-void amdgpu_ctx_mgr_entity_flush(struct amdgpu_ctx_mgr *mgr)
+long amdgpu_ctx_mgr_entity_flush(struct amdgpu_ctx_mgr *mgr, long timeout)
 {
 	unsigned num_entities = amdgput_ctx_total_num_entities();
 	struct amdgpu_ctx *ctx;
 	struct idr *idp;
 	uint32_t id, i;
-	long max_wait = MAX_WAIT_SCHED_ENTITY_Q_EMPTY;

 	idp = &mgr->ctx_handles;

 	mutex_lock(&mgr->lock);
 	idr_for_each_entry(idp, ctx, id) {
-
-		if (!ctx->adev) {
-			mutex_unlock(&mgr->lock);
-			return;
-		}
-
 		for (i = 0; i < num_entities; i++) {
 			struct drm_sched_entity *entity;

 			entity = &ctx->entities[0][i].entity;
-			max_wait = drm_sched_entity_flush(entity, max_wait);
+			timeout = drm_sched_entity_flush(entity, timeout);
 		}
 	}
 	mutex_unlock(&mgr->lock);
+	return timeout;
 }

 void amdgpu_ctx_mgr_entity_fini(struct amdgpu_ctx_mgr *mgr)
@ -579,10 +590,6 @@ void amdgpu_ctx_mgr_entity_fini(struct amdgpu_ctx_mgr *mgr)
 	idp = &mgr->ctx_handles;

 	idr_for_each_entry(idp, ctx, id) {
-
-		if (!ctx->adev)
-			return;
-
 		if (kref_read(&ctx->refcount) != 1) {
 			DRM_ERROR("ctx %p is still alive\n", ctx);
 			continue;
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.h
@ -49,6 +49,8 @@ struct amdgpu_ctx {
 	enum drm_sched_priority		override_priority;
 	struct mutex			lock;
 	atomic_t			guilty;
+	uint32_t			ras_counter_ce;
+	uint32_t			ras_counter_ue;
 };

 struct amdgpu_ctx_mgr {
@ -82,7 +84,7 @@ int amdgpu_ctx_wait_prev_fence(struct amdgpu_ctx *ctx,

 void amdgpu_ctx_mgr_init(struct amdgpu_ctx_mgr *mgr);
 void amdgpu_ctx_mgr_entity_fini(struct amdgpu_ctx_mgr *mgr);
-void amdgpu_ctx_mgr_entity_flush(struct amdgpu_ctx_mgr *mgr);
+long amdgpu_ctx_mgr_entity_flush(struct amdgpu_ctx_mgr *mgr, long timeout);
 void amdgpu_ctx_mgr_fini(struct amdgpu_ctx_mgr *mgr);

 #endif
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c
@ -568,10 +568,9 @@ static ssize_t amdgpu_debugfs_sensor_read(struct file *f, char __user *buf,
 	idx = *pos >> 2;

 	valuesize = sizeof(values);
-	if (adev->powerplay.pp_funcs && adev->powerplay.pp_funcs->read_sensor)
-		r = amdgpu_dpm_read_sensor(adev, idx, &values[0], &valuesize);
-	else
-		return -EINVAL;
+	r = amdgpu_dpm_read_sensor(adev, idx, &values[0], &valuesize);
+	if (r)
+		return r;

 	if (size > valuesize)
 		return -EINVAL;
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
@ -60,6 +60,7 @@
 #include "amdgpu_pm.h"

 #include "amdgpu_xgmi.h"
+#include "amdgpu_ras.h"

 MODULE_FIRMWARE("amdgpu/vega10_gpu_info.bin");
 MODULE_FIRMWARE("amdgpu/vega12_gpu_info.bin");
@ -1506,7 +1507,9 @@ static int amdgpu_device_ip_early_init(struct amdgpu_device *adev)
 			return -EAGAIN;
 	}

-	adev->powerplay.pp_feature = amdgpu_pp_feature_mask;
+	adev->pm.pp_feature = amdgpu_pp_feature_mask;
+	if (amdgpu_sriov_vf(adev))
+		adev->pm.pp_feature &= ~PP_GFXOFF_MASK;

 	for (i = 0; i < adev->num_ip_blocks; i++) {
 		if ((amdgpu_ip_block_mask & (1 << i)) == 0) {
@ -1638,6 +1641,10 @@ static int amdgpu_device_ip_init(struct amdgpu_device *adev)
 {
 	int i, r;

+	r = amdgpu_ras_init(adev);
+	if (r)
+		return r;
+
 	for (i = 0; i < adev->num_ip_blocks; i++) {
 		if (!adev->ip_blocks[i].status.valid)
 			continue;
@ -1681,6 +1688,13 @@ static int amdgpu_device_ip_init(struct amdgpu_device *adev)
 		}
 	}

+	r = amdgpu_ib_pool_init(adev);
+	if (r) {
+		dev_err(adev->dev, "IB initialization failed (%d).\n", r);
+		amdgpu_vf_error_put(adev, AMDGIM_ERROR_VF_IB_INIT_FAIL, 0, r);
+		goto init_failed;
+	}
+
 	r = amdgpu_ucode_create_bo(adev); /* create ucode bo when sw_init complete*/
 	if (r)
 		goto init_failed;
@ -1869,6 +1883,8 @@ static int amdgpu_device_ip_fini(struct amdgpu_device *adev)
 {
 	int i, r;

+	amdgpu_ras_pre_fini(adev);
+
 	if (adev->gmc.xgmi.num_physical_nodes > 1)
 		amdgpu_xgmi_remove_device(adev);

@ -1917,6 +1933,7 @@ static int amdgpu_device_ip_fini(struct amdgpu_device *adev)
 			amdgpu_free_static_csa(&adev->virt.csa_obj);
 			amdgpu_device_wb_fini(adev);
 			amdgpu_device_vram_scratch_fini(adev);
+			amdgpu_ib_pool_fini(adev);
 		}

 		r = adev->ip_blocks[i].version->funcs->sw_fini((void *)adev);
@ -1937,6 +1954,8 @@ static int amdgpu_device_ip_fini(struct amdgpu_device *adev)
 		adev->ip_blocks[i].status.late_initialized = false;
 	}

+	amdgpu_ras_fini(adev);
+
 	if (amdgpu_sriov_vf(adev))
 		if (amdgpu_virt_release_full_gpu(adev, false))
 			DRM_ERROR("failed to release exclusive mode on fini\n");
@ -1999,6 +2018,10 @@ static void amdgpu_device_ip_late_init_func_handler(struct work_struct *work)
 	r = amdgpu_device_enable_mgpu_fan_boost();
 	if (r)
 		DRM_ERROR("enable mgpu fan boost failed (%d).\n", r);
+
+	/*set to low pstate by default */
+	amdgpu_xgmi_set_pstate(adev, 0);
+
 }

 static void amdgpu_device_delay_enable_gfx_off(struct work_struct *work)
@ -2369,7 +2392,7 @@ static void amdgpu_device_xgmi_reset_func(struct work_struct *__work)

 	adev->asic_reset_res =  amdgpu_asic_reset(adev);
 	if (adev->asic_reset_res)
-		DRM_WARN("ASIC reset failed with err r, %d for drm dev, %s",
+		DRM_WARN("ASIC reset failed with error, %d for drm dev, %s",
 			 adev->asic_reset_res, adev->ddev->unique);
 }

@ -2448,6 +2471,7 @@ int amdgpu_device_init(struct amdgpu_device *adev,
 	mutex_init(&adev->virt.vf_errors.lock);
 	hash_init(adev->mn_hash);
 	mutex_init(&adev->lock_reset);
+	mutex_init(&adev->virt.dpm_mutex);

 	amdgpu_device_check_arguments(adev);

@ -2642,13 +2666,6 @@ fence_driver_init:
 	/* Get a log2 for easy divisions. */
 	adev->mm_stats.log2_max_MBps = ilog2(max(1u, max_MBps));

-	r = amdgpu_ib_pool_init(adev);
-	if (r) {
-		dev_err(adev->dev, "IB initialization failed (%d).\n", r);
-		amdgpu_vf_error_put(adev, AMDGIM_ERROR_VF_IB_INIT_FAIL, 0, r);
-		goto failed;
-	}
-
 	amdgpu_fbdev_init(adev);

 	r = amdgpu_pm_sysfs_init(adev);
@ -2694,6 +2711,9 @@ fence_driver_init:
 		goto failed;
 	}

+	/* must succeed. */
+	amdgpu_ras_post_init(adev);
+
 	return 0;

 failed:
@ -2726,7 +2746,6 @@ void amdgpu_device_fini(struct amdgpu_device *adev)
 		else
 			drm_atomic_helper_shutdown(adev->ddev);
 	}
-	amdgpu_ib_pool_fini(adev);
 	amdgpu_fence_driver_fini(adev);
 	amdgpu_pm_sysfs_fini(adev);
 	amdgpu_fbdev_fini(adev);
@ -3225,6 +3244,8 @@ static int amdgpu_device_reset_sriov(struct amdgpu_device *adev,
 	if (r)
 		return r;

+	amdgpu_amdkfd_pre_reset(adev);
+
 	/* Resume IP prior to SMC */
 	r = amdgpu_device_ip_reinit_early_sriov(adev);
 	if (r)
@ -3244,6 +3265,7 @@ static int amdgpu_device_reset_sriov(struct amdgpu_device *adev,

 	amdgpu_irq_gpu_reset_resume_helper(adev);
 	r = amdgpu_ib_ring_tests(adev);
+	amdgpu_amdkfd_post_reset(adev);

 error:
 	amdgpu_virt_init_data_exchange(adev);
@ -3376,7 +3398,7 @@ static int amdgpu_do_asic_reset(struct amdgpu_hive_info *hive,
 				r = amdgpu_asic_reset(tmp_adev);

 			if (r) {
-				DRM_ERROR("ASIC reset failed with err r, %d for drm dev, %s",
+				DRM_ERROR("ASIC reset failed with error, %d for drm dev, %s",
 					 r, tmp_adev->ddev->unique);
 				break;
 			}
@ -3393,6 +3415,11 @@ static int amdgpu_do_asic_reset(struct amdgpu_hive_info *hive,
 						break;
 				}
 			}
+
+			list_for_each_entry(tmp_adev, device_list_handle,
+					gmc.xgmi.head) {
+				amdgpu_ras_reserve_bad_pages(tmp_adev);
+			}
 		}
 	}

@ -3411,7 +3438,7 @@ static int amdgpu_do_asic_reset(struct amdgpu_hive_info *hive,

 				vram_lost = amdgpu_device_check_vram_lost(tmp_adev);
 				if (vram_lost) {
-					DRM_ERROR("VRAM is lost!\n");
+					DRM_INFO("VRAM is lost due to GPU reset!\n");
 					atomic_inc(&tmp_adev->vram_lost_counter);
 				}

--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.c
@ -904,3 +904,19 @@ amdgpu_get_vce_clock_state(void *handle, u32 idx)

 	return NULL;
 }
+
+int amdgpu_dpm_get_sclk(struct amdgpu_device *adev, bool low)
+{
+	if (is_support_sw_smu(adev))
+		return smu_get_sclk(&adev->smu, low);
+	else
+		return (adev)->powerplay.pp_funcs->get_sclk((adev)->powerplay.pp_handle, (low));
+}
+
+int amdgpu_dpm_get_mclk(struct amdgpu_device *adev, bool low)
+{
+	if (is_support_sw_smu(adev))
+		return smu_get_mclk(&adev->smu, low);
+	else
+		return (adev)->powerplay.pp_funcs->get_mclk((adev)->powerplay.pp_handle, (low));
+}
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h
@ -260,9 +260,6 @@ enum amdgpu_pcie_gen {
 #define amdgpu_dpm_enable_bapm(adev, e) \
 		((adev)->powerplay.pp_funcs->enable_bapm((adev)->powerplay.pp_handle, (e)))

-#define amdgpu_dpm_read_sensor(adev, idx, value, size) \
-		((adev)->powerplay.pp_funcs->read_sensor((adev)->powerplay.pp_handle, (idx), (value), (size)))
-
 #define amdgpu_dpm_set_fan_control_mode(adev, m) \
 		((adev)->powerplay.pp_funcs->set_fan_control_mode((adev)->powerplay.pp_handle, (m)))

@ -281,18 +278,18 @@ enum amdgpu_pcie_gen {
 #define amdgpu_dpm_set_fan_speed_rpm(adev, s) \
 		((adev)->powerplay.pp_funcs->set_fan_speed_rpm)((adev)->powerplay.pp_handle, (s))

-#define amdgpu_dpm_get_sclk(adev, l) \
-		((adev)->powerplay.pp_funcs->get_sclk((adev)->powerplay.pp_handle, (l)))
-
-#define amdgpu_dpm_get_mclk(adev, l)  \
-		((adev)->powerplay.pp_funcs->get_mclk((adev)->powerplay.pp_handle, (l)))
-
 #define amdgpu_dpm_force_performance_level(adev, l) \
 		((adev)->powerplay.pp_funcs->force_performance_level((adev)->powerplay.pp_handle, (l)))

 #define amdgpu_dpm_get_current_power_state(adev) \
 		((adev)->powerplay.pp_funcs->get_current_power_state((adev)->powerplay.pp_handle))

+#define amdgpu_smu_get_current_power_state(adev) \
+		((adev)->smu.ppt_funcs->get_current_power_state(&((adev)->smu)))
+
+#define amdgpu_smu_set_power_state(adev) \
+		((adev)->smu.ppt_funcs->set_power_state(&((adev)->smu)))
+
 #define amdgpu_dpm_get_pp_num_states(adev, data) \
 		((adev)->powerplay.pp_funcs->get_pp_num_states((adev)->powerplay.pp_handle, data))

@ -448,6 +445,9 @@ struct amdgpu_pm {
 	uint32_t                smu_prv_buffer_size;
 	struct amdgpu_bo        *smu_prv_buffer;
 	bool ac_power;
+	/* powerplay feature */
+	uint32_t pp_feature;
+
 };

 #define R600_SSTU_DFLT                               0
@ -486,6 +486,8 @@ void amdgpu_dpm_print_ps_status(struct amdgpu_device *adev,
 u32 amdgpu_dpm_get_vblank_time(struct amdgpu_device *adev);
 u32 amdgpu_dpm_get_vrefresh(struct amdgpu_device *adev);
 void amdgpu_dpm_get_active_displays(struct amdgpu_device *adev);
+int amdgpu_dpm_read_sensor(struct amdgpu_device *adev, enum amd_pp_sensors sensor,
+			   void *data, uint32_t *size);

 bool amdgpu_is_internal_thermal_sensor(enum amdgpu_int_thermal_type sensor);

@ -504,4 +506,8 @@ enum amdgpu_pcie_gen amdgpu_get_pcie_gen_support(struct amdgpu_device *adev,
 struct amd_vce_state*
 amdgpu_get_vce_clock_state(void *handle, u32 idx);

+extern int amdgpu_dpm_get_sclk(struct amdgpu_device *adev, bool low);
+
+extern int amdgpu_dpm_get_mclk(struct amdgpu_device *adev, bool low);
+
 #endif
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
@ -74,9 +74,11 @@
 * - 3.28.0 - Add AMDGPU_CHUNK_ID_SCHEDULED_DEPENDENCIES
 * - 3.29.0 - Add AMDGPU_IB_FLAG_RESET_GDS_MAX_WAVE_ID
 * - 3.30.0 - Add AMDGPU_SCHED_OP_CONTEXT_PRIORITY_OVERRIDE.
+ * - 3.31.0 - Add support for per-flip tiling attribute changes with DC
+ * - 3.32.0 - Add syncobj timeline support to AMDGPU_CS.
 */
 #define KMS_DRIVER_MAJOR	3
-#define KMS_DRIVER_MINOR	30
+#define KMS_DRIVER_MINOR	32
 #define KMS_DRIVER_PATCHLEVEL	0

 int amdgpu_vram_limit = 0;
@ -117,8 +119,8 @@ uint amdgpu_pg_mask = 0xffffffff;
 uint amdgpu_sdma_phase_quantum = 32;
 char *amdgpu_disable_cu = NULL;
 char *amdgpu_virtual_display = NULL;
-/* OverDrive(bit 14),gfxoff(bit 15),stutter mode(bit 17) disabled by default*/
-uint amdgpu_pp_feature_mask = 0xfffd3fff;
+/* OverDrive(bit 14) disabled by default*/
+uint amdgpu_pp_feature_mask = 0xffffbfff;
 int amdgpu_ngg = 0;
 int amdgpu_prim_buf_per_se = 0;
 int amdgpu_pos_buf_per_se = 0;
@ -136,6 +138,8 @@ uint amdgpu_dc_feature_mask = 0;
 struct amdgpu_mgpu_info mgpu_info = {
 	.mutex = __MUTEX_INITIALIZER(mgpu_info.mutex),
 };
+int amdgpu_ras_enable = -1;
+uint amdgpu_ras_mask = 0xffffffff;

 /**
 * DOC: vramlimit (int)
@ -494,6 +498,21 @@ module_param_named(gpu_recovery, amdgpu_gpu_recovery, int, 0444);
 MODULE_PARM_DESC(emu_mode, "Emulation mode, (1 = enable, 0 = disable)");
 module_param_named(emu_mode, amdgpu_emu_mode, int, 0444);

+/**
+ * DOC: ras_enable (int)
+ * Enable RAS features on the GPU (0 = disable, 1 = enable, -1 = auto (default))
+ */
+MODULE_PARM_DESC(ras_enable, "Enable RAS features on the GPU (0 = disable, 1 = enable, -1 = auto (default))");
+module_param_named(ras_enable, amdgpu_ras_enable, int, 0444);
+
+/**
+ * DOC: ras_mask (uint)
+ * Mask of RAS features to enable (default 0xffffffff), only valid when ras_enable == 1
+ * See the flags in drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h
+ */
+MODULE_PARM_DESC(ras_mask, "Mask of RAS features to enable (default 0xffffffff), only valid when ras_enable == 1");
+module_param_named(ras_mask, amdgpu_ras_mask, uint, 0444);
+
 /**
 * DOC: si_support (int)
 * Set SI support driver. This parameter works after set config CONFIG_DRM_AMDGPU_SI. For SI asic, when radeon driver is enabled,
@ -974,6 +993,7 @@ amdgpu_pci_remove(struct pci_dev *pdev)

 	DRM_ERROR("Device removal is currently not supported outside of fbcon\n");
 	drm_dev_unplug(dev);
+	drm_dev_put(dev);
 	pci_disable_device(pdev);
 	pci_set_drvdata(pdev, NULL);
 }
@ -1158,13 +1178,14 @@ static int amdgpu_flush(struct file *f, fl_owner_t id)
 {
 	struct drm_file *file_priv = f->private_data;
 	struct amdgpu_fpriv *fpriv = file_priv->driver_priv;
+	long timeout = MAX_WAIT_SCHED_ENTITY_Q_EMPTY;

-	amdgpu_ctx_mgr_entity_flush(&fpriv->ctx_mgr);
+	timeout = amdgpu_ctx_mgr_entity_flush(&fpriv->ctx_mgr, timeout);
+	timeout = amdgpu_vm_wait_idle(&fpriv->vm, timeout);

-	return 0;
+	return timeout >= 0 ? 0 : timeout;
 }

-
 static const struct file_operations amdgpu_driver_kms_fops = {
 	.owner = THIS_MODULE,
 	.open = drm_open,
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c
@ -49,12 +49,11 @@
 static int
 amdgpufb_open(struct fb_info *info, int user)
 {
-	struct amdgpu_fbdev *rfbdev = info->par;
-	struct amdgpu_device *adev = rfbdev->adev;
-	int ret = pm_runtime_get_sync(adev->ddev->dev);
+	struct drm_fb_helper *fb_helper = info->par;
+	int ret = pm_runtime_get_sync(fb_helper->dev->dev);
 	if (ret < 0 && ret != -EACCES) {
-		pm_runtime_mark_last_busy(adev->ddev->dev);
-		pm_runtime_put_autosuspend(adev->ddev->dev);
+		pm_runtime_mark_last_busy(fb_helper->dev->dev);
+		pm_runtime_put_autosuspend(fb_helper->dev->dev);
 		return ret;
 	}
 	return 0;
@ -63,11 +62,10 @@ amdgpufb_open(struct fb_info *info, int user)
 static int
 amdgpufb_release(struct fb_info *info, int user)
 {
-	struct amdgpu_fbdev *rfbdev = info->par;
-	struct amdgpu_device *adev = rfbdev->adev;
+	struct drm_fb_helper *fb_helper = info->par;

-	pm_runtime_mark_last_busy(adev->ddev->dev);
-	pm_runtime_put_autosuspend(adev->ddev->dev);
+	pm_runtime_mark_last_busy(fb_helper->dev->dev);
+	pm_runtime_put_autosuspend(fb_helper->dev->dev);
 	return 0;
 }

@ -233,9 +231,6 @@ static int amdgpufb_create(struct drm_fb_helper *helper,
 		goto out;
 	}

-	info->par = rfbdev;
-	info->skip_vt_switch = true;
-
 	ret = amdgpu_display_framebuffer_init(adev->ddev, &rfbdev->rfb,
 					      &mode_cmd, gobj);
 	if (ret) {
@ -248,10 +243,6 @@ static int amdgpufb_create(struct drm_fb_helper *helper,
 	/* setup helper */
 	rfbdev->helper.fb = fb;

-	strcpy(info->fix.id, "amdgpudrmfb");
-
-	drm_fb_helper_fill_fix(info, fb->pitches[0], fb->format->depth);
-
 	info->fbops = &amdgpufb_ops;

 	tmp = amdgpu_bo_gpu_offset(abo) - adev->gmc.vram_start;
@ -260,7 +251,7 @@ static int amdgpufb_create(struct drm_fb_helper *helper,
 	info->screen_base = amdgpu_bo_kptr(abo);
 	info->screen_size = amdgpu_bo_size(abo);

-	drm_fb_helper_fill_var(info, &rfbdev->helper, sizes->fb_width, sizes->fb_height);
+	drm_fb_helper_fill_info(info, &rfbdev->helper, sizes);

 	/* setup aperture base/size for vesafb takeover */
 	info->apertures->ranges[0].base = adev->ddev->mode_config.fb_base;
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c
@ -136,8 +136,9 @@ int amdgpu_fence_emit(struct amdgpu_ring *ring, struct dma_fence **f,
 {
 	struct amdgpu_device *adev = ring->adev;
 	struct amdgpu_fence *fence;
-	struct dma_fence *old, **ptr;
+	struct dma_fence __rcu **ptr;
 	uint32_t seq;
+	int r;

 	fence = kmem_cache_alloc(amdgpu_fence_slab, GFP_KERNEL);
 	if (fence == NULL)
@ -153,15 +154,24 @@ int amdgpu_fence_emit(struct amdgpu_ring *ring, struct dma_fence **f,
 			       seq, flags | AMDGPU_FENCE_FLAG_INT);

 	ptr = &ring->fence_drv.fences[seq & ring->fence_drv.num_fences_mask];
+	if (unlikely(rcu_dereference_protected(*ptr, 1))) {
+		struct dma_fence *old;
+
+		rcu_read_lock();
+		old = dma_fence_get_rcu_safe(ptr);
+		rcu_read_unlock();
+
+		if (old) {
+			r = dma_fence_wait(old, false);
+			dma_fence_put(old);
+			if (r)
+				return r;
+		}
+	}
+
 	/* This function can't be called concurrently anyway, otherwise
 	 * emitting the fence would mess up the hardware ring buffer.
 	 */
-	old = rcu_dereference_protected(*ptr, 1);
-	if (old && !dma_fence_is_signaled(old)) {
-		DRM_INFO("rcu slot is busy\n");
-		dma_fence_wait(old, false);
-	}
-
 	rcu_assign_pointer(*ptr, dma_fence_get(&fence->base));

 	*f = &fence->base;
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
@ -31,6 +31,7 @@
 #include <drm/amdgpu_drm.h>
 #include "amdgpu.h"
 #include "amdgpu_display.h"
+#include "amdgpu_xgmi.h"

 void amdgpu_gem_object_free(struct drm_gem_object *gobj)
 {
@ -627,11 +628,6 @@ int amdgpu_gem_va_ioctl(struct drm_device *dev, void *data,

 	switch (args->operation) {
 	case AMDGPU_VA_OP_MAP:
-		r = amdgpu_vm_alloc_pts(adev, bo_va->base.vm, args->va_address,
-					args->map_size);
-		if (r)
-			goto error_backoff;
-
 		va_flags = amdgpu_gmc_get_pte_flags(adev, args->flags);
 		r = amdgpu_vm_bo_map(adev, bo_va, args->va_address,
 				     args->offset_in_bo, args->map_size,
@ -647,11 +643,6 @@ int amdgpu_gem_va_ioctl(struct drm_device *dev, void *data,
 						args->map_size);
 		break;
 	case AMDGPU_VA_OP_REPLACE:
-		r = amdgpu_vm_alloc_pts(adev, bo_va->base.vm, args->va_address,
-					args->map_size);
-		if (r)
-			goto error_backoff;
-
 		va_flags = amdgpu_gmc_get_pte_flags(adev, args->flags);
 		r = amdgpu_vm_bo_replace_map(adev, bo_va, args->va_address,
 					     args->offset_in_bo, args->map_size,
@ -678,6 +669,7 @@ int amdgpu_gem_op_ioctl(struct drm_device *dev, void *data,
 	struct amdgpu_device *adev = dev->dev_private;
 	struct drm_amdgpu_gem_op *args = data;
 	struct drm_gem_object *gobj;
+	struct amdgpu_vm_bo_base *base;
 	struct amdgpu_bo *robj;
 	int r;

@ -716,6 +708,15 @@ int amdgpu_gem_op_ioctl(struct drm_device *dev, void *data,
 			amdgpu_bo_unreserve(robj);
 			break;
 		}
+		for (base = robj->vm_bo; base; base = base->next)
+			if (amdgpu_xgmi_same_hive(amdgpu_ttm_adev(robj->tbo.bdev),
+				amdgpu_ttm_adev(base->vm->root.base.bo->tbo.bdev))) {
+				r = -EINVAL;
+				amdgpu_bo_unreserve(robj);
+				goto out;
+			}
+
+
 		robj->preferred_domains = args->value & (AMDGPU_GEM_DOMAIN_VRAM |
 							AMDGPU_GEM_DOMAIN_GTT |
 							AMDGPU_GEM_DOMAIN_CPU);
@ -745,17 +746,25 @@ int amdgpu_mode_dumb_create(struct drm_file *file_priv,
 	struct amdgpu_device *adev = dev->dev_private;
 	struct drm_gem_object *gobj;
 	uint32_t handle;
+	u64 flags = AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED;
 	u32 domain;
 	int r;

+	/*
+	 * The buffer returned from this function should be cleared, but
+	 * it can only be done if the ring is enabled or we'll fail to
+	 * create the buffer.
+	 */
+	if (adev->mman.buffer_funcs_enabled)
+		flags |= AMDGPU_GEM_CREATE_VRAM_CLEARED;
+
 	args->pitch = amdgpu_align_pitch(adev, args->width,
 					 DIV_ROUND_UP(args->bpp, 8), 0);
 	args->size = (u64)args->pitch * args->height;
 	args->size = ALIGN(args->size, PAGE_SIZE);
 	domain = amdgpu_bo_get_preferred_pin_domain(adev,
 				amdgpu_display_supported_domains(adev));
-	r = amdgpu_gem_object_create(adev, args->size, 0, domain,
-				     AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED,
+	r = amdgpu_gem_object_create(adev, args->size, 0, domain, flags,
 				     ttm_bo_type_device, NULL, &gobj);
 	if (r)
 		return -ENOMEM;
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c
@ -390,7 +390,7 @@ void amdgpu_gfx_compute_mqd_sw_fini(struct amdgpu_device *adev)

 void amdgpu_gfx_off_ctrl(struct amdgpu_device *adev, bool enable)
 {
-	if (!(adev->powerplay.pp_feature & PP_GFXOFF_MASK))
+	if (!(adev->pm.pp_feature & PP_GFXOFF_MASK))
 		return;

 	if (!adev->powerplay.pp_funcs || !adev->powerplay.pp_funcs->set_powergating_by_smu)
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h
@ -258,6 +258,9 @@ struct amdgpu_gfx {
 	/* pipe reservation */
 	struct mutex			pipe_reserve_mutex;
 	DECLARE_BITMAP			(pipe_reserve_bitmap, AMDGPU_MAX_COMPUTE_QUEUES);
+
+	/*ras */
+	struct ras_common_if		*ras_if;
 };

 #define amdgpu_gfx_get_gpu_clock_counter(adev) (adev)->gfx.funcs->get_gpu_clock_counter((adev))
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c
@ -79,6 +79,33 @@ uint64_t amdgpu_gmc_pd_addr(struct amdgpu_bo *bo)
 	return pd_addr;
 }

+/**
+ * amdgpu_gmc_set_pte_pde - update the page tables using CPU
+ *
+ * @adev: amdgpu_device pointer
+ * @cpu_pt_addr: cpu address of the page table
+ * @gpu_page_idx: entry in the page table to update
+ * @addr: dst addr to write into pte/pde
+ * @flags: access flags
+ *
+ * Update the page tables using CPU.
+ */
+int amdgpu_gmc_set_pte_pde(struct amdgpu_device *adev, void *cpu_pt_addr,
+				uint32_t gpu_page_idx, uint64_t addr,
+				uint64_t flags)
+{
+	void __iomem *ptr = (void *)cpu_pt_addr;
+	uint64_t value;
+
+	/*
+	 * The following is for PTE only. GART does not have PDEs.
+	*/
+	value = addr & 0x0000FFFFFFFFF000ULL;
+	value |= flags;
+	writeq(value, ptr + (gpu_page_idx * 8));
+	return 0;
+}
+
 /**
 * amdgpu_gmc_agp_addr - return the address in the AGP address space
 *
@ -213,3 +240,58 @@ void amdgpu_gmc_agp_location(struct amdgpu_device *adev, struct amdgpu_gmc *mc)
 	dev_info(adev->dev, "AGP: %lluM 0x%016llX - 0x%016llX\n",
 			mc->agp_size >> 20, mc->agp_start, mc->agp_end);
 }
+
+/**
+ * amdgpu_gmc_filter_faults - filter VM faults
+ *
+ * @adev: amdgpu device structure
+ * @addr: address of the VM fault
+ * @pasid: PASID of the process causing the fault
+ * @timestamp: timestamp of the fault
+ *
+ * Returns:
+ * True if the fault was filtered and should not be processed further.
+ * False if the fault is a new one and needs to be handled.
+ */
+bool amdgpu_gmc_filter_faults(struct amdgpu_device *adev, uint64_t addr,
+			      uint16_t pasid, uint64_t timestamp)
+{
+	struct amdgpu_gmc *gmc = &adev->gmc;
+
+	uint64_t stamp, key = addr << 4 | pasid;
+	struct amdgpu_gmc_fault *fault;
+	uint32_t hash;
+
+	/* If we don't have space left in the ring buffer return immediately */
+	stamp = max(timestamp, AMDGPU_GMC_FAULT_TIMEOUT + 1) -
+		AMDGPU_GMC_FAULT_TIMEOUT;
+	if (gmc->fault_ring[gmc->last_fault].timestamp >= stamp)
+		return true;
+
+	/* Try to find the fault in the hash */
+	hash = hash_64(key, AMDGPU_GMC_FAULT_HASH_ORDER);
+	fault = &gmc->fault_ring[gmc->fault_hash[hash].idx];
+	while (fault->timestamp >= stamp) {
+		uint64_t tmp;
+
+		if (fault->key == key)
+			return true;
+
+		tmp = fault->timestamp;
+		fault = &gmc->fault_ring[fault->next];
+
+		/* Check if the entry was reused */
+		if (fault->timestamp >= tmp)
+			break;
+	}
+
+	/* Add the fault to the ring */
+	fault = &gmc->fault_ring[gmc->last_fault];
+	fault->key = key;
+	fault->timestamp = timestamp;
+
+	/* And update the hash */
+	fault->next = gmc->fault_hash[hash].idx;
+	gmc->fault_hash[hash].idx = gmc->last_fault++;
+	return false;
+}
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h
@ -43,8 +43,34 @@
 */
 #define AMDGPU_GMC_HOLE_MASK	0x0000ffffffffffffULL

+/*
+ * Ring size as power of two for the log of recent faults.
+ */
+#define AMDGPU_GMC_FAULT_RING_ORDER	8
+#define AMDGPU_GMC_FAULT_RING_SIZE	(1 << AMDGPU_GMC_FAULT_RING_ORDER)
+
+/*
+ * Hash size as power of two for the log of recent faults
+ */
+#define AMDGPU_GMC_FAULT_HASH_ORDER	8
+#define AMDGPU_GMC_FAULT_HASH_SIZE	(1 << AMDGPU_GMC_FAULT_HASH_ORDER)
+
+/*
+ * Number of IH timestamp ticks until a fault is considered handled
+ */
+#define AMDGPU_GMC_FAULT_TIMEOUT	5000ULL
+
 struct firmware;

+/*
+ * GMC page fault information
+ */
+struct amdgpu_gmc_fault {
+	uint64_t	timestamp;
+	uint64_t	next:AMDGPU_GMC_FAULT_RING_ORDER;
+	uint64_t	key:52;
+};
+
 /*
 * VMHUB structures, functions & helpers
 */
@ -71,12 +97,6 @@ struct amdgpu_gmc_funcs {
 	/* Change the VMID -> PASID mapping */
 	void (*emit_pasid_mapping)(struct amdgpu_ring *ring, unsigned vmid,
 				   unsigned pasid);
-	/* write pte/pde updates using the cpu */
-	int (*set_pte_pde)(struct amdgpu_device *adev,
-			   void *cpu_pt_addr, /* cpu addr of page table */
-			   uint32_t gpu_page_idx, /* pte/pde to update */
-			   uint64_t addr, /* addr to write into pte/pde */
-			   uint64_t flags); /* access flags */
 	/* enable/disable PRT support */
 	void (*set_prt)(struct amdgpu_device *adev, bool enable);
 	/* set pte flags based per asic */
@ -147,15 +167,22 @@ struct amdgpu_gmc {
 	struct kfd_vm_fault_info *vm_fault_info;
 	atomic_t		vm_fault_info_updated;

+	struct amdgpu_gmc_fault	fault_ring[AMDGPU_GMC_FAULT_RING_SIZE];
+	struct {
+		uint64_t	idx:AMDGPU_GMC_FAULT_RING_ORDER;
+	} fault_hash[AMDGPU_GMC_FAULT_HASH_SIZE];
+	uint64_t		last_fault:AMDGPU_GMC_FAULT_RING_ORDER;
+
 	const struct amdgpu_gmc_funcs	*gmc_funcs;

 	struct amdgpu_xgmi xgmi;
+	struct amdgpu_irq_src	ecc_irq;
+	struct ras_common_if    *ras_if;
 };

 #define amdgpu_gmc_flush_gpu_tlb(adev, vmid, type) (adev)->gmc.gmc_funcs->flush_gpu_tlb((adev), (vmid), (type))
 #define amdgpu_gmc_emit_flush_gpu_tlb(r, vmid, addr) (r)->adev->gmc.gmc_funcs->emit_flush_gpu_tlb((r), (vmid), (addr))
 #define amdgpu_gmc_emit_pasid_mapping(r, vmid, pasid) (r)->adev->gmc.gmc_funcs->emit_pasid_mapping((r), (vmid), (pasid))
-#define amdgpu_gmc_set_pte_pde(adev, pt, idx, addr, flags) (adev)->gmc.gmc_funcs->set_pte_pde((adev), (pt), (idx), (addr), (flags))
 #define amdgpu_gmc_get_vm_pde(adev, level, dst, flags) (adev)->gmc.gmc_funcs->get_vm_pde((adev), (level), (dst), (flags))
 #define amdgpu_gmc_get_pte_flags(adev, flags) (adev)->gmc.gmc_funcs->get_vm_pte_flags((adev),(flags))

@ -189,6 +216,9 @@ static inline uint64_t amdgpu_gmc_sign_extend(uint64_t addr)

 void amdgpu_gmc_get_pde_for_bo(struct amdgpu_bo *bo, int level,
 			       uint64_t *addr, uint64_t *flags);
+int amdgpu_gmc_set_pte_pde(struct amdgpu_device *adev, void *cpu_pt_addr,
+				uint32_t gpu_page_idx, uint64_t addr,
+				uint64_t flags);
 uint64_t amdgpu_gmc_pd_addr(struct amdgpu_bo *bo);
 uint64_t amdgpu_gmc_agp_addr(struct ttm_buffer_object *bo);
 void amdgpu_gmc_vram_location(struct amdgpu_device *adev, struct amdgpu_gmc *mc,
@ -197,5 +227,7 @@ void amdgpu_gmc_gart_location(struct amdgpu_device *adev,
 			      struct amdgpu_gmc *mc);
 void amdgpu_gmc_agp_location(struct amdgpu_device *adev,
 			     struct amdgpu_gmc *mc);
+bool amdgpu_gmc_filter_faults(struct amdgpu_device *adev, uint64_t addr,
+			      uint16_t pasid, uint64_t timestamp);

 #endif
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gtt_mgr.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gtt_mgr.c
@ -36,6 +36,47 @@ struct amdgpu_gtt_node {
 	struct ttm_buffer_object *tbo;
 };

+/**
+ * DOC: mem_info_gtt_total
+ *
+ * The amdgpu driver provides a sysfs API for reporting current total size of
+ * the GTT.
+ * The file mem_info_gtt_total is used for this, and returns the total size of
+ * the GTT block, in bytes
+ */
+static ssize_t amdgpu_mem_info_gtt_total_show(struct device *dev,
+		struct device_attribute *attr, char *buf)
+{
+	struct drm_device *ddev = dev_get_drvdata(dev);
+	struct amdgpu_device *adev = ddev->dev_private;
+
+	return snprintf(buf, PAGE_SIZE, "%llu\n",
+			(adev->mman.bdev.man[TTM_PL_TT].size) * PAGE_SIZE);
+}
+
+/**
+ * DOC: mem_info_gtt_used
+ *
+ * The amdgpu driver provides a sysfs API for reporting current total amount of
+ * used GTT.
+ * The file mem_info_gtt_used is used for this, and returns the current used
+ * size of the GTT block, in bytes
+ */
+static ssize_t amdgpu_mem_info_gtt_used_show(struct device *dev,
+		struct device_attribute *attr, char *buf)
+{
+	struct drm_device *ddev = dev_get_drvdata(dev);
+	struct amdgpu_device *adev = ddev->dev_private;
+
+	return snprintf(buf, PAGE_SIZE, "%llu\n",
+			amdgpu_gtt_mgr_usage(&adev->mman.bdev.man[TTM_PL_TT]));
+}
+
+static DEVICE_ATTR(mem_info_gtt_total, S_IRUGO,
+	           amdgpu_mem_info_gtt_total_show, NULL);
+static DEVICE_ATTR(mem_info_gtt_used, S_IRUGO,
+	           amdgpu_mem_info_gtt_used_show, NULL);
+
 /**
 * amdgpu_gtt_mgr_init - init GTT manager and DRM MM
 *
@ -50,6 +91,7 @@ static int amdgpu_gtt_mgr_init(struct ttm_mem_type_manager *man,
 	struct amdgpu_device *adev = amdgpu_ttm_adev(man->bdev);
 	struct amdgpu_gtt_mgr *mgr;
 	uint64_t start, size;
+	int ret;

 	mgr = kzalloc(sizeof(*mgr), GFP_KERNEL);
 	if (!mgr)
@ -61,6 +103,18 @@ static int amdgpu_gtt_mgr_init(struct ttm_mem_type_manager *man,
 	spin_lock_init(&mgr->lock);
 	atomic64_set(&mgr->available, p_size);
 	man->priv = mgr;
+
+	ret = device_create_file(adev->dev, &dev_attr_mem_info_gtt_total);
+	if (ret) {
+		DRM_ERROR("Failed to create device file mem_info_gtt_total\n");
+		return ret;
+	}
+	ret = device_create_file(adev->dev, &dev_attr_mem_info_gtt_used);
+	if (ret) {
+		DRM_ERROR("Failed to create device file mem_info_gtt_used\n");
+		return ret;
+	}
+
 	return 0;
 }

@ -74,12 +128,17 @@ static int amdgpu_gtt_mgr_init(struct ttm_mem_type_manager *man,
 */
 static int amdgpu_gtt_mgr_fini(struct ttm_mem_type_manager *man)
 {
+	struct amdgpu_device *adev = amdgpu_ttm_adev(man->bdev);
 	struct amdgpu_gtt_mgr *mgr = man->priv;
 	spin_lock(&mgr->lock);
 	drm_mm_takedown(&mgr->mm);
 	spin_unlock(&mgr->lock);
 	kfree(mgr);
 	man->priv = NULL;
+
+	device_remove_file(adev->dev, &dev_attr_mem_info_gtt_total);
+	device_remove_file(adev->dev, &dev_attr_mem_info_gtt_used);
+
 	return 0;
 }

--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ih.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ih.c
@ -142,6 +142,7 @@ void amdgpu_ih_ring_fini(struct amdgpu_device *adev, struct amdgpu_ih_ring *ih)
 */
 int amdgpu_ih_process(struct amdgpu_device *adev, struct amdgpu_ih_ring *ih)
 {
+	unsigned int count = AMDGPU_IH_MAX_NUM_IVS;
 	u32 wptr;

 	if (!ih->enabled || adev->shutdown)
@ -159,7 +160,7 @@ restart_ih:
 	/* Order reading of wptr vs. reading of IH ring data */
 	rmb();

-	while (ih->rptr != wptr) {
+	while (ih->rptr != wptr && --count) {
 		amdgpu_irq_dispatch(adev, ih);
 		ih->rptr &= ih->ptr_mask;
 	}
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ih.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ih.h
@ -24,6 +24,9 @@
 #ifndef __AMDGPU_IH_H__
 #define __AMDGPU_IH_H__

+/* Maximum number of IVs processed at once */
+#define AMDGPU_IH_MAX_NUM_IVS	32
+
 struct amdgpu_device;
 struct amdgpu_iv_entry;

--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
@ -39,6 +39,7 @@
 #include "amdgpu_amdkfd.h"
 #include "amdgpu_gem.h"
 #include "amdgpu_display.h"
+#include "amdgpu_ras.h"

 static void amdgpu_unregister_gpu_instance(struct amdgpu_device *adev)
 {
@ -296,6 +297,17 @@ static int amdgpu_firmware_info(struct drm_amdgpu_info_firmware *fw_info,
 		fw_info->ver = adev->pm.fw_version;
 		fw_info->feature = 0;
 		break;
+	case AMDGPU_INFO_FW_TA:
+		if (query_fw->index > 1)
+			return -EINVAL;
+		if (query_fw->index == 0) {
+			fw_info->ver = adev->psp.ta_fw_version;
+			fw_info->feature = adev->psp.ta_xgmi_ucode_version;
+		} else {
+			fw_info->ver = adev->psp.ta_fw_version;
+			fw_info->feature = adev->psp.ta_ras_ucode_version;
+		}
+		break;
 	case AMDGPU_INFO_FW_SDMA:
 		if (query_fw->index >= adev->sdma.num_instances)
 			return -EINVAL;
@ -684,6 +696,10 @@ static int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file
 		if (adev->pm.dpm_enabled) {
 			dev_info.max_engine_clock = amdgpu_dpm_get_sclk(adev, false) * 10;
 			dev_info.max_memory_clock = amdgpu_dpm_get_mclk(adev, false) * 10;
+		} else if (amdgpu_sriov_vf(adev) && amdgim_is_hwperf(adev) &&
+			   adev->virt.ops->get_pp_clk) {
+			dev_info.max_engine_clock = amdgpu_virt_get_sclk(adev, false) * 10;
+			dev_info.max_memory_clock = amdgpu_virt_get_mclk(adev, false) * 10;
 		} else {
 			dev_info.max_engine_clock = adev->clock.default_sclk * 10;
 			dev_info.max_memory_clock = adev->clock.default_mclk * 10;
@ -909,6 +925,18 @@ static int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file
 	case AMDGPU_INFO_VRAM_LOST_COUNTER:
 		ui32 = atomic_read(&adev->vram_lost_counter);
 		return copy_to_user(out, &ui32, min(size, 4u)) ? -EFAULT : 0;
+	case AMDGPU_INFO_RAS_ENABLED_FEATURES: {
+		struct amdgpu_ras *ras = amdgpu_ras_get_context(adev);
+		uint64_t ras_mask;
+
+		if (!ras)
+			return -EINVAL;
+		ras_mask = (uint64_t)ras->supported << 32 | ras->features;
+
+		return copy_to_user(out, &ras_mask,
+				min_t(u64, size, sizeof(ras_mask))) ?
+			-EFAULT : 0;
+	}
 	default:
 		DRM_DEBUG_KMS("Invalid request %d\n", info->query);
 		return -EINVAL;
@ -1328,6 +1356,16 @@ static int amdgpu_debugfs_firmware_info(struct seq_file *m, void *data)
 	seq_printf(m, "ASD feature version: %u, firmware version: 0x%08x\n",
 		   fw_info.feature, fw_info.ver);

+	query_fw.fw_type = AMDGPU_INFO_FW_TA;
+	for (i = 0; i < 2; i++) {
+		query_fw.index = i;
+		ret = amdgpu_firmware_info(&fw_info, &query_fw, adev);
+		if (ret)
+			continue;
+		seq_printf(m, "TA %s feature version: %u, firmware version: 0x%08x\n",
+				i ? "RAS" : "XGMI", fw_info.feature, fw_info.ver);
+	}
+
 	/* SMC */
 	query_fw.fw_type = AMDGPU_INFO_FW_SMC;
 	ret = amdgpu_firmware_info(&fw_info, &query_fw, adev);
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h
@ -58,7 +58,7 @@ struct amdgpu_hpd;
 #define to_amdgpu_encoder(x) container_of(x, struct amdgpu_encoder, base)
 #define to_amdgpu_framebuffer(x) container_of(x, struct amdgpu_framebuffer, base)

-#define to_dm_plane_state(x)	container_of(x, struct dm_plane_state, base);
+#define to_dm_plane_state(x)	container_of(x, struct dm_plane_state, base)

 #define AMDGPU_MAX_HPD_PINS 6
 #define AMDGPU_MAX_CRTCS 6
@ -406,7 +406,7 @@ struct amdgpu_crtc {
 	struct amdgpu_flip_work *pflip_works;
 	enum amdgpu_flip_status pflip_status;
 	int deferred_flip_completion;
-	u64 last_flip_vblank;
+	u32 last_flip_vblank;
 	/* pll sharing */
 	struct amdgpu_atom_ss ss;
 	bool ss_enabled;
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
@ -88,12 +88,14 @@ static void amdgpu_bo_destroy(struct ttm_buffer_object *tbo)
 	if (bo->gem_base.import_attach)
 		drm_prime_gem_destroy(&bo->gem_base, bo->tbo.sg);
 	drm_gem_object_release(&bo->gem_base);
-	amdgpu_bo_unref(&bo->parent);
+	/* in case amdgpu_device_recover_vram got NULL of bo->parent */
 	if (!list_empty(&bo->shadow_list)) {
 		mutex_lock(&adev->shadow_list_lock);
 		list_del_init(&bo->shadow_list);
 		mutex_unlock(&adev->shadow_list_lock);
 	}
+	amdgpu_bo_unref(&bo->parent);
+
 	kfree(bo->metadata);
 	kfree(bo);
 }
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h
@ -72,6 +72,8 @@ struct amdgpu_bo_va {

 	/* If the mappings are cleared or filled */
 	bool				cleared;
+
+	bool				is_xgmi;
 };

 struct amdgpu_bo {
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
@ -28,6 +28,7 @@
 #include "amdgpu_pm.h"
 #include "amdgpu_dpm.h"
 #include "amdgpu_display.h"
+#include "amdgpu_smu.h"
 #include "atom.h"
 #include <linux/power_supply.h>
 #include <linux/hwmon.h>
@ -80,6 +81,27 @@ void amdgpu_pm_acpi_event_handler(struct amdgpu_device *adev)
 	}
 }

+int amdgpu_dpm_read_sensor(struct amdgpu_device *adev, enum amd_pp_sensors sensor,
+			   void *data, uint32_t *size)
+{
+	int ret = 0;
+
+	if (!data || !size)
+		return -EINVAL;
+
+	if (is_support_sw_smu(adev))
+		ret = smu_read_sensor(&adev->smu, sensor, data, size);
+	else {
+		if (adev->powerplay.pp_funcs && adev->powerplay.pp_funcs->read_sensor)
+			ret = adev->powerplay.pp_funcs->read_sensor((adev)->powerplay.pp_handle,
+								    sensor, data, size);
+		else
+			ret = -EINVAL;
+	}
+
+	return ret;
+}
+
 /**
 * DOC: power_dpm_state
 *
@ -122,7 +144,9 @@ static ssize_t amdgpu_get_dpm_state(struct device *dev,
 	struct amdgpu_device *adev = ddev->dev_private;
 	enum amd_pm_state_type pm;

-	if (adev->powerplay.pp_funcs->get_current_power_state)
+	if (is_support_sw_smu(adev) && adev->smu.ppt_funcs->get_current_power_state)
+		pm = amdgpu_smu_get_current_power_state(adev);
+	else if (adev->powerplay.pp_funcs->get_current_power_state)
 		pm = amdgpu_dpm_get_current_power_state(adev);
 	else
 		pm = adev->pm.dpm.user_state;
@ -240,7 +264,9 @@ static ssize_t amdgpu_get_dpm_forced_performance_level(struct device *dev,
 	     (ddev->switch_power_state != DRM_SWITCH_POWER_ON))
 		return snprintf(buf, PAGE_SIZE, "off\n");

-	if (adev->powerplay.pp_funcs->get_performance_level)
+	if (is_support_sw_smu(adev))
+		level = smu_get_performance_level(&adev->smu);
+	else if (adev->powerplay.pp_funcs->get_performance_level)
 		level = amdgpu_dpm_get_performance_level(adev);
 	else
 		level = adev->pm.dpm.forced_level;
@ -273,7 +299,9 @@ static ssize_t amdgpu_set_dpm_forced_performance_level(struct device *dev,
 	     (ddev->switch_power_state != DRM_SWITCH_POWER_ON))
 		return -EINVAL;

-	if (adev->powerplay.pp_funcs->get_performance_level)
+	if (is_support_sw_smu(adev))
+		current_level = smu_get_performance_level(&adev->smu);
+	else if (adev->powerplay.pp_funcs->get_performance_level)
 		current_level = amdgpu_dpm_get_performance_level(adev);

 	if (strncmp("low", buf, strlen("low")) == 0) {
@ -299,10 +327,35 @@ static ssize_t amdgpu_set_dpm_forced_performance_level(struct device *dev,
 		goto fail;
 	}

+        if (amdgpu_sriov_vf(adev)) {
+                if (amdgim_is_hwperf(adev) &&
+                    adev->virt.ops->force_dpm_level) {
+                        mutex_lock(&adev->pm.mutex);
+                        adev->virt.ops->force_dpm_level(adev, level);
+                        mutex_unlock(&adev->pm.mutex);
+                        return count;
+                } else {
+                        return -EINVAL;
+		}
+        }
+
 	if (current_level == level)
 		return count;

-	if (adev->powerplay.pp_funcs->force_performance_level) {
+	if (is_support_sw_smu(adev)) {
+		mutex_lock(&adev->pm.mutex);
+		if (adev->pm.dpm.thermal_active) {
+			count = -EINVAL;
+			mutex_unlock(&adev->pm.mutex);
+			goto fail;
+		}
+		ret = smu_force_performance_level(&adev->smu, level);
+		if (ret)
+			count = -EINVAL;
+		else
+			adev->pm.dpm.forced_level = level;
+		mutex_unlock(&adev->pm.mutex);
+	} else if (adev->powerplay.pp_funcs->force_performance_level) {
 		mutex_lock(&adev->pm.mutex);
 		if (adev->pm.dpm.thermal_active) {
 			count = -EINVAL;
@ -328,9 +381,13 @@ static ssize_t amdgpu_get_pp_num_states(struct device *dev,
 	struct drm_device *ddev = dev_get_drvdata(dev);
 	struct amdgpu_device *adev = ddev->dev_private;
 	struct pp_states_info data;
-	int i, buf_len;
+	int i, buf_len, ret;

-	if (adev->powerplay.pp_funcs->get_pp_num_states)
+	if (is_support_sw_smu(adev)) {
+		ret = smu_get_power_num_states(&adev->smu, &data);
+		if (ret)
+			return ret;
+	} else if (adev->powerplay.pp_funcs->get_pp_num_states)
 		amdgpu_dpm_get_pp_num_states(adev, &data);

 	buf_len = snprintf(buf, PAGE_SIZE, "states: %d\n", data.nums);
@ -351,23 +408,29 @@ static ssize_t amdgpu_get_pp_cur_state(struct device *dev,
 	struct drm_device *ddev = dev_get_drvdata(dev);
 	struct amdgpu_device *adev = ddev->dev_private;
 	struct pp_states_info data;
+	struct smu_context *smu = &adev->smu;
 	enum amd_pm_state_type pm = 0;
-	int i = 0;
+	int i = 0, ret = 0;

-	if (adev->powerplay.pp_funcs->get_current_power_state
+	if (is_support_sw_smu(adev)) {
+		pm = smu_get_current_power_state(smu);
+		ret = smu_get_power_num_states(smu, &data);
+		if (ret)
+			return ret;
+	} else if (adev->powerplay.pp_funcs->get_current_power_state
 		 && adev->powerplay.pp_funcs->get_pp_num_states) {
 		pm = amdgpu_dpm_get_current_power_state(adev);
 		amdgpu_dpm_get_pp_num_states(adev, &data);
-
-		for (i = 0; i < data.nums; i++) {
-			if (pm == data.states[i])
-				break;
-		}
-
-		if (i == data.nums)
-			i = -EINVAL;
 	}

+	for (i = 0; i < data.nums; i++) {
+		if (pm == data.states[i])
+			break;
+	}
+
+	if (i == data.nums)
+		i = -EINVAL;
+
 	return snprintf(buf, PAGE_SIZE, "%d\n", i);
 }

@ -397,6 +460,8 @@ static ssize_t amdgpu_set_pp_force_state(struct device *dev,

 	if (strlen(buf) == 1)
 		adev->pp_force_state_enabled = false;
+	else if (is_support_sw_smu(adev))
+		adev->pp_force_state_enabled = false;
 	else if (adev->powerplay.pp_funcs->dispatch_tasks &&
 			adev->powerplay.pp_funcs->get_pp_num_states) {
 		struct pp_states_info data;
@ -442,7 +507,12 @@ static ssize_t amdgpu_get_pp_table(struct device *dev,
 	char *table = NULL;
 	int size;

-	if (adev->powerplay.pp_funcs->get_pp_table)
+	if (is_support_sw_smu(adev)) {
+		size = smu_sys_get_pp_table(&adev->smu, (void **)&table);
+		if (size < 0)
+			return size;
+	}
+	else if (adev->powerplay.pp_funcs->get_pp_table)
 		size = amdgpu_dpm_get_pp_table(adev, &table);
 	else
 		return 0;
@ -462,8 +532,13 @@ static ssize_t amdgpu_set_pp_table(struct device *dev,
 {
 	struct drm_device *ddev = dev_get_drvdata(dev);
 	struct amdgpu_device *adev = ddev->dev_private;
+	int ret = 0;

-	if (adev->powerplay.pp_funcs->set_pp_table)
+	if (is_support_sw_smu(adev)) {
+		ret = smu_sys_set_pp_table(&adev->smu, (void *)buf, count);
+		if (ret)
+			return ret;
+	} else if (adev->powerplay.pp_funcs->set_pp_table)
 		amdgpu_dpm_set_pp_table(adev, buf, count);

 	return count;
@ -586,19 +661,29 @@ static ssize_t amdgpu_set_pp_od_clk_voltage(struct device *dev,
 			tmp_str++;
 	}

-	if (adev->powerplay.pp_funcs->odn_edit_dpm_table)
-		ret = amdgpu_dpm_odn_edit_dpm_table(adev, type,
+	if (is_support_sw_smu(adev)) {
+		ret = smu_od_edit_dpm_table(&adev->smu, type,
+					    parameter, parameter_size);
+
+		if (ret)
+			return -EINVAL;
+	} else {
+		if (adev->powerplay.pp_funcs->odn_edit_dpm_table)
+			ret = amdgpu_dpm_odn_edit_dpm_table(adev, type,
 						parameter, parameter_size);

-	if (ret)
-		return -EINVAL;
-
-	if (type == PP_OD_COMMIT_DPM_TABLE) {
-		if (adev->powerplay.pp_funcs->dispatch_tasks) {
-			amdgpu_dpm_dispatch_task(adev, AMD_PP_TASK_READJUST_POWER_STATE, NULL);
-			return count;
-		} else {
+		if (ret)
 			return -EINVAL;
+
+		if (type == PP_OD_COMMIT_DPM_TABLE) {
+			if (adev->powerplay.pp_funcs->dispatch_tasks) {
+				amdgpu_dpm_dispatch_task(adev,
+						AMD_PP_TASK_READJUST_POWER_STATE,
+						NULL);
+				return count;
+			} else {
+				return -EINVAL;
+			}
 		}
 	}

@ -613,7 +698,13 @@ static ssize_t amdgpu_get_pp_od_clk_voltage(struct device *dev,
 	struct amdgpu_device *adev = ddev->dev_private;
 	uint32_t size = 0;

-	if (adev->powerplay.pp_funcs->print_clock_levels) {
+	if (is_support_sw_smu(adev)) {
+		size = smu_print_clk_levels(&adev->smu, OD_SCLK, buf);
+		size += smu_print_clk_levels(&adev->smu, OD_MCLK, buf+size);
+		size += smu_print_clk_levels(&adev->smu, OD_VDDC_CURVE, buf+size);
+		size += smu_print_clk_levels(&adev->smu, OD_RANGE, buf+size);
+		return size;
+	} else if (adev->powerplay.pp_funcs->print_clock_levels) {
 		size = amdgpu_dpm_print_clock_levels(adev, OD_SCLK, buf);
 		size += amdgpu_dpm_print_clock_levels(adev, OD_MCLK, buf+size);
 		size += amdgpu_dpm_print_clock_levels(adev, OD_VDDC_CURVE, buf+size);
@ -711,7 +802,13 @@ static ssize_t amdgpu_get_pp_dpm_sclk(struct device *dev,
 	struct drm_device *ddev = dev_get_drvdata(dev);
 	struct amdgpu_device *adev = ddev->dev_private;

-	if (adev->powerplay.pp_funcs->print_clock_levels)
+	if (amdgpu_sriov_vf(adev) && amdgim_is_hwperf(adev) &&
+	    adev->virt.ops->get_pp_clk)
+		return adev->virt.ops->get_pp_clk(adev, PP_SCLK, buf);
+
+	if (is_support_sw_smu(adev))
+		return smu_print_clk_levels(&adev->smu, PP_SCLK, buf);
+	else if (adev->powerplay.pp_funcs->print_clock_levels)
 		return amdgpu_dpm_print_clock_levels(adev, PP_SCLK, buf);
 	else
 		return snprintf(buf, PAGE_SIZE, "\n");
@ -767,7 +864,9 @@ static ssize_t amdgpu_set_pp_dpm_sclk(struct device *dev,
 	if (ret)
 		return ret;

-	if (adev->powerplay.pp_funcs->force_clock_level)
+	if (is_support_sw_smu(adev))
+		ret = smu_force_clk_levels(&adev->smu, PP_SCLK, mask);
+	else if (adev->powerplay.pp_funcs->force_clock_level)
 		ret = amdgpu_dpm_force_clock_level(adev, PP_SCLK, mask);

 	if (ret)
@ -783,7 +882,9 @@ static ssize_t amdgpu_get_pp_dpm_mclk(struct device *dev,
 	struct drm_device *ddev = dev_get_drvdata(dev);
 	struct amdgpu_device *adev = ddev->dev_private;

-	if (adev->powerplay.pp_funcs->print_clock_levels)
+	if (is_support_sw_smu(adev))
+		return smu_print_clk_levels(&adev->smu, PP_MCLK, buf);
+	else if (adev->powerplay.pp_funcs->print_clock_levels)
 		return amdgpu_dpm_print_clock_levels(adev, PP_MCLK, buf);
 	else
 		return snprintf(buf, PAGE_SIZE, "\n");
@ -803,7 +904,9 @@ static ssize_t amdgpu_set_pp_dpm_mclk(struct device *dev,
 	if (ret)
 		return ret;

-	if (adev->powerplay.pp_funcs->force_clock_level)
+	if (is_support_sw_smu(adev))
+		ret = smu_force_clk_levels(&adev->smu, PP_MCLK, mask);
+	else if (adev->powerplay.pp_funcs->force_clock_level)
 		ret = amdgpu_dpm_force_clock_level(adev, PP_MCLK, mask);

 	if (ret)
@ -819,7 +922,9 @@ static ssize_t amdgpu_get_pp_dpm_socclk(struct device *dev,
 	struct drm_device *ddev = dev_get_drvdata(dev);
 	struct amdgpu_device *adev = ddev->dev_private;

-	if (adev->powerplay.pp_funcs->print_clock_levels)
+	if (is_support_sw_smu(adev))
+		return smu_print_clk_levels(&adev->smu, PP_SOCCLK, buf);
+	else if (adev->powerplay.pp_funcs->print_clock_levels)
 		return amdgpu_dpm_print_clock_levels(adev, PP_SOCCLK, buf);
 	else
 		return snprintf(buf, PAGE_SIZE, "\n");
@ -839,7 +944,9 @@ static ssize_t amdgpu_set_pp_dpm_socclk(struct device *dev,
 	if (ret)
 		return ret;

-	if (adev->powerplay.pp_funcs->force_clock_level)
+	if (is_support_sw_smu(adev))
+		ret = smu_force_clk_levels(&adev->smu, PP_SOCCLK, mask);
+	else if (adev->powerplay.pp_funcs->force_clock_level)
 		ret = amdgpu_dpm_force_clock_level(adev, PP_SOCCLK, mask);

 	if (ret)
@ -855,7 +962,9 @@ static ssize_t amdgpu_get_pp_dpm_fclk(struct device *dev,
 	struct drm_device *ddev = dev_get_drvdata(dev);
 	struct amdgpu_device *adev = ddev->dev_private;

-	if (adev->powerplay.pp_funcs->print_clock_levels)
+	if (is_support_sw_smu(adev))
+		return smu_print_clk_levels(&adev->smu, PP_FCLK, buf);
+	else if (adev->powerplay.pp_funcs->print_clock_levels)
 		return amdgpu_dpm_print_clock_levels(adev, PP_FCLK, buf);
 	else
 		return snprintf(buf, PAGE_SIZE, "\n");
@ -875,7 +984,9 @@ static ssize_t amdgpu_set_pp_dpm_fclk(struct device *dev,
 	if (ret)
 		return ret;

-	if (adev->powerplay.pp_funcs->force_clock_level)
+	if (is_support_sw_smu(adev))
+		ret = smu_force_clk_levels(&adev->smu, PP_FCLK, mask);
+	else if (adev->powerplay.pp_funcs->force_clock_level)
 		ret = amdgpu_dpm_force_clock_level(adev, PP_FCLK, mask);

 	if (ret)
@ -891,7 +1002,9 @@ static ssize_t amdgpu_get_pp_dpm_dcefclk(struct device *dev,
 	struct drm_device *ddev = dev_get_drvdata(dev);
 	struct amdgpu_device *adev = ddev->dev_private;

-	if (adev->powerplay.pp_funcs->print_clock_levels)
+	if (is_support_sw_smu(adev))
+		return smu_print_clk_levels(&adev->smu, PP_DCEFCLK, buf);
+	else if (adev->powerplay.pp_funcs->print_clock_levels)
 		return amdgpu_dpm_print_clock_levels(adev, PP_DCEFCLK, buf);
 	else
 		return snprintf(buf, PAGE_SIZE, "\n");
@ -911,7 +1024,9 @@ static ssize_t amdgpu_set_pp_dpm_dcefclk(struct device *dev,
 	if (ret)
 		return ret;

-	if (adev->powerplay.pp_funcs->force_clock_level)
+	if (is_support_sw_smu(adev))
+		ret = smu_force_clk_levels(&adev->smu, PP_DCEFCLK, mask);
+	else if (adev->powerplay.pp_funcs->force_clock_level)
 		ret = amdgpu_dpm_force_clock_level(adev, PP_DCEFCLK, mask);

 	if (ret)
@ -927,7 +1042,9 @@ static ssize_t amdgpu_get_pp_dpm_pcie(struct device *dev,
 	struct drm_device *ddev = dev_get_drvdata(dev);
 	struct amdgpu_device *adev = ddev->dev_private;

-	if (adev->powerplay.pp_funcs->print_clock_levels)
+	if (is_support_sw_smu(adev))
+		return smu_print_clk_levels(&adev->smu, PP_PCIE, buf);
+	else if (adev->powerplay.pp_funcs->print_clock_levels)
 		return amdgpu_dpm_print_clock_levels(adev, PP_PCIE, buf);
 	else
 		return snprintf(buf, PAGE_SIZE, "\n");
@ -947,7 +1064,9 @@ static ssize_t amdgpu_set_pp_dpm_pcie(struct device *dev,
 	if (ret)
 		return ret;

-	if (adev->powerplay.pp_funcs->force_clock_level)
+	if (is_support_sw_smu(adev))
+		ret = smu_force_clk_levels(&adev->smu, PP_PCIE, mask);
+	else if (adev->powerplay.pp_funcs->force_clock_level)
 		ret = amdgpu_dpm_force_clock_level(adev, PP_PCIE, mask);

 	if (ret)
@ -964,7 +1083,9 @@ static ssize_t amdgpu_get_pp_sclk_od(struct device *dev,
 	struct amdgpu_device *adev = ddev->dev_private;
 	uint32_t value = 0;

-	if (adev->powerplay.pp_funcs->get_sclk_od)
+	if (is_support_sw_smu(adev))
+		value = smu_get_od_percentage(&(adev->smu), OD_SCLK);
+	else if (adev->powerplay.pp_funcs->get_sclk_od)
 		value = amdgpu_dpm_get_sclk_od(adev);

 	return snprintf(buf, PAGE_SIZE, "%d\n", value);
@ -986,14 +1107,19 @@ static ssize_t amdgpu_set_pp_sclk_od(struct device *dev,
 		count = -EINVAL;
 		goto fail;
 	}
-	if (adev->powerplay.pp_funcs->set_sclk_od)
-		amdgpu_dpm_set_sclk_od(adev, (uint32_t)value);

-	if (adev->powerplay.pp_funcs->dispatch_tasks) {
-		amdgpu_dpm_dispatch_task(adev, AMD_PP_TASK_READJUST_POWER_STATE, NULL);
+	if (is_support_sw_smu(adev)) {
+		value = smu_set_od_percentage(&(adev->smu), OD_SCLK, (uint32_t)value);
 	} else {
-		adev->pm.dpm.current_ps = adev->pm.dpm.boot_ps;
-		amdgpu_pm_compute_clocks(adev);
+		if (adev->powerplay.pp_funcs->set_sclk_od)
+			amdgpu_dpm_set_sclk_od(adev, (uint32_t)value);
+
+		if (adev->powerplay.pp_funcs->dispatch_tasks) {
+			amdgpu_dpm_dispatch_task(adev, AMD_PP_TASK_READJUST_POWER_STATE, NULL);
+		} else {
+			adev->pm.dpm.current_ps = adev->pm.dpm.boot_ps;
+			amdgpu_pm_compute_clocks(adev);
+		}
 	}

 fail:
@ -1008,7 +1134,9 @@ static ssize_t amdgpu_get_pp_mclk_od(struct device *dev,
 	struct amdgpu_device *adev = ddev->dev_private;
 	uint32_t value = 0;

-	if (adev->powerplay.pp_funcs->get_mclk_od)
+	if (is_support_sw_smu(adev))
+		value = smu_get_od_percentage(&(adev->smu), OD_MCLK);
+	else if (adev->powerplay.pp_funcs->get_mclk_od)
 		value = amdgpu_dpm_get_mclk_od(adev);

 	return snprintf(buf, PAGE_SIZE, "%d\n", value);
@ -1030,14 +1158,19 @@ static ssize_t amdgpu_set_pp_mclk_od(struct device *dev,
 		count = -EINVAL;
 		goto fail;
 	}
-	if (adev->powerplay.pp_funcs->set_mclk_od)
-		amdgpu_dpm_set_mclk_od(adev, (uint32_t)value);

-	if (adev->powerplay.pp_funcs->dispatch_tasks) {
-		amdgpu_dpm_dispatch_task(adev, AMD_PP_TASK_READJUST_POWER_STATE, NULL);
+	if (is_support_sw_smu(adev)) {
+		value = smu_set_od_percentage(&(adev->smu), OD_MCLK, (uint32_t)value);
 	} else {
-		adev->pm.dpm.current_ps = adev->pm.dpm.boot_ps;
-		amdgpu_pm_compute_clocks(adev);
+		if (adev->powerplay.pp_funcs->set_mclk_od)
+			amdgpu_dpm_set_mclk_od(adev, (uint32_t)value);
+
+		if (adev->powerplay.pp_funcs->dispatch_tasks) {
+			amdgpu_dpm_dispatch_task(adev, AMD_PP_TASK_READJUST_POWER_STATE, NULL);
+		} else {
+			adev->pm.dpm.current_ps = adev->pm.dpm.boot_ps;
+			amdgpu_pm_compute_clocks(adev);
+		}
 	}

 fail:
@ -1071,7 +1204,9 @@ static ssize_t amdgpu_get_pp_power_profile_mode(struct device *dev,
 	struct drm_device *ddev = dev_get_drvdata(dev);
 	struct amdgpu_device *adev = ddev->dev_private;

-	if (adev->powerplay.pp_funcs->get_power_profile_mode)
+	if (is_support_sw_smu(adev))
+		return smu_get_power_profile_mode(&adev->smu, buf);
+	else if (adev->powerplay.pp_funcs->get_power_profile_mode)
 		return amdgpu_dpm_get_power_profile_mode(adev, buf);

 	return snprintf(buf, PAGE_SIZE, "\n");
@ -1121,9 +1256,10 @@ static ssize_t amdgpu_set_pp_power_profile_mode(struct device *dev,
 		}
 	}
 	parameter[parameter_size] = profile_mode;
-	if (adev->powerplay.pp_funcs->set_power_profile_mode)
+	if (is_support_sw_smu(adev))
+		ret = smu_set_power_profile_mode(&adev->smu, parameter, parameter_size);
+	else if (adev->powerplay.pp_funcs->set_power_profile_mode)
 		ret = amdgpu_dpm_set_power_profile_mode(adev, parameter, parameter_size);
-
 	if (!ret)
 		return count;
 fail:
@ -1146,14 +1282,10 @@ static ssize_t amdgpu_get_busy_percent(struct device *dev,
 	struct amdgpu_device *adev = ddev->dev_private;
 	int r, value, size = sizeof(value);

-	/* sanity check PP is enabled */
-	if (!(adev->powerplay.pp_funcs &&
-	      adev->powerplay.pp_funcs->read_sensor))
-		return -EINVAL;
-
 	/* read the IP busy sensor */
 	r = amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_GPU_LOAD,
 				   (void *)&value, &size);
+
 	if (r)
 		return r;

@ -1247,11 +1379,6 @@ static ssize_t amdgpu_hwmon_show_temp(struct device *dev,
 	     (ddev->switch_power_state != DRM_SWITCH_POWER_ON))
 		return -EINVAL;

-	/* sanity check PP is enabled */
-	if (!(adev->powerplay.pp_funcs &&
-	      adev->powerplay.pp_funcs->read_sensor))
-		return -EINVAL;
-
 	/* get the temperature */
 	r = amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_GPU_TEMP,
 				   (void *)&temp, &size);
@ -1283,11 +1410,14 @@ static ssize_t amdgpu_hwmon_get_pwm1_enable(struct device *dev,
 {
 	struct amdgpu_device *adev = dev_get_drvdata(dev);
 	u32 pwm_mode = 0;
+	if (is_support_sw_smu(adev)) {
+		pwm_mode = smu_get_fan_control_mode(&adev->smu);
+	} else {
+		if (!adev->powerplay.pp_funcs->get_fan_control_mode)
+			return -EINVAL;

-	if (!adev->powerplay.pp_funcs->get_fan_control_mode)
-		return -EINVAL;
-
-	pwm_mode = amdgpu_dpm_get_fan_control_mode(adev);
+		pwm_mode = amdgpu_dpm_get_fan_control_mode(adev);
+	}

 	return sprintf(buf, "%i\n", pwm_mode);
 }
@ -1306,14 +1436,22 @@ static ssize_t amdgpu_hwmon_set_pwm1_enable(struct device *dev,
 	     (adev->ddev->switch_power_state != DRM_SWITCH_POWER_ON))
 		return -EINVAL;

-	if (!adev->powerplay.pp_funcs->set_fan_control_mode)
-		return -EINVAL;
+	if (is_support_sw_smu(adev)) {
+		err = kstrtoint(buf, 10, &value);
+		if (err)
+			return err;

-	err = kstrtoint(buf, 10, &value);
-	if (err)
-		return err;
+		smu_set_fan_control_mode(&adev->smu, value);
+	} else {
+		if (!adev->powerplay.pp_funcs->set_fan_control_mode)
+			return -EINVAL;

-	amdgpu_dpm_set_fan_control_mode(adev, value);
+		err = kstrtoint(buf, 10, &value);
+		if (err)
+			return err;
+
+		amdgpu_dpm_set_fan_control_mode(adev, value);
+	}

 	return count;
 }
@ -1345,8 +1483,10 @@ static ssize_t amdgpu_hwmon_set_pwm1(struct device *dev,
 	if  ((adev->flags & AMD_IS_PX) &&
 	     (adev->ddev->switch_power_state != DRM_SWITCH_POWER_ON))
 		return -EINVAL;
-
-	pwm_mode = amdgpu_dpm_get_fan_control_mode(adev);
+	if (is_support_sw_smu(adev))
+		pwm_mode = smu_get_fan_control_mode(&adev->smu);
+	else
+		pwm_mode = amdgpu_dpm_get_fan_control_mode(adev);
 	if (pwm_mode != AMD_FAN_CTRL_MANUAL) {
 		pr_info("manual fan speed control should be enabled first\n");
 		return -EINVAL;
@ -1358,7 +1498,11 @@ static ssize_t amdgpu_hwmon_set_pwm1(struct device *dev,

 	value = (value * 100) / 255;

-	if (adev->powerplay.pp_funcs->set_fan_speed_percent) {
+	if (is_support_sw_smu(adev)) {
+		err = smu_set_fan_speed_percent(&adev->smu, value);
+		if (err)
+			return err;
+	} else if (adev->powerplay.pp_funcs->set_fan_speed_percent) {
 		err = amdgpu_dpm_set_fan_speed_percent(adev, value);
 		if (err)
 			return err;
@ -1380,7 +1524,11 @@ static ssize_t amdgpu_hwmon_get_pwm1(struct device *dev,
 	     (adev->ddev->switch_power_state != DRM_SWITCH_POWER_ON))
 		return -EINVAL;

-	if (adev->powerplay.pp_funcs->get_fan_speed_percent) {
+	if (is_support_sw_smu(adev)) {
+		err = smu_get_fan_speed_percent(&adev->smu, &speed);
+		if (err)
+			return err;
+	} else if (adev->powerplay.pp_funcs->get_fan_speed_percent) {
 		err = amdgpu_dpm_get_fan_speed_percent(adev, &speed);
 		if (err)
 			return err;
@ -1404,7 +1552,11 @@ static ssize_t amdgpu_hwmon_get_fan1_input(struct device *dev,
 	     (adev->ddev->switch_power_state != DRM_SWITCH_POWER_ON))
 		return -EINVAL;

-	if (adev->powerplay.pp_funcs->get_fan_speed_rpm) {
+	if (is_support_sw_smu(adev)) {
+		err = smu_get_current_rpm(&adev->smu, &speed);
+		if (err)
+			return err;
+	} else if (adev->powerplay.pp_funcs->get_fan_speed_rpm) {
 		err = amdgpu_dpm_get_fan_speed_rpm(adev, &speed);
 		if (err)
 			return err;
@ -1422,9 +1574,6 @@ static ssize_t amdgpu_hwmon_get_fan1_min(struct device *dev,
 	u32 size = sizeof(min_rpm);
 	int r;

-	if (!adev->powerplay.pp_funcs->read_sensor)
-		return -EINVAL;
-
 	r = amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_MIN_FAN_RPM,
 				   (void *)&min_rpm, &size);
 	if (r)
@ -1442,9 +1591,6 @@ static ssize_t amdgpu_hwmon_get_fan1_max(struct device *dev,
 	u32 size = sizeof(max_rpm);
 	int r;

-	if (!adev->powerplay.pp_funcs->read_sensor)
-		return -EINVAL;
-
 	r = amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_MAX_FAN_RPM,
 				   (void *)&max_rpm, &size);
 	if (r)
@ -1466,7 +1612,11 @@ static ssize_t amdgpu_hwmon_get_fan1_target(struct device *dev,
 	     (adev->ddev->switch_power_state != DRM_SWITCH_POWER_ON))
 		return -EINVAL;

-	if (adev->powerplay.pp_funcs->get_fan_speed_rpm) {
+	if (is_support_sw_smu(adev)) {
+		err = smu_get_current_rpm(&adev->smu, &rpm);
+		if (err)
+			return err;
+	} else if (adev->powerplay.pp_funcs->get_fan_speed_rpm) {
 		err = amdgpu_dpm_get_fan_speed_rpm(adev, &rpm);
 		if (err)
 			return err;
@ -1484,7 +1634,11 @@ static ssize_t amdgpu_hwmon_set_fan1_target(struct device *dev,
 	u32 value;
 	u32 pwm_mode;

-	pwm_mode = amdgpu_dpm_get_fan_control_mode(adev);
+	if (is_support_sw_smu(adev))
+		pwm_mode = smu_get_fan_control_mode(&adev->smu);
+	else
+		pwm_mode = amdgpu_dpm_get_fan_control_mode(adev);
+
 	if (pwm_mode != AMD_FAN_CTRL_MANUAL)
 		return -ENODATA;

@ -1497,7 +1651,11 @@ static ssize_t amdgpu_hwmon_set_fan1_target(struct device *dev,
 	if (err)
 		return err;

-	if (adev->powerplay.pp_funcs->set_fan_speed_rpm) {
+	if (is_support_sw_smu(adev)) {
+		err = smu_set_fan_speed_rpm(&adev->smu, value);
+		if (err)
+			return err;
+	} else if (adev->powerplay.pp_funcs->set_fan_speed_rpm) {
 		err = amdgpu_dpm_set_fan_speed_rpm(adev, value);
 		if (err)
 			return err;
@ -1513,11 +1671,14 @@ static ssize_t amdgpu_hwmon_get_fan1_enable(struct device *dev,
 	struct amdgpu_device *adev = dev_get_drvdata(dev);
 	u32 pwm_mode = 0;

-	if (!adev->powerplay.pp_funcs->get_fan_control_mode)
-		return -EINVAL;
-
-	pwm_mode = amdgpu_dpm_get_fan_control_mode(adev);
+	if (is_support_sw_smu(adev)) {
+		pwm_mode = smu_get_fan_control_mode(&adev->smu);
+	} else {
+		if (!adev->powerplay.pp_funcs->get_fan_control_mode)
+			return -EINVAL;

+		pwm_mode = amdgpu_dpm_get_fan_control_mode(adev);
+	}
 	return sprintf(buf, "%i\n", pwm_mode == AMD_FAN_CTRL_AUTO ? 0 : 1);
 }

@ -1536,8 +1697,6 @@ static ssize_t amdgpu_hwmon_set_fan1_enable(struct device *dev,
 	     (adev->ddev->switch_power_state != DRM_SWITCH_POWER_ON))
 		return -EINVAL;

-	if (!adev->powerplay.pp_funcs->set_fan_control_mode)
-		return -EINVAL;

 	err = kstrtoint(buf, 10, &value);
 	if (err)
@ -1550,7 +1709,13 @@ static ssize_t amdgpu_hwmon_set_fan1_enable(struct device *dev,
 	else
 		return -EINVAL;

-	amdgpu_dpm_set_fan_control_mode(adev, pwm_mode);
+	if (is_support_sw_smu(adev)) {
+		smu_set_fan_control_mode(&adev->smu, pwm_mode);
+	} else {
+		if (!adev->powerplay.pp_funcs->set_fan_control_mode)
+			return -EINVAL;
+		amdgpu_dpm_set_fan_control_mode(adev, pwm_mode);
+	}

 	return count;
 }
@ -1569,11 +1734,6 @@ static ssize_t amdgpu_hwmon_show_vddgfx(struct device *dev,
 	     (ddev->switch_power_state != DRM_SWITCH_POWER_ON))
 		return -EINVAL;

-	/* sanity check PP is enabled */
-	if (!(adev->powerplay.pp_funcs &&
-	      adev->powerplay.pp_funcs->read_sensor))
-	      return -EINVAL;
-
 	/* get the voltage */
 	r = amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_VDDGFX,
 				   (void *)&vddgfx, &size);
@ -1608,11 +1768,6 @@ static ssize_t amdgpu_hwmon_show_vddnb(struct device *dev,
 	     (ddev->switch_power_state != DRM_SWITCH_POWER_ON))
 		return -EINVAL;

-	/* sanity check PP is enabled */
-	if (!(adev->powerplay.pp_funcs &&
-	      adev->powerplay.pp_funcs->read_sensor))
-	      return -EINVAL;
-
 	/* get the voltage */
 	r = amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_VDDNB,
 				   (void *)&vddnb, &size);
@ -1644,11 +1799,6 @@ static ssize_t amdgpu_hwmon_show_power_avg(struct device *dev,
 	     (ddev->switch_power_state != DRM_SWITCH_POWER_ON))
 		return -EINVAL;

-	/* sanity check PP is enabled */
-	if (!(adev->powerplay.pp_funcs &&
-	      adev->powerplay.pp_funcs->read_sensor))
-	      return -EINVAL;
-
 	/* get the voltage */
 	r = amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_GPU_POWER,
 				   (void *)&query, &size);
@ -1675,7 +1825,10 @@ static ssize_t amdgpu_hwmon_show_power_cap_max(struct device *dev,
 	struct amdgpu_device *adev = dev_get_drvdata(dev);
 	uint32_t limit = 0;

-	if (adev->powerplay.pp_funcs && adev->powerplay.pp_funcs->get_power_limit) {
+	if (is_support_sw_smu(adev)) {
+		smu_get_power_limit(&adev->smu, &limit, true);
+		return snprintf(buf, PAGE_SIZE, "%u\n", limit * 1000000);
+	} else if (adev->powerplay.pp_funcs && adev->powerplay.pp_funcs->get_power_limit) {
 		adev->powerplay.pp_funcs->get_power_limit(adev->powerplay.pp_handle, &limit, true);
 		return snprintf(buf, PAGE_SIZE, "%u\n", limit * 1000000);
 	} else {
@ -1690,7 +1843,10 @@ static ssize_t amdgpu_hwmon_show_power_cap(struct device *dev,
 	struct amdgpu_device *adev = dev_get_drvdata(dev);
 	uint32_t limit = 0;

-	if (adev->powerplay.pp_funcs && adev->powerplay.pp_funcs->get_power_limit) {
+	if (is_support_sw_smu(adev)) {
+		smu_get_power_limit(&adev->smu, &limit, false);
+		return snprintf(buf, PAGE_SIZE, "%u\n", limit * 1000000);
+	} else if (adev->powerplay.pp_funcs && adev->powerplay.pp_funcs->get_power_limit) {
 		adev->powerplay.pp_funcs->get_power_limit(adev->powerplay.pp_handle, &limit, false);
 		return snprintf(buf, PAGE_SIZE, "%u\n", limit * 1000000);
 	} else {
@ -1713,7 +1869,9 @@ static ssize_t amdgpu_hwmon_set_power_cap(struct device *dev,
 		return err;

 	value = value / 1000000; /* convert to Watt */
-	if (adev->powerplay.pp_funcs && adev->powerplay.pp_funcs->set_power_limit) {
+	if (is_support_sw_smu(adev)) {
+		adev->smu.funcs->set_power_limit(&adev->smu, value);
+	} else if (adev->powerplay.pp_funcs && adev->powerplay.pp_funcs->set_power_limit) {
 		err = adev->powerplay.pp_funcs->set_power_limit(adev->powerplay.pp_handle, value);
 		if (err)
 			return err;
@ -1967,18 +2125,20 @@ static umode_t hwmon_attributes_visible(struct kobject *kobj,
 	     attr == &sensor_dev_attr_fan1_enable.dev_attr.attr))
 		return 0;

-	/* mask fan attributes if we have no bindings for this asic to expose */
-	if ((!adev->powerplay.pp_funcs->get_fan_speed_percent &&
-	     attr == &sensor_dev_attr_pwm1.dev_attr.attr) || /* can't query fan */
-	    (!adev->powerplay.pp_funcs->get_fan_control_mode &&
-	     attr == &sensor_dev_attr_pwm1_enable.dev_attr.attr)) /* can't query state */
-		effective_mode &= ~S_IRUGO;
+	if (!is_support_sw_smu(adev)) {
+		/* mask fan attributes if we have no bindings for this asic to expose */
+		if ((!adev->powerplay.pp_funcs->get_fan_speed_percent &&
+		     attr == &sensor_dev_attr_pwm1.dev_attr.attr) || /* can't query fan */
+		    (!adev->powerplay.pp_funcs->get_fan_control_mode &&
+		     attr == &sensor_dev_attr_pwm1_enable.dev_attr.attr)) /* can't query state */
+			effective_mode &= ~S_IRUGO;

-	if ((!adev->powerplay.pp_funcs->set_fan_speed_percent &&
-	     attr == &sensor_dev_attr_pwm1.dev_attr.attr) || /* can't manage fan */
-	    (!adev->powerplay.pp_funcs->set_fan_control_mode &&
-	     attr == &sensor_dev_attr_pwm1_enable.dev_attr.attr)) /* can't manage state */
-		effective_mode &= ~S_IWUSR;
+		if ((!adev->powerplay.pp_funcs->set_fan_speed_percent &&
+		     attr == &sensor_dev_attr_pwm1.dev_attr.attr) || /* can't manage fan */
+		    (!adev->powerplay.pp_funcs->set_fan_control_mode &&
+		     attr == &sensor_dev_attr_pwm1_enable.dev_attr.attr)) /* can't manage state */
+			effective_mode &= ~S_IWUSR;
+	}

 	if ((adev->flags & AMD_IS_APU) &&
 	    (attr == &sensor_dev_attr_power1_average.dev_attr.attr ||
@ -1987,20 +2147,22 @@ static umode_t hwmon_attributes_visible(struct kobject *kobj,
 	     attr == &sensor_dev_attr_power1_cap.dev_attr.attr))
 		return 0;

-	/* hide max/min values if we can't both query and manage the fan */
-	if ((!adev->powerplay.pp_funcs->set_fan_speed_percent &&
-	     !adev->powerplay.pp_funcs->get_fan_speed_percent) &&
-	     (!adev->powerplay.pp_funcs->set_fan_speed_rpm &&
-	     !adev->powerplay.pp_funcs->get_fan_speed_rpm) &&
-	    (attr == &sensor_dev_attr_pwm1_max.dev_attr.attr ||
-	     attr == &sensor_dev_attr_pwm1_min.dev_attr.attr))
-		return 0;
+	if (!is_support_sw_smu(adev)) {
+		/* hide max/min values if we can't both query and manage the fan */
+		if ((!adev->powerplay.pp_funcs->set_fan_speed_percent &&
+		     !adev->powerplay.pp_funcs->get_fan_speed_percent) &&
+		     (!adev->powerplay.pp_funcs->set_fan_speed_rpm &&
+		     !adev->powerplay.pp_funcs->get_fan_speed_rpm) &&
+		    (attr == &sensor_dev_attr_pwm1_max.dev_attr.attr ||
+		     attr == &sensor_dev_attr_pwm1_min.dev_attr.attr))
+			return 0;

-	if ((!adev->powerplay.pp_funcs->set_fan_speed_rpm &&
-	     !adev->powerplay.pp_funcs->get_fan_speed_rpm) &&
-	    (attr == &sensor_dev_attr_fan1_max.dev_attr.attr ||
-	     attr == &sensor_dev_attr_fan1_min.dev_attr.attr))
-		return 0;
+		if ((!adev->powerplay.pp_funcs->set_fan_speed_rpm &&
+		     !adev->powerplay.pp_funcs->get_fan_speed_rpm) &&
+		    (attr == &sensor_dev_attr_fan1_max.dev_attr.attr ||
+		     attr == &sensor_dev_attr_fan1_min.dev_attr.attr))
+			return 0;
+	}

 	/* only APUs have vddnb */
 	if (!(adev->flags & AMD_IS_APU) &&
@ -2039,9 +2201,7 @@ void amdgpu_dpm_thermal_work_handler(struct work_struct *work)
 	if (!adev->pm.dpm_enabled)
 		return;

-	if (adev->powerplay.pp_funcs &&
-	    adev->powerplay.pp_funcs->read_sensor &&
-	    !amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_GPU_TEMP,
+	if (!amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_GPU_TEMP,
 				    (void *)&temp, &size)) {
 		if (temp < adev->pm.dpm.thermal.min_temp)
 			/* switch back the user state */
@ -2267,7 +2427,13 @@ static void amdgpu_dpm_change_power_state_locked(struct amdgpu_device *adev)

 void amdgpu_dpm_enable_uvd(struct amdgpu_device *adev, bool enable)
 {
-	if (adev->powerplay.pp_funcs->set_powergating_by_smu) {
+	int ret = 0;
+	if (is_support_sw_smu(adev)) {
+	    ret = smu_dpm_set_power_gate(&adev->smu, AMD_IP_BLOCK_TYPE_UVD, enable);
+	    if (ret)
+		DRM_ERROR("[SW SMU]: dpm enable uvd failed, state = %s, ret = %d. \n",
+			  enable ? "true" : "false", ret);
+	} else if (adev->powerplay.pp_funcs->set_powergating_by_smu) {
 		/* enable/disable UVD */
 		mutex_lock(&adev->pm.mutex);
 		amdgpu_dpm_set_powergating_by_smu(adev, AMD_IP_BLOCK_TYPE_UVD, !enable);
@ -2288,7 +2454,13 @@ void amdgpu_dpm_enable_uvd(struct amdgpu_device *adev, bool enable)

 void amdgpu_dpm_enable_vce(struct amdgpu_device *adev, bool enable)
 {
-	if (adev->powerplay.pp_funcs->set_powergating_by_smu) {
+	int ret = 0;
+	if (is_support_sw_smu(adev)) {
+	    ret = smu_dpm_set_power_gate(&adev->smu, AMD_IP_BLOCK_TYPE_VCE, enable);
+	    if (ret)
+		DRM_ERROR("[SW SMU]: dpm enable vce failed, state = %s, ret = %d. \n",
+			  enable ? "true" : "false", ret);
+	} else if (adev->powerplay.pp_funcs->set_powergating_by_smu) {
 		/* enable/disable VCE */
 		mutex_lock(&adev->pm.mutex);
 		amdgpu_dpm_set_powergating_by_smu(adev, AMD_IP_BLOCK_TYPE_VCE, !enable);
@ -2413,7 +2585,8 @@ int amdgpu_pm_sysfs_init(struct amdgpu_device *adev)
 				"pp_power_profile_mode\n");
 		return ret;
 	}
-	if (hwmgr->od_enabled) {
+	if ((is_support_sw_smu(adev) && adev->smu.od_enabled) ||
+	    (!is_support_sw_smu(adev) && hwmgr->od_enabled)) {
 		ret = device_create_file(adev->dev,
 				&dev_attr_pp_od_clk_voltage);
 		if (ret) {
@ -2489,7 +2662,8 @@ void amdgpu_pm_sysfs_fini(struct amdgpu_device *adev)
 	device_remove_file(adev->dev, &dev_attr_pp_mclk_od);
 	device_remove_file(adev->dev,
 			&dev_attr_pp_power_profile_mode);
-	if (hwmgr->od_enabled)
+	if ((is_support_sw_smu(adev) && adev->smu.od_enabled) ||
+	    (!is_support_sw_smu(adev) && hwmgr->od_enabled))
 		device_remove_file(adev->dev,
 				&dev_attr_pp_od_clk_voltage);
 	device_remove_file(adev->dev, &dev_attr_gpu_busy_percent);
@ -2516,28 +2690,38 @@ void amdgpu_pm_compute_clocks(struct amdgpu_device *adev)
 			amdgpu_fence_wait_empty(ring);
 	}

-	if (adev->powerplay.pp_funcs->dispatch_tasks) {
-		if (!amdgpu_device_has_dc_support(adev)) {
+	if (is_support_sw_smu(adev)) {
+		struct smu_context *smu = &adev->smu;
+		struct smu_dpm_context *smu_dpm = &adev->smu.smu_dpm;
+		mutex_lock(&(smu->mutex));
+		smu_handle_task(&adev->smu,
+				smu_dpm->dpm_level,
+				AMD_PP_TASK_DISPLAY_CONFIG_CHANGE);
+		mutex_unlock(&(smu->mutex));
+	} else {
+		if (adev->powerplay.pp_funcs->dispatch_tasks) {
+			if (!amdgpu_device_has_dc_support(adev)) {
+				mutex_lock(&adev->pm.mutex);
+				amdgpu_dpm_get_active_displays(adev);
+				adev->pm.pm_display_cfg.num_display = adev->pm.dpm.new_active_crtc_count;
+				adev->pm.pm_display_cfg.vrefresh = amdgpu_dpm_get_vrefresh(adev);
+				adev->pm.pm_display_cfg.min_vblank_time = amdgpu_dpm_get_vblank_time(adev);
+				/* we have issues with mclk switching with refresh rates over 120 hz on the non-DC code. */
+				if (adev->pm.pm_display_cfg.vrefresh > 120)
+					adev->pm.pm_display_cfg.min_vblank_time = 0;
+				if (adev->powerplay.pp_funcs->display_configuration_change)
+					adev->powerplay.pp_funcs->display_configuration_change(
+									adev->powerplay.pp_handle,
+									&adev->pm.pm_display_cfg);
+				mutex_unlock(&adev->pm.mutex);
+			}
+			amdgpu_dpm_dispatch_task(adev, AMD_PP_TASK_DISPLAY_CONFIG_CHANGE, NULL);
+		} else {
 			mutex_lock(&adev->pm.mutex);
 			amdgpu_dpm_get_active_displays(adev);
-			adev->pm.pm_display_cfg.num_display = adev->pm.dpm.new_active_crtc_count;
-			adev->pm.pm_display_cfg.vrefresh = amdgpu_dpm_get_vrefresh(adev);
-			adev->pm.pm_display_cfg.min_vblank_time = amdgpu_dpm_get_vblank_time(adev);
-			/* we have issues with mclk switching with refresh rates over 120 hz on the non-DC code. */
-			if (adev->pm.pm_display_cfg.vrefresh > 120)
-				adev->pm.pm_display_cfg.min_vblank_time = 0;
-			if (adev->powerplay.pp_funcs->display_configuration_change)
-				adev->powerplay.pp_funcs->display_configuration_change(
-								adev->powerplay.pp_handle,
-								&adev->pm.pm_display_cfg);
+			amdgpu_dpm_change_power_state_locked(adev);
 			mutex_unlock(&adev->pm.mutex);
 		}
-		amdgpu_dpm_dispatch_task(adev, AMD_PP_TASK_DISPLAY_CONFIG_CHANGE, NULL);
-	} else {
-		mutex_lock(&adev->pm.mutex);
-		amdgpu_dpm_get_active_displays(adev);
-		amdgpu_dpm_change_power_state_locked(adev);
-		mutex_unlock(&adev->pm.mutex);
 	}
 }

@ -2553,11 +2737,6 @@ static int amdgpu_debugfs_pm_info_pp(struct seq_file *m, struct amdgpu_device *a
 	uint32_t query = 0;
 	int size;

-	/* sanity check PP is enabled */
-	if (!(adev->powerplay.pp_funcs &&
-	      adev->powerplay.pp_funcs->read_sensor))
-	      return -EINVAL;
-
 	/* GPU Clocks */
 	size = sizeof(value);
 	seq_printf(m, "GFX Clocks and Power:\n");
@ -2649,7 +2828,7 @@ static int amdgpu_debugfs_pm_info(struct seq_file *m, void *data)
 	if  ((adev->flags & AMD_IS_PX) &&
 	     (ddev->switch_power_state != DRM_SWITCH_POWER_ON)) {
 		seq_printf(m, "PX asic powered off\n");
-	} else if (adev->powerplay.pp_funcs->debugfs_print_current_performance_level) {
+	} else if (!is_support_sw_smu(adev) && adev->powerplay.pp_funcs->debugfs_print_current_performance_level) {
 		mutex_lock(&adev->pm.mutex);
 		if (adev->powerplay.pp_funcs->debugfs_print_current_performance_level)
 			adev->powerplay.pp_funcs->debugfs_print_current_performance_level(adev, m);
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c
@ -120,6 +120,7 @@ psp_cmd_submit_buf(struct psp_context *psp,
 {
 	int ret;
 	int index;
+	int timeout = 2000;

 	memset(psp->cmd_buf_mem, 0, PSP_CMD_BUFFER_SIZE);

@ -133,8 +134,11 @@ psp_cmd_submit_buf(struct psp_context *psp,
 		return ret;
 	}

-	while (*((unsigned int *)psp->fence_buf) != index)
+	while (*((unsigned int *)psp->fence_buf) != index) {
+		if (--timeout == 0)
+			break;
 		msleep(1);
+	}

 	/* In some cases, psp response status is not 0 even there is no
 	 * problem while the command is submitted. Some version of PSP FW
@ -143,12 +147,14 @@ psp_cmd_submit_buf(struct psp_context *psp,
 	 * during psp initialization to avoid breaking hw_init and it doesn't
 	 * return -EINVAL.
 	 */
-	if (psp->cmd_buf_mem->resp.status) {
+	if (psp->cmd_buf_mem->resp.status || !timeout) {
 		if (ucode)
 			DRM_WARN("failed to load ucode id (%d) ",
 				  ucode->ucode_id);
 		DRM_WARN("psp command failed and response status is (%d)\n",
 			  psp->cmd_buf_mem->resp.status);
+		if (!timeout)
+			return -EINVAL;
 	}

 	/* get xGMI session id from response buffer */
@ -181,13 +187,13 @@ static int psp_tmr_init(struct psp_context *psp)
 	int ret;

 	/*
-	 * Allocate 3M memory aligned to 1M from Frame Buffer (local
-	 * physical).
+	 * According to HW engineer, they prefer the TMR address be "naturally
+	 * aligned" , e.g. the start address be an integer divide of TMR size.
 	 *
 	 * Note: this memory need be reserved till the driver
 	 * uninitializes.
 	 */
-	ret = amdgpu_bo_create_kernel(psp->adev, PSP_TMR_SIZE, 0x100000,
+	ret = amdgpu_bo_create_kernel(psp->adev, PSP_TMR_SIZE, PSP_TMR_SIZE,
 				      AMDGPU_GEM_DOMAIN_VRAM,
 				      &psp->tmr_bo, &psp->tmr_mc_addr, &psp->tmr_buf);

@ -466,6 +472,206 @@ static int psp_xgmi_initialize(struct psp_context *psp)
 	return ret;
 }

+// ras begin
+static void psp_prep_ras_ta_load_cmd_buf(struct psp_gfx_cmd_resp *cmd,
+		uint64_t ras_ta_mc, uint64_t ras_mc_shared,
+		uint32_t ras_ta_size, uint32_t shared_size)
+{
+	cmd->cmd_id = GFX_CMD_ID_LOAD_TA;
+	cmd->cmd.cmd_load_ta.app_phy_addr_lo = lower_32_bits(ras_ta_mc);
+	cmd->cmd.cmd_load_ta.app_phy_addr_hi = upper_32_bits(ras_ta_mc);
+	cmd->cmd.cmd_load_ta.app_len = ras_ta_size;
+
+	cmd->cmd.cmd_load_ta.cmd_buf_phy_addr_lo = lower_32_bits(ras_mc_shared);
+	cmd->cmd.cmd_load_ta.cmd_buf_phy_addr_hi = upper_32_bits(ras_mc_shared);
+	cmd->cmd.cmd_load_ta.cmd_buf_len = shared_size;
+}
+
+static int psp_ras_init_shared_buf(struct psp_context *psp)
+{
+	int ret;
+
+	/*
+	 * Allocate 16k memory aligned to 4k from Frame Buffer (local
+	 * physical) for ras ta <-> Driver
+	 */
+	ret = amdgpu_bo_create_kernel(psp->adev, PSP_RAS_SHARED_MEM_SIZE,
+			PAGE_SIZE, AMDGPU_GEM_DOMAIN_VRAM,
+			&psp->ras.ras_shared_bo,
+			&psp->ras.ras_shared_mc_addr,
+			&psp->ras.ras_shared_buf);
+
+	return ret;
+}
+
+static int psp_ras_load(struct psp_context *psp)
+{
+	int ret;
+	struct psp_gfx_cmd_resp *cmd;
+
+	/*
+	 * TODO: bypass the loading in sriov for now
+	 */
+	if (amdgpu_sriov_vf(psp->adev))
+		return 0;
+
+	cmd = kzalloc(sizeof(struct psp_gfx_cmd_resp), GFP_KERNEL);
+	if (!cmd)
+		return -ENOMEM;
+
+	memset(psp->fw_pri_buf, 0, PSP_1_MEG);
+	memcpy(psp->fw_pri_buf, psp->ta_ras_start_addr, psp->ta_ras_ucode_size);
+
+	psp_prep_ras_ta_load_cmd_buf(cmd, psp->fw_pri_mc_addr,
+			psp->ras.ras_shared_mc_addr,
+			psp->ta_ras_ucode_size, PSP_RAS_SHARED_MEM_SIZE);
+
+	ret = psp_cmd_submit_buf(psp, NULL, cmd,
+			psp->fence_buf_mc_addr);
+
+	if (!ret) {
+		psp->ras.ras_initialized = 1;
+		psp->ras.session_id = cmd->resp.session_id;
+	}
+
+	kfree(cmd);
+
+	return ret;
+}
+
+static void psp_prep_ras_ta_unload_cmd_buf(struct psp_gfx_cmd_resp *cmd,
+						uint32_t ras_session_id)
+{
+	cmd->cmd_id = GFX_CMD_ID_UNLOAD_TA;
+	cmd->cmd.cmd_unload_ta.session_id = ras_session_id;
+}
+
+static int psp_ras_unload(struct psp_context *psp)
+{
+	int ret;
+	struct psp_gfx_cmd_resp *cmd;
+
+	/*
+	 * TODO: bypass the unloading in sriov for now
+	 */
+	if (amdgpu_sriov_vf(psp->adev))
+		return 0;
+
+	cmd = kzalloc(sizeof(struct psp_gfx_cmd_resp), GFP_KERNEL);
+	if (!cmd)
+		return -ENOMEM;
+
+	psp_prep_ras_ta_unload_cmd_buf(cmd, psp->ras.session_id);
+
+	ret = psp_cmd_submit_buf(psp, NULL, cmd,
+			psp->fence_buf_mc_addr);
+
+	kfree(cmd);
+
+	return ret;
+}
+
+static void psp_prep_ras_ta_invoke_cmd_buf(struct psp_gfx_cmd_resp *cmd,
+		uint32_t ta_cmd_id,
+		uint32_t ras_session_id)
+{
+	cmd->cmd_id = GFX_CMD_ID_INVOKE_CMD;
+	cmd->cmd.cmd_invoke_cmd.session_id = ras_session_id;
+	cmd->cmd.cmd_invoke_cmd.ta_cmd_id = ta_cmd_id;
+	/* Note: cmd_invoke_cmd.buf is not used for now */
+}
+
+int psp_ras_invoke(struct psp_context *psp, uint32_t ta_cmd_id)
+{
+	int ret;
+	struct psp_gfx_cmd_resp *cmd;
+
+	/*
+	 * TODO: bypass the loading in sriov for now
+	 */
+	if (amdgpu_sriov_vf(psp->adev))
+		return 0;
+
+	cmd = kzalloc(sizeof(struct psp_gfx_cmd_resp), GFP_KERNEL);
+	if (!cmd)
+		return -ENOMEM;
+
+	psp_prep_ras_ta_invoke_cmd_buf(cmd, ta_cmd_id,
+			psp->ras.session_id);
+
+	ret = psp_cmd_submit_buf(psp, NULL, cmd,
+			psp->fence_buf_mc_addr);
+
+	kfree(cmd);
+
+	return ret;
+}
+
+int psp_ras_enable_features(struct psp_context *psp,
+		union ta_ras_cmd_input *info, bool enable)
+{
+	struct ta_ras_shared_memory *ras_cmd;
+	int ret;
+
+	if (!psp->ras.ras_initialized)
+		return -EINVAL;
+
+	ras_cmd = (struct ta_ras_shared_memory *)psp->ras.ras_shared_buf;
+	memset(ras_cmd, 0, sizeof(struct ta_ras_shared_memory));
+
+	if (enable)
+		ras_cmd->cmd_id = TA_RAS_COMMAND__ENABLE_FEATURES;
+	else
+		ras_cmd->cmd_id = TA_RAS_COMMAND__DISABLE_FEATURES;
+
+	ras_cmd->ras_in_message = *info;
+
+	ret = psp_ras_invoke(psp, ras_cmd->cmd_id);
+	if (ret)
+		return -EINVAL;
+
+	return ras_cmd->ras_status;
+}
+
+static int psp_ras_terminate(struct psp_context *psp)
+{
+	int ret;
+
+	if (!psp->ras.ras_initialized)
+		return 0;
+
+	ret = psp_ras_unload(psp);
+	if (ret)
+		return ret;
+
+	psp->ras.ras_initialized = 0;
+
+	/* free ras shared memory */
+	amdgpu_bo_free_kernel(&psp->ras.ras_shared_bo,
+			&psp->ras.ras_shared_mc_addr,
+			&psp->ras.ras_shared_buf);
+
+	return 0;
+}
+
+static int psp_ras_initialize(struct psp_context *psp)
+{
+	int ret;
+
+	if (!psp->ras.ras_initialized) {
+		ret = psp_ras_init_shared_buf(psp);
+		if (ret)
+			return ret;
+	}
+
+	ret = psp_ras_load(psp);
+	if (ret)
+		return ret;
+
+	return 0;
+}
+// ras end
+
 static int psp_hw_start(struct psp_context *psp)
 {
 	struct amdgpu_device *adev = psp->adev;
@ -473,25 +679,35 @@ static int psp_hw_start(struct psp_context *psp)

 	if (!amdgpu_sriov_vf(adev) || !adev->in_gpu_reset) {
 		ret = psp_bootloader_load_sysdrv(psp);
-		if (ret)
+		if (ret) {
+			DRM_ERROR("PSP load sysdrv failed!\n");
 			return ret;
+		}

 		ret = psp_bootloader_load_sos(psp);
-		if (ret)
+		if (ret) {
+			DRM_ERROR("PSP load sos failed!\n");
 			return ret;
+		}
 	}

 	ret = psp_ring_create(psp, PSP_RING_TYPE__KM);
-	if (ret)
+	if (ret) {
+		DRM_ERROR("PSP create ring failed!\n");
 		return ret;
+	}

 	ret = psp_tmr_load(psp);
-	if (ret)
+	if (ret) {
+		DRM_ERROR("PSP load tmr failed!\n");
 		return ret;
+	}

 	ret = psp_asd_load(psp);
-	if (ret)
+	if (ret) {
+		DRM_ERROR("PSP load asd failed!\n");
 		return ret;
+	}

 	if (adev->gmc.xgmi.num_physical_nodes > 1) {
 		ret = psp_xgmi_initialize(psp);
@ -502,6 +718,15 @@ static int psp_hw_start(struct psp_context *psp)
 			dev_err(psp->adev->dev,
 				"XGMI: Failed to initialize XGMI session\n");
 	}
+
+
+	if (psp->adev->psp.ta_fw) {
+		ret = psp_ras_initialize(psp);
+		if (ret)
+			dev_err(psp->adev->dev,
+					"RAS: Failed to initialize RAS\n");
+	}
+
 	return 0;
 }

@ -665,53 +890,52 @@ static int psp_load_fw(struct amdgpu_device *adev)
 					&psp->fence_buf_mc_addr,
 					&psp->fence_buf);
 	if (ret)
-		goto failed_mem2;
+		goto failed;

 	ret = amdgpu_bo_create_kernel(adev, PSP_CMD_BUFFER_SIZE, PAGE_SIZE,
 				      AMDGPU_GEM_DOMAIN_VRAM,
 				      &psp->cmd_buf_bo, &psp->cmd_buf_mc_addr,
 				      (void **)&psp->cmd_buf_mem);
 	if (ret)
-		goto failed_mem1;
+		goto failed;

 	memset(psp->fence_buf, 0, PSP_FENCE_BUFFER_SIZE);

 	ret = psp_ring_init(psp, PSP_RING_TYPE__KM);
-	if (ret)
-		goto failed_mem;
+	if (ret) {
+		DRM_ERROR("PSP ring init failed!\n");
+		goto failed;
+	}

 	ret = psp_tmr_init(psp);
-	if (ret)
-		goto failed_mem;
+	if (ret) {
+		DRM_ERROR("PSP tmr init failed!\n");
+		goto failed;
+	}

 	ret = psp_asd_init(psp);
-	if (ret)
-		goto failed_mem;
+	if (ret) {
+		DRM_ERROR("PSP asd init failed!\n");
+		goto failed;
+	}

 skip_memalloc:
 	ret = psp_hw_start(psp);
 	if (ret)
-		goto failed_mem;
+		goto failed;

 	ret = psp_np_fw_load(psp);
 	if (ret)
-		goto failed_mem;
+		goto failed;

 	return 0;

-failed_mem:
-	amdgpu_bo_free_kernel(&psp->cmd_buf_bo,
-			      &psp->cmd_buf_mc_addr,
-			      (void **)&psp->cmd_buf_mem);
-failed_mem1:
-	amdgpu_bo_free_kernel(&psp->fence_buf_bo,
-			      &psp->fence_buf_mc_addr, &psp->fence_buf);
-failed_mem2:
-	amdgpu_bo_free_kernel(&psp->fw_pri_bo,
-			      &psp->fw_pri_mc_addr, &psp->fw_pri_buf);
 failed:
-	kfree(psp->cmd);
-	psp->cmd = NULL;
+	/*
+	 * all cleanup jobs (xgmi terminate, ras terminate,
+	 * ring destroy, cmd/fence/fw buffers destory,
+	 * psp->cmd destory) are delayed to psp_hw_fini
+	 */
 	return ret;
 }

@ -753,6 +977,9 @@ static int psp_hw_fini(void *handle)
 	    psp->xgmi_context.initialized == 1)
                psp_xgmi_terminate(psp);

+	if (psp->adev->psp.ta_fw)
+		psp_ras_terminate(psp);
+
 	psp_ring_destroy(psp, PSP_RING_TYPE__KM);

 	amdgpu_bo_free_kernel(&psp->tmr_bo, &psp->tmr_mc_addr, &psp->tmr_buf);
@ -786,6 +1013,14 @@ static int psp_suspend(void *handle)
 		}
 	}

+	if (psp->adev->psp.ta_fw) {
+		ret = psp_ras_terminate(psp);
+		if (ret) {
+			DRM_ERROR("Failed to terminate ras ta\n");
+			return ret;
+		}
+	}
+
 	ret = psp_ring_stop(psp, PSP_RING_TYPE__KM);
 	if (ret) {
 		DRM_ERROR("PSP ring stop failed\n");
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h
@ -28,11 +28,13 @@
 #include "amdgpu.h"
 #include "psp_gfx_if.h"
 #include "ta_xgmi_if.h"
+#include "ta_ras_if.h"

 #define PSP_FENCE_BUFFER_SIZE	0x1000
 #define PSP_CMD_BUFFER_SIZE	0x1000
 #define PSP_ASD_SHARED_MEM_SIZE 0x4000
 #define PSP_XGMI_SHARED_MEM_SIZE 0x4000
+#define PSP_RAS_SHARED_MEM_SIZE 0x4000
 #define PSP_1_MEG		0x100000
 #define PSP_TMR_SIZE	0x400000

@ -88,6 +90,9 @@ struct psp_funcs
 	int (*xgmi_set_topology_info)(struct psp_context *psp, int number_devices,
 				      struct psp_xgmi_topology_info *topology);
 	bool (*support_vmr_ring)(struct psp_context *psp);
+	int (*ras_trigger_error)(struct psp_context *psp,
+			struct ta_ras_trigger_error_input *info);
+	int (*ras_cure_posion)(struct psp_context *psp, uint64_t *mode_ptr);
 };

 struct psp_xgmi_context {
@ -98,6 +103,16 @@ struct psp_xgmi_context {
 	void                            *xgmi_shared_buf;
 };

+struct psp_ras_context {
+	/*ras fw*/
+	bool			ras_initialized;
+	uint32_t		session_id;
+	struct amdgpu_bo	*ras_shared_bo;
+	uint64_t		ras_shared_mc_addr;
+	void			*ras_shared_buf;
+	struct amdgpu_ras	*ras;
+};
+
 struct psp_context
 {
 	struct amdgpu_device            *adev;
@ -150,10 +165,15 @@ struct psp_context

 	/* xgmi ta firmware and buffer */
 	const struct firmware		*ta_fw;
+	uint32_t			ta_fw_version;
 	uint32_t			ta_xgmi_ucode_version;
 	uint32_t			ta_xgmi_ucode_size;
 	uint8_t				*ta_xgmi_start_addr;
+	uint32_t			ta_ras_ucode_version;
+	uint32_t			ta_ras_ucode_size;
+	uint8_t				*ta_ras_start_addr;
 	struct psp_xgmi_context		xgmi_context;
+	struct psp_ras_context		ras;
 };

 struct amdgpu_psp_funcs {
@ -207,6 +227,13 @@ struct psp_xgmi_topology_info {

 #define amdgpu_psp_check_fw_loading_status(adev, i) (adev)->firmware.funcs->check_fw_loading_status((adev), (i))

+#define psp_ras_trigger_error(psp, info) \
+	((psp)->funcs->ras_trigger_error ? \
+	(psp)->funcs->ras_trigger_error((psp), (info)) : -EINVAL)
+#define psp_ras_cure_posion(psp, addr) \
+	((psp)->funcs->ras_cure_posion ? \
+	(psp)->funcs->ras_cure_posion(psp, (addr)) : -EINVAL)
+
 extern const struct amd_ip_funcs psp_ip_funcs;

 extern const struct amdgpu_ip_block_version psp_v3_1_ip_block;
@ -217,6 +244,11 @@ extern const struct amdgpu_ip_block_version psp_v10_0_ip_block;

 int psp_gpu_reset(struct amdgpu_device *adev);
 int psp_xgmi_invoke(struct psp_context *psp, uint32_t ta_cmd_id);
+
+int psp_ras_invoke(struct psp_context *psp, uint32_t ta_cmd_id);
+int psp_ras_enable_features(struct psp_context *psp,
+		union ta_ras_cmd_input *info, bool enable);
+
 extern const struct amdgpu_ip_block_version psp_v11_0_ip_block;

 #endif
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h
@ -0,0 +1,294 @@
+/*
+ * Copyright 2018 Advanced Micro Devices, Inc.
+ *
+ * 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 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 COPYRIGHT HOLDER(S) OR AUTHOR(S) 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 _AMDGPU_RAS_H
+#define _AMDGPU_RAS_H
+
+#include <linux/debugfs.h>
+#include <linux/list.h>
+#include "amdgpu.h"
+#include "amdgpu_psp.h"
+#include "ta_ras_if.h"
+
+enum amdgpu_ras_block {
+	AMDGPU_RAS_BLOCK__UMC = 0,
+	AMDGPU_RAS_BLOCK__SDMA,
+	AMDGPU_RAS_BLOCK__GFX,
+	AMDGPU_RAS_BLOCK__MMHUB,
+	AMDGPU_RAS_BLOCK__ATHUB,
+	AMDGPU_RAS_BLOCK__PCIE_BIF,
+	AMDGPU_RAS_BLOCK__HDP,
+	AMDGPU_RAS_BLOCK__XGMI_WAFL,
+	AMDGPU_RAS_BLOCK__DF,
+	AMDGPU_RAS_BLOCK__SMN,
+	AMDGPU_RAS_BLOCK__SEM,
+	AMDGPU_RAS_BLOCK__MP0,
+	AMDGPU_RAS_BLOCK__MP1,
+	AMDGPU_RAS_BLOCK__FUSE,
+
+	AMDGPU_RAS_BLOCK__LAST
+};
+
+#define AMDGPU_RAS_BLOCK_COUNT	AMDGPU_RAS_BLOCK__LAST
+#define AMDGPU_RAS_BLOCK_MASK	((1ULL << AMDGPU_RAS_BLOCK_COUNT) - 1)
+
+enum amdgpu_ras_error_type {
+	AMDGPU_RAS_ERROR__NONE							= 0,
+	AMDGPU_RAS_ERROR__PARITY						= 1,
+	AMDGPU_RAS_ERROR__SINGLE_CORRECTABLE					= 2,
+	AMDGPU_RAS_ERROR__MULTI_UNCORRECTABLE					= 4,
+	AMDGPU_RAS_ERROR__POISON						= 8,
+};
+
+enum amdgpu_ras_ret {
+	AMDGPU_RAS_SUCCESS = 0,
+	AMDGPU_RAS_FAIL,
+	AMDGPU_RAS_UE,
+	AMDGPU_RAS_CE,
+	AMDGPU_RAS_PT,
+};
+
+struct ras_common_if {
+	enum amdgpu_ras_block block;
+	enum amdgpu_ras_error_type type;
+	uint32_t sub_block_index;
+	/* block name */
+	char name[32];
+};
+
+typedef int (*ras_ih_cb)(struct amdgpu_device *adev,
+		struct amdgpu_iv_entry *entry);
+
+struct amdgpu_ras {
+	/* ras infrastructure */
+	/* for ras itself. */
+	uint32_t hw_supported;
+	/* for IP to check its ras ability. */
+	uint32_t supported;
+	uint32_t features;
+	struct list_head head;
+	/* debugfs */
+	struct dentry *dir;
+	/* debugfs ctrl */
+	struct dentry *ent;
+	/* sysfs */
+	struct device_attribute features_attr;
+	/* block array */
+	struct ras_manager *objs;
+
+	/* gpu recovery */
+	struct work_struct recovery_work;
+	atomic_t in_recovery;
+	struct amdgpu_device *adev;
+	/* error handler data */
+	struct ras_err_handler_data *eh_data;
+	struct mutex recovery_lock;
+
+	uint32_t flags;
+};
+
+/* interfaces for IP */
+
+struct ras_fs_if {
+	struct ras_common_if head;
+	char sysfs_name[32];
+	char debugfs_name[32];
+};
+
+struct ras_query_if {
+	struct ras_common_if head;
+	unsigned long ue_count;
+	unsigned long ce_count;
+};
+
+struct ras_inject_if {
+	struct ras_common_if head;
+	uint64_t address;
+	uint64_t value;
+};
+
+struct ras_cure_if {
+	struct ras_common_if head;
+	uint64_t address;
+};
+
+struct ras_ih_if {
+	struct ras_common_if head;
+	ras_ih_cb cb;
+};
+
+struct ras_dispatch_if {
+	struct ras_common_if head;
+	struct amdgpu_iv_entry *entry;
+};
+
+struct ras_debug_if {
+	union {
+		struct ras_common_if head;
+		struct ras_inject_if inject;
+	};
+	int op;
+};
+/* work flow
+ * vbios
+ * 1: ras feature enable (enabled by default)
+ * psp
+ * 2: ras framework init (in ip_init)
+ * IP
+ * 3: IH add
+ * 4: debugfs/sysfs create
+ * 5: query/inject
+ * 6: debugfs/sysfs remove
+ * 7: IH remove
+ * 8: feature disable
+ */
+
+#define amdgpu_ras_get_context(adev)		((adev)->psp.ras.ras)
+#define amdgpu_ras_set_context(adev, ras_con)	((adev)->psp.ras.ras = (ras_con))
+
+/* check if ras is supported on block, say, sdma, gfx */
+static inline int amdgpu_ras_is_supported(struct amdgpu_device *adev,
+		unsigned int block)
+{
+	struct amdgpu_ras *ras = amdgpu_ras_get_context(adev);
+
+	return ras && (ras->supported & (1 << block));
+}
+
+int amdgpu_ras_query_error_count(struct amdgpu_device *adev,
+		bool is_ce);
+
+/* error handling functions */
+int amdgpu_ras_add_bad_pages(struct amdgpu_device *adev,
+		unsigned long *bps, int pages);
+
+int amdgpu_ras_reserve_bad_pages(struct amdgpu_device *adev);
+
+static inline int amdgpu_ras_reset_gpu(struct amdgpu_device *adev,
+		bool is_baco)
+{
+	/* remove me when gpu reset works on vega20 A1. */
+#if 0
+	struct amdgpu_ras *ras = amdgpu_ras_get_context(adev);
+
+	if (atomic_cmpxchg(&ras->in_recovery, 0, 1) == 0)
+		schedule_work(&ras->recovery_work);
+#endif
+	return 0;
+}
+
+static inline enum ta_ras_block
+amdgpu_ras_block_to_ta(enum amdgpu_ras_block block) {
+	switch (block) {
+	case AMDGPU_RAS_BLOCK__UMC:
+		return TA_RAS_BLOCK__UMC;
+	case AMDGPU_RAS_BLOCK__SDMA:
+		return TA_RAS_BLOCK__SDMA;
+	case AMDGPU_RAS_BLOCK__GFX:
+		return TA_RAS_BLOCK__GFX;
+	case AMDGPU_RAS_BLOCK__MMHUB:
+		return TA_RAS_BLOCK__MMHUB;
+	case AMDGPU_RAS_BLOCK__ATHUB:
+		return TA_RAS_BLOCK__ATHUB;
+	case AMDGPU_RAS_BLOCK__PCIE_BIF:
+		return TA_RAS_BLOCK__PCIE_BIF;
+	case AMDGPU_RAS_BLOCK__HDP:
+		return TA_RAS_BLOCK__HDP;
+	case AMDGPU_RAS_BLOCK__XGMI_WAFL:
+		return TA_RAS_BLOCK__XGMI_WAFL;
+	case AMDGPU_RAS_BLOCK__DF:
+		return TA_RAS_BLOCK__DF;
+	case AMDGPU_RAS_BLOCK__SMN:
+		return TA_RAS_BLOCK__SMN;
+	case AMDGPU_RAS_BLOCK__SEM:
+		return TA_RAS_BLOCK__SEM;
+	case AMDGPU_RAS_BLOCK__MP0:
+		return TA_RAS_BLOCK__MP0;
+	case AMDGPU_RAS_BLOCK__MP1:
+		return TA_RAS_BLOCK__MP1;
+	case AMDGPU_RAS_BLOCK__FUSE:
+		return TA_RAS_BLOCK__FUSE;
+	default:
+		WARN_ONCE(1, "RAS ERROR: unexpected block id %d\n", block);
+		return TA_RAS_BLOCK__UMC;
+	}
+}
+
+static inline enum ta_ras_error_type
+amdgpu_ras_error_to_ta(enum amdgpu_ras_error_type error) {
+	switch (error) {
+	case AMDGPU_RAS_ERROR__NONE:
+		return TA_RAS_ERROR__NONE;
+	case AMDGPU_RAS_ERROR__PARITY:
+		return TA_RAS_ERROR__PARITY;
+	case AMDGPU_RAS_ERROR__SINGLE_CORRECTABLE:
+		return TA_RAS_ERROR__SINGLE_CORRECTABLE;
+	case AMDGPU_RAS_ERROR__MULTI_UNCORRECTABLE:
+		return TA_RAS_ERROR__MULTI_UNCORRECTABLE;
+	case AMDGPU_RAS_ERROR__POISON:
+		return TA_RAS_ERROR__POISON;
+	default:
+		WARN_ONCE(1, "RAS ERROR: unexpected error type %d\n", error);
+		return TA_RAS_ERROR__NONE;
+	}
+}
+
+/* called in ip_init and ip_fini */
+int amdgpu_ras_init(struct amdgpu_device *adev);
+void amdgpu_ras_post_init(struct amdgpu_device *adev);
+int amdgpu_ras_fini(struct amdgpu_device *adev);
+int amdgpu_ras_pre_fini(struct amdgpu_device *adev);
+
+int amdgpu_ras_feature_enable(struct amdgpu_device *adev,
+		struct ras_common_if *head, bool enable);
+
+int amdgpu_ras_feature_enable_on_boot(struct amdgpu_device *adev,
+		struct ras_common_if *head, bool enable);
+
+int amdgpu_ras_sysfs_create(struct amdgpu_device *adev,
+		struct ras_fs_if *head);
+
+int amdgpu_ras_sysfs_remove(struct amdgpu_device *adev,
+		struct ras_common_if *head);
+
+int amdgpu_ras_debugfs_create(struct amdgpu_device *adev,
+		struct ras_fs_if *head);
+
+int amdgpu_ras_debugfs_remove(struct amdgpu_device *adev,
+		struct ras_common_if *head);
+
+int amdgpu_ras_error_query(struct amdgpu_device *adev,
+		struct ras_query_if *info);
+
+int amdgpu_ras_error_inject(struct amdgpu_device *adev,
+		struct ras_inject_if *info);
+
+int amdgpu_ras_interrupt_add_handler(struct amdgpu_device *adev,
+		struct ras_ih_if *info);
+
+int amdgpu_ras_interrupt_remove_handler(struct amdgpu_device *adev,
+		struct ras_ih_if *info);
+
+int amdgpu_ras_interrupt_dispatch(struct amdgpu_device *adev,
+		struct ras_dispatch_if *info);
+#endif
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c
@ -248,6 +248,8 @@ int amdgpu_ring_init(struct amdgpu_device *adev, struct amdgpu_ring *ring,
 	 */
 	if (ring->funcs->type == AMDGPU_RING_TYPE_KIQ)
 		sched_hw_submission = max(sched_hw_submission, 256);
+	else if (ring == &adev->sdma.instance[0].page)
+		sched_hw_submission = 256;

 	if (ring->adev == NULL) {
 		if (adev->num_rings >= AMDGPU_MAX_RINGS)
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_sdma.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_sdma.h
@ -28,9 +28,8 @@
 #define AMDGPU_MAX_SDMA_INSTANCES		2

 enum amdgpu_sdma_irq {
-	AMDGPU_SDMA_IRQ_TRAP0 = 0,
-	AMDGPU_SDMA_IRQ_TRAP1,
-
+	AMDGPU_SDMA_IRQ_INSTANCE0  = 0,
+	AMDGPU_SDMA_IRQ_INSTANCE1,
 	AMDGPU_SDMA_IRQ_LAST
 };

@ -49,9 +48,11 @@ struct amdgpu_sdma {
 	struct amdgpu_sdma_instance instance[AMDGPU_MAX_SDMA_INSTANCES];
 	struct amdgpu_irq_src	trap_irq;
 	struct amdgpu_irq_src	illegal_inst_irq;
+	struct amdgpu_irq_src	ecc_irq;
 	int			num_instances;
 	uint32_t                    srbm_soft_reset;
 	bool			has_page_queue;
+	struct ras_common_if	*ras_if;
 };

 /*
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
@ -50,8 +50,6 @@
 #include "amdgpu_sdma.h"
 #include "bif/bif_4_1_d.h"

-#define DRM_FILE_PAGE_OFFSET (0x100000000ULL >> PAGE_SHIFT)
-
 static int amdgpu_map_buffer(struct ttm_buffer_object *bo,
 			     struct ttm_mem_reg *mem, unsigned num_pages,
 			     uint64_t offset, unsigned window,
@ -1424,6 +1422,13 @@ static bool amdgpu_ttm_bo_eviction_valuable(struct ttm_buffer_object *bo,
 	struct dma_fence *f;
 	int i;

+	/* Don't evict VM page tables while they are busy, otherwise we can't
+	 * cleanly handle page faults.
+	 */
+	if (bo->type == ttm_bo_type_kernel &&
+	    !reservation_object_test_signaled_rcu(bo->resv, true))
+		return false;
+
 	/* If bo is a KFD BO, check if the bo belongs to the current process.
 	 * If true, then return false as any KFD process needs all its BOs to
 	 * be resident to run successfully
@ -1671,7 +1676,6 @@ int amdgpu_ttm_init(struct amdgpu_device *adev)
 	r = ttm_bo_device_init(&adev->mman.bdev,
 			       &amdgpu_bo_driver,
 			       adev->ddev->anon_inode->i_mapping,
-			       DRM_FILE_PAGE_OFFSET,
 			       adev->need_dma32);
 	if (r) {
 		DRM_ERROR("failed initializing buffer object driver(%d).\n", r);
@ -1877,14 +1881,9 @@ void amdgpu_ttm_set_buffer_funcs_status(struct amdgpu_device *adev, bool enable)

 int amdgpu_mmap(struct file *filp, struct vm_area_struct *vma)
 {
-	struct drm_file *file_priv;
-	struct amdgpu_device *adev;
+	struct drm_file *file_priv = filp->private_data;
+	struct amdgpu_device *adev = file_priv->minor->dev->dev_private;

-	if (unlikely(vma->vm_pgoff < DRM_FILE_PAGE_OFFSET))
-		return -EINVAL;
-
-	file_priv = filp->private_data;
-	adev = file_priv->minor->dev->dev_private;
 	if (adev == NULL)
 		return -EINVAL;

--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c
@ -36,6 +36,7 @@ void amdgpu_virt_init_setting(struct amdgpu_device *adev)
 	/* enable virtual display */
 	adev->mode_info.num_crtc = 1;
 	adev->enable_virtual_display = true;
+	adev->ddev->driver->driver_features &= ~DRIVER_ATOMIC;
 	adev->cg_flags = 0;
 	adev->pg_flags = 0;
 }
@ -375,4 +376,53 @@ void amdgpu_virt_init_data_exchange(struct amdgpu_device *adev)
 	}
 }

+static uint32_t parse_clk(char *buf, bool min)
+{
+        char *ptr = buf;
+        uint32_t clk = 0;
+
+        do {
+                ptr = strchr(ptr, ':');
+                if (!ptr)
+                        break;
+                ptr+=2;
+                clk = simple_strtoul(ptr, NULL, 10);
+        } while (!min);
+
+        return clk * 100;
+}
+
+uint32_t amdgpu_virt_get_sclk(struct amdgpu_device *adev, bool lowest)
+{
+	char *buf = NULL;
+	uint32_t clk = 0;
+
+	buf = kzalloc(PAGE_SIZE, GFP_KERNEL);
+	if (!buf)
+		return -ENOMEM;
+
+	adev->virt.ops->get_pp_clk(adev, PP_SCLK, buf);
+	clk = parse_clk(buf, lowest);
+
+	kfree(buf);
+
+	return clk;
+}
+
+uint32_t amdgpu_virt_get_mclk(struct amdgpu_device *adev, bool lowest)
+{
+	char *buf = NULL;
+	uint32_t clk = 0;
+
+	buf = kzalloc(PAGE_SIZE, GFP_KERNEL);
+	if (!buf)
+		return -ENOMEM;
+
+	adev->virt.ops->get_pp_clk(adev, PP_MCLK, buf);
+	clk = parse_clk(buf, lowest);
+
+	kfree(buf);
+
+	return clk;
+}

--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.h
@ -57,6 +57,8 @@ struct amdgpu_virt_ops {
 	int (*reset_gpu)(struct amdgpu_device *adev);
 	int (*wait_reset)(struct amdgpu_device *adev);
 	void (*trans_msg)(struct amdgpu_device *adev, u32 req, u32 data1, u32 data2, u32 data3);
+	int (*get_pp_clk)(struct amdgpu_device *adev, u32 type, char *buf);
+	int (*force_dpm_level)(struct amdgpu_device *adev, u32 level);
 };

 /*
@ -83,6 +85,8 @@ enum AMDGIM_FEATURE_FLAG {
 	AMDGIM_FEATURE_GIM_LOAD_UCODES   = 0x2,
 	/* VRAM LOST by GIM */
 	AMDGIM_FEATURE_GIM_FLR_VRAMLOST = 0x4,
+	/* HW PERF SIM in GIM */
+	AMDGIM_FEATURE_HW_PERF_SIMULATION = (1 << 3),
 };

 struct amd_sriov_msg_pf2vf_info_header {
@ -252,6 +256,8 @@ struct amdgpu_virt {
 	struct amdgpu_vf_error_buffer   vf_errors;
 	struct amdgpu_virt_fw_reserve	fw_reserve;
 	uint32_t gim_feature;
+	/* protect DPM events to GIM */
+	struct mutex                    dpm_mutex;
 };

 #define amdgpu_sriov_enabled(adev) \
@ -278,6 +284,9 @@ static inline bool is_virtual_machine(void)
 #endif
 }

+#define amdgim_is_hwperf(adev) \
+	((adev)->virt.gim_feature & AMDGIM_FEATURE_HW_PERF_SIMULATION)
+
 bool amdgpu_virt_mmio_blocked(struct amdgpu_device *adev);
 void amdgpu_virt_init_setting(struct amdgpu_device *adev);
 uint32_t amdgpu_virt_kiq_rreg(struct amdgpu_device *adev, uint32_t reg);
@ -295,5 +304,7 @@ int amdgpu_virt_fw_reserve_get_checksum(void *obj, unsigned long obj_size,
 					unsigned int key,
 					unsigned int chksum);
 void amdgpu_virt_init_data_exchange(struct amdgpu_device *adev);
+uint32_t amdgpu_virt_get_sclk(struct amdgpu_device *adev, bool lowest);
+uint32_t amdgpu_virt_get_mclk(struct amdgpu_device *adev, bool lowest);

 #endif
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h
@ -30,7 +30,6 @@
 #include <drm/gpu_scheduler.h>
 #include <drm/drm_file.h>
 #include <drm/ttm/ttm_bo_driver.h>
-#include <linux/chash.h>

 #include "amdgpu_sync.h"
 #include "amdgpu_ring.h"
@ -140,7 +139,6 @@ struct amdgpu_vm_bo_base {

 struct amdgpu_vm_pt {
 	struct amdgpu_vm_bo_base	base;
-	bool				huge;

 	/* array of page tables, one for each directory entry */
 	struct amdgpu_vm_pt		*entries;
@ -167,11 +165,6 @@ struct amdgpu_vm_pte_funcs {
 			    uint32_t incr, uint64_t flags);
 };

-#define AMDGPU_VM_FAULT(pasid, addr) (((u64)(pasid) << 48) | (addr))
-#define AMDGPU_VM_FAULT_PASID(fault) ((u64)(fault) >> 48)
-#define AMDGPU_VM_FAULT_ADDR(fault)  ((u64)(fault) & 0xfffffffff000ULL)
-
-
 struct amdgpu_task_info {
 	char	process_name[TASK_COMM_LEN];
 	char	task_name[TASK_COMM_LEN];
@ -179,11 +172,52 @@ struct amdgpu_task_info {
 	pid_t	tgid;
 };

-#define AMDGPU_PAGEFAULT_HASH_BITS 8
-struct amdgpu_retryfault_hashtable {
-	DECLARE_CHASH_TABLE(hash, AMDGPU_PAGEFAULT_HASH_BITS, 8, 0);
-	spinlock_t	lock;
-	int		count;
+/**
+ * struct amdgpu_vm_update_params
+ *
+ * Encapsulate some VM table update parameters to reduce
+ * the number of function parameters
+ *
+ */
+struct amdgpu_vm_update_params {
+
+	/**
+	 * @adev: amdgpu device we do this update for
+	 */
+	struct amdgpu_device *adev;
+
+	/**
+	 * @vm: optional amdgpu_vm we do this update for
+	 */
+	struct amdgpu_vm *vm;
+
+	/**
+	 * @pages_addr:
+	 *
+	 * DMA addresses to use for mapping
+	 */
+	dma_addr_t *pages_addr;
+
+	/**
+	 * @job: job to used for hw submission
+	 */
+	struct amdgpu_job *job;
+
+	/**
+	 * @num_dw_left: number of dw left for the IB
+	 */
+	unsigned int num_dw_left;
+};
+
+struct amdgpu_vm_update_funcs {
+	int (*map_table)(struct amdgpu_bo *bo);
+	int (*prepare)(struct amdgpu_vm_update_params *p, void * owner,
+		       struct dma_fence *exclusive);
+	int (*update)(struct amdgpu_vm_update_params *p,
+		      struct amdgpu_bo *bo, uint64_t pe, uint64_t addr,
+		      unsigned count, uint32_t incr, uint64_t flags);
+	int (*commit)(struct amdgpu_vm_update_params *p,
+		      struct dma_fence **fence);
 };

 struct amdgpu_vm {
@ -221,7 +255,10 @@ struct amdgpu_vm {
 	struct amdgpu_vmid	*reserved_vmid[AMDGPU_MAX_VMHUBS];

 	/* Flag to indicate if VM tables are updated by CPU or GPU (SDMA) */
-	bool                    use_cpu_for_update;
+	bool					use_cpu_for_update;
+
+	/* Functions to use for VM table updates */
+	const struct amdgpu_vm_update_funcs	*update_funcs;

 	/* Flag to indicate ATS support from PTE for GFX9 */
 	bool			pte_support_ats;
@ -245,7 +282,6 @@ struct amdgpu_vm {
 	struct ttm_lru_bulk_move lru_bulk_move;
 	/* mark whether can do the bulk move */
 	bool			bulk_moveable;
-	struct amdgpu_retryfault_hashtable *fault_hash;
 };

 struct amdgpu_vm_manager {
@ -267,6 +303,7 @@ struct amdgpu_vm_manager {
 	const struct amdgpu_vm_pte_funcs	*vm_pte_funcs;
 	struct drm_sched_rq			*vm_pte_rqs[AMDGPU_MAX_RINGS];
 	unsigned				vm_pte_num_rqs;
+	struct amdgpu_ring			*page_fault;

 	/* partial resident texture handling */
 	spinlock_t				prt_lock;
@ -283,14 +320,23 @@ struct amdgpu_vm_manager {
 	 */
 	struct idr				pasid_idr;
 	spinlock_t				pasid_lock;
+
+	/* counter of mapped memory through xgmi */
+	uint32_t				xgmi_map_counter;
+	struct mutex				lock_pstate;
 };

 #define amdgpu_vm_copy_pte(adev, ib, pe, src, count) ((adev)->vm_manager.vm_pte_funcs->copy_pte((ib), (pe), (src), (count)))
 #define amdgpu_vm_write_pte(adev, ib, pe, value, count, incr) ((adev)->vm_manager.vm_pte_funcs->write_pte((ib), (pe), (value), (count), (incr)))
 #define amdgpu_vm_set_pte_pde(adev, ib, pe, addr, count, incr, flags) ((adev)->vm_manager.vm_pte_funcs->set_pte_pde((ib), (pe), (addr), (count), (incr), (flags)))

+extern const struct amdgpu_vm_update_funcs amdgpu_vm_cpu_funcs;
+extern const struct amdgpu_vm_update_funcs amdgpu_vm_sdma_funcs;
+
 void amdgpu_vm_manager_init(struct amdgpu_device *adev);
 void amdgpu_vm_manager_fini(struct amdgpu_device *adev);
+
+long amdgpu_vm_wait_idle(struct amdgpu_vm *vm, long timeout);
 int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm,
 		   int vm_context, unsigned int pasid);
 int amdgpu_vm_make_compute(struct amdgpu_device *adev, struct amdgpu_vm *vm, unsigned int pasid);
@ -303,9 +349,6 @@ bool amdgpu_vm_ready(struct amdgpu_vm *vm);
 int amdgpu_vm_validate_pt_bos(struct amdgpu_device *adev, struct amdgpu_vm *vm,
 			      int (*callback)(void *p, struct amdgpu_bo *bo),
 			      void *param);
-int amdgpu_vm_alloc_pts(struct amdgpu_device *adev,
-			struct amdgpu_vm *vm,
-			uint64_t saddr, uint64_t size);
 int amdgpu_vm_flush(struct amdgpu_ring *ring, struct amdgpu_job *job, bool need_pipe_sync);
 int amdgpu_vm_update_directories(struct amdgpu_device *adev,
 				 struct amdgpu_vm *vm);
@ -319,6 +362,7 @@ int amdgpu_vm_bo_update(struct amdgpu_device *adev,
 			bool clear);
 void amdgpu_vm_bo_invalidate(struct amdgpu_device *adev,
 			     struct amdgpu_bo *bo, bool evicted);
+uint64_t amdgpu_vm_map_gart(const dma_addr_t *pages_addr, uint64_t addr);
 struct amdgpu_bo_va *amdgpu_vm_bo_find(struct amdgpu_vm *vm,
 				       struct amdgpu_bo *bo);
 struct amdgpu_bo_va *amdgpu_vm_bo_add(struct amdgpu_device *adev,
@ -358,11 +402,6 @@ void amdgpu_vm_set_task_info(struct amdgpu_vm *vm);

 void amdgpu_vm_move_to_lru_tail(struct amdgpu_device *adev,
 				struct amdgpu_vm *vm);
-
-int amdgpu_vm_add_fault(struct amdgpu_retryfault_hashtable *fault_hash, u64 key);
-
-void amdgpu_vm_clear_fault(struct amdgpu_retryfault_hashtable *fault_hash, u64 key);
-
 void amdgpu_vm_del_from_lru_notify(struct ttm_buffer_object *bo);

 #endif
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_cpu.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_cpu.c
@ -0,0 +1,127 @@
+/*
+ * Copyright 2019 Advanced Micro Devices, Inc.
+ *
+ * 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 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 COPYRIGHT HOLDER(S) OR AUTHOR(S) 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 "amdgpu_vm.h"
+#include "amdgpu_object.h"
+#include "amdgpu_trace.h"
+
+/**
+ * amdgpu_vm_cpu_map_table - make sure new PDs/PTs are kmapped
+ *
+ * @table: newly allocated or validated PD/PT
+ */
+static int amdgpu_vm_cpu_map_table(struct amdgpu_bo *table)
+{
+	return amdgpu_bo_kmap(table, NULL);
+}
+
+/**
+ * amdgpu_vm_cpu_prepare - prepare page table update with the CPU
+ *
+ * @p: see amdgpu_vm_update_params definition
+ * @owner: owner we need to sync to
+ * @exclusive: exclusive move fence we need to sync to
+ *
+ * Returns:
+ * Negativ errno, 0 for success.
+ */
+static int amdgpu_vm_cpu_prepare(struct amdgpu_vm_update_params *p, void *owner,
+				 struct dma_fence *exclusive)
+{
+	int r;
+
+	/* Wait for PT BOs to be idle. PTs share the same resv. object
+	 * as the root PD BO
+	 */
+	r = amdgpu_bo_sync_wait(p->vm->root.base.bo, owner, true);
+	if (unlikely(r))
+		return r;
+
+	/* Wait for any BO move to be completed */
+	if (exclusive) {
+		r = dma_fence_wait(exclusive, true);
+		if (unlikely(r))
+			return r;
+	}
+
+	return 0;
+}
+
+/**
+ * amdgpu_vm_cpu_update - helper to update page tables via CPU
+ *
+ * @p: see amdgpu_vm_update_params definition
+ * @bo: PD/PT to update
+ * @pe: kmap addr of the page entry
+ * @addr: dst addr to write into pe
+ * @count: number of page entries to update
+ * @incr: increase next addr by incr bytes
+ * @flags: hw access flags
+ *
+ * Write count number of PT/PD entries directly.
+ */
+static int amdgpu_vm_cpu_update(struct amdgpu_vm_update_params *p,
+				struct amdgpu_bo *bo, uint64_t pe,
+				uint64_t addr, unsigned count, uint32_t incr,
+				uint64_t flags)
+{
+	unsigned int i;
+	uint64_t value;
+
+	pe += (unsigned long)amdgpu_bo_kptr(bo);
+
+	trace_amdgpu_vm_set_ptes(pe, addr, count, incr, flags);
+
+	for (i = 0; i < count; i++) {
+		value = p->pages_addr ?
+			amdgpu_vm_map_gart(p->pages_addr, addr) :
+			addr;
+		amdgpu_gmc_set_pte_pde(p->adev, (void *)(uintptr_t)pe,
+				       i, value, flags);
+		addr += incr;
+	}
+	return 0;
+}
+
+/**
+ * amdgpu_vm_cpu_commit - commit page table update to the HW
+ *
+ * @p: see amdgpu_vm_update_params definition
+ * @fence: unused
+ *
+ * Make sure that the hardware sees the page table updates.
+ */
+static int amdgpu_vm_cpu_commit(struct amdgpu_vm_update_params *p,
+				struct dma_fence **fence)
+{
+	/* Flush HDP */
+	mb();
+	amdgpu_asic_flush_hdp(p->adev, NULL);
+	return 0;
+}
+
+const struct amdgpu_vm_update_funcs amdgpu_vm_cpu_funcs = {
+	.map_table = amdgpu_vm_cpu_map_table,
+	.prepare = amdgpu_vm_cpu_prepare,
+	.update = amdgpu_vm_cpu_update,
+	.commit = amdgpu_vm_cpu_commit
+};
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_sdma.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_sdma.c
@ -0,0 +1,270 @@
+/*
+ * Copyright 2019 Advanced Micro Devices, Inc.
+ *
+ * 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 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 COPYRIGHT HOLDER(S) OR AUTHOR(S) 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 "amdgpu_vm.h"
+#include "amdgpu_job.h"
+#include "amdgpu_object.h"
+#include "amdgpu_trace.h"
+
+#define AMDGPU_VM_SDMA_MIN_NUM_DW	256u
+#define AMDGPU_VM_SDMA_MAX_NUM_DW	(16u * 1024u)
+
+/**
+ * amdgpu_vm_sdma_map_table - make sure new PDs/PTs are GTT mapped
+ *
+ * @table: newly allocated or validated PD/PT
+ */
+static int amdgpu_vm_sdma_map_table(struct amdgpu_bo *table)
+{
+	int r;
+
+	r = amdgpu_ttm_alloc_gart(&table->tbo);
+	if (r)
+		return r;
+
+	if (table->shadow)
+		r = amdgpu_ttm_alloc_gart(&table->shadow->tbo);
+
+	return r;
+}
+
+/**
+ * amdgpu_vm_sdma_prepare - prepare SDMA command submission
+ *
+ * @p: see amdgpu_vm_update_params definition
+ * @owner: owner we need to sync to
+ * @exclusive: exclusive move fence we need to sync to
+ *
+ * Returns:
+ * Negativ errno, 0 for success.
+ */
+static int amdgpu_vm_sdma_prepare(struct amdgpu_vm_update_params *p,
+				  void *owner, struct dma_fence *exclusive)
+{
+	struct amdgpu_bo *root = p->vm->root.base.bo;
+	unsigned int ndw = AMDGPU_VM_SDMA_MIN_NUM_DW;
+	int r;
+
+	r = amdgpu_job_alloc_with_ib(p->adev, ndw * 4, &p->job);
+	if (r)
+		return r;
+
+	r = amdgpu_sync_fence(p->adev, &p->job->sync, exclusive, false);
+	if (r)
+		return r;
+
+	r = amdgpu_sync_resv(p->adev, &p->job->sync, root->tbo.resv,
+			     owner, false);
+	if (r)
+		return r;
+
+	p->num_dw_left = ndw;
+	return 0;
+}
+
+/**
+ * amdgpu_vm_sdma_commit - commit SDMA command submission
+ *
+ * @p: see amdgpu_vm_update_params definition
+ * @fence: resulting fence
+ *
+ * Returns:
+ * Negativ errno, 0 for success.
+ */
+static int amdgpu_vm_sdma_commit(struct amdgpu_vm_update_params *p,
+				 struct dma_fence **fence)
+{
+	struct amdgpu_bo *root = p->vm->root.base.bo;
+	struct amdgpu_ib *ib = p->job->ibs;
+	struct amdgpu_ring *ring;
+	struct dma_fence *f;
+	int r;
+
+	ring = container_of(p->vm->entity.rq->sched, struct amdgpu_ring, sched);
+
+	WARN_ON(ib->length_dw == 0);
+	amdgpu_ring_pad_ib(ring, ib);
+	WARN_ON(ib->length_dw > p->num_dw_left);
+	r = amdgpu_job_submit(p->job, &p->vm->entity,
+			      AMDGPU_FENCE_OWNER_VM, &f);
+	if (r)
+		goto error;
+
+	amdgpu_bo_fence(root, f, true);
+	if (fence)
+		swap(*fence, f);
+	dma_fence_put(f);
+	return 0;
+
+error:
+	amdgpu_job_free(p->job);
+	return r;
+}
+
+
+/**
+ * amdgpu_vm_sdma_copy_ptes - copy the PTEs from mapping
+ *
+ * @p: see amdgpu_vm_update_params definition
+ * @bo: PD/PT to update
+ * @pe: addr of the page entry
+ * @count: number of page entries to copy
+ *
+ * Traces the parameters and calls the DMA function to copy the PTEs.
+ */
+static void amdgpu_vm_sdma_copy_ptes(struct amdgpu_vm_update_params *p,
+				     struct amdgpu_bo *bo, uint64_t pe,
+				     unsigned count)
+{
+	struct amdgpu_ib *ib = p->job->ibs;
+	uint64_t src = ib->gpu_addr;
+
+	src += p->num_dw_left * 4;
+
+	pe += amdgpu_bo_gpu_offset(bo);
+	trace_amdgpu_vm_copy_ptes(pe, src, count);
+
+	amdgpu_vm_copy_pte(p->adev, ib, pe, src, count);
+}
+
+/**
+ * amdgpu_vm_sdma_set_ptes - helper to call the right asic function
+ *
+ * @p: see amdgpu_vm_update_params definition
+ * @bo: PD/PT to update
+ * @pe: addr of the page entry
+ * @addr: dst addr to write into pe
+ * @count: number of page entries to update
+ * @incr: increase next addr by incr bytes
+ * @flags: hw access flags
+ *
+ * Traces the parameters and calls the right asic functions
+ * to setup the page table using the DMA.
+ */
+static void amdgpu_vm_sdma_set_ptes(struct amdgpu_vm_update_params *p,
+				    struct amdgpu_bo *bo, uint64_t pe,
+				    uint64_t addr, unsigned count,
+				    uint32_t incr, uint64_t flags)
+{
+	struct amdgpu_ib *ib = p->job->ibs;
+
+	pe += amdgpu_bo_gpu_offset(bo);
+	trace_amdgpu_vm_set_ptes(pe, addr, count, incr, flags);
+	if (count < 3) {
+		amdgpu_vm_write_pte(p->adev, ib, pe, addr | flags,
+				    count, incr);
+	} else {
+		amdgpu_vm_set_pte_pde(p->adev, ib, pe, addr,
+				      count, incr, flags);
+	}
+}
+
+/**
+ * amdgpu_vm_sdma_update - execute VM update
+ *
+ * @p: see amdgpu_vm_update_params definition
+ * @bo: PD/PT to update
+ * @pe: addr of the page entry
+ * @addr: dst addr to write into pe
+ * @count: number of page entries to update
+ * @incr: increase next addr by incr bytes
+ * @flags: hw access flags
+ *
+ * Reserve space in the IB, setup mapping buffer on demand and write commands to
+ * the IB.
+ */
+static int amdgpu_vm_sdma_update(struct amdgpu_vm_update_params *p,
+				 struct amdgpu_bo *bo, uint64_t pe,
+				 uint64_t addr, unsigned count, uint32_t incr,
+				 uint64_t flags)
+{
+	unsigned int i, ndw, nptes;
+	uint64_t *pte;
+	int r;
+
+	do {
+		ndw = p->num_dw_left;
+		ndw -= p->job->ibs->length_dw;
+
+		if (ndw < 32) {
+			r = amdgpu_vm_sdma_commit(p, NULL);
+			if (r)
+				return r;
+
+			/* estimate how many dw we need */
+			ndw = 32;
+			if (p->pages_addr)
+				ndw += count * 2;
+			ndw = max(ndw, AMDGPU_VM_SDMA_MIN_NUM_DW);
+			ndw = min(ndw, AMDGPU_VM_SDMA_MAX_NUM_DW);
+
+			r = amdgpu_job_alloc_with_ib(p->adev, ndw * 4, &p->job);
+			if (r)
+				return r;
+
+			p->num_dw_left = ndw;
+		}
+
+		if (!p->pages_addr) {
+			/* set page commands needed */
+			if (bo->shadow)
+				amdgpu_vm_sdma_set_ptes(p, bo->shadow, pe, addr,
+							count, incr, flags);
+			amdgpu_vm_sdma_set_ptes(p, bo, pe, addr, count,
+						incr, flags);
+			return 0;
+		}
+
+		/* copy commands needed */
+		ndw -= p->adev->vm_manager.vm_pte_funcs->copy_pte_num_dw *
+			(bo->shadow ? 2 : 1);
+
+		/* for padding */
+		ndw -= 7;
+
+		nptes = min(count, ndw / 2);
+
+		/* Put the PTEs at the end of the IB. */
+		p->num_dw_left -= nptes * 2;
+		pte = (uint64_t *)&(p->job->ibs->ptr[p->num_dw_left]);
+		for (i = 0; i < nptes; ++i, addr += incr) {
+			pte[i] = amdgpu_vm_map_gart(p->pages_addr, addr);
+			pte[i] |= flags;
+		}
+
+		if (bo->shadow)
+			amdgpu_vm_sdma_copy_ptes(p, bo->shadow, pe, nptes);
+		amdgpu_vm_sdma_copy_ptes(p, bo, pe, nptes);
+
+		pe += nptes * 8;
+		count -= nptes;
+	} while (count);
+
+	return 0;
+}
+
+const struct amdgpu_vm_update_funcs amdgpu_vm_sdma_funcs = {
+	.map_table = amdgpu_vm_sdma_map_table,
+	.prepare = amdgpu_vm_sdma_prepare,
+	.update = amdgpu_vm_sdma_update,
+	.commit = amdgpu_vm_sdma_commit
+};
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c
@ -32,6 +32,85 @@ struct amdgpu_vram_mgr {
 	atomic64_t vis_usage;
 };

+/**
+ * DOC: mem_info_vram_total
+ *
+ * The amdgpu driver provides a sysfs API for reporting current total VRAM
+ * available on the device
+ * The file mem_info_vram_total is used for this and returns the total
+ * amount of VRAM in bytes
+ */
+static ssize_t amdgpu_mem_info_vram_total_show(struct device *dev,
+		struct device_attribute *attr, char *buf)
+{
+	struct drm_device *ddev = dev_get_drvdata(dev);
+	struct amdgpu_device *adev = ddev->dev_private;
+
+	return snprintf(buf, PAGE_SIZE, "%llu\n", adev->gmc.real_vram_size);
+}
+
+/**
+ * DOC: mem_info_vis_vram_total
+ *
+ * The amdgpu driver provides a sysfs API for reporting current total
+ * visible VRAM available on the device
+ * The file mem_info_vis_vram_total is used for this and returns the total
+ * amount of visible VRAM in bytes
+ */
+static ssize_t amdgpu_mem_info_vis_vram_total_show(struct device *dev,
+		struct device_attribute *attr, char *buf)
+{
+	struct drm_device *ddev = dev_get_drvdata(dev);
+	struct amdgpu_device *adev = ddev->dev_private;
+
+	return snprintf(buf, PAGE_SIZE, "%llu\n", adev->gmc.visible_vram_size);
+}
+
+/**
+ * DOC: mem_info_vram_used
+ *
+ * The amdgpu driver provides a sysfs API for reporting current total VRAM
+ * available on the device
+ * The file mem_info_vram_used is used for this and returns the total
+ * amount of currently used VRAM in bytes
+ */
+static ssize_t amdgpu_mem_info_vram_used_show(struct device *dev,
+		struct device_attribute *attr, char *buf)
+{
+	struct drm_device *ddev = dev_get_drvdata(dev);
+	struct amdgpu_device *adev = ddev->dev_private;
+
+	return snprintf(buf, PAGE_SIZE, "%llu\n",
+		amdgpu_vram_mgr_usage(&adev->mman.bdev.man[TTM_PL_VRAM]));
+}
+
+/**
+ * DOC: mem_info_vis_vram_used
+ *
+ * The amdgpu driver provides a sysfs API for reporting current total of
+ * used visible VRAM
+ * The file mem_info_vis_vram_used is used for this and returns the total
+ * amount of currently used visible VRAM in bytes
+ */
+static ssize_t amdgpu_mem_info_vis_vram_used_show(struct device *dev,
+		struct device_attribute *attr, char *buf)
+{
+	struct drm_device *ddev = dev_get_drvdata(dev);
+	struct amdgpu_device *adev = ddev->dev_private;
+
+	return snprintf(buf, PAGE_SIZE, "%llu\n",
+		amdgpu_vram_mgr_vis_usage(&adev->mman.bdev.man[TTM_PL_VRAM]));
+}
+
+static DEVICE_ATTR(mem_info_vram_total, S_IRUGO,
+		   amdgpu_mem_info_vram_total_show, NULL);
+static DEVICE_ATTR(mem_info_vis_vram_total, S_IRUGO,
+		   amdgpu_mem_info_vis_vram_total_show,NULL);
+static DEVICE_ATTR(mem_info_vram_used, S_IRUGO,
+		   amdgpu_mem_info_vram_used_show, NULL);
+static DEVICE_ATTR(mem_info_vis_vram_used, S_IRUGO,
+		   amdgpu_mem_info_vis_vram_used_show, NULL);
+
 /**
 * amdgpu_vram_mgr_init - init VRAM manager and DRM MM
 *
@ -43,7 +122,9 @@ struct amdgpu_vram_mgr {
 static int amdgpu_vram_mgr_init(struct ttm_mem_type_manager *man,
 				unsigned long p_size)
 {
+	struct amdgpu_device *adev = amdgpu_ttm_adev(man->bdev);
 	struct amdgpu_vram_mgr *mgr;
+	int ret;

 	mgr = kzalloc(sizeof(*mgr), GFP_KERNEL);
 	if (!mgr)
@ -52,6 +133,29 @@ static int amdgpu_vram_mgr_init(struct ttm_mem_type_manager *man,
 	drm_mm_init(&mgr->mm, 0, p_size);
 	spin_lock_init(&mgr->lock);
 	man->priv = mgr;
+
+	/* Add the two VRAM-related sysfs files */
+	ret = device_create_file(adev->dev, &dev_attr_mem_info_vram_total);
+	if (ret) {
+		DRM_ERROR("Failed to create device file mem_info_vram_total\n");
+		return ret;
+	}
+	ret = device_create_file(adev->dev, &dev_attr_mem_info_vis_vram_total);
+	if (ret) {
+		DRM_ERROR("Failed to create device file mem_info_vis_vram_total\n");
+		return ret;
+	}
+	ret = device_create_file(adev->dev, &dev_attr_mem_info_vram_used);
+	if (ret) {
+		DRM_ERROR("Failed to create device file mem_info_vram_used\n");
+		return ret;
+	}
+	ret = device_create_file(adev->dev, &dev_attr_mem_info_vis_vram_used);
+	if (ret) {
+		DRM_ERROR("Failed to create device file mem_info_vis_vram_used\n");
+		return ret;
+	}
+
 	return 0;
 }

@ -65,6 +169,7 @@ static int amdgpu_vram_mgr_init(struct ttm_mem_type_manager *man,
 */
 static int amdgpu_vram_mgr_fini(struct ttm_mem_type_manager *man)
 {
+	struct amdgpu_device *adev = amdgpu_ttm_adev(man->bdev);
 	struct amdgpu_vram_mgr *mgr = man->priv;

 	spin_lock(&mgr->lock);
@ -72,6 +177,10 @@ static int amdgpu_vram_mgr_fini(struct ttm_mem_type_manager *man)
 	spin_unlock(&mgr->lock);
 	kfree(mgr);
 	man->priv = NULL;
+	device_remove_file(adev->dev, &dev_attr_mem_info_vram_total);
+	device_remove_file(adev->dev, &dev_attr_mem_info_vis_vram_total);
+	device_remove_file(adev->dev, &dev_attr_mem_info_vram_used);
+	device_remove_file(adev->dev, &dev_attr_mem_info_vis_vram_used);
 	return 0;
 }

--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c
@ -24,6 +24,7 @@
 #include <linux/list.h>
 #include "amdgpu.h"
 #include "amdgpu_xgmi.h"
+#include "amdgpu_smu.h"


 static DEFINE_MUTEX(xgmi_mutex);
@ -34,12 +35,132 @@ static DEFINE_MUTEX(xgmi_mutex);
 static struct amdgpu_hive_info xgmi_hives[AMDGPU_MAX_XGMI_HIVE];
 static unsigned hive_count = 0;

-
 void *amdgpu_xgmi_hive_try_lock(struct amdgpu_hive_info *hive)
 {
 	return &hive->device_list;
 }

+static ssize_t amdgpu_xgmi_show_hive_id(struct device *dev,
+		struct device_attribute *attr, char *buf)
+{
+	struct amdgpu_hive_info *hive =
+			container_of(attr, struct amdgpu_hive_info, dev_attr);
+
+	return snprintf(buf, PAGE_SIZE, "%llu\n", hive->hive_id);
+}
+
+static int amdgpu_xgmi_sysfs_create(struct amdgpu_device *adev,
+				    struct amdgpu_hive_info *hive)
+{
+	int ret = 0;
+
+	if (WARN_ON(hive->kobj))
+		return -EINVAL;
+
+	hive->kobj = kobject_create_and_add("xgmi_hive_info", &adev->dev->kobj);
+	if (!hive->kobj) {
+		dev_err(adev->dev, "XGMI: Failed to allocate sysfs entry!\n");
+		return -EINVAL;
+	}
+
+	hive->dev_attr = (struct device_attribute) {
+		.attr = {
+			.name = "xgmi_hive_id",
+			.mode = S_IRUGO,
+
+		},
+		.show = amdgpu_xgmi_show_hive_id,
+	};
+
+	ret = sysfs_create_file(hive->kobj, &hive->dev_attr.attr);
+	if (ret) {
+		dev_err(adev->dev, "XGMI: Failed to create device file xgmi_hive_id\n");
+		kobject_del(hive->kobj);
+		kobject_put(hive->kobj);
+		hive->kobj = NULL;
+	}
+
+	return ret;
+}
+
+static void amdgpu_xgmi_sysfs_destroy(struct amdgpu_device *adev,
+				    struct amdgpu_hive_info *hive)
+{
+	sysfs_remove_file(hive->kobj, &hive->dev_attr.attr);
+	kobject_del(hive->kobj);
+	kobject_put(hive->kobj);
+	hive->kobj = NULL;
+}
+
+static ssize_t amdgpu_xgmi_show_device_id(struct device *dev,
+				     struct device_attribute *attr,
+				     char *buf)
+{
+	struct drm_device *ddev = dev_get_drvdata(dev);
+	struct amdgpu_device *adev = ddev->dev_private;
+
+	return snprintf(buf, PAGE_SIZE, "%llu\n", adev->gmc.xgmi.node_id);
+
+}
+
+
+static DEVICE_ATTR(xgmi_device_id, S_IRUGO, amdgpu_xgmi_show_device_id, NULL);
+
+
+static int amdgpu_xgmi_sysfs_add_dev_info(struct amdgpu_device *adev,
+					 struct amdgpu_hive_info *hive)
+{
+	int ret = 0;
+	char node[10] = { 0 };
+
+	/* Create xgmi device id file */
+	ret = device_create_file(adev->dev, &dev_attr_xgmi_device_id);
+	if (ret) {
+		dev_err(adev->dev, "XGMI: Failed to create device file xgmi_device_id\n");
+		return ret;
+	}
+
+	/* Create sysfs link to hive info folder on the first device */
+	if (adev != hive->adev) {
+		ret = sysfs_create_link(&adev->dev->kobj, hive->kobj,
+					"xgmi_hive_info");
+		if (ret) {
+			dev_err(adev->dev, "XGMI: Failed to create link to hive info");
+			goto remove_file;
+		}
+	}
+
+	sprintf(node, "node%d", hive->number_devices);
+	/* Create sysfs link form the hive folder to yourself */
+	ret = sysfs_create_link(hive->kobj, &adev->dev->kobj, node);
+	if (ret) {
+		dev_err(adev->dev, "XGMI: Failed to create link from hive info");
+		goto remove_link;
+	}
+
+	goto success;
+
+
+remove_link:
+	sysfs_remove_link(&adev->dev->kobj, adev->ddev->unique);
+
+remove_file:
+	device_remove_file(adev->dev, &dev_attr_xgmi_device_id);
+
+success:
+	return ret;
+}
+
+static void amdgpu_xgmi_sysfs_rem_dev_info(struct amdgpu_device *adev,
+					  struct amdgpu_hive_info *hive)
+{
+	device_remove_file(adev->dev, &dev_attr_xgmi_device_id);
+	sysfs_remove_link(&adev->dev->kobj, adev->ddev->unique);
+	sysfs_remove_link(hive->kobj, adev->ddev->unique);
+}
+
+
+
 struct amdgpu_hive_info *amdgpu_get_xgmi_hive(struct amdgpu_device *adev, int lock)
 {
 	int i;
@ -66,18 +187,50 @@ struct amdgpu_hive_info *amdgpu_get_xgmi_hive(struct amdgpu_device *adev, int lo

 	/* initialize new hive if not exist */
 	tmp = &xgmi_hives[hive_count++];
+
+	if (amdgpu_xgmi_sysfs_create(adev, tmp)) {
+		mutex_unlock(&xgmi_mutex);
+		return NULL;
+	}
+
+	tmp->adev = adev;
 	tmp->hive_id = adev->gmc.xgmi.hive_id;
 	INIT_LIST_HEAD(&tmp->device_list);
 	mutex_init(&tmp->hive_lock);
 	mutex_init(&tmp->reset_lock);
+
 	if (lock)
 		mutex_lock(&tmp->hive_lock);
-
+	tmp->pstate = -1;
 	mutex_unlock(&xgmi_mutex);

 	return tmp;
 }

+int amdgpu_xgmi_set_pstate(struct amdgpu_device *adev, int pstate)
+{
+	int ret = 0;
+	struct amdgpu_hive_info *hive = amdgpu_get_xgmi_hive(adev, 0);
+
+	if (!hive)
+		return 0;
+
+	if (hive->pstate == pstate)
+		return 0;
+
+	dev_dbg(adev->dev, "Set xgmi pstate %d.\n", pstate);
+
+	if (is_support_sw_smu(adev))
+		ret = smu_set_xgmi_pstate(&adev->smu, pstate);
+	if (ret)
+		dev_err(adev->dev,
+			"XGMI: Set pstate failure on device %llx, hive %llx, ret %d",
+			adev->gmc.xgmi.node_id,
+			adev->gmc.xgmi.hive_id, ret);
+
+	return ret;
+}
+
 int amdgpu_xgmi_update_topology(struct amdgpu_hive_info *hive, struct amdgpu_device *adev)
 {
 	int ret = -EINVAL;
@ -156,8 +309,17 @@ int amdgpu_xgmi_add_device(struct amdgpu_device *adev)
 			break;
 	}

-	dev_info(adev->dev, "XGMI: Add node %d, hive 0x%llx.\n",
-		 adev->gmc.xgmi.physical_node_id, adev->gmc.xgmi.hive_id);
+	if (!ret)
+		ret = amdgpu_xgmi_sysfs_add_dev_info(adev, hive);
+
+	if (!ret)
+		dev_info(adev->dev, "XGMI: Add node %d, hive 0x%llx.\n",
+			 adev->gmc.xgmi.physical_node_id, adev->gmc.xgmi.hive_id);
+	else
+		dev_err(adev->dev, "XGMI: Failed to add node %d, hive 0x%llx ret: %d\n",
+			adev->gmc.xgmi.physical_node_id, adev->gmc.xgmi.hive_id,
+			ret);
+

 	mutex_unlock(&hive->hive_lock);
 exit:
@ -176,9 +338,11 @@ void amdgpu_xgmi_remove_device(struct amdgpu_device *adev)
 		return;

 	if (!(hive->number_devices--)) {
+		amdgpu_xgmi_sysfs_destroy(adev, hive);
 		mutex_destroy(&hive->hive_lock);
 		mutex_destroy(&hive->reset_lock);
 	} else {
+		amdgpu_xgmi_sysfs_rem_dev_info(adev, hive);
 		mutex_unlock(&hive->hive_lock);
 	}
 }
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.h
@ -29,13 +29,25 @@ struct amdgpu_hive_info {
 	struct list_head	device_list;
 	struct psp_xgmi_topology_info	topology_info;
 	int number_devices;
-	struct mutex hive_lock,
-		     reset_lock;
+	struct mutex hive_lock, reset_lock;
+	struct kobject *kobj;
+	struct device_attribute dev_attr;
+	struct amdgpu_device *adev;
+	int pstate; /*0 -- low , 1 -- high , -1 unknown*/
 };

 struct amdgpu_hive_info *amdgpu_get_xgmi_hive(struct amdgpu_device *adev, int lock);
 int amdgpu_xgmi_update_topology(struct amdgpu_hive_info *hive, struct amdgpu_device *adev);
 int amdgpu_xgmi_add_device(struct amdgpu_device *adev);
 void amdgpu_xgmi_remove_device(struct amdgpu_device *adev);
+int amdgpu_xgmi_set_pstate(struct amdgpu_device *adev, int pstate);
+
+static inline bool amdgpu_xgmi_same_hive(struct amdgpu_device *adev,
+		struct amdgpu_device *bo_adev)
+{
+	return (adev != bo_adev &&
+		adev->gmc.xgmi.hive_id &&
+		adev->gmc.xgmi.hive_id == bo_adev->gmc.xgmi.hive_id);
+}

 #endif
--- a/drivers/gpu/drm/amd/amdgpu/cik_sdma.c
+++ b/drivers/gpu/drm/amd/amdgpu/cik_sdma.c
@ -977,8 +977,8 @@ static int cik_sdma_sw_init(void *handle)
 		r = amdgpu_ring_init(adev, ring, 1024,
 				     &adev->sdma.trap_irq,
 				     (i == 0) ?
-				     AMDGPU_SDMA_IRQ_TRAP0 :
-				     AMDGPU_SDMA_IRQ_TRAP1);
+				     AMDGPU_SDMA_IRQ_INSTANCE0 :
+				     AMDGPU_SDMA_IRQ_INSTANCE1);
 		if (r)
 			return r;
 	}
@ -1114,7 +1114,7 @@ static int cik_sdma_set_trap_irq_state(struct amdgpu_device *adev,
 	u32 sdma_cntl;

 	switch (type) {
-	case AMDGPU_SDMA_IRQ_TRAP0:
+	case AMDGPU_SDMA_IRQ_INSTANCE0:
 		switch (state) {
 		case AMDGPU_IRQ_STATE_DISABLE:
 			sdma_cntl = RREG32(mmSDMA0_CNTL + SDMA0_REGISTER_OFFSET);
@ -1130,7 +1130,7 @@ static int cik_sdma_set_trap_irq_state(struct amdgpu_device *adev,
 			break;
 		}
 		break;
-	case AMDGPU_SDMA_IRQ_TRAP1:
+	case AMDGPU_SDMA_IRQ_INSTANCE1:
 		switch (state) {
 		case AMDGPU_IRQ_STATE_DISABLE:
 			sdma_cntl = RREG32(mmSDMA0_CNTL + SDMA1_REGISTER_OFFSET);
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c
@ -782,6 +782,25 @@ static void gfx_v6_0_tiling_mode_table_init(struct amdgpu_device *adev)
 				BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
 				BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
 				MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2);
+		tilemode[18] =  MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
+				ARRAY_MODE(ARRAY_1D_TILED_THICK) |
+				PIPE_CONFIG(ADDR_SURF_P4_8x16);
+		tilemode[19] =  MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
+				ARRAY_MODE(ARRAY_2D_TILED_XTHICK) |
+				PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+				BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+				BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+				MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) |
+				NUM_BANKS(ADDR_SURF_16_BANK) |
+				TILE_SPLIT(split_equal_to_row_size);
+		tilemode[20] =  MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
+				ARRAY_MODE(ARRAY_2D_TILED_THICK) |
+				PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+				BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+				BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+				MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) |
+				NUM_BANKS(ADDR_SURF_16_BANK) |
+				TILE_SPLIT(split_equal_to_row_size);
 		tilemode[21] =  MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
 				ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
 				PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
@ -3236,6 +3236,7 @@ static void gfx_v8_0_tiling_mode_table_init(struct amdgpu_device *adev)
 		dev_warn(adev->dev,
 			 "Unknown chip type (%d) in function gfx_v8_0_tiling_mode_table_init() falling through to CHIP_CARRIZO\n",
 			 adev->asic_type);
+		/* fall through */

 	case CHIP_CARRIZO:
 		modearray[0] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
@ -40,6 +40,8 @@

 #include "ivsrcid/gfx/irqsrcs_gfx_9_0.h"

+#include "amdgpu_ras.h"
+
 #define GFX9_NUM_GFX_RINGS     1
 #define GFX9_MEC_HPD_SIZE 4096
 #define RLCG_UCODE_LOADING_START_ADDRESS 0x00002000L
@ -576,6 +578,27 @@ static void gfx_v9_0_check_fw_write_wait(struct amdgpu_device *adev)
 	}
 }

+static void gfx_v9_0_check_if_need_gfxoff(struct amdgpu_device *adev)
+{
+	switch (adev->asic_type) {
+	case CHIP_VEGA10:
+	case CHIP_VEGA12:
+	case CHIP_VEGA20:
+		break;
+	case CHIP_RAVEN:
+		if (adev->rev_id >= 0x8 || adev->pdev->device == 0x15d8)
+			break;
+		if ((adev->gfx.rlc_fw_version < 531) ||
+		    (adev->gfx.rlc_fw_version == 53815) ||
+		    (adev->gfx.rlc_feature_version < 1) ||
+		    !adev->gfx.rlc.is_rlc_v2_1)
+			adev->pm.pp_feature &= ~PP_GFXOFF_MASK;
+		break;
+	default:
+		break;
+	}
+}
+
 static int gfx_v9_0_init_microcode(struct amdgpu_device *adev)
 {
 	const char *chip_name;
@ -828,6 +851,7 @@ static int gfx_v9_0_init_microcode(struct amdgpu_device *adev)
 	}

 out:
+	gfx_v9_0_check_if_need_gfxoff(adev);
 	gfx_v9_0_check_fw_write_wait(adev);
 	if (err) {
 		dev_err(adev->dev,
@ -1639,6 +1663,18 @@ static int gfx_v9_0_sw_init(void *handle)
 	if (r)
 		return r;

+	/* ECC error */
+	r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_GRBM_CP, GFX_9_0__SRCID__CP_ECC_ERROR,
+			      &adev->gfx.cp_ecc_error_irq);
+	if (r)
+		return r;
+
+	/* FUE error */
+	r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_GRBM_CP, GFX_9_0__SRCID__CP_FUE_ERROR,
+			      &adev->gfx.cp_ecc_error_irq);
+	if (r)
+		return r;
+
 	adev->gfx.gfx_current_status = AMDGPU_GFX_NORMAL_MODE;

 	gfx_v9_0_scratch_init(adev);
@ -1731,6 +1767,20 @@ static int gfx_v9_0_sw_fini(void *handle)
 	int i;
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;

+	if (amdgpu_ras_is_supported(adev, AMDGPU_RAS_BLOCK__GFX) &&
+			adev->gfx.ras_if) {
+		struct ras_common_if *ras_if = adev->gfx.ras_if;
+		struct ras_ih_if ih_info = {
+			.head = *ras_if,
+		};
+
+		amdgpu_ras_debugfs_remove(adev, ras_if);
+		amdgpu_ras_sysfs_remove(adev, ras_if);
+		amdgpu_ras_interrupt_remove_handler(adev,  &ih_info);
+		amdgpu_ras_feature_enable(adev, ras_if, 0);
+		kfree(ras_if);
+	}
+
 	amdgpu_bo_free_kernel(&adev->gds.oa_gfx_bo, NULL, NULL);
 	amdgpu_bo_free_kernel(&adev->gds.gws_gfx_bo, NULL, NULL);
 	amdgpu_bo_free_kernel(&adev->gds.gds_gfx_bo, NULL, NULL);
@ -3303,6 +3353,7 @@ static int gfx_v9_0_hw_fini(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;

+	amdgpu_irq_put(adev, &adev->gfx.cp_ecc_error_irq, 0);
 	amdgpu_irq_put(adev, &adev->gfx.priv_reg_irq, 0);
 	amdgpu_irq_put(adev, &adev->gfx.priv_inst_irq, 0);

@ -3492,6 +3543,80 @@ static int gfx_v9_0_early_init(void *handle)
 	return 0;
 }

+static int gfx_v9_0_process_ras_data_cb(struct amdgpu_device *adev,
+		struct amdgpu_iv_entry *entry);
+
+static int gfx_v9_0_ecc_late_init(void *handle)
+{
+	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+	struct ras_common_if **ras_if = &adev->gfx.ras_if;
+	struct ras_ih_if ih_info = {
+		.cb = gfx_v9_0_process_ras_data_cb,
+	};
+	struct ras_fs_if fs_info = {
+		.sysfs_name = "gfx_err_count",
+		.debugfs_name = "gfx_err_inject",
+	};
+	struct ras_common_if ras_block = {
+		.block = AMDGPU_RAS_BLOCK__GFX,
+		.type = AMDGPU_RAS_ERROR__MULTI_UNCORRECTABLE,
+		.sub_block_index = 0,
+		.name = "gfx",
+	};
+	int r;
+
+	if (!amdgpu_ras_is_supported(adev, AMDGPU_RAS_BLOCK__GFX)) {
+		amdgpu_ras_feature_enable_on_boot(adev, &ras_block, 0);
+		return 0;
+	}
+
+	if (*ras_if)
+		goto resume;
+
+	*ras_if = kmalloc(sizeof(**ras_if), GFP_KERNEL);
+	if (!*ras_if)
+		return -ENOMEM;
+
+	**ras_if = ras_block;
+
+	r = amdgpu_ras_feature_enable_on_boot(adev, *ras_if, 1);
+	if (r)
+		goto feature;
+
+	ih_info.head = **ras_if;
+	fs_info.head = **ras_if;
+
+	r = amdgpu_ras_interrupt_add_handler(adev, &ih_info);
+	if (r)
+		goto interrupt;
+
+	r = amdgpu_ras_debugfs_create(adev, &fs_info);
+	if (r)
+		goto debugfs;
+
+	r = amdgpu_ras_sysfs_create(adev, &fs_info);
+	if (r)
+		goto sysfs;
+resume:
+	r = amdgpu_irq_get(adev, &adev->gfx.cp_ecc_error_irq, 0);
+	if (r)
+		goto irq;
+
+	return 0;
+irq:
+	amdgpu_ras_sysfs_remove(adev, *ras_if);
+sysfs:
+	amdgpu_ras_debugfs_remove(adev, *ras_if);
+debugfs:
+	amdgpu_ras_interrupt_remove_handler(adev, &ih_info);
+interrupt:
+	amdgpu_ras_feature_enable(adev, *ras_if, 0);
+feature:
+	kfree(*ras_if);
+	*ras_if = NULL;
+	return -EINVAL;
+}
+
 static int gfx_v9_0_late_init(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
@ -3505,6 +3630,10 @@ static int gfx_v9_0_late_init(void *handle)
 	if (r)
 		return r;

+	r = gfx_v9_0_ecc_late_init(handle);
+	if (r)
+		return r;
+
 	return 0;
 }

@ -4541,6 +4670,45 @@ static int gfx_v9_0_set_priv_inst_fault_state(struct amdgpu_device *adev,
 	return 0;
 }

+#define ENABLE_ECC_ON_ME_PIPE(me, pipe)				\
+	WREG32_FIELD15(GC, 0, CP_ME##me##_PIPE##pipe##_INT_CNTL,\
+			CP_ECC_ERROR_INT_ENABLE, 1)
+
+#define DISABLE_ECC_ON_ME_PIPE(me, pipe)			\
+	WREG32_FIELD15(GC, 0, CP_ME##me##_PIPE##pipe##_INT_CNTL,\
+			CP_ECC_ERROR_INT_ENABLE, 0)
+
+static int gfx_v9_0_set_cp_ecc_error_state(struct amdgpu_device *adev,
+					      struct amdgpu_irq_src *source,
+					      unsigned type,
+					      enum amdgpu_interrupt_state state)
+{
+	switch (state) {
+	case AMDGPU_IRQ_STATE_DISABLE:
+		WREG32_FIELD15(GC, 0, CP_INT_CNTL_RING0,
+				CP_ECC_ERROR_INT_ENABLE, 0);
+		DISABLE_ECC_ON_ME_PIPE(1, 0);
+		DISABLE_ECC_ON_ME_PIPE(1, 1);
+		DISABLE_ECC_ON_ME_PIPE(1, 2);
+		DISABLE_ECC_ON_ME_PIPE(1, 3);
+		break;
+
+	case AMDGPU_IRQ_STATE_ENABLE:
+		WREG32_FIELD15(GC, 0, CP_INT_CNTL_RING0,
+				CP_ECC_ERROR_INT_ENABLE, 1);
+		ENABLE_ECC_ON_ME_PIPE(1, 0);
+		ENABLE_ECC_ON_ME_PIPE(1, 1);
+		ENABLE_ECC_ON_ME_PIPE(1, 2);
+		ENABLE_ECC_ON_ME_PIPE(1, 3);
+		break;
+	default:
+		break;
+	}
+
+	return 0;
+}
+
+
 static int gfx_v9_0_set_eop_interrupt_state(struct amdgpu_device *adev,
 					    struct amdgpu_irq_src *src,
 					    unsigned type,
@ -4657,6 +4825,34 @@ static int gfx_v9_0_priv_inst_irq(struct amdgpu_device *adev,
 	return 0;
 }

+static int gfx_v9_0_process_ras_data_cb(struct amdgpu_device *adev,
+		struct amdgpu_iv_entry *entry)
+{
+	/* TODO ue will trigger an interrupt. */
+	kgd2kfd_set_sram_ecc_flag(adev->kfd.dev);
+	amdgpu_ras_reset_gpu(adev, 0);
+	return AMDGPU_RAS_UE;
+}
+
+static int gfx_v9_0_cp_ecc_error_irq(struct amdgpu_device *adev,
+				  struct amdgpu_irq_src *source,
+				  struct amdgpu_iv_entry *entry)
+{
+	struct ras_common_if *ras_if = adev->gfx.ras_if;
+	struct ras_dispatch_if ih_data = {
+		.entry = entry,
+	};
+
+	if (!ras_if)
+		return 0;
+
+	ih_data.head = *ras_if;
+
+	DRM_ERROR("CP ECC ERROR IRQ\n");
+	amdgpu_ras_interrupt_dispatch(adev, &ih_data);
+	return 0;
+}
+
 static const struct amd_ip_funcs gfx_v9_0_ip_funcs = {
 	.name = "gfx_v9_0",
 	.early_init = gfx_v9_0_early_init,
@ -4818,6 +5014,12 @@ static const struct amdgpu_irq_src_funcs gfx_v9_0_priv_inst_irq_funcs = {
 	.process = gfx_v9_0_priv_inst_irq,
 };

+static const struct amdgpu_irq_src_funcs gfx_v9_0_cp_ecc_error_irq_funcs = {
+	.set = gfx_v9_0_set_cp_ecc_error_state,
+	.process = gfx_v9_0_cp_ecc_error_irq,
+};
+
+
 static void gfx_v9_0_set_irq_funcs(struct amdgpu_device *adev)
 {
 	adev->gfx.eop_irq.num_types = AMDGPU_CP_IRQ_LAST;
@ -4828,6 +5030,9 @@ static void gfx_v9_0_set_irq_funcs(struct amdgpu_device *adev)

 	adev->gfx.priv_inst_irq.num_types = 1;
 	adev->gfx.priv_inst_irq.funcs = &gfx_v9_0_priv_inst_irq_funcs;
+
+	adev->gfx.cp_ecc_error_irq.num_types = 2; /*C5 ECC error and C9 FUE error*/
+	adev->gfx.cp_ecc_error_irq.funcs = &gfx_v9_0_cp_ecc_error_irq_funcs;
 }

 static void gfx_v9_0_set_rlc_funcs(struct amdgpu_device *adev)
--- a/drivers/gpu/drm/amd/amdgpu/gfxhub_v1_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfxhub_v1_0.c
@ -143,7 +143,7 @@ static void gfxhub_v1_0_init_cache_regs(struct amdgpu_device *adev)
 	/* XXX for emulation, Refer to closed source code.*/
 	tmp = REG_SET_FIELD(tmp, VM_L2_CNTL, L2_PDE0_CACHE_TAG_GENERATION_MODE,
 			    0);
-	tmp = REG_SET_FIELD(tmp, VM_L2_CNTL, PDE_FAULT_CLASSIFICATION, 1);
+	tmp = REG_SET_FIELD(tmp, VM_L2_CNTL, PDE_FAULT_CLASSIFICATION, 0);
 	tmp = REG_SET_FIELD(tmp, VM_L2_CNTL, CONTEXT1_IDENTITY_ACCESS_MODE, 1);
 	tmp = REG_SET_FIELD(tmp, VM_L2_CNTL, IDENTITY_MODE_FRAGMENT_SIZE, 0);
 	WREG32_SOC15(GC, 0, mmVM_L2_CNTL, tmp);
@ -236,7 +236,7 @@ static void gfxhub_v1_0_setup_vmid_config(struct amdgpu_device *adev)
 				    block_size);
 		/* Send no-retry XNACK on fault to suppress VM fault storm. */
 		tmp = REG_SET_FIELD(tmp, VM_CONTEXT1_CNTL,
-				    RETRY_PERMISSION_OR_INVALID_PAGE_FAULT, 0);
+				    RETRY_PERMISSION_OR_INVALID_PAGE_FAULT, 1);
 		WREG32_SOC15_OFFSET(GC, 0, mmVM_CONTEXT1_CNTL, i, tmp);
 		WREG32_SOC15_OFFSET(GC, 0, mmVM_CONTEXT1_PAGE_TABLE_START_ADDR_LO32, i*2, 0);
 		WREG32_SOC15_OFFSET(GC, 0, mmVM_CONTEXT1_PAGE_TABLE_START_ADDR_HI32, i*2, 0);
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c
@ -225,7 +225,7 @@ static void gmc_v6_0_vram_gtt_location(struct amdgpu_device *adev,
 	u64 base = RREG32(mmMC_VM_FB_LOCATION) & 0xFFFF;
 	base <<= 24;

-	amdgpu_gmc_vram_location(adev, &adev->gmc, base);
+	amdgpu_gmc_vram_location(adev, mc, base);
 	amdgpu_gmc_gart_location(adev, mc);
 }

@ -383,20 +383,6 @@ static uint64_t gmc_v6_0_emit_flush_gpu_tlb(struct amdgpu_ring *ring,
 	return pd_addr;
 }

-static int gmc_v6_0_set_pte_pde(struct amdgpu_device *adev, void *cpu_pt_addr,
-				uint32_t gpu_page_idx, uint64_t addr,
-				uint64_t flags)
-{
-	void __iomem *ptr = (void *)cpu_pt_addr;
-	uint64_t value;
-
-	value = addr & 0xFFFFFFFFFFFFF000ULL;
-	value |= flags;
-	writeq(value, ptr + (gpu_page_idx * 8));
-
-	return 0;
-}
-
 static uint64_t gmc_v6_0_get_vm_pte_flags(struct amdgpu_device *adev,
 					  uint32_t flags)
 {
@ -886,7 +872,7 @@ static int gmc_v6_0_sw_init(void *handle)
 		pci_set_consistent_dma_mask(adev->pdev, DMA_BIT_MASK(32));
 		dev_warn(adev->dev, "amdgpu: No coherent DMA available.\n");
 	}
-	adev->need_swiotlb = drm_get_max_iomem() > ((u64)1 << dma_bits);
+	adev->need_swiotlb = drm_need_swiotlb(dma_bits);

 	r = gmc_v6_0_init_microcode(adev);
 	if (r) {
@ -1169,7 +1155,6 @@ static const struct amd_ip_funcs gmc_v6_0_ip_funcs = {
 static const struct amdgpu_gmc_funcs gmc_v6_0_gmc_funcs = {
 	.flush_gpu_tlb = gmc_v6_0_flush_gpu_tlb,
 	.emit_flush_gpu_tlb = gmc_v6_0_emit_flush_gpu_tlb,
-	.set_pte_pde = gmc_v6_0_set_pte_pde,
 	.set_prt = gmc_v6_0_set_prt,
 	.get_vm_pde = gmc_v6_0_get_vm_pde,
 	.get_vm_pte_flags = gmc_v6_0_get_vm_pte_flags
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c
@ -242,7 +242,7 @@ static void gmc_v7_0_vram_gtt_location(struct amdgpu_device *adev,
 	u64 base = RREG32(mmMC_VM_FB_LOCATION) & 0xFFFF;
 	base <<= 24;

-	amdgpu_gmc_vram_location(adev, &adev->gmc, base);
+	amdgpu_gmc_vram_location(adev, mc, base);
 	amdgpu_gmc_gart_location(adev, mc);
 }

@ -460,31 +460,6 @@ static void gmc_v7_0_emit_pasid_mapping(struct amdgpu_ring *ring, unsigned vmid,
 	amdgpu_ring_emit_wreg(ring, mmIH_VMID_0_LUT + vmid, pasid);
 }

-/**
- * gmc_v7_0_set_pte_pde - update the page tables using MMIO
- *
- * @adev: amdgpu_device pointer
- * @cpu_pt_addr: cpu address of the page table
- * @gpu_page_idx: entry in the page table to update
- * @addr: dst addr to write into pte/pde
- * @flags: access flags
- *
- * Update the page tables using the CPU.
- */
-static int gmc_v7_0_set_pte_pde(struct amdgpu_device *adev, void *cpu_pt_addr,
-				 uint32_t gpu_page_idx, uint64_t addr,
-				 uint64_t flags)
-{
-	void __iomem *ptr = (void *)cpu_pt_addr;
-	uint64_t value;
-
-	value = addr & 0xFFFFFFFFFFFFF000ULL;
-	value |= flags;
-	writeq(value, ptr + (gpu_page_idx * 8));
-
-	return 0;
-}
-
 static uint64_t gmc_v7_0_get_vm_pte_flags(struct amdgpu_device *adev,
 					  uint32_t flags)
 {
@ -1030,7 +1005,7 @@ static int gmc_v7_0_sw_init(void *handle)
 		pci_set_consistent_dma_mask(adev->pdev, DMA_BIT_MASK(32));
 		pr_warn("amdgpu: No coherent DMA available\n");
 	}
-	adev->need_swiotlb = drm_get_max_iomem() > ((u64)1 << dma_bits);
+	adev->need_swiotlb = drm_need_swiotlb(dma_bits);

 	r = gmc_v7_0_init_microcode(adev);
 	if (r) {
@ -1376,7 +1351,6 @@ static const struct amdgpu_gmc_funcs gmc_v7_0_gmc_funcs = {
 	.flush_gpu_tlb = gmc_v7_0_flush_gpu_tlb,
 	.emit_flush_gpu_tlb = gmc_v7_0_emit_flush_gpu_tlb,
 	.emit_pasid_mapping = gmc_v7_0_emit_pasid_mapping,
-	.set_pte_pde = gmc_v7_0_set_pte_pde,
 	.set_prt = gmc_v7_0_set_prt,
 	.get_vm_pte_flags = gmc_v7_0_get_vm_pte_flags,
 	.get_vm_pde = gmc_v7_0_get_vm_pde
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c
@ -433,7 +433,7 @@ static void gmc_v8_0_vram_gtt_location(struct amdgpu_device *adev,
 		base = RREG32(mmMC_VM_FB_LOCATION) & 0xFFFF;
 	base <<= 24;

-	amdgpu_gmc_vram_location(adev, &adev->gmc, base);
+	amdgpu_gmc_vram_location(adev, mc, base);
 	amdgpu_gmc_gart_location(adev, mc);
 }

@ -662,50 +662,26 @@ static void gmc_v8_0_emit_pasid_mapping(struct amdgpu_ring *ring, unsigned vmid,
 	amdgpu_ring_emit_wreg(ring, mmIH_VMID_0_LUT + vmid, pasid);
 }

-/**
- * gmc_v8_0_set_pte_pde - update the page tables using MMIO
+/*
+ * PTE format on VI:
+ * 63:40 reserved
+ * 39:12 4k physical page base address
+ * 11:7 fragment
+ * 6 write
+ * 5 read
+ * 4 exe
+ * 3 reserved
+ * 2 snooped
+ * 1 system
+ * 0 valid
 *
- * @adev: amdgpu_device pointer
- * @cpu_pt_addr: cpu address of the page table
- * @gpu_page_idx: entry in the page table to update
- * @addr: dst addr to write into pte/pde
- * @flags: access flags
- *
- * Update the page tables using the CPU.
+ * PDE format on VI:
+ * 63:59 block fragment size
+ * 58:40 reserved
+ * 39:1 physical base address of PTE
+ * bits 5:1 must be 0.
+ * 0 valid
 */
-static int gmc_v8_0_set_pte_pde(struct amdgpu_device *adev, void *cpu_pt_addr,
-				uint32_t gpu_page_idx, uint64_t addr,
-				uint64_t flags)
-{
-	void __iomem *ptr = (void *)cpu_pt_addr;
-	uint64_t value;
-
-	/*
-	 * PTE format on VI:
-	 * 63:40 reserved
-	 * 39:12 4k physical page base address
-	 * 11:7 fragment
-	 * 6 write
-	 * 5 read
-	 * 4 exe
-	 * 3 reserved
-	 * 2 snooped
-	 * 1 system
-	 * 0 valid
-	 *
-	 * PDE format on VI:
-	 * 63:59 block fragment size
-	 * 58:40 reserved
-	 * 39:1 physical base address of PTE
-	 * bits 5:1 must be 0.
-	 * 0 valid
-	 */
-	value = addr & 0x000000FFFFFFF000ULL;
-	value |= flags;
-	writeq(value, ptr + (gpu_page_idx * 8));
-
-	return 0;
-}

 static uint64_t gmc_v8_0_get_vm_pte_flags(struct amdgpu_device *adev,
 					  uint32_t flags)
@ -1155,7 +1131,7 @@ static int gmc_v8_0_sw_init(void *handle)
 		pci_set_consistent_dma_mask(adev->pdev, DMA_BIT_MASK(32));
 		pr_warn("amdgpu: No coherent DMA available\n");
 	}
-	adev->need_swiotlb = drm_get_max_iomem() > ((u64)1 << dma_bits);
+	adev->need_swiotlb = drm_need_swiotlb(dma_bits);

 	r = gmc_v8_0_init_microcode(adev);
 	if (r) {
@ -1743,7 +1719,6 @@ static const struct amdgpu_gmc_funcs gmc_v8_0_gmc_funcs = {
 	.flush_gpu_tlb = gmc_v8_0_flush_gpu_tlb,
 	.emit_flush_gpu_tlb = gmc_v8_0_emit_flush_gpu_tlb,
 	.emit_pasid_mapping = gmc_v8_0_emit_pasid_mapping,
-	.set_pte_pde = gmc_v8_0_set_pte_pde,
 	.set_prt = gmc_v8_0_set_prt,
 	.get_vm_pte_flags = gmc_v8_0_get_vm_pte_flags,
 	.get_vm_pde = gmc_v8_0_get_vm_pde
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
@ -47,6 +47,8 @@

 #include "ivsrcid/vmc/irqsrcs_vmc_1_0.h"

+#include "amdgpu_ras.h"
+
 /* add these here since we already include dce12 headers and these are for DCN */
 #define mmHUBP0_DCSURF_PRI_VIEWPORT_DIMENSION                                                          0x055d
 #define mmHUBP0_DCSURF_PRI_VIEWPORT_DIMENSION_BASE_IDX                                                 2
@ -84,121 +86,182 @@ static const struct soc15_reg_golden golden_settings_athub_1_0_0[] =
 	SOC15_REG_GOLDEN_VALUE(ATHUB, 0, mmRPB_ARB_CNTL2, 0x00ff00ff, 0x00080008)
 };

-/* Ecc related register addresses, (BASE + reg offset) */
-/* Universal Memory Controller caps (may be fused). */
-/* UMCCH:UmcLocalCap */
-#define UMCLOCALCAPS_ADDR0	(0x00014306 + 0x00000000)
-#define UMCLOCALCAPS_ADDR1	(0x00014306 + 0x00000800)
-#define UMCLOCALCAPS_ADDR2	(0x00014306 + 0x00001000)
-#define UMCLOCALCAPS_ADDR3	(0x00014306 + 0x00001800)
-#define UMCLOCALCAPS_ADDR4	(0x00054306 + 0x00000000)
-#define UMCLOCALCAPS_ADDR5	(0x00054306 + 0x00000800)
-#define UMCLOCALCAPS_ADDR6	(0x00054306 + 0x00001000)
-#define UMCLOCALCAPS_ADDR7	(0x00054306 + 0x00001800)
-#define UMCLOCALCAPS_ADDR8	(0x00094306 + 0x00000000)
-#define UMCLOCALCAPS_ADDR9	(0x00094306 + 0x00000800)
-#define UMCLOCALCAPS_ADDR10	(0x00094306 + 0x00001000)
-#define UMCLOCALCAPS_ADDR11	(0x00094306 + 0x00001800)
-#define UMCLOCALCAPS_ADDR12	(0x000d4306 + 0x00000000)
-#define UMCLOCALCAPS_ADDR13	(0x000d4306 + 0x00000800)
-#define UMCLOCALCAPS_ADDR14	(0x000d4306 + 0x00001000)
-#define UMCLOCALCAPS_ADDR15	(0x000d4306 + 0x00001800)
-
-/* Universal Memory Controller Channel config. */
-/* UMCCH:UMC_CONFIG */
-#define UMCCH_UMC_CONFIG_ADDR0	(0x00014040 + 0x00000000)
-#define UMCCH_UMC_CONFIG_ADDR1	(0x00014040 + 0x00000800)
-#define UMCCH_UMC_CONFIG_ADDR2	(0x00014040 + 0x00001000)
-#define UMCCH_UMC_CONFIG_ADDR3	(0x00014040 + 0x00001800)
-#define UMCCH_UMC_CONFIG_ADDR4	(0x00054040 + 0x00000000)
-#define UMCCH_UMC_CONFIG_ADDR5	(0x00054040 + 0x00000800)
-#define UMCCH_UMC_CONFIG_ADDR6	(0x00054040 + 0x00001000)
-#define UMCCH_UMC_CONFIG_ADDR7	(0x00054040 + 0x00001800)
-#define UMCCH_UMC_CONFIG_ADDR8	(0x00094040 + 0x00000000)
-#define UMCCH_UMC_CONFIG_ADDR9	(0x00094040 + 0x00000800)
-#define UMCCH_UMC_CONFIG_ADDR10	(0x00094040 + 0x00001000)
-#define UMCCH_UMC_CONFIG_ADDR11	(0x00094040 + 0x00001800)
-#define UMCCH_UMC_CONFIG_ADDR12	(0x000d4040 + 0x00000000)
-#define UMCCH_UMC_CONFIG_ADDR13	(0x000d4040 + 0x00000800)
-#define UMCCH_UMC_CONFIG_ADDR14	(0x000d4040 + 0x00001000)
-#define UMCCH_UMC_CONFIG_ADDR15	(0x000d4040 + 0x00001800)
-
-/* Universal Memory Controller Channel Ecc config. */
-/* UMCCH:EccCtrl */
-#define UMCCH_ECCCTRL_ADDR0	(0x00014053 + 0x00000000)
-#define UMCCH_ECCCTRL_ADDR1	(0x00014053 + 0x00000800)
-#define UMCCH_ECCCTRL_ADDR2	(0x00014053 + 0x00001000)
-#define UMCCH_ECCCTRL_ADDR3	(0x00014053 + 0x00001800)
-#define UMCCH_ECCCTRL_ADDR4	(0x00054053 + 0x00000000)
-#define UMCCH_ECCCTRL_ADDR5	(0x00054053 + 0x00000800)
-#define UMCCH_ECCCTRL_ADDR6	(0x00054053 + 0x00001000)
-#define UMCCH_ECCCTRL_ADDR7	(0x00054053 + 0x00001800)
-#define UMCCH_ECCCTRL_ADDR8	(0x00094053 + 0x00000000)
-#define UMCCH_ECCCTRL_ADDR9	(0x00094053 + 0x00000800)
-#define UMCCH_ECCCTRL_ADDR10	(0x00094053 + 0x00001000)
-#define UMCCH_ECCCTRL_ADDR11	(0x00094053 + 0x00001800)
-#define UMCCH_ECCCTRL_ADDR12	(0x000d4053 + 0x00000000)
-#define UMCCH_ECCCTRL_ADDR13	(0x000d4053 + 0x00000800)
-#define UMCCH_ECCCTRL_ADDR14	(0x000d4053 + 0x00001000)
-#define UMCCH_ECCCTRL_ADDR15	(0x000d4053 + 0x00001800)
-
-static const uint32_t ecc_umclocalcap_addrs[] = {
-	UMCLOCALCAPS_ADDR0,
-	UMCLOCALCAPS_ADDR1,
-	UMCLOCALCAPS_ADDR2,
-	UMCLOCALCAPS_ADDR3,
-	UMCLOCALCAPS_ADDR4,
-	UMCLOCALCAPS_ADDR5,
-	UMCLOCALCAPS_ADDR6,
-	UMCLOCALCAPS_ADDR7,
-	UMCLOCALCAPS_ADDR8,
-	UMCLOCALCAPS_ADDR9,
-	UMCLOCALCAPS_ADDR10,
-	UMCLOCALCAPS_ADDR11,
-	UMCLOCALCAPS_ADDR12,
-	UMCLOCALCAPS_ADDR13,
-	UMCLOCALCAPS_ADDR14,
-	UMCLOCALCAPS_ADDR15,
+static const uint32_t ecc_umc_mcumc_ctrl_addrs[] = {
+	(0x000143c0 + 0x00000000),
+	(0x000143c0 + 0x00000800),
+	(0x000143c0 + 0x00001000),
+	(0x000143c0 + 0x00001800),
+	(0x000543c0 + 0x00000000),
+	(0x000543c0 + 0x00000800),
+	(0x000543c0 + 0x00001000),
+	(0x000543c0 + 0x00001800),
+	(0x000943c0 + 0x00000000),
+	(0x000943c0 + 0x00000800),
+	(0x000943c0 + 0x00001000),
+	(0x000943c0 + 0x00001800),
+	(0x000d43c0 + 0x00000000),
+	(0x000d43c0 + 0x00000800),
+	(0x000d43c0 + 0x00001000),
+	(0x000d43c0 + 0x00001800),
+	(0x001143c0 + 0x00000000),
+	(0x001143c0 + 0x00000800),
+	(0x001143c0 + 0x00001000),
+	(0x001143c0 + 0x00001800),
+	(0x001543c0 + 0x00000000),
+	(0x001543c0 + 0x00000800),
+	(0x001543c0 + 0x00001000),
+	(0x001543c0 + 0x00001800),
+	(0x001943c0 + 0x00000000),
+	(0x001943c0 + 0x00000800),
+	(0x001943c0 + 0x00001000),
+	(0x001943c0 + 0x00001800),
+	(0x001d43c0 + 0x00000000),
+	(0x001d43c0 + 0x00000800),
+	(0x001d43c0 + 0x00001000),
+	(0x001d43c0 + 0x00001800),
 };

-static const uint32_t ecc_umcch_umc_config_addrs[] = {
-	UMCCH_UMC_CONFIG_ADDR0,
-	UMCCH_UMC_CONFIG_ADDR1,
-	UMCCH_UMC_CONFIG_ADDR2,
-	UMCCH_UMC_CONFIG_ADDR3,
-	UMCCH_UMC_CONFIG_ADDR4,
-	UMCCH_UMC_CONFIG_ADDR5,
-	UMCCH_UMC_CONFIG_ADDR6,
-	UMCCH_UMC_CONFIG_ADDR7,
-	UMCCH_UMC_CONFIG_ADDR8,
-	UMCCH_UMC_CONFIG_ADDR9,
-	UMCCH_UMC_CONFIG_ADDR10,
-	UMCCH_UMC_CONFIG_ADDR11,
-	UMCCH_UMC_CONFIG_ADDR12,
-	UMCCH_UMC_CONFIG_ADDR13,
-	UMCCH_UMC_CONFIG_ADDR14,
-	UMCCH_UMC_CONFIG_ADDR15,
+static const uint32_t ecc_umc_mcumc_ctrl_mask_addrs[] = {
+	(0x000143e0 + 0x00000000),
+	(0x000143e0 + 0x00000800),
+	(0x000143e0 + 0x00001000),
+	(0x000143e0 + 0x00001800),
+	(0x000543e0 + 0x00000000),
+	(0x000543e0 + 0x00000800),
+	(0x000543e0 + 0x00001000),
+	(0x000543e0 + 0x00001800),
+	(0x000943e0 + 0x00000000),
+	(0x000943e0 + 0x00000800),
+	(0x000943e0 + 0x00001000),
+	(0x000943e0 + 0x00001800),
+	(0x000d43e0 + 0x00000000),
+	(0x000d43e0 + 0x00000800),
+	(0x000d43e0 + 0x00001000),
+	(0x000d43e0 + 0x00001800),
+	(0x001143e0 + 0x00000000),
+	(0x001143e0 + 0x00000800),
+	(0x001143e0 + 0x00001000),
+	(0x001143e0 + 0x00001800),
+	(0x001543e0 + 0x00000000),
+	(0x001543e0 + 0x00000800),
+	(0x001543e0 + 0x00001000),
+	(0x001543e0 + 0x00001800),
+	(0x001943e0 + 0x00000000),
+	(0x001943e0 + 0x00000800),
+	(0x001943e0 + 0x00001000),
+	(0x001943e0 + 0x00001800),
+	(0x001d43e0 + 0x00000000),
+	(0x001d43e0 + 0x00000800),
+	(0x001d43e0 + 0x00001000),
+	(0x001d43e0 + 0x00001800),
 };

-static const uint32_t ecc_umcch_eccctrl_addrs[] = {
-	UMCCH_ECCCTRL_ADDR0,
-	UMCCH_ECCCTRL_ADDR1,
-	UMCCH_ECCCTRL_ADDR2,
-	UMCCH_ECCCTRL_ADDR3,
-	UMCCH_ECCCTRL_ADDR4,
-	UMCCH_ECCCTRL_ADDR5,
-	UMCCH_ECCCTRL_ADDR6,
-	UMCCH_ECCCTRL_ADDR7,
-	UMCCH_ECCCTRL_ADDR8,
-	UMCCH_ECCCTRL_ADDR9,
-	UMCCH_ECCCTRL_ADDR10,
-	UMCCH_ECCCTRL_ADDR11,
-	UMCCH_ECCCTRL_ADDR12,
-	UMCCH_ECCCTRL_ADDR13,
-	UMCCH_ECCCTRL_ADDR14,
-	UMCCH_ECCCTRL_ADDR15,
+static const uint32_t ecc_umc_mcumc_status_addrs[] = {
+	(0x000143c2 + 0x00000000),
+	(0x000143c2 + 0x00000800),
+	(0x000143c2 + 0x00001000),
+	(0x000143c2 + 0x00001800),
+	(0x000543c2 + 0x00000000),
+	(0x000543c2 + 0x00000800),
+	(0x000543c2 + 0x00001000),
+	(0x000543c2 + 0x00001800),
+	(0x000943c2 + 0x00000000),
+	(0x000943c2 + 0x00000800),
+	(0x000943c2 + 0x00001000),
+	(0x000943c2 + 0x00001800),
+	(0x000d43c2 + 0x00000000),
+	(0x000d43c2 + 0x00000800),
+	(0x000d43c2 + 0x00001000),
+	(0x000d43c2 + 0x00001800),
+	(0x001143c2 + 0x00000000),
+	(0x001143c2 + 0x00000800),
+	(0x001143c2 + 0x00001000),
+	(0x001143c2 + 0x00001800),
+	(0x001543c2 + 0x00000000),
+	(0x001543c2 + 0x00000800),
+	(0x001543c2 + 0x00001000),
+	(0x001543c2 + 0x00001800),
+	(0x001943c2 + 0x00000000),
+	(0x001943c2 + 0x00000800),
+	(0x001943c2 + 0x00001000),
+	(0x001943c2 + 0x00001800),
+	(0x001d43c2 + 0x00000000),
+	(0x001d43c2 + 0x00000800),
+	(0x001d43c2 + 0x00001000),
+	(0x001d43c2 + 0x00001800),
 };

+static int gmc_v9_0_ecc_interrupt_state(struct amdgpu_device *adev,
+		struct amdgpu_irq_src *src,
+		unsigned type,
+		enum amdgpu_interrupt_state state)
+{
+	u32 bits, i, tmp, reg;
+
+	bits = 0x7f;
+
+	switch (state) {
+	case AMDGPU_IRQ_STATE_DISABLE:
+		for (i = 0; i < ARRAY_SIZE(ecc_umc_mcumc_ctrl_addrs); i++) {
+			reg = ecc_umc_mcumc_ctrl_addrs[i];
+			tmp = RREG32(reg);
+			tmp &= ~bits;
+			WREG32(reg, tmp);
+		}
+		for (i = 0; i < ARRAY_SIZE(ecc_umc_mcumc_ctrl_mask_addrs); i++) {
+			reg = ecc_umc_mcumc_ctrl_mask_addrs[i];
+			tmp = RREG32(reg);
+			tmp &= ~bits;
+			WREG32(reg, tmp);
+		}
+		break;
+	case AMDGPU_IRQ_STATE_ENABLE:
+		for (i = 0; i < ARRAY_SIZE(ecc_umc_mcumc_ctrl_addrs); i++) {
+			reg = ecc_umc_mcumc_ctrl_addrs[i];
+			tmp = RREG32(reg);
+			tmp |= bits;
+			WREG32(reg, tmp);
+		}
+		for (i = 0; i < ARRAY_SIZE(ecc_umc_mcumc_ctrl_mask_addrs); i++) {
+			reg = ecc_umc_mcumc_ctrl_mask_addrs[i];
+			tmp = RREG32(reg);
+			tmp |= bits;
+			WREG32(reg, tmp);
+		}
+		break;
+	default:
+		break;
+	}
+
+	return 0;
+}
+
+static int gmc_v9_0_process_ras_data_cb(struct amdgpu_device *adev,
+		struct amdgpu_iv_entry *entry)
+{
+	kgd2kfd_set_sram_ecc_flag(adev->kfd.dev);
+	amdgpu_ras_reset_gpu(adev, 0);
+	return AMDGPU_RAS_UE;
+}
+
+static int gmc_v9_0_process_ecc_irq(struct amdgpu_device *adev,
+		struct amdgpu_irq_src *source,
+		struct amdgpu_iv_entry *entry)
+{
+	struct ras_common_if *ras_if = adev->gmc.ras_if;
+	struct ras_dispatch_if ih_data = {
+		.entry = entry,
+	};
+
+	if (!ras_if)
+		return 0;
+
+	ih_data.head = *ras_if;
+
+	amdgpu_ras_interrupt_dispatch(adev, &ih_data);
+	return 0;
+}
+
 static int gmc_v9_0_vm_fault_interrupt_state(struct amdgpu_device *adev,
 					struct amdgpu_irq_src *src,
 					unsigned type,
@ -244,62 +307,6 @@ static int gmc_v9_0_vm_fault_interrupt_state(struct amdgpu_device *adev,
 	return 0;
 }

-/**
- * vega10_ih_prescreen_iv - prescreen an interrupt vector
- *
- * @adev: amdgpu_device pointer
- *
- * Returns true if the interrupt vector should be further processed.
- */
-static bool gmc_v9_0_prescreen_iv(struct amdgpu_device *adev,
-				  struct amdgpu_iv_entry *entry,
-				  uint64_t addr)
-{
-	struct amdgpu_vm *vm;
-	u64 key;
-	int r;
-
-	/* No PASID, can't identify faulting process */
-	if (!entry->pasid)
-		return true;
-
-	/* Not a retry fault */
-	if (!(entry->src_data[1] & 0x80))
-		return true;
-
-	/* Track retry faults in per-VM fault FIFO. */
-	spin_lock(&adev->vm_manager.pasid_lock);
-	vm = idr_find(&adev->vm_manager.pasid_idr, entry->pasid);
-	if (!vm) {
-		/* VM not found, process it normally */
-		spin_unlock(&adev->vm_manager.pasid_lock);
-		return true;
-	}
-
-	key = AMDGPU_VM_FAULT(entry->pasid, addr);
-	r = amdgpu_vm_add_fault(vm->fault_hash, key);
-
-	/* Hash table is full or the fault is already being processed,
-	 * ignore further page faults
-	 */
-	if (r != 0) {
-		spin_unlock(&adev->vm_manager.pasid_lock);
-		return false;
-	}
-	/* No locking required with single writer and single reader */
-	r = kfifo_put(&vm->faults, key);
-	if (!r) {
-		/* FIFO is full. Ignore it until there is space */
-		amdgpu_vm_clear_fault(vm->fault_hash, key);
-		spin_unlock(&adev->vm_manager.pasid_lock);
-		return false;
-	}
-
-	spin_unlock(&adev->vm_manager.pasid_lock);
-	/* It's the first fault for this address, process it normally */
-	return true;
-}
-
 static int gmc_v9_0_process_interrupt(struct amdgpu_device *adev,
 				struct amdgpu_irq_src *source,
 				struct amdgpu_iv_entry *entry)
@ -312,9 +319,11 @@ static int gmc_v9_0_process_interrupt(struct amdgpu_device *adev,
 	addr = (u64)entry->src_data[0] << 12;
 	addr |= ((u64)entry->src_data[1] & 0xf) << 44;

-	if (!gmc_v9_0_prescreen_iv(adev, entry, addr))
+	if (retry_fault && amdgpu_gmc_filter_faults(adev, addr, entry->pasid,
+						    entry->timestamp))
 		return 1; /* This also prevents sending it to KFD */

+	/* If it's the first fault for this address, process it normally */
 	if (!amdgpu_sriov_vf(adev)) {
 		status = RREG32(hub->vm_l2_pro_fault_status);
 		WREG32_P(hub->vm_l2_pro_fault_cntl, 1, ~1);
@ -350,10 +359,19 @@ static const struct amdgpu_irq_src_funcs gmc_v9_0_irq_funcs = {
 	.process = gmc_v9_0_process_interrupt,
 };

+
+static const struct amdgpu_irq_src_funcs gmc_v9_0_ecc_funcs = {
+	.set = gmc_v9_0_ecc_interrupt_state,
+	.process = gmc_v9_0_process_ecc_irq,
+};
+
 static void gmc_v9_0_set_irq_funcs(struct amdgpu_device *adev)
 {
 	adev->gmc.vm_fault.num_types = 1;
 	adev->gmc.vm_fault.funcs = &gmc_v9_0_irq_funcs;
+
+	adev->gmc.ecc_irq.num_types = 1;
+	adev->gmc.ecc_irq.funcs = &gmc_v9_0_ecc_funcs;
 }

 static uint32_t gmc_v9_0_get_invalidate_req(unsigned int vmid,
@ -466,64 +484,37 @@ static void gmc_v9_0_emit_pasid_mapping(struct amdgpu_ring *ring, unsigned vmid,
 	amdgpu_ring_emit_wreg(ring, reg, pasid);
 }

-/**
- * gmc_v9_0_set_pte_pde - update the page tables using MMIO
+/*
+ * PTE format on VEGA 10:
+ * 63:59 reserved
+ * 58:57 mtype
+ * 56 F
+ * 55 L
+ * 54 P
+ * 53 SW
+ * 52 T
+ * 50:48 reserved
+ * 47:12 4k physical page base address
+ * 11:7 fragment
+ * 6 write
+ * 5 read
+ * 4 exe
+ * 3 Z
+ * 2 snooped
+ * 1 system
+ * 0 valid
 *
- * @adev: amdgpu_device pointer
- * @cpu_pt_addr: cpu address of the page table
- * @gpu_page_idx: entry in the page table to update
- * @addr: dst addr to write into pte/pde
- * @flags: access flags
- *
- * Update the page tables using the CPU.
+ * PDE format on VEGA 10:
+ * 63:59 block fragment size
+ * 58:55 reserved
+ * 54 P
+ * 53:48 reserved
+ * 47:6 physical base address of PD or PTE
+ * 5:3 reserved
+ * 2 C
+ * 1 system
+ * 0 valid
 */
-static int gmc_v9_0_set_pte_pde(struct amdgpu_device *adev, void *cpu_pt_addr,
-				uint32_t gpu_page_idx, uint64_t addr,
-				uint64_t flags)
-{
-	void __iomem *ptr = (void *)cpu_pt_addr;
-	uint64_t value;
-
-	/*
-	 * PTE format on VEGA 10:
-	 * 63:59 reserved
-	 * 58:57 mtype
-	 * 56 F
-	 * 55 L
-	 * 54 P
-	 * 53 SW
-	 * 52 T
-	 * 50:48 reserved
-	 * 47:12 4k physical page base address
-	 * 11:7 fragment
-	 * 6 write
-	 * 5 read
-	 * 4 exe
-	 * 3 Z
-	 * 2 snooped
-	 * 1 system
-	 * 0 valid
-	 *
-	 * PDE format on VEGA 10:
-	 * 63:59 block fragment size
-	 * 58:55 reserved
-	 * 54 P
-	 * 53:48 reserved
-	 * 47:6 physical base address of PD or PTE
-	 * 5:3 reserved
-	 * 2 C
-	 * 1 system
-	 * 0 valid
-	 */
-
-	/*
-	 * The following is for PTE only. GART does not have PDEs.
-	*/
-	value = addr & 0x0000FFFFFFFFF000ULL;
-	value |= flags;
-	writeq(value, ptr + (gpu_page_idx * 8));
-	return 0;
-}

 static uint64_t gmc_v9_0_get_vm_pte_flags(struct amdgpu_device *adev,
 						uint32_t flags)
@ -593,7 +584,6 @@ static const struct amdgpu_gmc_funcs gmc_v9_0_gmc_funcs = {
 	.flush_gpu_tlb = gmc_v9_0_flush_gpu_tlb,
 	.emit_flush_gpu_tlb = gmc_v9_0_emit_flush_gpu_tlb,
 	.emit_pasid_mapping = gmc_v9_0_emit_pasid_mapping,
-	.set_pte_pde = gmc_v9_0_set_pte_pde,
 	.get_vm_pte_flags = gmc_v9_0_get_vm_pte_flags,
 	.get_vm_pde = gmc_v9_0_get_vm_pde
 };
@ -620,85 +610,6 @@ static int gmc_v9_0_early_init(void *handle)
 	return 0;
 }

-static int gmc_v9_0_ecc_available(struct amdgpu_device *adev)
-{
-	uint32_t reg_val;
-	uint32_t reg_addr;
-	uint32_t field_val;
-	size_t i;
-	uint32_t fv2;
-	size_t lost_sheep;
-
-	DRM_DEBUG("ecc: gmc_v9_0_ecc_available()\n");
-
-	lost_sheep = 0;
-	for (i = 0; i < ARRAY_SIZE(ecc_umclocalcap_addrs); ++i) {
-		reg_addr = ecc_umclocalcap_addrs[i];
-		DRM_DEBUG("ecc: "
-			  "UMCCH_UmcLocalCap[%zu]: reg_addr: 0x%08x\n",
-			  i, reg_addr);
-		reg_val = RREG32(reg_addr);
-		field_val = REG_GET_FIELD(reg_val, UMCCH0_0_UmcLocalCap,
-					  EccDis);
-		DRM_DEBUG("ecc: "
-			  "reg_val: 0x%08x, "
-			  "EccDis: 0x%08x, ",
-			  reg_val, field_val);
-		if (field_val) {
-			DRM_ERROR("ecc: UmcLocalCap:EccDis is set.\n");
-			++lost_sheep;
-		}
-	}
-
-	for (i = 0; i < ARRAY_SIZE(ecc_umcch_umc_config_addrs); ++i) {
-		reg_addr = ecc_umcch_umc_config_addrs[i];
-		DRM_DEBUG("ecc: "
-			  "UMCCH0_0_UMC_CONFIG[%zu]: reg_addr: 0x%08x",
-			  i, reg_addr);
-		reg_val = RREG32(reg_addr);
-		field_val = REG_GET_FIELD(reg_val, UMCCH0_0_UMC_CONFIG,
-					  DramReady);
-		DRM_DEBUG("ecc: "
-			  "reg_val: 0x%08x, "
-			  "DramReady: 0x%08x\n",
-			  reg_val, field_val);
-
-		if (!field_val) {
-			DRM_ERROR("ecc: UMC_CONFIG:DramReady is not set.\n");
-			++lost_sheep;
-		}
-	}
-
-	for (i = 0; i < ARRAY_SIZE(ecc_umcch_eccctrl_addrs); ++i) {
-		reg_addr = ecc_umcch_eccctrl_addrs[i];
-		DRM_DEBUG("ecc: "
-			  "UMCCH_EccCtrl[%zu]: reg_addr: 0x%08x, ",
-			  i, reg_addr);
-		reg_val = RREG32(reg_addr);
-		field_val = REG_GET_FIELD(reg_val, UMCCH0_0_EccCtrl,
-					  WrEccEn);
-		fv2 = REG_GET_FIELD(reg_val, UMCCH0_0_EccCtrl,
-				    RdEccEn);
-		DRM_DEBUG("ecc: "
-			  "reg_val: 0x%08x, "
-			  "WrEccEn: 0x%08x, "
-			  "RdEccEn: 0x%08x\n",
-			  reg_val, field_val, fv2);
-
-		if (!field_val) {
-			DRM_DEBUG("ecc: WrEccEn is not set\n");
-			++lost_sheep;
-		}
-		if (!fv2) {
-			DRM_DEBUG("ecc: RdEccEn is not set\n");
-			++lost_sheep;
-		}
-	}
-
-	DRM_DEBUG("ecc: lost_sheep: %zu\n", lost_sheep);
-	return lost_sheep == 0;
-}
-
 static bool gmc_v9_0_keep_stolen_memory(struct amdgpu_device *adev)
 {

@ -751,10 +662,82 @@ static int gmc_v9_0_allocate_vm_inv_eng(struct amdgpu_device *adev)
 	return 0;
 }

+static int gmc_v9_0_ecc_late_init(void *handle)
+{
+	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+	struct ras_common_if **ras_if = &adev->gmc.ras_if;
+	struct ras_ih_if ih_info = {
+		.cb = gmc_v9_0_process_ras_data_cb,
+	};
+	struct ras_fs_if fs_info = {
+		.sysfs_name = "umc_err_count",
+		.debugfs_name = "umc_err_inject",
+	};
+	struct ras_common_if ras_block = {
+		.block = AMDGPU_RAS_BLOCK__UMC,
+		.type = AMDGPU_RAS_ERROR__MULTI_UNCORRECTABLE,
+		.sub_block_index = 0,
+		.name = "umc",
+	};
+	int r;
+
+	if (!amdgpu_ras_is_supported(adev, AMDGPU_RAS_BLOCK__UMC)) {
+		amdgpu_ras_feature_enable_on_boot(adev, &ras_block, 0);
+		return 0;
+	}
+	/* handle resume path. */
+	if (*ras_if)
+		goto resume;
+
+	*ras_if = kmalloc(sizeof(**ras_if), GFP_KERNEL);
+	if (!*ras_if)
+		return -ENOMEM;
+
+	**ras_if = ras_block;
+
+	r = amdgpu_ras_feature_enable_on_boot(adev, *ras_if, 1);
+	if (r)
+		goto feature;
+
+	ih_info.head = **ras_if;
+	fs_info.head = **ras_if;
+
+	r = amdgpu_ras_interrupt_add_handler(adev, &ih_info);
+	if (r)
+		goto interrupt;
+
+	r = amdgpu_ras_debugfs_create(adev, &fs_info);
+	if (r)
+		goto debugfs;
+
+	r = amdgpu_ras_sysfs_create(adev, &fs_info);
+	if (r)
+		goto sysfs;
+resume:
+	r = amdgpu_irq_get(adev, &adev->gmc.ecc_irq, 0);
+	if (r)
+		goto irq;
+
+	return 0;
+irq:
+	amdgpu_ras_sysfs_remove(adev, *ras_if);
+sysfs:
+	amdgpu_ras_debugfs_remove(adev, *ras_if);
+debugfs:
+	amdgpu_ras_interrupt_remove_handler(adev, &ih_info);
+interrupt:
+	amdgpu_ras_feature_enable(adev, *ras_if, 0);
+feature:
+	kfree(*ras_if);
+	*ras_if = NULL;
+	return -EINVAL;
+}
+
+
 static int gmc_v9_0_late_init(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
-	int r;
+	bool r;

 	if (!gmc_v9_0_keep_stolen_memory(adev))
 		amdgpu_bo_late_init(adev);
@ -762,20 +745,36 @@ static int gmc_v9_0_late_init(void *handle)
 	r = gmc_v9_0_allocate_vm_inv_eng(adev);
 	if (r)
 		return r;
+	/* Check if ecc is available */
+	if (!amdgpu_sriov_vf(adev)) {
+		switch (adev->asic_type) {
+		case CHIP_VEGA10:
+		case CHIP_VEGA20:
+			r = amdgpu_atomfirmware_mem_ecc_supported(adev);
+			if (!r) {
+				DRM_INFO("ECC is not present.\n");
+				if (adev->df_funcs->enable_ecc_force_par_wr_rmw)
+					adev->df_funcs->enable_ecc_force_par_wr_rmw(adev, false);
+			} else {
+				DRM_INFO("ECC is active.\n");
+			}

-	if (adev->asic_type == CHIP_VEGA10 && !amdgpu_sriov_vf(adev)) {
-		r = gmc_v9_0_ecc_available(adev);
-		if (r == 1) {
-			DRM_INFO("ECC is active.\n");
-		} else if (r == 0) {
-			DRM_INFO("ECC is not present.\n");
-			adev->df_funcs->enable_ecc_force_par_wr_rmw(adev, false);
-		} else {
-			DRM_ERROR("gmc_v9_0_ecc_available() failed. r: %d\n", r);
-			return r;
+			r = amdgpu_atomfirmware_sram_ecc_supported(adev);
+			if (!r) {
+				DRM_INFO("SRAM ECC is not present.\n");
+			} else {
+				DRM_INFO("SRAM ECC is active.\n");
+			}
+			break;
+		default:
+			break;
 		}
 	}

+	r = gmc_v9_0_ecc_late_init(handle);
+	if (r)
+		return r;
+
 	return amdgpu_irq_get(adev, &adev->gmc.vm_fault, 0);
 }

@ -787,7 +786,7 @@ static void gmc_v9_0_vram_gtt_location(struct amdgpu_device *adev,
 		base = mmhub_v1_0_get_fb_location(adev);
 	/* add the xgmi offset of the physical node */
 	base += adev->gmc.xgmi.physical_node_id * adev->gmc.xgmi.node_segment_size;
-	amdgpu_gmc_vram_location(adev, &adev->gmc, base);
+	amdgpu_gmc_vram_location(adev, mc, base);
 	amdgpu_gmc_gart_location(adev, mc);
 	if (!amdgpu_sriov_vf(adev))
 		amdgpu_gmc_agp_location(adev, mc);
@ -987,6 +986,12 @@ static int gmc_v9_0_sw_init(void *handle)
 	if (r)
 		return r;

+	/* interrupt sent to DF. */
+	r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_DF, 0,
+			&adev->gmc.ecc_irq);
+	if (r)
+		return r;
+
 	/* Set the internal MC address mask
 	 * This is the max address of the GPU's
 	 * internal address space.
@ -1011,7 +1016,7 @@ static int gmc_v9_0_sw_init(void *handle)
 		pci_set_consistent_dma_mask(adev->pdev, DMA_BIT_MASK(32));
 		printk(KERN_WARNING "amdgpu: No coherent DMA available.\n");
 	}
-	adev->need_swiotlb = drm_get_max_iomem() > ((u64)1 << dma_bits);
+	adev->need_swiotlb = drm_need_swiotlb(dma_bits);

 	if (adev->gmc.xgmi.supported) {
 		r = gfxhub_v1_1_get_xgmi_info(adev);
@ -1052,6 +1057,22 @@ static int gmc_v9_0_sw_fini(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;

+	if (amdgpu_ras_is_supported(adev, AMDGPU_RAS_BLOCK__UMC) &&
+			adev->gmc.ras_if) {
+		struct ras_common_if *ras_if = adev->gmc.ras_if;
+		struct ras_ih_if ih_info = {
+			.head = *ras_if,
+		};
+
+		/*remove fs first*/
+		amdgpu_ras_debugfs_remove(adev, ras_if);
+		amdgpu_ras_sysfs_remove(adev, ras_if);
+		/*remove the IH*/
+		amdgpu_ras_interrupt_remove_handler(adev, &ih_info);
+		amdgpu_ras_feature_enable(adev, ras_if, 0);
+		kfree(ras_if);
+	}
+
 	amdgpu_gem_force_release(adev);
 	amdgpu_vm_manager_fini(adev);

@ -1198,6 +1219,7 @@ static int gmc_v9_0_hw_fini(void *handle)
 		return 0;
 	}

+	amdgpu_irq_put(adev, &adev->gmc.ecc_irq, 0);
 	amdgpu_irq_put(adev, &adev->gmc.vm_fault, 0);
 	gmc_v9_0_gart_disable(adev);

--- a/drivers/gpu/drm/amd/amdgpu/kv_dpm.c
+++ b/drivers/gpu/drm/amd/amdgpu/kv_dpm.c
@ -2824,7 +2824,7 @@ static int kv_dpm_init(struct amdgpu_device *adev)
 		pi->caps_tcp_ramping = true;
 	}

-	if (adev->powerplay.pp_feature & PP_SCLK_DEEP_SLEEP_MASK)
+	if (adev->pm.pp_feature & PP_SCLK_DEEP_SLEEP_MASK)
 		pi->caps_sclk_ds = true;
 	else
 		pi->caps_sclk_ds = false;
--- a/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c
@ -163,7 +163,7 @@ static void mmhub_v1_0_init_cache_regs(struct amdgpu_device *adev)
 	/* XXX for emulation, Refer to closed source code.*/
 	tmp = REG_SET_FIELD(tmp, VM_L2_CNTL, L2_PDE0_CACHE_TAG_GENERATION_MODE,
 			    0);
-	tmp = REG_SET_FIELD(tmp, VM_L2_CNTL, PDE_FAULT_CLASSIFICATION, 1);
+	tmp = REG_SET_FIELD(tmp, VM_L2_CNTL, PDE_FAULT_CLASSIFICATION, 0);
 	tmp = REG_SET_FIELD(tmp, VM_L2_CNTL, CONTEXT1_IDENTITY_ACCESS_MODE, 1);
 	tmp = REG_SET_FIELD(tmp, VM_L2_CNTL, IDENTITY_MODE_FRAGMENT_SIZE, 0);
 	WREG32_SOC15(MMHUB, 0, mmVM_L2_CNTL, tmp);
@ -256,7 +256,7 @@ static void mmhub_v1_0_setup_vmid_config(struct amdgpu_device *adev)
 				    block_size);
 		/* Send no-retry XNACK on fault to suppress VM fault storm. */
 		tmp = REG_SET_FIELD(tmp, VM_CONTEXT1_CNTL,
-				    RETRY_PERMISSION_OR_INVALID_PAGE_FAULT, 0);
+				    RETRY_PERMISSION_OR_INVALID_PAGE_FAULT, 1);
 		WREG32_SOC15_OFFSET(MMHUB, 0, mmVM_CONTEXT1_CNTL, i, tmp);
 		WREG32_SOC15_OFFSET(MMHUB, 0, mmVM_CONTEXT1_PAGE_TABLE_START_ADDR_LO32, i*2, 0);
 		WREG32_SOC15_OFFSET(MMHUB, 0, mmVM_CONTEXT1_PAGE_TABLE_START_ADDR_HI32, i*2, 0);
--- a/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c
+++ b/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c
@ -157,6 +157,82 @@ static void xgpu_ai_mailbox_trans_msg (struct amdgpu_device *adev,
 	xgpu_ai_mailbox_set_valid(adev, false);
 }

+static int xgpu_ai_get_pp_clk(struct amdgpu_device *adev, u32 type, char *buf)
+{
+        int r = 0;
+        u32 req, val, size;
+
+        if (!amdgim_is_hwperf(adev) || buf == NULL)
+                return -EBADRQC;
+
+        switch(type) {
+        case PP_SCLK:
+                req = IDH_IRQ_GET_PP_SCLK;
+                break;
+        case PP_MCLK:
+                req = IDH_IRQ_GET_PP_MCLK;
+                break;
+        default:
+                return -EBADRQC;
+        }
+
+        mutex_lock(&adev->virt.dpm_mutex);
+
+        xgpu_ai_mailbox_trans_msg(adev, req, 0, 0, 0);
+
+        r = xgpu_ai_poll_msg(adev, IDH_SUCCESS);
+        if (!r && adev->fw_vram_usage.va != NULL) {
+                val = RREG32_NO_KIQ(
+                        SOC15_REG_OFFSET(NBIO, 0,
+                                         mmBIF_BX_PF0_MAILBOX_MSGBUF_RCV_DW1));
+                size = strnlen((((char *)adev->virt.fw_reserve.p_pf2vf) +
+                                val), PAGE_SIZE);
+
+                if (size < PAGE_SIZE)
+                        strcpy(buf,((char *)adev->virt.fw_reserve.p_pf2vf + val));
+                else
+                        size = 0;
+
+                r = size;
+                goto out;
+        }
+
+        r = xgpu_ai_poll_msg(adev, IDH_FAIL);
+        if(r)
+                pr_info("%s DPM request failed",
+                        (type == PP_SCLK)? "SCLK" : "MCLK");
+
+out:
+        mutex_unlock(&adev->virt.dpm_mutex);
+        return r;
+}
+
+static int xgpu_ai_force_dpm_level(struct amdgpu_device *adev, u32 level)
+{
+        int r = 0;
+        u32 req = IDH_IRQ_FORCE_DPM_LEVEL;
+
+        if (!amdgim_is_hwperf(adev))
+                return -EBADRQC;
+
+        mutex_lock(&adev->virt.dpm_mutex);
+        xgpu_ai_mailbox_trans_msg(adev, req, level, 0, 0);
+
+        r = xgpu_ai_poll_msg(adev, IDH_SUCCESS);
+        if (!r)
+                goto out;
+
+        r = xgpu_ai_poll_msg(adev, IDH_FAIL);
+        if (!r)
+                pr_info("DPM request failed");
+        else
+                pr_info("Mailbox is broken");
+
+out:
+        mutex_unlock(&adev->virt.dpm_mutex);
+        return r;
+}
+
 static int xgpu_ai_send_access_requests(struct amdgpu_device *adev,
 					enum idh_request req)
 {
@ -375,4 +451,6 @@ const struct amdgpu_virt_ops xgpu_ai_virt_ops = {
 	.reset_gpu = xgpu_ai_request_reset,
 	.wait_reset = NULL,
 	.trans_msg = xgpu_ai_mailbox_trans_msg,
+	.get_pp_clk = xgpu_ai_get_pp_clk,
+	.force_dpm_level = xgpu_ai_force_dpm_level,
 };
--- a/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.h
+++ b/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.h
@ -35,6 +35,10 @@ enum idh_request {
 	IDH_REL_GPU_FINI_ACCESS,
 	IDH_REQ_GPU_RESET_ACCESS,

+	IDH_IRQ_FORCE_DPM_LEVEL = 10,
+	IDH_IRQ_GET_PP_SCLK,
+	IDH_IRQ_GET_PP_MCLK,
+
 	IDH_LOG_VF_ERROR       = 200,
 };

@ -43,6 +47,8 @@ enum idh_event {
 	IDH_READY_TO_ACCESS_GPU,
 	IDH_FLR_NOTIFICATION,
 	IDH_FLR_NOTIFICATION_CMPL,
+	IDH_SUCCESS,
+	IDH_FAIL,
 	IDH_EVENT_MAX
 };

--- a/Show More
+++ b/Show More