Merge branch 'next' of git://git.kernel.org/pub/scm/linux/kernel/git/vkoul/dmaengine.git

2025-01-15 02:05:33 +00:00 · 2025-01-14 14:56:49 +11:00 · 2025-01-14 14:56:49 +11:00 · 41363b1d4f
commit 41363b1d4f
parent 6e9ab53e5e 98d187a989
47 changed files with 2059 additions and 375 deletions
--- a/Documentation/devicetree/bindings/dma/adi,axi-dmac.txt
+++ b/Documentation/devicetree/bindings/dma/adi,axi-dmac.txt
@ -1,61 +0,0 @@
-Analog Devices AXI-DMAC DMA controller
-
-Required properties:
- - compatible: Must be "adi,axi-dmac-1.00.a".
- - reg: Specification for the controllers memory mapped register map.
- - interrupts: Specification for the controllers interrupt.
- - clocks: Phandle and specifier to the controllers AXI interface clock
- - #dma-cells: Must be 1.
-
-Required sub-nodes:
- - adi,channels: This sub-node must contain a sub-node for each DMA channel. For
-   the channel sub-nodes the following bindings apply. They must match the
-   configuration options of the peripheral as it was instantiated.
-
-Required properties for adi,channels sub-node:
- - #size-cells: Must be 0
- - #address-cells: Must be 1
-
-Required channel sub-node properties:
- - reg: Which channel this node refers to.
- - adi,source-bus-width,
-   adi,destination-bus-width: Width of the source or destination bus in bits.
- - adi,source-bus-type,
-   adi,destination-bus-type: Type of the source or destination bus. Must be one
-   of the following:
-	0 (AXI_DMAC_TYPE_AXI_MM): Memory mapped AXI interface
-	1 (AXI_DMAC_TYPE_AXI_STREAM): Streaming AXI interface
-	2 (AXI_DMAC_TYPE_AXI_FIFO): FIFO interface
-
-Deprecated optional channel properties:
- - adi,length-width: Width of the DMA transfer length register.
- - adi,cyclic: Must be set if the channel supports hardware cyclic DMA
-   transfers.
- - adi,2d: Must be set if the channel supports hardware 2D DMA transfers.
-
-DMA clients connected to the AXI-DMAC DMA controller must use the format
-described in the dma.txt file using a one-cell specifier. The value of the
-specifier refers to the DMA channel index.
-
-Example:
-
-dma: dma@7c420000 {
-	compatible = "adi,axi-dmac-1.00.a";
-	reg = <0x7c420000 0x10000>;
-	interrupts = <0 57 0>;
-	clocks = <&clkc 16>;
-	#dma-cells = <1>;
-
-	adi,channels {
-		#size-cells = <0>;
-		#address-cells = <1>;
-
-		dma-channel@0 {
-			reg = <0>;
-			adi,source-bus-width = <32>;
-			adi,source-bus-type = <ADI_AXI_DMAC_TYPE_MM_AXI>;
-			adi,destination-bus-width = <64>;
-			adi,destination-bus-type = <ADI_AXI_DMAC_TYPE_FIFO>;
-		};
-	};
-};
--- a/Documentation/devicetree/bindings/dma/adi,axi-dmac.yaml
+++ b/Documentation/devicetree/bindings/dma/adi,axi-dmac.yaml
@ -0,0 +1,129 @@
+# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
+%YAML 1.2
+---
+$id: http://devicetree.org/schemas/dma/adi,axi-dmac.yaml#
+$schema: http://devicetree.org/meta-schemas/core.yaml#
+
+title: Analog Devices AXI-DMAC DMA controller
+
+description: |
+  FPGA-based DMA controller designed for use with high-speed converter hardware.
+
+  http://analogdevicesinc.github.io/hdl/library/axi_dmac/index.html
+
+maintainers:
+  - Nuno Sa <nuno.sa@analog.com>
+
+additionalProperties: false
+
+properties:
+  compatible:
+    const: adi,axi-dmac-1.00.a
+
+  reg:
+    maxItems: 1
+
+  interrupts:
+    maxItems: 1
+
+  clocks:
+    maxItems: 1
+
+  "#dma-cells":
+    const: 1
+
+  adi,channels:
+    deprecated: true
+    type: object
+    description:
+      This sub-node must contain a sub-node for each DMA channel. This node is
+      only required for IP versions older than 4.3.a and should otherwise be
+      omitted.
+    additionalProperties: false
+
+    properties:
+      "#size-cells":
+        const: 0
+      "#address-cells":
+        const: 1
+
+    patternProperties:
+      "^dma-channel@[0-9a-f]+$":
+        type: object
+        description:
+          DMA channel properties based on HDL compile-time configuration.
+        additionalProperties: false
+
+        properties:
+          reg:
+            maxItems: 1
+
+          adi,source-bus-width:
+            $ref: /schemas/types.yaml#/definitions/uint32
+            description: Width of the source bus in bits.
+            enum: [8, 16, 32, 64, 128]
+
+          adi,destination-bus-width:
+            $ref: /schemas/types.yaml#/definitions/uint32
+            description: Width of the destination bus in bits.
+            enum: [8, 16, 32, 64, 128]
+
+          adi,source-bus-type:
+            $ref: /schemas/types.yaml#/definitions/uint32
+            description: |
+              Type of the source bus.
+
+              0: Memory mapped AXI interface
+              1: Streaming AXI interface
+              2: FIFO interface
+            enum: [0, 1, 2]
+
+          adi,destination-bus-type:
+            $ref: /schemas/types.yaml#/definitions/uint32
+            description: Type of the destination bus (see adi,source-bus-type).
+            enum: [0, 1, 2]
+
+          adi,length-width:
+            deprecated: true
+            $ref: /schemas/types.yaml#/definitions/uint32
+            description: Width of the DMA transfer length register.
+
+          adi,cyclic:
+            deprecated: true
+            type: boolean
+            description:
+              Must be set if the channel supports hardware cyclic DMA transfers.
+
+          adi,2d:
+            deprecated: true
+            type: boolean
+            description:
+              Must be set if the channel supports hardware 2D DMA transfers.
+
+        required:
+          - reg
+          - adi,source-bus-width
+          - adi,destination-bus-width
+          - adi,source-bus-type
+          - adi,destination-bus-type
+
+    required:
+      - "#size-cells"
+      - "#address-cells"
+
+required:
+  - compatible
+  - reg
+  - interrupts
+  - clocks
+  - "#dma-cells"
+
+examples:
+  - |
+    dma-controller@7c420000 {
+        compatible = "adi,axi-dmac-1.00.a";
+        reg = <0x7c420000 0x10000>;
+        interrupts = <0 57 0>;
+        clocks = <&clkc 16>;
+        #dma-cells = <1>;
+    };
--- a/Documentation/devicetree/bindings/dma/allwinner,sun4i-a10-dma.yaml
+++ b/Documentation/devicetree/bindings/dma/allwinner,sun4i-a10-dma.yaml
@ -22,7 +22,9 @@ properties:
      number.

  compatible:
-    const: allwinner,sun4i-a10-dma
+    enum:
+      - allwinner,sun4i-a10-dma
+      - allwinner,suniv-f1c100s-dma

  reg:
    maxItems: 1
--- a/Documentation/devicetree/bindings/dma/atmel,sama5d4-dma.yaml
+++ b/Documentation/devicetree/bindings/dma/atmel,sama5d4-dma.yaml
@ -0,0 +1,79 @@
+# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
+%YAML 1.2
+---
+$id: http://devicetree.org/schemas/dma/atmel,sama5d4-dma.yaml#
+$schema: http://devicetree.org/meta-schemas/core.yaml#
+
+title: Microchip AT91 Extensible Direct Memory Access Controller
+
+maintainers:
+  - Nicolas Ferre <nicolas.ferre@microchip.com>
+  - Charan Pedumuru <charan.pedumuru@microchip.com>
+
+description:
+  The DMA Controller (XDMAC) is a AHB-protocol central direct memory access
+  controller. It performs peripheral data transfer and memory move operations
+  over one or two bus ports through the unidirectional communication
+  channel. Each channel is fully programmable and provides both peripheral
+  or memory-to-memory transfers. The channel features are configurable at
+  implementation.
+
+allOf:
+  - $ref: dma-controller.yaml#
+
+properties:
+  compatible:
+    oneOf:
+      - enum:
+          - atmel,sama5d4-dma
+          - microchip,sama7g5-dma
+      - items:
+          - enum:
+              - microchip,sam9x60-dma
+              - microchip,sam9x7-dma
+          - const: atmel,sama5d4-dma
+
+  "#dma-cells":
+    description: |
+      Represents the number of integer cells in the `dmas` property of client
+      devices. The single cell specifies the channel configuration register:
+        - bit 13: SIF (Source Interface Identifier) for memory interface.
+        - bit 14: DIF (Destination Interface Identifier) for peripheral interface.
+        - bit 30-24: PERID (Peripheral Identifier).
+    const: 1
+
+  reg:
+    maxItems: 1
+
+  interrupts:
+    maxItems: 1
+
+  clocks:
+    maxItems: 1
+
+  clock-names:
+    const: dma_clk
+
+required:
+  - compatible
+  - reg
+  - interrupts
+  - clocks
+  - clock-names
+  - "#dma-cells"
+
+unevaluatedProperties: false
+
+examples:
+  - |
+    #include <dt-bindings/clock/at91.h>
+    #include <dt-bindings/dma/at91.h>
+    #include <dt-bindings/interrupt-controller/irq.h>
+    dma-controller@f0008000 {
+        compatible = "atmel,sama5d4-dma";
+        reg = <0xf0008000 0x1000>;
+        interrupts = <20 IRQ_TYPE_LEVEL_HIGH 0>;
+        #dma-cells = <1>;
+        clocks = <&pmc PMC_TYPE_PERIPHERAL 20>;
+        clock-names = "dma_clk";
+    };
--- a/Documentation/devicetree/bindings/dma/atmel-xdma.txt
+++ b/Documentation/devicetree/bindings/dma/atmel-xdma.txt
@ -1,54 +0,0 @@
-* Atmel Extensible Direct Memory Access Controller (XDMAC)
-
-* XDMA Controller
-Required properties:
- compatible: Should be "atmel,sama5d4-dma", "microchip,sam9x60-dma" or
-  "microchip,sama7g5-dma" or
-  "microchip,sam9x7-dma", "atmel,sama5d4-dma".
- reg: Should contain DMA registers location and length.
- interrupts: Should contain DMA interrupt.
- #dma-cells: Must be <1>, used to represent the number of integer cells in
-the dmas property of client devices.
-  - The 1st cell specifies the channel configuration register:
-    - bit 13: SIF, source interface identifier, used to get the memory
-    interface identifier,
-    - bit 14: DIF, destination interface identifier, used to get the peripheral
-    interface identifier,
-    - bit 30-24: PERID, peripheral identifier.
-
-Example:
-
-dma1: dma-controller@f0004000 {
-	compatible = "atmel,sama5d4-dma";
-	reg = <0xf0004000 0x200>;
-	interrupts = <50 4 0>;
-	#dma-cells = <1>;
-};
-
-
-* DMA clients
-DMA clients connected to the Atmel XDMA controller must use the format
-described in the dma.txt file, using a one-cell specifier for each channel.
-The two cells in order are:
-1. A phandle pointing to the DMA controller.
-2. Channel configuration register. Configurable fields are:
-    - bit 13: SIF, source interface identifier, used to get the memory
-    interface identifier,
-    - bit 14: DIF, destination interface identifier, used to get the peripheral
-    interface identifier,
-  - bit 30-24: PERID, peripheral identifier.
-
-Example:
-
-i2c2: i2c@f8024000 {
-	compatible = "atmel,at91sam9x5-i2c";
-	reg = <0xf8024000 0x4000>;
-	interrupts = <34 4 6>;
-	dmas = <&dma1
-		(AT91_XDMAC_DT_MEM_IF(0) | AT91_XDMAC_DT_PER_IF(1)
-		 | AT91_XDMAC_DT_PERID(6))>,
-	       <&dma1
-		(AT91_XDMAC_DT_MEM_IF(0) | AT91_XDMAC_DT_PER_IF(1)
-		| AT91_XDMAC_DT_PERID(7))>;
-	dma-names = "tx", "rx";
-};
--- a/Documentation/devicetree/bindings/dma/fsl,edma.yaml
+++ b/Documentation/devicetree/bindings/dma/fsl,edma.yaml
@ -26,9 +26,13 @@ properties:
          - fsl,imx93-edma3
          - fsl,imx93-edma4
          - fsl,imx95-edma5
+          - nxp,s32g2-edma
      - items:
          - const: fsl,ls1028a-edma
          - const: fsl,vf610-edma
+      - items:
+          - const: nxp,s32g3-edma
+          - const: nxp,s32g2-edma

  reg:
    minItems: 1
@ -221,6 +225,36 @@ allOf:
      properties:
        power-domains: false

+  - if:
+      properties:
+        compatible:
+          contains:
+            const: nxp,s32g2-edma
+    then:
+      properties:
+        clocks:
+          minItems: 2
+          maxItems: 2
+        clock-names:
+          items:
+            - const: dmamux0
+            - const: dmamux1
+        interrupts:
+          minItems: 3
+          maxItems: 3
+        interrupt-names:
+          items:
+            - const: tx-0-15
+            - const: tx-16-31
+            - const: err
+        reg:
+          minItems: 3
+          maxItems: 3
+        "#dma-cells":
+          const: 2
+        dma-channels:
+          const: 32
+
 unevaluatedProperties: false

 examples:
--- a/Documentation/devicetree/bindings/dma/nvidia,tegra210-adma.yaml
+++ b/Documentation/devicetree/bindings/dma/nvidia,tegra210-adma.yaml
@ -13,9 +13,6 @@ description: |
 maintainers:
  - Jon Hunter <jonathanh@nvidia.com>

-allOf:
-  - $ref: dma-controller.yaml#
-
 properties:
  compatible:
    oneOf:
@ -29,7 +26,19 @@ properties:
          - const: nvidia,tegra186-adma

  reg:
-    maxItems: 1
+    description:
+      The 'page' region describes the address space of the page
+      used for accessing the DMA channel registers. The 'global'
+      region describes the address space of the global DMA registers.
+      In the absence of the 'reg-names' property, there must be a
+      single entry that covers the address space of the global DMA
+      registers and the DMA channel registers.
+    minItems: 1
+    maxItems: 2
+
+  reg-names:
+    minItems: 1
+    maxItems: 2

  interrupts:
    description: |
@ -63,6 +72,49 @@ required:
  - clocks
  - clock-names

+allOf:
+  - $ref: dma-controller.yaml#
+  - if:
+      properties:
+        compatible:
+          contains:
+            enum:
+              - nvidia,tegra210-adma
+    then:
+      properties:
+        reg:
+          items:
+            - description: Full address space range of DMA registers.
+
+  - if:
+      properties:
+        compatible:
+          contains:
+            enum:
+              - nvidia,tegra186-adma
+    then:
+      anyOf:
+        - properties:
+            reg:
+              items:
+                - description: Full address space range of DMA registers.
+        - properties:
+            reg:
+              items:
+                - description: Channel Page address space range of DMA registers.
+            reg-names:
+              items:
+                - const: page
+        - properties:
+            reg:
+              items:
+                - description: Channel Page address space range of DMA registers.
+                - description: Global Page address space range of DMA registers.
+            reg-names:
+              items:
+                - const: page
+                - const: global
+
 additionalProperties: false

 examples:
--- a/Documentation/devicetree/bindings/dma/qcom,gpi.yaml
+++ b/Documentation/devicetree/bindings/dma/qcom,gpi.yaml
@ -25,7 +25,9 @@ properties:
      - items:
          - enum:
              - qcom,qcm2290-gpi-dma
+              - qcom,qcs8300-gpi-dma
              - qcom,qdu1000-gpi-dma
+              - qcom,sa8775p-gpi-dma
              - qcom,sar2130p-gpi-dma
              - qcom,sc7280-gpi-dma
              - qcom,sdx75-gpi-dma
@ -35,10 +37,12 @@ properties:
              - qcom,sm8450-gpi-dma
              - qcom,sm8550-gpi-dma
              - qcom,sm8650-gpi-dma
+              - qcom,sm8750-gpi-dma
              - qcom,x1e80100-gpi-dma
          - const: qcom,sm6350-gpi-dma
      - items:
          - enum:
+              - qcom,qcs615-gpi-dma
              - qcom,sdm670-gpi-dma
              - qcom,sm6125-gpi-dma
              - qcom,sm8150-gpi-dma
--- a/Documentation/devicetree/bindings/dma/stm32/st,stm32-dmamux.yaml
+++ b/Documentation/devicetree/bindings/dma/stm32/st,stm32-dmamux.yaml
@ -15,6 +15,16 @@ allOf:
 properties:
  "#dma-cells":
    const: 3
+    description: |
+      Each cell represents the following:
+      1. The mux input number/line for the request
+      2. Bitfield representing DMA channel configuration that is passed
+         to the real DMA controller
+      3. Bitfield representing device dependent DMA features passed to
+         the real DMA controller
+
+      For bitfield definitions of cells 2 and 3, see the associated
+      bindings doc for the actual DMA controller in st,stm32-dma.yaml.

  compatible:
    const: st,stm32h7-dmamux
--- a/Documentation/devicetree/bindings/dma/ti/k3-bcdma.yaml
+++ b/Documentation/devicetree/bindings/dma/ti/k3-bcdma.yaml
@ -34,6 +34,7 @@ properties:
      - ti,am62a-dmss-bcdma-csirx
      - ti,am64-dmss-bcdma
      - ti,j721s2-dmss-bcdma-csi
+      - ti,j722s-dmss-bcdma-csi

  reg:
    minItems: 3
@ -196,7 +197,9 @@ allOf:
      properties:
        compatible:
          contains:
-            const: ti,j721s2-dmss-bcdma-csi
+            enum:
+              - ti,j721s2-dmss-bcdma-csi
+              - ti,j722s-dmss-bcdma-csi
    then:
      properties:
        ti,sci-rm-range-bchan: false
--- a/10
+++ b/10
@ -987,6 +987,12 @@ L:	linux-edac@vger.kernel.org
 S:	Supported
 F:	drivers/ras/amd/atl/*

+AMD AE4DMA DRIVER
+M:	Basavaraj Natikar <Basavaraj.Natikar@amd.com>
+L:	dmaengine@vger.kernel.org
+S:	Supported
+F:	drivers/dma/amd/ae4dma/
+
 AMD AXI W1 DRIVER
 M:	Kris Chaplin <kris.chaplin@amd.com>
 R:	Thomas Delev <thomas.delev@amd.com>
@ -1179,8 +1185,8 @@ F:	tools/power/x86/amd_pstate_tracer/amd_pstate_trace.py
 AMD PTDMA DRIVER
 M:	Basavaraj Natikar <Basavaraj.Natikar@amd.com>
 L:	dmaengine@vger.kernel.org
-S:	Maintained
-F:	drivers/dma/ptdma/
+S:	Supported
+F:	drivers/dma/amd/ptdma/

 AMD QDMA DRIVER
 M:	Nishad Saraf <nishads@amd.com>
--- a/drivers/dma/Kconfig
+++ b/drivers/dma/Kconfig
@ -162,8 +162,8 @@ config DMA_SA11X0

 config DMA_SUN4I
 	tristate "Allwinner A10 DMA SoCs support"
-	depends on MACH_SUN4I || MACH_SUN5I || MACH_SUN7I
-	default (MACH_SUN4I || MACH_SUN5I || MACH_SUN7I)
+	depends on MACH_SUN4I || MACH_SUN5I || MACH_SUN7I || MACH_SUNIV
+	default (MACH_SUN4I || MACH_SUN5I || MACH_SUN7I || MACH_SUNIV)
 	select DMA_ENGINE
 	select DMA_VIRTUAL_CHANNELS
 	help
@ -740,8 +740,6 @@ source "drivers/dma/bestcomm/Kconfig"

 source "drivers/dma/mediatek/Kconfig"

-source "drivers/dma/ptdma/Kconfig"
-
 source "drivers/dma/qcom/Kconfig"

 source "drivers/dma/dw/Kconfig"
--- a/drivers/dma/Makefile
+++ b/drivers/dma/Makefile
@ -16,7 +16,6 @@ obj-$(CONFIG_DMATEST) += dmatest.o
 obj-$(CONFIG_ALTERA_MSGDMA) += altera-msgdma.o
 obj-$(CONFIG_AMBA_PL08X) += amba-pl08x.o
 obj-$(CONFIG_AMCC_PPC440SPE_ADMA) += ppc4xx/
-obj-$(CONFIG_AMD_PTDMA) += ptdma/
 obj-$(CONFIG_APPLE_ADMAC) += apple-admac.o
 obj-$(CONFIG_AT_HDMAC) += at_hdmac.o
 obj-$(CONFIG_AT_XDMAC) += at_xdmac.o
--- a/drivers/dma/amd/Kconfig
+++ b/drivers/dma/amd/Kconfig
@ -1,4 +1,32 @@
 # SPDX-License-Identifier: GPL-2.0-only
+#
+
+config AMD_AE4DMA
+	tristate  "AMD AE4DMA Engine"
+	depends on (X86_64 || COMPILE_TEST) && PCI
+	depends on AMD_PTDMA
+	select DMA_ENGINE
+	select DMA_VIRTUAL_CHANNELS
+	help
+	  Enable support for the AMD AE4DMA controller. This controller
+	  provides DMA capabilities to perform high bandwidth memory to
+	  memory and IO copy operations. It performs DMA transfer through
+	  queue-based descriptor management. This DMA controller is intended
+	  to be used with AMD Non-Transparent Bridge devices and not for
+	  general purpose peripheral DMA.
+
+config AMD_PTDMA
+	tristate  "AMD PassThru DMA Engine"
+	depends on X86_64 && PCI
+	select DMA_ENGINE
+	select DMA_VIRTUAL_CHANNELS
+	help
+	  Enable support for the AMD PTDMA controller. This controller
+	  provides DMA capabilities to perform high bandwidth memory to
+	  memory and IO copy operations. It performs DMA transfer through
+	  queue-based descriptor management. This DMA controller is intended
+	  to be used with AMD Non-Transparent Bridge devices and not for
+	  general purpose peripheral DMA.

 config AMD_QDMA
 	tristate "AMD Queue-based DMA"
--- a/drivers/dma/amd/Makefile
+++ b/drivers/dma/amd/Makefile
@ -1,3 +1,5 @@
 # SPDX-License-Identifier: GPL-2.0

+obj-$(CONFIG_AMD_AE4DMA) += ae4dma/
+obj-$(CONFIG_AMD_PTDMA) += ptdma/
 obj-$(CONFIG_AMD_QDMA) += qdma/
--- a/drivers/dma/amd/ae4dma/Makefile
+++ b/drivers/dma/amd/ae4dma/Makefile
@ -0,0 +1,10 @@
+# SPDX-License-Identifier: GPL-2.0
+#
+# AMD AE4DMA driver
+#
+
+obj-$(CONFIG_AMD_AE4DMA) += ae4dma.o
+
+ae4dma-objs := ae4dma-dev.o
+
+ae4dma-$(CONFIG_PCI) += ae4dma-pci.o
--- a/drivers/dma/amd/ae4dma/ae4dma-dev.c
+++ b/drivers/dma/amd/ae4dma/ae4dma-dev.c
@ -0,0 +1,157 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * AMD AE4DMA driver
+ *
+ * Copyright (c) 2024, Advanced Micro Devices, Inc.
+ * All Rights Reserved.
+ *
+ * Author: Basavaraj Natikar <Basavaraj.Natikar@amd.com>
+ */
+
+#include "ae4dma.h"
+
+static unsigned int max_hw_q = 1;
+module_param(max_hw_q, uint, 0444);
+MODULE_PARM_DESC(max_hw_q, "max hw queues supported by engine (any non-zero value, default: 1)");
+
+static void ae4_pending_work(struct work_struct *work)
+{
+	struct ae4_cmd_queue *ae4cmd_q = container_of(work, struct ae4_cmd_queue, p_work.work);
+	struct pt_cmd_queue *cmd_q = &ae4cmd_q->cmd_q;
+	struct pt_cmd *cmd;
+	u32 cridx;
+
+	for (;;) {
+		wait_event_interruptible(ae4cmd_q->q_w,
+					 ((atomic64_read(&ae4cmd_q->done_cnt)) <
+					   atomic64_read(&ae4cmd_q->intr_cnt)));
+
+		atomic64_inc(&ae4cmd_q->done_cnt);
+
+		mutex_lock(&ae4cmd_q->cmd_lock);
+		cridx = readl(cmd_q->reg_control + AE4_RD_IDX_OFF);
+		while ((ae4cmd_q->dridx != cridx) && !list_empty(&ae4cmd_q->cmd)) {
+			cmd = list_first_entry(&ae4cmd_q->cmd, struct pt_cmd, entry);
+			list_del(&cmd->entry);
+
+			ae4_check_status_error(ae4cmd_q, ae4cmd_q->dridx);
+			cmd->pt_cmd_callback(cmd->data, cmd->ret);
+
+			ae4cmd_q->q_cmd_count--;
+			ae4cmd_q->dridx = (ae4cmd_q->dridx + 1) % CMD_Q_LEN;
+
+			complete_all(&ae4cmd_q->cmp);
+		}
+		mutex_unlock(&ae4cmd_q->cmd_lock);
+	}
+}
+
+static irqreturn_t ae4_core_irq_handler(int irq, void *data)
+{
+	struct ae4_cmd_queue *ae4cmd_q = data;
+	struct pt_cmd_queue *cmd_q;
+	struct pt_device *pt;
+	u32 status;
+
+	cmd_q = &ae4cmd_q->cmd_q;
+	pt = cmd_q->pt;
+
+	pt->total_interrupts++;
+	atomic64_inc(&ae4cmd_q->intr_cnt);
+
+	status = readl(cmd_q->reg_control + AE4_INTR_STS_OFF);
+	if (status & BIT(0)) {
+		status &= GENMASK(31, 1);
+		writel(status, cmd_q->reg_control + AE4_INTR_STS_OFF);
+	}
+
+	wake_up(&ae4cmd_q->q_w);
+
+	return IRQ_HANDLED;
+}
+
+void ae4_destroy_work(struct ae4_device *ae4)
+{
+	struct ae4_cmd_queue *ae4cmd_q;
+	int i;
+
+	for (i = 0; i < ae4->cmd_q_count; i++) {
+		ae4cmd_q = &ae4->ae4cmd_q[i];
+
+		if (!ae4cmd_q->pws)
+			break;
+
+		cancel_delayed_work_sync(&ae4cmd_q->p_work);
+		destroy_workqueue(ae4cmd_q->pws);
+	}
+}
+
+int ae4_core_init(struct ae4_device *ae4)
+{
+	struct pt_device *pt = &ae4->pt;
+	struct ae4_cmd_queue *ae4cmd_q;
+	struct device *dev = pt->dev;
+	struct pt_cmd_queue *cmd_q;
+	int i, ret = 0;
+
+	writel(max_hw_q, pt->io_regs);
+
+	for (i = 0; i < max_hw_q; i++) {
+		ae4cmd_q = &ae4->ae4cmd_q[i];
+		ae4cmd_q->id = ae4->cmd_q_count;
+		ae4->cmd_q_count++;
+
+		cmd_q = &ae4cmd_q->cmd_q;
+		cmd_q->pt = pt;
+
+		cmd_q->reg_control = pt->io_regs + ((i + 1) * AE4_Q_SZ);
+
+		ret = devm_request_irq(dev, ae4->ae4_irq[i], ae4_core_irq_handler, 0,
+				       dev_name(pt->dev), ae4cmd_q);
+		if (ret)
+			return ret;
+
+		cmd_q->qsize = Q_SIZE(sizeof(struct ae4dma_desc));
+
+		cmd_q->qbase = dmam_alloc_coherent(dev, cmd_q->qsize, &cmd_q->qbase_dma,
+						   GFP_KERNEL);
+		if (!cmd_q->qbase)
+			return -ENOMEM;
+	}
+
+	for (i = 0; i < ae4->cmd_q_count; i++) {
+		ae4cmd_q = &ae4->ae4cmd_q[i];
+
+		cmd_q = &ae4cmd_q->cmd_q;
+
+		cmd_q->reg_control = pt->io_regs + ((i + 1) * AE4_Q_SZ);
+
+		/* Update the device registers with queue information. */
+		writel(CMD_Q_LEN, cmd_q->reg_control + AE4_MAX_IDX_OFF);
+
+		cmd_q->qdma_tail = cmd_q->qbase_dma;
+		writel(lower_32_bits(cmd_q->qdma_tail), cmd_q->reg_control + AE4_Q_BASE_L_OFF);
+		writel(upper_32_bits(cmd_q->qdma_tail), cmd_q->reg_control + AE4_Q_BASE_H_OFF);
+
+		INIT_LIST_HEAD(&ae4cmd_q->cmd);
+		init_waitqueue_head(&ae4cmd_q->q_w);
+
+		ae4cmd_q->pws = alloc_ordered_workqueue("ae4dma_%d", WQ_MEM_RECLAIM, ae4cmd_q->id);
+		if (!ae4cmd_q->pws) {
+			ae4_destroy_work(ae4);
+			return -ENOMEM;
+		}
+		INIT_DELAYED_WORK(&ae4cmd_q->p_work, ae4_pending_work);
+		queue_delayed_work(ae4cmd_q->pws, &ae4cmd_q->p_work,  usecs_to_jiffies(100));
+
+		init_completion(&ae4cmd_q->cmp);
+	}
+
+	ret = pt_dmaengine_register(pt);
+	if (ret)
+		ae4_destroy_work(ae4);
+	else
+		ptdma_debugfs_setup(pt);
+
+	return ret;
+}
--- a/drivers/dma/amd/ae4dma/ae4dma-pci.c
+++ b/drivers/dma/amd/ae4dma/ae4dma-pci.c
@ -0,0 +1,158 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * AMD AE4DMA driver
+ *
+ * Copyright (c) 2024, Advanced Micro Devices, Inc.
+ * All Rights Reserved.
+ *
+ * Author: Basavaraj Natikar <Basavaraj.Natikar@amd.com>
+ */
+
+#include "ae4dma.h"
+
+static int ae4_get_irqs(struct ae4_device *ae4)
+{
+	struct ae4_msix *ae4_msix = ae4->ae4_msix;
+	struct pt_device *pt = &ae4->pt;
+	struct device *dev = pt->dev;
+	struct pci_dev *pdev;
+	int i, v, ret;
+
+	pdev = to_pci_dev(dev);
+
+	for (v = 0; v < ARRAY_SIZE(ae4_msix->msix_entry); v++)
+		ae4_msix->msix_entry[v].entry = v;
+
+	ret = pci_alloc_irq_vectors(pdev, v, v, PCI_IRQ_MSIX);
+	if (ret != v) {
+		if (ret > 0)
+			pci_free_irq_vectors(pdev);
+
+		dev_err(dev, "could not enable MSI-X (%d), trying MSI\n", ret);
+		ret = pci_alloc_irq_vectors(pdev, 1, 1, PCI_IRQ_MSI);
+		if (ret < 0) {
+			dev_err(dev, "could not enable MSI (%d)\n", ret);
+			return ret;
+		}
+
+		ret = pci_irq_vector(pdev, 0);
+		if (ret < 0) {
+			pci_free_irq_vectors(pdev);
+			return ret;
+		}
+
+		for (i = 0; i < MAX_AE4_HW_QUEUES; i++)
+			ae4->ae4_irq[i] = ret;
+
+	} else {
+		ae4_msix->msix_count = ret;
+		for (i = 0; i < MAX_AE4_HW_QUEUES; i++)
+			ae4->ae4_irq[i] = ae4_msix->msix_entry[i].vector;
+	}
+
+	return ret;
+}
+
+static void ae4_free_irqs(struct ae4_device *ae4)
+{
+	struct ae4_msix *ae4_msix = ae4->ae4_msix;
+	struct pt_device *pt = &ae4->pt;
+	struct device *dev = pt->dev;
+	struct pci_dev *pdev;
+
+	pdev = to_pci_dev(dev);
+
+	if (ae4_msix && (ae4_msix->msix_count || ae4->ae4_irq[MAX_AE4_HW_QUEUES - 1]))
+		pci_free_irq_vectors(pdev);
+}
+
+static void ae4_deinit(struct ae4_device *ae4)
+{
+	ae4_free_irqs(ae4);
+}
+
+static int ae4_pci_probe(struct pci_dev *pdev, const struct pci_device_id *id)
+{
+	struct device *dev = &pdev->dev;
+	struct ae4_device *ae4;
+	struct pt_device *pt;
+	int bar_mask;
+	int ret = 0;
+
+	ae4 = devm_kzalloc(dev, sizeof(*ae4), GFP_KERNEL);
+	if (!ae4)
+		return -ENOMEM;
+
+	ae4->ae4_msix = devm_kzalloc(dev, sizeof(struct ae4_msix), GFP_KERNEL);
+	if (!ae4->ae4_msix)
+		return -ENOMEM;
+
+	ret = pcim_enable_device(pdev);
+	if (ret)
+		goto ae4_error;
+
+	bar_mask = pci_select_bars(pdev, IORESOURCE_MEM);
+	ret = pcim_iomap_regions(pdev, bar_mask, "ae4dma");
+	if (ret)
+		goto ae4_error;
+
+	pt = &ae4->pt;
+	pt->dev = dev;
+	pt->ver = AE4_DMA_VERSION;
+
+	pt->io_regs = pcim_iomap_table(pdev)[0];
+	if (!pt->io_regs) {
+		ret = -ENOMEM;
+		goto ae4_error;
+	}
+
+	ret = ae4_get_irqs(ae4);
+	if (ret < 0)
+		goto ae4_error;
+
+	pci_set_master(pdev);
+
+	dma_set_mask_and_coherent(dev, DMA_BIT_MASK(48));
+
+	dev_set_drvdata(dev, ae4);
+
+	ret = ae4_core_init(ae4);
+	if (ret)
+		goto ae4_error;
+
+	return 0;
+
+ae4_error:
+	ae4_deinit(ae4);
+
+	return ret;
+}
+
+static void ae4_pci_remove(struct pci_dev *pdev)
+{
+	struct ae4_device *ae4 = dev_get_drvdata(&pdev->dev);
+
+	ae4_destroy_work(ae4);
+	ae4_deinit(ae4);
+}
+
+static const struct pci_device_id ae4_pci_table[] = {
+	{ PCI_VDEVICE(AMD, 0x14C8), },
+	{ PCI_VDEVICE(AMD, 0x14DC), },
+	{ PCI_VDEVICE(AMD, 0x149B), },
+	/* Last entry must be zero */
+	{ 0, }
+};
+MODULE_DEVICE_TABLE(pci, ae4_pci_table);
+
+static struct pci_driver ae4_pci_driver = {
+	.name = "ae4dma",
+	.id_table = ae4_pci_table,
+	.probe = ae4_pci_probe,
+	.remove = ae4_pci_remove,
+};
+
+module_pci_driver(ae4_pci_driver);
+
+MODULE_LICENSE("GPL");
+MODULE_DESCRIPTION("AMD AE4DMA driver");
--- a/drivers/dma/amd/ae4dma/ae4dma.h
+++ b/drivers/dma/amd/ae4dma/ae4dma.h
@ -0,0 +1,100 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * AMD AE4DMA driver
+ *
+ * Copyright (c) 2024, Advanced Micro Devices, Inc.
+ * All Rights Reserved.
+ *
+ * Author: Basavaraj Natikar <Basavaraj.Natikar@amd.com>
+ */
+#ifndef __AE4DMA_H__
+#define __AE4DMA_H__
+
+#include <linux/device.h>
+#include <linux/dmaengine.h>
+#include <linux/dmapool.h>
+#include <linux/list.h>
+#include <linux/mutex.h>
+#include <linux/pci.h>
+#include <linux/spinlock.h>
+#include <linux/wait.h>
+
+#include "../ptdma/ptdma.h"
+#include "../../virt-dma.h"
+
+#define MAX_AE4_HW_QUEUES		16
+
+#define AE4_DESC_COMPLETED		0x03
+
+#define AE4_MAX_IDX_OFF			0x08
+#define AE4_RD_IDX_OFF			0x0c
+#define AE4_WR_IDX_OFF			0x10
+#define AE4_INTR_STS_OFF		0x14
+#define AE4_Q_BASE_L_OFF		0x18
+#define AE4_Q_BASE_H_OFF		0x1c
+#define AE4_Q_SZ			0x20
+
+#define AE4_DMA_VERSION			4
+#define CMD_AE4_DESC_DW0_VAL		2
+
+struct ae4_msix {
+	int msix_count;
+	struct msix_entry msix_entry[MAX_AE4_HW_QUEUES];
+};
+
+struct ae4_cmd_queue {
+	struct ae4_device *ae4;
+	struct pt_cmd_queue cmd_q;
+	struct list_head cmd;
+	/* protect command operations */
+	struct mutex cmd_lock;
+	struct delayed_work p_work;
+	struct workqueue_struct *pws;
+	struct completion cmp;
+	wait_queue_head_t q_w;
+	atomic64_t intr_cnt;
+	atomic64_t done_cnt;
+	u64 q_cmd_count;
+	u32 dridx;
+	u32 tail_wi;
+	u32 id;
+};
+
+union dwou {
+	u32 dw0;
+	struct dword0 {
+	u8	byte0;
+	u8	byte1;
+	u16	timestamp;
+	} dws;
+};
+
+struct dword1 {
+	u8	status;
+	u8	err_code;
+	u16	desc_id;
+};
+
+struct ae4dma_desc {
+	union dwou dwouv;
+	struct dword1 dw1;
+	u32 length;
+	u32 rsvd;
+	u32 src_hi;
+	u32 src_lo;
+	u32 dst_hi;
+	u32 dst_lo;
+};
+
+struct ae4_device {
+	struct pt_device pt;
+	struct ae4_msix *ae4_msix;
+	struct ae4_cmd_queue ae4cmd_q[MAX_AE4_HW_QUEUES];
+	unsigned int ae4_irq[MAX_AE4_HW_QUEUES];
+	unsigned int cmd_q_count;
+};
+
+int ae4_core_init(struct ae4_device *ae4);
+void ae4_destroy_work(struct ae4_device *ae4);
+void ae4_check_status_error(struct ae4_cmd_queue *ae4cmd_q, int idx);
+#endif
--- a/drivers/dma/amd/ptdma/Makefile
+++ b/drivers/dma/amd/ptdma/Makefile
--- a/drivers/dma/amd/ptdma/ptdma-debugfs.c
+++ b/drivers/dma/amd/ptdma/ptdma-debugfs.c
@ -13,6 +13,7 @@
 #include <linux/seq_file.h>

 #include "ptdma.h"
+#include "../ae4dma/ae4dma.h"

 /* DebugFS helpers */
 #define	RI_VERSION_NUM	0x0000003F
@ -23,11 +24,19 @@
 static int pt_debugfs_info_show(struct seq_file *s, void *p)
 {
 	struct pt_device *pt = s->private;
+	struct ae4_device *ae4;
 	unsigned int regval;

 	seq_printf(s, "Device name: %s\n", dev_name(pt->dev));
-	seq_printf(s, "   # Queues: %d\n", 1);
-	seq_printf(s, "     # Cmds: %d\n", pt->cmd_count);
+
+	if (pt->ver == AE4_DMA_VERSION) {
+		ae4 = container_of(pt, struct ae4_device, pt);
+		seq_printf(s, "   # Queues: %d\n", ae4->cmd_q_count);
+		seq_printf(s, "     # Cmds per queue: %d\n", CMD_Q_LEN);
+	} else {
+		seq_printf(s, "   # Queues: %d\n", 1);
+		seq_printf(s, "     # Cmds: %d\n", pt->cmd_count);
+	}

 	regval = ioread32(pt->io_regs + CMD_PT_VERSION);

@ -55,6 +64,7 @@ static int pt_debugfs_stats_show(struct seq_file *s, void *p)
 static int pt_debugfs_queue_show(struct seq_file *s, void *p)
 {
 	struct pt_cmd_queue *cmd_q = s->private;
+	struct pt_device *pt;
 	unsigned int regval;

 	if (!cmd_q)
@ -62,18 +72,24 @@ static int pt_debugfs_queue_show(struct seq_file *s, void *p)

 	seq_printf(s, "               Pass-Thru: %ld\n", cmd_q->total_pt_ops);

-	regval = ioread32(cmd_q->reg_control + 0x000C);
+	pt = cmd_q->pt;
+	if (pt->ver == AE4_DMA_VERSION) {
+		regval = readl(cmd_q->reg_control + 0x4);
+		seq_printf(s, "     Enabled Interrupts:: status 0x%x\n", regval);
+	} else {
+		regval = ioread32(cmd_q->reg_control + 0x000C);

-	seq_puts(s, "      Enabled Interrupts:");
-	if (regval & INT_EMPTY_QUEUE)
-		seq_puts(s, " EMPTY");
-	if (regval & INT_QUEUE_STOPPED)
-		seq_puts(s, " STOPPED");
-	if (regval & INT_ERROR)
-		seq_puts(s, " ERROR");
-	if (regval & INT_COMPLETION)
-		seq_puts(s, " COMPLETION");
-	seq_puts(s, "\n");
+		seq_puts(s, "      Enabled Interrupts:");
+		if (regval & INT_EMPTY_QUEUE)
+			seq_puts(s, " EMPTY");
+		if (regval & INT_QUEUE_STOPPED)
+			seq_puts(s, " STOPPED");
+		if (regval & INT_ERROR)
+			seq_puts(s, " ERROR");
+		if (regval & INT_COMPLETION)
+			seq_puts(s, " COMPLETION");
+		seq_puts(s, "\n");
+	}

 	return 0;
 }
@ -84,8 +100,12 @@ DEFINE_SHOW_ATTRIBUTE(pt_debugfs_stats);

 void ptdma_debugfs_setup(struct pt_device *pt)
 {
-	struct pt_cmd_queue *cmd_q;
 	struct dentry *debugfs_q_instance;
+	struct ae4_cmd_queue *ae4cmd_q;
+	struct pt_cmd_queue *cmd_q;
+	struct ae4_device *ae4;
+	char name[30];
+	int i;

 	if (!debugfs_initialized())
 		return;
@ -96,11 +116,28 @@ void ptdma_debugfs_setup(struct pt_device *pt)
 	debugfs_create_file("stats", 0400, pt->dma_dev.dbg_dev_root, pt,
 			    &pt_debugfs_stats_fops);

-	cmd_q = &pt->cmd_q;

-	debugfs_q_instance =
-		debugfs_create_dir("q", pt->dma_dev.dbg_dev_root);
+	if (pt->ver == AE4_DMA_VERSION) {
+		ae4 = container_of(pt, struct ae4_device, pt);
+		for (i = 0; i < ae4->cmd_q_count; i++) {
+			ae4cmd_q = &ae4->ae4cmd_q[i];
+			cmd_q = &ae4cmd_q->cmd_q;

-	debugfs_create_file("stats", 0400, debugfs_q_instance, cmd_q,
-			    &pt_debugfs_queue_fops);
+			memset(name, 0, sizeof(name));
+			snprintf(name, 29, "q%d", ae4cmd_q->id);
+
+			debugfs_q_instance =
+				debugfs_create_dir(name, pt->dma_dev.dbg_dev_root);
+
+			debugfs_create_file("stats", 0400, debugfs_q_instance, cmd_q,
+					    &pt_debugfs_queue_fops);
+		}
+	} else {
+		debugfs_q_instance =
+			debugfs_create_dir("q", pt->dma_dev.dbg_dev_root);
+		cmd_q = &pt->cmd_q;
+		debugfs_create_file("stats", 0400, debugfs_q_instance, cmd_q,
+				    &pt_debugfs_queue_fops);
+	}
 }
+EXPORT_SYMBOL_GPL(ptdma_debugfs_setup);
--- a/drivers/dma/amd/ptdma/ptdma-dev.c
+++ b/drivers/dma/amd/ptdma/ptdma-dev.c
--- a/drivers/dma/amd/ptdma/ptdma-dmaengine.c
+++ b/drivers/dma/amd/ptdma/ptdma-dmaengine.c
@ -9,9 +9,58 @@
 * Author: Gary R Hook <gary.hook@amd.com>
 */

+#include <linux/bitfield.h>
 #include "ptdma.h"
-#include "../dmaengine.h"
-#include "../virt-dma.h"
+#include "../ae4dma/ae4dma.h"
+#include "../../dmaengine.h"
+
+static char *ae4_error_codes[] = {
+	"",
+	"ERR 01: INVALID HEADER DW0",
+	"ERR 02: INVALID STATUS",
+	"ERR 03: INVALID LENGTH - 4 BYTE ALIGNMENT",
+	"ERR 04: INVALID SRC ADDR - 4 BYTE ALIGNMENT",
+	"ERR 05: INVALID DST ADDR - 4 BYTE ALIGNMENT",
+	"ERR 06: INVALID ALIGNMENT",
+	"ERR 07: INVALID DESCRIPTOR",
+};
+
+static void ae4_log_error(struct pt_device *d, int e)
+{
+	/* ERR 01 - 07 represents Invalid AE4 errors */
+	if (e <= 7)
+		dev_info(d->dev, "AE4DMA error: %s (0x%x)\n", ae4_error_codes[e], e);
+	/* ERR 08 - 15 represents Invalid Descriptor errors */
+	else if (e > 7 && e <= 15)
+		dev_info(d->dev, "AE4DMA error: %s (0x%x)\n", "INVALID DESCRIPTOR", e);
+	/* ERR 16 - 31 represents Firmware errors */
+	else if (e > 15 && e <= 31)
+		dev_info(d->dev, "AE4DMA error: %s (0x%x)\n", "FIRMWARE ERROR", e);
+	/* ERR 32 - 63 represents Fatal errors */
+	else if (e > 31 && e <= 63)
+		dev_info(d->dev, "AE4DMA error: %s (0x%x)\n", "FATAL ERROR", e);
+	/* ERR 64 - 255 represents PTE errors */
+	else if (e > 63 && e <= 255)
+		dev_info(d->dev, "AE4DMA error: %s (0x%x)\n", "PTE ERROR", e);
+	else
+		dev_info(d->dev, "Unknown AE4DMA error");
+}
+
+void ae4_check_status_error(struct ae4_cmd_queue *ae4cmd_q, int idx)
+{
+	struct pt_cmd_queue *cmd_q = &ae4cmd_q->cmd_q;
+	struct ae4dma_desc desc;
+	u8 status;
+
+	memcpy(&desc, &cmd_q->qbase[idx], sizeof(struct ae4dma_desc));
+	status = desc.dw1.status;
+	if (status && status != AE4_DESC_COMPLETED) {
+		cmd_q->cmd_error = desc.dw1.err_code;
+		if (cmd_q->cmd_error)
+			ae4_log_error(cmd_q->pt, cmd_q->cmd_error);
+	}
+}
+EXPORT_SYMBOL_GPL(ae4_check_status_error);

 static inline struct pt_dma_chan *to_pt_chan(struct dma_chan *dma_chan)
 {
@ -45,7 +94,71 @@ static void pt_do_cleanup(struct virt_dma_desc *vd)
 	kmem_cache_free(pt->dma_desc_cache, desc);
 }

-static int pt_dma_start_desc(struct pt_dma_desc *desc)
+static struct pt_cmd_queue *pt_get_cmd_queue(struct pt_device *pt, struct pt_dma_chan *chan)
+{
+	struct ae4_cmd_queue *ae4cmd_q;
+	struct pt_cmd_queue *cmd_q;
+	struct ae4_device *ae4;
+
+	if (pt->ver == AE4_DMA_VERSION) {
+		ae4 = container_of(pt, struct ae4_device, pt);
+		ae4cmd_q = &ae4->ae4cmd_q[chan->id];
+		cmd_q = &ae4cmd_q->cmd_q;
+	} else {
+		cmd_q = &pt->cmd_q;
+	}
+
+	return cmd_q;
+}
+
+static int ae4_core_execute_cmd(struct ae4dma_desc *desc, struct ae4_cmd_queue *ae4cmd_q)
+{
+	bool soc = FIELD_GET(DWORD0_SOC, desc->dwouv.dw0);
+	struct pt_cmd_queue *cmd_q = &ae4cmd_q->cmd_q;
+
+	if (soc) {
+		desc->dwouv.dw0 |= FIELD_PREP(DWORD0_IOC, desc->dwouv.dw0);
+		desc->dwouv.dw0 &= ~DWORD0_SOC;
+	}
+
+	mutex_lock(&ae4cmd_q->cmd_lock);
+	memcpy(&cmd_q->qbase[ae4cmd_q->tail_wi], desc, sizeof(struct ae4dma_desc));
+	ae4cmd_q->q_cmd_count++;
+	ae4cmd_q->tail_wi = (ae4cmd_q->tail_wi + 1) % CMD_Q_LEN;
+	writel(ae4cmd_q->tail_wi, cmd_q->reg_control + AE4_WR_IDX_OFF);
+	mutex_unlock(&ae4cmd_q->cmd_lock);
+
+	wake_up(&ae4cmd_q->q_w);
+
+	return 0;
+}
+
+static int pt_core_perform_passthru_ae4(struct pt_cmd_queue *cmd_q,
+					struct pt_passthru_engine *pt_engine)
+{
+	struct ae4_cmd_queue *ae4cmd_q = container_of(cmd_q, struct ae4_cmd_queue, cmd_q);
+	struct ae4dma_desc desc;
+
+	cmd_q->cmd_error = 0;
+	cmd_q->total_pt_ops++;
+	memset(&desc, 0, sizeof(desc));
+	desc.dwouv.dws.byte0 = CMD_AE4_DESC_DW0_VAL;
+
+	desc.dw1.status = 0;
+	desc.dw1.err_code = 0;
+	desc.dw1.desc_id = 0;
+
+	desc.length = pt_engine->src_len;
+
+	desc.src_lo = upper_32_bits(pt_engine->src_dma);
+	desc.src_hi = lower_32_bits(pt_engine->src_dma);
+	desc.dst_lo = upper_32_bits(pt_engine->dst_dma);
+	desc.dst_hi = lower_32_bits(pt_engine->dst_dma);
+
+	return ae4_core_execute_cmd(&desc, ae4cmd_q);
+}
+
+static int pt_dma_start_desc(struct pt_dma_desc *desc, struct pt_dma_chan *chan)
 {
 	struct pt_passthru_engine *pt_engine;
 	struct pt_device *pt;
@ -56,13 +169,18 @@ static int pt_dma_start_desc(struct pt_dma_desc *desc)

 	pt_cmd = &desc->pt_cmd;
 	pt = pt_cmd->pt;
-	cmd_q = &pt->cmd_q;
+
+	cmd_q = pt_get_cmd_queue(pt, chan);
+
 	pt_engine = &pt_cmd->passthru;

 	pt->tdata.cmd = pt_cmd;

 	/* Execute the command */
-	pt_cmd->ret = pt_core_perform_passthru(cmd_q, pt_engine);
+	if (pt->ver == AE4_DMA_VERSION)
+		pt_cmd->ret = pt_core_perform_passthru_ae4(cmd_q, pt_engine);
+	else
+		pt_cmd->ret = pt_core_perform_passthru(cmd_q, pt_engine);

 	return 0;
 }
@ -151,7 +269,7 @@ static void pt_cmd_callback(void *data, int err)
 		if (!desc)
 			break;

-		ret = pt_dma_start_desc(desc);
+		ret = pt_dma_start_desc(desc, chan);
 		if (!ret)
 			break;

@ -186,7 +304,10 @@ static struct pt_dma_desc *pt_create_desc(struct dma_chan *dma_chan,
 {
 	struct pt_dma_chan *chan = to_pt_chan(dma_chan);
 	struct pt_passthru_engine *pt_engine;
+	struct pt_device *pt = chan->pt;
+	struct ae4_cmd_queue *ae4cmd_q;
 	struct pt_dma_desc *desc;
+	struct ae4_device *ae4;
 	struct pt_cmd *pt_cmd;

 	desc = pt_alloc_dma_desc(chan, flags);
@ -194,7 +315,7 @@ static struct pt_dma_desc *pt_create_desc(struct dma_chan *dma_chan,
 		return NULL;

 	pt_cmd = &desc->pt_cmd;
-	pt_cmd->pt = chan->pt;
+	pt_cmd->pt = pt;
 	pt_engine = &pt_cmd->passthru;
 	pt_cmd->engine = PT_ENGINE_PASSTHRU;
 	pt_engine->src_dma = src;
@ -205,6 +326,14 @@ static struct pt_dma_desc *pt_create_desc(struct dma_chan *dma_chan,

 	desc->len = len;

+	if (pt->ver == AE4_DMA_VERSION) {
+		ae4 = container_of(pt, struct ae4_device, pt);
+		ae4cmd_q = &ae4->ae4cmd_q[chan->id];
+		mutex_lock(&ae4cmd_q->cmd_lock);
+		list_add_tail(&pt_cmd->entry, &ae4cmd_q->cmd);
+		mutex_unlock(&ae4cmd_q->cmd_lock);
+	}
+
 	return desc;
 }

@ -258,24 +387,43 @@ static void pt_issue_pending(struct dma_chan *dma_chan)
 		pt_cmd_callback(desc, 0);
 }

+static void pt_check_status_trans_ae4(struct pt_device *pt, struct pt_cmd_queue *cmd_q)
+{
+	struct ae4_cmd_queue *ae4cmd_q = container_of(cmd_q, struct ae4_cmd_queue, cmd_q);
+	int i;
+
+	for (i = 0; i < CMD_Q_LEN; i++)
+		ae4_check_status_error(ae4cmd_q, i);
+}
+
 static enum dma_status
 pt_tx_status(struct dma_chan *c, dma_cookie_t cookie,
 		struct dma_tx_state *txstate)
 {
-	struct pt_device *pt = to_pt_chan(c)->pt;
-	struct pt_cmd_queue *cmd_q = &pt->cmd_q;
+	struct pt_dma_chan *chan = to_pt_chan(c);
+	struct pt_device *pt = chan->pt;
+	struct pt_cmd_queue *cmd_q;
+
+	cmd_q = pt_get_cmd_queue(pt, chan);
+
+	if (pt->ver == AE4_DMA_VERSION)
+		pt_check_status_trans_ae4(pt, cmd_q);
+	else
+		pt_check_status_trans(pt, cmd_q);

-	pt_check_status_trans(pt, cmd_q);
 	return dma_cookie_status(c, cookie, txstate);
 }

 static int pt_pause(struct dma_chan *dma_chan)
 {
 	struct pt_dma_chan *chan = to_pt_chan(dma_chan);
+	struct pt_device *pt = chan->pt;
+	struct pt_cmd_queue *cmd_q;
 	unsigned long flags;

 	spin_lock_irqsave(&chan->vc.lock, flags);
-	pt_stop_queue(&chan->pt->cmd_q);
+	cmd_q = pt_get_cmd_queue(pt, chan);
+	pt_stop_queue(cmd_q);
 	spin_unlock_irqrestore(&chan->vc.lock, flags);

 	return 0;
@ -285,10 +433,13 @@ static int pt_resume(struct dma_chan *dma_chan)
 {
 	struct pt_dma_chan *chan = to_pt_chan(dma_chan);
 	struct pt_dma_desc *desc = NULL;
+	struct pt_device *pt = chan->pt;
+	struct pt_cmd_queue *cmd_q;
 	unsigned long flags;

 	spin_lock_irqsave(&chan->vc.lock, flags);
-	pt_start_queue(&chan->pt->cmd_q);
+	cmd_q = pt_get_cmd_queue(pt, chan);
+	pt_start_queue(cmd_q);
 	desc = pt_next_dma_desc(chan);
 	spin_unlock_irqrestore(&chan->vc.lock, flags);

@ -302,11 +453,17 @@ static int pt_resume(struct dma_chan *dma_chan)
 static int pt_terminate_all(struct dma_chan *dma_chan)
 {
 	struct pt_dma_chan *chan = to_pt_chan(dma_chan);
+	struct pt_device *pt = chan->pt;
+	struct pt_cmd_queue *cmd_q;
 	unsigned long flags;
-	struct pt_cmd_queue *cmd_q = &chan->pt->cmd_q;
 	LIST_HEAD(head);

-	iowrite32(SUPPORTED_INTERRUPTS, cmd_q->reg_control + 0x0010);
+	cmd_q = pt_get_cmd_queue(pt, chan);
+	if (pt->ver == AE4_DMA_VERSION)
+		pt_stop_queue(cmd_q);
+	else
+		iowrite32(SUPPORTED_INTERRUPTS, cmd_q->reg_control + 0x0010);
+
 	spin_lock_irqsave(&chan->vc.lock, flags);
 	vchan_get_all_descriptors(&chan->vc, &head);
 	spin_unlock_irqrestore(&chan->vc.lock, flags);
@ -319,14 +476,24 @@ static int pt_terminate_all(struct dma_chan *dma_chan)

 int pt_dmaengine_register(struct pt_device *pt)
 {
-	struct pt_dma_chan *chan;
 	struct dma_device *dma_dev = &pt->dma_dev;
-	char *cmd_cache_name;
+	struct ae4_cmd_queue *ae4cmd_q = NULL;
+	struct ae4_device *ae4 = NULL;
+	struct pt_dma_chan *chan;
 	char *desc_cache_name;
-	int ret;
+	char *cmd_cache_name;
+	int ret, i;
+
+	if (pt->ver == AE4_DMA_VERSION)
+		ae4 = container_of(pt, struct ae4_device, pt);
+
+	if (ae4)
+		pt->pt_dma_chan = devm_kcalloc(pt->dev, ae4->cmd_q_count,
+					       sizeof(*pt->pt_dma_chan), GFP_KERNEL);
+	else
+		pt->pt_dma_chan = devm_kzalloc(pt->dev, sizeof(*pt->pt_dma_chan),
+					       GFP_KERNEL);

-	pt->pt_dma_chan = devm_kzalloc(pt->dev, sizeof(*pt->pt_dma_chan),
-				       GFP_KERNEL);
 	if (!pt->pt_dma_chan)
 		return -ENOMEM;

@ -368,9 +535,6 @@ int pt_dmaengine_register(struct pt_device *pt)

 	INIT_LIST_HEAD(&dma_dev->channels);

-	chan = pt->pt_dma_chan;
-	chan->pt = pt;
-
 	/* Set base and prep routines */
 	dma_dev->device_free_chan_resources = pt_free_chan_resources;
 	dma_dev->device_prep_dma_memcpy = pt_prep_dma_memcpy;
@ -382,8 +546,21 @@ int pt_dmaengine_register(struct pt_device *pt)
 	dma_dev->device_terminate_all = pt_terminate_all;
 	dma_dev->device_synchronize = pt_synchronize;

-	chan->vc.desc_free = pt_do_cleanup;
-	vchan_init(&chan->vc, dma_dev);
+	if (ae4) {
+		for (i = 0; i < ae4->cmd_q_count; i++) {
+			chan = pt->pt_dma_chan + i;
+			ae4cmd_q = &ae4->ae4cmd_q[i];
+			chan->id = ae4cmd_q->id;
+			chan->pt = pt;
+			chan->vc.desc_free = pt_do_cleanup;
+			vchan_init(&chan->vc, dma_dev);
+		}
+	} else {
+		chan = pt->pt_dma_chan;
+		chan->pt = pt;
+		chan->vc.desc_free = pt_do_cleanup;
+		vchan_init(&chan->vc, dma_dev);
+	}

 	ret = dma_async_device_register(dma_dev);
 	if (ret)
@ -399,6 +576,7 @@ err_cache:

 	return ret;
 }
+EXPORT_SYMBOL_GPL(pt_dmaengine_register);

 void pt_dmaengine_unregister(struct pt_device *pt)
 {
--- a/drivers/dma/amd/ptdma/ptdma-pci.c
+++ b/drivers/dma/amd/ptdma/ptdma-pci.c
--- a/drivers/dma/amd/ptdma/ptdma.h
+++ b/drivers/dma/amd/ptdma/ptdma.h
@ -22,7 +22,7 @@
 #include <linux/wait.h>
 #include <linux/dmapool.h>

-#include "../virt-dma.h"
+#include "../../virt-dma.h"

 #define MAX_PT_NAME_LEN			16
 #define MAX_DMAPOOL_NAME_LEN		32
@ -184,6 +184,7 @@ struct pt_dma_desc {
 struct pt_dma_chan {
 	struct virt_dma_chan vc;
 	struct pt_device *pt;
+	u32 id;
 };

 struct pt_cmd_queue {
@ -262,6 +263,7 @@ struct pt_device {
 	unsigned long total_interrupts;

 	struct pt_tasklet_data tdata;
+	int ver;
 };

 /*
--- a/drivers/dma/amd/qdma/qdma.c
+++ b/drivers/dma/amd/qdma/qdma.c
@ -283,16 +283,20 @@ static int qdma_check_queue_status(struct qdma_device *qdev,

 static int qdma_clear_queue_context(const struct qdma_queue *queue)
 {
-	enum qdma_ctxt_type h2c_types[] = { QDMA_CTXT_DESC_SW_H2C,
-					    QDMA_CTXT_DESC_HW_H2C,
-					    QDMA_CTXT_DESC_CR_H2C,
-					    QDMA_CTXT_PFTCH, };
-	enum qdma_ctxt_type c2h_types[] = { QDMA_CTXT_DESC_SW_C2H,
-					    QDMA_CTXT_DESC_HW_C2H,
-					    QDMA_CTXT_DESC_CR_C2H,
-					    QDMA_CTXT_PFTCH, };
+	static const enum qdma_ctxt_type h2c_types[] = {
+		QDMA_CTXT_DESC_SW_H2C,
+		QDMA_CTXT_DESC_HW_H2C,
+		QDMA_CTXT_DESC_CR_H2C,
+		QDMA_CTXT_PFTCH,
+	};
+	static const enum qdma_ctxt_type c2h_types[] = {
+		QDMA_CTXT_DESC_SW_C2H,
+		QDMA_CTXT_DESC_HW_C2H,
+		QDMA_CTXT_DESC_CR_C2H,
+		QDMA_CTXT_PFTCH,
+	};
 	struct qdma_device *qdev = queue->qdev;
-	enum qdma_ctxt_type *type;
+	const enum qdma_ctxt_type *type;
 	int ret, num, i;

 	if (queue->dir == DMA_MEM_TO_DEV) {
--- a/drivers/dma/bcm2835-dma.c
+++ b/drivers/dma/bcm2835-dma.c
@ -875,6 +875,27 @@ static struct dma_chan *bcm2835_dma_xlate(struct of_phandle_args *spec,
 	return chan;
 }

+static int bcm2835_dma_suspend_late(struct device *dev)
+{
+	struct bcm2835_dmadev *od = dev_get_drvdata(dev);
+	struct bcm2835_chan *c, *next;
+
+	list_for_each_entry_safe(c, next, &od->ddev.channels,
+				 vc.chan.device_node) {
+		void __iomem *chan_base = c->chan_base;
+
+		/* Check if DMA channel is busy */
+		if (readl(chan_base + BCM2835_DMA_ADDR))
+			return -EBUSY;
+	}
+
+	return 0;
+}
+
+static const struct dev_pm_ops bcm2835_dma_pm_ops = {
+	SET_LATE_SYSTEM_SLEEP_PM_OPS(bcm2835_dma_suspend_late, NULL)
+};
+
 static int bcm2835_dma_probe(struct platform_device *pdev)
 {
 	struct bcm2835_dmadev *od;
@ -1033,6 +1054,7 @@ static struct platform_driver bcm2835_dma_driver = {
 	.driver = {
 		.name = "bcm2835-dma",
 		.of_match_table = of_match_ptr(bcm2835_dma_of_match),
+		.pm = pm_ptr(&bcm2835_dma_pm_ops),
 	},
 };

--- a/drivers/dma/fsl-edma-common.c
+++ b/drivers/dma/fsl-edma-common.c
@ -480,8 +480,8 @@ void fsl_edma_fill_tcd(struct fsl_edma_chan *fsl_chan,
 		       bool disable_req, bool enable_sg)
 {
 	struct dma_slave_config *cfg = &fsl_chan->cfg;
+	u32 burst = 0;
 	u16 csr = 0;
-	u32 burst;

 	/*
 	 * eDMA hardware SGs require the TCDs to be stored in little
@ -496,16 +496,30 @@ void fsl_edma_fill_tcd(struct fsl_edma_chan *fsl_chan,

 	fsl_edma_set_tcd_to_le(fsl_chan, tcd, soff, soff);

-	if (fsl_chan->is_multi_fifo) {
-		/* set mloff to support multiple fifo */
-		burst = cfg->direction == DMA_DEV_TO_MEM ?
-				cfg->src_maxburst : cfg->dst_maxburst;
-		nbytes |= EDMA_V3_TCD_NBYTES_MLOFF(-(burst * 4));
-		/* enable DMLOE/SMLOE */
-		if (cfg->direction == DMA_MEM_TO_DEV) {
+	/* If we expect to have either multi_fifo or a port window size,
+	 * we will use minor loop offset, meaning bits 29-10 will be used for
+	 * address offset, while bits 9-0 will be used to tell DMA how much
+	 * data to read from addr.
+	 * If we don't have either of those, will use a major loop reading from addr
+	 * nbytes (29bits).
+	 */
+	if (cfg->direction == DMA_MEM_TO_DEV) {
+		if (fsl_chan->is_multi_fifo)
+			burst = cfg->dst_maxburst * 4;
+		if (cfg->dst_port_window_size)
+			burst = cfg->dst_port_window_size * cfg->dst_addr_width;
+		if (burst) {
+			nbytes |= EDMA_V3_TCD_NBYTES_MLOFF(-burst);
 			nbytes |= EDMA_V3_TCD_NBYTES_DMLOE;
 			nbytes &= ~EDMA_V3_TCD_NBYTES_SMLOE;
-		} else {
+		}
+	} else {
+		if (fsl_chan->is_multi_fifo)
+			burst = cfg->src_maxburst * 4;
+		if (cfg->src_port_window_size)
+			burst = cfg->src_port_window_size * cfg->src_addr_width;
+		if (burst) {
+			nbytes |= EDMA_V3_TCD_NBYTES_MLOFF(-burst);
 			nbytes |= EDMA_V3_TCD_NBYTES_SMLOE;
 			nbytes &= ~EDMA_V3_TCD_NBYTES_DMLOE;
 		}
@ -623,11 +637,15 @@ struct dma_async_tx_descriptor *fsl_edma_prep_dma_cyclic(
 			dst_addr = fsl_chan->dma_dev_addr;
 			soff = fsl_chan->cfg.dst_addr_width;
 			doff = fsl_chan->is_multi_fifo ? 4 : 0;
+			if (fsl_chan->cfg.dst_port_window_size)
+				doff = fsl_chan->cfg.dst_addr_width;
 		} else if (direction == DMA_DEV_TO_MEM) {
 			src_addr = fsl_chan->dma_dev_addr;
 			dst_addr = dma_buf_next;
 			soff = fsl_chan->is_multi_fifo ? 4 : 0;
 			doff = fsl_chan->cfg.src_addr_width;
+			if (fsl_chan->cfg.src_port_window_size)
+				soff = fsl_chan->cfg.src_addr_width;
 		} else {
 			/* DMA_DEV_TO_DEV */
 			src_addr = fsl_chan->cfg.src_addr;
--- a/drivers/dma/fsl-edma-common.h
+++ b/drivers/dma/fsl-edma-common.h
@ -68,6 +68,8 @@
 #define EDMA_V3_CH_CSR_EEI         BIT(2)
 #define EDMA_V3_CH_CSR_DONE        BIT(30)
 #define EDMA_V3_CH_CSR_ACTIVE      BIT(31)
+#define EDMA_V3_CH_ES_ERR          BIT(31)
+#define EDMA_V3_MP_ES_VLD          BIT(31)

 enum fsl_edma_pm_state {
 	RUNNING = 0,
@ -241,6 +243,7 @@ struct fsl_edma_engine {
 	const struct fsl_edma_drvdata *drvdata;
 	u32			n_chans;
 	int			txirq;
+	int			txirq_16_31;
 	int			errirq;
 	bool			big_endian;
 	struct edma_regs	regs;
--- a/drivers/dma/fsl-edma-main.c
+++ b/drivers/dma/fsl-edma-main.c
@ -3,10 +3,11 @@
 * drivers/dma/fsl-edma.c
 *
 * Copyright 2013-2014 Freescale Semiconductor, Inc.
+ * Copyright 2024 NXP
 *
 * Driver for the Freescale eDMA engine with flexible channel multiplexing
 * capability for DMA request sources. The eDMA block can be found on some
- * Vybrid and Layerscape SoCs.
+ * Vybrid, Layerscape and S32G SoCs.
 */

 #include <dt-bindings/dma/fsl-edma.h>
@ -72,6 +73,60 @@ static irqreturn_t fsl_edma2_tx_handler(int irq, void *devi_id)
 	return fsl_edma_tx_handler(irq, fsl_chan->edma);
 }

+static irqreturn_t fsl_edma3_or_tx_handler(int irq, void *dev_id,
+					   u8 start, u8 end)
+{
+	struct fsl_edma_engine *fsl_edma = dev_id;
+	struct fsl_edma_chan *chan;
+	int i;
+
+	end = min(end, fsl_edma->n_chans);
+
+	for (i = start; i < end; i++) {
+		chan = &fsl_edma->chans[i];
+
+		fsl_edma3_tx_handler(irq, chan);
+	}
+
+	return IRQ_HANDLED;
+}
+
+static irqreturn_t fsl_edma3_tx_0_15_handler(int irq, void *dev_id)
+{
+	return fsl_edma3_or_tx_handler(irq, dev_id, 0, 16);
+}
+
+static irqreturn_t fsl_edma3_tx_16_31_handler(int irq, void *dev_id)
+{
+	return fsl_edma3_or_tx_handler(irq, dev_id, 16, 32);
+}
+
+static irqreturn_t fsl_edma3_or_err_handler(int irq, void *dev_id)
+{
+	struct fsl_edma_engine *fsl_edma = dev_id;
+	struct edma_regs *regs = &fsl_edma->regs;
+	unsigned int err, ch, ch_es;
+	struct fsl_edma_chan *chan;
+
+	err = edma_readl(fsl_edma, regs->es);
+	if (!(err & EDMA_V3_MP_ES_VLD))
+		return IRQ_NONE;
+
+	for (ch = 0; ch < fsl_edma->n_chans; ch++) {
+		chan = &fsl_edma->chans[ch];
+
+		ch_es = edma_readl_chreg(chan, ch_es);
+		if (!(ch_es & EDMA_V3_CH_ES_ERR))
+			continue;
+
+		edma_writel_chreg(chan, EDMA_V3_CH_ES_ERR, ch_es);
+		fsl_edma_disable_request(chan);
+		fsl_edma->chans[ch].status = DMA_ERROR;
+	}
+
+	return IRQ_HANDLED;
+}
+
 static irqreturn_t fsl_edma_err_handler(int irq, void *dev_id)
 {
 	struct fsl_edma_engine *fsl_edma = dev_id;
@ -274,6 +329,49 @@ static int fsl_edma3_irq_init(struct platform_device *pdev, struct fsl_edma_engi
 	return 0;
 }

+static int fsl_edma3_or_irq_init(struct platform_device *pdev,
+				 struct fsl_edma_engine *fsl_edma)
+{
+	int ret;
+
+	fsl_edma->txirq = platform_get_irq_byname(pdev, "tx-0-15");
+	if (fsl_edma->txirq < 0)
+		return fsl_edma->txirq;
+
+	fsl_edma->txirq_16_31 = platform_get_irq_byname(pdev, "tx-16-31");
+	if (fsl_edma->txirq_16_31 < 0)
+		return fsl_edma->txirq_16_31;
+
+	fsl_edma->errirq = platform_get_irq_byname(pdev, "err");
+	if (fsl_edma->errirq < 0)
+		return fsl_edma->errirq;
+
+	ret = devm_request_irq(&pdev->dev, fsl_edma->txirq,
+			       fsl_edma3_tx_0_15_handler, 0, "eDMA tx0_15",
+			       fsl_edma);
+	if (ret)
+		return dev_err_probe(&pdev->dev, ret,
+			       "Can't register eDMA tx0_15 IRQ.\n");
+
+	if (fsl_edma->n_chans > 16) {
+		ret = devm_request_irq(&pdev->dev, fsl_edma->txirq_16_31,
+				       fsl_edma3_tx_16_31_handler, 0,
+				       "eDMA tx16_31", fsl_edma);
+		if (ret)
+			return dev_err_probe(&pdev->dev, ret,
+					"Can't register eDMA tx16_31 IRQ.\n");
+	}
+
+	ret = devm_request_irq(&pdev->dev, fsl_edma->errirq,
+			       fsl_edma3_or_err_handler, 0, "eDMA err",
+			       fsl_edma);
+	if (ret)
+		return dev_err_probe(&pdev->dev, ret,
+				     "Can't register eDMA err IRQ.\n");
+
+	return 0;
+}
+
 static int
 fsl_edma2_irq_init(struct platform_device *pdev,
 		   struct fsl_edma_engine *fsl_edma)
@ -404,6 +502,14 @@ static struct fsl_edma_drvdata imx95_data5 = {
 	.setup_irq = fsl_edma3_irq_init,
 };

+static const struct fsl_edma_drvdata s32g2_data = {
+	.dmamuxs = DMAMUX_NR,
+	.chreg_space_sz = EDMA_TCD,
+	.chreg_off = 0x4000,
+	.flags = FSL_EDMA_DRV_EDMA3 | FSL_EDMA_DRV_MUX_SWAP,
+	.setup_irq = fsl_edma3_or_irq_init,
+};
+
 static const struct of_device_id fsl_edma_dt_ids[] = {
 	{ .compatible = "fsl,vf610-edma", .data = &vf610_data},
 	{ .compatible = "fsl,ls1028a-edma", .data = &ls1028a_data},
@ -413,6 +519,7 @@ static const struct of_device_id fsl_edma_dt_ids[] = {
 	{ .compatible = "fsl,imx93-edma3", .data = &imx93_data3},
 	{ .compatible = "fsl,imx93-edma4", .data = &imx93_data4},
 	{ .compatible = "fsl,imx95-edma5", .data = &imx95_data5},
+	{ .compatible = "nxp,s32g2-edma", .data = &s32g2_data},
 	{ /* sentinel */ }
 };
 MODULE_DEVICE_TABLE(of, fsl_edma_dt_ids);
@ -545,10 +652,6 @@ static int fsl_edma_probe(struct platform_device *pdev)
 	for (i = 0; i < fsl_edma->drvdata->dmamuxs; i++) {
 		char clkname[32];

-		/* eDMAv3 mux register move to TCD area if ch_mux exist */
-		if (drvdata->flags & FSL_EDMA_DRV_SPLIT_REG)
-			break;
-
 		fsl_edma->muxbase[i] = devm_platform_ioremap_resource(pdev,
 								      1 + i);
 		if (IS_ERR(fsl_edma->muxbase[i])) {
@ -677,7 +780,7 @@ static int fsl_edma_probe(struct platform_device *pdev)
 	}

 	ret = of_dma_controller_register(np,
-			drvdata->flags & FSL_EDMA_DRV_SPLIT_REG ? fsl_edma3_xlate : fsl_edma_xlate,
+			drvdata->dmamuxs ? fsl_edma_xlate : fsl_edma3_xlate,
 			fsl_edma);
 	if (ret) {
 		dev_err(&pdev->dev,
--- a/drivers/dma/idxd/cdev.c
+++ b/drivers/dma/idxd/cdev.c
@ -28,7 +28,6 @@ struct idxd_cdev_context {
 * global to avoid conflict file names.
 */
 static DEFINE_IDA(file_ida);
-static DEFINE_MUTEX(ida_lock);

 /*
 * ictx is an array based off of accelerator types. enum idxd_type
@ -123,9 +122,7 @@ static void idxd_file_dev_release(struct device *dev)
 	struct idxd_device *idxd = wq->idxd;
 	int rc;

-	mutex_lock(&ida_lock);
 	ida_free(&file_ida, ctx->id);
-	mutex_unlock(&ida_lock);

 	/* Wait for in-flight operations to complete. */
 	if (wq_shared(wq)) {
@ -284,9 +281,7 @@ static int idxd_cdev_open(struct inode *inode, struct file *filp)
 	}

 	idxd_cdev = wq->idxd_cdev;
-	mutex_lock(&ida_lock);
 	ctx->id = ida_alloc(&file_ida, GFP_KERNEL);
-	mutex_unlock(&ida_lock);
 	if (ctx->id < 0) {
 		dev_warn(dev, "ida alloc failure\n");
 		goto failed_ida;
--- a/drivers/dma/idxd/idxd.h
+++ b/drivers/dma/idxd/idxd.h
@ -374,6 +374,17 @@ struct idxd_device {
 	struct dentry *dbgfs_evl_file;

 	bool user_submission_safe;
+
+	struct idxd_saved_states *idxd_saved;
+};
+
+struct idxd_saved_states {
+	struct idxd_device saved_idxd;
+	struct idxd_evl saved_evl;
+	struct idxd_engine **saved_engines;
+	struct idxd_wq **saved_wqs;
+	struct idxd_group **saved_groups;
+	unsigned long *saved_wq_enable_map;
 };

 static inline unsigned int evl_ent_size(struct idxd_device *idxd)
@ -725,8 +736,6 @@ static inline void idxd_desc_complete(struct idxd_desc *desc,
 				   &desc->txd, &status);
 }

-int idxd_register_bus_type(void);
-void idxd_unregister_bus_type(void);
 int idxd_register_devices(struct idxd_device *idxd);
 void idxd_unregister_devices(struct idxd_device *idxd);
 void idxd_wqs_quiesce(struct idxd_device *idxd);
@ -742,6 +751,8 @@ void idxd_unmask_error_interrupts(struct idxd_device *idxd);

 /* device control */
 int idxd_device_drv_probe(struct idxd_dev *idxd_dev);
+int idxd_pci_probe_alloc(struct idxd_device *idxd, struct pci_dev *pdev,
+			 const struct pci_device_id *id);
 void idxd_device_drv_remove(struct idxd_dev *idxd_dev);
 int idxd_drv_enable_wq(struct idxd_wq *wq);
 void idxd_drv_disable_wq(struct idxd_wq *wq);
--- a/drivers/dma/idxd/init.c
+++ b/drivers/dma/idxd/init.c
@ -78,6 +78,8 @@ static struct pci_device_id idxd_pci_tbl[] = {
 	{ PCI_DEVICE_DATA(INTEL, IAX_SPR0, &idxd_driver_data[IDXD_TYPE_IAX]) },
 	/* IAA on DMR platforms */
 	{ PCI_DEVICE_DATA(INTEL, IAA_DMR, &idxd_driver_data[IDXD_TYPE_IAX]) },
+	/* IAA PTL platforms */
+	{ PCI_DEVICE_DATA(INTEL, IAA_PTL, &idxd_driver_data[IDXD_TYPE_IAX]) },
 	{ 0, }
 };
 MODULE_DEVICE_TABLE(pci, idxd_pci_tbl);
@ -723,67 +725,464 @@ static void idxd_cleanup(struct idxd_device *idxd)
 		idxd_disable_sva(idxd->pdev);
 }

-static int idxd_pci_probe(struct pci_dev *pdev, const struct pci_device_id *id)
+/*
+ * Attach IDXD device to IDXD driver.
+ */
+static int idxd_bind(struct device_driver *drv, const char *buf)
 {
-	struct device *dev = &pdev->dev;
-	struct idxd_device *idxd;
-	struct idxd_driver_data *data = (struct idxd_driver_data *)id->driver_data;
+	const struct bus_type *bus = drv->bus;
+	struct device *dev;
+	int err = -ENODEV;
+
+	dev = bus_find_device_by_name(bus, NULL, buf);
+	if (dev)
+		err = device_driver_attach(drv, dev);
+
+	put_device(dev);
+
+	return err;
+}
+
+/*
+ * Detach IDXD device from driver.
+ */
+static void idxd_unbind(struct device_driver *drv, const char *buf)
+{
+	const struct bus_type *bus = drv->bus;
+	struct device *dev;
+
+	dev = bus_find_device_by_name(bus, NULL, buf);
+	if (dev && dev->driver == drv)
+		device_release_driver(dev);
+
+	put_device(dev);
+}
+
+#define idxd_free_saved_configs(saved_configs, count)	\
+	do {						\
+		int i;					\
+							\
+		for (i = 0; i < (count); i++)		\
+			kfree(saved_configs[i]);	\
+	} while (0)
+
+static void idxd_free_saved(struct idxd_group **saved_groups,
+			    struct idxd_engine **saved_engines,
+			    struct idxd_wq **saved_wqs,
+			    struct idxd_device *idxd)
+{
+	if (saved_groups)
+		idxd_free_saved_configs(saved_groups, idxd->max_groups);
+	if (saved_engines)
+		idxd_free_saved_configs(saved_engines, idxd->max_engines);
+	if (saved_wqs)
+		idxd_free_saved_configs(saved_wqs, idxd->max_wqs);
+}
+
+/*
+ * Save IDXD device configurations including engines, groups, wqs etc.
+ * The saved configurations can be restored when needed.
+ */
+static int idxd_device_config_save(struct idxd_device *idxd,
+				   struct idxd_saved_states *idxd_saved)
+{
+	struct device *dev = &idxd->pdev->dev;
+	int i;
+
+	memcpy(&idxd_saved->saved_idxd, idxd, sizeof(*idxd));
+
+	if (idxd->evl) {
+		memcpy(&idxd_saved->saved_evl, idxd->evl,
+		       sizeof(struct idxd_evl));
+	}
+
+	struct idxd_group **saved_groups __free(kfree) =
+			kcalloc_node(idxd->max_groups,
+				     sizeof(struct idxd_group *),
+				     GFP_KERNEL, dev_to_node(dev));
+	if (!saved_groups)
+		return -ENOMEM;
+
+	for (i = 0; i < idxd->max_groups; i++) {
+		struct idxd_group *saved_group __free(kfree) =
+			kzalloc_node(sizeof(*saved_group), GFP_KERNEL,
+				     dev_to_node(dev));
+
+		if (!saved_group) {
+			/* Free saved groups */
+			idxd_free_saved(saved_groups, NULL, NULL, idxd);
+
+			return -ENOMEM;
+		}
+
+		memcpy(saved_group, idxd->groups[i], sizeof(*saved_group));
+		saved_groups[i] = no_free_ptr(saved_group);
+	}
+
+	struct idxd_engine **saved_engines =
+			kcalloc_node(idxd->max_engines,
+				     sizeof(struct idxd_engine *),
+				     GFP_KERNEL, dev_to_node(dev));
+	if (!saved_engines) {
+		/* Free saved groups */
+		idxd_free_saved(saved_groups, NULL, NULL, idxd);
+
+		return -ENOMEM;
+	}
+	for (i = 0; i < idxd->max_engines; i++) {
+		struct idxd_engine *saved_engine __free(kfree) =
+				kzalloc_node(sizeof(*saved_engine), GFP_KERNEL,
+					     dev_to_node(dev));
+		if (!saved_engine) {
+			/* Free saved groups and engines */
+			idxd_free_saved(saved_groups, saved_engines, NULL,
+					idxd);
+
+			return -ENOMEM;
+		}
+
+		memcpy(saved_engine, idxd->engines[i], sizeof(*saved_engine));
+		saved_engines[i] = no_free_ptr(saved_engine);
+	}
+
+	unsigned long *saved_wq_enable_map __free(bitmap) =
+			bitmap_zalloc_node(idxd->max_wqs, GFP_KERNEL,
+					   dev_to_node(dev));
+	if (!saved_wq_enable_map) {
+		/* Free saved groups and engines */
+		idxd_free_saved(saved_groups, saved_engines, NULL, idxd);
+
+		return -ENOMEM;
+	}
+
+	bitmap_copy(saved_wq_enable_map, idxd->wq_enable_map, idxd->max_wqs);
+
+	struct idxd_wq **saved_wqs __free(kfree) =
+			kcalloc_node(idxd->max_wqs, sizeof(struct idxd_wq *),
+				     GFP_KERNEL, dev_to_node(dev));
+	if (!saved_wqs) {
+		/* Free saved groups and engines */
+		idxd_free_saved(saved_groups, saved_engines, NULL, idxd);
+
+		return -ENOMEM;
+	}
+
+	for (i = 0; i < idxd->max_wqs; i++) {
+		struct idxd_wq *saved_wq __free(kfree) =
+			kzalloc_node(sizeof(*saved_wq), GFP_KERNEL,
+				     dev_to_node(dev));
+		struct idxd_wq *wq;
+
+		if (!saved_wq) {
+			/* Free saved groups, engines, and wqs */
+			idxd_free_saved(saved_groups, saved_engines, saved_wqs,
+					idxd);
+
+			return -ENOMEM;
+		}
+
+		if (!test_bit(i, saved_wq_enable_map))
+			continue;
+
+		wq = idxd->wqs[i];
+		mutex_lock(&wq->wq_lock);
+		memcpy(saved_wq, wq, sizeof(*saved_wq));
+		saved_wqs[i] = no_free_ptr(saved_wq);
+		mutex_unlock(&wq->wq_lock);
+	}
+
+	/* Save configurations */
+	idxd_saved->saved_groups = no_free_ptr(saved_groups);
+	idxd_saved->saved_engines = no_free_ptr(saved_engines);
+	idxd_saved->saved_wq_enable_map = no_free_ptr(saved_wq_enable_map);
+	idxd_saved->saved_wqs = no_free_ptr(saved_wqs);
+
+	return 0;
+}
+
+/*
+ * Restore IDXD device configurations including engines, groups, wqs etc
+ * that were saved before.
+ */
+static void idxd_device_config_restore(struct idxd_device *idxd,
+				       struct idxd_saved_states *idxd_saved)
+{
+	struct idxd_evl *saved_evl = &idxd_saved->saved_evl;
+	int i;
+
+	idxd->rdbuf_limit = idxd_saved->saved_idxd.rdbuf_limit;
+
+	if (saved_evl)
+		idxd->evl->size = saved_evl->size;
+
+	for (i = 0; i < idxd->max_groups; i++) {
+		struct idxd_group *saved_group, *group;
+
+		saved_group = idxd_saved->saved_groups[i];
+		group = idxd->groups[i];
+
+		group->rdbufs_allowed = saved_group->rdbufs_allowed;
+		group->rdbufs_reserved = saved_group->rdbufs_reserved;
+		group->tc_a = saved_group->tc_a;
+		group->tc_b = saved_group->tc_b;
+		group->use_rdbuf_limit = saved_group->use_rdbuf_limit;
+
+		kfree(saved_group);
+	}
+	kfree(idxd_saved->saved_groups);
+
+	for (i = 0; i < idxd->max_engines; i++) {
+		struct idxd_engine *saved_engine, *engine;
+
+		saved_engine = idxd_saved->saved_engines[i];
+		engine = idxd->engines[i];
+
+		engine->group = saved_engine->group;
+
+		kfree(saved_engine);
+	}
+	kfree(idxd_saved->saved_engines);
+
+	bitmap_copy(idxd->wq_enable_map, idxd_saved->saved_wq_enable_map,
+		    idxd->max_wqs);
+	bitmap_free(idxd_saved->saved_wq_enable_map);
+
+	for (i = 0; i < idxd->max_wqs; i++) {
+		struct idxd_wq *saved_wq, *wq;
+		size_t len;
+
+		if (!test_bit(i, idxd->wq_enable_map))
+			continue;
+
+		saved_wq = idxd_saved->saved_wqs[i];
+		wq = idxd->wqs[i];
+
+		mutex_lock(&wq->wq_lock);
+
+		wq->group = saved_wq->group;
+		wq->flags = saved_wq->flags;
+		wq->threshold = saved_wq->threshold;
+		wq->size = saved_wq->size;
+		wq->priority = saved_wq->priority;
+		wq->type = saved_wq->type;
+		len = strlen(saved_wq->name) + 1;
+		strscpy(wq->name, saved_wq->name, len);
+		wq->max_xfer_bytes = saved_wq->max_xfer_bytes;
+		wq->max_batch_size = saved_wq->max_batch_size;
+		wq->enqcmds_retries = saved_wq->enqcmds_retries;
+		wq->descs = saved_wq->descs;
+		wq->idxd_chan = saved_wq->idxd_chan;
+		len = strlen(saved_wq->driver_name) + 1;
+		strscpy(wq->driver_name, saved_wq->driver_name, len);
+
+		mutex_unlock(&wq->wq_lock);
+
+		kfree(saved_wq);
+	}
+
+	kfree(idxd_saved->saved_wqs);
+}
+
+static void idxd_reset_prepare(struct pci_dev *pdev)
+{
+	struct idxd_device *idxd = pci_get_drvdata(pdev);
+	struct device *dev = &idxd->pdev->dev;
+	const char *idxd_name;
 	int rc;

+	dev = &idxd->pdev->dev;
+	idxd_name = dev_name(idxd_confdev(idxd));
+
+	struct idxd_saved_states *idxd_saved __free(kfree) =
+			kzalloc_node(sizeof(*idxd_saved), GFP_KERNEL,
+				     dev_to_node(&pdev->dev));
+	if (!idxd_saved) {
+		dev_err(dev, "HALT: no memory\n");
+
+		return;
+	}
+
+	/* Save IDXD configurations. */
+	rc = idxd_device_config_save(idxd, idxd_saved);
+	if (rc < 0) {
+		dev_err(dev, "HALT: cannot save %s configs\n", idxd_name);
+
+		return;
+	}
+
+	idxd->idxd_saved = no_free_ptr(idxd_saved);
+
+	/* Save PCI device state. */
+	pci_save_state(idxd->pdev);
+}
+
+static void idxd_reset_done(struct pci_dev *pdev)
+{
+	struct idxd_device *idxd = pci_get_drvdata(pdev);
+	const char *idxd_name;
+	struct device *dev;
+	int rc, i;
+
+	if (!idxd->idxd_saved)
+		return;
+
+	dev = &idxd->pdev->dev;
+	idxd_name = dev_name(idxd_confdev(idxd));
+
+	/* Restore PCI device state. */
+	pci_restore_state(idxd->pdev);
+
+	/* Unbind idxd device from driver. */
+	idxd_unbind(&idxd_drv.drv, idxd_name);
+
+	/*
+	 * Probe PCI device without allocating or changing
+	 * idxd software data which keeps the same as before FLR.
+	 */
+	idxd_pci_probe_alloc(idxd, NULL, NULL);
+
+	/* Restore IDXD configurations. */
+	idxd_device_config_restore(idxd, idxd->idxd_saved);
+
+	/* Re-configure IDXD device if allowed. */
+	if (test_bit(IDXD_FLAG_CONFIGURABLE, &idxd->flags)) {
+		rc = idxd_device_config(idxd);
+		if (rc < 0) {
+			dev_err(dev, "HALT: %s config fails\n", idxd_name);
+			goto out;
+		}
+	}
+
+	/* Bind IDXD device to driver. */
+	rc = idxd_bind(&idxd_drv.drv, idxd_name);
+	if (rc < 0) {
+		dev_err(dev, "HALT: binding %s to driver fails\n", idxd_name);
+		goto out;
+	}
+
+	/* Bind enabled wq in the IDXD device to driver. */
+	for (i = 0; i < idxd->max_wqs; i++) {
+		if (test_bit(i, idxd->wq_enable_map)) {
+			struct idxd_wq *wq = idxd->wqs[i];
+			char wq_name[32];
+
+			wq->state = IDXD_WQ_DISABLED;
+			sprintf(wq_name, "wq%d.%d", idxd->id, wq->id);
+			/*
+			 * Bind to user driver depending on wq type.
+			 *
+			 * Currently only support user type WQ. Will support
+			 * kernel type WQ in the future.
+			 */
+			if (wq->type == IDXD_WQT_USER)
+				rc = idxd_bind(&idxd_user_drv.drv, wq_name);
+			else
+				rc = -EINVAL;
+			if (rc < 0) {
+				clear_bit(i, idxd->wq_enable_map);
+				dev_err(dev,
+					"HALT: unable to re-enable wq %s\n",
+					dev_name(wq_confdev(wq)));
+			}
+		}
+	}
+out:
+	kfree(idxd->idxd_saved);
+}
+
+static const struct pci_error_handlers idxd_error_handler = {
+	.reset_prepare	= idxd_reset_prepare,
+	.reset_done	= idxd_reset_done,
+};
+
+/*
+ * Probe idxd PCI device.
+ * If idxd is not given, need to allocate idxd and set up its data.
+ *
+ * If idxd is given, idxd was allocated and setup already. Just need to
+ * configure device without re-allocating and re-configuring idxd data.
+ * This is useful for recovering from FLR.
+ */
+int idxd_pci_probe_alloc(struct idxd_device *idxd, struct pci_dev *pdev,
+			 const struct pci_device_id *id)
+{
+	bool alloc_idxd = idxd ? false : true;
+	struct idxd_driver_data *data;
+	struct device *dev;
+	int rc;
+
+	pdev = idxd ? idxd->pdev : pdev;
+	dev = &pdev->dev;
+	data = id ? (struct idxd_driver_data *)id->driver_data : NULL;
 	rc = pci_enable_device(pdev);
 	if (rc)
 		return rc;

-	dev_dbg(dev, "Alloc IDXD context\n");
-	idxd = idxd_alloc(pdev, data);
-	if (!idxd) {
-		rc = -ENOMEM;
-		goto err_idxd_alloc;
-	}
+	if (alloc_idxd) {
+		dev_dbg(dev, "Alloc IDXD context\n");
+		idxd = idxd_alloc(pdev, data);
+		if (!idxd) {
+			rc = -ENOMEM;
+			goto err_idxd_alloc;
+		}

-	dev_dbg(dev, "Mapping BARs\n");
-	idxd->reg_base = pci_iomap(pdev, IDXD_MMIO_BAR, 0);
-	if (!idxd->reg_base) {
-		rc = -ENOMEM;
-		goto err_iomap;
-	}
+		dev_dbg(dev, "Mapping BARs\n");
+		idxd->reg_base = pci_iomap(pdev, IDXD_MMIO_BAR, 0);
+		if (!idxd->reg_base) {
+			rc = -ENOMEM;
+			goto err_iomap;
+		}

-	dev_dbg(dev, "Set DMA masks\n");
-	rc = dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(64));
-	if (rc)
-		goto err;
+		dev_dbg(dev, "Set DMA masks\n");
+		rc = dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(64));
+		if (rc)
+			goto err;
+	}

 	dev_dbg(dev, "Set PCI master\n");
 	pci_set_master(pdev);
 	pci_set_drvdata(pdev, idxd);

-	idxd->hw.version = ioread32(idxd->reg_base + IDXD_VER_OFFSET);
-	rc = idxd_probe(idxd);
-	if (rc) {
-		dev_err(dev, "Intel(R) IDXD DMA Engine init failed\n");
-		goto err;
-	}
+	if (alloc_idxd) {
+		idxd->hw.version = ioread32(idxd->reg_base + IDXD_VER_OFFSET);
+		rc = idxd_probe(idxd);
+		if (rc) {
+			dev_err(dev, "Intel(R) IDXD DMA Engine init failed\n");
+			goto err;
+		}

-	if (data->load_device_defaults) {
-		rc = data->load_device_defaults(idxd);
+		if (data->load_device_defaults) {
+			rc = data->load_device_defaults(idxd);
+			if (rc)
+				dev_warn(dev, "IDXD loading device defaults failed\n");
+		}
+
+		rc = idxd_register_devices(idxd);
+		if (rc) {
+			dev_err(dev, "IDXD sysfs setup failed\n");
+			goto err_dev_register;
+		}
+
+		rc = idxd_device_init_debugfs(idxd);
 		if (rc)
-			dev_warn(dev, "IDXD loading device defaults failed\n");
+			dev_warn(dev, "IDXD debugfs failed to setup\n");
 	}

-	rc = idxd_register_devices(idxd);
-	if (rc) {
-		dev_err(dev, "IDXD sysfs setup failed\n");
-		goto err_dev_register;
-	}
+	if (!alloc_idxd) {
+		/* Release interrupts in the IDXD device. */
+		idxd_cleanup_interrupts(idxd);

-	rc = idxd_device_init_debugfs(idxd);
-	if (rc)
-		dev_warn(dev, "IDXD debugfs failed to setup\n");
+		/* Re-enable interrupts in the IDXD device. */
+		rc = idxd_setup_interrupts(idxd);
+		if (rc)
+			dev_warn(dev, "IDXD interrupts failed to setup\n");
+	}

 	dev_info(&pdev->dev, "Intel(R) Accelerator Device (v%x)\n",
 		 idxd->hw.version);

-	idxd->user_submission_safe = data->user_submission_safe;
+	if (data)
+		idxd->user_submission_safe = data->user_submission_safe;

 	return 0;

@ -798,6 +1197,11 @@ static int idxd_pci_probe(struct pci_dev *pdev, const struct pci_device_id *id)
 	return rc;
 }

+static int idxd_pci_probe(struct pci_dev *pdev, const struct pci_device_id *id)
+{
+	return idxd_pci_probe_alloc(NULL, pdev, id);
+}
+
 void idxd_wqs_quiesce(struct idxd_device *idxd)
 {
 	struct idxd_wq *wq;
@ -864,6 +1268,7 @@ static struct pci_driver idxd_pci_driver = {
 	.probe		= idxd_pci_probe,
 	.remove		= idxd_remove,
 	.shutdown	= idxd_shutdown,
+	.err_handler	= &idxd_error_handler,
 };

 static int __init idxd_init_module(void)
--- a/drivers/dma/idxd/irq.c
+++ b/drivers/dma/idxd/irq.c
@ -383,15 +383,65 @@ static void process_evl_entries(struct idxd_device *idxd)
 	mutex_unlock(&evl->lock);
 }

+static void idxd_device_flr(struct work_struct *work)
+{
+	struct idxd_device *idxd = container_of(work, struct idxd_device, work);
+	int rc;
+
+	/*
+	 * IDXD device requires a Function Level Reset (FLR).
+	 * pci_reset_function() will reset the device with FLR.
+	 */
+	rc = pci_reset_function(idxd->pdev);
+	if (rc)
+		dev_err(&idxd->pdev->dev, "FLR failed\n");
+}
+
+static irqreturn_t idxd_halt(struct idxd_device *idxd)
+{
+	union gensts_reg gensts;
+
+	gensts.bits = ioread32(idxd->reg_base + IDXD_GENSTATS_OFFSET);
+	if (gensts.state == IDXD_DEVICE_STATE_HALT) {
+		idxd->state = IDXD_DEV_HALTED;
+		if (gensts.reset_type == IDXD_DEVICE_RESET_SOFTWARE) {
+			/*
+			 * If we need a software reset, we will throw the work
+			 * on a system workqueue in order to allow interrupts
+			 * for the device command completions.
+			 */
+			INIT_WORK(&idxd->work, idxd_device_reinit);
+			queue_work(idxd->wq, &idxd->work);
+		} else if (gensts.reset_type == IDXD_DEVICE_RESET_FLR) {
+			idxd->state = IDXD_DEV_HALTED;
+			idxd_mask_error_interrupts(idxd);
+			dev_dbg(&idxd->pdev->dev,
+				"idxd halted, doing FLR. After FLR, configs are restored\n");
+			INIT_WORK(&idxd->work, idxd_device_flr);
+			queue_work(idxd->wq, &idxd->work);
+
+		} else {
+			idxd->state = IDXD_DEV_HALTED;
+			idxd_wqs_quiesce(idxd);
+			idxd_wqs_unmap_portal(idxd);
+			idxd_device_clear_state(idxd);
+			dev_err(&idxd->pdev->dev,
+				"idxd halted, need system reset");
+
+			return -ENXIO;
+		}
+	}
+
+	return IRQ_HANDLED;
+}
+
 irqreturn_t idxd_misc_thread(int vec, void *data)
 {
 	struct idxd_irq_entry *irq_entry = data;
 	struct idxd_device *idxd = ie_to_idxd(irq_entry);
 	struct device *dev = &idxd->pdev->dev;
-	union gensts_reg gensts;
 	u32 val = 0;
 	int i;
-	bool err = false;
 	u32 cause;

 	cause = ioread32(idxd->reg_base + IDXD_INTCAUSE_OFFSET);
@ -401,7 +451,7 @@ irqreturn_t idxd_misc_thread(int vec, void *data)
 	iowrite32(cause, idxd->reg_base + IDXD_INTCAUSE_OFFSET);

 	if (cause & IDXD_INTC_HALT_STATE)
-		goto halt;
+		return idxd_halt(idxd);

 	if (cause & IDXD_INTC_ERR) {
 		spin_lock(&idxd->dev_lock);
@ -435,7 +485,6 @@ irqreturn_t idxd_misc_thread(int vec, void *data)
 		for (i = 0; i < 4; i++)
 			dev_warn_ratelimited(dev, "err[%d]: %#16.16llx\n",
 					     i, idxd->sw_err.bits[i]);
-		err = true;
 	}

 	if (cause & IDXD_INTC_INT_HANDLE_REVOKED) {
@ -480,34 +529,6 @@ irqreturn_t idxd_misc_thread(int vec, void *data)
 		dev_warn_once(dev, "Unexpected interrupt cause bits set: %#x\n",
 			      val);

-	if (!err)
-		goto out;
-
-halt:
-	gensts.bits = ioread32(idxd->reg_base + IDXD_GENSTATS_OFFSET);
-	if (gensts.state == IDXD_DEVICE_STATE_HALT) {
-		idxd->state = IDXD_DEV_HALTED;
-		if (gensts.reset_type == IDXD_DEVICE_RESET_SOFTWARE) {
-			/*
-			 * If we need a software reset, we will throw the work
-			 * on a system workqueue in order to allow interrupts
-			 * for the device command completions.
-			 */
-			INIT_WORK(&idxd->work, idxd_device_reinit);
-			queue_work(idxd->wq, &idxd->work);
-		} else {
-			idxd->state = IDXD_DEV_HALTED;
-			idxd_wqs_quiesce(idxd);
-			idxd_wqs_unmap_portal(idxd);
-			idxd_device_clear_state(idxd);
-			dev_err(&idxd->pdev->dev,
-				"idxd halted, need %s.\n",
-				gensts.reset_type == IDXD_DEVICE_RESET_FLR ?
-				"FLR" : "system reset");
-		}
-	}
-
-out:
 	return IRQ_HANDLED;
 }

--- a/drivers/dma/idxd/registers.h
+++ b/drivers/dma/idxd/registers.h
@ -9,6 +9,7 @@
 #define PCI_DEVICE_ID_INTEL_DSA_GNRD	0x11fb
 #define PCI_DEVICE_ID_INTEL_DSA_DMR	0x1212
 #define PCI_DEVICE_ID_INTEL_IAA_DMR	0x1216
+#define PCI_DEVICE_ID_INTEL_IAA_PTL	0xb02d

 #define DEVICE_VERSION_1		0x100
 #define DEVICE_VERSION_2		0x200
--- a/drivers/dma/idxd/sysfs.c
+++ b/drivers/dma/idxd/sysfs.c
@ -1979,13 +1979,3 @@ void idxd_unregister_devices(struct idxd_device *idxd)
 		device_unregister(group_confdev(group));
 	}
 }
-
-int idxd_register_bus_type(void)
-{
-	return bus_register(&dsa_bus_type);
-}
-
-void idxd_unregister_bus_type(void)
-{
-	bus_unregister(&dsa_bus_type);
-}
--- a/drivers/dma/mv_xor.c
+++ b/drivers/dma/mv_xor.c
@ -1369,10 +1369,9 @@ static int mv_xor_probe(struct platform_device *pdev)
 		return 0;

 	if (pdev->dev.of_node) {
-		struct device_node *np;
 		int i = 0;

-		for_each_child_of_node(pdev->dev.of_node, np) {
+		for_each_child_of_node_scoped(pdev->dev.of_node, np) {
 			struct mv_xor_chan *chan;
 			dma_cap_mask_t cap_mask;
 			int irq;
@ -1388,7 +1387,6 @@ static int mv_xor_probe(struct platform_device *pdev)
 			irq = irq_of_parse_and_map(np, 0);
 			if (!irq) {
 				ret = -ENODEV;
-				of_node_put(np);
 				goto err_channel_add;
 			}

@ -1397,7 +1395,6 @@ static int mv_xor_probe(struct platform_device *pdev)
 			if (IS_ERR(chan)) {
 				ret = PTR_ERR(chan);
 				irq_dispose_mapping(irq);
-				of_node_put(np);
 				goto err_channel_add;
 			}

--- a/drivers/dma/ptdma/Kconfig
+++ b/drivers/dma/ptdma/Kconfig
@ -1,13 +0,0 @@
-# SPDX-License-Identifier: GPL-2.0-only
-config AMD_PTDMA
-	tristate  "AMD PassThru DMA Engine"
-	depends on X86_64 && PCI
-	select DMA_ENGINE
-	select DMA_VIRTUAL_CHANNELS
-	help
-	  Enable support for the AMD PTDMA controller. This controller
-	  provides DMA capabilities to perform high bandwidth memory to
-	  memory and IO copy operations. It performs DMA transfer through
-	  queue-based descriptor management. This DMA controller is intended
-	  to be used with AMD Non-Transparent Bridge devices and not for
-	  general purpose peripheral DMA.
--- a/drivers/dma/qcom/bam_dma.c
+++ b/drivers/dma/qcom/bam_dma.c
@ -59,6 +59,9 @@ struct bam_desc_hw {
 #define DESC_FLAG_NWD BIT(12)
 #define DESC_FLAG_CMD BIT(11)

+#define BAM_NDP_REVISION_START	0x20
+#define BAM_NDP_REVISION_END	0x27
+
 struct bam_async_desc {
 	struct virt_dma_desc vd;

@ -398,6 +401,7 @@ struct bam_device {

 	/* dma start transaction tasklet */
 	struct tasklet_struct task;
+	u32 bam_revision;
 };

 /**
@ -441,8 +445,10 @@ static void bam_reset(struct bam_device *bdev)
 	writel_relaxed(val, bam_addr(bdev, 0, BAM_CTRL));

 	/* set descriptor threshold, start with 4 bytes */
-	writel_relaxed(DEFAULT_CNT_THRSHLD,
-			bam_addr(bdev, 0, BAM_DESC_CNT_TRSHLD));
+	if (in_range(bdev->bam_revision, BAM_NDP_REVISION_START,
+		     BAM_NDP_REVISION_END))
+		writel_relaxed(DEFAULT_CNT_THRSHLD,
+			       bam_addr(bdev, 0, BAM_DESC_CNT_TRSHLD));

 	/* Enable default set of h/w workarounds, ie all except BAM_FULL_PIPE */
 	writel_relaxed(BAM_CNFG_BITS_DEFAULT, bam_addr(bdev, 0, BAM_CNFG_BITS));
@ -1000,9 +1006,10 @@ static void bam_apply_new_config(struct bam_chan *bchan,
 			maxburst = bchan->slave.src_maxburst;
 		else
 			maxburst = bchan->slave.dst_maxburst;
-
-		writel_relaxed(maxburst,
-			       bam_addr(bdev, 0, BAM_DESC_CNT_TRSHLD));
+		if (in_range(bdev->bam_revision, BAM_NDP_REVISION_START,
+			     BAM_NDP_REVISION_END))
+			writel_relaxed(maxburst,
+				       bam_addr(bdev, 0, BAM_DESC_CNT_TRSHLD));
 	}

 	bchan->reconfigure = 0;
@ -1192,10 +1199,11 @@ static int bam_init(struct bam_device *bdev)
 	u32 val;

 	/* read revision and configuration information */
-	if (!bdev->num_ees) {
-		val = readl_relaxed(bam_addr(bdev, 0, BAM_REVISION));
+	val = readl_relaxed(bam_addr(bdev, 0, BAM_REVISION));
+	if (!bdev->num_ees)
 		bdev->num_ees = (val >> NUM_EES_SHIFT) & NUM_EES_MASK;
-	}
+
+	bdev->bam_revision = val & REVISION_MASK;

 	/* check that configured EE is within range */
 	if (bdev->ee >= bdev->num_ees)
--- a/drivers/dma/qcom/gpi.c
+++ b/drivers/dma/qcom/gpi.c
@ -18,6 +18,7 @@
 #include "../virt-dma.h"

 #define TRE_TYPE_DMA		0x10
+#define TRE_TYPE_IMMEDIATE_DMA	0x11
 #define TRE_TYPE_GO		0x20
 #define TRE_TYPE_CONFIG0	0x22

@ -64,6 +65,7 @@

 /* DMA TRE */
 #define TRE_DMA_LEN		GENMASK(23, 0)
+#define TRE_DMA_IMMEDIATE_LEN	GENMASK(3, 0)

 /* Register offsets from gpi-top */
 #define GPII_n_CH_k_CNTXT_0_OFFS(n, k)	(0x20000 + (0x4000 * (n)) + (0x80 * (k)))
@ -1711,6 +1713,7 @@ static int gpi_create_spi_tre(struct gchan *chan, struct gpi_desc *desc,
 	dma_addr_t address;
 	struct gpi_tre *tre;
 	unsigned int i;
+	int len;

 	/* first create config tre if applicable */
 	if (direction == DMA_MEM_TO_DEV && spi->set_config) {
@ -1763,14 +1766,30 @@ static int gpi_create_spi_tre(struct gchan *chan, struct gpi_desc *desc,
 	tre_idx++;

 	address = sg_dma_address(sgl);
-	tre->dword[0] = lower_32_bits(address);
-	tre->dword[1] = upper_32_bits(address);
+	len = sg_dma_len(sgl);

-	tre->dword[2] = u32_encode_bits(sg_dma_len(sgl), TRE_DMA_LEN);
+	/* Support Immediate dma for write transfers for data length up to 8 bytes */
+	if (direction == DMA_MEM_TO_DEV && len <= 2 * sizeof(tre->dword[0])) {
+		/*
+		 * For Immediate dma, data length may not always be length of 8 bytes,
+		 * it can be length less than 8, hence initialize both dword's with 0
+		 */
+		tre->dword[0] = 0;
+		tre->dword[1] = 0;
+		memcpy(&tre->dword[0], sg_virt(sgl), len);

-	tre->dword[3] = u32_encode_bits(TRE_TYPE_DMA, TRE_FLAGS_TYPE);
-	if (direction == DMA_MEM_TO_DEV)
-		tre->dword[3] |= u32_encode_bits(1, TRE_FLAGS_IEOT);
+		tre->dword[2] = u32_encode_bits(len, TRE_DMA_IMMEDIATE_LEN);
+		tre->dword[3] = u32_encode_bits(TRE_TYPE_IMMEDIATE_DMA, TRE_FLAGS_TYPE);
+	} else {
+		tre->dword[0] = lower_32_bits(address);
+		tre->dword[1] = upper_32_bits(address);
+
+		tre->dword[2] = u32_encode_bits(len, TRE_DMA_LEN);
+		tre->dword[3] = u32_encode_bits(TRE_TYPE_DMA, TRE_FLAGS_TYPE);
+	}
+
+	tre->dword[3] |= u32_encode_bits(direction == DMA_MEM_TO_DEV,
+					 TRE_FLAGS_IEOT);

 	for (i = 0; i < tre_idx; i++)
 		dev_dbg(dev, "TRE:%d %x:%x:%x:%x\n", i, desc->tre[i].dword[0],
--- a/drivers/dma/sh/rcar-dmac.c
+++ b/drivers/dma/sh/rcar-dmac.c
@ -2023,6 +2023,10 @@ static const struct of_device_id rcar_dmac_of_ids[] = {
 		.compatible = "renesas,rcar-gen4-dmac",
 		.data = &rcar_gen4_dmac_data,
 	}, {
+		/*
+		 * Backward compatibility for between v5.12 - v5.19
+		 * which didn't combined with "renesas,rcar-gen4-dmac"
+		 */
 		.compatible = "renesas,dmac-r8a779a0",
 		.data = &rcar_gen4_dmac_data,
 	},
--- a/drivers/dma/sun4i-dma.c
+++ b/drivers/dma/sun4i-dma.c
@ -13,7 +13,9 @@
 #include <linux/interrupt.h>
 #include <linux/module.h>
 #include <linux/of_dma.h>
+#include <linux/of_device.h>
 #include <linux/platform_device.h>
+#include <linux/reset.h>
 #include <linux/slab.h>
 #include <linux/spinlock.h>

@ -31,12 +33,21 @@
 #define SUN4I_DMA_CFG_SRC_ADDR_MODE(mode)	((mode) << 5)
 #define SUN4I_DMA_CFG_SRC_DRQ_TYPE(type)	(type)

+#define SUNIV_DMA_CFG_DST_DATA_WIDTH(width)	((width) << 24)
+#define SUNIV_DMA_CFG_SRC_DATA_WIDTH(width)	((width) << 8)
+
+#define SUN4I_MAX_BURST	8
+#define SUNIV_MAX_BURST	4
+
 /** Normal DMA register values **/

 /* Normal DMA source/destination data request type values */
 #define SUN4I_NDMA_DRQ_TYPE_SDRAM		0x16
 #define SUN4I_NDMA_DRQ_TYPE_LIMIT		(0x1F + 1)

+#define SUNIV_NDMA_DRQ_TYPE_SDRAM		0x11
+#define SUNIV_NDMA_DRQ_TYPE_LIMIT		(0x17 + 1)
+
 /** Normal DMA register layout **/

 /* Dedicated DMA source/destination address mode values */
@ -50,6 +61,9 @@
 #define SUN4I_NDMA_CFG_BYTE_COUNT_MODE_REMAIN	BIT(15)
 #define SUN4I_NDMA_CFG_SRC_NON_SECURE		BIT(6)

+#define SUNIV_NDMA_CFG_CONT_MODE		BIT(29)
+#define SUNIV_NDMA_CFG_WAIT_STATE(n)		((n) << 26)
+
 /** Dedicated DMA register values **/

 /* Dedicated DMA source/destination address mode values */
@ -62,6 +76,9 @@
 #define SUN4I_DDMA_DRQ_TYPE_SDRAM		0x1
 #define SUN4I_DDMA_DRQ_TYPE_LIMIT		(0x1F + 1)

+#define SUNIV_DDMA_DRQ_TYPE_SDRAM		0x1
+#define SUNIV_DDMA_DRQ_TYPE_LIMIT		(0x9 + 1)
+
 /** Dedicated DMA register layout **/

 /* Dedicated DMA configuration register layout */
@ -115,6 +132,11 @@
 #define SUN4I_DMA_NR_MAX_VCHANS						\
 	(SUN4I_NDMA_NR_MAX_VCHANS + SUN4I_DDMA_NR_MAX_VCHANS)

+#define SUNIV_NDMA_NR_MAX_CHANNELS	4
+#define SUNIV_DDMA_NR_MAX_CHANNELS	4
+#define SUNIV_NDMA_NR_MAX_VCHANS	(24 * 2 - 1)
+#define SUNIV_DDMA_NR_MAX_VCHANS	10
+
 /* This set of SUN4I_DDMA timing parameters were found experimentally while
 * working with the SPI driver and seem to make it behave correctly */
 #define SUN4I_DDMA_MAGIC_SPI_PARAMETERS \
@ -132,6 +154,33 @@
 #define SUN4I_DDMA_MAX_SEG_SIZE		SZ_16M
 #define SUN4I_DMA_MAX_SEG_SIZE		SUN4I_NDMA_MAX_SEG_SIZE

+/*
+ * Hardware channels / ports representation
+ *
+ * The hardware is used in several SoCs, with differing numbers
+ * of channels and endpoints. This structure ties those numbers
+ * to a certain compatible string.
+ */
+struct sun4i_dma_config {
+	u32 ndma_nr_max_channels;
+	u32 ndma_nr_max_vchans;
+
+	u32 ddma_nr_max_channels;
+	u32 ddma_nr_max_vchans;
+
+	u32 dma_nr_max_channels;
+
+	void (*set_dst_data_width)(u32 *p_cfg, s8 data_width);
+	void (*set_src_data_width)(u32 *p_cfg, s8 data_width);
+	int (*convert_burst)(u32 maxburst);
+
+	u8 ndma_drq_sdram;
+	u8 ddma_drq_sdram;
+
+	u8 max_burst;
+	bool has_reset;
+};
+
 struct sun4i_dma_pchan {
 	/* Register base of channel */
 	void __iomem			*base;
@ -170,7 +219,7 @@ struct sun4i_dma_contract {
 };

 struct sun4i_dma_dev {
-	DECLARE_BITMAP(pchans_used, SUN4I_DMA_NR_MAX_CHANNELS);
+	unsigned long *pchans_used;
 	struct dma_device		slave;
 	struct sun4i_dma_pchan		*pchans;
 	struct sun4i_dma_vchan		*vchans;
@ -178,6 +227,8 @@ struct sun4i_dma_dev {
 	struct clk			*clk;
 	int				irq;
 	spinlock_t			lock;
+	const struct sun4i_dma_config *cfg;
+	struct reset_control *rst;
 };

 static struct sun4i_dma_dev *to_sun4i_dma_dev(struct dma_device *dev)
@ -200,7 +251,27 @@ static struct device *chan2dev(struct dma_chan *chan)
 	return &chan->dev->device;
 }

-static int convert_burst(u32 maxburst)
+static void set_dst_data_width_a10(u32 *p_cfg, s8 data_width)
+{
+	*p_cfg |= SUN4I_DMA_CFG_DST_DATA_WIDTH(data_width);
+}
+
+static void set_src_data_width_a10(u32 *p_cfg, s8 data_width)
+{
+	*p_cfg |= SUN4I_DMA_CFG_SRC_DATA_WIDTH(data_width);
+}
+
+static void set_dst_data_width_f1c100s(u32 *p_cfg, s8 data_width)
+{
+	*p_cfg |= SUNIV_DMA_CFG_DST_DATA_WIDTH(data_width);
+}
+
+static void set_src_data_width_f1c100s(u32 *p_cfg, s8 data_width)
+{
+	*p_cfg |= SUNIV_DMA_CFG_SRC_DATA_WIDTH(data_width);
+}
+
+static int convert_burst_a10(u32 maxburst)
 {
 	if (maxburst > 8)
 		return -EINVAL;
@ -209,6 +280,15 @@ static int convert_burst(u32 maxburst)
 	return (maxburst >> 2);
 }

+static int convert_burst_f1c100s(u32 maxburst)
+{
+	if (maxburst > 4)
+		return -EINVAL;
+
+	/* 1 -> 0, 4 -> 1 */
+	return (maxburst >> 2);
+}
+
 static int convert_buswidth(enum dma_slave_buswidth addr_width)
 {
 	if (addr_width > DMA_SLAVE_BUSWIDTH_4_BYTES)
@ -233,15 +313,15 @@ static struct sun4i_dma_pchan *find_and_use_pchan(struct sun4i_dma_dev *priv,
 	int i, max;

 	/*
-	 * pchans 0-SUN4I_NDMA_NR_MAX_CHANNELS are normal, and
-	 * SUN4I_NDMA_NR_MAX_CHANNELS+ are dedicated ones
+	 * pchans 0-priv->cfg->ndma_nr_max_channels are normal, and
+	 * priv->cfg->ndma_nr_max_channels+ are dedicated ones
 	 */
 	if (vchan->is_dedicated) {
-		i = SUN4I_NDMA_NR_MAX_CHANNELS;
-		max = SUN4I_DMA_NR_MAX_CHANNELS;
+		i = priv->cfg->ndma_nr_max_channels;
+		max = priv->cfg->dma_nr_max_channels;
 	} else {
 		i = 0;
-		max = SUN4I_NDMA_NR_MAX_CHANNELS;
+		max = priv->cfg->ndma_nr_max_channels;
 	}

 	spin_lock_irqsave(&priv->lock, flags);
@ -444,6 +524,7 @@ generate_ndma_promise(struct dma_chan *chan, dma_addr_t src, dma_addr_t dest,
 		      size_t len, struct dma_slave_config *sconfig,
 		      enum dma_transfer_direction direction)
 {
+	struct sun4i_dma_dev *priv = to_sun4i_dma_dev(chan->device);
 	struct sun4i_dma_promise *promise;
 	int ret;

@ -467,13 +548,13 @@ generate_ndma_promise(struct dma_chan *chan, dma_addr_t src, dma_addr_t dest,
 		sconfig->src_addr_width, sconfig->dst_addr_width);

 	/* Source burst */
-	ret = convert_burst(sconfig->src_maxburst);
+	ret = priv->cfg->convert_burst(sconfig->src_maxburst);
 	if (ret < 0)
 		goto fail;
 	promise->cfg |= SUN4I_DMA_CFG_SRC_BURST_LENGTH(ret);

 	/* Destination burst */
-	ret = convert_burst(sconfig->dst_maxburst);
+	ret = priv->cfg->convert_burst(sconfig->dst_maxburst);
 	if (ret < 0)
 		goto fail;
 	promise->cfg |= SUN4I_DMA_CFG_DST_BURST_LENGTH(ret);
@ -482,13 +563,13 @@ generate_ndma_promise(struct dma_chan *chan, dma_addr_t src, dma_addr_t dest,
 	ret = convert_buswidth(sconfig->src_addr_width);
 	if (ret < 0)
 		goto fail;
-	promise->cfg |= SUN4I_DMA_CFG_SRC_DATA_WIDTH(ret);
+	priv->cfg->set_src_data_width(&promise->cfg, ret);

 	/* Destination bus width */
 	ret = convert_buswidth(sconfig->dst_addr_width);
 	if (ret < 0)
 		goto fail;
-	promise->cfg |= SUN4I_DMA_CFG_DST_DATA_WIDTH(ret);
+	priv->cfg->set_dst_data_width(&promise->cfg, ret);

 	return promise;

@ -510,6 +591,7 @@ static struct sun4i_dma_promise *
 generate_ddma_promise(struct dma_chan *chan, dma_addr_t src, dma_addr_t dest,
 		      size_t len, struct dma_slave_config *sconfig)
 {
+	struct sun4i_dma_dev *priv = to_sun4i_dma_dev(chan->device);
 	struct sun4i_dma_promise *promise;
 	int ret;

@ -524,13 +606,13 @@ generate_ddma_promise(struct dma_chan *chan, dma_addr_t src, dma_addr_t dest,
 		SUN4I_DDMA_CFG_BYTE_COUNT_MODE_REMAIN;

 	/* Source burst */
-	ret = convert_burst(sconfig->src_maxburst);
+	ret = priv->cfg->convert_burst(sconfig->src_maxburst);
 	if (ret < 0)
 		goto fail;
 	promise->cfg |= SUN4I_DMA_CFG_SRC_BURST_LENGTH(ret);

 	/* Destination burst */
-	ret = convert_burst(sconfig->dst_maxburst);
+	ret = priv->cfg->convert_burst(sconfig->dst_maxburst);
 	if (ret < 0)
 		goto fail;
 	promise->cfg |= SUN4I_DMA_CFG_DST_BURST_LENGTH(ret);
@ -539,13 +621,13 @@ generate_ddma_promise(struct dma_chan *chan, dma_addr_t src, dma_addr_t dest,
 	ret = convert_buswidth(sconfig->src_addr_width);
 	if (ret < 0)
 		goto fail;
-	promise->cfg |= SUN4I_DMA_CFG_SRC_DATA_WIDTH(ret);
+	priv->cfg->set_src_data_width(&promise->cfg, ret);

 	/* Destination bus width */
 	ret = convert_buswidth(sconfig->dst_addr_width);
 	if (ret < 0)
 		goto fail;
-	promise->cfg |= SUN4I_DMA_CFG_DST_DATA_WIDTH(ret);
+	priv->cfg->set_dst_data_width(&promise->cfg, ret);

 	return promise;

@ -622,6 +704,7 @@ static struct dma_async_tx_descriptor *
 sun4i_dma_prep_dma_memcpy(struct dma_chan *chan, dma_addr_t dest,
 			  dma_addr_t src, size_t len, unsigned long flags)
 {
+	struct sun4i_dma_dev *priv = to_sun4i_dma_dev(chan->device);
 	struct sun4i_dma_vchan *vchan = to_sun4i_dma_vchan(chan);
 	struct dma_slave_config *sconfig = &vchan->cfg;
 	struct sun4i_dma_promise *promise;
@ -638,8 +721,8 @@ sun4i_dma_prep_dma_memcpy(struct dma_chan *chan, dma_addr_t dest,
 	 */
 	sconfig->src_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES;
 	sconfig->dst_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES;
-	sconfig->src_maxburst = 8;
-	sconfig->dst_maxburst = 8;
+	sconfig->src_maxburst = priv->cfg->max_burst;
+	sconfig->dst_maxburst = priv->cfg->max_burst;

 	if (vchan->is_dedicated)
 		promise = generate_ddma_promise(chan, src, dest, len, sconfig);
@ -654,11 +737,13 @@ sun4i_dma_prep_dma_memcpy(struct dma_chan *chan, dma_addr_t dest,

 	/* Configure memcpy mode */
 	if (vchan->is_dedicated) {
-		promise->cfg |= SUN4I_DMA_CFG_SRC_DRQ_TYPE(SUN4I_DDMA_DRQ_TYPE_SDRAM) |
-				SUN4I_DMA_CFG_DST_DRQ_TYPE(SUN4I_DDMA_DRQ_TYPE_SDRAM);
+		promise->cfg |=
+			SUN4I_DMA_CFG_SRC_DRQ_TYPE(priv->cfg->ddma_drq_sdram) |
+			SUN4I_DMA_CFG_DST_DRQ_TYPE(priv->cfg->ddma_drq_sdram);
 	} else {
-		promise->cfg |= SUN4I_DMA_CFG_SRC_DRQ_TYPE(SUN4I_NDMA_DRQ_TYPE_SDRAM) |
-				SUN4I_DMA_CFG_DST_DRQ_TYPE(SUN4I_NDMA_DRQ_TYPE_SDRAM);
+		promise->cfg |=
+			SUN4I_DMA_CFG_SRC_DRQ_TYPE(priv->cfg->ndma_drq_sdram) |
+			SUN4I_DMA_CFG_DST_DRQ_TYPE(priv->cfg->ndma_drq_sdram);
 	}

 	/* Fill the contract with our only promise */
@ -673,6 +758,7 @@ sun4i_dma_prep_dma_cyclic(struct dma_chan *chan, dma_addr_t buf, size_t len,
 			  size_t period_len, enum dma_transfer_direction dir,
 			  unsigned long flags)
 {
+	struct sun4i_dma_dev *priv = to_sun4i_dma_dev(chan->device);
 	struct sun4i_dma_vchan *vchan = to_sun4i_dma_vchan(chan);
 	struct dma_slave_config *sconfig = &vchan->cfg;
 	struct sun4i_dma_promise *promise;
@ -696,11 +782,11 @@ sun4i_dma_prep_dma_cyclic(struct dma_chan *chan, dma_addr_t buf, size_t len,
 	if (vchan->is_dedicated) {
 		io_mode = SUN4I_DDMA_ADDR_MODE_IO;
 		linear_mode = SUN4I_DDMA_ADDR_MODE_LINEAR;
-		ram_type = SUN4I_DDMA_DRQ_TYPE_SDRAM;
+		ram_type = priv->cfg->ddma_drq_sdram;
 	} else {
 		io_mode = SUN4I_NDMA_ADDR_MODE_IO;
 		linear_mode = SUN4I_NDMA_ADDR_MODE_LINEAR;
-		ram_type = SUN4I_NDMA_DRQ_TYPE_SDRAM;
+		ram_type = priv->cfg->ndma_drq_sdram;
 	}

 	if (dir == DMA_MEM_TO_DEV) {
@ -793,6 +879,7 @@ sun4i_dma_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
 			unsigned int sg_len, enum dma_transfer_direction dir,
 			unsigned long flags, void *context)
 {
+	struct sun4i_dma_dev *priv = to_sun4i_dma_dev(chan->device);
 	struct sun4i_dma_vchan *vchan = to_sun4i_dma_vchan(chan);
 	struct dma_slave_config *sconfig = &vchan->cfg;
 	struct sun4i_dma_promise *promise;
@ -818,11 +905,11 @@ sun4i_dma_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
 	if (vchan->is_dedicated) {
 		io_mode = SUN4I_DDMA_ADDR_MODE_IO;
 		linear_mode = SUN4I_DDMA_ADDR_MODE_LINEAR;
-		ram_type = SUN4I_DDMA_DRQ_TYPE_SDRAM;
+		ram_type = priv->cfg->ddma_drq_sdram;
 	} else {
 		io_mode = SUN4I_NDMA_ADDR_MODE_IO;
 		linear_mode = SUN4I_NDMA_ADDR_MODE_LINEAR;
-		ram_type = SUN4I_NDMA_DRQ_TYPE_SDRAM;
+		ram_type = priv->cfg->ndma_drq_sdram;
 	}

 	if (dir == DMA_MEM_TO_DEV)
@ -1150,6 +1237,10 @@ static int sun4i_dma_probe(struct platform_device *pdev)
 	if (!priv)
 		return -ENOMEM;

+	priv->cfg = of_device_get_match_data(&pdev->dev);
+	if (!priv->cfg)
+		return -ENODEV;
+
 	priv->base = devm_platform_ioremap_resource(pdev, 0);
 	if (IS_ERR(priv->base))
 		return PTR_ERR(priv->base);
@ -1164,6 +1255,13 @@ static int sun4i_dma_probe(struct platform_device *pdev)
 		return PTR_ERR(priv->clk);
 	}

+	if (priv->cfg->has_reset) {
+		priv->rst = devm_reset_control_get_exclusive_deasserted(&pdev->dev, NULL);
+		if (IS_ERR(priv->rst))
+			return dev_err_probe(&pdev->dev, PTR_ERR(priv->rst),
+					     "Failed to get reset control\n");
+	}
+
 	platform_set_drvdata(pdev, priv);
 	spin_lock_init(&priv->lock);

@ -1197,23 +1295,26 @@ static int sun4i_dma_probe(struct platform_device *pdev)

 	priv->slave.dev = &pdev->dev;

-	priv->pchans = devm_kcalloc(&pdev->dev, SUN4I_DMA_NR_MAX_CHANNELS,
+	priv->pchans = devm_kcalloc(&pdev->dev, priv->cfg->dma_nr_max_channels,
 				    sizeof(struct sun4i_dma_pchan), GFP_KERNEL);
 	priv->vchans = devm_kcalloc(&pdev->dev, SUN4I_DMA_NR_MAX_VCHANS,
 				    sizeof(struct sun4i_dma_vchan), GFP_KERNEL);
-	if (!priv->vchans || !priv->pchans)
+	priv->pchans_used = devm_kcalloc(&pdev->dev,
+					 BITS_TO_LONGS(priv->cfg->dma_nr_max_channels),
+					 sizeof(unsigned long), GFP_KERNEL);
+	if (!priv->vchans || !priv->pchans || !priv->pchans_used)
 		return -ENOMEM;

 	/*
-	 * [0..SUN4I_NDMA_NR_MAX_CHANNELS) are normal pchans, and
-	 * [SUN4I_NDMA_NR_MAX_CHANNELS..SUN4I_DMA_NR_MAX_CHANNELS) are
+	 * [0..priv->cfg->ndma_nr_max_channels) are normal pchans, and
+	 * [priv->cfg->ndma_nr_max_channels..priv->cfg->dma_nr_max_channels) are
 	 * dedicated ones
 	 */
-	for (i = 0; i < SUN4I_NDMA_NR_MAX_CHANNELS; i++)
+	for (i = 0; i < priv->cfg->ndma_nr_max_channels; i++)
 		priv->pchans[i].base = priv->base +
 			SUN4I_NDMA_CHANNEL_REG_BASE(i);

-	for (j = 0; i < SUN4I_DMA_NR_MAX_CHANNELS; i++, j++) {
+	for (j = 0; i < priv->cfg->dma_nr_max_channels; i++, j++) {
 		priv->pchans[i].base = priv->base +
 			SUN4I_DDMA_CHANNEL_REG_BASE(j);
 		priv->pchans[i].is_dedicated = 1;
@ -1284,8 +1385,51 @@ static void sun4i_dma_remove(struct platform_device *pdev)
 	clk_disable_unprepare(priv->clk);
 }

+static struct sun4i_dma_config sun4i_a10_dma_cfg = {
+	.ndma_nr_max_channels	= SUN4I_NDMA_NR_MAX_CHANNELS,
+	.ndma_nr_max_vchans	= SUN4I_NDMA_NR_MAX_VCHANS,
+
+	.ddma_nr_max_channels	= SUN4I_DDMA_NR_MAX_CHANNELS,
+	.ddma_nr_max_vchans	= SUN4I_DDMA_NR_MAX_VCHANS,
+
+	.dma_nr_max_channels	= SUN4I_DMA_NR_MAX_CHANNELS,
+
+	.set_dst_data_width	= set_dst_data_width_a10,
+	.set_src_data_width	= set_src_data_width_a10,
+	.convert_burst		= convert_burst_a10,
+
+	.ndma_drq_sdram		= SUN4I_NDMA_DRQ_TYPE_SDRAM,
+	.ddma_drq_sdram		= SUN4I_DDMA_DRQ_TYPE_SDRAM,
+
+	.max_burst		= SUN4I_MAX_BURST,
+	.has_reset		= false,
+};
+
+static struct sun4i_dma_config suniv_f1c100s_dma_cfg = {
+	.ndma_nr_max_channels	= SUNIV_NDMA_NR_MAX_CHANNELS,
+	.ndma_nr_max_vchans	= SUNIV_NDMA_NR_MAX_VCHANS,
+
+	.ddma_nr_max_channels	= SUNIV_DDMA_NR_MAX_CHANNELS,
+	.ddma_nr_max_vchans	= SUNIV_DDMA_NR_MAX_VCHANS,
+
+	.dma_nr_max_channels	= SUNIV_NDMA_NR_MAX_CHANNELS +
+		SUNIV_DDMA_NR_MAX_CHANNELS,
+
+	.set_dst_data_width	= set_dst_data_width_f1c100s,
+	.set_src_data_width	= set_src_data_width_f1c100s,
+	.convert_burst		= convert_burst_f1c100s,
+
+	.ndma_drq_sdram		= SUNIV_NDMA_DRQ_TYPE_SDRAM,
+	.ddma_drq_sdram		= SUNIV_DDMA_DRQ_TYPE_SDRAM,
+
+	.max_burst		= SUNIV_MAX_BURST,
+	.has_reset		= true,
+};
+
 static const struct of_device_id sun4i_dma_match[] = {
-	{ .compatible = "allwinner,sun4i-a10-dma" },
+	{ .compatible = "allwinner,sun4i-a10-dma", .data = &sun4i_a10_dma_cfg },
+	{ .compatible = "allwinner,suniv-f1c100s-dma",
+		.data = &suniv_f1c100s_dma_cfg },
 	{ /* sentinel */ },
 };
 MODULE_DEVICE_TABLE(of, sun4i_dma_match);
--- a/drivers/dma/tegra210-adma.c
+++ b/drivers/dma/tegra210-adma.c
@ -43,6 +43,10 @@
 #define ADMA_CH_CONFIG_MAX_BUFS				8
 #define TEGRA186_ADMA_CH_CONFIG_OUTSTANDING_REQS(reqs)	(reqs << 4)

+#define TEGRA186_ADMA_GLOBAL_PAGE_CHGRP			0x30
+#define TEGRA186_ADMA_GLOBAL_PAGE_RX_REQ		0x70
+#define TEGRA186_ADMA_GLOBAL_PAGE_TX_REQ		0x84
+
 #define ADMA_CH_FIFO_CTRL				0x2c
 #define ADMA_CH_TX_FIFO_SIZE_SHIFT			8
 #define ADMA_CH_RX_FIFO_SIZE_SHIFT			0
@ -96,6 +100,7 @@ struct tegra_adma_chip_data {
 	unsigned int ch_fifo_size_mask;
 	unsigned int sreq_index_offset;
 	bool has_outstanding_reqs;
+	void (*set_global_pg_config)(struct tegra_adma *tdma);
 };

 /*
@ -151,6 +156,7 @@ struct tegra_adma {
 	struct dma_device		dma_dev;
 	struct device			*dev;
 	void __iomem			*base_addr;
+	void __iomem			*ch_base_addr;
 	struct clk			*ahub_clk;
 	unsigned int			nr_channels;
 	unsigned long			*dma_chan_mask;
@ -159,6 +165,7 @@ struct tegra_adma {

 	/* Used to store global command register state when suspending */
 	unsigned int			global_cmd;
+	unsigned int			ch_page_no;

 	const struct tegra_adma_chip_data *cdata;

@ -176,6 +183,11 @@ static inline u32 tdma_read(struct tegra_adma *tdma, u32 reg)
 	return readl(tdma->base_addr + tdma->cdata->global_reg_offset + reg);
 }

+static inline void tdma_ch_global_write(struct tegra_adma *tdma, u32 reg, u32 val)
+{
+	writel(val, tdma->ch_base_addr + tdma->cdata->global_reg_offset + reg);
+}
+
 static inline void tdma_ch_write(struct tegra_adma_chan *tdc, u32 reg, u32 val)
 {
 	writel(val, tdc->chan_addr + reg);
@ -217,13 +229,30 @@ static int tegra_adma_slave_config(struct dma_chan *dc,
 	return 0;
 }

+static void tegra186_adma_global_page_config(struct tegra_adma *tdma)
+{
+	/*
+	 * Clear the default page1 channel group configs and program
+	 * the global registers based on the actual page usage
+	 */
+	tdma_write(tdma, TEGRA186_ADMA_GLOBAL_PAGE_CHGRP, 0);
+	tdma_write(tdma, TEGRA186_ADMA_GLOBAL_PAGE_RX_REQ, 0);
+	tdma_write(tdma, TEGRA186_ADMA_GLOBAL_PAGE_TX_REQ, 0);
+	tdma_write(tdma, TEGRA186_ADMA_GLOBAL_PAGE_CHGRP + (tdma->ch_page_no * 0x4), 0xff);
+	tdma_write(tdma, TEGRA186_ADMA_GLOBAL_PAGE_RX_REQ + (tdma->ch_page_no * 0x4), 0x1ffffff);
+	tdma_write(tdma, TEGRA186_ADMA_GLOBAL_PAGE_TX_REQ + (tdma->ch_page_no * 0x4), 0xffffff);
+}
+
 static int tegra_adma_init(struct tegra_adma *tdma)
 {
 	u32 status;
 	int ret;

-	/* Clear any interrupts */
-	tdma_write(tdma, tdma->cdata->ch_base_offset + tdma->cdata->global_int_clear, 0x1);
+	/* Clear any channels group global interrupts */
+	tdma_ch_global_write(tdma, tdma->cdata->global_int_clear, 0x1);
+
+	if (!tdma->base_addr)
+		return 0;

 	/* Assert soft reset */
 	tdma_write(tdma, ADMA_GLOBAL_SOFT_RESET, 0x1);
@ -237,6 +266,9 @@ static int tegra_adma_init(struct tegra_adma *tdma)
 	if (ret)
 		return ret;

+	if (tdma->cdata->set_global_pg_config)
+		tdma->cdata->set_global_pg_config(tdma);
+
 	/* Enable global ADMA registers */
 	tdma_write(tdma, ADMA_GLOBAL_CMD, 1);

@ -736,7 +768,9 @@ static int __maybe_unused tegra_adma_runtime_suspend(struct device *dev)
 	struct tegra_adma_chan *tdc;
 	int i;

-	tdma->global_cmd = tdma_read(tdma, ADMA_GLOBAL_CMD);
+	if (tdma->base_addr)
+		tdma->global_cmd = tdma_read(tdma, ADMA_GLOBAL_CMD);
+
 	if (!tdma->global_cmd)
 		goto clk_disable;

@ -777,7 +811,11 @@ static int __maybe_unused tegra_adma_runtime_resume(struct device *dev)
 		dev_err(dev, "ahub clk_enable failed: %d\n", ret);
 		return ret;
 	}
-	tdma_write(tdma, ADMA_GLOBAL_CMD, tdma->global_cmd);
+	if (tdma->base_addr) {
+		tdma_write(tdma, ADMA_GLOBAL_CMD, tdma->global_cmd);
+		if (tdma->cdata->set_global_pg_config)
+			tdma->cdata->set_global_pg_config(tdma);
+	}

 	if (!tdma->global_cmd)
 		return 0;
@ -817,6 +855,7 @@ static const struct tegra_adma_chip_data tegra210_chip_data = {
 	.ch_fifo_size_mask	= 0xf,
 	.sreq_index_offset	= 2,
 	.has_outstanding_reqs	= false,
+	.set_global_pg_config	= NULL,
 };

 static const struct tegra_adma_chip_data tegra186_chip_data = {
@ -833,6 +872,7 @@ static const struct tegra_adma_chip_data tegra186_chip_data = {
 	.ch_fifo_size_mask	= 0x1f,
 	.sreq_index_offset	= 4,
 	.has_outstanding_reqs	= true,
+	.set_global_pg_config	= tegra186_adma_global_page_config,
 };

 static const struct of_device_id tegra_adma_of_match[] = {
@ -846,7 +886,8 @@ static int tegra_adma_probe(struct platform_device *pdev)
 {
 	const struct tegra_adma_chip_data *cdata;
 	struct tegra_adma *tdma;
-	int ret, i;
+	struct resource *res_page, *res_base;
+	int ret, i, page_no;

 	cdata = of_device_get_match_data(&pdev->dev);
 	if (!cdata) {
@ -865,9 +906,35 @@ static int tegra_adma_probe(struct platform_device *pdev)
 	tdma->nr_channels = cdata->nr_channels;
 	platform_set_drvdata(pdev, tdma);

-	tdma->base_addr = devm_platform_ioremap_resource(pdev, 0);
-	if (IS_ERR(tdma->base_addr))
-		return PTR_ERR(tdma->base_addr);
+	res_page = platform_get_resource_byname(pdev, IORESOURCE_MEM, "page");
+	if (res_page) {
+		tdma->ch_base_addr = devm_ioremap_resource(&pdev->dev, res_page);
+		if (IS_ERR(tdma->ch_base_addr))
+			return PTR_ERR(tdma->ch_base_addr);
+
+		res_base = platform_get_resource_byname(pdev, IORESOURCE_MEM, "global");
+		if (res_base) {
+			page_no = (res_page->start - res_base->start) / cdata->ch_base_offset;
+			if (page_no <= 0)
+				return -EINVAL;
+			tdma->ch_page_no = page_no - 1;
+			tdma->base_addr = devm_ioremap_resource(&pdev->dev, res_base);
+			if (IS_ERR(tdma->base_addr))
+				return PTR_ERR(tdma->base_addr);
+		}
+	} else {
+		/* If no 'page' property found, then reg DT binding would be legacy */
+		res_base = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+		if (res_base) {
+			tdma->base_addr = devm_ioremap_resource(&pdev->dev, res_base);
+			if (IS_ERR(tdma->base_addr))
+				return PTR_ERR(tdma->base_addr);
+		} else {
+			return -ENODEV;
+		}
+
+		tdma->ch_base_addr = tdma->base_addr + cdata->ch_base_offset;
+	}

 	tdma->ahub_clk = devm_clk_get(&pdev->dev, "d_audio");
 	if (IS_ERR(tdma->ahub_clk)) {
@ -900,8 +967,7 @@ static int tegra_adma_probe(struct platform_device *pdev)
 		if (!test_bit(i, tdma->dma_chan_mask))
 			continue;

-		tdc->chan_addr = tdma->base_addr + cdata->ch_base_offset
-				 + (cdata->ch_reg_size * i);
+		tdc->chan_addr = tdma->ch_base_addr + (cdata->ch_reg_size * i);

 		tdc->irq = of_irq_get(pdev->dev.of_node, i);
 		if (tdc->irq <= 0) {
--- a/drivers/dma/ti/edma.c
+++ b/drivers/dma/ti/edma.c
@ -208,7 +208,6 @@ struct edma_desc {
 struct edma_cc;

 struct edma_tc {
-	struct device_node		*node;
 	u16				id;
 };

@ -2460,19 +2459,19 @@ static int edma_probe(struct platform_device *pdev)
 			goto err_reg1;
 		}

-		for (i = 0;; i++) {
+		for (i = 0; i < ecc->num_tc; i++) {
 			ret = of_parse_phandle_with_fixed_args(node, "ti,tptcs",
 							       1, i, &tc_args);
-			if (ret || i == ecc->num_tc)
+			if (ret)
 				break;

-			ecc->tc_list[i].node = tc_args.np;
 			ecc->tc_list[i].id = i;
 			queue_priority_mapping[i][1] = tc_args.args[0];
 			if (queue_priority_mapping[i][1] > lowest_priority) {
 				lowest_priority = queue_priority_mapping[i][1];
 				info->default_queue = i;
 			}
+			of_node_put(tc_args.np);
 		}

 		/* See if we have optional dma-channel-mask array */
--- a/drivers/dma/ti/k3-udma.c
+++ b/drivers/dma/ti/k3-udma.c
@ -4404,6 +4404,18 @@ static struct udma_match_data j721s2_bcdma_csi_data = {
 	.soc_data = &j721s2_bcdma_csi_soc_data,
 };

+static struct udma_match_data j722s_bcdma_csi_data = {
+	.type = DMA_TYPE_BCDMA,
+	.psil_base = 0x3100,
+	.enable_memcpy_support = false,
+	.burst_size = {
+		TI_SCI_RM_UDMAP_CHAN_BURST_SIZE_64_BYTES, /* Normal Channels */
+		0, /* No H Channels */
+		0, /* No UH Channels */
+	},
+	.soc_data = &j721s2_bcdma_csi_soc_data,
+};
+
 static const struct of_device_id udma_of_match[] = {
 	{
 		.compatible = "ti,am654-navss-main-udmap",
@ -4435,6 +4447,10 @@ static const struct of_device_id udma_of_match[] = {
 		.compatible = "ti,j721s2-dmss-bcdma-csi",
 		.data = &j721s2_bcdma_csi_data,
 	},
+	{
+		.compatible = "ti,j722s-dmss-bcdma-csi",
+		.data = &j722s_bcdma_csi_data,
+	},
 	{ /* Sentinel */ },
 };
 MODULE_DEVICE_TABLE(of, udma_of_match);
--- a/drivers/dma/xilinx/xdma.c
+++ b/drivers/dma/xilinx/xdma.c
@ -390,15 +390,11 @@ static int xdma_xfer_start(struct xdma_chan *xchan)
 */
 static int xdma_xfer_stop(struct xdma_chan *xchan)
 {
-	int ret;
 	struct xdma_device *xdev = xchan->xdev_hdl;

 	/* clear run stop bit to prevent any further auto-triggering */
-	ret = regmap_write(xdev->rmap, xchan->base + XDMA_CHAN_CONTROL_W1C,
-			   CHAN_CTRL_RUN_STOP);
-	if (ret)
-		return ret;
-	return ret;
+	return regmap_write(xdev->rmap, xchan->base + XDMA_CHAN_CONTROL_W1C,
+			    CHAN_CTRL_RUN_STOP);
 }

 /**
--- a/drivers/dma/xilinx/xilinx_dma.c
+++ b/drivers/dma/xilinx/xilinx_dma.c
@ -1404,16 +1404,18 @@ static void xilinx_vdma_start_transfer(struct xilinx_dma_chan *chan)

 	dma_ctrl_write(chan, XILINX_DMA_REG_DMACR, reg);

-	j = chan->desc_submitcount;
-	reg = dma_read(chan, XILINX_DMA_REG_PARK_PTR);
-	if (chan->direction == DMA_MEM_TO_DEV) {
-		reg &= ~XILINX_DMA_PARK_PTR_RD_REF_MASK;
-		reg |= j << XILINX_DMA_PARK_PTR_RD_REF_SHIFT;
-	} else {
-		reg &= ~XILINX_DMA_PARK_PTR_WR_REF_MASK;
-		reg |= j << XILINX_DMA_PARK_PTR_WR_REF_SHIFT;
+	if (config->park) {
+		j = chan->desc_submitcount;
+		reg = dma_read(chan, XILINX_DMA_REG_PARK_PTR);
+		if (chan->direction == DMA_MEM_TO_DEV) {
+			reg &= ~XILINX_DMA_PARK_PTR_RD_REF_MASK;
+			reg |= j << XILINX_DMA_PARK_PTR_RD_REF_SHIFT;
+		} else {
+			reg &= ~XILINX_DMA_PARK_PTR_WR_REF_MASK;
+			reg |= j << XILINX_DMA_PARK_PTR_WR_REF_SHIFT;
+		}
+		dma_write(chan, XILINX_DMA_REG_PARK_PTR, reg);
 	}
-	dma_write(chan, XILINX_DMA_REG_PARK_PTR, reg);

 	/* Start the hardware */
 	xilinx_dma_start(chan);