Merge tag 'spi-v6.17' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/spi

Pull spi updates from Mark Brown: "This release is almost entirely driver work, mostly new drivers with the usual smattering of per driver updates anf fixes, with only trivial changes in the core. Highlights include: - Quite a bit of maintainence work on the STM32 and Qualcomm drivers - Usage of the newly added devm_dma_request_chan() in the ateml driver, pulling in the relevant dmaengine change - Cleanups of our usage of the PM autosuspend functions, this pulls in some PM core changes on a shared tag - Support for ADI sigma-delta triggers, Amlogic SPISG, Mediatek MT6991 and MT8196, Renesas RZ/V2H(P) and SOPHGO SG2042" * tag 'spi-v6.17' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/spi: (62 commits) spi: SPISG: Fix less than zero comparison on a u32 variable spi: intel: Allow writeable MTD partition with module param spi: Add driver for the RZ/V2H(P) RSPI IP spi: dt-bindings: Document the RZ/V2H(P) RSPI MAINTAINERS: Add an entry for Amlogic spi driver spi: Add Amlogic SPISG driver spi: dt-bindings: Add binding document of Amlogic SPISG controller spi: spi-sg2044-nor: Add SPI-NOR controller for SG2042 spi: spi-sg2044-nor: Add configurable chip_info spi: dt-bindings: spi-sg2044-nor: Change SOPHGO SG2042 spi: spi-qpic-snand: simplify bad block marker duplication spi: spidev: Add an entry for the ABB spi sensors dt-bindings: trivial-devices: Document ABB sensors spi: stm32-ospi: Fix NULL vs IS_ERR() bug in stm32_ospi_get_resources() spi: gpio: Use explicit 'unsigned int' for parameter types spi: dt-bindings: spi-mux: Drop "spi-max-frequency" as required spi: st: Switch from CONFIG_PM_SLEEP guards to pm_sleep_ptr() spi: rspi: Convert to DEFINE_SIMPLE_DEV_PM_OPS() spi: sh-msiof: Convert to DEFINE_SIMPLE_DEV_PM_OPS() spi: xilinx: Fix block comment style and minor cleanups ...
2025-07-28 23:03:09 -07:00
parent bf977a9ad3 2d442a0c78
commit 0262163136
56 changed files with 2963 additions and 545 deletions
--- a/Documentation/devicetree/bindings/spi/amlogic,a4-spisg.yaml
+++ b/Documentation/devicetree/bindings/spi/amlogic,a4-spisg.yaml
@@ -0,0 +1,59 @@
+# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
+# Copyright (C) 2025 Amlogic, Inc. All rights reserved
+%YAML 1.2
+---
+$id: http://devicetree.org/schemas/spi/amlogic,a4-spisg.yaml#
+$schema: http://devicetree.org/meta-schemas/core.yaml#
+
+title: Amlogic SPI Scatter-Gather Controller
+
+maintainers:
+  - Xianwei Zhao <xianwei.zhao@amlogic.com>
+  - Sunny Luo <sunny.luo@amlogic.com>
+
+allOf:
+  - $ref: spi-controller.yaml#
+
+properties:
+  compatible:
+    const: amlogic,a4-spisg
+
+  reg:
+    maxItems: 1
+
+  interrupts:
+    maxItems: 1
+
+  clocks:
+    maxItems: 2
+
+  clock-names:
+    items:
+      - const: core
+      - const: pclk
+
+  resets:
+    maxItems: 1
+
+required:
+  - compatible
+  - reg
+  - interrupts
+  - clocks
+  - clock-names
+
+unevaluatedProperties: false
+
+examples:
+  - |
+    #include <dt-bindings/interrupt-controller/arm-gic.h>
+    spi@50000 {
+        compatible = "amlogic,a4-spisg";
+        reg = <0x50000 0x38>;
+        interrupts = <GIC_SPI 183 IRQ_TYPE_LEVEL_HIGH>;
+        clocks = <&clkc 37>,
+                 <&clkc 93>;
+        clock-names = "core", "pclk";
+        #address-cells = <1>;
+        #size-cells = <0>;
+    };
--- a/Documentation/devicetree/bindings/spi/fsl,dspi.yaml
+++ b/Documentation/devicetree/bindings/spi/fsl,dspi.yaml
@@ -23,6 +23,7 @@ properties:
          - fsl,ls2080a-dspi
          - fsl,ls2085a-dspi
          - fsl,lx2160a-dspi
+          - nxp,s32g2-dspi
      - items:
          - enum:
              - fsl,ls1012a-dspi
@@ -37,6 +38,9 @@ properties:
      - items:
          - const: fsl,lx2160a-dspi
          - const: fsl,ls2085a-dspi
+      - items:
+          - const: nxp,s32g3-dspi
+          - const: nxp,s32g2-dspi

  reg:
    maxItems: 1
@@ -114,3 +118,17 @@ examples:
            spi-cs-hold-delay-ns = <50>;
        };
    };
+  # S32G3 in target mode
+  - |
+    spi@401d4000 {
+        compatible = "nxp,s32g3-dspi", "nxp,s32g2-dspi";
+        reg = <0x401d4000 0x1000>;
+        interrupts = <GIC_SPI 85 IRQ_TYPE_LEVEL_HIGH>;
+        clocks = <&clks 26>;
+        clock-names = "dspi";
+        spi-num-chipselects = <8>;
+        bus-num = <0>;
+        dmas = <&edma0 0 7>, <&edma0 0 8>;
+        dma-names = "tx", "rx";
+        spi-slave;
+    };
--- a/Documentation/devicetree/bindings/spi/marvell,orion-spi.yaml
+++ b/Documentation/devicetree/bindings/spi/marvell,orion-spi.yaml
@@ -0,0 +1,102 @@
+# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
+%YAML 1.2
+---
+$id: http://devicetree.org/schemas/spi/marvell,orion-spi.yaml#
+$schema: http://devicetree.org/meta-schemas/core.yaml#
+
+title: Marvell Orion SPI controller
+
+maintainers:
+  - Andrew Lunn <andrew@lunn.ch>
+  - Gregory CLEMENT <gregory.clement@bootlin.com>
+
+allOf:
+  - $ref: /schemas/spi/spi-controller.yaml#
+
+properties:
+  compatible:
+    oneOf:
+      - enum:
+          - marvell,orion-spi
+          - marvell,armada-380-spi  # For ap80x and cp11x
+      - items:
+          - enum:
+              - marvell,armada-370-spi
+              - marvell,armada-375-spi
+              - marvell,armada-380-spi
+              - marvell,armada-390-spi
+              - marvell,armada-xp-spi
+          - const: marvell,orion-spi
+
+  cell-index:
+    description: Instance id for the SPI controller
+    deprecated: true
+
+  reg:
+    minItems: 1
+    items:
+      - description: control registers
+      - description: CS0 MBUS target/attribute registers for direct mode
+      - description: CS1 MBUS target/attribute registers for direct mode
+      - description: CS2 MBUS target/attribute registers for direct mode
+      - description: CS3 MBUS target/attribute registers for direct mode
+      - description: CS4 MBUS target/attribute registers for direct mode
+      - description: CS5 MBUS target/attribute registers for direct mode
+      - description: CS6 MBUS target/attribute registers for direct mode
+      - description: CS7 MBUS target/attribute registers for direct mode
+
+  clocks:
+    minItems: 1
+    maxItems: 2
+
+  clock-names:
+    items:
+      - const: core
+      - const: axi
+
+  interrupts:
+    maxItems: 1
+
+required:
+  - compatible
+  - reg
+  - clocks
+
+unevaluatedProperties: false
+
+examples:
+  - |
+    spi@10600 {
+      compatible = "marvell,orion-spi";
+      #address-cells = <1>;
+      #size-cells = <0>;
+      cell-index = <0>;
+      reg = <0x10600 0x28>;
+      clocks = <&coreclk 0>;
+      interrupts = <23>;
+    };
+  - |
+    #define MBUS_ID(target,attributes) (((target) << 24) | ((attributes) << 16))
+
+    bus {
+        #address-cells = <2>;
+        #size-cells = <1>;
+
+        spi@10600 {
+          compatible = "marvell,orion-spi";
+          #address-cells = <1>;
+          #size-cells = <0>;
+          cell-index = <0>;
+          reg = <MBUS_ID(0xf0, 0x01) 0x10600 0x28>, /* control */
+                <MBUS_ID(0x01, 0x1e) 0 0xffffffff>, /* CS0 */
+                <MBUS_ID(0x01, 0x5e) 0 0xffffffff>, /* CS1 */
+                <MBUS_ID(0x01, 0x9e) 0 0xffffffff>, /* CS2 */
+                <MBUS_ID(0x01, 0xde) 0 0xffffffff>, /* CS3 */
+                <MBUS_ID(0x01, 0x1f) 0 0xffffffff>, /* CS4 */
+                <MBUS_ID(0x01, 0x5f) 0 0xffffffff>, /* CS5 */
+                <MBUS_ID(0x01, 0x9f) 0 0xffffffff>, /* CS6 */
+                <MBUS_ID(0x01, 0xdf) 0 0xffffffff>; /* CS7 */
+          clocks = <&coreclk 0>;
+          interrupts = <23>;
+        };
+    };
--- a/Documentation/devicetree/bindings/spi/mediatek,spi-mt65xx.yaml
+++ b/Documentation/devicetree/bindings/spi/mediatek,spi-mt65xx.yaml
@@ -39,6 +39,10 @@ properties:
              - mediatek,mt7988-spi-single
              - mediatek,mt8188-spi-ipm
          - const: mediatek,spi-ipm
+      - items:
+          - enum:
+              - mediatek,mt8196-spi
+          - const: mediatek,mt6991-spi
      - items:
          - enum:
              - mediatek,mt2701-spi
@@ -46,6 +50,7 @@ properties:
              - mediatek,mt6589-spi
              - mediatek,mt6765-spi
              - mediatek,mt6893-spi
+              - mediatek,mt6991-spi
              - mediatek,mt7622-spi
              - mediatek,mt8135-spi
              - mediatek,mt8173-spi
--- a/Documentation/devicetree/bindings/spi/mxs-spi.yaml
+++ b/Documentation/devicetree/bindings/spi/mxs-spi.yaml
@@ -24,6 +24,9 @@ properties:
  interrupts:
    maxItems: 1

+  clocks:
+    maxItems: 1
+
  dmas:
    maxItems: 1

--- a/Documentation/devicetree/bindings/spi/nxp,lpc3220-spi.yaml
+++ b/Documentation/devicetree/bindings/spi/nxp,lpc3220-spi.yaml
@@ -0,0 +1,44 @@
+# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
+%YAML 1.2
+---
+$id: http://devicetree.org/schemas/spi/nxp,lpc3220-spi.yaml#
+$schema: http://devicetree.org/meta-schemas/core.yaml#
+
+title: NXP LPC3220 SPI controller
+
+maintainers:
+  - Frank Li <Frank.Li@nxp.com>
+
+properties:
+  compatible:
+    enum:
+      - nxp,lpc3220-spi
+
+  reg:
+    maxItems: 1
+
+  clocks:
+    maxItems: 1
+
+allOf:
+  - $ref: spi-controller.yaml#
+
+unevaluatedProperties: false
+
+required:
+  - compatible
+  - reg
+  - clocks
+
+examples:
+  - |
+    #include <dt-bindings/clock/lpc32xx-clock.h>
+
+    spi@20088000 {
+        compatible = "nxp,lpc3220-spi";
+        reg = <0x20088000 0x1000>;
+        clocks = <&clk LPC32XX_CLK_SPI1>;
+        #address-cells = <1>;
+        #size-cells = <0>;
+    };
+
--- a/Documentation/devicetree/bindings/spi/renesas,rzv2h-rspi.yaml
+++ b/Documentation/devicetree/bindings/spi/renesas,rzv2h-rspi.yaml
@@ -0,0 +1,96 @@
+# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
+%YAML 1.2
+---
+$id: http://devicetree.org/schemas/spi/renesas,rzv2h-rspi.yaml#
+$schema: http://devicetree.org/meta-schemas/core.yaml#
+
+title: Renesas RZ/V2H(P) Renesas Serial Peripheral Interface (RSPI)
+
+maintainers:
+  - Fabrizio Castro <fabrizio.castro.jz@renesas.com>
+
+allOf:
+  - $ref: spi-controller.yaml#
+
+properties:
+  compatible:
+    const: renesas,r9a09g057-rspi # RZ/V2H(P)
+
+  reg:
+    maxItems: 1
+
+  interrupts:
+    items:
+      - description: Idle Interrupt
+      - description: Error Interrupt
+      - description: Communication End Interrupt
+      - description: Receive Buffer Full Interrupt
+      - description: Transmit Buffer Empty Interrupt
+
+  interrupt-names:
+    items:
+      - const: idle
+      - const: error
+      - const: end
+      - const: rx
+      - const: tx
+
+  clocks:
+    maxItems: 3
+
+  clock-names:
+    items:
+      - const: pclk
+      - const: pclk_sfr
+      - const: tclk
+
+  resets:
+    maxItems: 2
+
+  reset-names:
+    items:
+      - const: presetn
+      - const: tresetn
+
+  power-domains:
+    maxItems: 1
+
+required:
+  - compatible
+  - reg
+  - interrupts
+  - interrupt-names
+  - clocks
+  - clock-names
+  - resets
+  - reset-names
+  - power-domains
+  - '#address-cells'
+  - '#size-cells'
+
+unevaluatedProperties: false
+
+examples:
+  - |
+    #include <dt-bindings/interrupt-controller/arm-gic.h>
+    #include <dt-bindings/clock/renesas-cpg-mssr.h>
+    spi@12800800 {
+        compatible = "renesas,r9a09g057-rspi";
+
+        reg = <0x12800800 0x400>;
+        interrupts = <GIC_SPI 111 IRQ_TYPE_LEVEL_HIGH>,
+                     <GIC_SPI 112 IRQ_TYPE_LEVEL_HIGH>,
+                     <GIC_SPI 113 IRQ_TYPE_EDGE_RISING>,
+                     <GIC_SPI 504 IRQ_TYPE_EDGE_RISING>,
+                     <GIC_SPI 505 IRQ_TYPE_EDGE_RISING>;
+        interrupt-names = "idle", "error", "end", "rx", "tx";
+        clocks = <&cpg CPG_MOD 0x5a>,
+                 <&cpg CPG_MOD 0x5b>,
+                 <&cpg CPG_MOD 0x5c>;
+        clock-names = "pclk", "pclk_sfr", "tclk";
+        resets = <&cpg 0x7f>, <&cpg 0x80>;
+        reset-names = "presetn", "tresetn";
+        power-domains = <&cpg>;
+        #address-cells = <1>;
+        #size-cells = <0>;
+    };
--- a/Documentation/devicetree/bindings/spi/spi-mux.yaml
+++ b/Documentation/devicetree/bindings/spi/spi-mux.yaml
@@ -46,7 +46,6 @@ properties:
 required:
  - compatible
  - reg
-  - spi-max-frequency
  - mux-controls

 unevaluatedProperties: false
--- a/Documentation/devicetree/bindings/spi/spi-orion.txt
+++ b/Documentation/devicetree/bindings/spi/spi-orion.txt
@@ -1,79 +0,0 @@
-Marvell Orion SPI device
-
-Required properties:
- compatible : should be on of the following:
-    - "marvell,orion-spi" for the Orion, mv78x00, Kirkwood and Dove SoCs
-    - "marvell,armada-370-spi", for the Armada 370 SoCs
-    - "marvell,armada-375-spi", for the Armada 375 SoCs
-    - "marvell,armada-380-spi", for the Armada 38x SoCs
-    - "marvell,armada-390-spi", for the Armada 39x SoCs
-    - "marvell,armada-xp-spi", for the Armada XP SoCs
- reg : offset and length of the register set for the device.
-	This property can optionally have additional entries to configure
-	the SPI direct access mode that some of the Marvell SoCs support
-	additionally to the normal indirect access (PIO) mode. The values
-	for the MBus "target" and "attribute" are defined in the Marvell
-	SoC "Functional Specifications" Manual in the chapter "Marvell
-	Core Processor Address Decoding".
-	The eight register sets following the control registers refer to
-	chip-select lines 0 through 7 respectively.
- cell-index : Which of multiple SPI controllers is this.
- clocks : pointers to the reference clocks for this device, the first
-	   one is the one used for the clock on the spi bus, the
-	   second one is optional and is the clock used for the
-	   functional part of the controller
-
-Optional properties:
- interrupts : Is currently not used.
- clock-names : names of used clocks, mandatory if the second clock is
-		used, the name must be "core", and "axi" (the latter
-		is only for Armada 7K/8K).
-
-
-Example:
-       spi@10600 {
-	       compatible = "marvell,orion-spi";
-	       #address-cells = <1>;
-	       #size-cells = <0>;
-	       cell-index = <0>;
-	       reg = <0x10600 0x28>;
-	       interrupts = <23>;
-       };
-
-Example with SPI direct mode support (optionally):
-	spi0: spi@10600 {
-		compatible = "marvell,orion-spi";
-		#address-cells = <1>;
-		#size-cells = <0>;
-		cell-index = <0>;
-		reg = <MBUS_ID(0xf0, 0x01) 0x10600 0x28>, /* control */
-		      <MBUS_ID(0x01, 0x1e) 0 0xffffffff>, /* CS0 */
-		      <MBUS_ID(0x01, 0x5e) 0 0xffffffff>, /* CS1 */
-		      <MBUS_ID(0x01, 0x9e) 0 0xffffffff>, /* CS2 */
-		      <MBUS_ID(0x01, 0xde) 0 0xffffffff>, /* CS3 */
-		      <MBUS_ID(0x01, 0x1f) 0 0xffffffff>, /* CS4 */
-		      <MBUS_ID(0x01, 0x5f) 0 0xffffffff>, /* CS5 */
-		      <MBUS_ID(0x01, 0x9f) 0 0xffffffff>, /* CS6 */
-		      <MBUS_ID(0x01, 0xdf) 0 0xffffffff>; /* CS7 */
-		interrupts = <23>;
-	};
-
-To enable the direct mode, the board specific 'ranges' property in the
-'soc' node needs to add the entries for the desired SPI controllers
-and its chip-selects that are used in the direct mode instead of PIO
-mode. Here an example for this (SPI controller 0, device 1 and SPI
-controller 1, device 2 are used in direct mode. All other SPI device
-are used in the default indirect (PIO) mode):
-	soc {
-		/*
-		 * Enable the SPI direct access by configuring an entry
-		 * here in the board-specific ranges property
-		 */
-		ranges = <MBUS_ID(0xf0, 0x01) 0 0 0xf1000000 0x100000>,	/* internal regs */
-			 <MBUS_ID(0x01, 0x1d) 0 0 0xfff00000 0x100000>,	/* BootROM       */
-			 <MBUS_ID(0x01, 0x5e) 0 0 0xf1100000 0x10000>,	/* SPI0-DEV1 */
-			 <MBUS_ID(0x01, 0x9a) 0 0 0xf1110000 0x10000>;	/* SPI1-DEV2 */
-
-For further information on the MBus bindings, please see the MBus
-DT documentation:
-Documentation/devicetree/bindings/bus/mvebu-mbus.txt
--- a/Documentation/devicetree/bindings/spi/spi-peripheral-props.yaml
+++ b/Documentation/devicetree/bindings/spi/spi-peripheral-props.yaml
@@ -115,6 +115,7 @@ properties:
    maxItems: 4

  st,spi-midi-ns:
+    deprecated: true
    description: |
      Only for STM32H7, (Master Inter-Data Idleness) minimum time
      delay in nanoseconds inserted between two consecutive data frames.
--- a/Documentation/devicetree/bindings/spi/spi-sg2044-nor.yaml
+++ b/Documentation/devicetree/bindings/spi/spi-sg2044-nor.yaml
@@ -14,12 +14,9 @@ allOf:

 properties:
  compatible:
-    oneOf:
-      - const: sophgo,sg2044-spifmc-nor
-      - items:
-          - enum:
-              - sophgo,sg2042-spifmc-nor
-          - const: sophgo,sg2044-spifmc-nor
+    enum:
+      - sophgo,sg2042-spifmc-nor
+      - sophgo,sg2044-spifmc-nor

  reg:
    maxItems: 1
--- a/Documentation/devicetree/bindings/spi/st,stm32-spi.yaml
+++ b/Documentation/devicetree/bindings/spi/st,stm32-spi.yaml
@@ -18,6 +18,38 @@ maintainers:

 allOf:
  - $ref: spi-controller.yaml#
+  - if:
+      properties:
+        compatible:
+          contains:
+            const: st,stm32f4-spi
+
+    then:
+      properties:
+        st,spi-midi-ns: false
+        sram: false
+        dmas:
+          maxItems: 2
+        dma-names:
+          items:
+            - const: rx
+            - const: tx
+
+  - if:
+      properties:
+        compatible:
+          contains:
+            const: st,stm32mp25-spi
+
+    then:
+      properties:
+        sram: false
+        dmas:
+          maxItems: 2
+        dma-names:
+          items:
+            - const: rx
+            - const: tx

 properties:
  compatible:
@@ -41,16 +73,28 @@ properties:

  dmas:
    description: |
-      DMA specifiers for tx and rx dma. DMA fifo mode must be used. See
-      the STM32 DMA controllers bindings Documentation/devicetree/bindings/dma/stm32/*.yaml.
+      DMA specifiers for tx and rx channels. DMA fifo mode must be used. See
+      the STM32 DMA bindings Documentation/devicetree/bindings/dma/stm32/st,*dma.yaml
+    minItems: 2
    items:
      - description: rx DMA channel
      - description: tx DMA channel
+      - description: rxm2m MDMA channel

  dma-names:
+    minItems: 2
    items:
      - const: rx
      - const: tx
+      - const: rxm2m
+
+  sram:
+    $ref: /schemas/types.yaml#/definitions/phandle
+    description: |
+      Phandles to a reserved SRAM region which is used as temporary
+      storage memory between DMA and MDMA engines.
+      The region should be defined as child node of the AHB SRAM node
+      as per the generic bindings in Documentation/devicetree/bindings/sram/sram.yaml

  access-controllers:
    minItems: 1
--- a/Documentation/devicetree/bindings/trigger-source/adi,util-sigma-delta-spi.yaml
+++ b/Documentation/devicetree/bindings/trigger-source/adi,util-sigma-delta-spi.yaml
@@ -0,0 +1,49 @@
+# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
+# Copyright (c) 2025 Analog Devices, Inc.
+# Copyright (c) 2025 BayLibre, SAS
+
+%YAML 1.2
+---
+$id: http://devicetree.org/schemas/trigger-source/adi,util-sigma-delta-spi.yaml#
+$schema: http://devicetree.org/meta-schemas/core.yaml#
+
+title: Analog Devices Util Sigma-Delta SPI IP Core
+
+maintainers:
+  - David Lechner <dlechner@baylibre.com>
+
+description:
+  The Util Sigma-Delta SPI is an FPGA IP core from Analog Devices that provides
+  a SPI offload trigger from the RDY signal of the combined DOUT/RDY pin of
+  the sigma-delta family of ADCs.
+  https://analogdevicesinc.github.io/hdl/library/util_sigma_delta_spi/index.html
+
+properties:
+  compatible:
+    const: adi,util-sigma-delta-spi
+
+  reg:
+    maxItems: 1
+
+  clocks:
+    maxItems: 1
+
+  '#trigger-source-cells':
+    const: 0
+
+required:
+  - compatible
+  - reg
+  - clocks
+  - '#trigger-source-cells'
+
+additionalProperties: false
+
+examples:
+  - |
+    trigger@40000 {
+        reg = <0x40000 0x1000>;
+        compatible = "adi,util-sigma-delta-spi";
+        clocks = <&clk 0>;
+        #trigger-source-cells = <0>;
+    };
--- a/Documentation/devicetree/bindings/trivial-devices.yaml
+++ b/Documentation/devicetree/bindings/trivial-devices.yaml
@@ -30,6 +30,8 @@ properties:
    items:
      # Entries are sorted alphanumerically by the compatible
      - enum:
+            # ABB register based spi sensors
+          - abb,spi-sensor
            # Acbel fsg032 power supply
          - acbel,fsg032
            # SMBus/I2C Digital Temperature Sensor in 6-Pin SOT with SMBus Alert and Over Temperature Pin
--- a/16
+++ b/16
@@ -1308,6 +1308,15 @@ S:	Maintained
 F:	Documentation/devicetree/bindings/rtc/amlogic,a4-rtc.yaml
 F:	drivers/rtc/rtc-amlogic-a4.c

+AMLOGIC SPISG DRIVER
+M:	Sunny Luo <sunny.luo@amlogic.com>
+M:	Xianwei Zhao <xianwei.zhao@amlogic.com>
+L:	linux-amlogic@lists.infradead.org
+L:	linux-spi@vger.kernel.org
+S:	Maintained
+F:	Documentation/devicetree/bindings/spi/amlogic,a4-spisg.yaml
+F:	drivers/spi/spi-amlogic-spisg.c
+
 AMPHENOL CHIPCAP 2 DRIVER
 M:	Javier Carrasco <javier.carrasco.cruz@gmail.com>
 L:	linux-hwmon@vger.kernel.org
@@ -23432,7 +23441,7 @@ F:	include/linux/mtd/spi-nor.h

 SPI OFFLOAD
 R:	David Lechner <dlechner@baylibre.com>
-F:	drivers/spi/spi-offload-trigger-pwm.c
+F:	drivers/spi/spi-offload-trigger-*.c
 F:	drivers/spi/spi-offload.c
 F:	include/linux/spi/offload/
 K:	spi_offload
@@ -25253,6 +25262,11 @@ W:	https://github.com/srcres258/linux-doc
 T:	git git://github.com/srcres258/linux-doc.git doc-zh-tw
 F:	Documentation/translations/zh_TW/

+TRIGGER SOURCE - ADI UTIL SIGMA DELTA SPI
+M:	David Lechner <dlechner@baylibre.com>
+S:	Maintained
+F:	Documentation/devicetree/bindings/trigger-source/adi,util-sigma-delta-spi.yaml
+
 TRIGGER SOURCE - PWM
 M:	David Lechner <dlechner@baylibre.com>
 S:	Maintained
--- a/drivers/dma/dmaengine.c
+++ b/drivers/dma/dmaengine.c
@@ -926,6 +926,36 @@ void dma_release_channel(struct dma_chan *chan)
 }
 EXPORT_SYMBOL_GPL(dma_release_channel);

+static void dmaenginem_release_channel(void *chan)
+{
+	dma_release_channel(chan);
+}
+
+/**
+ * devm_dma_request_chan - try to allocate an exclusive slave channel
+ * @dev:	pointer to client device structure
+ * @name:	slave channel name
+ *
+ * Returns pointer to appropriate DMA channel on success or an error pointer.
+ *
+ * The operation is managed and will be undone on driver detach.
+ */
+
+struct dma_chan *devm_dma_request_chan(struct device *dev, const char *name)
+{
+	struct dma_chan *chan = dma_request_chan(dev, name);
+	int ret = 0;
+
+	if (!IS_ERR(chan))
+		ret = devm_add_action_or_reset(dev, dmaenginem_release_channel, chan);
+
+	if (ret)
+		return ERR_PTR(ret);
+
+	return chan;
+}
+EXPORT_SYMBOL_GPL(devm_dma_request_chan);
+
 /**
 * dmaengine_get - register interest in dma_channels
 */
--- a/drivers/mtd/nand/raw/qcom_nandc.c
+++ b/drivers/mtd/nand/raw/qcom_nandc.c
@@ -1379,7 +1379,7 @@ static int qcom_nand_attach_chip(struct nand_chip *chip)
 	struct qcom_nand_controller *nandc = get_qcom_nand_controller(chip);
 	int cwperpage, bad_block_byte, ret;
 	bool wide_bus;
-	int ecc_mode = 1;
+	int ecc_mode = ECC_MODE_8BIT;

 	/* controller only supports 512 bytes data steps */
 	ecc->size = NANDC_STEP_SIZE;
@@ -1400,7 +1400,7 @@ static int qcom_nand_attach_chip(struct nand_chip *chip)
 	if (ecc->strength >= 8) {
 		/* 8 bit ECC defaults to BCH ECC on all platforms */
 		host->bch_enabled = true;
-		ecc_mode = 1;
+		ecc_mode = ECC_MODE_8BIT;

 		if (wide_bus) {
 			host->ecc_bytes_hw = 14;
@@ -1420,7 +1420,7 @@ static int qcom_nand_attach_chip(struct nand_chip *chip)
 		if (nandc->props->ecc_modes & ECC_BCH_4BIT) {
 			/* BCH */
 			host->bch_enabled = true;
-			ecc_mode = 0;
+			ecc_mode = ECC_MODE_4BIT;

 			if (wide_bus) {
 				host->ecc_bytes_hw = 8;
--- a/drivers/spi/Kconfig
+++ b/drivers/spi/Kconfig
@@ -99,6 +99,15 @@ config SPI_AMLOGIC_SPIFC_A1
 	  This enables master mode support for the SPIFC (SPI flash
 	  controller) available in Amlogic A1 (A113L SoC).

+config SPI_AMLOGIC_SPISG
+	tristate "Amlogic SPISG controller"
+	depends on COMMON_CLK
+	depends on ARCH_MESON || COMPILE_TEST
+	help
+	  This enables master mode support for the SPISG (SPI scatter-gather
+	  communication controller), which is available on platforms such as
+	  Amlogic A4 SoCs.
+
 config SPI_APPLE
 	tristate "Apple SoC SPI Controller platform driver"
 	depends on ARCH_APPLE || COMPILE_TEST
@@ -647,10 +656,10 @@ config SPI_FSL_SPI
 config SPI_FSL_DSPI
 	tristate "Freescale DSPI controller"
 	select REGMAP_MMIO
-	depends on SOC_VF610 || SOC_LS1021A || ARCH_LAYERSCAPE || M5441x || COMPILE_TEST
+	depends on ARCH_MXC || ARCH_NXP || M5441x || COMPILE_TEST
 	help
 	  This enables support for the Freescale DSPI controller in master
-	  mode. VF610, LS1021A and ColdFire platforms uses the controller.
+	  mode. S32, VF610, LS1021A and ColdFire platforms uses the controller.

 config SPI_FSL_ESPI
 	tristate "Freescale eSPI controller"
@@ -923,6 +932,14 @@ config SPI_RSPI
 	help
 	  SPI driver for Renesas RSPI and QSPI blocks.

+config SPI_RZV2H_RSPI
+	tristate "Renesas RZ/V2H RSPI controller"
+	depends on ARCH_RENESAS || COMPILE_TEST
+	help
+	  RSPI driver for the Renesas RZ/V2H Serial Peripheral Interface (RSPI).
+	  RSPI supports both SPI host and SPI target roles. This option only
+	  enables the SPI host role.
+
 config SPI_RZV2M_CSI
 	tristate "Renesas RZ/V2M CSI controller"
 	depends on ARCH_RENESAS || COMPILE_TEST
@@ -1355,6 +1372,11 @@ if SPI_OFFLOAD

 comment "SPI Offload triggers"

+config SPI_OFFLOAD_TRIGGER_ADI_UTIL_SD
+	tristate "SPI offload trigger using ADI sigma-delta utility"
+	help
+	  SPI offload trigger from ADI sigma-delta utility FPGA IP block.
+
 config SPI_OFFLOAD_TRIGGER_PWM
 	tristate "SPI offload trigger using PWM"
 	depends on PWM
--- a/drivers/spi/Makefile
+++ b/drivers/spi/Makefile
@@ -20,6 +20,7 @@ obj-$(CONFIG_SPI_ALTERA)		+= spi-altera-platform.o
 obj-$(CONFIG_SPI_ALTERA_CORE)		+= spi-altera-core.o
 obj-$(CONFIG_SPI_ALTERA_DFL)		+= spi-altera-dfl.o
 obj-$(CONFIG_SPI_AMLOGIC_SPIFC_A1)	+= spi-amlogic-spifc-a1.o
+obj-$(CONFIG_SPI_AMLOGIC_SPISG)		+= spi-amlogic-spisg.o
 obj-$(CONFIG_SPI_APPLE)			+= spi-apple.o
 obj-$(CONFIG_SPI_AR934X)		+= spi-ar934x.o
 obj-$(CONFIG_SPI_ARMADA_3700)		+= spi-armada-3700.o
@@ -126,6 +127,7 @@ obj-$(CONFIG_MACH_REALTEK_RTL)		+= spi-realtek-rtl.o
 obj-$(CONFIG_SPI_REALTEK_SNAND)		+= spi-realtek-rtl-snand.o
 obj-$(CONFIG_SPI_RPCIF)			+= spi-rpc-if.o
 obj-$(CONFIG_SPI_RSPI)			+= spi-rspi.o
+obj-$(CONFIG_SPI_RZV2H_RSPI)		+= spi-rzv2h-rspi.o
 obj-$(CONFIG_SPI_RZV2M_CSI)		+= spi-rzv2m-csi.o
 obj-$(CONFIG_SPI_S3C64XX)		+= spi-s3c64xx.o
 obj-$(CONFIG_SPI_SC18IS602)		+= spi-sc18is602.o
@@ -170,3 +172,4 @@ obj-$(CONFIG_SPI_SLAVE_SYSTEM_CONTROL)	+= spi-slave-system-control.o

 # SPI offload triggers
 obj-$(CONFIG_SPI_OFFLOAD_TRIGGER_PWM)	+= spi-offload-trigger-pwm.o
+obj-$(CONFIG_SPI_OFFLOAD_TRIGGER_ADI_UTIL_SD) += spi-offload-trigger-adi-util-sigma-delta.o
--- a/drivers/spi/atmel-quadspi.c
+++ b/drivers/spi/atmel-quadspi.c
@@ -965,7 +965,6 @@ static int atmel_qspi_exec_op(struct spi_mem *mem, const struct spi_mem_op *op)
 	err = aq->ops->transfer(mem, op, offset);

 pm_runtime_put:
-	pm_runtime_mark_last_busy(&aq->pdev->dev);
 	pm_runtime_put_autosuspend(&aq->pdev->dev);
 	return err;
 }
@@ -1168,7 +1167,6 @@ static int atmel_qspi_setup(struct spi_device *spi)
 	aq->scr |= QSPI_SCR_SCBR(scbr);
 	atmel_qspi_write(aq->scr, aq, QSPI_SCR);

-	pm_runtime_mark_last_busy(ctrl->dev.parent);
 	pm_runtime_put_autosuspend(ctrl->dev.parent);

 	return 0;
@@ -1230,7 +1228,6 @@ static int atmel_qspi_set_cs_timing(struct spi_device *spi)
 	aq->mr |= QSPI_MR_DLYBCT(cs_hold) | QSPI_MR_DLYCS(cs_inactive);
 	atmel_qspi_write(aq->mr, aq, QSPI_MR);

-	pm_runtime_mark_last_busy(ctrl->dev.parent);
 	pm_runtime_put_autosuspend(ctrl->dev.parent);

 	return 0;
@@ -1285,18 +1282,21 @@ static int atmel_qspi_dma_init(struct spi_controller *ctrl)
 	struct atmel_qspi *aq = spi_controller_get_devdata(ctrl);
 	int ret;

-	aq->rx_chan = dma_request_chan(&aq->pdev->dev, "rx");
+	aq->rx_chan = devm_dma_request_chan(&aq->pdev->dev, "rx");
 	if (IS_ERR(aq->rx_chan)) {
 		ret = dev_err_probe(&aq->pdev->dev, PTR_ERR(aq->rx_chan),
 				    "RX DMA channel is not available\n");
-		goto null_rx_chan;
+		aq->rx_chan = NULL;
+		return ret;
 	}

-	aq->tx_chan = dma_request_chan(&aq->pdev->dev, "tx");
+	aq->tx_chan = devm_dma_request_chan(&aq->pdev->dev, "tx");
 	if (IS_ERR(aq->tx_chan)) {
 		ret = dev_err_probe(&aq->pdev->dev, PTR_ERR(aq->tx_chan),
 				    "TX DMA channel is not available\n");
-		goto release_rx_chan;
+		aq->rx_chan = NULL;
+		aq->tx_chan = NULL;
+		return ret;
 	}

 	ctrl->dma_rx = aq->rx_chan;
@@ -1307,21 +1307,6 @@ static int atmel_qspi_dma_init(struct spi_controller *ctrl)
 		 dma_chan_name(aq->tx_chan), dma_chan_name(aq->rx_chan));

 	return 0;
-
-release_rx_chan:
-	dma_release_channel(aq->rx_chan);
-	aq->tx_chan = NULL;
-null_rx_chan:
-	aq->rx_chan = NULL;
-	return ret;
-}
-
-static void atmel_qspi_dma_release(struct atmel_qspi *aq)
-{
-	if (aq->rx_chan)
-		dma_release_channel(aq->rx_chan);
-	if (aq->tx_chan)
-		dma_release_channel(aq->tx_chan);
 }

 static const struct atmel_qspi_ops atmel_qspi_ops = {
@@ -1426,14 +1411,13 @@ static int atmel_qspi_probe(struct platform_device *pdev)

 	/* Request the IRQ */
 	irq = platform_get_irq(pdev, 0);
-	if (irq < 0) {
-		err = irq;
-		goto dma_release;
-	}
+	if (irq < 0)
+		return irq;
+
 	err = devm_request_irq(&pdev->dev, irq, atmel_qspi_interrupt,
 			       0, dev_name(&pdev->dev), aq);
 	if (err)
-		goto dma_release;
+		return err;

 	pm_runtime_set_autosuspend_delay(&pdev->dev, 500);
 	pm_runtime_use_autosuspend(&pdev->dev);
@@ -1442,22 +1426,15 @@ static int atmel_qspi_probe(struct platform_device *pdev)

 	err = atmel_qspi_init(aq);
 	if (err)
-		goto dma_release;
+		return err;

 	err = spi_register_controller(ctrl);
 	if (err)
-		goto dma_release;
+		return err;

-	pm_runtime_mark_last_busy(&pdev->dev);
 	pm_runtime_put_autosuspend(&pdev->dev);

 	return 0;
-
-dma_release:
-	if (aq->caps->has_dma)
-		atmel_qspi_dma_release(aq);
-
-	return err;
 }

 static int atmel_qspi_sama7g5_suspend(struct atmel_qspi *aq)
@@ -1507,9 +1484,6 @@ static void atmel_qspi_remove(struct platform_device *pdev)

 	ret = pm_runtime_get_sync(&pdev->dev);
 	if (ret >= 0) {
-		if (aq->caps->has_dma)
-			atmel_qspi_dma_release(aq);
-
 		if (aq->caps->has_gclk) {
 			ret = atmel_qspi_sama7g5_suspend(aq);
 			if (ret)
@@ -1582,7 +1556,6 @@ static int __maybe_unused atmel_qspi_resume(struct device *dev)

 	atmel_qspi_write(aq->scr, aq, QSPI_SCR);

-	pm_runtime_mark_last_busy(dev);
 	pm_runtime_put_autosuspend(dev);

 	return 0;
--- a/drivers/spi/spi-amlogic-spisg.c
+++ b/drivers/spi/spi-amlogic-spisg.c
@@ -0,0 +1,888 @@
+// SPDX-License-Identifier: GPL-2.0+
+/*
+ * Driver for Amlogic SPI communication Scatter-Gather Controller
+ *
+ * Copyright (C) 2025 Amlogic, Inc. All rights reserved
+ *
+ * Author: Sunny Luo <sunny.luo@amlogic.com>
+ * Author: Xianwei Zhao <xianwei.zhao@amlogic.com>
+ */
+
+#include <linux/bitfield.h>
+#include <linux/device.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/clk.h>
+#include <linux/clk-provider.h>
+#include <linux/dma-mapping.h>
+#include <linux/platform_device.h>
+#include <linux/pinctrl/consumer.h>
+#include <linux/pm_runtime.h>
+#include <linux/spi/spi.h>
+#include <linux/types.h>
+#include <linux/interrupt.h>
+#include <linux/reset.h>
+#include <linux/regmap.h>
+
+/* Register Map */
+#define SPISG_REG_CFG_READY		0x00
+
+#define SPISG_REG_CFG_SPI		0x04
+#define CFG_BUS64_EN			BIT(0)
+#define CFG_SLAVE_EN			BIT(1)
+#define CFG_SLAVE_SELECT		GENMASK(3, 2)
+#define CFG_SFLASH_WP			BIT(4)
+#define CFG_SFLASH_HD			BIT(5)
+/* start on vsync rising */
+#define CFG_HW_POS			BIT(6)
+/* start on vsync falling */
+#define CFG_HW_NEG			BIT(7)
+
+#define SPISG_REG_CFG_START		0x08
+#define CFG_BLOCK_NUM			GENMASK(19, 0)
+#define CFG_BLOCK_SIZE			GENMASK(22, 20)
+#define CFG_DATA_COMMAND		BIT(23)
+#define CFG_OP_MODE			GENMASK(25, 24)
+#define CFG_RXD_MODE			GENMASK(27, 26)
+#define CFG_TXD_MODE			GENMASK(29, 28)
+#define CFG_EOC				BIT(30)
+#define CFG_PEND			BIT(31)
+
+#define SPISG_REG_CFG_BUS		0x0C
+#define CFG_CLK_DIV			GENMASK(7, 0)
+#define CLK_DIV_WIDTH			8
+#define CFG_RX_TUNING			GENMASK(11, 8)
+#define CFG_TX_TUNING			GENMASK(15, 12)
+#define CFG_CS_SETUP			GENMASK(19, 16)
+#define CFG_LANE			GENMASK(21, 20)
+#define CFG_HALF_DUPLEX			BIT(22)
+#define CFG_B_L_ENDIAN			BIT(23)
+#define CFG_DC_MODE			BIT(24)
+#define CFG_NULL_CTL			BIT(25)
+#define CFG_DUMMY_CTL			BIT(26)
+#define CFG_READ_TURN			GENMASK(28, 27)
+#define CFG_KEEP_SS			BIT(29)
+#define CFG_CPHA			BIT(30)
+#define CFG_CPOL			BIT(31)
+
+#define SPISG_REG_PIO_TX_DATA_L		0x10
+#define SPISG_REG_PIO_TX_DATA_H		0x14
+#define SPISG_REG_PIO_RX_DATA_L		0x18
+#define SPISG_REG_PIO_RX_DATA_H		0x1C
+#define SPISG_REG_MEM_TX_ADDR_L		0x10
+#define SPISG_REG_MEM_TX_ADDR_H		0x14
+#define SPISG_REG_MEM_RX_ADDR_L		0x18
+#define SPISG_REG_MEM_RX_ADDR_H		0x1C
+#define SPISG_REG_DESC_LIST_L		0x20
+#define SPISG_REG_DESC_LIST_H		0x24
+#define LIST_DESC_PENDING		BIT(31)
+#define SPISG_REG_DESC_CURRENT_L	0x28
+#define SPISG_REG_DESC_CURRENT_H	0x2c
+#define SPISG_REG_IRQ_STS		0x30
+#define SPISG_REG_IRQ_ENABLE		0x34
+#define IRQ_RCH_DESC_EOC		BIT(0)
+#define IRQ_RCH_DESC_INVALID		BIT(1)
+#define IRQ_RCH_DESC_RESP		BIT(2)
+#define IRQ_RCH_DATA_RESP		BIT(3)
+#define IRQ_WCH_DESC_EOC		BIT(4)
+#define IRQ_WCH_DESC_INVALID		BIT(5)
+#define IRQ_WCH_DESC_RESP		BIT(6)
+#define IRQ_WCH_DATA_RESP		BIT(7)
+#define IRQ_DESC_ERR			BIT(8)
+#define IRQ_SPI_READY			BIT(9)
+#define IRQ_DESC_DONE			BIT(10)
+#define IRQ_DESC_CHAIN_DONE		BIT(11)
+
+#define SPISG_MAX_REG			0x40
+
+#define SPISG_BLOCK_MAX			0x100000
+
+#define SPISG_OP_MODE_WRITE_CMD		0
+#define SPISG_OP_MODE_READ_STS		1
+#define SPISG_OP_MODE_WRITE		2
+#define SPISG_OP_MODE_READ		3
+
+#define SPISG_DATA_MODE_NONE		0
+#define SPISG_DATA_MODE_PIO		1
+#define SPISG_DATA_MODE_MEM		2
+#define SPISG_DATA_MODE_SG		3
+
+#define SPISG_CLK_DIV_MAX		256
+/* recommended by specification */
+#define SPISG_CLK_DIV_MIN		4
+#define DIV_NUM (SPISG_CLK_DIV_MAX - SPISG_CLK_DIV_MIN + 1)
+
+#define SPISG_PCLK_RATE_MIN		24000000
+
+#define SPISG_SINGLE_SPI		0
+#define SPISG_DUAL_SPI			1
+#define SPISG_QUAD_SPI			2
+
+struct spisg_sg_link {
+#define LINK_ADDR_VALID		BIT(0)
+#define LINK_ADDR_EOC		BIT(1)
+#define LINK_ADDR_IRQ		BIT(2)
+#define LINK_ADDR_ACT		GENMASK(5, 3)
+#define LINK_ADDR_RING		BIT(6)
+#define LINK_ADDR_LEN		GENMASK(31, 8)
+	u32			addr;
+	u32			addr1;
+};
+
+struct spisg_descriptor {
+	u32				cfg_start;
+	u32				cfg_bus;
+	u64				tx_paddr;
+	u64				rx_paddr;
+};
+
+struct spisg_descriptor_extra {
+	struct spisg_sg_link		*tx_ccsg;
+	struct spisg_sg_link		*rx_ccsg;
+	int				tx_ccsg_len;
+	int				rx_ccsg_len;
+};
+
+struct spisg_device {
+	struct spi_controller		*controller;
+	struct platform_device		*pdev;
+	struct regmap			*map;
+	struct clk			*core;
+	struct clk			*pclk;
+	struct clk			*sclk;
+	struct clk_div_table		*tbl;
+	struct completion		completion;
+	u32				status;
+	u32				speed_hz;
+	u32				effective_speed_hz;
+	u32				bytes_per_word;
+	u32				cfg_spi;
+	u32				cfg_start;
+	u32				cfg_bus;
+};
+
+static int spi_delay_to_sclk(u32 slck_speed_hz, struct spi_delay *delay)
+{
+	s32 ns;
+
+	if (!delay)
+		return 0;
+
+	if (delay->unit == SPI_DELAY_UNIT_SCK)
+		return delay->value;
+
+	ns = spi_delay_to_ns(delay, NULL);
+	if (ns < 0)
+		return 0;
+
+	return DIV_ROUND_UP_ULL(slck_speed_hz * ns, NSEC_PER_SEC);
+}
+
+static inline u32 aml_spisg_sem_down_read(struct spisg_device *spisg)
+{
+	u32 ret;
+
+	regmap_read(spisg->map, SPISG_REG_CFG_READY, &ret);
+	if (ret)
+		regmap_write(spisg->map, SPISG_REG_CFG_READY, 0);
+
+	return ret;
+}
+
+static inline void aml_spisg_sem_up_write(struct spisg_device *spisg)
+{
+	regmap_write(spisg->map, SPISG_REG_CFG_READY, 1);
+}
+
+static int aml_spisg_set_speed(struct spisg_device *spisg, uint speed_hz)
+{
+	u32 cfg_bus;
+
+	if (!speed_hz || speed_hz == spisg->speed_hz)
+		return 0;
+
+	spisg->speed_hz = speed_hz;
+	clk_set_rate(spisg->sclk, speed_hz);
+	/* Store the div for the descriptor mode */
+	regmap_read(spisg->map, SPISG_REG_CFG_BUS, &cfg_bus);
+	spisg->cfg_bus &= ~CFG_CLK_DIV;
+	spisg->cfg_bus |= cfg_bus & CFG_CLK_DIV;
+	spisg->effective_speed_hz = clk_get_rate(spisg->sclk);
+	dev_dbg(&spisg->pdev->dev,
+		"desired speed %dHz, effective speed %dHz\n",
+		speed_hz, spisg->effective_speed_hz);
+
+	return 0;
+}
+
+static bool aml_spisg_can_dma(struct spi_controller *ctlr,
+			      struct spi_device *spi,
+			      struct spi_transfer *xfer)
+{
+	return true;
+}
+
+static void aml_spisg_sg_xlate(struct sg_table *sgt, struct spisg_sg_link *ccsg)
+{
+	struct scatterlist *sg;
+	int i;
+
+	for_each_sg(sgt->sgl, sg, sgt->nents, i) {
+		ccsg->addr = FIELD_PREP(LINK_ADDR_VALID, 1) |
+			     FIELD_PREP(LINK_ADDR_RING, 0) |
+			     FIELD_PREP(LINK_ADDR_EOC, sg_is_last(sg)) |
+			     FIELD_PREP(LINK_ADDR_LEN, sg_dma_len(sg));
+		ccsg->addr1 = (u32)sg_dma_address(sg);
+		ccsg++;
+	}
+}
+
+static int nbits_to_lane[] = {
+	SPISG_SINGLE_SPI,
+	SPISG_SINGLE_SPI,
+	SPISG_DUAL_SPI,
+	-EINVAL,
+	SPISG_QUAD_SPI
+};
+
+static int aml_spisg_setup_transfer(struct spisg_device *spisg,
+				    struct spi_transfer *xfer,
+				    struct spisg_descriptor *desc,
+				    struct spisg_descriptor_extra *exdesc)
+{
+	int block_size, blocks;
+	struct device *dev = &spisg->pdev->dev;
+	struct spisg_sg_link *ccsg;
+	int ccsg_len;
+	dma_addr_t paddr;
+	int ret;
+
+	memset(desc, 0, sizeof(*desc));
+	if (exdesc)
+		memset(exdesc, 0, sizeof(*exdesc));
+	aml_spisg_set_speed(spisg, xfer->speed_hz);
+	xfer->effective_speed_hz = spisg->effective_speed_hz;
+
+	desc->cfg_start = spisg->cfg_start;
+	desc->cfg_bus = spisg->cfg_bus;
+
+	block_size = xfer->bits_per_word >> 3;
+	blocks = xfer->len / block_size;
+
+	desc->cfg_start |= FIELD_PREP(CFG_EOC, 0);
+	desc->cfg_bus |= FIELD_PREP(CFG_KEEP_SS, !xfer->cs_change);
+	desc->cfg_bus |= FIELD_PREP(CFG_NULL_CTL, 0);
+
+	if (xfer->tx_buf || xfer->tx_dma) {
+		desc->cfg_bus |= FIELD_PREP(CFG_LANE, nbits_to_lane[xfer->tx_nbits]);
+		desc->cfg_start |= FIELD_PREP(CFG_OP_MODE, SPISG_OP_MODE_WRITE);
+	}
+	if (xfer->rx_buf || xfer->rx_dma) {
+		desc->cfg_bus |= FIELD_PREP(CFG_LANE, nbits_to_lane[xfer->rx_nbits]);
+		desc->cfg_start |= FIELD_PREP(CFG_OP_MODE, SPISG_OP_MODE_READ);
+	}
+
+	if (FIELD_GET(CFG_OP_MODE, desc->cfg_start) == SPISG_OP_MODE_READ_STS) {
+		desc->cfg_start |= FIELD_PREP(CFG_BLOCK_SIZE, blocks) |
+				   FIELD_PREP(CFG_BLOCK_NUM, 1);
+	} else {
+		blocks = min_t(int, blocks, SPISG_BLOCK_MAX);
+		desc->cfg_start |= FIELD_PREP(CFG_BLOCK_SIZE, block_size & 0x7) |
+				   FIELD_PREP(CFG_BLOCK_NUM, blocks);
+	}
+
+	if (xfer->tx_sg.nents && xfer->tx_sg.sgl) {
+		ccsg_len = xfer->tx_sg.nents * sizeof(struct spisg_sg_link);
+		ccsg = kzalloc(ccsg_len, GFP_KERNEL | GFP_DMA);
+		if (!ccsg) {
+			dev_err(dev, "alloc tx_ccsg failed\n");
+			return -ENOMEM;
+		}
+
+		aml_spisg_sg_xlate(&xfer->tx_sg, ccsg);
+		paddr = dma_map_single(dev, (void *)ccsg,
+				       ccsg_len, DMA_TO_DEVICE);
+		ret = dma_mapping_error(dev, paddr);
+		if (ret) {
+			kfree(ccsg);
+			dev_err(dev, "tx ccsg map failed\n");
+			return ret;
+		}
+
+		desc->tx_paddr = paddr;
+		desc->cfg_start |= FIELD_PREP(CFG_TXD_MODE, SPISG_DATA_MODE_SG);
+		exdesc->tx_ccsg = ccsg;
+		exdesc->tx_ccsg_len = ccsg_len;
+		dma_sync_sgtable_for_device(spisg->controller->cur_tx_dma_dev,
+					    &xfer->tx_sg, DMA_TO_DEVICE);
+	} else if (xfer->tx_buf || xfer->tx_dma) {
+		paddr = xfer->tx_dma;
+		if (!paddr) {
+			paddr = dma_map_single(dev, (void *)xfer->tx_buf,
+					       xfer->len, DMA_TO_DEVICE);
+			ret = dma_mapping_error(dev, paddr);
+			if (ret) {
+				dev_err(dev, "tx buf map failed\n");
+				return ret;
+			}
+		}
+		desc->tx_paddr = paddr;
+		desc->cfg_start |= FIELD_PREP(CFG_TXD_MODE, SPISG_DATA_MODE_MEM);
+	}
+
+	if (xfer->rx_sg.nents && xfer->rx_sg.sgl) {
+		ccsg_len = xfer->rx_sg.nents * sizeof(struct spisg_sg_link);
+		ccsg = kzalloc(ccsg_len, GFP_KERNEL | GFP_DMA);
+		if (!ccsg) {
+			dev_err(dev, "alloc rx_ccsg failed\n");
+			return -ENOMEM;
+		}
+
+		aml_spisg_sg_xlate(&xfer->rx_sg, ccsg);
+		paddr = dma_map_single(dev, (void *)ccsg,
+				       ccsg_len, DMA_TO_DEVICE);
+		ret = dma_mapping_error(dev, paddr);
+		if (ret) {
+			kfree(ccsg);
+			dev_err(dev, "rx ccsg map failed\n");
+			return ret;
+		}
+
+		desc->rx_paddr = paddr;
+		desc->cfg_start |= FIELD_PREP(CFG_RXD_MODE, SPISG_DATA_MODE_SG);
+		exdesc->rx_ccsg = ccsg;
+		exdesc->rx_ccsg_len = ccsg_len;
+		dma_sync_sgtable_for_device(spisg->controller->cur_rx_dma_dev,
+					    &xfer->rx_sg, DMA_FROM_DEVICE);
+	} else if (xfer->rx_buf || xfer->rx_dma) {
+		paddr = xfer->rx_dma;
+		if (!paddr) {
+			paddr = dma_map_single(dev, xfer->rx_buf,
+					       xfer->len, DMA_FROM_DEVICE);
+			ret = dma_mapping_error(dev, paddr);
+			if (ret) {
+				dev_err(dev, "rx buf map failed\n");
+				return ret;
+			}
+		}
+
+		desc->rx_paddr = paddr;
+		desc->cfg_start |= FIELD_PREP(CFG_RXD_MODE, SPISG_DATA_MODE_MEM);
+	}
+
+	return 0;
+}
+
+static void aml_spisg_cleanup_transfer(struct spisg_device *spisg,
+				       struct spi_transfer *xfer,
+				       struct spisg_descriptor *desc,
+				       struct spisg_descriptor_extra *exdesc)
+{
+	struct device *dev = &spisg->pdev->dev;
+
+	if (desc->tx_paddr) {
+		if (FIELD_GET(CFG_TXD_MODE, desc->cfg_start) == SPISG_DATA_MODE_SG) {
+			dma_unmap_single(dev, (dma_addr_t)desc->tx_paddr,
+					 exdesc->tx_ccsg_len, DMA_TO_DEVICE);
+			kfree(exdesc->tx_ccsg);
+			dma_sync_sgtable_for_cpu(spisg->controller->cur_tx_dma_dev,
+						 &xfer->tx_sg, DMA_TO_DEVICE);
+		} else if (!xfer->tx_dma) {
+			dma_unmap_single(dev, (dma_addr_t)desc->tx_paddr,
+					 xfer->len, DMA_TO_DEVICE);
+		}
+	}
+
+	if (desc->rx_paddr) {
+		if (FIELD_GET(CFG_RXD_MODE, desc->cfg_start) == SPISG_DATA_MODE_SG) {
+			dma_unmap_single(dev, (dma_addr_t)desc->rx_paddr,
+					 exdesc->rx_ccsg_len, DMA_TO_DEVICE);
+			kfree(exdesc->rx_ccsg);
+			dma_sync_sgtable_for_cpu(spisg->controller->cur_rx_dma_dev,
+						 &xfer->rx_sg, DMA_FROM_DEVICE);
+		} else if (!xfer->rx_dma) {
+			dma_unmap_single(dev, (dma_addr_t)desc->rx_paddr,
+					 xfer->len, DMA_FROM_DEVICE);
+		}
+	}
+}
+
+static void aml_spisg_setup_null_desc(struct spisg_device *spisg,
+				      struct spisg_descriptor *desc,
+				      u32 n_sclk)
+{
+	/* unit is the last xfer sclk */
+	desc->cfg_start = spisg->cfg_start;
+	desc->cfg_bus = spisg->cfg_bus;
+
+	desc->cfg_start |= FIELD_PREP(CFG_OP_MODE, SPISG_OP_MODE_WRITE) |
+			   FIELD_PREP(CFG_BLOCK_SIZE, 1) |
+			   FIELD_PREP(CFG_BLOCK_NUM, DIV_ROUND_UP(n_sclk, 8));
+
+	desc->cfg_bus |= FIELD_PREP(CFG_NULL_CTL, 1);
+}
+
+static void aml_spisg_pending(struct spisg_device *spisg,
+			      dma_addr_t desc_paddr,
+			      bool trig,
+			      bool irq_en)
+{
+	u32 desc_l, desc_h, cfg_spi, irq_enable;
+
+#ifdef	CONFIG_ARCH_DMA_ADDR_T_64BIT
+	desc_l = (u64)desc_paddr & 0xffffffff;
+	desc_h = (u64)desc_paddr >> 32;
+#else
+	desc_l = desc_paddr & 0xffffffff;
+	desc_h = 0;
+#endif
+
+	cfg_spi = spisg->cfg_spi;
+	if (trig)
+		cfg_spi |= CFG_HW_POS;
+	else
+		desc_h |= LIST_DESC_PENDING;
+
+	irq_enable = IRQ_RCH_DESC_INVALID | IRQ_RCH_DESC_RESP |
+		     IRQ_RCH_DATA_RESP | IRQ_WCH_DESC_INVALID |
+		     IRQ_WCH_DESC_RESP | IRQ_WCH_DATA_RESP |
+		     IRQ_DESC_ERR | IRQ_DESC_CHAIN_DONE;
+	regmap_write(spisg->map, SPISG_REG_IRQ_ENABLE, irq_en ? irq_enable : 0);
+	regmap_write(spisg->map, SPISG_REG_CFG_SPI, cfg_spi);
+	regmap_write(spisg->map, SPISG_REG_DESC_LIST_L, desc_l);
+	regmap_write(spisg->map, SPISG_REG_DESC_LIST_H, desc_h);
+}
+
+static irqreturn_t aml_spisg_irq(int irq, void *data)
+{
+	struct spisg_device *spisg = (void *)data;
+	u32 sts;
+
+	spisg->status = 0;
+	regmap_read(spisg->map, SPISG_REG_IRQ_STS, &sts);
+	regmap_write(spisg->map, SPISG_REG_IRQ_STS, sts);
+	if (sts & (IRQ_RCH_DESC_INVALID |
+		   IRQ_RCH_DESC_RESP |
+		   IRQ_RCH_DATA_RESP |
+		   IRQ_WCH_DESC_INVALID |
+		   IRQ_WCH_DESC_RESP |
+		   IRQ_WCH_DATA_RESP |
+		   IRQ_DESC_ERR))
+		spisg->status = sts;
+	else if (sts & IRQ_DESC_CHAIN_DONE)
+		spisg->status = 0;
+	else
+		return IRQ_NONE;
+
+	complete(&spisg->completion);
+
+	return IRQ_HANDLED;
+}
+
+static int aml_spisg_transfer_one_message(struct spi_controller *ctlr,
+					  struct spi_message *msg)
+{
+	struct spisg_device *spisg = spi_controller_get_devdata(ctlr);
+	struct device *dev = &spisg->pdev->dev;
+	unsigned long long ms = 0;
+	struct spi_transfer *xfer;
+	struct spisg_descriptor *descs, *desc;
+	struct spisg_descriptor_extra *exdescs, *exdesc;
+	dma_addr_t descs_paddr;
+	int desc_num = 1, descs_len;
+	u32 cs_hold_in_sclk = 0;
+	int ret = -EIO;
+
+	if (!aml_spisg_sem_down_read(spisg)) {
+		spi_finalize_current_message(ctlr);
+		dev_err(dev, "controller busy\n");
+		return -EBUSY;
+	}
+
+	/* calculate the desc num for all xfer */
+	list_for_each_entry(xfer, &msg->transfers, transfer_list)
+		desc_num++;
+
+	/* alloc descriptor/extra-descriptor table */
+	descs = kcalloc(desc_num, sizeof(*desc) + sizeof(*exdesc),
+			GFP_KERNEL | GFP_DMA);
+	if (!descs) {
+		spi_finalize_current_message(ctlr);
+		aml_spisg_sem_up_write(spisg);
+		return -ENOMEM;
+	}
+	descs_len = sizeof(*desc) * desc_num;
+	exdescs = (struct spisg_descriptor_extra *)(descs + desc_num);
+
+	/* config descriptor for each xfer */
+	desc = descs;
+	exdesc = exdescs;
+	list_for_each_entry(xfer, &msg->transfers, transfer_list) {
+		ret = aml_spisg_setup_transfer(spisg, xfer, desc, exdesc);
+		if (ret) {
+			dev_err(dev, "config descriptor failed\n");
+			goto end;
+		}
+
+		/* calculate cs-setup delay with the first xfer speed */
+		if (list_is_first(&xfer->transfer_list, &msg->transfers))
+			desc->cfg_bus |= FIELD_PREP(CFG_CS_SETUP,
+				spi_delay_to_sclk(xfer->effective_speed_hz, &msg->spi->cs_setup));
+
+		/* calculate cs-hold delay with the last xfer speed */
+		if (list_is_last(&xfer->transfer_list, &msg->transfers))
+			cs_hold_in_sclk =
+				spi_delay_to_sclk(xfer->effective_speed_hz, &msg->spi->cs_hold);
+
+		desc++;
+		exdesc++;
+		ms += DIV_ROUND_UP_ULL(8LL * MSEC_PER_SEC * xfer->len,
+				       xfer->effective_speed_hz);
+	}
+
+	if (cs_hold_in_sclk)
+		/* additional null-descriptor to achieve the cs-hold delay */
+		aml_spisg_setup_null_desc(spisg, desc, cs_hold_in_sclk);
+	else
+		desc--;
+
+	desc->cfg_bus |= FIELD_PREP(CFG_KEEP_SS, 0);
+	desc->cfg_start |= FIELD_PREP(CFG_EOC, 1);
+
+	/* some tolerances */
+	ms += ms + 20;
+	if (ms > UINT_MAX)
+		ms = UINT_MAX;
+
+	descs_paddr = dma_map_single(dev, (void *)descs,
+				     descs_len, DMA_TO_DEVICE);
+	ret = dma_mapping_error(dev, descs_paddr);
+	if (ret) {
+		dev_err(dev, "desc table map failed\n");
+		goto end;
+	}
+
+	reinit_completion(&spisg->completion);
+	aml_spisg_pending(spisg, descs_paddr, false, true);
+	if (wait_for_completion_timeout(&spisg->completion,
+					spi_controller_is_target(spisg->controller) ?
+					MAX_SCHEDULE_TIMEOUT : msecs_to_jiffies(ms)))
+		ret = spisg->status ? -EIO : 0;
+	else
+		ret = -ETIMEDOUT;
+
+	dma_unmap_single(dev, descs_paddr, descs_len, DMA_TO_DEVICE);
+end:
+	desc = descs;
+	exdesc = exdescs;
+	list_for_each_entry(xfer, &msg->transfers, transfer_list)
+		aml_spisg_cleanup_transfer(spisg, xfer, desc++, exdesc++);
+	kfree(descs);
+
+	if (!ret)
+		msg->actual_length = msg->frame_length;
+	msg->status = ret;
+	spi_finalize_current_message(ctlr);
+	aml_spisg_sem_up_write(spisg);
+
+	return ret;
+}
+
+static int aml_spisg_prepare_message(struct spi_controller *ctlr,
+				     struct spi_message *message)
+{
+	struct spisg_device *spisg = spi_controller_get_devdata(ctlr);
+	struct spi_device *spi = message->spi;
+
+	if (!spi->bits_per_word || spi->bits_per_word % 8) {
+		dev_err(&spisg->pdev->dev, "invalid wordlen %d\n", spi->bits_per_word);
+		return -EINVAL;
+	}
+
+	spisg->bytes_per_word = spi->bits_per_word >> 3;
+
+	spisg->cfg_spi &= ~CFG_SLAVE_SELECT;
+	spisg->cfg_spi |= FIELD_PREP(CFG_SLAVE_SELECT, spi_get_chipselect(spi, 0));
+
+	spisg->cfg_bus &= ~(CFG_CPOL | CFG_CPHA | CFG_B_L_ENDIAN | CFG_HALF_DUPLEX);
+	spisg->cfg_bus |= FIELD_PREP(CFG_CPOL, !!(spi->mode & SPI_CPOL)) |
+			  FIELD_PREP(CFG_CPHA, !!(spi->mode & SPI_CPHA)) |
+			  FIELD_PREP(CFG_B_L_ENDIAN, !!(spi->mode & SPI_LSB_FIRST)) |
+			  FIELD_PREP(CFG_HALF_DUPLEX, !!(spi->mode & SPI_3WIRE));
+
+	return 0;
+}
+
+static int aml_spisg_setup(struct spi_device *spi)
+{
+	if (!spi->controller_state)
+		spi->controller_state = spi_controller_get_devdata(spi->controller);
+
+	return 0;
+}
+
+static void aml_spisg_cleanup(struct spi_device *spi)
+{
+	spi->controller_state = NULL;
+}
+
+static int aml_spisg_target_abort(struct spi_controller *ctlr)
+{
+	struct spisg_device *spisg = spi_controller_get_devdata(ctlr);
+
+	spisg->status = 0;
+	regmap_write(spisg->map, SPISG_REG_DESC_LIST_H, 0);
+	complete(&spisg->completion);
+
+	return 0;
+}
+
+static int aml_spisg_clk_init(struct spisg_device *spisg, void __iomem *base)
+{
+	struct device *dev = &spisg->pdev->dev;
+	struct clk_init_data init;
+	struct clk_divider *div;
+	struct clk_div_table *tbl;
+	char name[32];
+	int ret, i;
+
+	spisg->core = devm_clk_get_enabled(dev, "core");
+	if (IS_ERR_OR_NULL(spisg->core)) {
+		dev_err(dev, "core clock request failed\n");
+		return PTR_ERR(spisg->core);
+	}
+
+	spisg->pclk = devm_clk_get_enabled(dev, "pclk");
+	if (IS_ERR_OR_NULL(spisg->pclk)) {
+		dev_err(dev, "pclk clock request failed\n");
+		return PTR_ERR(spisg->pclk);
+	}
+
+	clk_set_min_rate(spisg->pclk, SPISG_PCLK_RATE_MIN);
+
+	clk_disable_unprepare(spisg->pclk);
+
+	tbl = devm_kzalloc(dev, sizeof(struct clk_div_table) * (DIV_NUM + 1), GFP_KERNEL);
+	if (!tbl)
+		return -ENOMEM;
+
+	for (i = 0; i < DIV_NUM; i++) {
+		tbl[i].val = i + SPISG_CLK_DIV_MIN - 1;
+		tbl[i].div = i + SPISG_CLK_DIV_MIN;
+	}
+	spisg->tbl = tbl;
+
+	div = devm_kzalloc(dev, sizeof(*div), GFP_KERNEL);
+	if (!div)
+		return -ENOMEM;
+
+	div->flags = CLK_DIVIDER_ROUND_CLOSEST;
+	div->reg = base + SPISG_REG_CFG_BUS;
+	div->shift = __bf_shf(CFG_CLK_DIV);
+	div->width = CLK_DIV_WIDTH;
+	div->table = tbl;
+
+	/* Register value should not be outside of the table */
+	regmap_update_bits(spisg->map, SPISG_REG_CFG_BUS, CFG_CLK_DIV,
+			   FIELD_PREP(CFG_CLK_DIV, SPISG_CLK_DIV_MIN - 1));
+
+	/* Register clk-divider */
+	snprintf(name, sizeof(name), "%s_div", dev_name(dev));
+	init.name = name;
+	init.ops = &clk_divider_ops;
+	init.flags = CLK_SET_RATE_PARENT;
+	init.parent_data = &(const struct clk_parent_data) {
+				.fw_name = "pclk",
+			   };
+	init.num_parents = 1;
+	div->hw.init = &init;
+	ret = devm_clk_hw_register(dev, &div->hw);
+	if (ret) {
+		dev_err(dev, "clock registration failed\n");
+		return ret;
+	}
+
+	spisg->sclk = devm_clk_hw_get_clk(dev, &div->hw, NULL);
+	if (IS_ERR_OR_NULL(spisg->sclk)) {
+		dev_err(dev, "get clock failed\n");
+		return PTR_ERR(spisg->sclk);
+	}
+
+	clk_prepare_enable(spisg->sclk);
+
+	return 0;
+}
+
+static int aml_spisg_probe(struct platform_device *pdev)
+{
+	struct spi_controller *ctlr;
+	struct spisg_device *spisg;
+	struct device *dev = &pdev->dev;
+	void __iomem *base;
+	int ret, irq;
+
+	const struct regmap_config aml_regmap_config = {
+		.reg_bits = 32,
+		.val_bits = 32,
+		.reg_stride = 4,
+		.max_register = SPISG_MAX_REG,
+	};
+
+	if (of_property_read_bool(dev->of_node, "spi-slave"))
+		ctlr = spi_alloc_target(dev, sizeof(*spisg));
+	else
+		ctlr = spi_alloc_host(dev, sizeof(*spisg));
+	if (!ctlr)
+		return dev_err_probe(dev, -ENOMEM, "controller allocation failed\n");
+
+	spisg = spi_controller_get_devdata(ctlr);
+	spisg->controller = ctlr;
+
+	spisg->pdev = pdev;
+	platform_set_drvdata(pdev, spisg);
+
+	base = devm_platform_ioremap_resource(pdev, 0);
+	if (IS_ERR(base))
+		return dev_err_probe(dev, PTR_ERR(base), "resource ioremap failed\n");
+
+	spisg->map = devm_regmap_init_mmio(dev, base, &aml_regmap_config);
+	if (IS_ERR(spisg->map))
+		return dev_err_probe(dev, PTR_ERR(spisg->map), "regmap init failed\n");
+
+	irq = platform_get_irq(pdev, 0);
+	if (irq < 0) {
+		ret = irq;
+		goto out_controller;
+	}
+
+	ret = device_reset_optional(dev);
+	if (ret)
+		return dev_err_probe(dev, ret, "reset dev failed\n");
+
+	ret = aml_spisg_clk_init(spisg, base);
+	if (ret)
+		return dev_err_probe(dev, ret, "clock init failed\n");
+
+	spisg->cfg_spi = 0;
+	spisg->cfg_start = 0;
+	spisg->cfg_bus = 0;
+
+	spisg->cfg_spi = FIELD_PREP(CFG_SFLASH_WP, 1) |
+			 FIELD_PREP(CFG_SFLASH_HD, 1);
+	if (spi_controller_is_target(ctlr)) {
+		spisg->cfg_spi |= FIELD_PREP(CFG_SLAVE_EN, 1);
+		spisg->cfg_bus = FIELD_PREP(CFG_TX_TUNING, 0xf);
+	}
+	/* default pending */
+	spisg->cfg_start = FIELD_PREP(CFG_PEND, 1);
+
+	pm_runtime_set_active(&spisg->pdev->dev);
+	pm_runtime_enable(&spisg->pdev->dev);
+	pm_runtime_resume_and_get(&spisg->pdev->dev);
+
+	ctlr->num_chipselect = 4;
+	ctlr->dev.of_node = pdev->dev.of_node;
+	ctlr->mode_bits = SPI_CPHA | SPI_CPOL | SPI_LSB_FIRST |
+			  SPI_3WIRE | SPI_TX_QUAD | SPI_RX_QUAD;
+	ctlr->max_speed_hz = 1000 * 1000 * 100;
+	ctlr->min_speed_hz = 1000 * 10;
+	ctlr->setup = aml_spisg_setup;
+	ctlr->cleanup = aml_spisg_cleanup;
+	ctlr->prepare_message = aml_spisg_prepare_message;
+	ctlr->transfer_one_message = aml_spisg_transfer_one_message;
+	ctlr->target_abort = aml_spisg_target_abort;
+	ctlr->can_dma = aml_spisg_can_dma;
+	ctlr->max_dma_len = SPISG_BLOCK_MAX;
+	ctlr->auto_runtime_pm = true;
+
+	dma_set_max_seg_size(&pdev->dev, SPISG_BLOCK_MAX);
+
+	ret = devm_request_irq(&pdev->dev, irq, aml_spisg_irq, 0, NULL, spisg);
+	if (ret) {
+		dev_err(&pdev->dev, "irq request failed\n");
+		goto out_clk;
+	}
+
+	ret = devm_spi_register_controller(dev, ctlr);
+	if (ret) {
+		dev_err(&pdev->dev, "spi controller registration failed\n");
+		goto out_clk;
+	}
+
+	init_completion(&spisg->completion);
+
+	pm_runtime_put(&spisg->pdev->dev);
+
+	return 0;
+out_clk:
+	if (spisg->core)
+		clk_disable_unprepare(spisg->core);
+	clk_disable_unprepare(spisg->pclk);
+out_controller:
+	spi_controller_put(ctlr);
+
+	return ret;
+}
+
+static void aml_spisg_remove(struct platform_device *pdev)
+{
+	struct spisg_device *spisg = platform_get_drvdata(pdev);
+
+	if (!pm_runtime_suspended(&pdev->dev)) {
+		pinctrl_pm_select_sleep_state(&spisg->pdev->dev);
+		clk_disable_unprepare(spisg->core);
+		clk_disable_unprepare(spisg->pclk);
+	}
+}
+
+static int spisg_suspend_runtime(struct device *dev)
+{
+	struct spisg_device *spisg = dev_get_drvdata(dev);
+
+	pinctrl_pm_select_sleep_state(&spisg->pdev->dev);
+	clk_disable_unprepare(spisg->sclk);
+	clk_disable_unprepare(spisg->core);
+
+	return 0;
+}
+
+static int spisg_resume_runtime(struct device *dev)
+{
+	struct spisg_device *spisg = dev_get_drvdata(dev);
+
+	clk_prepare_enable(spisg->core);
+	clk_prepare_enable(spisg->sclk);
+	pinctrl_pm_select_default_state(&spisg->pdev->dev);
+
+	return 0;
+}
+
+static const struct dev_pm_ops amlogic_spisg_pm_ops = {
+	.runtime_suspend	= spisg_suspend_runtime,
+	.runtime_resume		= spisg_resume_runtime,
+};
+
+static const struct of_device_id amlogic_spisg_of_match[] = {
+	{
+		.compatible = "amlogic,a4-spisg",
+	},
+
+	{ /* sentinel */ }
+};
+MODULE_DEVICE_TABLE(of, amlogic_spisg_of_match);
+
+static struct platform_driver amlogic_spisg_driver = {
+	.probe = aml_spisg_probe,
+	.remove = aml_spisg_remove,
+	.driver  = {
+		.name = "amlogic-spisg",
+		.of_match_table = amlogic_spisg_of_match,
+		.pm = &amlogic_spisg_pm_ops,
+	},
+};
+
+module_platform_driver(amlogic_spisg_driver);
+
+MODULE_DESCRIPTION("Amlogic SPI Scatter-Gather Controller driver");
+MODULE_AUTHOR("Sunny Luo <sunny.luo@amlogic.com>");
+MODULE_LICENSE("GPL");
--- a/drivers/spi/spi-cadence-quadspi.c
+++ b/drivers/spi/spi-cadence-quadspi.c
@@ -1469,7 +1469,6 @@ static int cqspi_exec_mem_op(struct spi_mem *mem, const struct spi_mem_op *op)

 	ret = cqspi_mem_process(mem, op);

-	pm_runtime_mark_last_busy(dev);
 	pm_runtime_put_autosuspend(dev);

 	if (ret)
@@ -1970,7 +1969,6 @@ static int cqspi_probe(struct platform_device *pdev)
 		goto probe_setup_failed;
 	}

-	pm_runtime_mark_last_busy(dev);
 	pm_runtime_put_autosuspend(dev);

 	return 0;
--- a/drivers/spi/spi-cadence.c
+++ b/drivers/spi/spi-cadence.c
@@ -662,7 +662,6 @@ static int cdns_spi_probe(struct platform_device *pdev)
 		/* Set to default valid value */
 		ctlr->max_speed_hz = xspi->clk_rate / 4;
 		xspi->speed_hz = ctlr->max_speed_hz;
-		pm_runtime_mark_last_busy(&pdev->dev);
 		pm_runtime_put_autosuspend(&pdev->dev);
 	} else {
 		ctlr->mode_bits |= SPI_NO_CS;
--- a/drivers/spi/spi-falcon.c
+++ b/drivers/spi/spi-falcon.c
@@ -94,8 +94,9 @@ struct falcon_sflash {
 	struct spi_controller *host;
 };

-int falcon_sflash_xfer(struct spi_device *spi, struct spi_transfer *t,
-		unsigned long flags)
+static int
+falcon_sflash_xfer(struct spi_device *spi, struct spi_transfer *t,
+		   unsigned long flags)
 {
 	struct device *dev = &spi->dev;
 	struct falcon_sflash *priv = spi_controller_get_devdata(spi->controller);
--- a/drivers/spi/spi-fsl-dspi.c
+++ b/drivers/spi/spi-fsl-dspi.c
@@ -24,6 +24,7 @@

 #define SPI_MCR				0x00
 #define SPI_MCR_HOST			BIT(31)
+#define SPI_MCR_MTFE			BIT(26)
 #define SPI_MCR_PCSIS(x)		((x) << 16)
 #define SPI_MCR_CLR_TXF			BIT(11)
 #define SPI_MCR_CLR_RXF			BIT(10)
@@ -35,8 +36,9 @@
 #define SPI_TCR				0x08
 #define SPI_TCR_GET_TCNT(x)		(((x) & GENMASK(31, 16)) >> 16)

-#define SPI_CTAR(x)			(0x0c + (((x) & GENMASK(1, 0)) * 4))
+#define SPI_CTAR(x)			(0x0c + (((x) & GENMASK(2, 0)) * 4))
 #define SPI_CTAR_FMSZ(x)		(((x) << 27) & GENMASK(30, 27))
+#define SPI_CTAR_DBR			BIT(31)
 #define SPI_CTAR_CPOL			BIT(26)
 #define SPI_CTAR_CPHA			BIT(25)
 #define SPI_CTAR_LSBFE			BIT(24)
@@ -93,12 +95,14 @@
 #define SPI_TXFR1			0x40
 #define SPI_TXFR2			0x44
 #define SPI_TXFR3			0x48
+#define SPI_TXFR4			0x4C
 #define SPI_RXFR0			0x7c
 #define SPI_RXFR1			0x80
 #define SPI_RXFR2			0x84
 #define SPI_RXFR3			0x88
+#define SPI_RXFR4			0x8C

-#define SPI_CTARE(x)			(0x11c + (((x) & GENMASK(1, 0)) * 4))
+#define SPI_CTARE(x)			(0x11c + (((x) & GENMASK(2, 0)) * 4))
 #define SPI_CTARE_FMSZE(x)		(((x) & 0x1) << 16)
 #define SPI_CTARE_DTCP(x)		((x) & 0x7ff)

@@ -109,6 +113,8 @@

 #define DMA_COMPLETION_TIMEOUT		msecs_to_jiffies(3000)

+#define SPI_25MHZ			25000000
+
 struct chip_data {
 	u32			ctar_val;
 };
@@ -122,6 +128,7 @@ struct fsl_dspi_devtype_data {
 	enum dspi_trans_mode	trans_mode;
 	u8			max_clock_factor;
 	int			fifo_size;
+	const struct regmap_config *regmap;
 };

 enum {
@@ -135,6 +142,102 @@ enum {
 	LX2160A,
 	MCF5441X,
 	VF610,
+	S32G,
+	S32G_TARGET,
+};
+
+static const struct regmap_range dspi_yes_ranges[] = {
+	regmap_reg_range(SPI_MCR, SPI_MCR),
+	regmap_reg_range(SPI_TCR, SPI_CTAR(3)),
+	regmap_reg_range(SPI_SR, SPI_TXFR3),
+	regmap_reg_range(SPI_RXFR0, SPI_RXFR3),
+	regmap_reg_range(SPI_CTARE(0), SPI_CTARE(3)),
+	regmap_reg_range(SPI_SREX, SPI_SREX),
+};
+
+static const struct regmap_range s32g_dspi_yes_ranges[] = {
+	regmap_reg_range(SPI_MCR, SPI_MCR),
+	regmap_reg_range(SPI_TCR, SPI_CTAR(5)),
+	regmap_reg_range(SPI_SR, SPI_TXFR4),
+	regmap_reg_range(SPI_RXFR0, SPI_RXFR4),
+	regmap_reg_range(SPI_CTARE(0), SPI_CTARE(5)),
+	regmap_reg_range(SPI_SREX, SPI_SREX),
+};
+
+static const struct regmap_access_table dspi_access_table = {
+	.yes_ranges	= dspi_yes_ranges,
+	.n_yes_ranges	= ARRAY_SIZE(dspi_yes_ranges),
+};
+
+static const struct regmap_access_table s32g_dspi_access_table = {
+	.yes_ranges	= s32g_dspi_yes_ranges,
+	.n_yes_ranges	= ARRAY_SIZE(s32g_dspi_yes_ranges),
+};
+
+static const struct regmap_range dspi_volatile_ranges[] = {
+	regmap_reg_range(SPI_MCR, SPI_TCR),
+	regmap_reg_range(SPI_SR, SPI_SR),
+	regmap_reg_range(SPI_PUSHR, SPI_RXFR4),
+	regmap_reg_range(SPI_SREX, SPI_SREX),
+};
+
+static const struct regmap_access_table dspi_volatile_table = {
+	.yes_ranges	= dspi_volatile_ranges,
+	.n_yes_ranges	= ARRAY_SIZE(dspi_volatile_ranges),
+};
+
+enum {
+	DSPI_REGMAP,
+	S32G_DSPI_REGMAP,
+	DSPI_XSPI_REGMAP,
+	S32G_DSPI_XSPI_REGMAP,
+	DSPI_PUSHR,
+};
+
+static const struct regmap_config dspi_regmap_config[] = {
+	[DSPI_REGMAP] = {
+		.reg_bits	= 32,
+		.val_bits	= 32,
+		.reg_stride	= 4,
+		.max_register	= SPI_RXFR3,
+		.volatile_table	= &dspi_volatile_table,
+		.rd_table	= &dspi_access_table,
+		.wr_table	= &dspi_access_table,
+	},
+	[S32G_DSPI_REGMAP] = {
+		.reg_bits	= 32,
+		.val_bits	= 32,
+		.reg_stride	= 4,
+		.max_register	= SPI_RXFR4,
+		.volatile_table	= &dspi_volatile_table,
+		.wr_table	= &s32g_dspi_access_table,
+		.rd_table	= &s32g_dspi_access_table,
+	},
+	[DSPI_XSPI_REGMAP] = {
+		.reg_bits	= 32,
+		.val_bits	= 32,
+		.reg_stride	= 4,
+		.max_register	= SPI_SREX,
+		.volatile_table	= &dspi_volatile_table,
+		.rd_table	= &dspi_access_table,
+		.wr_table	= &dspi_access_table,
+	},
+	[S32G_DSPI_XSPI_REGMAP] = {
+		.reg_bits	= 32,
+		.val_bits	= 32,
+		.reg_stride	= 4,
+		.max_register	= SPI_SREX,
+		.volatile_table	= &dspi_volatile_table,
+		.wr_table	= &s32g_dspi_access_table,
+		.rd_table	= &s32g_dspi_access_table,
+	},
+	[DSPI_PUSHR] = {
+		.name		= "pushr",
+		.reg_bits	= 16,
+		.val_bits	= 16,
+		.reg_stride	= 2,
+		.max_register	= 0x2,
+	},
 };

 static const struct fsl_dspi_devtype_data devtype_data[] = {
@@ -142,55 +245,77 @@ static const struct fsl_dspi_devtype_data devtype_data[] = {
 		.trans_mode		= DSPI_DMA_MODE,
 		.max_clock_factor	= 2,
 		.fifo_size		= 4,
+		.regmap			= &dspi_regmap_config[DSPI_REGMAP],
 	},
 	[LS1021A] = {
 		/* Has A-011218 DMA erratum */
 		.trans_mode		= DSPI_XSPI_MODE,
 		.max_clock_factor	= 8,
 		.fifo_size		= 4,
+		.regmap			= &dspi_regmap_config[DSPI_XSPI_REGMAP],
 	},
 	[LS1012A] = {
 		/* Has A-011218 DMA erratum */
 		.trans_mode		= DSPI_XSPI_MODE,
 		.max_clock_factor	= 8,
 		.fifo_size		= 16,
+		.regmap			= &dspi_regmap_config[DSPI_XSPI_REGMAP],
 	},
 	[LS1028A] = {
 		.trans_mode		= DSPI_XSPI_MODE,
 		.max_clock_factor	= 8,
 		.fifo_size		= 4,
+		.regmap			= &dspi_regmap_config[DSPI_XSPI_REGMAP],
 	},
 	[LS1043A] = {
 		/* Has A-011218 DMA erratum */
 		.trans_mode		= DSPI_XSPI_MODE,
 		.max_clock_factor	= 8,
 		.fifo_size		= 16,
+		.regmap			= &dspi_regmap_config[DSPI_XSPI_REGMAP],
 	},
 	[LS1046A] = {
 		/* Has A-011218 DMA erratum */
 		.trans_mode		= DSPI_XSPI_MODE,
 		.max_clock_factor	= 8,
 		.fifo_size		= 16,
+		.regmap			= &dspi_regmap_config[DSPI_XSPI_REGMAP],
 	},
 	[LS2080A] = {
 		.trans_mode		= DSPI_XSPI_MODE,
 		.max_clock_factor	= 8,
 		.fifo_size		= 4,
+		.regmap			= &dspi_regmap_config[DSPI_XSPI_REGMAP],
 	},
 	[LS2085A] = {
 		.trans_mode		= DSPI_XSPI_MODE,
 		.max_clock_factor	= 8,
 		.fifo_size		= 4,
+		.regmap			= &dspi_regmap_config[DSPI_XSPI_REGMAP],
 	},
 	[LX2160A] = {
 		.trans_mode		= DSPI_XSPI_MODE,
 		.max_clock_factor	= 8,
 		.fifo_size		= 4,
+		.regmap			= &dspi_regmap_config[DSPI_XSPI_REGMAP],
 	},
 	[MCF5441X] = {
 		.trans_mode		= DSPI_DMA_MODE,
 		.max_clock_factor	= 8,
 		.fifo_size		= 16,
+		.regmap			= &dspi_regmap_config[DSPI_REGMAP],
+	},
+	[S32G] = {
+		.trans_mode	  = DSPI_XSPI_MODE,
+		.max_clock_factor = 1,
+		.fifo_size	  = 5,
+		.regmap		  = &dspi_regmap_config[S32G_DSPI_XSPI_REGMAP],
+	},
+	[S32G_TARGET] = {
+		.trans_mode	  = DSPI_DMA_MODE,
+		.max_clock_factor = 1,
+		.fifo_size	  = 5,
+		.regmap		  = &dspi_regmap_config[S32G_DSPI_REGMAP],
 	},
 };

@@ -225,6 +350,7 @@ struct fsl_dspi {
 	const void				*tx;
 	void					*rx;
 	u16					tx_cmd;
+	bool					mtf_enabled;
 	const struct fsl_dspi_devtype_data	*devtype_data;

 	struct completion			xfer_done;
@@ -247,6 +373,12 @@ struct fsl_dspi {
 	void (*dev_to_host)(struct fsl_dspi *dspi, u32 rxdata);
 };

+static bool is_s32g_dspi(struct fsl_dspi *data)
+{
+	return data->devtype_data == &devtype_data[S32G] ||
+	       data->devtype_data == &devtype_data[S32G_TARGET];
+}
+
 static void dspi_native_host_to_dev(struct fsl_dspi *dspi, u32 *txdata)
 {
 	switch (dspi->oper_word_size) {
@@ -595,7 +727,7 @@ static void dspi_release_dma(struct fsl_dspi *dspi)
 }

 static void hz_to_spi_baud(char *pbr, char *br, int speed_hz,
-			   unsigned long clkrate)
+			   unsigned long clkrate, bool mtf_enabled)
 {
 	/* Valid baud rate pre-scaler values */
 	int pbr_tbl[4] = {2, 3, 5, 7};
@@ -612,7 +744,13 @@ static void hz_to_spi_baud(char *pbr, char *br, int speed_hz,

 	for (i = 0; i < ARRAY_SIZE(brs); i++)
 		for (j = 0; j < ARRAY_SIZE(pbr_tbl); j++) {
-			scale = brs[i] * pbr_tbl[j];
+			if (mtf_enabled) {
+				/* In MTF mode DBR=1 so frequency is doubled */
+				scale = (brs[i] * pbr_tbl[j]) / 2;
+			} else {
+				scale = brs[i] * pbr_tbl[j];
+			}
+
 			if (scale >= scale_needed) {
 				if (scale < minscale) {
 					minscale = scale;
@@ -746,8 +884,12 @@ static void dspi_setup_accel(struct fsl_dspi *dspi)
 	struct spi_transfer *xfer = dspi->cur_transfer;
 	bool odd = !!(dspi->len & 1);

-	/* No accel for frames not multiple of 8 bits at the moment */
-	if (xfer->bits_per_word % 8)
+	/*
+	 * No accel for DMA transfers or frames not multiples of 8 bits at the
+	 * moment.
+	 */
+	if (dspi->devtype_data->trans_mode == DSPI_DMA_MODE ||
+	    xfer->bits_per_word % 8)
 		goto no_accel;

 	if (!odd && dspi->len <= dspi->devtype_data->fifo_size * 2) {
@@ -756,10 +898,7 @@ static void dspi_setup_accel(struct fsl_dspi *dspi)
 		dspi->oper_bits_per_word = 8;
 	} else {
 		/* Start off with maximum supported by hardware */
-		if (dspi->devtype_data->trans_mode == DSPI_XSPI_MODE)
-			dspi->oper_bits_per_word = 32;
-		else
-			dspi->oper_bits_per_word = 16;
+		dspi->oper_bits_per_word = 32;

 		/*
 		 * And go down only if the buffer can't be sent with
@@ -1027,6 +1166,20 @@ static int dspi_transfer_one_message(struct spi_controller *ctlr,
 	return status;
 }

+static int dspi_set_mtf(struct fsl_dspi *dspi)
+{
+	if (spi_controller_is_target(dspi->ctlr))
+		return 0;
+
+	if (dspi->mtf_enabled)
+		regmap_update_bits(dspi->regmap, SPI_MCR, SPI_MCR_MTFE,
+				   SPI_MCR_MTFE);
+	else
+		regmap_update_bits(dspi->regmap, SPI_MCR, SPI_MCR_MTFE, 0);
+
+	return 0;
+}
+
 static int dspi_setup(struct spi_device *spi)
 {
 	struct fsl_dspi *dspi = spi_controller_get_devdata(spi->controller);
@@ -1085,7 +1238,16 @@ static int dspi_setup(struct spi_device *spi)
 		cs_sck_delay, sck_cs_delay);

 	clkrate = clk_get_rate(dspi->clk);
-	hz_to_spi_baud(&pbr, &br, spi->max_speed_hz, clkrate);
+
+	if (is_s32g_dspi(dspi) && spi->max_speed_hz > SPI_25MHZ)
+		dspi->mtf_enabled = true;
+	else
+		dspi->mtf_enabled = false;
+
+	dspi_set_mtf(dspi);
+
+	hz_to_spi_baud(&pbr, &br, spi->max_speed_hz, clkrate,
+		       dspi->mtf_enabled);

 	/* Set PCS to SCK delay scale values */
 	ns_delay_scale(&pcssck, &cssck, cs_sck_delay, clkrate);
@@ -1107,6 +1269,9 @@ static int dspi_setup(struct spi_device *spi)
 				  SPI_CTAR_PBR(pbr) |
 				  SPI_CTAR_BR(br);

+		if (dspi->mtf_enabled)
+			chip->ctar_val |= SPI_CTAR_DBR;
+
 		if (spi->mode & SPI_LSB_FIRST)
 			chip->ctar_val |= SPI_CTAR_LSBFE;
 	}
@@ -1160,112 +1325,14 @@ static const struct of_device_id fsl_dspi_dt_ids[] = {
 	}, {
 		.compatible = "fsl,lx2160a-dspi",
 		.data = &devtype_data[LX2160A],
+	}, {
+		.compatible = "nxp,s32g2-dspi",
+		.data = &devtype_data[S32G],
 	},
 	{ /* sentinel */ }
 };
 MODULE_DEVICE_TABLE(of, fsl_dspi_dt_ids);

-#ifdef CONFIG_PM_SLEEP
-static int dspi_suspend(struct device *dev)
-{
-	struct fsl_dspi *dspi = dev_get_drvdata(dev);
-
-	if (dspi->irq)
-		disable_irq(dspi->irq);
-	spi_controller_suspend(dspi->ctlr);
-	clk_disable_unprepare(dspi->clk);
-
-	pinctrl_pm_select_sleep_state(dev);
-
-	return 0;
-}
-
-static int dspi_resume(struct device *dev)
-{
-	struct fsl_dspi *dspi = dev_get_drvdata(dev);
-	int ret;
-
-	pinctrl_pm_select_default_state(dev);
-
-	ret = clk_prepare_enable(dspi->clk);
-	if (ret)
-		return ret;
-	spi_controller_resume(dspi->ctlr);
-	if (dspi->irq)
-		enable_irq(dspi->irq);
-
-	return 0;
-}
-#endif /* CONFIG_PM_SLEEP */
-
-static SIMPLE_DEV_PM_OPS(dspi_pm, dspi_suspend, dspi_resume);
-
-static const struct regmap_range dspi_yes_ranges[] = {
-	regmap_reg_range(SPI_MCR, SPI_MCR),
-	regmap_reg_range(SPI_TCR, SPI_CTAR(3)),
-	regmap_reg_range(SPI_SR, SPI_TXFR3),
-	regmap_reg_range(SPI_RXFR0, SPI_RXFR3),
-	regmap_reg_range(SPI_CTARE(0), SPI_CTARE(3)),
-	regmap_reg_range(SPI_SREX, SPI_SREX),
-};
-
-static const struct regmap_access_table dspi_access_table = {
-	.yes_ranges	= dspi_yes_ranges,
-	.n_yes_ranges	= ARRAY_SIZE(dspi_yes_ranges),
-};
-
-static const struct regmap_range dspi_volatile_ranges[] = {
-	regmap_reg_range(SPI_MCR, SPI_TCR),
-	regmap_reg_range(SPI_SR, SPI_SR),
-	regmap_reg_range(SPI_PUSHR, SPI_RXFR3),
-};
-
-static const struct regmap_access_table dspi_volatile_table = {
-	.yes_ranges	= dspi_volatile_ranges,
-	.n_yes_ranges	= ARRAY_SIZE(dspi_volatile_ranges),
-};
-
-static const struct regmap_config dspi_regmap_config = {
-	.reg_bits	= 32,
-	.val_bits	= 32,
-	.reg_stride	= 4,
-	.max_register	= 0x88,
-	.volatile_table	= &dspi_volatile_table,
-	.rd_table	= &dspi_access_table,
-	.wr_table	= &dspi_access_table,
-};
-
-static const struct regmap_range dspi_xspi_volatile_ranges[] = {
-	regmap_reg_range(SPI_MCR, SPI_TCR),
-	regmap_reg_range(SPI_SR, SPI_SR),
-	regmap_reg_range(SPI_PUSHR, SPI_RXFR3),
-	regmap_reg_range(SPI_SREX, SPI_SREX),
-};
-
-static const struct regmap_access_table dspi_xspi_volatile_table = {
-	.yes_ranges	= dspi_xspi_volatile_ranges,
-	.n_yes_ranges	= ARRAY_SIZE(dspi_xspi_volatile_ranges),
-};
-
-static const struct regmap_config dspi_xspi_regmap_config[] = {
-	{
-		.reg_bits	= 32,
-		.val_bits	= 32,
-		.reg_stride	= 4,
-		.max_register	= 0x13c,
-		.volatile_table	= &dspi_xspi_volatile_table,
-		.rd_table	= &dspi_access_table,
-		.wr_table	= &dspi_access_table,
-	},
-	{
-		.name		= "pushr",
-		.reg_bits	= 16,
-		.val_bits	= 16,
-		.reg_stride	= 2,
-		.max_register	= 0x2,
-	},
-};
-
 static int dspi_init(struct fsl_dspi *dspi)
 {
 	unsigned int mcr;
@@ -1301,6 +1368,50 @@ static int dspi_init(struct fsl_dspi *dspi)
 	return 0;
 }

+#ifdef CONFIG_PM_SLEEP
+static int dspi_suspend(struct device *dev)
+{
+	struct fsl_dspi *dspi = dev_get_drvdata(dev);
+
+	if (dspi->irq)
+		disable_irq(dspi->irq);
+	spi_controller_suspend(dspi->ctlr);
+	clk_disable_unprepare(dspi->clk);
+
+	pinctrl_pm_select_sleep_state(dev);
+
+	return 0;
+}
+
+static int dspi_resume(struct device *dev)
+{
+	struct fsl_dspi *dspi = dev_get_drvdata(dev);
+	int ret;
+
+	pinctrl_pm_select_default_state(dev);
+
+	ret = clk_prepare_enable(dspi->clk);
+	if (ret)
+		return ret;
+	spi_controller_resume(dspi->ctlr);
+
+	ret = dspi_init(dspi);
+	if (ret) {
+		dev_err(dev, "failed to initialize dspi during resume\n");
+		return ret;
+	}
+
+	dspi_set_mtf(dspi);
+
+	if (dspi->irq)
+		enable_irq(dspi->irq);
+
+	return 0;
+}
+#endif /* CONFIG_PM_SLEEP */
+
+static SIMPLE_DEV_PM_OPS(dspi_pm, dspi_suspend, dspi_resume);
+
 static int dspi_target_abort(struct spi_controller *host)
 {
 	struct fsl_dspi *dspi = spi_controller_get_devdata(host);
@@ -1325,7 +1436,6 @@ static int dspi_target_abort(struct spi_controller *host)
 static int dspi_probe(struct platform_device *pdev)
 {
 	struct device_node *np = pdev->dev.of_node;
-	const struct regmap_config *regmap_config;
 	struct fsl_dspi_platform_data *pdata;
 	struct spi_controller *ctlr;
 	int ret, cs_num, bus_num = -1;
@@ -1338,7 +1448,10 @@ static int dspi_probe(struct platform_device *pdev)
 	if (!dspi)
 		return -ENOMEM;

-	ctlr = spi_alloc_host(&pdev->dev, 0);
+	if (of_property_read_bool(np, "spi-slave"))
+		ctlr = spi_alloc_target(&pdev->dev, 0);
+	else
+		ctlr = spi_alloc_host(&pdev->dev, 0);
 	if (!ctlr)
 		return -ENOMEM;

@@ -1377,9 +1490,6 @@ static int dspi_probe(struct platform_device *pdev)
 		of_property_read_u32(np, "bus-num", &bus_num);
 		ctlr->bus_num = bus_num;

-		if (of_property_read_bool(np, "spi-slave"))
-			ctlr->target = true;
-
 		dspi->devtype_data = of_device_get_match_data(&pdev->dev);
 		if (!dspi->devtype_data) {
 			dev_err(&pdev->dev, "can't get devtype_data\n");
@@ -1397,6 +1507,9 @@ static int dspi_probe(struct platform_device *pdev)
 		dspi->pushr_tx = 0;
 	}

+	if (spi_controller_is_target(ctlr) && is_s32g_dspi(dspi))
+		dspi->devtype_data = &devtype_data[S32G_TARGET];
+
 	if (dspi->devtype_data->trans_mode == DSPI_XSPI_MODE)
 		ctlr->bits_per_word_mask = SPI_BPW_RANGE_MASK(4, 32);
 	else
@@ -1408,11 +1521,8 @@ static int dspi_probe(struct platform_device *pdev)
 		goto out_ctlr_put;
 	}

-	if (dspi->devtype_data->trans_mode == DSPI_XSPI_MODE)
-		regmap_config = &dspi_xspi_regmap_config[0];
-	else
-		regmap_config = &dspi_regmap_config;
-	dspi->regmap = devm_regmap_init_mmio(&pdev->dev, base, regmap_config);
+	dspi->regmap = devm_regmap_init_mmio(&pdev->dev, base,
+					     dspi->devtype_data->regmap);
 	if (IS_ERR(dspi->regmap)) {
 		dev_err(&pdev->dev, "failed to init regmap: %ld\n",
 				PTR_ERR(dspi->regmap));
@@ -1423,7 +1533,7 @@ static int dspi_probe(struct platform_device *pdev)
 	if (dspi->devtype_data->trans_mode == DSPI_XSPI_MODE) {
 		dspi->regmap_pushr = devm_regmap_init_mmio(
 			&pdev->dev, base + SPI_PUSHR,
-			&dspi_xspi_regmap_config[1]);
+			&dspi_regmap_config[DSPI_PUSHR]);
 		if (IS_ERR(dspi->regmap_pushr)) {
 			dev_err(&pdev->dev,
 				"failed to init pushr regmap: %ld\n",
--- a/drivers/spi/spi-fsl-espi.c
+++ b/drivers/spi/spi-fsl-espi.c
@@ -513,7 +513,6 @@ static int fsl_espi_setup(struct spi_device *spi)

 	fsl_espi_setup_transfer(spi, NULL);

-	pm_runtime_mark_last_busy(espi->dev);
 	pm_runtime_put_autosuspend(espi->dev);

 	return 0;
@@ -726,7 +725,6 @@ static int fsl_espi_probe(struct device *dev, struct resource *mem,

 	dev_info(dev, "irq = %u\n", irq);

-	pm_runtime_mark_last_busy(dev);
 	pm_runtime_put_autosuspend(dev);

 	return 0;
--- a/drivers/spi/spi-fsl-lpspi.c
+++ b/drivers/spi/spi-fsl-lpspi.c
@@ -233,7 +233,6 @@ static int lpspi_unprepare_xfer_hardware(struct spi_controller *controller)
 	struct fsl_lpspi_data *fsl_lpspi =
 				spi_controller_get_devdata(controller);

-	pm_runtime_mark_last_busy(fsl_lpspi->dev);
 	pm_runtime_put_autosuspend(fsl_lpspi->dev);

 	return 0;
@@ -966,7 +965,6 @@ static int fsl_lpspi_probe(struct platform_device *pdev)
 		goto free_dma;
 	}

-	pm_runtime_mark_last_busy(fsl_lpspi->dev);
 	pm_runtime_put_autosuspend(fsl_lpspi->dev);

 	return 0;
--- a/drivers/spi/spi-gpio.c
+++ b/drivers/spi/spi-gpio.c
@@ -104,7 +104,7 @@ static inline int getmiso(const struct spi_device *spi)
 */

 static u32 spi_gpio_txrx_word_mode0(struct spi_device *spi,
-		unsigned nsecs, u32 word, u8 bits, unsigned flags)
+		unsigned int nsecs, u32 word, u8 bits, unsigned int flags)
 {
 	if (unlikely(spi->mode & SPI_LSB_FIRST))
 		return bitbang_txrx_le_cpha0(spi, nsecs, 0, flags, word, bits);
@@ -113,7 +113,7 @@ static u32 spi_gpio_txrx_word_mode0(struct spi_device *spi,
 }

 static u32 spi_gpio_txrx_word_mode1(struct spi_device *spi,
-		unsigned nsecs, u32 word, u8 bits, unsigned flags)
+		unsigned int nsecs, u32 word, u8 bits, unsigned int flags)
 {
 	if (unlikely(spi->mode & SPI_LSB_FIRST))
 		return bitbang_txrx_le_cpha1(spi, nsecs, 0, flags, word, bits);
@@ -122,7 +122,7 @@ static u32 spi_gpio_txrx_word_mode1(struct spi_device *spi,
 }

 static u32 spi_gpio_txrx_word_mode2(struct spi_device *spi,
-		unsigned nsecs, u32 word, u8 bits, unsigned flags)
+		unsigned int nsecs, u32 word, u8 bits, unsigned int flags)
 {
 	if (unlikely(spi->mode & SPI_LSB_FIRST))
 		return bitbang_txrx_le_cpha0(spi, nsecs, 1, flags, word, bits);
@@ -131,7 +131,7 @@ static u32 spi_gpio_txrx_word_mode2(struct spi_device *spi,
 }

 static u32 spi_gpio_txrx_word_mode3(struct spi_device *spi,
-		unsigned nsecs, u32 word, u8 bits, unsigned flags)
+		unsigned int nsecs, u32 word, u8 bits, unsigned int flags)
 {
 	if (unlikely(spi->mode & SPI_LSB_FIRST))
 		return bitbang_txrx_le_cpha1(spi, nsecs, 1, flags, word, bits);
@@ -150,7 +150,7 @@ static u32 spi_gpio_txrx_word_mode3(struct spi_device *spi,
 */

 static u32 spi_gpio_spec_txrx_word_mode0(struct spi_device *spi,
-		unsigned nsecs, u32 word, u8 bits, unsigned flags)
+		unsigned int nsecs, u32 word, u8 bits, unsigned int flags)
 {
 	flags = spi->controller->flags;
 	if (unlikely(spi->mode & SPI_LSB_FIRST))
@@ -160,7 +160,7 @@ static u32 spi_gpio_spec_txrx_word_mode0(struct spi_device *spi,
 }

 static u32 spi_gpio_spec_txrx_word_mode1(struct spi_device *spi,
-		unsigned nsecs, u32 word, u8 bits, unsigned flags)
+		unsigned int nsecs, u32 word, u8 bits, unsigned int flags)
 {
 	flags = spi->controller->flags;
 	if (unlikely(spi->mode & SPI_LSB_FIRST))
@@ -170,7 +170,7 @@ static u32 spi_gpio_spec_txrx_word_mode1(struct spi_device *spi,
 }

 static u32 spi_gpio_spec_txrx_word_mode2(struct spi_device *spi,
-		unsigned nsecs, u32 word, u8 bits, unsigned flags)
+		unsigned int nsecs, u32 word, u8 bits, unsigned int flags)
 {
 	flags = spi->controller->flags;
 	if (unlikely(spi->mode & SPI_LSB_FIRST))
@@ -180,7 +180,7 @@ static u32 spi_gpio_spec_txrx_word_mode2(struct spi_device *spi,
 }

 static u32 spi_gpio_spec_txrx_word_mode3(struct spi_device *spi,
-		unsigned nsecs, u32 word, u8 bits, unsigned flags)
+		unsigned int nsecs, u32 word, u8 bits, unsigned int flags)
 {
 	flags = spi->controller->flags;
 	if (unlikely(spi->mode & SPI_LSB_FIRST))
--- a/drivers/spi/spi-imx.c
+++ b/drivers/spi/spi-imx.c
@@ -1748,7 +1748,6 @@ spi_imx_prepare_message(struct spi_controller *controller, struct spi_message *m

 	ret = spi_imx->devtype_data->prepare_message(spi_imx, msg);
 	if (ret) {
-		pm_runtime_mark_last_busy(spi_imx->dev);
 		pm_runtime_put_autosuspend(spi_imx->dev);
 	}

@@ -1760,7 +1759,6 @@ spi_imx_unprepare_message(struct spi_controller *controller, struct spi_message
 {
 	struct spi_imx_data *spi_imx = spi_controller_get_devdata(controller);

-	pm_runtime_mark_last_busy(spi_imx->dev);
 	pm_runtime_put_autosuspend(spi_imx->dev);
 	return 0;
 }
@@ -1933,7 +1931,6 @@ static int spi_imx_probe(struct platform_device *pdev)
 		goto out_register_controller;
 	}

-	pm_runtime_mark_last_busy(spi_imx->dev);
 	pm_runtime_put_autosuspend(spi_imx->dev);

 	return ret;
--- a/drivers/spi/spi-intel.c
+++ b/drivers/spi/spi-intel.c
@@ -189,6 +189,11 @@ struct intel_spi_mem_op {
 static bool writeable;
 module_param(writeable, bool, 0);
 MODULE_PARM_DESC(writeable, "Enable write access to SPI flash chip (default=0)");
+static bool ignore_protection_status;
+module_param(ignore_protection_status, bool, 0);
+MODULE_PARM_DESC(
+	ignore_protection_status,
+	"Do not block SPI flash chip write access even if it is write-protected (default=0)");

 static void intel_spi_dump_regs(struct intel_spi *ispi)
 {
@@ -1248,13 +1253,15 @@ static void intel_spi_fill_partition(struct intel_spi *ispi,
 			continue;

 		/*
-		 * If any of the regions have protection bits set, make the
-		 * whole partition read-only to be on the safe side.
+		 * If any of the regions have protection bits set and
+		 * the ignore protection status parameter is not set,
+		 * make the whole partition read-only to be on the safe side.
 		 *
 		 * Also if the user did not ask the chip to be writeable
 		 * mask the bit too.
 		 */
-		if (!writeable || intel_spi_is_protected(ispi, base, limit)) {
+		if (!writeable || (!ignore_protection_status &&
+				   intel_spi_is_protected(ispi, base, limit))) {
 			part->mask_flags |= MTD_WRITEABLE;
 			ispi->protected = true;
 		}
--- a/drivers/spi/spi-microchip-core-qspi.c
+++ b/drivers/spi/spi-microchip-core-qspi.c
@@ -194,7 +194,7 @@ static inline void mchp_coreqspi_read_op(struct mchp_coreqspi *qspi)
 	}
 }

-static inline void mchp_coreqspi_write_op(struct mchp_coreqspi *qspi, bool word)
+static inline void mchp_coreqspi_write_op(struct mchp_coreqspi *qspi)
 {
 	u32 control, data;

@@ -222,6 +222,87 @@ static inline void mchp_coreqspi_write_op(struct mchp_coreqspi *qspi, bool word)
 	}
 }

+static inline void mchp_coreqspi_write_read_op(struct mchp_coreqspi *qspi)
+{
+	u32 control, data;
+
+	qspi->rx_len = qspi->tx_len;
+
+	control = readl_relaxed(qspi->regs + REG_CONTROL);
+	control |= CONTROL_FLAGSX4;
+	writel_relaxed(control, qspi->regs + REG_CONTROL);
+
+	while (qspi->tx_len >= 4) {
+		while (readl_relaxed(qspi->regs + REG_STATUS) & STATUS_TXFIFOFULL)
+			;
+
+		data = qspi->txbuf ? *((u32 *)qspi->txbuf) : 0xaa;
+		if (qspi->txbuf)
+			qspi->txbuf += 4;
+		qspi->tx_len -= 4;
+		writel_relaxed(data, qspi->regs + REG_X4_TX_DATA);
+
+		/*
+		 * The rx FIFO is twice the size of the tx FIFO, so there is
+		 * no requirement to block transmission if receive data is not
+		 * ready, and it is fine to let the tx FIFO completely fill
+		 * without reading anything from the rx FIFO. Once the tx FIFO
+		 * has been filled and becomes non-full due to a transmission
+		 * occurring there will always be something to receive.
+		 * IOW, this is safe as TX_FIFO_SIZE + 4 < 2 * TX_FIFO_SIZE
+		 */
+		if (qspi->rx_len >= 4) {
+			if (readl_relaxed(qspi->regs + REG_STATUS) & STATUS_RXAVAILABLE) {
+				data = readl_relaxed(qspi->regs + REG_X4_RX_DATA);
+				*(u32 *)qspi->rxbuf = data;
+				qspi->rxbuf += 4;
+				qspi->rx_len -= 4;
+			}
+		}
+	}
+
+	/*
+	 * Since transmission is not being blocked by clearing the rx FIFO,
+	 * loop here until all received data "leaked" by the loop above has
+	 * been dealt with.
+	 */
+	while (qspi->rx_len >= 4) {
+		while (readl_relaxed(qspi->regs + REG_STATUS) & STATUS_RXFIFOEMPTY)
+			;
+		data = readl_relaxed(qspi->regs + REG_X4_RX_DATA);
+		*(u32 *)qspi->rxbuf = data;
+		qspi->rxbuf += 4;
+		qspi->rx_len -= 4;
+	}
+
+	/*
+	 * Since rx_len and tx_len must be < 4 bytes at this point, there's no
+	 * concern about overflowing the rx or tx FIFOs any longer. It's
+	 * therefore safe to loop over the remainder of the transmit data before
+	 * handling the remaining receive data.
+	 */
+	if (!qspi->tx_len)
+		return;
+
+	control &= ~CONTROL_FLAGSX4;
+	writel_relaxed(control, qspi->regs + REG_CONTROL);
+
+	while (qspi->tx_len--) {
+		while (readl_relaxed(qspi->regs + REG_STATUS) & STATUS_TXFIFOFULL)
+			;
+		data = qspi->txbuf ? *qspi->txbuf : 0xaa;
+		qspi->txbuf++;
+		writel_relaxed(data, qspi->regs + REG_TX_DATA);
+	}
+
+	while (qspi->rx_len--) {
+		while (readl_relaxed(qspi->regs + REG_STATUS) & STATUS_RXFIFOEMPTY)
+			;
+		data = readl_relaxed(qspi->regs + REG_RX_DATA);
+		*qspi->rxbuf++ = (data & 0xFF);
+	}
+}
+
 static void mchp_coreqspi_enable_ints(struct mchp_coreqspi *qspi)
 {
 	u32 mask = IEN_TXDONE |
@@ -266,7 +347,7 @@ static irqreturn_t mchp_coreqspi_isr(int irq, void *dev_id)
 }

 static int mchp_coreqspi_setup_clock(struct mchp_coreqspi *qspi, struct spi_device *spi,
-				     const struct spi_mem_op *op)
+				     u32 max_freq)
 {
 	unsigned long clk_hz;
 	u32 control, baud_rate_val = 0;
@@ -275,11 +356,11 @@ static int mchp_coreqspi_setup_clock(struct mchp_coreqspi *qspi, struct spi_devi
 	if (!clk_hz)
 		return -EINVAL;

-	baud_rate_val = DIV_ROUND_UP(clk_hz, 2 * op->max_freq);
+	baud_rate_val = DIV_ROUND_UP(clk_hz, 2 * max_freq);
 	if (baud_rate_val > MAX_DIVIDER || baud_rate_val < MIN_DIVIDER) {
 		dev_err(&spi->dev,
 			"could not configure the clock for spi clock %d Hz & system clock %ld Hz\n",
-			op->max_freq, clk_hz);
+			max_freq, clk_hz);
 		return -EINVAL;
 	}

@@ -367,23 +448,13 @@ static inline void mchp_coreqspi_config_op(struct mchp_coreqspi *qspi, const str
 	writel_relaxed(frames, qspi->regs + REG_FRAMES);
 }

-static int mchp_qspi_wait_for_ready(struct spi_mem *mem)
+static int mchp_coreqspi_wait_for_ready(struct mchp_coreqspi *qspi)
 {
-	struct mchp_coreqspi *qspi = spi_controller_get_devdata
-				    (mem->spi->controller);
 	u32 status;
-	int ret;

-	ret = readl_poll_timeout(qspi->regs + REG_STATUS, status,
+	return readl_poll_timeout(qspi->regs + REG_STATUS, status,
 				 (status & STATUS_READY), 0,
 				 TIMEOUT_MS);
-	if (ret) {
-		dev_err(&mem->spi->dev,
-			"Timeout waiting on QSPI ready.\n");
-		return -ETIMEDOUT;
-	}
-
-	return ret;
 }

 static int mchp_coreqspi_exec_op(struct spi_mem *mem, const struct spi_mem_op *op)
@@ -396,11 +467,13 @@ static int mchp_coreqspi_exec_op(struct spi_mem *mem, const struct spi_mem_op *o
 	int err, i;

 	mutex_lock(&qspi->op_lock);
-	err = mchp_qspi_wait_for_ready(mem);
-	if (err)
+	err = mchp_coreqspi_wait_for_ready(qspi);
+	if (err) {
+		dev_err(&mem->spi->dev, "Timeout waiting on QSPI ready.\n");
 		goto error;
+	}

-	err = mchp_coreqspi_setup_clock(qspi, mem->spi, op);
+	err = mchp_coreqspi_setup_clock(qspi, mem->spi, op->max_freq);
 	if (err)
 		goto error;

@@ -415,7 +488,7 @@ static int mchp_coreqspi_exec_op(struct spi_mem *mem, const struct spi_mem_op *o
 		qspi->rxbuf = NULL;
 		qspi->tx_len = op->cmd.nbytes;
 		qspi->rx_len = 0;
-		mchp_coreqspi_write_op(qspi, false);
+		mchp_coreqspi_write_op(qspi);
 	}

 	qspi->txbuf = &opaddr[0];
@@ -426,7 +499,7 @@ static int mchp_coreqspi_exec_op(struct spi_mem *mem, const struct spi_mem_op *o
 		qspi->rxbuf = NULL;
 		qspi->tx_len = op->addr.nbytes;
 		qspi->rx_len = 0;
-		mchp_coreqspi_write_op(qspi, false);
+		mchp_coreqspi_write_op(qspi);
 	}

 	if (op->data.nbytes) {
@@ -435,7 +508,7 @@ static int mchp_coreqspi_exec_op(struct spi_mem *mem, const struct spi_mem_op *o
 			qspi->rxbuf = NULL;
 			qspi->rx_len = 0;
 			qspi->tx_len = op->data.nbytes;
-			mchp_coreqspi_write_op(qspi, true);
+			mchp_coreqspi_write_op(qspi);
 		} else {
 			qspi->txbuf = NULL;
 			qspi->rxbuf = (u8 *)op->data.buf.in;
@@ -515,6 +588,109 @@ static const struct spi_controller_mem_caps mchp_coreqspi_mem_caps = {
 	.per_op_freq = true,
 };

+static int mchp_coreqspi_unprepare_message(struct spi_controller *ctlr, struct spi_message *m)
+{
+	struct mchp_coreqspi *qspi = spi_controller_get_devdata(ctlr);
+
+	/*
+	 * This delay is required for the driver to function correctly,
+	 * but no explanation has been determined for why it is required.
+	 */
+	udelay(750);
+
+	mutex_unlock(&qspi->op_lock);
+
+	return 0;
+}
+
+static int mchp_coreqspi_prepare_message(struct spi_controller *ctlr, struct spi_message *m)
+{
+	struct mchp_coreqspi *qspi = spi_controller_get_devdata(ctlr);
+	struct spi_transfer *t = NULL;
+	u32 control, frames;
+	u32 total_bytes = 0, cmd_bytes = 0, idle_cycles = 0;
+	int ret;
+	bool quad = false, dual = false;
+
+	mutex_lock(&qspi->op_lock);
+	ret = mchp_coreqspi_wait_for_ready(qspi);
+	if (ret) {
+		mutex_unlock(&qspi->op_lock);
+		dev_err(&ctlr->dev, "Timeout waiting on QSPI ready.\n");
+		return ret;
+	}
+
+	ret = mchp_coreqspi_setup_clock(qspi, m->spi, m->spi->max_speed_hz);
+	if (ret) {
+		mutex_unlock(&qspi->op_lock);
+		return ret;
+	}
+
+	control = readl_relaxed(qspi->regs + REG_CONTROL);
+	control &= ~(CONTROL_MODE12_MASK | CONTROL_MODE0);
+	writel_relaxed(control, qspi->regs + REG_CONTROL);
+
+	reinit_completion(&qspi->data_completion);
+
+	list_for_each_entry(t, &m->transfers, transfer_list) {
+		total_bytes += t->len;
+		if (!cmd_bytes && !(t->tx_buf && t->rx_buf))
+			cmd_bytes = t->len;
+		if (!t->rx_buf)
+			cmd_bytes = total_bytes;
+		if (t->tx_nbits == SPI_NBITS_QUAD || t->rx_nbits == SPI_NBITS_QUAD)
+			quad = true;
+		else if (t->tx_nbits == SPI_NBITS_DUAL || t->rx_nbits == SPI_NBITS_DUAL)
+			dual = true;
+	}
+
+	control = readl_relaxed(qspi->regs + REG_CONTROL);
+	if (quad) {
+		control |= (CONTROL_MODE0 | CONTROL_MODE12_EX_RW);
+	} else if (dual) {
+		control &= ~CONTROL_MODE0;
+		control |= CONTROL_MODE12_FULL;
+	} else {
+		control &= ~(CONTROL_MODE12_MASK | CONTROL_MODE0);
+	}
+	writel_relaxed(control, qspi->regs + REG_CONTROL);
+
+	frames = total_bytes & BYTESUPPER_MASK;
+	writel_relaxed(frames, qspi->regs + REG_FRAMESUP);
+	frames = total_bytes & BYTESLOWER_MASK;
+	frames |= cmd_bytes << FRAMES_CMDBYTES_SHIFT;
+	frames |= idle_cycles << FRAMES_IDLE_SHIFT;
+	control = readl_relaxed(qspi->regs + REG_CONTROL);
+	if (control & CONTROL_MODE12_MASK)
+		frames |= (1 << FRAMES_SHIFT);
+
+	frames |= FRAMES_FLAGWORD;
+	writel_relaxed(frames, qspi->regs + REG_FRAMES);
+
+	return 0;
+};
+
+static int mchp_coreqspi_transfer_one(struct spi_controller *ctlr, struct spi_device *spi,
+				      struct spi_transfer *t)
+{
+	struct mchp_coreqspi *qspi = spi_controller_get_devdata(ctlr);
+
+	qspi->tx_len = t->len;
+
+	if (t->tx_buf)
+		qspi->txbuf = (u8 *)t->tx_buf;
+
+	if (!t->rx_buf) {
+		mchp_coreqspi_write_op(qspi);
+	} else {
+		qspi->rxbuf = (u8 *)t->rx_buf;
+		qspi->rx_len = t->len;
+		mchp_coreqspi_write_read_op(qspi);
+	}
+
+	return 0;
+}
+
 static int mchp_coreqspi_probe(struct platform_device *pdev)
 {
 	struct spi_controller *ctlr;
@@ -562,6 +738,12 @@ static int mchp_coreqspi_probe(struct platform_device *pdev)
 	ctlr->mode_bits = SPI_CPOL | SPI_CPHA | SPI_RX_DUAL | SPI_RX_QUAD |
 			  SPI_TX_DUAL | SPI_TX_QUAD;
 	ctlr->dev.of_node = np;
+	ctlr->min_speed_hz = clk_get_rate(qspi->clk) / 30;
+	ctlr->prepare_message = mchp_coreqspi_prepare_message;
+	ctlr->unprepare_message = mchp_coreqspi_unprepare_message;
+	ctlr->transfer_one = mchp_coreqspi_transfer_one;
+	ctlr->num_chipselect = 2;
+	ctlr->use_gpio_descriptors = true;

 	ret = devm_spi_register_controller(&pdev->dev, ctlr);
 	if (ret)
--- a/drivers/spi/spi-mt65xx.c
+++ b/drivers/spi/spi-mt65xx.c
@@ -220,6 +220,14 @@ static const struct mtk_spi_compatible mt6893_compat = {
 	.no_need_unprepare = true,
 };

+static const struct mtk_spi_compatible mt6991_compat = {
+	.need_pad_sel = true,
+	.must_tx = true,
+	.enhance_timing = true,
+	.dma_ext = true,
+	.ipm_design = true,
+};
+
 /*
 * A piece of default chip info unless the platform
 * supplies it.
@@ -245,6 +253,9 @@ static const struct of_device_id mtk_spi_of_match[] = {
 	{ .compatible = "mediatek,mt6765-spi",
 		.data = (void *)&mt6765_compat,
 	},
+	{ .compatible = "mediatek,mt6991-spi",
+		.data = (void *)&mt6991_compat,
+	},
 	{ .compatible = "mediatek,mt7622-spi",
 		.data = (void *)&mt7622_compat,
 	},
--- a/drivers/spi/spi-mtk-nor.c
+++ b/drivers/spi/spi-mtk-nor.c
@@ -918,7 +918,6 @@ static int mtk_nor_probe(struct platform_device *pdev)
 	if (ret < 0)
 		goto err_probe;

-	pm_runtime_mark_last_busy(&pdev->dev);
 	pm_runtime_put_autosuspend(&pdev->dev);

 	dev_info(&pdev->dev, "spi frequency: %d Hz\n", sp->spi_freq);
--- a/drivers/spi/spi-nxp-fspi.c
+++ b/drivers/spi/spi-nxp-fspi.c
@@ -968,7 +968,6 @@ static int nxp_fspi_exec_op(struct spi_mem *mem, const struct spi_mem_op *op)
 	/* Invalidate the data in the AHB buffer. */
 	nxp_fspi_invalid(f);

-	pm_runtime_mark_last_busy(f->dev);
 	pm_runtime_put_autosuspend(f->dev);

 	return err;
--- a/drivers/spi/spi-offload-trigger-adi-util-sigma-delta.c
+++ b/drivers/spi/spi-offload-trigger-adi-util-sigma-delta.c
@@ -0,0 +1,59 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Copyright (C) 2025 Analog Devices Inc.
+ * Copyright (C) 2025 BayLibre, SAS
+ */
+
+#include <linux/clk.h>
+#include <linux/device.h>
+#include <linux/mod_devicetable.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/property.h>
+#include <linux/spi/offload/provider.h>
+
+static bool adi_util_sigma_delta_match(struct spi_offload_trigger *trigger,
+				       enum spi_offload_trigger_type type,
+				       u64 *args, u32 nargs)
+{
+	return type == SPI_OFFLOAD_TRIGGER_DATA_READY && nargs == 0;
+}
+
+static const struct spi_offload_trigger_ops adi_util_sigma_delta_ops = {
+	.match = adi_util_sigma_delta_match,
+};
+
+static int adi_util_sigma_delta_probe(struct platform_device *pdev)
+{
+	struct device *dev = &pdev->dev;
+	struct spi_offload_trigger_info info = {
+		.fwnode = dev_fwnode(dev),
+		.ops = &adi_util_sigma_delta_ops,
+	};
+	struct clk *clk;
+
+	clk = devm_clk_get_enabled(dev, NULL);
+	if (IS_ERR(clk))
+		return dev_err_probe(dev, PTR_ERR(clk), "Failed to get clock\n");
+
+	return devm_spi_offload_trigger_register(dev, &info);
+}
+
+static const struct of_device_id adi_util_sigma_delta_of_match_table[] = {
+	{ .compatible = "adi,util-sigma-delta-spi", },
+	{ }
+};
+MODULE_DEVICE_TABLE(of, adi_util_sigma_delta_of_match_table);
+
+static struct platform_driver adi_util_sigma_delta_driver = {
+	.probe  = adi_util_sigma_delta_probe,
+	.driver = {
+		.name = "adi-util-sigma-delta-spi",
+		.of_match_table = adi_util_sigma_delta_of_match_table,
+	},
+};
+module_platform_driver(adi_util_sigma_delta_driver);
+
+MODULE_AUTHOR("David Lechner <dlechner@baylibre.com>");
+MODULE_DESCRIPTION("ADI Sigma-Delta SPI offload trigger utility driver");
+MODULE_LICENSE("GPL");
--- a/drivers/spi/spi-omap2-mcspi.c
+++ b/drivers/spi/spi-omap2-mcspi.c
@@ -272,7 +272,6 @@ static void omap2_mcspi_set_cs(struct spi_device *spi, bool enable)

 		mcspi_write_chconf0(spi, l);

-		pm_runtime_mark_last_busy(mcspi->dev);
 		pm_runtime_put_autosuspend(mcspi->dev);
 	}
 }
@@ -1102,7 +1101,6 @@ static int omap2_mcspi_setup(struct spi_device *spi)
 	if (ret && initial_setup)
 		omap2_mcspi_cleanup(spi);

-	pm_runtime_mark_last_busy(mcspi->dev);
 	pm_runtime_put_autosuspend(mcspi->dev);

 	return ret;
@@ -1379,7 +1377,6 @@ static int omap2_mcspi_controller_setup(struct omap2_mcspi *mcspi)
 	ctx->wakeupenable = OMAP2_MCSPI_WAKEUPENABLE_WKEN;

 	omap2_mcspi_set_mode(ctlr);
-	pm_runtime_mark_last_busy(mcspi->dev);
 	pm_runtime_put_autosuspend(mcspi->dev);
 	return 0;
 }
--- a/drivers/spi/spi-pci1xxxx.c
+++ b/drivers/spi/spi-pci1xxxx.c
@@ -23,6 +23,7 @@
 #define	SYS_FREQ_DEFAULT		(62500000)

 #define	PCI1XXXX_SPI_MAX_CLOCK_HZ	(30000000)
+#define	PCI1XXXX_SPI_CLK_25MHZ		(25000000)
 #define	PCI1XXXX_SPI_CLK_20MHZ		(20000000)
 #define	PCI1XXXX_SPI_CLK_15MHZ		(15000000)
 #define	PCI1XXXX_SPI_CLK_12MHZ		(12000000)
@@ -96,8 +97,8 @@
 #define SPI_DMA_CH1_DONE_INT		BIT(1)
 #define SPI_DMA_CH0_ABORT_INT		BIT(16)
 #define SPI_DMA_CH1_ABORT_INT		BIT(17)
-#define SPI_DMA_DONE_INT_MASK		(SPI_DMA_CH0_DONE_INT | SPI_DMA_CH1_DONE_INT)
-#define SPI_DMA_ABORT_INT_MASK		(SPI_DMA_CH0_ABORT_INT | SPI_DMA_CH1_ABORT_INT)
+#define SPI_DMA_DONE_INT_MASK(x)	(1 << (x))
+#define SPI_DMA_ABORT_INT_MASK(x)	(1 << (16 + (x)))
 #define DMA_CH_CONTROL_LIE		BIT(3)
 #define DMA_CH_CONTROL_RIE		BIT(4)
 #define DMA_INTR_EN			(DMA_CH_CONTROL_RIE | DMA_CH_CONTROL_LIE)
@@ -131,12 +132,15 @@
 #define SPI_SUSPEND_CONFIG 0x101
 #define SPI_RESUME_CONFIG 0x203

+#define NUM_VEC_PER_INST 3
+
 struct pci1xxxx_spi_internal {
 	u8 hw_inst;
 	u8 clkdiv;
-	int irq;
+	int irq[NUM_VEC_PER_INST];
 	int mode;
 	bool spi_xfer_in_progress;
+	atomic_t dma_completion_count;
 	void *rx_buf;
 	bool dma_aborted_rd;
 	u32 bytes_recvd;
@@ -160,8 +164,10 @@ struct pci1xxxx_spi {
 	u8 dev_rev;
 	void __iomem *reg_base;
 	void __iomem *dma_offset_bar;
-	/* lock to safely access the DMA registers in isr */
-	spinlock_t dma_reg_lock;
+	/* lock to safely access the DMA RD registers in isr */
+	spinlock_t dma_rd_reg_lock;
+	/* lock to safely access the DMA RD registers in isr */
+	spinlock_t dma_wr_reg_lock;
 	bool can_dma;
 	struct pci1xxxx_spi_internal *spi_int[] __counted_by(total_hw_instances);
 };
@@ -192,6 +198,9 @@ static const struct pci_device_id pci1xxxx_spi_pci_id_table[] = {

 MODULE_DEVICE_TABLE(pci, pci1xxxx_spi_pci_id_table);

+static irqreturn_t pci1xxxx_spi_isr_dma_rd(int irq, void *dev);
+static irqreturn_t pci1xxxx_spi_isr_dma_wr(int irq, void *dev);
+
 static int pci1xxxx_set_sys_lock(struct pci1xxxx_spi *par)
 {
 	writel(SPI_SYSLOCK, par->reg_base + SPI_SYSLOCK_REG);
@@ -212,13 +221,16 @@ static void pci1xxxx_release_sys_lock(struct pci1xxxx_spi *par)
 	writel(0x0, par->reg_base + SPI_SYSLOCK_REG);
 }

-static int pci1xxxx_check_spi_can_dma(struct pci1xxxx_spi *spi_bus, int irq)
+static int pci1xxxx_check_spi_can_dma(struct pci1xxxx_spi *spi_bus, int hw_inst, int num_vector)
 {
 	struct pci_dev *pdev = spi_bus->dev;
 	u32 pf_num;
 	u32 regval;
 	int ret;

+	if (num_vector != hw_inst * NUM_VEC_PER_INST)
+		return -EOPNOTSUPP;
+
 	/*
 	 * DEV REV Registers is a system register, HW Syslock bit
 	 * should be acquired before accessing the register
@@ -246,16 +258,6 @@ static int pci1xxxx_check_spi_can_dma(struct pci1xxxx_spi *spi_bus, int irq)
 	if (spi_bus->dev_rev < 0xC0 || pf_num)
 		return -EOPNOTSUPP;

-	/*
-	 * DMA Supported only with MSI Interrupts
-	 * One of the SPI instance's MSI vector address and data
-	 * is used for DMA Interrupt
-	 */
-	if (!irq_get_msi_desc(irq)) {
-		dev_warn(&pdev->dev, "Error MSI Interrupt not supported, will operate in PIO mode\n");
-		return -EOPNOTSUPP;
-	}
-
 	spi_bus->dma_offset_bar = pcim_iomap(pdev, 2, pci_resource_len(pdev, 2));
 	if (!spi_bus->dma_offset_bar) {
 		dev_warn(&pdev->dev, "Error failed to map dma bar, will operate in PIO mode\n");
@@ -272,29 +274,91 @@ static int pci1xxxx_check_spi_can_dma(struct pci1xxxx_spi *spi_bus, int irq)
 	return 0;
 }

-static int pci1xxxx_spi_dma_init(struct pci1xxxx_spi *spi_bus, int irq)
+static void pci1xxxx_spi_dma_config(struct pci1xxxx_spi *spi_bus)
 {
+	struct pci1xxxx_spi_internal *spi_sub_ptr;
+	u8 iter, irq_index;
 	struct msi_msg msi;
+	u32 regval;
+	u16 data;
+
+	irq_index = spi_bus->total_hw_instances;
+	for (iter = 0; iter < spi_bus->total_hw_instances; iter++) {
+		spi_sub_ptr = spi_bus->spi_int[iter];
+		get_cached_msi_msg(spi_sub_ptr->irq[1], &msi);
+		if (iter == 0) {
+			writel(msi.address_hi, spi_bus->dma_offset_bar +
+			       SPI_DMA_INTR_IMWR_WDONE_HIGH);
+			writel(msi.address_hi, spi_bus->dma_offset_bar +
+			       SPI_DMA_INTR_IMWR_WABORT_HIGH);
+			writel(msi.address_hi, spi_bus->dma_offset_bar +
+			       SPI_DMA_INTR_IMWR_RDONE_HIGH);
+			writel(msi.address_hi, spi_bus->dma_offset_bar +
+			       SPI_DMA_INTR_IMWR_RABORT_HIGH);
+			writel(msi.address_lo, spi_bus->dma_offset_bar +
+			       SPI_DMA_INTR_IMWR_WDONE_LOW);
+			writel(msi.address_lo, spi_bus->dma_offset_bar +
+			       SPI_DMA_INTR_IMWR_WABORT_LOW);
+			writel(msi.address_lo, spi_bus->dma_offset_bar +
+			       SPI_DMA_INTR_IMWR_RDONE_LOW);
+			writel(msi.address_lo, spi_bus->dma_offset_bar +
+			       SPI_DMA_INTR_IMWR_RABORT_LOW);
+			writel(0, spi_bus->dma_offset_bar + SPI_DMA_INTR_WR_IMWR_DATA);
+			writel(0, spi_bus->dma_offset_bar + SPI_DMA_INTR_RD_IMWR_DATA);
+		}
+		regval = readl(spi_bus->dma_offset_bar + SPI_DMA_INTR_WR_IMWR_DATA);
+		data = msi.data + irq_index;
+		writel((regval | (data << (iter * 16))), spi_bus->dma_offset_bar +
+		       SPI_DMA_INTR_WR_IMWR_DATA);
+		regval = readl(spi_bus->dma_offset_bar + SPI_DMA_INTR_WR_IMWR_DATA);
+		irq_index++;
+
+		data = msi.data + irq_index;
+		regval = readl(spi_bus->dma_offset_bar + SPI_DMA_INTR_RD_IMWR_DATA);
+		writel(regval | (data << (iter * 16)), spi_bus->dma_offset_bar +
+		       SPI_DMA_INTR_RD_IMWR_DATA);
+		regval = readl(spi_bus->dma_offset_bar + SPI_DMA_INTR_RD_IMWR_DATA);
+		irq_index++;
+	}
+}
+
+static int pci1xxxx_spi_dma_init(struct pci1xxxx_spi *spi_bus, int hw_inst, int num_vector)
+{
+	struct pci1xxxx_spi_internal *spi_sub_ptr;
+	u8 iter, irq_index;
 	int ret;

-	ret = pci1xxxx_check_spi_can_dma(spi_bus, irq);
+	irq_index = hw_inst;
+	ret = pci1xxxx_check_spi_can_dma(spi_bus, hw_inst, num_vector);
 	if (ret)
 		return ret;

-	spin_lock_init(&spi_bus->dma_reg_lock);
-	get_cached_msi_msg(irq, &msi);
+	spin_lock_init(&spi_bus->dma_rd_reg_lock);
+	spin_lock_init(&spi_bus->dma_wr_reg_lock);
 	writel(SPI_DMA_ENGINE_EN, spi_bus->dma_offset_bar + SPI_DMA_GLOBAL_WR_ENGINE_EN);
 	writel(SPI_DMA_ENGINE_EN, spi_bus->dma_offset_bar + SPI_DMA_GLOBAL_RD_ENGINE_EN);
-	writel(msi.address_hi, spi_bus->dma_offset_bar + SPI_DMA_INTR_IMWR_WDONE_HIGH);
-	writel(msi.address_hi, spi_bus->dma_offset_bar + SPI_DMA_INTR_IMWR_WABORT_HIGH);
-	writel(msi.address_hi, spi_bus->dma_offset_bar + SPI_DMA_INTR_IMWR_RDONE_HIGH);
-	writel(msi.address_hi, spi_bus->dma_offset_bar + SPI_DMA_INTR_IMWR_RABORT_HIGH);
-	writel(msi.address_lo, spi_bus->dma_offset_bar + SPI_DMA_INTR_IMWR_WDONE_LOW);
-	writel(msi.address_lo, spi_bus->dma_offset_bar + SPI_DMA_INTR_IMWR_WABORT_LOW);
-	writel(msi.address_lo, spi_bus->dma_offset_bar + SPI_DMA_INTR_IMWR_RDONE_LOW);
-	writel(msi.address_lo, spi_bus->dma_offset_bar + SPI_DMA_INTR_IMWR_RABORT_LOW);
-	writel(msi.data, spi_bus->dma_offset_bar + SPI_DMA_INTR_WR_IMWR_DATA);
-	writel(msi.data, spi_bus->dma_offset_bar + SPI_DMA_INTR_RD_IMWR_DATA);
+
+	for (iter = 0; iter < hw_inst; iter++) {
+		spi_sub_ptr = spi_bus->spi_int[iter];
+		spi_sub_ptr->irq[1] = pci_irq_vector(spi_bus->dev, irq_index);
+		ret = devm_request_irq(&spi_bus->dev->dev, spi_sub_ptr->irq[1],
+				       pci1xxxx_spi_isr_dma_wr, PCI1XXXX_IRQ_FLAGS,
+				       pci_name(spi_bus->dev), spi_sub_ptr);
+		if (ret < 0)
+			return ret;
+
+		irq_index++;
+
+		spi_sub_ptr->irq[2] = pci_irq_vector(spi_bus->dev, irq_index);
+		ret = devm_request_irq(&spi_bus->dev->dev, spi_sub_ptr->irq[2],
+				       pci1xxxx_spi_isr_dma_rd, PCI1XXXX_IRQ_FLAGS,
+				       pci_name(spi_bus->dev), spi_sub_ptr);
+		if (ret < 0)
+			return ret;
+
+		irq_index++;
+	}
+	pci1xxxx_spi_dma_config(spi_bus);
 	dma_set_max_seg_size(&spi_bus->dev->dev, PCI1XXXX_SPI_BUFFER_SIZE);
 	spi_bus->can_dma = true;
 	return 0;
@@ -318,12 +382,14 @@ static void pci1xxxx_spi_set_cs(struct spi_device *spi, bool enable)
 	writel(regval, par->reg_base + SPI_MST_CTL_REG_OFFSET(p->hw_inst));
 }

-static u8 pci1xxxx_get_clock_div(u32 hz)
+static u8 pci1xxxx_get_clock_div(struct pci1xxxx_spi *par, u32 hz)
 {
 	u8 val = 0;

 	if (hz >= PCI1XXXX_SPI_MAX_CLOCK_HZ)
 		val = 2;
+	else if (par->dev_rev >= 0xC0 && hz >= PCI1XXXX_SPI_CLK_25MHZ)
+		val = 1;
 	else if ((hz < PCI1XXXX_SPI_MAX_CLOCK_HZ) && (hz >= PCI1XXXX_SPI_CLK_20MHZ))
 		val = 3;
 	else if ((hz < PCI1XXXX_SPI_CLK_20MHZ) && (hz >= PCI1XXXX_SPI_CLK_15MHZ))
@@ -398,13 +464,14 @@ static void pci1xxxx_spi_setup(struct pci1xxxx_spi *par, u8 hw_inst, u32 mode,
 	writel(regval, par->reg_base + SPI_MST_CTL_REG_OFFSET(hw_inst));
 }

-static void pci1xxxx_start_spi_xfer(struct pci1xxxx_spi_internal *p, u8 hw_inst)
+static void pci1xxxx_start_spi_xfer(struct pci1xxxx_spi_internal *p)
 {
 	u32 regval;

-	regval = readl(p->parent->reg_base + SPI_MST_CTL_REG_OFFSET(hw_inst));
+	atomic_set(&p->dma_completion_count, 0);
+	regval = readl(p->parent->reg_base + SPI_MST_CTL_REG_OFFSET(p->hw_inst));
 	regval |= SPI_MST_CTL_GO;
-	writel(regval, p->parent->reg_base + SPI_MST_CTL_REG_OFFSET(hw_inst));
+	writel(regval, p->parent->reg_base + SPI_MST_CTL_REG_OFFSET(p->hw_inst));
 }

 static int pci1xxxx_spi_transfer_with_io(struct spi_controller *spi_ctlr,
@@ -423,7 +490,7 @@ static int pci1xxxx_spi_transfer_with_io(struct spi_controller *spi_ctlr,

 	p->spi_xfer_in_progress = true;
 	p->bytes_recvd = 0;
-	clkdiv = pci1xxxx_get_clock_div(xfer->speed_hz);
+	clkdiv = pci1xxxx_get_clock_div(par, xfer->speed_hz);
 	tx_buf = xfer->tx_buf;
 	rx_buf = xfer->rx_buf;
 	transfer_len = xfer->len;
@@ -448,7 +515,7 @@ static int pci1xxxx_spi_transfer_with_io(struct spi_controller *spi_ctlr,
 				    &tx_buf[bytes_transfered], len);
 			bytes_transfered += len;
 			pci1xxxx_spi_setup(par, p->hw_inst, spi->mode, clkdiv, len);
-			pci1xxxx_start_spi_xfer(p, p->hw_inst);
+			pci1xxxx_start_spi_xfer(p);

 			/* Wait for DMA_TERM interrupt */
 			result = wait_for_completion_timeout(&p->spi_xfer_done,
@@ -474,7 +541,6 @@ static int pci1xxxx_spi_transfer_with_dma(struct spi_controller *spi_ctlr,
 {
 	struct pci1xxxx_spi_internal *p = spi_controller_get_devdata(spi_ctlr);
 	struct pci1xxxx_spi *par = p->parent;
-	dma_addr_t rx_dma_addr = 0;
 	dma_addr_t tx_dma_addr = 0;
 	int ret = 0;
 	u32 regval;
@@ -483,6 +549,7 @@ static int pci1xxxx_spi_transfer_with_dma(struct spi_controller *spi_ctlr,
 	p->tx_sgl = xfer->tx_sg.sgl;
 	p->rx_sgl = xfer->rx_sg.sgl;
 	p->rx_buf = xfer->rx_buf;
+	atomic_set(&p->dma_completion_count, 1);
 	regval = readl(par->reg_base + SPI_MST_EVENT_REG_OFFSET(p->hw_inst));
 	writel(regval, par->reg_base + SPI_MST_EVENT_REG_OFFSET(p->hw_inst));

@@ -492,20 +559,16 @@ static int pci1xxxx_spi_transfer_with_dma(struct spi_controller *spi_ctlr,
 	}
 	p->xfer = xfer;
 	p->mode = spi->mode;
-	p->clkdiv = pci1xxxx_get_clock_div(xfer->speed_hz);
+	p->clkdiv = pci1xxxx_get_clock_div(par, xfer->speed_hz);
 	p->bytes_recvd = 0;
 	p->rx_buf = xfer->rx_buf;
 	regval = readl(par->reg_base + SPI_MST_EVENT_REG_OFFSET(p->hw_inst));
 	writel(regval, par->reg_base + SPI_MST_EVENT_REG_OFFSET(p->hw_inst));

 	tx_dma_addr = sg_dma_address(p->tx_sgl);
-	rx_dma_addr = sg_dma_address(p->rx_sgl);
 	p->tx_sgl_len = sg_dma_len(p->tx_sgl);
-	p->rx_sgl_len = sg_dma_len(p->rx_sgl);
 	pci1xxxx_spi_setup(par, p->hw_inst, p->mode, p->clkdiv, p->tx_sgl_len);
 	pci1xxxx_spi_setup_dma_to_io(p, (tx_dma_addr), p->tx_sgl_len);
-	if (rx_dma_addr)
-		pci1xxxx_spi_setup_dma_from_io(p, rx_dma_addr, p->rx_sgl_len);
 	writel(p->hw_inst, par->dma_offset_bar + SPI_DMA_RD_DOORBELL_REG);

 	reinit_completion(&p->spi_xfer_done);
@@ -595,83 +658,111 @@ static irqreturn_t pci1xxxx_spi_isr_io(int irq, void *dev)
 	return spi_int_fired;
 }

-static void pci1xxxx_spi_setup_next_dma_transfer(struct pci1xxxx_spi_internal *p)
+static void pci1xxxx_spi_setup_next_dma_to_io_transfer(struct pci1xxxx_spi_internal *p)
 {
 	dma_addr_t tx_dma_addr = 0;
-	dma_addr_t rx_dma_addr = 0;
 	u32 prev_len;

 	p->tx_sgl = sg_next(p->tx_sgl);
-	if (p->rx_sgl)
-		p->rx_sgl = sg_next(p->rx_sgl);
-	if (!p->tx_sgl) {
-		/* Clear xfer_done */
-		complete(&p->spi_xfer_done);
-	} else {
+	if (p->tx_sgl) {
 		tx_dma_addr = sg_dma_address(p->tx_sgl);
 		prev_len = p->tx_sgl_len;
 		p->tx_sgl_len = sg_dma_len(p->tx_sgl);
+		pci1xxxx_spi_setup_dma_to_io(p, tx_dma_addr, p->tx_sgl_len);
+		writel(p->hw_inst, p->parent->dma_offset_bar + SPI_DMA_RD_DOORBELL_REG);
 		if (prev_len != p->tx_sgl_len)
 			pci1xxxx_spi_setup(p->parent,
 					   p->hw_inst, p->mode, p->clkdiv, p->tx_sgl_len);
-		pci1xxxx_spi_setup_dma_to_io(p, tx_dma_addr, p->tx_sgl_len);
-		if (p->rx_sgl) {
-			rx_dma_addr = sg_dma_address(p->rx_sgl);
-			p->rx_sgl_len = sg_dma_len(p->rx_sgl);
-			pci1xxxx_spi_setup_dma_from_io(p, rx_dma_addr, p->rx_sgl_len);
-		}
-		writel(p->hw_inst, p->parent->dma_offset_bar + SPI_DMA_RD_DOORBELL_REG);
 	}
 }

-static irqreturn_t pci1xxxx_spi_isr_dma(int irq, void *dev)
+static void pci1xxxx_spi_setup_next_dma_from_io_transfer(struct pci1xxxx_spi_internal *p)
+{
+	dma_addr_t rx_dma_addr = 0;
+
+	if (p->rx_sgl) {
+		rx_dma_addr = sg_dma_address(p->rx_sgl);
+		p->rx_sgl_len = sg_dma_len(p->rx_sgl);
+		pci1xxxx_spi_setup_dma_from_io(p, rx_dma_addr, p->rx_sgl_len);
+		writel(p->hw_inst, p->parent->dma_offset_bar + SPI_DMA_WR_DOORBELL_REG);
+	}
+}
+
+static irqreturn_t pci1xxxx_spi_isr_dma_rd(int irq, void *dev)
 {
 	struct pci1xxxx_spi_internal *p = dev;
 	irqreturn_t spi_int_fired = IRQ_NONE;
 	unsigned long flags;
 	u32 regval;

-	spin_lock_irqsave(&p->parent->dma_reg_lock, flags);
 	/* Clear the DMA RD INT and start spi xfer*/
 	regval = readl(p->parent->dma_offset_bar + SPI_DMA_INTR_RD_STS);
-	if (regval & SPI_DMA_DONE_INT_MASK) {
-		if (regval & SPI_DMA_CH0_DONE_INT)
-			pci1xxxx_start_spi_xfer(p, SPI0);
-		if (regval & SPI_DMA_CH1_DONE_INT)
-			pci1xxxx_start_spi_xfer(p, SPI1);
-		spi_int_fired = IRQ_HANDLED;
+	if (regval) {
+		if (regval & SPI_DMA_DONE_INT_MASK(p->hw_inst)) {
+			/* Start the SPI transfer only if both DMA read and write are completed */
+			if (atomic_inc_return(&p->dma_completion_count) == 2)
+				pci1xxxx_start_spi_xfer(p);
+			spi_int_fired = IRQ_HANDLED;
+		}
+		if (regval & SPI_DMA_ABORT_INT_MASK(p->hw_inst)) {
+			p->dma_aborted_rd = true;
+			spi_int_fired = IRQ_HANDLED;
+		}
+		spin_lock_irqsave(&p->parent->dma_rd_reg_lock, flags);
+		writel((SPI_DMA_DONE_INT_MASK(p->hw_inst) | SPI_DMA_ABORT_INT_MASK(p->hw_inst)),
+		       p->parent->dma_offset_bar + SPI_DMA_INTR_RD_CLR);
+		spin_unlock_irqrestore(&p->parent->dma_rd_reg_lock, flags);
 	}
-	if (regval & SPI_DMA_ABORT_INT_MASK) {
-		p->dma_aborted_rd = true;
-		spi_int_fired = IRQ_HANDLED;
-	}
-	writel(regval, p->parent->dma_offset_bar + SPI_DMA_INTR_RD_CLR);
+	return spi_int_fired;
+}
+
+static irqreturn_t pci1xxxx_spi_isr_dma_wr(int irq, void *dev)
+{
+	struct pci1xxxx_spi_internal *p = dev;
+	irqreturn_t spi_int_fired = IRQ_NONE;
+	unsigned long flags;
+	u32 regval;

 	/* Clear the DMA WR INT */
 	regval = readl(p->parent->dma_offset_bar + SPI_DMA_INTR_WR_STS);
-	if (regval & SPI_DMA_DONE_INT_MASK) {
-		if (regval & SPI_DMA_CH0_DONE_INT)
-			pci1xxxx_spi_setup_next_dma_transfer(p->parent->spi_int[SPI0]);
+	if (regval) {
+		if (regval & SPI_DMA_DONE_INT_MASK(p->hw_inst)) {
+			spi_int_fired = IRQ_HANDLED;
+			if (sg_is_last(p->rx_sgl)) {
+				complete(&p->spi_xfer_done);
+			} else {
+				p->rx_sgl =  sg_next(p->rx_sgl);
+				if (atomic_inc_return(&p->dma_completion_count) == 2)
+					pci1xxxx_start_spi_xfer(p);
+			}

-		if (regval & SPI_DMA_CH1_DONE_INT)
-			pci1xxxx_spi_setup_next_dma_transfer(p->parent->spi_int[SPI1]);
+		}
+		if (regval & SPI_DMA_ABORT_INT_MASK(p->hw_inst)) {
+			p->dma_aborted_wr = true;
+			spi_int_fired = IRQ_HANDLED;
+		}
+		spin_lock_irqsave(&p->parent->dma_wr_reg_lock, flags);
+		writel((SPI_DMA_DONE_INT_MASK(p->hw_inst) | SPI_DMA_ABORT_INT_MASK(p->hw_inst)),
+		       p->parent->dma_offset_bar + SPI_DMA_INTR_WR_CLR);
+		spin_unlock_irqrestore(&p->parent->dma_wr_reg_lock, flags);
+	}
+	return spi_int_fired;
+}

-		spi_int_fired = IRQ_HANDLED;
-	}
-	if (regval & SPI_DMA_ABORT_INT_MASK) {
-		p->dma_aborted_wr = true;
-		spi_int_fired = IRQ_HANDLED;
-	}
-	writel(regval, p->parent->dma_offset_bar + SPI_DMA_INTR_WR_CLR);
-	spin_unlock_irqrestore(&p->parent->dma_reg_lock, flags);
+static irqreturn_t pci1xxxx_spi_isr_dma(int irq, void *dev)
+{
+	struct pci1xxxx_spi_internal *p = dev;
+	irqreturn_t spi_int_fired = IRQ_NONE;
+	u32 regval;

 	/* Clear the SPI GO_BIT Interrupt */
 	regval = readl(p->parent->reg_base + SPI_MST_EVENT_REG_OFFSET(p->hw_inst));
 	if (regval & SPI_INTR) {
-		writel(p->hw_inst, p->parent->dma_offset_bar + SPI_DMA_WR_DOORBELL_REG);
+		pci1xxxx_spi_setup_next_dma_from_io_transfer(p);
+		pci1xxxx_spi_setup_next_dma_to_io_transfer(p);
 		spi_int_fired = IRQ_HANDLED;
+		writel(regval, p->parent->reg_base + SPI_MST_EVENT_REG_OFFSET(p->hw_inst));
 	}
-	writel(regval, p->parent->reg_base + SPI_MST_EVENT_REG_OFFSET(p->hw_inst));
 	return spi_int_fired;
 }

@@ -761,7 +852,7 @@ static int pci1xxxx_spi_probe(struct pci_dev *pdev, const struct pci_device_id *
 			if (!spi_bus->reg_base)
 				return -EINVAL;

-			num_vector = pci_alloc_irq_vectors(pdev, 1, hw_inst_cnt,
+			num_vector = pci_alloc_irq_vectors(pdev, 1, hw_inst_cnt * NUM_VEC_PER_INST,
 							   PCI_IRQ_INTX | PCI_IRQ_MSI);
 			if (num_vector < 0) {
 				dev_err(&pdev->dev, "Error allocating MSI vectors\n");
@@ -775,27 +866,23 @@ static int pci1xxxx_spi_probe(struct pci_dev *pdev, const struct pci_device_id *
 			regval &= ~SPI_INTR;
 			writel(regval, spi_bus->reg_base +
 			       SPI_MST_EVENT_MASK_REG_OFFSET(spi_sub_ptr->hw_inst));
-			spi_sub_ptr->irq = pci_irq_vector(pdev, 0);
+			spi_sub_ptr->irq[0] = pci_irq_vector(pdev, 0);

 			if (num_vector >= hw_inst_cnt)
-				ret = devm_request_irq(&pdev->dev, spi_sub_ptr->irq,
+				ret = devm_request_irq(&pdev->dev, spi_sub_ptr->irq[0],
 						       pci1xxxx_spi_isr, PCI1XXXX_IRQ_FLAGS,
 						       pci_name(pdev), spi_sub_ptr);
 			else
-				ret = devm_request_irq(&pdev->dev, spi_sub_ptr->irq,
+				ret = devm_request_irq(&pdev->dev, spi_sub_ptr->irq[0],
 						       pci1xxxx_spi_shared_isr,
 						       PCI1XXXX_IRQ_FLAGS | IRQF_SHARED,
 						       pci_name(pdev), spi_bus);
 			if (ret < 0) {
 				dev_err(&pdev->dev, "Unable to request irq : %d",
-					spi_sub_ptr->irq);
+					spi_sub_ptr->irq[0]);
 				return -ENODEV;
 			}

-			ret = pci1xxxx_spi_dma_init(spi_bus, spi_sub_ptr->irq);
-			if (ret && ret != -EOPNOTSUPP)
-				return ret;
-
 			/* This register is only applicable for 1st instance */
 			regval = readl(spi_bus->reg_base + SPI_PCI_CTRL_REG_OFFSET(0));
 			if (!only_sec_inst)
@@ -817,13 +904,13 @@ static int pci1xxxx_spi_probe(struct pci_dev *pdev, const struct pci_device_id *
 			writel(regval, spi_bus->reg_base +
 			       SPI_MST_EVENT_MASK_REG_OFFSET(spi_sub_ptr->hw_inst));
 			if (num_vector >= hw_inst_cnt) {
-				spi_sub_ptr->irq = pci_irq_vector(pdev, iter);
-				ret = devm_request_irq(&pdev->dev, spi_sub_ptr->irq,
+				spi_sub_ptr->irq[0] = pci_irq_vector(pdev, iter);
+				ret = devm_request_irq(&pdev->dev, spi_sub_ptr->irq[0],
 						       pci1xxxx_spi_isr, PCI1XXXX_IRQ_FLAGS,
 						       pci_name(pdev), spi_sub_ptr);
 				if (ret < 0) {
 					dev_err(&pdev->dev, "Unable to request irq : %d",
-						spi_sub_ptr->irq);
+						spi_sub_ptr->irq[0]);
 					return -ENODEV;
 				}
 			}
@@ -846,6 +933,10 @@ static int pci1xxxx_spi_probe(struct pci_dev *pdev, const struct pci_device_id *
 		if (ret)
 			return ret;
 	}
+	ret = pci1xxxx_spi_dma_init(spi_bus, hw_inst_cnt, num_vector);
+	if (ret && ret != -EOPNOTSUPP)
+		return ret;
+
 	pci_set_drvdata(pdev, spi_bus);

 	return 0;
--- a/drivers/spi/spi-qpic-snand.c
+++ b/drivers/spi/spi-qpic-snand.c
@@ -59,12 +59,6 @@
 #define OOB_BUF_SIZE			128
 #define ecceng_to_qspi(eng)		container_of(eng, struct qpic_spi_nand, ecc_eng)

-struct qpic_snand_op {
-	u32 cmd_reg;
-	u32 addr1_reg;
-	u32 addr2_reg;
-};
-
 struct snandc_read_status {
 	__le32 snandc_flash;
 	__le32 snandc_buffer;
@@ -283,9 +277,22 @@ static int qcom_spi_ecc_init_ctx_pipelined(struct nand_device *nand)
 		goto err_free_ecc_cfg;
 	}

-	if (ecc_cfg->strength != 4) {
+	switch (ecc_cfg->strength) {
+	case 4:
+		ecc_cfg->ecc_mode = ECC_MODE_4BIT;
+		ecc_cfg->ecc_bytes_hw = 7;
+		ecc_cfg->spare_bytes = 4;
+		break;
+
+	case 8:
+		ecc_cfg->ecc_mode = ECC_MODE_8BIT;
+		ecc_cfg->ecc_bytes_hw = 13;
+		ecc_cfg->spare_bytes = 2;
+		break;
+
+	default:
 		dev_err(snandc->dev,
-			"only 4 bits ECC strength is supported\n");
+			"only 4 or 8 bits ECC strength is supported\n");
 		ret = -EOPNOTSUPP;
 		goto err_free_ecc_cfg;
 	}
@@ -302,8 +309,6 @@ static int qcom_spi_ecc_init_ctx_pipelined(struct nand_device *nand)
 	nand->ecc.ctx.priv = ecc_cfg;
 	snandc->qspi->mtd = mtd;

-	ecc_cfg->ecc_bytes_hw = 7;
-	ecc_cfg->spare_bytes = 4;
 	ecc_cfg->bbm_size = 1;
 	ecc_cfg->bch_enabled = true;
 	ecc_cfg->bytes = ecc_cfg->ecc_bytes_hw + ecc_cfg->spare_bytes + ecc_cfg->bbm_size;
@@ -365,7 +370,7 @@ static int qcom_spi_ecc_init_ctx_pipelined(struct nand_device *nand)
 			       FIELD_PREP(ECC_SW_RESET, 0) |
 			       FIELD_PREP(ECC_NUM_DATA_BYTES_MASK, ecc_cfg->cw_data) |
 			       FIELD_PREP(ECC_FORCE_CLK_OPEN, 1) |
-			       FIELD_PREP(ECC_MODE_MASK, 0) |
+			       FIELD_PREP(ECC_MODE_MASK, ecc_cfg->ecc_mode) |
 			       FIELD_PREP(ECC_PARITY_SIZE_BYTES_BCH_MASK, ecc_cfg->ecc_bytes_hw);

 	ecc_cfg->ecc_buf_cfg = FIELD_PREP(NUM_STEPS_MASK, 0x203);
@@ -608,10 +613,16 @@ static int qcom_spi_read_last_cw(struct qcom_nand_controller *snandc,

 	bbpos = mtd->writesize - ecc_cfg->cw_size * (num_cw - 1);

-	if (snandc->data_buffer[bbpos] == 0xff)
-		snandc->data_buffer[bbpos + 1] = 0xff;
-	if (snandc->data_buffer[bbpos] != 0xff)
-		snandc->data_buffer[bbpos + 1] = snandc->data_buffer[bbpos];
+	/*
+	 * TODO: The SPINAND code expects two bad block marker bytes
+	 * at the beginning of the OOB area, but the OOB layout used by
+	 * the driver has only one. Duplicate that for now in order to
+	 * avoid certain blocks to be marked as bad.
+	 *
+	 * This can be removed once single-byte bad block marker support
+	 * gets implemented in the SPINAND code.
+	 */
+	snandc->data_buffer[bbpos + 1] = snandc->data_buffer[bbpos];

 	memcpy(op->data.buf.in, snandc->data_buffer + bbpos, op->data.nbytes);

@@ -851,7 +862,7 @@ static int qcom_spi_read_page_ecc(struct qcom_nand_controller *snandc,
 		int data_size, oob_size;

 		if (i == (num_cw - 1)) {
-			data_size = 512 - ((num_cw - 1) << 2);
+			data_size = NANDC_STEP_SIZE - ((num_cw - 1) << 2);
 			oob_size = (num_cw << 2) + ecc_cfg->ecc_bytes_hw +
 				    ecc_cfg->spare_bytes;
 		} else {
@@ -1310,7 +1321,6 @@ static int qcom_spi_write_page(struct qcom_nand_controller *snandc,
 static int qcom_spi_send_cmdaddr(struct qcom_nand_controller *snandc,
 				 const struct spi_mem_op *op)
 {
-	struct qpic_snand_op s_op = {};
 	u32 cmd;
 	int ret, opcode;

@@ -1318,34 +1328,24 @@ static int qcom_spi_send_cmdaddr(struct qcom_nand_controller *snandc,
 	if (ret < 0)
 		return ret;

-	s_op.cmd_reg = cmd;
-	s_op.addr1_reg = op->addr.val;
-	s_op.addr2_reg = 0;
-
 	opcode = op->cmd.opcode;

 	switch (opcode) {
 	case SPINAND_WRITE_EN:
 		return 0;
 	case SPINAND_PROGRAM_EXECUTE:
-		s_op.addr1_reg = op->addr.val << 16;
-		s_op.addr2_reg = op->addr.val >> 16 & 0xff;
-		snandc->qspi->addr1 = cpu_to_le32(s_op.addr1_reg);
-		snandc->qspi->addr2 = cpu_to_le32(s_op.addr2_reg);
+		snandc->qspi->addr1 = cpu_to_le32(op->addr.val << 16);
+		snandc->qspi->addr2 = cpu_to_le32(op->addr.val >> 16 & 0xff);
 		snandc->qspi->cmd = cpu_to_le32(cmd);
 		return qcom_spi_program_execute(snandc, op);
 	case SPINAND_READ:
-		s_op.addr1_reg = (op->addr.val << 16);
-		s_op.addr2_reg = op->addr.val >> 16 & 0xff;
-		snandc->qspi->addr1 = cpu_to_le32(s_op.addr1_reg);
-		snandc->qspi->addr2 = cpu_to_le32(s_op.addr2_reg);
+		snandc->qspi->addr1 = cpu_to_le32(op->addr.val << 16);
+		snandc->qspi->addr2 = cpu_to_le32(op->addr.val >> 16 & 0xff);
 		snandc->qspi->cmd = cpu_to_le32(cmd);
 		return 0;
 	case SPINAND_ERASE:
-		s_op.addr2_reg = (op->addr.val >> 16) & 0xffff;
-		s_op.addr1_reg = op->addr.val;
-		snandc->qspi->addr1 = cpu_to_le32(s_op.addr1_reg << 16);
-		snandc->qspi->addr2 = cpu_to_le32(s_op.addr2_reg);
+		snandc->qspi->addr1 = cpu_to_le32(op->addr.val << 16);
+		snandc->qspi->addr2 = cpu_to_le32(op->addr.val >> 16 & 0xffff);
 		snandc->qspi->cmd = cpu_to_le32(cmd);
 		return qcom_spi_block_erase(snandc);
 	default:
@@ -1357,10 +1357,10 @@ static int qcom_spi_send_cmdaddr(struct qcom_nand_controller *snandc,
 	qcom_clear_read_regs(snandc);
 	qcom_clear_bam_transaction(snandc);

-	snandc->regs->cmd = cpu_to_le32(s_op.cmd_reg);
+	snandc->regs->cmd = cpu_to_le32(cmd);
 	snandc->regs->exec = cpu_to_le32(1);
-	snandc->regs->addr0 = cpu_to_le32(s_op.addr1_reg);
-	snandc->regs->addr1 = cpu_to_le32(s_op.addr2_reg);
+	snandc->regs->addr0 = cpu_to_le32(op->addr.val);
+	snandc->regs->addr1 = cpu_to_le32(0);

 	qcom_write_reg_dma(snandc, &snandc->regs->cmd, NAND_FLASH_CMD, 3, NAND_BAM_NEXT_SGL);
 	qcom_write_reg_dma(snandc, &snandc->regs->exec, NAND_EXEC_CMD, 1, NAND_BAM_NEXT_SGL);
--- a/drivers/spi/spi-rockchip-sfc.c
+++ b/drivers/spi/spi-rockchip-sfc.c
@@ -565,7 +565,6 @@ static int rockchip_sfc_exec_mem_op(struct spi_mem *mem, const struct spi_mem_op

 	ret = rockchip_sfc_xfer_done(sfc, 100000);
 out:
-	pm_runtime_mark_last_busy(sfc->dev);
 	pm_runtime_put_autosuspend(sfc->dev);

 	return ret;
@@ -712,7 +711,6 @@ static int rockchip_sfc_probe(struct platform_device *pdev)
 	if (ret)
 		goto err_register;

-	pm_runtime_mark_last_busy(dev);
 	pm_runtime_put_autosuspend(dev);

 	return 0;
@@ -799,7 +797,6 @@ static int rockchip_sfc_resume(struct device *dev)

 	rockchip_sfc_init(sfc);

-	pm_runtime_mark_last_busy(dev);
 	pm_runtime_put_autosuspend(dev);

 	return 0;
--- a/drivers/spi/spi-rspi.c
+++ b/drivers/spi/spi-rspi.c
@@ -1404,7 +1404,6 @@ static const struct platform_device_id spi_driver_ids[] = {

 MODULE_DEVICE_TABLE(platform, spi_driver_ids);

-#ifdef CONFIG_PM_SLEEP
 static int rspi_suspend(struct device *dev)
 {
 	struct rspi_data *rspi = dev_get_drvdata(dev);
@@ -1419,11 +1418,7 @@ static int rspi_resume(struct device *dev)
 	return spi_controller_resume(rspi->ctlr);
 }

-static SIMPLE_DEV_PM_OPS(rspi_pm_ops, rspi_suspend, rspi_resume);
-#define DEV_PM_OPS	&rspi_pm_ops
-#else
-#define DEV_PM_OPS	NULL
-#endif /* CONFIG_PM_SLEEP */
+static DEFINE_SIMPLE_DEV_PM_OPS(rspi_pm_ops, rspi_suspend, rspi_resume);

 static struct platform_driver rspi_driver = {
 	.probe =	rspi_probe,
@@ -1431,7 +1426,7 @@ static struct platform_driver rspi_driver = {
 	.id_table =	spi_driver_ids,
 	.driver		= {
 		.name = "renesas_spi",
-		.pm = DEV_PM_OPS,
+		.pm = pm_sleep_ptr(&rspi_pm_ops),
 		.of_match_table = of_match_ptr(rspi_of_match),
 	},
 };
--- a/drivers/spi/spi-rzv2h-rspi.c
+++ b/drivers/spi/spi-rzv2h-rspi.c
@@ -0,0 +1,466 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * Renesas RZ/V2H Renesas Serial Peripheral Interface (RSPI)
+ *
+ * Copyright (C) 2025 Renesas Electronics Corporation
+ */
+
+#include <linux/bitfield.h>
+#include <linux/bitops.h>
+#include <linux/bits.h>
+#include <linux/clk.h>
+#include <linux/interrupt.h>
+#include <linux/io.h>
+#include <linux/limits.h>
+#include <linux/log2.h>
+#include <linux/math.h>
+#include <linux/of.h>
+#include <linux/platform_device.h>
+#include <linux/property.h>
+#include <linux/reset.h>
+#include <linux/spi/spi.h>
+#include <linux/wait.h>
+
+/* Registers */
+#define RSPI_SPDR		0x00
+#define RSPI_SPCR		0x08
+#define RSPI_SSLP		0x10
+#define RSPI_SPBR		0x11
+#define RSPI_SPSCR		0x13
+#define RSPI_SPCMD		0x14
+#define RSPI_SPDCR2		0x44
+#define RSPI_SPSR		0x52
+#define RSPI_SPSRC		0x6a
+#define RSPI_SPFCR		0x6c
+
+/* Register SPCR */
+#define RSPI_SPCR_MSTR		BIT(30)
+#define RSPI_SPCR_SPRIE		BIT(17)
+#define RSPI_SPCR_SCKASE	BIT(12)
+#define RSPI_SPCR_SPE		BIT(0)
+
+/* Register SPBR */
+#define RSPI_SPBR_SPR_MIN	0
+#define RSPI_SPBR_SPR_MAX	255
+
+/* Register SPCMD */
+#define RSPI_SPCMD_SSLA		GENMASK(25, 24)
+#define RSPI_SPCMD_SPB		GENMASK(20, 16)
+#define RSPI_SPCMD_LSBF		BIT(12)
+#define RSPI_SPCMD_SSLKP	BIT(7)
+#define RSPI_SPCMD_BRDV		GENMASK(3, 2)
+#define RSPI_SPCMD_CPOL		BIT(1)
+#define RSPI_SPCMD_CPHA		BIT(0)
+
+#define RSPI_SPCMD_BRDV_MIN	0
+#define RSPI_SPCMD_BRDV_MAX	3
+
+/* Register SPDCR2 */
+#define RSPI_SPDCR2_TTRG	GENMASK(11, 8)
+#define RSPI_SPDCR2_RTRG	GENMASK(3, 0)
+#define RSPI_FIFO_SIZE		16
+
+/* Register SPSR */
+#define RSPI_SPSR_SPRF		BIT(15)
+
+/* Register RSPI_SPSRC */
+#define RSPI_SPSRC_CLEAR	0xfd80
+
+#define RSPI_RESET_NUM		2
+#define RSPI_CLK_NUM		3
+
+struct rzv2h_rspi_priv {
+	struct reset_control_bulk_data resets[RSPI_RESET_NUM];
+	struct spi_controller *controller;
+	void __iomem *base;
+	struct clk *tclk;
+	wait_queue_head_t wait;
+	unsigned int bytes_per_word;
+	u32 freq;
+	u16 status;
+};
+
+#define RZV2H_RSPI_TX(func, type)					\
+static inline void rzv2h_rspi_tx_##type(struct rzv2h_rspi_priv *rspi,	\
+					const void *txbuf,		\
+					unsigned int index) {		\
+	type buf = 0;							\
+									\
+	if (txbuf)							\
+		buf = ((type *)txbuf)[index];				\
+									\
+	func(buf, rspi->base + RSPI_SPDR);				\
+}
+
+#define RZV2H_RSPI_RX(func, type)					\
+static inline void rzv2h_rspi_rx_##type(struct rzv2h_rspi_priv *rspi,	\
+					void *rxbuf,			\
+					unsigned int index) {		\
+	type buf = func(rspi->base + RSPI_SPDR);			\
+									\
+	if (rxbuf)							\
+		((type *)rxbuf)[index] = buf;				\
+}
+
+RZV2H_RSPI_TX(writel, u32)
+RZV2H_RSPI_TX(writew, u16)
+RZV2H_RSPI_TX(writeb, u8)
+RZV2H_RSPI_RX(readl, u32)
+RZV2H_RSPI_RX(readw, u16)
+RZV2H_RSPI_RX(readl, u8)
+
+static void rzv2h_rspi_reg_rmw(const struct rzv2h_rspi_priv *rspi,
+				int reg_offs, u32 bit_mask, u32 value)
+{
+	u32 tmp;
+
+	value <<= __ffs(bit_mask);
+	tmp = (readl(rspi->base + reg_offs) & ~bit_mask) | value;
+	writel(tmp, rspi->base + reg_offs);
+}
+
+static inline void rzv2h_rspi_spe_disable(const struct rzv2h_rspi_priv *rspi)
+{
+	rzv2h_rspi_reg_rmw(rspi, RSPI_SPCR, RSPI_SPCR_SPE, 0);
+}
+
+static inline void rzv2h_rspi_spe_enable(const struct rzv2h_rspi_priv *rspi)
+{
+	rzv2h_rspi_reg_rmw(rspi, RSPI_SPCR, RSPI_SPCR_SPE, 1);
+}
+
+static inline void rzv2h_rspi_clear_fifos(const struct rzv2h_rspi_priv *rspi)
+{
+	writeb(1, rspi->base + RSPI_SPFCR);
+}
+
+static inline void rzv2h_rspi_clear_all_irqs(struct rzv2h_rspi_priv *rspi)
+{
+	writew(RSPI_SPSRC_CLEAR, rspi->base + RSPI_SPSRC);
+	rspi->status = 0;
+}
+
+static irqreturn_t rzv2h_rx_irq_handler(int irq, void *data)
+{
+	struct rzv2h_rspi_priv *rspi = data;
+
+	rspi->status = readw(rspi->base + RSPI_SPSR);
+	wake_up(&rspi->wait);
+
+	return IRQ_HANDLED;
+}
+
+static inline int rzv2h_rspi_wait_for_interrupt(struct rzv2h_rspi_priv *rspi,
+						u32 wait_mask)
+{
+	return wait_event_timeout(rspi->wait, (rspi->status & wait_mask),
+				  HZ) == 0 ? -ETIMEDOUT : 0;
+}
+
+static void rzv2h_rspi_send(struct rzv2h_rspi_priv *rspi, const void *txbuf,
+			    unsigned int index)
+{
+	switch (rspi->bytes_per_word) {
+	case 4:
+		rzv2h_rspi_tx_u32(rspi, txbuf, index);
+		break;
+	case 2:
+		rzv2h_rspi_tx_u16(rspi, txbuf, index);
+		break;
+	default:
+		rzv2h_rspi_tx_u8(rspi, txbuf, index);
+	}
+}
+
+static int rzv2h_rspi_receive(struct rzv2h_rspi_priv *rspi, void *rxbuf,
+			      unsigned int index)
+{
+	int ret;
+
+	ret = rzv2h_rspi_wait_for_interrupt(rspi, RSPI_SPSR_SPRF);
+	if (ret)
+		return ret;
+
+	switch (rspi->bytes_per_word) {
+	case 4:
+		rzv2h_rspi_rx_u32(rspi, rxbuf, index);
+		break;
+	case 2:
+		rzv2h_rspi_rx_u16(rspi, rxbuf, index);
+		break;
+	default:
+		rzv2h_rspi_rx_u8(rspi, rxbuf, index);
+	}
+
+	return 0;
+}
+
+static int rzv2h_rspi_transfer_one(struct spi_controller *controller,
+				  struct spi_device *spi,
+				  struct spi_transfer *transfer)
+{
+	struct rzv2h_rspi_priv *rspi = spi_controller_get_devdata(controller);
+	unsigned int words_to_transfer, i;
+	int ret = 0;
+
+	transfer->effective_speed_hz = rspi->freq;
+	words_to_transfer = transfer->len / rspi->bytes_per_word;
+
+	for (i = 0; i < words_to_transfer; i++) {
+		rzv2h_rspi_clear_all_irqs(rspi);
+
+		rzv2h_rspi_send(rspi, transfer->tx_buf, i);
+
+		ret = rzv2h_rspi_receive(rspi, transfer->rx_buf, i);
+		if (ret)
+			break;
+	}
+
+	rzv2h_rspi_clear_all_irqs(rspi);
+
+	if (ret)
+		transfer->error = SPI_TRANS_FAIL_IO;
+
+	spi_finalize_current_transfer(controller);
+
+	return ret;
+}
+
+static inline u32 rzv2h_rspi_calc_bitrate(unsigned long tclk_rate, u8 spr,
+					  u8 brdv)
+{
+	return DIV_ROUND_UP(tclk_rate, (2 * (spr + 1) * (1 << brdv)));
+}
+
+static u32 rzv2h_rspi_setup_clock(struct rzv2h_rspi_priv *rspi, u32 hz)
+{
+	unsigned long tclk_rate;
+	int spr;
+	u8 brdv;
+
+	/*
+	 * From the manual:
+	 * Bit rate = f(RSPI_n_TCLK)/(2*(n+1)*2^(N))
+	 *
+	 * Where:
+	 * * RSPI_n_TCLK is fixed to 200MHz on V2H
+	 * * n = SPR - is RSPI_SPBR.SPR (from 0 to 255)
+	 * * N = BRDV - is RSPI_SPCMD.BRDV (from 0 to 3)
+	 */
+	tclk_rate = clk_get_rate(rspi->tclk);
+	for (brdv = RSPI_SPCMD_BRDV_MIN; brdv <= RSPI_SPCMD_BRDV_MAX; brdv++) {
+		spr = DIV_ROUND_UP(tclk_rate, hz * (1 << (brdv + 1)));
+		spr--;
+		if (spr >= RSPI_SPBR_SPR_MIN && spr <= RSPI_SPBR_SPR_MAX)
+			goto clock_found;
+	}
+
+	return 0;
+
+clock_found:
+	rzv2h_rspi_reg_rmw(rspi, RSPI_SPCMD, RSPI_SPCMD_BRDV, brdv);
+	writeb(spr, rspi->base + RSPI_SPBR);
+
+	return rzv2h_rspi_calc_bitrate(tclk_rate, spr, brdv);
+}
+
+static int rzv2h_rspi_prepare_message(struct spi_controller *ctlr,
+				      struct spi_message *message)
+{
+	struct rzv2h_rspi_priv *rspi = spi_controller_get_devdata(ctlr);
+	const struct spi_device *spi = message->spi;
+	struct spi_transfer *xfer;
+	u32 speed_hz = U32_MAX;
+	u8 bits_per_word;
+	u32 conf32;
+	u16 conf16;
+
+	/* Make sure SPCR.SPE is 0 before amending the configuration */
+	rzv2h_rspi_spe_disable(rspi);
+
+	/* Configure the device to work in "host" mode */
+	conf32 = RSPI_SPCR_MSTR;
+
+	/* Auto-stop function */
+	conf32 |= RSPI_SPCR_SCKASE;
+
+	/* SPI receive buffer full interrupt enable */
+	conf32 |= RSPI_SPCR_SPRIE;
+
+	writel(conf32, rspi->base + RSPI_SPCR);
+
+	/* Use SPCMD0 only */
+	writeb(0x0, rspi->base + RSPI_SPSCR);
+
+	/* Setup mode */
+	conf32 = FIELD_PREP(RSPI_SPCMD_CPOL, !!(spi->mode & SPI_CPOL));
+	conf32 |= FIELD_PREP(RSPI_SPCMD_CPHA, !!(spi->mode & SPI_CPHA));
+	conf32 |= FIELD_PREP(RSPI_SPCMD_LSBF, !!(spi->mode & SPI_LSB_FIRST));
+	conf32 |= FIELD_PREP(RSPI_SPCMD_SSLKP, 1);
+	conf32 |= FIELD_PREP(RSPI_SPCMD_SSLA, spi_get_chipselect(spi, 0));
+	writel(conf32, rspi->base + RSPI_SPCMD);
+	if (spi->mode & SPI_CS_HIGH)
+		writeb(BIT(spi_get_chipselect(spi, 0)), rspi->base + RSPI_SSLP);
+	else
+		writeb(0, rspi->base + RSPI_SSLP);
+
+	/* Setup FIFO thresholds */
+	conf16 = FIELD_PREP(RSPI_SPDCR2_TTRG, RSPI_FIFO_SIZE - 1);
+	conf16 |= FIELD_PREP(RSPI_SPDCR2_RTRG, 0);
+	writew(conf16, rspi->base + RSPI_SPDCR2);
+
+	rzv2h_rspi_clear_fifos(rspi);
+
+	list_for_each_entry(xfer, &message->transfers, transfer_list) {
+		if (!xfer->speed_hz)
+			continue;
+
+		speed_hz = min(xfer->speed_hz, speed_hz);
+		bits_per_word = xfer->bits_per_word;
+	}
+
+	if (speed_hz == U32_MAX)
+		return -EINVAL;
+
+	rspi->bytes_per_word = roundup_pow_of_two(BITS_TO_BYTES(bits_per_word));
+	rzv2h_rspi_reg_rmw(rspi, RSPI_SPCMD, RSPI_SPCMD_SPB, bits_per_word - 1);
+
+	rspi->freq = rzv2h_rspi_setup_clock(rspi, speed_hz);
+	if (!rspi->freq)
+		return -EINVAL;
+
+	rzv2h_rspi_spe_enable(rspi);
+
+	return 0;
+}
+
+static int rzv2h_rspi_unprepare_message(struct spi_controller *ctlr,
+					struct spi_message *message)
+{
+	struct rzv2h_rspi_priv *rspi = spi_controller_get_devdata(ctlr);
+
+	rzv2h_rspi_spe_disable(rspi);
+
+	return 0;
+}
+
+static int rzv2h_rspi_probe(struct platform_device *pdev)
+{
+	struct spi_controller *controller;
+	struct device *dev = &pdev->dev;
+	struct rzv2h_rspi_priv *rspi;
+	struct clk_bulk_data *clks;
+	unsigned long tclk_rate;
+	int irq_rx, ret, i;
+
+	controller = devm_spi_alloc_host(dev, sizeof(*rspi));
+	if (!controller)
+		return -ENOMEM;
+
+	rspi = spi_controller_get_devdata(controller);
+	platform_set_drvdata(pdev, rspi);
+
+	rspi->controller = controller;
+
+	rspi->base = devm_platform_ioremap_resource(pdev, 0);
+	if (IS_ERR(rspi->base))
+		return PTR_ERR(rspi->base);
+
+	ret = devm_clk_bulk_get_all_enabled(dev, &clks);
+	if (ret != RSPI_CLK_NUM)
+		return dev_err_probe(dev, ret >= 0 ? -EINVAL : ret,
+				     "cannot get clocks\n");
+	for (i = 0; i < RSPI_CLK_NUM; i++) {
+		if (!strcmp(clks[i].id, "tclk")) {
+			rspi->tclk = clks[i].clk;
+			break;
+		}
+	}
+
+	if (!rspi->tclk)
+		return dev_err_probe(dev, -EINVAL, "Failed to get tclk\n");
+
+	tclk_rate = clk_get_rate(rspi->tclk);
+
+	rspi->resets[0].id = "presetn";
+	rspi->resets[1].id = "tresetn";
+	ret = devm_reset_control_bulk_get_exclusive(dev, RSPI_RESET_NUM,
+						    rspi->resets);
+	if (ret)
+		return dev_err_probe(dev, ret, "cannot get resets\n");
+
+	irq_rx = platform_get_irq_byname(pdev, "rx");
+	if (irq_rx < 0)
+		return dev_err_probe(dev, irq_rx, "cannot get IRQ 'rx'\n");
+
+	ret = reset_control_bulk_deassert(RSPI_RESET_NUM, rspi->resets);
+	if (ret)
+		return dev_err_probe(dev, ret, "failed to deassert resets\n");
+
+	init_waitqueue_head(&rspi->wait);
+
+	ret = devm_request_irq(dev, irq_rx, rzv2h_rx_irq_handler, 0,
+			       dev_name(dev), rspi);
+	if (ret) {
+		dev_err(dev, "cannot request `rx` IRQ\n");
+		goto quit_resets;
+	}
+
+	controller->mode_bits = SPI_CPHA | SPI_CPOL | SPI_CS_HIGH |
+				SPI_LSB_FIRST;
+	controller->bits_per_word_mask = SPI_BPW_RANGE_MASK(4, 32);
+	controller->prepare_message = rzv2h_rspi_prepare_message;
+	controller->unprepare_message = rzv2h_rspi_unprepare_message;
+	controller->num_chipselect = 4;
+	controller->transfer_one = rzv2h_rspi_transfer_one;
+	controller->min_speed_hz = rzv2h_rspi_calc_bitrate(tclk_rate,
+							   RSPI_SPBR_SPR_MAX,
+							   RSPI_SPCMD_BRDV_MAX);
+	controller->max_speed_hz = rzv2h_rspi_calc_bitrate(tclk_rate,
+							   RSPI_SPBR_SPR_MIN,
+							   RSPI_SPCMD_BRDV_MIN);
+
+	device_set_node(&controller->dev, dev_fwnode(dev));
+
+	ret = spi_register_controller(controller);
+	if (ret) {
+		dev_err(dev, "register controller failed\n");
+		goto quit_resets;
+	}
+
+	return 0;
+
+quit_resets:
+	reset_control_bulk_assert(RSPI_RESET_NUM, rspi->resets);
+
+	return ret;
+}
+
+static void rzv2h_rspi_remove(struct platform_device *pdev)
+{
+	struct rzv2h_rspi_priv *rspi = platform_get_drvdata(pdev);
+
+	spi_unregister_controller(rspi->controller);
+
+	reset_control_bulk_assert(RSPI_RESET_NUM, rspi->resets);
+}
+
+static const struct of_device_id rzv2h_rspi_match[] = {
+	{ .compatible = "renesas,r9a09g057-rspi" },
+	{ /* sentinel */ }
+};
+MODULE_DEVICE_TABLE(of, rzv2h_rspi_match);
+
+static struct platform_driver rzv2h_rspi_drv = {
+	.probe = rzv2h_rspi_probe,
+	.remove = rzv2h_rspi_remove,
+	.driver = {
+		.name = "rzv2h_rspi",
+		.of_match_table = rzv2h_rspi_match,
+	},
+};
+module_platform_driver(rzv2h_rspi_drv);
+
+MODULE_LICENSE("GPL");
+MODULE_AUTHOR("Fabrizio Castro <fabrizio.castro.jz@renesas.com>");
+MODULE_DESCRIPTION("Renesas RZ/V2H(P) Serial Peripheral Interface Driver");
--- a/drivers/spi/spi-s3c64xx.c
+++ b/drivers/spi/spi-s3c64xx.c
@@ -1045,14 +1045,12 @@ static int s3c64xx_spi_setup(struct spi_device *spi)
 		}
 	}

-	pm_runtime_mark_last_busy(&sdd->pdev->dev);
 	pm_runtime_put_autosuspend(&sdd->pdev->dev);
 	s3c64xx_spi_set_cs(spi, false);

 	return 0;

 setup_exit:
-	pm_runtime_mark_last_busy(&sdd->pdev->dev);
 	pm_runtime_put_autosuspend(&sdd->pdev->dev);
 	/* setup() returns with device de-selected */
 	s3c64xx_spi_set_cs(spi, false);
@@ -1384,7 +1382,6 @@ static int s3c64xx_spi_probe(struct platform_device *pdev)
 	dev_dbg(&pdev->dev, "\tIOmem=[%pR]\tFIFO %dbytes\n",
 		mem_res, sdd->fifo_depth);

-	pm_runtime_mark_last_busy(&pdev->dev);
 	pm_runtime_put_autosuspend(&pdev->dev);

 	return 0;
--- a/drivers/spi/spi-sg2044-nor.c
+++ b/drivers/spi/spi-sg2044-nor.c
@@ -84,12 +84,18 @@

 #define SPIFMC_MAX_READ_SIZE			0x10000

+struct sg204x_spifmc_chip_info {
+	bool has_opt_reg;
+	u32 rd_fifo_int_trigger_level;
+};
+
 struct sg2044_spifmc {
 	struct spi_controller *ctrl;
 	void __iomem *io_base;
 	struct device *dev;
 	struct mutex lock;
 	struct clk *clk;
+	const struct sg204x_spifmc_chip_info *chip_info;
 };

 static int sg2044_spifmc_wait_int(struct sg2044_spifmc *spifmc, u8 int_type)
@@ -139,7 +145,7 @@ static ssize_t sg2044_spifmc_read_64k(struct sg2044_spifmc *spifmc,

 	reg = sg2044_spifmc_init_reg(spifmc);
 	reg |= (op->addr.nbytes + op->dummy.nbytes) << SPIFMC_TRAN_CSR_ADDR_BYTES_SHIFT;
-	reg |= SPIFMC_TRAN_CSR_FIFO_TRG_LVL_8_BYTE;
+	reg |= spifmc->chip_info->rd_fifo_int_trigger_level;
 	reg |= SPIFMC_TRAN_CSR_WITH_CMD;
 	reg |= SPIFMC_TRAN_CSR_TRAN_MODE_RX;

@@ -335,7 +341,8 @@ static ssize_t sg2044_spifmc_trans_reg(struct sg2044_spifmc *spifmc,
 		reg |= SPIFMC_TRAN_CSR_TRAN_MODE_RX;
 		reg |= SPIFMC_TRAN_CSR_TRAN_MODE_TX;

-		writel(SPIFMC_OPT_DISABLE_FIFO_FLUSH, spifmc->io_base + SPIFMC_OPT);
+		if (spifmc->chip_info->has_opt_reg)
+			writel(SPIFMC_OPT_DISABLE_FIFO_FLUSH, spifmc->io_base + SPIFMC_OPT);
 	} else {
 		/*
 		 * If write values to the Status Register,
@@ -457,6 +464,11 @@ static int sg2044_spifmc_probe(struct platform_device *pdev)
 	ret = devm_mutex_init(dev, &spifmc->lock);
 	if (ret)
 		return ret;
+	spifmc->chip_info = device_get_match_data(&pdev->dev);
+	if (!spifmc->chip_info) {
+		dev_err(&pdev->dev, "Failed to get specific chip info\n");
+		return -EINVAL;
+	}

 	sg2044_spifmc_init(spifmc);
 	sg2044_spifmc_init_reg(spifmc);
@@ -468,8 +480,19 @@ static int sg2044_spifmc_probe(struct platform_device *pdev)
 	return 0;
 }

+static const struct sg204x_spifmc_chip_info sg2044_chip_info = {
+	.has_opt_reg = true,
+	.rd_fifo_int_trigger_level = SPIFMC_TRAN_CSR_FIFO_TRG_LVL_8_BYTE,
+};
+
+static const struct sg204x_spifmc_chip_info sg2042_chip_info = {
+	.has_opt_reg = false,
+	.rd_fifo_int_trigger_level = SPIFMC_TRAN_CSR_FIFO_TRG_LVL_1_BYTE,
+};
+
 static const struct of_device_id sg2044_spifmc_match[] = {
-	{ .compatible = "sophgo,sg2044-spifmc-nor" },
+	{ .compatible = "sophgo,sg2044-spifmc-nor", .data = &sg2044_chip_info },
+	{ .compatible = "sophgo,sg2042-spifmc-nor", .data = &sg2042_chip_info },
 	{ /* sentinel */ }
 };
 MODULE_DEVICE_TABLE(of, sg2044_spifmc_match);
--- a/drivers/spi/spi-sh-msiof.c
+++ b/drivers/spi/spi-sh-msiof.c
@@ -1320,7 +1320,6 @@ static const struct platform_device_id spi_driver_ids[] = {
 };
 MODULE_DEVICE_TABLE(platform, spi_driver_ids);

-#ifdef CONFIG_PM_SLEEP
 static int sh_msiof_spi_suspend(struct device *dev)
 {
 	struct sh_msiof_spi_priv *p = dev_get_drvdata(dev);
@@ -1335,12 +1334,8 @@ static int sh_msiof_spi_resume(struct device *dev)
 	return spi_controller_resume(p->ctlr);
 }

-static SIMPLE_DEV_PM_OPS(sh_msiof_spi_pm_ops, sh_msiof_spi_suspend,
-			 sh_msiof_spi_resume);
-#define DEV_PM_OPS	(&sh_msiof_spi_pm_ops)
-#else
-#define DEV_PM_OPS	NULL
-#endif /* CONFIG_PM_SLEEP */
+static DEFINE_SIMPLE_DEV_PM_OPS(sh_msiof_spi_pm_ops, sh_msiof_spi_suspend,
+				sh_msiof_spi_resume);

 static struct platform_driver sh_msiof_spi_drv = {
 	.probe		= sh_msiof_spi_probe,
@@ -1348,7 +1343,7 @@ static struct platform_driver sh_msiof_spi_drv = {
 	.id_table	= spi_driver_ids,
 	.driver		= {
 		.name		= "spi_sh_msiof",
-		.pm		= DEV_PM_OPS,
+		.pm		= pm_sleep_ptr(&sh_msiof_spi_pm_ops),
 		.of_match_table = of_match_ptr(sh_msiof_match),
 	},
 };
--- a/drivers/spi/spi-sprd.c
+++ b/drivers/spi/spi-sprd.c
@@ -982,7 +982,6 @@ static int sprd_spi_probe(struct platform_device *pdev)
 	if (ret)
 		goto err_rpm_put;

-	pm_runtime_mark_last_busy(&pdev->dev);
 	pm_runtime_put_autosuspend(&pdev->dev);

 	return 0;
--- a/drivers/spi/spi-st-ssc4.c
+++ b/drivers/spi/spi-st-ssc4.c
@@ -378,8 +378,7 @@ static void spi_st_remove(struct platform_device *pdev)
 	pinctrl_pm_select_sleep_state(&pdev->dev);
 }

-#ifdef CONFIG_PM
-static int spi_st_runtime_suspend(struct device *dev)
+static int __maybe_unused spi_st_runtime_suspend(struct device *dev)
 {
 	struct spi_controller *host = dev_get_drvdata(dev);
 	struct spi_st *spi_st = spi_controller_get_devdata(host);
@@ -392,7 +391,7 @@ static int spi_st_runtime_suspend(struct device *dev)
 	return 0;
 }

-static int spi_st_runtime_resume(struct device *dev)
+static int __maybe_unused spi_st_runtime_resume(struct device *dev)
 {
 	struct spi_controller *host = dev_get_drvdata(dev);
 	struct spi_st *spi_st = spi_controller_get_devdata(host);
@@ -403,10 +402,8 @@ static int spi_st_runtime_resume(struct device *dev)

 	return ret;
 }
-#endif

-#ifdef CONFIG_PM_SLEEP
-static int spi_st_suspend(struct device *dev)
+static int __maybe_unused spi_st_suspend(struct device *dev)
 {
 	struct spi_controller *host = dev_get_drvdata(dev);
 	int ret;
@@ -418,7 +415,7 @@ static int spi_st_suspend(struct device *dev)
 	return pm_runtime_force_suspend(dev);
 }

-static int spi_st_resume(struct device *dev)
+static int __maybe_unused spi_st_resume(struct device *dev)
 {
 	struct spi_controller *host = dev_get_drvdata(dev);
 	int ret;
@@ -429,7 +426,6 @@ static int spi_st_resume(struct device *dev)

 	return pm_runtime_force_resume(dev);
 }
-#endif

 static const struct dev_pm_ops spi_st_pm = {
 	SET_SYSTEM_SLEEP_PM_OPS(spi_st_suspend, spi_st_resume)
@@ -445,7 +441,7 @@ MODULE_DEVICE_TABLE(of, stm_spi_match);
 static struct platform_driver spi_st_driver = {
 	.driver = {
 		.name = "spi-st",
-		.pm = &spi_st_pm,
+		.pm = pm_sleep_ptr(&spi_st_pm),
 		.of_match_table = of_match_ptr(stm_spi_match),
 	},
 	.probe = spi_st_probe,
--- a/drivers/spi/spi-stm32-ospi.c
+++ b/drivers/spi/spi-stm32-ospi.c
@@ -547,7 +547,6 @@ static int stm32_ospi_poll_status(struct spi_mem *mem,
 	ret = stm32_ospi_send(mem->spi, op);
 	mutex_unlock(&ospi->lock);

-	pm_runtime_mark_last_busy(ospi->dev);
 	pm_runtime_put_autosuspend(ospi->dev);

 	return ret;
@@ -571,7 +570,6 @@ static int stm32_ospi_exec_op(struct spi_mem *mem, const struct spi_mem_op *op)
 	ret = stm32_ospi_send(mem->spi, op);
 	mutex_unlock(&ospi->lock);

-	pm_runtime_mark_last_busy(ospi->dev);
 	pm_runtime_put_autosuspend(ospi->dev);

 	return ret;
@@ -628,7 +626,6 @@ static ssize_t stm32_ospi_dirmap_read(struct spi_mem_dirmap_desc *desc,
 	ret = stm32_ospi_send(desc->mem->spi, &op);
 	mutex_unlock(&ospi->lock);

-	pm_runtime_mark_last_busy(ospi->dev);
 	pm_runtime_put_autosuspend(ospi->dev);

 	return ret ?: len;
@@ -713,7 +710,6 @@ end_of_transfer:
 	msg->status = ret;
 	spi_finalize_current_message(ctrl);

-	pm_runtime_mark_last_busy(ospi->dev);
 	pm_runtime_put_autosuspend(ospi->dev);

 	return ret;
@@ -750,7 +746,6 @@ static int stm32_ospi_setup(struct spi_device *spi)

 	mutex_unlock(&ospi->lock);

-	pm_runtime_mark_last_busy(ospi->dev);
 	pm_runtime_put_autosuspend(ospi->dev);

 	return 0;
@@ -771,9 +766,7 @@ static int stm32_ospi_get_resources(struct platform_device *pdev)
 {
 	struct device *dev = &pdev->dev;
 	struct stm32_ospi *ospi = platform_get_drvdata(pdev);
-	struct resource *res;
-	struct reserved_mem *rmem = NULL;
-	struct device_node *node;
+	struct resource *res, _res;
 	int ret;

 	ospi->regs_base = devm_platform_get_and_ioremap_resource(pdev, 0, &res);
@@ -825,18 +818,14 @@ static int stm32_ospi_get_resources(struct platform_device *pdev)
 			goto err_dma;
 	}

-	node = of_parse_phandle(dev->of_node, "memory-region", 0);
-	if (node)
-		rmem = of_reserved_mem_lookup(node);
-	of_node_put(node);
-
-	if (rmem) {
-		ospi->mm_size = rmem->size;
-		ospi->mm_base = devm_ioremap(dev, rmem->base, rmem->size);
-		if (!ospi->mm_base) {
-			dev_err(dev, "unable to map memory region: %pa+%pa\n",
-				&rmem->base, &rmem->size);
-			ret = -ENOMEM;
+	res = &_res;
+	ret = of_reserved_mem_region_to_resource(dev->of_node, 0, res);
+	if (!ret) {
+		ospi->mm_size = resource_size(res);
+		ospi->mm_base = devm_ioremap_resource(dev, res);
+		if (IS_ERR(ospi->mm_base)) {
+			dev_err(dev, "unable to map memory region: %pR\n", res);
+			ret = PTR_ERR(ospi->mm_base);
 			goto err_dma;
 		}

@@ -953,7 +942,6 @@ static int stm32_ospi_probe(struct platform_device *pdev)
 		goto err_pm_resume;
 	}

-	pm_runtime_mark_last_busy(ospi->dev);
 	pm_runtime_put_autosuspend(ospi->dev);

 	return 0;
@@ -1032,7 +1020,6 @@ static int __maybe_unused stm32_ospi_resume(struct device *dev)

 	writel_relaxed(ospi->cr_reg, regs_base + OSPI_CR);
 	writel_relaxed(ospi->dcr_reg, regs_base + OSPI_DCR1);
-	pm_runtime_mark_last_busy(ospi->dev);
 	pm_runtime_put_autosuspend(ospi->dev);

 	return 0;
--- a/drivers/spi/spi-stm32-qspi.c
+++ b/drivers/spi/spi-stm32-qspi.c
@@ -463,7 +463,6 @@ static int stm32_qspi_poll_status(struct spi_mem *mem, const struct spi_mem_op *
 	ret = stm32_qspi_send(mem->spi, op);
 	mutex_unlock(&qspi->lock);

-	pm_runtime_mark_last_busy(qspi->dev);
 	pm_runtime_put_autosuspend(qspi->dev);

 	return ret;
@@ -487,7 +486,6 @@ static int stm32_qspi_exec_op(struct spi_mem *mem, const struct spi_mem_op *op)
 	ret = stm32_qspi_send(mem->spi, op);
 	mutex_unlock(&qspi->lock);

-	pm_runtime_mark_last_busy(qspi->dev);
 	pm_runtime_put_autosuspend(qspi->dev);

 	return ret;
@@ -543,7 +541,6 @@ static ssize_t stm32_qspi_dirmap_read(struct spi_mem_dirmap_desc *desc,
 	ret = stm32_qspi_send(desc->mem->spi, &op);
 	mutex_unlock(&qspi->lock);

-	pm_runtime_mark_last_busy(qspi->dev);
 	pm_runtime_put_autosuspend(qspi->dev);

 	return ret ?: len;
@@ -627,7 +624,6 @@ end_of_transfer:
 	msg->status = ret;
 	spi_finalize_current_message(ctrl);

-	pm_runtime_mark_last_busy(qspi->dev);
 	pm_runtime_put_autosuspend(qspi->dev);

 	return ret;
@@ -684,7 +680,6 @@ static int stm32_qspi_setup(struct spi_device *spi)
 	writel_relaxed(qspi->dcr_reg, qspi->io_base + QSPI_DCR);
 	mutex_unlock(&qspi->lock);

-	pm_runtime_mark_last_busy(qspi->dev);
 	pm_runtime_put_autosuspend(qspi->dev);

 	return 0;
@@ -858,7 +853,6 @@ static int stm32_qspi_probe(struct platform_device *pdev)
 	if (ret)
 		goto err_pm_runtime_free;

-	pm_runtime_mark_last_busy(dev);
 	pm_runtime_put_autosuspend(dev);

 	return 0;
@@ -938,7 +932,6 @@ static int __maybe_unused stm32_qspi_resume(struct device *dev)
 	writel_relaxed(qspi->cr_reg, qspi->io_base + QSPI_CR);
 	writel_relaxed(qspi->dcr_reg, qspi->io_base + QSPI_DCR);

-	pm_runtime_mark_last_busy(dev);
 	pm_runtime_put_autosuspend(dev);

 	return 0;
--- a/drivers/spi/spi-stm32.c
+++ b/drivers/spi/spi-stm32.c
@@ -9,7 +9,9 @@
 #include <linux/debugfs.h>
 #include <linux/clk.h>
 #include <linux/delay.h>
+#include <linux/dma-mapping.h>
 #include <linux/dmaengine.h>
+#include <linux/genalloc.h>
 #include <linux/interrupt.h>
 #include <linux/iopoll.h>
 #include <linux/module.h>
@@ -154,6 +156,9 @@
 /* STM32H7_SPI_I2SCFGR bit fields */
 #define STM32H7_SPI_I2SCFGR_I2SMOD	BIT(0)

+/* STM32MP25_SPICFG2 bit fields */
+#define STM32MP25_SPI_CFG2_RDIOM	BIT(13)
+
 /* STM32MP25 SPI registers bit fields */
 #define STM32MP25_SPI_HWCFGR1			0x3F0

@@ -222,6 +227,7 @@ struct stm32_spi_reg {
 * @rx: SPI RX data register
 * @tx: SPI TX data register
 * @fullcfg: SPI full or limited feature set register
+ * @rdy_en: SPI ready feature register
 */
 struct stm32_spi_regspec {
 	const struct stm32_spi_reg en;
@@ -235,6 +241,7 @@ struct stm32_spi_regspec {
 	const struct stm32_spi_reg rx;
 	const struct stm32_spi_reg tx;
 	const struct stm32_spi_reg fullcfg;
+	const struct stm32_spi_reg rdy_en;
 };

 struct stm32_spi;
@@ -276,7 +283,7 @@ struct stm32_spi_cfg {
 	int (*config)(struct stm32_spi *spi);
 	void (*set_bpw)(struct stm32_spi *spi);
 	int (*set_mode)(struct stm32_spi *spi, unsigned int comm_type);
-	void (*set_data_idleness)(struct stm32_spi *spi, u32 length);
+	void (*set_data_idleness)(struct stm32_spi *spi, struct spi_transfer *xfer);
 	int (*set_number_of_data)(struct stm32_spi *spi, u32 length);
 	void (*write_tx)(struct stm32_spi *spi);
 	void (*read_rx)(struct stm32_spi *spi);
@@ -323,6 +330,11 @@ struct stm32_spi_cfg {
 * @dma_rx: dma channel for RX transfer
 * @phys_addr: SPI registers physical base address
 * @device_mode: the controller is configured as SPI device
+ * @sram_pool: SRAM pool for DMA transfers
+ * @sram_rx_buf_size: size of SRAM buffer for RX transfer
+ * @sram_rx_buf: SRAM buffer for RX transfer
+ * @sram_dma_rx_buf: SRAM buffer physical address for RX transfer
+ * @mdma_rx: MDMA channel for RX transfer
 */
 struct stm32_spi {
 	struct device *dev;
@@ -357,6 +369,12 @@ struct stm32_spi {
 	dma_addr_t phys_addr;

 	bool device_mode;
+
+	struct gen_pool *sram_pool;
+	size_t sram_rx_buf_size;
+	void *sram_rx_buf;
+	dma_addr_t sram_dma_rx_buf;
+	struct dma_chan *mdma_rx;
 };

 static const struct stm32_spi_regspec stm32fx_spi_regspec = {
@@ -415,6 +433,8 @@ static const struct stm32_spi_regspec stm32mp25_spi_regspec = {
 	.tx = { STM32H7_SPI_TXDR },

 	.fullcfg = { STM32MP25_SPI_HWCFGR1, STM32MP25_SPI_HWCFGR1_FULLCFG },
+
+	.rdy_en = { STM32H7_SPI_CFG2, STM32MP25_SPI_CFG2_RDIOM },
 };

 static inline void stm32_spi_set_bits(struct stm32_spi *spi,
@@ -878,8 +898,11 @@ static void stm32h7_spi_disable(struct stm32_spi *spi)

 	if (spi->cur_usedma && spi->dma_tx)
 		dmaengine_terminate_async(spi->dma_tx);
-	if (spi->cur_usedma && spi->dma_rx)
+	if (spi->cur_usedma && spi->dma_rx) {
 		dmaengine_terminate_async(spi->dma_rx);
+		if (spi->mdma_rx)
+			dmaengine_terminate_async(spi->mdma_rx);
+	}

 	stm32_spi_clr_bits(spi, STM32H7_SPI_CR1, STM32H7_SPI_CR1_SPE);

@@ -1091,10 +1114,13 @@ static irqreturn_t stm32h7_spi_irq_thread(int irq, void *dev_id)
 	}

 	if (sr & STM32H7_SPI_SR_EOT) {
+		dev_dbg(spi->dev, "End of transfer\n");
 		if (!spi->cur_usedma && (spi->rx_buf && (spi->rx_len > 0)))
 			stm32h7_spi_read_rxfifo(spi);
 		if (!spi->cur_usedma ||
-		    (spi->cur_comm == SPI_SIMPLEX_TX || spi->cur_comm == SPI_3WIRE_TX))
+		    (spi->cur_comm == SPI_SIMPLEX_TX || spi->cur_comm == SPI_3WIRE_TX) ||
+		    (spi->mdma_rx && (spi->cur_comm == SPI_SIMPLEX_RX ||
+		     spi->cur_comm == SPI_FULL_DUPLEX)))
 			end = true;
 	}

@@ -1111,6 +1137,11 @@ static irqreturn_t stm32h7_spi_irq_thread(int irq, void *dev_id)
 	spin_unlock_irqrestore(&spi->lock, flags);

 	if (end) {
+		if (spi->cur_usedma && spi->mdma_rx) {
+			dmaengine_pause(spi->dma_rx);
+			/* Wait for callback */
+			return IRQ_HANDLED;
+		}
 		stm32h7_spi_disable(spi);
 		spi_finalize_current_transfer(ctrl);
 	}
@@ -1172,15 +1203,21 @@ static int stm32_spi_prepare_msg(struct spi_controller *ctrl,
 	else
 		clrb |= spi->cfg->regs->cs_high.mask;

-	dev_dbg(spi->dev, "cpol=%d cpha=%d lsb_first=%d cs_high=%d\n",
+	if (spi_dev->mode & SPI_READY)
+		setb |= spi->cfg->regs->rdy_en.mask;
+	else
+		clrb |= spi->cfg->regs->rdy_en.mask;
+
+	dev_dbg(spi->dev, "cpol=%d cpha=%d lsb_first=%d cs_high=%d rdy=%d\n",
 		!!(spi_dev->mode & SPI_CPOL),
 		!!(spi_dev->mode & SPI_CPHA),
 		!!(spi_dev->mode & SPI_LSB_FIRST),
-		!!(spi_dev->mode & SPI_CS_HIGH));
+		!!(spi_dev->mode & SPI_CS_HIGH),
+		!!(spi_dev->mode & SPI_READY));

 	spin_lock_irqsave(&spi->lock, flags);

-	/* CPOL, CPHA and LSB FIRST bits have common register */
+	/* CPOL, CPHA, LSB FIRST, CS_HIGH and RDY_EN bits have common register */
 	if (clrb || setb)
 		writel_relaxed(
 			(readl_relaxed(spi->base + spi->cfg->regs->cpol.reg) &
@@ -1410,6 +1447,8 @@ static void stm32h7_spi_transfer_one_dma_start(struct stm32_spi *spi)
 	/* Enable the interrupts */
 	if (spi->cur_comm == SPI_SIMPLEX_TX || spi->cur_comm == SPI_3WIRE_TX)
 		ier |= STM32H7_SPI_IER_EOTIE | STM32H7_SPI_IER_TXTFIE;
+	if (spi->mdma_rx && (spi->cur_comm == SPI_SIMPLEX_RX || spi->cur_comm == SPI_FULL_DUPLEX))
+		ier |= STM32H7_SPI_IER_EOTIE;

 	stm32_spi_set_bits(spi, STM32H7_SPI_IER, ier);

@@ -1419,6 +1458,121 @@ static void stm32h7_spi_transfer_one_dma_start(struct stm32_spi *spi)
 		stm32_spi_set_bits(spi, STM32H7_SPI_CR1, STM32H7_SPI_CR1_CSTART);
 }

+/**
+ * stm32_spi_prepare_rx_dma_mdma_chaining - Prepare RX DMA and MDMA chaining
+ * @spi: pointer to the spi controller data structure
+ * @xfer: pointer to the spi transfer
+ * @rx_dma_conf: pointer to the DMA configuration for RX channel
+ * @rx_dma_desc: pointer to the RX DMA descriptor
+ * @rx_mdma_desc: pointer to the RX MDMA descriptor
+ *
+ * It must return 0 if the chaining is possible or an error code if not.
+ */
+static int stm32_spi_prepare_rx_dma_mdma_chaining(struct stm32_spi *spi,
+						  struct spi_transfer *xfer,
+						  struct dma_slave_config *rx_dma_conf,
+						  struct dma_async_tx_descriptor **rx_dma_desc,
+						  struct dma_async_tx_descriptor **rx_mdma_desc)
+{
+	struct dma_async_tx_descriptor *_mdma_desc = *rx_mdma_desc;
+	struct dma_async_tx_descriptor *_dma_desc = *rx_dma_desc;
+	struct dma_slave_config rx_mdma_conf = {0};
+	u32 sram_period, nents = 0, spi_s_len;
+	struct sg_table dma_sgt, mdma_sgt;
+	struct scatterlist *spi_s, *s;
+	dma_addr_t dma_buf;
+	int i, ret;
+
+	sram_period = spi->sram_rx_buf_size / 2;
+
+	/* Configure MDMA RX channel */
+	rx_mdma_conf.direction = rx_dma_conf->direction;
+	rx_mdma_conf.src_addr = spi->sram_dma_rx_buf;
+	rx_mdma_conf.peripheral_config = rx_dma_conf->peripheral_config;
+	rx_mdma_conf.peripheral_size = rx_dma_conf->peripheral_size;
+	dmaengine_slave_config(spi->mdma_rx, &rx_mdma_conf);
+
+	/* Count the number of entries needed */
+	for_each_sg(xfer->rx_sg.sgl, spi_s, xfer->rx_sg.nents, i)
+		if (sg_dma_len(spi_s) > sram_period)
+			nents += DIV_ROUND_UP(sg_dma_len(spi_s), sram_period);
+		else
+			nents++;
+
+	/* Prepare DMA slave_sg DBM transfer DEV_TO_MEM (RX>MEM=SRAM) */
+	ret = sg_alloc_table(&dma_sgt, nents, GFP_ATOMIC);
+	if (ret)
+		return ret;
+
+	spi_s = xfer->rx_sg.sgl;
+	spi_s_len = sg_dma_len(spi_s);
+	dma_buf = spi->sram_dma_rx_buf;
+	for_each_sg(dma_sgt.sgl, s, dma_sgt.nents, i) {
+		size_t bytes = min_t(size_t, spi_s_len, sram_period);
+
+		sg_dma_len(s) = bytes;
+		sg_dma_address(s) = dma_buf;
+		spi_s_len -= bytes;
+
+		if (!spi_s_len && sg_next(spi_s)) {
+			spi_s = sg_next(spi_s);
+			spi_s_len = sg_dma_len(spi_s);
+			dma_buf = spi->sram_dma_rx_buf;
+		} else { /* DMA configured in DBM: it will swap between the SRAM periods */
+			if (i & 1)
+				dma_buf += sram_period;
+			else
+				dma_buf = spi->sram_dma_rx_buf;
+		}
+	}
+
+	_dma_desc = dmaengine_prep_slave_sg(spi->dma_rx, dma_sgt.sgl,
+					    dma_sgt.nents, rx_dma_conf->direction,
+					    DMA_PREP_INTERRUPT);
+	sg_free_table(&dma_sgt);
+
+	if (!_dma_desc)
+		return -EINVAL;
+
+	/* Prepare MDMA slave_sg transfer MEM_TO_MEM (SRAM>DDR) */
+	ret = sg_alloc_table(&mdma_sgt, nents, GFP_ATOMIC);
+	if (ret) {
+		_dma_desc = NULL;
+		return ret;
+	}
+
+	spi_s = xfer->rx_sg.sgl;
+	spi_s_len = sg_dma_len(spi_s);
+	dma_buf = sg_dma_address(spi_s);
+	for_each_sg(mdma_sgt.sgl, s, mdma_sgt.nents, i) {
+		size_t bytes = min_t(size_t, spi_s_len, sram_period);
+
+		sg_dma_len(s) = bytes;
+		sg_dma_address(s) = dma_buf;
+		spi_s_len -= bytes;
+
+		if (!spi_s_len && sg_next(spi_s)) {
+			spi_s = sg_next(spi_s);
+			spi_s_len = sg_dma_len(spi_s);
+			dma_buf = sg_dma_address(spi_s);
+		} else {
+			dma_buf += bytes;
+		}
+	}
+
+	_mdma_desc = dmaengine_prep_slave_sg(spi->mdma_rx, mdma_sgt.sgl,
+					     mdma_sgt.nents, rx_mdma_conf.direction,
+					     DMA_PREP_INTERRUPT);
+	sg_free_table(&mdma_sgt);
+
+	if (!_mdma_desc) {
+		_dma_desc = NULL;
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
 /**
 * stm32_spi_transfer_one_dma - transfer a single spi_transfer using DMA
 * @spi: pointer to the spi controller data structure
@@ -1430,38 +1584,43 @@ static void stm32h7_spi_transfer_one_dma_start(struct stm32_spi *spi)
 static int stm32_spi_transfer_one_dma(struct stm32_spi *spi,
 				      struct spi_transfer *xfer)
 {
+	struct dma_async_tx_descriptor *rx_mdma_desc = NULL, *rx_dma_desc = NULL;
+	struct dma_async_tx_descriptor *tx_dma_desc = NULL;
 	struct dma_slave_config tx_dma_conf, rx_dma_conf;
-	struct dma_async_tx_descriptor *tx_dma_desc, *rx_dma_desc;
 	unsigned long flags;
+	int ret = 0;

 	spin_lock_irqsave(&spi->lock, flags);

-	rx_dma_desc = NULL;
 	if (spi->rx_buf && spi->dma_rx) {
 		stm32_spi_dma_config(spi, spi->dma_rx, &rx_dma_conf, DMA_DEV_TO_MEM);
-		dmaengine_slave_config(spi->dma_rx, &rx_dma_conf);
+		if (spi->mdma_rx) {
+			rx_dma_conf.peripheral_size = 1;
+			dmaengine_slave_config(spi->dma_rx, &rx_dma_conf);

-		/* Enable Rx DMA request */
-		stm32_spi_set_bits(spi, spi->cfg->regs->dma_rx_en.reg,
-				   spi->cfg->regs->dma_rx_en.mask);
-
-		rx_dma_desc = dmaengine_prep_slave_sg(
-					spi->dma_rx, xfer->rx_sg.sgl,
-					xfer->rx_sg.nents,
-					rx_dma_conf.direction,
-					DMA_PREP_INTERRUPT);
+			ret = stm32_spi_prepare_rx_dma_mdma_chaining(spi, xfer, &rx_dma_conf,
+								     &rx_dma_desc, &rx_mdma_desc);
+			if (ret) { /* RX DMA MDMA chaining not possible, fallback to DMA only */
+				rx_dma_conf.peripheral_config = 0;
+				rx_dma_desc = NULL;
+			}
+		}
+		if (!rx_dma_desc) {
+			dmaengine_slave_config(spi->dma_rx, &rx_dma_conf);
+			rx_dma_desc = dmaengine_prep_slave_sg(spi->dma_rx, xfer->rx_sg.sgl,
+							      xfer->rx_sg.nents,
+							      rx_dma_conf.direction,
+							      DMA_PREP_INTERRUPT);
+		}
 	}

-	tx_dma_desc = NULL;
 	if (spi->tx_buf && spi->dma_tx) {
 		stm32_spi_dma_config(spi, spi->dma_tx, &tx_dma_conf, DMA_MEM_TO_DEV);
 		dmaengine_slave_config(spi->dma_tx, &tx_dma_conf);
-
-		tx_dma_desc = dmaengine_prep_slave_sg(
-					spi->dma_tx, xfer->tx_sg.sgl,
-					xfer->tx_sg.nents,
-					tx_dma_conf.direction,
-					DMA_PREP_INTERRUPT);
+		tx_dma_desc = dmaengine_prep_slave_sg(spi->dma_tx, xfer->tx_sg.sgl,
+						      xfer->tx_sg.nents,
+						      tx_dma_conf.direction,
+						      DMA_PREP_INTERRUPT);
 	}

 	if ((spi->tx_buf && spi->dma_tx && !tx_dma_desc) ||
@@ -1472,9 +1631,25 @@ static int stm32_spi_transfer_one_dma(struct stm32_spi *spi,
 		goto dma_desc_error;

 	if (rx_dma_desc) {
-		rx_dma_desc->callback = spi->cfg->dma_rx_cb;
-		rx_dma_desc->callback_param = spi;
+		if (rx_mdma_desc) {
+			rx_mdma_desc->callback = spi->cfg->dma_rx_cb;
+			rx_mdma_desc->callback_param = spi;
+		} else {
+			rx_dma_desc->callback = spi->cfg->dma_rx_cb;
+			rx_dma_desc->callback_param = spi;
+		}

+		/* Enable Rx DMA request */
+		stm32_spi_set_bits(spi, spi->cfg->regs->dma_rx_en.reg,
+				   spi->cfg->regs->dma_rx_en.mask);
+		if (rx_mdma_desc) {
+			if (dma_submit_error(dmaengine_submit(rx_mdma_desc))) {
+				dev_err(spi->dev, "Rx MDMA submit failed\n");
+				goto dma_desc_error;
+			}
+			/* Enable Rx MDMA channel */
+			dma_async_issue_pending(spi->mdma_rx);
+		}
 		if (dma_submit_error(dmaengine_submit(rx_dma_desc))) {
 			dev_err(spi->dev, "Rx DMA submit failed\n");
 			goto dma_desc_error;
@@ -1509,6 +1684,8 @@ static int stm32_spi_transfer_one_dma(struct stm32_spi *spi,
 	return 1;

 dma_submit_error:
+	if (spi->mdma_rx)
+		dmaengine_terminate_sync(spi->mdma_rx);
 	if (spi->dma_rx)
 		dmaengine_terminate_sync(spi->dma_rx);

@@ -1520,6 +1697,9 @@ dma_desc_error:

 	dev_info(spi->dev, "DMA issue: fall back to irq transfer\n");

+	if (spi->sram_rx_buf)
+		memset(spi->sram_rx_buf, 0, spi->sram_rx_buf_size);
+
 	spi->cur_usedma = false;
 	return spi->cfg->transfer_one_irq(spi);
 }
@@ -1702,11 +1882,26 @@ static int stm32h7_spi_set_mode(struct stm32_spi *spi, unsigned int comm_type)
 * stm32h7_spi_data_idleness - configure minimum time delay inserted between two
 *			       consecutive data frames in host mode
 * @spi: pointer to the spi controller data structure
- * @len: transfer len
+ * @xfer: pointer to spi transfer
 */
-static void stm32h7_spi_data_idleness(struct stm32_spi *spi, u32 len)
+static void stm32h7_spi_data_idleness(struct stm32_spi *spi, struct spi_transfer *xfer)
 {
 	u32 cfg2_clrb = 0, cfg2_setb = 0;
+	u32 len = xfer->len;
+	u32 spi_delay_ns;
+
+	spi_delay_ns = spi_delay_to_ns(&xfer->word_delay, xfer);
+
+	if (spi->cur_midi != 0) {
+		dev_warn(spi->dev, "st,spi-midi-ns DT property is deprecated\n");
+		if (spi_delay_ns) {
+			dev_warn(spi->dev, "Overriding st,spi-midi-ns with word_delay_ns %d\n",
+				 spi_delay_ns);
+			spi->cur_midi = spi_delay_ns;
+		}
+	} else {
+		spi->cur_midi = spi_delay_ns;
+	}

 	cfg2_clrb |= STM32H7_SPI_CFG2_MIDI;
 	if ((len > 1) && (spi->cur_midi > 0)) {
@@ -1768,6 +1963,13 @@ static int stm32_spi_transfer_one_setup(struct stm32_spi *spi,
 	spi->cur_bpw = transfer->bits_per_word;
 	spi->cfg->set_bpw(spi);

+	if (spi_dev->mode & SPI_READY && spi->cur_bpw < 8) {
+		writel_relaxed(readl_relaxed(spi->base + spi->cfg->regs->rdy_en.reg) &
+				~spi->cfg->regs->rdy_en.mask,
+					spi->base + spi->cfg->regs->rdy_en.reg);
+		dev_dbg(spi->dev, "RDY logic disabled as bits per word < 8\n");
+	}
+
 	/* Update spi->cur_speed with real clock speed */
 	if (STM32_SPI_HOST_MODE(spi)) {
 		mbr = stm32_spi_prepare_mbr(spi, transfer->speed_hz,
@@ -1790,7 +1992,7 @@ static int stm32_spi_transfer_one_setup(struct stm32_spi *spi,
 	spi->cur_comm = comm_type;

 	if (STM32_SPI_HOST_MODE(spi) && spi->cfg->set_data_idleness)
-		spi->cfg->set_data_idleness(spi, transfer->len);
+		spi->cfg->set_data_idleness(spi, transfer);

 	if (spi->cur_bpw <= 8)
 		nb_words = transfer->len;
@@ -1871,6 +2073,9 @@ static int stm32_spi_unprepare_msg(struct spi_controller *ctrl,

 	spi->cfg->disable(spi);

+	if (spi->sram_rx_buf)
+		memset(spi->sram_rx_buf, 0, spi->sram_rx_buf_size);
+
 	return 0;
 }

@@ -2069,9 +2274,15 @@ static int stm32_spi_probe(struct platform_device *pdev)
 	struct resource *res;
 	struct reset_control *rst;
 	struct device_node *np = pdev->dev.of_node;
+	const struct stm32_spi_cfg *cfg;
 	bool device_mode;
 	int ret;
-	const struct stm32_spi_cfg *cfg = of_device_get_match_data(&pdev->dev);
+
+	cfg = of_device_get_match_data(&pdev->dev);
+	if (!cfg) {
+		dev_err(&pdev->dev, "Failed to get match data for platform\n");
+		return -ENODEV;
+	}

 	device_mode = of_property_read_bool(np, "spi-slave");
 	if (!cfg->has_device_mode && device_mode) {
@@ -2179,7 +2390,7 @@ static int stm32_spi_probe(struct platform_device *pdev)
 	ctrl->auto_runtime_pm = true;
 	ctrl->bus_num = pdev->id;
 	ctrl->mode_bits = SPI_CPHA | SPI_CPOL | SPI_CS_HIGH | SPI_LSB_FIRST |
-			  SPI_3WIRE;
+			  SPI_3WIRE | SPI_READY;
 	ctrl->bits_per_word_mask = spi->cfg->get_bpw_mask(spi);
 	ctrl->max_speed_hz = spi->clk_rate / spi->cfg->baud_rate_div_min;
 	ctrl->min_speed_hz = spi->clk_rate / spi->cfg->baud_rate_div_max;
@@ -2219,6 +2430,33 @@ static int stm32_spi_probe(struct platform_device *pdev)
 	if (spi->dma_tx || spi->dma_rx)
 		ctrl->can_dma = stm32_spi_can_dma;

+	spi->sram_pool = of_gen_pool_get(pdev->dev.of_node, "sram", 0);
+	if (spi->sram_pool) {
+		spi->sram_rx_buf_size = gen_pool_size(spi->sram_pool);
+		dev_info(&pdev->dev, "SRAM pool: %zu KiB for RX DMA/MDMA chaining\n",
+			 spi->sram_rx_buf_size / 1024);
+		spi->sram_rx_buf = gen_pool_dma_zalloc(spi->sram_pool, spi->sram_rx_buf_size,
+						       &spi->sram_dma_rx_buf);
+		if (!spi->sram_rx_buf) {
+			dev_err(&pdev->dev, "failed to allocate SRAM buffer\n");
+		} else {
+			spi->mdma_rx = dma_request_chan(spi->dev, "rxm2m");
+			if (IS_ERR(spi->mdma_rx)) {
+				ret = PTR_ERR(spi->mdma_rx);
+				spi->mdma_rx = NULL;
+				if (ret == -EPROBE_DEFER) {
+					goto err_pool_free;
+				} else {
+					gen_pool_free(spi->sram_pool,
+						      (unsigned long)spi->sram_rx_buf,
+						      spi->sram_rx_buf_size);
+					dev_warn(&pdev->dev,
+						 "failed to request rx mdma channel, DMA only\n");
+				}
+			}
+		}
+	}
+
 	pm_runtime_set_autosuspend_delay(&pdev->dev,
 					 STM32_SPI_AUTOSUSPEND_DELAY);
 	pm_runtime_use_autosuspend(&pdev->dev);
@@ -2233,7 +2471,6 @@ static int stm32_spi_probe(struct platform_device *pdev)
 		goto err_pm_disable;
 	}

-	pm_runtime_mark_last_busy(&pdev->dev);
 	pm_runtime_put_autosuspend(&pdev->dev);

 	dev_info(&pdev->dev, "driver initialized (%s mode)\n",
@@ -2246,6 +2483,13 @@ err_pm_disable:
 	pm_runtime_put_noidle(&pdev->dev);
 	pm_runtime_set_suspended(&pdev->dev);
 	pm_runtime_dont_use_autosuspend(&pdev->dev);
+
+	if (spi->mdma_rx)
+		dma_release_channel(spi->mdma_rx);
+err_pool_free:
+	if (spi->sram_pool)
+		gen_pool_free(spi->sram_pool, (unsigned long)spi->sram_rx_buf,
+			      spi->sram_rx_buf_size);
 err_dma_release:
 	if (spi->dma_tx)
 		dma_release_channel(spi->dma_tx);
@@ -2276,6 +2520,11 @@ static void stm32_spi_remove(struct platform_device *pdev)
 		dma_release_channel(ctrl->dma_tx);
 	if (ctrl->dma_rx)
 		dma_release_channel(ctrl->dma_rx);
+	if (spi->mdma_rx)
+		dma_release_channel(spi->mdma_rx);
+	if (spi->sram_rx_buf)
+		gen_pool_free(spi->sram_pool, (unsigned long)spi->sram_rx_buf,
+			      spi->sram_rx_buf_size);

 	clk_disable_unprepare(spi->clk);

@@ -2342,7 +2591,6 @@ static int __maybe_unused stm32_spi_resume(struct device *dev)

 	spi->cfg->config(spi);

-	pm_runtime_mark_last_busy(dev);
 	pm_runtime_put_autosuspend(dev);

 	return 0;
--- a/drivers/spi/spi-ti-qspi.c
+++ b/drivers/spi/spi-ti-qspi.c
@@ -158,7 +158,6 @@ static int ti_qspi_setup(struct spi_device *spi)
 		return ret;
 	}

-	pm_runtime_mark_last_busy(qspi->dev);
 	ret = pm_runtime_put_autosuspend(qspi->dev);
 	if (ret < 0) {
 		dev_err(qspi->dev, "pm_runtime_put_autosuspend() failed\n");
@@ -195,7 +194,6 @@ static void ti_qspi_setup_clk(struct ti_qspi *qspi, u32 speed_hz)
 		ctx_reg->clkctrl = clk_ctrl_new;
 	}

-	pm_runtime_mark_last_busy(qspi->dev);
 	pm_runtime_put_autosuspend(qspi->dev);
 }

--- a/drivers/spi/spi-xilinx.c
+++ b/drivers/spi/spi-xilinx.c
@@ -89,8 +89,8 @@ struct xilinx_spi {
 	u8 bytes_per_word;
 	int buffer_size;	/* buffer size in words */
 	u32 cs_inactive;	/* Level of the CS pins when inactive*/
-	unsigned int (*read_fn)(void __iomem *);
-	void (*write_fn)(u32, void __iomem *);
+	unsigned int (*read_fn)(void __iomem *addr);
+	void (*write_fn)(u32 val, void __iomem *addr);
 };

 static void xspi_write32(u32 val, void __iomem *addr)
@@ -251,6 +251,7 @@ static int xilinx_spi_txrx_bufs(struct spi_device *spi, struct spi_transfer *t)
 	if (xspi->irq >= 0 &&
 	    (xspi->force_irq || remaining_words > xspi->buffer_size)) {
 		u32 isr;
+
 		use_irq = true;
 		/* Inhibit irq to avoid spurious irqs on tx_empty*/
 		cr = xspi->read_fn(xspi->regs + XSPI_CR_OFFSET);
--- a/drivers/spi/spi-zynqmp-gqspi.c
+++ b/drivers/spi/spi-zynqmp-gqspi.c
@@ -1330,7 +1330,6 @@ static int zynqmp_qspi_probe(struct platform_device *pdev)
 		goto clk_dis_all;
 	}

-	pm_runtime_mark_last_busy(&pdev->dev);
 	pm_runtime_put_autosuspend(&pdev->dev);

 	return 0;
--- a/drivers/spi/spi.c
+++ b/drivers/spi/spi.c
@@ -1723,7 +1723,6 @@ EXPORT_SYMBOL_GPL(spi_finalize_current_transfer);
 static void spi_idle_runtime_pm(struct spi_controller *ctlr)
 {
 	if (ctlr->auto_runtime_pm) {
-		pm_runtime_mark_last_busy(ctlr->dev.parent);
 		pm_runtime_put_autosuspend(ctlr->dev.parent);
 	}
 }
@@ -3856,7 +3855,6 @@ static int spi_set_cs_timing(struct spi_device *spi)
 			}

 			status = spi->controller->set_cs_timing(spi);
-			pm_runtime_mark_last_busy(parent);
 			pm_runtime_put_autosuspend(parent);
 		} else {
 			status = spi->controller->set_cs_timing(spi);
@@ -3991,7 +3989,6 @@ int spi_setup(struct spi_device *spi)
 		status = 0;

 		spi_set_cs(spi, false, true);
-		pm_runtime_mark_last_busy(spi->controller->dev.parent);
 		pm_runtime_put_autosuspend(spi->controller->dev.parent);
 	} else {
 		spi_set_cs(spi, false, true);
--- a/drivers/spi/spidev.c
+++ b/drivers/spi/spidev.c
@@ -703,6 +703,7 @@ static const struct class spidev_class = {
 * spidev_dt_ids array below. Both arrays are kept in the same ordering.
 */
 static const struct spi_device_id spidev_spi_ids[] = {
+	{ .name = /* abb */ "spi-sensor" },
 	{ .name = /* cisco */ "spi-petra" },
 	{ .name = /* dh */ "dhcom-board" },
 	{ .name = /* elgin */ "jg10309-01" },
@@ -735,6 +736,7 @@ static int spidev_of_check(struct device *dev)
 }

 static const struct of_device_id spidev_dt_ids[] = {
+	{ .compatible = "abb,spi-sensor", .data = &spidev_of_check },
 	{ .compatible = "cisco,spi-petra", .data = &spidev_of_check },
 	{ .compatible = "dh,dhcom-board", .data = &spidev_of_check },
 	{ .compatible = "elgin,jg10309-01", .data = &spidev_of_check },
--- a/include/linux/dmaengine.h
+++ b/include/linux/dmaengine.h
@@ -1524,6 +1524,7 @@ struct dma_chan *__dma_request_channel(const dma_cap_mask_t *mask,

 struct dma_chan *dma_request_chan(struct device *dev, const char *name);
 struct dma_chan *dma_request_chan_by_mask(const dma_cap_mask_t *mask);
+struct dma_chan *devm_dma_request_chan(struct device *dev, const char *name);

 void dma_release_channel(struct dma_chan *chan);
 int dma_get_slave_caps(struct dma_chan *chan, struct dma_slave_caps *caps);
@@ -1560,6 +1561,12 @@ static inline struct dma_chan *dma_request_chan_by_mask(
 {
 	return ERR_PTR(-ENODEV);
 }
+
+static inline struct dma_chan *devm_dma_request_chan(struct device *dev, const char *name)
+{
+	return ERR_PTR(-ENODEV);
+}
+
 static inline void dma_release_channel(struct dma_chan *chan)
 {
 }
--- a/include/linux/mtd/nand-qpic-common.h
+++ b/include/linux/mtd/nand-qpic-common.h
@@ -101,6 +101,8 @@
 #define	ECC_SW_RESET			BIT(1)
 #define	ECC_MODE			4
 #define	ECC_MODE_MASK			GENMASK(5, 4)
+#define	ECC_MODE_4BIT			0
+#define	ECC_MODE_8BIT			1
 #define	ECC_PARITY_SIZE_BYTES_BCH	8
 #define	ECC_PARITY_SIZE_BYTES_BCH_MASK	GENMASK(12, 8)
 #define	ECC_NUM_DATA_BYTES		16