hollalinux
/
linux-sg2042
mirror of https://github.com/sophgo/linux-riscv.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

drm next pull request for 5.1 

-----BEGIN PGP SIGNATURE-----
 
 iQIcBAABAgAGBQJcfzdFAAoJEAx081l5xIa+QxsP/A6QP+gx4vQ8XXikaJMNz89e
 59TAbXHW/2qFMHRtUesuB2bc1a2cw2ppFsrryG7c4HqjKDDHna7Yx2JzZYL0MmNh
 SpJYL4yMuu/2TmyCouaAYzzP+5Supdosfif4LRn3269DH0i5MWXL+NVrbeB47blG
 XwjQTu46yfn06IFAo5bI2jMqSuPCDd4Hzpyixpvmjt+r16XwoH5nGUrDCHG8t/jV
 +PUZCYAjn71in6Z66MKZv/EVCVFfTnaVJ2KEgw7e+vWxnERkRh/xnRO6KIXMD5O1
 vo2qc2vbxkGpjaE6pDzC/2e5pRJT8Ks0t50jYjbVF+6nHpP5XIPvAXH4R2QdTA7B
 Jiu8N0oz6wj0H3AJ/V38rEHWW8zgOfXkhbRBfmfQ9NfgiEfwxqCVgspIOwei4oVw
 hvMXYUBM1CU+JIfW6w7ZT4oHALUlnCpnr5DQRdCNRm8zjClyNfIAoJIJrOtqmX44
 qjEzSgxb89ZtS7c0yislSBaovgAmcM3I+aq5I4xokdY0hFEZ6QomuKunyuQ8pBYa
 3gsvMEReLxETffhhYpjBt5+b5IgB49nf3Y38CKFurv32Sp0p0YgK0qVo8qRQHclj
 QIJ+3+zQMCX20swYpCWXhOPUIwtQppdKhWzg12my8rL2VgTlYhjlEbL4EL+Wk+hv
 6Ipulthzn0RyrSK9Dojh
 =GlRQ
 -----END PGP SIGNATURE-----

Merge tag 'drm-next-2019-03-06' of git://anongit.freedesktop.org/drm/drm

Pull drm updates from Dave Airlie:
 "This is the main drm pull request for the 5.1 merge window.

  The big changes I'd highlight are:
   - nouveau has HMM support now, there is finally an in-tree user so we
     can quieten down the rip it out people.
   - i915 now enables fastboot by default on Skylake+
   - Displayport Multistream support has been refactored and should
     hopefully be more reliable.

  Core:
   - header cleanups aiming towards removing drmP.h
   - dma-buf fence seqnos to 64-bits
   - common helper for DP mst hotplug for radeon,i915,amdgpu + new
     refcounting scheme
   - MST i2c improvements
   - drm_syncobj_cb removal
   - ARM FB compression fourcc
   - P010 + P016 fourcc
   - allwinner tiled format modifier
   - i2c over aux I2C_M_STOP support
   - DRM_AUTH handling fixes

  TTM:
   - ref/unref renaming

  New driver:
   - ARM komeda display driver

  scheduler:
   - refactor mirror list handling
   - rework hw fence processing
   - 0 run queue entity fix

  bridge:
   - TI DS90C185 LVDS bridge
   - thc631lvdm83d bridge improvements
   - cadence + allwinner DSI ported to generic phy

  panels:
   - Sitronix ST7701 panel
   - Kingdisplay KD097D04
   - LeMaker BL035-RGB-002
   - PDA 91-00156-A0
   - Innolux EE101IA-01D

  i915:
   - Enable fastboot by default on SKL+/VLV/CHV
   - Export RPCS configuration for ICL media driver
   - Coffelake PCI ID
   - CNL clocks setup fixes
   - ACPI/PMIC support for MIPI/DSI
   - Per-engine WA init for all engines
   - Shrinker locking fixes
   - Kerneldoc updates
   - Lots of ring improvements and reset fixes
   - Coffeelake GVT Support
   - VFIO GVT EDID Region support
   - runtime PM wakeref tracking
   - ILK->IVB primary plane enable delays
   - userptr mutex locking fixes
   - DSI fixes
   - LVDS/TV cleanups
   - HW readout fixes
   - LUT robustness fixes
   - ICL display and watermark fixes
   - gem mmap race fix

  amdgpu:
   - add scheduled dependencies interface
   - DCC on scanout surfaces
   - vega10/20 BACO support
   - Multiple IH rings on soc15
   - XGMI locking fixes
   - DC i2c/aux cleanups
   - runtime SMU debug interface
   - Kexec improvmeents
   - SR-IOV fixes
   - DC freesync + ABM fixes
   - GDS fixes
   - GPUVM fixes
   - vega20 PCIE DPM switching fixes
   - Context priority handling fixes

  radeon:
   - fix missing break in evergreen parser

  nouveau:
   - SVM support via HMM

  msm:
   - QCOM Compressed modifier support

  exynos:
   - s5pv210 rotator support

  imx:
   - zpos property support
   - pending update fixes

  v3d:
   - cache flush improvments

  vc4:
   - reflection support
   - HDMI overscan support

  tegra:
   - CEC refactoring
   - HDMI audio fixes
   - Tegra186 prep work
   - SOR crossbar device tree fixes

  sun4i:
   - implicit fencing support
   - YUV and scalar support improvements
   - A23 support
   - tiling fixes

  atmel-hlcdc:
   - clipping and rotation property fixes

  qxl:
   - BO and PRIME improvements
   - generic fbdev emulation

  dw-hdmi:
   - HDMI 2.0 2160p
   - YUV420 ouput

  rockchip:
   - implicit fencing support
   - reflection proerties

  virtio-gpu:
   - use generic fbdev emulation

  tilcdc:
   - cpufreq vs crtc init fix

  rcar-du:
   - R8A774C0 support
   - D3/E3 RGB output routing fixes and DPAD0 support
   - RA87744 LVDS support

  bochs:
   - atomic and generic fbdev emulation
   - ID mismatch error on bochs load

  meson:
   - remove firmware fbs"

* tag 'drm-next-2019-03-06' of git://anongit.freedesktop.org/drm/drm: (1130 commits)
  drm/amd/display: Use vrr friendly pageflip throttling in DC.
  drm/imx: only send commit done event when all state has been applied
  drm/imx: allow building under COMPILE_TEST
  drm/imx: imx-tve: depend on COMMON_CLK
  drm/imx: ipuv3-plane: add zpos property
  drm/imx: ipuv3-plane: add function to query atomic update status
  gpu: ipu-v3: prg: add function to get channel configure status
  gpu: ipu-v3: pre: add double buffer status readback
  drm/amdgpu: Bump amdgpu version for context priority override.
  drm/amdgpu/powerplay: fix typo in BACO header guards
  drm/amdgpu/powerplay: fix return codes in BACO code
  drm/amdgpu: add missing license on baco files
  drm/bochs: Fix the ID mismatch error
  drm/nouveau/dmem: use dma addresses during migration copies
  drm/nouveau/dmem: use physical vram addresses during migration copies
  drm/nouveau/dmem: extend copy function to allow direct use of physical addresses
  drm/nouveau/svm: new ioctl to migrate process memory to GPU memory
  drm/nouveau/dmem: device memory helpers for SVM
  drm/nouveau/svm: initial support for shared virtual memory
  drm/nouveau: prepare for enabling svm with existing userspace interfaces
  ...
This commit is contained in:
Linus Torvalds 2019-03-08 08:23:15 -08:00
parent b5dd0c658c 4b057e73f2
commit 851ca779d1
1141 changed files with 39790 additions and 38139 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

73
Documentation/devicetree/bindings/display/arm,komeda.txt Normal file
View File

@ -0,0 +1,73 @@
Device Tree bindings for Arm Komeda display driver
Required properties:
- compatible: Should be "arm,mali-d71"
- reg: Physical base address and length of the registers in the system
- interrupts: the interrupt line number of the device in the system
- clocks: A list of phandle + clock-specifier pairs, one for each entry
in 'clock-names'
- clock-names: A list of clock names. It should contain:
- "mclk": for the main processor clock
- "pclk": for the APB interface clock
- #address-cells: Must be 1
- #size-cells: Must be 0
Required properties for sub-node: pipeline@nq
Each device contains one or two pipeline sub-nodes (at least one), each
pipeline node should provide properties:
- reg: Zero-indexed identifier for the pipeline
- clocks: A list of phandle + clock-specifier pairs, one for each entry
in 'clock-names'
- clock-names: should contain:
- "pxclk": pixel clock
- "aclk": AXI interface clock
- port: each pipeline connect to an encoder input port. The connection is
modeled using the OF graph bindings specified in
Documentation/devicetree/bindings/graph.txt
Optional properties:
- memory-region: phandle to a node describing memory (see
Documentation/devicetree/bindings/reserved-memory/reserved-memory.txt)
to be used for the framebuffer; if not present, the framebuffer may
be located anywhere in memory.
Example:
/ {
...
dp0: display@c00000 {
#address-cells = <1>;
#size-cells = <0>;
compatible = "arm,mali-d71";
reg = <0xc00000 0x20000>;
interrupts = <0 168 4>;
clocks = <&dpu_mclk>, <&dpu_aclk>;
clock-names = "mclk", "pclk";
dp0_pipe0: pipeline@0 {
clocks = <&fpgaosc2>, <&dpu_aclk>;
clock-names = "pxclk", "aclk";
reg = <0>;
port {
dp0_pipe0_out: endpoint {
remote-endpoint = <&db_dvi0_in>;
};
};
};
dp0_pipe1: pipeline@1 {
clocks = <&fpgaosc2>, <&dpu_aclk>;
clock-names = "pxclk", "aclk";
reg = <1>;
port {
dp0_pipe1_out: endpoint {
remote-endpoint = <&db_dvi1_in>;
};
};
};
};
...
};

12
Documentation/devicetree/bindings/display/bridge/lvds-transmitter.txt
View File

@ -22,13 +22,11 @@ among others.
Required properties:
- compatible: Must be one or more of the following
- "ti,ds90c185" for the TI DS90C185 FPD-Link Serializer
- "lvds-encoder" for a generic LVDS encoder device
- compatible: Must be "lvds-encoder"
When compatible with the generic version, nodes must list the
device-specific version corresponding to the device first
followed by the generic version.
Any encoder compatible with this generic binding, but with additional
properties not listed here, must list a device specific compatible first
followed by this generic compatible.
Required nodes:
@ -44,8 +42,6 @@ Example
lvds-encoder {
compatible = "lvds-encoder";
#address-cells = <1>;
#size-cells = <0>;
ports {
#address-cells = <1>;

4
Documentation/devicetree/bindings/display/bridge/renesas,lvds.txt
View File

@ -8,6 +8,8 @@ Required properties:
- compatible : Shall contain one of
- "renesas,r8a7743-lvds" for R8A7743 (RZ/G1M) compatible LVDS encoders
- "renesas,r8a7744-lvds" for R8A7744 (RZ/G1N) compatible LVDS encoders
- "renesas,r8a774c0-lvds" for R8A774C0 (RZ/G2E) compatible LVDS encoders
- "renesas,r8a7790-lvds" for R8A7790 (R-Car H2) compatible LVDS encoders
- "renesas,r8a7791-lvds" for R8A7791 (R-Car M2-W) compatible LVDS encoders
- "renesas,r8a7793-lvds" for R8A7793 (R-Car M2-N) compatible LVDS encoders
@ -25,7 +27,7 @@ Required properties:
- clock-names: Name of the clocks. This property is model-dependent.
- The functional clock, which mandatory for all models, shall be listed
first, and shall be named "fck".
- On R8A77990 and R8A77995, the LVDS encoder can use the EXTAL or
- On R8A77990, R8A77995 and R8A774C0, the LVDS encoder can use the EXTAL or
DU_DOTCLKINx clocks. Those clocks are optional. When supplied they must be
named "extal" and "dclkin.x" respectively, with "x" being the DU_DOTCLKIN
numerical index.

2
Documentation/devicetree/bindings/display/bridge/thine,thc63lvdm83d.txt
View File

@ -10,7 +10,7 @@ Required properties:
Optional properties:
- pwdn-gpios: Power down control GPIO
- powerdown-gpios: Power down control GPIO (the /PWDN pin, active low).
Required nodes:

55
Documentation/devicetree/bindings/display/bridge/ti,ds90c185.txt Normal file
View File

@ -0,0 +1,55 @@
Texas Instruments FPD-Link (LVDS) Serializer
--------------------------------------------
The DS90C185 and DS90C187 are low-power serializers for portable
battery-powered applications that reduces the size of the RGB
interface between the host GPU and the display.
Required properties:
- compatible: Should be
"ti,ds90c185", "lvds-encoder" for the TI DS90C185 FPD-Link Serializer
"ti,ds90c187", "lvds-encoder" for the TI DS90C187 FPD-Link Serializer
Optional properties:
- powerdown-gpios: Power down control GPIO (the PDB pin, active-low)
Required nodes:
The devices have two video ports. Their connections are modeled using the OF
graph bindings specified in Documentation/devicetree/bindings/graph.txt.
- Video port 0 for parallel input
- Video port 1 for LVDS output
Example
-------
lvds-encoder {
compatible = "ti,ds90c185", "lvds-encoder";
powerdown-gpios = <&gpio 17 GPIO_ACTIVE_LOW>;
ports {
#address-cells = <1>;
#size-cells = <0>;
port@0 {
reg = <0>;
lvds_enc_in: endpoint {
remote-endpoint = <&lcdc_out_rgb>;
};
};
port@1 {
reg = <1>;
lvds_enc_out: endpoint {
remote-endpoint = <&lvds_panel_in>;
};
};
};
};

59
Documentation/devicetree/bindings/display/msm/gmu.txt Normal file
View File

@ -0,0 +1,59 @@
Qualcomm adreno/snapdragon GMU (Graphics management unit)
The GMU is a programmable power controller for the GPU. the CPU controls the
GMU which in turn handles power controls for the GPU.
Required properties:
- compatible: "qcom,adreno-gmu-XYZ.W", "qcom,adreno-gmu"
for example: "qcom,adreno-gmu-630.2", "qcom,adreno-gmu"
Note that you need to list the less specific "qcom,adreno-gmu"
for generic matches and the more specific identifier to identify
the specific device.
- reg: Physical base address and length of the GMU registers.
- reg-names: Matching names for the register regions
* "gmu"
* "gmu_pdc"
* "gmu_pdc_seg"
- interrupts: The interrupt signals from the GMU.
- interrupt-names: Matching names for the interrupts
* "hfi"
* "gmu"
- clocks: phandles to the device clocks
- clock-names: Matching names for the clocks
* "gmu"
* "cxo"
* "axi"
* "mnoc"
- power-domains: should be <&clock_gpucc GPU_CX_GDSC>
- iommus: phandle to the adreno iommu
- operating-points-v2: phandle to the OPP operating points
Example:
/ {
...
gmu: gmu@506a000 {
compatible="qcom,adreno-gmu-630.2", "qcom,adreno-gmu";
reg = <0x506a000 0x30000>,
<0xb280000 0x10000>,
<0xb480000 0x10000>;
reg-names = "gmu", "gmu_pdc", "gmu_pdc_seq";
interrupts = <GIC_SPI 304 IRQ_TYPE_LEVEL_HIGH>,
<GIC_SPI 305 IRQ_TYPE_LEVEL_HIGH>;
interrupt-names = "hfi", "gmu";
clocks = <&gpucc GPU_CC_CX_GMU_CLK>,
<&gpucc GPU_CC_CXO_CLK>,
<&gcc GCC_DDRSS_GPU_AXI_CLK>,
<&gcc GCC_GPU_MEMNOC_GFX_CLK>;
clock-names = "gmu", "cxo", "axi", "memnoc";
power-domains = <&gpucc GPU_CX_GDSC>;
iommus = <&adreno_smmu 5>;
operating-points-v2 = <&gmu_opp_table>;
};
};

42
Documentation/devicetree/bindings/display/msm/gpu.txt
View File

@ -10,14 +10,23 @@ Required properties:
If "amd,imageon" is used, there should be no top level msm device.
- reg: Physical base address and length of the controller's registers.
- interrupts: The interrupt signal from the gpu.
- clocks: device clocks
- clocks: device clocks (if applicable)
See ../clocks/clock-bindings.txt for details.
- clock-names: the following clocks are required:
- clock-names: the following clocks are required by a3xx, a4xx and a5xx
cores:
* "core"
* "iface"
* "mem_iface"
For GMU attached devices the GPU clocks are not used and are not required. The
following devices should not list clocks:
- qcom,adreno-630.2
- iommus: optional phandle to an adreno iommu instance
- operating-points-v2: optional phandle to the OPP operating points
- qcom,gmu: For GMU attached devices a phandle to the GMU device that will
control the power for the GPU. Applicable targets:
- qcom,adreno-630.2
Example:
Example 3xx/4xx/a5xx:
/ {
...
@ -37,3 +46,30 @@ Example:
<&mmcc MMSS_IMEM_AHB_CLK>;
};
};
Example a6xx (with GMU):
/ {
...
gpu@5000000 {
compatible = "qcom,adreno-630.2", "qcom,adreno";
#stream-id-cells = <16>;
reg = <0x5000000 0x40000>, <0x509e000 0x10>;
reg-names = "kgsl_3d0_reg_memory", "cx_mem";
/*
* Look ma, no clocks! The GPU clocks and power are
* controlled entirely by the GMU
*/
interrupts = <GIC_SPI 300 IRQ_TYPE_LEVEL_HIGH>;
iommus = <&adreno_smmu 0>;
operating-points-v2 = <&gpu_opp_table>;
qcom,gmu = <&gmu>;
};
};

0
Documentation/devicetree/bindings/display/panel/auo,g101evn010 → Documentation/devicetree/bindings/display/panel/auo,g101evn010.txt
View File

7
Documentation/devicetree/bindings/display/panel/innolux,ee101ia-01d.txt Normal file
View File

@ -0,0 +1,7 @@
Innolux Corporation 10.1" EE101IA-01D WXGA (1280x800) LVDS panel
Required properties:
- compatible: should be "innolux,ee101ia-01d"
This binding is compatible with the lvds-panel binding, which is specified
in panel-lvds.txt in this directory.

12
Documentation/devicetree/bindings/display/panel/lemaker,bl035-rgb-002.txt Normal file
View File

@ -0,0 +1,12 @@
LeMaker BL035-RGB-002 3.5" QVGA TFT LCD panel
Required properties:
- compatible: should be "lemaker,bl035-rgb-002"
- power-supply: as specified in the base binding
Optional properties:
- backlight: as specified in the base binding
- enable-gpios: as specified in the base binding
This binding is compatible with the simple-panel binding, which is specified
in simple-panel.txt in this directory.

14
Documentation/devicetree/bindings/display/panel/pda,91-00156-a0.txt Normal file
View File

@ -0,0 +1,14 @@
PDA 91-00156-A0 5.0" WVGA TFT LCD panel
Required properties:
- compatible: should be "pda,91-00156-a0"
- power-supply: this panel requires a single power supply. A phandle to a
regulator needs to be specified here. Compatible with panel-common binding which
is specified in the panel-common.txt in this directory.
- backlight: this panel's backlight is controlled by an external backlight
controller. A phandle to this controller needs to be specified here.
Compatible with panel-common binding which is specified in the panel-common.txt
in this directory.
This binding is compatible with the simple-panel binding, which is specified
in simple-panel.txt in this directory.

30
Documentation/devicetree/bindings/display/panel/sitronix,st7701.txt Normal file
View File

@ -0,0 +1,30 @@
Sitronix ST7701 based LCD panels
ST7701 designed for small and medium sizes of TFT LCD display, is
capable of supporting up to 480RGBX864 in resolution. It provides
several system interfaces like MIPI/RGB/SPI.
Techstar TS8550B is 480x854, 2-lane MIPI DSI LCD panel which has
inbuilt ST7701 chip.
Required properties:
- compatible: must be "sitronix,st7701" and one of
* "techstar,ts8550b"
- reset-gpios: a GPIO phandle for the reset pin
Required properties for techstar,ts8550b:
- reg: DSI virtual channel used by that screen
- VCC-supply: analog regulator for MIPI circuit
- IOVCC-supply: I/O system regulator
Optional properties:
- backlight: phandle for the backlight control.
panel@0 {
compatible = "techstar,ts8550b", "sitronix,st7701";
reg = <0>;
VCC-supply = <&reg_dldo2>;
IOVCC-supply = <&reg_dldo2>;
reset-gpios = <&pio 3 24 GPIO_ACTIVE_HIGH>; /* LCD-RST: PD24 */
backlight = <&backlight>;
};

2
Documentation/devicetree/bindings/display/renesas,du.txt
View File

@ -7,6 +7,7 @@ Required Properties:
- "renesas,du-r8a7744" for R8A7744 (RZ/G1N) compatible DU
- "renesas,du-r8a7745" for R8A7745 (RZ/G1E) compatible DU
- "renesas,du-r8a77470" for R8A77470 (RZ/G1C) compatible DU
- "renesas,du-r8a774c0" for R8A774C0 (RZ/G2E) compatible DU
- "renesas,du-r8a7779" for R8A7779 (R-Car H1) compatible DU
- "renesas,du-r8a7790" for R8A7790 (R-Car H2) compatible DU
- "renesas,du-r8a7791" for R8A7791 (R-Car M2-W) compatible DU
@ -57,6 +58,7 @@ corresponding to each DU output.
R8A7744 (RZ/G1N) DPAD 0 LVDS 0 - -
R8A7745 (RZ/G1E) DPAD 0 DPAD 1 - -
R8A77470 (RZ/G1C) DPAD 0 DPAD 1 LVDS 0 -
R8A774C0 (RZ/G2E) DPAD 0 LVDS 0 LVDS 1 -
R8A7779 (R-Car H1) DPAD 0 DPAD 1 - -
R8A7790 (R-Car H2) DPAD 0 LVDS 0 LVDS 1 -
R8A7791 (R-Car M2-W) DPAD 0 LVDS 0 - -

1
Documentation/devicetree/bindings/display/rockchip/rockchip-vop.txt
View File

@ -10,6 +10,7 @@ Required properties:
"rockchip,rk3126-vop";
"rockchip,px30-vop-lit";
"rockchip,px30-vop-big";
"rockchip,rk3066-vop";
"rockchip,rk3188-vop";
"rockchip,rk3288-vop";
"rockchip,rk3368-vop";

5
Documentation/devicetree/bindings/display/sunxi/sun4i-drm.txt
View File

@ -156,6 +156,7 @@ Required properties:
* allwinner,sun6i-a31-tcon
* allwinner,sun6i-a31s-tcon
* allwinner,sun7i-a20-tcon
* allwinner,sun8i-a23-tcon
* allwinner,sun8i-a33-tcon
* allwinner,sun8i-a83t-tcon-lcd
* allwinner,sun8i-a83t-tcon-tv
@ -276,6 +277,7 @@ Required properties:
- compatible: value must be one of:
* allwinner,sun6i-a31-drc
* allwinner,sun6i-a31s-drc
* allwinner,sun8i-a23-drc
* allwinner,sun8i-a33-drc
* allwinner,sun9i-a80-drc
- reg: base address and size of the memory-mapped region.
@ -303,6 +305,7 @@ Required properties:
* allwinner,sun5i-a13-display-backend
* allwinner,sun6i-a31-display-backend
* allwinner,sun7i-a20-display-backend
* allwinner,sun8i-a23-display-backend
* allwinner,sun8i-a33-display-backend
* allwinner,sun9i-a80-display-backend
- reg: base address and size of the memory-mapped region.
@ -360,6 +363,7 @@ Required properties:
* allwinner,sun5i-a13-display-frontend
* allwinner,sun6i-a31-display-frontend
* allwinner,sun7i-a20-display-frontend
* allwinner,sun8i-a23-display-frontend
* allwinner,sun8i-a33-display-frontend
* allwinner,sun9i-a80-display-frontend
- reg: base address and size of the memory-mapped region.
@ -419,6 +423,7 @@ Required properties:
* allwinner,sun6i-a31-display-engine
* allwinner,sun6i-a31s-display-engine
* allwinner,sun7i-a20-display-engine
* allwinner,sun8i-a23-display-engine
* allwinner,sun8i-a33-display-engine
* allwinner,sun8i-a83t-display-engine
* allwinner,sun8i-h3-display-engine

3
Documentation/devicetree/bindings/display/tegra/nvidia,tegra20-host1x.txt
View File

@ -238,6 +238,9 @@ of the following host1x client modules:
- nvidia,hpd-gpio: specifies a GPIO used for hotplug detection
- nvidia,edid: supplies a binary EDID blob
- nvidia,panel: phandle of a display panel
- nvidia,xbar-cfg: 5 cells containing the crossbar configuration. Each lane
of the SOR, identified by the cell's index, is mapped via the crossbar to
the pad specified by the cell's value.
Optional properties when driving an eDP output:
- nvidia,dpaux: phandle to a DispayPort AUX interface

7
Documentation/devicetree/bindings/gpu/samsung-rotator.txt
View File

@ -2,9 +2,10 @@
Required properties:
- compatible : value should be one of the following:
(a) "samsung,exynos4210-rotator" for Rotator IP in Exynos4210
(b) "samsung,exynos4212-rotator" for Rotator IP in Exynos4212/4412
(c) "samsung,exynos5250-rotator" for Rotator IP in Exynos5250
* "samsung,s5pv210-rotator" for Rotator IP in S5PV210
* "samsung,exynos4210-rotator" for Rotator IP in Exynos4210
* "samsung,exynos4212-rotator" for Rotator IP in Exynos4212/4412
* "samsung,exynos5250-rotator" for Rotator IP in Exynos5250
- reg : Physical base address of the IP registers and length of memory
mapped region.

2
Documentation/devicetree/bindings/vendor-prefixes.txt
View File

@ -216,6 +216,7 @@ laird Laird PLC
lantiq Lantiq Semiconductor
lattice Lattice Semiconductor
lego LEGO Systems A/S
lemaker Shenzhen LeMaker Technology Co., Ltd.
lenovo Lenovo Group Ltd.
lg LG Corporation
libretech Shenzhen Libre Technology Co., Ltd
@ -303,6 +304,7 @@ ovti OmniVision Technologies
oxsemi Oxford Semiconductor, Ltd.
panasonic Panasonic Corporation
parade Parade Technologies Inc.
pda Precision Design Associates, Inc.
pericom Pericom Technology Inc.
pervasive Pervasive Displays, Inc.
phicomm PHICOMM Co., Ltd.

235
Documentation/gpu/afbc.rst Normal file
View File

@ -0,0 +1,235 @@
.. SPDX-License-Identifier: GPL-2.0+
===================================
Arm Framebuffer Compression (AFBC)
===================================
AFBC is a proprietary lossless image compression protocol and format.
It provides fine-grained random access and minimizes the amount of
data transferred between IP blocks.
AFBC can be enabled on drivers which support it via use of the AFBC
format modifiers defined in drm_fourcc.h. See DRM_FORMAT_MOD_ARM_AFBC(*).
All users of the AFBC modifiers must follow the usage guidelines laid
out in this document, to ensure compatibility across different AFBC
producers and consumers.
Components and Ordering
=======================
AFBC streams can contain several components - where a component
corresponds to a color channel (i.e. R, G, B, X, A, Y, Cb, Cr).
The assignment of input/output color channels must be consistent
between the encoder and the decoder for correct operation, otherwise
the consumer will interpret the decoded data incorrectly.
Furthermore, when the lossless colorspace transform is used
(AFBC_FORMAT_MOD_YTR, which should be enabled for RGB buffers for
maximum compression efficiency), the component order must be:
* Component 0: R
* Component 1: G
* Component 2: B
The component ordering is communicated via the fourcc code in the
fourcc:modifier pair. In general, component '0' is considered to
reside in the least-significant bits of the corresponding linear
format. For example, COMP(bits):
* DRM_FORMAT_ABGR8888
* Component 0: R(8)
* Component 1: G(8)
* Component 2: B(8)
* Component 3: A(8)
* DRM_FORMAT_BGR888
* Component 0: R(8)
* Component 1: G(8)
* Component 2: B(8)
* DRM_FORMAT_YUYV
* Component 0: Y(8)
* Component 1: Cb(8, 2x1 subsampled)
* Component 2: Cr(8, 2x1 subsampled)
In AFBC, 'X' components are not treated any differently from any other
component. Therefore, an AFBC buffer with fourcc DRM_FORMAT_XBGR8888
encodes with 4 components, like so:
* DRM_FORMAT_XBGR8888
* Component 0: R(8)
* Component 1: G(8)
* Component 2: B(8)
* Component 3: X(8)
Please note, however, that the inclusion of a "wasted" 'X' channel is
bad for compression efficiency, and so it's recommended to avoid
formats containing 'X' bits. If a fourth component is
required/expected by the encoder/decoder, then it is recommended to
instead use an equivalent format with alpha, setting all alpha bits to
'1'. If there is no requirement for a fourth component, then a format
which doesn't include alpha can be used, e.g. DRM_FORMAT_BGR888.
Number of Planes
================
Formats which are typically multi-planar in linear layouts (e.g. YUV
420), can be encoded into one, or multiple, AFBC planes. As with
component order, the encoder and decoder must agree about the number
of planes in order to correctly decode the buffer. The fourcc code is
used to determine the number of encoded planes in an AFBC buffer,
matching the number of planes for the linear (unmodified) format.
Within each plane, the component ordering also follows the fourcc
code:
For example:
* DRM_FORMAT_YUYV: nplanes = 1
* Plane 0:
* Component 0: Y(8)
* Component 1: Cb(8, 2x1 subsampled)
* Component 2: Cr(8, 2x1 subsampled)
* DRM_FORMAT_NV12: nplanes = 2
* Plane 0:
* Component 0: Y(8)
* Plane 1:
* Component 0: Cb(8, 2x1 subsampled)
* Component 1: Cr(8, 2x1 subsampled)
Cross-device interoperability
=============================
For maximum compatibility across devices, the table below defines
canonical formats for use between AFBC-enabled devices. Formats which
are listed here must be used exactly as specified when using the AFBC
modifiers. Formats which are not listed should be avoided.
.. flat-table:: AFBC formats
* - Fourcc code
- Description
- Planes/Components
* - DRM_FORMAT_ABGR2101010
- 10-bit per component RGB, with 2-bit alpha
- Plane 0: 4 components
* Component 0: R(10)
* Component 1: G(10)
* Component 2: B(10)
* Component 3: A(2)
* - DRM_FORMAT_ABGR8888
- 8-bit per component RGB, with 8-bit alpha
- Plane 0: 4 components
* Component 0: R(8)
* Component 1: G(8)
* Component 2: B(8)
* Component 3: A(8)
* - DRM_FORMAT_BGR888
- 8-bit per component RGB
- Plane 0: 3 components
* Component 0: R(8)
* Component 1: G(8)
* Component 2: B(8)
* - DRM_FORMAT_BGR565
- 5/6-bit per component RGB
- Plane 0: 3 components
* Component 0: R(5)
* Component 1: G(6)
* Component 2: B(5)
* - DRM_FORMAT_ABGR1555
- 5-bit per component RGB, with 1-bit alpha
- Plane 0: 4 components
* Component 0: R(5)
* Component 1: G(5)
* Component 2: B(5)
* Component 3: A(1)
* - DRM_FORMAT_VUY888
- 8-bit per component YCbCr 444, single plane
- Plane 0: 3 components
* Component 0: Y(8)
* Component 1: Cb(8)
* Component 2: Cr(8)
* - DRM_FORMAT_VUY101010
- 10-bit per component YCbCr 444, single plane
- Plane 0: 3 components
* Component 0: Y(10)
* Component 1: Cb(10)
* Component 2: Cr(10)
* - DRM_FORMAT_YUYV
- 8-bit per component YCbCr 422, single plane
- Plane 0: 3 components
* Component 0: Y(8)
* Component 1: Cb(8, 2x1 subsampled)
* Component 2: Cr(8, 2x1 subsampled)
* - DRM_FORMAT_NV16
- 8-bit per component YCbCr 422, two plane
- Plane 0: 1 component
* Component 0: Y(8)
Plane 1: 2 components
* Component 0: Cb(8, 2x1 subsampled)
* Component 1: Cr(8, 2x1 subsampled)
* - DRM_FORMAT_Y210
- 10-bit per component YCbCr 422, single plane
- Plane 0: 3 components
* Component 0: Y(10)
* Component 1: Cb(10, 2x1 subsampled)
* Component 2: Cr(10, 2x1 subsampled)
* - DRM_FORMAT_P210
- 10-bit per component YCbCr 422, two plane
- Plane 0: 1 component
* Component 0: Y(10)
Plane 1: 2 components
* Component 0: Cb(10, 2x1 subsampled)
* Component 1: Cr(10, 2x1 subsampled)
* - DRM_FORMAT_YUV420_8BIT
- 8-bit per component YCbCr 420, single plane
- Plane 0: 3 components
* Component 0: Y(8)
* Component 1: Cb(8, 2x2 subsampled)
* Component 2: Cr(8, 2x2 subsampled)
* - DRM_FORMAT_YUV420_10BIT
- 10-bit per component YCbCr 420, single plane
- Plane 0: 3 components
* Component 0: Y(10)
* Component 1: Cb(10, 2x2 subsampled)
* Component 2: Cr(10, 2x2 subsampled)
* - DRM_FORMAT_NV12
- 8-bit per component YCbCr 420, two plane
- Plane 0: 1 component
* Component 0: Y(8)
Plane 1: 2 components
* Component 0: Cb(8, 2x2 subsampled)
* Component 1: Cr(8, 2x2 subsampled)
* - DRM_FORMAT_P010
- 10-bit per component YCbCr 420, two plane
- Plane 0: 1 component
* Component 0: Y(10)
Plane 1: 2 components
* Component 0: Cb(10, 2x2 subsampled)
* Component 1: Cr(10, 2x2 subsampled)

52
Documentation/gpu/dp-mst/topology-figure-1.dot Normal file
View File

@ -0,0 +1,52 @@
digraph T {
/* Make sure our payloads are always drawn below the driver node */
subgraph cluster_driver {
fillcolor = grey;
style = filled;
driver -> {payload1, payload2} [dir=none];
}
/* Driver malloc references */
edge [style=dashed];
driver -> port1;
driver -> port2;
driver -> port3:e;
driver -> port4;
payload1:s -> port1:e;
payload2:s -> port3:e;
edge [style=""];
subgraph cluster_topology {
label="Topology Manager";
labelloc=bottom;
/* Topology references */
mstb1 -> {port1, port2};
port1 -> mstb2;
port2 -> mstb3 -> {port3, port4};
port3 -> mstb4;
/* Malloc references */
edge [style=dashed;dir=back];
mstb1 -> {port1, port2};
port1 -> mstb2;
port2 -> mstb3 -> {port3, port4};
port3 -> mstb4;
}
driver [label="DRM driver";style=filled;shape=box;fillcolor=lightblue];
payload1 [label="Payload #1";style=filled;shape=box;fillcolor=lightblue];
payload2 [label="Payload #2";style=filled;shape=box;fillcolor=lightblue];
mstb1 [label="MSTB #1";style=filled;fillcolor=palegreen;shape=oval];
mstb2 [label="MSTB #2";style=filled;fillcolor=palegreen;shape=oval];
mstb3 [label="MSTB #3";style=filled;fillcolor=palegreen;shape=oval];
mstb4 [label="MSTB #4";style=filled;fillcolor=palegreen;shape=oval];
port1 [label="Port #1";shape=oval];
port2 [label="Port #2";shape=oval];
port3 [label="Port #3";shape=oval];
port4 [label="Port #4";shape=oval];
}

56
Documentation/gpu/dp-mst/topology-figure-2.dot Normal file
View File

@ -0,0 +1,56 @@
digraph T {
/* Make sure our payloads are always drawn below the driver node */
subgraph cluster_driver {
fillcolor = grey;
style = filled;
driver -> {payload1, payload2} [dir=none];
}
/* Driver malloc references */
edge [style=dashed];
driver -> port1;
driver -> port2;
driver -> port3:e;
driver -> port4 [color=red];
payload1:s -> port1:e;
payload2:s -> port3:e;
edge [style=""];
subgraph cluster_topology {
label="Topology Manager";
labelloc=bottom;
/* Topology references */
mstb1 -> {port1, port2};
port1 -> mstb2;
edge [color=red];
port2 -> mstb3 -> {port3, port4};
port3 -> mstb4;
edge [color=""];
/* Malloc references */
edge [style=dashed;dir=back];
mstb1 -> {port1, port2};
port1 -> mstb2;
port2 -> mstb3 -> port3;
edge [color=red];
mstb3 -> port4;
port3 -> mstb4;
}
mstb1 [label="MSTB #1";style=filled;fillcolor=palegreen];
mstb2 [label="MSTB #2";style=filled;fillcolor=palegreen];
mstb3 [label="MSTB #3";style=filled;fillcolor=palegreen];
mstb4 [label="MSTB #4";style=filled;fillcolor=grey];
port1 [label="Port #1"];
port2 [label="Port #2"];
port3 [label="Port #3"];
port4 [label="Port #4";style=filled;fillcolor=grey];
driver [label="DRM driver";style=filled;shape=box;fillcolor=lightblue];
payload1 [label="Payload #1";style=filled;shape=box;fillcolor=lightblue];
payload2 [label="Payload #2";style=filled;shape=box;fillcolor=lightblue];
}

59
Documentation/gpu/dp-mst/topology-figure-3.dot Normal file
View File

@ -0,0 +1,59 @@
digraph T {
/* Make sure our payloads are always drawn below the driver node */
subgraph cluster_driver {
fillcolor = grey;
style = filled;
edge [dir=none];
driver -> payload1;
driver -> payload2 [penwidth=3];
edge [dir=""];
}
/* Driver malloc references */
edge [style=dashed];
driver -> port1;
driver -> port2;
driver -> port3:e;
driver -> port4 [color=grey];
payload1:s -> port1:e;
payload2:s -> port3:e [penwidth=3];
edge [style=""];
subgraph cluster_topology {
label="Topology Manager";
labelloc=bottom;
/* Topology references */
mstb1 -> {port1, port2};
port1 -> mstb2;
edge [color=grey];
port2 -> mstb3 -> {port3, port4};
port3 -> mstb4;
edge [color=""];
/* Malloc references */
edge [style=dashed;dir=back];
mstb1 -> {port1, port2};
port1 -> mstb2;
port2 -> mstb3 [penwidth=3];
mstb3 -> port3 [penwidth=3];
edge [color=grey];
mstb3 -> port4;
port3 -> mstb4;
}
mstb1 [label="MSTB #1";style=filled;fillcolor=palegreen];
mstb2 [label="MSTB #2";style=filled;fillcolor=palegreen];
mstb3 [label="MSTB #3";style=filled;fillcolor=palegreen;penwidth=3];
mstb4 [label="MSTB #4";style=filled;fillcolor=grey];
port1 [label="Port #1"];
port2 [label="Port #2";penwidth=5];
port3 [label="Port #3";penwidth=3];
port4 [label="Port #4";style=filled;fillcolor=grey];
driver [label="DRM driver";style=filled;shape=box;fillcolor=lightblue];
payload1 [label="Payload #1";style=filled;shape=box;fillcolor=lightblue];
payload2 [label="Payload #2";style=filled;shape=box;fillcolor=lightblue;penwidth=3];
}

2
Documentation/gpu/drivers.rst
View File

@ -17,6 +17,8 @@ GPU Driver Documentation
vkms
bridge/dw-hdmi
xen-front
afbc
komeda-kms
.. only:: subproject and html

74
Documentation/gpu/drm-internals.rst
View File

@ -39,68 +39,6 @@ sections.
Driver Information
------------------
Driver Features
~~~~~~~~~~~~~~~
Drivers inform the DRM core about their requirements and supported
features by setting appropriate flags in the driver_features field.
Since those flags influence the DRM core behaviour since registration
time, most of them must be set to registering the :c:type:`struct
drm_driver <drm_driver>` instance.
u32 driver_features;
DRIVER_USE_AGP
Driver uses AGP interface, the DRM core will manage AGP resources.
DRIVER_LEGACY
Denote a legacy driver using shadow attach. Don't use.
DRIVER_KMS_LEGACY_CONTEXT
Used only by nouveau for backwards compatibility with existing userspace.
Don't use.
DRIVER_PCI_DMA
Driver is capable of PCI DMA, mapping of PCI DMA buffers to
userspace will be enabled. Deprecated.
DRIVER_SG
Driver can perform scatter/gather DMA, allocation and mapping of
scatter/gather buffers will be enabled. Deprecated.
DRIVER_HAVE_DMA
Driver supports DMA, the userspace DMA API will be supported.
Deprecated.
DRIVER_HAVE_IRQ; DRIVER_IRQ_SHARED
DRIVER_HAVE_IRQ indicates whether the driver has an IRQ handler
managed by the DRM Core. The core will support simple IRQ handler
installation when the flag is set. The installation process is
described in ?.
DRIVER_IRQ_SHARED indicates whether the device & handler support
shared IRQs (note that this is required of PCI drivers).
DRIVER_GEM
Driver use the GEM memory manager.
DRIVER_MODESET
Driver supports mode setting interfaces (KMS).
DRIVER_PRIME
Driver implements DRM PRIME buffer sharing.
DRIVER_RENDER
Driver supports dedicated render nodes.
DRIVER_ATOMIC
Driver supports atomic properties. In this case the driver must
implement appropriate obj->atomic_get_property() vfuncs for any
modeset objects with driver specific properties.
DRIVER_SYNCOBJ
Driver support drm sync objects.
Major, Minor and Patchlevel
~~~~~~~~~~~~~~~~~~~~~~~~~~~
@ -143,6 +81,9 @@ Device Instance and Driver Handling
.. kernel-doc:: drivers/gpu/drm/drm_drv.c
:doc: driver instance overview
.. kernel-doc:: include/drm/drm_device.h
:internal:
.. kernel-doc:: include/drm/drm_drv.h
:internal:
@ -230,6 +171,15 @@ Printer
.. kernel-doc:: drivers/gpu/drm/drm_print.c
:export:
Utilities
---------
.. kernel-doc:: include/drm/drm_util.h
:doc: drm utils
.. kernel-doc:: include/drm/drm_util.h
:internal:
Legacy Support Code
===================

54
Documentation/gpu/drm-kms-helpers.rst
View File

@ -116,8 +116,6 @@ Framebuffer CMA Helper Functions Reference
.. kernel-doc:: drivers/gpu/drm/drm_fb_cma_helper.c
:export:
.. _drm_bridges:
Framebuffer GEM Helper Reference
================================
@ -127,6 +125,8 @@ Framebuffer GEM Helper Reference
.. kernel-doc:: drivers/gpu/drm/drm_gem_framebuffer_helper.c
:export:
.. _drm_bridges:
Bridges
=======
@ -208,18 +208,40 @@ Display Port Dual Mode Adaptor Helper Functions Reference
.. kernel-doc:: drivers/gpu/drm/drm_dp_dual_mode_helper.c
:export:
Display Port MST Helper Functions Reference
===========================================
Display Port MST Helpers
========================
Overview
--------
.. kernel-doc:: drivers/gpu/drm/drm_dp_mst_topology.c
:doc: dp mst helper
.. kernel-doc:: drivers/gpu/drm/drm_dp_mst_topology.c
:doc: Branch device and port refcounting
Functions Reference
-------------------
.. kernel-doc:: include/drm/drm_dp_mst_helper.h
:internal:
.. kernel-doc:: drivers/gpu/drm/drm_dp_mst_topology.c
:export:
Topology Lifetime Internals
---------------------------
These functions aren't exported to drivers, but are documented here to help make
the MST topology helpers easier to understand
.. kernel-doc:: drivers/gpu/drm/drm_dp_mst_topology.c
:functions: drm_dp_mst_topology_try_get_mstb drm_dp_mst_topology_get_mstb
drm_dp_mst_topology_put_mstb
drm_dp_mst_topology_try_get_port drm_dp_mst_topology_get_port
drm_dp_mst_topology_put_port
drm_dp_mst_get_mstb_malloc drm_dp_mst_put_mstb_malloc
MIPI DSI Helper Functions Reference
===================================
@ -274,18 +296,6 @@ SCDC Helper Functions Reference
.. kernel-doc:: drivers/gpu/drm/drm_scdc_helper.c
:export:
Rectangle Utilities Reference
=============================
.. kernel-doc:: include/drm/drm_rect.h
:doc: rect utils
.. kernel-doc:: include/drm/drm_rect.h
:internal:
.. kernel-doc:: drivers/gpu/drm/drm_rect.c
:export:
HDMI Infoframes Helper Reference
================================
@ -300,6 +310,18 @@ libraries and hence is also included here.
.. kernel-doc:: drivers/video/hdmi.c
:export:
Rectangle Utilities Reference
=============================
.. kernel-doc:: include/drm/drm_rect.h
:doc: rect utils
.. kernel-doc:: include/drm/drm_rect.h
:internal:
.. kernel-doc:: drivers/gpu/drm/drm_rect.c
:export:
Flip-work Helper Reference
==========================

96
Documentation/gpu/drm-kms.rst
View File

@ -410,102 +410,6 @@ Encoder Functions Reference
.. kernel-doc:: drivers/gpu/drm/drm_encoder.c
:export:
KMS Initialization and Cleanup
==============================
A KMS device is abstracted and exposed as a set of planes, CRTCs,
encoders and connectors. KMS drivers must thus create and initialize all
those objects at load time after initializing mode setting.
CRTCs (:c:type:`struct drm_crtc <drm_crtc>`)
--------------------------------------------
A CRTC is an abstraction representing a part of the chip that contains a
pointer to a scanout buffer. Therefore, the number of CRTCs available
determines how many independent scanout buffers can be active at any
given time. The CRTC structure contains several fields to support this:
a pointer to some video memory (abstracted as a frame buffer object), a
display mode, and an (x, y) offset into the video memory to support
panning or configurations where one piece of video memory spans multiple
CRTCs.
CRTC Initialization
~~~~~~~~~~~~~~~~~~~
A KMS device must create and register at least one struct
:c:type:`struct drm_crtc <drm_crtc>` instance. The instance is
allocated and zeroed by the driver, possibly as part of a larger
structure, and registered with a call to :c:func:`drm_crtc_init()`
with a pointer to CRTC functions.
Cleanup
-------
The DRM core manages its objects' lifetime. When an object is not needed
anymore the core calls its destroy function, which must clean up and
free every resource allocated for the object. Every
:c:func:`drm_\*_init()` call must be matched with a corresponding
:c:func:`drm_\*_cleanup()` call to cleanup CRTCs
(:c:func:`drm_crtc_cleanup()`), planes
(:c:func:`drm_plane_cleanup()`), encoders
(:c:func:`drm_encoder_cleanup()`) and connectors
(:c:func:`drm_connector_cleanup()`). Furthermore, connectors that
have been added to sysfs must be removed by a call to
:c:func:`drm_connector_unregister()` before calling
:c:func:`drm_connector_cleanup()`.
Connectors state change detection must be cleanup up with a call to
:c:func:`drm_kms_helper_poll_fini()`.
Output discovery and initialization example
-------------------------------------------
.. code-block:: c
void intel_crt_init(struct drm_device *dev)
{
struct drm_connector *connector;
struct intel_output *intel_output;
intel_output = kzalloc(sizeof(struct intel_output), GFP_KERNEL);
if (!intel_output)
return;
connector = &intel_output->base;
drm_connector_init(dev, &intel_output->base,
&intel_crt_connector_funcs, DRM_MODE_CONNECTOR_VGA);
drm_encoder_init(dev, &intel_output->enc, &intel_crt_enc_funcs,
DRM_MODE_ENCODER_DAC);
drm_connector_attach_encoder(&intel_output->base,
&intel_output->enc);
/* Set up the DDC bus. */
intel_output->ddc_bus = intel_i2c_create(dev, GPIOA, "CRTDDC_A");
if (!intel_output->ddc_bus) {
dev_printk(KERN_ERR, &dev->pdev->dev, "DDC bus registration "
"failed.\n");
return;
}
intel_output->type = INTEL_OUTPUT_ANALOG;
connector->interlace_allowed = 0;
connector->doublescan_allowed = 0;
drm_encoder_helper_add(&intel_output->enc, &intel_crt_helper_funcs);
drm_connector_helper_add(connector, &intel_crt_connector_helper_funcs);
drm_connector_register(connector);
}
In the example above (taken from the i915 driver), a CRTC, connector and
encoder combination is created. A device-specific i2c bus is also
created for fetching EDID data and performing monitor detection. Once
the process is complete, the new connector is registered with sysfs to
make its properties available to applications.
KMS Locking
===========

8
Documentation/gpu/drm-uapi.rst
View File

@ -238,6 +238,14 @@ DRM specific patterns. Note that ENOTTY has the slightly unintuitive meaning of
Testing and validation
======================
Testing Requirements for userspace API
--------------------------------------
New cross-driver userspace interface extensions, like new IOCTL, new KMS
properties, new files in sysfs or anything else that constitutes an API change
should have driver-agnostic testcases in IGT for that feature, if such a test
can be reasonably made using IGT for the target hardware.
Validating changes with IGT
---------------------------

488
Documentation/gpu/komeda-kms.rst Normal file
View File

@ -0,0 +1,488 @@
.. SPDX-License-Identifier: GPL-2.0
==============================
drm/komeda Arm display driver
==============================
The drm/komeda driver supports the Arm display processor D71 and later products,
this document gives a brief overview of driver design: how it works and why
design it like that.
Overview of D71 like display IPs
================================
From D71, Arm display IP begins to adopt a flexible and modularized
architecture. A display pipeline is made up of multiple individual and
functional pipeline stages called components, and every component has some
specific capabilities that can give the flowed pipeline pixel data a
particular processing.
Typical D71 components:
Layer
-----
Layer is the first pipeline stage, which prepares the pixel data for the next
stage. It fetches the pixel from memory, decodes it if it's AFBC, rotates the
source image, unpacks or converts YUV pixels to the device internal RGB pixels,
then adjusts the color_space of pixels if needed.
Scaler
------
As its name suggests, scaler takes responsibility for scaling, and D71 also
supports image enhancements by scaler.
The usage of scaler is very flexible and can be connected to layer output
for layer scaling, or connected to compositor and scale the whole display
frame and then feed the output data into wb_layer which will then write it
into memory.
Compositor (compiz)
-------------------
Compositor blends multiple layers or pixel data flows into one single display
frame. its output frame can be fed into post image processor for showing it on
the monitor or fed into wb_layer and written to memory at the same time.
user can also insert a scaler between compositor and wb_layer to down scale
the display frame first and and then write to memory.
Writeback Layer (wb_layer)
--------------------------
Writeback layer does the opposite things of Layer, which connects to compiz
and writes the composition result to memory.
Post image processor (improc)
-----------------------------
Post image processor adjusts frame data like gamma and color space to fit the
requirements of the monitor.
Timing controller (timing_ctrlr)
--------------------------------
Final stage of display pipeline, Timing controller is not for the pixel
handling, but only for controlling the display timing.
Merger
------
D71 scaler mostly only has the half horizontal input/output capabilities
compared with Layer, like if Layer supports 4K input size, the scaler only can
support 2K input/output in the same time. To achieve the ful frame scaling, D71
introduces Layer Split, which splits the whole image to two half parts and feeds
them to two Layers A and B, and does the scaling independently. After scaling
the result need to be fed to merger to merge two part images together, and then
output merged result to compiz.
Splitter
--------
Similar to Layer Split, but Splitter is used for writeback, which splits the
compiz result to two parts and then feed them to two scalers.
Possible D71 Pipeline usage
===========================
Benefitting from the modularized architecture, D71 pipelines can be easily
adjusted to fit different usages. And D71 has two pipelines, which support two
types of working mode:
- Dual display mode
Two pipelines work independently and separately to drive two display outputs.
- Single display mode
Two pipelines work together to drive only one display output.
On this mode, pipeline_B doesn't work indenpendently, but outputs its
composition result into pipeline_A, and its pixel timing also derived from
pipeline_A.timing_ctrlr. The pipeline_B works just like a "slave" of
pipeline_A(master)
Single pipeline data flow
-------------------------
.. kernel-render:: DOT
:alt: Single pipeline digraph
:caption: Single pipeline data flow
digraph single_ppl {
rankdir=LR;
subgraph {
"Memory";
"Monitor";
}
subgraph cluster_pipeline {
style=dashed
node [shape=box]
{
node [bgcolor=grey style=dashed]
"Scaler-0";
"Scaler-1";
"Scaler-0/1"
}
node [bgcolor=grey style=filled]
"Layer-0" -> "Scaler-0"
"Layer-1" -> "Scaler-0"
"Layer-2" -> "Scaler-1"
"Layer-3" -> "Scaler-1"
"Layer-0" -> "Compiz"
"Layer-1" -> "Compiz"
"Layer-2" -> "Compiz"
"Layer-3" -> "Compiz"
"Scaler-0" -> "Compiz"
"Scaler-1" -> "Compiz"
"Compiz" -> "Scaler-0/1" -> "Wb_layer"
"Compiz" -> "Improc" -> "Timing Controller"
}
"Wb_layer" -> "Memory"
"Timing Controller" -> "Monitor"
}
Dual pipeline with Slave enabled
--------------------------------
.. kernel-render:: DOT
:alt: Slave pipeline digraph
:caption: Slave pipeline enabled data flow
digraph slave_ppl {
rankdir=LR;
subgraph {
"Memory";
"Monitor";
}
node [shape=box]
subgraph cluster_pipeline_slave {
style=dashed
label="Slave Pipeline_B"
node [shape=box]
{
node [bgcolor=grey style=dashed]
"Slave.Scaler-0";
"Slave.Scaler-1";
}
node [bgcolor=grey style=filled]
"Slave.Layer-0" -> "Slave.Scaler-0"
"Slave.Layer-1" -> "Slave.Scaler-0"
"Slave.Layer-2" -> "Slave.Scaler-1"
"Slave.Layer-3" -> "Slave.Scaler-1"
"Slave.Layer-0" -> "Slave.Compiz"
"Slave.Layer-1" -> "Slave.Compiz"
"Slave.Layer-2" -> "Slave.Compiz"
"Slave.Layer-3" -> "Slave.Compiz"
"Slave.Scaler-0" -> "Slave.Compiz"
"Slave.Scaler-1" -> "Slave.Compiz"
}
subgraph cluster_pipeline_master {
style=dashed
label="Master Pipeline_A"
node [shape=box]
{
node [bgcolor=grey style=dashed]
"Scaler-0";
"Scaler-1";
"Scaler-0/1"
}
node [bgcolor=grey style=filled]
"Layer-0" -> "Scaler-0"
"Layer-1" -> "Scaler-0"
"Layer-2" -> "Scaler-1"
"Layer-3" -> "Scaler-1"
"Slave.Compiz" -> "Compiz"
"Layer-0" -> "Compiz"
"Layer-1" -> "Compiz"
"Layer-2" -> "Compiz"
"Layer-3" -> "Compiz"
"Scaler-0" -> "Compiz"
"Scaler-1" -> "Compiz"
"Compiz" -> "Scaler-0/1" -> "Wb_layer"
"Compiz" -> "Improc" -> "Timing Controller"
}
"Wb_layer" -> "Memory"
"Timing Controller" -> "Monitor"
}
Sub-pipelines for input and output
----------------------------------
A complete display pipeline can be easily divided into three sub-pipelines
according to the in/out usage.
Layer(input) pipeline
~~~~~~~~~~~~~~~~~~~~~
.. kernel-render:: DOT
:alt: Layer data digraph
:caption: Layer (input) data flow
digraph layer_data_flow {
rankdir=LR;
node [shape=box]
{
node [bgcolor=grey style=dashed]
"Scaler-n";
}
"Layer-n" -> "Scaler-n" -> "Compiz"
}
.. kernel-render:: DOT
:alt: Layer Split digraph
:caption: Layer Split pipeline
digraph layer_data_flow {
rankdir=LR;
node [shape=box]
"Layer-0/1" -> "Scaler-0" -> "Merger"
"Layer-2/3" -> "Scaler-1" -> "Merger"
"Merger" -> "Compiz"
}
Writeback(output) pipeline
~~~~~~~~~~~~~~~~~~~~~~~~~~
.. kernel-render:: DOT
:alt: writeback digraph
:caption: Writeback(output) data flow
digraph writeback_data_flow {
rankdir=LR;
node [shape=box]
{
node [bgcolor=grey style=dashed]
"Scaler-n";
}
"Compiz" -> "Scaler-n" -> "Wb_layer"
}
.. kernel-render:: DOT
:alt: split writeback digraph
:caption: Writeback(output) Split data flow
digraph writeback_data_flow {
rankdir=LR;
node [shape=box]
"Compiz" -> "Splitter"
"Splitter" -> "Scaler-0" -> "Merger"
"Splitter" -> "Scaler-1" -> "Merger"
"Merger" -> "Wb_layer"
}
Display output pipeline
~~~~~~~~~~~~~~~~~~~~~~~
.. kernel-render:: DOT
:alt: display digraph
:caption: display output data flow
digraph single_ppl {
rankdir=LR;
node [shape=box]
"Compiz" -> "Improc" -> "Timing Controller"
}
In the following section we'll see these three sub-pipelines will be handled
by KMS-plane/wb_conn/crtc respectively.
Komeda Resource abstraction
===========================
struct komeda_pipeline/component
--------------------------------
To fully utilize and easily access/configure the HW, the driver side also uses
a similar architecture: Pipeline/Component to describe the HW features and
capabilities, and a specific component includes two parts:
- Data flow controlling.
- Specific component capabilities and features.
So the driver defines a common header struct komeda_component to describe the
data flow control and all specific components are a subclass of this base
structure.
.. kernel-doc:: drivers/gpu/drm/arm/display/komeda/komeda_pipeline.h
:internal:
Resource discovery and initialization
=====================================
Pipeline and component are used to describe how to handle the pixel data. We
still need a @struct komeda_dev to describe the whole view of the device, and
the control-abilites of device.
We have &komeda_dev, &komeda_pipeline, &komeda_component. Now fill devices with
pipelines. Since komeda is not for D71 only but also intended for later products,
of course we’d better share as much as possible between different products. To
achieve this, split the komeda device into two layers: CORE and CHIP.
- CORE: for common features and capabilities handling.
- CHIP: for register programing and HW specific feature (limitation) handling.
CORE can access CHIP by three chip function structures:
- struct komeda_dev_funcs
- struct komeda_pipeline_funcs
- struct komeda_component_funcs
.. kernel-doc:: drivers/gpu/drm/arm/display/komeda/komeda_dev.h
:internal:
Format handling
===============
.. kernel-doc:: drivers/gpu/drm/arm/display/komeda/komeda_format_caps.h
:internal:
.. kernel-doc:: drivers/gpu/drm/arm/display/komeda/komeda_framebuffer.h
:internal:
Attach komeda_dev to DRM-KMS
============================
Komeda abstracts resources by pipeline/component, but DRM-KMS uses
crtc/plane/connector. One KMS-obj cannot represent only one single component,
since the requirements of a single KMS object cannot simply be achieved by a
single component, usually that needs multiple components to fit the requirement.
Like set mode, gamma, ctm for KMS all target on CRTC-obj, but komeda needs
compiz, improc and timing_ctrlr to work together to fit these requirements.
And a KMS-Plane may require multiple komeda resources: layer/scaler/compiz.
So, one KMS-Obj represents a sub-pipeline of komeda resources.
- Plane: `Layer(input) pipeline`_
- Wb_connector: `Writeback(output) pipeline`_
- Crtc: `Display output pipeline`_
So, for komeda, we treat KMS crtc/plane/connector as users of pipeline and
component, and at any one time a pipeline/component only can be used by one
user. And pipeline/component will be treated as private object of DRM-KMS; the
state will be managed by drm_atomic_state as well.
How to map plane to Layer(input) pipeline
-----------------------------------------
Komeda has multiple Layer input pipelines, see:
- `Single pipeline data flow`_
- `Dual pipeline with Slave enabled`_
The easiest way is binding a plane to a fixed Layer pipeline, but consider the
komeda capabilities:
- Layer Split, See `Layer(input) pipeline`_
Layer_Split is quite complicated feature, which splits a big image into two
parts and handles it by two layers and two scalers individually. But it
imports an edge problem or effect in the middle of the image after the split.
To avoid such a problem, it needs a complicated Split calculation and some
special configurations to the layer and scaler. We'd better hide such HW
related complexity to user mode.
- Slave pipeline, See `Dual pipeline with Slave enabled`_
Since the compiz component doesn't output alpha value, the slave pipeline
only can be used for bottom layers composition. The komeda driver wants to
hide this limitation to the user. The way to do this is to pick a suitable
Layer according to plane_state->zpos.
So for komeda, the KMS-plane doesn't represent a fixed komeda layer pipeline,
but multiple Layers with same capabilities. Komeda will select one or more
Layers to fit the requirement of one KMS-plane.
Make component/pipeline to be drm_private_obj
---------------------------------------------
Add :c:type:`drm_private_obj` to :c:type:`komeda_component`, :c:type:`komeda_pipeline`
.. code-block:: c
struct komeda_component {
struct drm_private_obj obj;
...
}
struct komeda_pipeline {
struct drm_private_obj obj;
...
}
Tracking component_state/pipeline_state by drm_atomic_state
-----------------------------------------------------------
Add :c:type:`drm_private_state` and user to :c:type:`komeda_component_state`,
:c:type:`komeda_pipeline_state`
.. code-block:: c
struct komeda_component_state {
struct drm_private_state obj;
void *binding_user;
...
}
struct komeda_pipeline_state {
struct drm_private_state obj;
struct drm_crtc *crtc;
...
}
komeda component validation
---------------------------
Komeda has multiple types of components, but the process of validation are
similar, usually including the following steps:
.. code-block:: c
int komeda_xxxx_validate(struct komeda_component_xxx xxx_comp,
struct komeda_component_output *input_dflow,
struct drm_plane/crtc/connector *user,
struct drm_plane/crtc/connector_state, *user_state)
{
setup 1: check if component is needed, like the scaler is optional depending
on the user_state; if unneeded, just return, and the caller will
put the data flow into next stage.
Setup 2: check user_state with component features and capabilities to see
if requirements can be met; if not, return fail.
Setup 3: get component_state from drm_atomic_state, and try set to set
user to component; fail if component has been assigned to another
user already.
Setup 3: configure the component_state, like set its input component,
convert user_state to component specific state.
Setup 4: adjust the input_dflow and prepare it for the next stage.
}
komeda_kms Abstraction
----------------------
.. kernel-doc:: drivers/gpu/drm/arm/display/komeda/komeda_kms.h
:internal:
komde_kms Functions
-------------------
.. kernel-doc:: drivers/gpu/drm/arm/display/komeda/komeda_crtc.c
:internal:
.. kernel-doc:: drivers/gpu/drm/arm/display/komeda/komeda_plane.c
:internal:
Build komeda to be a Linux module driver
========================================
Now we have two level devices:
- komeda_dev: describes the real display hardware.
- komeda_kms_dev: attachs or connects komeda_dev to DRM-KMS.
All komeda operations are supplied or operated by komeda_dev or komeda_kms_dev,
the module driver is only a simple wrapper to pass the Linux command
(probe/remove/pm) into komeda_dev or komeda_kms_dev.

110
Documentation/gpu/todo.rst
View File

@ -82,30 +82,6 @@ events for atomic commits correctly. But fixing these bugs is good anyway.
Contact: Daniel Vetter, respective driver maintainers
Better manual-upload support for atomic
---------------------------------------
This would be especially useful for tinydrm:
- Add a struct drm_rect dirty_clip to drm_crtc_state. When duplicating the
crtc state, clear that to the max values, x/y = 0 and w/h = MAX_INT, in
__drm_atomic_helper_crtc_duplicate_state().
- Move tinydrm_merge_clips into drm_framebuffer.c, dropping the tinydrm\_
prefix ofc and using drm_fb\_. drm_framebuffer.c makes sense since this
is a function useful to implement the fb->dirty function.
- Create a new drm_fb_dirty function which does essentially what e.g.
mipi_dbi_fb_dirty does. You can use e.g. drm_atomic_helper_update_plane as the
template. But instead of doing a simple full-screen plane update, this new
helper also sets crtc_state->dirty_clip to the right coordinates. And of
course it needs to check whether the fb is actually active (and maybe where),
so there's some book-keeping involved. There's also some good fun involved in
scaling things appropriately. For that case we might simply give up and
declare the entire area covered by the plane as dirty.
Contact: Noralf Trønnes, Daniel Vetter
Fallout from atomic KMS
-----------------------
@ -209,6 +185,36 @@ Would be great to refactor this all into a set of small common helpers.
Contact: Daniel Vetter
Generic fbdev defio support
---------------------------
The defio support code in the fbdev core has some very specific requirements,
which means drivers need to have a special framebuffer for fbdev. Which prevents
us from using the generic fbdev emulation code everywhere. The main issue is
that it uses some fields in struct page itself, which breaks shmem gem objects
(and other things).
Possible solution would be to write our own defio mmap code in the drm fbdev
emulation. It would need to fully wrap the existing mmap ops, forwarding
everything after it has done the write-protect/mkwrite trickery:
- In the drm_fbdev_fb_mmap helper, if we need defio, change the
default page prots to write-protected with something like this::
vma->vm_page_prot = pgprot_wrprotect(vma->vm_page_prot);
- Set the mkwrite and fsync callbacks with similar implementions to the core
fbdev defio stuff. These should all work on plain ptes, they don't actually
require a struct page. uff. These should all work on plain ptes, they don't
actually require a struct page.
- Track the dirty pages in a separate structure (bitfield with one bit per page
should work) to avoid clobbering struct page.
Might be good to also have some igt testcases for this.
Contact: Daniel Vetter, Noralf Tronnes
Put a reservation_object into drm_gem_object
--------------------------------------------
@ -256,6 +262,44 @@ As a reference, take a look at the conversions already completed in drm core.
Contact: Sean Paul, respective driver maintainers
Rename CMA helpers to DMA helpers
---------------------------------
CMA (standing for contiguous memory allocator) is really a bit an accident of
what these were used for first, a much better name would be DMA helpers. In the
text these should even be called coherent DMA memory helpers (so maybe CDM, but
no one knows what that means) since underneath they just use dma_alloc_coherent.
Contact: Laurent Pinchart, Daniel Vetter
Convert direct mode.vrefresh accesses to use drm_mode_vrefresh()
----------------------------------------------------------------
drm_display_mode.vrefresh isn't guaranteed to be populated. As such, using it
is risky and has been known to cause div-by-zero bugs. Fortunately, drm core
has helper which will use mode.vrefresh if it's !0 and will calculate it from
the timings when it's 0.
Use simple search/replace, or (more fun) cocci to replace instances of direct
vrefresh access with a call to the helper. Check out
https://lists.freedesktop.org/archives/dri-devel/2019-January/205186.html for
inspiration.
Once all instances of vrefresh have been converted, remove vrefresh from
drm_display_mode to avoid future use.
Contact: Sean Paul
Remove drm_display_mode.hsync
-----------------------------
We have drm_mode_hsync() to calculate this from hsync_start/end, since drivers
shouldn't/don't use this, remove this member to avoid any temptations to use it
in the future. If there is any debug code using drm_display_mode.hsync, convert
it to use drm_mode_hsync() instead.
Contact: Sean Paul
Core refactorings
=================
@ -354,13 +398,6 @@ KMS cleanups
Some of these date from the very introduction of KMS in 2008 ...
- drm_mode_config.crtc_idr is misnamed, since it contains all KMS object. Should
be renamed to drm_mode_config.object_idr.
- drm_display_mode doesn't need to be derived from drm_mode_object. That's
leftovers from older (never merged into upstream) KMS designs where modes
where set using their ID, including support to add/remove modes.
- Make ->funcs and ->helper_private vtables optional. There's a bunch of empty
function tables in drivers, but before we can remove them we need to make sure
that all the users in helpers and drivers do correctly check for a NULL
@ -432,21 +469,10 @@ those drivers as simple as possible, so lots of room for refactoring:
one of the ideas for having a shared dsi/dbi helper, abstracting away the
transport details more.
- tinydrm_gem_cma_prime_import_sg_table should probably go into the cma
helpers, as a _vmapped variant (since not every driver needs the vmap).
And tinydrm_gem_cma_free_object could the be merged into
drm_gem_cma_free_object().
- tinydrm_fb_create we could move into drm_simple_pipe, only need to add
the fb_create hook to drm_simple_pipe_funcs, which would again simplify a
bunch of things (since it gives you a one-stop vfunc for simple drivers).
- Quick aside: The unregister devm stuff is kinda getting the lifetimes of
a drm_device wrong. Doesn't matter, since everyone else gets it wrong
too :-)
- also rework the drm_framebuffer_funcs->dirty hook wire-up, see above.
Contact: Noralf Trønnes, Daniel Vetter
AMD DC Display Driver

11
Documentation/gpu/vkms.rst
View File

@ -23,17 +23,6 @@ CRC API Improvements
- Add igt test to check extreme alpha values i.e. fully opaque and fully
transparent (intermediate values are affected by hw-specific rounding modes).
Vblank issues
-------------
Some IGT test cases are failing. Need to analyze why and fix the issues:
- plain-flip-fb-recreate
- plain-flip-ts-check
- flip-vs-blocking-wf-vblank
- plain-flip-fb-recreate-interruptible
- flip-vs-wf_vblank-interruptible
Runtime Configuration
---------------------

43
MAINTAINERS
View File

@ -1149,13 +1149,26 @@ S: Supported
F: drivers/gpu/drm/arm/hdlcd_*
F: Documentation/devicetree/bindings/display/arm,hdlcd.txt
ARM KOMEDA DRM-KMS DRIVER
M: James (Qian) Wang <james.qian.wang@arm.com>
M: Liviu Dudau <liviu.dudau@arm.com>
L: Mali DP Maintainers <malidp@foss.arm.com>
S: Supported
T: git git://linux-arm.org/linux-ld.git for-upstream/mali-dp
F: drivers/gpu/drm/arm/display/include/
F: drivers/gpu/drm/arm/display/komeda/
F: Documentation/devicetree/bindings/display/arm/arm,komeda.txt
F: Documentation/gpu/komeda-kms.rst
ARM MALI-DP DRM DRIVER
M: Liviu Dudau <liviu.dudau@arm.com>
M: Brian Starkey <brian.starkey@arm.com>
M: Mali DP Maintainers <malidp@foss.arm.com>
L: Mali DP Maintainers <malidp@foss.arm.com>
S: Supported
T: git git://linux-arm.org/linux-ld.git for-upstream/mali-dp
F: drivers/gpu/drm/arm/
F: Documentation/devicetree/bindings/display/arm,malidp.txt
F: Documentation/gpu/afbc.rst
ARM MFM AND FLOPPY DRIVERS
M: Ian Molton <spyro@f2s.com>
@ -4900,10 +4913,11 @@ F: Documentation/devicetree/bindings/display/multi-inno,mi0283qt.txt
DRM DRIVER FOR MSM ADRENO GPU
M: Rob Clark <robdclark@gmail.com>
M: Sean Paul <sean@poorly.run>
L: linux-arm-msm@vger.kernel.org
L: dri-devel@lists.freedesktop.org
L: freedreno@lists.freedesktop.org
T: git git://people.freedesktop.org/~robclark/linux
T: git https://gitlab.freedesktop.org/drm/msm.git
S: Maintained
F: drivers/gpu/drm/msm/
F: include/uapi/drm/msm_drm.h
@ -4943,6 +4957,7 @@ DRM DRIVER FOR QXL VIRTUAL GPU
M: Dave Airlie <airlied@redhat.com>
M: Gerd Hoffmann <kraxel@redhat.com>
L: virtualization@lists.linux-foundation.org
L: spice-devel@lists.freedesktop.org
T: git git://anongit.freedesktop.org/drm/drm-misc
S: Maintained
F: drivers/gpu/drm/qxl/
@ -4963,6 +4978,12 @@ S: Orphan / Obsolete
F: drivers/gpu/drm/sis/
F: include/uapi/drm/sis_drm.h
DRM DRIVER FOR SITRONIX ST7701 PANELS
M: Jagan Teki <jagan@amarulasolutions.com>
S: Maintained
F: drivers/gpu/drm/panel/panel-sitronix-st7701.c
F: Documentation/devicetree/bindings/display/panel/sitronix,st7701.txt
DRM DRIVER FOR SITRONIX ST7586 PANELS
M: David Lechner <david@lechnology.com>
S: Maintained
@ -4979,6 +5000,13 @@ DRM DRIVER FOR TDFX VIDEO CARDS
S: Orphan / Obsolete
F: drivers/gpu/drm/tdfx/
DRM DRIVER FOR TPO TPG110 PANELS
M: Linus Walleij <linus.walleij@linaro.org>
T: git git://anongit.freedesktop.org/drm/drm-misc
S: Maintained
F: drivers/gpu/drm/panel/panel-tpo-tpg110.c
F: Documentation/devicetree/bindings/display/panel/tpo,tpg110.txt
DRM DRIVER FOR USB DISPLAYLINK VIDEO ADAPTERS
M: Dave Airlie <airlied@redhat.com>
R: Sean Paul <sean@poorly.run>
@ -4987,6 +5015,16 @@ S: Odd Fixes
F: drivers/gpu/drm/udl/
T: git git://anongit.freedesktop.org/drm/drm-misc
DRM DRIVER FOR VIRTUAL KERNEL MODESETTING (VKMS)
M: Rodrigo Siqueira <rodrigosiqueiramelo@gmail.com>
R: Haneen Mohammed <hamohammed.sa@gmail.com>
R: Daniel Vetter <daniel@ffwll.ch>
T: git git://anongit.freedesktop.org/drm/drm-misc
S: Maintained
L: dri-devel@lists.freedesktop.org
F: drivers/gpu/drm/vkms/
F: Documentation/gpu/vkms.rst
DRM DRIVER FOR VMWARE VIRTUAL GPU
M: "VMware Graphics" <linux-graphics-maintainer@vmware.com>
M: Thomas Hellstrom <thellstrom@vmware.com>
@ -5056,7 +5094,6 @@ F: Documentation/devicetree/bindings/display/atmel/
T: git git://anongit.freedesktop.org/drm/drm-misc
DRM DRIVERS FOR BRIDGE CHIPS
M: Archit Taneja <architt@codeaurora.org>
M: Andrzej Hajda <a.hajda@samsung.com>
R: Laurent Pinchart <Laurent.pinchart@ideasonboard.com>
S: Maintained

61
drivers/acpi/pmic/intel_pmic.c
View File

@ -15,6 +15,7 @@
#include <linux/export.h>
#include <linux/acpi.h>
#include <linux/mfd/intel_soc_pmic.h>
#include <linux/regmap.h>
#include <acpi/acpi_lpat.h>
#include "intel_pmic.h"
@ -36,6 +37,8 @@ struct intel_pmic_opregion {
struct intel_pmic_regs_handler_ctx ctx;
};
static struct intel_pmic_opregion *intel_pmic_opregion;
static int pmic_get_reg_bit(int address, struct pmic_table *table,
int count, int *reg, int *bit)
{
@ -304,6 +307,7 @@ int intel_pmic_install_opregion_handler(struct device *dev, acpi_handle handle,
}
opregion->data = d;
intel_pmic_opregion = opregion;
return 0;
out_remove_thermal_handler:
@ -319,3 +323,60 @@ out_error:
return ret;
}
EXPORT_SYMBOL_GPL(intel_pmic_install_opregion_handler);
/**
* intel_soc_pmic_exec_mipi_pmic_seq_element - Execute PMIC MIPI sequence
* @i2c_address: I2C client address for the PMIC
* @reg_address: PMIC register address
* @value: New value for the register bits to change
* @mask: Mask indicating which register bits to change
*
* DSI LCD panels describe an initialization sequence in the i915 VBT (Video
* BIOS Tables) using so called MIPI sequences. One possible element in these
* sequences is a PMIC specific element of 15 bytes.
*
* This function executes these PMIC specific elements sending the embedded
* commands to the PMIC.
*
* Return 0 on success, < 0 on failure.
*/
int intel_soc_pmic_exec_mipi_pmic_seq_element(u16 i2c_address, u32 reg_address,
u32 value, u32 mask)
{
struct intel_pmic_opregion_data *d;
int ret;
if (!intel_pmic_opregion) {
pr_warn("%s: No PMIC registered\n", __func__);
return -ENXIO;
}
d = intel_pmic_opregion->data;
mutex_lock(&intel_pmic_opregion->lock);
if (d->exec_mipi_pmic_seq_element) {
ret = d->exec_mipi_pmic_seq_element(intel_pmic_opregion->regmap,
i2c_address, reg_address,
value, mask);
} else if (d->pmic_i2c_address) {
if (i2c_address == d->pmic_i2c_address) {
ret = regmap_update_bits(intel_pmic_opregion->regmap,
reg_address, mask, value);
} else {
pr_err("%s: Unexpected i2c-addr: 0x%02x (reg-addr 0x%x value 0x%x mask 0x%x)\n",
__func__, i2c_address, reg_address, value, mask);
ret = -ENXIO;
}
} else {
pr_warn("%s: Not implemented\n", __func__);
pr_warn("%s: i2c-addr: 0x%x reg-addr 0x%x value 0x%x mask 0x%x\n",
__func__, i2c_address, reg_address, value, mask);
ret = -EOPNOTSUPP;
}
mutex_unlock(&intel_pmic_opregion->lock);
return ret;
}
EXPORT_SYMBOL_GPL(intel_soc_pmic_exec_mipi_pmic_seq_element);

4
drivers/acpi/pmic/intel_pmic.h
View File

@ -15,10 +15,14 @@ struct intel_pmic_opregion_data {
int (*update_aux)(struct regmap *r, int reg, int raw_temp);
int (*get_policy)(struct regmap *r, int reg, int bit, u64 *value);
int (*update_policy)(struct regmap *r, int reg, int bit, int enable);
int (*exec_mipi_pmic_seq_element)(struct regmap *r, u16 i2c_address,
u32 reg_address, u32 value, u32 mask);
struct pmic_table *power_table;
int power_table_count;
struct pmic_table *thermal_table;
int thermal_table_count;
/* For generic exec_mipi_pmic_seq_element handling */
int pmic_i2c_address;
};
int intel_pmic_install_opregion_handler(struct device *dev, acpi_handle handle, struct regmap *regmap, struct intel_pmic_opregion_data *d);

19
drivers/acpi/pmic/intel_pmic_chtwc.c
View File

@ -231,6 +231,24 @@ static int intel_cht_wc_pmic_update_power(struct regmap *regmap, int reg,
return regmap_update_bits(regmap, reg, bitmask, on ? 1 : 0);
}
static int intel_cht_wc_exec_mipi_pmic_seq_element(struct regmap *regmap,
u16 i2c_client_address,
u32 reg_address,
u32 value, u32 mask)
{
u32 address;
if (i2c_client_address > 0xff || reg_address > 0xff) {
pr_warn("%s warning addresses too big client 0x%x reg 0x%x\n",
__func__, i2c_client_address, reg_address);
return -ERANGE;
}
address = (i2c_client_address << 8) | reg_address;
return regmap_update_bits(regmap, address, mask, value);
}
/*
* The thermal table and ops are empty, we do not support the Thermal opregion
* (DPTF) due to lacking documentation.
@ -238,6 +256,7 @@ static int intel_cht_wc_pmic_update_power(struct regmap *regmap, int reg,
static struct intel_pmic_opregion_data intel_cht_wc_pmic_opregion_data = {
.get_power = intel_cht_wc_pmic_get_power,
.update_power = intel_cht_wc_pmic_update_power,
.exec_mipi_pmic_seq_element = intel_cht_wc_exec_mipi_pmic_seq_element,
.power_table = power_table,
.power_table_count = ARRAY_SIZE(power_table),
};

1
drivers/acpi/pmic/intel_pmic_xpower.c
View File

@ -265,6 +265,7 @@ static struct intel_pmic_opregion_data intel_xpower_pmic_opregion_data = {
.power_table_count = ARRAY_SIZE(power_table),
.thermal_table = thermal_table,
.thermal_table_count = ARRAY_SIZE(thermal_table),
.pmic_i2c_address = 0x34,
};
static acpi_status intel_xpower_pmic_gpio_handler(u32 function,

12
drivers/dma-buf/dma-buf.c
View File

@ -1093,17 +1093,7 @@ static int dma_buf_debug_show(struct seq_file *s, void *unused)
return 0;
}
static int dma_buf_debug_open(struct inode *inode, struct file *file)
{
return single_open(file, dma_buf_debug_show, NULL);
}
static const struct file_operations dma_buf_debug_fops = {
.open = dma_buf_debug_open,
.read = seq_read,
.llseek = seq_lseek,
.release = single_release,
};
DEFINE_SHOW_ATTRIBUTE(dma_buf_debug);
static struct dentry *dma_buf_debugfs_dir;

2
drivers/dma-buf/dma-fence.c
View File

@ -649,7 +649,7 @@ EXPORT_SYMBOL(dma_fence_wait_any_timeout);
*/
void
dma_fence_init(struct dma_fence *fence, const struct dma_fence_ops *ops,
spinlock_t *lock, u64 context, unsigned seqno)
spinlock_t *lock, u64 context, u64 seqno)
{
BUG_ON(!lock);
BUG_ON(!ops || !ops->get_driver_name || !ops->get_timeline_name);

2
drivers/dma-buf/sw_sync.c
View File

@ -172,7 +172,7 @@ static bool timeline_fence_enable_signaling(struct dma_fence *fence)
static void timeline_fence_value_str(struct dma_fence *fence,
char *str, int size)
{
snprintf(str, size, "%d", fence->seqno);
snprintf(str, size, "%lld", fence->seqno);
}
static void timeline_fence_timeline_value_str(struct dma_fence *fence,

16
drivers/dma-buf/sync_debug.c
View File

@ -147,7 +147,7 @@ static void sync_print_sync_file(struct seq_file *s,
}
}
static int sync_debugfs_show(struct seq_file *s, void *unused)
static int sync_info_debugfs_show(struct seq_file *s, void *unused)
{
struct list_head *pos;
@ -178,17 +178,7 @@ static int sync_debugfs_show(struct seq_file *s, void *unused)
return 0;
}
static int sync_info_debugfs_open(struct inode *inode, struct file *file)
{
return single_open(file, sync_debugfs_show, inode->i_private);
}
static const struct file_operations sync_info_debugfs_fops = {
.open = sync_info_debugfs_open,
.read = seq_read,
.llseek = seq_lseek,
.release = single_release,
};
DEFINE_SHOW_ATTRIBUTE(sync_info_debugfs);
static __init int sync_debugfs_init(void)
{
@ -218,7 +208,7 @@ void sync_dump(void)
};
int i;
sync_debugfs_show(&s, NULL);
sync_info_debugfs_show(&s, NULL);
for (i = 0; i < s.count; i += DUMP_CHUNK) {
if ((s.count - i) > DUMP_CHUNK) {

4
drivers/dma-buf/sync_file.c
View File

@ -144,7 +144,7 @@ char *sync_file_get_name(struct sync_file *sync_file, char *buf, int len)
} else {
struct dma_fence *fence = sync_file->fence;
snprintf(buf, len, "%s-%s%llu-%d",
snprintf(buf, len, "%s-%s%llu-%lld",
fence->ops->get_driver_name(fence),
fence->ops->get_timeline_name(fence),
fence->context,
@ -258,7 +258,7 @@ static struct sync_file *sync_file_merge(const char *name, struct sync_file *a,
i_b++;
} else {
if (pt_a->seqno - pt_b->seqno <= INT_MAX)
if (__dma_fence_is_later(pt_a->seqno, pt_b->seqno))
add_fence(fences, &i, pt_a);
else
add_fence(fences, &i, pt_b);

4
drivers/gpu/drm/Kconfig
View File

@ -170,10 +170,6 @@ config DRM_KMS_CMA_HELPER
bool
depends on DRM
select DRM_GEM_CMA_HELPER
select DRM_KMS_FB_HELPER
select FB_SYS_FILLRECT
select FB_SYS_COPYAREA
select FB_SYS_IMAGEBLIT
help
Choose this if you need the KMS CMA helper functions

4
drivers/gpu/drm/Makefile
View File

@ -51,7 +51,7 @@ obj-$(CONFIG_DRM_DEBUG_SELFTEST) += selftests/
obj-$(CONFIG_DRM) += drm.o
obj-$(CONFIG_DRM_MIPI_DSI) += drm_mipi_dsi.o
obj-$(CONFIG_DRM_PANEL_ORIENTATION_QUIRKS) += drm_panel_orientation_quirks.o
obj-$(CONFIG_DRM_ARM) += arm/
obj-y += arm/
obj-$(CONFIG_DRM_TTM) += ttm/
obj-$(CONFIG_DRM_SCHED) += scheduler/
obj-$(CONFIG_DRM_TDFX) += tdfx/
@ -81,7 +81,7 @@ obj-$(CONFIG_DRM_UDL) += udl/
obj-$(CONFIG_DRM_AST) += ast/
obj-$(CONFIG_DRM_ARMADA) += armada/
obj-$(CONFIG_DRM_ATMEL_HLCDC) += atmel-hlcdc/
obj-$(CONFIG_DRM_RCAR_DU) += rcar-du/
obj-y += rcar-du/
obj-$(CONFIG_DRM_SHMOBILE) +=shmobile/
obj-y += omapdrm/
obj-$(CONFIG_DRM_SUN4I) += sun4i/

2
drivers/gpu/drm/amd/amdgpu/Makefile
View File

@ -57,7 +57,7 @@ amdgpu-y += amdgpu_device.o amdgpu_kms.o \
# add asic specific block
amdgpu-$(CONFIG_DRM_AMDGPU_CIK)+= cik.o cik_ih.o kv_smc.o kv_dpm.o \
ci_smc.o ci_dpm.o dce_v8_0.o gfx_v7_0.o cik_sdma.o uvd_v4_2.o vce_v2_0.o
dce_v8_0.o gfx_v7_0.o cik_sdma.o uvd_v4_2.o vce_v2_0.o
amdgpu-$(CONFIG_DRM_AMDGPU_SI)+= si.o gmc_v6_0.o gfx_v6_0.o si_ih.o si_dma.o dce_v6_0.o si_dpm.o si_smc.o

11
drivers/gpu/drm/amd/amdgpu/amdgpu.h
View File

@ -411,6 +411,8 @@ struct amdgpu_fpriv {
struct amdgpu_ctx_mgr ctx_mgr;
};
int amdgpu_file_to_fpriv(struct file *filp, struct amdgpu_fpriv **fpriv);
int amdgpu_ib_get(struct amdgpu_device *adev, struct amdgpu_vm *vm,
unsigned size, struct amdgpu_ib *ib);
void amdgpu_ib_free(struct amdgpu_device *adev, struct amdgpu_ib *ib,
@ -542,6 +544,11 @@ struct amdgpu_asic_funcs {
bool (*need_full_reset)(struct amdgpu_device *adev);
/* initialize doorbell layout for specific asic*/
void (*init_doorbell_index)(struct amdgpu_device *adev);
/* PCIe bandwidth usage */
void (*get_pcie_usage)(struct amdgpu_device *adev, uint64_t *count0,
uint64_t *count1);
/* do we need to reset the asic at init time (e.g., kexec) */
bool (*need_reset_on_init)(struct amdgpu_device *adev);
};
/*
@ -634,7 +641,7 @@ struct amdgpu_nbio_funcs {
void (*hdp_flush)(struct amdgpu_device *adev, struct amdgpu_ring *ring);
u32 (*get_memsize)(struct amdgpu_device *adev);
void (*sdma_doorbell_range)(struct amdgpu_device *adev, int instance,
bool use_doorbell, int doorbell_index);
bool use_doorbell, int doorbell_index, int doorbell_size);
void (*enable_doorbell_aperture)(struct amdgpu_device *adev,
bool enable);
void (*enable_doorbell_selfring_aperture)(struct amdgpu_device *adev,
@ -1042,6 +1049,8 @@ int emu_soc_asic_init(struct amdgpu_device *adev);
#define amdgpu_asic_invalidate_hdp(adev, r) (adev)->asic_funcs->invalidate_hdp((adev), (r))
#define amdgpu_asic_need_full_reset(adev) (adev)->asic_funcs->need_full_reset((adev))
#define amdgpu_asic_init_doorbell_index(adev) (adev)->asic_funcs->init_doorbell_index((adev))
#define amdgpu_asic_get_pcie_usage(adev, cnt0, cnt1) ((adev)->asic_funcs->get_pcie_usage((adev), (cnt0), (cnt1)))
#define amdgpu_asic_need_reset_on_init(adev) (adev)->asic_funcs->need_reset_on_init((adev))
/* Common functions */
bool amdgpu_device_should_recover_gpu(struct amdgpu_device *adev);

121
drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c
View File

@ -28,8 +28,6 @@
#include <linux/module.h>
#include <linux/dma-buf.h>
const struct kgd2kfd_calls *kgd2kfd;
static const unsigned int compute_vmid_bitmap = 0xFF00;
/* Total memory size in system memory and all GPU VRAM. Used to
@ -47,12 +45,9 @@ int amdgpu_amdkfd_init(void)
amdgpu_amdkfd_total_mem_size *= si.mem_unit;
#ifdef CONFIG_HSA_AMD
ret = kgd2kfd_init(KFD_INTERFACE_VERSION, &kgd2kfd);
if (ret)
kgd2kfd = NULL;
ret = kgd2kfd_init();
amdgpu_amdkfd_gpuvm_init_mem_limits();
#else
kgd2kfd = NULL;
ret = -ENOENT;
#endif
@ -61,17 +56,13 @@ int amdgpu_amdkfd_init(void)
void amdgpu_amdkfd_fini(void)
{
if (kgd2kfd)
kgd2kfd->exit();
kgd2kfd_exit();
}
void amdgpu_amdkfd_device_probe(struct amdgpu_device *adev)
{
const struct kfd2kgd_calls *kfd2kgd;
if (!kgd2kfd)
return;
switch (adev->asic_type) {
#ifdef CONFIG_DRM_AMDGPU_CIK
case CHIP_KAVERI:
@ -98,8 +89,8 @@ void amdgpu_amdkfd_device_probe(struct amdgpu_device *adev)
return;
}
adev->kfd.dev = kgd2kfd->probe((struct kgd_dev *)adev,
adev->pdev, kfd2kgd);
adev->kfd.dev = kgd2kfd_probe((struct kgd_dev *)adev,
adev->pdev, kfd2kgd);
if (adev->kfd.dev)
amdgpu_amdkfd_total_mem_size += adev->gmc.real_vram_size;
@ -140,7 +131,7 @@ static void amdgpu_doorbell_get_kfd_info(struct amdgpu_device *adev,
void amdgpu_amdkfd_device_init(struct amdgpu_device *adev)
{
int i, n;
int i;
int last_valid_bit;
if (adev->kfd.dev) {
@ -151,7 +142,9 @@ void amdgpu_amdkfd_device_init(struct amdgpu_device *adev)
.gpuvm_size = min(adev->vm_manager.max_pfn
<< AMDGPU_GPU_PAGE_SHIFT,
AMDGPU_GMC_HOLE_START),
.drm_render_minor = adev->ddev->render->index
.drm_render_minor = adev->ddev->render->index,
.sdma_doorbell_idx = adev->doorbell_index.sdma_engine,
};
/* this is going to have a few of the MSBs set that we need to
@ -181,44 +174,29 @@ void amdgpu_amdkfd_device_init(struct amdgpu_device *adev)
&gpu_resources.doorbell_aperture_size,
&gpu_resources.doorbell_start_offset);
if (adev->asic_type < CHIP_VEGA10) {
kgd2kfd->device_init(adev->kfd.dev, &gpu_resources);
return;
}
n = (adev->asic_type < CHIP_VEGA20) ? 2 : 8;
for (i = 0; i < n; i += 2) {
/* On SOC15 the BIF is involved in routing
* doorbells using the low 12 bits of the
* address. Communicate the assignments to
* KFD. KFD uses two doorbell pages per
* process in case of 64-bit doorbells so we
* can use each doorbell assignment twice.
*/
gpu_resources.sdma_doorbell[0][i] =
adev->doorbell_index.sdma_engine0 + (i >> 1);
gpu_resources.sdma_doorbell[0][i+1] =
adev->doorbell_index.sdma_engine0 + 0x200 + (i >> 1);
gpu_resources.sdma_doorbell[1][i] =
adev->doorbell_index.sdma_engine1 + (i >> 1);
gpu_resources.sdma_doorbell[1][i+1] =
adev->doorbell_index.sdma_engine1 + 0x200 + (i >> 1);
}
/* Doorbells 0x0e0-0ff and 0x2e0-2ff are reserved for
* SDMA, IH and VCN. So don't use them for the CP.
/* Since SOC15, BIF starts to statically use the
* lower 12 bits of doorbell addresses for routing
* based on settings in registers like
* SDMA0_DOORBELL_RANGE etc..
* In order to route a doorbell to CP engine, the lower
* 12 bits of its address has to be outside the range
* set for SDMA, VCN, and IH blocks.
*/
gpu_resources.reserved_doorbell_mask = 0x1e0;
gpu_resources.reserved_doorbell_val = 0x0e0;
if (adev->asic_type >= CHIP_VEGA10) {
gpu_resources.non_cp_doorbells_start =
adev->doorbell_index.first_non_cp;
gpu_resources.non_cp_doorbells_end =
adev->doorbell_index.last_non_cp;
}
kgd2kfd->device_init(adev->kfd.dev, &gpu_resources);
kgd2kfd_device_init(adev->kfd.dev, &gpu_resources);
}
}
void amdgpu_amdkfd_device_fini(struct amdgpu_device *adev)
{
if (adev->kfd.dev) {
kgd2kfd->device_exit(adev->kfd.dev);
kgd2kfd_device_exit(adev->kfd.dev);
adev->kfd.dev = NULL;
}
}
@ -227,13 +205,13 @@ void amdgpu_amdkfd_interrupt(struct amdgpu_device *adev,
const void *ih_ring_entry)
{
if (adev->kfd.dev)
kgd2kfd->interrupt(adev->kfd.dev, ih_ring_entry);
kgd2kfd_interrupt(adev->kfd.dev, ih_ring_entry);
}
void amdgpu_amdkfd_suspend(struct amdgpu_device *adev)
{
if (adev->kfd.dev)
kgd2kfd->suspend(adev->kfd.dev);
kgd2kfd_suspend(adev->kfd.dev);
}
int amdgpu_amdkfd_resume(struct amdgpu_device *adev)
@ -241,7 +219,7 @@ int amdgpu_amdkfd_resume(struct amdgpu_device *adev)
int r = 0;
if (adev->kfd.dev)
r = kgd2kfd->resume(adev->kfd.dev);
r = kgd2kfd_resume(adev->kfd.dev);
return r;
}
@ -251,7 +229,7 @@ int amdgpu_amdkfd_pre_reset(struct amdgpu_device *adev)
int r = 0;
if (adev->kfd.dev)
r = kgd2kfd->pre_reset(adev->kfd.dev);
r = kgd2kfd_pre_reset(adev->kfd.dev);
return r;
}
@ -261,7 +239,7 @@ int amdgpu_amdkfd_post_reset(struct amdgpu_device *adev)
int r = 0;
if (adev->kfd.dev)
r = kgd2kfd->post_reset(adev->kfd.dev);
r = kgd2kfd_post_reset(adev->kfd.dev);
return r;
}
@ -619,4 +597,47 @@ struct kfd2kgd_calls *amdgpu_amdkfd_gfx_9_0_get_functions(void)
{
return NULL;
}
struct kfd_dev *kgd2kfd_probe(struct kgd_dev *kgd, struct pci_dev *pdev,
const struct kfd2kgd_calls *f2g)
{
return NULL;
}
bool kgd2kfd_device_init(struct kfd_dev *kfd,
const struct kgd2kfd_shared_resources *gpu_resources)
{
return false;
}
void kgd2kfd_device_exit(struct kfd_dev *kfd)
{
}
void kgd2kfd_exit(void)
{
}
void kgd2kfd_suspend(struct kfd_dev *kfd)
{
}
int kgd2kfd_resume(struct kfd_dev *kfd)
{
return 0;
}
int kgd2kfd_pre_reset(struct kfd_dev *kfd)
{
return 0;
}
int kgd2kfd_post_reset(struct kfd_dev *kfd)
{
return 0;
}
void kgd2kfd_interrupt(struct kfd_dev *kfd, const void *ih_ring_entry)
{
}
#endif

19
drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h
View File

@ -33,7 +33,6 @@
#include "amdgpu_sync.h"
#include "amdgpu_vm.h"
extern const struct kgd2kfd_calls *kgd2kfd;
extern uint64_t amdgpu_amdkfd_total_mem_size;
struct amdgpu_device;
@ -214,4 +213,22 @@ int amdgpu_amdkfd_gpuvm_import_dmabuf(struct kgd_dev *kgd,
void amdgpu_amdkfd_gpuvm_init_mem_limits(void);
void amdgpu_amdkfd_unreserve_memory_limit(struct amdgpu_bo *bo);
/* KGD2KFD callbacks */
int kgd2kfd_init(void);
void kgd2kfd_exit(void);
struct kfd_dev *kgd2kfd_probe(struct kgd_dev *kgd, struct pci_dev *pdev,
const struct kfd2kgd_calls *f2g);
bool kgd2kfd_device_init(struct kfd_dev *kfd,
const struct kgd2kfd_shared_resources *gpu_resources);
void kgd2kfd_device_exit(struct kfd_dev *kfd);
void kgd2kfd_suspend(struct kfd_dev *kfd);
int kgd2kfd_resume(struct kfd_dev *kfd);
int kgd2kfd_pre_reset(struct kfd_dev *kfd);
int kgd2kfd_post_reset(struct kfd_dev *kfd);
void kgd2kfd_interrupt(struct kfd_dev *kfd, const void *ih_ring_entry);
int kgd2kfd_quiesce_mm(struct mm_struct *mm);
int kgd2kfd_resume_mm(struct mm_struct *mm);
int kgd2kfd_schedule_evict_and_restore_process(struct mm_struct *mm,
struct dma_fence *fence);
#endif /* AMDGPU_AMDKFD_H_INCLUDED */

2
drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_fence.c
View File

@ -122,7 +122,7 @@ static bool amdkfd_fence_enable_signaling(struct dma_fence *f)
if (dma_fence_is_signaled(f))
return true;
if (!kgd2kfd->schedule_evict_and_restore_process(fence->mm, f))
if (!kgd2kfd_schedule_evict_and_restore_process(fence->mm, f))
return true;
return false;

142
drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c
View File

@ -204,38 +204,25 @@ void amdgpu_amdkfd_unreserve_memory_limit(struct amdgpu_bo *bo)
}
/* amdgpu_amdkfd_remove_eviction_fence - Removes eviction fence(s) from BO's
/* amdgpu_amdkfd_remove_eviction_fence - Removes eviction fence from BO's
* reservation object.
*
* @bo: [IN] Remove eviction fence(s) from this BO
* @ef: [IN] If ef is specified, then this eviction fence is removed if it
* @ef: [IN] This eviction fence is removed if it
* is present in the shared list.
* @ef_list: [OUT] Returns list of eviction fences. These fences are removed
* from BO's reservation object shared list.
* @ef_count: [OUT] Number of fences in ef_list.
*
* NOTE: If called with ef_list, then amdgpu_amdkfd_add_eviction_fence must be
* called to restore the eviction fences and to avoid memory leak. This is
* useful for shared BOs.
* NOTE: Must be called with BO reserved i.e. bo->tbo.resv->lock held.
*/
static int amdgpu_amdkfd_remove_eviction_fence(struct amdgpu_bo *bo,
struct amdgpu_amdkfd_fence *ef,
struct amdgpu_amdkfd_fence ***ef_list,
unsigned int *ef_count)
struct amdgpu_amdkfd_fence *ef)
{
struct reservation_object *resv = bo->tbo.resv;
struct reservation_object_list *old, *new;
unsigned int i, j, k;
if (!ef && !ef_list)
if (!ef)
return -EINVAL;
if (ef_list) {
*ef_list = NULL;
*ef_count = 0;
}
old = reservation_object_get_list(resv);
if (!old)
return 0;
@ -254,8 +241,7 @@ static int amdgpu_amdkfd_remove_eviction_fence(struct amdgpu_bo *bo,
f = rcu_dereference_protected(old->shared[i],
reservation_object_held(resv));
if ((ef && f->context == ef->base.context) ||
(!ef && to_amdgpu_amdkfd_fence(f)))
if (f->context == ef->base.context)
RCU_INIT_POINTER(new->shared[--j], f);
else
RCU_INIT_POINTER(new->shared[k++], f);
@ -263,21 +249,6 @@ static int amdgpu_amdkfd_remove_eviction_fence(struct amdgpu_bo *bo,
new->shared_max = old->shared_max;
new->shared_count = k;
if (!ef) {
unsigned int count = old->shared_count - j;
/* Alloc memory for count number of eviction fence pointers.
* Fill the ef_list array and ef_count
*/
*ef_list = kcalloc(count, sizeof(**ef_list), GFP_KERNEL);
*ef_count = count;
if (!*ef_list) {
kfree(new);
return -ENOMEM;
}
}
/* Install the new fence list, seqcount provides the barriers */
preempt_disable();
write_seqcount_begin(&resv->seq);
@ -291,46 +262,13 @@ static int amdgpu_amdkfd_remove_eviction_fence(struct amdgpu_bo *bo,
f = rcu_dereference_protected(new->shared[i],
reservation_object_held(resv));
if (!ef)
(*ef_list)[k++] = to_amdgpu_amdkfd_fence(f);
else
dma_fence_put(f);
dma_fence_put(f);
}
kfree_rcu(old, rcu);
return 0;
}
/* amdgpu_amdkfd_add_eviction_fence - Adds eviction fence(s) back into BO's
* reservation object.
*
* @bo: [IN] Add eviction fences to this BO
* @ef_list: [IN] List of eviction fences to be added
* @ef_count: [IN] Number of fences in ef_list.
*
* NOTE: Must call amdgpu_amdkfd_remove_eviction_fence before calling this
* function.
*/
static void amdgpu_amdkfd_add_eviction_fence(struct amdgpu_bo *bo,
struct amdgpu_amdkfd_fence **ef_list,
unsigned int ef_count)
{
int i;
if (!ef_list || !ef_count)
return;
for (i = 0; i < ef_count; i++) {
amdgpu_bo_fence(bo, &ef_list[i]->base, true);
/* Re-adding the fence takes an additional reference. Drop that
* reference.
*/
dma_fence_put(&ef_list[i]->base);
}
kfree(ef_list);
}
static int amdgpu_amdkfd_bo_validate(struct amdgpu_bo *bo, uint32_t domain,
bool wait)
{
@ -346,18 +284,8 @@ static int amdgpu_amdkfd_bo_validate(struct amdgpu_bo *bo, uint32_t domain,
ret = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx);
if (ret)
goto validate_fail;
if (wait) {
struct amdgpu_amdkfd_fence **ef_list;
unsigned int ef_count;
ret = amdgpu_amdkfd_remove_eviction_fence(bo, NULL, &ef_list,
&ef_count);
if (ret)
goto validate_fail;
ttm_bo_wait(&bo->tbo, false, false);
amdgpu_amdkfd_add_eviction_fence(bo, ef_list, ef_count);
}
if (wait)
amdgpu_bo_sync_wait(bo, AMDGPU_FENCE_OWNER_KFD, false);
validate_fail:
return ret;
@ -444,7 +372,6 @@ static int add_bo_to_vm(struct amdgpu_device *adev, struct kgd_mem *mem,
{
int ret;
struct kfd_bo_va_list *bo_va_entry;
struct amdgpu_bo *pd = vm->root.base.bo;
struct amdgpu_bo *bo = mem->bo;
uint64_t va = mem->va;
struct list_head *list_bo_va = &mem->bo_va_list;
@ -484,14 +411,8 @@ static int add_bo_to_vm(struct amdgpu_device *adev, struct kgd_mem *mem,
*p_bo_va_entry = bo_va_entry;
/* Allocate new page tables if needed and validate
* them. Clearing of new page tables and validate need to wait
* on move fences. We don't want that to trigger the eviction
* fence, so remove it temporarily.
* them.
*/
amdgpu_amdkfd_remove_eviction_fence(pd,
vm->process_info->eviction_fence,
NULL, NULL);
ret = amdgpu_vm_alloc_pts(adev, vm, va, amdgpu_bo_size(bo));
if (ret) {
pr_err("Failed to allocate pts, err=%d\n", ret);
@ -504,13 +425,9 @@ static int add_bo_to_vm(struct amdgpu_device *adev, struct kgd_mem *mem,
goto err_alloc_pts;
}
/* Add the eviction fence back */
amdgpu_bo_fence(pd, &vm->process_info->eviction_fence->base, true);
return 0;
err_alloc_pts:
amdgpu_bo_fence(pd, &vm->process_info->eviction_fence->base, true);
amdgpu_vm_bo_rmv(adev, bo_va_entry->bo_va);
list_del(&bo_va_entry->bo_list);
err_vmadd:
@ -809,24 +726,11 @@ static int unmap_bo_from_gpuvm(struct amdgpu_device *adev,
{
struct amdgpu_bo_va *bo_va = entry->bo_va;
struct amdgpu_vm *vm = bo_va->base.vm;
struct amdgpu_bo *pd = vm->root.base.bo;
/* Remove eviction fence from PD (and thereby from PTs too as
* they share the resv. object). Otherwise during PT update
* job (see amdgpu_vm_bo_update_mapping), eviction fence would
* get added to job->sync object and job execution would
* trigger the eviction fence.
*/
amdgpu_amdkfd_remove_eviction_fence(pd,
vm->process_info->eviction_fence,
NULL, NULL);
amdgpu_vm_bo_unmap(adev, bo_va, entry->va);
amdgpu_vm_clear_freed(adev, vm, &bo_va->last_pt_update);
/* Add the eviction fence back */
amdgpu_bo_fence(pd, &vm->process_info->eviction_fence->base, true);
amdgpu_sync_fence(NULL, sync, bo_va->last_pt_update, false);
return 0;
@ -1002,7 +906,7 @@ static int init_kfd_vm(struct amdgpu_vm *vm, void **process_info,
pr_err("validate_pt_pd_bos() failed\n");
goto validate_pd_fail;
}
ret = ttm_bo_wait(&vm->root.base.bo->tbo, false, false);
amdgpu_bo_sync_wait(vm->root.base.bo, AMDGPU_FENCE_OWNER_KFD, false);
if (ret)
goto wait_pd_fail;
amdgpu_bo_fence(vm->root.base.bo,
@ -1389,8 +1293,7 @@ int amdgpu_amdkfd_gpuvm_free_memory_of_gpu(
* attached
*/
amdgpu_amdkfd_remove_eviction_fence(mem->bo,
process_info->eviction_fence,
NULL, NULL);
process_info->eviction_fence);
pr_debug("Release VA 0x%llx - 0x%llx\n", mem->va,
mem->va + bo_size * (1 + mem->aql_queue));
@ -1617,8 +1520,7 @@ int amdgpu_amdkfd_gpuvm_unmap_memory_from_gpu(
if (mem->mapped_to_gpu_memory == 0 &&
!amdgpu_ttm_tt_get_usermm(mem->bo->tbo.ttm) && !mem->bo->pin_count)
amdgpu_amdkfd_remove_eviction_fence(mem->bo,
process_info->eviction_fence,
NULL, NULL);
process_info->eviction_fence);
unreserve_out:
unreserve_bo_and_vms(&ctx, false, false);
@ -1679,7 +1581,7 @@ int amdgpu_amdkfd_gpuvm_map_gtt_bo_to_kernel(struct kgd_dev *kgd,
}
amdgpu_amdkfd_remove_eviction_fence(
bo, mem->process_info->eviction_fence, NULL, NULL);
bo, mem->process_info->eviction_fence);
list_del_init(&mem->validate_list.head);
if (size)
@ -1790,7 +1692,7 @@ int amdgpu_amdkfd_evict_userptr(struct kgd_mem *mem,
evicted_bos = atomic_inc_return(&process_info->evicted_bos);
if (evicted_bos == 1) {
/* First eviction, stop the queues */
r = kgd2kfd->quiesce_mm(mm);
r = kgd2kfd_quiesce_mm(mm);
if (r)
pr_err("Failed to quiesce KFD\n");
schedule_delayed_work(&process_info->restore_userptr_work,
@ -1945,16 +1847,6 @@ static int validate_invalid_user_pages(struct amdkfd_process_info *process_info)
amdgpu_sync_create(&sync);
/* Avoid triggering eviction fences when unmapping invalid
* userptr BOs (waits for all fences, doesn't use
* FENCE_OWNER_VM)
*/
list_for_each_entry(peer_vm, &process_info->vm_list_head,
vm_list_node)
amdgpu_amdkfd_remove_eviction_fence(peer_vm->root.base.bo,
process_info->eviction_fence,
NULL, NULL);
ret = process_validate_vms(process_info);
if (ret)
goto unreserve_out;
@ -2015,10 +1907,6 @@ static int validate_invalid_user_pages(struct amdkfd_process_info *process_info)
ret = process_update_pds(process_info, &sync);
unreserve_out:
list_for_each_entry(peer_vm, &process_info->vm_list_head,
vm_list_node)
amdgpu_bo_fence(peer_vm->root.base.bo,
&process_info->eviction_fence->base, true);
ttm_eu_backoff_reservation(&ticket, &resv_list);
amdgpu_sync_wait(&sync, false);
amdgpu_sync_free(&sync);
@ -2082,7 +1970,7 @@ static void amdgpu_amdkfd_restore_userptr_worker(struct work_struct *work)
evicted_bos)
goto unlock_out;
evicted_bos = 0;
if (kgd2kfd->resume_mm(mm)) {
if (kgd2kfd_resume_mm(mm)) {
pr_err("%s: Failed to resume KFD\n", __func__);
/* No recovery from this failure. Probably the CP is
* hanging. No point trying again.

2
drivers/gpu/drm/amd/amdgpu/amdgpu_connectors.c
View File

@ -25,8 +25,8 @@
*/
#include <drm/drmP.h>
#include <drm/drm_edid.h>
#include <drm/drm_crtc_helper.h>
#include <drm/drm_fb_helper.h>
#include <drm/drm_probe_helper.h>
#include <drm/amdgpu_drm.h>
#include "amdgpu.h"
#include "atom.h"

13
drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
View File

@ -214,6 +214,7 @@ static int amdgpu_cs_parser_init(struct amdgpu_cs_parser *p, union drm_amdgpu_cs
case AMDGPU_CHUNK_ID_DEPENDENCIES:
case AMDGPU_CHUNK_ID_SYNCOBJ_IN:
case AMDGPU_CHUNK_ID_SYNCOBJ_OUT:
case AMDGPU_CHUNK_ID_SCHEDULED_DEPENDENCIES:
break;
default:
@ -1090,6 +1091,15 @@ static int amdgpu_cs_process_fence_dep(struct amdgpu_cs_parser *p,
fence = amdgpu_ctx_get_fence(ctx, entity,
deps[i].handle);
if (chunk->chunk_id == AMDGPU_CHUNK_ID_SCHEDULED_DEPENDENCIES) {
struct drm_sched_fence *s_fence = to_drm_sched_fence(fence);
struct dma_fence *old = fence;
fence = dma_fence_get(&s_fence->scheduled);
dma_fence_put(old);
}
if (IS_ERR(fence)) {
r = PTR_ERR(fence);
amdgpu_ctx_put(ctx);
@ -1177,7 +1187,8 @@ static int amdgpu_cs_dependencies(struct amdgpu_device *adev,
chunk = &p->chunks[i];
if (chunk->chunk_id == AMDGPU_CHUNK_ID_DEPENDENCIES) {
if (chunk->chunk_id == AMDGPU_CHUNK_ID_DEPENDENCIES ||
chunk->chunk_id == AMDGPU_CHUNK_ID_SCHEDULED_DEPENDENCIES) {
r = amdgpu_cs_process_fence_dep(p, chunk);
if (r)
return r;

11
drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c
View File

@ -124,6 +124,7 @@ static int amdgpu_ctx_init(struct amdgpu_device *adev,
struct amdgpu_ring *rings[AMDGPU_MAX_RINGS];
struct drm_sched_rq *rqs[AMDGPU_MAX_RINGS];
unsigned num_rings;
unsigned num_rqs = 0;
switch (i) {
case AMDGPU_HW_IP_GFX:
@ -166,12 +167,16 @@ static int amdgpu_ctx_init(struct amdgpu_device *adev,
break;
}
for (j = 0; j < num_rings; ++j)
rqs[j] = &rings[j]->sched.sched_rq[priority];
for (j = 0; j < num_rings; ++j) {
if (!rings[j]->adev)
continue;
rqs[num_rqs++] = &rings[j]->sched.sched_rq[priority];
}
for (j = 0; j < amdgpu_ctx_num_entities[i]; ++j)
r = drm_sched_entity_init(&ctx->entities[i][j].entity,
rqs, num_rings, &ctx->guilty);
rqs, num_rqs, &ctx->guilty);
if (r)
goto error_cleanup_entities;
}

3
drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c
View File

@ -158,9 +158,6 @@ static int amdgpu_debugfs_process_reg_op(bool read, struct file *f,
while (size) {
uint32_t value;
if (*pos > adev->rmmio_size)
goto end;
if (read) {
value = RREG32(*pos >> 2);
r = put_user(value, (uint32_t *)buf);

129
drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
View File

@ -30,8 +30,8 @@
#include <linux/console.h>
#include <linux/slab.h>
#include <drm/drmP.h>
#include <drm/drm_crtc_helper.h>
#include <drm/drm_atomic_helper.h>
#include <drm/drm_probe_helper.h>
#include <drm/amdgpu_drm.h>
#include <linux/vgaarb.h>
#include <linux/vga_switcheroo.h>
@ -1645,7 +1645,7 @@ static int amdgpu_device_ip_init(struct amdgpu_device *adev)
if (r) {
DRM_ERROR("sw_init of IP block <%s> failed %d\n",
adev->ip_blocks[i].version->funcs->name, r);
return r;
goto init_failed;
}
adev->ip_blocks[i].status.sw = true;
@ -1654,17 +1654,17 @@ static int amdgpu_device_ip_init(struct amdgpu_device *adev)
r = amdgpu_device_vram_scratch_init(adev);
if (r) {
DRM_ERROR("amdgpu_vram_scratch_init failed %d\n", r);
return r;
goto init_failed;
}
r = adev->ip_blocks[i].version->funcs->hw_init((void *)adev);
if (r) {
DRM_ERROR("hw_init %d failed %d\n", i, r);
return r;
goto init_failed;
}
r = amdgpu_device_wb_init(adev);
if (r) {
DRM_ERROR("amdgpu_device_wb_init failed %d\n", r);
return r;
goto init_failed;
}
adev->ip_blocks[i].status.hw = true;
@ -1675,7 +1675,7 @@ static int amdgpu_device_ip_init(struct amdgpu_device *adev)
AMDGPU_CSA_SIZE);
if (r) {
DRM_ERROR("allocate CSA failed %d\n", r);
return r;
goto init_failed;
}
}
}
@ -1683,30 +1683,32 @@ static int amdgpu_device_ip_init(struct amdgpu_device *adev)
r = amdgpu_ucode_create_bo(adev); /* create ucode bo when sw_init complete*/
if (r)
return r;
goto init_failed;
r = amdgpu_device_ip_hw_init_phase1(adev);
if (r)
return r;
goto init_failed;
r = amdgpu_device_fw_loading(adev);
if (r)
return r;
goto init_failed;
r = amdgpu_device_ip_hw_init_phase2(adev);
if (r)
return r;
goto init_failed;
if (adev->gmc.xgmi.num_physical_nodes > 1)
amdgpu_xgmi_add_device(adev);
amdgpu_amdkfd_device_init(adev);
init_failed:
if (amdgpu_sriov_vf(adev)) {
amdgpu_virt_init_data_exchange(adev);
if (!r)
amdgpu_virt_init_data_exchange(adev);
amdgpu_virt_release_full_gpu(adev, true);
}
return 0;
return r;
}
/**
@ -2133,7 +2135,7 @@ static int amdgpu_device_ip_reinit_early_sriov(struct amdgpu_device *adev)
continue;
r = block->version->funcs->hw_init(adev);
DRM_INFO("RE-INIT: %s %s\n", block->version->funcs->name, r?"failed":"succeeded");
DRM_INFO("RE-INIT-early: %s %s\n", block->version->funcs->name, r?"failed":"succeeded");
if (r)
return r;
}
@ -2167,7 +2169,7 @@ static int amdgpu_device_ip_reinit_late_sriov(struct amdgpu_device *adev)
continue;
r = block->version->funcs->hw_init(adev);
DRM_INFO("RE-INIT: %s %s\n", block->version->funcs->name, r?"failed":"succeeded");
DRM_INFO("RE-INIT-late: %s %s\n", block->version->funcs->name, r?"failed":"succeeded");
if (r)
return r;
}
@ -2548,6 +2550,17 @@ int amdgpu_device_init(struct amdgpu_device *adev,
/* detect if we are with an SRIOV vbios */
amdgpu_device_detect_sriov_bios(adev);
/* check if we need to reset the asic
* E.g., driver was not cleanly unloaded previously, etc.
*/
if (!amdgpu_sriov_vf(adev) && amdgpu_asic_need_reset_on_init(adev)) {
r = amdgpu_asic_reset(adev);
if (r) {
dev_err(adev->dev, "asic reset on init failed\n");
goto failed;
}
}
/* Post card if necessary */
if (amdgpu_device_need_post(adev)) {
if (!adev->bios) {
@ -2612,6 +2625,8 @@ fence_driver_init:
}
dev_err(adev->dev, "amdgpu_device_ip_init failed\n");
amdgpu_vf_error_put(adev, AMDGIM_ERROR_VF_AMDGPU_INIT_FAIL, 0, 0);
if (amdgpu_virt_request_full_gpu(adev, false))
amdgpu_virt_release_full_gpu(adev, false);
goto failed;
}
@ -2707,7 +2722,7 @@ void amdgpu_device_fini(struct amdgpu_device *adev)
amdgpu_irq_disable_all(adev);
if (adev->mode_info.mode_config_initialized){
if (!amdgpu_device_has_dc_support(adev))
drm_crtc_force_disable_all(adev->ddev);
drm_helper_force_disable_all(adev->ddev);
else
drm_atomic_helper_shutdown(adev->ddev);
}
@ -3298,17 +3313,15 @@ static int amdgpu_device_pre_asic_reset(struct amdgpu_device *adev,
if (!ring || !ring->sched.thread)
continue;
kthread_park(ring->sched.thread);
if (job && job->base.sched != &ring->sched)
continue;
drm_sched_hw_job_reset(&ring->sched, job ? &job->base : NULL);
drm_sched_stop(&ring->sched);
/* after all hw jobs are reset, hw fence is meaningless, so force_completion */
amdgpu_fence_driver_force_completion(ring);
}
if(job)
drm_sched_increase_karma(&job->base);
if (!amdgpu_sriov_vf(adev)) {
@ -3454,14 +3467,10 @@ static void amdgpu_device_post_asic_reset(struct amdgpu_device *adev,
if (!ring || !ring->sched.thread)
continue;
/* only need recovery sched of the given job's ring
* or all rings (in the case @job is NULL)
* after above amdgpu_reset accomplished
*/
if ((!job || job->base.sched == &ring->sched) && !adev->asic_reset_res)
drm_sched_job_recovery(&ring->sched);
if (!adev->asic_reset_res)
drm_sched_resubmit_jobs(&ring->sched);
kthread_unpark(ring->sched.thread);
drm_sched_start(&ring->sched, !adev->asic_reset_res);
}
if (!amdgpu_device_has_dc_support(adev)) {
@ -3521,9 +3530,9 @@ int amdgpu_device_gpu_recover(struct amdgpu_device *adev,
* by different nodes. No point also since the one node already executing
* reset will also reset all the other nodes in the hive.
*/
hive = amdgpu_get_xgmi_hive(adev);
hive = amdgpu_get_xgmi_hive(adev, 0);
if (hive && adev->gmc.xgmi.num_physical_nodes > 1 &&
!mutex_trylock(&hive->hive_lock))
!mutex_trylock(&hive->reset_lock))
return 0;
/* Start with adev pre asic reset first for soft reset check.*/
@ -3602,13 +3611,45 @@ retry: /* Rest of adevs pre asic reset from XGMI hive. */
}
if (hive && adev->gmc.xgmi.num_physical_nodes > 1)
mutex_unlock(&hive->hive_lock);
mutex_unlock(&hive->reset_lock);
if (r)
dev_info(adev->dev, "GPU reset end with ret = %d\n", r);
return r;
}
static void amdgpu_device_get_min_pci_speed_width(struct amdgpu_device *adev,
enum pci_bus_speed *speed,
enum pcie_link_width *width)
{
struct pci_dev *pdev = adev->pdev;
enum pci_bus_speed cur_speed;
enum pcie_link_width cur_width;
*speed = PCI_SPEED_UNKNOWN;
*width = PCIE_LNK_WIDTH_UNKNOWN;
while (pdev) {
cur_speed = pcie_get_speed_cap(pdev);
cur_width = pcie_get_width_cap(pdev);
if (cur_speed != PCI_SPEED_UNKNOWN) {
if (*speed == PCI_SPEED_UNKNOWN)
*speed = cur_speed;
else if (cur_speed < *speed)
*speed = cur_speed;
}
if (cur_width != PCIE_LNK_WIDTH_UNKNOWN) {
if (*width == PCIE_LNK_WIDTH_UNKNOWN)
*width = cur_width;
else if (cur_width < *width)
*width = cur_width;
}
pdev = pci_upstream_bridge(pdev);
}
}
/**
* amdgpu_device_get_pcie_info - fence pcie info about the PCIE slot
*
@ -3621,8 +3662,8 @@ retry: /* Rest of adevs pre asic reset from XGMI hive. */
static void amdgpu_device_get_pcie_info(struct amdgpu_device *adev)
{
struct pci_dev *pdev;
enum pci_bus_speed speed_cap;
enum pcie_link_width link_width;
enum pci_bus_speed speed_cap, platform_speed_cap;
enum pcie_link_width platform_link_width;
if (amdgpu_pcie_gen_cap)
adev->pm.pcie_gen_mask = amdgpu_pcie_gen_cap;
@ -3639,6 +3680,12 @@ static void amdgpu_device_get_pcie_info(struct amdgpu_device *adev)
return;
}
if (adev->pm.pcie_gen_mask && adev->pm.pcie_mlw_mask)
return;
amdgpu_device_get_min_pci_speed_width(adev, &platform_speed_cap,
&platform_link_width);
if (adev->pm.pcie_gen_mask == 0) {
/* asic caps */
pdev = adev->pdev;
@ -3664,22 +3711,20 @@ static void amdgpu_device_get_pcie_info(struct amdgpu_device *adev)
adev->pm.pcie_gen_mask |= CAIL_ASIC_PCIE_LINK_SPEED_SUPPORT_GEN1;
}
/* platform caps */
pdev = adev->ddev->pdev->bus->self;
speed_cap = pcie_get_speed_cap(pdev);
if (speed_cap == PCI_SPEED_UNKNOWN) {
if (platform_speed_cap == PCI_SPEED_UNKNOWN) {
adev->pm.pcie_gen_mask |= (CAIL_PCIE_LINK_SPEED_SUPPORT_GEN1 |
CAIL_PCIE_LINK_SPEED_SUPPORT_GEN2);
} else {
if (speed_cap == PCIE_SPEED_16_0GT)
if (platform_speed_cap == PCIE_SPEED_16_0GT)
adev->pm.pcie_gen_mask |= (CAIL_PCIE_LINK_SPEED_SUPPORT_GEN1 |
CAIL_PCIE_LINK_SPEED_SUPPORT_GEN2 |
CAIL_PCIE_LINK_SPEED_SUPPORT_GEN3 |
CAIL_PCIE_LINK_SPEED_SUPPORT_GEN4);
else if (speed_cap == PCIE_SPEED_8_0GT)
else if (platform_speed_cap == PCIE_SPEED_8_0GT)
adev->pm.pcie_gen_mask |= (CAIL_PCIE_LINK_SPEED_SUPPORT_GEN1 |
CAIL_PCIE_LINK_SPEED_SUPPORT_GEN2 |
CAIL_PCIE_LINK_SPEED_SUPPORT_GEN3);
else if (speed_cap == PCIE_SPEED_5_0GT)
else if (platform_speed_cap == PCIE_SPEED_5_0GT)
adev->pm.pcie_gen_mask |= (CAIL_PCIE_LINK_SPEED_SUPPORT_GEN1 |
CAIL_PCIE_LINK_SPEED_SUPPORT_GEN2);
else
@ -3688,12 +3733,10 @@ static void amdgpu_device_get_pcie_info(struct amdgpu_device *adev)
}
}
if (adev->pm.pcie_mlw_mask == 0) {
pdev = adev->ddev->pdev->bus->self;
link_width = pcie_get_width_cap(pdev);
if (link_width == PCIE_LNK_WIDTH_UNKNOWN) {
if (platform_link_width == PCIE_LNK_WIDTH_UNKNOWN) {
adev->pm.pcie_mlw_mask |= AMDGPU_DEFAULT_PCIE_MLW_MASK;
} else {
switch (link_width) {
switch (platform_link_width) {
case PCIE_LNK_X32:
adev->pm.pcie_mlw_mask = (CAIL_PCIE_LINK_WIDTH_SUPPORT_X32 |
CAIL_PCIE_LINK_WIDTH_SUPPORT_X16 |

20
drivers/gpu/drm/amd/amdgpu/amdgpu_doorbell.h
View File

@ -51,14 +51,7 @@ struct amdgpu_doorbell_index {
uint32_t userqueue_start;
uint32_t userqueue_end;
uint32_t gfx_ring0;
uint32_t sdma_engine0;
uint32_t sdma_engine1;
uint32_t sdma_engine2;
uint32_t sdma_engine3;
uint32_t sdma_engine4;
uint32_t sdma_engine5;
uint32_t sdma_engine6;
uint32_t sdma_engine7;
uint32_t sdma_engine[8];
uint32_t ih;
union {
struct {
@ -78,7 +71,11 @@ struct amdgpu_doorbell_index {
uint32_t vce_ring6_7;
} uvd_vce;
};
uint32_t first_non_cp;
uint32_t last_non_cp;
uint32_t max_assignment;
/* Per engine SDMA doorbell size in dword */
uint32_t sdma_doorbell_range;
};
typedef enum _AMDGPU_DOORBELL_ASSIGNMENT
@ -148,6 +145,10 @@ typedef enum _AMDGPU_VEGA20_DOORBELL_ASSIGNMENT
AMDGPU_VEGA20_DOORBELL64_VCE_RING2_3 = 0x18D,
AMDGPU_VEGA20_DOORBELL64_VCE_RING4_5 = 0x18E,
AMDGPU_VEGA20_DOORBELL64_VCE_RING6_7 = 0x18F,
AMDGPU_VEGA20_DOORBELL64_FIRST_NON_CP = AMDGPU_VEGA20_DOORBELL_sDMA_ENGINE0,
AMDGPU_VEGA20_DOORBELL64_LAST_NON_CP = AMDGPU_VEGA20_DOORBELL64_VCE_RING6_7,
AMDGPU_VEGA20_DOORBELL_MAX_ASSIGNMENT = 0x18F,
AMDGPU_VEGA20_DOORBELL_INVALID = 0xFFFF
} AMDGPU_VEGA20_DOORBELL_ASSIGNMENT;
@ -227,6 +228,9 @@ typedef enum _AMDGPU_DOORBELL64_ASSIGNMENT
AMDGPU_DOORBELL64_VCE_RING4_5 = 0xFE,
AMDGPU_DOORBELL64_VCE_RING6_7 = 0xFF,
AMDGPU_DOORBELL64_FIRST_NON_CP = AMDGPU_DOORBELL64_sDMA_ENGINE0,
AMDGPU_DOORBELL64_LAST_NON_CP = AMDGPU_DOORBELL64_VCE_RING6_7,
AMDGPU_DOORBELL64_MAX_ASSIGNMENT = 0xFF,
AMDGPU_DOORBELL64_INVALID = 0xFFFF
} AMDGPU_DOORBELL64_ASSIGNMENT;

88
drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.c
View File

@ -184,61 +184,6 @@ u32 amdgpu_dpm_get_vrefresh(struct amdgpu_device *adev)
return vrefresh;
}
void amdgpu_calculate_u_and_p(u32 i, u32 r_c, u32 p_b,
u32 *p, u32 *u)
{
u32 b_c = 0;
u32 i_c;
u32 tmp;
i_c = (i * r_c) / 100;
tmp = i_c >> p_b;
while (tmp) {
b_c++;
tmp >>= 1;
}
*u = (b_c + 1) / 2;
*p = i_c / (1 << (2 * (*u)));
}
int amdgpu_calculate_at(u32 t, u32 h, u32 fh, u32 fl, u32 *tl, u32 *th)
{
u32 k, a, ah, al;
u32 t1;
if ((fl == 0) || (fh == 0) || (fl > fh))
return -EINVAL;
k = (100 * fh) / fl;
t1 = (t * (k - 100));
a = (1000 * (100 * h + t1)) / (10000 + (t1 / 100));
a = (a + 5) / 10;
ah = ((a * t) + 5000) / 10000;
al = a - ah;
*th = t - ah;
*tl = t + al;
return 0;
}
bool amdgpu_is_uvd_state(u32 class, u32 class2)
{
if (class & ATOM_PPLIB_CLASSIFICATION_UVDSTATE)
return true;
if (class & ATOM_PPLIB_CLASSIFICATION_HD2STATE)
return true;
if (class & ATOM_PPLIB_CLASSIFICATION_HDSTATE)
return true;
if (class & ATOM_PPLIB_CLASSIFICATION_SDSTATE)
return true;
if (class2 & ATOM_PPLIB_CLASSIFICATION2_MVC)
return true;
return false;
}
bool amdgpu_is_internal_thermal_sensor(enum amdgpu_int_thermal_type sensor)
{
switch (sensor) {
@ -949,39 +894,6 @@ enum amdgpu_pcie_gen amdgpu_get_pcie_gen_support(struct amdgpu_device *adev,
return AMDGPU_PCIE_GEN1;
}
u16 amdgpu_get_pcie_lane_support(struct amdgpu_device *adev,
u16 asic_lanes,
u16 default_lanes)
{
switch (asic_lanes) {
case 0:
default:
return default_lanes;
case 1:
return 1;
case 2:
return 2;
case 4:
return 4;
case 8:
return 8;
case 12:
return 12;
case 16:
return 16;
}
}
u8 amdgpu_encode_pci_lane_width(u32 lanes)
{
u8 encoded_lanes[] = { 0, 1, 2, 0, 3, 0, 0, 0, 4, 0, 0, 0, 5, 0, 0, 0, 6 };
if (lanes > 16)
return 0;
return encoded_lanes[lanes];
}
struct amd_vce_state*
amdgpu_get_vce_clock_state(void *handle, u32 idx)
{

17
drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h
View File

@ -364,6 +364,14 @@ enum amdgpu_pcie_gen {
((adev)->powerplay.pp_funcs->enable_mgpu_fan_boost(\
(adev)->powerplay.pp_handle))
#define amdgpu_dpm_get_ppfeature_status(adev, buf) \
((adev)->powerplay.pp_funcs->get_ppfeature_status(\
(adev)->powerplay.pp_handle, (buf)))
#define amdgpu_dpm_set_ppfeature_status(adev, ppfeatures) \
((adev)->powerplay.pp_funcs->set_ppfeature_status(\
(adev)->powerplay.pp_handle, (ppfeatures)))
struct amdgpu_dpm {
struct amdgpu_ps *ps;
/* number of valid power states */
@ -478,10 +486,6 @@ void amdgpu_dpm_print_ps_status(struct amdgpu_device *adev,
u32 amdgpu_dpm_get_vblank_time(struct amdgpu_device *adev);
u32 amdgpu_dpm_get_vrefresh(struct amdgpu_device *adev);
void amdgpu_dpm_get_active_displays(struct amdgpu_device *adev);
bool amdgpu_is_uvd_state(u32 class, u32 class2);
void amdgpu_calculate_u_and_p(u32 i, u32 r_c, u32 p_b,
u32 *p, u32 *u);
int amdgpu_calculate_at(u32 t, u32 h, u32 fh, u32 fl, u32 *tl, u32 *th);
bool amdgpu_is_internal_thermal_sensor(enum amdgpu_int_thermal_type sensor);
@ -497,11 +501,6 @@ enum amdgpu_pcie_gen amdgpu_get_pcie_gen_support(struct amdgpu_device *adev,
enum amdgpu_pcie_gen asic_gen,
enum amdgpu_pcie_gen default_gen);
u16 amdgpu_get_pcie_lane_support(struct amdgpu_device *adev,
u16 asic_lanes,
u16 default_lanes);
u8 amdgpu_encode_pci_lane_width(u32 lanes);
struct amd_vce_state*
amdgpu_get_vce_clock_state(void *handle, u32 idx);

25
drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
View File

@ -32,7 +32,7 @@
#include <linux/module.h>
#include <linux/pm_runtime.h>
#include <linux/vga_switcheroo.h>
#include <drm/drm_crtc_helper.h>
#include <drm/drm_probe_helper.h>
#include "amdgpu.h"
#include "amdgpu_irq.h"
@ -71,9 +71,12 @@
* - 3.25.0 - Add support for sensor query info (stable pstate sclk/mclk).
* - 3.26.0 - GFX9: Process AMDGPU_IB_FLAG_TC_WB_NOT_INVALIDATE.
* - 3.27.0 - Add new chunk to to AMDGPU_CS to enable BO_LIST creation.
* - 3.28.0 - Add AMDGPU_CHUNK_ID_SCHEDULED_DEPENDENCIES
* - 3.29.0 - Add AMDGPU_IB_FLAG_RESET_GDS_MAX_WAVE_ID
* - 3.30.0 - Add AMDGPU_SCHED_OP_CONTEXT_PRIORITY_OVERRIDE.
*/
#define KMS_DRIVER_MAJOR 3
#define KMS_DRIVER_MINOR 27
#define KMS_DRIVER_MINOR 30
#define KMS_DRIVER_PATCHLEVEL 0
int amdgpu_vram_limit = 0;
@ -1176,6 +1179,22 @@ static const struct file_operations amdgpu_driver_kms_fops = {
#endif
};
int amdgpu_file_to_fpriv(struct file *filp, struct amdgpu_fpriv **fpriv)
{
struct drm_file *file;
if (!filp)
return -EINVAL;
if (filp->f_op != &amdgpu_driver_kms_fops) {
return -EINVAL;
}
file = filp->private_data;
*fpriv = file->driver_priv;
return 0;
}
static bool
amdgpu_get_crtc_scanout_position(struct drm_device *dev, unsigned int pipe,
bool in_vblank_irq, int *vpos, int *hpos,
@ -1189,7 +1208,7 @@ amdgpu_get_crtc_scanout_position(struct drm_device *dev, unsigned int pipe,
static struct drm_driver kms_driver = {
.driver_features =
DRIVER_USE_AGP | DRIVER_ATOMIC |
DRIVER_HAVE_IRQ | DRIVER_IRQ_SHARED | DRIVER_GEM |
DRIVER_GEM |
DRIVER_PRIME | DRIVER_RENDER | DRIVER_MODESET | DRIVER_SYNCOBJ,
.load = amdgpu_driver_load_kms,
.open = amdgpu_driver_open_kms,

2
drivers/gpu/drm/amd/amdgpu/amdgpu_gds.h
View File

@ -37,6 +37,8 @@ struct amdgpu_gds {
struct amdgpu_gds_asic_info mem;
struct amdgpu_gds_asic_info gws;
struct amdgpu_gds_asic_info oa;
uint32_t gds_compute_max_wave_id;
/* At present, GDS, GWS and OA resources for gfx (graphics)
* is always pre-allocated and available for graphics operation.
* Such resource is shared between all gfx clients.

7
drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
View File

@ -54,10 +54,6 @@ int amdgpu_gem_object_create(struct amdgpu_device *adev, unsigned long size,
memset(&bp, 0, sizeof(bp));
*obj = NULL;
/* At least align on page size */
if (alignment < PAGE_SIZE) {
alignment = PAGE_SIZE;
}
bp.size = size;
bp.byte_align = alignment;
@ -244,9 +240,6 @@ int amdgpu_gem_create_ioctl(struct drm_device *dev, void *data,
return -EINVAL;
}
flags |= AMDGPU_GEM_CREATE_NO_CPU_ACCESS;
/* GDS allocations must be DW aligned */
if (args->in.domains & AMDGPU_GEM_DOMAIN_GDS)
size = ALIGN(size, 4);
}
if (flags & AMDGPU_GEM_CREATE_VM_ALWAYS_VALID) {

10
drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c
View File

@ -202,12 +202,12 @@ int amdgpu_ib_schedule(struct amdgpu_ring *ring, unsigned num_ibs,
amdgpu_asic_flush_hdp(adev, ring);
}
if (need_ctx_switch)
status |= AMDGPU_HAVE_CTX_SWITCH;
skip_preamble = ring->current_ctx == fence_ctx;
if (job && ring->funcs->emit_cntxcntl) {
if (need_ctx_switch)
status |= AMDGPU_HAVE_CTX_SWITCH;
status |= job->preamble_status;
amdgpu_ring_emit_cntxcntl(ring, status);
}
@ -221,8 +221,8 @@ int amdgpu_ib_schedule(struct amdgpu_ring *ring, unsigned num_ibs,
!amdgpu_sriov_vf(adev)) /* for SRIOV preemption, Preamble CE ib must be inserted anyway */
continue;
amdgpu_ring_emit_ib(ring, job, ib, need_ctx_switch);
need_ctx_switch = false;
amdgpu_ring_emit_ib(ring, job, ib, status);
status &= ~AMDGPU_HAVE_CTX_SWITCH;
}
if (ring->funcs->emit_tmz)

46
drivers/gpu/drm/amd/amdgpu/amdgpu_ih.c
View File

@ -52,6 +52,8 @@ int amdgpu_ih_ring_init(struct amdgpu_device *adev, struct amdgpu_ih_ring *ih,
ih->use_bus_addr = use_bus_addr;
if (use_bus_addr) {
dma_addr_t dma_addr;
if (ih->ring)
return 0;
@ -59,21 +61,26 @@ int amdgpu_ih_ring_init(struct amdgpu_device *adev, struct amdgpu_ih_ring *ih,
* add them to the end of the ring allocation.
*/
ih->ring = dma_alloc_coherent(adev->dev, ih->ring_size + 8,
&ih->rb_dma_addr, GFP_KERNEL);
&dma_addr, GFP_KERNEL);
if (ih->ring == NULL)
return -ENOMEM;
memset((void *)ih->ring, 0, ih->ring_size + 8);
ih->wptr_offs = (ih->ring_size / 4) + 0;
ih->rptr_offs = (ih->ring_size / 4) + 1;
ih->gpu_addr = dma_addr;
ih->wptr_addr = dma_addr + ih->ring_size;
ih->wptr_cpu = &ih->ring[ih->ring_size / 4];
ih->rptr_addr = dma_addr + ih->ring_size + 4;
ih->rptr_cpu = &ih->ring[(ih->ring_size / 4) + 1];
} else {
r = amdgpu_device_wb_get(adev, &ih->wptr_offs);
unsigned wptr_offs, rptr_offs;
r = amdgpu_device_wb_get(adev, &wptr_offs);
if (r)
return r;
r = amdgpu_device_wb_get(adev, &ih->rptr_offs);
r = amdgpu_device_wb_get(adev, &rptr_offs);
if (r) {
amdgpu_device_wb_free(adev, ih->wptr_offs);
amdgpu_device_wb_free(adev, wptr_offs);
return r;
}
@ -82,10 +89,15 @@ int amdgpu_ih_ring_init(struct amdgpu_device *adev, struct amdgpu_ih_ring *ih,
&ih->ring_obj, &ih->gpu_addr,
(void **)&ih->ring);
if (r) {
amdgpu_device_wb_free(adev, ih->rptr_offs);
amdgpu_device_wb_free(adev, ih->wptr_offs);
amdgpu_device_wb_free(adev, rptr_offs);
amdgpu_device_wb_free(adev, wptr_offs);
return r;
}
ih->wptr_addr = adev->wb.gpu_addr + wptr_offs * 4;
ih->wptr_cpu = &adev->wb.wb[wptr_offs];
ih->rptr_addr = adev->wb.gpu_addr + rptr_offs * 4;
ih->rptr_cpu = &adev->wb.wb[rptr_offs];
}
return 0;
}
@ -109,13 +121,13 @@ void amdgpu_ih_ring_fini(struct amdgpu_device *adev, struct amdgpu_ih_ring *ih)
* add them to the end of the ring allocation.
*/
dma_free_coherent(adev->dev, ih->ring_size + 8,
(void *)ih->ring, ih->rb_dma_addr);
(void *)ih->ring, ih->gpu_addr);
ih->ring = NULL;
} else {
amdgpu_bo_free_kernel(&ih->ring_obj, &ih->gpu_addr,
(void **)&ih->ring);
amdgpu_device_wb_free(adev, ih->wptr_offs);
amdgpu_device_wb_free(adev, ih->rptr_offs);
amdgpu_device_wb_free(adev, (ih->wptr_addr - ih->gpu_addr) / 4);
amdgpu_device_wb_free(adev, (ih->rptr_addr - ih->gpu_addr) / 4);
}
}
@ -128,16 +140,14 @@ void amdgpu_ih_ring_fini(struct amdgpu_device *adev, struct amdgpu_ih_ring *ih)
* Interrupt hander (VI), walk the IH ring.
* Returns irq process return code.
*/
int amdgpu_ih_process(struct amdgpu_device *adev, struct amdgpu_ih_ring *ih,
void (*callback)(struct amdgpu_device *adev,
struct amdgpu_ih_ring *ih))
int amdgpu_ih_process(struct amdgpu_device *adev, struct amdgpu_ih_ring *ih)
{
u32 wptr;
if (!ih->enabled || adev->shutdown)
return IRQ_NONE;
wptr = amdgpu_ih_get_wptr(adev);
wptr = amdgpu_ih_get_wptr(adev, ih);
restart_ih:
/* is somebody else already processing irqs? */
@ -150,15 +160,15 @@ restart_ih:
rmb();
while (ih->rptr != wptr) {
callback(adev, ih);
amdgpu_irq_dispatch(adev, ih);
ih->rptr &= ih->ptr_mask;
}
amdgpu_ih_set_rptr(adev);
amdgpu_ih_set_rptr(adev, ih);
atomic_set(&ih->lock, 0);
/* make sure wptr hasn't changed while processing */
wptr = amdgpu_ih_get_wptr(adev);
wptr = amdgpu_ih_get_wptr(adev, ih);
if (wptr != ih->rptr)
goto restart_ih;

40
drivers/gpu/drm/amd/amdgpu/amdgpu_ih.h
View File

@ -31,40 +31,44 @@ struct amdgpu_iv_entry;
* R6xx+ IH ring
*/
struct amdgpu_ih_ring {
struct amdgpu_bo *ring_obj;
volatile uint32_t *ring;
unsigned rptr;
unsigned ring_size;
uint64_t gpu_addr;
uint32_t ptr_mask;
atomic_t lock;
bool enabled;
unsigned wptr_offs;
unsigned rptr_offs;
u32 doorbell_index;
bool use_doorbell;
bool use_bus_addr;
dma_addr_t rb_dma_addr; /* only used when use_bus_addr = true */
struct amdgpu_bo *ring_obj;
volatile uint32_t *ring;
uint64_t gpu_addr;
uint64_t wptr_addr;
volatile uint32_t *wptr_cpu;
uint64_t rptr_addr;
volatile uint32_t *rptr_cpu;
bool enabled;
unsigned rptr;
atomic_t lock;
};
/* provided by the ih block */
struct amdgpu_ih_funcs {
/* ring read/write ptr handling, called from interrupt context */
u32 (*get_wptr)(struct amdgpu_device *adev);
void (*decode_iv)(struct amdgpu_device *adev,
u32 (*get_wptr)(struct amdgpu_device *adev, struct amdgpu_ih_ring *ih);
void (*decode_iv)(struct amdgpu_device *adev, struct amdgpu_ih_ring *ih,
struct amdgpu_iv_entry *entry);
void (*set_rptr)(struct amdgpu_device *adev);
void (*set_rptr)(struct amdgpu_device *adev, struct amdgpu_ih_ring *ih);
};
#define amdgpu_ih_get_wptr(adev) (adev)->irq.ih_funcs->get_wptr((adev))
#define amdgpu_ih_decode_iv(adev, iv) (adev)->irq.ih_funcs->decode_iv((adev), (iv))
#define amdgpu_ih_set_rptr(adev) (adev)->irq.ih_funcs->set_rptr((adev))
#define amdgpu_ih_get_wptr(adev, ih) (adev)->irq.ih_funcs->get_wptr((adev), (ih))
#define amdgpu_ih_decode_iv(adev, iv) \
(adev)->irq.ih_funcs->decode_iv((adev), (ih), (iv))
#define amdgpu_ih_set_rptr(adev, ih) (adev)->irq.ih_funcs->set_rptr((adev), (ih))
int amdgpu_ih_ring_init(struct amdgpu_device *adev, struct amdgpu_ih_ring *ih,
unsigned ring_size, bool use_bus_addr);
void amdgpu_ih_ring_fini(struct amdgpu_device *adev, struct amdgpu_ih_ring *ih);
int amdgpu_ih_process(struct amdgpu_device *adev, struct amdgpu_ih_ring *ih,
void (*callback)(struct amdgpu_device *adev,
struct amdgpu_ih_ring *ih));
int amdgpu_ih_process(struct amdgpu_device *adev, struct amdgpu_ih_ring *ih);
#endif

75
drivers/gpu/drm/amd/amdgpu/amdgpu_irq.c
View File

@ -130,27 +130,6 @@ void amdgpu_irq_disable_all(struct amdgpu_device *adev)
spin_unlock_irqrestore(&adev->irq.lock, irqflags);
}
/**
* amdgpu_irq_callback - callback from the IH ring
*
* @adev: amdgpu device pointer
* @ih: amdgpu ih ring
*
* Callback from IH ring processing to handle the entry at the current position
* and advance the read pointer.
*/
static void amdgpu_irq_callback(struct amdgpu_device *adev,
struct amdgpu_ih_ring *ih)
{
u32 ring_index = ih->rptr >> 2;
struct amdgpu_iv_entry entry;
entry.iv_entry = (const uint32_t *)&ih->ring[ring_index];
amdgpu_ih_decode_iv(adev, &entry);
amdgpu_irq_dispatch(adev, &entry);
}
/**
* amdgpu_irq_handler - IRQ handler
*
@ -168,12 +147,42 @@ irqreturn_t amdgpu_irq_handler(int irq, void *arg)
struct amdgpu_device *adev = dev->dev_private;
irqreturn_t ret;
ret = amdgpu_ih_process(adev, &adev->irq.ih, amdgpu_irq_callback);
ret = amdgpu_ih_process(adev, &adev->irq.ih);
if (ret == IRQ_HANDLED)
pm_runtime_mark_last_busy(dev->dev);
return ret;
}
/**
* amdgpu_irq_handle_ih1 - kick of processing for IH1
*
* @work: work structure in struct amdgpu_irq
*
* Kick of processing IH ring 1.
*/
static void amdgpu_irq_handle_ih1(struct work_struct *work)
{
struct amdgpu_device *adev = container_of(work, struct amdgpu_device,
irq.ih1_work);
amdgpu_ih_process(adev, &adev->irq.ih1);
}
/**
* amdgpu_irq_handle_ih2 - kick of processing for IH2
*
* @work: work structure in struct amdgpu_irq
*
* Kick of processing IH ring 2.
*/
static void amdgpu_irq_handle_ih2(struct work_struct *work)
{
struct amdgpu_device *adev = container_of(work, struct amdgpu_device,
irq.ih2_work);
amdgpu_ih_process(adev, &adev->irq.ih2);
}
/**
* amdgpu_msi_ok - check whether MSI functionality is enabled
*
@ -238,6 +247,9 @@ int amdgpu_irq_init(struct amdgpu_device *adev)
amdgpu_hotplug_work_func);
}
INIT_WORK(&adev->irq.ih1_work, amdgpu_irq_handle_ih1);
INIT_WORK(&adev->irq.ih2_work, amdgpu_irq_handle_ih2);
adev->irq.installed = true;
r = drm_irq_install(adev->ddev, adev->ddev->pdev->irq);
if (r) {
@ -359,15 +371,22 @@ int amdgpu_irq_add_id(struct amdgpu_device *adev,
* Dispatches IRQ to IP blocks.
*/
void amdgpu_irq_dispatch(struct amdgpu_device *adev,
struct amdgpu_iv_entry *entry)
struct amdgpu_ih_ring *ih)
{
unsigned client_id = entry->client_id;
unsigned src_id = entry->src_id;
u32 ring_index = ih->rptr >> 2;
struct amdgpu_iv_entry entry;
unsigned client_id, src_id;
struct amdgpu_irq_src *src;
bool handled = false;
int r;
trace_amdgpu_iv(entry);
entry.iv_entry = (const uint32_t *)&ih->ring[ring_index];
amdgpu_ih_decode_iv(adev, &entry);
trace_amdgpu_iv(ih - &adev->irq.ih, &entry);
client_id = entry.client_id;
src_id = entry.src_id;
if (client_id >= AMDGPU_IRQ_CLIENTID_MAX) {
DRM_DEBUG("Invalid client_id in IV: %d\n", client_id);
@ -383,7 +402,7 @@ void amdgpu_irq_dispatch(struct amdgpu_device *adev,
client_id, src_id);
} else if ((src = adev->irq.client[client_id].sources[src_id])) {
r = src->funcs->process(adev, src, entry);
r = src->funcs->process(adev, src, &entry);
if (r < 0)
DRM_ERROR("error processing interrupt (%d)\n", r);
else if (r)
@ -395,7 +414,7 @@ void amdgpu_irq_dispatch(struct amdgpu_device *adev,
/* Send it to amdkfd as well if it isn't already handled */
if (!handled)
amdgpu_amdkfd_interrupt(adev, entry->iv_entry);
amdgpu_amdkfd_interrupt(adev, entry.iv_entry);
}
/**

10
drivers/gpu/drm/amd/amdgpu/amdgpu_irq.h
View File

@ -87,9 +87,11 @@ struct amdgpu_irq {
/* status, etc. */
bool msi_enabled; /* msi enabled */
/* interrupt ring */
struct amdgpu_ih_ring ih;
const struct amdgpu_ih_funcs *ih_funcs;
/* interrupt rings */
struct amdgpu_ih_ring ih, ih1, ih2;
const struct amdgpu_ih_funcs *ih_funcs;
struct work_struct ih1_work, ih2_work;
struct amdgpu_irq_src self_irq;
/* gen irq stuff */
struct irq_domain *domain; /* GPU irq controller domain */
@ -106,7 +108,7 @@ int amdgpu_irq_add_id(struct amdgpu_device *adev,
unsigned client_id, unsigned src_id,
struct amdgpu_irq_src *source);
void amdgpu_irq_dispatch(struct amdgpu_device *adev,
struct amdgpu_iv_entry *entry);
struct amdgpu_ih_ring *ih);
int amdgpu_irq_update(struct amdgpu_device *adev, struct amdgpu_irq_src *src,
unsigned type);
int amdgpu_irq_get(struct amdgpu_device *adev, struct amdgpu_irq_src *src,

2
drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
View File

@ -207,7 +207,7 @@ int amdgpu_driver_load_kms(struct drm_device *dev, unsigned long flags)
if (!r) {
acpi_status = amdgpu_acpi_init(adev);
if (acpi_status)
dev_dbg(&dev->pdev->dev,
dev_dbg(&dev->pdev->dev,
"Error during ACPI methods call\n");
}

1
drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h
View File

@ -38,6 +38,7 @@
#include <drm/drm_crtc_helper.h>
#include <drm/drm_fb_helper.h>
#include <drm/drm_plane_helper.h>
#include <drm/drm_probe_helper.h>
#include <linux/i2c.h>
#include <linux/i2c-algo-bit.h>
#include <linux/hrtimer.h>

40
drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
View File

@ -426,12 +426,20 @@ static int amdgpu_bo_do_create(struct amdgpu_device *adev,
size_t acc_size;
int r;
page_align = roundup(bp->byte_align, PAGE_SIZE) >> PAGE_SHIFT;
if (bp->domain & (AMDGPU_GEM_DOMAIN_GDS | AMDGPU_GEM_DOMAIN_GWS |
AMDGPU_GEM_DOMAIN_OA))
/* Note that GDS/GWS/OA allocates 1 page per byte/resource. */
if (bp->domain & (AMDGPU_GEM_DOMAIN_GWS | AMDGPU_GEM_DOMAIN_OA)) {
/* GWS and OA don't need any alignment. */
page_align = bp->byte_align;
size <<= PAGE_SHIFT;
else
} else if (bp->domain & AMDGPU_GEM_DOMAIN_GDS) {
/* Both size and alignment must be a multiple of 4. */
page_align = ALIGN(bp->byte_align, 4);
size = ALIGN(size, 4) << PAGE_SHIFT;
} else {
/* Memory should be aligned at least to a page size. */
page_align = ALIGN(bp->byte_align, PAGE_SIZE) >> PAGE_SHIFT;
size = ALIGN(size, PAGE_SIZE);
}
if (!amdgpu_bo_validate_size(adev, size, bp->domain))
return -ENOMEM;
@ -1276,6 +1284,30 @@ void amdgpu_bo_fence(struct amdgpu_bo *bo, struct dma_fence *fence,
reservation_object_add_excl_fence(resv, fence);
}
/**
* amdgpu_sync_wait_resv - Wait for BO reservation fences
*
* @bo: buffer object
* @owner: fence owner
* @intr: Whether the wait is interruptible
*
* Returns:
* 0 on success, errno otherwise.
*/
int amdgpu_bo_sync_wait(struct amdgpu_bo *bo, void *owner, bool intr)
{
struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev);
struct amdgpu_sync sync;
int r;
amdgpu_sync_create(&sync);
amdgpu_sync_resv(adev, &sync, bo->tbo.resv, owner, false);
r = amdgpu_sync_wait(&sync, intr);
amdgpu_sync_free(&sync);
return r;
}
/**
* amdgpu_bo_gpu_offset - return GPU offset of bo
* @bo: amdgpu object for which we query the offset

1
drivers/gpu/drm/amd/amdgpu/amdgpu_object.h
View File

@ -266,6 +266,7 @@ void amdgpu_bo_move_notify(struct ttm_buffer_object *bo,
int amdgpu_bo_fault_reserve_notify(struct ttm_buffer_object *bo);
void amdgpu_bo_fence(struct amdgpu_bo *bo, struct dma_fence *fence,
bool shared);
int amdgpu_bo_sync_wait(struct amdgpu_bo *bo, void *owner, bool intr);
u64 amdgpu_bo_gpu_offset(struct amdgpu_bo *bo);
int amdgpu_bo_validate(struct amdgpu_bo *bo);
int amdgpu_bo_restore_shadow(struct amdgpu_bo *shadow,

354
drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
View File

@ -626,11 +626,71 @@ static ssize_t amdgpu_get_pp_od_clk_voltage(struct device *dev,
}
/**
* DOC: pp_dpm_sclk pp_dpm_mclk pp_dpm_pcie
* DOC: ppfeatures
*
* The amdgpu driver provides a sysfs API for adjusting what powerplay
* features to be enabled. The file ppfeatures is used for this. And
* this is only available for Vega10 and later dGPUs.
*
* Reading back the file will show you the followings:
* - Current ppfeature masks
* - List of the all supported powerplay features with their naming,
* bitmasks and enablement status('Y'/'N' means "enabled"/"disabled").
*
* To manually enable or disable a specific feature, just set or clear
* the corresponding bit from original ppfeature masks and input the
* new ppfeature masks.
*/
static ssize_t amdgpu_set_ppfeature_status(struct device *dev,
struct device_attribute *attr,
const char *buf,
size_t count)
{
struct drm_device *ddev = dev_get_drvdata(dev);
struct amdgpu_device *adev = ddev->dev_private;
uint64_t featuremask;
int ret;
ret = kstrtou64(buf, 0, &featuremask);
if (ret)
return -EINVAL;
pr_debug("featuremask = 0x%llx\n", featuremask);
if (adev->powerplay.pp_funcs->set_ppfeature_status) {
ret = amdgpu_dpm_set_ppfeature_status(adev, featuremask);
if (ret)
return -EINVAL;
}
return count;
}
static ssize_t amdgpu_get_ppfeature_status(struct device *dev,
struct device_attribute *attr,
char *buf)
{
struct drm_device *ddev = dev_get_drvdata(dev);
struct amdgpu_device *adev = ddev->dev_private;
if (adev->powerplay.pp_funcs->get_ppfeature_status)
return amdgpu_dpm_get_ppfeature_status(adev, buf);
return snprintf(buf, PAGE_SIZE, "\n");
}
/**
* DOC: pp_dpm_sclk pp_dpm_mclk pp_dpm_socclk pp_dpm_fclk pp_dpm_dcefclk
* pp_dpm_pcie
*
* The amdgpu driver provides a sysfs API for adjusting what power levels
* are enabled for a given power state. The files pp_dpm_sclk, pp_dpm_mclk,
* and pp_dpm_pcie are used for this.
* pp_dpm_socclk, pp_dpm_fclk, pp_dpm_dcefclk and pp_dpm_pcie are used for
* this.
*
* pp_dpm_socclk and pp_dpm_dcefclk interfaces are only available for
* Vega10 and later ASICs.
* pp_dpm_fclk interface is only available for Vega20 and later ASICs.
*
* Reading back the files will show you the available power levels within
* the power state and the clock information for those levels.
@ -640,6 +700,8 @@ static ssize_t amdgpu_get_pp_od_clk_voltage(struct device *dev,
* Secondly,Enter a new value for each level by inputing a string that
* contains " echo xx xx xx > pp_dpm_sclk/mclk/pcie"
* E.g., echo 4 5 6 to > pp_dpm_sclk will enable sclk levels 4, 5, and 6.
*
* NOTE: change to the dcefclk max dpm level is not supported now
*/
static ssize_t amdgpu_get_pp_dpm_sclk(struct device *dev,
@ -750,6 +812,114 @@ static ssize_t amdgpu_set_pp_dpm_mclk(struct device *dev,
return count;
}
static ssize_t amdgpu_get_pp_dpm_socclk(struct device *dev,
struct device_attribute *attr,
char *buf)
{
struct drm_device *ddev = dev_get_drvdata(dev);
struct amdgpu_device *adev = ddev->dev_private;
if (adev->powerplay.pp_funcs->print_clock_levels)
return amdgpu_dpm_print_clock_levels(adev, PP_SOCCLK, buf);
else
return snprintf(buf, PAGE_SIZE, "\n");
}
static ssize_t amdgpu_set_pp_dpm_socclk(struct device *dev,
struct device_attribute *attr,
const char *buf,
size_t count)
{
struct drm_device *ddev = dev_get_drvdata(dev);
struct amdgpu_device *adev = ddev->dev_private;
int ret;
uint32_t mask = 0;
ret = amdgpu_read_mask(buf, count, &mask);
if (ret)
return ret;
if (adev->powerplay.pp_funcs->force_clock_level)
ret = amdgpu_dpm_force_clock_level(adev, PP_SOCCLK, mask);
if (ret)
return -EINVAL;
return count;
}
static ssize_t amdgpu_get_pp_dpm_fclk(struct device *dev,
struct device_attribute *attr,
char *buf)
{
struct drm_device *ddev = dev_get_drvdata(dev);
struct amdgpu_device *adev = ddev->dev_private;
if (adev->powerplay.pp_funcs->print_clock_levels)
return amdgpu_dpm_print_clock_levels(adev, PP_FCLK, buf);
else
return snprintf(buf, PAGE_SIZE, "\n");
}
static ssize_t amdgpu_set_pp_dpm_fclk(struct device *dev,
struct device_attribute *attr,
const char *buf,
size_t count)
{
struct drm_device *ddev = dev_get_drvdata(dev);
struct amdgpu_device *adev = ddev->dev_private;
int ret;
uint32_t mask = 0;
ret = amdgpu_read_mask(buf, count, &mask);
if (ret)
return ret;
if (adev->powerplay.pp_funcs->force_clock_level)
ret = amdgpu_dpm_force_clock_level(adev, PP_FCLK, mask);
if (ret)
return -EINVAL;
return count;
}
static ssize_t amdgpu_get_pp_dpm_dcefclk(struct device *dev,
struct device_attribute *attr,
char *buf)
{
struct drm_device *ddev = dev_get_drvdata(dev);
struct amdgpu_device *adev = ddev->dev_private;
if (adev->powerplay.pp_funcs->print_clock_levels)
return amdgpu_dpm_print_clock_levels(adev, PP_DCEFCLK, buf);
else
return snprintf(buf, PAGE_SIZE, "\n");
}
static ssize_t amdgpu_set_pp_dpm_dcefclk(struct device *dev,
struct device_attribute *attr,
const char *buf,
size_t count)
{
struct drm_device *ddev = dev_get_drvdata(dev);
struct amdgpu_device *adev = ddev->dev_private;
int ret;
uint32_t mask = 0;
ret = amdgpu_read_mask(buf, count, &mask);
if (ret)
return ret;
if (adev->powerplay.pp_funcs->force_clock_level)
ret = amdgpu_dpm_force_clock_level(adev, PP_DCEFCLK, mask);
if (ret)
return -EINVAL;
return count;
}
static ssize_t amdgpu_get_pp_dpm_pcie(struct device *dev,
struct device_attribute *attr,
char *buf)
@ -990,6 +1160,31 @@ static ssize_t amdgpu_get_busy_percent(struct device *dev,
return snprintf(buf, PAGE_SIZE, "%d\n", value);
}
/**
* DOC: pcie_bw
*
* The amdgpu driver provides a sysfs API for estimating how much data
* has been received and sent by the GPU in the last second through PCIe.
* The file pcie_bw is used for this.
* The Perf counters count the number of received and sent messages and return
* those values, as well as the maximum payload size of a PCIe packet (mps).
* Note that it is not possible to easily and quickly obtain the size of each
* packet transmitted, so we output the max payload size (mps) to allow for
* quick estimation of the PCIe bandwidth usage
*/
static ssize_t amdgpu_get_pcie_bw(struct device *dev,
struct device_attribute *attr,
char *buf)
{
struct drm_device *ddev = dev_get_drvdata(dev);
struct amdgpu_device *adev = ddev->dev_private;
uint64_t count0, count1;
amdgpu_asic_get_pcie_usage(adev, &count0, &count1);
return snprintf(buf, PAGE_SIZE, "%llu %llu %i\n",
count0, count1, pcie_get_mps(adev->pdev));
}
static DEVICE_ATTR(power_dpm_state, S_IRUGO | S_IWUSR, amdgpu_get_dpm_state, amdgpu_set_dpm_state);
static DEVICE_ATTR(power_dpm_force_performance_level, S_IRUGO | S_IWUSR,
amdgpu_get_dpm_forced_performance_level,
@ -1008,6 +1203,15 @@ static DEVICE_ATTR(pp_dpm_sclk, S_IRUGO | S_IWUSR,
static DEVICE_ATTR(pp_dpm_mclk, S_IRUGO | S_IWUSR,
amdgpu_get_pp_dpm_mclk,
amdgpu_set_pp_dpm_mclk);
static DEVICE_ATTR(pp_dpm_socclk, S_IRUGO | S_IWUSR,
amdgpu_get_pp_dpm_socclk,
amdgpu_set_pp_dpm_socclk);
static DEVICE_ATTR(pp_dpm_fclk, S_IRUGO | S_IWUSR,
amdgpu_get_pp_dpm_fclk,
amdgpu_set_pp_dpm_fclk);
static DEVICE_ATTR(pp_dpm_dcefclk, S_IRUGO | S_IWUSR,
amdgpu_get_pp_dpm_dcefclk,
amdgpu_set_pp_dpm_dcefclk);
static DEVICE_ATTR(pp_dpm_pcie, S_IRUGO | S_IWUSR,
amdgpu_get_pp_dpm_pcie,
amdgpu_set_pp_dpm_pcie);
@ -1025,6 +1229,10 @@ static DEVICE_ATTR(pp_od_clk_voltage, S_IRUGO | S_IWUSR,
amdgpu_set_pp_od_clk_voltage);
static DEVICE_ATTR(gpu_busy_percent, S_IRUGO,
amdgpu_get_busy_percent, NULL);
static DEVICE_ATTR(pcie_bw, S_IRUGO, amdgpu_get_pcie_bw, NULL);
static DEVICE_ATTR(ppfeatures, S_IRUGO | S_IWUSR,
amdgpu_get_ppfeature_status,
amdgpu_set_ppfeature_status);
static ssize_t amdgpu_hwmon_show_temp(struct device *dev,
struct device_attribute *attr,
@ -1516,6 +1724,75 @@ static ssize_t amdgpu_hwmon_set_power_cap(struct device *dev,
return count;
}
static ssize_t amdgpu_hwmon_show_sclk(struct device *dev,
struct device_attribute *attr,
char *buf)
{
struct amdgpu_device *adev = dev_get_drvdata(dev);
struct drm_device *ddev = adev->ddev;
uint32_t sclk;
int r, size = sizeof(sclk);
/* Can't get voltage when the card is off */
if ((adev->flags & AMD_IS_PX) &&
(ddev->switch_power_state != DRM_SWITCH_POWER_ON))
return -EINVAL;
/* sanity check PP is enabled */
if (!(adev->powerplay.pp_funcs &&
adev->powerplay.pp_funcs->read_sensor))
return -EINVAL;
/* get the sclk */
r = amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_GFX_SCLK,
(void *)&sclk, &size);
if (r)
return r;
return snprintf(buf, PAGE_SIZE, "%d\n", sclk * 10 * 1000);
}
static ssize_t amdgpu_hwmon_show_sclk_label(struct device *dev,
struct device_attribute *attr,
char *buf)
{
return snprintf(buf, PAGE_SIZE, "sclk\n");
}
static ssize_t amdgpu_hwmon_show_mclk(struct device *dev,
struct device_attribute *attr,
char *buf)
{
struct amdgpu_device *adev = dev_get_drvdata(dev);
struct drm_device *ddev = adev->ddev;
uint32_t mclk;
int r, size = sizeof(mclk);
/* Can't get voltage when the card is off */
if ((adev->flags & AMD_IS_PX) &&
(ddev->switch_power_state != DRM_SWITCH_POWER_ON))
return -EINVAL;
/* sanity check PP is enabled */
if (!(adev->powerplay.pp_funcs &&
adev->powerplay.pp_funcs->read_sensor))
return -EINVAL;
/* get the sclk */
r = amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_GFX_MCLK,
(void *)&mclk, &size);
if (r)
return r;
return snprintf(buf, PAGE_SIZE, "%d\n", mclk * 10 * 1000);
}
static ssize_t amdgpu_hwmon_show_mclk_label(struct device *dev,
struct device_attribute *attr,
char *buf)
{
return snprintf(buf, PAGE_SIZE, "mclk\n");
}
/**
* DOC: hwmon
@ -1532,6 +1809,10 @@ static ssize_t amdgpu_hwmon_set_power_cap(struct device *dev,
*
* - GPU fan
*
* - GPU gfx/compute engine clock
*
* - GPU memory clock (dGPU only)
*
* hwmon interfaces for GPU temperature:
*
* - temp1_input: the on die GPU temperature in millidegrees Celsius
@ -1576,6 +1857,12 @@ static ssize_t amdgpu_hwmon_set_power_cap(struct device *dev,
*
* - fan[1-*]_enable: Enable or disable the sensors.1: Enable 0: Disable
*
* hwmon interfaces for GPU clocks:
*
* - freq1_input: the gfx/compute clock in hertz
*
* - freq2_input: the memory clock in hertz
*
* You can use hwmon tools like sensors to view this information on your system.
*
*/
@ -1600,6 +1887,10 @@ static SENSOR_DEVICE_ATTR(power1_average, S_IRUGO, amdgpu_hwmon_show_power_avg,
static SENSOR_DEVICE_ATTR(power1_cap_max, S_IRUGO, amdgpu_hwmon_show_power_cap_max, NULL, 0);
static SENSOR_DEVICE_ATTR(power1_cap_min, S_IRUGO, amdgpu_hwmon_show_power_cap_min, NULL, 0);
static SENSOR_DEVICE_ATTR(power1_cap, S_IRUGO | S_IWUSR, amdgpu_hwmon_show_power_cap, amdgpu_hwmon_set_power_cap, 0);
static SENSOR_DEVICE_ATTR(freq1_input, S_IRUGO, amdgpu_hwmon_show_sclk, NULL, 0);
static SENSOR_DEVICE_ATTR(freq1_label, S_IRUGO, amdgpu_hwmon_show_sclk_label, NULL, 0);
static SENSOR_DEVICE_ATTR(freq2_input, S_IRUGO, amdgpu_hwmon_show_mclk, NULL, 0);
static SENSOR_DEVICE_ATTR(freq2_label, S_IRUGO, amdgpu_hwmon_show_mclk_label, NULL, 0);
static struct attribute *hwmon_attributes[] = {
&sensor_dev_attr_temp1_input.dev_attr.attr,
@ -1622,6 +1913,10 @@ static struct attribute *hwmon_attributes[] = {
&sensor_dev_attr_power1_cap_max.dev_attr.attr,
&sensor_dev_attr_power1_cap_min.dev_attr.attr,
&sensor_dev_attr_power1_cap.dev_attr.attr,
&sensor_dev_attr_freq1_input.dev_attr.attr,
&sensor_dev_attr_freq1_label.dev_attr.attr,
&sensor_dev_attr_freq2_input.dev_attr.attr,
&sensor_dev_attr_freq2_label.dev_attr.attr,
NULL
};
@ -1713,6 +2008,12 @@ static umode_t hwmon_attributes_visible(struct kobject *kobj,
attr == &sensor_dev_attr_in1_label.dev_attr.attr))
return 0;
/* no mclk on APUs */
if ((adev->flags & AMD_IS_APU) &&
(attr == &sensor_dev_attr_freq2_input.dev_attr.attr ||
attr == &sensor_dev_attr_freq2_label.dev_attr.attr))
return 0;
return effective_mode;
}
@ -2071,6 +2372,25 @@ int amdgpu_pm_sysfs_init(struct amdgpu_device *adev)
DRM_ERROR("failed to create device file pp_dpm_mclk\n");
return ret;
}
if (adev->asic_type >= CHIP_VEGA10) {
ret = device_create_file(adev->dev, &dev_attr_pp_dpm_socclk);
if (ret) {
DRM_ERROR("failed to create device file pp_dpm_socclk\n");
return ret;
}
ret = device_create_file(adev->dev, &dev_attr_pp_dpm_dcefclk);
if (ret) {
DRM_ERROR("failed to create device file pp_dpm_dcefclk\n");
return ret;
}
}
if (adev->asic_type >= CHIP_VEGA20) {
ret = device_create_file(adev->dev, &dev_attr_pp_dpm_fclk);
if (ret) {
DRM_ERROR("failed to create device file pp_dpm_fclk\n");
return ret;
}
}
ret = device_create_file(adev->dev, &dev_attr_pp_dpm_pcie);
if (ret) {
DRM_ERROR("failed to create device file pp_dpm_pcie\n");
@ -2109,12 +2429,31 @@ int amdgpu_pm_sysfs_init(struct amdgpu_device *adev)
"gpu_busy_level\n");
return ret;
}
/* PCIe Perf counters won't work on APU nodes */
if (!(adev->flags & AMD_IS_APU)) {
ret = device_create_file(adev->dev, &dev_attr_pcie_bw);
if (ret) {
DRM_ERROR("failed to create device file pcie_bw\n");
return ret;
}
}
ret = amdgpu_debugfs_pm_init(adev);
if (ret) {
DRM_ERROR("Failed to register debugfs file for dpm!\n");
return ret;
}
if ((adev->asic_type >= CHIP_VEGA10) &&
!(adev->flags & AMD_IS_APU)) {
ret = device_create_file(adev->dev,
&dev_attr_ppfeatures);
if (ret) {
DRM_ERROR("failed to create device file "
"ppfeatures\n");
return ret;
}
}
adev->pm.sysfs_initialized = true;
return 0;
@ -2139,7 +2478,13 @@ void amdgpu_pm_sysfs_fini(struct amdgpu_device *adev)
device_remove_file(adev->dev, &dev_attr_pp_dpm_sclk);
device_remove_file(adev->dev, &dev_attr_pp_dpm_mclk);
if (adev->asic_type >= CHIP_VEGA10) {
device_remove_file(adev->dev, &dev_attr_pp_dpm_socclk);
device_remove_file(adev->dev, &dev_attr_pp_dpm_dcefclk);
}
device_remove_file(adev->dev, &dev_attr_pp_dpm_pcie);
if (adev->asic_type >= CHIP_VEGA20)
device_remove_file(adev->dev, &dev_attr_pp_dpm_fclk);
device_remove_file(adev->dev, &dev_attr_pp_sclk_od);
device_remove_file(adev->dev, &dev_attr_pp_mclk_od);
device_remove_file(adev->dev,
@ -2148,6 +2493,11 @@ void amdgpu_pm_sysfs_fini(struct amdgpu_device *adev)
device_remove_file(adev->dev,
&dev_attr_pp_od_clk_voltage);
device_remove_file(adev->dev, &dev_attr_gpu_busy_percent);
if (!(adev->flags & AMD_IS_APU))
device_remove_file(adev->dev, &dev_attr_pcie_bw);
if ((adev->asic_type >= CHIP_VEGA10) &&
!(adev->flags & AMD_IS_APU))
device_remove_file(adev->dev, &dev_attr_ppfeatures);
}
void amdgpu_pm_compute_clocks(struct amdgpu_device *adev)

141
drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c
View File

@ -67,9 +67,6 @@ static int psp_sw_init(void *handle)
psp->adev = adev;
if (adev->firmware.load_type != AMDGPU_FW_LOAD_PSP)
return 0;
ret = psp_init_microcode(psp);
if (ret) {
DRM_ERROR("Failed to load psp firmware!\n");
@ -83,9 +80,6 @@ static int psp_sw_fini(void *handle)
{
struct amdgpu_device *adev = (struct amdgpu_device *)handle;
if (adev->firmware.load_type != AMDGPU_FW_LOAD_PSP)
return 0;
release_firmware(adev->psp.sos_fw);
adev->psp.sos_fw = NULL;
release_firmware(adev->psp.asd_fw);
@ -142,13 +136,24 @@ psp_cmd_submit_buf(struct psp_context *psp,
while (*((unsigned int *)psp->fence_buf) != index)
msleep(1);
/* the status field must be 0 after FW is loaded */
if (ucode && psp->cmd_buf_mem->resp.status) {
DRM_ERROR("failed loading with status (%d) and ucode id (%d)\n",
psp->cmd_buf_mem->resp.status, ucode->ucode_id);
return -EINVAL;
/* In some cases, psp response status is not 0 even there is no
* problem while the command is submitted. Some version of PSP FW
* doesn't write 0 to that field.
* So here we would like to only print a warning instead of an error
* during psp initialization to avoid breaking hw_init and it doesn't
* return -EINVAL.
*/
if (psp->cmd_buf_mem->resp.status) {
if (ucode)
DRM_WARN("failed to load ucode id (%d) ",
ucode->ucode_id);
DRM_WARN("psp command failed and response status is (%d)\n",
psp->cmd_buf_mem->resp.status);
}
/* get xGMI session id from response buffer */
cmd->resp.session_id = psp->cmd_buf_mem->resp.session_id;
if (ucode) {
ucode->tmr_mc_addr_lo = psp->cmd_buf_mem->resp.fw_addr_lo;
ucode->tmr_mc_addr_hi = psp->cmd_buf_mem->resp.fw_addr_hi;
@ -500,6 +505,98 @@ static int psp_hw_start(struct psp_context *psp)
return 0;
}
static int psp_get_fw_type(struct amdgpu_firmware_info *ucode,
enum psp_gfx_fw_type *type)
{
switch (ucode->ucode_id) {
case AMDGPU_UCODE_ID_SDMA0:
*type = GFX_FW_TYPE_SDMA0;
break;
case AMDGPU_UCODE_ID_SDMA1:
*type = GFX_FW_TYPE_SDMA1;
break;
case AMDGPU_UCODE_ID_CP_CE:
*type = GFX_FW_TYPE_CP_CE;
break;
case AMDGPU_UCODE_ID_CP_PFP:
*type = GFX_FW_TYPE_CP_PFP;
break;
case AMDGPU_UCODE_ID_CP_ME:
*type = GFX_FW_TYPE_CP_ME;
break;
case AMDGPU_UCODE_ID_CP_MEC1:
*type = GFX_FW_TYPE_CP_MEC;
break;
case AMDGPU_UCODE_ID_CP_MEC1_JT:
*type = GFX_FW_TYPE_CP_MEC_ME1;
break;
case AMDGPU_UCODE_ID_CP_MEC2:
*type = GFX_FW_TYPE_CP_MEC;
break;
case AMDGPU_UCODE_ID_CP_MEC2_JT:
*type = GFX_FW_TYPE_CP_MEC_ME2;
break;
case AMDGPU_UCODE_ID_RLC_G:
*type = GFX_FW_TYPE_RLC_G;
break;
case AMDGPU_UCODE_ID_RLC_RESTORE_LIST_CNTL:
*type = GFX_FW_TYPE_RLC_RESTORE_LIST_SRM_CNTL;
break;
case AMDGPU_UCODE_ID_RLC_RESTORE_LIST_GPM_MEM:
*type = GFX_FW_TYPE_RLC_RESTORE_LIST_GPM_MEM;
break;
case AMDGPU_UCODE_ID_RLC_RESTORE_LIST_SRM_MEM:
*type = GFX_FW_TYPE_RLC_RESTORE_LIST_SRM_MEM;
break;
case AMDGPU_UCODE_ID_SMC:
*type = GFX_FW_TYPE_SMU;
break;
case AMDGPU_UCODE_ID_UVD:
*type = GFX_FW_TYPE_UVD;
break;
case AMDGPU_UCODE_ID_UVD1:
*type = GFX_FW_TYPE_UVD1;
break;
case AMDGPU_UCODE_ID_VCE:
*type = GFX_FW_TYPE_VCE;
break;
case AMDGPU_UCODE_ID_VCN:
*type = GFX_FW_TYPE_VCN;
break;
case AMDGPU_UCODE_ID_DMCU_ERAM:
*type = GFX_FW_TYPE_DMCU_ERAM;
break;
case AMDGPU_UCODE_ID_DMCU_INTV:
*type = GFX_FW_TYPE_DMCU_ISR;
break;
case AMDGPU_UCODE_ID_MAXIMUM:
default:
return -EINVAL;
}
return 0;
}
static int psp_prep_load_ip_fw_cmd_buf(struct amdgpu_firmware_info *ucode,
struct psp_gfx_cmd_resp *cmd)
{
int ret;
uint64_t fw_mem_mc_addr = ucode->mc_addr;
memset(cmd, 0, sizeof(struct psp_gfx_cmd_resp));
cmd->cmd_id = GFX_CMD_ID_LOAD_IP_FW;
cmd->cmd.cmd_load_ip_fw.fw_phy_addr_lo = lower_32_bits(fw_mem_mc_addr);
cmd->cmd.cmd_load_ip_fw.fw_phy_addr_hi = upper_32_bits(fw_mem_mc_addr);
cmd->cmd.cmd_load_ip_fw.fw_size = ucode->ucode_size;
ret = psp_get_fw_type(ucode, &cmd->cmd.cmd_load_ip_fw.fw_type);
if (ret)
DRM_ERROR("Unknown firmware type\n");
return ret;
}
static int psp_np_fw_load(struct psp_context *psp)
{
int i, ret;
@ -521,7 +618,7 @@ static int psp_np_fw_load(struct psp_context *psp)
/*skip ucode loading in SRIOV VF */
continue;
ret = psp_prep_cmd_buf(ucode, psp->cmd);
ret = psp_prep_load_ip_fw_cmd_buf(ucode, psp->cmd);
if (ret)
return ret;
@ -546,7 +643,7 @@ static int psp_load_fw(struct amdgpu_device *adev)
struct psp_context *psp = &adev->psp;
if (amdgpu_sriov_vf(adev) && adev->in_gpu_reset) {
psp_ring_destroy(psp, PSP_RING_TYPE__KM);
psp_ring_stop(psp, PSP_RING_TYPE__KM); /* should not destroy ring, only stop */
goto skip_memalloc;
}
@ -623,10 +720,6 @@ static int psp_hw_init(void *handle)
int ret;
struct amdgpu_device *adev = (struct amdgpu_device *)handle;
if (adev->firmware.load_type != AMDGPU_FW_LOAD_PSP)
return 0;
mutex_lock(&adev->firmware.mutex);
/*
* This sequence is just used on hw_init only once, no need on
@ -656,9 +749,6 @@ static int psp_hw_fini(void *handle)
struct amdgpu_device *adev = (struct amdgpu_device *)handle;
struct psp_context *psp = &adev->psp;
if (adev->firmware.load_type != AMDGPU_FW_LOAD_PSP)
return 0;
if (adev->gmc.xgmi.num_physical_nodes > 1 &&
psp->xgmi_context.initialized == 1)
psp_xgmi_terminate(psp);
@ -687,9 +777,6 @@ static int psp_suspend(void *handle)
struct amdgpu_device *adev = (struct amdgpu_device *)handle;
struct psp_context *psp = &adev->psp;
if (adev->firmware.load_type != AMDGPU_FW_LOAD_PSP)
return 0;
if (adev->gmc.xgmi.num_physical_nodes > 1 &&
psp->xgmi_context.initialized == 1) {
ret = psp_xgmi_terminate(psp);
@ -714,9 +801,6 @@ static int psp_resume(void *handle)
struct amdgpu_device *adev = (struct amdgpu_device *)handle;
struct psp_context *psp = &adev->psp;
if (adev->firmware.load_type != AMDGPU_FW_LOAD_PSP)
return 0;
DRM_INFO("PSP is resuming...\n");
mutex_lock(&adev->firmware.mutex);
@ -752,11 +836,6 @@ static bool psp_check_fw_loading_status(struct amdgpu_device *adev,
{
struct amdgpu_firmware_info *ucode = NULL;
if (adev->firmware.load_type != AMDGPU_FW_LOAD_PSP) {
DRM_INFO("firmware is not loaded by PSP\n");
return true;
}
if (!adev->firmware.fw_size)
return false;

3
drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h
View File

@ -65,8 +65,6 @@ struct psp_funcs
int (*init_microcode)(struct psp_context *psp);
int (*bootloader_load_sysdrv)(struct psp_context *psp);
int (*bootloader_load_sos)(struct psp_context *psp);
int (*prep_cmd_buf)(struct amdgpu_firmware_info *ucode,
struct psp_gfx_cmd_resp *cmd);
int (*ring_init)(struct psp_context *psp, enum psp_ring_type ring_type);
int (*ring_create)(struct psp_context *psp,
enum psp_ring_type ring_type);
@ -176,7 +174,6 @@ struct psp_xgmi_topology_info {
struct psp_xgmi_node_info nodes[AMDGPU_XGMI_MAX_CONNECTED_NODES];
};
#define psp_prep_cmd_buf(ucode, type) (psp)->funcs->prep_cmd_buf((ucode), (type))
#define psp_ring_init(psp, type) (psp)->funcs->ring_init((psp), (type))
#define psp_ring_create(psp, type) (psp)->funcs->ring_create((psp), (type))
#define psp_ring_stop(psp, type) (psp)->funcs->ring_stop((psp), (type))

4
drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h
View File

@ -131,7 +131,7 @@ struct amdgpu_ring_funcs {
void (*emit_ib)(struct amdgpu_ring *ring,
struct amdgpu_job *job,
struct amdgpu_ib *ib,
bool ctx_switch);
uint32_t flags);
void (*emit_fence)(struct amdgpu_ring *ring, uint64_t addr,
uint64_t seq, unsigned flags);
void (*emit_pipeline_sync)(struct amdgpu_ring *ring);
@ -229,7 +229,7 @@ struct amdgpu_ring {
#define amdgpu_ring_get_rptr(r) (r)->funcs->get_rptr((r))
#define amdgpu_ring_get_wptr(r) (r)->funcs->get_wptr((r))
#define amdgpu_ring_set_wptr(r) (r)->funcs->set_wptr((r))
#define amdgpu_ring_emit_ib(r, job, ib, c) ((r)->funcs->emit_ib((r), (job), (ib), (c)))
#define amdgpu_ring_emit_ib(r, job, ib, flags) ((r)->funcs->emit_ib((r), (job), (ib), (flags)))
#define amdgpu_ring_emit_pipeline_sync(r) (r)->funcs->emit_pipeline_sync((r))
#define amdgpu_ring_emit_vm_flush(r, vmid, addr) (r)->funcs->emit_vm_flush((r), (vmid), (addr))
#define amdgpu_ring_emit_fence(r, addr, seq, flags) (r)->funcs->emit_fence((r), (addr), (seq), (flags))

2
drivers/gpu/drm/amd/amdgpu/amdgpu_sa.c
View File

@ -388,7 +388,7 @@ void amdgpu_sa_bo_dump_debug_info(struct amdgpu_sa_manager *sa_manager,
soffset, eoffset, eoffset - soffset);
if (i->fence)
seq_printf(m, " protected by 0x%08x on context %llu",
seq_printf(m, " protected by 0x%016llx on context %llu",
i->fence->seqno, i->fence->context);
seq_printf(m, "\n");

51
drivers/gpu/drm/amd/amdgpu/amdgpu_sched.c
View File

@ -54,16 +54,20 @@ static int amdgpu_sched_process_priority_override(struct amdgpu_device *adev,
enum drm_sched_priority priority)
{
struct file *filp = fget(fd);
struct drm_file *file;
struct amdgpu_fpriv *fpriv;
struct amdgpu_ctx *ctx;
uint32_t id;
int r;
if (!filp)
return -EINVAL;
file = filp->private_data;
fpriv = file->driver_priv;
r = amdgpu_file_to_fpriv(filp, &fpriv);
if (r) {
fput(filp);
return r;
}
idr_for_each_entry(&fpriv->ctx_mgr.ctx_handles, ctx, id)
amdgpu_ctx_priority_override(ctx, priority);
@ -72,6 +76,39 @@ static int amdgpu_sched_process_priority_override(struct amdgpu_device *adev,
return 0;
}
static int amdgpu_sched_context_priority_override(struct amdgpu_device *adev,
int fd,
unsigned ctx_id,
enum drm_sched_priority priority)
{
struct file *filp = fget(fd);
struct amdgpu_fpriv *fpriv;
struct amdgpu_ctx *ctx;
int r;
if (!filp)
return -EINVAL;
r = amdgpu_file_to_fpriv(filp, &fpriv);
if (r) {
fput(filp);
return r;
}
ctx = amdgpu_ctx_get(fpriv, ctx_id);
if (!ctx) {
fput(filp);
return -EINVAL;
}
amdgpu_ctx_priority_override(ctx, priority);
amdgpu_ctx_put(ctx);
fput(filp);
return 0;
}
int amdgpu_sched_ioctl(struct drm_device *dev, void *data,
struct drm_file *filp)
{
@ -81,7 +118,7 @@ int amdgpu_sched_ioctl(struct drm_device *dev, void *data,
int r;
priority = amdgpu_to_sched_priority(args->in.priority);
if (args->in.flags || priority == DRM_SCHED_PRIORITY_INVALID)
if (priority == DRM_SCHED_PRIORITY_INVALID)
return -EINVAL;
switch (args->in.op) {
@ -90,6 +127,12 @@ int amdgpu_sched_ioctl(struct drm_device *dev, void *data,
args->in.fd,
priority);
break;
case AMDGPU_SCHED_OP_CONTEXT_PRIORITY_OVERRIDE:
r = amdgpu_sched_context_priority_override(adev,
args->in.fd,
args->in.ctx_id,
priority);
break;
default:
DRM_ERROR("Invalid sched op specified: %d\n", args->in.op);
r = -EINVAL;

11
drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h
View File

@ -76,9 +76,10 @@ TRACE_EVENT(amdgpu_mm_wreg,
);
TRACE_EVENT(amdgpu_iv,
TP_PROTO(struct amdgpu_iv_entry *iv),
TP_ARGS(iv),
TP_PROTO(unsigned ih, struct amdgpu_iv_entry *iv),
TP_ARGS(ih, iv),
TP_STRUCT__entry(
__field(unsigned, ih)
__field(unsigned, client_id)
__field(unsigned, src_id)
__field(unsigned, ring_id)
@ -90,6 +91,7 @@ TRACE_EVENT(amdgpu_iv,
__array(unsigned, src_data, 4)
),
TP_fast_assign(
__entry->ih = ih;
__entry->client_id = iv->client_id;
__entry->src_id = iv->src_id;
__entry->ring_id = iv->ring_id;
@ -103,8 +105,9 @@ TRACE_EVENT(amdgpu_iv,
__entry->src_data[2] = iv->src_data[2];
__entry->src_data[3] = iv->src_data[3];
),
TP_printk("client_id:%u src_id:%u ring:%u vmid:%u timestamp: %llu pasid:%u src_data: %08x %08x %08x %08x",
__entry->client_id, __entry->src_id,
TP_printk("ih:%u client_id:%u src_id:%u ring:%u vmid:%u "
"timestamp: %llu pasid:%u src_data: %08x %08x %08x %08x",
__entry->ih, __entry->client_id, __entry->src_id,
__entry->ring_id, __entry->vmid,
__entry->timestamp, __entry->pasid,
__entry->src_data[0], __entry->src_data[1],

9
drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
View File

@ -1546,7 +1546,8 @@ static struct ttm_bo_driver amdgpu_bo_driver = {
.io_mem_reserve = &amdgpu_ttm_io_mem_reserve,
.io_mem_free = &amdgpu_ttm_io_mem_free,
.io_mem_pfn = amdgpu_ttm_io_mem_pfn,
.access_memory = &amdgpu_ttm_access_memory
.access_memory = &amdgpu_ttm_access_memory,
.del_from_lru_notify = &amdgpu_vm_del_from_lru_notify
};
/*
@ -1755,7 +1756,7 @@ int amdgpu_ttm_init(struct amdgpu_device *adev)
}
r = amdgpu_bo_create_kernel(adev, adev->gds.mem.gfx_partition_size,
PAGE_SIZE, AMDGPU_GEM_DOMAIN_GDS,
4, AMDGPU_GEM_DOMAIN_GDS,
&adev->gds.gds_gfx_bo, NULL, NULL);
if (r)
return r;
@ -1768,7 +1769,7 @@ int amdgpu_ttm_init(struct amdgpu_device *adev)
}
r = amdgpu_bo_create_kernel(adev, adev->gds.gws.gfx_partition_size,
PAGE_SIZE, AMDGPU_GEM_DOMAIN_GWS,
1, AMDGPU_GEM_DOMAIN_GWS,
&adev->gds.gws_gfx_bo, NULL, NULL);
if (r)
return r;
@ -1781,7 +1782,7 @@ int amdgpu_ttm_init(struct amdgpu_device *adev)
}
r = amdgpu_bo_create_kernel(adev, adev->gds.oa.gfx_partition_size,
PAGE_SIZE, AMDGPU_GEM_DOMAIN_OA,
1, AMDGPU_GEM_DOMAIN_OA,
&adev->gds.oa_gfx_bo, NULL, NULL);
if (r)
return r;

2
drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c
View File

@ -1035,7 +1035,7 @@ out:
void amdgpu_vce_ring_emit_ib(struct amdgpu_ring *ring,
struct amdgpu_job *job,
struct amdgpu_ib *ib,
bool ctx_switch)
uint32_t flags)
{
amdgpu_ring_write(ring, VCE_CMD_IB);
amdgpu_ring_write(ring, lower_32_bits(ib->gpu_addr));

2
drivers/gpu/drm/amd/amdgpu/amdgpu_vce.h
View File

@ -66,7 +66,7 @@ void amdgpu_vce_free_handles(struct amdgpu_device *adev, struct drm_file *filp);
int amdgpu_vce_ring_parse_cs(struct amdgpu_cs_parser *p, uint32_t ib_idx);
int amdgpu_vce_ring_parse_cs_vm(struct amdgpu_cs_parser *p, uint32_t ib_idx);
void amdgpu_vce_ring_emit_ib(struct amdgpu_ring *ring, struct amdgpu_job *job,
struct amdgpu_ib *ib, bool ctx_switch);
struct amdgpu_ib *ib, uint32_t flags);
void amdgpu_vce_ring_emit_fence(struct amdgpu_ring *ring, u64 addr, u64 seq,
unsigned flags);
int amdgpu_vce_ring_test_ring(struct amdgpu_ring *ring);

133
drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
View File

@ -107,14 +107,6 @@ struct amdgpu_pte_update_params {
* DMA addresses to use for mapping, used during VM update by CPU
*/
dma_addr_t *pages_addr;
/**
* @kptr:
*
* Kernel pointer of PD/PT BO that needs to be updated,
* used during VM update by CPU
*/
void *kptr;
};
/**
@ -623,6 +615,28 @@ void amdgpu_vm_get_pd_bo(struct amdgpu_vm *vm,
list_add(&entry->tv.head, validated);
}
void amdgpu_vm_del_from_lru_notify(struct ttm_buffer_object *bo)
{
struct amdgpu_bo *abo;
struct amdgpu_vm_bo_base *bo_base;
if (!amdgpu_bo_is_amdgpu_bo(bo))
return;
if (bo->mem.placement & TTM_PL_FLAG_NO_EVICT)
return;
abo = ttm_to_amdgpu_bo(bo);
if (!abo->parent)
return;
for (bo_base = abo->vm_bo; bo_base; bo_base = bo_base->next) {
struct amdgpu_vm *vm = bo_base->vm;
if (abo->tbo.resv == vm->root.base.bo->tbo.resv)
vm->bulk_moveable = false;
}
}
/**
* amdgpu_vm_move_to_lru_tail - move all BOs to the end of LRU
*
@ -686,8 +700,6 @@ int amdgpu_vm_validate_pt_bos(struct amdgpu_device *adev, struct amdgpu_vm *vm,
struct amdgpu_vm_bo_base *bo_base, *tmp;
int r = 0;
vm->bulk_moveable &= list_empty(&vm->evicted);
list_for_each_entry_safe(bo_base, tmp, &vm->evicted, vm_status) {
struct amdgpu_bo *bo = bo_base->bo;
@ -801,15 +813,22 @@ static int amdgpu_vm_clear_bo(struct amdgpu_device *adev,
addr += ats_entries * 8;
}
if (entries)
if (entries) {
uint64_t value = 0;
/* Workaround for fault priority problem on GMC9 */
if (level == AMDGPU_VM_PTB && adev->asic_type >= CHIP_VEGA10)
value = AMDGPU_PTE_EXECUTABLE;
amdgpu_vm_set_pte_pde(adev, &job->ibs[0], addr, 0,
entries, 0, 0);
entries, 0, value);
}
amdgpu_ring_pad_ib(ring, &job->ibs[0]);
WARN_ON(job->ibs[0].length_dw > 64);
r = amdgpu_sync_resv(adev, &job->sync, bo->tbo.resv,
AMDGPU_FENCE_OWNER_UNDEFINED, false);
AMDGPU_FENCE_OWNER_KFD, false);
if (r)
goto error_free;
@ -1313,31 +1332,6 @@ static void amdgpu_vm_cpu_set_ptes(struct amdgpu_pte_update_params *params,
}
}
/**
* amdgpu_vm_wait_pd - Wait for PT BOs to be free.
*
* @adev: amdgpu_device pointer
* @vm: related vm
* @owner: fence owner
*
* Returns:
* 0 on success, errno otherwise.
*/
static int amdgpu_vm_wait_pd(struct amdgpu_device *adev, struct amdgpu_vm *vm,
void *owner)
{
struct amdgpu_sync sync;
int r;
amdgpu_sync_create(&sync);
amdgpu_sync_resv(adev, &sync, vm->root.base.bo->tbo.resv, owner, false);
r = amdgpu_sync_wait(&sync, true);
amdgpu_sync_free(&sync);
return r;
}
/**
* amdgpu_vm_update_func - helper to call update function
*
@ -1432,7 +1426,8 @@ restart:
params.adev = adev;
if (vm->use_cpu_for_update) {
r = amdgpu_vm_wait_pd(adev, vm, AMDGPU_FENCE_OWNER_VM);
r = amdgpu_bo_sync_wait(vm->root.base.bo,
AMDGPU_FENCE_OWNER_VM, true);
if (unlikely(r))
return r;
@ -1505,20 +1500,27 @@ error:
}
/**
* amdgpu_vm_update_huge - figure out parameters for PTE updates
* amdgpu_vm_update_flags - figure out flags for PTE updates
*
* Make sure to set the right flags for the PTEs at the desired level.
*/
static void amdgpu_vm_update_huge(struct amdgpu_pte_update_params *params,
struct amdgpu_bo *bo, unsigned level,
uint64_t pe, uint64_t addr,
unsigned count, uint32_t incr,
uint64_t flags)
static void amdgpu_vm_update_flags(struct amdgpu_pte_update_params *params,
struct amdgpu_bo *bo, unsigned level,
uint64_t pe, uint64_t addr,
unsigned count, uint32_t incr,
uint64_t flags)
{
if (level != AMDGPU_VM_PTB) {
flags |= AMDGPU_PDE_PTE;
amdgpu_gmc_get_vm_pde(params->adev, level, &addr, &flags);
} else if (params->adev->asic_type >= CHIP_VEGA10 &&
!(flags & AMDGPU_PTE_VALID) &&
!(flags & AMDGPU_PTE_PRT)) {
/* Workaround for fault priority problem on GMC9 */
flags |= AMDGPU_PTE_EXECUTABLE;
}
amdgpu_vm_update_func(params, bo, pe, addr, count, incr, flags);
@ -1675,9 +1677,9 @@ static int amdgpu_vm_update_ptes(struct amdgpu_pte_update_params *params,
uint64_t upd_end = min(entry_end, frag_end);
unsigned nptes = (upd_end - frag_start) >> shift;
amdgpu_vm_update_huge(params, pt, cursor.level,
pe_start, dst, nptes, incr,
flags | AMDGPU_PTE_FRAG(frag));
amdgpu_vm_update_flags(params, pt, cursor.level,
pe_start, dst, nptes, incr,
flags | AMDGPU_PTE_FRAG(frag));
pe_start += nptes * 8;
dst += (uint64_t)nptes * AMDGPU_GPU_PAGE_SIZE << shift;
@ -1746,22 +1748,29 @@ static int amdgpu_vm_bo_update_mapping(struct amdgpu_device *adev,
params.adev = adev;
params.vm = vm;
/* sync to everything on unmapping */
/* sync to everything except eviction fences on unmapping */
if (!(flags & AMDGPU_PTE_VALID))
owner = AMDGPU_FENCE_OWNER_UNDEFINED;
owner = AMDGPU_FENCE_OWNER_KFD;
if (vm->use_cpu_for_update) {
/* params.src is used as flag to indicate system Memory */
if (pages_addr)
params.src = ~0;
/* Wait for PT BOs to be free. PTs share the same resv. object
/* Wait for PT BOs to be idle. PTs share the same resv. object
* as the root PD BO
*/
r = amdgpu_vm_wait_pd(adev, vm, owner);
r = amdgpu_bo_sync_wait(vm->root.base.bo, owner, true);
if (unlikely(r))
return r;
/* Wait for any BO move to be completed */
if (exclusive) {
r = dma_fence_wait(exclusive, true);
if (unlikely(r))
return r;
}
params.func = amdgpu_vm_cpu_set_ptes;
params.pages_addr = pages_addr;
return amdgpu_vm_update_ptes(&params, start, last + 1,
@ -1775,13 +1784,12 @@ static int amdgpu_vm_bo_update_mapping(struct amdgpu_device *adev,
/*
* reserve space for two commands every (1 << BLOCK_SIZE)
* entries or 2k dwords (whatever is smaller)
*
* The second command is for the shadow pagetables.
*/
ncmds = ((nptes >> min(adev->vm_manager.block_size, 11u)) + 1);
/* The second command is for the shadow pagetables. */
if (vm->root.base.bo->shadow)
ncmds = ((nptes >> min(adev->vm_manager.block_size, 11u)) + 1) * 2;
else
ncmds = ((nptes >> min(adev->vm_manager.block_size, 11u)) + 1);
ncmds *= 2;
/* padding, etc. */
ndw = 64;
@ -1800,10 +1808,11 @@ static int amdgpu_vm_bo_update_mapping(struct amdgpu_device *adev,
ndw += ncmds * 10;
/* extra commands for begin/end fragments */
ncmds = 2 * adev->vm_manager.fragment_size;
if (vm->root.base.bo->shadow)
ndw += 2 * 10 * adev->vm_manager.fragment_size * 2;
else
ndw += 2 * 10 * adev->vm_manager.fragment_size;
ncmds *= 2;
ndw += 10 * ncmds;
params.func = amdgpu_vm_do_set_ptes;
}
@ -3005,7 +3014,7 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm,
}
DRM_DEBUG_DRIVER("VM update mode is %s\n",
vm->use_cpu_for_update ? "CPU" : "SDMA");
WARN_ONCE((vm->use_cpu_for_update & !amdgpu_gmc_vram_full_visible(&adev->gmc)),
WARN_ONCE((vm->use_cpu_for_update && !amdgpu_gmc_vram_full_visible(&adev->gmc)),
"CPU update of VM recommended only for large BAR system\n");
vm->last_update = NULL;
@ -3135,7 +3144,7 @@ int amdgpu_vm_make_compute(struct amdgpu_device *adev, struct amdgpu_vm *vm, uns
vm->pte_support_ats = pte_support_ats;
DRM_DEBUG_DRIVER("VM update mode is %s\n",
vm->use_cpu_for_update ? "CPU" : "SDMA");
WARN_ONCE((vm->use_cpu_for_update & !amdgpu_gmc_vram_full_visible(&adev->gmc)),
WARN_ONCE((vm->use_cpu_for_update && !amdgpu_gmc_vram_full_visible(&adev->gmc)),
"CPU update of VM recommended only for large BAR system\n");
if (vm->pasid) {

2
drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h
View File

@ -363,4 +363,6 @@ int amdgpu_vm_add_fault(struct amdgpu_retryfault_hashtable *fault_hash, u64 key)
void amdgpu_vm_clear_fault(struct amdgpu_retryfault_hashtable *fault_hash, u64 key);
void amdgpu_vm_del_from_lru_notify(struct ttm_buffer_object *bo);
#endif

54
drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c
View File

@ -40,26 +40,40 @@ void *amdgpu_xgmi_hive_try_lock(struct amdgpu_hive_info *hive)
return &hive->device_list;
}
struct amdgpu_hive_info *amdgpu_get_xgmi_hive(struct amdgpu_device *adev)
struct amdgpu_hive_info *amdgpu_get_xgmi_hive(struct amdgpu_device *adev, int lock)
{
int i;
struct amdgpu_hive_info *tmp;
if (!adev->gmc.xgmi.hive_id)
return NULL;
mutex_lock(&xgmi_mutex);
for (i = 0 ; i < hive_count; ++i) {
tmp = &xgmi_hives[i];
if (tmp->hive_id == adev->gmc.xgmi.hive_id)
if (tmp->hive_id == adev->gmc.xgmi.hive_id) {
if (lock)
mutex_lock(&tmp->hive_lock);
mutex_unlock(&xgmi_mutex);
return tmp;
}
}
if (i >= AMDGPU_MAX_XGMI_HIVE)
if (i >= AMDGPU_MAX_XGMI_HIVE) {
mutex_unlock(&xgmi_mutex);
return NULL;
}
/* initialize new hive if not exist */
tmp = &xgmi_hives[hive_count++];
tmp->hive_id = adev->gmc.xgmi.hive_id;
INIT_LIST_HEAD(&tmp->device_list);
mutex_init(&tmp->hive_lock);
mutex_init(&tmp->reset_lock);
if (lock)
mutex_lock(&tmp->hive_lock);
mutex_unlock(&xgmi_mutex);
return tmp;
}
@ -77,10 +91,6 @@ int amdgpu_xgmi_update_topology(struct amdgpu_hive_info *hive, struct amdgpu_dev
"XGMI: Set topology failure on device %llx, hive %llx, ret %d",
adev->gmc.xgmi.node_id,
adev->gmc.xgmi.hive_id, ret);
else
dev_info(adev->dev, "XGMI: Set topology for node %d, hive 0x%llx.\n",
adev->gmc.xgmi.physical_node_id,
adev->gmc.xgmi.hive_id);
return ret;
}
@ -111,10 +121,14 @@ int amdgpu_xgmi_add_device(struct amdgpu_device *adev)
return ret;
}
mutex_lock(&xgmi_mutex);
hive = amdgpu_get_xgmi_hive(adev);
if (!hive)
hive = amdgpu_get_xgmi_hive(adev, 1);
if (!hive) {
ret = -EINVAL;
dev_err(adev->dev,
"XGMI: node 0x%llx, can not match hive 0x%llx in the hive list.\n",
adev->gmc.xgmi.node_id, adev->gmc.xgmi.hive_id);
goto exit;
}
hive_topology = &hive->topology_info;
@ -142,8 +156,11 @@ int amdgpu_xgmi_add_device(struct amdgpu_device *adev)
break;
}
dev_info(adev->dev, "XGMI: Add node %d, hive 0x%llx.\n",
adev->gmc.xgmi.physical_node_id, adev->gmc.xgmi.hive_id);
mutex_unlock(&hive->hive_lock);
exit:
mutex_unlock(&xgmi_mutex);
return ret;
}
@ -154,15 +171,14 @@ void amdgpu_xgmi_remove_device(struct amdgpu_device *adev)
if (!adev->gmc.xgmi.supported)
return;
mutex_lock(&xgmi_mutex);
hive = amdgpu_get_xgmi_hive(adev);
hive = amdgpu_get_xgmi_hive(adev, 1);
if (!hive)
goto exit;
return;
if (!(hive->number_devices--))
if (!(hive->number_devices--)) {
mutex_destroy(&hive->hive_lock);
exit:
mutex_unlock(&xgmi_mutex);
mutex_destroy(&hive->reset_lock);
} else {
mutex_unlock(&hive->hive_lock);
}
}

5
drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.h
View File

@ -29,10 +29,11 @@ struct amdgpu_hive_info {
struct list_head device_list;
struct psp_xgmi_topology_info topology_info;
int number_devices;
struct mutex hive_lock;
struct mutex hive_lock,
reset_lock;
};
struct amdgpu_hive_info *amdgpu_get_xgmi_hive(struct amdgpu_device *adev);
struct amdgpu_hive_info *amdgpu_get_xgmi_hive(struct amdgpu_device *adev, int lock);
int amdgpu_xgmi_update_topology(struct amdgpu_hive_info *hive, struct amdgpu_device *adev);
int amdgpu_xgmi_add_device(struct amdgpu_device *adev);
void amdgpu_xgmi_remove_device(struct amdgpu_device *adev);

2
drivers/gpu/drm/amd/amdgpu/atom.c
View File

@ -27,6 +27,8 @@
#include <linux/slab.h>
#include <asm/unaligned.h>
#include <drm/drm_util.h>
#define ATOM_DEBUG
#include "atom.h"

6844
drivers/gpu/drm/amd/amdgpu/ci_dpm.c
View File

File diff suppressed because it is too large Load Diff

349
drivers/gpu/drm/amd/amdgpu/ci_dpm.h
View File

@ -1,349 +0,0 @@
/*
* Copyright 2013 Advanced Micro Devices, Inc.
*
* Permission is hereby granted, free of charge, to any person obtaining a
* copy of this software and associated documentation files (the "Software"),
* to deal in the Software without restriction, including without limitation
* the rights to use, copy, modify, merge, publish, distribute, sublicense,
* and/or sell copies of the Software, and to permit persons to whom the
* Software is furnished to do so, subject to the following conditions:
*
* The above copyright notice and this permission notice shall be included in
* all copies or substantial portions of the Software.
*
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
* THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
* OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
* ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
* OTHER DEALINGS IN THE SOFTWARE.
*
*/
#ifndef __CI_DPM_H__
#define __CI_DPM_H__
#include "amdgpu_atombios.h"
#include "ppsmc.h"
#define SMU__NUM_SCLK_DPM_STATE 8
#define SMU__NUM_MCLK_DPM_LEVELS 6
#define SMU__NUM_LCLK_DPM_LEVELS 8
#define SMU__NUM_PCIE_DPM_LEVELS 8
#include "smu7_discrete.h"
#define CISLANDS_MAX_HARDWARE_POWERLEVELS 2
#define CISLANDS_UNUSED_GPIO_PIN 0x7F
struct ci_pl {
u32 mclk;
u32 sclk;
enum amdgpu_pcie_gen pcie_gen;
u16 pcie_lane;
};
struct ci_ps {
u16 performance_level_count;
bool dc_compatible;
u32 sclk_t;
struct ci_pl performance_levels[CISLANDS_MAX_HARDWARE_POWERLEVELS];
};
struct ci_dpm_level {
bool enabled;
u32 value;
u32 param1;
};
#define CISLAND_MAX_DEEPSLEEP_DIVIDER_ID 5
#define MAX_REGULAR_DPM_NUMBER 8
#define CISLAND_MINIMUM_ENGINE_CLOCK 800
struct ci_single_dpm_table {
u32 count;
struct ci_dpm_level dpm_levels[MAX_REGULAR_DPM_NUMBER];
};
struct ci_dpm_table {
struct ci_single_dpm_table sclk_table;
struct ci_single_dpm_table mclk_table;
struct ci_single_dpm_table pcie_speed_table;
struct ci_single_dpm_table vddc_table;
struct ci_single_dpm_table vddci_table;
struct ci_single_dpm_table mvdd_table;
};
struct ci_mc_reg_entry {
u32 mclk_max;
u32 mc_data[SMU7_DISCRETE_MC_REGISTER_ARRAY_SIZE];
};
struct ci_mc_reg_table {
u8 last;
u8 num_entries;
u16 valid_flag;
struct ci_mc_reg_entry mc_reg_table_entry[MAX_AC_TIMING_ENTRIES];
SMU7_Discrete_MCRegisterAddress mc_reg_address[SMU7_DISCRETE_MC_REGISTER_ARRAY_SIZE];
};
struct ci_ulv_parm
{
bool supported;
u32 cg_ulv_parameter;
u32 volt_change_delay;
struct ci_pl pl;
};
#define CISLANDS_MAX_LEAKAGE_COUNT 8
struct ci_leakage_voltage {
u16 count;
u16 leakage_id[CISLANDS_MAX_LEAKAGE_COUNT];
u16 actual_voltage[CISLANDS_MAX_LEAKAGE_COUNT];
};
struct ci_dpm_level_enable_mask {
u32 uvd_dpm_enable_mask;
u32 vce_dpm_enable_mask;
u32 acp_dpm_enable_mask;
u32 samu_dpm_enable_mask;
u32 sclk_dpm_enable_mask;
u32 mclk_dpm_enable_mask;
u32 pcie_dpm_enable_mask;
};
struct ci_vbios_boot_state
{
u16 mvdd_bootup_value;
u16 vddc_bootup_value;
u16 vddci_bootup_value;
u32 sclk_bootup_value;
u32 mclk_bootup_value;
u16 pcie_gen_bootup_value;
u16 pcie_lane_bootup_value;
};
struct ci_clock_registers {
u32 cg_spll_func_cntl;
u32 cg_spll_func_cntl_2;
u32 cg_spll_func_cntl_3;
u32 cg_spll_func_cntl_4;
u32 cg_spll_spread_spectrum;
u32 cg_spll_spread_spectrum_2;
u32 dll_cntl;
u32 mclk_pwrmgt_cntl;
u32 mpll_ad_func_cntl;
u32 mpll_dq_func_cntl;
u32 mpll_func_cntl;
u32 mpll_func_cntl_1;
u32 mpll_func_cntl_2;
u32 mpll_ss1;
u32 mpll_ss2;
};
struct ci_thermal_temperature_setting {
s32 temperature_low;
s32 temperature_high;
s32 temperature_shutdown;
};
struct ci_pcie_perf_range {
u16 max;
u16 min;
};
enum ci_pt_config_reg_type {
CISLANDS_CONFIGREG_MMR = 0,
CISLANDS_CONFIGREG_SMC_IND,
CISLANDS_CONFIGREG_DIDT_IND,
CISLANDS_CONFIGREG_CACHE,
CISLANDS_CONFIGREG_MAX
};
#define POWERCONTAINMENT_FEATURE_BAPM 0x00000001
#define POWERCONTAINMENT_FEATURE_TDCLimit 0x00000002
#define POWERCONTAINMENT_FEATURE_PkgPwrLimit 0x00000004
struct ci_pt_config_reg {
u32 offset;
u32 mask;
u32 shift;
u32 value;
enum ci_pt_config_reg_type type;
};
struct ci_pt_defaults {
u8 svi_load_line_en;
u8 svi_load_line_vddc;
u8 tdc_vddc_throttle_release_limit_perc;
u8 tdc_mawt;
u8 tdc_waterfall_ctl;
u8 dte_ambient_temp_base;
u32 display_cac;
u32 bapm_temp_gradient;
u16 bapmti_r[SMU7_DTE_ITERATIONS * SMU7_DTE_SOURCES * SMU7_DTE_SINKS];
u16 bapmti_rc[SMU7_DTE_ITERATIONS * SMU7_DTE_SOURCES * SMU7_DTE_SINKS];
};
#define DPMTABLE_OD_UPDATE_SCLK 0x00000001
#define DPMTABLE_OD_UPDATE_MCLK 0x00000002
#define DPMTABLE_UPDATE_SCLK 0x00000004
#define DPMTABLE_UPDATE_MCLK 0x00000008
struct ci_power_info {
struct ci_dpm_table dpm_table;
struct ci_dpm_table golden_dpm_table;
u32 voltage_control;
u32 mvdd_control;
u32 vddci_control;
u32 active_auto_throttle_sources;
struct ci_clock_registers clock_registers;
u16 acpi_vddc;
u16 acpi_vddci;
enum amdgpu_pcie_gen force_pcie_gen;
enum amdgpu_pcie_gen acpi_pcie_gen;
struct ci_leakage_voltage vddc_leakage;
struct ci_leakage_voltage vddci_leakage;
u16 max_vddc_in_pp_table;
u16 min_vddc_in_pp_table;
u16 max_vddci_in_pp_table;
u16 min_vddci_in_pp_table;
u32 mclk_strobe_mode_threshold;
u32 mclk_stutter_mode_threshold;
u32 mclk_edc_enable_threshold;
u32 mclk_edc_wr_enable_threshold;
struct ci_vbios_boot_state vbios_boot_state;
/* smc offsets */
u32 sram_end;
u32 dpm_table_start;
u32 soft_regs_start;
u32 mc_reg_table_start;
u32 fan_table_start;
u32 arb_table_start;
/* smc tables */
SMU7_Discrete_DpmTable smc_state_table;
SMU7_Discrete_MCRegisters smc_mc_reg_table;
SMU7_Discrete_PmFuses smc_powertune_table;
/* other stuff */
struct ci_mc_reg_table mc_reg_table;
struct atom_voltage_table vddc_voltage_table;
struct atom_voltage_table vddci_voltage_table;
struct atom_voltage_table mvdd_voltage_table;
struct ci_ulv_parm ulv;
u32 power_containment_features;
const struct ci_pt_defaults *powertune_defaults;
u32 dte_tj_offset;
bool vddc_phase_shed_control;
struct ci_thermal_temperature_setting thermal_temp_setting;
struct ci_dpm_level_enable_mask dpm_level_enable_mask;
u32 need_update_smu7_dpm_table;
u32 sclk_dpm_key_disabled;
u32 mclk_dpm_key_disabled;
u32 pcie_dpm_key_disabled;
u32 thermal_sclk_dpm_enabled;
struct ci_pcie_perf_range pcie_gen_performance;
struct ci_pcie_perf_range pcie_lane_performance;
struct ci_pcie_perf_range pcie_gen_powersaving;
struct ci_pcie_perf_range pcie_lane_powersaving;
u32 activity_target[SMU7_MAX_LEVELS_GRAPHICS];
u32 mclk_activity_target;
u32 low_sclk_interrupt_t;
u32 last_mclk_dpm_enable_mask;
u32 sys_pcie_mask;
/* caps */
bool caps_power_containment;
bool caps_cac;
bool caps_sq_ramping;
bool caps_db_ramping;
bool caps_td_ramping;
bool caps_tcp_ramping;
bool caps_fps;
bool caps_sclk_ds;
bool caps_sclk_ss_support;
bool caps_mclk_ss_support;
bool caps_uvd_dpm;
bool caps_vce_dpm;
bool caps_samu_dpm;
bool caps_acp_dpm;
bool caps_automatic_dc_transition;
bool caps_sclk_throttle_low_notification;
bool caps_dynamic_ac_timing;
bool caps_od_fuzzy_fan_control_support;
/* flags */
bool thermal_protection;
bool pcie_performance_request;
bool dynamic_ss;
bool dll_default_on;
bool cac_enabled;
bool uvd_enabled;
bool battery_state;
bool pspp_notify_required;
bool enable_bapm_feature;
bool enable_tdc_limit_feature;
bool enable_pkg_pwr_tracking_feature;
bool use_pcie_performance_levels;
bool use_pcie_powersaving_levels;
bool uvd_power_gated;
/* driver states */
struct amdgpu_ps current_rps;
struct ci_ps current_ps;
struct amdgpu_ps requested_rps;
struct ci_ps requested_ps;
/* fan control */
bool fan_ctrl_is_in_default_mode;
bool fan_is_controlled_by_smc;
u32 t_min;
u32 fan_ctrl_default_mode;
};
#define CISLANDS_VOLTAGE_CONTROL_NONE 0x0
#define CISLANDS_VOLTAGE_CONTROL_BY_GPIO 0x1
#define CISLANDS_VOLTAGE_CONTROL_BY_SVID2 0x2
#define CISLANDS_Q88_FORMAT_CONVERSION_UNIT 256
#define CISLANDS_VRC_DFLT0 0x3FFFC000
#define CISLANDS_VRC_DFLT1 0x000400
#define CISLANDS_VRC_DFLT2 0xC00080
#define CISLANDS_VRC_DFLT3 0xC00200
#define CISLANDS_VRC_DFLT4 0xC01680
#define CISLANDS_VRC_DFLT5 0xC00033
#define CISLANDS_VRC_DFLT6 0xC00033
#define CISLANDS_VRC_DFLT7 0x3FFFC000
#define CISLANDS_CGULVPARAMETER_DFLT 0x00040035
#define CISLAND_TARGETACTIVITY_DFLT 30
#define CISLAND_MCLK_TARGETACTIVITY_DFLT 10
#define PCIE_PERF_REQ_REMOVE_REGISTRY 0
#define PCIE_PERF_REQ_FORCE_LOWPOWER 1
#define PCIE_PERF_REQ_PECI_GEN1 2
#define PCIE_PERF_REQ_PECI_GEN2 3
#define PCIE_PERF_REQ_PECI_GEN3 4
#define CISLANDS_SSTU_DFLT 0
#define CISLANDS_SST_DFLT 0x00C8
/* XXX are these ok? */
#define CISLANDS_TEMP_RANGE_MIN (90 * 1000)
#define CISLANDS_TEMP_RANGE_MAX (120 * 1000)
int amdgpu_ci_copy_bytes_to_smc(struct amdgpu_device *adev,
u32 smc_start_address,
const u8 *src, u32 byte_count, u32 limit);
void amdgpu_ci_start_smc(struct amdgpu_device *adev);
void amdgpu_ci_reset_smc(struct amdgpu_device *adev);
int amdgpu_ci_program_jump_on_start(struct amdgpu_device *adev);
void amdgpu_ci_stop_smc_clock(struct amdgpu_device *adev);
void amdgpu_ci_start_smc_clock(struct amdgpu_device *adev);
bool amdgpu_ci_is_smc_running(struct amdgpu_device *adev);
PPSMC_Result amdgpu_ci_send_msg_to_smc(struct amdgpu_device *adev, PPSMC_Msg msg);
PPSMC_Result amdgpu_ci_wait_for_smc_inactive(struct amdgpu_device *adev);
int amdgpu_ci_load_smc_ucode(struct amdgpu_device *adev, u32 limit);
int amdgpu_ci_read_smc_sram_dword(struct amdgpu_device *adev,
u32 smc_address, u32 *value, u32 limit);
int amdgpu_ci_write_smc_sram_dword(struct amdgpu_device *adev,
u32 smc_address, u32 value, u32 limit);
#endif

279
drivers/gpu/drm/amd/amdgpu/ci_smc.c
View File

@ -1,279 +0,0 @@
/*
* Copyright 2011 Advanced Micro Devices, Inc.
*
* Permission is hereby granted, free of charge, to any person obtaining a
* copy of this software and associated documentation files (the "Software"),
* to deal in the Software without restriction, including without limitation
* the rights to use, copy, modify, merge, publish, distribute, sublicense,
* and/or sell copies of the Software, and to permit persons to whom the
* Software is furnished to do so, subject to the following conditions:
*
* The above copyright notice and this permission notice shall be included in
* all copies or substantial portions of the Software.
*
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
* THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
* OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
* ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
* OTHER DEALINGS IN THE SOFTWARE.
*
* Authors: Alex Deucher
*/
#include <linux/firmware.h>
#include <drm/drmP.h>
#include "amdgpu.h"
#include "cikd.h"
#include "ppsmc.h"
#include "amdgpu_ucode.h"
#include "ci_dpm.h"
#include "smu/smu_7_0_1_d.h"
#include "smu/smu_7_0_1_sh_mask.h"
static int ci_set_smc_sram_address(struct amdgpu_device *adev,
u32 smc_address, u32 limit)
{
if (smc_address & 3)
return -EINVAL;
if ((smc_address + 3) > limit)
return -EINVAL;
WREG32(mmSMC_IND_INDEX_0, smc_address);
WREG32_P(mmSMC_IND_ACCESS_CNTL, 0, ~SMC_IND_ACCESS_CNTL__AUTO_INCREMENT_IND_0_MASK);
return 0;
}
int amdgpu_ci_copy_bytes_to_smc(struct amdgpu_device *adev,
u32 smc_start_address,
const u8 *src, u32 byte_count, u32 limit)
{
unsigned long flags;
u32 data, original_data;
u32 addr;
u32 extra_shift;
int ret = 0;
if (smc_start_address & 3)
return -EINVAL;
if ((smc_start_address + byte_count) > limit)
return -EINVAL;
addr = smc_start_address;
spin_lock_irqsave(&adev->smc_idx_lock, flags);
while (byte_count >= 4) {
/* SMC address space is BE */
data = (src[0] << 24) | (src[1] << 16) | (src[2] << 8) | src[3];
ret = ci_set_smc_sram_address(adev, addr, limit);
if (ret)
goto done;
WREG32(mmSMC_IND_DATA_0, data);
src += 4;
byte_count -= 4;
addr += 4;
}
/* RMW for the final bytes */
if (byte_count > 0) {
data = 0;
ret = ci_set_smc_sram_address(adev, addr, limit);
if (ret)
goto done;
original_data = RREG32(mmSMC_IND_DATA_0);
extra_shift = 8 * (4 - byte_count);
while (byte_count > 0) {
data = (data << 8) + *src++;
byte_count--;
}
data <<= extra_shift;
data |= (original_data & ~((~0UL) << extra_shift));
ret = ci_set_smc_sram_address(adev, addr, limit);
if (ret)
goto done;
WREG32(mmSMC_IND_DATA_0, data);
}
done:
spin_unlock_irqrestore(&adev->smc_idx_lock, flags);
return ret;
}
void amdgpu_ci_start_smc(struct amdgpu_device *adev)
{
u32 tmp = RREG32_SMC(ixSMC_SYSCON_RESET_CNTL);
tmp &= ~SMC_SYSCON_RESET_CNTL__rst_reg_MASK;
WREG32_SMC(ixSMC_SYSCON_RESET_CNTL, tmp);
}
void amdgpu_ci_reset_smc(struct amdgpu_device *adev)
{
u32 tmp = RREG32_SMC(ixSMC_SYSCON_RESET_CNTL);
tmp |= SMC_SYSCON_RESET_CNTL__rst_reg_MASK;
WREG32_SMC(ixSMC_SYSCON_RESET_CNTL, tmp);
}
int amdgpu_ci_program_jump_on_start(struct amdgpu_device *adev)
{
static u8 data[] = { 0xE0, 0x00, 0x80, 0x40 };
return amdgpu_ci_copy_bytes_to_smc(adev, 0x0, data, 4, sizeof(data)+1);
}
void amdgpu_ci_stop_smc_clock(struct amdgpu_device *adev)
{
u32 tmp = RREG32_SMC(ixSMC_SYSCON_CLOCK_CNTL_0);
tmp |= SMC_SYSCON_CLOCK_CNTL_0__ck_disable_MASK;
WREG32_SMC(ixSMC_SYSCON_CLOCK_CNTL_0, tmp);
}
void amdgpu_ci_start_smc_clock(struct amdgpu_device *adev)
{
u32 tmp = RREG32_SMC(ixSMC_SYSCON_CLOCK_CNTL_0);
tmp &= ~SMC_SYSCON_CLOCK_CNTL_0__ck_disable_MASK;
WREG32_SMC(ixSMC_SYSCON_CLOCK_CNTL_0, tmp);
}
bool amdgpu_ci_is_smc_running(struct amdgpu_device *adev)
{
u32 clk = RREG32_SMC(ixSMC_SYSCON_CLOCK_CNTL_0);
u32 pc_c = RREG32_SMC(ixSMC_PC_C);
if (!(clk & SMC_SYSCON_CLOCK_CNTL_0__ck_disable_MASK) && (0x20100 <= pc_c))
return true;
return false;
}
PPSMC_Result amdgpu_ci_send_msg_to_smc(struct amdgpu_device *adev, PPSMC_Msg msg)
{
u32 tmp;
int i;
if (!amdgpu_ci_is_smc_running(adev))
return PPSMC_Result_Failed;
WREG32(mmSMC_MESSAGE_0, msg);
for (i = 0; i < adev->usec_timeout; i++) {
tmp = RREG32(mmSMC_RESP_0);
if (tmp != 0)
break;
udelay(1);
}
tmp = RREG32(mmSMC_RESP_0);
return (PPSMC_Result)tmp;
}
PPSMC_Result amdgpu_ci_wait_for_smc_inactive(struct amdgpu_device *adev)
{
u32 tmp;
int i;
if (!amdgpu_ci_is_smc_running(adev))
return PPSMC_Result_OK;
for (i = 0; i < adev->usec_timeout; i++) {
tmp = RREG32_SMC(ixSMC_SYSCON_CLOCK_CNTL_0);
if ((tmp & SMC_SYSCON_CLOCK_CNTL_0__cken_MASK) == 0)
break;
udelay(1);
}
return PPSMC_Result_OK;
}
int amdgpu_ci_load_smc_ucode(struct amdgpu_device *adev, u32 limit)
{
const struct smc_firmware_header_v1_0 *hdr;
unsigned long flags;
u32 ucode_start_address;
u32 ucode_size;
const u8 *src;
u32 data;
if (!adev->pm.fw)
return -EINVAL;
hdr = (const struct smc_firmware_header_v1_0 *)adev->pm.fw->data;
amdgpu_ucode_print_smc_hdr(&hdr->header);
adev->pm.fw_version = le32_to_cpu(hdr->header.ucode_version);
ucode_start_address = le32_to_cpu(hdr->ucode_start_addr);
ucode_size = le32_to_cpu(hdr->header.ucode_size_bytes);
src = (const u8 *)
(adev->pm.fw->data + le32_to_cpu(hdr->header.ucode_array_offset_bytes));
if (ucode_size & 3)
return -EINVAL;
spin_lock_irqsave(&adev->smc_idx_lock, flags);
WREG32(mmSMC_IND_INDEX_0, ucode_start_address);
WREG32_P(mmSMC_IND_ACCESS_CNTL, SMC_IND_ACCESS_CNTL__AUTO_INCREMENT_IND_0_MASK,
~SMC_IND_ACCESS_CNTL__AUTO_INCREMENT_IND_0_MASK);
while (ucode_size >= 4) {
/* SMC address space is BE */
data = (src[0] << 24) | (src[1] << 16) | (src[2] << 8) | src[3];
WREG32(mmSMC_IND_DATA_0, data);
src += 4;
ucode_size -= 4;
}
WREG32_P(mmSMC_IND_ACCESS_CNTL, 0, ~SMC_IND_ACCESS_CNTL__AUTO_INCREMENT_IND_0_MASK);
spin_unlock_irqrestore(&adev->smc_idx_lock, flags);
return 0;
}
int amdgpu_ci_read_smc_sram_dword(struct amdgpu_device *adev,
u32 smc_address, u32 *value, u32 limit)
{
unsigned long flags;
int ret;
spin_lock_irqsave(&adev->smc_idx_lock, flags);
ret = ci_set_smc_sram_address(adev, smc_address, limit);
if (ret == 0)
*value = RREG32(mmSMC_IND_DATA_0);
spin_unlock_irqrestore(&adev->smc_idx_lock, flags);
return ret;
}
int amdgpu_ci_write_smc_sram_dword(struct amdgpu_device *adev,
u32 smc_address, u32 value, u32 limit)
{
unsigned long flags;
int ret;
spin_lock_irqsave(&adev->smc_idx_lock, flags);
ret = ci_set_smc_sram_address(adev, smc_address, limit);
if (ret == 0)
WREG32(mmSMC_IND_DATA_0, value);
spin_unlock_irqrestore(&adev->smc_idx_lock, flags);
return ret;
}

75
drivers/gpu/drm/amd/amdgpu/cik.c
View File

@ -1741,6 +1741,69 @@ static bool cik_need_full_reset(struct amdgpu_device *adev)
return true;
}
static void cik_get_pcie_usage(struct amdgpu_device *adev, uint64_t *count0,
uint64_t *count1)
{
uint32_t perfctr = 0;
uint64_t cnt0_of, cnt1_of;
int tmp;
/* This reports 0 on APUs, so return to avoid writing/reading registers
* that may or may not be different from their GPU counterparts
*/
if (adev->flags & AMD_IS_APU)
return;
/* Set the 2 events that we wish to watch, defined above */
/* Reg 40 is # received msgs, Reg 104 is # of posted requests sent */
perfctr = REG_SET_FIELD(perfctr, PCIE_PERF_CNTL_TXCLK, EVENT0_SEL, 40);
perfctr = REG_SET_FIELD(perfctr, PCIE_PERF_CNTL_TXCLK, EVENT1_SEL, 104);
/* Write to enable desired perf counters */
WREG32_PCIE(ixPCIE_PERF_CNTL_TXCLK, perfctr);
/* Zero out and enable the perf counters
* Write 0x5:
* Bit 0 = Start all counters(1)
* Bit 2 = Global counter reset enable(1)
*/
WREG32_PCIE(ixPCIE_PERF_COUNT_CNTL, 0x00000005);
msleep(1000);
/* Load the shadow and disable the perf counters
* Write 0x2:
* Bit 0 = Stop counters(0)
* Bit 1 = Load the shadow counters(1)
*/
WREG32_PCIE(ixPCIE_PERF_COUNT_CNTL, 0x00000002);
/* Read register values to get any >32bit overflow */
tmp = RREG32_PCIE(ixPCIE_PERF_CNTL_TXCLK);
cnt0_of = REG_GET_FIELD(tmp, PCIE_PERF_CNTL_TXCLK, COUNTER0_UPPER);
cnt1_of = REG_GET_FIELD(tmp, PCIE_PERF_CNTL_TXCLK, COUNTER1_UPPER);
/* Get the values and add the overflow */
*count0 = RREG32_PCIE(ixPCIE_PERF_COUNT0_TXCLK) | (cnt0_of << 32);
*count1 = RREG32_PCIE(ixPCIE_PERF_COUNT1_TXCLK) | (cnt1_of << 32);
}
static bool cik_need_reset_on_init(struct amdgpu_device *adev)
{
u32 clock_cntl, pc;
if (adev->flags & AMD_IS_APU)
return false;
/* check if the SMC is already running */
clock_cntl = RREG32_SMC(ixSMC_SYSCON_CLOCK_CNTL_0);
pc = RREG32_SMC(ixSMC_PC_C);
if ((0 == REG_GET_FIELD(clock_cntl, SMC_SYSCON_CLOCK_CNTL_0, ck_disable)) &&
(0x20100 <= pc))
return true;
return false;
}
static const struct amdgpu_asic_funcs cik_asic_funcs =
{
.read_disabled_bios = &cik_read_disabled_bios,
@ -1756,6 +1819,8 @@ static const struct amdgpu_asic_funcs cik_asic_funcs =
.invalidate_hdp = &cik_invalidate_hdp,
.need_full_reset = &cik_need_full_reset,
.init_doorbell_index = &legacy_doorbell_index_init,
.get_pcie_usage = &cik_get_pcie_usage,
.need_reset_on_init = &cik_need_reset_on_init,
};
static int cik_common_early_init(void *handle)
@ -2005,10 +2070,7 @@ int cik_set_ip_blocks(struct amdgpu_device *adev)
amdgpu_device_ip_block_add(adev, &cik_ih_ip_block);
amdgpu_device_ip_block_add(adev, &gfx_v7_2_ip_block);
amdgpu_device_ip_block_add(adev, &cik_sdma_ip_block);
if (amdgpu_dpm == -1)
amdgpu_device_ip_block_add(adev, &pp_smu_ip_block);
else
amdgpu_device_ip_block_add(adev, &ci_smu_ip_block);
amdgpu_device_ip_block_add(adev, &pp_smu_ip_block);
if (adev->enable_virtual_display)
amdgpu_device_ip_block_add(adev, &dce_virtual_ip_block);
#if defined(CONFIG_DRM_AMD_DC)
@ -2026,10 +2088,7 @@ int cik_set_ip_blocks(struct amdgpu_device *adev)
amdgpu_device_ip_block_add(adev, &cik_ih_ip_block);
amdgpu_device_ip_block_add(adev, &gfx_v7_3_ip_block);
amdgpu_device_ip_block_add(adev, &cik_sdma_ip_block);
if (amdgpu_dpm == -1)
amdgpu_device_ip_block_add(adev, &pp_smu_ip_block);
else
amdgpu_device_ip_block_add(adev, &ci_smu_ip_block);
amdgpu_device_ip_block_add(adev, &pp_smu_ip_block);
if (adev->enable_virtual_display)
amdgpu_device_ip_block_add(adev, &dce_virtual_ip_block);
#if defined(CONFIG_DRM_AMD_DC)

1
drivers/gpu/drm/amd/amdgpu/cik_dpm.h
View File

@ -24,7 +24,6 @@
#ifndef __CIK_DPM_H__
#define __CIK_DPM_H__
extern const struct amdgpu_ip_block_version ci_smu_ip_block;
extern const struct amdgpu_ip_block_version kv_smu_ip_block;
#endif

36
drivers/gpu/drm/amd/amdgpu/cik_ih.c
View File

@ -103,9 +103,9 @@ static void cik_ih_disable_interrupts(struct amdgpu_device *adev)
*/
static int cik_ih_irq_init(struct amdgpu_device *adev)
{
struct amdgpu_ih_ring *ih = &adev->irq.ih;
int rb_bufsz;
u32 interrupt_cntl, ih_cntl, ih_rb_cntl;
u64 wptr_off;
/* disable irqs */
cik_ih_disable_interrupts(adev);
@ -131,9 +131,8 @@ static int cik_ih_irq_init(struct amdgpu_device *adev)
ih_rb_cntl |= IH_RB_CNTL__WPTR_WRITEBACK_ENABLE_MASK;
/* set the writeback address whether it's enabled or not */
wptr_off = adev->wb.gpu_addr + (adev->irq.ih.wptr_offs * 4);
WREG32(mmIH_RB_WPTR_ADDR_LO, lower_32_bits(wptr_off));
WREG32(mmIH_RB_WPTR_ADDR_HI, upper_32_bits(wptr_off) & 0xFF);
WREG32(mmIH_RB_WPTR_ADDR_LO, lower_32_bits(ih->wptr_addr));
WREG32(mmIH_RB_WPTR_ADDR_HI, upper_32_bits(ih->wptr_addr) & 0xFF);
WREG32(mmIH_RB_CNTL, ih_rb_cntl);
@ -183,11 +182,12 @@ static void cik_ih_irq_disable(struct amdgpu_device *adev)
* Used by cik_irq_process().
* Returns the value of the wptr.
*/
static u32 cik_ih_get_wptr(struct amdgpu_device *adev)
static u32 cik_ih_get_wptr(struct amdgpu_device *adev,
struct amdgpu_ih_ring *ih)
{
u32 wptr, tmp;
wptr = le32_to_cpu(adev->wb.wb[adev->irq.ih.wptr_offs]);
wptr = le32_to_cpu(*ih->wptr_cpu);
if (wptr & IH_RB_WPTR__RB_OVERFLOW_MASK) {
wptr &= ~IH_RB_WPTR__RB_OVERFLOW_MASK;
@ -196,13 +196,13 @@ static u32 cik_ih_get_wptr(struct amdgpu_device *adev)
* this should allow us to catchup.
*/
dev_warn(adev->dev, "IH ring buffer overflow (0x%08X, 0x%08X, 0x%08X)\n",
wptr, adev->irq.ih.rptr, (wptr + 16) & adev->irq.ih.ptr_mask);
adev->irq.ih.rptr = (wptr + 16) & adev->irq.ih.ptr_mask;
wptr, ih->rptr, (wptr + 16) & ih->ptr_mask);
ih->rptr = (wptr + 16) & ih->ptr_mask;
tmp = RREG32(mmIH_RB_CNTL);
tmp |= IH_RB_CNTL__WPTR_OVERFLOW_CLEAR_MASK;
WREG32(mmIH_RB_CNTL, tmp);
}
return (wptr & adev->irq.ih.ptr_mask);
return (wptr & ih->ptr_mask);
}
/* CIK IV Ring
@ -237,16 +237,17 @@ static u32 cik_ih_get_wptr(struct amdgpu_device *adev)
* position and also advance the position.
*/
static void cik_ih_decode_iv(struct amdgpu_device *adev,
struct amdgpu_ih_ring *ih,
struct amdgpu_iv_entry *entry)
{
/* wptr/rptr are in bytes! */
u32 ring_index = adev->irq.ih.rptr >> 2;
u32 ring_index = ih->rptr >> 2;
uint32_t dw[4];
dw[0] = le32_to_cpu(adev->irq.ih.ring[ring_index + 0]);
dw[1] = le32_to_cpu(adev->irq.ih.ring[ring_index + 1]);
dw[2] = le32_to_cpu(adev->irq.ih.ring[ring_index + 2]);
dw[3] = le32_to_cpu(adev->irq.ih.ring[ring_index + 3]);
dw[0] = le32_to_cpu(ih->ring[ring_index + 0]);
dw[1] = le32_to_cpu(ih->ring[ring_index + 1]);
dw[2] = le32_to_cpu(ih->ring[ring_index + 2]);
dw[3] = le32_to_cpu(ih->ring[ring_index + 3]);
entry->client_id = AMDGPU_IRQ_CLIENTID_LEGACY;
entry->src_id = dw[0] & 0xff;
@ -256,7 +257,7 @@ static void cik_ih_decode_iv(struct amdgpu_device *adev,
entry->pasid = (dw[2] >> 16) & 0xffff;
/* wptr/rptr are in bytes! */
adev->irq.ih.rptr += 16;
ih->rptr += 16;
}
/**
@ -266,9 +267,10 @@ static void cik_ih_decode_iv(struct amdgpu_device *adev,
*
* Set the IH ring buffer rptr.
*/
static void cik_ih_set_rptr(struct amdgpu_device *adev)
static void cik_ih_set_rptr(struct amdgpu_device *adev,
struct amdgpu_ih_ring *ih)
{
WREG32(mmIH_RB_RPTR, adev->irq.ih.rptr);
WREG32(mmIH_RB_RPTR, ih->rptr);
}
static int cik_ih_early_init(void *handle)

2
drivers/gpu/drm/amd/amdgpu/cik_sdma.c
View File

@ -220,7 +220,7 @@ static void cik_sdma_ring_insert_nop(struct amdgpu_ring *ring, uint32_t count)
static void cik_sdma_ring_emit_ib(struct amdgpu_ring *ring,
struct amdgpu_job *job,
struct amdgpu_ib *ib,
bool ctx_switch)
uint32_t flags)
{
unsigned vmid = AMDGPU_JOB_GET_VMID(job);
u32 extra_bits = vmid & 0xf;

40
drivers/gpu/drm/amd/amdgpu/cz_ih.c
View File

@ -103,9 +103,9 @@ static void cz_ih_disable_interrupts(struct amdgpu_device *adev)
*/
static int cz_ih_irq_init(struct amdgpu_device *adev)
{
int rb_bufsz;
struct amdgpu_ih_ring *ih = &adev->irq.ih;
u32 interrupt_cntl, ih_cntl, ih_rb_cntl;
u64 wptr_off;
int rb_bufsz;
/* disable irqs */
cz_ih_disable_interrupts(adev);
@ -133,9 +133,8 @@ static int cz_ih_irq_init(struct amdgpu_device *adev)
ih_rb_cntl = REG_SET_FIELD(ih_rb_cntl, IH_RB_CNTL, WPTR_WRITEBACK_ENABLE, 1);
/* set the writeback address whether it's enabled or not */
wptr_off = adev->wb.gpu_addr + (adev->irq.ih.wptr_offs * 4);
WREG32(mmIH_RB_WPTR_ADDR_LO, lower_32_bits(wptr_off));
WREG32(mmIH_RB_WPTR_ADDR_HI, upper_32_bits(wptr_off) & 0xFF);
WREG32(mmIH_RB_WPTR_ADDR_LO, lower_32_bits(ih->wptr_addr));
WREG32(mmIH_RB_WPTR_ADDR_HI, upper_32_bits(ih->wptr_addr) & 0xFF);
WREG32(mmIH_RB_CNTL, ih_rb_cntl);
@ -185,11 +184,12 @@ static void cz_ih_irq_disable(struct amdgpu_device *adev)
* Used by cz_irq_process(VI).
* Returns the value of the wptr.
*/
static u32 cz_ih_get_wptr(struct amdgpu_device *adev)
static u32 cz_ih_get_wptr(struct amdgpu_device *adev,
struct amdgpu_ih_ring *ih)
{
u32 wptr, tmp;
wptr = le32_to_cpu(adev->wb.wb[adev->irq.ih.wptr_offs]);
wptr = le32_to_cpu(*ih->wptr_cpu);
if (REG_GET_FIELD(wptr, IH_RB_WPTR, RB_OVERFLOW)) {
wptr = REG_SET_FIELD(wptr, IH_RB_WPTR, RB_OVERFLOW, 0);
@ -198,13 +198,13 @@ static u32 cz_ih_get_wptr(struct amdgpu_device *adev)
* this should allow us to catchup.
*/
dev_warn(adev->dev, "IH ring buffer overflow (0x%08X, 0x%08X, 0x%08X)\n",
wptr, adev->irq.ih.rptr, (wptr + 16) & adev->irq.ih.ptr_mask);
adev->irq.ih.rptr = (wptr + 16) & adev->irq.ih.ptr_mask;
wptr, ih->rptr, (wptr + 16) & ih->ptr_mask);
ih->rptr = (wptr + 16) & ih->ptr_mask;
tmp = RREG32(mmIH_RB_CNTL);
tmp = REG_SET_FIELD(tmp, IH_RB_CNTL, WPTR_OVERFLOW_CLEAR, 1);
WREG32(mmIH_RB_CNTL, tmp);
}
return (wptr & adev->irq.ih.ptr_mask);
return (wptr & ih->ptr_mask);
}
/**
@ -216,16 +216,17 @@ static u32 cz_ih_get_wptr(struct amdgpu_device *adev)
* position and also advance the position.
*/
static void cz_ih_decode_iv(struct amdgpu_device *adev,
struct amdgpu_iv_entry *entry)
struct amdgpu_ih_ring *ih,
struct amdgpu_iv_entry *entry)
{
/* wptr/rptr are in bytes! */
u32 ring_index = adev->irq.ih.rptr >> 2;
u32 ring_index = ih->rptr >> 2;
uint32_t dw[4];
dw[0] = le32_to_cpu(adev->irq.ih.ring[ring_index + 0]);
dw[1] = le32_to_cpu(adev->irq.ih.ring[ring_index + 1]);
dw[2] = le32_to_cpu(adev->irq.ih.ring[ring_index + 2]);
dw[3] = le32_to_cpu(adev->irq.ih.ring[ring_index + 3]);
dw[0] = le32_to_cpu(ih->ring[ring_index + 0]);
dw[1] = le32_to_cpu(ih->ring[ring_index + 1]);
dw[2] = le32_to_cpu(ih->ring[ring_index + 2]);
dw[3] = le32_to_cpu(ih->ring[ring_index + 3]);
entry->client_id = AMDGPU_IRQ_CLIENTID_LEGACY;
entry->src_id = dw[0] & 0xff;
@ -235,7 +236,7 @@ static void cz_ih_decode_iv(struct amdgpu_device *adev,
entry->pasid = (dw[2] >> 16) & 0xffff;
/* wptr/rptr are in bytes! */
adev->irq.ih.rptr += 16;
ih->rptr += 16;
}
/**
@ -245,9 +246,10 @@ static void cz_ih_decode_iv(struct amdgpu_device *adev,
*
* Set the IH ring buffer rptr.
*/
static void cz_ih_set_rptr(struct amdgpu_device *adev)
static void cz_ih_set_rptr(struct amdgpu_device *adev,
struct amdgpu_ih_ring *ih)
{
WREG32(mmIH_RB_RPTR, adev->irq.ih.rptr);
WREG32(mmIH_RB_RPTR, ih->rptr);
}
static int cz_ih_early_init(void *handle)

2
drivers/gpu/drm/amd/amdgpu/dce_v10_0.c
View File

@ -1682,7 +1682,7 @@ static void dce_v10_0_afmt_setmode(struct drm_encoder *encoder,
dce_v10_0_audio_write_sad_regs(encoder);
dce_v10_0_audio_write_latency_fields(encoder, mode);
err = drm_hdmi_avi_infoframe_from_display_mode(&frame, mode, false);
err = drm_hdmi_avi_infoframe_from_display_mode(&frame, connector, mode);
if (err < 0) {
DRM_ERROR("failed to setup AVI infoframe: %zd\n", err);
return;

2
drivers/gpu/drm/amd/amdgpu/dce_v11_0.c
View File

@ -1724,7 +1724,7 @@ static void dce_v11_0_afmt_setmode(struct drm_encoder *encoder,
dce_v11_0_audio_write_sad_regs(encoder);
dce_v11_0_audio_write_latency_fields(encoder, mode);
err = drm_hdmi_avi_infoframe_from_display_mode(&frame, mode, false);
err = drm_hdmi_avi_infoframe_from_display_mode(&frame, connector, mode);
if (err < 0) {
DRM_ERROR("failed to setup AVI infoframe: %zd\n", err);
return;

5
drivers/gpu/drm/amd/amdgpu/dce_v6_0.c
View File

@ -1423,6 +1423,7 @@ static void dce_v6_0_audio_set_avi_infoframe(struct drm_encoder *encoder,
struct amdgpu_device *adev = dev->dev_private;
struct amdgpu_encoder *amdgpu_encoder = to_amdgpu_encoder(encoder);
struct amdgpu_encoder_atom_dig *dig = amdgpu_encoder->enc_priv;
struct drm_connector *connector = amdgpu_get_connector_for_encoder(encoder);
struct hdmi_avi_infoframe frame;
u8 buffer[HDMI_INFOFRAME_HEADER_SIZE + HDMI_AVI_INFOFRAME_SIZE];
uint8_t *payload = buffer + 3;
@ -1430,7 +1431,7 @@ static void dce_v6_0_audio_set_avi_infoframe(struct drm_encoder *encoder,
ssize_t err;
u32 tmp;
err = drm_hdmi_avi_infoframe_from_display_mode(&frame, mode, false);
err = drm_hdmi_avi_infoframe_from_display_mode(&frame, connector, mode);
if (err < 0) {
DRM_ERROR("failed to setup AVI infoframe: %zd\n", err);
return;
@ -2979,7 +2980,7 @@ static int dce_v6_0_pageflip_irq(struct amdgpu_device *adev,
struct amdgpu_irq_src *source,
struct amdgpu_iv_entry *entry)
{
unsigned long flags;
unsigned long flags;
unsigned crtc_id;
struct amdgpu_crtc *amdgpu_crtc;
struct amdgpu_flip_work *works;

2
drivers/gpu/drm/amd/amdgpu/dce_v8_0.c
View File

@ -1616,7 +1616,7 @@ static void dce_v8_0_afmt_setmode(struct drm_encoder *encoder,
dce_v8_0_audio_write_sad_regs(encoder);
dce_v8_0_audio_write_latency_fields(encoder, mode);
err = drm_hdmi_avi_infoframe_from_display_mode(&frame, mode, false);
err = drm_hdmi_avi_infoframe_from_display_mode(&frame, connector, mode);
if (err < 0) {
DRM_ERROR("failed to setup AVI infoframe: %zd\n", err);
return;

4
drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c
View File

@ -1842,13 +1842,13 @@ static void gfx_v6_0_ring_emit_fence(struct amdgpu_ring *ring, u64 addr,
static void gfx_v6_0_ring_emit_ib(struct amdgpu_ring *ring,
struct amdgpu_job *job,
struct amdgpu_ib *ib,
bool ctx_switch)
uint32_t flags)
{
unsigned vmid = AMDGPU_JOB_GET_VMID(job);
u32 header, control = 0;
/* insert SWITCH_BUFFER packet before first IB in the ring frame */
if (ctx_switch) {
if (flags & AMDGPU_HAVE_CTX_SWITCH) {
amdgpu_ring_write(ring, PACKET3(PACKET3_SWITCH_BUFFER, 0));
amdgpu_ring_write(ring, 0);
}

25
drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c
View File

@ -2228,13 +2228,13 @@ static void gfx_v7_0_ring_emit_fence_compute(struct amdgpu_ring *ring,
static void gfx_v7_0_ring_emit_ib_gfx(struct amdgpu_ring *ring,
struct amdgpu_job *job,
struct amdgpu_ib *ib,
bool ctx_switch)
uint32_t flags)
{
unsigned vmid = AMDGPU_JOB_GET_VMID(job);
u32 header, control = 0;
/* insert SWITCH_BUFFER packet before first IB in the ring frame */
if (ctx_switch) {
if (flags & AMDGPU_HAVE_CTX_SWITCH) {
amdgpu_ring_write(ring, PACKET3(PACKET3_SWITCH_BUFFER, 0));
amdgpu_ring_write(ring, 0);
}
@ -2259,11 +2259,27 @@ static void gfx_v7_0_ring_emit_ib_gfx(struct amdgpu_ring *ring,
static void gfx_v7_0_ring_emit_ib_compute(struct amdgpu_ring *ring,
struct amdgpu_job *job,
struct amdgpu_ib *ib,
bool ctx_switch)
uint32_t flags)
{
unsigned vmid = AMDGPU_JOB_GET_VMID(job);
u32 control = INDIRECT_BUFFER_VALID | ib->length_dw | (vmid << 24);
/* Currently, there is a high possibility to get wave ID mismatch
* between ME and GDS, leading to a hw deadlock, because ME generates
* different wave IDs than the GDS expects. This situation happens
* randomly when at least 5 compute pipes use GDS ordered append.
* The wave IDs generated by ME are also wrong after suspend/resume.
* Those are probably bugs somewhere else in the kernel driver.
*
* Writing GDS_COMPUTE_MAX_WAVE_ID resets wave ID counters in ME and
* GDS to 0 for this ring (me/pipe).
*/
if (ib->flags & AMDGPU_IB_FLAG_RESET_GDS_MAX_WAVE_ID) {
amdgpu_ring_write(ring, PACKET3(PACKET3_SET_CONFIG_REG, 1));
amdgpu_ring_write(ring, mmGDS_COMPUTE_MAX_WAVE_ID - PACKET3_SET_CONFIG_REG_START);
amdgpu_ring_write(ring, ring->adev->gds.gds_compute_max_wave_id);
}
amdgpu_ring_write(ring, PACKET3(PACKET3_INDIRECT_BUFFER, 2));
amdgpu_ring_write(ring,
#ifdef __BIG_ENDIAN
@ -5000,7 +5016,7 @@ static const struct amdgpu_ring_funcs gfx_v7_0_ring_funcs_compute = {
7 + /* gfx_v7_0_ring_emit_pipeline_sync */
CIK_FLUSH_GPU_TLB_NUM_WREG * 5 + 7 + /* gfx_v7_0_ring_emit_vm_flush */
7 + 7 + 7, /* gfx_v7_0_ring_emit_fence_compute x3 for user fence, vm fence */
.emit_ib_size = 4, /* gfx_v7_0_ring_emit_ib_compute */
.emit_ib_size = 7, /* gfx_v7_0_ring_emit_ib_compute */
.emit_ib = gfx_v7_0_ring_emit_ib_compute,
.emit_fence = gfx_v7_0_ring_emit_fence_compute,
.emit_pipeline_sync = gfx_v7_0_ring_emit_pipeline_sync,
@ -5057,6 +5073,7 @@ static void gfx_v7_0_set_gds_init(struct amdgpu_device *adev)
adev->gds.mem.total_size = RREG32(mmGDS_VMID0_SIZE);
adev->gds.gws.total_size = 64;
adev->gds.oa.total_size = 16;
adev->gds.gds_compute_max_wave_id = RREG32(mmGDS_COMPUTE_MAX_WAVE_ID);
if (adev->gds.mem.total_size == 64 * 1024) {
adev->gds.mem.gfx_partition_size = 4096;

25
drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
View File

@ -6047,7 +6047,7 @@ static void gfx_v8_0_ring_emit_vgt_flush(struct amdgpu_ring *ring)
static void gfx_v8_0_ring_emit_ib_gfx(struct amdgpu_ring *ring,
struct amdgpu_job *job,
struct amdgpu_ib *ib,
bool ctx_switch)
uint32_t flags)
{
unsigned vmid = AMDGPU_JOB_GET_VMID(job);
u32 header, control = 0;
@ -6079,11 +6079,27 @@ static void gfx_v8_0_ring_emit_ib_gfx(struct amdgpu_ring *ring,
static void gfx_v8_0_ring_emit_ib_compute(struct amdgpu_ring *ring,
struct amdgpu_job *job,
struct amdgpu_ib *ib,
bool ctx_switch)
uint32_t flags)
{
unsigned vmid = AMDGPU_JOB_GET_VMID(job);
u32 control = INDIRECT_BUFFER_VALID | ib->length_dw | (vmid << 24);
/* Currently, there is a high possibility to get wave ID mismatch
* between ME and GDS, leading to a hw deadlock, because ME generates
* different wave IDs than the GDS expects. This situation happens
* randomly when at least 5 compute pipes use GDS ordered append.
* The wave IDs generated by ME are also wrong after suspend/resume.
* Those are probably bugs somewhere else in the kernel driver.
*
* Writing GDS_COMPUTE_MAX_WAVE_ID resets wave ID counters in ME and
* GDS to 0 for this ring (me/pipe).
*/
if (ib->flags & AMDGPU_IB_FLAG_RESET_GDS_MAX_WAVE_ID) {
amdgpu_ring_write(ring, PACKET3(PACKET3_SET_CONFIG_REG, 1));
amdgpu_ring_write(ring, mmGDS_COMPUTE_MAX_WAVE_ID - PACKET3_SET_CONFIG_REG_START);
amdgpu_ring_write(ring, ring->adev->gds.gds_compute_max_wave_id);
}
amdgpu_ring_write(ring, PACKET3(PACKET3_INDIRECT_BUFFER, 2));
amdgpu_ring_write(ring,
#ifdef __BIG_ENDIAN
@ -6890,7 +6906,7 @@ static const struct amdgpu_ring_funcs gfx_v8_0_ring_funcs_compute = {
7 + /* gfx_v8_0_ring_emit_pipeline_sync */
VI_FLUSH_GPU_TLB_NUM_WREG * 5 + 7 + /* gfx_v8_0_ring_emit_vm_flush */
7 + 7 + 7, /* gfx_v8_0_ring_emit_fence_compute x3 for user fence, vm fence */
.emit_ib_size = 4, /* gfx_v8_0_ring_emit_ib_compute */
.emit_ib_size = 7, /* gfx_v8_0_ring_emit_ib_compute */
.emit_ib = gfx_v8_0_ring_emit_ib_compute,
.emit_fence = gfx_v8_0_ring_emit_fence_compute,
.emit_pipeline_sync = gfx_v8_0_ring_emit_pipeline_sync,
@ -6920,7 +6936,7 @@ static const struct amdgpu_ring_funcs gfx_v8_0_ring_funcs_kiq = {
7 + /* gfx_v8_0_ring_emit_pipeline_sync */
17 + /* gfx_v8_0_ring_emit_vm_flush */
7 + 7 + 7, /* gfx_v8_0_ring_emit_fence_kiq x3 for user fence, vm fence */
.emit_ib_size = 4, /* gfx_v8_0_ring_emit_ib_compute */
.emit_ib_size = 7, /* gfx_v8_0_ring_emit_ib_compute */
.emit_fence = gfx_v8_0_ring_emit_fence_kiq,
.test_ring = gfx_v8_0_ring_test_ring,
.insert_nop = amdgpu_ring_insert_nop,
@ -6996,6 +7012,7 @@ static void gfx_v8_0_set_gds_init(struct amdgpu_device *adev)
adev->gds.mem.total_size = RREG32(mmGDS_VMID0_SIZE);
adev->gds.gws.total_size = 64;
adev->gds.oa.total_size = 16;
adev->gds.gds_compute_max_wave_id = RREG32(mmGDS_COMPUTE_MAX_WAVE_ID);
if (adev->gds.mem.total_size == 64 * 1024) {
adev->gds.mem.gfx_partition_size = 4096;

44
drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
View File

@ -3972,7 +3972,7 @@ static void gfx_v9_0_ring_emit_hdp_flush(struct amdgpu_ring *ring)
static void gfx_v9_0_ring_emit_ib_gfx(struct amdgpu_ring *ring,
struct amdgpu_job *job,
struct amdgpu_ib *ib,
bool ctx_switch)
uint32_t flags)
{
unsigned vmid = AMDGPU_JOB_GET_VMID(job);
u32 header, control = 0;
@ -4005,11 +4005,27 @@ static void gfx_v9_0_ring_emit_ib_gfx(struct amdgpu_ring *ring,
static void gfx_v9_0_ring_emit_ib_compute(struct amdgpu_ring *ring,
struct amdgpu_job *job,
struct amdgpu_ib *ib,
bool ctx_switch)
uint32_t flags)
{
unsigned vmid = AMDGPU_JOB_GET_VMID(job);
u32 control = INDIRECT_BUFFER_VALID | ib->length_dw | (vmid << 24);
/* Currently, there is a high possibility to get wave ID mismatch
* between ME and GDS, leading to a hw deadlock, because ME generates
* different wave IDs than the GDS expects. This situation happens
* randomly when at least 5 compute pipes use GDS ordered append.
* The wave IDs generated by ME are also wrong after suspend/resume.
* Those are probably bugs somewhere else in the kernel driver.
*
* Writing GDS_COMPUTE_MAX_WAVE_ID resets wave ID counters in ME and
* GDS to 0 for this ring (me/pipe).
*/
if (ib->flags & AMDGPU_IB_FLAG_RESET_GDS_MAX_WAVE_ID) {
amdgpu_ring_write(ring, PACKET3(PACKET3_SET_CONFIG_REG, 1));
amdgpu_ring_write(ring, mmGDS_COMPUTE_MAX_WAVE_ID);
amdgpu_ring_write(ring, ring->adev->gds.gds_compute_max_wave_id);
}
amdgpu_ring_write(ring, PACKET3(PACKET3_INDIRECT_BUFFER, 2));
BUG_ON(ib->gpu_addr & 0x3); /* Dword align */
amdgpu_ring_write(ring,
@ -4729,7 +4745,7 @@ static const struct amdgpu_ring_funcs gfx_v9_0_ring_funcs_compute = {
SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT * 7 +
2 + /* gfx_v9_0_ring_emit_vm_flush */
8 + 8 + 8, /* gfx_v9_0_ring_emit_fence x3 for user fence, vm fence */
.emit_ib_size = 4, /* gfx_v9_0_ring_emit_ib_compute */
.emit_ib_size = 7, /* gfx_v9_0_ring_emit_ib_compute */
.emit_ib = gfx_v9_0_ring_emit_ib_compute,
.emit_fence = gfx_v9_0_ring_emit_fence,
.emit_pipeline_sync = gfx_v9_0_ring_emit_pipeline_sync,
@ -4764,7 +4780,7 @@ static const struct amdgpu_ring_funcs gfx_v9_0_ring_funcs_kiq = {
SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT * 7 +
2 + /* gfx_v9_0_ring_emit_vm_flush */
8 + 8 + 8, /* gfx_v9_0_ring_emit_fence_kiq x3 for user fence, vm fence */
.emit_ib_size = 4, /* gfx_v9_0_ring_emit_ib_compute */
.emit_ib_size = 7, /* gfx_v9_0_ring_emit_ib_compute */
.emit_fence = gfx_v9_0_ring_emit_fence_kiq,
.test_ring = gfx_v9_0_ring_test_ring,
.insert_nop = amdgpu_ring_insert_nop,
@ -4846,6 +4862,26 @@ static void gfx_v9_0_set_gds_init(struct amdgpu_device *adev)
break;
}
switch (adev->asic_type) {
case CHIP_VEGA10:
case CHIP_VEGA20:
adev->gds.gds_compute_max_wave_id = 0x7ff;
break;
case CHIP_VEGA12:
adev->gds.gds_compute_max_wave_id = 0x27f;
break;
case CHIP_RAVEN:
if (adev->rev_id >= 0x8)
adev->gds.gds_compute_max_wave_id = 0x77; /* raven2 */
else
adev->gds.gds_compute_max_wave_id = 0x15f; /* raven1 */
break;
default:
/* this really depends on the chip */
adev->gds.gds_compute_max_wave_id = 0x7ff;
break;
}
adev->gds.gws.total_size = 64;
adev->gds.oa.total_size = 16;

3
drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c
View File

@ -1471,8 +1471,9 @@ static int gmc_v8_0_process_interrupt(struct amdgpu_device *adev,
gmc_v8_0_set_fault_enable_default(adev, false);
if (printk_ratelimit()) {
struct amdgpu_task_info task_info = { 0 };
struct amdgpu_task_info task_info;
memset(&task_info, 0, sizeof(struct amdgpu_task_info));
amdgpu_vm_get_task_info(adev, entry->pasid, &task_info);
dev_err(adev->dev, "GPU fault detected: %d 0x%08x for process %s pid %d thread %s pid %d\n",

Some files were not shown because too many files have changed in this diff Show More