[Solved by NeddySeagoon gorg86] AMDGPU Rocm Does't work

Message

4nn13 · Post by **4nn13** » Sun Sep 03, 2023 11:47 am

after emerge the following packages Rocm still doesn't work

Packages:

dev-libs/rocm-opencl-runtime
dev-libs/rocr-runtime
dev-libs/roct-thunk-interface
dev-util/rocm-smi
x11-drivers/xf86-video-amdgpu
x11-drivers/xf86-video-ati

Kernel:

Code: Select all

6.1.46-gentoo-dist

Firmware:

Code: Select all

CONFIG_EXTRA_FIRMWARE="radeon/R300_cp.bin"
CONFIG_EXTRA_FIRMWARE_DIR="/lib/firmware"

dmesg kfd:

Code: Select all

$ ~ dmesg |grep kfd
$ ~ dmesg |grep KFD
$ ~

Radeon dmesg:

Code: Select all

[   10.951530] [drm] radeon kernel modesetting enabled.
[   10.951958] radeon 0000:01:00.0: enabling device (0000 -> 0003)
[   10.970889] radeon 0000:01:00.0: VRAM: 2048M 0x0000000000000000 - 0x000000007FFFFFFF (2048M used)
[   10.970892] radeon 0000:01:00.0: GTT: 2048M 0x0000000080000000 - 0x00000000FFFFFFFF
[   10.970913] [drm] radeon: 2048M of VRAM memory ready
[   10.970914] [drm] radeon: 2048M of GTT memory ready.
[   10.970922] Loading firmware: radeon/hainan_pfp.bin
[   11.422505] Loading firmware: radeon/hainan_me.bin
[   11.532384] Loading firmware: radeon/hainan_ce.bin
[   11.553628] Loading firmware: radeon/hainan_rlc.bin
[   11.555590] Loading firmware: radeon/si58_mc.bin
[   11.592777] Loading firmware: radeon/banks_k_2_smc.bin
[   11.632724] [drm] radeon: dpm initialized
[   11.640218] radeon 0000:01:00.0: WB enabled
[   11.640220] radeon 0000:01:00.0: fence driver on ring 0 use gpu addr 0x0000000080000c00
[   11.640222] radeon 0000:01:00.0: fence driver on ring 1 use gpu addr 0x0000000080000c04
[   11.640223] radeon 0000:01:00.0: fence driver on ring 2 use gpu addr 0x0000000080000c08
[   11.640224] radeon 0000:01:00.0: fence driver on ring 3 use gpu addr 0x0000000080000c0c
[   11.640225] radeon 0000:01:00.0: fence driver on ring 4 use gpu addr 0x0000000080000c10
[   11.640227] radeon 0000:01:00.0: radeon: MSI limited to 32-bit
[   11.640271] radeon 0000:01:00.0: radeon: using MSI.
[   11.640286] [drm] radeon: irq initialized.
[   11.884494] [drm] Initialized radeon 2.50.0 20080528 for 0000:01:00.0 on minor 0
[   38.172863] radeon 0000:01:00.0: WB enabled
[   38.172866] radeon 0000:01:00.0: fence driver on ring 0 use gpu addr 0x0000000080000c00
[   38.172868] radeon 0000:01:00.0: fence driver on ring 1 use gpu addr 0x0000000080000c04
[   38.172870] radeon 0000:01:00.0: fence driver on ring 2 use gpu addr 0x0000000080000c08
[   38.172871] radeon 0000:01:00.0: fence driver on ring 3 use gpu addr 0x0000000080000c0c
[   38.172872] radeon 0000:01:00.0: fence driver on ring 4 use gpu addr 0x0000000080000c10
[   38.173516] debugfs: File 'radeon_ring_gfx' in directory '0' already present!
[   38.173519] debugfs: File 'radeon_ring_cp1' in directory '0' already present!
[   38.173520] debugfs: File 'radeon_ring_cp2' in directory '0' already present!
[   38.173521] debugfs: File 'radeon_ring_dma1' in directory '0' already present!
[   38.173522] debugfs: File 'radeon_ring_dma2' in directory '0' already present!
[ 5631.071941] radeon 0000:01:00.0: WB enabled
[ 5631.071944] radeon 0000:01:00.0: fence driver on ring 0 use gpu addr 0x0000000080000c00
[ 5631.071946] radeon 0000:01:00.0: fence driver on ring 1 use gpu addr 0x0000000080000c04
[ 5631.071947] radeon 0000:01:00.0: fence driver on ring 2 use gpu addr 0x0000000080000c08
[ 5631.071948] radeon 0000:01:00.0: fence driver on ring 3 use gpu addr 0x0000000080000c0c
[ 5631.071949] radeon 0000:01:00.0: fence driver on ring 4 use gpu addr 0x0000000080000c10
[ 5631.072586] debugfs: File 'radeon_ring_gfx' in directory '0' already present!
[ 5631.072589] debugfs: File 'radeon_ring_cp1' in directory '0' already present!
[ 5631.072591] debugfs: File 'radeon_ring_cp2' in directory '0' already present!
[ 5631.072592] debugfs: File 'radeon_ring_dma1' in directory '0' already present!
[ 5631.072594] debugfs: File 'radeon_ring_dma2' in directory '0' already present!
[ 6702.618787] radeon 0000:01:00.0: WB enabled
[ 6702.618789] radeon 0000:01:00.0: fence driver on ring 0 use gpu addr 0x0000000080000c00
[ 6702.618791] radeon 0000:01:00.0: fence driver on ring 1 use gpu addr 0x0000000080000c04
[ 6702.618793] radeon 0000:01:00.0: fence driver on ring 2 use gpu addr 0x0000000080000c08
[ 6702.618794] radeon 0000:01:00.0: fence driver on ring 3 use gpu addr 0x0000000080000c0c
[ 6702.618795] radeon 0000:01:00.0: fence driver on ring 4 use gpu addr 0x0000000080000c10
[ 6702.619411] debugfs: File 'radeon_ring_gfx' in directory '0' already present!
[ 6702.619413] debugfs: File 'radeon_ring_cp1' in directory '0' already present!
[ 6702.619415] debugfs: File 'radeon_ring_cp2' in directory '0' already present!
[ 6702.619416] debugfs: File 'radeon_ring_dma1' in directory '0' already present!
[ 6702.619417] debugfs: File 'radeon_ring_dma2' in directory '0' already present!

rocm-smi:

Code: Select all

======================= ROCm System Management Interface =======================
WARNING: No AMD GPUs specified
================================= Concise Info =================================
GPU  Temp (DieEdge)  AvgPwr  SCLK  MCLK  Fan  Perf  PwrCap  VRAM%  GPU%  
================================================================================
============================= End of ROCm SMI Log ==============================

so what is the problem overhere?

Post by **NeddySeagoon** » Sun Sep 03, 2023 11:50 am

4nn13,

Please post the output of

Code: Select all

lspci -nnk

4nn13 · Post by **4nn13** » Sun Sep 03, 2023 11:51 am

NeddySeagoon wrote:4nn13,

Please post the output of
Code: Select all
lspci -nnk

Thanks for the quickly reply

here is the output

Code: Select all

00:00.0 Host bridge [0600]: Intel Corporation Coffee Lake HOST and DRAM Controller [8086:3e34] (rev 0b)
	DeviceName: Onboard - Other
	Subsystem: Dell Coffee Lake HOST and DRAM Controller [1028:08c9]
	Kernel driver in use: skl_uncore
00:02.0 VGA compatible controller [0300]: Intel Corporation WhiskeyLake-U GT2 [UHD Graphics 620] [8086:3ea0]
	DeviceName: Onboard - Video
	Subsystem: Dell WhiskeyLake-U GT2 [UHD Graphics 620] [1028:08c9]
	Kernel driver in use: i915
	Kernel modules: i915
00:04.0 Signal processing controller [1180]: Intel Corporation Xeon E3-1200 v5/E3-1500 v5/6th Gen Core Processor Thermal Subsystem [8086:1903] (rev 0b)
	DeviceName: Onboard - Other
	Subsystem: Dell Xeon E3-1200 v5/E3-1500 v5/6th Gen Core Processor Thermal Subsystem [1028:08c9]
	Kernel driver in use: proc_thermal
	Kernel modules: processor_thermal_device_pci_legacy
00:08.0 System peripheral [0880]: Intel Corporation Xeon E3-1200 v5/v6 / E3-1500 v5 / 6th/7th/8th Gen Core Processor Gaussian Mixture Model [8086:1911]
	DeviceName: Onboard - Other
	Subsystem: Dell Xeon E3-1200 v5/v6 / E3-1500 v5 / 6th/7th/8th Gen Core Processor Gaussian Mixture Model [1028:08c9]
00:12.0 Signal processing controller [1180]: Intel Corporation Cannon Point-LP Thermal Controller [8086:9df9] (rev 30)
	DeviceName: Onboard - Other
	Subsystem: Dell Cannon Point-LP Thermal Controller [1028:08c9]
	Kernel driver in use: intel_pch_thermal
	Kernel modules: intel_pch_thermal
00:14.0 USB controller [0c03]: Intel Corporation Cannon Point-LP USB 3.1 xHCI Controller [8086:9ded] (rev 30)
	DeviceName: Onboard - Other
	Subsystem: Dell Cannon Point-LP USB 3.1 xHCI Controller [1028:08c9]
	Kernel driver in use: xhci_hcd
00:14.2 RAM memory [0500]: Intel Corporation Cannon Point-LP Shared SRAM [8086:9def] (rev 30)
	DeviceName: Onboard - Other
	Subsystem: Dell Cannon Point-LP Shared SRAM [1028:08c9]
00:15.0 Serial bus controller [0c80]: Intel Corporation Cannon Point-LP Serial IO I2C Controller #0 [8086:9de8] (rev 30)
	DeviceName: Onboard - Other
	Subsystem: Dell Cannon Point-LP Serial IO I2C Controller [1028:08c9]
	Kernel driver in use: intel-lpss
00:15.1 Serial bus controller [0c80]: Intel Corporation Cannon Point-LP Serial IO I2C Controller #1 [8086:9de9] (rev 30)
	DeviceName: Onboard - Other
	Subsystem: Dell Cannon Point-LP Serial IO I2C Controller [1028:08c9]
	Kernel driver in use: intel-lpss
00:17.0 RAID bus controller [0104]: Intel Corporation 82801 Mobile SATA Controller [RAID mode] [8086:282a] (rev 30)
	DeviceName: Onboard - Other
	Subsystem: Dell 82801 Mobile SATA Controller [RAID mode] [1028:08c9]
	Kernel driver in use: ahci
00:19.0 Serial bus controller [0c80]: Intel Corporation Cannon Point-LP Serial IO I2C Host Controller [8086:9dc5] (rev 30)
	DeviceName: Onboard - Other
	Subsystem: Dell Cannon Point-LP Serial IO I2C Host Controller [1028:08c9]
	Kernel driver in use: intel-lpss
00:1c.0 PCI bridge [0604]: Intel Corporation Cannon Point-LP PCI Express Root Port #5 [8086:9dbc] (rev f0)
	Subsystem: Dell Cannon Point-LP PCI Express Root Port [1028:08c9]
	Kernel driver in use: pcieport
00:1d.0 PCI bridge [0604]: Intel Corporation Cannon Point-LP PCI Express Root Port #9 [8086:9db0] (rev f0)
	Subsystem: Dell Cannon Point-LP PCI Express Root Port [1028:08c9]
	Kernel driver in use: pcieport
00:1d.1 PCI bridge [0604]: Intel Corporation Cannon Point-LP PCI Express Root Port #10 [8086:9db1] (rev f0)
	Subsystem: Dell Cannon Point-LP PCI Express Root Port [1028:08c9]
	Kernel driver in use: pcieport
00:1f.0 ISA bridge [0601]: Intel Corporation Cannon Point-LP LPC Controller [8086:9d84] (rev 30)
	DeviceName: Onboard - Other
	Subsystem: Dell Cannon Point-LP LPC Controller [1028:08c9]
00:1f.3 Audio device [0403]: Intel Corporation Cannon Point-LP High Definition Audio Controller [8086:9dc8] (rev 30)
	DeviceName: Onboard - Sound
	Subsystem: Dell Cannon Point-LP High Definition Audio Controller [1028:08c9]
	Kernel driver in use: snd_hda_intel
	Kernel modules: snd_hda_intel, snd_soc_skl, snd_sof_pci_intel_cnl
00:1f.4 SMBus [0c05]: Intel Corporation Cannon Point-LP SMBus Controller [8086:9da3] (rev 30)
	DeviceName: Onboard - Other
	Subsystem: Dell Cannon Point-LP SMBus Controller [1028:08c9]
	Kernel driver in use: i801_smbus
	Kernel modules: i2c_i801
00:1f.5 Serial bus controller [0c80]: Intel Corporation Cannon Point-LP SPI Controller [8086:9da4] (rev 30)
	DeviceName: Onboard - Other
	Subsystem: Dell Cannon Point-LP SPI Controller [1028:08c9]
01:00.0 Display controller [0380]: Advanced Micro Devices, Inc. [AMD/ATI] Jet PRO [Radeon R5 M230 / R7 M260DX / Radeon 520/610 Mobile] [1002:6665] (rev c3)
	Subsystem: Dell Jet PRO [Radeon R5 M230 / R7 M260DX / Radeon 520/610 Mobile] [1028:08c9]
	Kernel driver in use: radeon
	Kernel modules: radeon, amdgpu
02:00.0 Ethernet controller [0200]: Realtek Semiconductor Co., Ltd. RTL810xE PCI Express Fast Ethernet controller [10ec:8136] (rev 07)
	Subsystem: Dell RTL810xE PCI Express Fast Ethernet controller [1028:08c9]
	Kernel driver in use: r8169
	Kernel modules: r8169
03:00.0 Network controller [0280]: Qualcomm Atheros QCA9377 802.11ac Wireless Network Adapter [168c:0042] (rev 31)
	Subsystem: Dell QCA9377 802.11ac Wireless Network Adapter [1028:1810]
	Kernel driver in use: ath10k_pci
	Kernel modules: ath10k_pci

Post by **NeddySeagoon** » Sun Sep 03, 2023 2:25 pm

4nn13,

Code: Select all

01:00.0 Display controller [0380]: Advanced Micro Devices, Inc. [AMD/ATI] Jet PRO [Radeon R5 M230 / R7 M260DX / Radeon 520/610 Mobile] [1002:6665] (rev c3)
   Subsystem: Dell Jet PRO [Radeon R5 M230 / R7 M260DX / Radeon 520/610 Mobile] [1028:08c9]
   Kernel driver in use: radeon
   Kernel modules: radeon, amdgpu

The AMDGPU driver is not in use. The PCI ID 1002:6665 says that its possible to use amdgpu.
The R7 M260 is a HAINAN chipset. That's Southern Islands, support is optional. See the AMDGPU wiki page.

That does not mean that a GCN1.0+ GPU will work with AMDGPU Rocm

4nn13 · Post by **4nn13** » Sun Sep 03, 2023 3:16 pm

NeddySeagoon wrote:4nn13,
Code: Select all
01:00.0 Display controller [0380]: Advanced Micro Devices, Inc. [AMD/ATI] Jet PRO [Radeon R5 M230 / R7 M260DX / Radeon 520/610 Mobile] [1002:6665] (rev c3)
   Subsystem: Dell Jet PRO [Radeon R5 M230 / R7 M260DX / Radeon 520/610 Mobile] [1028:08c9]
   Kernel driver in use: radeon
   Kernel modules: radeon, amdgpu
The AMDGPU driver is not in use. The PCI ID 1002:6665 says that its possible to use amdgpu.
The R7 M260 is a HAINAN chipset. That's Southern Islands, support is optional. See the AMDGPU wiki page.

That does not mean that a GCN1.0+ GPU will work with AMDGPU Rocm

I've actually read the AMDGPU page many time and I have no a clue what should I do next, so can you tell me more details?

Post by **NeddySeagoon** » Sun Sep 03, 2023 3:44 pm

4nn13,

Rebuild your kernel removing DRM radon support.

It its place add back AMDGPU support (it looks like that is there already) including the optional Southern Islands support for your HAINAN chipset.
That's

Code: Select all

Enable amdgpu support for SI parts

kernel option on the wiki
Your GPU is a Southern Islands part and amdgpu support is both optional and experimental.
That means you need to ask for it and if it breaks, you can keep all the pieces.

Set your VIDEO_CARDS="amdgpu radeonsi" and rebuild mesa and the Xorg drivers.
That should install the amdgpu Xorg driver and add 3D support for it to mesa.

Do a --depclean too. The ati/radeon video driver should be removed.

When you reboot, lspci must not list

Code: Select all

Kernel driver in use: radeon

for your card.
amdgpu is what you want to see.

Radeon is the recommended driver for you but amdgpu can work too,

4nn13 · Post by **4nn13** » Sun Sep 03, 2023 3:59 pm

NeddySeagoon wrote:4nn13,

Rebuild your kernel removing DRM radon support.

It its place add back AMDGPU support (it looks like that is there already) including the optional Southern Islands support for your HAINAN chipset.
That's
Code: Select all
Enable amdgpu support for SI parts
kernel option on the wiki
Your GPU is a Southern Islands part and amdgpu support is both optional and experimental.
That means you need to ask for it and if it breaks, you can keep all the pieces.

Set your VIDEO_CARDS="amdgpu radeonsi" and rebuild mesa and the Xorg drivers.
That should install the amdgpu Xorg driver and add 3D support for it to mesa.

Do a --depclean too. The ati/radeon video driver should be removed.

When you reboot, lspci must not list
Code: Select all
Kernel driver in use: radeon
for your card.
amdgpu is what you want to see.

Radeon is the recommended driver for you but amdgpu can work too,

The SI its already enabled, and there is no option called DRM radon
The options what i found related to DRM are
1, DRM DP AUX Interface
2. DRM support for bochs .......
3, DRM for solomon

another question is should i remove the extra_firmware option from .config? do I need to add another .bin file(s)?
I'm going to remove ati radeon now

Post by **NeddySeagoon** » Sun Sep 03, 2023 4:26 pm

4nn13,

Code: Select all

  │ │    < > ATI Radeon                                                   │ │  
  │ │    <M> AMD GPU                                                      │ │  
  │ │    [*]   Enable amdgpu support for SI parts                         │ │

The kernel symbol is CONFIG_DRM_RADEON.
The menu text is ATI Radeon. Turn that off.

If AMD GPU is made as a module, it will load its firmware from /lib/firmware.

If AMD GPU is made as built in, the firmware must be included too. That's all of

Code: Select all

$ ls /lib/firmware/amdgpu/hain*
/lib/firmware/amdgpu/hainan_ce.bin     /lib/firmware/amdgpu/hainan_pfp.bin
/lib/firmware/amdgpu/hainan_k_smc.bin  /lib/firmware/amdgpu/hainan_rlc.bin
/lib/firmware/amdgpu/hainan_mc.bin     /lib/firmware/amdgpu/hainan_smc.bin
/lib/firmware/amdgpu/hainan_me.bin

4nn13 · Post by **4nn13** » Sun Sep 03, 2023 4:34 pm

NeddySeagoon wrote:4nn13,
Code: Select all
  │ │    < > ATI Radeon                                                   │ │  
  │ │    <M> AMD GPU                                                      │ │  
  │ │    [*]   Enable amdgpu support for SI parts                         │ │  
The kernel symbol is CONFIG_DRM_RADEON.
The menu text is ATI Radeon. Turn that off.

If AMD GPU is made as a module, it will load its firmware from /lib/firmware.

If AMD GPU is made as built in, the firmware must be included too. That's all of
Code: Select all
$ ls /lib/firmware/amdgpu/hain*
/lib/firmware/amdgpu/hainan_ce.bin     /lib/firmware/amdgpu/hainan_pfp.bin
/lib/firmware/amdgpu/hainan_k_smc.bin  /lib/firmware/amdgpu/hainan_rlc.bin
/lib/firmware/amdgpu/hainan_mc.bin     /lib/firmware/amdgpu/hainan_smc.bin
/lib/firmware/amdgpu/hainan_me.bin

Okay I did what you said and it will talke some time to re-compile the kernel.

4nn13 · Post by **4nn13** » Sun Sep 03, 2023 6:41 pm

4nn13 wrote:
NeddySeagoon wrote:4nn13,
Code: Select all
  │ │    < > ATI Radeon                                                   │ │  
  │ │    <M> AMD GPU                                                      │ │  
  │ │    [*]   Enable amdgpu support for SI parts                         │ │  
The kernel symbol is CONFIG_DRM_RADEON.
The menu text is ATI Radeon. Turn that off.

If AMD GPU is made as a module, it will load its firmware from /lib/firmware.

If AMD GPU is made as built in, the firmware must be included too. That's all of
Code: Select all
$ ls /lib/firmware/amdgpu/hain*
/lib/firmware/amdgpu/hainan_ce.bin     /lib/firmware/amdgpu/hainan_pfp.bin
/lib/firmware/amdgpu/hainan_k_smc.bin  /lib/firmware/amdgpu/hainan_rlc.bin
/lib/firmware/amdgpu/hainan_mc.bin     /lib/firmware/amdgpu/hainan_smc.bin
/lib/firmware/amdgpu/hainan_me.bin
Okay I did what you said and it will talke some time to re-compile the kernel.

I re-compiled kernel mesa x11-driver and xserver now I can see KDF its showed-up but rocm-smi has an error

Code: Select all

$ ~ dmesg |grep KDF
[    0.627969] alg: self-tests for CTR-KDF (hmac(sha256)) passed
$ ~ dmesg |grep hainan
[   10.485312] [drm] Loading hainan Microcode
[   10.485315] Loading firmware: radeon/hainan_pfp.bin
[   10.699570] Loading firmware: radeon/hainan_me.bin
[   11.106299] Loading firmware: radeon/hainan_ce.bin
[   11.127633] Loading firmware: radeon/hainan_rlc.bin
$ ~
$ ~ rocm-smi 
======================= ROCm System Management Interface =======================
WARNING: No AMD GPUs specified
================================= Concise Info =================================
GPU  Temp (DieEdge)  AvgPwr  SCLK  MCLK  Fan  Perf  PwrCap  VRAM%  GPU%  
================================================================================
============================= End of ROCm SMI Log ==============================[
$ ~ clinfo 
Number of platforms                               2
  Platform Name                                   AMD Accelerated Parallel Processing
  Platform Vendor                                 Advanced Micro Devices, Inc.
  Platform Version                                OpenCL 2.1 AMD-APP.dbg (3513.0)
  Platform Profile                                FULL_PROFILE
  Platform Extensions                             cl_khr_icd cl_amd_event_callback 
  Platform Extensions function suffix             AMD
  Platform Host timer resolution                  1ns

  Platform Name                                   rusticl
  Platform Vendor                                 Mesa/X.org
  Platform Version                                OpenCL 3.0 
  Platform Profile                                FULL_PROFILE
  Platform Extensions                             cl_khr_icd cl_khr_il_program
  Platform Extensions with Version                cl_khr_icd                                                       0x400000 (1.0.0)
                                                  cl_khr_il_program                                                0x400000 (1.0.0)
  Platform Numeric Version                        0xc00000 (3.0.0)
  Platform Extensions function suffix             MESA
  Platform Host timer resolution                  0ns

  Platform Name                                   AMD Accelerated Parallel Processing
Number of devices                                 0

  Platform Name                                   rusticl
Number of devices                                 0

NULL platform behavior
  clGetPlatformInfo(NULL, CL_PLATFORM_NAME, ...)  No platform
  clGetDeviceIDs(NULL, CL_DEVICE_TYPE_ALL, ...)   No platform
  clCreateContext(NULL, ...) [default]            No platform
  clCreateContext(NULL, ...) [other]              <error: no devices in non-default plaforms>
  clCreateContextFromType(NULL, CL_DEVICE_TYPE_DEFAULT)  No devices found in platform
  clCreateContextFromType(NULL, CL_DEVICE_TYPE_CPU)  No devices found in platform
  clCreateContextFromType(NULL, CL_DEVICE_TYPE_GPU)  No devices found in platform
  clCreateContextFromType(NULL, CL_DEVICE_TYPE_ACCELERATOR)  No devices found in platform
  clCreateContextFromType(NULL, CL_DEVICE_TYPE_CUSTOM)  No devices found in platform
  clCreateContextFromType(NULL, CL_DEVICE_TYPE_ALL)  No devices found in platform

ICD loader properties
  ICD loader Name                                 Khronos OpenCL ICD Loaderns
  ICD loader Vendor                               Khronos Groupns
  ICD loader Version                              3.0.6ns
  ICD loader Profile                              OpenCL 3.0ns

Post by **NeddySeagoon** » Sun Sep 03, 2023 6:59 pm

4nn13,

What does

Code: Select all

lspci -nnk

say about the kernel dirvers for your video card now?

Loading firmware: radeon/hainan_pfp.bin looks wrong. It still says radeon/

I was expecting to see amdgpu/ in the firmware path.

gorg86 · Post by **gorg86** » Sun Sep 03, 2023 8:54 pm

Your GPU is not supported by ROCm even if PCI-E atomics worked (which for some reason is not available on your PC even though your platform should support it).

4nn13 · Post by **4nn13** » Sun Sep 03, 2023 11:06 pm

NeddySeagoon wrote:4nn13,

What does
Code: Select all
lspci -nnk
say about the kernel dirvers for your video card now?

Loading firmware: radeon/hainan_pfp.bin looks wrong. It still says radeon/

I was expecting to see amdgpu/ in the firmware path.

Here is the output :

Code: Select all

00:00.0 Host bridge [0600]: Intel Corporation Coffee Lake HOST and DRAM Controller [8086:3e34] (rev 0b)
	DeviceName: Onboard - Other
	Subsystem: Dell Coffee Lake HOST and DRAM Controller [1028:08c9]
	Kernel driver in use: skl_uncore
00:02.0 VGA compatible controller [0300]: Intel Corporation WhiskeyLake-U GT2 [UHD Graphics 620] [8086:3ea0]
	DeviceName: Onboard - Video
	Subsystem: Dell WhiskeyLake-U GT2 [UHD Graphics 620] [1028:08c9]
	Kernel driver in use: i915
	Kernel modules: i915
00:04.0 Signal processing controller [1180]: Intel Corporation Xeon E3-1200 v5/E3-1500 v5/6th Gen Core Processor Thermal Subsystem [8086:1903] (rev 0b)
	DeviceName: Onboard - Other
	Subsystem: Dell Xeon E3-1200 v5/E3-1500 v5/6th Gen Core Processor Thermal Subsystem [1028:08c9]
	Kernel driver in use: proc_thermal
	Kernel modules: processor_thermal_device_pci_legacy
00:08.0 System peripheral [0880]: Intel Corporation Xeon E3-1200 v5/v6 / E3-1500 v5 / 6th/7th/8th Gen Core Processor Gaussian Mixture Model [8086:1911]
	DeviceName: Onboard - Other
	Subsystem: Dell Xeon E3-1200 v5/v6 / E3-1500 v5 / 6th/7th/8th Gen Core Processor Gaussian Mixture Model [1028:08c9]
00:12.0 Signal processing controller [1180]: Intel Corporation Cannon Point-LP Thermal Controller [8086:9df9] (rev 30)
	DeviceName: Onboard - Other
	Subsystem: Dell Cannon Point-LP Thermal Controller [1028:08c9]
	Kernel driver in use: intel_pch_thermal
	Kernel modules: intel_pch_thermal
00:14.0 USB controller [0c03]: Intel Corporation Cannon Point-LP USB 3.1 xHCI Controller [8086:9ded] (rev 30)
	DeviceName: Onboard - Other
	Subsystem: Dell Cannon Point-LP USB 3.1 xHCI Controller [1028:08c9]
	Kernel driver in use: xhci_hcd
00:14.2 RAM memory [0500]: Intel Corporation Cannon Point-LP Shared SRAM [8086:9def] (rev 30)
	DeviceName: Onboard - Other
	Subsystem: Dell Cannon Point-LP Shared SRAM [1028:08c9]
00:15.0 Serial bus controller [0c80]: Intel Corporation Cannon Point-LP Serial IO I2C Controller #0 [8086:9de8] (rev 30)
	DeviceName: Onboard - Other
	Subsystem: Dell Cannon Point-LP Serial IO I2C Controller [1028:08c9]
	Kernel driver in use: intel-lpss
00:15.1 Serial bus controller [0c80]: Intel Corporation Cannon Point-LP Serial IO I2C Controller #1 [8086:9de9] (rev 30)
	DeviceName: Onboard - Other
	Subsystem: Dell Cannon Point-LP Serial IO I2C Controller [1028:08c9]
	Kernel driver in use: intel-lpss
00:17.0 RAID bus controller [0104]: Intel Corporation 82801 Mobile SATA Controller [RAID mode] [8086:282a] (rev 30)
	DeviceName: Onboard - Other
	Subsystem: Dell 82801 Mobile SATA Controller [RAID mode] [1028:08c9]
	Kernel driver in use: ahci
00:19.0 Serial bus controller [0c80]: Intel Corporation Cannon Point-LP Serial IO I2C Host Controller [8086:9dc5] (rev 30)
	DeviceName: Onboard - Other
	Subsystem: Dell Cannon Point-LP Serial IO I2C Host Controller [1028:08c9]
	Kernel driver in use: intel-lpss
00:1c.0 PCI bridge [0604]: Intel Corporation Cannon Point-LP PCI Express Root Port #5 [8086:9dbc] (rev f0)
	Subsystem: Dell Cannon Point-LP PCI Express Root Port [1028:08c9]
	Kernel driver in use: pcieport
00:1d.0 PCI bridge [0604]: Intel Corporation Cannon Point-LP PCI Express Root Port #9 [8086:9db0] (rev f0)
	Subsystem: Dell Cannon Point-LP PCI Express Root Port [1028:08c9]
	Kernel driver in use: pcieport
00:1d.1 PCI bridge [0604]: Intel Corporation Cannon Point-LP PCI Express Root Port #10 [8086:9db1] (rev f0)
	Subsystem: Dell Cannon Point-LP PCI Express Root Port [1028:08c9]
	Kernel driver in use: pcieport
00:1f.0 ISA bridge [0601]: Intel Corporation Cannon Point-LP LPC Controller [8086:9d84] (rev 30)
	DeviceName: Onboard - Other
	Subsystem: Dell Cannon Point-LP LPC Controller [1028:08c9]
00:1f.3 Audio device [0403]: Intel Corporation Cannon Point-LP High Definition Audio Controller [8086:9dc8] (rev 30)
	DeviceName: Onboard - Sound
	Subsystem: Dell Cannon Point-LP High Definition Audio Controller [1028:08c9]
	Kernel driver in use: snd_hda_intel
	Kernel modules: snd_hda_intel, snd_soc_skl, snd_sof_pci_intel_cnl
00:1f.4 SMBus [0c05]: Intel Corporation Cannon Point-LP SMBus Controller [8086:9da3] (rev 30)
	DeviceName: Onboard - Other
	Subsystem: Dell Cannon Point-LP SMBus Controller [1028:08c9]
	Kernel driver in use: i801_smbus
	Kernel modules: i2c_i801
00:1f.5 Serial bus controller [0c80]: Intel Corporation Cannon Point-LP SPI Controller [8086:9da4] (rev 30)
	DeviceName: Onboard - Other
	Subsystem: Dell Cannon Point-LP SPI Controller [1028:08c9]
01:00.0 Display controller [0380]: Advanced Micro Devices, Inc. [AMD/ATI] Jet PRO [Radeon R5 M230 / R7 M260DX / Radeon 520/610 Mobile] [1002:6665] (rev c3)
	Subsystem: Dell Jet PRO [Radeon R5 M230 / R7 M260DX / Radeon 520/610 Mobile] [1028:08c9]
	Kernel driver in use: radeon
	Kernel modules: radeon, amdgpu
02:00.0 Ethernet controller [0200]: Realtek Semiconductor Co., Ltd. RTL810xE PCI Express Fast Ethernet controller [10ec:8136] (rev 07)
	Subsystem: Dell RTL810xE PCI Express Fast Ethernet controller [1028:08c9]
	Kernel driver in use: r8169
	Kernel modules: r8169
03:00.0 Network controller [0280]: Qualcomm Atheros QCA9377 802.11ac Wireless Network Adapter [168c:0042] (rev 31)
	Subsystem: Dell QCA9377 802.11ac Wireless Network Adapter [1028:1810]
	Kernel driver in use: ath10k_pci
	Kernel modules: ath10k_pci

grknight · Post by **grknight** » Sun Sep 03, 2023 11:10 pm

radeon.si_support=0 amdgpu.si_support=1 as kernel command line arguments will force AMDGPU for any Southern Islands cards.

This is needed when both radeon and amdgpu are enabled in the kernel.

Goverp · Post by **Goverp** » Sun Sep 03, 2023 11:32 pm

gorg86 wrote:Your GPU is not supported by ROCm even if PCI-E atomics worked (which for some reason is not available on your PC even though your platform should support it).

Probably irrelevant, but I had no PCI-E atomics 'cos my video card was in the wrong slot (two available, and the right one blocked the airflow to a motherboard fan), but this was on much later hardware.

4nn13 · Post by **4nn13** » Sun Sep 03, 2023 11:36 pm

Goverp wrote:
gorg86 wrote:Your GPU is not supported by ROCm even if PCI-E atomics worked (which for some reason is not available on your PC even though your platform should support it).
Probably irrelevant, but I had no PCI-E atomics 'cos my video card was in the wrong slot (two available, and the right one blocked the airflow to a motherboard fan), but this was on much later hardware.

It's same like i forgot to modifie /etc/portage/make.conf..... also need to re-compile X and mesa again.......

4nn13 · Post by **4nn13** » Sun Sep 03, 2023 11:47 pm

Goverp wrote:
gorg86 wrote:Your GPU is not supported by ROCm even if PCI-E atomics worked (which for some reason is not available on your PC even though your platform should support it).
Probably irrelevant, but I had no PCI-E atomics 'cos my video card was in the wrong slot (two available, and the right one blocked the airflow to a motherboard fan), but this was on much later hardware.

I modified /etc/portage/make.conf (VIDEO_CARDS="amdgpu radeonsi intel" )and then rebuild mesa, xorg-drivers&server

and here is the output

Code: Select all

01:00.0 Display controller [0380]: Advanced Micro Devices, Inc. [AMD/ATI] Jet PRO [Radeon R5 M230 / R7 M260DX / Radeon 520/610 Mobile] [1002:6665] (rev c3)
	Subsystem: Dell Jet PRO [Radeon R5 M230 / R7 M260DX / Radeon 520/610 Mobile] [1028:08c9]
	Kernel driver in use: radeon
	Kernel modules: radeon, amdgpu

Code: Select all

$ ~ rocm-smi 


======================= ROCm System Management Interface =======================
WARNING: No AMD GPUs specified
================================= Concise Info =================================
GPU  Temp (DieEdge)  AvgPwr  SCLK  MCLK  Fan  Perf  PwrCap  VRAM%  GPU%  
================================================================================
============================= End of ROCm SMI Log ==============================

gorg86 · Post by **gorg86** » Mon Sep 04, 2023 1:35 am

rocm-smi will always output nothing, because it only lists supported devices. I don't know what you are trying to do, but even if ROCm worked, you won't be able to do any useful computing with it anyway with 2GB VRAM.

4nn13 · Post by **4nn13** » Mon Sep 04, 2023 1:39 am

gorg86 wrote:rocm-smi will always output nothing, because it only lists supported devices. I don't know what you are trying to do, but even if ROCm worked, you won't be able to do any useful computing with it anyway with 2GB VRAM.

I'm trying to make hashcat works

gorg86 · Post by **gorg86** » Mon Sep 04, 2023 2:09 am

https://en.wikipedia.org/wiki/ROCm#Hardware_support

You need a Polaris GPU as a minimum, even though AMD officially dropped Polaris support it can work if your platform supports atomic transactions and if you compiled rocBLAS with legacy support (AMDGPU_TARGETS="gfx803").
People managed to get ROCm to work on 390 cards (without atomic transactions), but it's dodgy.
A NAVI GPU works with and without atomics.

Here are some links about the confusing atomics situation, this was clear as mud for a long time:
https://github.com/RadeonOpenCompute/RO ... orted-cpus
https://github.com/RadeonOpenCompute/RO ... -422836032

Hashcat can utilize opencl afaik,
Try

Code: Select all

hashcat -I

to list devices and

Code: Select all

hashcat -b

to see if it works.

If it does not work you'll have to compile mesa without opencl and install dev-libs/amdgpu-pro-opencl

4nn13 · Post by **4nn13** » Mon Sep 04, 2023 2:37 am

gorg86 wrote:https://en.wikipedia.org/wiki/ROCm#Hardware_support

You need a Polaris GPU as a minimum, even though AMD officially dropped Polaris support it can work if your platform supports atomic transactions and if you compiled rocBLAS with legacy support (AMDGPU_TARGETS="gfx803").
People managed to get ROCm to work on 390 cards (without atomic transactions), but it's dodgy.
A NAVI GPU works with and without atomics.

Here are some links about the confusing atomics situation, this was clear as mud for a long time:
https://github.com/RadeonOpenCompute/RO ... orted-cpus
https://github.com/RadeonOpenCompute/RO ... -422836032

Hashcat can utilize opencl afaik,
Try
Code: Select all
hashcat -I
to list devices and
Code: Select all
hashcat -b
to see if it works.

If it does not work you'll have to compile mesa without opencl and install dev-libs/amdgpu-pro-opencl

Code: Select all

$ ~ hashcat -I
hashcat (v6.2.6) starting in backend information mode

No devices found/left.

$ ~

looks like i need to install dev-libs/amdgpu-pro-opencl

gorg86 · Post by **gorg86** » Mon Sep 04, 2023 2:57 am

Don't forget to do what grknight said, that's important. Check dmesg if amdgpu is actually used.

4nn13 · Post by **4nn13** » Mon Sep 04, 2023 5:34 am

gorg86 wrote:Don't forget to do what grknight said, that's important. Check dmesg if amdgpu is actually used.

There is an error

Code: Select all

$ ~ sudo emerge -av dev-libs/amdgpu-pro-opencl

These are the packages that would be merged, in order:

Calculating dependencies               ... done!
Dependency resolution took 0.88 s.

[ebuild  N F  ~] dev-libs/amdgpu-pro-opencl-20.40.1147286::gentoo  ABI_X86="(64) -32" 325,207 KiB

Total: 1 package (1 new), Size of downloads: 325,207 KiB
Fetch Restriction: 1 package (1 unsatisfied)

Fetch instructions for dev-libs/amdgpu-pro-opencl-20.40.1147286:
 * Please download Radeon Software for Linux version 20.40 for Ubuntu 20.04.1 from
 *     https://www.amd.com/en/support/kb/release-notes/rn-amdgpu-unified-linux-20-40
 * The archive should then be placed into your distfiles directory.

Would you like to merge these packages? [Yes/No] 
>>> Verifying ebuild manifests
>>> Emerging (1 of 1) dev-libs/amdgpu-pro-opencl-20.40.1147286::gentoo
>>> Failed to emerge dev-libs/amdgpu-pro-opencl-20.40.1147286, Log file:
>>>  '/var/tmp/portage/dev-libs/amdgpu-pro-opencl-20.40.1147286/temp/build.log'
>>> Jobs: 0 of 1 complete, 1 failed                 Load avg: 0.13, 0.38, 0.35

!!! dev-libs/amdgpu-pro-opencl-20.40.1147286 has fetch restriction turned on.
!!! This probably means that this ebuild's files must be downloaded
!!! manually.  See the comments in the ebuild for more information.

 * Fetch failed for 'dev-libs/amdgpu-pro-opencl-20.40.1147286', Log file:
 *  '/var/tmp/portage/dev-libs/amdgpu-pro-opencl-20.40.1147286/temp/build.log'
 * Package:    dev-libs/amdgpu-pro-opencl-20.40.1147286:0
 * Repository: gentoo
 * USE:        abi_x86_64 amd64 elibc_glibc kernel_linux
 * FEATURES:   network-sandbox preserve-libs sandbox userpriv usersandbox
 * Please download Radeon Software for Linux version 20.40 for Ubuntu 20.04.1 from
 *     https://www.amd.com/en/support/kb/release-notes/rn-amdgpu-unified-linux-20-40
 * The archive should then be placed into your distfiles directory.

 * Messages for package dev-libs/amdgpu-pro-opencl-20.40.1147286:

 * Fetch failed for 'dev-libs/amdgpu-pro-opencl-20.40.1147286', Log file:
 *  '/var/tmp/portage/dev-libs/amdgpu-pro-opencl-20.40.1147286/temp/build.log'
$ ~ sudo cat /var/tmp/portage/dev-libs/amdgpu-pro-opencl-20.40.1147286/temp/build.log

!!! dev-libs/amdgpu-pro-opencl-20.40.1147286 has fetch restriction turned on.
!!! This probably means that this ebuild's files must be downloaded
!!! manually.  See the comments in the ebuild for more information.

 * Fetch failed for 'dev-libs/amdgpu-pro-opencl-20.40.1147286', Log file:
 *  '/var/tmp/portage/dev-libs/amdgpu-pro-opencl-20.40.1147286/temp/build.log'
 * Package:    dev-libs/amdgpu-pro-opencl-20.40.1147286:0
 * Repository: gentoo
 * USE:        abi_x86_64 amd64 elibc_glibc kernel_linux
 * FEATURES:   network-sandbox preserve-libs sandbox userpriv usersandbox
 * Please download Radeon Software for Linux version 20.40 for Ubuntu 20.04.1 from
 *     https://www.amd.com/en/support/kb/release-notes/rn-amdgpu-unified-linux-20-40
 * The archive should then be placed into your distfiles directory.
$ ~

gorg86 · Post by **gorg86** » Mon Sep 04, 2023 6:14 am

you have to download the package from https://www.amd.com/en/support/kb/relea ... inux-20-40 and place it in distfiles like instructed in the error message.

4nn13 · Post by **4nn13** » Mon Sep 04, 2023 6:37 am

gorg86 wrote:you have to download the package from https://www.amd.com/en/support/kb/relea ... inux-20-40 and place it in distfiles like instructed in the error message.

Which package should I download? there are only the driver ubuntu and redhat