[SOLVED] Slackware64 current (15.0) OpenCL added to stock Mesa amdgpu - call for testing

kingbeowulf · 02-23-2021, 07:46 PM

Just for giggles I fired up PhoenixMiner and it seems it found the correct RX5700XT OpenCL platform (Slackware64-current mulilib kernel 5.10.17, default settings, with just -amd -acm) Although, at 50 MH/s I won't be expecting much "profit" ;-)

So I'd say there might be something screwy on your end.

dchmelik · 02-23-2021, 08:25 PM

Quote:

Originally Posted by kingbeowulf

What happens when you move '/etc/OpenCL/vendors/mesa.icd' somewhere else?

When I did that, nsgminer started working and the pool accepted a share. PhoenixMiner is taking longer to start, but it might be that the pool I'm on tests your GPU for a while first. Could I just rename it mesa.icd.bak or would it still be detected?

I'm wondering if by Slackware 15 we won't have to make this minor tweak anymore. I'd never have figured it out without expert help...

kingbeowulf · 02-23-2021, 09:05 PM

Quote:

Originally Posted by dchmelik

... Could I just rename it mesa.icd.bak or would it still be detected?

I'm wondering if by Slackware 15 we won't have to make this minor tweak anymore. I'd never have figured it out without expert help...

Renaming "might" work...not sure. Don't forget too that mesa.icd will be replaced on each Mesa upgrade. There should be a setting in these miners to pick the OpenCL platform.

heh "expert". much appreciated, though befuddled. I'm just a hobbyist that like to tinker!

jedrek.b · 08-03-2021, 09:42 AM

Hi, it looks like AMD ROCm starting from version 3.7 has open-source OpenCL image module included. I have used rocm few years ago and this post help me with install https://www.linuxquestions.org/quest...re-4175638428/ but that was 3 years ago. A lot has change since then and I remember it was pain to get romc opencl running. There was bug in rocm image support back then so I switched to binary opencl package created by kingbeowulf.

Did anyone tried latest ROCm ? I think it would be helpful to have slackbuild script to build ROCm from source, but this is beyond my skills.

What do you think ? I would like to hear your opinions.

kingbeowulf · 08-15-2021, 01:10 PM

I'm looking into it, as an option to replace the repackaging of the OpenCL bits from amdgpu-pro. Supposedly ROCm finally added Navi support (?) which was holding me back. There are some other limitations and requirements as well that make ROCm unsuitable for a GUI desktop. I just wish they would just release a simple OpenCL library and runtime without all the ROCm API and SDK complexity.

dchmelik · 10-19-2021, 11:56 PM

Renaming/moving that configuration file didn't work to get PhoneixMiner nor ethminer working for ethereum--they started but just never continued. I still was able to mine with ryominer (for ryo), nsgminer (for feathercoin), but can't also test lolMiner (for bitcoin gold) yet. I am unable to try ROCm but it'd be preferable, except I use desktops. AMD's drivers had several more versions since this thread continued, up to at least version 21.30. Will we see builds for these?

I think it's important to keep all versions of the builds around. Some older drivers work better, but if you have too new a kernel, and were using some OS it installs DKMS, you will no longer have a display (had this problem on KDE Neon GNU/Linux family PCs). A lot of the newer drivers are shoddier and all the miners just exit (may be a too low-Watt power supply unit (PSU) issue) but if you have a new enough kernel, some of them end up being okay, but if you do something as simple as upgrade a card, miners may start exiting again. It just takes trial & error to determine which will work at this point. Any particular driver won't necessarily work for everyone's situations.

denydias · 12-19-2021, 03:35 PM

Quote:

Originally Posted by kingbeowulf

Using amdgpu-pro-20.40-1147286-ubuntu-20.04.tar.xz, I was able to get an updated OpenCL slackbuild working on Slackware64-current (15.0, Sun Nov 15 00:02:28 UTC 2020).

I've just built and installed the resulting package. Works like a charm using AMD OpenCL 20.40-1147286 in Slackware64-current Sun Dec 19 18:57:11 UTC 2021.

But AMD announced this on Radeon™ Software for Linux® 21.40.1 Release Notes (dated 11/10/21):

Quote:

Introducing support for RHEL 8.5.
Release 21.40.1 is the first Radeon™ Software for Linux that uses unified ROCm™ and graphics drivers. This release has not been fully validated for Machine Learning use cases. Users are recommended to use https://rocmdocs.amd.com/en/latest/C...ase-Notes.html for ROCm™ use cases.
This release introduces a new driver distribution model that is based on public package repository instead of tarball. The installer is posted on amd.com.
The amdgpu-install and amdgpu-uninstall scripts have been modified to support the above unified stack and new distribution model. Additional information on amdgpu-install and amdgpu-uninstall can be found in the installation documentation at https://amdgpu-install.readthedocs.io/en/21.40/ .
The installer streamlines the installation process by:
- Abstracting the distribution-specific package installation logic,
- Performing the repository set-up,
- Allowing users to specify the use case and automating the installation of all the required packages,
- Downloading only packages that are needed for the chosen use case and reducing download size, and
- Performing post-install checks to verify whether the installation was performed successfully.
EULA acceptance for proprietary components has been moved from amd.com to the installer.

So I ask to @kingbeowulf: do you plan to upgrade your amdgpu-opencl.SlackBuild to use the new structure?

I've tried to do it by following AUR opencl-amd model but failed miserably.

EDIT: here's OpenCL working on my box with the now outdated 20.40-1147286...

Code:

$ neofetch
                  :::::::                      deny@minsky.local 
            :::::::::::::::::::                ----------------- 
         :::::::::::::::::::::::::             OS: Slackware 15.0 x86_64 x86_64 
       ::::::::cllcccccllllllll::::::          Host: LENOVO LNVNB161216 
    :::::::::lc               dc:::::::        Kernel: 5.15.10 
   ::::::::cl   clllccllll    oc:::::::::      Uptime: 1 hour, 52 mins 
  :::::::::o   lc::::::::co   oc::::::::::     Packages: 1591 (pkgtool) 
 ::::::::::o    cccclc:::::clcc::::::::::::    Shell: bash 5.1.12 
 :::::::::::lc        cclccclc:::::::::::::    Resolution: 1920x1080, 1920x1080 
::::::::::::::lcclcc          lc::::::::::::   DE: Plasma 5.23.4 (Wayland) 
::::::::::cclcc:::::lccclc     oc:::::::::::   Theme: Breeze Light [Plasma], Breeze [GTK3] 
::::::::::o    l::::::::::l    lc:::::::::::   Icons: [Plasma], breeze-dark [GTK2/3] 
 :::::cll:o     clcllcccll     o:::::::::::    Terminal: yakuake 
 :::::occ:o                  clc:::::::::::    Terminal Font: Perplexed 12 
  ::::ocl:ccslclccclclccclclc:::::::::::::     CPU: AMD Ryzen 7 5700U with Radeon Graphics (16) @ 1.800GHz 
   :::oclcccccccccccccllllllllllllll:::::      GPU: AMD ATI Lucienne 
    ::lcc1lcccccccccccccccccccccccco::::       Memory: 3270MiB / 17925MiB 
      ::::::::::::::::::::::::::::::::
        ::::::::::::::::::::::::::::                                   
           ::::::::::::::::::::::                                      
                ::::::::::::

$ inxi -G
Graphics:
  Device-1: AMD Lucienne driver: amdgpu v: kernel
  Display: wayland server: Slackware Linux Project 1.20.14 driver:
    loaded: amdgpu,ati unloaded: modesetting,vesa resolution: 1: 1920x1080~60Hz
    2: 1920x1080~60Hz
  OpenGL: renderer: AMD RENOIR (DRM 3.42.0 5.15.10 LLVM 13.0.0)
    v: 4.6 Mesa 21.2.6

$ clinfo 
Number of platforms:                             2
  Platform Profile:                              FULL_PROFILE
  Platform Version:                              OpenCL 2.1 AMD-APP (3180.7)
  Platform Name:                                 AMD Accelerated Parallel Processing
  Platform Vendor:                               Advanced Micro Devices, Inc.
  Platform Extensions:                           cl_khr_icd cl_amd_event_callback cl_amd_offline_devices 
  Platform Profile:                              FULL_PROFILE
  Platform Version:                              OpenCL 1.1 Mesa 21.2.6
  Platform Name:                                 Clover
  Platform Vendor:                               Mesa
  Platform Extensions:                           cl_khr_icd


  Platform Name:                                 AMD Accelerated Parallel Processing
Number of devices:                               1
  Device Type:                                   CL_DEVICE_TYPE_GPU
  Vendor ID:                                     1002h
  Board name:                                    Unknown AMD GPU
  Device Topology:                               PCI[ B#3, D#0, F#0 ]
  Max compute units:                             28
  Max work items dimensions:                     3
    Max work items[0]:                           1024
    Max work items[1]:                           1024
    Max work items[2]:                           1024
  Max work group size:                           256
  Preferred vector width char:                   4
  Preferred vector width short:                  2
  Preferred vector width int:                    1
  Preferred vector width long:                   1
  Preferred vector width float:                  1
  Preferred vector width double:                 1
  Native vector width char:                      4
  Native vector width short:                     2
  Native vector width int:                       1
  Native vector width long:                      1
  Native vector width float:                     1
  Native vector width double:                    1
  Max clock frequency:                           1900Mhz
  Address bits:                                  64
  Max memory allocation:                         1825361100
  Image support:                                 Yes
  Max number of images read arguments:           128
  Max number of images write arguments:          64
  Max image 2D width:                            16384
  Max image 2D height:                           16384
  Max image 3D width:                            2048
  Max image 3D height:                           2048
  Max image 3D depth:                            2048
  Max samplers within kernel:                    16
  Max size of kernel argument:                   1024
  Alignment (bits) of base address:              2048
  Minimum alignment (bytes) for any datatype:    128
  Single precision floating point capability
    Denorms:                                     No
    Quiet NaNs:                                  Yes
    Round to nearest even:                       Yes
    Round to zero:                               Yes
    Round to +ve and infinity:                   Yes
    IEEE754-2008 fused multiply-add:             Yes
  Cache type:                                    Read/Write
  Cache line size:                               64
  Cache size:                                    16384
  Global memory size:                            3758096384
  Constant buffer size:                          1825361100
  Max number of constant args:                   8
  Local memory type:                             Scratchpad
  Local memory size:                             65536
  Max pipe arguments:                            16
  Max pipe active reservations:                  16
  Max pipe packet size:                          1825361100
  Max global variable size:                      1642824960
  Max global variable preferred total size:      3758096384
  Max read/write image args:                     64
  Max on device events:                          1024
  Queue on device max size:                      8388608
  Max on device queues:                          1
  Queue on device preferred size:                262144
  SVM capabilities:                              
    Coarse grain buffer:                         Yes
    Fine grain buffer:                           Yes
    Fine grain system:                           No
    Atomics:                                     No
  Preferred platform atomic alignment:           0
  Preferred global atomic alignment:             0
  Preferred local atomic alignment:              0
  Kernel Preferred work group size multiple:     64
  Error correction support:                      0
  Unified memory for Host and Device:            1
  Profiling timer resolution:                    1
  Device endianess:                              Little
  Available:                                     Yes
  Compiler available:                            Yes
  Execution capabilities:                                
    Execute OpenCL kernels:                      Yes
    Execute native function:                     No
  Queue on Host properties:                              
    Out-of-Order:                                No
    Profiling :                                  Yes
  Queue on Device properties:                            
    Out-of-Order:                                Yes
    Profiling :                                  Yes
  Platform ID:                                   0x7f3161890f30
  Name:                                          gfx902
  Vendor:                                        Advanced Micro Devices, Inc.
  Device OpenCL C version:                       OpenCL C 2.0 
  Driver version:                                3180.7 (PAL,HSAIL)
  Profile:                                       FULL_PROFILE
  Version:                                       OpenCL 2.0 AMD-APP (3180.7)
  Extensions:                                    cl_khr_fp64 cl_amd_fp64 cl_khr_global_int32_base_atomics cl_khr_global_int32_extended_atomics
cl_khr_local_int32_base_atomics cl_khr_local_int32_extended_atomics cl_khr_int64_base_atomics cl_khr_int64_extended_atomics cl_khr_3d_image_writes
cl_khr_byte_addressable_store cl_khr_fp16 cl_khr_gl_sharing cl_khr_gl_depth_images cl_amd_device_attribute_query cl_amd_vec3 cl_amd_printf cl_amd_media_ops
cl_amd_media_ops2 cl_amd_popcnt cl_khr_image2d_from_buffer cl_khr_subgroups cl_khr_gl_event cl_khr_depth_images cl_khr_mipmap_image cl_khr_mipmap_image_writes
cl_amd_copy_buffer_p2p 


  Platform Name:                                 Clover
Number of devices:                               1
  Device Type:                                   CL_DEVICE_TYPE_GPU
  Vendor ID:                                     1002h
  Max compute units:                             28
  Max work items dimensions:                     3
    Max work items[0]:                           256
    Max work items[1]:                           256
    Max work items[2]:                           256
  Max work group size:                           256
  Preferred vector width char:                   16
  Preferred vector width short:                  8
  Preferred vector width int:                    4
  Preferred vector width long:                   2
  Preferred vector width float:                  4
  Preferred vector width double:                 2
  Native vector width char:                      16
  Native vector width short:                     8
  Native vector width int:                       4
  Native vector width long:                      2
  Native vector width float:                     4
  Native vector width double:                    2
  Max clock frequency:                           1900Mhz
  Address bits:                                  64
  Max memory allocation:                         2254857830
  Image support:                                 No
  Max size of kernel argument:                   1024
  Alignment (bits) of base address:              32768
  Minimum alignment (bytes) for any datatype:    128
  Single precision floating point capability
    Denorms:                                     No
    Quiet NaNs:                                  Yes
    Round to nearest even:                       Yes
    Round to zero:                               No
    Round to +ve and infinity:                   No
    IEEE754-2008 fused multiply-add:             No
  Cache type:                                    None
  Cache line size:                               0
  Cache size:                                    0
  Global memory size:                            3221225472
  Constant buffer size:                          67108864
  Max number of constant args:                   16
  Local memory type:                             Scratchpad
  Local memory size:                             32768

# cat /usr/local/sbin/mixbench 
#!/bin/sh
cd /root/src/amdgpu-opencl/mixbench/build
./mixbench-ocl-ro --help|tail -3
./mixbench-ocl-ro
./mixbench-ocl-alt

# mixbench
Available OpenCL devices:
  1. gfx902/AMD Accelerated Parallel Processing
  2. AMD RENOIR (DRM 3.42.0, 5.15.10, LLVM 13.0.0)/Clover
mixbench-ocl/read-only (v0.03-19-gf47d11e)
Use "-h" argument to see available options
------------------------ Device specifications ------------------------
Platform:            AMD Accelerated Parallel Processing
Device:              gfx902/Advanced Micro Devices, Inc.
Driver version:      3180.7 (PAL,HSAIL)
Address bits:        64
GPU clock rate:      1900 MHz
Total global mem:    3584 MB
Max allowed buffer:  1740 MB
OpenCL version:      OpenCL 2.0 AMD-APP (3180.7)
Total CUs:           28
-----------------------------------------------------------------------
Buffer size:            256MB
Workgroup size:         256
Elements per workitem:  8
Workitem fusion degree: 4
Workitem stride:        NDRange
Buffer allocation:      Device allocated
Timer:                  CL event based
Loading kernel source file...
Precompilation of kernels... [>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>]
----------------------------------------------------------------------------- CSV data -----------------------------------------------------------------------------
Experiment ID, Single Precision ops,,,,              Double precision ops,,,,              Half precision ops,,,,                Integer operations,,, 
Compute iters, Flops/byte, ex.time,  GFLOPS, GB/sec, Flops/byte, ex.time,  GFLOPS, GB/sec, Flops/byte, ex.time,  GFLOPS, GB/sec, Iops/byte, ex.time,   GIOPS, GB/sec
            0,      0.250,    4.74,    7.07,  28.30,      0.125,    7.40,    4.53,  36.28,      0.500,    3.92,   17.11,  34.22,     0.250,    5.59,    6.01,  24.03
            1,      0.750,    3.35,   30.03,  40.04,      0.375,    6.77,   14.87,  39.65,      1.500,    5.06,   39.79,  26.53,     0.750,    3.37,   29.85,  39.80
            2,      1.250,    5.48,   30.62,  24.50,      0.625,    8.17,   20.53,  32.85,      2.500,    3.56,   94.28,  37.71,     1.250,    5.13,   32.72,  26.18
            3,      1.750,    3.36,   69.95,  39.97,      0.875,    6.70,   35.04,  40.04,      3.500,    5.12,   91.76,  26.22,     1.750,    4.89,   48.05,  27.46
            4,      2.250,    4.20,   71.86,  31.94,      1.125,    6.94,   43.51,  38.67,      4.500,    4.87,  124.07,  27.57,     2.250,    5.49,   55.00,  24.44
            5,      2.750,    3.35,  110.18,  40.07,      1.375,    6.74,   54.76,  39.82,      5.500,    5.41,  136.43,  24.81,     2.750,    3.94,   93.68,  34.07
            6,      3.250,    3.36,  129.96,  39.99,      1.625,   10.82,   40.32,  24.81,      6.500,    3.46,  252.36,  38.82,     3.250,    4.09,  106.72,  32.84
            7,      3.750,    4.63,  108.82,  29.02,      1.875,    7.58,   66.41,  35.42,      7.500,    4.00,  251.49,  33.53,     3.750,    5.56,   90.45,  24.12
            8,      4.250,    3.33,  171.22,  40.29,      2.125,    6.71,   84.97,  39.99,      8.500,    5.22,  218.61,  25.72,     4.250,    3.43,  166.17,  39.10
            9,      4.750,    4.17,  152.96,  32.20,      2.375,    9.55,   66.74,  28.10,      9.500,    3.63,  351.22,  36.97,     4.750,    5.08,  125.47,  26.41
           10,      5.250,    3.34,  210.83,  40.16,      2.625,    6.83,  103.19,  39.31,     10.500,    4.99,  282.26,  26.88,     5.250,    5.40,  130.51,  24.86
           11,      5.750,    3.38,  228.56,  39.75,      2.875,    6.98,  110.50,  38.44,     11.500,    4.87,  317.20,  27.58,     5.750,    3.42,  225.91,  39.29
           12,      6.250,    5.60,  149.92,  23.99,      3.125,    9.23,   90.88,  29.08,     12.500,    3.45,  486.17,  38.89,     6.250,    5.26,  159.53,  25.53
           13,      6.750,    3.35,  270.40,  40.06,      3.375,    8.64,  104.90,  31.08,     13.500,    3.84,  472.27,  34.98,     6.750,    3.37,  268.95,  39.84
           14,      7.250,    5.56,  175.10,  24.15,      3.625,    9.96,   97.72,  26.96,     14.500,    5.70,  341.39,  23.54,     7.250,    3.37,  288.41,  39.78
           15,      7.750,    3.36,  309.99,  40.00,      3.875,   10.43,   99.74,  25.74,     15.500,    3.51,  593.38,  38.28,     7.750,    3.77,  275.81,  35.59
           16,      8.250,    4.61,  240.13,  29.11,      4.125,   10.85,  102.02,  24.73,     16.500,    5.37,  412.69,  25.01,     8.250,    3.48,  318.34,  38.59
           17,      8.750,    3.35,  350.15,  40.02,      4.375,   11.13,  105.51,  24.12,     17.500,    3.49,  672.87,  38.45,     8.750,    4.50,  260.97,  29.83
           18,      9.250,    4.04,  307.05,  33.19,      4.625,   11.87,  104.56,  22.61,     18.500,    5.72,  434.18,  23.47,     9.250,    3.75,  330.67,  35.75
           20,     10.250,    3.36,  410.04,  40.00,      5.125,   12.50,  110.06,  21.48,     20.500,    4.47,  615.78,  30.04,    10.250,    5.26,  261.48,  25.51
           22,     11.250,    3.35,  450.54,  40.05,      5.625,   13.60,  110.99,  19.73,     22.500,    3.37,  895.78,  39.81,    11.250,    6.27,  240.84,  21.41
           24,     12.250,    3.35,  490.36,  40.03,      6.125,   14.54,  113.10,  18.47,     24.500,    4.67,  703.81,  28.73,    12.250,    4.69,  350.33,  28.60
           28,     14.250,    3.37,  567.94,  39.86,      7.125,   17.05,  112.17,  15.74,     28.500,    3.79, 1008.80,  35.40,    14.250,    5.36,  356.92,  25.05
           32,     16.250,    3.36,  649.47,  39.97,      8.125,   19.68,  110.84,  13.64,     32.500,    3.39, 1286.76,  39.59,    16.250,    6.04,  360.92,  22.21
           40,     20.250,    3.36,  807.75,  39.89,     10.125,   24.72,  109.93,  10.86,     40.500,    3.50, 1554.90,  38.39,    20.250,    7.43,  365.62,  18.06
           48,     24.250,    3.39,  959.39,  39.56,     12.125,   29.28,  111.16,   9.17,     48.500,    3.96, 1644.34,  33.90,    24.250,    8.80,  369.75,  15.25
           56,     28.250,    3.40, 1115.41,  39.48,     14.125,   33.52,  113.10,   8.01,     56.500,    4.46, 1699.94,  30.09,    28.250,   10.19,  372.26,  13.18
           64,     32.250,    3.38, 1279.02,  39.66,     16.125,   37.58,  115.17,   7.14,     64.500,    5.02, 1725.21,  26.75,    32.250,   11.56,  374.46,  11.61
           80,     40.250,    3.46, 1559.89,  38.75,     20.125,   47.50,  113.73,   5.65,     80.500,    6.11, 1768.38,  21.97,    40.250,   14.32,  377.24,   9.37
           96,     48.250,    3.93, 1646.63,  34.13,     24.125,   56.95,  113.72,   4.71,     96.500,    9.15, 1414.96,  14.66,    48.250,   17.08,  379.13,   7.86
          128,     64.250,    4.96, 1739.76,  27.08,     32.125,   74.66,  115.51,   3.60,    128.500,    9.76, 1767.41,  13.75,    64.250,   22.60,  381.62,   5.94
          192,     96.250,    9.10, 1419.50,  14.75,     48.125,  116.26,  111.12,   2.31,    192.500,   13.76, 1877.04,   9.75,    96.250,   33.64,  384.00,   3.99
          256,    128.250,    9.53, 1805.36,  14.08,     64.125,  149.35,  115.26,   1.80,    256.500,   18.19, 1893.07,   7.38,   128.250,   44.68,  385.26,   3.00
--------------------------------------------------------------------------------------------------------------------------------------------------------------------
mixbench-ocl/alternating (v0.03-19-gf47d11e)
Use "-h" argument to see available options
------------------------ Device specifications ------------------------
Platform:            AMD Accelerated Parallel Processing
Device:              gfx902/Advanced Micro Devices, Inc.
Driver version:      3180.7 (PAL,HSAIL)
Address bits:        64
GPU clock rate:      1900 MHz
Total global mem:    3584 MB
Max allowed buffer:  1740 MB
OpenCL version:      OpenCL 2.0 AMD-APP (3180.7)
Total CUs:           28
-----------------------------------------------------------------------
Buffer size:            64MB
Workgroup size:         256
Workitem stride:        NDRange
Buffer allocation:      Device allocated
Timer:                  CL event based
Loading kernel source file...
Precompilation of kernels... [>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>]
----------------------------------------------------------------------------- CSV data -----------------------------------------------------------------------------
Experiment ID, Single Precision ops,,,,              Double precision ops,,,,              Half precision ops,,,,                Integer operations,,, 
Compute iters, Flops/byte, ex.time,  GFLOPS, GB/sec, Flops/byte, ex.time,  GFLOPS, GB/sec, Flops/byte, ex.time,  GFLOPS, GB/sec, Iops/byte, ex.time,   GIOPS, GB/sec
            0,      0.000,  269.94,    0.00,  31.82,      0.000,  588.99,    0.00,  29.17,      0.000,  214.27,    0.00,  40.09,     0.000,  270.42,    0.00,  31.77
            1,      0.129,  201.65,    5.32,  41.27,      0.065,  574.42,    1.87,  28.97,      0.258,  200.25,   10.72,  41.56,     0.129,  202.25,    5.31,  41.14
            2,      0.267,  182.18,   11.79,  44.20,      0.133,  552.18,    3.89,  29.17,      0.533,  179.98,   23.86,  44.74,     0.267,  182.54,   11.76,  44.12
            3,      0.414,  170.48,   18.90,  45.66,      0.207,  536.01,    6.01,  29.05,      0.828,  168.12,   38.32,  46.30,     0.414,  168.59,   19.11,  46.18
            4,      0.571,  154.24,   27.85,  48.73,      0.286,  510.65,    8.41,  29.44,      1.143,  154.09,   55.75,  48.78,     0.571,  150.79,   28.48,  49.84
            5,      0.741,  143.51,   37.41,  50.50,      0.370,  494.03,   10.87,  29.34,      1.481,   86.00,  124.85,  84.28,     0.741,  141.45,   37.95,  51.24
            6,      0.923,  130.96,   49.19,  53.29,      0.462,  467.91,   13.77,  29.83,      1.846,  129.48,   99.51,  53.90,     0.923,  126.02,   51.12,  55.38
            7,      1.120,  120.09,   62.59,  55.88,      0.560,  447.22,   16.81,  30.01,      2.240,  119.36,  125.94,  56.23,     1.120,  116.96,   64.26,  57.38
            8,      1.333,  106.32,   80.79,  60.59,      0.667,  422.33,   20.34,  30.51,      2.667,  106.95,  160.63,  60.24,     1.333,  101.32,   84.78,  63.58
            9,      1.565,   97.78,   98.84,  63.15,      0.783,  378.60,   25.52,  32.62,      3.130,   50.05,  386.20, 123.37,     1.565,   92.08,  104.95,  67.05
           10,      1.818,   88.04,  121.96,  67.08,      0.909,  359.01,   29.91,  32.90,      3.636,  118.41,  181.36,  49.87,     1.818,   79.72,  134.69,  74.08
           11,      2.095,   80.29,  147.11,  70.21,      1.048,  318.47,   37.09,  35.40,      4.190,   79.33,  297.77,  71.06,     2.095,   49.12,  240.45, 114.76
           12,      2.400,   71.08,  181.28,  75.53,      1.200,  300.32,   42.90,  35.75,      4.800,   67.46,  381.98,  79.58,     2.400,   51.15,  251.89, 104.95
           13,      2.737,   64.09,  217.80,  79.58,      1.368,  151.32,   92.25,  67.41,      5.474,   59.40,  469.96,  85.86,     2.737,   53.26,  262.07,  95.76
           14,      3.111,   55.73,  269.75,  86.70,      1.556,  141.07,  106.56,  68.50,      6.222,   32.83,  915.82, 147.19,     3.111,   55.47,  270.99,  87.10
           15,      3.529,   45.44,  354.44, 100.42,      1.765,  148.93,  108.15,  61.28,      7.059,   32.86,  980.17, 138.86,     3.529,   56.92,  282.98,  80.18
           16,      4.000,  106.63,  161.11,  40.28,      2.000,  224.16,   76.64,  38.32,      8.000,  102.77,  334.34,  41.79,     4.000,   59.59,  288.28,  72.07
           17,      4.533,   86.93,  209.97,  46.32,      2.267,  165.66,  110.19,  48.61,      9.067,   82.00,  445.23,  49.11,     4.533,   59.61,  306.24,  67.55
           18,      5.143,   79.65,  242.66,  47.18,      2.571,  172.66,  111.94,  43.53,     10.286,   74.54,  518.59,  50.42,     5.143,   62.25,  310.48,  60.37
           19,      5.846,   61.84,  329.90,  56.43,      2.923,  182.00,  112.09,  38.35,     11.692,   36.81, 1108.53,  94.81,     5.846,   64.09,  318.33,  54.45
           20,      6.667,   56.78,  378.18,  56.73,      3.333,  188.41,  113.98,  34.19,     13.333,   35.85, 1198.20,  89.87,     6.667,   66.07,  325.05,  48.76
           21,      7.636,   63.94,  352.67,  46.18,      3.818,  198.39,  113.66,  29.77,     15.273,   58.37,  772.66,  50.59,     7.636,   67.99,  331.64,  43.43
           22,      8.800,   82.43,  286.56,  32.56,      4.400,  207.73,  113.71,  25.84,     17.600,   53.89,  876.72,  49.81,     8.800,   69.90,  337.94,  38.40
           23,     10.222,   53.94,  457.88,  44.79,      5.111,  213.83,  115.49,  22.60,     20.444,   49.20, 1003.86,  49.10,    10.222,   71.84,  343.79,  33.63
           24,     12.000,   72.29,  356.49,  29.71,      6.000,  221.29,  116.45,  19.41,     24.000,   38.00, 1356.31,  56.51,    12.000,   73.13,  352.40,  29.37
           25,     14.286,   42.57,  630.58,  44.14,      7.143,  227.61,  117.93,  16.51,     28.571,   35.18, 1525.93,  53.41,    14.286,   75.07,  357.58,  25.03
           26,     17.333,   57.26,  487.52,  28.13,      8.667,  240.25,  116.20,  13.41,     34.667,   49.88, 1119.47,  32.29,    17.333,   77.61,  359.73,  20.75
           27,     21.600,   36.91,  785.37,  36.36,     10.800,  246.09,  117.81,  10.91,     43.200,   35.34, 1640.65,  37.98,    21.600,   79.53,  364.53,  16.88
           28,     28.000,   39.37,  763.61,  27.27,     14.000,  251.36,  119.61,   8.54,     56.000,   34.87, 1724.35,  30.79,    28.000,   80.83,  371.96,  13.28
           29,     38.667,   19.58, 1590.01,  41.12,     19.333,  261.44,  119.10,   6.16,     77.333,   35.97, 1731.53,  22.39,    38.667,   82.76,  376.24,   9.73
           30,     60.000,   18.69, 1723.09,  28.72,     30.000,  270.07,  119.27,   3.98,    120.000,   36.06, 1786.74,  14.89,    60.000,   84.68,  380.41,   6.34
           31,    124.000,   18.37, 1811.81,  14.61,     62.000,  279.49,  119.10,   1.92,    248.000,   35.43, 1878.83,   7.58,   124.000,   88.59,  375.72,   3.03
           32,        inf,   17.69, 1942.26,   0.00,        inf,  299.14,  114.86,   0.00,        inf,   36.96, 1859.47,   0.00,       inf,   88.33,  389.01,   0.00
--------------------------------------------------------------------------------------------------------------------------------------------------------------------

kingbeowulf · 12-23-2021, 12:26 AM

I'm waiting for Slackware-current to settle down a bit more, as well as work, home etc has been a bit hectic. Rumor has it that "real soon now" might actually be "real soon"! LOL.

ROCm still does not yet "officially" support NAVI and NAVI2 GPUs, yet now ROCm is slipped into AMDGPU-PRO. Bunch of weirdos. I'm going to take a stab at a buildscript for ROCm to get the runtime (ROCr) bits working. The AMDGPU-PRO installation as in my amdgpu-opencl-20.40 script will be problematic from a maintenance standpoint, and the their installer uses those weird package managers (apt, yum etc). I'll try packaging 21.40.1 as well.

@denydias thanks for the feedback. Since you have Ryzen integrated GPU (Vega 8 I think), you may be better off with 20.40 than trying ROCm/ROCr until AMD clarifies support.

@dchmelik with amdgpu-opencl-20.40, on my RX5700XT and RX590, PhoenixMiner (eth, etc) work with about 51 and 27 MH/s respectively, as do various BOINC projects, over various 5.14.x and 5.15.x kernels. The RX590 does crash on occasion, but I think that is my card. Both boxes have 850 W power supplies. Not sure why you can't run PhoenixMiner.

I'll keep versions on my vanity website. Once Slackbuilds.org opens up, people can use their cgit to access older versions as well.

denydias · 12-23-2021, 01:00 AM

Quote:

Originally Posted by kingbeowulf

I'm waiting for Slackware-current to settle down a bit more, as well as work, home etc has been a bit hectic. Rumor has it that "real soon now" might actually be "real soon"! LOL.

Dunno... if it doesn't come in two days, then I don't now when it will come. :P

Quote:

Originally Posted by kingbeowulf

@denydias thanks for the feedback. Since you have Ryzen integrated GPU (Vega 8 I think), you may be better off with 20.40 than trying ROCm/ROCr until AMD clarifies support.

My pleasure. 20.40 is rock solid here. The only issue is imagemagick compiled with `--disable-opencl`, but gimp, libreoffice's calc and others are working like a charm.

Whenever you release a new ROCm/ROCr SBo, just ping us here.

Thanks for the effort, @kingbeowulf.

dchmelik · 12-23-2021, 01:21 AM

Quote:

Originally Posted by kingbeowulf

@dchmelik with amdgpu-opencl-20.40, on my RX5700XT and RX590, PhoenixMiner (eth, etc) work with about 51 and 27 MH/s respectively, as do various BOINC projects, over various 5.14.x and 5.15.x kernels. The RX590 does crash on occasion, but I think that is my card. Both boxes have 850 W power supplies. Not sure why you can't run PhoenixMiner.

I was able to run some non-ethereum (ETH) miners, but as for ETH, I switched to TeamRedMiner, which is specifically for AMD GPUs.
I'm looking forward to your next SlackBuild--the one I'm using now might be a bit old...

dchmelik · 04-09-2022, 02:21 AM

Aren't there several new releases to test?

kingbeowulf · 04-18-2022, 12:02 AM

I'm getting there "real soon now." Instead of a binary repack, I thought I'd wade through the ROCm runtime stuff. Either I'm an old man stuck in the '80s, or AMD is making a basic, simple OPENCL runtime build as difficult as possible. Extremely convoluted. I'll sift through the new AMDGPU-PRO soon. I promise!

bassmadrigal · 04-18-2022, 03:36 PM

Quote:

Originally Posted by kingbeowulf

I'm getting there "real soon now." Instead of a binary repack, I thought I'd wade through the ROCm runtime stuff. Either I'm an old man stuck in the '80s, or AMD is making a basic, simple OPENCL runtime build as difficult as possible. Extremely convoluted. I'll sift through the new AMDGPU-PRO soon. I promise!

ROCm is really complex... much more than it seems it should be. If it's not tackled before I get my system finally upgraded to 15.0, I may try and get a SlackBuild going. However, I also have an impending move to a new area of the country, so depending on when I find time to upgrade, it could be after the move. So others will probably have the opportunity to do it before I get a chance.

denydias · 04-18-2022, 05:51 PM

@kingbeowulf
Then I'm and old man stuck in '80s like you too. I told you that I tried to build the new ROCm thing from arc templates before using your script. That attempt took me some precious hours and failed miserably. Unfortunately I'm still too busy with the project I'm working on to give proper attention to this right now, so I'm just sticking with whatever you come up with... as a proof of my appreciation, I'm willing to help you and others on reviewing and testing anything you guys need.

kingbeowulf · 05-02-2022, 06:16 PM

Hello AMDGPU OpenCL testers.

I have a new amdgpu-opencl based on AMDGPU-PRO-22.10.1 for the proprietary (legacy, non-ROCm) OpenCL stack. This "should" work for Vega and older cards. No-go so far for newer Navi since the llvm and libclc included with Slackware is missing, for example, /usr/share/clc/gfx1010-amdgcn-mesa-mesa3d.bc for my Radeon RX 5700 XT (NAVI 1O). I have an old RX 590 (Polaris 30) somewhere - I'll try it as soon as I find it - but it may be too "new" as well. The buildscript is based on Redhat's rpms this time. The Ubuntu debs were a mess.

The SBo style build archive is here: https://www.linuxgalaxy.org/files/sb.../SBo-archives/

note: The included clinfo may not work. You can try this better one: https://github.com/Oblomov/clinfo

I think I have the build sequence for the ROCm OpenCL runtime (rocr). Hopefully I can build the bits with the existing Slackware llvm, etc.

Edit: Whoops. Corrected AMDGPU-PRO-22.10.1 version.

Have fun !