GPU WU failing after update to AMD Linux drivers 20.40

Paul
Paul
Joined: 3 May 07
Posts: 123
Credit: 1785671287
RAC: 265239
Topic 226656

I just upgraded my system, so I have fresh AMD Linux drivers.  All of the things I know to check look okay and the same as they did before, when it was working, albeit w/ incremented minor version numbers.

Mon 27 Dec 2021 06:39:26 PM PST |  | Starting BOINC client version 7.16.16 for x86_64-pc-linux-gnu
Mon 27 Dec 2021 06:39:26 PM PST |  | log flags: file_xfer, sched_ops, task, coproc_debug
Mon 27 Dec 2021 06:39:26 PM PST |  | Libraries: libcurl/7.79.1 OpenSSL/1.1.1l-fips zlib/1.2.11 brotli/1.0.9 libidn2/2.3.2 libpsl/0.21.1 (+libidn2/2.3.2) libssh/0.9.6/openssl/zlib nghttp2/1.45.1 OpenLDAP/2.4.59
Mon 27 Dec 2021 06:39:26 PM PST |  | OpenCL: AMD/ATI GPU 0: AMD Radeon RX 5700 XT (NAVI10, DRM 3.42.0, 5.15.11-200.fc35.x86_ (driver version 21.3.2, device version OpenCL 1.1 Mesa 21.3.2, 8192MB, 8192MB available, 6720 GFLOPS peak)
Mon 27 Dec 2021 06:39:26 PM PST |  | [coproc] NVIDIA: libcuda.so: cannot open shared object file: No such file or directory
Mon 27 Dec 2021 06:39:26 PM PST |  | [coproc] ATI: libaticalrt.so: cannot open shared object file: No such file or directory
Mon 27 Dec 2021 06:39:26 PM PST |  | libc: GNU libc version 2.34
Mon 27 Dec 2021 06:39:26 PM PST |  | Processor: 24 AuthenticAMD AMD Ryzen 9 5900X 12-Core Processor [Family 25 Model 33 Stepping 0]
Mon 27 Dec 2021 06:39:26 PM PST |  | OS: Linux Fedora: Fedora release 35 (Thirty Five) [5.15.11-200.fc35.x86_64|libc 2.34 (GNU libc)]

Just noticed this, so I don't see the error WU reports posted to the server yet.

Paul
Paul
Joined: 3 May 07
Posts: 123
Credit: 1785671287
RAC: 265239

Okay, I see now that the

Okay, I see now that the OpenCL target that I was using with the old driver isn't recognized by BOINC.  However, it shows up in clinfo, which, in the past, has been equivalent to BOINC's check, but I guess even that is broken now.

OpenCL compiling FAILED! : -11 . Error message: fatal error: cannot open file '/usr/lib64/clc/gfx1010-amdgcn-mesa-mesa3d.bc': No such file or directory
Paul
Paul
Joined: 3 May 07
Posts: 123
Credit: 1785671287
RAC: 265239

I go it working, again.  I

I go it working, again.  I just kept trying to install more and more packages, ignoring installation errors, and rebooting, repeatedly, until it started working.

hsakmt-1.0.6-17.rocm3.9.0.fc35.x86_64
rocm-runtime-3.9.0-2.fc35.x86_64
hsakmt-devel-1.0.6-17.rocm3.9.0.fc35.x86_64
xorg-x11-drv-amdgpu-21.0.0-1.fc35.x86_64
python3-setproctitle-1.2.2-3.fc35.x86_64
rocminfo-3.9.0-2.fc35.x86_64
rocm-runtime-devel-3.9.0-2.fc35.x86_64
amdgpu-install-21.40.1.40501-1337797.el8.noarch
rocm-core-4.5.1.40501-84.el8.x86_64
rocm-device-libs-1.0.0.40501-84.el8.x86_64
mesa-amdgpu-libglapi-21.3.0.40501-1337797.el8.x86_64
llvm-amdgpu-libs-12.0.40501-1337797.el8.x86_64
libwayland-amdgpu-server-1.18.0.40501-1337797.el8.x86_64
mesa-amdgpu-libgbm-21.3.0.40501-1337797.el8.x86_64
mesa-amdgpu-filesystem-21.3.0.40501-1337797.el8.x86_64
libwayland-amdgpu-client-1.18.0.40501-1337797.el8.x86_64
hsakmt-roct-devel-20210902.7.5.40501-84.el8.x86_64
hsa-rocr-1.4.0.40501-84.el8.x86_64
hsa-rocr-devel-1.4.0.40501-84.el8.x86_64
rocm-language-runtime-4.5.1.40501-84.el8.x86_64
mesa-amdgpu-libEGL-21.3.0.40501-1337797.el8.x86_64
mesa-amdgpu-dri-drivers-21.3.0.40501-1337797.el8.x86_64
mesa-amdgpu-vdpau-drivers-21.3.0.40501-1337797.el8.x86_64
mesa-amdgpu-libxatracker-21.3.0.40501-1337797.el8.x86_64
mesa-amdgpu-libGL-21.3.0.40501-1337797.el8.x86_64
mesa-amdgpu-libGLES-21.3.0.40501-1337797.el8.x86_64
libdrm-amdgpu-common-1.0.0.40501-1337797.el8.noarch
libwayland-amdgpu-egl-1.18.0.40501-1337797.el8.x86_64
rocm-llvm-13.0.0.21432.40501-84.el8.x86_64
hip-runtime-amd-4.4.21432.40501-84.el8.x86_64
rocm-hip-runtime-4.5.1.40501-84.el8.x86_64
vulkan-amdgpu-21.40.1-1337797.el8.x86_64
amdgpu-pro-core-21.40.1-1337797.el8.noarch
ocl-icd-amdgpu-pro-21.40.1-1337797.el8.x86_64
clinfo-amdgpu-pro-21.40.1-1337797.el8.x86_64
rocm-opencl-2.0.0.40501-84.el8.x86_64
rocm-opencl-runtime-4.5.1.40501-84.el8.x86_64
amdgpu-pro-rocr-opencl-21.40.1-1337797.el8.x86_64
libdrm-amdgpu-2.4.107.40501-1337797.el8.x86_64
xorg-x11-amdgpu-drv-amdgpu-24.1.0-1337797.el8.x86_64
llvm120-amdgpu-12.0.40501-1337797.el8.x86_64
llvm-amdgpu-12.0.40501-1337797.el8.x86_64
llvm120-amdgpu-devel-12.0.40501-1337797.el8.x86_64
llvm-amdgpu-devel-12.0.40501-1337797.el8.x86_64
llvm-amdgpu-static-12.0.40501-1337797.el8.x86_64
amdgpu-lib-21.40.1.40501-1337797.el8.x86_64
libva-amdgpu-2.8.0.40501-1337797.el8.x86_64
opencl-legacy-amdgpu-pro-icd-21.40.1-1337797.el8.x86_64
libwayland-amdgpu-cursor-1.18.0.40501-1337797.el8.x86_64
amdgpu-versionlist-21.40.1.40501-1337797.el8.noarch
drm-utils-amdgpu-2.4.107.40501-1337797.el8.x86_64

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.