search

fireice-uk / xmr-stak

Free Monero RandomX Miner and unified CryptoNight miner
GNU General Public License v3.0
4.05k stars 1.79k forks source link

Segmentation fault (AMD+clang7) #1758

Closed minzak closed 6 years ago

minzak commented 6 years ago

Hm, very strange, all was worked, but no today. Latest Debian 9 x64. NO opencl-mesa I Add Clang7 in system (i read issues, where clang7 is better), i think/guess this reason of Segmentation fault:

Commandline: apt-get upgrade
Upgrade: linux-libc-dev:amd64 (4.16.16-2~bpo9+1, 4.17.8-1~bpo9+1), linux-compiler-gcc-6-x86:amd64 (4.16.16-2~bpo9+1, 4.17.8-1~bpo9+1), libfuse2:amd64 (2.9.7-1, 2.9.7-1+deb9u1)
apt-get install clang-7 clang-tools-7 libclang-common-7-dev libclang-7-dev libclang1-7 libllvm7 lldb-7 llvm-7 llvm-7-dev llvm-7-examples llvm-7-runtime clang-format-7 python-clang-7 lld-7 libfuzzer-7-dev clang-7-doc llvm-7-doc

My build is:

root@ferma:/usr/src# git clone https://github.com/fireice-uk/xmr-stak
Cloning into 'xmr-stak'...
remote: Counting objects: 5091, done.
remote: Total 5091 (delta 0), reused 0 (delta 0), pack-reused 5091
Receiving objects: 100% (5091/5091), 1.48 MiB | 2.89 MiB/s, done.
Resolving deltas: 100% (3634/3634), done.
root@ferma:/usr/src#
root@ferma:/usr/src/xmr-stak/build# CC=/usr/lib/llvm-7/bin/clang CXX=/usr/lib/llvm-7/bin/clang++ cmake -DCMAKE_BUILD_TYPE=Debug -DCUDA_ENABLE=OFF -DOpenCL_ENABLE=ON -DCPU_ENABL                                      E=ON ..
-- The C compiler identification is Clang 7.0.0
-- The CXX compiler identification is Clang 7.0.0
-- Check for working C compiler: /usr/lib/llvm-7/bin/clang
-- Check for working C compiler: /usr/lib/llvm-7/bin/clang -- works
-- Detecting C compiler ABI info
-- Detecting C compiler ABI info - done
-- Detecting C compile features
-- Detecting C compile features - done
-- Check for working CXX compiler: /usr/lib/llvm-7/bin/clang++
-- Check for working CXX compiler: /usr/lib/llvm-7/bin/clang++ -- works
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Detecting CXX compile features
-- Detecting CXX compile features - done
-- Looking for CL_VERSION_2_0
-- Looking for CL_VERSION_2_0 - found
-- Found OpenCL: /opt/amd-app-sdk/lib/x86_64/libOpenCL.so (found version "2.0")
-- Looking for pthread.h
-- Looking for pthread.h - found
-- Looking for pthread_create
-- Looking for pthread_create - not found
-- Looking for pthread_create in pthreads
-- Looking for pthread_create in pthreads - not found
-- Looking for pthread_create in pthread
-- Looking for pthread_create in pthread - found
-- Found Threads: TRUE
-- Found OpenSSL: /usr/lib/x86_64-linux-gnu/libssl.so;/usr/lib/x86_64-linux-gnu/libcrypto.so (found version "1.1.0f")
-- Configuring done
-- Generating done
-- Build files have been written to: /usr/src/xmr-stak/build
root@ferma:/usr/src/xmr-stak/build# make -j 3
Scanning dependencies of target xmr-stak-c
[ 10%] Building C object CMakeFiles/xmr-stak-c.dir/xmrstak/backend/cpu/crypto/c_jh.c.o
[ 10%] Building C object CMakeFiles/xmr-stak-c.dir/xmrstak/backend/cpu/crypto/c_groestl.c.o
[ 10%] Building C object CMakeFiles/xmr-stak-c.dir/xmrstak/backend/cpu/crypto/c_blake256.c.o
[ 13%] Building C object CMakeFiles/xmr-stak-c.dir/xmrstak/backend/cpu/crypto/c_keccak.c.o
[ 16%] Building C object CMakeFiles/xmr-stak-c.dir/xmrstak/backend/cpu/crypto/c_skein.c.o
[ 20%] Linking C static library bin/libxmr-stak-c.a
[ 20%] Built target xmr-stak-c
Scanning dependencies of target xmr-stak-backend
[ 30%] Building CXX object CMakeFiles/xmr-stak-backend.dir/xmrstak/version.cpp.o
[ 30%] Building CXX object CMakeFiles/xmr-stak-backend.dir/xmrstak/jconf.cpp.o
[ 30%] Building CXX object CMakeFiles/xmr-stak-backend.dir/xmrstak/backend/cpu/hwlocMemory.cpp.o
[ 33%] Building CXX object CMakeFiles/xmr-stak-backend.dir/xmrstak/backend/cpu/jconf.cpp.o
[ 36%] Building CXX object CMakeFiles/xmr-stak-backend.dir/xmrstak/backend/cpu/minethd.cpp.o
[ 43%] Building CXX object CMakeFiles/xmr-stak-backend.dir/xmrstak/backend/backendConnector.cpp.o
[ 43%] Building CXX object CMakeFiles/xmr-stak-backend.dir/xmrstak/backend/globalStates.cpp.o
[ 46%] Building CXX object CMakeFiles/xmr-stak-backend.dir/xmrstak/backend/cpu/crypto/cryptonight_common.cpp.o
[ 50%] Building CXX object CMakeFiles/xmr-stak-backend.dir/xmrstak/http/httpd.cpp.o
[ 53%] Building CXX object CMakeFiles/xmr-stak-backend.dir/xmrstak/http/webdesign.cpp.o
[ 56%] Building CXX object CMakeFiles/xmr-stak-backend.dir/xmrstak/misc/console.cpp.o
[ 60%] Building CXX object CMakeFiles/xmr-stak-backend.dir/xmrstak/misc/executor.cpp.o
[ 63%] Building CXX object CMakeFiles/xmr-stak-backend.dir/xmrstak/misc/telemetry.cpp.o
[ 66%] Building CXX object CMakeFiles/xmr-stak-backend.dir/xmrstak/misc/uac.cpp.o
[ 70%] Building CXX object CMakeFiles/xmr-stak-backend.dir/xmrstak/misc/utility.cpp.o
[ 73%] Building CXX object CMakeFiles/xmr-stak-backend.dir/xmrstak/net/jpsock.cpp.o
[ 76%] Building CXX object CMakeFiles/xmr-stak-backend.dir/xmrstak/net/socket.cpp.o
[ 80%] Linking CXX static library bin/libxmr-stak-backend.a
[ 80%] Built target xmr-stak-backend
Scanning dependencies of target xmr-stak
Scanning dependencies of target xmrstak_opencl_backend
[ 83%] Building CXX object CMakeFiles/xmr-stak.dir/xmrstak/cli/cli-miner.cpp.o
[ 86%] Building CXX object CMakeFiles/xmrstak_opencl_backend.dir/xmrstak/backend/amd/amd_gpu/gpu.cpp.o
[ 90%] Building CXX object CMakeFiles/xmrstak_opencl_backend.dir/xmrstak/backend/amd/jconf.cpp.o
[ 93%] Building CXX object CMakeFiles/xmrstak_opencl_backend.dir/xmrstak/backend/amd/minethd.cpp.o
[ 96%] Linking CXX executable bin/xmr-stak
[ 96%] Built target xmr-stak
[100%] Linking CXX shared library bin/libxmrstak_opencl_backend.so
[100%] Built target xmrstak_opencl_backend
root@ferma:/usr/src/xmr-stak/build#

I run xmr with deleted amd.txt

root@ferma:/opt/xmr-stak# ./xmr-stak
-------------------------------------------------------------------
xmr-stak 2.4.7 c5f0505

Brought to you by fireice_uk and psychocrypt under GPLv3.
Based on CPU mining code by wolf9466 (heavily optimized by fireice_uk).
Based on OpenCL mining code by wolf9466.

Configurable dev donation level is set to 2.0%

You can use following keys to display reports:
'h' - hashrate
'r' - results
'c' - connection
-------------------------------------------------------------------
[2018-07-30 22:16:51] : Mining coin: cryptonight_v7
[2018-07-30 22:16:51] : Found AMD platform index id = 0, name = Advanced Micro Devices, Inc.
[2018-07-30 22:16:51] : Found OpenCL GPU Ellesmere.
[2018-07-30 22:16:51] : Found OpenCL GPU Ellesmere.
[2018-07-30 22:16:51] : Found OpenCL GPU Ellesmere.
[2018-07-30 22:16:51] : Found OpenCL GPU Ellesmere.
[2018-07-30 22:16:51] : Found OpenCL GPU Ellesmere.
[2018-07-30 22:16:51] : Found OpenCL GPU Ellesmere.
[2018-07-30 22:16:51] : Found OpenCL GPU Ellesmere.
[2018-07-30 22:16:51] : Found OpenCL GPU Ellesmere.
[2018-07-30 22:16:51] : AMD: GPU (OpenCL) configuration stored in file 'amd.txt'
[2018-07-30 22:16:51] : Compiling code and initializing GPUs. This will take a while...
[2018-07-30 22:16:51] : Device 0 work size 8 / 32.
[2018-07-30 22:16:51] : OpenCL device 0 - Load precompiled code from file /root/.openclcache/8979c28ee48993be74e0dff567217807eb53d517c4620575576267503777c566.openclbin
Segmentation fault

root@ferma:/opt/xmr-stak#

In system opencl-amdgpu:

root@ferma:/opt/xmr-stak# dpkg-query -l | grep opencl-amdgpu
ii  opencl-amdgpu-pro-icd:amd64                     17.40-514569                                          amd64        non-free AMD OpenCL ICD Loaders
root@ferma:/opt/xmr-stak#

HelloWorld from example OpenCL is work.

Also Add my clinfo clinfo.txt

How to check it, what wrong? Thanks.

Spudz76 commented 6 years ago

When recompiling or running any different version (new/changed at all) please rm -rf /root/.openclcache/ before running changed executable.

It is loading a precompiled (with the other version) OpenCL kernel binary which must be incompatible with changed driver/SDK/compiler. Delete them and it will recompile and probably not crash.

That cache thing was probably the worst new feature in a while. Nice idea, causes too much dumb, saves only 10 seconds per GPU. Probably need to drop a .version file in there and automatically delete everything if you run some other version than last one that created the .version file, to make it less of a "just delete the cache constantly out of just making sure"

minzak commented 6 years ago

Yes, /.openclcache/ - always delete before run, previous i had bad screen

[2018-07-30 22:50:06] : Mining coin: cryptonight_v7
[2018-07-30 22:50:06] : Compiling code and initializing GPUs. This will take a while...
[2018-07-30 22:50:06] : Device 3 work size 8 / 32.
[2018-07-30 22:50:06] : OpenCL device 3 - Precompiled code /root/.openclcache/ec3329e6dc4ed1a7eb265d882b810a43ec443fcf29c456e67987c0ffdc57e3f6.openclbin not found. Compiling ...
[2018-07-30 22:50:18] : OpenCL device 3 - Precompiled code stored in file /root/.openclcache/ec3329e6dc4ed1a7eb265d882b810a43ec443fcf29c456e67987c0ffdc57e3f6.openclbin
[2018-07-30 22:50:18] : Device 3 work size 8 / 32.
[2018-07-30 22:50:18] : OpenCL device 3 - Load precompiled code from file /root/.openclcache/ec3329e6dc4ed1a7eb265d882b810a43ec443fcf29c456e67987c0ffdc57e3f6.openclbin
Segmentation fault

Ok, i read about something not compatible with clang, but not fully understand (Ukrainian/Russian is my native language)

Ok i have 3 way, i'm right? 1- just apt purge previous clang - it must helps. 2- compile OpenCL and XMR-STAK with same compiler (gcc or clang7 or else..) 3- use Gentoo and recompile all whole system ))

Spudz76 commented 6 years ago

Which way are you providing the alternate compiler to cmake?

I use similar to:

CC=/usr/bin/clang-7 CXX=/usr/bin/clang++-7 \
 cmake -DCMAKE_BUILD_TYPE=Release -DCUDA_ENABLE=OFF -DOpenCL_ENABLE=OFF ..

I will test clang-7 and OpenCL it may not work / I only tested alternative compilers for CPU backend.

No reason to purge, I have 7 compilers installed, all work perfectly. Do try clang-3.8 though I think it does work with OpenCL. Maybe clang-7 is too new for AMD stuff.

EDIT: definitely using clang-3.8 for one of my xenial AMD rigs.

Spudz76 commented 6 years ago

You could also try running with command line arg --noAMDCache

The entire cache hack is somewhat undocumented. It compiles on one GPU then dumps the memory using unofficial methods to a file, and to load a cache it just dumps that binary onto the other GPU. That could be easily broken while the rest of "normal" OpenCL usage is not, the technique is not AMD supported and happened to work but maybe not with other compilers/drivers than the standards.

Spudz76 commented 6 years ago

clang-6.0 worked, I have trouble getting clang-7 on Ubuntu Xenial

Which driver are you trying? I use 17.50-552542 with this build command:

CC=/usr/bin/clang-6.0 CXX=/usr/bin/clang++-6.0 \
cmake -DCMAKE_BUILD_TYPE=Release \
      -DCUDA_ENABLE=OFF \
      -DOpenCL_ENABLE=ON \
      -DOpenCL_LIBRARY=/opt/amdgpu-pro/lib/x86_64-linux-gnu/libOpenCL.so \
   ..

My successful (with cache too) output:

[2018-07-30 14:16:46] : Mining coin: monero7
!!!! Doing only a benchmark and exiting. To mine, remove the '--benchmark' option. !!!!
[2018-07-30 14:16:46] : Prepare benchmark for block version 7
[2018-07-30 14:16:46] : Compiling code and initializing GPUs. This will take a while...
[2018-07-30 14:16:46] : Device 0 work size 8 / 32.
[2018-07-30 14:16:46] : OpenCL device 0 - Precompiled code /root/.openclcache/5a904bcadf5b454083f9787e13b8eae5206e090ee44af17d7721e5a2ecfce484.openclbin not found. Compiling ...
[2018-07-30 14:17:20] : OpenCL device 0 - Precompiled code stored in file /root/.openclcache/5a904bcadf5b454083f9787e13b8eae5206e090ee44af17d7721e5a2ecfce484.openclbin
[2018-07-30 14:17:20] : Device 1 work size 8 / 32.
[2018-07-30 14:17:21] : OpenCL device 1 - Load precompiled code from file /root/.openclcache/5a904bcadf5b454083f9787e13b8eae5206e090ee44af17d7721e5a2ecfce484.openclbin
[2018-07-30 14:17:21] : Device 2 work size 8 / 32.
[2018-07-30 14:17:21] : OpenCL device 2 - Load precompiled code from file /root/.openclcache/5a904bcadf5b454083f9787e13b8eae5206e090ee44af17d7721e5a2ecfce484.openclbin
[2018-07-30 14:17:21] : Device 3 work size 8 / 32.
[2018-07-30 14:17:21] : OpenCL device 3 - Load precompiled code from file /root/.openclcache/5a904bcadf5b454083f9787e13b8eae5206e090ee44af17d7721e5a2ecfce484.openclbin
[2018-07-30 14:17:21] : Device 4 work size 8 / 32.
[2018-07-30 14:17:22] : OpenCL device 4 - Precompiled code /root/.openclcache/54896cf94f10944a77d20e4f720146c0b5a35b7935080c784a64424cfcd5c084.openclbin not found. Compiling ...
[2018-07-30 14:17:58] : OpenCL device 4 - Precompiled code stored in file /root/.openclcache/54896cf94f10944a77d20e4f720146c0b5a35b7935080c784a64424cfcd5c084.openclbin
[2018-07-30 14:17:58] : Device 5 work size 8 / 32.
[2018-07-30 14:17:59] : OpenCL device 5 - Precompiled code /root/.openclcache/e1b56274bc8080c47cade4f242ff7d3799f68456adfc2d869a41a7c3961a95f3.openclbin not found. Compiling ...
[2018-07-30 14:18:35] : OpenCL device 5 - Precompiled code stored in file /root/.openclcache/e1b56274bc8080c47cade4f242ff7d3799f68456adfc2d869a41a7c3961a95f3.openclbin
[2018-07-30 14:18:35] : Starting AMD GPU (OpenCL) thread 0, no affinity.
[2018-07-30 14:18:35] : Starting AMD GPU (OpenCL) thread 1, no affinity.
[2018-07-30 14:18:35] : Starting AMD GPU (OpenCL) thread 2, no affinity.
[2018-07-30 14:18:35] : Starting AMD GPU (OpenCL) thread 3, no affinity.
[2018-07-30 14:18:35] : Starting AMD GPU (OpenCL) thread 4, no affinity.
[2018-07-30 14:18:35] : Starting AMD GPU (OpenCL) thread 5, no affinity.
[2018-07-30 14:18:35] : WARNING: backend CPU disabled.
[2018-07-30 14:18:35] : Wait 5 sec until all backends are initialized
[2018-07-30 14:18:40] : Start a 15 second benchmark...
[2018-07-30 14:18:55] : Benchmark Thread 0 amd: 706.8 H/S
[2018-07-30 14:18:55] : Benchmark Thread 1 amd: 731.7 H/S
[2018-07-30 14:18:55] : Benchmark Thread 2 amd: 677.2 H/S
[2018-07-30 14:18:55] : Benchmark Thread 3 amd: 678.3 H/S
[2018-07-30 14:18:55] : Benchmark Thread 4 amd: 328.9 H/S
[2018-07-30 14:18:55] : Benchmark Thread 5 amd: 269.1 H/S
[2018-07-30 14:18:55] : Benchmark Total: 3392.0 H/S

That was running at the same time as ethdcrminer64 -- kind of good rates really (I think it made the ETH rates really low though, during)

minzak commented 6 years ago

yes, i write previous, i build with this string: CC=/usr/lib/llvm-7/bin/clang CXX=/usr/lib/llvm-7/bin/clang++ cmake -DCMAKE_BUILD_TYPE=Debug -DCUDA_ENABLE=OFF -DOpenCL_ENABLE=ON -DCPU_ENABLE=ON .. my driver is 17.40-514569, ok will try use latest 17.50.xx

Hm, purge clang, and recompile with no clang - not helps

apt purge clang-7 clang-tools-7 libclang-common-7-dev libclang-7-dev libclang1-7 libllvm7 lldb-7 llvm-7 llvm-7-dev llvm-7-examples llvm-7-runtime clang-format-7 python-clang-7 lld-7 libfuzzer-7-dev clang-7-doc llvm-7-doc -y
apt autoermove -y
reboot

Ok. i try to use clang6 on Debian

Hm, very intresting, when i delete cache and run - i see that cache build but not work, Maybe something leave in system from clang7 ?

root@ferma:/opt/xmr-stak# ./xmr-stak
-------------------------------------------------------------------
xmr-stak 2.4.7 c5f0505

Brought to you by fireice_uk and psychocrypt under GPLv3.
Based on CPU mining code by wolf9466 (heavily optimized by fireice_uk).
Based on OpenCL mining code by wolf9466.

Configurable dev donation level is set to 2.0%

You can use following keys to display reports:
'h' - hashrate
'r' - results
'c' - connection
-------------------------------------------------------------------
[2018-07-30 23:18:02] : Mining coin: cryptonight_v7
[2018-07-30 23:18:02] : Compiling code and initializing GPUs. This will take a while...
[2018-07-30 23:18:02] : Device 3 work size 8 / 32.
[2018-07-30 23:18:02] : OpenCL device 3 - Precompiled code /root/.openclcache/ec3329e6dc4ed1a7eb265d882b810a43ec443fcf29c456e67987c0ffdc57e3f6.openclbin not found. Compiling ...
[2018-07-30 23:18:13] : OpenCL device 3 - Precompiled code stored in file /root/.openclcache/ec3329e6dc4ed1a7eb265d882b810a43ec443fcf29c456e67987c0ffdc57e3f6.openclbin
[2018-07-30 23:18:13] : Device 3 work size 8 / 32.
[2018-07-30 23:18:13] : OpenCL device 3 - Load precompiled code from file /root/.openclcache/ec3329e6dc4ed1a7eb265d882b810a43ec443fcf29c456e67987c0ffdc57e3f6.openclbin
Segmentation fault

But when i use --noAMDCache - it is works! But it is just every time compile cache, and work..

root@ferma:/opt/xmr-stak# ./xmr-stak --noAMDCache
-------------------------------------------------------------------
xmr-stak 2.4.7 c5f0505

Brought to you by fireice_uk and psychocrypt under GPLv3.
Based on CPU mining code by wolf9466 (heavily optimized by fireice_uk).
Based on OpenCL mining code by wolf9466.

Configurable dev donation level is set to 2.0%

You can use following keys to display reports:
'h' - hashrate
'r' - results
'c' - connection
-------------------------------------------------------------------
[2018-07-30 23:20:11] : Mining coin: cryptonight_v7
[2018-07-30 23:20:11] : Compiling code and initializing GPUs. This will take a while...
[2018-07-30 23:20:11] : Device 3 work size 8 / 32.
[2018-07-30 23:20:23] : Device 3 work size 8 / 32.
[2018-07-30 23:20:34] : Device 4 work size 8 / 32.
[2018-07-30 23:20:46] : Device 4 work size 8 / 32.
[2018-07-30 23:20:57] : Device 0 work size 8 / 32.
[2018-07-30 23:21:08] : Device 0 work size 8 / 32.
[2018-07-30 23:21:20] : Device 1 work size 8 / 32.
[2018-07-30 23:21:31] : Device 1 work size 8 / 32.
[2018-07-30 23:21:43] : Device 2 work size 8 / 32.
[2018-07-30 23:21:55] : Device 2 work size 8 / 32.
[2018-07-30 23:22:06] : Device 5 work size 8 / 32.
[2018-07-30 23:22:17] : Device 5 work size 8 / 32.
[2018-07-30 23:22:29] : Device 6 work size 8 / 32.
[2018-07-30 23:22:40] : Device 6 work size 8 / 32.
[2018-07-30 23:22:51] : Device 7 work size 8 / 32.
[2018-07-30 23:23:03] : Device 7 work size 8 / 32.
[2018-07-30 23:23:14] : Starting AMD GPU (OpenCL) thread 0, no affinity.
[2018-07-30 23:23:14] : Starting AMD GPU (OpenCL) thread 1, no affinity.
[2018-07-30 23:23:14] : Starting AMD GPU (OpenCL) thread 2, no affinity.
[2018-07-30 23:23:14] : Starting AMD GPU (OpenCL) thread 3, no affinity.
[2018-07-30 23:23:14] : Starting AMD GPU (OpenCL) thread 4, no affinity.
[2018-07-30 23:23:14] : Starting AMD GPU (OpenCL) thread 5, no affinity.
[2018-07-30 23:23:14] : Starting AMD GPU (OpenCL) thread 6, no affinity.
[2018-07-30 23:23:14] : Starting AMD GPU (OpenCL) thread 7, no affinity.
[2018-07-30 23:23:14] : Starting AMD GPU (OpenCL) thread 8, no affinity.
[2018-07-30 23:23:14] : Starting AMD GPU (OpenCL) thread 9, no affinity.
[2018-07-30 23:23:14] : Starting AMD GPU (OpenCL) thread 10, no affinity.
[2018-07-30 23:23:14] : Starting AMD GPU (OpenCL) thread 11, no affinity.
[2018-07-30 23:23:14] : Starting AMD GPU (OpenCL) thread 12, no affinity.
[2018-07-30 23:23:14] : Starting AMD GPU (OpenCL) thread 13, no affinity.
[2018-07-30 23:23:14] : Starting AMD GPU (OpenCL) thread 14, no affinity.
[2018-07-30 23:23:14] : Starting AMD GPU (OpenCL) thread 15, no affinity.
[2018-07-30 23:23:14] : Starting 1x thread, affinity: 0.
[2018-07-30 23:23:14] : hwloc: memory pinned
[2018-07-30 23:23:14] : Starting 1x thread, affinity: 1.
[2018-07-30 23:23:14] : hwloc: memory pinned
[2018-07-30 23:23:14] : Starting 1x thread, affinity: 2.
[2018-07-30 23:23:14] : hwloc: memory pinned
[2018-07-30 23:23:14] : Fast-connecting to monerov.ingest.cryptoknight.cc:9222 pool ...
[2018-07-30 23:23:14] : Fast-connecting to graft.ingest.cryptoknight.cc:9112 pool ...
[2018-07-30 23:23:14] : Pool monerov.ingest.cryptoknight.cc:9222 connected. Logging in...
[2018-07-30 23:23:14] : Pool graft.ingest.cryptoknight.cc:9112 connected. Logging in...
[2018-07-30 23:23:15] : Difficulty changed. Now: 150000.
[2018-07-30 23:23:15] : Pool logged in.
[2018-07-30 23:23:28] : Result accepted by the pool.
minzak commented 6 years ago

Oh, no, not work, but run and has a slow hashrate:

HASHRATE REPORT - AMD
| ID |    10s |    60s |    15m | ID |    10s |    60s |    15m |
|  0 |   (na) |   63.4 |   (na) |  1 |   (na) |   (na) |   (na) |
|  2 |   (na) |   63.4 |   (na) |  3 |   (na) |   (na) |   (na) |
Totals (AMD):     0.0  126.7    0.0 H/s
-----------------------------------------------------------------
Totals (ALL):    193.7  320.7    0.0 H/s
Highest:     0.0 H/s
-----------------------------------------------------------------

It is like AMD bug, when Card in 16xPCI slot, but i not use 16x.

How to see logs what OpenCL is compiled? I think It helps to detect this issue.

psychocrypt commented 6 years ago

It looks like an issue we had also on OSX. As I remember the problem was a broken OpenCL driver.

minzak commented 6 years ago

@psychocrypt You right, it something wrong with driver, but many OpenCL work, But also interesting to see how almost half Cards are freezing, 0MHz CLCs and etc. Ok, thanks to all, i will reinstall whole system. And try to use clang 6)