fireice-uk / xmr-stak Goto Github PK
View Code? Open in Web Editor NEWFree Monero RandomX Miner and unified CryptoNight miner
License: GNU General Public License v3.0
Free Monero RandomX Miner and unified CryptoNight miner
License: GNU General Public License v3.0
There is no point in having separate files for config. JSON values should be renamed and integrated as optional (kNullType) values.
cpu_threads_conf
(no rename)
gpu_threads_conf
-> amd_threads_conf
platform_index
-> opencl_platform_index
gpu_threads_conf
-> nvida_threads_conf
Autoconfig should build the config.txt by appending the relevant templates.
I compiled the new 2.0 unified and get the message I am used to on the CPU only version, 'hwloc: memory pinned. But I also get 'MEMORY ALLOC FAILED: VirtualAlloc failed.' and my CPU mining is slow. Confused by getting both messages.
Any help would be much appreciated.
[2017-10-28 22:08:32] : Compiling code and initializing GPUs. This will take a while...
[2017-10-28 22:08:32] : Device 0 work size 8 / 256.
[2017-10-28 22:08:37] : Device 1 work size 8 / 256.
[2017-10-28 22:08:41] : Device 2 work size 8 / 256.
[2017-10-28 22:08:46] : Starting GPU thread, no affinity.
[2017-10-28 22:08:46] : MEMORY ALLOC FAILED: VirtualAlloc failed.
[2017-10-28 22:08:46] : MEMORY ALLOC FAILED: VirtualAlloc failed.
[2017-10-28 22:08:46] : Starting GPU thread, no affinity.
[2017-10-28 22:08:46] : Starting GPU thread, no affinity.
[2017-10-28 22:08:46] : MEMORY ALLOC FAILED: VirtualAlloc failed.
[2017-10-28 22:08:46] : Starting single thread, affinity: 0.
[2017-10-28 22:08:46] : Starting single thread, affinity: 2.
[2017-10-28 22:08:46] : Starting single thread, affinity: 4.
[2017-10-28 22:08:46] : Starting single thread, affinity: 6.
[2017-10-28 22:08:46] : hwloc: memory pinned
[2017-10-28 22:08:46] : hwloc: memory pinned
[2017-10-28 22:08:46] : MEMORY ALLOC FAILED: VirtualAlloc failed.
[2017-10-28 22:08:46] : Connecting to pool pool.supportxmr.com:7777 ...
[2017-10-28 22:08:46] : MEMORY ALLOC FAILED: VirtualAlloc failed.
[2017-10-28 22:08:46] : hwloc: memory pinned
[2017-10-28 22:08:46] : MEMORY ALLOC FAILED: VirtualAlloc failed.
[2017-10-28 22:08:46] : hwloc: memory pinned
[2017-10-28 22:08:46] : MEMORY ALLOC FAILED: VirtualAlloc failed.
[2017-10-28 22:08:46] : Connected. Logging in...
[2017-10-28 22:08:46] : Difficulty changed. Now: 25000.
[2017-10-28 22:08:46] : New block detected.
It seems to be failing building the back-end DLLs. I am using the same process as before without the AMD SDK. I have tried CUDA8 & 9. I can compile the previous build successfully.
Am I doing something incorrectly?
This is how I would see the current xmr-stak-cpu, xmr-stak-amd and xmr-stak-nvidia repositories.
This allows are a couple of critical things:
Flavours as I see them are:
List I (pick one or more):
List II (pick one)
So eventually we will be able to add xmr-stak-cpu-gui simply as a another wrapper around xmr-stak and push the pre-built binaries there.
This is a fairly critical idea going forward so please confirm if you grok and agree.
During the refactoring I have remove the nicehash pool support, I added it back with 1f40f88.
I have not tested if it runs without issues, therefore a runtime test is needed.
Hey people,
I see that fireice merged a new dev-version (#77). but I do not unterstand GIT. how can I download this new src-bundle?
Sorry, I am new at GIT.
Network errors get counted twice after the first error.
After the first error each followed error is counted as two errors.
This is can be reproduced by disconnect the system from the network during the mining. After the first error is triggert plug the system in again and after the miner get a new job unplug the miner again. Than if the next share is found the miner is counting the error as two connection errors.
Hi !
First, uber nice work on this all-in-one refactoring, thanks
I gave it a shot yesterday on debian sid + cuda 9.0 and hashrates seems more stable than 5 separate miners !
Machine A :
Machine B (laptop):
I had to adapt the autosuggest on the first machine for the 1060, first it gave :
{ "index" : 0,
"threads" : 52, "blocks" : 27,
"bfactor" : 0, "bsleep" : 0,
"affine_to_cpu" : false,
},
but this was giving ~10% rejected shares ( at 520/Hs ), dividing by 2
{ "index" : 0,
"threads" : 26, "blocks" : 27,
"bfactor" : 0, "bsleep" : 0,
"affine_to_cpu" : false,
},
made all the rejects go away and we're now at 100% accepted for 12h :)
Hope it helps !
With xmr-stak release proxies and pools will loose the ability to differentiate between cpu, amd and nvidia miners on the same machine. This can cause problems for user interfaces. As such we would like to implement extended statistics mode.
To signal its readiness to accept extended statistics the pool / proxy should add following field to the login reply
{"id" : 0, "jsonrpc" : "2.0", "error" : null, "result" : {"id" : "...", "job" :
{"blob" : "...", "job_id" : "...", "target" : "..."}, "extensions" : ["backend", "hashcount", "algo", "motd"], "status" : "OK"}}
If the server specifies motd extension, the miner will store a message of the day from the pool operator for display with the hashrate results, unless disabled in the config. Message of the day needs only to be sent once per miner connection, unless it changes.
The field needs to be hex-encoded and optionally added to the job dictionary (either in the login or job packet) as such:
{"blob" : "...", "job_id" : "...", "target" : "...", "motd" : "596F757220706F6F6C206E6565647320796F7521" }
When the miner submits a result, it will add following fields:
{"method" : "submit", "params" : {"id" : "...", "job_id" : "...", "nonce" : "...", "result" : "...", "backend" : "amd", "hashcount" : 15550, "algo" : "cryptonight" }, "id" : 1}
backend is a JSON string value, which can be "cpu", "amd" or "nvidia" depending on which backend generated the result
hashcount is a JSON integer value. This is the total number of hashes that the backend in question has done so far. Please note, this number always increases and is not reset on pool switches or disconnects. To calculate a fairly accurate hashrate estimate from pool code you can divide the difference of two hashcounts by the time between them.
algo is a JSON string specifying the algorithm used to generate the hash ( currently "cryptonight", "cryptonight-lite" ).
Hi,
I have a pc with dual GPU and dual CPU, the GPUs are the same, but they are performing very different.
Here the hashrate:
HASHRATE REPORT
| ID | 10s | 60s | 15m | ID | 10s | 60s | 15m |
| 0 | 287.7 | 282.8 | 282.8 | 1 | 393.2 | 393.5 | 393.6 |
| 2 | 38.5 | 38.6 | 38.6 | 3 | 33.8 | 33.8 | 33.8 |
| 4 | 41.2 | 41.2 | 41.2 | 5 | 41.8 | 41.8 | 41.8 |
| 6 | 45.1 | 45.1 | 45.1 | 7 | 41.0 | 41.0 | 41.0 |
| 8 | 34.3 | 34.4 | 34.3 | 9 | 38.9 | 39.0 | 38.9 |
| 10 | 33.9 | 33.9 | 33.9 | 11 | 35.6 | 35.6 | 35.6 |
| 12 | 44.6 | 44.6 | 44.6 | 13 | 41.3 | 41.3 | 41.3 |
| 14 | 44.8 | 44.8 | 44.8 | 15 | 41.1 | 41.0 | 41.0 |
| 16 | 38.9 | 38.9 | 38.9 | 17 | 34.3 | 34.2 | 34.2 |
-----------------------------------------------------
Here the nvidia.txt:
"gpu_threads_conf" :
[
// gpu: Tesla K20c architecture: 35
// memory: 4669/4742 MiB
{ "index" : 0,
"threads" : 26, "blocks" : 39,
"bfactor" : 0, "bsleep" : 0,
"affine_to_cpu" : true,
},
// gpu: Tesla K20c architecture: 35
// memory: 4669/4742 MiB
{ "index" : 2,
"threads" : 26, "blocks" : 39,
"bfactor" : 0, "bsleep" : 0,
"affine_to_cpu" : true,
},
],
Here the nvidia-smi output:
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 384.81 Driver Version: 384.81 |
|-------------------------------+----------------------+----------------------+
| GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. |
|===============================+======================+======================|
| 0 Quadro K600 Off | 00000000:03:00.0 On | N/A |
| 27% 55C P8 N/A / N/A | 49MiB / 978MiB | 0% Default |
+-------------------------------+----------------------+----------------------+
| 1 Tesla K20c Off | 00000000:04:00.0 Off | 0 |
| 45% 58C P0 107W / 225W | 2104MiB / 4742MiB | 100% Default |
+-------------------------------+----------------------+----------------------+
| 2 Tesla K20c Off | 00000000:81:00.0 Off | 0 |
| 62% 66C P0 115W / 225W | 2104MiB / 4742MiB | 100% Default |
+-------------------------------+----------------------+----------------------+
+-----------------------------------------------------------------------------+
| Processes: GPU Memory |
| GPU PID Type Process name Usage |
|=============================================================================|
| 0 1800 G /usr/bin/X 37MiB |
| 1 8254 C ./xmr-stak 2093MiB |
| 2 8254 C ./xmr-stak 2093MiB |
+-----------------------------------------------------------------------------+
Can you explain why?
Thanks
This text is copied out of #28 (review)
The compile still fails, but we are making progress. I googled for a fix and adding add_definitions(-D_MWAITXINTRIN_H_INCLUDED)
seems to fix the problem.
/usr/lib/gcc/x86_64-linux-gnu/5/include/mwaitxintrin.h(36): error: identifier "__builtin_ia32_monitorx" is undefined
/usr/lib/gcc/x86_64-linux-gnu/5/include/mwaitxintrin.h(42): error: identifier "__builtin_ia32_mwaitx" is undefined
2 errors detected in the compilation of "/tmp/tmpxft_00007a1d_00000000-22_cuda_extra.compute_20.cpp1.ii".
CMake Error at xmrstak_cuda_backend_generated_cuda_extra.cu.o.cmake:266 (message):
Error generating file
/home/fireice/github/xmr-stak/build/CMakeFiles/xmrstak_cuda_backend.dir/xmrstak/backend/nvidia/nvcc_code/./xmrstak_cuda_backend_generated_cuda_extra.cu.o
Can you make sure if that flag needs to be targeted at GCC or not.
added by psychocrypt: This error will only occur if the unsupported compiler gcc 5.X is enabled by manually editing of the file host_config.h
All old miner versions contains a self test of the cpu miner. The code is currently not called and returned always true.
Hi,
Thank your for developing this great xmr miner!
After successful build on Linux, I would like to build on my MacBook Pro Retina 2012 (with Nvidia inside).
My laptop begins to be to old to support CUDA 9, so I use the latest CUDA 8 with Xcode 8.2
sudo xcode-select -s /Applications/Xcode82.app/Contents/Developer/
cc --version
Apple LLVM version 8.0.0 (clang-800.0.42.1)
Target: x86_64-apple-darwin17.0.0
Thread model: posix
InstalledDir: /Applications/Xcode82.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin
I choose to use the cmake options in the .travis.yml for Mac :
cmake -DMICROHTTPD_ENABLE=OFF -DOPENSSL_ROOT_DIR=/usr/local/opt/openssl -DOpenCL_ENABLE=OFF ..
in xmr-stak/build folder.
No issue when configuring the build with cmake, then after make clean and make, I've got this issue:
Scanning dependencies of target xmr-stak-c
[ 3%] Building C object CMakeFiles/xmr-stak-c.dir/xmrstak/backend/cpu/crypto/c_blake256.c.o
[ 7%] Building C object CMakeFiles/xmr-stak-c.dir/xmrstak/backend/cpu/crypto/c_groestl.c.o
[ 11%] Building C object CMakeFiles/xmr-stak-c.dir/xmrstak/backend/cpu/crypto/c_jh.c.o
[ 14%] Building C object CMakeFiles/xmr-stak-c.dir/xmrstak/backend/cpu/crypto/c_keccak.c.o
[ 18%] Building C object CMakeFiles/xmr-stak-c.dir/xmrstak/backend/cpu/crypto/c_skein.c.o
[ 22%] Linking C static library bin/libxmr-stak-c.a
[ 22%] Built target xmr-stak-c
Scanning dependencies of target xmr-stak-backend
[ 25%] Building CXX object CMakeFiles/xmr-stak-backend.dir/xmrstak/jconf.cpp.o
[ 29%] Building CXX object CMakeFiles/xmr-stak-backend.dir/xmrstak/backend/cpu/jconf.cpp.o
[ 33%] Building CXX object CMakeFiles/xmr-stak-backend.dir/xmrstak/backend/cpu/minethd.cpp.o
[ 37%] Building CXX object CMakeFiles/xmr-stak-backend.dir/xmrstak/backend/backendConnector.cpp.o
[ 40%] Building CXX object CMakeFiles/xmr-stak-backend.dir/xmrstak/backend/globalStates.cpp.o
[ 44%] Building CXX object CMakeFiles/xmr-stak-backend.dir/xmrstak/backend/cpu/crypto/cryptonight_common.cpp.o
[ 48%] Building CXX object CMakeFiles/xmr-stak-backend.dir/xmrstak/http/httpd.cpp.o
[ 51%] Building CXX object CMakeFiles/xmr-stak-backend.dir/xmrstak/http/webdesign.cpp.o
[ 55%] Building CXX object CMakeFiles/xmr-stak-backend.dir/xmrstak/misc/console.cpp.o
[ 59%] Building CXX object CMakeFiles/xmr-stak-backend.dir/xmrstak/misc/executor.cpp.o
[ 62%] Building CXX object CMakeFiles/xmr-stak-backend.dir/xmrstak/misc/telemetry.cpp.o
[ 66%] Building CXX object CMakeFiles/xmr-stak-backend.dir/xmrstak/net/jpsock.cpp.o
[ 70%] Building CXX object CMakeFiles/xmr-stak-backend.dir/xmrstak/net/socket.cpp.o
[ 74%] Linking CXX static library bin/libxmr-stak-backend.a
/Applications/Xcode82.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/ranlib: file: bin/libxmr-stak-backend.a(httpd.cpp.o) has no symbols
/Applications/Xcode82.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/ranlib: file: bin/libxmr-stak-backend.a(httpd.cpp.o) has no symbols
[ 74%] Built target xmr-stak-backend
Scanning dependencies of target xmr-stak
[ 77%] Building CXX object CMakeFiles/xmr-stak.dir/xmrstak/cli/cli-miner.cpp.o
[ 81%] Linking CXX executable bin/xmr-stak
[ 81%] Built target xmr-stak
[ 85%] Building NVCC (Device) object CMakeFiles/xmrstak_cuda_backend.dir/xmrstak/backend/nvidia/nvcc_code/xmrstak_cuda_backend_generated_cuda_extra.cu.o
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
[ 88%] Building NVCC (Device) object CMakeFiles/xmrstak_cuda_backend.dir/xmrstak/backend/nvidia/nvcc_code/xmrstak_cuda_backend_generated_cuda_core.cu.o
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
Scanning dependencies of target xmrstak_cuda_backend
[ 92%] Building CXX object CMakeFiles/xmrstak_cuda_backend.dir/xmrstak/backend/nvidia/jconf.cpp.o
[ 96%] Building CXX object CMakeFiles/xmrstak_cuda_backend.dir/xmrstak/backend/nvidia/minethd.cpp.o
[100%] Linking CXX shared library bin/libxmrstak_cuda_backend.dylib
Undefined symbols for architecture x86_64:
"win_exit()", referenced from:
xmrstak::nvidia::minethd::thread_starter(unsigned int, xmrstak::miner_work&) in minethd.cpp.o
"jconf::jconf()", referenced from:
xmrstak::nvidia::minethd::work_main() in minethd.cpp.o
xmrstak::nvidia::minethd::self_test() in minethd.cpp.o
"printer::print_msg(verbosity, char const*, ...)", referenced from:
xmrstak::nvidia::jconf::parse_config(char const*) in jconf.cpp.o
xmrstak::nvidia::minethd::work_main() in minethd.cpp.o
xmrstak::nvidia::minethd::thread_starter(unsigned int, xmrstak::miner_work&) in minethd.cpp.o
xmrstak::nvidia::autoAdjust::printConfig() in minethd.cpp.o
xmrstak::nvidia::autoAdjust::generateThreadConfig() in minethd.cpp.o
"printer::printer()", referenced from:
xmrstak::nvidia::jconf::parse_config(char const*) in jconf.cpp.o
xmrstak::nvidia::minethd::thread_starter(unsigned int, xmrstak::miner_work&) in minethd.cpp.o
xmrstak::nvidia::autoAdjust::printConfig() in minethd.cpp.o
printer::inst() in minethd.cpp.o
xmrstak::nvidia::autoAdjust::generateThreadConfig() in minethd.cpp.o
"xmrstak::cpu::minethd::func_selector(bool, bool)", referenced from:
xmrstak::nvidia::minethd::work_main() in minethd.cpp.o
"xmrstak::cpu::minethd::thd_setaffinity(_opaque_pthread_t*, unsigned long long)", referenced from:
xmrstak::nvidia::minethd::thread_starter(unsigned int, xmrstak::miner_work&) in minethd.cpp.o
"xmrstak::cpu::minethd::minethd_alloc_ctx()", referenced from:
xmrstak::nvidia::minethd::work_main() in minethd.cpp.o
"executor::log_result_error(std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> >&&)", referenced from:
xmrstak::nvidia::minethd::work_main() in minethd.cpp.o
"executor::executor()", referenced from:
xmrstak::nvidia::minethd::work_main() in minethd.cpp.o
ld: symbol(s) not found for architecture x86_64
clang: error: linker command failed with exit code 1 (use -v to see invocation)
make[2]: *** [bin/libxmrstak_cuda_backend.dylib] Error 1
make[1]: *** [CMakeFiles/xmrstak_cuda_backend.dir/all] Error 2
make: *** [all] Error 2
I don't know how to resolve this issue ...
Thanks in advance for your help! 👍
I tried to compile on Windows 10 with Visual Studio 2017 and Cuda 9.0, and i am getting the following error:
Building NVCC (Device) object CMakeFiles/xmrstak_cuda_backend.dir/xmrstak/backend/nvidia/nvcc_code/Release/xmrstak_cu da_backend_generated_cuda_core.cu.obj CMake Error at xmrstak_cuda_backend_generated_cuda_core.cu.obj.Release.cmake:219 (message): Error generating C:/Users/Antreas/xmr-stak/CMakeFiles/xmrstak_cuda_backend.dir/xmrstak/backend/nvidia/nvcc_code/Release/xmrstak_cuda _backend_generated_cuda_core.cu.obj
The full logs running the CMake generation and compiling is in the attached file.
log.txt
At some point GCC optimisation flags got removed. Can you fix that?
For the release build the proper settings should be:
-O3 and NDEBUG defined
for the debug build
-O0 and -g
I tested the thread pinning under linux. Something went sometimes wrong.
Normally I use thread 0 and 1 on my laptop. To get a higher chance to trigger the bug I configuread it to 0, 3.
Than I started the miner and looked to htop
. In 20 -50% the second thread is going to core 1 instead of core 3.
It looks like a race condition somewhere.
I checked if hwloc is a problem but also without hwloc I get this behavior.
I also tried to pin the thread with hwloc but with the same result.
Compiling with options:
cmake -G "Visual Studio 15 2017 Win64" -T v140,host=x64 -DXMR-STAK_CURRENCY=monero -DMICROHTTPD_ENABLE=OFF -WIN_UAC=OFF ..
But binary still triggers UAC.
One of my test produced a race condition. The nvidia config was written to cpu.txt
instead to nvidia.txt
I used #55 rebased to the dev 6701b0c.
I can't reproduce it (but this is the nature of a race condition), it could be that #44 is introducing this race condition.
I will investigate some time to this soon.
[2017-10-17 21:46:14] : NVIDIA: GPU configuration stored in file 'cpu.txt'
[2017-10-17 21:46:14] : Starting GPU thread, no affinity.
[2017-10-17 21:46:14] : Starting GPU thread, no affinity.
...
[2017-10-17 21:46:14] : MEMORY ALLOC FAILED: mmap failed
[2017-10-17 21:46:15] : WARNING: No AMD OpenCL platform found. Possible driver issues or wrong vendor driver.
[2017-10-17 21:46:15] : WARNING: backend AMD disabled.
[2017-10-17 21:46:15] : Invalid config file. Missing value "cpu_threads_conf".
[CUDA] Error gpu [CUDA] Error gpu 2: </xmr-stak-nvidia/xmrstak/backend/nvidia/nvcc_code/cuda_extra.cu>:1920: </xmr-stak-nvidia/xmrstak/backend/nvidia/nvcc_code/cuda_extra.cu>:192[CUDA] Error gpu 4: <xmr-stak-nvidia/xmrstak/backend/nvidia/nvcc_code/cuda_extra.cu>:192
cc-ing: @fireice-uk
First of all you guys are doing an awesome job, making possible for all of us to mine with different platforms.
I'm getting the following message:
Compiling code and initializing GPUs. This will take a while...
Error CL_DEVICE_NOT_FOUND when calling clGetDeviceIDs for number of devices.
WARNING: AMD device not found
System: 2 x XEON L5630 12MB L3 24GB RAM HP server DL380 G6 Firepro V5800 installed as add on card with motherboard integrated graphics disabled on BIOS with OpenSUSE LEAP 42.1 running on metal.
I'm using FGLRX drivers since I couldn't install the AMDPRO in any opensuse version (42.1 is not compatible) and (42.2/ 42.3 the system freezes after install. First the Xorg and a few seconds after the entire system).
libOpenCL1 2.2.11-2.1 (from opensuse)
opencl-headers 2.2+git.20170617-1.2 (from opensuse)
fglrx64 15.302.3-1 (http://geeko.ioda.net/mirror/amd-fglrx/openSUSE_Leap_42.1/)
xmr-stak compiled with GCC 5.3.1, cmake 3.3.2-1.2 both from SUSE repo's
lspci -nnk | grep -A3 VGA report:
10:00.0 VGA compatible controller [0300]: Advanced Micro Devices, Inc. [AMD/ATI] Juniper XT [FirePro V5800] [1002:68a9]
Subsystem: Advanced Micro Devices, Inc. [AMD/ATI] Device [1002:2303]
Kernel driver in use: fglrx_pci
Kernel modules: radeon, fglrx
clinfo report attached:
clinfo-report.txt
During the compiling process I can see:
-- Looking for CL_VERSION_2_0
-- Looking for CL_VERSION_2_0 - found
-- Found OpenCL: /usr/lib64/libOpenCL.so (found version "2.0")
The thing is: I installed the opensuse normally and after I did the installation of libOpenCL1 and opencl-headers. Then I installed the fglrx driver that seems to overwrite the libOpenCL1, and maybe this is promoting the error.
If I remove the libOpenCL1 and headers and install only the fglrx drivers, the xmr-stak cannot compile because it needs the CL/cl.h file included in the package libOpenCL1 from opensuse....since I know, fglrx doesn't install CL/cl.h or am I missing something ?
PS: this is happing with xmr-stak-amd too !
Any ideas ?
Thanks a lot !!!
For xmr-stak release I would like to have a pre-built Linux binary. @psychocrypt can you look into this one? I'm fairly busy with multipool and protocol extentions.
Suggested option: AppImage
Is there a plan to add CryptoNight-Lite support in this miner so it can be used for AEON mining?
Another useful feature for CryptoNight-Lite will be quad thread support (4 hashes per thread) in low power mode (but also keep double thread mode for maximum flexibility).
Hi, having issue on card #12 when using 12 GPU rig.
Totals: 5020.6 5018.8 5017.6 H/s
Highest: 5021.7 H/s
Using Ubuntu 16.04 with current release xmr-stak. Have no issues on xmr-stak-amd current release.
uname -a Linux mm-miner01 4.4.0-96-generic #119-Ubuntu SMP Tue Sep 12 14:59:54 UTC 2017 x86_64 x86_64 x86_64 GNU/Linux
Thanks
When xmr-stak prompt UAC dialog to ask you, if you select yes, new process will be created and it will ignore all command line parameter.
Beside that, UAC handle is very annoying with who having correct setting about large page (you can even use large page when answer no), it's should have a switch or config option to enable or disable it. If it cannot turn on/off by switch/option please leave it as xmr-stak-cpu
The CPU backend is loaded as last backend. Each backend is provided with a thread offset (number of threads from other backends started before the backend).
If the thread offset is not zero than we have GPUs in the system, in this case the CPU auto suggestion should prefer low power mode
over hyper threads to keep cores free to handle the gpus.
The large page currently not working, it always show "MEMORY ALLOC FAILED: VirtualAlloc falled.", the older version work fine.
Normally with old version I can get about 90H/s each double thread (lower_power_mode = true - 2 x CPU E5-2670v2),
but for universal version it only get about 50H/s, and it's say: MEMORY ALLOC FAILED: VirtualAlloc falled.
Heyho,
before I mined Monero with xmr-stak-cpu and xmr-stak-nvidia. Now i tried xmr-stak. The compiling went like a charm (except i have not the AMD sources whats OK because I only use NVIDIA)
System: i5-4 Gen. (4 Core no HT) with 4x GTX 1060 3GB
Then I start the software and this happends:
Please enter:
- currency: 'monero' or 'aeon'
monero
- pool address: e.g. pool.usxmrpool.com:3333
pool.supportxmr.com:5555
- user name (wallet address or pool login):
[...]BMYCBFbLU52L[...] # shorten
- password (mostly empty or x):
mpm01:<my email>
Configuration stored in file 'config.txt'
-------------------------------------------------------------------
xmr-stak 2.0.0-predev mining software.
Based on CPU mining code by wolf9466 (heavily optimized by fireice_uk).
NVIDIA mining code was written by KlausT and psychocrypt.
Brought to you by fireice_uk and psychocrypt under GPLv3.
Configurable dev donation level is set to 1.0 %
You can use following keys to display reports:
'h' - hashrate
'r' - results
'c' - connection
-------------------------------------------------------------------
[2017-10-29 01:01:17] : Start mining: MONERO
[2017-10-29 01:01:17] : NVIDIA: GPU configuration stored in file 'nvidia.txt'
[2017-10-29 01:01:17] : Starting NVIDIA GPU thread 0, no affinity.
[2017-10-29 01:01:17] : Starting NVIDIA GPU thread 1, no affinity.
[2017-10-29 01:01:17] : Starting NVIDIA GPU thread 2, no affinity.
[2017-10-29 01:01:17] : Starting NVIDIA GPU thread 3, no affinity.
[2017-10-29 01:01:17] : CPU configuration stored in file 'cpu.txt'
[2017-10-29 01:01:17] : Starting single thread, affinity: 0.
[2017-10-29 01:01:17] : hwloc: memory pinned
[2017-10-29 01:01:17] : Starting single thread, affinity: 1.
[2017-10-29 01:01:17] : hwloc: memory pinned
[2017-10-29 01:01:17] : Starting single thread, affinity: 2.
[2017-10-29 01:01:17] : hwloc: memory pinned
[2017-10-29 01:01:17] : Connecting to pool pool.supportxmr.com:5555 ...
[2017-10-29 01:01:17] : Connected. Logging in...
[2017-10-29 01:01:17] : Difficulty changed. Now: 5000.
[2017-10-29 01:01:17] : Pool switched.
./xmr-stak: symbol lookup error: ./libxmrstak_cuda_backend.so: undefined symbol: _Z25cryptonight_core_cpu_hashILm524288ELm2097136ELj19EEvP8nvid_ctx
root@mpm01:/work/Monero/xmr-stak#
Annd.. Will the other projects be continued, or is this the leading project?
The low power mode (or double hash) calculation is broken since #19. If I enable it the miner will hang at the beginning. The commit 9c3a71e is working.
CC-ing: @fireice-uk
Hey,
I compiled the xmr-stak version for windows and tested it.
My mainworkingstation (Testsystem):
Windows 10 Pro 64 Bit
Intel i7-4790K 4Ghz [email protected] Ghz
Nvidia Geforce GTX 980 4GB
configs (generated automaticly):
"cpu_threads_conf" :
[
{ "low_power_mode" : false, "no_prefetch" : true, "affine_to_cpu" : 0 },
{ "low_power_mode" : false, "no_prefetch" : true, "affine_to_cpu" : 2 },
{ "low_power_mode" : false, "no_prefetch" : true, "affine_to_cpu" : 4 },
{ "low_power_mode" : false, "no_prefetch" : true, "affine_to_cpu" : 6 },
],
"gpu_threads_conf" :
[
// gpu: GeForce GTX 980 architecture: 52
// memory: 3376/4096 MiB
{ "index" : 0,
"threads" : 20, "blocks" : 48,
"bfactor" : 6, "bsleep" : 25,
"affine_to_cpu" : false,
},
],
I changed nvidia config like this:
bfactor: 8
bsleep: 100
result: program runs but the System was very laggy.
I changed nvidia config again like this:
bfactor: 12
bsleep: 100
result: program runs and system is only a little bit laggy (an vlc-video runs after 2 sec smootly)
after 15 minutes:
HASHRATE REPORT
| ID | 10s | 60s | 15m | ID | 10s | 60s | 15m |
| 0 | (na) | 84.3 | 84.8 | 1 | 31.1 | 31.5 | 32.9 |
| 2 | 38.9 | 37.2 | 37.9 | 3 | 38.2 | 36.3 | 36.1 |
| 4 | 21.8 | 22.0 | 22.9 |
-----------------------------------------------------
Totals: (na) 211.2 214.5 H/s
Highest: 0.0 H/s
So i tweaked the config a little bit:
"cpu_threads_conf" :
[
{ "low_power_mode" : false, "no_prefetch" : true, "affine_to_cpu" : 2 },
{ "low_power_mode" : false, "no_prefetch" : true, "affine_to_cpu" : 3 },
{ "low_power_mode" : false, "no_prefetch" : true, "affine_to_cpu" : 4 },
{ "low_power_mode" : false, "no_prefetch" : true, "affine_to_cpu" : 5 },
{ "low_power_mode" : false, "no_prefetch" : true, "affine_to_cpu" : 6 },
{ "low_power_mode" : false, "no_prefetch" : true, "affine_to_cpu" : 7 },
],
// 2,4,6 are real cores
// 3,5,7 are HT cores
"gpu_threads_conf" :
[
{ "index" : 0, "threads" : 32, "blocks" : 16, "bfactor" : 12, "bsleep" : 5, "affine_to_cpu" : true},
],
result:
HASHRATE REPORT
| ID | 10s | 60s | 15m | ID | 10s | 60s | 15m |
| 0 | 467.4 | 461.6 | 459.6 | 1 | 20.4 | 19.4 | 19.1 |
| 2 | 25.2 | 25.2 | 25.8 | 3 | 12.5 | 15.2 | 13.9 |
| 4 | 23.1 | 22.4 | 23.0 | 5 | 16.3 | 16.3 | 17.1 |
| 6 | 33.5 | 30.3 | 31.3 |
-----------------------------------------------------
Totals: 598.3 590.4 589.7 H/s
Highest: 642.1 H/s
here is how i find the best values of my cards:
xmr-stat-gpu values.xlsx
Conclusion:
It would be usefull to allow that a list of pools with there login data (username, pool passwd, address) can be configured.
A json array like for cpu_thread_conf
would be the best.
If we add a value
to each pool entry that can be set by the user for each pool and a additional flag ( a new variable those is not part of the pool array) what the value means we can allow two cases:
100min - dev pool time
. Each pool is than used for mining depending of the time slice within e.g. 98min. This allows for example the user to provide other developer (e.g. monero devs) with a donation, or the user likes to split the hash power to different pools)/* This defines in which order the pools are use:
* "failover" - means the value defines the priority of the pool (higher is used first),
* if the connection get lost the pool is switched to the next
* (after X min the miner tries to fall back to the original pool)
* "weighted" - all pool values will be accumulated and the hash rate
* is given to all pools (within 100 minutes) depending on the weight
*/
"pool_order" : "weighted",
"pool_list" :
[
// if the dev donation is 2%, this pool is used 42 minute
{ "pool_address" : "pool1", "wallet_address" : "myXMRAddress",
"pool_password" : "x", "value" : 60 },
// this pool is used 56 minutes
{ "pool_address" : "pool2", "wallet_address" : "myXMRAddress",
"pool_password" : "", "value" : 80 },
],
Like in PR #24 early returns within a kernel are not defined for CUDA. To avoid side effects with other compiler e.g. Clang all early return should be removed.
This will not increase the runtime!
CMake output doesn't seem to mention CUDA:
-- The C compiler identification is GNU 5.4.0
-- The CXX compiler identification is GNU 5.4.0
-- Check for working C compiler: /usr/bin/cc
-- Check for working C compiler: /usr/bin/cc -- 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/bin/c++
-- Check for working CXX compiler: /usr/bin/c++ -- works
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Detecting CXX compile features
-- Detecting CXX compile features - done
-- 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 "1.0.2g")
-- Configuring done
-- Generating done
-- Build files have been written to: /home/fireice/github/xmr-stak/build
Build then fails with:
nvcc fatal : Unsupported gpu architecture 'compute_60'
CMake Error at xmrstak_cuda_backend_generated_cuda_core.cu.o.cmake:207 (message):
Error generating
/home/fireice/github/xmr-stak/build/CMakeFiles/xmrstak_cuda_backend.dir/xmrstak/backend/nvidia/nvcc_code/./xmrstak_cuda_backend_generated_cuda_core.cu.o
CMakeFiles/xmrstak_cuda_backend.dir/build.make:70: recipe for target 'CMakeFiles/xmrstak_cuda_backend.dir/xmrstak/backend/nvidia/nvcc_code/xmrstak_cuda_backend_generated_cuda_core.cu.o' failed
make[2]: *** [CMakeFiles/xmrstak_cuda_backend.dir/xmrstak/backend/nvidia/nvcc_code/xmrstak_cuda_backend_generated_cuda_core.cu.o] Error 1
CMakeFiles/Makefile2:67: recipe for target 'CMakeFiles/xmrstak_cuda_backend.dir/all' failed
make[1]: *** [CMakeFiles/xmrstak_cuda_backend.dir/all] Error 2
Makefile:127: recipe for target 'all' failed
make: *** [all] Error 2
CMake output from xmr-stak-nvidia:
-- The C compiler identification is GNU 5.4.0
-- The CXX compiler identification is GNU 5.4.0
-- Check for working C compiler: /usr/bin/cc
-- Check for working C compiler: /usr/bin/cc -- 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/bin/c++
-- Check for working CXX compiler: /usr/bin/c++ -- works
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Detecting CXX compile features
-- Detecting CXX compile features - done
-- 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 CUDA: /usr (found suitable version "7.5", minimum required is "7.5")
-- Found OpenSSL: /usr/lib/x86_64-linux-gnu/libssl.so;/usr/lib/x86_64-linux-gnu/libcrypto.so (found version "1.0.2g")
-- Configuring done
-- Generating done
-- Build files have been written to: /home/fireice/github/xmr-stak-nvidia/build
When I have only AMD SDK installed and disabled NVIDIA backend or compile XMR-STAK-AMD it has no problem detect OpenCL 2.0, but with CUDA 9 installed it only find 1.2 version.
-- Looking for CL_VERSION_2_0
-- Looking for CL_VERSION_2_0 - not found
-- Looking for CL_VERSION_1_2
-- Looking for CL_VERSION_1_2 - found
-- Found OpenCL: C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v9.0/lib/x64/OpenCL.lib (found version "1.2")
-- Found CUDA: C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v9.0 (found suitable version "9.0", minimum required is "7.5")
Hi,
you wrote in the cpu.txt that on hyperthreading systems it is better to assign threads to physical cores, but I have a dual XEON wiht 10 physical cores each for a total of 40 logical cores and the autoconfiguratior gave me this:
"cpu_threads_conf" :
[
{ "low_power_mode" : false, "no_prefetch" : true, "affine_to_cpu" : 0 },
{ "low_power_mode" : false, "no_prefetch" : true, "affine_to_cpu" : 1 },
{ "low_power_mode" : false, "no_prefetch" : true, "affine_to_cpu" : 2 },
{ "low_power_mode" : false, "no_prefetch" : true, "affine_to_cpu" : 3 },
{ "low_power_mode" : false, "no_prefetch" : true, "affine_to_cpu" : 4 },
{ "low_power_mode" : false, "no_prefetch" : true, "affine_to_cpu" : 5 },
{ "low_power_mode" : false, "no_prefetch" : true, "affine_to_cpu" : 6 },
{ "low_power_mode" : false, "no_prefetch" : true, "affine_to_cpu" : 7 },
{ "low_power_mode" : false, "no_prefetch" : true, "affine_to_cpu" : 8 },
{ "low_power_mode" : false, "no_prefetch" : true, "affine_to_cpu" : 9 },
{ "low_power_mode" : false, "no_prefetch" : true, "affine_to_cpu" : 20 },
{ "low_power_mode" : false, "no_prefetch" : true, "affine_to_cpu" : 21 },
{ "low_power_mode" : false, "no_prefetch" : true, "affine_to_cpu" : 22 },
{ "low_power_mode" : false, "no_prefetch" : true, "affine_to_cpu" : 10 },
{ "low_power_mode" : false, "no_prefetch" : true, "affine_to_cpu" : 11 },
{ "low_power_mode" : false, "no_prefetch" : true, "affine_to_cpu" : 12 },
{ "low_power_mode" : false, "no_prefetch" : true, "affine_to_cpu" : 13 },
{ "low_power_mode" : false, "no_prefetch" : true, "affine_to_cpu" : 14 },
{ "low_power_mode" : false, "no_prefetch" : true, "affine_to_cpu" : 15 },
{ "low_power_mode" : false, "no_prefetch" : true, "affine_to_cpu" : 16 },
{ "low_power_mode" : false, "no_prefetch" : true, "affine_to_cpu" : 17 },
{ "low_power_mode" : false, "no_prefetch" : true, "affine_to_cpu" : 18 },
{ "low_power_mode" : false, "no_prefetch" : true, "affine_to_cpu" : 19 },
{ "low_power_mode" : false, "no_prefetch" : true, "affine_to_cpu" : 30 },
{ "low_power_mode" : false, "no_prefetch" : true, "affine_to_cpu" : 31 },
{ "low_power_mode" : false, "no_prefetch" : true, "affine_to_cpu" : 32 },
],
So the miner is using 26 logical cores, 20 physical cores + 6 HTT cores.
Can you explain me this discrepancy?
I think the HTT core IDs are: 20,21,22,30,31,32, am I right?
So I can test which configuration is better.
This is my hwloc-ls:
Machine (128GB)
NUMANode L#0 (P#0 64GB)
Socket L#0 + L3 L#0 (25MB)
L2 L#0 (256KB) + L1d L#0 (32KB) + L1i L#0 (32KB) + Core L#0
PU L#0 (P#0)
PU L#1 (P#20)
L2 L#1 (256KB) + L1d L#1 (32KB) + L1i L#1 (32KB) + Core L#1
PU L#2 (P#1)
PU L#3 (P#21)
L2 L#2 (256KB) + L1d L#2 (32KB) + L1i L#2 (32KB) + Core L#2
PU L#4 (P#2)
PU L#5 (P#22)
L2 L#3 (256KB) + L1d L#3 (32KB) + L1i L#3 (32KB) + Core L#3
PU L#6 (P#3)
PU L#7 (P#23)
L2 L#4 (256KB) + L1d L#4 (32KB) + L1i L#4 (32KB) + Core L#4
PU L#8 (P#4)
PU L#9 (P#24)
L2 L#5 (256KB) + L1d L#5 (32KB) + L1i L#5 (32KB) + Core L#5
PU L#10 (P#5)
PU L#11 (P#25)
L2 L#6 (256KB) + L1d L#6 (32KB) + L1i L#6 (32KB) + Core L#6
PU L#12 (P#6)
PU L#13 (P#26)
L2 L#7 (256KB) + L1d L#7 (32KB) + L1i L#7 (32KB) + Core L#7
PU L#14 (P#7)
PU L#15 (P#27)
L2 L#8 (256KB) + L1d L#8 (32KB) + L1i L#8 (32KB) + Core L#8
PU L#16 (P#8)
PU L#17 (P#28)
L2 L#9 (256KB) + L1d L#9 (32KB) + L1i L#9 (32KB) + Core L#9
PU L#18 (P#9)
PU L#19 (P#29)
NUMANode L#1 (P#1 64GB)
Socket L#1 + L3 L#1 (25MB)
L2 L#10 (256KB) + L1d L#10 (32KB) + L1i L#10 (32KB) + Core L#10
PU L#20 (P#10)
PU L#21 (P#30)
L2 L#11 (256KB) + L1d L#11 (32KB) + L1i L#11 (32KB) + Core L#11
PU L#22 (P#11)
PU L#23 (P#31)
L2 L#12 (256KB) + L1d L#12 (32KB) + L1i L#12 (32KB) + Core L#12
PU L#24 (P#12)
PU L#25 (P#32)
L2 L#13 (256KB) + L1d L#13 (32KB) + L1i L#13 (32KB) + Core L#13
PU L#26 (P#13)
PU L#27 (P#33)
L2 L#14 (256KB) + L1d L#14 (32KB) + L1i L#14 (32KB) + Core L#14
PU L#28 (P#14)
PU L#29 (P#34)
L2 L#15 (256KB) + L1d L#15 (32KB) + L1i L#15 (32KB) + Core L#15
PU L#30 (P#15)
PU L#31 (P#35)
L2 L#16 (256KB) + L1d L#16 (32KB) + L1i L#16 (32KB) + Core L#16
PU L#32 (P#16)
PU L#33 (P#36)
L2 L#17 (256KB) + L1d L#17 (32KB) + L1i L#17 (32KB) + Core L#17
PU L#34 (P#17)
PU L#35 (P#37)
L2 L#18 (256KB) + L1d L#18 (32KB) + L1i L#18 (32KB) + Core L#18
PU L#36 (P#18)
PU L#37 (P#38)
L2 L#19 (256KB) + L1d L#19 (32KB) + L1i L#19 (32KB) + Core L#19
PU L#38 (P#19)
PU L#39 (P#39)
This struct makes singleton pattern objects very confusing. Think of a better way to do it.
Pardon if formatting is messy. I've never used GitHub. All this data is in the attachment,
Hit me up on SupportXMR Chatango with questions or requests.
##########################################
GPU0: EVGA GeForce GTX 980 SC GAMING ACX 2.0
Spec Sheet: https://www.evga.com/Products/Specs/GPU.aspx?pn=472A80BC-9A78-45B1-8560-6FA1916330E8
Overclocked to GPU Freq: 1457MHz, MEM Freq: 4001
My config (Yields about 570H/s):
"gpu_threads_conf" :
[
// gpu: GeForce GTX 980 architecture: 52
// memory: 3376/4096 MiB
{ "index" : 0,
"threads" : 32, "blocks" : 16,
"bfactor" : 8, "bsleep" : 25,
"affine_to_cpu" : false,
},
],
Default Config (Crashes miner software):
"gpu_threads_conf" :
[
// gpu: GeForce GTX 980 architecture: 52
// memory: 3376/4096 MiB
{ "index" : 0,
"threads" : 20, "blocks" : 48,
"bfactor" : 6, "bsleep" : 25,
"affine_to_cpu" : false,
},
],
##########################################
GPU1: SAPPHIRE Radeon™ RX Vega64 8G HBM2
Manufacturers Site: http://www.sapphiretech.com/productdetial.asp?pid=AE163547-6889-4073-BBF6-9DB2C45BB233&lang=eng
If stock clocks are used then best result is with Thread1: "intensity" : 1920, "worksize" : 8, Thread2: "intensity" : 1920, "worksize" : 8 yields about 1950H/s
Underclocked GPU/Overclocked Memeory GPU Freq: 1270MHz, MEM Freq: 1100MHz
My Config (Yields about 1890H/s):
"gpu_threads_conf" : [
// gpu: gfx901 memory:3984
// compute units: 64
{ "index" : 0,
"intensity" : 2016, "worksize" : 8,
"affine_to_cpu" : false,
},
{ "index" : 0,
"intensity" : 1800, "worksize" : 8,
"affine_to_cpu" : false,
},
],
Default configuration (Untested):
"gpu_threads_conf" : [
// gpu: gfx901 memory:3984
// compute units: 64
{ "index" : 0,
"intensity" : 512, "worksize" : 8,
"affine_to_cpu" : false,
},
],
#########################################
CPU: Intel® Core™ i7-5820K Processor
Spec Sheet: https://ark.intel.com/products/82932/Intel-Core-i7-5820K-Processor-15M-Cache-up-to-3_60-GHz
Overclocked CPU Freq: 4.3GHz, Cache Multiplier: 40
RAM: 32GB DDR4
My Config (Yields about 430H/s):
"cpu_threads_conf" :
[
{ "low_power_mode" : false, "no_prefetch" : true, "affine_to_cpu" : 0 },
{ "low_power_mode" : false, "no_prefetch" : true, "affine_to_cpu" : 2 },
{ "low_power_mode" : false, "no_prefetch" : true, "affine_to_cpu" : 4 },
{ "low_power_mode" : false, "no_prefetch" : true, "affine_to_cpu" : 6 },
{ "low_power_mode" : false, "no_prefetch" : true, "affine_to_cpu" : 8 },
{ "low_power_mode" : false, "no_prefetch" : true, "affine_to_cpu" : 10 },
{ "low_power_mode" : false, "no_prefetch" : true, "affine_to_cpu" : 1 },
],
Default Config (Untested):
"cpu_threads_conf" :
[
{ "low_power_mode" : false, "no_prefetch" : true, "affine_to_cpu" : 0 },
{ "low_power_mode" : false, "no_prefetch" : true, "affine_to_cpu" : 2 },
{ "low_power_mode" : false, "no_prefetch" : true, "affine_to_cpu" : 4 },
{ "low_power_mode" : false, "no_prefetch" : true, "affine_to_cpu" : 6 },
{ "low_power_mode" : false, "no_prefetch" : true, "affine_to_cpu" : 8 },
{ "low_power_mode" : false, "no_prefetch" : true, "affine_to_cpu" : 10 },
{ "low_power_mode" : false, "no_prefetch" : true, "affine_to_cpu" : 1 },
{ "low_power_mode" : false, "no_prefetch" : true, "affine_to_cpu" : 3 },
],
#########################################
Currently the hash report shows all backends within a table. By adding a type to each backend (in iBackend.hpp
) we can split the table into a overview per backend.
You have the following text in "Usage.md" but it's the last line in the file and you forgot to actually add your wallet :-)
"If you want to donate directly to support further development, here is my wallet"
The error Error CL_INVALID_WORK_GROUP_SIZE when calling clEnqueueNDRangeKernel for kernel 0.
still exists and was described in issue fireice-uk/xmr-stak-amd#93.
This is a reminder for me, I already found the reason for this bug.
Hello,
I was wondering if someone had an already compiled version for download available?
@psychocrypt can you confirm that the autosuggestion on this backend is good enough to remove the gpu_thread_num safety?
I try to compile the xmr-stak from source in Windows 7 to support CPU, AMD and NVIDIA, but I only found xmr-stak-cpu bin. How to do that?
cmake -G "Visual Studio 15 2017 Win64" -T v140,host=x64 -DCUDA_ENABLE=ON -DOpenCL_ENABLE=ON --DCMAKE_LINK_STATIC=ON ..
Not really an issue, but a request asking you if it would be possible to add a sum of the hashrate by thread_id in the html report please ?
I tried to do it but I couldn't find how/where I have to calculate the sum to print it in webdesign.
We have a large mix of workers now with capabilities ranging in two orders of magnitude from 10H/s to 1000H/s
If we are going to support nicehash then we are working in a very restricted space - we have 16.7 million nonces to split.
Moreover if we stop, and then resume at a pool without dropping connection we need to include that in the nonce budget, otherwise we will get collisions.
The maths just doesn't add up, 8 workers at 2500H/s and 1 resume is 3.6 million hashes. There is no way to that we can fit 64 cpu workers without making the nonce calculation differentiate between cpu and gpu workers
Statistic for gpu threads is randomly show/hide for each threads, it's will change with each refresh (this threads hide, other threads show).
Beside that, when having gpu, total hashrate of 10s is always hide. Highest hastrate alway 0.
On CLI only having zero highest hashrate problem.
Hi,
yesterday I cloned and compiled the miner on linux and it went all good.
Today, on a different but identical pc, I have cloned and compiled the miner and it doesn't work, I made a lot of tries, the only way to make it works was to use the source code cloned yesterday.
The only thing I get is this:
-------------------------------------------------------------------
xmr-stak 2.0.0-predev mining software.
Based on CPU mining code by wolf9466 (heavily optimized by fireice_uk).
Brought to you by fireice_uk and psychocrypt under GPLv3.
Configurable dev donation level is set to 2.0 %
You can use following keys to display reports:
'h' - hashrate
'r' - results
'c' - connection
-------------------------------------------------------------------
[2017-10-18 14:42:43] : hwloc: memory pinned
then it stucks.
With the miner compiled from the old code (using the same config.txt and cpu.txt) I get:
-------------------------------------------------------------------
xmr-stak 2.0.0-predev mining software.
Based on CPU mining code by wolf9466 (heavily optimized by fireice_uk).
Brought to you by fireice_uk and psychocrypt under GPLv3.
Configurable dev donation level is set to 2.0 %
You can use following keys to display reports:
'h' - hashrate
'r' - results
'c' - connection
-------------------------------------------------------------------
[2017-10-18 14:49:22] : Starting double thread, affinity: 0.
[2017-10-18 14:49:22] : Starting single thread, affinity: 1.
[2017-10-18 14:49:22] : Starting double thread, affinity: 2.
[2017-10-18 14:49:22] : Starting single thread, affinity: 3.
[2017-10-18 14:49:22] : Connecting to pool xxxx:yyy ...
[2017-10-18 14:49:22] : Connected. Logging in...
[2017-10-18 14:49:22] : hwloc: memory pinned
[2017-10-18 14:49:22] : hwloc: memory pinned
[2017-10-18 14:49:22] : hwloc: memory pinned
[2017-10-18 14:49:22] : hwloc: memory pinned
[2017-10-18 14:49:22] : Difficulty changed. Now: 120001.
[2017-10-18 14:49:22] : New block detected.
I saw you made some change on the code in the last 24 hours, can you check the new code?
Thanks
The auto suggestion for the AMD backend will add the name of the GPU as comment into the config.
What is the name of the VEGA gpu provided by OpenCL?
Could someone from the communityplease provide us with the config created by the auto suggestion!
This information is needed to improve the auto suggestion for VEGA gpus.
I've been trying to get this to compile on Win10 64bit today, and keep running into errors stating the following;
C:\xmr-stak\xmr-stak\build>cmake -G "Visual Studio 15 2017 Win64" -T v140,host=x64 -DOpenCL_ENABLE=ON ..
-- Selecting Windows SDK version 10.0.16299.0 to target Windows 10.0.16278.
CMake Error at CMakeLists.txt:173 (message):
OpenCL NOT found: use -DOpenCL_ENABLE=OFF
to build without OpenCL support
for AMD gpu's
I'm running the AMD 17.x ReLive series drivers, hoping that's not the issue. I've also installed the 3.x AMD APP SDK thinking I might be missing OpenCL libraries.
I had the same issue running the prebuilt xmr-stak-amd binary on this machine. That said, it's got an HD 7950 that runs the Claymore AMD GPU Monero miner just fine... so I'm at a bit of a loss as to what I'm overlooking or missing for the build process.
@fireice-uk Can we remove
Lines 118 to 125 in 22039b2
Running latest xmr-stak-dev, on linux with rx480( not OC), compile without problems, cpu backend i5 works flawlessly , but having problem with GPU.
After first start, and auto configuration, in amd.txt, suggested configuration is:
"intensity" : 0, "worksize": 8
And if i increase intensity more than 32, get many "AMD invalid result" errors.
Also tried playing with worksize , but no difference.
Sometimes most errors are upon miner start, and sometimes start without errors, and after some time starts throwing errors
RESULT REPORT
Difficulty : 100001
Good results : 316 / 403 (73.8 %)
Avg result time : 100.1 sec
Pool-side hashes : 2958631
Top 10 best results found:
| 0 | 17327791 | 1 | 6890190 |
| 2 | 5178865 | 3 | 4649874 |
| 4 | 4548370 | 5 | 3991239 |
| 6 | 3690780 | 7 | 3579746 |
| 8 | 3028025 | 9 | 2505256 |
Error details:
| Count | Error text | Last seen |
| 87 | AMD Invalid Result | 2017-10-30 07:49:40 |
Note, tried different GPU's, all rx480 and almost new, i took them from my eth rig (works without problems) to test aeon mining. Same configuration both my rig and testing computer, ubuntu/drivers, so GPU should be fine
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.