nvcc fatal : Option ‘–generate-code arch=compute_60’, missing code

[aseidler]

Hello
I am a beginner with OpenLB.

So I managed to get a few things done with the help of the forum and the developer guide to run my simulation. But trying it on a single gpu only is still not possible. I get this error message after trying to compile my code and have no clue to solve this.
So I am happy about any help.

make:
make -C ../../../external
make[1]: Entering directory ‘/home/aseid/olb-1.6r0/olb-1.6r0/external’
make -C zlib
make[2]: Entering directory ‘/home/aseid/olb-1.6r0/olb-1.6r0/external/zlib’
make[2]: Nothing to be done for ‘all’.
make[2]: Leaving directory ‘/home/aseid/olb-1.6r0/olb-1.6r0/external/zlib’
cp zlib/build/libz.a lib/
make -C tinyxml
make[2]: Entering directory ‘/home/aseid/olb-1.6r0/olb-1.6r0/external/tinyxml’
make[2]: Nothing to be done for ‘all’.
make[2]: Leaving directory ‘/home/aseid/olb-1.6r0/olb-1.6r0/external/tinyxml’
cp tinyxml/build/libtinyxml.a lib/
make[1]: Leaving directory ‘/home/aseid/olb-1.6r0/olb-1.6r0/external’
make -C ../../.. core
make[1]: Entering directory ‘/home/aseid/olb-1.6r0/olb-1.6r0’
make[1]: Nothing to be done for ‘core’.
make[1]: Leaving directory ‘/home/aseid/olb-1.6r0/olb-1.6r0’
nvcc -O3 -std=c++17 –forward-unknown-to-host-compiler -pthread –forward-unknown-to-host-compiler -x cu -O3 -std=c++17 –generate-code=arch=compute_60 ,code=[compute_60 ,sm_60 ] –extended-lambda –expt-relaxed-constexpr -rdc=true -Xcudafe “–diag_suppress=implicit_return_from_non_void_function –display_error_number –diag_suppress=20014 –diag_suppress=20011” -DPLATFORM_CPU_SISD -DPLATFORM_GPU_CUDA -DDEFAULT_FLOATING_POINT_TYPE=float -I../../../src -I../../../external/zlib -I../../../external/tinyxml -c -o vpetorustube.o vpetorustube.cpp
nvcc fatal : Option ‘–generate-code arch=compute_60’, missing code
make: *** [../../../default.single.mk:35: vpetorustube.o] Error 1

config.mk:

CXX := nvcc
CC := nvcc

CXXFLAGS := -O3
CXXFLAGS += -std=c++17 –forward-unknown-to-host-compiler

PARALLEL_MODE := NONE

PLATFORMS := CPU_SISD GPU_CUDA

# for e.g. RTX 30* (Ampere), see table in rules.mk for other options
CUDA_ARCH := 62

FLOATING_POINT_TYPE := float

USE_EMBEDDED_DEPENDENCIES := ON

Setup:

I am using a NVIDIA QUADRO P2200

with nvidia-smi:
+—————————————————————————————+
| NVIDIA-SMI 545.23.08 Driver Version: 516.01 CUDA Version: 11.7 |
|—————————————–+———————-+———————-+
| GPU Name Persistence-M | Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap | Memory-Usage | GPU-Util Compute M. |
| | | MIG M. |

nvcc -V:
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2022 NVIDIA Corporation
Built on Wed_Jun__8_16:49:14_PDT_2022
Cuda compilation tools, release 11.7, V11.7.99
Build cuda_11.7.r11.7/compiler.31442593_0

Any suggestions and help will be greatly appreciated

-Alex

[Adrian]

The error is for a different config – your listing states CUDA_ARCH as 62 whereas the error mentions 60. Is this the only change? If so, the issue is that you have a space after the cuda arch number that doesn’t get sanitized (will be fixed in 1.7).

[aseidler]

Hello Adrian,

Thank you for the quick reply.

Indeed the space in the CUDA_ARCH definition was the gist of the matter.
Still, my code isn’t running but the examples like venturi3d are now running on the GPU.

All the best,
Alex

[Adrian]

Happy to hear that the examples are working now.

What is the issue with your case? Does it work on the CPU?

[aseidler]

Hey Adrian,

i checked my code in respect to the examples. But still I got this error where I am a bit clueless.

[ThreadPool] Sucessfully initialized, numThreads=1
[main] Start Simulation …
[UnitConverter] —————– UnitConverter information —————–
[UnitConverter] — Parameters:
[UnitConverter] Resolution: N= 50
[UnitConverter] Lattice velocity: latticeU= 0.00712282
[UnitConverter] Lattice relaxation frequency: omega= 1.95718
[UnitConverter] Lattice relaxation time: tau= 0.510939
[UnitConverter] Characteristical length(m): charL= 0.00980004
[UnitConverter] Characteristical speed(m/s): charU= 0.01
[UnitConverter] Phys. kinematic viscosity(m^2/s): charNu= 1.0034e-06
[UnitConverter] Phys. density(kg/m^d): charRho= 998
[UnitConverter] Characteristical pressure(N/m^2): charPressure= 0
[UnitConverter] Mach number: machNumber= 0.0123371
[UnitConverter] Reynolds number: reynoldsNumber= 97.6683
[UnitConverter] Knudsen number: knudsenNumber= 0.000126316
[UnitConverter]
[UnitConverter] — Conversion factors:
[UnitConverter] Voxel length(m): physDeltaX= 0.000196001
[UnitConverter] Time step(s): physDeltaT= 0.000139608
[UnitConverter] Velocity factor(m/s): physVelocity= 1.40394
[UnitConverter] Density factor(kg/m^3): physDensity= 998
[UnitConverter] Mass factor(kg): physMass= 7.51457e-09
[UnitConverter] Viscosity factor(m^2/s): physViscosity= 0.000275173
[UnitConverter] Force factor(N): physForce= 7.55688e-05
[UnitConverter] Pressure factor(N/m^2): physPressure= 1967.1
[UnitConverter] ————————————————————-
[STLreader] Voxelizing …
[STLmesh] nTriangles=21792; maxDist2=9.57151e-07
[STLmesh] minPhysR(StlMesh)=(-0.00531261,-0.0120206,-1.79526e-06); maxPhysR(StlMesh)=(0.00531261,0.0120206,0.0500018)
[Octree] radius=0.0501762; center=(-4.90002e-05,-4.90002e-05,0.024951)
[STLreader] voxelSize=0.000196001; stlSize=0.001
[STLreader] minPhysR(VoxelMesh)=(-0.00524302,-0.011907,0.000156916); maxPhysR(VoxelMesh)=(0.00514502,0.012005,0.0499412)
[STLreader] Voxelizing … OK
[main] STL was read …
[prepareGeometry] Prepare Geometry …
[SuperGeometry3D] cleaned 0 outer boundary voxel(s)
[SuperGeometry3D] cleaned 0 outer boundary voxel(s)
[SuperGeometry3D] cleaned 0 inner boundary voxel(s) of Type 3
[SuperGeometryStatistics3D] updated
[SuperGeometry3D] the model is correct!
[CuboidGeometry3D] —Cuboid Stucture Statistics—
[CuboidGeometry3D] Number of Cuboids: 1
[CuboidGeometry3D] Delta (min): 0.000196001
[CuboidGeometry3D] (max): 0.000196001
[CuboidGeometry3D] Ratio (min): 0.448
[CuboidGeometry3D] (max): 4.58929
[CuboidGeometry3D] Nodes (min): 1799000
[CuboidGeometry3D] (max): 1799000
[CuboidGeometry3D] Weight (min): 1799000
[CuboidGeometry3D] (max): 1799000
[CuboidGeometry3D] ——————————–
[SuperGeometryStatistics3D] materialNumber=0; count=1216568; minPhysR=(-0.00543902,-0.012103,-3.90851e-05); maxPhysR=(0.00534102,0.0122011,0.0501371)
[SuperGeometryStatistics3D] materialNumber=1; count=490960; minPhysR=(-0.00543902,-0.011907,0.000156916); maxPhysR=(0.00514502,0.012005,0.0499411)
[SuperGeometryStatistics3D] materialNumber=2; count=87399; minPhysR=(-0.00543902,-0.012103,-3.90851e-05); maxPhysR=(0.00534102,0.0122011,0.0501371)
[SuperGeometryStatistics3D] materialNumber=3; count=2036; minPhysR=(-0.00485102,-0.00485102,-3.90851e-05); maxPhysR=(0.00494902,0.00494902,-3.90851e-05)
[SuperGeometryStatistics3D] materialNumber=4; count=2037; minPhysR=(-0.00485102,-0.00485102,0.0501371); maxPhysR=(0.00494902,0.00494902,0.0501371)
[prepareGeometry] Prepare Geometry … OK
[main] Define SuperLattice …
terminate called after throwing an instance of ‘thrust::system::system_error’
what(): parallel_for failed: cudaErrorSymbolNotFound: named symbol not found
Aborted

Thank you for your help so far!

-Alex

[aseidler]

To the second question:
Yes it is running without compilation errors on only cpu and mpi.

[aseidler]

I managed to reproduce this error the CUDA_ARCH number was to high

[Adrian]

You mean you managed to fix it?

In any case, generating device code for a later generation than the one a GPU has would explain the Thrust exception.

[aseidler]

Yes and thank you for your help!

[Author]

Posts