Contribute
Register

[Solved] CUDA driver in MacOS High Sierra

Status
Not open for further replies.
What do you get when you run GPUSniffer as shown above?
Code:
--- OpenGL Info ---
Vendor: NVIDIA Corporation
Renderer: NVIDIA GeForce GTX 1080 Ti OpenGL Engine
OpenGL Version: 2.1 NVIDIA-10.25.28 378.10.10.10.15.120
GLSL Version: 1.20
Monitors: 3
Monitor 0 properties -
   Size: (0, 0, 1920, 1200)
   Max texture size: 16384
   Supports non-power of two: 1
   Shaders 444: 1
   Shaders 422: 1
   Shaders 420: 1
Monitor 1 properties -
   Size: (1920, 0, 1280, 1024)
   Max texture size: 16384
   Supports non-power of two: 1
   Shaders 444: 1
   Shaders 422: 1
   Shaders 420: 1
Monitor 2 properties -
   Size: (-539, -2160, 3840, 2160)
   Max texture size: 16384
   Supports non-power of two: 1
   Shaders 444: 1
   Shaders 422: 1
   Shaders 420: 1


--- GPU Computation Info ---
Found 3 devices supporting GPU computation.
CUDA Device 0 -
   Name: GeForce GTX 1080 Ti
   Vendor: NVIDIA
   Capability: 6.1
   Driver: 9
   Total Video Memory: 11263MB
<140735523697472> <Assert> <0> Debug Assert failed!
<140735523697472> <Assert> <0> Failed to load module: Fills, error: CUDA_ERROR_UNKNOWN


Expression: result == CUDA_SUCCESS
Process: GPUSniffer
Process ID: 926
Thread ID: 140735523697472
Thread Name:
File: /PPro11.1.1/releases/2017.03-1/shared/adobe/MediaCore/GPUFoundation/Src/ModuleManager.cpp
Line: 424
Function: void *GF::LoadCUDAModule(const dvacore::StdString &, const KernelStrings &)

<140735523697472> <Assert> <0> GF::LoadModule(GF::DeviceFramework, unsigned int, GF::Vendor, double, bool, _cl_device_id*, _cl_context*, void*, std::string const&)
GF::Device::FindModule(std::string)
GF::FillWithColor(boost::shared_ptr<GF::Device> const&, void const*, int, dvamediatypes::PixelFormat, int, int, int, int, float4)
GF::(anonymous namespace)::GPUVideoFrame::ClearToBlack()
DS::(anonymous namespace)::GPUSnifferInner(std::ostream&, unsigned int)
DS::GPUSnifferMain(std::ostream&, unsigned int)
main
start
??? <1>

Debug Assert failed!Failed to load module: Fills, error: CUDA_ERROR_UNKNOWN


Expression: result == CUDA_SUCCESS
Process: GPUSniffer
Process ID: 926
Thread ID: 140735523697472
Thread Name:
File: /PPro11.1.1/releases/2017.03-1/shared/adobe/MediaCore/GPUFoundation/Src/ModuleManager.cpp
Line: 424
Function: void *GF::LoadCUDAModule(const dvacore::StdString &, const KernelStrings &)
GF::LoadModule(GF::DeviceFramework, unsigned int, GF::Vendor, double, bool, _cl_device_id*, _cl_context*, void*, std::string const&)
GF::Device::FindModule(std::string)
GF::FillWithColor(boost::shared_ptr<GF::Device> const&, void const*, int, dvamediatypes::PixelFormat, int, int, int, int, float4)
GF::(anonymous namespace)::GPUVideoFrame::ClearToBlack()
DS::(anonymous namespace)::GPUSnifferInner(std::ostream&, unsigned int)
DS::GPUSnifferMain(std::ostream&, unsigned int)
main
start
??? <1>
<140735523697472> <GPU> <4> Cannot find module: Fills
   * Not chosen because of kernel failure.
   * Not chosen because of kernel failure.
OpenCL Device 1 -
   Name: GeForce GTX 1080 Ti
   Vendor: NVIDIA (Apple platform)
   Capability: 1.2
   Driver: 1.2
   Total Video Memory: 11264MB
Metal Device 2 -
   Name: GeForce GTX 1080 Ti
   Vendor: NVIDIA (Apple platform)
   Capability: 10000
   Driver: 0
   Total Video Memory: 11264MB
 
i'm having same issue as above. I did do the hack at

https://www.tonymacx86.com/threads/...-12-1-367-15-10-15.205568/page-2#post-1357600

but reinstalled cuda with the newest one for high sierra and now getting the same as below
Code:
--- GPU Computation Info ---
Found 3 devices supporting GPU computation.
CUDA Device 0 -
   Name: GeForce GTX 1080 Ti
   Vendor: NVIDIA
   Capability: 6.1
   Driver: 9
   Total Video Memory: 11263MB
<140735876715328> <Assert> <0> Debug Assert failed!
<140735876715328> <Assert> <0> Failed to load module: Fills, error: CUDA_ERROR_UNKNOWN


Expression: result == CUDA_SUCCESS
Process: GPUSniffer
Process ID: 1012
Thread ID: 140735876715328
Thread Name:
File: /Haberdasher/releases/2017.08/shared/adobe/MediaCore/GPUFoundation/Src/CUDA/CUDAModuleManager.cpp
Line: 43
Function: virtual void *GF::CUDADeviceModuleManager::LoadModuleImpl(const dvacore::StdString &, const KernelStrings &)

<140735876715328> <Assert> <0> GF::CUDADeviceModuleManager::LoadModuleImpl(std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> > const&, std::__1::vector<std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> >, dvacore::utility::SmallBlockAllocator::STLAllocator<std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> > > > const&)
GF::DeviceModuleManager::LoadModuleGuarded(std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> > const&, std::__1::vector<std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> >, dvacore::utility::SmallBlockAllocator::STLAllocator<std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> > > > const&)
GF::DeviceModuleManager::LoadModule(std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> > const&)
GF::Device::FindModule(std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> >)
GF::LoadKernel(boost::shared_ptr<GF::Device> const&, char const*, char const*)
GF::FillWithColor(boost::shared_ptr<GF::Device> const&, void const*, int, dvamediatypes::PixelFormat, int, int, int, int, float4)
GF::(anonymous namespace)::GPUVideoFrame::ClearToBlack()
DS::(anonymous namespace)::GPUSnifferInner(std::__1::basic_ostream<char, std::__1::char_traits<char> >&, unsigned int)
DS::GPUSnifferMain(std::__1::basic_ostream<char, std::__1::char_traits<char> >&, unsigned int)
main
start

Debug Assert failed!Failed to load module: Fills, error: CUDA_ERROR_UNKNOWN


Expression: result == CUDA_SUCCESS
Process: GPUSniffer
Process ID: 1012
Thread ID: 140735876715328
Thread Name:
File: /Haberdasher/releases/2017.08/shared/adobe/MediaCore/GPUFoundation/Src/CUDA/CUDAModuleManager.cpp
Line: 43
Function: virtual void *GF::CUDADeviceModuleManager::LoadModuleImpl(const dvacore::StdString &, const KernelStrings &)
GF::CUDADeviceModuleManager::LoadModuleImpl(std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> > const&, std::__1::vector<std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> >, dvacore::utility::SmallBlockAllocator::STLAllocator<std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> > > > const&)
GF::DeviceModuleManager::LoadModuleGuarded(std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> > const&, std::__1::vector<std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> >, dvacore::utility::SmallBlockAllocator::STLAllocator<std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> > > > const&)
GF::DeviceModuleManager::LoadModule(std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> > const&)
GF::Device::FindModule(std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> >)
GF::LoadKernel(boost::shared_ptr<GF::Device> const&, char const*, char const*)
GF::FillWithColor(boost::shared_ptr<GF::Device> const&, void const*, int, dvamediatypes::PixelFormat, int, int, int, int, float4)
GF::(anonymous namespace)::GPUVideoFrame::ClearToBlack()
DS::(anonymous namespace)::GPUSnifferInner(std::__1::basic_ostream<char, std::__1::char_traits<char> >&, unsigned int)
DS::GPUSnifferMain(std::__1::basic_ostream<char, std::__1::char_traits<char> >&, unsigned int)
main
start
<140735876715328> <GPU> <4> Cannot find module: Fills
   * Not chosen because of kernel failure.
   * Not chosen because of kernel failure.
OpenCL Device 1 -
   Name: GeForce GTX 1080 Ti
   Vendor: NVIDIA (Apple platform)
   Capability: 1.2
   Driver: 1.2
   Total Video Memory: 11264MB
Metal Device 2 -
   Name: GeForce GTX 1080 Ti
   Vendor: NVIDIA (Apple platform)
   Capability: 10000
   Driver: 0
   Total Video Memory: 11264MB

Any ideas?

i hope i dont need to reinstall everything....
 
No cuda ... help please
 

Attachments

  • Schermata 2017-11-01 alle 11.28.16 AM.png
    Schermata 2017-11-01 alle 11.28.16 AM.png
    139.7 KB · Views: 447
Same thing...
Installed latest CUDA and it is not recognized.
Using native Apple drivers for GTX650, without nVidia web drivers.
 
Installed, same problem.
 
Status
Not open for further replies.
Back
Top