Open ishitachaturvedi opened 2 years ago
Yep, today I have the same. I'm trying to run CUDA sample "vectorAdd" (CUDA 11.6) in release mode at Ubuntu. I have no video card (integrated Intel one).
sorry, looks like the issue is duplicated can someone kindly point to a good and final resolution of the issue?
Me too, I also run over this question, I run on the wls2 with Ubuntu 18.04 and Cuda 11.06, using the RTX2060s' gpgpusim-config
I can't run GPGPU-Sim in Cuda version 11.0 or above, too. My workaround was to run GPGPU-Sim on Cuda 10.0 (with Ubuntu 20.04 and g/g++ 7).
I solve the problem now by reinstalling the dependency, and dealing with the GCC g++ version. I found that the dependency of the dependency has the version conflict, hence making the segfaults. otherwise, I also found that my gcc and g++ were different versions. I thought that maybe the two points are why the problems happened. I solve the problem under the env of wls2 ubuntu 18.04, cuda 11.06.
I solve the problem now by reinstalling the dependency, and dealing with the GCC g++ version. I found that the dependency of the dependency has the version conflict, hence making the segfaults. otherwise, I also found that my gcc and g++ were different versions. I thought that maybe the two points are why the problems happened. I solve the problem under the env of wls2 ubuntu 18.04, cuda 11.06.
Can you elaborate the the dependency of the dependency has the version conflict
?
I also have the segfaults problem with same version g++ and gcc in 11.2.0, cuda in 11.6. Thank you!
Hey, can you tell exactly which dependency was the issue with? I am running gpgpus-sim in wsl ubuntu 20.06, cuda 11.4. I am getting similar errors. My g++/gcc are both version 9.4.0
I solve the problem now by reinstalling the dependency, and dealing with the GCC g++ version. I found that the dependency of the dependency has the version conflict, hence making the segfaults. otherwise, I also found that my gcc and g++ were different versions. I thought that maybe the two points are why the problems happened. I solve the problem under the env of wls2 ubuntu 18.04, cuda 11.06.
Can you elaborate the
the dependency of the dependency has the version conflict
? I also have the segfaults problem with same version g++ and gcc in 11.2.0, cuda in 11.6. Thank you!
like this
The following packages have unmet dependencies: libglu1-mesa-dev : Depends: libglu1-mesa (= 9.0.0-2.1build1) but 9.0.1-1build1 is to be installed zlib1g-dev : Depends: zlib1g (= 1:1.2.11.dfsg-0ubuntu2) but 1:1.2.11.dfsg-2ubuntu1.2 is to be installed E: Unable to correct problems, you have held broken packages.
and also i change the dependency of libglut3-dev into freeglut3-dev cause i couldn't install the libglut3-dev successfully
I wrote a tutorial of the installation of gpgpusim, but is in chinese, the link is https://gitee.com/dazhuzhu-gitee/gpgpusim-document/blob/master/document/GPGPU-Sim%E5%AE%89%E8%A3%85%E7%AE%80%E6%98%93%E6%95%99%E7%A8%8B_Au.md
Hey, Thanks a lot! I figured it out mostly but still thanks
Gyanesh
On Fri, Apr 22, 2022, 04:00 dazhuzhu zheng @.***> wrote:
I wrote a tutorial of the installation of gpgpusim, but is in chinese, the link is
— Reply to this email directly, view it on GitHub https://github.com/gpgpu-sim/gpgpu-sim_distribution/issues/249#issuecomment-1106136902, or unsubscribe https://github.com/notifications/unsubscribe-auth/AFIHWMVZBFQCDJRDWZFE7JLVGJMBZANCNFSM5NMVWYFQ . You are receiving this because you commented.Message ID: @.***>
Hey, Thanks a lot! I figured it out mostly but still thanks Gyanesh
Hello Every one & @gyaneshg96 I am stuck in the same issue? Could you please let me know how did you solve it? My setup Ubuntu 20.04 CUDA - 11.8
I have tried gcc & g++ version 7.5.0 & 7.5.0 But no luck.
Thanks in advance! Pallab
I used a docker environment to run the simulator. It has a lower version of cuda and gcc, but was enough for my purposes.
Gyanesh
On Thu, Nov 24, 2022 at 11:22 AM pal-stdr @.***> wrote:
Hey, Thanks a lot! I figured it out mostly but still thanks Gyanesh … <#m-6181088633648215842> On Fri, Apr 22, 2022, 04:00 dazhuzhu zheng @.> wrote: I wrote a tutorial of the installation of gpgpusim, but is in chinese, the link is https://gitee.com/dazhuzhu-gitee/gpgpusim-document/blob/master/document/GPGPU-Sim%E5%AE%89%E8%A3%85%E7%AE%80%E6%98%93%E6%95%99%E7%A8%8B_Au.md https://gitee.com/dazhuzhu-gitee/gpgpusim-document/blob/master/document/GPGPU-Sim%E5%AE%89%E8%A3%85%E7%AE%80%E6%98%93%E6%95%99%E7%A8%8B_Au.md — Reply to this email directly, view it on GitHub <#249 (comment) https://github.com/gpgpu-sim/gpgpu-sim_distribution/issues/249#issuecomment-1106136902>, or unsubscribe https://github.com/notifications/unsubscribe-auth/AFIHWMVZBFQCDJRDWZFE7JLVGJMBZANCNFSM5NMVWYFQ https://github.com/notifications/unsubscribe-auth/AFIHWMVZBFQCDJRDWZFE7JLVGJMBZANCNFSM5NMVWYFQ . You are receiving this because you commented.Message ID: @.>
Hello I am stuck in the same issue? Could you please let me know how did you solve it? My setup Ubuntu 20.04 CUDA - 11.8
I have tried gcc & g++ version 7 & 9 But no luck.
Thanks in advance! Pallab
— Reply to this email directly, view it on GitHub https://github.com/gpgpu-sim/gpgpu-sim_distribution/issues/249#issuecomment-1326652364, or unsubscribe https://github.com/notifications/unsubscribe-auth/AFIHWMWKEI7SMD4V34YMQ3DWJ6I2RANCNFSM5NMVWYFQ . You are receiving this because you commented.Message ID: @.***>
I met the same issue and resolved with re-compiled gpgpu-sim with debug option. If you see the same issue, please try with:
source setup_environment debug make
My set is: WSL2 Ubuntu 20.04 g++ 9.3.0 cuda 11.0
Still experiencing this error Ubuntu 20.04 CUDA 11.0 gcc/g++ 9.4
Anyone have suggestions?
FIXED: source setup_environment debug
Are there any fixes for this issue other than compiling GPGPU-Sim in debug mode?
My setup: Ubuntu 20.04 CUDA 11.8 gcc/g++ 8.4
I need to run GPGPU-Sim for multiple time-consuming benchmarks and debug mode just takes too long to simulate.
It seems to be because this line lacks a return
statement, adding it fixes this issue with a release build
Before:
cudaConfigureCallInternal(gridDim, blockDim, sharedMem, stream);
After:
return cudaConfigureCallInternal(gridDim, blockDim, sharedMem, stream);
After building GPGPUSim and adding --cudart shared on running GPGPUSim it prints the system configuration followed by multiple- "GPGPU-Sim PTX: CUDA API function "unsigned int __cudaPushCallConfiguration(dim3, dim3, size_t, CUstream_st*)" has been called" and seg faults