error at the beginning of running gem5 gpu

[Veydan Wu]

Hi, all

I encountered an error in running gem5-gpu SE mode at the very beginning. 

... ...

GPGPU-Sim PTX: ... done pre-decoding instructions for '_Z7RecoverPmS_S_S_S_iii'.

GPGPU-Sim PTX: finished parsing EMBEDDED .ptx file _1.ptx

GPGPU-Sim PTX: extracting embedded .ptx to temporary file "_ptx_RGXBDQ"

Running: cat _ptx_RGXBDQ | sed 's/.version 1.5/.version 1.4/' | sed 's/, texmode_independent//' | sed 's/\(\.extern \.const\[1\] .b8 \w\+\)\[\]/\1\[1\]/' | sed 's/const\[.\]/const\[0\]/g' > _ptx2_jnAv0W

GPGPU-Sim PTX: ERROR ** while loading PTX (a) -1

               Ensure you have write access to simulation directory

               and have 'cat' and 'sed' in your path.

and the simulation just exit. I launched several instances of gem5-gpu, and every time there are 1 or 2 failed with this error, and it is random in terms of which one(s) failed. If I move the failed one(s) to another machine, they could run without such error. But no matter on which machine, there are always some failed.  Has anybody encountered this error before? Thanks a lot. 

--

Regards,

Veydan

[Jason Power]

Hi Veydan,

What kind of platform are you running on? This looks like an issue with your file system. Things like this could happen using NFS or AFS, though I haven't seen this specific problem. Since it only happens on some simulations, I'd guess it's a problem with your platform, not with gem5-gpu.

Let us know if you find anything.

Cheers,

Jason

[nirmal kumbhare]

Hello Veydan,

Were you able to get around the above problem? I am hitting some what similar issue while running "rodinia/backprop" benchmark. It seems system function call from c program is returning (-1) due to which i am getting following failing signature:

GPGPU-Sim PTX: extracting embedded .ptx to temporary file "_ptx_KQ2D2i"

Running: cat _ptx_KQ2D2i | sed 's/.version 1.5/.version 1.4/' | sed 's/, texmode_independent//' | sed 's/\(\.extern \.const\[1\] .b8 \w\+\)\[\]/\1\[1\]/' | sed 's/const\[.\]/const\[0\]/g' > _ptx2_h4vIU3

GPGPU-Sim PTX: ERROR ** while loading PTX (a) -1

               Ensure you have write access to simulation directory

               and have 'cat' and 'sed' in your path.

If you were able to get around this please share the fix.

Hello Jason,

I am running on "Ubuntu 14.04.1 LTS". and my file system is ext4.

Thanks,

Nirmal K

[Jason Power]

Hi Nirmal,

I'd check to be cure that you have permission to write to your simulation directory and that you can run cat and sed. This step is doing some slight changes to the PTX code so that GPGPU-Sim can work with it. It's using simple bash commands, so I'm not sure why you're seeing the error.

Hopefully this helps,

Jason

[Josue Quiroga]

Hi guys,

      I do have the same problem, I already check the permissions and it stills appears the same error:

GPGPU-Sim PTX: finished parsing EMBEDDED .ptx file _1.ptx

GPGPU-Sim PTX: extracting embedded .ptx to temporary file "_ptx_S08MRf"

Running: cat _ptx_S08MRf | sed 's/.version 1.5/.version 1.4/' | sed 's/, texmode_independent//' | sed 's/\(\.extern \.const\[1\] .b8 \w\+\)\[\]/\1\[1\]/' | sed 's/const\[.\]/const\[0\]/g' > _ptx2_a9sPrL

GPGPU-Sim PTX: ERROR ** while loading PTX (a) -1

               Ensure you have write access to simulation directory

               and have 'cat' and 'sed' in your path.

Do you have any idea of what else we can try? Thanks in advance.

Best regards

Josué Vladimir Quiroga Esparza.

[Victor Garcia]

Hi Josue,

This is something I've seen sometimes when running jobs in our cluster. It seems to be an problem with NFS as Jason says. 

It's a rare error that I've been unable to consistently replicate, so I just end up resubmitting the jobs that failed and eventually the error goes away.

Sorry not to be of more help, but you can test at least if it's an NFS problem by running the simulation in a local machine.

Good luck!

[Josue Quiroga]

Hi Victor,

    The interesting thing is that I'm trying the simulator on my computer, I'm going to install it directly on the cluster. Thank you for the quick answer. 

Best regards

Josué Vladimir Quiroga Esparza.

[Josue Quiroga]

It totally works in the cluster, I don't know why it does not work in my computer. Thank you.

[Xuhao Chen]

I've got the same error when running the latest gem5-gpu on Ubuntu 14.04 64-bit in a virtual machine.

The thing is, when I run it like this:

build/X86_VI_hammer_GPU/gem5.opt ../gem5-gpu/configs/se_fusion.py --total-mem-size=512MB -c ../benchmarks/rodinia/backprop/gem5_fusion_backprop -o 16

it works, however, when I increase the memory size to 1024MB, like this:

build/X86_VI_hammer_GPU/gem5.opt ../gem5-gpu/configs/se_fusion.py --total-mem-size=1024MB -c ../benchmarks/rodinia/backprop/gem5_fusion_backprop -o 16

I got the error:

Running: cat _ptx_4B46NS | sed 's/.version 1.5/.version 1.4/' | sed 's/, texmode_independent//' | sed 's/\(\.extern \.const\[1\] .b8 \w\+\)\[\]/\1\[1\]/' | sed 's/const\[.\]/const\[0\]/g' > _ptx2_D2nkxE

GPGPU-Sim PTX: ERROR ** while loading PTX (a) -1

               Ensure you have write access to simulation directory

               and have 'cat' and 'sed' in your path.

so it looks like a bug instead of just a system config problem. What do you think?