Modified SETI MB CUDA + opt AP package for full GPU utilization

[Raistmer]

Well, 6.06 sources differ form my own.
For example:

Mine:
          // TODO: Deallocate these at the end of the function
#ifdef USE_CUDA
         if(gSetiUseCudaDevice)
            cudaAcc_fftwf_plan_dft_1d(FftNum, FftLen, NumDataPoints);
   
From Berkeley's repository:

#ifdef USE_CUDA
            // Create FFT plan configuration
         if(gSetiUseCudaDevice)
            {
            if(cudaAcc_fftwf_plan_dft_1d(FftNum, FftLen, NumDataPoints))
                {
                    // If were here, something went wrong in the plan.
                    // Possibly ran out of video memory.  Destroy what we
                    // have and fallback to host CPU processing.
                    fprintf(stderr, "setiathome_CUDA: CUDA runtime ERROR in plan FFT. Falling back to HOST CPU processing...\n");
                    cudaAcc_free();
                    gSetiUseCudaDevice = 0;

                }
            }

#endif

Will rebuild mod with new sources and re-enable access to it for testing.

[Raistmer]

Ok, here is a new package, based on current (380) revision of Berkeley's SETI CUDA branch.


[attachment deleted by admin]

[Jason G]

Does 6.06 fix the memory leak at least ?

[Raistmer]

Does 6.06 fix the memory leak at least ?

I think it fixes nothing. But wanna check this for sure.
Joe's proposal about leak fixing still pending. I didn't check this in sources still but pretty sure that modification not included in sources by Berkeley.
rev 380 from 17 December ....

ADDON: now I will use this "up to date"mod to check broken results standalone too.
I didn't see any problems with PG* VHARs before, only with VLARs....

And first E-mailed sources differ from Berkeley's rep ones... So it's worth to try...

[Raistmer]

Here is new build with logging of AR of overflowed tasks.
It will create and append later file r_debug.txt on C:\ and will write AR of overflowed task there.
Sure there will be "legal" overflows too, but any overflows not validated against CPU result worth to report.

Just replace CUDA MB executable from my last package with this one. Name remains same.

Logging mod is:

  if (e == RESULT_OVERFLOW) {
        fprintf(stderr, "SETI@Home Informational message -9 result_overflow\n");
        fprintf(stderr, "NOTE: The number of results detected exceeds the storage space allocated.\n");
      {//R: Logging of AR of overflowed task
         FILE* f=fopen("C:\\r_debug.txt","a");
         if(f){
            fprintf(f,"AR of overflowed task is\t%f\n",swi.angle_range);
            fclose(f);
         }
      }


[attachment deleted by admin]

[wndrbr3d]

Why don't you have the debug file write to the BOINC project folder?

People running Vista won't have access to write to root unless they're running BOINC and it's child processes as administrator. This way everyone can get a debug, not just the XP users 

[popandbob]

Also might be an idea to log the AR's of the wu's that don't overflow to see if there are any ranges that for whatever reason don't overflow...
Like for me 2.71 overflows 2.72 doesn't (for the most part)

[Raistmer]

Why don't you have the debug file write to the BOINC project folder?

People running Vista won't have access to write to root unless they're running BOINC and it's child processes as administrator. This way everyone can get a debug, not just the XP users 

To lazy to figure out where to get SETI project directory on BOINC installation. Any ideas?

[Raistmer]

Also might be an idea to log the AR's of the wu's that don't overflow to see if there are any ranges that for whatever reason don't overflow...
Like for me 2.71 overflows 2.72 doesn't (for the most part)

Don't wanna be blaming for hard drive overloading

[Josef W. Segur]

Why don't you have the debug file write to the BOINC project folder?

People running Vista won't have access to write to root unless they're running BOINC and it's child processes as administrator. This way everyone can get a debug, not just the XP users 

To lazy to figure out where to get SETI project directory on BOINC installation. Any ideas?

The project_dir in the APP_INIT_DATA structure. The structure is defined in the BOINC app_ipc.h file, seti.h has a "extern APP_INIT_DATA app_init_data;" line, and part of initialization is to read the init_data.xml file in the slot directory and fill in the structure. For standalone testing it will probably be a null string, then maybe just using the current directory might make sense.

That's assuming the app will have permission to write in the project directory, I'm not sure about that.
                                                                         Joe

[Raistmer]

Thanks again, Joe
Well, writing in current directory not suitable for BOINC runs (slot will be deleted). Will try to change to project directory and check on VIsta.
C:\r_debug.txt still doesn't exist here although host completed few VHAR tasks. SO, BOINC-summoned app can't access to C:\ indeed 

[efmer (fred)]

This works always as iit is the Windows Temp Directory GetTempPath(MAX_PATH,wTempPath);

[Raistmer]

Thanks here too and welcome on Lunatics forums

[Slawek]

Hi

What news in CUDA APP ? new version?

[Raistmer]

Hi

What news in CUDA APP ? new version?

Ok, will build one with log in temp folder. And available memory log at task startup. It will illustrate memory leak problem I hope. So, only cosmetic changes.
Nothing new in Berkeley's repository still.

Addon: here some cosmetic improved version
Now will be this string in stder:

Total GPU memory 402653184    free GPU memory 355926016

If the GPU memory value will decrease from task to task - it's better to reboot before CUDA failures begins.


[attachment deleted by admin]