@r05: try compiling the older version of the algorithm, which consumes much less memory:

make clean
CFLAGS="-DDEBUG_GPUV3" make -f makefile.unix


and then try several memory amounts from -m 20 (4Mb) to -m 27 (512MB) (default).

If you're still with the same error, I'm lost :( Which graphics card are you running it on? How much RAM on your machine?

Quote from: girino on January 14, 2014, 01:11:38 AM
Ok, first error is easy and should be fixed on the repository. AMD SDK does not like "#pragma unroll (8)" directive. Just change all occurrences of this to just "#pragma unroll" in file opencl/OpenCLMomentumV4.cl

The second one seems like lack of memory. What did you change in order for it to happen?

I reinstalled the AMD drivers from their website directly instead of the Debian repos.

Your precompiled binary says:

Quotelibboost_system.so.1.48.0: cannot open shared object file: No such file or directory

I'm running Debian 7 so I currently have 1.49.0 installed, which would explain the binary not running. I'm still at a loss about the compiled miner, though..

Quote from: carloss on January 14, 2014, 09:16:33 AM

Quote from: Komik on January 14, 2014, 01:41:57 AM
yvg1900,
Thank you very much!
My result: AMD FX 8350 (OC 4600Mhz) 345 CPM!

P.S: 7 Thread!

Very nice. What version (for which architecture) and what av do you use?

yam-yvg1900-M7i-win64-bdver1 AV=1

https://bitsharestalk.org/index.php?topic=1928.0

CPU: AMD FX-8350 8-core 4.2GHz (OC 4.80GHz)
RAM: 16 GB DDR3 RAM 2133 Mhz
OS: Windows 7 (64bit)
Software: yam-yvg1900-M7i-win64-bdver1

PTS Agg. CPM (1024 MB): ~ 395-402 (AV=1)
PTS Agg. CPM (1024 MB): 386-390 (AV=2)

Quote from: Komik on January 14, 2014, 01:41:57 AM
yvg1900,
Thank you very much!
My result: AMD FX 8350 (OC 4600Mhz) 345 CPM!

P.S: 7 Thread!

Very nice. What version (for which architecture) and what av do you use?

Quote from: dga on January 14, 2014, 02:12:34 AM

Quote from: Komik on January 14, 2014, 01:41:57 AM
yvg1900,
Thank you very much!
My result: AMD FX 8350 (OC 4600Mhz) 345 CPM!

P.S: 7 Thread!

Out of curiosity, would you mind letting me know how this compares to the open source version I just released, if you're mining on beeeeeer? :)

Pts Pool Miner v 0.7 RC 2 = 199.2 c/m

Quote from: Komik on January 14, 2014, 01:41:57 AM
yvg1900,
Thank you very much!
My result: AMD FX 8350 (OC 4600Mhz) 345 CPM!

P.S: 7 Thread!

Out of curiosity, would you mind letting me know how this compares to the open source version I just released, if you're mining on beeeeeer? :)

yvg1900,
Thank you very much!
My result: AMD FX 8350 (OC 4600Mhz) 345 CPM!

P.S: 7 Thread!

try downloading this compiled version (i compiled with NVidia SDK, but should work with AMD as well): http://girino.org/ptsminer/ptsminer-linux-x86_64.tgz

Ok, first error is easy and should be fixed on the repository. AMD SDK does not like "#pragma unroll (8)" directive. Just change all occurrences of this to just "#pragma unroll" in file opencl/OpenCLMomentumV4.cl

The second one seems like lack of memory. What did you change in order for it to happen?

Quote from: girino on January 13, 2014, 05:54:49 PM
@r05 try getting the code from the repository again. The makefile.unix was bogus and i corrected it sometime ago (yesterday i think). make sure you are building with:

Code Select Expand

make -f makefile.unix

Also, the new algorithm is almost 50% faster.

Finally got it to compile, another (different) error:

Quote********************************************
*** ptsminer - Pts Pool Miner v0.7 RC2 <experimental>
*** by xolokram/TB - www.beeeeer.org - glhf
***
*** performance improvements by girino
***    if you like, donate:
***    PTS: PkyeQNn1yGV5psGeZ4sDu6nz2vWHTujf4h
***    BTC: 1GiRiNoKznfGbt8bkU1Ley85TgVV7ZTXce
*** thanks to wjchen for SSE4 improvements.
***
*** press CTRL+C to exit
********************************************
using GPU
spawning 1 worker thread(s)
[WORKER0] Hello, World!
[WORKER0] GoGoGo!
connecting to 54.201.26.128:1337
[MASTER] work received - sharetarget: 03ffffffffffffffffffffffffffffffffffffffffffffffffffffffbeefde4d
[STATS] 2014-Jan-13 22:33:14 | 0.0 c/m | 0.0 sh/m | VL: 0 (0.0%), RJ: 0 (0.0%), ST: 0 (0.0%)
Platform: AMD Accelerated Parallel Processing
  Device: Barts
  Device: Intel(R) Core(TM)2 Quad CPU    Q9550  @ 2.83GHz
Starting OpenCLMomentum V4
Device: Barts
Max work group size: 256
ERROR: -11, CL_BUILD_PROGRAM_FAILURE, if there is a failure to build the program executable. This error will be returned if clBuildProgram does not return until the build has completed.
--- Build log ---
▒▒c
---  End log  ---
ptsminer: OpenCLObjects.cpp:289: OpenCLProgram* OpenCLContext::loadProgramFromStrings(std::vector<std::basic_string<char> >): Assertion `!error' failed.
Aborted

After some fiddling I now get this error:

Quote*** ptsminer - Pts Pool Miner v0.7 RC2 <experimental>
*** by xolokram/TB - www.beeeeer.org - glhf
***
*** performance improvements by girino
***    if you like, donate:
***    PTS: PkyeQNn1yGV5psGeZ4sDu6nz2vWHTujf4h
***    BTC: 1GiRiNoKznfGbt8bkU1Ley85TgVV7ZTXce
*** thanks to wjchen for SSE4 improvements.
***
*** press CTRL+C to exit
********************************************
using GPU
spawning 1 worker thread(s)
[WORKER0] Hello, World!
[WORKER0] GoGoGo!
connecting to 54.201.26.128:1337
[MASTER] work received - sharetarget: 03ffffffffffffffffffffffffffffffffffffffffffffffffffffffbeefde4d
[STATS] 2014-Jan-13 23:32:40 | VL: 0 (0.0%), RJ: 0 (0.0%), ST: 0 (0.0%)
Platform: AMD Accelerated Parallel Processing
  Device: Barts
  Device: Intel(R) Core(TM)2 Quad CPU    Q9550  @ 2.83GHz
Starting OpenCLMomentum V4
Device: Barts
Max work group size: 256

[STATS] 2014-Jan-13 23:33:10 | 0.0 c/m | 0.0 sh/m | VL: 0 (0.0%), RJ: 0 (0.0%), ST: 0 (0.0%)
ERROR: -5, CL_OUT_OF_RESOURCES, if there is a failure to allocate resources required by the OpenCL implementation on the device.
ptsminer: OpenCLObjects.cpp:421: _cl_event* OpenCLCommandQueue::enqueueKernel1D(OpenCLKernel*, size_t, size_t, _cl_event**, size_t): Assertion `_MY_ERR_X == 0' failed.

Ok just released cudaPTSwin - Beeeeer - PTS GPU miner for Windows  -> https://bitsharestalk.org/index.php?topic=2311.0

Cheers!

Issues with running the AMD card without an X windows session - my own fault, not that of the application. Standby, I will feedback to you :)

Quote from: girino on January 13, 2014, 05:54:49 PM
@r05 try getting the code from the repository again. The makefile.unix was bogus and i corrected it sometime ago (yesterday i think). make sure you are building with:

Code Select Expand

make -f makefile.unix

Also, the new algorithm is almost 50% faster.

Compiled after getting a fresh copy from the repo - give me 10 minutes and I'll put the AMD card back in and report back. :)

@r05 try getting the code from the repository again. The makefile.unix was bogus and i corrected it sometime ago (yesterday i think). make sure you are building with:

make -f makefile.unix

Also, the new algorithm is almost 50% faster.

Quote from: girino on January 13, 2014, 01:42:00 AM
you need to add the SDK to the library path. try running something like:

export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/path/to/SDK/lib

where "/path/to/SDK/lib" is the place where the SDK installed the libOpenCL.so file. On NVidia its

/usr/local/cuda/lib/x86_64/

not sure where AMD installs it.

Sorry I'm still having no luck.. Same set of errors despite setting that environment variable.