Poclbm Source Code HELP!!!

[betatest512]

can someone show me the line of code in the poclbm source code where the nonce is set

and the line of code where is nonce is incremented please show me the exact line and not only the function it is it and please provide as much detail as you can.

and can anyone tell me what does poclbm do after it has gone through all the nonce from 1 billion to 4 billion. what does it increment next. (and can you point out the line in code where it is incremented)

[1714218001]

1714218001

[1714218001]

1714218001

[1714218001]

1714218001

[1714218001]

1714218001

[crazyates]

Of all the people on this forum, you're the last one who should be digging around the poclbm source code. Please stop, and just mine away like a good little boy (or girl, not sure).

[kuzetsa]

can someone show me the line of code in the poclbm source code where the nonce is set

and the line of code where is nonce is incremented please show me the exact line and not only the function it is it and please provide as much detail as you can.

and can anyone tell me what does poclbm do after it has gone through all the nonce from 1 billion to 4 billion. what does it increment next. (and can you point out the line in code where it is incremented)

**edited because I was using dismissive language / hadn't considered that you might be planning a patch of some sort

After 2^32 (4 billion) the "difficulty 1" share is submitted. This is because difficulty "1" only has that many possible nonce in addition to the base for your share (which is set by the pool or your copy of bitcoind via getwork RPC, depending on use case)

... Curious / wondering what change (if any) you plan to make here.

Sorry, I know I really didn't answer -- OpenCL breaks my head and I've not looked in the poclbm kernel yet.


Edited to add:

From looking at poclbm120823 (from cgminer) it seems that unless you're using the standalone version (or at least some version other than the one used in cgminer) the kernel itself doesn't seem to use counter checking to stop after 2^32 hashes. Really not sure what sort of control loop is used by cgminer, but the opencl kernel itself doesn't seem to handle it. Sorry. I'm more confused than I expected from looking at it.

[-ck]

From looking at poclbm120823 (from cgminer) it seems that unless you're using the standalone version (or at least some version other than the one used in cgminer)

I simply kept the poclbm name in that kernel because that's where it evolved from. The poclbm kernel in cgminer is nothing like poclbm's own kernel any more.

[kuzetsa]

((...snip...))

The poclbm kernel in cgminer is nothing like poclbm's own kernel any more.

I stand corrected.

Was I correct in my reverse engineering? The control loop for cgminer resides in the cgminer CLI program itself? (not in the open CL kernel)

Edited to clarify:

@betatest512

Still pretty sure my post isn't fulfilling the exact specifics of the request you're making... I'm still really not sure how the standalone poclbm thingy compares to cgminer's use of an opencl kernel "inspired by" poclbm...

... But this seems like the sort of code which could be roughly traced to, and found near the edges at or around the glue in NEARLY ANY bitcoin miner implementation.