topic Do you mean kernel like this: in OpenCL* for CPU

VS Memory/Watchlists window content inaccurate during debugging

PCox — Mon, 16 Mar 2015 23:18:33 GMT

Using code buiider 1.3.0.92 under windows 7 and VS 2010 sp1....

If I run a hack kernel like this...

__kernel void test_local()
{
__local int s_testValue[10][10];
__local int* rowStartThisThread = &(s_mcuValues[0][0]);

s_mcuValues[0][0] = 5;
}

and set a breakpoint after the setting of the 5, then the VS memory and watchlists don't see the '5' (they only see zeros), however the rowStartThisThread pointer does show it correctly. Am I making some bad assumptions here?

Do you mean kernel like this:

Robert_I_Intel — Mon, 16 Mar 2015 23:24:44 GMT

Do you mean kernel like this:

__kernel void test_local()
{
__local int s_testValue[10][10];
__local int* rowStartThisThread = &(s_testValue[0][0]);

s_testValue[0][0] = 5;
}

What happens if you put the barrier after assigning 5 and observe things after the barrier?

Hi Robert, thanks for the

PCox — Tue, 17 Mar 2015 00:07:07 GMT

Hi Robert, thanks for the impressively quick response! Yes sorry I changed the variable names inconsistently at the last minute before posting.

But sigh.

I put a barrier(CLK_LOCAL_MEM_FENCE) in, and, yes, it all sprang to life as I would have expected. But I then restored the code back to what it was, only to have it continue working. I really dislike heisenberg faults.....

I can't think of any unaccounted variables in play here (btw I'm only launching a single thread). Meanwhile the problem in the real part of the code continues, with or without a barrier. Very frustrating.