parameters not being passed to the CUDA kernel

Hello.

I have the following kernel function definiton:

global void calc_covMatrix(float4 params, cuComplex* d_cov, cuComplex* d_hhBuff,cuComplex* d_vvBuff,cuComplex* d_vhBuff)

I am passing in a float4 params value, where I have set params.x, params.y and params.z

But it seems like my kernel is not getting any of the values.

What could be the cause of this?

Also weirtd is, that I am calling the kernel like so:

float4 params;
params.x = rows;
params.y = cols;
params.z = 0;
calc_covMatrix<<<dimGrid, dimBlock>>>(params, d_cov, d_hhBuff, d_vvBuff, d_vhBuff);
cutilSafeCall(cudaMemcpy(h_cov, d_cov, memsize, cudaMemcpyDeviceToHost));
params.z = 2;
calc_covMatrix<<<dimGrid, dimBlock>>>(params,d_cov, d_hhBuff, d_vvBuff, d_vhBuff);
cutilSafeCall(cudaMemcpy(h_cov, d_cov, memsize, cudaMemcpyDeviceToHost));

params.z = 4;
calc_covMatrix<<<dimGrid, dimBlock>>>(params, d_cov, d_hhBuff, d_vvBuff, d_vhBuff);
cutilSafeCall(cudaMemcpy(h_cov, d_cov, memsize, cudaMemcpyDeviceToHost));

And the only one that seems to get ANY value other than zero is the third call to the kernel, and it is only seeming to get the value 2!