Pragma Unroll with variable loop bound

[Altera_Forum]

Is it compulsory for the loop bound to be constant for unrolling? 

 Would the below kernel generate 50 times hw replication? What will happen if host sets the k = 60?    __kernel __attribute__((task)) void test_multiplier(global char *restrict in, global char *restrict weights, global int *restrict out, int k) {   int output = 0; #pragma unroll 50 for(int i=0; i<k; i++){ output += in * weights; }

[Altera_Forum]

--- Quote Start ---  Is it compulsory for the loop bound to be constant for unrolling?  --- Quote End ---   No. However, for minimum area overhead and maximum performance, it is best if the loop bound is known at compile-time and is divisible by the unroll factor.    --- Quote Start ---  Would the below kernel generate 50 times hw replication? What will happen if host sets the k = 60?   --- Quote End ---   The compiler will create a pipeline that can process 50 iteration per clock by using more FPGA resources. If you set k = 60, then the pipeline will work for two clocks + pipeline latency; nothing will go wrong, just the pipeline utilization will be low in the last clock.  ...