I’ve implemented a CUDA program.
And in this program, a debug error has occured.
The error message is like below
HEAP CORRUPTION DETECTED: after normal block(#62) at 0x02485F48.
CRT detected that the application wrote to memory after end of heap buffer.”
A block contains 88 threads, and a grid contains 6633 blocks.
dim3 threads( BSIZE, BSIZE ); dim3 grid( T2/BSIZE, T1/BSIZE );
(Here, BSIZE=8, T1=264 and T2=528)
Has the heap corruption occured because of too many blocks(or threads)?
How can I solve this problem?