a little bug
Closed this issue · 3 comments
jiaqiwang969 commented
Hi, maybe there is a little bug:
change eplison to eps?
davidem88 commented
No, why do you say so? It seems correct to me. epsilon is read from input and it's a scalar
jiaqiwang969 commented
That because when I am debugging, I found the bug:
—> FOAM FATAL ERROR:
LHS and RHS of + have different dimensions
dimensions : [0 0 -2 0 0 0 0] + [0 0 0 0 0 0 0]
divU2[icell] + rotU2[icell] + epsilon?
epsilon is dimensions [0 0 0 0 0 0 0], and it sends the value to eps, with its dimension [0 0 -2 0 0 0 0].
I am not sure why in cpu, it not show bug, In GPU version debugging, I found it.
davidem88 commented
As far as I know, in the CPU version of OF physical dimensions are not checked in the loops. That is why it is working anyway.