On 8/10/24 23:13, Herbert Xu wrote:
On Sat, Aug 10, 2024 at 11:11:07PM -0400, Waiman Long wrote:
Unless I'm missing something chunk_size cannot be zero before the
division because that's the first thing we check upon entry into
this function.
chunk_size is initialized as
ps.chunk_size = job->size / (ps.nworks * load_balance_factor);
chunk_size will be 0 if job->size < (ps.nworks * load_balance_factor). If
min_chunk is 0, chunk_size will remain 0.
That's why I was suggesting that you replace the division by
DIV_ROUND_UP. That should ensure that ps.chunk_size is not zero.
Now I see what you mean. Yes, we can probably change that to a
DIV_ROUND_UP operation to make sure that chunk_size is at least one
unless job->size is 0. I still think the current patch is a bit more
fail-safe.
Cheers, Longman
