On Thu, Jan 21, 2016 at 11:44:13AM +0300, Andrey Ryabinin wrote:
> The main idea is following:
> * for background works we check all UBs for exceeding dirty limit.
> * background work goes on if any UB has exceed dirty limit.
> * In that case, writeback will skip inodes if those belong to
>
The main idea is following:
* for background works we check all UBs for exceeding dirty limit.
* background work goes on if any UB has exceed dirty limit.
* In that case, writeback will skip inodes if those belong to
"within dirty-limits UB"
writeback_inodes_wb() gain an new 'struct user_