于 2013年09月13日 11:15, Qu Wenruo 写道:
于 2013年09月13日 09:47, Liu Bo 写道:
On Thu, Sep 12, 2013 at 04:08:24PM +0800, Qu Wenruo wrote:
The original btrfs_workers uses the fs_info-thread_pool_size as the
max_active, and the previous patches followed this way.
But the kernel workqueue has the default
The original btrfs_workers uses the fs_info-thread_pool_size as the
max_active, and the previous patches followed this way.
But the kernel workqueue has the default value(0) for workqueue,
and workqueue itself has some threshold mechanism to prevent creating
too many threads, so we should use the
On Thu, Sep 12, 2013 at 04:08:24PM +0800, Qu Wenruo wrote:
The original btrfs_workers uses the fs_info-thread_pool_size as the
max_active, and the previous patches followed this way.
But the kernel workqueue has the default value(0) for workqueue,
and workqueue itself has some threshold
于 2013年09月13日 09:47, Liu Bo 写道:
On Thu, Sep 12, 2013 at 04:08:24PM +0800, Qu Wenruo wrote:
The original btrfs_workers uses the fs_info-thread_pool_size as the
max_active, and the previous patches followed this way.
But the kernel workqueue has the default value(0) for workqueue,
and workqueue