At Tue, 3 May 2011 12:02:06 +0100, Mark Brown wrote:
On Tue, May 03, 2011 at 12:50:03PM +0200, Takashi Iwai wrote:
Mark Brown wrote:
This isn't about CPU usage, it's about I/O bandwidth which is a big concern in situations like resume where you can be bringing the device back up from cold.
Hm, but how do these patches achieve it? I see no change in the I/O access side.
There's none directly but we need to get the data into blocks before we can do bulk I/O (or do complicated gather bulk I/O).
So, this is the preliminary work for implementing the bulk I/O? If so, it's worth to consider once whether implementing in the rb-tree cache code is the right choice. Can it be implemented in the cache management core, since you'll need an API anyway for getting the bulk register array via cache manager?
It can be that implementing in each cache backend code is the best choice in the end, but an overlook from a high place is always good before going forward.
Takashi