On 04/01/2012 14:58, Simon Schwarz wrote:

Hi Stefano,

Hi Simon,

Just FYI: I already did this with DMA - the patch was rejected and I haven't worked on it since (as you already said it is some work...).

Yes, of course - it makes sense to find a solution that will be accepted before implementing it - this is the reason I prepared a way to enable caches, but I have not yet done and I will not do if this solution will be rejected.

It was really speeding up the copy process.

I am sure it is a big improvement...

I have some meassurments on this and it would be really interesting to compare them with the d-cache solution.

If you are interested I can send you my data - so you can compare it when you have d-cache ready.

Ok, understood. Let's see if we can merge soon the patchset into mainline - that opens also the possibility to add further SOCs, not only TI. And then we can proceed optimizing with DMA or d-cache.

Best regards, Stefano