On Mon, Jun 08, 2015 at 04:41:23PM +0530, Vinod Koul wrote: > On Tue, Jun 02, 2015 at 03:59:48PM +0200, Maxime Ripard wrote: > > On Tue, Jun 02, 2015 at 06:39:48PM +0530, Vinod Koul wrote: > > > > > Do we have a real world example where we would want to DMA without a kernel > > > > > client, IOW a usermode client? > > > > > > > > memset/memcpy on large buffers ? > > > > > > > > And like I mentionned on the memset reintroduction serie, we do have a > > > > platform where we exactly want to do framebuffer-related operations > > > > (blitting for example, but also memset on the framebuffer itself) > > > > which usually are accelerated by the GPU, but that platform doesn't > > > > have any GPU while it has a DMA controller able to do just that. > > > > > > So you dont have a framebuffer driver which can invoke dmaengine ops? > > > > We do, but that would require adding possibly controversial ioctls, > > without covering the general use case (ie. offloading the copying > > and/or setting of arbitrarily large buffers from userspace). > > Yes but without a client who would like the dma arbitrarily large buffers? > That part is not clear to me yet. The "client" would be the application itself. -- Maxime Ripard, Free Electrons Embedded Linux, Kernel and Android engineering http://free-electrons.com
Attachment:
signature.asc
Description: Digital signature