On 11/21/2014 03:51 PM, Chris Wilson wrote: > On Fri, Nov 21, 2014 at 03:48:33PM +0100, Thomas Hellstrom wrote: >> On 11/21/2014 03:14 PM, Chris Wilson wrote: >>> On Thu, Nov 20, 2014 at 09:56:25AM +0100, Thomas Hellstrom wrote: >>>> It happens on occasion that developers of generic user-space applications >>>> abuse the dumb buffer API to get hold of drm buffers that they can both >>>> mmap() and use for GPU acceleration, using the assumptions that dumb buffers >>>> and buffers available for GPU are >>>> a) The same type and can be aribtrarily type-casted. >>>> b) fully coherent. >>> Both (a) and (b) are true for intel and it turns out to be a requirement >>> for smooth transitions from the boot splash screens into X, and relied >>> upon by userspace. >>> -Chris >>> >> So when you say relied upon by user-space, do you mean generic >> user-space or driver-specific user-space? >> >> With that, I mean what component is responsible for deciding that the >> dumb buffer can be accelerated? The Intel xorg driver? > There is no way for the driver to know it has a dumb buffer. It copies > the contents of the current framebuffer onto its replacement framebuffer > so that it can do a seamless switch. Sure, but inside the driver is the only place this is happening, right? It's not happening in generic code? If it's in the driver, it's legitimate, and my patch incorrect, because the driver should really be allowed to typecast any buffer... /Thomas > -Chris > _______________________________________________ dri-devel mailing list dri-devel@xxxxxxxxxxxxxxxxxxxxx http://lists.freedesktop.org/mailman/listinfo/dri-devel
