On Fri, Dec 7, 2012 at 9:33 PM, Aaron Plattner wrote:
> Ah, I see what you mean. This would also need either a pointer to the drv
> directly so the helpers can lock drv->struct_mutex, or a helper function to
> convert from a drm_prime_helper_obj* to a gem_buffer_object* in a
> driver-specific way
On Fri, Dec 07, 2012 at 09:58:38AM -0800, Aaron Plattner wrote:
> On 12/06/2012 01:46 PM, Daniel Vetter wrote:
[snip]
> >- Just an idea for all the essential noop cpu helpers: Maybe just call
> > them dma_buf*noop and shovel them as convenience helpers into the
> > dma-buf.c code in the core,
Op 06-12-12 22:50, Daniel Vetter schreef:
> On Thu, Dec 06, 2012 at 10:46:30PM +0100, Daniel Vetter wrote:
>> On Thu, Dec 06, 2012 at 10:07:48AM -0800, Aaron Plattner wrote:
>>> Instead of reimplementing all of the dma_buf functionality in every driver,
>>> create helpers drm_prime_import and drm_p
On 12/07/2012 01:38 PM, Daniel Vetter wrote:
On Fri, Dec 7, 2012 at 9:33 PM, Aaron Plattner wrote:
Ah, I see what you mean. This would also need either a pointer to the drv
directly so the helpers can lock drv->struct_mutex, or a helper function to
convert from a drm_prime_helper_obj* to a gem
On 12/07/2012 01:38 PM, Daniel Vetter wrote:
> On Fri, Dec 7, 2012 at 9:33 PM, Aaron Plattner
> wrote:
>> Ah, I see what you mean. This would also need either a pointer to the drv
>> directly so the helpers can lock drv->struct_mutex, or a helper function to
>> convert from a drm_prime_helper_ob
On Fri, Dec 7, 2012 at 9:33 PM, Aaron Plattner wrote:
> Ah, I see what you mean. This would also need either a pointer to the drv
> directly so the helpers can lock drv->struct_mutex, or a helper function to
> convert from a drm_prime_helper_obj* to a gem_buffer_object* in a
> driver-specific way
On 12/07/2012 10:48 AM, Daniel Vetter wrote:
On Fri, Dec 07, 2012 at 09:58:38AM -0800, Aaron Plattner wrote:
On 12/06/2012 01:46 PM, Daniel Vetter wrote:
- To make it a first-class helper and remove any and all midlayer stints
from this (I truly loath midlayers ...) maybe shovel this into a
On 12/07/2012 10:48 AM, Daniel Vetter wrote:
> On Fri, Dec 07, 2012 at 09:58:38AM -0800, Aaron Plattner wrote:
>> On 12/06/2012 01:46 PM, Daniel Vetter wrote:
>>> - To make it a first-class helper and remove any and all midlayer stints
>>>from this (I truly loath midlayers ...) maybe shovel thi
On Fri, Dec 07, 2012 at 09:58:38AM -0800, Aaron Plattner wrote:
> On 12/06/2012 01:46 PM, Daniel Vetter wrote:
[snip]
> >- Just an idea for all the essential noop cpu helpers: Maybe just call
> > them dma_buf*noop and shovel them as convenience helpers into the
> > dma-buf.c code in the core,
On 12/06/2012 01:46 PM, Daniel Vetter wrote:
On Thu, Dec 06, 2012 at 10:07:48AM -0800, Aaron Plattner wrote:
Instead of reimplementing all of the dma_buf functionality in every driver,
create helpers drm_prime_import and drm_prime_export that implement them in
terms of new, lower-level hook func
On 12/06/2012 01:46 PM, Daniel Vetter wrote:
> On Thu, Dec 06, 2012 at 10:07:48AM -0800, Aaron Plattner wrote:
>> Instead of reimplementing all of the dma_buf functionality in every driver,
>> create helpers drm_prime_import and drm_prime_export that implement them in
>> terms of new, lower-level h
Op 06-12-12 22:50, Daniel Vetter schreef:
> On Thu, Dec 06, 2012 at 10:46:30PM +0100, Daniel Vetter wrote:
>> On Thu, Dec 06, 2012 at 10:07:48AM -0800, Aaron Plattner wrote:
>>> Instead of reimplementing all of the dma_buf functionality in every driver,
>>> create helpers drm_prime_import and drm_p
On Thu, Dec 06, 2012 at 10:46:30PM +0100, Daniel Vetter wrote:
> On Thu, Dec 06, 2012 at 10:07:48AM -0800, Aaron Plattner wrote:
> > Instead of reimplementing all of the dma_buf functionality in every driver,
> > create helpers drm_prime_import and drm_prime_export that implement them in
> > terms
On Thu, Dec 06, 2012 at 10:07:48AM -0800, Aaron Plattner wrote:
> Instead of reimplementing all of the dma_buf functionality in every driver,
> create helpers drm_prime_import and drm_prime_export that implement them in
> terms of new, lower-level hook functions:
>
> gem_prime_pin: callback when
On Thu, Dec 06, 2012 at 10:46:30PM +0100, Daniel Vetter wrote:
> On Thu, Dec 06, 2012 at 10:07:48AM -0800, Aaron Plattner wrote:
> > Instead of reimplementing all of the dma_buf functionality in every driver,
> > create helpers drm_prime_import and drm_prime_export that implement them in
> > terms
On Thu, Dec 06, 2012 at 10:07:48AM -0800, Aaron Plattner wrote:
> Instead of reimplementing all of the dma_buf functionality in every driver,
> create helpers drm_prime_import and drm_prime_export that implement them in
> terms of new, lower-level hook functions:
>
> gem_prime_pin: callback when
Instead of reimplementing all of the dma_buf functionality in every driver,
create helpers drm_prime_import and drm_prime_export that implement them in
terms of new, lower-level hook functions:
gem_prime_pin: callback when a buffer is created, used to pin buffers into GTT
gem_prime_get_pages:
Instead of reimplementing all of the dma_buf functionality in every driver,
create helpers drm_prime_import and drm_prime_export that implement them in
terms of new, lower-level hook functions:
gem_prime_pin: callback when a buffer is created, used to pin buffers into GTT
gem_prime_get_pages:
18 matches
Mail list logo
