On 03/05/16 01:10, Kenneth Graunke wrote:
> On Friday, April 29, 2016 1:29:53 PM PDT Samuel Iglesias Gonsálvez wrote:
>> When there is a mix of definitions of uniforms with 32-bit or 64-bit
>> data type sizes, the driver ends up doing misaligned access to double
>> based variables in the push
On Monday, May 2, 2016 4:10:38 PM PDT Kenneth Graunke wrote:
> The idea of "upload all the 64-bit things first, add 0 or 1 padding
> slots, then upload all the 32-bit things" also seems like it could
> simplify this code a lot.
Sorry, I misspoke - if you upload all the 64-bit things first, then
On Friday, April 29, 2016 1:29:53 PM PDT Samuel Iglesias Gonsálvez wrote:
> When there is a mix of definitions of uniforms with 32-bit or 64-bit
> data type sizes, the driver ends up doing misaligned access to double
> based variables in the push constant buffer.
>
> To fix this, the driver adds
On Mon, May 02, 2016 at 04:28:18PM +0300, Pohjolainen, Topi wrote:
> On Fri, Apr 29, 2016 at 01:29:53PM +0200, Samuel Iglesias Gons?lvez wrote:
> > When there is a mix of definitions of uniforms with 32-bit or 64-bit
> > data type sizes, the driver ends up doing misaligned access to double
> >
On Fri, Apr 29, 2016 at 01:29:53PM +0200, Samuel Iglesias Gons?lvez wrote:
> When there is a mix of definitions of uniforms with 32-bit or 64-bit
> data type sizes, the driver ends up doing misaligned access to double
> based variables in the push constant buffer.
>
> To fix this, the driver adds
On Sun, May 01, 2016 at 07:42:42PM -0700, Jordan Justen wrote:
> On 2016-04-29 04:29:53, Samuel Iglesias Gonsálvez wrote:
> > When there is a mix of definitions of uniforms with 32-bit or 64-bit
> > data type sizes, the driver ends up doing misaligned access to double
> > based variables in the
On 2016-04-29 04:29:53, Samuel Iglesias Gonsálvez wrote:
> When there is a mix of definitions of uniforms with 32-bit or 64-bit
> data type sizes, the driver ends up doing misaligned access to double
> based variables in the push constant buffer.
>
> To fix this, the driver adds padding when
When there is a mix of definitions of uniforms with 32-bit or 64-bit
data type sizes, the driver ends up doing misaligned access to double
based variables in the push constant buffer.
To fix this, the driver adds padding when needed in the push constant
buffer and takes it into account to avoid