> You're always calling gen8_write_pdp_ctx() and gen8_write_tail_ctx()
> together, kmapping the page twice is a bit wastful.
You are totally right... I´ll join them in the next version.
-- Oscar
___
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
On Thu, Mar 27, 2014 at 06:00:08PM +, oscar.ma...@intel.com wrote:
> +static int gen8_write_pdp_ctx(struct i915_hw_context *ctx,
> +struct i915_hw_ppgtt *ppgtt)
> +{
> + struct page *page;
> + uint32_t *reg_state;
> +
> + page = i915_gem_object_get_pa
Bummer. I´ll fix it on the next version.
Thanks!
Oscar
> -Original Message-
> From: Lespiau, Damien
> Sent: Monday, March 31, 2014 5:43 PM
> To: Mateo Lozano, Oscar
> Cc: intel-gfx@lists.freedesktop.org
> Subject: Re: [Intel-gfx] [PATCH 39/49] drm/i915/bdw: Swap the
On Thu, Mar 27, 2014 at 06:00:08PM +, oscar.ma...@intel.com wrote:
> + reg_state[CTX_PDP3_UDW+1] = ppgtt->pd_dma_addr[3] >> 32;
> + reg_state[CTX_PDP3_LDW+1] = ppgtt->pd_dma_addr[3];
> + reg_state[CTX_PDP2_UDW+1] = ppgtt->pd_dma_addr[2] >> 32;
> + reg_state[CTX_PDP2_LDW+1] = ppg
From: Oscar Mateo
Each logical ring context has the PDPs in the context object, so update
them before submission. This should work both for Aliasing PPGTT
(nothing will be changed) and Full PPGTT.
Also, don't write PDP in the legacy way when using logical ring contexts
(this is mostly for correc
