On Mon, 2020-02-10 at 12:28 -0800, Matt Roper wrote:
> We need to explicitly set the TLB Request Timer initial value in the
> BW_BUDDY registers to 0x8 rather than relying on the hardware
> default.
>
> Bspec: 52890
> Bspec: 50044
> Cc: Stanislav Lisovskiy <stanislav.lisovs...@intel.com>
> Signed-off-by: Matt Roper <matthew.d.ro...@intel.com>
> ---
> drivers/gpu/drm/i915/display/intel_display_power.c | 6 ++++++
> drivers/gpu/drm/i915/i915_reg.h | 1 +
> 2 files changed, 7 insertions(+)
>
> diff --git a/drivers/gpu/drm/i915/display/intel_display_power.c
> b/drivers/gpu/drm/i915/display/intel_display_power.c
> index 6e55e88ca0d0..2ad2694a28e6 100644
> --- a/drivers/gpu/drm/i915/display/intel_display_power.c
> +++ b/drivers/gpu/drm/i915/display/intel_display_power.c
> @@ -5001,6 +5001,12 @@ static void tgl_bw_buddy_init(struct
> drm_i915_private *dev_priv)
> table[i].page_mask);
> intel_de_write(dev_priv, BW_BUDDY2_PAGE_MASK,
> table[i].page_mask);
> +
> + /* Wa_22010178259:tgl */
> + intel_de_rmw(dev_priv, BW_BUDDY1_CTL,
> + BW_BUDDY_TLB_REQ_TIMER_MASK, 0x8);
> + intel_de_rmw(dev_priv, BW_BUDDY2_CTL,
> + BW_BUDDY_TLB_REQ_TIMER_MASK, 0x8);
This will clear BW_BUDDY_TLB_REQ_TIMER_MASK and set 0x8, while it
should be 0x8 << 16(adding a macro for that)
> }
> }
>
> diff --git a/drivers/gpu/drm/i915/i915_reg.h
> b/drivers/gpu/drm/i915/i915_reg.h
> index a6847f26678c..d829b9ae8402 100644
> --- a/drivers/gpu/drm/i915/i915_reg.h
> +++ b/drivers/gpu/drm/i915/i915_reg.h
> @@ -7766,6 +7766,7 @@ enum {
> #define BW_BUDDY1_CTL _MMIO(0x45140)
> #define BW_BUDDY2_CTL _MMIO(0x45150)
> #define BW_BUDDY_DISABLE REG_BIT(31)
> +#define BW_BUDDY_TLB_REQ_TIMER_MASK REG_GENMASK(21, 16)
>
> #define BW_BUDDY1_PAGE_MASK _MMIO(0x45144)
> #define BW_BUDDY2_PAGE_MASK _MMIO(0x45154)
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
