[Intel-gfx] [PATCH] drm/i915/execlists: Terminate the context image with BB_END

Lionel Landwerlin lionel.g.landwerlin at intel.com
Tue Jul 31 12:47:32 UTC 2018


On 30/07/18 17:43, Chris Wilson wrote:
> In the aub trace utility, the context images are terminated with a
> MI_BATCH_BUFFER_END; the simulator is reported as complaining otherwise.
> Do the same for our protocontext image for completeness, and in passing
> apply the magic bit for gen10 to mark the end of the context image.
>
> Reported-by: Lionel Landwerlin <lionel.g.landwerlin at intel.com>
> Signed-off-by: Chris Wilson <chris at chris-wilson.co.uk>
> Cc: Lionel Landwerlin <lionel.g.landwerlin at intel.com>

Doesn't look like anything exploded.
Have you noticed any improvement maybe in the benchmarks?

Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin at intel.com>

> ---
>   drivers/gpu/drm/i915/intel_lrc.c     | 4 ++++
>   drivers/gpu/drm/i915/intel_lrc_reg.h | 2 +-
>   2 files changed, 5 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
> index fad689efb67a..b0be180c6294 100644
> --- a/drivers/gpu/drm/i915/intel_lrc.c
> +++ b/drivers/gpu/drm/i915/intel_lrc.c
> @@ -2653,6 +2653,10 @@ static void execlists_init_reg_state(u32 *regs,
>   
>   		i915_oa_init_reg_state(engine, ctx, regs);
>   	}
> +
> +	regs[CTX_END] = MI_BATCH_BUFFER_END;
> +	if (INTEL_GEN(dev_priv) >= 10)
> +		regs[CTX_END] |= BIT(0);
>   }
>   
>   static int
> diff --git a/drivers/gpu/drm/i915/intel_lrc_reg.h b/drivers/gpu/drm/i915/intel_lrc_reg.h
> index 169a2239d6c7..5ef932d810a7 100644
> --- a/drivers/gpu/drm/i915/intel_lrc_reg.h
> +++ b/drivers/gpu/drm/i915/intel_lrc_reg.h
> @@ -37,7 +37,7 @@
>   #define CTX_PDP0_LDW			0x32
>   #define CTX_LRI_HEADER_2		0x41
>   #define CTX_R_PWR_CLK_STATE		0x42
> -#define CTX_GPGPU_CSR_BASE_ADDRESS	0x44
> +#define CTX_END				0x44
>   
>   #define CTX_REG(reg_state, pos, reg, val) do { \
>   	u32 *reg_state__ = (reg_state); \




More information about the Intel-gfx mailing list