Hi Sebastian,
On Fri, Aug 01, 2025 at 09:48:54AM +0000, Sebastian Brzezinka wrote:
> Relocated the workaround for "Disable Repacking for Compression" from
> rcs_engine_wa_init() to icl_ctx_workarounds_init() for Jasper Lake and
> Elkhart Lake platforms. This ensures the WA is applied during context
> initialization.
I would re-write this to make it a bit more understandable:
CACHE_MODE_0 registers should be saved and restored as part of
the context, not during engine reset. Move the related workaround
from rcs_engine_wa_init() to icl_ctx_workarounds_init() for
Jasper Lake and Elkhart Lake platforms. This ensures the WA is
applied during context initialisation.
How does it look to you?
> Signed-off-by: Sebastian Brzezinka <sebastian.brzezi...@intel.com>
BSPEC: 55884
Fixes: 0ddae025ab6c ("drm/i915: Disable compression tricks on JSL")
Closes: Fixes: 0ddae025ab6c ("drm/i915: Disable compression tricks on JSL")
Cc: <sta...@vger.kernel.org> # v6.13+
> ---
> drivers/gpu/drm/i915/gt/intel_workarounds.c | 20 +++++++++++---------
> 1 file changed, 11 insertions(+), 9 deletions(-)
>
> diff --git a/drivers/gpu/drm/i915/gt/intel_workarounds.c
> b/drivers/gpu/drm/i915/gt/intel_workarounds.c
> index b37e400f74e5..79298618cda6 100644
> --- a/drivers/gpu/drm/i915/gt/intel_workarounds.c
> +++ b/drivers/gpu/drm/i915/gt/intel_workarounds.c
> @@ -634,6 +634,8 @@ static void cfl_ctx_workarounds_init(struct
> intel_engine_cs *engine,
> static void icl_ctx_workarounds_init(struct intel_engine_cs *engine,
> struct i915_wa_list *wal)
> {
> + struct drm_i915_private *i915 = engine->i915;
> +
> /* Wa_1406697149 (WaDisableBankHangMode:icl) */
> wa_write(wal, GEN8_L3CNTLREG, GEN8_ERRDETBCTRL);
>
> @@ -669,6 +671,15 @@ static void icl_ctx_workarounds_init(struct
> intel_engine_cs *engine,
>
> /* Wa_1406306137:icl,ehl */
> wa_mcr_masked_en(wal, GEN9_ROW_CHICKEN4, GEN11_DIS_PICK_2ND_EU);
> +
> + if (IS_JASPERLAKE(i915) || IS_ELKHARTLAKE(i915)) {
> + /*
> + * "Disable Repacking for Compression (masked R/W access)
> + * before rendering compressed surfaces for display."
As we are here, I would remove the quotes here, they don't mean
much.
With all the above:
Reviewed-by: Andi Shyti <andi.sh...@linux.intel.com>
Thanks,
Andi
