diff options
| author | Chris Wilson <chris@chris-wilson.co.uk> | 2020-06-11 09:01:36 +0100 |
|---|---|---|
| committer | Joonas Lahtinen <joonas.lahtinen@linux.intel.com> | 2020-06-16 11:34:24 +0300 |
| commit | 7237b190add0794bd95979018a23eda698f2705d (patch) | |
| tree | 4ed83f249e21bd6b437e71a0695a75e004ad0ad4 /drivers/gpu/drm/i915/gt | |
| parent | ef50fa9bd17d13d0611e39e13b37bbd3e1ea50bf (diff) | |
| download | linux-7237b190add0794bd95979018a23eda698f2705d.tar.bz2 | |
drm/i915/gt: Move ivb GT workarounds from init_clock_gating to workarounds
Rescue the GT workarounds from being buried inside init_clock_gating so
that we remember to apply them after a GT reset, and that they are
included in our verification that the workarounds are applied.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: stable@vger.kernel.org
Link: https://patchwork.freedesktop.org/patch/msgid/20200611080140.30228-2-chris@chris-wilson.co.uk
(cherry picked from commit 19f1f627b33385a2f0855cbc7d33d86d7f4a1e78)
Signed-off-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Diffstat (limited to 'drivers/gpu/drm/i915/gt')
| -rw-r--r-- | drivers/gpu/drm/i915/gt/intel_workarounds.c | 62 |
1 files changed, 62 insertions, 0 deletions
diff --git a/drivers/gpu/drm/i915/gt/intel_workarounds.c b/drivers/gpu/drm/i915/gt/intel_workarounds.c index b5d0fbe3b211..d9b72679ca82 100644 --- a/drivers/gpu/drm/i915/gt/intel_workarounds.c +++ b/drivers/gpu/drm/i915/gt/intel_workarounds.c @@ -693,6 +693,66 @@ int intel_engine_emit_ctx_wa(struct i915_request *rq) } static void +ivb_gt_workarounds_init(struct drm_i915_private *i915, struct i915_wa_list *wal) +{ + /* WaDisableEarlyCull:ivb */ + wa_masked_en(wal, _3D_CHICKEN3, _3D_CHICKEN_SF_DISABLE_OBJEND_CULL); + + /* WaDisablePSDDualDispatchEnable:ivb */ + if (IS_IVB_GT1(i915)) + wa_masked_en(wal, + GEN7_HALF_SLICE_CHICKEN1, + GEN7_PSD_SINGLE_PORT_DISPATCH_ENABLE); + + /* WaDisable_RenderCache_OperationalFlush:ivb */ + wa_masked_dis(wal, CACHE_MODE_0_GEN7, RC_OP_FLUSH_ENABLE); + + /* Apply the WaDisableRHWOOptimizationForRenderHang:ivb workaround. */ + wa_masked_dis(wal, + GEN7_COMMON_SLICE_CHICKEN1, + GEN7_CSC1_RHWO_OPT_DISABLE_IN_RCC); + + /* WaApplyL3ControlAndL3ChickenMode:ivb */ + wa_write(wal, GEN7_L3CNTLREG1, GEN7_WA_FOR_GEN7_L3_CONTROL); + wa_write(wal, GEN7_L3_CHICKEN_MODE_REGISTER, GEN7_WA_L3_CHICKEN_MODE); + + /* WaForceL3Serialization:ivb */ + wa_write_clr(wal, GEN7_L3SQCREG4, L3SQ_URB_READ_CAM_MATCH_DISABLE); + + /* + * WaVSThreadDispatchOverride:ivb,vlv + * + * This actually overrides the dispatch + * mode for all thread types. + */ + wa_write_masked_or(wal, GEN7_FF_THREAD_MODE, + GEN7_FF_SCHED_MASK, + GEN7_FF_TS_SCHED_HW | + GEN7_FF_VS_SCHED_HW | + GEN7_FF_DS_SCHED_HW); + + if (0) { /* causes HiZ corruption on ivb:gt1 */ + /* enable HiZ Raw Stall Optimization */ + wa_masked_dis(wal, CACHE_MODE_0_GEN7, HIZ_RAW_STALL_OPT_DISABLE); + } + + /* WaDisable4x2SubspanOptimization:ivb */ + wa_masked_en(wal, CACHE_MODE_1, PIXEL_SUBSPAN_COLLECT_OPT_DISABLE); + + /* + * BSpec recommends 8x4 when MSAA is used, + * however in practice 16x4 seems fastest. + * + * Note that PS/WM thread counts depend on the WIZ hashing + * disable bit, which we don't touch here, but it's good + * to keep in mind (see 3DSTATE_PS and 3DSTATE_WM). + */ + wa_add(wal, GEN7_GT_MODE, 0, + _MASKED_FIELD(GEN6_WIZ_HASHING_MASK, GEN6_WIZ_HASHING_16x4), + GEN6_WIZ_HASHING_16x4); +} + +static void hsw_gt_workarounds_init(struct drm_i915_private *i915, struct i915_wa_list *wal) { /* L3 caching of data atomics doesn't work -- disable it. */ @@ -1011,6 +1071,8 @@ gt_init_workarounds(struct drm_i915_private *i915, struct i915_wa_list *wal) skl_gt_workarounds_init(i915, wal); else if (IS_HASWELL(i915)) hsw_gt_workarounds_init(i915, wal); + else if (IS_IVYBRIDGE(i915)) + ivb_gt_workarounds_init(i915, wal); else if (INTEL_GEN(i915) <= 8) return; else |