Workarounds for Icelake
Subsystems (2)
v4 cover Workarounds for Icelake
List of GT workarounds for Icelake that we have been carrying in internal. Can we get eyes on these please?
Oscar Mateo (22):
drm/i915/icl: Introduce initial Icelake Workarounds drm/i915/icl: Enable Sampler DFR drm/i915/icl: WaGAPZPriorityScheme drm/i915/icl: WaL3BankAddressHashing drm/i915/icl: WaModifyGamTlbPartitioning drm/i915/icl: WaDisableCleanEvicts drm/i915/icl: WaCL2SFHalfMaxAlloc drm/i915/icl: WaDisCtxReload
drm/i915/icl: Wa_1405779004
drm/i915/icl: Wa_1406680159 and Wa_2201832410 drm/i915/icl: Wa_1604302699
drm/i915/icl: Wa_1406838659
drm/i915/icl: WaForwardProgressSoftReset drm/i915/icl: WaDisableImprovedTdlClkGating drm/i915/icl: WaEnableStateCacheRedirectToCS drm/i915/icl: Wa_2006665173
drm/i915/icl: WaEnableFloatBlendOptimization drm/i915/icl: WaSendPushConstantsFromMMIO drm/i915/icl: WaAllowUMDToModifyHalfSliceChicken2 drm/i915/icl: WaAllowUMDToModifyHalfSliceChicken7 drm/i915/icl: WaAllowUmdWriteTRTTRootTable drm/i915/icl: WaAllowUMDToModifySamplerMode
drivers/gpu/drm/i915/i915_drv.h | 9 ++ drivers/gpu/drm/i915/i915_gem_gtt.c | 4 +- drivers/gpu/drm/i915/i915_reg.h | 85 +++++++++++++---- drivers/gpu/drm/i915/intel_lrc.c | 2 + drivers/gpu/drm/i915/intel_pm.c | 4 +- drivers/gpu/drm/i915/intel_uncore.c | 7 +- drivers/gpu/drm/i915/intel_workarounds.c | 157 +++++++++++++++++++++++++++++++ 7 files changed, 245 insertions(+), 23 deletions(-)
Signature · 6 lines
1.9.1 _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
PATCH v4 1/22 drm/i915/icl: Introduce initial Icelake Workarounds
Inherit workarounds from previous platforms that are still valid for Icelake.
v2: GEN7_ROW_CHICKEN2 is masked
v3:
- Since it has been fixed already in upstream, removed the TODO comment about WA_SET_BIT for WaInPlaceDecompressionHang.
- Squashed with this patch:
drm/i915/icl: add icelake_init_clock_gating()
from Paulo Zanoni <[email protected]>
- Squashed with this patch:
drm/i915/icl: WaForceEnableNonCoherent
from Oscar Mateo <[email protected]>
- WaPushConstantDereferenceHoldDisable is now Wa_1604370585 and applies to B0 as well.
- WaPipeControlBefore3DStateSamplePattern WABB was being applied to ICL incorrectly.
v4:
- Wrap the commit message
- s/dev_priv/p to please checkpatch
v5: Rebased on top of the WA refactoring
v6: Rebased on top of further whitelist registers refactoring (Michel)
v7: Added WaRsForcewakeAddDelayForAck
--- drivers/gpu/drm/i915/i915_drv.h | 9 +++++++ drivers/gpu/drm/i915/i915_gem_gtt.c | 4 +-- drivers/gpu/drm/i915/i915_reg.h | 1 + drivers/gpu/drm/i915/intel_lrc.c | 2 ++ drivers/gpu/drm/i915/intel_pm.c | 4 ++- drivers/gpu/drm/i915/intel_uncore.c | 7 +++-- drivers/gpu/drm/i915/intel_workarounds.c | 46 ++++++++++++++++++++++++++++++++ 7 files changed, 68 insertions(+), 5 deletions(-)
drivers/gpu/drm/i915/i915_drv.h +9 -0
@@ -2462,6 +2462,15 @@ static inline unsigned int i915_sg_segment_size(void)#define IS_CNL_REVID (p, since, until) \ (IS_CANNONLAKE(p) && IS_REVID(p, since, until)) +#define ICL_REVID_A0 0x0 +#define ICL_REVID_A2 0x1 +#define ICL_REVID_B0 0x3 +#define ICL_REVID_B2 0x4 +#define ICL_REVID_C0 0x5 + +#define IS_ICL_REVID (p, since, until) \ + (IS_ICELAKE(p) && IS_REVID(p, since, until)) +/* * The genX designation typically refers to the render engine, so render * capability related checks should use IS_GEN, while display and other checks
drivers/gpu/drm/i915/i915_gem_gtt.c +2 -2
@@ -2140,12 +2140,12 @@ static void gtt_write_workarounds(struct drm_i915_private *dev_priv)* called on driver load and after a GPU reset, so you can place * workarounds here even if they get overwritten by GPU reset. */ -/* WaIncreaseDefaultTLBEntries:chv,bdw,skl,bxt,kbl,glk,cfl,cnl */ +/* WaIncreaseDefaultTLBEntries:chv,bdw,skl,bxt,kbl,glk,cfl,cnl,icl */ if (IS_BROADWELL (dev_priv))I915_WRITE (GEN8_L3_LRA_1_GPGPU ,GEN8_L3_LRA_1_GPGPU_DEFAULT_VALUE_BDW );else if (IS_CHERRYVIEW (dev_priv))I915_WRITE (GEN8_L3_LRA_1_GPGPU ,GEN8_L3_LRA_1_GPGPU_DEFAULT_VALUE_CHV ); -else if (IS_GEN9_BC (dev_priv) ||IS_GEN10 (dev_priv)) +else if (IS_GEN9_BC (dev_priv) ||IS_GEN10 (dev_priv) ||IS_GEN11 (dev_priv))I915_WRITE (GEN8_L3_LRA_1_GPGPU ,GEN9_L3_LRA_1_GPGPU_DEFAULT_VALUE_SKL );else if (IS_GEN9_LP (dev_priv))I915_WRITE (GEN8_L3_LRA_1_GPGPU ,GEN9_L3_LRA_1_GPGPU_DEFAULT_VALUE_BXT );
drivers/gpu/drm/i915/i915_reg.h +1 -0
@@ -7203,6 +7203,7 @@ enum {/* GEN8 chicken */ #define HDC_CHICKEN0 _MMIO(0x7300)#define CNL_HDC_CHICKEN0 _MMIO(0xE5F0) +#define ICL_HDC_CHICKEN0 _MMIO(0xE5F4)#define HDC_FORCE_CSR_NON_COHERENT_OVR_DISABLE (1<<15)#define HDC_FENCE_DEST_SLM_DISABLE (1<<14)#define HDC_DONOT_FETCH_MEM_WHEN_MASKED (1<<11)
drivers/gpu/drm/i915/intel_lrc.c +2 -0
@@ -1636,6 +1636,8 @@ static int intel_init_workaround_bb(struct intel_engine_cs *engine)return -EINVAL ;switch (INTEL_GEN (engine->i915)) { +case 11 : +return 0 ;case 10 : wa_bb_fn[0 ] = gen10_init_indirectctx_bb; wa_bb_fn[1 ] =NULL ;
drivers/gpu/drm/i915/intel_pm.c +3 -1
@@ -9123,7 +9123,9 @@ static void nop_init_clock_gating(struct drm_i915_private *dev_priv)*/ void intel_init_clock_gating_hooks (struct drm_i915_private *dev_priv) { -if (IS_CANNONLAKE (dev_priv)) +if (IS_ICELAKE (dev_priv)) + dev_priv->display.init_clock_gating = nop_init_clock_gating; +else if (IS_CANNONLAKE (dev_priv)) dev_priv->display.init_clock_gating = cnl_init_clock_gating;else if (IS_COFFEELAKE (dev_priv)) dev_priv->display.init_clock_gating = cfl_init_clock_gating;
drivers/gpu/drm/i915/intel_uncore.c +5 -2
@@ -139,7 +139,9 @@ enum ack_type {* in the hope that the original ack will be delivered along with * the fallback ack. * -* This workaround is described in HSDES #1604254524 +* This workaround is described in HSDES #1604254524 and it's known as: +* WaRsForcewakeAddDelayForAck:skl,bxt,kbl,glk,cfl,cnl,icl +* although the name is a bit misleading. */ pass =1 ;
@@ -1394,7 +1396,8 @@ static void intel_uncore_fw_domains_init(struct drm_i915_private *dev_priv)if (INTEL_GEN (dev_priv) >=11 ) {int i; - dev_priv->uncore.funcs.force_wake_get = fw_domains_get; + dev_priv->uncore.funcs.force_wake_get = + fw_domains_get_with_fallback; dev_priv->uncore.funcs.force_wake_put = fw_domains_put; fw_domain_init(dev_priv, FW_DOMAIN_ID_RENDER, FORCEWAKE_RENDER_GEN9,
drivers/gpu/drm/i915/intel_workarounds.c +46 -0
82 lines not on this page · raw message
Signature · 6 lines
1.9.1 _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Re: [PATCH 01/22] drm/i915/icl: Introduce initial Icelake Workarounds
Quoted 42 lines · Oscar Mateo
Inherit workarounds from previous platforms that are still valid for Icelake.
v2:
GEN7_ROW_CHICKEN2is masked
v3:
- Since it has been fixed already in upstream, removed the TODO comment aboutWA_SET_BITfor WaInPlaceDecompressionHang.
- Squashed with this patch:
drm/i915/icl: addicelake_init_clock_gating()
from Paulo Zanoni <[email protected]>
- Squashed with this patch:
drm/i915/icl: WaForceEnableNonCoherent
from Oscar Mateo <[email protected]>
- WaPushConstantDereferenceHoldDisable is nowWa_1604370585and applies to B0 as well.
- WaPipeControlBefore3DStateSamplePattern WABB was being applied to ICL incorrectly.
v4:
- Wrap the commit message
- s/dev_priv/p to please checkpatch
v5: Rebased on top of the WA refactoring
v6: Rebased on top of further whitelist registers refactoring (Michel)
v7: Added WaRsForcewakeAddDelayForAckCc Tomasz Lis <[email protected]>Cc Mika Kuoppala <[email protected]>Signed-off-by Oscar Mateo <[email protected]>--- drivers/gpu/drm/i915/i915_drv.h | 9 +++++++ drivers/gpu/drm/i915/i915_gem_gtt.c | 4 +-- drivers/gpu/drm/i915/i915_reg.h | 1 + drivers/gpu/drm/i915/intel_lrc.c | 2 ++ drivers/gpu/drm/i915/intel_pm.c | 4 ++- drivers/gpu/drm/i915/intel_uncore.c | 7 +++-- drivers/gpu/drm/i915/intel_workarounds.c | 46 ++++++++++++++++++++++++++++++++ 7 files changed, 68 insertions(+), 5 deletions(-)drivers/gpu/drm/i915/i915_drv.h +0
-0
drivers/gpu/drm/i915/i915_drv.h static inline unsigned int i915_sg_segment_size(void) · line 2466
Quoted 5 lines
drivers/gpu/drm/i915/i915_drv.h +2
-0@@ -2462,6 +2462,15 @@ static inline unsigned int i915_sg_segment_size(void)#define IS_CNL_REVID (p, since, until) \ (IS_CANNONLAKE(p) && IS_REVID(p, since, until)) +#define ICL_REVID_A0 0x0 +#define ICL_REVID_A2 0x1
Just noted that for some reason bspec puts A0 and A2 under same revid. Bspec err?
Quoted 74 lines
drivers/gpu/drm/i915/i915_drv.h +7
-0+#define ICL_REVID_B0 0x3 +#define ICL_REVID_B2 0x4 +#define ICL_REVID_C0 0x5 + +#define IS_ICL_REVID (p, since, until) \ + (IS_ICELAKE(p) && IS_REVID(p, since, until)) +/* * The genX designation typically refers to the render engine, so render * capability related checks should use IS_GEN, while display and other checks drivers/gpu/drm/i915/i915_gem_gtt.c +2
-2@@ -2140,12 +2140,12 @@ static void gtt_write_workarounds(struct drm_i915_private *dev_priv)* called on driver load and after a GPU reset, so you can place * workarounds here even if they get overwritten by GPU reset. */ -/* WaIncreaseDefaultTLBEntries:chv,bdw,skl,bxt,kbl,glk,cfl,cnl */ +/* WaIncreaseDefaultTLBEntries:chv,bdw,skl,bxt,kbl,glk,cfl,cnl,icl */ if (IS_BROADWELL (dev_priv))I915_WRITE (GEN8_L3_LRA_1_GPGPU ,GEN8_L3_LRA_1_GPGPU_DEFAULT_VALUE_BDW );else if (IS_CHERRYVIEW (dev_priv))I915_WRITE (GEN8_L3_LRA_1_GPGPU ,GEN8_L3_LRA_1_GPGPU_DEFAULT_VALUE_CHV ); -else if (IS_GEN9_BC (dev_priv) ||IS_GEN10 (dev_priv)) +else if (IS_GEN9_BC (dev_priv) ||IS_GEN10 (dev_priv) ||IS_GEN11 (dev_priv))I915_WRITE (GEN8_L3_LRA_1_GPGPU ,GEN9_L3_LRA_1_GPGPU_DEFAULT_VALUE_SKL );else if (IS_GEN9_LP (dev_priv))I915_WRITE (GEN8_L3_LRA_1_GPGPU ,GEN9_L3_LRA_1_GPGPU_DEFAULT_VALUE_BXT );drivers/gpu/drm/i915/i915_reg.h +1
-0@@ -7203,6 +7203,7 @@ enum {/* GEN8 chicken */ #define HDC_CHICKEN0 _MMIO(0x7300)#define CNL_HDC_CHICKEN0 _MMIO(0xE5F0) +#define ICL_HDC_CHICKEN0 _MMIO(0xE5F4)#define HDC_FORCE_CSR_NON_COHERENT_OVR_DISABLE (1<<15)#define HDC_FENCE_DEST_SLM_DISABLE (1<<14)#define HDC_DONOT_FETCH_MEM_WHEN_MASKED (1<<11)drivers/gpu/drm/i915/intel_lrc.c +2
-0@@ -1636,6 +1636,8 @@ static int intel_init_workaround_bb(struct intel_engine_cs *engine)return -EINVAL ;switch (INTEL_GEN (engine->i915)) { +case 11 : +return 0 ;case 10 : wa_bb_fn[0 ] = gen10_init_indirectctx_bb; wa_bb_fn[1 ] =NULL ;drivers/gpu/drm/i915/intel_pm.c +3
-1@@ -9123,7 +9123,9 @@ static void nop_init_clock_gating(struct drm_i915_private *dev_priv)*/ void intel_init_clock_gating_hooks (struct drm_i915_private *dev_priv) { -if (IS_CANNONLAKE (dev_priv)) +if (IS_ICELAKE (dev_priv)) + dev_priv->display.init_clock_gating = nop_init_clock_gating; +else if (IS_CANNONLAKE (dev_priv)) dev_priv->display.init_clock_gating = cnl_init_clock_gating;else if (IS_COFFEELAKE (dev_priv)) dev_priv->display.init_clock_gating = cfl_init_clock_gating;drivers/gpu/drm/i915/intel_uncore.c +0
-0
drivers/gpu/drm/i915/intel_uncore.c enum ack_type { · line 144
Quoted 7 lines
drivers/gpu/drm/i915/intel_uncore.c +3
-1@@ -139,7 +139,9 @@ enum ack_type {* in the hope that the original ack will be delivered along with * the fallback ack. * -* This workaround is described in HSDES #1604254524 +* This workaround is described in HSDES #1604254524 and it's known as: +* WaRsForcewakeAddDelayForAck:skl,bxt,kbl,glk,cfl,cnl,icl +* although the name is a bit misleading.
Just for the record:
When I implemented this there was recommendation to do both, delaying for ack and then this method of using a reserver bit. My interpretation was that the delay was used as a first weapon to combat the issue. And then later, reserve bit method appeared.
I did not use WaRsForcewakeAddDelayForAck as I thought that this will be named differently. And also I think this method is a superset, making delaying irrelevant. As we fallback to reserve is we miss ack so no need to delay before polling. And adding delay to hotpath should be the last resort anyways.
I think this is the evolution of WaRsForcewakeAddDelayForAck (v2) and there is no better name, we should keep it.
Quoted 19 lines
drivers/gpu/drm/i915/intel_uncore.c +2
-1*/ pass =1 ;@@ -1394,7 +1396,8 @@ static void intel_uncore_fw_domains_init(struct drm_i915_private *dev_priv)if (INTEL_GEN (dev_priv) >=11 ) {int i; - dev_priv->uncore.funcs.force_wake_get = fw_domains_get; + dev_priv->uncore.funcs.force_wake_get = + fw_domains_get_with_fallback; dev_priv->uncore.funcs.force_wake_put = fw_domains_put; fw_domain_init(dev_priv, FW_DOMAIN_ID_RENDER, FORCEWAKE_RENDER_GEN9,drivers/gpu/drm/i915/intel_workarounds.c +0
-0@@ -441,6 +441,27 @@ static int cnl_ctx_workarounds_init(struct drm_i915_private *dev_priv)return 0 ;
drivers/gpu/drm/i915/intel_workarounds.c static int cnl_ctx_workarounds_init(struct drm_i915_private *dev_priv) · line 451
Quoted 10 lines
drivers/gpu/drm/i915/intel_workarounds.c +8
-0} +static int icl_ctx_workarounds_init (struct drm_i915_private *dev_priv) +{ +/* Wa_1604370585:icl (pre-prod) +* Formerly known as WaPushConstantDereferenceHoldDisable +*/ +if (IS_ICL_REVID (dev_priv,ICL_REVID_A0 ,ICL_REVID_B0 )) +WA_SET_BIT_MASKED (GEN7_ROW_CHICKEN2 , +PUSH_CONSTANT_DEREF_DISABLE );
Inherited from CNL and had to check if we have that on cnl. We do.
drivers/gpu/drm/i915/intel_workarounds.c static int cnl_ctx_workarounds_init(struct drm_i915_private *dev_priv) · line 460
Quoted 9 lines
drivers/gpu/drm/i915/intel_workarounds.c +9
-0+ +/* WaForceEnableNonCoherent:icl +* This is not the same workaround as in early Gen9 platforms, where +* lacking this could cause system hangs, but coherency performance +* overhead is high and only a few compute workloads really need it +* (the register is whitelisted in hardware now, so UMDs can opt in +* for coherency if they have a good reason). +*/ +WA_SET_BIT_MASKED (ICL_HDC_CHICKEN0 ,HDC_FORCE_NON_COHERENT );
Right, but the register name should be ICL_HDC_MODE.
Quoted 23 lines
drivers/gpu/drm/i915/intel_workarounds.c +9
-0+ +return 0 ; +} +int intel_ctx_workarounds_init (struct drm_i915_private *dev_priv) {int err =0 ;@@ -465,6 +486,8 @@ int intel_ctx_workarounds_init(struct drm_i915_private *dev_priv) err =cfl_ctx_workarounds_init (dev_priv); else if (IS_CANNONLAKE(dev_priv)) err =cnl_ctx_workarounds_init (dev_priv); + else if (IS_ICELAKE(dev_priv)) + err =icl_ctx_workarounds_init (dev_priv); elseMISSING_CASE (INTEL_GEN (dev_priv ));if (err)@@ -663,6 +686,21 @@ static void cnl_gt_workarounds_apply(struct drm_i915_private *dev_priv) _MASKED_BIT_ENABLE(GEN9_FFSC_PERCTX_PREEMPT_CTRL)); } +static void icl_gt_workarounds_apply (struct drm_i915_private *dev_priv) +{ +/* This is not an Wa. Enable for better image quality */
drivers/gpu/drm/i915/intel_workarounds.c static void cnl_gt_workarounds_apply(struct drm_i915_private *dev_priv) · line 701
Quoted 10 lines
drivers/gpu/drm/i915/intel_workarounds.c +10
-0+I915_WRITE (_3D_CHICKEN3, +_MASKED_BIT_ENABLE (_3D_CHICKEN3_AA_LINE_QUALITY_FIX_ENABLE)); + +/* WaInPlaceDecompressionHang:icl */ +I915_WRITE (GEN9_GAMT_ECO_REG_RW_IA , (I915_READ (GEN9_GAMT_ECO_REG_RW_IA ) | +GAMT_ECO_ENABLE_IN_PLACE_DECOMPRESS )); + +/* WaPipelineFlushCoherentLines:icl */ +I915_WRITE (GEN8_L3SQCREG4 , (I915_READ (GEN8_L3SQCREG4 ) | +GEN8_LQSC_FLUSH_COHERENT_LINES ));
Didn't find a HSDES entry for this. The workaround name and the reg/bit matches tho.
But the real question in here is that do we need to set this through indirect bb like we do with gen[8,9].
And just to note that cnl is missing this too. But that can be done as a followup when we first figure out that should we use the indirect bb for all >= gen8.
-Mika
Quoted 36 lines
drivers/gpu/drm/i915/intel_workarounds.c +10
-0+} +void intel_gt_workarounds_apply (struct drm_i915_private *dev_priv) {if (INTEL_GEN (dev_priv) <8 )@@ -683,6 +721,8 @@ void intel_gt_workarounds_apply(struct drm_i915_private *dev_priv)cfl_gt_workarounds_apply (dev_priv); else if (IS_CANNONLAKE(dev_priv)) cnl_gt_workarounds_apply(dev_priv); + else if (IS_ICELAKE(dev_priv)) + icl_gt_workarounds_apply(dev_priv); elseMISSING_CASE (INTEL_GEN (dev_priv )); }@@ -761,6 +801,10 @@ static void cnl_whitelist_build(struct whitelist *w)whitelist_reg (w,GEN8_CS_CHICKEN1 ); } +static void icl_whitelist_build (struct whitelist *w) +{ +} +static struct whitelist *whitelist_build (struct intel_engine_cs *engine,struct whitelist *w) {@@ -789,6 +833,8 @@ static struct whitelist *whitelist_build(struct intel_engine_cs *engine,cfl_whitelist_build (w); else if (IS_CANNONLAKE(i915)) cnl_whitelist_build(w); + else if (IS_ICELAKE(i915)) + icl_whitelist_build(w); elseMISSING_CASE (INTEL_GEN (i915 ));Signature · 1 line
1.9.1
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Re: [PATCH 01/22] drm/i915/icl: Introduce initial Icelake Workarounds
Quoted 47 lines · Mika Kuoppala
Quoted 45 lines · Oscar Mateo
Inherit workarounds from previous platforms that are still valid for Icelake.
v2:
GEN7_ROW_CHICKEN2is masked
v3:
- Since it has been fixed already in upstream, removed the TODO comment aboutWA_SET_BITfor WaInPlaceDecompressionHang.
- Squashed with this patch:
drm/i915/icl: addicelake_init_clock_gating()
from Paulo Zanoni <[email protected]>
- Squashed with this patch:
drm/i915/icl: WaForceEnableNonCoherent
from Oscar Mateo <[email protected]>
- WaPushConstantDereferenceHoldDisable is nowWa_1604370585and applies to B0 as well.
- WaPipeControlBefore3DStateSamplePattern WABB was being applied to ICL incorrectly.
v4:
- Wrap the commit message
- s/dev_priv/p to please checkpatch
v5: Rebased on top of the WA refactoring
v6: Rebased on top of further whitelist registers refactoring (Michel)
v7: Added WaRsForcewakeAddDelayForAckCc Tomasz Lis <[email protected]>Cc Mika Kuoppala <[email protected]>Signed-off-by Oscar Mateo <[email protected]>--- drivers/gpu/drm/i915/i915_drv.h | 9 +++++++ drivers/gpu/drm/i915/i915_gem_gtt.c | 4 +-- drivers/gpu/drm/i915/i915_reg.h | 1 + drivers/gpu/drm/i915/intel_lrc.c | 2 ++ drivers/gpu/drm/i915/intel_pm.c | 4 ++- drivers/gpu/drm/i915/intel_uncore.c | 7 +++-- drivers/gpu/drm/i915/intel_workarounds.c | 46 ++++++++++++++++++++++++++++++++ 7 files changed, 68 insertions(+), 5 deletions(-)drivers/gpu/drm/i915/i915_drv.h +0
-0@@ -2462,6 +2462,15 @@ static inline unsigned int i915_sg_segment_size(void)#define IS_CNL_REVID (p, since, until) \ (IS_CANNONLAKE(p) && IS_REVID(p, since, until))
Quoted 4 lines
drivers/gpu/drm/i915/i915_drv.h static inline unsigned int i915_sg_segment_size(void) · line 2466
Quoted 2 lines
drivers/gpu/drm/i915/i915_drv.h +2
-0+#define ICL_REVID_A0 0x0 +#define ICL_REVID_A2 0x1Just noted that for some reason bspec puts A0 and A2 under same revid. Bspec err?
That's what I hope. I have opened a bug against the BSpec to be 100% sure.
Quoted 88 lines
Quoted 74 lines
drivers/gpu/drm/i915/i915_drv.h +7
-0+#define ICL_REVID_B0 0x3 +#define ICL_REVID_B2 0x4 +#define ICL_REVID_C0 0x5 + +#define IS_ICL_REVID (p, since, until) \ + (IS_ICELAKE(p) && IS_REVID(p, since, until)) +/* * The genX designation typically refers to the render engine, so render * capability related checks should use IS_GEN, while display and other checks drivers/gpu/drm/i915/i915_gem_gtt.c +2
-2@@ -2140,12 +2140,12 @@ static void gtt_write_workarounds(struct drm_i915_private *dev_priv)* called on driver load and after a GPU reset, so you can place * workarounds here even if they get overwritten by GPU reset. */ -/* WaIncreaseDefaultTLBEntries:chv,bdw,skl,bxt,kbl,glk,cfl,cnl */ +/* WaIncreaseDefaultTLBEntries:chv,bdw,skl,bxt,kbl,glk,cfl,cnl,icl */ if (IS_BROADWELL (dev_priv))I915_WRITE (GEN8_L3_LRA_1_GPGPU ,GEN8_L3_LRA_1_GPGPU_DEFAULT_VALUE_BDW );else if (IS_CHERRYVIEW (dev_priv))I915_WRITE (GEN8_L3_LRA_1_GPGPU ,GEN8_L3_LRA_1_GPGPU_DEFAULT_VALUE_CHV ); -else if (IS_GEN9_BC (dev_priv) ||IS_GEN10 (dev_priv)) +else if (IS_GEN9_BC (dev_priv) ||IS_GEN10 (dev_priv) ||IS_GEN11 (dev_priv))I915_WRITE (GEN8_L3_LRA_1_GPGPU ,GEN9_L3_LRA_1_GPGPU_DEFAULT_VALUE_SKL );else if (IS_GEN9_LP (dev_priv))I915_WRITE (GEN8_L3_LRA_1_GPGPU ,GEN9_L3_LRA_1_GPGPU_DEFAULT_VALUE_BXT );drivers/gpu/drm/i915/i915_reg.h +1
-0@@ -7203,6 +7203,7 @@ enum {/* GEN8 chicken */ #define HDC_CHICKEN0 _MMIO(0x7300)#define CNL_HDC_CHICKEN0 _MMIO(0xE5F0) +#define ICL_HDC_CHICKEN0 _MMIO(0xE5F4)#define HDC_FORCE_CSR_NON_COHERENT_OVR_DISABLE (1<<15)#define HDC_FENCE_DEST_SLM_DISABLE (1<<14)#define HDC_DONOT_FETCH_MEM_WHEN_MASKED (1<<11)drivers/gpu/drm/i915/intel_lrc.c +2
-0@@ -1636,6 +1636,8 @@ static int intel_init_workaround_bb(struct intel_engine_cs *engine)return -EINVAL ;switch (INTEL_GEN (engine->i915)) { +case 11 : +return 0 ;case 10 : wa_bb_fn[0 ] = gen10_init_indirectctx_bb; wa_bb_fn[1 ] =NULL ;drivers/gpu/drm/i915/intel_pm.c +3
-1@@ -9123,7 +9123,9 @@ static void nop_init_clock_gating(struct drm_i915_private *dev_priv)*/ void intel_init_clock_gating_hooks (struct drm_i915_private *dev_priv) { -if (IS_CANNONLAKE (dev_priv)) +if (IS_ICELAKE (dev_priv)) + dev_priv->display.init_clock_gating = nop_init_clock_gating; +else if (IS_CANNONLAKE (dev_priv)) dev_priv->display.init_clock_gating = cnl_init_clock_gating;else if (IS_COFFEELAKE (dev_priv)) dev_priv->display.init_clock_gating = cfl_init_clock_gating;drivers/gpu/drm/i915/intel_uncore.c +0
-0drivers/gpu/drm/i915/intel_uncore.c enum ack_type { · line 144
Quoted 7 lines
drivers/gpu/drm/i915/intel_uncore.c +3
-1@@ -139,7 +139,9 @@ enum ack_type {* in the hope that the original ack will be delivered along with * the fallback ack. * -* This workaround is described in HSDES #1604254524 +* This workaround is described in HSDES #1604254524 and it's known as: +* WaRsForcewakeAddDelayForAck:skl,bxt,kbl,glk,cfl,cnl,icl +* although the name is a bit misleading. Just for the record:
When I implemented this there was recommendation to do both, delaying for ack and then this method of using a reserver bit. My interpretation was that the delay was used as a first weapon to combat the issue. And then later, reserve bit method appeared.
Quoted 8 lines
I did not use WaRsForcewakeAddDelayForAck as I thought that this will be named differently. And also I think this method is a superset, making delaying irrelevant. As we fallback to reserve is we miss ack so no need to delay before polling. And adding delay to hotpath should be the last resort anyways.
I think this is the evolution of WaRsForcewakeAddDelayForAck (v2) and there is no better name, we should keep it.
Yes, I read the mailing lists comments about this. AFAICT, there is no better name.
Quoted 31 lines
Quoted 19 lines
drivers/gpu/drm/i915/intel_uncore.c +2
-1*/ pass =1 ;@@ -1394,7 +1396,8 @@ static void intel_uncore_fw_domains_init(struct drm_i915_private *dev_priv)if (INTEL_GEN (dev_priv) >=11 ) {int i; - dev_priv->uncore.funcs.force_wake_get = fw_domains_get; + dev_priv->uncore.funcs.force_wake_get = + fw_domains_get_with_fallback; dev_priv->uncore.funcs.force_wake_put = fw_domains_put; fw_domain_init(dev_priv, FW_DOMAIN_ID_RENDER, FORCEWAKE_RENDER_GEN9,drivers/gpu/drm/i915/intel_workarounds.c +0
-0@@ -441,6 +441,27 @@ static int cnl_ctx_workarounds_init(struct drm_i915_private *dev_priv)return 0 ;drivers/gpu/drm/i915/intel_workarounds.c static int cnl_ctx_workarounds_init(struct drm_i915_private *dev_priv) · line 451
Quoted 10 lines
drivers/gpu/drm/i915/intel_workarounds.c +8
-0} +static int icl_ctx_workarounds_init (struct drm_i915_private *dev_priv) +{ +/* Wa_1604370585:icl (pre-prod) +* Formerly known as WaPushConstantDereferenceHoldDisable +*/ +if (IS_ICL_REVID (dev_priv,ICL_REVID_A0 ,ICL_REVID_B0 )) +WA_SET_BIT_MASKED (GEN7_ROW_CHICKEN2 , +PUSH_CONSTANT_DEREF_DISABLE );Inherited from CNL and had to check if we have that on cnl. We do.
Quoted 10 lines
drivers/gpu/drm/i915/intel_workarounds.c static int cnl_ctx_workarounds_init(struct drm_i915_private *dev_priv) · line 460
Quoted 9 lines
drivers/gpu/drm/i915/intel_workarounds.c +9
-0+ +/* WaForceEnableNonCoherent:icl +* This is not the same workaround as in early Gen9 platforms, where +* lacking this could cause system hangs, but coherency performance +* overhead is high and only a few compute workloads really need it +* (the register is whitelisted in hardware now, so UMDs can opt in +* for coherency if they have a good reason). +*/ +WA_SET_BIT_MASKED (ICL_HDC_CHICKEN0 ,HDC_FORCE_NON_COHERENT );Right, but the register name should be
ICL_HDC_MODE.
ACK
Quoted 36 lines
Quoted 23 lines
drivers/gpu/drm/i915/intel_workarounds.c +9
-0+ +return 0 ; +} +int intel_ctx_workarounds_init (struct drm_i915_private *dev_priv) {int err =0 ;@@ -465,6 +486,8 @@ int intel_ctx_workarounds_init(struct drm_i915_private *dev_priv) err =cfl_ctx_workarounds_init (dev_priv); else if (IS_CANNONLAKE(dev_priv)) err =cnl_ctx_workarounds_init (dev_priv); + else if (IS_ICELAKE(dev_priv)) + err =icl_ctx_workarounds_init (dev_priv); elseMISSING_CASE (INTEL_GEN (dev_priv ));if (err)@@ -663,6 +686,21 @@ static void cnl_gt_workarounds_apply(struct drm_i915_private *dev_priv) _MASKED_BIT_ENABLE(GEN9_FFSC_PERCTX_PREEMPT_CTRL)); } +static void icl_gt_workarounds_apply (struct drm_i915_private *dev_priv) +{ +/* This is not an Wa. Enable for better image quality */ drivers/gpu/drm/i915/intel_workarounds.c static void cnl_gt_workarounds_apply(struct drm_i915_private *dev_priv) · line 701
Quoted 10 lines
drivers/gpu/drm/i915/intel_workarounds.c +10
-0+I915_WRITE (_3D_CHICKEN3, +_MASKED_BIT_ENABLE (_3D_CHICKEN3_AA_LINE_QUALITY_FIX_ENABLE)); + +/* WaInPlaceDecompressionHang:icl */ +I915_WRITE (GEN9_GAMT_ECO_REG_RW_IA , (I915_READ (GEN9_GAMT_ECO_REG_RW_IA ) | +GAMT_ECO_ENABLE_IN_PLACE_DECOMPRESS )); + +/* WaPipelineFlushCoherentLines:icl */ +I915_WRITE (GEN8_L3SQCREG4 , (I915_READ (GEN8_L3SQCREG4 ) | +GEN8_LQSC_FLUSH_COHERENT_LINES ));Didn't find a HSDES entry for this. The workaround name and the reg/bit matches tho.
Quoted 8 lines
But the real question in here is that do we need to set this through indirect bb like we do with gen[8,9].
And just to note that cnl is missing this too. But that can be done as a followup when we first figure out that should we use the indirect bb for all >= gen8.
-Mika
Hmmmm... looking at it more carefully: isn't that a slightly different WA: WaPipelineFlushCoherentLines (not needed starting CNL)? This is equivalent to a different WA: WaOCLCoherentLineFlush.
But indeed, it seems to be required for CNL as well. In fact, I'm pretty sure I added it to this patch because I saw it in CNL. Did it get lost?
Quoted 34 lines
Quoted 34 lines
drivers/gpu/drm/i915/intel_workarounds.c +10
-0+} +void intel_gt_workarounds_apply (struct drm_i915_private *dev_priv) {if (INTEL_GEN (dev_priv) <8 )@@ -683,6 +721,8 @@ void intel_gt_workarounds_apply(struct drm_i915_private *dev_priv)cfl_gt_workarounds_apply (dev_priv); else if (IS_CANNONLAKE(dev_priv)) cnl_gt_workarounds_apply(dev_priv); + else if (IS_ICELAKE(dev_priv)) + icl_gt_workarounds_apply(dev_priv); elseMISSING_CASE (INTEL_GEN (dev_priv )); }@@ -761,6 +801,10 @@ static void cnl_whitelist_build(struct whitelist *w)whitelist_reg (w,GEN8_CS_CHICKEN1 ); } +static void icl_whitelist_build (struct whitelist *w) +{ +} +static struct whitelist *whitelist_build (struct intel_engine_cs *engine,struct whitelist *w) {@@ -789,6 +833,8 @@ static struct whitelist *whitelist_build(struct intel_engine_cs *engine,cfl_whitelist_build (w); else if (IS_CANNONLAKE(i915)) cnl_whitelist_build(w); + else if (IS_ICELAKE(i915)) + icl_whitelist_build(w); elseMISSING_CASE (INTEL_GEN (i915 ));
Quoted 2 lines
Quoted 2 lines
Signature · 1 line
1.9.1
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
PATCH v4 2/22 drm/i915/icl: Enable Sampler DFR
In reply to Oscar MateoSampler Dynamic Frequency Rebalancing (DFR) aims to reduce Sampler power by dynamically changing its clock frequency in low-throughput conditions. This patches enables it by default on Gen11.
v2: Wrong operation to clear the bit (Praveen)
v3: Rebased on top of the WA refactoring
--- drivers/gpu/drm/i915/i915_reg.h | 3 +++ drivers/gpu/drm/i915/intel_workarounds.c | 4 ++++ 2 files changed, 7 insertions(+)
drivers/gpu/drm/i915/i915_reg.h +3 -0
@@ -8218,6 +8218,9 @@ enum {#define GEN8_GARBCNTL _MMIO(0xB004)#define GEN9_GAPS_TSV_CREDIT_DISABLE (1<<7) +#define GEN10_DFR_RATIO_EN_AND_CHICKEN _MMIO(0x9550) +#define DFR_DISABLE (1 << 9) +/* IVYBRIDGE DPF */ #define GEN7_L3CDERRST1 (slice) _MMIO(0xB008 + (slice) * 0x200)/* L3CD Error Status 1 */ #define GEN7_L3CDERRST1_ROW_MASK (0x7ff<<14)
drivers/gpu/drm/i915/intel_workarounds.c +4 -0
@@ -692,6 +692,10 @@ static void icl_gt_workarounds_apply(struct drm_i915_private *dev_priv)I915_WRITE (_3D_CHICKEN3,_MASKED_BIT_ENABLE (_3D_CHICKEN3_AA_LINE_QUALITY_FIX_ENABLE)); +/* This is not an Wa. Enable to reduce Sampler power */ +I915_WRITE (GEN10_DFR_RATIO_EN_AND_CHICKEN , + (I915_READ (GEN10_DFR_RATIO_EN_AND_CHICKEN ) & ~DFR_DISABLE )); +/* WaInPlaceDecompressionHang:icl */ I915_WRITE (GEN9_GAMT_ECO_REG_RW_IA , (I915_READ (GEN9_GAMT_ECO_REG_RW_IA ) |GAMT_ECO_ENABLE_IN_PLACE_DECOMPRESS ));
Signature · 6 lines
1.9.1 _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Rodrigo Vivi
maintainerRe: [PATCH 02/22] drm/i915/icl: Enable Sampler DFR
Quoted 35 lines · Oscar Mateo
Sampler Dynamic Frequency Rebalancing (DFR) aims to reduce Sampler power by dynamically changing its clock frequency in low-throughput conditions. This patches enables it by default on Gen11.
v2: Wrong operation to clear the bit (Praveen)
v3: Rebased on top of the WA refactoringCc Praveen Paneri <[email protected]>Cc Mika Kuoppala <[email protected]>Signed-off-by Oscar Mateo <[email protected]>Reviewed-by Sagar Arun Kamble <[email protected]>--- drivers/gpu/drm/i915/i915_reg.h | 3 +++ drivers/gpu/drm/i915/intel_workarounds.c | 4 ++++ 2 files changed, 7 insertions(+)drivers/gpu/drm/i915/i915_reg.h +3
-0@@ -8218,6 +8218,9 @@ enum {#define GEN8_GARBCNTL _MMIO(0xB004)#define GEN9_GAPS_TSV_CREDIT_DISABLE (1<<7) +#define GEN10_DFR_RATIO_EN_AND_CHICKEN _MMIO(0x9550) +#define DFR_DISABLE (1 << 9) +/* IVYBRIDGE DPF */ #define GEN7_L3CDERRST1 (slice) _MMIO(0xB008 + (slice) * 0x200)/* L3CD Error Status 1 */ #define GEN7_L3CDERRST1_ROW_MASK (0x7ff<<14)drivers/gpu/drm/i915/intel_workarounds.c +0
-0
drivers/gpu/drm/i915/intel_workarounds.c static void icl_gt_workarounds_apply(struct drm_i915_private *dev_priv) · line 695
Quoted 4 lines
drivers/gpu/drm/i915/intel_workarounds.c +1
-0@@ -692,6 +692,10 @@ static void icl_gt_workarounds_apply(struct drm_i915_private *dev_priv)I915_WRITE (_3D_CHICKEN3,_MASKED_BIT_ENABLE (_3D_CHICKEN3_AA_LINE_QUALITY_FIX_ENABLE)); +/* This is not an Wa. Enable to reduce Sampler power */
First of all it is strange that a feature is under a chicken bit,
but if it is not an wa, but a PM feature, shouldn't we add it to some sort of init clock gating function, or a new specific function called when it makes sense?
Quoted 13 lines
drivers/gpu/drm/i915/intel_workarounds.c +3
-0+I915_WRITE (GEN10_DFR_RATIO_EN_AND_CHICKEN , + (I915_READ (GEN10_DFR_RATIO_EN_AND_CHICKEN ) & ~DFR_DISABLE )); +/* WaInPlaceDecompressionHang:icl */ I915_WRITE (GEN9_GAMT_ECO_REG_RW_IA , (I915_READ (GEN9_GAMT_ECO_REG_RW_IA ) |GAMT_ECO_ENABLE_IN_PLACE_DECOMPRESS ));Signature · 6 lines
1.9.1 _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Re: [PATCH 02/22] drm/i915/icl: Enable Sampler DFR
Quoted 39 lines · Rodrigo Vivi
Quoted 38 lines · Oscar Mateo
Sampler Dynamic Frequency Rebalancing (DFR) aims to reduce Sampler power by dynamically changing its clock frequency in low-throughput conditions. This patches enables it by default on Gen11.
v2: Wrong operation to clear the bit (Praveen)
v3: Rebased on top of the WA refactoringCc Praveen Paneri <[email protected]>Cc Mika Kuoppala <[email protected]>Signed-off-by Oscar Mateo <[email protected]>Reviewed-by Sagar Arun Kamble <[email protected]>--- drivers/gpu/drm/i915/i915_reg.h | 3 +++ drivers/gpu/drm/i915/intel_workarounds.c | 4 ++++ 2 files changed, 7 insertions(+)drivers/gpu/drm/i915/i915_reg.h +3
-0@@ -8218,6 +8218,9 @@ enum {#define GEN8_GARBCNTL _MMIO(0xB004)#define GEN9_GAPS_TSV_CREDIT_DISABLE (1<<7) +#define GEN10_DFR_RATIO_EN_AND_CHICKEN _MMIO(0x9550) +#define DFR_DISABLE (1 << 9) +/* IVYBRIDGE DPF */ #define GEN7_L3CDERRST1 (slice) _MMIO(0xB008 + (slice) * 0x200)/* L3CD Error Status 1 */ #define GEN7_L3CDERRST1_ROW_MASK (0x7ff<<14)drivers/gpu/drm/i915/intel_workarounds.c +0
-0@@ -692,6 +692,10 @@ static void icl_gt_workarounds_apply(struct drm_i915_private *dev_priv)I915_WRITE (_3D_CHICKEN3,_MASKED_BIT_ENABLE (_3D_CHICKEN3_AA_LINE_QUALITY_FIX_ENABLE));
Quoted 2 lines
drivers/gpu/drm/i915/intel_workarounds.c static void icl_gt_workarounds_apply(struct drm_i915_private *dev_priv) · line 695
Quoted 1 line
drivers/gpu/drm/i915/intel_workarounds.c +1
-0+/* This is not an Wa. Enable to reduce Sampler power */ First of all it is strange that a feature is under a chicken bit,
There is a precedent: _3D_CHICKEN3_AA_LINE_QUALITY_FIX_ENABLE
Quoted 2 lines
but if it is not an wa, but a PM feature, shouldn't we add it to some sort of init clock gating function, or a new specific function called when it makes sense?
I'm open to suggestions
Quoted 9 lines
Quoted 9 lines
drivers/gpu/drm/i915/intel_workarounds.c +3
-0+I915_WRITE (GEN10_DFR_RATIO_EN_AND_CHICKEN , + (I915_READ (GEN10_DFR_RATIO_EN_AND_CHICKEN ) & ~DFR_DISABLE )); +/* WaInPlaceDecompressionHang:icl */ I915_WRITE (GEN9_GAMT_ECO_REG_RW_IA , (I915_READ (GEN9_GAMT_ECO_REG_RW_IA ) |GAMT_ECO_ENABLE_IN_PLACE_DECOMPRESS ));Signature · 1 line
1.9.1
Quoted 4 lines
Quoted 4 lines
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Rodrigo Vivi
maintainerRe: [PATCH 02/22] drm/i915/icl: Enable Sampler DFR
Quoted 43 lines · Oscar Mateo
Quoted 31 lines · Rodrigo Vivi
Quoted 30 lines · Oscar Mateo
Sampler Dynamic Frequency Rebalancing (DFR) aims to reduce Sampler power by dynamically changing its clock frequency in low-throughput conditions. This patches enables it by default on Gen11.
v2: Wrong operation to clear the bit (Praveen)
v3: Rebased on top of the WA refactoringCc Praveen Paneri <[email protected]>Cc Mika Kuoppala <[email protected]>Signed-off-by Oscar Mateo <[email protected]>Reviewed-by Sagar Arun Kamble <[email protected]>--- drivers/gpu/drm/i915/i915_reg.h | 3 +++ drivers/gpu/drm/i915/intel_workarounds.c | 4 ++++ 2 files changed, 7 insertions(+)drivers/gpu/drm/i915/i915_reg.h +3
-0@@ -8218,6 +8218,9 @@ enum {#define GEN8_GARBCNTL _MMIO(0xB004)#define GEN9_GAPS_TSV_CREDIT_DISABLE (1<<7) +#define GEN10_DFR_RATIO_EN_AND_CHICKEN _MMIO(0x9550) +#define DFR_DISABLE (1 << 9) +/* IVYBRIDGE DPF */ #define GEN7_L3CDERRST1 (slice) _MMIO(0xB008 + (slice) * 0x200)/* L3CD Error Status 1 */ #define GEN7_L3CDERRST1_ROW_MASK (0x7ff<<14)drivers/gpu/drm/i915/intel_workarounds.c +0
-0Quoted 8 lines
drivers/gpu/drm/i915/intel_workarounds.c static void icl_gt_workarounds_apply(struct drm_i915_private *dev_priv) · line 694
Quoted 7 lines
index 3f00623..60a5b1d 100644
drivers/gpu/drm/i915/intel_workarounds.c +1
-0@@ -692,6 +692,10 @@ static void icl_gt_workarounds_apply(struct drm_i915_private *dev_priv)I915_WRITE (_3D_CHICKEN3,_MASKED_BIT_ENABLE (_3D_CHICKEN3_AA_LINE_QUALITY_FIX_ENABLE)); +/* This is not an Wa. Enable to reduce Sampler power */ First of all it is strange that a feature is under a chicken bit,
Quoted 1 line
There is a precedent:
_3D_CHICKEN3_AA_LINE_QUALITY_FIX_ENABLE
Oh, true!
Actually that never got same question from me because that function is called init clock gating, although this function nowadays is basically workarounds related to clock gatings :/)
Quoted 5 lines
Quoted 3 lines
but if it is not an wa, but a PM feature, shouldn't we add it to some sort of init clock gating function, or a new specific function called when it makes sense?
I'm open to suggestions
moving to clock gating is an option?
or needs to be called from context?
Quoted 10 lines
Quoted 10 lines
Quoted 10 lines
drivers/gpu/drm/i915/intel_workarounds.c +3
-0+I915_WRITE (GEN10_DFR_RATIO_EN_AND_CHICKEN , + (I915_READ (GEN10_DFR_RATIO_EN_AND_CHICKEN ) & ~DFR_DISABLE )); +/* WaInPlaceDecompressionHang:icl */ I915_WRITE (GEN9_GAMT_ECO_REG_RW_IA , (I915_READ (GEN9_GAMT_ECO_REG_RW_IA ) |GAMT_ECO_ENABLE_IN_PLACE_DECOMPRESS ));Signature · 1 line
1.9.1
Quoted 9 lines
Quoted 4 lines
Quoted 4 lines
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Re: [PATCH 02/22] drm/i915/icl: Enable Sampler DFR
Quoted 49 lines · Rodrigo Vivi
Quoted 41 lines · Oscar Mateo
Quoted 31 lines · Rodrigo Vivi
Quoted 30 lines · Oscar Mateo
Sampler Dynamic Frequency Rebalancing (DFR) aims to reduce Sampler power by dynamically changing its clock frequency in low-throughput conditions. This patches enables it by default on Gen11.
v2: Wrong operation to clear the bit (Praveen)
v3: Rebased on top of the WA refactoringCc Praveen Paneri <[email protected]>Cc Mika Kuoppala <[email protected]>Signed-off-by Oscar Mateo <[email protected]>Reviewed-by Sagar Arun Kamble <[email protected]>--- drivers/gpu/drm/i915/i915_reg.h | 3 +++ drivers/gpu/drm/i915/intel_workarounds.c | 4 ++++ 2 files changed, 7 insertions(+)drivers/gpu/drm/i915/i915_reg.h +3
-0@@ -8218,6 +8218,9 @@ enum {#define GEN8_GARBCNTL _MMIO(0xB004)#define GEN9_GAPS_TSV_CREDIT_DISABLE (1<<7) +#define GEN10_DFR_RATIO_EN_AND_CHICKEN _MMIO(0x9550) +#define DFR_DISABLE (1 << 9) +/* IVYBRIDGE DPF */ #define GEN7_L3CDERRST1 (slice) _MMIO(0xB008 + (slice) * 0x200)/* L3CD Error Status 1 */ #define GEN7_L3CDERRST1_ROW_MASK (0x7ff<<14)drivers/gpu/drm/i915/intel_workarounds.c +0
-0Quoted 8 lines
drivers/gpu/drm/i915/intel_workarounds.c static void icl_gt_workarounds_apply(struct drm_i915_private *dev_priv) · line 694
Quoted 7 lines
index 3f00623..60a5b1d 100644
drivers/gpu/drm/i915/intel_workarounds.c +1
-0@@ -692,6 +692,10 @@ static void icl_gt_workarounds_apply(struct drm_i915_private *dev_priv)I915_WRITE (_3D_CHICKEN3,_MASKED_BIT_ENABLE (_3D_CHICKEN3_AA_LINE_QUALITY_FIX_ENABLE)); +/* This is not an Wa. Enable to reduce Sampler power */ First of all it is strange that a feature is under a chicken bit,
Quoted 1 line
There is a precedent:
_3D_CHICKEN3_AA_LINE_QUALITY_FIX_ENABLEOh, true!
Actually that never got same question from me because that function is called init clock gating, although this function nowadays is basically workarounds related to clock gatings :/)
Quoted 5 lines
Quoted 3 lines
Quoted 2 lines
but if it is not an wa, but a PM feature, shouldn't we add it to some sort of init clock gating function, or a new specific function called when it makes sense?
I'm open to suggestions
moving to clock gating is an option?
or needs to be called from context?
No, moving it to init_clock_gating is definitely an option. It is not truly related to clock gating, but at least it is related to Power Management... (the _3D_CHICKEN3_AA_LINE_QUALITY_FIX_ENABLE thing is more difficult to swallow...)
Quoted 9 lines
Quoted 9 lines
Quoted 9 lines
Quoted 9 lines
drivers/gpu/drm/i915/intel_workarounds.c +3
-0+I915_WRITE (GEN10_DFR_RATIO_EN_AND_CHICKEN , + (I915_READ (GEN10_DFR_RATIO_EN_AND_CHICKEN ) & ~DFR_DISABLE )); +/* WaInPlaceDecompressionHang:icl */ I915_WRITE (GEN9_GAMT_ECO_REG_RW_IA , (I915_READ (GEN9_GAMT_ECO_REG_RW_IA ) |GAMT_ECO_ENABLE_IN_PLACE_DECOMPRESS ));Signature · 1 line
1.9.1
Quoted 8 lines
Quoted 8 lines
Quoted 4 lines
Quoted 4 lines
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Rodrigo Vivi
maintainerRe: [PATCH 02/22] drm/i915/icl: Enable Sampler DFR
Quoted 54 lines · Oscar Mateo
Quoted 51 lines · Rodrigo Vivi
Quoted 41 lines · Oscar Mateo
Quoted 31 lines · Rodrigo Vivi
Quoted 30 lines · Oscar Mateo
Sampler Dynamic Frequency Rebalancing (DFR) aims to reduce Sampler power by dynamically changing its clock frequency in low-throughput conditions. This patches enables it by default on Gen11.
v2: Wrong operation to clear the bit (Praveen)
v3: Rebased on top of the WA refactoringCc Praveen Paneri <[email protected]>Cc Mika Kuoppala <[email protected]>Signed-off-by Oscar Mateo <[email protected]>Reviewed-by Sagar Arun Kamble <[email protected]>--- drivers/gpu/drm/i915/i915_reg.h | 3 +++ drivers/gpu/drm/i915/intel_workarounds.c | 4 ++++ 2 files changed, 7 insertions(+)drivers/gpu/drm/i915/i915_reg.h +3
-0@@ -8218,6 +8218,9 @@ enum {#define GEN8_GARBCNTL _MMIO(0xB004)#define GEN9_GAPS_TSV_CREDIT_DISABLE (1<<7) +#define GEN10_DFR_RATIO_EN_AND_CHICKEN _MMIO(0x9550) +#define DFR_DISABLE (1 << 9) +/* IVYBRIDGE DPF */ #define GEN7_L3CDERRST1 (slice) _MMIO(0xB008 + (slice) * 0x200)/* L3CD Error Status 1 */ #define GEN7_L3CDERRST1_ROW_MASK (0x7ff<<14)drivers/gpu/drm/i915/intel_workarounds.c +0
-0Quoted 8 lines
drivers/gpu/drm/i915/intel_workarounds.c static void icl_gt_workarounds_apply(struct drm_i915_private *dev_priv) · line 694
Quoted 7 lines
index 3f00623..60a5b1d 100644
drivers/gpu/drm/i915/intel_workarounds.c +1
-0@@ -692,6 +692,10 @@ static void icl_gt_workarounds_apply(struct drm_i915_private *dev_priv)I915_WRITE (_3D_CHICKEN3,_MASKED_BIT_ENABLE (_3D_CHICKEN3_AA_LINE_QUALITY_FIX_ENABLE)); +/* This is not an Wa. Enable to reduce Sampler power */ First of all it is strange that a feature is under a chicken bit,
Quoted 1 line
There is a precedent:
_3D_CHICKEN3_AA_LINE_QUALITY_FIX_ENABLEOh, true!
Actually that never got same question from me because that function is called init clock gating, although this function nowadays is basically workarounds related to clock gatings :/)
Quoted 2 lines
Quoted 2 lines
but if it is not an wa, but a PM feature, shouldn't we add it to some sort of init clock gating function, or a new specific function called when it makes sense?
Quoted 8 lines
Quoted 3 lines
Quoted 1 line
I'm open to suggestions
moving to clock gating is an option?
or needs to be called from context?No, moving it to
init_clock_gatingis definitely an option. It is not truly related to clock gating, but at least it is related to Power Management... (the_3D_CHICKEN3_AA_LINE_QUALITY_FIX_ENABLEthing is more difficult to swallow...)
I agree hehe... my bad... :$
Quoted 19 lines
Quoted 10 lines
Quoted 10 lines
Quoted 10 lines
Quoted 10 lines
drivers/gpu/drm/i915/intel_workarounds.c +3
-0+I915_WRITE (GEN10_DFR_RATIO_EN_AND_CHICKEN , + (I915_READ (GEN10_DFR_RATIO_EN_AND_CHICKEN ) & ~DFR_DISABLE )); +/* WaInPlaceDecompressionHang:icl */ I915_WRITE (GEN9_GAMT_ECO_REG_RW_IA , (I915_READ (GEN9_GAMT_ECO_REG_RW_IA ) |GAMT_ECO_ENABLE_IN_PLACE_DECOMPRESS ));Signature · 1 line
1.9.1Quoted 8 lines
Quoted 8 lines
Quoted 4 lines
Quoted 4 lines
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Quoted 4 lines
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
PATCH v4 3/22 drm/i915/icl: WaGAPZPriorityScheme
In reply to Oscar MateoThe default GAPZ arbitrer priority value at power-on has been found to be incorrect.
v2: Now renamed to Wa_1405543622
v3: Rebased on top of the WA refactoring
--- drivers/gpu/drm/i915/i915_reg.h | 5 +++-- drivers/gpu/drm/i915/intel_workarounds.c | 6 ++++++ 2 files changed, 9 insertions(+), 2 deletions(-)
drivers/gpu/drm/i915/i915_reg.h +3 -2
@@ -8215,8 +8215,9 @@ enum {#define GEN8_DOP_CLOCK_GATE_GUC_ENABLE (1<<4)#define GEN8_DOP_CLOCK_GATE_MEDIA_ENABLE (1<<6) -#define GEN8_GARBCNTL _MMIO(0xB004) -#define GEN9_GAPS_TSV_CREDIT_DISABLE (1<<7) +#define GEN8_GARBCNTL _MMIO(0xB004) +#define GEN9_GAPS_TSV_CREDIT_DISABLE (1 << 7) +#define GEN11_ARBITRATION_PRIO_ORDER_MASK (0x3f << 22)#define GEN10_DFR_RATIO_EN_AND_CHICKEN _MMIO(0x9550)#define DFR_DISABLE (1 << 9)
drivers/gpu/drm/i915/intel_workarounds.c +6 -0
@@ -703,6 +703,12 @@ static void icl_gt_workarounds_apply(struct drm_i915_private *dev_priv)/* WaPipelineFlushCoherentLines:icl */ I915_WRITE (GEN8_L3SQCREG4 , (I915_READ (GEN8_L3SQCREG4 ) |GEN8_LQSC_FLUSH_COHERENT_LINES )); + +/* Wa_1405543622:icl +* Formerly known as WaGAPZPriorityScheme +*/ +I915_WRITE (GEN8_GARBCNTL , (I915_READ (GEN8_GARBCNTL ) | +GEN11_ARBITRATION_PRIO_ORDER_MASK )); }void intel_gt_workarounds_apply (struct drm_i915_private *dev_priv)
Signature · 6 lines
1.9.1 _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Re: [PATCH 03/22] drm/i915/icl: WaGAPZPriorityScheme
Quoted 7 lines · Oscar Mateo
The default GAPZ arbitrer priority value at power-on has been found to be incorrect.
v2: Now renamed to
Wa_1405543622v3: Rebased on top of the WA refactoring
I have suggested that when implementing workarounds, authors, for example, add
References: HSDES#1405542622, HSDES#1207131207
as a first tagline to ease up the review.
I do that again :) No need to resend,
Quoted 42 lines
Cc Mika Kuoppala <[email protected]>Signed-off-by Oscar Mateo <[email protected]>--- drivers/gpu/drm/i915/i915_reg.h | 5 +++-- drivers/gpu/drm/i915/intel_workarounds.c | 6 ++++++ 2 files changed, 9 insertions(+), 2 deletions(-)drivers/gpu/drm/i915/i915_reg.h +3
-2@@ -8215,8 +8215,9 @@ enum {#define GEN8_DOP_CLOCK_GATE_GUC_ENABLE (1<<4)#define GEN8_DOP_CLOCK_GATE_MEDIA_ENABLE (1<<6) -#define GEN8_GARBCNTL _MMIO(0xB004) -#define GEN9_GAPS_TSV_CREDIT_DISABLE (1<<7) +#define GEN8_GARBCNTL _MMIO(0xB004) +#define GEN9_GAPS_TSV_CREDIT_DISABLE (1 << 7) +#define GEN11_ARBITRATION_PRIO_ORDER_MASK (0x3f << 22)#define GEN10_DFR_RATIO_EN_AND_CHICKEN _MMIO(0x9550)#define DFR_DISABLE (1 << 9)drivers/gpu/drm/i915/intel_workarounds.c +6
-0@@ -703,6 +703,12 @@ static void icl_gt_workarounds_apply(struct drm_i915_private *dev_priv)/* WaPipelineFlushCoherentLines:icl */ I915_WRITE (GEN8_L3SQCREG4 , (I915_READ (GEN8_L3SQCREG4 ) |GEN8_LQSC_FLUSH_COHERENT_LINES )); + +/* Wa_1405543622:icl +* Formerly known as WaGAPZPriorityScheme +*/ +I915_WRITE (GEN8_GARBCNTL , (I915_READ (GEN8_GARBCNTL ) | +GEN11_ARBITRATION_PRIO_ORDER_MASK )); }void intel_gt_workarounds_apply (struct drm_i915_private *dev_priv)Signature · 1 line
1.9.1
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Re: [PATCH 03/22] drm/i915/icl: WaGAPZPriorityScheme
Quoted 5 lines · Mika Kuoppala
Quoted 3 lines · Oscar Mateo
The default GAPZ arbitrer priority value at power-on has been found to be incorrect.
Quoted 8 lines
Quoted 4 lines
v2: Now renamed to
Wa_1405543622v3: Rebased on top of the WA refactoring
I have suggested that when implementing workarounds, authors, for example, add
References: HSDES#1405542622, HSDES#1207131207
The thing is, nowadays, the name of the WA itself *is* the HSDES number (e.g. this is Wa_1405543622 because it is documented in HSDES#1405543622). At least for hardware workarounds. But I have no problem adding the "References" tag as well, as long as we have consensus.
Quoted 4 lines
as a first tagline to ease up the review.
I do that again :) No need to resend,Reviewed-by Mika Kuoppala <[email protected]>
Thanks!
Quoted 39 lines
Quoted 39 lines
Cc Mika Kuoppala <[email protected]>Signed-off-by Oscar Mateo <[email protected]>--- drivers/gpu/drm/i915/i915_reg.h | 5 +++-- drivers/gpu/drm/i915/intel_workarounds.c | 6 ++++++ 2 files changed, 9 insertions(+), 2 deletions(-)drivers/gpu/drm/i915/i915_reg.h +3
-2@@ -8215,8 +8215,9 @@ enum {#define GEN8_DOP_CLOCK_GATE_GUC_ENABLE (1<<4)#define GEN8_DOP_CLOCK_GATE_MEDIA_ENABLE (1<<6) -#define GEN8_GARBCNTL _MMIO(0xB004) -#define GEN9_GAPS_TSV_CREDIT_DISABLE (1<<7) +#define GEN8_GARBCNTL _MMIO(0xB004) +#define GEN9_GAPS_TSV_CREDIT_DISABLE (1 << 7) +#define GEN11_ARBITRATION_PRIO_ORDER_MASK (0x3f << 22)#define GEN10_DFR_RATIO_EN_AND_CHICKEN _MMIO(0x9550)#define DFR_DISABLE (1 << 9)drivers/gpu/drm/i915/intel_workarounds.c +6
-0@@ -703,6 +703,12 @@ static void icl_gt_workarounds_apply(struct drm_i915_private *dev_priv)/* WaPipelineFlushCoherentLines:icl */ I915_WRITE (GEN8_L3SQCREG4 , (I915_READ (GEN8_L3SQCREG4 ) |GEN8_LQSC_FLUSH_COHERENT_LINES )); + +/* Wa_1405543622:icl +* Formerly known as WaGAPZPriorityScheme +*/ +I915_WRITE (GEN8_GARBCNTL , (I915_READ (GEN8_GARBCNTL ) | +GEN11_ARBITRATION_PRIO_ORDER_MASK )); }
Quoted 3 lines
Quoted 3 lines
drivers/gpu/drm/i915/intel_workarounds.c +0
-0void intel_gt_workarounds_apply (struct drm_i915_private *dev_priv)Signature · 1 line
1.9.1
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
PATCH v4 4/22 drm/i915/icl: WaL3BankAddressHashing
In reply to Oscar MateoRevert to an L3 non-hash model, for performance reasons.
v2:
- Place the WA name above the actual change
- Improve the register naming
v3:
- Rebased
- Renamed to Wa_1604223664
v4: Rebased on top of the WA refactoring
--- drivers/gpu/drm/i915/i915_reg.h | 6 ++++++ drivers/gpu/drm/i915/intel_workarounds.c | 20 ++++++++++++++++---- 2 files changed, 22 insertions(+), 4 deletions(-)
drivers/gpu/drm/i915/i915_reg.h +6 -0
@@ -8218,6 +8218,12 @@ enum {#define GEN8_GARBCNTL _MMIO(0xB004)#define GEN9_GAPS_TSV_CREDIT_DISABLE (1 << 7)#define GEN11_ARBITRATION_PRIO_ORDER_MASK (0x3f << 22) +#define GEN11_HASH_CTRL_EXCL_MASK (0x7f << 0) +#define GEN11_HASH_CTRL_EXCL_BIT0 (1 << 0) + +#define GEN11_GLBLINVL _MMIO(0xB404) +#define GEN11_BANK_HASH_ADDR_EXCL_BIT0 (1 << 0) +#define GEN11_BANK_HASH_ADDR_EXCL_MASK (0x3f << 5)#define GEN10_DFR_RATIO_EN_AND_CHICKEN _MMIO(0x9550)#define DFR_DISABLE (1 << 9)
drivers/gpu/drm/i915/intel_workarounds.c +16 -4
@@ -704,11 +704,23 @@ static void icl_gt_workarounds_apply(struct drm_i915_private *dev_priv)I915_WRITE (GEN8_L3SQCREG4 , (I915_READ (GEN8_L3SQCREG4 ) |GEN8_LQSC_FLUSH_COHERENT_LINES )); -/* Wa_1405543622:icl -* Formerly known as WaGAPZPriorityScheme +I915_WRITE (GEN8_GARBCNTL , +/* Wa_1604223664:icl +* Formerly known as WaL3BankAddressHashing +*/ + ((I915_READ (GEN8_GARBCNTL ) & ~GEN11_HASH_CTRL_EXCL_MASK ) | +GEN11_HASH_CTRL_EXCL_BIT0 | +/* Wa_1405543622:icl +* Formerly known as WaGAPZPriorityScheme +*/ +GEN11_ARBITRATION_PRIO_ORDER_MASK )); + +/* Wa_1604223664:icl +* Formerly known as WaL3BankAddressHashing */ -I915_WRITE (GEN8_GARBCNTL , (I915_READ (GEN8_GARBCNTL ) | -GEN11_ARBITRATION_PRIO_ORDER_MASK )); +I915_WRITE (GEN11_GLBLINVL , + ((I915_READ (GEN11_GLBLINVL ) & ~GEN11_BANK_HASH_ADDR_EXCL_MASK ) | +GEN11_BANK_HASH_ADDR_EXCL_BIT0 )); }void intel_gt_workarounds_apply (struct drm_i915_private *dev_priv)
Signature · 6 lines
1.9.1 _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Re: [PATCH 04/22] drm/i915/icl: WaL3BankAddressHashing
Quoted 22 lines · Oscar Mateo
Revert to an L3 non-hash model, for performance reasons.
v2:
- Place the WA name above the actual change
- Improve the register naming
v3:
- Rebased
- Renamed toWa_1604223664
v4: Rebased on top of the WA refactoringCc Mika Kuoppala <[email protected]>Signed-off-by Oscar Mateo <[email protected]>--- drivers/gpu/drm/i915/i915_reg.h | 6 ++++++ drivers/gpu/drm/i915/intel_workarounds.c | 20 ++++++++++++++++---- 2 files changed, 22 insertions(+), 4 deletions(-)drivers/gpu/drm/i915/i915_reg.h +0
-0
drivers/gpu/drm/i915/i915_reg.h enum { · line 8225
Quoted 8 lines
drivers/gpu/drm/i915/i915_reg.h +5
-0@@ -8218,6 +8218,12 @@ enum {#define GEN8_GARBCNTL _MMIO(0xB004)#define GEN9_GAPS_TSV_CREDIT_DISABLE (1 << 7)#define GEN11_ARBITRATION_PRIO_ORDER_MASK (0x3f << 22) +#define GEN11_HASH_CTRL_EXCL_MASK (0x7f << 0) +#define GEN11_HASH_CTRL_EXCL_BIT0 (1 << 0) + +#define GEN11_GLBLINVL _MMIO(0xB404) +#define GEN11_BANK_HASH_ADDR_EXCL_BIT0 (1 << 0)
(1 << 5)
drivers/gpu/drm/i915/i915_reg.h enum { · line 8226
Quoted 1 line
drivers/gpu/drm/i915/i915_reg.h +1
-0+#define GEN11_BANK_HASH_ADDR_EXCL_MASK (0x3f << 5)
0x7f
-Mika
Quoted 37 lines
drivers/gpu/drm/i915/i915_reg.h +0
-0#define GEN10_DFR_RATIO_EN_AND_CHICKEN _MMIO(0x9550)#define DFR_DISABLE (1 << 9)drivers/gpu/drm/i915/intel_workarounds.c +16
-4@@ -704,11 +704,23 @@ static void icl_gt_workarounds_apply(struct drm_i915_private *dev_priv)I915_WRITE (GEN8_L3SQCREG4 , (I915_READ (GEN8_L3SQCREG4 ) |GEN8_LQSC_FLUSH_COHERENT_LINES )); -/* Wa_1405543622:icl -* Formerly known as WaGAPZPriorityScheme +I915_WRITE (GEN8_GARBCNTL , +/* Wa_1604223664:icl +* Formerly known as WaL3BankAddressHashing +*/ + ((I915_READ (GEN8_GARBCNTL ) & ~GEN11_HASH_CTRL_EXCL_MASK ) | +GEN11_HASH_CTRL_EXCL_BIT0 | +/* Wa_1405543622:icl +* Formerly known as WaGAPZPriorityScheme +*/ +GEN11_ARBITRATION_PRIO_ORDER_MASK )); + +/* Wa_1604223664:icl +* Formerly known as WaL3BankAddressHashing */ -I915_WRITE (GEN8_GARBCNTL , (I915_READ (GEN8_GARBCNTL ) | -GEN11_ARBITRATION_PRIO_ORDER_MASK )); +I915_WRITE (GEN11_GLBLINVL , + ((I915_READ (GEN11_GLBLINVL ) & ~GEN11_BANK_HASH_ADDR_EXCL_MASK ) | +GEN11_BANK_HASH_ADDR_EXCL_BIT0 )); }void intel_gt_workarounds_apply (struct drm_i915_private *dev_priv)Signature · 1 line
1.9.1
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Re: [PATCH 04/22] drm/i915/icl: WaL3BankAddressHashing
Quoted 34 lines · Mika Kuoppala
Quoted 22 lines · Oscar Mateo
Revert to an L3 non-hash model, for performance reasons.
v2:
- Place the WA name above the actual change
- Improve the register naming
v3:
- Rebased
- Renamed toWa_1604223664
v4: Rebased on top of the WA refactoringCc Mika Kuoppala <[email protected]>Signed-off-by Oscar Mateo <[email protected]>--- drivers/gpu/drm/i915/i915_reg.h | 6 ++++++ drivers/gpu/drm/i915/intel_workarounds.c | 20 ++++++++++++++++---- 2 files changed, 22 insertions(+), 4 deletions(-)drivers/gpu/drm/i915/i915_reg.h +0
-0drivers/gpu/drm/i915/i915_reg.h enum { · line 8225
Quoted 8 lines
drivers/gpu/drm/i915/i915_reg.h +5
-0@@ -8218,6 +8218,12 @@ enum {#define GEN8_GARBCNTL _MMIO(0xB004)#define GEN9_GAPS_TSV_CREDIT_DISABLE (1 << 7)#define GEN11_ARBITRATION_PRIO_ORDER_MASK (0x3f << 22) +#define GEN11_HASH_CTRL_EXCL_MASK (0x7f << 0) +#define GEN11_HASH_CTRL_EXCL_BIT0 (1 << 0) + +#define GEN11_GLBLINVL _MMIO(0xB404) +#define GEN11_BANK_HASH_ADDR_EXCL_BIT0 (1 << 0)(1 << 5)
Quoted 3 lines
drivers/gpu/drm/i915/i915_reg.h enum { · line 8226
Quoted 1 line
drivers/gpu/drm/i915/i915_reg.h +1
-0+#define GEN11_BANK_HASH_ADDR_EXCL_MASK (0x3f << 5)
0x7f
-Mika
Lordy! Thanks for spotting this. Also, I went back to the HSD looking for the source of the mistake and the WA has been updated with more steps, so I'll send a new version.
-- Oscar
Quoted 35 lines
Quoted 35 lines
drivers/gpu/drm/i915/i915_reg.h +0
-0#define GEN10_DFR_RATIO_EN_AND_CHICKEN _MMIO(0x9550)#define DFR_DISABLE (1 << 9)drivers/gpu/drm/i915/intel_workarounds.c +16
-4@@ -704,11 +704,23 @@ static void icl_gt_workarounds_apply(struct drm_i915_private *dev_priv)I915_WRITE (GEN8_L3SQCREG4 , (I915_READ (GEN8_L3SQCREG4 ) |GEN8_LQSC_FLUSH_COHERENT_LINES )); -/* Wa_1405543622:icl -* Formerly known as WaGAPZPriorityScheme +I915_WRITE (GEN8_GARBCNTL , +/* Wa_1604223664:icl +* Formerly known as WaL3BankAddressHashing +*/ + ((I915_READ (GEN8_GARBCNTL ) & ~GEN11_HASH_CTRL_EXCL_MASK ) | +GEN11_HASH_CTRL_EXCL_BIT0 | +/* Wa_1405543622:icl +* Formerly known as WaGAPZPriorityScheme +*/ +GEN11_ARBITRATION_PRIO_ORDER_MASK )); + +/* Wa_1604223664:icl +* Formerly known as WaL3BankAddressHashing */ -I915_WRITE (GEN8_GARBCNTL , (I915_READ (GEN8_GARBCNTL ) | -GEN11_ARBITRATION_PRIO_ORDER_MASK )); +I915_WRITE (GEN11_GLBLINVL , + ((I915_READ (GEN11_GLBLINVL ) & ~GEN11_BANK_HASH_ADDR_EXCL_MASK ) | +GEN11_BANK_HASH_ADDR_EXCL_BIT0 )); }
Quoted 3 lines
Quoted 3 lines
drivers/gpu/drm/i915/intel_workarounds.c +0
-0void intel_gt_workarounds_apply (struct drm_i915_private *dev_priv)Signature · 1 line
1.9.1
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
PATCH v4 5/22 drm/i915/icl: WaModifyGamTlbPartitioning
In reply to Oscar MateoAdjust default GAM TLB partitioning for performance reasons.
v2: Only touch the bits that we really need
v3: Rebased on top of the WA refactoring
--- drivers/gpu/drm/i915/i915_reg.h | 5 +++++ drivers/gpu/drm/i915/intel_workarounds.c | 5 +++++ 2 files changed, 10 insertions(+)
drivers/gpu/drm/i915/i915_reg.h +5 -0
@@ -8228,6 +8228,11 @@ enum {#define GEN10_DFR_RATIO_EN_AND_CHICKEN _MMIO(0x9550)#define DFR_DISABLE (1 << 9) +#define GEN11_GACB_PERF_CTRL _MMIO(0x4B80) +#define GEN11_HASH_CTRL_MASK (0x3 << 12 | 0xf << 0) +#define GEN11_HASH_CTRL_BIT0 (1 << 0) +#define GEN11_HASH_CTRL_BIT4 (1 << 12) +/* IVYBRIDGE DPF */ #define GEN7_L3CDERRST1 (slice) _MMIO(0xB008 + (slice) * 0x200)/* L3CD Error Status 1 */ #define GEN7_L3CDERRST1_ROW_MASK (0x7ff<<14)
drivers/gpu/drm/i915/intel_workarounds.c +5 -0
@@ -721,6 +721,11 @@ static void icl_gt_workarounds_apply(struct drm_i915_private *dev_priv)I915_WRITE (GEN11_GLBLINVL , ((I915_READ (GEN11_GLBLINVL ) & ~GEN11_BANK_HASH_ADDR_EXCL_MASK ) |GEN11_BANK_HASH_ADDR_EXCL_BIT0 )); + +/* WaModifyGamTlbPartitioning:icl */ +I915_WRITE (GEN11_GACB_PERF_CTRL , + ((I915_READ (GEN11_GACB_PERF_CTRL ) & ~GEN11_HASH_CTRL_MASK ) | +GEN11_HASH_CTRL_BIT0 |GEN11_HASH_CTRL_BIT4 )); }void intel_gt_workarounds_apply (struct drm_i915_private *dev_priv)
Signature · 6 lines
1.9.1 _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
PATCH v4 6/22 drm/i915/icl: WaDisableCleanEvicts
In reply to Oscar MateoAvoids an undefined LLC behavior.
BSpec: 9613
v2: Renamed to Wa_1405733216
v3: Spaces around '<<' and fix surrounding code
v4: Rebased on top of the WA refactoring
--- drivers/gpu/drm/i915/i915_reg.h | 5 +++-- drivers/gpu/drm/i915/intel_workarounds.c | 6 ++++++ 2 files changed, 9 insertions(+), 2 deletions(-)
drivers/gpu/drm/i915/i915_reg.h +3 -2
@@ -7197,8 +7197,9 @@ enum {#define L3SQ_URB_READ_CAM_MATCH_DISABLE (1<<27)#define GEN8_L3SQCREG4 _MMIO(0xb118) -#define GEN8_LQSC_RO_PERF_DIS (1<<27) -#define GEN8_LQSC_FLUSH_COHERENT_LINES (1<<21) +#define GEN11_LQSC_CLEAN_EVICT_DISABLE (1 << 6) +#define GEN8_LQSC_RO_PERF_DIS (1 << 27) +#define GEN8_LQSC_FLUSH_COHERENT_LINES (1 << 21)/* GEN8 chicken */ #define HDC_CHICKEN0 _MMIO(0x7300)
drivers/gpu/drm/i915/intel_workarounds.c +6 -0
@@ -726,6 +726,12 @@ static void icl_gt_workarounds_apply(struct drm_i915_private *dev_priv)I915_WRITE (GEN11_GACB_PERF_CTRL , ((I915_READ (GEN11_GACB_PERF_CTRL ) & ~GEN11_HASH_CTRL_MASK ) |GEN11_HASH_CTRL_BIT0 |GEN11_HASH_CTRL_BIT4 )); + +/* Wa_1405733216:icl +* Formerly known as WaDisableCleanEvicts +*/ +I915_WRITE (GEN8_L3SQCREG4 , (I915_READ (GEN8_L3SQCREG4 ) | +GEN11_LQSC_CLEAN_EVICT_DISABLE )); }void intel_gt_workarounds_apply (struct drm_i915_private *dev_priv)
Signature · 6 lines
1.9.1 _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
PATCH v4 7/22 drm/i915/icl: WaCL2SFHalfMaxAlloc
In reply to Oscar MateoThis workarounds an issue with insufficient storage for the CL2 and SF units.
v2: Renamed to Wa_1405766107
v3: Wrapped the commit message
v4: Rebased on top of the WA refactoring
--- drivers/gpu/drm/i915/i915_reg.h | 4 ++++ drivers/gpu/drm/i915/intel_workarounds.c | 7 +++++++ 2 files changed, 11 insertions(+)
drivers/gpu/drm/i915/i915_reg.h +4 -0
@@ -8234,6 +8234,10 @@ enum {#define GEN11_HASH_CTRL_BIT0 (1 << 0)#define GEN11_HASH_CTRL_BIT4 (1 << 12) +#define GEN11_LSN_UNSLCVC _MMIO(0xB43C) +#define GEN11_LSN_UNSLCVC_GAFS_HALF_CL2_MACALLOC (1 << 9) +#define GEN11_LSN_UNSLCVC_GAFS_HALF_SF_MAXALLOC (1 << 7) +/* IVYBRIDGE DPF */ #define GEN7_L3CDERRST1 (slice) _MMIO(0xB008 + (slice) * 0x200)/* L3CD Error Status 1 */ #define GEN7_L3CDERRST1_ROW_MASK (0x7ff<<14)
drivers/gpu/drm/i915/intel_workarounds.c +7 -0
@@ -732,6 +732,13 @@ static void icl_gt_workarounds_apply(struct drm_i915_private *dev_priv)*/ I915_WRITE (GEN8_L3SQCREG4 , (I915_READ (GEN8_L3SQCREG4 ) |GEN11_LQSC_CLEAN_EVICT_DISABLE )); + +/* Wa_1405766107:icl +* Formerly known as WaCL2SFHalfMaxAlloc +*/ +I915_WRITE (GEN11_LSN_UNSLCVC , (I915_READ (GEN11_LSN_UNSLCVC ) | +GEN11_LSN_UNSLCVC_GAFS_HALF_SF_MAXALLOC | +GEN11_LSN_UNSLCVC_GAFS_HALF_CL2_MACALLOC )); }void intel_gt_workarounds_apply (struct drm_i915_private *dev_priv)
Signature · 6 lines
1.9.1 _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
PATCH v4 8/22 drm/i915/icl: WaDisCtxReload
In reply to Oscar MateoRevert to the legacy implementation to avoid a system hang.
v2: Correct the address for GAMW_ECO_DEV_RW_IA_REG
v3: Renamed to Wa_220166154
v4: Rebased on top of the WA refactoring
--- drivers/gpu/drm/i915/i915_reg.h | 3 +++ drivers/gpu/drm/i915/intel_workarounds.c | 6 ++++++ 2 files changed, 9 insertions(+)
drivers/gpu/drm/i915/i915_reg.h +3 -0
@@ -8238,6 +8238,9 @@ enum {#define GEN11_LSN_UNSLCVC_GAFS_HALF_CL2_MACALLOC (1 << 9)#define GEN11_LSN_UNSLCVC_GAFS_HALF_SF_MAXALLOC (1 << 7) +#define GAMW_ECO_DEV_RW_IA_REG _MMIO(0x4080) +#define GAMW_ECO_DEV_CTX_RELOAD_DISABLE (1 << 7) +/* IVYBRIDGE DPF */ #define GEN7_L3CDERRST1 (slice) _MMIO(0xB008 + (slice) * 0x200)/* L3CD Error Status 1 */ #define GEN7_L3CDERRST1_ROW_MASK (0x7ff<<14)
drivers/gpu/drm/i915/intel_workarounds.c +6 -0
@@ -739,6 +739,12 @@ static void icl_gt_workarounds_apply(struct drm_i915_private *dev_priv)I915_WRITE (GEN11_LSN_UNSLCVC , (I915_READ (GEN11_LSN_UNSLCVC ) |GEN11_LSN_UNSLCVC_GAFS_HALF_SF_MAXALLOC |GEN11_LSN_UNSLCVC_GAFS_HALF_CL2_MACALLOC )); + +/* Wa_220166154:icl +* Formerly known as WaDisCtxReload +*/ +I915_WRITE (GAMW_ECO_DEV_RW_IA_REG , (I915_READ (GAMW_ECO_DEV_RW_IA_REG ) | +GAMW_ECO_DEV_CTX_RELOAD_DISABLE )); }void intel_gt_workarounds_apply (struct drm_i915_private *dev_priv)
Signature · 6 lines
1.9.1 _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
PATCH v4 9/22 drm/i915/icl: Wa_1405779004
In reply to Oscar MateoDisable MSC clock gating to prevent data corruption.
BSpec: 19257
v2: Rebased on top of the WA refactoring
--- drivers/gpu/drm/i915/i915_reg.h | 1 + drivers/gpu/drm/i915/intel_workarounds.c | 6 ++++++ 2 files changed, 7 insertions(+)
drivers/gpu/drm/i915/i915_reg.h +1 -0
@@ -3840,6 +3840,7 @@ enum {#define SLICE_UNIT_LEVEL_CLKGATE _MMIO(0x94d4)#define SARBUNIT_CLKGATE_DIS (1 << 5)#define RCCUNIT_CLKGATE_DIS (1 << 7) +#define MSCUNIT_CLKGATE_DIS (1 << 10)#define SUBSLICE_UNIT_LEVEL_CLKGATE _MMIO(0x9524)#define GWUNIT_CLKGATE_DIS (1 << 16)
drivers/gpu/drm/i915/intel_workarounds.c +6 -0
@@ -745,6 +745,12 @@ static void icl_gt_workarounds_apply(struct drm_i915_private *dev_priv)*/ I915_WRITE (GAMW_ECO_DEV_RW_IA_REG , (I915_READ (GAMW_ECO_DEV_RW_IA_REG ) |GAMW_ECO_DEV_CTX_RELOAD_DISABLE )); + +/* Wa_1405779004:icl (pre-prod) */ +if (IS_ICL_REVID (dev_priv,ICL_REVID_A0 ,ICL_REVID_B0 )) +I915_WRITE (SLICE_UNIT_LEVEL_CLKGATE , + (I915_READ (SLICE_UNIT_LEVEL_CLKGATE ) | +MSCUNIT_CLKGATE_DIS )); }void intel_gt_workarounds_apply (struct drm_i915_private *dev_priv)
Signature · 6 lines
1.9.1 _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
PATCH v4 10/22 drm/i915/icl: Wa_1406680159 and Wa_2201832410
In reply to Oscar MateoDisable GWL clock gating to prevent two different issues that might cause hangs.
Please notice that one of the issues is pre-production only.
v2: Rebased on top of the WA refactoring
--- drivers/gpu/drm/i915/intel_workarounds.c | 6 ++++++ 1 file changed, 6 insertions(+)
drivers/gpu/drm/i915/intel_workarounds.c +6 -0
@@ -751,6 +751,12 @@ static void icl_gt_workarounds_apply(struct drm_i915_private *dev_priv)I915_WRITE (SLICE_UNIT_LEVEL_CLKGATE , (I915_READ (SLICE_UNIT_LEVEL_CLKGATE ) |MSCUNIT_CLKGATE_DIS )); + +/* Wa_1406680159:icl */ +/* Wa_2201832410:icl (pre-prod, only until C0) */ +I915_WRITE (SUBSLICE_UNIT_LEVEL_CLKGATE , + (I915_READ (SUBSLICE_UNIT_LEVEL_CLKGATE ) | +GWUNIT_CLKGATE_DIS )); }void intel_gt_workarounds_apply (struct drm_i915_private *dev_priv)
Signature · 6 lines
1.9.1 _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Rodrigo Vivi
maintainerRe: [PATCH 10/22] drm/i915/icl: Wa_1406680159 and Wa_2201832410
Quoted 18 lines · Oscar Mateo
Disable GWL clock gating to prevent two different issues that might cause hangs.
Please notice that one of the issues is pre-production only.
v2: Rebased on top of the WA refactoring
Cc Mika Kuoppala <[email protected]>Signed-off-by Oscar Mateo <[email protected]>--- drivers/gpu/drm/i915/intel_workarounds.c | 6 ++++++ 1 file changed, 6 insertions(+)drivers/gpu/drm/i915/intel_workarounds.c +0
-0
drivers/gpu/drm/i915/intel_workarounds.c static void icl_gt_workarounds_apply(struct drm_i915_private *dev_priv) · line 756
Quoted 6 lines
drivers/gpu/drm/i915/intel_workarounds.c +3
-0@@ -751,6 +751,12 @@ static void icl_gt_workarounds_apply(struct drm_i915_private *dev_priv)I915_WRITE (SLICE_UNIT_LEVEL_CLKGATE , (I915_READ (SLICE_UNIT_LEVEL_CLKGATE ) |MSCUNIT_CLKGATE_DIS )); + +/* Wa_1406680159:icl */ +/* Wa_2201832410:icl (pre-prod, only until C0) */
what about adding the REVID checks?
but, well, why one is pre-prod and other is forever?
Quoted 13 lines
drivers/gpu/drm/i915/intel_workarounds.c +3
-0+I915_WRITE (SUBSLICE_UNIT_LEVEL_CLKGATE , + (I915_READ (SUBSLICE_UNIT_LEVEL_CLKGATE ) | +GWUNIT_CLKGATE_DIS )); }void intel_gt_workarounds_apply (struct drm_i915_private *dev_priv)Signature · 6 lines
1.9.1 _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Re: [PATCH 10/22] drm/i915/icl: Wa_1406680159 and Wa_2201832410
Quoted 27 lines · Rodrigo Vivi
Quoted 18 lines · Oscar Mateo
Disable GWL clock gating to prevent two different issues that might cause hangs.
Please notice that one of the issues is pre-production only.
v2: Rebased on top of the WA refactoring
Cc Mika Kuoppala <[email protected]>Signed-off-by Oscar Mateo <[email protected]>--- drivers/gpu/drm/i915/intel_workarounds.c | 6 ++++++ 1 file changed, 6 insertions(+)drivers/gpu/drm/i915/intel_workarounds.c +0
-0drivers/gpu/drm/i915/intel_workarounds.c static void icl_gt_workarounds_apply(struct drm_i915_private *dev_priv) · line 756
Quoted 6 lines
drivers/gpu/drm/i915/intel_workarounds.c +3
-0@@ -751,6 +751,12 @@ static void icl_gt_workarounds_apply(struct drm_i915_private *dev_priv)I915_WRITE (SLICE_UNIT_LEVEL_CLKGATE , (I915_READ (SLICE_UNIT_LEVEL_CLKGATE ) |MSCUNIT_CLKGATE_DIS )); + +/* Wa_1406680159:icl */ +/* Wa_2201832410:icl (pre-prod, only until C0) */ what about adding the REVID checks?
Quoted 1 line
but, well, why one is pre-prod and other is forever?
That's the thing: this is the same workaround for two different HW problems. One is fixed post-C0, but the other is not, so I cannot add REVID checks :(
Quoted 6 lines
Quoted 6 lines
drivers/gpu/drm/i915/intel_workarounds.c +3
-0+I915_WRITE (SUBSLICE_UNIT_LEVEL_CLKGATE , + (I915_READ (SUBSLICE_UNIT_LEVEL_CLKGATE ) | +GWUNIT_CLKGATE_DIS )); }
Quoted 8 lines
Quoted 8 lines
drivers/gpu/drm/i915/intel_workarounds.c +0
-0void intel_gt_workarounds_apply (struct drm_i915_private *dev_priv)Signature · 6 lines
1.9.1 _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Rodrigo Vivi
maintainerRe: [PATCH 10/22] drm/i915/icl: Wa_1406680159 and Wa_2201832410
Quoted 30 lines · Oscar Mateo
Quoted 27 lines · Rodrigo Vivi
Quoted 18 lines · Oscar Mateo
Disable GWL clock gating to prevent two different issues that might cause hangs.
Please notice that one of the issues is pre-production only.
v2: Rebased on top of the WA refactoring
Cc Mika Kuoppala <[email protected]>Signed-off-by Oscar Mateo <[email protected]>--- drivers/gpu/drm/i915/intel_workarounds.c | 6 ++++++ 1 file changed, 6 insertions(+)drivers/gpu/drm/i915/intel_workarounds.c +0
-0drivers/gpu/drm/i915/intel_workarounds.c static void icl_gt_workarounds_apply(struct drm_i915_private *dev_priv) · line 756
Quoted 6 lines
drivers/gpu/drm/i915/intel_workarounds.c +3
-0@@ -751,6 +751,12 @@ static void icl_gt_workarounds_apply(struct drm_i915_private *dev_priv)I915_WRITE (SLICE_UNIT_LEVEL_CLKGATE , (I915_READ (SLICE_UNIT_LEVEL_CLKGATE ) |MSCUNIT_CLKGATE_DIS )); + +/* Wa_1406680159:icl */ +/* Wa_2201832410:icl (pre-prod, only until C0) */ what about adding the REVID checks?
Quoted 4 lines
Quoted 1 line
but, well, why one is pre-prod and other is forever?
That's the thing: this is the same workaround for two different HW problems. One is fixed post-C0, but the other is not, so I cannot add REVID checks :(
Oh of course... Makes sense. Thanks ;)
Quoted 10 lines
Quoted 10 lines
Quoted 10 lines
drivers/gpu/drm/i915/intel_workarounds.c +3
-0+I915_WRITE (SUBSLICE_UNIT_LEVEL_CLKGATE , + (I915_READ (SUBSLICE_UNIT_LEVEL_CLKGATE ) | +GWUNIT_CLKGATE_DIS )); }void intel_gt_workarounds_apply (struct drm_i915_private *dev_priv)Signature · 1 line
1.9.1
Quoted 9 lines
Quoted 4 lines
Quoted 4 lines
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Re: [PATCH 10/22] drm/i915/icl: Wa_1406680159 and Wa_2201832410
Quoted 30 lines · Rodrigo Vivi
Quoted 29 lines · Oscar Mateo
Quoted 27 lines · Rodrigo Vivi
Quoted 18 lines · Oscar Mateo
Disable GWL clock gating to prevent two different issues that might cause hangs.
Please notice that one of the issues is pre-production only.
v2: Rebased on top of the WA refactoring
Cc Mika Kuoppala <[email protected]>Signed-off-by Oscar Mateo <[email protected]>--- drivers/gpu/drm/i915/intel_workarounds.c | 6 ++++++ 1 file changed, 6 insertions(+)drivers/gpu/drm/i915/intel_workarounds.c +0
-0drivers/gpu/drm/i915/intel_workarounds.c static void icl_gt_workarounds_apply(struct drm_i915_private *dev_priv) · line 756
Quoted 6 lines
drivers/gpu/drm/i915/intel_workarounds.c +3
-0@@ -751,6 +751,12 @@ static void icl_gt_workarounds_apply(struct drm_i915_private *dev_priv)I915_WRITE (SLICE_UNIT_LEVEL_CLKGATE , (I915_READ (SLICE_UNIT_LEVEL_CLKGATE ) |MSCUNIT_CLKGATE_DIS )); + +/* Wa_1406680159:icl */ +/* Wa_2201832410:icl (pre-prod, only until C0) */ what about adding the REVID checks?
Quoted 4 lines
Quoted 3 lines
Quoted 1 line
but, well, why one is pre-prod and other is forever?
That's the thing: this is the same workaround for two different HW problems. One is fixed post-C0, but the other is not, so I cannot add REVID checks :(
Oh of course... Makes sense. Thanks ;)
Good news: Wa_2201832410 has been officially merged with Wa_1406680159, so I can remove the confusing comment \o/
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
drivers/gpu/drm/i915/intel_workarounds.c +3
-0+I915_WRITE (SUBSLICE_UNIT_LEVEL_CLKGATE , + (I915_READ (SUBSLICE_UNIT_LEVEL_CLKGATE ) | +GWUNIT_CLKGATE_DIS )); }void intel_gt_workarounds_apply (struct drm_i915_private *dev_priv)Signature · 1 line
1.9.1
Quoted 8 lines
Quoted 8 lines
Quoted 4 lines
Quoted 4 lines
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
PATCH v4 11/22 drm/i915/icl: Wa_1604302699
In reply to Oscar MateoDisable I2M Write for performance reasons.
v2: Rebased on top of the WA refactoring
--- drivers/gpu/drm/i915/i915_reg.h | 4 +++- drivers/gpu/drm/i915/intel_workarounds.c | 5 +++++ 2 files changed, 8 insertions(+), 1 deletion(-)
drivers/gpu/drm/i915/i915_reg.h +3 -1
@@ -7192,7 +7192,9 @@ enum {#define GEN7_L3CNTLREG3 _MMIO(0xB024)#define GEN7_L3_CHICKEN_MODE_REGISTER _MMIO(0xB030) -#define GEN7_WA_L3_CHICKEN_MODE 0x20000000 +#define GEN10_L3_CHICKEN_MODE_REGISTER _MMIO(0xB114) +#define GEN7_WA_L3_CHICKEN_MODE 0x20000000 +#define GEN11_I2M_WRITE_DISABLE (1 << 28)#define GEN7_L3SQCREG4 _MMIO(0xb034)#define L3SQ_URB_READ_CAM_MATCH_DISABLE (1<<27)
drivers/gpu/drm/i915/intel_workarounds.c +5 -0
@@ -757,6 +757,11 @@ static void icl_gt_workarounds_apply(struct drm_i915_private *dev_priv)I915_WRITE (SUBSLICE_UNIT_LEVEL_CLKGATE , (I915_READ (SUBSLICE_UNIT_LEVEL_CLKGATE ) |GWUNIT_CLKGATE_DIS )); + +/* Wa_1604302699:icl */ +I915_WRITE (GEN10_L3_CHICKEN_MODE_REGISTER , + (I915_READ (GEN10_L3_CHICKEN_MODE_REGISTER ) | +GEN11_I2M_WRITE_DISABLE )); }void intel_gt_workarounds_apply (struct drm_i915_private *dev_priv)
Signature · 6 lines
1.9.1 _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
PATCH v4 12/22 drm/i915/icl: Wa_1406838659
In reply to Oscar MateoDisable CGPSF unit clock gating to prevent an issue.
v2: Rebased on top of the WA refactoring
--- drivers/gpu/drm/i915/i915_reg.h | 13 ++++++++----- drivers/gpu/drm/i915/intel_workarounds.c | 6 ++++++ 2 files changed, 14 insertions(+), 5 deletions(-)
drivers/gpu/drm/i915/i915_reg.h +8 -5
@@ -3838,15 +3838,18 @@ enum {* GEN10 clock gating regs */ #define SLICE_UNIT_LEVEL_CLKGATE _MMIO(0x94d4) -#define SARBUNIT_CLKGATE_DIS (1 << 5) -#define RCCUNIT_CLKGATE_DIS (1 << 7) -#define MSCUNIT_CLKGATE_DIS (1 << 10) +#define SARBUNIT_CLKGATE_DIS (1 << 5) +#define RCCUNIT_CLKGATE_DIS (1 << 7) +#define MSCUNIT_CLKGATE_DIS (1 << 10)#define SUBSLICE_UNIT_LEVEL_CLKGATE _MMIO(0x9524) -#define GWUNIT_CLKGATE_DIS (1 << 16) +#define GWUNIT_CLKGATE_DIS (1 << 16)#define UNSLICE_UNIT_LEVEL_CLKGATE _MMIO(0x9434) -#define VFUNIT_CLKGATE_DIS (1 << 20) +#define VFUNIT_CLKGATE_DIS (1 << 20) + +#define INF_UNIT_LEVEL_CLKGATE _MMIO(0x9560) +#define CGPSF_CLKGATE_DIS (1 << 3)/* * Display engine regs
drivers/gpu/drm/i915/intel_workarounds.c +6 -0
@@ -762,6 +762,12 @@ static void icl_gt_workarounds_apply(struct drm_i915_private *dev_priv)I915_WRITE (GEN10_L3_CHICKEN_MODE_REGISTER , (I915_READ (GEN10_L3_CHICKEN_MODE_REGISTER ) |GEN11_I2M_WRITE_DISABLE )); + +/* Wa_1406838659:icl (pre-prod) */ +if (IS_ICL_REVID (dev_priv,ICL_REVID_A0 ,ICL_REVID_C0 )) +I915_WRITE (INF_UNIT_LEVEL_CLKGATE , + (I915_READ (INF_UNIT_LEVEL_CLKGATE ) | +CGPSF_CLKGATE_DIS )); }void intel_gt_workarounds_apply (struct drm_i915_private *dev_priv)
Signature · 6 lines
1.9.1 _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
PATCH v4 13/22 drm/i915/icl: WaForwardProgressSoftReset
In reply to Oscar MateoAvoids a hang during soft reset.
v2: Rebased on top of the WA refactoring
--- drivers/gpu/drm/i915/i915_reg.h | 5 +++++ drivers/gpu/drm/i915/intel_workarounds.c | 8 ++++++++ 2 files changed, 13 insertions(+)
drivers/gpu/drm/i915/i915_reg.h +5 -0
@@ -9711,6 +9711,11 @@ enum skl_power_gate {#define GEN9_VEBOX_MOCS (i) _MMIO(0xcb00 + (i) * 4)/* Video MOCS registers */ #define GEN9_BLT_MOCS (i) _MMIO(0xcc00 + (i) * 4)/* Blitter MOCS registers */ +#define GEN10_SCRATCH_LNCF2 _MMIO(0xb0a0) +#define PMFLUSHDONE_LNICRSDROP (1 << 20) +#define PMFLUSH_GAPL3UNBLOCK (1 << 21) +#define PMFLUSHDONE_LNEBLK (1 << 22) +/* gamt regs */ #define GEN8_L3_LRA_1_GPGPU _MMIO(0x4dd4)#define GEN8_L3_LRA_1_GPGPU_DEFAULT_VALUE_BDW 0x67F1427F/* max/min for LRA1/2 */
drivers/gpu/drm/i915/intel_workarounds.c +8 -0
@@ -768,6 +768,14 @@ static void icl_gt_workarounds_apply(struct drm_i915_private *dev_priv)I915_WRITE (INF_UNIT_LEVEL_CLKGATE , (I915_READ (INF_UNIT_LEVEL_CLKGATE ) |CGPSF_CLKGATE_DIS )); + +/* WaForwardProgressSoftReset:icl (pre-prod) */ +if (IS_ICL_REVID (dev_priv,ICL_REVID_A0 ,ICL_REVID_B0 )) +I915_WRITE (GEN10_SCRATCH_LNCF2 , + (I915_READ (GEN10_SCRATCH_LNCF2 ) | +PMFLUSHDONE_LNICRSDROP | +PMFLUSH_GAPL3UNBLOCK | +PMFLUSHDONE_LNEBLK )); }void intel_gt_workarounds_apply (struct drm_i915_private *dev_priv)
Signature · 6 lines
1.9.1 _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
PATCH v4 14/22 drm/i915/icl: WaDisableImprovedTdlClkGating
In reply to Oscar MateoRevert to the legacy implementation.
v2: GEN7_ROW_CHICKEN2 is masked
v3:
- Rebased
- Renamed to Wa_2006611047
- A0 and B0 only
v4:
- Add spaces around '<<' (and fix the surrounding code as well)
- Mark the WA as pre-prod
v5: Rebased on top of the WA refactoring
--- drivers/gpu/drm/i915/i915_reg.h | 5 +++-- drivers/gpu/drm/i915/intel_workarounds.c | 7 +++++++ 2 files changed, 10 insertions(+), 2 deletions(-)
drivers/gpu/drm/i915/i915_reg.h +3 -2
@@ -8284,8 +8284,9 @@ enum {#define GEN7_ROW_CHICKEN2 _MMIO(0xe4f4)#define GEN7_ROW_CHICKEN2_GT2 _MMIO(0xf4f4) -#define DOP_CLOCK_GATING_DISABLE (1<<0) -#define PUSH_CONSTANT_DEREF_DISABLE (1<<8) +#define DOP_CLOCK_GATING_DISABLE (1 << 0) +#define PUSH_CONSTANT_DEREF_DISABLE (1 << 8) +#define GEN11_TDL_CLOCK_GATING_FIX_DISABLE (1 << 1)#define HSW_ROW_CHICKEN3 _MMIO(0xe49c)#define HSW_ROW_CHICKEN3_L3_GLOBAL_ATOMICS_DISABLE (1 << 6)
drivers/gpu/drm/i915/intel_workarounds.c +7 -0
@@ -459,6 +459,13 @@ static int icl_ctx_workarounds_init(struct drm_i915_private *dev_priv)*/ WA_SET_BIT_MASKED (ICL_HDC_CHICKEN0 ,HDC_FORCE_NON_COHERENT ); +/* Wa_2006611047:icl (pre-prod) +* Formerly known as WaDisableImprovedTdlClkGating +*/ +if (IS_ICL_REVID (dev_priv,ICL_REVID_A0 ,ICL_REVID_B0 )) +WA_SET_BIT_MASKED (GEN7_ROW_CHICKEN2 , +GEN11_TDL_CLOCK_GATING_FIX_DISABLE ); +return 0 ; }
Signature · 6 lines
1.9.1 _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
PATCH v4 15/22 drm/i915/icl: WaEnableStateCacheRedirectToCS
In reply to Oscar MateoRedirects the state cache to the CS Command buffer section for performance reasons.
v2: Rebased
v3: Rebased on top of the WA refactoring
--- drivers/gpu/drm/i915/i915_reg.h | 1 + drivers/gpu/drm/i915/intel_workarounds.c | 4 ++++ 2 files changed, 5 insertions(+)
drivers/gpu/drm/i915/i915_reg.h +1 -0
@@ -7173,6 +7173,7 @@ enum {#define DISABLE_PIXEL_MASK_CAMMING (1<<14)#define GEN9_SLICE_COMMON_ECO_CHICKEN1 _MMIO(0x731c) +#define GEN11_STATE_CACHE_REDIRECT_TO_CS (1 << 11)#define GEN7_L3SQCREG1 _MMIO(0xB010)#define VLV_B0_WA_L3SQCREG1_VALUE 0x00D30000
drivers/gpu/drm/i915/intel_workarounds.c +4 -0
@@ -466,6 +466,10 @@ static int icl_ctx_workarounds_init(struct drm_i915_private *dev_priv)WA_SET_BIT_MASKED (GEN7_ROW_CHICKEN2 ,GEN11_TDL_CLOCK_GATING_FIX_DISABLE ); +/* WaEnableStateCacheRedirectToCS:icl */ +WA_SET_BIT_MASKED (GEN9_SLICE_COMMON_ECO_CHICKEN1 , +GEN11_STATE_CACHE_REDIRECT_TO_CS ); +return 0 ; }
Signature · 6 lines
1.9.1 _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
PATCH v4 16/22 drm/i915/icl: Wa_2006665173
In reply to Oscar MateoDisable blend embellishment in RCC.
v2: Rebased on top of the WA refactoring
--- drivers/gpu/drm/i915/i915_reg.h | 18 +++++++++++------- drivers/gpu/drm/i915/intel_workarounds.c | 5 +++++ 2 files changed, 16 insertions(+), 7 deletions(-)
drivers/gpu/drm/i915/i915_reg.h +11 -7
@@ -7157,13 +7157,17 @@ enum {/* GEN7 chicken */ #define GEN7_COMMON_SLICE_CHICKEN1 _MMIO(0x7010) -# define GEN7_CSC1_RHWO_OPT_DISABLE_IN_RCC ((1<<10) | (1<<26)) -# define GEN9_RHWO_OPTIMIZATION_DISABLE (1<<14) -#define COMMON_SLICE_CHICKEN2 _MMIO(0x7014) -# define GEN9_PBE_COMPRESSED_HASH_SELECTION (1<<13) -# define GEN9_DISABLE_GATHER_AT_SET_SHADER_COMMON_SLICE (1<<12) -# define GEN8_SBE_DISABLE_REPLAY_BUF_OPTIMIZATION (1<<8) -# define GEN8_CSC2_SBE_VUE_CACHE_CONSERVATIVE (1<<0) +#define GEN7_CSC1_RHWO_OPT_DISABLE_IN_RCC ((1 << 10) | (1 << 26)) +#define GEN9_RHWO_OPTIMIZATION_DISABLE (1 << 14) + +#define COMMON_SLICE_CHICKEN2 _MMIO(0x7014) +#define GEN9_PBE_COMPRESSED_HASH_SELECTION (1 << 13) +#define GEN9_DISABLE_GATHER_AT_SET_SHADER_COMMON_SLICE (1 << 12) +#define GEN8_SBE_DISABLE_REPLAY_BUF_OPTIMIZATION (1 << 8) +#define GEN8_CSC2_SBE_VUE_CACHE_CONSERVATIVE (1 << 0) + +#define GEN11_COMMON_SLICE_CHICKEN3 _MMIO(0x7304) +#define GEN11_BLEND_EMB_FIX_DISABLE_IN_RCC (1 << 11)#define HIZ_CHICKEN _MMIO(0x7018)# define CHV_HZ_8X8_MODE_IN_1X (1<<15)
drivers/gpu/drm/i915/intel_workarounds.c +5 -0
@@ -470,6 +470,11 @@ static int icl_ctx_workarounds_init(struct drm_i915_private *dev_priv)WA_SET_BIT_MASKED (GEN9_SLICE_COMMON_ECO_CHICKEN1 ,GEN11_STATE_CACHE_REDIRECT_TO_CS ); +/* Wa_2006665173:icl (pre-prod) */ +if (IS_ICL_REVID (dev_priv,ICL_REVID_A0 ,ICL_REVID_B0 )) +WA_SET_BIT_MASKED (GEN11_COMMON_SLICE_CHICKEN3 , +GEN11_BLEND_EMB_FIX_DISABLE_IN_RCC ); +return 0 ; }
Signature · 6 lines
1.9.1 _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
PATCH v4 17/22 drm/i915/icl: WaEnableFloatBlendOptimization
In reply to Oscar MateoEnables blend optimization for floating point RTs
v2: Rebased on top of the WA refactoring
--- drivers/gpu/drm/i915/i915_reg.h | 3 +++ drivers/gpu/drm/i915/intel_workarounds.c | 3 +++ 2 files changed, 6 insertions(+)
drivers/gpu/drm/i915/i915_reg.h +3 -0
@@ -2663,6 +2663,9 @@ enum i915_power_well_id {#define GEN8_4x4_STC_OPTIMIZATION_DISABLE (1<<6)#define GEN9_PARTIAL_RESOLVE_IN_VC_DISABLE (1<<1) +#define GEN10_CACHE_MODE_SS _MMIO(0xe420) +#define FLOAT_BLEND_OPTIMIZATION_ENABLE (1 << 4) +#define GEN6_BLITTER_ECOSKPD _MMIO(0x221d0)#define GEN6_BLITTER_LOCK_SHIFT 16#define GEN6_BLITTER_FBC_NOTIFY (1<<3)
drivers/gpu/drm/i915/intel_workarounds.c +3 -0
@@ -475,6 +475,9 @@ static int icl_ctx_workarounds_init(struct drm_i915_private *dev_priv)WA_SET_BIT_MASKED (GEN11_COMMON_SLICE_CHICKEN3 ,GEN11_BLEND_EMB_FIX_DISABLE_IN_RCC ); +/* WaEnableFloatBlendOptimization:icl */ +WA_SET_BIT_MASKED (GEN10_CACHE_MODE_SS ,FLOAT_BLEND_OPTIMIZATION_ENABLE ); +return 0 ; }
Signature · 6 lines
1.9.1 _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
PATCH v4 18/22 drm/i915/icl: WaSendPushConstantsFromMMIO
In reply to Oscar MateoAllows UMDs to set 'Disable Gather at Set Shader Common Slice'.
Do Linux UMDs make use of this? This change has been security reviewed and the whitelisting approved. Virtualization of other OSes could certainly use it...
v2: Rebased
v3: Rebased on top of the WA refactoring
v4: Rebased on top of the WA whitelist reg refactoring (Michel)
--- drivers/gpu/drm/i915/intel_workarounds.c | 2 ++ 1 file changed, 2 insertions(+)
drivers/gpu/drm/i915/intel_workarounds.c +2 -0
@@ -899,6 +899,8 @@ static void cnl_whitelist_build(struct whitelist *w)static void icl_whitelist_build (struct whitelist *w) { +/* WaSendPushConstantsFromMMIO:icl */ +whitelist_reg (w,COMMON_SLICE_CHICKEN2 ); } static struct whitelist *whitelist_build(struct intel_engine_cs *engine,
Signature · 6 lines
1.9.1 _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
PATCH v4 19/22 drm/i915/icl: WaAllowUMDToModifyHalfSliceChicken2
In reply to Oscar MateoRequired to dinamically set 'Small PL Lossless Fix Enable'
Do Linux UMDs make use of this? This change has been security reviewed and the whitelisting approved. Virtualization of other OSes could certainly use it.
v2: For whatever reason, this ended up in KBL (??!!)
v3: Rebased on top of the WA refactoring
v4: Rebased on top of whitelist reg refactoring (Michel)
--- drivers/gpu/drm/i915/intel_workarounds.c | 3 +++ 1 file changed, 3 insertions(+)
drivers/gpu/drm/i915/intel_workarounds.c +3 -0
@@ -901,6 +901,9 @@ static void icl_whitelist_build(struct whitelist *w) {/* WaSendPushConstantsFromMMIO:icl */ whitelist_reg (w,COMMON_SLICE_CHICKEN2 ); + +/* WaAllowUMDToModifyHalfSliceChicken2:icl */ +whitelist_reg (w,HALF_SLICE_CHICKEN2 ); } static struct whitelist *whitelist_build(struct intel_engine_cs *engine,
Signature · 6 lines
1.9.1 _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
PATCH v4 20/22 drm/i915/icl: WaAllowUMDToModifyHalfSliceChicken7
In reply to Oscar MateoRequired to dinamically set 'Trilinear Filter Quality Mode'
Do Linux UMDs make use of this? This change has been security reviewed and the whitelisting approved. Virtualization of other OSes could certainly use it.
v2: For whatever reason, this ended up in KBL (??!!)
v3: Rebased on top of the WA refactoring
v4: Rebased on top of the whitelist reg refactoring (Michel)
--- drivers/gpu/drm/i915/intel_workarounds.c | 3 +++ 1 file changed, 3 insertions(+)
drivers/gpu/drm/i915/intel_workarounds.c +3 -0
@@ -904,6 +904,9 @@ static void icl_whitelist_build(struct whitelist *w)/* WaAllowUMDToModifyHalfSliceChicken2:icl */ whitelist_reg (w,HALF_SLICE_CHICKEN2 ); + +/* WaAllowUMDToModifyHalfSliceChicken7:icl */ +whitelist_reg (w,GEN9_HALF_SLICE_CHICKEN7 ); } static struct whitelist *whitelist_build(struct intel_engine_cs *engine,
Signature · 6 lines
1.9.1 _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
PATCH v4 21/22 drm/i915/icl: WaAllowUmdWriteTRTTRootTable
In reply to Oscar MateoRequired for TR-TT (Tiled Resource Translation Table) support.
Do Linux UMDs make use of this? This change has been security reviewed and the whitelisting approved. Virtualization of other OSes could certainly use it.
v2: For whatever reason, this ended up in KBL (??!!)
v3: Rebased on top of the WA refactoring
v4: Rebased on top of whitelist reg refactoring (Michel)
--- drivers/gpu/drm/i915/i915_reg.h | 3 +++ drivers/gpu/drm/i915/intel_workarounds.c | 4 ++++ 2 files changed, 7 insertions(+)
drivers/gpu/drm/i915/i915_reg.h +3 -0
@@ -8255,6 +8255,9 @@ enum {#define GAMW_ECO_DEV_RW_IA_REG _MMIO(0x4080)#define GAMW_ECO_DEV_CTX_RELOAD_DISABLE (1 << 7) +#define TR_VA_TTL3_PTR_DW0 _MMIO(0x4DE0) +#define TR_VA_TTL3_PTR_DW1 _MMIO(0x4DE4) +/* IVYBRIDGE DPF */ #define GEN7_L3CDERRST1 (slice) _MMIO(0xB008 + (slice) * 0x200)/* L3CD Error Status 1 */ #define GEN7_L3CDERRST1_ROW_MASK (0x7ff<<14)
drivers/gpu/drm/i915/intel_workarounds.c +4 -0
@@ -907,6 +907,10 @@ static void icl_whitelist_build(struct whitelist *w)/* WaAllowUMDToModifyHalfSliceChicken7:icl */ whitelist_reg (w,GEN9_HALF_SLICE_CHICKEN7 ); + +/* WaAllowUmdWriteTRTTRootTable:icl */ +whitelist_reg (w,TR_VA_TTL3_PTR_DW0 ); +whitelist_reg (w,TR_VA_TTL3_PTR_DW1 ); } static struct whitelist *whitelist_build(struct intel_engine_cs *engine,
Signature · 6 lines
1.9.1 _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
PATCH v4 22/22 drm/i915/icl: WaAllowUMDToModifySamplerMode
In reply to Oscar MateoRequired for Bindless samplers.
Do Linux UMDs make use of this? This change has been security reviewed and the whitelisting approved. Virtualization of other OSes could certainly use it.
v2: Rebased on top of the WA refactoring (Michel)
--- drivers/gpu/drm/i915/i915_reg.h | 2 ++ drivers/gpu/drm/i915/intel_workarounds.c | 3 +++ 2 files changed, 5 insertions(+)
drivers/gpu/drm/i915/i915_reg.h +2 -0
@@ -8258,6 +8258,8 @@ enum {#define TR_VA_TTL3_PTR_DW0 _MMIO(0x4DE0)#define TR_VA_TTL3_PTR_DW1 _MMIO(0x4DE4) +#define GEN10_SAMPLER_MODE _MMIO(0xE18C) +/* IVYBRIDGE DPF */ #define GEN7_L3CDERRST1 (slice) _MMIO(0xB008 + (slice) * 0x200)/* L3CD Error Status 1 */ #define GEN7_L3CDERRST1_ROW_MASK (0x7ff<<14)
drivers/gpu/drm/i915/intel_workarounds.c +3 -0
@@ -911,6 +911,9 @@ static void icl_whitelist_build(struct whitelist *w)/* WaAllowUmdWriteTRTTRootTable:icl */ whitelist_reg (w,TR_VA_TTL3_PTR_DW0 );whitelist_reg (w,TR_VA_TTL3_PTR_DW1 ); + +/* WaAllowUMDToModifySamplerMode:icl */ +whitelist_reg (w,GEN10_SAMPLER_MODE ); } static struct whitelist *whitelist_build(struct intel_engine_cs *engine,
Signature · 6 lines
1.9.1 _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Patchwork
CI✗ Fi.CI.CHECKPATCH: warning for Workarounds for Icelake
In reply to Oscar Mateo== Series Details ==
== Series Details ==
Series: Workarounds for Icelake
URL : https://patchwork.freedesktop.org/series/42055/
State : warning
== Summary ==
$ dim checkpatch origin/drm-tip abf7b0421b03 drm/i915/icl: Introduce initial Icelake Workarounds -:48: CHECK:MACRO_ARG_REUSE: Macro argument reuse 'p' - possible side-effects? #48: FILE: drivers/gpu/drm/i915/i915_drv.h:2471: +#define IS_ICL_REVID(p, since, until) \ + (IS_ICELAKE(p) && IS_REVID(p, since, until))
total: 0 errors, 0 warnings, 1 checks, 155 lines checked 5ee197f360ed drm/i915/icl: Enable Sampler DFR 64f72e593847 drm/i915/icl: WaGAPZPriorityScheme 363ae679b17f drm/i915/icl: WaL3BankAddressHashing f4c1b4bb9090 drm/i915/icl: WaModifyGamTlbPartitioning bd0addeea3e4 drm/i915/icl: WaDisableCleanEvicts 614afb61f95c drm/i915/icl: WaCL2SFHalfMaxAlloc 816ee27c52c2 drm/i915/icl: WaDisCtxReload ed2ec7a07267 drm/i915/icl: Wa_1405779004 4adc5c7395a1 drm/i915/icl: Wa_1406680159 and Wa_2201832410 be63d1d1ad53 drm/i915/icl: Wa_1604302699 9702c4288bac drm/i915/icl: Wa_1406838659 c4772140f55f drm/i915/icl: WaForwardProgressSoftReset 7ddbeb426ced drm/i915/icl: WaDisableImprovedTdlClkGating d397ada30a53 drm/i915/icl: WaEnableStateCacheRedirectToCS 8e678b2a3d72 drm/i915/icl: Wa_2006665173 6416305b9e4a drm/i915/icl: WaEnableFloatBlendOptimization 1afd609b2026 drm/i915/icl: WaSendPushConstantsFromMMIO cfbc51ef5c4a drm/i915/icl: WaAllowUMDToModifyHalfSliceChicken2 a506e91880c4 drm/i915/icl: WaAllowUMDToModifyHalfSliceChicken7 3c8350e7a85b drm/i915/icl: WaAllowUmdWriteTRTTRootTable b3e83155f633 drm/i915/icl: WaAllowUMDToModifySamplerMode
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Patchwork
CI✗ Fi.CI.SPARSE: warning for Workarounds for Icelake
In reply to Oscar Mateo== Series Details ==
== Series Details ==
Series: Workarounds for Icelake
URL : https://patchwork.freedesktop.org/series/42055/
State : warning
== Summary ==
$ dim sparse origin/drm-tip Commit: drm/i915/icl: Introduce initial Icelake Workarounds -drivers/gpu/drm/i915/selftests/../i915_drv.h:3656:16: warning: expression using sizeof(void) +drivers/gpu/drm/i915/selftests/../i915_drv.h:3665:16: warning: expression using sizeof(void)
Commit: drm/i915/icl: Enable Sampler DFR
Okay!
Commit: drm/i915/icl: WaGAPZPriorityScheme
Okay!
Commit: drm/i915/icl: WaL3BankAddressHashing
Okay!
Commit: drm/i915/icl: WaModifyGamTlbPartitioning
Okay!
Commit: drm/i915/icl: WaDisableCleanEvicts
Okay!
Commit: drm/i915/icl: WaCL2SFHalfMaxAlloc
Okay!
Commit: drm/i915/icl: WaDisCtxReload
Okay!
Commit: drm/i915/icl: Wa_1405779004
Okay!
Commit: drm/i915/icl: Wa_1406680159 and Wa_2201832410 Okay!
Commit: drm/i915/icl: Wa_1604302699
Okay!
Commit: drm/i915/icl: Wa_1406838659
Okay!
Commit: drm/i915/icl: WaForwardProgressSoftReset
Okay!
Commit: drm/i915/icl: WaDisableImprovedTdlClkGating Okay!
Commit: drm/i915/icl: WaEnableStateCacheRedirectToCS Okay!
Commit: drm/i915/icl: Wa_2006665173
Okay!
Commit: drm/i915/icl: WaEnableFloatBlendOptimization Okay!
Commit: drm/i915/icl: WaSendPushConstantsFromMMIO
Okay!
Commit: drm/i915/icl: WaAllowUMDToModifyHalfSliceChicken2 Okay!
Commit: drm/i915/icl: WaAllowUMDToModifyHalfSliceChicken7 Okay!
Commit: drm/i915/icl: WaAllowUmdWriteTRTTRootTable Okay!
Commit: drm/i915/icl: WaAllowUMDToModifySamplerMode Okay!
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Patchwork
CI✓ Fi.CI.BAT: success for Workarounds for Icelake
In reply to Oscar Mateo== Series Details ==
== Series Details ==
Series: Workarounds for Icelake
URL : https://patchwork.freedesktop.org/series/42055/
State : success
== Summary ==
= CI Bug Log - changes from CI_DRM_4072 -> Patchwork_8768 =
== Summary - SUCCESS ==
No regressions found.
External URL: https://patchwork.freedesktop.org/api/1.0/series/42055/revisions/1/mbox/
== Known issues ==
Here are the changes found in Patchwork_8768 that come from known issues:
=== IGT changes ===
==== Issues hit ====
igt@debugfs_test@read_all_entries:
fi-snb-2520m: PASS -> INCOMPLETE (fdo#103713)
igt@drv_module_reload@basic-no-display:
fi-cnl-psr: NOTRUN -> DMESG-WARN (fdo#105395) +2
igt@gem_exec_suspend@basic-s4-devices:
fi-kbl-7500u: PASS -> DMESG-WARN (fdo#105128)
igt@gem_mmap_gtt@basic-small-bo-tiledx:
fi-gdg-551: PASS -> FAIL (fdo#102575)
==== Possible fixes ====
igt@kms_pipe_crc_basic@suspend-read-crc-pipe-c:
fi-ivb-3520m: DMESG-WARN (fdo#106084) -> PASS fdo#102575 https://bugs.freedesktop.org/show_bug.cgi?id=102575
fdo#103713 https://bugs.freedesktop.org/show_bug.cgi?id=103713
fdo#105128 https://bugs.freedesktop.org/show_bug.cgi?id=105128
fdo#105395 https://bugs.freedesktop.org/show_bug.cgi?id=105395
fdo#106084 https://bugs.freedesktop.org/show_bug.cgi?id=106084
== Participating hosts (33 -> 33) ==
Additional (2): fi-kbl-7560u fi-cnl-psr
Missing (2): fi-ilk-m540 fi-skl-6700hq
== Build changes ==
* Linux: CI_DRM_4072 -> Patchwork_8768
CI_DRM_4072: b35e59e5c6a9cae11d5183d2bf9c5c99ceedbc7c @ git://anongit.freedesktop.org/gfx-ci/linux
IGT_4442: 8168bb65d5e64d4df4e5d847d448bab2d2825d73 @ git://anongit.freedesktop.org/xorg/app/intel-gpu-tools
Patchwork_8768: b3e83155f6333c0a9db115f82db0290028c52055 @ git://anongit.freedesktop.org/gfx-ci/linux
piglit_4442: e60d247eb359f044caf0c09904da14e39d7adca1 @ git://anongit.freedesktop.org/piglit
== Linux commits ==
b3e83155f633 drm/i915/icl: WaAllowUMDToModifySamplerMode 3c8350e7a85b drm/i915/icl: WaAllowUmdWriteTRTTRootTable a506e91880c4 drm/i915/icl: WaAllowUMDToModifyHalfSliceChicken7 cfbc51ef5c4a drm/i915/icl: WaAllowUMDToModifyHalfSliceChicken2 1afd609b2026 drm/i915/icl: WaSendPushConstantsFromMMIO 6416305b9e4a drm/i915/icl: WaEnableFloatBlendOptimization 8e678b2a3d72 drm/i915/icl: Wa_2006665173d397ada30a53 drm/i915/icl: WaEnableStateCacheRedirectToCS 7ddbeb426ced drm/i915/icl: WaDisableImprovedTdlClkGating c4772140f55f drm/i915/icl: WaForwardProgressSoftReset 9702c4288bac drm/i915/icl: Wa_1406838659be63d1d1ad53 drm/i915/icl: Wa_16043026994adc5c7395a1 drm/i915/icl: Wa_1406680159 and Wa_2201832410 ed2ec7a07267 drm/i915/icl: Wa_1405779004816ee27c52c2 drm/i915/icl: WaDisCtxReload614afb61f95c drm/i915/icl: WaCL2SFHalfMaxAlloc bd0addeea3e4 drm/i915/icl: WaDisableCleanEvicts f4c1b4bb9090 drm/i915/icl: WaModifyGamTlbPartitioning 363ae679b17f drm/i915/icl: WaL3BankAddressHashing 64f72e593847 drm/i915/icl: WaGAPZPriorityScheme 5ee197f360ed drm/i915/icl: Enable Sampler DFR abf7b0421b03 drm/i915/icl: Introduce initial Icelake Workarounds
== Logs ==
For more details see: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_8768/issues.html
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Patchwork
CI✓ Fi.CI.IGT: success for Workarounds for Icelake
In reply to Oscar Mateo== Series Details ==
== Series Details ==
Series: Workarounds for Icelake
URL : https://patchwork.freedesktop.org/series/42055/
State : success
== Summary ==
= CI Bug Log - changes from CI_DRM_4072_full -> Patchwork_8768_full =
== Summary - WARNING ==
Minor unknown changes coming with Patchwork_8768_full need to be verified manually.
If you think the reported changes have nothing to do with the changes introduced in Patchwork_8768_full, please notify your bug team to allow them to document this new failure mode, which will reduce false positives in CI.
External URL: https://patchwork.freedesktop.org/api/1.0/series/42055/revisions/1/mbox/
== Possible new issues ==
Here are the unknown changes that may have been introduced in Patchwork_8768_full:
=== IGT changes ===
==== Warnings ====
igt@gem_exec_schedule@deep-bsd1:
shard-kbl: SKIP -> PASS +1
igt@gem_exec_schedule@deep-bsd2:
shard-kbl: PASS -> SKIP
igt@kms_busy@extended-modeset-hang-oldfb-render-a:
shard-glk: PASS -> SKIP +76
igt@kms_vblank@pipe-b-wait-forked-busy-hang:
shard-glk: SKIP -> PASS +104== Known issues ==
Here are the changes found in Patchwork_8768_full that come from known issues:
=== IGT changes ===
==== Issues hit ====
igt@gem_exec_schedule@wide-bsd2:
shard-kbl: PASS -> DMESG-WARN (fdo#105602, fdo#103558)
igt@gem_ppgtt@blt-vs-render-ctxn:
shard-kbl: PASS -> INCOMPLETE (fdo#106023, fdo#103665)
igt@kms_flip@2x-flip-vs-expired-vblank:
shard-hsw: PASS -> FAIL (fdo#102887)
igt@kms_flip@flip-vs-expired-vblank:
shard-hsw: PASS -> FAIL (fdo#105707)
==== Possible fixes ====
igt@kms_flip@dpms-vs-vblank-race-interruptible:
shard-glk: FAIL (fdo#103060) -> PASS
igt@kms_flip@plain-flip-fb-recreate:
shard-hsw: FAIL (fdo#100368) -> PASS
shard-glk: FAIL (fdo#100368) -> PASS +1
igt@kms_sysfs_edid_timing:
shard-apl: WARN (fdo#100047) -> PASS
igt@perf@blocking:
shard-hsw: FAIL (fdo#102252) -> PASS fdo#100047 https://bugs.freedesktop.org/show_bug.cgi?id=100047
fdo#100368 https://bugs.freedesktop.org/show_bug.cgi?id=100368
fdo#102252 https://bugs.freedesktop.org/show_bug.cgi?id=102252
fdo#102887 https://bugs.freedesktop.org/show_bug.cgi?id=102887
fdo#103060 https://bugs.freedesktop.org/show_bug.cgi?id=103060
fdo#103558 https://bugs.freedesktop.org/show_bug.cgi?id=103558
fdo#103665 https://bugs.freedesktop.org/show_bug.cgi?id=103665
fdo#105602 https://bugs.freedesktop.org/show_bug.cgi?id=105602
fdo#105707 https://bugs.freedesktop.org/show_bug.cgi?id=105707
fdo#106023 https://bugs.freedesktop.org/show_bug.cgi?id=106023
== Participating hosts (6 -> 5) ==
Missing (1): shard-glkb
== Build changes ==
* Linux: CI_DRM_4072 -> Patchwork_8768
CI_DRM_4072: b35e59e5c6a9cae11d5183d2bf9c5c99ceedbc7c @ git://anongit.freedesktop.org/gfx-ci/linux
IGT_4442: 8168bb65d5e64d4df4e5d847d448bab2d2825d73 @ git://anongit.freedesktop.org/xorg/app/intel-gpu-tools
Patchwork_8768: b3e83155f6333c0a9db115f82db0290028c52055 @ git://anongit.freedesktop.org/gfx-ci/linux
piglit_4442: e60d247eb359f044caf0c09904da14e39d7adca1 @ git://anongit.freedesktop.org/piglit
== Logs ==
For more details see: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_8768/shards.html
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx