Re: [PATCH 2/2] drm/i915/gt: Set default CCS mode '1'

[Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com>]

On 19/02/2024 11:16, Tvrtko Ursulin wrote:
> 
> On 15/02/2024 13:59, Andi Shyti wrote:
>> Since CCS automatic load balancing is disabled, we will impose a
>> fixed balancing policy that involves setting all the CCS engines
>> to work together on the same load.
>>
>> Simultaneously, the user will see only 1 CCS rather than the
>> actual number. As of now, this change affects only DG2.
>>
>> Fixes: d2eae8e98d59 ("drm/i915/dg2: Drop force_probe requirement")
>> Signed-off-by: Andi Shyti <andi.shyti@linux.intel.com>
>> Cc: Chris Wilson <chris.p.wilson@linux.intel.com>
>> Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
>> Cc: Matt Roper <matthew.d.roper@intel.com>
>> Cc: <stable@vger.kernel.org> # v6.2+
>> ---
>>   drivers/gpu/drm/i915/gt/intel_gt.c      | 11 +++++++++++
>>   drivers/gpu/drm/i915/gt/intel_gt_regs.h |  2 ++
>>   drivers/gpu/drm/i915/i915_drv.h         | 17 +++++++++++++++++
>>   drivers/gpu/drm/i915/i915_query.c       |  5 +++--
>>   4 files changed, 33 insertions(+), 2 deletions(-)
>>
>> diff --git a/drivers/gpu/drm/i915/gt/intel_gt.c 
>> b/drivers/gpu/drm/i915/gt/intel_gt.c
>> index a425db5ed3a2..e19df4ef47f6 100644
>> --- a/drivers/gpu/drm/i915/gt/intel_gt.c
>> +++ b/drivers/gpu/drm/i915/gt/intel_gt.c
>> @@ -168,6 +168,14 @@ static void init_unused_rings(struct intel_gt *gt)
>>       }
>>   }
>> +static void intel_gt_apply_ccs_mode(struct intel_gt *gt)
>> +{
>> +    if (!IS_DG2(gt->i915))
>> +        return;
>> +
>> +    intel_uncore_write(gt->uncore, XEHP_CCS_MODE, 0);
>> +}
>> +
>>   int intel_gt_init_hw(struct intel_gt *gt)
>>   {
>>       struct drm_i915_private *i915 = gt->i915;
>> @@ -195,6 +203,9 @@ int intel_gt_init_hw(struct intel_gt *gt)
>>       intel_gt_init_swizzling(gt);
>> +    /* Configure CCS mode */
>> +    intel_gt_apply_ccs_mode(gt);
>> +
>>       /*
>>        * At least 830 can leave some of the unused rings
>>        * "active" (ie. head != tail) after resume which
>> diff --git a/drivers/gpu/drm/i915/gt/intel_gt_regs.h 
>> b/drivers/gpu/drm/i915/gt/intel_gt_regs.h
>> index cf709f6c05ae..c148113770ea 100644
>> --- a/drivers/gpu/drm/i915/gt/intel_gt_regs.h
>> +++ b/drivers/gpu/drm/i915/gt/intel_gt_regs.h
>> @@ -1605,6 +1605,8 @@
>>   #define   GEN12_VOLTAGE_MASK            REG_GENMASK(10, 0)
>>   #define   GEN12_CAGF_MASK            REG_GENMASK(19, 11)
>> +#define XEHP_CCS_MODE                          _MMIO(0x14804)
>> +
>>   #define GEN11_GT_INTR_DW(x)            _MMIO(0x190018 + ((x) * 4))
>>   #define   GEN11_CSME                (31)
>>   #define   GEN12_HECI_2                (30)
>> diff --git a/drivers/gpu/drm/i915/i915_drv.h 
>> b/drivers/gpu/drm/i915/i915_drv.h
>> index e81b3b2858ac..0853ffd3cb8d 100644
>> --- a/drivers/gpu/drm/i915/i915_drv.h
>> +++ b/drivers/gpu/drm/i915/i915_drv.h
>> @@ -396,6 +396,23 @@ static inline struct intel_gt *to_gt(const struct 
>> drm_i915_private *i915)
>>            (engine__); \
>>            (engine__) = 
>> rb_to_uabi_engine(rb_next(&(engine__)->uabi_node)))
>> +/*
>> + * Exclude unavailable engines.
>> + *
>> + * Only the first CCS engine is utilized due to the disabling of CCS 
>> auto load
>> + * balancing. As a result, all CCS engines operate collectively, 
>> functioning
>> + * essentially as a single CCS engine, hence the count of active CCS 
>> engines is
>> + * considered '1'.
>> + * Currently, this applies to platforms with more than one CCS engine,
>> + * specifically DG2.
>> + */
>> +#define for_each_available_uabi_engine(engine__, i915__) \
>> +    for_each_uabi_engine(engine__, i915__) \
>> +        if ((IS_DG2(i915__)) && \
>> +            ((engine__)->uabi_class == I915_ENGINE_CLASS_COMPUTE) && \
>> +            ((engine__)->uabi_instance)) { } \
>> +        else
>> +
> 
> If you don't want userspace to see some engines, just don't add them to 
> the uabi list in intel_engines_driver_register or thereabouts?
> 
> Similar as we do for gsc which uses I915_NO_UABI_CLASS, although for ccs 
> you can choose a different approach, whatever is more elegant.
> 
> That is also needed for i915->engine_uabi_class_count to be right, so 
> userspace stats which rely on it are correct.

I later realized it is more than that - everything that uses 
intel_engine_lookup_user to look up class instance passed in from 
userspace relies on the engine not being on the user list otherwise 
userspace could bypass the fact engine query does not list it. Like PMU, 
Perf/POA, context engine map and SSEU context query.

Regards,

Tvrtko

> 
> Regards,
> 
> Tvrtko
> 
>>   #define INTEL_INFO(i915)    ((i915)->__info)
>>   #define RUNTIME_INFO(i915)    (&(i915)->__runtime)
>>   #define DRIVER_CAPS(i915)    (&(i915)->caps)
>> diff --git a/drivers/gpu/drm/i915/i915_query.c 
>> b/drivers/gpu/drm/i915/i915_query.c
>> index fa3e937ed3f5..2d41bda626a6 100644
>> --- a/drivers/gpu/drm/i915/i915_query.c
>> +++ b/drivers/gpu/drm/i915/i915_query.c
>> @@ -124,6 +124,7 @@ static int query_geometry_subslices(struct 
>> drm_i915_private *i915,
>>       return fill_topology_info(sseu, query_item, 
>> sseu->geometry_subslice_mask);
>>   }
>> +
>>   static int
>>   query_engine_info(struct drm_i915_private *i915,
>>             struct drm_i915_query_item *query_item)
>> @@ -140,7 +141,7 @@ query_engine_info(struct drm_i915_private *i915,
>>       if (query_item->flags)
>>           return -EINVAL;
>> -    for_each_uabi_engine(engine, i915)
>> +    for_each_available_uabi_engine(engine, i915)
>>           num_uabi_engines++;
>>       len = struct_size(query_ptr, engines, num_uabi_engines);
>> @@ -155,7 +156,7 @@ query_engine_info(struct drm_i915_private *i915,
>>       info_ptr = &query_ptr->engines[0];
>> -    for_each_uabi_engine(engine, i915) {
>> +    for_each_available_uabi_engine(engine, i915) {
>>           info.engine.engine_class = engine->uabi_class;
>>           info.engine.engine_instance = engine->uabi_instance;
>>           info.flags = I915_ENGINE_INFO_HAS_LOGICAL_INSTANCE;