[PATCH 07/14] drm/i915: Add bigjoiner and uncompressed joiner hw readout sanity checks

[Nautiyal, Ankit K]

On 9/9/2024 2:01 PM, Nautiyal, Ankit K wrote:
>
> On 9/6/2024 9:09 PM, Ville Syrjälä wrote:
>> On Fri, Sep 06, 2024 at 06:28:00PM +0530, Ankit Nautiyal wrote:
>>> From: Stanislav Lisovskiy <stanislav.lisovskiy at intel.com>
>>>
>>> Adding sanity checks for primary and secondary bigjoiner/uncompressed
>>> bitmasks, should make it easier to spot possible issues.
>>>
>>> Signed-off-by: Stanislav Lisovskiy <stanislav.lisovskiy at intel.com>
>>> Reviewed-by: Suraj Kandpal <suraj.kandpal at intel.com>
>>> ---
>>>   drivers/gpu/drm/i915/display/intel_display.c | 24 
>>> ++++++++++++++++++++
>>>   1 file changed, 24 insertions(+)
>>>
>>> diff --git a/drivers/gpu/drm/i915/display/intel_display.c 
>>> b/drivers/gpu/drm/i915/display/intel_display.c
>>> index cdc7531311fc..6f098383479f 100644
>>> --- a/drivers/gpu/drm/i915/display/intel_display.c
>>> +++ b/drivers/gpu/drm/i915/display/intel_display.c
>>> @@ -3638,11 +3638,17 @@ static void enabled_bigjoiner_pipes(struct 
>>> drm_i915_private *dev_priv,
>>>       }
>>>   }
>>>   +static u8 expected_secondary_pipes(u8 primary_pipes)
>>> +{
>>> +    return primary_pipes << 1;
>>> +}
>> Extracting this should be a separate patch really since it
>> was already in the code until the previous patch erroneously
>> removed it.
>
> Sure, will move this to another patch.
>
>
>>
>>> +
>>>   static void enabled_joiner_pipes(struct drm_i915_private *dev_priv,
>>>                    u8 *primary_pipes, u8 *secondary_pipes)
>>>   {
>>>       u8 primary_uncompressed_joiner_pipes, primary_bigjoiner_pipes;
>>>       u8 secondary_uncompressed_joiner_pipes, 
>>> secondary_bigjoiner_pipes;
>>> +    u8 uncompressed_joiner_pipes, bigjoiner_pipes;
>>>         enabled_uncompressed_joiner_pipes(dev_priv, 
>>> &primary_uncompressed_joiner_pipes,
>>> &secondary_uncompressed_joiner_pipes);
>>> @@ -3650,6 +3656,24 @@ static void enabled_joiner_pipes(struct 
>>> drm_i915_private *dev_priv,
>>>       enabled_bigjoiner_pipes(dev_priv, &primary_bigjoiner_pipes,
>>>                   &secondary_bigjoiner_pipes);
>>>   +    uncompressed_joiner_pipes = primary_uncompressed_joiner_pipes |
>>> +                    secondary_uncompressed_joiner_pipes;
>>> +    bigjoiner_pipes = primary_bigjoiner_pipes | 
>>> secondary_bigjoiner_pipes;
>>> +
>>> +    drm_WARN(&dev_priv->drm, (uncompressed_joiner_pipes & 
>>> bigjoiner_pipes) != 0,
>>> +         "Uncomressed joiner pipes(%x) and bigjoiner pipes(%x) 
>>> can't intersect\n",
>>> +         uncompressed_joiner_pipes, bigjoiner_pipes);
>> Please use 0x%x for printing hex numbers
>>
>> Side note: maybe we should switch to the standard %#x instead?
>> Though we'd need to double check whether prink() follows the
>> idiotic C standard or not:
>>   printf("%#x", 1) -> 0x1, for any non-zero value
>>   bs.
>>   printf("%#x", 0) -> 0 (sigh)
>>
>> If printk() behaves the same way then I want to see %#x anywhere
>> (pretty sure we do use it in a few places already though).
>
> I will check how %#x behaves with 0 with printk, accordingly will use 
> 0x%x or %x#x, (we do use it in some places).

Just checked, with printk(), %x#x with 0, prints 0x0 unlike printf() 
which prints 0, so %x#x it is then.


>
> Regards,
>
> Ankit
>
>>
>>
>>> +    drm_WARN(&dev_priv->drm, secondary_bigjoiner_pipes !=
>>> +         expected_secondary_pipes(primary_bigjoiner_pipes),
>>> +         "Wrong secondary bigjoiner pipes(expected %x, current %x)\n",
>>> +         expected_secondary_pipes(primary_bigjoiner_pipes),
>>> +         secondary_bigjoiner_pipes);
>>> +    drm_WARN(&dev_priv->drm, secondary_uncompressed_joiner_pipes !=
>>> + expected_secondary_pipes(primary_uncompressed_joiner_pipes),
>>> +         "Wrong secondary uncompressed joiner pipes(expected %x, 
>>> current %x)\n",
>>> + expected_secondary_pipes(primary_uncompressed_joiner_pipes),
>>> +         secondary_uncompressed_joiner_pipes);
>>> +
>>>       *primary_pipes = 0;
>>>       *secondary_pipes = 0;
>>>   --
>>> 2.45.2