[Intel-gfx] [PATCH v4 7/9] drm/i915/guc: Remove i915_guc_log_unregister

[Sagar Arun Kamble]

On 9/22/2017 2:09 AM, Michal Wajdeczko wrote:
> On Thu, 21 Sep 2017 19:40:04 +0200, Sagar Arun Kamble 
> <sagar.a.kamble at intel.com> wrote:
>
>>
>>
>> On 9/21/2017 11:01 PM, Sagar Arun Kamble wrote:
>>>
>>>
>>> On 9/21/2017 2:28 AM, Michal Wajdeczko wrote:
>>>> On Wed, 20 Sep 2017 19:38:22 +0200, Sagar Arun Kamble 
>>>> <sagar.a.kamble at intel.com> wrote:
>>>>
>>>>> Functionality needed to disable GuC interrupts and cleanup the
>>>>> runtime/relay data structures is already covered in the unload path
>>>>> via intel_guc_fini_hw and intel_guc_cleanup hence remove
>>>>> i915_guc_log_unregister
>>>>>
>>>>> v2: Removed the function i915_guc_log_unregister.
>>>>>
>>>>> v3: Rebase as intel_guc.h is removed.
>>>>>
>>>>> v4: Rebase as intel_guc.h is created again. :)
>>>>>
>>>>> v5: Rebase as intel_guc.h is removed.
>>>>>
>>>>> Cc: Michal Wajdeczko <michal.wajdeczko at intel.com>
>>>>> Reviewed-by: Michal Wajdeczko <michal.wajdeczko at intel.com>
>>>>
>>>> I'm afraid I've to revoke my r-b as with removal of the 
>>>> log_unregister()
>>>> we will loose symmetry with log_register() where 
>>>> relay_late_setup_files()
>>>> was hidden, and we should still clean it up in 
>>>> i915_driver_unregister()
>>>>
>>>> Michal
>>> Agree with you. We will need to maintain symmetry. Will revisit with 
>>> separate patches.
>>> Better to integrate in the CT series I guess as there are multiple 
>>> logging related
>>> interrupt enable/disable occurrences.
>> Without this patch we are getting RPM WARN which I had floated patch 
>> initally for.
>> https://patchwork.freedesktop.org/patch/176538/
>> Kindly let me know if we should get that patch in till we fix the 
>> logging/CT interrupt interaction.
>>
>
> Can we only call relay_close() here (without any irq/runtime cleanup,
> as that will be covered in unload path) - that way we will keep
> symmetry and avoid RPM issue ...
>
> Michal
Yes. Will need to handle the sync. between unload closing the relay_chan 
and reads from relay_chan.
Will share patch for this.
>
>> [173707.659868] RPM wakelock ref not held during HW access
>> [173707.659897] ------------[ cut here ]------------
>> [173707.659969] WARNING: CPU: 1 PID: 20769 at 
>> drivers/gpu/drm/i915/intel_drv.h:1800 fwtable_write32+0x1ad/0x1d0 [i915]
>> [173707.660228]  gen6_disable_pm_irq+0x47/0x50 [i915]
>> [173707.660269]  gen9_disable_guc_interrupts+0x33/0x60 [i915]
>> [173707.660332]  i915_guc_log_unregister+0x2e/0x50 [i915]
>> [173707.660373]  i915_driver_unload+0x4d/0x180 [i915]
>> [173707.660415]  i915_pci_remove+0x19/0x30 [i915]
>> [173707.660423]  pci_device_remove+0x39/0xc0
>>
>>>
>>>
>>>>
>>>>> Signed-off-by: Sagar Arun Kamble <sagar.a.kamble at intel.com>
>>>>> ---
>>>>>  drivers/gpu/drm/i915/i915_drv.c      |  1 -
>>>>>  drivers/gpu/drm/i915/intel_guc_log.c | 12 ------------
>>>>>  drivers/gpu/drm/i915/intel_uc.h      |  1 -
>>>>>  3 files changed, 14 deletions(-)
>>>>>
>>>>> diff --git a/drivers/gpu/drm/i915/i915_drv.c 
>>>>> b/drivers/gpu/drm/i915/i915_drv.c
>>>>> index 6f36ced..c69a30a 100644
>>>>> --- a/drivers/gpu/drm/i915/i915_drv.c
>>>>> +++ b/drivers/gpu/drm/i915/i915_drv.c
>>>>> @@ -1252,7 +1252,6 @@ static void i915_driver_unregister(struct 
>>>>> drm_i915_private *dev_priv)
>>>>>      i915_perf_unregister(dev_priv);
>>>>>     i915_teardown_sysfs(dev_priv);
>>>>> -    i915_guc_log_unregister(dev_priv);
>>>>>      drm_dev_unregister(&dev_priv->drm);
>>>>>     i915_gem_shrinker_cleanup(dev_priv);
>>>>> diff --git a/drivers/gpu/drm/i915/intel_guc_log.c 
>>>>> b/drivers/gpu/drm/i915/intel_guc_log.c
>>>>> index 16d3b87..3c45681 100644
>>>>> --- a/drivers/gpu/drm/i915/intel_guc_log.c
>>>>> +++ b/drivers/gpu/drm/i915/intel_guc_log.c
>>>>> @@ -648,15 +648,3 @@ void i915_guc_log_register(struct 
>>>>> drm_i915_private *dev_priv)
>>>>>      guc_log_late_setup(&dev_priv->guc);
>>>>>      mutex_unlock(&dev_priv->drm.struct_mutex);
>>>>>  }
>>>>> -
>>>>> -void i915_guc_log_unregister(struct drm_i915_private *dev_priv)
>>>>> -{
>>>>> -    if (!i915.enable_guc_submission)
>>>>> -        return;
>>>>> -
>>>>> -    mutex_lock(&dev_priv->drm.struct_mutex);
>>>>> -    /* GuC logging is currently the only user of Guc2Host 
>>>>> interrupts */
>>>>> -    gen9_disable_guc_interrupts(dev_priv);
>>>>> -    guc_log_runtime_destroy(&dev_priv->guc);
>>>>> -    mutex_unlock(&dev_priv->drm.struct_mutex);
>>>>> -}
>>>>> diff --git a/drivers/gpu/drm/i915/intel_uc.h 
>>>>> b/drivers/gpu/drm/i915/intel_uc.h
>>>>> index 8557e33..c2c104a 100644
>>>>> --- a/drivers/gpu/drm/i915/intel_uc.h
>>>>> +++ b/drivers/gpu/drm/i915/intel_uc.h
>>>>> @@ -244,7 +244,6 @@ static inline void intel_guc_notify(struct 
>>>>> intel_guc *guc)
>>>>>  void intel_guc_log_destroy(struct intel_guc *guc);
>>>>>  int i915_guc_log_control(struct drm_i915_private *dev_priv, u64 
>>>>> control_val);
>>>>>  void i915_guc_log_register(struct drm_i915_private *dev_priv);
>>>>> -void i915_guc_log_unregister(struct drm_i915_private *dev_priv);
>>>>> static inline u32 guc_ggtt_offset(struct i915_vma *vma)
>>>>>  {
>>>