[PATCH v7 2/2] drm/i915/gvt: Use sched_lock to protect gvt scheduler logic.

[Colin Xu]

On 05/14/2018 10:29 AM, Zhenyu Wang wrote:
> On 2018.04.21 13:10:55 +0800, colin.xu at intel.com wrote:
>> From: Pei Zhang <pei.zhang at intel.com>
>>
>> The scheduler lock(gvt->sched_lock) is used to protect gvt
>> scheduler logic, including the gvt scheduler structure(gvt->scheduler
>> and per vgpu schedule data(vgpu->sched_data, vgpu->sched_ctl).
>>
>> v7:
>>    - Rebase to latest gvt-staging.
>>    - Remove gtt_lock since already proteced by gvt_lock and vgpu_lock.
>>    - Fix a typo in intel_gvt_deactivate_vgpu, unlock the wrong lock.
>> v6:
>>    - Rebase to latest gvt-staging.
>> v5:
>>    - Rebase to latest gvt-staging.
>> v4:
>>    - Rebase to latest gvt-staging.
>> v3: update to latest code base
>>
>> Signed-off-by: Pei Zhang <pei.zhang at intel.com>
>> Signed-off-by: Colin Xu <colin.xu at intel.com>
>> ---
> Reviewed-by: Zhenyu Wang <zhenyuw at linux.intel.com>
>
>> diff --git a/drivers/gpu/drm/i915/gvt/sched_policy.c b/drivers/gpu/drm/i915/gvt/sched_policy.c
>> index d053cbe1dc94..62c720be9877 100644
>> --- a/drivers/gpu/drm/i915/gvt/sched_policy.c
>> +++ b/drivers/gpu/drm/i915/gvt/sched_policy.c
>> @@ -228,7 +228,7 @@ void intel_gvt_schedule(struct intel_gvt *gvt)
>>   	struct gvt_sched_data *sched_data = gvt->scheduler.sched_data;
>>   	ktime_t cur_time;
>>   
>> -	mutex_lock(&gvt->lock);
>> +	mutex_lock(&gvt->sched_lock);
>>   	cur_time = ktime_get();
>>   
>>   	if (test_and_clear_bit(INTEL_GVT_REQUEST_SCHED,
>> @@ -241,10 +241,11 @@ void intel_gvt_schedule(struct intel_gvt *gvt)
>>   	}
>>   	clear_bit(INTEL_GVT_REQUEST_EVENT_SCHED, (void *)&gvt->service_request);
>>   
>> +
>>   	vgpu_update_timeslice(gvt->scheduler.current_vgpu, cur_time);
>>   	tbs_sched_func(sched_data);
> no extra blank line.
>
> Pls fix style issue, rebase to staging and resend for merge.
>
> thanks
>
Hi Zhenyu,

Thanks for your comments.
I've revised the code based on your suggestions and send out as v8.

-- 
Best Regards,
Colin Xu