[Intel-gfx] [PATCH 6/8] drm/i915: add free_unused_pages support to i915

[Arunpravin]

On 04/11/21 12:48 am, Matthew Auld wrote:
> On 25/10/2021 14:00, Arunpravin wrote:
>> add drm_buddy_free_unused_pages() support on
>> contiguous allocation
>>
>> Signed-off-by: Arunpravin <Arunpravin.PaneerSelvam at amd.com>
>> ---
>>   drivers/gpu/drm/i915/i915_ttm_buddy_manager.c | 8 ++++++++
>>   1 file changed, 8 insertions(+)
>>
>> diff --git a/drivers/gpu/drm/i915/i915_ttm_buddy_manager.c b/drivers/gpu/drm/i915/i915_ttm_buddy_manager.c
>> index 963468228392..162947af8e04 100644
>> --- a/drivers/gpu/drm/i915/i915_ttm_buddy_manager.c
>> +++ b/drivers/gpu/drm/i915/i915_ttm_buddy_manager.c
>> @@ -98,6 +98,14 @@ static int i915_ttm_buddy_man_alloc(struct ttm_resource_manager *man,
>>   	if (unlikely(err))
>>   		goto err_free_blocks;
>>   
>> +	if (place->flags & TTM_PL_FLAG_CONTIGUOUS) {
>> +		err = drm_buddy_free_unused_pages(mm, (uint64_t)n_pages << PAGE_SHIFT,
>> +						       &bman_res->blocks);
>> +
>> +		if (unlikely(err))
>> +			goto err_free_blocks;
> 
> That needs some locking, mark_free/mark_split are all globally visible. 
> Some concurrent user might steal the block, or worse.
[Arun] I will lock using mutex
> 
>> +	}
>> +
>>   	*res = &bman_res->base;
>>   	return 0;
>>   
>>