Re: [PATCH v2] media: intel/ipu6: optimize the IPU6 MMU mapping and unmapping flow

[Bingbu Cao <bingbu.cao@linux.intel.com>]

Sakari,

Thank you for the review.

On 10/21/24 4:36 PM, Sakari Ailus wrote:
> Hi Bingbu,
> 
> On Fri, Aug 16, 2024 at 11:31:21AM +0800, bingbu.cao@intel.com wrote:
>> From: Bingbu Cao <bingbu.cao@intel.com>
>>
>> ipu6_mmu_map() and ipu6_mmu_unmap() operated on a per-page basis,
>> leading to frequent calls to spin_locks/unlocks and
>> clflush_cache_range for each page. This will cause inefficiencies,
>> especially when handling large dma-bufs with hundreds of pages.
>>
>> This change enhances ipu6_mmu_map()/ipu6_mmu_unmap() with batching
>> process multiple contiguous pages. This significantly reduces calls
>> for spin_lock/unlock and clflush_cache_range() and improve the
>> performance.
>>
>> Signed-off-by: Jianhui Dai <jianhui.j.dai@intel.com>
>> Signed-off-by: Bingbu Cao <bingbu.cao@intel.com>
> 
> Thanks for the patch.
> 
> Could you split this into three patches (at least) to make it more
> reviewable:
> 
> - Move l2_unmap() up to its new location.
> - Add unmapping optimisation.
> - Add mapping optimisation.
> 

Yes, I will split this.

-- 
Best regards,
Bingbu Cao