[PATCH v2 1/1] drm/mm: optimize rb_hole_addr rbtree search

Das, Nirmoy nirmoy.das at amd.com
Sat May 2 11:25:49 UTC 2020


On 5/2/2020 10:00 AM, Chris Wilson wrote:
> Quoting Nirmoy (2020-04-30 11:30:43)
>> On 4/30/20 12:15 PM, Chris Wilson wrote:
>>> Quoting Nirmoy Das (2020-04-30 10:58:39)
>>>> +void insert_hole_addr(struct rb_root *root, struct drm_mm_node *node)
>>> ^ static
>>
>> Ah I forgot!
>>
>>> (sparse [make C=1] or make W=1 will spot this)
>>
>> Thanks for the tip :)
>>
>> Nirmoy
>>
>>> Looks good, I'll check more carefully later.
> Reviewed-by: Chris Wilson <chris at chris-wilson.co.uk>
>
> If you do find some time to add some more tests, that would be great!
>
> Even converting your example into a test-drm_mm benchmark [spending a
> few seconds runtime max!] will be very useful.


Thanks Chris for reviewing this. I already started looking into adding 
some fragmentation tests :)


Regards,

Nirmoy


> -Chris
> _______________________________________________
> dri-devel mailing list
> dri-devel at lists.freedesktop.org
> https://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Flists.freedesktop.org%2Fmailman%2Flistinfo%2Fdri-devel&data=02%7C01%7Cnirmoy.das%40amd.com%7Ce73eb7b87f8a4c4c180108d7ee77389c%7C3dd8961fe4884e608e11a82d994e183d%7C0%7C0%7C637240068162932922&sdata=%2FthV3slEerwVUYVwGzfiOb%2BtpboXQwYca%2BlAJ3L4Ad0%3D&reserved=0


More information about the dri-devel mailing list