AMD-GFX Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: Felix Kuehling <felix.kuehling@amd.com>
To: Shashank Sharma <shashank.sharma@amd.com>, amd-gfx@lists.freedesktop.org
Cc: Alex Deucher <alexander.deucher@amd.com>,
	mukul.joshi@amd.com, contactshashanksharma@gmail.com,
	Christian Koenig <christian.koenig@amd.com>,
	arvind.yadav@amd.com
Subject: Re: [PATCH v2 08/12] drm/amdgpu: use doorbell manager for kfd kernel doorbells
Date: Fri, 9 Jun 2023 15:33:16 -0400	[thread overview]
Message-ID: <bc736154-f727-a1f6-d83e-826122cd702e@amd.com> (raw)
In-Reply-To: <659d9452-7828-5b14-edf6-a3b307d4319e@amd.com>


On 2023-04-25 15:59, Shashank Sharma wrote:
>
> On 24/04/2023 21:56, Felix Kuehling wrote:
>> On 2023-04-22 2:39, Shashank Sharma wrote:
>>> - KFD process level doorbells: doorbell pages which are allocated by 
>>> kernel but mapped and written by userspace processes, saved in 
>>> struct pdd->qpd->doorbells
>>>
>>> size = kfd_doorbell_process_slice.
>>>
>>> We realized that we only need 1-2 doorbells for KFD kernel level 
>>> stuff (so kept it one page), but need 2-page of doorbells for KFD 
>>> process, so they are sized accordingly.
>>>
>>> We have also run kfd_test_suit and verified the changes for any 
>>> regression. Hope this helps in explaining the design.
>>
>> Right, I missed that this was only for kernel doorbells. I wonder 
>> whether KFD really needs its own page here. I think we only need a 
>> doorbell for HWS. And when we use MES, I think even that isn't needed 
>> because MES packet submissions go through amdgpu. So maybe KFD 
>> doesn't need its own kernel-mode doorbell page any more on systems 
>> with user graphics mode queues.
>
> Yeah, for any IP with MES enabled, KFD doesn't need kernel level 
> doorbells. But I still allocated a page just to make sure we do not 
> break any non-MES platforms or use cases where MES is deliberately 
> disabled from kernel command line. Hope that works for you.

Even without MES, we still only need one doorbell for HWS. Allocating a 
whole page for that is wasteful. Anyway, I'm OK with cleaning that up later.

Regards,
   Felix


>
> - Shashank
>
>>
>> Regards,
>>   Felix
>>
>>
>>>
>>> - Shashank 

  reply	other threads:[~2023-06-09 19:33 UTC|newest]

Thread overview: 28+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-04-12 16:25 [PATCH v2 00/12] AMDGPU doorbell manager Shashank Sharma
2023-04-12 16:25 ` [PATCH v2 01/12] drm/amdgpu: create a new file for " Shashank Sharma
2023-04-13 10:48   ` Christian König
2023-04-13 10:51     ` Christian König
2023-04-12 16:25 ` [PATCH v2 02/12] drm/amdgpu: don't modify num_doorbells for mes Shashank Sharma
2023-04-12 16:25 ` [PATCH v2 03/12] drm/amdgpu: add UAPI for allocating doorbell memory Shashank Sharma
2023-04-12 16:25 ` [PATCH v2 04/12] drm/amdgpu: accommodate DOMAIN/PL_DOORBELL Shashank Sharma
2023-04-12 16:25 ` [PATCH v2 05/12] drm/amdgpu: initialize ttm for doorbells Shashank Sharma
2023-04-12 16:25 ` [PATCH v2 06/12] drm/amdgpu: create kernel doorbell pages Shashank Sharma
2023-04-12 16:25 ` [PATCH v2 07/12] get absolute offset from doorbell index Shashank Sharma
2023-04-13 10:59   ` Christian König
2023-04-12 16:25 ` [PATCH v2 08/12] drm/amdgpu: use doorbell manager for kfd kernel doorbells Shashank Sharma
2023-04-13 11:02   ` Christian König
2023-04-13 13:08     ` Shashank Sharma
2023-04-21 19:58   ` Felix Kuehling
2023-04-22  6:39     ` Shashank Sharma
2023-04-24 19:56       ` Felix Kuehling
2023-04-25 19:59         ` Shashank Sharma
2023-06-09 19:33           ` Felix Kuehling [this message]
2023-04-12 16:25 ` [PATCH v2 09/12] drm/amdgpu: use doorbell manager for kfd process doorbells Shashank Sharma
2023-04-13 11:07   ` Christian König
2023-04-21 20:11   ` Felix Kuehling
2023-04-22  6:45     ` Shashank Sharma
2023-04-12 16:25 ` [PATCH v2 10/12] drm/amdgpu: remove unused functions and variables Shashank Sharma
2023-06-09 19:34   ` Felix Kuehling
2023-04-12 16:25 ` [PATCH v2 11/12] drm/amdgpu: use doorbell mgr for MES kernel doorbells Shashank Sharma
2023-04-12 16:25 ` [PATCH v2 12/12] drm/amdgpu: cleanup MES process level doorbells Shashank Sharma
2023-04-13 11:19   ` Christian König

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=bc736154-f727-a1f6-d83e-826122cd702e@amd.com \
    --to=felix.kuehling@amd.com \
    --cc=alexander.deucher@amd.com \
    --cc=amd-gfx@lists.freedesktop.org \
    --cc=arvind.yadav@amd.com \
    --cc=christian.koenig@amd.com \
    --cc=contactshashanksharma@gmail.com \
    --cc=mukul.joshi@amd.com \
    --cc=shashank.sharma@amd.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox