From: Nirmoy <nirmodas@amd.com>
To: "Christian König" <christian.koenig@amd.com>,
"Nirmoy Das" <nirmoy.aiemd@gmail.com>,
amd-gfx@lists.freedesktop.org
Cc: alexander.deucher@amd.com, kenny.ho@amd.com, nirmoy.das@amd.com,
pierre-eric.pelloux-prayer@amd.com
Subject: Re: [PATCH] drm/scheduler: fix race condition in load balancer
Date: Wed, 15 Jan 2020 14:24:06 +0100 [thread overview]
Message-ID: <a23d0e09-dd32-fd18-9ea2-a5bf724bfe7f@amd.com> (raw)
In-Reply-To: <8b4d2ea2-a28d-6eb4-2d50-02b5c450922f@amd.com>
>> I think I know why it happens. At init all entity's rq gets assigned
>> to sched_list[0]. I put some prints to check what we compare in
>> drm_sched_entity_get_free_sched.
>>
>> It turns out most of the time it compares zero values(num_jobs(0) <
>> min_jobs(0)) so most of the time 1st rq(sdma0, comp_1.0.0) was picked
>> by drm_sched_entity_get_free_sched.
>
> Well that is expected because the unit tests always does
> submission,wait,submission,wait,submission,wait.... So the number of
> jobs in the scheduler becomes zero in between.
Even with multiple parallel instances of amdgpu_test, I haven't seen any
improvement in the load balance.
>
>> This patch was not correct , had an extra atomic_inc(num_jobs) in
>> drm_sched_job_init. This probably added bit of randomness I think,
>> which helped in better job distribution.
>
> Mhm, that might not be a bad idea after all. We could rename num_jobs
> into something like like score and do a +1 in
> drm_sched_rq_add_entity() and a -1 in drm_sched_rq_remove_entity().
>
> That should have pretty much the effect we want to have.
That's sounds good as well. I will create a patch.
>
>> I've updated my previous RFC patch which uses time consumed by each
>> sched for load balance with a twist of ignoring previously scheduled
>> sched/rq. Let me know what do you think.
>
> I didn't had time yet to wrap my head around that in detail, but at
> least of hand Luben is right that the locking looks really awkward.
I was unable to find a better way to do the locking part. My mail client
might've missed Luben's review, can't find it :/
Regards,
Nirmoy
_______________________________________________
amd-gfx mailing list
amd-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
prev parent reply other threads:[~2020-01-15 13:23 UTC|newest]
Thread overview: 10+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-01-14 15:43 [PATCH] drm/scheduler: fix race condition in load balancer Nirmoy Das
2020-01-14 16:01 ` Christian König
2020-01-14 16:13 ` Nirmoy
2020-01-14 16:20 ` Christian König
2020-01-14 16:20 ` Nirmoy
2020-01-14 16:23 ` Christian König
2020-01-14 16:27 ` Nirmoy
2020-01-15 11:04 ` Nirmoy
2020-01-15 12:52 ` Christian König
2020-01-15 13:24 ` Nirmoy [this message]
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=a23d0e09-dd32-fd18-9ea2-a5bf724bfe7f@amd.com \
--to=nirmodas@amd.com \
--cc=alexander.deucher@amd.com \
--cc=amd-gfx@lists.freedesktop.org \
--cc=christian.koenig@amd.com \
--cc=kenny.ho@amd.com \
--cc=nirmoy.aiemd@gmail.com \
--cc=nirmoy.das@amd.com \
--cc=pierre-eric.pelloux-prayer@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;
as well as URLs for NNTP newsgroup(s).