From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from gabe.freedesktop.org (gabe.freedesktop.org [131.252.210.177]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id 10F84C3ABDE for ; Tue, 13 May 2025 08:16:04 +0000 (UTC) Received: from gabe.freedesktop.org (localhost [127.0.0.1]) by gabe.freedesktop.org (Postfix) with ESMTP id 7630510E54B; Tue, 13 May 2025 08:16:02 +0000 (UTC) Authentication-Results: gabe.freedesktop.org; dkim=pass (2048-bit key; secure) header.d=mailbox.org header.i=@mailbox.org header.b="q47bmBE7"; dkim-atps=neutral Received: from mout-p-202.mailbox.org (mout-p-202.mailbox.org [80.241.56.172]) by gabe.freedesktop.org (Postfix) with ESMTPS id E5C9110E181; Mon, 12 May 2025 12:50:06 +0000 (UTC) Received: from smtp1.mailbox.org (smtp1.mailbox.org [10.196.197.1]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature RSA-PSS (4096 bits) server-digest SHA256) (No client certificate requested) by mout-p-202.mailbox.org (Postfix) with ESMTPS id 4Zwzw705Bdz9sss; Mon, 12 May 2025 14:49:59 +0200 (CEST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=mailbox.org; s=mail20150812; t=1747054199; h=from:from:reply-to:reply-to:subject:subject:date:date: message-id:message-id:to:to:cc:cc:mime-version:mime-version: content-type:content-type: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=RucK/CUg7Z+XGAmZS46R/ajUuyzusOAdYiFs/Ok7uCk=; b=q47bmBE76SXgdFHB5F2/VBCE6kWNYRDxIRLen84vGzAiwZ2OJHh8zujTFpmsvRcz7xUXhZ 5PHQqHaeL5ykn5tqxpF8u5CSbE8s6u5EfSkBJa7XycDe4TsXH5qB4D5OnXZHSvZN08fyLq KcYetILN+GafwNm6SIpmLO+/MjXycwTydFUH5/mu9dUAHx8u4IhJixlJC9cFM8aAFT/7/n 16Bnr99P45+olTKRWSqC8UPRqDwpC7/CPU3nAuP+K7yqKHmQDPxi/JJcZGupt0HaW1CesM yvOfXQV5QMBaBIE+B0Xhew8umNbxXhL4r4okgXqIFQ9oKg4oq383ig4qzPLNBw== Message-ID: <657c053d7cd443ff310dfff19d03ab11e0f17289.camel@mailbox.org> Subject: Re: [RFC v4 04/16] drm/sched: Avoid double re-lock on the job free path From: Philipp Stanner To: Tvrtko Ursulin , amd-gfx@lists.freedesktop.org, dri-devel@lists.freedesktop.org Cc: kernel-dev@igalia.com, Christian =?ISO-8859-1?Q?K=F6nig?= , Danilo Krummrich , Matthew Brost , Philipp Stanner Date: Mon, 12 May 2025 14:49:55 +0200 In-Reply-To: <20250425102034.85133-5-tvrtko.ursulin@igalia.com> References: <20250425102034.85133-1-tvrtko.ursulin@igalia.com> <20250425102034.85133-5-tvrtko.ursulin@igalia.com> Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable MIME-Version: 1.0 X-MBO-RS-ID: a51891704627a892095 X-MBO-RS-META: 7tnqyyomak5yfkowpqcj38ha19666u8n X-Mailman-Approved-At: Tue, 13 May 2025 08:15:53 +0000 X-BeenThere: amd-gfx@lists.freedesktop.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: Discussion list for AMD gfx List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Reply-To: phasta@kernel.org Errors-To: amd-gfx-bounces@lists.freedesktop.org Sender: "amd-gfx" On Fri, 2025-04-25 at 11:20 +0100, Tvrtko Ursulin wrote: > Currently the job free work item will lock sched->job_list_lock first > time > to see if there are any jobs, free a single job, and then lock again > to > decide whether to re-queue itself if there are more finished jobs. >=20 > Since drm_sched_get_finished_job() already looks at the second job in > the > queue we can simply add the signaled check and have it return the > presence > of more jobs to free to the caller. That way the work item does not > have > to lock the list again and repeat the signaled check. Are you convinced that this is worth it? I'm torn. It's rare that one returns a status through a boolean by reference. Independently from that, this is a candidate which certainly can be branched out from this series, to make the series completely about the new scheduling policy, not general other improvements. P. >=20 > Signed-off-by: Tvrtko Ursulin > Cc: Christian K=C3=B6nig > Cc: Danilo Krummrich > Cc: Matthew Brost > Cc: Philipp Stanner > --- > =C2=A0drivers/gpu/drm/scheduler/sched_main.c | 39 +++++++++++------------= - > -- > =C2=A01 file changed, 16 insertions(+), 23 deletions(-) >=20 > diff --git a/drivers/gpu/drm/scheduler/sched_main.c > b/drivers/gpu/drm/scheduler/sched_main.c > index 86e40157b09b..a45b02fd2af3 100644 > --- a/drivers/gpu/drm/scheduler/sched_main.c > +++ b/drivers/gpu/drm/scheduler/sched_main.c > @@ -365,22 +365,6 @@ static void __drm_sched_run_free_queue(struct > drm_gpu_scheduler *sched) > =C2=A0 queue_work(sched->submit_wq, &sched->work_free_job); > =C2=A0} > =C2=A0 > -/** > - * drm_sched_run_free_queue - enqueue free-job work if ready > - * @sched: scheduler instance > - */ > -static void drm_sched_run_free_queue(struct drm_gpu_scheduler > *sched) > -{ > - struct drm_sched_job *job; > - > - spin_lock(&sched->job_list_lock); > - job =3D list_first_entry_or_null(&sched->pending_list, > - =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 struct drm_sched_job, list); > - if (job && dma_fence_is_signaled(&job->s_fence->finished)) > - __drm_sched_run_free_queue(sched); > - spin_unlock(&sched->job_list_lock); > -} > - > =C2=A0/** > =C2=A0 * drm_sched_job_done - complete a job > =C2=A0 * @s_job: pointer to the job which is done > @@ -1097,12 +1081,13 @@ drm_sched_select_entity(struct > drm_gpu_scheduler *sched) > =C2=A0 * drm_sched_get_finished_job - fetch the next finished job to be > destroyed > =C2=A0 * > =C2=A0 * @sched: scheduler instance > + * @have_more: are there more finished jobs on the list > =C2=A0 * > =C2=A0 * Returns the next finished job from the pending list (if there is > one) > =C2=A0 * ready for it to be destroyed. > =C2=A0 */ > =C2=A0static struct drm_sched_job * > -drm_sched_get_finished_job(struct drm_gpu_scheduler *sched) > +drm_sched_get_finished_job(struct drm_gpu_scheduler *sched, bool > *have_more) > =C2=A0{ > =C2=A0 struct drm_sched_job *job, *next; > =C2=A0 > @@ -1110,22 +1095,27 @@ drm_sched_get_finished_job(struct > drm_gpu_scheduler *sched) > =C2=A0 > =C2=A0 job =3D list_first_entry_or_null(&sched->pending_list, > =C2=A0 =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 struct drm_sched_job, list= ); > - > =C2=A0 if (job && dma_fence_is_signaled(&job->s_fence->finished)) { > =C2=A0 /* remove job from pending_list */ > =C2=A0 list_del_init(&job->list); > =C2=A0 > =C2=A0 /* cancel this job's TO timer */ > =C2=A0 cancel_delayed_work(&sched->work_tdr); > - /* make the scheduled timestamp more accurate */ > + > + *have_more =3D false; > =C2=A0 next =3D list_first_entry_or_null(&sched- > >pending_list, > =C2=A0 typeof(*next), > list); > - > =C2=A0 if (next) { > + /* make the scheduled timestamp more > accurate */ > =C2=A0 if (test_bit(DMA_FENCE_FLAG_TIMESTAMP_BIT, > =C2=A0 =C2=A0=C2=A0=C2=A0=C2=A0 &next->s_fence- > >scheduled.flags)) > =C2=A0 next->s_fence->scheduled.timestamp =3D > =C2=A0 dma_fence_timestamp(&job- > >s_fence->finished); > + > + if (test_bit(DMA_FENCE_FLAG_SIGNALED_BIT, > + =C2=A0=C2=A0=C2=A0=C2=A0 &next->s_fence- > >finished.flags)) > + *have_more =3D true; > + > =C2=A0 /* start TO timer for next job */ > =C2=A0 drm_sched_start_timeout(sched); > =C2=A0 } > @@ -1184,12 +1174,15 @@ static void drm_sched_free_job_work(struct > work_struct *w) > =C2=A0 struct drm_gpu_scheduler *sched =3D > =C2=A0 container_of(w, struct drm_gpu_scheduler, > work_free_job); > =C2=A0 struct drm_sched_job *job; > + bool have_more; > =C2=A0 > - job =3D drm_sched_get_finished_job(sched); > - if (job) > + job =3D drm_sched_get_finished_job(sched, &have_more); > + if (job) { > =C2=A0 sched->ops->free_job(job); > + if (have_more) > + __drm_sched_run_free_queue(sched); > + } > =C2=A0 > - drm_sched_run_free_queue(sched); > =C2=A0 drm_sched_run_job_queue(sched); > =C2=A0} > =C2=A0