Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
From:   Rob Clark <robdclark@gmail.com>
To:     dri-devel@lists.freedesktop.org
Cc:     Daniel Vetter <daniel@ffwll.ch>,
        Rob Clark <robdclark@chromium.org>,
        Rob Clark <robdclark@gmail.com>, Sean Paul <sean@poorly.run>,
        David Airlie <airlied@linux.ie>,
        linux-arm-msm@vger.kernel.org (open list:DRM DRIVER FOR MSM ADRENO GPU),
        freedreno@lists.freedesktop.org (open list:DRM DRIVER FOR MSM ADRENO
        GPU), linux-kernel@vger.kernel.org (open list)
Subject: [PATCH v2 07/22] drm/msm: Do rpm get sooner in the submit path
Date:   Sun, 11 Oct 2020 19:09:34 -0700
Message-Id: <20201012020958.229288-8-robdclark@gmail.com>
In-Reply-To: <20201012020958.229288-1-robdclark@gmail.com>
References: <20201012020958.229288-1-robdclark@gmail.com>
MIME-Version: 1.0
Content-Transfer-Encoding: 8bit
Precedence: bulk

From: Rob Clark <robdclark@chromium.org>

Unfortunately, due to an dev_pm_opp locking interaction with
mm->mmap_sem, we need to do pm get before aquiring obj locks,
otherwise we can have anger lockdep with the chain:

  opp_table_lock --> &mm->mmap_sem --> reservation_ww_class_mutex

For an explicit fencing userspace, the impact should be minimal
as we do all the fence waits before this point.  It could result
in some needless resumes in error cases, etc.

Signed-off-by: Rob Clark <robdclark@chromium.org>
---
 drivers/gpu/drm/msm/msm_gem_submit.c | 15 +++++++++++++--
 1 file changed, 13 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/msm/msm_gem_submit.c b/drivers/gpu/drm/msm/msm_gem_submit.c
index 002130d826aa..a9422d043bfe 100644
--- a/drivers/gpu/drm/msm/msm_gem_submit.c
+++ b/drivers/gpu/drm/msm/msm_gem_submit.c
@@ -744,11 +744,20 @@ int msm_ioctl_gem_submit(struct drm_device *dev, void *data,
 
 	ret = submit_lookup_objects(submit, args, file);
 	if (ret)
-		goto out;
+		goto out_pre_pm;
 
 	ret = submit_lookup_cmds(submit, args, file);
 	if (ret)
-		goto out;
+		goto out_pre_pm;
+
+	/*
+	 * Thanks to dev_pm_opp opp_table_lock interactions with mm->mmap_sem
+	 * in the resume path, we need to to rpm get before we lock objs.
+	 * Which unfortunately might involve powering up the GPU sooner than
+	 * is necessary.  But at least in the explicit fencing case, we will
+	 * have already done all the fence waiting.
+	 */
+	pm_runtime_get_sync(&gpu->pdev->dev);
 
 	/* copy_*_user while holding a ww ticket upsets lockdep */
 	ww_acquire_init(&submit->ticket, &reservation_ww_class);
@@ -825,6 +834,8 @@ int msm_ioctl_gem_submit(struct drm_device *dev, void *data,
 
 
 out:
+	pm_runtime_put(&gpu->pdev->dev);
+out_pre_pm:
 	submit_cleanup(submit);
 	if (has_ww_ticket)
 		ww_acquire_fini(&submit->ticket);
-- 
2.26.2