linux-kernel - [RFC 12/17] drm/amdgpu/dc: Stop dma_resv_lock inversion in commit

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-Id: <20200512085944.222637-13-daniel.vetter@ffwll.ch>
Date:   Tue, 12 May 2020 10:59:39 +0200
From:   Daniel Vetter <daniel.vetter@...ll.ch>
To:     DRI Development <dri-devel@...ts.freedesktop.org>
Cc:     LKML <linux-kernel@...r.kernel.org>,
        Daniel Vetter <daniel.vetter@...ll.ch>,
        linux-media@...r.kernel.org, linaro-mm-sig@...ts.linaro.org,
        linux-rdma@...r.kernel.org, amd-gfx@...ts.freedesktop.org,
        intel-gfx@...ts.freedesktop.org,
        Chris Wilson <chris@...is-wilson.co.uk>,
        Maarten Lankhorst <maarten.lankhorst@...ux.intel.com>,
        Christian König <christian.koenig@....com>,
        Daniel Vetter <daniel.vetter@...el.com>
Subject: [RFC 12/17] drm/amdgpu/dc: Stop dma_resv_lock inversion in commit_tail

Trying to grab dma_resv_lock while in commit_tail before we've done
all the code that leads to the eventual signalling of the vblank event
(which can be a dma_fence) is deadlock-y. Don't do that.

Here the solution is easy because just grabbing locks to read
something races anyway. We don't need to bother, READ_ONCE is
equivalent. And avoids the locking issue.

Cc: linux-media@...r.kernel.org
Cc: linaro-mm-sig@...ts.linaro.org
Cc: linux-rdma@...r.kernel.org
Cc: amd-gfx@...ts.freedesktop.org
Cc: intel-gfx@...ts.freedesktop.org
Cc: Chris Wilson <chris@...is-wilson.co.uk>
Cc: Maarten Lankhorst <maarten.lankhorst@...ux.intel.com>
Cc: Christian König <christian.koenig@....com>
Signed-off-by: Daniel Vetter <daniel.vetter@...el.com>
---
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 10 ++++++++++
 1 file changed, 10 insertions(+)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
index 9bfaa4cad483..28e1af9f823c 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -6699,7 +6699,11 @@ static void amdgpu_dm_commit_planes(struct drm_atomic_state *state,
 		 * explicitly on fences instead
 		 * and in general should be called for
 		 * blocking commit to as per framework helpers
+		 *
+		 * Yes, this deadlocks, since you're calling dma_resv_lock in a
+		 * path that leads to a dma_fence_signal(). Don't do that.
 		 */
+#if 0
 		r = amdgpu_bo_reserve(abo, true);
 		if (unlikely(r != 0))
 			DRM_ERROR("failed to reserve buffer before flip\n");
@@ -6709,6 +6713,12 @@ static void amdgpu_dm_commit_planes(struct drm_atomic_state *state,
 		tmz_surface = amdgpu_bo_encrypted(abo);
 
 		amdgpu_bo_unreserve(abo);
+#endif
+		/*
+		 * this races anyway, so READ_ONCE isn't any better or worse
+		 * than the stuff above. Except the stuff above can deadlock.
+		 */
+		tiling_flags = READ_ONCE(abo->tiling_flags);
 
 		fill_dc_plane_info_and_addr(
 			dm->adev, new_plane_state, tiling_flags,
-- 
2.26.2