linux-kernel - [PATCH 1/2] media: vb2: Keep dma-buf buffers mapped until they are freed

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-next>] [day] [month] [year] [list]

Date:   Thu, 24 Jan 2019 10:51:55 +0100
From:   Paul Kocialkowski <paul.kocialkowski@...tlin.com>
To:     linux-media@...r.kernel.org, linux-kernel@...r.kernel.org,
        devel@...verdev.osuosl.org, linux-arm-kernel@...ts.infradead.org,
        linux-sunxi@...glegroups.com
Cc:     Pawel Osciak <pawel@...iak.com>,
        Mauro Carvalho Chehab <mchehab@...nel.org>,
        Maxime Ripard <maxime.ripard@...tlin.com>,
        Paul Kocialkowski <paul.kocialkowski@...tlin.com>,
        Thomas Petazzoni <thomas.petazzoni@...tlin.com>,
        Hans Verkuil <hverkuil@...all.nl>,
        Ezequiel Garcia <ezequiel@...labora.com>,
        Tomasz Figa <tfiga@...omium.org>,
        Laurent Pinchart <laurent.pinchart@...asonboard.com>,
        Sakari Ailus <sakari.ailus@...ux.intel.com>,
        Pawel Osciak <posciak@...omium.org>
Subject: [PATCH 1/2] media: vb2: Keep dma-buf buffers mapped until they are freed

From: Pawel Osciak <posciak@...omium.org>

When using vb2 for video decoding, dequeued capture buffers may still
be accessed by the hardware: this is the case when they are used as
reference frames for decoding subsequent frames.

When the buffer is imported with dma-buf, it needs to be mapped before
access. Until now, it was mapped when queuing and unmapped when
dequeuing, which doesn't work for access as a reference frames.

One way to solve this would be to map the buffer again when it is
needed as a reference, but the mapping/unmapping operations can
seriously impact performance. As a result, map the buffer once (when it
is first needed when queued) and keep it mapped until it is freed.

Signed-off-by: Pawel Osciak <posciak@...omium.org>
Reviewed-on: https://chromium-review.googlesource.com/334103
Signed-off-by: Paul Kocialkowski <paul.kocialkowski@...tlin.com>
[Paul: Updated for mainline and changed commit message]
---
 drivers/media/common/videobuf2/videobuf2-core.c | 11 +++--------
 1 file changed, 3 insertions(+), 8 deletions(-)

diff --git a/drivers/media/common/videobuf2/videobuf2-core.c b/drivers/media/common/videobuf2/videobuf2-core.c
index 70e8c3366f9c..ce9294a635cc 100644
--- a/drivers/media/common/videobuf2/videobuf2-core.c
+++ b/drivers/media/common/videobuf2/videobuf2-core.c
@@ -1196,6 +1196,9 @@ static int __prepare_dmabuf(struct vb2_buffer *vb)
 	 * userspace knows sooner rather than later if the dma-buf map fails.
 	 */
 	for (plane = 0; plane < vb->num_planes; ++plane) {
+		if (vb->planes[plane].dbuf_mapped)
+			continue;
+
 		ret = call_memop(vb, map_dmabuf, vb->planes[plane].mem_priv);
 		if (ret) {
 			dprintk(1, "failed to map dmabuf for plane %d\n",
@@ -1758,14 +1761,6 @@ static void __vb2_dqbuf(struct vb2_buffer *vb)
 
 	vb->state = VB2_BUF_STATE_DEQUEUED;
 
-	/* unmap DMABUF buffer */
-	if (q->memory == VB2_MEMORY_DMABUF)
-		for (i = 0; i < vb->num_planes; ++i) {
-			if (!vb->planes[i].dbuf_mapped)
-				continue;
-			call_void_memop(vb, unmap_dmabuf, vb->planes[i].mem_priv);
-			vb->planes[i].dbuf_mapped = 0;
-		}
 	call_void_bufop(q, init_buffer, vb);
 }
 
-- 
2.20.1