All of lore.kernel.org
 help / color / mirror / Atom feed
From: Dan Carpenter <dan.carpenter-QHcLZuEGTsvQT0dZR+AlfA@public.gmane.org>
To: tiffany.lin-NuS5LvNUpcJWk0Htik3J/w@public.gmane.org
Cc: linux-mediatek-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r@public.gmane.org,
	linux-media-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
Subject: [bug report] [media] vcodec: mediatek: Add Mediatek V4L2 Video Decoder Driver
Date: Wed, 9 Nov 2016 16:28:20 +0300	[thread overview]
Message-ID: <20161109132820.GA26677@mwanda> (raw)

Hello Tiffany Lin,

The patch 590577a4e525: "[media] vcodec: mediatek: Add Mediatek V4L2
Video Decoder Driver" from Sep 2, 2016, leads to the following static
checker warning:

	drivers/media/platform/mtk-vcodec/mtk_vcodec_dec.c:536 vidioc_vdec_qbuf()
	error: buffer overflow 'vq->bufs' 32 <= u32max

drivers/media/platform/mtk-vcodec/mtk_vcodec_dec.c
   520  static int vidioc_vdec_qbuf(struct file *file, void *priv,
   521                              struct v4l2_buffer *buf)
   522  {
   523          struct mtk_vcodec_ctx *ctx = fh_to_ctx(priv);
   524          struct vb2_queue *vq;
   525          struct vb2_buffer *vb;
   526          struct mtk_video_dec_buf *mtkbuf;
   527          struct vb2_v4l2_buffer  *vb2_v4l2;
   528  
   529          if (ctx->state == MTK_STATE_ABORT) {
   530                  mtk_v4l2_err("[%d] Call on QBUF after unrecoverable error",
   531                                  ctx->id);
   532                  return -EIO;
   533          }
   534  
   535          vq = v4l2_m2m_get_vq(ctx->m2m_ctx, buf->type);
   536          vb = vq->bufs[buf->index];

Smatch thinks that "buf->index" comes straight from the user without
being checked and that this is a buffer overflow.  It seems simple
enough to analyse the call tree.

__video_do_ioctl()
->  v4l_qbuf()
  -> vidioc_vdec_qbuf()

It seems like Smatch is correct.  I looked at a different implementation
of this and that one wasn't checked either so maybe there is something
I am not seeing.

This has obvious security implications.  Can someone take a look at
this?

   537          vb2_v4l2 = container_of(vb, struct vb2_v4l2_buffer, vb2_buf);
   538          mtkbuf = container_of(vb2_v4l2, struct mtk_video_dec_buf, vb);
   539  
   540          if ((buf->type == V4L2_BUF_TYPE_VIDEO_OUTPUT_MPLANE) &&
   541              (buf->m.planes[0].bytesused == 0)) {
   542                  mtkbuf->lastframe = true;
   543                  mtk_v4l2_debug(1, "[%d] (%d) id=%d lastframe=%d (%d,%d, %d) vb=%p",
   544                           ctx->id, buf->type, buf->index,
   545                           mtkbuf->lastframe, buf->bytesused,
   546                           buf->m.planes[0].bytesused, buf->length,
   547                           vb);
   548          }
   549  
   550          return v4l2_m2m_qbuf(file, ctx->m2m_ctx, buf);
   551  }

regards,
dan carpenter

WARNING: multiple messages have this Message-ID (diff)
From: Dan Carpenter <dan.carpenter@oracle.com>
To: tiffany.lin@mediatek.com
Cc: linux-media@vger.kernel.org, linux-mediatek@lists.infradead.org
Subject: [bug report] [media] vcodec: mediatek: Add Mediatek V4L2 Video Decoder Driver
Date: Wed, 9 Nov 2016 16:28:20 +0300	[thread overview]
Message-ID: <20161109132820.GA26677@mwanda> (raw)

Hello Tiffany Lin,

The patch 590577a4e525: "[media] vcodec: mediatek: Add Mediatek V4L2
Video Decoder Driver" from Sep 2, 2016, leads to the following static
checker warning:

	drivers/media/platform/mtk-vcodec/mtk_vcodec_dec.c:536 vidioc_vdec_qbuf()
	error: buffer overflow 'vq->bufs' 32 <= u32max

drivers/media/platform/mtk-vcodec/mtk_vcodec_dec.c
   520  static int vidioc_vdec_qbuf(struct file *file, void *priv,
   521                              struct v4l2_buffer *buf)
   522  {
   523          struct mtk_vcodec_ctx *ctx = fh_to_ctx(priv);
   524          struct vb2_queue *vq;
   525          struct vb2_buffer *vb;
   526          struct mtk_video_dec_buf *mtkbuf;
   527          struct vb2_v4l2_buffer  *vb2_v4l2;
   528  
   529          if (ctx->state == MTK_STATE_ABORT) {
   530                  mtk_v4l2_err("[%d] Call on QBUF after unrecoverable error",
   531                                  ctx->id);
   532                  return -EIO;
   533          }
   534  
   535          vq = v4l2_m2m_get_vq(ctx->m2m_ctx, buf->type);
   536          vb = vq->bufs[buf->index];

Smatch thinks that "buf->index" comes straight from the user without
being checked and that this is a buffer overflow.  It seems simple
enough to analyse the call tree.

__video_do_ioctl()
->  v4l_qbuf()
  -> vidioc_vdec_qbuf()

It seems like Smatch is correct.  I looked at a different implementation
of this and that one wasn't checked either so maybe there is something
I am not seeing.

This has obvious security implications.  Can someone take a look at
this?

   537          vb2_v4l2 = container_of(vb, struct vb2_v4l2_buffer, vb2_buf);
   538          mtkbuf = container_of(vb2_v4l2, struct mtk_video_dec_buf, vb);
   539  
   540          if ((buf->type == V4L2_BUF_TYPE_VIDEO_OUTPUT_MPLANE) &&
   541              (buf->m.planes[0].bytesused == 0)) {
   542                  mtkbuf->lastframe = true;
   543                  mtk_v4l2_debug(1, "[%d] (%d) id=%d lastframe=%d (%d,%d, %d) vb=%p",
   544                           ctx->id, buf->type, buf->index,
   545                           mtkbuf->lastframe, buf->bytesused,
   546                           buf->m.planes[0].bytesused, buf->length,
   547                           vb);
   548          }
   549  
   550          return v4l2_m2m_qbuf(file, ctx->m2m_ctx, buf);
   551  }

regards,
dan carpenter

             reply	other threads:[~2016-11-09 13:28 UTC|newest]

Thread overview: 6+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-11-09 13:28 Dan Carpenter [this message]
2016-11-09 13:28 ` [bug report] [media] vcodec: mediatek: Add Mediatek V4L2 Video Decoder Driver Dan Carpenter
2016-11-09 13:45 ` Hans Verkuil
2016-11-10  4:31   ` Tiffany Lin
2016-11-10  4:31     ` Tiffany Lin
2016-11-10 13:30   ` Dan Carpenter

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=20161109132820.GA26677@mwanda \
    --to=dan.carpenter-qhclzuegtsvqt0dzr+alfa@public.gmane.org \
    --cc=linux-media-u79uwXL29TY76Z2rM5mHXA@public.gmane.org \
    --cc=linux-mediatek-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r@public.gmane.org \
    --cc=tiffany.lin-NuS5LvNUpcJWk0Htik3J/w@public.gmane.org \
    /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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.