Alsa-Devel Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: Dan Carpenter <dan.carpenter@oracle.com>
To: Daniel Mack <zonque@gmail.com>
Cc: alsa-devel@alsa-project.org
Subject: [smatch stuff] reading beyond the end of the array in build_feature_ctl()
Date: Thu, 6 Oct 2011 16:39:48 +0300	[thread overview]
Message-ID: <20111006133948.GA5217@elgon.mountain> (raw)

Hi Daniel,

Smatch complains about a potential read past the end of the array in
build_feature_ctl().

sound/usb/mixer.c +1035 build_feature_ctl(34)
	error: buffer overflow 'audio_feature_info' 13 <= 14

This is in the init code so probably it's correct, and I've just
missed how it works.  Can you take a look?

The problem is that audio_feature_info[] has 13 elements, but Smatch
thinks control - 1 can go up to 14 so we're two elements past the end.

  1035          cval->val_type = audio_feature_info[control-1].type;

Smatch get the value for "control" from "i" in parse_audio_feature_unit().

sound/usb/mixer.c
  1261          } else { /* UAC_VERSION_2 */
  1262                  for (i = 0; i < 30/2; i++) {
  1263                          unsigned int ch_bits = 0;
  1264                          unsigned int ch_read_only = 0;
  1265  
  1266                          for (j = 0; j < channels; j++) {
  1267                                  unsigned int mask = snd_usb_combine_bytes(bmaControls + csize * (j+1), csize);
  1268                                  if (uac2_control_is_readable(mask, i)) {
  1269                                          ch_bits |= (1 << j);
  1270                                          if (!uac2_control_is_writeable(mask, i))
  1271                                                  ch_read_only |= (1 << j);
  1272                                  }
  1273                          }
  1274  
  1275                          /* NOTE: build_feature_ctl() will mark the control read-only if all channels
  1276                           * are marked read-only in the descriptors. Otherwise, the control will be
  1277                           * reported as writeable, but the driver will not actually issue a write
  1278                           * command for read-only channels */
  1279                          if (ch_bits & 1) /* the first channel must be set (for ease of programming) */
  1280                                  build_feature_ctl(state, _ftr, ch_bits, i, &iterm, unitid, ch_read_only);
                                                                                ^
"i" can be 13 here.

  1281                          if (uac2_control_is_readable(master_bits, i))
  1282                                  build_feature_ctl(state, _ftr, 0, i, &iterm, unitid,
                                                                          ^
"i" maybe can go up to 14 here, I'm not sure.

  1283                                                    !uac2_control_is_writeable(master_bits, i));
  1284                  }
  1285          }

regards,
dan carpenter

             reply	other threads:[~2011-10-06 13:40 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-10-06 13:39 Dan Carpenter [this message]
2011-10-13  6:14 ` [smatch stuff] reading beyond the end of the array in build_feature_ctl() Takashi Iwai

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=20111006133948.GA5217@elgon.mountain \
    --to=dan.carpenter@oracle.com \
    --cc=alsa-devel@alsa-project.org \
    --cc=zonque@gmail.com \
    /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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox