linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To: linux-kernel@vger.kernel.org
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
	stable@vger.kernel.org, Takashi Iwai <tiwai@suse.de>,
	Mauro Carvalho Chehab <mchehab@s-opensource.com>,
	Amit Pundir <amit.pundir@linaro.org>
Subject: [PATCH 3.18 48/59] xc2028: Fix use-after-free bug properly
Date: Tue, 23 May 2017 22:10:16 +0200	[thread overview]
Message-ID: <20170523200853.175607354@linuxfoundation.org> (raw)
In-Reply-To: <20170523200849.241966497@linuxfoundation.org>

3.18-stable review patch.  If anyone has any objections, please let me know.

------------------

From: Takashi Iwai <tiwai@suse.de>

commit 22a1e7783e173ab3d86018eb590107d68df46c11 upstream.

The commit 8dfbcc4351a0 ("[media] xc2028: avoid use after free") tried
to address the reported use-after-free by clearing the reference.

However, it's clearing the wrong pointer; it sets NULL to
priv->ctrl.fname, but it's anyway overwritten by the next line
memcpy(&priv->ctrl, p, sizeof(priv->ctrl)).

OTOH, the actual code accessing the freed string is the strcmp() call
with priv->fname:
	if (!firmware_name[0] && p->fname &&
	    priv->fname && strcmp(p->fname, priv->fname))
		free_firmware(priv);

where priv->fname points to the previous file name, and this was
already freed by kfree().

For fixing the bug properly, this patch does the following:

- Keep the copy of firmware file name in only priv->fname,
  priv->ctrl.fname isn't changed;
- The allocation is done only when the firmware gets loaded;
- The kfree() is called in free_firmware() commonly

Fixes: commit 8dfbcc4351a0 ('[media] xc2028: avoid use after free')
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
Signed-off-by: Amit Pundir <amit.pundir@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

---
 drivers/media/tuners/tuner-xc2028.c |   37 +++++++++++++++---------------------
 1 file changed, 16 insertions(+), 21 deletions(-)

--- a/drivers/media/tuners/tuner-xc2028.c
+++ b/drivers/media/tuners/tuner-xc2028.c
@@ -281,6 +281,14 @@ static void free_firmware(struct xc2028_
 	int i;
 	tuner_dbg("%s called\n", __func__);
 
+	/* free allocated f/w string */
+	if (priv->fname != firmware_name)
+		kfree(priv->fname);
+	priv->fname = NULL;
+
+	priv->state = XC2028_NO_FIRMWARE;
+	memset(&priv->cur_fw, 0, sizeof(priv->cur_fw));
+
 	if (!priv->firm)
 		return;
 
@@ -291,9 +299,6 @@ static void free_firmware(struct xc2028_
 
 	priv->firm = NULL;
 	priv->firm_size = 0;
-	priv->state = XC2028_NO_FIRMWARE;
-
-	memset(&priv->cur_fw, 0, sizeof(priv->cur_fw));
 }
 
 static int load_all_firmwares(struct dvb_frontend *fe,
@@ -884,9 +889,8 @@ read_not_reliable:
 	return 0;
 
 fail:
-	priv->state = XC2028_NO_FIRMWARE;
+	free_firmware(priv);
 
-	memset(&priv->cur_fw, 0, sizeof(priv->cur_fw));
 	if (retry_count < 8) {
 		msleep(50);
 		retry_count++;
@@ -1332,11 +1336,8 @@ static int xc2028_dvb_release(struct dvb
 	mutex_lock(&xc2028_list_mutex);
 
 	/* only perform final cleanup if this is the last instance */
-	if (hybrid_tuner_report_instance_count(priv) == 1) {
+	if (hybrid_tuner_report_instance_count(priv) == 1)
 		free_firmware(priv);
-		kfree(priv->ctrl.fname);
-		priv->ctrl.fname = NULL;
-	}
 
 	if (priv)
 		hybrid_tuner_release_state(priv);
@@ -1399,19 +1400,8 @@ static int xc2028_set_config(struct dvb_
 
 	/*
 	 * Copy the config data.
-	 * For the firmware name, keep a local copy of the string,
-	 * in order to avoid troubles during device release.
 	 */
-	kfree(priv->ctrl.fname);
-	priv->ctrl.fname = NULL;
 	memcpy(&priv->ctrl, p, sizeof(priv->ctrl));
-	if (p->fname) {
-		priv->ctrl.fname = kstrdup(p->fname, GFP_KERNEL);
-		if (priv->ctrl.fname == NULL) {
-			rc = -ENOMEM;
-			goto unlock;
-		}
-	}
 
 	/*
 	 * If firmware name changed, frees firmware. As free_firmware will
@@ -1426,10 +1416,15 @@ static int xc2028_set_config(struct dvb_
 
 	if (priv->state == XC2028_NO_FIRMWARE) {
 		if (!firmware_name[0])
-			priv->fname = priv->ctrl.fname;
+			priv->fname = kstrdup(p->fname, GFP_KERNEL);
 		else
 			priv->fname = firmware_name;
 
+		if (!priv->fname) {
+			rc = -ENOMEM;
+			goto unlock;
+		}
+
 		rc = request_firmware_nowait(THIS_MODULE, 1,
 					     priv->fname,
 					     priv->i2c_props.adap->dev.parent,

  parent reply	other threads:[~2017-05-23 21:02 UTC|newest]

Thread overview: 59+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-05-23 20:09 [PATCH 3.18 00/59] 3.18.55-stable review Greg Kroah-Hartman
2017-05-23 20:09 ` [PATCH 3.18 01/59] USB: ene_usb6250: fix DMA to the stack Greg Kroah-Hartman
2017-05-23 20:09 ` [PATCH 3.18 02/59] watchdog: pcwd_usb: fix NULL-deref at probe Greg Kroah-Hartman
2017-05-23 20:09 ` [PATCH 3.18 03/59] char: lp: fix possible integer overflow in lp_setup() Greg Kroah-Hartman
2017-05-23 20:09 ` [PATCH 3.18 04/59] USB: core: replace %p with %pK Greg Kroah-Hartman
2017-05-23 20:09 ` [PATCH 3.18 05/59] dm btree: fix for dm_btree_find_lowest_key() Greg Kroah-Hartman
2017-05-23 20:09 ` [PATCH 3.18 06/59] dm bufio: avoid a possible ABBA deadlock Greg Kroah-Hartman
2017-05-23 20:09 ` [PATCH 3.18 07/59] dm thin metadata: call precommit before saving the roots Greg Kroah-Hartman
2017-05-23 20:09 ` [PATCH 3.18 08/59] dm space map disk: fix some book keeping in the disk space map Greg Kroah-Hartman
2017-05-23 20:09 ` [PATCH 3.18 09/59] mwifiex: pcie: fix cmd_buf use-after-free in remove/reset Greg Kroah-Hartman
2017-05-23 20:09 ` [PATCH 3.18 11/59] regulator: tps65023: Fix inverted core enable logic Greg Kroah-Hartman
2017-05-23 20:09 ` [PATCH 3.18 12/59] ath9k_htc: fix NULL-deref at probe Greg Kroah-Hartman
2017-05-23 20:09 ` [PATCH 3.18 13/59] cdc-acm: fix possible invalid access when processing notification Greg Kroah-Hartman
2017-05-23 20:09 ` [PATCH 3.18 14/59] of: fix sparse warning in of_pci_range_parser_one Greg Kroah-Hartman
2017-05-23 20:09 ` [PATCH 3.18 15/59] of: fdt: add missing allocation-failure check Greg Kroah-Hartman
2017-05-23 20:09 ` [PATCH 3.18 16/59] iio: dac: ad7303: fix channel description Greg Kroah-Hartman
2017-05-23 20:09 ` [PATCH 3.18 17/59] pid_ns: Sleep in TASK_INTERRUPTIBLE in zap_pid_ns_processes Greg Kroah-Hartman
2017-05-23 20:09 ` [PATCH 3.18 18/59] USB: serial: ftdi_sio: fix setting latency for unprivileged users Greg Kroah-Hartman
2017-05-23 20:09 ` [PATCH 3.18 19/59] USB: serial: ftdi_sio: add Olimex ARM-USB-TINY(H) PIDs Greg Kroah-Hartman
2017-05-23 20:09 ` [PATCH 3.18 20/59] usb: host: xhci-plat: propagate return value of platform_get_irq() Greg Kroah-Hartman
2017-05-23 20:09 ` [PATCH 3.18 21/59] usb: host: xhci-mem: allocate zeroed Scratchpad Buffer Greg Kroah-Hartman
2017-05-23 20:09 ` [PATCH 3.18 22/59] net: irda: irda-usb: fix firmware name on big-endian hosts Greg Kroah-Hartman
2017-05-23 20:09 ` [PATCH 3.18 23/59] [media] usbvision: fix NULL-deref at probe Greg Kroah-Hartman
2017-05-23 20:09 ` [PATCH 3.18 24/59] [media] mceusb: " Greg Kroah-Hartman
2017-05-23 20:09 ` [PATCH 3.18 25/59] [media] ttusb2: limit messages to buffer size Greg Kroah-Hartman
2017-05-23 20:09 ` [PATCH 3.18 26/59] usb: musb: tusb6010_omap: Do not reset the other directions packet size Greg Kroah-Hartman
2017-05-23 20:09 ` [PATCH 3.18 27/59] USB: iowarrior: fix info ioctl on big-endian hosts Greg Kroah-Hartman
2017-05-23 20:09 ` [PATCH 3.18 28/59] usb: serial: option: add Telit ME910 support Greg Kroah-Hartman
2017-05-23 20:09 ` [PATCH 3.18 30/59] USB: serial: mct_u232: fix big-endian baud-rate handling Greg Kroah-Hartman
2017-05-23 20:09 ` [PATCH 3.18 31/59] USB: serial: io_ti: fix div-by-zero in set_termios Greg Kroah-Hartman
2017-05-23 20:10 ` [PATCH 3.18 32/59] USB: hub: fix SS hub-descriptor handling Greg Kroah-Hartman
2017-05-23 20:10 ` [PATCH 3.18 33/59] USB: hub: fix non-SS " Greg Kroah-Hartman
2017-05-23 20:10 ` [PATCH 3.18 34/59] tty: Prevent ldisc drivers from re-using stale tty fields Greg Kroah-Hartman
2017-05-24 13:44   ` Alan Cox
2017-05-24 15:03     ` Greg Kroah-Hartman
2017-05-23 20:10 ` [PATCH 3.18 36/59] iio: proximity: as3935: fix as3935_write Greg Kroah-Hartman
2017-05-23 20:10 ` [PATCH 3.18 37/59] [media] gspca: konica: add missing endpoint sanity check Greg Kroah-Hartman
2017-05-23 20:10 ` [PATCH 3.18 38/59] [media] s5p-mfc: Fix unbalanced call to clock management Greg Kroah-Hartman
2017-05-23 20:10 ` [PATCH 3.18 39/59] [media] dib0700: fix NULL-deref at probe Greg Kroah-Hartman
2017-05-23 20:10 ` [PATCH 3.18 40/59] [media] zr364xx: enforce minimum size when reading header Greg Kroah-Hartman
2017-05-23 20:10 ` [PATCH 3.18 41/59] [media] cx231xx-cards: fix NULL-deref at probe Greg Kroah-Hartman
2017-05-23 20:10 ` [PATCH 3.18 42/59] [media] cx231xx-audio: " Greg Kroah-Hartman
2017-05-23 20:10 ` [PATCH 3.18 43/59] powerpc/pseries: Fix of_node_put() underflow during DLPAR remove Greg Kroah-Hartman
2017-05-23 20:10 ` [PATCH 3.18 44/59] ARM: dts: at91: sama5d3_xplained: fix ADC vref Greg Kroah-Hartman
2017-05-23 20:10 ` [PATCH 3.18 45/59] ARM: dts: at91: sama5d3_xplained: not all ADC channels are available Greg Kroah-Hartman
2017-05-23 20:10 ` [PATCH 3.18 46/59] arm64: uaccess: ensure extension of access_ok() addr Greg Kroah-Hartman
2017-05-23 20:10 ` [PATCH 3.18 47/59] arm64: documentation: document tagged pointer stack constraints Greg Kroah-Hartman
2017-05-23 20:10 ` Greg Kroah-Hartman [this message]
2017-05-23 20:10 ` [PATCH 3.18 49/59] mm/huge_memory.c: respect FOLL_FORCE/FOLL_COW for thp Greg Kroah-Hartman
2017-05-23 20:10 ` [PATCH 3.18 50/59] metag/uaccess: Fix access_ok() Greg Kroah-Hartman
2017-05-23 20:10 ` [PATCH 3.18 51/59] metag/uaccess: Check access_ok in strncpy_from_user Greg Kroah-Hartman
2017-05-23 20:10 ` [PATCH 3.18 52/59] stackprotector: Increase the per-task stack canarys random range from 32 bits to 64 bits on 64-bit platforms Greg Kroah-Hartman
2017-05-23 20:10 ` [PATCH 3.18 53/59] uwb: fix device quirk on big-endian hosts Greg Kroah-Hartman
2017-05-23 20:10 ` [PATCH 3.18 54/59] osf_wait4(): fix infoleak Greg Kroah-Hartman
2017-05-23 20:10 ` [PATCH 3.18 55/59] tracing/kprobes: Enforce kprobes teardown after testing Greg Kroah-Hartman
2017-05-23 20:10 ` [PATCH 3.18 56/59] PCI: Fix pci_mmap_fits() for HAVE_PCI_RESOURCE_TO_USER platforms Greg Kroah-Hartman
2017-05-23 20:10 ` [PATCH 3.18 58/59] drivers: char: mem: Check for address space wraparound with mmap() Greg Kroah-Hartman
2017-05-23 20:10 ` [PATCH 3.18 59/59] usb: misc: legousbtower: Fix memory leak Greg Kroah-Hartman
2017-05-24 15:45 ` [PATCH 3.18 00/59] 3.18.55-stable review Guenter Roeck

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=20170523200853.175607354@linuxfoundation.org \
    --to=gregkh@linuxfoundation.org \
    --cc=amit.pundir@linaro.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mchehab@s-opensource.com \
    --cc=stable@vger.kernel.org \
    --cc=tiwai@suse.de \
    /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;
as well as URLs for NNTP newsgroup(s).