stable.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, JongHwan Kim <zzoru007@gmail.com>,
	Dmitry Vyukov <dvyukov@google.com>,
	Andrei Vagin <avagin@gmail.com>,
	"Eric W. Biederman" <ebiederm@xmission.com>
Subject: [PATCH 4.10 18/48] ucount: Remove the atomicity from ucount->count
Date: Thu, 16 Mar 2017 23:30:02 +0900	[thread overview]
Message-ID: <20170316142921.699574681@linuxfoundation.org> (raw)
In-Reply-To: <20170316142920.761502205@linuxfoundation.org>

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

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

From: Eric W. Biederman <ebiederm@xmission.com>

commit 040757f738e13caaa9c5078bca79aa97e11dde88 upstream.

Always increment/decrement ucount->count under the ucounts_lock.  The
increments are there already and moving the decrements there means the
locking logic of the code is simpler.  This simplification in the
locking logic fixes a race between put_ucounts and get_ucounts that
could result in a use-after-free because the count could go zero then
be found by get_ucounts and then be freed by put_ucounts.

A bug presumably this one was found by a combination of syzkaller and
KASAN.  JongWhan Kim reported the syzkaller failure and Dmitry Vyukov
spotted the race in the code.

Fixes: f6b2db1a3e8d ("userns: Make the count of user namespaces per user")
Reported-by: JongHwan Kim <zzoru007@gmail.com>
Reported-by: Dmitry Vyukov <dvyukov@google.com>
Reviewed-by: Andrei Vagin <avagin@gmail.com>
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

---
 include/linux/user_namespace.h |    2 +-
 kernel/ucount.c                |   18 +++++++++++-------
 2 files changed, 12 insertions(+), 8 deletions(-)

--- a/include/linux/user_namespace.h
+++ b/include/linux/user_namespace.h
@@ -65,7 +65,7 @@ struct ucounts {
 	struct hlist_node node;
 	struct user_namespace *ns;
 	kuid_t uid;
-	atomic_t count;
+	int count;
 	atomic_t ucount[UCOUNT_COUNTS];
 };
 
--- a/kernel/ucount.c
+++ b/kernel/ucount.c
@@ -139,7 +139,7 @@ static struct ucounts *get_ucounts(struc
 
 		new->ns = ns;
 		new->uid = uid;
-		atomic_set(&new->count, 0);
+		new->count = 0;
 
 		spin_lock_irq(&ucounts_lock);
 		ucounts = find_ucounts(ns, uid, hashent);
@@ -150,8 +150,10 @@ static struct ucounts *get_ucounts(struc
 			ucounts = new;
 		}
 	}
-	if (!atomic_add_unless(&ucounts->count, 1, INT_MAX))
+	if (ucounts->count == INT_MAX)
 		ucounts = NULL;
+	else
+		ucounts->count += 1;
 	spin_unlock_irq(&ucounts_lock);
 	return ucounts;
 }
@@ -160,13 +162,15 @@ static void put_ucounts(struct ucounts *
 {
 	unsigned long flags;
 
-	if (atomic_dec_and_test(&ucounts->count)) {
-		spin_lock_irqsave(&ucounts_lock, flags);
+	spin_lock_irqsave(&ucounts_lock, flags);
+	ucounts->count -= 1;
+	if (!ucounts->count)
 		hlist_del_init(&ucounts->node);
-		spin_unlock_irqrestore(&ucounts_lock, flags);
+	else
+		ucounts = NULL;
+	spin_unlock_irqrestore(&ucounts_lock, flags);
 
-		kfree(ucounts);
-	}
+	kfree(ucounts);
 }
 
 static inline bool atomic_inc_below(atomic_t *v, int u)

  parent reply	other threads:[~2017-03-16 14:36 UTC|newest]

Thread overview: 51+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-03-16 14:29 [PATCH 4.10 00/48] 4.10.4-stable review Greg Kroah-Hartman
2017-03-16 14:29 ` [PATCH 4.10 01/48] iio: 104-quad-8: Fix off-by-one error when addressing flag register Greg Kroah-Hartman
2017-03-16 14:29 ` [PATCH 4.10 02/48] ARM: qcom_defconfig: Enable RPM/RPM-SMD clocks Greg Kroah-Hartman
2017-03-16 14:29 ` [PATCH 4.10 03/48] USB: serial: digi_acceleport: fix OOB data sanity check Greg Kroah-Hartman
2017-03-16 14:29 ` [PATCH 4.10 04/48] USB: serial: digi_acceleport: fix OOB-event processing Greg Kroah-Hartman
2017-03-16 14:29 ` [PATCH 4.10 05/48] crypto: improve gcc optimization flags for serpent and wp512 Greg Kroah-Hartman
2017-03-16 14:29 ` [PATCH 4.10 06/48] MIPS: Update defconfigs for NF_CT_PROTO_DCCP/UDPLITE change Greg Kroah-Hartman
2017-03-16 14:29 ` [PATCH 4.10 07/48] MIPS: VDSO: avoid duplicate CAC_BASE definition Greg Kroah-Hartman
2017-03-16 14:29 ` [PATCH 4.10 08/48] MIPS: ip27: Disable qlge driver in defconfig Greg Kroah-Hartman
2017-03-16 14:29 ` [PATCH 4.10 09/48] MIPS: Update ip27_defconfig for SCSI_DH change Greg Kroah-Hartman
2017-03-16 14:29 ` [PATCH 4.10 10/48] MIPS: ip22: Fix ip28 build for modern gcc Greg Kroah-Hartman
2017-03-16 14:29 ` [PATCH 4.10 11/48] MIPS: Update lemote2f_defconfig for CPU_FREQ_STAT change Greg Kroah-Hartman
2017-03-16 14:29 ` [PATCH 4.10 12/48] mtd: pmcmsp: use kstrndup instead of kmalloc+strncpy Greg Kroah-Hartman
2017-03-16 14:29 ` [PATCH 4.10 13/48] MIPS: ralink: Cosmetic change to prom_init() Greg Kroah-Hartman
2017-03-16 14:29 ` [PATCH 4.10 14/48] MIPS: ralink: Remove unused timer functions Greg Kroah-Hartman
2017-03-16 14:29 ` [PATCH 4.10 15/48] MIPS: ralink: Remove unused rt*_wdt_reset functions Greg Kroah-Hartman
2017-03-16 14:30 ` [PATCH 4.10 17/48] tracing: Add #undef to fix compile error Greg Kroah-Hartman
2017-03-16 14:30 ` Greg Kroah-Hartman [this message]
2017-03-16 14:30 ` [PATCH 4.10 19/48] efi/arm: Fix boot crash with CONFIG_CPUMASK_OFFSTACK=y Greg Kroah-Hartman
2017-03-16 14:30 ` [PATCH 4.10 20/48] [media] dw2102: dont do DMA on stack Greg Kroah-Hartman
2017-03-16 14:30 ` [PATCH 4.10 21/48] i2c: add missing of_node_put in i2c_mux_del_adapters Greg Kroah-Hartman
2017-03-16 14:30 ` [PATCH 4.10 22/48] powerpc: Emulation support for load/store instructions on LE Greg Kroah-Hartman
2017-03-16 14:30 ` [PATCH 4.10 23/48] powerpc/booke: Fix boot crash due to null hugepd Greg Kroah-Hartman
2017-03-16 14:30 ` [PATCH 4.10 24/48] powerpc/xics: Work around limitations of OPAL XICS priority handling Greg Kroah-Hartman
2017-03-16 14:30 ` [PATCH 4.10 25/48] PCI: Prevent VPD access for QLogic ISP2722 Greg Kroah-Hartman
2017-03-16 14:30 ` [PATCH 4.10 26/48] usb: gadget: dummy_hcd: clear usb_gadget region before registration Greg Kroah-Hartman
2017-03-16 14:30 ` [PATCH 4.10 27/48] usb: dwc3: gadget: make Set Endpoint Configuration macros safe Greg Kroah-Hartman
2017-03-16 14:30 ` [PATCH 4.10 28/48] usb: dwc3-omap: Fix missing break in dwc3_omap_set_mailbox() Greg Kroah-Hartman
2017-03-16 14:30 ` [PATCH 4.10 29/48] usb: ohci-at91: Do not drop unhandled USB suspend control requests Greg Kroah-Hartman
2017-03-16 14:30 ` [PATCH 4.10 30/48] usb: gadget: function: f_fs: pass companion descriptor along Greg Kroah-Hartman
2017-03-16 14:30 ` [PATCH 4.10 31/48] Revert "usb: gadget: uvc: Add missing call for additional setup data" Greg Kroah-Hartman
2017-03-16 14:30 ` [PATCH 4.10 32/48] usb: host: xhci-dbg: HCIVERSION should be a binary number Greg Kroah-Hartman
2017-03-16 14:30 ` [PATCH 4.10 33/48] usb: host: xhci-plat: Fix timeout on removal of hot pluggable xhci controllers Greg Kroah-Hartman
2017-03-16 14:30 ` [PATCH 4.10 34/48] USB: serial: safe_serial: fix information leak in completion handler Greg Kroah-Hartman
2017-03-16 14:30 ` [PATCH 4.10 35/48] USB: serial: omninet: fix reference leaks at open Greg Kroah-Hartman
2017-03-16 14:30 ` [PATCH 4.10 36/48] USB: iowarrior: fix NULL-deref at probe Greg Kroah-Hartman
2017-03-16 14:30 ` [PATCH 4.10 37/48] USB: iowarrior: fix NULL-deref in write Greg Kroah-Hartman
2017-03-16 14:30 ` [PATCH 4.10 38/48] USB: serial: io_ti: fix NULL-deref in interrupt callback Greg Kroah-Hartman
2017-03-16 14:30 ` [PATCH 4.10 39/48] USB: serial: io_ti: fix information leak in completion handler Greg Kroah-Hartman
2017-03-16 14:30 ` [PATCH 4.10 41/48] KVM: s390: Fix guest migration for huge guests resulting in panic Greg Kroah-Hartman
2017-03-16 14:30 ` [PATCH 4.10 42/48] KVM: arm/arm64: Let vcpu thread modify its own active state Greg Kroah-Hartman
2017-03-16 14:30 ` [PATCH 4.10 43/48] drm/i915/gvt: Fix superfluous newline in GVT_DISPLAY_READY env var Greg Kroah-Hartman
2017-03-16 14:30 ` [PATCH 4.10 44/48] [media] serial_ir: ensure were ready to receive interrupts Greg Kroah-Hartman
2017-03-16 14:30 ` [PATCH 4.10 45/48] dm: flush queued bios when process blocks to avoid deadlock Greg Kroah-Hartman
2017-03-16 14:30 ` [PATCH 4.10 46/48] [media] rc: raw decoder for keymap protocol is not loaded on register Greg Kroah-Hartman
2017-03-16 14:30 ` [PATCH 4.10 47/48] ext4: dont BUG when truncating encrypted inodes on the orphan list Greg Kroah-Hartman
2017-03-16 14:30 ` [PATCH 4.10 48/48] IB/mlx5: Verify that Q counters are supported Greg Kroah-Hartman
2017-03-16 19:21 ` [PATCH 4.10 00/48] 4.10.4-stable review Shuah Khan
2017-03-17  1:42   ` Greg Kroah-Hartman
2017-03-16 22:37 ` Guenter Roeck
2017-03-17  1:42   ` Greg Kroah-Hartman

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=20170316142921.699574681@linuxfoundation.org \
    --to=gregkh@linuxfoundation.org \
    --cc=avagin@gmail.com \
    --cc=dvyukov@google.com \
    --cc=ebiederm@xmission.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=stable@vger.kernel.org \
    --cc=zzoru007@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;
as well as URLs for NNTP newsgroup(s).