public inbox for stable@vger.kernel.org
 help / color / mirror / Atom feed
From: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To: stable@vger.kernel.org
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
	patches@lists.linux.dev,
	Mikhail Gavrilov <mikhail.v.gavrilov@gmail.com>,
	Dmitry Torokhov <dmitry.torokhov@gmail.com>
Subject: [PATCH 6.1 05/55] Input: uinput - fix circular locking dependency with ff-core
Date: Mon, 13 Apr 2026 18:00:39 +0200	[thread overview]
Message-ID: <20260413155725.027691280@linuxfoundation.org> (raw)
In-Reply-To: <20260413155724.820472494@linuxfoundation.org>

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

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

From: Mikhail Gavrilov <mikhail.v.gavrilov@gmail.com>

commit 4cda78d6f8bf2b700529f2fbccb994c3e826d7c2 upstream.

A lockdep circular locking dependency warning can be triggered
reproducibly when using a force-feedback gamepad with uinput (for
example, playing ELDEN RING under Wine with a Flydigi Vader 5
controller):

  ff->mutex -> udev->mutex -> input_mutex -> dev->mutex -> ff->mutex

The cycle is caused by four lock acquisition paths:

1. ff upload: input_ff_upload() holds ff->mutex and calls
   uinput_dev_upload_effect() -> uinput_request_submit() ->
   uinput_request_send(), which acquires udev->mutex.

2. device create: uinput_ioctl_handler() holds udev->mutex and calls
   uinput_create_device() -> input_register_device(), which acquires
   input_mutex.

3. device register: input_register_device() holds input_mutex and
   calls kbd_connect() -> input_register_handle(), which acquires
   dev->mutex.

4. evdev release: evdev_release() calls input_flush_device() under
   dev->mutex, which calls input_ff_flush() acquiring ff->mutex.

Fix this by introducing a new state_lock spinlock to protect
udev->state and udev->dev access in uinput_request_send() instead of
acquiring udev->mutex.  The function only needs to atomically check
device state and queue an input event into the ring buffer via
uinput_dev_event() -- both operations are safe under a spinlock
(ktime_get_ts64() and wake_up_interruptible() do not sleep).  This
breaks the ff->mutex -> udev->mutex link since a spinlock is a leaf in
the lock ordering and cannot form cycles with mutexes.

To keep state transitions visible to uinput_request_send(), protect
writes to udev->state in uinput_create_device() and
uinput_destroy_device() with the same state_lock spinlock.

Additionally, move init_completion(&request->done) from
uinput_request_send() to uinput_request_submit() before
uinput_request_reserve_slot().  Once the slot is allocated,
uinput_flush_requests() may call complete() on it at any time from
the destroy path, so the completion must be initialised before the
request becomes visible.

Lock ordering after the fix:

  ff->mutex -> state_lock (spinlock, leaf)
  udev->mutex -> state_lock (spinlock, leaf)
  udev->mutex -> input_mutex -> dev->mutex -> ff->mutex (no back-edge)

Fixes: ff462551235d ("Input: uinput - switch to the new FF interface")
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/all/CABXGCsMoxag+kEwHhb7KqhuyxfmGGd0P=tHZyb1uKE0pLr8Hkg@mail.gmail.com/
Signed-off-by: Mikhail Gavrilov <mikhail.v.gavrilov@gmail.com>
Link: https://patch.msgid.link/20260407075031.38351-1-mikhail.v.gavrilov@gmail.com
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 drivers/input/misc/uinput.c |   28 +++++++++++++++++++++-------
 1 file changed, 21 insertions(+), 7 deletions(-)

--- a/drivers/input/misc/uinput.c
+++ b/drivers/input/misc/uinput.c
@@ -56,6 +56,7 @@ struct uinput_device {
 	struct input_dev	*dev;
 	struct mutex		mutex;
 	enum uinput_state	state;
+	spinlock_t		state_lock;
 	wait_queue_head_t	waitq;
 	unsigned char		ready;
 	unsigned char		head;
@@ -145,19 +146,15 @@ static void uinput_request_release_slot(
 static int uinput_request_send(struct uinput_device *udev,
 			       struct uinput_request *request)
 {
-	int retval;
+	int retval = 0;
 
-	retval = mutex_lock_interruptible(&udev->mutex);
-	if (retval)
-		return retval;
+	spin_lock(&udev->state_lock);
 
 	if (udev->state != UIST_CREATED) {
 		retval = -ENODEV;
 		goto out;
 	}
 
-	init_completion(&request->done);
-
 	/*
 	 * Tell our userspace application about this new request
 	 * by queueing an input event.
@@ -165,7 +162,7 @@ static int uinput_request_send(struct ui
 	uinput_dev_event(udev->dev, EV_UINPUT, request->code, request->id);
 
  out:
-	mutex_unlock(&udev->mutex);
+	spin_unlock(&udev->state_lock);
 	return retval;
 }
 
@@ -174,6 +171,13 @@ static int uinput_request_submit(struct
 {
 	int retval;
 
+	/*
+	 * Initialize completion before allocating the request slot.
+	 * Once the slot is allocated, uinput_flush_requests() may
+	 * complete it at any time, so it must be initialized first.
+	 */
+	init_completion(&request->done);
+
 	retval = uinput_request_reserve_slot(udev, request);
 	if (retval)
 		return retval;
@@ -288,7 +292,14 @@ static void uinput_destroy_device(struct
 	struct input_dev *dev = udev->dev;
 	enum uinput_state old_state = udev->state;
 
+	/*
+	 * Update state under state_lock so that concurrent
+	 * uinput_request_send() sees the state change before we
+	 * flush pending requests and tear down the device.
+	 */
+	spin_lock(&udev->state_lock);
 	udev->state = UIST_NEW_DEVICE;
+	spin_unlock(&udev->state_lock);
 
 	if (dev) {
 		name = dev->name;
@@ -365,7 +376,9 @@ static int uinput_create_device(struct u
 	if (error)
 		goto fail2;
 
+	spin_lock(&udev->state_lock);
 	udev->state = UIST_CREATED;
+	spin_unlock(&udev->state_lock);
 
 	return 0;
 
@@ -383,6 +396,7 @@ static int uinput_open(struct inode *ino
 		return -ENOMEM;
 
 	mutex_init(&newdev->mutex);
+	spin_lock_init(&newdev->state_lock);
 	spin_lock_init(&newdev->requests_lock);
 	init_waitqueue_head(&newdev->requests_waitq);
 	init_waitqueue_head(&newdev->waitq);



  parent reply	other threads:[~2026-04-13 16:16 UTC|newest]

Thread overview: 68+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-04-13 16:00 [PATCH 6.1 00/55] 6.1.169-rc1 review Greg Kroah-Hartman
2026-04-13 16:00 ` [PATCH 6.1 01/55] lib/crypto: chacha: Zeroize permuted_state before it leaves scope Greg Kroah-Hartman
2026-04-13 16:00 ` [PATCH 6.1 02/55] wifi: rt2x00usb: fix devres lifetime Greg Kroah-Hartman
2026-04-13 16:00 ` [PATCH 6.1 03/55] xfrm_user: fix info leak in build_report() Greg Kroah-Hartman
2026-04-13 16:00 ` [PATCH 6.1 04/55] mptcp: fix slab-use-after-free in __inet_lookup_established Greg Kroah-Hartman
2026-04-13 16:00 ` Greg Kroah-Hartman [this message]
2026-04-13 16:00 ` [PATCH 6.1 06/55] Input: uinput - take event lock when submitting FF request "event" Greg Kroah-Hartman
2026-04-13 16:00 ` [PATCH 6.1 07/55] MIPS: Always record SEGBITS in cpu_data.vmbits Greg Kroah-Hartman
2026-04-13 16:00 ` [PATCH 6.1 08/55] MIPS: mm: Suppress TLB uniquification on EHINV hardware Greg Kroah-Hartman
2026-04-13 16:00 ` [PATCH 6.1 09/55] MIPS: mm: Rewrite TLB uniquification for the hidden bit feature Greg Kroah-Hartman
2026-04-13 16:00 ` [PATCH 6.1 10/55] media: uvcvideo: Mark invalid entities with id UVC_INVALID_ENTITY_ID Greg Kroah-Hartman
2026-04-13 16:00 ` [PATCH 6.1 11/55] media: uvcvideo: Use heuristic to find stream entity Greg Kroah-Hartman
2026-04-13 16:00 ` [PATCH 6.1 12/55] apparmor: validate DFA start states are in bounds in unpack_pdb Greg Kroah-Hartman
2026-04-13 16:00 ` [PATCH 6.1 13/55] apparmor: fix memory leak in verify_header Greg Kroah-Hartman
2026-04-13 16:00 ` [PATCH 6.1 14/55] apparmor: replace recursive profile removal with iterative approach Greg Kroah-Hartman
2026-04-13 16:00 ` [PATCH 6.1 15/55] apparmor: fix: limit the number of levels of policy namespaces Greg Kroah-Hartman
2026-04-13 16:00 ` [PATCH 6.1 16/55] apparmor: fix side-effect bug in match_char() macro usage Greg Kroah-Hartman
2026-04-13 16:00 ` [PATCH 6.1 17/55] apparmor: fix missing bounds check on DEFAULT table in verify_dfa() Greg Kroah-Hartman
2026-04-13 16:00 ` [PATCH 6.1 18/55] apparmor: Fix double free of ns_name in aa_replace_profiles() Greg Kroah-Hartman
2026-04-13 16:00 ` [PATCH 6.1 19/55] apparmor: fix unprivileged local user can do privileged policy management Greg Kroah-Hartman
2026-04-13 16:00 ` [PATCH 6.1 20/55] apparmor: fix differential encoding verification Greg Kroah-Hartman
2026-04-13 16:00 ` [PATCH 6.1 21/55] apparmor: fix race on rawdata dereference Greg Kroah-Hartman
2026-04-13 16:00 ` [PATCH 6.1 22/55] apparmor: fix race between freeing data and fs accessing it Greg Kroah-Hartman
2026-04-13 16:00 ` [PATCH 6.1 23/55] usb: gadget: u_ether: Fix race between gether_disconnect and eth_stop Greg Kroah-Hartman
2026-04-13 16:00 ` [PATCH 6.1 24/55] Revert "ACPI: EC: Evaluate orphan _REG under EC device" Greg Kroah-Hartman
2026-04-13 16:00 ` [PATCH 6.1 25/55] ACPICA: Add a depth argument to acpi_execute_reg_methods() Greg Kroah-Hartman
2026-04-13 16:01 ` [PATCH 6.1 26/55] ACPI: EC: Evaluate _REG outside the EC scope more carefully Greg Kroah-Hartman
2026-04-13 16:01 ` [PATCH 6.1 27/55] usb: gadget: f_hid: move list and spinlock inits from bind to alloc Greg Kroah-Hartman
2026-04-13 16:01 ` [PATCH 6.1 28/55] rfkill: Use sysfs_emit() to instead of sprintf() Greg Kroah-Hartman
2026-04-13 16:01 ` [PATCH 6.1 29/55] rfkill: sync before userspace visibility/changes Greg Kroah-Hartman
2026-04-13 16:01 ` [PATCH 6.1 30/55] net: rfkill: reduce data->mtx scope in rfkill_fop_open Greg Kroah-Hartman
2026-04-13 16:01 ` [PATCH 6.1 31/55] net: rfkill: prevent unlimited numbers of rfkill events from being created Greg Kroah-Hartman
2026-04-13 16:01 ` [PATCH 6.1 32/55] seg6: separate dst_cache for input and output paths in seg6 lwtunnel Greg Kroah-Hartman
2026-04-13 16:01 ` [PATCH 6.1 33/55] Revert "mptcp: add needs_id for netlink appending addr" Greg Kroah-Hartman
2026-04-13 16:01 ` [PATCH 6.1 34/55] drm/scheduler: signal scheduled fence when kill job Greg Kroah-Hartman
2026-04-13 16:01 ` [PATCH 6.1 35/55] netfilter: nft_set_pipapo: do not rely on ZERO_SIZE_PTR Greg Kroah-Hartman
2026-04-13 16:01 ` [PATCH 6.1 36/55] netfilter: nft_ct: fix use-after-free in timeout object destroy Greg Kroah-Hartman
2026-04-13 16:01 ` [PATCH 6.1 37/55] xfrm: clear trailing padding in build_polexpire() Greg Kroah-Hartman
2026-04-13 16:01 ` [PATCH 6.1 38/55] tipc: fix bc_ackers underflow on duplicate GRP_ACK_MSG Greg Kroah-Hartman
2026-04-13 16:01 ` [PATCH 6.1 39/55] wifi: brcmsmac: Fix dma_free_coherent() size Greg Kroah-Hartman
2026-04-13 16:01 ` [PATCH 6.1 40/55] arm64: dts: hisilicon: poplar: Correct PCIe reset GPIO polarity Greg Kroah-Hartman
2026-04-13 16:01 ` [PATCH 6.1 41/55] arm64: dts: hisilicon: hi3798cv200: Add missing dma-ranges Greg Kroah-Hartman
2026-04-13 16:01 ` [PATCH 6.1 42/55] nfc: pn533: allocate rx skb before consuming bytes Greg Kroah-Hartman
2026-04-13 16:01 ` [PATCH 6.1 43/55] batman-adv: reject oversized global TT response buffers Greg Kroah-Hartman
2026-04-13 16:01 ` [PATCH 6.1 44/55] EDAC/mc: Fix error path ordering in edac_mc_alloc() Greg Kroah-Hartman
2026-04-13 16:01 ` [PATCH 6.1 45/55] net/tls: fix use-after-free in -EBUSY error path of tls_do_encryption Greg Kroah-Hartman
2026-04-13 16:01 ` [PATCH 6.1 46/55] net: altera-tse: fix skb leak on DMA mapping error in tse_start_xmit() Greg Kroah-Hartman
2026-04-13 16:01 ` [PATCH 6.1 47/55] batman-adv: hold claim backbone gateways by reference Greg Kroah-Hartman
2026-04-13 16:01 ` [PATCH 6.1 48/55] drm/i915/gt: fix refcount underflow in intel_engine_park_heartbeat Greg Kroah-Hartman
2026-04-13 16:01 ` [PATCH 6.1 49/55] net/mlx5: Update the list of the PCI supported devices Greg Kroah-Hartman
2026-04-13 16:01 ` [PATCH 6.1 50/55] mmc: vub300: fix NULL-deref on disconnect Greg Kroah-Hartman
2026-04-13 16:01 ` [PATCH 6.1 51/55] net: qualcomm: qca_uart: report the consumed byte on RX skb allocation failure Greg Kroah-Hartman
2026-04-13 16:01 ` [PATCH 6.1 52/55] net: stmmac: fix integer underflow in chain mode Greg Kroah-Hartman
2026-04-13 16:01 ` [PATCH 6.1 53/55] rxrpc: fix reference count leak in rxrpc_server_keyring() Greg Kroah-Hartman
2026-04-13 16:01 ` [PATCH 6.1 54/55] rxrpc: Fix key/keyring checks in setsockopt(RXRPC_SECURITY_KEY/KEYRING) Greg Kroah-Hartman
2026-04-13 16:01 ` [PATCH 6.1 55/55] Revert "PCI: Enable ACS after configuring IOMMU for OF platforms" Greg Kroah-Hartman
2026-04-13 17:38 ` [PATCH 6.1 00/55] 6.1.169-rc1 review Brett A C Sheffield
2026-04-13 18:51 ` Florian Fainelli
2026-04-14  7:53 ` Jon Hunter
2026-04-14  8:13 ` Pavel Machek
2026-04-14  8:20 ` Pavel Machek
2026-04-14  9:06 ` Peter Schneider
2026-04-14 11:49 ` Ron Economos
2026-04-14 12:30 ` Francesco Dolcini
2026-04-14 17:44 ` Shuah Khan
2026-04-14 17:44 ` Miguel Ojeda
2026-04-14 18:18 ` Mark Brown
2026-04-16 22:09 ` Barry K. Nathan

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=20260413155725.027691280@linuxfoundation.org \
    --to=gregkh@linuxfoundation.org \
    --cc=dmitry.torokhov@gmail.com \
    --cc=mikhail.v.gavrilov@gmail.com \
    --cc=patches@lists.linux.dev \
    --cc=stable@vger.kernel.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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox