Skip to contentSkip to conversation

7.0.12-rc1 review

PATCH In review Greg Kroah-Hartman 19 participants 361 messages Jun 7 – Oct 11 Page 2 of 3

Subsystems (4)
lore
Share & tools
lore
author maintainer
#140 lore raw

PATCH 126/332 usb: typec: wcove: dont write past struct pd_message in wcove_read_rx_buffer()

In reply to Greg Kroah-Hartman

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

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

From: Greg Kroah-Hartman <[email protected]>

commit 4af7ad0e6d7aa4403dbb1dac7b9659b0421efcaa upstream.

wcove_read_rx_buffer() copies the PD RX FIFO into the caller's struct pd_message with

	for (i = 0; i < USBC_RXINFO_RXBYTES(info); i++)
		regmap_read(wcove->regmap, USBC_RX_DATA + i, msg + i);

which has two problems:

USBC_RXINFO_RXBYTES() is a 5-bit field (max 31) while struct pd_message is 30 bytes (__le16 header + __le32 payload[PD_MAX_PAYLOAD], packed). The byte count latched in RXINFO is the number of bytes the port partner put on the wire, so a malicious partner that transmits a 31-byte frame can drive the loop one byte past the destination if the WCOVE BMC receiver does not enforce the PD object-count limit in hardware. The existing FIXME flagged this as unverified.

Independently, regmap_read() takes an unsigned int * and stores a full unsigned int at the destination. Passing the byte pointer msg + i means each iteration writes four bytes; the high three are zero (val_bits is
8) and are normally overwritten by the next iteration, but the final iteration's high bytes are not. With RXBYTES == 30 the i == 29 iteration already writes three zero bytes past msg, which sits on the IRQ thread's stack in wcove_typec_irq().

Clamp the loop to sizeof(struct pd_message) and read each register into a local before storing only its low byte, so the copy can never exceed the destination regardless of what RXINFO reports.

---
 drivers/usb/typec/tcpm/wcove.c |   13 ++++++++-----
 1 file changed, 8 insertions(+), 5 deletions(-)
drivers/usb/typec/tcpm/wcove.c +8 -5
@@ -444,9 +444,11 @@ static int wcove_start_toggling(struct t
 	return regmap_write(wcove->regmap, USBC_CONTROL1, usbc_ctrl);
 }
 
-static int wcove_read_rx_buffer(struct wcove_typec *wcove, void *msg)
+static int wcove_read_rx_buffer(struct wcove_typec *wcove,
+				struct pd_message *msg)
 {
-	unsigned int info;
+	unsigned int info, val, len;
+	u8 *buf = (u8 *)msg;
 	int ret;
 	int i;
 
@@ -454,12 +456,13 @@ static int wcove_read_rx_buffer(struct w
 	if (ret)
 		return ret;
 
-	/* FIXME: Check that USBC_RXINFO_RXBYTES(info) matches the header */
+	len = min(USBC_RXINFO_RXBYTES(info), sizeof(*msg));
 
-	for (i = 0; i < USBC_RXINFO_RXBYTES(info); i++) {
-		ret = regmap_read(wcove->regmap, USBC_RX_DATA + i, msg + i);
+	for (i = 0; i < len; i++) {
+		ret = regmap_read(wcove->regmap, USBC_RX_DATA + i, &val);
 		if (ret)
 			return ret;
+		buf[i] = val;
 	}
 
 	return regmap_write(wcove->regmap, USBC_RXSTATUS,
author maintainer
#141 lore raw

PATCH 127/332 usb: typec: tcpm/tcpci_maxim: validate header NDO against RX_BYTE_CNT

In reply to Greg Kroah-Hartman

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

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

From: Greg Kroah-Hartman <[email protected]>

commit aa2f716327be1818e1cb156da8a2844804aaec2f upstream.

A broken/malicious port can transmit a CRC-valid frame whose header advertises up to seven data objects but whose body carries fewer than that. Check for this, and rightfully reject the message, instead of reading from uninitialized stack memory.

Assisted-by gkh_clanker_t1000
Cc Heikki Krogerus <[email protected]>
Cc "André Draszik" <[email protected]>
Cc Badhri Jagan Sridharan <[email protected]>
Cc Amit Sunil Dhamne <[email protected]>
Cc stable <[email protected]>
---
 drivers/usb/typec/tcpm/tcpci_maxim_core.c |    9 +++++++++
 1 file changed, 9 insertions(+)
drivers/usb/typec/tcpm/tcpci_maxim_core.c +9 -0
@@ -186,6 +186,15 @@ static void process_rx(struct max_tcpci_
 	rx_buf_ptr = rx_buf + TCPC_RECEIVE_BUFFER_RX_BYTE_BUF_OFFSET;
 	msg.header = cpu_to_le16(*(u16 *)rx_buf_ptr);
 	rx_buf_ptr = rx_buf_ptr + sizeof(msg.header);
+
+	if (count < TCPC_RECEIVE_BUFFER_RX_BYTE_BUF_OFFSET + sizeof(msg.header) +
+		    pd_header_cnt_le(msg.header) * sizeof(msg.payload[0])) {
+		max_tcpci_write16(chip, TCPC_ALERT, TCPC_ALERT_RX_STATUS);
+		dev_err(chip->dev, "Invalid TCPC_RX_BYTE_CNT %d for header cnt %d\n",
+			count, pd_header_cnt_le(msg.header));
+		return;
+	}
+
 	for (payload_index = 0; payload_index < pd_header_cnt_le(msg.header); payload_index++,
 	     rx_buf_ptr += sizeof(msg.payload[0]))
 		msg.payload[payload_index] = cpu_to_le32(*(u32 *)rx_buf_ptr);
author maintainer
#142 lore raw

PATCH 128/332 usb: typec: ucsi: validate connector number in ucsi_connector_change()

In reply to Greg Kroah-Hartman

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

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

From: Greg Kroah-Hartman <[email protected]>

commit 288a81a8507052bcfbf884d39a463c44c42c5fd9 upstream.

The connector number in a UCSI CCI notification is a 7-bit field supplied by the PPM. ucsi_connector_change() uses it to index the ucsi->connector[] array without checking it against the number of connectors the PPM reported at init time, so a buggy or malicious PPM (EC firmware, or an I2C-attached UCSI controller on the ccg / stm32g0 / glink transports) can drive schedule_work() on memory past the end of the array.

Reject connector numbers that are zero or exceed cap.num_connectors before dereferencing the array.

---
 drivers/usb/typec/ucsi/ucsi.c |   11 ++++++++++-
 1 file changed, 10 insertions(+), 1 deletion(-)
drivers/usb/typec/ucsi/ucsi.c +10 -1
@@ -1370,13 +1370,22 @@ out_unlock:
  */
 void ucsi_connector_change(struct ucsi *ucsi, u8 num)
 {
-	struct ucsi_connector *con = &ucsi->connector[num - 1];
+	struct ucsi_connector *con;
 
 	if (!(ucsi->ntfy & UCSI_ENABLE_NTFY_CONNECTOR_CHANGE)) {
 		dev_dbg(ucsi->dev, "Early connector change event\n");
 		return;
 	}
 
+	if (!num || num > ucsi->cap.num_connectors) {
+		dev_warn_ratelimited(ucsi->dev,
+				     "Bogus connector change on %u (max %u)\n",
+				     num, ucsi->cap.num_connectors);
+		return;
+	}
+
+	con = &ucsi->connector[num - 1];
+
 	if (!test_and_set_bit(EVENT_PENDING, &ucsi->flags))
 		schedule_work(&con->work);
 }
author maintainer
#143 lore raw

PATCH 129/332 USB: serial: safe_serial: fix memory corruption with small endpoint

In reply to Greg Kroah-Hartman

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

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

From: Johan Hovold <[email protected]>

commit 438061ed1ad85e6743e2dce826671772d81089ec upstream.

Make sure that the bulk-out buffer size is at least eight bytes to avoid user-controlled slab corruption in "safe" mode should a malicious device report a smaller size.

---
 drivers/usb/serial/safe_serial.c |   11 +++++++++++
 1 file changed, 11 insertions(+)
drivers/usb/serial/safe_serial.c +11 -0
@@ -259,6 +259,7 @@ static int safe_prepare_write_buffer(str
 static int safe_startup(struct usb_serial *serial)
 {
 	struct usb_interface_descriptor	*desc;
+	int bulk_out_size;
 
 	if (serial->dev->descriptor.bDeviceClass != CDC_DEVICE_CLASS)
 		return -ENODEV;
@@ -279,6 +280,16 @@ static int safe_startup(struct usb_seria
 	default:
 		return -EINVAL;
 	}
+
+	/*
+	 * The bulk-out buffer needs to be large enough for the two-byte
+	 * trailer in safe mode, but assume anything smaller than eight bytes
+	 * is broken.
+	 */
+	bulk_out_size = serial->port[0]->bulk_out_size;
+	if (bulk_out_size > 0 && bulk_out_size < 8)
+		return -EINVAL;
+
 	return 0;
 }
 
author maintainer
#144 lore raw

PATCH 130/332 media: rc: igorplugusb: fix control request setup packet

In reply to Greg Kroah-Hartman

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

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

From: Henri A <[email protected]>

commit 171022c7d594c133a45f92357a2a91475edabe20 upstream.

Commit eac69475b01f ("media: rc: igorplugusb: heed coherency rules") changed the control request storage from an embedded struct to an allocated pointer so it can obey DMA coherency rules.

However, the driver still passes &ir->request to usb_fill_control_urb(). That points the URB setup packet at the pointer field itself rather than at the allocated struct usb_ctrlrequest.

USB core then interprets pointer bytes as the setup packet. This can produce an invalid bRequestType and trigger the control direction warning reported by syzbot:

usb 2-1: BOGUS control dir, pipe 80003580 doesn't match bRequestType 0

Pass ir->request itself as the setup packet.

Fixes eac69475b01f ("media: rc: igorplugusb: heed coherency rules")
Reported-by [email protected]
Assisted-by Codex:GPT-5.5
---
 drivers/media/rc/igorplugusb.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/media/rc/igorplugusb.c +1 -1
@@ -184,7 +184,7 @@ static int igorplugusb_probe(struct usb_
 	if (!ir->buf_in)
 		goto fail;
 	usb_fill_control_urb(ir->urb, udev,
-		usb_rcvctrlpipe(udev, 0), (uint8_t *)&ir->request,
+		usb_rcvctrlpipe(udev, 0), (uint8_t *)ir->request,
 		ir->buf_in, MAX_PACKET, igorplugusb_callback, ir);
 
 	usb_make_path(udev, ir->phys, sizeof(ir->phys));
author maintainer
#145 lore raw

PATCH 131/332 Input: ims-pcu - fix usb_free_coherent() size in ims_pcu_buffers_free()

In reply to Greg Kroah-Hartman

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

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

From: Thomas Fourier <[email protected]>

commit dab48a7e74e6a394f3aa0461a2b1fb0c7b38fcb8 upstream.

The input buffer size is pcu->max_in_size, but pcu->max_out_size is passed to usb_free_coherent().

Change size to match the allocation size.

---
 drivers/input/misc/ims-pcu.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/input/misc/ims-pcu.c +1 -1
@@ -1604,7 +1604,7 @@ static void ims_pcu_buffers_free(struct
 	usb_kill_urb(pcu->urb_in);
 	usb_free_urb(pcu->urb_in);
 
-	usb_free_coherent(pcu->udev, pcu->max_out_size,
+	usb_free_coherent(pcu->udev, pcu->max_in_size,
 			  pcu->urb_in_buf, pcu->read_dma);
 
 	kfree(pcu->urb_out_buf);
author maintainer
#146 lore raw

PATCH 132/332 USB: serial: cypress_m8: fix memory corruption with small endpoint

In reply to Greg Kroah-Hartman

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

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

From: Johan Hovold <[email protected]>

commit e1a9d791fd66ab2431b9e6f6f835823809869047 upstream.

Make sure that the interrupt-out endpoint max packet size is at least eight bytes to avoid user-controlled slab corruption or NULL-pointer dereference should a malicious device report a smaller size.

Fixes 3416eaa1f8f8 ("USB: cypress_m8: Packet format is separate from characteristic size")
Cc [email protected] # 2.6.26
---
 drivers/usb/serial/cypress_m8.c |    8 ++++++++
 1 file changed, 8 insertions(+)
drivers/usb/serial/cypress_m8.c +8 -0
@@ -445,6 +445,14 @@ static int cypress_generic_port_probe(st
 		return -ENODEV;
 	}
 
+	/*
+	 * The buffer must be large enough for the one or two-byte header (and
+	 * following data), but assume anything smaller than eight bytes is
+	 * broken.
+	 */
+	if (port->interrupt_out_size < 8)
+		return -EINVAL;
+
 	priv = kzalloc_obj(struct cypress_private);
 	if (!priv)
 		return -ENOMEM;
author maintainer
#147 lore raw

PATCH 133/332 HID: quirks: Add ALWAYS_POLL quirk for SIGMACHIP USB mouse

In reply to Greg Kroah-Hartman

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

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

From: hlleng <[email protected]>

commit 07466fc91c55532edcfb5c6a7ccd2ea52728d6bd upstream.

The SIGMACHIP USB mouse with VID/PID 1c4f:0034 can disconnect and re-enumerate repeatedly after it has been enumerated if its interrupt endpoint is not continuously polled.

This was observed with the device reporting itself as "SIGMACHIP Usb Mouse". Keeping the input event device open avoids the disconnects.

Add HID_QUIRK_ALWAYS_POLL for this device so the HID core keeps polling it even when there is no userspace input consumer.

---
 drivers/hid/hid-ids.h    |    1 +
 drivers/hid/hid-quirks.c |    1 +
 2 files changed, 2 insertions(+)
drivers/hid/hid-ids.h +1 -0
@@ -1261,6 +1261,7 @@
 
 #define USB_VENDOR_ID_SIGMA_MICRO	0x1c4f
 #define USB_DEVICE_ID_SIGMA_MICRO_KEYBOARD	0x0002
+#define USB_DEVICE_ID_SIGMA_MICRO_USB_MOUSE	0x0034
 #define USB_DEVICE_ID_SIGMA_MICRO_KEYBOARD2	0x0059
 
 #define USB_VENDOR_ID_SIGMATEL		0x066F
drivers/hid/hid-quirks.c +1 -0
@@ -186,6 +186,7 @@ static const struct hid_device_id hid_qu
 	{ HID_USB_DEVICE(USB_VENDOR_ID_SEMICO, USB_DEVICE_ID_SEMICO_USB_KEYKOARD), HID_QUIRK_NO_INIT_REPORTS },
 	{ HID_USB_DEVICE(USB_VENDOR_ID_SENNHEISER, USB_DEVICE_ID_SENNHEISER_BTD500USB), HID_QUIRK_NOGET },
 	{ HID_USB_DEVICE(USB_VENDOR_ID_SIGMA_MICRO, USB_DEVICE_ID_SIGMA_MICRO_KEYBOARD), HID_QUIRK_NO_INIT_REPORTS },
+	{ HID_USB_DEVICE(USB_VENDOR_ID_SIGMA_MICRO, USB_DEVICE_ID_SIGMA_MICRO_USB_MOUSE), HID_QUIRK_ALWAYS_POLL },
 	{ HID_USB_DEVICE(USB_VENDOR_ID_SIGMATEL, USB_DEVICE_ID_SIGMATEL_STMP3780), HID_QUIRK_NOGET },
 	{ HID_USB_DEVICE(USB_VENDOR_ID_SIS_TOUCH, USB_DEVICE_ID_SIS1030_TOUCH), HID_QUIRK_NOGET },
 	{ HID_USB_DEVICE(USB_VENDOR_ID_SIS_TOUCH, USB_DEVICE_ID_SIS817_TOUCH), HID_QUIRK_NOGET },
author maintainer
#148 lore raw

PATCH 134/332 Bluetooth: btusb: Allow firmware re-download when version matches

In reply to Greg Kroah-Hartman

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

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

From: Shuai Zhang <[email protected]>

commit 82855073c1081732656734b74d7d1d5e4cfd0da7 upstream.

The Bluetooth host decides whether to download firmware by reading the controller firmware download completion flag and firmware version information.

If a USB error occurs during the firmware download process (for example due to a USB disconnect), the download is aborted immediately. An incomplete firmware transfer does not cause the controller to set the download completion flag, but the firmware version information may be updated at an early stage of the download process.

In this case, after USB reconnection, the host attempts to re-download the firmware because the download completion flag is not set. However, since the controller reports the same firmware version as the target firmware, the download is skipped. This ultimately results in the firmware not being properly updated on the controller.

This change removes the restriction that skips firmware download when the versions are equal. It covers scenarios where the USB connection can be disconnected at any time and ensures that firmware download can be retriggered after USB reconnection, allowing the Bluetooth firmware to be correctly and completely updated.

Fixes 3267c884cefa ("Bluetooth: btusb: Add support for QCA ROME chipset family")
---
 drivers/bluetooth/btusb.c |    8 +++++++-
 1 file changed, 7 insertions(+), 1 deletion(-)
drivers/bluetooth/btusb.c +7 -1
@@ -3511,7 +3511,13 @@ static int btusb_setup_qca_load_rampatch
 		    "firmware rome 0x%x build 0x%x",
 		    rver_rom, rver_patch, ver_rom, ver_patch);
 
-	if (rver_rom != ver_rom || rver_patch <= ver_patch) {
+	/* Allow rampatch when the patch version equals the firmware version.
+	 * A firmware download may be aborted by a transient USB error (e.g.
+	 * disconnect) after the controller updates version info but before
+	 * completion.
+	 * Allowing equal versions enables re-flashing during recovery.
+	 */
+	if (rver_rom != ver_rom || rver_patch < ver_patch) {
 		bt_dev_err(hdev, "rampatch file version did not match with firmware");
 		err = -EINVAL;
 		goto done;
author maintainer
#149 lore raw

PATCH 135/332 mm/vmalloc: do not trigger BUG() on BH disabled context

In reply to Greg Kroah-Hartman

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

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

From: Uladzislau Rezki (Sony) <[email protected]>

commit 04aa71da5f35aacdc9ae9cb5150947daa624f641 upstream.

__get_vm_area_node() currently triggers a BUG() if in_interrupt() returns true. However, in_interrupt() also reports true when BH are disabled.

The bridge code can call rhashtable_lookup_insert_fast() with bottom halves disabled:

__vlan_add()
-> br_fdb_add_local()
spin_lock_bh(&br->hash_lock); <-- Disable BH
-> fdb_add_local()

    -> fdb_create()
     -> rhashtable_lookup_insert_fast()
      -> kvmalloc()
       -> vmalloc()
        -> __get_vm_area_node()
         -> BUG_ON(in_interrupt())

spin_unlock_bh(&br->hash_lock)

this triggers the BUG() despite the caller not being in NMI or hard IRQ context.

Replace the in_interrupt() check with in_nmi() || in_hardirq().

---
 mm/vmalloc.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
mm/vmalloc.c +1 -1
@@ -3209,7 +3209,7 @@ struct vm_struct *__get_vm_area_node(uns
 	struct vm_struct *area;
 	unsigned long requested_size = size;
 
-	BUG_ON(in_interrupt());
+	BUG_ON(in_nmi() || in_hardirq());
 	size = ALIGN(size, 1ul << shift);
 	if (unlikely(!size))
 		return NULL;
author maintainer
#150 lore raw

PATCH 136/332 hpfs: fix a crash if hpfs_map_dnode_bitmap fails

In reply to Greg Kroah-Hartman

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

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

From: Mikulas Patocka <[email protected]>

commit 974820a59efde7c1a7e1260bcfe9bb81f833cc9f upstream.

If hpfs_map_dnode_bitmap fails, the code would call hpfs_brelse4 on uninitialized quad buffer head, causing a crash.

---
 fs/hpfs/alloc.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
fs/hpfs/alloc.c +1 -1
@@ -372,8 +372,8 @@ int hpfs_check_free_dnodes(struct super_
 				return 0;
 			}
 		}
+		hpfs_brelse4(&qbh);
 	}
-	hpfs_brelse4(&qbh);
 	i = 0;
 	if (hpfs_sb(s)->sb_c_bitmap != -1) {
 		bmp = hpfs_map_bitmap(s, b, &qbh, "chkdn1");
author maintainer
#151 lore raw

PATCH 137/332 mm/damon/sysfs-schemes: delete tried region in regions_rmdirs()

In reply to Greg Kroah-Hartman

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

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

From: SeongJae Park <[email protected]>

commit 441f92f7d386b85bad16de49db95a307cba048a2 upstream.

DAMON sysfs maintains the DAMOS tried region directory objects via a linked list. When the user requests refresh of the directories, DAMON sysfs removes all the region directories first, and then generate updated regions directory on the empty space. The removal function (damon_sysfs_scheme_regions_rm_dirs()) only puts the kobj objects. Deletion of the container region object from the linked list is done inside the kobj release callback function.

If somehow the callback invocation is delayed, the list will contain regions list that gonna be freed. If the updated region directories creation is started in this situation, the list can be corrupted and use-after-free can happen.

Because the kobj objects are managed by only DAMON sysfs, the issue cannot happen in normal situation. But, such delays can be made on kernels that built with CONFIG_DEBUG_KOBJECT_RELEASE. On the kernel, the issue can indeed be reproduced like below.

    # damo start --damos_action stat
    # cd /sys/kernel/mm/damon/admin/kdamonds/0/
    # for i in {1..10}; do echo update_schemes_tried_regions > state; done
    # dmesg | grep underflow
    [   89.296152] refcount_t: underflow; use-after-free.

Fix the issue by removing the region object from the list when decrementing the reference count.

Also update damos_sysfs_populate_region_dir() to add the region object to the list only after the kobject_init_and_add() is success, so that fail of kobject_init_and_add() is not leaving the deallocated object on the list.

The issue was discovered [1] by Sashiko.

---
 mm/damon/sysfs-schemes.c |    8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)
mm/damon/sysfs-schemes.c +4 -4
@@ -88,7 +88,6 @@ static void damon_sysfs_scheme_region_re
 	struct damon_sysfs_scheme_region *region = container_of(kobj,
 			struct damon_sysfs_scheme_region, kobj);
 
-	list_del(&region->list);
 	kfree(region);
 }
 
@@ -164,7 +163,7 @@ static void damon_sysfs_scheme_regions_r
 	struct damon_sysfs_scheme_region *r, *next;
 
 	list_for_each_entry_safe(r, next, &regions->regions_list, list) {
-		/* release function deletes it from the list */
+		list_del(&r->list);
 		kobject_put(&r->kobj);
 		regions->nr_regions--;
 	}
@@ -2870,14 +2869,15 @@ void damos_sysfs_populate_region_dir(str
 	if (!region)
 		return;
 	region->sz_filter_passed = sz_filter_passed;
-	list_add_tail(&region->list, &sysfs_regions->regions_list);
-	sysfs_regions->nr_regions++;
 	if (kobject_init_and_add(&region->kobj,
 				&damon_sysfs_scheme_region_ktype,
 				&sysfs_regions->kobj, "%d",
 				sysfs_regions->nr_regions++)) {
 		kobject_put(&region->kobj);
+		return;
 	}
+	list_add_tail(&region->list, &sysfs_regions->regions_list);
+	sysfs_regions->nr_regions++;
 }
 
 int damon_sysfs_schemes_clear_regions(
author maintainer
#152 lore raw

PATCH 138/332 ipc: limit next_id allocation to the valid ID range

In reply to Greg Kroah-Hartman

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

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

From: Linpu Yu <[email protected]>

commit fa0b9b2b7ae3539908d69c2b9ac0d144d9bc5139 upstream.

The checkpoint/restore sysctl path can request the next SysV IPC id through ids->next_id. ipc_idr_alloc() currently forwards that request to idr_alloc() with an open-ended upper bound.

If the valid tail of the SysV IPC id space is full, the allocation can spill beyond ipc_mni. The returned SysV IPC id still uses the normal index encoding, so later lookup and removal can target the wrong slot. This leaves the real IDR entry behind and breaks the IDR state for the object.

The bug is in ipc_idr_alloc() in the checkpoint/restore path.

1. ids->next_id is passed to:

       idr_alloc(&ids->ipcs_idr, new, ipcid_to_idx(next_id), 0, ...)

2. The zero upper bound makes the allocation effectively open-ended. Once the valid SysV IPC tail is occupied, idr_alloc() can spill past ipc_mni and allocate an entry beyond the valid IPC id range.

3. The new object id is still encoded with the narrower SysV IPC index width:

       new->id = (new->seq << ipcmni_seq_shift()) + idx

4. Later removal goes through ipc_rmid(), which uses:

       ipcid_to_idx(ipcp->id)

That truncates the real IDR index. An object actually stored at a high index can then be removed as if it lived at a low in-range index.

5. For shared memory, shm_destroy() frees the current object anyway, but the real high IDR slot is left behind as a dangling pointer.

6. A subsequent walk of /proc/sysvipc/shm reaches the stale IDR entry and dereferences freed memory.

Prevent this by bounding the requested allocation to ipc_mni so the checkpoint/restore path fails once the valid range is exhausted.

---
 ipc/util.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
ipc/util.c +1 -1
@@ -253,7 +253,7 @@ static inline int ipc_idr_alloc(struct i
 	} else {
 		new->seq = ipcid_to_seqx(next_id);
 		idx = idr_alloc(&ids->ipcs_idr, new, ipcid_to_idx(next_id),
-				0, GFP_NOWAIT);
+				ipc_mni, GFP_NOWAIT);
 	}
 	if (idx >= 0)
 		new->id = (new->seq << ipcmni_seq_shift()) + idx;
author maintainer
#153 lore raw

PATCH 139/332 mm: memcontrol: propagate NMI slab stats to memcg vmstats

In reply to Greg Kroah-Hartman

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

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

From: Alexandre Ghiti <[email protected]>

commit e16f17a9c5af50221184d1ef4be4056bf3c4209e upstream.

flush_nmi_stats() drains per-node NMI slab atomics into the per-node lruvec_stats, but does not propagate them to the memcg-level vmstats.

For non NMI case, account_slab_nmi_safe() calls mod_memcg_lruvec_state() which updates both per-node lruvec_stats and memcg-level vmstats, so flush_nmi_stats() needs to flush to per-node lruvec_stats as well as memcg-level vmstats.

So fix this by flushing to the memcg-level vmstats for NMI too.

---
 mm/memcontrol.c |    6 ++++++
 1 file changed, 6 insertions(+)
mm/memcontrol.c +6 -0
@@ -4085,6 +4085,9 @@ static void flush_nmi_stats(struct mem_c
 			lstats->state[index] += slab;
 			if (plstats)
 				plstats->state_pending[index] += slab;
+			memcg->vmstats->state[index] += slab;
+			if (parent)
+				parent->vmstats->state_pending[index] += slab;
 		}
 		if (atomic_read(&pn->slab_unreclaimable)) {
 			int slab = atomic_xchg(&pn->slab_unreclaimable, 0);
@@ -4093,6 +4096,9 @@ static void flush_nmi_stats(struct mem_c
 			lstats->state[index] += slab;
 			if (plstats)
 				plstats->state_pending[index] += slab;
+			memcg->vmstats->state[index] += slab;
+			if (parent)
+				parent->vmstats->state_pending[index] += slab;
 		}
 	}
 }
author maintainer
#154 lore raw

PATCH 140/332 mm/migrate_device: fix pgtable leak in migrate_vma_insert_huge_pmd_page

In reply to Greg Kroah-Hartman

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

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

From: Sunny Patel <[email protected]>

commit 2c6f81d58741349298f51ff697d988cb42881453 upstream.

When migrate_vma_insert_huge_pmd_page() jumps to unlock_abort due to a PMD check failure, the pgtable allocated earlier via pte_alloc_one() is never freed, causing a memory leak.

Added free_abort label to release the pgtable in error path.

Fixes a30b48bf1b24 ("mm/migrate_device: implement THP migration of zone device pages")
Cc Alistair Popple <[email protected]>
Cc Balbir Singh <[email protected]>
Cc Byungchul Park <[email protected]>
Cc Gregory Price <[email protected]>
Cc Joshua Hahn <[email protected]>
Cc Matthew Brost <[email protected]>
Cc Rakie Kim <[email protected]>
Cc Zi Yan <[email protected]>
---
 mm/migrate_device.c |    4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
mm/migrate_device.c +3 -1
@@ -846,7 +846,7 @@ static int migrate_vma_insert_huge_pmd_p
 	} else {
 		if (folio_is_zone_device(folio) &&
 		    !folio_is_device_coherent(folio)) {
-			goto abort;
+			goto free_abort;
 		}
 		entry = folio_mk_pmd(folio, vma->vm_page_prot);
 		if (vma->vm_flags & VM_WRITE)
@@ -899,6 +899,8 @@ static int migrate_vma_insert_huge_pmd_p
 
 unlock_abort:
 	spin_unlock(ptl);
+free_abort:
+	pte_free(vma->vm_mm, pgtable);
 abort:
 	for (i = 0; i < HPAGE_PMD_NR; i++)
 		src[i] &= ~MIGRATE_PFN_MIGRATE;
author maintainer
#155 lore raw

PATCH 141/332 memfd: deny writeable mappings when implying SEAL_WRITE

In reply to Greg Kroah-Hartman

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

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

From: Pratyush Yadav (Google) <[email protected]>

commit 3b041514cb6eae45869b020f743c14d983363222 upstream.

When SEAL_EXEC is added, SEAL_WRITE is implied to make W^X. But the implied seal is set after the check that makes sure the memfd can not have any writable mappings. This means one can use SEAL_EXEC to apply SEAL_WRITE while having writeable mappings.

This breaks the contract that SEAL_WRITE provides and can be used by an attacker to pass a memfd that appears to be write sealed but can still be modified arbitrarily.

Fix this by adding the implied seals before the call for mapping_deny_writable() is done.

Fixes c4f75bc8bd6b ("mm/memfd: add write seals when apply SEAL_EXEC to executable memfd")
Cc Baolin Wang <[email protected]>
Cc Brendan Jackman <[email protected]>
Cc Greg Thelen <[email protected]>
Cc Hugh Dickins <[email protected]>
Cc Kees Cook <[email protected]>
Cc "David Hildenbrand (Arm)" <[email protected]>
---
 mm/memfd.c |   12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)
mm/memfd.c +6 -6
@@ -283,6 +283,12 @@ static int memfd_add_seals(struct file *
 		goto unlock;
 	}
 
+	/*
+	 * SEAL_EXEC implies SEAL_WRITE, making W^X from the start.
+	 */
+	if (seals & F_SEAL_EXEC && inode->i_mode & 0111)
+		seals |= F_SEAL_SHRINK|F_SEAL_GROW|F_SEAL_WRITE|F_SEAL_FUTURE_WRITE;
+
 	if ((seals & F_SEAL_WRITE) && !(*file_seals & F_SEAL_WRITE)) {
 		error = mapping_deny_writable(file->f_mapping);
 		if (error)
@@ -295,12 +301,6 @@ static int memfd_add_seals(struct file *
 		}
 	}
 
-	/*
-	 * SEAL_EXEC implies SEAL_WRITE, making W^X from the start.
-	 */
-	if (seals & F_SEAL_EXEC && inode->i_mode & 0111)
-		seals |= F_SEAL_SHRINK|F_SEAL_GROW|F_SEAL_WRITE|F_SEAL_FUTURE_WRITE;
-
 	*file_seals |= seals;
 	error = 0;
 
author maintainer
#156 lore raw

PATCH 142/332 zram: fix use-after-free in zram_writeback_endio

In reply to Greg Kroah-Hartman

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

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

From: Richard Chang <[email protected]>

commit bf62f69574b19720ae5fbbbcdf24a0c4e3e05e43 upstream.

A crash was observed in zram_writeback_endio due to a NULL pointer dereference in wake_up. The root cause is a race condition between the bio completion handler (zram_writeback_endio) and the writeback task.

In zram_writeback_endio, wake_up() is called on &wb_ctl->done_wait after releasing wb_ctl->done_lock. This creates a race window where the writeback task can see num_inflight become 0, return, and free wb_ctl before zram_writeback_endio calls wake_up().

CPU 0 (zram_writeback_endio) CPU 1 (writeback_store)
============================ ============================

                                 zram_writeback_slots
                                   zram_submit_wb_request
                                   zram_submit_wb_request
                                   wait_event(wb_ctl->done_wait)

spin_lock(&wb_ctl->done_lock); list_add(&req->entry, &wb_ctl->done_reqs); spin_unlock(&wb_ctl->done_lock); wake_up(&wb_ctl->done_wait);
zram_complete_done_reqs
spin_lock(&wb_ctl->done_lock); list_add(&req->entry, &wb_ctl->done_reqs);

spin_unlock(&wb_ctl->done_lock);
                                   while (num_inflight) > 0)
                                     spin_lock(&wb_ctl->done_lock);
                                     list_del(&req->entry);
                                     spin_unlock(&wb_ctl->done_lock);
                                     // num_inflight becomes 0
                                     atomic_dec(num_inflight);

                                 // Leave zram_writeback_slots
                                 // Free wb_ctl
                                 release_wb_ctl(wb_ctl);

// UAF crash!
wake_up(&wb_ctl->done_wait);

This patch fixes this race by using RCU. By protecting wb_ctl with rcu_read_lock() in zram_writeback_endio and using kfree_rcu() to free it, we ensure that wb_ctl remains valid during the execution of zram_writeback_endio.

---
 drivers/block/zram/zram_drv.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)
drivers/block/zram/zram_drv.c +5 -1
index aebc710f0d6a..07111455eecf 100644
@@ -33,6 +33,7 @@
 #include <linux/cpuhotplug.h>
 #include <linux/part_stat.h>
 #include <linux/kernel_read_file.h>
+#include <linux/rcupdate.h>
 
 #include "zram_drv.h"
 
@@ -504,6 +505,7 @@ struct zram_wb_ctl {
 	wait_queue_head_t done_wait;
 	spinlock_t done_lock;
 	atomic_t num_inflight;
+	struct rcu_head rcu;
 };
 
 struct zram_wb_req {
@@ -847,7 +849,7 @@ static void release_wb_ctl(struct zram_wb_ctl *wb_ctl)
 		release_wb_req(req);
 	}
 
-	kfree(wb_ctl);
+	kfree_rcu(wb_ctl, rcu);
 }
 
 static struct zram_wb_ctl *init_wb_ctl(struct zram *zram)
@@ -964,11 +966,13 @@ static void zram_writeback_endio(struct bio *bio)
 	struct zram_wb_ctl *wb_ctl = bio->bi_private;
 	unsigned long flags;
 
+	rcu_read_lock();
 	spin_lock_irqsave(&wb_ctl->done_lock, flags);
 	list_add(&req->entry, &wb_ctl->done_reqs);
 	spin_unlock_irqrestore(&wb_ctl->done_lock, flags);
 
 	wake_up(&wb_ctl->done_wait);
+	rcu_read_unlock();
 }
 
 static void zram_submit_wb_request(struct zram *zram,
Signature · 1 line
2.54.0
author maintainer
#157 lore raw

PATCH 143/332 mm/rmap: initialize nr_pages to 1 at loop start in try_to_unmap_one

In reply to Greg Kroah-Hartman

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

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

From: Dev Jain <[email protected]>

commit 3f8968e9cbf95d5d87d32218906cab0b9b9eddbe upstream.

Initialize nr_pages to 1 at the start of each loop iteration, like folio_referenced_one() does.

Without this, nr_pages computed by a previous folio_unmap_pte_batch() call can be reused on a later iteration that does not run folio_unmap_pte_batch() again.

mmap a 64K large folio with MAP_ANONYMOUS | MAP_DROPPABLE, then call madvise(MADV_FREE), then make the last page device-exclusive via HMM_DMIRROR_EXCLUSIVE.

Trigger node reclaim through sysfs. Now, in try_to_unmap_one(), we will first clear the first 15 out of 16 entries mapping the lazyfree folio. This will set nr_pages to 15. In the next pvmw walk, this nr_pages gets reused on a device-exclusive pte, thus potentially corrupting folio refcount/mapcount.

At the moment, I have a userspace program which can make the kernel spit out a trace, but the blow up is in folio_referenced_one(), because there are existing bugs in the interaction between device-private and rmap (which too I am investigating). I did a one liner kernel change to avoid going into folio_referenced_one(), and the kernel blows up at folio_remove_rmap_ptes in try_to_unmap_one which is what I wanted.

Note that the bug is there not since file folio batching but lazyfree folio batching, since device-exclusive only works for anonymous folios.

Userspace visible effect is simply kernel crashing somewhere due to refcount/mapcount corruption.

Fixes 354dffd29575 ("mm: support batched unmap for lazyfree large folios during reclamation")
Cc Anshuman Khandual <[email protected]>
Cc Barry Song <[email protected]>
Cc Dev Jain <[email protected]>
Cc Harry Yoo <[email protected]>
Cc Jann Horn <[email protected]>
Cc Liam R. Howlett <[email protected]>
Cc Rik van Riel <[email protected]>
Cc Ryan Roberts <[email protected]>
Cc Vlastimil Babka <[email protected]>
---
 mm/rmap.c |    2 ++
 1 file changed, 2 insertions(+)
mm/rmap.c +2 -0
@@ -2024,6 +2024,8 @@ static bool try_to_unmap_one(struct foli
 	mmu_notifier_invalidate_range_start(&range);
 
 	while (page_vma_mapped_walk(&pvmw)) {
+		nr_pages = 1;
+
 		/*
 		 * If the folio is in an mlock()d vma, we must not swap it out.
 		 */
author maintainer
#158 lore raw

PATCH 144/332 auxdisplay: line-display: fix OOB read on zero-length message_store()

In reply to Greg Kroah-Hartman

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

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

From: Stepan Ionichev <[email protected]>

commit a7511dcd9dd4bc55d123f9b800c8a4ed2662e5c6 upstream.

linedisp_display() unconditionally reads msg[count - 1] before checking whether count is zero, so a write of zero bytes to the message sysfs attribute hits msg[-1]:

	write(fd, "", 0);

	-> message_store(..., buf, count=0)
	   -> linedisp_display(linedisp, buf, count=0)
	      -> msg[count - 1] == '\n'  ; OOB read

The kernfs write buffer for that store is a 1-byte allocation (kernfs_fop_write_iter() does kmalloc(len + 1) with len == 0), so msg[-1] is a 1-byte read before the slab object. On a KASAN-enabled kernel this trips an out-of-bounds report and panics; on stock kernels it silently reads adjacent slab data and, if that byte happens to be '\n', the following count-- wraps ssize_t 0 to -1 and is then passed to kmemdup_nul().

linedisp_display() is reached from the message_store() sysfs callback (drivers/auxdisplay/line-display.c message attribute, mode 0644) and from the in-tree initial-message setup with count == -1, so the OOB path is only userspace-triggerable via zero-byte writes; vfs_write() does not short-circuit on count == 0 and kernfs_fop_write_iter() dispatches the store callback regardless.

Guard the trailing-newline trim with a count check. The existing if (!count) block then takes the clear-display path unchanged.

Affects every auxdisplay driver that registers via linedisp_register() / linedisp_attach(): ht16k33, max6959, img-ascii-lcd, seg-led-gpio.

Fixes 7e76aece6f03 ("auxdisplay: Extract character line display core support")
---
 drivers/auxdisplay/line-display.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/auxdisplay/line-display.c +1 -1
@@ -173,7 +173,7 @@ static int linedisp_display(struct lined
 		count = strlen(msg);
 
 	/* if the string ends with a newline, trim it */
-	if (msg[count - 1] == '\n')
+	if (count && msg[count - 1] == '\n')
 		count--;
 
 	if (!count) {
author maintainer
#159 lore raw

PATCH 145/332 smb: client: fix uninitialized variable in smb2_writev_callback

In reply to Greg Kroah-Hartman

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

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

From: Steve French <[email protected]>

commit 9d2491197a00acf8c423512078458c2855102b66 upstream.

compiling with W=2 pointed out that "written may be used uninitialized"

Fixes 20d72b00ca81 ("netfs: Fix the request's work item to not require a ref")
---
 fs/smb/client/smb2pdu.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
fs/smb/client/smb2pdu.c +1 -1
@@ -4943,7 +4943,7 @@ smb2_writev_callback(struct TCP_Server_I
 	unsigned int rreq_debug_id = wdata->rreq->debug_id;
 	unsigned int subreq_debug_index = wdata->subreq.debug_index;
 	ssize_t result = 0;
-	size_t written;
+	size_t written = 0;
 
 	WARN_ONCE(wdata->server != server,
 		  "wdata server %p != mid server %p",
author maintainer
#160 lore raw

PATCH 146/332 Bluetooth: L2CAP: use chan timer to close channels in cleanup_listen()

In reply to Greg Kroah-Hartman

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

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

From: Siwei Zhang <[email protected]>

commit 8c8e620467a7b51562dbcefbd1f09f288d7d710d upstream.

l2cap_chan_close() removes the channel from conn->chan_l, which must be done under conn->lock. cleanup_listen() runs under the parent sk_lock, so acquiring conn->lock would invert the established conn->lock -> chan->lock -> sk_lock order.

Instead of calling l2cap_chan_close() directly, schedule l2cap_chan_timeout with delay 0 to close the channel asynchronously. The timeout handler already acquires conn->lock and chan->lock in the correct order.

The timer is only armed when chan->conn is still set: if it is already NULL, l2cap_conn_del() has already processed this channel (l2cap_chan_del + l2cap_sock_teardown_cb + l2cap_sock_close_cb), so there is nothing left to do. If l2cap_conn_del() races in after the timer is armed, __clear_chan_timer() inside l2cap_chan_del() cancels it; if the timer has already fired, the handler returns harmlessly because chan->conn was cleared.

Fixes 3df91ea20e74 ("Bluetooth: Revert to mutexes from RCU list")
Cc <[email protected]> # 0b58004: Bluetooth: fix UAF in l2cap_sock_cleanup_listen() vs l2cap_conn_del()
---
 net/bluetooth/l2cap_sock.c |   16 +++++++++-------
 1 file changed, 9 insertions(+), 7 deletions(-)
net/bluetooth/l2cap_sock.c +9 -7
@@ -1499,6 +1499,10 @@ static void l2cap_sock_cleanup_listen(st
 	 * pin it (hold_unless_zero() additionally skips a chan already past
 	 * its last reference).  We then drop the sk lock before taking
 	 * chan->lock, so sk and chan locks are never held together.
+	 *
+	 * Since we cannot call l2cap_chan_close() without conn->lock,
+	 * schedule l2cap_chan_timeout to close the channel; it already
+	 * acquires conn->lock -> chan->lock in the correct order.
 	 */
 	while ((sk = bt_accept_dequeue(parent, NULL))) {
 		struct l2cap_chan *chan;
@@ -1516,14 +1520,12 @@ static void l2cap_sock_cleanup_listen(st
 		       state_to_string(chan->state));
 
 		l2cap_chan_lock(chan);
-		__clear_chan_timer(chan);
-		l2cap_chan_close(chan, ECONNRESET);
-		/* l2cap_conn_del() may already have killed this socket
-		 * (it sets SOCK_DEAD); skip the duplicate to avoid a
-		 * double sock_put()/l2cap_chan_put().
+		/* Since we cannot call l2cap_chan_close() without
+		 * conn->lock, schedule its timer to trigger the close
+		 * and cleanup of this channel.
 		 */
-		if (!sock_flag(sk, SOCK_DEAD))
-			l2cap_sock_kill(sk);
+		if (chan->conn)
+			__set_chan_timer(chan, 0);
 		l2cap_chan_unlock(chan);
 
 		l2cap_chan_put(chan);
author maintainer
#161 lore raw

PATCH 147/332 Bluetooth: L2CAP: fix chan ref leak in l2cap_chan_timeout() on !conn

In reply to Greg Kroah-Hartman

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

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

From: Siwei Zhang <[email protected]>

commit 9dbd84990394c51f5cee1e8871bb5ff8af5ed939 upstream.

__set_chan_timer() takes a l2cap_chan reference via l2cap_chan_hold() before scheduling the delayed work. The normal path in l2cap_chan_timeout() drops this reference with l2cap_chan_put() at the end, but the early return when chan->conn is NULL skips the put, leaking the reference.

Add the missing l2cap_chan_put() before the early return.

Fixes adf0398cee86 ("Bluetooth: l2cap: fix null-ptr-deref in l2cap_chan_timeout")
---
 net/bluetooth/l2cap_core.c |    4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
net/bluetooth/l2cap_core.c +3 -1
@@ -411,8 +411,10 @@ static void l2cap_chan_timeout(struct wo
 
 	BT_DBG("chan %p state %s", chan, state_to_string(chan->state));
 
-	if (!conn)
+	if (!conn) {
+		l2cap_chan_put(chan);
 		return;
+	}
 
 	mutex_lock(&conn->lock);
 	/* __set_chan_timer() calls l2cap_chan_hold(chan) while scheduling
author maintainer
#162 lore raw

PATCH 148/332 Bluetooth: HIDP: fix missing length checks in hidp_input_report()

In reply to Greg Kroah-Hartman

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

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

From: Muhammad Bilal <[email protected]>

commit 2a3ac9ee11dbb9845f3947cef4a79dba658cf6f6 upstream.

hidp_input_report() reads keyboard and mouse payload data from an skb without first verifying that skb->len contains enough data.

hidp_recv_intr_frame() pulls the 1-byte HIDP header before dispatching to hidp_input_report(). If a paired device sends a truncated packet, the handler reads beyond the valid skb data, resulting in an out-of-bounds read of skb data. The OOB bytes may be interpreted as phantom key presses or spurious mouse movement.

Replace the open-coded length tracking and pointer arithmetic with skb_pull_data() calls. skb_pull_data() returns NULL if the requested bytes are not present, eliminating the need for a manual size variable and the separate skb->len guard.

---
 net/bluetooth/hidp/core.c |   23 ++++++++++++++++++-----
 1 file changed, 18 insertions(+), 5 deletions(-)
net/bluetooth/hidp/core.c +18 -5
@@ -179,12 +179,21 @@ static void hidp_input_report(struct hid
 {
 	struct input_dev *dev = session->input;
 	unsigned char *keys = session->keys;
-	unsigned char *udata = skb->data + 1;
-	signed char *sdata = skb->data + 1;
-	int i, size = skb->len - 1;
+	unsigned char *udata;
+	signed char *sdata;
+	u8 *hdr;
+	int i;
+
+	hdr = skb_pull_data(skb, 1);
+	if (!hdr)
+		return;
 
-	switch (skb->data[0]) {
+	switch (*hdr) {
 	case 0x01:	/* Keyboard report */
+		udata = skb_pull_data(skb, 8);
+		if (!udata)
+			break;
+
 		for (i = 0; i < 8; i++)
 			input_report_key(dev, hidp_keycode[i + 224], (udata[0] >> i) & 1);
 
@@ -213,6 +222,10 @@ static void hidp_input_report(struct hid
 		break;
 
 	case 0x02:	/* Mouse report */
+		sdata = skb_pull_data(skb, 3);
+		if (!sdata)
+			break;
+
 		input_report_key(dev, BTN_LEFT,   sdata[0] & 0x01);
 		input_report_key(dev, BTN_RIGHT,  sdata[0] & 0x02);
 		input_report_key(dev, BTN_MIDDLE, sdata[0] & 0x04);
@@ -222,7 +235,7 @@ static void hidp_input_report(struct hid
 		input_report_rel(dev, REL_X, sdata[1]);
 		input_report_rel(dev, REL_Y, sdata[2]);
 
-		if (size > 3)
+		if (skb->len > 0)
 			input_report_rel(dev, REL_WHEEL, sdata[3]);
 		break;
 	}
author maintainer
#163 lore raw

PATCH 149/332 Bluetooth: ISO: fix UAF in iso_recv_frame

In reply to Greg Kroah-Hartman

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

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

From: Muhammad Bilal <[email protected]>

commit 47f23a259517abbdb8032c057a1e8a6bf3734878 upstream.

iso_recv_frame reads conn->sk under iso_conn_lock but releases the lock before using sk, with no reference held. A concurrent iso_sock_kill() can free sk in that window, causing use-after-free on sk->sk_state and sock_queue_rcv_skb().

Fix by replacing the bare pointer read with iso_sock_hold(conn), which calls sock_hold() while the spinlock is held, atomically elevating the refcount before the lock drops. Add a drop_put label so sock_put() is called on all exit paths where the hold succeeded.

Fixes ccf74f2390d60a2f9a75ef496d2564abb478f46a ("Bluetooth: Add BTPROTO_ISO socket type")
---
 net/bluetooth/iso.c |   10 +++++++---
 1 file changed, 7 insertions(+), 3 deletions(-)
net/bluetooth/iso.c +7 -3
@@ -572,7 +572,7 @@ static void iso_recv_frame(struct iso_co
 	struct sock *sk;
 
 	iso_conn_lock(conn);
-	sk = conn->sk;
+	sk = iso_sock_hold(conn);
 	iso_conn_unlock(conn);
 
 	if (!sk)
@@ -581,11 +581,15 @@ static void iso_recv_frame(struct iso_co
 	BT_DBG("sk %p len %d", sk, skb->len);
 
 	if (sk->sk_state != BT_CONNECTED)
-		goto drop;
+		goto drop_put;
 
-	if (!sock_queue_rcv_skb(sk, skb))
+	if (!sock_queue_rcv_skb(sk, skb)) {
+		sock_put(sk);
 		return;
+	}
 
+drop_put:
+	sock_put(sk);
 drop:
 	kfree_skb(skb);
 }
author maintainer
#164 lore raw

PATCH 150/332 Bluetooth: ISO: serialize iso_sock_clear_timer with socket lock

In reply to Greg Kroah-Hartman

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

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

From: Muhammad Bilal <[email protected]>

commit 4b5f8e608749b7e8fa386c6e4301cf9272595859 upstream.

iso_sock_close() calls iso_sock_clear_timer() before acquiring lock_sock(sk).

iso_sock_clear_timer() reads iso_pi(sk)->conn twice without the socket lock held:

    if (!iso_pi(sk)->conn)
        return;
    cancel_delayed_work(&iso_pi(sk)->conn->timeout_work);

Concurrently, iso_conn_del() executes under lock_sock(sk) and calls iso_chan_del(), which sets iso_pi(sk)->conn to NULL and may result in the final reference to the connection being dropped:

    CPU0                         CPU1
    ----                         ----
    iso_sock_clear_timer()
      if (conn != NULL) ...      lock_sock(sk)
                                   iso_chan_del()
                                   iso_pi(sk)->conn = NULL
      cancel_delayed_work(conn)  /* NULL deref or UAF */

iso_pi(sk)->conn is not stable across the unlock window, causing a NULL pointer dereference or use-after-free.

Serialize iso_sock_clear_timer() with the socket lock by moving it inside lock_sock()/release_sock(), matching the pattern used in iso_conn_del() and all other call sites.

Fixes ccf74f2390d60a2f9a75ef496d2564abb478f46a ("Bluetooth: Add BTPROTO_ISO socket type")
---
 net/bluetooth/iso.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
net/bluetooth/iso.c +1 -1
@@ -872,8 +872,8 @@ static void __iso_sock_close(struct sock
 /* Must be called on unlocked socket. */
 static void iso_sock_close(struct sock *sk)
 {
-	iso_sock_clear_timer(sk);
 	lock_sock(sk);
+	iso_sock_clear_timer(sk);
 	__iso_sock_close(sk);
 	release_sock(sk);
 	iso_sock_kill(sk);
author maintainer
#165 lore raw

PATCH 151/332 Bluetooth: hci_conn: Fix memory leak in hci_le_big_terminate()

In reply to Greg Kroah-Hartman

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

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

From: Pavitra Jha <[email protected]>

commit bfa9d28960ed677d556bdf097073bc3129686229 upstream.

hci_le_big_terminate() allocates iso_list_data via kzalloc_obj but returns 0 without freeing it when neither pa_sync_term nor big_sync_term flags are set after evaluating the PA and BIG sync connection state.

This early-return path was introduced when hci_le_big_terminate() was refactored to take struct hci_conn instead of raw u8 parameters, adding PA/BIG flag evaluation logic. The existing kfree() on hci_cmd_sync_queue failure does not cover this path.

Fixes a7bcffc673de ("Bluetooth: Add PA_LINK to distinguish BIG sync and PA sync connections")
---
 net/bluetooth/hci_conn.c |    4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
net/bluetooth/hci_conn.c +3 -1
@@ -803,8 +803,10 @@ static int hci_le_big_terminate(struct h
 			d->big_sync_term = true;
 	}
 
-	if (!d->pa_sync_term && !d->big_sync_term)
+	if (!d->pa_sync_term && !d->big_sync_term) {
+		kfree(d);
 		return 0;
+	}
 
 	ret = hci_cmd_sync_queue(hdev, big_terminate_sync, d,
 				 terminate_big_destroy);
author maintainer
#166 lore raw

PATCH 152/332 Bluetooth: hci_qca: Use 100 ms SSR delay for rampatch and NVM loading

In reply to Greg Kroah-Hartman

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

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

From: Shuai Zhang <[email protected]>

commit fa21e86caba2347e89eb65af926205a36a097c53 upstream.

When bt_en is pulled high by hardware, the host does not re-download the firmware after SSR. The controller loads the rampatch and NVM internally.

On HMT chip, the rampatch is ~264 KB and the NVM is ~9.4 KB. The loading process takes approximately 70 ms. The previous 50 ms delay is too short, causing the controller to not respond to the reset command sent by the host, which leads to BT initialization failure:

Bluetooth: hci0: QCA memdump Done, received 458752, total 458752
Bluetooth: hci0: mem_dump_status: 2
Bluetooth: hci0: Opcode 0x0c03 failed: -110

Increase the delay to 100 ms, which was confirmed as a safe value by the controller, to ensure the controller has finished loading the firmware before the host sends commands.

Steps to reproduce:
1. Trigger SSR and wait for SSR to complete:
hcitool cmd 0x3f 0c 26
2. Run "bluetoothctl power on" and observe that BT fails to start.

Fixes fce1a9244a0f ("Bluetooth: hci_qca: Fix SSR (SubSystem Restart) fail when BT_EN is pulled up by hw")
---
 drivers/bluetooth/hci_qca.c |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/bluetooth/hci_qca.c +2 -2
@@ -1677,8 +1677,8 @@ static void qca_hw_error(struct hci_dev
 		mod_timer(&qca->tx_idle_timer, jiffies +
 				  msecs_to_jiffies(qca->tx_idle_delay));
 
-		/* Controller reset completion time is 50ms */
-		msleep(50);
+		/* Wait for the controller to load the rampatch and NVM. */
+		msleep(100);
 
 		clear_bit(QCA_SSR_TRIGGERED, &qca->flags);
 		clear_bit(QCA_IBS_DISABLED, &qca->flags);
author maintainer
#167 lore raw

PATCH 153/332 Bluetooth: hci_sync: fix UAF in hci_le_create_cis_sync

In reply to Greg Kroah-Hartman

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

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

From: Doruk Tan Ozturk <[email protected]>

commit bfea6091e0fffb270c20e74384b660910277eb6c upstream.

hci_le_create_cis_sync() dereferences conn->conn_timeout after releasing both rcu_read_lock() and hci_dev_lock(hdev). The conn pointer was obtained from an RCU-protected iteration over hdev->conn_hash.list and is not valid once these locks are dropped. A concurrent disconnect can free the hci_conn between the unlock and the dereference, causing a use-after-free read.

The cancellation mechanism in hci_conn_del() cannot prevent this because hci_le_create_cis_pending() queues hci_create_cis_sync with data=NULL:

    hci_cmd_sync_queue(hdev, hci_create_cis_sync, NULL, NULL);

While hci_conn_del() dequeues with data=conn:

    hci_cmd_sync_dequeue(hdev, NULL, conn, NULL);

Since NULL != conn, the lookup in _hci_cmd_sync_lookup_entry() never matches, and the pending work item is not cancelled.

Fix this by saving conn->conn_timeout into a local variable while the locks are still held, so the stale conn pointer is never dereferenced after unlock.

This is the same class of bug as the one fixed by commit 035c25007c9e ("Bluetooth: hci_sync: Fix UAF on le_read_features_complete") which addressed the identical pattern in a different function.

This vulnerability was identified using 0sec.ai, an open-source automated security auditing platform (https://github.com/0sec-labs).

Fixes c09b80be6ffc ("Bluetooth: hci_conn: Fix not waiting for HCI_EVT_LE_CIS_ESTABLISHED")
---
 net/bluetooth/hci_sync.c |    4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
net/bluetooth/hci_sync.c +3 -1
@@ -6711,6 +6711,7 @@ int hci_le_create_cis_sync(struct hci_de
 	DEFINE_FLEX(struct hci_cp_le_create_cis, cmd, cis, num_cis, 0x1f);
 	size_t aux_num_cis = 0;
 	struct hci_conn *conn;
+	u16 timeout = 0;
 	u8 cig = BT_ISO_QOS_CIG_UNSET;
 
 	/* The spec allows only one pending LE Create CIS command at a time. If
@@ -6781,6 +6782,7 @@ int hci_le_create_cis_sync(struct hci_de
 		set_bit(HCI_CONN_CREATE_CIS, &conn->flags);
 		cis->acl_handle = cpu_to_le16(conn->parent->handle);
 		cis->cis_handle = cpu_to_le16(conn->handle);
+		timeout = conn->conn_timeout;
 		aux_num_cis++;
 
 		if (aux_num_cis >= cmd->num_cis)
@@ -6800,7 +6802,7 @@ done:
 	return __hci_cmd_sync_status_sk(hdev, HCI_OP_LE_CREATE_CIS,
 					struct_size(cmd, cis, cmd->num_cis),
 					cmd, HCI_EVT_LE_CIS_ESTABLISHED,
-					conn->conn_timeout, NULL);
+					timeout, NULL);
 }
 
 int hci_le_remove_cig_sync(struct hci_dev *hdev, u8 handle)
author maintainer
#168 lore raw

PATCH 154/332 Input: xpad - fix out-of-bounds access for Share button

In reply to Greg Kroah-Hartman

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

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

From: Dmitry Torokhov <[email protected]>

commit 6cdc46b38cf146ce81d4831b6472dbf7731849a2 upstream.

xpadone_process_packet() receives len directly from urb->actual_length and uses it to index the share-button byte at data[len - 18] or data[len - 26]. Since both len and data[0] are under the device's control, a broken controller can send a GIP_CMD_INPUT packet with actual_length < 18 (e.g. 5 bytes) and reach this code path, causing accesses beyond the actual array.

Fix this by calculating the offset and checking bounds against the packet length.

Fixes 4ef46367073b ("Input: xpad - fix Share button on Xbox One controllers")
---
 drivers/input/joystick/xpad.c |    8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)
drivers/input/joystick/xpad.c +4 -4
@@ -1110,10 +1110,10 @@ static void xpadone_process_packet(struc
 		input_report_key(dev, BTN_START,  data[4] & BIT(2));
 		input_report_key(dev, BTN_SELECT, data[4] & BIT(3));
 		if (xpad->mapping & MAP_SHARE_BUTTON) {
-			if (xpad->mapping & MAP_SHARE_OFFSET)
-				input_report_key(dev, KEY_RECORD, data[len - 26] & BIT(0));
-			else
-				input_report_key(dev, KEY_RECORD, data[len - 18] & BIT(0));
+			u32 offset = (xpad->mapping & MAP_SHARE_OFFSET) ? 26 : 18;
+
+			if (len >= offset)
+				input_report_key(dev, KEY_RECORD, data[len - offset] & BIT(0));
 		}
 
 		/* buttons A,B,X,Y */
author maintainer
#169 lore raw

PATCH 155/332 parport: Fix race between port and client registration

In reply to Greg Kroah-Hartman

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

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

From: Ben Hutchings <[email protected]>

commit ef15ccbb3e8640a723c42ad90eaf81d66ae02017 upstream.

The parport subsystem registers port devices before they are fully initialised, resulting in a race condition where client drivers such as lp can attach to ports that are not completely initialised or even being torn down.

When the port and client drivers are built as modules and loaded around the same time during boot, this occasionally results in a crash. I was able to make this happen reliably in a VM with a PC-style parallel port by patching parport_pc to fail probing:

drivers/parport/parport_pc.c static struct parport *__parport_pc_probe_port(unsigned long int base, · line 2075

Quoted 11 lines
drivers/parport/parport_pc.c +1 -1
@@ -2069,7 +2069,7 @@ static struct parport *__parport_pc_probe_port(unsigned long int base,
 	if (!p)
 		goto out3;
 
-	base_res = request_region(base, 3, p->name);
+	base_res = NULL;
 	if (!base_res)
 		goto out4;
 

and then running:

while true; do
modprobe lp & modprobe parport_pc
wait
rmmod lp parport_pc
done

for a few seconds.

In the long term I think port registration should be changed to put the call to device_add() inside parport_announce_port(), but since the latter currently cannot fail this will require changing all port drivers.

For now, add a flag to indicate whether a port has been "announced" and only try to attach client drivers to ports when the flag is set.

---
 drivers/parport/share.c |   11 +++++++++--
 include/linux/parport.h |    1 +
 2 files changed, 10 insertions(+), 2 deletions(-)
drivers/parport/share.c +9 -2
@@ -214,10 +214,14 @@ static void get_lowlevel_driver(void)
 static int port_check(struct device *dev, void *dev_drv)
 {
 	struct parport_driver *drv = dev_drv;
+	struct parport *port;
 
 	/* only send ports, do not send other devices connected to bus */
-	if (is_parport(dev))
-		drv->match_port(to_parport_dev(dev));
+	if (is_parport(dev)) {
+		port = to_parport_dev(dev);
+		if (test_bit(PARPORT_ANNOUNCED, &port->devflags))
+			drv->match_port(port);
+	}
 	return 0;
 }
 
@@ -532,6 +536,7 @@ void parport_announce_port(struct parpor
 		if (slave)
 			attach_driver_chain(slave);
 	}
+	set_bit(PARPORT_ANNOUNCED, &port->devflags);
 	mutex_unlock(&registration_lock);
 }
 EXPORT_SYMBOL(parport_announce_port);
@@ -561,6 +566,8 @@ void parport_remove_port(struct parport
 
 	mutex_lock(&registration_lock);
 
+	clear_bit(PARPORT_ANNOUNCED, &port->devflags);
+
 	/* Spread the word. */
 	detach_driver_chain(port);
 
include/linux/parport.h +1 -0
@@ -240,6 +240,7 @@ struct parport {
 
 	unsigned long devflags;
 #define PARPORT_DEVPROC_REGISTERED	0
+#define PARPORT_ANNOUNCED		1
 	struct pardevice *proc_device;	/* Currently register proc device */
 
 	struct list_head full_list;
author maintainer
#170 lore raw

PATCH 156/332 rust_binder: Avoid holding lock when dropping delivered_death

In reply to Greg Kroah-Hartman

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

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

From: Matthew Maurer <[email protected]>

commit f6d8fea9e3953151a4adb4f603503dc3dc9c69da upstream.

In 6c37bebd8c926, we switched to looping over the list and dropping each individual node, ostensibly without the lock held in the loop body.

If the kernel were using Rust Edition 2024, the comment would be accurate, and the lock would not be held across the drop. However, the kernel is currently using 2021, so tail expression lifetime extension results in the lock being held across the drop. Explicitly binding the expression result to a variable makes the lockguard no longer part of a tail expression, causing the lock to be dropped before entering the loop body.

This was detected via CONFIG_PROVE_LOCKING identifying an invalid wait context at the drop site.

---
 drivers/android/binder/process.rs |    7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)
drivers/android/binder/process.rs +6 -1
@@ -1402,7 +1402,12 @@ impl Process {
         // Clear delivered_deaths list.
         //
         // Scope ensures that MutexGuard is dropped while executing the body.
-        while let Some(delivered_death) = { self.inner.lock().delivered_deaths.pop_front() } {
+        while let Some(delivered_death) = {
+            // Explicitly bind to avoid tail expression lifetime extension of the lockguard
+            // Can be removed when the kernel moves to edition 2024
+            let maybe_death = self.inner.lock().delivered_deaths.pop_front();
+            maybe_death
+        } {
             drop(delivered_death);
         }
 
author maintainer
#171 lore raw

PATCH 157/332 rust_binder: avoid calling pending_oneway_finished() on TF_UPDATE_TXN

In reply to Greg Kroah-Hartman

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

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

From: Alice Ryhl <[email protected]>

commit 4c19719eb8b8df08c5bec7c499f73ddaea6f09fc upstream.

When an outdated transaction is removed from oneway_todo due to TF_UPDATE_TXN, its Allocation is dropped. The current implementation of Allocation::drop calls pending_oneway_finished(), assuming the transaction was executed. This leads to premature execution of the next queued one-way transaction.

Fix this by taking the oneway_node from the Allocation of the outdated transaction before it is dropped. This prevents Allocation::drop from signaling completion.

We do not call take_oneway_node() from Transaction::cancel because it's actually correct to call pending_oneway_finished() on cancel if the transaction did not come from oneway_todo. This ensures that if BINDER_THREAD_EXIT is invoked and cancels a oneway transaction, then the next transaction is taken from oneway_todo.

This bug does not lead to any issues in the kernel, but may lead to Binder delivering transactions to userspace earlier than userspace expected to receive them.

Cc stable <[email protected]>
Fixes eafedbc7c050 ("rust_binder: add Rust Binder driver")
Assisted-by Antigravity:gemini
---
 drivers/android/binder/allocation.rs  |    8 ++++++++
 drivers/android/binder/transaction.rs |   11 ++++++++++-
 2 files changed, 18 insertions(+), 1 deletion(-)
drivers/android/binder/allocation.rs +8 -0
@@ -160,6 +160,14 @@ impl Allocation {
         self.get_or_init_info().target_node = Some(target_node);
     }
 
+    pub(crate) fn take_oneway_node(&mut self) -> Option<DArc<Node>> {
+        if let Some(info) = self.allocation_info.as_mut() {
+            info.oneway_node.take()
+        } else {
+            None
+        }
+    }
+
     /// Reserve enough space to push at least `num_fds` fds.
     pub(crate) fn info_add_fd_reserve(&mut self, num_fds: usize) -> Result {
         self.get_or_init_info()
drivers/android/binder/transaction.rs +10 -1
@@ -250,7 +250,8 @@ impl Transaction {
     /// Not used for replies.
     pub(crate) fn submit(self: DLArc<Self>) -> BinderResult {
         // Defined before `process_inner` so that the destructor runs after releasing the lock.
-        let mut _t_outdated;
+        let _t_outdated;
+        let _oneway_node;
 
         let oneway = self.flags & TF_ONE_WAY != 0;
         let process = self.to.clone();
@@ -267,6 +268,14 @@ impl Transaction {
                         if let Some(t_outdated) =
                             target_node.take_outdated_transaction(&self, &mut process_inner)
                         {
+                            let mut alloc_guard = t_outdated.allocation.lock();
+                            if let Some(alloc) = (*alloc_guard).as_mut() {
+                                // Take the oneway node to prevent `Allocation::drop` from calling
+                                // `pending_oneway_finished()`, which would be incorrect as this
+                                // transaction is not being submitted.
+                                _oneway_node = alloc.take_oneway_node();
+                            }
+                            drop(alloc_guard);
                             // Save the transaction to be dropped after locks are released.
                             _t_outdated = t_outdated;
                         }
author maintainer
#172 lore raw

PATCH 158/332 USB: cdc-acm: Fix bit overlap and move quirk definitions to header

In reply to Greg Kroah-Hartman

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

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

From: Wentao Guan <[email protected]>

commit 5eb070769ea5e18405535609d1d3f6886f3755bd upstream.

The VENDOR_CLASS_DATA_IFACE and ALWAYS_POLL_CTRL quirk flags added in commit f58752ebcb35 ("USB: cdc-acm: Add quirks for Yoga Book 9 14IAH10 INGENIC touchscreen") were placed inside the acm_ctrl_msg() function rather than in the header with the other quirk flags. Then, their values (BIT(9) and BIT(10)) collided with NO_UNION_12 which is already BIT(9).

Move the definitions to drivers/usb/class/cdc-acm.h where they belong and shift them to BIT(10) and BIT(11) to avoid the overlap.

Fixes f58752ebcb35 ("USB: cdc-acm: Add quirks for Yoga Book 9 14IAH10 INGENIC touchscreen")
Cc stable <[email protected]>
---
 drivers/usb/class/cdc-acm.c |    2 --
 drivers/usb/class/cdc-acm.h |    2 ++
 2 files changed, 2 insertions(+), 2 deletions(-)
drivers/usb/class/cdc-acm.c +0 -2
@@ -114,8 +114,6 @@ static int acm_ctrl_msg(struct acm *acm,
 	int retval;
 
 	retval = usb_autopm_get_interface(acm->control);
-#define VENDOR_CLASS_DATA_IFACE		BIT(9)  /* data interface uses vendor-specific class */
-#define ALWAYS_POLL_CTRL		BIT(10) /* keep ctrl URB active even without an open TTY */
 	if (retval)
 		return retval;
 
drivers/usb/class/cdc-acm.h +2 -0
@@ -115,3 +115,5 @@ struct acm {
 #define DISABLE_ECHO			BIT(7)
 #define MISSING_CAP_BRK			BIT(8)
 #define NO_UNION_12			BIT(9)
+#define VENDOR_CLASS_DATA_IFACE		BIT(10)  /* data interface uses vendor-specific class */
+#define ALWAYS_POLL_CTRL		BIT(11) /* keep ctrl URB active even without an open TTY */
author maintainer
#173 lore raw

PATCH 160/332 KVM: arm64: PMU: Preserve AArch32 counter low bits

In reply to Greg Kroah-Hartman

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

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

From: Qiang Ma <[email protected]>

commit 1750ad1388e03fb27068cd1f22c9c8b4590fe936 upstream.

AArch32 writes to PMU event counters cannot update the top 32 bits, even when PMUv3p5 makes the counters 64-bit. KVM therefore needs to preserve the existing high half and only update the low half written by the guest, unless the caller explicitly forces a full reset through PMCR.P.

The current code masks @val down to the old high half before taking lower_32_bits(val), which means the low half is always zero. As a result, AArch32 writes to event counters discard the guest-provided low 32 bits instead of storing them.

Build the new value from the old high 32 bits and the low 32 bits of the value supplied by the guest.

Fixes 26d2d0594d70 ("KVM: arm64: PMU: Do not let AArch32 change the counters' top 32 bits")
---
 arch/arm64/kvm/pmu-emul.c |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
arch/arm64/kvm/pmu-emul.c +2 -2
@@ -174,8 +174,8 @@ static void kvm_pmu_set_pmc_value(struct
 		 * action is to use PMCR.P, which will reset them to
 		 * 0 (the only use of the 'force' parameter).
 		 */
-		val  = __vcpu_sys_reg(vcpu, reg) & GENMASK(63, 32);
-		val |= lower_32_bits(val);
+		val = (__vcpu_sys_reg(vcpu, reg) & GENMASK(63, 32)) |
+		      lower_32_bits(val);
 	}
 
 	__vcpu_assign_sys_reg(vcpu, reg, val);
author maintainer
#174 lore raw

PATCH 161/332 KVM: SVM: Flush the current TLB when transitioning from xAVIC => x2AVIC

In reply to Greg Kroah-Hartman

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

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

From: Sean Christopherson <[email protected]>

commit a9e18aa3263f356edae305e29830e5fe63d8597a upstream.

Flush the current TLB when xAVIC *or* x2AVIC is activated, as KVM is (apparently) responsible for purging TLB entries when transitioning from xAVIC to x2AVIC. The APM says a whole lot of nothing about TLB flushing with respect to (x2)AVIC, but empirical data strongly suggests hardware also does a whole lot of nothing.

Failure to flush the TLB when enabling x2AVIC can lead to guest accesses to the APIC base address getting incorrectly redirected to the virtual APIC page. The flaw most visibly manifests as failures in KVM-Unit-Test's verify_disabled_apic_mmio() testcase when x2APIC is enabled (though for reasons unknown, the test only reliably fails with EFI builds).

Fixes 0ccf3e7cb95a ("KVM: SVM: Flush the "current" TLB when activating AVIC")
Fixes 4d1d7942e36a ("KVM: SVM: Introduce logic to (de)activate x2AVIC mode")
Cc Naveen N Rao (AMD) <[email protected]>
---
 arch/x86/kvm/svm/avic.c |   35 +++++++++++++++++++++++++++++------
 1 file changed, 29 insertions(+), 6 deletions(-)
arch/x86/kvm/svm/avic.c +29 -6
@@ -196,6 +196,35 @@ static void avic_activate_vmcb(struct vc
 	svm_clr_intercept(svm, INTERCEPT_CR8_WRITE);
 
 	/*
+	 * Flush the TLB when enabling (x2)AVIC and when transitioning between
+	 * xAVIC and x2AVIC, as the CPU may have inserted a TLB entry for the
+	 * "wrong" mapping.
+	 *
+	 * KVM uses a per-VM "scratch" page to back the APIC memslot, because
+	 * KVM also uses per-VM page tables *and* maintains the page table (NPT
+	 * or shadow page) mappings for said memslot even if one or more vCPUs
+	 * have their local APIC hardware-disabled or are in x2APIC mode, i.e.
+	 * even if one or more vCPUs' APIC MMIO BAR is effectively disabled.
+	 *
+	 * If xAVIC is fully enabled, hardware ignores the physical address in
+	 * KVM's page tables, i.e. in the leaf SPTE for the APIC memslot, and
+	 * instead redirects the access to the AVIC backing page, i.e. to the
+	 * vCPU's virtual APIC page.  If xAVIC is not enabled (APIC is either
+	 * hardware-disabled or in x2APIC mode), then guest accesses will use
+	 * the page table mapping verbatim, i.e. will access the per-VM scratch
+	 * page, as normal memory.
+	 *
+	 * In both cases, the CPU is allowed to cache TLB entries for the APIC
+	 * base GPA.  So, KVM needs to flush the TLB when enabling xAVIC, as
+	 * accesses need to be redirected to the virtual APIC page, but the TLB
+	 * may contain entries pointing at the scratch page.  KVM also needs to
+	 * flush the TLB when enabling x2AVIC, as accesses need to go to the
+	 * scratch page, but the TLB may contain entries tagged as xAVIC, i.e.
+	 * entries pointing to the vCPU's virtual APIC page.
+	 */
+	kvm_make_request(KVM_REQ_TLB_FLUSH_CURRENT, &svm->vcpu);
+
+	/*
 	 * Note: KVM supports hybrid-AVIC mode, where KVM emulates x2APIC MSR
 	 * accesses, while interrupt injection to a running vCPU can be
 	 * achieved using AVIC doorbell.  KVM disables the APIC access page
@@ -208,12 +237,6 @@ static void avic_activate_vmcb(struct vc
 		/* Disabling MSR intercept for x2APIC registers */
 		avic_set_x2apic_msr_interception(svm, false);
 	} else {
-		/*
-		 * Flush the TLB, the guest may have inserted a non-APIC
-		 * mapping into the TLB while AVIC was disabled.
-		 */
-		kvm_make_request(KVM_REQ_TLB_FLUSH_CURRENT, &svm->vcpu);
-
 		/* Enabling MSR intercept for x2APIC registers */
 		avic_set_x2apic_msr_interception(svm, true);
 	}
author maintainer
#175 lore raw

PATCH 162/332 KVM: SEV: Require in-GHCB scratch area if GHCB v2+ is in use

In reply to Greg Kroah-Hartman

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

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

From: Michael Roth <[email protected]>

commit db3f2195d29344a3cf1e9dd9ab7f21ced7308cf7 upstream.

As per the GHCB spec, when using GHCB v2+ require the software scratch area to reside in the GHCB's shared buffer. Note, things like Page State Change (PSC) requests _rely_ on this behavior, as the guest can't provide a length when making the request, i.e. the size of the guest payload is bounded by the size of the shared buffer.

Failure to force usage of the GHCB, and a slew of other flaws, lets a malicious SNP guest corrupt host kernel heap memory, and leak host heap layout information.

setup_vmgexit_scratch() allocates a buffer via kvzalloc(exit_info_2), where exit_info_2 is guest-controlled. With exit_info_2=24, this yields a 24-byte allocation in kmalloc-cg-32 (32-byte slab objects). The buffer holds an 8-byte psc_hdr followed by 8-byte psc_entry structs, so only entries[0] and entries[1] are in-bounds.

snp_begin_psc() validates end_entry against VMGEXIT_PSC_MAX_COUNT (253) but NOT against the actual buffer size:

      idx_end = hdr->end_entry;

      if (idx_end >= VMGEXIT_PSC_MAX_COUNT) {   // checks 253, not buffer
          snp_complete_psc(svm, ...);
          return 1;
      }

      for (idx = idx_start; idx <= idx_end; idx++) {
          entry_start = entries[idx];           // OOB when idx >= 2

The guest sets end_entry=10+, causing the host to iterate entries[2+] which are OOB into adjacent slab objects. For each OOB entry:

- The host reads 8 bytes (OOB READ / info leak oracle)
- If the data passes PSC validation, __snp_complete_one_psc() writes

    cur_page = 1 or 512 into the entry (OOB WRITE, sev.c:3806)

- If validation fails, the error response reveals whether adjacent memory is zero vs non-zero (information disclosure to guest)

The guest controls allocation size (exit_info_2), entry range (cur_entry/end_entry), and can fire unlimited VMGEXITs to repeatedly hit different slab positions.

By exploiting the variety of bugs, a malicious SEV-SNP guest can:
- OOB read adjacent kmalloc-cg-32 objects (heap layout disclosure)
- OOB write cur_page bits into adjacent objects (heap corruption)
- Trigger use-after-free conditions across VMGEXITs

E.g. with KASAN enabled, a single insmod of the PoC guest module produces 73 KASAN reports:

    BUG: KASAN: slab-out-of-bounds in snp_begin_psc+0x126/0x890
    Read of size 8 at addr ffff888219ffb5e0 by task qemu-system-x86/2199

    BUG: KASAN: slab-out-of-bounds in snp_begin_psc+0x468/0x890
    Write of size 8 at addr ffff888351566648 by task qemu-system-x86/2199

    The buggy address belongs to the object at ffff888XXXXXXXXX
     which belongs to the cache kmalloc-cg-32 of size 32
    The buggy address is located N bytes to the right of
     allocated 32-byte region [ffff888XXXXXXXXX, ffff888XXXXXXXXX)

Breakdown:
62 slab-out-of-bounds (reads + writes past allocation)
7 slab-use-after-free
4 use-after-free

All credit to Stan for the wonderful description and reproducer!

Cc Michael Roth <[email protected]>
Cc Tom Lendacky <[email protected]>
Cc Peter Gonda <[email protected]>
Cc Jacky Li <[email protected]>
Fixes 4af663c2f64a ("KVM: SEV: Allow per-guest configuration of GHCB protocol version")
[sean: write changelog]

Message-ID: <[email protected]>

---
 arch/x86/kvm/svm/sev.c |    4 ++++
 1 file changed, 4 insertions(+)
arch/x86/kvm/svm/sev.c +4 -0
@@ -3658,6 +3658,10 @@ static int setup_vmgexit_scratch(struct
 		scratch_va = (void *)svm->sev_es.ghcb;
 		scratch_va += (scratch_gpa_beg - control->ghcb_gpa);
 	} else {
+		/* GHCB v2 requires the scratch area to be within the GHCB. */
+		if (to_kvm_sev_info(svm->vcpu.kvm)->ghcb_version >= 2)
+			goto e_scratch;
+
 		/*
 		 * The guest memory must be read into a kernel buffer, so
 		 * limit the size
author maintainer
#176 lore raw

PATCH 163/332 KVM: SEV: Ignore Port I/O requests of length 0

In reply to Greg Kroah-Hartman

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

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

From: Sean Christopherson <[email protected]>

commit 3988bd2723de407ae90fa7a6f6029b4e60238c58 upstream.

Explicitly ignore Port I/O requests of length '0' (or count '0'), so that setting up the software scratch area (and other code) doesn't have to worry about underflowing the length, and to allow for WARNing on trying to configure the scratch area with len==0.

Fixes 291bd20d5d88 ("KVM: SVM: Add initial support for a VMGEXIT VMEXIT")

Message-ID: <[email protected]>

---
 arch/x86/kvm/svm/sev.c |    8 ++++++++
 1 file changed, 8 insertions(+)
arch/x86/kvm/svm/sev.c +8 -0
@@ -4545,6 +4545,11 @@ int sev_handle_vmgexit(struct kvm_vcpu *
 			    control->exit_info_1, control->exit_info_2);
 		ret = -EINVAL;
 		break;
+	case SVM_EXIT_IOIO:
+		if (!((control->exit_info_1 & SVM_IOIO_SIZE_MASK) >> SVM_IOIO_SIZE_SHIFT))
+			return 1;
+
+		fallthrough;
 	default:
 		ret = svm_invoke_exit_handler(vcpu, control->exit_code);
 	}
@@ -4565,6 +4570,9 @@ int sev_es_string_io(struct vcpu_svm *sv
 	if (unlikely(check_mul_overflow(count, size, &bytes)))
 		return -EINVAL;
 
+	if (!bytes)
+		return 1;
+
 	r = setup_vmgexit_scratch(svm, in, bytes);
 	if (r)
 		return r;
author maintainer
#177 lore raw

PATCH 164/332 KVM: SEV: Use the size of the PSC header as the minimum size for PSC requests

In reply to Greg Kroah-Hartman

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

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

From: Sean Christopherson <[email protected]>

commit 2be54670bdc017004c4a4b8bddb6ff02ebe7dbe2 upstream.

When handling a Page State Change (PSC) #VMGEXIT use the size of the PSC header as the minimum size for the scratch area. Per the GHCB spec, PSC requests do NOT provide the length, i.e. using control->exit_info_2 for the length is completely made up behavior. The existing code "works", e.g. even though Linux-as-a-guest always passes '0', because KVM doesn't do anything with the length when the request is in the GHCB's shared buffer.

Use the header as the min length. Once the header is retrieved, KVM can use the specified indices to compute the full size of the request.

Fixes 9b54e248d264 ("KVM: SEV: Add support to handle Page State Change VMGEXIT")

Message-ID: <[email protected]>

---
 arch/x86/kvm/svm/sev.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/x86/kvm/svm/sev.c +1 -1
@@ -4519,7 +4519,7 @@ int sev_handle_vmgexit(struct kvm_vcpu *
 		vcpu->run->system_event.data[0] = control->ghcb_gpa;
 		break;
 	case SVM_VMGEXIT_PSC:
-		ret = setup_vmgexit_scratch(svm, true, control->exit_info_2);
+		ret = setup_vmgexit_scratch(svm, true, sizeof(struct psc_hdr));
 		if (ret)
 			break;
 
author maintainer
#178 lore raw

PATCH 165/332 KVM: SEV: WARN if KVM attempts to setup scratch area with min_len==0

In reply to Greg Kroah-Hartman

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

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

From: Sean Christopherson <[email protected]>

commit f185e05dce6f170f83c4ba602e969b1c3c7a22e6 upstream.

Now that all paths in KVM properly validate the length needed for the scratch area, and are guaranteed to pass in a non-zero length, WARN if KVM attempts to configured the scratch area with min_len==0 to guard against future bugs.

Message-ID: <[email protected]>

---
 arch/x86/kvm/svm/sev.c |    3 +++
 1 file changed, 3 insertions(+)
arch/x86/kvm/svm/sev.c +3 -0
@@ -3624,6 +3624,9 @@ static int setup_vmgexit_scratch(struct
 	u64 scratch_gpa_beg, scratch_gpa_end;
 	void *scratch_va;
 
+	if (WARN_ON_ONCE(!min_len))
+		goto e_scratch;
+
 	scratch_gpa_beg = svm->sev_es.sw_scratch;
 	if (!scratch_gpa_beg) {
 		pr_err("vmgexit: scratch gpa not provided\n");
author maintainer
#179 lore raw

PATCH 166/332 KVM: SEV: Compute the correct max length of the in-GHCB scratch area

In reply to Greg Kroah-Hartman

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

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

From: Sean Christopherson <[email protected]>

commit 5867d7e202e09f037cefe77f7af4413c7c0fa088 upstream.

When setting the length of the GHCB scratch area, and the area is in the GHCB shared buffer, set the effective length of the scratch area to the max possible size given the start of the guest-provided pointer, and the end of the shared buffer.

The code was "fine" when first introduced, as KVM doesn't consult the length of the buffer when emulating MMIO, because the passed in @len always specifies the *max* size required. But for PSC requests, the incoming @len is just the minimum length (to process the header), and KVM needs to know the full size of the scratch area to avoid buffer overflows (spoiler alert).

Opportunistically rename @len => @min_len to better reflect its role.

Fixes 9b54e248d264 ("KVM: SEV: Add support to handle Page State Change VMGEXIT")

Message-ID: <[email protected]>

---
 arch/x86/kvm/svm/sev.c |   19 ++++++++++---------
 1 file changed, 10 insertions(+), 9 deletions(-)
arch/x86/kvm/svm/sev.c +10 -9
@@ -3617,7 +3617,7 @@ int pre_sev_run(struct vcpu_svm *svm, in
 }
 
 #define GHCB_SCRATCH_AREA_LIMIT		(16ULL * PAGE_SIZE)
-static int setup_vmgexit_scratch(struct vcpu_svm *svm, bool sync, u64 len)
+static int setup_vmgexit_scratch(struct vcpu_svm *svm, bool sync, u64 min_len)
 {
 	struct vmcb_control_area *control = &svm->vmcb->control;
 	u64 ghcb_scratch_beg, ghcb_scratch_end;
@@ -3633,10 +3633,10 @@ static int setup_vmgexit_scratch(struct
 		goto e_scratch;
 	}
 
-	scratch_gpa_end = scratch_gpa_beg + len;
+	scratch_gpa_end = scratch_gpa_beg + min_len;
 	if (scratch_gpa_end < scratch_gpa_beg) {
 		pr_err("vmgexit: scratch length (%#llx) not valid for scratch address (%#llx)\n",
-		       len, scratch_gpa_beg);
+		       min_len, scratch_gpa_beg);
 		goto e_scratch;
 	}
 
@@ -3660,6 +3660,8 @@ static int setup_vmgexit_scratch(struct
 
 		scratch_va = (void *)svm->sev_es.ghcb;
 		scratch_va += (scratch_gpa_beg - control->ghcb_gpa);
+
+		svm->sev_es.ghcb_sa_len = ghcb_scratch_end - scratch_gpa_beg;
 	} else {
 		/* GHCB v2 requires the scratch area to be within the GHCB. */
 		if (to_kvm_sev_info(svm->vcpu.kvm)->ghcb_version >= 2)
@@ -3669,16 +3671,16 @@ static int setup_vmgexit_scratch(struct
 		 * The guest memory must be read into a kernel buffer, so
 		 * limit the size
 		 */
-		if (len > GHCB_SCRATCH_AREA_LIMIT) {
+		if (min_len > GHCB_SCRATCH_AREA_LIMIT) {
 			pr_err("vmgexit: scratch area exceeds KVM limits (%#llx requested, %#llx limit)\n",
-			       len, GHCB_SCRATCH_AREA_LIMIT);
+			       min_len, GHCB_SCRATCH_AREA_LIMIT);
 			goto e_scratch;
 		}
-		scratch_va = kvzalloc(len, GFP_KERNEL_ACCOUNT);
+		scratch_va = kvzalloc(min_len, GFP_KERNEL_ACCOUNT);
 		if (!scratch_va)
 			return -ENOMEM;
 
-		if (kvm_read_guest(svm->vcpu.kvm, scratch_gpa_beg, scratch_va, len)) {
+		if (kvm_read_guest(svm->vcpu.kvm, scratch_gpa_beg, scratch_va, min_len)) {
 			/* Unable to copy scratch area from guest */
 			pr_err("vmgexit: kvm_read_guest for scratch area failed\n");
 
@@ -3694,11 +3696,10 @@ static int setup_vmgexit_scratch(struct
 		 */
 		svm->sev_es.ghcb_sa_sync = sync;
 		svm->sev_es.ghcb_sa_free = true;
+		svm->sev_es.ghcb_sa_len = min_len;
 	}
 
 	svm->sev_es.ghcb_sa = scratch_va;
-	svm->sev_es.ghcb_sa_len = len;
-
 	return 0;
 
 e_scratch:
author maintainer
#180 lore raw

PATCH 167/332 KVM: SEV: Check PSC request indices against the actual size of the buffer

In reply to Greg Kroah-Hartman

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

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

From: Sean Christopherson <[email protected]>

commit 121d88de56bc5c0ba0ce2f6381af67f948a7e7c1 upstream.

When processing Page State Change (PSC) requests, validate the PSC buffer against the effective size of the scratch area, which could be less than the maximum size if the guest provided a pointer that isn't exactly at the start of the GHCB shared buffer.

Fixes 9b54e248d264 ("KVM: SEV: Add support to handle Page State Change VMGEXIT")

Message-ID: <[email protected]>

---
 arch/x86/kvm/svm/sev.c |   17 +++++++++++++++--
 1 file changed, 15 insertions(+), 2 deletions(-)
arch/x86/kvm/svm/sev.c +15 -2
@@ -3857,7 +3857,7 @@ static int snp_begin_psc(struct vcpu_svm
 	struct kvm_vcpu *vcpu = &svm->vcpu;
 	struct psc_hdr *hdr = &psc->hdr;
 	struct psc_entry entry_start;
-	u16 idx, idx_start, idx_end;
+	u16 idx, idx_start, idx_end, max_nr_entries;
 	int npages;
 	bool huge;
 	u64 gfn;
@@ -3867,6 +3867,19 @@ static int snp_begin_psc(struct vcpu_svm
 		return 1;
 	}
 
+	/*
+	 * GHCB v2 requires the scratch area to reside within the GHCB itself,
+	 * and PSC requests are only supported for GHCB v2+.  Thus it should be
+	 * impossible to exceed the max PSC entry count (which is derived from
+	 * the size of the shared GHCB buffer).
+	 */
+	max_nr_entries = (sev_es->ghcb_sa_len - sizeof(struct psc_hdr)) /
+			 sizeof(struct psc_entry);
+	if (WARN_ON_ONCE(max_nr_entries > VMGEXIT_PSC_MAX_COUNT)) {
+		snp_complete_psc(svm, VMGEXIT_PSC_ERROR_GENERIC);
+		return 1;
+	}
+
 next_range:
 	/* There should be no other PSCs in-flight at this point. */
 	if (WARN_ON_ONCE(svm->sev_es.psc_inflight)) {
@@ -3882,7 +3895,7 @@ next_range:
 	idx_start = hdr->cur_entry;
 	idx_end = hdr->end_entry;
 
-	if (idx_end >= VMGEXIT_PSC_MAX_COUNT) {
+	if (idx_end >= max_nr_entries) {
 		snp_complete_psc(svm, VMGEXIT_PSC_ERROR_INVALID_HDR);
 		return 1;
 	}
author maintainer
#181 lore raw

PATCH 168/332 KVM: SEV: Use READ_ONCE() when reading entries/indices from PSC buffer

In reply to Greg Kroah-Hartman

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

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

From: Sean Christopherson <[email protected]>

commit c8cc238093ca6c99267032f6cfe78f59389f3157 upstream.

Use READ_ONCE() when reading entries/indices from the guest-accessible Page State Change buffer to defend against TOCTOU bugs.

Don't bother with READ_ONCE()/WRITE_ONCE() for cases where KVM is writing (and not consuming the result!), as the guest isn't supposed to touch the buffer while it's being processed. I.e. using READ_ONCE() is all about protecting against misbehaving guests.

Fixes 9b54e248d264 ("KVM: SEV: Add support to handle Page State Change VMGEXIT")

Message-ID: <[email protected]>

---
 arch/x86/kvm/svm/sev.c |   12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)
arch/x86/kvm/svm/sev.c +6 -6
@@ -3827,9 +3827,9 @@ static void __snp_complete_one_psc(struc
 	 */
 	for (idx = svm->sev_es.psc_idx; svm->sev_es.psc_inflight;
 	     svm->sev_es.psc_inflight--, idx++) {
-		struct psc_entry *entry = &entries[idx];
+		struct psc_entry entry = READ_ONCE(entries[idx]);
 
-		entry->cur_page = entry->pagesize ? 512 : 1;
+		entries[idx].cur_page = entry.pagesize ? 512 : 1;
 	}
 
 	hdr->cur_entry = idx;
@@ -3892,8 +3892,8 @@ next_range:
 	 * validation, so take care to only use validated copies of values used
 	 * for things like array indexing.
 	 */
-	idx_start = hdr->cur_entry;
-	idx_end = hdr->end_entry;
+	idx_start = READ_ONCE(hdr->cur_entry);
+	idx_end = READ_ONCE(hdr->end_entry);
 
 	if (idx_end >= max_nr_entries) {
 		snp_complete_psc(svm, VMGEXIT_PSC_ERROR_INVALID_HDR);
@@ -3902,7 +3902,7 @@ next_range:
 
 	/* Find the start of the next range which needs processing. */
 	for (idx = idx_start; idx <= idx_end; idx++, hdr->cur_entry++) {
-		entry_start = entries[idx];
+		entry_start = READ_ONCE(entries[idx]);
 
 		gfn = entry_start.gfn;
 		huge = entry_start.pagesize;
@@ -3946,7 +3946,7 @@ next_range:
 	 * KVM_HC_MAP_GPA_RANGE exit.
 	 */
 	while (++idx <= idx_end) {
-		struct psc_entry entry = entries[idx];
+		struct psc_entry entry = READ_ONCE(entries[idx]);
 
 		if (entry.operation != entry_start.operation ||
 		    entry.gfn != entry_start.gfn + npages ||
author maintainer
#182 lore raw

PATCH 169/332 KVM: SEV: Dont explicitly pass PSC buffer to snp_begin_psc()

In reply to Greg Kroah-Hartman

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

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

From: Sean Christopherson <[email protected]>

commit ebe4b2dc9cfbfb2d8f665667c4d08f4c6c9bec05 upstream.

Stop explicitly passing the PSC buffer to snp_begin_psc(): it *must* be the scratch area. This will allow fixing a variety of bugs without further complicating the code.

No functional change intended.

Message-ID: <[email protected]>

---
 arch/x86/kvm/svm/sev.c |   11 ++++++-----
 1 file changed, 6 insertions(+), 5 deletions(-)
arch/x86/kvm/svm/sev.c +6 -5
@@ -3796,7 +3796,7 @@ struct psc_buffer {
 	struct psc_entry entries[];
 } __packed;
 
-static int snp_begin_psc(struct vcpu_svm *svm, struct psc_buffer *psc);
+static int snp_begin_psc(struct vcpu_svm *svm);
 
 static void snp_complete_psc(struct vcpu_svm *svm, u64 psc_ret)
 {
@@ -3838,7 +3838,6 @@ static void __snp_complete_one_psc(struc
 static int snp_complete_one_psc(struct kvm_vcpu *vcpu)
 {
 	struct vcpu_svm *svm = to_svm(vcpu);
-	struct psc_buffer *psc = svm->sev_es.ghcb_sa;
 
 	if (vcpu->run->hypercall.ret) {
 		snp_complete_psc(svm, VMGEXIT_PSC_ERROR_GENERIC);
@@ -3848,11 +3847,13 @@ static int snp_complete_one_psc(struct k
 	__snp_complete_one_psc(svm);
 
 	/* Handle the next range (if any). */
-	return snp_begin_psc(svm, psc);
+	return snp_begin_psc(svm);
 }
 
-static int snp_begin_psc(struct vcpu_svm *svm, struct psc_buffer *psc)
+static int snp_begin_psc(struct vcpu_svm *svm)
 {
+	struct vcpu_sev_es_state *sev_es = &svm->sev_es;
+	struct psc_buffer *psc = sev_es->ghcb_sa;
 	struct psc_entry *entries = psc->entries;
 	struct kvm_vcpu *vcpu = &svm->vcpu;
 	struct psc_hdr *hdr = &psc->hdr;
@@ -4540,7 +4541,7 @@ int sev_handle_vmgexit(struct kvm_vcpu *
 		if (ret)
 			break;
 
-		ret = snp_begin_psc(svm, svm->sev_es.ghcb_sa);
+		ret = snp_begin_psc(svm);
 		break;
 	case SVM_VMGEXIT_AP_CREATION:
 		ret = sev_snp_ap_creation(svm);
author maintainer
#183 lore raw

PATCH 170/332 gpio: shared: undo the vote of the proxy on GPIO free

In reply to Greg Kroah-Hartman

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

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

From: Bartosz Golaszewski <[email protected]>

commit bbec30f7e19d9a1c604da7164b8057ccee590e72 upstream.

When the user of a shared GPIO managed by gpio-shared-proxy calls gpiod_put() to release it, we never undo the potential "vote" for driving the shared line "high". In the free() callback, check if this proxy voted for "high" and - if so - decrease the number of votes and potentially revert the value to low if this is the last user.

---
 drivers/gpio/gpio-shared-proxy.c |    9 +++++++++
 1 file changed, 9 insertions(+)
drivers/gpio/gpio-shared-proxy.c +9 -0
@@ -103,9 +103,18 @@ static void gpio_shared_proxy_free(struc
 {
 	struct gpio_shared_proxy_data *proxy = gpiochip_get_data(gc);
 	struct gpio_shared_desc *shared_desc = proxy->shared_desc;
+	int ret;
 
 	guard(gpio_shared_desc_lock)(shared_desc);
 
+	if (proxy->voted_high) {
+		ret = gpio_shared_proxy_set_unlocked(proxy,
+			shared_desc->can_sleep ? gpiod_set_value_cansleep : gpiod_set_value, 0);
+		if (ret)
+			dev_err(proxy->dev,
+				"Failed to unset the shared GPIO value on release: %d\n", ret);
+	}
+
 	proxy->shared_desc->usecnt--;
 
 	dev_dbg(proxy->dev, "Shared GPIO freed, number of users: %u\n",
author maintainer
#184 lore raw

PATCH 171/332 gpio: shared: fix deadlock on shared proxys parent removal

In reply to Greg Kroah-Hartman

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

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

From: Bartosz Golaszewski <[email protected]>

commit a1b836607304f71051f9f9dcccf8b5097b86a1fb upstream.

Commit 710abda58055 ("gpio: shared: call gpio_chip::of_xlate() if set") used the mutex embedded in struct gpio_shared_entry to protect the offset field which now can be modified after assignment. The critical section however is too wide and introduced a potential deadlock on the removal of the shared GPIO proxy's parent.

Make the critical section shorter - only protect the offset when it's being read.

While at it: mention the fact that the entry lock is now also used to protect against concurrent access to the offset field in the structure's documentation.

---
 drivers/gpio/gpiolib-shared.c | 7 +++----
 1 file changed, 3 insertions(+), 4 deletions(-)
drivers/gpio/gpiolib-shared.c +3 -4
index e02d6b93a4ab..087b64c06c9f 100644
@@ -53,7 +53,7 @@ struct gpio_shared_entry {
 	unsigned int offset;
 	/* Index in the property value array. */
 	size_t index;
-	/* Synchronizes the modification of shared_desc. */
+	/* Synchronizes the modification of shared_desc and offset. */
 	struct mutex lock;
 	struct gpio_shared_desc *shared_desc;
 	struct kref ref;
@@ -598,12 +598,11 @@ void gpio_device_teardown_shared(struct gpio_device *gdev)
 	struct gpio_shared_ref *ref;
 
 	list_for_each_entry(entry, &gpio_shared_list, list) {
-		guard(mutex)(&entry->lock);
-
 		if (!device_match_fwnode(&gdev->dev, entry->fwnode))
 			continue;
 
-		gpiod_free_commit(&gdev->descs[entry->offset]);
+		scoped_guard(mutex, &entry->lock)
+			gpiod_free_commit(&gdev->descs[entry->offset]);
 
 		list_for_each_entry(ref, &entry->refs, list) {
 			guard(mutex)(&ref->lock);
Signature · 1 line
2.54.0
author maintainer
#185 lore raw

PATCH 172/332 gpio: shared: fix lockdep false positive by removing unneeded lock

In reply to Greg Kroah-Hartman

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

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

From: Bartosz Golaszewski <[email protected]>

commit 9d7697fabbc72428f981c01ddbe0a6be0ce8b6fa upstream.

By the time gpio_device_teardown_shared() is called, the parent device is gone from the global list of GPIO devices and all outstanding SRCU read-side critical sections have completed. That means that no concurrent gpio_find_and_request() can call gpio_shared_add_proxy_lookup() for this device at this time. There's also no risk of the parent device being re-bound to the driver before the unbinding completes (including the child devices).

Lockdep produces a false-positive report about a possible circular dependency as it doesn't know the ordering guarantee. Not taking the ref->lock in gpio_device_teardown_shared() silences it and is safe to do.

---
 drivers/gpio/gpiolib-shared.c |    2 --
 1 file changed, 2 deletions(-)
drivers/gpio/gpiolib-shared.c +0 -2
@@ -605,8 +605,6 @@ void gpio_device_teardown_shared(struct
 			gpiod_free_commit(&gdev->descs[entry->offset]);
 
 		list_for_each_entry(ref, &entry->refs, list) {
-			guard(mutex)(&ref->lock);
-
 			if (ref->lookup) {
 				gpiod_remove_lookup_table(ref->lookup);
 				kfree(ref->lookup->table[0].key);
author maintainer
#186 lore raw

PATCH 173/332 Disable -Wattribute-alias for clang-23 and newer

In reply to Greg Kroah-Hartman

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

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

From: Nathan Chancellor <[email protected]>

commit 175db11786bde9061db526bf1ac5107d915f5163 upstream.

Clang recently added support for -Wattribute-alias [1], which results in the same warnings that necessitated commit bee20031772a ("disable -Wattribute-alias warning for SYSCALL_DEFINEx()") for GCC.

kernel/time/itimer.c:325:1: error: alias and aliasee have different types 'long (unsigned int)' and 'long (typeof (__builtin_choose_expr((__builtin_types_compatible_p(typeof ((unsigned int)0), typeof (0LL)) || __builtin_types_compatible_p(typeof ((unsigned int)0), typeof (0ULL))), 0LL, 0L)))' (aka 'long (long)') [-Werror,-Wattribute-alias]
325 | SYSCALL_DEFINE1(alarm, unsigned int, seconds)
| ^
include/linux/syscalls.h:225:36: note: expanded from macro 'SYSCALL_DEFINE1'

    225 | #define SYSCALL_DEFINE1(name, ...) SYSCALL_DEFINEx(1, _##name, __VA_ARGS__)
        |                                    ^
  include/linux/syscalls.h:236:2: note: expanded from macro 'SYSCALL_DEFINEx'
    236 |         __SYSCALL_DEFINEx(x, sname, __VA_ARGS__)
        |         ^
  include/linux/syscalls.h:251:18: note: expanded from macro '__SYSCALL_DEFINEx'
    251 |                 __attribute__((alias(__stringify(__se_sys##name))));    \
        |                                ^
  kernel/time/itimer.c:325:1: note: aliasee is declared here
  include/linux/syscalls.h:225:36: note: expanded from macro 'SYSCALL_DEFINE1'
    225 | #define SYSCALL_DEFINE1(name, ...) SYSCALL_DEFINEx(1, _##name, __VA_ARGS__)
        |                                    ^
  include/linux/syscalls.h:236:2: note: expanded from macro 'SYSCALL_DEFINEx'
    236 |         __SYSCALL_DEFINEx(x, sname, __VA_ARGS__)
        |         ^
  include/linux/syscalls.h:255:18: note: expanded from macro '__SYSCALL_DEFINEx'
    255 |         asmlinkage long __se_sys##name(__MAP(x,__SC_LONG,__VA_ARGS__))  \
        |                         ^
  <scratch space>:16:1: note: expanded from here
     16 | __se_sys_alarm
        | ^

Disable the warnings in the same way for clang-23 and newer. Disable the warning about unknown warning options to avoid breaking the build for versions of clang-23 that do not have -Wattribute-alias, such as ones deployed by vendors like Android or CI systems or when bisecting LLVM between llvmorg-23-init and release/23.x.

---
 arch/riscv/include/asm/syscall_wrapper.h |    4 ++++
 include/linux/compat.h                   |    4 ++++
 include/linux/compiler-clang.h           |    6 ++++++
 include/linux/compiler_types.h           |    4 ++++
 include/linux/syscalls.h                 |    4 ++++
 5 files changed, 22 insertions(+)
arch/riscv/include/asm/syscall_wrapper.h +4 -0
@@ -32,6 +32,10 @@ asmlinkage long __riscv_sys_ni_syscall(c
 	__diag_push();									\
 	__diag_ignore(GCC, 8, "-Wattribute-alias",					\
 			"Type aliasing is used to sanitize syscall arguments");		\
+	__diag_ignore(clang, 23, "-Wunknown-warning-option",				\
+		      "Avoid breaking versions without -Wattribute-alias");		\
+	__diag_ignore(clang, 23, "-Wattribute-alias",					\
+			"Type aliasing is used to sanitize syscall arguments");		\
 	static long __se_##prefix##name(ulong, ulong, ulong, ulong, ulong, ulong, 	\
 					ulong)						\
 			__attribute__((alias(__stringify(___se_##prefix##name))));	\
include/linux/compat.h +4 -0
@@ -72,6 +72,10 @@
 	__diag_push();								\
 	__diag_ignore(GCC, 8, "-Wattribute-alias",				\
 		      "Type aliasing is used to sanitize syscall arguments");\
+	__diag_ignore(clang, 23, "-Wunknown-warning-option",			\
+		      "Avoid breaking versions without -Wattribute-alias");	\
+	__diag_ignore(clang, 23, "-Wattribute-alias",				\
+		      "Type aliasing is used to sanitize syscall arguments");	\
 	asmlinkage long compat_sys##name(__MAP(x,__SC_DECL,__VA_ARGS__))	\
 		__attribute__((alias(__stringify(__se_compat_sys##name))));	\
 	ALLOW_ERROR_INJECTION(compat_sys##name, ERRNO);				\
include/linux/compiler-clang.h +6 -0
@@ -131,6 +131,12 @@
 #define __diag_str(s)		__diag_str1(s)
 #define __diag(s)		_Pragma(__diag_str(clang diagnostic s))
 
+#if CONFIG_CLANG_VERSION >= 230000
+#define __diag_clang_23(s)	__diag(s)
+#else
+#define __diag_clang_23(s)
+#endif
+
 #define __diag_clang_13(s)	__diag(s)
 
 #define __diag_ignore_all(option, comment) \
include/linux/compiler_types.h +4 -0
@@ -718,6 +718,10 @@ struct ftrace_likely_data {
 #define __diag_GCC(version, severity, string)
 #endif
 
+#ifndef __diag_clang
+#define __diag_clang(version, severity, string)
+#endif
+
 #define __diag_push()	__diag(push)
 #define __diag_pop()	__diag(pop)
 
include/linux/syscalls.h +4 -0
@@ -247,6 +247,10 @@ static inline int is_syscall_trace_event
 	__diag_push();							\
 	__diag_ignore(GCC, 8, "-Wattribute-alias",			\
 		      "Type aliasing is used to sanitize syscall arguments");\
+	__diag_ignore(clang, 23, "-Wunknown-warning-option",		\
+		      "Avoid breaking versions without -Wattribute-alias");\
+	__diag_ignore(clang, 23, "-Wattribute-alias",			\
+		      "Type aliasing is used to sanitize syscall arguments");\
 	asmlinkage long sys##name(__MAP(x,__SC_DECL,__VA_ARGS__))	\
 		__attribute__((alias(__stringify(__se_sys##name))));	\
 	ALLOW_ERROR_INJECTION(sys##name, ERRNO);			\
author maintainer
#187 lore raw

PATCH 174/332 iio: adc: xilinx-xadc: Fix sequencer mode in postdisable for dual mux

In reply to Greg Kroah-Hartman

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

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

From: Christofer Jonason <[email protected]>

commit 852534744c2d35626a604f128ff0b8ec12805591 upstream.

xadc_postdisable() unconditionally sets the sequencer to continuous mode. For dual external multiplexer configurations this is incorrect:
simultaneous sampling mode is required so that ADC-A samples through the mux on VAUX[0-7] while ADC-B simultaneously samples through the mux on VAUX[8-15]. In continuous mode only ADC-A is active, so VAUX[8-15] channels return incorrect data.

Since postdisable is also called from xadc_probe() to set the initial idle state, the wrong sequencer mode is active from the moment the driver loads.

The preenable path already uses xadc_get_seq_mode() which returns SIMULTANEOUS for dual mux. Fix postdisable to do the same.

---
 drivers/iio/adc/xilinx-xadc-core.c |   11 +++++++++--
 1 file changed, 9 insertions(+), 2 deletions(-)
drivers/iio/adc/xilinx-xadc-core.c +9 -2
@@ -817,6 +817,7 @@ static int xadc_postdisable(struct iio_d
 {
 	struct xadc *xadc = iio_priv(indio_dev);
 	unsigned long scan_mask;
+	int seq_mode;
 	int ret;
 	int i;
 
@@ -824,6 +825,12 @@ static int xadc_postdisable(struct iio_d
 	for (i = 0; i < indio_dev->num_channels; i++)
 		scan_mask |= BIT(indio_dev->channels[i].scan_index);
 
+	/*
+	 * Use the correct sequencer mode for the idle state: simultaneous
+	 * mode for dual external mux configurations, continuous otherwise.
+	 */
+	seq_mode = xadc_get_seq_mode(xadc, scan_mask);
+
 	/* Enable all channels and calibration */
 	ret = xadc_write_adc_reg(xadc, XADC_REG_SEQ(0), scan_mask & 0xffff);
 	if (ret)
@@ -834,11 +841,11 @@ static int xadc_postdisable(struct iio_d
 		return ret;
 
 	ret = xadc_update_adc_reg(xadc, XADC_REG_CONF1, XADC_CONF1_SEQ_MASK,
-		XADC_CONF1_SEQ_CONTINUOUS);
+				  seq_mode);
 	if (ret)
 		return ret;
 
-	return xadc_power_adc_b(xadc, XADC_CONF1_SEQ_CONTINUOUS);
+	return xadc_power_adc_b(xadc, seq_mode);
 }
 
 static int xadc_preenable(struct iio_dev *indio_dev)
author maintainer
#188 lore raw

PATCH 175/332 iio: adc: npcm: fix unbalanced clk_disable_unprepare()

In reply to Greg Kroah-Hartman

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

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

From: David Carlier <[email protected]>

commit 0d42e2c0bd6ceb89e44c6e065f9bdf9b1df3ef0c upstream.

The driver acquired the ADC clock with devm_clk_get() and read its rate, but never called clk_prepare_enable(). The probe error path and npcm_adc_remove() both called clk_disable_unprepare() unconditionally, causing the clk framework's enable/prepare counts to underflow on probe failure or module unbind.

The issue went unnoticed because NPCM BMC firmware leaves the ADC clock enabled at boot, so the driver happened to work in practice.

Switch to devm_clk_get_enabled() so the clock is properly enabled during probe and automatically released by the device-managed cleanup, and drop the now-redundant clk_disable_unprepare() from both the probe error path and remove().

While at it, drop the duplicate error message on devm_request_irq() failure since the IRQ core already logs it.

Fixes 9bf85fbc9d8f ("iio: adc: add NPCM ADC driver")
---
 drivers/iio/adc/npcm_adc.c |   25 ++++++++-----------------
 1 file changed, 8 insertions(+), 17 deletions(-)
drivers/iio/adc/npcm_adc.c +8 -17
@@ -231,7 +231,7 @@ static int npcm_adc_probe(struct platfor
 	if (IS_ERR(info->reset))
 		return PTR_ERR(info->reset);
 
-	info->adc_clk = devm_clk_get(&pdev->dev, NULL);
+	info->adc_clk = devm_clk_get_enabled(&pdev->dev, NULL);
 	if (IS_ERR(info->adc_clk)) {
 		dev_warn(&pdev->dev, "ADC clock failed: can't read clk\n");
 		return PTR_ERR(info->adc_clk);
@@ -244,17 +244,13 @@ static int npcm_adc_probe(struct platfor
 	info->adc_sample_hz = clk_get_rate(info->adc_clk) / ((div + 1) * 2);
 
 	irq = platform_get_irq(pdev, 0);
-	if (irq < 0) {
-		ret = irq;
-		goto err_disable_clk;
-	}
+	if (irq < 0)
+		return irq;
 
 	ret = devm_request_irq(&pdev->dev, irq, npcm_adc_isr, 0,
 			       "NPCM_ADC", indio_dev);
-	if (ret < 0) {
-		dev_err(dev, "failed requesting interrupt\n");
-		goto err_disable_clk;
-	}
+	if (ret < 0)
+		return ret;
 
 	reg_con = ioread32(info->regs + NPCM_ADCCON);
 	info->vref = devm_regulator_get_optional(&pdev->dev, "vref");
@@ -262,7 +258,7 @@ static int npcm_adc_probe(struct platfor
 		ret = regulator_enable(info->vref);
 		if (ret) {
 			dev_err(&pdev->dev, "Can't enable ADC reference voltage\n");
-			goto err_disable_clk;
+			return ret;
 		}
 
 		iowrite32(reg_con & ~NPCM_ADCCON_REFSEL,
@@ -272,10 +268,8 @@ static int npcm_adc_probe(struct platfor
 		 * Any error which is not ENODEV indicates the regulator
 		 * has been specified and so is a failure case.
 		 */
-		if (PTR_ERR(info->vref) != -ENODEV) {
-			ret = PTR_ERR(info->vref);
-			goto err_disable_clk;
-		}
+		if (PTR_ERR(info->vref) != -ENODEV)
+			return PTR_ERR(info->vref);
 
 		/* Use internal reference */
 		iowrite32(reg_con | NPCM_ADCCON_REFSEL,
@@ -314,8 +308,6 @@ err_iio_register:
 	iowrite32(reg_con & ~NPCM_ADCCON_ADC_EN, info->regs + NPCM_ADCCON);
 	if (!IS_ERR(info->vref))
 		regulator_disable(info->vref);
-err_disable_clk:
-	clk_disable_unprepare(info->adc_clk);
 
 	return ret;
 }
@@ -332,7 +324,6 @@ static void npcm_adc_remove(struct platf
 	iowrite32(regtemp & ~NPCM_ADCCON_ADC_EN, info->regs + NPCM_ADCCON);
 	if (!IS_ERR(info->vref))
 		regulator_disable(info->vref);
-	clk_disable_unprepare(info->adc_clk);
 }
 
 static struct platform_driver npcm_adc_driver = {
author maintainer
#189 lore raw

PATCH 176/332 iio: dac: ad3530r: Fix AD3531/AD3531R powerdown mode strings

In reply to Greg Kroah-Hartman

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

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

From: Kim Seer Paller <[email protected]>

commit ebd250c2581ec46c64c73fdfa918c9a7f757505e upstream.

The AD3531/AD3531R has different output operating modes from the AD3530/AD3530R. According to the AD3531/AD3531R datasheet, the powerdown modes are:
01: 500 Ohm output impedance
10: 3.85 kOhm output impedance
11: 16 kOhm output impedance

The driver currently uses the AD3530R modes (1k, 7.7k, 32k) for all variants, which is incorrect for AD3531/AD3531R.

Add AD3531R-specific powerdown mode strings and assign them to the AD3531/AD3531R chip variants.

Fixes 93583174a3df ("iio: dac: ad3530r: Add driver for AD3530R and AD3531R")
---
 drivers/iio/dac/ad3530r.c |   54 ++++++++++++++++++++++++++++++++++------------
 1 file changed, 40 insertions(+), 14 deletions(-)
drivers/iio/dac/ad3530r.c +40 -14
@@ -105,6 +105,12 @@ static const char * const ad3530r_powerd
 	"32kohm_to_gnd",
 };
 
+static const char * const ad3531r_powerdown_modes[] = {
+	"500ohm_to_gnd",
+	"3.85kohm_to_gnd",
+	"16kohm_to_gnd",
+};
+
 static int ad3530r_get_powerdown_mode(struct iio_dev *indio_dev,
 				      const struct iio_chan_spec *chan)
 {
@@ -133,6 +139,13 @@ static const struct iio_enum ad3530r_pow
 	.set = ad3530r_set_powerdown_mode,
 };
 
+static const struct iio_enum ad3531r_powerdown_mode_enum = {
+	.items = ad3531r_powerdown_modes,
+	.num_items = ARRAY_SIZE(ad3531r_powerdown_modes),
+	.get = ad3530r_get_powerdown_mode,
+	.set = ad3530r_set_powerdown_mode,
+};
+
 static ssize_t ad3530r_get_dac_powerdown(struct iio_dev *indio_dev,
 					 uintptr_t private,
 					 const struct iio_chan_spec *chan,
@@ -276,7 +289,20 @@ static const struct iio_chan_spec_ext_in
 	{ }
 };
 
-#define AD3530R_CHAN(_chan)					\
+static const struct iio_chan_spec_ext_info ad3531r_ext_info[] = {
+	{
+		.name = "powerdown",
+		.shared = IIO_SEPARATE,
+		.read = ad3530r_get_dac_powerdown,
+		.write = ad3530r_set_dac_powerdown,
+	},
+	IIO_ENUM("powerdown_mode", IIO_SEPARATE, &ad3531r_powerdown_mode_enum),
+	IIO_ENUM_AVAILABLE("powerdown_mode", IIO_SHARED_BY_TYPE,
+			   &ad3531r_powerdown_mode_enum),
+	{ }
+};
+
+#define AD3530R_CHAN(_chan, _ext_info)				\
 {								\
 	.type = IIO_VOLTAGE,					\
 	.indexed = 1,						\
@@ -284,25 +310,25 @@ static const struct iio_chan_spec_ext_in
 	.output = 1,						\
 	.info_mask_separate = BIT(IIO_CHAN_INFO_RAW) |		\
 			      BIT(IIO_CHAN_INFO_SCALE),		\
-	.ext_info = ad3530r_ext_info,				\
+	.ext_info = _ext_info,					\
 }
 
 static const struct iio_chan_spec ad3530r_channels[] = {
-	AD3530R_CHAN(0),
-	AD3530R_CHAN(1),
-	AD3530R_CHAN(2),
-	AD3530R_CHAN(3),
-	AD3530R_CHAN(4),
-	AD3530R_CHAN(5),
-	AD3530R_CHAN(6),
-	AD3530R_CHAN(7),
+	AD3530R_CHAN(0, ad3530r_ext_info),
+	AD3530R_CHAN(1, ad3530r_ext_info),
+	AD3530R_CHAN(2, ad3530r_ext_info),
+	AD3530R_CHAN(3, ad3530r_ext_info),
+	AD3530R_CHAN(4, ad3530r_ext_info),
+	AD3530R_CHAN(5, ad3530r_ext_info),
+	AD3530R_CHAN(6, ad3530r_ext_info),
+	AD3530R_CHAN(7, ad3530r_ext_info),
 };
 
 static const struct iio_chan_spec ad3531r_channels[] = {
-	AD3530R_CHAN(0),
-	AD3530R_CHAN(1),
-	AD3530R_CHAN(2),
-	AD3530R_CHAN(3),
+	AD3530R_CHAN(0, ad3531r_ext_info),
+	AD3530R_CHAN(1, ad3531r_ext_info),
+	AD3530R_CHAN(2, ad3531r_ext_info),
+	AD3530R_CHAN(3, ad3531r_ext_info),
 };
 
 static const struct ad3530r_chip_info ad3530_chip = {
author maintainer
#190 lore raw

PATCH 177/332 iio: dac: max5821: fix return value check in powerdown sync

In reply to Greg Kroah-Hartman

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

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

From: Salah Triki <[email protected]>

commit d0a228d903425e653f18a4341e60c0538afb6d41 upstream.

The function max5821_sync_powerdown_mode() returned the result of i2c_master_send() directly. If a partial transfer occurred, it would be incorrectly treated as a success by the caller.

While the caller currently handles the positive return value of 2 as success, this patch refactors the function to return 0 on full success and -EIO on short writes. This ensures robust error handling for incomplete transfers and improves code maintainability by using sizeof(outbuf).

Fixes 472988972737 ("iio: add support of the max5821")
---
 drivers/iio/dac/max5821.c |    9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)
drivers/iio/dac/max5821.c +8 -1
@@ -90,6 +90,7 @@ static int max5821_sync_powerdown_mode(s
 				       const struct iio_chan_spec *chan)
 {
 	u8 outbuf[2];
+	int ret;
 
 	outbuf[0] = MAX5821_EXTENDED_COMMAND_MODE;
 
@@ -103,7 +104,13 @@ static int max5821_sync_powerdown_mode(s
 	else
 		outbuf[1] |= MAX5821_EXTENDED_POWER_UP;
 
-	return i2c_master_send(data->client, outbuf, 2);
+	ret = i2c_master_send(data->client, outbuf, sizeof(outbuf));
+	if (ret < 0)
+		return ret;
+	if (ret != sizeof(outbuf))
+		return -EIO;
+
+	return 0;
 }
 
 static ssize_t max5821_write_dac_powerdown(struct iio_dev *indio_dev,
author maintainer
#191 lore raw

PATCH 178/332 iio: dac: ad5686: fix ref bit initialization for single-channel parts

In reply to Greg Kroah-Hartman

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

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

From: Rodrigo Alencar <[email protected]>

commit ecae2ae606d493cf11457946436335bd0e726663 upstream.

The reference bit position was ignored when writing the register at the probe() function (!!val was used). When such bit is 1, internal voltage reference is disabled so that an external one can be used. For multi-channel devices, bit 0 of the Internal Reference Setup command behaves the same way, so AD5686_REF_BIT_MSK is created. The issue exists since support for single-channel devices were first introduced.

Fixes be1b24d24541 ("iio:dac:ad5686: Add AD5691R/AD5692R/AD5693/AD5693R support")
---
 drivers/iio/dac/ad5686.c |    6 +++---
 drivers/iio/dac/ad5686.h |    1 +
 2 files changed, 4 insertions(+), 3 deletions(-)
drivers/iio/dac/ad5686.c +3 -3
@@ -509,7 +509,7 @@ int ad5686_probe(struct device *dev,
 		break;
 	case AD5686_REGMAP:
 		cmd = AD5686_CMD_INTERNAL_REFER_SETUP;
-		ref_bit_msk = 0;
+		ref_bit_msk = AD5686_REF_BIT_MSK;
 		break;
 	case AD5693_REGMAP:
 		cmd = AD5686_CMD_CONTROL_REG;
@@ -520,9 +520,9 @@ int ad5686_probe(struct device *dev,
 		return -EINVAL;
 	}
 
-	val = (has_external_vref | ref_bit_msk);
+	val = has_external_vref ? ref_bit_msk : 0;
 
-	ret = st->write(st, cmd, 0, !!val);
+	ret = st->write(st, cmd, 0, val);
 	if (ret)
 		return ret;
 
drivers/iio/dac/ad5686.h +1 -0
@@ -46,6 +46,7 @@
 
 #define AD5310_REF_BIT_MSK			BIT(8)
 #define AD5683_REF_BIT_MSK			BIT(12)
+#define AD5686_REF_BIT_MSK			BIT(0)
 #define AD5693_REF_BIT_MSK			BIT(12)
 
 /**
author maintainer
#192 lore raw

PATCH 179/332 iio: dac: ad5686: fix input raw value check

In reply to Greg Kroah-Hartman

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

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

From: Rodrigo Alencar <[email protected]>

commit d01220ee5e43c65a206df827b39bf5cf5f7b9dce upstream.

Fix range check for input raw value, which is off by one, i.e., for a 10-bit DAC the max valid value is 1023, but 1 << 10 equals 1024, which passes the previous check, allowing an out-of-range write. The issue exists since the ad5686 driver was first introduced.

Fixes c2f37c8dcadc ("iio: dac: New driver for AD5686R, AD5685R, AD5684R Digital to analog converters")
---
 drivers/iio/dac/ad5686.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/iio/dac/ad5686.c +1 -1
@@ -154,7 +154,7 @@ static int ad5686_write_raw(struct iio_d
 
 	switch (mask) {
 	case IIO_CHAN_INFO_RAW:
-		if (val > (1 << chan->scan_type.realbits) || val < 0)
+		if (val >= (1 << chan->scan_type.realbits) || val < 0)
 			return -EINVAL;
 
 		mutex_lock(&st->lock);
author maintainer
#193 lore raw

PATCH 180/332 iio: dac: ad5686: acquire lock when doing powerdown control

In reply to Greg Kroah-Hartman

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

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

From: Rodrigo Alencar <[email protected]>

commit 5237c3175cae5ab05f18878cec3301a04403859e upstream.

Protect access of pwr_down_mode and pwr_down_mask fields with existing mutex lock. Each channel exposes their own attributes for controlling powerdown modes and powerdown state. This fixes potential race conditions as those the write functions perform non-atomic read-modify-write operations to those pwr_down_* fields. This issue exists since the ad5686 driver was first introduced.

Fixes c2f37c8dcadc ("iio: dac: New driver for AD5686R, AD5685R, AD5684R Digital to analog converters")
---
 drivers/iio/dac/ad5686.c |    8 ++++++++
 1 file changed, 8 insertions(+)
drivers/iio/dac/ad5686.c +8 -0
@@ -30,6 +30,8 @@ static int ad5686_get_powerdown_mode(str
 {
 	struct ad5686_state *st = iio_priv(indio_dev);
 
+	guard(mutex)(&st->lock);
+
 	return ((st->pwr_down_mode >> (chan->channel * 2)) & 0x3) - 1;
 }
 
@@ -39,6 +41,8 @@ static int ad5686_set_powerdown_mode(str
 {
 	struct ad5686_state *st = iio_priv(indio_dev);
 
+	guard(mutex)(&st->lock);
+
 	st->pwr_down_mode &= ~(0x3 << (chan->channel * 2));
 	st->pwr_down_mode |= ((mode + 1) << (chan->channel * 2));
 
@@ -57,6 +61,8 @@ static ssize_t ad5686_read_dac_powerdown
 {
 	struct ad5686_state *st = iio_priv(indio_dev);
 
+	guard(mutex)(&st->lock);
+
 	return sysfs_emit(buf, "%d\n", !!(st->pwr_down_mask &
 				       (0x3 << (chan->channel * 2))));
 }
@@ -77,6 +83,8 @@ static ssize_t ad5686_write_dac_powerdow
 	if (ret)
 		return ret;
 
+	guard(mutex)(&st->lock);
+
 	if (readin)
 		st->pwr_down_mask |= (0x3 << (chan->channel * 2));
 	else
author maintainer
#194 lore raw

PATCH 181/332 iio: dac: ad5686: fix powerdown control on dual-channel devices

In reply to Greg Kroah-Hartman

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

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

From: Rodrigo Alencar <[email protected]>

commit 8aeaf25a85263a7a43357e16ad78ab969f6f8aeb upstream.

Fix powerdown control by using a proper bit shift for the powerdown mask values. During initialization, powerdown bits are initialized so that unused bits are set to 1 and the correct bit shift is used. Dual-channel devices use one-hot encoding in the address and that reflects on the position of the powerdown bits, which are not channel-index based for that case. Quad-channel devices also use one-hot encoding for the channel address but the result of log2(address) coincides with the channel index value. Mask as 0x3U is used rather than 0x3, because shift can reach value of 30 (last channel of a 16-channel device), which would mess with the sign bit. The issue was introduced when first adding support for dual-channel devices, which overlooked powerdown control differences.

Fixes 7dc8faeab3e3 ("iio: dac: ad5686: add support for AD5338R")
---
 drivers/iio/dac/ad5686.c |   40 ++++++++++++++++++++++++++++++----------
 1 file changed, 30 insertions(+), 10 deletions(-)
drivers/iio/dac/ad5686.c +30 -10
@@ -25,26 +25,37 @@ static const char * const ad5686_powerdo
 	"three_state"
 };
 
+static inline unsigned int ad5686_pd_mask_shift(const struct iio_chan_spec *chan)
+{
+	if (chan->channel == chan->address)
+		return chan->channel * 2;
+
+	/* one-hot encoding is used in dual/quad channel devices */
+	return __ffs(chan->address) * 2;
+}
+
 static int ad5686_get_powerdown_mode(struct iio_dev *indio_dev,
 				     const struct iio_chan_spec *chan)
 {
+	unsigned int shift = ad5686_pd_mask_shift(chan);
 	struct ad5686_state *st = iio_priv(indio_dev);
 
 	guard(mutex)(&st->lock);
 
-	return ((st->pwr_down_mode >> (chan->channel * 2)) & 0x3) - 1;
+	return ((st->pwr_down_mode >> shift) & 0x3U) - 1;
 }
 
 static int ad5686_set_powerdown_mode(struct iio_dev *indio_dev,
 				     const struct iio_chan_spec *chan,
 				     unsigned int mode)
 {
+	unsigned int shift = ad5686_pd_mask_shift(chan);
 	struct ad5686_state *st = iio_priv(indio_dev);
 
 	guard(mutex)(&st->lock);
 
-	st->pwr_down_mode &= ~(0x3 << (chan->channel * 2));
-	st->pwr_down_mode |= ((mode + 1) << (chan->channel * 2));
+	st->pwr_down_mode &= ~(0x3U << shift);
+	st->pwr_down_mode |= (mode + 1) << shift;
 
 	return 0;
 }
@@ -59,12 +70,12 @@ static const struct iio_enum ad5686_powe
 static ssize_t ad5686_read_dac_powerdown(struct iio_dev *indio_dev,
 		uintptr_t private, const struct iio_chan_spec *chan, char *buf)
 {
+	unsigned int shift = ad5686_pd_mask_shift(chan);
 	struct ad5686_state *st = iio_priv(indio_dev);
 
 	guard(mutex)(&st->lock);
 
-	return sysfs_emit(buf, "%d\n", !!(st->pwr_down_mask &
-				       (0x3 << (chan->channel * 2))));
+	return sysfs_emit(buf, "%d\n", !!(st->pwr_down_mask & (0x3U << shift)));
 }
 
 static ssize_t ad5686_write_dac_powerdown(struct iio_dev *indio_dev,
@@ -86,9 +97,9 @@ static ssize_t ad5686_write_dac_powerdow
 	guard(mutex)(&st->lock);
 
 	if (readin)
-		st->pwr_down_mask |= (0x3 << (chan->channel * 2));
+		st->pwr_down_mask |= 0x3U << ad5686_pd_mask_shift(chan);
 	else
-		st->pwr_down_mask &= ~(0x3 << (chan->channel * 2));
+		st->pwr_down_mask &= ~(0x3U << ad5686_pd_mask_shift(chan));
 
 	switch (st->chip_info->regmap_type) {
 	case AD5310_REGMAP:
@@ -468,7 +479,7 @@ int ad5686_probe(struct device *dev,
 {
 	struct ad5686_state *st;
 	struct iio_dev *indio_dev;
-	unsigned int val, ref_bit_msk;
+	unsigned int val, ref_bit_msk, shift;
 	bool has_external_vref;
 	u8 cmd;
 	int ret, i;
@@ -492,9 +503,18 @@ int ad5686_probe(struct device *dev,
 	has_external_vref = ret != -ENODEV;
 	st->vref_mv = has_external_vref ? ret / 1000 : st->chip_info->int_vref_mv;
 
+	/* Initialize masks to all ones provided the max shift (last channel) */
+	shift = ad5686_pd_mask_shift(&st->chip_info->channels[st->chip_info->num_channels - 1]);
+	st->pwr_down_mask = GENMASK(shift + 1, 0);
+	st->pwr_down_mode = GENMASK(shift + 1, 0);
+
 	/* Set all the power down mode for all channels to 1K pulldown */
-	for (i = 0; i < st->chip_info->num_channels; i++)
-		st->pwr_down_mode |= (0x01 << (i * 2));
+	for (i = 0; i < st->chip_info->num_channels; i++) {
+		shift = ad5686_pd_mask_shift(&st->chip_info->channels[i]);
+		st->pwr_down_mask &= ~(0x3U << shift); /* powered up state */
+		st->pwr_down_mode &= ~(0x3U << shift);
+		st->pwr_down_mode |= 0x01U << shift;
+	}
 
 	indio_dev->name = name;
 	indio_dev->info = &ad5686_info;
author maintainer
#195 lore raw

PATCH 182/332 iio: adc: mt6359: fix unchecked return value in mt6358_read_imp

In reply to Greg Kroah-Hartman

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

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

From: Salah Triki <[email protected]>

commit f9bbd943c34a9ad60e593a4b99ce2394e4e2381b upstream.

In mt6358_read_imp(), the variable val_v is passed to regmap_read() but the return value is not checked. If the read fails, val_v remains uninitialized and its random stack content is subsequently reported as a measurement result.

Initialize val_v to zero to ensure a predictable value is reported in case of bus failure and to prevent potential stack data leakage. This also satisfies static analyzers that might otherwise flag the variable as used uninitialized.

Fixes 3587914bf61d ("iio: adc: Add support for MediaTek MT6357/8/9 Auxiliary ADC")
---
 drivers/iio/adc/mt6359-auxadc.c |    1 +
 1 file changed, 1 insertion(+)
drivers/iio/adc/mt6359-auxadc.c +1 -0
@@ -497,6 +497,7 @@ static int mt6358_read_imp(struct mt6359
 		return ret;
 
 	/* Read the params before stopping */
+	val_v = 0;
 	regmap_read(regmap, reg_adc0 + (cinfo->imp_adc_num << 1), &val_v);
 
 	mt6358_stop_imp_conv(adc_dev);
author maintainer
#196 lore raw

PATCH 183/332 iio: adc: viperboard: Fix error handling in vprbrd_iio_read_raw

In reply to Greg Kroah-Hartman

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

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

From: Salah Triki <[email protected]>

commit 422b5bbf333f75fb486855ad0eedc23cf21f3277 upstream.

The driver proceeds to the reception phase even if the preceding transmission fails.

This uses a goto error label for an early bail out and ensures the mutex is properly unlocked in case of failure.

---
 drivers/iio/adc/viperboard_adc.c |    4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
drivers/iio/adc/viperboard_adc.c +3 -1
@@ -70,8 +70,10 @@ static int vprbrd_iio_read_raw(struct ii
 			VPRBRD_USB_TYPE_OUT, 0x0000, 0x0000, admsg,
 			sizeof(struct vprbrd_adc_msg), VPRBRD_USB_TIMEOUT_MS);
 		if (ret != sizeof(struct vprbrd_adc_msg)) {
-			dev_err(&iio_dev->dev, "usb send error on adc read\n");
+			mutex_unlock(&vb->lock);
 			error = -EREMOTEIO;
+			dev_err(&iio_dev->dev, "usb send error on adc read\n");
+			goto error;
 		}
 
 		ret = usb_control_msg(vb->usb_dev,
author maintainer
#197 lore raw

PATCH 184/332 iio: adc: ad4695: Fix call ordering in offload buffer postenable

In reply to Greg Kroah-Hartman

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

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

From: Radu Sabau <[email protected]>

commit 1a772719318c11e146f6fbe621fffd230a6f456a upstream.

ad4695_enter_advanced_sequencer_mode() was called after spi_offload_trigger_enable(). That is wrong because ad4695_enter_advanced_sequencer_mode() issues regular SPI transfers to put the ADC into advanced sequencer mode, and not all SPI offload capable controllers support regular SPI transfers while offloading is enabled.

Fix this by calling ad4695_enter_advanced_sequencer_mode() before spi_offload_trigger_enable(), so the ADC is fully configured before the first CNV pulse can occur. This is consistent with the same constraint that already applies to the BUSY_GP_EN write above it.

Update the error unwind labels accordingly: add err_exit_conversion_mode so that a failure of spi_offload_trigger_enable() correctly exits conversion mode before clearing BUSY_GP_EN.

Fixes f09f140e3ea8 ("iio: adc: ad4695: Add support for SPI offload")
---
 drivers/iio/adc/ad4695.c |   23 ++++++++---------------
 1 file changed, 8 insertions(+), 15 deletions(-)
drivers/iio/adc/ad4695.c +8 -15
@@ -876,14 +876,14 @@ static int ad4695_offload_buffer_postena
 	if (ret)
 		goto err_unoptimize_message;
 
-	ret = spi_offload_trigger_enable(st->offload, st->offload_trigger,
-					 &config);
+	ret = ad4695_enter_advanced_sequencer_mode(st, num_slots);
 	if (ret)
 		goto err_disable_busy_output;
 
-	ret = ad4695_enter_advanced_sequencer_mode(st, num_slots);
+	ret = spi_offload_trigger_enable(st->offload, st->offload_trigger,
+					 &config);
 	if (ret)
-		goto err_offload_trigger_disable;
+		goto err_exit_conversion_mode;
 
 	mutex_lock(&st->cnv_pwm_lock);
 	pwm_get_state(st->cnv_pwm, &state);
@@ -895,23 +895,16 @@ static int ad4695_offload_buffer_postena
 	ret = pwm_apply_might_sleep(st->cnv_pwm, &state);
 	mutex_unlock(&st->cnv_pwm_lock);
 	if (ret)
-		goto err_offload_exit_conversion_mode;
+		goto err_offload_trigger_disable;
 
 	return 0;
 
-err_offload_exit_conversion_mode:
-	/*
-	 * We have to unwind in a different order to avoid triggering offload.
-	 * ad4695_exit_conversion_mode() triggers a conversion, so it has to be
-	 * done after spi_offload_trigger_disable().
-	 */
-	spi_offload_trigger_disable(st->offload, st->offload_trigger);
-	ad4695_exit_conversion_mode(st);
-	goto err_disable_busy_output;
-
 err_offload_trigger_disable:
 	spi_offload_trigger_disable(st->offload, st->offload_trigger);
 
+err_exit_conversion_mode:
+	ad4695_exit_conversion_mode(st);
+
 err_disable_busy_output:
 	regmap_clear_bits(st->regmap, AD4695_REG_GP_MODE,
 			  AD4695_REG_GP_MODE_BUSY_GP_EN);
author maintainer
#198 lore raw

PATCH 185/332 iio: adc: nxp-sar-adc: fix division by zero in write_raw

In reply to Greg Kroah-Hartman

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

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

From: Antoniu Miclaus <[email protected]>

commit a9aba21a539c668a66b58eeb08ad3909e5a54c2a upstream.

Add a validation check for the sampling frequency value before using it as a divisor. A user writing zero or a negative value to the sampling_frequency sysfs attribute triggers a division by zero in the kernel.

Also prevent unsigned integer underflow when the computed cycle count is smaller than NXP_SAR_ADC_CONV_TIME, which would wrap the u32 inpsamp to a huge value.

Fixes 4434072a893e ("iio: adc: Add the NXP SAR ADC support for the s32g2/3 platforms")
---
 drivers/iio/adc/nxp-sar-adc.c |    9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)
drivers/iio/adc/nxp-sar-adc.c +8 -1
@@ -559,6 +559,9 @@ static int nxp_sar_adc_write_raw(struct
 
 	switch (mask) {
 	case IIO_CHAN_INFO_SAMP_FREQ:
+		if (val <= 0)
+			return -EINVAL;
+
 		/*
 		 * Configures the sample period duration in terms of the SAR
 		 * controller clock. The minimum acceptable value is 8.
@@ -567,7 +570,11 @@ static int nxp_sar_adc_write_raw(struct
 		 * sampling timing which gives us the number of cycles expected.
 		 * The value is 8-bit wide, consequently the max value is 0xFF.
 		 */
-		inpsamp = clk_get_rate(info->clk) / val - NXP_SAR_ADC_CONV_TIME;
+		inpsamp = clk_get_rate(info->clk) / val;
+		if (inpsamp < NXP_SAR_ADC_CONV_TIME)
+			return -EINVAL;
+
+		inpsamp -= NXP_SAR_ADC_CONV_TIME;
 		nxp_sar_adc_conversion_timing_set(info, inpsamp);
 		return 0;
 
author maintainer
#199 lore raw

PATCH 186/332 iio: adc: nxp-sar-adc: Avoid division by zero

In reply to Greg Kroah-Hartman

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

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

From: Andy Shevchenko <[email protected]>

commit 7e5c0f97c66ad538b87c04a640573371fb434b4f upstream.

When Common Clock Framework is disabled, clk_get_rate() returns 0. This is used as part of the divisor to perform nanosecond delays with help of ndelay(). When the above condition occurs the compiler, due to unspecified behaviour, is free to do what it wants to. Here it saturates the value, which is logical from mathematics point of view. However, the ndelay() implementation has set a reasonable upper threshold and refuses to provide anything for such a long delay. That's why code may not be linked under these circumstances.

To solve the issue, provide a wrapper that calls ndelay() when the value is known not to be zero.

---
 drivers/iio/adc/nxp-sar-adc.c |   13 +++++++++++--
 1 file changed, 11 insertions(+), 2 deletions(-)
drivers/iio/adc/nxp-sar-adc.c +11 -2
@@ -198,6 +198,15 @@ static void nxp_sar_adc_irq_cfg(struct n
 		writel(0, NXP_SAR_ADC_IMR(info->regs));
 }
 
+static void nxp_sar_adc_wait_for(struct nxp_sar_adc *info, unsigned int cycles)
+{
+	u64 rate;
+
+	rate = clk_get_rate(info->clk);
+	if (rate)
+		ndelay(div64_u64(NSEC_PER_SEC, rate * cycles));
+}
+
 static bool nxp_sar_adc_set_enabled(struct nxp_sar_adc *info, bool enable)
 {
 	u32 mcr;
@@ -221,7 +230,7 @@ static bool nxp_sar_adc_set_enabled(stru
 	 * configuration of NCMR and the setting of NSTART.
 	 */
 	if (enable)
-		ndelay(div64_u64(NSEC_PER_SEC, clk_get_rate(info->clk) * 3));
+		nxp_sar_adc_wait_for(info, 3);
 
 	return pwdn;
 }
@@ -468,7 +477,7 @@ static void nxp_sar_adc_stop_conversion(
 	 * only when the capture finishes. The delay will be very
 	 * short, usec-ish, which is acceptable in the atomic context.
 	 */
-	ndelay(div64_u64(NSEC_PER_SEC, clk_get_rate(info->clk)) * 80);
+	nxp_sar_adc_wait_for(info, 80);
 }
 
 static int nxp_sar_adc_start_conversion(struct nxp_sar_adc *info, bool raw)
author maintainer
#200 lore raw

PATCH 187/332 iio: adc: nxp-sar-adc: zero-initialize dma_slave_config

In reply to Greg Kroah-Hartman

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

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

From: Shuvam Pandey <[email protected]>

commit 8ce176501f836634f9c0419c0820140f968e9dc5 upstream.

nxp_sar_adc_start_cyclic_dma() only fills the RX-side members of dma_slave_config before passing it to dmaengine_slave_config().

Zero-initialize the structure so unused members do not contain stack garbage. Some DMA engines consult optional dma_slave_config fields, so leaving them uninitialized can cause DMA setup failures.

Fixes 4434072a893e ("iio: adc: Add the NXP SAR ADC support for the s32g2/3 platforms")
---
 drivers/iio/adc/nxp-sar-adc.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/iio/adc/nxp-sar-adc.c +1 -1
@@ -675,7 +675,7 @@ static void nxp_sar_adc_dma_cb(void *dat
 static int nxp_sar_adc_start_cyclic_dma(struct iio_dev *indio_dev)
 {
 	struct nxp_sar_adc *info = iio_priv(indio_dev);
-	struct dma_slave_config config;
+	struct dma_slave_config config = { };
 	struct dma_async_tx_descriptor *desc;
 	int ret;
 
author maintainer
#201 lore raw

PATCH 188/332 iio: gyro: itg3200: fix i2c read into the wrong stack location

In reply to Greg Kroah-Hartman

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

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

From: David Carlier <[email protected]>

commit 6bdc3023d62ed5c7d591f0eb27a5adb37fb892ae upstream.

itg3200_read_all_channels() takes __be16 *buf' as a parameter and fills the i2c_msg destination as (char *)&buf'. Since buf' is the parameter (a pointer), &buf' is the address of the local pointer slot on the stack of itg3200_read_all_channels(), not the address of the caller's scan buffer. The (char *) cast hides the type mismatch.

i2c_transfer() therefore writes ITG3200_SCAN_ELEMENTS * sizeof(s16) = 8 bytes into the parameter's stack slot, which is discarded when the function returns. The caller's scan buffer in itg3200_trigger_handler() is never written to, so iio_push_to_buffers_with_timestamp() pushes uninitialised stack contents to userspace via /dev/iio:deviceX every scan -- both a functional bug (no actual gyroscope or temperature data is delivered through the triggered buffer) and an information leak.

The non-buffered read_raw() path is unaffected: it goes through itg3200_read_reg_s16() which uses `&out' on a local s16 value, where that is correct.

Drop the spurious `&' so the i2c read writes into the caller's buffer.

---
 drivers/iio/gyro/itg3200_buffer.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/iio/gyro/itg3200_buffer.c +1 -1
@@ -34,7 +34,7 @@ static int itg3200_read_all_channels(str
 			.addr = i2c->addr,
 			.flags = i2c->flags | I2C_M_RD,
 			.len = ITG3200_SCAN_ELEMENTS * sizeof(s16),
-			.buf = (char *)&buf,
+			.buf = (char *)buf,
 		},
 	};
 
author maintainer
#202 lore raw

PATCH 189/332 iio: gyro: adis16260: fix division by zero in write_raw

In reply to Greg Kroah-Hartman

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

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

From: Antoniu Miclaus <[email protected]>

commit 761e8b489e6cf166c574034b70637f8a7eadd0ee upstream.

Add a validation check for the sampling frequency value before using it as a divisor. A user writing zero to the sampling_frequency sysfs attribute triggers a division by zero in the kernel.

Fixes 089a41985c6c ("staging: iio: adis16260 digital gyro driver")
---
 drivers/iio/gyro/adis16260.c |    3 +++
 1 file changed, 3 insertions(+)
drivers/iio/gyro/adis16260.c +3 -0
@@ -287,6 +287,9 @@ static int adis16260_write_raw(struct ii
 		addr = adis16260_addresses[chan->scan_index][1];
 		return adis_write_reg_16(adis, addr, val);
 	case IIO_CHAN_INFO_SAMP_FREQ:
+		if (val <= 0)
+			return -EINVAL;
+
 		if (spi_get_device_id(adis->spi)->driver_data)
 			t = 256 / val;
 		else
author maintainer
#203 lore raw

PATCH 190/332 iio: ssp_sensors: cancel delayed work_refresh on remove

In reply to Greg Kroah-Hartman

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

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

From: Sanjay Chitroda <[email protected]>

commit eedf7602fbd929e97e0c480da501dc7a34beb2a8 upstream.

The work_refresh may still be pending or running when the device is removed, cancel the delayed work_refresh in remove path.

Fixes 50dd64d57eee ("iio: common: ssp_sensors: Add sensorhub driver")
---
 drivers/iio/common/ssp_sensors/ssp_dev.c |    1 +
 1 file changed, 1 insertion(+)
drivers/iio/common/ssp_sensors/ssp_dev.c +1 -0
@@ -590,6 +590,7 @@ static void ssp_remove(struct spi_device
 	ssp_clean_pending_list(data);
 
 	free_irq(data->spi->irq, data);
+	cancel_delayed_work_sync(&data->work_refresh);
 
 	timer_delete_sync(&data->wdt_timer);
 	cancel_work_sync(&data->work_wdt);
author maintainer
#204 lore raw

PATCH 191/332 iio: temperature: tsys01: fix broken PROM checksum validation

In reply to Greg Kroah-Hartman

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

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

From: Salah Triki <[email protected]>

commit 4701e471c16866e7aa8f5e6a3a6b0d31e097e2c9 upstream.

The current implementation of tsys01_crc_valid() incorrectly sums the first word (n_prom[0]) repeatedly instead of iterating over the 8 words retrieved from the PROM. This leads to a checksum mismatch and probe failure on hardware.

According to the TSYS01 datasheet, the PROM consists of 8 words. A valid check must iterate through all 8 words to verify the integrity of the calibration data. The current driver only checks the first word 8 times.

Note: This fix was identified during a code audit and is based on datasheet specifications. It has not been tested on real hardware.

Fixes 43e53407f680 ("Add tsys01 meas-spec driver support")
---
 drivers/iio/temperature/tsys01.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/iio/temperature/tsys01.c +1 -1
@@ -119,7 +119,7 @@ static bool tsys01_crc_valid(u16 *n_prom
 	u8 sum = 0;
 
 	for (cnt = 0; cnt < TSYS01_PROM_WORDS_NB; cnt++)
-		sum += ((n_prom[0] >> 8) + (n_prom[0] & 0xFF));
+		sum += ((n_prom[cnt] >> 8) + (n_prom[cnt] & 0xFF));
 
 	return (sum == 0);
 }
author maintainer
#205 lore raw

PATCH 192/332 iio: magnetometer: st_magn: fix default DRDY pin selection for LIS2MDL

In reply to Greg Kroah-Hartman

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

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

From: Advait Dhamorikar <[email protected]>

commit 49f79cd28f1e3333cbe0d616ce59ead0b24bf34e upstream.

The device tree binding for st,lis2mdl does not support st,drdy-int-pin property. However, when no platform data is provided and the property is absent, the driver falls back to default_magn_pdata which hardcodes drdy_int_pin = 2. This causes st_sensors_set_drdy_int_pin to fail with -EINVAL because the LIS2MDL sensor settings have no INT2 DRDY mask defined.

Fix this by checking the sensor's INT2 DRDY mask availability at probe time and selecting the appropriate default pin. Sensors that do not support INT2 DRDY will default to INT1, while all others retain the existing default of INT2.

Fixes 38934daf7b5c ("iio: magnetometer: st_magn: Provide default platform data")
---
 drivers/iio/magnetometer/st_magn_core.c |   13 +++++++++++--
 1 file changed, 11 insertions(+), 2 deletions(-)
drivers/iio/magnetometer/st_magn_core.c +11 -2
@@ -506,6 +506,11 @@ static const struct st_sensors_platform_
 	.drdy_int_pin = 2,
 };
 
+/* LIS2MDL only supports DRDY on INT1 */
+static const struct st_sensors_platform_data alt_magn_pdata = {
+	.drdy_int_pin = 1,
+};
+
 static int st_magn_read_raw(struct iio_dev *indio_dev,
 			struct iio_chan_spec const *ch, int *val,
 							int *val2, long mask)
@@ -628,8 +633,12 @@ int st_magn_common_probe(struct iio_dev
 	mdata->current_fullscale = &mdata->sensor_settings->fs.fs_avl[0];
 	mdata->odr = mdata->sensor_settings->odr.odr_avl[0].hz;
 
-	if (!pdata)
-		pdata = (struct st_sensors_platform_data *)&default_magn_pdata;
+	if (!pdata) {
+		if (mdata->sensor_settings->drdy_irq.int2.mask)
+			pdata = (struct st_sensors_platform_data *)&default_magn_pdata;
+		else
+			pdata = (struct st_sensors_platform_data *)&alt_magn_pdata;
+	}
 
 	err = st_sensors_init_sensor(indio_dev, pdata);
 	if (err < 0)
author maintainer
#206 lore raw

PATCH 193/332 iio: light: veml6070: Fix resource leak in probe error path

In reply to Greg Kroah-Hartman

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

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

From: Felix Gu <[email protected]>

commit b66f922f6a4fa92840f662fbcfeb4f8a0f774bcc upstream.

The driver calls i2c_new_dummy_device() to create a dummy device, then calls i2c_smbus_write_byte(). If i2c_smbus_write_byte() fails and returns, the cleanup via devm_add_action_or_reset() was never registered, so the dummy device leaks.

Switch to devm_i2c_new_dummy_device() which registers cleanup atomically with device creation, eliminating the error-path window.

Fixes 7501bff87c3e ("iio: light: veml6070: add action for i2c_unregister_device")
---
 drivers/iio/light/veml6070.c |   14 ++------------
 1 file changed, 2 insertions(+), 12 deletions(-)
drivers/iio/light/veml6070.c +2 -12
@@ -245,13 +245,6 @@ static const struct iio_info veml6070_in
 	.write_raw = veml6070_write_raw,
 };
 
-static void veml6070_i2c_unreg(void *p)
-{
-	struct veml6070_data *data = p;
-
-	i2c_unregister_device(data->client2);
-}
-
 static int veml6070_probe(struct i2c_client *client)
 {
 	struct veml6070_data *data;
@@ -281,7 +274,8 @@ static int veml6070_probe(struct i2c_cli
 	if (ret < 0)
 		return ret;
 
-	data->client2 = i2c_new_dummy_device(client->adapter, VEML6070_ADDR_DATA_LSB);
+	data->client2 = devm_i2c_new_dummy_device(&client->dev, client->adapter,
+						  VEML6070_ADDR_DATA_LSB);
 	if (IS_ERR(data->client2))
 		return dev_err_probe(&client->dev, PTR_ERR(data->client2),
 				     "i2c device for second chip address failed\n");
@@ -292,10 +286,6 @@ static int veml6070_probe(struct i2c_cli
 	if (ret < 0)
 		return ret;
 
-	ret = devm_add_action_or_reset(&client->dev, veml6070_i2c_unreg, data);
-	if (ret < 0)
-		return ret;
-
 	return devm_iio_device_register(&client->dev, indio_dev);
 }
 
author maintainer
#207 lore raw

PATCH 194/332 iio: Fix iio_multiply_value use in iio_read_channel_processed_scale

In reply to Greg Kroah-Hartman

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

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

From: Svyatoslav Ryhel <[email protected]>

commit bb21ee31f5753a7972148798fd7dfb841dd33bdb upstream.

The function iio_multiply_value returns IIO_VAL_INT (1) on success or a negative error number on failure, while iio_read_channel_processed_scale should return an error code or 0. This creates a situation where the expected result is treated as an error. Fix this by checking the iio_multiply_value result separately, instead of passing it as a return value.

Fixes 05f958d003c9 ("iio: Improve iio_read_channel_processed_scale() precision")
---
 drivers/iio/inkern.c |    6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)
drivers/iio/inkern.c +5 -1
@@ -738,7 +738,11 @@ int iio_read_channel_processed_scale(str
 		if (ret < 0)
 			return ret;
 
-		return iio_multiply_value(val, scale, ret, pval, pval2);
+		ret = iio_multiply_value(val, scale, ret, pval, pval2);
+		if (ret < 0)
+			return ret;
+
+		return 0;
 	} else {
 		ret = iio_channel_read(chan, val, NULL, IIO_CHAN_INFO_RAW);
 		if (ret < 0)
author maintainer
#208 lore raw

PATCH 195/332 iio: chemical: mhz19b: reject oversized serial replies

In reply to Greg Kroah-Hartman

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

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

From: Pengpeng Hou <[email protected]>

commit 673478bc29cf72010faaf293c1c8c667393335a0 upstream.

mhz19b_receive_buf() appends each serdev chunk into the fixed MHZ19B_CMD_SIZE receive buffer and advances buf_idx by len without checking that the chunk fits in the remaining space. A large callback can therefore overflow st->buf before the command path validates the reply.

Reset the reply state before each command and reject oversized serial replies before copying them into the fixed buffer. When an oversized reply is detected, wake the waiter and report -EMSGSIZE instead of overwriting st->buf.

Fixes 4572a70b3681 ("iio: chemical: Add support for Winsen MHZ19B CO2 sensor")
---
 drivers/iio/chemical/mhz19b.c |   17 +++++++++++++++++
 1 file changed, 17 insertions(+)
drivers/iio/chemical/mhz19b.c +17 -0
@@ -52,6 +52,8 @@ struct mhz19b_state {
 	struct completion buf_ready;
 
 	u8 buf_idx;
+	bool buf_overflow;
+
 	/*
 	 * Serdev receive buffer.
 	 * When data is received from the MH-Z19B,
@@ -106,6 +108,10 @@ static int mhz19b_serdev_cmd(struct iio_
 	cmd_buf[8] = mhz19b_get_checksum(cmd_buf);
 
 	/* Write buf to uart ctrl synchronously */
+	st->buf_idx = 0;
+	st->buf_overflow = false;
+	reinit_completion(&st->buf_ready);
+
 	ret = serdev_device_write(serdev, cmd_buf, MHZ19B_CMD_SIZE, 0);
 	if (ret < 0)
 		return ret;
@@ -121,6 +127,9 @@ static int mhz19b_serdev_cmd(struct iio_
 		if (!ret)
 			return -ETIMEDOUT;
 
+		if (st->buf_overflow)
+			return -EMSGSIZE;
+
 		if (st->buf[8] != mhz19b_get_checksum(st->buf)) {
 			dev_err(dev, "checksum err");
 			return -EINVAL;
@@ -240,6 +249,14 @@ static size_t mhz19b_receive_buf(struct
 {
 	struct iio_dev *indio_dev = dev_get_drvdata(&serdev->dev);
 	struct mhz19b_state *st = iio_priv(indio_dev);
+	size_t remaining = MHZ19B_CMD_SIZE - st->buf_idx;
+
+	if (len > remaining) {
+		st->buf_idx = 0;
+		st->buf_overflow = true;
+		complete(&st->buf_ready);
+		return len;
+	}
 
 	memcpy(st->buf + st->buf_idx, data, len);
 	st->buf_idx += len;
author maintainer
#209 lore raw

PATCH 196/332 iio: chemical: scd30: fix division by zero in write_raw

In reply to Greg Kroah-Hartman

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

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

From: Antoniu Miclaus <[email protected]>

commit 5aba4f94b225617a55fed442a70329b2ee19c0a5 upstream.

Add a zero check for val2 before using it as a divisor when setting the sampling frequency. A user writing a zero fractional part to the sampling_frequency sysfs attribute triggers a division by zero in the kernel.

Fixes 64b3d8b1b0f5 ("iio: chemical: scd30: add core driver")
---
 drivers/iio/chemical/scd30_core.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/iio/chemical/scd30_core.c +1 -1
@@ -256,7 +256,7 @@ static int scd30_write_raw(struct iio_de
 	guard(mutex)(&state->lock);
 	switch (mask) {
 	case IIO_CHAN_INFO_SAMP_FREQ:
-		if (val)
+		if (val || !val2)
 			return -EINVAL;
 
 		val = 1000000000 / val2;
author maintainer
#210 lore raw

PATCH 197/332 iio: light: cm3323: fix reg_conf not being initialized correctly

In reply to Greg Kroah-Hartman

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

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

From: Aldo Conte <[email protected]>

commit 1f4f0bcc5255dec5c4c3a1551bf49d8c33b69b20 upstream.

The code stores the return value of i2c_smbus_write_word_data() in data->reg_conf; however, this value represents the result of the write operation and not the value actually written to the configuration register. This meant that the contents of data->reg_conf did not truly reflect the contents
of the hardware register.

Instead, save the value of the register before the write and use this value in the I2C write.

The bug was found by code inspection: i2c_smbus_write_word_data() returns 0 on success, not the value written to the register.

Tested using i2c-stub on a Raspberry Pi 3B running a custom 6.19.10 kernel. Before loading the driver, the configuration register 0x00 CM3323_CMD_CONF was populated with 0x0030 using
i2cset -y 11 0x10 0x00 0x0030 w, encoding an integration time of 320ms in bits[6:4].

Due to incorrect initialization of data->reg_conf in cm3323_init(), the print of integration_time returns 0.040000 instead of the expected 0.320000. This happens because the read of the integration_time depends on cm3323_get_it_bits() that is based on the value of data->reg_conf, which is erroneously set to 0.

With this fix applied, data->reg_conf correctly saves 0x0030 after init and the successive integration_time reports 0.320000 as expected.

Fixes 8b0544263761 ("iio: light: Add support for Capella CM3323 color sensor")
---
 drivers/iio/light/cm3323.c |    5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)
drivers/iio/light/cm3323.c +2 -3
@@ -89,15 +89,14 @@ static int cm3323_init(struct iio_dev *i
 
 	/* enable sensor and set auto force mode */
 	ret &= ~(CM3323_CONF_SD_BIT | CM3323_CONF_AF_BIT);
+	data->reg_conf = ret;
 
-	ret = i2c_smbus_write_word_data(data->client, CM3323_CMD_CONF, ret);
+	ret = i2c_smbus_write_word_data(data->client, CM3323_CMD_CONF, data->reg_conf);
 	if (ret < 0) {
 		dev_err(&data->client->dev, "Error writing reg_conf\n");
 		return ret;
 	}
 
-	data->reg_conf = ret;
-
 	return 0;
 }
 
author maintainer
#211 lore raw

PATCH 198/332 iio: buffer: hw-consumer: fix use-after-free in error path

In reply to Greg Kroah-Hartman

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

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

From: Felix Gu <[email protected]>

commit 6f5ed4f2c7c83f33344e0ba179f72a12e5dad4a4 upstream.

In the err_put_buffers cleanup path of iio_hw_consumer_alloc(), the code was using list_for_each_entry() to iterate through buffers while calling iio_buffer_put() which can free the current buffer if refcount drops to 0. The list_for_each_entry() loop macro then evaluates buf->head.next to continue iteration, accessing the freed buffer.

Fix this by using list_for_each_entry_safe().

---
 drivers/iio/buffer/industrialio-hw-consumer.c |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/iio/buffer/industrialio-hw-consumer.c +2 -2
@@ -82,7 +82,7 @@ static struct hw_consumer_buffer *iio_hw
  */
 struct iio_hw_consumer *iio_hw_consumer_alloc(struct device *dev)
 {
-	struct hw_consumer_buffer *buf;
+	struct hw_consumer_buffer *buf, *tmp;
 	struct iio_hw_consumer *hwc;
 	struct iio_channel *chan;
 	int ret;
@@ -113,7 +113,7 @@ struct iio_hw_consumer *iio_hw_consumer_
 	return hwc;
 
 err_put_buffers:
-	list_for_each_entry(buf, &hwc->buffers, head)
+	list_for_each_entry_safe(buf, tmp, &hwc->buffers, head)
 		iio_buffer_put(&buf->buffer);
 	iio_channel_release_all(hwc->channels);
 err_free_hwc:
author maintainer
#212 lore raw

PATCH 199/332 iio: buffer: Fix DMA fence leak in iio_buffer_enqueue_dmabuf()

In reply to Greg Kroah-Hartman

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

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

From: Benoît Monin <[email protected]>

commit a093999355084bdbfe6e97f1dd232e58a1525f0b upstream.

iio_buffer_enqueue_dmabuf() allocates a struct iio_dma_fence (104 bytes, kmalloc-128) via kmalloc_obj()+dma_fence_init(), which sets the initial kref to 1. It then calls dma_resv_add_fence() which takes a second reference (kref=2), and stores a raw pointer in block->fence.

On the success path the function returns without calling dma_fence_put() to release the initial reference, so every buffer enqueue permanently leaks one kmalloc-128 allocation.

The iio_buffer_cleanup() work item only releases the temporary reference taken during completion signalling by iio_buffer_signal_dmabuf_done(); the initial reference from dma_fence_init() is never released.

With four iio_rwdev instances at 240kHz and 512 samples per buffer, this produces ~1875 kmalloc-128 allocations per second matching the observed slab growth exactly. A test with ftrace confirmed that the dma_fence_destroy event was never triggered.

Fix by calling dma_fence_put() after dma_resv_add_fence(), transferring ownership of the fence to the DMA reservation object. The DMA fence then gets properly discarded after being signalled.

Fixes 3e26d9f08fbe0 ("iio: core: Add new DMABUF interface infrastructure")
---
 drivers/iio/industrialio-buffer.c |    1 +
 1 file changed, 1 insertion(+)
drivers/iio/industrialio-buffer.c +1 -0
@@ -1909,6 +1909,7 @@ static int iio_buffer_enqueue_dmabuf(str
 
 	dma_resv_add_fence(dmabuf->resv, &fence->base,
 			   dma_to_ram ? DMA_RESV_USAGE_WRITE : DMA_RESV_USAGE_READ);
+	dma_fence_put(&fence->base);
 	dma_resv_unlock(dmabuf->resv);
 
 	cookie = dma_fence_begin_signalling();
author maintainer
#213 lore raw

PATCH 200/332 USB: serial: omninet: fix memory corruption with small endpoint

In reply to Greg Kroah-Hartman

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

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

From: Johan Hovold <[email protected]>

commit 60df93d30f9bdd27db17c4d80ed80ef718d7226b upstream.

Make sure that the bulk-out buffers are at least as large as the hardcoded transfer size to avoid user-controlled slab corruption should a malicious device report a smaller endpoint max packet size than expected.

---
 drivers/usb/serial/omninet.c |    9 +++++----
 1 file changed, 5 insertions(+), 4 deletions(-)
drivers/usb/serial/omninet.c +5 -4
@@ -30,6 +30,10 @@
 /* This one seems to be a re-branded ZyXEL device */
 #define BT_IGNITIONPRO_ID	0x2000
 
+#define OMNINET_HEADERLEN	4
+#define OMNINET_BULKOUTSIZE	64
+#define OMNINET_PAYLOADSIZE	(OMNINET_BULKOUTSIZE - OMNINET_HEADERLEN)
+
 /* function prototypes */
 static void omninet_process_read_urb(struct urb *urb);
 static int omninet_prepare_write_buffer(struct usb_serial_port *port,
@@ -54,6 +58,7 @@ static struct usb_serial_driver zyxel_om
 	.description =		"ZyXEL - omni.net usb",
 	.id_table =		id_table,
 	.num_bulk_out =		2,
+	.bulk_out_size =	OMNINET_BULKOUTSIZE,
 	.calc_num_ports =	omninet_calc_num_ports,
 	.port_probe =		omninet_port_probe,
 	.port_remove =		omninet_port_remove,
@@ -130,10 +135,6 @@ static void omninet_port_remove(struct u
 	kfree(od);
 }
 
-#define OMNINET_HEADERLEN	4
-#define OMNINET_BULKOUTSIZE	64
-#define OMNINET_PAYLOADSIZE	(OMNINET_BULKOUTSIZE - OMNINET_HEADERLEN)
-
 static void omninet_process_read_urb(struct urb *urb)
 {
 	struct usb_serial_port *port = urb->context;
author maintainer
#214 lore raw

PATCH 201/332 usb: cdns3: gadget: fix request skipping after clearing halt

In reply to Greg Kroah-Hartman

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

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

From: Yongchao Wu <[email protected]>

commit c8778ff817a7047d6848fefba99dcb27b1bf01fe upstream.

According to the cdns3 datasheet, the EPRST (Endpoint Reset) command causes the DMA engine to reposition its internal pointer to the next Transfer Descriptor (TD) if it was already processing one.

This issue is consistently observed during the ADB identification process on macOS hosts, where the host issues a Clear_Halt. Although commit 4bf2dd65135a ("usb: cdns3: gadget: toggle cycle bit before reset endpoint") attempted to avoid DMA advance by toggling the cycle bit, trace logs show that on certain hosts like macOS, the DMA pointer (EP_TRADDR) still shifts after EPRST:

cdns3_ctrl_req: Clear Endpoint Feature(Halt ep1out)
cdns3_doorbell_epx: ep1out, ep_trbaddr f9c04030 <-- Should be f9c04000
cdns3_gadget_giveback: ep1out: req: ... length: 16384/16384

As shown above, the DMA pointer jumped to the next TD, causing the controller to skip the initial TRBs of the request. This leads to data misalignment and ADB protocol hangs on macOS.

Fix this by manually restoring the EP_TRADDR register to the starting physical address of the current request after the EPRST operation is complete.

---
 drivers/usb/cdns3/cdns3-gadget.c |   12 +++++++++++-
 1 file changed, 11 insertions(+), 1 deletion(-)
drivers/usb/cdns3/cdns3-gadget.c +11 -1
@@ -2817,9 +2817,19 @@ int __cdns3_gadget_ep_clear_halt(struct
 	priv_ep->flags &= ~(EP_STALLED | EP_STALL_PENDING);
 
 	if (request) {
-		if (trb)
+		if (trb) {
 			*trb = trb_tmp;
 
+			/*
+			 * Per datasheet, EPRST causes DMA to reposition to the next TD.
+			 * Manually reset EP_TRADDR to the current TRB to prevent
+			 * the hardware from skipping the interrupted request.
+			 */
+			writel(EP_TRADDR_TRADDR(priv_ep->trb_pool_dma +
+						priv_req->start_trb * TRB_SIZE),
+						&priv_dev->regs->ep_traddr);
+		}
+
 		cdns3_rearm_transfer(priv_ep, 1);
 	}
 
author maintainer
#215 lore raw

PATCH 202/332 usb: cdns3: plat: fix leaked usb2_phy initialization on usb3_phy acquisition failure

In reply to Greg Kroah-Hartman

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

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

From: Peter Chen <[email protected]>

commit e6970cda63fd4b4546aeed9d0e2f53a7c95cd09c upstream.

Move usb2_phy initialization after usb3_phy acquisition.

---
 drivers/usb/cdns3/cdns3-plat.c |    8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)
drivers/usb/cdns3/cdns3-plat.c +4 -4
@@ -126,15 +126,15 @@ static int cdns3_plat_probe(struct platf
 		return dev_err_probe(dev, PTR_ERR(cdns->usb2_phy),
 				     "Failed to get cdn3,usb2-phy\n");
 
-	ret = phy_init(cdns->usb2_phy);
-	if (ret)
-		return ret;
-
 	cdns->usb3_phy = devm_phy_optional_get(dev, "cdns3,usb3-phy");
 	if (IS_ERR(cdns->usb3_phy))
 		return dev_err_probe(dev, PTR_ERR(cdns->usb3_phy),
 				     "Failed to get cdn3,usb3-phy\n");
 
+	ret = phy_init(cdns->usb2_phy);
+	if (ret)
+		return ret;
+
 	ret = phy_init(cdns->usb3_phy);
 	if (ret)
 		goto err_phy3_init;
author maintainer
#216 lore raw

PATCH 203/332 usb: cdns3: plat: fix unbalanced pm_runtime_forbid() call permanently leaks the runtime PM usage counter across bind/unbind cycles

In reply to Greg Kroah-Hartman

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

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

From: Peter Chen <[email protected]>

commit ae6f3b82324e4f39ad8443c9020787e6fc889637 upstream.

Call pm_runtime_allow(dev) conditionally at cdns3_plat_remove.

---
 drivers/usb/cdns3/cdns3-plat.c |    3 +++
 1 file changed, 3 insertions(+)
drivers/usb/cdns3/cdns3-plat.c +3 -0
@@ -186,6 +186,9 @@ static void cdns3_plat_remove(struct pla
 	struct device *dev = cdns->dev;
 
 	pm_runtime_get_sync(dev);
+	if (!(cdns->pdata && (cdns->pdata->quirks & CDNS3_DEFAULT_PM_RUNTIME_ALLOW)))
+		pm_runtime_allow(dev);
+
 	pm_runtime_disable(dev);
 	pm_runtime_put_noidle(dev);
 	cdns_remove(cdns);
author maintainer
#217 lore raw

PATCH 204/332 usb: dwc2: Fix use after free in debug code

In reply to Greg Kroah-Hartman

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

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

From: Dan Carpenter <[email protected]>

commit 9ea06a3fbf9f16e0d98c52cb3b99642be15ec281 upstream.

We're not allowed to dereference "urb" after calling usb_hcd_giveback_urb() so save the urb->status ahead of time.

---
 drivers/usb/dwc2/hcd.c |    4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
drivers/usb/dwc2/hcd.c +3 -1
@@ -4804,6 +4804,7 @@ static int _dwc2_hcd_urb_dequeue(struct
 	struct dwc2_hsotg *hsotg = dwc2_hcd_to_hsotg(hcd);
 	int rc;
 	unsigned long flags;
+	int urb_status;
 
 	dev_dbg(hsotg->dev, "DWC OTG HCD URB Dequeue\n");
 	dwc2_dump_urb_info(hcd, urb, "urb_dequeue");
@@ -4828,11 +4829,12 @@ static int _dwc2_hcd_urb_dequeue(struct
 
 	/* Higher layer software sets URB status */
 	spin_unlock(&hsotg->lock);
+	urb_status = urb->status;
 	usb_hcd_giveback_urb(hcd, urb, status);
 	spin_lock(&hsotg->lock);
 
 	dev_dbg(hsotg->dev, "Called usb_hcd_giveback_urb()\n");
-	dev_dbg(hsotg->dev, "  urb->status = %d\n", urb->status);
+	dev_dbg(hsotg->dev, "  urb->status = %d\n", urb_status);
 out:
 	spin_unlock_irqrestore(&hsotg->lock, flags);
 
author maintainer
#218 lore raw

PATCH 205/332 Input: elan_i2c - validate firmware size before use

In reply to Greg Kroah-Hartman

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

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

From: Dmitry Torokhov <[email protected]>

commit 76b0d0baa9ae9c60e726bbe1b6ff0bec2c993634 upstream.

Ensure that the firmware file is large enough to contain the expected number of pages and the signature (which resides at the end of the firmware blob) before accessing them to prevent potential out-of-bounds reads.

---
 drivers/input/mouse/elan_i2c_core.c |    5 +++++
 1 file changed, 5 insertions(+)
drivers/input/mouse/elan_i2c_core.c +5 -0
@@ -645,6 +645,11 @@ static ssize_t elan_sysfs_update_fw(stru
 		return error;
 	}
 
+	if (fw->size < data->fw_signature_address + sizeof(signature)) {
+		dev_err(dev, "firmware file too small\n");
+		return -EBADF;
+	}
+
 	/* Firmware file must match signature data */
 	fw_signature = &fw->data[data->fw_signature_address];
 	if (memcmp(fw_signature, signature, sizeof(signature)) != 0) {
author maintainer
#219 lore raw

PATCH 206/332 i2c: davinci: fix division by zero on missing clock-frequency

In reply to Greg Kroah-Hartman

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

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

From: Chaitanya Sabnis <[email protected]>

commit 030675aa54cf757769b3db65642433d626b3ed7c upstream.

When the 'clock-frequency' property is missing from the device tree, the driver falls back to DAVINCI_I2C_DEFAULT_BUS_FREQ. However, this macro was defined in kHz (100), whereas the device tree property is expected in Hz.

The probe function divided the fallback value by 1000, causing integer truncation that resulted in dev->bus_freq = 0. This triggered a deterministic division-by-zero kernel panic when calculating clock dividers later in the probe sequence.

Fix this by redefining DAVINCI_I2C_DEFAULT_BUS_FREQ in Hz (100000) to match the expected device tree property unit, allowing the existing division logic to work correctly for both cases.

---
 drivers/i2c/busses/i2c-davinci.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/i2c/busses/i2c-davinci.c +1 -1
@@ -117,7 +117,7 @@
 /* timeout for pm runtime autosuspend */
 #define DAVINCI_I2C_PM_TIMEOUT	1000	/* ms */
 
-#define DAVINCI_I2C_DEFAULT_BUS_FREQ	100
+#define DAVINCI_I2C_DEFAULT_BUS_FREQ	100000
 
 struct davinci_i2c_dev {
 	struct device           *dev;
author maintainer
#220 lore raw

PATCH 207/332 x86/ftrace: Relocate %rip-relative percpu refs in dynamic trampolines

In reply to Greg Kroah-Hartman

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

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

From: Alexis Lothoré (eBPF Foundation) <[email protected]>

commit a17dc12bfed8868e6a86f3b45c16065a70641acb upstream.

With CONFIG_CALL_DEPTH_TRACKING enabled on an x86 retbleed-affected platform (eg: Skylake), with retbleed=stuff, registering a dynamic ftrace trampoline crashes on the first call into the traced function:

  BUG: unable to handle page fault for address: ffff88817ae18880
  #PF: supervisor write access in kernel mode
  #PF: error_code(0x0002) - not-present page
  PGD 4b53067 P4D 4b53067 PUD 0
  Oops: Oops: 0002 [#1] SMP PTI
  CPU: 3 UID: 0 PID: 187 Comm: usleep Not tainted 7.0.10 #243 PREEMPT(full)
  Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS Arch Linux 1.17.0-2-2 04/01/2014
  Code: 24 78 00 00 00 00 48 89 ea 48 89 54 24 20 48 8b b4 24 b8 00 00 00 48 8b bc 24 b0 00 00 00 48 89 bc 24 80 00 00 00 48 83 ef 05 <65> 48 c1 3d 1f a8 b6 02 05 48 8b 15 f6 00 00 00 4c 89 3c 24 4c 89
  Call Trace:
   <TASK>
   ? find_held_lock
   ? exc_page_fault
   ? lock_release
   ? __x64_sys_clock_nanosleep
   ? lockdep_hardirqs_on_prepare
   ? trace_hardirqs_on
   __x64_sys_clock_nanosleep
   do_syscall_64
   ? exc_page_fault
   ? call_depth_return_thunk
   entry_SYSCALL_64_after_hwframe
  ...
  Kernel panic - not syncing: Fatal exception

This small reproducer allows to easily trigger the crash:

  # echo 'p __x64_sys_clock_nanosleep' > /sys/kernel/tracing/kprobe_events
  # echo 1 > /sys/kernel/tracing/events/kprobes/p___x64_sys_clock_nanosleep_0/enable
  # usleep 1

Monitoring the crash under GDB points to the exact instruction in charge of incrementing the call depth:

sarq $5, %gs:__x86_call_depth(%rip)

This instruction matches the one inserted by the ftrace_regs_caller from ftrace_64.S. This emitted code was likely working fine until the introduction of

59bec00ace28 ("x86/percpu: Introduce %rip-relative addressing to PER_CPU_VAR()"):

it has made the call depth accounting addressing relative to $rip, instead of being based on an absolute address.

As this code exact location depends on where the trampoline lives in memory, the corresponding displacement needs to be adjusted at runtime to actually correctly find the per-cpu __x86_call_depth value, otherwise the targeted address is wrong, leading to the page fault seen above.

Fix the %rip-relative displacement of the copied CALL_DEPTH_ACCOUNT instruction (from ftrace_regs_caller) by calling text_poke_apply_relocation(), as it is done for example by the x86 BPF JIT compiler through x86_call_depth_emit_accounting(). This corrects both CALL_DEPTH_ACCOUNT slots, in ftrace_caller and ftrace_regs_caller.

[ bp: Massage. ]

---
 arch/x86/kernel/ftrace.c |    7 +++++++
 1 file changed, 7 insertions(+)
arch/x86/kernel/ftrace.c +7 -0
@@ -376,6 +376,13 @@ create_trampoline(struct ftrace_ops *ops
 	}
 
 	/*
+	 * Generated trampoline may contain rIP-relative addressing which
+	 * displacement needs to be fixed.
+	 */
+	text_poke_apply_relocation(trampoline, trampoline, size,
+				   (void *)start_offset, size);
+
+	/*
 	 * The address of the ftrace_ops that is used for this trampoline
 	 * is stored at the end of the trampoline. This will be used to
 	 * load the third parameter for the callback. Basically, that
author maintainer
#221 lore raw

PATCH 208/332 wireguard: send: append trailer after expanding head

In reply to Greg Kroah-Hartman

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

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

From: Jason A. Donenfeld <[email protected]>

commit f75e3eb08fe31d30a9af6ed80cdd22e6772837e2 upstream.

With how this is currently written, we add the trailer, zero it out, and then add the header space on. If that header space requires a reallocation + copy, the zeros in the trailer aren't copied, because the skb len hasn't actually been yet expanded to cover that. Instead add the padding at the end of the process rather than at the beginning.

---
 drivers/net/wireguard/send.c |   20 ++++++++++----------
 1 file changed, 10 insertions(+), 10 deletions(-)
drivers/net/wireguard/send.c +10 -10
@@ -177,16 +177,6 @@ static bool encrypt_packet(struct sk_buf
 	trailer_len = padding_len + noise_encrypted_len(0);
 	plaintext_len = skb->len + padding_len;
 
-	/* Expand data section to have room for padding and auth tag. */
-	num_frags = skb_cow_data(skb, trailer_len, &trailer);
-	if (unlikely(num_frags < 0 || num_frags > ARRAY_SIZE(sg)))
-		return false;
-
-	/* Set the padding to zeros, and make sure it and the auth tag are part
-	 * of the skb.
-	 */
-	memset(skb_tail_pointer(trailer), 0, padding_len);
-
 	/* Expand head section to have room for our header and the network
 	 * stack's headers.
 	 */
@@ -198,6 +188,16 @@ static bool encrypt_packet(struct sk_buf
 		     skb_checksum_help(skb)))
 		return false;
 
+	/* Expand data section to have room for padding and auth tag. */
+	num_frags = skb_cow_data(skb, trailer_len, &trailer);
+	if (unlikely(num_frags < 0 || num_frags > ARRAY_SIZE(sg)))
+		return false;
+
+	/* Set the padding to zeros, and make sure it and the auth tag are part
+	 * of the skb.
+	 */
+	memset(skb_tail_pointer(trailer), 0, padding_len);
+
 	/* Only after checksumming can we safely add on the padding at the end
 	 * and the header.
 	 */
author maintainer
#222 lore raw

PATCH 209/332 bpf: sockmap: fix tail fragment offset in bpf_msg_push_data

In reply to Greg Kroah-Hartman

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

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

From: Yuqi Xu <[email protected]>

commit f72eed9b84fb771019a955908132410a9ba9ea3f upstream.

When bpf_msg_push_data() inserts data in the middle of a scatterlist entry, it splits the original entry into a left fragment and a right fragment.

The right fragment offset is page-local, but the code advances it with start, which is the message-global insertion point. For inserts into a non-first SG entry, this over-advances the offset and leaves the split layout inconsistent.

Advance the right fragment offset by the fragment-local delta, start - offset, which matches the length removed from the front of the original entry.

---
 net/core/filter.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
net/core/filter.c +1 -1
@@ -2865,7 +2865,7 @@ BPF_CALL_4(bpf_msg_push_data, struct sk_
 
 		psge->length = start - offset;
 		rsge.length -= psge->length;
-		rsge.offset += start;
+		rsge.offset += start - offset;
 
 		sk_msg_iter_var_next(i);
 		sg_unmark_end(psge);
author maintainer
#223 lore raw

PATCH 210/332 macsec: fix replay protection at XPN lower-PN wrap

In reply to Greg Kroah-Hartman

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

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

From: Junrui Luo <[email protected]>

commit e68842b3356471ba56c882209f324613dac47f64 upstream.

In macsec_post_decrypt(), when pn is U32_MAX, pn + 1 overflows u32 to 0 and the first branch never fires. If next_pn_halves.lower is also in the upper half, pn_same_half(pn, lower) is true and the XPN else-if does not fire either, leaving next_pn_halves unchanged. An attacker that captures the legitimate frame carrying pn == 0xFFFFFFFF on an XPN association can then replay it indefinitely, since lowest_pn never rises above the captured pn and macsec_decrypt() reconstructs the same IV.

Extend the XPN else-if to also fire when pn + 1 wraps to 0, so receipt of pn == U32_MAX advances next_pn_halves to (upper + 1, 0).

---
 drivers/net/macsec.c |    3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/net/macsec.c +2 -1
@@ -804,7 +804,8 @@ static bool macsec_post_decrypt(struct s
 		if (pn + 1 > rx_sa->next_pn_halves.lower) {
 			rx_sa->next_pn_halves.lower = pn + 1;
 		} else if (secy->xpn &&
-			   !pn_same_half(pn, rx_sa->next_pn_halves.lower)) {
+			   (pn + 1 == 0 ||
+			    !pn_same_half(pn, rx_sa->next_pn_halves.lower))) {
 			rx_sa->next_pn_halves.upper++;
 			rx_sa->next_pn_halves.lower = pn + 1;
 		}
author maintainer
#224 lore raw

PATCH 211/332 ipv6: exthdrs: refresh nh pointer after ipv6_hop_jumbo()

In reply to Greg Kroah-Hartman

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

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

From: Justin Iurman <[email protected]>

commit d47548a36639095939f4747d4c43f2271366f565 upstream.

ipv6_hop_jumbo() calls pskb_trim_rcsum(), which can change skb pointers. Let's recompute nh pointer to make sure any change won't mess things up.

---
 net/ipv6/exthdrs.c |    2 ++
 1 file changed, 2 insertions(+)
net/ipv6/exthdrs.c +2 -0
@@ -184,6 +184,8 @@ static bool ip6_parse_tlv(bool hopbyhop,
 				case IPV6_TLV_JUMBO:
 					if (!ipv6_hop_jumbo(skb, off))
 						return false;
+
+					nh = skb_network_header(skb);
 					break;
 				case IPV6_TLV_CALIPSO:
 					if (!ipv6_hop_calipso(skb, off))
author maintainer
#225 lore raw

PATCH 212/332 ASoC: qcom: q6asm-dai: fix error handling in prepare and set_params

In reply to Greg Kroah-Hartman

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

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

From: Srinivas Kandagatla <[email protected]>

commit 4b4db09f283df65d780bc7cee66cb4a7e9bf4770 upstream.

Fix error handling in q6asm_dai_compr_set_params() and q6asm_dai_prepare() for both CMD_CLOSE and q6asm_unmap_memory_regions().

In both the functions, we are doing q6asm_audio_client_free in failure cases, which means if prepare or set_params fail, we can never recover. Now open and close are done in respective dai_open/close functions.

---
 sound/soc/qcom/qdsp6/q6asm-dai.c |   24 ++++++++++++++++--------
 1 file changed, 16 insertions(+), 8 deletions(-)
sound/soc/qcom/qdsp6/q6asm-dai.c +16 -8
@@ -227,9 +227,19 @@ static int q6asm_dai_prepare(struct snd_
 	/* rate and channels are sent to audio driver */
 	if (prtd->state == Q6ASM_STREAM_RUNNING) {
 		/* clear the previous setup if any  */
-		q6asm_cmd(prtd->audio_client, prtd->stream_id, CMD_CLOSE);
-		q6asm_unmap_memory_regions(substream->stream,
-					   prtd->audio_client);
+		ret = q6asm_cmd(prtd->audio_client, prtd->stream_id, CMD_CLOSE);
+		if (ret < 0) {
+			dev_err(dev, "Failed to close q6asm stream %d\n", prtd->stream_id);
+			return ret;
+		}
+
+		ret = q6asm_unmap_memory_regions(substream->stream, prtd->audio_client);
+		if (ret < 0) {
+			dev_err(dev, "Failed to unmap memory regions for q6asm stream %d\n",
+				prtd->stream_id);
+			return ret;
+		}
+
 		q6routing_stream_close(soc_prtd->dai_link->id,
 					 substream->stream);
 		prtd->state = Q6ASM_STREAM_STOPPED;
@@ -297,8 +307,6 @@ routing_err:
 	q6asm_cmd(prtd->audio_client, prtd->stream_id,  CMD_CLOSE);
 open_err:
 	q6asm_unmap_memory_regions(substream->stream, prtd->audio_client);
-	q6asm_audio_client_free(prtd->audio_client);
-	prtd->audio_client = NULL;
 
 	return ret;
 }
@@ -916,7 +924,7 @@ static int q6asm_dai_compr_set_params(st
 			      prtd->session_id, dir);
 	if (ret) {
 		dev_err(dev, "Stream reg failed ret:%d\n", ret);
-		goto q6_err;
+		goto routing_err;
 	}
 
 	ret = __q6asm_dai_compr_set_codec_params(component, stream,
@@ -942,11 +950,11 @@ static int q6asm_dai_compr_set_params(st
 	return 0;
 
 q6_err:
+	q6routing_stream_close(rtd->dai_link->id, dir);
+routing_err:
 	q6asm_cmd(prtd->audio_client, prtd->stream_id, CMD_CLOSE);
 
 open_err:
-	q6asm_audio_client_free(prtd->audio_client);
-	prtd->audio_client = NULL;
 	return ret;
 }
 
author maintainer
#226 lore raw

PATCH 213/332 octeontx2-af: validate body pcifunc in rvu_mbox_handler_rep_event_notify

In reply to Greg Kroah-Hartman

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

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

From: Michael Bommarito <[email protected]>

commit 2156a29aecfffa2eb7c558255690084efbe9f3b0 upstream.

rvu_mbox_handler_rep_event_notify() in drivers/net/ethernet/marvell/ octeontx2/af/rvu_rep.c queues a sender-controlled REP_EVENT_NOTIFY request body verbatim, and rvu_rep_up_notify() then forwards event->pcifunc (the nested body field, distinct from the AF-normalised header pcifunc) into rvu_get_pfvf(), rvu_get_pf() and the AF->PF mailbox device index without any bounds check.

A VF attached to a PF that has been put into switchdev representor mode reaches this path: the VF mailbox handler otx2_pfvf_mbox_handler() forwards every message id including MBOX_MSG_REP_EVENT_NOTIFY to AF without an allowlist, and the AF dispatcher rewrites only msg->pcifunc, leaving struct rep_event::pcifunc attacker-controlled. The sibling rvu_mbox_handler_esw_cfg() refuses requests whose header pcifunc is not rvu->rep_pcifunc; this handler has no equivalent gate.

An out-of-range body pcifunc selects an &rvu->pf[]/&rvu->hwvf[] element past the allocated array and, for RVU_EVENT_MAC_ADDR_CHANGE, turns into a six-byte attacker-chosen OOB ether_addr_copy() target inside the queued worker; KASAN reports a slab-out-of-bounds write in rvu_rep_wq_handler.

Reject malformed requests at the handler entry by gating on is_pf_func_valid(), which is already the canonical PF/VF range check in this driver; expose it via rvu.h so callers in rvu_rep.c can use it instead of open-coding the same range arithmetic.

Fixes b8fea84a0468 ("octeontx2-pf: Add support to sync link state between representor and VFs")
---
 drivers/net/ethernet/marvell/octeontx2/af/rvu.c     |    2 +-
 drivers/net/ethernet/marvell/octeontx2/af/rvu.h     |    1 +
 drivers/net/ethernet/marvell/octeontx2/af/rvu_rep.c |    8 ++++++++
 3 files changed, 10 insertions(+), 1 deletion(-)
drivers/net/ethernet/marvell/octeontx2/af/rvu.c +1 -1
@@ -435,7 +435,7 @@ struct rvu_pfvf *rvu_get_pfvf(struct rvu
 		return &rvu->pf[rvu_get_pf(rvu->pdev, pcifunc)];
 }
 
-static bool is_pf_func_valid(struct rvu *rvu, u16 pcifunc)
+bool is_pf_func_valid(struct rvu *rvu, u16 pcifunc)
 {
 	int pf, vf, nvfs;
 	u64 cfg;
drivers/net/ethernet/marvell/octeontx2/af/rvu.h +1 -0
@@ -910,6 +910,7 @@ u16 rvu_get_rsrc_mapcount(struct rvu_pfv
 struct rvu_pfvf *rvu_get_pfvf(struct rvu *rvu, int pcifunc);
 void rvu_get_pf_numvfs(struct rvu *rvu, int pf, int *numvfs, int *hwvf);
 bool is_block_implemented(struct rvu_hwinfo *hw, int blkaddr);
+bool is_pf_func_valid(struct rvu *rvu, u16 pcifunc);
 bool is_pffunc_map_valid(struct rvu *rvu, u16 pcifunc, int blktype);
 int rvu_get_lf(struct rvu *rvu, struct rvu_block *block, u16 pcifunc, u16 slot);
 int rvu_lf_reset(struct rvu *rvu, struct rvu_block *block, int lf);
drivers/net/ethernet/marvell/octeontx2/af/rvu_rep.c +8 -0
@@ -97,6 +97,14 @@ int rvu_mbox_handler_rep_event_notify(st
 {
 	struct rep_evtq_ent *qentry;
 
+	/* The mailbox dispatcher normalises only the header pcifunc; the
+	 * nested struct rep_event::pcifunc body field is sender-controlled
+	 * and is later used by rvu_rep_up_notify() to index rvu->pf[] /
+	 * rvu->hwvf[].  Reject out-of-range body selectors before queueing.
+	 */
+	if (!is_pf_func_valid(rvu, req->pcifunc))
+		return -EINVAL;
+
 	qentry = kmalloc_obj(*qentry, GFP_ATOMIC);
 	if (!qentry)
 		return -ENOMEM;
author maintainer
#227 lore raw

PATCH 214/332 ipv6: exthdrs: refresh nh after handling HAO option

In reply to Greg Kroah-Hartman

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

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

From: Zhengchuan Liang <[email protected]>

commit f7b52afe3592eae66e160586b45a3f2242972c63 upstream.

ip6_parse_tlv() caches skb_network_header(skb) in nh while walking IPv6 TLVs.

ipv6_dest_hao() may call pskb_expand_head() for a cloned skb, which can move the skb head and invalidate the cached network header pointer. Refresh nh after ipv6_dest_hao() returns so any trailing padding or TLVs are parsed from the current skb head.

This matches the existing pattern used in ip6_parse_tlv() after helpers that can modify skb header storage.

---
 net/ipv6/exthdrs.c |    2 ++
 1 file changed, 2 insertions(+)
net/ipv6/exthdrs.c +2 -0
@@ -203,6 +203,8 @@ static bool ip6_parse_tlv(bool hopbyhop,
 				case IPV6_TLV_HAO:
 					if (!ipv6_dest_hao(skb, off))
 						return false;
+
+					nh = skb_network_header(skb);
 					break;
 #endif
 				default:
author maintainer
#228 lore raw

PATCH 215/332 ip6: vti: Use ip6_tnl.net in vti6_siocdevprivate().

In reply to Greg Kroah-Hartman

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

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

From: Maoyi Xie <[email protected]>

commit 8b484efd5cb4eeef9021a661e198edc5349dacf6 upstream.

After patch 1/2 in this series, vti6_update() unlinks and relinks the tunnel through t->net. vti6_siocdevprivate() still uses dev_net(dev) for the collision lookup. For a tunnel moved through IFLA_NET_NS_FD, dev_net(dev) is the new netns, not t->net.

SIOCCHGTUNNEL on a migrated tunnel then runs:

net = dev_net(dev) /* migrated netns */
t = vti6_locate(net, &p1, false) /* misses target in t->net */
...
t = netdev_priv(dev)
vti6_update(t, &p1, false) /* mutates t->net's hash */

A caller in the migrated netns picks params that match a tunnel in the creation netns. The lookup in dev_net(dev) finds nothing. vti6_update() prepends the migrated tunnel at the head of the creation netns hash bucket for those params. Later lookups in the creation netns resolve to the migrated device. xfrm receive delivers the matched packets through a device the caller controls.

Reachable from an unprivileged user namespace (unshare --user --map-root-user --net). Cross tenant scope on container hosts.

Switch the SIOCCHGTUNNEL path on a non fallback device to use t->net for the lookup. The lookup now matches the netns vti6_update() operates on.

Also add ns_capable(self->net->user_ns, CAP_NET_ADMIN) before the lookup. The check at the top of the case is against dev_net(dev)->user_ns, which after migration is the attacker's netns. A caller there can pick params absent from self->net, the lookup returns NULL, t becomes self, and vti6_update() inserts the device into the creation netns hash. The new check requires CAP_NET_ADMIN in the creation netns user_ns too.

SIOCADDTUNNEL and SIOCCHGTUNNEL on the fallback device keep dev_net(dev), which equals init_net there.

---
 net/ipv6/ip6_vti.c |   11 +++++++++--
 1 file changed, 9 insertions(+), 2 deletions(-)
net/ipv6/ip6_vti.c +9 -2
@@ -834,17 +834,24 @@ vti6_siocdevprivate(struct net_device *d
 		if (p.proto != IPPROTO_IPV6  && p.proto != 0)
 			break;
 		vti6_parm_from_user(&p1, &p);
-		t = vti6_locate(net, &p1, cmd == SIOCADDTUNNEL);
 		if (dev != ip6n->fb_tnl_dev && cmd == SIOCCHGTUNNEL) {
+			struct ip6_tnl *self = netdev_priv(dev);
+
+			err = -EPERM;
+			if (!ns_capable(self->net->user_ns, CAP_NET_ADMIN))
+				break;
+			t = vti6_locate(self->net, &p1, false);
 			if (t) {
 				if (t->dev != dev) {
 					err = -EEXIST;
 					break;
 				}
 			} else
-				t = netdev_priv(dev);
+				t = self;
 
 			err = vti6_update(t, &p1, false);
+		} else {
+			t = vti6_locate(net, &p1, cmd == SIOCADDTUNNEL);
 		}
 		if (t) {
 			err = 0;
author maintainer
#229 lore raw

PATCH 216/332 ipv6: validate extension header length before copying to cmsg

In reply to Greg Kroah-Hartman

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

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

From: Qi Tang <[email protected]>

commit dd433671fef381fdaf7b530c631e6b782d66e224 upstream.

ip6_datagram_recv_specific_ctl() builds IPV6_{HOPOPTS,DSTOPTS,RTHDR} cmsgs (and their IPV6_2292* legacy counterparts) by trusting the on-wire hdrlen byte (ptr[1]) when computing the put_cmsg() length. The length was validated only at parse time (ipv6_parse_hopopts(), etc.). An nftables payload-write expression can rewrite hdrlen after parsing and before the skb reaches recvmsg; the write itself is in-bounds but put_cmsg() then reads up to ((hdrlen+1) << 3) = 2040 bytes from an 8-byte header. nftables is reachable from an unprivileged user namespace, so this is an unprivileged slab-out-of-bounds read:

  BUG: KASAN: slab-out-of-bounds in put_cmsg+0x3ac/0x540
   put_cmsg+0x3ac/0x540
   udpv6_recvmsg+0xca0/0x1250
   sock_recvmsg+0xdf/0x190
   ____sys_recvmsg+0x1b1/0x620

Add ipv6_get_exthdr_len() which validates that at least two bytes are accessible before reading the hdrlen field, then checks the computed length against skb_tail_pointer(skb), returning 0 on failure. Extension headers are kept in the linear skb area by pskb_may_pull() during input, so skb_tail_pointer() is the correct bound.

Use ipv6_get_exthdr_len() at all non-AH call sites: the five standalone cmsg blocks (HbH, 2292HbH, 2292DSTOPTS x2, 2292RTHDR) and the three standard cases in the extension-header walk loop (DSTOPTS, ROUTING, default). AH retains an inline bounds check because its length formula differs ((ptr[1]+2)<<2).

The walk loop also gets a pre-read bounds check at the top to validate ptr before any case accesses ptr[0] or ptr[1].

When the walk loop detects a corrupted header, return from the function instead of continuing to process later socket options.

---
 net/ipv6/datagram.c |   54 ++++++++++++++++++++++++++++++++++++++++++++--------
 1 file changed, 46 insertions(+), 8 deletions(-)
net/ipv6/datagram.c +46 -8
@@ -618,6 +618,18 @@ void ip6_datagram_recv_common_ctl(struct
 	}
 }
 
+static u16 ipv6_get_exthdr_len(const struct sk_buff *skb, const u8 *ptr)
+{
+	u16 len;
+
+	if (ptr + 2 > skb_tail_pointer(skb))
+		return 0;
+
+	len = (ptr[1] + 1) << 3;
+
+	return (len <= skb_tail_pointer(skb) - ptr) ? len : 0;
+}
+
 void ip6_datagram_recv_specific_ctl(struct sock *sk, struct msghdr *msg,
 				    struct sk_buff *skb)
 {
@@ -644,7 +656,10 @@ void ip6_datagram_recv_specific_ctl(stru
 	/* HbH is allowed only once */
 	if (np->rxopt.bits.hopopts && (opt->flags & IP6SKB_HOPBYHOP)) {
 		u8 *ptr = nh + sizeof(struct ipv6hdr);
-		put_cmsg(msg, SOL_IPV6, IPV6_HOPOPTS, (ptr[1]+1)<<3, ptr);
+		u16 len = ipv6_get_exthdr_len(skb, ptr);
+
+		if (len)
+			put_cmsg(msg, SOL_IPV6, IPV6_HOPOPTS, len, ptr);
 	}
 
 	if (opt->lastopt &&
@@ -665,26 +680,37 @@ void ip6_datagram_recv_specific_ctl(stru
 			unsigned int len;
 			u8 *ptr = nh + off;
 
+			if (ptr + 2 > skb_tail_pointer(skb))
+				return;
+
 			switch (nexthdr) {
 			case IPPROTO_DSTOPTS:
 				nexthdr = ptr[0];
-				len = (ptr[1] + 1) << 3;
+				len = ipv6_get_exthdr_len(skb, ptr);
+				if (!len)
+					return;
 				if (np->rxopt.bits.dstopts)
 					put_cmsg(msg, SOL_IPV6, IPV6_DSTOPTS, len, ptr);
 				break;
 			case IPPROTO_ROUTING:
 				nexthdr = ptr[0];
-				len = (ptr[1] + 1) << 3;
+				len = ipv6_get_exthdr_len(skb, ptr);
+				if (!len)
+					return;
 				if (np->rxopt.bits.srcrt)
 					put_cmsg(msg, SOL_IPV6, IPV6_RTHDR, len, ptr);
 				break;
 			case IPPROTO_AH:
 				nexthdr = ptr[0];
 				len = (ptr[1] + 2) << 2;
+				if (ptr + len > skb_tail_pointer(skb))
+					return;
 				break;
 			default:
 				nexthdr = ptr[0];
-				len = (ptr[1] + 1) << 3;
+				len = ipv6_get_exthdr_len(skb, ptr);
+				if (!len)
+					return;
 				break;
 			}
 
@@ -706,19 +732,31 @@ void ip6_datagram_recv_specific_ctl(stru
 	}
 	if (np->rxopt.bits.ohopopts && (opt->flags & IP6SKB_HOPBYHOP)) {
 		u8 *ptr = nh + sizeof(struct ipv6hdr);
-		put_cmsg(msg, SOL_IPV6, IPV6_2292HOPOPTS, (ptr[1]+1)<<3, ptr);
+		u16 len = ipv6_get_exthdr_len(skb, ptr);
+
+		if (len)
+			put_cmsg(msg, SOL_IPV6, IPV6_2292HOPOPTS, len, ptr);
 	}
 	if (np->rxopt.bits.odstopts && opt->dst0) {
 		u8 *ptr = nh + opt->dst0;
-		put_cmsg(msg, SOL_IPV6, IPV6_2292DSTOPTS, (ptr[1]+1)<<3, ptr);
+		u16 len = ipv6_get_exthdr_len(skb, ptr);
+
+		if (len)
+			put_cmsg(msg, SOL_IPV6, IPV6_2292DSTOPTS, len, ptr);
 	}
 	if (np->rxopt.bits.osrcrt && opt->srcrt) {
 		struct ipv6_rt_hdr *rthdr = (struct ipv6_rt_hdr *)(nh + opt->srcrt);
-		put_cmsg(msg, SOL_IPV6, IPV6_2292RTHDR, (rthdr->hdrlen+1) << 3, rthdr);
+		u16 len = ipv6_get_exthdr_len(skb, (u8 *)rthdr);
+
+		if (len)
+			put_cmsg(msg, SOL_IPV6, IPV6_2292RTHDR, len, rthdr);
 	}
 	if (np->rxopt.bits.odstopts && opt->dst1) {
 		u8 *ptr = nh + opt->dst1;
-		put_cmsg(msg, SOL_IPV6, IPV6_2292DSTOPTS, (ptr[1]+1)<<3, ptr);
+		u16 len = ipv6_get_exthdr_len(skb, ptr);
+
+		if (len)
+			put_cmsg(msg, SOL_IPV6, IPV6_2292DSTOPTS, len, ptr);
 	}
 	if (np->rxopt.bits.rxorigdstaddr) {
 		struct sockaddr_in6 sin6;
author maintainer
#230 lore raw

PATCH 217/332 xfrm: input: hold netns during deferred transport reinjection

In reply to Greg Kroah-Hartman

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

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

From: Zhengchuan Liang <[email protected]>

commit c16f74dc1d75d0e2e7670076d5375deda110ebeb upstream.

Transport-mode reinjection stores a struct net pointer in skb->cb and uses it later from xfrm_trans_reinject(). That pointer must stay valid until the deferred callback runs.

Take a netns reference when queueing deferred reinjection work and drop it after the callback completes. Use maybe_get_net() so the queueing path does not revive a namespace that is already being torn down.

This keeps the existing workqueue design and fixes the netns lifetime handling in one place for all users of xfrm_trans_queue_net().

Fixes 7b3801927e52 ("xfrm: introduce xfrm_trans_queue_net")
Assisted-by Codex:gpt-5.4
---
 net/xfrm/xfrm_input.c |   16 ++++++++++++----
 1 file changed, 12 insertions(+), 4 deletions(-)
net/xfrm/xfrm_input.c +12 -4
@@ -797,9 +797,12 @@ static void xfrm_trans_reinject(struct w
 	spin_unlock_bh(&trans->queue_lock);
 
 	local_bh_disable();
-	while ((skb = __skb_dequeue(&queue)))
-		XFRM_TRANS_SKB_CB(skb)->finish(XFRM_TRANS_SKB_CB(skb)->net,
-					       NULL, skb);
+	while ((skb = __skb_dequeue(&queue))) {
+		struct net *net = XFRM_TRANS_SKB_CB(skb)->net;
+
+		XFRM_TRANS_SKB_CB(skb)->finish(net, NULL, skb);
+		put_net(net);
+	}
 	local_bh_enable();
 }
 
@@ -808,6 +811,7 @@ int xfrm_trans_queue_net(struct net *net
 				       struct sk_buff *))
 {
 	struct xfrm_trans_tasklet *trans;
+	struct net *hold_net;
 
 	trans = this_cpu_ptr(&xfrm_trans_tasklet);
 
@@ -816,8 +820,12 @@ int xfrm_trans_queue_net(struct net *net
 
 	BUILD_BUG_ON(sizeof(struct xfrm_trans_cb) > sizeof(skb->cb));
 
+	hold_net = maybe_get_net(net);
+	if (!hold_net)
+		return -ENODEV;
+
 	XFRM_TRANS_SKB_CB(skb)->finish = finish;
-	XFRM_TRANS_SKB_CB(skb)->net = net;
+	XFRM_TRANS_SKB_CB(skb)->net = hold_net;
 	spin_lock_bh(&trans->queue_lock);
 	__skb_queue_tail(&trans->queue, skb);
 	spin_unlock_bh(&trans->queue_lock);
author maintainer
#231 lore raw

PATCH 218/332 l2tp: use refcount_inc_not_zero in l2tp_session_get_by_ifname

In reply to Greg Kroah-Hartman

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

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

From: Michael Bommarito <[email protected]>

commit 05f95729ca844704d15e49ce14868af4b403b32b upstream.

A reader in l2tp_session_get_by_ifname() can return a pointer to a session whose refcount has reached zero. The getter takes its reference with plain refcount_inc(), but every other session getter in the same file (l2tp_v2_session_get, l2tp_v3_session_get, and the corresponding _get_next variants) uses refcount_inc_not_zero() because the IDR/RCU lookup can race with refcount_dec_and_test() -> l2tp_session_free() -> kfree_rcu(). The ifname getter is the only outlier; the inconsistency was raised on-list after 979c017803c4 ("l2tp: use list_del_rcu in l2tp_session_unhash").

A reader inside rcu_read_lock_bh() that matches session->ifname can be preempted between the strcmp() and the refcount_inc(). If the last reference drops on another CPU in that window, the reader's refcount_inc() runs on a counter that has reached zero. refcount_t catches the addition-on-zero, prints "refcount_t: addition on 0; use-after-free", saturates the counter, and returns the saturated pointer to the caller. Session memory is held live by the in-flight RCU read section, but the kfree_rcu() callback queued from l2tp_session_free() will free it once the grace period closes; a caller that dereferences the returned session past that point hits a slab-use-after-free. On PREEMPT_RT local_bh_disable() is a per-CPU sleeping lock and the preemption window is real; on stock PREEMPT kernels local_bh_disable() is a preempt_count increment that closes the cross-CPU race in practice (see below).

Use refcount_inc_not_zero() and continue the list walk on failure, matching the other session getters in the file. The ifname getter is the only session getter in net/l2tp/ that still uses the bare refcount_inc() pattern; this change restores file-internal consistency. The success path is unchanged.

---
 net/l2tp/l2tp_core.c |   11 ++++++-----
 1 file changed, 6 insertions(+), 5 deletions(-)
net/l2tp/l2tp_core.c +6 -5
@@ -441,12 +441,13 @@ struct l2tp_session *l2tp_session_get_by
 	idr_for_each_entry_ul(&pn->l2tp_tunnel_idr, tunnel, tmp, tunnel_id) {
 		if (tunnel) {
 			list_for_each_entry_rcu(session, &tunnel->session_list, list) {
-				if (!strcmp(session->ifname, ifname)) {
-					refcount_inc(&session->ref_count);
-					rcu_read_unlock_bh();
+				if (strcmp(session->ifname, ifname))
+					continue;
+				if (!refcount_inc_not_zero(&session->ref_count))
+					continue;
+				rcu_read_unlock_bh();
 
-					return session;
-				}
+				return session;
 			}
 		}
 	}
author maintainer
#232 lore raw

PATCH 219/332 ip6: vti: Use ip6_tnl.net in vti6_changelink().

In reply to Greg Kroah-Hartman

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

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

From: Kuniyuki Iwashima <[email protected]>

commit 11b326fb0a374f4654f9be22d0f0f7abd9f7d3fe upstream.

ip netns add ns1
ip netns add ns2
ip -n ns1 link add vti6_test type vti6 remote ::1 local ::2 key 7 ip -n ns1 link set vti6_test netns ns2
ip -n ns2 link set vti6_test type vti6 remote ::3 local ::4 key 9 ip netns del ns2
ip netns del ns1

[  132.495484] ------------[ cut here ]------------
[  132.497609] kernel BUG at net/core/dev.c:12376!

Commit 61220ab34948 ("vti6: Enable namespace changing") dropped NETIF_F_NETNS_LOCAL from vti6 devices. A vti6 tunnel can then move through IFLA_NET_NS_FD. After the move dev_net(dev) points at the new netns while t->net stays at the creation netns.

vti6_changelink() and vti6_update() still use dev_net(dev) and dev_net(t->dev). They unlink from one per netns hash and relink into another. The creation netns is left with a stale entry. cleanup_net() of that netns later walks freed memory.

Reachable from an unprivileged user namespace (unshare --user --map-root-user --net). Cross tenant scope on container hosts.

---
 net/ipv6/ip6_vti.c |   12 +++++++-----
 1 file changed, 7 insertions(+), 5 deletions(-)
net/ipv6/ip6_vti.c +7 -5
@@ -722,10 +722,11 @@ vti6_tnl_change(struct ip6_tnl *t, const
 static int vti6_update(struct ip6_tnl *t, struct __ip6_tnl_parm *p,
 		       bool keep_mtu)
 {
-	struct net *net = dev_net(t->dev);
-	struct vti6_net *ip6n = net_generic(net, vti6_net_id);
+	struct net *net = t->net;
+	struct vti6_net *ip6n;
 	int err;
 
+	ip6n = net_generic(net, vti6_net_id);
 	vti6_tnl_unlink(ip6n, t);
 	synchronize_net();
 	err = vti6_tnl_change(t, p, keep_mtu);
@@ -1038,11 +1039,12 @@ static int vti6_changelink(struct net_de
 			   struct nlattr *data[],
 			   struct netlink_ext_ack *extack)
 {
-	struct ip6_tnl *t;
+	struct ip6_tnl *t = netdev_priv(dev);
+	struct net *net = t->net;
 	struct __ip6_tnl_parm p;
-	struct net *net = dev_net(dev);
-	struct vti6_net *ip6n = net_generic(net, vti6_net_id);
+	struct vti6_net *ip6n;
 
+	ip6n = net_generic(net, vti6_net_id);
 	if (dev == ip6n->fb_tnl_dev)
 		return -EINVAL;
 
author maintainer
#233 lore raw

PATCH 220/332 net: skbuff: fix missing zerocopy reference in pskb_carve helpers

In reply to Greg Kroah-Hartman

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

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

From: Minh Nguyen <[email protected]>

commit 98d0912e9f841e5529a5b89a972805f34cb1c69d upstream.

pskb_carve_inside_header() and pskb_carve_inside_nonlinear() both copy the old skb_shared_info header into a new buffer via memcpy(), which includes the destructor_arg pointer (uarg) for MSG_ZEROCOPY skbs. Neither function calls net_zcopy_get() for the new shinfo, creating an unaccounted holder: every skb_shared_info with destructor_arg set will call skb_zcopy_clear() once when freed, but the corresponding net_zcopy_get() was never called for the new copy. Repeated calls drive uarg->refcnt to zero prematurely, freeing ubuf_info_msgzc while TX skbs still hold live destructor_arg pointers.

KASAN reports use-after-free on a freed ubuf_info_msgzc:

  BUG: KASAN: slab-use-after-free in skb_release_data+0x77b/0x810
  Read of size 8 at addr ffff88801574d3e8 by task poc/220

  Call Trace:
   skb_release_data+0x77b/0x810
   kfree_skb_list_reason+0x13e/0x610
   skb_release_data+0x4cd/0x810
   sk_skb_reason_drop+0xf3/0x340
   skb_queue_purge_reason+0x282/0x440
   rds_tcp_inc_free+0x1e/0x30
   rds_recvmsg+0x354/0x1780
   __sys_recvmsg+0xdf/0x180

  Allocated by task 219:
   msg_zerocopy_realloc+0x157/0x7b0
   tcp_sendmsg_locked+0x2892/0x3ba0

  Freed by task 219:
   ip_recv_error+0x74a/0xb10
   tcp_recvmsg+0x475/0x530

The skb consuming the late access still referenced the same uarg via shinfo->destructor_arg copied by pskb_carve_inside_nonlinear() without a refcount bump. This has been verified to be reliably exploitable: a working proof-of-concept achieves full root privilege escalation from an unprivileged local user on a default kernel configuration.

The fix follows the pattern of pskb_expand_head() which has the same memcpy/cloned structure. For pskb_carve_inside_header(), net_zcopy_get() is placed after skb_orphan_frags() succeeds, so the orphan error path needs no cleanup. For pskb_carve_inside_nonlinear(), net_zcopy_get() is placed after all failure points and just before skb_release_data(), so no error path needs cleanup at all -- matching pskb_expand_head() more closely and avoiding the need for a balancing net_zcopy_put().

Fixes 6fa01ccd8830 ("skbuff: Add pskb_extract() helper function")
Assisted-by Claude:claude-sonnet-4-6
---
 net/core/skbuff.c |    4 ++++
 1 file changed, 4 insertions(+)
net/core/skbuff.c +4 -0
@@ -6862,6 +6862,8 @@ static int pskb_carve_inside_header(stru
 			skb_kfree_head(data, size);
 			return -ENOMEM;
 		}
+		if (skb_zcopy(skb))
+			net_zcopy_get(skb_zcopy(skb));
 		for (i = 0; i < skb_shinfo(skb)->nr_frags; i++)
 			skb_frag_ref(skb, i);
 		if (skb_has_frag_list(skb))
@@ -7010,6 +7012,8 @@ static int pskb_carve_inside_nonlinear(s
 		skb_kfree_head(data, size);
 		return -ENOMEM;
 	}
+	if (skb_zcopy(skb))
+		net_zcopy_get(skb_zcopy(skb));
 	skb_release_data(skb, SKB_CONSUMED);
 
 	skb->head = data;
author maintainer
#234 lore raw

PATCH 221/332 spi: spi-mem: avoid mutating op template in spi_mem_supports_op()

In reply to Greg Kroah-Hartman

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

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

From: Santhosh Kumar K <[email protected]>

commit 79378db6a86c7014cce40b65252e6c18f5b8bcc2 upstream.

spi_mem_supports_op() accepts a const struct spi_mem_op pointer but casts away const internally to call spi_mem_adjust_op_freq(). This mutates the caller's op template, which causes stale max_freq values when callers reuse persistent templates - subsequent calls won't re-apply the device frequency cap since spi_mem_adjust_op_freq() skips non-zero values.

Fix by operating on a stack-local copy instead.

Fixes a4f8e70d75dd ("spi: spi-mem: add spi_mem_adjust_op_freq() in spi_mem_supports_op()")
Cc Tianyu Xu <[email protected]>
---
 drivers/spi/spi-mem.c |   15 +++++++++++----
 1 file changed, 11 insertions(+), 4 deletions(-)
drivers/spi/spi-mem.c +11 -4
@@ -279,13 +279,20 @@ static bool spi_mem_internal_supports_op
  */
 bool spi_mem_supports_op(struct spi_mem *mem, const struct spi_mem_op *op)
 {
-	/* Make sure the operation frequency is correct before going futher */
-	spi_mem_adjust_op_freq(mem, (struct spi_mem_op *)op);
+	struct spi_mem_op eval_op = *op;
 
-	if (spi_mem_check_op(op))
+	/*
+	 * Work on a local copy; this is a pure capability check and must
+	 * not modify the caller's op. Stored templates with max_freq == 0
+	 * must remain unset so their frequency is always re-capped to the
+	 * current device maximum at execution time.
+	 */
+	spi_mem_adjust_op_freq(mem, &eval_op);
+
+	if (spi_mem_check_op(&eval_op))
 		return false;
 
-	return spi_mem_internal_supports_op(mem, op);
+	return spi_mem_internal_supports_op(mem, &eval_op);
 }
 EXPORT_SYMBOL_GPL(spi_mem_supports_op);
 
author maintainer
#235 lore raw

PATCH 222/332 HID: wacom: Fix OOB write in wacom_hid_set_device_mode()

In reply to Greg Kroah-Hartman

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

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

From: Lee Jones <[email protected]>

commit c0a8899e02ddebd51e2589835182c239c2e224ae upstream.

wacom_hid_set_device_mode() currently assumes that the HID_DG_INPUTMODE usage is always located in the first field (field[0]) of the feature report. However, a device can specify HID_DG_INPUTMODE in a different field.

If HID_DG_INPUTMODE is in a field other than the first one and the first field has a report_count smaller than the usage_index of HID_DG_INPUTMODE, this leads to an out-of-bounds write to r->field[0]->value.

Fix this by storing the field index of HID_DG_INPUTMODE in 'struct hid_data' during feature mapping. In wacom_hid_set_device_mode(), use this stored field index to access the correct field and add bounds checks to ensure both the field index and the value index are within valid ranges before writing.

Fixes 5ae6e89f7409 ("HID: wacom: implement the finger part of the HID generic handling")
---
 drivers/hid/wacom_sys.c |   13 ++++++++++---
 drivers/hid/wacom_wac.h |    1 +
 2 files changed, 11 insertions(+), 3 deletions(-)
drivers/hid/wacom_sys.c +10 -3
@@ -356,6 +356,7 @@ static void wacom_feature_mapping(struct
 
 		hid_data->inputmode = field->report->id;
 		hid_data->inputmode_index = usage->usage_index;
+		hid_data->inputmode_field_index = field->index;
 		break;
 
 	case HID_UP_DIGITIZER:
@@ -571,9 +572,14 @@ static int wacom_hid_set_device_mode(str
 
 	re = &(hdev->report_enum[HID_FEATURE_REPORT]);
 	r = re->report_id_hash[hid_data->inputmode];
-	if (r) {
-		r->field[0]->value[hid_data->inputmode_index] = 2;
-		hid_hw_request(hdev, r, HID_REQ_SET_REPORT);
+	if (r && hid_data->inputmode_field_index >= 0 &&
+	    hid_data->inputmode_field_index < r->maxfield) {
+		struct hid_field *field = r->field[hid_data->inputmode_field_index];
+
+		if (field && hid_data->inputmode_index < field->report_count) {
+			field->value[hid_data->inputmode_index] = 2;
+			hid_hw_request(hdev, r, HID_REQ_SET_REPORT);
+		}
 	}
 	return 0;
 }
@@ -2846,6 +2852,7 @@ static int wacom_probe(struct hid_device
 		return -ENODEV;
 
 	wacom_wac->hid_data.inputmode = -1;
+	wacom_wac->hid_data.inputmode_field_index = -1;
 	wacom_wac->mode_report = -1;
 
 	if (hid_is_usb(hdev)) {
drivers/hid/wacom_wac.h +1 -0
@@ -295,6 +295,7 @@ struct wacom_shared {
 struct hid_data {
 	__s16 inputmode;	/* InputMode HID feature, -1 if non-existent */
 	__s16 inputmode_index;	/* InputMode HID feature index in the report */
+	__s16 inputmode_field_index; /* InputMode HID feature field index in the report */
 	bool sense_state;
 	bool inrange_state;
 	bool eraser;
author maintainer
#236 lore raw

PATCH 223/332 iommu, debugobjects: avoid gcc-16.1 section mismatch warnings

In reply to Greg Kroah-Hartman

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

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

From: Arnd Bergmann <[email protected]>

commit 4c9ad387aa2d6785299722e54224d34764edaeb3 upstream.

gcc-16 has gained some more advanced inter-procedual optimization techniques that enable it to inline the dummy_tlb_add_page() and dummy_tlb_flush() function pointers into a specialized version of __arm_v7s_unmap:

WARNING: modpost: vmlinux: section mismatch in reference: __arm_v7s_unmap+0x2cc (section: .text) -> dummy_tlb_add_page (section: .init.text)
ERROR: modpost: Section mismatches detected.

Quoted 1 line

From what I can tell, the transformation is correct, as this is only

called when __arm_v7s_unmap() is called from arm_v7s_do_selftests(), which is also __init. Since __arm_v7s_unmap() however is not __init, gcc cannot inline the inner function calls directly.

In debug_objects_selftest(), the same thing happens. Both the caller and the leaf function are __init, but the IPA pulls it into a non-init one:

WARNING: modpost: vmlinux: section mismatch in reference: lookup_object_or_alloc+0x7c (section: .text.lookup_object_or_alloc) -> is_static_object (section: .init.text)

Marking the affected functions as not "__init" would reliably avoid this issue but is not a good solution because it removes an otherwise correct annotation. I tried marking the functions as 'noinline', but that ended up not covering all the affected configurations.

With some more experimenting, I found that marking these functions as __attribute__((noipa)) is both logical and reliable.

In order to keep the syntax readable, add a custom macro for this in include/linux/compiler_attributes.h next to other related macros and use it to annotate both files.

---
 drivers/iommu/io-pgtable-arm-v7s.c  |   18 ++++++++++++------
 include/linux/compiler_attributes.h |   11 +++++++++++
 lib/debugobjects.c                  |    2 +-
 3 files changed, 24 insertions(+), 7 deletions(-)
drivers/iommu/io-pgtable-arm-v7s.c +12 -6
@@ -777,21 +777,27 @@ struct io_pgtable_init_fns io_pgtable_ar
 
 static struct io_pgtable_cfg *cfg_cookie __initdata;
 
-static void __init dummy_tlb_flush_all(void *cookie)
+/*
+ * __noipa prevents gcc from turning indirect iommu_flush_ops calls
+ * into direct calls from a specialized __arm_v7s_unmap() that triggers
+ * a build time section mismatch assertion.
+ */
+static __noipa void __init dummy_tlb_flush_all(void *cookie)
 {
 	WARN_ON(cookie != cfg_cookie);
 }
 
-static void __init dummy_tlb_flush(unsigned long iova, size_t size,
-				   size_t granule, void *cookie)
+static __noipa void __init dummy_tlb_flush(unsigned long iova, size_t size,
+					   size_t granule, void *cookie)
 {
 	WARN_ON(cookie != cfg_cookie);
 	WARN_ON(!(size & cfg_cookie->pgsize_bitmap));
 }
 
-static void __init dummy_tlb_add_page(struct iommu_iotlb_gather *gather,
-				      unsigned long iova, size_t granule,
-				      void *cookie)
+static __noipa void __init dummy_tlb_add_page(struct iommu_iotlb_gather *gather,
+					      unsigned long iova,
+					      size_t granule,
+					      void *cookie)
 {
 	dummy_tlb_flush(iova, granule, granule, cookie);
 }
include/linux/compiler_attributes.h +11 -0
@@ -397,6 +397,17 @@
 #endif
 
 /*
+ * Optional: not supported by clang
+ *
+ *   gcc: https://gcc.gnu.org/onlinedocs/gcc/Common-Attributes.html#index-noipa
+ */
+#if __has_attribute(noipa)
+# define __noipa __attribute__((noipa))
+#else
+# define __noipa
+#endif
+
+/*
  *   gcc: https://gcc.gnu.org/onlinedocs/gcc/Common-Function-Attributes.html#index-weak-function-attribute
  *   gcc: https://gcc.gnu.org/onlinedocs/gcc/Common-Variable-Attributes.html#index-weak-variable-attribute
  */
lib/debugobjects.c +1 -1
@@ -1212,7 +1212,7 @@ struct self_test {
 
 static __initconst const struct debug_obj_descr descr_type_test;
 
-static bool __init is_static_object(void *addr)
+static __noipa bool __init is_static_object(void *addr)
 {
 	struct self_test *obj = addr;
 
author maintainer
#237 lore raw

PATCH 224/332 nfc: hci: fix out-of-bounds read in HCP header parsing

In reply to Greg Kroah-Hartman

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

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

From: Ashutosh Desai <[email protected]>

commit f040e590c035bfd9553fe79ee9585caf1b14d67b upstream.

Both nfc_hci_recv_from_llc() and nci_hci_data_received_cb() read packet->header from skb->data at function entry without first checking that the buffer holds at least one byte. A malicious NFC peer can send a 0-byte HCP frame that passes through the SHDLC layer and reaches these functions, causing an out-of-bounds heap read of packet->header. The same 0-byte frame, if queued as a non-final fragment, also causes the reassembly loop to underflow msg_len to UINT_MAX, triggering skb_over_panic() when the reassembled skb is written.

Fix this by adding a pskb_may_pull() check at the entry of each function before packet->header is first accessed. The existing pskb_may_pull() checks before the reassembled hcp_skb is cast to struct hcp_packet remain in place to guard the 2-byte HCP message header.

Fixes 8b8d2e08bf0d ("NFC: HCI support")
Fixes 11f54f228643 ("NFC: nci: Add HCI over NCI protocol support")
---
 net/nfc/hci/core.c |   10 ++++++++++
 net/nfc/nci/hci.c  |   10 ++++++++++
 2 files changed, 20 insertions(+)
net/nfc/hci/core.c +10 -0
@@ -861,6 +861,11 @@ static void nfc_hci_recv_from_llc(struct
 	struct sk_buff *frag_skb;
 	int msg_len;
 
+	if (!pskb_may_pull(skb, NFC_HCI_HCP_PACKET_HEADER_LEN)) {
+		kfree_skb(skb);
+		return;
+	}
+
 	packet = (struct hcp_packet *)skb->data;
 	if ((packet->header & ~NFC_HCI_FRAGMENT) == 0) {
 		skb_queue_tail(&hdev->rx_hcp_frags, skb);
@@ -904,6 +909,11 @@ static void nfc_hci_recv_from_llc(struct
 	 * unblock waiting cmd context. Otherwise, enqueue to dispatch
 	 * in separate context where handler can also execute command.
 	 */
+	if (!pskb_may_pull(hcp_skb, NFC_HCI_HCP_HEADER_LEN)) {
+		kfree_skb(hcp_skb);
+		return;
+	}
+
 	packet = (struct hcp_packet *)hcp_skb->data;
 	type = HCP_MSG_GET_TYPE(packet->message.header);
 	if (type == NFC_HCI_HCP_RESPONSE) {
net/nfc/nci/hci.c +10 -0
@@ -439,6 +439,11 @@ void nci_hci_data_received_cb(void *cont
 		return;
 	}
 
+	if (!pskb_may_pull(skb, NCI_HCI_HCP_PACKET_HEADER_LEN)) {
+		kfree_skb(skb);
+		return;
+	}
+
 	packet = (struct nci_hcp_packet *)skb->data;
 	if ((packet->header & ~NCI_HCI_FRAGMENT) == 0) {
 		skb_queue_tail(&ndev->hci_dev->rx_hcp_frags, skb);
@@ -482,6 +487,11 @@ void nci_hci_data_received_cb(void *cont
 	 * unblock waiting cmd context. Otherwise, enqueue to dispatch
 	 * in separate context where handler can also execute command.
 	 */
+	if (!pskb_may_pull(hcp_skb, NCI_HCI_HCP_HEADER_LEN)) {
+		kfree_skb(hcp_skb);
+		return;
+	}
+
 	packet = (struct nci_hcp_packet *)hcp_skb->data;
 	type = NCI_HCP_MSG_GET_TYPE(packet->message.header);
 	if (type == NCI_HCI_HCP_RESPONSE) {
author maintainer
#238 lore raw

PATCH 225/332 xfrm: route MIGRATE notifications to callers netns

In reply to Greg Kroah-Hartman

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

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

From: Maoyi Xie <[email protected]>

commit 7e2a4f7ca0952820731ef7bdadfc9a9e9d3571b4 upstream.

xfrm_send_migrate() in net/xfrm/xfrm_user.c and pfkey_send_migrate() in net/key/af_key.c both hardcode &init_net for the multicast that announces a successful XFRM_MSG_MIGRATE / SADB_X_MIGRATE.

XFRM_MSG_MIGRATE arrives on a per-netns NETLINK_XFRM socket, and the rest of the xfrm/af_key netlink path was made netns-aware in 2008. The other 14 multicast paths in xfrm_user.c route their event using xs_net(x), xp_net(xp) or sock_net(skb->sk); only the migrate path was missed.

Two consequences of the init_net hardcoding:

1. The notification (selector, old/new endpoint addresses, and the km_address) is delivered to listeners on init_net's XFRMNLGRP_MIGRATE / pfkey BROADCAST_ALL groups rather than on the issuing netns. An IKE daemon running in init_net therefore receives migration notifications originating from any other netns on the host.

2. An IKE daemon running inside a non-init netns and subscribed to its own XFRMNLGRP_MIGRATE / pfkey groups never receives the notification of its own migration. IKEv2 MOBIKE / address-update handling inside a netns is silently broken.

Thread struct net through km_migrate() and the xfrm_mgr.migrate function pointer, drop the &init_net override in xfrm_send_migrate() and pfkey_send_migrate(), and pass the caller's net (already in scope in xfrm_migrate() via sock_net(skb->sk)) all the way down. struct xfrm_mgr is in-tree only and not exported as a stable API, so the function-pointer signature change is internal.

pfkey_broadcast() is already netns-aware via net_generic(net, pfkey_net_id) since the pernet conversion. The five other pfkey_broadcast() callers in af_key.c already pass xs_net(x), sock_net(sk) or a per-netns net, so this only removes the &init_net outlier.

Fixes 5c79de6e79cd ("[XFRM]: User interface for handling XFRM_MSG_MIGRATE")
Cc [email protected] # v5.15+
---
 include/net/xfrm.h     |    3 ++-
 net/key/af_key.c       |    6 +++---
 net/xfrm/xfrm_policy.c |    2 +-
 net/xfrm/xfrm_state.c  |    4 ++--
 net/xfrm/xfrm_user.c   |    5 ++---
 5 files changed, 10 insertions(+), 10 deletions(-)
include/net/xfrm.h +2 -1
@@ -715,6 +715,7 @@ struct xfrm_mgr {
 					   const struct xfrm_migrate *m,
 					   int num_bundles,
 					   const struct xfrm_kmaddress *k,
+					   struct net *net,
 					   const struct xfrm_encap_tmpl *encap);
 	bool			(*is_alive)(const struct km_event *c);
 };
@@ -1891,7 +1892,7 @@ int xfrm_sk_policy_insert(struct sock *s
 #ifdef CONFIG_XFRM_MIGRATE
 int km_migrate(const struct xfrm_selector *sel, u8 dir, u8 type,
 	       const struct xfrm_migrate *m, int num_bundles,
-	       const struct xfrm_kmaddress *k,
+	       const struct xfrm_kmaddress *k, struct net *net,
 	       const struct xfrm_encap_tmpl *encap);
 struct xfrm_state *xfrm_migrate_state_find(struct xfrm_migrate *m, struct net *net,
 						u32 if_id);
net/key/af_key.c +3 -3
@@ -3564,7 +3564,7 @@ static int set_ipsecrequest(struct sk_bu
 #ifdef CONFIG_NET_KEY_MIGRATE
 static int pfkey_send_migrate(const struct xfrm_selector *sel, u8 dir, u8 type,
 			      const struct xfrm_migrate *m, int num_bundles,
-			      const struct xfrm_kmaddress *k,
+			      const struct xfrm_kmaddress *k, struct net *net,
 			      const struct xfrm_encap_tmpl *encap)
 {
 	int i;
@@ -3669,7 +3669,7 @@ static int pfkey_send_migrate(const stru
 	}
 
 	/* broadcast migrate message to sockets */
-	pfkey_broadcast(skb, GFP_ATOMIC, BROADCAST_ALL, NULL, &init_net);
+	pfkey_broadcast(skb, GFP_ATOMIC, BROADCAST_ALL, NULL, net);
 
 	return 0;
 
@@ -3680,7 +3680,7 @@ err:
 #else
 static int pfkey_send_migrate(const struct xfrm_selector *sel, u8 dir, u8 type,
 			      const struct xfrm_migrate *m, int num_bundles,
-			      const struct xfrm_kmaddress *k,
+			      const struct xfrm_kmaddress *k, struct net *net,
 			      const struct xfrm_encap_tmpl *encap)
 {
 	return -ENOPROTOOPT;
net/xfrm/xfrm_policy.c +1 -1
@@ -4704,7 +4704,7 @@ int xfrm_migrate(const struct xfrm_selec
 	}
 
 	/* Stage 5 - announce */
-	km_migrate(sel, dir, type, m, num_migrate, k, encap);
+	km_migrate(sel, dir, type, m, num_migrate, k, net, encap);
 
 	xfrm_pol_put(pol);
 
net/xfrm/xfrm_state.c +2 -2
@@ -2837,7 +2837,7 @@ EXPORT_SYMBOL(km_policy_expired);
 #ifdef CONFIG_XFRM_MIGRATE
 int km_migrate(const struct xfrm_selector *sel, u8 dir, u8 type,
 	       const struct xfrm_migrate *m, int num_migrate,
-	       const struct xfrm_kmaddress *k,
+	       const struct xfrm_kmaddress *k, struct net *net,
 	       const struct xfrm_encap_tmpl *encap)
 {
 	int err = -EINVAL;
@@ -2848,7 +2848,7 @@ int km_migrate(const struct xfrm_selecto
 	list_for_each_entry_rcu(km, &xfrm_km_list, list) {
 		if (km->migrate) {
 			ret = km->migrate(sel, dir, type, m, num_migrate, k,
-					  encap);
+					  net, encap);
 			if (!ret)
 				err = ret;
 		}
net/xfrm/xfrm_user.c +2 -3
@@ -3271,10 +3271,9 @@ out_cancel:
 
 static int xfrm_send_migrate(const struct xfrm_selector *sel, u8 dir, u8 type,
 			     const struct xfrm_migrate *m, int num_migrate,
-			     const struct xfrm_kmaddress *k,
+			     const struct xfrm_kmaddress *k, struct net *net,
 			     const struct xfrm_encap_tmpl *encap)
 {
-	struct net *net = &init_net;
 	struct sk_buff *skb;
 	int err;
 
@@ -3292,7 +3291,7 @@ static int xfrm_send_migrate(const struc
 #else
 static int xfrm_send_migrate(const struct xfrm_selector *sel, u8 dir, u8 type,
 			     const struct xfrm_migrate *m, int num_migrate,
-			     const struct xfrm_kmaddress *k,
+			     const struct xfrm_kmaddress *k, struct net *net,
 			     const struct xfrm_encap_tmpl *encap)
 {
 	return -ENOPROTOOPT;
author maintainer
#239 lore raw

PATCH 226/332 xfrm: ipcomp: Free destination pages on acomp errors

In reply to Greg Kroah-Hartman

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

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

From: Herbert Xu <[email protected]>

commit 7dbac7680eb629b3b4dc7e98c34f943b8814c0c8 upstream.

Move the out_free_req label up by a couple of lines so that the allocated dst SG list gets freed on error as well as success.

---
 net/xfrm/xfrm_ipcomp.c |   12 ++++++++----
 1 file changed, 8 insertions(+), 4 deletions(-)
net/xfrm/xfrm_ipcomp.c +8 -4
@@ -51,11 +51,15 @@ static int ipcomp_post_acomp(struct sk_b
 	struct scatterlist *dsg;
 	int len, dlen;
 
-	if (unlikely(err))
-		goto out_free_req;
+	if (unlikely(!req))
+		return err;
 
 	extra = acomp_request_extra(req);
 	dsg = extra->sg;
+
+	if (unlikely(err))
+		goto out_free_req;
+
 	dlen = req->dlen;
 
 	pskb_trim_unique(skb, 0);
@@ -84,10 +88,10 @@ static int ipcomp_post_acomp(struct sk_b
 		skb_shinfo(skb)->nr_frags++;
 	} while ((dlen -= len));
 
-	for (; dsg; dsg = sg_next(dsg))
+out_free_req:
+	for (; dsg && sg_page(dsg); dsg = sg_next(dsg))
 		__free_page(sg_page(dsg));
 
-out_free_req:
 	acomp_request_free(req);
 	return err;
 }
author maintainer
#240 lore raw

PATCH 227/332 xfrm: ah: use skb_to_full_sk in async output callbacks

In reply to Greg Kroah-Hartman

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

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

From: Michael Bommarito <[email protected]>

commit 79d8be262377f7112cfa3088dfc4142d5a2533f3 upstream.

When AH output is offloaded to an asynchronous crypto provider (hardware accelerators such as AMD CCP, or a forced-async software shim used for testing), the digest completion fires ah_output_done() / ah6_output_done() on a workqueue. The egress skb at that point may have been originated by a TCP listener sending a SYN-ACK, which sets skb->sk to a request_sock via skb_set_owner_edemux(); it may also have been originated by an inet_timewait_sock retransmit. Neither is a full struct sock, and passing the raw skb->sk to xfrm_output_resume() then forwards a non-full socket through the rest of the xfrm output chain.

xfrm_output_resume() and its downstream consumers expect a full sk where they dereference at all. The natural egress path through ah_output_done() does not crash today because the consumers that read past sock_common are either gated by sk_fullsock() or short-circuit on flags that are clear on a fresh request_sock; an exhaustive walk of the 50 most plausible consumers under sch_fq, dev_queue_xmit, netfilter, tc-egress and cgroup-egress BPF found no current unguarded deref. The bug is still a real type confusion that future consumer changes could turn into a memory-corruption primitive.

This is the same bug class fixed for ESP in commit 1620c88887b1 ("xfrm: Fix the usage of skb->sk"). Apply the analogous fix to
AH: convert skb->sk to a full socket pointer (or NULL) via skb_to_full_sk() before handing it to xfrm_output_resume().

The same async AH callbacks were touched recently for an independent ESN-related ICV layout bug in commit ec54093e6a8f ("xfrm: ah: account for ESN high bits in async callbacks"); the sk type-confusion addressed here is orthogonal. This patch is part of an ongoing audit of the AH callback paths; an ah_output ihl-validation hardening series is also currently under review on netdev.

Reproduced under UML + KASAN + lockdep with a forced-async hmac(sha1) shim that registers at priority 9999 and wraps the sync in-tree hmac-sha1-lib. With the shim loaded, ah_output_done runs on every SYN-ACK egress through a transport-mode AH SA and skb->sk arrives as a request_sock (TCP_NEW_SYN_RECV); after this patch, xfrm_output_resume() receives the listener (the result of sk_to_full_sk()) and consumer derefs land on full-sock fields as intended.

Fixes 9ab1265d5231 ("xfrm: Use actual socket sk instead of skb socket for xfrm_output_resume")
Assisted-by Claude:claude-opus-4-7
---
 net/ipv4/ah4.c |    2 +-
 net/ipv6/ah6.c |    2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)
net/ipv4/ah4.c +1 -1
@@ -143,7 +143,7 @@ static void ah_output_done(void *data, i
 	}
 
 	kfree(AH_SKB_CB(skb)->tmp);
-	xfrm_output_resume(skb->sk, skb, err);
+	xfrm_output_resume(skb_to_full_sk(skb), skb, err);
 }
 
 static int ah_output(struct xfrm_state *x, struct sk_buff *skb)
net/ipv6/ah6.c +1 -1
@@ -337,7 +337,7 @@ static void ah6_output_done(void *data,
 	ah6_restore_hdrs(top_iph, iph_ext, extlen);
 
 	kfree(AH_SKB_CB(skb)->tmp);
-	xfrm_output_resume(skb->sk, skb, err);
+	xfrm_output_resume(skb_to_full_sk(skb), skb, err);
 }
 
 static int ah6_output(struct xfrm_state *x, struct sk_buff *skb)
author maintainer
#241 lore raw

PATCH 228/332 ALSA: scarlett2: Fix 2i2 Gen 4 direct monitor gain on firmware 2417

In reply to Greg Kroah-Hartman

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

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

From: Geoffrey D. Bennett <[email protected]>

commit db37cf47b67e38ade40de5cd74a4d4d772ff1416 upstream.

Firmware 2417 for the Scarlett 4th Gen 2i2 moved the direct monitor gain parameter by 4 bytes, from offset 0x2a0 to 0x2a4, breaking the "Direct Monitor X Mix Y" controls.

Special-case the offset in the get/set config helpers when the running firmware is 2417 or later.

Fixes 4e809a299677 ("ALSA: scarlett2: Add support for Solo, 2i2, and 4i4 Gen 4")
---
 sound/usb/mixer_scarlett2.c |   33 +++++++++++++++++++++++++++++----
 1 file changed, 29 insertions(+), 4 deletions(-)
sound/usb/mixer_scarlett2.c +29 -4
@@ -2504,6 +2504,27 @@ static int scarlett2_has_config_item(
 	return !!private->config_set->items[config_item_num].offset;
 }
 
+/* Return the configuration item's offset, applying any per-firmware
+ * overrides.
+ *
+ * Firmware 2417 for the 2i2 Gen 4 moved DIRECT_MONITOR_GAIN by 4
+ * bytes. Apply that shift here so that the rest of the driver can
+ * keep using the single config set. This override can be removed
+ * once the multi-config-set framework lands.
+ */
+static int scarlett2_config_item_offset(
+	struct scarlett2_data *private, int config_item_num)
+{
+	int offset = private->config_set->items[config_item_num].offset;
+
+	if (config_item_num == SCARLETT2_CONFIG_DIRECT_MONITOR_GAIN &&
+	    private->info == &s2i2_gen4_info &&
+	    private->firmware_version >= 2417)
+		offset = 0x2a4;
+
+	return offset;
+}
+
 /* Send a USB message to get configuration parameters; result placed in *buf */
 static int scarlett2_usb_get_config(
 	struct usb_mixer_interface *mixer,
@@ -2513,6 +2534,7 @@ static int scarlett2_usb_get_config(
 	const struct scarlett2_config *config_item =
 		&private->config_set->items[config_item_num];
 	int size, err, i;
+	int item_offset;
 	u8 *buf_8;
 	u8 value;
 
@@ -2522,13 +2544,15 @@ static int scarlett2_usb_get_config(
 	if (!config_item->offset)
 		return -EFAULT;
 
+	item_offset = scarlett2_config_item_offset(private, config_item_num);
+
 	/* Writes to the parameter buffer are always 1 byte */
 	size = config_item->size ? config_item->size : 8;
 
 	/* For byte-sized parameters, retrieve directly into buf */
 	if (size >= 8) {
 		size = size / 8 * count;
-		err = scarlett2_usb_get(mixer, config_item->offset, buf, size);
+		err = scarlett2_usb_get(mixer, item_offset, buf, size);
 		if (err < 0)
 			return err;
 		if (config_item->size == 16) {
@@ -2546,7 +2570,7 @@ static int scarlett2_usb_get_config(
 	}
 
 	/* For bit-sized parameters, retrieve into value */
-	err = scarlett2_usb_get(mixer, config_item->offset, &value, 1);
+	err = scarlett2_usb_get(mixer, item_offset, &value, 1);
 	if (err < 0)
 		return err;
 
@@ -2696,7 +2720,8 @@ static int scarlett2_usb_set_config(
 	 */
 	if (config_item->size >= 8) {
 		size = config_item->size / 8;
-		offset = config_item->offset + index * size;
+		offset = scarlett2_config_item_offset(private, config_item_num) +
+			 index * size;
 
 	/* If updating a bit, retrieve the old value, set/clear the
 	 * bit as needed, and update value
@@ -2705,7 +2730,7 @@ static int scarlett2_usb_set_config(
 		u8 tmp;
 
 		size = 1;
-		offset = config_item->offset;
+		offset = scarlett2_config_item_offset(private, config_item_num);
 
 		err = scarlett2_usb_get(mixer, offset, &tmp, 1);
 		if (err < 0)
author maintainer
#242 lore raw

PATCH 229/332 ALSA: firewire-motu: Protect register DSP event queue positions

In reply to Greg Kroah-Hartman

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

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

From: Cássio Gabriel <[email protected]>

commit 98fb1c1bb11e29eb609b7200a25e136e05aa4498 upstream.

The register DSP event queue is updated under parser->lock, but snd_motu_register_dsp_message_parser_count_event() reads pull_pos and push_pos without the lock. snd_motu_register_dsp_message_parser_copy_event() also reads both queue positions before taking the lock.

Protect these accesses with parser->lock as well. This keeps the hwdep poll/read path consistent with the producer side and with the cached meter/parameter accessors.

---
 sound/firewire/motu/motu-register-dsp-message-parser.c |   11 +++++++----
 1 file changed, 7 insertions(+), 4 deletions(-)
sound/firewire/motu/motu-register-dsp-message-parser.c +7 -4
@@ -386,6 +386,8 @@ unsigned int snd_motu_register_dsp_messa
 {
 	struct msg_parser *parser = motu->message_parser;
 
+	guard(spinlock_irqsave)(&parser->lock);
+
 	if (parser->pull_pos > parser->push_pos)
 		return EVENT_QUEUE_SIZE - parser->pull_pos + parser->push_pos;
 	else
@@ -395,13 +397,14 @@ unsigned int snd_motu_register_dsp_messa
 bool snd_motu_register_dsp_message_parser_copy_event(struct snd_motu *motu, u32 *event)
 {
 	struct msg_parser *parser = motu->message_parser;
-	unsigned int pos = parser->pull_pos;
-
-	if (pos == parser->push_pos)
-		return false;
+	unsigned int pos;
 
 	guard(spinlock_irqsave)(&parser->lock);
 
+	if (parser->pull_pos == parser->push_pos)
+		return false;
+
+	pos = parser->pull_pos;
 	*event = parser->event_queue[pos];
 
 	++pos;
author maintainer
#243 lore raw

PATCH 230/332 netfilter: conntrack: tcp: do not force CLOSE on invalid-seq RST without direction check

In reply to Greg Kroah-Hartman

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

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

From: Hamza Mahfooz <[email protected]>

commit bed6e04be8e6b9133d8b16d5a42d0e0ce674fa9a upstream.

An unintended behavior in the TCP conntrack state machine allows a connection to be forced into the CLOSE state using an RST packet with an invalid sequence number.

Specifically, after a SYN packet is observed, an RST with an invalid SEQ can transition the conntrack entry to TCP_CONNTRACK_CLOSE, regardless of whether the RST corresponds to the expected reply direction. The relevant code path assumes the RST is a response to an outgoing SYN, but does not validate packet direction or ensure that a matching SYN was actually sent in the opposite direction.

As a result, a crafted packet sequence consisting of a SYN followed by an invalid-sequence RST can prematurely terminate an active NAT entry. This makes connection teardown easier than intended.

So, tighten the state transition logic to ensure that RST-triggered CLOSE transitions only occur when the RST is a valid response to a previously observed SYN in the correct direction.

Fixes 9fb9cbb1082d ("[NETFILTER]: Add nf_conntrack subsystem.")
---
 net/netfilter/nf_conntrack_proto_tcp.c |    3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
net/netfilter/nf_conntrack_proto_tcp.c +2 -1
@@ -1221,7 +1221,8 @@ int nf_conntrack_tcp_packet(struct nf_co
 			new_state = old_state;
 		}
 		if (((test_bit(IPS_SEEN_REPLY_BIT, &ct->status)
-			 && ct->proto.tcp.last_index == TCP_SYN_SET)
+			 && ct->proto.tcp.last_index == TCP_SYN_SET
+			 && ct->proto.tcp.last_dir != dir)
 			|| (!test_bit(IPS_ASSURED_BIT, &ct->status)
 			    && ct->proto.tcp.last_index == TCP_ACK_SET))
 		    && ntohl(th->ack_seq) == ct->proto.tcp.last_end) {
author maintainer
#244 lore raw

PATCH 231/332 ASoC: qcom: q6asm-dai: close stream only when running

In reply to Greg Kroah-Hartman

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

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

From: Srinivas Kandagatla <[email protected]>

commit 048c540ee76ded666bda74f9dae1ca3254e0633c upstream.

q6asm_dai_close() and q6asm_dai_compr_free() currently issue CMD_CLOSE whenever prtd->state is non-zero.

After prepare() closes an existing stream, the state is updated to Q6ASM_STREAM_STOPPED. Since this state is also non-zero, the close and free paths can send CMD_CLOSE again for a stream that has already been closed.

Restrict CMD_CLOSE to the Q6ASM_STREAM_RUNNING state so the command is sent only when the ASM stream is still active.

---
 sound/soc/qcom/qdsp6/q6asm-dai.c |   14 +++++++-------
 1 file changed, 7 insertions(+), 7 deletions(-)
sound/soc/qcom/qdsp6/q6asm-dai.c +7 -7
@@ -465,12 +465,12 @@ static int q6asm_dai_close(struct snd_so
 	struct q6asm_dai_rtd *prtd = runtime->private_data;
 
 	if (prtd->audio_client) {
-		if (prtd->state)
+		if (prtd->state == Q6ASM_STREAM_RUNNING) {
 			q6asm_cmd(prtd->audio_client, prtd->stream_id,
 				  CMD_CLOSE);
-
-		q6asm_unmap_memory_regions(substream->stream,
+			q6asm_unmap_memory_regions(substream->stream,
 					   prtd->audio_client);
+		}
 		q6asm_audio_client_free(prtd->audio_client);
 		prtd->audio_client = NULL;
 	}
@@ -682,7 +682,7 @@ static int q6asm_dai_compr_free(struct s
 	struct snd_soc_pcm_runtime *rtd = stream->private_data;
 
 	if (prtd->audio_client) {
-		if (prtd->state) {
+		if (prtd->state == Q6ASM_STREAM_RUNNING) {
 			q6asm_cmd(prtd->audio_client, prtd->stream_id,
 				  CMD_CLOSE);
 			if (prtd->next_track_stream_id) {
@@ -690,11 +690,11 @@ static int q6asm_dai_compr_free(struct s
 					  prtd->next_track_stream_id,
 					  CMD_CLOSE);
 			}
-		}
 
-		snd_dma_free_pages(&prtd->dma_buffer);
-		q6asm_unmap_memory_regions(stream->direction,
+			q6asm_unmap_memory_regions(stream->direction,
 					   prtd->audio_client);
+		}
+		snd_dma_free_pages(&prtd->dma_buffer);
 		q6asm_audio_client_free(prtd->audio_client);
 		prtd->audio_client = NULL;
 	}
author maintainer
#245 lore raw

PATCH 232/332 ASoC: qcom: q6asm-dai: do not set stream state in event and trigger callbacks

In reply to Greg Kroah-Hartman

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

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

From: Srinivas Kandagatla <[email protected]>

commit cee3e63e7106c3c81b2053371fdf14240bfba2fc upstream.

The q6asm-dai stream state is used by prepare() to decide whether an existing stream setup needs to be closed before opening/configuring a new one. Updating the state from trigger or asynchronous DSP callbacks can make that state stale or incorrect relative to the actual setup lifetime.

In particular, setting Q6ASM_STREAM_STOPPED on STOP or EOS completion can make prepare() believe there is no active setup to close, which can result in opening/configuring the same stream more than once.

Keep stream state updates tied to prepare(), where the stream is actually closed and reopened, and stop changing it from trigger and EOS callbacks.

---
 sound/soc/qcom/qdsp6/q6asm-dai.c |    5 -----
 1 file changed, 5 deletions(-)
sound/soc/qcom/qdsp6/q6asm-dai.c +0 -5
@@ -186,7 +186,6 @@ static void event_handler(uint32_t opcod
 	case ASM_CLIENT_EVENT_CMD_RUN_DONE:
 		break;
 	case ASM_CLIENT_EVENT_CMD_EOS_DONE:
-		prtd->state = Q6ASM_STREAM_STOPPED;
 		break;
 	case ASM_CLIENT_EVENT_DATA_WRITE_DONE: {
 		snd_pcm_period_elapsed(substream);
@@ -349,7 +348,6 @@ static int q6asm_dai_trigger(struct snd_
 				       0, 0, 0);
 		break;
 	case SNDRV_PCM_TRIGGER_STOP:
-		prtd->state = Q6ASM_STREAM_STOPPED;
 		ret = q6asm_cmd_nowait(prtd->audio_client, prtd->stream_id,
 				       CMD_EOS);
 		break;
@@ -563,8 +561,6 @@ static void compress_event_handler(uint3
 			snd_compr_drain_notify(prtd->cstream);
 			prtd->notify_on_drain = false;
 
-		} else {
-			prtd->state = Q6ASM_STREAM_STOPPED;
 		}
 		break;
 
@@ -1022,7 +1018,6 @@ static int q6asm_dai_compr_trigger(struc
 				       0, 0, 0);
 		break;
 	case SNDRV_PCM_TRIGGER_STOP:
-		prtd->state = Q6ASM_STREAM_STOPPED;
 		ret = q6asm_cmd_nowait(prtd->audio_client, prtd->stream_id,
 				       CMD_EOS);
 		break;
author maintainer
#246 lore raw

PATCH 233/332 xfrm: esp: restore combined single-frag length gate

In reply to Greg Kroah-Hartman

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

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

From: Jingguo Tan <[email protected]>

commit dfa0d7b0ff1eb6b2c416b8fdb9b4f2cefba57a40 upstream.

The ESP out-of-place fast path appends the trailer in esp_output_head() before esp_output_tail() allocates the destination page frag. The head-side gate currently checks skb->data_len and tailen separately, but the tail code allocates a single destination frag from the combined post-trailer skb->data_len.

Reject the page-frag fast path when the combined aligned length exceeds a page. Otherwise skb_page_frag_refill() may fall back to a single page while the destination sg still spans the combined skb->data_len.

Restore this combined-length page gate for both IPv4 and IPv6.

Fixes 5bd8baab087d ("esp: limit skb_page_frag_refill use to a single page")
---
 net/ipv4/esp4.c |    4 ++--
 net/ipv6/esp6.c |    4 ++--
 2 files changed, 4 insertions(+), 4 deletions(-)
net/ipv4/esp4.c +2 -2
@@ -419,8 +419,8 @@ int esp_output_head(struct xfrm_state *x
 			return err;
 	}
 
-	if (ALIGN(tailen, L1_CACHE_BYTES) > PAGE_SIZE ||
-	    ALIGN(skb->data_len, L1_CACHE_BYTES) > PAGE_SIZE)
+	if (ALIGN(skb->data_len + tailen, L1_CACHE_BYTES) >
+	    PAGE_SIZE)
 		goto cow;
 
 	if (!skb_cloned(skb)) {
net/ipv6/esp6.c +2 -2
@@ -448,8 +448,8 @@ int esp6_output_head(struct xfrm_state *
 			return err;
 	}
 
-	if (ALIGN(tailen, L1_CACHE_BYTES) > PAGE_SIZE ||
-	    ALIGN(skb->data_len, L1_CACHE_BYTES) > PAGE_SIZE)
+	if (ALIGN(skb->data_len + tailen, L1_CACHE_BYTES) >
+	    PAGE_SIZE)
 		goto cow;
 
 	if (!skb_cloned(skb)) {
author maintainer
#247 lore raw

PATCH 234/332 ALSA: hda/realtek: Fix speaker output on ASUS ROG Strix G615LP

In reply to Greg Kroah-Hartman

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

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

From: Zhang Heng <[email protected]>

commit 20587302f8d700f26ee2c8a60ffb0a69ae0edf16 upstream.

Add quirk for ALC294 codec on ASUS ROG Strix G615LP (SSID 1043:1214) using ALC287_FIXUP_TXNW2781_I2C_ASUS to fix speaker output.

---
 sound/hda/codecs/realtek/alc269.c |    1 +
 1 file changed, 1 insertion(+)
sound/hda/codecs/realtek/alc269.c +1 -0
@@ -7260,6 +7260,7 @@ static const struct hda_quirk alc269_fix
 	SND_PCI_QUIRK(0x1043, 0x11c0, "ASUS X556UR", ALC255_FIXUP_ASUS_MIC_NO_PRESENCE),
 	HDA_CODEC_QUIRK(0x1043, 0x1204, "ASUS Strix G16 G615JMR", ALC287_FIXUP_TXNW2781_I2C_ASUS),
 	SND_PCI_QUIRK(0x1043, 0x1204, "ASUS Strix G615JHR_JMR_JPR", ALC287_FIXUP_TAS2781_I2C),
+	HDA_CODEC_QUIRK(0x1043, 0x1214, "ASUS ROG Strix G615LP", ALC287_FIXUP_TXNW2781_I2C_ASUS),
 	SND_PCI_QUIRK(0x1043, 0x1214, "ASUS Strix G615LH_LM_LP", ALC287_FIXUP_TAS2781_I2C),
 	SND_PCI_QUIRK(0x1043, 0x125e, "ASUS Q524UQK", ALC255_FIXUP_ASUS_MIC_NO_PRESENCE),
 	SND_PCI_QUIRK(0x1043, 0x1271, "ASUS X430UN", ALC256_FIXUP_ASUS_MIC_NO_PRESENCE),
author maintainer
#248 lore raw

PATCH 235/332 xfrm: iptfs: reset runtime state when cloning SAs

In reply to Greg Kroah-Hartman

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

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

From: Shaomin Chen <[email protected]>

commit 7f83d174073234839aea176f265e517e0d50a1d2 upstream.

iptfs_clone_state() clones the IPTFS mode data with kmemdup(). This copies runtime objects which must not be shared with the original SA, including the embedded sk_buff_head, hrtimers, spinlock, and in-flight reassembly/reorder state.

If xfrm_state_migrate() fails after clone_state() but before the later init_state() call has reinitialized those fields, the cloned state can be destroyed by xfrm_state_gc_task() with list and timer state copied from the original SA. With queued packets this lets the clone splice and free skbs owned by the original IPTFS queue, leading to use-after-free and double-free reports in iptfs_destroy_state() and skb release paths.

Reinitialize the clone's runtime state before publishing it through x->mode_data. Because clone_state() now publishes a destroyable mode_data object before init_state(), take the mode callback module reference there. Avoid taking it again from __iptfs_init_state() for the same object.

Fixes 0e4fbf013fa5 ("xfrm: iptfs: add user packet (tunnel ingress) handling")
---
 net/xfrm/xfrm_iptfs.c |   28 +++++++++++++++++++++++-----
 1 file changed, 23 insertions(+), 5 deletions(-)
net/xfrm/xfrm_iptfs.c +23 -5
@@ -2650,7 +2650,8 @@ static void __iptfs_init_state(struct xf
 	x->props.enc_hdr_len = sizeof(struct ip_iptfs_hdr);
 
 	/* Always keep a module reference when x->mode_data is set */
-	__module_get(x->mode_cbs->owner);
+	if (x->mode_data != xtfs)
+		__module_get(x->mode_cbs->owner);
 
 	x->mode_data = xtfs;
 	xtfs->x = x;
@@ -2658,22 +2659,39 @@ static void __iptfs_init_state(struct xf
 
 static int iptfs_clone_state(struct xfrm_state *x, struct xfrm_state *orig)
 {
+	struct skb_wseq *w_saved = NULL;
 	struct xfrm_iptfs_data *xtfs;
 
 	xtfs = kmemdup(orig->mode_data, sizeof(*xtfs), GFP_KERNEL);
 	if (!xtfs)
 		return -ENOMEM;
 
-	xtfs->ra_newskb = NULL;
 	if (xtfs->cfg.reorder_win_size) {
-		xtfs->w_saved = kzalloc_objs(*xtfs->w_saved,
-					     xtfs->cfg.reorder_win_size);
-		if (!xtfs->w_saved) {
+		w_saved = kzalloc_objs(*w_saved, xtfs->cfg.reorder_win_size);
+		if (!w_saved) {
 			kfree_sensitive(xtfs);
 			return -ENOMEM;
 		}
 	}
+	xtfs->w_saved = w_saved;
 
+	__skb_queue_head_init(&xtfs->queue);
+	xtfs->queue_size = 0;
+	hrtimer_setup(&xtfs->iptfs_timer, iptfs_delay_timer, CLOCK_MONOTONIC,
+		      IPTFS_HRTIMER_MODE);
+
+	spin_lock_init(&xtfs->drop_lock);
+	hrtimer_setup(&xtfs->drop_timer, iptfs_drop_timer, CLOCK_MONOTONIC,
+		      IPTFS_HRTIMER_MODE);
+
+	xtfs->w_seq_set = false;
+	xtfs->w_wantseq = 0;
+	xtfs->w_savedlen = 0;
+	xtfs->ra_newskb = NULL;
+	xtfs->ra_wantseq = 0;
+	xtfs->ra_runtlen = 0;
+
+	__module_get(x->mode_cbs->owner);
 	x->mode_data = xtfs;
 	xtfs->x = x;
 
author maintainer
#249 lore raw

PATCH 236/332 dma-buf: fix UAF in dma_buf_fd() tracepoint

In reply to Greg Kroah-Hartman

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

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

From: David Carlier <[email protected]>

commit ead6680f354f83966c796fc7f9463a3171789616 upstream.

Once FD_ADD() returns, the fd is live in the file descriptor table and a thread sharing that table can close() it before DMA_BUF_TRACE() runs. The close drops the last reference, __fput() frees the dma_buf, and the tracepoint then dereferences dmabuf to take dmabuf->name_lock -- slab-use-after-free.

Split FD_ADD() back into get_unused_fd_flags() + fd_install() and emit the tracepoint between them. While the fdtable slot is reserved with a NULL file pointer, a racing close() returns -EBADF without entering __fput(), so the dma_buf stays alive across the trace. Same approach as commit 2d76319c4cbb ("dma-buf: fix UAF in dma_buf_put() tracepoint").

This undoes the FD_ADD() conversion done in commit 34dfce523c90 ("dma: convert dma_buf_fd() to FD_ADD()"); FD_ADD() has no place to hook the tracepoint safely.

---
 drivers/dma-buf/dma-buf.c |    6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)
drivers/dma-buf/dma-buf.c +5 -1
@@ -792,9 +792,13 @@ int dma_buf_fd(struct dma_buf *dmabuf, i
 	if (!dmabuf || !dmabuf->file)
 		return -EINVAL;
 
-	fd = FD_ADD(flags, dmabuf->file);
+	fd = get_unused_fd_flags(flags);
+	if (fd < 0)
+		return fd;
+
 	DMA_BUF_TRACE(trace_dma_buf_fd, dmabuf, fd);
 
+	fd_install(fd, dmabuf->file);
 	return fd;
 }
 EXPORT_SYMBOL_NS_GPL(dma_buf_fd, "DMA_BUF");
author maintainer
#250 lore raw

PATCH 237/332 Input: xpad - add "Nova 2 Lite" from GameSir

In reply to Greg Kroah-Hartman

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

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

From: Qbeliw Tanaka <[email protected]>

commit 1f6ac0f8441c48c4cc250141e1da8486c13512ba upstream.

Add support for the gamepad "Nova 2 Lite" from GameSir, compatible with the Xbox 360 gamepad.

---
 drivers/input/joystick/xpad.c |    1 +
 1 file changed, 1 insertion(+)
drivers/input/joystick/xpad.c +1 -0
@@ -425,6 +425,7 @@ static const struct xpad_device {
 	{ 0x3285, 0x0662, "Nacon Revolution5 Pro", 0, XTYPE_XBOX360 },
 	{ 0x3285, 0x0663, "Nacon Evol-X", 0, XTYPE_XBOXONE },
 	{ 0x3537, 0x1004, "GameSir T4 Kaleid", 0, XTYPE_XBOX360 },
+	{ 0x3537, 0x100f, "GameSir Nova 2 Lite", 0, XTYPE_XBOX360 },
 	{ 0x3537, 0x1010, "GameSir G7 SE", 0, XTYPE_XBOXONE },
 	{ 0x3651, 0x1000, "CRKD SG", 0, XTYPE_XBOX360 },
 	{ 0x366c, 0x0005, "ByoWave Proteus Controller", MAP_SHARE_BUTTON, XTYPE_XBOXONE, FLAG_DELAY_INIT },
author maintainer
#251 lore raw

PATCH 238/332 Input: xpad - add support for ASUS ROG RAIKIRI II

In reply to Greg Kroah-Hartman

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

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

From: Dmitriy Zharov <[email protected]>

commit c897cf120696b94f56ed0f3197ba9a77071a59ec upstream.

Add the VID/PIDs for the ASUS ROG RAIKIRI II controller to xpad_device and the VID to xpad_table. The controller has a physical PC/XBOX toggle which switches between XBOX360 and XBOXONE protocols.

---
 drivers/input/joystick/xpad.c |    5 +++++
 1 file changed, 5 insertions(+)
drivers/input/joystick/xpad.c +5 -0
@@ -220,6 +220,10 @@ static const struct xpad_device {
 	{ 0x07ff, 0xffff, "Mad Catz GamePad", 0, XTYPE_XBOX360 },
 	{ 0x0b05, 0x1a38, "ASUS ROG RAIKIRI", MAP_SHARE_BUTTON, XTYPE_XBOXONE },
 	{ 0x0b05, 0x1abb, "ASUS ROG RAIKIRI PRO", 0, XTYPE_XBOXONE },
+	{ 0x0b05, 0x1c91, "ASUS ROG RAIKIRI II", 0, XTYPE_XBOX360 },
+	{ 0x0b05, 0x1c92, "ASUS ROG RAIKIRI II WIRELESS", 0, XTYPE_XBOX360 },
+	{ 0x0b05, 0x1c96, "ASUS ROG RAIKIRI II XBOX", MAP_SHARE_BUTTON, XTYPE_XBOXONE },
+	{ 0x0b05, 0x1d04, "ASUS ROG RAIKIRI II XBOX WIRELESS", MAP_SHARE_BUTTON, XTYPE_XBOXONE },
 	{ 0x0c12, 0x0005, "Intec wireless", 0, XTYPE_XBOX },
 	{ 0x0c12, 0x8801, "Nyko Xbox Controller", 0, XTYPE_XBOX },
 	{ 0x0c12, 0x8802, "Zeroplus Xbox Controller", 0, XTYPE_XBOX },
@@ -542,6 +546,7 @@ static const struct usb_device_id xpad_t
 	{ USB_DEVICE(0x0738, 0x4540) },		/* Mad Catz Beat Pad */
 	XPAD_XBOXONE_VENDOR(0x0738),		/* Mad Catz FightStick TE 2 */
 	XPAD_XBOX360_VENDOR(0x07ff),		/* Mad Catz Gamepad */
+	XPAD_XBOX360_VENDOR(0x0b05),		/* ASUS controllers */
 	XPAD_XBOXONE_VENDOR(0x0b05),		/* ASUS controllers */
 	XPAD_XBOX360_VENDOR(0x0c12),		/* Zeroplus X-Box 360 controllers */
 	XPAD_XBOX360_VENDOR(0x0db0),		/* Micro Star International X-Box 360 controllers */
author maintainer
#252 lore raw

PATCH 239/332 ksmbd: OOB read regression in smb_check_perm_dacl() ACE-walk loops

In reply to Greg Kroah-Hartman

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

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

From: Ali Ganiyev <[email protected]>

commit 0e60dafe97eca61721f3db456f97d97a80c6c8ae upstream.

Commit d07b26f39246 ("ksmbd: require minimum ACE size in smb_check_perm_dacl()") introduced a transposed bounds check:

if (offsetof(struct smb_ace, sid) + aces_size < CIFS_SID_BASE_SIZE)

Since offsetof(..sid) is 8 and CIFS_SID_BASE_SIZE is 8, this evaluates to aces_size < 0. Because aces_size is always non-negative, this check becomes dead code and never breaks the loop.

Worse, that commit removed the old 4-byte guard, meaning the loop now reads ace->size (offset 2) even when aces_size is 0-3 bytes. This re-opens a 2-byte heap out-of-bounds (OOB) read past the pntsd allocation during subsequent SMB2_CREATE operations.

Fix this by properly transposing the comparison to require at least 16 bytes (8-byte offset + 8-byte SID base), matching the correct form used in smb_inherit_dacl().

Fixes d07b26f39246 ("ksmbd: require minimum ACE size in smb_check_perm_dacl()")
---
 fs/smb/server/smbacl.c |    8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)
fs/smb/server/smbacl.c +4 -4
@@ -1446,8 +1446,8 @@ int smb_check_perm_dacl(struct ksmbd_con
 		ace = (struct smb_ace *)((char *)pdacl + sizeof(struct smb_acl));
 		aces_size = acl_size - sizeof(struct smb_acl);
 		for (i = 0; i < le16_to_cpu(pdacl->num_aces); i++) {
-			if (offsetof(struct smb_ace, sid) +
-			    aces_size < CIFS_SID_BASE_SIZE)
+			if (aces_size < offsetof(struct smb_ace, sid) +
+			    CIFS_SID_BASE_SIZE)
 				break;
 			ace_size = le16_to_cpu(ace->size);
 			if (ace_size > aces_size ||
@@ -1467,8 +1467,8 @@ int smb_check_perm_dacl(struct ksmbd_con
 	ace = (struct smb_ace *)((char *)pdacl + sizeof(struct smb_acl));
 	aces_size = acl_size - sizeof(struct smb_acl);
 	for (i = 0; i < le16_to_cpu(pdacl->num_aces); i++) {
-		if (offsetof(struct smb_ace, sid) +
-		    aces_size < CIFS_SID_BASE_SIZE)
+		if (aces_size < offsetof(struct smb_ace, sid) +
+		    CIFS_SID_BASE_SIZE)
 			break;
 		ace_size = le16_to_cpu(ace->size);
 		if (ace_size > aces_size ||
author maintainer
#253 lore raw

PATCH 240/332 misc: rp1: Send IACK on IRQ activate to fix kdump/kexec

In reply to Greg Kroah-Hartman

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

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

From: Xiaolei Wang <[email protected]>

commit 36770417153644bc88281c7284730ef1d14d8d3c upstream.

After a kexec/kdump reboot, the macb Ethernet controller fails to receive any packets, causing DHCP to hang indefinitely and the network interface to be unusable despite link being up.

The root cause is that RP1's level-triggered MSI-X interrupt sources (such as macb on hwirq 6) may have their internal state machines stuck in the "waiting for IACK" state. This happens because the previous kernel crashed before sending the acknowledgment for a pending level interrupt.

In this stuck state, RP1 will not generate new MSI-X writes even though the interrupt source remains asserted. Since no new MSI-X is sent, the GIC never sees a new edge, the chained IRQ handler is never invoked, and the interrupt is permanently lost.

Fix this by sending MSIX_CFG_IACK in rp1_irq_activate(). This unconditionally resets the MSI-X state machine back to idle when a child device requests its interrupt. If the interrupt source is still asserted, RP1 will immediately issue a new MSI-X with the freshly configured msg_addr/msg_data, and normal interrupt delivery resumes.

Writing IACK when the state machine is already idle (i.e., on a normal cold boot) is harmless — it has no effect.

Fixes 49d63971f963 ("misc: rp1: RaspberryPi RP1 misc driver")
Cc stable <[email protected]>
---
 drivers/misc/rp1/rp1_pci.c |    1 +
 1 file changed, 1 insertion(+)
drivers/misc/rp1/rp1_pci.c +1 -0
@@ -143,6 +143,7 @@ static int rp1_irq_activate(struct irq_d
 	struct rp1_dev *rp1 = d->host_data;
 
 	msix_cfg_set(rp1, (unsigned int)irqd->hwirq, MSIX_CFG_ENABLE);
+	msix_cfg_set(rp1, (unsigned int)irqd->hwirq, MSIX_CFG_IACK);
 
 	return 0;
 }
author maintainer
#254 lore raw

PATCH 241/332 Input: atmel_mxt_ts - fix boundary check in mxt_prepare_cfg_mem

In reply to Greg Kroah-Hartman

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

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

From: Dmitry Torokhov <[email protected]>

commit baa0210fb6a9dc3882509a9411b6d284d88fe30e upstream.

When a configuration file provides an object size that is larger than the driver's known mxt_obj_size(object), the driver intends to discard the extra bytes.

The loop iterates using for (i = 0; i < size; i++). Inside the loop, the condition to skip processing extra bytes is:

if (i > mxt_obj_size(object))
continue;

Since i is a 0-based index, the valid indices for the object are 0 through mxt_obj_size(object) - 1.

When i == mxt_obj_size(object), the condition evaluates to false, and the code processes the byte instead of discarding it.

This causes the code to calculate byte_offset = reg + i - cfg->start_ofs and writes the byte there, overwriting exactly one byte of the adjacent instance or object.

Update the boundary check to skip extra bytes correctly by using >=.

Fixes 50a77c658b80 ("Input: atmel_mxt_ts - download device config using firmware loader")
Assisted-by Gemini:gemini-3.1-pro
---
 drivers/input/touchscreen/atmel_mxt_ts.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/input/touchscreen/atmel_mxt_ts.c +1 -1
@@ -1477,7 +1477,7 @@ static int mxt_prepare_cfg_mem(struct mx
 			}
 			cfg->raw_pos += offset;
 
-			if (i > mxt_obj_size(object))
+			if (i >= mxt_obj_size(object))
 				continue;
 
 			byte_offset = reg + i - cfg->start_ofs;
author maintainer
#255 lore raw

PATCH 242/332 Input: synaptics - add LEN2058 to SMBus passlist for ThinkPad E490

In reply to Greg Kroah-Hartman

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

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

From: Nicolás Bazaes <[email protected]>

commit 16ca52bc209fa4bf9239cd9e5643e95533476b58 upstream.

The Lenovo ThinkPad E490 (PNP ID: LEN2058) has a Synaptics TM3471-020 touchpad that supports SMBus/RMI4 mode but is not listed in smbus_pnp_ids[]. Without this entry, RMI4 over SMBus is not enabled by default, and the touchpad falls back to PS/2 mode.

Adding LEN2058 to the passlist enables automatic RMI4 detection without requiring the psmouse.synaptics_intertouch parameter, and matches the behavior of similar ThinkPad models already in the list (E480/LEN2054, E580/LEN2055).

Tested on ThinkPad E490 with kernel 7.0.5-zen1 and Arch Linux. RMI4 over SMBus is confirmed working without any kernel parameters.

---
 drivers/input/mouse/synaptics.c |    1 +
 1 file changed, 1 insertion(+)
drivers/input/mouse/synaptics.c +1 -0
@@ -190,6 +190,7 @@ static const char * const smbus_pnp_ids[
 	"LEN2044", /* L470  */
 	"LEN2054", /* E480 */
 	"LEN2055", /* E580 */
+	"LEN2058", /* E490 */
 	"LEN2068", /* T14 Gen 1 */
 	"SYN1221", /* TUXEDO InfinityBook Pro 14 v5 */
 	"SYN3003", /* HP EliteBook 850 G1 */
author maintainer
#256 lore raw

PATCH 243/332 gpib: cb7210: Fix region leak when request_irq fails

In reply to Greg Kroah-Hartman

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

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

From: Hongling Zeng <[email protected]>

commit 2eae90a457baa0048a96ed38ad93090ee38c8b2f upstream.

When request_irq() fails, the region allocated by request_region() is not released. Fix this by adding an error handling path with proper goto labels to release the region.

---
 drivers/gpib/cb7210/cb7210.c |   10 ++++++++--
 1 file changed, 8 insertions(+), 2 deletions(-)
drivers/gpib/cb7210/cb7210.c +8 -2
@@ -1049,7 +1049,8 @@ static int cb_isa_attach(struct gpib_boa
 	if (!request_region(config->ibbase, cb7210_iosize, DRV_NAME)) {
 		dev_err(board->gpib_dev, "ioports starting at 0x%x are already in use\n",
 			config->ibbase);
-		return -EBUSY;
+		retval = -EBUSY;
+		goto err_release_region;
 	}
 	nec_priv->iobase = config->ibbase;
 	cb_priv->fifo_iobase = nec7210_iobase(cb_priv);
@@ -1062,11 +1063,16 @@ static int cb_isa_attach(struct gpib_boa
 	// install interrupt handler
 	if (request_irq(config->ibirq, cb7210_interrupt, isr_flags, DRV_NAME, board)) {
 		dev_err(board->gpib_dev, "failed to obtain IRQ %d\n", config->ibirq);
-		return -EBUSY;
+		retval = -EBUSY;
+		goto err_release_region;
 	}
 	cb_priv->irq = config->ibirq;
 
 	return cb7210_init(cb_priv, board);
+
+err_release_region:
+	release_region(nec7210_iobase(cb_priv), cb7210_iosize);
+	return retval;
 }
 
 static void cb_isa_detach(struct gpib_board *board)
#257 lore raw

Re: [PATCH 7.0 243/332] gpib: cb7210: Fix region leak when request_irq fails

Quoted 7 lines · Greg Kroah-Hartman

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

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

From: Hongling Zeng <[email protected]>

commit 2eae90a457baa0048a96ed38ad93090ee38c8b2f upstream.

This was reverted by 05d5d79440c2cc0784f91b61f2012753e66be472.

thanks,

Signature · 2 lines
js
suse labs
maintainer
#258 lore raw

Re: [PATCH 7.0 243/332] gpib: cb7210: Fix region leak when request_irq fails

Quoted 1 line

This was reverted by 05d5d79440c2cc0784f91b61f2012753e66be472.

Dropped from the 7.0 and 6.18 queues, thanks.

--
Thanks,
Sasha

author maintainer
#259 lore raw

PATCH 244/332 dt-bindings: usb: Fix EIC7700 USB resets issue

In reply to Greg Kroah-Hartman

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

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

From: Hang Cao <[email protected]>

commit f1ecb0e563595d4ba9a3b8e39ed52a3dc2d8e328 upstream.

The EIC7700 USB requires a USB PHY reset operation; otherwise, the USB will not work. The reason why the USB driver that was applied can work properly is that the USB PHY has already been reset in ESWIN's U-Boot.

However, the proper functioning of the USB driver should not be dependent on the bootloader. Therefore, it is necessary to incorporate the USB PHY reset signal into the DT bindings.

This patch does not introduce any backward incompatibility since the dts is not upstream yet. As array of reset operations are used in the driver, no modifications to the USB controller driver are needed.

Fixes c640a4239db5 ("dt-bindings: usb: Add ESWIN EIC7700 USB controller")
Cc stable <[email protected]>
---
 Documentation/devicetree/bindings/usb/eswin,eic7700-usb.yaml |    7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)
Documentation/devicetree/bindings/usb/eswin,eic7700-usb.yaml +4 -3
@@ -41,12 +41,13 @@ properties:
       - const: usb_en
 
   resets:
-    maxItems: 2
+    maxItems: 3
 
   reset-names:
     items:
       - const: vaux
       - const: usb_rst
+      - const: usb_phy
 
   eswin,hsp-sp-csr:
     description:
@@ -85,8 +86,8 @@ examples:
         interrupt-parent = <&plic>;
         interrupts = <85>;
         interrupt-names = "peripheral";
-        resets = <&reset 84>, <&hspcrg 2>;
-        reset-names = "vaux", "usb_rst";
+        resets = <&reset 84>, <&hspcrg 2>, <&hspcrg 4>;
+        reset-names = "vaux", "usb_rst", "usb_phy";
         dr_mode = "peripheral";
         maximum-speed = "high-speed";
         phy_type = "utmi";
author maintainer
#260 lore raw

PATCH 245/332 comedi: comedi_test: fix check for valid scan_begin_src in waveform_ai_cmdtest()

In reply to Greg Kroah-Hartman

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

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

From: Ian Abbott <[email protected]>

commit 542f5248cb481073203e0dadab5bcbd28aeae308 upstream.

Commit 783ddaebd397 ("staging: comedi: comedi_test: support scan_begin_src == TRIG_FOLLOW") neglected to add a test that scan_begin_src has only one bit set. The allowed values are TRIG_FOLLOW and TRIG_TIMER, but the code incorrectly also allows TRIG_FOLLOW | TRIG_TIMER. Add a call to comedi_check_trigger_is_unique() to check that only one trigger source bit is set.

Fixes 783ddaebd397 ("staging: comedi: comedi_test: support scan_begin_src == TRIG_FOLLOW")
Cc stable <[email protected]>
---
 drivers/comedi/drivers/comedi_test.c |    1 +
 1 file changed, 1 insertion(+)
drivers/comedi/drivers/comedi_test.c +1 -0
@@ -274,6 +274,7 @@ static int waveform_ai_cmdtest(struct co
 	/* Step 2a : make sure trigger sources are unique */
 
 	err |= comedi_check_trigger_is_unique(cmd->convert_src);
+	err |= comedi_check_trigger_is_unique(cmd->scan_begin_src);
 	err |= comedi_check_trigger_is_unique(cmd->stop_src);
 
 	/* Step 2b : and mutually compatible */
author maintainer
#261 lore raw

PATCH 246/332 comedi: comedi_test: Fix limiting of convert_arg in waveform_ai_cmdtest()

In reply to Greg Kroah-Hartman

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

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

From: Ian Abbott <[email protected]>

commit 8a3bee801d420be8a7a0bae4a26547b353b8fe22 upstream.

The function checks and possibly modifies the description of an asynchronous command to be run on the analog input subdevice of a comedi device attached to the "comedi_test" driver, returning 0 if no modifications were required, or a positive value that indicates which step of the checking process it failed on. Step 4 fixes up various argument values for various trigger sources.

There are two bugs in the fixing up of the convert_arg value to keep the scan_begin_arg value within the range of unsigned int when scan_begin_src and convert_src both have the value TRIG_TIMER, which indicates that the corresponding _arg values hold a time period in nanoseconds. The code also uses scan_end_arg which hold the number of "conversions" within each "scan". The goal is to end up with the scan period being less than or equal to the convert period multiplied by the number of conversions per scan. It intends to do that by clamping the convert_arg value to a maximum value of UINT_MAX / scan_end_arg rounded down to a multiple of 1000 (NSEC_PER_USEC).

(The rounding from nanoseconds to microseconds is because the driver is modelling a device that uses a 1 MHz clock for timing. This is partly because that is a more typical timing base for real hardware devices driven by comedi, and partly because the driver used to use struct timeval internally.)

The first bug is that the code checks if scan_begin_arg == TRIG_TIMER when it should be checking if scan_begin_src == TRIG_TIMER. The bugged check will always fail because if scan_begin_src == TRIG_TIMER, then scan_begin_arg will be at least 1000 (NSEC_PER_USEC), otherwise scan_begin_src == TRIG_FOLLOW and scan_begin_arg will be 0. (N.B TRIG_TIMER is defined as 0x10.) The second bug is that is rounding the maximum value down to a multiple of 1000000000 (NSEC_PER_SEC) instead of 1000 (NSEC_PER_USEC), however this bug is not reached due to the first bug. This patch fixes both bugs.

Fixes 783ddaebd397 ("staging: comedi: comedi_test: support scan_begin_src == TRIG_FOLLOW")
Fixes 5afdcad2f818 ("staging: comedi: comedi_test: limit maximum convert_arg")
Cc stable <[email protected]>
---
 drivers/comedi/drivers/comedi_test.c |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/comedi/drivers/comedi_test.c +2 -2
@@ -325,10 +325,10 @@ static int waveform_ai_cmdtest(struct co
 		arg = min(arg,
 			  rounddown(UINT_MAX, (unsigned int)NSEC_PER_USEC));
 		arg = NSEC_PER_USEC * DIV_ROUND_CLOSEST(arg, NSEC_PER_USEC);
-		if (cmd->scan_begin_arg == TRIG_TIMER) {
+		if (cmd->scan_begin_src == TRIG_TIMER) {
 			/* limit convert_arg to keep scan_begin_arg in range */
 			limit = UINT_MAX / cmd->scan_end_arg;
-			limit = rounddown(limit, (unsigned int)NSEC_PER_SEC);
+			limit = rounddown(limit, (unsigned int)NSEC_PER_USEC);
 			arg = min(arg, limit);
 		}
 		err |= comedi_check_trigger_arg_is(&cmd->convert_arg, arg);
author maintainer
#262 lore raw

PATCH 247/332 counter: Fix refcount leak in counter_alloc() error path

In reply to Greg Kroah-Hartman

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

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

From: Guangshuo Li <[email protected]>

commit d9eeb0ea0d2de658663bfaa9c26eccdd8fd64440 upstream.

After device_initialize(), the lifetime of the embedded struct device is expected to be managed through the device core reference counting.

In counter_alloc(), if dev_set_name() fails after device_initialize(), the error path removes the chrdev, frees the ID, and frees the backing allocation directly instead of releasing the device reference with put_device(). This bypasses the normal device lifetime rules and may leave the reference count of the embedded struct device unbalanced, resulting in a refcount leak.

The issue was identified by a static analysis tool I developed and confirmed by manual review.

Fix this by using put_device() in the dev_set_name() failure path and let counter_device_release() handle the final cleanup.

---
 drivers/counter/counter-core.c |    3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/counter/counter-core.c +2 -1
@@ -124,7 +124,8 @@ struct counter_device *counter_alloc(siz
 
 err_dev_set_name:
 
-	counter_chrdev_remove(counter);
+	put_device(dev);
+	return NULL;
 err_chrdev_add:
 
 	ida_free(&counter_ida, dev->id);
author maintainer
#263 lore raw

PATCH 248/332 tty: serial: pch_uart: add check for dma_alloc_coherent()

In reply to Greg Kroah-Hartman

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

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

From: Zhaoyang Yu <[email protected]>

commit 6fe472c1bbbe238e91141f7cabc1226e96a60d43 upstream.

Add a check for dma_alloc_coherent() failure to prevent a potential NULL pointer dereference in dma_handle_rx(). Properly release DMA channels and the PCI device reference using a goto ladder if the allocation fails.

---
 drivers/tty/serial/pch_uart.c |   19 +++++++++++++------
 1 file changed, 13 insertions(+), 6 deletions(-)
drivers/tty/serial/pch_uart.c +13 -6
@@ -689,8 +689,7 @@ static void pch_request_dma(struct uart_
 	if (!chan) {
 		dev_err(priv->port.dev, "%s:dma_request_channel FAILS(Tx)\n",
 			__func__);
-		pci_dev_put(dma_dev);
-		return;
+		goto err_pci_get;
 	}
 	priv->chan_tx = chan;
 
@@ -704,18 +703,26 @@ static void pch_request_dma(struct uart_
 	if (!chan) {
 		dev_err(priv->port.dev, "%s:dma_request_channel FAILS(Rx)\n",
 			__func__);
-		dma_release_channel(priv->chan_tx);
-		priv->chan_tx = NULL;
-		pci_dev_put(dma_dev);
-		return;
+		goto err_req_tx;
 	}
 
 	/* Get Consistent memory for DMA */
 	priv->rx_buf_virt = dma_alloc_coherent(port->dev, port->fifosize,
 				    &priv->rx_buf_dma, GFP_KERNEL);
+	if (!priv->rx_buf_virt)
+		goto err_req_rx;
 	priv->chan_rx = chan;
 
 	pci_dev_put(dma_dev);
+	return;
+
+err_req_rx:
+	dma_release_channel(chan);
+err_req_tx:
+	dma_release_channel(priv->chan_tx);
+	priv->chan_tx = NULL;
+err_pci_get:
+	pci_dev_put(dma_dev);
 }
 
 static void pch_dma_rx_complete(void *arg)
author maintainer
#264 lore raw

PATCH 249/332 tty: serial: samsung: Remove redundant port lock acquisition in rx helpers

In reply to Greg Kroah-Hartman

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

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

From: Tudor Ambarus <[email protected]>

commit a3bb136bff5e6a5e48cdd813246c9c4686feaaa9 upstream.

Sashiko identified a deadlock when the console flow is engaged [1].

When console flow control is enabled (UPF_CONS_FLOW), s3c24xx_serial_stop_tx() calls s3c24xx_serial_rx_enable() and s3c24xx_serial_start_tx() calls s3c24xx_serial_rx_disable().

The serial core framework invokes the .stop_tx() and .start_tx() callbacks with the port->lock spinlock already held. Furthermore, all internal driver paths that invoke stop_tx (such as the DMA TX completion handler s3c24xx_serial_tx_dma_complete() or the PIO TX IRQ handler s3c24xx_serial_tx_irq()) also acquire port->lock prior to calling it. (Note that s3c24xx_serial_start_tx() is only invoked by the serial core).

However, s3c24xx_serial_rx_enable() and s3c24xx_serial_rx_disable() unconditionally attempt to acquire port->lock again using uart_port_lock_irqsave(). Since spinlocks are not recursive, this causes a deadlock on the same CPU when console flow control is engaged.

Remove the redundant lock acquisition from both rx helper functions.

---
 drivers/tty/serial/samsung_tty.c |    8 --------
 1 file changed, 8 deletions(-)
drivers/tty/serial/samsung_tty.c +0 -8
@@ -245,12 +245,9 @@ static bool s3c24xx_serial_txempty_nofif
 static void s3c24xx_serial_rx_enable(struct uart_port *port)
 {
 	struct s3c24xx_uart_port *ourport = to_ourport(port);
-	unsigned long flags;
 	int count = 10000;
 	u32 ucon, ufcon;
 
-	uart_port_lock_irqsave(port, &flags);
-
 	while (--count && !s3c24xx_serial_txempty_nofifo(port))
 		udelay(100);
 
@@ -263,23 +260,18 @@ static void s3c24xx_serial_rx_enable(str
 	wr_regl(port, S3C2410_UCON, ucon);
 
 	ourport->rx_enabled = 1;
-	uart_port_unlock_irqrestore(port, flags);
 }
 
 static void s3c24xx_serial_rx_disable(struct uart_port *port)
 {
 	struct s3c24xx_uart_port *ourport = to_ourport(port);
-	unsigned long flags;
 	u32 ucon;
 
-	uart_port_lock_irqsave(port, &flags);
-
 	ucon = rd_regl(port, S3C2410_UCON);
 	ucon &= ~S3C2410_UCON_RXIRQMODE;
 	wr_regl(port, S3C2410_UCON, ucon);
 
 	ourport->rx_enabled = 0;
-	uart_port_unlock_irqrestore(port, flags);
 }
 
 static void s3c24xx_serial_stop_tx(struct uart_port *port)
author maintainer
#265 lore raw

PATCH 250/332 uio: uio_pci_generic_sva: fix double free of devm_kzalloc() memory

In reply to Greg Kroah-Hartman

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

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

From: Guangshuo Li <[email protected]>

commit f74c8696f14149d5e43cc28b015326a759c48f00 upstream.

uio_pci_sva allocates struct uio_pci_sva_dev with devm_kzalloc() in probe(), but then calls kfree(udev) both on the probe() error path (label out_free) and again in remove().

Because devm_kzalloc() allocations are devres-managed and are freed automatically when the device is detached (including after a failing probe() and during driver unbind), the explicit kfree() can lead to a double free.

If probe() fails after devm_kzalloc(), the error path frees udev and devres cleanup will free it again when the core unwinds the partially bound device. On normal driver removal, remove() frees udev and devres will free it again when the device is detached.

This issue was identified by a static analysis tool I developed and confirmed by manual review. Fix by removing the manual kfree() calls and dropping the now-unused label.

Fixes 3397c3cd859a2 ("uio: Add SVA support for PCI devices via uio_pci_generic_sva.c")
Cc stable <[email protected]>
---
 drivers/uio/uio_pci_generic_sva.c |    7 +------
 1 file changed, 1 insertion(+), 6 deletions(-)
drivers/uio/uio_pci_generic_sva.c +1 -6
@@ -129,15 +129,13 @@ static int probe(struct pci_dev *pdev, c
 	ret = devm_uio_register_device(&pdev->dev, &udev->info);
 	if (ret) {
 		dev_err(&pdev->dev, "Failed to register uio device\n");
-		goto out_free;
+		goto out_disable;
 	}
 
 	pci_set_drvdata(pdev, udev);
 
 	return 0;
 
-out_free:
-	kfree(udev);
 out_disable:
 	pci_disable_device(pdev);
 
@@ -146,11 +144,8 @@ out_disable:
 
 static void remove(struct pci_dev *pdev)
 {
-	struct uio_pci_sva_dev *udev = pci_get_drvdata(pdev);
-
 	pci_release_regions(pdev);
 	pci_disable_device(pdev);
-	kfree(udev);
 }
 
 static ssize_t pasid_show(struct device *dev,
author maintainer
#266 lore raw

PATCH 251/332 usb: chipidea: core: convert ci_role_switch to local variable

In reply to Greg Kroah-Hartman

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

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

From: Xu Yang <[email protected]>

commit 8f6aa392653e52a45858cff5c063df550028836b upstream.

When a system contains multiple USB controllers, the global ci_role_switch variable may be overwritten by subsequent driver initialization code.

This can cause issues in the following cases:
- The 2nd ci_hdrc_probe() sees ci_role_switch.fwnode as non-NULL even though the "usb-role-switch" property is not present for the controller.
- When the ci_hdrc device is unbound and bound again, ci_role_switch fwnode will not be reassigned, and the old value will be used instead.

Convert ci_role_switch to a local variable to fix these issues.

Fixes 05559f10ed79 ("usb: chipidea: add role switch class support")
Cc stable <[email protected]>
---
 drivers/usb/chipidea/core.c |   16 ++++++----------
 1 file changed, 6 insertions(+), 10 deletions(-)
drivers/usb/chipidea/core.c +6 -10
@@ -670,12 +670,6 @@ static enum ci_role ci_get_role(struct c
 	return role;
 }
 
-static struct usb_role_switch_desc ci_role_switch = {
-	.set = ci_usb_role_switch_set,
-	.get = ci_usb_role_switch_get,
-	.allow_userspace_control = true,
-};
-
 static int ci_get_platdata(struct device *dev,
 		struct ci_hdrc_platform_data *platdata)
 {
@@ -802,9 +796,6 @@ static int ci_get_platdata(struct device
 			cable->connected = false;
 	}
 
-	if (device_property_read_bool(dev, "usb-role-switch"))
-		ci_role_switch.fwnode = dev->fwnode;
-
 	platdata->pctl = devm_pinctrl_get(dev);
 	if (!IS_ERR(platdata->pctl)) {
 		struct pinctrl_state *p;
@@ -1048,6 +1039,7 @@ ATTRIBUTE_GROUPS(ci);
 
 static int ci_hdrc_probe(struct platform_device *pdev)
 {
+	struct usb_role_switch_desc ci_role_switch = {};
 	struct device	*dev = &pdev->dev;
 	struct ci_hdrc	*ci;
 	struct resource	*res;
@@ -1194,7 +1186,11 @@ static int ci_hdrc_probe(struct platform
 		}
 	}
 
-	if (ci_role_switch.fwnode) {
+	if (device_property_read_bool(dev, "usb-role-switch")) {
+		ci_role_switch.set = ci_usb_role_switch_set;
+		ci_role_switch.get = ci_usb_role_switch_get;
+		ci_role_switch.allow_userspace_control = true;
+		ci_role_switch.fwnode = dev_fwnode(dev);
 		ci_role_switch.driver_data = ci;
 		ci->role_switch = usb_role_switch_register(dev,
 					&ci_role_switch);
author maintainer
#267 lore raw

PATCH 252/332 usb: core: Fix up Interrupt IN endpoints with bogus wBytesPerInterval

In reply to Greg Kroah-Hartman

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

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

From: Michal Pecio <[email protected]>

commit 727d045d064b7c9a24db3bce9c0485a382cb768b upstream.

Tao Xue found that some common devices violate USB 3.x section 9.6.7 by reporting wBytesPerInterval lower than the size of packets they actually send. I confirmed that AX88179 may set it to 0 and RTL8153 CDC configuration sets it to 8 but sends both 8 and 16 byte packets:

S Ii:11:007:3 -115:128 16 <
C Ii:11:007:3 0:128 8 = a1000000 01000000
S Ii:11:007:3 -115:128 16 <
C Ii:11:007:3 0:128 16 = a12a0000 01000800 00000000 00000000

Most xHCI host controllers neglect interrupt bandwidth reservations and let such devices exceed theirs, some fail the URB with EOVERFLOW.

Assume that wBytesPerInterval lower than wMaxPacketSize is bogus and increase it to the worst case maximum on interrupt IN endpoints. This solves xHCI problems and appears to have no other effect. Interrupt transfers are not limited to one interval and drivers submit URBs of class defined size without looking at wBytesPerInterval. Any multi- interval transfer is considered terminated by a packet shorter than wMaxPacketSize regardless of wBytesPerInterval - see USB3 8.10.3.

Stay in spec on OUT endpoints and isochronous. No buggy devices are known and we don't want to risk sending more data than the device is prepared to handle or confusing isoc drivers regarding altsetting capacities guaranteed by the device itself. And don't complain when wMaxPacketSize <= wBytesPerInterval < wMaxPacketSize * (bMaxBurst+1) because enabling this seems to be the exact goal of the spec.

---
 drivers/usb/core/config.c |    9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)
drivers/usb/core/config.c +8 -1
@@ -191,7 +191,14 @@ static void usb_parse_ss_endpoint_compan
 			(desc->bMaxBurst + 1);
 	else
 		max_tx = 999999;
-	if (le16_to_cpu(desc->wBytesPerInterval) > max_tx) {
+	/*
+	 * wBytesPerInterval > max_tx is bogus, but USB3 spec doesn't forbid the opposite.
+	 * Experience shows that wBytesPerInterval < wMaxPacketSize on common interrupt IN
+	 * endpoints is usually bogus too, and recent HCs enforce interrupt BW limits.
+	 */
+	if (le16_to_cpu(desc->wBytesPerInterval) > max_tx ||
+	    (le16_to_cpu(desc->wBytesPerInterval) < usb_endpoint_maxp(&ep->desc) &&
+	     usb_endpoint_is_int_in(&ep->desc))) {
 		dev_notice(ddev, "%s endpoint with wBytesPerInterval of %d in "
 				"config %d interface %d altsetting %d ep %d: "
 				"setting to %d\n",
author maintainer
#268 lore raw

PATCH 253/332 usb: dwc3: xilinx: fix error handling in zynqmp init error paths

In reply to Greg Kroah-Hartman

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

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

From: Radhey Shyam Pandey <[email protected]>

commit c1a0ecbf32c4b397353204e2ec94c5bb9f3300ed upstream.

Fix error handling and resource cleanup i.e remove invalid phy_exit() after failed phy_init(), route failures through proper cleanup paths and return 0 explicitly on success.

---
 drivers/usb/dwc3/dwc3-xilinx.c |   27 +++++++++++++++------------
 1 file changed, 15 insertions(+), 12 deletions(-)
drivers/usb/dwc3/dwc3-xilinx.c +15 -12
@@ -184,15 +184,13 @@ static int dwc3_xlnx_init_zynqmp(struct
 	}
 
 	ret = phy_init(priv_data->usb3_phy);
-	if (ret < 0) {
-		phy_exit(priv_data->usb3_phy);
+	if (ret < 0)
 		goto err;
-	}
 
 	ret = reset_control_deassert(apbrst);
 	if (ret < 0) {
 		dev_err(dev, "Failed to release APB reset\n");
-		goto err;
+		goto err_phy_exit;
 	}
 
 	if (priv_data->usb3_phy) {
@@ -208,26 +206,24 @@ static int dwc3_xlnx_init_zynqmp(struct
 	ret = reset_control_deassert(crst);
 	if (ret < 0) {
 		dev_err(dev, "Failed to release core reset\n");
-		goto err;
+		goto err_phy_exit;
 	}
 
 	ret = reset_control_deassert(hibrst);
 	if (ret < 0) {
 		dev_err(dev, "Failed to release hibernation reset\n");
-		goto err;
+		goto err_phy_exit;
 	}
 
 	ret = phy_power_on(priv_data->usb3_phy);
-	if (ret < 0) {
-		phy_exit(priv_data->usb3_phy);
-		goto err;
-	}
+	if (ret < 0)
+		goto err_phy_exit;
 
 	/* ulpi reset via gpio-modepin or gpio-framework driver */
 	reset_gpio = devm_gpiod_get_optional(dev, "reset", GPIOD_OUT_HIGH);
 	if (IS_ERR(reset_gpio)) {
-		return dev_err_probe(dev, PTR_ERR(reset_gpio),
-				     "Failed to request reset GPIO\n");
+		ret = PTR_ERR(reset_gpio);
+		goto err_phy_power_off;
 	}
 
 	if (reset_gpio) {
@@ -237,6 +233,13 @@ static int dwc3_xlnx_init_zynqmp(struct
 	}
 
 	dwc3_xlnx_set_coherency(priv_data, XLNX_USB_TRAFFIC_ROUTE_CONFIG);
+
+	return 0;
+
+err_phy_power_off:
+	phy_power_off(priv_data->usb3_phy);
+err_phy_exit:
+	phy_exit(priv_data->usb3_phy);
 err:
 	return ret;
 }
author maintainer
#269 lore raw

PATCH 254/332 usb: musb: omap2430: Fix use-after-free in omap2430_probe()

In reply to Greg Kroah-Hartman

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

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

From: Wentao Liang <[email protected]>

commit e194ce048f5a6c549b3a23a8c568c6470f40f772 upstream.

In omap2430_probe(), of_node_put(np) is called prematurely before the last access to np, leading to a use-after-free if the node's reference count drops to zero. Move the of_node_put() calls after the last use of np in both the success and error paths.

Fixes ffbe2feac59b ("usb: musb: omap2430: Fix probe regression for missing resources")
Cc stable <[email protected]>
---
 drivers/usb/musb/omap2430.c |    3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/usb/musb/omap2430.c +2 -1
@@ -337,7 +337,6 @@ static int omap2430_probe(struct platfor
 	} else {
 		device_set_of_node_from_dev(&musb->dev, &pdev->dev);
 	}
-	of_node_put(np);
 
 	glue->dev			= &pdev->dev;
 	glue->musb			= musb;
@@ -455,6 +454,7 @@ static int omap2430_probe(struct platfor
 		dev_err(&pdev->dev, "failed to register musb device\n");
 		goto err_disable_rpm;
 	}
+	of_node_put(np);
 
 	return 0;
 
@@ -464,6 +464,7 @@ err_put_control_otghs:
 	if (!IS_ERR(glue->control_otghs))
 		put_device(glue->control_otghs);
 err_put_musb:
+	of_node_put(np);
 	platform_device_put(musb);
 
 	return ret;
author maintainer
#270 lore raw

PATCH 255/332 USB: quirks: add NO_LPM for Lenovo ThinkPad USB-C Dock Gen2 hub controllers

In reply to Greg Kroah-Hartman

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

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

From: Stephen J. Fuhry <[email protected]>

commit 9ddb9c0deca48d2c2a22ebf4d2f35c925a520328 upstream.

The Lenovo ThinkPad USB-C Dock Gen2 (17ef:a391, 17ef:a392) hub controllers exhibit link instability when USB Link Power Management is enabled, similar to the dock's Ethernet adapter (17ef:a387) which already carries USB_QUIRK_NO_LPM.

When the dock reconnects after a transient disconnect, the hub controllers enter LPM states between re-enumeration retries, causing repeated disconnect/reconnect cycles lasting up to two minutes. Disabling LPM for these devices restores stable enumeration.

---
 drivers/usb/core/quirks.c |    4 ++++
 1 file changed, 4 insertions(+)
drivers/usb/core/quirks.c +4 -0
@@ -513,6 +513,10 @@ static const struct usb_device_id usb_qu
 	/* Lenovo ThinkPad USB-C Dock Gen2 Ethernet (RTL8153 GigE) */
 	{ USB_DEVICE(0x17ef, 0xa387), .driver_info = USB_QUIRK_NO_LPM },
 
+	/* Lenovo ThinkPad USB-C Dock Gen2 USB 3.1 and USB 2.0 hub controllers */
+	{ USB_DEVICE(0x17ef, 0xa391), .driver_info = USB_QUIRK_NO_LPM },
+	{ USB_DEVICE(0x17ef, 0xa392), .driver_info = USB_QUIRK_NO_LPM },
+
 	/* BUILDWIN Photo Frame */
 	{ USB_DEVICE(0x1908, 0x1315), .driver_info =
 			USB_QUIRK_HONOR_BNUMINTERFACES },
author maintainer
#271 lore raw

PATCH 256/332 usb: storage: Add quirks for PNY Elite Portable SSD

In reply to Greg Kroah-Hartman

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

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

From: Sam Burkels <[email protected]>

commit b53ebb811e00be50a779ce4e7aee604178b4a825 upstream.

The PNY Elite Portable SSD (USB ID 154b:f009) is a sibling of the already-quirked PNY Pro Elite SSDs (154b:f00b and 154b:f00d). Like its siblings, it uses a Phison-based USB-SATA bridge that exhibits firmware bugs when bound to the uas driver.

Without quirks, the device fails to complete READ CAPACITY commands when accessed over UAS on a SuperSpeed (USB 3) port. The device enumerates and reports as a SCSI direct-access device, but reports zero logical blocks and never finishes spin-up:

usb 2-3: new SuperSpeed USB device number 8 using xhci_hcd usb 2-3: New USB device found, idVendor=154b, idProduct=f009 usb 2-3: Product: PNY ELITE PSSD
usb 2-3: Manufacturer: PNY
scsi host0: uas
scsi 0:0:0:0: Direct-Access PNY PNY ELITE PSSD 0
sd 0:0:0:0: [sda] Spinning up disk...
[...10+ seconds of polling, no progress...]
sd 0:0:0:0: [sda] Read Capacity(16) failed: hostbyte=DID_ERROR sd 0:0:0:0: [sda] Read Capacity(10) failed: hostbyte=DID_ERROR sd 0:0:0:0: [sda] 0 512-byte logical blocks: (0 B/0 B)

Tested each individual quirk to find the minimum that fixes this:
- US_FL_NO_ATA_1X alone: device hangs on spin-up
- US_FL_NO_REPORT_OPCODES alone: works on USB 2.0, hangs on USB 3.0
- US_FL_NO_ATA_1X | US_FL_NO_REPORT_OPCODES: works on both

With both quirks the device enumerates correctly while still using the uas driver, and delivers full UAS throughput (~281 MB/s sequential read on a USB 3.0 Gen 1 port).

The existing PNY Pro Elite entries (f00b, f00d) only set NO_ATA_1X, but this device additionally chokes on REPORT OPCODES under SuperSpeed.

---
 drivers/usb/storage/unusual_uas.h |    7 +++++++
 1 file changed, 7 insertions(+)
drivers/usb/storage/unusual_uas.h +7 -0
@@ -132,6 +132,13 @@ UNUSUAL_DEV(0x152d, 0x0583, 0x0000, 0x99
 		USB_SC_DEVICE, USB_PR_DEVICE, NULL,
 		US_FL_NO_REPORT_OPCODES),
 
+/* Reported-by: Sam Burkels <[email protected]> */
+UNUSUAL_DEV(0x154b, 0xf009, 0x0000, 0x9999,
+		"PNY",
+		"PNY ELITE PSSD",
+		USB_SC_DEVICE, USB_PR_DEVICE, NULL,
+		US_FL_NO_ATA_1X | US_FL_NO_REPORT_OPCODES),
+
 /* Reported-by: Thinh Nguyen <[email protected]> */
 UNUSUAL_DEV(0x154b, 0xf00b, 0x0000, 0x9999,
 		"PNY",
author maintainer
#272 lore raw

PATCH 257/332 usbip: vudc: Fix use after free bug in vudc_remove due to race condition

In reply to Greg Kroah-Hartman

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

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

From: Michael Bommarito <[email protected]>

commit d96209626a29ea64666be98c30b30ac82e5f1be6 upstream.

This patch follows up Zheng Wang's 2023 report of a use-after-free in vudc_remove(). The original thread stalled on Shuah Khan's request for runtime testing of the unplug/unbind path. This patch supplies that testing and keeps Zheng's original fix shape.

In vudc_probe(), v_init_timer() binds udc->tr_timer.timer to v_timer(). usbip_sockfd_store() starts the timer via v_start_timer()/v_kick_timer(). vudc_remove() can then free the containing struct vudc while the timer is still pending or executing.

KASAN confirms the race on an unpatched x86_64 QEMU guest with CONFIG_KASAN=y, CONFIG_USBIP_VUDC=y, CONFIG_USB_ZERO=y, and a tight loop that repeatedly writes a socket fd to usbip_sockfd, closes the socket pair, and unbinds/rebinds usbip-vudc.0:

  BUG: KASAN: slab-use-after-free in __run_timer_base.part.0+0x8ba/0x8e0
  Write of size 8 at addr ffff888001b80740 by task trigger_and_unb/239
  Allocated by task 239:
    vudc_probe+0x4d/0xaa0
  Freed by task 239:
    kfree+0x18f/0x520
    device_release_driver_internal+0x388/0x540
    unbind_store+0xd9/0x100

This lands in the timer core rather than v_timer() itself because the embedded timer_list is being walked after its containing struct vudc has already been freed. The underlying lifetime bug is the same one Zheng reported.

With v_stop_timer() called from vudc_remove() and the timer deleted synchronously, the same harness completed 5000 bind/unbind iterations with no KASAN report.

---
 drivers/usb/usbip/vudc_dev.c      |    1 +
 drivers/usb/usbip/vudc_transfer.c |    3 ++-
 2 files changed, 3 insertions(+), 1 deletion(-)
drivers/usb/usbip/vudc_dev.c +1 -0
@@ -632,6 +632,7 @@ void vudc_remove(struct platform_device
 {
 	struct vudc *udc = platform_get_drvdata(pdev);
 
+	v_stop_timer(udc);
 	usb_del_gadget_udc(&udc->gadget);
 	cleanup_vudc_hw(udc);
 	kfree(udc);
drivers/usb/usbip/vudc_transfer.c +2 -1
@@ -490,7 +490,8 @@ void v_stop_timer(struct vudc *udc)
 {
 	struct transfer_timer *t = &udc->tr_timer;
 
-	/* timer itself will take care of stopping */
+	/* Delete the timer synchronously before teardown frees udc. */
 	dev_dbg(&udc->pdev->dev, "timer stop");
+	timer_delete_sync(&t->timer);
 	t->state = VUDC_TR_STOPPED;
 }
author maintainer
#273 lore raw

PATCH 258/332 usb: usbtmc: check URB actual_length for interrupt-IN notifications

In reply to Greg Kroah-Hartman

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

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

From: Heitor Alves de Siqueira <[email protected]>

commit 52f2ad3f7e5eb3b5908e1d685d4342519dc9cfcd upstream.

USBTMC devices can use an optional interrupt endpoint for notification messages. These typically contain two-byte headers indicating the payload format, but the driver does not check if these headers are present before accessing the data buffers. In cases where the URB actual_length is not enough to fit these headers, the driver will either cause an out-of-bounds read, or consume stale leftover data from a previous notification.

Fix by checking if actual_data contains enough bytes for the headers, otherwise resubmit URB to the interrupt endpoint.

---
 drivers/usb/class/usbtmc.c |    8 ++++++++
 1 file changed, 8 insertions(+)
drivers/usb/class/usbtmc.c +8 -0
@@ -2306,6 +2306,14 @@ static void usbtmc_interrupt(struct urb
 
 	switch (status) {
 	case 0: /* SUCCESS */
+		/* ensure at least two bytes of headers were transferred */
+		if (urb->actual_length < 2) {
+			dev_warn(dev,
+				"actual length %d not sufficient for interrupt headers\n",
+				urb->actual_length);
+			goto exit;
+		}
+
 		/* check for valid STB notification */
 		if (data->iin_buffer[0] > 0x81) {
 			data->bNotify1 = data->iin_buffer[0];
author maintainer
#274 lore raw

PATCH 259/332 usb: usbtmc: reject interrupt endpoints with small wMaxPacketSize

In reply to Greg Kroah-Hartman

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

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

From: Heitor Alves de Siqueira <[email protected]>

commit 121d2f682ba912b1427cddca7cf84840f41cc620 upstream.

The USB488 subclass specification requires interrupt wMaxPacketSize to be 0x02, unless the device sends vendor-specific notifications. Endpoints that advertise less than 2 bytes for wMaxPacketSize are unlikely to work with the current driver, as URBs will not have enough space for interrupt headers. Considering that any notification URBs will be ignored by the driver, reject these endpoints early during probe.

---
 drivers/usb/class/usbtmc.c |    6 ++++++
 1 file changed, 6 insertions(+)
drivers/usb/class/usbtmc.c +6 -0
@@ -2440,6 +2440,12 @@ static int usbtmc_probe(struct usb_inter
 		data->iin_ep = int_in->bEndpointAddress;
 		data->iin_wMaxPacketSize = usb_endpoint_maxp(int_in);
 		data->iin_interval = int_in->bInterval;
+		/* wMaxPacketSize should be 0x02 or more as per USB488 Table 22 */
+		if (iface_desc->desc.bInterfaceProtocol == 1 &&
+		    data->iin_wMaxPacketSize < 2) {
+			retcode = -EINVAL;
+			goto err_put;
+		}
 		dev_dbg(&intf->dev, "Found Int in endpoint at %u\n",
 				data->iin_ep);
 	}
author maintainer
#275 lore raw

PATCH 260/332 usb: typec: tipd: Fix error code in tps6598x_probe()

In reply to Greg Kroah-Hartman

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

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

From: Dan Carpenter <[email protected]>

commit b02900c85a6423cf9b3dcc6b47bf060c85075e69 upstream.

Set the error code on these two error paths. The existing code returns success.

Fixes 77ed2f4538da ("usb: typec: tipd: Use read_power_status function in probe")
Fixes 04041fd7d6ec ("usb: typec: tipd: Read data status in probe and cache its value")
Cc stable <[email protected]>
---
 drivers/usb/typec/tipd/core.c |    1 +
 1 file changed, 1 insertion(+)
drivers/usb/typec/tipd/core.c +1 -0
@@ -1835,6 +1835,7 @@ static int tps6598x_probe(struct i2c_cli
 		goto err_role_put;
 
 	if (status & TPS_STATUS_PLUG_PRESENT) {
+		ret = -EINVAL;
 		if (!tps6598x_read_power_status(tps))
 			goto err_unregister_port;
 		if (!tps->data->read_data_status(tps))
author maintainer
#276 lore raw

PATCH 261/332 usb: typec: tcpm: improve handling of DISCOVER_MODES failures

In reply to Greg Kroah-Hartman

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

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

From: Sebastian Reichel <[email protected]>

commit c06e6cd488194e37ed4dc29d1488d1ffb760de60 upstream.

UGREEN USB-C Multifunction Adapter Model CM512 (AKA "Revodok 107") exposes two SVIDs: 0xff01 (DP Alt Mode) and 0x1d5c. The DISCOVER_MODES step succeeds for 0xff01 and gets a NAK for 0x1d5c. Currently this results in DP Alt Mode not being registered either, since the modes are only registered once all of them have been discovered. The NAK results in the processing being stopped and thus no Alt modes being registered.

Improve the situation by handling the NAK gracefully and continue processing the other modes.

Before this change, the TCPM log ends like this:

(more log entries before this)

[    5.028287] AMS DISCOVER_SVIDS finished
[    5.028291] cc:=4
[    5.040040] SVID 1: 0xff01
[    5.040054] SVID 2: 0x1d5c
[    5.040082] AMS DISCOVER_MODES start
[    5.040096] PD TX, header: 0x1b6f
[    5.050946] PD TX complete, status: 0
[    5.059609] PD RX, header: 0x264f [1]
[    5.059626] Rx VDM cmd 0xff018043 type 1 cmd 3 len 2
[    5.059640] AMS DISCOVER_MODES finished
[    5.059644] cc:=4
[    5.069994]  Alternate mode 0: SVID 0xff01, VDO 1: 0x000c0045
[    5.070029] AMS DISCOVER_MODES start
[    5.070043] PD TX, header: 0x1d6f
[    5.081139] PD TX complete, status: 0
[    5.087498] PD RX, header: 0x184f [1]
[    5.087515] Rx VDM cmd 0x1d5c8083 type 2 cmd 3 len 1
[    5.087529] AMS DISCOVER_MODES finished
[    5.087534] cc:=4

(no further log entries after this point)

After this patch the TCPM log looks exactly the same, but then continues like this:

[    5.100222] Skip SVID 0x1d5c (failed to discover mode)
[    5.101699] AMS DFP_TO_UFP_ENTER_MODE start

(log goes on as the system initializes DP AltMode)

---
 drivers/usb/typec/tcpm/tcpm.c |   97 ++++++++++++++++++++++++++----------------
 1 file changed, 61 insertions(+), 36 deletions(-)
drivers/usb/typec/tcpm/tcpm.c +61 -36
@@ -2002,6 +2002,55 @@ static bool tcpm_cable_vdm_supported(str
 	       tcpm_can_communicate_sop_prime(port);
 }
 
+static int tcpm_handle_discover_mode(struct tcpm_port *port, u32 *response,
+				     enum tcpm_transmit_type rx_sop_type,
+				     enum tcpm_transmit_type *response_tx_sop_type)
+{
+	struct typec_port *typec = port->typec_port;
+	struct pd_mode_data *modep;
+
+	if (rx_sop_type == TCPC_TX_SOP) {
+		modep = &port->mode_data;
+		modep->svid_index++;
+
+		if (modep->svid_index < modep->nsvids) {
+			u16 svid = modep->svids[modep->svid_index];
+			*response_tx_sop_type = TCPC_TX_SOP;
+			response[0] = VDO(svid, 1,
+					  typec_get_negotiated_svdm_version(typec),
+					  CMD_DISCOVER_MODES);
+			return 1;
+		}
+
+		if (tcpm_cable_vdm_supported(port)) {
+			*response_tx_sop_type = TCPC_TX_SOP_PRIME;
+			response[0] = VDO(USB_SID_PD, 1,
+					  typec_get_cable_svdm_version(typec),
+					  CMD_DISCOVER_SVID);
+			return 1;
+		}
+
+		tcpm_register_partner_altmodes(port);
+	} else if (rx_sop_type == TCPC_TX_SOP_PRIME) {
+		modep = &port->mode_data_prime;
+		modep->svid_index++;
+
+		if (modep->svid_index < modep->nsvids) {
+			u16 svid = modep->svids[modep->svid_index];
+			*response_tx_sop_type = TCPC_TX_SOP_PRIME;
+			response[0] = VDO(svid, 1,
+					  typec_get_cable_svdm_version(typec),
+					  CMD_DISCOVER_MODES);
+			return 1;
+		}
+
+		tcpm_register_plug_altmodes(port);
+		tcpm_register_partner_altmodes(port);
+	}
+
+	return 0;
+}
+
 static int tcpm_pd_svdm(struct tcpm_port *port, struct typec_altmode *adev,
 			const u32 *p, int cnt, u32 *response,
 			enum adev_actions *adev_action,
@@ -2259,41 +2308,11 @@ static int tcpm_pd_svdm(struct tcpm_port
 			}
 			break;
 		case CMD_DISCOVER_MODES:
-			if (rx_sop_type == TCPC_TX_SOP) {
-				/* 6.4.4.3.3 */
-				svdm_consume_modes(port, p, cnt, rx_sop_type);
-				modep->svid_index++;
-				if (modep->svid_index < modep->nsvids) {
-					u16 svid = modep->svids[modep->svid_index];
-					*response_tx_sop_type = TCPC_TX_SOP;
-					response[0] = VDO(svid, 1, svdm_version,
-							  CMD_DISCOVER_MODES);
-					rlen = 1;
-				} else if (tcpm_cable_vdm_supported(port)) {
-					*response_tx_sop_type = TCPC_TX_SOP_PRIME;
-					response[0] = VDO(USB_SID_PD, 1,
-							  typec_get_cable_svdm_version(typec),
-							  CMD_DISCOVER_SVID);
-					rlen = 1;
-				} else {
-					tcpm_register_partner_altmodes(port);
-				}
-			} else if (rx_sop_type == TCPC_TX_SOP_PRIME) {
-				/* 6.4.4.3.3 */
-				svdm_consume_modes(port, p, cnt, rx_sop_type);
-				modep_prime->svid_index++;
-				if (modep_prime->svid_index < modep_prime->nsvids) {
-					u16 svid = modep_prime->svids[modep_prime->svid_index];
-					*response_tx_sop_type = TCPC_TX_SOP_PRIME;
-					response[0] = VDO(svid, 1,
-							  typec_get_cable_svdm_version(typec),
-							  CMD_DISCOVER_MODES);
-					rlen = 1;
-				} else {
-					tcpm_register_plug_altmodes(port);
-					tcpm_register_partner_altmodes(port);
-				}
-			}
+			/* 6.4.4.3.3 */
+			svdm_consume_modes(port, p, cnt, rx_sop_type);
+			rlen = tcpm_handle_discover_mode(port, response,
+							 rx_sop_type,
+							 response_tx_sop_type);
 			break;
 		case CMD_ENTER_MODE:
 			*response_tx_sop_type = rx_sop_type;
@@ -2336,9 +2355,15 @@ static int tcpm_pd_svdm(struct tcpm_port
 		switch (cmd) {
 		case CMD_DISCOVER_IDENT:
 		case CMD_DISCOVER_SVID:
-		case CMD_DISCOVER_MODES:
 		case VDO_CMD_VENDOR(0) ... VDO_CMD_VENDOR(15):
 			break;
+		case CMD_DISCOVER_MODES:
+			tcpm_log(port, "Skip SVID 0x%04x (failed to discover mode)",
+				 PD_VDO_SVID_SVID0(p[0]));
+			rlen = tcpm_handle_discover_mode(port, response,
+							 rx_sop_type,
+							 response_tx_sop_type);
+			break;
 		case CMD_ENTER_MODE:
 			/* Back to USB Operation */
 			*adev_action = ADEV_NOTIFY_USB_AND_QUEUE_VDM;
author maintainer
#277 lore raw

PATCH 262/332 usb: typec: ucsi: Check if power role change actually happened before handling

In reply to Greg Kroah-Hartman

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

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

From: Myrrh Periwinkle <[email protected]>

commit b80e7d34c7ea6a564525119d6138fbb577a23dba upstream.

The CrOS EC may send a connector status change event with the power direction changed flag set even if the power direction hasn't actually changed after initiating a SET_PDR command internally [1]. In practice this happens on every system suspend due to other changes performed by the EC [2][3][4], causing suspend to fail.

Fix this by checking if the power role change actually happened before handling it.

[1]: https://source.chromium.org/chromiumos/chromiumos/codesearch/+/main:src/platform/ec/zephyr/subsys/pd_controller/pdc_power_mgmt.c;l=1689;drc=2d5a1cffce4e5ac8a39442cb3b764d2d5e1cf794
[2]: https://source.chromium.org/chromiumos/chromiumos/codesearch/+/main:src/platform/ec/zephyr/subsys/pd_controller/pdc_power_mgmt.c;l=3923;drc=2d5a1cffce4e5ac8a39442cb3b764d2d5e1cf794
[3]: https://source.chromium.org/chromiumos/chromiumos/codesearch/+/main:src/platform/ec/zephyr/subsys/pd_controller/pdc_power_mgmt.c;l=5094;drc=2d5a1cffce4e5ac8a39442cb3b764d2d5e1cf794
[4]: https://source.chromium.org/chromiumos/chromiumos/codesearch/+/main:src/platform/ec/zephyr/subsys/pd_controller/pdc_power_mgmt.c;l=2229;drc=2d5a1cffce4e5ac8a39442cb3b764d2d5e1cf794

Cc stable <[email protected]>
Fixes 7616f006db07 ("usb: typec: ucsi: Update power_supply on power role change")
---
 drivers/usb/typec/ucsi/ucsi.c |    6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)
drivers/usb/typec/ucsi/ucsi.c +4 -2
@@ -1267,7 +1267,7 @@ static void ucsi_handle_connector_change
 						  work);
 	struct ucsi *ucsi = con->ucsi;
 	u8 curr_scale, volt_scale;
-	enum typec_role role;
+	enum typec_role role, prev_role;
 	u16 change;
 	int ret;
 	u32 val;
@@ -1278,6 +1278,8 @@ static void ucsi_handle_connector_change
 		dev_err_once(ucsi->dev, "%s entered without EVENT_PENDING\n",
 			     __func__);
 
+	prev_role = UCSI_CONSTAT(con, PWR_DIR);
+
 	ret = ucsi_get_connector_status(con, true);
 	if (ret) {
 		dev_err(ucsi->dev, "%s: GET_CONNECTOR_STATUS failed (%d)\n",
@@ -1294,7 +1296,7 @@ static void ucsi_handle_connector_change
 	change = UCSI_CONSTAT(con, CHANGE);
 	role = UCSI_CONSTAT(con, PWR_DIR);
 
-	if (change & UCSI_CONSTAT_POWER_DIR_CHANGE) {
+	if ((change & UCSI_CONSTAT_POWER_DIR_CHANGE) && role != prev_role) {
 		typec_set_pwr_role(con->port, role);
 		ucsi_port_psy_changed(con);
 
author maintainer
#278 lore raw

PATCH 263/332 usb: typec: ucsi: Dont update power_supply on power role change if not connected

In reply to Greg Kroah-Hartman

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

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

From: Myrrh Periwinkle <[email protected]>

commit d98d413ca65d0790a8f3695d0a5845538958ab84 upstream.

We only need to update the power_supply on power role change if the port is connected, because otherwise the online status should be the same for both cases.

Cc stable <[email protected]>
Fixes 7616f006db07 ("usb: typec: ucsi: Update power_supply on power role change")
---
 drivers/usb/typec/ucsi/ucsi.c |    7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)
drivers/usb/typec/ucsi/ucsi.c +6 -1
@@ -1298,7 +1298,12 @@ static void ucsi_handle_connector_change
 
 	if ((change & UCSI_CONSTAT_POWER_DIR_CHANGE) && role != prev_role) {
 		typec_set_pwr_role(con->port, role);
-		ucsi_port_psy_changed(con);
+
+		/* Some power_supply properties vary depending on the power direction when
+		 * connected
+		 */
+		if (UCSI_CONSTAT(con, CONNECTED))
+			ucsi_port_psy_changed(con);
 
 		/* Complete pending power role swap */
 		if (!completion_done(&con->complete))
author maintainer
#279 lore raw

PATCH 264/332 USB: serial: option: add MeiG SRM813Q

In reply to Greg Kroah-Hartman

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

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

From: Jan Volckaert <[email protected]>

commit 7d2b37d3e42d19071b62f4ddbee6e16e905efbf1 upstream.

Add support for the Qualcomm Technology Snapdragon X35-based MeiG SRM813Q module.

The module can be put in different modes via AT commands to enable/disable GPS functionality:

MODEM - PPP mode(2dee:4d63): AT+SER=1,1

If#= 0: RMNET
If#= 1: DIAG/ADB
If#= 2: MODEM
If#= 3: AT

P:  Vendor=2dee ProdID=4d63 Rev=05.15
S:  Manufacturer=MEIG
S:  Product=LTE-A Module
S:  SerialNumber=1bd51f0e
C:  #Ifs= 4 Cfg#= 1 Atr=80 MxPwr=500mA
I:  If#= 0 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=ff Prot=50 Driver=qmi_wwan
E:  Ad=01(O) Atr=02(Bulk) MxPS= 512 Ivl=0ms
E:  Ad=81(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms
E:  Ad=82(I) Atr=03(Int.) MxPS=   8 Ivl=32ms
I:  If#= 1 Alt= 0 #EPs= 2 Cls=ff(vend.) Sub=ff Prot=30 Driver=option
E:  Ad=02(O) Atr=02(Bulk) MxPS= 512 Ivl=0ms
E:  Ad=83(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms
I:  If#= 2 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=ff Prot=40 Driver=option
E:  Ad=03(O) Atr=02(Bulk) MxPS= 512 Ivl=0ms
E:  Ad=84(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms
E:  Ad=85(I) Atr=03(Int.) MxPS=  10 Ivl=32ms
I:  If#= 3 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=ff Prot=40 Driver=option
E:  Ad=04(O) Atr=02(Bulk) MxPS= 512 Ivl=0ms
E:  Ad=86(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms
E:  Ad=87(I) Atr=03(Int.) MxPS=  10 Ivl=32ms

NMEA mode(2dee:4d64): AT+SER=51,1

If#= 0: RMNET
If#= 1: DIAG/ADB
If#= 2: NMEA
If#= 3: AT

P:  Vendor=2dee ProdID=4d64 Rev=05.15
S:  Manufacturer=MEIG
S:  Product=LTE-A Module
S:  SerialNumber=1bd51f0e
C:  #Ifs= 4 Cfg#= 1 Atr=80 MxPwr=500mA
I:  If#= 0 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=ff Prot=50 Driver=qmi_wwan
E:  Ad=01(O) Atr=02(Bulk) MxPS= 512 Ivl=0ms
E:  Ad=81(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms
E:  Ad=82(I) Atr=03(Int.) MxPS=   8 Ivl=32ms
I:  If#= 1 Alt= 0 #EPs= 2 Cls=ff(vend.) Sub=ff Prot=30 Driver=option
E:  Ad=02(O) Atr=02(Bulk) MxPS= 512 Ivl=0ms
E:  Ad=83(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms
I:  If#= 2 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=ff Prot=60 Driver=option
E:  Ad=03(O) Atr=02(Bulk) MxPS= 512 Ivl=0ms
E:  Ad=84(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms
E:  Ad=85(I) Atr=03(Int.) MxPS=  10 Ivl=32ms
I:  If#= 3 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=ff Prot=40 Driver=option
E:  Ad=04(O) Atr=02(Bulk) MxPS= 512 Ivl=0ms
E:  Ad=86(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms
E:  Ad=87(I) Atr=03(Int.) MxPS=  10 Ivl=32ms
---
 drivers/usb/serial/option.c |    6 ++++++
 1 file changed, 6 insertions(+)
drivers/usb/serial/option.c +6 -0
@@ -2450,6 +2450,12 @@ static const struct usb_device_id option
 	{ USB_DEVICE_AND_INTERFACE_INFO(0x2dee, 0x4d38, 0xff, 0xff, 0x30) },	/* MeiG Smart SRM825WN (Diag) */
 	{ USB_DEVICE_AND_INTERFACE_INFO(0x2dee, 0x4d38, 0xff, 0xff, 0x40) },	/* MeiG Smart SRM825WN (AT) */
 	{ USB_DEVICE_AND_INTERFACE_INFO(0x2dee, 0x4d38, 0xff, 0xff, 0x60) },	/* MeiG Smart SRM825WN (NMEA) */
+	{ USB_DEVICE_AND_INTERFACE_INFO(0x2dee, 0x4d63, 0xff, 0xff, 0x30) },	/* MeiG SRM813Q (Diag) */
+	{ USB_DEVICE_AND_INTERFACE_INFO(0x2dee, 0x4d63, 0xff, 0xff, 0x40) },	/* MeiG SRM813Q (AT) */
+	{ USB_DEVICE_AND_INTERFACE_INFO(0x2dee, 0x4d64, 0xff, 0xff, 0x30) },	/* MeiG SRM813Q (Diag) */
+	{ USB_DEVICE_AND_INTERFACE_INFO(0x2dee, 0x4d64, 0xff, 0xff, 0x40) },	/* MeiG SRM813Q (AT) */
+	{ USB_DEVICE_AND_INTERFACE_INFO(0x2dee, 0x4d64, 0xff, 0xff, 0x60) },	/* MeiG SRM813Q (NMEA) */
+
 	{ USB_DEVICE_INTERFACE_CLASS(0x2df3, 0x9d03, 0xff) },			/* LongSung M5710 */
 	{ USB_DEVICE_INTERFACE_CLASS(0x305a, 0x1404, 0xff) },			/* GosunCn GM500 RNDIS */
 	{ USB_DEVICE_INTERFACE_CLASS(0x305a, 0x1405, 0xff) },			/* GosunCn GM500 MBIM */
author maintainer
#280 lore raw

PATCH 265/332 USB: serial: option: add missing RSVD(5) flag for Rolling RW135R-GL

In reply to Greg Kroah-Hartman

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

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

From: Wanquan Zhong <[email protected]>

commit 689f2facc689c8add11d7ff69fbbad17d65ee596 upstream.

The RW135R-GL entry added in commit 01e8d0f74222 ("USB: serial: option:
add support for Rolling Wireless RW135R-GL") was missing the .driver_info = RSVD(5) flag used by other Rolling Wireless MBIM laptop modules (e.g. RW135-GL and RW350-GL).

Without this flag, the option driver incorrectly binds to the reserved ADB interface (If#5) in multi-interface USB modes, causing AT/MBIM communication failures after mode switching. This matches the handling of other Rolling Wireless MBIM devices.

- VID:PID 33f8:1003, RW135R-GL for laptop debug M.2 cards (with MBIM interface for Linux/Chrome OS)

0x1003: mbim, diag, AT, pipe

Here are the outputs of usb-devices:

T:  Bus=03 Lev=01 Prnt=01 Port=04 Cnt=02 Dev#=  8 Spd=480  MxCh= 0
D:  Ver= 2.00 Cls=00(>ifc ) Sub=00 Prot=00 MxPS=64 #Cfgs=  1
P:  Vendor=33f8 ProdID=1003 Rev= 5.15
S:  Manufacturer=Rolling Wireless S.a.r.l.
S:  Product=Rolling RW135R-GL Module
S:  SerialNumber=12345678
C:* #Ifs= 5 Cfg#= 1 Atr=a0 MxPwr=500mA
A:  FirstIf#= 0 IfCount= 2 Cls=02(comm.) Sub=0e Prot=00
I:* If#= 0 Alt= 0 #EPs= 1 Cls=02(comm.) Sub=0e Prot=00 Driver=cdc_mbim
E:  Ad=82(I) Atr=03(Int.) MxPS=  64 Ivl=32ms
I:  If#= 1 Alt= 0 #EPs= 0 Cls=0a(data ) Sub=00 Prot=02 Driver=cdc_mbim
I:* If#= 1 Alt= 1 #EPs= 2 Cls=0a(data ) Sub=00 Prot=02 Driver=cdc_mbim
E:  Ad=81(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms
E:  Ad=01(O) Atr=02(Bulk) MxPS= 512 Ivl=0ms
I:* If#= 2 Alt= 0 #EPs= 2 Cls=ff(vend.) Sub=00 Prot=00 Driver=option
E:  Ad=83(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms
E:  Ad=02(O) Atr=02(Bulk) MxPS= 512 Ivl=0ms
I:* If#= 3 Alt= 0 #EPs= 2 Cls=ff(vend.) Sub=ff Prot=30 Driver=option
E:  Ad=03(O) Atr=02(Bulk) MxPS= 512 Ivl=0ms
E:  Ad=84(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms
I:* If#= 4 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=ff Prot=40 Driver=option
E:  Ad=86(I) Atr=03(Int.) MxPS=  10 Ivl=32ms
E:  Ad=85(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms
E:  Ad=04(O) Atr=02(Bulk) MxPS= 512 Ivl=0ms

- VID:PID 33f8:1003, RW135R-GL for laptop debug M.2 cards (with MBIM interface for Linux/Chrome OS)

0x1003: mbim, diag, AT, ADB, pipe

Here are the outputs of usb-devices:

T:  Bus=03 Lev=01 Prnt=01 Port=04 Cnt=02 Dev#=  7 Spd=480  MxCh= 0
D:  Ver= 2.00 Cls=00(>ifc ) Sub=00 Prot=00 MxPS=64 #Cfgs=  1
P:  Vendor=33f8 ProdID=1003 Rev= 5.15
S:  Manufacturer=Rolling Wireless S.a.r.l.
S:  Product=Rolling RW135R-GL Module
S:  SerialNumber=12345678
C:* #Ifs= 6 Cfg#= 1 Atr=a0 MxPwr=500mA
A:  FirstIf#= 0 IfCount= 2 Cls=02(comm.) Sub=0e Prot=00
I:* If#= 0 Alt= 0 #EPs= 1 Cls=02(comm.) Sub=0e Prot=00 Driver=cdc_mbim
E:  Ad=82(I) Atr=03(Int.) MxPS=  64 Ivl=32ms
I:  If#= 1 Alt= 0 #EPs= 0 Cls=0a(data ) Sub=00 Prot=02 Driver=cdc_mbim
I:* If#= 1 Alt= 1 #EPs= 2 Cls=0a(data ) Sub=00 Prot=02 Driver=cdc_mbim
E:  Ad=81(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms
E:  Ad=01(O) Atr=02(Bulk) MxPS= 512 Ivl=0ms
I:* If#= 2 Alt= 0 #EPs= 2 Cls=ff(vend.) Sub=00 Prot=00 Driver=option
E:  Ad=83(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms
E:  Ad=02(O) Atr=02(Bulk) MxPS= 512 Ivl=0ms
I:* If#= 3 Alt= 0 #EPs= 2 Cls=ff(vend.) Sub=ff Prot=30 Driver=option
E:  Ad=03(O) Atr=02(Bulk) MxPS= 512 Ivl=0ms
E:  Ad=84(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms
I:* If#= 4 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=ff Prot=40 Driver=option
E:  Ad=86(I) Atr=03(Int.) MxPS=  10 Ivl=32ms
E:  Ad=85(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms
E:  Ad=04(O) Atr=02(Bulk) MxPS= 512 Ivl=0ms
I:* If#= 5 Alt= 0 #EPs= 2 Cls=ff(vend.) Sub=42 Prot=01 Driver=(none)
E:  Ad=05(O) Atr=02(Bulk) MxPS= 512 Ivl=0ms
E:  Ad=87(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms

- VID:PID 33f8:1003, RW135R-GL for laptop debug M.2 cards (with MBIM interface for Linux/Chrome OS)

0x1003: mbim, pipe

Here are the outputs of usb-devices:

T:  Bus=03 Lev=01 Prnt=01 Port=04 Cnt=02 Dev#=  9 Spd=480  MxCh= 0
D:  Ver= 2.00 Cls=00(>ifc ) Sub=00 Prot=00 MxPS=64 #Cfgs=  1
P:  Vendor=33f8 ProdID=1003 Rev= 5.15
S:  Manufacturer=Rolling Wireless S.a.r.l.
S:  Product=Rolling RW135R-GL Module
S:  SerialNumber=12345678
C:* #Ifs= 3 Cfg#= 1 Atr=a0 MxPwr=500mA
A:  FirstIf#= 0 IfCount= 2 Cls=02(comm.) Sub=0e Prot=00
I:* If#= 0 Alt= 0 #EPs= 1 Cls=02(comm.) Sub=0e Prot=00 Driver=cdc_mbim
E:  Ad=82(I) Atr=03(Int.) MxPS=  64 Ivl=32ms
I:  If#= 1 Alt= 0 #EPs= 0 Cls=0a(data ) Sub=00 Prot=02 Driver=cdc_mbim
I:* If#= 1 Alt= 1 #EPs= 2 Cls=0a(data ) Sub=00 Prot=02 Driver=cdc_mbim
E:  Ad=81(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms
E:  Ad=01(O) Atr=02(Bulk) MxPS= 512 Ivl=0ms
I:* If#= 2 Alt= 0 #EPs= 2 Cls=ff(vend.) Sub=00 Prot=00 Driver=option
E:  Ad=83(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms
E:  Ad=02(O) Atr=02(Bulk) MxPS= 512 Ivl=0ms
Fixes 01e8d0f74222 ("USB: serial: option: add support for Rolling Wireless RW135R-GL")
---
 drivers/usb/serial/option.c |    3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/usb/serial/option.c +2 -1
@@ -2476,7 +2476,8 @@ static const struct usb_device_id option
 	{ USB_DEVICE_INTERFACE_CLASS(0x33f8, 0x0302, 0xff) },			/* Rolling RW101R-GL (laptop MBIM) */
 	{ USB_DEVICE_INTERFACE_CLASS(0x33f8, 0x0802, 0xff),			/* Rolling RW350-GL (laptop MBIM) */
 	  .driver_info = RSVD(5) },
-	{ USB_DEVICE_INTERFACE_CLASS(0x33f8, 0x1003, 0xff) },			/* Rolling RW135R-GL (laptop MBIM) */
+	{ USB_DEVICE_INTERFACE_CLASS(0x33f8, 0x1003, 0xff),			/* Rolling RW135R-GL (laptop MBIM) */
+	  .driver_info = RSVD(5) },
 	{ USB_DEVICE_AND_INTERFACE_INFO(0x3731, 0x0100, 0xff, 0xff, 0x30) },	/* NetPrisma LCUK54-WWD for Global */
 	{ USB_DEVICE_AND_INTERFACE_INFO(0x3731, 0x0100, 0xff, 0x00, 0x40) },
 	{ USB_DEVICE_AND_INTERFACE_INFO(0x3731, 0x0100, 0xff, 0xff, 0x40) },
author maintainer
#281 lore raw

PATCH 266/332 USB: serial: belkin_sa: validate interrupt status length

In reply to Greg Kroah-Hartman

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

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

From: Zhang Cen <[email protected]>

commit 4ce058df2ee02cc2a0f0fd5cd64ce6f1482a0b65 upstream.

The Belkin interrupt callback treats interrupt data as a four-byte status report and reads LSR/MSR fields at offsets 2 and 3. The interrupt-in buffer length is derived from endpoint wMaxPacketSize, and short interrupt transfers may complete successfully with a smaller actual_length.

Check the completed interrupt packet length before parsing status fields so short interrupt endpoints and short successful packets are ignored instead of causing out-of-bounds or stale status-byte reads.

KASAN report as below:

BUG: KASAN: slab-out-of-bounds in belkin_sa_read_int_callback()
Read of size 1
Call trace:
  belkin_sa_read_int_callback() (drivers/usb/serial/belkin_sa.c:202)
  __usb_hcd_giveback_urb() (drivers/usb/core/hcd.c:1630)
  dummy_timer() (?:?)
Fixes 1da177e4c3f4 ("Linux-2.6.12-rc2")
Assisted-by Codex:gpt-5.5
---
 drivers/usb/serial/belkin_sa.c |    3 +++
 1 file changed, 3 insertions(+)
drivers/usb/serial/belkin_sa.c +3 -0
@@ -194,6 +194,9 @@ static void belkin_sa_read_int_callback(
 
 	usb_serial_debug_data(&port->dev, __func__, urb->actual_length, data);
 
+	if (urb->actual_length < BELKIN_SA_MSR_INDEX + 1)
+		goto exit;
+
 	/* Handle known interrupt data */
 	/* ignore data[0] and data[1] */
 
author maintainer
#282 lore raw

PATCH 267/332 USB: serial: cypress_m8: validate interrupt packet headers

In reply to Greg Kroah-Hartman

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

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

From: Zhang Cen <[email protected]>

commit 9f9bfc80c67f35a275820da7e83a35dface08281 upstream.

cypress_read_int_callback() parses the interrupt-in buffer according to the selected Cypress packet format. Format 1 has a two-byte status/count header and format 2 has a one-byte combined status/count header. The usb-serial core sizes the interrupt-in buffer from the endpoint descriptor's wMaxPacketSize, and successful interrupt transfers can complete short when URB_SHORT_NOT_OK is not set.

Check that the completed packet contains the selected header before reading it. Malformed short reports are ignored and the interrupt URB is resubmitted through the existing retry path, preventing out-of-bounds header-byte reads.

KASAN report as below:
KASAN slab-out-of-bounds in cypress_read_int_callback+0x240/0x7f0

Read of size 1
Call trace:
  cypress_read_int_callback() (drivers/usb/serial/cypress_m8.c:1009)
  __usb_hcd_giveback_urb()
  dummy_timer()
Fixes 3416eaa1f8f8 ("USB: cypress_m8: Packet format is separate from characteristic size")
Assisted-by Codex:gpt-5.5
Fixes 3416eaa1f8f8 ("USB: cypress_m8: Packet format is separate from characteristic size")
Cc [email protected] # 2.6.26
[ johan: use constants in header length sanity checks ]
---
 drivers/usb/serial/cypress_m8.c |   12 +++++++++++-
 1 file changed, 11 insertions(+), 1 deletion(-)
drivers/usb/serial/cypress_m8.c +11 -1
@@ -1025,8 +1025,8 @@ static void cypress_read_int_callback(st
 	char tty_flag = TTY_NORMAL;
 	int bytes = 0;
 	int result;
-	int i = 0;
 	int status = urb->status;
+	int i;
 
 	switch (status) {
 	case 0: /* success */
@@ -1064,22 +1064,32 @@ static void cypress_read_int_callback(st
 
 	spin_lock_irqsave(&priv->lock, flags);
 	result = urb->actual_length;
+	i = 0;
 	switch (priv->pkt_fmt) {
 	default:
 	case packet_format_1:
 		/* This is for the CY7C64013... */
+		if (result < 2)
+			break;
 		priv->current_status = data[0] & 0xF8;
 		bytes = data[1] + 2;
 		i = 2;
 		break;
 	case packet_format_2:
 		/* This is for the CY7C63743... */
+		if (result < 1)
+			break;
 		priv->current_status = data[0] & 0xF8;
 		bytes = (data[0] & 0x07) + 1;
 		i = 1;
 		break;
 	}
 	spin_unlock_irqrestore(&priv->lock, flags);
+	if (i == 0) {
+		dev_dbg(dev, "%s - short packet received: %d bytes\n",
+			__func__, result);
+		goto continue_read;
+	}
 	if (result < bytes) {
 		dev_dbg(dev,
 			"%s - wrong packet size - received %d bytes but packet said %d bytes\n",
author maintainer
#283 lore raw

PATCH 268/332 USB: serial: digi_acceleport: fix memory corruption with small endpoints

In reply to Greg Kroah-Hartman

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

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

From: Johan Hovold <[email protected]>

commit cb3560e8eab1dfa1cac1ed52631adf8ec6ff2cd5 upstream.

Add the missing bulk-out buffer size sanity checks to avoid out-of-bounds memory accesses or slab corruption should a malicious device report smaller buffers than expected.

---
 drivers/usb/serial/digi_acceleport.c |   23 +++++++++++++++++++++--
 1 file changed, 21 insertions(+), 2 deletions(-)
drivers/usb/serial/digi_acceleport.c +21 -2
@@ -1229,15 +1229,34 @@ static int digi_port_init(struct usb_ser
 static int digi_startup(struct usb_serial *serial)
 {
 	struct digi_serial *serial_priv;
+	int oob_port_num;
 	int ret;
+	int i;
+
+	/*
+	 * The port bulk-out buffers must be large enough for header and
+	 * buffered data.
+	 */
+	for (i = 0; i < serial->type->num_ports; i++) {
+		if (serial->port[i]->bulk_out_size < DIGI_OUT_BUF_SIZE + 2)
+			return -EINVAL;
+	}
+
+	/*
+	 * The OOB port bulk-out buffer must be large enough for the two
+	 * commands in digi_set_modem_signals().
+	 */
+	oob_port_num = serial->type->num_ports;
+	if (serial->port[oob_port_num]->bulk_out_size < 8)
+		return -EINVAL;
 
 	serial_priv = kzalloc_obj(*serial_priv);
 	if (!serial_priv)
 		return -ENOMEM;
 
 	spin_lock_init(&serial_priv->ds_serial_lock);
-	serial_priv->ds_oob_port_num = serial->type->num_ports;
-	serial_priv->ds_oob_port = serial->port[serial_priv->ds_oob_port_num];
+	serial_priv->ds_oob_port_num = oob_port_num;
+	serial_priv->ds_oob_port = serial->port[oob_port_num];
 
 	ret = digi_port_init(serial_priv->ds_oob_port,
 						serial_priv->ds_oob_port_num);
author maintainer
#284 lore raw

PATCH 269/332 USB: serial: keyspan: fix missing indat transfer sanity check

In reply to Greg Kroah-Hartman

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

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

From: Johan Hovold <[email protected]>

commit ab8336a7e414f018430aa1af3a46944032f7ff96 upstream.

Add the missing sanity check on the size of usa49wg indat transfers to avoid parsing stale or uninitialised slab data.

Fixes 0ca1268e109a ("USB Serial Keyspan: add support for USA-49WG & USA-28XG")
Cc [email protected] # 2.6.23
---
 drivers/usb/serial/keyspan.c |    4 ++++
 1 file changed, 4 insertions(+)
drivers/usb/serial/keyspan.c +4 -0
@@ -1187,6 +1187,10 @@ static void usa49wg_indat_callback(struc
 	len = 0;
 
 	while (i < urb->actual_length) {
+		if (urb->actual_length - i < 3) {
+			dev_warn_ratelimited(&urb->dev->dev, "malformed indat packet\n");
+			break;
+		}
 
 		/* Check port number from message */
 		if (data[i] >= serial->num_ports) {
author maintainer
#285 lore raw

PATCH 270/332 USB: serial: mxuport: fix memory corruption with small endpoint

In reply to Greg Kroah-Hartman

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

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

From: Johan Hovold <[email protected]>

commit 4085f0dbb1ce2251c9a5938d693de6593f0ab2bd upstream.

Make sure that the bulk-out endpoint max packet size is at least eight bytes to avoid user-controlled slab corruption should a malicious device report a smaller size.

Fixes ee467a1f2066 ("USB: serial: add Moxa UPORT 12XX/14XX/16XX driver")
Cc Andrew Lunn <[email protected]>
---
 drivers/usb/serial/mxuport.c |    8 ++++++++
 1 file changed, 8 insertions(+)
drivers/usb/serial/mxuport.c +8 -0
@@ -962,6 +962,14 @@ static int mxuport_calc_num_ports(struct
 	 */
 	BUILD_BUG_ON(ARRAY_SIZE(epds->bulk_out) < 16);
 
+	/*
+	 * The bulk-out buffers must be large enough for the four-byte header
+	 * (and following data), but assume anything smaller than eight bytes
+	 * is broken.
+	 */
+	if (usb_endpoint_maxp(epds->bulk_out[0]) < 8)
+		return -EINVAL;
+
 	for (i = 1; i < num_ports; ++i)
 		epds->bulk_out[i] = epds->bulk_out[0];
 
author maintainer
#286 lore raw

PATCH 271/332 USB: serial: mct_u232: fix memory corruption with small endpoint

In reply to Greg Kroah-Hartman

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

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

From: Johan Hovold <[email protected]>

commit 915b36d701950503c4ea0f6e314b10868e59fce3 upstream.

The driver overrides the maximum transfer size for a specific device which only accepts 16 byte packets for its 32 byte bulk-out endpoint.

Make sure to never increase the maximum transfer size to prevent slab corruption should a malicious device report a smaller endpoint max packet size than expected.

---
 drivers/usb/serial/mct_u232.c |   21 +++++++++++----------
 1 file changed, 11 insertions(+), 10 deletions(-)
drivers/usb/serial/mct_u232.c +11 -10
@@ -378,6 +378,7 @@ static int mct_u232_port_probe(struct us
 {
 	struct usb_serial *serial = port->serial;
 	struct mct_u232_private *priv;
+	u16 pid;
 
 	/* check first to simplify error handling */
 	if (!serial->port[1] || !serial->port[1]->interrupt_in_urb) {
@@ -385,6 +386,16 @@ static int mct_u232_port_probe(struct us
 		return -ENODEV;
 	}
 
+	/*
+	 * Compensate for a hardware bug: although the Sitecom U232-P25
+	 * device reports a maximum output packet size of 32 bytes,
+	 * it seems to be able to accept only 16 bytes (and that's what
+	 * SniffUSB says too...)
+	 */
+	pid = le16_to_cpu(serial->dev->descriptor.idProduct);
+	if (pid == MCT_U232_SITECOM_PID)
+		port->bulk_out_size = min(16, port->bulk_out_size);
+
 	priv = kzalloc_obj(*priv);
 	if (!priv)
 		return -ENOMEM;
@@ -410,7 +421,6 @@ static void mct_u232_port_remove(struct
 
 static int  mct_u232_open(struct tty_struct *tty, struct usb_serial_port *port)
 {
-	struct usb_serial *serial = port->serial;
 	struct mct_u232_private *priv = usb_get_serial_port_data(port);
 	int retval = 0;
 	unsigned int control_state;
@@ -418,15 +428,6 @@ static int  mct_u232_open(struct tty_str
 	unsigned char last_lcr;
 	unsigned char last_msr;
 
-	/* Compensate for a hardware bug: although the Sitecom U232-P25
-	 * device reports a maximum output packet size of 32 bytes,
-	 * it seems to be able to accept only 16 bytes (and that's what
-	 * SniffUSB says too...)
-	 */
-	if (le16_to_cpu(serial->dev->descriptor.idProduct)
-						== MCT_U232_SITECOM_PID)
-		port->bulk_out_size = 16;
-
 	/* Do a defined restart: the normal serial device seems to
 	 * always turn on DTR and RTS here, so do the same. I'm not
 	 * sure if this is really necessary. But it should not harm
author maintainer
#287 lore raw

PATCH 272/332 USB: serial: mct_u232: fix missing interrupt-in transfer sanity check

In reply to Greg Kroah-Hartman

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

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

From: Johan Hovold <[email protected]>

commit 245aba83e3c288e176ed037a1f6b618b09e92ed8 upstream.

Add the missing sanity check on the size of interrupt-in transfers to avoid parsing stale or uninitialised slab data (and leaking it to user space).

---
 drivers/usb/serial/mct_u232.c |    5 +++++
 1 file changed, 5 insertions(+)
drivers/usb/serial/mct_u232.c +5 -0
@@ -544,6 +544,11 @@ static void mct_u232_read_int_callback(s
 		goto exit;
 	}
 
+	if (urb->actual_length < 2) {
+		dev_warn_ratelimited(&port->dev, "short interrupt-in packet\n");
+		goto exit;
+	}
+
 	/*
 	 * The interrupt-in pipe signals exceptional conditions (modem line
 	 * signal changes and errors). data[0] holds MSR, data[1] holds LSR.
author maintainer
#288 lore raw

PATCH 273/332 usb: gadget: uvc: hold opts->lock across XU walks in uvc_function_bind

In reply to Greg Kroah-Hartman

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

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

From: Kai Aizen <[email protected]>

commit 68aa70648b625fa684bc0b71bbfd905f4943ca20 upstream.

uvc_function_bind() walks &opts->extension_units twice without holding opts->lock:

- directly, for the iExtension string-descriptor fixup loop;
- indirectly, four times via uvc_copy_descriptors() (once per speed),

    where the helper iterates uvc->desc.extension_units (which aliases
    &opts->extension_units) to size and emit XU descriptors.

The configfs side (uvcg_extension_make / uvcg_extension_drop, in drivers/usb/gadget/function/uvc_configfs.c) takes opts->lock around its list_add_tail / list_del operations. A privileged userspace process that holds the configfs subtree open and writes the gadget UDC name to bind the function while concurrently rmdir()'ing an extensions subdir can race uvcg_extension_drop() against the bind-time list walks and dereference a freed struct uvcg_extension.

Hold opts->lock from the start of the XU string-descriptor fixup through the last uvc_copy_descriptors() call, releasing on the descriptor-error path via a new error_unlock label that drops the lock before falling through to the existing error label. This matches the locking discipline of the configfs callbacks and removes the only remaining unsynchronised reader of the XU list during bind.

Reachability: only privileged processes that can mount configfs and write to gadget UDC files can trigger the race, so this is a correctness fix rather than a security boundary.

Fixes 0525210c9840 ("usb: gadget: uvc: Allow definition of XUs in configfs")
Cc stable <[email protected]>
---
 drivers/usb/gadget/function/f_uvc.c |   28 +++++++++++++++++++++-------
 1 file changed, 21 insertions(+), 7 deletions(-)
drivers/usb/gadget/function/f_uvc.c +21 -7
@@ -769,6 +769,16 @@ uvc_function_bind(struct usb_configurati
 	uvc_ss_streaming_ep.bEndpointAddress = uvc->video.ep->address;
 
 	/*
+	 * Hold opts->lock across both the XU string-descriptor fixup below and
+	 * the descriptor-copy block further down.  Without this, configfs
+	 * uvcg_extension_drop() (which takes opts->lock) can race with the
+	 * list_for_each_entry() walks here and inside uvc_copy_descriptors(),
+	 * leading to a UAF on a freed struct uvcg_extension.  See
+	 * drivers/usb/gadget/function/uvc_configfs.c::uvcg_extension_drop().
+	 */
+	mutex_lock(&opts->lock);
+
+	/*
 	 * XUs can have an arbitrary string descriptor describing them. If they
 	 * have one pick up the ID.
 	 */
@@ -785,7 +795,7 @@ uvc_function_bind(struct usb_configurati
 				 ARRAY_SIZE(uvc_en_us_strings));
 	if (IS_ERR(us)) {
 		ret = PTR_ERR(us);
-		goto error;
+		goto error_unlock;
 	}
 
 	uvc_iad.iFunction = opts->iad_index ? cdev->usb_strings[opts->iad_index].id :
@@ -799,14 +809,14 @@ uvc_function_bind(struct usb_configurati
 
 	/* Allocate interface IDs. */
 	if ((ret = usb_interface_id(c, f)) < 0)
-		goto error;
+		goto error_unlock;
 	uvc_iad.bFirstInterface = ret;
 	uvc_control_intf.bInterfaceNumber = ret;
 	uvc->control_intf = ret;
 	opts->control_interface = ret;
 
 	if ((ret = usb_interface_id(c, f)) < 0)
-		goto error;
+		goto error_unlock;
 	uvc_streaming_intf_alt0.bInterfaceNumber = ret;
 	uvc_streaming_intf_alt1.bInterfaceNumber = ret;
 	uvc->streaming_intf = ret;
@@ -817,30 +827,32 @@ uvc_function_bind(struct usb_configurati
 	if (IS_ERR(f->fs_descriptors)) {
 		ret = PTR_ERR(f->fs_descriptors);
 		f->fs_descriptors = NULL;
-		goto error;
+		goto error_unlock;
 	}
 
 	f->hs_descriptors = uvc_copy_descriptors(uvc, USB_SPEED_HIGH);
 	if (IS_ERR(f->hs_descriptors)) {
 		ret = PTR_ERR(f->hs_descriptors);
 		f->hs_descriptors = NULL;
-		goto error;
+		goto error_unlock;
 	}
 
 	f->ss_descriptors = uvc_copy_descriptors(uvc, USB_SPEED_SUPER);
 	if (IS_ERR(f->ss_descriptors)) {
 		ret = PTR_ERR(f->ss_descriptors);
 		f->ss_descriptors = NULL;
-		goto error;
+		goto error_unlock;
 	}
 
 	f->ssp_descriptors = uvc_copy_descriptors(uvc, USB_SPEED_SUPER_PLUS);
 	if (IS_ERR(f->ssp_descriptors)) {
 		ret = PTR_ERR(f->ssp_descriptors);
 		f->ssp_descriptors = NULL;
-		goto error;
+		goto error_unlock;
 	}
 
+	mutex_unlock(&opts->lock);
+
 	/* Preallocate control endpoint request. */
 	uvc->control_req = usb_ep_alloc_request(cdev->gadget->ep0, GFP_KERNEL);
 	uvc->control_buf = kmalloc(UVC_MAX_REQUEST_SIZE, GFP_KERNEL);
@@ -872,6 +884,8 @@ uvc_function_bind(struct usb_configurati
 
 	return 0;
 
+error_unlock:
+	mutex_unlock(&opts->lock);
 v4l2_error:
 	v4l2_device_unregister(&uvc->v4l2_dev);
 error:
author maintainer
#289 lore raw

PATCH 274/332 usb: gadget: net2280: Fix double free in probe error path

In reply to Greg Kroah-Hartman

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

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

From: Guangshuo Li <[email protected]>

commit c8547c74988e0b5f4cbb1b895e2a57aae084f070 upstream.

usb_initialize_gadget() installs gadget_release() as the release callback for the embedded gadget device. The struct net2280 instance is therefore released through gadget_release() when the gadget device's last reference is dropped.

The probe error path calls net2280_remove(), which tears down the partially initialized device and drops the gadget reference with usb_put_gadget(). Calling kfree(dev) afterwards can free the same object again.

Drop the explicit kfree() and let the gadget device release callback handle the final free. This issue was found by a static analysis tool I am developing.

---
 drivers/usb/gadget/udc/net2280.c |    4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)
drivers/usb/gadget/udc/net2280.c +1 -3
@@ -3790,10 +3790,8 @@ static int net2280_probe(struct pci_dev
 	return 0;
 
 done:
-	if (dev) {
+	if (dev)
 		net2280_remove(pdev);
-		kfree(dev);
-	}
 	return retval;
 }
 
author maintainer
#290 lore raw

PATCH 275/332 usb: gadget: f_hid: fix device reference leak in hidg_alloc()

In reply to Greg Kroah-Hartman

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

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

From: Guangshuo Li <[email protected]>

commit 4f88d65def6f3c90121601b4f62a4c967f3063a6 upstream.

hidg_alloc() initializes hidg->dev with device_initialize() before calling dev_set_name(). If dev_set_name() fails, the function currently jumps to err_unlock and returns without calling put_device().

This leaves the device reference unbalanced and prevents hidg_release() from being called. Calling put_device() here is also safe, since hidg_release() only frees resources owned by hidg.

The issue was identified by a static analysis tool I developed and confirmed by manual review.

Route the dev_set_name() failure path through err_put_device so the device reference is dropped properly.

Fixes 89ff3dfac604 ("usb: gadget: f_hid: fix f_hidg lifetime vs cdev")
Cc stable <[email protected]>
---
 drivers/usb/gadget/function/f_hid.c |    3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)
drivers/usb/gadget/function/f_hid.c +1 -2
@@ -1620,7 +1620,7 @@ static struct usb_function *hidg_alloc(s
 	hidg->dev.devt = MKDEV(major, opts->minor);
 	ret = dev_set_name(&hidg->dev, "hidg%d", opts->minor);
 	if (ret)
-		goto err_unlock;
+		goto err_put_device;
 
 	hidg->bInterfaceSubClass = opts->subclass;
 	hidg->bInterfaceProtocol = opts->protocol;
@@ -1657,7 +1657,6 @@ static struct usb_function *hidg_alloc(s
 
 err_put_device:
 	put_device(&hidg->dev);
-err_unlock:
 	mutex_unlock(&opts->lock);
 	return ERR_PTR(ret);
 }
author maintainer
#291 lore raw

PATCH 276/332 usb: gadget: composite: fix integer underflow in WebUSB GET_URL handling

In reply to Greg Kroah-Hartman

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

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

From: Jeremy Erazo <[email protected]>

commit 6c5dbc104dadd79fc2923497c20bae759a18758c upstream.

The WebUSB GET_URL handler in composite_setup() narrows landing_page_length to fit the host-supplied wLength using

landing_page_length = w_length
- WEBUSB_URL_DESCRIPTOR_HEADER_LENGTH + landing_page_offset;

If wLength is smaller than WEBUSB_URL_DESCRIPTOR_HEADER_LENGTH the unsigned subtraction wraps, and the subsequent

	memcpy(url_descriptor->URL,
	       cdev->landing_page + landing_page_offset,
	       landing_page_length - landing_page_offset);

ends up copying close to UINT_MAX bytes from cdev->landing_page into cdev->req->buf. KASAN reports a slab-out-of-bounds in composite_setup on the kmalloc-2k gadget_info allocation, and FORTIFY_SOURCE traps the memcpy as a 4294967293-byte field-spanning write into url_descriptor->URL (size 252).

A USB host can reach this from a single SETUP packet against any gadget that has webusb/use=1 and a landingPage configured.

Handle the small-wLength case before the math: when the host requested fewer bytes than the URL descriptor header, only the header is meaningful and no URL bytes need to be copied. Setting landing_page_length to landing_page_offset makes the existing memcpy a no-op and leaves the descriptor returned to the host unchanged for all larger wLength values.

Fixes 93c473948c58 ("usb: gadget: add WebUSB landing page support")
Cc stable <[email protected]>
---
 drivers/usb/gadget/composite.c |    5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)
drivers/usb/gadget/composite.c +4 -1
@@ -2172,7 +2172,10 @@ unknown:
 				sizeof(url_descriptor->URL)
 				- WEBUSB_URL_DESCRIPTOR_HEADER_LENGTH + landing_page_offset);
 
-			if (w_length < WEBUSB_URL_DESCRIPTOR_HEADER_LENGTH + landing_page_length)
+			if (w_length < WEBUSB_URL_DESCRIPTOR_HEADER_LENGTH)
+				landing_page_length = landing_page_offset;
+			else if (w_length <
+				 WEBUSB_URL_DESCRIPTOR_HEADER_LENGTH + landing_page_length)
 				landing_page_length = w_length
 				- WEBUSB_URL_DESCRIPTOR_HEADER_LENGTH + landing_page_offset;
 
author maintainer
#292 lore raw

PATCH 277/332 usb: gadget: dummy_hcd: Reject hub port requests for non-existent ports

In reply to Greg Kroah-Hartman

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

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

From: Seungjin Bae <[email protected]>

commit 7d9633528dd40e33964d2dc74a5abbf5c4d116ce upstream.

The dummy_hub_control() function handles USB hub class requests to the virtual root hub. The GetPortStatus case returns -EPIPE for requests with wIndex != 1, since the virtual root hub has only a single port. However, the ClearPortFeature and SetPortFeature cases lack the same check.

Fix this by extending the wIndex != 1 rejection to both cases, matching the existing behavior of GetPortStatus.

---
 drivers/usb/gadget/udc/dummy_hcd.c |    4 ++++
 1 file changed, 4 insertions(+)
drivers/usb/gadget/udc/dummy_hcd.c +4 -0
@@ -2134,6 +2134,8 @@ static int dummy_hub_control(
 	case ClearHubFeature:
 		break;
 	case ClearPortFeature:
+		if (wIndex != 1)
+			goto error;
 		switch (wValue) {
 		case USB_PORT_FEAT_SUSPEND:
 			if (hcd->speed == HCD_USB3) {
@@ -2248,6 +2250,8 @@ static int dummy_hub_control(
 		retval = -EPIPE;
 		break;
 	case SetPortFeature:
+		if (wIndex != 1)
+			goto error;
 		switch (wValue) {
 		case USB_PORT_FEAT_LINK_STATE:
 			if (hcd->speed != HCD_USB3) {
author maintainer
#293 lore raw

PATCH 278/332 usb: gadget: f_fs: copy only received bytes on short ep0 read

In reply to Greg Kroah-Hartman

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

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

From: Michael Bommarito <[email protected]>

commit 4e036c10e7f4df5d951c69cc3697bc8e209c6d02 upstream.

ffs_ep0_read() allocates its control-OUT data buffer with kmalloc() (not kzalloc) at the Length value from the Setup packet, then copies that full len to userspace regardless of how many bytes were actually received:

    data = kmalloc(len, GFP_KERNEL);
    ...
    ret = __ffs_ep0_queue_wait(ffs, data, len);
    if ((ret > 0) && (copy_to_user(buf, data, len)))
            ret = -EFAULT;

__ffs_ep0_queue_wait() returns req->actual, which on a short control OUT transfer is strictly less than len. The copy_to_user() call still copies len bytes, so on a short OUT the last (len - ret) bytes of the kmalloc() buffer -- uninitialised slab residue -- are delivered to the FunctionFS daemon.

Short ep0 OUT completions are specified USB control-transfer behavior and are produced by in-tree UDCs:

* dwc2 continues on req->actual < req->length for ep0 DATA OUT (short-not-ok is the only ep0-OUT stall path).
* aspeed_udc ends ep0 OUT on rx_len < ep->ep.maxpacket.
* renesas_usbf logs "ep0 short packet" and completes the request.
* dwc3 stalls on short IN but not on short OUT.

A short ep0 OUT is therefore not evidence of a broken UDC; it is a normal condition f_fs has to cope with. The sibling gadgetfs implementation in drivers/usb/gadget/legacy/inode.c already does this correctly via min(len, dev->req->actual) before copy_to_user(). This patch brings f_fs.c to the same safe pattern rather than trimming at a defensive layer.

The bug is reached from the FunctionFS device node, which in real deployments is owned by the privileged gadget daemon (adbd, UMS, composite gadget services, etc.); it is not reachable from unprivileged userspace. Linux host stacks normally reject short-wLength control OUTs before they reach the gadget, so reproducing this required a build that bypasses that host-side check. With the bypass in place, a 1-byte payload on a 64-byte Setup produces 63 bytes of non-canary slab residue in the daemon's read buffer.

Fix by copying only ret (actually received) bytes to userspace.

Fixes ddf8abd25994 ("USB: f_fs: the FunctionFS driver")
Cc stable <[email protected]>
Assisted-by Claude:claude-opus-4-7
---
 drivers/usb/gadget/function/f_fs.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/usb/gadget/function/f_fs.c +1 -1
@@ -619,7 +619,7 @@ static ssize_t ffs_ep0_read(struct file
 
 		/* unlocks spinlock */
 		ret = __ffs_ep0_queue_wait(ffs, data, len);
-		if ((ret > 0) && (copy_to_user(buf, data, len)))
+		if ((ret > 0) && (copy_to_user(buf, data, ret)))
 			ret = -EFAULT;
 		goto done_mutex;
 
author maintainer
#294 lore raw

PATCH 279/332 usb: gadget: f_fs: serialize DMABUF cancel against request completion

In reply to Greg Kroah-Hartman

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

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

From: Michael Bommarito <[email protected]>

commit 2796646f6d892c1eb6818c7ca41fdfa12568e8d1 upstream.

ffs_epfile_dmabuf_io_complete() calls usb_ep_free_request() on the completed request but leaves priv->req, the back-pointer that ffs_dmabuf_transfer() set on submission, pointing at the freed memory. A later FUNCTIONFS_DMABUF_DETACH ioctl or ffs_epfile_release() on the close path still sees priv->req non-NULL under ffs->eps_lock:

    if (priv->ep && priv->req)
            usb_ep_dequeue(priv->ep, priv->req);

so usb_ep_dequeue() is called on a freed usb_request.

On dummy_hcd the dequeue path only walks a live queue and pointer-compares, so the freed pointer reads without faulting and KASAN requires an explicit check at the FunctionFS call site to surface the use-after-free. On SG-capable in-tree UDCs the dequeue path dereferences the supplied request immediately:

* chipidea's ep_dequeue() does
container_of(req, struct ci_hw_req, req) and reads hwreq->req.status before acquiring its own lock.
* cdnsp's cdnsp_gadget_ep_dequeue() reads request->status first.

The narrower option of clearing priv->req via cmpxchg() in the completion does not close the race: the completion runs without eps_lock, so a cancel path holding eps_lock can still observe priv->req non-NULL, race a concurrent completion that clears and frees, and pass the freed pointer to usb_ep_dequeue(). A slightly longer fix that moves the free into the cleanup work is needed.

Same class of lifetime race as the recent usbip-vudc timer fix [1].

Take eps_lock in the sole place that mutates priv->req from the callback direction by moving usb_ep_free_request() out of the completion into ffs_dmabuf_cleanup(), the existing work handler scheduled by ffs_dmabuf_signal_done() on ffs->io_completion_wq. Clear priv->req there under eps_lock before freeing, and only clear if priv->req still names our request (a subsequent ffs_dmabuf_transfer() on the same attachment may have queued a new one).

This keeps the existing dummy_hcd sync-dequeue invariant: the completion callback is still invoked by the UDC without eps_lock held (dummy_hcd drops its own lock before calling the callback), and the callback now takes no f_fs lock at all. Serialization against the cancel path happens in cleanup, which runs from the workqueue with no f_fs lock held on entry.

The priv ref count protects the containing ffs_dmabuf_priv:
ffs_dmabuf_transfer() takes a ref via ffs_dmabuf_get(), cleanup drops it via ffs_dmabuf_put(), so priv stays live for the cleanup even after the cancel path's list_del + ffs_dmabuf_put.

The ffs_dmabuf_transfer() error path no longer frees usb_req
inline: fence->req and fence->ep are set before usb_ep_queue(), so ffs_dmabuf_cleanup() (scheduled by the error-path ffs_dmabuf_signal_done()) owns the free regardless of whether the queue succeeded.

Reproduced under KASAN on both detach and close paths against dummy_hcd with an observability hook (kasan_check_byte(priv->req) immediately before usb_ep_dequeue) at the two FunctionFS cancel sites to surface the stale-pointer access; the hook is not part of this patch. The KASAN allocator / free stacks in the captured splats identify the same request: alloc in dummy_alloc_request, free in dummy_timer, fault reached from ffs_epfile_release (close) and from the FUNCTIONFS_DMABUF_DETACH ioctl (detach). With the patch applied, both paths are silent under the same hook.

The bug is reached from the FunctionFS device node, which in real deployments is owned by the privileged gadget daemon (adbd, UMS, composite gadget services, etc.); it is not reachable from unprivileged userspace or from a USB host on the cable. FunctionFS mounts default to GLOBAL_ROOT_UID, but the filesystem supports uid=, gid=, and fmode= delegation to a non-root gadget daemon, so on real deployments the attacker may be a less-privileged service rather than root.

Fixes 7b07a2a7ca02 ("usb: gadget: functionfs: Add DMABUF import interface")
Cc stable <[email protected]>
Assisted-by Claude:claude-opus-4-7
---
 drivers/usb/gadget/function/f_fs.c |   24 ++++++++++++++++++++++--
 1 file changed, 22 insertions(+), 2 deletions(-)
drivers/usb/gadget/function/f_fs.c +22 -2
@@ -150,6 +150,8 @@ struct ffs_dma_fence {
 	struct dma_fence base;
 	struct ffs_dmabuf_priv *priv;
 	struct work_struct work;
+	struct usb_ep *ep;
+	struct usb_request *req;
 };
 
 struct ffs_epfile {
@@ -1385,6 +1387,21 @@ static void ffs_dmabuf_cleanup(struct wo
 	struct ffs_dmabuf_priv *priv = dma_fence->priv;
 	struct dma_buf_attachment *attach = priv->attach;
 	struct dma_fence *fence = &dma_fence->base;
+	struct usb_request *req = dma_fence->req;
+	struct usb_ep *ep = dma_fence->ep;
+
+	/*
+	 * eps_lock pairs with the cancel paths so they cannot pass a freed
+	 * req to usb_ep_dequeue().  Only clear if priv->req still names ours;
+	 * a re-queue on the same attachment may have taken that slot.
+	 */
+	spin_lock_irq(&priv->ffs->eps_lock);
+	if (priv->req == req)
+		priv->req = NULL;
+	spin_unlock_irq(&priv->ffs->eps_lock);
+
+	if (ep && req)
+		usb_ep_free_request(ep, req);
 
 	ffs_dmabuf_put(attach);
 	dma_fence_put(fence);
@@ -1414,8 +1431,8 @@ static void ffs_epfile_dmabuf_io_complet
 					  struct usb_request *req)
 {
 	pr_vdebug("FFS: DMABUF transfer complete, status=%d\n", req->status);
+	/* req is freed by ffs_dmabuf_cleanup() under eps_lock. */
 	ffs_dmabuf_signal_done(req->context, req->status);
-	usb_ep_free_request(ep, req);
 }
 
 static const char *ffs_dmabuf_get_driver_name(struct dma_fence *fence)
@@ -1699,6 +1716,10 @@ static int ffs_dmabuf_transfer(struct fi
 	usb_req->context  = fence;
 	usb_req->complete = ffs_epfile_dmabuf_io_complete;
 
+	/* ffs_dmabuf_cleanup() frees usb_req via these two fields. */
+	fence->req = usb_req;
+	fence->ep = ep->ep;
+
 	cookie = dma_fence_begin_signalling();
 	ret = usb_ep_queue(ep->ep, usb_req, GFP_ATOMIC);
 	dma_fence_end_signalling(cookie);
@@ -1708,7 +1729,6 @@ static int ffs_dmabuf_transfer(struct fi
 	} else {
 		pr_warn("FFS: Failed to queue DMABUF: %d\n", ret);
 		ffs_dmabuf_signal_done(fence, ret);
-		usb_ep_free_request(ep->ep, usb_req);
 	}
 
 	spin_unlock_irq(&epfile->ffs->eps_lock);
author maintainer
#295 lore raw

PATCH 280/332 thunderbolt: property: Reject u32 wrap in tb_property_entry_valid()

In reply to Greg Kroah-Hartman

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

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

From: Michael Bommarito <[email protected]>

commit 01deda0152066c6c955f0619114ea6afa070aaec upstream.

entry->value is u32 and entry->length is u16; the sum is performed in u32 and wraps. A malicious XDomain peer can pick
value = 0xffffff00, length = 0x100 so the sum 0x100000000 wraps to 0 and passes the > block_len check. tb_property_parse() then passes entry->value to parse_dwdata() as a dword offset into the property block, reading attacker-directed memory far past the allocation.

For TEXT-typed entries with the "deviceid" or "vendorid" keys this lands in xd->device_name / xd->vendor_name and is readable back via the per-XDomain device_name / vendor_name sysfs attributes; the leak is NUL-bounded (kstrdup() stops at the first zero byte) and untargeted (the attacker picks a delta, not an absolute address). DATA-typed entries are parsed into property->value.data but not generically surfaced to userspace.

Use check_add_overflow() so a wrapped sum is rejected.

Fixes cdae7c07e3e3 ("thunderbolt: Add support for XDomain properties")
Assisted-by Claude:claude-opus-4-6
Assisted-by Codex:gpt-5-4
---
 drivers/thunderbolt/property.c |    6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)
drivers/thunderbolt/property.c +5 -1
@@ -8,6 +8,7 @@
  */
 
 #include <linux/err.h>
+#include <linux/overflow.h>
 #include <linux/slab.h>
 #include <linux/string.h>
 #include <linux/uuid.h>
@@ -52,13 +53,16 @@ static inline void format_dwdata(void *d
 static bool tb_property_entry_valid(const struct tb_property_entry *entry,
 				  size_t block_len)
 {
+	u32 end;
+
 	switch (entry->type) {
 	case TB_PROPERTY_TYPE_DIRECTORY:
 	case TB_PROPERTY_TYPE_DATA:
 	case TB_PROPERTY_TYPE_TEXT:
 		if (entry->length > block_len)
 			return false;
-		if (entry->value + entry->length > block_len)
+		if (check_add_overflow(entry->value, entry->length, &end) ||
+		    end > block_len)
 			return false;
 		break;
 
author maintainer
#296 lore raw

PATCH 281/332 thunderbolt: property: Reject dir_len < 4 to prevent size_t underflow

In reply to Greg Kroah-Hartman

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

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

From: Michael Bommarito <[email protected]>

commit de21b59c29e31c5108ddc04210631bbfab81b997 upstream.

On the non-root path, __tb_property_parse_dir() takes dir_len from entry->length (u16 widened to size_t). Two distinct OOB conditions follow when entry->length < 4:

1. The non-root path begins with kmemdup(&block[dir_offset], sizeof(*dir->uuid), ...) which always reads 4 dwords from dir_offset. tb_property_entry_valid() only enforces dir_offset + entry->length <= block_len, so a crafted entry with dir_offset close to the end of the property block and entry->length in 0..3 passes that gate but lets the UUID copy run off the block (e.g. dir_offset = 497, dir_len = 3 in a 500-dword block reads block[497..501]).

2. After the kmemdup, content_len = dir_len - 4 underflows size_t to ~SIZE_MAX, nentries becomes SIZE_MAX / 4, and the entry walk runs OOB on each iteration until an entry fails validation or the kernel oopses on an unmapped page.

Reject dir_len < 4 on the non-root path *before* the UUID kmemdup, which closes both holes.

Also move INIT_LIST_HEAD(&dir->properties) up to immediately after the dir allocation so the new error-return path (and the existing uuid-alloc failure path) calling tb_property_free_dir() sees a walkable list rather than the zero-initialized NULL next/prev that list_for_each_entry_safe() would oops on.

Fixes cdae7c07e3e3 ("thunderbolt: Add support for XDomain properties")
Assisted-by Claude:claude-opus-4-6
Assisted-by Codex:gpt-5-4
---
 drivers/thunderbolt/property.c |    8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)
drivers/thunderbolt/property.c +6 -2
@@ -174,10 +174,16 @@ static struct tb_property_dir *__tb_prop
 	if (!dir)
 		return NULL;
 
+	INIT_LIST_HEAD(&dir->properties);
+
 	if (is_root) {
 		content_offset = dir_offset + 2;
 		content_len = dir_len;
 	} else {
+		if (dir_len < 4) {
+			tb_property_free_dir(dir);
+			return NULL;
+		}
 		dir->uuid = kmemdup(&block[dir_offset], sizeof(*dir->uuid),
 				    GFP_KERNEL);
 		if (!dir->uuid) {
@@ -191,8 +197,6 @@ static struct tb_property_dir *__tb_prop
 	entries = (const struct tb_property_entry *)&block[content_offset];
 	nentries = content_len / (sizeof(*entries) / 4);
 
-	INIT_LIST_HEAD(&dir->properties);
-
 	for (i = 0; i < nentries; i++) {
 		struct tb_property *property;
 
author maintainer
#297 lore raw

PATCH 282/332 thunderbolt: property: Cap recursion depth in __tb_property_parse_dir()

In reply to Greg Kroah-Hartman

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

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

From: Michael Bommarito <[email protected]>

commit 928abe19fbf0127003abcb1ea69cabc1c897d0ab upstream.

A DIRECTORY entry's value field is used as the dir_offset for a recursive call into __tb_property_parse_dir() with no depth counter. A crafted peer that chains DIRECTORY entries into a back-reference loop drives the parser until the kernel stack is exhausted and the guard page fires. Any untrusted XDomain peer (cable, dock, in-line inspector, adjacent host) that reaches the PROPERTIES_REQUEST control-plane exchange can trigger this without authentication.

Thread a depth counter through tb_property_parse() and __tb_property_parse_dir(), and reject blocks that exceed TB_PROPERTY_MAX_DEPTH = 8. That is comfortably larger than any observed legitimate XDomain layout.

Operators who do not need XDomain host-to-host discovery can disable the path entirely with thunderbolt.xdomain=0 on the kernel command line.

Fixes cdae7c07e3e3 ("thunderbolt: Add support for XDomain properties")
Assisted-by Claude:claude-opus-4-6
Assisted-by Codex:gpt-5-4
---
 drivers/thunderbolt/property.c |   18 ++++++++++++------
 1 file changed, 12 insertions(+), 6 deletions(-)
drivers/thunderbolt/property.c +12 -6
@@ -35,10 +35,11 @@ struct tb_property_dir_entry {
 };
 
 #define TB_PROPERTY_ROOTDIR_MAGIC	0x55584401
+#define TB_PROPERTY_MAX_DEPTH		8
 
 static struct tb_property_dir *__tb_property_parse_dir(const u32 *block,
 	size_t block_len, unsigned int dir_offset, size_t dir_len,
-	bool is_root);
+	bool is_root, unsigned int depth);
 
 static inline void parse_dwdata(void *dst, const void *src, size_t dwords)
 {
@@ -97,7 +98,8 @@ tb_property_alloc(const char *key, enum
 }
 
 static struct tb_property *tb_property_parse(const u32 *block, size_t block_len,
-					const struct tb_property_entry *entry)
+					const struct tb_property_entry *entry,
+					unsigned int depth)
 {
 	char key[TB_PROPERTY_KEY_SIZE + 1];
 	struct tb_property *property;
@@ -118,7 +120,7 @@ static struct tb_property *tb_property_p
 	switch (property->type) {
 	case TB_PROPERTY_TYPE_DIRECTORY:
 		dir = __tb_property_parse_dir(block, block_len, entry->value,
-					      entry->length, false);
+					      entry->length, false, depth + 1);
 		if (!dir) {
 			kfree(property);
 			return NULL;
@@ -163,13 +165,17 @@ static struct tb_property *tb_property_p
 }
 
 static struct tb_property_dir *__tb_property_parse_dir(const u32 *block,
-	size_t block_len, unsigned int dir_offset, size_t dir_len, bool is_root)
+	size_t block_len, unsigned int dir_offset, size_t dir_len, bool is_root,
+	unsigned int depth)
 {
 	const struct tb_property_entry *entries;
 	size_t i, content_len, nentries;
 	unsigned int content_offset;
 	struct tb_property_dir *dir;
 
+	if (depth > TB_PROPERTY_MAX_DEPTH)
+		return NULL;
+
 	dir = kzalloc_obj(*dir);
 	if (!dir)
 		return NULL;
@@ -200,7 +206,7 @@ static struct tb_property_dir *__tb_prop
 	for (i = 0; i < nentries; i++) {
 		struct tb_property *property;
 
-		property = tb_property_parse(block, block_len, &entries[i]);
+		property = tb_property_parse(block, block_len, &entries[i], depth);
 		if (!property) {
 			tb_property_free_dir(dir);
 			return NULL;
@@ -239,7 +245,7 @@ struct tb_property_dir *tb_property_pars
 		return NULL;
 
 	return __tb_property_parse_dir(block, block_len, 0, rootdir->length,
-				       true);
+				       true, 0);
 }
 
 /**
author maintainer
#298 lore raw

PATCH 283/332 scsi: fcoe: Reject FIP descriptors with zero fip_dlen in CVL walker

In reply to Greg Kroah-Hartman

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

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

From: Michael Bommarito <[email protected]>

commit 9eed1bd59937e6828b00d2f2dfef631d964f3636 upstream.

drivers/scsi/fcoe/fcoe_ctlr.c::fcoe_ctlr_recv_clr_vlink() advanced the descriptor cursor by an attacker-supplied fip_dlen without ever requiring dlen >= sizeof(struct fip_desc) in the default branch. The named descriptor cases (FIP_DT_MAC, FIP_DT_NAME, FIP_DT_VN_ID) checked their per-type minimum lengths, but a FIP_DT_NON_CRITICAL descriptor (fip_dtype >= 128, which the standard requires receivers to silently ignore) skipped that check entirely.

An unauthenticated L2 peer on the FCoE control VLAN could hang fcoe_ctlr_recv_work on an fcoe, qedf, or bnx2fc initiator indefinitely by emitting one FIP CVL frame whose single descriptor had fip_dtype == FIP_DT_NON_CRITICAL and fip_dlen == 0: the cursor advanced zero bytes per iteration and the loop condition rlen >= sizeof(*desc) stayed true forever, blocking every subsequent FIP frame on that controller.

Tighten the outer dlen guard to also reject dlen < sizeof(struct fip_desc), so a malformed descriptor whose length cannot even cover the descriptor header is rejected before the switch. This is the same lower-bound the named cases already apply and is the minimum scope that closes the loop.

Fixes 97c8389d54b9 ("[SCSI] fcoe, libfcoe: Add support for FIP. FCoE discovery and keep-alive.")
Assisted-by Claude:claude-opus-4-7
---
 drivers/scsi/fcoe/fcoe_ctlr.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/scsi/fcoe/fcoe_ctlr.c +1 -1
@@ -1385,7 +1385,7 @@ static void fcoe_ctlr_recv_clr_vlink(str
 
 	while (rlen >= sizeof(*desc)) {
 		dlen = desc->fip_dlen * FIP_BPW;
-		if (dlen > rlen)
+		if (dlen < sizeof(*desc) || dlen > rlen)
 			goto err;
 		/* Drop CVL if there are duplicate critical descriptors */
 		if ((desc->fip_dtype < 32) &&
author maintainer
#299 lore raw

PATCH 284/332 scsi: scsi_transport_fc: Widen FPIN pname walker counter to u32

In reply to Greg Kroah-Hartman

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

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

From: Michael Bommarito <[email protected]>

commit a9a39233ec1fc9f97ea1340a4d09bb7ec2be5153 upstream.

An adjacent Fibre Channel fabric actor that can deliver an FPIN ELS frame to an lpfc or qla2xxx Linux initiator can trigger a non-return in the generic FC transport. This is not a local userspace or IP network path; the attacker must be able to inject fabric traffic, for example as a compromised switch or fabric controller, or as a same-zone N_Port on a fabric that permits source spoofing.

The Link-Integrity and Peer-Congestion FPIN walkers used a u8 loop counter against the 32-bit on-wire pname_count field, and did not bound pname_count by the descriptor body already validated by the TLV walker. A pname_count of 256 therefore wraps the counter and keeps the loop condition true indefinitely.

Factor the shared pname_list[] walk into one helper, widen the counter to u32, and clamp pname_count against the entries that fit in the descriptor body before iterating.

Fixes 3dcfe0de5a97 ("scsi: fc: Parse FPIN packets and update statistics")
Assisted-by Claude:claude-opus-4-7
---
 drivers/scsi/scsi_transport_fc.c |   77 ++++++++++++++++++++-------------------
 1 file changed, 41 insertions(+), 36 deletions(-)
drivers/scsi/scsi_transport_fc.c +41 -36
@@ -737,6 +737,37 @@ fc_cn_stats_update(u16 event_type, struc
 	}
 }
 
+static void
+fc_fpin_pname_stats_update(struct Scsi_Host *shost,
+			   struct fc_rport *attach_rport, u16 event_type,
+			   u32 desc_len, u32 fixed_len, u32 pname_count,
+			   __be64 *pname_list,
+			   void (*stats_update)(u16 event_type,
+						struct fc_fpin_stats *stats))
+{
+	u32 i;
+	struct fc_rport *rport;
+	u64 wwpn;
+
+	if (desc_len < fixed_len)
+		pname_count = 0;
+	else
+		pname_count = min(pname_count, (desc_len - fixed_len) /
+				   sizeof(pname_list[0]));
+
+	for (i = 0; i < pname_count; i++) {
+		wwpn = be64_to_cpu(pname_list[i]);
+		rport = fc_find_rport_by_wwpn(shost, wwpn);
+		if (rport &&
+		    (rport->roles & FC_PORT_ROLE_FCP_TARGET ||
+		     rport->roles & FC_PORT_ROLE_NVME_TARGET)) {
+			if (rport == attach_rport)
+				continue;
+			stats_update(event_type, &rport->fpin_stats);
+		}
+	}
+}
+
 /*
  * fc_fpin_li_stats_update - routine to update Link Integrity
  * event statistics.
@@ -747,13 +778,11 @@ fc_cn_stats_update(u16 event_type, struc
 static void
 fc_fpin_li_stats_update(struct Scsi_Host *shost, struct fc_tlv_desc *tlv)
 {
-	u8 i;
 	struct fc_rport *rport = NULL;
 	struct fc_rport *attach_rport = NULL;
 	struct fc_host_attrs *fc_host = shost_to_fc_host(shost);
 	struct fc_fn_li_desc *li_desc = (struct fc_fn_li_desc *)tlv;
 	u16 event_type = be16_to_cpu(li_desc->event_type);
-	u64 wwpn;
 
 	rport = fc_find_rport_by_wwpn(shost,
 				      be64_to_cpu(li_desc->attached_wwpn));
@@ -764,22 +793,11 @@ fc_fpin_li_stats_update(struct Scsi_Host
 		fc_li_stats_update(event_type, &attach_rport->fpin_stats);
 	}
 
-	if (be32_to_cpu(li_desc->pname_count) > 0) {
-		for (i = 0;
-		    i < be32_to_cpu(li_desc->pname_count);
-		    i++) {
-			wwpn = be64_to_cpu(li_desc->pname_list[i]);
-			rport = fc_find_rport_by_wwpn(shost, wwpn);
-			if (rport &&
-			    (rport->roles & FC_PORT_ROLE_FCP_TARGET ||
-			    rport->roles & FC_PORT_ROLE_NVME_TARGET)) {
-				if (rport == attach_rport)
-					continue;
-				fc_li_stats_update(event_type,
-						   &rport->fpin_stats);
-			}
-		}
-	}
+	fc_fpin_pname_stats_update(shost, attach_rport, event_type,
+				   be32_to_cpu(li_desc->desc_len),
+				   FC_TLV_DESC_LENGTH_FROM_SZ(*li_desc),
+				   be32_to_cpu(li_desc->pname_count),
+				   li_desc->pname_list, fc_li_stats_update);
 
 	if (fc_host->port_name == be64_to_cpu(li_desc->attached_wwpn))
 		fc_li_stats_update(event_type, &fc_host->fpin_stats);
@@ -827,13 +845,11 @@ static void
 fc_fpin_peer_congn_stats_update(struct Scsi_Host *shost,
 				struct fc_tlv_desc *tlv)
 {
-	u8 i;
 	struct fc_rport *rport = NULL;
 	struct fc_rport *attach_rport = NULL;
 	struct fc_fn_peer_congn_desc *pc_desc =
 	    (struct fc_fn_peer_congn_desc *)tlv;
 	u16 event_type = be16_to_cpu(pc_desc->event_type);
-	u64 wwpn;
 
 	rport = fc_find_rport_by_wwpn(shost,
 				      be64_to_cpu(pc_desc->attached_wwpn));
@@ -844,22 +860,11 @@ fc_fpin_peer_congn_stats_update(struct S
 		fc_cn_stats_update(event_type, &attach_rport->fpin_stats);
 	}
 
-	if (be32_to_cpu(pc_desc->pname_count) > 0) {
-		for (i = 0;
-		    i < be32_to_cpu(pc_desc->pname_count);
-		    i++) {
-			wwpn = be64_to_cpu(pc_desc->pname_list[i]);
-			rport = fc_find_rport_by_wwpn(shost, wwpn);
-			if (rport &&
-			    (rport->roles & FC_PORT_ROLE_FCP_TARGET ||
-			     rport->roles & FC_PORT_ROLE_NVME_TARGET)) {
-				if (rport == attach_rport)
-					continue;
-				fc_cn_stats_update(event_type,
-						   &rport->fpin_stats);
-			}
-		}
-	}
+	fc_fpin_pname_stats_update(shost, attach_rport, event_type,
+				   be32_to_cpu(pc_desc->desc_len),
+				   FC_TLV_DESC_LENGTH_FROM_SZ(*pc_desc),
+				   be32_to_cpu(pc_desc->pname_count),
+				   pc_desc->pname_list, fc_cn_stats_update);
 }
 
 /*
author maintainer
#300 lore raw

PATCH 285/332 scsi: target: iscsi: Fix CRC overread and double-free in iscsit_handle_text_cmd()

In reply to Greg Kroah-Hartman

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

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

From: Michael Bommarito <[email protected]>

commit 778c2ab142c625a8a8afa570e0f9b7873f445d99 upstream.

Two latent bugs in the Text-phase handler, both present since the original LIO integration in commit e48354ce078c ("iscsi-target: Add iSCSI fabric support for target v4.1"):

1) DataDigest CRC buffer overread (4 bytes past text_in).

text_in is kzalloc()'d at ALIGN(payload_length, 4). rx_size is then incremented by ISCSI_CRC_LEN to make room for the received DataDigest in the iovec, but the same (now-bumped) rx_size is passed as the buffer length to iscsit_crc_buf():

       if (conn->conn_ops->DataDigest) {
               ...
               rx_size += ISCSI_CRC_LEN;
       }
       ...
       if (conn->conn_ops->DataDigest) {
               data_crc = iscsit_crc_buf(text_in, rx_size, 0, NULL);

iscsit_crc_buf() walks rx_size bytes of text_in with crc32c(), so when DataDigest is negotiated it reads 4 bytes past the end of the text_in allocation. KASAN reproduces this directly on the unpatched mainline tree as slab-out-of-bounds in crc32c() called from the Text PDU path. The OOB bytes feed crc32c() and are then compared against the initiator-supplied checksum, so the value does not flow back to the attacker, but the kernel does read past the buffer on every Text PDU with DataDigest=CRC32C.

Fix by passing the actual padded payload length (ALIGN(payload_length, 4)) that was used for the kzalloc().

2) Stale cmd->text_in_ptr re-free (double-free) on ERL>0 bad DataDigest drop.

On DataDigest mismatch with ErrorRecoveryLevel > 0 the handler silently drops the PDU and lets the initiator plug the CmdSN gap:

               kfree(text_in);
               return 0;

cmd->text_in_ptr still points at the freed buffer. The next Text Request on the same ITT re-enters iscsit_setup_text_cmd(), which unconditionally does

       kfree(cmd->text_in_ptr);
       cmd->text_in_ptr = NULL;

freeing the same pointer a second time. Session teardown via iscsit_release_cmd() has the same shape and hits the same double-free if the connection is dropped before a second Text Request arrives.

On an unmodified mainline tree the bug-1 CRC overread fires first on the initial valid Text Request and perturbs the subsequent state, so #4 was isolated by building a kernel with only the bug-1 hunk of this patch applied plus temporary printk() observability around the three relevant kfree() sites. The observability prints are not part of this patch. On that build, a three-PDU Text Request sequence after login produces two back-to-back splats:

       BUG: KASAN: double-free in iscsit_setup_text_cmd+0x??
       BUG: KASAN: double-free in iscsit_release_cmd+0x??

showing the same pointer freed in the ERL>0 drop path and again in iscsit_setup_text_cmd() (next Text Request on the same ITT) and once more in iscsit_release_cmd() (session teardown). On distro kernels with CONFIG_SLAB_FREELIST_HARDENED=y (default) the double-free becomes a remote kernel BUG(); on non-hardened kernels it corrupts the slab freelist.

Fix by clearing cmd->text_in_ptr after the kfree() in the ERL>0 drop path. With both hunks applied #4 is directly observable on the stock tree without observability printks; fixing bug-1 alone would mask #4 less, not more, so the hunks are submitted together.

Both fixes are one-liners. The Text PDU state machine is unchanged and the wire protocol is unaffected.

Fixes e48354ce078c ("iscsi-target: Add iSCSI fabric support for target v4.1")
Assisted-by Claude:claude-opus-4-7
---
 drivers/target/iscsi/iscsi_target.c |    5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)
drivers/target/iscsi/iscsi_target.c +4 -1
@@ -2281,7 +2281,9 @@ iscsit_handle_text_cmd(struct iscsit_con
 			goto reject;
 
 		if (conn->conn_ops->DataDigest) {
-			data_crc = iscsit_crc_buf(text_in, rx_size, 0, NULL);
+			data_crc = iscsit_crc_buf(text_in,
+						  ALIGN(payload_length, 4),
+						  0, NULL);
 			if (checksum != data_crc) {
 				pr_err("Text data CRC32C DataDigest"
 					" 0x%08x does not match computed"
@@ -2300,6 +2302,7 @@ iscsit_handle_text_cmd(struct iscsit_con
 					" Command CmdSN: 0x%08x due to"
 					" DataCRC error.\n", hdr->cmdsn);
 					kfree(text_in);
+					cmd->text_in_ptr = NULL;
 					return 0;
 				}
 			} else {