In the Linux kernel, the following vulnerability has been resolved:
drm/i915/ttm: fix CCS handling
Crucible + recent Mesa seems to sometimes hit:
GEM_BUG_ON(num_ccs_blks > NUM_CCS_BLKS_PER_XFER)
And it looks like we can also trigger this with gem_lmem_swapping, if we
modify the test to use slightly larger object sizes.
Looking closer it looks like we have the following issues in
migrate_copy():
- We are using plain integer in various places, which we can easily
overflow with a large object.
- We pass the entire object size (when the src is lmem) into
emit_pte() and then try to copy it, which doesn't work, since we
only have a few fixed sized windows in which to map the pages and
perform the copy. With an object > 8M we therefore aren't properly
copying the pages. And then with an object > 64M we trigger the
GEM_BUG_ON(num_ccs_blks > NUM_CCS_BLKS_PER_XFER).
So it looks like our copy handling for any object > 8M (which is our
CHUNK_SZ) is currently broken on DG2.
Testcase: igt@gem_lmem_swapping
(cherry picked from commit 8676145eb2f53a9940ff70910caf0125bd8a4bc2)
In the Linux kernel, the following vulnerability has been resolved:
xhci: Fix null pointer dereference in remove if xHC has only one roothub
The remove path in xhci platform driver tries to remove and put both main
and shared hcds even if only a main hcd exists (one roothub)
This causes a null pointer dereference in reboot for those controllers.
Check that the shared_hcd exists before trying to remove it.
In the Linux kernel, the following vulnerability has been resolved:
bpf: Do mark_chain_precision for ARG_CONST_ALLOC_SIZE_OR_ZERO
Precision markers need to be propagated whenever we have an ARG_CONST_*
style argument, as the verifier cannot consider imprecise scalars to be
equivalent for the purposes of states_equal check when such arguments
refine the return value (in this case, set mem_size for PTR_TO_MEM). The
resultant mem_size for the R0 is derived from the constant value, and if
the verifier incorrectly prunes states considering them equivalent where
such arguments exist (by seeing that both registers have reg->precise as
false in regsafe), we can end up with invalid programs passing the
verifier which can do access beyond what should have been the correct
mem_size in that explored state.
To show a concrete example of the problem:
0000000000000000 <prog>:
0: r2 = *(u32 *)(r1 + 80)
1: r1 = *(u32 *)(r1 + 76)
2: r3 = r1
3: r3 += 4
4: if r3 > r2 goto +18 <LBB5_5>
5: w2 = 0
6: *(u32 *)(r1 + 0) = r2
7: r1 = *(u32 *)(r1 + 0)
8: r2 = 1
9: if w1 == 0 goto +1 <LBB5_3>
10: r2 = -1
0000000000000058 <LBB5_3>:
11: r1 = 0 ll
13: r3 = 0
14: call bpf_ringbuf_reserve
15: if r0 == 0 goto +7 <LBB5_5>
16: r1 = r0
17: r1 += 16777215
18: w2 = 0
19: *(u8 *)(r1 + 0) = r2
20: r1 = r0
21: r2 = 0
22: call bpf_ringbuf_submit
00000000000000b8 <LBB5_5>:
23: w0 = 0
24: exit
For the first case, the single line execution's exploration will prune
the search at insn 14 for the branch insn 9's second leg as it will be
verified first using r2 = -1 (UINT_MAX), while as w1 at insn 9 will
always be 0 so at runtime we don't get error for being greater than
UINT_MAX/4 from bpf_ringbuf_reserve. The verifier during regsafe just
sees reg->precise as false for both r2 registers in both states, hence
considers them equal for purposes of states_equal.
If we propagated precise markers using the backtracking support, we
would use the precise marking to then ensure that old r2 (UINT_MAX) was
within the new r2 (1) and this would never be true, so the verification
would rightfully fail.
The end result is that the out of bounds access at instruction 19 would
be permitted without this fix.
Note that reg->precise is always set to true when user does not have
CAP_BPF (or when subprog count is greater than 1 (i.e. use of any static
or global functions)), hence this is only a problem when precision marks
need to be explicitly propagated (i.e. privileged users with CAP_BPF).
A simplified test case has been included in the next patch to prevent
future regressions.
In the Linux kernel, the following vulnerability has been resolved:
net/sched: fix netdevice reference leaks in attach_default_qdiscs()
In attach_default_qdiscs(), if a dev has multiple queues and queue 0 fails
to attach qdisc because there is no memory in attach_one_default_qdisc().
Then dev->qdisc will be noop_qdisc by default. But the other queues may be
able to successfully attach to default qdisc.
In this case, the fallback to noqueue process will be triggered. If the
original attached qdisc is not released and a new one is directly
attached, this will cause netdevice reference leaks.
The following is the bug log:
veth0: default qdisc (fq_codel) fail, fallback to noqueue
unregister_netdevice: waiting for veth0 to become free. Usage count = 32
leaked reference.
qdisc_alloc+0x12e/0x210
qdisc_create_dflt+0x62/0x140
attach_one_default_qdisc.constprop.41+0x44/0x70
dev_activate+0x128/0x290
__dev_open+0x12a/0x190
__dev_change_flags+0x1a2/0x1f0
dev_change_flags+0x23/0x60
do_setlink+0x332/0x1150
__rtnl_newlink+0x52f/0x8e0
rtnl_newlink+0x43/0x70
rtnetlink_rcv_msg+0x140/0x3b0
netlink_rcv_skb+0x50/0x100
netlink_unicast+0x1bb/0x290
netlink_sendmsg+0x37c/0x4e0
sock_sendmsg+0x5f/0x70
____sys_sendmsg+0x208/0x280
Fix this bug by clearing any non-noop qdiscs that may have been assigned
before trying to re-attach.
In the Linux kernel, the following vulnerability has been resolved:
kcm: fix strp_init() order and cleanup
strp_init() is called just a few lines above this csk->sk_user_data
check, it also initializes strp->work etc., therefore, it is
unnecessary to call strp_done() to cancel the freshly initialized
work.
And if sk_user_data is already used by KCM, psock->strp should not be
touched, particularly strp->work state, so we need to move strp_init()
after the csk->sk_user_data check.
This also makes a lockdep warning reported by syzbot go away.
In the Linux kernel, the following vulnerability has been resolved:
staging: rtl8712: fix use after free bugs
_Read/Write_MACREG callbacks are NULL so the read/write_macreg_hdl()
functions don't do anything except free the "pcmd" pointer. It
results in a use after free. Delete them.
In the Linux kernel, the following vulnerability has been resolved:
powerpc/rtas: Fix RTAS MSR[HV] handling for Cell
The semi-recent changes to MSR handling when entering RTAS (firmware)
cause crashes on IBM Cell machines. An example trace:
kernel tried to execute user page (2fff01a8) - exploit attempt? (uid: 0)
BUG: Unable to handle kernel instruction fetch
Faulting instruction address: 0x2fff01a8
Oops: Kernel access of bad area, sig: 11 [#1]
BE PAGE_SIZE=64K MMU=Hash SMP NR_CPUS=4 NUMA Cell
Modules linked in:
CPU: 0 PID: 0 Comm: swapper/0 Tainted: G W 6.0.0-rc2-00433-gede0a8d3307a #207
NIP: 000000002fff01a8 LR: 0000000000032608 CTR: 0000000000000000
REGS: c0000000015236b0 TRAP: 0400 Tainted: G W (6.0.0-rc2-00433-gede0a8d3307a)
MSR: 0000000008001002 <ME,RI> CR: 00000000 XER: 20000000
...
NIP 0x2fff01a8
LR 0x32608
Call Trace:
0xc00000000143c5f8 (unreliable)
.rtas_call+0x224/0x320
.rtas_get_boot_time+0x70/0x150
.read_persistent_clock64+0x114/0x140
.read_persistent_wall_and_boot_offset+0x24/0x80
.timekeeping_init+0x40/0x29c
.start_kernel+0x674/0x8f0
start_here_common+0x1c/0x50
Unlike PAPR platforms where RTAS is only used in guests, on the IBM Cell
machines Linux runs with MSR[HV] set but also uses RTAS, provided by
SLOF.
Fix it by copying the MSR[HV] bit from the MSR value we've just read
using mfmsr into the value used for RTAS.
It seems like we could also fix it using an #ifdef CELL to set MSR[HV],
but that doesn't work because it's possible to build a single kernel
image that runs on both Cell native and pseries.
In the Linux kernel, the following vulnerability has been resolved:
Input: iforce - wake up after clearing IFORCE_XMIT_RUNNING flag
syzbot is reporting hung task at __input_unregister_device() [1], for
iforce_close() waiting at wait_event_interruptible() with dev->mutex held
is blocking input_disconnect_device() from __input_unregister_device().
It seems that the cause is simply that commit c2b27ef672992a20 ("Input:
iforce - wait for command completion when closing the device") forgot to
call wake_up() after clear_bit().
Fix this problem by introducing a helper that calls clear_bit() followed
by wake_up_all().
In the Linux kernel, the following vulnerability has been resolved:
iio: light: cm3605: Fix an error handling path in cm3605_probe()
The commit in Fixes also introduced a new error handling path which should
goto the existing error handling path.
Otherwise some resources leak.
In the Linux kernel, the following vulnerability has been resolved:
misc: fastrpc: fix memory corruption on probe
Add the missing sanity check on the probed-session count to avoid
corrupting memory beyond the fixed-size slab-allocated session array
when there are more than FASTRPC_MAX_SESSIONS sessions defined in the
devicetree.
In the Linux kernel, the following vulnerability has been resolved:
firmware_loader: Fix use-after-free during unregister
In the following code within firmware_upload_unregister(), the call to
device_unregister() could result in the dev_release function freeing the
fw_upload_priv structure before it is dereferenced for the call to
module_put(). This bug was found by the kernel test robot using
CONFIG_KASAN while running the firmware selftests.
device_unregister(&fw_sysfs->dev);
module_put(fw_upload_priv->module);
The problem is fixed by copying fw_upload_priv->module to a local variable
for use when calling device_unregister().
In the Linux kernel, the following vulnerability has been resolved:
misc: fastrpc: fix memory corruption on open
The probe session-duplication overflow check incremented the session
count also when there were no more available sessions so that memory
beyond the fixed-size slab-allocated session array could be corrupted in
fastrpc_session_alloc() on open().
In the Linux kernel, the following vulnerability has been resolved:
firmware_loader: Fix memory leak in firmware upload
In the case of firmware-upload, an instance of struct fw_upload is
allocated in firmware_upload_register(). This data needs to be freed
in fw_dev_release(). Create a new fw_upload_free() function in
sysfs_upload.c to handle the firmware-upload specific memory frees
and incorporate the missing kfree call for the fw_upload structure.
In the Linux kernel, the following vulnerability has been resolved:
vt: Clear selection before changing the font
When changing the console font with ioctl(KDFONTOP) the new font size
can be bigger than the previous font. A previous selection may thus now
be outside of the new screen size and thus trigger out-of-bounds
accesses to graphics memory if the selection is removed in
vc_do_resize().
Prevent such out-of-memory accesses by dropping the selection before the
various con_font_set() console handlers are called.
In the Linux kernel, the following vulnerability has been resolved:
binder: fix alloc->vma_vm_mm null-ptr dereference
Syzbot reported a couple issues introduced by commit 44e602b4e52f
("binder_alloc: add missing mmap_lock calls when using the VMA"), in
which we attempt to acquire the mmap_lock when alloc->vma_vm_mm has not
been initialized yet.
This can happen if a binder_proc receives a transaction without having
previously called mmap() to setup the binder_proc->alloc space in [1].
Also, a similar issue occurs via binder_alloc_print_pages() when we try
to dump the debugfs binder stats file in [2].
Sample of syzbot's crash report:
==================================================================
KASAN: null-ptr-deref in range [0x0000000000000128-0x000000000000012f]
CPU: 0 PID: 3755 Comm: syz-executor229 Not tainted 6.0.0-rc1-next-20220819-syzkaller #0
syz-executor229[3755] cmdline: ./syz-executor2294415195
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 07/22/2022
RIP: 0010:__lock_acquire+0xd83/0x56d0 kernel/locking/lockdep.c:4923
[...]
Call Trace:
<TASK>
lock_acquire kernel/locking/lockdep.c:5666 [inline]
lock_acquire+0x1ab/0x570 kernel/locking/lockdep.c:5631
down_read+0x98/0x450 kernel/locking/rwsem.c:1499
mmap_read_lock include/linux/mmap_lock.h:117 [inline]
binder_alloc_new_buf_locked drivers/android/binder_alloc.c:405 [inline]
binder_alloc_new_buf+0xa5/0x19e0 drivers/android/binder_alloc.c:593
binder_transaction+0x242e/0x9a80 drivers/android/binder.c:3199
binder_thread_write+0x664/0x3220 drivers/android/binder.c:3986
binder_ioctl_write_read drivers/android/binder.c:5036 [inline]
binder_ioctl+0x3470/0x6d00 drivers/android/binder.c:5323
vfs_ioctl fs/ioctl.c:51 [inline]
__do_sys_ioctl fs/ioctl.c:870 [inline]
__se_sys_ioctl fs/ioctl.c:856 [inline]
__x64_sys_ioctl+0x193/0x200 fs/ioctl.c:856
do_syscall_x64 arch/x86/entry/common.c:50 [inline]
do_syscall_64+0x35/0xb0 arch/x86/entry/common.c:80
entry_SYSCALL_64_after_hwframe+0x63/0xcd
[...]
==================================================================
Fix these issues by setting up alloc->vma_vm_mm pointer during open()
and caching directly from current->mm. This guarantees we have a valid
reference to take the mmap_lock during scenarios described above.
[1] https://syzkaller.appspot.com/bug?extid=f7dc54e5be28950ac459
[2] https://syzkaller.appspot.com/bug?extid=a75ebe0452711c9e56d9
In the Linux kernel, the following vulnerability has been resolved:
clk: bcm: rpi: Prevent out-of-bounds access
The while loop in raspberrypi_discover_clocks() relies on the assumption
that the id of the last clock element is zero. Because this data comes
from the Videocore firmware and it doesn't guarantuee such a behavior
this could lead to out-of-bounds access. So fix this by providing
a sentinel element.
In the Linux kernel, the following vulnerability has been resolved:
Revert "usb: typec: ucsi: add a common function ucsi_unregister_connectors()"
The recent commit 87d0e2f41b8c ("usb: typec: ucsi: add a common
function ucsi_unregister_connectors()") introduced a regression that
caused NULL dereference at reading the power supply sysfs. It's a
stale sysfs entry that should have been removed but remains with NULL
ops. The commit changed the error handling to skip the entries after
a NULL con->wq, and this leaves the power device unreleased.
For addressing the regression, the straight revert is applied here.
Further code improvements can be done from the scratch again.
In the Linux kernel, the following vulnerability has been resolved:
USB: gadget: Fix obscure lockdep violation for udc_mutex
A recent commit expanding the scope of the udc_lock mutex in the
gadget core managed to cause an obscure and slightly bizarre lockdep
violation. In abbreviated form:
======================================================
WARNING: possible circular locking dependency detected
5.19.0-rc7+ #12510 Not tainted
------------------------------------------------------
udevadm/312 is trying to acquire lock:
ffff80000aae1058 (udc_lock){+.+.}-{3:3}, at: usb_udc_uevent+0x54/0xe0
but task is already holding lock:
ffff000002277548 (kn->active#4){++++}-{0:0}, at: kernfs_seq_start+0x34/0xe0
which lock already depends on the new lock.
the existing dependency chain (in reverse order) is:
-> #3 (kn->active#4){++++}-{0:0}:
lock_acquire+0x68/0x84
__kernfs_remove+0x268/0x380
kernfs_remove_by_name_ns+0x58/0xac
sysfs_remove_file_ns+0x18/0x24
device_del+0x15c/0x440
-> #2 (device_links_lock){+.+.}-{3:3}:
lock_acquire+0x68/0x84
__mutex_lock+0x9c/0x430
mutex_lock_nested+0x38/0x64
device_link_remove+0x3c/0xa0
_regulator_put.part.0+0x168/0x190
regulator_put+0x3c/0x54
devm_regulator_release+0x14/0x20
-> #1 (regulator_list_mutex){+.+.}-{3:3}:
lock_acquire+0x68/0x84
__mutex_lock+0x9c/0x430
mutex_lock_nested+0x38/0x64
regulator_lock_dependent+0x54/0x284
regulator_enable+0x34/0x80
phy_power_on+0x24/0x130
__dwc2_lowlevel_hw_enable+0x100/0x130
dwc2_lowlevel_hw_enable+0x18/0x40
dwc2_hsotg_udc_start+0x6c/0x2f0
gadget_bind_driver+0x124/0x1f4
-> #0 (udc_lock){+.+.}-{3:3}:
__lock_acquire+0x1298/0x20cc
lock_acquire.part.0+0xe0/0x230
lock_acquire+0x68/0x84
__mutex_lock+0x9c/0x430
mutex_lock_nested+0x38/0x64
usb_udc_uevent+0x54/0xe0
Evidently this was caused by the scope of udc_mutex being too large.
The mutex is only meant to protect udc->driver along with a few other
things. As far as I can tell, there's no reason for the mutex to be
held while the gadget core calls a gadget driver's ->bind or ->unbind
routine, or while a UDC is being started or stopped. (This accounts
for link #1 in the chain above, where the mutex is held while the
dwc2_hsotg_udc is started as part of driver probing.)
Gadget drivers' ->disconnect callbacks are problematic. Even though
usb_gadget_disconnect() will now acquire the udc_mutex, there's a
window in usb_gadget_bind_driver() between the times when the mutex is
released and the ->bind callback is invoked. If a disconnect occurred
during that window, we could call the driver's ->disconnect routine
before its ->bind routine. To prevent this from happening, it will be
necessary to prevent a UDC from connecting while it has no gadget
driver. This should be done already but it doesn't seem to be;
currently usb_gadget_connect() has no check for this. Such a check
will have to be added later.
Some degree of mutual exclusion is required in soft_connect_store(),
which can dereference udc->driver at arbitrary times since it is a
sysfs callback. The solution here is to acquire the gadget's device
lock rather than the udc_mutex. Since the driver core guarantees that
the device lock is always held during driver binding and unbinding,
this will make the accesses in soft_connect_store() mutually exclusive
with any changes to udc->driver.
Lastly, it turns out there is one place which should hold the
udc_mutex but currently does not: The function_show() routine needs
protection while it dereferences udc->driver. The missing lock and
unlock calls are added.
In the Linux kernel, the following vulnerability has been resolved:
wifi: mac80211: Don't finalize CSA in IBSS mode if state is disconnected
When we are not connected to a channel, sending channel "switch"
announcement doesn't make any sense.
The BSS list is empty in that case. This causes the for loop in
cfg80211_get_bss() to be bypassed, so the function returns NULL
(check line 1424 of net/wireless/scan.c), causing the WARN_ON()
in ieee80211_ibss_csa_beacon() to get triggered (check line 500
of net/mac80211/ibss.c), which was consequently reported on the
syzkaller dashboard.
Thus, check if we have an existing connection before generating
the CSA beacon in ieee80211_ibss_finish_csa().
In the Linux kernel, the following vulnerability has been resolved:
tty: n_gsm: add sanity check for gsm->receive in gsm_receive_buf()
A null pointer dereference can happen when attempting to access the
"gsm->receive()" function in gsmld_receive_buf(). Currently, the code
assumes that gsm->recieve is only called after MUX activation.
Since the gsmld_receive_buf() function can be accessed without the need to
initialize the MUX, the gsm->receive() function will not be set and a
NULL pointer dereference will occur.
Fix this by avoiding the call to "gsm->receive()" in case the function is
not initialized by adding a sanity check.
Call Trace:
<TASK>
gsmld_receive_buf+0x1c2/0x2f0 drivers/tty/n_gsm.c:2861
tiocsti drivers/tty/tty_io.c:2293 [inline]
tty_ioctl+0xa75/0x15d0 drivers/tty/tty_io.c:2692
vfs_ioctl fs/ioctl.c:51 [inline]
__do_sys_ioctl fs/ioctl.c:870 [inline]
__se_sys_ioctl fs/ioctl.c:856 [inline]
__x64_sys_ioctl+0x193/0x200 fs/ioctl.c:856
do_syscall_x64 arch/x86/entry/common.c:50 [inline]
do_syscall_64+0x35/0xb0 arch/x86/entry/common.c:80
entry_SYSCALL_64_after_hwframe+0x63/0xcd
In the Linux kernel, the following vulnerability has been resolved:
binder: fix UAF of ref->proc caused by race condition
A transaction of type BINDER_TYPE_WEAK_HANDLE can fail to increment the
reference for a node. In this case, the target proc normally releases
the failed reference upon close as expected. However, if the target is
dying in parallel the call will race with binder_deferred_release(), so
the target could have released all of its references by now leaving the
cleanup of the new failed reference unhandled.
The transaction then ends and the target proc gets released making the
ref->proc now a dangling pointer. Later on, ref->node is closed and we
attempt to take spin_lock(&ref->proc->inner_lock), which leads to the
use-after-free bug reported below. Let's fix this by cleaning up the
failed reference on the spot instead of relying on the target to do so.
==================================================================
BUG: KASAN: use-after-free in _raw_spin_lock+0xa8/0x150
Write of size 4 at addr ffff5ca207094238 by task kworker/1:0/590
CPU: 1 PID: 590 Comm: kworker/1:0 Not tainted 5.19.0-rc8 #10
Hardware name: linux,dummy-virt (DT)
Workqueue: events binder_deferred_func
Call trace:
dump_backtrace.part.0+0x1d0/0x1e0
show_stack+0x18/0x70
dump_stack_lvl+0x68/0x84
print_report+0x2e4/0x61c
kasan_report+0xa4/0x110
kasan_check_range+0xfc/0x1a4
__kasan_check_write+0x3c/0x50
_raw_spin_lock+0xa8/0x150
binder_deferred_func+0x5e0/0x9b0
process_one_work+0x38c/0x5f0
worker_thread+0x9c/0x694
kthread+0x188/0x190
ret_from_fork+0x10/0x20
In the Linux kernel, the following vulnerability has been resolved:
cifs: fix small mempool leak in SMB2_negotiate()
In some cases of failure (dialect mismatches) in SMB2_negotiate(), after
the request is sent, the checks would return -EIO when they should be
rather setting rc = -EIO and jumping to neg_exit to free the response
buffer from mempool.
In the Linux kernel, the following vulnerability has been resolved:
media: mceusb: Use new usb_control_msg_*() routines
Automatic kernel fuzzing led to a WARN about invalid pipe direction in
the mceusb driver:
------------[ cut here ]------------
usb 6-1: BOGUS control dir, pipe 80000380 doesn't match bRequestType 40
WARNING: CPU: 0 PID: 2465 at drivers/usb/core/urb.c:410
usb_submit_urb+0x1326/0x1820 drivers/usb/core/urb.c:410
Modules linked in:
CPU: 0 PID: 2465 Comm: kworker/0:2 Not tainted 5.19.0-rc4-00208-g69cb6c6556ad #1
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS
1.13.0-1ubuntu1.1 04/01/2014
Workqueue: usb_hub_wq hub_event
RIP: 0010:usb_submit_urb+0x1326/0x1820 drivers/usb/core/urb.c:410
Code: 7c 24 40 e8 ac 23 91 fd 48 8b 7c 24 40 e8 b2 70 1b ff 45 89 e8
44 89 f1 4c 89 e2 48 89 c6 48 c7 c7 a0 30 a9 86 e8 48 07 11 02 <0f> 0b
e9 1c f0 ff ff e8 7e 23 91 fd 0f b6 1d 63 22 83 05 31 ff 41
RSP: 0018:ffffc900032becf0 EFLAGS: 00010282
RAX: 0000000000000000 RBX: ffff8881100f3058 RCX: 0000000000000000
RDX: ffffc90004961000 RSI: ffff888114c6d580 RDI: fffff52000657d90
RBP: ffff888105ad90f0 R08: ffffffff812c3638 R09: 0000000000000000
R10: 0000000000000005 R11: ffffed1023504ef1 R12: ffff888105ad9000
R13: 0000000000000040 R14: 0000000080000380 R15: ffff88810ba96500
FS: 0000000000000000(0000) GS:ffff88811a800000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00007ffe810bda58 CR3: 000000010b720000 CR4: 0000000000350ef0
Call Trace:
<TASK>
usb_start_wait_urb+0x101/0x4c0 drivers/usb/core/message.c:58
usb_internal_control_msg drivers/usb/core/message.c:102 [inline]
usb_control_msg+0x31c/0x4a0 drivers/usb/core/message.c:153
mceusb_gen1_init drivers/media/rc/mceusb.c:1431 [inline]
mceusb_dev_probe+0x258e/0x33f0 drivers/media/rc/mceusb.c:1807
The reason for the warning is clear enough; the driver sends an
unusual read request on endpoint 0 but does not set the USB_DIR_IN bit
in the bRequestType field.
More importantly, the whole situation can be avoided and the driver
simplified by converting it over to the relatively new
usb_control_msg_recv() and usb_control_msg_send() routines. That's
what this fix does.
In the Linux kernel, the following vulnerability has been resolved:
USB: core: Prevent nested device-reset calls
Automatic kernel fuzzing revealed a recursive locking violation in
usb-storage:
============================================
WARNING: possible recursive locking detected
5.18.0 #3 Not tainted
--------------------------------------------
kworker/1:3/1205 is trying to acquire lock:
ffff888018638db8 (&us_interface_key[i]){+.+.}-{3:3}, at:
usb_stor_pre_reset+0x35/0x40 drivers/usb/storage/usb.c:230
but task is already holding lock:
ffff888018638db8 (&us_interface_key[i]){+.+.}-{3:3}, at:
usb_stor_pre_reset+0x35/0x40 drivers/usb/storage/usb.c:230
...
stack backtrace:
CPU: 1 PID: 1205 Comm: kworker/1:3 Not tainted 5.18.0 #3
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS
1.13.0-1ubuntu1.1 04/01/2014
Workqueue: usb_hub_wq hub_event
Call Trace:
<TASK>
__dump_stack lib/dump_stack.c:88 [inline]
dump_stack_lvl+0xcd/0x134 lib/dump_stack.c:106
print_deadlock_bug kernel/locking/lockdep.c:2988 [inline]
check_deadlock kernel/locking/lockdep.c:3031 [inline]
validate_chain kernel/locking/lockdep.c:3816 [inline]
__lock_acquire.cold+0x152/0x3ca kernel/locking/lockdep.c:5053
lock_acquire kernel/locking/lockdep.c:5665 [inline]
lock_acquire+0x1ab/0x520 kernel/locking/lockdep.c:5630
__mutex_lock_common kernel/locking/mutex.c:603 [inline]
__mutex_lock+0x14f/0x1610 kernel/locking/mutex.c:747
usb_stor_pre_reset+0x35/0x40 drivers/usb/storage/usb.c:230
usb_reset_device+0x37d/0x9a0 drivers/usb/core/hub.c:6109
r871xu_dev_remove+0x21a/0x270 drivers/staging/rtl8712/usb_intf.c:622
usb_unbind_interface+0x1bd/0x890 drivers/usb/core/driver.c:458
device_remove drivers/base/dd.c:545 [inline]
device_remove+0x11f/0x170 drivers/base/dd.c:537
__device_release_driver drivers/base/dd.c:1222 [inline]
device_release_driver_internal+0x1a7/0x2f0 drivers/base/dd.c:1248
usb_driver_release_interface+0x102/0x180 drivers/usb/core/driver.c:627
usb_forced_unbind_intf+0x4d/0xa0 drivers/usb/core/driver.c:1118
usb_reset_device+0x39b/0x9a0 drivers/usb/core/hub.c:6114
This turned out not to be an error in usb-storage but rather a nested
device reset attempt. That is, as the rtl8712 driver was being
unbound from a composite device in preparation for an unrelated USB
reset (that driver does not have pre_reset or post_reset callbacks),
its ->remove routine called usb_reset_device() -- thus nesting one
reset call within another.
Performing a reset as part of disconnect processing is a questionable
practice at best. However, the bug report points out that the USB
core does not have any protection against nested resets. Adding a
reset_in_progress flag and testing it will prevent such errors in the
future.
In the Linux kernel, the following vulnerability has been resolved:
dma-buf/dma-resv: check if the new fence is really later
Previously when we added a fence to a dma_resv object we always
assumed the the newer than all the existing fences.
With Jason's work to add an UAPI to explicit export/import that's not
necessary the case any more. So without this check we would allow
userspace to force the kernel into an use after free error.
Since the change is very small and defensive it's probably a good
idea to backport this to stable kernels as well just in case others
are using the dma_resv object in the same way.
In the Linux kernel, the following vulnerability has been resolved:
wifi: mac80211: Fix UAF in ieee80211_scan_rx()
ieee80211_scan_rx() tries to access scan_req->flags after a
null check, but a UAF is observed when the scan is completed
and __ieee80211_scan_completed() executes, which then calls
cfg80211_scan_done() leading to the freeing of scan_req.
Since scan_req is rcu_dereference()'d, prevent the racing in
__ieee80211_scan_completed() by ensuring that from mac80211's
POV it is no longer accessed from an RCU read critical section
before we call cfg80211_scan_done().
In the Linux kernel, the following vulnerability has been resolved:
gpio: virtuser: fix potential out-of-bound write
If the caller wrote more characters, count is truncated to the max
available space in "simple_write_to_buffer". Check that the input
size does not exceed the buffer size. Write a zero termination
afterwards.
In the Linux kernel, the following vulnerability has been resolved:
spi-rockchip: Fix register out of bounds access
Do not write native chip select stuff for GPIO chip selects.
GPIOs can be numbered much higher than native CS.
Also, it makes no sense.
In the Linux kernel, the following vulnerability has been resolved:
drm/amd/display: Increase block_sequence array size
[Why]
It's possible to generate more than 50 steps in hwss_build_fast_sequence,
for example with a 6-pipe asic where all pipes are in one MPC chain. This
overflows the block_sequence buffer and corrupts block_sequence_steps,
causing a crash.
[How]
Expand block_sequence to 100 items. A naive upper bound on the possible
number of steps for a 6-pipe asic, ignoring the potential for steps to be
mutually exclusive, is 91 with current code, therefore 100 is sufficient.
In the Linux kernel, the following vulnerability has been resolved:
crypto: algif_hash - fix double free in hash_accept
If accept(2) is called on socket type algif_hash with
MSG_MORE flag set and crypto_ahash_import fails,
sk2 is freed. However, it is also freed in af_alg_release,
leading to slab-use-after-free error.
In the Linux kernel, the following vulnerability has been resolved:
ALSA: pcm: Fix race of buffer access at PCM OSS layer
The PCM OSS layer tries to clear the buffer with the silence data at
initialization (or reconfiguration) of a stream with the explicit call
of snd_pcm_format_set_silence() with runtime->dma_area. But this may
lead to a UAF because the accessed runtime->dma_area might be freed
concurrently, as it's performed outside the PCM ops.
For avoiding it, move the code into the PCM core and perform it inside
the buffer access lock, so that it won't be changed during the
operation.
In the Linux kernel, the following vulnerability has been resolved:
platform/x86: dell-wmi-sysman: Avoid buffer overflow in current_password_store()
If the 'buf' array received from the user contains an empty string, the
'length' variable will be zero. Accessing the 'buf' array element with
index 'length - 1' will result in a buffer overflow.
Add a check for an empty string.
Found by Linux Verification Center (linuxtesting.org) with SVACE.
In the Linux kernel, the following vulnerability has been resolved:
alloc_tag: allocate percpu counters for module tags dynamically
When a module gets unloaded it checks whether any of its tags are still in
use and if so, we keep the memory containing module's allocation tags
alive until all tags are unused. However percpu counters referenced by
the tags are freed by free_module(). This will lead to UAF if the memory
allocated by a module is accessed after module was unloaded.
To fix this we allocate percpu counters for module allocation tags
dynamically and we keep it alive for tags which are still in use after
module unloading. This also removes the requirement of a larger
PERCPU_MODULE_RESERVE when memory allocation profiling is enabled because
percpu memory for counters does not need to be reserved anymore.
In the Linux kernel, the following vulnerability has been resolved:
scsi: target: iscsi: Fix timeout on deleted connection
NOPIN response timer may expire on a deleted connection and crash with
such logs:
Did not receive response to NOPIN on CID: 0, failing connection for I_T Nexus (null),i,0x00023d000125,iqn.2017-01.com.iscsi.target,t,0x3d
BUG: Kernel NULL pointer dereference on read at 0x00000000
NIP strlcpy+0x8/0xb0
LR iscsit_fill_cxn_timeout_err_stats+0x5c/0xc0 [iscsi_target_mod]
Call Trace:
iscsit_handle_nopin_response_timeout+0xfc/0x120 [iscsi_target_mod]
call_timer_fn+0x58/0x1f0
run_timer_softirq+0x740/0x860
__do_softirq+0x16c/0x420
irq_exit+0x188/0x1c0
timer_interrupt+0x184/0x410
That is because nopin response timer may be re-started on nopin timer
expiration.
Stop nopin timer before stopping the nopin response timer to be sure
that no one of them will be re-started.
In the Linux kernel, the following vulnerability has been resolved:
vhost-scsi: protect vq->log_used with vq->mutex
The vhost-scsi completion path may access vq->log_base when vq->log_used is
already set to false.
vhost-thread QEMU-thread
vhost_scsi_complete_cmd_work()
-> vhost_add_used()
-> vhost_add_used_n()
if (unlikely(vq->log_used))
QEMU disables vq->log_used
via VHOST_SET_VRING_ADDR.
mutex_lock(&vq->mutex);
vq->log_used = false now!
mutex_unlock(&vq->mutex);
QEMU gfree(vq->log_base)
log_used()
-> log_write(vq->log_base)
Assuming the VMM is QEMU. The vq->log_base is from QEMU userpace and can be
reclaimed via gfree(). As a result, this causes invalid memory writes to
QEMU userspace.
The control queue path has the same issue.
In the Linux kernel, the following vulnerability has been resolved:
libnvdimm/labels: Fix divide error in nd_label_data_init()
If a faulty CXL memory device returns a broken zero LSA size in its
memory device information (Identify Memory Device (Opcode 4000h), CXL
spec. 3.1, 8.2.9.9.1.1), a divide error occurs in the libnvdimm
driver:
Oops: divide error: 0000 [#1] PREEMPT SMP NOPTI
RIP: 0010:nd_label_data_init+0x10e/0x800 [libnvdimm]
Code and flow:
1) CXL Command 4000h returns LSA size = 0
2) config_size is assigned to zero LSA size (CXL pmem driver):
drivers/cxl/pmem.c: .config_size = mds->lsa_size,
3) max_xfer is set to zero (nvdimm driver):
drivers/nvdimm/label.c: max_xfer = min_t(size_t, ndd->nsarea.max_xfer, config_size);
4) A subsequent DIV_ROUND_UP() causes a division by zero:
drivers/nvdimm/label.c: /* Make our initial read size a multiple of max_xfer size */
drivers/nvdimm/label.c: read_size = min(DIV_ROUND_UP(read_size, max_xfer) * max_xfer,
drivers/nvdimm/label.c- config_size);
Fix this by checking the config size parameter by extending an
existing check.
In the Linux kernel, the following vulnerability has been resolved:
x86/mm: Check return value from memblock_phys_alloc_range()
At least with CONFIG_PHYSICAL_START=0x100000, if there is < 4 MiB of
contiguous free memory available at this point, the kernel will crash
and burn because memblock_phys_alloc_range() returns 0 on failure,
which leads memblock_phys_free() to throw the first 4 MiB of physical
memory to the wolves.
At a minimum it should fail gracefully with a meaningful diagnostic,
but in fact everything seems to work fine without the weird reserve
allocation.
In the Linux kernel, the following vulnerability has been resolved:
ASoC: sma1307: Add NULL check in sma1307_setting_loaded()
All varibale allocated by kzalloc and devm_kzalloc could be NULL.
Multiple pointer checks and their cleanup are added.
This issue is found by our static analysis tool
In the Linux kernel, the following vulnerability has been resolved:
PCI: endpoint: pci-epf-test: Fix double free that causes kernel to oops
Fix a kernel oops found while testing the stm32_pcie Endpoint driver
with handling of PERST# deassertion:
During EP initialization, pci_epf_test_alloc_space() allocates all BARs,
which are further freed if epc_set_bar() fails (for instance, due to no
free inbound window).
However, when pci_epc_set_bar() fails, the error path:
pci_epc_set_bar() ->
pci_epf_free_space()
does not clear the previous assignment to epf_test->reg[bar].
Then, if the host reboots, the PERST# deassertion restarts the BAR
allocation sequence with the same allocation failure (no free inbound
window), creating a double free situation since epf_test->reg[bar] was
deallocated and is still non-NULL.
Thus, make sure that pci_epf_alloc_space() and pci_epf_free_space()
invocations are symmetric, and as such, set epf_test->reg[bar] to NULL
when memory is freed.
[kwilczynski: commit log]
In the Linux kernel, the following vulnerability has been resolved:
crypto: lzo - Fix compression buffer overrun
Unlike the decompression code, the compression code in LZO never
checked for output overruns. It instead assumes that the caller
always provides enough buffer space, disregarding the buffer length
provided by the caller.
Add a safe compression interface that checks for the end of buffer
before each write. Use the safe interface in crypto/lzo.
In the Linux kernel, the following vulnerability has been resolved:
rseq: Fix segfault on registration when rseq_cs is non-zero
The rseq_cs field is documented as being set to 0 by user-space prior to
registration, however this is not currently enforced by the kernel. This
can result in a segfault on return to user-space if the value stored in
the rseq_cs field doesn't point to a valid struct rseq_cs.
The correct solution to this would be to fail the rseq registration when
the rseq_cs field is non-zero. However, some older versions of glibc
will reuse the rseq area of previous threads without clearing the
rseq_cs field and will also terminate the process if the rseq
registration fails in a secondary thread. This wasn't caught in testing
because in this case the leftover rseq_cs does point to a valid struct
rseq_cs.
What we can do is clear the rseq_cs field on registration when it's
non-zero which will prevent segfaults on registration and won't break
the glibc versions that reuse rseq areas on thread creation.
In the Linux kernel, the following vulnerability has been resolved:
dm cache: prevent BUG_ON by blocking retries on failed device resumes
A cache device failing to resume due to mapping errors should not be
retried, as the failure leaves a partially initialized policy object.
Repeating the resume operation risks triggering BUG_ON when reloading
cache mappings into the incomplete policy object.
Reproduce steps:
1. create a cache metadata consisting of 512 or more cache blocks,
with some mappings stored in the first array block of the mapping
array. Here we use cache_restore v1.0 to build the metadata.
cat <<EOF >> cmeta.xml
<superblock uuid="" block_size="128" nr_cache_blocks="512" \
policy="smq" hint_width="4">
<mappings>
<mapping cache_block="0" origin_block="0" dirty="false"/>
</mappings>
</superblock>
EOF
dmsetup create cmeta --table "0 8192 linear /dev/sdc 0"
cache_restore -i cmeta.xml -o /dev/mapper/cmeta --metadata-version=2
dmsetup remove cmeta
2. wipe the second array block of the mapping array to simulate
data degradations.
mapping_root=$(dd if=/dev/sdc bs=1c count=8 skip=192 \
2>/dev/null | hexdump -e '1/8 "%u\n"')
ablock=$(dd if=/dev/sdc bs=1c count=8 skip=$((4096*mapping_root+2056)) \
2>/dev/null | hexdump -e '1/8 "%u\n"')
dd if=/dev/zero of=/dev/sdc bs=4k count=1 seek=$ablock
3. try bringing up the cache device. The resume is expected to fail
due to the broken array block.
dmsetup create cmeta --table "0 8192 linear /dev/sdc 0"
dmsetup create cdata --table "0 65536 linear /dev/sdc 8192"
dmsetup create corig --table "0 524288 linear /dev/sdc 262144"
dmsetup create cache --notable
dmsetup load cache --table "0 524288 cache /dev/mapper/cmeta \
/dev/mapper/cdata /dev/mapper/corig 128 2 metadata2 writethrough smq 0"
dmsetup resume cache
4. try resuming the cache again. An unexpected BUG_ON is triggered
while loading cache mappings.
dmsetup resume cache
Kernel logs:
(snip)
------------[ cut here ]------------
kernel BUG at drivers/md/dm-cache-policy-smq.c:752!
Oops: invalid opcode: 0000 [#1] PREEMPT SMP KASAN NOPTI
CPU: 0 UID: 0 PID: 332 Comm: dmsetup Not tainted 6.13.4 #3
RIP: 0010:smq_load_mapping+0x3e5/0x570
Fix by disallowing resume operations for devices that failed the
initial attempt.
In the Linux kernel, the following vulnerability has been resolved:
orangefs: Do not truncate file size
'len' is used to store the result of i_size_read(), so making 'len'
a size_t results in truncation to 4GiB on 32-bit systems.
In the Linux kernel, the following vulnerability has been resolved:
virtio: break and reset virtio devices on device_shutdown()
Hongyu reported a hang on kexec in a VM. QEMU reported invalid memory
accesses during the hang.
Invalid read at addr 0x102877002, size 2, region '(null)', reason: rejected
Invalid write at addr 0x102877A44, size 2, region '(null)', reason: rejected
...
It was traced down to virtio-console. Kexec works fine if virtio-console
is not in use.
The issue is that virtio-console continues to write to the MMIO even after
underlying virtio-pci device is reset.
Additionally, Eric noticed that IOMMUs are reset before devices, if
devices are not reset on shutdown they continue to poke at guest memory
and get errors from the IOMMU. Some devices get wedged then.
The problem can be solved by breaking all virtio devices on virtio
bus shutdown, then resetting them.
In the Linux kernel, the following vulnerability has been resolved:
dm: fix unconditional IO throttle caused by REQ_PREFLUSH
When a bio with REQ_PREFLUSH is submitted to dm, __send_empty_flush()
generates a flush_bio with REQ_OP_WRITE | REQ_PREFLUSH | REQ_SYNC,
which causes the flush_bio to be throttled by wbt_wait().
An example from v5.4, similar problem also exists in upstream:
crash> bt 2091206
PID: 2091206 TASK: ffff2050df92a300 CPU: 109 COMMAND: "kworker/u260:0"
#0 [ffff800084a2f7f0] __switch_to at ffff80004008aeb8
#1 [ffff800084a2f820] __schedule at ffff800040bfa0c4
#2 [ffff800084a2f880] schedule at ffff800040bfa4b4
#3 [ffff800084a2f8a0] io_schedule at ffff800040bfa9c4
#4 [ffff800084a2f8c0] rq_qos_wait at ffff8000405925bc
#5 [ffff800084a2f940] wbt_wait at ffff8000405bb3a0
#6 [ffff800084a2f9a0] __rq_qos_throttle at ffff800040592254
#7 [ffff800084a2f9c0] blk_mq_make_request at ffff80004057cf38
#8 [ffff800084a2fa60] generic_make_request at ffff800040570138
#9 [ffff800084a2fae0] submit_bio at ffff8000405703b4
#10 [ffff800084a2fb50] xlog_write_iclog at ffff800001280834 [xfs]
#11 [ffff800084a2fbb0] xlog_sync at ffff800001280c3c [xfs]
#12 [ffff800084a2fbf0] xlog_state_release_iclog at ffff800001280df4 [xfs]
#13 [ffff800084a2fc10] xlog_write at ffff80000128203c [xfs]
#14 [ffff800084a2fcd0] xlog_cil_push at ffff8000012846dc [xfs]
#15 [ffff800084a2fda0] xlog_cil_push_work at ffff800001284a2c [xfs]
#16 [ffff800084a2fdb0] process_one_work at ffff800040111d08
#17 [ffff800084a2fe00] worker_thread at ffff8000401121cc
#18 [ffff800084a2fe70] kthread at ffff800040118de4
After commit 2def2845cc33 ("xfs: don't allow log IO to be throttled"),
the metadata submitted by xlog_write_iclog() should not be throttled.
But due to the existence of the dm layer, throttling flush_bio indirectly
causes the metadata bio to be throttled.
Fix this by conditionally adding REQ_IDLE to flush_bio.bi_opf, which makes
wbt_should_throttle() return false to avoid wbt_wait().
In the Linux kernel, the following vulnerability has been resolved:
genirq/msi: Store the IOMMU IOVA directly in msi_desc instead of iommu_cookie
The IOMMU translation for MSI message addresses has been a 2-step process,
separated in time:
1) iommu_dma_prepare_msi(): A cookie pointer containing the IOVA address
is stored in the MSI descriptor when an MSI interrupt is allocated.
2) iommu_dma_compose_msi_msg(): this cookie pointer is used to compute a
translated message address.
This has an inherent lifetime problem for the pointer stored in the cookie
that must remain valid between the two steps. However, there is no locking
at the irq layer that helps protect the lifetime. Today, this works under
the assumption that the iommu domain is not changed while MSI interrupts
being programmed. This is true for normal DMA API users within the kernel,
as the iommu domain is attached before the driver is probed and cannot be
changed while a driver is attached.
Classic VFIO type1 also prevented changing the iommu domain while VFIO was
running as it does not support changing the "container" after starting up.
However, iommufd has improved this so that the iommu domain can be changed
during VFIO operation. This potentially allows userspace to directly race
VFIO_DEVICE_ATTACH_IOMMUFD_PT (which calls iommu_attach_group()) and
VFIO_DEVICE_SET_IRQS (which calls into iommu_dma_compose_msi_msg()).
This potentially causes both the cookie pointer and the unlocked call to
iommu_get_domain_for_dev() on the MSI translation path to become UAFs.
Fix the MSI cookie UAF by removing the cookie pointer. The translated IOVA
address is already known during iommu_dma_prepare_msi() and cannot change.
Thus, it can simply be stored as an integer in the MSI descriptor.
The other UAF related to iommu_get_domain_for_dev() will be addressed in
patch "iommu: Make iommu_dma_prepare_msi() into a generic operation" by
using the IOMMU group mutex.