commit 68035c80e1 upstream.
Way back in 2017, fuzzing the 4.14-rc2 USB stack with syzkaller kicked
up the following WARNING from the UVC chain scanning code:
| list_add double add: new=ffff880069084010, prev=ffff880069084010,
| next=ffff880067d22298.
| ------------[ cut here ]------------
| WARNING: CPU: 1 PID: 1846 at lib/list_debug.c:31 __list_add_valid+0xbd/0xf0
| Modules linked in:
| CPU: 1 PID: 1846 Comm: kworker/1:2 Not tainted
| 4.14.0-rc2-42613-g1488251d1a98 #238
| Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS Bochs 01/01/2011
| Workqueue: usb_hub_wq hub_event
| task: ffff88006b01ca40 task.stack: ffff880064358000
| RIP: 0010:__list_add_valid+0xbd/0xf0 lib/list_debug.c:29
| RSP: 0018:ffff88006435ddd0 EFLAGS: 00010286
| RAX: 0000000000000058 RBX: ffff880067d22298 RCX: 0000000000000000
| RDX: 0000000000000058 RSI: ffffffff85a58800 RDI: ffffed000c86bbac
| RBP: ffff88006435dde8 R08: 1ffff1000c86ba52 R09: 0000000000000000
| R10: 0000000000000002 R11: 0000000000000000 R12: ffff880069084010
| R13: ffff880067d22298 R14: ffff880069084010 R15: ffff880067d222a0
| FS: 0000000000000000(0000) GS:ffff88006c900000(0000) knlGS:0000000000000000
| CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
| CR2: 0000000020004ff2 CR3: 000000006b447000 CR4: 00000000000006e0
| Call Trace:
| __list_add ./include/linux/list.h:59
| list_add_tail+0x8c/0x1b0 ./include/linux/list.h:92
| uvc_scan_chain_forward.isra.8+0x373/0x416
| drivers/media/usb/uvc/uvc_driver.c:1471
| uvc_scan_chain drivers/media/usb/uvc/uvc_driver.c:1585
| uvc_scan_device drivers/media/usb/uvc/uvc_driver.c:1769
| uvc_probe+0x77f2/0x8f00 drivers/media/usb/uvc/uvc_driver.c:2104
Looking into the output from usbmon, the interesting part is the
following data packet:
ffff880069c63e00 30710169 C Ci:1:002:0 0 143 = 09028f00 01030080
00090403 00000e01 00000924 03000103 7c003328 010204db
If we drop the lead configuration and interface descriptors, we're left
with an output terminal descriptor describing a generic display:
/* Output terminal descriptor */
buf[0] 09
buf[1] 24
buf[2] 03 /* UVC_VC_OUTPUT_TERMINAL */
buf[3] 00 /* ID */
buf[4] 01 /* type == 0x0301 (UVC_OTT_DISPLAY) */
buf[5] 03
buf[6] 7c
buf[7] 00 /* source ID refers to self! */
buf[8] 33
The problem with this descriptor is that it is self-referential: the
source ID of 0 matches itself! This causes the 'struct uvc_entity'
representing the display to be added to its chain list twice during
'uvc_scan_chain()': once via 'uvc_scan_chain_entity()' when it is
processed directly from the 'dev->entities' list and then again
immediately afterwards when trying to follow the source ID in
'uvc_scan_chain_forward()'
Add a check before adding an entity to a chain list to ensure that the
entity is not already part of a chain.
Link: https://lore.kernel.org/linux-media/CAAeHK+z+Si69jUR+N-SjN9q4O+o5KFiNManqEa-PjUta7EOb7A@mail.gmail.com/
Cc: <stable@vger.kernel.org>
Fixes: c0efd23292 ("V4L/DVB (8145a): USB Video Class driver")
Reported-by: Andrey Konovalov <andreyknvl@google.com>
Signed-off-by: Will Deacon <will@kernel.org>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit d407302895 ]
The current logic that calls pci_disable_device() in __bnxt_close_nic()
during firmware reset is flawed. If firmware is still alive, we're
disabling the device too early, causing some firmware commands to
not reach the firmware.
Fix it by moving the logic to bnxt_reset_close(). If firmware is
in fatal condition, we call pci_disable_device() before we free
any of the rings to prevent DMA corruption of the freed rings. If
firmware is still alive, we call pci_disable_device() after the
last firmware message has been sent.
Fixes: 3bc7d4a352 ("bnxt_en: Add BNXT_STATE_IN_FW_RESET state.")
Signed-off-by: Vasundhara Volam <vasundhara-v.volam@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit dff6bc1bfd ]
Commit 6d97985072 ("isdn: move capi drivers to staging") cleaned up the
isdn drivers and split the MAINTAINERS section for ISDN, but missed to add
the terminal slash for the two directories mISDN and hardware. Hence, all
files in those directories were not part of the new ISDN/mISDN SUBSYSTEM,
but were considered to be part of "THE REST".
Rectify the situation, and while at it, also complete the section with two
further build files that belong to that subsystem.
This was identified with a small script that finds all files belonging to
"THE REST" according to the current MAINTAINERS file, and I investigated
upon its output.
Fixes: 6d97985072 ("isdn: move capi drivers to staging")
Signed-off-by: Lukas Bulwahn <lukas.bulwahn@gmail.com>
Acked-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit b5ce31b5e1 ]
Be sure to include all the packet type bits in the mask.
Fixes: fbfb803153 ("ionic: Add hardware init and device commands")
Signed-off-by: Shannon Nelson <snelson@pensando.io>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit 12de2eadf8 ]
bnxt_ulp_start() needs to be called before SRIOV is re-enabled after
firmware reset. Re-enabling SRIOV may consume all the resources and
may cause the RDMA driver to fail to get MSIX and other resources.
Fix it by calling bnxt_ulp_start() first before calling
bnxt_reenable_sriov().
We re-arrange the logic so that we call bnxt_ulp_start() and
bnxt_reenable_sriov() in proper sequence in bnxt_fw_reset_task() and
bnxt_open(). The former is the normal coordinated firmware reset sequence
and the latter is firmware reset while the function is down. This new
logic is now more straight forward and will now fix both scenarios.
Fixes: f3a6d206c2 ("bnxt_en: Call bnxt_ulp_stop()/bnxt_ulp_start() during error recovery.")
Reported-by: Vasundhara Volam <vasundhara-v.volam@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit 2318ca8aef ]
The probe() might enable a VDDIO regulator, which needs to be disabled
again before calling regulator_put(). Add a remove() function.
Fixes: 2f664823a4 ("net: phy: at803x: add device tree binding")
Signed-off-by: Michael Walle <michael@walle.cc>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit c16d4ee0e3 ]
Put the current logic in bnxt_open() to re-enable SRIOV after detecting
firmware reset into a new function bnxt_reenable_sriov(). This call
needs to be invoked in the firmware reset path also in the next patch.
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit 5273a191dc ]
When a call is disconnected, the connection pointer from the call is
cleared to make sure it isn't used again and to prevent further attempted
transmission for the call. Unfortunately, there might be a daemon trying
to use it at the same time to transmit a packet.
Fix this by keeping call->conn set, but setting a flag on the call to
indicate disconnection instead.
Remove also the bits in the transmission functions where the conn pointer is
checked and a ref taken under spinlock as this is now redundant.
Fixes: 8d94aa381d ("rxrpc: Calls shouldn't hold socket refs")
Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit 04d36d748f ]
The introduction of a split between the reference count on rxrpc_local
objects and the usage count didn't quite go far enough. A number of kernel
work items need to make use of the socket to perform transmission. These
also need to get an active count on the local object to prevent the socket
from being closed.
Fix this by getting the active count in those places.
Also split out the raw active count get/put functions as these places tend
to hold refs on the rxrpc_local object already, so getting and putting an
extra object ref is just a waste of time.
The problem can lead to symptoms like:
BUG: kernel NULL pointer dereference, address: 0000000000000018
..
CPU: 2 PID: 818 Comm: kworker/u9:0 Not tainted 5.5.0-fscache+ #51
...
RIP: 0010:selinux_socket_sendmsg+0x5/0x13
...
Call Trace:
security_socket_sendmsg+0x2c/0x3e
sock_sendmsg+0x1a/0x46
rxrpc_send_keepalive+0x131/0x1ae
rxrpc_peer_keepalive_worker+0x219/0x34b
process_one_work+0x18e/0x271
worker_thread+0x1a3/0x247
kthread+0xe6/0xeb
ret_from_fork+0x1f/0x30
Fixes: 730c5fd42c ("rxrpc: Fix local endpoint refcounting")
Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit f71dbf2fb2 ]
In rxrpc_input_data(), rxrpc_notify_socket() is called if the base sequence
number of the packet is immediately following the hard-ack point at the end
of the function. However, this isn't sufficient, since the recvmsg side
may have been advancing the window and then overrun the position in which
we're adding - at which point rx_hard_ack >= seq0 and no notification is
generated.
Fix this by always generating a notification at the end of the input
function.
Without this, a long call may stall, possibly indefinitely.
Fixes: 248f219cb8 ("rxrpc: Rewrite the data and ack handling code")
Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit fac20b9e73 ]
Fix rxrpc_put_local() to not access local->debug_id after calling
atomic_dec_return() as, unless that returned n==0, we no longer have the
right to access the object.
Fixes: 06d9532fa6 ("rxrpc: Fix read-after-free in rxrpc_queue_local()")
Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit 784f8344de ]
tp->segs_in and tp->segs_out need to be cleared in tcp_disconnect().
tcp_disconnect() is rarely used, but it is worth fixing it.
Fixes: 2efd055c53 ("tcp: add tcpi_segs_in and tcpi_segs_out to tcp_info")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Marcelo Ricardo Leitner <mleitner@redhat.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit db7ffee6f3 ]
tp->data_segs_in and tp->data_segs_out need to be cleared
in tcp_disconnect().
tcp_disconnect() is rarely used, but it is worth fixing it.
Fixes: a44d6eacda ("tcp: Add RFC4898 tcpEStatsPerfDataSegsOut/In")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Martin KaFai Lau <kafai@fb.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit c13c48c00a ]
total_retrans needs to be cleared in tcp_disconnect().
tcp_disconnect() is rarely used, but it is worth fixing it.
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: SeongJae Park <sjpark@amazon.de>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit 14b41a2959 ]
When running v5.5 with a rootfs on NFS, memory abort may happen in
the system resume stage:
Unable to handle kernel paging request at virtual address dead00000000012a
[dead00000000012a] address between user and kernel address ranges
pc : run_timer_softirq+0x334/0x3d8
lr : run_timer_softirq+0x244/0x3d8
x1 : ffff800011cafe80 x0 : dead000000000122
Call trace:
run_timer_softirq+0x334/0x3d8
efi_header_end+0x114/0x234
irq_exit+0xd0/0xd8
__handle_domain_irq+0x60/0xb0
gic_handle_irq+0x58/0xa8
el1_irq+0xb8/0x180
arch_cpu_idle+0x10/0x18
do_idle+0x1d8/0x2b0
cpu_startup_entry+0x24/0x40
secondary_start_kernel+0x1b4/0x208
Code: f9000693 a9400660 f9000020 b4000040 (f9000401)
---[ end trace bb83ceeb4c482071 ]---
Kernel panic - not syncing: Fatal exception in interrupt
SMP: stopping secondary CPUs
SMP: failed to stop secondary CPUs 2-3
Kernel Offset: disabled
CPU features: 0x00002,2300aa30
Memory Limit: none
---[ end Kernel panic - not syncing: Fatal exception in interrupt ]---
It's found that stmmac_xmit() and stmmac_resume() sometimes might
run concurrently, possibly resulting in a race condition between
mod_timer() and setup_timer(), being called by stmmac_xmit() and
stmmac_resume() respectively.
Since the resume() runs setup_timer() every time, it'd be safer to
have del_timer_sync() in the suspend() as the counterpart.
Signed-off-by: Nicolin Chen <nicoleotsuka@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit 18e4960c18 ]
The driver currently only calls netdev_set_tc_queue when the number of
TCs is greater than 1. Instead, the comparison should be greater than
or equal to 1. Even with 1 TC, we need to set the queue mapping.
This bug can cause warnings when the number of TCs is changed back to 1.
Fixes: 7809592d3e ("bnxt_en: Enable MSIX early in bnxt_init_one().")
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit 44efc78d0e ]
In order to fix XDP support if sw buffer management is used as fallback
for hw bm devices, define MVNETA_SKB_HEADROOM as maximum between
XDP_PACKET_HEADROOM and NET_SKB_PAD and let the hw aligns the IP header
to 4-byte boundary.
Fix rx_offset_correction initialization if mvneta_bm_port_init fails in
mvneta_resume routine
Fixes: 0db51da7a8 ("net: mvneta: add basic XDP support")
Tested-by: Sven Auhagen <sven.auhagen@voleatech.de>
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit 9fbf082f56 ]
VF numbers were assigned to node_guid and port_guid, but cleared
right before such query calls were issued. It caused to return
node/port GUIDs of VF index 0 for all VFs.
Fixes: 30aad41721 ("net/core: Add support for getting VF GUIDs")
Reported-by: Adrian Chiris <adrianc@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit 599be01ee5 ]
As Eric noticed, tcindex_alloc_perfect_hash() uses cp->hash
to compute the size of memory allocation, but cp->hash is
set again after the allocation, this caused an out-of-bound
access.
So we have to move all cp->hash initialization and computation
before the memory allocation. Move cp->mask and cp->shift together
as cp->hash may need them for computation too.
Reported-and-tested-by: syzbot+35d4dea36c387813ed31@syzkaller.appspotmail.com
Fixes: 331b72922c ("net: sched: RCU cls_tcindex")
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Cc: John Fastabend <john.fastabend@gmail.com>
Cc: Jamal Hadi Salim <jhs@mojatatu.com>
Cc: Jiri Pirko <jiri@resnulli.us>
Cc: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit 0d0d9a388a ]
In the past it was possible to create multiple L2TPv3 sessions with the
same session id as long as the sessions belonged to different tunnels.
The resulting sessions had issues when used with IP encapsulated tunnels,
but worked fine with UDP encapsulated ones. Some applications began to
rely on this behaviour to avoid having to negotiate unique session ids.
Some time ago a change was made to require session ids to be unique across
all tunnels, breaking the applications making use of this "feature".
This change relaxes the duplicate session id check to allow duplicates
if both of the colliding sessions belong to UDP encapsulated tunnels.
Fixes: dbdbc73b44 ("l2tp: fix duplicate session creation")
Signed-off-by: Ridge Kennedy <ridge.kennedy@alliedtelesis.co.nz>
Acked-by: James Chapman <jchapman@katalix.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit cda2cab077 ]
Latest kernels get the phys_port_name via devlink, if
ndo_get_phys_port_name is not defined. To provide the phys_port_name
correctly, register devlink before registering netdev.
Also call devlink_port_type_eth_set() after registering netdev as
devlink port updates the netdev structure and notifies user.
Cc: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: Vasundhara Volam <vasundhara-v.volam@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit 34ca70ef7d ]
As discussed in the strace issue tracker, it appears that the sparc32
sysvipc support has been broken for the past 11 years. It was however
working in compat mode, which is how it must have escaped most of the
regular testing.
The problem is that a cleanup patch inadvertently changed the uid/gid
fields in struct ipc64_perm from 32-bit types to 16-bit types in uapi
headers.
Both glibc and uclibc-ng still use the original types, so they should
work fine with compat mode, but not natively. Change the definitions
to use __kernel_uid32_t and __kernel_gid32_t again.
Fixes: 83c86984bf ("sparc: unify ipcbuf.h")
Link: https://github.com/strace/strace/issues/116
Cc: <stable@vger.kernel.org> # v2.6.29
Cc: Sam Ravnborg <sam@ravnborg.org>
Cc: "Dmitry V . Levin" <ldv@altlinux.org>
Cc: Rich Felker <dalias@libc.org>
Cc: libc-alpha@sourceware.org
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit d55966c427 upstream.
There was some logic added a while ago to clear out f_bavail in statfs()
if we did not have enough free metadata space to satisfy our global
reserve. This was incorrect at the time, however didn't really pose a
problem for normal file systems because we would often allocate chunks
if we got this low on free metadata space, and thus wouldn't really hit
this case unless we were actually full.
Fast forward to today and now we are much better about not allocating
metadata chunks all of the time. Couple this with d792b0f197 ("btrfs:
always reserve our entire size for the global reserve") which now means
we'll easily have a larger global reserve than our free space, we are
now more likely to trip over this while still having plenty of space.
Fix this by skipping this logic if the global rsv's space_info is not
full. space_info->full is 0 unless we've attempted to allocate a chunk
for that space_info and that has failed. If this happens then the space
for the global reserve is definitely sacred and we need to report
b_avail == 0, but before then we can just use our calculated b_avail.
Reported-by: Martin Steigerwald <martin@lichtvoll.de>
Fixes: ca8a51b3a9 ("btrfs: statfs: report zero available if metadata are exhausted")
CC: stable@vger.kernel.org # 4.5+
Reviewed-by: Qu Wenruo <wqu@suse.com>
Tested-By: Martin Steigerwald <martin@lichtvoll.de>
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 3bc0bb36fa upstream.
The test_cgcore_no_internal_process_constraint_on_threads selftest when
running with subsystem controlling noise triggers two warnings:
> [ 597.443115] WARNING: CPU: 1 PID: 28167 at kernel/cgroup/cgroup.c:3131 cgroup_apply_control_enable+0xe0/0x3f0
> [ 597.443413] WARNING: CPU: 1 PID: 28167 at kernel/cgroup/cgroup.c:3177 cgroup_apply_control_disable+0xa6/0x160
Both stem from a call to cgroup_type_write. The first warning was also
triggered by syzkaller.
When we're switching cgroup to threaded mode shortly after a subsystem
was disabled on it, we can see the respective subsystem css dying there.
The warning in cgroup_apply_control_enable is harmless in this case
since we're not adding new subsys anyway.
The warning in cgroup_apply_control_disable indicates an attempt to kill
css of recently disabled subsystem repeatedly.
The commit prevents these situations by making cgroup_type_write wait
for all dying csses to go away before re-applying subtree controls.
When at it, the locations of WARN_ON_ONCE calls are moved so that
warning is triggered only when we are about to misuse the dying css.
Reported-by: syzbot+5493b2a54d31d6aea629@syzkaller.appspotmail.com
Reported-by: Christian Brauner <christian.brauner@ubuntu.com>
Signed-off-by: Michal Koutný <mkoutny@suse.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 11eb85ec42 upstream.
Syzbot managed to trigger a use after free "KASAN: use-after-free Write
in hci_sock_bind". I have reviewed the code manually and one possibly
cause I have found is that we are not holding lock_sock(sk) when we do
the hci_dev_put(hdev) in hci_sock_release(). My theory is that the bind
and the release are racing against each other which results in this use
after free.
Reported-by: syzbot+eba992608adf3d796bcc@syzkaller.appspotmail.com
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 9a655c77ff upstream.
tpk_write()/tpk_close() could be interrupted when holding a mutex, then
in timer handler tpk_write() may be called again trying to acquire same
mutex, lead to deadlock.
Google syzbot reported this issue with CONFIG_DEBUG_ATOMIC_SLEEP
enabled:
BUG: sleeping function called from invalid context at
kernel/locking/mutex.c:938
in_atomic(): 1, irqs_disabled(): 0, non_block: 0, pid: 0, name: swapper/1
1 lock held by swapper/1/0:
...
Call Trace:
<IRQ>
dump_stack+0x197/0x210
___might_sleep.cold+0x1fb/0x23e
__might_sleep+0x95/0x190
__mutex_lock+0xc5/0x13c0
mutex_lock_nested+0x16/0x20
tpk_write+0x5d/0x340
resync_tnc+0x1b6/0x320
call_timer_fn+0x1ac/0x780
run_timer_softirq+0x6c3/0x1790
__do_softirq+0x262/0x98c
irq_exit+0x19b/0x1e0
smp_apic_timer_interrupt+0x1a3/0x610
apic_timer_interrupt+0xf/0x20
</IRQ>
See link https://syzkaller.appspot.com/bug?extid=2eeef62ee31f9460ad65 for
more details.
Fix it by using spinlock in process context instead of mutex and having
interrupt disabled in critical section.
Reported-by: syzbot+2eeef62ee31f9460ad65@syzkaller.appspotmail.com
Signed-off-by: Zhenzhong Duan <zhenzhong.duan@gmail.com>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Link: https://lore.kernel.org/r/20200113034842.435-1-zhenzhong.duan@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit a8772fad01 upstream.
syzbot is reporting that there is a race at tomoyo_stat_update() [1].
Although it is acceptable to fail to track exact number of times policy
was updated, convert to atomic_t because this is not a hot path.
[1] https://syzkaller.appspot.com/bug?id=a4d7b973972eeed410596e6604580e0133b0fc04
Reported-by: syzbot <syzbot+efea72d4a0a1d03596cd@syzkaller.appspotmail.com>
Signed-off-by: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 26cff63712 upstream.
It is not a fatal error if reading the mac address or the remote control
decoder state fails.
Reported-by: syzbot+ec869945d3dde5f33b43@syzkaller.appspotmail.com
Signed-off-by: Sean Young <sean@mess.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 51d0c99b39 upstream.
If usb_bulk_msg() fails, actual_length can be uninitialized.
Reported-by: syzbot+9d42b7773d2fecd983ab@syzkaller.appspotmail.com
Signed-off-by: Sean Young <sean@mess.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit eecc70d22a upstream.
This results in an uninitialized variable read.
Reported-by: syzbot+6bf9606ee955b646c0e1@syzkaller.appspotmail.com
Signed-off-by: Sean Young <sean@mess.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 5474ca7da6 upstream.
When a filesystem is mounted with jdev mount option, we store the
journal device name in an allocated string in superblock. However we
fail to ever free that string. Fix it.
Reported-by: syzbot+1c6756baf4b16b94d2a6@syzkaller.appspotmail.com
Fixes: c3aa077648 ("reiserfs: Properly display mount options in /proc/mounts")
CC: stable@vger.kernel.org
Signed-off-by: Jan Kara <jack@suse.cz>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit c7a91bc7c2 upstream.
What we are trying to do is change the '=' character to a NUL terminator
and then at the end of the function we restore it back to an '='. The
problem is there are two error paths where we jump to the end of the
function before we have replaced the '=' with NUL.
We end up putting the '=' in the wrong place (possibly one element
before the start of the buffer).
Link: http://lkml.kernel.org/r/20200115055426.vdjwvry44nfug7yy@kili.mountain
Reported-by: syzbot+e64a13c5369a194d67df@syzkaller.appspotmail.com
Fixes: 095f1fc4eb ("mempolicy: rework shmem mpol parsing and display")
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Acked-by: Vlastimil Babka <vbabka@suse.cz>
Dmitry Vyukov <dvyukov@google.com>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Dan Carpenter <dan.carpenter@oracle.com>
Cc: Lee Schermerhorn <lee.schermerhorn@hp.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit d7bbd6c1b0 upstream.
Since v4.3-rc1 commit 0723c05fb7 ("arm64: enable more compressed
Image formats"), it is possible to build Image.{bz2,lz4,lzma,lzo}
AArch64 images. However, the commit missed adding support for removing
those images on 'make ARCH=arm64 (dist)clean'.
Fix this by adding them to the target list.
Make sure to match the order of the recipes in the makefile.
Cc: stable@vger.kernel.org # v4.3+
Fixes: 0723c05fb7 ("arm64: enable more compressed Image formats")
Signed-off-by: Dirk Behme <dirk.behme@de.bosch.com>
Signed-off-by: Eugeniu Rosca <erosca@de.adit-jv.com>
Reviewed-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Signed-off-by: Will Deacon <will@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 6c4798d3f0 upstream.
Disable a couple of compilation warnings (which are treated as errors)
on strlcpy() definition and declaration, allowing users to compile perf
and kernel (objtool) when:
1. glibc have strlcpy() (such as in ALT Linux since 2004) objtool and
perf build fails with this (in gcc):
In file included from exec-cmd.c:3:
tools/include/linux/string.h:20:15: error: redundant redeclaration of ‘strlcpy’ [-Werror=redundant-decls]
20 | extern size_t strlcpy(char *dest, const char *src, size_t size);
2. clang ignores `-Wredundant-decls', but produces another warning when
building perf:
CC util/string.o
../lib/string.c:99:8: error: attribute declaration must precede definition [-Werror,-Wignored-attributes]
size_t __weak strlcpy(char *dest, const char *src, size_t size)
../../tools/include/linux/compiler.h:66:34: note: expanded from macro '__weak'
# define __weak __attribute__((weak))
/usr/include/bits/string_fortified.h:151:8: note: previous definition is here
__NTH (strlcpy (char *__restrict __dest, const char *__restrict __src,
Committer notes:
The
#pragma GCC diagnostic
directive was introduced in gcc 4.6, so check for that as well.
Fixes: ce99091 ("perf tools: Move strlcpy() from perf to tools/lib/string.c")
Fixes: 0215d59 ("tools lib: Reinstate strlcpy() header guard with __UCLIBC__")
Resolves: https://bugzilla.kernel.org/show_bug.cgi?id=118481
Signed-off-by: Vitaly Chikunov <vt@altlinux.org>
Reviewed-by: Dmitry Levin <ldv@altlinux.org>
Cc: Dmitry Levin <ldv@altlinux.org>
Cc: Josh Poimboeuf <jpoimboe@redhat.com>
Cc: kbuild test robot <lkp@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: stable@vger.kernel.org
Cc: Vineet Gupta <vineet.gupta1@synopsys.com>
Link: http://lore.kernel.org/lkml/20191224172029.19690-1-vt@altlinux.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 2fee1a7cc6 upstream.
The commit 4585fbcb53 ("PM / devfreq: Modify the device name as devfreq(X) for
sysfs") changed the node name to devfreq(x). After this commit, it is not
possible to get the device name through /sys/class/devfreq/devfreq(X)/*.
Add new name attribute in order to get device name.
Cc: stable@vger.kernel.org
Fixes: 4585fbcb53 ("PM / devfreq: Modify the device name as devfreq(X) for sysfs")
Signed-off-by: Chanwoo Choi <cw00.choi@samsung.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 69334ca530 upstream.
On 32-bit platform the size of long is only 32 bits which makes wrong
offset in the array of 64 bit size.
Calculate offset based on BITS_PER_LONG.
Link: http://lkml.kernel.org/r/20200109103601.45929-1-andriy.shevchenko@linux.intel.com
Fixes: 30544ed5de ("lib/bitmap: introduce bitmap_replace() helper")
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Reported-by: Guenter Roeck <linux@roeck-us.net>
Cc: Rasmus Villemoes <linux@rasmusvillemoes.dk>
Cc: Yury Norov <yury.norov@gmail.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit eed0f953b9 upstream.
On filesystems with a block size smaller than the page size,
gfs2_find_jhead can split a page across two bios (for example, when
blocks are not allocated consecutively). When that happens, the first
bio that completes will unlock the page in its bi_end_io handler even
though the page hasn't been read completely yet. Fix that by using a
chained bio for the rest of the page.
While at it, clean up the sector calculation logic in
gfs2_log_alloc_bio. In gfs2_find_jhead, simplify the disk block and
offset calculation logic and fix a variable name.
Fixes: f4686c26ec ("gfs2: read journal in large chunks")
Cc: stable@vger.kernel.org # v5.2+
Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit 074fadee59 ]
There is a race condition in the following scenario which results in an
use-after-free issue when reading a monitoring file and deleting the
parent ctrl_mon group concurrently:
Thread 1 calls atomic_inc() to take refcount of rdtgrp and then calls
kernfs_break_active_protection() to drop the active reference of kernfs
node in rdtgroup_kn_lock_live().
In Thread 2, kernfs_remove() is a blocking routine. It waits on all sub
kernfs nodes to drop the active reference when removing all subtree
kernfs nodes recursively. Thread 2 could block on kernfs_remove() until
Thread 1 calls kernfs_break_active_protection(). Only after
kernfs_remove() completes the refcount of rdtgrp could be trusted.
Before Thread 1 calls atomic_inc() and kernfs_break_active_protection(),
Thread 2 could call kfree() when the refcount of rdtgrp (sentry) is 0
instead of 1 due to the race.
In Thread 1, in rdtgroup_kn_unlock(), referring to earlier rdtgrp memory
(rdtgrp->waitcount) which was already freed in Thread 2 results in
use-after-free issue.
Thread 1 (rdtgroup_mondata_show) Thread 2 (rdtgroup_rmdir)
-------------------------------- -------------------------
rdtgroup_kn_lock_live
/*
* kn active protection until
* kernfs_break_active_protection(kn)
*/
rdtgrp = kernfs_to_rdtgroup(kn)
rdtgroup_kn_lock_live
atomic_inc(&rdtgrp->waitcount)
mutex_lock
rdtgroup_rmdir_ctrl
free_all_child_rdtgrp
/*
* sentry->waitcount should be 1
* but is 0 now due to the race.
*/
kfree(sentry)*[1]
/*
* Only after kernfs_remove()
* completes, the refcount of
* rdtgrp could be trusted.
*/
atomic_inc(&rdtgrp->waitcount)
/* kn->active-- */
kernfs_break_active_protection(kn)
rdtgroup_ctrl_remove
rdtgrp->flags = RDT_DELETED
/*
* Blocking routine, wait for
* all sub kernfs nodes to drop
* active reference in
* kernfs_break_active_protection.
*/
kernfs_remove(rdtgrp->kn)
rdtgroup_kn_unlock
mutex_unlock
atomic_dec_and_test(
&rdtgrp->waitcount)
&& (flags & RDT_DELETED)
kernfs_unbreak_active_protection(kn)
kfree(rdtgrp)
mutex_lock
mon_event_read
rdtgroup_kn_unlock
mutex_unlock
/*
* Use-after-free: refer to earlier rdtgrp
* memory which was freed in [1].
*/
atomic_dec_and_test(&rdtgrp->waitcount)
&& (flags & RDT_DELETED)
/* kn->active++ */
kernfs_unbreak_active_protection(kn)
kfree(rdtgrp)
Fix it by moving free_all_child_rdtgrp() to after kernfs_remove() in
rdtgroup_rmdir_ctrl() to ensure it has the accurate refcount of rdtgrp.
Fixes: f3cbeacaa0 ("x86/intel_rdt/cqm: Add rmdir support")
Suggested-by: Reinette Chatre <reinette.chatre@intel.com>
Signed-off-by: Xiaochen Shen <xiaochen.shen@intel.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Reviewed-by: Reinette Chatre <reinette.chatre@intel.com>
Reviewed-by: Tony Luck <tony.luck@intel.com>
Acked-by: Thomas Gleixner <tglx@linutronix.de>
Cc: stable@vger.kernel.org
Link: https://lkml.kernel.org/r/1578500886-21771-3-git-send-email-xiaochen.shen@intel.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit b8511ccc75 ]
A resource group (rdtgrp) contains a reference count (rdtgrp->waitcount)
that indicates how many waiters expect this rdtgrp to exist. Waiters
could be waiting on rdtgroup_mutex or some work sitting on a task's
workqueue for when the task returns from kernel mode or exits.
The deletion of a rdtgrp is intended to have two phases:
(1) while holding rdtgroup_mutex the necessary cleanup is done and
rdtgrp->flags is set to RDT_DELETED,
(2) after releasing the rdtgroup_mutex, the rdtgrp structure is freed
only if there are no waiters and its flag is set to RDT_DELETED. Upon
gaining access to rdtgroup_mutex or rdtgrp, a waiter is required to check
for the RDT_DELETED flag.
When unmounting the resctrl file system or deleting ctrl_mon groups,
all of the subdirectories are removed and the data structure of rdtgrp
is forcibly freed without checking rdtgrp->waitcount. If at this point
there was a waiter on rdtgrp then a use-after-free issue occurs when the
waiter starts running and accesses the rdtgrp structure it was waiting
on.
See kfree() calls in [1], [2] and [3] in these two call paths in
following scenarios:
(1) rdt_kill_sb() -> rmdir_all_sub() -> free_all_child_rdtgrp()
(2) rdtgroup_rmdir() -> rdtgroup_rmdir_ctrl() -> free_all_child_rdtgrp()
There are several scenarios that result in use-after-free issue in
following:
Scenario 1:
-----------
In Thread 1, rdtgroup_tasks_write() adds a task_work callback
move_myself(). If move_myself() is scheduled to execute after Thread 2
rdt_kill_sb() is finished, referring to earlier rdtgrp memory
(rdtgrp->waitcount) which was already freed in Thread 2 results in
use-after-free issue.
Thread 1 (rdtgroup_tasks_write) Thread 2 (rdt_kill_sb)
------------------------------- ----------------------
rdtgroup_kn_lock_live
atomic_inc(&rdtgrp->waitcount)
mutex_lock
rdtgroup_move_task
__rdtgroup_move_task
/*
* Take an extra refcount, so rdtgrp cannot be freed
* before the call back move_myself has been invoked
*/
atomic_inc(&rdtgrp->waitcount)
/* Callback move_myself will be scheduled for later */
task_work_add(move_myself)
rdtgroup_kn_unlock
mutex_unlock
atomic_dec_and_test(&rdtgrp->waitcount)
&& (flags & RDT_DELETED)
mutex_lock
rmdir_all_sub
/*
* sentry and rdtgrp are freed
* without checking refcount
*/
free_all_child_rdtgrp
kfree(sentry)*[1]
kfree(rdtgrp)*[2]
mutex_unlock
/*
* Callback is scheduled to execute
* after rdt_kill_sb is finished
*/
move_myself
/*
* Use-after-free: refer to earlier rdtgrp
* memory which was freed in [1] or [2].
*/
atomic_dec_and_test(&rdtgrp->waitcount)
&& (flags & RDT_DELETED)
kfree(rdtgrp)
Scenario 2:
-----------
In Thread 1, rdtgroup_tasks_write() adds a task_work callback
move_myself(). If move_myself() is scheduled to execute after Thread 2
rdtgroup_rmdir() is finished, referring to earlier rdtgrp memory
(rdtgrp->waitcount) which was already freed in Thread 2 results in
use-after-free issue.
Thread 1 (rdtgroup_tasks_write) Thread 2 (rdtgroup_rmdir)
------------------------------- -------------------------
rdtgroup_kn_lock_live
atomic_inc(&rdtgrp->waitcount)
mutex_lock
rdtgroup_move_task
__rdtgroup_move_task
/*
* Take an extra refcount, so rdtgrp cannot be freed
* before the call back move_myself has been invoked
*/
atomic_inc(&rdtgrp->waitcount)
/* Callback move_myself will be scheduled for later */
task_work_add(move_myself)
rdtgroup_kn_unlock
mutex_unlock
atomic_dec_and_test(&rdtgrp->waitcount)
&& (flags & RDT_DELETED)
rdtgroup_kn_lock_live
atomic_inc(&rdtgrp->waitcount)
mutex_lock
rdtgroup_rmdir_ctrl
free_all_child_rdtgrp
/*
* sentry is freed without
* checking refcount
*/
kfree(sentry)*[3]
rdtgroup_ctrl_remove
rdtgrp->flags = RDT_DELETED
rdtgroup_kn_unlock
mutex_unlock
atomic_dec_and_test(
&rdtgrp->waitcount)
&& (flags & RDT_DELETED)
kfree(rdtgrp)
/*
* Callback is scheduled to execute
* after rdt_kill_sb is finished
*/
move_myself
/*
* Use-after-free: refer to earlier rdtgrp
* memory which was freed in [3].
*/
atomic_dec_and_test(&rdtgrp->waitcount)
&& (flags & RDT_DELETED)
kfree(rdtgrp)
If CONFIG_DEBUG_SLAB=y, Slab corruption on kmalloc-2k can be observed
like following. Note that "0x6b" is POISON_FREE after kfree(). The
corrupted bits "0x6a", "0x64" at offset 0x424 correspond to
waitcount member of struct rdtgroup which was freed:
Slab corruption (Not tainted): kmalloc-2k start=ffff9504c5b0d000, len=2048
420: 6b 6b 6b 6b 6a 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b kkkkjkkkkkkkkkkk
Single bit error detected. Probably bad RAM.
Run memtest86+ or a similar memory test tool.
Next obj: start=ffff9504c5b0d800, len=2048
000: 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b kkkkkkkkkkkkkkkk
010: 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b kkkkkkkkkkkkkkkk
Slab corruption (Not tainted): kmalloc-2k start=ffff9504c58ab800, len=2048
420: 6b 6b 6b 6b 64 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b kkkkdkkkkkkkkkkk
Prev obj: start=ffff9504c58ab000, len=2048
000: 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b kkkkkkkkkkkkkkkk
010: 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b kkkkkkkkkkkkkkkk
Fix this by taking reference count (waitcount) of rdtgrp into account in
the two call paths that currently do not do so. Instead of always
freeing the resource group it will only be freed if there are no waiters
on it. If there are waiters, the resource group will have its flags set
to RDT_DELETED.
It will be left to the waiter to free the resource group when it starts
running and finding that it was the last waiter and the resource group
has been removed (rdtgrp->flags & RDT_DELETED) since. (1) rdt_kill_sb()
-> rmdir_all_sub() -> free_all_child_rdtgrp() (2) rdtgroup_rmdir() ->
rdtgroup_rmdir_ctrl() -> free_all_child_rdtgrp()
Fixes: f3cbeacaa0 ("x86/intel_rdt/cqm: Add rmdir support")
Fixes: 60cf5e101f ("x86/intel_rdt: Add mkdir to resctrl file system")
Suggested-by: Reinette Chatre <reinette.chatre@intel.com>
Signed-off-by: Xiaochen Shen <xiaochen.shen@intel.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Reviewed-by: Reinette Chatre <reinette.chatre@intel.com>
Reviewed-by: Tony Luck <tony.luck@intel.com>
Acked-by: Thomas Gleixner <tglx@linutronix.de>
Cc: stable@vger.kernel.org
Link: https://lkml.kernel.org/r/1578500886-21771-2-git-send-email-xiaochen.shen@intel.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 334b0f4e9b ]
There is a race condition which results in a deadlock when rmdir and
mkdir execute concurrently:
$ ls /sys/fs/resctrl/c1/mon_groups/m1/
cpus cpus_list mon_data tasks
Thread 1: rmdir /sys/fs/resctrl/c1
Thread 2: mkdir /sys/fs/resctrl/c1/mon_groups/m1
3 locks held by mkdir/48649:
#0: (sb_writers#17){.+.+}, at: [<ffffffffb4ca2aa0>] mnt_want_write+0x20/0x50
#1: (&type->i_mutex_dir_key#8/1){+.+.}, at: [<ffffffffb4c8c13b>] filename_create+0x7b/0x170
#2: (rdtgroup_mutex){+.+.}, at: [<ffffffffb4a4389d>] rdtgroup_kn_lock_live+0x3d/0x70
4 locks held by rmdir/48652:
#0: (sb_writers#17){.+.+}, at: [<ffffffffb4ca2aa0>] mnt_want_write+0x20/0x50
#1: (&type->i_mutex_dir_key#8/1){+.+.}, at: [<ffffffffb4c8c3cf>] do_rmdir+0x13f/0x1e0
#2: (&type->i_mutex_dir_key#8){++++}, at: [<ffffffffb4c86d5d>] vfs_rmdir+0x4d/0x120
#3: (rdtgroup_mutex){+.+.}, at: [<ffffffffb4a4389d>] rdtgroup_kn_lock_live+0x3d/0x70
Thread 1 is deleting control group "c1". Holding rdtgroup_mutex,
kernfs_remove() removes all kernfs nodes under directory "c1"
recursively, then waits for sub kernfs node "mon_groups" to drop active
reference.
Thread 2 is trying to create a subdirectory "m1" in the "mon_groups"
directory. The wrapper kernfs_iop_mkdir() takes an active reference to
the "mon_groups" directory but the code drops the active reference to
the parent directory "c1" instead.
As a result, Thread 1 is blocked on waiting for active reference to drop
and never release rdtgroup_mutex, while Thread 2 is also blocked on
trying to get rdtgroup_mutex.
Thread 1 (rdtgroup_rmdir) Thread 2 (rdtgroup_mkdir)
(rmdir /sys/fs/resctrl/c1) (mkdir /sys/fs/resctrl/c1/mon_groups/m1)
------------------------- -------------------------
kernfs_iop_mkdir
/*
* kn: "m1", parent_kn: "mon_groups",
* prgrp_kn: parent_kn->parent: "c1",
*
* "mon_groups", parent_kn->active++: 1
*/
kernfs_get_active(parent_kn)
kernfs_iop_rmdir
/* "c1", kn->active++ */
kernfs_get_active(kn)
rdtgroup_kn_lock_live
atomic_inc(&rdtgrp->waitcount)
/* "c1", kn->active-- */
kernfs_break_active_protection(kn)
mutex_lock
rdtgroup_rmdir_ctrl
free_all_child_rdtgrp
sentry->flags = RDT_DELETED
rdtgroup_ctrl_remove
rdtgrp->flags = RDT_DELETED
kernfs_get(kn)
kernfs_remove(rdtgrp->kn)
__kernfs_remove
/* "mon_groups", sub_kn */
atomic_add(KN_DEACTIVATED_BIAS, &sub_kn->active)
kernfs_drain(sub_kn)
/*
* sub_kn->active == KN_DEACTIVATED_BIAS + 1,
* waiting on sub_kn->active to drop, but it
* never drops in Thread 2 which is blocked
* on getting rdtgroup_mutex.
*/
Thread 1 hangs here ---->
wait_event(sub_kn->active == KN_DEACTIVATED_BIAS)
...
rdtgroup_mkdir
rdtgroup_mkdir_mon(parent_kn, prgrp_kn)
mkdir_rdt_prepare(parent_kn, prgrp_kn)
rdtgroup_kn_lock_live(prgrp_kn)
atomic_inc(&rdtgrp->waitcount)
/*
* "c1", prgrp_kn->active--
*
* The active reference on "c1" is
* dropped, but not matching the
* actual active reference taken
* on "mon_groups", thus causing
* Thread 1 to wait forever while
* holding rdtgroup_mutex.
*/
kernfs_break_active_protection(
prgrp_kn)
/*
* Trying to get rdtgroup_mutex
* which is held by Thread 1.
*/
Thread 2 hangs here ----> mutex_lock
...
The problem is that the creation of a subdirectory in the "mon_groups"
directory incorrectly releases the active protection of its parent
directory instead of itself before it starts waiting for rdtgroup_mutex.
This is triggered by the rdtgroup_mkdir() flow calling
rdtgroup_kn_lock_live()/rdtgroup_kn_unlock() with kernfs node of the
parent control group ("c1") as argument. It should be called with kernfs
node "mon_groups" instead. What is currently missing is that the
kn->priv of "mon_groups" is NULL instead of pointing to the rdtgrp.
Fix it by pointing kn->priv to rdtgrp when "mon_groups" is created. Then
it could be passed to rdtgroup_kn_lock_live()/rdtgroup_kn_unlock()
instead. And then it operates on the same rdtgroup structure but handles
the active reference of kernfs node "mon_groups" to prevent deadlock.
The same changes are also made to the "mon_data" directories.
This results in some unused function parameters that will be cleaned up
in follow-up patch as the focus here is on the fix only in support of
backporting efforts.
Fixes: c7d9aac613 ("x86/intel_rdt/cqm: Add mkdir support for RDT monitoring")
Suggested-by: Reinette Chatre <reinette.chatre@intel.com>
Signed-off-by: Xiaochen Shen <xiaochen.shen@intel.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Reviewed-by: Reinette Chatre <reinette.chatre@intel.com>
Reviewed-by: Tony Luck <tony.luck@intel.com>
Acked-by: Thomas Gleixner <tglx@linutronix.de>
Cc: stable@vger.kernel.org
Link: https://lkml.kernel.org/r/1578500886-21771-4-git-send-email-xiaochen.shen@intel.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
