Carlos Llamas [Fri, 4 Nov 2022 17:54:49 +0000 (17:54 +0000)]
binder: fix UAF of alloc->vma in race with munmap()
In commit 720c24192404 ("ANDROID: binder: change down_write to
down_read") binder assumed the mmap read lock is sufficient to protect
alloc->vma inside binder_update_page_range(). This used to be accurate
until commit dd2283f2605e ("mm: mmap: zap pages with read mmap_sem in
munmap"), which now downgrades the mmap_lock after detaching the vma
from the rbtree in munmap(). Then it proceeds to teardown and free the
vma with only the read lock held.
This means that accesses to alloc->vma in binder_update_page_range() now
will race with vm_area_free() in munmap() and can cause a UAF as shown
in the following KASAN trace:
==================================================================
BUG: KASAN: use-after-free in vm_insert_page+0x7c/0x1f0
Read of size 8 at addr ffff16204ad00600 by task server/558
To prevent the race above, revert back to taking the mmap write lock
inside binder_update_page_range(). One might expect an increase of mmap
lock contention. However, binder already serializes these calls via top
level alloc->mutex. Also, there was no performance impact shown when
running the binder benchmark tests.
Note this patch is specific to stable branches 5.4 and 5.10. Since in
newer kernel releases binder no longer caches a pointer to the vma.
Instead, it has been refactored to use vma_lookup() which avoids the
issue described here. This switch was introduced in commit a43cfc87caaf
("android: binder: stop saving a pointer to the VMA").
Fixes: dd2283f2605e ("mm: mmap: zap pages with read mmap_sem in munmap") Reported-by: Jann Horn <jannh@google.com> Cc: <stable@vger.kernel.org> # 5.10.x Cc: Minchan Kim <minchan@kernel.org> Cc: Yang Shi <yang.shi@linux.alibaba.com> Cc: Liam Howlett <liam.howlett@oracle.com> Signed-off-by: Carlos Llamas <cmllamas@google.com> Acked-by: Todd Kjos <tkjos@google.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
06781a5026350 Fixes the calculation of the DEVICE_BUSY_TIMEOUT register
value from busy_timeout_cycles. busy_timeout_cycles is calculated wrong
though: It is calculated based on the maximum page read time, but the
timeout is also used for page write and block erase operations which
require orders of magnitude bigger timeouts.
Fix this by calculating busy_timeout_cycles from the maximum of
tBERS_max and tPROG_max.
This is for now the easiest and most obvious way to fix the driver.
There's room for improvements though: The NAND_OP_WAITRDY_INSTR tells us
the desired timeout for the current operation, so we could program the
timeout dynamically for each operation instead of setting a fixed
timeout. Also we could wire up the interrupt handler to actually detect
and forward timeouts occurred when waiting for the chip being ready.
As a sidenote I verified that the change in 06781a5026350 is really
correct. I wired up the interrupt handler in my tree and measured the
time between starting the operation and the timeout interrupt handler
coming in. The time increases 41us with each step in the timeout
register which corresponds to 4096 clock cycles with the 99MHz clock
that I have.
Fixes: 06781a5026350 ("mtd: rawnand: gpmi: Fix setting busy timeout setting") Fixes: b1206122069aa ("mtd: rawniand: gpmi: use core timings instead of an empirical derivation") Cc: stable@vger.kernel.org Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de> Acked-by: Han Xu <han.xu@nxp.com> Tested-by: Tomasz Moń <tomasz.mon@camlingroup.com> Signed-off-by: Richard Weinberger <richard@nod.at> Signed-off-by: Tim Harvey <tharvey@gateworks.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
syzbot reported a memory leak [0] related to IPV6_ADDRFORM.
The scenario is that while one thread is converting an IPv6 socket into
IPv4 with IPV6_ADDRFORM, another thread calls do_ipv6_setsockopt() and
allocates memory to inet6_sk(sk)->XXX after conversion.
Then, the converted sk with (tcp|udp)_prot never frees the IPv6 resources,
which inet6_destroy_sock() should have cleaned up.
setsockopt(IPV6_ADDRFORM) setsockopt(IPV6_DSTOPTS)
+-----------------------+ +----------------------+
- do_ipv6_setsockopt(sk, ...)
- sockopt_lock_sock(sk) - do_ipv6_setsockopt(sk, ...)
- lock_sock(sk) ^._ called via tcpv6_prot
- WRITE_ONCE(sk->sk_prot, &tcp_prot) before WRITE_ONCE()
- xchg(&np->opt, NULL)
- txopt_put(opt)
- sockopt_release_sock(sk)
- release_sock(sk) - sockopt_lock_sock(sk)
- lock_sock(sk)
- ipv6_set_opt_hdr(sk, ...)
- ipv6_update_options(sk, opt)
- xchg(&inet6_sk(sk)->opt, opt)
^._ opt is never freed.
- sockopt_release_sock(sk)
- release_sock(sk)
Since IPV6_DSTOPTS allocates options under lock_sock(), we can avoid this
memory leak by testing whether sk_family is changed by IPV6_ADDRFORM after
acquiring the lock.
This issue exists from the initial commit between IPV6_ADDRFORM and
IPV6_PKTOPTIONS.
Commit d7e7b9af104c ("fscrypt: stop using keyrings subsystem for
fscrypt_master_key") moved the keyring destruction from __put_super() to
generic_shutdown_super() so that the filesystem's block device(s) are
still available. Unfortunately, this causes a memory leak in the case
where a mount is attempted with the test_dummy_encryption mount option,
but the mount fails after the option has already been processed.
To fix this, attempt the keyring destruction in both places.
Reported-by: syzbot+104c2a89561289cec13e@syzkaller.appspotmail.com Fixes: d7e7b9af104c ("fscrypt: stop using keyrings subsystem for fscrypt_master_key") Signed-off-by: Eric Biggers <ebiggers@google.com> Reviewed-by: Christian Brauner (Microsoft) <brauner@kernel.org> Link: https://lore.kernel.org/r/20221011213838.209879-1-ebiggers@kernel.org Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The approach of fs/crypto/ internally managing the fscrypt_master_key
structs as the payloads of "struct key" objects contained in a
"struct key" keyring has outlived its usefulness. The original idea was
to simplify the code by reusing code from the keyrings subsystem.
However, several issues have arisen that can't easily be resolved:
- When a master key struct is destroyed, blk_crypto_evict_key() must be
called on any per-mode keys embedded in it. (This started being the
case when inline encryption support was added.) Yet, the keyrings
subsystem can arbitrarily delay the destruction of keys, even past the
time the filesystem was unmounted. Therefore, currently there is no
easy way to call blk_crypto_evict_key() when a master key is
destroyed. Currently, this is worked around by holding an extra
reference to the filesystem's request_queue(s). But it was overlooked
that the request_queue reference is *not* guaranteed to pin the
corresponding blk_crypto_profile too; for device-mapper devices that
support inline crypto, it doesn't. This can cause a use-after-free.
- When the last inode that was using an incompletely-removed master key
is evicted, the master key removal is completed by removing the key
struct from the keyring. Currently this is done via key_invalidate().
Yet, key_invalidate() takes the key semaphore. This can deadlock when
called from the shrinker, since in fscrypt_ioctl_add_key(), memory is
allocated with GFP_KERNEL under the same semaphore.
- More generally, the fact that the keyrings subsystem can arbitrarily
delay the destruction of keys (via garbage collection delay, or via
random processes getting temporary key references) is undesirable, as
it means we can't strictly guarantee that all secrets are ever wiped.
- Doing the master key lookups via the keyrings subsystem results in the
key_permission LSM hook being called. fscrypt doesn't want this, as
all access control for encrypted files is designed to happen via the
files themselves, like any other files. The workaround which SELinux
users are using is to change their SELinux policy to grant key search
access to all domains. This works, but it is an odd extra step that
shouldn't really have to be done.
The fix for all these issues is to change the implementation to what I
should have done originally: don't use the keyrings subsystem to keep
track of the filesystem's fscrypt_master_key structs. Instead, just
store them in a regular kernel data structure, and rework the reference
counting, locking, and lifetime accordingly. Retain support for
RCU-mode key lookups by using a hash table. Replace fscrypt_sb_free()
with fscrypt_sb_delete(), which releases the keys synchronously and runs
a bit earlier during unmount, so that block devices are still available.
A side effect of this patch is that neither the master keys themselves
nor the filesystem keyrings will be listed in /proc/keys anymore.
("Master key users" and the master key users keyrings will still be
listed.) However, this was mostly an implementation detail, and it was
intended just for debugging purposes. I don't know of anyone using it.
This patch does *not* change how "master key users" (->mk_users) works;
that still uses the keyrings subsystem. That is still needed for key
quotas, and changing that isn't necessary to solve the issues listed
above. If we decide to change that too, it would be a separate patch.
I've marked this as fixing the original commit that added the fscrypt
keyring, but as noted above the most important issue that this patch
fixes wasn't introduced until the addition of inline encryption support.
The stated reasons for separating fscrypt_master_key::mk_secret_sem from
the standard semaphore contained in every 'struct key' no longer apply.
First, due to commit a992b20cd4ee ("fscrypt: add
fscrypt_prepare_new_inode() and fscrypt_set_context()"),
fscrypt_get_encryption_info() is no longer called from within a
filesystem transaction.
Second, due to commit d3ec10aa9581 ("KEYS: Don't write out to userspace
while holding key semaphore"), the semaphore for the "keyring" key type
no longer ranks above page faults.
That leaves performance as the only possible reason to keep the separate
mk_secret_sem. Specifically, having mk_secret_sem reduces the
contention between setup_file_encryption_key() and
FS_IOC_{ADD,REMOVE}_ENCRYPTION_KEY. However, these ioctls aren't
executed often, so this doesn't seem to be worth the extra complexity.
Therefore, simplify the locking design by just using key->sem instead of
mk_secret_sem.
Treat the claimed 96kHz 1ch in the descriptors as 48kHz 2ch, so that
the audio stream doesn't sound mono. Also fix initial stream
alignment, so that left and right channels are in the correct order.
If bfq_schedule_dispatch() is called from bfq_idle_slice_timer_body(),
then 'bfqd->queued' is read without holding 'bfqd->lock'. This is
wrong since it can be wrote concurrently.
Fix the problem by holding 'bfqd->lock' in such case.
Based on the probed device type, piix4_add_adapters_sb800() or single
piix4_add_adapter() will be called.
For the former case, piix4_adapter_count is set as the number of adapters,
while for antoher case it is not set and kept default *zero*.
When piix4 is removed, piix4_remove() removes the adapters added in
piix4_probe(), basing on the piix4_adapter_count value.
Because the count is zero for the single adapter case, the adapter won't
be removed and makes the sources allocated for adapter leaked, such as
the i2c client and device.
These sources can still be accessed by i2c or bus and cause problems.
An easily reproduced case is that if a new adapter is registered, i2c
will get the leaked adapter and try to call smbus_algorithm, which was
already freed:
When thermnal zones are defined, trip points definitions are mandatory.
Define a couple of critical trip points for monitoring of existing
PMIC and SOC thermal zones.
This was lost between txt to yaml conversion and was re-enforced recently
via the commit 8c596324232d ("dt-bindings: thermal: Fix missing required property")
Cc: Rob Herring <robh+dt@kernel.org> Cc: Krzysztof Kozlowski <krzysztof.kozlowski+dt@linaro.org> Cc: devicetree@vger.kernel.org Signed-off-by: Cristian Marussi <cristian.marussi@arm.com> Fixes: f7b636a8d83c ("arm64: dts: juno: add thermal zones for scpi sensors") Link: https://lore.kernel.org/r/20221028140833.280091-8-cristian.marussi@arm.com Signed-off-by: Sudeep Holla <sudeep.holla@arm.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
memblock_reserve() expects a physical address, but the address being
passed for the TPM final events log is what was returned from
early_memremap(). This results in something like the following:
For asic with VF MMIO access protection avoid using CPU for VM table updates.
CPU pagetable updates have issues with HDP flush as VF MMIO access protection
blocks write to mmBIF_BX_DEV0_EPF0_VF0_HDP_MEM_COHERENCY_FLUSH_CNTL register
during sriov runtime.
v3: introduce virtualization capability flag AMDGPU_VF_MMIO_ACCESS_PROTECT
which indicates that VF MMIO write access is not allowed in sriov runtime
Signed-off-by: Danijel Slivka <danijel.slivka@amd.com> Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
Userspace can currently write to sysfs to transition sdev_state to RUNNING
or OFFLINE from any source state. This causes issues because proper
transitioning out of some states involves steps besides just changing
sdev_state, so allowing userspace to change sdev_state regardless of the
source state can result in inconsistencies; e.g. with ISCSI we can end up
with sdev_state == SDEV_RUNNING while the device queue is quiesced. Any
task attempting I/O on the device will then hang, and in more recent
kernels, iscsid will hang as well.
More detail about this bug is provided in my first attempt:
Change num_ghes from int to unsigned int, preventing an overflow
and causing subsequent vmalloc() to fail.
The overflow happens in ghes_estatus_pool_init() when calculating
len during execution of the statement below as both multiplication
operands here are signed int:
len += (num_ghes * GHES_ESOURCE_PREALLOC_MAX_SIZE);
The following call trace is observed because of this bug:
drivers/media/dvb-frontends/drxk_hard.c: In function 'drxk_read_ucblocks':
drivers/media/dvb-frontends/drxk_hard.c:6673:21: warning: 'err' may be used uninitialized [-Wmaybe-uninitialized]
6673 | *ucblocks = (u32) err;
| ^~~~~~~~~
drivers/media/dvb-frontends/drxk_hard.c:6663:13: note: 'err' was declared here
6663 | u16 err;
| ^~~
The local sd_fmt variable in rkisp1_capture_link_validate() has
uninitialized fields, which causes random failures when calling the
subdev .get_fmt() operation. Fix it by initializing the variable when
declaring it, which zeros all other fields.
Initialize the four color space fields on the sink and source video pads
of the resizer in the .init_cfg() operation. The resizer can't perform
any color space conversion, so set the sink and source color spaces to
the same defaults, which match the ISP source video pad default.
For optimal power consumption of USB4 routers the XHCI PCIe endpoint
used for tunneling must be in D3. Historically this is accomplished
by a long list of PCIe IDs that correspond to these endpoints because
the xhci_hcd driver will not default to allowing runtime PM for all
devices.
As both AMD and Intel have released new products with new XHCI controllers
this list continues to grow. In reviewing the XHCI specification v1.2 on
page 607 there is already a requirement that the PCI power management
states D3hot and D3cold must be supported.
In the quirk list, use this to indicate that runtime PM should be allowed
on XHCI controllers. The following controllers are known to be xHC 1.2 and
dropped explicitly:
* AMD Yellow Carp
* Intel Alder Lake
* Intel Meteor Lake
* Intel Raptor Lake
[keep PCI ID for Alder Lake PCH for recently added quirk -Mathias]
There is some code in the parser that tries to read 0x8000
bytes into a block to "read in the middle" of the block. Well
that only works if the block is also 0x10000 bytes all the time,
else we get these parse errors as we reach the end of the flash:
spi-nor spi0.0: mx25l1606e (2048 Kbytes)
mtd_read error while parsing (offset: 0x200000): -22
mtd_read error while parsing (offset: 0x201000): -22
(...)
When the text console is scrolling text upwards it calls the fillrect()
function to empty the new line. The current implementation doesn't seem
to work correctly on HCRX cards in 32-bit mode and leave garbage in that
line instead. Fix it by falling back to standard cfb_fillrect() in that
case.
The stifb driver (for Artist/HCRX graphics on PA-RISC) was missing
the fillrect function.
Tested on a 715/64 PA-RISC machine and in qemu.
Signed-off-by: Helge Deller <deller@gmx.de>
Stable-dep-of: 776d875fd4cb ("fbdev: stifb: Fall back to cfb_fillrect() on 32-bit HCRX cards") Signed-off-by: Sasha Levin <sashal@kernel.org>
Enhanced Strobe (ES) does not work correctly on the ASUS 1100 series of
devices. Jasper Lake eMMCs (pci_id 8086:4dc4) are supposed to support
ES. There are also two system families under the series, thus this is
being scoped to the ASUS BIOS.
The failing ES prevents the installer from writing to disk. Falling back
to HS400 without ES fixes the issue.
Signed-off-by: Patrick Thompson <ptf@google.com> Fixes: 315e3bd7ac19 ("mmc: sdhci-pci: Add support for Intel JSL") Acked-by: Adrian Hunter <adrian.hunter@intel.com> Cc: stable@vger.kernel.org Link: https://lore.kernel.org/r/20221013210017.3751025-1-ptf@google.com Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
The core issues the warning "drop HS400 support since no 8-bit bus" when
one of the ESDHC_FLAG_HS400* flags is set on a non 8bit capable host. To
avoid this warning set these flags only on hosts that actually can do
8bit, i.e. have bus-width = <8> set in the device tree.
Device-managed resources allocated post component bind must be tied to
the lifetime of the aggregate DRM device or they will not necessarily be
released when binding of the aggregate device is deferred.
This is specifically true for the HDMI IRQ, which will otherwise remain
requested so that the next bind attempt fails when requesting the IRQ a
second time.
Fix this by tying the device-managed lifetime of the HDMI IRQ to the DRM
device so that it is released when bind fails.
Quoting the header comments, IRQF_ONESHOT is "Used by threaded interrupts
which need to keep the irq line disabled until the threaded handler has
been run.". When applied to an interrupt that doesn't request a threaded
irq then IRQF_ONESHOT has a lesser known (undocumented?) side effect,
which it to disable the forced threading of irqs. For "normal" kernels
if there is no thread_fn then IRQF_ONESHOT is a nop.
In this case disabling forced threading is not appropriate because the
driver calls wake_up_all() (via msm_hdmi_i2c_irq) and also directly uses
the regular spinlock API for locking (in msm_hdmi_hdcp_irq() ). Neither
of these APIs can be called from no-thread interrupt handlers on
PREEMPT_RT systems.
During the initialization of ip6_route_net_init_late(), if file
ipv6_route or rt6_stats fails to be created, the initialization is
successful by default. Therefore, the ipv6_route or rt6_stats file
doesn't be found during the remove in ip6_route_net_exit_late(). It
will cause WRNING.
The following is the stack information:
name 'rt6_stats'
WARNING: CPU: 0 PID: 9 at fs/proc/generic.c:712 remove_proc_entry+0x389/0x460
Modules linked in:
Workqueue: netns cleanup_net
RIP: 0010:remove_proc_entry+0x389/0x460
PKRU: 55555554
Call Trace:
<TASK>
ops_exit_list+0xb0/0x170
cleanup_net+0x4ea/0xb00
process_one_work+0x9bf/0x1710
worker_thread+0x665/0x1080
kthread+0x2e4/0x3a0
ret_from_fork+0x1f/0x30
</TASK>
Fixes: cdb1876192db ("[NETNS][IPV6] route6 - create route6 proc files for the namespace") Signed-off-by: Zhengchao Shao <shaozhengchao@huawei.com> Reviewed-by: Eric Dumazet <edumazet@google.com> Link: https://lore.kernel.org/r/20221102020610.351330-1-shaozhengchao@huawei.com Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
Fix it by passing NULL to pneigh_queue_purge() in neigh_ifdown() if dev
is NULL, to make kernel not panic immediately.
Fixes: 66ba215cb513 ("neigh: fix possible DoS due to net iface start/stop loop") Signed-off-by: Chen Zhongjin <chenzhongjin@huawei.com> Reviewed-by: Eric Dumazet <edumazet@google.com> Reviewed-by: Denis V. Lunev <den@openvz.org> Link: https://lore.kernel.org/r/20221101121552.21890-1-chenzhongjin@huawei.com Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
Shifting signed 32-bit value by 31 bits is undefined, so changing
significant bit to unsigned. The UBSAN warning calltrace like below:
UBSAN: shift-out-of-bounds in drivers/net/phy/mdio_bus.c:586:27
left shift of 1 by 31 places cannot be represented in type 'int'
Call Trace:
<TASK>
dump_stack_lvl+0x7d/0xa5
dump_stack+0x15/0x1b
ubsan_epilogue+0xe/0x4e
__ubsan_handle_shift_out_of_bounds+0x1e7/0x20c
__mdiobus_register+0x49d/0x4e0
fixed_mdio_bus_init+0xd8/0x12d
do_one_initcall+0x76/0x430
kernel_init_freeable+0x3b3/0x422
kernel_init+0x24/0x1e0
ret_from_fork+0x1f/0x30
</TASK>
Fixes: 4fd5f812c23c ("phylib: allow incremental scanning of an mii bus") Signed-off-by: Gaosheng Cui <cuigaosheng1@huawei.com> Reviewed-by: Andrew Lunn <andrew@lunn.ch> Link: https://lore.kernel.org/r/20221031132645.168421-1-cuigaosheng1@huawei.com Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
When l2cap_recv_frame() is invoked to receive data, and the cid is
L2CAP_CID_A2MP, if the channel does not exist, it will create a channel.
However, after a channel is created, the hold operation of the channel
is not performed. In this case, the value of channel reference counting
is 1. As a result, after hci_error_reset() is triggered, l2cap_conn_del()
invokes the close hook function of A2MP to release the channel. Then
l2cap_chan_unlock(chan) will trigger UAF issue.
The process is as follows:
Receive data:
l2cap_data_channel()
a2mp_channel_create() --->channel ref is 2
l2cap_chan_put() --->channel ref is 1
Triger event:
hci_error_reset()
hci_dev_do_close()
...
l2cap_disconn_cfm()
l2cap_conn_del()
l2cap_chan_hold() --->channel ref is 2
l2cap_chan_del() --->channel ref is 1
a2mp_chan_close_cb() --->channel ref is 0, release channel
l2cap_chan_unlock() --->UAF of channel
The detailed Call Trace is as follows:
BUG: KASAN: use-after-free in __mutex_unlock_slowpath+0xa6/0x5e0
Read of size 8 at addr ffff8880160664b8 by task kworker/u11:1/7593
Workqueue: hci0 hci_error_reset
Call Trace:
<TASK>
dump_stack_lvl+0xcd/0x134
print_report.cold+0x2ba/0x719
kasan_report+0xb1/0x1e0
kasan_check_range+0x140/0x190
__mutex_unlock_slowpath+0xa6/0x5e0
l2cap_conn_del+0x404/0x7b0
l2cap_disconn_cfm+0x8c/0xc0
hci_conn_hash_flush+0x11f/0x260
hci_dev_close_sync+0x5f5/0x11f0
hci_dev_do_close+0x2d/0x70
hci_error_reset+0x9e/0x140
process_one_work+0x98a/0x1620
worker_thread+0x665/0x1080
kthread+0x2e4/0x3a0
ret_from_fork+0x1f/0x30
</TASK>
Last potentially related work creation:
kasan_save_stack+0x1e/0x40
__kasan_record_aux_stack+0xbe/0xd0
call_rcu+0x99/0x740
netlink_release+0xe6a/0x1cf0
__sock_release+0xcd/0x280
sock_close+0x18/0x20
__fput+0x27c/0xa90
task_work_run+0xdd/0x1a0
exit_to_user_mode_prepare+0x23c/0x250
syscall_exit_to_user_mode+0x19/0x50
do_syscall_64+0x42/0x80
entry_SYSCALL_64_after_hwframe+0x63/0xcd
Second to last potentially related work creation:
kasan_save_stack+0x1e/0x40
__kasan_record_aux_stack+0xbe/0xd0
call_rcu+0x99/0x740
netlink_release+0xe6a/0x1cf0
__sock_release+0xcd/0x280
sock_close+0x18/0x20
__fput+0x27c/0xa90
task_work_run+0xdd/0x1a0
exit_to_user_mode_prepare+0x23c/0x250
syscall_exit_to_user_mode+0x19/0x50
do_syscall_64+0x42/0x80
entry_SYSCALL_64_after_hwframe+0x63/0xcd
Fixes: d0be8347c623 ("Bluetooth: L2CAP: Fix use-after-free caused by l2cap_chan_put") Signed-off-by: Zhengchao Shao <shaozhengchao@huawei.com> Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
An SKB can be queued by the first flow and immediately dequeued and
freed by the second flow, therefore the callers of l2cap_reassemble_sdu
can't use the SKB after that function returns. However, some places
continue accessing struct l2cap_ctrl that resides in the SKB's CB for a
short time after l2cap_reassemble_sdu returns, leading to a
use-after-free condition (the stack trace is below, line numbers for
kernel 5.19.8).
Fix it by keeping a local copy of struct l2cap_ctrl.
BUG: KASAN: use-after-free in l2cap_rx_state_recv (net/bluetooth/l2cap_core.c:6906) bluetooth
Read of size 1 at addr ffff88812025f2f0 by task kworker/u17:3/43169
Workqueue: hci0 hci_rx_work [bluetooth]
Call Trace:
<TASK>
dump_stack_lvl (lib/dump_stack.c:107 (discriminator 4))
print_report.cold (mm/kasan/report.c:314 mm/kasan/report.c:429)
? l2cap_rx_state_recv (net/bluetooth/l2cap_core.c:6906) bluetooth
kasan_report (mm/kasan/report.c:162 mm/kasan/report.c:493)
? l2cap_rx_state_recv (net/bluetooth/l2cap_core.c:6906) bluetooth
l2cap_rx_state_recv (net/bluetooth/l2cap_core.c:6906) bluetooth
l2cap_rx (net/bluetooth/l2cap_core.c:7236 net/bluetooth/l2cap_core.c:7271) bluetooth
ret_from_fork (arch/x86/entry/entry_64.S:306)
</TASK>
Allocated by task 43169:
kasan_save_stack (mm/kasan/common.c:39)
__kasan_slab_alloc (mm/kasan/common.c:45 mm/kasan/common.c:436 mm/kasan/common.c:469)
kmem_cache_alloc_node (mm/slab.h:750 mm/slub.c:3243 mm/slub.c:3293)
__alloc_skb (net/core/skbuff.c:414)
l2cap_recv_frag (./include/net/bluetooth/bluetooth.h:425 net/bluetooth/l2cap_core.c:8329) bluetooth
l2cap_recv_acldata (net/bluetooth/l2cap_core.c:8442) bluetooth
hci_rx_work (net/bluetooth/hci_core.c:3642 net/bluetooth/hci_core.c:3832) bluetooth
process_one_work (kernel/workqueue.c:2289)
worker_thread (./include/linux/list.h:292 kernel/workqueue.c:2437)
kthread (kernel/kthread.c:376)
ret_from_fork (arch/x86/entry/entry_64.S:306)
In the test_no_shared_qgroup() and test_multiple_refs() qgroup self tests,
if we fail to add the tree ref, remove the extent item or remove the
extent ref, we are returning from the test function without freeing the
"old_roots" ulist that was allocated by the previous calls to
btrfs_find_all_roots(). Fix that by calling ulist_free() before returning.
Fixes: 442244c96332 ("btrfs: qgroup: Switch self test to extent-oriented qgroup mechanism.") Signed-off-by: Filipe Manana <fdmanana@suse.com> Signed-off-by: David Sterba <dsterba@suse.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
During backref walking, at find_parent_nodes(), if we are dealing with a
data extent and we get an error while resolving the indirect backrefs, at
resolve_indirect_refs(), or in the while loop that iterates over the refs
in the direct refs rbtree, we end up leaking the inode lists attached to
the direct refs we have in the direct refs rbtree that were not yet added
to the refs ulist passed as argument to find_parent_nodes(). Since they
were not yet added to the refs ulist and prelim_release() does not free
the lists, on error the caller can only free the lists attached to the
refs that were added to the refs ulist, all the remaining refs get their
inode lists never freed, therefore leaking their memory.
Fix this by having prelim_release() always free any attached inode list
to each ref found in the rbtree, and have find_parent_nodes() set the
ref's inode list to NULL once it transfers ownership of the inode list
to a ref added to the refs ulist passed to find_parent_nodes().
Fixes: 86d5f9944252 ("btrfs: convert prelimary reference tracking to use rbtrees") Signed-off-by: Filipe Manana <fdmanana@suse.com> Signed-off-by: David Sterba <dsterba@suse.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
During backref walking, at resolve_indirect_refs(), if we get an error
we jump to the 'out' label and call ulist_free() on the 'parents' ulist,
which frees all the elements in the ulist - however that does not free
any inode lists that may be attached to elements, through the 'aux' field
of a ulist node, so we end up leaking lists if we have any attached to
the unodes.
Fix this by calling free_leaf_list() instead of ulist_free() when we exit
from resolve_indirect_refs(). The static function free_leaf_list() is
moved up for this to be possible and it's slightly simplified by removing
unnecessary code.
Fixes: 3301958b7c1d ("Btrfs: add inodes before dropping the extent lock in find_all_leafs") Signed-off-by: Filipe Manana <fdmanana@suse.com> Signed-off-by: David Sterba <dsterba@suse.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
The class is set in mISDN_register_device(), but if device_add() returns
error, it will lead to delete a device without added, fix this by using
device_is_registered() to check if the device is registered.
Fixes: a900845e5661 ("mISDN: Add support for Traverse Technologies NETJet PCI cards") Signed-off-by: Yang Yingliang <yangyingliang@huawei.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sashal@kernel.org>
Afer commit 1fa5ae857bb1 ("driver core: get rid of struct device's
bus_id string array"), the name of device is allocated dynamically,
add put_device() to give up the reference, so that the name can be
freed in kobject_cleanup() when the refcount is 0.
Set device class before put_device() to avoid null release() function
WARN message in device_release().
Fixes: 1fa5ae857bb1 ("driver core: get rid of struct device's bus_id string array") Signed-off-by: Yang Yingliang <yangyingliang@huawei.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sashal@kernel.org>
KASAN: null-ptr-deref in range [0x0000000000000380-0x0000000000000387]
CPU: 0 PID: 4069 Comm: kworker/0:15 Not tainted 6.0.0-syzkaller-02734-g0326074ff465 #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 09/22/2022
Workqueue: rcu_gp srcu_invoke_callbacks
RIP: 0010:rose_send_frame+0x1dd/0x2f0 net/rose/rose_link.c:101
Call Trace:
<IRQ>
rose_transmit_clear_request+0x1d5/0x290 net/rose/rose_link.c:255
rose_rx_call_request+0x4c0/0x1bc0 net/rose/af_rose.c:1009
rose_loopback_timer+0x19e/0x590 net/rose/rose_loopback.c:111
call_timer_fn+0x1a0/0x6b0 kernel/time/timer.c:1474
expire_timers kernel/time/timer.c:1519 [inline]
__run_timers.part.0+0x674/0xa80 kernel/time/timer.c:1790
__run_timers kernel/time/timer.c:1768 [inline]
run_timer_softirq+0xb3/0x1d0 kernel/time/timer.c:1803
__do_softirq+0x1d0/0x9c8 kernel/softirq.c:571
[...]
</IRQ>
It triggers NULL pointer dereference when 'neigh->dev->dev_addr' is
called in the rose_send_frame(). It's the first occurrence of the
`neigh` is in rose_loopback_timer() as `rose_loopback_neigh', and
the 'dev' in 'rose_loopback_neigh' is initialized sa nullptr.
During the initialization of ip_vs_app_net_init(), if file ip_vs_app
fails to be created, the initialization is successful by default.
Therefore, the ip_vs_app file doesn't be found during the remove in
ip_vs_app_net_cleanup(). It will cause WRNING.
The following is the stack information:
name 'ip_vs_app'
WARNING: CPU: 1 PID: 9 at fs/proc/generic.c:712 remove_proc_entry+0x389/0x460
Modules linked in:
Workqueue: netns cleanup_net
RIP: 0010:remove_proc_entry+0x389/0x460
Call Trace:
<TASK>
ops_exit_list+0x125/0x170
cleanup_net+0x4ea/0xb00
process_one_work+0x9bf/0x1710
worker_thread+0x665/0x1080
kthread+0x2e4/0x3a0
ret_from_fork+0x1f/0x30
</TASK>
Fixes: 457c4cbc5a3d ("[NET]: Make /proc/net per network namespace") Signed-off-by: Zhengchao Shao <shaozhengchao@huawei.com> Acked-by: Julian Anastasov <ja@ssi.bg> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
During the initialization of ip_vs_conn_net_init(), if file ip_vs_conn
or ip_vs_conn_sync fails to be created, the initialization is successful
by default. Therefore, the ip_vs_conn or ip_vs_conn_sync file doesn't
be found during the remove.
The following is the stack information:
name 'ip_vs_conn_sync'
WARNING: CPU: 3 PID: 9 at fs/proc/generic.c:712
remove_proc_entry+0x389/0x460
Modules linked in:
Workqueue: netns cleanup_net
RIP: 0010:remove_proc_entry+0x389/0x460
Call Trace:
<TASK>
__ip_vs_cleanup_batch+0x7d/0x120
ops_exit_list+0x125/0x170
cleanup_net+0x4ea/0xb00
process_one_work+0x9bf/0x1710
worker_thread+0x665/0x1080
kthread+0x2e4/0x3a0
ret_from_fork+0x1f/0x30
</TASK>
The `char` type with no explicit sign is sometimes signed and sometimes
unsigned. This code will break on platforms such as arm, where char is
unsigned. So mark it here as explicitly signed, so that the
todrop_counter decrement and subsequent comparison is correct.
Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com> Acked-by: Julian Anastasov <ja@ssi.bg> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
No need to postpone this to the commit release path, since no packets
are walking over this object, this is accessed from control plane only.
This helped uncovered UAF triggered by races with the netlink notifier.
Recently, we got two syzkaller problems because of oversize packet
when napi frags enabled.
One of the problems is because the first seg size of the iov_iter
from user space is very big, it is 2147479538 which is bigger than
the threshold value for bail out early in __alloc_pages(). And
skb->pfmemalloc is true, __kmalloc_reserve() would use pfmemalloc
reserves without __GFP_NOWARN flag. Thus we got a warning as following:
The other problem is because odd IPv6 packets without NEXTHDR_NONE
extension header and have big packet length, it is 2127925 which is
bigger than ETH_MAX_MTU(65535). After ipv6_gso_pull_exthdrs() in
ipv6_gro_receive(), network_header offset and transport_header offset
are all bigger than U16_MAX. That would trigger skb->network_header
and skb->transport_header overflow error, because they are all '__u16'
type. Eventually, it would affect the value for __skb_push(skb, value),
and make it be a big value. After __skb_push() in ipv6_gro_receive(),
skb->data would less than skb->head, an out of bounds memory bug occurred.
That would trigger the problem as following:
To fix the problems, restrict the packet size less than
(ETH_MAX_MTU - NET_SKB_PAD - NET_IP_ALIGN) which has considered reserved
skb space in napi_alloc_skb() because transport_header is an offset from
skb->head. Add len check in tun_napi_alloc_frags() simply.
Fixes: 90e33d459407 ("tun: enable napi_gro_frags() for TUN/TAP driver") Signed-off-by: Ziyang Xuan <william.xuanziyang@huawei.com> Reviewed-by: Eric Dumazet <edumazet@google.com> Link: https://lore.kernel.org/r/20221029094101.1653855-1-william.xuanziyang@huawei.com Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
We can't use "skb" again after passing it to qdisc_enqueue(). This is
basically identical to commit 2f09707d0c97 ("sch_sfb: Also store skb
len before calling child enqueue").
Fixes: d7f4f332f082 ("sch_red: update backlog as well") Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> Reviewed-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sashal@kernel.org>
Clang gives a warning when compiling pata_legacy.c with 'make W=1' about
the 'rt' local variable in pdc20230_set_piomode() being set but unused.
Quite obviously, there is an outb() call missing to write back the updated
variable. Moreover, checking the docs by Petr Soucek revealed that bitwise
AND should have been done with a negated timing mask and the master/slave
timing masks were swapped while updating...
Fixes: 669a5db411d8 ("[libata] Add a bunch of PATA drivers.") Reported-by: Damien Le Moal <damien.lemoal@opensource.wdc.com> Signed-off-by: Sergey Shtylyov <s.shtylyov@omp.ru> Signed-off-by: Damien Le Moal <damien.lemoal@opensource.wdc.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
nfcmrvl_i2c_nci_send() will be called by nfcmrvl_nci_send(), and skb
should be freed in nfcmrvl_i2c_nci_send(). However, nfcmrvl_nci_send()
will only free skb when i2c_master_send() return >=0, which means skb
will memleak when i2c_master_send() failed. Free skb no matter whether
i2c_master_send() succeeds.
Fixes: b5b3e23e4cac ("NFC: nfcmrvl: add i2c driver") Signed-off-by: Shang XiaoJing <shangxiaojing@huawei.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sashal@kernel.org>
s3fwrn5_nci_send() will call s3fwrn5_i2c_write() or s3fwrn82_uart_write(),
and free the skb if write() failed. However, even if the write() run
succeeds, the skb will not be freed in write(). As the result, the skb
will memleak. s3fwrn5_nci_send() should also free the skb when write()
succeeds.
Fixes: c04c674fadeb ("nfc: s3fwrn5: Add driver for Samsung S3FWRN5 NFC Chip") Signed-off-by: Shang XiaoJing <shangxiaojing@huawei.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sashal@kernel.org>
nxp_nci_send() will call nxp_nci_i2c_write(), and only free skb when
nxp_nci_i2c_write() failed. However, even if the nxp_nci_i2c_write()
run succeeds, the skb will not be freed in nxp_nci_i2c_write(). As the
result, the skb will memleak. nxp_nci_send() should also free the skb
when nxp_nci_i2c_write() succeeds.
Fixes: dece45855a8b ("NFC: nxp-nci: Add support for NXP NCI chips") Signed-off-by: Shang XiaoJing <shangxiaojing@huawei.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sashal@kernel.org>
fdp_nci_send() will call fdp_nci_i2c_write that will not free skb in
the function. As a result, when fdp_nci_i2c_write() finished, the skb
will memleak. fdp_nci_send() should free skb after fdp_nci_i2c_write()
finished.
Fixes: a06347c04c13 ("NFC: Add Intel Fields Peak NFC solution driver") Signed-off-by: Shang XiaoJing <shangxiaojing@huawei.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sashal@kernel.org>
Now that the kernel has ftrace, any debugging calls that just do "made
it to this function!" and "leaving this function!" can be removed.
Better to use standard debugging tools.
This allows also to remove several local variables and entire
fdp_nci_recv_frame() function (whose purpose was only to log).
There are two reasons for memleak in dsa_loop_init().
First, fixed_phy_register() create and register phy_device:
fixed_phy_register()
get_phy_device()
phy_device_create() # freed by phy_device_free()
phy_device_register() # freed by phy_device_remove()
But fixed_phy_unregister() only calls phy_device_remove().
So the memory allocated in phy_device_create() is leaked.
Second, when mdio_driver_register() fail in dsa_loop_init(),
it just returns and there is no cleanup for phydevs.
Fix the problems by catching the error of mdio_driver_register()
in dsa_loop_init(), then calling both fixed_phy_unregister() and
phy_device_free() to release phydevs.
Also add a function for phydevs cleanup to avoid duplacate.
Fixes: 98cd1552ea27 ("net: dsa: Mock-up driver") Signed-off-by: Chen Zhongjin <chenzhongjin@huawei.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sashal@kernel.org>
Currently, we are only guaranteed to send RECLAIM_COMPLETE if we have
open state to recover. Fix the client to always send RECLAIM_COMPLETE
after setting up the lease.
If the server reboots while we are engaged in a delegation return, and
there is a pNFS layout with return-on-close set, then the current code
can end up deadlocking in pnfs_roc() when nfs_inode_set_delegation()
tries to return the old delegation.
Now that delegreturn actually uses its own copy of the stateid, it
should be safe to just always update the delegation stateid in place.
Fixes: 078000d02d57 ("pNFS: We want return-on-close to complete when evicting the inode") Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com> Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
Commit 13bac861952a ("IB/hfi1: Fix abba locking issue with sc_disable()")
incorrectly tries to move a list from one list head to another. The
result is a kernel crash.
The crash is triggered when a link goes down and there are waiters for a
send to complete. The following signature is seen:
Commit 27cfde795a96 ("RDMA/cma: Fix arguments order in net device
validation") swapped the src and dst addresses in the call to
validate_net_dev().
As a consequence, the test in validate_ipv4_net_dev() to see if the
net_dev is the right one, is incorrect for port 1 <-> 2 communication when
the ports are on the same sub-net. This is fixed by denoting the
flowi4_oif as the device instead of the incoming one.
The bug has not been observed using IPv6 addresses.
Fixes: 27cfde795a96 ("RDMA/cma: Fix arguments order in net device validation") Signed-off-by: Håkon Bugge <haakon.bugge@oracle.com> Link: https://lore.kernel.org/r/20221012141542.16925-1-haakon.bugge@oracle.com Reviewed-by: Leon Romanovsky <leonro@nvidia.com> Signed-off-by: Leon Romanovsky <leon@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
The KVM_X86_SET_MSR_FILTER ioctls contains a pointer in the passed in
struct which means it has a different struct size depending on whether
it gets called from 32bit or 64bit code.
This patch introduces compat code that converts from the 32bit struct to
its 64bit counterpart which then gets used going forward internally.
With this applied, 32bit QEMU can successfully set MSR bitmaps when
running on 64bit kernels.
Reported-by: Andrew Randrianasulu <randrianasulu@gmail.com> Fixes: 1a155254ff937 ("KVM: x86: Introduce MSR filtering") Signed-off-by: Alexander Graf <graf@amazon.com>
Message-Id: <20221017184541.2658-4-graf@amazon.com> Cc: stable@vger.kernel.org Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
In the next patch we want to introduce a second caller to
set_msr_filter() which constructs its own filter list on the stack.
Refactor the original function so it takes it as argument instead of
reading it through copy_from_user().
Signed-off-by: Alexander Graf <graf@amazon.com>
Message-Id: <20221017184541.2658-3-graf@amazon.com> Cc: stable@vger.kernel.org Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
The flags for KVM_CAP_X86_USER_SPACE_MSR and KVM_X86_SET_MSR_FILTER
have no protection for their unused bits. Without protection, future
development for these features will be difficult. Add the protection
needed to make it possible to extend these features in the future.
Today, core ID is assumed to be unique within each package.
But an AlderLake-N platform adds a Module level between core and package,
Linux excludes the unknown modules bits from the core ID, resulting in
duplicate core ID's.
To keep core ID unique within a package, Linux must include all APIC-ID
bits for known or unknown levels above the core and below the package
in the core ID.
It is important to understand that core ID's have always come directly
from the APIC-ID encoding, which comes from the BIOS. Thus there is no
guarantee that they start at 0, or that they are contiguous.
As such, naively using them for array indexes can be problematic.
[ dhansen: un-known -> unknown ]
Fixes: 7745f03eb395 ("x86/topology: Add CPUID.1F multi-die/package support") Suggested-by: Len Brown <len.brown@intel.com> Signed-off-by: Zhang Rui <rui.zhang@intel.com> Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com> Reviewed-by: Len Brown <len.brown@intel.com> Cc: stable@vger.kernel.org Link: https://lkml.kernel.org/r/20221014090147.1836-5-rui.zhang@intel.com Signed-off-by: Sasha Levin <sashal@kernel.org>
CPUID.1F/B does not enumerate Package level explicitly, instead, all the
APIC-ID bits above the enumerated levels are assumed to be package ID
bits.
Current code gets package ID by shifting out all the APIC-ID bits that
Linux supports, rather than shifting out all the APIC-ID bits that
CPUID.1F enumerates. This introduces problems when CPUID.1F enumerates a
level that Linux does not support.
For example, on a single package AlderLake-N, there are 2 Ecore Modules
with 4 atom cores in each module. Linux does not support the Module
level and interprets the Module ID bits as package ID and erroneously
reports a multi module system as a multi-package system.
Fix this by using APIC-ID bits above all the CPUID.1F enumerated levels
as package ID.
[ dhansen: spelling fix ]
Fixes: 7745f03eb395 ("x86/topology: Add CPUID.1F multi-die/package support") Suggested-by: Len Brown <len.brown@intel.com> Signed-off-by: Zhang Rui <rui.zhang@intel.com> Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com> Reviewed-by: Len Brown <len.brown@intel.com> Cc: stable@vger.kernel.org Link: https://lkml.kernel.org/r/20221014090147.1836-4-rui.zhang@intel.com Signed-off-by: Sasha Levin <sashal@kernel.org>
CPUID Leaf 0x1F defines a DIE_TYPE level (nb: ECX[8:15] level type == 0x5),
but CPUID Leaf 0xB does not. However, detect_extended_topology() will
set struct cpuinfo_x86.cpu_die_id regardless of whether a valid Die ID
was found.
Only set cpu_die_id if a DIE_TYPE level is found. CPU topology code may
use another value for cpu_die_id, e.g. the AMD NodeId on AMD-based
systems. Code ordering should be maintained so that the CPUID Leaf 0x1F
Die ID value will take precedence on systems that may use another value.
Add a dedicated "exception type" for #DBs, as #DBs can be fault-like or
trap-like depending the sub-type of #DB, and effectively defer the
decision of what to do with the #DB to the caller.
For the emulator's two calls to exception_type(), treat the #DB as
fault-like, as the emulator handles only code breakpoint and general
detect #DBs, both of which are fault-like.
For event injection, which uses exception_type() to determine whether to
set EFLAGS.RF=1 on the stack, keep the current behavior of not setting
RF=1 for #DBs. Intel and AMD explicitly state RF isn't set on code #DBs,
so exempting by failing the "== EXCPT_FAULT" check is correct. The only
other fault-like #DB is General Detect, and despite Intel and AMD both
strongly implying (through omission) that General Detect #DBs should set
RF=1, hardware (multiple generations of both Intel and AMD), in fact does
not. Through insider knowledge, extreme foresight, sheer dumb luck, or
some combination thereof, KVM correctly handled RF for General Detect #DBs.
Fixes: 38827dbd3fb8 ("KVM: x86: Do not update EFLAGS on faulting emulation") Cc: stable@vger.kernel.org Signed-off-by: Sean Christopherson <seanjc@google.com> Reviewed-by: Maxim Levitsky <mlevitsk@redhat.com> Link: https://lore.kernel.org/r/20220830231614.3580124-9-seanjc@google.com Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
Trace exceptions that are re-injected, not just those that KVM is
injecting for the first time. Debugging re-injection bugs is painful
enough as is, not having visibility into what KVM is doing only makes
things worse.
Delay propagating pending=>injected in the non-reinjection path so that
the tracing can properly identify reinjected exceptions.
Signed-off-by: Sean Christopherson <seanjc@google.com> Reviewed-by: Maxim Levitsky <mlevitsk@redhat.com> Signed-off-by: Maciej S. Szmigiero <maciej.szmigiero@oracle.com>
Message-Id: <25470690a38b4d2b32b6204875dd35676c65c9f2.1651440202.git.maciej.szmigiero@oracle.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Stable-dep-of: 5623f751bd9c ("KVM: x86: Treat #DBs from the emulator as fault-like (code and DR7.GD=1)") Signed-off-by: Sasha Levin <sashal@kernel.org>
Don't propagate vmcs12's VM_ENTRY_LOAD_IA32_PERF_GLOBAL_CTRL to vmcs02.
KVM doesn't disallow L1 from using VM_ENTRY_LOAD_IA32_PERF_GLOBAL_CTRL
even when KVM itself doesn't use the control, e.g. due to the various
CPU errata that where the MSR can be corrupted on VM-Exit.
Preserve KVM's (vmcs01) setting to hopefully avoid having to toggle the
bit in vmcs02 at a later point. E.g. if KVM is loading PERF_GLOBAL_CTRL
when running L1, then odds are good KVM will also load the MSR when
running L2.
Fixes: 8bf00a529967 ("KVM: VMX: add support for switching of PERF_GLOBAL_CTRL") Cc: stable@vger.kernel.org Signed-off-by: Sean Christopherson <seanjc@google.com> Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com> Link: https://lore.kernel.org/r/20220830133737.1539624-18-vkuznets@redhat.com Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
When preparing controls for vmcs02, grab KVM's desired controls from
vmcs01's shadow state instead of recalculating the controls from scratch,
or in the secondary execution controls, instead of using the dedicated
cache. Calculating secondary exec controls is eye-poppingly expensive
due to the guest CPUID checks, hence the dedicated cache, but the other
calculations aren't exactly free either.
Explicitly clear several bits (x2APIC, DESC exiting, and load EFER on
exit) as appropriate as they may be set in vmcs01, whereas the previous
implementation relied on dynamic bits being cleared in the calculator.
Intentionally propagate VM_{ENTRY,EXIT}_LOAD_IA32_PERF_GLOBAL_CTRL from
vmcs01 to vmcs02. Whether or not PERF_GLOBAL_CTRL is loaded depends on
whether or not perf itself is active, so unless perf stops between the
exit from L1 and entry to L2, vmcs01 will hold the desired value. This
is purely an optimization as atomic_switch_perf_msrs() will set/clear
the control as needed at VM-Enter, i.e. it avoids two extra VMWRITEs in
the case where perf is active (versus starting with the bits clear in
vmcs02, which was the previous behavior).
Cc: Zeng Guang <guang.zeng@intel.com> Signed-off-by: Sean Christopherson <seanjc@google.com>
Message-Id: <20210810171952.2758100-3-seanjc@google.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Stable-dep-of: def9d705c05e ("KVM: nVMX: Don't propagate vmcs12's PERF_GLOBAL_CTRL settings to vmcs02") Signed-off-by: Sasha Levin <sashal@kernel.org>
A SERIAL_8250_16550A_VARIANTS configuration option has been recently
defined that lets one request the 8250 driver not to probe for 16550A
device features so as to reduce the driver's device startup time in
virtual machines.
Some actual hardware devices require these features to have been fully
determined however for their driver to work correctly, so define a flag
to let drivers request full 16550A feature probing on a device-by-device
basis if required regardless of the SERIAL_8250_16550A_VARIANTS option
setting chosen.
Fixes: dc56ecb81a0a ("serial: 8250: Support disabling mdelay-filled probes of 16550A variants") Cc: stable@vger.kernel.org # v5.6+ Reported-by: Anders Blomdell <anders.blomdell@control.lth.se> Signed-off-by: Maciej W. Rozycki <macro@orcam.me.uk> Link: https://lore.kernel.org/r/alpine.DEB.2.21.2209202357520.41633@angie.orcam.me.uk Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
When a UART port is newly registered, uart_configure_port() seeks to
deassert RS485 Transmit Enable by setting the RTS bit in port->mctrl.
However a number of UART drivers interpret a set RTS bit as *assertion*
instead of deassertion: Affected drivers include those using
serial8250_em485_config() (except 8250_bcm2835aux.c) and some using
mctrl_gpio (e.g. imx.c).
Since the interpretation of the RTS bit is driver-specific, it is not
suitable as a means to centrally deassert Transmit Enable in the serial
core. Instead, the serial core must call on drivers to deassert it in
their driver-specific way. One way to achieve that is to call
->rs485_config(). It implicitly deasserts Transmit Enable.
So amend uart_configure_port() and uart_resume_port() to invoke
uart_rs485_config(). That allows removing calls to uart_rs485_config()
from drivers' ->probe() hooks and declaring the function static.
Skip any invocation of ->set_mctrl() if RS485 is enabled. RS485 has no
hardware flow control, so the modem control lines are irrelevant and
need not be touched. When leaving RS485 mode, reset the modem control
lines to the state stored in port->mctrl. That way, UARTs which are
muxed between RS485 and RS232 transceivers drive the lines correctly
when switched to RS232. (serial8250_do_startup() historically raises
the OUT1 modem signal because otherwise interrupts are not signaled on
ancient PC UARTs, but I believe that no longer applies to modern,
RS485-capable UARTs and is thus safe to be skipped.)
imx.c modifies port->mctrl whenever Transmit Enable is asserted and
deasserted. Stop it from doing that so port->mctrl reflects the RS232
line state.
8250_omap.c deasserts Transmit Enable on ->runtime_resume() by calling
->set_mctrl(). Because that is now a no-op in RS485 mode, amend the
function to call serial8250_em485_stop_tx().
fsl_lpuart.c retrieves and applies the RS485 device tree properties
after registering the UART port. Because applying now happens on
registration in uart_configure_port(), move retrieval of the properties
ahead of uart_add_one_port().
Several drivers that support setting the RS485 configuration via userspace
implement one or more of the following tasks:
- in case of an invalid RTS configuration (both RTS after send and RTS on
send set or both unset) fall back to enable RTS on send and disable RTS
after send
- nullify the padding field of the returned serial_rs485 struct
- copy the configuration into the uart port struct
- limit RTS delays to 100 ms
Move these tasks into the serial core to make them generic and to provide
a consistent behaviour among all drivers.
We are seeing an IRQ storm on the global receive IRQ line under heavy
CAN bus load conditions with both CAN channels enabled.
Conditions:
The global receive IRQ line is shared between can0 and can1, either of
the channels can trigger interrupt while the other channel's IRQ line
is disabled (RFIE).
When global a receive IRQ interrupt occurs, we mask the interrupt in
the IRQ handler. Clearing and unmasking of the interrupt is happening
in rx_poll(). There is a race condition where rx_poll() unmasks the
interrupt, but the next IRQ handler does not mask the IRQ due to
NAPIF_STATE_MISSED flag (e.g.: can0 RX FIFO interrupt is disabled and
can1 is triggering RX interrupt, the delay in rx_poll() processing
results in setting NAPIF_STATE_MISSED flag) leading to an IRQ storm.
This patch fixes the issue by checking IRQ active and enabled before
handling the IRQ on a particular channel.
Fixes: dd3bd23eb438 ("can: rcar_canfd: Add Renesas R-Car CAN FD driver") Suggested-by: Marc Kleine-Budde <mkl@pengutronix.de> Signed-off-by: Biju Das <biju.das.jz@bp.renesas.com> Link: https://lore.kernel.org/all/20221025155657.1426948-2-biju.das.jz@bp.renesas.com Cc: stable@vger.kernel.org
[mkl: adjust commit message] Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
[biju: removed gpriv from RCANFD_RFCC_RFIE macro] Signed-off-by: Biju Das <biju.das.jz@bp.renesas.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The commit 26f55386f964 ("arm64/mm: Fix __enable_mmu() for new TGRAN range
values") had already switched into testing ID_AA64MMFR0_TGRAN range values.
This just changes system_supports_[4|16|64]kb_granule() helpers to perform
similar range tests as well. While here, it standardizes page size specific
supported min and max TGRAN values.
As per ARM ARM DDI 0487G.a, when FEAT_LPA2 is implemented, ID_AA64MMFR0_EL1
might contain a range of values to describe supported translation granules
(4K and 16K pages sizes in particular) instead of just enabled or disabled
values. This changes __enable_mmu() function to handle complete acceptable
range of values (depending on whether the field is signed or unsigned) now
represented with ID_AA64MMFR0_TGRAN_SUPPORTED_[MIN..MAX] pair. While here,
also fix similar situations in EFI stub and KVM as well.
Cc: Catalin Marinas <catalin.marinas@arm.com> Cc: Will Deacon <will@kernel.org> Cc: Marc Zyngier <maz@kernel.org> Cc: James Morse <james.morse@arm.com> Cc: Suzuki K Poulose <suzuki.poulose@arm.com> Cc: Ard Biesheuvel <ardb@kernel.org> Cc: Mark Rutland <mark.rutland@arm.com> Cc: linux-arm-kernel@lists.infradead.org Cc: kvmarm@lists.cs.columbia.edu Cc: linux-efi@vger.kernel.org Cc: linux-kernel@vger.kernel.org Acked-by: Marc Zyngier <maz@kernel.org> Signed-off-by: James Morse <james.morse@arm.com> Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com> Link: https://lore.kernel.org/r/1615355590-21102-1-git-send-email-anshuman.khandual@arm.com Signed-off-by: Will Deacon <will@kernel.org> Signed-off-by: Zenghui Yu <yuzenghui@huawei.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Because it introduces a problem that rq->__data_len is set to the wrong
value.
before the patch:
1) nr_bytes = rq->__data_len
2) rq->__data_len = sdp->sector_size
3) scsi_init_io()
4) rq->__data_len = nr_bytes
after the patch:
1) rq->__data_len = sdp->sector_size
2) scsi_init_io()
3) rq->__data_len = rq->__data_len -> __data_len is wrong
It will cause that io can only complete one segment each time, and the io
will requeue in scsi_io_completion_action(), which will cause severe
performance degradation.
Scsi write same is removed in commit e383e16e84e9 ("scsi: sd: Remove
WRITE_SAME support") from mainline, hence this patch is only needed for
stable kernels.
Fixes: 84f7a9de0602 ("scsi: sd: Remove a local variable") Signed-off-by: Yu Kuai <yukuai3@huawei.com> Reviewed-by: Bart Van Assche <bvanassche@acm.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Per AmpereOne erratum AC03_CPU_12, "Branch history may allow control of
speculative execution across software contexts," the AMPERE1 core needs the
bhb clearing loop to mitigate Spectre-BHB, with a loop iteration count of
11.
Under memory pressure, enetc_refill_rx_ring() may fail, and when called
during the enetc_open() -> enetc_setup_rxbdr() procedure, this is not
checked for.
An extreme case of memory pressure will result in exactly zero buffers
being allocated for the RX ring, and in such a case it is expected that
hardware drops all RX packets due to lack of buffers.
This does not happen, because the reset-default value of the consumer
and produces index is 0, and this makes the ENETC think that all buffers
have been initialized and that it owns them (when in reality none were).
The hardware guide explains this best:
| Configure the receive ring producer index register RBaPIR with a value
| of 0. The producer index is initially configured by software but owned
| by hardware after the ring has been enabled. Hardware increments the
| index when a frame is received which may consume one or more BDs.
| Hardware is not allowed to increment the producer index to match the
| consumer index since it is used to indicate an empty condition. The ring
| can hold at most RBLENR[LENGTH]-1 received BDs.
|
| Configure the receive ring consumer index register RBaCIR. The
| consumer index is owned by software and updated during operation of the
| of the BD ring by software, to indicate that any receive data occupied
| in the BD has been processed and it has been prepared for new data.
| - If consumer index and producer index are initialized to the same
| value, it indicates that all BDs in the ring have been prepared and
| hardware owns all of the entries.
| - If consumer index is initialized to producer index plus N, it would
| indicate N BDs have been prepared. Note that hardware cannot start if
| only a single buffer is prepared due to the restrictions described in
| (2).
| - Software may write consumer index to match producer index anytime
| while the ring is operational to indicate all received BDs prior have
| been processed and new BDs prepared for hardware.
Normally, the value of rx_ring->rcir (consumer index) is brought in sync
with the rx_ring->next_to_use software index, but this only happens if
page allocation ever succeeded.
When PI==CI==0, the hardware appears to receive frames and write them to
DMA address 0x0 (?!), then set the READY bit in the BD.
The enetc_clean_rx_ring() function (and its XDP derivative) is naturally
not prepared to handle such a condition. It will attempt to process
those frames using the rx_swbd structure associated with index i of the
RX ring, but that structure is not fully initialized (enetc_new_page()
does all of that). So what happens next is undefined behavior.
To operate using no buffer, we must initialize the CI to PI + 1, which
will block the hardware from advancing the CI any further, and drop
everything.
The issue was seen while adding support for zero-copy AF_XDP sockets,
where buffer memory comes from user space, which can even decide to
supply no buffers at all (example: "xdpsock --txonly"). However, the bug
is present also with the network stack code, even though it would take a
very determined person to trigger a page allocation failure at the
perfect time (a series of ifup/ifdown under memory pressure should
eventually reproduce it given enough retries).
Fixes: d4fd0404c1c9 ("enetc: Introduce basic PF and VF ENETC ethernet drivers") Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Claudiu Manoil <claudiu.manoil@nxp.com> Link: https://lore.kernel.org/r/20221027182925.3256653-1-vladimir.oltean@nxp.com Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
When setting Bluefield to DPU NIC mode using mlxconfig tool + sync
firmware reset flow, we run into scenario where the host was not
eswitch manager at the time of mlx5 driver load but becomes eswitch manager
after the sync firmware reset flow. This results in null pointer
access of mpfs structure during mac filter add. This change prevents null
pointer access but mpfs table entries will not be added.
Fixes: 5ec697446f46 ("net/mlx5: Add support for devlink reload action fw activate") Signed-off-by: Suresh Devarakonda <ramad@nvidia.com> Reviewed-by: Moshe Shemesh <moshe@nvidia.com> Reviewed-by: Bodong Wang <bodong@nvidia.com> Signed-off-by: Saeed Mahameed <saeedm@nvidia.com> Link: https://lore.kernel.org/r/20221026135153.154807-12-saeed@kernel.org Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
mlx5_cmd_cleanup_async_ctx should return only after all its callback
handlers were completed. Before this patch, the below race between
mlx5_cmd_cleanup_async_ctx and mlx5_cmd_exec_cb_handler was possible and
lead to a use-after-free:
1. mlx5_cmd_cleanup_async_ctx is called while num_inflight is 2 (i.e.
elevated by 1, a single inflight callback).
2. mlx5_cmd_cleanup_async_ctx decreases num_inflight to 1.
3. mlx5_cmd_exec_cb_handler is called, decreases num_inflight to 0 and
is about to call wake_up().
4. mlx5_cmd_cleanup_async_ctx calls wait_event, which returns
immediately as the condition (num_inflight == 0) holds.
5. mlx5_cmd_cleanup_async_ctx returns.
6. The caller of mlx5_cmd_cleanup_async_ctx frees the mlx5_async_ctx
object.
7. mlx5_cmd_exec_cb_handler goes on and calls wake_up() on the freed
object.
Fix it by syncing using a completion object. Mark it completed when
num_inflight reaches 0.
Trace:
BUG: KASAN: use-after-free in do_raw_spin_lock+0x23d/0x270
Read of size 4 at addr ffff888139cd12f4 by task swapper/5/0
An offloaded SA stops receiving after about 2^32 + replay_window
packets. For example, when SA reaches <seq-hi 0x1, seq 0x2c>, all
subsequent packets get dropped with SA-icv-failure (integrity_failed).
To reproduce the bug:
- ConnectX-6 Dx with crypto enabled (FW 22.30.1004)
- ipsec.conf:
nic-offload = yes
replay-window = 32
esn = yes
salifetime=24h
- Run netperf for a long time to send more than 2^32 packets
netperf -H <device-under-test> -t TCP_STREAM -l 20000
When 2^32 + replay_window packets are received, the replay window
moves from the 2nd half of subspace (overlap=1) to the 1st half
(overlap=0). The driver then updates the 'esn' value in NIC
(i.e. seq_hi) as follows.
seq_hi = xfrm_replay_seqhi(seq_bottom)
new esn in NIC = seq_hi + 1
The +1 increment is wrong, as seq_hi already contains the correct
seq_hi. For example, when seq_hi=1, the driver actually tells NIC to
use seq_hi=2 (esn). This incorrect esn value causes all subsequent
packets to fail integrity checks (SA-icv-failure). So, do not
increment.
Fixes: cb01008390bb ("net/mlx5: IPSec, Add support for ESN") Signed-off-by: Hyong Youb Kim <hyonkim@cisco.com> Acked-by: Leon Romanovsky <leonro@nvidia.com> Signed-off-by: Saeed Mahameed <saeedm@nvidia.com> Link: https://lore.kernel.org/r/20221026135153.154807-2-saeed@kernel.org Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
As explained by Julian, fib_nh_scope is related to fib_nh_gw4, but
fib_info_update_nhc_saddr() needs the scope of the route, which is
the scope "before" fib_nh_scope, ie fib_nh_scope - 1.
This patch fixes the problem described in commit 747c14307214 ("ip: fix
dflt addr selection for connected nexthop").
Fixes: 597cfe4fc339 ("nexthop: Add support for IPv4 nexthops") Link: https://lore.kernel.org/netdev/6c8a44ba-c2d5-cdf-c5c7-5baf97cba38@ssi.bg/ Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Reviewed-by: Julian Anastasov <ja@ssi.bg> Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
If of_device_register() returns error, the of node and the
name allocated in dev_set_name() is leaked, call put_device()
to give up the reference that was set in device_initialize(),
so that of node is put in logical_port_release() and the name
is freed in kobject_cleanup().
As noted by Paolo Abeni, pr_warn doesn't generate any splat and can still
preserve the warning to the user that feature downgrade occurred. We
likely cannot introduce other kinds of checks / enforcement here because
syzbot can generate different genl versions to the datapath.
Reported-by: syzbot+31cde0bef4bbf8ba2d86@syzkaller.appspotmail.com Fixes: 44da5ae5fbea ("openvswitch: Drop user features if old user space attempted to create datapath") Cc: Thomas Graf <tgraf@suug.ch> Signed-off-by: Aaron Conole <aconole@redhat.com> Acked-by: Ilya Maximets <i.maximets@ovn.org> Signed-off-by: Paolo Abeni <pabeni@redhat.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
projects / users / dwmw2 / linux.git / log