Replace deprecated one-element arrays with flexible array members.
CONFIG_FORTIFY_SOURCE reports the following warning when
one-element arrays are used as variable-length buffers:
sta_cmd.c:1033 mwifiex_sta_prepare_cmd
memcpy: detected field-spanning write (size 84) of single field
"domain->triplet" at .../marvell/mwifiex/sta_cmd.c:1033 (size 3)
Convert affected structs to use flexible array members.
- Preserve existing wire layouts.
- Use DECLARE_FLEX_ARRAY() for structs inside affected unions.
Tested-on: WRT3200ACM, OpenWrt
Signed-off-by: Georgi Valkov <gvalkov@gmail.com>
Reviewed-by: Francesco Dolcini <francesco.dolcini@toradex.com>
Link: https://patch.msgid.link/20260716001728.57799-1-gvalkov@gmail.com
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
at76_guess_freq() checks only that the received frame is at least a bare
802.11 header (24 bytes) before subtracting the fixed management-body
offset:
len -= el_off;
For both beacon and probe response frames, el_off is 36. If the frame is
shorter than el_off, subtracting it causes the calculated IE length to
wrap. The length is eventually passed to cfg80211_find_elem_match() as a
very large unsigned value, so the element walk runs beyond the RX skb.
This path is reached from at76_rx_tasklet() while scanning. If the device
delivers a truncated beacon or probe response, the oversized IE length
causes an out-of-bounds read during scanning.
Skip the IE lookup if the frame does not reach the variable elements,
before subtracting el_off.
Fixes: 1264b95146 ("at76c50x-usb: add driver")
Cc: stable@vger.kernel.org
Assisted-by: Claude:claude-opus-4-8
Signed-off-by: Huihui Huang <hhhuang@smu.edu.sg>
Link: https://patch.msgid.link/20260715140815.1242033-1-hhhuang@smu.edu.sg
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Hardware implementations report Scalable-Mode Page-walk Coherency Support
via the SMPWCS field in the extended capability register. If the hardware
does not support page-walk coherency, a clflush is required every time
the page table entries (which are walked by the IOMMU hardware) are
updated.
In the SVA case, page tables are managed by the CPU mm core, not by the
IOMMU driver. Because the IOMMU driver has no way of knowing whether the
CPU page table management code has ensured coherency via clflush, the
driver must deny SVA if the hardware does not support coherent paging.
Fixes: ff3dc6521f ("iommu/vt-d: Fix CPU and IOMMU SVM feature matching checks")
Cc: stable@vger.kernel.org
Signed-off-by: Lu Baolu <baolu.lu@linux.intel.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Samiullah Khawaja <skhawaja@google.com>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Will Deacon <will@kernel.org>
mwifiex_process_uap_event() handles EVENT_UAP_STA_ASSOC by exposing the
(re)association request IEs that the firmware copies into the event:
sinfo->assoc_req_ies = &event->data[len];
len = (u8 *)sinfo->assoc_req_ies - (u8 *)&event->frame_control;
sinfo->assoc_req_ies_len = le16_to_cpu(event->len) - (u16)len;
event->len is supplied by the device firmware and is never validated,
and the subtraction is unchecked. assoc_req_ies points into
adapter->event_body[MAX_EVENT_SIZE], a fixed-size array embedded in the
kmalloc()'d struct mwifiex_adapter.
On the ap_11n_enabled path mwifiex_set_sta_ht_cap() walks these IEs with
cfg80211_find_ie(), whose for_each_element() loop dereferences each
element header. A firmware-reported event->len larger than the bytes
actually received makes assoc_req_ies_len describe IEs that extend past
event_body, so the walk reads out of the adapter slab object, a
slab-out-of-bounds read (KASAN: slab-out-of-bounds in cfg80211_find_ie).
An event->len smaller than the header instead makes the int subtraction
negative, which wraps to a huge size_t when stored in assoc_req_ies_len.
The same length is handed to cfg80211_new_sta(), so a more modest
over-claim can also copy stale event_body bytes into the
NL80211_CMD_NEW_STATION notification.
A malicious or malfunctioning mwifiex device (USB/SDIO/PCIe) can deliver
such an event while the interface is in AP/uAP mode.
Validate event->len before use: reject a length that underflows the
header or that would place the IEs outside the event_body[] buffer the
event was copied into. event->len here is struct mwifiex_assoc_event.len,
a payload field internal to this event, not the transport frame length,
so it is validated in this handler rather than at the generic
MWIFIEX_TYPE_EVENT receive path, which only sees the event cause and the
transport frame length. The bound is against event_body[MAX_EVENT_SIZE]
rather than the actually-received length because the transports store the
event differently (USB and SDIO leave the 4-byte event header in
event_skb, PCIe strips it via skb_pull), whereas event_body is the single
fixed buffer all of them copy the event into. This is the event-path
analogue of the receive-path bounds checks added in commit 1195852816
("wifi: mwifiex: Fix OOB and integer underflow when rx packets").
Fixes: e568634ae7 ("mwifiex: add AP event handling framework")
Signed-off-by: HE WEI (ギカク) <skyexpoc@gmail.com>
Reviewed-by: Francesco Dolcini <francesco.dolcini@toradex.com>
Link: https://patch.msgid.link/20260715135711.34688-1-skyexpoc@gmail.com
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
This information can be considered part of the capabilities and should
also be copied to the NAN data station.
Fixes: 27e9b326b6 ("wifi: mac80211: support NAN stations")
Signed-off-by: Benjamin Berg <benjamin.berg@intel.com>
Signed-off-by: Miri Korenblit <miriam.rachel.korenblit@intel.com>
Link: https://patch.msgid.link/20260714141038.15620aa5324b.I049254b854ac91c32e0768eb7c819f32eda34218@changeid
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
When vhost owns the virtio-net header, i.e. when
VHOST_NET_F_VIRTIO_NET_HDR is negotiated, sock_hlen is 0,
meaning that no header will be forwarded to the TAP device.
In the current vhost_net_build_xdp() implementation,
when sock_hlen == 0, the gso pointer can point at the start of the
Ethernet frame instead of a virtio-net header.
This results in a wrong interpretation of the destination MAC address
bytes as struct virtio_net_hdr fields.
This can, for some MAC addresses, trigger -EINVAL and return early
before the TX descriptor is completed, which can stall vhost-net TX.
Before 97b2409f28, the gso pointer was set to the zeroed padding area,
using it as a synthetic virtio-net header. Restore that behavior.
Fixes: 97b2409f28 ("vhost-net: reduce one userspace copy when building XDP buff")
Signed-off-by: Enrico Zanda <enrico.zanda@arm.com>
Acked-by: Michael S. Tsirkin <mst@redhat.com>
Link: https://patch.msgid.link/20260708152242.2268848-1-enrico.zanda@arm.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
wilc_parse_assoc_resp_info() computes the trailing IE length as
ies_len = buffer_len - sizeof(*res);
without first checking that buffer_len is at least sizeof(struct
wilc_assoc_resp) (6 bytes). buffer_len is the length reported for a
received association response (host_int_parse_assoc_resp_info() passes
hif_drv->assoc_resp / assoc_resp_info_len straight in) and must be
validated before the driver accesses the fixed header.
For a frame shorter than the 6-byte fixed header, the subtraction wraps.
For a four-byte response the result is truncated to a u16 ies_len of
65534, so kmemdup() then attempts to copy 65534 bytes starting at
buffer + sizeof(*res), beyond the valid association-response data
(CWE-125). A response shorter than four bytes can also cause an
out-of-bounds read of res->status_code at offsets 2 and 3.
Reject frames too short to hold the fixed header before touching the
header or computing ies_len. Also set the connection status to a failure
on this path: the caller falls through to a
"conn_info->status == WLAN_STATUS_SUCCESS" check after the parser
returns, so leaving the status untouched could let a malformed short
response be treated as a successful association.
Fixes: c5c77ba18e ("staging: wilc1000: Add SDIO/SPI 802.11 driver")
Cc: stable@vger.kernel.org
Assisted-by: Claude:claude-opus-4-8
Signed-off-by: Huihui Huang <hhhuang@smu.edu.sg>
Link: https://patch.msgid.link/20260714091811.3596126-1-hhhuang@smu.edu.sg
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
The most significant change is to fix an ath12k regression which led
to low MLO RX throughput on WCN7850.
The remainder are an assortment of minor bug fixes spread across many
of the ath drivers.
-----BEGIN PGP SIGNATURE-----
iHUEABYKAB0WIQQ/mtSHzPUi16IfDEksFbugiYzLewUCalUE+AAKCRAsFbugiYzL
e41eAQD9fk23dZo0OlT3lWl4AO2VmUECeXFZ0SqCGMQ828t07wEA1NF/1Nzyf7xV
6S4gOVCX+lX8lm9AISx8lBcA6llb5g8=
=aNPp
-----END PGP SIGNATURE-----
Merge tag 'ath-current-20260713' of git://git.kernel.org/pub/scm/linux/kernel/git/ath/ath
Jeff Johnson says:
==================
ath.git update for v7.2-rc4
The most significant change is to fix an ath12k regression which led
to low MLO RX throughput on WCN7850.
The remainder are an assortment of minor bug fixes spread across many
of the ath drivers.
==================
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
When userspace writes 1 to /sys/bus/pci/devices/XX/remove to remove
HDA PCI device, the HDA hardware control path is shut down and devres
unmaps the BAR virtual address bus->remap_addr automatically during
driver removal.
If a delayed HDA verb command arrives after the MMIO region is
unmapped, the driver will access invalid virtual addresses and trigger
a page fault splat.
So add an error check right after constructing snd_hda_power_pm.
Signed-off-by: Bob Song <songxiebing@kylinos.cn>
Link: https://patch.msgid.link/20260717024948.506335-1-songxiebing@kylinos.cn
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Fix regression in rgpower table loading, caused by using
request_firmware(): when the requested firmware does not exist, e.g.
nxp/rgpower_WW.bin does not exist on OpenWRT builds for WRT3200ACM,
request_firmware() falls back to firmware_fallback_sysfs(), which expects
the firmware to be provided by user space using SYSFS. No such utility is
provided in this configuration, so the entire system locks up for 60
seconds, until the request times out. During this time, no other log
messages are observed, and the device does not respond to commands over
UART.
The request_firmware() call is performed in the following context:
current->comm kworker/1:2 in_task 1 irqs_disabled 0 in_atomic 0
Fixed by using request_firmware_direct(). This prevents fallback to SYSFS,
and avoids delay. The rgpower table is optional. The driver falls back
to the device tree power table if the firmware is not present.
The error code is printed for debugging and returned to the caller,
which only cares for success or failure, so there are no side effects.
Fixes: 7b6f16a258 ("wifi: mwifiex: add rgpower table loading support")
Signed-off-by: Georgi Valkov <gvalkov@gmail.com>
Reviewed-by: Francesco Dolcini <francesco.dolcini@toradex.com>
Link: https://patch.msgid.link/20260712221709.7099-1-gvalkov@gmail.com
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
The program header scan handles PT_INTERP from a switch nested in the
scan loop, so its break leaves the switch and not the loop. A binary
carrying more than one PT_INTERP runs the case again and overwrites both
interpreter_name and interpreter. The previous name allocation leaks and
so does the previous interpreter reference, along with the write denial
open_exec() took on it. The denial is never released, so the file stays
unwritable for as long as the system runs.
An unprivileged caller reaches this with a crafted binary and repeats it
at will. binfmt_elf stops at the first PT_INTERP. Do the same here.
The flaw dates back to the driver's introduction in the pre-git history
tree introduced in v2.6.11 by 91808d6ebe39 ("[PATCH] FRV: Add FDPIC ELF
binary format driver").
Link: https://patch.msgid.link/20260721-gezittert-medium-kreide-b41fc1f0277e@brauner
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Cc: stable@vger.kernel.org
Reviewed-by: Jori Koolstra <jkoolstra@xs4all.nl>
Signed-off-by: Christian Brauner (Amutable) <brauner@kernel.org>
wanxl_pci_init_one() stores the freshly allocated card in driver data
before the PLX BAR is mapped. Several early probe failures then unwind
through wanxl_pci_remove_one(), including failure to allocate the coherent
status area or to restore the DMA mask.
wanxl_pci_remove_one() unconditionally calls wanxl_reset(), and
wanxl_reset() dereferences card->plx. On those early failures card->plx
is still NULL, so the error path can dereference a NULL MMIO pointer.
Only issue the hardware reset once the BAR mapping exists. The remaining
cleanup in wanxl_pci_remove_one() already checks whether later resources
were allocated.
This issue was found by a static analysis checker and confirmed by
manual source review.
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Signed-off-by: Ruoyu Wang <ruoyuw560@gmail.com>
Link: https://patch.msgid.link/20260708143415.3169358-1-ruoyuw560@gmail.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
nfp_cpp_resource_find() allocates a CPP mutex handle for the matching
resource-table entry and then reports success. nfp_resource_try_acquire()
immediately passes that handle to nfp_cpp_mutex_trylock().
However, nfp_cpp_mutex_alloc() returns NULL on failure. If that happens
for a matching table entry, the resource lookup still returns success and
the following trylock dereferences a NULL mutex pointer while opening the
resource.
nfp_resource_acquire() already treats failure to allocate the table mutex
as -ENOMEM. Do the same for the resource mutex and fail the lookup before
publishing the rest of the resource handle.
This issue was found by a static analysis checker and confirmed by
manual source review.
Fixes: f01a216157 ("nfp: add support for resources")
Signed-off-by: Ruoyu Wang <ruoyuw560@gmail.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/20260708143408.3168425-1-ruoyuw560@gmail.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
When ieee80211_vif_update_links() adds new links it allocates a link
container for each and calls ieee80211_link_init() (which registers the
per-link debugfs files with file->private_data pointing into the container)
and ieee80211_link_setup(). If the subsequent drv_change_vif_links() fails,
the error path restores the old pointers and jumps to 'free', which frees
the new containers but never removes their debugfs entries or stops the
links. The debugfs files survive with file->private_data dangling at the
freed container, so a later open()+read() (e.g. link-1/txpower)
dereferences freed memory in ieee80211_if_read_link(), a use-after-free.
The removal path already dismantles links correctly via
ieee80211_tear_down_links(), which removes each link's keys and debugfs
entries and calls ieee80211_link_stop(); the add path on the error branch
does not. Commit be1ba9ed22 ("wifi: mac80211: avoid weird state in error
path") hardened this same error path for the link-removal case
(new_links == 0) but left the newly-added links' teardown unaddressed.
drv_change_vif_links() can fail at runtime on MLO drivers (internal
allocation / queue / firmware command failures).
Remove the new links' debugfs entries and stop them before freeing.
BUG: KASAN: slab-use-after-free in ieee80211_if_read_link (net/mac80211/debugfs_netdev.c:127)
Read of size 8 at addr ffff888011290000 by task exploit/145
Call Trace:
...
ieee80211_if_read_link (net/mac80211/debugfs_netdev.c:127)
short_proxy_read (fs/debugfs/file.c:373)
vfs_read (fs/read_write.c:572)
ksys_read (fs/read_write.c:716)
do_syscall_64 (arch/x86/entry/syscall_64.c:94)
entry_SYSCALL_64_after_hwframe (arch/x86/entry/entry_64.S:121)
...
Oops: general protection fault, probably for non-canonical address 0xdffffc000000000a
RIP: 0010:ieee80211_if_read_link (net/mac80211/debugfs_netdev.c:127)
Kernel panic - not syncing: Fatal exception
Fixes: 170cd6a66d ("wifi: mac80211: add netdev per-link debugfs data and driver hook")
Reported-by: Weiming Shi <bestswngs@gmail.com>
Assisted-by: Claude:claude-opus-4-8
Signed-off-by: Xiang Mei <xmei5@asu.edu>
Link: https://patch.msgid.link/20260711210302.2098404-1-xmei5@asu.edu
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
The stop value is derived from bprm->p >> PAGE_SHIFT. The index variable
is an unsigned long. If bprm->p drops below PAGE_SIZE and stop becomes
zero the loop condition index >= stop is always true.
After the index == 0 iteration the decrement wraps to ULONG_MAX and
bprm->page[ULONG_MAX] reads sizeof(void *) bytes in front of the array.
The pointer has wrapped to -1. That garbage pointer is then passed to
kmap_local_page() and PAGE_SIZE bytes are copied from wherever that
lands into the stack of the process being created. And the loop doesn't
terminate either...
Getting there only requires bprm->p < PAGE_SIZE. On !MMU
bprm_set_stack_limit() and bprm_hit_stack_limit() are empty. So the only
constraint on how far bprm->p is pushed down is valid_arg_len(), i.e.
that each individual string still fits in what is left.
bprm->p starts at PAGE_SIZE * MAX_ARG_PAGES - sizeof(void *) so a
single argument or environment string of a little over 31 pages leaves
it in the first page:
Oops - load access fault [#1]
CPU: 0 UID: 0 PID: 1 Comm: victim Not tainted 7.2.0-rc4 #1
epc : __memcpy+0xd4/0xf8
ra : transfer_args_to_stack+0xaa/0xae
s4 : ffffffffffffffff s2 : 0000000000000000
a1 : ffffffdc98000000 a2 : 0000000000001000
status: 0000000a00001880 badaddr: ffffffdc98000000 cause: 0000000000000005
[<801a5324>] __memcpy+0xd4/0xf8
[<800d5f6a>] load_flat_binary+0x43a/0x65e
[<800a2de4>] bprm_execve+0x1d4/0x316
[<800a351a>] do_execveat_common+0x12e/0x138
[<800a3d44>] __riscv_sys_execve+0x38/0x4e
Kernel panic - not syncing: Fatal exception in interrupt
This is an arcane bug but we should still fix it.
Count down from MAX_ARG_PAGES so the loop ends when index reaches stop,
stop == 0 included. The iterations performed are unchanged for every
other value of stop.
Only CONFIG_MMU=n builds are affected, transfer_args_to_stack() is used
by binfmt_flat and binfmt_elf_fdpic on nommu only.
The loop predates git history. commit 7e7ec6a934
("elf_fdpic_transfer_args_to_stack(): make it generic") only moved it
from binfmt_elf_fdpic.c into fs/exec.c and narrowed the copy to the used
part of the first page. The condition and the decrement are unchanged
from 2.6.12-rc2.
Link: https://patch.msgid.link/20260721-hochachtung-staumauer-pigmente-15d71f7d7d04@brauner
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Cc: stable@vger.kernel.org
Reviewed-by: David Hildenbrand (Arm) <david@kernel.org>
Signed-off-by: Christian Brauner (Amutable) <brauner@kernel.org>
need_sync is a per-IOMMU flag shared by all domains and devices behind
that IOMMU. It is set whenever a command is queued with sync == true and
cleared when a completion-wait (CWAIT) command is queued. However, a
cleared need_sync only means that a covering CWAIT has been queued, not
that all previously queued commands have actually completed in hardware.
iommu_completion_wait() read need_sync locklessly and returned early
when it was false. This breaks the "block until all previously queued
commands have completed" contract in a multi-CPU scenario:
CPU2: queue inv-B => need_sync = true
CPU1: queue CWAIT(N); need_sync = false; then wait_on_sem(N)
CPU2: read need_sync == false => return 0 (no wait!)
CPU2 returns without waiting for any sequence number even though its
inv-B may not have completed yet (CWAIT(N), queued after inv-B, has not
been signaled). CPU2 then proceeds to, for example, free page-table
pages while the IOMMU can still walk stale translations, opening a
use-after-free window. This is a logical race in the meaning of the
flag, not a memory-visibility issue, so barriers alone do not help.
Fix it without losing the optimization of avoiding redundant CWAIT
commands: take iommu->lock before testing need_sync, and when it is
false do not return early but wait for the last allocated sequence
number (cmd_sem_val). Since need_sync == false implies no sync command
was queued after the last CWAIT, that CWAIT is FIFO-ordered after every
not-yet-completed command, so waiting for its sequence number guarantees
all prior commands (possibly queued by another CPU) have completed. The
common path with pending work is unchanged and no extra hardware command
is issued.
Signed-off-by: Guanghui Feng <guanghuifeng@linux.alibaba.com>
Fixes: 815b33fdc2 ("x86/amd-iommu: Cleanup completion-wait handling")
Reviewed-by: Vasant Hegde <vasant.hegde@amd.com>
Signed-off-by: Will Deacon <will@kernel.org>
Add timer_shutdown_sync() call at the free callback for systimer
backend, in order to make sure that we can release the resources.
This is only for hardening, and there shouldn't be any actual issue
that requires this change for now.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Link: https://patch.msgid.link/20260720135356.1779857-4-tiwai@suse.de
The systimer backend invokes timer_delete() at stopping the PCM, but
it misses its synchronization, which might lead to concurrent changes
or releases at PCM prepare or free. Use the sync_stop callback to
assure the synchronization of timer deletion.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Link: https://patch.msgid.link/20260720135356.1779857-3-tiwai@suse.de
The hrtimer backend code in ALSA dummy driver calls explicitly the
synchronization of hrtimer cancel from prepare and free callbacks, and
this is exactly what the standard PCM sync_stop callback serves for.
Replace the open-code with the standard PCM sync_stop callback.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Link: https://patch.msgid.link/20260720135356.1779857-2-tiwai@suse.de
Ido Schimmel says:
====================
mlxsw: Make the driver ops-locked
Make the driver ops-locked in order to allow ethtool operations to be
invoked without RTNL being held.
An ops-locked driver has most of its NDOs, all of its ethtool operations
and some net device notifications run with the netdev instance lock
held.
In the specific case of mlxsw, the driver is not using any functions
that acquire this lock nor functions that expect the lock to be held for
an ops-locked driver. Therefore, converting its NDOs to run with the
lock being held is trivial except for a small quirk which is handled in
patch #1.
The driver does not generate any net device notifications, so there is
no risk of nested notifications of the ops-locked types. For the
notifications that run under the instance lock, RTNL is also held, and
the driver does not acquire the instance lock itself, so no changes are
required in its notifier handling.
Ethtool operations can be invoked without RTNL except for two operations
that are annotated in patch #2.
Lastly, patch #3 converts the driver to be ops-locked.
A probe on rtnl_lock() shows it is no longer taken when dumping
statistics:
# perf probe --add rtnl_lock
Before:
# perf stat -e probe:rtnl_lock -- ethtool -S swp1 --all-groups
[...]
1 probe:rtnl_lock
After:
# perf stat -e probe:rtnl_lock -- ethtool -S swp1 --all-groups
[...]
0 probe:rtnl_lock
No issues were reported after running a full regression with a debug
config that has lockdep enabled.
====================
Link: https://patch.msgid.link/20260708123933.1303291-1-idosch@nvidia.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
After the previous changes the driver is now ready to have its net
device and ethtool operations invoked with the netdev instance lock
held.
Tell the core about it by setting request_ops_lock to true.
Reviewed-by: Danielle Ratson <danieller@nvidia.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Link: https://patch.msgid.link/20260708123933.1303291-4-idosch@nvidia.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
A subsequent patch is going to make the driver ops-locked and allow
ethtool operations to run without RTNL. In preparation for this change,
tell the core about a couple of ethtool operations that should remain
under RTNL:
1. Set pause parameters: Configures the port's headroom buffer which is
also configured by RTNL-only paths such as DCB and qdisc. These paths
can probably be converted to acquire the netdev instance lock, but this
operation in not frequently called (unlike stats query), so avoid the
added complexity for now.
2. Get link state: Calls ethtool_op_get_link() which requires RTNL. See
commit 1105ef941c ("net: ethtool: keep rtnl_lock for ops using
ethtool_op_get_link()").
All the other operations do not access shared resources, do not invoke
helpers that require RTNL or already have the appropriate locking in
place.
Reviewed-by: Danielle Ratson <danieller@nvidia.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Link: https://patch.msgid.link/20260708123933.1303291-3-idosch@nvidia.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Commit c5b9b518ad ("mlxsw: spectrum: Add set_rx_mode ndo stub") added
a stub for ndo_set_rx_mode to prevent dev_ifsioc() from returning an
error for the SIOCADDMULTI and SIOCDELMULTI cases.
Since then dev_ifsioc() was taught to also accept ndo_set_rx_mode_async
and commit 3cbd229388 ("net: warn ops-locked drivers still using
ndo_set_rx_mode") modified register_netdevice() to warn when registering
an ops-locked net device that still uses ndo_set_rx_mode instead of
ndo_set_rx_mode_async.
In preparation for converting the driver to be ops-locked, convert the
ndo_set_rx_mode stub to a ndo_set_rx_mode_async stub.
Reviewed-by: Danielle Ratson <danieller@nvidia.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Link: https://patch.msgid.link/20260708123933.1303291-2-idosch@nvidia.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
After the previous patch, __br_multicast_send_query() no longer relies
on br_multicast_alloc_query() to determine the IGMP type of the query.
Remove the argument.
Reviewed-by: Petr Machata <petrm@nvidia.com>
Acked-by: Nikolay Aleksandrov <nikolay@nvidia.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Link: https://patch.msgid.link/20260708122820.1298718-3-idosch@nvidia.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Connecting two bridges on the same system [1] can result in a lockdep
splat [2].
The report is a false positive. Multicast queries are built and
transmitted under the bridge multicast lock. When the outgoing port of
one bridge is configured on top of another bridge, the transmit path
re-enters bridge code and acquires the other bridge's multicast lock in
order to snoop the query. Both lock instances share a single lockdep
class, so lockdep flags the nested acquisition as an AA deadlock.
Giving each bridge its own lock class will not solve the problem: the
reverse topology would produce an ABBA splat with the same pair of
classes. It also consumes a lockdep key per bridge.
Instead, fix the problem by deferring the transmission of the queries to
a workqueue. Build the skb and update querier state under the lock as
before, then enqueue the skb on a per multicast context queue and
schedule the work.
Purge the queue when the multicast context is de-initialized. At this
stage the work cannot be requeued. There is no need to take a reference
on skb->dev since the work cannot outlive the bridge or the bridge port.
Use the high priority workqueue to reduce the delay between the enqueue
time and the transmission time. With default settings (i.e., querier
interval - 255 seconds, query interval - 125 seconds) the extra delay
should not be a problem.
Avoid the unlikely case of the queue growing endlessly by limiting it to
1,000 skbs. Use this number for the simple reason that this is the
default Tx queue length.
Use local_bh_{disable,enable}() to disable/enable softIRQs and migration
in order to avoid corrupting the multicast statistics (per-CPU
u64_stats).
[1]
ip link add name br1 up type bridge mcast_snooping 1 mcast_querier 1
ip link add name br0 up type bridge mcast_snooping 1 mcast_querier 1
ip link add link br0 name br0.10 up master br1 type vlan id 10
[2]
WARNING: possible recursive locking detected
7.0.0-virtme-gb50c64a58a90 #1 Not tainted
[...]
ip/339 is trying to acquire lock:
ffff888104f0b480 (&br->multicast_lock){+.-.}-{3:3}, at: br_ip6_multicast_query (net/bridge/br_multicast.c:3584)
but task is already holding lock:
ffff888104f03480 (&br->multicast_lock){+.-.}-{3:3}, at: br_multicast_port_query_expired (net/bridge/br_multicast.c:1904)
[...]
Call Trace:
[...]
br_ip6_multicast_query (net/bridge/br_multicast.c:3584)
br_multicast_ipv6_rcv (net/bridge/br_multicast.c:3988)
br_dev_xmit (net/bridge/br_device.c:98 (discriminator 1))
dev_hard_start_xmit (net/core/dev.c:3904)
__dev_queue_xmit (net/core/dev.c:4871)
vlan_dev_hard_start_xmit (net/8021q/vlan_dev.c:131 (discriminator 1))
dev_hard_start_xmit (net/core/dev.c:3904)
__dev_queue_xmit (net/core/dev.c:4871)
br_dev_queue_push_xmit (net/bridge/br_forward.c:60)
__br_multicast_send_query (net/bridge/br_multicast.c:1811 (discriminator 1))
br_multicast_send_query (net/bridge/br_multicast.c:1889)
br_multicast_port_query_expired (net/bridge/br_multicast.c:1914)
call_timer_fn (kernel/time/timer.c:1749)
[...]
Reported-by: syzbot+d7b7f1412c02134efa6d@syzkaller.appspotmail.com
Closes: https://lore.kernel.org/netdev/000000000000c4c9d405f2643e01@google.com/
Reviewed-by: Petr Machata <petrm@nvidia.com>
Acked-by: Nikolay Aleksandrov <nikolay@nvidia.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Link: https://patch.msgid.link/20260708122820.1298718-2-idosch@nvidia.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Pass source address, DSCP and output interface from the tunnel
configuration to ip_route_output() in ipip_fill_forward_path(), aligning
the route lookup with the slow path in ipip_tunnel_xmit().
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Reviewed-by: David Ahern <dsahern@kernel.org>
Link: https://patch.msgid.link/20260708-ipip-route-lookup-fill_forward_path-v1-1-b77df74822ed@kernel.org
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
The eDMA/HDMA transfers the driver issues carry a requester function
number in their TLPs, but nothing ever programs it: eDMA v0 leaves the
FUNC_NUM field of the channel control word zero and HDMA leaves the
per-channel func_num register at its reset value, so every transfer is
attributed to function 0. That is invisible in single-function setups,
but once the DMA block serves a non-zero endpoint function, its
requests must carry that function's number for the host to attribute
and translate them correctly.
Record the function number in the chip data (PCI_FUNC() of the probing
device for dw-edma-pcie) and program it per channel.
Endpoint-local chip instances keep func_no at 0, so transfers issued by
the endpoint-side driver remain PF0-attributed. Delegated channels are
programmed by the host-side dw-edma-pcie instance when it takes over the
channel, using that instance's PCI_FUNC().
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Signed-off-by: Koichiro Den <den@valinux.co.jp>
Link: https://patch.msgid.link/20260721062815.4117887-15-den@valinux.co.jp
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Allow PCI ID match data to pass dw_edma_chip flags into dw_edma_probe().
This keeps per-device policy in the match data instead of open-coding it
in probe().
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Signed-off-by: Koichiro Den <den@valinux.co.jp>
Link: https://patch.msgid.link/20260721062815.4117887-14-den@valinux.co.jp
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Skip data block BAR mapping and debug output when a channel has no data
block size. This lets future providers describe channels that only need
descriptor memory exposed.
No functional change intended for existing Synopsys EDDA and
AMD (Xilinx) MDB/CPM6 devices. Their static channel descriptions still
provide data block sizes where data block windows are used. A zero-sized
data block now means "not present" for future metadata providers.
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Signed-off-by: Koichiro Den <den@valinux.co.jp>
Link: https://patch.msgid.link/20260721062815.4117887-13-den@valinux.co.jp
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Add an optional physical address override to struct dw_edma_block and
use a helper to compute descriptor block addresses.
No functional change intended. Existing Synopsys EDDA and AMD (Xilinx)
MDB/CPM6 block descriptors leave the override unset, so the helper still
returns the same values as before.
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Signed-off-by: Koichiro Den <den@valinux.co.jp>
Link: https://patch.msgid.link/20260721062815.4117887-12-den@valinux.co.jp
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Add a match-data flag for devices whose DMA register block starts at an
offset inside the mapped BAR. Existing Synopsys EDDA and AMD (Xilinx)
MDB/CPM6 matches keep using the BAR mapping base directly.
The offsets handled here come from static, trusted match data. Later
metadata-derived offsets are validated when the endpoint DMA metadata
path parses them at that trust boundary.
No functional change intended.
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Signed-off-by: Koichiro Den <den@valinux.co.jp>
Link: https://patch.msgid.link/20260721062815.4117887-11-den@valinux.co.jp
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Move the platform ops pointer into match data. Existing EDDA/MDB/CPM6
matches keep using dw_edma_pcie_plat_ops.
No functional changes intended.
Suggested-by: Frank Li <Frank.Li@nxp.com>
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Signed-off-by: Koichiro Den <den@valinux.co.jp>
Link: https://patch.msgid.link/20260721062815.4117887-10-den@valinux.co.jp
Signed-off-by: Vinod Koul <vkoul@kernel.org>
dw_edma_pcie_probe() now obtains DMA layout data through device-specific
capability callbacks, not only from PCIe Vendor-Specific Extended
Capabilities. Rename the local data copy from vsec_data to dma_data
before adding endpoint DMA BAR metadata discovery, which does not rely
on VSEC.
No functional change intended.
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Signed-off-by: Koichiro Den <den@valinux.co.jp>
Link: https://patch.msgid.link/20260721062815.4117887-9-den@valinux.co.jp
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Move device-specific capability parsing behind per-device match data.
The existing probe path mixes two decisions: which static template a PCI
ID uses, and which device-specific capability parser adjusts that
template. Split those decisions so device-specific discovery can be
added through match data instead of adding more vendor checks to
dw_edma_pcie_probe().
No functional change is intended for the existing Synopsys EDDA and
AMD (Xilinx) MDB/CPM6 matches. They still copy the same static template
data and run the same capability parsing logic before BAR mapping. The
AMD (Xilinx) MDB/CPM6 entries also keep using endpoint memory physical
addresses for descriptor windows through a new match-data flag.
Suggested-by: Frank Li <Frank.Li@nxp.com>
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Signed-off-by: Koichiro Den <den@valinux.co.jp>
Link: https://patch.msgid.link/20260721062815.4117887-8-den@valinux.co.jp
Signed-off-by: Vinod Koul <vkoul@kernel.org>
The dw-edma-pcie driver copies static template data into a mutable
dw_edma_pcie_data instance before applying capability-derived updates.
Keep the derived non-LL mode in that copy as well, instead of only
tracking it in a local variable in dw_edma_pcie_probe().
This prepares for keeping capability parsing behind match data without a
separate non-LL output parameter.
No functional change intended.
Suggested-by: Frank Li <Frank.Li@nxp.com>
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Signed-off-by: Koichiro Den <den@valinux.co.jp>
Link: https://patch.msgid.link/20260721062815.4117887-7-den@valinux.co.jp
Signed-off-by: Vinod Koul <vkoul@kernel.org>
A DesignWare eDMA instance may represent only a subset of channels that
is also initialized by another OS instance, such as an endpoint-side OS.
Add a partial ownership flag for instances that must preserve
controller-wide state owned by that peer.
In partial ownership mode, dw-edma skips the initial core reset and uses
the limited quiesce path in probe() and remove() instead of the full
core-off path. The flag also makes the driver validate the ownership
granularity required by each register layout before registering
channels.
Probe fails if the limited quiesce cannot stop its resources; remove
reports the error after completing the remaining driver teardown.
Partial instances also skip interrupt-emulation doorbell allocation: the
emulated doorbell is a controller-level resource, and a partial owner
must not claim it on behalf of the whole block.
For EDMA_MF_EDMA_UNROLL and EDMA_MF_HDMA_COMPAT, the driver programs
per-direction registers, such as DMA_{WRITE,READ}_INT_MASK_OFF and
DMA_{WRITE,READ}_INT_CLEAR_OFF. These register layouts have at most
EDMA_MAX_{WR,RD}_CH channels per direction, so the capped hardware
channel count still represents the whole direction. A partial instance
can therefore expose write or read channels only if it owns every
channel in that direction; otherwise two OS instances could update the
same direction-wide registers without a shared locking protocol.
In contrast, HDMA native uses per-channel registers, so it can be owned
at channel granularity.
Signed-off-by: Koichiro Den <den@valinux.co.jp>
Link: https://patch.msgid.link/20260721062815.4117887-6-den@valinux.co.jp
Signed-off-by: Vinod Koul <vkoul@kernel.org>
dw_edma_irq_request() passes struct dw_edma_irq to request_irq() before
dw_edma_channel_setup() fills the back pointer. A shared interrupt can
therefore enter the handler with dw_irq->dw still NULL, leading to a
NULL pointer dereference.
Set the back pointer before installing each handler.
Fixes: e63d79d1ff ("dmaengine: Add Synopsys eDMA IP core driver")
Cc: stable@vger.kernel.org
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Signed-off-by: Koichiro Den <den@valinux.co.jp>
Link: https://patch.msgid.link/20260721062815.4117887-5-den@valinux.co.jp
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Add core operations that quiesce only the resources represented by a
dw-edma instance, separate from the existing full controller off path.
For unrolled eDMA, quiesce masks the direction, disables ENGINE_EN,
waits for it to read back clear, then clears pending interrupt status.
If ENGINE_EN does not clear, return the timeout to the caller.
HDMA compatibility mode does not implement ENGINE_EN, so quiesce clears
CHi_PWR_EN for each represented channel instead. Both register maps
share interrupt control per direction, so quiescing one channel
quiesces the whole direction. Callers must own that direction and stop
the peer from programming it first.
HDMA native has per-channel registers and can quiesce the represented
channel directly.
No caller is added yet, so this is a no-functional-change preparation
for delegated channel reclaim and partial-owned remove paths.
Signed-off-by: Koichiro Den <den@valinux.co.jp>
Link: https://patch.msgid.link/20260721062815.4117887-4-den@valinux.co.jp
Signed-off-by: Vinod Koul <vkoul@kernel.org>
DesignWare eDMA can signal completion locally through edma_int[] and
remotely through IMWr/MSI. When channels are delegated to a remote
frontend, the local endpoint side and the remote host side must not both
service the same DONE/ABORT status.
Add channel interrupt routing state and initialize it from the controller
instance configuration. Update the eDMA and HDMA native paths so
linked-list interrupt generation, HDMA non-linked-list interrupt enables,
and DONE/ABORT masking follow the selected mode. For HDMA native
non-linked-list channels, keep the local stop/abort enables set so status
is latched. In remote mode, also enable remote signaling and mask the local
interrupt pins.
Keep the existing dw-edma-pcie host-side instances in remote interrupt
routing mode so their IMWr/MSI completion model remains unchanged after
local routing becomes the zero value.
Note:
- The routing mode describes where a channel should report completion.
It does not by itself say whether this dw-edma instance owns the
interrupt status. A local instance must ignore remote-only channels,
and a remote instance must ignore local-only channels, even if such
interrupts are unexpectedly delivered. Otherwise the non-owner side
could steal the interrupt from the owner by clearing shared DONE/ABORT
status.
Cc: Devendra K Verma <devendra.verma@amd.com>
Suggested-by: Frank Li <Frank.Li@nxp.com>
Signed-off-by: Koichiro Den <den@valinux.co.jp>
Link: https://patch.msgid.link/20260721062815.4117887-3-den@valinux.co.jp
Signed-off-by: Vinod Koul <vkoul@kernel.org>
The HDMA linked-list and non-linked-list start paths both program the
stop/abort interrupt setup register using the same local/remote enable
policy. Only the interrupt-mask handling differs by transfer mode.
Factor the common setup into dw_hdma_v0_core_int_setup() before adding
per-channel interrupt routing support. No functional change intended.
Suggested-by: Frank Li <Frank.Li@nxp.com>
Signed-off-by: Koichiro Den <den@valinux.co.jp>
Link: https://patch.msgid.link/20260721062815.4117887-2-den@valinux.co.jp
Signed-off-by: Vinod Koul <vkoul@kernel.org>
- Avoid redundant page-table allocations in ioremap pcache topup
- Apply SBI FWFT LOCK flag only on successful set
- Bound SBI PMU counter mask scan to BITS_PER_LONG
- Skip TLB flush when G-stage PTE becomes valid with Svvptc
- Zicbo[m|z|p] block sizes should be always present in ONE_REG
- Inject instruction access fault on unmapped guest fetch
- Serialize virtual interrupt pending state updates using raw spinlock
- Fix Spectre-v1 in vector register access via ONE_REG
-----BEGIN PGP SIGNATURE-----
iQIzBAABCgAdFiEEZdn75s5e6LHDQ+f/rUjsVaLHLAcFAmpbMVwACgkQrUjsVaLH
LAdENhAAi5eWE2WACOp50yy+e9/K/n7WFyGsnAg1u157HR0ywBIfOKA9Q3fGnHs/
WUeqXjtjnVLJFB/LfG3aF9O4PLyU+iZ2g7On5O4ZXQMY2wpueHBZ2fyvUFmrWKl6
JwwgluMxeX1IIhJdhSe0hUU8sxcurYtmuxMKO/2lkYcqxfn6UGpBHONkNGVBncsJ
RcaU2lQBnv+8hJ0wlCb0hep9R34YkjF1rmti71tru7XqnF3rq2Z5PBITjfBHXmeB
6iYvVnpSvPiEvxF8454dvQ9bbXmh3OS+ncI/qjS1u2fkrQrM00eZs2bKzUpW7LlG
7ZA5Rm+EQPbHoPV1EKHXXip7uSZ7U8nm9CxMCADI7zZi1K3qvKnnRtA4MahLxO2o
WSxWEW7ONHskZWz1YZUyqpPCutJQOtKaIu7ITzS0aHep9UB/3Lrs1sA9hRTEINUl
9O1Nusro+wdlG1SsycWXPNoyHvEnN1h2FjWLRpr54X9mEQtCCg0/qNvwo6CPZNwH
E9lWgx7rcFnmy24ln01fHOtTD5IKRNis2MUswqNVyc3X7gOft+w8/1QE/MHYC9cl
LF7b12xCOFp1PMkicgVYwWRQ+6qrvidFvWOWqVtcRmfMw6TjXIrB5Fr/MYg8gBL/
N6uzyk3+UmSiFehbElXww3YEiqd81KzKgeEsU74/T2yyAEflVYA=
=jcT0
-----END PGP SIGNATURE-----
Merge tag 'kvm-riscv-fixes-7.2-1' of https://github.com/kvm-riscv/linux into HEAD
KVM/riscv fixes for 7.2, take #1
- Avoid redundant page-table allocations in ioremap pcache topup
- Apply SBI FWFT LOCK flag only on successful set
- Bound SBI PMU counter mask scan to BITS_PER_LONG
- Skip TLB flush when G-stage PTE becomes valid with Svvptc
- Zicbo[m|z|p] block sizes should be always present in ONE_REG
- Inject instruction access fault on unmapped guest fetch
- Serialize virtual interrupt pending state updates using raw spinlock
- Fix Spectre-v1 in vector register access via ONE_REG
On Intel platforms with a VMX preemption timer and APICv, if a VMM
calls KVM_GET_LAPIC before KVM_GET_MSRS to save the vCPU state, it is
possible to lose a pending timer interrupt.
If the thread running these ioctls is migrated to another core after
calling KVM_GET_LAPIC but before KVM_GET_MSRS and the guest is using
their LAPIC timer in TSC-deadline mode, not only does the save LAPIC
state not carry the pending interrupt, the TSCDEADLINE MSR will be
zeroed.
After migration across CPUs, KVM_GET_MSRS calls vcpu_load, posting the
interrupt and clearing the MSR:
vcpu_load() ->
kvm_arch_vcpu_load() ->
kvm_lapic_restart_hv_timer() ->
start_hv_timer() ->
apic_timer_expired() ->
kvm_apic_inject_pending_timer_irqs()
. post interrupt into the LAPIC state
. clear IA32_TSCDEADLINE
The saved LAPIC state will be missing the pending interrupt and the saved
MSR will be zero. Oops.
Fix by only posting an interrupt when we're attempting to enter the guest
(vcpu->wants_to_run == true), not for vcpu_load from other paths.
Assisted-by: gemini:gemini-3.1-pro-preview
Debugged-by: David Matlack <dmatlack@google.com>
Debugged-by: Sean Christopherson <seanjc@google.com>
Debugged-by: Jim Mattson <jmattson@google.com>
Debugged-by: James Houghton <jthoughton@google.com>
Signed-off-by: Venkatesh Srinivas <venkateshs@chromium.org>
Message-ID: <20260715234234.15382-2-venkateshs@chromium.org>
Reviewed-by: James Houghton <jthoughton@google.com>
Reviewed-by: Chao Gao <chao.gao@intel.com>
Cc: stable@vger.kernel.org
Fixes: ae95f566b3 ("KVM: X86: TSCDEADLINE MSR emulation fastpath", 2020-05-15)
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
The test asserted that the X86_FEATURE_SEV CPUID bit exactly matches
whether KVM offers KVM_X86_SEV_VM. That is not an invariant: when all
SEV ASIDs are assigned to SEV-SNP, KVM does not offer the SEV VM type
even though CPUID reports SEV, so the test aborts on an SNP-only host.
Derive SEV availability from KVM_CAP_VM_TYPES (as already done for SEV-ES
and SNP), assert only the one-way implication that a type offered by KVM
is also reported in CPUID, and TEST_REQUIRE() the SEV VM type so the test
skips cleanly when it is unavailable.
Reviewed-by: Tycho Andersen (AMD) <tycho@kernel.org>
Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
Message-ID: <5d3c345113748f39b7982e365d241abaf3e11086.1784545391.git.dwmw@amazon.co.uk>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
sev_smoke_test ran the plain SEV subtest unconditionally, gated only on
the X86_FEATURE_SEV CPUID bit, while gating SEV-ES and SNP on the
KVM_CAP_VM_TYPES bits. CPUID reporting SEV does not mean KVM offers the
SEV VM type: when all SEV ASIDs are assigned to SEV-SNP, KVM_X86_SEV_VM
is unavailable even though X86_FEATURE_SEV is set. On such a host the
test aborts in KVM_CREATE_VM instead of exercising the available modes.
Gate the SEV subtest on KVM_CAP_VM_TYPES like the others, so the test
runs the VM types the host actually offers.
Reviewed-by: Tycho Andersen (AMD) <tycho@kernel.org>
Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
Message-ID: <2b5e7a83d277134294199a455469bb436196b902.1784545391.git.dwmw@amazon.co.uk>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
mmu_destroy_caches() destroys pte_list_desc_cache and
mmu_page_header_cache, but leaves both pointers unchanged. The pointers
live in kvm.ko, and therefore survive when a vendor module is unloaded
while kvm.ko remains loaded.
If creation of pte_list_desc_cache fails during a subsequent vendor
module load, its assignment sets pte_list_desc_cache to NULL and the
error path calls mmu_destroy_caches(). mmu_page_header_cache still
points to the cache destroyed during the preceding vendor module
unload. Passing that stale pointer to kmem_cache_destroy() causes a
slab use-after-free.
Reproduce the issue on a v7.1.3 kernel with CONFIG_KASAN=y,
CONFIG_KASAN_GENERIC=y, CONFIG_KVM=m, and CONFIG_KVM_INTEL=m. A
one-shot test hook forces pte_list_desc_cache to NULL on the second
invocation of kvm_mmu_vendor_module_init():
1. Load kvm.ko and kvm-intel.ko, creating both caches.
2. Unload only kvm_intel, leaving kvm.ko loaded.
3. Reload kvm_intel and force initialization through the -ENOMEM path.
KASAN reports:
BUG: KASAN: slab-use-after-free in
kvm_mmu_vendor_module_init+0x5b/0x170 [kvm]
...
kmem_cache_destroy+0x21/0x1d0
kvm_mmu_vendor_module_init+0x5b/0x170 [kvm]
...
Allocated by task 16817:
__kmem_cache_create_args+0x12c/0x3b0
__kmem_cache_create.constprop.0+0xb6/0xf0 [kvm]
kvm_mmu_vendor_module_init+0x13b/0x170 [kvm]
...
Freed by task 16820:
kmem_cache_destroy+0x117/0x1d0
kvm_mmu_vendor_module_exit+0x21/0x30 [kvm]
Clear both pointers immediately after destroying their caches so that
the stored state reflects the caches' lifetime and repeated cleanup is
safe.
With the fix applied, the same injected vendor module reload fails with
-ENOMEM as expected and produces no KASAN report.
Fixes: cb498ea2ce ("KVM: Portability: Combine kvm_init and kvm_init_x86")
Cc: stable@vger.kernel.org
Signed-off-by: Phil Rosenthal <phil@phil.gs>
Message-ID: <20260718-kvm-mmu-cache-uaf-v3-1-e103b93c74e1@phil.gs>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
kvm_mmu_zap_oldest_mmu_pages() excludes a shadow page whose root_count
is non-zero from top-level reclaim, because such a page cannot be
freed. The path in mmu_page_zap_pte() that recursively zaps a parentless
nested TDP child has no such check. As a result, a shadow page can
be zapped even if the page itself can't be freed; as the comment in
kvm_mmu_zap_oldest_mmu_pages() notes, zapping it will just force vCPUs
to rebuild the page.
As in top-level reclaim, do not recursively prepare zapping of a
nested TDP child whose root_count is non-zero.
Fixes: 2de4085ccc ("KVM: x86/MMU: Recursively zap nested TDP SPs when zapping last/only parent")
Signed-off-by: Hyunwoo Kim <imv4bel@gmail.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Check for a "stale" page fault, i.e. for an invalid and/or obsolete root,
after making MMU pages available for the shadow MMU. If reclaiming shadow
pages zaps an in-use root, i.e. marks it invalid, then KVM will attempt to
map memory into an invalid root. On its own, populating an invalid root is
"fine", but because child shadow pages inherit their parent's role, any
children created during the map/fetch will be created as invalid pages,
thus violating KVM's invariant that invalid pages are never on the list of
active MMU pages.
Note, the underlying flaw has existed since KVM first started tracking
invalid roots in 2008 (commit 2e53d63acb, "KVM: MMU: ignore zapped root
pagetables"), but the true badness only came along in 2020 (Linux 5.9)
with the invariant that invalid shadow pages can't be on the list of
active pages.
Note #2, inheriting role.invalid when creating child shadow pages is also
far from ideal; that flaw will be addressed separately.
Reported-by: Hyunwoo Kim <imv4bel@gmail.com>
Fixes: f95eec9bed ("KVM: x86/mmu: Don't put invalid SPs back on the list of active pages")
Cc: stable@vger.kernel.org
Signed-off-by: Sean Christopherson <seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
free_nested() frees the shadow VMCS while vmcs01 still points to it. But
because it is asynchronous with respect to loaded_vmcs_clear(), the vCPU
might migrate before the pointer is cleared and __loaded_vmcs_clear()
may then execute VMCLEAR.
The VMCS needs to stay attached until its explicit VMCLEAR completes, but
then it can be hidden and the page safely freed.
Fixes: 355f4fb140 ("kvm: nVMX: VMCLEAR an active shadow VMCS after last use")
Cc: stable@vger.kernel.org
Signed-off-by: Hyunwoo Kim <imv4bel@gmail.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
airoha_npu_send_msg() always maps the mailbox buffer with DMA_TO_DEVICE,
but some callers expect the NPU to write response data back into the
same buffer:
- airoha_npu_wlan_msg_get() (NPU_OP_GET): NPU writes response into
the buffer, then the caller reads it via memcpy()
- airoha_npu_ppe_stats_setup() (NPU_OP_SET): NPU writes back
npu_stats_addr field in the response
On non-cache-coherent architectures like EN7581 (Cortex-A53 without
hardware cache coherency for NPU DMA), DMA_TO_DEVICE unmap is a no-op
— it does not invalidate the CPU cache. If the NPU-written cache line
is still present in the CPU cache when the caller reads the buffer,
the CPU observes stale data instead of the NPU response.
This is a timing-sensitive bug: small mailbox buffers (~24 bytes)
typically fit in a single cache line and may survive in the cache
until the caller reads them, producing silent data corruption rather
than a crash. The bug is more likely to trigger when the caller reads
the response immediately after dma_unmap_single() without intervening
cache-evicting operations.
Fix by using DMA_BIDIRECTIONAL for both map and unmap, which ensures
dma_unmap_single() invalidates the CPU cache on non-coherent systems.
The mailbox buffers are small so there is no performance concern.
Fixes: c52918744e ("net: airoha: npu: Move memory allocation in airoha_npu_send_msg() caller")
Signed-off-by: Wayen Yan <win847@gmail.com>
Acked-by: Lorenzo Bianconi <lorenzo@kernel.org>
Link: https://patch.msgid.link/178351055214.98729.11403147818632027428@gmail.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
