Since the previous commit, STP works on bridge in netns.
# unshare -n
# ip link add br0 type bridge
# ip link add veth0 type veth peer name veth1
# ip link set veth0 master br0 up
[ 50.558135] br0: port 1(veth0) entered blocking state
[ 50.558366] br0: port 1(veth0) entered disabled state
[ 50.558798] veth0: entered allmulticast mode
[ 50.564401] veth0: entered promiscuous mode
# ip link set veth1 master br0 up
[ 54.215487] br0: port 2(veth1) entered blocking state
[ 54.215657] br0: port 2(veth1) entered disabled state
[ 54.215848] veth1: entered allmulticast mode
[ 54.219577] veth1: entered promiscuous mode
# ip link set br0 type bridge stp_state 1
# ip link set br0 up
[ 61.960726] br0: port 2(veth1) entered blocking state
[ 61.961097] br0: port 2(veth1) entered listening state
[ 61.961495] br0: port 1(veth0) entered blocking state
[ 61.961653] br0: port 1(veth0) entered listening state
[ 63.998835] br0: port 2(veth1) entered blocking state
[ 77.437113] br0: port 1(veth0) entered learning state
[ 86.653501] br0: received packet on veth0 with own address as source address (addr:6e:0f:e7:6f:5f:5f, vlan:0)
[ 92.797095] br0: port 1(veth0) entered forwarding state
[ 92.797398] br0: topology change detected, propagating
llc: Check netns in llc_estab_match() and llc_listener_match().
We will remove this restriction in llc_rcv() in the following patch,
which means that the protocol handler must be aware of netns.
if (!net_eq(dev_net(dev), &init_net))
goto drop;
llc_rcv() fetches llc_type_handlers[llc_pdu_type(skb) - 1] and calls it
if not NULL.
If the PDU type is LLC_DEST_CONN, llc_conn_handler() is called to pass
skb to corresponding sockets. Then, we must look up a proper socket in
the same netns with skb->dev.
llc_conn_handler() calls __llc_lookup() to look up a established or
litening socket by __llc_lookup_established() and llc_lookup_listener().
Both functions iterate on a list and call llc_estab_match() or
llc_listener_match() to check if the socket is the correct destination.
However, these functions do not check netns.
Also, bind() and connect() call llc_establish_connection(), which
finally calls __llc_lookup_established(), to check if there is a
conflicting socket.
Let's test netns in llc_estab_match() and llc_listener_match().
We will remove this restriction in llc_rcv() soon, which means that the
protocol handler must be aware of netns.
if (!net_eq(dev_net(dev), &init_net))
goto drop;
llc_rcv() fetches llc_type_handlers[llc_pdu_type(skb) - 1] and calls it
if not NULL.
If the PDU type is LLC_DEST_SAP, llc_sap_handler() is called to pass skb
to corresponding sockets. Then, we must look up a proper socket in the
same netns with skb->dev.
If the destination is a multicast address, llc_sap_handler() calls
llc_sap_mcast(). It calculates a hash based on DSAP and skb->dev->ifindex,
iterates on a socket list, and calls llc_mcast_match() to check if the
socket is the correct destination. Then, llc_mcast_match() checks if
skb->dev matches with llc_sk(sk)->dev. So, we need not check netns here.
OTOH, if the destination is a unicast address, llc_sap_handler() calls
llc_lookup_dgram() to look up a socket, but it does not check the netns.
Therefore, we need to add netns check in llc_lookup_dgram().
entries and bind debugfs files would display wrong data on NETSYS_V2 and
later because instead of using mtk_get_ib1_pkt_type the driver would use
MTK_FOE_IB1_PACKET_TYPE which corresponds to NETSYS_V1(.x) SoCs.
Use mtk_get_ib1_pkt_type so entries and bind records display correctly.
r8169: revert 2ab19de62d67 ("r8169: remove ASPM restrictions now that ASPM is disabled during NAPI poll")
There have been reports that on a number of systems this change breaks
network connectivity. Therefore effectively revert it. Mainly affected
seem to be systems where BIOS denies ASPM access to OS.
Due to later changes we can't do a direct revert.
Commit 3f4ca5fafc08 ("tcp: avoid the lookup process failing to get sk in
ehash table") reversed the order in how a socket is inserted into ehash
to fix an issue that ehash-lookup could fail when reqsk/full sk/twsk are
swapped. However, it introduced another lookup failure.
The full socket in ehash is allocated from a slab with SLAB_TYPESAFE_BY_RCU
and does not have SOCK_RCU_FREE, so the socket could be reused even while
it is being referenced on another CPU doing RCU lookup.
Let's say a socket is reused and inserted into the same hash bucket during
lookup. After the blamed commit, a new socket is inserted at the end of
the list. If that happens, we will skip sockets placed after the previous
position of the reused socket, resulting in ehash lookup failure.
As described in Documentation/RCU/rculist_nulls.rst, we should insert a
new socket at the head of the list to avoid such an issue.
This issue, the swap-lookup-failure, and another variant reported in [0]
can all be handled properly by adding a locked ehash lookup suggested by
Eric Dumazet [1].
However, this issue could occur for every packet, thus more likely than
the other two races, so let's revert the change for now.
Jakub Kicinski [Wed, 19 Jul 2023 22:01:10 +0000 (15:01 -0700)]
Merge tag 'for-netdev' of https://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf
Alexei Starovoitov says:
====================
pull-request: bpf 2023-07-19
We've added 4 non-merge commits during the last 1 day(s) which contain
a total of 3 files changed, 55 insertions(+), 10 deletions(-).
The main changes are:
1) Fix stack depth check in presence of async callbacks,
from Kumar Kartikeya Dwivedi.
2) Fix BTI type used for freplace attached functions,
from Alexander Duyck.
* tag 'for-netdev' of https://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf:
bpf, arm64: Fix BTI type used for freplace attached functions
selftests/bpf: Add more tests for check_max_stack_depth bug
bpf: Repeat check_max_stack_depth for async callbacks
bpf: Fix subprog idx logic in check_max_stack_depth
====================
Jakub Kicinski [Wed, 19 Jul 2023 02:49:08 +0000 (19:49 -0700)]
Merge branch '40GbE' of git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/net-queue
Tony Nguyen says:
====================
Intel Wired LAN Driver Updates 2023-07-17 (iavf)
This series contains updates to iavf driver only.
Ding Hui fixes use-after-free issue by calling netif_napi_del() for all
allocated q_vectors. He also resolves out-of-bounds issue by not
updating to new values when timeout is encountered.
Marcin and Ahmed change the way resets are handled so that the callback
operating under the RTNL lock will wait for the reset to finish, the
rtnl_lock sensitive functions in reset flow will schedule the netdev update
for later in order to remove circular dependency with the critical lock.
* '40GbE' of git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/net-queue:
iavf: fix reset task race with iavf_remove()
iavf: fix a deadlock caused by rtnl and driver's lock circular dependencies
Revert "iavf: Do not restart Tx queues after reset task failure"
Revert "iavf: Detach device during reset task"
iavf: Wait for reset in callbacks which trigger it
iavf: use internal state to free traffic IRQs
iavf: Fix out-of-bounds when setting channels on remove
iavf: Fix use-after-free in free_netdev
====================
Reported by Kernel Concurrency Sanitizer on:
CPU: 0 PID: 32078 Comm: syz-executor.4 Not tainted 6.5.0-rc1-syzkaller-00033-geb26cbb1a754 #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 07/03/2023
octeontx2-pf: mcs: Generate hash key using ecb(aes)
Hardware generated encryption and ICV tags are found to
be wrong when tested with IEEE MACSEC test vectors.
This is because as per the HRM, the hash key (derived by
AES-ECB block encryption of an all 0s block with the SAK)
has to be programmed by the software in
MCSX_RS_MCS_CPM_TX_SLAVE_SA_PLCY_MEM_4X register.
Hence fix this by generating hash key in software and
configuring in hardware.
In normal operation, each populated queue item has
next_to_watch pointing to the last TX desc of the packet,
while each cleaned item has it set to 0. In particular,
next_to_use that points to the next (necessarily clean)
item to use has next_to_watch set to 0.
When the TX queue is used both by an application using
AF_XDP with ZEROCOPY as well as a second non-XDP application
generating high traffic, the queue pointers can get in
an invalid state where next_to_use points to an item
where next_to_watch is NOT set to 0.
However, the implementation assumes at several places
that this is never the case, so if it does hold,
bad things happen. In particular, within the loop inside
of igc_clean_tx_irq(), next_to_clean can overtake next_to_use.
Finally, this prevents any further transmission via
this queue and it never gets unblocked or signaled.
Secondly, if the queue is in this garbled state,
the inner loop of igc_clean_tx_ring() will never terminate,
completely hogging a CPU core.
The reason is that igc_xdp_xmit_zc() reads next_to_use
before acquiring the lock, and writing it back
(potentially unmodified) later. If it got modified
before locking, the outdated next_to_use is written
pointing to an item that was already used elsewhere
(and thus next_to_watch got written).
Jakub Kicinski [Wed, 19 Jul 2023 01:33:34 +0000 (18:33 -0700)]
Merge tag 'linux-can-fixes-for-6.5-20230717' of git://git.kernel.org/pub/scm/linux/kernel/git/mkl/linux-can
Marc Kleine-Budde says:
====================
pull-request: can 2023-07-17
The 1st patch is by Ziyang Xuan and fixes a possible memory leak in
the receiver handling in the CAN RAW protocol.
YueHaibing contributes a use after free in bcm_proc_show() of the
Broad Cast Manager (BCM) CAN protocol.
The next 2 patches are by me and fix a possible null pointer
dereference in the RX path of the gs_usb driver with activated
hardware timestamps and the candlelight firmware.
The last patch is by Fedor Ross, Marek Vasut and me and targets the
mcp251xfd driver. The polling timeout of __mcp251xfd_chip_set_mode()
is increased to fix bus joining on busy CAN buses and very low bit
rate.
* tag 'linux-can-fixes-for-6.5-20230717' of git://git.kernel.org/pub/scm/linux/kernel/git/mkl/linux-can:
can: mcp251xfd: __mcp251xfd_chip_set_mode(): increase poll timeout
can: gs_usb: fix time stamp counter initialization
can: gs_usb: gs_can_open(): improve error handling
can: bcm: Fix UAF in bcm_proc_show()
can: raw: fix receiver memory leak
====================
====================
selftests: tc: increase timeout and add missing kconfig
When looking for something else in LKFT reports [1], I noticed that the
TC selftest ended with a timeout error:
not ok 1 selftests: tc-testing: tdc.sh # TIMEOUT 45 seconds
I also noticed most of the tests were skipped because the "teardown
stage" did not complete successfully. It was due to missing kconfig.
These patches fix these two errors plus an extra one because this
selftest reads info from "/proc/net/nf_conntrack". Thank you Pedro for
having helped me fixing these issues [2].
When looking at the TC selftest reports, I noticed one test was failing
because /proc/net/nf_conntrack was not available.
not ok 373 3992 - Add ct action triggering DNAT tuple conflict
Could not match regex pattern. Verify command output:
cat: /proc/net/nf_conntrack: No such file or directory
It is only available if NF_CONNTRACK_PROCFS kconfig is set. So the issue
can be fixed simply by adding it to the list of required kconfig.
When looking for something else in LKFT reports [1], I noticed most of
the tests were skipped because the "teardown stage" did not complete
successfully.
Pedro found out this is due to the fact CONFIG_NF_FLOW_TABLE is required
but not listed in the 'config' file. Adding it to the list fixes the
issues on LKFT side. CONFIG_NET_ACT_CT is now set to 'm' in the final
kconfig.
When looking for something else in LKFT reports [1], I noticed that the
TC selftest ended with a timeout error:
not ok 1 selftests: tc-testing: tdc.sh # TIMEOUT 45 seconds
The timeout had been introduced 3 years ago, see the Fixes commit below.
This timeout is only in place when executing the selftests via the
kselftests runner scripts. I guess this is not what most TC devs are
using and nobody noticed the issue before.
The new timeout is set to 15 minutes as suggested by Pedro [2]. It looks
like it is plenty more time than what it takes in "normal" conditions.
Alexander Duyck [Thu, 13 Jul 2023 16:49:31 +0000 (09:49 -0700)]
bpf, arm64: Fix BTI type used for freplace attached functions
When running an freplace attached bpf program on an arm64 system w were
seeing the following issue:
Unhandled 64-bit el1h sync exception on CPU47, ESR 0x0000000036000003 -- BTI
After a bit of work to track it down I determined that what appeared to be
happening is that the 'bti c' at the start of the program was somehow being
reached after a 'br' instruction. Further digging pointed me toward the
fact that the function was attached via freplace. This in turn led me to
build_plt which I believe is invoking the long jump which is triggering
this error.
To resolve it we can replace the 'bti c' with 'bti jc' and add a comment
explaining why this has to be modified as such.
====================
Two more fixes for check_max_stack_depth
I noticed two more bugs while reviewing the code, description and
examples available in the patches.
One leads to incorrect subprog index to be stored in the frame stack
maintained by the function (leading to incorrect tail_call_reachable
marks, among other things).
The other problem is missing exploration pass of other async callbacks
when they are not called from the main prog. Call chains rooted at them
can thus bypass the stack limits (32 call frames * max permitted stack
depth per function).
selftests/bpf: Add more tests for check_max_stack_depth bug
Another test which now exercies the path of the verifier where it will
explore call chains rooted at the async callback. Without the prior
fixes, this program loads successfully, which is incorrect.
bpf: Repeat check_max_stack_depth for async callbacks
While the check_max_stack_depth function explores call chains emanating
from the main prog, which is typically enough to cover all possible call
chains, it doesn't explore those rooted at async callbacks unless the
async callback will have been directly called, since unlike non-async
callbacks it skips their instruction exploration as they don't
contribute to stack depth.
It could be the case that the async callback leads to a callchain which
exceeds the stack depth, but this is never reachable while only
exploring the entry point from main subprog. Hence, repeat the check for
the main subprog *and* all async callbacks marked by the symbolic
execution pass of the verifier, as execution of the program may begin at
any of them.
Consider functions with following stack depths:
main: 256
async: 256
foo: 256
main:
rX = async
bpf_timer_set_callback(...)
async:
foo()
Here, async is not descended as it does not contribute to stack depth of
main (since it is referenced using bpf_pseudo_func and not
bpf_pseudo_call). However, when async is invoked asynchronously, it will
end up breaching the MAX_BPF_STACK limit by calling foo.
Hence, in addition to main, we also need to explore call chains
beginning at all async callback subprogs in a program.
bpf: Fix subprog idx logic in check_max_stack_depth
The assignment to idx in check_max_stack_depth happens once we see a
bpf_pseudo_call or bpf_pseudo_func. This is not an issue as the rest of
the code performs a few checks and then pushes the frame to the frame
stack, except the case of async callbacks. If the async callback case
causes the loop iteration to be skipped, the idx assignment will be
incorrect on the next iteration of the loop. The value stored in the
frame stack (as the subprogno of the current subprog) will be incorrect.
This leads to incorrect checks and incorrect tail_call_reachable
marking. Save the target subprog in a new variable and only assign to
idx once we are done with the is_async_cb check which may skip pushing
of frame to the frame stack and subsequent stack depth checks and tail
call markings.
octeontx2-pf: Dont allocate BPIDs for LBK interfaces
Current driver enables backpressure for LBK interfaces.
But these interfaces do not support this feature.
Hence, this patch fixes the issue by skipping the
backpressure configuration for these interfaces.
Cited commit converted the neighbour code to use the standard RCU
variant instead of the RCU-bh variant, but the VRF code still uses
rcu_read_lock_bh() / rcu_read_unlock_bh() around the neighbour lookup
code in its IPv4 and IPv6 output paths, resulting in lockdep splats
[1][2]. Can be reproduced using [3].
Fix by switching to rcu_read_lock() / rcu_read_unlock().
ip link add name vrf-red up numtxqueues 2 type vrf table 10
ip link add name swp1 up master vrf-red type dummy
ip address add 192.0.2.1/24 dev swp1
ip address add 2001:db8:1::1/64 dev swp1
ip neigh add 192.0.2.2 lladdr 00:11:22:33:44:55 nud perm dev swp1
ip neigh add 2001:db8:1::2 lladdr 00:11:22:33:44:55 nud perm dev swp1
ip vrf exec vrf-red ping 192.0.2.2 -c 1 &> /dev/null
ip vrf exec vrf-red ping6 2001:db8:1::2 -c 1 &> /dev/null
The mcp251xfd controller needs an idle bus to enter 'Normal CAN 2.0
mode' or . The maximum length of a CAN frame is 736 bits (64 data
bytes, CAN-FD, EFF mode, worst case bit stuffing and interframe
spacing). For low bit rates like 10 kbit/s the arbitrarily chosen
MCP251XFD_POLL_TIMEOUT_US of 1 ms is too small.
Otherwise during polling for the CAN controller to enter 'Normal CAN
2.0 mode' the timeout limit is exceeded and the configuration fails
with:
| $ ip link set dev can1 up type can bitrate 10000
| [ 731.911072] mcp251xfd spi2.1 can1: Controller failed to enter mode CAN 2.0 Mode (6) and stays in Configuration Mode (4) (con=0x068b0760, osc=0x00000468).
| [ 731.927192] mcp251xfd spi2.1 can1: CRC read error at address 0x0e0c (length=4, data=00 00 00 00, CRC=0x0000) retrying.
| [ 731.938101] A link change request failed with some changes committed already. Interface can1 may have been left with an inconsistent configuration, please check.
| RTNETLINK answers: Connection timed out
Make MCP251XFD_POLL_TIMEOUT_US timeout calculation dynamic. Use
maximum of 1ms and bit time of 1 full 64 data bytes CAN-FD frame in
EFF mode, worst case bit stuffing and interframe spacing at the
current bit rate.
For easier backporting define the macro MCP251XFD_FRAME_LEN_MAX_BITS
that holds the max frame length in bits, which is 736. This can be
replaced by can_frame_bits(true, true, true, true, CANFD_MAX_DLEN) in
a cleanup patch later.
Ahmed Zaki [Mon, 5 Jun 2023 14:52:26 +0000 (10:52 -0400)]
iavf: fix reset task race with iavf_remove()
The reset task is currently scheduled from the watchdog or adminq tasks.
First, all direct calls to schedule the reset task are replaced with the
iavf_schedule_reset(), which is modified to accept the flag showing the
type of reset.
To prevent the reset task from starting once iavf_remove() starts, we need
to check the __IAVF_IN_REMOVE_TASK bit before we schedule it. This is now
easily added to iavf_schedule_reset().
Finally, remove the check for IAVF_FLAG_RESET_NEEDED in the watchdog task.
It is redundant since all callers who set the flag immediately schedules
the reset task.
Fixes: 3ccd54ef44eb ("iavf: Fix init state closure on remove") Fixes: 14756b2ae265 ("iavf: Fix __IAVF_RESETTING state usage") Signed-off-by: Ahmed Zaki <[email protected]> Signed-off-by: Mateusz Palczewski <[email protected]> Tested-by: Rafal Romanowski <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
Ahmed Zaki [Mon, 5 Jun 2023 14:52:25 +0000 (10:52 -0400)]
iavf: fix a deadlock caused by rtnl and driver's lock circular dependencies
A driver's lock (crit_lock) is used to serialize all the driver's tasks.
Lockdep, however, shows a circular dependency between rtnl and
crit_lock. This happens when an ndo that already holds the rtnl requests
the driver to reset, since the reset task (in some paths) tries to grab
rtnl to either change real number of queues of update netdev features.
[566.241851] ======================================================
[566.241893] WARNING: possible circular locking dependency detected
[566.241936] 6.2.14-100.fc36.x86_64+debug #1 Tainted: G OE
[566.241984] ------------------------------------------------------
[566.242025] repro.sh/2604 is trying to acquire lock:
[566.242061] ffff9280fc5ceee8 (&adapter->crit_lock){+.+.}-{3:3}, at: iavf_close+0x3c/0x240 [iavf]
[566.242167]
but task is already holding lock:
[566.242209] ffffffff9976d350 (rtnl_mutex){+.+.}-{3:3}, at: iavf_remove+0x6b5/0x730 [iavf]
[566.242300]
which lock already depends on the new lock.
The deadlock can be triggered by a script that is continuously resetting
the VF adapter while doing other operations requiring RTNL, e.g:
while :; do
ip link set $VF up
ethtool --set-channels $VF combined 2
ip link set $VF down
ip link set $VF up
ethtool --set-channels $VF combined 4
ip link set $VF down
done
Any operation that triggers a reset can substitute "ethtool --set-channles"
As a fix, add a new task "finish_config" that do all the work which
needs rtnl lock. With the exception of iavf_remove(), all work that
require rtnl should be called from this task.
As for iavf_remove(), at the point where we need to call
unregister_netdevice() (and grab rtnl_lock), we make sure the finish_config
task is not running (cancel_work_sync()) to safely grab rtnl. Subsequent
finish_config work cannot restart after that since the task is guarded
by the __IAVF_IN_REMOVE_TASK bit in iavf_schedule_finish_config().
Detaching device during reset was not fully fixing the rtnl locking issue,
as there could be a situation where callback was already in progress before
detaching netdev.
Furthermore, detaching netdevice causes TX timeouts if traffic is running.
To reproduce:
ip netns exec ns1 iperf3 -c $PEER_IP -t 600 --logfile /dev/null &
while :; do
for i in 200 7000 400 5000 300 3000 ; do
ip netns exec ns1 ip link set $VF1 mtu $i
sleep 2
done
sleep 10
done
Currently, callbacks such as iavf_change_mtu() wait for the reset.
If the reset fails to acquire the rtnl_lock, they schedule the netdev
update for later while continuing the reset flow. Operations like MTU
changes are performed under the rtnl_lock. Therefore, when the operation
finishes, another callback that uses rtnl_lock can start.
Marcin Szycik [Mon, 5 Jun 2023 14:52:22 +0000 (10:52 -0400)]
iavf: Wait for reset in callbacks which trigger it
There was a fail when trying to add the interface to bonding
right after changing the MTU on the interface. It was caused
by bonding interface unable to open the interface due to
interface being in __RESETTING state because of MTU change.
Add new reset_waitqueue to indicate that reset has finished.
Add waiting for reset to finish in callbacks which trigger hw reset:
iavf_set_priv_flags(), iavf_change_mtu() and iavf_set_ringparam().
We use a 5000ms timeout period because on Hyper-V based systems,
this operation takes around 3000-4000ms. In normal circumstances,
it doesn't take more than 500ms to complete.
Add a function iavf_wait_for_reset() to reuse waiting for reset code and
use it also in iavf_set_channels(), which already waits for reset.
We don't use error handling in iavf_set_channels() as this could
cause the device to be in incorrect state if the reset was scheduled
but hit timeout or the waitng function was interrupted by a signal.
Ahmed Zaki [Fri, 19 May 2023 21:46:02 +0000 (15:46 -0600)]
iavf: use internal state to free traffic IRQs
If the system tries to close the netdev while iavf_reset_task() is
running, __LINK_STATE_START will be cleared and netif_running() will
return false in iavf_reinit_interrupt_scheme(). This will result in
iavf_free_traffic_irqs() not being called and a leak as follows:
[7632.489326] remove_proc_entry: removing non-empty directory 'irq/999', leaking at least 'iavf-enp24s0f0v0-TxRx-0'
[7632.490214] WARNING: CPU: 0 PID: 10 at fs/proc/generic.c:718 remove_proc_entry+0x19b/0x1b0
is shown when pci_disable_msix() is later called. Fix by using the
internal adapter state. The traffic IRQs will always exist if
state == __IAVF_RUNNING.
Fixes: 5b36e8d04b44 ("i40evf: Enable VF to request an alternate queue allocation") Signed-off-by: Ahmed Zaki <[email protected]> Tested-by: Rafal Romanowski <[email protected]> Signed-off-by: Tony Nguyen <[email protected]>
Ding Hui [Tue, 9 May 2023 11:11:48 +0000 (19:11 +0800)]
iavf: Fix out-of-bounds when setting channels on remove
If we set channels greater during iavf_remove(), and waiting reset done
would be timeout, then returned with error but changed num_active_queues
directly, that will lead to OOB like the following logs. Because the
num_active_queues is greater than tx/rx_rings[] allocated actually.
Ding Hui [Tue, 9 May 2023 11:11:47 +0000 (19:11 +0800)]
iavf: Fix use-after-free in free_netdev
We do netif_napi_add() for all allocated q_vectors[], but potentially
do netif_napi_del() for part of them, then kfree q_vectors and leave
invalid pointers at dev->napi_list.
Although the patch #2 (of 2) can avoid the issue triggered by this
repro.sh, there still are other potential risks that if num_active_queues
is changed to less than allocated q_vectors[] by unexpected, the
mismatched netif_napi_add/del() can also cause UAF.
Since we actually call netif_napi_add() for all allocated q_vectors
unconditionally in iavf_alloc_q_vectors(), so we should fix it by
letting netif_napi_del() match to netif_napi_add().
During testing I noticed a crash if unloading/loading the gs_usb
driver during high CAN bus load.
The current version of the candlelight firmware doesn't flush the
queues of the received CAN frames during the reset command. This leads
to a crash if hardware timestamps are enabled, and an URB from the
device is received before the cycle counter/time counter
infrastructure has been setup.
First clean up then error handling in gs_can_open().
Then, fix the problem by converting the cycle counter/time counter
infrastructure from a per-channel to per-device and set it up before
submitting RX-URBs to the USB stack.
can: gs_usb: fix time stamp counter initialization
If the gs_usb device driver is unloaded (or unbound) before the
interface is shut down, the USB stack first calls the struct
usb_driver::disconnect and then the struct net_device_ops::ndo_stop
callback.
In gs_usb_disconnect() all pending bulk URBs are killed, i.e. no more
RX'ed CAN frames are send from the USB device to the host. Later in
gs_can_close() a reset control message is send to each CAN channel to
remove the controller from the CAN bus. In this race window the USB
device can still receive CAN frames from the bus and internally queue
them to be send to the host.
At least in the current version of the candlelight firmware, the queue
of received CAN frames is not emptied during the reset command. After
loading (or binding) the gs_usb driver, new URBs are submitted during
the struct net_device_ops::ndo_open callback and the candlelight
firmware starts sending its already queued CAN frames to the host.
However, this scenario was not considered when implementing the
hardware timestamp function. The cycle counter/time counter
infrastructure is set up (gs_usb_timestamp_init()) after the USBs are
submitted, resulting in a NULL pointer dereference if
timecounter_cyc2time() (via the call chain:
gs_usb_receive_bulk_callback() -> gs_usb_set_timestamp() ->
gs_usb_skb_set_timestamp()) is called too early.
Move the gs_usb_timestamp_init() function before the URBs are
submitted to fix this problem.
For a comprehensive solution, we need to consider gs_usb devices with
more than 1 channel. The cycle counter/time counter infrastructure is
setup per channel, but the RX URBs are per device. Once gs_can_open()
of _a_ channel has been called, and URBs have been submitted, the
gs_usb_receive_bulk_callback() can be called for _all_ available
channels, even for channels that are not running, yet. As cycle
counter/time counter has not set up, this will again lead to a NULL
pointer dereference.
Convert the cycle counter/time counter from a "per channel" to a "per
device" functionality. Also set it up, before submitting any URBs to
the device.
Further in gs_usb_receive_bulk_callback(), don't process any URBs for
not started CAN channels, only resubmit the URB.
The gs_usb driver handles USB devices with more than 1 CAN channel.
The RX path for all channels share the same bulk endpoint (the
transmitted bulk data encodes the channel number). These per-device
resources are allocated and submitted by the first opened channel.
During this allocation, the resources are either released immediately
in case of a failure or the URBs are anchored. All anchored URBs are
finally killed with gs_usb_disconnect().
Currently, gs_can_open() returns with an error if the allocation of a
URB or a buffer fails. However, if usb_submit_urb() fails, the driver
continues with the URBs submitted so far, even if no URBs were
successfully submitted.
Treat every error as fatal and free all allocated resources
immediately.
Switch to goto-style error handling, to prepare the driver for more
per-device resource allocation.
It's a bug in the concurrent scenario of unregister_netdevice_many()
and raw_release() as following:
cpu0 cpu1
unregister_netdevice_many(can_dev)
unlist_netdevice(can_dev) // dev_get_by_index() return NULL after this
net_set_todo(can_dev)
raw_release(can_socket)
dev = dev_get_by_index(, ro->ifindex); // dev == NULL
if (dev) { // receivers in dev_rcv_lists not free because dev is NULL
raw_disable_allfilters(, dev, );
dev_put(dev);
}
...
ro->bound = 0;
...
call_netdevice_notifiers(NETDEV_UNREGISTER, )
raw_notify(, NETDEV_UNREGISTER, )
if (ro->bound) // invalid because ro->bound has been set 0
raw_disable_allfilters(, dev, ); // receivers in dev_rcv_lists will never be freed
Add a net_device pointer member in struct raw_sock to record bound
can_dev, and use rtnl_lock to serialize raw_socket members between
raw_bind(), raw_release(), raw_setsockopt() and raw_notify(). Use
ro->dev to decide whether to free receivers in dev_rcv_lists.
r8169: fix ASPM-related problem for chip version 42 and 43
Referenced commit missed that for chip versions 42 and 43 ASPM
remained disabled in the respective rtl_hw_start_...() routines.
This resulted in problems as described in the referenced bug
ticket. Therefore re-instantiate the previous logic.
Fixes: 5fc3f6c90cca ("r8169: consolidate disabling ASPM before EPHY access") Closes: https://bugzilla.kernel.org/show_bug.cgi?id=217635 Signed-off-by: Heiner Kallweit <[email protected]> Signed-off-by: David S. Miller <[email protected]>
Tristram Ha [Fri, 14 Jul 2023 00:46:22 +0000 (17:46 -0700)]
net: dsa: microchip: correct KSZ8795 static MAC table access
The KSZ8795 driver code was modified to use on KSZ8863/73, which has
different register definitions. Some of the new KSZ8795 register
information are wrong compared to previous code.
KSZ8795 also behaves differently in that the STATIC_MAC_TABLE_USE_FID
and STATIC_MAC_TABLE_FID bits are off by 1 when doing MAC table reading
than writing. To compensate that a special code was added to shift the
register value by 1 before applying those bits. This is wrong when the
code is running on KSZ8863, so this special code is only executed when
KSZ8795 is detected.
Fixes: 4b20a07e103f ("net: dsa: microchip: ksz8795: add support for ksz88xx chips") Signed-off-by: Tristram Ha <[email protected]> Reviewed-by: Horatiu Vultur <[email protected]> Reviewed-by: Simon Horman <[email protected]> Signed-off-by: David S. Miller <[email protected]>
David S. Miller [Mon, 17 Jul 2023 06:33:39 +0000 (07:33 +0100)]
Merge branch 'sched-fixes'
Victor Nogueira says:
====================
net: sched: Fixes for classifiers
Four different classifiers (bpf, u32, matchall, and flower) are
calling tcf_bind_filter in their callbacks, but arent't undoing it by
calling tcf_unbind_filter if their was an error after binding.
This patch set fixes all this by calling tcf_unbind_filter in such
cases.
This set also undoes a refcount decrement in cls_u32 when an update
fails under specific conditions which are described in patch #3.
v1 -> v2:
* Remove blank line after fixes tag
* Fix reverse xmas tree issues pointed out by Simon
v2 -> v3:
* Inline functions cls_bpf_set_parms and fl_set_parms to avoid adding
yet another parameter (and a return value at it) to them.
* Remove similar fixes for u32 and matchall, which will be sent soon,
once we find a way to do the fixes without adding a return parameter
to their set_parms functions.
v3 -> v4:
* Inline mall_set_parms to avoid adding yet another parameter.
* Remove set_flags parameter from u32_set_parms and create a separate
function for calling tcf_bind_filter and tcf_unbind_filter in case of
failure.
* Change cover letter title to also encompass refcnt fix for u32
v4 -> v5:
* Change back tag to net
====================
Victor Nogueira [Thu, 13 Jul 2023 18:05:14 +0000 (15:05 -0300)]
net: sched: cls_flower: Undo tcf_bind_filter in case of an error
If TCA_FLOWER_CLASSID is specified in the netlink message, the code will
call tcf_bind_filter. However, if any error occurs after that, the code
should undo this by calling tcf_unbind_filter.
Victor Nogueira [Thu, 13 Jul 2023 18:05:12 +0000 (15:05 -0300)]
net: sched: cls_u32: Undo refcount decrement in case update failed
In the case of an update, when TCA_U32_LINK is set, u32_set_parms will
decrement the refcount of the ht_down (struct tc_u_hnode) pointer
present in the older u32 filter which we are replacing. However, if
u32_replace_hw_knode errors out, the update command fails and that
ht_down pointer continues decremented. To fix that, when
u32_replace_hw_knode fails, check if ht_down's refcount was decremented
and undo the decrement.
Randy Dunlap [Fri, 14 Jul 2023 04:51:27 +0000 (21:51 -0700)]
rsi: remove kernel-doc comment marker
Change an errant kernel-doc comment marker (/**) to a regular
comment to prevent a kernel-doc warning.
rsi_91x.h:3: warning: This comment starts with '/**', but isn't a kernel-doc comment. Refer Documentation/doc-guide/kernel-doc.rst
* Copyright (c) 2017 Redpine Signals Inc.
Randy Dunlap [Fri, 14 Jul 2023 04:51:24 +0000 (21:51 -0700)]
net: llc: fix kernel-doc notation warnings
Use the corrent function parameter name or format to prevent
kernel-doc warnings.
Add 2 function parameter descriptions to prevent kernel-doc warnings.
llc_pdu.h:278: warning: Function parameter or member 'da' not described in 'llc_pdu_decode_da'
llc_pdu.h:278: warning: Excess function parameter 'sa' description in 'llc_pdu_decode_da'
llc_pdu.h:330: warning: Function parameter or member 'skb' not described in 'llc_pdu_init_as_test_cmd'
llc_pdu.h:379: warning: Function parameter or member 'svcs_supported' not described in 'llc_pdu_init_as_xid_cmd'
llc_pdu.h:379: warning: Function parameter or member 'rx_window' not described in 'llc_pdu_init_as_xid_cmd'
Randy Dunlap [Fri, 14 Jul 2023 04:51:22 +0000 (21:51 -0700)]
devlink: fix kernel-doc notation warnings
Spell function or struct member names correctly.
Use ':' instead of '-' for struct member entries.
Mark one field as private in kernel-doc.
Add a few entries that were missing.
Fix a typo.
These changes prevent kernel-doc warnings:
devlink.h:252: warning: Function parameter or member 'field_id' not described in 'devlink_dpipe_match'
devlink.h:267: warning: Function parameter or member 'field_id' not described in 'devlink_dpipe_action'
devlink.h:310: warning: Function parameter or member 'match_values_count' not described in 'devlink_dpipe_entry'
devlink.h:355: warning: Function parameter or member 'list' not described in 'devlink_dpipe_table'
devlink.h:374: warning: Function parameter or member 'actions_dump' not described in 'devlink_dpipe_table_ops'
devlink.h:374: warning: Function parameter or member 'matches_dump' not described in 'devlink_dpipe_table_ops'
devlink.h:374: warning: Function parameter or member 'entries_dump' not described in 'devlink_dpipe_table_ops'
devlink.h:374: warning: Function parameter or member 'counters_set_update' not described in 'devlink_dpipe_table_ops'
devlink.h:374: warning: Function parameter or member 'size_get' not described in 'devlink_dpipe_table_ops'
devlink.h:384: warning: Function parameter or member 'headers' not described in 'devlink_dpipe_headers'
devlink.h:384: warning: Function parameter or member 'headers_count' not described in 'devlink_dpipe_headers'
devlink.h:398: warning: Function parameter or member 'unit' not described in 'devlink_resource_size_params'
devlink.h:487: warning: Function parameter or member 'id' not described in 'devlink_param'
devlink.h:645: warning: Function parameter or member 'overwrite_mask' not described in 'devlink_flash_update_params'
Fixes: 1555d204e743 ("devlink: Support for pipeline debug (dpipe)") Fixes: d9f9b9a4d05f ("devlink: Add support for resource abstraction") Fixes: eabaef1896bc ("devlink: Add devlink_param register and unregister") Fixes: 5d5b4128c4ca ("devlink: introduce flash update overwrite mask") Signed-off-by: Randy Dunlap <[email protected]> Cc: Jiri Pirko <[email protected]> Cc: Moshe Shemesh <[email protected]> Cc: Jacob Keller <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Jakub Kicinski <[email protected]>
Randy Dunlap [Fri, 14 Jul 2023 04:51:21 +0000 (21:51 -0700)]
codel: fix kernel-doc notation warnings
Use '@' before the struct member names in kernel-doc notation
to prevent kernel-doc warnings.
codel.h:158: warning: Function parameter or member 'ecn_mark' not described in 'codel_stats'
codel.h:158: warning: Function parameter or member 'ce_mark' not described in 'codel_stats'
Randy Dunlap [Fri, 14 Jul 2023 04:51:20 +0000 (21:51 -0700)]
net: cfg802154: fix kernel-doc notation warnings
Add an enum heading to the kernel-doc comments to prevent
kernel-doc warnings.
cfg802154.h:174: warning: Cannot understand * @WPAN_PHY_FLAG_TRANSMIT_POWER: Indicates that transceiver will support
on line 174 - I thought it was a doc line
cfg802154.h:192: warning: Enum value 'WPAN_PHY_FLAG_TXPOWER' not described in enum 'wpan_phy_flags'
cfg802154.h:192: warning: Excess enum value 'WPAN_PHY_FLAG_TRANSMIT_POWER' description in 'wpan_phy_flags'
Randy Dunlap [Fri, 14 Jul 2023 04:51:19 +0000 (21:51 -0700)]
net: bonding: remove kernel-doc comment marker
Change an errant kernel-doc comment marker (/**) to a regular
comment to prevent a kernel-doc warning.
bonding.h:282: warning: This comment starts with '/**', but isn't a kernel-doc comment. Refer Documentation/doc-guide/kernel-doc.rst
* Returns NULL if the net_device does not belong to any of the bond's slaves
Calling ethtool during reload can lead to call trace, because VSI isn't
configured for some time, but netdev is alive.
To fix it add rtnl lock for VSI deconfig and config. Set ::num_q_vectors
to 0 after freeing and add a check for ::tx/rx_rings in ring related
ethtool ops.
Petr Oros [Mon, 19 Jun 2023 10:58:13 +0000 (12:58 +0200)]
ice: Unregister netdev and devlink_port only once
Since commit 6624e780a577fc ("ice: split ice_vsi_setup into smaller
functions") ice_vsi_release does things twice. There is unregister
netdev which is unregistered in ice_deinit_eth also.
It also unregisters the devlink_port twice which is also unregistered
in ice_deinit_eth(). This double deregistration is hidden because
devl_port_unregister ignores the return value of xa_erase.
Yan Zhai [Thu, 13 Jul 2023 17:28:00 +0000 (10:28 -0700)]
gso: fix dodgy bit handling for GSO_UDP_L4
Commit 1fd54773c267 ("udp: allow header check for dodgy GSO_UDP_L4
packets.") checks DODGY bit for UDP, but for packets that can be fed
directly to the device after gso_segs reset, it actually falls through
to fragmentation:
Daniel Golle [Thu, 13 Jul 2023 02:42:29 +0000 (03:42 +0100)]
net: ethernet: mtk_eth_soc: handle probe deferral
Move the call to of_get_ethdev_address to mtk_add_mac which is part of
the probe function and can hence itself return -EPROBE_DEFER should
of_get_ethdev_address return -EPROBE_DEFER. This allows us to entirely
get rid of the mtk_init function.
The problem of of_get_ethdev_address returning -EPROBE_DEFER surfaced
in situations in which the NVMEM provider holding the MAC address has
not yet be loaded at the time mtk_eth_soc is initially probed. In this
case probing of mtk_eth_soc should be deferred instead of falling back
to use a random MAC address, so once the NVMEM provider becomes
available probing can be repeated.
Fixes: 656e705243fd ("net-next: mediatek: add support for MT7623 ethernet") Signed-off-by: Daniel Golle <[email protected]> Signed-off-by: David S. Miller <[email protected]>
bridge: Add extack warning when enabling STP in netns.
When we create an L2 loop on a bridge in netns, we will see packets storm
even if STP is enabled.
# unshare -n
# ip link add br0 type bridge
# ip link add veth0 type veth peer name veth1
# ip link set veth0 master br0 up
# ip link set veth1 master br0 up
# ip link set br0 type bridge stp_state 1
# ip link set br0 up
# sleep 30
# ip -s link show br0
2: br0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP mode DEFAULT group default qlen 1000
link/ether b6:61:98:1c:1c:b5 brd ff:ff:ff:ff:ff:ff
RX: bytes packets errors dropped missed mcast 95655376812861249 0 0 0 12861249 <-. Keep
TX: bytes packets errors dropped carrier collsns | increasing 1027834 11951 0 0 0 0 <-' rapidly
This is because llc_rcv() drops all packets in non-root netns and BPDU
is dropped.
Let's add extack warning when enabling STP in netns.
# unshare -n
# ip link add br0 type bridge
# ip link set br0 type bridge stp_state 1
Warning: bridge: STP does not work in non-root netns.
Note this commit will be reverted later when we namespacify the whole LLC
infra.
CPSW ALE has 75 bit ALE entries which are stored within three 32 bit words.
The cpsw_ale_get_field() and cpsw_ale_set_field() functions assume that the
field will be strictly contained within one word. However, this is not
guaranteed to be the case and it is possible for ALE field entries to span
across up to two words at the most.
Fix the methods to handle getting/setting fields spanning up to two words.
Fixes: db82173f23c5 ("netdev: driver: ethernet: add cpsw address lookup engine support") Signed-off-by: Tanmay Patil <[email protected]>
[[email protected]: rephrased commit message and added Fixes tag] Signed-off-by: Siddharth Vadapalli <[email protected]> Signed-off-by: David S. Miller <[email protected]>
Mark Brown [Wed, 12 Jul 2023 11:16:16 +0000 (12:16 +0100)]
net: dsa: ar9331: Use explict flags for regmap single read/write
The at9331 is only able to read or write a single register at once. The
driver has a custom regmap bus and chooses to tell the regmap core about
this by reporting the maximum transfer sizes rather than the explicit
flags that exist at the regmap level. Since there are a number of
problems with the raw transfer limits and the regmap level flags are
better integrated anyway convert the driver to use the flags.
This bug is caused by the fact that usbnet trusts the bulk endpoint
addresses its probe routine receives in the driver_info structure, and
it does not check to see that these endpoints actually exist and have
the expected type and directions.
dsa: mv88e6xxx: Do a final check before timing out
I get sporadic timeouts from the driver when using the
MV88E6352. Reading the status again after the loop fixes the
problem: the operation is successful but goes undetected.
Some added prints show things like this:
[ 58.356209] mv88e6085 mdio_mux-0.1:00: Timeout while waiting
for switch, addr 1b reg 0b, mask 8000, val 0000, data c000
[ 58.367487] mv88e6085 mdio_mux-0.1:00: Timeout waiting for
ATU op 4000, fid 0001
(...)
[ 61.826293] mv88e6085 mdio_mux-0.1:00: Timeout while waiting
for switch, addr 1c reg 18, mask 8000, val 0000, data 9860
[ 61.837560] mv88e6085 mdio_mux-0.1:00: Timeout waiting
for PHY command 1860 to complete
The reason is probably not the commands: I think those are
mostly fine with the 50+50ms timeout, but the problem
appears when OpenWrt brings up several interfaces in
parallel on a system with 7 populated ports: if one of
them take more than 50 ms and waits one or more of the
others can get stuck on the mutex for the switch and then
this can easily multiply.
As we sleep and wait, the function loop needs a final
check after exiting the loop if we were successful.
Merge tag 'net-6.5-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net
Pull networking fixes from Paolo Abeni:
"Including fixes from netfilter, wireless and ebpf.
Current release - regressions:
- netfilter: conntrack: gre: don't set assured flag for clash entries
- wifi: iwlwifi: remove 'use_tfh' config to fix crash
Previous releases - regressions:
- ipv6: fix a potential refcount underflow for idev
- icmp6: ifix null-ptr-deref of ip6_null_entry->rt6i_idev in
icmp6_dev()
- bpf: fix max stack depth check for async callbacks
- eth: mlx5e:
- check for NOT_READY flag state after locking
- fix page_pool page fragment tracking for XDP
- eth: igc:
- fix tx hang issue when QBV gate is closed
- fix corner cases for TSN offload
- eth: octeontx2-af: Move validation of ptp pointer before its usage
- eth: ena: fix shift-out-of-bounds in exponential backoff
Previous releases - always broken:
- core: prevent skb corruption on frag list segmentation
- sched:
- cls_fw: fix improper refcount update leads to use-after-free
- sch_qfq: account for stab overhead in qfq_enqueue
- netfilter:
- report use refcount overflow
- prevent OOB access in nft_byteorder_eval
- wifi: mt7921e: fix init command fail with enabled device
- eth: ocelot: fix oversize frame dropping for preemptible TCs
- eth: fec: recycle pages for transmitted XDP frames"
* tag 'net-6.5-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (79 commits)
selftests: tc-testing: add test for qfq with stab overhead
net/sched: sch_qfq: account for stab overhead in qfq_enqueue
selftests: tc-testing: add tests for qfq mtu sanity check
net/sched: sch_qfq: reintroduce lmax bound check for MTU
wifi: cfg80211: fix receiving mesh packets without RFC1042 header
wifi: rtw89: debug: fix error code in rtw89_debug_priv_send_h2c_set()
net: txgbe: fix eeprom calculation error
net/sched: make psched_mtu() RTNL-less safe
net: ena: fix shift-out-of-bounds in exponential backoff
netdevsim: fix uninitialized data in nsim_dev_trap_fa_cookie_write()
net/sched: flower: Ensure both minimum and maximum ports are specified
MAINTAINERS: Add another mailing list for QUALCOMM ETHQOS ETHERNET DRIVER
docs: netdev: update the URL of the status page
wifi: iwlwifi: remove 'use_tfh' config to fix crash
xdp: use trusted arguments in XDP hints kfuncs
bpf: cpumap: Fix memory leak in cpu_map_update_elem
wifi: airo: avoid uninitialized warning in airo_get_rate()
octeontx2-pf: Add additional check for MCAM rules
net: dsa: Removed unneeded of_node_put in felix_parse_ports_node
net: fec: use netdev_err_once() instead of netdev_err()
...
Merge tag 'trace-v6.5-rc1-3' of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace
Pull tracing fixes from Steven Rostedt:
- Fix some missing-prototype warnings
- Fix user events struct args (did not include size of struct)
When creating a user event, the "struct" keyword is to denote that
the size of the field will be passed in. But the parsing failed to
handle this case.
- Add selftest to struct sizes for user events
- Fix sample code for direct trampolines.
The sample code for direct trampolines attached to handle_mm_fault().
But the prototype changed and the direct trampoline sample code was
not updated. Direct trampolines needs to have the arguments correct
otherwise it can fail or crash the system.
Due to backward compatibility, the structure used to save stack
traces in the kernel had a fixed size of 8. This structure is
exported to user space via the tracing format file. A change was made
to allow more than 8 functions to be recorded, and user space now
uses the size field to know how many functions are actually in the
stack.
But the structure still has size of 8 (even though it points into the
ring buffer that has the required amount allocated to hold a full
stack.
This was fine until the fortifier noticed that the
memcpy(&entry->caller, stack, size) was greater than the 8 functions
and would complain at runtime about it.
Hide this by using a pointer to the stack location on the ring buffer
instead of using the address of the entry structure caller field.
- Fix a deadloop in reading trace_pipe that was caused by a mismatch
between ring_buffer_empty() returning false which then asked to read
the data, but the read code uses rb_num_of_entries() that returned
zero, and causing a infinite "retry".
- Fix a warning caused by not using all pages allocated to store ftrace
functions, where this can happen if the linker inserts a bunch of
"NULL" entries, causing the accounting of how many pages needed to be
off.
- Fix histogram synthetic event crashing when the start event is
removed and the end event is still using a variable from it
- Fix memory leak in freeing iter->temp in tracing_release_pipe()
* tag 'trace-v6.5-rc1-3' of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace:
tracing: Fix memory leak of iter->temp when reading trace_pipe
tracing/histograms: Add histograms to hist_vars if they have referenced variables
tracing: Stop FORTIFY_SOURCE complaining about stack trace caller
ftrace: Fix possible warning on checking all pages used in ftrace_process_locs()
ring-buffer: Fix deadloop issue on reading trace_pipe
tracing: arm64: Avoid missing-prototype warnings
selftests/user_events: Test struct size match cases
tracing/user_events: Fix struct arg size match check
x86/ftrace: Remove unsued extern declaration ftrace_regs_caller_ret()
arm64: ftrace: Add direct call trampoline samples support
samples: ftrace: Save required argument registers in sample trampolines
Merge tag 'for-linus-6.5-rc2-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/xen/tip
Pull xen fixes from Juergen Gross:
- a cleanup of the Xen related ELF-notes
- a fix for virtio handling in Xen dom0 when running Xen in a VM
* tag 'for-linus-6.5-rc2-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/xen/tip:
xen/virtio: Fix NULL deref when a bridge of PCI root bus has no parent
x86/Xen: tidy xen-head.S
Merge tag 'sh-for-v6.5-tag2' of git://git.kernel.org/pub/scm/linux/kernel/git/glaubitz/sh-linux
Pull sh fixes from John Paul Adrian Glaubitz:
"The sh updates introduced multiple regressions.
In particular, the change a8ac2961148e ("sh: Avoid using IRQ0 on SH3
and SH4") causes several boards to hang during boot due to incorrect
IRQ numbers.
Geert Uytterhoeven has contributed patches that handle the virq offset
in the IRQ code for the dreamcast, highlander and r2d boards while
Artur Rojek has contributed a patch which handles the virq offset for
the hd64461 companion chip"
* tag 'sh-for-v6.5-tag2' of git://git.kernel.org/pub/scm/linux/kernel/git/glaubitz/sh-linux:
sh: hd64461: Handle virq offset for offchip IRQ base and HD64461 IRQ
sh: mach-dreamcast: Handle virq offset in cascaded IRQ demux
sh: mach-highlander: Handle virq offset in cascaded IRL demux
sh: mach-r2d: Handle virq offset in cascaded IRL demux
Pedro Tammela [Tue, 11 Jul 2023 21:01:02 +0000 (18:01 -0300)]
net/sched: sch_qfq: account for stab overhead in qfq_enqueue
Lion says:
-------
In the QFQ scheduler a similar issue to CVE-2023-31436
persists.
Consider the following code in net/sched/sch_qfq.c:
static int qfq_enqueue(struct sk_buff *skb, struct Qdisc *sch,
struct sk_buff **to_free)
{
unsigned int len = qdisc_pkt_len(skb), gso_segs;
// ...
if (unlikely(cl->agg->lmax < len)) {
pr_debug("qfq: increasing maxpkt from %u to %u for class %u",
cl->agg->lmax, len, cl->common.classid);
err = qfq_change_agg(sch, cl, cl->agg->class_weight, len);
if (err) {
cl->qstats.drops++;
return qdisc_drop(skb, sch, to_free);
}
// ...
}
Similarly to CVE-2023-31436, "lmax" is increased without any bounds
checks according to the packet length "len". Usually this would not
impose a problem because packet sizes are naturally limited.
This is however not the actual packet length, rather the
"qdisc_pkt_len(skb)" which might apply size transformations according to
"struct qdisc_size_table" as created by "qdisc_get_stab()" in
net/sched/sch_api.c if the TCA_STAB option was set when modifying the qdisc.
A user may choose virtually any size using such a table.
As a result the same issue as in CVE-2023-31436 can occur, allowing heap
out-of-bounds read / writes in the kmalloc-8192 cache.
-------
We can create the issue with the following commands:
tc qdisc add dev $DEV root handle 1: stab mtu 2048 tsize 512 mpu 0 \
overhead 999999999 linklayer ethernet qfq
tc class add dev $DEV parent 1: classid 1:1 htb rate 6mbit burst 15k
tc filter add dev $DEV parent 1: matchall classid 1:1
ping -I $DEV 1.1.1.2
This is caused by incorrectly assuming that qdisc_pkt_len() returns a
length within the QFQ_MIN_LMAX < len < QFQ_MAX_LMAX.
projects / linux.git / log