linux-stable.git/include/net, branch v6.9.3 Linux kernel stable tree Bluetooth: hci_core: Fix not handling hdev->le_num_of_adv_sets=1 2024-05-30T07:44:41+00:00 Luiz Augusto von Dentz luiz.von.dentz@intel.com 2024-05-13T20:07:55+00:00 0baa80343e78b9b69bb14d23d1dfe032dc95ad93 [ Upstream commit e77f43d531af41e9ce299eab10dcae8fa5dbc293 ] If hdev->le_num_of_adv_sets is set to 1 it means that only handle 0x00 can be used, but since the MGMT interface instances start from 1 (instance 0 means all instances in case of MGMT_OP_REMOVE_ADVERTISING) the code needs to map the instance to handle otherwise users will not be able to advertise as instance 1 would attempt to use handle 0x01. Fixes: 1d0fac2c38ed ("Bluetooth: Use controller sets when available") Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com> Signed-off-by: Sasha Levin <sashal@kernel.org> 
[ Upstream commit e77f43d531af41e9ce299eab10dcae8fa5dbc293 ]

If hdev->le_num_of_adv_sets is set to 1 it means that only handle 0x00
can be used, but since the MGMT interface instances start from 1
(instance 0 means all instances in case of MGMT_OP_REMOVE_ADVERTISING)
the code needs to map the instance to handle otherwise users will not be
able to advertise as instance 1 would attempt to use handle 0x01.

Fixes: 1d0fac2c38ed ("Bluetooth: Use controller sets when available")
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Bluetooth: hci_conn, hci_sync: Use __counted_by() to avoid -Wfamnae warnings 2024-05-30T07:44:41+00:00 Gustavo A. R. Silva gustavoars@kernel.org 2024-04-26T22:52:46+00:00 ec339f0c3472445003b47f28da951e119446146c [ Upstream commit c4585edf708edb5277a3cc4b8581ccb833f3307d ] Prepare for the coming implementation by GCC and Clang of the __counted_by attribute. Flexible array members annotated with __counted_by can have their accesses bounds-checked at run-time via CONFIG_UBSAN_BOUNDS (for array indexing) and CONFIG_FORTIFY_SOURCE (for strcpy/memcpy-family functions). Also, -Wflex-array-member-not-at-end is coming in GCC-14, and we are getting ready to enable it globally. So, use the `DEFINE_FLEX()` helper for multiple on-stack definitions of a flexible structure where the size of the flexible-array member is known at compile-time, and refactor the rest of the code, accordingly. Notice that, due to the use of `__counted_by()` in `struct hci_cp_le_create_cis`, the for loop in function `hci_cs_le_create_cis()` had to be modified. Once the index `i`, through which `cp->cis[i]` is accessed, falls in the interval [0, cp->num_cis), `cp->num_cis` cannot be decremented all the way down to zero while accessing `cp->cis[]`: net/bluetooth/hci_event.c:4310: 4310 for (i = 0; cp->num_cis; cp->num_cis--, i++) { ... 4314 handle = __le16_to_cpu(cp->cis[i].cis_handle); otherwise, only half (one iteration before `cp->num_cis == i`) or half plus one (one iteration before `cp->num_cis < i`) of the items in the array will be accessed before running into an out-of-bounds issue. So, in order to avoid this, set `cp->num_cis` to zero just after the for loop. Also, make use of `aux_num_cis` variable to update `cmd->num_cis` after a `list_for_each_entry_rcu()` loop. With these changes, fix the following warnings: net/bluetooth/hci_sync.c:1239:56: warning: structure containing a flexible array member is not at the end of another structure [-Wflex-array-member-not-at-end] net/bluetooth/hci_sync.c:1415:51: warning: structure containing a flexible array member is not at the end of another structure [-Wflex-array-member-not-at-end] net/bluetooth/hci_sync.c:1731:51: warning: structure containing a flexible array member is not at the end of another structure [-Wflex-array-member-not-at-end] net/bluetooth/hci_sync.c:6497:45: warning: structure containing a flexible array member is not at the end of another structure [-Wflex-array-member-not-at-end] Link: https://github.com/KSPP/linux/issues/202 Signed-off-by: Gustavo A. R. Silva <gustavoars@kernel.org> Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com> Stable-dep-of: e77f43d531af ("Bluetooth: hci_core: Fix not handling hdev->le_num_of_adv_sets=1") Signed-off-by: Sasha Levin <sashal@kernel.org> 
[ Upstream commit c4585edf708edb5277a3cc4b8581ccb833f3307d ]

Prepare for the coming implementation by GCC and Clang of the
__counted_by attribute. Flexible array members annotated with
__counted_by can have their accesses bounds-checked at run-time
via CONFIG_UBSAN_BOUNDS (for array indexing) and CONFIG_FORTIFY_SOURCE
(for strcpy/memcpy-family functions).

Also, -Wflex-array-member-not-at-end is coming in GCC-14, and we are
getting ready to enable it globally.

So, use the `DEFINE_FLEX()` helper for multiple on-stack definitions
of a flexible structure where the size of the flexible-array member
is known at compile-time, and refactor the rest of the code,
accordingly.

Notice that, due to the use of `__counted_by()` in `struct
hci_cp_le_create_cis`, the for loop in function `hci_cs_le_create_cis()`
had to be modified. Once the index `i`, through which `cp->cis[i]` is
accessed, falls in the interval [0, cp->num_cis), `cp->num_cis` cannot
be decremented all the way down to zero while accessing `cp->cis[]`:

net/bluetooth/hci_event.c:4310:
4310    for (i = 0; cp->num_cis; cp->num_cis--, i++) {
                ...
4314            handle = __le16_to_cpu(cp->cis[i].cis_handle);

otherwise, only half (one iteration before `cp->num_cis == i`) or half
plus one (one iteration before `cp->num_cis < i`) of the items in the
array will be accessed before running into an out-of-bounds issue. So,
in order to avoid this, set `cp->num_cis` to zero just after the for
loop.

Also, make use of `aux_num_cis` variable to update `cmd->num_cis` after
a `list_for_each_entry_rcu()` loop.

With these changes, fix the following warnings:
net/bluetooth/hci_sync.c:1239:56: warning: structure containing a flexible
array member is not at the end of another structure
[-Wflex-array-member-not-at-end]
net/bluetooth/hci_sync.c:1415:51: warning: structure containing a flexible
array member is not at the end of another structure
[-Wflex-array-member-not-at-end]
net/bluetooth/hci_sync.c:1731:51: warning: structure containing a flexible
array member is not at the end of another structure
[-Wflex-array-member-not-at-end]
net/bluetooth/hci_sync.c:6497:45: warning: structure containing a flexible
array member is not at the end of another structure
[-Wflex-array-member-not-at-end]

Link: https://github.com/KSPP/linux/issues/202
Signed-off-by: Gustavo A. R. Silva <gustavoars@kernel.org>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
Stable-dep-of: e77f43d531af ("Bluetooth: hci_core: Fix not handling hdev->le_num_of_adv_sets=1")
Signed-off-by: Sasha Levin <sashal@kernel.org>
Bluetooth: HCI: Remove HCI_AMP support 2024-05-30T07:44:41+00:00 Luiz Augusto von Dentz luiz.von.dentz@intel.com 2024-05-06T22:33:52+00:00 af1d425b6dc67cd67809f835dd7afb6be4d43e03 [ Upstream commit 84a4bb6548a29326564f0e659fb8064503ecc1c7 ] Since BT_HS has been remove HCI_AMP controllers no longer has any use so remove it along with the capability of creating AMP controllers. Since we no longer need to differentiate between AMP and Primary controllers, as only HCI_PRIMARY is left, this also remove hdev->dev_type altogether. Fixes: e7b02296fb40 ("Bluetooth: Remove BT_HS") Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com> Signed-off-by: Sasha Levin <sashal@kernel.org> 
[ Upstream commit 84a4bb6548a29326564f0e659fb8064503ecc1c7 ]

Since BT_HS has been remove HCI_AMP controllers no longer has any use so
remove it along with the capability of creating AMP controllers.

Since we no longer need to differentiate between AMP and Primary
controllers, as only HCI_PRIMARY is left, this also remove
hdev->dev_type altogether.

Fixes: e7b02296fb40 ("Bluetooth: Remove BT_HS")
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Bluetooth: ISO: Make iso_get_sock_listen generic 2024-05-30T07:44:40+00:00 Iulia Tanasescu iulia.tanasescu@nxp.com 2024-04-02T11:39:30+00:00 c581c331100ad5fe1d332e541b6b854a56450898 [ Upstream commit 311527e9dafdcae0c5a20d62f4f84ad01b33b5f4 ] This makes iso_get_sock_listen more generic, to return matching socket in the state provided as argument. Signed-off-by: Iulia Tanasescu <iulia.tanasescu@nxp.com> Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com> Stable-dep-of: a5b862c6a221 ("Bluetooth: L2CAP: Fix div-by-zero in l2cap_le_flowctl_init()") Signed-off-by: Sasha Levin <sashal@kernel.org> 
[ Upstream commit 311527e9dafdcae0c5a20d62f4f84ad01b33b5f4 ]

This makes iso_get_sock_listen more generic, to return matching socket
in the state provided as argument.

Signed-off-by: Iulia Tanasescu <iulia.tanasescu@nxp.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
Stable-dep-of: a5b862c6a221 ("Bluetooth: L2CAP: Fix div-by-zero in l2cap_le_flowctl_init()")
Signed-off-by: Sasha Levin <sashal@kernel.org>
Bluetooth: compute LE flow credits based on recvbuf space 2024-05-30T07:44:40+00:00 Sebastian Urban surban@surban.net 2024-05-01T10:08:58+00:00 0c287656fce837201c295fc4f9747603593615b2 [ Upstream commit ce60b9231b66710b6ee24042ded26efee120ecfc ] Previously LE flow credits were returned to the sender even if the socket's receive buffer was full. This meant that no back-pressure was applied to the sender, thus it continued to send data, resulting in data loss without any error being reported. Furthermore, the amount of credits was essentially fixed to a small amount, leading to reduced performance. This is fixed by computing the number of returned LE flow credits based on the estimated available space in the receive buffer of an L2CAP socket. Consequently, if the receive buffer is full, no credits are returned until the buffer is read and thus cleared by user-space. Since the computation of available receive buffer space can only be performed approximately (due to sk_buff overhead) and the receive buffer size may be changed by user-space after flow credits have been sent, superfluous received data is temporary stored within l2cap_pinfo. This is necessary because Bluetooth LE provides no retransmission mechanism once the data has been acked by the physical layer. If receive buffer space estimation is not possible at the moment, we fall back to providing credits for one full packet as before. This is currently the case during connection setup, when MPS is not yet available. Fixes: b1c325c23d75 ("Bluetooth: Implement returning of LE L2CAP credits") Signed-off-by: Sebastian Urban <surban@surban.net> Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com> Signed-off-by: Sasha Levin <sashal@kernel.org> 
[ Upstream commit ce60b9231b66710b6ee24042ded26efee120ecfc ]

Previously LE flow credits were returned to the
sender even if the socket's receive buffer was
full. This meant that no back-pressure
was applied to the sender, thus it continued to
send data, resulting in data loss without any
error being reported. Furthermore, the amount
of credits was essentially fixed to a small
amount, leading to reduced performance.

This is fixed by computing the number of returned
LE flow credits based on the estimated available
space in the receive buffer of an L2CAP socket.
Consequently, if the receive buffer is full, no
credits are returned until the buffer is read and
thus cleared by user-space.

Since the computation of available receive buffer
space can only be performed approximately (due to
sk_buff overhead) and the receive buffer size may
be changed by user-space after flow credits have
been sent, superfluous received data is temporary
stored within l2cap_pinfo. This is necessary
because Bluetooth LE provides no retransmission
mechanism once the data has been acked by the
physical layer.

If receive buffer space estimation is not possible
at the moment, we fall back to providing credits
for one full packet as before. This is currently
the case during connection setup, when MPS is not
yet available.

Fixes: b1c325c23d75 ("Bluetooth: Implement returning of LE L2CAP credits")
Signed-off-by: Sebastian Urban <surban@surban.net>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
ax25: Use kernel universal linked list to implement ax25_dev_list 2024-05-30T07:44:39+00:00 Duoming Zhou duoming@zju.edu.cn 2024-05-09T09:36:33+00:00 5954945db960bfb8edad4234c1d1a13fb48e7424 [ Upstream commit a7d6e36b9ad052926ba2ecba3a59d8bb67dabcb4 ] The origin ax25_dev_list implements its own single linked list, which is complicated and error-prone. For example, when deleting the node of ax25_dev_list in ax25_dev_device_down(), we have to operate on the head node and other nodes separately. This patch uses kernel universal linked list to replace original ax25_dev_list, which make the operation of ax25_dev_list easier. We should do "dev->ax25_ptr = ax25_dev;" and "dev->ax25_ptr = NULL;" while holding the spinlock, otherwise the ax25_dev_device_up() and ax25_dev_device_down() could race. Suggested-by: Dan Carpenter <dan.carpenter@linaro.org> Signed-off-by: Duoming Zhou <duoming@zju.edu.cn> Reviewed-by: Dan Carpenter <dan.carpenter@linaro.org> Link: https://lore.kernel.org/r/85bba3af651ca0e1a519da8d0d715b949891171c.1715247018.git.duoming@zju.edu.cn Signed-off-by: Jakub Kicinski <kuba@kernel.org> Stable-dep-of: b505e0319852 ("ax25: Fix reference count leak issues of ax25_dev") Signed-off-by: Sasha Levin <sashal@kernel.org> 
[ Upstream commit a7d6e36b9ad052926ba2ecba3a59d8bb67dabcb4 ]

The origin ax25_dev_list implements its own single linked list,
which is complicated and error-prone. For example, when deleting
the node of ax25_dev_list in ax25_dev_device_down(), we have to
operate on the head node and other nodes separately.

This patch uses kernel universal linked list to replace original
ax25_dev_list, which make the operation of ax25_dev_list easier.

We should do "dev->ax25_ptr = ax25_dev;" and "dev->ax25_ptr = NULL;"
while holding the spinlock, otherwise the ax25_dev_device_up() and
ax25_dev_device_down() could race.

Suggested-by: Dan Carpenter <dan.carpenter@linaro.org>
Signed-off-by: Duoming Zhou <duoming@zju.edu.cn>
Reviewed-by: Dan Carpenter <dan.carpenter@linaro.org>
Link: https://lore.kernel.org/r/85bba3af651ca0e1a519da8d0d715b949891171c.1715247018.git.duoming@zju.edu.cn
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Stable-dep-of: b505e0319852 ("ax25: Fix reference count leak issues of ax25_dev")
Signed-off-by: Sasha Levin <sashal@kernel.org>
tcp: increase the default TCP scaling ratio 2024-05-30T07:44:20+00:00 Hechao Li hli@netflix.com 2024-04-09T16:43:55+00:00 43bfd1952216d240d22810e4adc323c32cf64223 [ Upstream commit 697a6c8cec03c2299f850fa50322641a8bf6b915 ] After commit dfa2f0483360 ("tcp: get rid of sysctl_tcp_adv_win_scale"), we noticed an application-level timeout due to reduced throughput. Before the commit, for a client that sets SO_RCVBUF to 65k, it takes around 22 seconds to transfer 10M data. After the commit, it takes 40 seconds. Because our application has a 30-second timeout, this regression broke the application. The reason that it takes longer to transfer data is that tp->scaling_ratio is initialized to a value that results in ~0.25 of rcvbuf. In our case, SO_RCVBUF is set to 65536 by the application, which translates to 2 * 65536 = 131,072 bytes in rcvbuf and hence a ~28k initial receive window. Later, even though the scaling_ratio is updated to a more accurate skb->len/skb->truesize, which is ~0.66 in our environment, the window stays at ~0.25 * rcvbuf. This is because tp->window_clamp does not change together with the tp->scaling_ratio update when autotuning is disabled due to SO_RCVBUF. As a result, the window size is capped at the initial window_clamp, which is also ~0.25 * rcvbuf, and never grows bigger. Most modern applications let the kernel do autotuning, and benefit from the increased scaling_ratio. But there are applications such as kafka that has a default setting of SO_RCVBUF=64k. This patch increases the initial scaling_ratio from ~25% to 50% in order to make it backward compatible with the original default sysctl_tcp_adv_win_scale for applications setting SO_RCVBUF. Fixes: dfa2f0483360 ("tcp: get rid of sysctl_tcp_adv_win_scale") Signed-off-by: Hechao Li <hli@netflix.com> Reviewed-by: Tycho Andersen <tycho@tycho.pizza> Reviewed-by: Eric Dumazet <edumazet@google.com> Link: https://lore.kernel.org/netdev/20240402215405.432863-1-hli@netflix.com/ Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sashal@kernel.org> 
[ Upstream commit 697a6c8cec03c2299f850fa50322641a8bf6b915 ]

After commit dfa2f0483360 ("tcp: get rid of sysctl_tcp_adv_win_scale"),
we noticed an application-level timeout due to reduced throughput.

Before the commit, for a client that sets SO_RCVBUF to 65k, it takes
around 22 seconds to transfer 10M data. After the commit, it takes 40
seconds. Because our application has a 30-second timeout, this
regression broke the application.

The reason that it takes longer to transfer data is that
tp->scaling_ratio is initialized to a value that results in ~0.25 of
rcvbuf. In our case, SO_RCVBUF is set to 65536 by the application, which
translates to 2 * 65536 = 131,072 bytes in rcvbuf and hence a ~28k
initial receive window.

Later, even though the scaling_ratio is updated to a more accurate
skb->len/skb->truesize, which is ~0.66 in our environment, the window
stays at ~0.25 * rcvbuf. This is because tp->window_clamp does not
change together with the tp->scaling_ratio update when autotuning is
disabled due to SO_RCVBUF. As a result, the window size is capped at the
initial window_clamp, which is also ~0.25 * rcvbuf, and never grows
bigger.

Most modern applications let the kernel do autotuning, and benefit from
the increased scaling_ratio. But there are applications such as kafka
that has a default setting of SO_RCVBUF=64k.

This patch increases the initial scaling_ratio from ~25% to 50% in order
to make it backward compatible with the original default
sysctl_tcp_adv_win_scale for applications setting SO_RCVBUF.

Fixes: dfa2f0483360 ("tcp: get rid of sysctl_tcp_adv_win_scale")
Signed-off-by: Hechao Li <hli@netflix.com>
Reviewed-by: Tycho Andersen <tycho@tycho.pizza>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Link: https://lore.kernel.org/netdev/20240402215405.432863-1-hli@netflix.com/
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Bluetooth: L2CAP: Fix div-by-zero in l2cap_le_flowctl_init() 2024-05-25T14:30:54+00:00 Sungwoo Kim iam@sung-woo.kim 2024-05-04T19:23:29+00:00 d2b2f7d3936dc5990549bc36ab7ac7ac37f22c30 commit a5b862c6a221459d54e494e88965b48dcfa6cc44 upstream. l2cap_le_flowctl_init() can cause both div-by-zero and an integer overflow since hdev->le_mtu may not fall in the valid range. Move MTU from hci_dev to hci_conn to validate MTU and stop the connection process earlier if MTU is invalid. Also, add a missing validation in read_buffer_size() and make it return an error value if the validation fails. Now hci_conn_add() returns ERR_PTR() as it can fail due to the both a kzalloc failure and invalid MTU value. divide error: 0000 [#1] PREEMPT SMP KASAN NOPTI CPU: 0 PID: 67 Comm: kworker/u5:0 Tainted: G W 6.9.0-rc5+ #20 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.15.0-1 04/01/2014 Workqueue: hci0 hci_rx_work RIP: 0010:l2cap_le_flowctl_init+0x19e/0x3f0 net/bluetooth/l2cap_core.c:547 Code: e8 17 17 0c 00 66 41 89 9f 84 00 00 00 bf 01 00 00 00 41 b8 02 00 00 00 4c 89 fe 4c 89 e2 89 d9 e8 27 17 0c 00 44 89 f0 31 d2 <66> f7 f3 89 c3 ff c3 4d 8d b7 88 00 00 00 4c 89 f0 48 c1 e8 03 42 RSP: 0018:ffff88810bc0f858 EFLAGS: 00010246 RAX: 00000000000002a0 RBX: 0000000000000000 RCX: dffffc0000000000 RDX: 0000000000000000 RSI: ffff88810bc0f7c0 RDI: ffffc90002dcb66f RBP: ffff88810bc0f880 R08: aa69db2dda70ff01 R09: 0000ffaaaaaaaaaa R10: 0084000000ffaaaa R11: 0000000000000000 R12: ffff88810d65a084 R13: dffffc0000000000 R14: 00000000000002a0 R15: ffff88810d65a000 FS: 0000000000000000(0000) GS:ffff88811ac00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 0000000020000100 CR3: 0000000103268003 CR4: 0000000000770ef0 PKRU: 55555554 Call Trace: <TASK> l2cap_le_connect_req net/bluetooth/l2cap_core.c:4902 [inline] l2cap_le_sig_cmd net/bluetooth/l2cap_core.c:5420 [inline] l2cap_le_sig_channel net/bluetooth/l2cap_core.c:5486 [inline] l2cap_recv_frame+0xe59d/0x11710 net/bluetooth/l2cap_core.c:6809 l2cap_recv_acldata+0x544/0x10a0 net/bluetooth/l2cap_core.c:7506 hci_acldata_packet net/bluetooth/hci_core.c:3939 [inline] hci_rx_work+0x5e5/0xb20 net/bluetooth/hci_core.c:4176 process_one_work kernel/workqueue.c:3254 [inline] process_scheduled_works+0x90f/0x1530 kernel/workqueue.c:3335 worker_thread+0x926/0xe70 kernel/workqueue.c:3416 kthread+0x2e3/0x380 kernel/kthread.c:388 ret_from_fork+0x5c/0x90 arch/x86/kernel/process.c:147 ret_from_fork_asm+0x1a/0x30 arch/x86/entry/entry_64.S:244 </TASK> Modules linked in: ---[ end trace 0000000000000000 ]--- Fixes: 6ed58ec520ad ("Bluetooth: Use LE buffers for LE traffic") Suggested-by: Luiz Augusto von Dentz <luiz.dentz@gmail.com> Signed-off-by: Sungwoo Kim <iam@sung-woo.kim> Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> 
commit a5b862c6a221459d54e494e88965b48dcfa6cc44 upstream.

l2cap_le_flowctl_init() can cause both div-by-zero and an integer
overflow since hdev->le_mtu may not fall in the valid range.

Move MTU from hci_dev to hci_conn to validate MTU and stop the connection
process earlier if MTU is invalid.
Also, add a missing validation in read_buffer_size() and make it return
an error value if the validation fails.
Now hci_conn_add() returns ERR_PTR() as it can fail due to the both a
kzalloc failure and invalid MTU value.

divide error: 0000 [#1] PREEMPT SMP KASAN NOPTI
CPU: 0 PID: 67 Comm: kworker/u5:0 Tainted: G        W          6.9.0-rc5+ #20
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.15.0-1 04/01/2014
Workqueue: hci0 hci_rx_work
RIP: 0010:l2cap_le_flowctl_init+0x19e/0x3f0 net/bluetooth/l2cap_core.c:547
Code: e8 17 17 0c 00 66 41 89 9f 84 00 00 00 bf 01 00 00 00 41 b8 02 00 00 00 4c
89 fe 4c 89 e2 89 d9 e8 27 17 0c 00 44 89 f0 31 d2 <66> f7 f3 89 c3 ff c3 4d 8d
b7 88 00 00 00 4c 89 f0 48 c1 e8 03 42
RSP: 0018:ffff88810bc0f858 EFLAGS: 00010246
RAX: 00000000000002a0 RBX: 0000000000000000 RCX: dffffc0000000000
RDX: 0000000000000000 RSI: ffff88810bc0f7c0 RDI: ffffc90002dcb66f
RBP: ffff88810bc0f880 R08: aa69db2dda70ff01 R09: 0000ffaaaaaaaaaa
R10: 0084000000ffaaaa R11: 0000000000000000 R12: ffff88810d65a084
R13: dffffc0000000000 R14: 00000000000002a0 R15: ffff88810d65a000
FS:  0000000000000000(0000) GS:ffff88811ac00000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 0000000020000100 CR3: 0000000103268003 CR4: 0000000000770ef0
PKRU: 55555554
Call Trace:
 <TASK>
 l2cap_le_connect_req net/bluetooth/l2cap_core.c:4902 [inline]
 l2cap_le_sig_cmd net/bluetooth/l2cap_core.c:5420 [inline]
 l2cap_le_sig_channel net/bluetooth/l2cap_core.c:5486 [inline]
 l2cap_recv_frame+0xe59d/0x11710 net/bluetooth/l2cap_core.c:6809
 l2cap_recv_acldata+0x544/0x10a0 net/bluetooth/l2cap_core.c:7506
 hci_acldata_packet net/bluetooth/hci_core.c:3939 [inline]
 hci_rx_work+0x5e5/0xb20 net/bluetooth/hci_core.c:4176
 process_one_work kernel/workqueue.c:3254 [inline]
 process_scheduled_works+0x90f/0x1530 kernel/workqueue.c:3335
 worker_thread+0x926/0xe70 kernel/workqueue.c:3416
 kthread+0x2e3/0x380 kernel/kthread.c:388
 ret_from_fork+0x5c/0x90 arch/x86/kernel/process.c:147
 ret_from_fork_asm+0x1a/0x30 arch/x86/entry/entry_64.S:244
 </TASK>
Modules linked in:
---[ end trace 0000000000000000 ]---

Fixes: 6ed58ec520ad ("Bluetooth: Use LE buffers for LE traffic")
Suggested-by: Luiz Augusto von Dentz <luiz.dentz@gmail.com>
Signed-off-by: Sungwoo Kim <iam@sung-woo.kim>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Merge tag 'ipsec-2024-05-02' of git://git.kernel.org/pub/scm/linux/kernel/git/klassert/ipsec 2024-05-03T22:56:15+00:00 Jakub Kicinski kuba@kernel.org 2024-05-03T22:56:15+00:00 d0de61673908765f4e189fe26f4de6e87caf00e5 Steffen Klassert says: ==================== pull request (net): ipsec 2024-05-02 1) Fix an error pointer dereference in xfrm_in_fwd_icmp. From Antony Antony. 2) Preserve vlan tags for ESP transport mode software GRO. From Paul Davey. 3) Fix a spelling mistake in an uapi xfrm.h comment. From Anotny Antony. * tag 'ipsec-2024-05-02' of git://git.kernel.org/pub/scm/linux/kernel/git/klassert/ipsec: xfrm: Correct spelling mistake in xfrm.h comment xfrm: Preserve vlan tags for transport mode software GRO xfrm: fix possible derferencing in error path ==================== Link: https://lore.kernel.org/r/20240502084838.2269355-1-steffen.klassert@secunet.com Signed-off-by: Jakub Kicinski <kuba@kernel.org> 
Steffen Klassert says:

====================
pull request (net): ipsec 2024-05-02

1) Fix an error pointer dereference in xfrm_in_fwd_icmp.
   From Antony Antony.

2) Preserve vlan tags for ESP transport mode software GRO.
   From Paul Davey.

3) Fix a spelling mistake in an uapi xfrm.h comment.
   From Anotny Antony.

* tag 'ipsec-2024-05-02' of git://git.kernel.org/pub/scm/linux/kernel/git/klassert/ipsec:
  xfrm: Correct spelling mistake in xfrm.h comment
  xfrm: Preserve vlan tags for transport mode software GRO
  xfrm: fix possible derferencing in error path
====================

Link: https://lore.kernel.org/r/20240502084838.2269355-1-steffen.klassert@secunet.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
net: gro: fix udp bad offset in socket lookup by adding {inner_}network_offset to napi_gro_cb 2024-05-02T09:02:48+00:00 Richard Gobert richardbgobert@gmail.com 2024-04-30T14:35:54+00:00 5ef31ea5d053a8f493a772ebad3f3ce82c35d845 Commits a602456 ("udp: Add GRO functions to UDP socket") and 57c67ff ("udp: additional GRO support") introduce incorrect usage of {ip,ipv6}_hdr in the complete phase of gro. The functions always return skb->network_header, which in the case of encapsulated packets at the gro complete phase, is always set to the innermost L3 of the packet. That means that calling {ip,ipv6}_hdr for skbs which completed the GRO receive phase (both in gro_list and *_gro_complete) when parsing an encapsulated packet's _outer_ L3/L4 may return an unexpected value. This incorrect usage leads to a bug in GRO's UDP socket lookup. udp{4,6}_lib_lookup_skb functions use ip_hdr/ipv6_hdr respectively. These *_hdr functions return network_header which will point to the innermost L3, resulting in the wrong offset being used in __udp{4,6}_lib_lookup with encapsulated packets. This patch adds network_offset and inner_network_offset to napi_gro_cb, and makes sure both are set correctly. To fix the issue, network_offsets union is used inside napi_gro_cb, in which both the outer and the inner network offsets are saved. Reproduction example: Endpoint configuration example (fou + local address bind) # ip fou add port 6666 ipproto 4 # ip link add name tun1 type ipip remote 2.2.2.1 local 2.2.2.2 encap fou encap-dport 5555 encap-sport 6666 mode ipip # ip link set tun1 up # ip a add 1.1.1.2/24 dev tun1 Netperf TCP_STREAM result on net-next before patch is applied: net-next main, GRO enabled: $ netperf -H 1.1.1.2 -t TCP_STREAM -l 5 Recv Send Send Socket Socket Message Elapsed Size Size Size Time Throughput bytes bytes bytes secs. 10^6bits/sec 131072 16384 16384 5.28 2.37 net-next main, GRO disabled: $ netperf -H 1.1.1.2 -t TCP_STREAM -l 5 Recv Send Send Socket Socket Message Elapsed Size Size Size Time Throughput bytes bytes bytes secs. 10^6bits/sec 131072 16384 16384 5.01 2745.06 patch applied, GRO enabled: $ netperf -H 1.1.1.2 -t TCP_STREAM -l 5 Recv Send Send Socket Socket Message Elapsed Size Size Size Time Throughput bytes bytes bytes secs. 10^6bits/sec 131072 16384 16384 5.01 2877.38 Fixes: a6024562ffd7 ("udp: Add GRO functions to UDP socket") Signed-off-by: Richard Gobert <richardbgobert@gmail.com> Reviewed-by: Eric Dumazet <edumazet@google.com> Reviewed-by: Willem de Bruijn <willemb@google.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com> 
Commits a602456 ("udp: Add GRO functions to UDP socket") and 57c67ff ("udp:
additional GRO support") introduce incorrect usage of {ip,ipv6}_hdr in the
complete phase of gro. The functions always return skb->network_header,
which in the case of encapsulated packets at the gro complete phase, is
always set to the innermost L3 of the packet. That means that calling
{ip,ipv6}_hdr for skbs which completed the GRO receive phase (both in
gro_list and *_gro_complete) when parsing an encapsulated packet's _outer_
L3/L4 may return an unexpected value.

This incorrect usage leads to a bug in GRO's UDP socket lookup.
udp{4,6}_lib_lookup_skb functions use ip_hdr/ipv6_hdr respectively. These
*_hdr functions return network_header which will point to the innermost L3,
resulting in the wrong offset being used in __udp{4,6}_lib_lookup with
encapsulated packets.

This patch adds network_offset and inner_network_offset to napi_gro_cb, and
makes sure both are set correctly.

To fix the issue, network_offsets union is used inside napi_gro_cb, in
which both the outer and the inner network offsets are saved.

Reproduction example:

Endpoint configuration example (fou + local address bind)

    # ip fou add port 6666 ipproto 4
    # ip link add name tun1 type ipip remote 2.2.2.1 local 2.2.2.2 encap fou encap-dport 5555 encap-sport 6666 mode ipip
    # ip link set tun1 up
    # ip a add 1.1.1.2/24 dev tun1

Netperf TCP_STREAM result on net-next before patch is applied:

net-next main, GRO enabled:
    $ netperf -H 1.1.1.2 -t TCP_STREAM -l 5
    Recv   Send    Send
    Socket Socket  Message  Elapsed
    Size   Size    Size     Time     Throughput
    bytes  bytes   bytes    secs.    10^6bits/sec

    131072  16384  16384    5.28        2.37

net-next main, GRO disabled:
    $ netperf -H 1.1.1.2 -t TCP_STREAM -l 5
    Recv   Send    Send
    Socket Socket  Message  Elapsed
    Size   Size    Size     Time     Throughput
    bytes  bytes   bytes    secs.    10^6bits/sec

    131072  16384  16384    5.01     2745.06

patch applied, GRO enabled:
    $ netperf -H 1.1.1.2 -t TCP_STREAM -l 5
    Recv   Send    Send
    Socket Socket  Message  Elapsed
    Size   Size    Size     Time     Throughput
    bytes  bytes   bytes    secs.    10^6bits/sec

    131072  16384  16384    5.01     2877.38

Fixes: a6024562ffd7 ("udp: Add GRO functions to UDP socket")
Signed-off-by: Richard Gobert <richardbgobert@gmail.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>