linux.git/net, branch v6.1.98

inet_diag: Initialize pad field in struct inet_diag_req_v2

2024-07-11T10:47:12+00:00

[ Upstream commit 61cf1c739f08190a4cbf047b9fbb192a94d87e3f ]

KMSAN reported uninit-value access in raw_lookup() [1]. Diag for raw
sockets uses the pad field in struct inet_diag_req_v2 for the
underlying protocol. This field corresponds to the sdiag_raw_protocol
field in struct inet_diag_req_raw.

inet_diag_get_exact_compat() converts inet_diag_req to
inet_diag_req_v2, but leaves the pad field uninitialized. So the issue
occurs when raw_lookup() accesses the sdiag_raw_protocol field.

Fix this by initializing the pad field in
inet_diag_get_exact_compat(). Also, do the same fix in
inet_diag_dump_compat() to avoid the similar issue in the future.

[1]
BUG: KMSAN: uninit-value in raw_lookup net/ipv4/raw_diag.c:49 [inline]
BUG: KMSAN: uninit-value in raw_sock_get+0x657/0x800 net/ipv4/raw_diag.c:71
 raw_lookup net/ipv4/raw_diag.c:49 [inline]
 raw_sock_get+0x657/0x800 net/ipv4/raw_diag.c:71
 raw_diag_dump_one+0xa1/0x660 net/ipv4/raw_diag.c:99
 inet_diag_cmd_exact+0x7d9/0x980
 inet_diag_get_exact_compat net/ipv4/inet_diag.c:1404 [inline]
 inet_diag_rcv_msg_compat+0x469/0x530 net/ipv4/inet_diag.c:1426
 sock_diag_rcv_msg+0x23d/0x740 net/core/sock_diag.c:282
 netlink_rcv_skb+0x537/0x670 net/netlink/af_netlink.c:2564
 sock_diag_rcv+0x35/0x40 net/core/sock_diag.c:297
 netlink_unicast_kernel net/netlink/af_netlink.c:1335 [inline]
 netlink_unicast+0xe74/0x1240 net/netlink/af_netlink.c:1361
 netlink_sendmsg+0x10c6/0x1260 net/netlink/af_netlink.c:1905
 sock_sendmsg_nosec net/socket.c:730 [inline]
 __sock_sendmsg+0x332/0x3d0 net/socket.c:745
 ____sys_sendmsg+0x7f0/0xb70 net/socket.c:2585
 ___sys_sendmsg+0x271/0x3b0 net/socket.c:2639
 __sys_sendmsg net/socket.c:2668 [inline]
 __do_sys_sendmsg net/socket.c:2677 [inline]
 __se_sys_sendmsg net/socket.c:2675 [inline]
 __x64_sys_sendmsg+0x27e/0x4a0 net/socket.c:2675
 x64_sys_call+0x135e/0x3ce0 arch/x86/include/generated/asm/syscalls_64.h:47
 do_syscall_x64 arch/x86/entry/common.c:52 [inline]
 do_syscall_64+0xd9/0x1e0 arch/x86/entry/common.c:83
 entry_SYSCALL_64_after_hwframe+0x77/0x7f

Uninit was stored to memory at:
 raw_sock_get+0x650/0x800 net/ipv4/raw_diag.c:71
 raw_diag_dump_one+0xa1/0x660 net/ipv4/raw_diag.c:99
 inet_diag_cmd_exact+0x7d9/0x980
 inet_diag_get_exact_compat net/ipv4/inet_diag.c:1404 [inline]
 inet_diag_rcv_msg_compat+0x469/0x530 net/ipv4/inet_diag.c:1426
 sock_diag_rcv_msg+0x23d/0x740 net/core/sock_diag.c:282
 netlink_rcv_skb+0x537/0x670 net/netlink/af_netlink.c:2564
 sock_diag_rcv+0x35/0x40 net/core/sock_diag.c:297
 netlink_unicast_kernel net/netlink/af_netlink.c:1335 [inline]
 netlink_unicast+0xe74/0x1240 net/netlink/af_netlink.c:1361
 netlink_sendmsg+0x10c6/0x1260 net/netlink/af_netlink.c:1905
 sock_sendmsg_nosec net/socket.c:730 [inline]
 __sock_sendmsg+0x332/0x3d0 net/socket.c:745
 ____sys_sendmsg+0x7f0/0xb70 net/socket.c:2585
 ___sys_sendmsg+0x271/0x3b0 net/socket.c:2639
 __sys_sendmsg net/socket.c:2668 [inline]
 __do_sys_sendmsg net/socket.c:2677 [inline]
 __se_sys_sendmsg net/socket.c:2675 [inline]
 __x64_sys_sendmsg+0x27e/0x4a0 net/socket.c:2675
 x64_sys_call+0x135e/0x3ce0 arch/x86/include/generated/asm/syscalls_64.h:47
 do_syscall_x64 arch/x86/entry/common.c:52 [inline]
 do_syscall_64+0xd9/0x1e0 arch/x86/entry/common.c:83
 entry_SYSCALL_64_after_hwframe+0x77/0x7f

Local variable req.i created at:
 inet_diag_get_exact_compat net/ipv4/inet_diag.c:1396 [inline]
 inet_diag_rcv_msg_compat+0x2a6/0x530 net/ipv4/inet_diag.c:1426
 sock_diag_rcv_msg+0x23d/0x740 net/core/sock_diag.c:282

CPU: 1 PID: 8888 Comm: syz-executor.6 Not tainted 6.10.0-rc4-00217-g35bb670d65fc #32
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.16.3-2.fc40 04/01/2014

Fixes: 432490f9d455 ("net: ip, diag -- Add diag interface for raw sockets")
Reported-by: syzkaller 
Signed-off-by: Shigeru Yoshida 
Reviewed-by: Eric Dumazet 
Link: https://patch.msgid.link/20240703091649.111773-1-syoshida@redhat.com
Signed-off-by: Paolo Abeni 
Signed-off-by: Sasha Levin

netfilter: nf_tables: unconditionally flush pending work before notifier

2024-07-11T10:47:12+00:00

[ Upstream commit 9f6958ba2e902f9820c594869bd710ba74b7c4c0 ]

syzbot reports:

KASAN: slab-uaf in nft_ctx_update include/net/netfilter/nf_tables.h:1831
KASAN: slab-uaf in nft_commit_release net/netfilter/nf_tables_api.c:9530
KASAN: slab-uaf int nf_tables_trans_destroy_work+0x152b/0x1750 net/netfilter/nf_tables_api.c:9597
Read of size 2 at addr ffff88802b0051c4 by task kworker/1:1/45
[..]
Workqueue: events nf_tables_trans_destroy_work
Call Trace:
 nft_ctx_update include/net/netfilter/nf_tables.h:1831 [inline]
 nft_commit_release net/netfilter/nf_tables_api.c:9530 [inline]
 nf_tables_trans_destroy_work+0x152b/0x1750 net/netfilter/nf_tables_api.c:9597

Problem is that the notifier does a conditional flush, but its possible
that the table-to-be-removed is still referenced by transactions being
processed by the worker, so we need to flush unconditionally.

We could make the flush_work depend on whether we found a table to delete
in nf-next to avoid the flush for most cases.

AFAICS this problem is only exposed in nf-next, with
commit e169285f8c56 ("netfilter: nf_tables: do not store nft_ctx in transaction objects"),
with this commit applied there is an unconditional fetch of
table->family which is whats triggering the above splat.

Fixes: 2c9f0293280e ("netfilter: nf_tables: flush pending destroy work before netlink notifier")
Reported-and-tested-by: syzbot+4fd66a69358fc15ae2ad@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=4fd66a69358fc15ae2ad
Signed-off-by: Florian Westphal 
Signed-off-by: Pablo Neira Ayuso 
Signed-off-by: Sasha Levin

net: allow skb_datagram_iter to be called from any context

2024-07-11T10:47:11+00:00

[ Upstream commit d2d30a376d9cc94c6fb730c58b3e5b7426ecb6de ]

We only use the mapping in a single context, so kmap_local is sufficient
and cheaper. Make sure to use skb_frag_foreach_page as skb frags may
contain compound pages and we need to map page by page.

Reported-by: kernel test robot 
Closes: https://lore.kernel.org/oe-lkp/202406161539.b5ff7b20-oliver.sang@intel.com
Fixes: 950fcaecd5cc ("datagram: consolidate datagram copy to iter helpers")
Signed-off-by: Sagi Grimberg 
Link: https://patch.msgid.link/20240626100008.831849-1-sagi@grimberg.me
Signed-off-by: Paolo Abeni 
Signed-off-by: Sasha Levin

tcp_metrics: validate source addr length

2024-07-11T10:47:11+00:00

[ Upstream commit 66be40e622e177316ae81717aa30057ba9e61dff ]

I don't see anything checking that TCP_METRICS_ATTR_SADDR_IPV4
is at least 4 bytes long, and the policy doesn't have an entry
for this attribute at all (neither does it for IPv6 but v6 is
manually validated).

Reviewed-by: Eric Dumazet 
Fixes: 3e7013ddf55a ("tcp: metrics: Allow selective get/del of tcp-metrics based on src IP")
Signed-off-by: Jakub Kicinski 
Signed-off-by: David S. Miller 
Signed-off-by: Sasha Levin

UPSTREAM: tcp: fix DSACK undo in fast recovery to call tcp_try_to_open()

2024-07-11T10:47:10+00:00

[ Upstream commit a6458ab7fd4f427d4f6f54380453ad255b7fde83 ]

In some production workloads we noticed that connections could
sometimes close extremely prematurely with ETIMEDOUT after
transmitting only 1 TLP and RTO retransmission (when we would normally
expect roughly tcp_retries2 = TCP_RETR2 = 15 RTOs before a connection
closes with ETIMEDOUT).

From tracing we determined that these workloads can suffer from a
scenario where in fast recovery, after some retransmits, a DSACK undo
can happen at a point where the scoreboard is totally clear (we have
retrans_out == sacked_out == lost_out == 0). In such cases, calling
tcp_try_keep_open() means that we do not execute any code path that
clears tp->retrans_stamp to 0. That means that tp->retrans_stamp can
remain erroneously set to the start time of the undone fast recovery,
even after the fast recovery is undone. If minutes or hours elapse,
and then a TLP/RTO/RTO sequence occurs, then the start_ts value in
retransmits_timed_out() (which is from tp->retrans_stamp) will be
erroneously ancient (left over from the fast recovery undone via
DSACKs). Thus this ancient tp->retrans_stamp value can cause the
connection to die very prematurely with ETIMEDOUT via
tcp_write_err().

The fix: we change DSACK undo in fast recovery (TCP_CA_Recovery) to
call tcp_try_to_open() instead of tcp_try_keep_open(). This ensures
that if no retransmits are in flight at the time of DSACK undo in fast
recovery then we properly zero retrans_stamp. Note that calling
tcp_try_to_open() is more consistent with other loss recovery
behavior, since normal fast recovery (CA_Recovery) and RTO recovery
(CA_Loss) both normally end when tp->snd_una meets or exceeds
tp->high_seq and then in tcp_fastretrans_alert() the "default" switch
case executes tcp_try_to_open(). Also note that by inspection this
change to call tcp_try_to_open() implies at least one other nice bug
fix, where now an ECE-marked DSACK that causes an undo will properly
invoke tcp_enter_cwr() rather than ignoring the ECE mark.

Fixes: c7d9d6a185a7 ("tcp: undo on DSACK during recovery")
Signed-off-by: Neal Cardwell 
Signed-off-by: Yuchung Cheng 
Signed-off-by: Eric Dumazet 
Signed-off-by: David S. Miller 
Signed-off-by: Sasha Levin

mac802154: fix time calculation in ieee802154_configure_durations()

2024-07-11T10:47:10+00:00

[ Upstream commit 07aa33988ad92fef79056f5ec30b9a0e4364b616 ]

Since 'symbol_duration' of 'struct wpan_phy' is in nanoseconds but
'lifs_period' and 'sifs_period' are both in microseconds, fix time
calculation in 'ieee802154_configure_durations()' and use convenient
'NSEC_PER_USEC' in 'ieee802154_setup_wpan_phy_pib()' as well.
Compile tested only.

Found by Linux Verification Center (linuxtesting.org) with SVACE.

Fixes: 781830c800dd ("net: mac802154: Set durations automatically")
Signed-off-by: Dmitry Antipov 
Acked-by: Miquel Raynal 
Message-ID: <20240508114010.219527-1-dmantipov@yandex.ru>
Signed-off-by: Stefan Schmidt 
Signed-off-by: Sasha Levin

sctp: prefer struct_size over open coded arithmetic

2024-07-11T10:47:07+00:00

[ Upstream commit e5c5f3596de224422561d48eba6ece5210d967b3 ]

This is an effort to get rid of all multiplications from allocation
functions in order to prevent integer overflows [1][2].

As the "ids" variable is a pointer to "struct sctp_assoc_ids" and this
structure ends in a flexible array:

struct sctp_assoc_ids {
	[...]
	sctp_assoc_t	gaids_assoc_id[];
};

the preferred way in the kernel is to use the struct_size() helper to
do the arithmetic instead of the calculation "size + size * count" in
the kmalloc() function.

Also, refactor the code adding the "ids_size" variable to avoid sizing
twice.

This way, the code is more readable and safer.

This code was detected with the help of Coccinelle, and audited and
modified manually.

Link: https://www.kernel.org/doc/html/latest/process/deprecated.html#open-coded-arithmetic-in-allocator-arguments [1]
Link: https://github.com/KSPP/linux/issues/160 [2]
Signed-off-by: Erick Archer 
Acked-by: Xin Long 
Reviewed-by: Kees Cook 
Link: https://lore.kernel.org/r/PAXPR02MB724871DB78375AB06B5171C88B152@PAXPR02MB7248.eurprd02.prod.outlook.com
Signed-off-by: Paolo Abeni 
Signed-off-by: Sasha Levin

batman-adv: Don't accept TT entries for out-of-spec VIDs

2024-07-05T07:31:58+00:00

commit 537a350d14321c8cca5efbf0a33a404fec3a9f9e upstream.

The internal handling of VLAN IDs in batman-adv is only specified for
following encodings:

* VLAN is used
  - bit 15 is 1
  - bit 11 - bit 0 is the VLAN ID (0-4095)
  - remaining bits are 0
* No VLAN is used
  - bit 15 is 0
  - remaining bits are 0

batman-adv was only preparing new translation table entries (based on its
soft interface information) using this encoding format. But the receive
path was never checking if entries in the roam or TT TVLVs were also
following this encoding.

It was therefore possible to create more than the expected maximum of 4096
+ 1 entries in the originator VLAN list. Simply by setting the "remaining
bits" to "random" values in corresponding TVLV.

Cc: stable@vger.kernel.org
Fixes: 7ea7b4a14275 ("batman-adv: make the TT CRC logic VLAN specific")
Reported-by: Linus Lüssing 
Signed-off-by: Sven Eckelmann 
Signed-off-by: Simon Wunderlich 
Signed-off-by: Greg Kroah-Hartman

net: can: j1939: enhanced error handling for tightly received RTS messages in xtp_rx_rts_session_new

2024-07-05T07:31:56+00:00

commit d3e2904f71ea0fe7eaff1d68a2b0363c888ea0fb upstream.

This patch enhances error handling in scenarios with RTS (Request to
Send) messages arriving closely. It replaces the less informative WARN_ON_ONCE
backtraces with a new error handling method. This provides clearer error
messages and allows for the early termination of problematic sessions.
Previously, sessions were only released at the end of j1939_xtp_rx_rts().

Potentially this could be reproduced with something like:
testj1939 -r vcan0:0x80 &
while true; do
	# send first RTS
	cansend vcan0 18EC8090#1014000303002301;
	# send second RTS
	cansend vcan0 18EC8090#1014000303002301;
	# send abort
	cansend vcan0 18EC8090#ff00000000002301;
done

Fixes: 9d71dd0c7009 ("can: add support of SAE J1939 protocol")
Reported-by: syzbot+daa36413a5cedf799ae4@syzkaller.appspotmail.com
Cc: stable@vger.kernel.org
Signed-off-by: Oleksij Rempel 
Link: https://lore.kernel.org/all/20231117124959.961171-1-o.rempel@pengutronix.de
Signed-off-by: Marc Kleine-Budde 
Signed-off-by: Greg Kroah-Hartman

net: can: j1939: recover socket queue on CAN bus error during BAM transmission

2024-07-05T07:31:56+00:00

commit 9ad1da14ab3bf23087ae45fe399d84a109ddb81a upstream.

Addresses an issue where a CAN bus error during a BAM transmission
could stall the socket queue, preventing further transmissions even
after the bus error is resolved. The fix activates the next queued
session after the error recovery, allowing communication to continue.

Fixes: 9d71dd0c70099 ("can: add support of SAE J1939 protocol")
Cc: stable@vger.kernel.org
Reported-by: Alexander Hölzl 
Tested-by: Alexander Hölzl 
Signed-off-by: Oleksij Rempel 
Link: https://lore.kernel.org/all/20240528070648.1947203-1-o.rempel@pengutronix.de
Cc: stable@vger.kernel.org
Signed-off-by: Marc Kleine-Budde 
Signed-off-by: Greg Kroah-Hartman