[ Upstream commit 7ffc748115 ]
rhashtable does not provide stable walk, duplicated elements are
possible in case of resizing. I considered that checking for errors when
calling rhashtable_walk_next() was sufficient to detect the resizing.
However, rhashtable_walk_next() returns -EAGAIN only at the end of the
iteration, which is too late, because a gc work containing duplicated
elements could have been already scheduled for removal to the worker.
Add a u32 gc worker sequence number per set, bump it on every workqueue
run. Annotate gc worker sequence number on the expired element. Use it
to skip those already seen in this gc workqueue run.
Note that this new field is never reset in case gc transaction fails, so
next gc worker run on the expired element overrides it. Wraparound of gc
worker sequence number should not be an issue with stale gc worker
sequence number in the element, that would just postpone the element
removal in one gc run.
Note that it is not possible to use flags to annotate that element is
pending gc run to detect duplicates, given that gc transaction can be
invalidated in case of update from the control plane, therefore, not
allowing to clear such flag.
On x86_64, pahole reports no changes in the size of nft_rhash_elem.
Fixes: f6c383b8c3 ("netfilter: nf_tables: adapt set backend to use GC transaction API")
Reported-by: Laurent Fasnacht <laurent.fasnacht@proton.ch>
Tested-by: Laurent Fasnacht <laurent.fasnacht@proton.ch>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 456f010bfa ]
User space may unload ip_set.ko while it is itself requesting a set type
backend module, leading to a kernel crash. The race condition may be
provoked by inserting an mdelay() right after the nfnl_unlock() call.
Fixes: a7b4f989a6 ("netfilter: ipset: IP set core support")
Signed-off-by: Phil Sutter <phil@nwl.cc>
Acked-by: Jozsef Kadlecsik <kadlec@netfilter.org>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 2922078094 ]
When matching erspan_opt in cls_flower, only the (version, dir, hwid)
fields are relevant. However, in fl_set_erspan_opt() it initializes
all bits of erspan_opt and its mask to 1. This inadvertently requires
packets to match not only the (version, dir, hwid) fields but also the
other fields that are unexpectedly set to 1.
This patch resolves the issue by ensuring that only the (version, dir,
hwid) fields are configured in fl_set_erspan_opt(), leaving the other
fields to 0 in erspan_opt.
Fixes: 79b1011cb3 ("net: sched: allow flower to match erspan options")
Reported-by: Shuang Li <shuali@redhat.com>
Signed-off-by: Xin Long <lucien.xin@gmail.com>
Reviewed-by: Cong Wang <cong.wang@bytedance.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 2c7f14ed9c ]
We encountered a LGR/link use-after-free issue, which manifested as
the LGR/link refcnt reaching 0 early and entering the clear process,
making resource access unsafe.
refcount_t: addition on 0; use-after-free.
WARNING: CPU: 14 PID: 107447 at lib/refcount.c:25 refcount_warn_saturate+0x9c/0x140
Workqueue: events smc_lgr_terminate_work [smc]
Call trace:
refcount_warn_saturate+0x9c/0x140
__smc_lgr_terminate.part.45+0x2a8/0x370 [smc]
smc_lgr_terminate_work+0x28/0x30 [smc]
process_one_work+0x1b8/0x420
worker_thread+0x158/0x510
kthread+0x114/0x118
or
refcount_t: underflow; use-after-free.
WARNING: CPU: 6 PID: 93140 at lib/refcount.c:28 refcount_warn_saturate+0xf0/0x140
Workqueue: smc_hs_wq smc_listen_work [smc]
Call trace:
refcount_warn_saturate+0xf0/0x140
smcr_link_put+0x1cc/0x1d8 [smc]
smc_conn_free+0x110/0x1b0 [smc]
smc_conn_abort+0x50/0x60 [smc]
smc_listen_find_device+0x75c/0x790 [smc]
smc_listen_work+0x368/0x8a0 [smc]
process_one_work+0x1b8/0x420
worker_thread+0x158/0x510
kthread+0x114/0x118
It is caused by repeated release of LGR/link refcnt. One suspect is that
smc_conn_free() is called repeatedly because some smc_conn_free() from
server listening path are not protected by sock lock.
e.g.
Calls under socklock | smc_listen_work
-------------------------------------------------------
lock_sock(sk) | smc_conn_abort
smc_conn_free | \- smc_conn_free
\- smcr_link_put | \- smcr_link_put (duplicated)
release_sock(sk)
So here add sock lock protection in smc_listen_work() path, making it
exclusive with other connection operations.
Fixes: 3b2dec2603 ("net/smc: restructure client and server code in af_smc")
Co-developed-by: Guangguan Wang <guangguan.wang@linux.alibaba.com>
Signed-off-by: Guangguan Wang <guangguan.wang@linux.alibaba.com>
Co-developed-by: Kai <KaiShen@linux.alibaba.com>
Signed-off-by: Kai <KaiShen@linux.alibaba.com>
Signed-off-by: Wen Gu <guwen@linux.alibaba.com>
Reviewed-by: Wenjia Zhang <wenjia@linux.ibm.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 3301ab7d5a ]
Dst objects get leaked in ip6_negative_advice() when this function is
executed for an expired IPv6 route located in the exception table. There
are several conditions that must be fulfilled for the leak to occur:
* an ICMPv6 packet indicating a change of the MTU for the path is received,
resulting in an exception dst being created
* a TCP connection that uses the exception dst for routing packets must
start timing out so that TCP begins retransmissions
* after the exception dst expires, the FIB6 garbage collector must not run
before TCP executes ip6_negative_advice() for the expired exception dst
When TCP executes ip6_negative_advice() for an exception dst that has
expired and if no other socket holds a reference to the exception dst, the
refcount of the exception dst is 2, which corresponds to the increment
made by dst_init() and the increment made by the TCP socket for which the
connection is timing out. The refcount made by the socket is never
released. The refcount of the dst is decremented in sk_dst_reset() but
that decrement is counteracted by a dst_hold() intentionally placed just
before the sk_dst_reset() in ip6_negative_advice(). After
ip6_negative_advice() has finished, there is no other object tied to the
dst. The socket lost its reference stored in sk_dst_cache and the dst is
no longer in the exception table. The exception dst becomes a leaked
object.
As a result of this dst leak, an unbalanced refcount is reported for the
loopback device of a net namespace being destroyed under kernels that do
not contain e5f80fcf86 ("ipv6: give an IPv6 dev to blackhole_netdev"):
unregister_netdevice: waiting for lo to become free. Usage count = 2
Fix the dst leak by removing the dst_hold() in ip6_negative_advice(). The
patch that introduced the dst_hold() in ip6_negative_advice() was
92f1655aa2 ("net: fix __dst_negative_advice() race"). But 92f1655aa2
merely refactored the code with regards to the dst refcount so the issue
was present even before 92f1655aa2. The bug was introduced in
54c1a859ef ("ipv6: Don't drop cache route entry unless timer actually
expired.") where the expired cached route is deleted and the sk_dst_cache
member of the socket is set to NULL by calling dst_negative_advice() but
the refcount belonging to the socket is left unbalanced.
The IPv4 version - ipv4_negative_advice() - is not affected by this bug.
When the TCP connection times out ipv4_negative_advice() merely resets the
sk_dst_cache of the socket while decrementing the refcount of the
exception dst.
Fixes: 92f1655aa2 ("net: fix __dst_negative_advice() race")
Fixes: 54c1a859ef ("ipv6: Don't drop cache route entry unless timer actually expired.")
Link: https://lore.kernel.org/netdev/20241113105611.GA6723@incl/T/#u
Signed-off-by: Jiri Wiesner <jwiesner@suse.de>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Link: https://patch.msgid.link/20241128085950.GA4505@incl
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 488b6d91b0 ]
When SOF_TIMESTAMPING_OPT_ID is used to ambiguate timestamped datagrams,
the sk_tskey can become unpredictable in case of any error happened
during sendmsg(). Move increment later in the code and make decrement of
sk_tskey in error path. This solution is still racy in case of multiple
threads doing snedmsg() over the very same socket in parallel, but still
makes error path much more predictable.
Fixes: 09c2d251b7 ("net-timestamp: add key to disambiguate concurrent datagrams")
Reported-by: Andy Lutomirski <luto@amacapital.net>
Signed-off-by: Vadim Fedorenko <vadfed@meta.com>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Link: https://lore.kernel.org/r/20240213110428.1681540-1-vadfed@meta.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Stable-dep-of: 3301ab7d5a ("net/ipv6: release expired exception dst cached in socket")
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 1596a135e3 ]
When the length of a GSO packet in the tbf qdisc is larger than the burst
size configured the packet will be segmented by the tbf_segment function.
Whenever this function is used to enqueue SKBs, the backlog statistic of
the tbf is not increased correctly. This can lead to underflows of the
'backlog' byte-statistic value when these packets are dequeued from tbf.
Reproduce the bug:
Ensure that the sender machine has GSO enabled. Configured the tbf on
the outgoing interface of the machine as follows (burstsize = 1 MTU):
$ tc qdisc add dev <oif> root handle 1: tbf rate 50Mbit burst 1514 latency 50ms
Send bulk TCP traffic out via this interface, e.g., by running an iPerf3
client on this machine. Check the qdisc statistics:
$ tc -s qdisc show dev <oif>
The 'backlog' byte-statistic has incorrect values while traffic is
transferred, e.g., high values due to u32 underflows. When the transfer
is stopped, the value is != 0, which should never happen.
This patch fixes this bug by updating the statistics correctly, even if
single SKBs of a GSO SKB cannot be enqueued.
Fixes: e43ac79a4b ("sch_tbf: segment too big GSO packets")
Signed-off-by: Martin Ottens <martin.ottens@fau.de>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Link: https://patch.msgid.link/20241125174608.1484356-1-martin.ottens@fau.de
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit b7529880cb ]
cgroup maximum depth is INT_MAX by default, there is a cgroup toggle to
restrict this maximum depth to a more reasonable value not to harm
performance. Remove unnecessary WARN_ON_ONCE which is reachable from
userspace.
Fixes: 7f3287db65 ("netfilter: nft_socket: make cgroupsv2 matching work with namespaces")
Reported-by: syzbot+57bac0866ddd99fe47c0@syzkaller.appspotmail.com
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 146b6f1112 ]
Under certain kernel configurations when building with Clang/LLVM, the
compiler does not generate a return or jump as the terminator
instruction for ip_vs_protocol_init(), triggering the following objtool
warning during build time:
vmlinux.o: warning: objtool: ip_vs_protocol_init() falls through to next function __initstub__kmod_ip_vs_rr__935_123_ip_vs_rr_init6()
At runtime, this either causes an oops when trying to load the ipvs
module or a boot-time panic if ipvs is built-in. This same issue has
been reported by the Intel kernel test robot previously.
Digging deeper into both LLVM and the kernel code reveals this to be a
undefined behavior problem. ip_vs_protocol_init() uses a on-stack buffer
of 64 chars to store the registered protocol names and leaves it
uninitialized after definition. The function calls strnlen() when
concatenating protocol names into the buffer. With CONFIG_FORTIFY_SOURCE
strnlen() performs an extra step to check whether the last byte of the
input char buffer is a null character (commit 3009f891bb ("fortify:
Allow strlen() and strnlen() to pass compile-time known lengths")).
This, together with possibly other configurations, cause the following
IR to be generated:
define hidden i32 @ip_vs_protocol_init() local_unnamed_addr #5 section ".init.text" align 16 !kcfi_type !29 {
%1 = alloca [64 x i8], align 16
...
14: ; preds = %11
%15 = getelementptr inbounds i8, ptr %1, i64 63
%16 = load i8, ptr %15, align 1
%17 = tail call i1 @llvm.is.constant.i8(i8 %16)
%18 = icmp eq i8 %16, 0
%19 = select i1 %17, i1 %18, i1 false
br i1 %19, label %20, label %23
20: ; preds = %14
%21 = call i64 @strlen(ptr noundef nonnull dereferenceable(1) %1) #23
...
23: ; preds = %14, %11, %20
%24 = call i64 @strnlen(ptr noundef nonnull dereferenceable(1) %1, i64 noundef 64) #24
...
}
The above code calculates the address of the last char in the buffer
(value %15) and then loads from it (value %16). Because the buffer is
never initialized, the LLVM GVN pass marks value %16 as undefined:
%13 = getelementptr inbounds i8, ptr %1, i64 63
br i1 undef, label %14, label %17
This gives later passes (SCCP, in particular) more DCE opportunities by
propagating the undef value further, and eventually removes everything
after the load on the uninitialized stack location:
define hidden i32 @ip_vs_protocol_init() local_unnamed_addr #0 section ".init.text" align 16 !kcfi_type !11 {
%1 = alloca [64 x i8], align 16
...
12: ; preds = %11
%13 = getelementptr inbounds i8, ptr %1, i64 63
unreachable
}
In this way, the generated native code will just fall through to the
next function, as LLVM does not generate any code for the unreachable IR
instruction and leaves the function without a terminator.
Zero the on-stack buffer to avoid this possible UB.
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Reported-by: kernel test robot <lkp@intel.com>
Closes: https://lore.kernel.org/oe-kbuild-all/202402100205.PWXIz1ZK-lkp@intel.com/
Co-developed-by: Ruowen Qin <ruqin@redhat.com>
Signed-off-by: Ruowen Qin <ruqin@redhat.com>
Signed-off-by: Jinghao Jia <jinghao7@illinois.edu>
Acked-by: Julian Anastasov <ja@ssi.bg>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 4db9ad82a6 ]
Since transport->sock has been set to NULL during reset transport,
XPRT_SOCK_UPD_TIMEOUT also needs to be cleared. Otherwise, the
xs_tcp_set_socket_timeouts() may be triggered in xs_tcp_send_request()
to dereference the transport->sock that has been set to NULL.
Fixes: 7196dbb02e ("SUNRPC: Allow changing of the TCP timeout parameters on the fly")
Signed-off-by: Li Lingfeng <lilingfeng3@huawei.com>
Signed-off-by: Liu Jian <liujian56@huawei.com>
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 7ef3ae82a6 ]
Large amount of mount hangs observed during hotplugging of 9pfs devices. The
9pfs Xen driver attempts to initialize itself more than once, causing the
frontend and backend to disagree: the backend listens on a channel that the
frontend does not send on, resulting in stalled processing.
Only allow initialization of 9p frontend once.
Fixes: c15fe55d14 ("9p/xen: fix connection sequence")
Signed-off-by: Alex Zenla <alex@edera.dev>
Signed-off-by: Alexander Merritt <alexander@edera.dev>
Signed-off-by: Ariadne Conill <ariadne@ariadne.space>
Reviewed-by: Juergen Gross <jgross@suse.com>
Message-ID: <20241119211633.38321-1-alexander@edera.dev>
Signed-off-by: Dominique Martinet <asmadeus@codewreck.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
commit 2862eee078 upstream.
The function `c_show` was called with protection from RCU. This only
ensures that `cp` will not be freed. Therefore, the reference count for
`cp` can drop to zero, which will trigger a refcount use-after-free
warning when `cache_get` is called. To resolve this issue, use
`cache_get_rcu` to ensure that `cp` remains active.
------------[ cut here ]------------
refcount_t: addition on 0; use-after-free.
WARNING: CPU: 7 PID: 822 at lib/refcount.c:25
refcount_warn_saturate+0xb1/0x120
CPU: 7 UID: 0 PID: 822 Comm: cat Not tainted 6.12.0-rc3+ #1
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS
1.16.1-2.fc37 04/01/2014
RIP: 0010:refcount_warn_saturate+0xb1/0x120
Call Trace:
<TASK>
c_show+0x2fc/0x380 [sunrpc]
seq_read_iter+0x589/0x770
seq_read+0x1e5/0x270
proc_reg_read+0xe1/0x140
vfs_read+0x125/0x530
ksys_read+0xc1/0x160
do_syscall_64+0x5f/0x170
entry_SYSCALL_64_after_hwframe+0x76/0x7e
Cc: stable@vger.kernel.org # v4.20+
Signed-off-by: Yang Erkun <yangerkun@huawei.com>
Reviewed-by: Jeff Layton <jlayton@kernel.org>
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 35f56c554e upstream.
When tb[IPSET_ATTR_IP_TO] is not present but tb[IPSET_ATTR_CIDR] exists,
the values of ip and ip_to are slightly swapped. Therefore, the range check
for ip should be done later, but this part is missing and it seems that the
vulnerability occurs.
So we should add missing range checks and remove unnecessary range checks.
Cc: <stable@vger.kernel.org>
Reported-by: syzbot+58c872f7790a4d2ac951@syzkaller.appspotmail.com
Fixes: 72205fc68b ("netfilter: ipset: bitmap:ip set type support")
Signed-off-by: Jeongjun Park <aha310510@gmail.com>
Acked-by: Jozsef Kadlecsik <kadlec@blackhole.kfki.hu>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 5fe6caa62b upstream.
Commit 9bf4e919cc worked around an issue introduced after an innocuous
optimisation change in LLVM main:
> len is defined as an 'int' because it is assigned from
> '__user int *optlen'. However, it is clamped against the result of
> sizeof(), which has a type of 'size_t' ('unsigned long' for 64-bit
> platforms). This is done with min_t() because min() requires compatible
> types, which results in both len and the result of sizeof() being casted
> to 'unsigned int', meaning len changes signs and the result of sizeof()
> is truncated. From there, len is passed to copy_to_user(), which has a
> third parameter type of 'unsigned long', so it is widened and changes
> signs again. This excessive casting in combination with the KCSAN
> instrumentation causes LLVM to fail to eliminate the __bad_copy_from()
> call, failing the build.
The same issue occurs in rfcomm in functions rfcomm_sock_getsockopt and
rfcomm_sock_getsockopt_old.
Change the type of len to size_t in both rfcomm_sock_getsockopt and
rfcomm_sock_getsockopt_old and replace min_t() with min().
Cc: stable@vger.kernel.org
Co-authored-by: Aleksei Vetrov <vvvvvv@google.com>
Improves: 9bf4e919cc ("Bluetooth: Fix type of len in {l2cap,sco}_sock_getsockopt_old()")
Link: https://github.com/ClangBuiltLinux/linux/issues/2007
Link: https://github.com/llvm/llvm-project/issues/85647
Signed-off-by: Andrej Shadura <andrew.shadura@collabora.co.uk>
Reviewed-by: Nathan Chancellor <nathan@kernel.org>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit fc9c273d6d ]
Similar to the previous patch, plumb the RCU lock inside
the ipmr_get_table(), provided a lockless variant and apply
the latter in the few spots were the lock is already held.
Fixes: 709b46e8d9 ("net: Add compat ioctl support for the ipv4 multicast ioctl SIOCGETSGCNT")
Fixes: f0ad0860d0 ("ipv4: ipmr: support multiple tables")
Reviewed-by: David Ahern <dsahern@kernel.org>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit f1553c9894 ]
Several places call ip6mr_get_table() with no RCU nor RTNL lock.
Add RCU protection inside such helper and provide a lockless variant
for the few callers that already acquired the relevant lock.
Note that some users additionally reference the table outside the RCU
lock. That is actually safe as the table deletion can happen only
after all table accesses are completed.
Fixes: e2d57766e6 ("net: Provide compat support for SIOCGETMIFCNT_IN6 and SIOCGETSGCNT_IN6.")
Fixes: d7c31cbde4 ("net: ip6mr: add RTM_GETROUTE netlink op")
Reviewed-by: David Ahern <dsahern@kernel.org>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit c31e72d021 ]
The cited commit replaced inet_csk_reqsk_queue_drop_and_put() with
__inet_csk_reqsk_queue_drop() and reqsk_put() in reqsk_timer_handler().
Then, oreq should be passed to reqsk_put() instead of req; otherwise
use-after-free of nreq could happen when reqsk is migrated but the
retry attempt failed (e.g. due to timeout).
Let's pass oreq to reqsk_put().
Fixes: e8c526f2bd ("tcp/dccp: Don't use timer_pending() in reqsk_queue_unlink().")
Reported-by: Liu Jian <liujian56@huawei.com>
Closes: https://lore.kernel.org/netdev/1284490f-9525-42ee-b7b8-ccadf6606f6d@huawei.com/
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Reviewed-by: Vadim Fedorenko <vadim.fedorenko@linux.dev>
Reviewed-by: Liu Jian <liujian56@huawei.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Martin KaFai Lau <martin.lau@kernel.org>
Link: https://patch.msgid.link/20241123174236.62438-1-kuniyu@amazon.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 0202005664 ]
copy_from_sockptr() does not return negative value on error; instead, it
reports the number of bytes that failed to copy. Since it's deprecated,
switch to copy_safe_from_sockptr().
Note: Keeping the `optlen != sizeof(unsigned int)` check as
copy_safe_from_sockptr() by itself would also accept
optlen > sizeof(unsigned int). Which would allow a more lenient handling
of inputs.
Fixes: 17926a7932 ("[AF_RXRPC]: Provide secure RxRPC sockets for use by userspace and kernel both")
Signed-off-by: Michal Luczaj <mhal@rbox.co>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 1465036b10 ]
copy_from_sockptr() is used incorrectly: return value is the number of
bytes that could not be copied. Since it's deprecated, switch to
copy_safe_from_sockptr().
Note: Keeping the `optlen != sizeof(int)` check as copy_safe_from_sockptr()
by itself would also accept optlen > sizeof(int). Which would allow a more
lenient handling of inputs.
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Suggested-by: David Wei <dw@davidwei.uk>
Signed-off-by: Michal Luczaj <mhal@rbox.co>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 9cfb5e7f0d ]
Following sequence in hsr_init_sk() is invalid :
skb_reset_mac_header(skb);
skb_reset_mac_len(skb);
skb_reset_network_header(skb);
skb_reset_transport_header(skb);
It is invalid because skb_reset_mac_len() needs the correct
network header, which should be after the mac header.
This patch moves the skb_reset_network_header()
and skb_reset_transport_header() before
the call to dev_hard_header().
As a result skb->mac_len is no longer set to a value
close to 65535.
Fixes: 48b491a5cc ("net: hsr: fix mac_len checks")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: George McCollister <george.mccollister@gmail.com>
Link: https://patch.msgid.link/20241122171343.897551-1-edumazet@google.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 00b5b7aab9 ]
RFC8981 section 3.4 says that existing temporary addresses must have their
lifetimes adjusted so that no temporary addresses should ever remain "valid"
or "preferred" longer than the incoming SLAAC Prefix Information. This would
strongly imply in Linux's case that if the "mngtmpaddr" address is deleted or
un-flagged as such, its corresponding temporary addresses must be cleared out
right away.
But now the temporary address is renewed even after ‘mngtmpaddr’ is removed
or becomes unmanaged as manage_tempaddrs() set temporary addresses
prefered/valid time to 0, and later in addrconf_verify_rtnl() all checkings
failed to remove the addresses. Fix this by deleting the temporary address
directly for these situations.
Fixes: 778964f2fd ("ipv6/addrconf: fix timing bug in tempaddr regen")
Signed-off-by: Hangbin Liu <liuhangbin@gmail.com>
Reviewed-by: David Ahern <dsahern@kernel.org>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit ebaf81317e ]
Passing MSG_PEEK flag to skb_recv_datagram() increments skb refcount
(skb->users) and iucv_sock_recvmsg() does not decrement skb refcount
at exit.
This results in skb memory leak in skb_queue_purge() and WARN_ON in
iucv_sock_destruct() during socket close. To fix this decrease
skb refcount by one if MSG_PEEK is set in order to prevent memory
leak and WARN_ON.
WARNING: CPU: 2 PID: 6292 at net/iucv/af_iucv.c:286 iucv_sock_destruct+0x144/0x1a0 [af_iucv]
CPU: 2 PID: 6292 Comm: afiucv_test_msg Kdump: loaded Tainted: G W 6.10.0-rc7 #1
Hardware name: IBM 3931 A01 704 (z/VM 7.3.0)
Call Trace:
[<001587c682c4aa98>] iucv_sock_destruct+0x148/0x1a0 [af_iucv]
[<001587c682c4a9d0>] iucv_sock_destruct+0x80/0x1a0 [af_iucv]
[<001587c704117a32>] __sk_destruct+0x52/0x550
[<001587c704104a54>] __sock_release+0xa4/0x230
[<001587c704104c0c>] sock_close+0x2c/0x40
[<001587c702c5f5a8>] __fput+0x2e8/0x970
[<001587c7024148c4>] task_work_run+0x1c4/0x2c0
[<001587c7023b0716>] do_exit+0x996/0x1050
[<001587c7023b13aa>] do_group_exit+0x13a/0x360
[<001587c7023b1626>] __s390x_sys_exit_group+0x56/0x60
[<001587c7022bccca>] do_syscall+0x27a/0x380
[<001587c7049a6a0c>] __do_syscall+0x9c/0x160
[<001587c7049ce8a8>] system_call+0x70/0x98
Last Breaking-Event-Address:
[<001587c682c4a9d4>] iucv_sock_destruct+0x84/0x1a0 [af_iucv]
Fixes: eac3731bd0 ("[S390]: Add AF_IUCV socket support")
Reviewed-by: Alexandra Winter <wintera@linux.ibm.com>
Reviewed-by: Thorsten Winkler <twinkler@linux.ibm.com>
Signed-off-by: Sidraya Jayagond <sidraya@linux.ibm.com>
Signed-off-by: Alexandra Winter <wintera@linux.ibm.com>
Reviewed-by: David Wei <dw@davidwei.uk>
Link: https://patch.msgid.link/20241119152219.3712168-1-wintera@linux.ibm.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit ce89e742a4 ]
There's issue as follows:
RPC: Registered rdma transport module.
RPC: Registered rdma backchannel transport module.
RPC: Unregistered rdma transport module.
RPC: Unregistered rdma backchannel transport module.
BUG: unable to handle page fault for address: fffffbfff80c609a
PGD 123fee067 P4D 123fee067 PUD 123fea067 PMD 10c624067 PTE 0
Oops: Oops: 0000 [#1] PREEMPT SMP KASAN NOPTI
RIP: 0010:percpu_counter_destroy_many+0xf7/0x2a0
Call Trace:
<TASK>
__die+0x1f/0x70
page_fault_oops+0x2cd/0x860
spurious_kernel_fault+0x36/0x450
do_kern_addr_fault+0xca/0x100
exc_page_fault+0x128/0x150
asm_exc_page_fault+0x26/0x30
percpu_counter_destroy_many+0xf7/0x2a0
mmdrop+0x209/0x350
finish_task_switch.isra.0+0x481/0x840
schedule_tail+0xe/0xd0
ret_from_fork+0x23/0x80
ret_from_fork_asm+0x1a/0x30
</TASK>
If register_sysctl() return NULL, then svc_rdma_proc_cleanup() will not
destroy the percpu counters which init in svc_rdma_proc_init().
If CONFIG_HOTPLUG_CPU is enabled, residual nodes may be in the
'percpu_counters' list. The above issue may occur once the module is
removed. If the CONFIG_HOTPLUG_CPU configuration is not enabled, memory
leakage occurs.
To solve above issue just destroy all percpu counters when
register_sysctl() return NULL.
Fixes: 1e7e557316 ("svcrdma: Restore read and write stats")
Fixes: 22df5a2246 ("svcrdma: Convert rdma_stat_sq_starve to a per-CPU counter")
Fixes: df971cd853 ("svcrdma: Convert rdma_stat_recv to a per-CPU counter")
Signed-off-by: Ye Bin <yebin10@huawei.com>
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 376bcd9b37 ]
There is no need to declare two tables to just create directories,
this can be easily be done with a prefix path with register_sysctl().
Simplify this registration.
Signed-off-by: Luis Chamberlain <mcgrof@kernel.org>
Reviewed-by: Jeff Layton <jlayton@kernel.org>
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Stable-dep-of: ce89e742a4 ("svcrdma: fix miss destroy percpu_counter in svc_rdma_proc_init()")
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 8ca2a1eead ]
When the stream_verdict program returns SK_PASS, it places the received skb
into its own receive queue, but a recursive lock eventually occurs, leading
to an operating system deadlock. This issue has been present since v6.9.
'''
sk_psock_strp_data_ready
write_lock_bh(&sk->sk_callback_lock)
strp_data_ready
strp_read_sock
read_sock -> tcp_read_sock
strp_recv
cb.rcv_msg -> sk_psock_strp_read
# now stream_verdict return SK_PASS without peer sock assign
__SK_PASS = sk_psock_map_verd(SK_PASS, NULL)
sk_psock_verdict_apply
sk_psock_skb_ingress_self
sk_psock_skb_ingress_enqueue
sk_psock_data_ready
read_lock_bh(&sk->sk_callback_lock) <= dead lock
'''
This topic has been discussed before, but it has not been fixed.
Previous discussion:
https://lore.kernel.org/all/6684a5864ec86_403d20898@john.notmuch
Fixes: 6648e61322 ("bpf, skmsg: Fix NULL pointer dereference in sk_psock_skb_ingress_enqueue")
Reported-by: Vincent Whitchurch <vincent.whitchurch@datadoghq.com>
Signed-off-by: Jiayuan Chen <mrpre@163.com>
Signed-off-by: John Fastabend <john.fastabend@gmail.com>
Acked-by: Martin KaFai Lau <martin.lau@kernel.org>
Link: https://patch.msgid.link/20241118030910.36230-2-mrpre@163.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit eb02688c5c ]
The CI is hitting some aperiodic hangup at device removal time in the
pmtu.sh self-test:
unregister_netdevice: waiting for veth_A-R1 to become free. Usage count = 6
ref_tracker: veth_A-R1@ffff888013df15d8 has 1/5 users at
dst_init+0x84/0x4a0
dst_alloc+0x97/0x150
ip6_dst_alloc+0x23/0x90
ip6_rt_pcpu_alloc+0x1e6/0x520
ip6_pol_route+0x56f/0x840
fib6_rule_lookup+0x334/0x630
ip6_route_output_flags+0x259/0x480
ip6_dst_lookup_tail.constprop.0+0x5c2/0x940
ip6_dst_lookup_flow+0x88/0x190
udp_tunnel6_dst_lookup+0x2a7/0x4c0
vxlan_xmit_one+0xbde/0x4a50 [vxlan]
vxlan_xmit+0x9ad/0xf20 [vxlan]
dev_hard_start_xmit+0x10e/0x360
__dev_queue_xmit+0xf95/0x18c0
arp_solicit+0x4a2/0xe00
neigh_probe+0xaa/0xf0
While the first suspect is the dst_cache, explicitly tracking the dst
owing the last device reference via probes proved such dst is held by
the nexthop in the originating fib6_info.
Similar to commit f5b51fe804 ("ipv6: route: purge exception on
removal"), we need to explicitly release the originating fib info when
disconnecting a to-be-removed device from a live ipv6 dst: move the
fib6_info cleanup into ip6_dst_ifdown().
Tested running:
./pmtu.sh cleanup_ipv6_exception
in a tight loop for more than 400 iterations with no spat, running an
unpatched kernel I observed a splat every ~10 iterations.
Fixes: f88d8ea67f ("ipv6: Plumb support for nexthop object in a fib6_info")
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: David Ahern <dsahern@kernel.org>
Link: https://patch.msgid.link/604c45c188c609b732286b47ac2a451a40f6cf6d.1730828007.git.pabeni@redhat.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit b4cb4a1391 ]
Toke mentioned unrcu_pointer() existence, allowing
to remove some of the ugly casts we have when using
xchg() for rcu protected pointers.
Also make inet_rcv_compat const.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Toke Høiland-Jørgensen <toke@redhat.com>
Reviewed-by: Toke Høiland-Jørgensen <toke@redhat.com>
Link: https://lore.kernel.org/r/20240604111603.45871-1-edumazet@google.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Stable-dep-of: eb02688c5c ("ipv6: release nexthop on device removal")
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 86e8921df0 ]
TCPDIAG_GETSOCK and DCCPDIAG_GETSOCK diag are serialized
on sock_diag_table_mutex.
This is to make sure inet_diag module is not unloaded
while diag was ongoing.
It is time to get rid of this mutex and use RCU protection,
allowing full parallelism.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Guillaume Nault <gnault@redhat.com>
Reviewed-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Stable-dep-of: eb02688c5c ("ipv6: release nexthop on device removal")
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 1d55a69747 ]
sock_diag_broadcast_destroy_work() and __sock_diag_cmd()
are currently using sock_diag_table_mutex to protect
against concurrent sock_diag_handlers[] changes.
This makes inet_diag dump serialized, thus less scalable
than legacy /proc files.
It is time to switch to full RCU protection.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Guillaume Nault <gnault@redhat.com>
Reviewed-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Stable-dep-of: eb02688c5c ("ipv6: release nexthop on device removal")
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 114b4bb1cc ]
Following patch is going to use RCU instead of
sock_diag_table_mutex acquisition.
This patch is a preparation, no change of behavior yet.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Guillaume Nault <gnault@redhat.com>
Reviewed-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Stable-dep-of: eb02688c5c ("ipv6: release nexthop on device removal")
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 955afd57dc ]
Found in the test_txmsg_pull in test_sockmap,
```
txmsg_cork = 512; // corking is importrant here
opt->iov_length = 3;
opt->iov_count = 1;
opt->rate = 512; // sendmsg will be invoked 512 times
```
The first sendmsg will send an sk_msg with size 3, and bpf_msg_pull_data
will be invoked the first time. sk_msg_reset_curr will reset the copybreak
from 3 to 0. In the second sendmsg, since we are in the stage of corking,
psock->cork will be reused in func sk_msg_alloc. msg->sg.copybreak is 0
now, the second msg will overwrite the first msg. As a result, we could
not pass the data integrity test.
The same problem happens in push and pop test. Thus, fix sk_msg_reset_curr
to restore the correct copybreak.
Fixes: bb9aefde5b ("bpf: sockmap, updating the sg structure should also update curr")
Signed-off-by: Zijian Zhang <zijianzhang@bytedance.com>
Link: https://lore.kernel.org/r/20241106222520.527076-9-zijianzhang@bytedance.com
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 5d609ba262 ]
Several fixes to bpf_msg_pop_data,
1. In sk_msg_shift_left, we should put_page
2. if (len == 0), return early is better
3. pop the entire sk_msg (last == msg->sg.size) should be supported
4. Fix for the value of variable "a"
5. In sk_msg_shift_left, after shifting, i has already pointed to the next
element. Addtional sk_msg_iter_var_next may result in BUG.
Fixes: 7246d8ed4d ("bpf: helper to pop data from messages")
Signed-off-by: Zijian Zhang <zijianzhang@bytedance.com>
Reviewed-by: John Fastabend <john.fastabend@gmail.com>
Link: https://lore.kernel.org/r/20241106222520.527076-8-zijianzhang@bytedance.com
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 15ab0548e3 ]
Several fixes to bpf_msg_push_data,
1. test_sockmap has tests where bpf_msg_push_data is invoked to push some
data at the end of a message, but -EINVAL is returned. In this case, in
bpf_msg_push_data, after the first loop, i will be set to msg->sg.end, add
the logic to handle it.
2. In the code block of "if (start - offset)", it's possible that "i"
points to the last of sk_msg_elem. In this case, "sk_msg_iter_next(msg,
end)" might still be called twice, another invoking is in "if (!copy)"
code block, but actually only one is needed. Add the logic to handle it,
and reconstruct the code to make the logic more clear.
Fixes: 6fff607e2f ("bpf: sk_msg program helper bpf_msg_push_data")
Signed-off-by: Zijian Zhang <zijianzhang@bytedance.com>
Link: https://lore.kernel.org/r/20241106222520.527076-7-zijianzhang@bytedance.com
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit cddc04275f ]
Update of stateful object triggers:
WARNING: suspicious RCU usage
net/netfilter/nf_tables_api.c:7759 RCU-list traversed in non-reader section!!
other info that might help us debug this:
rcu_scheduler_active = 2, debug_locks = 1
1 lock held by nft/3060:
#0: ffff88810f0578c8 (&nft_net->commit_mutex){+.+.}-{4:4}, [..]
... but this list is not protected by the transaction mutex but the
nfnl nftables subsystem mutex.
Switch to nft_obj_type_get which will acquire rcu read lock,
bump refcount, and returns the result.
v3: Dan Carpenter points out nft_obj_type_get returns error pointer, not
NULL, on error.
Fixes: dad3bdeef4 ("netfilter: nf_tables: fix memory leak during stateful obj update").
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 84b1a0c014 ]
Turn update into noop as a follow up for:
9fedd894b4 ("netfilter: nf_tables: fix unexpected EOPNOTSUPP error")
instead of adding a transaction object which is simply discarded at a
later stage of the commit protocol.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Stable-dep-of: cddc04275f ("netfilter: nf_tables: must hold rcu read lock while iterating object type list")
Signed-off-by: Sasha Levin <sashal@kernel.org>
commit b169e76eba upstream.
In 'mptcp_reset_tout_timer', promote 'probe_timestamp' to unsigned long
to avoid possible integer overflow. Compile tested only.
Found by Linux Verification Center (linuxtesting.org) with SVACE.
Signed-off-by: Dmitry Kandybka <d.kandybka@gmail.com>
Link: https://patch.msgid.link/20241107103657.1560536-1-d.kandybka@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
[ Conflict in this version because commit d866ae9aaa ("mptcp: add a
new sysctl for make after break timeout") is not in this version, and
replaced TCP_TIMEWAIT_LEN in the expression. The fix can still be
applied the same way: by forcing a cast to unsigned long for the first
item. ]
Signed-off-by: Matthieu Baerts (NGI0) <matttbe@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit f53e1c9c72 ]
If mgmt_index_removed is called while there are commands queued on
cmd_sync it could lead to crashes like the bellow trace:
0x0000053D: __list_del_entry_valid_or_report+0x98/0xdc
0x0000053D: mgmt_pending_remove+0x18/0x58 [bluetooth]
0x0000053E: mgmt_remove_adv_monitor_complete+0x80/0x108 [bluetooth]
0x0000053E: hci_cmd_sync_work+0xbc/0x164 [bluetooth]
So while handling mgmt_index_removed this attempts to dequeue
commands passed as user_data to cmd_sync.
Fixes: 7cf5c2978f ("Bluetooth: hci_sync: Refactor remove Adv Monitor")
Reported-by: jiaymao <quic_jiaymao@quicinc.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
[Xiangyu: BP to fix CVE: CVE-2024-49951, Minor conflict resolution]
Signed-off-by: Xiangyu Chen <xiangyu.chen@windriver.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 505ea2b295 ]
This adds functions to queue, dequeue and lookup into the cmd_sync
list.
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
Signed-off-by: Xiangyu Chen <xiangyu.chen@windriver.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 9ab5cf19fb ]
Config a small gso_max_size/gso_ipv4_max_size will lead to an underflow
in sk_dst_gso_max_size(), which may trigger a BUG_ON crash,
because sk->sk_gso_max_size would be much bigger than device limits.
Call Trace:
tcp_write_xmit
tso_segs = tcp_init_tso_segs(skb, mss_now);
tcp_set_skb_tso_segs
tcp_skb_pcount_set
// skb->len = 524288, mss_now = 8
// u16 tso_segs = 524288/8 = 65535 -> 0
tso_segs = DIV_ROUND_UP(skb->len, mss_now)
BUG_ON(!tso_segs)
Add check for the minimum value of gso_max_size and gso_ipv4_max_size.
Fixes: 46e6b992c2 ("rtnetlink: allow GSO maximums to be set on device creation")
Fixes: 9eefedd58a ("net: add gso_ipv4_max_size and gro_ipv4_max_size per device")
Signed-off-by: Wang Liang <wangliang74@huawei.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Link: https://patch.msgid.link/20241023035213.517386-1-wangliang74@huawei.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
[ Resolve minor conflicts to fix CVE-2024-50258 ]
Signed-off-by: Bin Lan <bin.lan.cn@windriver.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit e28acc9c1c ]
Accessing `mr_table->mfc_cache_list` is protected by an RCU lock. In the
following code flow, the RCU read lock is not held, causing the
following error when `RCU_PROVE` is not held. The same problem might
show up in the IPv6 code path.
6.12.0-rc5-kbuilder-01145-gbac17284bdcb #33 Tainted: G E N
-----------------------------
net/ipv4/ipmr_base.c:313 RCU-list traversed in non-reader section!!
rcu_scheduler_active = 2, debug_locks = 1
2 locks held by RetransmitAggre/3519:
#0: ffff88816188c6c0 (nlk_cb_mutex-ROUTE){+.+.}-{3:3}, at: __netlink_dump_start+0x8a/0x290
#1: ffffffff83fcf7a8 (rtnl_mutex){+.+.}-{3:3}, at: rtnl_dumpit+0x6b/0x90
stack backtrace:
lockdep_rcu_suspicious
mr_table_dump
ipmr_rtm_dumproute
rtnl_dump_all
rtnl_dumpit
netlink_dump
__netlink_dump_start
rtnetlink_rcv_msg
netlink_rcv_skb
netlink_unicast
netlink_sendmsg
This is not a problem per see, since the RTNL lock is held here, so, it
is safe to iterate in the list without the RCU read lock, as suggested
by Eric.
To alleviate the concern, modify the code to use
list_for_each_entry_rcu() with the RTNL-held argument.
The annotation will raise an error only if RTNL or RCU read lock are
missing during iteration, signaling a legitimate problem, otherwise it
will avoid this false positive.
This will solve the IPv6 case as well, since ip6mr_rtm_dumproute() calls
this function as well.
Signed-off-by: Breno Leitao <leitao@debian.org>
Reviewed-by: David Ahern <dsahern@kernel.org>
Link: https://patch.msgid.link/20241108-ipmr_rcu-v2-1-c718998e209b@debian.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
