| Age | Commit message (Collapse) | Author |
|---|
|
The dev_close() call was added in commit c9eb3e0f8701 ("net: dsa: Add
support for learning FDB through notification") "to indicate inconsistent
situation" when we could not delete an FDB entry from the port.
bridge fdb del d8:58:d7:00:ca:6d dev swp0 self master
It is a bit drastic and at the same time not helpful if the above fails
to only print with netdev_dbg log level, but on the other hand to bring
the interface down.
So increase the verbosity of the error message, and drop dev_close().
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
Currently the bridge emits atomic switchdev notifications for
dynamically learnt FDB entries. Monitoring these notifications works
wonders for switchdev drivers that want to keep their hardware FDB in
sync with the bridge's FDB.
For example station A wants to talk to station B in the diagram below,
and we are concerned with the behavior of the bridge on the DUT device:
DUT
+-------------------------------------+
| br0 |
| +------+ +------+ +------+ +------+ |
| | | | | | | | | |
| | swp0 | | swp1 | | swp2 | | eth0 | |
+-------------------------------------+
| | |
Station A | |
| |
+--+------+--+ +--+------+--+
| | | | | | | |
| | swp0 | | | | swp0 | |
Another | +------+ | | +------+ | Another
switch | br0 | | br0 | switch
| +------+ | | +------+ |
| | | | | | | |
| | swp1 | | | | swp1 | |
+--+------+--+ +--+------+--+
|
Station B
Interfaces swp0, swp1, swp2 are handled by a switchdev driver that has
the following property: frames injected from its control interface bypass
the internal address analyzer logic, and therefore, this hardware does
not learn from the source address of packets transmitted by the network
stack through it. So, since bridging between eth0 (where Station B is
attached) and swp0 (where Station A is attached) is done in software,
the switchdev hardware will never learn the source address of Station B.
So the traffic towards that destination will be treated as unknown, i.e.
flooded.
This is where the bridge notifications come in handy. When br0 on the
DUT sees frames with Station B's MAC address on eth0, the switchdev
driver gets these notifications and can install a rule to send frames
towards Station B's address that are incoming from swp0, swp1, swp2,
only towards the control interface. This is all switchdev driver private
business, which the notification makes possible.
All is fine until someone unplugs Station B's cable and moves it to the
other switch:
DUT
+-------------------------------------+
| br0 |
| +------+ +------+ +------+ +------+ |
| | | | | | | | | |
| | swp0 | | swp1 | | swp2 | | eth0 | |
+-------------------------------------+
| | |
Station A | |
| |
+--+------+--+ +--+------+--+
| | | | | | | |
| | swp0 | | | | swp0 | |
Another | +------+ | | +------+ | Another
switch | br0 | | br0 | switch
| +------+ | | +------+ |
| | | | | | | |
| | swp1 | | | | swp1 | |
+--+------+--+ +--+------+--+
|
Station B
Luckily for the use cases we care about, Station B is noisy enough that
the DUT hears it (on swp1 this time). swp1 receives the frames and
delivers them to the bridge, who enters the unlikely path in br_fdb_update
of updating an existing entry. It moves the entry in the software bridge
to swp1 and emits an addition notification towards that.
As far as the switchdev driver is concerned, all that it needs to ensure
is that traffic between Station A and Station B is not forever broken.
If it does nothing, then the stale rule to send frames for Station B
towards the control interface remains in place. But Station B is no
longer reachable via the control interface, but via a port that can
offload the bridge port learning attribute. It's just that the port is
prevented from learning this address, since the rule overrides FDB
updates. So the rule needs to go. The question is via what mechanism.
It sure would be possible for this switchdev driver to keep track of all
addresses which are sent to the control interface, and then also listen
for bridge notifier events on its own ports, searching for the ones that
have a MAC address which was previously sent to the control interface.
But this is cumbersome and inefficient. Instead, with one small change,
the bridge could notify of the address deletion from the old port, in a
symmetrical manner with how it did for the insertion. Then the switchdev
driver would not be required to monitor learn/forget events for its own
ports. It could just delete the rule towards the control interface upon
bridge entry migration. This would make hardware address learning be
possible again. Then it would take a few more packets until the hardware
and software FDB would be in sync again.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Acked-by: Nikolay Aleksandrov <nikolay@nvidia.com>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
Conntrack reassembly records the largest fragment size seen in IPCB.
However, when this gets forwarded/transmitted, fragmentation will only
be forced if one of the fragmented packets had the DF bit set.
In that case, a flag in IPCB will force fragmentation even if the
MTU is large enough.
This should work fine, but this breaks with ip tunnels.
Consider client that sends a UDP datagram of size X to another host.
The client fragments the datagram, so two packets, of size y and z, are
sent. DF bit is not set on any of these packets.
Middlebox netfilter reassembles those packets back to single size-X
packet, before routing decision.
packet-size-vs-mtu checks in ip_forward are irrelevant, because DF bit
isn't set. At output time, ip refragmentation is skipped as well
because x is still smaller than the mtu of the output device.
If ttransmit device is an ip tunnel, the packet size increases to
x+overhead.
Also, tunnel might be configured to force DF bit on outer header.
In this case, packet will be dropped (exceeds MTU) and an ICMP error is
generated back to sender.
But sender already respects the announced MTU, all the packets that
it sent did fit the announced mtu.
Force refragmentation as per original sizes unconditionally so ip tunnel
will encapsulate the fragments instead.
The only other solution I see is to place ip refragmentation in
the ip_tunnel code to handle this case.
Fixes: d6b915e29f4ad ("ip_fragment: don't forward defragmented DF packet")
Reported-by: Christian Perle <christian.perle@secunet.com>
Signed-off-by: Florian Westphal <fw@strlen.de>
Acked-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
For some reason ip_tunnel insist on setting the DF bit anyway when the
inner header has the DF bit set, EVEN if the tunnel was configured with
'nopmtudisc'.
This means that the script added in the previous commit
cannot be made to work by adding the 'nopmtudisc' flag to the
ip tunnel configuration. Doing so breaks connectivity even for the
without-conntrack/netfilter scenario.
When nopmtudisc is set, the tunnel will skip the mtu check, so no
icmp error is sent to client. Then, because inner header has DF set,
the outer header gets added with DF bit set as well.
IP stack then sends an error to itself because the packet exceeds
the device MTU.
Fixes: 23a3647bc4f93 ("ip_tunnels: Use skb-len to PMTU check.")
Cc: Stefano Brivio <sbrivio@redhat.com>
Signed-off-by: Florian Westphal <fw@strlen.de>
Acked-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
Move the NETIF_F_RX_UDP_TUNNEL_PORT feature check into
udp_tunnel_nic_*_port() helpers, since they're always
done right before the call.
Add similar checks before calling the notifier.
udp_tunnel_nic invokes the notifier without checking
features which could result in some wasted cycles.
Reviewed-by: Alexander Duyck <alexanderduyck@fb.com>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
All drivers use udp_tunnel_nic_*_port() helpers, prepare for
NDO removal by invoking those helpers directly.
The helpers are safe to call on all devices, they check if
device has the UDP tunnel state initialized.
Reviewed-by: Alexander Duyck <alexanderduyck@fb.com>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
Route removal is handled by two code paths. The main removal path is via
fib6_del_route() which will handle purging any PMTU exceptions from the
cache, removing all per-cpu copies of the DST entry used by the route, and
releasing the fib6_info struct.
The second removal location is during fib6_add_rt2node() during a route
replacement operation. This path also calls fib6_purge_rt() to handle
cleaning up the per-cpu copies of the DST entries and releasing the
fib6_info associated with the older route, but it does not flush any PMTU
exceptions that the older route had. Since the older route is removed from
the tree during the replacement, we lose any way of accessing it again.
As these lingering DSTs and the fib6_info struct are holding references to
the underlying netdevice struct as well, unregistering that device from the
kernel can never complete.
Fixes: 2b760fcf5cfb3 ("ipv6: hook up exception table to store dst cache")
Signed-off-by: Sean Tranchetti <stranche@codeaurora.org>
Reviewed-by: David Ahern <dsahern@kernel.org>
Link: https://lore.kernel.org/r/1609892546-11389-1-git-send-email-stranche@quicinc.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/mkl/linux-can-next
Marc Kleine-Budde says:
====================
pull-request: can-next 2021-01-06
The first 16 patches are by me and target the tcan4x5x SPI glue driver for the
m_can CAN driver. First there are a several cleanup commits, then the SPI
regmap part is converted to 8 bits per word, to make it possible to use that
driver on SPI controllers that only support the 8 bit per word mode (such as
the SPI cores on the raspberry pi).
Oliver Hartkopp contributes a patch for the CAN_RAW protocol. The getsockopt()
for CAN_RAW_FILTER is changed to return -ERANGE if the filterset does not fit
into the provided user space buffer.
The last two patches are by Joakim Zhang and add wakeup support to the flexcan
driver for the i.MX8QM SoC. The dt-bindings docs are extended to describe the
added property.
* tag 'linux-can-next-for-5.12-20210106' of git://git.kernel.org/pub/scm/linux/kernel/git/mkl/linux-can-next:
can: flexcan: add CAN wakeup function for i.MX8QM
dt-bindings: can: fsl,flexcan: add fsl,scu-index property to indicate a resource
can: raw: return -ERANGE when filterset does not fit into user space buffer
can: tcan4x5x: add support for half-duplex controllers
can: tcan4x5x: rework SPI access
can: tcan4x5x: add {wr,rd}_table
can: tcan4x5x: add max_raw_{read,write} of 256
can: tcan4x5x: tcan4x5x_regmap: set reg_stride to 4
can: tcan4x5x: fix max register value
can: tcan4x5x: tcan4x5x_regmap_init(): use spi as context pointer
can: tcan4x5x: tcan4x5x_regmap_write(): remove not needed casts and replace 4 by sizeof
can: tcan4x5x: rename regmap_spi_gather_write() -> tcan4x5x_regmap_gather_write()
can: tcan4x5x: remove regmap async support
can: tcan4x5x: tcan4x5x_bus: remove not needed read_flag_mask
can: tcan4x5x: mark struct regmap_bus tcan4x5x_bus as constant
can: tcan4x5x: move regmap code into seperate file
can: tcan4x5x: rename tcan4x5x.c -> tcan4x5x-core.c
can: tcan4x5x: beautify indention of tcan4x5x_of_match and tcan4x5x_id_table
can: tcan4x5x: replace DEVICE_NAME by KBUILD_MODNAME
====================
Link: https://lore.kernel.org/r/20210107094900.173046-1-mkl@pengutronix.de
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
Looking for an -EINVAL all over the dsa code could take hours for
inexperienced DSA users.
Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Link: https://lore.kernel.org/r/20210106090915.21439-1-zajec5@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
Multiple filters (struct can_filter) can be set with the setsockopt()
function, which was originally intended as a write-only operation.
As getsockopt() also provides a CAN_RAW_FILTER option to read back the
given filters, the caller has to provide an appropriate user space buffer.
In the case this buffer is too small the getsockopt() silently truncates
the filter information and gives no information about the needed space.
This is safe but not convenient for the programmer.
In net/core/sock.c the SO_PEERGROUPS sockopt had a similar requirement
and solved it by returning -ERANGE in the case that the provided data
does not fit into the given user space buffer and fills the required size
into optlen, so that the caller can retry with a matching buffer length.
This patch adopts this approach for CAN_RAW_FILTER getsockopt().
Reported-by: Phillip Schichtel <phillip@schich.tel>
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
Tested-By: Phillip Schichtel <phillip@schich.tel>
Link: https://lore.kernel.org/r/20201216174928.21663-1-socketcan@hartkopp.net
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
|
|
A null-ptr-deref bug is reported by Hulk Robot like this:
--------------
KASAN: null-ptr-deref in range [0x0000000000000128-0x000000000000012f]
Call Trace:
qrtr_ns_remove+0x22/0x40 [ns]
qrtr_proto_fini+0xa/0x31 [qrtr]
__x64_sys_delete_module+0x337/0x4e0
do_syscall_64+0x34/0x80
entry_SYSCALL_64_after_hwframe+0x44/0xa9
RIP: 0033:0x468ded
--------------
When qrtr_ns_init fails in qrtr_proto_init, qrtr_ns_remove which would
be called later on would raise a null-ptr-deref because qrtr_ns.workqueue
has been destroyed.
Fix it by making qrtr_ns_init have a return value and adding a check in
qrtr_proto_init.
Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Qinglang Miao <miaoqinglang@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
VLAN checks for NETREG_UNINITIALIZED to distinguish between
registration failure and unregistration in progress.
Since commit cb626bf566eb ("net-sysfs: Fix reference count leak")
registration failure may, however, result in NETREG_UNREGISTERED
as well as NETREG_UNINITIALIZED.
This fix is similer to cebb69754f37 ("rtnetlink: Fix
memory(net_device) leak when ->newlink fails")
Fixes: cb626bf566eb ("net-sysfs: Fix reference count leak")
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
If there is a NCI command in work queue after closing the NCI device at
nci_unregister_device, The NCI command timer starts at flush_workqueue
function and then NCI command timeout handler would be called 5 second
after flushing the NCI command work queue and destroying the queue.
At that time, the timeout handler would try to use NCI command work queue
that is destroyed already. it will causes the problem. To avoid this
abnormal situation, change the sequence to prevent the NCI command timeout
handler from being called after destroying the NCI command work queue.
Signed-off-by: Bongsu Jeon <bongsu.jeon@samsung.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
The function sockfd_lookup uses fget on the value that is stored in
the file field of the returned structure, so fput should ultimately be
applied to this value. This can be done directly, but it seems better
to use the specific macro sockfd_put, which does the same thing.
Perform a source code refactoring by using the following semantic patch.
// <smpl>
@@
expression s;
@@
s = sockfd_lookup(...)
...
+ sockfd_put(s);
- fput(s->file);
// </smpl>
Signed-off-by: Zheng Yongjun <zhengyongjun3@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Without crc32 support, this fails to link:
arm-linux-gnueabi-ld: net/wireless/scan.o: in function `cfg80211_scan_6ghz':
scan.c:(.text+0x928): undefined reference to `crc32_le'
Fixes: c8cb5b854b40 ("nl80211/cfg80211: support 6 GHz scanning")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Use the existing offsetof() macro instead of duplicating code.
Signed-off-by: Zheng Yongjun <zhengyongjun3@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net
Pull networking fixes from Jakub Kicinski:
"Networking fixes, including fixes from netfilter, wireless and bpf
trees.
Current release - regressions:
- mt76: fix NULL pointer dereference in mt76u_status_worker and
mt76s_process_tx_queue
- net: ipa: fix interconnect enable bug
Current release - always broken:
- netfilter: fixes possible oops in mtype_resize in ipset
- ath11k: fix number of coding issues found by static analysis tools
and spurious error messages
Previous releases - regressions:
- e1000e: re-enable s0ix power saving flows for systems with the
Intel i219-LM Ethernet controllers to fix power use regression
- virtio_net: fix recursive call to cpus_read_lock() to avoid a
deadlock
- ipv4: ignore ECN bits for fib lookups in fib_compute_spec_dst()
- sysfs: take the rtnl lock around XPS configuration
- xsk: fix memory leak for failed bind and rollback reservation at
NETDEV_TX_BUSY
- r8169: work around power-saving bug on some chip versions
Previous releases - always broken:
- dcb: validate netlink message in DCB handler
- tun: fix return value when the number of iovs exceeds MAX_SKB_FRAGS
to prevent unnecessary retries
- vhost_net: fix ubuf refcount when sendmsg fails
- bpf: save correct stopping point in file seq iteration
- ncsi: use real net-device for response handler
- neighbor: fix div by zero caused by a data race (TOCTOU)
- bareudp: fix use of incorrect min_headroom size and a false
positive lockdep splat from the TX lock
- mvpp2:
- clear force link UP during port init procedure in case
bootloader had set it
- add TCAM entry to drop flow control pause frames
- fix PPPoE with ipv6 packet parsing
- fix GoP Networking Complex Control config of port 3
- fix pkt coalescing IRQ-threshold configuration
- xsk: fix race in SKB mode transmit with shared cq
- ionic: account for vlan tag len in rx buffer len
- stmmac: ignore the second clock input, current clock framework does
not handle exclusive clock use well, other drivers may reconfigure
the second clock
Misc:
- ppp: change PPPIOCUNBRIDGECHAN ioctl request number to follow
existing scheme"
* tag 'net-5.11-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (99 commits)
net: dsa: lantiq_gswip: Fix GSWIP_MII_CFG(p) register access
net: dsa: lantiq_gswip: Enable GSWIP_MII_CFG_EN also for internal PHYs
net: lapb: Decrease the refcount of "struct lapb_cb" in lapb_device_event
r8169: work around power-saving bug on some chip versions
net: usb: qmi_wwan: add Quectel EM160R-GL
selftests: mlxsw: Set headroom size of correct port
net: macb: Correct usage of MACB_CAPS_CLK_HW_CHG flag
ibmvnic: fix: NULL pointer dereference.
docs: networking: packet_mmap: fix old config reference
docs: networking: packet_mmap: fix formatting for C macros
vhost_net: fix ubuf refcount incorrectly when sendmsg fails
bareudp: Fix use of incorrect min_headroom size
bareudp: set NETIF_F_LLTX flag
net: hdlc_ppp: Fix issues when mod_timer is called while timer is running
atlantic: remove architecture depends
erspan: fix version 1 check in gre_parse_header()
net: hns: fix return value check in __lb_other_process()
net: sched: prevent invalid Scell_log shift count
net: neighbor: fix a crash caused by mod zero
ipv4: Ignore ECN bits for fib lookups in fib_compute_spec_dst()
...
|
|
Pablo Neira Ayuso says:
====================
Netfilter fixes for net
The following patchset contains Netfilter fixes for net:
1) Missing sanitization of rateest userspace string, bug has been
triggered by syzbot, patch from Florian Westphal.
2) Report EOPNOTSUPP on missing set features in nft_dynset, otherwise
error reporting to userspace via EINVAL is misleading since this is
reserved for malformed netlink requests.
3) New binaries with old kernels might silently accept several set
element expressions. New binaries set on the NFT_SET_EXPR and
NFT_DYNSET_F_EXPR flags to request for several expressions per
element, hence old kernels which do not support for this bail out
with EOPNOTSUPP.
* git://git.kernel.org/pub/scm/linux/kernel/git/pablo/nf:
netfilter: nftables: add set expression flags
netfilter: nft_dynset: report EOPNOTSUPP on missing set feature
netfilter: xt_RATEEST: reject non-null terminated string from userspace
====================
Link: https://lore.kernel.org/r/20210103192920.18639-1-pablo@netfilter.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
In lapb_device_event, lapb_devtostruct is called to get a reference to
an object of "struct lapb_cb". lapb_devtostruct increases the refcount
of the object and returns a pointer to it. However, we didn't decrease
the refcount after we finished using the pointer. This patch fixes this
problem.
Fixes: a4989fa91110 ("net/lapb: support netdev events")
Cc: Martin Schiller <ms@dev.tdt.de>
Signed-off-by: Xie He <xie.he.0141@gmail.com>
Link: https://lore.kernel.org/r/20201231174331.64539-1-xie.he.0141@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
Since a few years, kernel addresses are no longer included in oops
dumps, at least on x86. All we get is a symbol name with offset and
size.
This is a problem for ceph_connection_operations handlers, especially
con->ops->dispatch(). All three handlers have the same name and there
is little context to disambiguate between e.g. monitor and OSD clients
because almost everything is inlined. gdb sneakily stops at the first
matching symbol, so one has to resort to nm and addr2line.
Some of these are already prefixed with mon_, osd_ or mds_. Let's do
the same for all others.
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
Acked-by: Jeff Layton <jlayton@kernel.org>
|
|
Try and avoid leaving bits and pieces of session key and connection
secret (gets split into GCM key and a pair of GCM IVs) around.
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
Reviewed-by: Jeff Layton <jlayton@kernel.org>
|
|
Use three-way comparison for address components to avoid integer
wraparound in the result of xfrm_policy_addr_delta(). This ensures
that the search trees are built and traversed correctly.
Treat IPv4 and IPv6 similarly by returning 0 when prefixlen == 0.
Prefix /0 has only one equivalence class.
Fixes: 9cf545ebd591d ("xfrm: policy: store inexact policies in a tree ordered by destination address")
Signed-off-by: Visa Hankala <visa@hankala.org>
Acked-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
|
|
xfrm_probe_algs() probes kernel crypto modules and changes the
availability of struct xfrm_algo_desc. But there is a small window
where ealg->available and aalg->available get changed between
count_ah_combs()/count_esp_combs() and dump_ah_combs()/dump_esp_combs(),
in this case we may allocate a smaller skb but later put a larger
amount of data and trigger the panic in skb_put().
Fix this by relaxing the checks when counting the size, that is,
skipping the test of ->available. We may waste some memory for a few
of sizeof(struct sadb_comb), but it is still much better than a panic.
Reported-by: syzbot+b2bf2652983d23734c5c@syzkaller.appspotmail.com
Cc: Steffen Klassert <steffen.klassert@secunet.com>
Cc: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: Cong Wang <cong.wang@bytedance.com>
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
|
|
The disable_xfrm flag signals that xfrm should not be performed during
routing towards a device before reaching device xmit.
For xfrm interfaces this is usually desired as they perform the outbound
policy lookup as part of their xmit using their if_id.
Before this change enabling this flag on xfrm interfaces prevented them
from xmitting as xfrm_lookup_with_ifid() would not perform a policy lookup
in case the original dst had the DST_NOXFRM flag.
This optimization is incorrect when the lookup is done by the xfrm
interface xmit logic.
Fix by performing policy lookup when invoked by xfrmi as if_id != 0.
Similarly it's unlikely for the 'no policy exists on net' check to yield
any performance benefits when invoked from xfrmi.
Fixes: f203b76d7809 ("xfrm: Add virtual xfrm interfaces")
Signed-off-by: Eyal Birger <eyal.birger@gmail.com>
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
|
|
Daniel Borkmann says:
====================
pull-request: bpf 2020-12-28
The following pull-request contains BPF updates for your *net* tree.
There is a small merge conflict between bpf tree commit 69ca310f3416
("bpf: Save correct stopping point in file seq iteration") and net tree
commit 66ed594409a1 ("bpf/task_iter: In task_file_seq_get_next use
task_lookup_next_fd_rcu"). The get_files_struct() does not exist anymore
in net, so take the hunk in HEAD and add the `info->tid = curr_tid` to
the error path:
[...]
curr_task = task_seq_get_next(ns, &curr_tid, true);
if (!curr_task) {
info->task = NULL;
info->tid = curr_tid;
return NULL;
}
/* set info->task and info->tid */
[...]
We've added 10 non-merge commits during the last 9 day(s) which contain
a total of 11 files changed, 75 insertions(+), 20 deletions(-).
The main changes are:
1) Various AF_XDP fixes such as fill/completion ring leak on failed bind and
fixing a race in skb mode's backpressure mechanism, from Magnus Karlsson.
2) Fix latency spikes on lockdep enabled kernels by adding a rescheduling
point to BPF hashtab initialization, from Eric Dumazet.
3) Fix a splat in task iterator by saving the correct stopping point in the
seq file iteration, from Jonathan Lemon.
4) Fix BPF maps selftest by adding retries in case hashtab returns EBUSY
errors on update/deletes, from Andrii Nakryiko.
5) Fix BPF selftest error reporting to something more user friendly if the
vmlinux BTF cannot be found, from Kamal Mostafa.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Both version 0 and version 1 use ETH_P_ERSPAN, but version 0 does not
have an erspan header. So the check in gre_parse_header() is wrong,
we have to distinguish version 1 from version 0.
We can just check the gre header length like is_erspan_type1().
Fixes: cb73ee40b1b3 ("net: ip_gre: use erspan key field for tunnel lookup")
Reported-by: syzbot+f583ce3d4ddf9836b27a@syzkaller.appspotmail.com
Cc: William Tu <u9012063@gmail.com>
Cc: Lorenzo Bianconi <lorenzo.bianconi@redhat.com>
Signed-off-by: Cong Wang <cong.wang@bytedance.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Check Scell_log shift size in red_check_params() and modify all callers
of red_check_params() to pass Scell_log.
This prevents a shift out-of-bounds as detected by UBSAN:
UBSAN: shift-out-of-bounds in ./include/net/red.h:252:22
shift exponent 72 is too large for 32-bit type 'int'
Fixes: 8afa10cbe281 ("net_sched: red: Avoid illegal values")
Signed-off-by: Randy Dunlap <rdunlap@infradead.org>
Reported-by: syzbot+97c5bd9cc81eca63d36e@syzkaller.appspotmail.com
Cc: Nogah Frankel <nogahf@mellanox.com>
Cc: Jamal Hadi Salim <jhs@mojatatu.com>
Cc: Cong Wang <xiyou.wangcong@gmail.com>
Cc: Jiri Pirko <jiri@resnulli.us>
Cc: netdev@vger.kernel.org
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
pneigh_enqueue() tries to obtain a random delay by mod
NEIGH_VAR(p, PROXY_DELAY). However, NEIGH_VAR(p, PROXY_DELAY)
migth be zero at that point because someone could write zero
to /proc/sys/net/ipv4/neigh/[device]/proxy_delay after the
callers check it.
This patch uses prandom_u32_max() to get a random delay instead
which avoids potential division by zero.
Signed-off-by: weichenchen <weichen.chen@linux.alibaba.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
RT_TOS() only clears one of the ECN bits. Therefore, when
fib_compute_spec_dst() resorts to a fib lookup, it can return
different results depending on the value of the second ECN bit.
For example, ECT(0) and ECT(1) packets could be treated differently.
$ ip netns add ns0
$ ip netns add ns1
$ ip link add name veth01 netns ns0 type veth peer name veth10 netns ns1
$ ip -netns ns0 link set dev lo up
$ ip -netns ns1 link set dev lo up
$ ip -netns ns0 link set dev veth01 up
$ ip -netns ns1 link set dev veth10 up
$ ip -netns ns0 address add 192.0.2.10/24 dev veth01
$ ip -netns ns1 address add 192.0.2.11/24 dev veth10
$ ip -netns ns1 address add 192.0.2.21/32 dev lo
$ ip -netns ns1 route add 192.0.2.10/32 tos 4 dev veth10 src 192.0.2.21
$ ip netns exec ns1 sysctl -wq net.ipv4.icmp_echo_ignore_broadcasts=0
With TOS 4 and ECT(1), ns1 replies using source address 192.0.2.21
(ping uses -Q to set all TOS and ECN bits):
$ ip netns exec ns0 ping -c 1 -b -Q 5 192.0.2.255
[...]
64 bytes from 192.0.2.21: icmp_seq=1 ttl=64 time=0.544 ms
But with TOS 4 and ECT(0), ns1 replies using source address 192.0.2.11
because the "tos 4" route isn't matched:
$ ip netns exec ns0 ping -c 1 -b -Q 6 192.0.2.255
[...]
64 bytes from 192.0.2.11: icmp_seq=1 ttl=64 time=0.597 ms
After this patch the ECN bits don't affect the result anymore:
$ ip netns exec ns0 ping -c 1 -b -Q 6 192.0.2.255
[...]
64 bytes from 192.0.2.21: icmp_seq=1 ttl=64 time=0.591 ms
Fixes: 35ebf65e851c ("ipv4: Create and use fib_compute_spec_dst() helper.")
Signed-off-by: Guillaume Nault <gnault@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
the following syzkaller reproducer:
r0 = socket$inet_mptcp(0x2, 0x1, 0x106)
bind$inet(r0, &(0x7f0000000080)={0x2, 0x4e24, @multicast2}, 0x10)
connect$inet(r0, &(0x7f0000000480)={0x2, 0x4e24, @local}, 0x10)
sendto$inet(r0, &(0x7f0000000100)="f6", 0xffffffe7, 0xc000, 0x0, 0x0)
systematically triggers the following warning:
WARNING: CPU: 2 PID: 8618 at net/core/stream.c:208 sk_stream_kill_queues+0x3fa/0x580
Modules linked in:
CPU: 2 PID: 8618 Comm: syz-executor Not tainted 5.10.0+ #334
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.11.1-4.module+el8.1.0+4066+0f1aadab 04/04
RIP: 0010:sk_stream_kill_queues+0x3fa/0x580
Code: df 48 c1 ea 03 0f b6 04 02 84 c0 74 04 3c 03 7e 40 8b ab 20 02 00 00 e9 64 ff ff ff e8 df f0 81 2
RSP: 0018:ffffc9000290fcb0 EFLAGS: 00010293
RAX: ffff888011cb8000 RBX: 0000000000000000 RCX: ffffffff86eecf0e
RDX: 0000000000000000 RSI: ffffffff86eecf6a RDI: 0000000000000005
RBP: 0000000000000e28 R08: ffff888011cb8000 R09: fffffbfff1f48139
R10: ffffffff8fa409c7 R11: fffffbfff1f48138 R12: ffff8880215e6220
R13: ffffffff8fa409c0 R14: ffffc9000290fd30 R15: 1ffff92000521fa2
FS: 00007f41c78f4800(0000) GS:ffff88802d000000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00007f95c803d088 CR3: 0000000025ed2000 CR4: 00000000000006f0
Call Trace:
__mptcp_destroy_sock+0x4f5/0x8e0
mptcp_close+0x5e2/0x7f0
inet_release+0x12b/0x270
__sock_release+0xc8/0x270
sock_close+0x18/0x20
__fput+0x272/0x8e0
task_work_run+0xe0/0x1a0
exit_to_user_mode_prepare+0x1df/0x200
syscall_exit_to_user_mode+0x19/0x50
entry_SYSCALL_64_after_hwframe+0x44/0xa9
userspace programs provide arbitrarily high values of 'len' in sendmsg():
this is causing integer overflow of 'amount'. Cap forward allocation to 1
megabyte: higher values are not really useful.
Suggested-by: Paolo Abeni <pabeni@redhat.com>
Fixes: e93da92896bc ("mptcp: implement wmem reservation")
Signed-off-by: Davide Caratti <dcaratti@redhat.com>
Link: https://lore.kernel.org/r/3334d00d8b2faecafdfab9aa593efcbf61442756.1608584474.git.dcaratti@redhat.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
Accesses to dev->xps_rxqs_map (when using dev->num_tc) should be
protected by the rtnl lock, like we do for netif_set_xps_queue. I didn't
see an actual bug being triggered, but let's be safe here and take the
rtnl lock while accessing the map in sysfs.
Fixes: 8af2c06ff4b1 ("net-sysfs: Add interface for Rx queue(s) map per Tx queue")
Signed-off-by: Antoine Tenart <atenart@kernel.org>
Reviewed-by: Alexander Duyck <alexanderduyck@fb.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
Two race conditions can be triggered when storing xps rxqs, resulting in
various oops and invalid memory accesses:
1. Calling netdev_set_num_tc while netif_set_xps_queue:
- netif_set_xps_queue uses dev->tc_num as one of the parameters to
compute the size of new_dev_maps when allocating it. dev->tc_num is
also used to access the map, and the compiler may generate code to
retrieve this field multiple times in the function.
- netdev_set_num_tc sets dev->tc_num.
If new_dev_maps is allocated using dev->tc_num and then dev->tc_num
is set to a higher value through netdev_set_num_tc, later accesses to
new_dev_maps in netif_set_xps_queue could lead to accessing memory
outside of new_dev_maps; triggering an oops.
2. Calling netif_set_xps_queue while netdev_set_num_tc is running:
2.1. netdev_set_num_tc starts by resetting the xps queues,
dev->tc_num isn't updated yet.
2.2. netif_set_xps_queue is called, setting up the map with the
*old* dev->num_tc.
2.3. netdev_set_num_tc updates dev->tc_num.
2.4. Later accesses to the map lead to out of bound accesses and
oops.
A similar issue can be found with netdev_reset_tc.
One way of triggering this is to set an iface up (for which the driver
uses netdev_set_num_tc in the open path, such as bnx2x) and writing to
xps_rxqs in a concurrent thread. With the right timing an oops is
triggered.
Both issues have the same fix: netif_set_xps_queue, netdev_set_num_tc
and netdev_reset_tc should be mutually exclusive. We do that by taking
the rtnl lock in xps_rxqs_store.
Fixes: 8af2c06ff4b1 ("net-sysfs: Add interface for Rx queue(s) map per Tx queue")
Signed-off-by: Antoine Tenart <atenart@kernel.org>
Reviewed-by: Alexander Duyck <alexanderduyck@fb.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
Accesses to dev->xps_cpus_map (when using dev->num_tc) should be
protected by the rtnl lock, like we do for netif_set_xps_queue. I didn't
see an actual bug being triggered, but let's be safe here and take the
rtnl lock while accessing the map in sysfs.
Fixes: 184c449f91fe ("net: Add support for XPS with QoS via traffic classes")
Signed-off-by: Antoine Tenart <atenart@kernel.org>
Reviewed-by: Alexander Duyck <alexanderduyck@fb.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
Two race conditions can be triggered when storing xps cpus, resulting in
various oops and invalid memory accesses:
1. Calling netdev_set_num_tc while netif_set_xps_queue:
- netif_set_xps_queue uses dev->tc_num as one of the parameters to
compute the size of new_dev_maps when allocating it. dev->tc_num is
also used to access the map, and the compiler may generate code to
retrieve this field multiple times in the function.
- netdev_set_num_tc sets dev->tc_num.
If new_dev_maps is allocated using dev->tc_num and then dev->tc_num
is set to a higher value through netdev_set_num_tc, later accesses to
new_dev_maps in netif_set_xps_queue could lead to accessing memory
outside of new_dev_maps; triggering an oops.
2. Calling netif_set_xps_queue while netdev_set_num_tc is running:
2.1. netdev_set_num_tc starts by resetting the xps queues,
dev->tc_num isn't updated yet.
2.2. netif_set_xps_queue is called, setting up the map with the
*old* dev->num_tc.
2.3. netdev_set_num_tc updates dev->tc_num.
2.4. Later accesses to the map lead to out of bound accesses and
oops.
A similar issue can be found with netdev_reset_tc.
One way of triggering this is to set an iface up (for which the driver
uses netdev_set_num_tc in the open path, such as bnx2x) and writing to
xps_cpus in a concurrent thread. With the right timing an oops is
triggered.
Both issues have the same fix: netif_set_xps_queue, netdev_set_num_tc
and netdev_reset_tc should be mutually exclusive. We do that by taking
the rtnl lock in xps_cpus_store.
Fixes: 184c449f91fe ("net: Add support for XPS with QoS via traffic classes")
Signed-off-by: Antoine Tenart <atenart@kernel.org>
Reviewed-by: Alexander Duyck <alexanderduyck@fb.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
crypto_shash_setkey() and crypto_aead_setkey() will do a (small)
GFP_ATOMIC allocation to align the key if it isn't suitably aligned.
It's not a big deal, but at the same time easy to avoid.
The actual alignment requirement is dynamic, queryable with
crypto_shash_alignmask() and crypto_aead_alignmask(), but shouldn't
be stricter than 16 bytes for our algorithms.
Fixes: cd1a677cad99 ("libceph, ceph: implement msgr2.1 protocol (crc and secure modes)")
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
|
|
auth_signature frame is 68 bytes in plain mode and 96 bytes in
secure mode but we are requesting 68 bytes in both modes. By luck,
this doesn't actually result in any invalid memory accesses because
the allocation is satisfied out of kmalloc-96 slab and so exactly
96 bytes are allocated, but KASAN rightfully complains.
Fixes: cd1a677cad99 ("libceph, ceph: implement msgr2.1 protocol (crc and secure modes)")
Reported-by: Luis Henriques <lhenriques@suse.de>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
|
|
The set flag NFT_SET_EXPR provides a hint to the kernel that userspace
supports for multiple expressions per set element. In the same
direction, NFT_DYNSET_F_EXPR specifies that dynset expression defines
multiple expressions per set element.
This allows new userspace software with old kernels to bail out with
EOPNOTSUPP. This update is similar to ef516e8625dd ("netfilter:
nf_tables: reintroduce the NFT_SET_CONCAT flag"). The NFT_SET_EXPR flag
needs to be set on when the NFTA_SET_EXPRESSIONS attribute is specified.
The NFT_SET_EXPR flag is not set on with NFTA_SET_EXPR to retain
backward compatibility in old userspace binaries.
Fixes: 48b0ae046ee9 ("netfilter: nftables: netlink support for several set element expressions")
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
If userspace requests a feature which is not available the original set
definition, then bail out with EOPNOTSUPP. If userspace sends
unsupported dynset flags (new feature not supported by this kernel),
then report EOPNOTSUPP to userspace. EINVAL should be only used to
report malformed netlink messages from userspace.
Fixes: 22fe54d5fefc ("netfilter: nf_tables: add support for dynamic set updates")
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
syzbot reports:
detected buffer overflow in strlen
[..]
Call Trace:
strlen include/linux/string.h:325 [inline]
strlcpy include/linux/string.h:348 [inline]
xt_rateest_tg_checkentry+0x2a5/0x6b0 net/netfilter/xt_RATEEST.c:143
strlcpy assumes src is a c-string. Check info->name before its used.
Reported-by: syzbot+e86f7c428c8c50db65b4@syzkaller.appspotmail.com
Fixes: 5859034d7eb8793 ("[NETFILTER]: x_tables: add RATEEST target")
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
When aggregating ncsi interfaces and dedicated interfaces to bond
interfaces, the ncsi response handler will use the wrong net device to
find ncsi_dev, so that the ncsi interface will not work properly.
Here, we use the original net device to fix it.
Fixes: 138635cc27c9 ("net/ncsi: NCSI response packet handler")
Signed-off-by: John Wang <wangzhiqiang.bj@bytedance.com>
Link: https://lore.kernel.org/r/20201223055523.2069-1-wangzhiqiang.bj@bytedance.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
DCB uses the same handler function for both RTM_GETDCB and RTM_SETDCB
messages. dcb_doit() bounces RTM_SETDCB mesasges if the user does not have
the CAP_NET_ADMIN capability.
However, the operation to be performed is not decided from the DCB message
type, but from the DCB command. Thus DCB_CMD_*_GET commands are used for
reading DCB objects, the corresponding SET and DEL commands are used for
manipulation.
The assumption is that set-like commands will be sent via an RTM_SETDCB
message, and get-like ones via RTM_GETDCB. However, this assumption is not
enforced.
It is therefore possible to manipulate DCB objects without CAP_NET_ADMIN
capability by sending the corresponding command in an RTM_GETDCB message.
That is a bug. Fix it by validating the type of the request message against
the type used for the response.
Fixes: 2f90b8657ec9 ("ixgbe: this patch adds support for DCB to the kernel and ixgbe driver")
Signed-off-by: Petr Machata <me@pmachata.org>
Link: https://lore.kernel.org/r/a2a9b88418f3a58ef211b718f2970128ef9e3793.1608673640.git.me@pmachata.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
Pull 9p update from Dominique Martinet:
- fix long-standing limitation on open-unlink-fop pattern
- add refcount to p9_fid (fixes the above and will allow for more
cleanups and simplifications in the future)
* tag '9p-for-5.11-rc1' of git://github.com/martinetd/linux:
9p: Remove unnecessary IS_ERR() check
9p: Uninitialized variable in v9fs_writeback_fid()
9p: Fix writeback fid incorrectly being attached to dentry
9p: apply review requests for fid refcounting
9p: add refcount to p9_fid struct
fs/9p: search open fids first
fs/9p: track open fids
fs/9p: fix create-unlink-getattr idiom
|
|
When setting xfrm replay_window to values higher than 32, a rare
page-fault occurs in xfrm_replay_advance_bmp:
BUG: unable to handle page fault for address: ffff8af350ad7920
#PF: supervisor write access in kernel mode
#PF: error_code(0x0002) - not-present page
PGD ad001067 P4D ad001067 PUD 0
Oops: 0002 [#1] SMP PTI
CPU: 3 PID: 30 Comm: ksoftirqd/3 Kdump: loaded Not tainted 5.4.52-050452-generic #202007160732
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.11.0-2.el7 04/01/2014
RIP: 0010:xfrm_replay_advance_bmp+0xbb/0x130
RSP: 0018:ffffa1304013ba40 EFLAGS: 00010206
RAX: 000000000000010d RBX: 0000000000000002 RCX: 00000000ffffff4b
RDX: 0000000000000018 RSI: 00000000004c234c RDI: 00000000ffb3dbff
RBP: ffffa1304013ba50 R08: ffff8af330ad7920 R09: 0000000007fffffa
R10: 0000000000000800 R11: 0000000000000010 R12: ffff8af29d6258c0
R13: ffff8af28b95c700 R14: 0000000000000000 R15: ffff8af29d6258fc
FS: 0000000000000000(0000) GS:ffff8af339ac0000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: ffff8af350ad7920 CR3: 0000000015ee4000 CR4: 00000000001406e0
Call Trace:
xfrm_input+0x4e5/0xa10
xfrm4_rcv_encap+0xb5/0xe0
xfrm4_udp_encap_rcv+0x140/0x1c0
Analysis revealed offending code is when accessing:
replay_esn->bmp[nr] |= (1U << bitnr);
with 'nr' being 0x07fffffa.
This happened in an SMP system when reordering of packets was present;
A packet arrived with a "too old" sequence number (outside the window,
i.e 'diff > replay_window'), and therefore the following calculation:
bitnr = replay_esn->replay_window - (diff - pos);
yields a negative result, but since bitnr is u32 we get a large unsigned
quantity (in crash dump above: 0xffffff4b seen in ecx).
This was supposed to be protected by xfrm_input()'s former call to:
if (x->repl->check(x, skb, seq)) {
However, the state's spinlock x->lock is *released* after '->check()'
is performed, and gets re-acquired before '->advance()' - which gives a
chance for a different core to update the xfrm state, e.g. by advancing
'replay_esn->seq' when it encounters more packets - leading to a
'diff > replay_window' situation when original core continues to
xfrm_replay_advance_bmp().
An attempt to fix this issue was suggested in commit bcf66bf54aab
("xfrm: Perform a replay check after return from async codepaths"),
by calling 'x->repl->recheck()' after lock is re-acquired, but fix
applied only to asyncronous crypto algorithms.
Augment the fix, by *always* calling 'recheck()' - irrespective if we're
using async crypto.
Fixes: 0ebea8ef3559 ("[IPSEC]: Move state lock into x->type->input")
Signed-off-by: Shmulik Ladkani <shmulik.ladkani@gmail.com>
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
|
|
Pablo Neira Ayuso says:
====================
Netfilter fixes for net
1) Incorrect loop in error path of nft_set_elem_expr_clone(),
from Colin Ian King.
2) Missing xt_table_get_private_protected() to access table
private data in x_tables, from Subash Abhinov Kasiviswanathan.
3) Possible oops in ipset hash type resize, from Vasily Averin.
4) Fix shift-out-of-bounds in ipset hash type, also from Vasily.
* git://git.kernel.org/pub/scm/linux/kernel/git/pablo/nf:
netfilter: ipset: fix shift-out-of-bounds in htable_bits()
netfilter: ipset: fixes possible oops in mtype_resize
netfilter: x_tables: Update remaining dereference to RCU
netfilter: nftables: fix incorrect increment of loop counter
====================
Link: https://lore.kernel.org/r/20201218120409.3659-1-pablo@netfilter.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
taprio_graft() can insert a NULL element in the array of child qdiscs. As
a consquence, taprio_reset() might not reset child qdiscs completely, and
taprio_destroy() might leak resources. Fix it by ensuring that loops that
iterate over q->qdiscs[] don't end when they find the first NULL item.
Fixes: 44d4775ca518 ("net/sched: sch_taprio: reset child qdiscs before freeing them")
Fixes: 5a781ccbd19e ("tc: Add support for configuring the taprio scheduler")
Suggested-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Davide Caratti <dcaratti@redhat.com>
Link: https://lore.kernel.org/r/13edef6778fef03adc751582562fba4a13e06d6a.1608240532.git.dcaratti@redhat.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
On 64-bit systems the packet procfs header field names following 'sk'
are not aligned correctly:
sk RefCnt Type Proto Iface R Rmem User Inode
00000000605d2c64 3 3 0003 7 1 450880 0 16643
00000000080e9b80 2 2 0000 0 0 0 0 17404
00000000b23b8a00 2 2 0000 0 0 0 0 17421
...
With this change field names are correctly aligned:
sk RefCnt Type Proto Iface R Rmem User Inode
000000005c3b1d97 3 3 0003 7 1 21568 0 16178
000000007be55bb7 3 3 fbce 8 1 0 0 16250
00000000be62127d 3 3 fbcd 8 1 0 0 16254
...
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Link: https://lore.kernel.org/r/54917251d8433735d9a24e935a6cb8eb88b4058a.1608103684.git.baruch@tkos.co.il
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
Daire Byrne reports a ~50% aggregrate throughput regression on his
Linux NFS server after commit da1661b93bf4 ("SUNRPC: Teach server to
use xprt_sock_sendmsg for socket sends"), which replaced
kernel_send_page() calls in NFSD's socket send path with calls to
sock_sendmsg() using iov_iter.
Investigation showed that tcp_sendmsg() was not using zero-copy to
send the xdr_buf's bvec pages, but instead was relying on memcpy.
This means copying every byte of a large NFS READ payload.
It looks like TLS sockets do indeed support a ->sendpage method,
so it's really not necessary to use xprt_sock_sendmsg() to support
TLS fully on the server. A mechanical reversion of da1661b93bf4 is
not possible at this point, but we can re-implement the server's
TCP socket sendmsg path using kernel_sendpage().
Reported-by: Daire Byrne <daire@dneg.com>
BugLink: https://bugzilla.kernel.org/show_bug.cgi?id=209439
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
|
|
Rollback the reservation in the completion ring when we get a
NETDEV_TX_BUSY. When this error is received from the driver, we are
supposed to let the user application retry the transmit again. And in
order to do this, we need to roll back the failed send so it can be
retried. Unfortunately, we did not cancel the reservation we had made
in the completion ring. By not doing this, we actually make the
completion ring one entry smaller per NETDEV_TX_BUSY error we get, and
after enough of these errors the completion ring will be of size zero
and transmit will stop working.
Fix this by cancelling the reservation when we get a NETDEV_TX_BUSY
error.
Fixes: 642e450b6b59 ("xsk: Do not discard packet when NETDEV_TX_BUSY")
Reported-by: Xuan Zhuo <xuanzhuo@linux.alibaba.com>
Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Björn Töpel <bjorn.topel@intel.com>
Link: https://lore.kernel.org/bpf/20201218134525.13119-3-magnus.karlsson@gmail.com
|
|
Fix a race when multiple sockets are simultaneously calling sendto()
when the completion ring is shared in the SKB case. This is the case
when you share the same netdev and queue id through the
XDP_SHARED_UMEM bind flag. The problem is that multiple processes can
be in xsk_generic_xmit() and call the backpressure mechanism in
xskq_prod_reserve(xs->pool->cq). As this is a shared resource in this
specific scenario, a race might occur since the rings are
single-producer single-consumer.
Fix this by moving the tx_completion_lock from the socket to the pool
as the pool is shared between the sockets that share the completion
ring. (The pool is not shared when this is not the case.) And then
protect the accesses to xskq_prod_reserve() with this lock. The
tx_completion_lock is renamed cq_lock to better reflect that it
protects accesses to the potentially shared completion ring.
Fixes: 35fcde7f8deb ("xsk: support for Tx")
Reported-by: Xuan Zhuo <xuanzhuo@linux.alibaba.com>
Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Björn Töpel <bjorn.topel@intel.com>
Link: https://lore.kernel.org/bpf/20201218134525.13119-2-magnus.karlsson@gmail.com
|
|
Fix a possible memory leak when a bind of an AF_XDP socket fails. When
the fill and completion rings are created, they are tied to the
socket. But when the buffer pool is later created at bind time, the
ownership of these two rings are transferred to the buffer pool as
they might be shared between sockets (and the buffer pool cannot be
created until we know what we are binding to). So, before the buffer
pool is created, these two rings are cleaned up with the socket, and
after they have been transferred they are cleaned up together with
the buffer pool.
The problem is that ownership was transferred before it was absolutely
certain that the buffer pool could be created and initialized
correctly and when one of these errors occurred, the fill and
completion rings did neither belong to the socket nor the pool and
where therefore leaked. Solve this by moving the ownership transfer
to the point where the buffer pool has been completely set up and
there is no way it can fail.
Fixes: 7361f9c3d719 ("xsk: Move fill and completion rings to buffer pool")
Reported-by: syzbot+cfa88ddd0655afa88763@syzkaller.appspotmail.com
Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Björn Töpel <bjorn.topel@intel.com>
Link: https://lore.kernel.org/bpf/20201214085127.3960-1-magnus.karlsson@gmail.com
|
