We have identified a problem with the "oversubscription" policy in the
link transmission code.
When small messages are transmitted, and the sending link has reached
the transmit window limit, those messages will be bundled and put into
the link backlog queue. However, bundles of data messages are counted
at the 'CRITICAL' level, so that the counter for that level, instead of
the counter for the real, bundled message's level is the one being
increased.
Subsequent, to-be-bundled data messages at non-CRITICAL levels continue
to be tested against the unchanged counter for their own level, while
contributing to an unrestrained increase at the CRITICAL backlog level.
This leaves a gap in congestion control algorithm for small messages
that can result in starvation for other users or a "real" CRITICAL
user. Even that eventually can lead to buffer exhaustion & link reset.
We fix this by keeping a 'target_bskb' buffer pointer at each levels,
then when bundling, we only bundle messages at the same importance
level only. This way, we know exactly how many slots a certain level
have occupied in the queue, so can manage level congestion accurately.
By bundling messages at the same level, we even have more benefits. Let
consider this:
- One socket sends 64-byte messages at the 'CRITICAL' level;
- Another sends 4096-byte messages at the 'LOW' level;
When a 64-byte message comes and is bundled the first time, we put the
overhead of message bundle to it (+ 40-byte header, data copy, etc.)
for later use, but the next message can be a 4096-byte one that cannot
be bundled to the previous one. This means the last bundle carries only
one payload message which is totally inefficient, as for the receiver
also! Later on, another 64-byte message comes, now we make a new bundle
and the same story repeats...
With the new bundling algorithm, this will not happen, the 64-byte
messages will be bundled together even when the 4096-byte message(s)
comes in between. However, if the 4096-byte messages are sent at the
same level i.e. 'CRITICAL', the bundling algorithm will again cause the
same overhead.
Also, the same will happen even with only one socket sending small
messages at a rate close to the link transmit's one, so that, when one
message is bundled, it's transmitted shortly. Then, another message
comes, a new bundle is created and so on...
We will solve this issue radically by another patch.
Fixes: 365ad353c2 ("tipc: reduce risk of user starvation during link congestion")
Reported-by: Hoang Le <hoang.h.le@dektech.com.au>
Acked-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: Tuong Lien <tuong.t.lien@dektech.com.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
xennet_fill_frags() uses ~0U as return value when the sk_buff is not able
to cache extra fragments. This is incorrect because the return type of
xennet_fill_frags() is RING_IDX and 0xffffffff is an expected value for
ring buffer index.
In the situation when the rsp_cons is approaching 0xffffffff, the return
value of xennet_fill_frags() may become 0xffffffff which xennet_poll() (the
caller) would regard as error. As a result, queue->rx.rsp_cons is set
incorrectly because it is updated only when there is error. If there is no
error, xennet_poll() would be responsible to update queue->rx.rsp_cons.
Finally, queue->rx.rsp_cons would point to the rx ring buffer entries whose
queue->rx_skbs[i] and queue->grant_rx_ref[i] are already cleared to NULL.
This leads to NULL pointer access in the next iteration to process rx ring
buffer entries.
The symptom is similar to the one fixed in
commit 00b368502d ("xen-netfront: do not assume sk_buff_head list is
empty in error handling").
This patch changes the return type of xennet_fill_frags() to indicate
whether it is successful or failed. The queue->rx.rsp_cons will be
always updated inside this function.
Fixes: ad4f15dc2c ("xen/netfront: don't bug in case of too many frags")
Signed-off-by: Dongli Zhang <dongli.zhang@oracle.com>
Reviewed-by: Juergen Gross <jgross@suse.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Rajendra reported a kernel panic when a link was taken down:
[ 6870.263084] BUG: unable to handle kernel NULL pointer dereference at 00000000000000a8
[ 6870.271856] IP: [<ffffffff8efc5764>] __ipv6_ifa_notify+0x154/0x290
<snip>
[ 6870.570501] Call Trace:
[ 6870.573238] [<ffffffff8efc58c6>] ? ipv6_ifa_notify+0x26/0x40
[ 6870.579665] [<ffffffff8efc98ec>] ? addrconf_dad_completed+0x4c/0x2c0
[ 6870.586869] [<ffffffff8efe70c6>] ? ipv6_dev_mc_inc+0x196/0x260
[ 6870.593491] [<ffffffff8efc9c6a>] ? addrconf_dad_work+0x10a/0x430
[ 6870.600305] [<ffffffff8f01ade4>] ? __switch_to_asm+0x34/0x70
[ 6870.606732] [<ffffffff8ea93a7a>] ? process_one_work+0x18a/0x430
[ 6870.613449] [<ffffffff8ea93d6d>] ? worker_thread+0x4d/0x490
[ 6870.619778] [<ffffffff8ea93d20>] ? process_one_work+0x430/0x430
[ 6870.626495] [<ffffffff8ea99dd9>] ? kthread+0xd9/0xf0
[ 6870.632145] [<ffffffff8f01ade4>] ? __switch_to_asm+0x34/0x70
[ 6870.638573] [<ffffffff8ea99d00>] ? kthread_park+0x60/0x60
[ 6870.644707] [<ffffffff8f01ae77>] ? ret_from_fork+0x57/0x70
[ 6870.650936] Code: 31 c0 31 d2 41 b9 20 00 08 02 b9 09 00 00 0
addrconf_dad_work is kicked to be scheduled when a device is brought
up. There is a race between addrcond_dad_work getting scheduled and
taking the rtnl lock and a process taking the link down (under rtnl).
The latter removes the host route from the inet6_addr as part of
addrconf_ifdown which is run for NETDEV_DOWN. The former attempts
to use the host route in ipv6_ifa_notify. If the down event removes
the host route due to the race to the rtnl, then the BUG listed above
occurs.
This scenario does not occur when the ipv6 address is not kept
(net.ipv6.conf.all.keep_addr_on_down = 0) as addrconf_ifdown sets the
state of the ifp to DEAD. Handle when the addresses are kept by checking
IF_READY which is reset by addrconf_ifdown.
The 'dead' flag for an inet6_addr is set only under rtnl, in
addrconf_ifdown and it means the device is getting removed (or IPv6 is
disabled). The interesting cases for changing the idev flag are
addrconf_notify (NETDEV_UP and NETDEV_CHANGE) and addrconf_ifdown
(reset the flag). The former does not have the idev lock - only rtnl;
the latter has both. Based on that the existing dead + IF_READY check
can be moved to right after the rtnl_lock in addrconf_dad_work.
Fixes: f1705ec197 ("net: ipv6: Make address flushing on ifdown optional")
Reported-by: Rajendra Dendukuri <rajendra.dendukuri@broadcom.com>
Signed-off-by: David Ahern <dsahern@gmail.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The cited commit exposed an old retransmits_timed_out() bug
which assumed it could call tcp_model_timeout() with
TCP_RTO_MIN as rto_base for all states.
But flows in SYN_SENT or SYN_RECV state uses a different
RTO base (1 sec instead of 200 ms, unless BPF choses
another value)
This caused a reduction of SYN retransmits from 6 to 4 with
the default /proc/sys/net/ipv4/tcp_syn_retries value.
Fixes: a41e8a88b0 ("tcp: better handle TCP_USER_TIMEOUT in SYN_SENT state")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Marek Majkowski <marek@cloudflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Lockdep is unhappy if two locks from the same class are held.
Fix the below warning for hyperv and virtio sockets (vmci socket code
doesn't have the issue) by using lock_sock_nested() when __vsock_release()
is called recursively:
============================================
WARNING: possible recursive locking detected
5.3.0+ #1 Not tainted
--------------------------------------------
server/1795 is trying to acquire lock:
ffff8880c5158990 (sk_lock-AF_VSOCK){+.+.}, at: hvs_release+0x10/0x120 [hv_sock]
but task is already holding lock:
ffff8880c5158150 (sk_lock-AF_VSOCK){+.+.}, at: __vsock_release+0x2e/0xf0 [vsock]
other info that might help us debug this:
Possible unsafe locking scenario:
CPU0
----
lock(sk_lock-AF_VSOCK);
lock(sk_lock-AF_VSOCK);
*** DEADLOCK ***
May be due to missing lock nesting notation
2 locks held by server/1795:
#0: ffff8880c5d05ff8 (&sb->s_type->i_mutex_key#10){+.+.}, at: __sock_release+0x2d/0xa0
#1: ffff8880c5158150 (sk_lock-AF_VSOCK){+.+.}, at: __vsock_release+0x2e/0xf0 [vsock]
stack backtrace:
CPU: 5 PID: 1795 Comm: server Not tainted 5.3.0+ #1
Call Trace:
dump_stack+0x67/0x90
__lock_acquire.cold.67+0xd2/0x20b
lock_acquire+0xb5/0x1c0
lock_sock_nested+0x6d/0x90
hvs_release+0x10/0x120 [hv_sock]
__vsock_release+0x24/0xf0 [vsock]
__vsock_release+0xa0/0xf0 [vsock]
vsock_release+0x12/0x30 [vsock]
__sock_release+0x37/0xa0
sock_close+0x14/0x20
__fput+0xc1/0x250
task_work_run+0x98/0xc0
do_exit+0x344/0xc60
do_group_exit+0x47/0xb0
get_signal+0x15c/0xc50
do_signal+0x30/0x720
exit_to_usermode_loop+0x50/0xa0
do_syscall_64+0x24e/0x270
entry_SYSCALL_64_after_hwframe+0x49/0xbe
RIP: 0033:0x7f4184e85f31
Tested-by: Stefano Garzarella <sgarzare@redhat.com>
Signed-off-by: Dexuan Cui <decui@microsoft.com>
Reviewed-by: Stefano Garzarella <sgarzare@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Fix NULL-pointer dereference on tty open due to a failure to handle a
missing interrupt-in endpoint when probing modem ports:
BUG: kernel NULL pointer dereference, address: 0000000000000006
...
RIP: 0010:tiocmget_submit_urb+0x1c/0xe0 [hso]
...
Call Trace:
hso_start_serial_device+0xdc/0x140 [hso]
hso_serial_open+0x118/0x1b0 [hso]
tty_open+0xf1/0x490
Fixes: 542f548236 ("tty: Modem functions for the HSO driver")
Signed-off-by: Johan Hovold <johan@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch is syncing driver with actual devicetree documentation:
Documentation/devicetree/bindings/net/qca,ar71xx.txt
|Optional subnodes:
|- mdio : specifies the mdio bus, used as a container for phy nodes
| according to phy.txt in the same directory
The driver was working with fixed phy without any noticeable issues. This bug
was uncovered by introducing dsa ar9331-switch driver.
Since no one reported this bug until now, I assume no body is using it
and this patch should not brake existing system.
Fixes: d51b6ce441 ("net: ethernet: add ag71xx driver")
Signed-off-by: Oleksij Rempel <o.rempel@pengutronix.de>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
Jose Abreu says:
====================
net: stmmac: Fixes for -net
Misc fixes for -net tree. More info in commit logs.
v2 is just a rebase of v1 against -net and we added a new patch (09/09) to
fix RSS feature.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Commit b6b6cc9acd, changed the call to dwxgmac2_rss_write_reg()
passing it the variable cfg->key[i].
As key is an u8 but we write 32 bits at a time we need to cast it into
an u32 so that the correct key values are written. Notice that the for
loop already takes this into account so we don't try to write past the
keys size.
Fixes: b6b6cc9acd ("net: stmmac: selftest: avoid large stack usage")
Signed-off-by: Jose Abreu <Jose.Abreu@synopsys.com>
Reviewed-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
The sizeof(cfg->key) is != ARRAY_SIZE(cfg->key). Fix it. This warning is
triggered when running with cc flag -Wsizeof-array-div.
Reported-by: kbuild test robot <lkp@intel.com>
Reported-by: Nick Desaulniers <ndesaulniers@google.com>
Reported-by: Nathan Chancellor <natechancellor@gmail.com>
Fixes: 76067459c6 ("net: stmmac: Implement RSS and enable it in XGMAC core")
Reviewed-by: Nick Desaulniers <ndesaulniers@google.com>
Signed-off-by: Jose Abreu <Jose.Abreu@synopsys.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
We don't use it anyway as XGMAC only supports polling for timestamp (in
current SW implementation). This greatly reduces the system load by
reducing the number of interrupts.
Fixes: 2142754f8b ("net: stmmac: Add MAC related callbacks for XGMAC2")
Signed-off-by: Jose Abreu <Jose.Abreu@synopsys.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
If WoL is enabled we can't really stop the PHY, otherwise we will not
receive the WoL packet. Fix this by telling phylink that only the MAC is
down and only stop the PHY if WoL is not enabled.
Fixes: 74371272f9 ("net: stmmac: Convert to phylink and remove phylib logic")
Signed-off-by: Jose Abreu <Jose.Abreu@synopsys.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The case for PTPV2_EVENT requires event packets to be captured so add
this setting to the list of enabled captures.
Fixes: 891434b18e ("stmmac: add IEEE PTPv1 and PTPv2 support.")
Signed-off-by: Jose Abreu <Jose.Abreu@synopsys.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
We need to always update the MAC Hash Filter so that previous entries
are invalidated.
Found out while running stmmac selftests.
Fixes: b8ef7020d6 ("net: stmmac: add support for hash table size 128/256 in dwmac4")
Signed-off-by: Jose Abreu <Jose.Abreu@synopsys.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Although some XGMAC setups support frames larger than DMA size, some of
them may not. As we can't know before-hand which ones support let's use
the maximum DMA buffer size in the Jumbo Tests.
User can always reconfigure the MTU to achieve larger frames.
Fixes: 427849e8c3 ("net: stmmac: selftests: Add Jumbo Frame tests")
Signed-off-by: Jose Abreu <Jose.Abreu@synopsys.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Since commit b8ef7020d6 ("net: stmmac: add support for hash table size
128/256 in dwmac4"), we can detect the Hash Table dinamically.
Let's implement this feature in XGMAC cores and fix possible setups that
don't support the maximum size for Hash Table.
Signed-off-by: Jose Abreu <Jose.Abreu@synopsys.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Some setups may not have all Unicast addresses filters available. Let's
check this before trying to setup filters.
Fixes: 0efedbf11f ("net: stmmac: xgmac: Fix XGMAC selftests")
Signed-off-by: Jose Abreu <Jose.Abreu@synopsys.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
If any of the param or info_get op returns error, dumpit cb is
skipping to dump remaining params or info_get ops for all the
drivers.
Fix to not return if any of the param/info_get op returns error
as not supported and continue to dump remaining information.
v2: Modify the patch to return error, except for params/info_get
op that return -EOPNOTSUPP as suggested by Andrew Lunn. Also, modify
commit message to reflect the same.
Cc: Andrew Lunn <andrew@lunn.ch>
Cc: Jiri Pirko <jiri@mellanox.com>
Cc: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: Vasundhara Volam <vasundhara-v.volam@broadcom.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
of_node_put needs to be called when the device node which is got
from of_get_child_by_name finished using.
irq_domain_add_linear() also calls of_node_get() to increase refcount,
so irq_domain will not be affected when it is released.
Fixes: d8652956cf ("net: dsa: realtek-smi: Add Realtek SMI driver")
Signed-off-by: Wen Yang <wenyang@linux.alibaba.com>
Cc: Linus Walleij <linus.walleij@linaro.org>
Cc: Andrew Lunn <andrew@lunn.ch>
Cc: Vivien Didelot <vivien.didelot@gmail.com>
Cc: Florian Fainelli <f.fainelli@gmail.com>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: netdev@vger.kernel.org
Cc: linux-kernel@vger.kernel.org
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
of_node_put needs to be called when the device node which is got
from of_get_child_by_name finished using.
In both cases of success and failure, we need to release 'ports',
so clean up the code using goto.
fixes: a556c76adc ("net: mscc: Add initial Ocelot switch support")
Signed-off-by: Wen Yang <wenyang@linux.alibaba.com>
Cc: Alexandre Belloni <alexandre.belloni@bootlin.com>
Cc: Microchip Linux Driver Support <UNGLinuxDriver@microchip.com>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: netdev@vger.kernel.org
Cc: linux-kernel@vger.kernel.org
Signed-off-by: David S. Miller <davem@davemloft.net>
As explained in the "net: sched: taprio: Avoid division by zero on
invalid link speed" commit, it is legal for the ethtool API to return
zero as a link speed. So guard against it to ensure we don't perform a
division by zero in kernel.
Fixes: e0a7683d30 ("net/sched: cbs: fix port_rate miscalculation")
Signed-off-by: Vladimir Oltean <olteanv@gmail.com>
Acked-by: Vinicius Costa Gomes <vinicius.gomes@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The check in taprio_set_picos_per_byte is currently not robust enough
and will trigger this division by zero, due to e.g. PHYLINK not setting
kset->base.speed when there is no PHY connected:
[ 27.109992] Division by zero in kernel.
[ 27.113842] CPU: 1 PID: 198 Comm: tc Not tainted 5.3.0-rc5-01246-gc4006b8c2637-dirty #212
[ 27.121974] Hardware name: Freescale LS1021A
[ 27.126234] [<c03132e0>] (unwind_backtrace) from [<c030d8b8>] (show_stack+0x10/0x14)
[ 27.133938] [<c030d8b8>] (show_stack) from [<c10b21b0>] (dump_stack+0xb0/0xc4)
[ 27.141124] [<c10b21b0>] (dump_stack) from [<c10af97c>] (Ldiv0_64+0x8/0x18)
[ 27.148052] [<c10af97c>] (Ldiv0_64) from [<c0700260>] (div64_u64+0xcc/0xf0)
[ 27.154978] [<c0700260>] (div64_u64) from [<c07002d0>] (div64_s64+0x4c/0x68)
[ 27.161993] [<c07002d0>] (div64_s64) from [<c0f3d890>] (taprio_set_picos_per_byte+0xe8/0xf4)
[ 27.170388] [<c0f3d890>] (taprio_set_picos_per_byte) from [<c0f3f614>] (taprio_change+0x668/0xcec)
[ 27.179302] [<c0f3f614>] (taprio_change) from [<c0f2bc24>] (qdisc_create+0x1fc/0x4f4)
[ 27.187091] [<c0f2bc24>] (qdisc_create) from [<c0f2c0c8>] (tc_modify_qdisc+0x1ac/0x6f8)
[ 27.195055] [<c0f2c0c8>] (tc_modify_qdisc) from [<c0ee9604>] (rtnetlink_rcv_msg+0x268/0x2dc)
[ 27.203449] [<c0ee9604>] (rtnetlink_rcv_msg) from [<c0f4fef0>] (netlink_rcv_skb+0xe0/0x114)
[ 27.211756] [<c0f4fef0>] (netlink_rcv_skb) from [<c0f4f6cc>] (netlink_unicast+0x1b4/0x22c)
[ 27.219977] [<c0f4f6cc>] (netlink_unicast) from [<c0f4fa84>] (netlink_sendmsg+0x284/0x340)
[ 27.228198] [<c0f4fa84>] (netlink_sendmsg) from [<c0eae5fc>] (sock_sendmsg+0x14/0x24)
[ 27.235988] [<c0eae5fc>] (sock_sendmsg) from [<c0eaedf8>] (___sys_sendmsg+0x214/0x228)
[ 27.243863] [<c0eaedf8>] (___sys_sendmsg) from [<c0eb015c>] (__sys_sendmsg+0x50/0x8c)
[ 27.251652] [<c0eb015c>] (__sys_sendmsg) from [<c0301000>] (ret_fast_syscall+0x0/0x54)
[ 27.259524] Exception stack(0xe8045fa8 to 0xe8045ff0)
[ 27.264546] 5fa0: b6f608c8 000000f8 00000003 bed7e2f0 00000000 00000000
[ 27.272681] 5fc0: b6f608c8 000000f8 004ce54c 00000128 5d3ce8c7 00000000 00000026 00505c9c
[ 27.280812] 5fe0: 00000070 bed7e298 004ddd64 b6dd1e64
Russell King points out that the ethtool API says zero is a valid return
value of __ethtool_get_link_ksettings:
* If it is enabled then they are read-only; if the link
* is up they represent the negotiated link mode; if the link is down,
* the speed is 0, %SPEED_UNKNOWN or the highest enabled speed and
* @duplex is %DUPLEX_UNKNOWN or the best enabled duplex mode.
So, it seems that taprio is not following the API... I'd suggest either
fixing taprio, or getting agreement to change the ethtool API.
The chosen path was to fix taprio.
Fixes: 7b9eba7ba0 ("net/sched: taprio: fix picos_per_byte miscalculation")
Signed-off-by: Vladimir Oltean <olteanv@gmail.com>
Acked-by: Vinicius Costa Gomes <vinicius.gomes@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
* two null pointer dereference fixes
* a fix for preempt-enabled/BHs-enabled (lockdep) splats
(that correctly pointed out a bug)
* a fix for multi-BSSID ordering assumptions
* a fix for the EDMG support, on-stack chandefs need to
be initialized properly (now that they're bigger)
* beacon (head) data from userspace should be validated
-----BEGIN PGP SIGNATURE-----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=nhNe
-----END PGP SIGNATURE-----
Merge tag 'mac80211-for-davem-2019-10-01' of git://git.kernel.org/pub/scm/linux/kernel/git/jberg/mac80211
Johannes Berg says:
====================
A small list of fixes this time:
* two null pointer dereference fixes
* a fix for preempt-enabled/BHs-enabled (lockdep) splats
(that correctly pointed out a bug)
* a fix for multi-BSSID ordering assumptions
* a fix for the EDMG support, on-stack chandefs need to
be initialized properly (now that they're bigger)
* beacon (head) data from userspace should be validated
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
When no other driver selects the devlink library code, ionic
produces a link failure:
drivers/net/ethernet/pensando/ionic/ionic_devlink.o: In function `ionic_devlink_alloc':
ionic_devlink.c:(.text+0xd): undefined reference to `devlink_alloc'
drivers/net/ethernet/pensando/ionic/ionic_devlink.o: In function `ionic_devlink_register':
ionic_devlink.c:(.text+0x71): undefined reference to `devlink_register'
Add the same 'select' statement that the other drivers use here.
Fixes: fbfb803153 ("ionic: Add hardware init and device commands")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: Shannon Nelson <snelson@pensando.io>
Signed-off-by: David S. Miller <davem@davemloft.net>
Resolving a couple of Sphinx documentation warnings
that are generated in the networking section.
- WARNING: document isn't included in any toctree
- WARNING: Title underline too short.
Signed-off-by: Adam Zerella <adam.zerella@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Always acquire tx descriptor spinlock even if a xdp program is not loaded
on the netsec device since ndo_xdp_xmit can run concurrently with
netsec_netdev_start_xmit and netsec_clean_tx_dring. This can happen
loading a xdp program on a different device (e.g virtio-net) and
xdp_do_redirect_map/xdp_do_redirect_slow can redirect to netsec even if
we do not have a xdp program on it.
Fixes: ba2b232108 ("net: netsec: add XDP support")
Tested-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Reviewed-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Acked-by: Toke Høiland-Jørgensen <toke@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
If the interface type is P2P_DEVICE or NAN, read the file of
'/sys/kernel/debug/ieee80211/phyx/netdev:wlanx/aqm' will get a
NULL pointer dereference. As for those interface type, the
pointer sdata->vif.txq is NULL.
Unable to handle kernel NULL pointer dereference at virtual address 00000011
CPU: 1 PID: 30936 Comm: cat Not tainted 4.14.104 #1
task: ffffffc0337e4880 task.stack: ffffff800cd20000
PC is at ieee80211_if_fmt_aqm+0x34/0xa0 [mac80211]
LR is at ieee80211_if_fmt_aqm+0x34/0xa0 [mac80211]
[...]
Process cat (pid: 30936, stack limit = 0xffffff800cd20000)
[...]
[<ffffff8000b7cd00>] ieee80211_if_fmt_aqm+0x34/0xa0 [mac80211]
[<ffffff8000b7c414>] ieee80211_if_read+0x60/0xbc [mac80211]
[<ffffff8000b7ccc4>] ieee80211_if_read_aqm+0x28/0x30 [mac80211]
[<ffffff80082eff94>] full_proxy_read+0x2c/0x48
[<ffffff80081eef00>] __vfs_read+0x2c/0xd4
[<ffffff80081ef084>] vfs_read+0x8c/0x108
[<ffffff80081ef494>] SyS_read+0x40/0x7c
Signed-off-by: Miaoqing Pan <miaoqing@codeaurora.org>
Acked-by: Toke Høiland-Jørgensen <toke@redhat.com>
Link: https://lore.kernel.org/r/1569549796-8223-1-git-send-email-miaoqing@codeaurora.org
[trim useless data from commit message]
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
If the interface is not in MESH mode, the command 'iw wlanx mpath del'
will cause kernel panic.
The root cause is null pointer access in mpp_flush_by_proxy(), as the
pointer 'sdata->u.mesh.mpp_paths' is NULL for non MESH interface.
Unable to handle kernel NULL pointer dereference at virtual address 00000068
[...]
PC is at _raw_spin_lock_bh+0x20/0x5c
LR is at mesh_path_del+0x1c/0x17c [mac80211]
[...]
Process iw (pid: 4537, stack limit = 0xd83e0238)
[...]
[<c021211c>] (_raw_spin_lock_bh) from [<bf8c7648>] (mesh_path_del+0x1c/0x17c [mac80211])
[<bf8c7648>] (mesh_path_del [mac80211]) from [<bf6cdb7c>] (extack_doit+0x20/0x68 [compat])
[<bf6cdb7c>] (extack_doit [compat]) from [<c05c309c>] (genl_rcv_msg+0x274/0x30c)
[<c05c309c>] (genl_rcv_msg) from [<c05c25d8>] (netlink_rcv_skb+0x58/0xac)
[<c05c25d8>] (netlink_rcv_skb) from [<c05c2e14>] (genl_rcv+0x20/0x34)
[<c05c2e14>] (genl_rcv) from [<c05c1f90>] (netlink_unicast+0x11c/0x204)
[<c05c1f90>] (netlink_unicast) from [<c05c2420>] (netlink_sendmsg+0x30c/0x370)
[<c05c2420>] (netlink_sendmsg) from [<c05886d0>] (sock_sendmsg+0x70/0x84)
[<c05886d0>] (sock_sendmsg) from [<c0589f4c>] (___sys_sendmsg.part.3+0x188/0x228)
[<c0589f4c>] (___sys_sendmsg.part.3) from [<c058add4>] (__sys_sendmsg+0x4c/0x70)
[<c058add4>] (__sys_sendmsg) from [<c0208c80>] (ret_fast_syscall+0x0/0x44)
Code: e2822c02 e2822001 e5832004 f590f000 (e1902f9f)
---[ end trace bbd717600f8f884d ]---
Signed-off-by: Miaoqing Pan <miaoqing@codeaurora.org>
Link: https://lore.kernel.org/r/1569485810-761-1-git-send-email-miaoqing@codeaurora.org
[trim useless data from commit message]
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
In a few places we don't properly initialize on-stack chandefs,
resulting in EDMG data to be non-zero, which broke things.
Additionally, in a few places we rely on the driver to init the
data completely, but perhaps we shouldn't as non-EDMG drivers
may not initialize the EDMG data, also initialize it there.
Cc: stable@vger.kernel.org
Fixes: 2a38075cd0 ("nl80211: Add support for EDMG channels")
Reported-by: Dmitry Osipenko <digetx@gmail.com>
Tested-by: Dmitry Osipenko <digetx@gmail.com>
Link: https://lore.kernel.org/r/1569239475-I2dcce394ecf873376c386a78f31c2ec8b538fa25@changeid
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
The code copying the data assumes that the SSID element is
before the MBSSID element, but since the data is untrusted
from the AP, this cannot be guaranteed.
Validate that this is indeed the case and ignore the MBSSID
otherwise, to avoid having to deal with both cases for the
copy of data that should be between them.
Cc: stable@vger.kernel.org
Fixes: 0b8fb8235b ("cfg80211: Parsing of Multiple BSSID information in scanning")
Link: https://lore.kernel.org/r/1569009255-I1673911f5eae02964e21bdc11b2bf58e5e207e59@changeid
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
We currently don't validate the beacon head, i.e. the header,
fixed part and elements that are to go in front of the TIM
element. This means that the variable elements there can be
malformed, e.g. have a length exceeding the buffer size, but
most downstream code from this assumes that this has already
been checked.
Add the necessary checks to the netlink policy.
Cc: stable@vger.kernel.org
Fixes: ed1b6cc7f8 ("cfg80211/nl80211: add beacon settings")
Link: https://lore.kernel.org/r/1569009255-I7ac7fbe9436e9d8733439eab8acbbd35e55c74ef@changeid
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
The speed divisor is used in a context expecting an s64, but it is
evaluated using 32-bit arithmetic.
To avoid that happening, instead of multiplying by 1,000,000 in the
first place, simplify the fraction and do a standard 32 bit division
instead.
Fixes: f04b514c0c ("taprio: Set default link speed to 10 Mbps in taprio_set_picos_per_byte")
Reported-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Suggested-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: Vladimir Oltean <olteanv@gmail.com>
Acked-by: Vinicius Costa Gomes <vinicius.gomes@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In sja1105_static_config_upload, in two cases memory is leaked: when
static_config_buf_prepare_for_upload fails and when sja1105_inhibit_tx
fails. In both cases config_buf should be released.
Fixes: 8aa9ebccae ("net: dsa: Introduce driver for NXP SJA1105 5-port L2 switch")
Fixes: 1a4c69406c ("net: dsa: sja1105: Prevent PHY jabbering during switch reset")
Signed-off-by: Navid Emamdoost <navid.emamdoost@gmail.com>
Signed-off-by: Vladimir Oltean <olteanv@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Sometimes the PTP synchronization on the switch 'jumps':
ptp4l[11241.155]: rms 8 max 16 freq -21732 +/- 11 delay 742 +/- 0
ptp4l[11243.157]: rms 7 max 17 freq -21731 +/- 10 delay 744 +/- 0
ptp4l[11245.160]: rms 33592410 max 134217731 freq +192422 +/- 8530253 delay 743 +/- 0
ptp4l[11247.163]: rms 811631 max 964131 freq +10326 +/- 557785 delay 743 +/- 0
ptp4l[11249.166]: rms 261936 max 533876 freq -304323 +/- 126371 delay 744 +/- 0
ptp4l[11251.169]: rms 48700 max 57740 freq -20218 +/- 30532 delay 744 +/- 0
ptp4l[11253.171]: rms 14570 max 30163 freq -5568 +/- 7563 delay 742 +/- 0
ptp4l[11255.174]: rms 2914 max 3440 freq -22001 +/- 1667 delay 744 +/- 1
ptp4l[11257.177]: rms 811 max 1710 freq -22653 +/- 451 delay 744 +/- 1
ptp4l[11259.180]: rms 177 max 218 freq -21695 +/- 89 delay 741 +/- 0
ptp4l[11261.182]: rms 45 max 92 freq -21677 +/- 32 delay 742 +/- 0
ptp4l[11263.186]: rms 14 max 32 freq -21733 +/- 11 delay 742 +/- 0
ptp4l[11265.188]: rms 9 max 14 freq -21725 +/- 12 delay 742 +/- 0
ptp4l[11267.191]: rms 9 max 16 freq -21727 +/- 13 delay 742 +/- 0
ptp4l[11269.194]: rms 6 max 15 freq -21726 +/- 9 delay 743 +/- 0
ptp4l[11271.197]: rms 8 max 15 freq -21728 +/- 11 delay 743 +/- 0
ptp4l[11273.200]: rms 6 max 12 freq -21727 +/- 8 delay 743 +/- 0
ptp4l[11275.202]: rms 9 max 17 freq -21720 +/- 11 delay 742 +/- 0
ptp4l[11277.205]: rms 9 max 18 freq -21725 +/- 12 delay 742 +/- 0
Background: the switch only offers partial RX timestamps (24 bits) and
it is up to the driver to read the PTP clock to fill those timestamps up
to 64 bits. But the PTP clock readout needs to happen quickly enough (in
0.135 seconds, in fact), otherwise the PTP clock will wrap around 24
bits, condition which cannot be detected.
Looking at the 'max 134217731' value on output line 3, one can see that
in hex it is 0x8000003. Because the PTP clock resolution is 8 ns,
that means 0x1000000 in ticks, which is exactly 2^24. So indeed this is
a PTP clock wraparound, but the reason might be surprising.
What is going on is that sja1105_tstamp_reconstruct(priv, now, ts)
expects a "now" time that is later than the "ts" was snapshotted at.
This, of course, is obvious: we read the PTP time _after_ the partial RX
timestamp was received. However, the workqueue is processing frames from
a skb queue and reuses the same PTP time, read once at the beginning.
Normally the skb queue only contains one frame and all goes well. But
when the skb queue contains two frames, the second frame that gets
dequeued might have been partially timestamped by the RX MAC _after_ we
had read our PTP time initially.
The code was originally like that due to concerns that SPI access for
PTP time readout is a slow process, and we are time-constrained anyway
(aka: premature optimization). But some timing analysis reveals that the
time spent until the RX timestamp is completely reconstructed is 1 order
of magnitude lower than the 0.135 s deadline even under worst-case
conditions. So we can afford to read the PTP time for each frame in the
RX timestamping queue, which of course ensures that the full PTP time is
in the partial timestamp's future.
Fixes: f3097be21b ("net: dsa: sja1105: Add a state machine for RX timestamping")
Signed-off-by: Vladimir Oltean <olteanv@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Stefan Schmidt says:
====================
pull-request: ieee802154 for net 2019-09-28
An update from ieee802154 for your *net* tree.
Three driver fixes. Navid Emamdoost fixed a memory leak on an error
path in the ca8210 driver, Johan Hovold fixed a use-after-free found
by syzbot in the atusb driver and Christophe JAILLET makes sure
__skb_put_data is used instead of memcpy in the mcr20a driver
I switched from branches to tags here to be pulled from. So far not
annotated and not signed. Once I fixed my scripts it should contain
this messages as annotations. If you want it signed as well just tell
me. If there are any problems let me know.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
The "reuse->sock[]" array is shared by multiple sockets. The going away
sk must unpublish itself from "reuse->sock[]" before making call_rcu()
call. However, this unpublish-action is currently done after a grace
period and it may cause use-after-free.
The fix is to move reuseport_detach_sock() to sk_destruct().
Due to the above reason, any socket with sk_reuseport_cb has
to go through the rcu grace period before freeing it.
It is a rather old bug (~3 yrs). The Fixes tag is not necessary
the right commit but it is the one that introduced the SOCK_RCU_FREE
logic and this fix is depending on it.
Fixes: a4298e4522 ("net: add SOCK_RCU_FREE socket flag")
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Suggested-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
erspan driver calls ether_setup(), after commit 61e84623ac
("net: centralize net_device min/max MTU checking"), the range
of mtu is [min_mtu, max_mtu], which is [68, 1500] by default.
It causes the dev mtu of the erspan device to not be greater
than 1500, this limit value is not correct for ipgre tap device.
Tested:
Before patch:
# ip link set erspan0 mtu 1600
Error: mtu greater than device maximum.
After patch:
# ip link set erspan0 mtu 1600
# ip -d link show erspan0
21: erspan0@NONE: <BROADCAST,MULTICAST> mtu 1600 qdisc noop state DOWN
mode DEFAULT group default qlen 1000
link/ether 00:00:00:00:00:00 brd ff:ff:ff:ff:ff:ff promiscuity 0 minmtu 68 maxmtu 0
Fixes: 61e84623ac ("net: centralize net_device min/max MTU checking")
Signed-off-by: Haishuang Yan <yanhaishuang@cmss.chinamobile.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The QCA8K family supports up to 7 ports. So use the existing
QCA8K_NUM_PORTS define to allocate the switch structure and limit all
operations with the switch ports.
This was not an issue until commit 0394a63acf ("net: dsa: enable and
disable all ports") disabled all unused ports. Since the unused ports 7-11
are outside of the correct register range on this switch some registers
were rewritten with invalid content.
Fixes: 6b93fb4648 ("net-next: dsa: add new driver for qca8xxx family")
Fixes: a0c02161ec ("net: dsa: variable number of ports")
Fixes: 0394a63acf ("net: dsa: enable and disable all ports")
Signed-off-by: Michal Vokáč <michal.vokac@ysoft.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
Pull networking fixes from David Miller:
1) Sanity check URB networking device parameters to avoid divide by
zero, from Oliver Neukum.
2) Disable global multicast filter in NCSI, otherwise LLDP and IPV6
don't work properly. Longer term this needs a better fix tho. From
Vijay Khemka.
3) Small fixes to selftests (use ping when ping6 is not present, etc.)
from David Ahern.
4) Bring back rt_uses_gateway member of struct rtable, it's semantics
were not well understood and trying to remove it broke things. From
David Ahern.
5) Move usbnet snaity checking, ignore endpoints with invalid
wMaxPacketSize. From Bjørn Mork.
6) Missing Kconfig deps for sja1105 driver, from Mao Wenan.
7) Various small fixes to the mlx5 DR steering code, from Alaa Hleihel,
Alex Vesker, and Yevgeny Kliteynik
8) Missing CAP_NET_RAW checks in various places, from Ori Nimron.
9) Fix crash when removing sch_cbs entry while offloading is enabled,
from Vinicius Costa Gomes.
10) Signedness bug fixes, generally in looking at the result given by
of_get_phy_mode() and friends. From Dan Crapenter.
11) Disable preemption around BPF_PROG_RUN() calls, from Eric Dumazet.
12) Don't create VRF ipv6 rules if ipv6 is disabled, from David Ahern.
13) Fix quantization code in tcp_bbr, from Kevin Yang.
* git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (127 commits)
net: tap: clean up an indentation issue
nfp: abm: fix memory leak in nfp_abm_u32_knode_replace
tcp: better handle TCP_USER_TIMEOUT in SYN_SENT state
sk_buff: drop all skb extensions on free and skb scrubbing
tcp_bbr: fix quantization code to not raise cwnd if not probing bandwidth
mlxsw: spectrum_flower: Fail in case user specifies multiple mirror actions
Documentation: Clarify trap's description
mlxsw: spectrum: Clear VLAN filters during port initialization
net: ena: clean up indentation issue
NFC: st95hf: clean up indentation issue
net: phy: micrel: add Asym Pause workaround for KSZ9021
net: socionext: ave: Avoid using netdev_err() before calling register_netdev()
ptp: correctly disable flags on old ioctls
lib: dimlib: fix help text typos
net: dsa: microchip: Always set regmap stride to 1
nfp: flower: fix memory leak in nfp_flower_spawn_vnic_reprs
nfp: flower: prevent memory leak in nfp_flower_spawn_phy_reprs
net/sched: Set default of CONFIG_NET_TC_SKB_EXT to N
vrf: Do not attempt to create IPv6 mcast rule if IPv6 is disabled
net: sched: sch_sfb: don't call qdisc_put() while holding tree lock
...
Merge hugepage allocation updates from David Rientjes:
"We (mostly Linus, Andrea, and myself) have been discussing offlist how
to implement a sane default allocation strategy for hugepages on NUMA
platforms.
With these reverts in place, the page allocator will happily allocate
a remote hugepage immediately rather than try to make a local hugepage
available. This incurs a substantial performance degradation when
memory compaction would have otherwise made a local hugepage
available.
This series reverts those reverts and attempts to propose a more sane
default allocation strategy specifically for hugepages. Andrea
acknowledges this is likely to fix the swap storms that he originally
reported that resulted in the patches that removed __GFP_THISNODE from
hugepage allocations.
The immediate goal is to return 5.3 to the behavior the kernel has
implemented over the past several years so that remote hugepages are
not immediately allocated when local hugepages could have been made
available because the increased access latency is untenable.
The next goal is to introduce a sane default allocation strategy for
hugepages allocations in general regardless of the configuration of
the system so that we prevent thrashing of local memory when
compaction is unlikely to succeed and can prefer remote hugepages over
remote native pages when the local node is low on memory."
Note on timing: this reverts the hugepage VM behavior changes that got
introduced fairly late in the 5.3 cycle, and that fixed a huge
performance regression for certain loads that had been around since
4.18.
Andrea had this note:
"The regression of 4.18 was that it was taking hours to start a VM
where 3.10 was only taking a few seconds, I reported all the details
on lkml when it was finally tracked down in August 2018.
https://lore.kernel.org/linux-mm/20180820032640.9896-2-aarcange@redhat.com/
__GFP_THISNODE in MADV_HUGEPAGE made the above enterprise vfio
workload degrade like in the "current upstream" above. And it still
would have been that bad as above until 5.3-rc5"
where the bad behavior ends up happening as you fill up a local node,
and without that change, you'd get into the nasty swap storm behavior
due to compaction working overtime to make room for more memory on the
nodes.
As a result 5.3 got the two performance fix reverts in rc5.
However, David Rientjes then noted that those performance fixes in turn
regressed performance for other loads - although not quite to the same
degree. He suggested reverting the reverts and instead replacing them
with two small changes to how hugepage allocations are done (patch
descriptions rephrased by me):
- "avoid expensive reclaim when compaction may not succeed": just admit
that the allocation failed when you're trying to allocate a huge-page
and compaction wasn't successful.
- "allow hugepage fallback to remote nodes when madvised": when that
node-local huge-page allocation failed, retry without forcing the
local node.
but by then I judged it too late to replace the fixes for a 5.3 release.
So 5.3 was released with behavior that harked back to the pre-4.18 logic.
But now we're in the merge window for 5.4, and we can see if this
alternate model fixes not just the horrendous swap storm behavior, but
also restores the performance regression that the late reverts caused.
Fingers crossed.
* emailed patches from David Rientjes <rientjes@google.com>:
mm, page_alloc: allow hugepage fallback to remote nodes when madvised
mm, page_alloc: avoid expensive reclaim when compaction may not succeed
Revert "Revert "Revert "mm, thp: consolidate THP gfp handling into alloc_hugepage_direct_gfpmask""
Revert "Revert "mm, thp: restore node-local hugepage allocations""
For systems configured to always try hard to allocate transparent
hugepages (thp defrag setting of "always") or for memory that has been
explicitly madvised to MADV_HUGEPAGE, it is often better to fallback to
remote memory to allocate the hugepage if the local allocation fails
first.
The point is to allow the initial call to __alloc_pages_node() to attempt
to defragment local memory to make a hugepage available, if possible,
rather than immediately fallback to remote memory. Local hugepages will
always have a better access latency than remote (huge)pages, so an attempt
to make a hugepage available locally is always preferred.
If memory compaction cannot be successful locally, however, it is likely
better to fallback to remote memory. This could take on two forms: either
allow immediate fallback to remote memory or do per-zone watermark checks.
It would be possible to fallback only when per-zone watermarks fail for
order-0 memory, since that would require local reclaim for all subsequent
faults so remote huge allocation is likely better than thrashing the local
zone for large workloads.
In this case, it is assumed that because the system is configured to try
hard to allocate hugepages or the vma is advised to explicitly want to try
hard for hugepages that remote allocation is better when local allocation
and memory compaction have both failed.
Signed-off-by: David Rientjes <rientjes@google.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Stefan Priebe - Profihost AG <s.priebe@profihost.ag>
Cc: "Kirill A. Shutemov" <kirill@shutemov.name>
Cc: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Memory compaction has a couple significant drawbacks as the allocation
order increases, specifically:
- isolate_freepages() is responsible for finding free pages to use as
migration targets and is implemented as a linear scan of memory
starting at the end of a zone,
- failing order-0 watermark checks in memory compaction does not account
for how far below the watermarks the zone actually is: to enable
migration, there must be *some* free memory available. Per the above,
watermarks are not always suffficient if isolate_freepages() cannot
find the free memory but it could require hundreds of MBs of reclaim to
even reach this threshold (read: potentially very expensive reclaim with
no indication compaction can be successful), and
- if compaction at this order has failed recently so that it does not even
run as a result of deferred compaction, looping through reclaim can often
be pointless.
For hugepage allocations, these are quite substantial drawbacks because
these are very high order allocations (order-9 on x86) and falling back to
doing reclaim can potentially be *very* expensive without any indication
that compaction would even be successful.
Reclaim itself is unlikely to free entire pageblocks and certainly no
reliance should be put on it to do so in isolation (recall lumpy reclaim).
This means we should avoid reclaim and simply fail hugepage allocation if
compaction is deferred.
It is also not helpful to thrash a zone by doing excessive reclaim if
compaction may not be able to access that memory. If order-0 watermarks
fail and the allocation order is sufficiently large, it is likely better
to fail the allocation rather than thrashing the zone.
Signed-off-by: David Rientjes <rientjes@google.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Stefan Priebe - Profihost AG <s.priebe@profihost.ag>
Cc: "Kirill A. Shutemov" <kirill@shutemov.name>
Cc: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This reverts commit 92717d429b.
Since commit a8282608c8 ("Revert "mm, thp: restore node-local hugepage
allocations"") is reverted in this series, it is better to restore the
previous 5.2 behavior between the thp allocation and the page allocator
rather than to attempt any consolidation or cleanup for a policy that is
now reverted. It's less risky during an rc cycle and subsequent patches
in this series further modify the same policy that the pre-5.3 behavior
implements.
Consolidation and cleanup can be done subsequent to a sane default page
allocation strategy, so this patch reverts a cleanup done on a strategy
that is now reverted and thus is the least risky option.
Signed-off-by: David Rientjes <rientjes@google.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Stefan Priebe - Profihost AG <s.priebe@profihost.ag>
Cc: "Kirill A. Shutemov" <kirill@shutemov.name>
Cc: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This reverts commit a8282608c8.
The commit references the original intended semantic for MADV_HUGEPAGE
which has subsequently taken on three unique purposes:
- enables or disables thp for a range of memory depending on the system's
config (is thp "enabled" set to "always" or "madvise"),
- determines the synchronous compaction behavior for thp allocations at
fault (is thp "defrag" set to "always", "defer+madvise", or "madvise"),
and
- reverts a previous MADV_NOHUGEPAGE (there is no madvise mode to only
clear previous hugepage advice).
These are the three purposes that currently exist in 5.2 and over the
past several years that userspace has been written around. Adding a
NUMA locality preference adds a fourth dimension to an already conflated
advice mode.
Based on the semantic that MADV_HUGEPAGE has provided over the past
several years, there exist workloads that use the tunable based on these
principles: specifically that the allocation should attempt to
defragment a local node before falling back. It is agreed that remote
hugepages typically (but not always) have a better access latency than
remote native pages, although on Naples this is at parity for
intersocket.
The revert commit that this patch reverts allows hugepage allocation to
immediately allocate remotely when local memory is fragmented. This is
contrary to the semantic of MADV_HUGEPAGE over the past several years:
that is, memory compaction should be attempted locally before falling
back.
The performance degradation of remote hugepages over local hugepages on
Rome, for example, is 53.5% increased access latency. For this reason,
the goal is to revert back to the 5.2 and previous behavior that would
attempt local defragmentation before falling back. With the patch that
is reverted by this patch, we see performance degradations at the tail
because the allocator happily allocates the remote hugepage rather than
even attempting to make a local hugepage available.
zone_reclaim_mode is not a solution to this problem since it does not
only impact hugepage allocations but rather changes the memory
allocation strategy for *all* page allocations.
Signed-off-by: David Rientjes <rientjes@google.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Stefan Priebe - Profihost AG <s.priebe@profihost.ag>
Cc: "Kirill A. Shutemov" <kirill@shutemov.name>
Cc: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
An assortment of fixes that were either missed by me, or didn't arrive quite in
time for the first v5.4 pull.
Most notable is a fix for an issue with tlbie (broadcast TLB invalidation) on
Power9, when using the Radix MMU. The tlbie can race with an mtpid (move to PID
register, essentially MMU context switch) on another thread of the core, which
can cause stores to continue to go to a page after it's unmapped.
A fix in our KVM code to add a missing barrier, the lack of which has been
observed to cause missed IPIs and subsequently stuck CPUs in the host.
A change to the way we initialise PCR (Processor Compatibility Register) to make
it forward compatible with future CPUs.
On some older PowerVM systems our H_BLOCK_REMOVE support could oops, fix it to
detect such systems and fallback to the old invalidation method.
A fix for an oops seen on some machines when using KASAN on 32-bit.
A handful of other minor fixes, and two new selftests.
Thanks to:
Alistair Popple, Aneesh Kumar K.V, Christophe Leroy, Gustavo Romero, Joel
Stanley, Jordan Niethe, Laurent Dufour, Michael Roth, Oliver O'Halloran.
-----BEGIN PGP SIGNATURE-----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=dF42
-----END PGP SIGNATURE-----
Merge tag 'powerpc-5.4-2' of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux
Pull powerpc fixes from Michael Ellerman:
"An assortment of fixes that were either missed by me, or didn't arrive
quite in time for the first v5.4 pull.
- Most notable is a fix for an issue with tlbie (broadcast TLB
invalidation) on Power9, when using the Radix MMU. The tlbie can
race with an mtpid (move to PID register, essentially MMU context
switch) on another thread of the core, which can cause stores to
continue to go to a page after it's unmapped.
- A fix in our KVM code to add a missing barrier, the lack of which
has been observed to cause missed IPIs and subsequently stuck CPUs
in the host.
- A change to the way we initialise PCR (Processor Compatibility
Register) to make it forward compatible with future CPUs.
- On some older PowerVM systems our H_BLOCK_REMOVE support could
oops, fix it to detect such systems and fallback to the old
invalidation method.
- A fix for an oops seen on some machines when using KASAN on 32-bit.
- A handful of other minor fixes, and two new selftests.
Thanks to: Alistair Popple, Aneesh Kumar K.V, Christophe Leroy,
Gustavo Romero, Joel Stanley, Jordan Niethe, Laurent Dufour, Michael
Roth, Oliver O'Halloran"
* tag 'powerpc-5.4-2' of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux:
powerpc/eeh: Fix eeh eeh_debugfs_break_device() with SRIOV devices
powerpc/nvdimm: use H_SCM_QUERY hcall on H_OVERLAP error
powerpc/nvdimm: Use HCALL error as the return value
selftests/powerpc: Add test case for tlbie vs mtpidr ordering issue
powerpc/mm: Fixup tlbie vs mtpidr/mtlpidr ordering issue on POWER9
powerpc/book3s64/radix: Rename CPU_FTR_P9_TLBIE_BUG feature flag
powerpc/book3s64/mm: Don't do tlbie fixup for some hardware revisions
powerpc/pseries: Call H_BLOCK_REMOVE when supported
powerpc/pseries: Read TLB Block Invalidate Characteristics
KVM: PPC: Book3S HV: use smp_mb() when setting/clearing host_ipi flag
powerpc/mm: Fix an Oops in kasan_mmu_init()
powerpc/mm: Add a helper to select PAGE_KERNEL_RO or PAGE_READONLY
powerpc/64s: Set reserved PCR bits
powerpc: Fix definition of PCR bits to work with old binutils
powerpc/book3s64/radix: Remove WARN_ON in destroy_context()
powerpc/tm: Add tm-poison test
Pull x86 fix from Ingo Molnar:
"A kexec fix for the case when GCC_PLUGIN_STACKLEAK=y is enabled"
* 'x86-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
x86/purgatory: Disable the stackleak GCC plugin for the purgatory
Pull scheduler fixes from Ingo Molnar:
- Apply a number of membarrier related fixes and cleanups, which fixes
a use-after-free race in the membarrier code
- Introduce proper RCU protection for tasks on the runqueue - to get
rid of the subtle task_rcu_dereference() interface that was easy to
get wrong
- Misc fixes, but also an EAS speedup
* 'sched-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
sched/fair: Avoid redundant EAS calculation
sched/core: Remove double update_max_interval() call on CPU startup
sched/core: Fix preempt_schedule() interrupt return comment
sched/fair: Fix -Wunused-but-set-variable warnings
sched/core: Fix migration to invalid CPU in __set_cpus_allowed_ptr()
sched/membarrier: Return -ENOMEM to userspace on memory allocation failure
sched/membarrier: Skip IPIs when mm->mm_users == 1
selftests, sched/membarrier: Add multi-threaded test
sched/membarrier: Fix p->mm->membarrier_state racy load
sched/membarrier: Call sync_core only before usermode for same mm
sched/membarrier: Remove redundant check
sched/membarrier: Fix private expedited registration check
tasks, sched/core: RCUify the assignment of rq->curr
tasks, sched/core: With a grace period after finish_task_switch(), remove unnecessary code
tasks, sched/core: Ensure tasks are available for a grace period after leaving the runqueue
tasks: Add a count of task RCU users
sched/core: Convert vcpu_is_preempted() from macro to an inline function
sched/fair: Remove unused cfs_rq_clock_task() function
Pull kernel lockdown mode from James Morris:
"This is the latest iteration of the kernel lockdown patchset, from
Matthew Garrett, David Howells and others.
From the original description:
This patchset introduces an optional kernel lockdown feature,
intended to strengthen the boundary between UID 0 and the kernel.
When enabled, various pieces of kernel functionality are restricted.
Applications that rely on low-level access to either hardware or the
kernel may cease working as a result - therefore this should not be
enabled without appropriate evaluation beforehand.
The majority of mainstream distributions have been carrying variants
of this patchset for many years now, so there's value in providing a
doesn't meet every distribution requirement, but gets us much closer
to not requiring external patches.
There are two major changes since this was last proposed for mainline:
- Separating lockdown from EFI secure boot. Background discussion is
covered here: https://lwn.net/Articles/751061/
- Implementation as an LSM, with a default stackable lockdown LSM
module. This allows the lockdown feature to be policy-driven,
rather than encoding an implicit policy within the mechanism.
The new locked_down LSM hook is provided to allow LSMs to make a
policy decision around whether kernel functionality that would allow
tampering with or examining the runtime state of the kernel should be
permitted.
The included lockdown LSM provides an implementation with a simple
policy intended for general purpose use. This policy provides a coarse
level of granularity, controllable via the kernel command line:
lockdown={integrity|confidentiality}
Enable the kernel lockdown feature. If set to integrity, kernel features
that allow userland to modify the running kernel are disabled. If set to
confidentiality, kernel features that allow userland to extract
confidential information from the kernel are also disabled.
This may also be controlled via /sys/kernel/security/lockdown and
overriden by kernel configuration.
New or existing LSMs may implement finer-grained controls of the
lockdown features. Refer to the lockdown_reason documentation in
include/linux/security.h for details.
The lockdown feature has had signficant design feedback and review
across many subsystems. This code has been in linux-next for some
weeks, with a few fixes applied along the way.
Stephen Rothwell noted that commit 9d1f8be5cf ("bpf: Restrict bpf
when kernel lockdown is in confidentiality mode") is missing a
Signed-off-by from its author. Matthew responded that he is providing
this under category (c) of the DCO"
* 'next-lockdown' of git://git.kernel.org/pub/scm/linux/kernel/git/jmorris/linux-security: (31 commits)
kexec: Fix file verification on S390
security: constify some arrays in lockdown LSM
lockdown: Print current->comm in restriction messages
efi: Restrict efivar_ssdt_load when the kernel is locked down
tracefs: Restrict tracefs when the kernel is locked down
debugfs: Restrict debugfs when the kernel is locked down
kexec: Allow kexec_file() with appropriate IMA policy when locked down
lockdown: Lock down perf when in confidentiality mode
bpf: Restrict bpf when kernel lockdown is in confidentiality mode
lockdown: Lock down tracing and perf kprobes when in confidentiality mode
lockdown: Lock down /proc/kcore
x86/mmiotrace: Lock down the testmmiotrace module
lockdown: Lock down module params that specify hardware parameters (eg. ioport)
lockdown: Lock down TIOCSSERIAL
lockdown: Prohibit PCMCIA CIS storage when the kernel is locked down
acpi: Disable ACPI table override if the kernel is locked down
acpi: Ignore acpi_rsdp kernel param when the kernel has been locked down
ACPI: Limit access to custom_method when the kernel is locked down
x86/msr: Restrict MSR access when the kernel is locked down
x86: Lock down IO port access when the kernel is locked down
...
Tuong Lien