Selftests are reporting this failure in test_lwt_seg6local.sh:
+ ip netns exec ns2 ip -6 route add fb00::6 encap bpf in obj test_lwt_seg6local.o sec encap_srh dev veth2
Error fetching program/map!
Failed to parse eBPF program: Operation not permitted
The problem is __attribute__((always_inline)) alone is not enough to prevent
clang from inserting those functions in .text. In that case, .text is not
marked as relocateable.
See the output of objdump -h test_lwt_seg6local.o:
Idx Name Size VMA LMA File off Algn
0 .text 00003530 0000000000000000 0000000000000000 00000040 2**3
CONTENTS, ALLOC, LOAD, READONLY, CODE
This causes the iproute bpf loader to fail in bpf_fetch_prog_sec:
bpf_has_call_data returns true but bpf_fetch_prog_relo fails as there's no
relocateable .text section in the file.
To fix this, convert to 'static __always_inline'.
v2: Use 'static __always_inline' instead of 'static inline
__attribute__((always_inline))'
Fixes: c99a84eac0 ("selftests/bpf: test for seg6local End.BPF action")
Signed-off-by: Jiri Benc <jbenc@redhat.com>
Acked-by: Yonghong Song <yhs@fb.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
There are currently no tests for ALU64 shift operations when the shift
amount is 0. This adds 6 new tests to make sure they are equivalent
to a no-op. The x32 JIT had such bugs that could have been caught by
these tests.
Cc: Xi Wang <xi.wang@gmail.com>
Signed-off-by: Luke Nelson <luke.r.nels@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
The current x32 BPF JIT does not correctly compile shift operations when
the immediate shift amount is 0. The expected behavior is for this to
be a no-op.
The following program demonstrates the bug. The expexceted result is 1,
but the current JITed code returns 2.
r0 = 1
r1 = 1
r1 <<= 0
if r1 == 1 goto end
r0 = 2
end:
exit
This patch simplifies the code and fixes the bug.
Fixes: 03f5781be2 ("bpf, x86_32: add eBPF JIT compiler for ia32")
Co-developed-by: Xi Wang <xi.wang@gmail.com>
Signed-off-by: Xi Wang <xi.wang@gmail.com>
Signed-off-by: Luke Nelson <luke.r.nels@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
The current x32 BPF JIT for shift operations is not correct when the
shift amount in a register is 0. The expected behavior is a no-op, whereas
the current implementation changes bits in the destination register.
The following example demonstrates the bug. The expected result of this
program is 1, but the current JITed code returns 2.
r0 = 1
r1 = 1
r2 = 0
r1 <<= r2
if r1 == 1 goto end
r0 = 2
end:
exit
The bug is caused by an incorrect assumption by the JIT that a shift by
32 clear the register. On x32 however, shifts use the lower 5 bits of
the source, making a shift by 32 equivalent to a shift by 0.
This patch fixes the bug using double-precision shifts, which also
simplifies the code.
Fixes: 03f5781be2 ("bpf, x86_32: add eBPF JIT compiler for ia32")
Co-developed-by: Xi Wang <xi.wang@gmail.com>
Signed-off-by: Xi Wang <xi.wang@gmail.com>
Signed-off-by: Luke Nelson <luke.r.nels@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Merge commit 1c8c5a9d38 ("Merge
git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next") undid the
fix from commit 36f9814a49 ("bpf: fix uapi hole for 32 bit compat
applications") by taking the gpl_compatible 1-bit field definition from
commit b85fab0e67 ("bpf: Add gpl_compatible flag to struct
bpf_prog_info") as is. That breaks architectures with 16-bit alignment
like m68k. Add 31-bit pad after gpl_compatible to restore alignment of
following fields.
Thanks to Dmitry V. Levin his analysis of this bug history.
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Acked-by: Song Liu <songliubraving@fb.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Daniel Borkmann <daniel@iogearbox.net>
Cc: Geert Uytterhoeven <geert@linux-m68k.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Yauheni reported the following code do not work correctly on BE arches:
ALU_ARSH_X:
DST = (u64) (u32) ((*(s32 *) &DST) >> SRC);
CONT;
ALU_ARSH_K:
DST = (u64) (u32) ((*(s32 *) &DST) >> IMM);
CONT;
and are causing failure of test_verifier test 'arsh32 on imm 2' on BE
arches.
The code is taking address and interpreting memory directly, so is not
endianness neutral. We should instead perform standard C type casting on
the variable. A u64 to s32 conversion will drop the high 32-bit and reserve
the low 32-bit as signed integer, this is all we want.
Fixes: 2dc6b100f9 ("bpf: interpreter support BPF_ALU | BPF_ARSH")
Reported-by: Yauheni Kaliuta <yauheni.kaliuta@redhat.com>
Reviewed-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Quentin Monnet <quentin.monnet@netronome.com>
Signed-off-by: Jiong Wang <jiong.wang@netronome.com>
Acked-by: Song Liu <songliubraving@fb.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
cgroup code tries to use argv[0] as the cgroup path,
but if it fails uses argv[1] to report errors.
Fixes: 5ccda64d38 ("bpftool: implement cgroup bpf operations")
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Quentin Monnet <quentin.monnet@netronome.com>
Acked-by: Roman Gushchin <guro@fb.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Clang warns:
In file included from net/xdp/xsk_queue.c:10:
net/xdp/xsk_queue.h:292:2: warning: expression result unused
[-Wunused-value]
WRITE_ONCE(q->ring->producer, q->prod_tail);
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
include/linux/compiler.h:284:6: note: expanded from macro 'WRITE_ONCE'
__u.__val; \
~~~ ^~~~~
1 warning generated.
The q->prod_tail assignment has a comma at the end, not a semi-colon.
Fix that so clang no longer warns and everything works as expected.
Fixes: c497176cb2 ("xsk: add Rx receive functions and poll support")
Link: https://github.com/ClangBuiltLinux/linux/issues/544
Signed-off-by: Nathan Chancellor <natechancellor@gmail.com>
Acked-by: Nick Desaulniers <ndesaulniers@google.com>
Acked-by: Jonathan Lemon <jonathan.lemon@gmail.com>
Acked-by: Björn Töpel <bjorn.topel@intel.com>
Acked-by: Song Liu <songliubraving@fb.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Jonathan Lemon has volunteered as an official AF_XDP reviewer. Thank
you, Jonathan!
Signed-off-by: Björn Töpel <bjorn.topel@intel.com>
Acked-by: Jonathan Lemon <jonathan.lemon@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
When we terminate xdp_redirect, it ends up with following message:
"Program on iface OUT changed, not removing"
This results in dummy prog still attached to OUT interface.
It is because signal handler checks if the programs are the same that
we had attached. But while fetching dummy_prog_id, current code uses
prog_fd instead of dummy_prog_fd. This patch passes the correct fd.
Fixes: 3b7a8ec2de ("samples/bpf: Check the prog id before exiting")
Signed-off-by: Prashant Bhole <prashantbhole.linux@gmail.com>
Acked-by: Toke Høiland-Jørgensen <toke@redhat.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Commit 1dc9285184 ("bpf: kernel side support for BTF Var and DataSec")
added invocations of btf_type_is_resolve_source_only before
btf_type_nosize_or_null which checks for the NULL pointer.
Swap the order of btf_type_nosize_or_null and
btf_type_is_resolve_source_only to make sure the do the NULL pointer
check first.
Fixes: 1dc9285184 ("bpf: kernel side support for BTF Var and DataSec")
Reported-by: syzbot <syzkaller@googlegroups.com>
Signed-off-by: Stanislav Fomichev <sdf@google.com>
Acked-by: Andrii Nakryiko <andriin@fb.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
The bpf_ipv6_fib_lookup function should return BPF_FIB_LKUP_RET_FWD_DISABLED
when forwarding is disabled for the input device. However instead of checking
if forwarding is enabled on the input device, it checked the global
net->ipv6.devconf_all->forwarding flag. Change it to behave as expected.
Fixes: 87f5fc7e48 ("bpf: Provide helper to do forwarding lookups in kernel FIB table")
Signed-off-by: Anton Protopopov <a.s.protopopov@gmail.com>
Acked-by: Toke Høiland-Jørgensen <toke@redhat.com>
Reviewed-by: David Ahern <dsahern@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Eric Dumazet says:
====================
tcp: make sack processing more robust
Jonathan Looney brought to our attention multiple problems
in TCP stack at the sender side.
SACK processing can be abused by malicious peers to either
cause overflows, or increase of memory usage.
First two patches fix the immediate problems.
Since the malicious peers abuse senders by advertizing a very
small MSS in their SYN or SYNACK packet, the last two
patches add a new sysctl so that admins can chose a higher
limit for MSS clamping.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
lapb_register calls lapb_create_cb, which initializes the control-
block's ref-count to one, and __lapb_insert_cb, which increments it when
adding the new block to the list of blocks.
lapb_unregister calls __lapb_remove_cb, which decrements the ref-count
when removing control-block from the list of blocks, and calls lapb_put
itself to decrement the ref-count before returning.
However, lapb_unregister also calls __lapb_devtostruct to look up the
right control-block for the given net_device, and __lapb_devtostruct
also bumps the ref-count, which means that when lapb_unregister returns
the ref-count is still 1 and the control-block is leaked.
Call lapb_put after __lapb_devtostruct to fix leak.
Reported-by: syzbot+afb980676c836b4a0afa@syzkaller.appspotmail.com
Signed-off-by: Jeremy Sowden <jeremy@azazel.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
Syzbot reported a memleak caused by grp members' deferredq list not
purged when the grp is be deleted.
The issue occurs when more(msg_grp_bc_seqno(hdr), m->bc_rcv_nxt) in
tipc_group_filter_msg() and the skb will stay in deferredq.
So fix it by calling __skb_queue_purge for each member's deferredq
in tipc_group_delete() when a tipc sk leaves the grp.
Fixes: b87a5ea31c ("tipc: guarantee group unicast doesn't bypass group broadcast")
Reported-by: syzbot+78fbe679c8ca8d264a8d@syzkaller.appspotmail.com
Signed-off-by: Xin Long <lucien.xin@gmail.com>
Acked-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
tcp_tx_skb_cache_key and tcp_rx_skb_cache_key must be available
even if CONFIG_SYSCTL is not set.
Fixes: 0b7d7f6b22 ("tcp: add tcp_tx_skb_cache sysctl")
Fixes: ede61ca474 ("tcp: add tcp_rx_skb_cache sysctl")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
gcc 8.2.0 may report these bogus warnings under some condition:
warning: ‘vnew’ may be used uninitialized in this function
warning: ‘hvs_new’ may be used uninitialized in this function
Actually, the 2 pointers are only initialized and used if the variable
"conn_from_host" is true. The code is not buggy here.
Signed-off-by: Dexuan Cui <decui@microsoft.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Number of Rx queues used for flow hashing returned by the driver is
incorrect and this bug prevents user to use the last Rx queue in
indirection table.
Let's say we have a NIC with 6 combined queues:
[root@sm-03 ~]# ethtool -l enp4s0f0
Channel parameters for enp4s0f0:
Pre-set maximums:
RX: 5
TX: 5
Other: 0
Combined: 6
Current hardware settings:
RX: 0
TX: 0
Other: 0
Combined: 6
Default indirection table maps all (6) queues equally but the driver
reports only 5 rings available.
[root@sm-03 ~]# ethtool -x enp4s0f0
RX flow hash indirection table for enp4s0f0 with 5 RX ring(s):
0: 0 1 2 3 4 5 0 1
8: 2 3 4 5 0 1 2 3
16: 4 5 0 1 2 3 4 5
24: 0 1 2 3 4 5 0 1
...
Now change indirection table somehow:
[root@sm-03 ~]# ethtool -X enp4s0f0 weight 1 1
[root@sm-03 ~]# ethtool -x enp4s0f0
RX flow hash indirection table for enp4s0f0 with 6 RX ring(s):
0: 0 0 0 0 0 0 0 0
...
64: 1 1 1 1 1 1 1 1
...
Now it is not possible to change mapping back to equal (default) state:
[root@sm-03 ~]# ethtool -X enp4s0f0 equal 6
Cannot set RX flow hash configuration: Invalid argument
Fixes: 594ad54a2c ("be2net: Add support for setting and getting rx flow hash options")
Reported-by: Tianhao <tizhao@redhat.com>
Signed-off-by: Ivan Vecera <ivecera@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When stack receives pkt: [802.1P vlan 0][802.1AD vlan 100][IPv4],
vlan_do_receive() returns false if it does not find vlan_dev. Later
__netif_receive_skb_core() fails to find packet type handler for
skb->protocol 801.1AD and drops the packet.
801.1P header with vlan id 0 should be handled as untagged packets.
This patch fixes it by checking if vlan_id is 0 and processes next vlan
header.
Signed-off-by: Govindarajulu Varadarajan <gvaradar@cisco.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
If mtu probing is enabled tcp_mtu_probing() could very well end up
with a too small MSS.
Use the new sysctl tcp_min_snd_mss to make sure MSS search
is performed in an acceptable range.
CVE-2019-11479 -- tcp mss hardcoded to 48
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: Jonathan Lemon <jonathan.lemon@gmail.com>
Cc: Jonathan Looney <jtl@netflix.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Tyler Hicks <tyhicks@canonical.com>
Cc: Bruce Curtis <brucec@netflix.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Some TCP peers announce a very small MSS option in their SYN and/or
SYN/ACK messages.
This forces the stack to send packets with a very high network/cpu
overhead.
Linux has enforced a minimal value of 48. Since this value includes
the size of TCP options, and that the options can consume up to 40
bytes, this means that each segment can include only 8 bytes of payload.
In some cases, it can be useful to increase the minimal value
to a saner value.
We still let the default to 48 (TCP_MIN_SND_MSS), for compatibility
reasons.
Note that TCP_MAXSEG socket option enforces a minimal value
of (TCP_MIN_MSS). David Miller increased this minimal value
in commit c39508d6f1 ("tcp: Make TCP_MAXSEG minimum more correct.")
from 64 to 88.
We might in the future merge TCP_MIN_SND_MSS and TCP_MIN_MSS.
CVE-2019-11479 -- tcp mss hardcoded to 48
Signed-off-by: Eric Dumazet <edumazet@google.com>
Suggested-by: Jonathan Looney <jtl@netflix.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Tyler Hicks <tyhicks@canonical.com>
Cc: Bruce Curtis <brucec@netflix.com>
Cc: Jonathan Lemon <jonathan.lemon@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Jonathan Looney reported that a malicious peer can force a sender
to fragment its retransmit queue into tiny skbs, inflating memory
usage and/or overflow 32bit counters.
TCP allows an application to queue up to sk_sndbuf bytes,
so we need to give some allowance for non malicious splitting
of retransmit queue.
A new SNMP counter is added to monitor how many times TCP
did not allow to split an skb if the allowance was exceeded.
Note that this counter might increase in the case applications
use SO_SNDBUF socket option to lower sk_sndbuf.
CVE-2019-11478 : tcp_fragment, prevent fragmenting a packet when the
socket is already using more than half the allowed space
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: Jonathan Looney <jtl@netflix.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Reviewed-by: Tyler Hicks <tyhicks@canonical.com>
Cc: Bruce Curtis <brucec@netflix.com>
Cc: Jonathan Lemon <jonathan.lemon@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Jonathan Looney reported that TCP can trigger the following crash
in tcp_shifted_skb() :
BUG_ON(tcp_skb_pcount(skb) < pcount);
This can happen if the remote peer has advertized the smallest
MSS that linux TCP accepts : 48
An skb can hold 17 fragments, and each fragment can hold 32KB
on x86, or 64KB on PowerPC.
This means that the 16bit witdh of TCP_SKB_CB(skb)->tcp_gso_segs
can overflow.
Note that tcp_sendmsg() builds skbs with less than 64KB
of payload, so this problem needs SACK to be enabled.
SACK blocks allow TCP to coalesce multiple skbs in the retransmit
queue, thus filling the 17 fragments to maximal capacity.
CVE-2019-11477 -- u16 overflow of TCP_SKB_CB(skb)->tcp_gso_segs
Fixes: 832d11c5cd ("tcp: Try to restore large SKBs while SACK processing")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: Jonathan Looney <jtl@netflix.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Reviewed-by: Tyler Hicks <tyhicks@canonical.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Bruce Curtis <brucec@netflix.com>
Cc: Jonathan Lemon <jonathan.lemon@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Alexei Starovoitov says:
====================
pull-request: bpf 2019-06-15
The following pull-request contains BPF updates for your *net* tree.
The main changes are:
1) fix stack layout of JITed x64 bpf code, from Alexei.
2) fix out of bounds memory access in bpf_sk_storage, from Arthur.
3) fix lpm trie walk, from Jonathan.
4) fix nested bpf_perf_event_output, from Matt.
5) and several other fixes.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
This reverts commit ef7bfa8472.
Russell King espressed some strong opposition to this
change, explaining that this is trying to make phylink
behave outside of how it has been designed.
Signed-off-by: David S. Miller <davem@davemloft.net>
BPF_PROG_TYPE_RAW_TRACEPOINTs can be executed nested on the same CPU, as
they do not increment bpf_prog_active while executing.
This enables three levels of nesting, to support
- a kprobe or raw tp or perf event,
- another one of the above that irq context happens to call, and
- another one in nmi context
(at most one of which may be a kprobe or perf event).
Fixes: 20b9d7ac48 ("bpf: avoid excessive stack usage for perf_sample_data")
Signed-off-by: Matt Mullins <mmullins@fb.com>
Acked-by: Andrii Nakryiko <andriin@fb.com>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
bpf_sk_storage maps use multiple spin locks to reduce contention.
The number of locks to use is determined by the number of possible CPUs.
With only 1 possible CPU, bucket_log == 0, and 2^0 = 1 locks are used.
When updating elements, the correct lock is determined with hash_ptr().
Calling hash_ptr() with 0 bits is undefined behavior, as it does:
x >> (64 - bits)
Using the value results in an out of bounds memory access.
In my case, this manifested itself as a page fault when raw_spin_lock_bh()
is called later, when running the self tests:
./tools/testing/selftests/bpf/test_verifier 773 775
[ 16.366342] BUG: unable to handle page fault for address: ffff8fe7a66f93f8
Force the minimum number of locks to two.
Signed-off-by: Arthur Fabre <afabre@cloudflare.com>
Fixes: 6ac99e8f23 ("bpf: Introduce bpf sk local storage")
Acked-by: Andrii Nakryiko <andriin@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Set the SOCK_DONE flag to match the TCP_CLOSING state when a peer has
shut down and there is nothing left to read.
This fixes the following bug:
1) Peer sends SHUTDOWN(RDWR).
2) Socket enters TCP_CLOSING but SOCK_DONE is not set.
3) read() returns -ENOTCONN until close() is called, then returns 0.
Signed-off-by: Stephen Barber <smbarber@chromium.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
We get this regression when using RTL8366RB as part of a bridge
with OpenWrt:
WARNING: CPU: 0 PID: 1347 at net/switchdev/switchdev.c:291
switchdev_port_attr_set_now+0x80/0xa4
lan0: Commit of attribute (id=7) failed.
(...)
realtek-smi switch lan0: failed to initialize vlan filtering on this port
This is because it is trying to disable VLAN filtering
on VLAN0, as we have forgot to add 1 to the port number
to get the right VLAN in rtl8366_vlan_filtering(): when
we initialize the VLAN we associate VLAN1 with port 0,
VLAN2 with port 1 etc, so we need to add 1 to the port
offset.
Fixes: d8652956cf ("net: dsa: realtek-smi: Add Realtek SMI driver")
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
The phy_state field of phylink should carry only valid information
especially when this can be passed to the .mac_config callback.
Update the an_enabled field with the autoneg state in the
phylink_phy_change function.
Fixes: 9525ae8395 ("phylink: add phylink infrastructure")
Signed-off-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Eric Dumazet says:
====================
tcp: add three static keys
Recent addition of per TCP socket rx/tx cache brought
regressions for some workloads, as reported by Feng Tang.
It seems better to make them opt-in, before we adopt better
heuristics.
The last patch adds high_order_alloc_disable sysctl
to ask TCP sendmsg() to exclusively use order-0 allocations,
as mm layer has specific optimizations.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
>From linux-3.7, (commit 5640f76858 "net: use a per task frag
allocator") TCP sendmsg() has preferred using order-3 allocations.
While it gives good results for most cases, we had reports
that heavy uses of TCP over loopback were hitting a spinlock
contention in page allocations/freeing.
This commits adds a sysctl so that admins can opt-in
for order-0 allocations. Hopefully mm layer might optimize
order-3 allocations in the future since it could give us
a nice boost (see 8 lines of following benchmark)
The following benchmark shows a win when more than 8 TCP_STREAM
threads are running (56 x86 cores server in my tests)
for thr in {1..30}
do
sysctl -wq net.core.high_order_alloc_disable=0
T0=`./super_netperf $thr -H 127.0.0.1 -l 15`
sysctl -wq net.core.high_order_alloc_disable=1
T1=`./super_netperf $thr -H 127.0.0.1 -l 15`
echo $thr:$T0:$T1
done
1: 49979: 37267
2: 98745: 76286
3: 141088: 110051
4: 177414: 144772
5: 197587: 173563
6: 215377: 208448
7: 241061: 234087
8: 267155: 263373
9: 295069: 297402
10: 312393: 335213
11: 340462: 368778
12: 371366: 403954
13: 412344: 443713
14: 426617: 473580
15: 474418: 507861
16: 503261: 538539
17: 522331: 563096
18: 532409: 567084
19: 550824: 605240
20: 525493: 641988
21: 564574: 665843
22: 567349: 690868
23: 583846: 710917
24: 588715: 736306
25: 603212: 763494
26: 604083: 792654
27: 602241: 796450
28: 604291: 797993
29: 611610: 833249
30: 577356: 841062
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Feng Tang reported a performance regression after introduction
of per TCP socket tx/rx caches, for TCP over loopback (netperf)
There is high chance the regression is caused by a change on
how well the 32 KB per-thread page (current->task_frag) can
be recycled, and lack of pcp caches for order-3 pages.
I could not reproduce the regression myself, cpus all being
spinning on the mm spinlocks for page allocs/freeing, regardless
of enabling or disabling the per tcp socket caches.
It seems best to disable the feature by default, and let
admins enabling it.
MM layer either needs to provide scalable order-3 pages
allocations, or could attempt a trylock on zone->lock if
the caller only attempts to get a high-order page and is
able to fallback to order-0 ones in case of pressure.
Tests run on a 56 cores host (112 hyper threads)
- 35.49% netperf [kernel.vmlinux] [k] queued_spin_lock_slowpath
- 35.49% queued_spin_lock_slowpath
- 18.18% get_page_from_freelist
- __alloc_pages_nodemask
- 18.18% alloc_pages_current
skb_page_frag_refill
sk_page_frag_refill
tcp_sendmsg_locked
tcp_sendmsg
inet_sendmsg
sock_sendmsg
__sys_sendto
__x64_sys_sendto
do_syscall_64
entry_SYSCALL_64_after_hwframe
__libc_send
+ 17.31% __free_pages_ok
+ 31.43% swapper [kernel.vmlinux] [k] intel_idle
+ 9.12% netperf [kernel.vmlinux] [k] copy_user_enhanced_fast_string
+ 6.53% netserver [kernel.vmlinux] [k] copy_user_enhanced_fast_string
+ 0.69% netserver [kernel.vmlinux] [k] queued_spin_lock_slowpath
+ 0.68% netperf [kernel.vmlinux] [k] skb_release_data
+ 0.52% netperf [kernel.vmlinux] [k] tcp_sendmsg_locked
0.46% netperf [kernel.vmlinux] [k] _raw_spin_lock_irqsave
Fixes: 472c2e07ee ("tcp: add one skb cache for tx")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: Feng Tang <feng.tang@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Instead of relying on rps_needed, it is safer to use a separate
static key, since we do not want to enable TCP rx_skb_cache
by default. This feature can cause huge increase of memory
usage on hosts with millions of sockets.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Convert proc_dointvec_minmax_bpf_stats() into a more generic
helper, since we are going to use jump labels more often.
Note that sysctl_bpf_stats_enabled is removed, since
it is no longer needed/used.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
For better consistency of synthetic NIC names, we set the probe mode to
PROBE_FORCE_SYNCHRONOUS. So the names can be aligned with the vmbus
channel offer sequence.
Fixes: af0a5646cb ("use the new async probing feature for the hyperv drivers")
Signed-off-by: Haiyang Zhang <haiyangz@microsoft.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Current flower mask creating code assumes that temporary mask that is used
when inserting new filter is stack allocated. To prevent race condition
with data patch synchronize_rcu() is called every time fl_create_new_mask()
replaces temporary stack allocated mask. As reported by Jiri, this
increases runtime of creating 20000 flower classifiers from 4 seconds to
163 seconds. However, this design is no longer necessary since temporary
mask was converted to be dynamically allocated by commit 2cddd20147
("net/sched: cls_flower: allocate mask dynamically in fl_change()").
Remove synchronize_rcu() calls from mask creation code. Instead, refactor
fl_change() to always deallocate temporary mask with rcu grace period.
Fixes: 195c234d15 ("net: sched: flower: handle concurrent mask insertion")
Reported-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: Vlad Buslov <vladbu@mellanox.com>
Tested-by: Jiri Pirko <jiri@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When building with CONFIG_NET_DSA_REALTEK_SMI and CONFIG_REALTEK_PHY
enabled as loadable modules, we see the following warning:
warning: same module names found:
drivers/net/phy/realtek.ko
drivers/net/dsa/realtek.ko
Rework so the driver name is realtek-smi instead of realtek.
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Anders Roxell <anders.roxell@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Based on comments from Xin, even after fixes for our recent syzbot
report of cookie memory leaks, its possible to get a resend of an INIT
chunk which would lead to us leaking cookie memory.
To ensure that we don't leak cookie memory, free any previously
allocated cookie first.
Change notes
v1->v2
update subsystem tag in subject (davem)
repeat kfree check for peer_random and peer_hmacs (xin)
v2->v3
net->sctp
also free peer_chunks
v3->v4
fix subject tags
v4->v5
remove cut line
Signed-off-by: Neil Horman <nhorman@tuxdriver.com>
Reported-by: syzbot+f7e9153b037eac9b1df8@syzkaller.appspotmail.com
CC: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
CC: Xin Long <lucien.xin@gmail.com>
CC: "David S. Miller" <davem@davemloft.net>
CC: netdev@vger.kernel.org
Acked-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
If some of the switch ports were not listed in the device tree, due to
being unused, the ksz_mib_read_work function ended up accessing a NULL
dp->slave pointer and causing an oops. Skip checking statistics for any
unused ports.
Fixes: 7c6ff470aa ("net: dsa: microchip: add MIB counter reading support")
Signed-off-by: Robert Hancock <hancock@sedsystems.ca>
Reviewed-by: Vivien Didelot <vivien.didelot@gmail.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Permit mux_id values up to 254 to be used in qmimux_register_device()
for compatibility with ip(8) and the rmnet driver.
Fixes: c6adf77953 ("net: usb: qmi_wwan: add qmap mux protocol support")
Cc: Daniele Palmas <dnlplm@gmail.com>
Signed-off-by: Reinhard Speyerer <rspmn@arcor.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
Switch qmimux_unregister_device() and qmi_wwan_disconnect() to
use unregister_netdevice_queue() and unregister_netdevice_many()
instead of unregister_netdevice(). This avoids RCU stalls which
have been observed on device disconnect in certain setups otherwise.
Fixes: c6adf77953 ("net: usb: qmi_wwan: add qmap mux protocol support")
Cc: Daniele Palmas <dnlplm@gmail.com>
Signed-off-by: Reinhard Speyerer <rspmn@arcor.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
The QMAP code in the qmi_wwan driver is based on the CodeAurora GobiNet
driver which does not process QMAP padding in the RX path correctly.
Add support for QMAP padding to qmimux_rx_fixup() according to the
description of the rmnet driver.
Fixes: c6adf77953 ("net: usb: qmi_wwan: add qmap mux protocol support")
Cc: Daniele Palmas <dnlplm@gmail.com>
Signed-off-by: Reinhard Speyerer <rspmn@arcor.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
Since commit 177366bf7c the %rbp stopped pointing to %rbp of the
previous stack frame. That broke frame pointer based stack unwinding.
This commit is a partial revert of it.
Note that the location of tail_call_cnt is fixed, since the verifier
enforces MAX_BPF_STACK stack size for programs with tail calls.
Fixes: 177366bf7c ("bpf: change x86 JITed program stack layout")
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
.ndo_xdp_xmit() assumes it is called under RCU. For example virtio_net
uses RCU to detect it has setup the resources for tx. The assumption
accidentally broke when introducing bulk queue in devmap.
Fixes: 5d053f9da4 ("bpf: devmap prepare xdp frames for bulking")
Reported-by: David Ahern <dsahern@gmail.com>
Signed-off-by: Toshiaki Makita <toshiaki.makita1@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
