The following pull-request contains BPF updates for your *net* tree.
The main changes are:
1) Fix a out-of-bounds access in __bpf_redirect_no_mac, from Willem.
2) Fix bpf_setsockopt to reset sock dst on SO_MARK changes, from Peter.
3) Fix map in map masking to prevent out-of-bounds access under
speculative execution, from Daniel.
4) Fix bpf_setsockopt's SO_MAX_PACING_RATE to support TCP internal
pacing, from Yuchung.
5) Fix json writer license in bpftool, from Thomas.
6) Fix AF_XDP to check if an actually queue exists during umem
setup, from Krzysztof.
7) Several fixes to BPF stackmap's build id handling. Another fix
for bpftool build to account for libbfd variations wrt linking
requirements, from Stanislav.
8) Fix BPF samples build with clang by working around missing asm
goto, from Yonghong.
9) Fix libbpf to retry program load on signal interrupt, from Lorenz.
10) Various minor compile warning fixes in BPF code, from Mathieu.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Willem de Bruijn [Wed, 16 Jan 2019 01:19:22 +0000 (20:19 -0500)]
bpf: in __bpf_redirect_no_mac pull mac only if present
Syzkaller was able to construct a packet of negative length by
redirecting from bpf_prog_test_run_skb with BPF_PROG_TYPE_LWT_XMIT:
BUG: KASAN: slab-out-of-bounds in memcpy include/linux/string.h:345 [inline]
BUG: KASAN: slab-out-of-bounds in skb_copy_from_linear_data include/linux/skbuff.h:3421 [inline]
BUG: KASAN: slab-out-of-bounds in __pskb_copy_fclone+0x2dd/0xeb0 net/core/skbuff.c:1395
Read of size 4294967282 at addr ffff8801d798009c by task syz-executor2/12942
The generated test constructs a packet with mac header, network
header, skb->data pointing to network header and skb->len 0.
Redirecting to a sit0 through __bpf_redirect_no_mac pulls the
mac length, even though skb->data already is at skb->network_header.
bpf_prog_test_run_skb has already pulled it as LWT_XMIT !is_l2.
Update the offset calculation to pull only if skb->data differs
from skb->network_header, which is not true in this case.
The test itself can be run only from commit 1cf1cae963c2 ("bpf:
introduce BPF_PROG_TEST_RUN command"), but the same type of packets
with skb at network header could already be built from lwt xmit hooks,
so this fix is more relevant to that commit.
Also set the mac header on redirect from LWT_XMIT, as even after this
change to __bpf_redirect_no_mac that field is expected to be set, but
is not yet in ip_finish_output2.
Fixes: 3a0af8fd61f9 ("bpf: BPF for lightweight tunnel infrastructure") Reported-by: syzbot <syzkaller@googlegroups.com> Signed-off-by: Willem de Bruijn <willemb@google.com> Acked-by: Martin KaFai Lau <kafai@fb.com> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Camelia Groza [Thu, 17 Jan 2019 12:22:36 +0000 (14:22 +0200)]
net: phy: phy driver features are mandatory
Since phy driver features became a link_mode bitmap, phy drivers that
don't have a list of features configured will cause the kernel to crash
when probed.
Prevent the phy driver from registering if the features field is missing.
Fixes: 719655a14971 ("net: phy: Replace phy driver features u32 with link_mode bitmap") Reported-by: Scott Wood <oss@buserror.net> Signed-off-by: Camelia Groza <camelia.groza@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Nathan Chancellor [Thu, 10 Jan 2019 05:41:08 +0000 (22:41 -0700)]
isdn: avm: Fix string plus integer warning from Clang
A recent commit in Clang expanded the -Wstring-plus-int warning, showing
some odd behavior in this file.
drivers/isdn/hardware/avm/b1.c:426:30: warning: adding 'int' to a string does not append to the string [-Wstring-plus-int]
cinfo->version[j] = "\0\0" + 1;
~~~~~~~^~~
drivers/isdn/hardware/avm/b1.c:426:30: note: use array indexing to silence this warning
cinfo->version[j] = "\0\0" + 1;
^
& [ ]
1 warning generated.
This is equivalent to just "\0". Nick pointed out that it is smarter to
use "" instead of "\0" because "" is used elsewhere in the kernel and
can be deduplicated at the linking stage.
Link: https://github.com/ClangBuiltLinux/linux/issues/309 Suggested-by: Nick Desaulniers <ndesaulniers@google.com> Signed-off-by: Nathan Chancellor <natechancellor@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Eli Britstein [Wed, 19 Dec 2018 05:36:51 +0000 (07:36 +0200)]
net/mlx5e: Fix cb_ident duplicate in indirect block register
Previously the identifier used for indirect block callback registry
and for block rule cb registry (when done via indirect blocks) was the
pointer to the tunnel netdev we were interested in receiving updates on.
This worked fine if a single PF existed that registered one callback for
the tunnel netdev of interest. However, if multiple PFs are in place then
the 2nd PF tries to register with the same tunnel netdev identifier. This
leads to EEXIST errors and/or incorrect cb deletions.
Prevent this conflict by using the rpriv pointer as the identifier for
netdev indirect block cb registry, allowing each PF to register a unique
callback per tunnel netdev. For block cb registry, the same PF may
register multiple cbs to the same block if using TC shared blocks.
Instead of the rpriv, use the pointer to the allocated indr_priv data as
the identifier here. This means that there can be a unique block callback
for each PF/tunnel netdev combo.
Fixes: f5bc2c5de101 ("net/mlx5e: Support TC indirect block notifications
for eswitch uplink reprs") Signed-off-by: Eli Britstein <elibr@mellanox.com> Reviewed-by: Oz Shlomo <ozsh@mellanox.com> Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Cong Wang [Tue, 4 Dec 2018 06:14:04 +0000 (22:14 -0800)]
net/mlx5e: Force CHECKSUM_UNNECESSARY for short ethernet frames
When an ethernet frame is padded to meet the minimum ethernet frame
size, the padding octets are not covered by the hardware checksum.
Fortunately the padding octets are usually zero's, which don't affect
checksum. However, we have a switch which pads non-zero octets, this
causes kernel hardware checksum fault repeatedly.
Prior to:
commit '88078d98d1bb ("net: pskb_trim_rcsum() and CHECKSUM_COMPLETE ...")'
skb checksum was forced to be CHECKSUM_NONE when padding is detected.
After it, we need to keep skb->csum updated, like what we do for RXFCS.
However, fixing up CHECKSUM_COMPLETE requires to verify and parse IP
headers, it is not worthy the effort as the packets are so small that
CHECKSUM_COMPLETE can't save anything.
Fixes: 88078d98d1bb ("net: pskb_trim_rcsum() and CHECKSUM_COMPLETE are friends"), Cc: Eric Dumazet <edumazet@google.com> Cc: Tariq Toukan <tariqt@mellanox.com> Cc: Nikola Ciprich <nikola.ciprich@linuxbox.cz> Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com> Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Thomas Gleixner [Thu, 17 Jan 2019 23:14:24 +0000 (00:14 +0100)]
tools: bpftool: Cleanup license mess
Precise and non-ambiguous license information is important. The recent
relicensing of the bpftools introduced a license conflict.
The files have now:
SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause
and
* This program is free software; you can redistribute it and/or
* modify it under the terms of the GNU General Public License
* as published by the Free Software Foundation; either version
* 2 of the License, or (at your option) any later version
Amazingly about 20 people acked that change and neither they nor the
committer noticed. Oh well.
Digging deeper: The files were imported from the iproute2 repository with
the GPL V2 or later boiler plate text in commit b66e907cfee2 ("tools:
bpftool: copy JSON writer from iproute2 repository")
That commit explicitly removes the boiler plate and relicenses the code
uner GPL-2.0-only and BSD-2-Clause. As Steven wrote the original code and
also the relicensing commit, it's assumed that the relicensing was intended
to do exaclty that. Just the kernel side update failed to remove the boiler
plate. Do so now.
Fixes: 907b22365115 ("tools: bpftool: dual license all files") Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Cc: Jakub Kicinski <jakub.kicinski@netronome.com> Cc: Roman Gushchin <guro@fb.com> Cc: YueHaibing <yuehaibing@huawei.com> Cc: Yonghong Song <yhs@fb.com> Cc: Stanislav Fomichev <sdf@google.com> Cc: Sean Young <sean@mess.org> Cc: Jiri Benc <jbenc@redhat.com> Cc: David Calavera <david.calavera@gmail.com> Cc: Andrey Ignatov <rdna@fb.com> Cc: Joe Stringer <joe@wand.net.nz> Cc: David Ahern <dsahern@gmail.com> Cc: Alexei Starovoitov <alexei.starovoitov@gmail.com> Cc: Petar Penkov <ppenkov@stanford.edu> Cc: Sandipan Das <sandipan@linux.ibm.com> Cc: Prashant Bhole <bhole_prashant_q7@lab.ntt.co.jp> Cc: Stephen Hemminger <stephen@networkplumber.org> Cc: John Fastabend <john.fastabend@gmail.com> Cc: Taeung Song <treeze.taeung@gmail.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Alexei Starovoitov <ast@kernel.org> Cc: Daniel Borkmann <daniel@iogearbox.net> Cc: Quentin Monnet <quentin.monnet@netronome.com> CC: okash.khawaja@gmail.com Cc: netdev@vger.kernel.org Acked-by: Jakub Kicinski <jakub.kicinski@netronome.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Daniel Borkmann [Thu, 17 Jan 2019 15:34:45 +0000 (16:34 +0100)]
bpf: fix inner map masking to prevent oob under speculation
During review I noticed that inner meta map setup for map in
map is buggy in that it does not propagate all needed data
from the reference map which the verifier is later accessing.
In particular one such case is index masking to prevent out of
bounds access under speculative execution due to missing the
map's unpriv_array/index_mask field propagation. Fix this such
that the verifier is generating the correct code for inlined
lookups in case of unpriviledged use.
Before patch (test_verifier's 'map in map access' dump):
Daniel Borkmann [Thu, 17 Jan 2019 15:15:09 +0000 (16:15 +0100)]
bpf: pull in pkt_sched.h header for tooling to fix bpftool build
Dan reported that bpftool does not compile for him:
$ make tools/bpf
DESCEND bpf
Auto-detecting system features:
.. libbfd: [ on ]
.. disassembler-four-args: [ OFF ]
DESCEND bpftool
Auto-detecting system features:
.. libbfd: [ on ]
.. disassembler-four-args: [ OFF ]
CC /opt/linux.git/tools/bpf/bpftool/net.o
In file included from /opt/linux.git/tools/include/uapi/linux/pkt_cls.h:6:0,
from /opt/linux.git/tools/include/uapi/linux/tc_act/tc_bpf.h:14,
from net.c:13:
net.c: In function 'show_dev_tc_bpf':
net.c:164:21: error: 'TC_H_CLSACT' undeclared (first use in this function)
handle = TC_H_MAKE(TC_H_CLSACT, TC_H_MIN_INGRESS);
[...]
Fix it by importing pkt_sched.h header copy into tooling
infrastructure.
Fixes: 49a249c38726 ("tools/bpftool: copy a few net uapi headers to tools directory") Fixes: f6f3bac08ff9 ("tools/bpf: bpftool: add net support") Reported-by: Dan Gilson <dan_gilson@yahoo.com>
Reference: https://bugzilla.kernel.org/show_bug.cgi?id=202315 Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Acked-by: Martin KaFai Lau <kafai@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org>
David S. Miller [Fri, 18 Jan 2019 23:12:16 +0000 (15:12 -0800)]
Merge branch 'mlxsw-fixes'
Ido Schimmel says:
====================
mlxsw: Various fixes
This patchset contains small fixes in mlxsw and one fix in the bridge
driver.
Patches #1-#4 perform small adjustments in PCI and FID code following
recent tests that were performed on the Spectrum-2 ASIC.
Patch #5 fixes the bridge driver to mark FDB entries that were added by
user as such. Otherwise, these entries will be ignored by underlying
switch drivers.
Patch #6 fixes a long standing issue in mlxsw where the driver
incorrectly programmed static FDB entries as both static and sticky.
Patches #7-#8 add test cases for above mentioned bugs.
Please consider patches #1, #2 and #4 for stable.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Ido Schimmel [Fri, 18 Jan 2019 15:58:01 +0000 (15:58 +0000)]
mlxsw: spectrum_switchdev: Do not treat static FDB entries as sticky
The driver currently treats static FDB entries as both static and
sticky. This is incorrect and prevents such entries from being roamed to
a different port via learning.
Fix this by configuring static entries with ageing disabled and roaming
enabled.
In net-next we can add proper support for the newly introduced 'sticky'
flag.
Fixes: 56ade8fe3fe1 ("mlxsw: spectrum: Add initial support for Spectrum ASIC") Signed-off-by: Ido Schimmel <idosch@mellanox.com> Reported-by: Alexander Petrovskiy <alexpe@mellanox.com> Reviewed-by: Petr Machata <petrm@mellanox.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Ido Schimmel [Fri, 18 Jan 2019 15:58:00 +0000 (15:58 +0000)]
net: bridge: Mark FDB entries that were added by user as such
Externally learned entries can be added by a user or by a switch driver
that is notifying the bridge driver about entries that were learned in
hardware.
In the first case, the entries are not marked with the 'added_by_user'
flag, which causes switch drivers to ignore them and not offload them.
The 'added_by_user' flag can be set on externally learned FDB entries
based on the 'swdev_notify' parameter in br_fdb_external_learn_add(),
which effectively means if the created / updated FDB entry was added by
a user or not.
Fixes: 816a3bed9549 ("switchdev: Add fdb.added_by_user to switchdev notifications") Signed-off-by: Ido Schimmel <idosch@mellanox.com> Reported-by: Alexander Petrovskiy <alexpe@mellanox.com> Reviewed-by: Petr Machata <petrm@mellanox.com> Cc: Roopa Prabhu <roopa@cumulusnetworks.com> Cc: Nikolay Aleksandrov <nikolay@cumulusnetworks.com> Cc: bridge@lists.linux-foundation.org Signed-off-by: David S. Miller <davem@davemloft.net>
Nir Dotan [Fri, 18 Jan 2019 15:57:59 +0000 (15:57 +0000)]
mlxsw: spectrum_fid: Update dummy FID index
When using a tc flower action of egress mirred redirect, the driver adds
an implicit FID setting action. This implicit action sets a dummy FID to
the packet and is used as part of a design for trapping unmatched flows
in OVS. While this implicit FID setting action is supposed to be a NOP
when a redirect action is added, in Spectrum-2 the FID record is
consulted as the dummy FID index is an 802.1D FID index and the packet
is dropped instead of being redirected.
Set the dummy FID index value to be within 802.1Q range. This satisfies
both Spectrum-1 which ignores the FID and Spectrum-2 which identifies it
as an 802.1Q FID and will then follow the redirect action.
Fixes: c3ab435466d5 ("mlxsw: spectrum: Extend to support Spectrum-2 ASIC") Signed-off-by: Nir Dotan <nird@mellanox.com> Signed-off-by: Ido Schimmel <idosch@mellanox.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Nir Dotan [Fri, 18 Jan 2019 15:57:56 +0000 (15:57 +0000)]
mlxsw: pci: Increase PCI SW reset timeout
Spectrum-2 PHY layer introduces a calibration period which is a part of the
Spectrum-2 firmware boot process. Hence increase the SW timeout waiting for
the firmware to come out of boot. This does not increase system boot time
in cases where the firmware PHY calibration process is done quickly.
Fixes: c3ab435466d5 ("mlxsw: spectrum: Extend to support Spectrum-2 ASIC") Signed-off-by: Nir Dotan <nird@mellanox.com> Acked-by: Jiri Pirko <jiri@mellanox.com> Signed-off-by: Ido Schimmel <idosch@mellanox.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Ido Schimmel [Fri, 18 Jan 2019 15:57:55 +0000 (15:57 +0000)]
mlxsw: pci: Ring CQ's doorbell before RDQ's
When a packet should be trapped to the CPU the device consumes a WQE
(work queue element) from an RDQ (receive descriptor queue) and copies
the packet to the address specified in the WQE. The device then tries to
post a CQE (completion queue element) that contains various metadata
(e.g., ingress port) about the packet to a CQ (completion queue).
In case the device managed to consume a WQE, but did not manage to post
the corresponding CQE, it will get stuck. This unlikely situation can be
triggered due to the scheme the driver is currently using to process
CQEs.
The driver will consume up to 512 CQEs at a time and after processing
each corresponding WQE it will ring the RDQ's doorbell, letting the
device know that a new WQE was posted for it to consume. Only after
processing all the CQEs (up to 512), the driver will ring the CQ's
doorbell, letting the device know that new ones can be posted.
Fix this by having the driver ring the CQ's doorbell for every processed
CQE, but before ringing the RDQ's doorbell. This guarantees that
whenever we post a new WQE, there is a corresponding CQE available. Copy
the currently processed CQE to prevent the device from overwriting it
with a new CQE after ringing the doorbell.
Note that the driver still arms the CQ only after processing all the
pending CQEs, so that interrupts for this CQ will only be delivered
after the driver finished its processing.
Before commit 8404f6f2e8ed ("mlxsw: pci: Allow to use CQEs of version 1
and version 2") the issue was virtually impossible to trigger since the
number of CQEs was twice the number of WQEs and the number of CQEs
processed at a time was equal to the number of available WQEs.
Fixes: 8404f6f2e8ed ("mlxsw: pci: Allow to use CQEs of version 1 and version 2") Signed-off-by: Ido Schimmel <idosch@mellanox.com> Reported-by: Semion Lisyansky <semionl@mellanox.com> Tested-by: Semion Lisyansky <semionl@mellanox.com> Acked-by: Jiri Pirko <jiri@mellanox.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Felix Manlunas [Thu, 17 Jan 2019 18:07:45 +0000 (18:07 +0000)]
MAINTAINERS: update email addresses of liquidio driver maintainers
Update email addresses of liquidio driver maintainers. Also remove a
former maintainer.
Signed-off-by: Felix Manlunas <fmanlunas@marvell.com> Acked-by: Derek Chickles <dchickles@marvell.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Jonathan Neuschäfer [Thu, 17 Jan 2019 17:02:18 +0000 (18:02 +0100)]
net: Fix typo in NET_FAILOVER help text
"also enables" should not be spelled as one word.
Fixes: cfc80d9a1163 ("net: Introduce net_failover driver") Signed-off-by: Jonathan Neuschäfer <j.neuschaefer@gmx.net> Signed-off-by: David S. Miller <davem@davemloft.net>
Ross Lagerwall [Thu, 17 Jan 2019 15:34:38 +0000 (15:34 +0000)]
net: Fix usage of pskb_trim_rcsum
In certain cases, pskb_trim_rcsum() may change skb pointers.
Reinitialize header pointers afterwards to avoid potential
use-after-frees. Add a note in the documentation of
pskb_trim_rcsum(). Found by KASAN.
Signed-off-by: Ross Lagerwall <ross.lagerwall@citrix.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Thomas Petazzoni [Wed, 16 Jan 2019 09:53:58 +0000 (10:53 +0100)]
net: phy: mdio_bus: add missing device_del() in mdiobus_register() error handling
The current code in __mdiobus_register() doesn't properly handle
failures returned by the devm_gpiod_get_optional() call: it returns
immediately, without unregistering the device that was added by the
call to device_register() earlier in the function.
This leaves a stale device, which then causes a NULL pointer
dereference in the code that handles deferred probing:
The actual error that we had from devm_gpiod_get_optional() was
-EPROBE_DEFER, due to the GPIO being provided by a driver that is
probed later than the Ethernet controller driver.
To fix this, we simply add the missing device_del() invocation in the
error path.
Fixes: 69226896ad636 ("mdio_bus: Issue GPIO RESET to PHYs") Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com> Reviewed-by: Andrew Lunn <andrew@lunn.ch> Signed-off-by: David S. Miller <davem@davemloft.net>
Otto Sabart [Mon, 14 Jan 2019 11:56:36 +0000 (12:56 +0100)]
doc: net: fix bad references to network drivers
Fix "reference to nonexisting document" warnings.
Fixes: b255e500c8dc ("net: documentation: build a directory structure for drivers") Signed-off-by: Otto Sabart <ottosabart@seberm.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Lendacky, Thomas [Thu, 17 Jan 2019 14:20:14 +0000 (14:20 +0000)]
amd-xgbe: Fix mdio access for non-zero ports and clause 45 PHYs
The XGBE hardware has support for performing MDIO operations using an
MDIO command request. The driver mistakenly uses the mdio port address
as the MDIO command request device address instead of the MDIO command
request port address. Additionally, the driver does not properly check
for and create a clause 45 MDIO command.
Check the supplied MDIO register to determine if the request is a clause
45 operation (MII_ADDR_C45). For a clause 45 operation, extract the device
address and register number from the supplied MDIO register and use them
to set the MDIO command request device address and register number fields.
For a clause 22 operation, the MDIO request device address is set to zero
and the MDIO command request register number is set to the supplied MDIO
register. In either case, the supplied MDIO port address is used as the
MDIO command request port address.
Fixes: 732f2ab7afb9 ("amd-xgbe: Add support for MDIO attached PHYs") Signed-off-by: Tom Lendacky <thomas.lendacky@amd.com> Tested-by: Shyam Sundar S K <Shyam-sundar.S-k@amd.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Camelia Groza [Thu, 17 Jan 2019 12:33:33 +0000 (14:33 +0200)]
net: phy: add missing phy driver features
The phy drivers for CS4340 and TN2020 are missing their
features attributes. Add them.
Fixes: 719655a14971 ("net: phy: Replace phy driver features u32 with link_mode bitmap") Reported-by: Scott Wood <oss@buserror.net> Signed-off-by: Camelia Groza <camelia.groza@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Yunjian Wang [Thu, 17 Jan 2019 01:46:41 +0000 (09:46 +0800)]
net: bridge: Fix ethernet header pointer before check skb forwardable
The skb header should be set to ethernet header before using
is_skb_forwardable. Because the ethernet header length has been
considered in is_skb_forwardable(including dev->hard_header_len
length).
To reproduce the issue:
1, add 2 ports on linux bridge br using following commands:
$ brctl addbr br
$ brctl addif br eth0
$ brctl addif br eth1
2, the MTU of eth0 and eth1 is 1500
3, send a packet(Data 1480, UDP 8, IP 20, Ethernet 14, VLAN 4)
from eth0 to eth1
So the expect result is packet larger than 1500 cannot pass through
eth0 and eth1. But currently, the packet passes through success, it
means eth1's MTU limit doesn't take effect.
Fixes: f6367b4660dd ("bridge: use is_skb_forwardable in forward path") Cc: bridge@lists.linux-foundation.org Cc: Nkolay Aleksandrov <nikolay@cumulusnetworks.com> Cc: Roopa Prabhu <roopa@cumulusnetworks.com> Cc: Stephen Hemminger <stephen@networkplumber.org> Signed-off-by: Yunjian Wang <wangyunjian@huawei.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Jason Wang [Wed, 16 Jan 2019 08:54:42 +0000 (16:54 +0800)]
vhost: log dirty page correctly
Vhost dirty page logging API is designed to sync through GPA. But we
try to log GIOVA when device IOTLB is enabled. This is wrong and may
lead to missing data after migration.
To solve this issue, when logging with device IOTLB enabled, we will:
1) reuse the device IOTLB translation result of GIOVA->HVA mapping to
get HVA, for writable descriptor, get HVA through iovec. For used
ring update, translate its GIOVA to HVA
2) traverse the GPA->HVA mapping to get the possible GPA and log
through GPA. Pay attention this reverse mapping is not guaranteed
to be unique, so we should log each possible GPA in this case.
This fix the failure of scp to guest during migration. In -next, we
will probably support passing GIOVA->GPA instead of GIOVA->HVA.
Fixes: 6b1e6cc7855b ("vhost: new device IOTLB API") Reported-by: Jintack Lim <jintack@cs.columbia.edu> Cc: Jintack Lim <jintack@cs.columbia.edu> Signed-off-by: Jason Wang <jasowang@redhat.com> Acked-by: Michael S. Tsirkin <mst@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Nicolas Dichtel [Thu, 17 Jan 2019 10:27:22 +0000 (11:27 +0100)]
af_packet: fix raw sockets over 6in4 tunnel
Since commit cb9f1b783850, scapy (which uses an AF_PACKET socket in
SOCK_RAW mode) is unable to send a basic icmp packet over a sit tunnel:
Here is a example of the setup:
$ ip link set ntfp2 up
$ ip addr add 10.125.0.1/24 dev ntfp2
$ ip tunnel add tun1 mode sit ttl 64 local 10.125.0.1 remote 10.125.0.2 dev ntfp2
$ ip addr add fd00:cafe:cafe::1/128 dev tun1
$ ip link set dev tun1 up
$ ip route add fd00:200::/64 dev tun1
$ scapy
>>> p = []
>>> p += IPv6(src='fd00:100::1', dst='fd00:200::1')/ICMPv6EchoRequest()
>>> send(p, count=1, inter=0.1)
>>> quit()
$ ip -s link ls dev tun1 | grep -A1 "TX.*errors"
TX: bytes packets errors dropped carrier collsns
0 0 1 0 0 0
The problem is that the network offset is set to the hard_header_len of the
output device (tun1, ie 14 + 20) and in our case, because the packet is
small (48 bytes) the pskb_inet_may_pull() fails (it tries to pull 40 bytes
(ipv6 header) starting from the network offset).
This problem is more generally related to device with variable hard header
length. To avoid a too intrusive patch in the current release, a (ugly)
workaround is proposed in this patch. It has to be cleaned up in net-next.
Yuchung Cheng [Thu, 17 Jan 2019 16:51:01 +0000 (08:51 -0800)]
bpf: fix SO_MAX_PACING_RATE to support TCP internal pacing
If sch_fq packet scheduler is not used, TCP can fallback to
internal pacing, but this requires sk_pacing_status to
be properly set.
Fixes: 8c4b4c7e9ff0 ("bpf: Add setsockopt helper function to bpf") Signed-off-by: Yuchung Cheng <ycheng@google.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Lawrence Brakmo <brakmo@fb.com> Acked-by: Martin KaFai Lau <kafai@fb.com> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Peter Oskolkov [Wed, 16 Jan 2019 16:47:54 +0000 (08:47 -0800)]
bpf: bpf_setsockopt: reset sock dst on SO_MARK changes
In sock_setsockopt() (net/core/sock.h), when SO_MARK option is used
to change sk_mark, sk_dst_reset(sk) is called. The same should be
done in bpf_setsockopt().
Fixes: 8c4b4c7e9ff0 ("bpf: Add setsockopt helper function to bpf") Reported-by: Maciej Żenczykowski <maze@google.com> Signed-off-by: Peter Oskolkov <posk@google.com> Acked-by: Martin KaFai Lau <kafai@fb.com> Reviewed-by: Maciej Żenczykowski <maze@google.com> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Alexey Kodanev [Wed, 16 Jan 2019 16:17:45 +0000 (19:17 +0300)]
udp6: add missing rehash callback to udplite
After commit 23b0269e58ae ("net: udp6: prefer listeners bound to an
address"), UDP-Lite only works when specifying a local address for
the sockets.
This is related to the problem addressed in the commit 719f835853a9
("udp: add rehash on connect()"). Moreover, __udp6_lib_lookup() now
looks for a socket immediately in the secondary hash table.
And this issue was found with LTP/network tests as well.
Fixes: 23b0269e58ae ("net: udp6: prefer listeners bound to an address") Signed-off-by: Alexey Kodanev <alexey.kodanev@oracle.com> Reviewed-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Alexey Kodanev [Wed, 16 Jan 2019 16:17:44 +0000 (19:17 +0300)]
udp: add missing rehash callback to udplite
After commit 4cdeeee9252a ("net: udp: prefer listeners bound to an
address"), UDP-Lite only works when specifying a local address for
the sockets.
This is related to the problem addressed in the commit 719f835853a9
("udp: add rehash on connect()"). Moreover, __udp4_lib_lookup() now
looks for a socket immediately in the secondary hash table.
The issue was found with LTP/network tests (UDP-Lite test-cases).
Fixes: 4cdeeee9252a ("net: udp: prefer listeners bound to an address") Signed-off-by: Alexey Kodanev <alexey.kodanev@oracle.com> Reviewed-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Ivan Vecera [Wed, 16 Jan 2019 15:53:52 +0000 (16:53 +0100)]
net/sched: cls_flower: allocate mask dynamically in fl_change()
Recent changes (especially 05cd271fd61a ("cls_flower: Support multiple
masks per priority")) in the fl_flow_mask structure grow it and its
current size e.g. on x86_64 with defconfig is 760 bytes and more than
1024 bytes with some debug options enabled. Prior the mentioned commit
its size was 176 bytes (using defconfig on x86_64).
With regard to this fact it's reasonable to allocate this structure
dynamically in fl_change() to reduce its stack size.
v2:
- use kzalloc() instead of kcalloc()
Fixes: 05cd271fd61a ("cls_flower: Support multiple masks per priority") Cc: Jiri Pirko <jiri@resnulli.us> Cc: Paul Blakey <paulb@mellanox.com> Acked-by: Jiri Pirko <jiri@mellanox.com> Signed-off-by: Ivan Vecera <ivecera@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Colin Ian King [Tue, 15 Jan 2019 18:03:38 +0000 (18:03 +0000)]
atm: he: fix sign-extension overflow on large shift
Shifting the 1 by exp by an int can lead to sign-extension overlow when
exp is 31 since 1 is an signed int and sign-extending this result to an
unsigned long long will set the upper 32 bits. Fix this by shifting an
unsigned long.
Detected by cppcheck:
(warning) Shifting signed 32-bit value by 31 bits is undefined behaviour
Signed-off-by: Colin Ian King <colin.king@canonical.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Mathieu Malaterre [Wed, 16 Jan 2019 19:29:40 +0000 (20:29 +0100)]
bpf: Make function btf_name_offset_valid static
Initially in commit 69b693f0aefa ("bpf: btf: Introduce BPF Type Format
(BTF)") the function 'btf_name_offset_valid' was introduced as static
function it was later on changed to a non-static one, and then finally
in commit 23127b33ec80 ("bpf: Create a new btf_name_by_offset() for
non type name use case") the function prototype was removed.
Revert back to original implementation and make the function static.
Remove warning triggered with W=1:
kernel/bpf/btf.c:470:6: warning: no previous prototype for 'btf_name_offset_valid' [-Wmissing-prototypes]
Fixes: 23127b33ec80 ("bpf: Create a new btf_name_by_offset() for non type name use case") Signed-off-by: Mathieu Malaterre <malat@debian.org> Acked-by: Martin KaFai Lau <kafai@fb.com> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Stanislav Fomichev [Wed, 16 Jan 2019 22:03:17 +0000 (14:03 -0800)]
selftests/bpf: retry tests that expect build-id
While running test_progs in a loop I found out that I'm sometimes hitting
"Didn't find expected build ID from the map" error.
Looking at stack_map_get_build_id_offset() it seems that it is racy (by
design) and can sometimes return BPF_STACK_BUILD_ID_IP (i.e. can't trylock
current->mm->mmap_sem).
Let's retry this test a single time.
Fixes: 13790d1cc72c ("bpf: add selftest for stackmap with build_id in NMI context") Acked-by: Song Liu <songliubraving@fb.com> Signed-off-by: Stanislav Fomichev <sdf@google.com> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Stanislav Fomichev [Wed, 16 Jan 2019 22:03:16 +0000 (14:03 -0800)]
bpf: zero out build_id for BPF_STACK_BUILD_ID_IP
When returning BPF_STACK_BUILD_ID_IP from stack_map_get_build_id_offset,
make sure that build_id field is empty. Since we are using percpu
free list, there is a possibility that we might reuse some previous
bpf_stack_build_id with non-zero build_id.
Fixes: 615755a77b24 ("bpf: extend stackmap to save binary_build_id+offset instead of address") Acked-by: Song Liu <songliubraving@fb.com> Signed-off-by: Stanislav Fomichev <sdf@google.com> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Stanislav Fomichev [Wed, 16 Jan 2019 22:03:15 +0000 (14:03 -0800)]
bpf: don't assume build-id length is always 20 bytes
Build-id length is not fixed to 20, it can be (`man ld` /--build-id):
* 128-bit (uuid)
* 160-bit (sha1)
* any length specified in ld --build-id=0xhexstring
To fix the issue of missing BPF_STACK_BUILD_ID_VALID for shorter build-ids,
assume that build-id is somewhere in the range of 1 .. 20.
Set the remaining bytes to zero.
v2:
* don't introduce new "len = min(BPF_BUILD_ID_SIZE, nhdr->n_descsz)",
we already know that nhdr->n_descsz <= BPF_BUILD_ID_SIZE if we enter
this 'if' condition
Fixes: 615755a77b24 ("bpf: extend stackmap to save binary_build_id+offset instead of address") Acked-by: Song Liu <songliubraving@fb.com> Signed-off-by: Stanislav Fomichev <sdf@google.com> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Willem de Bruijn [Tue, 15 Jan 2019 16:40:02 +0000 (11:40 -0500)]
udp: with udp_segment release on error path
Failure __ip_append_data triggers udp_flush_pending_frames, but these
tests happen later. The skb must be freed directly.
Fixes: bec1f6f697362 ("udp: generate gso with UDP_SEGMENT") Reported-by: Eric Dumazet <edumazet@google.com> Signed-off-by: Willem de Bruijn <willemb@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Taehee Yoo [Tue, 15 Jan 2019 16:35:22 +0000 (01:35 +0900)]
net: bpfilter: change section name of bpfilter UMH blob.
The section of bpfilter UMH blob is the ".bpfilter_umh". but this is not
an explicit section. so linking warning occurred at compile time for the
powerpc.
So, this patch makes use of the ".rodata" instead of the ".bpfilter_umh".
Config condition:
CONFIG_BPFILTER=y
CONFIG_BPFILTER_UMH=y
Result:
ld: warning: orphan section `.bpfilter_umh' from
`net/bpfilter/bpfilter_umh_blob.o' being placed in section `.bpfilter_umh'
Fixes: 61fbf5933d42 ("net: bpfilter: restart bpfilter_umh when error occurred") Reported-by: Stephen Rothwell <sfr@canb.auug.org.au> Signed-off-by: Taehee Yoo <ap420073@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Andrew Lunn [Tue, 15 Jan 2019 15:55:30 +0000 (16:55 +0100)]
net: phy: Add missing features to PHY drivers
The bcm87xx and micrel driver has PHYs which are missing the .features
value. Add them. The bcm87xx is a 10G FEC only PHY. Add the needed
features definition of this PHY.
Fixes: 719655a14971 ("net: phy: Replace phy driver features u32 with link_mode bitmap") Reported-by: Scott Wood <oss@buserror.net> Reported-by: Camelia Groza <camelia.groza@nxp.com> Signed-off-by: Andrew Lunn <andrew@lunn.ch> Signed-off-by: David S. Miller <davem@davemloft.net>
Jakub Kicinski [Mon, 14 Jan 2019 18:52:45 +0000 (10:52 -0800)]
ipv6: route: place a warning with duplicated string with correct extack
"IPv6: " prefix is already added by pr_fmt, no need to include
it again in the pr_warn() format. The message predates extack
support, we can replace the whole thing with an extack message.
Suggested-by: David Ahern <dsahern@gmail.com> Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com> Reviewed-by: David Ahern <dsahern@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Davide Caratti [Mon, 14 Jan 2019 17:16:44 +0000 (18:16 +0100)]
selftests: tc-testing: fix parsing of ife type
In iproute2 commit 90c5c969f0b9 ("fix print_0xhex on 32 bit"), the format
specifier for the ife type changed from 0x%X to %#llX, causing systematic
failures in the following TDC test cases:
7682 - Create valid ife encode action with mark and pass control
ef47 - Create valid ife encode action with mark and pipe control
df43 - Create valid ife encode action with mark and continue control
e4cf - Create valid ife encode action with mark and drop control
ccba - Create valid ife encode action with mark and reclassify control
a1cf - Create valid ife encode action with mark and jump control
cb3d - Create valid ife encode action with mark value at 32-bit maximum
95ed - Create valid ife encode action with prio and pass control
aa17 - Create valid ife encode action with prio and pipe control
74c7 - Create valid ife encode action with prio and continue control
7a97 - Create valid ife encode action with prio and drop control
f66b - Create valid ife encode action with prio and reclassify control
3056 - Create valid ife encode action with prio and jump control
7dd3 - Create valid ife encode action with prio value at 32-bit maximum
05bb - Create valid ife encode action with tcindex and pass control
ce65 - Create valid ife encode action with tcindex and pipe control
09cd - Create valid ife encode action with tcindex and continue control
8eb5 - Create valid ife encode action with tcindex and continue control
451a - Create valid ife encode action with tcindex and drop control
d76c - Create valid ife encode action with tcindex and reclassify control
e731 - Create valid ife encode action with tcindex and jump control
b7b8 - Create valid ife encode action with tcindex value at 16-bit maximum
2a9c - Create valid ife encode action with mac src parameter
cf5c - Create valid ife encode action with mac dst parameter
2353 - Create valid ife encode action with mac src and mac dst parameters
552c - Create valid ife encode action with mark and type parameters
0421 - Create valid ife encode action with prio and type parameters
4017 - Create valid ife encode action with tcindex and type parameters
fac3 - Create valid ife encode action with index at 32-bit maximnum
7c25 - Create valid ife decode action with pass control
dccb - Create valid ife decode action with pipe control
7bb9 - Create valid ife decode action with continue control
d9ad - Create valid ife decode action with drop control
219f - Create valid ife decode action with reclassify control
8f44 - Create valid ife decode action with jump control
b330 - Create ife encode action with cookie
Change 'matchPattern' values, allowing '0' and '0x0' if ife type is equal
to 0, and accepting both '0x' and '0X' otherwise, to let these tests pass
both with old and new tc binaries.
While at it, fix a small typo in test case fac3 ('maximnum'->'maximum').
Signed-off-by: Davide Caratti <dcaratti@redhat.com> Acked-by: Stephen Hemminger <stephen@networkplumber.org> Signed-off-by: David S. Miller <davem@davemloft.net>
Konstantin Khlebnikov [Mon, 14 Jan 2019 10:38:43 +0000 (13:38 +0300)]
net/core/neighbour: fix kmemleak minimal reference count for hash tables
This should be 1 for normal allocations, 0 disables leak reporting.
Signed-off-by: Konstantin Khlebnikov <khlebnikov@yandex-team.ru> Reported-by: Cong Wang <xiyou.wangcong@gmail.com> Fixes: 85704cb8dcfd ("net/core/neighbour: tell kmemleak about hash tables") Signed-off-by: David S. Miller <davem@davemloft.net>
Xin Long [Mon, 14 Jan 2019 10:34:02 +0000 (18:34 +0800)]
sctp: allocate sctp_sockaddr_entry with kzalloc
The similar issue as fixed in Commit 4a2eb0c37b47 ("sctp: initialize
sin6_flowinfo for ipv6 addrs in sctp_inet6addr_event") also exists
in sctp_inetaddr_event, as Alexander noticed.
To fix it, allocate sctp_sockaddr_entry with kzalloc for both sctp
ipv4 and ipv6 addresses, as does in sctp_v4/6_copy_addrlist().
Reported-by: Alexander Potapenko <glider@google.com> Signed-off-by: Xin Long <lucien.xin@gmail.com> Reported-by: syzbot+ae0c70c0c2d40c51bb92@syzkaller.appspotmail.com Acked-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com> Acked-by: Neil Horman <nhorman@tuxdriver.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Xin Long [Mon, 14 Jan 2019 10:10:06 +0000 (18:10 +0800)]
erspan: build the header with the right proto according to erspan_ver
As said in draft-foschiano-erspan-03#section4:
Different frame variants known as "ERSPAN Types" can be
distinguished based on the GRE "Protocol Type" field value: Type I
and II's value is 0x88BE while Type III's is 0x22EB [ETYPES].
So set it properly in erspan_xmit() according to erspan_ver. While at
it, also remove the unused parameter 'proto' in erspan_fb_xmit().
Fixes: 94d7d8f29287 ("ip6_gre: add erspan v2 support") Reported-by: Jianlin Shi <jishi@redhat.com> Signed-off-by: Xin Long <lucien.xin@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Ross Lagerwall [Mon, 14 Jan 2019 09:16:56 +0000 (09:16 +0000)]
openvswitch: Avoid OOB read when parsing flow nlattrs
For nested and variable attributes, the expected length of an attribute
is not known and marked by a negative number. This results in an OOB
read when the expected length is later used to check if the attribute is
all zeros. Fix this by using the actual length of the attribute rather
than the expected length.
Signed-off-by: Ross Lagerwall <ross.lagerwall@citrix.com> Acked-by: Pravin B Shelar <pshelar@ovn.org> Signed-off-by: David S. Miller <davem@davemloft.net>
Dmitry Bezrukov [Mon, 14 Jan 2019 08:14:41 +0000 (08:14 +0000)]
net: usb: aqc111: Extend HWID table by TRENDnet device
New device of TRENDnet based on aqc111u
Add this ID to blacklist of cdc_ether driver as well
Signed-off-by: Dmitry Bezrukov <dmitry.bezrukov@aquantia.com> Signed-off-by: Igor Russkikh <igor.russkikh@aquantia.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Cong Wang [Sat, 12 Jan 2019 02:55:42 +0000 (18:55 -0800)]
net_sched: refetch skb protocol for each filter
Martin reported a set of filters don't work after changing
from reclassify to continue. Looking into the code, it
looks like skb protocol is not always fetched for each
iteration of the filters. But, as demonstrated by Martin,
TC actions could modify skb->protocol, for example act_vlan,
this means we have to refetch skb protocol in each iteration,
rather than using the one we fetch in the beginning of the loop.
This bug is _not_ introduced by commit 3b3ae880266d
("net: sched: consolidate tc_classify{,_compat}"), technically,
if act_vlan is the only action that modifies skb protocol, then
it is commit c7e2b9689ef8 ("sched: introduce vlan action") which
introduced this bug.
Reported-by: Martin Olsson <martin.olsson+netdev@sentorsecurity.com> Cc: Jamal Hadi Salim <jhs@mojatatu.com> Cc: Jiri Pirko <jiri@resnulli.us> Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com> Acked-by: Jamal Hadi Salim <jhs@mojatatu.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Timotej Lazar [Sun, 13 Jan 2019 00:22:55 +0000 (01:22 +0100)]
net: phy: meson-gxl: Use the genphy_soft_reset callback
Since the referenced commit, Ethernet fails to come up at boot on the
board meson-gxl-s905x-libretech-cc. Fix this by re-enabling the
genphy_soft_reset callback for the Amlogic Meson GXL PHY driver.
Fixes: 6e2d85ec0559 ("net: phy: Stop with excessive soft reset") Signed-off-by: Timotej Lazar <timotej.lazar@araneo.si> Reviewed-by: Florian Fainelli <f.fainelli@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Fixes: b8a51b38e4d4 ("fou, fou6: ICMP error handlers for FoU and GUE") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: syzbot <syzkaller@googlegroups.com> Cc: Stefano Brivio <sbrivio@redhat.com> Cc: Sabrina Dubroca <sd@queasysnail.net> Signed-off-by: David S. Miller <davem@davemloft.net>
Davide Caratti [Fri, 11 Jan 2019 14:08:23 +0000 (15:08 +0100)]
selftests: tc-testing: fix tunnel_key failure if dst_port is unspecified
After commit 1c25324caf82 ("net/sched: act_tunnel_key: Don't dump dst port
if it wasn't set"), act_tunnel_key doesn't dump anymore the destination
port, unless it was explicitly configured. This caused systematic failures
in the following TDC test case:
7a88 - Add tunnel_key action with cookie parameter
Avoid matching zero values of TCA_TUNNEL_KEY_ENC_DST_PORT to let the test
pass again.
Signed-off-by: Davide Caratti <dcaratti@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Davide Caratti [Fri, 11 Jan 2019 10:49:58 +0000 (11:49 +0100)]
selftests: tc-testing: drop test on missing tunnel key id
After merge of commit 80ef0f22ceda ("net/sched: act_tunnel_key: Allow
key-less tunnels"), act_tunnel_key does not reject anymore requests to
install 'set' rules where the key id is missing. Therefore, drop the
following TDC testcase:
ba4e - Add tunnel_key set action with missing mandatory id parameter
because it's going to become a systematic fail as soon as userspace
iproute2 will start supporting key-less tunnels.
Signed-off-by: Davide Caratti <dcaratti@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Andrew Lunn [Thu, 10 Jan 2019 21:48:36 +0000 (22:48 +0100)]
net: phy: marvell: Errata for mv88e6390 internal PHYs
The VOD can be out of spec, unless some magic value is poked into an
undocumented register in an undocumented page.
Fixes: e4cf8a38fc0d ("net: phy: Marvell: Add mv88e6390 internal PHY") Signed-off-by: Andrew Lunn <andrew@lunn.ch> Signed-off-by: David S. Miller <davem@davemloft.net>
Greg Kroah-Hartman [Thu, 10 Jan 2019 20:24:13 +0000 (21:24 +0100)]
IN_BADCLASS: fix macro to actually work
Commit 65cab850f0ee ("net: Allow class-e address assignment via ifconfig
ioctl") modified the IN_BADCLASS macro a bit, but unfortunatly one too
many '(' characters were added to the line, making any code that used
it, not build properly.
Also, the macro now compares an unsigned with a signed value, which
isn't ok, so fix that up by making both types match properly.
Reported-by: Christopher Ferris <cferris@google.com> Fixes: 65cab850f0ee ("net: Allow class-e address assignment via ifconfig ioctl") Cc: Dave Taht <dave.taht@gmail.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Signed-off-by: David S. Miller <davem@davemloft.net>
Heiner Kallweit [Thu, 10 Jan 2019 19:22:26 +0000 (20:22 +0100)]
net: phy: micrel: set soft_reset callback to genphy_soft_reset for KSZ9031
So far genphy_soft_reset was used automatically if the PHY driver
didn't implement the soft_reset callback. This changed with the
mentioned commit and broke KSZ9031. To fix this configure the
KSZ9031 PHY driver to use genphy_soft_reset.
Fixes: 6e2d85ec0559 ("net: phy: Stop with excessive soft reset") Reported-by: Tony Lindgren <tony@atomide.com> Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com> Tested-by: Tony Lindgren <tony@atomide.com> Tested-by: Sekhar Nori <nsekhar@ti.com> Reviewed-by: Florian Fainelli <f.fainelli@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
when the tunnel_key action is replaced, the kernel forgets to release the
dst metadata: ensure they are released by tunnel_key_init(), the same way
it's done in tunnel_key_release().
Fixes: d0f6dd8a914f4 ("net/sched: Introduce act_tunnel_key") Signed-off-by: Davide Caratti <dcaratti@redhat.com> Acked-by: Cong Wang <xiyou.wangcong@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
David Howells [Thu, 10 Jan 2019 16:59:13 +0000 (16:59 +0000)]
Revert "rxrpc: Allow failed client calls to be retried"
The changes introduced to allow rxrpc calls to be retried creates an issue
when it comes to refcounting afs_call structs. The problem is that when
rxrpc_send_data() queues the last packet for an asynchronous call, the
following sequence can occur:
(1) The notify_end_tx callback is invoked which causes the state in the
afs_call to be changed from AFS_CALL_CL_REQUESTING or
AFS_CALL_SV_REPLYING.
(2) afs_deliver_to_call() can then process event notifications from rxrpc
on the async_work queue.
(3) Delivery of events, such as an abort from the server, can cause the
afs_call state to be changed to AFS_CALL_COMPLETE on async_work.
(4) For an asynchronous call, afs_process_async_call() notes that the call
is complete and tried to clean up all the refs on async_work.
(5) rxrpc_send_data() might return the amount of data transferred
(success) or an error - which could in turn reflect a local error or a
received error.
Synchronising the clean up after rxrpc_kernel_send_data() returns an error
with the asynchronous cleanup is then tricky to get right.
Mostly revert commit c038a58ccfd6704d4d7d60ed3d6a0fca13cf13a4. The two API
functions the original commit added aren't currently used. This makes
rxrpc_kernel_send_data() always return successfully if it queued the data
it was given.
Note that this doesn't affect synchronous calls since their Rx notification
function merely pokes a wait queue and does not refcounting. The
asynchronous call notification function *has* to do refcounting and pass a
ref over the work item to avoid the need to sync the workqueue in call
cleanup.
Signed-off-by: David Howells <dhowells@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Wed, 16 Jan 2019 04:29:21 +0000 (20:29 -0800)]
Merge branch 'tipc-uninit-values'
Ying Xue says:
====================
tipc: fix uninit-value issues reported by syzbot
Recently, syzbot complained that TIPC module exits several issues
associated with uninit-value type. So, in this series, we try to
fix them as many as possible.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
In tipc_nl_compat_recv(), when the len variable returned by
nlmsg_attrlen() is 0, the message is still treated as a valid one,
which is obviously unresonable. When len is zero, it means the
message not only doesn't contain any valid TLV payload, but also
TLV header is not included. Under this stituation, tlv_type field
in TLV header is still accessed in tipc_nl_compat_dumpit() or
tipc_nl_compat_doit(), but the field space is obviously illegal.
Of course, it is not initialized.
Reported-by: syzbot+bca0dc46634781f08b38@syzkaller.appspotmail.com Reported-by: syzbot+6bdb590321a7ae40c1a6@syzkaller.appspotmail.com Signed-off-by: Ying Xue <ying.xue@windriver.com> Signed-off-by: David S. Miller <davem@davemloft.net>
We cannot take for granted the thing that the length of data contained
in TLV is longer than the size of struct tipc_name_table_query in
tipc_nl_compat_name_table_dump().
Reported-by: syzbot+06e771a754829716a327@syzkaller.appspotmail.com Signed-off-by: Ying Xue <ying.xue@windriver.com> Signed-off-by: David S. Miller <davem@davemloft.net>
The uninitialised access happened in
nla_put_string(skb, TIPC_NLA_LINK_NAME, lc->name)
This is because lc->name string is not validated before it's used.
Reported-by: syzbot+d78b8a29241a195aefb8@syzkaller.appspotmail.com Signed-off-by: Ying Xue <ying.xue@windriver.com> Signed-off-by: David S. Miller <davem@davemloft.net>
The root cause is that we don't validate whether bear name is a valid
string in tipc_nl_compat_bearer_enable().
Meanwhile, we also fix the same issue in the following functions:
tipc_nl_compat_bearer_disable()
tipc_nl_compat_link_stat_dump()
tipc_nl_compat_media_set()
tipc_nl_compat_bearer_set()
Reported-by: syzbot+b33d5cae0efd35dbfe77@syzkaller.appspotmail.com Signed-off-by: Ying Xue <ying.xue@windriver.com> Signed-off-by: David S. Miller <davem@davemloft.net>
The uninitialised access happened in tipc_nl_compat_link_reset_stats:
nla_put_string(skb, TIPC_NLA_LINK_NAME, name)
This is because name string is not validated before it's used.
Reported-by: syzbot+e01d94b5a4c266be6e4c@syzkaller.appspotmail.com Signed-off-by: Ying Xue <ying.xue@windriver.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Local variable description: ----s.i@tipc_conn_recv_work
Variable was created at:
tipc_conn_recv_work+0x65/0x560 net/tipc/topsrv.c:419
process_one_work+0x12c6/0x1f60 kernel/workqueue.c:2145
In tipc_conn_rcv_from_sock(), it always supposes the length of message
received from sock_recvmsg() is not smaller than the size of struct
tipc_subscr. However, this assumption is false. Especially when the
length of received message is shorter than struct tipc_subscr size,
we will end up touching uninitialized fields in tipc_conn_rcv_sub().
Reported-by: syzbot+8951a3065ee7fd6d6e23@syzkaller.appspotmail.com Reported-by: syzbot+75e6e042c5bbf691fc82@syzkaller.appspotmail.com Signed-off-by: Ying Xue <ying.xue@windriver.com> Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Wed, 16 Jan 2019 04:12:01 +0000 (20:12 -0800)]
Merge branch 'sch_cake-leaf-qdisc-fixes'
Toke Høiland-Jørgensen says:
====================
sched: Fix qdisc interactions exposed by using sch_cake as a leaf qdisc
This series fixes a couple of issues exposed by running sch_cake as a
leaf qdisc in an HFSC tree, which were discovered and reported by Pete
Heist. The interaction between CAKE's GSO splitting and the parent
qdisc's notion of its own queue length could cause queue stalls. While
investigating the report, I also noticed that several qdiscs would
dereference the skb pointer after dequeue, which is potentially
problematic since the GSO splitting code also frees the original skb.
See the individual patches in the series for details.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Toke Høiland-Jørgensen [Wed, 9 Jan 2019 16:09:44 +0000 (17:09 +0100)]
sch_cake: Correctly update parent qlen when splitting GSO packets
To ensure parent qdiscs have the same notion of the number of enqueued
packets even after splitting a GSO packet, update the qdisc tree with the
number of packets that was added due to the split.
Reported-by: Pete Heist <pete@heistp.net> Tested-by: Pete Heist <pete@heistp.net> Signed-off-by: Toke Høiland-Jørgensen <toke@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Toke Høiland-Jørgensen [Wed, 9 Jan 2019 16:09:43 +0000 (17:09 +0100)]
sched: Fix detection of empty queues in child qdiscs
Several qdiscs check on enqueue whether the packet was enqueued to a class
with an empty queue, in which case the class is activated. This is done by
checking if the qlen is exactly 1 after enqueue. However, if GSO splitting
is enabled in the child qdisc, a single packet can result in a qlen longer
than 1. This means the activation check fails, leading to a stalled queue.
Fix this by checking if the queue is empty *before* enqueue, and running
the activation logic if this was the case.
Reported-by: Pete Heist <pete@heistp.net> Signed-off-by: Toke Høiland-Jørgensen <toke@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Toke Høiland-Jørgensen [Wed, 9 Jan 2019 16:09:42 +0000 (17:09 +0100)]
sched: Avoid dereferencing skb pointer after child enqueue
Parent qdiscs may dereference the pointer to the enqueued skb after
enqueue. However, both CAKE and TBF call consume_skb() on the original skb
when splitting GSO packets, leading to a potential use-after-free in the
parent. Fix this by avoiding dereferencing the skb pointer after enqueueing
to the child.
Signed-off-by: Toke Høiland-Jørgensen <toke@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Stanislav Fomichev [Tue, 15 Jan 2019 22:03:27 +0000 (14:03 -0800)]
tools/bpf: properly account for libbfd variations
On some platforms, in order to link against libbfd, we need to
link against liberty and even possibly libz. Account for that
in the bpftool Makefile. We now have proper feature detection
for each case, so handle each one separately.
See recent commit 14541b1e7e72 ("perf build: Don't unconditionally link the
libbfd feature test to -liberty and -lz") where I fixed feature
detection.
v2 (addressed Jakub's nits):
* better syntax for 'else ifeq'
* no space between ifeq args
v3:
* use LIBS, not EXTLIBS for -DHAVE_LIBBFD_SUPPORT
Fixes: 29a9c10e4110 ("bpftool: make libbfd optional") Signed-off-by: Stanislav Fomichev <sdf@google.com> Acked-by: Jakub Kicinski <jakub.kicinski@netronome.com> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Hangbin Liu [Thu, 10 Jan 2019 03:17:42 +0000 (11:17 +0800)]
ip6_gre: update version related info when changing link
We forgot to update ip6erspan version related info when changing link,
which will cause setting new hwid failed.
Reported-by: Jianlin Shi <jishi@redhat.com> Fixes: 94d7d8f292870 ("ip6_gre: add erspan v2 support") Signed-off-by: Hangbin Liu <liuhangbin@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Heiner Kallweit [Wed, 9 Jan 2019 19:34:56 +0000 (20:34 +0100)]
net: phy: fix too strict check in phy_start_aneg
When adding checks to detect wrong usage of the phylib API we added
a check to phy_start_aneg() which is too strict. If the phylib
state machine is in state PHY_HALTED we should allow reconfiguring
and restarting aneg, and just don't touch the state.
Fixes: 2b3e88ea6528 ("net: phy: improve phy state checking") Reported-by: Chris Wilson <chris@chris-wilson.co.uk> Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com> Tested-by: Chris Wilson <chris@chris-wilson.co.uk> Reviewed-by: Florian Fainelli <f.fainelli@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Due to problems found in additional testing, this causes an illegal
context switch in the RCU read-side critical section.
CC: Dave Jones <davej@codemonkey.org.uk> CC: Cong Wang <xiyou.wangcong@gmail.com> CC: Jan Jablonsky <jan.jablonsky@thalesgroup.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Dan Carpenter [Wed, 9 Jan 2019 10:50:07 +0000 (13:50 +0300)]
selftests/txtimestamp: Fix an equals vs assign bug
This should be == instead of =.
Fixes: b52354aa068e ("selftests: expand txtimestamp with ipv6 dgram + raw and pf_packet") Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> Acked-by: Willem de Bruijn <willemb@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Ido Schimmel [Wed, 9 Jan 2019 09:57:39 +0000 (09:57 +0000)]
net: ipv4: Fix memory leak in network namespace dismantle
IPv4 routing tables are flushed in two cases:
1. In response to events in the netdev and inetaddr notification chains
2. When a network namespace is being dismantled
In both cases only routes associated with a dead nexthop group are
flushed. However, a nexthop group will only be marked as dead in case it
is populated with actual nexthops using a nexthop device. This is not
the case when the route in question is an error route (e.g.,
'blackhole', 'unreachable').
Therefore, when a network namespace is being dismantled such routes are
not flushed and leaked [1].
To reproduce:
# ip netns add blue
# ip -n blue route add unreachable 192.0.2.0/24
# ip netns del blue
Fix this by not skipping error routes that are not marked with
RTNH_F_DEAD when flushing the routing tables.
To prevent the flushing of such routes in case #1, add a parameter to
fib_table_flush() that indicates if the table is flushed as part of
namespace dismantle or not.
Note that this problem does not exist in IPv6 since error routes are
associated with the loopback device.
Fixes: 8cced9eff1d4 ("[NETNS]: Enable routing configuration in non-initial namespace.") Signed-off-by: Ido Schimmel <idosch@mellanox.com> Reviewed-by: David Ahern <dsahern@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Olivier Matz [Wed, 9 Jan 2019 09:57:21 +0000 (10:57 +0100)]
ip6_gre: fix tunnel list corruption for x-netns
In changelink ops, the ip6gre_net pointer is retrieved from
dev_net(dev), which is wrong in case of x-netns. Thus, the tunnel is not
unlinked from its current list and is relinked into another net
namespace. This corrupts the tunnel lists and can later trigger a kernel
oops.
Fix this by retrieving the netns from device private area.
Fixes: c8632fc30bb0 ("net: ip6_gre: Split up ip6gre_changelink()") Cc: Petr Machata <petrm@mellanox.com> Signed-off-by: Olivier Matz <olivier.matz@6wind.com> Acked-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Signed-off-by: David S. Miller <davem@davemloft.net>
This is the first batch of Netfilter fixes for your net tree:
1) Fix endless loop in nf_tables rules netlink dump, from Phil Sutter.
2) Reference counter leak in object from the error path, from Taehee Yoo.
3) Selective rule dump requires table and chain.
4) Fix DNAT with nft_flow_offload reverse route lookup, from wenxu.
5) Use GFP_KERNEL_ACCOUNT in vmalloc allocation from ebtables, from
Shakeel Butt.
6) Set ifindex from route to fix interaction with VRF slave device,
also from wenxu.
7) Use nfct_help() to check for conntrack helper, IPS_HELPER status
flag is only set from explicit helpers via -j CT, from Henry Yen.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Lorenz Bauer [Tue, 8 Jan 2019 13:58:00 +0000 (13:58 +0000)]
bpf: libbpf: retry loading program on EAGAIN
Commit c3494801cd17 ("bpf: check pending signals while
verifying programs") makes it possible for the BPF_PROG_LOAD
to fail with EAGAIN. Retry unconditionally in this case.
Fixes: c3494801cd17 ("bpf: check pending signals while verifying programs") Signed-off-by: Lorenz Bauer <lmb@cloudflare.com> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
x86 compilation has required asm goto support since 4.17.
Since clang does not support asm goto, at 4.17,
Commit b1ae32dbab50 ("x86/cpufeature: Guard asm_volatile_goto usage
for BPF compilation") worked around the issue by permitting an
alternative implementation without asm goto for clang.
Compiling samples/bpf directories, most bpf programs failed
compilation with error messages like:
In file included from /home/yhs/work/bpf-next/samples/bpf/xdp_sample_pkts_kern.c:2:
In file included from /home/yhs/work/bpf-next/include/linux/ptrace.h:6:
In file included from /home/yhs/work/bpf-next/include/linux/sched.h:15:
In file included from /home/yhs/work/bpf-next/include/linux/sem.h:5:
In file included from /home/yhs/work/bpf-next/include/uapi/linux/sem.h:5:
In file included from /home/yhs/work/bpf-next/include/linux/ipc.h:9:
In file included from /home/yhs/work/bpf-next/include/linux/refcount.h:72:
/home/yhs/work/bpf-next/arch/x86/include/asm/refcount.h:70:9: error: 'asm goto' constructs are not supported yet
return GEN_BINARY_SUFFIXED_RMWcc(LOCK_PREFIX "subl",
^
/home/yhs/work/bpf-next/arch/x86/include/asm/rmwcc.h:67:2: note: expanded from macro 'GEN_BINARY_SUFFIXED_RMWcc'
__GEN_RMWcc(op " %[val], %[var]\n\t" suffix, var, cc, \
^
/home/yhs/work/bpf-next/arch/x86/include/asm/rmwcc.h:21:2: note: expanded from macro '__GEN_RMWcc'
asm_volatile_goto (fullop "; j" #cc " %l[cc_label]" \
^
/home/yhs/work/bpf-next/include/linux/compiler_types.h:188:37: note: expanded from macro 'asm_volatile_goto'
#define asm_volatile_goto(x...) asm goto(x)
Most implementation does not even provide an alternative
implementation. And it is also not practical to make changes
for each call site.
This patch workarounded the asm goto issue by redefining the macro like below:
#define asm_volatile_goto(x...) asm volatile("invalid use of asm_volatile_goto")
If asm_volatile_goto is not used by bpf programs, which is typically the case, nothing bad
will happen. If asm_volatile_goto is used by bpf programs, which is incorrect, the compiler
will issue an error since "invalid use of asm_volatile_goto" is not valid assembly codes.
With this patch, all bpf programs under samples/bpf can pass compilation.
Note that bpf programs under tools/testing/selftests/bpf/ compiled fine as
they do not access kernel internal headers.
Fixes: e769742d3584 ("Revert "x86/jump-labels: Macrofy inline assembly code to work around GCC inlining bugs"") Fixes: 18fe58229d80 ("x86, asm: change the GEN_*_RMWcc() macros to not quote the condition") Acked-by: Alexei Starovoitov <ast@kernel.org> Signed-off-by: Yonghong Song <yhs@fb.com> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Krzysztof Kazimierczak [Thu, 10 Jan 2019 19:29:02 +0000 (20:29 +0100)]
xsk: Check if a queue exists during umem setup
In the xdp_umem_assign_dev() path, the xsk code does not
check if a queue for which umem is to be created exists.
It leads to a situation where umem is not assigned to any
Tx/Rx queue of a netdevice, without notifying the stack
about an error. This affects both XDP_SKB and XDP_DRV
modes - in case of XDP_DRV_ZC, queue index is checked by
the driver.
This patch fixes xsk code, so that in both XDP_SKB and
XDP_DRV mode of AF_XDP, an error is returned when requested
queue index exceedes an existing maximum.
Fixes: c9b47cc1fabca ("xsk: fix bug when trying to use both copy and zero-copy on one queue id") Reported-by: Jakub Spizewski <jakub.spizewski@intel.com> Signed-off-by: Krzysztof Kazimierczak <krzysztof.kazimierczak@intel.com> Acked-by: Björn Töpel <bjorn.topel@intel.com> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Linus Torvalds [Tue, 15 Jan 2019 17:28:26 +0000 (05:28 +1200)]
Merge tag 'trace-v5.0-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace
Pull tracing fix from Steven Rostedt:
"Andrea Righi fixed a NULL pointer dereference in trace_kprobe_create()
It is possible to trigger a NULL pointer dereference by writing an
incorrectly formatted string to the krpobe_events file"
* tag 'trace-v5.0-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace:
tracing/kprobes: Fix NULL pointer dereference in trace_kprobe_create()
1) Fix regression in multi-SKB responses to RTM_GETADDR, from Arthur
Gautier.
2) Fix ipv6 frag parsing in openvswitch, from Yi-Hung Wei.
3) Unbounded recursion in ipv4 and ipv6 GUE tunnels, from Stefano
Brivio.
4) Use after free in hns driver, from Yonglong Liu.
5) icmp6_send() needs to handle the case of NULL skb, from Eric
Dumazet.
6) Missing rcu read lock in __inet6_bind() when operating on mapped
addresses, from David Ahern.
7) Memory leak in tipc-nl_compat_publ_dump(), from Gustavo A. R. Silva.
8) Fix PHY vs r8169 module loading ordering issues, from Heiner
Kallweit.
9) Fix bridge vlan memory leak, from Ido Schimmel.
10) Dev refcount leak in AF_PACKET, from Jason Gunthorpe.
11) Infoleak in ipv6_local_error(), flow label isn't completely
initialized. From Eric Dumazet.
12) Handle mv88e6390 errata, from Andrew Lunn.
13) Making vhost/vsock CID hashing consistent, from Zha Bin.
14) Fix lack of UMH cleanup when it unexpectedly exits, from Taehee Yoo.
15) Bridge forwarding must clear skb->tstamp, from Paolo Abeni.
* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net: (87 commits)
bnxt_en: Fix context memory allocation.
bnxt_en: Fix ring checking logic on 57500 chips.
mISDN: hfcsusb: Use struct_size() in kzalloc()
net: clear skb->tstamp in bridge forwarding path
net: bpfilter: disallow to remove bpfilter module while being used
net: bpfilter: restart bpfilter_umh when error occurred
net: bpfilter: use cleanup callback to release umh_info
umh: add exit routine for UMH process
isdn: i4l: isdn_tty: Fix some concurrency double-free bugs
vhost/vsock: fix vhost vsock cid hashing inconsistent
net: stmmac: Prevent RX starvation in stmmac_napi_poll()
net: stmmac: Fix the logic of checking if RX Watchdog must be enabled
net: stmmac: Check if CBS is supported before configuring
net: stmmac: dwxgmac2: Only clear interrupts that are active
net: stmmac: Fix PCI module removal leak
tools/bpf: fix bpftool map dump with bitfields
tools/bpf: test btf bitfield with >=256 struct member offset
bpf: fix bpffs bitfield pretty print
net: ethernet: mediatek: fix warning in phy_start_aneg
tcp: change txhash on SYN-data timeout
...
Andrea Righi [Fri, 11 Jan 2019 06:01:13 +0000 (07:01 +0100)]
tracing/kprobes: Fix NULL pointer dereference in trace_kprobe_create()
It is possible to trigger a NULL pointer dereference by writing an
incorrectly formatted string to krpobe_events (trying to create a
kretprobe omitting the symbol).
Ming Lei [Tue, 15 Jan 2019 03:59:52 +0000 (11:59 +0800)]
sbitmap: Protect swap_lock from hardirq
Because we may call blk_mq_get_driver_tag() directly from
blk_mq_dispatch_rq_list() without holding any lock, then HARDIRQ may
come and the above DEADLOCK is triggered.
Commit ab53dcfb3e7b ("sbitmap: Protect swap_lock from hardirq") tries to
fix this issue by using 'spin_lock_bh', which isn't enough because we
complete request from hardirq context direclty in case of multiqueue.
Cc: Clark Williams <williams@redhat.com> Fixes: ab53dcfb3e7b ("sbitmap: Protect swap_lock from hardirq") Cc: Jens Axboe <axboe@kernel.dk> Cc: Ming Lei <ming.lei@redhat.com> Cc: Guenter Roeck <linux@roeck-us.net> Cc: Steven Rostedt (VMware) <rostedt@goodmis.org> Signed-off-by: Ming Lei <ming.lei@redhat.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
projects / users / jedix / linux-maple.git / log