Pablo Neira Ayuso says:
====================
Netfilter updates for net-next
A small batch with accumulated updates in nf-next, mostly IPVS updates,
they are:
1) Add 64-bits stats counters to IPVS, from Julian Anastasov.
2) Move NETFILTER_XT_MATCH_ADDRTYPE out of NETFILTER_ADVANCED as docker
seem to require this, from Anton Blanchard.
3) Use boolean instead of numeric value in set_match_v*(), from
coccinelle via Fengguang Wu.
4) Allows rescheduling of new connections in IPVS when port reuse is
detected, from Marcelo Ricardo Leitner.
5) Add missing bits to support arptables extensions from nft_compat,
from Arturo Borrero.
Patrick is preparing a large batch to enhance the set infrastructure,
named expressions among other things, that should follow up soon after
this batch.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Both sk_attach_filter() and sk_attach_bpf() are setting up sk_filter,
charging skmem and attaching it to the socket after we got the eBPF
prog up and ready. Lets refactor that into a common helper.
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
Johan Hedberg says:
====================
pull request: bluetooth-next 2015-03-02
Here's the first bluetooth-next pull request targeting the 4.1 kernel:
- ieee802154/6lowpan cleanups
- SCO routing to host interface support for the btmrvl driver
- AMP code cleanups
- Fixes to AMP HCI init sequence
- Refactoring of the HCI callback mechanism
- Added shutdown routine for Intel controllers in the btusb driver
- New config option to enable/disable Bluetooth debugfs information
- Fix for early data reception on L2CAP fixed channels
Please let me know if there are any issues pulling. Thanks.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
The switch_mm function does nothing in case the prev and next mm
are the same. It can happen that a crst_table_downgrade has changed
the top-level pgd in the meantime on a different CPU. Always store
the new ASCE to be picked up in entry.S.
[heiko.carstens@de.ibm.com]: Bug was introduced with git commit
53e857f308 ("s390/mm,tlb: race of lazy TLB flush vs. recreation
of TLB entries") and causes random crashes due to broken page tables
being used.
Reported-by: Dominik Vogt <vogt@linux.vnet.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
With git commit 4d92f50249 ("s390: reintroduce diag 44 calls for
cpu_relax()") I reintroduced a non-trivial cpu_relax() variant on s390.
The difference to the previous variant however is that the new version is
an out-of-line function, which will be traced if function tracing is enabled.
Switching to different tracers includes instruction patching. Therefore this
is done within stop_machine() "context" to prevent that any function tracing
is going on while instructions are being patched.
With the new out-of-line variant of cpu_relax() this is not true anymore,
since cpu_relax() gets called in a busy loop by all waiting cpus within
stop_machine() until function patching is finished.
Therefore cpu_relax() must be marked notrace.
This fixes kernel crashes when frequently switching between "function" and
"function_graph" tracers.
Moving cpu_relax() to a header file again, doesn't work because of header
include order dependencies.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Ying Xue says:
====================
net: Remove iocb argument from sendmsg and recvmsg
Currently there is only one user - TIPC whose sendmsg() instances
using iocb argument. Meanwhile, there is no user using iocb argument
in its recvmsg() instance. Therefore, if we eliminate the werid usage
of iobc argument from TIPC, the iocb argument can be removed from
all sendmsg() and recvmsg() instances of the whole networking stack.
Reference:
https://patchwork.ozlabs.org/patch/433960/
Changes:
v2:
* Fix compile errors of DCCP module pointed by David
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
After TIPC doesn't depend on iocb argument in its internal
implementations of sendmsg() and recvmsg() hooks defined in proto
structure, no any user is using iocb argument in them at all now.
Then we can drop the redundant iocb argument completely from kinds of
implementations of both sendmsg() and recvmsg() in the entire
networking stack.
Cc: Christoph Hellwig <hch@lst.de>
Suggested-by: Al Viro <viro@ZenIV.linux.org.uk>
Signed-off-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Currently the iocb argument is used to idenfiy whether or not socket
lock is hold before tipc_sendmsg()/tipc_send_stream() is called. But
this usage prevents iocb argument from being dropped through sendmsg()
at socket common layer. Therefore, in the commit we introduce two new
functions called __tipc_sendmsg() and __tipc_send_stream(). When they
are invoked, it assumes that their callers have taken socket lock,
thereby avoiding the weird usage of iocb argument.
Cc: Al Viro <viro@ZenIV.linux.org.uk>
Cc: Christoph Hellwig <hch@lst.de>
Reviewed-by: Erik Hugne <erik.hugne@ericsson.com>
Reviewed-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The driver was ignoring limits requested by libata.force. The output
would look like:
fsl-sata ffe18000.sata: Sata FSL Platform/CSB Driver init
ata1: FORCE: PHY spd limit set to 1.5Gbps
ata1: SATA max UDMA/133 irq 74
ata1: Signature Update detected @ 0 msecs
ata1: SATA link up 3.0 Gbps (SStatus 123 SControl 310)
Signed-off-by: Martin Hicks <mort@bork.org>
Signed-off-by: Tejun Heo <tj@kernel.org>
The cpuset.sched_relax_domain_level can control how far we do
immediate load balancing on a system. However, it was found on recent
kernels that echo'ing a value into cpuset.sched_relax_domain_level
did not reduce any immediate load balancing.
The reason this occurred was because the update_domain_attr_tree() traversal
did not update for the "top_cpuset". This resulted in nothing being changed
when modifying the sched_relax_domain_level parameter.
This patch is able to address that problem by having update_domain_attr_tree()
allow updates for the root in the cpuset traversal.
Fixes: fc560a26ac ("cpuset: replace cpuset->stack_list with cpuset_for_each_descendant_pre()")
Cc: <stable@vger.kernel.org> # 3.9+
Signed-off-by: Jason Low <jason.low2@hp.com>
Signed-off-by: Zefan Li <lizefan@huawei.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
Tested-by: Serge Hallyn <serge.hallyn@canonical.com>
If clone_children is enabled, effective masks won't be initialized
due to the bug:
# mount -t cgroup -o cpuset xxx /mnt
# echo 1 > cgroup.clone_children
# mkdir /mnt/tmp
# cat /mnt/tmp/
# cat cpuset.effective_cpus
# cat cpuset.cpus
0-15
And then this cpuset won't constrain the tasks in it.
Either the bug or the fix has no effect on unified hierarchy, as
there's no clone_chidren flag there any more.
Reported-by: Christian Brauner <christianvanbrauner@gmail.com>
Reported-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Cc: <stable@vger.kernel.org> # 3.17+
Signed-off-by: Zefan Li <lizefan@huawei.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
Tested-by: Serge Hallyn <serge.hallyn@canonical.com>
During system reboot, the sh-dma-engine device may be runtime-suspended,
causing a crash:
Unhandled fault: imprecise external abort (0x1406) at 0x0002c02c
Internal error: : 1406 [#1] SMP ARM
...
PC is at sh_dmae_ctl_stop+0x28/0x64
LR is at sh_dmae_ctl_stop+0x24/0x64
If the sh-dma-engine is runtime-suspended, its module clock is turned
off, and its registers cannot be accessed.
To fix this, move the call to sh_dmae_ctl_stop(), which touches the
DMAOR register, to the sh_dmae_suspend() and sh_dmae_runtime_suspend()
callbacks. This makes PM operations more symmetric, as both
sh_dmae_resume() and sh_dmae_runtime_resume() already call sh_dmae_rst()
to re-initialize the DMAOR register.
Remove sh_dmae_shutdown(), as it became empty.
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Atmel based boards can now only be used with device tree. Drop non DT
initialization.
Signed-off-by: Alexandre Belloni <alexandre.belloni@free-electrons.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
It doesn't operate on PCI core, but PCI host device, so there is no
point of passing core related struct.
Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Bringing PCIe hosted bus up requires operating on host-related core.
Since we plan to support PCIe Gen 2 devices we should provide a helper
picking the correct one (PCIE or PCIE2).
Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Starting with kernel 3.19 reason is provided by cfg80211.
Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
Acked-by: Arend van Spriel <arend@broadcom.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
and a type bug that could lead to integer overflow - Ivan Khoronzhuk
* Fix boundary checking in efi_high_alloc() which can lead to memory
corruption in the EFI boot stubs - Yinghai Lu
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1
iQIcBAABAgAGBQJU9FtlAAoJEC84WcCNIz1VjfsP/jnZPtkSapSsFP9c7AfV/vpg
i4PLGk+18QhXpNrCVC1U4sdx3y+zefqImrDNEv72BLX6YDb10RvtydxEy4Kg2aaE
XzCRinHWu3+IEwv4fKAmNj2HORTl+jn79JDZ97jm1PN5sOxVcRG9e3QBg6aTVhHr
MdTXRMAKHYD+ZX5hrCMrbFXi1dboxVsUb1zwMTbJcmPSVPWToqNKCruSwp29LNfP
/2ZsJJSHgFP3tobk37JHDTHxjXaN/GUIwQC9cIWUQMPiwU3+WeOvROBPeKUTFNv7
kS4CXY5Q6eKz+pWYqG+FhbfHM71GTWPyFEJNeLtALg2DSKbgL6lJbtkrPpBVXrcU
TeHlHnYTlqEpcMqHW3JtrVb0Of0/8X/9YfWjpmdxNcNbbp7KvzTtoBcP8MjGdbIq
CztyB4clFsiyy1bEoGHFTVArzch5nn7sRCL3mYhTNQaeyN6TZc0wMXOFF/JU7N5a
GCn9VO6T396L/7WdzG0B/Uo01xw11OS/R0jZVoDvtGfAregO+NU+yLunTEYaRtkC
prxQ62Bu21EjLKJcdr/toFkEG8sT08XJnGTixRJnJlw+hmsK8WaigBrdpirXT5SV
TDJJNyo6A/drfjcPoTI4lCR1CpPV3QXjCTmhh+K6tbvX5/npuWN/i4KJh54WuwT4
BKouS5gjrgYcHH/XJjsQ
=GJnM
-----END PGP SIGNATURE-----
Merge tag 'efi-urgent' of git://git.kernel.org/pub/scm/linux/kernel/git/mfleming/efi into x86/urgent
Pull EFI fixes from Matt Fleming:
" - Fix regression in DMI sysfs code for handling "End of Table" entry
and a type bug that could lead to integer overflow. (Ivan Khoronzhuk)
- Fix boundary checking in efi_high_alloc() which can lead to memory
corruption in the EFI boot stubs. (Yinghai Lu)"
Signed-off-by: Ingo Molnar <mingo@kernel.org>
This patch adds support to arptables extensions from nft_compat.
Signed-off-by: Arturo Borrero Gonzalez <arturo.borrero.glez@gmail.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
This patch adds entry for SAMSUNG THERMAL DRIVER in the MAINTAINERS file.
It has been agreed, that pull request are going to be sent to Eduardo
Valentin.
Signed-off-by: Lukasz Majewski <l.majewski@samsung.com>
Acked-by: Eduardo Valentin <edubezval@gmail.com>
Commit: e725d26c48 provided possibility to
use device tree to asses if cpu can be used as cooling device. Since the
code was somewhat awkward, simpler approach has been proposed.
Test HW: Exynos 4412 - Odroid U3.
Suggested-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Lukasz Majewski <l.majewski@samsung.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
This patch fixes the wrong control of PD_DET_EN (power down detection mode)
for Exynos7 because exynos7_tmu_control() always enables the power down detection
mode regardless 'on' parameter.
Cc: Zhang Rui <rui.zhang@intel.com>
Cc: Eduardo Valentin <edubezval@gmail.com>
Signed-off-by: Chanwoo Choi <cw00.choi@samsung.com>
Tested-by: Abhilash Kesavan <a.kesavan@samsung.com>
Signed-off-by: Lukasz Majewski <l.majewski@samsung.com>
The ch341_set_baudrate() function initialize the device baud speed
according to the value on priv->baud_rate. By default the ch341_open() set
it to a hardcoded value (DEFAULT_BAUD_RATE 9600). Unfortunately, the
tty_struct is not initialized with the same default value. (usually 56700)
This means that the tty_struct and the device baud rate generator are not
synchronized after opening the port.
Fixup is done by calling ch341_set_termios() if tty exist.
Remove unnecessary variable priv->baud_rate setup as it's already done by
ch341_port_probe().
Remove unnecessary call to ch341_set_{handshake,baudrate}() in
ch341_open() as there already called in ch341_configure() and
ch341_set_termios()
Signed-off-by: Nicolas PLANEL <nicolas.planel@enovance.com>
Signed-off-by: Johan Hovold <johan@kernel.org>
Currently scan params structure has only active or passive dwell time
fields, passive one is used for fragmented scans too. FW needs the
passive dwell time even when performing fragmented scan for calculating
time between channels. Add a separate parameter for fragmented dwell time
and pass both fragmented and passive to FW.
Signed-off-by: David Spinadel <david.spinadel@intel.com>
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Lots of issues can be caught when the RSSI drops. Add the
ability to collect the firmware data at that point.
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
It can be very useful to monitor the statistics and trigger
a firmware dump when a certain value hits a certain offset.
Since the statistics are huge, add a generic trigger. When
the DWORD at offset X reaches value Y.
Since there is another trigger before this one I can't add
right now because of a dependency on mac80211, add a
reserved entry to keep the enum in place.
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Sometimes the firmware will have a hard coded configuration.
In this case, the driver won't find any configuration
in the firmware file, and it will have to re-start
recording in case it has been stopped. This can't be done
by the configuration host command since there is no such
host command configured. Do that with the registers instead.
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
This will allow to collect the data as soon the firmware
sends a specific notification of command response.
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
We fire the trigger when the channel switch starts, but
the delay is configurable. That makes is easier to catch
channel switches that fail.
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Missing beacons is a good indication that something is going
wrong in the firmware. Add a trigger to be able to collect
data when we start missing beacons with a configurable
threshold.
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Now that the firmware dump can be triggered by events in
the code and not only the user or an firmware ASSERT, we
need a way to know why the firmware dump was triggered.
Add a section in the dump file for that.
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Most of the time, the issues we want to debug with the
firmware dump mechanism are transient. It is then very
hard to stop the recording on time and get meaningful
data.
In order to solve this, I add here an infrastucture
of triggers. The user will supply a list of triggers
that will start / stop the recording. We have two types
of triggers: start and stop. Start triggers can start a
specific configuration. The stop triggers will be able to
kick the collection of the data with the currently running
configuration. These triggers are given to the driver by
the .ucode file - just like the configuration.
In the next patches, I'll add triggers in the code.
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
20 ms fragments are no longer required by system.
Signed-off-by: David Spinadel <david.spinadel@intel.com>
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Packet Level Co-Running is a BT Coex feature which is
supported on certain devices only, hence the need for
a TLV flag for it.
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
We shouldn't call iwl_mvm_d3_disconnect_iter() on the running
interfaces when we are woken up due to net-detect, because it doesn't
make sense. Additionally, this seems to set the
IEEE80211_SDATA_DISCONNECT_RESUME flag that will cause a disconnection
on the next resume (if a normal WoWLAN is used).
To solve this, skip the iteration loop when net-detect is set.
Signed-off-by: Luciano Coelho <luciano.coelho@intel.com>
Reported-by: Samuel Tan <samueltan@chromium.org>
Reviewed-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Eyal Birger says:
====================
net: move skb->dropcount to skb->cb[]
Commit 977750076d ("af_packet: add interframe drop cmsg (v6)")
unionized skb->mark and skb->dropcount in order to allow recording
of the socket drop count while maintaining struct sk_buff size.
skb->dropcount was introduced since there was no available room
in skb->cb[] in packet sockets. However, its introduction led to
the inability to export skb->mark to userspace.
It was considered to alias skb->priority instead of skb->mark.
However, that would lead to the inabilty to export skb->priority
to userspace if desired. Such change may also lead to hard-to-find
issues as skb->priority is assumed to be alias free, and, as noted
by Shmulik Ladkani, is not 'naturally orthogonal' with other skb
fields.
This patch series follows the suggestions made by Eric Dumazet moving
the dropcount metric to skb->cb[], eliminating this problem
at the expense of 4 bytes less in skb->cb[] for protocol families
using it.
The patch series include compactization of bluetooth and packet
use of skb->cb[] as well as the infrastructure for placing dropcount
in skb->cb[].
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Commit 977750076d ("af_packet: add interframe drop cmsg (v6)")
unionized skb->mark and skb->dropcount in order to allow recording
of the socket drop count while maintaining struct sk_buff size.
skb->dropcount was introduced since there was no available room
in skb->cb[] in packet sockets. However, its introduction led to
the inability to export skb->mark, or any other aliased field to
userspace if so desired.
Moving the dropcount metric to skb->cb[] eliminates this problem
at the expense of 4 bytes less in skb->cb[] for protocol families
using it.
Signed-off-by: Eyal Birger <eyal.birger@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
As part of an effort to move skb->dropcount to skb->cb[], use
a common function in order to set dropcount in struct sk_buff.
Signed-off-by: Eyal Birger <eyal.birger@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
As part of an effort to move skb->dropcount to skb->cb[] use a common
macro in protocol families using skb->cb[] for ancillary data to
validate available room in skb->cb[].
Signed-off-by: Eyal Birger <eyal.birger@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
As part of an effort to move skb->dropcount to skb->cb[], 4 bytes
of additional room are needed in skb->cb[] in packet sockets.
Store the skb original length in the first two fields of sockaddr_ll
(sll_family and sll_protocol) as they can be derived from the skb when
needed.
Signed-off-by: Eyal Birger <eyal.birger@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Commit 3b885787ea ("net: Generalize socket rx gap / receive queue overflow cmsg")
allowed receiving packet dropcount information as a socket level option.
RXRPC sockets recvmsg function was changed to support this by calling
sock_recv_ts_and_drops() instead of sock_recv_timestamp().
However, protocol families wishing to receive dropcount should call
sock_queue_rcv_skb() or set the dropcount specifically (as done
in packet_rcv()). This was not done for rxrpc and thus this feature
never worked on these sockets.
Formalizing this by not calling sock_recv_ts_and_drops() in rxrpc as
part of an effort to move skb->dropcount into skb->cb[]
Signed-off-by: Eyal Birger <eyal.birger@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Convert boolean fields incoming and req_start to bit fields and move
force_active in order save space in bt_skb_cb in an effort to use
a portion of skb->cb[] for storing skb->dropcount.
Signed-off-by: Eyal Birger <eyal.birger@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
struct hci_req_ctrl is never used outside of struct bt_skb_cb;
Inlining it frees 8 bytes on a 64 bit system in skb->cb[] allowing
the addition of more ancillary data.
Signed-off-by: Eyal Birger <eyal.birger@gmail.com>
Reviewed-by: Shmulik Ladkani <shmulik.ladkani@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When a PADT frame is received, the socket may not be in a good state to
close down the PPP interface. The current implementation handles this by
simply blocking all further PPP traffic, and hoping that the lack of traffic
will trigger the user to investigate.
Use schedule_work to get to a process context from which we clear down the
PPP interface, in a fashion analogous to hangup on a TTY-based PPP
interface. This causes pppd to disconnect immediately, and allows tools to
take immediate corrective action.
Note that pppd's rp_pppoe.so plugin has code in it to disable the session
when it disconnects; however, as a consequence of this patch, the session is
already disabled before rp_pppoe.so is asked to disable the session. The
result is a harmless error message:
Failed to disconnect PPPoE socket: 114 Operation already in progress
This message is safe to ignore, as long as the error is 114 Operation
already in progress; in that specific case, it means that the PPPoE session
has already been disabled before pppd tried to disable it.
Signed-off-by: Simon Farnsworth <simon@farnz.org.uk>
Tested-by: Dan Williams <dcbw@redhat.com>
Tested-by: Christoph Schulz <develop@kristov.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
The bnx2 driver uses .ndo_fix_features to force enable of Rx VLAN tag
stripping when the card cannot disable it. The driver should remove
NETIF_F_HW_VLAN_CTAG_RX flag from hw_features instead so it is fixed
for the ethtool.
Cc: Sony Chacko <sony.chacko@qlogic.com>
Cc: Dept-HSGLinuxNICDev@qlogic.com
Signed-off-by: Ivan Vecera <ivecera@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
If the server does not return a valid set of attributes that we can
use to either create a file or refresh the inode, then there is no
value in calling nfs_prime_dcache().
However if we're just refreshing the inode using the attributes that
the server returned, then it shouldn't matter whether or not we have
a filehandle, as long as we check the fsid+fileid combination.
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
