Vivien Didelot [Fri, 6 Jan 2017 21:42:00 +0000 (16:42 -0500)]
net: dsa: move HWMON support to its own file
Isolate the HWMON support in DSA in its own file. Currently only the
legacy DSA code is concerned.
Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.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>
David S. Miller [Sun, 8 Jan 2017 02:03:51 +0000 (21:03 -0500)]
Merge branch 'netcp-next'
Murali Karicheri says:
====================
netcp: enhancements and minor fixes
This series is for net-next. This propagates enhancements and minor
bug fixes from internal version of the driver to keep the upstream
in sync. Please review and apply if this looks good.
Tested on all of K2HK/E/L boards with nfs rootfs.
Test logs below
K2HK-EVM: http://pastebin.ubuntu.com/23754106/
k2L-EVM: http://pastebin.ubuntu.com/23754143/
K2E-EVM: http://pastebin.ubuntu.com/23754159/
History:
v1 - dropped 1/10 amd 2/10 of v0 based on comments from Rob as
it needs more work before submission
v0 - Initial version
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
net: netcp: ale: add proper ale entry mask bits for netcp switch ALE
For NetCP NU Switch ALE, some of the mask bits are different than
defaults used in the driver. Add a new macro DEFINE_ALE_FIELD1 that use
a configurable mask bits and use it in the driver. These bits are set to
correct values by using the new variables added to cpsw_ale structure
and re-used in the macros. The parameter nu_switch_ale is configured by
the caller driver to indicate the ALE is for that switch and is used in
the ALE driver to do customization as needed.
Signed-off-by: Murali Karicheri <m-karicheri2@ti.com> Signed-off-by: Sekhar Nori <nsekhar@ti.com> Signed-off-by: David S. Miller <davem@davemloft.net>
net: netcp: ale: use ale_status to size the ale table
ALE h/w on newer version of NetCP (K2E/L/G) does provide a ALE_STATUS
register for the size of the ALE Table implemented in h/w. Currently
for example we set ALE Table size to 1024 for NetCP ALE on
K2E even though the ALE Status/Documentation shows it has 8192 entries.
So take advantage of this register to read the size of ALE table supported
and use that value in the driver for the newer version of NetCP ALE.
For NetCP lite, ALE Table size is much less (64) and indicated by a size
of zero in ALE_STATUS. So use that as a default for now. While at it,
also fix the ale table size on 10G switch to 2048 per User guide
http://www.ti.com/lit/ug/spruhj5/spruhj5.pdf
Signed-off-by: Murali Karicheri <m-karicheri2@ti.com> Signed-off-by: Sekhar Nori <nsekhar@ti.com> Signed-off-by: David S. Miller <davem@davemloft.net>
net: netcp: ale: update to support unknown vlan controls for NU switch
In NU Ethernet switch used on some of the Keystone SoCs, there is
separate UNKNOWNVLAN register for membership, unreg mcast flood, reg
mcast flood and force untag egress bits in ALE. So control for these
fields require different address offset, shift and size of field.
As this ALE has the same version number as ALE in CPSW found on other
SoCs, customization based on version number is not possible. So
use a configuration parameter, nu_switch_ale, to identify the ALE
ALE found in NU Switch. Different treatment is needed for NU Switch
ALE due to difference in the ale table bits, separate unknown vlan
registers etc. The register information available in ale_controls,
needs to be updated to support the netcp NU switch h/w. So it is not
constant array any more since it needs to be updated based
on ALE type. The header of the file is also updated to indicate it
supports N port switch ALE, not just 3 port. The version mask is
3 bits in NU Switch ALE vs 8 bits on other ALE types.
While at it, change the debug print to info print so that ALE
version gets displayed in boot log.
Signed-off-by: Murali Karicheri <m-karicheri2@ti.com> Signed-off-by: Sekhar Nori <nsekhar@ti.com> Signed-off-by: David S. Miller <davem@davemloft.net>
net: netcp: use hw capability to remove FCS word from rx packets
Some of the newer Ethernet switch hw (such as that on k2e/l/g) can
strip the Etherenet FCS from packet at the port 0 egress of the switch.
So use this capability instead of doing it in software.
Signed-off-by: Murali Karicheri <m-karicheri2@ti.com> Signed-off-by: Sekhar Nori <nsekhar@ti.com> Signed-off-by: David S. Miller <davem@davemloft.net>
net: netcp: ethss: get phy-handle only if link interface is MAC-to-PHY
Currently to parse phy-handle, driver doesn't check if the interface is
MAC to PHY. This patch add this check for all MAC to PHY interface types
supported by the driver.
Signed-off-by: Murali Karicheri <m-karicheri2@ti.com> Signed-off-by: Sekhar Nori <nsekhar@ti.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Previously the network statistics were stored in 32 bit variable
which can cause some stats to roll over after several minutes of
high traffic. This implements 64 bit storage so larger numbers
can be stored.
Signed-off-by: Michael Scherban <m-scherban@ti.com> Signed-off-by: Murali Karicheri <m-karicheri2@ti.com> Signed-off-by: Sekhar Nori <nsekhar@ti.com> Signed-off-by: David S. Miller <davem@davemloft.net>
The psdata is populated with command data by netcp modules
to the tail of the buffer and set_words() copy the same
to the front of the psdata. So remove the redundant memmov
function call.
Signed-off-by: Murali Karicheri <m-karicheri2@ti.com> Signed-off-by: David S. Miller <davem@davemloft.net>
net: netcp: extract eflag from desc for rx_hook handling
Extract the eflag bits from the received desc and pass it down
the rx_hook chain to be available for netcp modules. Also the
psdata and epib data has to be inspected by the netcp modules.
So the desc can be freed only after returning from the rx_hook.
So move knav_pool_desc_put() after the rx_hook processing.
Signed-off-by: Murali Karicheri <m-karicheri2@ti.com> Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Sun, 8 Jan 2017 01:48:16 +0000 (20:48 -0500)]
Merge branch 'cpsw-cpdma-DDR'
Grygorii Strashko says:
====================
net: ethernet: ti: cpsw: support placing CPDMA descriptors into DDR
This series intended to add support for placing CPDMA descriptors into
DDR by introducing new module parameter "descs_pool_size" to specify
size of descriptor's pool. The "descs_pool_size" defines total number
of CPDMA CPPI descriptors to be used for both ingress/egress packets
processing. If not specified - the default value 256 will be used
which will allow to place descriptor's pool into the internal CPPI
RAM.
In addition, added ability to re-split CPDMA pool of descriptors
between RX and TX path via ethtool '-G' command wich will allow to
configure and fix number of descriptors used by RX and TX path, which,
then, will be split between RX/TX channels proportionally depending on
number of RX/TX channels and its weight.
This allows significantly to reduce UDP packets drop rate for
bandwidth >301 Mbits/sec (am57x).
Before enabling this feature, the am437x SoC has to be fixed as it's
proved that it's not working when CPDMA descriptors placed in DDR.
So, the patch 1 fixes this issue.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Even if no_bd_ram property is described in TI CPSW bindings the support for
it has never been introduced in CPSW driver, so there are no real users of
it. Hence, remove no_bd_ram property from documentation and DT files.
Cc: 'Rob Herring <robh+dt@kernel.org>' Signed-off-by: Grygorii Strashko <grygorii.strashko@ti.com> Signed-off-by: David S. Miller <davem@davemloft.net>
net: ethernet: ti: cpsw: add support for ringparam configuration
The CPDMA uses one pool of descriptors for both RX and TX which by default
split between all channels proportionally depending on total number of
CPDMA channels and number of TX and RX channels. As result, more
descriptors will be consumed by TX path if there are more TX channels and
there is no way now to dedicate more descriptors for RX path.
So, add the ability to re-split CPDMA pool of descriptors between RX and TX
path via ethtool '-G' command wich will allow to configure and fix number
of descriptors used by RX and TX path, which, then, will be split between
RX/TX channels proportionally depending on RX/TX channels number and
weight. ethtool '-G' command will accept only number of RX entries and rest
of descriptors will be arranged for TX automatically.
Command:
ethtool -G <devname> rx <number of descriptors>
defaults and limitations:
- minimum number of rx descriptors is 10% of total number of descriptors in
CPDMA pool
- maximum number of rx descriptors is 90% of total number of descriptors in
CPDMA pool
- by default, descriptors will be split equally between RX/TX path
- any values passed in "tx" parameter will be ignored
net: ethernet: ti: cpsw: add support for descs pool size configuration
The CPSW CPDMA can process buffer descriptors placed as in internal
CPPI RAM as in DDR. This patch adds support in CPSW and CPDMA for
descs_pool_size mudule parameter, which defines total number of CPDMA CPPI
descriptors to be used for both ingress/egress packets processing:
- memory size, required for CPDMA descriptor pool, is calculated basing
on number of descriptors specified by user in descs_pool_size and
CPDMA descriptor size and allocated from coherent memory (CMA area);
- CPDMA descriptor pool will be allocated in DDR if pool memory size >
internal CPPI RAM or use internal CPPI RAM otherwise;
- if descs_pool_size not specified in DT - the default value 256 will
be used which will allow to place CPDMA descriptors pool into the
internal CPPI RAM (current default behaviour);
- CPDMA will ignore descs_pool_size if descs_pool_size = 0 for
backward comaptiobility with davinci_emac.
descs_pool_size is boot time setting and can't be changed once
CPSW/CPDMA is initialized.
Signed-off-by: Grygorii Strashko <grygorii.strashko@ti.com> Signed-off-by: David S. Miller <davem@davemloft.net>
net: ethernet: ti: cpdma: minimize number of parameters in cpdma_desc_pool_create/destroy()
Update cpdma_desc_pool_create/destroy() to accept only one parameter
struct cpdma_ctlr*, as this structure contains all required
information for pool creation/destruction.
Signed-off-by: Grygorii Strashko <grygorii.strashko@ti.com> Signed-off-by: David S. Miller <davem@davemloft.net>
The currently processing cpdma descriptor with EOQ flag set may
contain two values in Next Descriptor Pointer field:
- valid pointer: means CPDMA missed addition of new desc in queue;
- null: no more descriptors in queue.
In the later case, it's not required to write to HDP register, but now
CPDMA does it.
Hence, add additional check for Next Descriptor Pointer != null in
cpdma_chan_process() function before writing in HDP register.
Signed-off-by: Grygorii Strashko <grygorii.strashko@ti.com> Signed-off-by: David S. Miller <davem@davemloft.net>
net: ethernet: ti: cpdma: am437x: allow descs to be plased in ddr
It's observed that cpsw/cpdma is not working properly when CPPI
descriptors are placed in DDR instead of internal CPPI RAM on am437x
SoC:
- rx/tx silently stops processing packets;
- or - after boot it's working for sometime, but stuck once Network
load is increased (ping is working, but iperf is not).
(The same issue has not been reproduced on am335x and am57xx).
It seems that write to HDP register processed faster by interconnect
than writing of descriptor memory buffer in DDR, which is probably
caused by store buffer / write buffer differences as these functions
are implemented differently across devices. So, to fix this i come up
with two minimal, required changes:
1) all accesses to the channel register HDP/CP/RXFREE registers should
be done using sync IO accessors readl()/writel(), because all previous
memory writes writes have to be completed before starting channel
(write to HDP) or completing desc processing.
2) the change 1 only doesn't work on am437x and additional reading of
desc's field is required right after the new descriptor was filled
with data and before pointer on it will be stored in
prev_desc->hw_next field or HDP register.
In addition, to above changes this patch eliminates all relaxed ordering
I/O accessors in this driver as suggested by David Miller to avoid such
kind of issues in the future, but with one exception - relaxed IO accessors
will still be used to fill desc in cpdma_chan_submit(), which is safe as
there is read barrier at the end of write sequence, and because sync IO
accessors usage here will affect on net performance.
Signed-off-by: Grygorii Strashko <grygorii.strashko@ti.com> Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Sat, 7 Jan 2017 03:18:57 +0000 (22:18 -0500)]
Merge branch 'l2tp-cleanup-socket-lookup-code'
Guillaume Nault says:
====================
l2tp: cleanup socket lookup code in l2tp_ip and l2tp_ip6
First three patches remove redundant tests and add missing "const"
qualifiers.
Fourth patch splits the conditionals found in __l2tp_ip*_bind_lookup(),
to make these functions easier to review. In the process, I found that
some corner cases were still not handled properly. So I've added the
missing tests in this patch too, because they're pretty simple and the
whole "if" statements are modified anyway.
I expect it to be easier to review this way. If not, I can split up
patch #4, post the missing tests separately to -net, and later repost
this series as pure cleanup. Just let me know.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Guillaume Nault [Fri, 6 Jan 2017 19:03:57 +0000 (20:03 +0100)]
l2tp: rework socket comparison in __l2tp_ip*_bind_lookup()
Split conditions, so that each test becomes clearer.
Also, for l2tp_ip, check if "laddr" is 0. This prevents a socket from
binding to the unspecified address when other sockets are already bound
using the same device (if any), connection ID and namespace.
Same thing for l2tp_ip6: add ipv6_addr_any(laddr) and
ipv6_addr_any(raddr) tests to ensure that an IPv6 unspecified address
passed as parameter is properly treated a wildcard.
Signed-off-by: Guillaume Nault <g.nault@alphalink.fr> Signed-off-by: David S. Miller <davem@davemloft.net>
RDS: validate the requested traces user input against max supported
Larger than supported value can lead to array read/write overflow.
Reported-by: Colin Ian King <colin.king@canonical.com> Signed-off-by: Santosh Shilimkar <santosh.shilimkar@oracle.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Xin Long [Fri, 6 Jan 2017 14:18:33 +0000 (22:18 +0800)]
sctp: prepare asoc stream for stream reconf
sctp stream reconf, described in RFC 6525, needs a structure to
save per stream information in assoc, like stream state.
In the future, sctp stream scheduler also needs it to save some
stream scheduler params and queues.
This patchset is to prepare the stream array in assoc for stream
reconf. It defines sctp_stream that includes stream arrays inside
to replace ssnmap.
Note that we use different structures for IN and OUT streams, as
the members in per OUT stream will get more and more different
from per IN stream.
v1->v2:
- put these patches into a smaller group.
v2->v3:
- define sctp_stream to contain stream arrays, and create stream.c
to put stream-related functions.
- merge 3 patches into 1, as new sctp_stream has the same name
with before.
Signed-off-by: Xin Long <lucien.xin@gmail.com> Reviewed-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Eric Garver [Fri, 6 Jan 2017 01:22:36 +0000 (20:22 -0500)]
udp: inuse checks can quit early for reuseport
UDP lib inuse checks will walk the entire hash bucket to check if the
portaddr is in use. In the case of reuseport we can stop searching when
we find a matching reuseport.
On a 16-core VM a test program that spawns 16 threads that each bind to
1024 sockets (one per 10ms) takes 1m45s. With this change it takes 11s.
Also add a cond_resched() when the port is not specified.
Signed-off-by: Eric Garver <e@erig.me> Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Fri, 6 Jan 2017 21:05:37 +0000 (16:05 -0500)]
Merge branch '1GbE' of git://git.kernel.org/pub/scm/linux/kernel/git/jkirsher/next-queue
Jeff Kirsher says:
====================
1GbE Intel Wired LAN Driver Updates 2017-01-06
This series contains updates/fixes to igb and e1000e.
Joe fixes indentation and improper line wrapping in igb.
David Singleton fixes an issue in e1000e where in systemd, where things
are done in parallel and can create a condition where e1000_shutdown is
called after e1000_close, hitting BUG_ON assert in free_msi_irqs.
Cao Jin fixes a code comment on the wakeup status register. Also fixes
a possible NULL pointer dereference by using igb_adapter->io_addr
instead of e1000_hw->hw_addr in igb_configure_tx_ring().
Chris Arges works around a firmware issue, which can cause probe of i210
NIC to fail, so zero the page select register during igb_get_phy_id() to
workaround the issue. Aaron Sierra adds also a check for this issue
during the initialization of PHY parameters to ensure that this same
issue happens after probe.
Todd fixes a possible race condition in close/suspend by extending
the rtnl_lock() to protect the call to netif_device_detach() and
igb_clear_interrupt_scheme(). Also adds i211 to a known i210/i211
workaround.
Hannu Lounento fixes inverted logic on a debug statement.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
David Ahern [Fri, 6 Jan 2017 03:32:46 +0000 (19:32 -0800)]
net: ipv4: Simplify rt_fill_info
rt_fill_info has only 1 caller and both of the last 2 args -- nowait
and flags -- are hardcoded to 0. Given that remove them as input arguments
and simplify rt_fill_info accordingly.
Signed-off-by: David Ahern <dsa@cumulusnetworks.com> Signed-off-by: David S. Miller <davem@davemloft.net>
The issue comes when there are multiple threads attempting to use
the mailbox facility at the same time.
When DCB operations and interface up/down is run in a loop for every
0.1 sec, we observed mailbox collisions. And out of the two commands
one would fail with the present code, since we don't queue the second
command.
To overcome the above issue, added a queue to access the mailbox.
Whenever a mailbox command is issued add it to the queue. If its at
the head issue the mailbox command, else wait for the existing command
to complete. Usually command takes less than a milli-second to
complete.
Also timeout from the loop, if the command under execution takes
long time to run.
In reality, the number of mailbox access collisions is going to be
very rare since no one runs such abusive script.
Signed-off-by: Hariprasad Shenai <hariprasad@chelsio.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Weilin Chang [Thu, 5 Jan 2017 00:18:50 +0000 (16:18 -0800)]
liquidio: fix wrong information about channels reported to ethtool
Information reported to ethtool about channels is sometimes wrong for PF,
and always wrong for VF. Fix them by getting the information from the
right fields from the right structs.
Signed-off-by: Weilin Chang <weilin.chang@cavium.com> Signed-off-by: Felix Manlunas <felix.manlunas@cavium.com> Signed-off-by: Derek Chickles <derek.chickles@cavium.com> Signed-off-by: Satanand Burla <satananda.burla@cavium.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Mahesh Bandewar [Wed, 4 Jan 2017 23:01:01 +0000 (15:01 -0800)]
ipv6: do not send RTM_DELADDR for tentative addresses
RTM_NEWADDR notification is sent when IFA_F_TENTATIVE is cleared from
the address. So if the address is added and deleted before DAD probes
completes, the RTM_DELADDR will be sent for which there was no
RTM_NEWADDR causing asymmetry in notification. However if the same
logic is used while sending RTM_DELADDR notification, this asymmetry
can be avoided.
Signed-off-by: Mahesh Bandewar <maheshb@google.com> CC: Hideaki YOSHIFUJI <yoshfuji@linux-ipv6.org> CC: Patrick McHardy <kaber@trash.net> CC: Hannes Frederic Sowa <hannes@stressinduktion.org> Signed-off-by: David S. Miller <davem@davemloft.net>
The VF driver is using the wrong struct when sending commands to the NIC
firmware, sometimes causing adverse effects in the firmware. The right
struct is the one that the PF is using, so make the VF use that as well.
Signed-off-by: Prasad Kanneganti <prasad.kanneganti@cavium.com> Signed-off-by: Felix Manlunas <felix.manlunas@cavium.com> Signed-off-by: Derek Chickles <derek.chickles@cavium.com> Signed-off-by: Satanand Burla <satananda.burla@cavium.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Hannu Lounento [Mon, 2 Jan 2017 17:26:06 +0000 (18:26 +0100)]
igb: Fix hw_dbg logging in igb_update_flash_i210
Fix an if statement with hw_dbg lines where the logic was inverted with
regards to the corresponding return value used in the if statement.
Signed-off-by: Hannu Lounento <hannu.lounento@ge.com> Signed-off-by: Peter Senna Tschudin <peter.senna@collabora.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Todd Fujinaka [Tue, 15 Nov 2016 16:54:26 +0000 (08:54 -0800)]
igb: close/suspend race in netif_device_detach
Similar to ixgbe, when an interface is part of a namespace it is
possible that igb_close() may be called while __igb_shutdown() is
running which ends up in a double free WARN and/or a BUG in
free_msi_irqs().
Extend the rtnl_lock() to protect the call to netif_device_detach() and
igb_clear_interrupt_scheme() in __igb_shutdown() and check for
netif_device_present() to avoid calling igb_clear_interrupt_scheme() a
second time in igb_close().
Also extend the rtnl lock in igb_resume() to netif_device_attach().
Signed-off-by: Todd Fujinaka <todd.fujinaka@intel.com> Acked-by: Alexander Duyck <alexander.h.duyck@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
igb: re-assign hw address pointer on reset after PCI error
Whenever the igb driver detects the result of a read operation returns
a value composed only by F's (like 0xFFFFFFFF), it will detach the
net_device, clear the hw_addr pointer and warn to the user that adapter's
link is lost - those steps happen on igb_rd32().
In case a PCI error happens on Power architecture, there's a recovery
mechanism called EEH, that will reset the PCI slot and call driver's
handlers to reset the adapter and network functionality as well.
We observed that once hw_addr is NULL after the error is detected on
igb_rd32(), it's never assigned back, so in the process of resetting
the network functionality we got a NULL pointer dereference in both
igb_configure_tx_ring() and igb_configure_rx_ring(). In order to avoid
such bug, this patch re-assigns the hw_addr value in the slot_reset
handler.
Reported-by: Anthony H Thai <ahthai@us.ibm.com> Reported-by: Harsha Thyagaraja <hathyaga@in.ibm.com> Signed-off-by: Guilherme G Piccoli <gpiccoli@linux.vnet.ibm.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Aaron Sierra [Tue, 29 Nov 2016 16:03:56 +0000 (10:03 -0600)]
igb: reset the PHY before reading the PHY ID
Several people have reported firmware leaving the I210/I211 PHY's page
select register set to something other than the default of zero. This
causes the first accesses, PHY_IDx register reads, to access something
else, resulting in device probe failure:
igb: Intel(R) Gigabit Ethernet Network Driver - version 5.4.0-k
igb: Copyright (c) 2007-2014 Intel Corporation.
igb: probe of 0000:01:00.0 failed with error -2
This problem began for them after a previous patch I submitted was
applied:
I personally experienced this problem after attempting to PXE boot from
I210 devices using this firmware:
Intel(R) Boot Agent GE v1.5.78
Copyright (C) 1997-2014, Intel Corporation
Resetting the PHY before reading from it, ensures the page select
register is in its default state and doesn't make assumptions about
the PHY's register set before the PHY has been probed.
Cc: Matwey V. Kornilov <matwey@sai.msu.ru> Cc: Chris Arges <carges@vectranetworks.com> Cc: Jochen Henneberg <jh@henneberg-systemdesign.com> Signed-off-by: Aaron Sierra <asierra@xes-inc.com> Tested-by: Matwey V. Kornilov <matwey@sai.msu.ru> Tested-by: Chris J Arges <christopherarges@gmail.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Cao jin [Tue, 8 Nov 2016 07:06:20 +0000 (15:06 +0800)]
igb: use igb_adapter->io_addr instead of e1000_hw->hw_addr
When running as guest, under certain condition, it will oops as following.
writel() in igb_configure_tx_ring() results in oops, because hw->hw_addr
is NULL. While other register access won't oops kernel because they use
wr32/rd32 which have a defense against NULL pointer.
A similar issue & solution could be found at:
http://patchwork.ozlabs.org/patch/689592/
Signed-off-by: Cao jin <caoj.fnst@cn.fujitsu.com> Acked-by: Alexander Duyck <alexander.h.duyck@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Chris J Arges [Wed, 2 Nov 2016 14:13:42 +0000 (09:13 -0500)]
igb: Workaround for igb i210 firmware issue
Sometimes firmware may not properly initialize I347AT4_PAGE_SELECT causing
the probe of an igb i210 NIC to fail. This patch adds an addition zeroing
of this register during igb_get_phy_id to workaround this issue.
Thanks for Jochen Henneberg for the idea and original patch.
Signed-off-by: Chris J Arges <christopherarges@gmail.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Cao jin [Wed, 2 Nov 2016 07:20:15 +0000 (15:20 +0800)]
igb: correct register comments
Signed-off-by: Cao jin <caoj.fnst@cn.fujitsu.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
khalidm [Mon, 17 Oct 2016 16:51:08 +0000 (09:51 -0700)]
e1000e: driver trying to free already-free irq
During systemd reboot sequence network driver interface is shutdown
by e1000_close. The PCI driver interface is shut by e1000_shutdown.
The e1000_shutdown checks for netif_running status, if still up it
brings down driver. But it disables msi outside of this if statement,
regardless of netif status. All this is OK when e1000_close happens
after shutdown. However, by default, everything in systemd is done
in parallel. This creates a conditions where e1000_shutdown is called
after e1000_close, therefore hitting BUG_ON assert in free_msi_irqs.
CC: xe-kernel@external.cisco.com Signed-off-by: khalidm <khalidm@cisco.com> Signed-off-by: David Singleton <davsingl@cisco.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Joe Perches [Tue, 27 Sep 2016 03:46:25 +0000 (20:46 -0700)]
igb: Realign bad indentation
Statements should start on tabstops.
Use a single statement and test instead of multiple tests.
Signed-off-by: Joe Perches <joe@perches.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
tools: psock_tpacket: block Rx until socket filter has been added and socket has been bound to loopback.
Packets from any/all interfaces may be queued up on the PF_PACKET socket
before it is bound to the loopback interface by psock_tpacket, and
when these are passed up by the kernel, they could interfere
with the Rx tests.
Avoid interference from spurious packet by blocking Rx until the
socket filter has been set up, and the packet has been bound to the
desired (lo) interface. The effective sequence is
socket(PF_PACKET, SOCK_RAW, 0);
set up ring
Invoke SO_ATTACH_FILTER
bind to sll_protocol set to ETH_P_ALL, sll_ifindex for lo
After this sequence, the only packets that will be passed up are
those received on loopback that pass the attached filter.
Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com> Signed-off-by: David S. Miller <davem@davemloft.net>
For TCP sockets, TX timestamps are only captured when the user data
is successfully and fully written to the socket. In many cases,
however, TCP writes can be partial for which no timestamp is
collected.
Collect timestamps whenever any user data is (fully or partially)
copied into the socket. Pass tcp_write_queue_tail to tcp_tx_timestamp
instead of the local skb pointer since it can be set to NULL on
the error path.
Note that tcp_write_queue_tail can be NULL, even if bytes have been
copied to the socket. This is because acknowledgements are being
processed in tcp_sendmsg(), and by the time tcp_tx_timestamp is
called tcp_write_queue_tail can be NULL. For such cases, this patch
does not collect any timestamps (i.e., it is best-effort).
This patch is written with suggestions from Willem de Bruijn and
Eric Dumazet.
Change-log V1 -> V2:
- Use sockc.tsflags instead of sk->sk_tsflags.
- Use the same code path for normal writes and errors.
Signed-off-by: Soheil Hassas Yeganeh <soheil@google.com> Acked-by: Yuchung Cheng <ycheng@google.com> Cc: Willem de Bruijn <willemb@google.com> Cc: Eric Dumazet <edumazet@google.com> Cc: Neal Cardwell <ncardwell@google.com> Cc: Martin KaFai Lau <kafai@fb.com> Acked-by: Willem de Bruijn <willemb@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Thu, 5 Jan 2017 17:43:28 +0000 (12:43 -0500)]
Merge tag 'rxrpc-rewrite-20170105' of git://git.kernel.org/pub/scm/linux/kernel/git/dhowells/linux-fs
David Howells says:
====================
rxrpc: Update tracing and proc interfaces
This set of patches fixes and extends tracing:
(1) Fix the handling of enum-to-string translations so that external
tracing tools can make use of it by using TRACE_DEFINE_ENUM.
(2) Extend a couple of tracepoints to export some extra available
information and add three new tracepoints to allow monitoring of
received DATA packets, call disconnection and improper/implicit call
termination.
and adds a bit more procfs-exported information:
(3) Show a call's hard-ACK cursors in /proc/net/rxrpc_calls.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
net:dsa: check for EPROBE_DEFER from dsa_dst_parse()
Since there can be multiple dsa switches stacked together but
not all of devicetree nodes available at the time of calling
dsa_dst_parse(), EPROBE_DEFER can be returned by it. When this
happens, only the last dsa switch has to be deleted by
dsa_dst_del_ds(), but not the whole list, because next time linux
cames back to this function it will try to add only the last dsa
switch which returned EPROBE_DEFER.
Signed-off-by: Volodymyr Bendiuga <volodymyr.bendiuga@gmail.com> Reviewed-by: Andrew Lunn <andrew@lunn.ch> Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Volodymyr Bendiuga <volodymyr.bendiuga@gmail.com> Reviewed-by: Andrew Lunn <andrew@lunn.ch> Signed-off-by: David S. Miller <davem@davemloft.net>
xemaclite_remove_ndev() is a simple wrapper around free_netdev()
checking for NULL before the call. All possible paths calling
it are guaranteed to pass a non-NULL argument, so rather call
free_netdev() directly.
Signed-off-by: Tobias Klauser <tklauser@distanz.ch> Signed-off-by: David S. Miller <davem@davemloft.net>
David Howells [Thu, 5 Jan 2017 10:38:34 +0000 (10:38 +0000)]
rxrpc: Add some more tracing
Add the following extra tracing information:
(1) Modify the rxrpc_transmit tracepoint to record the Tx window size as
this is varied by the slow-start algorithm.
(2) Modify the rxrpc_rx_ack tracepoint to record more information from
received ACK packets.
(3) Add an rxrpc_rx_data tracepoint to record the information in DATA
packets.
(4) Add an rxrpc_disconnect_call tracepoint to record call disconnection,
including the reason the call was disconnected.
(5) Add an rxrpc_improper_term tracepoint to record implicit termination
of a call by a client either by starting a new call on a particular
connection channel without first transmitting the final ACK for the
previous call.
Signed-off-by: David Howells <dhowells@redhat.com>
David Howells [Thu, 5 Jan 2017 10:38:33 +0000 (10:38 +0000)]
rxrpc: Fix handling of enums-to-string translation in tracing
Fix the way enum values are translated into strings in AF_RXRPC
tracepoints. The problem with just doing a lookup in a normal flat array
of strings or chars is that external tracing infrastructure can't find it.
Rather, TRACE_DEFINE_ENUM must be used.
Also sort the enums and string tables to make it easier to keep them in
order so that a future patch to __print_symbolic() can be optimised to try
a direct lookup into the table first before iterating over it.
A couple of _proto() macro calls are removed because they refered to tables
that got moved to the tracing infrastructure. The relevant data can be
found by way of tracing.
Signed-off-by: David Howells <dhowells@redhat.com>
Daniel Borkmann [Thu, 5 Jan 2017 01:34:28 +0000 (02:34 +0100)]
packet: fix panic in __packet_set_timestamp on tpacket_v3 in tx mode
When TX timestamping is in use with TPACKET_V3's TX ring, then we'll
hit the BUG() in __packet_set_timestamp() when ring buffer slot is
returned to user space via tpacket_destruct_skb(). This is due to v3
being assumed as unreachable here, but since 7f953ab2ba46 ("af_packet:
TX_RING support for TPACKET_V3") it's not anymore. Fix it by filling
the timestamp back into the ring slot.
Fixes: 7f953ab2ba46 ("af_packet: TX_RING support for TPACKET_V3") Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Acked-by: Sowmini Varadhan <sowmini.varadhan@oracle.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Linus Torvalds [Thu, 5 Jan 2017 02:33:35 +0000 (18:33 -0800)]
Merge tag 'xfs-for-linus-4.10-rc3' of git://git.kernel.org/pub/scm/fs/xfs/xfs-linux
Pull xfs fixes from Darrick Wong:
- fixes for crashes and double-cleanup errors
- XFS maintainership handover
- fix to prevent absurdly large block reservations
- fix broken sysfs getter/setters
* tag 'xfs-for-linus-4.10-rc3' of git://git.kernel.org/pub/scm/fs/xfs/xfs-linux:
xfs: fix max_retries _show and _store functions
xfs: update MAINTAINERS
xfs: fix crash and data corruption due to removal of busy COW extents
xfs: use the actual AG length when reserving blocks
xfs: fix double-cleanup when CUI recovery fails
1) stmmac_drv_probe() can race with stmmac_open() because we register
the netdevice too early. Fix from Florian Fainelli.
2) UFO handling in __ip6_append_data() and ip6_finish_output() use
different tests for deciding whether a frame will be fragmented or
not, put them in sync. Fix from Zheng Li.
3) The rtnetlink getstats handlers need to validate that the netlink
request is large enough, fix from Mathias Krause.
4) Use after free in mlx4 driver, from Jack Morgenstein.
5) Fix setting of garbage UID value in sockets during setattr() calls,
from Eric Biggers.
6) Packet drop_monitor doesn't format the netlink messages properly
such that nlmsg_next fails to work, fix from Reiter Wolfgang.
7) Fix handling of wildcard addresses in l2tp lookups, from Guillaume
Nault.
8) __skb_flow_dissect() can crash on pptp packets, from Ian Kumlien.
9) IGMP code doesn't reset group query timers properly, from Michal
Tesar.
10) Fix overzealous MAIN/LOCAL route table combining in ipv4, from
Alexander Duyck.
11) vxlan offload check needs to be more strict in be2net driver, from
Sabrina Dubroca.
12) Moving l3mdev to packet hooks lost RX stat counters unintentionally,
fix from David Ahern.
* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net: (52 commits)
sh_eth: enable RX descriptor word 0 shift on SH7734
sfc: don't report RX hash keys to ethtool when RSS wasn't enabled
dpaa_eth: Initialize CGR structure before init
dpaa_eth: cleanup after init_phy() failure
net: systemport: Pad packet before inserting TSB
net: systemport: Utilize skb_put_padto()
LiquidIO VF: s/select/imply/ for PTP_1588_CLOCK
libcxgb: fix error check for ip6_route_output()
net: usb: asix_devices: add .reset_resume for USB PM
net: vrf: Add missing Rx counters
drop_monitor: consider inserted data in genlmsg_end
benet: stricter vxlan offloading check in be_features_check
ipv4: Do not allow MAIN to be alias for new LOCAL w/ custom rules
net: macb: Updated resource allocation function calls to new version of API.
net: stmmac: dwmac-oxnas: use generic pm implementation
net: stmmac: dwmac-oxnas: fix fixed-link-phydev leaks
net: stmmac: dwmac-oxnas: fix of-node leak
Documentation/networking: fix typo in mpls-sysctl
igmp: Make igmp group member RFC 3376 compliant
flow_dissector: Update pptp handling to avoid null pointer deref.
...
Andrew Lunn [Wed, 4 Jan 2017 18:56:24 +0000 (19:56 +0100)]
dsa: mv88e6xxx: Optimise atu_get
Lookup in the ATU can be performed starting from a given MAC
address. This is faster than starting with the first possible MAC
address and iterating all entries.
Entries are returned in numeric order. So if the MAC address returned
is bigger than what we are searching for, we know it is not in the
ATU.
Using the benchmark provided by Volodymyr Bendiuga
<volodymyr.bendiuga@gmail.com>,
Sergei Shtylyov [Wed, 4 Jan 2017 20:10:23 +0000 (23:10 +0300)]
sh_eth: enable RX descriptor word 0 shift on SH7734
The RX descriptor word 0 on SH7734 has the RFS[9:0] field in bits 16-25
(bits 0-15 usually used for that are occupied by the packet checksum).
Thus we need to set the 'shift_rd0' field in the SH7734 SoC data...
Fixes: f0e81fecd4f8 ("net: sh_eth: Add support SH7734") Signed-off-by: Sergei Shtylyov <sergei.shtylyov@cogentembedded.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Edward Cree [Wed, 4 Jan 2017 15:10:56 +0000 (15:10 +0000)]
sfc: don't report RX hash keys to ethtool when RSS wasn't enabled
If we failed to set up RSS on EF10 (e.g. because firmware declared
RX_RSS_LIMITED), ethtool --show-nfc $dev rx-flow-hash ... should report
no fields, rather than confusingly reporting what fields we _would_ be
hashing on if RSS was working.
Fixes: dcb4123cbec0 ("sfc: disable RSS when unsupported") Signed-off-by: Edward Cree <ecree@solarflare.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Here's a set of 3 patches against DaveM's 'net-next.git' repo. I'm cleaning
up the E-MAC interrupt handling with the main goal of factoring out the E-MAC
interrupt handler into a separate function.
[1/3] sh_eth: handle only enabled E-MAC interrupts
[2/3] sh_eth: no need for *else* after *goto*
[3/3] sh_eth: factor out sh_eth_emac_interrupt()
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Sergei Shtylyov [Wed, 4 Jan 2017 12:11:21 +0000 (15:11 +0300)]
sh_eth: factor out sh_eth_emac_interrupt()
The E-MAC interrupt (EESR.ECI) is not always caused by an error condition,
so it really shouldn't be handled by sh_eth_error(). Factor out the E-MAC
interrupt handler, sh_eth_emac_interrupt(), removing the ECI bit from the
EESR's values throughout the driver...
Update Cogent Embedded's copyright and clean up the whitespace in Renesas'
copyright, while at it...
Signed-off-by: Sergei Shtylyov <sergei.shtylyov@cogentembedded.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Sergei Shtylyov [Wed, 4 Jan 2017 12:10:50 +0000 (15:10 +0300)]
sh_eth: no need for *else* after *goto*
Well, checkpatch.pl complains about *else* after *return* and *break* but
not after *goto*... and it probably should have complained about the code
in sh_eth_error(). Win couple LoCs by removing that *else*. :-)
Signed-off-by: Sergei Shtylyov <sergei.shtylyov@cogentembedded.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Sergei Shtylyov [Wed, 4 Jan 2017 12:10:21 +0000 (15:10 +0300)]
sh_eth: handle only enabled E-MAC interrupts
The driver should only handle the enabled E-MAC interrupts, like it does
for the E-DMAC interrupts since commit 3893b27345ac ("sh_eth: workaround
for spurious ECI interrupt"), so mask ECSR with ECSIPR when reading it
in sh_eth_error().
Signed-off-by: Sergei Shtylyov <sergei.shtylyov@cogentembedded.com> Signed-off-by: David S. Miller <davem@davemloft.net>
====================
net: systemport: Fix padding vs. TSB insertion
This patch series fixes how we pad the packets submitted to the SYSTEMPORT
adapter, and how the transmit status block (prepended 8 bytes) fits in the
picture. The first patch is not technically a bug fix, but is required for the
second path to be applied and to greatly simplify the skb length calculation.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Inserting the TSB means adding an extra 8 bytes in front the of packet
that is going to be used as metadata information by the TDMA engine, but
stripped off, so it does not really help with the packet padding.
For some odd packet sizes that fall below the 60 bytes payload (e.g: ARP)
we can end-up padding them after the TSB insertion, thus making them 64
bytes, but with the TDMA stripping off the first 8 bytes, they could
still be smaller than 64 bytes which is required to ingress the switch.
Fix this by swapping the padding and TSB insertion, guaranteeing that
the packets have the right sizes.
Fixes: 80105befdb4b ("net: systemport: add Broadcom SYSTEMPORT Ethernet MAC driver") Signed-off-by: Florian Fainelli <f.fainelli@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Since we need to pad our packets, utilize skb_put_padto() which
increases skb->len by how much we need to pad, allowing us to eliminate
the test on skb->len right below.
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Mahesh Bandewar [Tue, 3 Jan 2017 20:47:16 +0000 (12:47 -0800)]
ipvlan: assign unique dev-id for each slave device.
IPvlan setup uses one mac-address (of master). The IPv6 link-local
addresses are derived using the mac-address on the link. Lack of
dev-ids makes these link-local addresses same for all slaves including
that of master device. dev-ids are necessary to add differentiation
when L2 address is shared.
Signed-off-by: Mahesh Bandewar <maheshb@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Vivien Didelot [Tue, 3 Jan 2017 19:31:49 +0000 (14:31 -0500)]
net: dsa: remove out label in dsa_switch_setup_one
The "out" label in dsa_switch_setup_one() is useless, thus remove it.
Reviewed-by: Andrew Lunn <andrew@lunn.ch> Signed-off-by: Vivien Didelot <vivien.didelot@savoirfairelinux.com> Reviewed-by: Florian Fainelli <f.fainelli@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
liquidio driver incorrectly indicates that PTP is supported in 23XX
adapters; this patch fixes that. PTP is supported in 66XX and 68XX
adapters, and the driver correctly indicates that.
Signed-off-by: Prasad Kanneganti <prasad.kanneganti@cavium.com> Signed-off-by: Felix Manlunas <felix.manlunas@cavium.com> Signed-off-by: Derek Chickles <derek.chickles@cavium.com> Signed-off-by: Satanand Burla <satananda.burla@cavium.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Nicolas Pitre [Tue, 3 Jan 2017 18:57:00 +0000 (13:57 -0500)]
LiquidIO VF: s/select/imply/ for PTP_1588_CLOCK
Fix a minor fallout from the merge of the timers and the networking
trees. The following error may result if the PTP_1588_CLOCK
prerequisites are not available:
drivers/built-in.o: In function `ptp_clock_unregister':
(.text+0x40e0a5): undefined reference to `pps_unregister_source'
drivers/built-in.o: In function `ptp_clock_unregister':
(.text+0x40e0cc): undefined reference to `posix_clock_unregister'
drivers/built-in.o: In function `ptp_clock_event':
(.text+0x40e249): undefined reference to `pps_event'
drivers/built-in.o: In function `ptp_clock_register':
(.text+0x40e5e1): undefined reference to `pps_register_source'
drivers/built-in.o: In function `ptp_clock_register':
(.text+0x40e62c): undefined reference to `posix_clock_register'
drivers/built-in.o: In function `ptp_clock_register':
(.text+0x40e68d): undefined reference to `pps_unregister_source'
Signed-off-by: Nicolas Pitre <nico@linaro.org> Acked-by: Richard Cochran <richardcochran@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Peter Chen [Tue, 3 Jan 2017 09:22:20 +0000 (17:22 +0800)]
net: usb: asix_devices: add .reset_resume for USB PM
The USB core may call reset_resume when it fails to resume asix device.
And USB core can recovery this abnormal resume at low level driver,
the same .resume at asix driver can work too. Add .reset_resume can
avoid disconnecting after backing from system resume, and NFS can
still be mounted after this commit.
Signed-off-by: Peter Chen <peter.chen@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Linus Torvalds [Wed, 4 Jan 2017 17:03:37 +0000 (09:03 -0800)]
Merge branch 'for-linus' of git://git.kernel.dk/linux-block
Pull block layer fixes from Jens Axboe:
"A set of fixes for the current series, one fixing a regression with
block size < page cache size in the alias series from Jan. Outside of
that, two small cleanups for wbt from Bart, a nvme pull request from
Christoph, and a few small fixes of documentation updates"
* 'for-linus' of git://git.kernel.dk/linux-block:
block: fix up io_poll documentation
block: Avoid that sparse complains about context imbalance in __wbt_wait()
block: Make wbt_wait() definition consistent with declaration
clean_bdev_aliases: Prevent cleaning blocks that are not in block range
genhd: remove dead and duplicated scsi code
block: add back plugging in __blkdev_direct_IO
nvmet/fcloop: remove some logically dead code performing redundant ret checks
nvmet: fix KATO offset in Set Features
nvme/fc: simplify error handling of nvme_fc_create_hw_io_queues
nvme/fc: correct some printk information
nvme/scsi: Remove START STOP emulation
nvme/pci: Delete misleading queue-wrap comment
nvme/pci: Fix whitespace problem
nvme: simplify stripe quirk
nvme: update maintainers information
Linus Torvalds [Wed, 4 Jan 2017 17:00:57 +0000 (09:00 -0800)]
Merge tag 'fbdev-v4.10-rc2' of git://github.com/bzolnier/linux
Pull fbdev fixes from Bartlomiej Zolnierkiewicz:
- bring fbdev subsystem back into Maintained mode
- add missing devm_ioremap() error checking to cobalt_lcdfb driver
* tag 'fbdev-v4.10-rc2' of git://github.com/bzolnier/linux:
video: fbdev: cobalt_lcdfb: Handle return NULL error from devm_ioremap
MAINTAINERS: add myself as maintainer of fbdev
Linus Torvalds [Wed, 4 Jan 2017 16:56:05 +0000 (08:56 -0800)]
Merge tag 'gcc-plugins-v4.10-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/kees/linux
Pull gcc-plugins fixes from Kees Cook:
"Small fixes for gcc-plugins when using certain gcc versions:
- update gcc-common.h for gcc 7 (Emese Revfy)
- fix latent_entropy type for early gcc on ARM (PaX Team)"
* tag 'gcc-plugins-v4.10-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/kees/linux:
gcc-plugins: update gcc-common.h for gcc-7
latent_entropy: fix ARM build error on earlier gcc
Arvind Yadav [Tue, 13 Dec 2016 08:20:52 +0000 (13:50 +0530)]
video: fbdev: cobalt_lcdfb: Handle return NULL error from devm_ioremap
Here, If devm_ioremap will fail. It will return NULL.
Kernel can run into a NULL-pointer dereference.
This error check will avoid NULL pointer dereference.
Carlos Maiolino [Wed, 4 Jan 2017 04:34:17 +0000 (20:34 -0800)]
xfs: fix max_retries _show and _store functions
max_retries _show and _store functions should test against cfg->max_retries,
not cfg->retry_timeout
Signed-off-by: Carlos Maiolino <cmaiolino@redhat.com> Reviewed-by: Eric Sandeen <sandeen@redhat.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
xfs: fix crash and data corruption due to removal of busy COW extents
There is a race window between write_cache_pages calling
clear_page_dirty_for_io and XFS calling set_page_writeback, in which
the mapping for an inode is tagged neither as dirty, nor as writeback.
If the COW shrinker hits in exactly that window we'll remove the delayed
COW extents and writepages trying to write it back, which in release
kernels will manifest as corruption of the bmap btree, and in debug
kernels will trip the ASSERT about now calling xfs_bmapi_write with the
COWFORK flag for holes. A complex customer load manages to hit this
window fairly reliably, probably by always having COW writeback in flight
while the cow shrinker runs.
This patch adds another check for having the I_DIRTY_PAGES flag set,
which is still set during this race window. While this fixes the problem
I'm still not overly happy about the way the COW shrinker works as it
still seems a bit fragile.
Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Darrick J. Wong [Wed, 4 Jan 2017 02:39:33 +0000 (18:39 -0800)]
xfs: use the actual AG length when reserving blocks
We need to use the actual AG length when making per-AG reservations,
since we could otherwise end up reserving more blocks out of the last
AG than there are actual blocks.
Complained-about-by: Brian Foster <bfoster@redhat.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com> Reviewed-by: Christoph Hellwig <hch@lst.de>
Yotam Gigi [Tue, 3 Jan 2017 17:20:24 +0000 (19:20 +0200)]
net/sched: cls_matchall: Fix error path
Fix several error paths in matchall:
- Release reference to actions in case the hardware fails offloading
(relevant to skip_sw only)
- Fix error path in case tcf_exts initialization/validation fail
Fixes: bf3994d2ed31 ("net/sched: introduce Match-all classifier") Signed-off-by: Yotam Gigi <yotamg@mellanox.com> Signed-off-by: David S. Miller <davem@davemloft.net>
David Ahern [Tue, 3 Jan 2017 17:37:55 +0000 (09:37 -0800)]
net: vrf: Add missing Rx counters
The move from rx-handler to L3 receive handler inadvertantly dropped the
rx counters. Restore them.
Fixes: 74b20582ac38 ("net: l3mdev: Add hook in ip and ipv6") Reported-by: Dinesh Dutt <ddutt@cumulusnetworks.com> Signed-off-by: David Ahern <dsa@cumulusnetworks.com> Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Tue, 3 Jan 2017 21:23:01 +0000 (16:23 -0500)]
Merge branch '10GbE' of git://git.kernel.org/pub/scm/linux/kernel/git/jkirsher/next-queue
Jeff Kirsher says:
====================
10GbE Intel Wired LAN Driver Updates 2017-01-03
This series contains updates to ixgbe and ixgbevf only.
Emil fixes ixgbe to use the NVM settings for FEC, so do not override the
settings. Fixed the indirection table for x550, where newer devices can
support up to 64 RSS queues. Extends the rtnl_lock() to protect the call
to netif_device_detach() and ixgbe_clear_interrupt_scheme() to avoid
against a double free WARN and/or a BUG in free_msi_irqs(). Fixed AER
error handling by making sure that the driver frees the IRQs in
ixgbe_io_error_detected() when responding to a PCIe AER error, and to
restore them when the interface recovers.
Tony updates the driver to report the driver version to the firmware using
the host interface command for x550 devices. Fixed the PHY reset check
for x550em_ext_t PHY type. Fixed bounds checking for x540 devices to
ensure the index is valid for the LED function. Fixed the BaseT adapters
which support 100Mb capability and were not reporting the capability.
Ken Cox adds a missing check for the trusted bit before trying to set the
MACVLAN MAC address.
Yusuke Suzuki fixes an issue with 82599 and x540 devices where receive
timestamps were not working becase the bitwise operation for RX_HWSTAMP
falg was incorrect.
Don ensures that x553 KR/KX devices correctly advertise link speeds.
Adds the mailbox message to allow for VF promiscuous mode support.
Mark fixes two issues with EEPROM access, where the semaphore was not
being held until the entire response was read and the acquiring/releasing
of the semaphore is slow. Cleaned up firmware version method and
functions which are no longer used. Added new interfaces for firmware
commands to access some new PHYs.
v2: fixed tab indentation in patch 12 and mis-spelled words in patch 15
based on feedback from Sergei Shtylyov and Rami Rosen.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
projects / mirror_ubuntu-jammy-kernel.git / log