In v3.9 6fd6ce2056 ("ipv6: Do not depend on rt->n in
ip6_finish_output2()." changed the behaviour of ip6_finish_output2()
such that the recently introduced rt6_nexthop() is used
instead of an assigned neighbor.
As rt6_nexthop() prefers rt6i_gateway only for gatewayed
routes this causes a problem for users like IPVS, xt_TEE and
RAW(hdrincl) if they want to use different address for routing
compared to the destination address.
Another case is when redirect can create RTF_DYNAMIC
route without RTF_GATEWAY flag, we ignore the rt6i_gateway
in rt6_nexthop().
Fix the above problems by considering the rt6i_gateway if
present, so that traffic routed to address on local subnet is
not wrongly diverted to the destination address.
Thanks to Simon Horman and Phil Oester for spotting the
problematic commit.
Thanks to Hannes Frederic Sowa for his review and help in testing.
Reported-by: Phil Oester <kernel@linuxace.com>
Reported-by: Mark Brooks <mark@loadbalancer.org>
Signed-off-by: Julian Anastasov <ja@ssi.bg>
Acked-by: Hannes Frederic Sowa <hannes@stressinduktion.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Yuval Mintz says:
====================
bnx2x: Bug fixes patch series
This patch series contains fixes for various flows - several SR-IOV issues
are fixed, ethtool callbacks (coalescing and register dump) are corrected,
null pointer dereference on error flows is prevented, etc.
Changes from V1
---------------
- Patch 2 "bnx2x: Prevent an illegal pointer dereference during panic"
is revised, with improved handling of edge cases.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Current driver implementation incorrectly sets the flag only if 64-bit
DMA mask succeeded.
Signed-off-by: Merav Sicron <meravs@broadcom.com>
Signed-off-by: Yuval Mintz <yuvalmin@broadcom.com>
Signed-off-by: Ariel Elior <ariele@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
As part of a register dump, the interface pretends to have the identity
of other interfaces of the same physical device in order to perform
HW configuration for them - specifically, it needs to prevent attentions
from generating on those functions as the register dump accesses registers
in common blocks which whose reading might generate an attention.
However, such pretension is unsafe - unlike other flows in which the driver
uses pretend, during register dump there is no guarantee no other HW access
will take place (by other flows). If such access will take place, the HW will
be accessed by the wrong interface, and leave both functions in an incorrect
state.
This patch removes all pretensions from the register dump flow. Instead, it
changes initial configuration of attentions such that no fatal attention will
be generated for other functions as a result of the register dump
(notice however, a debug print claiming an attention from other functions IS
possible during the register dump)
Signed-off-by: Dmitry Kravkov <dmitry@broadcom.com>
Signed-off-by: Yuval Mintz <yuvalmin@broadcom.com>
Signed-off-by: Ariel Elior <ariele@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
bnx2x has several clients to its DMAE machines - all of them with the exception
of the statistics flow used the same locking mechanisms to synchronize the DMAE
machines' usage.
Since statistics (which are periodically entered) use DMAE without taking the
locks, they may erase the commands which were previously set -
e.g., it may cause a VF to timeout while waiting for a PF answer on the VF-PF
channel as that command header would have been overwritten by the statistics'
header.
This patch makes certain that all flows utilizing DMAE will use the same
API, assuring that the locking scheme will be kept by all said flows.
Signed-off-by: Ariel Elior <ariele@broadcom.com>
Signed-off-by: Yuval Mintz <yuvalmin@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
If debug message is open and bnx2x_vfop_qdtor_cmd() were to fail,
the resulting print would have caused a null pointer dereference.
Signed-off-by: Yuval Mintz <yuvalmin@broadcom.com>
Signed-off-by: Ariel Elior <ariele@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Starting with commit b9871bc "bnx2x: VF RSS support - PF side", if a PF will
have SR-IOV supported in its PCI configuration space, storage drivers will not
work for that interface.
This patch fixes the resource calculation to allow such a configuration to
properly work.
Signed-off-by: Ariel Elior <ariele@broadcom.com>
Signed-off-by: Yuval Mintz <yuvalmin@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
bnx2x drivers configure coalescing incorrectly (e.g., as a result of a call
to 'ethtool -c'). Although this is almost invisible to the user (due to NAPI)
designated tests will show the configuration is incorrect.
Signed-off-by: Dmitry Kravkov <dmitry@broadcom.com>
Signed-off-by: Yuval Mintz <yuvalmin@broadcom.com>
Signed-off-by: Ariel Elior <ariele@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Current code returns upon failure, leaving the VF-PF in an unusable state;
This patch adds the missing release so further commands could pass between
PF and VF.
Signed-off-by: Ariel Elior <ariele@broadcom.com>
Signed-off-by: Yuval Mintz <yuvalmin@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
During a panic, the driver tries to print the Management FW buffer of recent
commands. To do so, the driver reads the address of that buffer from a known
address. If the buffer is unavailable (e.g., PCI reads don't work, MCP is
failing, etc.), the driver will try to access the address it has read, possibly
causing a kernel panic.
This check 'sanitizes' the access, validating the read value is indeed a valid
address inside the management FW's buffers.
The patch also removes a read outside the scope of the buffer, which resulted
in some unrelated chraracters appearing in the log.
Signed-off-by: Yuval Mintz <yuvalmin@broadcom.com>
Signed-off-by: Dmitry Kravkov <dmitry@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
bnx2x VFs do not support Multi-CoS; Current implementation
erroneously sets the VFs maximal number of CoS to be > 1.
This will cause the driver to call alloc_etherdev_mqs() with
a number of queues it cannot possibly support and reflects
in 'odd' driver prints.
Signed-off-by: Yuval Mintz <yuvalmin@broadcom.com>
Signed-off-by: Ariel Elior <ariele@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The create_flow/destroy_flow uverbs and the associated extensions to
the user-kernel verbs ABI are under review and are too experimental to
freeze at this point.
So userspace is not exposed to experimental features and an uinstable
ABI, temporarily disable this for v3.12 (with a Kconfig option behind
staging to reenable it if desired).
The feature will be enabled after proper cleanup for v3.13.
Signed-off-by: Yann Droneaud <ydroneaud@opteya.com>
Link: http://marc.info/?i=cover.1381351016.git.ydroneaud@opteya.com
Link: http://marc.info/?i=cover.1381177342.git.ydroneaud@opteya.com
[ Add a Kconfig option to reenable these verbs. - Roland ]
Signed-off-by: Roland Dreier <roland@purestorage.com>
The LEGO-wiimote uses a different VID than the Nintendo ID. The device is
technically the same so add the ID.
Cc: <stable@vger.kernel.org> # 3.11+
Signed-off-by: David Herrmann <dh.herrmann@gmail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
Commit b82715fdd4 introduces
a 'device' subdirectory under /sys/class/hwmon/hwmonX/ directory,
for the thermal_zone hwmon devices. And this results in different
handling by libsensors.
The problem is reported and discussed in this thread
http://marc.info/?l=linux-pm&m=138229306109596&w=2
This patch reverts commit b82715fdd4.
Reported-by: Arnaud Ebalard <arno@natisbad.org>
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Pull parsic fixes from Helge Deller:
"There are just two small fixes in here:
- Revert a commit which exported the flush_cache_page function. This
was noticed by Christoph Hellwig.
- Enable the DEVTMPFS, DEVTMPFS_MOUNT and BLK_DEV_INITRD config
options in the parisc defconfigs so that latest udev/initrd finds
the root disk at boot"
* 'parisc-3.12' of git://git.kernel.org/pub/scm/linux/kernel/git/deller/parisc-linux:
parisc: enable DEVTMPFS, DEVTMPFS_MOUNT and BLK_DEV_INITRD in defconfigs
Revert "parisc: Export flush_cache_page() (needed by lustre)"
Background: nfsd v[23] had throughput regression since delayed fput
went in; every read or write ends up doing fput() and we get a pair
of extra context switches out of that (plus quite a bit of work
in queue_work itselfi, apparently). Use of schedule_delayed_work()
gives it a chance to accumulate a bit before we do __fput() on all
of them. I'm not too happy about that solution, but... on at least
one real-world setup it reverts about 10% throughput loss we got from
switch to delayed fput.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Jiri Pirko says:
====================
UFO fixes
Couple of patches fixing UFO functionality in different situations.
v1->v2:
- minor if{}else{} coding style adjustment suggested by Sergei Shtylyov
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Now, if user application does:
sendto len<mtu flag MSG_MORE
sendto len>mtu flag 0
The skb is not treated as fragmented one because it is not initialized
that way. So move the initialization to fix this.
introduced by:
commit e89e9cf539 "[IPv4/IPv6]: UFO Scatter-gather approach"
Signed-off-by: Jiri Pirko <jiri@resnulli.us>
Acked-by: Hannes Frederic Sowa <hannes@stressinduktion.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Now, if user application does:
sendto len<mtu flag MSG_MORE
sendto len>mtu flag 0
The skb is not treated as fragmented one because it is not initialized
that way. So move the initialization to fix this.
introduced by:
commit e89e9cf539 "[IPv4/IPv6]: UFO Scatter-gather approach"
Signed-off-by: Jiri Pirko <jiri@resnulli.us>
Acked-by: Hannes Frederic Sowa <hannes@stressinduktion.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
if up->pending != 0 dontfrag is left with default value -1. That
causes that application that do:
sendto len>mtu flag MSG_MORE
sendto len>mtu flag 0
will receive EMSGSIZE errno as the result of the second sendto.
This patch fixes it by respecting IPV6_DONTFRAG socket option.
introduced by:
commit 4b340ae20d "IPv6: Complete IPV6_DONTFRAG support"
Signed-off-by: Jiri Pirko <jiri@resnulli.us>
Acked-by: Hannes Frederic Sowa <hannes@stressinduktion.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
When CONFIG_NETLABEL is disabled, the cipso_v4_validate() function could loop
forever in the main loop if opt[opt_iter +1] == 0, this will causing a kernel
crash in an SMP system, since the CPU executing this function will
stall /not respond to IPIs.
This problem can be reproduced by running the IP Stack Integrity Checker
(http://isic.sourceforge.net) using the following command on a Linux machine
connected to DUT:
"icmpsic -s rand -d <DUT IP address> -r 123456"
wait (1-2 min)
Signed-off-by: Seif Mazareeb <seif@marvell.com>
Acked-by: Paul Moore <paul@paul-moore.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In the case of credentials passing in unix stream sockets (dgram
sockets seem not affected), we get a rather sparse race after
commit 16e5726 ("af_unix: dont send SCM_CREDENTIALS by default").
We have a stream server on receiver side that requests credential
passing from senders (e.g. nc -U). Since we need to set SO_PASSCRED
on each spawned/accepted socket on server side to 1 first (as it's
not inherited), it can happen that in the time between accept() and
setsockopt() we get interrupted, the sender is being scheduled and
continues with passing data to our receiver. At that time SO_PASSCRED
is neither set on sender nor receiver side, hence in cmsg's
SCM_CREDENTIALS we get eventually pid:0, uid:65534, gid:65534
(== overflow{u,g}id) instead of what we actually would like to see.
On the sender side, here nc -U, the tests in maybe_add_creds()
invoked through unix_stream_sendmsg() would fail, as at that exact
time, as mentioned, the sender has neither SO_PASSCRED on his side
nor sees it on the server side, and we have a valid 'other' socket
in place. Thus, sender believes it would just look like a normal
connection, not needing/requesting SO_PASSCRED at that time.
As reverting 16e5726 would not be an option due to the significant
performance regression reported when having creds always passed,
one way/trade-off to prevent that would be to set SO_PASSCRED on
the listener socket and allow inheriting these flags to the spawned
socket on server side in accept(). It seems also logical to do so
if we'd tell the listener socket to pass those flags onwards, and
would fix the race.
Before, strace:
recvmsg(4, {msg_name(0)=NULL, msg_iov(1)=[{"blub\n", 4096}],
msg_controllen=32, {cmsg_len=28, cmsg_level=SOL_SOCKET,
cmsg_type=SCM_CREDENTIALS{pid=0, uid=65534, gid=65534}},
msg_flags=0}, 0) = 5
After, strace:
recvmsg(4, {msg_name(0)=NULL, msg_iov(1)=[{"blub\n", 4096}],
msg_controllen=32, {cmsg_len=28, cmsg_level=SOL_SOCKET,
cmsg_type=SCM_CREDENTIALS{pid=11580, uid=1000, gid=1000}},
msg_flags=0}, 0) = 5
Signed-off-by: Daniel Borkmann <dborkman@redhat.com>
Cc: Eric Dumazet <edumazet@google.com>
Cc: Eric W. Biederman <ebiederm@xmission.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Help ensure that Lars-Peter gets CCed on dmaengine related patches by
adding a MAINTAINERS entry for the helpers.
Signed-off-by: Mark Brown <broonie@kernel.org>
Acked-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
A number of small, fairly unexciting, fixes for drivers - nothing stand
out.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.15 (GNU/Linux)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=aZuy
-----END PGP SIGNATURE-----
Merge tag 'asoc-v3.12-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/sound into for-linus
ASoC: Fixes for v3.12
A number of small, fairly unexciting, fixes for drivers - nothing stand
out.
Latest udev requires that DEVTMPFS and DEVTMPFS_MOUNT are enabled, else
initrd will fail to find root filesystem. Enable missing BLK_DEV_INITRD
for B180 and C3000 machines.
Signed-off-by: Helge Deller <deller@gmx.de>
This reverts commit 320c90be7b.
Christoph Hellwig <hch@infradead.org> commented:
This one shouldn't go in - Geert sent it a bit prematurely, as Lustre
shouldn't use it just to reimplement core VM functionality (which it
shouldn't use either, but that's a separate story).
Signed-off-by: Helge Deller <deller@gmx.de>
Pull btrfs fix from Chris Mason:
"Sage hit a deadlock with ceph on btrfs, and Josef tracked it down to a
regression in our initial rc1 pull. When doing nocow writes we were
sometimes starting a transaction with locks held"
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs:
Btrfs: release path before starting transaction in can_nocow_extent
- intel_pstate fix for misbehavior after system resume if sysfs
attributes are set in a specific way before the corresponding
suspend from Dirk Brandewie.
- A recent intel_pstate fix has no effect if unsigned long is 32-bit,
so fix it up to cover that case as well.
- The s3c64xx cpufreq driver was not updated when the index field of
struct cpufreq_frequency_table was replaced with driver_data, so
update it now. From Charles Keepax.
- The Kconfig help text for ACPI_BUTTON still refers to /proc/acpi/event
that has been dropped recently, so modify it to remove that reference.
From Krzysztof Mazur.
- A Lan Tianyu's change adds a missing mutex unlock to an error code
path in acpi_resume_power_resources().
- Some code related to ACPI power resources, whose very purpose is
questionable to put it lightly, turns out to cause problems to
happen during testing on real systems, so remove it completely
(we may revisit that in the future if there's a compelling enough
reason). From Rafael J Wysocki and Aaron Lu.
/
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.19 (GNU/Linux)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=MV4Q
-----END PGP SIGNATURE-----
Merge tag 'pm+acpi-3.12-rc6' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm
Pull ACPI and power management fixes from Rafael Wysocki:
- intel_pstate fix for misbehavior after system resume if sysfs
attributes are set in a specific way before the corresponding suspend
from Dirk Brandewie.
- A recent intel_pstate fix has no effect if unsigned long is 32-bit,
so fix it up to cover that case as well.
- The s3c64xx cpufreq driver was not updated when the index field of
struct cpufreq_frequency_table was replaced with driver_data, so
update it now. From Charles Keepax.
- The Kconfig help text for ACPI_BUTTON still refers to
/proc/acpi/event that has been dropped recently, so modify it to
remove that reference. From Krzysztof Mazur.
- A Lan Tianyu's change adds a missing mutex unlock to an error code
path in acpi_resume_power_resources().
- Some code related to ACPI power resources, whose very purpose is
questionable to put it lightly, turns out to cause problems to happen
during testing on real systems, so remove it completely (we may
revisit that in the future if there's a compelling enough reason).
From Rafael J Wysocki and Aaron Lu.
* tag 'pm+acpi-3.12-rc6' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:
ACPI / PM: Drop two functions that are not used any more
ATA / ACPI: remove power dependent device handling
cpufreq: s3c64xx: Rename index to driver_data
ACPI / power: Drop automaitc resume of power resource dependent devices
intel_pstate: Fix type mismatch warning
cpufreq / intel_pstate: Fix max_perf_pct on resume
ACPI: remove /proc/acpi/event from ACPI_BUTTON help
ACPI / power: Release resource_lock after acpi_power_get_state() return error
o validate Tx queue only in case of adapters which supports
multi Tx queue.
This patch is to fix regression introduced in commit
aa4a1f7df7
"qlcnic: Enable Tx queue changes using ethtool for 82xx Series adapter"
Signed-off-by: Himanshu Madhani <himanshu.madhani@qlogic.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
It is a required field for all TX_CREATE cmd versions > 0.
This fixes a driver initialization failure, caused by recent SH-R Firmwares
(versions > 10.0.639.0) failing the TX_CREATE cmd when if_id field is
not passed.
Signed-off-by: Sathya Perla <sathya.perla@emulex.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In 3.12 I changed audio to be enabled by default,
but you still had to turn it on via xrandr. This
was confusing to users so change it to minic the
previous behavior:
- audio option is set to -1 (auto) by default which is
the current 3.12 behavior (audio is enabled but requires
xrandr to turn it on).
- if audio = 1, the audio is enabled without needing
to mess with xrandr (previous behavior)
- audio = 0 disables audio
It retains the new feature of allowing the user to enable
audio on the fly with xrandr, but turns audio on
automatically if radeon.audio=1 is set which is what
most users expect.
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
This only seem to work for H.264 but not for VC-1 streams.
Need to investigate further why exactly.
This reverts commit 4b40e59212.
Signed-off-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Stop leaking IB memory and scratch register space when the test fails.
Signed-off-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The wanxl_ioctl() code fails to initialize the two padding bytes of
struct sync_serial_settings after the ->loopback member. Add an explicit
memset(0) before filling the structure to avoid the info leak.
Signed-off-by: Salva Peiró <speiro@ai2.upv.es>
Signed-off-by: David S. Miller <davem@davemloft.net>
Toshiaki Makita says:
====================
bridge: Fix problems around the PVID
There seem to be some undesirable behaviors related with PVID.
1. It has no effect assigning PVID to a port. PVID cannot be applied
to any frame regardless of whether we set it or not.
2. FDB entries learned via frames applied PVID are registered with
VID 0 rather than VID value of PVID.
3. We can set 0 or 4095 as a PVID that are not allowed in IEEE 802.1Q.
This leads interoperational problems such as sending frames with VID
4095, which is not allowed in IEEE 802.1Q, and treating frames with VID
0 as they belong to VLAN 0, which is expected to be handled as they have
no VID according to IEEE 802.1Q.
Note: 2nd and 3rd problems are potential and not exposed unless 1st problem
is fixed, because we cannot activate PVID due to it.
This is my analysis for each behavior.
1. We are using VLAN_TAG_PRESENT bit when getting PVID, and not when
adding/deleting PVID.
It can be fixed in either way using or not using VLAN_TAG_PRESENT,
but I think the latter is slightly more efficient.
2. We are setting skb->vlan_tci with the value of PVID but the variable
vid, which is used in FDB later, is set to 0 at br_allowed_ingress()
when untagged frames arrive at a port with PVID valid. I'm afraid that
vid should be updated to the value of PVID if PVID is valid.
3. According to IEEE 802.1Q-2011 (6.9.1 and Table 9-2), we cannot use
VID 0 or 4095 as a PVID.
It looks like that there are more stuff to consider.
- VID 0:
VID 0 shall not be configured in any FDB entry and used in a tag header
to indicate it is a 802.1p priority-tagged frame.
Priority-tagged frames should be applied PVID (from IEEE 802.1Q 6.9.1).
In my opinion, since we can filter incomming priority-tagged frames by
deleting PVID, we don't need to filter them by vlan_bitmap.
In other words, priority-tagged frames don't have VID 0 but have no VID,
which is the same as untagged frames, and should be filtered by unsetting
PVID.
So, not only we cannot set PVID as 0, but also we don't need to add 0 to
vlan_bitmap, which enables us to simply forbid to add vlan 0.
- VID 4095:
VID 4095 shall not be transmitted in a tag header. This VID value may be
used to indicate a wildcard match for the VID in management operations or
FDB entries (from IEEE 802.1Q Table 9-2).
In current implementation, we can create a static FDB entry with all
existing VIDs by not specifying any VID when creating it.
I don't think this way to add wildcard-like entries needs to change,
and VID 4095 looks no use and can be unacceptable to add.
Consequently, I believe what we should do for 3rd problem is below:
- Not allowing VID 0 and 4095 to be added.
- Applying PVID to priority-tagged (VID 0) frames.
Note: It has been descovered that another problem related to priority-tags
remains. If we use vlan 0 interface such as eth0.0, we cannot communicate
with another end station via a linux bridge.
This problem exists regardless of whether this patch set is applied or not
because we might receive untagged frames from another end station even if we
are sending priority-tagged frames.
This issue will be addressed by another patch set introducing an additional
egress policy, on which Vlad Yasevich is working.
See http://marc.info/?t=137880893800001&r=1&w=2 for detailed discussion.
Patch set follows this mail.
The order of patches is not the same as described above, because the way
to fix 1st problem is based on the assumption that we don't use VID 0 as
a PVID, which is realized by fixing 3rd problem.
(1/4)(2/4): Fix 3rd problem.
(3/4): Fix 1st problem.
(4/4): Fix 2nd probelm.
v2:
- Add descriptions about the problem related to priority-tags in cover letter.
- Revise patch comments to reference the newest spec.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
We currently set the value that variable vid is pointing, which will be
used in FDB later, to 0 at br_allowed_ingress() when we receive untagged
or priority-tagged frames, even though the PVID is valid.
This leads to FDB updates in such a wrong way that they are learned with
VID 0.
Update the value to that of PVID if the PVID is applied.
Signed-off-by: Toshiaki Makita <makita.toshiaki@lab.ntt.co.jp>
Reviewed-by: Vlad Yasevich <vyasevic@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
We are using the VLAN_TAG_PRESENT bit to detect whether the PVID is
set or not at br_get_pvid(), while we don't care about the bit in
adding/deleting the PVID, which makes it impossible to forward any
incomming untagged frame with vlan_filtering enabled.
Since vid 0 cannot be used for the PVID, we can use vid 0 to indicate
that the PVID is not set, which is slightly more efficient than using
the VLAN_TAG_PRESENT.
Fix the problem by getting rid of using the VLAN_TAG_PRESENT.
Signed-off-by: Toshiaki Makita <makita.toshiaki@lab.ntt.co.jp>
Reviewed-by: Vlad Yasevich <vyasevic@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
IEEE 802.1Q says that when we receive priority-tagged (VID 0) frames
use the PVID for the port as its VID.
(See IEEE 802.1Q-2011 6.9.1 and Table 9-2)
Apply the PVID to not only untagged frames but also priority-tagged frames.
Signed-off-by: Toshiaki Makita <makita.toshiaki@lab.ntt.co.jp>
Reviewed-by: Vlad Yasevich <vyasevic@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
IEEE 802.1Q says that:
- VID 0 shall not be configured as a PVID, or configured in any Filtering
Database entry.
- VID 4095 shall not be configured as a PVID, or transmitted in a tag
header. This VID value may be used to indicate a wildcard match for the VID
in management operations or Filtering Database entries.
(See IEEE 802.1Q-2011 6.9.1 and Table 9-2)
Don't accept adding these VIDs in the vlan_filtering implementation.
Signed-off-by: Toshiaki Makita <makita.toshiaki@lab.ntt.co.jp>
Reviewed-by: Vlad Yasevich <vyasevic@redhat.com>
Acked-by: Stephen Hemminger <stephen@networkplumber.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Commit 040a0a37 ("mutex: Add support for wound/wait style locks")
used "!__builtin_constant_p(p == NULL)" but gcc 3.x cannot
handle such expression correctly, leading to boot failure when
built with CONFIG_DEBUG_MUTEXES=y.
Fix it by explicitly passing a bool which tells whether p != NULL
or not.
[ PeterZ: This is a sad patch, but provided it actually generates
similar code I suppose its the best we can do bar whole
sale deprecating gcc-3. ]
Signed-off-by: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Acked-by: Maarten Lankhorst <maarten.lankhorst@canonical.com>
Cc: peterz@infradead.org
Cc: imirkin@alum.mit.edu
Cc: daniel.vetter@ffwll.ch
Cc: robdclark@gmail.com
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Link: http://lkml.kernel.org/r/201310171945.AGB17114.FSQVtHOJFOOFML@I-love.SAKURA.ne.jp
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Julian Anastasov
David S. Miller
Merav Sicron
Dmitry Kravkov
Ariel Elior
Yuval Mintz
Mugunthan V N
Yann Droneaud
David Herrmann
Zhang Rui
Linus Torvalds
Al Viro
Ingo Molnar
Jiri Pirko
Seif Mazareeb
Daniel Borkmann
Mark Brown
Takashi Iwai
Helge Deller
Alex Deucher
Himanshu Madhani
Vasundhara Volam
Christian König
Salva Peiró
Toshiaki Makita
Tetsuo Handa