John W. Linville [Fri, 28 Sep 2012 14:55:00 +0000 (10:55 -0400)]
Merge tag 'nfc-next-3.7-2' of git://git.kernel.org/pub/scm/linux/kernel/git/sameo/nfc-3.0
So says Samuel Ortiz <sameo@linux.intel.com>:
The 2nd NFC pull request for 3.7.
- A couple of wrong context sleep fixes.
- An LLCP rwlock intizialisation fix.
- A missing mutex unlocking for pn533.
- LLCP raw sockets support. This is going to be used for NFC sniffing.
- A build fix for llc_shdlc. It fixes a build error triggered by code that's
living in wireless-next.
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Eric Dumazet [Wed, 26 Sep 2012 06:46:57 +0000 (06:46 +0000)]
net: use bigger pages in __netdev_alloc_frag
We currently use percpu order-0 pages in __netdev_alloc_frag
to deliver fragments used by __netdev_alloc_skb()
Depending on NIC driver and arch being 32 or 64 bit, it allows a page to
be split in several fragments (between 1 and 8), assuming PAGE_SIZE=4096
Switching to bigger pages (32768 bytes for PAGE_SIZE=4096 case) allows :
- Better filling of space (the ending hole overhead is less an issue)
- Less calls to page allocator or accesses to page->_count
- Could allow struct skb_shared_info futures changes without major
performance impact.
This patch implements a transparent fallback to smaller
pages in case of memory pressure.
It also uses a standard "struct page_frag" instead of a custom one.
Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Alexander Duyck <alexander.h.duyck@intel.com> Cc: Benjamin LaHaise <bcrl@kvack.org> Signed-off-by: David S. Miller <davem@davemloft.net>
Eric Dumazet [Wed, 26 Sep 2012 01:39:20 +0000 (01:39 +0000)]
ipv6: gre: remove ip6gre_header_parse()
dev_parse_header() callers provide 8 bytes of storage,
so it's not possible to store an IPv6 address.
Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Thu, 27 Sep 2012 22:35:47 +0000 (18:35 -0400)]
pkt_sched: Fix warning false positives.
GCC refuses to recognize that all error control flows do in fact
set err to something.
Add an explicit initialization to shut it up.
net/sched/sch_drr.c: In function ‘drr_enqueue’:
net/sched/sch_drr.c:359:11: warning: ‘err’ may be used uninitialized in this function [-Wmaybe-uninitialized]
net/sched/sch_qfq.c: In function ‘qfq_enqueue’:
net/sched/sch_qfq.c:885:11: warning: ‘err’ may be used uninitialized in this function [-Wmaybe-uninitialized]
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Thu, 27 Sep 2012 22:31:58 +0000 (18:31 -0400)]
bna: Fix warning false positive.
GCC can't see that in all non-error-return paths we do in fact
set *using_dac to something.
Add an explicit initialization to remove this warning:
drivers/net/ethernet/brocade/bna/bnad.c: In function ‘bnad_pci_probe’:
drivers/net/ethernet/brocade/bna/bnad.c:3079:5: warning: ‘using_dac’ may be used uninitialized in this function [-Wmaybe-uninitialized]
drivers/net/ethernet/brocade/bna/bnad.c:3233:7: note: ‘using_dac’ was declared here
Signed-off-by: David S. Miller <davem@davemloft.net>
fix copy-paste error introduced in linux-next commit
"ipv6: add a new namespace for nf_conntrack_reasm"
Signed-off-by: Konstantin Khlebnikov <khlebnikov@openvz.org> Cc: Amerigo Wang <amwang@redhat.com> Cc: David S. Miller <davem@davemloft.net> Acked-by: Cong Wang <amwang@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Ivan Vecera [Tue, 25 Sep 2012 02:50:42 +0000 (02:50 +0000)]
be2net: fix vfs enumeration
Current VFs enumeration algorithm used in be_find_vfs does not take domain
number into the match. The match found in igb/ixgbe is more elegant and
safe.
Signed-off-by: Ivan Vecera <ivecera@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Linux tunnels were written before RFC6040 and therefore never
implemented the corner case of ECN getting set in the outer header
and the inner header not being ready for it.
Section 4.2. Default Tunnel Egress Behaviour.
o If the inner ECN field is Not-ECT, the decapsulator MUST NOT
propagate any other ECN codepoint onwards. This is because the
inner Not-ECT marking is set by transports that rely on dropped
packets as an indication of congestion and would not understand or
respond to any other ECN codepoint [RFC4774]. Specifically:
* If the inner ECN field is Not-ECT and the outer ECN field is
CE, the decapsulator MUST drop the packet.
* If the inner ECN field is Not-ECT and the outer ECN field is
Not-ECT, ECT(0), or ECT(1), the decapsulator MUST forward the
outgoing packet with the ECN field cleared to Not-ECT.
This patch moves the ECN decap logic out of the individual tunnels
into a common place.
It also adds logging to allow detecting broken systems that
set ECN bits incorrectly when tunneling (or an intermediate
router might be changing the header).
Overloads rx_frame_error to keep track of ECN related error.
Thanks to Chris Wright who caught this while reviewing the new VXLAN
tunnel.
This code was tested by injecting faulty logic in other end GRE
to send incorrectly encapsulated packets.
Signed-off-by: Stephen Hemminger <shemminger@vyatta.com> Signed-off-by: David S. Miller <davem@davemloft.net>
GRE driver incorrectly uses zero as a flag value. Zero is a perfectly
valid value for key, and the tunnel should match packets with no key only
with tunnels created without key, and vice versa.
This is a slightly visible change since previously it might be possible to
construct a working tunnel that sent key 0 and received only because
of the key wildcard of zero. I.e the sender sent key of zero, but tunnel
was defined without key.
Note: using gre key 0 requires iproute2 utilities v3.2 or later.
The original utility code was broken as well.
Signed-off-by: Stephen Hemminger <shemminger@vyatta.com> Signed-off-by: David S. Miller <davem@davemloft.net>
This patch adds proper handling of the buggy revision A2 of LXT973 phy, adding
precautions linked to ERRATA Item 4:
Revision A2 of LXT973 chip randomly returns the contents of the previous even
register when you read a odd register regularly
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr> Acked-by: Richard Cochran <richardcochran@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
cxgb4: Don't attempt to upgrade T4 firmware when cxgb4 will end up as a slave
This patch adds a new common code routine to upgrade an adapter's
firmware. This routine handles all of the complexities of working with the
the existing adapter firmware in order to quiesce the adapter and uP, etc.
For an automatic upgrade it will send a HELLO command to check if cxgb4
want/can upgrade firmware, i.e. if cxgb4 is MASTER and has newer firmware
that it wants to load and call the new common code routine t4_fw_upgrade.
Note that it should not issue a RESET command after a successful firmware
upgrade.
Signed-off-by: Jay Hernandez <jay@chelsio.com> Signed-off-by: Vipul Pandya <vipul@chelsio.com> Signed-off-by: David S. Miller <davem@davemloft.net>
cxgb4: Inform caller if driver didn't upgrade firmware
If a card had already been initialized, on reloading cxgb4 driver firmware
required an upgrade but the upgrade did not happen. In that case a mailbox
timeout would occur during T4 configuration file stuff. The fix is to let the
caller know the firmware was not upgraded so a reset would be issued before
starting the T4 config stuff.
Signed-off-by: Jay Hernandez <jay@chelsio.com> Signed-off-by: Vipul Pandya <vipul@chelsio.com> Signed-off-by: David S. Miller <davem@davemloft.net>
cxgb4: Add support for T4 hardwired driver configuration settings
In case if user defined configuration file at /lib/firmware/cxgb4/t4-config.txt
location and also factory default configuration file written to FLASH are not
present then driver will use hardwired configuration settings.
Signed-off-by: Jay Hernandez <jay@chelsio.com> Signed-off-by: Vipul Pandya <vipul@chelsio.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Starting with T4 firmware version 1.3.11.0 the firmware now supports device
configuration via a Firmware Configuration File. The Firmware Configuration
File was primarily developed in order to centralize all of the configuration,
resource allocation, etc. for Unified Wire operation where multiple
Physical / Virtual Function Drivers would be using a T4 adapter simultaneously.
The Firmware Configuration file can live in three locations as shown below
in order of precedence.
1) User defined configuration file: /lib/firmware/cxgb4/t4-config.txt
2) Factory Default configuration file written to FLASH within
the manufacturing process.
3) Hardwired driver configuration.
Signed-off-by: Jay Hernandez <jay@chelsio.com> Signed-off-by: Vipul Pandya <vipul@chelsio.com> Signed-off-by: David S. Miller <davem@davemloft.net>
cxgb4/cxgb4vf: Code cleanup to enable T4 Configuration File support
This patch adds new enums and macros to enable T4 configuration file support. It
also removes duplicate macro definitions.
It fixes the build failure in cxgb4vf driver introduced because of old macro
definition removal.
It also performs SGE initialization based on T4 configuration file is provided
or not. If it is provided then it uses the parameters provided in it otherwise
it uses hard coded values.
Signed-off-by: Jay Hernandez <jay@chelsio.com> Signed-off-by: Vipul Pandya <vipul@chelsio.com> Signed-off-by: David S. Miller <davem@davemloft.net>
cxgb4: Add functions to read memory via PCIE memory window
This patch implements two new functions t4_mem_win_read and t4_memory_read.
These new functions can be used to read memory via the PCIE memory window.
Please note, for proper execution of these functions PCIE_MEM_ACCESS_BASE_WIN
registers must be setup correctly like how setup_memwin in the cxgb4 driver
does it.
Signed-off-by: Jay Hernandez <jay@chelsio.com> Signed-off-by: Vipul Pandya <vipul@chelsio.com> Signed-off-by: David S. Miller <davem@davemloft.net>
NFC: Fix sleeping in invalid context when netlink socket is closed
netlink_register_notifier requires notify functions to not sleep.
nfc_stop_poll locks device mutex and must not be called from notifier.
Create workqueue that will handle this for all devices.
John W. Linville [Wed, 26 Sep 2012 17:39:10 +0000 (13:39 -0400)]
NFC: Add dummy nfc_llc_shdlc_register definition
This is used when CONFIG_NFC_SHDLC is disabled.
Reported-by: Randy Dunlap <rdunlap@xenotime.net> Signed-off-by: John W. Linville <linville@tuxdriver.com> Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
This adds support for socket of type SOCK_RAW to LLCP.
sk_buff are copied and sent to raw sockets with a 2 bytes extra header:
The first byte header contains the nfc adapter index.
The second one contains flags:
- 0x01 - Direction (0=RX, 1=TX)
- 0x02-0x80 - Reserved
A raw socket has to be explicitly bound to a nfc adapter. This is achieved
by specifying the adapter index to be bound to in the dev_idx field of the
sockaddr_nfc_llcp struct passed to bind().
Signed-off-by: Thierry Escande <thierry.escande@linux.intel.com> Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
Whenever both WLAN and BT in/out sleep mode, sometimes WLAN
is not able to take back the shared LNA control after resumes
from sleep mode. The idea is that for WLAN to check if BT owns
LNA control and BT is in sleep mode when WLAN just resumes from
sleep mode. If the condition is true, do a BTCOEX_RC_WARM_RESET
for WLAN to take back the control of shared LNA.
Now the issue is the BT sleep value read from MCI register is
overlooked by assigning u32 into u8. Hence the above condition never
be met so that WLAN used to report beacon losses and frequent
connection drops.
Signed-off-by: Rajkumar Manoharan <rmanohar@qca.qualcomm.com> Tested-by: Paul Stewart <pstew@chromium.org> Signed-off-by: John W. Linville <linville@tuxdriver.com>
rt2x00: Replace open coded interface checking with interface combinations.
Mac80211 has formal infrastructure to specify which interface combinations
are supported. Make use of this facility in favor of open coding it
ourselves.
So far we only have to specify we can support multiple AP interfaces,
no other combinations are supported.
Inspired by an earlier patch from Paul Fertser.
Signed-off-by: Gertjan van Wingerde <gwingerde@gmail.com> Cc: Paul Fertser <fercerpav@gmail.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
rt2x00: Deprecate max_sta_intf field of struct rt2x00_ops.
All drivers set this value to 1, so there is no need (currently) to let
drivers set this.
Therefor, remove the field; we can always add it back when it is needed.
Inspired by an earlier patch from Paul Fertser.
Signed-off-by: Gertjan van Wingerde <gwingerde@gmail.com> Cc: Paul Fertser <fercerpav@gmail.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
Dan Carpenter [Sun, 23 Sep 2012 16:33:00 +0000 (19:33 +0300)]
mwifiex: potential corruption in mwifiex_update_uap_custom_ie()
ap_custom_ie is a struct mwifiex_ie_list which is quite different and
also larger than struct mwifiex_ie. It's a difference between 4196
bytes and 262.
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> Tested-by: Stone Piao <piaoyun@marvell.com> Acked-by: Bing Zhao <bzhao@marvell.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
Currently, ASPM is disabled for all WLAN+BT combo chipsets
when BTCOEX is enabled. This is incorrect since the workaround
is required only for WB195, which is a AR9285+AR3011 combo
solution. Fix this by checking for the HW version when enabling
the workaround.
Cc: stable@vger.kernel.org Signed-off-by: Sujith Manoharan <c_manoha@qca.qualcomm.com> Tested-by: Paul Stewart <pstew@chromium.org> Signed-off-by: John W. Linville <linville@tuxdriver.com>
John W. Linville [Tue, 25 Sep 2012 19:50:56 +0000 (15:50 -0400)]
Merge tag 'nfc-next-3.7-1' of git://git.kernel.org/pub/scm/linux/kernel/git/sameo/nfc-3.0
So says Samuel Ortiz <sameo@linux.intel.com>:
This is the first NFC pull request for the 3.7 merge window.
With this one we get:
- HCI and LLC layers separation. We now can support various LLC
protocols for HCI drivers, SHDLC being one of them. This will be needed as
we're planning to support raw HCI chipsets that do the SHDLC encapsulation
in firmware. So for now we have an SHDLC and a NOP LLC layers.
- pn533 command queueing implementation. This simplifies the pn533 locking
logic and fixes a kernel warning.
- NCI p2p initiator mode implementation.
- Replace custom workqueues with system ones, for HCI and LLCP.
- Raw pn544 driver removal, as scheduled on the features-removal.txt file.
- A few HCI, SHDLC and LLCP fixes.
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Andy Shevchenko [Mon, 24 Sep 2012 22:09:58 +0000 (22:09 +0000)]
ipconfig: fix trivial build error
The commit 5e953778a2aab04929a5e7b69f53dc26e39b079e ("ipconfig: add nameserver
IPs to kernel-parameter ip=") introduces ic_nameservers_predef() that defined
only for BOOTP. However it is used by ip_auto_config_setup() as well. This
patch moves it outside of #ifdef BOOTP.
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Cc: Christoph Fritz <chf.fritz@googlemail.com> Cc: David S. Miller <davem@davemloft.net> Signed-off-by: David S. Miller <davem@davemloft.net>
cfg80211: Fix regulatory check for 60GHz band frequencies
The current regulatory code on cfg80211 performs a check to
see if a regulatory rule belongs to an IEEE band so that if
a Country IE is received and no rules are specified for a
band (which is allowed by IEEE) those bands are left intact.
The current band check assumes a rule is bound to a band
if the rule's start or end frequency is less than 2 GHz
apart from the center of frequency being inspected.
In order to support 60 GHz for 802.11ad we need to increase
this to account for the channel spacing of 2160 MHz whereby
a channel somewhere in the middle of a regulatory rule may
be more than 2 GHz apart from either the beginning or
end of the frequency rule.
Without a fix for this even though channels 1-3 are allowed world
wide on the rule (57240 - 63720 @ 2160), channel 2 at 60480 MHz
will end up getting disabled given that it is 3240 MHz from
both the frequency rule start and end frequency. Fix this by
using 2 GHz separation assumption for the 2.4 and 5 GHz bands
but for 60 GHz use a 10 GHz separation before assuming a rule
is not part of the band.
Since we have no 802.11ad drivers yet merged this change has
no impact to existing Linux upstream device drivers.
Signed-off-by: Vladimir Kondratiev <qca_vkondrat@qca.qualcomm.com> Acked-by: Luis R. Rodriguez <mcgrof@do-not-panic.com> Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Eric Dumazet [Tue, 25 Sep 2012 06:21:16 +0000 (08:21 +0200)]
net: raw: revert unrelated change
Commit 5640f7685831 ("net: use a per task frag allocator")
accidentally contained an unrelated change to net/ipv4/raw.c,
later committed (without the pr_err() debugging bits) in
net tree as commit ab43ed8b749 (ipv4: raw: fix icmp_filter())
This patch reverts this glitch, noticed by Stephen Rothwell.
Reported-by: Stephen Rothwell <sfr@canb.auug.org.au> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
During processing incoming RSET frame chip, possibly due to
its internal timout, can retrnasmit an another RSET which
is next queued for processing in shdlc layer.
In case when we accept processed RSET skip those remaining on
the rcv queue until chip will send it's first S or I frame.
This will mean the chip completed connection as well.
Signed-off-by: Waldemar Rymarkiewicz <waldemar.rymarkiewicz@tieto.com> Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
As queue_work() does not guarantee immediate execution of sm_work it
can happen in crossover RSET usecase that connect timer will constantly
change the shdlc state from NEGOTIATING to CONNECTING before shdlc has
chance to handle incoming frame.
Signed-off-by: Waldemar Rymarkiewicz <waldemar.rymarkiewicz@tieto.com> Acked-by: Eric Lapuyade <eric.lapuyade@intel.com> Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
xmit callback provided by a driver encapsulates upper layers
data and sends it to the hardware. So, HCI does not know the
exact amount of data being sent and thus can't handle partially
sent frames properly.
Therefore, the driver must return 0 for completely sent frame or
negative for failure.
Signed-off-by: Waldemar Rymarkiewicz <waldemar.rymarkiewicz@tieto.com> Acked-by: Eric Lapuyade <eric.lapuyade@intel.com> Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
Eric Lapuyade [Tue, 18 Sep 2012 17:45:48 +0000 (19:45 +0200)]
NFC: Changed HCI and PN544 HCI driver to use the new HCI LLC Core
The previous shdlc HCI driver and its header are removed from the tree.
PN544 now registers directly with HCI and passes the name of the llc it
requires (shdlc).
HCI instantiation now allocates the required llc instance. The llc is
started when the HCI device is brought up.
Signed-off-by: Eric Lapuyade <eric.lapuyade@intel.com> Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
Eric Lapuyade [Thu, 13 Sep 2012 15:10:48 +0000 (17:10 +0200)]
NFC: Add a nop (passthrough) llc module to llc core
This is a passthrough llc. It can be used by HCI drivers that don't
need link layer control. HCI will then write directly to the driver, and
driver will deliver incoming frames directly to HCI without any
processing.
Signed-off-by: Eric Lapuyade <eric.lapuyade@intel.com> Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
Eric Lapuyade [Thu, 13 Sep 2012 15:10:00 +0000 (17:10 +0200)]
NFC: Add an LLC Core layer to HCI
The LLC layer manages modules that control the link layer protocol (such
as shdlc) between HCI and an HCI driver. The driver must simply specify
the required llc when it registers with HCI.
Signed-off-by: Eric Lapuyade <eric.lapuyade@intel.com> Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
Eric Lapuyade [Tue, 11 Sep 2012 08:43:50 +0000 (10:43 +0200)]
NFC: Modified hci_transceive to become an asynchronous operation
This enables the completion callback to be called from a different
context, preventing a possible deadlock if the callback resulted in the
invocation of a nested call to the currently locked nfc_dev.
This is also more in line with the im_transceive nfc_ops for NFC Core or
NCI drivers which already behave asynchronously.
Signed-off-by: Eric Lapuyade <eric.lapuyade@intel.com> Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
nfc_llcp_build_tlv() malloced the memory and should be free in
nfc_llcp_build_gb() after used, and the same in the error handling
case, otherwise it will cause memory leak.
spatch with a semantic match is used to found this problem.
(http://coccinelle.lip6.fr/)
Signed-off-by: Wei Yongjun <yongjun_wei@trendmicro.com.cn> Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
Tejun Heo [Wed, 22 Aug 2012 23:28:46 +0000 (16:28 -0700)]
NFC: Don't use WQ_MEM_RECLAIM for pn533
NFC driver doesn't sit in memory reclaim path and has no reason to use
WQ_MEM_RECLAIM. Drop WQ_MEM_RECLAIM from pn533->wq and use
alloc_ordered_workqueue() instead of WQ_UNBOUND w/ max_active == 1.
Signed-off-by: Tejun Heo <tj@kernel.org> Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
Tejun Heo [Wed, 22 Aug 2012 23:22:16 +0000 (16:22 -0700)]
NFC: Use system_nrt_wq instead of custom ones
NFC is using a number of custom ordered workqueues w/ WQ_MEM_RECLAIM.
WQ_MEM_RECLAIM is unnecessary unless NFC is gonna be used as transport
for storage device, and all use cases match one work item to one
ordered workqueue - IOW, there's no actual ordering going on at all
and using system_nrt_wq gives the same behavior.
There's nothing to be gained by using custom workqueues. Use
system_nrt_wq instead and drop all the custom ones.
Signed-off-by: Tejun Heo <tj@kernel.org> Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
Ilan Elias [Wed, 15 Aug 2012 08:46:24 +0000 (11:46 +0300)]
NFC: Implement NCI dep_link_up and dep_link_down
During NFC-DEP target activation, store the remote
general bytes to be used later in dep_link_up.
When dep_link_up is called, activate the NFC-DEP target,
and forward the remote general bytes.
When dep_link_down is called, deactivate the target.
Signed-off-by: Ilan Elias <ilane@ti.com> Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
Samuel Ortiz [Fri, 17 Aug 2012 21:47:54 +0000 (23:47 +0200)]
NFC: Queue pn533 commands
Instead of returning EBUSY when getting a command while another one is
running, we queue them. Upon completion of the pending command, the next
one is processed.
Besides the fact that it simplifies the pn533 locking scheme, it also
comes with the nice side effect of fixing the following warning:
Daniel Borkmann [Mon, 24 Sep 2012 07:34:51 +0000 (07:34 +0000)]
x86: bpf_jit_comp: add XOR instruction for BPF JIT
This patch is a follow-up for patch "filter: add XOR instruction for use
with X/K" that implements BPF x86 JIT parts for the BPF XOR operation.
Signed-off-by: Daniel Borkmann <daniel.borkmann@tik.ee.ethz.ch> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Daniel Borkmann [Mon, 24 Sep 2012 02:23:59 +0000 (02:23 +0000)]
filter: add XOR instruction for use with X/K
SKF_AD_ALU_XOR_X has been added a while ago, but as an 'ancillary'
operation that is invoked through a negative offset in K within BPF
load operations. Since BPF_MOD has recently been added, BPF_XOR should
also be part of the common ALU operations. Removing SKF_AD_ALU_XOR_X
might not be an option since this is exposed to user space.
Signed-off-by: Daniel Borkmann <daniel.borkmann@tik.ee.ethz.ch> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Eric Dumazet [Sun, 23 Sep 2012 23:04:42 +0000 (23:04 +0000)]
net: use a per task frag allocator
We currently use a per socket order-0 page cache for tcp_sendmsg()
operations.
This page is used to build fragments for skbs.
Its done to increase probability of coalescing small write() into
single segments in skbs still in write queue (not yet sent)
But it wastes a lot of memory for applications handling many mostly
idle sockets, since each socket holds one page in sk->sk_sndmsg_page
Its also quite inefficient to build TSO 64KB packets, because we need
about 16 pages per skb on arches where PAGE_SIZE = 4096, so we hit
page allocator more than wanted.
This patch adds a per task frag allocator and uses bigger pages,
if available. An automatic fallback is done in case of memory pressure.
(up to 32768 bytes per frag, thats order-3 pages on x86)
This increases TCP stream performance by 20% on loopback device,
but also benefits on other network devices, since 8x less frags are
mapped on transmit and unmapped on tx completion. Alexander Duyck
mentioned a probable performance win on systems with IOMMU enabled.
Its possible some SG enabled hardware cant cope with bigger fragments,
but their ndo_start_xmit() should already handle this, splitting a
fragment in sub fragments, since some arches have PAGE_SIZE=65536
Successfully tested on various ethernet devices.
(ixgbe, igb, bnx2x, tg3, mellanox mlx4)
Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Ben Hutchings <bhutchings@solarflare.com> Cc: Vijay Subramanian <subramanian.vijay@gmail.com> Cc: Alexander Duyck <alexander.h.duyck@intel.com> Tested-by: Vijay Subramanian <subramanian.vijay@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
This is primarily to address transmission timeout occurrences, when
multiple H/W Tx queues are being used concurrently. Because in
the priority scheduling mode the controller does not service the
Tx queues equally (but in ascending index order), Tx timeouts are
being triggered rightaway for a basic test with multiple simultaneous
connections like:
iperf -c <server_ip> -n 100M -P 8
resulting in kernel trace:
NETDEV WATCHDOG: eth1 (fsl-gianfar): transmit queue <X> timed out
------------[ cut here ]------------
WARNING: at net/sched/sch_generic.c:255
...
and controller reset during intense traffic, and possibly further
complications.
This patch changes the default H/W Tx scheduling setting (TXSCHED)
for multi-queue devices, from priority scheduling mode to a weighted
round robin mode with equal weights for all H/W Tx queues, and
addresses the issue above.
Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Eric Dumazet [Sun, 23 Sep 2012 22:28:59 +0000 (22:28 +0000)]
net: loopback: set default mtu to 64K
loopback current mtu of 16436 bytes allows no more than 3 MSS TCP
segments per frame, or 48 Kbytes. Changing mtu to 64K allows TCP
stack to build large frames and significantly reduces stack overhead.
Performance boost on bulk TCP transferts can be up to 30 %, partly
because we now have one ACK message for two 64KB segments, and a lower
probability of hitting /proc/sys/net/ipv4/tcp_reordering default limit.
Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
This patch changes the definition of bnx2x_tests_str_arr from static char
pointer to static const char bi-directional array. Also the
bnx2x_get_strings function is simplified.
Reported-by: Joe Perches <joe@perches.com> Reported-by: David Laight <David.Laight@ACULAB.COM> Signed-off-by: Merav Sicron <meravs@broadcom.com> Signed-off-by: Eilon Greenstein <eilong@broadcom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Convert a nonnegative error return code to a negative one, as returned
elsewhere in the function.
A simplified version of the semantic match that finds this problem is as
follows: (http://coccinelle.lip6.fr/)
// <smpl>
(
if@p1 (\(ret < 0\|ret != 0\))
{ ... return ret; }
|
ret@p1 = 0
)
... when != ret = e1
when != &ret
*if(...)
{
... when != ret = e2
when forall
return ret;
}
// </smpl>
Signed-off-by: Peter Senna Tschudin <peter.senna@gmail.com> Signed-off-by: Ursula Braun <ursula.braun@de.ibm.com> Signed-off-by: Frank Blaschka <frank.blaschka@de.ibm.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Peter Senna Tschudin <peter.senna@gmail.com> Signed-off-by: Ursula Braun <ursula.braun@de.ibm.com> Signed-off-by: Frank Blaschka <frank.blaschka@de.ibm.com> Signed-off-by: David S. Miller <davem@davemloft.net>
qeth: fix possible memory leak in qeth_l3_add_[vipa|rxip]()
ipaddr has been allocated in function qeth_l3_add_vipa() but
does not free before leaving from the error handling cases. The
same problem also exists in function qeth_l3_add_rxip().
spatch with a semantic match is used to found this problem.
(http://coccinelle.lip6.fr/)
Signed-off-by: Wei Yongjun <yongjun_wei@trendmicro.com.cn> Signed-off-by: Frank Blaschka <frank.blaschka@de.ibm.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Sebastian Ott [Mon, 24 Sep 2012 04:24:24 +0000 (04:24 +0000)]
lcs: ensure proper ccw setup
Make sure that all ccws used for writing are initialized with
zeros - especially since the last ccw contains a TIC for which
the unused fields have to be zeros.
Signed-off-by: Sebastian Ott <sebott@linux.vnet.ibm.com> Signed-off-by: Frank Blaschka <frank.blaschka@de.ibm.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Sebastian Ott [Mon, 24 Sep 2012 04:24:23 +0000 (04:24 +0000)]
qeth: cleanup channel path descriptor function
Cleanup the qeth_get_channel_path_desc function and rename it
to qeth_update_from_chp_desc. No functional change.
Signed-off-by: Sebastian Ott <sebott@linux.vnet.ibm.com> Signed-off-by: Frank Blaschka <frank.blaschka@de.ibm.com> Acked-by: Ursula Braun <ursula.braun@de.ibm.com> Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Mon, 24 Sep 2012 19:36:53 +0000 (15:36 -0400)]
Merge branch 'master' of git://1984.lsi.us.es/nf-next
Pablo Neira Ayuso says:
====================
This patchset contains updates for your net-next tree, they are:
* Mostly fixes for the recently pushed IPv6 NAT support:
- Fix crash while removing nf_nat modules from Patrick McHardy.
- Fix unbalanced rcu_read_unlock from Ulrich Weber.
- Merge NETMAP and REDIRECT into one single xt_target module, from
Jan Engelhardt.
- Fix Kconfig for IPv6 NAT, which allows inconsistent configurations,
from myself.
* Updates for ipset, all of the from Jozsef Kadlecsik:
- Add the new "nomatch" option to obtain reverse set matching.
- Support for /0 CIDR in hash:net,iface set type.
- One non-critical fix for a rare crash due to pass really
wrong configuration parameters.
- Coding style cleanups.
- Sparse fixes.
- Add set revision supported via modinfo.i
* One extension for the xt_time match, to support matching during
the transition between two days with one single rule, from
Florian Westphal.
* Fix maximum packet length supported by nfnetlink_queue and add
NFQA_CAP_LEN attribute, from myself.
You can notice that this batch contains a couple of fixes that may
go to 3.6-rc but I don't consider them critical to push them:
* The ipset fix for the /0 cidr case, which is triggered with one
inconsistent command line invocation of ipset.
* The nfnetlink_queue maximum packet length supported since it requires
the new NFQA_CAP_LEN attribute to provide a full workaround for the
described problem.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Previously, it was not possible to connect to
networks which requires 11w to be supported by
the stations.
While the documentation hints that there's some
hardware support for offloading MFP "decryption",
this simple implementation relies on the mac80211
stack to do the actual crypto operations.
Signed-off-by: Christian Lamparter <chunkeey@googlemail.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
Previously, it was not possible to connect to
networks which requires 11w to be supported by
the stations.
Note:
As all current (and old) firmwares corrupt
incoming, protected management frames, the
decryption offloading needs to be disabled.
This will be done automatically if needed.
Signed-off-by: Christian Lamparter <chunkeey@googlemail.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
Paul Bolle [Fri, 21 Sep 2012 10:02:02 +0000 (12:02 +0200)]
ipw2x00: silence GCC warning for unused variable 'dev'
Building the libipw component without CONFIG_LIBIPW_DEBUG set triggers this GCC
warning:
drivers/net/wireless/ipw2x00/libipw_wx.c:526:21: warning: unused variable 'dev' [-Wunused-variable]
The cause of this warning is that, without CONFIG_LIBIPW_DEBUG set,
LIBIPW_DEBUG_WX compiles away. Fix it by substituting ieee->dev for (its
equivalent) dev.
Signed-off-by: Paul Bolle <pebolle@tiscali.nl> Signed-off-by: John W. Linville <linville@tuxdriver.com>
mwifiex: block scan request during heavy Tx traffic
Currently scan operation is delayed/aborted based on Tx traffic
consistency. This decision is taken after receiving scan
response of first scan command from FW. But when heavy traffic
is running, we can not even afford to send first scan command
and go off channel for 30msec. We will block scan request in
this case.
Signed-off-by: Amitkumar Karwar <akarwar@marvell.com> Signed-off-by: Bing Zhao <bzhao@marvell.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
mwifiex: disconnect the device before entering suspend state
By default, device is disconnected before entering suspend state.
User can keep the connection alive by using module parameter
"disconect_on_suspend=0".
Signed-off-by: Amitkumar Karwar <akarwar@marvell.com> Signed-off-by: Bing Zhao <bzhao@marvell.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
brcmfmac: get rid of void pointer in struct brcmf_cfg80211_priv
Field 'pub' in struct brcmf_cfg80211_priv was types as void pointer
and filled with parameter passed in brcmf_cfg80211_attach(). This
patch makes the type specific, ie. struct brcmf_pub.
Reviewed-by: Hante Meuleman <meuleman@broadcom.com> Signed-off-by: Arend van Spriel <arend@broadcom.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
brcmfmac: change struct brcmf_cfg80211_priv comments to kernel-doc
Small step to fix structure commenting using kernel-doc syntax.
Reviewed-by: Hante Meuleman <meuleman@broadcom.com> Signed-off-by: Arend van Spriel <arend@broadcom.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
Reviewed-by: Arend Van Spriel <arend@broadcom.com> Signed-off-by: Hante Meuleman <meuleman@broadcom.com> Signed-off-by: Arend van Spriel <arend@broadcom.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
projects / ~andy / linux / log