This reverts commit 22604c86.

We can't fix this issue in this way, because we now can try
to take the dev_base_lock rwlock as a writer in software interrupt
context and that is not allowed without major surgery elsewhere.

This initial link state problem needs to be solved in some other
way.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

In splice TCP receive, the SPLICE_F_NONBLOCK flag is used
to compute the "timeo" value.  So checking it again inside
of the main receive loop to trigger -EAGAIN processing is
entirely unnecessary.

Noticed by Jarek P. and Lennert Buytenhek.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Currently, setting SPLICE_F_NONBLOCK on splice from a TCP socket
results in masking of EOF (RDHUP) and error conditions on the socket
by an -EAGAIN return.  Move the NONBLOCK check in tcp_splice_read()
to be after the EOF and error checks to fix this.
Signed-off-by: NLennert Buytenhek <buytenh@marvell.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch integrates the TFRC library, which is a dependency of CCID-3 (and
CCID-4), with the new use of CCIDs in the DCCP module.		
Signed-off-by: NGerrit Renker <gerrit@erg.abdn.ac.uk>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch cleans up after integrating the CCID modules and, in addition,

 * moves the if/else cases from ccid_delete() into ccid_hc_{tx,rx}_delete();
 * removes the 'gfp' argument to ccid_new() - since it is always gfp_any().
Signed-off-by: NGerrit Renker <gerrit@erg.abdn.ac.uk>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Based on Arnaldo's earlier patch, this patch integrates the standardised
CCID congestion control plugins (CCID-2 and CCID-3) of DCCP with dccp.ko:

 * enables a faster connection path by eliminating the need to always go 
   through the CCID registration lock;

 * updates the implementation to use only a single array whose size equals
   the number of configured CCIDs instead of the maximum (256);

 * since the CCIDs are now fixed array elements, synchronization is no
   longer needed, simplifying use and implementation.

CCID-2 is suggested as minimum for a basic DCCP implementation (RFC 4340, 10);
CCID-3 is a standards-track CCID supported by RFC 4342 and RFC 5348.
Signed-off-by: NGerrit Renker <gerrit@erg.abdn.ac.uk>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

drivers/s390/net/qeth_l3_main.c: In function 'qeth_l3_setadapter_parms':
drivers/s390/net/qeth_l3_main.c:1049: warning: too many arguments for format
Signed-off-by: NHeiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: NFrank Blaschka <frank.blaschka@de.ibm.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

From: Klaus-Dieter Wacker <kdwacker@de.ibm.com>

The device driver qeth dos not support large send using EDDP for 
HiperSockets.
Signed-off-by: NKlaus-Dieter Wacker <kdwacker@de.ibm.com>
Signed-off-by: NFrank Blaschka <frank.blaschka@de.ibm.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The ip assist hw command for setting an IP address last unacceptable
long so we can not spin while we waiting for the irq. Since we can
ensure process context for all occurrences of this command we can use
wait.
Signed-off-by: NFrank Blaschka <frank.blaschka@de.ibm.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

For z/VM GuestLAN or VSWITCH devices the transport layer is
configured in z/VM. The layer2 attribute of a participating Linux
device has to match the z/VM definition. In case of a mismatch
Linux currently crashes in qeth recovery due to a reference to the
not yet existing net_device.
Solution: add a check for existence of net_device and add a message
pointing to the mismatch of layer definitions in Linux and z/VM.
Signed-off-by: NUrsula Braun <ursula.braun@de.ibm.com>
Signed-off-by: NFrank Blaschka <frank.blaschka@de.ibm.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

OSA-devices operating in layer3 mode offer adding of the source MAC
address to the QDIO header of inbound packets. The qeth driver can
exploit this functionality to replace FAKELL-entries in the ethernet
header of received packets.
Signed-off-by: NUrsula Braun <ursula.braun@de.ibm.com>
Signed-off-by: NFrank Blaschka <frank.blaschka@de.ibm.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The pre z9 machines provide an mcl string in EBCDIC format,
z9 or later provide string in ASCII format.
Signed-off-by: NKlaus-Dieter Wacker <kdwacker@de.ibm.com>
Signed-off-by: NFrank Blaschka <frank.blaschka@de.ibm.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Since commit ca109491 ("hrtimer:
removing all ur callback modes") the hrtimer callbacks are processed
only in hardirq context.

This patch moves some functionality into tasklets to run in softirq
context.

Additionally some duplicated code was removed in bcm_rx_thr_flush()
and an avoidable memcpy was removed from bcm_rx_handler().
Signed-off-by: NOliver Hartkopp <oliver@hartkopp.net>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Use kfree_skb instead of kfree for struct sk_buff pointers.
Signed-off-by: NRoel Kluin <roel.kluin@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NIlpo Järvinen <ilpo.jarvinen@helsinki.fi>
Reported-by: NEric Sesterhenn <snakebyte@gmx.de>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

From: Michael Marineau <mike@marineau.org>

Commit b4730016 "Do not fire linkwatch
events until the device is registered." was made as a workaround for
drivers that call netif_carrier_off before registering the device.
Unfortunately this causes these drivers to incorrectly report their
link status as IF_OPER_UNKNOWN which can falsely set the IFF_RUNNING
flag when the interface is first brought up. This issues was
previously pointed out[1] but was dismissed saying that IFF_RUNNING is
not related to the link status. From my digging IFF_RUNNING, as
reported to userspace, is based on the link state. It is set based on
__LINK_STATE_START and IF_OPER_UP or IF_OPER_UNKNOWN. See [2], [3],
and [4]. (Whether or not the kernel has IFF_RUNNING set in flags is
not reported to user space so it may well be independent of the link,
I don't know if and when it may get set.)

The end result depends slightly depending on the driver. The the two I
tested were e1000e and b44. With e1000e if the system is booted
without a network cable attached the interface will falsely report
RUNNING when it is brought up causing NetworkManager to attempt to
start it and eventually time out. With b44 when the system is booted
with a network cable attached and brought up with dhcpcd it will time
out the first time.

The attached patch that will still set the operstate variable
correctly to IF_OPER_UP/DOWN/etc when linkwatch_fire_event is called
but then return rather than skipping the linkwatch_fire_event call
entirely as the previous fix did. (sorry it isn't inline, I don't have
a patch friendly email client at the moment)
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

register_pernet_gen_device() expects 'int*', found via sparse.

 CHECK   drivers/net/tun.c
 drivers/net/tun.c:1245:36: warning: incorrect type in argument 1 (different signedness)
 drivers/net/tun.c:1245:36:    expected int *id
 drivers/net/tun.c:1245:36:    got unsigned int static [toplevel] *<noident>
Signed-off-by: NGerrit Renker <gerrit@erg.abdn.ac.uk>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Add missing space after if, switch, for and while keywords.
Signed-off-by: NBruce Allan <bruce.w.allan@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

commit 4dec9b80 ("rfkill: strip pointless
notifier chain") removed the only user of rfkill_led_trigger() that was not
guarded by #ifdef CONFIG_RFKILL_LEDS. Therefore, move rfkill_led_trigger()
completely inside #ifdef CONFIG_RFKILL_LEDS and avoid the compile time
warning:

net/rfkill/rfkill.c:59: warning: 'rfkill_led_trigger' defined but not used
Signed-off-by: NSimon Holm Thøgersen <odie@cs.aau.dk>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Some indexed registers do not have error bits. In these cases a
value of zero should be used for error checking.
Signed-off-by: NRon Mercer <ron.mercer@qlogic.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The length field for these rings is 16-bits.  If the length is
the max supported 65536 then the setting should be zero.
Signed-off-by: NRon Mercer <ron.mercer@qlogic.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Shadow registers are consistent memory locations to which the chip
echos ring indexes in little endian format.  These values need to
be endian swapped before referencing.

Note:
The register pointer declaration uses the volatile modifier which
causes warnings in checkpatch.
Per Documentation/volatile-considered-harmful.txt:
  - Pointers to data structures in coherent memory which might be modified
    by I/O devices can, sometimes, legitimately be volatile.  A ring buffer
    used by a network adapter, where that adapter changes pointers to
    indicate which descriptors have been processed, is an example of this
    type of situation.
Signed-off-by: NRon Mercer <ron.mercer@qlogic.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NRon Mercer <ron.mercer@qlogic.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NRon Mercer <ron.mercer@qlogic.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The enc28j60 driver doesn't check whether the length of the packet as reported 
by the hardware fits into the preallocated buffer. When stressed, the hardware 
may report insanely large packets even tough the "Receive OK" bit is set. Fix 
this.
Signed-off-by: NBaruch Siach <baruch@tkos.co.il>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Correct two typos.
Signed-off-by: NRoel Kluin <roel.kluin@gmail.com>
Acked-by: NKarsten Keil <kkeil@suse.de>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch allows GRO to merge page frags (skb_shinfo(skb)->frags)
in one skb, rather than using the less efficient frag_list.

It also adds a new interface, napi_gro_frags to allow drivers
to inject page frags directly into the stack without allocating
an skb.  This is intended to be the GRO equivalent for LRO's
lro_receive_frags interface.

The existing GSO interface can already handle page frags with
or without an appended frag_list so nothing needs to be changed
there.

The merging itself is rather simple.  We store any new frag entries
after the last existing entry, without checking whether the first
new entry can be merged with the last existing entry.  Making this
check would actually be easy but since no existing driver can
produce contiguous frags anyway it would just be mental masturbation.

If the total number of entries would exceed the capacity of a
single skb, we simply resort to using frag_list as we do now.
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

In order to allow GRO packets without frag_list at all, we need to
store the MSS in the packet itself.  The obvious place is gso_size.
The only thing to watch out for is if the packet ends up not being
GRO then we need to clear gso_size before pushing the packet into
the stack.
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Firmware blob is big endian
Signed-off-by: NJaswinder Singh Rajput <jaswinderrajput@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Firmware blob looks like this...
        u8 firmware_major
        u8 firmware_minor
        u8 firmware_fix
        u8 pad
        __be32 start_address
        __be32 length (total, including BSS sections to be zeroed)
        data... (in __be32 words, which is native for the firmware)
Signed-off-by: NJaswinder Singh Rajput <jaswinderrajput@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

We store the firmware in its native big-endian form now, so the loop in
ace_copy() is modified to use be32_to_cpup() when writing it out.

We can forget the BSS,SBSS sections of the firmware, since we were
clearing all the device's RAM anyway. And the text,rodata,data sections
can all be loaded as a single chunk since they're contiguous (give or
take a few dozen bytes in between).
Signed-off-by: NJaswinder Singh <jaswinder@infradead.org>
Signed-off-by: NDavid Woodhouse <dwmw2@infradead.org>
Acked-by: NJes Sorensen <jes@sgi.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Thanks to excellent diagnosis by Eduard Guzovsky.

The core problem is that on a network with lots of active
multicast traffic, the neighbour cache can fill up.  If
we try to allocate a new route and thus neighbour cache
entry, the bog-standard GC attempt the neighbour layer does
in ineffective because route entries hold a reference
to the existing neighbour entries and GC can only liberate
entries with no references.

IPV4 already has a way to handle this, by doing a route cache
GC in such situations (when neigh attach returns -ENOBUFS).

So simply mimick this on the ipv6 side.
Tested-by: NEduard Guzovsky <eguzovsky@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

When we converted the protocol atomic counters such as the orphan
count and the total socket count deadlocks were introduced due to
the mismatch in BH status of the spots that used the percpu counter
operations.

Based on the diagnosis and patch by Peter Zijlstra, this patch
fixes these issues by disabling BH where we may be in process
context.
Reported-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
Tested-by: NIngo Molnar <mingo@elte.hu>
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

In future all cpumask ops will only be valid (in general) for bit
numbers < nr_cpu_ids.  So use that instead of NR_CPUS in iterators
and other comparisons.

This is always safe: no cpu number can be >= nr_cpu_ids, and
nr_cpu_ids is initialized to NR_CPUS at boot.
Signed-off-by: NRusty Russell <rusty@rustcorp.com.au>
Signed-off-by: NMike Travis <travis@sgi.com>
Acked-by: NIngo Molnar <mingo@elte.hu>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This set of patches introduces calls to the following set of functions:

usb_endpoint_dir_in(epd)
usb_endpoint_dir_out(epd)
usb_endpoint_is_bulk_in(epd)
usb_endpoint_is_bulk_out(epd)
usb_endpoint_is_int_in(epd)
usb_endpoint_is_int_out(epd)
usb_endpoint_num(epd)
usb_endpoint_type(epd)
usb_endpoint_xfer_bulk(epd)
usb_endpoint_xfer_control(epd)
usb_endpoint_xfer_int(epd)
usb_endpoint_xfer_isoc(epd)

In some cases, introducing one of these functions is not possible, and it
just replaces an explicit integer value by one of the following constants:

USB_ENDPOINT_XFER_BULK
USB_ENDPOINT_XFER_CONTROL
USB_ENDPOINT_XFER_INT
USB_ENDPOINT_XFER_ISOC

In drivers/net/wireless/zd1211rw/zd_usb.c the code:

(endpoint->bEndpointAddress & USB_TYPE_MASK) == USB_DIR_OUT

is suspicious.  If it is intended to use USB_ENDPOINT_DIR_MASK rather than
USB_TYPE_MASK, then the whole conditional test could be converted to a call
to usb_endpoint_is_bulk_in.

An extract of the semantic patch that makes these changes is as follows:
(http://www.emn.fr/x-info/coccinelle/)

// <smpl>
@r1@ struct usb_endpoint_descriptor *epd; @@

- ((epd->bmAttributes & \(USB_ENDPOINT_XFERTYPE_MASK\|3\)) ==
- \(USB_ENDPOINT_XFER_CONTROL\|0\))
+ usb_endpoint_xfer_control(epd)

@r5@ struct usb_endpoint_descriptor *epd; @@

- ((epd->bEndpointAddress & \(USB_ENDPOINT_DIR_MASK\|0x80\)) ==
-  \(USB_DIR_IN\|0x80\))
+ usb_endpoint_dir_in(epd)

@inc@
@@

#include <linux/usb.h>

@depends on !inc && (r1||r5)@
@@

+ #include <linux/usb.h>
  #include <linux/usb/...>
// </smpl>
Signed-off-by: NJulia Lawall <julia@diku.dk>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

cls_cgroup can't be compiled as a module, since it's not supported by
cgroup.
Signed-off-by: NLi Zefan <lizf@cn.fujitsu.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

- It's better to use container_of() instead of casting cgroup_subsys_state *
  to cgroup_cls_state *.
- Add helper function task_cls_state().
- Rename net_cls_state() to cgrp_cls_state().
Signed-off-by: NLi Zefan <lizf@cn.fujitsu.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

When removing a cgroup, an oops was triggered immediately. The cause
is wrong kfree() in cgrp_destroy().
Signed-off-by: NLi Zefan <lizf@cn.fujitsu.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

in drivers/net/eexpress.c:558, function unstick_cu()

while (!SCB_complete(rsst=scb_status(dev))) {
	...
	if (...)
		printk(KERN_WARNING "%s: Reset timed out status %04x, retrying...\n",
                                                       dev->name,rsst);
}

but this will become 

while (!((rsst = scb_status(dev) & 0x8000) != 0) ...

because of the macro:

#define SCB_complete(s) ((s&0x8000)!=0)

so rsst can only become either 0x8000 or 0, but in the latter case the
loop ends, I think the wrong timed out status is printed. This also
cleans up similar macros.
Signed-off-by: NRoel Kluin <roel.kluin@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Now using Ethtool to determine ring sizes, removed the module parameters
that controlled those values.
Modifying ring size requires restart of the interface.
Signed-off-by: NYevgeny Petrilin <yevgenyp@mellanox.co.il>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>