pcmcia_request_irq() and pcmcia_enable_device() are intended
to be called from process context (first function allocate memory
with GFP_KERNEL, second take a mutex). We can not take spin lock
and call them.

It's safe to move spin lock after pcmcia_enable_device() as we
still hold off IRQ until dev->base_addr is 0 and driver will
not proceed with interrupts when is not ready.

Patch resolves:
https://bugzilla.redhat.com/show_bug.cgi?id=643758

Reported-and-tested-by: rbugz@biobind.com
Cc: stable@kernel.org  # 2.6.34+
Signed-off-by: NStanislaw Gruszka <sgruszka@redhat.com>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

This is an attempt to fix a long standing open bug:
http://bugzilla.intellinuxwireless.org/show_bug.cgi?id=1334

The interrupt handler checks for INTA being -1, apparently that means that the
hardware is gone. But the interrupt handler defers actual interrupt processing
to a tasklet. By the time the tasklet is run and checks INTA again, the
hardware might be gone and INTA be -1, which confuses the driver because all
event bits are set.

The patch applies to 2.6.37.
Signed-off-by: NIndan Zupancic <indan@nul.nu>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

P54_HDR_FLAG_DATA_OUT_SEQNR is meant to tell the
firmware that "the frame's sequence number has
already been set by the application."

Whereas IEEE80211_TX_CTL_ASSIGN_SEQ is set for
frames which lack a valid sequence number and
either the driver or firmware has to assign one.

Yup, it's the exact opposite!

Cc: <stable@kernel.org>
Signed-off-by: NChristian Lamparter <chunkeey@googlemail.com>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

The chainmask value along with other configuration has to be set
on the target for packet injection. Fix this and also move the monitor
interface addition before the channel set segment to ensure that
the opmode is updated properly.
Signed-off-by: NSujith Manoharan <Sujith.Manoharan@atheros.com>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

Commit "ath9k_hw: Abort rx if hw is not coming out of full sleep in reset"
uncondionally added aborting RX DMA in a HW reset, though it is a bit
unclear as to why this is needed.

Anyway, RX DMA is handled in the target for USB devices, and this would
interfere with normal operations (scanning etc.), so fix this.
Signed-off-by: NSujith Manoharan <Sujith.Manoharan@atheros.com>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

Hardcode the output voltage of x-PA bias LDO to the lowest
value for UB94. The card doesn't get too hot now.
Signed-off-by: NSujith Manoharan <Sujith.Manoharan@atheros.com>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

AR9287 based devices have issues with ADC gain calibration
which would cause uplink throughput drops in HT40 mode.
Remove ADC gain from the supported calibration algorithms.
Signed-off-by: NSujith Manoharan <Sujith.Manoharan@atheros.com>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

USB devices do not require the chip test routine.
Signed-off-by: NSujith Manoharan <Sujith.Manoharan@atheros.com>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

This change fixes several issues found in ntuple filtering while I was
doing the ATR refactor.

Specifically I updated the masks to work correctly with the latest version
of ethtool, I cleaned up the exception handling and added detailed error
output when a filter is rejected, and corrected several bits that were set
incorrectly in ixgbe_type.h.

The previous version of this patch included a printk that was left over from
me fixing the filter setup.  This patch does not include that printk.
Signed-off-by: NAlexander Duyck <alexander.h.duyck@intel.com>
Tested-by: NStephen Ko <stephen.s.ko@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This change adds a compressed input type for atr signature hash
computation.  It also drops the use of the set functions when setting up
the ATR input since we can then directly setup the hash input as two dwords
that can be stored and passed as registers.

With these changes the cost of computing the has is low enough that we can
perform a hash computation on each TCP SYN flagged packet allowing us to
drop the number of flow director misses considerably in tests such as
netperf TCP_CRR.
Signed-off-by: NAlexander Duyck <alexander.h.duyck@intel.com>
Tested-by: NStephen Ko <stephen.s.ko@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This change cleans up the layout of the flow director data, and the
algorithm used to calculate the hash resulting in a 35x / 3500% performance
increase versus the old flow director hash computation.  The overall effect
is only a 1% increase in transactions per second though due to the fact
that only 1 packet in 20 are actually hashed upon.

TCP_RR before:
Socket Size   Request  Resp.   Elapsed  Trans.
Send   Recv   Size     Size    Time     Rate
bytes  Bytes  bytes    bytes   secs.    per sec

16384  87380  1        1       60.00    23059.27
16384  87380

TCP_RR after:
Socket Size   Request  Resp.   Elapsed  Trans.
Send   Recv   Size     Size    Time     Rate
bytes  Bytes  bytes    bytes   secs.    per sec

16384  87380  1        1       60.00    23239.98
16384  87380
Signed-off-by: NAlexander Duyck <alexander.h.duyck@intel.com>
Tested-by: NStephen Ko <stephen.s.ko@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

When disable the Rx logic globally, we would also want to disable the per Rx
queue receive logic by per queue Rx control register RXDCTL so no more DMA is
happening from the packet buffer to the receive buffer associated with the Rx
ring, before we start unmapping Rx ring receive buffer. The hardware may take
max of 100us before the corresponding Rx queue is really disabled. Added
ixgbe_disable_rx_queue() for this purpose.
Signed-off-by: NYi Zou <yi.zou@intel.com>
Tested-by: NStephen Ko <stephen.s.ko@intel.com>
Tested-by: NRoss Brattain <ross.b.brattain@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch adds support for the gigabit phys present on the CE4100 reference
platforms.
Signed-off-by: NDirk Brandewie <dirk.j.brandewie@intel.com>
Tested-by: NJeff Pieper <jeffrey.e.pieper@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

82574 needs to configure Low Power Link Up (or LPLU) differently than
the other parts in the 8257x family supported by the driver.
Signed-off-by: NBruce Allan <bruce.w.allan@intel.com>
Tested-by: NEmil Tantilov <emil.s.tantilov@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Some Phys supported by the driver do not remain powered off across a reset
of the device when the interface is down, e.g. on 82571, but not on 82574.
This patch powers down (only when WoL is disabled) the PHY after a reset if
the interface is down and the ethtool diagnostics are not currently running.

The ethtool diagnostic function required a minor re-factor as a result, and
the e1000_[get|put]_hw_control() functions are renamed since they are no
longer static to netdev.c as they are needed by the ethtool diagnostics.
A couple minor whitespace issues were cleaned up, too.
Reported-by: NArthur Jones <ajones@riverbed.com>
Signed-off-by: NBruce Allan <bruce.w.allan@intel.com>
Tested-by: NJeff Pieper <jeffrey.e.pieper@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

For the 82579 jumbo frame workaround, there is no need to re-write the CRC
calculation functionality already found in the kernel's ether_crc_le().
Signed-off-by: NBruce Allan <bruce.w.allan@intel.com>
Tested-by: NJeff Pieper <jeffrey.e.pieper@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Use string functions with bounds checking rather than their non-bounds
checking counterparts, and do not hard code these boundaries.
Signed-off-by: NBruce Allan <bruce.w.allan@intel.com>
Tested-by: NEmil Tantilov <emil.s.tantilov@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Cleans up the code a bit by using the driver-specific e1e_rphy and
e1e_wphy macros instead of the full function pointer variants.  Fix
a couple whitespace issue with two already existing calls to e1e_wphy.
Signed-off-by: NBruce Allan <bruce.w.allan@intel.com>
Tested-by: NJeff Pieper <jeffrey.e.pieper@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The ICR register is clear on read and we don't care what the returned value
is when resetting the hardware so the icr variable(s) can be removed.  We
should not ignore the return from e1000_lv_jumbo_workaround_ich8lan() and
from e1000_get_phy_id_82571() (dump a debug message when it fails and when
an unknown Phy id is returned).
Signed-off-by: NBruce Allan <bruce.w.allan@intel.com>
Tested-by: NEmil Tantilov <emil.s.tantilov@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Now that there is a single function that can compute the device
features relevant to a packet, we don't want to run it for each
offload.  This converts netif_needs_gso() to take the features
of the device, rather than computing them itself.
Signed-off-by: NJesse Gross <jesse@nicira.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

It workarounds the 60s firmware load failure timeout for the
non-modular case.
Signed-off-by: NFrancois Romieu <romieu@fr.zoreil.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The forcedeth driver uses the legacy PCI power management, so it has
to do PCI-specific things in its ->suspend() and ->resume() callbacks
and some of them are not done correctly.

Convert forcedeth to the new PCI power management framework and make
it let the PCI subsystem take care of all the PCI-specific aspects of
device handling during system power transitions.

Tested with nVidia Corporation MCP55 Ethernet (rev a2).
Signed-off-by: NRafael J. Wysocki <rjw@sisk.pl>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This converts sky2 to new VLAN offload flags control via ethtool.
It also allows for transmit offload of vlan tagged frames which
was not possible before.
Signed-off-by: NStephen Hemminger <shemminger@vyatta.com>
Reviewed-by: NJesse Gross <jesse@nicira.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The sky2 driver would always try all possible supported speeds even
if the user only asked for a limited set of speed/duplex combinations.
Reported-by: NMohsen Hariri <m.hariri@gmail.com>
Signed-off-by: NStephen Hemminger <shemminger@vyatta.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Fix the race on bp->stats_pending between the timer and a LINK_UP event
handler.
Signed-off-by: NVladislav Zolotarov <vladz@broadcom.com>
Signed-off-by: NEilon Greenstein <eilong@broadcom.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Move to D0 before clearing MSI/MSI-X configuration. Otherwise MSI/MSI-X
won't be cleared.
Signed-off-by: NVladislav Zolotarov <vladz@broadcom.com>
Signed-off-by: NEilon Greenstein <eilong@broadcom.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Fixes in registers dump:
        - Properly calculate dump length for 57712.
        - Prevent HW blocks parity attentions when dumping registers in order to
prevent false parity errors handling.
        - Update the bnx2x_dump.h file: old one had a few bugs that could cause
fatal HW error as a result of a registers dump.
Signed-off-by: NVladislav Zolotarov <vladz@broadcom.com>
Signed-off-by: NEilon Greenstein <eilong@broadcom.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Don't prevent RSS configuration in INT#x and MSI interrupt modes. Otherwise
Rx hash key won't be available.
Signed-off-by: NVladislav Zolotarov <vladz@broadcom.com>
Signed-off-by: NEilon Greenstein <eilong@broadcom.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Failure to release_firmware() in drivers/net/hamradio/yam.c::add_mcs()
causes memory leak.
This patch should fix it.
Signed-off-by: NJesper Juhl <jj@chaosbits.net>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch is to add mx28 dual fec support. Here are some key notes
for mx28 fec controller.

 - The mx28 fec controller naming ENET-MAC is a different IP from FEC
   used on other i.mx variants.  But they are basically compatible
   on software interface, so it's possible to share the same driver.
 - ENET-MAC design on mx28 made an improper assumption that it runs
   on a big-endian system. As the result, driver has to swap every
   frame going to and coming from the controller.
 - The external phys can only be configured by fec0, which means fec1
   can not work independently and both phys need to be configured by
   mii_bus attached on fec0.
 - ENET-MAC reset will get mac address registers reset too.
 - ENET-MAC MII/RMII mode and 10M/100M speed are configured
   differently FEC.
 - ETHER_EN bit must be set to get ENET-MAC interrupt work.
Signed-off-by: NShawn Guo <shawn.guo@freescale.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The following commit made a fix to use fec_enet_open/fec_enet_close
over fec_enet_init/fec_stop for suspend/resume, because fec_enet_init
does not allow to have a working network interface at resume.

  e3fe8558
  net/fec: fix pm to survive to suspend/resume

This fix works for i.mx/mxc fec controller, but fails on mx28 fec
which gets a different interrupt logic design. On i.mx fec, interrupt
can be triggered even bit ETHER_EN of ECR register is not set. But
on mx28 fec, ETHER_EN must be set to get interrupt work. Meanwhile,
MII interrupt is mandatory to resume the driver, because MDIO
read/write changed to interrupt mode by commit below.

  97b72e43
  fec: use interrupt for MDIO completion indication

fec_restart/fec_stop comes out as the solution working for both
cases.
Signed-off-by: NShawn Guo <shawn.guo@freescale.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Add mac field into fec_platform_data and consolidate function
fec_get_mac to get mac address in following order.

 1) module parameter via kernel command line fec.macaddr=0x00,0x04,...
 2) from flash in case of CONFIG_M5272 or fec_platform_data mac
    field for others, which typically have mac stored in fuse
 3) fec mac address registers set by bootloader
Signed-off-by: NShawn Guo <shawn.guo@freescale.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The "index" becomes legacy since fep->pdev->id starts working
to identify the instance.

Moreover, the call of fec_enet_init(ndev, 0) always passes 0
to fep->index. This makes the following code in fec_get_mac buggy.

	/* Adjust MAC if using default MAC address */
	if (iap == fec_mac_default)
		dev->dev_addr[ETH_ALEN-1] = fec_mac_default[ETH_ALEN-1] + fep->index;

It may be the time to remove "index" and use fep->pdev->id instead.
Signed-off-by: NShawn Guo <shawn.guo@freescale.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

FEC_MMFR_OP_WRITE should be used than FEC_MMFR_OP_READ in
a mdio write operation.

It's probably a typo introduced by commit:

e6b043d5
netdev/fec.c: add phylib supporting to enable carrier detection (v2)
Signed-off-by: NShawn Guo <shawn.guo@freescale.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Fallback on the local-mac-address prom property if the Cassini device
does not have an address programmed in the VPD ROM. This uses the same
technique as implemented by the sungem driver.

The problem was reported by Frans van Berckel using Debian kernel 2.6.34-7
on Sun Fire V440. udev was assigning a new eth<n> device name on each reboot
because the cassini driver was using a random MAC address.

Fix tested on 2.6.34-7 and 2.6.37 Sun Fire V440. Compile tested against
2.6.36 davem/sparc-2.6.git
Reported-by: NFrans van Berckel <fberckel@xs4all.nl>
Tested-by: NFrans van Berckel <fberckel@xs4all.nl>
Reviewed-by: NJulian Calaby <julian.calaby@gmail.com>
Reviewed-by: NSam Ravnborg <sam@ravnborg.org>
Signed-off-by: NRichard Mortimer <richm@oldelvet.org.uk>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NFerenc Wagner <wferi@niif.hu>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NFerenc Wagner <wferi@niif.hu>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The new firmware interface requires each Slow Path Queue (SPQ) message's
type field to include the function number.  The existing code does not
do this consistently.  We fix this by OR'ing in the function number
into the type field centrally in cnic_submit_kwqe_16().
Signed-off-by: NMichael Chan <mchan@broadcom.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Return -ENOMEM instead of 0 for the case of mdiobus_alloc and kmalloc failure.
Signed-off-by: NAxel Lin <axel.lin@gmail.com>
Acked-by: NFlorian Fainelli <florian@openwrt.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Return PTR_ERR(port->phydev) instead of 1 if phy_connect failed.
Signed-off-by: NAxel Lin <axel.lin@gmail.com>
Acked-by: NKrzysztof Halasa <khc@pm.waw.pl>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>