Currenlty multicast_set and promiscuous_config cmds -- that may be called in BH context --
use the blocking MCC mbox to post cmds.
An mbox cmd is protected via a spin_lock(cmd_lock) and not spin_lock_bh() as it is undesirable
to disable BHs while a blocking mbox cmd is in progress (and take long to finish.)
This can lockup a cmd in progress in process context.
So, these two cmds in BH context must use the MCC queue to post cmds.
Signed-off-by: Sathya Perla <sathyap@serverengines.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Currenlty all cmds use the blocking MCC mbox to post cmds. An mbox cmd is protected
via a spin_lock(cmd_lock) and not spin_lock_bh() as it is undesirable
to disable BHs while a blocking mbox cmd is in progress (and take long to finish.)
This can lockup a cmd in progress in process context. Instead cmds that may be
called in BH context must use the MCC queue to post cmds. The cmd completions
are rcvd in a separate completion queue and the events are placed in the tx-event
queue.
Signed-off-by: Sathya Perla <sathyap@serverengines.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This fixes various endianness bugs. Some harmless and some real ones.
This is tested on a PowerPC-64 machine.
Signed-off-by: Michael Buesch <mb@bu3sch.de>
Cc: Stable <stable@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

If the iucv message limit for a communication path is exceeded,
sendmsg() returns -EAGAIN instead of -EPIPE.
The calling application can then handle this error situtation,
e.g. to try again after waiting some time.

For blocking sockets, sendmsg() waits up to the socket timeout
before returning -EAGAIN. For the new wait condition, a macro
has been introduced and the iucv_sock_wait_state() has been
refactored to this macro.
Signed-off-by: Hendrik Brueckner <brueckner@linux.vnet.ibm.com>
Signed-off-by: Ursula Braun <ursula.braun@de.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Change the if condition to exit sendmsg() if the socket in not connected.
Signed-off-by: Hendrik Brueckner <brueckner@linux.vnet.ibm.com>
Signed-off-by: Ursula Braun <ursula.braun@de.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This forces every update of tx ring producer to check for
availability of space for next full TSO command. Earlier
firmware control commands didn't care to pause tx queue.

Stop the tx queue if there's not enough space to transmit one full
LSO command left on the tx ring after current transmit. This avoids
returning NETDEV_TX_BUSY after checking distance between producer
and consumer on every cpu.

Restart the tx queue only if we have cleaned up enough tx
descriptors.
Signed-off-by: Dhananjay Phadke <dhananjay@netxen.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Fix the detection of cut-thru mode of the hardware (direct dma
to host) to mode configured in SRE (ingress block) rather than
onboard memory control.
Signed-off-by: Dhananjay Phadke <dhananjay@netxen.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

forcedeth doesnt use properly dma api in its tx completion path
and in nv_loopback_test()

pci_map_single() should be paired with pci_unmap_single()
pci_map_page() should be paired with pci_unmap_page()

forcedeth xmit path uses pci_map_single() & pci_map_page(),
but tx completion path only uses pci_unmap_single()

nv_loopback_test() uses pci_map_single() & pci_unmap_page()

Add a dma_single field in struct nv_skb_map, and
define a helper function nv_unmap_txskb
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
CC: Ayaz Abdulla <aabdulla@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

commit 2b85a34e
(net: No more expensive sock_hold()/sock_put() on each tx)
changed initial sk_wmem_alloc value.

This broke net/atm since this protocol assumed a null
initial value. This patch makes necessary changes.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

commit 2b85a34e
(net: No more expensive sock_hold()/sock_put() on each tx)
changed initial sk_wmem_alloc value.

We need to take into account this offset when reporting
sk_wmem_alloc to user, in PROC_FS files or various
ioctls (SIOCOUTQ/TIOCOUTQ)
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

on boot, link is always up.
Signed-off-by: Seguier Regis <rseguier@e-teleport.net>
Signed-off-by: David S. Miller <davem@davemloft.net>

If the adapter is not power-manageable using either ACPI, or the
native PCI PM interface, __e100_power_off() returns error code, which
causes every attempt to suspend to fail, although it should return 0
in such a case.  Fix this problem by ignoring the return value of
pci_set_power_state() in __e100_power_off().
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
Acked-by: Andreas Mohr <andi@lisas.de>
Signed-off-by: David S. Miller <davem@davemloft.net>

clk_disable was called twice in the remove function.
Correct this so that the driver module unloads without error.
Signed-off-by: Chaithrika U S <chaithrika@ti.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This patch is inspired by patch recently posted by Johannes Berg. Basically what
my patch does is to group list and a count of addresses into newly introduced
structure netdev_hw_addr_list. This brings us two benefits:
1) struct net_device becames a bit nicer.
2) in the future there will be a possibility to operate with lists independently
   on netdevices (with exporting right functions).
I wanted to introduce this patch before I'll post a multicast lists conversion.
Signed-off-by: Jiri Pirko <jpirko@redhat.com>

 drivers/net/bnx2.c              |    4 +-
 drivers/net/e1000/e1000_main.c  |    4 +-
 drivers/net/ixgbe/ixgbe_main.c  |    6 +-
 drivers/net/mv643xx_eth.c       |    2 +-
 drivers/net/niu.c               |    4 +-
 drivers/net/virtio_net.c        |   10 ++--
 drivers/s390/net/qeth_l2_main.c |    2 +-
 include/linux/netdevice.h       |   17 +++--
 net/core/dev.c                  |  130 ++++++++++++++++++--------------------
 9 files changed, 89 insertions(+), 90 deletions(-)
Signed-off-by: David S. Miller <davem@davemloft.net>

My previous patch, which explicitly delays freeing of tnodes by adding
them to the list to flush them after the update is finished, isn't
strict enough. It treats exceptionally tnodes without parent, assuming
they are newly created, so "invisible" for the read side yet.

But the top tnode doesn't have parent as well, so we have to exclude
all exceptions (at least until a better way is found). Additionally we
need to move rcu assignment of this node before flushing, so the
return type of the trie_rebalance() function is changed.
Reported-by: Yan Zheng <zheng.yan@oracle.com>
Signed-off-by: Jarek Poplawski <jarkao2@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Action police statistics could be misleading because drops are not
shown when expected.

With feedback from: Jamal Hadi Salim <hadi@cyberus.ca>
Reported-by: Pawel Staszewski <pstaszewski@itcare.pl>
Signed-off-by: Jarek Poplawski <jarkao2@gmail.com>
Acked-by: Jamal Hadi Salim <hadi@cyberus.ca>
Signed-off-by: David S. Miller <davem@davemloft.net>

Version bump.
Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Add support for generic receive offload.
Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This patch implements skb recycling. It reclaims transmitted skb's
for use in the receive ring.
Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Reduce the size of the driver transmit ring to reduce latency
and allow qdisc to do better rate control.  Also make it
obvious what the minimum transmit ring allowed is and why.
Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Since it is likely that there are multiple packets received per
interrupt, only update the receive counters once after all
packets are processed.
Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The logic in sky2_down was incorrect. Receiver could report status
after rx_stop was called.

The steps need to be:
   * stop new frames from being transmitted
   * shut off transmit/receive logic
   * synchronize with NAPI to process status info about transmitter
     and receiver
Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Add some read's to avoid any PCI posting issues when controlling
irq's.
Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Reset more parts of the receive path when device is take offline.
Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This unblocks the chip if it is stuck in pause cycle during
shutdown.
Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Stopping all activity through ChipCmd and blindly acking the irqs
is neither nice nor completely needed: the transition to low-power
mode does enough work and it apparently keeps the device in a sane
state.

Patch suggested by a fix for http://bugzilla.kernel.org/show_bug.cgi?id=9512

The rtl_shutdown path is kept unchanged so far.
Signed-off-by: Francois Romieu <romieu@fr.zoreil.com>
Tested-by: Anders Eriksson <aeriksson@fastmail.fm>
Cc: Edward Hsu <edward_hsu@realtek.com.tw>
Signed-off-by: David S. Miller <davem@davemloft.net>

When sis190 driver is trying to get default phy, if it doesn't find home
or lan phy, it falls back to the first phy in the phy list but list_entry()
points to a bogus entry. list_first_entry() should be used instead.
Signed-off-by: Arnaud Patard <apatard@mandriva.com>
Acked-off-by: Francois Romieu <romieu@fr.zoreil.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

-- derived from reverted commit 047584ce
-- reworked by Grant Likely to play nice with commit:
   "net: Rework ucc_geth driver to use of_mdio infrastructure"
   (0b9da337)
Signed-off-by: Haiying Wang <Haiying.Wang@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
Signed-off-by: David S. Miller <davem@davemloft.net>

This reverts commit 047584ce.

This patch meshes badly with "net: Rework ucc_geth driver to use
of_mdio infrastructure" (0b9da337).
Since most of the patch needs to be reworked, it is clearer to revert
the patch and then apply the corrected version
Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
Signed-off-by: David S. Miller <davem@davemloft.net>

The skb mac_header field is sometimes NULL (or ~0u) as a sentinel
value. The places where skb is expanded add an offset which would
change this flag into an invalid pointer (or offset).
Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Looking at the crash in log_martians(), one suspect is that the check for
mac header being set is not correct.  The value of mac_header defaults to
0 on allocation, therefore skb_mac_header_was_set will always be true on
platforms using NET_SKBUFF_USES_OFFSET.
Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Acked-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

commit 2b85a34e
(net: No more expensive sock_hold()/sock_put() on each tx)
changed initial sk_wmem_alloc value.

Some protocols check sk_wmem_alloc value to determine if a timer
must delay socket deallocation. We must take care of the sk_wmem_alloc
value being one instead of zero when no write allocations are pending.

Reported by Ingo Molnar, and full diagnostic from David Miller.

This patch introduces three helpers to get read/write allocations
and a followup patch will use these helpers to report correct
write allocations to user.
Reported-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Fix kernel-doc parameter name in phy_device.c.
Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Fix up remaining drivers returning a magic or an errno value from their
ndo_start_xmit() functions that were missed in the first pass:

- isdn_net: missed conversion
- bpqether: missed conversion: skb is freed, so return NETDEV_TX_OK
- hp100: intention appears to be to resubmit skb once resources are
  available, but due to no queue handling it is dropped for now.
- lapbether: skb is freed, so return NETDEV_TX_OK
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>

- Enabled SRIOV support in the driver.
- Call __vxge_hw_verify_pci_e_info() for the PF only. This function
  verifies the negotiated link width and current link speed in the
  Link Status Register (offset 12h) which are reserved fields for VFs
  as per the SRIOV specification, section 3.5.8.
- Implemented David Miller's comment to remove the #ifdef CONFIG_PCI_IOV as
  these intefaces have NOP versions declared when the defintion is not set.
Signed-off-by: Sivakumar Subramani <sivakumar.subramani@neterion.com>
Signed-off-by: Rastapur Santosh <santosh.rastapur@neterion.com>
Signed-off-by: Ramkrishna Vepa <ram.vepa@neterion.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Fix kernel-doc warnings (missing + extra entries) in skbuff.h.
Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Noticed by Stephen Rothwell.
Signed-off-by: David S. Miller <davem@davemloft.net>

If socket destuction gets delayed to a timer, we try to
lock_sock() from that timer which won't work.

Use bh_lock_sock() in that case.
Signed-off-by: David S. Miller <davem@davemloft.net>
Tested-by: Ingo Molnar <mingo@elte.hu>

The code to compute VPD size didn't handle some systems that use
chip without VPD. Also some of the newer chips use some additional
registers to store the actual size, and wasn't worth putting the
additional complexity in, so just remove the code.

No big loss since the code to set the VPD size was only a
convenience so that utilities would not read the extra space past
the end of the available VPD.

Move the first PCI config read earlier to detect bad hardware
where it returns all ones and refuse loading driver before furthur
damage.
Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Tested-by: Andy Whitcroft <apw@canonical.com>
Signed-off-by: David S. Miller <davem@davemloft.net>