Commit ee49bd93 ("mlx4_core: Reset device when internal error is
detected") introduced some section mismatch problems when
CONFIG_HOTPLUG=n, because the error recovery code tears down and
reinitializes the device after everything is loaded, which ends up
calling into lots of code marked __devinit and __devexit from regular
.text.  Fix this by getting rid of these now-incorrect section
markers.
Signed-off-by: NRoland Dreier <rolandd@cisco.com>

Firmware commands are sent to the HCA by writing multiple words to a
command register block.  Access to this block of registers is
serialized with a mutex.  However, on large SGI systems writes to the
register block may be reordered within the system interconnect and
reach the HCA in a different order than they were issued (even with
the mutex).  Fix this by adding an mmiowb() before dropping the mutex.

This bug was observed with real workloads with the similar FW command
code in the mthca driver, and adding the mmiowb() as in commit
66547550 ("IB/mthca: Use mmiowb() to avoid firmware commands getting
jumbled up") was confirmed to fix the problems, so we should add the
same fix to mlx4.
Signed-off-by: NRoland Dreier <rolandd@cisco.com>

Increase the number of QPs allowed per multicast group from 8 to 56.
This allows for one QP per core on 16-core systems, which are now
quite common, and allows some space for future growth.
Signed-off-by: NJack Morgenstein <jackm@dev.mellanox.co.il>
Signed-off-by: NRoland Dreier <rolandd@cisco.com>

Implement FMRs for mlx4.  This is an adaptation of code from mthca.
Signed-off-by: NJack Morgenstein <jackm@dev.mellanox.co.il>
Signed-off-by: NMichael S. Tsirkin <mst@dev.mellanox.co.il>
Signed-off-by: NRoland Dreier <rolandd@cisco.com>

Write MTT entries directly to ICM from the driver (eliminating use of
WRITE_MTT command).  This reduces the number of FW commands needed to
register an MR by at least a factor of 2 and speeds up memory
registration significantly.  This code will also be used to implement
FMRs.
Signed-off-by: NJack Morgenstein <jackm@dev.mellanox.co.il>
Signed-off-by: NMichael S. Tsirkin <mst@dev.mellanox.co.il>
Signed-off-by: NRoland Dreier <rolandd@cisco.com>

Everything that uses caps.reserved_mtts expects it to be a count of MTT
segments, not MTT entries.  So convert the value that the FW gives us to
a count of segments.
Signed-off-by: NRoland Dreier <rolandd@cisco.com>

Taking ilog2(dev->caps.reserved_mtts) to find out the order to pass to
the MTT buddy allocator will do the wrong thing if reserved_mtts is ever
not a power of 2.  Be safe and use fls(dev->caps.reserved_mtts - 1).
Signed-off-by: NRoland Dreier <rolandd@cisco.com>

Enable having ICM tables in coherent memory, and use coherent memory
for the dMPT table.  This will allow writing MPT entries for MRs both
via the SW2HW_MPT command and also directly by the driver for FMR
remapping without needing to flush or worry about cacheline boundaries.
Signed-off-by: NJack Morgenstein <jackm@dev.mellanox.co.il>
Signed-off-by: NMichael S. Tsirkin <mst@dev.mellanox.co.il>
Signed-off-by: NRoland Dreier <rolandd@cisco.com>

display the following device information under /sys/class/infiniband/mlx4_X:
board_id, fw_ver, hw_rev, hca_type.

This patch makes this information available to userspace utilities
such as ibstat and ibv_devinfo.
Signed-off-by: NJack Morgenstein <jackm@dev.mellanox.co.il>
Signed-off-by: NRoland Dreier <rolandd@cisco.com>

The SRC ("scalable RC") transport has been renamed to XRC ("extended 
RC"), to avoid having an abbreviation that is so easily confused with an 
abbreviation for "source."  Update the HCA capability decoding output to 
use the new name.
Signed-off-by: NRoland Dreier <rolandd@cisco.com>

mlx4_srq_query() returns a big-endian 16-bit value through an int *,
which screws up sparse checking.  Fix this so that a CPU-endian value
is returned.
Signed-off-by: NRoland Dreier <rolandd@cisco.com>

Recover from MSI-X errors by automatically falling back on regular
interrupt, instead of asking the user to do this manually.  This makes
it possible to enable MSI-X by default, and will make it possible to
get rid of the msi_x module option in the future.
Signed-off-by: NMichael S. Tsirkin <mst@dev.mellanox.co.il>
Signed-off-by: NRoland Dreier <rolandd@cisco.com>

Special QPs are not allocated using the regular QP number bitmap, so
when they are destroyed, their QP number should not be freed in the
bitmap.

Found by Dotan Barak of Mellanox.
Signed-off-by: NRoland Dreier <rolandd@cisco.com>

Rename GO_BIT_TIMEOUT to GO_BIT_TIMEOUT_MSECS for clarity, and
actually use it as the go bit timeout (instead of having the define
but then ignoring it and using a hard-coded 10 * HZ for the actual
timeout).
Signed-off-by: NDotan Barak <dotanb@dev.mellanox.co.il>
Signed-off-by: NRoland Dreier <rolandd@cisco.com>

The 8169/8110SC currently announces itself as:
[...]
eth0: RTL8169sc/8110sc at 0x........, ..:..:..:..:..:.., XID 18000000 IRQ ..
                                                             ^^^^^^^^
It uses RTL_GIGA_MAC_VER_05 and this part of the changeset can cut
its performance by a factor of 2~2.5 as reported by Timo.

(the driver includes code just before the hunk to write the ChipCmd
register when mac_version == RTL_GIGA_MAC_VER_0[1-4])
Signed-off-by: NFrancois Romieu <romieu@fr.zoreil.com>
Cc: Timo Jantunen <jeti@welho.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

Remove unneeded check that caused problems with jumbo frame sizes.
The check was recently added and is wrong.
When using jumbo frames the sky2 driver does fragmentation, so
rx_data_size is less than mtu.
Signed-off-by: NStephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

Correct printk with PFX before KERN_ in bcm43xx_wx.c
Signed-off-by: NJoe Perches <joe@perches.com>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

dm9601 didn't take the ethernet header into account when calculating
RX MTU, causing packets bigger than 1486 to fail.
Signed-off-by: NPeter Korsgaard <jacmet@sunsite.dk>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

This driver erroneously zeros dev->tx_queue_len, since
mp->tx_ring_size has not yet been initialized.  Actually,
the driver shouldn't modify tx_queue_len at all and should
leave the value set by alloc_etherdev(), currently 1000.
Signed-off-by: NDale Farnsworth <dale@farnsworth.org>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

Fix 4032 chip undocumented "feature" where bit-8 is set
if the inbound completion is for a VLAN.
Signed-off-by: NRon Mercer <ron.mercer@qlogic.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

Signed-off-by: NRon Mercer <ron.mercer@qlogic.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

Fix missing symbols in libertas USB driver when it is modular and rest
of libertas is built-in.
Signed-off-by: NRandy Dunlap <randy.dunlap@oracle.com>
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>

Commit 468d09f8 masked the "state"
interrupt (bit 20 of the cause register). This results in Radstone's
PPC7D repeatedly re-entering the interrupt routine, locking up the
board. The following patch returns the required handling for this
interrupt.
Signed-off-by: NMartyn Welch <martyn.welch@radstone.co.uk>
Signed-off-by: NDale Farnsworth <dale@farnsworth.org>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

This blade-specific board form factor is identical to the 82571EB
board.
Signed-off-by: NAuke Kok <auke-jan.h.kok@intel.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

This should fix http://bugzilla.kernel.org/show_bug.cgi?id=8667

After resume, driver has reset the chip so the current state
of transmit checksum offload state machine and DMA state machine
will be undefined.

The fix is to set the state so that first Tx will set MSS and offset
values.
Signed-off-by: NStephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

The FE+ workaround means the driver can no longer trust the status register
to indicate VLAN tagged frames.  The fix for this is to just disable VLAN
acceleration for that chip version. Tested and works fine.
Signed-off-by: NStephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

The Yukon FE+ chip appears to have a hardware glitch that causes bogus
receive status values to be posted. The data in the packet is good, but
the status value is random garbage.  As a temporary workaround until the
problem is better understood, implement the workaround the vendor driver
used of ignoring the status value on this chip.

Since this means trusting dodgy hardware values; add additional checking
of the receive packet length.
Signed-off-by: NStephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Signed-off-by: NMichal Schmidt <mschmidt@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This reverts commit fadacb1b.

The change being reverted made the driver consistent with
include/linux/netdevice.h, but then inconsistent with the other PCMCIA
ethernet drivers.
Signed-off-by: NJeff Garzik <jeff@garzik.org>

Be more selective about when to enable the ram buffer watchdog code.
It is unnecessary on XL A3 or later revs, and with Yukon FE
the buffer is so small (4K) that the watchdog detects false positives.
Signed-off-by: NStephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

One more snippet of PHY initialization required for FE+ chips.
Discovered in latest sk98lin 10.21.1.3 driver.

Please apply to 2.6.23.
Signed-off-by: NStephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

Update version number
Signed-off-by: NStephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

A driver writer from another operating system hinted that
the versions of Yukon 2 chip with rambuffer (EC and XL) have
a hardware bug that if the FIFO ever gets completely full it
will hang. Sounds like a classic ring full vs ring empty wrap around
bug.

As a workaround, use the existing watchdog timer to check for
ring full lockup.
Signed-off-by: NStephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

Add support for newest Marvell chips.
The Yukon FE plus chip is found in some not yet released laptops.
Tested on hardware evaluation boards.

This version of the patch is for 2.6.23. It supersedes
the two previous patches that are sitting in netdev-2.6 (upstream branch).
Signed-off-by: NStephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

This patch should cause no functional changes in driver behaviour.
There are (too) many revisions of the Yukon 2 chip now. Instead of
adding more conditionals based on chip revision; rerganize into a
set of feature flags so adding new versions is less problematic.
Signed-off-by: NStephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

On 100mbit versions, the driver always reports gigabit speed
available. The correct modes are already computed, then overwritten.
Signed-off-by: NStephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

The length check for truncated frames was not correctly handling
the case where VLAN acceleration had already read the tag.
Also, the Yukon EX has some features that use high bit of status
as security tag.
Signed-off-by: NPierre-Yves Ritschard <pyr@spootnik.org>
Signed-off-by: NStephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

Add the DIS_EARLY_DAC PHY workaround for 5709 A1.  Without it, link
sometimes does not come up.

Update version to 1.6.5.
Signed-off-by: NMichael Chan <mchan@broadcom.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch makes pppol2tp_xmit call skb_cow_head so that we don't modify
cloned skb data.  It also gets rid of skb2 we only need to preserve the
original skb for congestion notification, which is only applicable for
ppp_async and ppp_sync.

The other semantic change made here is the removal of socket accounting
for data tranmitted out of pppol2tp_xmit.  The original code leaked any
existing socket skb accounting.  We could fix this by dropping the
original skb owner.  However, this is undesirable as the packet has not
physically left the host yet.

In fact, all other tunnels in the kernel do not account skb's passing
through to their own socket.  In partciular, ESP over UDP does not do
so and it is the closest tunnel type to PPPoL2TP.  So this patch simply
removes the socket accounting in pppol2tp_xmit.  The accounting still
applies to control packets of course.

I've also added a reminder that the outgoing checksum here doesn't work.
I suppose existing deployments don't actually enable checksums.
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>