Core of the problem is that phy_suspend() suspends the PHY when it
should not because of WoL. phy_suspend() checks for WoL already, but
this works only if the PHY driver handles WoL (what is rarely the case).
Typically WoL is handled by the MAC driver.
This patch uses new member wol_enabled of struct net_device as
additional criteria in the check when not to suspend the PHY because
of WoL.
Last but not least change phy_detach() to call phy_suspend() before
attached_dev is set to NULL. phy_suspend() accesses attached_dev
when checking whether the MAC driver activated WoL.
Fixes: f1e911d5d0 ("r8169: add basic phylib support")
Fixes: e8cfd9d6c7 ("net: phy: call state machine synchronously in phy_stop")
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Actually there's nothing wrong with the two changes marked as "Fixes",
they just revealed a problem which has been existing before.
After having switched r8169 to phylib it was reported that WoL from
shutdown doesn't work any longer (WoL from suspend isn't affected).
Reason is that during shutdown phy_disconnect()->phy_detach()->
phy_suspend() is called.
A similar issue occurs when the phylib state machine calls
phy_suspend() when handling state PHY_HALTED.
Core of the problem is that phy_suspend() suspends the PHY when it
should not due to WoL. phy_suspend() checks for WoL already, but this
works only if the PHY driver handles WoL (what is rarely the case).
Typically WoL is handled by the MAC driver.
phylib knows about this and handles it in mdio_bus_phy_may_suspend(),
but that's used only when suspending the system, not in other cases
like shutdown.
Therefore factor out the relevant check from
mdio_bus_phy_may_suspend() to a new function phy_may_suspend() and
use it in phy_suspend().
Last but not least change phy_detach() to call phy_suspend() before
attached_dev is set to NULL. phy_suspend() accesses attached_dev
when checking whether the MAC driver activated WoL.
Fixes: f1e911d5d0 ("r8169: add basic phylib support")
Fixes: e8cfd9d6c7 ("net: phy: call state machine synchronously in phy_stop")
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The method ndo_start_xmit() is defined as returning an 'netdev_tx_t',
which is a typedef for an enum type, so make sure the implementation in
this driver has returns 'netdev_tx_t' value, and change the function
return type to netdev_tx_t.
Found by coccinelle.
Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The method ndo_start_xmit() is defined as returning an 'netdev_tx_t',
which is a typedef for an enum type, so make sure the implementation in
this driver has returns 'netdev_tx_t' value, and change the function
return type to netdev_tx_t.
Found by coccinelle.
Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Trival cleanup, list_move_tail will implement the same function that
list_del() + list_add_tail() will do. hence just replace them.
Signed-off-by: zhong jiang <zhongjiang@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Trival cleanup, list_move_tail will implement the same function that
list_del() + list_add_tail() will do. hence just replace them.
Signed-off-by: zhong jiang <zhongjiang@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The allocation of hwsim radio identifiers uses a post-increment from 0,
so the first radio has idx 0. This idx is explicitly excluded from
multicast announcements ever since, but it is unclear why.
Drop that idx check and announce the first radio as well. This makes
userspace happy if it relies on these events.
Signed-off-by: Martin Willi <martin@strongswan.org>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
The asynchronous destruction from a work-queue of radios tagged with
destroy-on-close may race with the owning namespace about to exit,
resulting in potential use-after-free of that namespace.
Instead of using a work-queue, move radios about to destroy to a
temporary list, which can be worked on synchronously after releasing
the lock. This should be safe to do from the netlink socket notifier,
as the namespace is guaranteed to not get released.
Signed-off-by: Martin Willi <martin@strongswan.org>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
The cleanup of radios during namespace exit has recently been reworked
to directly delete a radio while temporarily releasing the spinlock,
fixing a race condition between the work-queue execution and namespace
exits. However, the temporary unlock allows unsafe modifications on the
iterated list, resulting in a potential crash when continuing the
iteration of additional radios.
Move radios about to destroy to a temporary list, and clean that up
after releasing the spinlock once iteration is complete.
Fixes: 8cfd36a0b5 ("mac80211_hwsim: fix use-after-free bug in hwsim_exit_net")
Signed-off-by: Martin Willi <martin@strongswan.org>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Local variable 'autoneg' doesn't even exist:
drivers/net/phy/marvell.c: In function 'm88e1121_config_aneg':
drivers/net/phy/marvell.c:468:25: error: 'autoneg' undeclared (first use in this function); did you mean 'put_net'?
if (phydev->autoneg != autoneg || changed) {
^~~~~~~
Fixes: d6ab933647 ("net: phy: marvell: Avoid unnecessary soft reset")
Reported-by:Vakul Garg <vakul.garg@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The driver uses devm_ioremap_resource() which is only available when
CONFIG_HAS_IOMEM is set, make the driver depend on this config option.
User mode Linux does not have CONFIG_HAS_IOMEM set and the driver was
failing on this architecture.
Fixes: 14fceff477 ("net: dsa: Add Lantiq / Intel DSA driver for vrx200")
Reported-by: kbuild test robot <lkp@intel.com>
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
The BMCR.RESET bit on the Marvell PHYs has a special meaning in that
it commits the register writes into the HW for it to latch and be
configured appropriately. Doing software resets causes link drops, and
this is unnecessary disruption if nothing changed.
Determine from marvell_set_polarity()'s return code whether the register value
was changed and if it was, propagate that to the logic that hits the software
reset bit.
This avoids doing unnecessary soft reset if the PHY is configured in
the same state it was previously, this also eliminates the need for a
m88e1111_config_aneg() function since it now is the same as
marvell_config_aneg().
Tested-by: Wang, Dongsheng <dongsheng.wang@hxt-semitech.com>
Tested-by: Chris Healy <cphealy@gmail.com>
Tested-by: Andrew Lunn <andrew@lunn.ch>
Tested-by: Clemens Gruber <clemens.gruber@pqgruber.com>
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
While consolidating the PHY reset in phy_init_hw() an unconditionaly
BMCR soft-reset I became quite trigger happy with those. This was later
on deactivated for the Generic PHY driver on the premise that a prior
software entity (e.g: bootloader) might have applied workarounds in
commit 0878fff1f4 ("net: phy: Do not perform software reset for
Generic PHY").
Since we have a hook to wire-up a soft_reset callback, just use that and
get rid of the call to genphy_soft_reset() entirely. This speeds up
initialization and link establishment for most PHYs out there that do
not require a reset.
Fixes: 87aa9f9c61 ("net: phy: consolidate PHY reset in phy_init_hw()")
Tested-by: Wang, Dongsheng <dongsheng.wang@hxt-semitech.com>
Tested-by: Chris Healy <cphealy@gmail.com>
Tested-by: Andrew Lunn <andrew@lunn.ch>
Tested-by: Clemens Gruber <clemens.gruber@pqgruber.com>
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Jeff Kirsher says:
====================
40GbE Intel Wired LAN Driver Updates 2018-09-25
This series contains updates to i40e and xsk.
Mariusz fixes an issue where the VF link state was not being updated
properly when the PF is down or up. Also cleaned up the promiscuous
configuration during a VF reset.
Patryk simplifies the code a bit to use the variables for PF and HW that
are declared, rather than using the VSI pointers. Cleaned up the
message length parameter to several virtchnl functions, since it was not
being used (or needed).
Harshitha fixes two potential race conditions when trying to change VF
settings by creating a helper function to validate that the VF is
enabled and that the VSI is set up.
Sergey corrects a double "link down" message by putting in a check for
whether or not the link is up or going down.
Björn addresses an AF_XDP zero-copy issue that buffers passed
from userspace to the kernel was leaked when the hardware descriptor
ring was torn down. A zero-copy capable driver picks buffers off the
fill ring and places them on the hardware receive ring to be completed at
a later point when DMA is complete. Similar on the transmit side; The
driver picks buffers off the transmit ring and places them on the
transmit hardware ring.
In the typical flow, the receive buffer will be placed onto an receive
ring (completed to the user), and the transmit buffer will be placed on
the completion ring to notify the user that the transfer is done.
However, if the driver needs to tear down the hardware rings for some
reason (interface goes down, reconfiguration and such), the userspace
buffers cannot be leaked. They have to be reused or completed back to
userspace.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
When an AF_XDP UMEM is attached to any of the Rx rings, we disallow a
user to change the number of descriptors via e.g. "ethtool -G IFNAME".
Otherwise, the size of the stash/reuse queue can grow unbounded, which
would result in OOM or leaking userspace buffers.
Signed-off-by: Björn Töpel <bjorn.topel@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Outstanding Rx descriptors are temporarily stored on a stash/reuse
queue. When/if the HW rings comes up again, entries from the stash are
used to re-populate the ring.
The latter required some restructuring of the allocation scheme for
the AF_XDP zero-copy implementation. There is now a fast, and a slow
allocation. The "fast allocation" is used from the fast-path and
obtains free buffers from the fill ring and the internal recycle
mechanism. The "slow allocation" is only used in ring setup, and
obtains buffers from the fill ring and the stash (if any).
Signed-off-by: Björn Töpel <bjorn.topel@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
When the zero-copy enabled XDP Tx ring is torn down, due to
configuration changes, outstanding frames on the hardware descriptor
ring are queued on the completion ring.
The completion ring has a back-pressure mechanism that will guarantee
that there is sufficient space on the ring.
Signed-off-by: Björn Töpel <bjorn.topel@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
When we are trying to change VF settings, it is possible for 2 race
conditions to happen. One, when the VF is created but not yet enabled.
Second, the VF is enabled but the VSI is still not created or not yet
re-created in the VF reset flow.
This patch introduces a helper function to validate that the VF is
enabled and that the VSI is set up. This patch also calls this
function from other functions which could get into these race conditions.
While we are poking around here, remove unnecessary parenthesis that
checkpatch was complaining about.
Signed-off-by: Harshitha Ramamurthy <harshitha.ramamurthy@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
This patch cleans up promiscuous configuration when a VF reset occurs.
Previously the promiscuous mode settings were still there after the VF
driver removal.
Signed-off-by: Mariusz Stachura <mariusz.stachura@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
This resolves an issue where the VF link state was not being updated
when the PF is down or up, and the VF link state would always show
that it is running.
Signed-off-by: Mariusz Stachura <mariusz.stachura@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
If SMMU is on, there is more likely that skb_shinfo(skb)->frags[i]
can not send by a single BD. when this happen, the
hns_nic_net_xmit_hw function map the whole data in a frags using
skb_frag_dma_map, but unmap each BD' data individually when tx is
done, which causes problem when SMMU is on.
This patch fixes this problem by ummapping the whole data in a
frags when tx is done.
Signed-off-by: Yunsheng Lin <linyunsheng@huawei.com>
Signed-off-by: Peng Li <lipeng321@huawei.com>
Reviewed-by: Yisen Zhuang <yisen.zhuang@huawei.com>
Signed-off-by: Salil Mehta <salil.mehta@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
There's no guarantee that the mapping array doesn't cross a page
boundary. Use a second grant copy operation if necessary.
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Wei Liu <wei.liu2@citrix.com>
Reviewed-by: Paul Durrant <paul.durrant@citrix.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Checking them before the grant copy means nothing as to the validity of
the incoming request. As we shouldn't make the new data live before
having validated it, introduce a second instance of the mapping array.
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Paul Durrant <paul.durrant@citrix.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Both len and off are frontend specified values, so we need to make
sure there's no overflow when adding the two for the bounds check. We
also want to avoid undefined behavior and hence use off to index into
->hash.mapping[] only after bounds checking. This at the same time
allows to take care of not applying off twice for the bounds checking
against vif->num_queues.
It is also insufficient to bounds check copy_op.len, as this is len
truncated to 16 bits.
This is XSA-270 / CVE-2018-15471.
Reported-by: Felix Wilhelm <fwilhelm@google.com>
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Paul Durrant <paul.durrant@citrix.com>
Tested-by: Paul Durrant <paul.durrant@citrix.com>
Cc: stable@vger.kernel.org [4.7 onwards]
Signed-off-by: David S. Miller <davem@davemloft.net>
Version bump conflict in batman-adv, take what's in net-next.
iavf conflict, adjustment of netdev_ops in net-next conflicting
with poll controller method removal in net.
Signed-off-by: David S. Miller <davem@davemloft.net>
Clear ADDR64 dma bit in DMACFG register in case that HW_DMA_CAP_64B is
not detected on 64bit system.
The issue was observed when bootloader(u-boot) does not check macb
feature at DCFG6 register (DAW64_OFFSET) and enabling 64bit dma support
by default. Then macb driver is reading DMACFG register back and only
adding 64bit dma configuration but not cleaning it out.
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
Acked-by: Nicolas Ferre <nicolas.ferre@microchip.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Clear ADDR64 dma bit in DMACFG register in case that HW_DMA_CAP_64B is
not detected on 64bit system.
The issue was observed when bootloader(u-boot) does not check macb
feature at DCFG6 register (DAW64_OFFSET) and enabling 64bit dma support
by default. Then macb driver is reading DMACFG register back and only
adding 64bit dma configuration but not cleaning it out.
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
Acked-by: Nicolas Ferre <nicolas.ferre@microchip.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The check for pci_is_pcie() is redundant here because all
chip versions >=18 are PCIe only anyway. In addition use
dma_set_mask_and_coherent() instead of separate calls to
pci_set_dma_mask() and pci_set_consistent_dma_mask().
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Code can be slightly simplified by acking even events we're not
interested in. In addition add a comment making clear that the
read has no functional purpose and is just a PCI commit.
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The networking core has a default watchdog timeout of 5s. I see no
need to define an own timeout of 6s which is basically the same.
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Set uid as part of DCT commands so that the firmware can manage the
DCT object in a secured way.
That will enable using a DCT that was created by verbs application
to be used by the DEVX flow in case the uid is equal.
Signed-off-by: Yishai Hadas <yishaih@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Set uid as part of SRQ commands so that the firmware can manage the
SRQ object in a secured way.
That will enable using an SRQ that was created by verbs application
to be used by the DEVX flow in case the uid is equal.
Signed-off-by: Yishai Hadas <yishaih@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Set uid as part of SQ commands so that the firmware can manage the
SQ object in a secured way.
That will enable using an SQ that was created by verbs application
to be used by the DEVX flow in case the uid is equal.
Signed-off-by: Yishai Hadas <yishaih@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Set uid as part of RQ commands so that the firmware can manage the
RQ object in a secured way.
That will enable using an RQ that was created by verbs application
to be used by the DEVX flow in case the uid is equal.
Signed-off-by: Yishai Hadas <yishaih@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Set uid as part of QP commands so that the firmware can manage the
QP object in a secured way.
That will enable using a QP that was created by verbs application to
be used by the DEVX flow in case the uid is equal.
Signed-off-by: Yishai Hadas <yishaih@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Set uid as part of CQ commands so that the firmware can manage the CQ
object in a secured way.
The firmware should mark this CQ with the given uid so that it can
be used later on only by objects with the same uid.
Upon DEVX flows that use this CQ (e.g. create QP command), the
pointed CQ must have the same uid as of the issuer uid command.
When a command is issued with uid=0 it means that the issuer of the
command is trusted (i.e. kernel), in that case any pointed object
can be used regardless of its uid.
Signed-off-by: Yishai Hadas <yishaih@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Until now, the Rx flow hash key was a 5-tuple (IP src, IP dst,
IP nextproto, L4 src port, L4 dst port) fixed value that we
configured at probe.
Add support for configuring this hash key at runtime.
We support all standard header fields configurable through ethtool,
but cannot differentiate between flow types, so the same hash key
is applied regardless of protocol.
We also don't support the discard option.
Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
With CONFIG_DMA_API_DEBUG enabled we get DMA unmapping warning in
various places of the mvneta driver, for example when putting down an
interface while traffic is passing through.
The issue is when using s/w buffer management, the Rx buffers are mapped
using dma_map_page but unmapped with dma_unmap_single. This patch fixes
this by using the right unmapping function.
Fixes: 562e2f467e ("net: mvneta: Improve the buffer allocation method for SWBM")
Signed-off-by: Antoine Tenart <antoine.tenart@bootlin.com>
Reviewed-by: Gregory CLEMENT <gregory.clement@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The SPI protocol for the QCA7000 doesn't have any fault detection.
In order to increase the drivers reliability in noisy environments,
we could implement a write verification inspired by the enc28j60.
This should avoid situations were the driver wrongly assumes the
receive interrupt is enabled and miss all incoming packets.
This function is disabled per default and can be controlled via module
parameter wr_verify.
Signed-off-by: Michael Heimpold <michael.heimpold@i2se.com>
Signed-off-by: Stefan Wahren <stefan.wahren@i2se.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This commit allows each TXQ to be picked in a round-robin fashion by
the PPv2 transmit scheduling mechanism. This is opposed to the default
behaviour that prioritizes the highest numbered queues.
Suggested-by: Yan Markman <ymarkman@marvell.com>
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Since the PPv2 controller has multiple TX queues, we can spread traffic
by assining TX queues to CPUs, allowing to use XPS to balance egress
traffic between CPUs.
Suggested-by : Yan Markman <ymarkman@marvell.com>
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch fixes skb_shared area, which will be corrupted
upon reception of 4K jumbo packets.
Originally build_skb usage purpose was to reuse page for skb to eliminate
needs of extra fragments. But that logic does not take into account that
skb_shared_info should be reserved at the end of skb data area.
In case packet data consumes all the page (4K), skb_shinfo location
overflows the page. As a consequence, __build_skb zeroed shinfo data above
the allocated page, corrupting next page.
The issue is rarely seen in real life because jumbo are normally larger
than 4K and that causes another code path to trigger.
But it 100% reproducible with simple scapy packet, like:
sendp(IP(dst="192.168.100.3") / TCP(dport=443) \
/ Raw(RandString(size=(4096-40))), iface="enp1s0")
Fixes: 018423e90b ("net: ethernet: aquantia: Add ring support code")
Reported-by: Friedemann Gerold <f.gerold@b-c-s.de>
Reported-by: Michael Rauch <michael@rauch.be>
Signed-off-by: Friedemann Gerold <f.gerold@b-c-s.de>
Tested-by: Nikita Danilov <nikita.danilov@aquantia.com>
Signed-off-by: Igor Russkikh <igor.russkikh@aquantia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
As diagnosed by Song Liu, ndo_poll_controller() can
be very dangerous on loaded hosts, since the cpu
calling ndo_poll_controller() might steal all NAPI
contexts (for all RX/TX queues of the NIC). This capture
can last for unlimited amount of time, since one
cpu is generally not able to drain all the queues under load.
tun uses NAPI for TX completions, so we better let core
networking stack call the napi->poll() to avoid the capture.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
As diagnosed by Song Liu, ndo_poll_controller() can
be very dangerous on loaded hosts, since the cpu
calling ndo_poll_controller() might steal all NAPI
contexts (for all RX/TX queues of the NIC). This capture
can last for unlimited amount of time, since one
cpu is generally not able to drain all the queues under load.
nfp uses NAPI for TX completions, so we better let core
networking stack call the napi->poll() to avoid the capture.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Jakub Kicinski <jakub.kicinski@netronome.com>
Acked-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Tested-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
As diagnosed by Song Liu, ndo_poll_controller() can
be very dangerous on loaded hosts, since the cpu
calling ndo_poll_controller() might steal all NAPI
contexts (for all RX/TX queues of the NIC). This capture
can last for unlimited amount of time, since one
cpu is generally not able to drain all the queues under load.
bnxt uses NAPI for TX completions, so we better let core
networking stack call the napi->poll() to avoid the capture.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
