There is a threshold now used to also limit the skb allocation
when use zero-copy. This is to avoid that there are incoherence
in the ring due to a failure on skb allocation under very
aggressive testing and under low memory conditions.
Signed-off-by: NGiuseppe Cavallaro <peppe.cavallaro@st.com>
Signed-off-by: NAlexandre TORGUE <alexandre.torgue@st.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch is to allow this driver to copy tiny frames during the reception
process. This is giving more stability while stressing the driver on STi
embedded systems.
Signed-off-by: NGiuseppe Cavallaro <peppe.cavallaro@st.com>
Signed-off-by: NAlexandre TORGUE <alexandre.torgue@st.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

phy_bus_name can be NULL when "fixed-link" property isn't used.
Then, since "stmmac: do not poll phy handler when attach a switch",
phy_bus_name ptr needs to be checked before strcmp is called.
Signed-off-by: NFabrice Gasnier <fabrice.gasnier@st.com>
Signed-off-by: NGiuseppe Cavallaro <peppe.cavallaro@st.com>
Signed-off-by: NAlexandre TORGUE <alexandre.torgue@st.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch avoids to call the stmmac_adjust_link when
the driver is connected to a switch by using the FIXED_PHY
support. Prior this patch the phydev->irq was set as PHY_POLL
so periodically the phy handler was invoked spending useless
time because the link cannot actually change.
Note that the stmmac_adjust_link will be called just one
time and this guarantees that the ST glue logic will be
setup according to the mode and speed fixed.
Signed-off-by: NGiuseppe Cavallaro <peppe.cavallaro@st.com>
Signed-off-by: NAlexandre TORGUE <alexandre.torgue@st.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch is to fill the first descriptor just before granting
the DMA engine so at the end of the xmit.
The patch takes care about the algorithm adopted to mitigate the
interrupts, then it fixes the last segment in case of no fragments.
Moreover, this new implementation does not pass any "ter" field when
prepare the descriptors because this is not necessary.
The patch also details the memory barrier in the xmit.

As final results, this patch guarantees the same performances
but fixing a case if small datagram are sent. In fact, this
kind of test is impacted if no coalesce is done.
Signed-off-by: NFabrice Gasnier <fabrice.gasnier@st.com>
Signed-off-by: NGiuseppe Cavallaro <peppe.cavallaro@st.com>
Signed-off-by: NAlexandre TORGUE <alexandre.torgue@st.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The dirty index can be updated out of the loop where all the
tx resources are claimed. This will help on performances too.
Also a useless debug printk has been removed from the main loop.
Signed-off-by: NGiuseppe Cavallaro <peppe.cavallaro@st.com>
Signed-off-by: NAlexandre TORGUE <alexandre.torgue@st.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch "inline" get_tx_owner and get_ls routines. It Results in a
unique read to tdes0, instead of three, to check TX_OWN and LS bits,
and other status bits.

It helps improve driver TX path by removing two uncached read/writes
inside TX clean loop for enhanced descriptors but not for normal ones
because the des1 must be read in any case.
Signed-off-by: NFabrice Gasnier <fabrice.gasnier@st.com>
Acked-by: NGiuseppe Cavallaro <peppe.cavallaro@st.com>
Signed-off-by: NAlexandre TORGUE <alexandre.torgue@st.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch is to optimize the way to manage the TDES inside the
xmit function. When prepare the frame, some settings (e.g. OWN
bit) can be merged. This has been reworked to improve the tx
performances.
Signed-off-by: NFabrice Gasnier <fabrice.gasnier@st.com>
Signed-off-by: NGiuseppe Cavallaro <peppe.cavallaro@st.com>
Signed-off-by: NAlexandre TORGUE <alexandre.torgue@st.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The RDES0 register can be read several times while doing RX of a
packet.
This patch slightly improves RX path performance by reading rdes0
once for two operation: check rx owner, get rx status bits.
Signed-off-by: NFabrice Gasnier <fabrice.gasnier@st.com>
Acked-by: NGiuseppe Cavallaro <peppe.cavallaro@st.com>
Signed-off-by: NAlexandre TORGUE <alexandre.torgue@st.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Optimize tx_clean by avoiding a des3 read in stmmac_clean_desc3().

In ring mode, TX, des3 seems only used when xmit a jumbo frame.
In case of normal descriptors, it may also be used for time
stamping.
Clean it in the above two case, without reading it.
Signed-off-by: NFabrice Gasnier <fabrice.gasnier@st.com>
Signed-off-by: NGiuseppe Cavallaro <peppe.cavallaro@st.com>
Signed-off-by: NAlexandre TORGUE <alexandre.torgue@st.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

last_segment field is read twice from dma descriptors in stmmac_clean().
Add last_segment to dma data so that this flag is from priv
structure in cache instead of memory.
It avoids reading twice from memory for each loop in stmmac_clean().
Signed-off-by: NFabrice Gasnier <fabrice.gasnier@st.com>
Signed-off-by: NGiuseppe Cavallaro <peppe.cavallaro@st.com>
Signed-off-by: NAlexandre TORGUE <alexandre.torgue@st.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Currently, the code pulls out the length field when
unmapping a buffer directly from the descriptor. This will result
in an uncached read to a dma_alloc_coherent() region. There is no
need to do this, so this patch simply puts the value directly into
a data structure which will hit the cache.
Signed-off-by: NFabrice Gasnier <fabrice.gasnier@st.com>
Signed-off-by: NGiuseppe Cavallaro <peppe.cavallaro@st.com>
Signed-off-by: NAlexandre TORGUE <alexandre.torgue@st.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch is to rework the ring management now optimized.
The indexes into the ring buffer are always incremented, and
the entry is accessed via doing a modulo to find the "real"
position in the ring.
It is inefficient, modulo is an expensive operation.

The formula [(entry + 1) & (size - 1)] is now adopted on
a ring that is power-of-2 in size.
Then, the number of elements cannot be set by command line but
it is fixed.
Signed-off-by: NGiuseppe Cavallaro <peppe.cavallaro@st.com>
Signed-off-by: NAlexandre TORGUE <alexandre.torgue@st.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch completely changes the descriptor layout to improve
the whole performances due to the single read usage of the
descriptors in critical paths.
Signed-off-by: NGiuseppe Cavallaro <peppe.cavallaro@st.com>
Signed-off-by: NAlexandre TORGUE <alexandre.torgue@st.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch restructures the DMA bus settings and this is done
by introducing a new platform structure used for programming
the AXI Bus Mode Register inside the DMA module.
This structure can be populated from device-tree as documented in the
binding txt file.

After initializing the DMA, the AXI register can be optionally tuned
for platform drivers based.
This patch also reworks some parameters to make coherent the DMA
configuration now that AXI register is introduced.
For example, the burst_len is managed by using the mentioned axi
support above; so the snps,burst-len parameter has been removed.
It makes sense to provide the AAL parameter from DT to Address-Aligned
Beats inside the Register0 and review the PBL settings when initialize
the engine.

For PCI glue, rebuilding the story of this setting, it
was added to align a configuration so not for fixing some
known problem. No issue raised after this patch.
It is safe to use the default burst length instead of
tuning it to the maximum value
Signed-off-by: NGiuseppe Cavallaro <peppe.cavallaro@st.com>
Signed-off-by: NAlexandre TORGUE <alexandre.torgue@st.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch is to share the same reset procedure between dwmac100 and
dwmac1000 chips.
This will also help on enhancing the driver and support new chips.
Signed-off-by: NGiuseppe Cavallaro <peppe.cavallaro@st.com>
Signed-off-by: NAlexandre TORGUE <alexandre.torgue@st.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Remove 2 unused fields from driver code.
Signed-off-by: NYuval Mintz <Yuval.Mintz@qlogic.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

In case of problems when initializing the chip, the error flows aren't
being properly done. Specifically, it's possible that the chip would be
left in a configuration allowing it [internally] to access the host
memory, causing fatal problems in the device that would require power
cycle to overcome.
Signed-off-by: NYuval Mintz <Yuval.Mintz@qlogic.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Current statistics logic is meant for L2, not for all future protocols.
Move this content to the proper designated file.
Signed-off-by: NYuval Mintz <Yuval.Mintz@qlogic.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

BB_A0 is a development model that is will not reach actual clients.
In fact, future firmware would simply fail to initialize such chip.

This changes the configuration into B0 instead of A0, and adds a safeguard
against the slim chance someone would actually try this with an A0 adapter
in which case probe would gracefully fail.
Signed-off-by: NYuval Mintz <Yuval.Mintz@qlogic.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Driver learns the inner bar sized from a register configured by management
firmware, but older versions are not setting this register.
But since we know which values were configured back then, use them instead.
Signed-off-by: NRam Amrani <Ram.Amrani@qlogic.com>
Signed-off-by: NYuval Mintz <Yuval.Mintz@qlogic.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch utilizes the attention infrastructure to log additional
information that relates only to specific HW blocks.
For some of those HW blocks, it also stops automatically disabling the
attention generation as the attention is considered benign and thus
should only be logged; No fear of it flooding the system.
Signed-off-by: NYuval Mintz <Yuval.Mintz@qlogic.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Each HW block contains common information about attention reasons,
raising a bit for each one of the different sub-reasons that caused it
to raise an attention.

This patch extends the infrastructure by allowing logging of the various
reasons causing the HW blocks to generate an attention.
Signed-off-by: NYuval Mintz <Yuval.Mintz@qlogic.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

HW is capable of generating attentnions for a multitude of reasons,
but current driver is enabling attention generation only for management
firmware [required for link notifications].

This patch enables almost all of the possible reasons for HW attentions,
logging the HW block generating the attention and preventing further
attentions from that source [to prevent possible attention flood].
It also lays the infrastructure for additional exploration of the various
attentions.
Signed-off-by: NYuval Mintz <Yuval.Mintz@qlogic.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NYuval Mintz <Yuval.Mintz@qlogic.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Avoid double mapping of io mapped memory, Device page may be
mapped to non-cached(NC) or to write-combining(WC).
The code before this fix tries to map it both to WC and NC
contrary to what stated in Intel's software developer manual.

Here we remove the global WC mapping of all UARS
"dev->priv.bf_mapping", since UAR mapping should be decided
per UAR (e.g we want different mappings for EQs, CQs vs QPs).

Caller will now have to choose whether to map via
write-combining API or not.

mlx5e SQs will choose write-combining in order to perform
BlueFlame writes.

Fixes: 88a85f99 ('TX latency optimization to save DMA reads')
Signed-off-by: NMoshe Lazer <moshel@mellanox.com>
Reviewed-by: NAchiad Shochat <achiad@mellanox.com>
Signed-off-by: NSaeed Mahameed <saeedm@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Calling mlx5e_set_coalesce while the interface is down will result in
modifying CQs that don't exist.

Fixes: f62b8bb8 ('net/mlx5: Extend mlx5_core to support ConnectX-4
Ethernet functionality')
Signed-off-by: NGal Pressman <galp@mellanox.com>
Signed-off-by: NSaeed Mahameed <saeedm@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

If CQ moderation is not supported by the device, print a warning on
netdevice load, and return error when trying to modify/query cq
moderation via ethtool.

Fixes: f62b8bb8 ('net/mlx5: Extend mlx5_core to support ConnectX-4
Ethernet functionality')
Signed-off-by: NGal Pressman <galp@mellanox.com>
Signed-off-by: NSaeed Mahameed <saeedm@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

By its role, there is no need to set all the other parameters
for the drop RQ.
Signed-off-by: NTariq Toukan <tariqt@mellanox.com>
Signed-off-by: NSaeed Mahameed <saeedm@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

For data cache locality considerations, we moved the nop and
csum_offload_inner within sq_stats struct as they are more
commonly accessed in xmit path.
Signed-off-by: NTariq Toukan <tariqt@mellanox.com>
Signed-off-by: NSaeed Mahameed <saeedm@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Instead of the pair (channel, tc), we now use a single number that
goes over all tx queues of a TC, for all TCs.
Signed-off-by: NTariq Toukan <tariqt@mellanox.com>
Signed-off-by: NSaeed Mahameed <saeedm@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

More proper to declare carrier state UP only after the channels
are ready for traffic.
Signed-off-by: NTariq Toukan <tariqt@mellanox.com>
Signed-off-by: NSaeed Mahameed <saeedm@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

We only need to flush the irq handler to make sure it does not
queue a work into the global work queue after we start to flush it.
So using synchronize_irq() is more appropriate than a spin lock.
Signed-off-by: NTariq Toukan <tariqt@mellanox.com>
Signed-off-by: NSaeed Mahameed <saeedm@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Allow a user to split or unsplit a port using the newly introduced
devlink ops.

Once split, the original netdev is destroyed and 2 or 4 others are
created, according to user configuration. The new ports are like any
other port, with the sole difference of supporting a lower maximum
speed. When unsplit, the reverse process takes place.
Signed-off-by: NIdo Schimmel <idosch@mellanox.com>
Signed-off-by: NJiri Pirko <jiri@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

When splitting and unsplitting we'll destroy usable ports on the fly, so
mark them using a NULL pointer to indicate that their local port number
is free and can be re-used.
Signed-off-by: NIdo Schimmel <idosch@mellanox.com>
Signed-off-by: NJiri Pirko <jiri@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The port netdevs are each associated with a different local port number
in the device. These local ports are grouped into groups of 4 (e.g.
(1-4), (5-8)) called clusters. The cluster constitutes the one of two
possible modules they can be mapped to. This mapping is board-specific
and done by the device's firmware during init.

When splitting a port by 4, the device requires us to first unmap all
the ports in the cluster and then map each to a single lane in the module
associated with the port netdev used as the handle for the operation.
This means that two port netdevs will disappear, as only 100Gb/s (4
lanes) ports can be split and we are guaranteed to have two of these
((1, 3), (5, 7) etc.) in a cluster.

When unsplit occurs we need to reinstantiate the two original 100Gb/s
ports and map each to its origianl module. Therefore, during driver init
store the initial local port to module mapping, so it can be used later
during unsplitting.

Note that a by 2 split doesn't require us to store the mapping, as we
only need to reinstantiate one port whose module is known.
Signed-off-by: NIdo Schimmel <idosch@mellanox.com>
Signed-off-by: NJiri Pirko <jiri@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

When splitting a port we replace it with 2 or 4 other ports. To be able
to do that we need to remove the original port netdev and unmap it from
its module. However, we first mark it as disabled, as active ports
cannot be unmapped.
Signed-off-by: NIdo Schimmel <idosch@mellanox.com>
Signed-off-by: NJiri Pirko <jiri@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Add middle layer in mlxsw core code to forward port split/unsplit calls
into specific ASIC drivers.
Signed-off-by: NJiri Pirko <jiri@mellanox.com>
Signed-off-by: NIdo Schimmel <idosch@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Implement newly introduced devlink interface. Add devlink port instances
for every port and set the port types accordingly.
Signed-off-by: NJiri Pirko <jiri@mellanox.com>
Signed-off-by: NIdo Schimmel <idosch@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

So far, there has been an mlx4-specific sysfs file allowing user to
change port type to either Ethernet of InfiniBand. This is very
inconvenient.

Allow to expose the same ability to set port type in a generic way
using devlink interface.
Signed-off-by: NJiri Pirko <jiri@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>