Extended RX buffer descriptors are necessary if they carry RX
timestamps, which will be true when PTP timestamping is enabled.
Right now, the rx_ring->ext_en is set from the function that allocates
ring resources (enetc_alloc_rx_resources() -> enetc_alloc_rxbdr()), and
also used later, in enetc_setup_rxbdr(). It is also used in the
enetc_rxbd() and enetc_rxbd_next() fast path helpers.
We want to decouple resource allocation from BD ring setup, but both
procedures depend on BD size (extended or not). Move the "extended"
boolean to enetc_open() and pass it both to the RX allocation procedure
as well as to the RX ring setup procedure. The latter will set
rx_ring->ext_en from now on.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The call path in enetc_close() is:
enetc_close()
-> enetc_free_rxtx_rings()
-> enetc_free_tx_ring()
-> enetc_free_tx_frame()
-> enetc_free_tx_resources()
-> enetc_free_txbdr()
-> enetc_free_tx_frame()
The enetc_free_tx_frame() function is written such that the second call
exits without doing anything, but nonetheless, it is completely
redundant. Delete it. This makes the TX teardown path more similar to
the RX one, where rx_swbd freeing is done in enetc_free_rx_ring(), not
in enetc_free_rxbdr().
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The call path in enetc_close() is:
enetc_close()
-> enetc_free_rxtx_rings()
-> enetc_free_rx_ring()
-> tests whether rx_ring->rx_swbd is NULL
-> enetc_free_tx_ring()
-> tests whether tx_ring->tx_swbd is NULL
-> enetc_free_rx_resources()
-> enetc_free_rxbdr()
-> sets rxr->rx_swbd to NULL
-> enetc_free_tx_resources()
-> enetc_free_txbdr()
-> setx txr->tx_swbd to NULL
From the above, it is clear that due to the function ordering, the
checks for NULL are redundant, since the software buffer descriptor
arrays have not yet been set to NULL. Drop these checks.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
This is a refactoring change which introduces the opposite function of
enetc_dma_alloc_bdr().
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
There is only one place which needs to set up indices in the RX ring.
Be consistent with what was done in the TX path and do this in
enetc_setup_rxbdr().
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
enetc_alloc_txbdr() deals with allocating resources necessary for a TX
ring to work (the array of software BDs and the array of TSO headers).
The next_to_clean and next_to_use pointers are overwritten with proper
values which are read from hardware here:
enetc_open
-> enetc_alloc_tx_resources
-> enetc_alloc_txbdr
-> set to zero
-> enetc_setup_bdrs
-> enetc_setup_txbdr
-> read from hardware
So their initialization with zeroes is pointless and confusing.
Delete it.
Consequently, since enetc_setup_txbdr() has no opposite cleanup
function, also delete the resetting of these indices from
enetc_free_tx_ring().
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The enetc MDIO bus driver can perform both C22 and C45 transfers.
Create separate functions for each and register the C45 versions using
the new API calls where appropriate.
This driver is shared with the Felix DSA switch, so update that at the
same time.
Signed-off-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Michael Walle <michael@walle.cc>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
This lockdep splat says it better than I could:
================================
WARNING: inconsistent lock state
6.2.0-rc2-07010-ga9b9500ffaac-dirty #967 Not tainted
--------------------------------
inconsistent {IN-SOFTIRQ-W} -> {SOFTIRQ-ON-W} usage.
kworker/1:3/179 [HC0[0]:SC0[0]:HE1:SE1] takes:
ffff3ec4036ce098 (_xmit_ETHER#2){+.?.}-{3:3}, at: netif_freeze_queues+0x5c/0xc0
{IN-SOFTIRQ-W} state was registered at:
_raw_spin_lock+0x5c/0xc0
sch_direct_xmit+0x148/0x37c
__dev_queue_xmit+0x528/0x111c
ip6_finish_output2+0x5ec/0xb7c
ip6_finish_output+0x240/0x3f0
ip6_output+0x78/0x360
ndisc_send_skb+0x33c/0x85c
ndisc_send_rs+0x54/0x12c
addrconf_rs_timer+0x154/0x260
call_timer_fn+0xb8/0x3a0
__run_timers.part.0+0x214/0x26c
run_timer_softirq+0x3c/0x74
__do_softirq+0x14c/0x5d8
____do_softirq+0x10/0x20
call_on_irq_stack+0x2c/0x5c
do_softirq_own_stack+0x1c/0x30
__irq_exit_rcu+0x168/0x1a0
irq_exit_rcu+0x10/0x40
el1_interrupt+0x38/0x64
irq event stamp: 7825
hardirqs last enabled at (7825): [<ffffdf1f7200cae4>] exit_to_kernel_mode+0x34/0x130
hardirqs last disabled at (7823): [<ffffdf1f708105f0>] __do_softirq+0x550/0x5d8
softirqs last enabled at (7824): [<ffffdf1f7081050c>] __do_softirq+0x46c/0x5d8
softirqs last disabled at (7811): [<ffffdf1f708166e0>] ____do_softirq+0x10/0x20
other info that might help us debug this:
Possible unsafe locking scenario:
CPU0
----
lock(_xmit_ETHER#2);
<Interrupt>
lock(_xmit_ETHER#2);
*** DEADLOCK ***
3 locks held by kworker/1:3/179:
#0: ffff3ec400004748 ((wq_completion)events){+.+.}-{0:0}, at: process_one_work+0x1f4/0x6c0
#1: ffff80000a0bbdc8 ((work_completion)(&priv->tx_onestep_tstamp)){+.+.}-{0:0}, at: process_one_work+0x1f4/0x6c0
#2: ffff3ec4036cd438 (&dev->tx_global_lock){+.+.}-{3:3}, at: netif_tx_lock+0x1c/0x34
Workqueue: events enetc_tx_onestep_tstamp
Call trace:
print_usage_bug.part.0+0x208/0x22c
mark_lock+0x7f0/0x8b0
__lock_acquire+0x7c4/0x1ce0
lock_acquire.part.0+0xe0/0x220
lock_acquire+0x68/0x84
_raw_spin_lock+0x5c/0xc0
netif_freeze_queues+0x5c/0xc0
netif_tx_lock+0x24/0x34
enetc_tx_onestep_tstamp+0x20/0x100
process_one_work+0x28c/0x6c0
worker_thread+0x74/0x450
kthread+0x118/0x11c
but I'll say it anyway: the enetc_tx_onestep_tstamp() work item runs in
process context, therefore with softirqs enabled (i.o.w., it can be
interrupted by a softirq). If we hold the netif_tx_lock() when there is
an interrupt, and the NET_TX softirq then gets scheduled, this will take
the netif_tx_lock() a second time and deadlock the kernel.
To solve this, use netif_tx_lock_bh(), which blocks softirqs from
running.
Fixes: 7294380c52 ("enetc: support PTP Sync packet one-step timestamping")
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Alexander Duyck <alexanderduyck@fb.com>
Link: https://lore.kernel.org/r/20230112105440.1786799-1-vladimir.oltean@nxp.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
While reviewing dependencies in some Kconfig files, I noticed the redundant
dependency "depends on PCI && PCI_MSI". The config PCI_MSI has always,
since its introduction, been dependent on the config PCI. So, it is
sufficient to just depend on PCI_MSI, and know that the dependency on PCI
is implicitly implied.
Reduce the dependencies of some network driver configs.
No functional change and effective change of Kconfig dependendencies.
Signed-off-by: Lukas Bulwahn <lukas.bulwahn@gmail.com>
Acked-by: Simon Horman <simon.horman@corigine.com>
Acked-by: Dimitris Michailidis <dmichail@fungible.com>
Link: https://lore.kernel.org/r/20230111125855.19020-1-lukas.bulwahn@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The fec MDIO bus driver can perform both C22 and C45 transfers.
Create separate functions for each and register the C45 versions using
the new API calls where appropriate.
Signed-off-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Michael Walle <michael@walle.cc>
Reviewed-by: Wei Fang <wei.fang@nxp.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The xgmac MDIO bus driver can perform both C22 and C45 transfers.
Create separate functions for each and register the C45 versions using
the new API calls where appropriate.
While at it, remove the misleading comment. According to Vladimir
Oltean:
- miimcom is a register accessed by fsl_pq_mdio.c, not by xgmac_mdio.c
- "device dev" doesn't really refer to anything (maybe "dev_addr").
- I don't understand what is meant by the comment "All PHY
configuration has to be done through the TSEC1 MIIM regs". Or rather
said, I think I understand, but it is irrelevant to the driver for 2
reasons:
* TSEC devices use the fsl_pq_mdio.c driver, not this one
* It doesn't matter to this driver whose TSEC registers are used for
MDIO access. The driver just works with the registers it's given,
which is a concern for the device tree.
- barring the above, the rest just describes the MDIO bus API, which is
superfluous
Signed-off-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Michael Walle <michael@walle.cc>
Tested-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Move XDP skb_shared_info structure initialization in from
enetc_map_rx_buff_to_xdp() to enetc_add_rx_buff_to_xdp() and do not always
access skb_shared_info in the xdp_buff/xdp_frame since it is located in a
different cacheline with respect to hard_start and data xdp pointers.
Rely on XDP_FLAGS_HAS_FRAGS flag to check if it really necessary to access
non-linear part of the xdp_buff/xdp_frame.
Reviewed-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Tested-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Remove xdp_redirect_sg counter and the related ethtool entry since it is
no longer used.
Tested-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Even if full XDP_REDIRECT is not supported yet for non-linear XDP buffers
since we allow redirecting just into CPUMAPs, unlock XDP_REDIRECT for
S/G XDP buffer and rely on XDP stack to properly take care of the
frames.
Tested-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
We want to fail if the PCS is not available, not if it is available. Fix
this condition.
Fixes: 5d93cfcf73 ("net: dpaa: Convert to phylink")
Reported-by: Christian Zigotzky <info@xenosoft.de>
Signed-off-by: Sean Anderson <seanga2@gmail.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
A remove callback just returning 0 is equivalent to no remove callback
at all. So drop the useless function.
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
The build_skb might return a null pointer but there is no check on the
return value in the fec_enet_rx_queue(). So a null pointer dereference
might occur. To avoid this, we check the return value of build_skb. If
the return value is a null pointer, the driver will recycle the page and
update the statistic of ndev. Then jump to rx_processing_done to clear
the status flags of the BD so that the hardware can recycle the BD.
Fixes: 95698ff617 ("net: fec: using page pool to manage RX buffers")
Signed-off-by: Wei Fang <wei.fang@nxp.com>
Reviewed-by: Shenwei Wang <Shenwei.wang@nxp.com>
Reviewed-by: Alexander Duyck <alexanderduyck@fb.com>
Link: https://lore.kernel.org/r/20221219022755.1047573-1-wei.fang@nxp.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Before enetc_clean_rx_ring_xdp() calls xdp_do_redirect(), each software
BD in the RX ring between index orig_i and i can have one of 2 refcount
values on its page.
We are the owner of the current buffer that is being processed, so the
refcount will be at least 1.
If the current owner of the buffer at the diametrically opposed index
in the RX ring (i.o.w, the other half of this page) has not yet called
kfree(), this page's refcount could even be 2.
enetc_page_reusable() in enetc_flip_rx_buff() tests for the page
refcount against 1, and [ if it's 2 ] does not attempt to reuse it.
But if enetc_flip_rx_buff() is put after the xdp_do_redirect() call,
the page refcount can have one of 3 values. It can also be 0, if there
is no owner of the other page half, and xdp_do_redirect() for this
buffer ran so far that it triggered a flush of the devmap/cpumap bulk
queue, and the consumers of those bulk queues also freed the buffer,
all by the time xdp_do_redirect() returns the execution back to enetc.
This is the reason why enetc_flip_rx_buff() is called before
xdp_do_redirect(), but there is a big flaw with that reasoning:
enetc_flip_rx_buff() will set rx_swbd->page = NULL on both sides of the
enetc_page_reusable() branch, and if xdp_do_redirect() returns an error,
we call enetc_xdp_free(), which does not deal gracefully with that.
In fact, what happens is quite special. The page refcounts start as 1.
enetc_flip_rx_buff() figures they're reusable, transfers these
rx_swbd->page pointers to a different rx_swbd in enetc_reuse_page(), and
bumps the refcount to 2. When xdp_do_redirect() later returns an error,
we call the no-op enetc_xdp_free(), but we still haven't lost the
reference to that page. A copy of it is still at rx_ring->next_to_alloc,
but that has refcount 2 (and there are no concurrent owners of it in
flight, to drop the refcount). What really kills the system is when
we'll flip the rx_swbd->page the second time around. With an updated
refcount of 2, the page will not be reusable and we'll really leak it.
Then enetc_new_page() will have to allocate more pages, which will then
eventually leak again on further errors from xdp_do_redirect().
The problem, summarized, is that we zeroize rx_swbd->page before we're
completely done with it, and this makes it impossible for the error path
to do something with it.
Since the packet is potentially multi-buffer and therefore the
rx_swbd->page is potentially an array, manual passing of the old
pointers between enetc_flip_rx_buff() and enetc_xdp_free() is a bit
difficult.
For the sake of going with a simple solution, we accept the possibility
of racing with xdp_do_redirect(), and we move the flip procedure to
execute only on the redirect success path. By racing, I mean that the
page may be deemed as not reusable by enetc (having a refcount of 0),
but there will be no leak in that case, either.
Once we accept that, we have something better to do with buffers on
XDP_REDIRECT failure. Since we haven't performed half-page flipping yet,
we won't, either (and this way, we can avoid enetc_xdp_free()
completely, which gives the entire page to the slab allocator).
Instead, we'll call enetc_xdp_drop(), which will recycle this half of
the buffer back to the RX ring.
Fixes: 9d2b68cc10 ("net: enetc: add support for XDP_REDIRECT")
Suggested-by: Lorenzo Bianconi <lorenzo.bianconi@redhat.com>
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Link: https://lore.kernel.org/r/20221213001908.2347046-1-vladimir.oltean@nxp.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The cmd_buff needs to be freed when error happened in
dpaa2_switch_acl_entry_add() and dpaa2_switch_acl_entry_remove().
Fixes: 1110318d83 ("dpaa2-switch: add tc flower hardware offload on ingress traffic")
Signed-off-by: Yuan Can <yuancan@huawei.com>
Link: https://lore.kernel.org/r/20221205061515.115012-1-yuancan@huawei.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Prior to the Fixes: commit, the initialization code went through the
same fec_enet_set_coalesce() function as used by ethtool, and that
function correctly checks whether the current variant has support for
irq coalescing.
Now that the initialization code instead calls fec_enet_itr_coal_set()
directly, that call needs to be guarded by a check for the
FEC_QUIRK_HAS_COALESCE bit.
Fixes: df727d4547 (net: fec: don't reset irq coalesce settings to defaults on "ip link up")
Reported-by: Greg Ungerer <gregungerer@westnet.com.au>
Signed-off-by: Rasmus Villemoes <linux@rasmusvillemoes.dk>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Link: https://lore.kernel.org/r/20221205204604.869853-1-linux@rasmusvillemoes.dk
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
After the introduction of a private mac_lock that serializes access to
priv->mac (and port_priv->mac in the switch), the only remaining purpose
of rtnl_lock() is to satisfy the locking requirements of
phylink_fwnode_phy_connect() and phylink_disconnect_phy().
But the functions these live in, dpaa2_mac_connect() and
dpaa2_mac_disconnect(), have contradictory locking requirements.
While phylink_fwnode_phy_connect() wants rtnl_lock() to be held,
phylink_create() wants it to not be held.
Move the rtnl_lock() from top-level (in the dpaa2-eth and dpaa2-switch
drivers) to only surround the phylink calls that require it, in the
dpaa2-mac library code.
This is possible because dpaa2_mac_connect() and dpaa2_mac_disconnect()
run unlocked, and there isn't any danger of an AB/BA deadlock between
the rtnl_mutex and other private locks.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Tested-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
The dpaa2-switch driver uses a DPMAC in the same way as the dpaa2-eth
driver, so we need to duplicate the locking solution established by the
previous change to the switch driver as well.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Tested-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
The dpaa2 architecture permits dynamic connections between objects on
the fsl-mc bus, specifically between a DPNI object (represented by a
struct net_device) and a DPMAC object (represented by a struct phylink).
The DPNI driver is notified when those connections are created/broken
through the dpni_irq0_handler_thread() method. To ensure that ethtool
operations, as well as netdev up/down operations serialize with the
connection/disconnection of the DPNI with a DPMAC,
dpni_irq0_handler_thread() takes the rtnl_lock() to block those other
operations from taking place.
There is code called by dpaa2_mac_connect() which wants to acquire the
rtnl_mutex once again, see phylink_create() -> phylink_register_sfp() ->
sfp_bus_add_upstream() -> rtnl_lock(). So the strategy doesn't quite
work out, even though it's fairly simple.
Create a different strategy, where all code paths in the dpaa2-eth
driver access priv->mac only while they are holding priv->mac_lock.
The phylink instance is not created or connected to the PHY under the
priv->mac_lock, but only assigned to priv->mac then. This will eliminate
the reliance on the rtnl_mutex.
Add lockdep annotations and put comments where holding the lock is not
necessary, and priv->mac can be dereferenced freely.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Tested-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
dpaa2_eth_connect_mac() is called both from dpaa2_eth_probe() and from
dpni_irq0_handler_thread().
It could happen that the DPNI gets connected to a DPMAC on the fsl-mc
bus exactly during probe, as soon as the "endpoint change" interrupt is
requested in dpaa2_eth_setup_irqs(). This will cause the
dpni_irq0_handler_thread() to register a phylink instance for that DPMAC.
Then, the probing function will also try to register a phylink instance
for the same DPMAC, operation which should fail (and this will fail the
probing of the driver).
Reorder dpaa2_eth_setup_irqs() and dpaa2_eth_connect_mac(), such that
dpni_irq0_handler_thread() never races with the DPMAC-related portion of
the probing path.
Also reorder dpaa2_eth_disconnect_mac() to be in the mirror position of
dpaa2_eth_connect_mac() in the teardown path.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Tested-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
The helper function will gain a lockdep annotation in a future patch.
Make sure to benefit from it.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Tested-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
DPNIs and DPSW objects can connect and disconnect at runtime from DPMAC
objects on the same fsl-mc bus. The DPMAC object also holds "ethtool -S"
unstructured counters. Those counters are only shown for the entity
owning the netdev (DPNI, DPSW) if it's connected to a DPMAC.
The ethtool stringset code path is split into multiple callbacks, but
currently, connecting and disconnecting the DPMAC takes the rtnl_lock().
This blocks the entire ethtool code path from running, see
ethnl_default_doit() -> rtnl_lock() -> ops->prepare_data() ->
strset_prepare_data().
This is going to be a problem if we are going to no longer require
rtnl_lock() when connecting/disconnecting the DPMAC, because the DPMAC
could appear between ops->get_sset_count() and ops->get_strings().
If it appears out of the blue, we will provide a stringset into an array
that was dimensioned thinking the DPMAC wouldn't be there => array
accessed out of bounds.
There isn't really a good way to work around that, and I don't want to
put too much pressure on the ethtool framework by playing locking games.
Just make the DPMAC counters be always available. They'll be zeroes if
the DPNI or DPSW isn't connected to a DPMAC.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Reviewed-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Tested-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
The dpaa2-switch has the exact same locking requirements when connected
to a DPMAC, so it needs port_priv->mac to always point either to NULL,
or to a DPMAC with a fully initialized phylink instance.
Make the same preparatory change in the dpaa2-switch driver as in the
dpaa2-eth one.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Tested-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
There are 2 requirements for correct code:
- Any time the driver accesses the priv->mac pointer at runtime, it
either holds NULL to indicate a DPNI-DPNI connection (or unconnected
DPNI), or a struct dpaa2_mac whose phylink instance was fully
initialized (created and connected to the PHY). No changes are made to
priv->mac while it is being used. Currently, rtnl_lock() watches over
the call to dpaa2_eth_connect_mac(), so it serves the purpose of
serializing this with all readers of priv->mac.
- dpaa2_mac_connect() should run unlocked, because inside it are 2
phylink calls with incompatible locking requirements: phylink_create()
requires that the rtnl_mutex isn't held, and phylink_fwnode_phy_connect()
requires that the rtnl_mutex is held. The only way to solve those
contradictory requirements is to let dpaa2_mac_connect() take
rtnl_lock() when it needs to.
To solve both requirements, we need to identify the writer side of the
priv->mac pointer, which can be wrapped in a mutex private to the driver
in a future patch. The dpaa2_mac_connect() cannot be part of the writer
side critical section, because of an AB/BA deadlock with rtnl_lock().
So the strategy needs to be that where we prepare the DPMAC by calling
dpaa2_mac_connect(), and only make priv->mac point to it once it's fully
prepared. This ensures that the writer side critical section has the
absolute minimum surface it can.
The reverse strategy is adopted in the dpaa2_eth_disconnect_mac() code
path. This makes sure that priv->mac is NULL when we start tearing down
the DPMAC that we disconnected from, and concurrent code will simply not
see it.
No locking changes in this patch (concurrent code is still blocked by
the rtnl_mutex).
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Tested-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
dpaa2_mac_disconnect() will only be called with a NULL mac->phylink if
dpaa2_mac_connect() failed, or was never called.
The callers are these:
dpaa2_eth_disconnect_mac():
if (dpaa2_eth_is_type_phy(priv))
dpaa2_mac_disconnect(priv->mac);
dpaa2_switch_port_disconnect_mac():
if (dpaa2_switch_port_is_type_phy(port_priv))
dpaa2_mac_disconnect(port_priv->mac);
priv->mac can be NULL, but in that case, dpaa2_eth_is_type_phy() returns
false, and dpaa2_mac_disconnect() is never called. Similar for
dpaa2-switch.
When priv->mac is non-NULL, it means that dpaa2_mac_connect() returned
zero (success), and therefore, priv->mac->phylink is also a valid
pointer.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Reviewed-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Tested-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
The phylink handling is intended to be hidden inside the dpaa2_mac
object. Move the phylink_start() call into dpaa2_mac_start(), and
phylink_stop() into dpaa2_mac_stop().
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Reviewed-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Tested-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
dpaa2_mac_is_type_fixed() is a header with no implementation and no
callers, which is referenced from the documentation though. It can be
deleted.
On the other hand, it would be useful to reuse the code between
dpaa2_eth_is_type_phy() and dpaa2_switch_port_is_type_phy(). That common
code should be called dpaa2_mac_is_type_phy(), so let's create that.
The removal and the addition are merged into the same patch because,
in fact, is_type_phy() is the logical opposite of is_type_fixed().
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Reviewed-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Tested-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
dpaa2_eth_setup_dpni() is called from the probe path and
dpaa2_eth_set_link_ksettings() is propagated to user space.
include/linux/errno.h says that ENOTSUPP is "Defined for the NFSv3
protocol". Conventional wisdom has it to not use it in networking
drivers. Replace it with -EOPNOTSUPP.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Reviewed-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Tested-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
The driver name is available in device_driver::name. Right now,
drivers still have to report this piece of information themselves in
their devlink_ops::info_get callback function.
In order to factorize code, make devlink_nl_info_fill() add the driver
name attribute.
Now that the core sets the driver name attribute, drivers are not
supposed to call devlink_info_driver_name_put() anymore. Remove
devlink_info_driver_name_put() and clean-up all the drivers using this
function in their callback.
Signed-off-by: Vincent Mailhol <mailhol.vincent@wanadoo.fr>
Tested-by: Ido Schimmel <idosch@nvidia.com> # mlxsw
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Currently, when a FEC device is brought up, the irq coalesce settings
are reset to their default values (1000us, 200 frames). That's
unexpected, and breaks for example use of an appropriate .link file to
make systemd-udev apply the desired
settings (https://www.freedesktop.org/software/systemd/man/systemd.link.html),
or any other method that would do a one-time setup during early boot.
Refactor the code so that fec_restart() instead uses
fec_enet_itr_coal_set(), which simply applies the settings that are
stored in the private data, and initialize that private data with the
default values.
Signed-off-by: Rasmus Villemoes <linux@rasmusvillemoes.dk>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
In the blamed commit, a rudimentary reallocation procedure for RX buffer
descriptors was implemented, for the situation when their format changes
between normal (no PTP) and extended (PTP).
enetc_hwtstamp_set() calls enetc_close() and enetc_open() in a sequence,
and this sequence loses information which was previously configured in
the TX BDR Mode Register, specifically via the enetc_set_bdr_prio() call.
The TX ring priority is configured by tc-mqprio and tc-taprio, and
affects important things for TSN such as the TX time of packets. The
issue manifests itself most visibly by the fact that isochron --txtime
reports premature packet transmissions when PTP is first enabled on an
enetc interface.
Save the TX ring priority in a new field in struct enetc_bdr (occupies a
2 byte hole on arm64) in order to make this survive a ring reconfiguration.
Fixes: 434cebabd3 ("enetc: Add dynamic allocation of extended Rx BD rings")
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Alexander Lobakin <alexandr.lobakin@intel.com>
Link: https://lore.kernel.org/r/20221122130936.1704151-1-vladimir.oltean@nxp.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Commit a7c2a32e7f ("net: fman: memac: Use lynx pcs driver") makes the
Freescale Data-Path Acceleration Architecture Frame Manager use lynx pcs
driver by selecting PCS_LYNX.
It also selects the non-existing config PCS as well, which has no effect.
Remove this select to a non-existing config.
Signed-off-by: Lukas Bulwahn <lukas.bulwahn@gmail.com>
Link: https://lore.kernel.org/r/20221116102450.13928-1-lukas.bulwahn@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Nothing in these file needs anything from linux/msi.h
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Ioana Ciornei <ioana.ciornei@nxp.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Added xdp and page pool statistics.
In order to make the implementation simple and compatible, the patch
uses the 32bit integer to record the XDP statistics.
Signed-off-by: Shenwei Wang <shenwei.wang@nxp.com>
Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Convert all remaining drivers that still use .adjfreq to the newer .adjfine
implementation. These drivers are not straightforward, as they use
non-standard methods of programming their hardware. They are all converted
to use scaled_ppm_to_ppb to get the parts per billion value that their
logic depends on.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Cc: Ariel Elior <aelior@marvell.com>
Cc: Sudarsana Kalluru <skalluru@marvell.com>
Cc: Manish Chopra <manishc@marvell.com>
Cc: Derek Chickles <dchickles@marvell.com>
Cc: Satanand Burla <sburla@marvell.com>
Cc: Felix Manlunas <fmanlunas@marvell.com>
Cc: Raju Rangoju <rajur@chelsio.com>
Cc: Joakim Zhang <qiangqing.zhang@nxp.com>
Cc: Edward Cree <ecree.xilinx@gmail.com>
Cc: Martin Habets <habetsm.xilinx@gmail.com>
Cc: Richard Cochran <richardcochran@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Virtually all conventional network drivers are now converted to use
phylink_generic_validate() - only DSA drivers and fman_memac remain,
so lets remove the necessity for network drivers to explicitly set
this member, and default to phylink_generic_validate() when unset.
This is possible as .validate must currently be set.
Any remaining instances that have not been addressed by this patch can
be fixed up later.
Signed-off-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk>
Reviewed-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Link: https://lore.kernel.org/r/E1or0FZ-001tRa-DI@rmk-PC.armlinux.org.uk
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Simplify the code logic of handling the quirk of FEC_QUIRK_HAS_RACC.
If a SoC has the RACC quirk, the driver will enable the 16bit shift
by default in the probe function for a better performance.
This patch handles the logic in one place to make the logic simple
and clean. The patch optimizes the fec_enet_xdp_get_tx_queue function
according to Paolo Abeni's comments, and it also exludes the SoCs that
require to do frame swap from XDP support.
Signed-off-by: Shenwei Wang <shenwei.wang@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When the mac device gets removed, it leaves behind the ethernet device.
This will result in a segfault next time the ethernet device accesses
mac_dev. Remove the ethernet device when we get removed to prevent
this. This is not completely reversible, since some resources aren't
cleaned up properly, but that can be addressed later.
Fixes: 3933961682 ("fsl/fman: Add FMan MAC driver")
Signed-off-by: Sean Anderson <sean.anderson@seco.com>
Link: https://lore.kernel.org/r/20221103182831.2248833-1-sean.anderson@seco.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Benefit from the previously implemented tracking of netdev events in
devlink code and instead of calling devlink_port_type_eth_set() and
devlink_port_type_clear() to set devlink port type and link to related
netdev, use SET_NETDEV_DEVLINK_PORT() macro to assign devlink_port
pointer to netdevice which is about to be registered.
Signed-off-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
This patch adds the initial XDP support to Freescale driver. It supports
XDP_PASS, XDP_DROP and XDP_REDIRECT actions. Upcoming patches will add
support for XDP_TX and Zero Copy features.
As the patch is rather large, the part of codes to collect the
statistics is separated and will prepare a dedicated patch for that
part.
I just tested with the application of xdpsock.
-- Native here means running command of "xdpsock -i eth0"
-- SKB-Mode means running command of "xdpsock -S -i eth0"
The following are the testing result relating to XDP mode:
root@imx8qxpc0mek:~/bpf# ./xdpsock -i eth0
sock0@eth0:0 rxdrop xdp-drv
pps pkts 1.00
rx 371347 2717794
tx 0 0
root@imx8qxpc0mek:~/bpf# ./xdpsock -S -i eth0
sock0@eth0:0 rxdrop xdp-skb
pps pkts 1.00
rx 202229 404528
tx 0 0
root@imx8qxpc0mek:~/bpf# ./xdp2 eth0
proto 0: 496708 pkt/s
proto 0: 505469 pkt/s
proto 0: 505283 pkt/s
proto 0: 505443 pkt/s
proto 0: 505465 pkt/s
root@imx8qxpc0mek:~/bpf# ./xdp2 -S eth0
proto 0: 0 pkt/s
proto 17: 118778 pkt/s
proto 17: 118989 pkt/s
proto 0: 1 pkt/s
proto 17: 118987 pkt/s
proto 0: 0 pkt/s
proto 17: 118943 pkt/s
proto 17: 118976 pkt/s
proto 0: 1 pkt/s
proto 17: 119006 pkt/s
proto 0: 0 pkt/s
proto 17: 119071 pkt/s
proto 17: 119092 pkt/s
Signed-off-by: Shenwei Wang <shenwei.wang@nxp.com>
Reported-by: kernel test robot <lkp@intel.com>
Link: https://lore.kernel.org/r/20221031185350.2045675-1-shenwei.wang@nxp.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
When this device is deferred, there is often no way to determine what
the cause was. Add some debug prints to make it easier to figure out
what is blocking the probe.
Signed-off-by: Sean Anderson <sean.anderson@seco.com>
Link: https://lore.kernel.org/r/20221027190005.400839-1-sean.anderson@seco.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The ndo_start_xmit() method must not free skb when returning
NETDEV_TX_BUSY, since caller is going to requeue freed skb.
Fix it by returning NETDEV_TX_OK in case of dma_map_single() fails.
Fixes: 79f339125e ("net: fec: Add software TSO support")
Signed-off-by: Zhang Changzhong <zhangchangzhong@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
