Currently MAC filters are not altered during a VF reset event. This may
lead to a stale filter when an administratively set MAC is forced by the
PF.
For an administratively set MAC the PF driver deletes the VFs filters,
overwrites the VFs MAC address and triggers a VF reset. However
the VF driver itself is not aware of the filter removal, which is what
the VF reset is for.
The VF reset queues all filters present in the VF driver to be re-added
to the PF filter list (including the filter for the now stale VF MAC
address) and triggers a VIRTCHNL_OP_GET_VF_RESOURCES event, which
provides the new MAC address to the VF.
When this happens i40e will complain and reject the stale MAC filter,
at least in the untrusted VF case.
i40e 0000:08:00.0: Setting MAC 3c:fa:fa:fa:fa:01 on VF 0
iavf 0000:08:02.0: Reset warning received from the PF
iavf 0000:08:02.0: Scheduling reset task
i40e 0000:08:00.0: Bring down and up the VF interface to make this change effective.
i40e 0000:08:00.0: VF attempting to override administratively set MAC address, bring down and up the VF interface to resume normal operation
i40e 0000:08:00.0: VF 0 failed opcode 10, retval: -1
iavf 0000:08:02.0: Failed to add MAC filter, error IAVF_ERR_NVM
To avoid re-adding the stale MAC filter it needs to be removed from the
VF driver's filter list before queuing the existing filters. Then during
the VIRTCHNL_OP_GET_VF_RESOURCES event the correct filter needs to be
added again, at which point the MAC address has been updated.
As a bonus this change makes bringing the VF down and up again
superfluous for the administratively set MAC case.
Signed-off-by: Stefan Assmann <sassmann@kpanic.de>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Changing the link mode should also be done for 100BaseFX SGMII modules,
otherwise they just don't work when the default link mode in CTRL_EXT
coming from the EEPROM is SERDES.
Additionally 100Base-LX SGMII SFP modules are also supported now, which
was not the case before.
Tested with an i210 using Flexoptix S.1303.2M.G 100FX and
S.1303.10.G 100LX SGMII SFP modules.
Signed-off-by: Manfred Rudigier <manfred.rudigier@omicronenergy.com>
Tested-by: Aaron Brown <aaron.f.brown@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
This patch fixes the calculation of queue when we restore flow director
filters after resetting adapter. In ixgbe_fdir_filter_restore(), filter's
vf may be zero which makes the queue outside of the rx_ring array.
The calculation is changed to the same as ixgbe_add_ethtool_fdir_entry().
Signed-off-by: Cambda Zhu <cambda@linux.alibaba.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Currently, though the FDB entry is added to VF, it does not appear in
RAR filters. VF driver only allows to add 10 entries. Attempting to add
another causes an error. This patch removes limitation and allows use of
all free RAR entries for the FDB if needed.
Fixes: 46ec20ff7d ("ixgbevf: Add macvlan support in the set rx mode op")
Signed-off-by: Radoslaw Tyl <radoslawx.tyl@intel.com>
Acked-by: Paul Menzel <pmenzel@molgen.mpg.de>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Currently in i40e_vc_disable_queues_msg() we are incorrectly
validating the virtchnl queue select bitmaps. The
virtchnl_queue_select rx_queues and tx_queue bitmap is being
compared against ICE_MAX_VF_QUEUES, but the problem is that
these bitmaps can have a value greater than I40E_MAX_VF_QUEUES.
Fix this by comparing the bitmaps against BIT(I40E_MAX_VF_QUEUES).
Also, add the function i40e_vc_validate_vqs_bitmaps() that checks to see
if both virtchnl_queue_select bitmaps are empty along with checking that
the bitmaps only have valid bits set. This function can then be used in
both the queue enable and disable flows.
Suggested-by: Arkady Gilinksky <arkady.gilinsky@harmonicinc.com>
Signed-off-by: Brett Creeley <brett.creeley@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
The following patch will change PRIO to replace a removed Qdisc with an
invisible FIFO, instead of NOOP. mlxsw will see this replacement due to the
graft message that is generated. But because FIFO does not issue its own
REPLACE message, when the graft operation takes place, the Qdisc that mlxsw
tracks under the indicated band is still the old one. The child
handle (0:0) therefore does not match, and mlxsw rejects the graft
operation, which leads to an extack message:
Warning: Offloading graft operation failed.
Fix by ignoring the invisible children in the PRIO graft handler. The
DESTROY message of the removed Qdisc is going to follow shortly and handle
the removal.
Fixes: 32dc5efc6c ("mlxsw: spectrum: qdiscs: prio: Handle graft command")
Signed-off-by: Petr Machata <petrm@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Allow all the RGMII modes to be used. This would allow us to represent
the hardware better in the device tree with RGMII_ID where in most
cases the PHY's internal delay for both RX and TX are used.
Fixes: 9f93ac8d40 ("net-next: stmmac: Add dwmac-sun8i")
Signed-off-by: Chen-Yu Tsai <wens@csie.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Allow all the RGMII modes to be used. This would allow us to represent
the hardware better in the device tree with RGMII_ID where in most
cases the PHY's internal delay for both RX and TX are used.
Fixes: af0bd4e9ba ("net: stmmac: sunxi platform extensions for GMAC in Allwinner A20 SoC's")
Signed-off-by: Chen-Yu Tsai <wens@csie.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
When using fixed link we don't need the MDIO bus support.
Reported-by: Heiko Stuebner <heiko@sntech.de>
Reported-by: kernelci.org bot <bot@kernelci.org>
Fixes: d3e014ec7d ("net: stmmac: platform: Fix MDIO init for platforms without PHY")
Signed-off-by: Jose Abreu <Jose.Abreu@synopsys.com>
Acked-by: Sriram Dash <Sriram.dash@samsung.com>
Tested-by: Patrice Chotard <patrice.chotard@st.com>
Tested-by: Heiko Stuebner <heiko@sntech.de>
Acked-by: Neil Armstrong <narmstrong@baylibre.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Tested-by: Florian Fainelli <f.fainelli@gmail> # Lamobo R1 (fixed-link + MDIO sub node for roboswitch).
Signed-off-by: David S. Miller <davem@davemloft.net>
Add one notifier for udev changes net device name.
Fixes: b6601323ef9e ("net: stmmac: debugfs entry name is not be changed when udev rename")
Signed-off-by: Jiping Ma <jiping.ma2@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Whenever adding new member of rule object we attach it to 2 lists,
These 2 lists should be initialized first.
Fixes: 41d0707415 ("net/mlx5: DR, Expose steering rule functionality")
Signed-off-by: Erez Shitrit <erezsh@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Set hairpin table size to the corret size, based on the groups that
would be created in it. Groups are laid out on the table such that a
group occupies a range of entries in the table. This implies that the
group ranges should have correspondence to the table they are laid upon.
The patch cited below made group 1's size to grow hence causing
overflow of group range laid on the table.
Fixes: a795d8db2a ("net/mlx5e: Support RSS for IP-in-IP and IPv6 tunneled packets")
Signed-off-by: Eli Cohen <eli@mellanox.com>
Signed-off-by: Mark Bloch <markb@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
No need for an atomic refcounter for the STE and hashtables.
These are internal SW steering resources and they are always
under domain mutex.
This also fixes the following refcount error:
refcount_t: addition on 0; use-after-free.
WARNING: CPU: 9 PID: 3527 at lib/refcount.c:25 refcount_warn_saturate+0x81/0xe0
Call Trace:
dr_table_init_nic+0x10d/0x110 [mlx5_core]
mlx5dr_table_create+0xb4/0x230 [mlx5_core]
mlx5_cmd_dr_create_flow_table+0x39/0x120 [mlx5_core]
__mlx5_create_flow_table+0x221/0x5f0 [mlx5_core]
esw_create_offloads_fdb_tables+0x180/0x5a0 [mlx5_core]
...
Fixes: 26d688e33f ("net/mlx5: DR, Add Steering entry (STE) utilities")
Signed-off-by: Yevgeny Kliteynik <kliteyn@mellanox.com>
Reviewed-by: Alex Vesker <valex@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Register devlink before interfaces are added.
This will allow interfaces to use devlink while initalizing. For example,
call mlx5_is_roce_enabled.
Fixes: aba25279c1 ("net/mlx5e: Add TX reporter support")
Signed-off-by: Michael Guralnik <michaelgur@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
In case a reporter exists, error message is logged only to the devlink
tracer. The devlink tracer is a visibility utility only, which user can
choose not to monitor.
After cited patch, 3rd party monitoring tools that tracks these error
message will no longer find them in dmesg, causing a regression.
With this patch, error messages are also logged into the dmesg.
Fixes: c50de4af1d ("net/mlx5e: Generalize tx reporter's functionality")
Signed-off-by: Eran Ben Elisha <eranbe@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Function entries were duplicated accidentally, removing the dups.
Fixes: ea4b4d7fc1 ("net: atlantic: loopback tests via private flags")
Signed-off-by: Igor Russkikh <irusskikh@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Initial loopback configuration should be called earlier, before
starting traffic on HW blocks. Otherwise depending on race conditions
it could be kept disabled.
Fixes: ea4b4d7fc1 ("net: atlantic: loopback tests via private flags")
Signed-off-by: Igor Russkikh <irusskikh@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Last code/checkpatch cleanup did a copy paste error where code from
firmware 3 API logic was moved to firmware 1 logic.
This resulted in FW1.x users would never see the link state as active.
Fixes: 7b0c342f1f ("net: atlantic: code style cleanup")
Signed-off-by: Igor Russkikh <irusskikh@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Fix up inconsistent usage of upper and lowercase letters in "Samsung"
name.
"SAMSUNG" is not an abbreviation but a regular trademarked name.
Therefore it should be written with lowercase letters starting with
capital letter.
Although advertisement materials usually use uppercase "SAMSUNG", the
lowercase version is used in all legal aspects (e.g. on Wikipedia and in
privacy/legal statements on
https://www.samsung.com/semiconductor/privacy-global/).
Signed-off-by: Krzysztof Kozlowski <krzk@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
The only clk init function in this driver that register a clk is
fu540_c000_clk_init(), and thus we need to unregister the clk when this
driver is removed on that platform. Other init functions, for example
macb_clk_init(), don't register clks and therefore we shouldn't
unregister the clks when this driver is removed. Convert this
registration path to devm so it gets auto-unregistered when this driver
is removed and drop the clk_unregister() calls in driver remove (and
error paths) so that we don't erroneously remove a clk from the system
that isn't registered by this driver.
Otherwise we get strange crashes with a use-after-free when the
devm_clk_get() call in macb_clk_init() calls clk_put() on a clk pointer
that has become invalid because it is freed in clk_unregister().
Cc: Nicolas Ferre <nicolas.ferre@microchip.com>
Cc: Yash Shah <yash.shah@sifive.com>
Reported-by: Guenter Roeck <linux@roeck-us.net>
Fixes: c218ad5590 ("macb: Add support for SiFive FU540-C000")
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Atomic operations that span cache lines are super-expensive on x86
(not just to the current processor, but also to other processes as all
memory operations are blocked until the operation completes). Upcoming
x86 processors have a switch to cause such operations to generate a #AC
trap. It is expected that some real time systems will enable this mode
in BIOS.
In preparation for this, it is necessary to fix code that may execute
atomic instructions with operands that cross cachelines because the #AC
trap will crash the kernel.
Since "pwol_mask" is local and never exposed to concurrency, there is
no need to set bits in pwol_mask using atomic operations.
Directly operate on the byte which contains the bit instead of using
__set_bit() to avoid any big endian concern due to type cast to
unsigned long in __set_bit().
Suggested-by: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Fenghua Yu <fenghua.yu@intel.com>
Signed-off-by: Tony Luck <tony.luck@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Current code use dma_wmb() to ensure Rx/Tx descriptors are visible
to device before writing to doorbell.
However, these dma_wmb() are wrong and unnecessary. Therefore,
they should be removed.
iowrite32be() called from gve_rx_write_doorbell()/gve_tx_put_doorbell()
should guaratee that all previous writes to WB/UC memory is visible to
device before the write done by iowrite32be().
E.g. On ARM64, iowrite32be() calls __iowmb() which expands to dma_wmb()
and only then calls __raw_writel().
Reviewed-by: Si-Wei Liu <si-wei.liu@oracle.com>
Signed-off-by: Liran Alon <liran.alon@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In order to dump the FECs registers the clocks have to be ticking,
otherwise a data abort occurs. Add calls to runtime PM so they are
enabled and later disabled.
Fixes: e8fcfcd568 ("net: fec: optimize the clock management to save power")
Reported-by: Chris Healy <Chris.Healy@zii.aero>
Signed-off-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
As per 802.3-2005, Section Two, Annex 28B, Table 28B-2 [1], when
_only_ Rx pause is enabled, both symmetric and asymmetric pause
towards local device must be enabled. Also, firmware returns the local
device's flow control pause params as part of advertised capabilities
and negotiated params as part of current link attributes. So, fix up
ethtool's flow control pause params fetch logic to read from acaps,
instead of linkattr.
[1] https://standards.ieee.org/standard/802_3-2005.html
Fixes: c3168cabe1 ("cxgb4/cxgbvf: Handle 32-bit fw port capabilities")
Signed-off-by: Surendra Mobiya <surendra@chelsio.com>
Signed-off-by: Rahul Lakkireddy <rahul.lakkireddy@chelsio.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Currently, VRRP packets and packets that hit exceptions during routing
(e.g., MTU error) are policed using the same policer towards the CPU.
This means, for example, that misconfiguration of the MTU on a routed
interface can prevent VRRP packets from reaching the CPU, which in turn
can cause the VRRP daemon to assume it is the Master router.
Fix this by using a dedicated policer for VRRP packets.
Fixes: 11566d34f8 ("mlxsw: spectrum: Add VRRP traps")
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Reported-by: Alex Veber <alexve@mellanox.com>
Tested-by: Alex Veber <alexve@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When a router interface (RIF) is created the MAC address of the backing
netdev is verified to have the same MSBs as existing RIFs. This is
required in order to avoid changing existing RIF MAC addresses that all
share the same MSBs.
Loopback RIFs are special in this regard as they do not have a MAC
address, given they are only used to loop packets from the overlay to
the underlay.
Without this change, an error is returned when trying to create a RIF
after the creation of a GRE tunnel that is represented by a loopback
RIF. 'rif->dev->dev_addr' points to the GRE device's local IP, which
does not share the same MSBs as physical interfaces. Adding an IP
address to any physical interface results in:
Error: mlxsw_spectrum: All router interface MAC addresses must have the
same prefix.
Fix this by skipping loopback RIFs during MAC validation.
Fixes: 74bc993974 ("mlxsw: spectrum_router: Veto unsupported RIF MAC addresses")
Signed-off-by: Amit Cohen <amitc@mellanox.com>
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
GXBB and newer SoCs use the fixed FCLK_DIV2 (1GHz) clock as input for
the m250_sel clock. Meson8b and Meson8m2 use MPLL2 instead, whose rate
can be adjusted at runtime.
So far we have been running MPLL2 with ~250MHz (and the internal
m250_div with value 1), which worked enough that we could transfer data
with an TX delay of 4ns. Unfortunately there is high packet loss with
an RGMII PHY when transferring data (receiving data works fine though).
Odroid-C1's u-boot is running with a TX delay of only 2ns as well as
the internal m250_div set to 2 - no lost (TX) packets can be observed
with that setting in u-boot.
Manual testing has shown that the TX packet loss goes away when using
the following settings in Linux (the vendor kernel uses the same
settings):
- MPLL2 clock set to ~500MHz
- m250_div set to 2
- TX delay set to 2ns on the MAC side
Update the m250_div divider settings to only accept dividers greater or
equal 2 to fix the TX delay generated by the MAC.
iperf3 results before the change:
[ ID] Interval Transfer Bitrate Retr
[ 5] 0.00-10.00 sec 182 MBytes 153 Mbits/sec 514 sender
[ 5] 0.00-10.00 sec 182 MBytes 152 Mbits/sec receiver
iperf3 results after the change (including an updated TX delay of 2ns):
[ ID] Interval Transfer Bitrate Retr Cwnd
[ 5] 0.00-10.00 sec 927 MBytes 778 Mbits/sec 0 sender
[ 5] 0.00-10.01 sec 927 MBytes 777 Mbits/sec receiver
Fixes: 4f6a71b84e ("net: stmmac: dwmac-meson8b: fix internal RGMII clock configuration")
Signed-off-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
While testing max vlan configuration on the PF, firmware gets
assert as driver was configuring number of vlans more than what
is supported per port/engine, it was figured out that there is an
implicit vlan (hidden default vlan consuming hardware cam entry resource)
which is configured default for all the clients (PF/VFs) on client_init
ramrod by the adapter implicitly, so when allocating resources among the
PFs this implicit vlan should be considered or total vlan entries should
be reduced by one to accommodate that default/implicit vlan entry.
Signed-off-by: Manish Chopra <manishc@marvell.com>
Signed-off-by: Ariel Elior <aelior@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Although it has same value as MAX_MAC_CREDIT_E2,
use MAX_VLAN_CREDIT_E2 appropriately.
Signed-off-by: Manish Chopra <manishc@marvell.com>
Signed-off-by: Ariel Elior <aelior@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
On the error path some fragments remain DMA mapped. Adding a fix
that unmaps all the fragments. Rework cleanup path to be simpler.
Fixes: 8151ee88ba ("dpaa_eth: use page backed rx buffers")
Signed-off-by: Madalin Bucur <madalin.bucur@oss.nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The burning process requires to perform internal allocations of large
chunks of memory. This memory doesn't need to be contiguous and can be
safely allocated by vzalloc() instead of kzalloc(). This patch changes
such allocation to avoid possible out-of-memory failure.
Fixes: 410ed13cae ("Add the mlxfw module for Mellanox firmware flash process")
Signed-off-by: Vladyslav Tarasiuk <vladyslavt@mellanox.com>
Reviewed-by: Aya Levin <ayal@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Tested-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Pull networking fixes from David Miller:
1) Several nf_flow_table_offload fixes from Pablo Neira Ayuso,
including adding a missing ipv6 match description.
2) Several heap overflow fixes in mwifiex from qize wang and Ganapathi
Bhat.
3) Fix uninit value in bond_neigh_init(), from Eric Dumazet.
4) Fix non-ACPI probing of nxp-nci, from Stephan Gerhold.
5) Fix use after free in tipc_disc_rcv(), from Tuong Lien.
6) Enforce limit of 33 tail calls in mips and riscv JIT, from Paul
Chaignon.
7) Multicast MAC limit test is off by one in qede, from Manish Chopra.
8) Fix established socket lookup race when socket goes from
TCP_ESTABLISHED to TCP_LISTEN, because there lacks an intervening
RCU grace period. From Eric Dumazet.
9) Don't send empty SKBs from tcp_write_xmit(), also from Eric Dumazet.
10) Fix active backup transition after link failure in bonding, from
Mahesh Bandewar.
11) Avoid zero sized hash table in gtp driver, from Taehee Yoo.
12) Fix wrong interface passed to ->mac_link_up(), from Russell King.
13) Fix DSA egress flooding settings in b53, from Florian Fainelli.
14) Memory leak in gmac_setup_txqs(), from Navid Emamdoost.
15) Fix double free in dpaa2-ptp code, from Ioana Ciornei.
16) Reject invalid MTU values in stmmac, from Jose Abreu.
17) Fix refcount leak in error path of u32 classifier, from Davide
Caratti.
18) Fix regression causing iwlwifi firmware crashes on boot, from Anders
Kaseorg.
19) Fix inverted return value logic in llc2 code, from Chan Shu Tak.
20) Disable hardware GRO when XDP is attached to qede, frm Manish
Chopra.
21) Since we encode state in the low pointer bits, dst metrics must be
at least 4 byte aligned, which is not necessarily true on m68k. Add
annotations to fix this, from Geert Uytterhoeven.
* git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (160 commits)
sfc: Include XDP packet headroom in buffer step size.
sfc: fix channel allocation with brute force
net: dst: Force 4-byte alignment of dst_metrics
selftests: pmtu: fix init mtu value in description
hv_netvsc: Fix unwanted rx_table reset
net: phy: ensure that phy IDs are correctly typed
mod_devicetable: fix PHY module format
qede: Disable hardware gro when xdp prog is installed
net: ena: fix issues in setting interrupt moderation params in ethtool
net: ena: fix default tx interrupt moderation interval
net/smc: unregister ib devices in reboot_event
net: stmmac: platform: Fix MDIO init for platforms without PHY
llc2: Fix return statement of llc_stat_ev_rx_null_dsap_xid_c (and _test_c)
net: hisilicon: Fix a BUG trigered by wrong bytes_compl
net: dsa: ksz: use common define for tag len
s390/qeth: don't return -ENOTSUPP to userspace
s390/qeth: fix promiscuous mode after reset
s390/qeth: handle error due to unsupported transport mode
cxgb4: fix refcount init for TC-MQPRIO offload
tc-testing: initial tdc selftests for cls_u32
...
Correct a mismatch between rx_page_buf_step and the actual step size
used when filling buffer pages.
This patch fixes the page overrun that occured when the MTU was set to
anything bigger than 1692.
Fixes: 3990a8fffb ("sfc: allocate channels for XDP tx queues")
Signed-off-by: Charles McLachlan <cmclachlan@solarflare.com>
Signed-off-by: Edward Cree <ecree@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
It was possible for channel allocation logic to get confused between what
it had and what it wanted, and end up trying to use the same channel for
both PTP and regular TX. This led to a kernel panic:
BUG: unable to handle page fault for address: 0000000000047635
#PF: supervisor write access in kernel mode
#PF: error_code(0x0002) - not-present page
PGD 0 P4D 0
Oops: 0002 [#1] SMP PTI
CPU: 0 PID: 0 Comm: swapper/0 Tainted: G W 5.4.0-rc3-ehc14+ #900
Hardware name: Dell Inc. PowerEdge R710/0M233H, BIOS 6.4.0 07/23/2013
RIP: 0010:native_queued_spin_lock_slowpath+0x188/0x1e0
Code: f3 90 48 8b 32 48 85 f6 74 f6 eb e8 c1 ee 12 83 e0 03 83 ee 01 48 c1 e0 05 48 63 f6 48 05 c0 98 02 00 48 03 04 f5 a0 c6 ed 81 <48> 89 10 8b 42 08 85 c0 75 09 f3 90 8b 42 08 85 c0 74 f7 48 8b 32
RSP: 0018:ffffc90000003d28 EFLAGS: 00010006
RAX: 0000000000047635 RBX: 0000000000000246 RCX: 0000000000040000
RDX: ffff888627a298c0 RSI: 0000000000003ffe RDI: ffff88861f6b8dd4
RBP: ffff8886225c6e00 R08: 0000000000040000 R09: 0000000000000000
R10: 0000000616f080c6 R11: 00000000000000c0 R12: ffff88861f6b8dd4
R13: ffffc90000003dc8 R14: ffff88861942bf00 R15: ffff8886150f2000
FS: 0000000000000000(0000) GS:ffff888627a00000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 0000000000047635 CR3: 000000000200a000 CR4: 00000000000006f0
Call Trace:
<IRQ>
_raw_spin_lock_irqsave+0x22/0x30
skb_queue_tail+0x1b/0x50
sock_queue_err_skb+0x9d/0xf0
__skb_complete_tx_timestamp+0x9d/0xc0
efx_dequeue_buffer+0x126/0x180 [sfc]
efx_xmit_done+0x73/0x1c0 [sfc]
efx_ef10_ev_process+0x56a/0xfe0 [sfc]
? tick_sched_do_timer+0x60/0x60
? timerqueue_add+0x5d/0x70
? enqueue_hrtimer+0x39/0x90
efx_poll+0x111/0x380 [sfc]
? rcu_accelerate_cbs+0x50/0x160
net_rx_action+0x14a/0x400
__do_softirq+0xdd/0x2d0
irq_exit+0xa0/0xb0
do_IRQ+0x53/0xe0
common_interrupt+0xf/0xf
</IRQ>
In the long run we intend to rewrite the channel allocation code, but for
'net' fix this by allocating extra_channels, and giving them TX queues,
even if we do not in fact need them (e.g. on NICs without MAC TX
timestamping), and thereby using simpler logic to assign the channels
once they're allocated.
Fixes: 3990a8fffb ("sfc: allocate channels for XDP tx queues")
Signed-off-by: Edward Cree <ecree@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
commit 18c602dee4 ("qede: Use NETIF_F_GRO_HW.") introduced
a regression in driver that when xdp program is installed on
qede device, device's aggregation feature (hardware GRO) is not
getting disabled, which is unexpected with xdp.
Fixes: 18c602dee4 ("qede: Use NETIF_F_GRO_HW.")
Signed-off-by: Manish Chopra <manishc@marvell.com>
Signed-off-by: Ariel Elior <aelior@marvell.com>
Reviewed-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Issue 1:
--------
Reproduction steps:
1. sudo ethtool -C eth0 rx-usecs 128
2. sudo ethtool -C eth0 adaptive-rx on
3. sudo ethtool -C eth0 adaptive-rx off
4. ethtool -c eth0
expected output: rx-usecs 128
actual output: rx-usecs 0
Reason for issue:
In stage 3, ethtool userspace calls first the ena_get_coalesce() handler
to get the current value of all properties, and then the ena_set_coalesce()
handler. When ena_get_coalesce() is called the adaptive interrupt
moderation is still on. There is an if in the code that returns the
rx_coalesce_usecs only if the adaptive interrupt moderation is off.
And since it is still on, rx_coalesce_usecs is not set, meaning it
stays 0.
Solution to issue:
Remove this if static interrupt moderation intervals have nothing to do
with dynamic ones.
Issue 2:
--------
Reproduction steps:
1. sudo ethtool -C eth0 adaptive-rx on
2. sudo ethtool -C eth0 rx-usecs 128
3. ethtool -c eth0
expected output: rx-usecs 128
actual output: rx-usecs 0
Reason for issue:
In stage 2, when ena_set_coalesce() is called, the handler tests if
rx adaptive interrupt moderation is on, and if it is, it returns before
getting to the part in the function that sets the rx non-adaptive
interrupt moderation interval.
Solution to issue:
Remove the return from the function when rx adaptive interrupt moderation
is on.
Also cleaned up the fixed code in ena_set_coalesce by grouping together
adaptive interrupt moderation toggling, and using && instead of nested
ifs.
Fixes: b3db86dc4b ("net: ena: reimplement set/get_coalesce()")
Fixes: 0eda847953 ("net: ena: fix retrieval of nonadaptive interrupt moderation intervals")
Fixes: 1738cd3ed3 ("net: ena: Add a driver for Amazon Elastic Network Adapters (ENA)")
Signed-off-by: Arthur Kiyanovski <akiyano@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Current default non-adaptive tx interrupt moderation interval is 196 us.
This value is too high and might cause the tx queue to fill up.
In this commit we set the default non-adaptive tx interrupt moderation
interval to 64 us in order to:
1. Reduce the probability of the queue filling-up (when compared to the
current default value of 196 us).
2. Reduce unnecessary tx interrupt overhead (which happens if we set the
default tx interval to 0).
We determined experimentally that 64 us is an optimal value that
reduces interrupt rate by more than 20% without affecting performance.
Fixes: 1738cd3ed3 ("net: ena: Add a driver for Amazon Elastic Network Adapters (ENA)")
Signed-off-by: Arthur Kiyanovski <akiyano@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The current implementation of "stmmac_dt_phy" function initializes
the MDIO platform bus data, even in the absence of PHY. This fix
will skip MDIO initialization if there is no PHY present.
Fixes: 7437127 ("net: stmmac: Convert to phylink and remove phylib logic")
Acked-by: Jayati Sahu <jayati.sahu@samsung.com>
Signed-off-by: Sriram Dash <sriram.dash@samsung.com>
Signed-off-by: Padmanabhan Rajanbabu <p.rajanbabu@samsung.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When doing stress test, we get the following trace:
kernel BUG at lib/dynamic_queue_limits.c:26!
Internal error: Oops - BUG: 0 [#1] SMP ARM
Modules linked in: hip04_eth
CPU: 0 PID: 2003 Comm: tDblStackPcap0 Tainted: G O L 4.4.197 #1
Hardware name: Hisilicon A15
task: c3637668 task.stack: de3bc000
PC is at dql_completed+0x18/0x154
LR is at hip04_tx_reclaim+0x110/0x174 [hip04_eth]
pc : [<c041abfc>] lr : [<bf0003a8>] psr: 800f0313
sp : de3bdc2c ip : 00000000 fp : c020fb10
r10: 00000000 r9 : c39b4224 r8 : 00000001
r7 : 00000046 r6 : c39b4000 r5 : 0078f392 r4 : 0078f392
r3 : 00000047 r2 : 00000000 r1 : 00000046 r0 : df5d5c80
Flags: Nzcv IRQs on FIQs on Mode SVC_32 ISA ARM Segment user
Control: 32c5387d Table: 1e189b80 DAC: 55555555
Process tDblStackPcap0 (pid: 2003, stack limit = 0xde3bc190)
Stack: (0xde3bdc2c to 0xde3be000)
[<c041abfc>] (dql_completed) from [<bf0003a8>] (hip04_tx_reclaim+0x110/0x174 [hip04_eth])
[<bf0003a8>] (hip04_tx_reclaim [hip04_eth]) from [<bf0012c0>] (hip04_rx_poll+0x20/0x388 [hip04_eth])
[<bf0012c0>] (hip04_rx_poll [hip04_eth]) from [<c04c8d9c>] (net_rx_action+0x120/0x374)
[<c04c8d9c>] (net_rx_action) from [<c021eaf4>] (__do_softirq+0x218/0x318)
[<c021eaf4>] (__do_softirq) from [<c021eea0>] (irq_exit+0x88/0xac)
[<c021eea0>] (irq_exit) from [<c0240130>] (msa_irq_exit+0x11c/0x1d4)
[<c0240130>] (msa_irq_exit) from [<c0267ba8>] (__handle_domain_irq+0x110/0x148)
[<c0267ba8>] (__handle_domain_irq) from [<c0201588>] (gic_handle_irq+0xd4/0x118)
[<c0201588>] (gic_handle_irq) from [<c0558360>] (__irq_svc+0x40/0x58)
Exception stack(0xde3bdde0 to 0xde3bde28)
dde0: 00000000 00008001 c3637668 00000000 00000000 a00f0213 dd3627a0 c0af6380
de00: c086d380 a00f0213 c0a22a50 de3bde6c 00000002 de3bde30 c0558138 c055813c
de20: 600f0213 ffffffff
[<c0558360>] (__irq_svc) from [<c055813c>] (_raw_spin_unlock_irqrestore+0x44/0x54)
Kernel panic - not syncing: Fatal exception in interrupt
Pre-modification code:
int hip04_mac_start_xmit(struct sk_buff *skb, struct net_device *ndev)
{
[...]
[1] priv->tx_head = TX_NEXT(tx_head);
[2] count++;
[3] netdev_sent_queue(ndev, skb->len);
[...]
}
An rx interrupt occurs if hip04_mac_start_xmit just executes to the line 2,
tx_head has been updated, but corresponding 'skb->len' has not been
added to dql_queue.
And then
hip04_mac_interrupt->__napi_schedule->hip04_rx_poll->hip04_tx_reclaim
In hip04_tx_reclaim, because tx_head has been updated,
bytes_compl will plus an additional "skb-> len"
which has not been added to dql_queue. And then
trigger the BUG_ON(bytes_compl > num_queued - dql->num_completed).
To solve the problem described above, we put
"netdev_sent_queue(ndev, skb->len);"
before
"priv->tx_head = TX_NEXT(tx_head);"
Fixes: a41ea46a9a ("net: hisilicon: new hip04 ethernet driver")
Signed-off-by: Jiangfeng Xiao <xiaojiangfeng@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch fixes the case where the PHY isn't described in the device
tree. This is due to the way the MDIO bus is registered in the driver:
whether the PHY is described in the device tree or not, the bus is
registered through of_mdiobus_register. The function masks all the PHYs
and only allow probing the ones described in the device tree. Prior to
the Phylink conversion this was also done but later on in the driver
the MDIO bus was manually scanned to circumvent the fact that the PHY
wasn't described.
This patch fixes it in a proper way, by registering the MDIO bus based
on if the PHY attached to a given interface is described in the device
tree or not.
Fixes: 7897b071ac ("net: macb: convert to phylink")
Signed-off-by: Antoine Tenart <antoine.tenart@bootlin.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Daniel Borkmann says:
====================
pull-request: bpf 2019-12-19
The following pull-request contains BPF updates for your *net* tree.
We've added 10 non-merge commits during the last 8 day(s) which contain
a total of 21 files changed, 269 insertions(+), 108 deletions(-).
The main changes are:
1) Fix lack of synchronization between xsk wakeup and destroying resources
used by xsk wakeup, from Maxim Mikityanskiy.
2) Fix pruning with tail call patching, untrack programs in case of verifier
error and fix a cgroup local storage tracking bug, from Daniel Borkmann.
3) Fix clearing skb->tstamp in bpf_redirect() when going from ingress to
egress which otherwise cause issues e.g. on fq qdisc, from Lorenz Bauer.
4) Fix compile warning of unused proc_dointvec_minmax_bpf_restricted() when
only cBPF is present, from Alexander Lobakin.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Use synchronize_rcu to wait until the XSK wakeup function finishes
before destroying the resources it uses:
1. ixgbe_down already calls synchronize_rcu after setting __IXGBE_DOWN.
2. After switching the XDP program, call synchronize_rcu to let
ixgbe_xsk_wakeup exit before the XDP program is freed.
3. Changing the number of channels brings the interface down.
4. Disabling UMEM sets __IXGBE_TX_DISABLED before closing hardware
resources and resetting xsk_umem. Check that bit in ixgbe_xsk_wakeup to
avoid using the XDP ring when it's already destroyed. synchronize_rcu is
called from ixgbe_txrx_ring_disable.
Signed-off-by: Maxim Mikityanskiy <maximmi@mellanox.com>
Signed-off-by: Björn Töpel <bjorn.topel@intel.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Link: https://lore.kernel.org/bpf/20191217162023.16011-5-maximmi@mellanox.com
Use synchronize_rcu to wait until the XSK wakeup function finishes
before destroying the resources it uses:
1. i40e_down already calls synchronize_rcu. On i40e_down either
__I40E_VSI_DOWN or __I40E_CONFIG_BUSY is set. Check the latter in
i40e_xsk_wakeup (the former is already checked there).
2. After switching the XDP program, call synchronize_rcu to let
i40e_xsk_wakeup exit before the XDP program is freed.
3. Changing the number of channels brings the interface down (see
i40e_prep_for_reset and i40e_pf_quiesce_all_vsi).
4. Disabling UMEM sets __I40E_CONFIG_BUSY, too.
Signed-off-by: Maxim Mikityanskiy <maximmi@mellanox.com>
Signed-off-by: Björn Töpel <bjorn.topel@intel.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Link: https://lore.kernel.org/bpf/20191217162023.16011-4-maximmi@mellanox.com
After disabling resources necessary for XSK (the XDP program, channels,
XSK queues), use synchronize_rcu to wait until the XSK wakeup function
finishes, before freeing the resources.
Suspend XSK wakeups during switching channels. If the XDP program is
being removed, synchronize_rcu before closing the old channels to allow
XSK wakeup to complete.
Signed-off-by: Maxim Mikityanskiy <maximmi@mellanox.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Link: https://lore.kernel.org/bpf/20191217162023.16011-3-maximmi@mellanox.com
If TX Coalesce timer is enabled we should always arm it, otherwise we
may hit the case where an interrupt is missed and the TX Queue will
timeout.
Arming the timer does not necessarly mean it will run the tx_clean()
because this function is wrapped around NAPI launcher.
Fixes: 9125cdd1be ("stmmac: add the initial tx coalesce schema")
Signed-off-by: Jose Abreu <Jose.Abreu@synopsys.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
XGMAC supports maximum MTU that can go to 16KB. Lets add this check in
the calculation of RX buffer size.
Fixes: 7ac6653a08 ("stmmac: Move the STMicroelectronics driver")
Signed-off-by: Jose Abreu <Jose.Abreu@synopsys.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The 16KB RX Buffer must also be 16 byte aligned. Fix it.
Fixes: 7ac6653a08 ("stmmac: Move the STMicroelectronics driver")
Signed-off-by: Jose Abreu <Jose.Abreu@synopsys.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Stefan Assmann