Only use the TROCR register on R-Car Gen3 as it is not present on other
SoCs.

Offsets used for the undocumented registers are considered reserved and
should not be written to. After some internal investigation with Renesas it
remains unclear why this driver accesses these fields on R-Car Gen2 but
regardless of what the historical reasons are the current code is
considered incorrect.
Signed-off-by: NSimon Horman <horms+renesas@verge.net.au>
Reviewed-by: NYoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
Acked-by: NSergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch removes the use of the undocumented BOC bit of the CCC register.

Current documentation for EtherAVB (ravb) describes the offset of what the
driver uses as the BOC bit as reserved and that only a value of 0 should be
written. After some internal investigation with Renesas it remains unclear
why this driver accesses these fields but regardless of what the historical
reasons are the current code is considered incorrect.

Based on work by Kazuya Mizuguchi <kazuya.mizuguchi.ks@renesas.com>
Signed-off-by: NSimon Horman <horms+renesas@verge.net.au>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch removes the use of the undocumented counter registers
CDCR, LCCR, CERCR, CEECR.

Offsets used for undocumented registers are considered reserved and
should not be written to. After some internal investigation with Renesas
it remains unclear why this driver accesses these fields but regardless of
what the historical reasons are the current code is considered incorrect.

Based on work by Kazuya Mizuguchi <kazuya.mizuguchi.ks@renesas.com>
Signed-off-by: NSimon Horman <horms+renesas@verge.net.au>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

When a Tx timestamp is requested, a pointer to the skb is stored in the
ravb_tstamp_skb struct. This was done without an skb_get. There exists
the possibility that the skb could be freed by ravb_tx_free (when
ravb_tx_free is called from ravb_start_xmit) before the timestamp was
processed, leading to a use-after-free bug.

Use skb_get when filling a ravb_tstamp_skb struct, and add appropriate
frees/consumes when a ravb_tstamp_skb struct is freed.

Fixes: c156633f ("Renesas Ethernet AVB driver proper")
Signed-off-by: NTho Vu <tho.vu.wh@rvc.renesas.com>
Signed-off-by: NKazuya Mizuguchi <kazuya.mizuguchi.ks@renesas.com>
Signed-off-by: NSimon Horman <horms+renesas@verge.net.au>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

There was NVMEM support added to of_get_mac_address, so it could now
return ERR_PTR encoded error values, so we need to adjust all current
users of of_get_mac_address to this new fact.

While at it, remove superfluous is_valid_ether_addr as the MAC address
returned from of_get_mac_address is always valid and checked by
is_valid_ether_addr anyway.

Fixes: d01f449c ("of_net: add NVMEM support to of_get_mac_address")
Signed-off-by: NPetr Štetiar <ynezz@true.cz>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

According to the R-Car Gen3 Hardware Manual Rev 1.50 of Nov 30, 2018, the
TX clock internal delay mode isn't supported on R-Car E3 (r8a77990) or D3
(r8a77995). And by extension it is also not supported by RZ/G2E (r9a774c0).

This matches all ES versions of the affected SoCs as it is
not clear if this problem will be resolved in newer chips.
This can be revisited, as necessary.

This patch does not error-out if PHY_INTERFACE_MODE_RGMII_ID or
PHY_INTERFACE_MODE_RGMII_TXID are used on SoCs where TX clock delay
mode is not supported as there is a risk of introducing a regression
when used in conjunction with older DT blobs present in the field.
Rather, a warning is logged in such cases.

Based on work by Kazuya Mizuguchi.
Signed-off-by: NSimon Horman <horms+renesas@verge.net.au>
Reviewed-by: NAndrew Lunn <andrew@lunn.ch>
Reviewed-by: NSergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

mmiowb() is now implied by spin_unlock() on architectures that require
it, so there is no reason to call it from driver code. This patch was
generated using coccinelle:

	@mmiowb@
	@@
	- mmiowb();

and invoked as:

$ for d in drivers include/linux/qed sound; do \
spatch --include-headers --sp-file mmiowb.cocci --dir $d --in-place; done

NOTE: mmiowb() has only ever guaranteed ordering in conjunction with
spin_unlock(). However, pairing each mmiowb() removal in this patch with
the corresponding call to spin_unlock() is not at all trivial, so there
is a small chance that this change may regress any drivers incorrectly
relying on mmiowb() to order MMIO writes between CPUs using lock-free
synchronisation. If you've ended up bisecting to this commit, you can
reintroduce the mmiowb() calls using wmb() instead, which should restore
the old behaviour on all architectures other than some esoteric ia64
systems.
Acked-by: NLinus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: NWill Deacon <will.deacon@arm.com>

After the previous patch, all the callers of ndo_select_queue()
provide as a 'fallback' argument netdev_pick_tx.
The only exceptions are nested calls to ndo_select_queue(),
which pass down the 'fallback' available in the current scope
- still netdev_pick_tx.

We can drop such argument and replace fallback() invocation with
netdev_pick_tx(). This avoids an indirect call per xmit packet
in some scenarios (TCP syn, UDP unconnected, XDP generic, pktgen)
with device drivers implementing such ndo. It also clean the code
a bit.

Tested with ixgbe and CONFIG_FCOE=m

With pktgen using queue xmit:
threads		vanilla 	patched
		(kpps)		(kpps)
1		2334		2428
2		4166		4278
4		7895		8100

 v1 -> v2:
 - rebased after helper's name change
Signed-off-by: NPaolo Abeni <pabeni@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Hardware has the CBS (Credit Based Shaper) which affects only Q3
and Q2. When updating the CBS settings, even if the driver does so
after waiting for Tx DMA finished, there is a possibility that frame
data still remains in TxFIFO.

To avoid this, decrease TxFIFO depth of Q3 and Q2 to one.

This patch has been exercised this using netperf TCP_MAERTS, TCP_STREAM
and UDP_STREAM tests run on an Ebisu board. No performance change was
detected, outside of noise in the tests, both in terms of throughput and
CPU utilisation.

Fixes: c156633f ("Renesas Ethernet AVB driver proper")
Signed-off-by: NMasaru Nagai <masaru.nagai.vx@renesas.com>
Signed-off-by: NKazuya Mizuguchi <kazuya.mizuguchi.ks@renesas.com>
[simon: updated changelog]
Signed-off-by: NSimon Horman <horms+renesas@verge.net.au>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

EtherAVB may provide a checksum of packet data appended to packet data. In
order to allow this checksum to be received by the host descriptor data
needs to be enlarged by 2 bytes to accommodate the checksum.

In the case of MTU-sized packets without a VLAN tag the
checksum were already accommodated by virtue of the space reserved for the
VLAN tag. However, a packet of MTU-size with a  VLAN tag consumed all
packet data space provided by a descriptor leaving no space for the
trailing checksum.

This was not detected by the driver which incorrectly used the last two
bytes of packet data as the checksum and truncate the packet by two bytes.
This resulted all such packets being dropped.

A work around is to disable RX checksum offload
 # ethtool -K eth0 rx off

This patch resolves this problem by increasing the size available for
packet data in RX descriptors by two bytes.

Tested on R-Car E3 (r8a77990) ES1.0 based Ebisu-4D board

v2
* Use sizeof(__sum16) directly rather than adding a driver-local
  #define for the size of the checksum provided by the hw (2 bytes).

Fixes: 4d86d381 ("ravb: RX checksum offload")
Signed-off-by: NSimon Horman <horms+renesas@verge.net.au>
Reviewed-by: NSergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Since only full-duplex operation is supported by the
hardware, remove duplex handling code and keep the
register setting of ECMR.DM fixed at 1.

This updates the driver implementation to follow the
data sheet text "This bit should always be set to 1."

Fixes: c156633f ("Renesas Ethernet AVB driver proper")
Signed-off-by: NMagnus Damm <damm+renesas@opensource.se>
Reviewed-by: NSergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

According to the data sheet the Ethernet-AVB hardware in R-Car Gen3
and R-Car Gen2 SoCs do not support half duplex operation. So update
the driver to mark 100Mbit and 1Gbps HDX as unsupported.

Fixes: c156633f ("Renesas Ethernet AVB driver proper")
Signed-off-by: NMagnus Damm <damm+renesas@opensource.se>
Reviewed-by: NSergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The previous commit to ravb had the side effect of making the PHY
advertise Pause and Asym Pause, which previously did not happen.  By
default, phydev->supported has both forms of pause enabled, but
phydev->advertising does not. The new phy_remove_link_mode() copies
phydev->supported to phydev->advertising after removing the requested
link mode. These Pause configuration bits appears it stops the PHY
from completing Auto-Neg and the link remains down.  Be explicit and
remove the Pause and Asym Pause modes, so restoring the old behavior.

Fixes: 41124fa6 ("net: ethernet: Add helper to remove a supported link mode")
Reported-by: NSimon Horman <horms@verge.net.au>
Signed-off-by: NAndrew Lunn <andrew@lunn.ch>
Reviewed-by: NSergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch sets from two descriptor to one descriptor because R-Car Gen3
does not have the 4 bytes alignment restriction of the transmission buffer.
Signed-off-by: NKazuya Mizuguchi <kazuya.mizuguchi.ks@renesas.com>
Signed-off-by: NSimon Horman <horms+renesas@verge.net.au>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

EtherAVB hardware requires 0 to be written to status register bits in
order to clear them, however, care must be taken not to:

1. Clear other bits, by writing zero to them
2. Write one to reserved bits

This patch corrects the ravb driver with respect to the second point above.
This is done by defining reserved bit masks for the affected registers and,
after auditing the code, ensure all sites that may write a one to a
reserved bit use are suitably masked.
Signed-off-by: NKazuya Mizuguchi <kazuya.mizuguchi.ks@renesas.com>
Signed-off-by: NSimon Horman <horms+renesas@verge.net.au>
Reviewed-by: NSergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Some MAC hardware cannot support a subset of link modes. e.g. often
1Gbps Full duplex is supported, but Half duplex is not. Add a helper
to remove such a link mode.
Signed-off-by: NAndrew Lunn <andrew@lunn.ch>
Reviewed-by: NFlorian Fainelli <f.fainelli@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NWolfram Sang <wsa+renesas@sang-engineering.com>
Acked-by: NSergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Reviewed-by: NSimon Horman <horms+renesas@verge.net.au>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The wrong helper is used to swap the bytes when adding the lower bits of
the TX descriptors tag field in the shared ds_tagl variable. The
variable contains the DS[11:0] field and then the TAG[3:0] bits.

The mistake was highlighted by the sparse warning:

ravb_main.c:1622:31:    left side has type restricted __le16
ravb_main.c:1622:31:    right side has type unsigned short
ravb_main.c:1622:31: warning: invalid assignment: |=
ravb_main.c:1622:34: warning: cast to restricted __le16
Signed-off-by: NNiklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Reviewed-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Acked-by: NSergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This fixes sparse warning:

ravb_main.c:1257 ravb_get_strings() error: memcpy() '*ravb_gstrings_stats' too small (32 vs 960)
Signed-off-by: NNiklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Reviewed-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Acked-by: NSergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Inside a loop in ravb_get_ethtool_stats() a variable 'stats' is declared
resulting in the argument also named 'stats' to be shadowed. Fix this
warning by renaming the unused argument 'stats' to 'estats'.

This fixes the sparse warning:

ravb_main.c:1225:36: originally declared here
ravb_main.c:1233:41: warning: symbol 'stats' shadows an earlier one
Signed-off-by: NNiklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Reviewed-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Acked-by: NSergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch makes it so that instead of passing a void pointer as the
accel_priv we instead pass a net_device pointer as sb_dev. Making this
change allows us to pass the subordinate device through to the fallback
function eventually so that we can keep the actual code in the
ndo_select_queue call as focused on possible on the exception cases.
Signed-off-by: NAlexander Duyck <alexander.h.duyck@intel.com>
Tested-by: NAndrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

The generic phy_ethtool_set_link_ksettings() function from phylib can
be used instead of in-house ravb_set_link_ksettings().
Signed-off-by: NVladimir Zapolskiy <vladimir_zapolskiy@mentor.com>
Reviewed-by: NSergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The generic phy_ethtool_get_link_ksettings() function from phylib can be
used instead of in-house ravb_get_link_ksettings().
Signed-off-by: NVladimir Zapolskiy <vladimir_zapolskiy@mentor.com>
Reviewed-by: NSergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

phy_ethtool_ksettings_get() call does not modify device state or device
driver state, hence there is no need to utilize a driver specific
spinlock.
Signed-off-by: NVladimir Zapolskiy <vladimir_zapolskiy@mentor.com>
Reviewed-by: NSergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The generic phy_ethtool_nway_reset() function from phylib can be used
instead of in-house ravb_nway_reset().
Signed-off-by: NVladimir Zapolskiy <vladimir_zapolskiy@mentor.com>
Reviewed-by: NSergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

There is no need to call a heavyweight phy_start_aneg() for phy
auto-negotiation by ethtool, the phy is already initialized and
link auto-negotiation is started by calling phy_start() from
ravb_phy_start() when a network device is opened.
Signed-off-by: NVladimir Zapolskiy <vladimir_zapolskiy@mentor.com>
Reviewed-by: NSergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The change fixes sleep in atomic context bug, which is encountered
every time when link settings are changed by ethtool.

Since commit 35b5f6b1 ("PHYLIB: Locking fixes for PHY I/O
potentially sleeping") phy_start_aneg() function utilizes a mutex
to serialize changes to phy state, however that helper function is
called in atomic context under a grabbed spinlock, because
phy_start_aneg() is called by phy_ethtool_ksettings_set() and by
replaced phy_ethtool_sset() helpers from phylib.

Now duplex mode setting is enforced in ravb_adjust_link() only, also
now RX/TX is disabled when link is put down or modifications to E-MAC
registers ECMR and GECMR are expected for both cases of checked and
ignored link status pin state from E-MAC interrupt handler.

Fixes: c156633f ("Renesas Ethernet AVB driver proper")
Signed-off-by: NVladimir Zapolskiy <vladimir_zapolskiy@mentor.com>
Reviewed-by: NSergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Since commit 35b5f6b1 ("PHYLIB: Locking fixes for PHY I/O
potentially sleeping") phy_start_aneg() function utilizes a mutex
to serialize changes to phy state, however the helper function is
called in atomic context.

The bug can be reproduced by running "ethtool -r" command, the bug
is reported if CONFIG_DEBUG_ATOMIC_SLEEP build option is enabled.

Fixes: c156633f ("Renesas Ethernet AVB driver proper")
Signed-off-by: NVladimir Zapolskiy <vladimir_zapolskiy@mentor.com>
Reviewed-by: NSergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

When addressing a review comment in a early version of the offending
patch a comment where left in which should have been removed. Remove the
comment to keep it consistent with the code.

Fixes: 75efa06f ("ravb: add support for changing MTU")
Reported-by: NSergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Signed-off-by: NNiklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Acked-by: NSergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Allow for changing the MTU within the limit of the maximum size of a
descriptor (2048 bytes). Add the callback to change MTU from user-space
and take the configurable MTU into account when configuring the
hardware.
Signed-off-by: NNiklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Currently, if Wake-on-LAN is enabled, the EtherAVB device's module clock
is manually kept running during system suspend, to make sure the device
stays active.

Since commit 91c719f5 ("soc: renesas: rcar-sysc: Keep wakeup
sources active during system suspend") , this workaround is no longer
needed.  Hence remove all explicit clock handling to keep the device
active.
Signed-off-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Reviewed-by: NNiklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Reviewed-by: NSergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Browsing thru the driver disassembly, I noticed that gcc was  able to
figure  out  that the 'ndev' pointer is always non-NULL when calling
free_netdev()  on the probe() method's  error path and  thus skip that
redundant NULL check... gcc is smart, be like gcc! :-)
Signed-off-by: NSergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Reviewed-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This reverts commit fbf3d034.

As of commit 56086910 ("clk: renesas: cpg-mssr: Restore module
clocks during resume"), the workaround is no longer needed.
Signed-off-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Reviewed-by: NNiklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Acked-by: NSergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The module clock is used for two purposes:
  - Wake-on-LAN (WoL), which is optional,
  - gPTP Timer Increment (GTI) configuration, which is mandatory.

As the clock is needed for GTI configuration anyway, WoL is always
available.  Hence remove duplication and repeated obtaining of the clock
by making GTI use the stored clock for WoL use.
Signed-off-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Reviewed-by: NNiklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Reviewed-by: NSergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Add support for RX checksum offload. This is enabled by default and
may be disabled and re-enabled using ethtool:

 # ethtool -K eth0 rx off
 # ethtool -K eth0 rx on

The RAVB provides a simple checksumming scheme which appears to be
completely compatible with CHECKSUM_COMPLETE: sum of all packet data after
the L2 header is appended to packet data; this may be trivially read by the
driver and used to update the skb accordingly.

In terms of performance throughput is close to gigabit line-rate both with
and without RX checksum offload enabled. Perf output, however, appears to
indicate that significantly less time is spent in do_csum(). This is as
expected.

Test results with RX checksum offload enabled:
 # /usr/bin/perf_3.16 record -o /run/perf.data -a netperf -t TCP_MAERTS -H 10.4.3.162
 MIGRATED TCP MAERTS TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 10.4.3.162 () port 0 AF_INET : demo
 enable_enobufs failed: getprotobyname
 Recv   Send    Send
 Socket Socket  Message  Elapsed
 Size   Size    Size     Time     Throughput
 bytes  bytes   bytes    secs.    10^6bits/sec

  87380  16384  16384    10.00     937.54

 Summary of output of perf report:
    18.28%      ksoftirqd/0  [kernel.kallsyms]  [k] _raw_spin_unlock_irqrestore
    10.34%      ksoftirqd/0  [kernel.kallsyms]  [k] __pi_memcpy
     9.83%      ksoftirqd/0  [kernel.kallsyms]  [k] ravb_poll
     7.89%      ksoftirqd/0  [kernel.kallsyms]  [k] skb_put
     4.01%      ksoftirqd/0  [kernel.kallsyms]  [k] dev_gro_receive
     3.37%          netperf  [kernel.kallsyms]  [k] __arch_copy_to_user
     3.17%          swapper  [kernel.kallsyms]  [k] arch_cpu_idle
     2.55%          swapper  [kernel.kallsyms]  [k] tick_nohz_idle_enter
     2.04%      ksoftirqd/0  [kernel.kallsyms]  [k] __pi___inval_dcache_area
     2.03%          swapper  [kernel.kallsyms]  [k] _raw_spin_unlock_irq
     1.96%      ksoftirqd/0  [kernel.kallsyms]  [k] __netdev_alloc_skb
     1.59%      ksoftirqd/0  [kernel.kallsyms]  [k] __slab_alloc.isra.83

Test results without RX checksum offload enabled:
 # /usr/bin/perf_3.16 record -o /run/perf.data -a netperf -t TCP_MAERTS -H 10.4.3.162
 MIGRATED TCP MAERTS TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 10.4.3.162 () port 0 AF_INET : demo
 enable_enobufs failed: getprotobyname
 Recv   Send    Send
 Socket Socket  Message  Elapsed
 Size   Size    Size     Time     Throughput
 bytes  bytes   bytes    secs.    10^6bits/sec

  87380  16384  16384    10.00     940.20

 Summary of output of perf report:
    17.10%    ksoftirqd/0  [kernel.kallsyms]  [k] _raw_spin_unlock_irqrestore
    10.99%    ksoftirqd/0  [kernel.kallsyms]  [k] __pi_memcpy
     8.87%    ksoftirqd/0  [kernel.kallsyms]  [k] ravb_poll
     8.16%    ksoftirqd/0  [kernel.kallsyms]  [k] skb_put
     7.42%    ksoftirqd/0  [kernel.kallsyms]  [k] do_csum
     3.91%    ksoftirqd/0  [kernel.kallsyms]  [k] dev_gro_receive
     2.31%        swapper  [kernel.kallsyms]  [k] arch_cpu_idle
     2.16%    ksoftirqd/0  [kernel.kallsyms]  [k] __pi___inval_dcache_area
     2.14%    ksoftirqd/0  [kernel.kallsyms]  [k] __netdev_alloc_skb
     1.93%        netperf  [kernel.kallsyms]  [k] __arch_copy_to_user
     1.79%        swapper  [kernel.kallsyms]  [k] tick_nohz_idle_enter
     1.63%    ksoftirqd/0  [kernel.kallsyms]  [k] __slab_alloc.isra.83

Above results collected on an R-Car Gen 3 Salvator-X/r8a7796 ES1.0.
Also tested on a R-Car Gen 3 Salvator-X/r8a7795 ES1.0.

By inspection this also appears to be compatible with the ravb found
on R-Car Gen 2 SoCs, however, this patch is currently untested on such
hardware.
Signed-off-by: NSimon Horman <horms+renesas@verge.net.au>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The renesas-cpg-mssr clock driver are not yet aware of PSCI sleep where
power is cut to the SoC. When resuming from this state with WoL enabled
the enable count of the ravb clock is 1 and the clock driver thinks the
clock is already on when PM core enables the clock and increments the
enable count to 2. This will result in the ravb driver failing to talk
to the hardware since the module clock is off. Work around this by
forcing the enable count to 0 and then back to 2 when resuming with WoL
enabled.

This workaround should be reverted once the renesas-cpg-mssr clock
driver becomes aware of this PSCI sleep behavior.
Signed-off-by: NNiklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Acked-by: NSergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

WoL is enabled in the suspend callback by setting MagicPacket detection
and disabling all interrupts expect MagicPacket. In the resume path the
driver needs to reset the hardware to rearm the WoL logic, this prevents
the driver from simply restoring the registers and to take advantage of
that ravb was not suspended to reduce resume time. To reset the
hardware the driver closes the device, sets it in reset mode and reopens
the device just like it would do in a normal suspend/resume scenario
without WoL enabled, but it both closes and opens the device in the
resume callback since the device needs to be reset for WoL to work.

One quirk needed for WoL is that the module clock needs to be prevented
from being switched off by Runtime PM. To keep the clock alive the
suspend callback need to call clk_enable() directly to increase the
usage count of the clock. Then when Runtime PM decreases the clock usage
count it won't reach 0 and be switched off.
Signed-off-by: NNiklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Acked-by: NSergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Make return value void since function never return meaningfull value
Signed-off-by: NYuval Shaia <yuval.shaia@oracle.com>
Acked-by: NSergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Commit a47b70ea ("ravb: unmap descriptors when freeing rings") has
introduced the issue seen in [1] reproduced on H3ULCB board.

Fix this by relocating the RX skb ringbuffer free operation, so that
swiotlb page unmapping can be done first. Freeing of aligned TX buffers
is not relevant to the issue seen in [1]. Still, reposition TX free
calls as well, to have all kfree() operations performed consistently
_after_ dma_unmap_*()/dma_free_*().

[1] Console screenshot with the problem reproduced:

salvator-x login: root
root@salvator-x:~# ifconfig eth0 up
Micrel KSZ9031 Gigabit PHY e6800000.ethernet-ffffffff:00: \
       attached PHY driver [Micrel KSZ9031 Gigabit PHY]   \
       (mii_bus:phy_addr=e6800000.ethernet-ffffffff:00, irq=235)
IPv6: ADDRCONF(NETDEV_UP): eth0: link is not ready
root@salvator-x:~#
root@salvator-x:~# ifconfig eth0 down

==================================================================
BUG: KASAN: use-after-free in swiotlb_tbl_unmap_single+0xc4/0x35c
Write of size 1538 at addr ffff8006d884f780 by task ifconfig/1649

CPU: 0 PID: 1649 Comm: ifconfig Not tainted 4.12.0-rc4-00004-g112eb072 #32
Hardware name: Renesas H3ULCB board based on r8a7795 (DT)
Call trace:
[<ffff20000808f11c>] dump_backtrace+0x0/0x3a4
[<ffff20000808f4d4>] show_stack+0x14/0x1c
[<ffff20000865970c>] dump_stack+0xf8/0x150
[<ffff20000831f8b0>] print_address_description+0x7c/0x330
[<ffff200008320010>] kasan_report+0x2e0/0x2f4
[<ffff20000831eac0>] check_memory_region+0x20/0x14c
[<ffff20000831f054>] memcpy+0x48/0x68
[<ffff20000869ed50>] swiotlb_tbl_unmap_single+0xc4/0x35c
[<ffff20000869fcf4>] unmap_single+0x90/0xa4
[<ffff20000869fd14>] swiotlb_unmap_page+0xc/0x14
[<ffff2000080a2974>] __swiotlb_unmap_page+0xcc/0xe4
[<ffff2000088acdb8>] ravb_ring_free+0x514/0x870
[<ffff2000088b25dc>] ravb_close+0x288/0x36c
[<ffff200008aaf8c4>] __dev_close_many+0x14c/0x174
[<ffff200008aaf9b4>] __dev_close+0xc8/0x144
[<ffff200008ac2100>] __dev_change_flags+0xd8/0x194
[<ffff200008ac221c>] dev_change_flags+0x60/0xb0
[<ffff200008ba2dec>] devinet_ioctl+0x484/0x9d4
[<ffff200008ba7b78>] inet_ioctl+0x190/0x194
[<ffff200008a78c44>] sock_do_ioctl+0x78/0xa8
[<ffff200008a7a128>] sock_ioctl+0x110/0x3c4
[<ffff200008365a70>] vfs_ioctl+0x90/0xa0
[<ffff200008365dbc>] do_vfs_ioctl+0x148/0xc38
[<ffff2000083668f0>] SyS_ioctl+0x44/0x74
[<ffff200008083770>] el0_svc_naked+0x24/0x28

The buggy address belongs to the page:
page:ffff7e001b6213c0 count:0 mapcount:0 mapping:          (null) index:0x0
flags: 0x4000000000000000()
raw: 4000000000000000 0000000000000000 0000000000000000 00000000ffffffff
raw: 0000000000000000 ffff7e001b6213e0 0000000000000000 0000000000000000
page dumped because: kasan: bad access detected

Memory state around the buggy address:
 ffff8006d884f680: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
 ffff8006d884f700: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
>ffff8006d884f780: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
                   ^
 ffff8006d884f800: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
 ffff8006d884f880: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
==================================================================
Disabling lock debugging due to kernel taint
root@salvator-x:~#

Fixes: a47b70ea ("ravb: unmap descriptors when freeing rings")
Signed-off-by: NEugeniu Rosca <erosca@de.adit-jv.com>
Acked-by: NSergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

If skb_put_padto() fails then it frees the skb.  I shifted that code
up a bit to make my error handling a little simpler.

Fixes: a0d2f206 ("Renesas Ethernet AVB PTP clock driver")
Signed-off-by: NDan Carpenter <dan.carpenter@oracle.com>
Acked-by: NSergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>