Cc: davem@davemloft.net
Cc: netdev@vger.kernel.org
Acked-by: NGreg Kroah-Hartman <gregkh@suse.de>
Signed-off-by: NKay Sievers <kay.sievers@vrfy.org>

When mv643xx_eth_open() is called to up an interface, port_start()
will first re-program the unicast address filter, and then
re-initialise the PORT_CONFIG register, but that will disable unicast
promiscuous mode if it was enabled by the unicast address filter setup.

This isn't a problem on ifconfig up, as ->set_rx_mode() will be called
shortly afterwards which will program the filters again, but it does
trigger when changing the MTU, which calls mv643xx_eth_stop() and then
mv643xx_eth_open() by hand to repopulate the receive rings with skbuffs
of the new size.

Swap the initialisation of the PORT_START register and the call to
the unicast filter setup function to fix this.
Signed-off-by: NLennert Buytenhek <buytenh@marvell.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

mib_counters_update() also restarts the timer.
So the timer is dequeued, the stats are read and then the timer is
enqueued again. This is "okay" unless someone unloads the module.
The locking here is also broken:
mib_counters_update() grabs just a simple spinlock. The only thing the
lock is good for is to protect the timer func against other callers
namely mv643xx_eth_stop() && mv643xx_eth_get_ethtool_stats(). That means
if the spinlock is taken via the ethtool path and than the timer kicks
in then the box will lock up.
Signed-off-by: NSebastian Andrzej Siewior <sebastian@breakpoint.cc>
Acked-by: NLennert Buytenhek <buytenh@marvell.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

dev_set_rx_mode() grabs netif_addr_lock_bh():

|BUG: sleeping function called from invalid context at /home/bigeasy/git/cryptodev-2.6/mm/slub.c:1599
|in_atomic(): 1, irqs_disabled(): 0, pid: 859, name: ifconfig
|2 locks held by ifconfig/859:
| #0:  (rtnl_mutex){--..}, at: [<c0239ccc>] rtnl_lock+0x18/0x20
| #1:  (_xmit_ETHER){-...}, at: [<c022d094>] dev_set_rx_mode+0x1c/0x30
|[<c029f118>] (dump_stack+0x0/0x14) from [<c003df28>] (__might_sleep+0x11c/0x13c)
|[<c003de0c>] (__might_sleep+0x0/0x13c) from [<c00a8854>] (kmem_cache_alloc+0x30/0xd4)
| r5:c78093a0 r4:c034a47c
|[<c00a8824>] (kmem_cache_alloc+0x0/0xd4) from [<c01a5fd0>] (mv643xx_eth_set_rx_mode+0x70/0x188)
|[<c01a5f60>] (mv643xx_eth_set_rx_mode+0x0/0x188) from [<c022ced0>] (__dev_set_rx_mode+0x40/0xac)
|[<c022ce90>] (__dev_set_rx_mode+0x0/0xac) from [<c022d09c>] (dev_set_rx_mode+0x24/0x30)
| r6:00001043 r5:c78090f8 r4:c7809000
|[<c022d078>] (dev_set_rx_mode+0x0/0x30) from [<c02304c4>] (dev_open+0xe4/0x114)
| r5:c7809350 r4:c7809000
|[<c02303e0>] (dev_open+0x0/0x114) from [<c022fd18>] (dev_change_flags+0xb0/0x190)
| r5:00000041 r4:c7809000
|[<c022fc68>] (dev_change_flags+0x0/0x190) from [<c0270250>] (devinet_ioctl+0x2f0/0x710)
| r7:c7221e70 r6:c7aadb00 r5:00000000 r4:00000001
|[<c026ff60>] (devinet_ioctl+0x0/0x710) from [<c02717c8>] (inet_ioctl+0xd4/0x110)
|[<c02716f4>] (inet_ioctl+0x0/0x110) from [<c021fb74>] (sock_ioctl+0x1f4/0x254)
| r4:c7242b40
|[<c021f980>] (sock_ioctl+0x0/0x254) from [<c00b8160>] (vfs_ioctl+0x38/0x98)
| r6:beec9bb8 r5:00008914 r4:c7242b40
|[<c00b8128>] (vfs_ioctl+0x0/0x98) from [<c00b873c>] (do_vfs_ioctl+0x484/0x4d4)
| r6:00008914 r5:c7242b40 r4:c74db1c0
|[<c00b82b8>] (do_vfs_ioctl+0x0/0x4d4) from [<c00b87cc>] (sys_ioctl+0x40/0x64)
|[<c00b878c>] (sys_ioctl+0x0/0x64) from [<c00269a0>] (ret_fast_syscall+0x0/0x2c)
|[42949399.520000]  r7:00000036 r6:beec9c80 r5:00000041 r4:beec9bb8
Signed-off-by: NSebastian Andrzej Siewior <sebastian@breakpoint.cc>
Acked-by: NLennert Buytenhek <buytenh@marvell.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Contrary to what the docs say, the 'extended interrupt cause' bit in
the interrupt cause register (bit 1) appears to not be maskable on at
least some of the mv643xx_eth platforms, making writing zeroes to the
interrupt mask register but not the extended interrupt mask register
insufficient to stop interrupts from occuring.  Therefore, also write
zeroes to the extended interrupt mask register when shutting down the
port.

This fixes the interrupt storm seen on the Pegasos board when shutting
down the interface.
Signed-off-by: NLennert Buytenhek <buytenh@marvell.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Commit 66e63ffb ("mv643xx_eth:
implement ->set_rx_mode()") cleaned up mv643xx_eth's multicast filter
programming, but broke it as well.

The non-special multicast filter table (for multicast addresses that
are not of the form 01:00:5e:00:00:xx) consists of 256 hash table
buckets organised as 64 32-bit words, where the 'accept' bits are
in the LSB of each byte, so in bits 24 16 8 0 of each 32-bit word.
The old code got this right, but the referenced commit broke this by
using bits 3 2 1 0 instead.  This commit fixes this up.
Signed-off-by: NLennert Buytenhek <buytenh@marvell.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This reverts commit cd4ccf76.

On the Pegasos board, we can't do DMA burst that are longer than
one cache line.  For now, go back to using 32 byte DMA bursts for
all mv643xx_eth platforms -- we can switch the ARM-based platforms
back to doing long 128 byte bursts in the next development cycle.
Signed-off-by: NLennert Buytenhek <buytenh@marvell.com>
Reported-by: NAlan Curry <pacman@kosh.dhis.org>
Reported-by: NGabriel Paubert <paubert@iram.es>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NLennert Buytenhek <buytenh@marvell.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NLennert Buytenhek <buytenh@marvell.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Currently, if multiple unicast addresses are programmed into a
mv643xx_eth interface, the core networking will resort to enabling
promiscuous mode on the interface, as mv643xx_eth does not implement
->set_rx_mode().

This patch switches mv643xx_eth over from ->set_multicast_list()
to ->set_rx_mode(), and implements support for secondary unicast
addresses.  The hardware can handle multiple unicast addresses as
long as their first 11 nibbles are the same (i.e. are of the form
xx:xx:xx:xx:xx:xy where the x part is the same for all addresses), so
if that is the case, we use that mode.  If it's not the case, we enable
unicast promiscuous mode in the hardware, which is slightly better than
enabling promiscuous mode for multicasts as well, which is what would
happen before.

While we are at it, change the programming sequence so that we
don't clear all filter bits first, so we don't lose all incoming
packets while the filter is being reprogrammed.
Signed-off-by: NLennert Buytenhek <buytenh@marvell.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Since txq_alloc_desc_index() is a very simple function, and since
descriptor ring index handling for transmit reclaim, receive
processing and receive refill is already handled inline as well,
inline txq_alloc_desc_index() into its two call sites.
Signed-off-by: NLennert Buytenhek <buytenh@marvell.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The mv643xx_eth driver uses the rdl()/wrl() macros to read and
write hardware registers.  Per-port registers are accessed in the
following way:

	#define PORT_STATUS(p)			(0x0444 + ((p) << 10))

	[...]

	static inline u32 rdl(struct mv643xx_eth_private *mp, int offset)
	{
		return readl(mp->shared->base + offset);
	}

	[...]

	port_status = rdl(mp, PORT_STATUS(mp->port_num));

By giving the per-port 'struct mv643xx_eth_private' its own
'void __iomem *base' pointer that points to the per-port register
area, we can get rid of both the double indirection and the << 10
that is done for every per-port register access -- this patch does
that.
Signed-off-by: NLennert Buytenhek <buytenh@marvell.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Fix up a couple of coding style issues caught by checkpatch.
Signed-off-by: NLennert Buytenhek <buytenh@marvell.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

When mv643xx_eth allocates skbuffs, it adds
'dma_get_cache_alignment() - 1' to the length it needs, so that it can
align the skb's ->data pointer to a cache boundary.  When checking
whether a transmitted skbuff can be reused as a receive buffer, these
bytes needs to be included into the minimum bound for the recycle check.
Signed-off-by: NLennert Buytenhek <buytenh@marvell.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NLennert Buytenhek <buytenh@marvell.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The generic packet receive code takes care of setting
netdev->last_rx when necessary, for the sake of the
bonding ARP monitor.

Drivers need not do it any more.

Some cases had to be skipped over because the drivers
were making use of the ->last_rx value themselves.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The mv643xx_eth mii bus implementation uses wait_event_timeout() to
wait for SMI completion interrupts.

If wait_event_timeout() would return zero, mv643xx_eth would conclude
that the SMI access timed out, but this is not necessarily true --
wait_event_timeout() can also return zero in the case where the SMI
completion interrupt did happen in time but where it took longer than
the requested timeout for the process performing the SMI access to be
scheduled again.  This would lead to occasional SMI access timeouts
when the system would be under heavy load.

The fix is to ignore the return value of wait_event_timeout(), and
to re-check the SMI done bit after wait_event_timeout() returns to
determine whether or not the SMI access timed out.
Signed-off-by: NLennert Buytenhek <buytenh@marvell.com>
Signed-off-by: NJeff Garzik <jgarzik@redhat.com>

This converts pretty much everything to print_mac. There were
a few things that had conflicts which I have just dropped for
now, no harm done.

I've built an allyesconfig with this and looked at the files
that weren't built very carefully, but it's a huge patch.
Signed-off-by: NJohannes Berg <johannes@sipsolutions.net>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

mv643xx_eth uses ip_hdr() (defined in linux/ip.h), but relied on
another header file to include the needed header file indirectly.
In latest net-next this indirect include chain is gone, so the
driver fails to build.  Include linux/ip.h explicitly to fix this.
Signed-off-by: NLennert Buytenhek <buytenh@marvell.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch introduces mdiobus_alloc() and mdiobus_free(), and
makes all mdio bus drivers use these functions to allocate their
struct mii_bus'es dynamically.
Signed-off-by: NLennert Buytenhek <buytenh@marvell.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>
Acked-by: NAndy Fleming <afleming@freescale.com>

In preparation of giving mii_bus objects a device tree presence of
their own, rename struct mii_bus's ->dev argument to ->parent, since
having a 'struct device *dev' that points to our parent device
conflicts with introducing a 'struct device dev' representing our own
device.
Signed-off-by: NLennert Buytenhek <buytenh@marvell.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>
Acked-by: NAndy Fleming <afleming@freescale.com>

This gives a nice increase in the maximum loss-free packet forwarding
rate in routing workloads.
Signed-off-by: NLennert Buytenhek <buytenh@marvell.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NLennert Buytenhek <buytenh@marvell.com>

Switch mv643xx_eth from using drivers/net/mii.c to using phylib.

Since the mv643xx_eth hardware does all the link state handling and
PHY polling, the driver will use phylib in the "Doing it all yourself"
mode described in the phylib documentation.
Signed-off-by: NLennert Buytenhek <buytenh@marvell.com>
Acked-by: NAndy Fleming <afleming@freescale.com>

If we don't poll the hardware statistics counters at least once every
~34 seconds, overflow might occur without us noticing.  So, set up a
timer to poll the statistics counters at least once every 30 seconds.
Signed-off-by: NLennert Buytenhek <buytenh@marvell.com>

When the IP header doesn't start 14, 18, 22 or 26 bytes into the packet
(which are the only four cases that the hardware can deal with if asked
to do IP checksumming on transmit), invoke the software checksum helper
instead of letting the packet go out with a corrupt checksum inserted
into the packet in the wrong place.
Signed-off-by: NLennert Buytenhek <buytenh@marvell.com>

We have to explicitly tell the hardware to include the pseudo-header
when doing receive checksumming, otherwise hardware checksumming will
fail for every received packet and we'll end up setting CHECKSUM_NONE
on every received packet.

While we're at it, when skb->ip_summed is set to CHECKSUM_UNNECESSARY
on received packets, skb->csum is supposed to be undefined, and thus
there is no need to set it.
Signed-off-by: NLennert Buytenhek <buytenh@marvell.com>

Add support for mv643xx_eth versions that have no transmit bandwidth
control registers at all, such as the ethernet block found in the
Marvell 88F6183 ARM SoC.
Signed-off-by: NLennert Buytenhek <buytenh@marvell.com>

Currently, the receive processing reads ->byte_cnt twice (once to
update interface statistics and once to properly size the data area
of the received skb), but since receive descriptors live in uncached
memory, caching this value in a local variable saves one uncached
access, and increases routing performance a tiny little bit more.
Signed-off-by: NLennert Buytenhek <buytenh@marvell.com>

Since the size of the receive queue is directly related to the data
cache footprint of the driver (between refilling a receive ring entry
with a fresh skb and receiving a packet in that entry, queue_size - 1
other skbs will have been touched), shrink the default receive queue
size to a saner number of entries, as 400 is definite overkill for
almost all workloads.

While we are at it, trim the default transmit queue size a bit as well.
Signed-off-by: NLennert Buytenhek <buytenh@marvell.com>

Get rid of the skb pointer array that we currently use for transmit
reclaim, and replace it with an skb queue, to which skbuffs are appended
when they are passed to the xmit function, and removed from the front
and freed when we do transmit queue reclaim and hit a descriptor with
the 'owned by device' bit clear and 'last descriptor' bit set.
Signed-off-by: NLennert Buytenhek <buytenh@marvell.com>

By moving DMA unmapping during transmit reclaim back under the netif
tx lock, we avoid the situation where we read the DMA address and buffer
length from the descriptor under the lock and then not do anything with
that data after dropping the lock on platforms where the DMA unmapping
routines are all NOPs (which is the case on all ARM platforms that
mv643xx_eth is used on at least).

This saves two uncached reads, which makes a small but measurable
performance difference in routing benchmarks.
Signed-off-by: NLennert Buytenhek <buytenh@marvell.com>

Since our ->hard_start_xmit() method is already called under spinlock
protection (the netif tx queue lock), we can simply make that lock
cover the private transmit state (descriptor ring indexes et al.) as
well, which avoids having to use a private lock to protect that state.

Since this was the last user of the driver-private spinlock, it can
be killed off.
Signed-off-by: NLennert Buytenhek <buytenh@marvell.com>

Move link status handling, transmit reclaim and TX_END handling from
the interrupt handler to the napi poll handler.  This allows switching
->lock over to a non-IRQ-safe lock and removes all explicit interrupt
disabling from the driver.
Signed-off-by: NLennert Buytenhek <buytenh@marvell.com>

As all the infrastructure for multiple transmit queues already exists
in the driver, this patch is entirely trivial.

The individual transmit queues are still serialised by the driver's
per-port private spinlock, but that will disappear (i.e. be replaced
by the per-subqueue ->_xmit_lock) in a subsequent patch.
Signed-off-by: NLennert Buytenhek <buytenh@marvell.com>

Delete a couple of unused and uninteresting interrupt source mask bits:
- The receive resource underrun interrupt sources are uninteresting
  because if we are in out-of-memory mode, we are already dealing with
  the issue, and we don't need the hardware to remind us again that we
  are out of memory.
- The LINK and PHY interrupt sources can be coalesced into one define,
  since we always use them together.
- The transmit resource underrun interrupt source can be disabled since
  we never activate the head descriptor of a paged skb until the
  fragments are all activated, so transmit underrun during a packet
  should never happen.
- The INT_EXT_TX_0 define is never used.
Signed-off-by: NLennert Buytenhek <buytenh@marvell.com>

There is no need to call netif_{stop,wake}_queue() when the link goes
down/up, as the networking already takes care of this internally.
Signed-off-by: NLennert Buytenhek <buytenh@marvell.com>

Currently, there are two different fields in the
mv643xx_eth_platform_data struct that together describe the PHY
address -- one field (phy_addr) has the address of the PHY, but if
that address is zero, a second field (force_phy_addr) needs to be
set to distinguish the actual address zero from a zero due to not
having filled in the PHY address explicitly (which should mean
'use the default PHY address').

If we are a bit smarter about the encoding of the phy_addr field,
we can avoid the need for a second field -- this patch does that.
Signed-off-by: NLennert Buytenhek <buytenh@marvell.com>

Which top-level unit's SMI interface to use should be a property of
the top-level unit, not of the individual ports.  This patch moves the
->shared_smi pointer from the per-port platform data to the global
platform data.
Signed-off-by: NLennert Buytenhek <buytenh@marvell.com>

Simplify receive and transmit queue handling by requiring the set
of queue numbers to be contiguous starting from zero.
Signed-off-by: NLennert Buytenhek <buytenh@marvell.com>