Report link up/down in standard ethtool way
Signed-off-by: NBrice Goglin <brice@myri.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

Remove the aligned-completion whitelist, and replace it by using the 1.4.16
firmware's auto-detection features to choose which firmware to load.
The driver now loads the aligned firmware, performs a MXGEFW_CMD_UNALIGNED_TEST,
and falls back to using the unaligned firmware if:
- The firmware is too old (ie, MXGEFW_CMD_UNALIGNED_TEST is an unknown command).
- The MXGEFW_CMD_UNALIGNED_TEST returns MXGEFW_CMD_ERROR_UNALIGNED, meaning
  that it has seen an unaligned completion during the DMA test.
Signed-off-by: NBrice Goglin <brice@myri.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

Move the DMA test code into its own function.
Signed-off-by: NBrice Goglin <brice@myri.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

Don't count on whatever implementation artifact preserves the
multicast list across a reset cmd, and setup multicast filtering
as part of our reset routine.

The setting of allmulti when adopting firmware with the rx-filter
broadcast bug is also moved into the multicast setup routine where
it belongs.
Signed-off-by: NBrice Goglin <brice@myri.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

Add dropped_pause, dropped_bad_phy, dropped_bad_crc32,
dropped_unicast_filtered to the set of ethtool counters.
Signed-off-by: NBrice Goglin <brice@myri.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

To clearly state the intent of copying to linear sk_buffs, _offset being a
overly long variant but interesting for the sake of saving some bytes.
Signed-off-by: NArnaldo Carvalho de Melo <acme@ghostprotocols.net>

The ip_hdrlen() buddy, created to reduce the number of skb->h.th-> uses and to
avoid the longer, open coded equivalent.

Ditched a no-op in bnx2 in the process.

I wonder if we should have a BUG_ON(skb->h.th->doff < 5) in tcp_optlen()...
Signed-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

For the quite common 'skb->h.raw - skb->data' sequence.
Signed-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

One less thing for drivers writers to worry about.
Signed-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Update the myri10ge driver version number to 1.3.0-1.233.
Signed-off-by: NBrice Goglin <brice@myri.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

Add the Intel 5000 southbridge (aka Intel 6310/6311/6321ESB) PCIe ports
and the Intel E30x0 chipsets to the whitelist of aligned PCIe completion.
Signed-off-by: NBrice Goglin <brice@myri.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

Simpler way of dealing with the firmware 4KB boundary crossing
restriction for rx buffers.  This fixes a variety of memory
corruption issues when using an "uncommon" MTU with a 16KB
page size.
Signed-off-by: NBrice Goglin <brice@myri.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

Correctly detect when TSO should be used on transmit by looking at the
skb->gso_size rather than seeing if the frame was larger than our MTU.
The old method causes problems when a host with a large (jumbo) MTU is
sending to a host with a small (standard) MTU.
Signed-off-by: NBrice Goglin <brice@myri.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

Driver version is now 1.3.0-1.226.
Signed-off-by: NBrice Goglin <brice@myri.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

Fix management of allocated physical pages when the architecture
page size is not 4kB since the firmware cannot cross 4K boundary.
Signed-off-by: NBrice Goglin <brice@myri.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

Update the default value of 2 module parameters:
* wcfifo disabled
* intr_coal_delay 75us
Signed-off-by: NBrice Goglin <brice@myri.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

[PATCH 1/4] myri10ge: Serverworks HT2100 provides aligned PCIe completion

Use the regular firmware on Serverworks HT2100 PCIe ports since this
chipset provides aligned PCIe completion.
Signed-off-by: NBrice Goglin <brice@myri.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

Add a wc_enabled flag in the myri10ge_priv instead of relying
on mtrr >= 0.
Signed-off-by: NBrice Goglin <brice@myri.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

Do not use 4k rdma request on SGI TIOCE chipset since this
bridge does not support it.
Signed-off-by: NBrice Goglin <brice@myri.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

Allocate a specific page and use pci_map_page for dma test instead
of relying on another existing buffer.
Signed-off-by: NBrice Goglin <brice@myri.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

Fix a missing error check in myri10ge_allocate_rings() and set status
to -ENOMEM before all actual allocations so that the error path returns
what it should.
Signed-off-by: NBrice Goglin <brice@myri.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

Fix copyright and license ("regents" should not have ever been used).
Signed-off-by: NBrice Goglin <brice@myri.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

Work around a bug which occurs when adopting firmware versions
1.4.4 though 1.4.11 where broadcasts are filtered as if they
were multicasts.
Signed-off-by: NBrice Goglin <brice@myri.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

Remove the NETIF_F_TSO #ifdef-ery in drivers/net; this was
for old-old-2.4 compat (even current 2.4 has NETIF_F_TSO)
but it's time to get rid of it by now.
Signed-off-by: NArjan van de Ven <arjan@linux.intel.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

Driver is now at version 1.2.0.
Signed-off-by: NBrice Goglin <brice@myri.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

Now that IRQ allocation is done in myri10ge_open(), we want to still
check when loading the driver that IRQ allocation could succeed later.

Additionaly, we fix the initialization and printing of netdev->irq.
Signed-off-by: NBrice Goglin <brice@myri.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

Under some circumstances, using WC without the WC fifo is faster.
So we make it possible to tune wc_fifo with a module parameter.
Signed-off-by: NBrice Goglin <brice@myri.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

On suspend, handle pci_set_power_state errors, and on resume
handle failures in pci_resume_state().
Signed-off-by: NBrice Goglin <brice@myri.com>
Signed-off-by: NStephen Hemminger <shemminger@osdl.org>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

The PCI MSI and express state are already saved and restored by the
current versions of pci_save_state/pci_restore_state.
Therefore it is no longer necessary for the driver to do it.
Signed-off-by: NBrice Goglin <brice@myri.com>
Signed-off-by: NStephen Hemminger <shemminger@osdl.org>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

Now that IRQ are requested is called on open() and freed on close(),
we can safely switch from/to MSI without unloading the module.

We are guaranteed to correctly free IRQ even if the sysfs file got
written in the meantime since the MSI initialization is stored in
mgp->msi_enabled.
Signed-off-by: NBrice Goglin <brice@myri.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

Request IRQ in myri10ge_open() and free in close() instead of probe()
and remove() to eliminate potential race between the watchdog and the
interrupt handler. Additionaly, the interrupt handler won't get called
on shared irq anymore when the interface is down.
Signed-off-by: NBrice Goglin <brice@myri.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

Since pci_save_state() pushes MSI and PCIe states on a kind of stack,
myri10ge saving the state in advance for parity recovery will push the
state again on the stack on suspend. This leads to some memory leak.
We add a couple additional calls to save_state and restore_state so
that we don't leak anymore.

For the future, we are thinking of a better way to recover from parity
error without using pci_save_state().
Signed-off-by: NBrice Goglin <brice@myri.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

Update driver version to 1.1.0.
Signed-off-by: NBrice Goglin <brice@myri.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

Fix sizing of big_bytes in the case of vlan frames. The 4
VLAN_HLEN bytes were omitted, leading to sizing the big buffer
4 bytes smaller than it should be.  Due to how rx buffers are
carved from pages, this was harmless for the common (9000, 1500)
byte MTUs, but could lead to data corruption for some MTUs.
Signed-off-by: NBrice Goglin <brice@myri.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

Receive full vlan frames into smalls when running with a jumbo MTU.
Signed-off-by: NBrice Goglin <brice@myri.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

Drop the old routines that used the physically contigous skb now
that we use the physical pages. And rename myri10ge_page_rx_done()
to myri10ge_rx_done() as it was previously.
Signed-off-by: NBrice Goglin <brice@myri.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

Switch to physical page skb, by calling the new page-based
allocation routines and using myri10ge_page_rx_done().
Signed-off-by: NBrice Goglin <brice@myri.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

Add physical page skb allocation routines and page based rx_done,
to be used by upcoming patches.
Signed-off-by: NBrice Goglin <brice@myri.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

Indentation cleanups to synchronize to our tree which is automatically
indent'ed.
Signed-off-by: NBrice Goglin <brice@myri.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>

In the myri10ge_submit_8rx() routine, write the 64 byte request block as
2 32-byte blocks so that it is handled by the hardware pio write handler
if write-combining is enabled.
Signed-off-by: NBrice Goglin <brice@myri.com>
Signed-off-by: NJeff Garzik <jeff@garzik.org>