Currently, we can only print the phy_status. Factor out the routine to
get the status, so we can query it from fec drivers and configure
accordingly. Needed because mx28 needs a special bit set for 10Mbit.
Signed-off-by: Wolfram Sang <w.sang@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Q: "the linux driver add these bits, why not we?"
A: Because nobody activated the bits?
Signed-off-by: Wolfram Sang <w.sang@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Introduce a variable which gets updated when needed and only written
once. Will make further additions easier.
Signed-off-by: Wolfram Sang <w.sang@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
remove double include, remove unused (and double in case of RCNTRL)
defines, sort the includes at least somewhat.
Signed-off-by: Wolfram Sang <w.sang@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Support for the old DM9000E device is now part of the new dm9k.c driver. So,
remove the old driver source and switch all users to the new driver.
Signed-off-by: Juergen Beisert <jbe@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
This patch adds support for the more recent DM9000A and DM9000B types, and keeps
support for the older DM9000E device. As this patch is more or less a complete
re-wrote of the existing driver I add a new source file instead of fixing the
existing one. In a later patch the old driver will be removed.
Signed-off-by: Juergen Beisert <jbe@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
We never had interrupt support in barebox and we have no plans to
add interrupt support. Even if we do I doubt the current fragments
of irq support are helpful, so remove them.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
To allow for some generic io accessors introduce io.h and use
this instead of asm/io.h throughout the tree.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
The driver uses a static int once variable to alloc the
rx packets. remove this to make the driver multi instance
safe. While at it, remove the crappy selfmade alignment.
dma_alloc_coherent returns sufficiently aligned memory.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
as barebox can be run as a second stage the first stage may have init it
before so re-used it
Signed-off-by: Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Without this patch (against next) I get this:
rsc@thebe:barebox$ make ARCH=arm CROSS_COMPILE=<path>/arm-v5te-linux-gnueabi- menuconfig
HOSTCC scripts/kconfig/lxdialog/checklist.o
HOSTCC scripts/kconfig/lxdialog/inputbox.o
HOSTCC scripts/kconfig/lxdialog/menubox.o
HOSTCC scripts/kconfig/lxdialog/textbox.o
HOSTCC scripts/kconfig/lxdialog/util.o
HOSTCC scripts/kconfig/lxdialog/yesno.o
HOSTCC scripts/kconfig/mconf.o
HOSTLD scripts/kconfig/mconf
scripts/kconfig/mconf Kconfig
warning: (ARCH_IMX25 && ARCH_IMX27 && ARCH_IMX35 && ARCH_IMX51 && ARCH_IMX28) selects ARCH_HAS_FEC_IMX which has unmet direct dependencies (ARCH_IMX)
warning: (ARCH_IMX25 && ARCH_IMX27 && ARCH_IMX35 && ARCH_IMX51 && ARCH_IMX28) selects ARCH_HAS_FEC_IMX which has unmet direct dependencies (ARCH_IMX)
Add the option for the FEC on MX28 as well.
Signed-off-by: Robert Schwebel <r.schwebel@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
We never supported the sigmatel stm chips and probably
never will. So do the first step and rename the architecture
to mxs just like in the kernel.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Ethernet PHY devices have 32 registers, each 16 bits wide. Thus, the size of
the phy device should be 64 bytes.
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
The idea is to panic() when there is no memory available for normal
operation. Exception: code which can consume arbitrary amount of RAM
(example: files allocated in ramfs) must report error instead of panic().
This patch also fixes code which didn't check for NULL from malloc() etc.
Usage: malloc(), memalign() return NULL when out of RAM.
xmalloc(), xmemalign() always return non-NULL or panic().
Signed-off-by: Krzysztof Hałasa <khc@pm.waw.pl>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
The i.MX28 comes with an extended ethernet controller (ENET/FEC) which
is backwards compatible to the FEC known from other i.MX CPUs. Add a few
adaptions to the existing driver to make it work with the MX28 FEC.
Signed-off-by: Juergen Beisert <jbe@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Some register bits are not defined in the IMX CPUs using this FEC (at least
I did not find anything about them in the related datasheets) and they will
get (partially) a new meaning in the i.MX28 CPU. So, removing them from the
source seems the best way to avoid any accidental usage on i.MX28.
Signed-off-by: Juergen Beisert <jbe@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
now we need to set the dev id to -1
Signed-off-by: Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Patch fix all similar warnings found by sparse:
drivers/net/smc91111.c:488:9:
warning: incorrect type in argument 1 (different base types)
expected void const volatile [noderef] <asn:2>*<noident>
got unsigned long [unsigned] offset
Signed-off-by: Marek Belisko <marek.belisko@open-nandra.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
this rework is done in order to add a phylib and allow to have phy driver support
Signed-off-by: Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
The mac address information is taken from the imx_iim driver if it's present.
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
this implementation is based on linux one (v2.6.35-rc5-76-gd0c6f62)
it will calculate all the clock dynamically instead of statictly
this will use also the new clock framework
it will also print the clock status after the console init
Signed-off-by: Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
if fec was allocated using malloc. if fec->miiphy->dev->driver != 0
miiphy_probe is not called
Signed-off-by: Jan Weitzel <J.Weitzel@phytec.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
ep93xx eth driver: Remove unnecessary parentheses in definition of the constants
MII_ADDRESS_MAX and MII_REGISTER_MAX
Signed-off-by: Matthias Kaehlcke <matthias@kaehlcke.net>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
ep93xx eth driver: Define ep93xx_eth_send_packet() and ep93xx_eth_rcv_packet()
before ep93xx_eth_probe(), and eliminate their prototype declarations
Signed-off-by: Matthias Kaehlcke <matthias@kaehlcke.net>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
On my little endian PXA270, the ethernet address is byte swapped:
correct ethernet address: 00:50:c2:80:a7:bd
broken ethernet address: 50:00:80:c2:bd:a7
The correct value is what the sticker on the baoard and the linux driver
says. This patch fixes the problem by reading the ethaddr byte-wise from
the eeprom.
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
edev was allocated using malloc. On one board (NOR boot), we met data abort
in eth_register (at dev_add_param(dev, &edev->param_ip);).
Allocating edev with xzalloc (as in several other drivers) fixed the problem.
Signed-off-by: Eric Benard <eric@eukrea.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
We need to redo buffer initialization each time we open
the device, not only once on init time. Otherwise network
does not work if the last transfer was interrupted with
ctrl-c
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
If a NIC's MII isn't connected to a PHY but directly to a switch port,
the switch doesn't sent a link signal to the NIC. (Because strictly
speaking, there isn't any ethernet link at all.)
This patch adds a force link feature to the phy, to cope with this
situation.
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
When operating in plain MII mode, the driver assumes the RMII bit not
set. Although this is the power on reset value, clearing this bit helps
when playing with different network setups.
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
This driver is not only used on i.MX27 but on i.MX35 and i.MX25, so
rename it to a more generic name
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
The driver was not using the phy address supplied by the platform
data. With this patch the miiphy.address is set phy addr supplied
by the mac platform data.
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
The FEC driver used to typedef it's own variable types. Use standard
U-Boot types instead
Signed-off-by: Carsten Schlote <schlote@vahanus.net>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
fixes:
drivers/net/smc911x.c:334:1: warning: "FLOW" redefined
In file included from include/asm/blackfin.h:33,
from include/asm/system.h:29,
from include/asm/bitops.h:33,
from include/linux/bitops.h:69,
from include/common.h:31,
from drivers/net/smc911x.c:29:
include/asm/cpu/defBF561.h:1654:1: warning: this is the location of the previous definition
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
do autongotiation on probe and not on open, so this is done
when we actually need the phy.
We do not disable the phy now, that's not good. But at least
networking is much more responsive now.
Currently U-Boot uses globally defined eth_* functions. This is
horribly unflexible. This patch replaces the global functions
with pointers from structs. We could also use CONFIG_NET_MULTI,
but this has other implications, though we should merge this
some day.
Also, U-Boot has no unique way to handle MAC addresses. Each and
every board and network driver uses it's own mechanism to set the
MAC address. There are several problems which I've for too often.
For example everything goes well if we boot from network, but when
we boot from flash U-Boot forgets to set the MAC address and the
linux network driver has none.
This patch adds [gs]et_mac_address to the eth_device struct and
handles it as follows:
- First try to get a valid MAC address from the EEPROM and set
'ethaddr' accordingly.
- If no valid MAC address is found in the EEPROM (or no EEPROM is
connected), we set the devices MAC address from 'ethaddr'
This is done in eth_initialize which is called on startup for
every board.
Sascha Hauer
Wolfram Sang
Juergen Beisert
Hubert Feurstein
Jean-Christophe PLAGNIOL-VILLARD
Franck Jullien
Robert Schwebel
Baruch Siach
Krzysztof Halasa
Marek Belisko
Jan Weitzel
Matthias Kaehlcke
Marc Kleine-Budde
Eric Benard
Sascha Hauer
Juergen Beisert
Ivo Clarysse
Enrik Berkhan
Juergen Beisert
sascha
Sascha Hauer