The tsec driver had a bunch of PHY drivers already written. This
converts them all into PHY Lib drivers, and serves as the first
set of PHY drivers for PHY Lib.
While doing that, cleaned up a number of magic numbers (though
not all of them, as PHY vendors like to keep their numbers as
magical as possible). Also, noticed that almost all of the
vitesse/cicada PHYs had the same config/parse/startup functions,
so those have been collapsed into one.
Signed-off-by: Andy Fleming <afleming@freescale.com>
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Acked-by: Detlev Zundel <dzu@denx.de>
Extends the mii_dev structure to participate in a full-blown MDIO and
PHY driver scheme. The mii_dev structure and miiphy calls are modified
in such a way to allow the original mii command and miiphy
infrastructure to work as before, but also to support a new set of APIs
which allow (among other things) sharing of PHY driver code and 10G support
The mii command will continue to support normal PHY management functions
(Clause 22 of 802.3), but will not be changed to support 10G
(Clause 45).
The basic design is similar to PHY Lib from Linux, but simplified for
U-Boot's network and driver infrastructure.
We now have MDIO drivers and PHY drivers
An MDIO driver provides:
read
write
reset
A PHY driver provides:
(optionally): probe
config - initial setup, starting of auto-negotiation
startup - waiting for AN, and reading link state
shutdown - any cleanup needed
The ethernet drivers interact with the PHY Lib using these functions:
phy_connect()
phy_config()
phy_startup()
phy_shutdown()
Each PHY driver can be configured separately, or all at once using
config_phylib_all_drivers.h (added in the patch which adds the drivers)
We also provide generic drivers for Clause 22 (10/100/1000), and
Clause 45 (10G) PHYs.
We also implement phy_reset(), and call it in phy_connect(). Because
phy_reset() is essentially the same as miiphy_reset, but:
a) must support 10G PHYs, and
b) should use the phylib primitives,
we implement miiphy_reset, using phy_reset(), but only when
CONFIG_PHYLIB is set. Otherwise, we just use the old version. In this
way, we save on compile size, even if we don't manage to save code size.
Pulled ethtool.h and mdio.h from:
git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next-2.6
782d640afd15af7a1faf01cfe566ca4ac511319d
With many, many deletions so as to enable compilation under u-boot
Signed-off-by: Andy Fleming <afleming@freescale.com>
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Acked-by: Detlev Zundel <dzu@denx.de>
The include/miiphy.h header duplicates a lot of things from linux/mii.h.
So punt all the things that overlap to keep the API simple and to make
merging between U-Boot and Linux simpler.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Before this commit, weak symbols were not overridden by non-weak symbols
found in archive libraries when linking with recent versions of
binutils. As stated in the System V ABI, "the link editor does not
extract archive members to resolve undefined weak symbols".
This commit changes all Makefiles to use partial linking (ld -r) instead
of creating library archives, which forces all symbols to participate in
linking, allowing non-weak symbols to override weak symbols as intended.
This approach is also used by Linux, from which the gmake function
cmd_link_o_target (defined in config.mk and used in all Makefiles) is
inspired.
The name of each former library archive is preserved except for
extensions which change from ".a" to ".o". This commit updates
references accordingly where needed, in particular in some linker
scripts.
This commit reveals board configurations that exclude some features but
include source files that depend these disabled features in the build,
resulting in undefined symbols. Known such cases include:
- disabling CMD_NET but not CMD_NFS;
- enabling CONFIG_OF_LIBFDT but not CONFIG_QE.
Signed-off-by: Sebastien Carlier <sebastien.carlier@gmail.com>
By now, the majority of architectures have working relocation
support, so the few remaining architectures have become exceptions.
To make this more obvious, we make working relocation now the default
case, and flag the remaining cases with CONFIG_NEEDS_MANUAL_RELOC.
Signed-off-by: Wolfgang Denk <wd@denx.de>
Tested-by: Heiko Schocher <hs@denx.de>
Tested-by: Reinhard Meyer <u-boot@emk-elektronik.de>
Commit 6e37b1a3a25004d3df5867de49fff6b3fc9c4f04 modifies several net calls
to take a (const char *) parameter instead of (char *), but in some cases
the modified functions call other functions taking (char *). The end result
is warnings about discarding the const qualifier.
This patch fixes these other function signatures.
Signed-off-by: Ben Warren <biggerbadderben@gmail.com>
Add missing newline.
Signed-off-by: Wolfgang Denk <wd@denx.de>
Cc: Luigi Mantellini <luigi.mantellini@idf-hit.com>
Cc: Ben Warren <biggerbadderben@gmail.com>
This feature is useful when your board uses different mii buses for different
phys and all (or a part) of these buses are implemented via bit-banging mode.
The driver requires that the following macros should be defined into the board
configuration file:
CONFIG_BITBANGMII - Enable the miiphybb driver
CONFIG_BITBANGMII_MULTI - Enable the multi bus support
If the CONFIG_BITBANGMII_MULTI is not defined, the board's config file needs
to define at least the following macros:
MII_INIT - Generic code to enable the MII bus (optional)
MDIO_DECLARE - Declaration needed to access to the MDIO pin (optional)
MDIO_ACTIVE - Activate the MDIO pin as out pin
MDIO_TRISTATE - Activate the MDIO pin as input/tristate pin
MDIO_READ - Read the MDIO pin
MDIO(v) - Write v on the MDIO pin
MDC_DECLARE - Declaration needed to access to the MDC pin (optional)
MDC(v) - Write v on the MDC pin
The previous macros make the driver compatible with the previous version
(that didn't support the multi-bus).
When the CONFIG_BITBANGMII_MULTI is also defined, the board code needs to fill
the bb_miiphy_buses[] array with a record for each required bus and declare
the bb_miiphy_buses_num variable with the number of mii buses.
The record (struct bb_miiphy_bus) has the following fields/callbacks (see
miiphy.h for details):
char name[] - The symbolic name that must be equal to the MII bus
registered name
int (*init)() - Initialization function called at startup time (just
before the Ethernet initialization)
int (*mdio_active)() - Activate the MDIO pin as output
int (*mdio_tristate)() - Activate the MDIO pin as input/tristate pin
int (*set_mdio)() - Write the MDIO pin
int (*get_mdio)() - Read the MDIO pin
int (*set_mdc)() - Write the MDC pin
int (*delay)() - Delay function
void *priv - Private data used by board specific code
The board code will look like:
struct bb_miiphy_bus bb_miiphy_buses[] = {
{ .name = miibus#1, .init = b1_init, .mdio_active = b1_mdio_active, ... },
{ .name = miibus#2, .init = b2_init, .mdio_active = b2_mdio_active, ... },
...
int bb_miiphy_buses_num = sizeof(bb_miiphy_buses) /
sizeof(bb_miiphy_buses[0]);
Signed-off-by: Luigi 'Comio' Mantellini <luigi.mantellini@idf-hit.com>
Signed-off-by: Ben Warren <biggerbadderben@gmail.com>
following build warning was observed
mv88e61xx.c: In function ‘mv88e61xx_busychk’:
mv88e61xx.c:208: warning: dereferencing type-punned pointer will break strict-aliasing rules
This patch fixes the same
Patch tested for rd6281a board build
Signed-off-by: Prafulla Wadaskar <prafulla@marvell.com>
Signed-off-by: Ben Warren <biggerbadderben@gmail.com>
1. mv88E61xx driver compiler warnings fixed
2. idstr if-else statements changed to switch() construct
and added default case too.
This fixed idstr may be uninitialized warning
Signed-off-by: Prafulla Wadaskar <prafulla@marvell.com>
Signed-off-by: Ben Warren <biggerbadderben@gmail.com>
With these fixes, this driver works properly for multi chip
addressging mode
Bugfixes:
1. Build error fixed for function mv88e61xx_busychk_multic-fixed
2. PHY dev address error detection- fixed
3. wrong busy bit was refered in function mv88e61xx_busychk -fixed
4. invalid data read ptr was refered for RD_PHY in case of
multichip addressing mode -fixed
The Multichip Address mode is tested with RD6281A board having
MV88E6165 switch on it
Signed-off-by: Prafulla Wadaskar <prafulla@marvell.com>
Signed-off-by: Ben Warren <biggerbadderben@gmail.com>
This patch sets the returned value to 0xFFFF if the PHY does not exist
and does not assert Transfer Acknowledge. A NULL check for the value
pointer is also added for buffer overflow protection.
Without this patch 'mii info' will show 'phantom' devices because the
value will be not be initialized and return with some random value.
Signed-off-by: Richard Retanubun <RichardRetanubun@RuggedCom.com>
Signed-off-by: Ben Warren <biggerbadderben@gmail.com>
Wolfgang Denk
Andy Fleming
Mike Frysinger
Sebastien Carlier
Ben Warren
Luigi 'Comio' Mantellini
Prafulla Wadaskar
Richard Retanubun
Jean-Christophe PLAGNIOL-VILLARD