Only install headers which are actually used by our userland. This
saves about 2.2MB under /usr/include/dev/. Discussed on tech-kern@
recently.

I HOPE to get the list right. The headers I left in are ones
used for MI tools and those whose usage I discovered by grep over tree sources.
Feel free to put needed includes back in if you encounter anything which
should not be removed from lists.

branches: 1.7.6;
Add a driver for the National Semiconductor DP83843 `PHYTER' 10/100
Ethernet PHY.

branches: 1.6.2; 1.6.8;
Spacing.

Install *reg.h

TDK Semiconductor 78Q2120 PHY, commonly used in low-power Cardbus cards.

Add PHY driver the the Intel 82553 found on some fxp's.
Some variants of this chip need a little extra MII fix before
being usable, so this is not yet put into GENERIC's.

Adapt to MII layer changes.

Rework the way kernel include files are installed. In the new method,
as with user-land programs, include files are installed by each directory
in the tree that has includes to install. (This allows more flexibility
as to what gets installed, makes 'partial installs' easier, and gives us
more options as to which machines' includes get installed at any given
time.) The old SYS_INCLUDES={symlinks,copies} behaviours are _both_
still supported, though at least one bug in the 'symlinks' case is
fixed by this change. Include files can't be build before installation,
so directories that have includes as targets (e.g. dev/pci) have to move
those targets into a different Makefile.

Rework the xxxVERBOSE option to share the common module-hook-based
verbose mechanism with MIIVERBOSE. This reduces some duplicated code
and allows us to once again permit auto-unload of MIIVERBOSE.

Change details:
* Update dev/devlist2h.awk to accomodate miidevs, including generation
of MII_STR_oui_model definitions and use of oui and model rather than
vendor and product. This also changes the compressed data in the
xxxdevs_data.h files to uint32_t (since mii oui's are up to 6 hex
digits long)
* Update a couple of phy drivers to use new calls to get verbose data
* Regen all of the xxxdevs{,_data}.h files (separate commit, coming
very soon)
* Update mii/mii_verbose.[ch] and mii/mii_physubr.c to use the various
DEV_VERBOSE_xxx macros
* Update the pci, usb, and hdaudio code as needed, to #include the
xxxdevs.h files (in order to get the proper printf format strings)
* Since dev/dev_verbose.c now uses non-literal printf format strings,
(to deal with the vendor/product vs oui/model issue), we need to
make sure it gets compiled with -Wno-error=format-nonliteral, even
in userland's libpci and librumpdev!
* Bump kernel version for the change in module interfaces

Welcome to 9.99.86!

XXX It might be useful in the future to extend the MII_STR_oui_model
XXX definitions to PCI as well (and perhaps USB and HDAUDIO). This
XXX would allow for a single centralized location for the products'
XXX descriptions, rather than being dispersed among individual
XXX drivers' xxx_match tables.

branches: 1.3.100;
Use ${TOOL_AWK} instead of ${AWK} or plain "awk" in make commands.
Pass AWK=${TOOL_AWK:Q} to shell scripts that use awk.

branches: 1.2.126; 1.2.130; 1.2.136;
Allow a verbose description to be printed with the option MIIVERBOSE.

branches: 1.1.12;
Near complete rewrite of the MII layer to present an interface compatible
with BSDI's MII layer to device drivers.

Define and implement a locking protocol for the ifmedia / mii layers:
- MP-safe drivers provide a mutex to ifmedia that is used to serialize
access to media-related structures / hardware regsiters. Converted
drivers use the new ifmedia_init_with_lock() function for this. The
new name is provided to ease the transition.
- Un-converted drivers continue to call ifmedia_init(), which will supply
a compatibility lock to be used instead. Several media-related entry
points must be aware of this compatibility lock, and are able to acquire
it recursively a limited number of times, if needed. This is a SPIN
mutex with priority IPL_NET.
- This same lock is used to serialize access to PHY registers and other
MII-related data structures.

The PHY drivers are modified to acquire and release the lock, as needed,
and assert the lock is held as a diagnostic aid.

The "usbnet" framework has had an overhaul of its internal locking
protocols to fit in with the media / mii changes, and the drivers adapted.

USB wifi drivers have been changed to provide their own adaptive mutex
to the ifmedia later via a new ieee80211_media_init_with_lock() function.
This is required because the USB drivers need an adaptive mutex.

Besised "usbnet", a few other drivers are converted: vmx, wm, ixgbe / ixv.

mcx also now calls ifmedia_init_with_lock() because it needs to also use
an adaptive mutex. The mcx driver still needs to be fully converted to
NET_MPSAFE.

- Simplify sc->mii_anegticks setting. Same as FreeBSD.

Don't set the default value not in the attach function. Instead, set the
default value (MII_ANEGTICKS) first in the beginning of the
mii_phy_add_media(). The function already has the code to change the value to
MII_ANEGTICKS_GIGE if it's gigabit capable.

- Remove extra pmf_device_register() call. It's done in mii_phy_add_media().

KNF. No functional change.

use a macro to tidy up the phydesc array initialization, from FreeBSD

Change MII PHY read/write API from:

int (*mii_readreg_t)(device_t, int, int);
void (*mii_writereg_t)(device_t, int, int, int);
to:

int (*mii_readreg_t)(device_t, int, int, uint16_t *);
int (*mii_writereg_t)(device_t, int, int, uint16_t);

Now we can test if a read/write operation failed or not by the return value.

In 802.3 spec says that the PHY shall not respond to read/write transaction
to the unimplemented register(22.2.4.3). Detecting timeout can be used to
check whether a register is implemented or not (if the register conforms to
the spec). ukphy(4) can be used this for MII_MMDACR and MII_MMDAADR.

Note that I noticed that the following code do infinite loop in the
read/wirte function. If it accesses unimplemented PHY register, it will hang.
It should be fixed:

arm/at91/at91emac.c
arm/ep93xx/epe.c
arm/omap/omapl1x_emac.c
mips/ralink/ralink_eth.c
arch/powerpc/booke/dev/pq3etsec.c(read)
dev/cadence/if_cemac.c <- hkenken
dev/ic/lan9118.c


Tested with the following device:

axe+ukphy
axe+rgephy
axen+rgephy (tested by Andrius V)
wm+atphy
wm+ukphy
wm+igphy
wm+ihphy
wm+makphy
sk+makphy
sk+brgphy
sk+gentbi
msk+makphy
sip+icsphy
sip+ukphy
re+rgephy
bge+brgphy
bnx+brgphy
gsip+gphyter
rtk+rlphy
fxp+inphy (tested by Andrius V)
tlp+acphy
ex+exphy
epic+qsphy
vge+ciphy (tested by Andrius V)
vr+ukphy (tested by Andrius V)
vte+ukphy (tested by Andrius V)

Not tested (MAC):
arm:at91emac
arm:cemac
arm:epe
arm:geminigmac
arm:enet
arm:cpsw
arm:emac(omac)
arm:emac(sunxi)
arm:npe
evbppc:temac
macppc:bm
macppc:gm
mips:aumac
mips:ae
mips:cnmac
mips:reth
mips:sbmac
playstation2:smap
powerpc:tsec
powerpc:emac(ibm4xx)
sgimips:mec
sparc:be
sf
ne(ax88190, dl10019)
awge
ep
gem
hme
smsh
mtd
sm
age
alc
ale
bce
cas
et
jme
lii
nfe
pcn
ste
stge
tl
xi
aue
mue
smsc
udav
url

Not tested (PHY):
amhphy
bmtphy
dmphy
etphy
glxtphy
ikphy
iophy
lxtphy
nsphyter
pnaphy
rdcphy
sqphy
tlphy
tqphy
urlphy

branches: 1.25.16; 1.25.18;
KNF. Remove extra spaces. No functional change.

branches: 1.24.4;
IFM_FDX and IFM_HDX use different bit, so set IFM_HDX bit if it's not full
duplex. For many drivers, it recognize half duplex if IFM_FDX isn't set,
but not for others. Same as {Free|Open}BSD.

branches: 1.23.22; 1.23.26; 1.23.40;
Reduce code duplication: most PHY drivers call mii_phy_add_media()
when they attach to the device tree, so call pmf_device_register(9)
once there instead of once in more than twenty drivers.

branches: 1.22.6; 1.22.8;
device_t/softc split for all mii(4) devices, and other related
cosmetic changes.

branches: 1.21.2; 1.21.4;
use aprint_*_dev and device_xname

branches: 1.20.10;
Merge jmcneill-pm branch.

branches: 1.19.22; 1.19.24; 1.19.34; 1.19.36;
- don't hardcode 5, 10 define and use MII_ANEGTICKS{,_GIGE}
- instead of != limit, use <= limit (conservative)

__unused removal on arguments; approved by core.

- sprinkle __unused on function decls.
- fix a couple of unused bugs
- no more -Wno-unused for i386

branches: 1.16.8; 1.16.10;
Use device_private().

branches: 1.15.4; 1.15.6; 1.15.8; 1.15.10; 1.15.12;
merge ktrace-lwp.

branches: 1.14.12;
Make use of static.

branches: 1.13.2;
Use aprint*().

Add support for Am79C87[45].

Add altima (broadcom) AC101L, phy for 3com 3c905cx-tx on nForce2 MCP-T

Add trailing ; to CFATTACH_DECL.

Use CFATTACH_DECL().

Declare all cfattach structures const.

Don't include <malloc.h> - no memory management style functions are use
in any of the MII drivers.

branches: 1.6.2;
Fix typo in comment. kern/16382 by Sean Davis.

Get flags passed down to PHY drivers correctly. This was done on
an adhoc basis in a couple of PHY drivers, this fixes it more generally.

Per a discussion w/ Cliff Neighbors <cliff@allegronetworks.com>.

Add another Altima PHY. From Peter Seebach.

add RCSID

branches: 1.2.4;
ANSI'ify.

branches: 1.1.2;
Driver for the Altima AC101 10/100 Ethernet PHY.

branches: 1.2.2; 1.2.6;
Note compatibility with the Am79c874, and document some more
registers.

Driver for the Altima AC101 10/100 Ethernet PHY.

s/should't/shouldn't/ and s/mistmatch/mismatch/ in comments.

Define and implement a locking protocol for the ifmedia / mii layers:
- MP-safe drivers provide a mutex to ifmedia that is used to serialize
access to media-related structures / hardware regsiters. Converted
drivers use the new ifmedia_init_with_lock() function for this. The
new name is provided to ease the transition.
- Un-converted drivers continue to call ifmedia_init(), which will supply
a compatibility lock to be used instead. Several media-related entry
points must be aware of this compatibility lock, and are able to acquire
it recursively a limited number of times, if needed. This is a SPIN
mutex with priority IPL_NET.
- This same lock is used to serialize access to PHY registers and other
MII-related data structures.

The PHY drivers are modified to acquire and release the lock, as needed,
and assert the lock is held as a diagnostic aid.

The "usbnet" framework has had an overhaul of its internal locking
protocols to fit in with the media / mii changes, and the drivers adapted.

USB wifi drivers have been changed to provide their own adaptive mutex
to the ifmedia later via a new ieee80211_media_init_with_lock() function.
This is required because the USB drivers need an adaptive mutex.

Besised "usbnet", a few other drivers are converted: vmx, wm, ixgbe / ixv.

mcx also now calls ifmedia_init_with_lock() because it needs to also use
an adaptive mutex. The mcx driver still needs to be fully converted to
NET_MPSAFE.

- Simplify sc->mii_anegticks setting. Same as FreeBSD.

Don't set the default value not in the attach function. Instead, set the
default value (MII_ANEGTICKS) first in the beginning of the
mii_phy_add_media(). The function already has the code to change the value to
MII_ANEGTICKS_GIGE if it's gigabit capable.

- Remove extra pmf_device_register() call. It's done in mii_phy_add_media().

Fix typo in comment.

KNF. No functional change.

use a macro to tidy up the phydesc array initialization, from FreeBSD

Change MII PHY read/write API from:

int (*mii_readreg_t)(device_t, int, int);
void (*mii_writereg_t)(device_t, int, int, int);
to:

int (*mii_readreg_t)(device_t, int, int, uint16_t *);
int (*mii_writereg_t)(device_t, int, int, uint16_t);

Now we can test if a read/write operation failed or not by the return value.

In 802.3 spec says that the PHY shall not respond to read/write transaction
to the unimplemented register(22.2.4.3). Detecting timeout can be used to
check whether a register is implemented or not (if the register conforms to
the spec). ukphy(4) can be used this for MII_MMDACR and MII_MMDAADR.

Note that I noticed that the following code do infinite loop in the
read/wirte function. If it accesses unimplemented PHY register, it will hang.
It should be fixed:

arm/at91/at91emac.c
arm/ep93xx/epe.c
arm/omap/omapl1x_emac.c
mips/ralink/ralink_eth.c
arch/powerpc/booke/dev/pq3etsec.c(read)
dev/cadence/if_cemac.c <- hkenken
dev/ic/lan9118.c


Tested with the following device:

axe+ukphy
axe+rgephy
axen+rgephy (tested by Andrius V)
wm+atphy
wm+ukphy
wm+igphy
wm+ihphy
wm+makphy
sk+makphy
sk+brgphy
sk+gentbi
msk+makphy
sip+icsphy
sip+ukphy
re+rgephy
bge+brgphy
bnx+brgphy
gsip+gphyter
rtk+rlphy
fxp+inphy (tested by Andrius V)
tlp+acphy
ex+exphy
epic+qsphy
vge+ciphy (tested by Andrius V)
vr+ukphy (tested by Andrius V)
vte+ukphy (tested by Andrius V)

Not tested (MAC):
arm:at91emac
arm:cemac
arm:epe
arm:geminigmac
arm:enet
arm:cpsw
arm:emac(omac)
arm:emac(sunxi)
arm:npe
evbppc:temac
macppc:bm
macppc:gm
mips:aumac
mips:ae
mips:cnmac
mips:reth
mips:sbmac
playstation2:smap
powerpc:tsec
powerpc:emac(ibm4xx)
sgimips:mec
sparc:be
sf
ne(ax88190, dl10019)
awge
ep
gem
hme
smsh
mtd
sm
age
alc
ale
bce
cas
et
jme
lii
nfe
pcn
ste
stge
tl
xi
aue
mue
smsc
udav
url

Not tested (PHY):
amhphy
bmtphy
dmphy
etphy
glxtphy
ikphy
iophy
lxtphy
nsphyter
pnaphy
rdcphy
sqphy
tlphy
tqphy
urlphy

branches: 1.20.26; 1.20.28;
IFM_FDX and IFM_HDX use different bit, so set IFM_HDX bit if it's not full
duplex. For many drivers, it recognize half duplex if IFM_FDX isn't set,
but not for others. Same as {Free|Open}BSD.

branches: 1.19.22; 1.19.26; 1.19.40;
Reduce code duplication: most PHY drivers call mii_phy_add_media()
when they attach to the device tree, so call pmf_device_register(9)
once there instead of once in more than twenty drivers.

branches: 1.18.6; 1.18.8;
device_t/softc split for all mii(4) devices, and other related
cosmetic changes.

branches: 1.17.2; 1.17.4;
use aprint_*_dev and device_xname

branches: 1.16.10;
Merge jmcneill-pm branch.

branches: 1.15.22; 1.15.24; 1.15.34; 1.15.36;
- don't hardcode 5, 10 define and use MII_ANEGTICKS{,_GIGE}
- instead of != limit, use <= limit (conservative)

__unused removal on arguments; approved by core.

- sprinkle __unused on function decls.
- fix a couple of unused bugs
- no more -Wno-unused for i386

branches: 1.12.8; 1.12.10;
Use device_private().

branches: 1.11.4; 1.11.6; 1.11.8; 1.11.10; 1.11.12;
merge ktrace-lwp.

branches: 1.10.12;
Make use of static.

branches: 1.9.2;
Use aprint*().

Add trailing ; to CFATTACH_DECL.

Use CFATTACH_DECL().

Declare all cfattach structures const.

Don't include <malloc.h> - no memory management style functions are use
in any of the MII drivers.

branches: 1.4.2;
Get flags passed down to PHY drivers correctly. This was done on
an adhoc basis in a couple of PHY drivers, this fixes it more generally.

Per a discussion w/ Cliff Neighbors <cliff@allegronetworks.com>.

add RCSID

branches: 1.2.4; 1.2.6;
ANSI'ify.

Driver for the 10BASE-T portion of the AMD 79c901 PHY.

s/chane/change/ in comment.

branches: 1.1.4; 1.1.6;
Driver for the 10BASE-T portion of the AMD 79c901 PHY.

Use device_{get,set}prop_bool() for rgmii-related properties.

fix typos in comments, mainly s/harware/hardware/.

branches: 1.31.6;
Retry autonegotiation every mii_anegticks seconds instead of mii_anegticks+1.

branches: 1.30.22;
Define and implement a locking protocol for the ifmedia / mii layers:
- MP-safe drivers provide a mutex to ifmedia that is used to serialize
access to media-related structures / hardware regsiters. Converted
drivers use the new ifmedia_init_with_lock() function for this. The
new name is provided to ease the transition.
- Un-converted drivers continue to call ifmedia_init(), which will supply
a compatibility lock to be used instead. Several media-related entry
points must be aware of this compatibility lock, and are able to acquire
it recursively a limited number of times, if needed. This is a SPIN
mutex with priority IPL_NET.
- This same lock is used to serialize access to PHY registers and other
MII-related data structures.

The PHY drivers are modified to acquire and release the lock, as needed,
and assert the lock is held as a diagnostic aid.

The "usbnet" framework has had an overhaul of its internal locking
protocols to fit in with the media / mii changes, and the drivers adapted.

USB wifi drivers have been changed to provide their own adaptive mutex
to the ifmedia later via a new ieee80211_media_init_with_lock() function.
This is required because the USB drivers need an adaptive mutex.

Besised "usbnet", a few other drivers are converted: vmx, wm, ixgbe / ixv.

mcx also now calls ifmedia_init_with_lock() because it needs to also use
an adaptive mutex. The mcx driver still needs to be fully converted to
NET_MPSAFE.

Fix a bug that atphy(4) doesn't work with Attansic L2 rev. 1.
Reported by Rocky Hotas.

- On ASUS M2N-MX SE Plus (NVIDIA MCP61 with Attansic L2 rev. 1), changing
debug port 0x29's value makes the next PHY read fail with error. Read any
register to ignore this problem if the PHY is Attansic L2 revision 1.
I don't know if this problem is from L2 rev. 1 itself or from the
combination because I have only one machine which has L2 rev. "1".
At least, ASUS eee pc 900 (Attansic L2 rev. "2") has no this problem.
- Add comment. AR8021 document has no description about the power saving
control register(debug port 0x29).
- Add comment. AR8031 document says the lower 14 bits are reserved and the
default value is 0x36d0. Shouldn't we clear those bits?
- I have no document neither L1(F1) nor L2(F2), so I don't know whether the
debug port access is correct or not.

Tested with the following machines:
- ASUS P5B SE, L1 rev. 5, age(4)
- ASUS K50IJ, L1 rev. 9, ale(4)
- ASUS eee pc 900, L2 rev. 2, lii(4)
- ASUS M2N-MX SE Plus, L2 rev. 1, nfe(4)
- Intel DP55WB, 82578(AR8021 rev. 2), wm(4)
- Dell inspiron 14z, AR0835 rev. 9, alc(4)

0x001374 is non-bitreversed value of Attansic OUI(0x00c82e).

Attansic/Atheros correctly uses ID1 and ID2 register, so delete all 0x001374
related entries.

KNF. No functional change.

- Simplify sc->mii_anegticks setting. Same as FreeBSD.

Don't set the default value not in the attach function. Instead, set the
default value (MII_ANEGTICKS) first in the beginning of the
mii_phy_add_media(). The function already has the code to change the value to
MII_ANEGTICKS_GIGE if it's gigabit capable.

- Remove extra pmf_device_register() call. It's done in mii_phy_add_media().

Fix a bug that atphy(4) can't negotiate correctly when the media setting is
neither auto nor 1000baseT. Use correct index for mii_media_table[].

History: mii_anar() is first added in OpenBSD and ported to NetBSD. On NetBSD,
only atphy(4) use this function. mii_physubr.c rev. 1.75 changed mii_anar()
for simplify. It changed the argument from the ifmedia word to ifm_data used
in our MII API, but the caller have not been changed. And then, PR kern/50206
was reported and the caller was modified by me to prevent panic but it was not
correct fix.

Modified atphy(4)

* Support CLK_25M clock out.
* Support internal delay for RGMII interface.

s/etphy/atphy/. No functional change.

branches: 1.22.4;
Simplify. No functional change.

KNF. No functional change.

use a macro to tidy up the phydesc array initialization, from FreeBSD

Change MII PHY read/write API from:

int (*mii_readreg_t)(device_t, int, int);
void (*mii_writereg_t)(device_t, int, int, int);
to:

int (*mii_readreg_t)(device_t, int, int, uint16_t *);
int (*mii_writereg_t)(device_t, int, int, uint16_t);

Now we can test if a read/write operation failed or not by the return value.

In 802.3 spec says that the PHY shall not respond to read/write transaction
to the unimplemented register(22.2.4.3). Detecting timeout can be used to
check whether a register is implemented or not (if the register conforms to
the spec). ukphy(4) can be used this for MII_MMDACR and MII_MMDAADR.

Note that I noticed that the following code do infinite loop in the
read/wirte function. If it accesses unimplemented PHY register, it will hang.
It should be fixed:

arm/at91/at91emac.c
arm/ep93xx/epe.c
arm/omap/omapl1x_emac.c
mips/ralink/ralink_eth.c
arch/powerpc/booke/dev/pq3etsec.c(read)
dev/cadence/if_cemac.c <- hkenken
dev/ic/lan9118.c


Tested with the following device:

axe+ukphy
axe+rgephy
axen+rgephy (tested by Andrius V)
wm+atphy
wm+ukphy
wm+igphy
wm+ihphy
wm+makphy
sk+makphy
sk+brgphy
sk+gentbi
msk+makphy
sip+icsphy
sip+ukphy
re+rgephy
bge+brgphy
bnx+brgphy
gsip+gphyter
rtk+rlphy
fxp+inphy (tested by Andrius V)
tlp+acphy
ex+exphy
epic+qsphy
vge+ciphy (tested by Andrius V)
vr+ukphy (tested by Andrius V)
vte+ukphy (tested by Andrius V)

Not tested (MAC):
arm:at91emac
arm:cemac
arm:epe
arm:geminigmac
arm:enet
arm:cpsw
arm:emac(omac)
arm:emac(sunxi)
arm:npe
evbppc:temac
macppc:bm
macppc:gm
mips:aumac
mips:ae
mips:cnmac
mips:reth
mips:sbmac
playstation2:smap
powerpc:tsec
powerpc:emac(ibm4xx)
sgimips:mec
sparc:be
sf
ne(ax88190, dl10019)
awge
ep
gem
hme
smsh
mtd
sm
age
alc
ale
bce
cas
et
jme
lii
nfe
pcn
ste
stge
tl
xi
aue
mue
smsc
udav
url

Not tested (PHY):
amhphy
bmtphy
dmphy
etphy
glxtphy
ikphy
iophy
lxtphy
nsphyter
pnaphy
rdcphy
sqphy
tlphy
tqphy
urlphy

branches: 1.18.8; 1.18.14; 1.18.16;
Set mii_mpd_{oui,model,rev}.

branches: 1.17.2;
Fix incorrect argument of mii_anar(). Fixes PR#50206.
XXX pullup -[567]

branches: 1.16.2; 1.16.4; 1.16.6;
Fix a bug that atphy doesn't set the capability of pause function correctly.
This bug was added in "mii.h" rev. 1.17.

branches: 1.15.6;
The bit location of link ability is different between 1000Base-X and others
(See annex 28B.2 and 28D). The old mii.h defined ANAR_X_PAUSE_* macros. Those
macros were named for 1000Base-X, but the bit definitions were not for
1000Base-X but for others (e.g. 1000BaseT). So there was bug in auto
negotiation for 1000Base-X, but there was no bug for other media. Define new
macro named ANAR_PAUSE_* and ANLPAR_PAUSE_* for other than 1000Base-X and fix
the bit definitions of ANAR_X_PAUSE_* and ANLPPAR_X_*. Change some PHY drivers
to use true macro. Same as other *BSDs.

atphy, ciphy, ihphy and rgephy are gigabit PHY, so not only check IFM_AUTO
for autonegotiation but also check IFM_1000_T, too.

Fix a bug that the auto negotiation timer isn't cleard as we expected.
To not to forget clearing the timer and to reduce the code duplication,
clear mii_ticks in *mii_phy_auto().

branches: 1.12.2;
Make it match AR8035

branches: 1.11.2; 1.11.8; 1.11.12; 1.11.14;
Attansic L2 is 10/100 but the status registers report 1000 Mbps, so don't
bother reading MII_EXTSR for this part. Also use MII_ANEGTICKS instead of
MII_ANEGTICKS_GIGE for this part.

match on MII_MODEL_ATTANSIC_L2

alc0 at pci3 dev 0 function 0: Atheros AR8152 v2.0 PCIe Fast Ethernet
alc0: ioapic0 pin 17
alc0: Ethernet address 00:26:6c:9e:d4:c1
atphy0 at alc0 phy 0: L2 10/100 PHY, rev. 5
atphy0: 10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, 1000baseT-FDX, auto

fix comment. from PR kern/44395

Fixes from PR kern/44395 Masanori Kanaoka:
- Fix register address in ale_phy_reset(). from linux
- Fix mask value in ale_stop_mac().
- Fix multicast handling. from openbsd
- Fix phy reset handling.

branches: 1.7.2; 1.7.4;
Recognize AR8021

branches: 1.6.4; 1.6.6;
- replace magic number with proper define from mii.h. From OpenBSD.
- fix typo in comment. From OpenBSD.

branches: 1.5.2; 1.5.4;
use auto-negotiation for manual media type selection.
This should fix establishment of 10/100Mbps link on L1E.
Helpful to reduce watchdog timeout problems on L1 (age(4)) reported by Jochen Kunz.
From FreeBSD via OpenBSD.

pretty print atphy's supported medias in dmesg.

Attach atphy(4) on ATTANSIC oui.
ok cube@

branches: 1.2.2; 1.2.4; 1.2.6;
add RCSID

Driver for Attansic Technology F1 10/100/1000 Ethernet PHY.

Written by Pyun YongHyeon for FreeBSD, ported to OpenBSD by Kevin Lo, ported to
NetBSD by me.

Define and implement a locking protocol for the ifmedia / mii layers:
- MP-safe drivers provide a mutex to ifmedia that is used to serialize
access to media-related structures / hardware regsiters. Converted
drivers use the new ifmedia_init_with_lock() function for this. The
new name is provided to ease the transition.
- Un-converted drivers continue to call ifmedia_init(), which will supply
a compatibility lock to be used instead. Several media-related entry
points must be aware of this compatibility lock, and are able to acquire
it recursively a limited number of times, if needed. This is a SPIN
mutex with priority IPL_NET.
- This same lock is used to serialize access to PHY registers and other
MII-related data structures.

The PHY drivers are modified to acquire and release the lock, as needed,
and assert the lock is held as a diagnostic aid.

The "usbnet" framework has had an overhaul of its internal locking
protocols to fit in with the media / mii changes, and the drivers adapted.

USB wifi drivers have been changed to provide their own adaptive mutex
to the ifmedia later via a new ieee80211_media_init_with_lock() function.
This is required because the USB drivers need an adaptive mutex.

Besised "usbnet", a few other drivers are converted: vmx, wm, ixgbe / ixv.

mcx also now calls ifmedia_init_with_lock() because it needs to also use
an adaptive mutex. The mcx driver still needs to be fully converted to
NET_MPSAFE.

- Simplify sc->mii_anegticks setting. Same as FreeBSD.

Don't set the default value not in the attach function. Instead, set the
default value (MII_ANEGTICKS) first in the beginning of the
mii_phy_add_media(). The function already has the code to change the value to
MII_ANEGTICKS_GIGE if it's gigabit capable.

- Remove extra pmf_device_register() call. It's done in mii_phy_add_media().

KNF. No functional change.

use a macro to tidy up the phydesc array initialization, from FreeBSD

Change MII PHY read/write API from:

int (*mii_readreg_t)(device_t, int, int);
void (*mii_writereg_t)(device_t, int, int, int);
to:

int (*mii_readreg_t)(device_t, int, int, uint16_t *);
int (*mii_writereg_t)(device_t, int, int, uint16_t);

Now we can test if a read/write operation failed or not by the return value.

In 802.3 spec says that the PHY shall not respond to read/write transaction
to the unimplemented register(22.2.4.3). Detecting timeout can be used to
check whether a register is implemented or not (if the register conforms to
the spec). ukphy(4) can be used this for MII_MMDACR and MII_MMDAADR.

Note that I noticed that the following code do infinite loop in the
read/wirte function. If it accesses unimplemented PHY register, it will hang.
It should be fixed:

arm/at91/at91emac.c
arm/ep93xx/epe.c
arm/omap/omapl1x_emac.c
mips/ralink/ralink_eth.c
arch/powerpc/booke/dev/pq3etsec.c(read)
dev/cadence/if_cemac.c <- hkenken
dev/ic/lan9118.c


Tested with the following device:

axe+ukphy
axe+rgephy
axen+rgephy (tested by Andrius V)
wm+atphy
wm+ukphy
wm+igphy
wm+ihphy
wm+makphy
sk+makphy
sk+brgphy
sk+gentbi
msk+makphy
sip+icsphy
sip+ukphy
re+rgephy
bge+brgphy
bnx+brgphy
gsip+gphyter
rtk+rlphy
fxp+inphy (tested by Andrius V)
tlp+acphy
ex+exphy
epic+qsphy
vge+ciphy (tested by Andrius V)
vr+ukphy (tested by Andrius V)
vte+ukphy (tested by Andrius V)

Not tested (MAC):
arm:at91emac
arm:cemac
arm:epe
arm:geminigmac
arm:enet
arm:cpsw
arm:emac(omac)
arm:emac(sunxi)
arm:npe
evbppc:temac
macppc:bm
macppc:gm
mips:aumac
mips:ae
mips:cnmac
mips:reth
mips:sbmac
playstation2:smap
powerpc:tsec
powerpc:emac(ibm4xx)
sgimips:mec
sparc:be
sf
ne(ax88190, dl10019)
awge
ep
gem
hme
smsh
mtd
sm
age
alc
ale
bce
cas
et
jme
lii
nfe
pcn
ste
stge
tl
xi
aue
mue
smsc
udav
url

Not tested (PHY):
amhphy
bmtphy
dmphy
etphy
glxtphy
ikphy
iophy
lxtphy
nsphyter
pnaphy
rdcphy
sqphy
tlphy
tqphy
urlphy

branches: 1.32.16; 1.32.18;
KNF. Remove extra spaces. No functional change.

branches: 1.31.4;
No functional change:
- Fix typo.
- Remove trailing white spaces.
- Capitalize comments.
- Tabify.
- KNF.

branches: 1.30.18; 1.30.22; 1.30.36;
Remove closes 3 & 4 from my licence. Lots of thanks to Soren Jacobsen
for the booring work !

Enable flow control and poll the status appropriately.

contributed anonymously.

Reduce code duplication: most PHY drivers call mii_phy_add_media()
when they attach to the device tree, so call pmf_device_register(9)
once there instead of once in more than twenty drivers.

branches: 1.27.6; 1.27.8;
device_t/softc split for all mii(4) devices, and other related
cosmetic changes.

Remove clause 3 and 4 from TNF licenses

branches: 1.25.2; 1.25.4;
use aprint_*_dev and device_xname

branches: 1.24.6;
Updates from OpenBSD:
-- r1.15
Add some DSP init code for BCM5221's.

Values derived from Apple's GMAC driver, same init
code also exists in Linux's sungem_phy driver.

tested by miod@ pedro@ and otto@ on a few systems.
-- r1.16
Always explicitly set IFM_HDX for half-duplex.

From brad@
--

branches: 1.23.2;
Merge jmcneill-pm branch.

branches: 1.22.2; 1.22.16; 1.22.22; 1.22.24; 1.22.30; 1.22.34; 1.22.36;
- don't hardcode 5, 10 define and use MII_ANEGTICKS{,_GIGE}
- instead of != limit, use <= limit (conservative)

__unused removal on arguments; approved by core.

- sprinkle __unused on function decls.
- fix a couple of unused bugs
- no more -Wno-unused for i386

branches: 1.19.8; 1.19.10;
Use device_private().

branches: 1.18.4; 1.18.6; 1.18.8; 1.18.10; 1.18.12;
merge ktrace-lwp.

branches: 1.17.8;
Support Broadcom BCM5222 dual 10/100 media interface.

branches: 1.16.4;
Add BCM 4401 phy (from FreeBSD).

Make use of static.

branches: 1.14.2;
Only read media status bits when they are valid.
From Makoto Fujiwara in PR kern/21810.

Match Broadcom's BCM5214 Quad 10/100 media interface.

Use aprint*().

Add trailing ; to CFATTACH_DECL.

Use CFATTACH_DECL().

Declare all cfattach structures const.

Don't include <malloc.h> - no memory management style functions are use
in any of the MII drivers.

branches: 1.7.2;
Get flags passed down to PHY drivers correctly. This was done on
an adhoc basis in a couple of PHY drivers, this fixes it more generally.

Per a discussion w/ Cliff Neighbors <cliff@allegronetworks.com>.

add RCSID

ANSI'ify.

Use the bit-swapped OUI for the 5221. (Have I mentioned that I really
don't like how the current MII_OUI() works?)

Add a 3Com 3c905B internal PHY.

branches: 1.2.2; 1.2.4;
Initialize sc->sc_anegticks to 5 on attach to prevent panic due to
assertion failure.

Driver for the Broadcom BCM5201/BCM5202 "Mini-Theta" Ethernet
PHYs. This is also the same PHY core as in the 3Com 3c905C.
The PHY is documented in the 3c905C manual.

s/dupled/duplex/ in comments.

fix various typos in comments and messages.

Remove clause 3 and 4 from TNF licenses

branches: 1.1.2; 1.1.122; 1.1.124; 1.1.126;
Driver for the Broadcom BCM5201/BCM5202 "Mini-Theta" Ethernet
PHYs. This is also the same PHY core as in the 3Com 3c905C.
The PHY is documented in the 3c905C manual.

Retry autonegotiation every mii_anegticks seconds instead of mii_anegticks+1.

branches: 1.90.20;
Add support for BCM54213PE RGMII clock delays, from OpenBSD

Don't set DVF_DETACH_SHUTDOWN. The MII layer wants to manage the lifecycle
of the PHY devices, and if a NIC driver chooses not to detach its PHYs
at shutdown, that's the driver's business.

PR kern/55121.

Define and implement a locking protocol for the ifmedia / mii layers:
- MP-safe drivers provide a mutex to ifmedia that is used to serialize
access to media-related structures / hardware regsiters. Converted
drivers use the new ifmedia_init_with_lock() function for this. The
new name is provided to ease the transition.
- Un-converted drivers continue to call ifmedia_init(), which will supply
a compatibility lock to be used instead. Several media-related entry
points must be aware of this compatibility lock, and are able to acquire
it recursively a limited number of times, if needed. This is a SPIN
mutex with priority IPL_NET.
- This same lock is used to serialize access to PHY registers and other
MII-related data structures.

The PHY drivers are modified to acquire and release the lock, as needed,
and assert the lock is held as a diagnostic aid.

The "usbnet" framework has had an overhaul of its internal locking
protocols to fit in with the media / mii changes, and the drivers adapted.

USB wifi drivers have been changed to provide their own adaptive mutex
to the ifmedia later via a new ieee80211_media_init_with_lock() function.
This is required because the USB drivers need an adaptive mutex.

Besised "usbnet", a few other drivers are converted: vmx, wm, ixgbe / ixv.

mcx also now calls ifmedia_init_with_lock() because it needs to also use
an adaptive mutex. The mcx driver still needs to be fully converted to
NET_MPSAFE.

Match BCM54213PE

branches: 1.86.2;
- Simplify sc->mii_anegticks setting. Same as FreeBSD.

Don't set the default value not in the attach function. Instead, set the
default value (MII_ANEGTICKS) first in the beginning of the
mii_phy_add_media(). The function already has the code to change the value to
MII_ANEGTICKS_GIGE if it's gigabit capable.

- Remove extra pmf_device_register() call. It's done in mii_phy_add_media().

Fix typo in comment.

branches: 1.84.4;
Fix a bug that the duplex of manual media setting may be wrong
when the IFM_GMASK bit other than IFM_[FH]DX is set. Same as a part of
FreeBSD r217413.

KNF. No functional change.

- Add Broadcom BCM540[24], BCM5424 and BCM5466.
- Sort lines.

use a macro to tidy up the phydesc array initialization, from FreeBSD

- Add support for BCM5762 ASIC devices.
- Add BCM5762, BCM5725, BCM5727, BCM57764, BCM57767 and BCM57787.

Change MII PHY read/write API from:

int (*mii_readreg_t)(device_t, int, int);
void (*mii_writereg_t)(device_t, int, int, int);
to:

int (*mii_readreg_t)(device_t, int, int, uint16_t *);
int (*mii_writereg_t)(device_t, int, int, uint16_t);

Now we can test if a read/write operation failed or not by the return value.

In 802.3 spec says that the PHY shall not respond to read/write transaction
to the unimplemented register(22.2.4.3). Detecting timeout can be used to
check whether a register is implemented or not (if the register conforms to
the spec). ukphy(4) can be used this for MII_MMDACR and MII_MMDAADR.

Note that I noticed that the following code do infinite loop in the
read/wirte function. If it accesses unimplemented PHY register, it will hang.
It should be fixed:

arm/at91/at91emac.c
arm/ep93xx/epe.c
arm/omap/omapl1x_emac.c
mips/ralink/ralink_eth.c
arch/powerpc/booke/dev/pq3etsec.c(read)
dev/cadence/if_cemac.c <- hkenken
dev/ic/lan9118.c


Tested with the following device:

axe+ukphy
axe+rgephy
axen+rgephy (tested by Andrius V)
wm+atphy
wm+ukphy
wm+igphy
wm+ihphy
wm+makphy
sk+makphy
sk+brgphy
sk+gentbi
msk+makphy
sip+icsphy
sip+ukphy
re+rgephy
bge+brgphy
bnx+brgphy
gsip+gphyter
rtk+rlphy
fxp+inphy (tested by Andrius V)
tlp+acphy
ex+exphy
epic+qsphy
vge+ciphy (tested by Andrius V)
vr+ukphy (tested by Andrius V)
vte+ukphy (tested by Andrius V)

Not tested (MAC):
arm:at91emac
arm:cemac
arm:epe
arm:geminigmac
arm:enet
arm:cpsw
arm:emac(omac)
arm:emac(sunxi)
arm:npe
evbppc:temac
macppc:bm
macppc:gm
mips:aumac
mips:ae
mips:cnmac
mips:reth
mips:sbmac
playstation2:smap
powerpc:tsec
powerpc:emac(ibm4xx)
sgimips:mec
sparc:be
sf
ne(ax88190, dl10019)
awge
ep
gem
hme
smsh
mtd
sm
age
alc
ale
bce
cas
et
jme
lii
nfe
pcn
ste
stge
tl
xi
aue
mue
smsc
udav
url

Not tested (PHY):
amhphy
bmtphy
dmphy
etphy
glxtphy
ikphy
iophy
lxtphy
nsphyter
pnaphy
rdcphy
sqphy
tlphy
tqphy
urlphy

Remove extra PHY read in brgphy_mii_phy_auto().

Whitespace fixes. No functional change.

branches: 1.76.20; 1.76.26; 1.76.28;
Add BCM5706.

- 5720S and 5709S share the same PHY id. Assume 5720S PHY if parent device
is bge(4). Same as FreeBSD. Tested on HP Moonshot.
- Fix media detect on some Fiber chips. Tested on my own BCM5709 card.
Same as {Free,Open}BSD.

Don't check BNX_CHIP_NUM. Check MII PHY ID because this check is used
for both bge(4) and bnx(4).

- Remove extra delay in brgphy_mii_phy_auto. Same as {Free,Open}BSD.
- s/u_int/uint/
- KNF.

Fix uninitiazed variable problem found on zaurus. I don't know why
only zaurus found it.

- Fix detection of BGEPHYF_FIBER_{MII|TBI}
- Add BCM5708S support in brgphy(4). The auto negotiation may have some bugs.
- Add 2500SX support (not tested).
- Fix bit definition of BRGPHY_MRBE_MSG_PG5_NP_T2 from FreeBSD.

IFM_FDX and IFM_HDX use different bit, so set IFM_HDX bit if it's not full
duplex. For many drivers, it recognize half duplex if IFM_FDX isn't set,
but not for others. Same as {Free|Open}BSD.

No functional change:
- Remove BRGPHY_SERDES_ANAR_* and BRGPHY_SERDES_ANLPAR_*. Those registers
are the same as MII_ANAR_* and MII_ANLPAR_*.
- Fix typo.
- Remove trailing white spaces.
- KNF.

branches: 1.68.2;
Cleanup flag setting. No functional change.
- Split flags into bge_flags and bge_phy_flags.
- Rename flags. PHY related flags are prefixed with BGEPHYF_*.
Other flags are prefixed with BGEF_*.

branches: 1.67.2;
Call brgpy specific autonego function in MII_TICK. Before this commit,
only MII_MEDIACHG calls brgphy_mii_phy_auto() and MII_TICK calls MI
mii_phy_auto(). That was not intended.

The bit location of link ability is different between 1000Base-X and others
(See annex 28B.2 and 28D). The old mii.h defined ANAR_X_PAUSE_* macros. Those
macros were named for 1000Base-X, but the bit definitions were not for
1000Base-X but for others (e.g. 1000BaseT). So there was bug in auto
negotiation for 1000Base-X, but there was no bug for other media. Define new
macro named ANAR_PAUSE_* and ANLPAR_PAUSE_* for other than 1000Base-X and fix
the bit definitions of ANAR_X_PAUSE_* and ANLPPAR_X_*. Change some PHY drivers
to use true macro. Same as other *BSDs.

KNF.

Fix a bug that the auto negotiation timer isn't cleard as we expected.
To not to forget clearing the timer and to reduce the code duplication,
clear mii_ticks in *mii_phy_auto().

In brgphyattach(), set sc_isbge, sc_isbnx and sc_phyflags before PHY_RESET()
because brgphy_reset() refers those flags.

- Rename PHY related flags for the consistency. It's the same as FreeBSD.
- Remove BGE_10_100_ONLY flag because this was not used.
For 10/100 devices, when calling mii_attach(), mask BMSR_EXTSTAT flag to
not to check Gigabit flags. It's the same as FreeBSD.

Add BCM5756, BCM5717C, BCM5719C, BCM5720C and BCM57780.

Add support for BCM57765. PR kern/46961

branches: 1.59.2; 1.59.8; 1.59.12;
add BCM5785 phy:

brgphy0 at bge0 phy 1: BCM5785 1000BASE-T media interface, rev. 3
brgphy0: 10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, 1000baseT, 1000baseT-FD , auto

branches: 1.58.2;
Fix BCM5709 PHY detection for ethernet PHYs (the SerDes case being already
handled):
- export bge(4) and bnx(4) CHIP ID and PHY flags to brgphy(4). Move to
"unsigned int" rather than "int", and reuse the same softc members for
chipid and phyflags (behavior controlled by the sc_isbge/isbnx boolean).
- apply bug fix for revisions A and B, so that autonegotiation can
complete (from OpenBSD).

Bug reported by Rivo Nurges via private mail, patch tested and
confirmed working by him (with thanks!)

branches: 1.57.2;
Add support for BCM5709S (SerDes PHY) in brgphy(4).

From FreeBSD, via OpenBSD, with some minor adaptations.

No comments, no objections on tech-kern@.

The driver was tested on a Dell M710. Access was kindly provided by
Uwe Toenjes, whom I thank for this.

See also http://mail-index.netbsd.org/tech-kern/2010/12/01/msg009478.html

Use device_private() instead of casting softc to (void *).

Include support for BCM5481. From OpenBSD and confirmed to work by
myself on XFX nForce 790i motherboard.

For a couple of flags, use bool/true/false instead of int/1/0.

branches: 1.53.2;
Add Broadcom BCM5482.
It tested on OpenBlockS600. (coming soon ;-)
And reorder.

branches: 1.52.2;
Add support for BCM5461,BCM5784 and BCM5761.
Enable brgphy_jumbo_settings().
Enable the Ethernet@Wirespeed function.

Support BCM5709CAX and BCM5709C PHYs.

Remove closes 3 & 4 from my licence. Lots of thanks to Soren Jacobsen
for the booring work !

Match the Broadcom BCM5464 PHY.

Unwrap not so long lines.

- no need to use device_parent() because device_t parent is passed
via config_attach(9)
- use device_is_a() rather than strcmp() for readability

struct device * -> device_t, no functional changes intended.

struct cfdata * -> cfdata_t, no functional changes intended.

use proplib

Some fixes and enhancements:

Both if_bge* and brgphy.[ch]:
Check the parent's MAC and use the quirk code for the bug like OpenBSD and
FreeBSD. Some bugs can't identify the PHY ID. For example, 5704 Ax has the
ADC bug, but 5704 A3 and 5704 B0 have the same PHY ID and revision. Add
BGE_PHY_CRC_BUG, BGE_PHY_ADC_BUG, BGE_PHY_5704_A0_BUG, BGE_PHY_JITTER_BUG,
BGE_PHY_ADJUST_TRIM and BGE_PHY_BER_BUG for this flag. Some of the DSP
patches are newly taken from OpenBSD and FreeBSD.

if_bge*:
Remove duplicated BGE_CHIPID_BCM5714_A0 entry in the known CHIPID table.
Fix obsolete comments.

brgphy.[ch]:
Add some PHY IDs.

TODO:
Add more three quirk code into bge and brgphy (the brgphy side's are
currently #if0'ed).
Add support for bnx into brgphy (currently #if0'ed)

Detach brgphy(4) at shutdown.

branches: 1.41.4;
Reduce code duplication: most PHY drivers call mii_phy_add_media()
when they attach to the device tree, so call pmf_device_register(9)
once there instead of once in more than twenty drivers.

branches: 1.40.2; 1.40.4; 1.40.8;
PR 39241: Add support for Broadcom BCM5906(M) from Karl Uwe Lockhoff

branches: 1.39.2; 1.39.6;
device_t/softc split for all mii(4) devices, and other related
cosmetic changes.

Remove clause 3 and 4 from TNF licenses

branches: 1.37.2; 1.37.4;
use aprint_*_dev and device_xname

branches: 1.36.6;
Add BCM5708C. Treating it like 5750/5780 is a guess but my Dell
PowerEdge 1950 now successfully negotiates 1000baseT full-duplex
where it didn't previously.

Merge jmcneill-pm branch.

branches: 1.34.2; 1.34.10; 1.34.12; 1.34.14;
Add support for BCM5754 and BCM5755. From suggestions from Jonathan
Stone.

branches: 1.33.6; 1.33.10;
tabify

branches: 1.32.2; 1.32.4;
Commit from AsiaBSDCon 2007's place

Fix the problem "ifconfig bge0 media auto" negotiates to 100Mbps on 1000BaseT
HUB.

branches: 1.31.2; 1.31.4;
Match BCM5752 PHY. From Matthias Drochner and FreeBSD/OpenBSD.

- don't hardcode 5, 10 define and use MII_ANEGTICKS{,_GIGE}
- instead of != limit, use <= limit (conservative)

__unused removal on arguments; approved by core.

- sprinkle __unused on function decls.
- fix a couple of unused bugs
- no more -Wno-unused for i386

branches: 1.27.8; 1.27.10;
Add PHY ID for the integral 1000base-T PHY in the bcm5780 (also known
as Broadcom/Serverworks HT-2000). Following the Broadcom-supplied
Linux driver, update brgphy.c to treat the bcm5780 integral copper PHY
just like the 5714/5715 PHY.

For NetBSB-3 pullup purposes, this commit includes:

src/sys/dev/mii/miidevs revision: 1.65 -> 1.66
src/sys/dev/mii/brgphy.c revision: 1.26 -> 1.27

plus a regen of src/sys/dev/mii/{miidevs.h,miidevs_data.h}.

Use device_private().

branches: 1.25.4; 1.25.6; 1.25.8; 1.25.10; 1.25.12;
Oops! Restore brgphys[]z stanza for bcm5705 integral PHY, somehow
lost in the previous addition of the bcm5714 integral phy.

Add PHY mii-id for the Broadcom 1000base-T PHY integral to the
bcm5714, added yesterday to sys/dev/pci/if_bge.c in revision 1.98,
since it appears the 5714 has a new PHY revision. Pending further
details, follow the FreeBSD code (as submitted by
John Cagle <john dot cagle at hp dot com> and committed by
ps@FreeBSD.ORG), and treat the 5714 integral PHY the same way as we
treat the bcm5750 integral PHY.

branches: 1.23.2; 1.23.4;
nuke trailing whitespace

branches: 1.22.4; 1.22.6;
Add support for the PHY found with BCM575x chips. Beware the magic
numbers!

From FreeBSD.

Make use of static.

gmii_phy_getflowcontrol() -> mii_phy_flowstatus(). From HITOSHI Osada.

Fetch negotiated flow control parameters. From HITOSHI Osada.

Add support for the BCM5705:
- Only BCM5705M asic rev A1 was tested.

Thanks to Bill Paul (wpaul@freebsd.org) for help and support.

Approved by: Frank van der Linden <fvdl@netbsd.org>

branches: 1.17.2;
Use aprint*().

Update first 16-bit word of 5401 DSP patch to match latest Linux drivers.
Add comment saying what the magic values do: disable tap power management.

Add PHY entry for bcm5704c (copper) integral phy to miidevs.
Add 5704 phy and dsp patch, and untestsed dsp patch for 5703, to brgphy.c.

Recognize BCM5703 phy. (with this and the bge changes, the bge0 on my
ASUS P4PE m/b is now working)

Add trailing ; to CFATTACH_DECL.

Use CFATTACH_DECL().

Declare all cfattach structures const.

Move the BCM5401 DSP patch out of the bge driver and into
the brgphy driver; all users of the BCM5400 and BCM5401 need
the DSP patch and the sledgehammer-reset-at-media-set-time.

Also add a DSP patch for the BCM5411 gleaned from Apple's
GMAC driver for Darwin.

Tested with a 3Com 3c996-T (BCM5700 + BCM5401).

Match the BCM5421, found on the Apple Xserve. From Matt Thomas.

Don't include <malloc.h> - no memory management style functions are use
in any of the MII drivers.

Match BCM5701

branches: 1.6.2; 1.6.4;
Get flags passed down to PHY drivers correctly. This was done on
an adhoc basis in a couple of PHY drivers, this fixes it more generally.

Per a discussion w/ Cliff Neighbors <cliff@allegronetworks.com>.

add RCSID

ANSI'ify.

branches: 1.3.2;
IFM_1000_TX -> IFM_1000_T, as (breifly) discussed on tech-net.

branches: 1.2.2;
Make PHY matching all table-driven.

Driver for Broadcom BCM5400-family Gig-E PHYs. Programming
information gleaned from FreeBSD.

s/Regiser/Register/ and s/regester/register/ in comments.

Add support for BCM54213PE RGMII clock delays, from OpenBSD

KNF. No functional change.

Remove the definition of BRGPHY_MII_1000CTL and use MII_GTCR.

branches: 1.9.26; 1.9.28;
- Fix detection of BGEPHYF_FIBER_{MII|TBI}
- Add BCM5708S support in brgphy(4). The auto negotiation may have some bugs.
- Add 2500SX support (not tested).
- Fix bit definition of BRGPHY_MRBE_MSG_PG5_NP_T2 from FreeBSD.

No functional change:
- Fix typo.
- Remove trailing white spaces.
- Capitalize comments.
- Tabify.
- KNF.

No functional change:
- Remove BRGPHY_SERDES_ANAR_* and BRGPHY_SERDES_ANLPAR_*. Those registers
are the same as MII_ANAR_* and MII_ANLPAR_*.
- Fix typo.
- Remove trailing white spaces.
- KNF.

branches: 1.6.8;
Sort. No functional change.

branches: 1.5.8; 1.5.14; 1.5.18;
Add support for BCM5709S (SerDes PHY) in brgphy(4).

From FreeBSD, via OpenBSD, with some minor adaptations.

No comments, no objections on tech-kern@.

The driver was tested on a Dell M710. Access was kindly provided by
Uwe Toenjes, whom I thank for this.

See also http://mail-index.netbsd.org/tech-kern/2010/12/01/msg009478.html

branches: 1.4.4;
Some fixes and enhancements:

Both if_bge* and brgphy.[ch]:
Check the parent's MAC and use the quirk code for the bug like OpenBSD and
FreeBSD. Some bugs can't identify the PHY ID. For example, 5704 Ax has the
ADC bug, but 5704 A3 and 5704 B0 have the same PHY ID and revision. Add
BGE_PHY_CRC_BUG, BGE_PHY_ADC_BUG, BGE_PHY_5704_A0_BUG, BGE_PHY_JITTER_BUG,
BGE_PHY_ADJUST_TRIM and BGE_PHY_BER_BUG for this flag. Some of the DSP
patches are newly taken from OpenBSD and FreeBSD.

if_bge*:
Remove duplicated BGE_CHIPID_BCM5714_A0 entry in the known CHIPID table.
Fix obsolete comments.

brgphy.[ch]:
Add some PHY IDs.

TODO:
Add more three quirk code into bge and brgphy (the brgphy side's are
currently #if0'ed).
Add support for bnx into brgphy (currently #if0'ed)

branches: 1.3.2; 1.3.4; 1.3.8; 1.3.12;
PR 39241: Add support for Broadcom BCM5906(M) from Karl Uwe Lockhoff

branches: 1.2.58; 1.2.104; 1.2.108; 1.2.110; 1.2.114;
Add more register definitions. From FreeBSD.

branches: 1.1.2; 1.1.4; 1.1.16; 1.1.18;
Driver for Broadcom BCM5400-family Gig-E PHYs. Programming
information gleaned from FreeBSD.

Retry autonegotiation every mii_anegticks seconds instead of mii_anegticks+1.

branches: 1.41.20;
Don't do full initialization for autonego when just restarting autonego
because it's not required.

This change reduce extra initialization which include PHY_RESET() which
caused long delay(max 500ms).

- Remove the waitfor argument from mii_phy_auto().
- Whitespace fix.

Define and implement a locking protocol for the ifmedia / mii layers:
- MP-safe drivers provide a mutex to ifmedia that is used to serialize
access to media-related structures / hardware regsiters. Converted
drivers use the new ifmedia_init_with_lock() function for this. The
new name is provided to ease the transition.
- Un-converted drivers continue to call ifmedia_init(), which will supply
a compatibility lock to be used instead. Several media-related entry
points must be aware of this compatibility lock, and are able to acquire
it recursively a limited number of times, if needed. This is a SPIN
mutex with priority IPL_NET.
- This same lock is used to serialize access to PHY registers and other
MII-related data structures.

The PHY drivers are modified to acquire and release the lock, as needed,
and assert the lock is held as a diagnostic aid.

The "usbnet" framework has had an overhaul of its internal locking
protocols to fit in with the media / mii changes, and the drivers adapted.

USB wifi drivers have been changed to provide their own adaptive mutex
to the ifmedia later via a new ieee80211_media_init_with_lock() function.
This is required because the USB drivers need an adaptive mutex.

Besised "usbnet", a few other drivers are converted: vmx, wm, ixgbe / ixv.

mcx also now calls ifmedia_init_with_lock() because it needs to also use
an adaptive mutex. The mcx driver still needs to be fully converted to
NET_MPSAFE.

- Simplify sc->mii_anegticks setting. Same as FreeBSD.

Don't set the default value not in the attach function. Instead, set the
default value (MII_ANEGTICKS) first in the beginning of the
mii_phy_add_media(). The function already has the code to change the value to
MII_ANEGTICKS_GIGE if it's gigabit capable.

- Remove extra pmf_device_register() call. It's done in mii_phy_add_media().

- Match a lot of Cicada and Vitesse devices correctly. This change also fixes
a bug that ciphy_fixup() didn't work.
- Match VSC8221, VSC8234 and VSC8641.

Call mii_phy_flowstatus() to show the flow setting.

- Indicat master mode if the negotiated result say so.
- KNF

branches: 1.34.4;
KNF. No functional change.

Fix a bug that the duplex of manual media setting may be wrong
when the IFM_GMASK bit other than IFM_[FH]DX is set. Same as a part of
FreeBSD r217413.

KNF. No functional change.

use a macro to tidy up the phydesc array initialization, from FreeBSD

Add CS8204, CS8244 VSC8211 and VSC8601 support from {Free,Open}BSD.

Change MII PHY read/write API from:

int (*mii_readreg_t)(device_t, int, int);
void (*mii_writereg_t)(device_t, int, int, int);
to:

int (*mii_readreg_t)(device_t, int, int, uint16_t *);
int (*mii_writereg_t)(device_t, int, int, uint16_t);

Now we can test if a read/write operation failed or not by the return value.

In 802.3 spec says that the PHY shall not respond to read/write transaction
to the unimplemented register(22.2.4.3). Detecting timeout can be used to
check whether a register is implemented or not (if the register conforms to
the spec). ukphy(4) can be used this for MII_MMDACR and MII_MMDAADR.

Note that I noticed that the following code do infinite loop in the
read/wirte function. If it accesses unimplemented PHY register, it will hang.
It should be fixed:

arm/at91/at91emac.c
arm/ep93xx/epe.c
arm/omap/omapl1x_emac.c
mips/ralink/ralink_eth.c
arch/powerpc/booke/dev/pq3etsec.c(read)
dev/cadence/if_cemac.c <- hkenken
dev/ic/lan9118.c


Tested with the following device:

axe+ukphy
axe+rgephy
axen+rgephy (tested by Andrius V)
wm+atphy
wm+ukphy
wm+igphy
wm+ihphy
wm+makphy
sk+makphy
sk+brgphy
sk+gentbi
msk+makphy
sip+icsphy
sip+ukphy
re+rgephy
bge+brgphy
bnx+brgphy
gsip+gphyter
rtk+rlphy
fxp+inphy (tested by Andrius V)
tlp+acphy
ex+exphy
epic+qsphy
vge+ciphy (tested by Andrius V)
vr+ukphy (tested by Andrius V)
vte+ukphy (tested by Andrius V)

Not tested (MAC):
arm:at91emac
arm:cemac
arm:epe
arm:geminigmac
arm:enet
arm:cpsw
arm:emac(omac)
arm:emac(sunxi)
arm:npe
evbppc:temac
macppc:bm
macppc:gm
mips:aumac
mips:ae
mips:cnmac
mips:reth
mips:sbmac
playstation2:smap
powerpc:tsec
powerpc:emac(ibm4xx)
sgimips:mec
sparc:be
sf
ne(ax88190, dl10019)
awge
ep
gem
hme
smsh
mtd
sm
age
alc
ale
bce
cas
et
jme
lii
nfe
pcn
ste
stge
tl
xi
aue
mue
smsc
udav
url

Not tested (PHY):
amhphy
bmtphy
dmphy
etphy
glxtphy
ikphy
iophy
lxtphy
nsphyter
pnaphy
rdcphy
sqphy
tlphy
tqphy
urlphy

CIPHY_MII_BMCR -> MII_BMCR in #ifdef foo.

The register bit definitions from register 0 to 15 in ciphyreg.h conform to
the 802.3 spec, so remove them and use mii.h's definition.
No functional change.

branches: 1.26.10; 1.26.16; 1.26.18;
KNF. Remove extra spaces. No functional change.

branches: 1.25.4;
IFM_FDX and IFM_HDX use different bit, so set IFM_HDX bit if it's not full
duplex. For many drivers, it recognize half duplex if IFM_FDX isn't set,
but not for others. Same as {Free|Open}BSD.

branches: 1.24.6;
Update link status quickly. This change is a part of mii_physubr.c rev.
1.77-1.78.

Update link status quickly. This change is a part of FreeBSD's mii_physubr.c
r158649.

ciphy is gigabit PHY, so use MII_ANEGTICKS_GIGE instead of MII_ANEGTICKS.

atphy, ciphy, ihphy and rgephy are gigabit PHY, so not only check IFM_AUTO
for autonegotiation but also check IFM_1000_T, too.

Fix a bug that the auto negotiation timer isn't cleard as we expected.
To not to forget clearing the timer and to reduce the code duplication,
clear mii_ticks in *mii_phy_auto().

branches: 1.19.12; 1.19.18; 1.19.22;
struct device * -> device_t, no functional changes intended.

struct cfdata * -> cfdata_t, no functional changes intended.

Change about 4500 of the K&R function definitions to ANSI ones.
There are still about 1600 left, but they have ',' or /* ... */
in the actual variable definitions - which my awk script doesn't handle.
There are also many that need () -> (void).
(The script does handle misordered arguments.)

branches: 1.16.4;
Reduce code duplication: most PHY drivers call mii_phy_add_media()
when they attach to the device tree, so call pmf_device_register(9)
once there instead of once in more than twenty drivers.

branches: 1.15.6; 1.15.8;
device_t/softc split for all mii(4) devices, and other related
cosmetic changes.

branches: 1.14.2; 1.14.4;
use aprint_*_dev and device_xname

branches: 1.13.10;
Merge jmcneill-pm branch.

branches: 1.12.4; 1.12.6;
machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h

branches: 1.11.8; 1.11.22; 1.11.24; 1.11.28;
Include <sys/device.h> for struct device.

- don't hardcode 5, 10 define and use MII_ANEGTICKS{,_GIGE}
- instead of != limit, use <= limit (conservative)

__unused removal on arguments; approved by core.

- sprinkle __unused on function decls.
- fix a couple of unused bugs
- no more -Wno-unused for i386

branches: 1.7.8; 1.7.10;
integrate kauth.

Use device_private().

branches: 1.5.2;
from OpenBSD:
We need to special case ciphy to ask for RGMII interface at 2.5V for nfe.
This makes Cicada/Vitesse PHYs attached to nfe work.
Many thanks to Bill Paul for diagnosing the problem and suggesting a solution.

branches: 1.4.4; 1.4.6; 1.4.8; 1.4.10;
Match all model numbers with both Cicada OUIs; this allows the proper PHY to
attach on my ZyXEL GN650-T.

branches: 1.3.2; 1.3.4; 1.3.6; 1.3.8;
Changes to match xxCICADA 8201B found in a VT6122.
Match at > 1 (10) to ensure we get ciphy instead of ukphy.
Use aprint(9) on attach.
Minor stylistic changes.
All from Tomokazu HARADA in PR kern/29705.

branches: 1.2.2;
nuke trailing whitespace

Add driver for 10/100/1000 copper PHY integrated on VIA Networking Technologies
VT6122 gigabit ethernet chips

From FreeBSD

fix spelling mistakes, mainly in comments and log messages.

The register bit definitions from register 0 to 15 in ciphyreg.h conform to
the 802.3 spec, so remove them and use mii.h's definition.
No functional change.

branches: 1.5.20; 1.5.26; 1.5.28;
No functional change:
- Fix typo.
- Remove trailing white spaces.
- Capitalize comments.
- Tabify.
- KNF.

branches: 1.4.106; 1.4.110; 1.4.126;
from OpenBSD:
We need to special case ciphy to ask for RGMII interface at 2.5V for nfe.
This makes Cicada/Vitesse PHYs attached to nfe work.
Many thanks to Bill Paul for diagnosing the problem and suggesting a solution.

branches: 1.3.4; 1.3.6; 1.3.8; 1.3.10;
merge ktrace-lwp.

branches: 1.2.2; 1.2.4; 1.2.6; 1.2.8; 1.2.10;
nuke trailing whitespace

Add driver for 10/100/1000 copper PHY integrated on VIA Networking Technologies
VT6122 gigabit ethernet chips

From FreeBSD

Rework the xxxVERBOSE option to share the common module-hook-based
verbose mechanism with MIIVERBOSE. This reduces some duplicated code
and allows us to once again permit auto-unload of MIIVERBOSE.

Change details:
* Update dev/devlist2h.awk to accomodate miidevs, including generation
of MII_STR_oui_model definitions and use of oui and model rather than
vendor and product. This also changes the compressed data in the
xxxdevs_data.h files to uint32_t (since mii oui's are up to 6 hex
digits long)
* Update a couple of phy drivers to use new calls to get verbose data
* Regen all of the xxxdevs{,_data}.h files (separate commit, coming
very soon)
* Update mii/mii_verbose.[ch] and mii/mii_physubr.c to use the various
DEV_VERBOSE_xxx macros
* Update the pci, usb, and hdaudio code as needed, to #include the
xxxdevs.h files (in order to get the proper printf format strings)
* Since dev/dev_verbose.c now uses non-literal printf format strings,
(to deal with the vendor/product vs oui/model issue), we need to
make sure it gets compiled with -Wno-error=format-nonliteral, even
in userland's libpci and librumpdev!
* Bump kernel version for the change in module interfaces

Welcome to 9.99.86!

XXX It might be useful in the future to extend the MII_STR_oui_model
XXX definitions to PCI as well (and perhaps USB and HDAUDIO). This
XXX would allow for a single centralized location for the products'
XXX descriptions, rather than being dispersed among individual
XXX drivers' xxx_match tables.

branches: 1.11.6;
mii_knowndevs[] is de facto const, define it as such.

This time for real. On my first try I did it in the generated
miidevs_data.h file.

Fix s/product/model/ search-and-destroy accident that affected license.

Back in 1998 thorpej@ adapted this from one of the several versions we
have in the tree but needed s/product/model/ in the text of the
script. Unfortunately that also affected the license text: "This
model includes software..." and "...or promote models".

branches: 1.9.4;
Fix compile error when MIIVERBOSE is set. Reported by Kurt Schreiner.

branches: 1.8.10;
remove my copyright.

branches: 1.7.44; 1.7.64;
Move TNF licenses to 2 clause form

branches: 1.6.70; 1.6.72; 1.6.74;
merge ktrace-lwp.

Don't leave blanks at the end of the "generated from:"
Also, don't print a space after "},"s at the end of line in one of the
files.

branches: 1.4.8; 1.4.10;
Fixes from PR#23177. Various lint/logic fixes:
Fix some non-initialized variables
close the output files when done
Redo the printing for RCS strings so they don't expand in the awk script too
Do proper tests for variables existance before accessing

Verified output from all scripts is identical to original versions

branches: 1.3.28;
Allow a verbose description to be printed with the option MIIVERBOSE.

branches: 1.2.12;
Assign copyright to TNF.

Near complete rewrite of the MII layer to present an interface compatible
with BSDI's MII layer to device drivers.

Define and implement a locking protocol for the ifmedia / mii layers:
- MP-safe drivers provide a mutex to ifmedia that is used to serialize
access to media-related structures / hardware regsiters. Converted
drivers use the new ifmedia_init_with_lock() function for this. The
new name is provided to ease the transition.
- Un-converted drivers continue to call ifmedia_init(), which will supply
a compatibility lock to be used instead. Several media-related entry
points must be aware of this compatibility lock, and are able to acquire
it recursively a limited number of times, if needed. This is a SPIN
mutex with priority IPL_NET.
- This same lock is used to serialize access to PHY registers and other
MII-related data structures.

The PHY drivers are modified to acquire and release the lock, as needed,
and assert the lock is held as a diagnostic aid.

The "usbnet" framework has had an overhaul of its internal locking
protocols to fit in with the media / mii changes, and the drivers adapted.

USB wifi drivers have been changed to provide their own adaptive mutex
to the ifmedia later via a new ieee80211_media_init_with_lock() function.
This is required because the USB drivers need an adaptive mutex.

Besised "usbnet", a few other drivers are converted: vmx, wm, ixgbe / ixv.

mcx also now calls ifmedia_init_with_lock() because it needs to also use
an adaptive mutex. The mcx driver still needs to be fully converted to
NET_MPSAFE.

unbreak the previous commit error

determine FDX flow control abilities

branches: 1.44.2;
- Simplify sc->mii_anegticks setting. Same as FreeBSD.

Don't set the default value not in the attach function. Instead, set the
default value (MII_ANEGTICKS) first in the beginning of the
mii_phy_add_media(). The function already has the code to change the value to
MII_ANEGTICKS_GIGE if it's gigabit capable.

- Remove extra pmf_device_register() call. It's done in mii_phy_add_media().

KNF. No functional change.

KNF. No functional change.

Add Davicom DM9161,DM9161A,DM9161B and DM9601 support.

Fix previos to not to duplicate DM9101 entry. One is DAVICOM and another
is xxDAVICOM.

use a macro to tidy up the phydesc array initialization, from FreeBSD

Fix typo in comment.

Change MII PHY read/write API from:

int (*mii_readreg_t)(device_t, int, int);
void (*mii_writereg_t)(device_t, int, int, int);
to:

int (*mii_readreg_t)(device_t, int, int, uint16_t *);
int (*mii_writereg_t)(device_t, int, int, uint16_t);

Now we can test if a read/write operation failed or not by the return value.

In 802.3 spec says that the PHY shall not respond to read/write transaction
to the unimplemented register(22.2.4.3). Detecting timeout can be used to
check whether a register is implemented or not (if the register conforms to
the spec). ukphy(4) can be used this for MII_MMDACR and MII_MMDAADR.

Note that I noticed that the following code do infinite loop in the
read/wirte function. If it accesses unimplemented PHY register, it will hang.
It should be fixed:

arm/at91/at91emac.c
arm/ep93xx/epe.c
arm/omap/omapl1x_emac.c
mips/ralink/ralink_eth.c
arch/powerpc/booke/dev/pq3etsec.c(read)
dev/cadence/if_cemac.c <- hkenken
dev/ic/lan9118.c


Tested with the following device:

axe+ukphy
axe+rgephy
axen+rgephy (tested by Andrius V)
wm+atphy
wm+ukphy
wm+igphy
wm+ihphy
wm+makphy
sk+makphy
sk+brgphy
sk+gentbi
msk+makphy
sip+icsphy
sip+ukphy
re+rgephy
bge+brgphy
bnx+brgphy
gsip+gphyter
rtk+rlphy
fxp+inphy (tested by Andrius V)
tlp+acphy
ex+exphy
epic+qsphy
vge+ciphy (tested by Andrius V)
vr+ukphy (tested by Andrius V)
vte+ukphy (tested by Andrius V)

Not tested (MAC):
arm:at91emac
arm:cemac
arm:epe
arm:geminigmac
arm:enet
arm:cpsw
arm:emac(omac)
arm:emac(sunxi)
arm:npe
evbppc:temac
macppc:bm
macppc:gm
mips:aumac
mips:ae
mips:cnmac
mips:reth
mips:sbmac
playstation2:smap
powerpc:tsec
powerpc:emac(ibm4xx)
sgimips:mec
sparc:be
sf
ne(ax88190, dl10019)
awge
ep
gem
hme
smsh
mtd
sm
age
alc
ale
bce
cas
et
jme
lii
nfe
pcn
ste
stge
tl
xi
aue
mue
smsc
udav
url

Not tested (PHY):
amhphy
bmtphy
dmphy
etphy
glxtphy
ikphy
iophy
lxtphy
nsphyter
pnaphy
rdcphy
sqphy
tlphy
tqphy
urlphy

branches: 1.36.16; 1.36.18;
KNF. Remove extra spaces. No functional change.

branches: 1.35.4;
IFM_FDX and IFM_HDX use different bit, so set IFM_HDX bit if it's not full
duplex. For many drivers, it recognize half duplex if IFM_FDX isn't set,
but not for others. Same as {Free|Open}BSD.

branches: 1.34.18; 1.34.22; 1.34.36;
Remove closes 3 & 4 from my licence. Lots of thanks to Soren Jacobsen
for the booring work !

Reduce code duplication: most PHY drivers call mii_phy_add_media()
when they attach to the device tree, so call pmf_device_register(9)
once there instead of once in more than twenty drivers.

branches: 1.32.6; 1.32.8;
device_t/softc split for all mii(4) devices, and other related
cosmetic changes.

Remove clause 3 and 4 from TNF licenses

branches: 1.30.2; 1.30.4;
use aprint_*_dev and device_xname

branches: 1.29.6;
Remove the device_is_active() check from each individual PHY's
service routine. Add a wrapper for PHY_SERVICE(), called phy_service(),
and check device_is_active() there. Make the mii_*() routines call
the PHY service routines thorugh phy_service() instead of PHY_SERVICE().

In mii_phy_resume(), restore a PHY's state after resetting it by
sending a MII_MEDIACHG command. This change makes bnx(4) resume
more reliably and more quickly; it should help other NICs to resume,
too, if they attach PHYs through MII.

branches: 1.28.2;
Merge jmcneill-pm branch.

branches: 1.27.22; 1.27.24; 1.27.30; 1.27.34; 1.27.36;
- don't hardcode 5, 10 define and use MII_ANEGTICKS{,_GIGE}
- instead of != limit, use <= limit (conservative)

__unused removal on arguments; approved by core.

- sprinkle __unused on function decls.
- fix a couple of unused bugs
- no more -Wno-unused for i386

branches: 1.24.8; 1.24.10;
Use device_private().

branches: 1.23.2; 1.23.4; 1.23.6;
Use device_is_active() rather than testing dv_flags for DVF_ACTIVE
directly.

branches: 1.22.2; 1.22.4; 1.22.6;
merge ktrace-lwp.

Add DAVICOM DM9102.

branches: 1.20.2; 1.20.4;
nuke trailing whitespace

branches: 1.19.4; 1.19.6;
Make use of static.

branches: 1.18.2;
Use aprint*().

Add trailing ; to CFATTACH_DECL.

Use CFATTACH_DECL().

Declare all cfattach structures const.

Don't include <malloc.h> - no memory management style functions are use
in any of the MII drivers.

branches: 1.13.2;
Get flags passed down to PHY drivers correctly. This was done on
an adhoc basis in a couple of PHY drivers, this fixes it more generally.

Per a discussion w/ Cliff Neighbors <cliff@allegronetworks.com>.

add RCSID

ANSI'ify.

branches: 1.10.2;
Make PHY matching all table-driven.

Make the number of ticks between auto-negotiation tries PHY-specific,
and default every PHY to 5 seconds (what we used before). If we find
Gig-E media on a PHY, bump it to 10 seconds, since it can take 5
seconds just to negotiate a Gig-E link.

branches: 1.8.2; 1.8.4;
Restructure the PHY entry points to use a structure of entry points
instead of discrete function pointers, and extend this to include
a "reset" entry point. Make sure any PHY-specific reset routine is
always used, and provide one for the LXT-970 which disables MII
interrupts (as is done for a few other PHYs we have drivers for).

branches: 1.7.4;
Make to match also MII_OUI_DAVICOM.
(This seems to work, but any GENERIC does not have dmphy...)

Factor out the tick handling code into a common function, and send
rt_ifmsg's when the link speed or link status changes.

Let mii_phy_setmedia() handle the IFM_AUTO case, too.

Make PHY drivers provide a pointer to their status routine, and add
a generic mii_phy_status() that calls back into the PHY. It doesn't
do anything interesting yet, but it will soon.

Don't dry to diving MIIF_NOISOLATE in the PHY drivers. Instead, pass
flags down from the parent to child vi mii_attach().

Bring some order to the chaos which was the MII code function naming
"conventions".

Driver for the Davicom DM9101 10/100 Ethernet PHY. This PHY is found
on several inexpensive 10/100 Ethernet boards (e.g. D-Link DFE-530TX,
Winbond 10/100 boards), and is also the built-in PHY on the DM9102
10/100 Ethernet chip.

Remove clause 3 and 4 from TNF licenses

branches: 1.1.6; 1.1.128; 1.1.130; 1.1.132;
Driver for the Davicom DM9101 10/100 Ethernet PHY. This PHY is found
on several inexpensive 10/100 Ethernet boards (e.g. D-Link DFE-530TX,
Winbond 10/100 boards), and is also the built-in PHY on the DM9102
10/100 Ethernet chip.

- Backout some changes done in rev. 1.44. OK'd by jdolecek@:
- Remove e1000phyreg.h and use makphyreg.h again.
- Remove Energy detect stuff. Leave it by default.
- Remove MDI crossover configuraton. Leave it by default. I think it's OK to
add new API to change the mode.
- PHY_RESET() is used to commit some changes, so I think it's not good to
always clear BMCR_AUTOEN before reset in makphy_reset().
- It's not required to read MII_100T2SR twice because this register has no
any bit which is latched.
- Use mii_phy_reset() instead of extracted code.
- Don't set PSCR_CRS_ON_TX on newer. Those chips have no this bit.
- Add some "XXX FIXME" comment. Non GMII mode uses different page except
very old chip. Some bits are at the same location but others are not.
- Remove obsolete comment.
- s/MII_MAKPHY_/MAKPHY_/

XXX rev. 1.44 changed the setting of mii_media_active when the interface is
set to other than auto. I suspect the intention of the change is to reflect
the link up/down status. But, it didn't work because non auto setting makes
PSSR_LINK bit alwasy set "without modifing some registers".
This commit won't fix the change and it'll be fixed in the next commit.

branches: 1.2.2; 1.2.4;
- All of Marvell PHY's registers from addr 0x0 to 0xf conforms IEEE 802.3
specification, so those register definitions are not required. Use mii.h's
definitions. Note that E1000_ER_PAR_DETECT_FAULT should be 0x0010 instead of
0x0100(typo).
- Restore code for BMCR_ISO which was removed in makphy.c rev. 1.44
- Whitespace fix.

switch to using OpenBSD eehpy(4) code to drive maphy(4), synchronizing support
for several special PHY conditions, particularly:
- Properly re-initialise the PHY upon resume
- Store next page in the Link Partner Next Page register for compatibility
with 802.3ab on 88E3016 PHYs. Fixes some autonegotiation problems on msk(4)
- Make 88E3016 actually work
- Make sure page 0 is selected when we initialize the PHY. Fixes problems
with the eephy(4) that attaches to nfe(4) on machines like the Sun Ultra 40.
(we had condition for this, now the page 0 is selected for any PHY type)
- Disable fiber/copper auto-selection on the 88E1111 if it is in RGMII mode, to
work around the fact that the onboard PHYs attached to nfe(4) on the Sun
X4100 M2 have fiber/copper auto-selection enabled even though the interfaces
are clearly copper-only

make sure to also add appropriate licenses, since basically nothing
really significant remains out of previous code

use FreeBSD <dev/mii/e1000reg.h> for register definitions as a base instead
of OpenBSD <dev/mii/eephyreg.h>, since it has some extra definitions for some
3016 bits, but add the several extra bits from OpenBSD needed by the code;
removed no longed used <dev/mii/makphyreg.h>

tested with PHY 88E1111, there no particular change observed - the
link status works as it did before, just now it does media nego
even before the interface is up

should however fix 88E3016 support and hence PR kern/49270 and PR kern/53301

Define and implement a locking protocol for the ifmedia / mii layers:
- MP-safe drivers provide a mutex to ifmedia that is used to serialize
access to media-related structures / hardware regsiters. Converted
drivers use the new ifmedia_init_with_lock() function for this. The
new name is provided to ease the transition.
- Un-converted drivers continue to call ifmedia_init(), which will supply
a compatibility lock to be used instead. Several media-related entry
points must be aware of this compatibility lock, and are able to acquire
it recursively a limited number of times, if needed. This is a SPIN
mutex with priority IPL_NET.
- This same lock is used to serialize access to PHY registers and other
MII-related data structures.

The PHY drivers are modified to acquire and release the lock, as needed,
and assert the lock is held as a diagnostic aid.

The "usbnet" framework has had an overhaul of its internal locking
protocols to fit in with the media / mii changes, and the drivers adapted.

USB wifi drivers have been changed to provide their own adaptive mutex
to the ifmedia later via a new ieee80211_media_init_with_lock() function.
This is required because the USB drivers need an adaptive mutex.

Besised "usbnet", a few other drivers are converted: vmx, wm, ixgbe / ixv.

mcx also now calls ifmedia_init_with_lock() because it needs to also use
an adaptive mutex. The mcx driver still needs to be fully converted to
NET_MPSAFE.

Move PHY_RESET() in the media change path from if_et.c to etphy.c.

I don't know if the reset is required or not. I tested some media
transitions without the reset and it worked. It might be OK to remove
but Linux does it only in et1011c_config_aneg(). So move the reset to
etphy_service(,,MII_MEDIACHG)'s autonego case. This change is also
required for future locking changes.

branches: 1.7.2;
- Simplify sc->mii_anegticks setting. Same as FreeBSD.

Don't set the default value not in the attach function. Instead, set the
default value (MII_ANEGTICKS) first in the beginning of the
mii_phy_add_media(). The function already has the code to change the value to
MII_ANEGTICKS_GIGE if it's gigabit capable.

- Remove extra pmf_device_register() call. It's done in mii_phy_add_media().

From FreeBSD:
- Support ET1011.
- Use mii_phy_flowstatus() to reflect flow status from negotiated result.

No functional change:
- Use static.
- KNF.

branches: 1.4.4;
KNF. No functional change.

use a macro to tidy up the phydesc array initialization, from FreeBSD

Change MII PHY read/write API from:

int (*mii_readreg_t)(device_t, int, int);
void (*mii_writereg_t)(device_t, int, int, int);
to:

int (*mii_readreg_t)(device_t, int, int, uint16_t *);
int (*mii_writereg_t)(device_t, int, int, uint16_t);

Now we can test if a read/write operation failed or not by the return value.

In 802.3 spec says that the PHY shall not respond to read/write transaction
to the unimplemented register(22.2.4.3). Detecting timeout can be used to
check whether a register is implemented or not (if the register conforms to
the spec). ukphy(4) can be used this for MII_MMDACR and MII_MMDAADR.

Note that I noticed that the following code do infinite loop in the
read/wirte function. If it accesses unimplemented PHY register, it will hang.
It should be fixed:

arm/at91/at91emac.c
arm/ep93xx/epe.c
arm/omap/omapl1x_emac.c
mips/ralink/ralink_eth.c
arch/powerpc/booke/dev/pq3etsec.c(read)
dev/cadence/if_cemac.c <- hkenken
dev/ic/lan9118.c


Tested with the following device:

axe+ukphy
axe+rgephy
axen+rgephy (tested by Andrius V)
wm+atphy
wm+ukphy
wm+igphy
wm+ihphy
wm+makphy
sk+makphy
sk+brgphy
sk+gentbi
msk+makphy
sip+icsphy
sip+ukphy
re+rgephy
bge+brgphy
bnx+brgphy
gsip+gphyter
rtk+rlphy
fxp+inphy (tested by Andrius V)
tlp+acphy
ex+exphy
epic+qsphy
vge+ciphy (tested by Andrius V)
vr+ukphy (tested by Andrius V)
vte+ukphy (tested by Andrius V)

Not tested (MAC):
arm:at91emac
arm:cemac
arm:epe
arm:geminigmac
arm:enet
arm:cpsw
arm:emac(omac)
arm:emac(sunxi)
arm:npe
evbppc:temac
macppc:bm
macppc:gm
mips:aumac
mips:ae
mips:cnmac
mips:reth
mips:sbmac
playstation2:smap
powerpc:tsec
powerpc:emac(ibm4xx)
sgimips:mec
sparc:be
sf
ne(ax88190, dl10019)
awge
ep
gem
hme
smsh
mtd
sm
age
alc
ale
bce
cas
et
jme
lii
nfe
pcn
ste
stge
tl
xi
aue
mue
smsc
udav
url

Not tested (PHY):
amhphy
bmtphy
dmphy
etphy
glxtphy
ikphy
iophy
lxtphy
nsphyter
pnaphy
rdcphy
sqphy
tlphy
tqphy
urlphy

branches: 1.1.6; 1.1.54; 1.1.60; 1.1.62;
PR/39094 - Kaspar Brand -- Add et (Agere ET1310/ET1301) network driver

The et(4) driver supports PCI Express Ethernet adapters based on
the Agere/LSI ET1310/ET1301 integrated MAC/PHY.

The et(4) driver was written by Sepherosa Ziehau for DragonFlyBSD,
ported to OpenBSD by Jonathan Gray and subsequently ported to NetBSD
by Kaspar Brand.

Define and implement a locking protocol for the ifmedia / mii layers:
- MP-safe drivers provide a mutex to ifmedia that is used to serialize
access to media-related structures / hardware regsiters. Converted
drivers use the new ifmedia_init_with_lock() function for this. The
new name is provided to ease the transition.
- Un-converted drivers continue to call ifmedia_init(), which will supply
a compatibility lock to be used instead. Several media-related entry
points must be aware of this compatibility lock, and are able to acquire
it recursively a limited number of times, if needed. This is a SPIN
mutex with priority IPL_NET.
- This same lock is used to serialize access to PHY registers and other
MII-related data structures.

The PHY drivers are modified to acquire and release the lock, as needed,
and assert the lock is held as a diagnostic aid.

The "usbnet" framework has had an overhaul of its internal locking
protocols to fit in with the media / mii changes, and the drivers adapted.

USB wifi drivers have been changed to provide their own adaptive mutex
to the ifmedia later via a new ieee80211_media_init_with_lock() function.
This is required because the USB drivers need an adaptive mutex.

Besised "usbnet", a few other drivers are converted: vmx, wm, ixgbe / ixv.

mcx also now calls ifmedia_init_with_lock() because it needs to also use
an adaptive mutex. The mcx driver still needs to be fully converted to
NET_MPSAFE.

- Simplify sc->mii_anegticks setting. Same as FreeBSD.

Don't set the default value not in the attach function. Instead, set the
default value (MII_ANEGTICKS) first in the beginning of the
mii_phy_add_media(). The function already has the code to change the value to
MII_ANEGTICKS_GIGE if it's gigabit capable.

- Remove extra pmf_device_register() call. It's done in mii_phy_add_media().

KNF. No functional change.

Change MII PHY read/write API from:

int (*mii_readreg_t)(device_t, int, int);
void (*mii_writereg_t)(device_t, int, int, int);
to:

int (*mii_readreg_t)(device_t, int, int, uint16_t *);
int (*mii_writereg_t)(device_t, int, int, uint16_t);

Now we can test if a read/write operation failed or not by the return value.

In 802.3 spec says that the PHY shall not respond to read/write transaction
to the unimplemented register(22.2.4.3). Detecting timeout can be used to
check whether a register is implemented or not (if the register conforms to
the spec). ukphy(4) can be used this for MII_MMDACR and MII_MMDAADR.

Note that I noticed that the following code do infinite loop in the
read/wirte function. If it accesses unimplemented PHY register, it will hang.
It should be fixed:

arm/at91/at91emac.c
arm/ep93xx/epe.c
arm/omap/omapl1x_emac.c
mips/ralink/ralink_eth.c
arch/powerpc/booke/dev/pq3etsec.c(read)
dev/cadence/if_cemac.c <- hkenken
dev/ic/lan9118.c


Tested with the following device:

axe+ukphy
axe+rgephy
axen+rgephy (tested by Andrius V)
wm+atphy
wm+ukphy
wm+igphy
wm+ihphy
wm+makphy
sk+makphy
sk+brgphy
sk+gentbi
msk+makphy
sip+icsphy
sip+ukphy
re+rgephy
bge+brgphy
bnx+brgphy
gsip+gphyter
rtk+rlphy
fxp+inphy (tested by Andrius V)
tlp+acphy
ex+exphy
epic+qsphy
vge+ciphy (tested by Andrius V)
vr+ukphy (tested by Andrius V)
vte+ukphy (tested by Andrius V)

Not tested (MAC):
arm:at91emac
arm:cemac
arm:epe
arm:geminigmac
arm:enet
arm:cpsw
arm:emac(omac)
arm:emac(sunxi)
arm:npe
evbppc:temac
macppc:bm
macppc:gm
mips:aumac
mips:ae
mips:cnmac
mips:reth
mips:sbmac
playstation2:smap
powerpc:tsec
powerpc:emac(ibm4xx)
sgimips:mec
sparc:be
sf
ne(ax88190, dl10019)
awge
ep
gem
hme
smsh
mtd
sm
age
alc
ale
bce
cas
et
jme
lii
nfe
pcn
ste
stge
tl
xi
aue
mue
smsc
udav
url

Not tested (PHY):
amhphy
bmtphy
dmphy
etphy
glxtphy
ikphy
iophy
lxtphy
nsphyter
pnaphy
rdcphy
sqphy
tlphy
tqphy
urlphy

branches: 1.54.16; 1.54.18;
KNF. Remove extra spaces. No functional change.

branches: 1.53.10;
Don't "return" but "break" in MII_TICK when the interface isn't doing
autonegotiaton or the link is up to update the status.

branches: 1.52.12; 1.52.22;
Remove closes 3 & 4 from my licence. Lots of thanks to Soren Jacobsen
for the booring work !

Reduce code duplication: most PHY drivers call mii_phy_add_media()
when they attach to the device tree, so call pmf_device_register(9)
once there instead of once in more than twenty drivers.

branches: 1.50.6; 1.50.8;
device_t/softc split for all mii(4) devices, and other related
cosmetic changes.

Remove clause 3 and 4 from TNF licenses

branches: 1.48.2; 1.48.4;
pretty up if_ex: pmf'ify, device split, and improve printing a bit too.

Reviewed and function-tested on if_ex_cardbus by cube (thanks).
Function-tested on if_ex_pci by myself.

use aprint_*_dev and device_xname

branches: 1.46.6;
Remove the device_is_active() check from each individual PHY's
service routine. Add a wrapper for PHY_SERVICE(), called phy_service(),
and check device_is_active() there. Make the mii_*() routines call
the PHY service routines thorugh phy_service() instead of PHY_SERVICE().

In mii_phy_resume(), restore a PHY's state after resetting it by
sending a MII_MEDIACHG command. This change makes bnx(4) resume
more reliably and more quickly; it should help other NICs to resume,
too, if they attach PHYs through MII.

branches: 1.45.2;
Merge jmcneill-pm branch.

branches: 1.44.22; 1.44.24; 1.44.30; 1.44.34; 1.44.36;
- don't hardcode 5, 10 define and use MII_ANEGTICKS{,_GIGE}
- instead of != limit, use <= limit (conservative)

__unused removal on arguments; approved by core.

- sprinkle __unused on function decls.
- fix a couple of unused bugs
- no more -Wno-unused for i386

branches: 1.41.8; 1.41.10;
Use device_private().

branches: 1.40.2; 1.40.4; 1.40.6;
Reduce previous (!is(...) != 0) to (!is(...)).

Use device_is_a().

Use device_is_active() rather than testing dv_flags for DVF_ACTIVE
directly.

branches: 1.37.2; 1.37.4; 1.37.6;
merge ktrace-lwp.

branches: 1.36.12;
Make use of static.

branches: 1.35.2;
Use aprint*().

Add trailing ; to CFATTACH_DECL.

Use CFATTACH_DECL().

Declare all cfattach structures const.

Rather than referencing the cfdriver directly in the cfdata entries,
instead use a string naming the driver. The cfdriver is then looked
up in a list which is built at run-time.

Don't include <malloc.h> - no memory management style functions are use
in any of the MII drivers.

branches: 1.29.2;
Get flags passed down to PHY drivers correctly. This was done on
an adhoc basis in a couple of PHY drivers, this fixes it more generally.

Per a discussion w/ Cliff Neighbors <cliff@allegronetworks.com>.

add RCSID

ANSI'ify.

branches: 1.26.2;
Make the number of ticks between auto-negotiation tries PHY-specific,
and default every PHY to 5 seconds (what we used before). If we find
Gig-E media on a PHY, bump it to 10 seconds, since it can take 5
seconds just to negotiate a Gig-E link.

branches: 1.25.2;
Restructure the PHY entry points to use a structure of entry points
instead of discrete function pointers, and extend this to include
a "reset" entry point. Make sure any PHY-specific reset routine is
always used, and provide one for the LXT-970 which disables MII
interrupts (as is done for a few other PHYs we have drivers for).

branches: 1.24.4;
Factor out the tick handling code into a common function, and send
rt_ifmsg's when the link speed or link status changes.

Let mii_phy_setmedia() handle the IFM_AUTO case, too.

Make PHY drivers provide a pointer to their status routine, and add
a generic mii_phy_status() that calls back into the PHY. It doesn't
do anything interesting yet, but it will soon.

Don't dry to diving MIIF_NOISOLATE in the PHY drivers. Instead, pass
flags down from the parent to child vi mii_attach().

Bring some order to the chaos which was the MII code function naming
"conventions".

Add detachment support to the MII layer.

Add mii_down(), which is used by MAC drivers to inform PHYs that the
interface is now down. PHYs use this to cancel pending asynchronous
operations.

Clean up the code that adds media a little, and make media selection
table-driven in preparation for some other changes to be made.

branches: 1.16.2; 1.16.4; 1.16.8;
Implement asynchronous autonegotiation when driven by the MII tick
(one-second clock). Prevents .5s delays every 5 seconds when the interface
is up but there is no link. Fixes PR 7361.

branches: 1.15.6;
Place the essentially common "ticks" and "active" members into the
mii_softc (generic phy goo), and just switch all of the PHY drivers
(except tlphy, which really does have special stuff) to use an mii_softc
instead of a private one.

Routines common between the Uknown PHY driver (forthcoming) and other
PHY drivers. This file has ukphy_status(), which is what PHY drivers
should use if media detection is done by decoding NWay.

Return a higher-priority match than 1. These are specific drivers. This
allows for a "generic unknown PHY" driver to be implemented.

Factor out the common reset code. Use the NOISOLATE flags as appropriate
to prevent a PHY from being isolated in the event the MAC can't deal with
this. (3Com 3c905B-TX, and Intel i82557).

Factor out some common code from all the PHY drivers (autonegotiation, for
now.)

Define and use generic PHY read/write reg macros.

3Com PHY can't be isolated, so don't allow IFM_NONE.

When doing a media change service request, use the media word from the
current ifmedia_entry, not from the user-supplied media word. The
user supplied media word may not necessarily match e.g. instance (if
the parent MAC driver is intentionally ignoring instance if its expecting
multiple PHYs with non-overlapping media, e.g. TI ThunderLAN) the media
word we are actually switching to.

Since PHY drivers use `instance' to determine if they should isolate
themselves, the ThunderLAN PHY was sometimes being incorrectly isolated
when in fact the user attempted to select that PHY (for e.g. BNC operation).

Don't print the revision of the 3c905B-TX's internal PHY; it's meaningless.

Make sure the parent is an `ex'. The 3c905B-TX's internal PHY isn't
the only one w/ non-existent ID registers. SIGH.

Add ETHER|NONE to default capabilities, or we'll panic when attaching
the forthcoming ex driver.

Make a variable name reflect a register name.

If autonegotiation is enabled in the BMSR, and the BMCR doens't claim
that it has completed, report "none" as the active media type.

Ugh, don't print out "autonegotiation failed to complete" if it did.
We get these once a second if we're in auto mode, the interface is up,
and there's no carrier.

Driver for the 3Com internal PHY, found on 3c905B-TX cards (and maybe
some 3c905-TX?). Partially from Frank van der Linden.

Add driver for Motorcomm YT8511 GbE PHY

Add ukphy_subr flag to ipgphy to make kernel compilable with ipgphy and
without ukphy. Pointed out by Hauke.

Port SMSC LAN87xx 10/100 Ethernet PHY driver from FreeBSD with some cleanup
and IFM_NONE support.

Add jmphy(4) from OpenBSD.

Add support IC Plus IP1000* PHYs. PR/kern 42314 first reported by
Tomokazu HARADA and patch provided by Andrius V.

branches: 1.50.20; 1.50.26;
centralize the dialout/call unit macros.

Add Micrel PHY (KSZ9021RN)

The new driver micphy is almost same as ukphy except that
micphy has a fixup for cpsw; a PHY with cpsw has to adjust
RGMII signal timing.

Reviewed by christos@

branches: 1.48.4; 1.48.14; 1.48.18;
Add a driver for the RDC Semiconductor RDC R6040 10/100 Ethernet controller,
as found in the vortex86 SoCs (http://www.vortex86dx.com).
Ported from freebsd.
Not added to amd64's GENERIC because this CPU is 32bit only.

thanks to DM&P Electronics, Inc for providing documentation and sample
devices for this work.

branches: 1.47.2; 1.47.4;
PR/44080: Grégoire Sutre: Support for the i82577 ethernet PHY

PR/39094 - Kaspar Brand -- Add et (Agere ET1310/ET1301) network driver

The et(4) driver supports PCI Express Ethernet adapters based on
the Agere/LSI ET1310/ET1301 integrated MAC/PHY.

The et(4) driver was written by Sepherosa Ziehau for DragonFlyBSD,
ported to OpenBSD by Jonathan Gray and subsequently ported to NetBSD
by Kaspar Brand.

Extract MIIVERBOSE into a kernel module. The module can be builtin
by defining 'options MIIVERBOSE' in the kernel config file (no change
from current behavior), or it can be loaded at boot time on those
architectures that support the boot loader's "load" command.

branches: 1.44.2; 1.44.4;
Oooops. Back out some of previous my commit.

Some fixes and enhancements:

Both if_bge* and brgphy.[ch]:
Check the parent's MAC and use the quirk code for the bug like OpenBSD and
FreeBSD. Some bugs can't identify the PHY ID. For example, 5704 Ax has the
ADC bug, but 5704 A3 and 5704 B0 have the same PHY ID and revision. Add
BGE_PHY_CRC_BUG, BGE_PHY_ADC_BUG, BGE_PHY_5704_A0_BUG, BGE_PHY_JITTER_BUG,
BGE_PHY_ADJUST_TRIM and BGE_PHY_BER_BUG for this flag. Some of the DSP
patches are newly taken from OpenBSD and FreeBSD.

if_bge*:
Remove duplicated BGE_CHIPID_BCM5714_A0 entry in the known CHIPID table.
Fix obsolete comments.

brgphy.[ch]:
Add some PHY IDs.

TODO:
Add more three quirk code into bge and brgphy (the brgphy side's are
currently #if0'ed).
Add support for bnx into brgphy (currently #if0'ed)

branches: 1.42.2;
Driver for Attansic Technology F1 10/100/1000 Ethernet PHY.

Written by Pyun YongHyeon for FreeBSD, ported to OpenBSD by Kevin Lo, ported to
NetBSD by me.

branches: 1.41.6; 1.41.10; 1.41.18; 1.41.20; 1.41.26;
Fix XEN2_DOMU (and amd64?) builds: move ether_mediastatus(),
ether_mediachange() to their own module that we compile only if
the kernel configuration demands support for both MII buses and
ethernet. Thanks to Tom Spindler for suggesting that these routines
move to dev/mii/.

branches: 1.40.18; 1.40.24; 1.40.30;
Enable mvphy(4)

branches: 1.39.4;
Add support for the Intel 80003 Gigabit Ethernet controller (found e.g. in
newer server chipsets) to wm(4), from the FreeBSD em(4) driver.
While there, add a few other Intel Ethernet controller that should work as
is.
Properly update the RX error and TX collision counters.
Add ikphy(4), a driver for the Intel i82563 Kumeran 10/100/1000 Ethernet PHYs

branches: 1.38.18; 1.38.20;
Add rlphy(4) (Realtek 8139/8201L PHY) from OpenBSD.