CVS log for src/sys/dev/mii/nsphy.c

Up to [cvs.NetBSD.org] / src / sys / dev / mii

Request diff between arbitrary revisions

Keyword substitution: kv
Default branch: MAIN

Merge changes from current as of 20200406

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().

Sync with HEAD

 KNF. No functional change.

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

Sync with HEAD

 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

update from HEAD

Sync with HEAD

Do not deref a NULL pointer if no current media has been selected.
This error condition does not happen with properly working hardware, but it is
no good reason for a kernel panic either.

Sync with HEAD

KNF. Remove extra spaces. No functional change.

Pull up following revision(s) (requested by msaitoh in ticket #1189):
	sys/dev/pci/if_wm.c: revision 1.270
	sys/dev/mii/inphy.c: revision 1.53
	sys/dev/mii/glxtphy.c: revision 1.24
	sys/dev/mii/tlphy.c: revision 1.62
	sys/dev/mii/iophy.c: revision 1.37
	sys/dev/mii/brgphy.c: revision 1.70
	sys/dev/mii/ihphy.c: revision 1.8
	sys/dev/mii/bmtphy.c: revision 1.31
	sys/dev/mii/urlphy.c: revision 1.30
	sys/dev/mii/makphy.c: revision 1.40
	sys/dev/mii/qsphy.c: revision 1.48
	sys/dev/mii/igphy.c: revision 1.23
	sys/dev/mii/nsphy.c: revision 1.58
	sys/dev/mii/mvphy.c: revision 1.10
	sys/dev/pci/if_txp.c: revision 1.41
	sys/dev/mii/nsphy.c: revision 1.59
	sys/dev/mii/rlphy.c: revision 1.28
	sys/dev/mii/icsphy.c: revision 1.49
	sys/dev/mii/rlphy.c: revision 1.29
	sys/dev/mii/lxtphy.c: revision 1.49
	sys/dev/mii/ciphyreg.h: revision 1.5
	sys/dev/mii/nsphyter.c: revision 1.38
	sys/dev/mii/sqphy.c: revision 1.50
	sys/dev/mii/gentbi.c: revision 1.26
	sys/dev/mii/gentbi.c: revision 1.27
	sys/dev/mii/tqphy.c: revision 1.39
	sys/dev/mii/ikphy.c: revision 1.10
	sys/dev/mii/dmphy.c: revision 1.35
	sys/dev/mii/amhphy.c: revision 1.20
	sys/dev/mii/acphy.c: revision 1.24
	sys/dev/mii/ciphy.c: revision 1.25
	sys/dev/mii/brgphyreg.h: revision 1.8
	sys/dev/mii/ukphy_subr.c: revision 1.12
	sys/dev/ic/rtl80x9.c: revision 1.16
	sys/arch/mips/adm5120/dev/if_admsw.c: revision 1.12
	sys/dev/pci/if_kse.c: revision 1.28
	sys/dev/mii/ukphy_subr.c: revision 1.13
	sys/dev/mii/mii.h: revision 1.18
	sys/dev/mii/gphyter.c: revision 1.29
No functional change:
- Fix typo.
- Remove trailing white spaces.
- Capitalize comments.
- Tabify.
- KNF.
 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.

Rebase to HEAD as of a few days ago.

Rebase.

 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:
- Fix typo.
- Remove trailing white spaces.
- Capitalize comments.
- Tabify.
- KNF.

sync to netbsd-5

sync with head

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

Sync with HEAD.

Commit is split, to avoid a "too many arguments" protocol error.

sync with head.

Pull up following revision(s) (requested by cegger in ticket #474):
	sys/dev/mii/inphy.c: revision 1.51
	sys/dev/mii/iophy.c: revision 1.35
	sys/dev/mii/nsphy.c: revision 1.56
	sys/dev/mii/rlphy.c: revision 1.25
	sys/dev/mii/ukphy_subr.c: revision 1.11
	sys/dev/pci/if_txp.c: revision 1.28
fix media priorities:
IEEE 802.3 Annex 28B.3 specifies the following relative
priorities of the technologies supported by
802.3 Selector Field value:
1000BASE-T full duplex
1000BASE-T
100BASE-T2 full duplex
100BASE-TX full duplex
100BASE-T2
100BASE-T4
100BASE-TX
10BASE-T full duplex
10BAST-T
Our drivers give 100BASE-T4 a higher priority than
100BASE-TX full duplex.
Fix this. This patch is based on changes in FreeBSD and OpenBSD.
Patch presented on tech-kern and tech-net:
http://mail-index.netbsd.org/tech-kern/2009/02/15/msg004397.html
http://mail-index.netbsd.org/tech-net/2009/02/15/msg001064.html
got no comments, no objections.

Sync with HEAD.

fix media priorities:

IEEE 802.3 Annex 28B.3 specifies the following relative
priorities of the technologies supported by
802.3 Selector Field value:

1000BASE-T full duplex
1000BASE-T
100BASE-T2 full duplex
100BASE-TX full duplex
100BASE-T2
100BASE-T4
100BASE-TX
10BASE-T full duplex
10BAST-T

Our drivers give 100BASE-T4 a higher priority than
100BASE-TX full duplex.
Fix this. This patch is based on changes in FreeBSD and OpenBSD.

Patch presented on tech-kern and tech-net:
http://mail-index.netbsd.org/tech-kern/2009/02/15/msg004397.html
http://mail-index.netbsd.org/tech-net/2009/02/15/msg001064.html

got no comments, no objections.

Sync with HEAD.

Sync with HEAD.

Update haad-dm branch to haad-dm-base2.

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.

Sync with HEAD.

sync with head.

sync with head.

device_t/softc split for all mii(4) devices, and other related
cosmetic changes.

Remove clause 3 and 4 from TNF licenses

use aprint_*_dev and device_xname

Sync with HEAD.

sync with head

sync with HEAD

Sync with HEAD

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.

Sync with head.

sync with head.

Merge jmcneill-pm branch.

Rename pnp(9) -> pmf(9), as requested by many.

Register MII PHY drives with the PM framework. Nothing to do on suspend,
just reset the PHY on resume.

sync with head.

sync with head.

Sync with head.

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

__unused removal on arguments; approved by core.

sync with head

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

sync with head

sync with head.

Sync with head.

sync with head.

sync with head.

Merge 2006-03-31 NetBSD-current into the "peter-altq" branch.

Use device_private().

sync with head.

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

merge ktrace-lwp.

sync with -current

sync with head.  xen and whitespace.  xen part is not finished.

Sync with HEAD.

Hi Perry!

nuke trailing whitespace

Fix the sync with head I botched.

Sync with HEAD.

Sync with HEAD.

Make use of static.

Sync with HEAD

Fix some typos. From Tom Cosgrove via jmc@openbsd.

Implement a separate nsphy_reset().  There are two reasons for this:

	1) This PHY can take an inordinate amount of time to reset if
	   media is attached.  Under fairly normal circumstances, up
	   to near one second.  This is because it appears to go through
	   an implicit autonegotiation cycle as part of the reset.

	2) During reset and autonegotiation, the BMCR will clear the reset
	   bit before the process is complete.  It will return 0 until the
	   process is complete and it's safe to access the PHY again.

This gets the on-board pcnet + nsphy ethernet working properly for me on
an IBM PC Server/325.  Fixes PR/16346.

Use aprint*().

Catch up to -current.

sync kqueue with -current; this includes merge of gehenna-devsw branch,
merge of i386 MP branch, and part of autoconf rototil work

Add trailing ; to CFATTACH_DECL.

Use CFATTACH_DECL().

Declare all cfattach structures const.

sync kqueue branch with HEAD

Catch up to -current.

catch up with -current.

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

catch up with -current on kqueue branch

Catch up to -current.
(CVS: It's not just a program. It's an adventure!)

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>.

Sync kqueue branch with -current.

Catch up to -current.

add RCSID

Catch up to -current.

Update the kqueue branch to HEAD.

ANSI'ify.

Catch up to -current.

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.

Sync with HEAD

Catch up with -current.

update for miidevs changes

Update thorpej_scsipi to -current as of a month ago
A i386 GENERIC kernel compiles without the siop, ahc and bha drivers
(will be updated later). i386 IDE/ATAPI and ncr work, as well as
sparc/esp_sbus. alpha should work as well (untested yet).
siop, ahc and bha will be updated once I've updated the branch to current
-current, as well as machine-dependant code.

Pull up revision 1.28 (requested by martin):
  Fix autonegotiation problem with some fxp cards.  Fixes PR#10550.

Pullup revision 1.28 (approved by Jason Thorpe)

  Fix to autonegotiation problem with some fxp cards. Fix found and reported
  by Erich T. Enke et al. on current-users.

Fixes PR 10550.

Fix to autonegotiation problem with some fxp cards. Fix found and reported
by Erich T. Enke et al. on current-users. This fixes PR 10550.

Update from trunk:
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).

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).

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.

Pull up to last week's -current.

Sync with -current

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.

Update from trunk.

Correct the sense of the PCR_FLINK100 bit.  From Dave Sainty, PR #7832.

Sync w/ -current.

pullup 1.16->1.17 (thorpej): Fix PR7361

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.

Common code for media-from-bmcr.

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.

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.

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).

Make sure CIMDIS is set and FLINK is clear in the PCR.

If the autonegotiation is enabled, the PAR doesn't indicate duplex mode
properly!  Work around this by determing current active media by taking
the highest-order common bit of our advertised capabilites and the link
partner's.

If the link partner doens't do autonegotiation, then parallel detection will
pick up the media type, which will never be full-duplex, so reading the PAR
is ok in this case.

Bug pointed out by Matthias Drochner, work-around inspired by reading
the dp83840 manual, section 3.9 (IEEE 802.3u auto-negotiation).

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.

Adapt to the new MII layer.  This PHY driver now properly supports
802.3u autonegotiation, and will report the precise media selected
by autonegotiation by reading PHY-specific registers.

Nuke __BROKEN_INDIRECT_CONFIG.

Adjust for config changes.

KNF.

Generic functions for the MII layer. mii attaches to mii-capable network
adapters, and provide media-selections and status to the parent.
This includes drivers for the Thunderland integrated PHY, and
National Semiconductor's DP83840A.
Note: The interface betweeen network controller and mii is not complete yet,
and will need some minors tweaks to support media auto-selection.