d/freebsd-skq
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
94889417b9
freebsd-skq/sys/dev/mii/brgphy.c

1077 lines
31 KiB
C
Raw Normal View History

Start each of the license/copyright comments with /*-, minor shuffle of lines
2005-01-06 01:43:34 +00:00
/*-
Add PHY drivers for the XMAC II's internal PHY and the Broadcom BCM5400 1000baseTX PHY. These will be used by the SysKonnect gigabit ethernet driver shortly.
2000-04-22 01:58:18 +00:00
* Copyright (c) 2000
* Bill Paul <wpaul@ee.columbia.edu>. All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
* 3. All advertising materials mentioning features or use of this software
* must display the following acknowledgement:
* This product includes software developed by Bill Paul.
* 4. Neither the name of the author nor the names of any co-contributors
* may be used to endorse or promote products derived from this software
* without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY Bill Paul AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL Bill Paul OR THE VOICES IN HIS HEAD
* BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
* CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
* SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
* INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
* CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
* ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF
* THE POSSIBILITY OF SUCH DAMAGE.
*/
Use __FBSDID(). Also some minor style cleanups.
2003-08-24 17:55:58 +00:00
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
Add PHY drivers for the XMAC II's internal PHY and the Broadcom BCM5400 1000baseTX PHY. These will be used by the SysKonnect gigabit ethernet driver shortly.
2000-04-22 01:58:18 +00:00
/*
Fix typos in comments while I am here.
2007-02-13 00:34:32 +00:00
* Driver for the Broadcom BCM54xx/57xx 1000baseTX PHY.
Add PHY drivers for the XMAC II's internal PHY and the Broadcom BCM5400 1000baseTX PHY. These will be used by the SysKonnect gigabit ethernet driver shortly.
2000-04-22 01:58:18 +00:00
*/
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/kernel.h>
Add some missing <sys/module.h> includes which are masked by the one on death-row in <sys/kernel.h>
2004-05-30 17:57:46 +00:00
#include <sys/module.h>
Add PHY drivers for the XMAC II's internal PHY and the Broadcom BCM5400 1000baseTX PHY. These will be used by the SysKonnect gigabit ethernet driver shortly.
2000-04-22 01:58:18 +00:00
#include <sys/socket.h>
#include <sys/bus.h>
#include <net/if.h>
The MII layer shouldn't care about administrative status of an interface. Make MII drivers forget about 'struct ifnet'. Later plan is to provide an administrative downcall from ifnet layer into drivers, to inform them about administrative status change. If someone thinks that processing MII events for an administratively down interface is a big problem, then drivers would turn MII processing off. The following MII drivers do evil things, like strcmp() on driver name, so they still need knowledge of ifnet and thus include if_var.h. They all need to be fixed: sys/dev/mii/brgphy.c sys/dev/mii/e1000phy.c sys/dev/mii/ip1000phy.c sys/dev/mii/jmphy.c sys/dev/mii/nsphy.c sys/dev/mii/rgephy.c sys/dev/mii/truephy.c Sponsored by: Netflix Sponsored by: Nginx, Inc.
2013-10-26 18:40:17 +00:00
#include <net/if_var.h>
Add a driver for the Broadcom NetXtreme II (BCM5706/BCM5708) PCI/PCIe Gigabit Ethernet adapeter. Submitted by: David Christensen
2006-04-10 19:55:23 +00:00
#include <net/ethernet.h>
Add PHY drivers for the XMAC II's internal PHY and the Broadcom BCM5400 1000baseTX PHY. These will be used by the SysKonnect gigabit ethernet driver shortly.
2000-04-22 01:58:18 +00:00
#include <net/if_media.h>
#include <dev/mii/mii.h>
#include <dev/mii/miivar.h>
Remove miidevs.h and generate it from miidevs at compile time. The devlist2h.awk tool to do this has been repocopied to sys/tools/.
2003-01-19 02:59:34 +00:00
#include "miidevs.h"
Add PHY drivers for the XMAC II's internal PHY and the Broadcom BCM5400 1000baseTX PHY. These will be used by the SysKonnect gigabit ethernet driver shortly.
2000-04-22 01:58:18 +00:00
#include <dev/mii/brgphyreg.h>
Add support for the BCM5705 and its ilk. Changes: - 5705 doesn't support jumbo frames - Statistics must be read from registers - RX return ring must be capped at 512 entries - Omit initialization of certain device blocks - Acknowledge link change interrupts by setting the 'link changed' bit in the status register (used to have no effect) - Remember to toggle the MI completion bit too - Set the mbuf low watermark differently (on-chip memory buffers, not BSD mbufs) - Don't enable Ethernet@WireSpeed feature for certain 5705 chip revs - Add additional PCI IDs for 5705 and 5782 parts - Add a forgotten 5704 PCI ID Most changes ripped kicking and screaming from the Broadcom linux driver. Thanks to Paul Saab for sanity testing. (My lack of sanity has been confirmed.)
2003-07-16 00:09:56 +00:00
#include <net/if_arp.h>
#include <machine/bus.h>
#include <dev/bge/if_bgereg.h>
Add a driver for the Broadcom NetXtreme II (BCM5706/BCM5708) PCI/PCIe Gigabit Ethernet adapeter. Submitted by: David Christensen
2006-04-10 19:55:23 +00:00
#include <dev/bce/if_bcereg.h>
Add PHY drivers for the XMAC II's internal PHY and the Broadcom BCM5400 1000baseTX PHY. These will be used by the SysKonnect gigabit ethernet driver shortly.
2000-04-22 01:58:18 +00:00
Prefer new location of pci include files (which have only been in the tree for two or more years now), except in a few places where there's code to be compatible with older versions of FreeBSD.
2003-08-22 06:42:59 +00:00
#include <dev/pci/pcireg.h>
#include <dev/pci/pcivar.h>
Add support for the Broadcom BCM5901 and BCM5901 rev A2 chips. These are 10/100 only NICs found on the IBM Thinkpad R40E and G40. These seem to be based on the BCM5705 MAC but with a PHY that doesn't support 1000Mbps modes. Submitted by: Igor Sviridov <sia@nest.org>
2003-08-12 05:18:51 +00:00
Add PHY drivers for the XMAC II's internal PHY and the Broadcom BCM5400 1000baseTX PHY. These will be used by the SysKonnect gigabit ethernet driver shortly.
2000-04-22 01:58:18 +00:00
#include "miibus_if.h"
Put function return types on a line by themselves. Cleanup my earlier de-__P sweep and remove whitespace between function names and paramters.
2002-10-14 22:31:52 +00:00
static int brgphy_probe(device_t);
static int brgphy_attach(device_t);
Add PHY drivers for the XMAC II's internal PHY and the Broadcom BCM5400 1000baseTX PHY. These will be used by the SysKonnect gigabit ethernet driver shortly.
2000-04-22 01:58:18 +00:00
Move MII model and revision into softc.
2007-01-16 00:52:26 +00:00
struct brgphy_softc {
struct mii_softc mii_sc;
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
int serdes_flags; /* Keeps track of the serdes type used */
Provide workaround for the ages old bug affecting certain BCM5708S chip revision often found in the blades and resulting in interfaces not sensing carrier signal. Looking at all problem reports it appears that it only affects some very specific silicon revision (ASIC (0x57081021); Rev (B2)) and version of the PHY that supports 1000baseSX-FDX media only. Therefore, narrow the scope of workaround to combination of that revision and media type. Given that the first report on this issue is dated back to 2007, there is not much hope that this issue will ever be properly resolved. Among affected systems are IBM HS21, Intel SBXD132 and HP BL460c. PR: 118238, 122551, 140970 MFC after: 1 month
2010-03-10 05:19:14 +00:00
#define BRGPHY_5706S 0x0001
#define BRGPHY_5708S 0x0002
#define BRGPHY_NOANWAIT 0x0004
- Added support for 5709S/5716S PHYs. Submitted by: pyunyh MFC after: 2 weeks
2010-03-18 20:57:57 +00:00
#define BRGPHY_5709S 0x0008
- Added support for BCM5709 and BCM5716. MFC after: 2 weeks
2008-06-13 01:20:29 +00:00
int bce_phy_flags; /* PHY flags transferred from the MAC driver */
Move MII model and revision into softc.
2007-01-16 00:52:26 +00:00
};
Add PHY drivers for the XMAC II's internal PHY and the Broadcom BCM5400 1000baseTX PHY. These will be used by the SysKonnect gigabit ethernet driver shortly.
2000-04-22 01:58:18 +00:00
static device_method_t brgphy_methods[] = {
/* device interface */
DEVMETHOD(device_probe, brgphy_probe),
DEVMETHOD(device_attach, brgphy_attach),
Make one generic mii_phy_detach() to replace 19 slightly different ones. Rename mii_phy_auto_stop() mii_phy_down(). Introduce mii_down(), use it from nge. Do not indirect it to 19 identical case's in 19 switchstatements like NetBSD did.
2002-04-29 13:07:38 +00:00
DEVMETHOD(device_detach, mii_phy_detach),
Add PHY drivers for the XMAC II's internal PHY and the Broadcom BCM5400 1000baseTX PHY. These will be used by the SysKonnect gigabit ethernet driver shortly.
2000-04-22 01:58:18 +00:00
DEVMETHOD(device_shutdown, bus_generic_shutdown),
Use DEVMETHOD_END.
2011-11-23 20:27:26 +00:00
DEVMETHOD_END
Add PHY drivers for the XMAC II's internal PHY and the Broadcom BCM5400 1000baseTX PHY. These will be used by the SysKonnect gigabit ethernet driver shortly.
2000-04-22 01:58:18 +00:00
};
static devclass_t brgphy_devclass;
static driver_t brgphy_driver = {
"brgphy",
brgphy_methods,
Correct driver_t brgphy_driver, which was forgotten from the last commit.
2007-01-16 17:48:57 +00:00
sizeof(struct brgphy_softc)
Add PHY drivers for the XMAC II's internal PHY and the Broadcom BCM5400 1000baseTX PHY. These will be used by the SysKonnect gigabit ethernet driver shortly.
2000-04-22 01:58:18 +00:00
};
DRIVER_MODULE(brgphy, miibus, brgphy_driver, brgphy_devclass, 0, 0);
Cleanup pass for mii drivers. . Make internal service routines static. . Use a consistent ordering of checks in MII_TICK. Do the work in the mii_phy_tick() subroutine if appropriate. . Call mii_phy_update() to trigger the callbacks.
2001-09-29 19:18:52 +00:00
static int brgphy_service(struct mii_softc *, struct mii_data *, int);
o Flesh out the generic IEEE 802.3 annex 31B full duplex flow control support in mii(4): - Merge generic flow control advertisement (which can be enabled by passing by MIIF_DOPAUSE to mii_attach(9)) and parsing support from NetBSD into mii_physubr.c and ukphy_subr.c. Unlike as in NetBSD, IFM_FLOW isn't implemented as a global option via the "don't care mask" but instead as a media specific option this. This has the following advantages: o allows flow control advertisement with autonegotiation to be turned on and off via ifconfig(8) with the default typically being off (though MIIF_FORCEPAUSE has been added causing flow control to be always advertised, allowing to easily MFC this changes for drivers that previously used home-grown support for flow control that behaved that way without breaking POLA) o allows to deal with PHY drivers where flow control advertisement with manual selection doesn't work or at least isn't implemented, like it's the case with brgphy(4), e1000phy(4) and ip1000phy(4), by setting MIIF_NOMANPAUSE o the available combinations of media options are readily available from the `ifconfig -m` output - Add IFM_FLOW to IFM_SHARED_OPTION_DESCRIPTIONS and IFM_ETH_RXPAUSE and IFM_ETH_TXPAUSE to IFM_SUBTYPE_ETHERNET_OPTION_DESCRIPTIONS so these are understood by ifconfig(8). o Make the master/slave support in mii(4) actually usable: - Change IFM_ETH_MASTER from being implemented as a global option via the "don't care mask" to a media specific one as it actually is only applicable to IFM_1000_T to date. - Let mii_phy_setmedia() set GTCR_MAN_MS in IFM_1000_T slave mode to actually configure manually selected slave mode (like we also do in the PHY specific implementations). - Add IFM_ETH_MASTER to IFM_SUBTYPE_ETHERNET_OPTION_DESCRIPTIONS so it is understood by ifconfig(8). o Switch bge(4), bce(4), msk(4), nfe(4) and stge(4) along with brgphy(4), e1000phy(4) and ip1000phy(4) to use the generic flow control support instead of home-grown solutions via IFM_FLAGs. This includes changing these PHY drivers and smcphy(4) to no longer unconditionally advertise support for flow control but only if the selected media has IFM_FLOW set (or MIIF_FORCEPAUSE is set) and implemented for these media variants, i.e. typically only for copper. o Switch brgphy(4), ciphy(4), e1000phy(4) and ip1000phy(4) to report and set IFM_1000_T master mode via IFM_ETH_MASTER instead of via IFF_LINK0 and some IFM_FLAGn. o Switch brgphy(4) to add at least the the supported copper media based on the contents of the BMSR via mii_phy_add_media() instead of hardcoding them. The latter approach seems to have developed historically, besides causing unnecessary code duplication it was also undesirable because brgphy_mii_phy_auto() already based the capability advertisement on the contents of the BMSR though. o Let brgphy(4) set IFM_1000_T master mode on all supported PHY and not just BCM5701. Apparently this was a misinterpretation of a workaround in the Linux tg3 driver; BCM5701 seem to require RGPHY_1000CTL_MSE and BRGPHY_1000CTL_MSC to be set when configuring autonegotiation but this doesn't mean we can't set these as well on other PHYs for manual media selection. o Let ukphy_status() report IFM_1000_T master mode via IFM_ETH_MASTER so IFM_1000_T master mode support now is generally available with all PHY drivers. o Don't let e1000phy(4) set master/slave bits for IFM_1000_SX as it's not applicable there. Reviewed by: yongari (plus additional testing) Obtained from: NetBSD (partially), OpenBSD (partially) MFC after: 2 weeks
2010-11-14 13:26:10 +00:00
static void brgphy_setmedia(struct mii_softc *, int);
Cleanup pass for mii drivers. . Make internal service routines static. . Use a consistent ordering of checks in MII_TICK. Do the work in the mii_phy_tick() subroutine if appropriate. . Call mii_phy_update() to trigger the callbacks.
2001-09-29 19:18:52 +00:00
static void brgphy_status(struct mii_softc *);
o Flesh out the generic IEEE 802.3 annex 31B full duplex flow control support in mii(4): - Merge generic flow control advertisement (which can be enabled by passing by MIIF_DOPAUSE to mii_attach(9)) and parsing support from NetBSD into mii_physubr.c and ukphy_subr.c. Unlike as in NetBSD, IFM_FLOW isn't implemented as a global option via the "don't care mask" but instead as a media specific option this. This has the following advantages: o allows flow control advertisement with autonegotiation to be turned on and off via ifconfig(8) with the default typically being off (though MIIF_FORCEPAUSE has been added causing flow control to be always advertised, allowing to easily MFC this changes for drivers that previously used home-grown support for flow control that behaved that way without breaking POLA) o allows to deal with PHY drivers where flow control advertisement with manual selection doesn't work or at least isn't implemented, like it's the case with brgphy(4), e1000phy(4) and ip1000phy(4), by setting MIIF_NOMANPAUSE o the available combinations of media options are readily available from the `ifconfig -m` output - Add IFM_FLOW to IFM_SHARED_OPTION_DESCRIPTIONS and IFM_ETH_RXPAUSE and IFM_ETH_TXPAUSE to IFM_SUBTYPE_ETHERNET_OPTION_DESCRIPTIONS so these are understood by ifconfig(8). o Make the master/slave support in mii(4) actually usable: - Change IFM_ETH_MASTER from being implemented as a global option via the "don't care mask" to a media specific one as it actually is only applicable to IFM_1000_T to date. - Let mii_phy_setmedia() set GTCR_MAN_MS in IFM_1000_T slave mode to actually configure manually selected slave mode (like we also do in the PHY specific implementations). - Add IFM_ETH_MASTER to IFM_SUBTYPE_ETHERNET_OPTION_DESCRIPTIONS so it is understood by ifconfig(8). o Switch bge(4), bce(4), msk(4), nfe(4) and stge(4) along with brgphy(4), e1000phy(4) and ip1000phy(4) to use the generic flow control support instead of home-grown solutions via IFM_FLAGs. This includes changing these PHY drivers and smcphy(4) to no longer unconditionally advertise support for flow control but only if the selected media has IFM_FLOW set (or MIIF_FORCEPAUSE is set) and implemented for these media variants, i.e. typically only for copper. o Switch brgphy(4), ciphy(4), e1000phy(4) and ip1000phy(4) to report and set IFM_1000_T master mode via IFM_ETH_MASTER instead of via IFF_LINK0 and some IFM_FLAGn. o Switch brgphy(4) to add at least the the supported copper media based on the contents of the BMSR via mii_phy_add_media() instead of hardcoding them. The latter approach seems to have developed historically, besides causing unnecessary code duplication it was also undesirable because brgphy_mii_phy_auto() already based the capability advertisement on the contents of the BMSR though. o Let brgphy(4) set IFM_1000_T master mode on all supported PHY and not just BCM5701. Apparently this was a misinterpretation of a workaround in the Linux tg3 driver; BCM5701 seem to require RGPHY_1000CTL_MSE and BRGPHY_1000CTL_MSC to be set when configuring autonegotiation but this doesn't mean we can't set these as well on other PHYs for manual media selection. o Let ukphy_status() report IFM_1000_T master mode via IFM_ETH_MASTER so IFM_1000_T master mode support now is generally available with all PHY drivers. o Don't let e1000phy(4) set master/slave bits for IFM_1000_SX as it's not applicable there. Reviewed by: yongari (plus additional testing) Obtained from: NetBSD (partially), OpenBSD (partially) MFC after: 2 weeks
2010-11-14 13:26:10 +00:00
static void brgphy_mii_phy_auto(struct mii_softc *, int);
- Move bge_phy_hack into the phy code and implement the various DSP patch workarounds for each phy revision. Obtained from: NetBSD & Broadcom Linux driver - Disable AUTOPOLL when accessing the PHY as it may cause PCI errors. Obtained from: NetBSD - Check the UPDATED bit in the status block so the driver knows that the status block as indeed changed since the last access. Broadcom documentation states drivers should unset the UPDATED/CHANGED bits after reading them. - When changing media types, first loop the phy then set the media. Broadcom documentation and Linux drivers do this and I observed much better handling of link after this change. - Broadcom documentation states that for 1000BaseT operation, autonegotiation must be enabled. Fix hard coding of media so that the driver only advertises 1000BaseT as the supported media type and enable autonegotition. - Only set Master/Slave on the 5701. Obtained from Broadcom Linux driver.
2003-05-03 19:06:50 +00:00
static void brgphy_reset(struct mii_softc *);
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
static void brgphy_enable_loopback(struct mii_softc *);
- Move bge_phy_hack into the phy code and implement the various DSP patch workarounds for each phy revision. Obtained from: NetBSD & Broadcom Linux driver - Disable AUTOPOLL when accessing the PHY as it may cause PCI errors. Obtained from: NetBSD - Check the UPDATED bit in the status block so the driver knows that the status block as indeed changed since the last access. Broadcom documentation states drivers should unset the UPDATED/CHANGED bits after reading them. - When changing media types, first loop the phy then set the media. Broadcom documentation and Linux drivers do this and I observed much better handling of link after this change. - Broadcom documentation states that for 1000BaseT operation, autonegotiation must be enabled. Fix hard coding of media so that the driver only advertises 1000BaseT as the supported media type and enable autonegotition. - Only set Master/Slave on the 5701. Obtained from Broadcom Linux driver.
2003-05-03 19:06:50 +00:00
static void bcm5401_load_dspcode(struct mii_softc *);
static void bcm5411_load_dspcode(struct mii_softc *);
Add support for BCM54K2 found in combination with Apple K2 GMAC. Submitted by: Andreas Tobler Obtained from: OpenBSD MFC after: 1 week
2010-02-20 22:01:24 +00:00
static void bcm54k2_load_dspcode(struct mii_softc *);
- Fix BCM5754 support found in Dell PowerEdge SC440. - Move some PHY bug detections from brgphy.c to if_bge.c. - Do not penalize working PHYs. - Re-arrange bge_flags roughly by their categories. - Fix minor style(9) nits. PR: kern/107257 Obtained from: OpenBSD Tested by: Mike Hibler <mike at flux dot utah dot edu>
2007-01-15 21:43:43 +00:00
static void brgphy_fixup_5704_a0_bug(struct mii_softc *);
Fix style(9). Pointed out by: many
2007-02-12 23:33:05 +00:00
static void brgphy_fixup_adc_bug(struct mii_softc *);
Add PHY DSP code for BCM5755M. Obtained from: OpenBSD
2007-02-12 22:51:25 +00:00
static void brgphy_fixup_adjust_trim(struct mii_softc *);
- Fix BCM5754 support found in Dell PowerEdge SC440. - Move some PHY bug detections from brgphy.c to if_bge.c. - Do not penalize working PHYs. - Re-arrange bge_flags roughly by their categories. - Fix minor style(9) nits. PR: kern/107257 Obtained from: OpenBSD Tested by: Mike Hibler <mike at flux dot utah dot edu>
2007-01-15 21:43:43 +00:00
static void brgphy_fixup_ber_bug(struct mii_softc *);
Add BCM5701 A0/B0 CRC bug workaround. Magic values taken from Linux driver.
2007-02-12 23:58:52 +00:00
static void brgphy_fixup_crc_bug(struct mii_softc *);
- Fix BCM5754 support found in Dell PowerEdge SC440. - Move some PHY bug detections from brgphy.c to if_bge.c. - Do not penalize working PHYs. - Re-arrange bge_flags roughly by their categories. - Fix minor style(9) nits. PR: kern/107257 Obtained from: OpenBSD Tested by: Mike Hibler <mike at flux dot utah dot edu>
2007-01-15 21:43:43 +00:00
static void brgphy_fixup_jitter_bug(struct mii_softc *);
- Move Ethernet@WireSpeed and jumbo frame configurations to separate functions. The idea is taken from OpenBSD. - Set/clear jumbo frame configurations for bge(4). - Re-add BCM5750 PHY workaround for bce(4), which was mistakenly removed from the previous commit.
2007-01-15 22:21:44 +00:00
static void brgphy_ethernet_wirespeed(struct mii_softc *);
static void brgphy_jumbo_settings(struct mii_softc *, u_long);
Add PHY drivers for the XMAC II's internal PHY and the Broadcom BCM5400 1000baseTX PHY. These will be used by the SysKonnect gigabit ethernet driver shortly.
2000-04-22 01:58:18 +00:00
Use mii_phy_match to reduce duplicated code. Reviewed by: glebius
2006-07-03 08:01:27 +00:00
static const struct mii_phydesc brgphys[] = {
- Remove attempts to implement setting of BMCR_LOOP/MIIF_NOLOOP (reporting IFM_LOOP based on BMCR_LOOP is left in place though as it might provide useful for debugging). For most mii(4) drivers it was unclear whether the PHYs driven by them actually support loopback or not. Moreover, typically loopback mode also needs to be activated on the MAC, which none of the Ethernet drivers using mii(4) implements. Given that loopback media has no real use (and obviously hardly had a chance to actually work) besides for driver development (which just loopback mode should be sufficient for though, i.e one doesn't necessary need support for loopback media) support for it is just dropped as both NetBSD and OpenBSD already did quite some time ago. - Let mii_phy_add_media() also announce the support of IFM_NONE. - 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 lxtphy(4) which disables MII interrupts (as is done for a few other PHYs we have drivers for). This includes changing NIC drivers which previously just called the generic mii_phy_reset() to now actually call the PHY-specific reset routine, which might be crucial in some cases. While at it, the redundant checks in these NIC drivers for mii->mii_instance not being zero before calling the reset routines were removed because as soon as one PHY driver attaches mii->mii_instance is incremented and we hardly can end up in their media change callbacks etc if no PHY driver has attached as mii_attach() would have failed in that case and not attach a miibus(4) instance. Consequently, NIC drivers now no longer should call mii_phy_reset() directly, so it was removed from EXPORT_SYMS. - Add a mii_phy_dev_attach() as a companion helper to mii_phy_dev_probe(). The purpose of that function is to perform the common steps to attach a PHY driver instance and to hook it up to the miibus(4) instance and to optionally also handle the probing, addition and initialization of the supported media. So all a PHY driver without any special requirements has to do in its bus attach method is to call mii_phy_dev_attach() along with PHY-specific MIIF_* flags, a pointer to its PHY functions and the add_media set to one. All PHY drivers were updated to take advantage of mii_phy_dev_attach() as appropriate. Along with these changes the capability mask was added to the mii_softc structure so PHY drivers taking advantage of mii_phy_dev_attach() but still handling media on their own do not need to fiddle with the MII attach arguments anyway. - Keep track of the PHY offset in the mii_softc structure. This is done for compatibility with NetBSD/OpenBSD. - Keep track of the PHY's OUI, model and revision in the mii_softc structure. Several PHY drivers require this information also after attaching and previously had to wrap their own softc around mii_softc. NetBSD/OpenBSD also keep track of the model and revision on their mii_softc structure. All PHY drivers were updated to take advantage as appropriate. - Convert the mebers of the MII data structure to unsigned where appropriate. This is partly inspired by NetBSD/OpenBSD. - According to IEEE 802.3-2002 the bits actually have to be reversed when mapping an OUI to the MII ID registers. All PHY drivers and miidevs where changed as necessary. Actually this now again allows to largely share miidevs with NetBSD, which fixed this problem already 9 years ago. Consequently miidevs was synced as far as possible. - Add MIIF_NOMANPAUSE and mii_phy_flowstatus() calls to drivers that weren't explicitly converted to support flow control before. It's unclear whether flow control actually works with these but typically it should and their net behavior should be more correct with these changes in place than without if the MAC driver sets MIIF_DOPAUSE. Obtained from: NetBSD (partially) Reviewed by: yongari (earlier version), silence on arch@ and net@
2011-05-03 19:51:29 +00:00
MII_PHY_DESC(BROADCOM, BCM5400),
MII_PHY_DESC(BROADCOM, BCM5401),
MII_PHY_DESC(BROADCOM, BCM5411),
MII_PHY_DESC(BROADCOM, BCM54K2),
MII_PHY_DESC(BROADCOM, BCM5701),
MII_PHY_DESC(BROADCOM, BCM5703),
MII_PHY_DESC(BROADCOM, BCM5704),
MII_PHY_DESC(BROADCOM, BCM5705),
MII_PHY_DESC(BROADCOM, BCM5706),
MII_PHY_DESC(BROADCOM, BCM5714),
MII_PHY_DESC(BROADCOM, BCM5421),
MII_PHY_DESC(BROADCOM, BCM5750),
MII_PHY_DESC(BROADCOM, BCM5752),
MII_PHY_DESC(BROADCOM, BCM5780),
MII_PHY_DESC(BROADCOM, BCM5708C),
MII_PHY_DESC(BROADCOM2, BCM5482),
MII_PHY_DESC(BROADCOM2, BCM5708S),
MII_PHY_DESC(BROADCOM2, BCM5709C),
MII_PHY_DESC(BROADCOM2, BCM5709S),
MII_PHY_DESC(BROADCOM2, BCM5709CAX),
MII_PHY_DESC(BROADCOM2, BCM5722),
MII_PHY_DESC(BROADCOM2, BCM5755),
MII_PHY_DESC(BROADCOM2, BCM5754),
MII_PHY_DESC(BROADCOM2, BCM5761),
MII_PHY_DESC(BROADCOM2, BCM5784),
Wrap BCM5785 in #ifdef notyet for now. According to yongari@ there are issues probably needing workarounds in bge(4) when brgphy(4) handles this PHY. Letting ukphy(4) handle it instead results in a working configuration, although likely with performance penalties.
2011-11-23 22:05:44 +00:00
#ifdef notyet /* better handled by ukphy(4) until WARs are implemented */
Probe the BCM5785. Obtained from: NetBSD
2011-11-23 20:09:34 +00:00
MII_PHY_DESC(BROADCOM2, BCM5785),
Wrap BCM5785 in #ifdef notyet for now. According to yongari@ there are issues probably needing workarounds in bge(4) when brgphy(4) handles this PHY. Letting ukphy(4) handle it instead results in a working configuration, although likely with performance penalties.
2011-11-23 22:05:44 +00:00
#endif
- Remove attempts to implement setting of BMCR_LOOP/MIIF_NOLOOP (reporting IFM_LOOP based on BMCR_LOOP is left in place though as it might provide useful for debugging). For most mii(4) drivers it was unclear whether the PHYs driven by them actually support loopback or not. Moreover, typically loopback mode also needs to be activated on the MAC, which none of the Ethernet drivers using mii(4) implements. Given that loopback media has no real use (and obviously hardly had a chance to actually work) besides for driver development (which just loopback mode should be sufficient for though, i.e one doesn't necessary need support for loopback media) support for it is just dropped as both NetBSD and OpenBSD already did quite some time ago. - Let mii_phy_add_media() also announce the support of IFM_NONE. - 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 lxtphy(4) which disables MII interrupts (as is done for a few other PHYs we have drivers for). This includes changing NIC drivers which previously just called the generic mii_phy_reset() to now actually call the PHY-specific reset routine, which might be crucial in some cases. While at it, the redundant checks in these NIC drivers for mii->mii_instance not being zero before calling the reset routines were removed because as soon as one PHY driver attaches mii->mii_instance is incremented and we hardly can end up in their media change callbacks etc if no PHY driver has attached as mii_attach() would have failed in that case and not attach a miibus(4) instance. Consequently, NIC drivers now no longer should call mii_phy_reset() directly, so it was removed from EXPORT_SYMS. - Add a mii_phy_dev_attach() as a companion helper to mii_phy_dev_probe(). The purpose of that function is to perform the common steps to attach a PHY driver instance and to hook it up to the miibus(4) instance and to optionally also handle the probing, addition and initialization of the supported media. So all a PHY driver without any special requirements has to do in its bus attach method is to call mii_phy_dev_attach() along with PHY-specific MIIF_* flags, a pointer to its PHY functions and the add_media set to one. All PHY drivers were updated to take advantage of mii_phy_dev_attach() as appropriate. Along with these changes the capability mask was added to the mii_softc structure so PHY drivers taking advantage of mii_phy_dev_attach() but still handling media on their own do not need to fiddle with the MII attach arguments anyway. - Keep track of the PHY offset in the mii_softc structure. This is done for compatibility with NetBSD/OpenBSD. - Keep track of the PHY's OUI, model and revision in the mii_softc structure. Several PHY drivers require this information also after attaching and previously had to wrap their own softc around mii_softc. NetBSD/OpenBSD also keep track of the model and revision on their mii_softc structure. All PHY drivers were updated to take advantage as appropriate. - Convert the mebers of the MII data structure to unsigned where appropriate. This is partly inspired by NetBSD/OpenBSD. - According to IEEE 802.3-2002 the bits actually have to be reversed when mapping an OUI to the MII ID registers. All PHY drivers and miidevs where changed as necessary. Actually this now again allows to largely share miidevs with NetBSD, which fixed this problem already 9 years ago. Consequently miidevs was synced as far as possible. - Add MIIF_NOMANPAUSE and mii_phy_flowstatus() calls to drivers that weren't explicitly converted to support flow control before. It's unclear whether flow control actually works with these but typically it should and their net behavior should be more correct with these changes in place than without if the MAC driver sets MIIF_DOPAUSE. Obtained from: NetBSD (partially) Reviewed by: yongari (earlier version), silence on arch@ and net@
2011-05-03 19:51:29 +00:00
MII_PHY_DESC(BROADCOM3, BCM5717C),
Recognize BCM5719C PHY. Submitted by: Geans Pin at Broadcom
2011-05-09 20:20:43 +00:00
MII_PHY_DESC(BROADCOM3, BCM5719C),
Recognize BCM5720C PHY.
2011-10-28 00:40:19 +00:00
MII_PHY_DESC(BROADCOM3, BCM5720C),
- Remove attempts to implement setting of BMCR_LOOP/MIIF_NOLOOP (reporting IFM_LOOP based on BMCR_LOOP is left in place though as it might provide useful for debugging). For most mii(4) drivers it was unclear whether the PHYs driven by them actually support loopback or not. Moreover, typically loopback mode also needs to be activated on the MAC, which none of the Ethernet drivers using mii(4) implements. Given that loopback media has no real use (and obviously hardly had a chance to actually work) besides for driver development (which just loopback mode should be sufficient for though, i.e one doesn't necessary need support for loopback media) support for it is just dropped as both NetBSD and OpenBSD already did quite some time ago. - Let mii_phy_add_media() also announce the support of IFM_NONE. - 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 lxtphy(4) which disables MII interrupts (as is done for a few other PHYs we have drivers for). This includes changing NIC drivers which previously just called the generic mii_phy_reset() to now actually call the PHY-specific reset routine, which might be crucial in some cases. While at it, the redundant checks in these NIC drivers for mii->mii_instance not being zero before calling the reset routines were removed because as soon as one PHY driver attaches mii->mii_instance is incremented and we hardly can end up in their media change callbacks etc if no PHY driver has attached as mii_attach() would have failed in that case and not attach a miibus(4) instance. Consequently, NIC drivers now no longer should call mii_phy_reset() directly, so it was removed from EXPORT_SYMS. - Add a mii_phy_dev_attach() as a companion helper to mii_phy_dev_probe(). The purpose of that function is to perform the common steps to attach a PHY driver instance and to hook it up to the miibus(4) instance and to optionally also handle the probing, addition and initialization of the supported media. So all a PHY driver without any special requirements has to do in its bus attach method is to call mii_phy_dev_attach() along with PHY-specific MIIF_* flags, a pointer to its PHY functions and the add_media set to one. All PHY drivers were updated to take advantage of mii_phy_dev_attach() as appropriate. Along with these changes the capability mask was added to the mii_softc structure so PHY drivers taking advantage of mii_phy_dev_attach() but still handling media on their own do not need to fiddle with the MII attach arguments anyway. - Keep track of the PHY offset in the mii_softc structure. This is done for compatibility with NetBSD/OpenBSD. - Keep track of the PHY's OUI, model and revision in the mii_softc structure. Several PHY drivers require this information also after attaching and previously had to wrap their own softc around mii_softc. NetBSD/OpenBSD also keep track of the model and revision on their mii_softc structure. All PHY drivers were updated to take advantage as appropriate. - Convert the mebers of the MII data structure to unsigned where appropriate. This is partly inspired by NetBSD/OpenBSD. - According to IEEE 802.3-2002 the bits actually have to be reversed when mapping an OUI to the MII ID registers. All PHY drivers and miidevs where changed as necessary. Actually this now again allows to largely share miidevs with NetBSD, which fixed this problem already 9 years ago. Consequently miidevs was synced as far as possible. - Add MIIF_NOMANPAUSE and mii_phy_flowstatus() calls to drivers that weren't explicitly converted to support flow control before. It's unclear whether flow control actually works with these but typically it should and their net behavior should be more correct with these changes in place than without if the MAC driver sets MIIF_DOPAUSE. Obtained from: NetBSD (partially) Reviewed by: yongari (earlier version), silence on arch@ and net@
2011-05-03 19:51:29 +00:00
MII_PHY_DESC(BROADCOM3, BCM57765),
- Probe BCM57780. - In case the parent is bge(4), don't set the Jumbo frame settings unless the MAC actually is Jumbo capable as otherwise the PHY might not have the corresponding registers implemented. This is also in line with what the Linux tg3 driver does. PR: 165032 Submitted by: Alexander Milanov Obtained from: OpenBSD MFC after: 3 days
2012-02-19 12:09:17 +00:00
MII_PHY_DESC(BROADCOM3, BCM57780),
Recognize BCM5725C PHY.
2013-07-20 07:24:01 +00:00
MII_PHY_DESC(BROADCOM4, BCM5725C),
- Remove attempts to implement setting of BMCR_LOOP/MIIF_NOLOOP (reporting IFM_LOOP based on BMCR_LOOP is left in place though as it might provide useful for debugging). For most mii(4) drivers it was unclear whether the PHYs driven by them actually support loopback or not. Moreover, typically loopback mode also needs to be activated on the MAC, which none of the Ethernet drivers using mii(4) implements. Given that loopback media has no real use (and obviously hardly had a chance to actually work) besides for driver development (which just loopback mode should be sufficient for though, i.e one doesn't necessary need support for loopback media) support for it is just dropped as both NetBSD and OpenBSD already did quite some time ago. - Let mii_phy_add_media() also announce the support of IFM_NONE. - 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 lxtphy(4) which disables MII interrupts (as is done for a few other PHYs we have drivers for). This includes changing NIC drivers which previously just called the generic mii_phy_reset() to now actually call the PHY-specific reset routine, which might be crucial in some cases. While at it, the redundant checks in these NIC drivers for mii->mii_instance not being zero before calling the reset routines were removed because as soon as one PHY driver attaches mii->mii_instance is incremented and we hardly can end up in their media change callbacks etc if no PHY driver has attached as mii_attach() would have failed in that case and not attach a miibus(4) instance. Consequently, NIC drivers now no longer should call mii_phy_reset() directly, so it was removed from EXPORT_SYMS. - Add a mii_phy_dev_attach() as a companion helper to mii_phy_dev_probe(). The purpose of that function is to perform the common steps to attach a PHY driver instance and to hook it up to the miibus(4) instance and to optionally also handle the probing, addition and initialization of the supported media. So all a PHY driver without any special requirements has to do in its bus attach method is to call mii_phy_dev_attach() along with PHY-specific MIIF_* flags, a pointer to its PHY functions and the add_media set to one. All PHY drivers were updated to take advantage of mii_phy_dev_attach() as appropriate. Along with these changes the capability mask was added to the mii_softc structure so PHY drivers taking advantage of mii_phy_dev_attach() but still handling media on their own do not need to fiddle with the MII attach arguments anyway. - Keep track of the PHY offset in the mii_softc structure. This is done for compatibility with NetBSD/OpenBSD. - Keep track of the PHY's OUI, model and revision in the mii_softc structure. Several PHY drivers require this information also after attaching and previously had to wrap their own softc around mii_softc. NetBSD/OpenBSD also keep track of the model and revision on their mii_softc structure. All PHY drivers were updated to take advantage as appropriate. - Convert the mebers of the MII data structure to unsigned where appropriate. This is partly inspired by NetBSD/OpenBSD. - According to IEEE 802.3-2002 the bits actually have to be reversed when mapping an OUI to the MII ID registers. All PHY drivers and miidevs where changed as necessary. Actually this now again allows to largely share miidevs with NetBSD, which fixed this problem already 9 years ago. Consequently miidevs was synced as far as possible. - Add MIIF_NOMANPAUSE and mii_phy_flowstatus() calls to drivers that weren't explicitly converted to support flow control before. It's unclear whether flow control actually works with these but typically it should and their net behavior should be more correct with these changes in place than without if the MAC driver sets MIIF_DOPAUSE. Obtained from: NetBSD (partially) Reviewed by: yongari (earlier version), silence on arch@ and net@
2011-05-03 19:51:29 +00:00
MII_PHY_DESC(xxBROADCOM_ALT1, BCM5906),
Use mii_phy_match to reduce duplicated code. Reviewed by: glebius
2006-07-03 08:01:27 +00:00
MII_PHY_END
};
- Remove attempts to implement setting of BMCR_LOOP/MIIF_NOLOOP (reporting IFM_LOOP based on BMCR_LOOP is left in place though as it might provide useful for debugging). For most mii(4) drivers it was unclear whether the PHYs driven by them actually support loopback or not. Moreover, typically loopback mode also needs to be activated on the MAC, which none of the Ethernet drivers using mii(4) implements. Given that loopback media has no real use (and obviously hardly had a chance to actually work) besides for driver development (which just loopback mode should be sufficient for though, i.e one doesn't necessary need support for loopback media) support for it is just dropped as both NetBSD and OpenBSD already did quite some time ago. - Let mii_phy_add_media() also announce the support of IFM_NONE. - 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 lxtphy(4) which disables MII interrupts (as is done for a few other PHYs we have drivers for). This includes changing NIC drivers which previously just called the generic mii_phy_reset() to now actually call the PHY-specific reset routine, which might be crucial in some cases. While at it, the redundant checks in these NIC drivers for mii->mii_instance not being zero before calling the reset routines were removed because as soon as one PHY driver attaches mii->mii_instance is incremented and we hardly can end up in their media change callbacks etc if no PHY driver has attached as mii_attach() would have failed in that case and not attach a miibus(4) instance. Consequently, NIC drivers now no longer should call mii_phy_reset() directly, so it was removed from EXPORT_SYMS. - Add a mii_phy_dev_attach() as a companion helper to mii_phy_dev_probe(). The purpose of that function is to perform the common steps to attach a PHY driver instance and to hook it up to the miibus(4) instance and to optionally also handle the probing, addition and initialization of the supported media. So all a PHY driver without any special requirements has to do in its bus attach method is to call mii_phy_dev_attach() along with PHY-specific MIIF_* flags, a pointer to its PHY functions and the add_media set to one. All PHY drivers were updated to take advantage of mii_phy_dev_attach() as appropriate. Along with these changes the capability mask was added to the mii_softc structure so PHY drivers taking advantage of mii_phy_dev_attach() but still handling media on their own do not need to fiddle with the MII attach arguments anyway. - Keep track of the PHY offset in the mii_softc structure. This is done for compatibility with NetBSD/OpenBSD. - Keep track of the PHY's OUI, model and revision in the mii_softc structure. Several PHY drivers require this information also after attaching and previously had to wrap their own softc around mii_softc. NetBSD/OpenBSD also keep track of the model and revision on their mii_softc structure. All PHY drivers were updated to take advantage as appropriate. - Convert the mebers of the MII data structure to unsigned where appropriate. This is partly inspired by NetBSD/OpenBSD. - According to IEEE 802.3-2002 the bits actually have to be reversed when mapping an OUI to the MII ID registers. All PHY drivers and miidevs where changed as necessary. Actually this now again allows to largely share miidevs with NetBSD, which fixed this problem already 9 years ago. Consequently miidevs was synced as far as possible. - Add MIIF_NOMANPAUSE and mii_phy_flowstatus() calls to drivers that weren't explicitly converted to support flow control before. It's unclear whether flow control actually works with these but typically it should and their net behavior should be more correct with these changes in place than without if the MAC driver sets MIIF_DOPAUSE. Obtained from: NetBSD (partially) Reviewed by: yongari (earlier version), silence on arch@ and net@
2011-05-03 19:51:29 +00:00
static const struct mii_phy_funcs brgphy_funcs = {
brgphy_service,
brgphy_status,
brgphy_reset
};
further narrow down no carrier workaround, since it appears to only affect very specific IBM hardware and other machines with the same BCM ASIC chip id 0x57081021 are just fine. MFC after: 1 month
2010-03-10 23:00:15 +00:00
#define HS21_PRODUCT_ID "IBM eServer BladeCenter HS21"
#define HS21_BCM_CHIPID 0x57081021
static int
detect_hs21(struct bce_softc *bce_sc)
{
char *sysenv;
Plug memory leakage introduced in r204989. Reported by: yongari
2010-11-15 23:13:25 +00:00
int found;
found = 0;
if (bce_sc->bce_chipid == HS21_BCM_CHIPID) {
sysenv = getenv("smbios.system.product");
if (sysenv != NULL) {
Restore the previous behaviour of substring match.
2010-11-15 23:38:52 +00:00
if (strncmp(sysenv, HS21_PRODUCT_ID,
strlen(HS21_PRODUCT_ID)) == 0)
Plug memory leakage introduced in r204989. Reported by: yongari
2010-11-15 23:13:25 +00:00
found = 1;
freeenv(sysenv);
}
}
return (found);
further narrow down no carrier workaround, since it appears to only affect very specific IBM hardware and other machines with the same BCM ASIC chip id 0x57081021 are just fine. MFC after: 1 month
2010-03-10 23:00:15 +00:00
}
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
/* Search for our PHY in the list of known PHYs */
Put function return types on a line by themselves. Cleanup my earlier de-__P sweep and remove whitespace between function names and paramters.
2002-10-14 22:31:52 +00:00
static int
Use ansi function definitions in preference to K&R to reduce diffs with NetBSD (and cause it looks cooler).
2005-09-30 19:39:27 +00:00
brgphy_probe(device_t dev)
Add PHY drivers for the XMAC II's internal PHY and the Broadcom BCM5400 1000baseTX PHY. These will be used by the SysKonnect gigabit ethernet driver shortly.
2000-04-22 01:58:18 +00:00
{
o Flesh out the generic IEEE 802.3 annex 31B full duplex flow control support in mii(4): - Merge generic flow control advertisement (which can be enabled by passing by MIIF_DOPAUSE to mii_attach(9)) and parsing support from NetBSD into mii_physubr.c and ukphy_subr.c. Unlike as in NetBSD, IFM_FLOW isn't implemented as a global option via the "don't care mask" but instead as a media specific option this. This has the following advantages: o allows flow control advertisement with autonegotiation to be turned on and off via ifconfig(8) with the default typically being off (though MIIF_FORCEPAUSE has been added causing flow control to be always advertised, allowing to easily MFC this changes for drivers that previously used home-grown support for flow control that behaved that way without breaking POLA) o allows to deal with PHY drivers where flow control advertisement with manual selection doesn't work or at least isn't implemented, like it's the case with brgphy(4), e1000phy(4) and ip1000phy(4), by setting MIIF_NOMANPAUSE o the available combinations of media options are readily available from the `ifconfig -m` output - Add IFM_FLOW to IFM_SHARED_OPTION_DESCRIPTIONS and IFM_ETH_RXPAUSE and IFM_ETH_TXPAUSE to IFM_SUBTYPE_ETHERNET_OPTION_DESCRIPTIONS so these are understood by ifconfig(8). o Make the master/slave support in mii(4) actually usable: - Change IFM_ETH_MASTER from being implemented as a global option via the "don't care mask" to a media specific one as it actually is only applicable to IFM_1000_T to date. - Let mii_phy_setmedia() set GTCR_MAN_MS in IFM_1000_T slave mode to actually configure manually selected slave mode (like we also do in the PHY specific implementations). - Add IFM_ETH_MASTER to IFM_SUBTYPE_ETHERNET_OPTION_DESCRIPTIONS so it is understood by ifconfig(8). o Switch bge(4), bce(4), msk(4), nfe(4) and stge(4) along with brgphy(4), e1000phy(4) and ip1000phy(4) to use the generic flow control support instead of home-grown solutions via IFM_FLAGs. This includes changing these PHY drivers and smcphy(4) to no longer unconditionally advertise support for flow control but only if the selected media has IFM_FLOW set (or MIIF_FORCEPAUSE is set) and implemented for these media variants, i.e. typically only for copper. o Switch brgphy(4), ciphy(4), e1000phy(4) and ip1000phy(4) to report and set IFM_1000_T master mode via IFM_ETH_MASTER instead of via IFF_LINK0 and some IFM_FLAGn. o Switch brgphy(4) to add at least the the supported copper media based on the contents of the BMSR via mii_phy_add_media() instead of hardcoding them. The latter approach seems to have developed historically, besides causing unnecessary code duplication it was also undesirable because brgphy_mii_phy_auto() already based the capability advertisement on the contents of the BMSR though. o Let brgphy(4) set IFM_1000_T master mode on all supported PHY and not just BCM5701. Apparently this was a misinterpretation of a workaround in the Linux tg3 driver; BCM5701 seem to require RGPHY_1000CTL_MSE and BRGPHY_1000CTL_MSC to be set when configuring autonegotiation but this doesn't mean we can't set these as well on other PHYs for manual media selection. o Let ukphy_status() report IFM_1000_T master mode via IFM_ETH_MASTER so IFM_1000_T master mode support now is generally available with all PHY drivers. o Don't let e1000phy(4) set master/slave bits for IFM_1000_SX as it's not applicable there. Reviewed by: yongari (plus additional testing) Obtained from: NetBSD (partially), OpenBSD (partially) MFC after: 2 weeks
2010-11-14 13:26:10 +00:00
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
return (mii_phy_dev_probe(dev, brgphys, BUS_PROBE_DEFAULT));
Add PHY drivers for the XMAC II's internal PHY and the Broadcom BCM5400 1000baseTX PHY. These will be used by the SysKonnect gigabit ethernet driver shortly.
2000-04-22 01:58:18 +00:00
}
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
/* Attach the PHY to the MII bus */
Put function return types on a line by themselves. Cleanup my earlier de-__P sweep and remove whitespace between function names and paramters.
2002-10-14 22:31:52 +00:00
static int
Use ansi function definitions in preference to K&R to reduce diffs with NetBSD (and cause it looks cooler).
2005-09-30 19:39:27 +00:00
brgphy_attach(device_t dev)
Add PHY drivers for the XMAC II's internal PHY and the Broadcom BCM5400 1000baseTX PHY. These will be used by the SysKonnect gigabit ethernet driver shortly.
2000-04-22 01:58:18 +00:00
{
Move MII model and revision into softc.
2007-01-16 00:52:26 +00:00
struct brgphy_softc *bsc;
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
struct bge_softc *bge_sc = NULL;
struct bce_softc *bce_sc = NULL;
Add PHY drivers for the XMAC II's internal PHY and the Broadcom BCM5400 1000baseTX PHY. These will be used by the SysKonnect gigabit ethernet driver shortly.
2000-04-22 01:58:18 +00:00
struct mii_softc *sc;
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
struct ifnet *ifp;
Add PHY drivers for the XMAC II's internal PHY and the Broadcom BCM5400 1000baseTX PHY. These will be used by the SysKonnect gigabit ethernet driver shortly.
2000-04-22 01:58:18 +00:00
Move MII model and revision into softc.
2007-01-16 00:52:26 +00:00
bsc = device_get_softc(dev);
sc = &bsc->mii_sc;
- In the spirit of previous simplifications factor out the checks for a different PHY instance being selected and isolation out into the wrappers around the service methods rather than duplicating them over and over again (besides, a PHY driver shouldn't need to care about which instance it actually is). - Centralize the check for the need to isolate a non-zero PHY instance not supporting isolation in mii_mediachg() and just ignore it rather than panicing, which should sufficient given that a) things are likely to just work anyway if one doesn't plug in more than one port at a time and b) refusing to attach in this case just leaves us in a unknown but most likely also not exactly correct configuration (besides several drivers setting MIIF_NOISOLATE didn't care about these anyway, probably due to setting this flag for no real reason). - Minor fixes like removing unnecessary setting of sc->mii_anegticks, using sc->mii_anegticks instead of hardcoded values etc.
2010-10-02 18:53:12 +00:00
- Remove attempts to implement setting of BMCR_LOOP/MIIF_NOLOOP (reporting IFM_LOOP based on BMCR_LOOP is left in place though as it might provide useful for debugging). For most mii(4) drivers it was unclear whether the PHYs driven by them actually support loopback or not. Moreover, typically loopback mode also needs to be activated on the MAC, which none of the Ethernet drivers using mii(4) implements. Given that loopback media has no real use (and obviously hardly had a chance to actually work) besides for driver development (which just loopback mode should be sufficient for though, i.e one doesn't necessary need support for loopback media) support for it is just dropped as both NetBSD and OpenBSD already did quite some time ago. - Let mii_phy_add_media() also announce the support of IFM_NONE. - 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 lxtphy(4) which disables MII interrupts (as is done for a few other PHYs we have drivers for). This includes changing NIC drivers which previously just called the generic mii_phy_reset() to now actually call the PHY-specific reset routine, which might be crucial in some cases. While at it, the redundant checks in these NIC drivers for mii->mii_instance not being zero before calling the reset routines were removed because as soon as one PHY driver attaches mii->mii_instance is incremented and we hardly can end up in their media change callbacks etc if no PHY driver has attached as mii_attach() would have failed in that case and not attach a miibus(4) instance. Consequently, NIC drivers now no longer should call mii_phy_reset() directly, so it was removed from EXPORT_SYMS. - Add a mii_phy_dev_attach() as a companion helper to mii_phy_dev_probe(). The purpose of that function is to perform the common steps to attach a PHY driver instance and to hook it up to the miibus(4) instance and to optionally also handle the probing, addition and initialization of the supported media. So all a PHY driver without any special requirements has to do in its bus attach method is to call mii_phy_dev_attach() along with PHY-specific MIIF_* flags, a pointer to its PHY functions and the add_media set to one. All PHY drivers were updated to take advantage of mii_phy_dev_attach() as appropriate. Along with these changes the capability mask was added to the mii_softc structure so PHY drivers taking advantage of mii_phy_dev_attach() but still handling media on their own do not need to fiddle with the MII attach arguments anyway. - Keep track of the PHY offset in the mii_softc structure. This is done for compatibility with NetBSD/OpenBSD. - Keep track of the PHY's OUI, model and revision in the mii_softc structure. Several PHY drivers require this information also after attaching and previously had to wrap their own softc around mii_softc. NetBSD/OpenBSD also keep track of the model and revision on their mii_softc structure. All PHY drivers were updated to take advantage as appropriate. - Convert the mebers of the MII data structure to unsigned where appropriate. This is partly inspired by NetBSD/OpenBSD. - According to IEEE 802.3-2002 the bits actually have to be reversed when mapping an OUI to the MII ID registers. All PHY drivers and miidevs where changed as necessary. Actually this now again allows to largely share miidevs with NetBSD, which fixed this problem already 9 years ago. Consequently miidevs was synced as far as possible. - Add MIIF_NOMANPAUSE and mii_phy_flowstatus() calls to drivers that weren't explicitly converted to support flow control before. It's unclear whether flow control actually works with these but typically it should and their net behavior should be more correct with these changes in place than without if the MAC driver sets MIIF_DOPAUSE. Obtained from: NetBSD (partially) Reviewed by: yongari (earlier version), silence on arch@ and net@
2011-05-03 19:51:29 +00:00
mii_phy_dev_attach(dev, MIIF_NOISOLATE | MIIF_NOMANPAUSE,
&brgphy_funcs, 0);
Add PHY drivers for the XMAC II's internal PHY and the Broadcom BCM5400 1000baseTX PHY. These will be used by the SysKonnect gigabit ethernet driver shortly.
2000-04-22 01:58:18 +00:00
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
bsc->serdes_flags = 0;
Recognize 5720S PHY and treat it as 5708S PHY. Unfortunately 5720S uses 5709S PHY id so add a hack to detect 5720S PHY by checking parent device name. 5720S PHY does not support 2500SX. Tested by: Geans Pin < geanspin <> broadcom dot com >
2012-12-20 05:02:12 +00:00
ifp = sc->mii_pdata->mii_ifp;
/* Find the MAC driver associated with this PHY. */
if (strcmp(ifp->if_dname, "bge") == 0)
bge_sc = ifp->if_softc;
else if (strcmp(ifp->if_dname, "bce") == 0)
bce_sc = ifp->if_softc;
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
/* Handle any special cases based on the PHY ID */
- Remove attempts to implement setting of BMCR_LOOP/MIIF_NOLOOP (reporting IFM_LOOP based on BMCR_LOOP is left in place though as it might provide useful for debugging). For most mii(4) drivers it was unclear whether the PHYs driven by them actually support loopback or not. Moreover, typically loopback mode also needs to be activated on the MAC, which none of the Ethernet drivers using mii(4) implements. Given that loopback media has no real use (and obviously hardly had a chance to actually work) besides for driver development (which just loopback mode should be sufficient for though, i.e one doesn't necessary need support for loopback media) support for it is just dropped as both NetBSD and OpenBSD already did quite some time ago. - Let mii_phy_add_media() also announce the support of IFM_NONE. - 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 lxtphy(4) which disables MII interrupts (as is done for a few other PHYs we have drivers for). This includes changing NIC drivers which previously just called the generic mii_phy_reset() to now actually call the PHY-specific reset routine, which might be crucial in some cases. While at it, the redundant checks in these NIC drivers for mii->mii_instance not being zero before calling the reset routines were removed because as soon as one PHY driver attaches mii->mii_instance is incremented and we hardly can end up in their media change callbacks etc if no PHY driver has attached as mii_attach() would have failed in that case and not attach a miibus(4) instance. Consequently, NIC drivers now no longer should call mii_phy_reset() directly, so it was removed from EXPORT_SYMS. - Add a mii_phy_dev_attach() as a companion helper to mii_phy_dev_probe(). The purpose of that function is to perform the common steps to attach a PHY driver instance and to hook it up to the miibus(4) instance and to optionally also handle the probing, addition and initialization of the supported media. So all a PHY driver without any special requirements has to do in its bus attach method is to call mii_phy_dev_attach() along with PHY-specific MIIF_* flags, a pointer to its PHY functions and the add_media set to one. All PHY drivers were updated to take advantage of mii_phy_dev_attach() as appropriate. Along with these changes the capability mask was added to the mii_softc structure so PHY drivers taking advantage of mii_phy_dev_attach() but still handling media on their own do not need to fiddle with the MII attach arguments anyway. - Keep track of the PHY offset in the mii_softc structure. This is done for compatibility with NetBSD/OpenBSD. - Keep track of the PHY's OUI, model and revision in the mii_softc structure. Several PHY drivers require this information also after attaching and previously had to wrap their own softc around mii_softc. NetBSD/OpenBSD also keep track of the model and revision on their mii_softc structure. All PHY drivers were updated to take advantage as appropriate. - Convert the mebers of the MII data structure to unsigned where appropriate. This is partly inspired by NetBSD/OpenBSD. - According to IEEE 802.3-2002 the bits actually have to be reversed when mapping an OUI to the MII ID registers. All PHY drivers and miidevs where changed as necessary. Actually this now again allows to largely share miidevs with NetBSD, which fixed this problem already 9 years ago. Consequently miidevs was synced as far as possible. - Add MIIF_NOMANPAUSE and mii_phy_flowstatus() calls to drivers that weren't explicitly converted to support flow control before. It's unclear whether flow control actually works with these but typically it should and their net behavior should be more correct with these changes in place than without if the MAC driver sets MIIF_DOPAUSE. Obtained from: NetBSD (partially) Reviewed by: yongari (earlier version), silence on arch@ and net@
2011-05-03 19:51:29 +00:00
switch (sc->mii_mpd_oui) {
Remove whitespace at the end of line.
2008-08-12 00:52:10 +00:00
case MII_OUI_BROADCOM:
- Remove attempts to implement setting of BMCR_LOOP/MIIF_NOLOOP (reporting IFM_LOOP based on BMCR_LOOP is left in place though as it might provide useful for debugging). For most mii(4) drivers it was unclear whether the PHYs driven by them actually support loopback or not. Moreover, typically loopback mode also needs to be activated on the MAC, which none of the Ethernet drivers using mii(4) implements. Given that loopback media has no real use (and obviously hardly had a chance to actually work) besides for driver development (which just loopback mode should be sufficient for though, i.e one doesn't necessary need support for loopback media) support for it is just dropped as both NetBSD and OpenBSD already did quite some time ago. - Let mii_phy_add_media() also announce the support of IFM_NONE. - 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 lxtphy(4) which disables MII interrupts (as is done for a few other PHYs we have drivers for). This includes changing NIC drivers which previously just called the generic mii_phy_reset() to now actually call the PHY-specific reset routine, which might be crucial in some cases. While at it, the redundant checks in these NIC drivers for mii->mii_instance not being zero before calling the reset routines were removed because as soon as one PHY driver attaches mii->mii_instance is incremented and we hardly can end up in their media change callbacks etc if no PHY driver has attached as mii_attach() would have failed in that case and not attach a miibus(4) instance. Consequently, NIC drivers now no longer should call mii_phy_reset() directly, so it was removed from EXPORT_SYMS. - Add a mii_phy_dev_attach() as a companion helper to mii_phy_dev_probe(). The purpose of that function is to perform the common steps to attach a PHY driver instance and to hook it up to the miibus(4) instance and to optionally also handle the probing, addition and initialization of the supported media. So all a PHY driver without any special requirements has to do in its bus attach method is to call mii_phy_dev_attach() along with PHY-specific MIIF_* flags, a pointer to its PHY functions and the add_media set to one. All PHY drivers were updated to take advantage of mii_phy_dev_attach() as appropriate. Along with these changes the capability mask was added to the mii_softc structure so PHY drivers taking advantage of mii_phy_dev_attach() but still handling media on their own do not need to fiddle with the MII attach arguments anyway. - Keep track of the PHY offset in the mii_softc structure. This is done for compatibility with NetBSD/OpenBSD. - Keep track of the PHY's OUI, model and revision in the mii_softc structure. Several PHY drivers require this information also after attaching and previously had to wrap their own softc around mii_softc. NetBSD/OpenBSD also keep track of the model and revision on their mii_softc structure. All PHY drivers were updated to take advantage as appropriate. - Convert the mebers of the MII data structure to unsigned where appropriate. This is partly inspired by NetBSD/OpenBSD. - According to IEEE 802.3-2002 the bits actually have to be reversed when mapping an OUI to the MII ID registers. All PHY drivers and miidevs where changed as necessary. Actually this now again allows to largely share miidevs with NetBSD, which fixed this problem already 9 years ago. Consequently miidevs was synced as far as possible. - Add MIIF_NOMANPAUSE and mii_phy_flowstatus() calls to drivers that weren't explicitly converted to support flow control before. It's unclear whether flow control actually works with these but typically it should and their net behavior should be more correct with these changes in place than without if the MAC driver sets MIIF_DOPAUSE. Obtained from: NetBSD (partially) Reviewed by: yongari (earlier version), silence on arch@ and net@
2011-05-03 19:51:29 +00:00
switch (sc->mii_mpd_model) {
case MII_MODEL_BROADCOM_BCM5706:
case MII_MODEL_BROADCOM_BCM5714:
- Added support for 5709S/5716S PHYs. Submitted by: pyunyh MFC after: 2 weeks
2010-03-18 20:57:57 +00:00
/*
* The 5464 PHY used in the 5706 supports both copper
* and fiber interfaces over GMII. Need to check the
* shadow registers to see which mode is actually
* in effect, and therefore whether we have 5706C or
* 5706S.
*/
PHY_WRITE(sc, BRGPHY_MII_SHADOW_1C,
BRGPHY_SHADOW_1C_MODE_CTRL);
if (PHY_READ(sc, BRGPHY_MII_SHADOW_1C) &
BRGPHY_SHADOW_1C_ENA_1000X) {
bsc->serdes_flags |= BRGPHY_5706S;
sc->mii_flags |= MIIF_HAVEFIBER;
}
break;
- Probe BCM57780. - In case the parent is bge(4), don't set the Jumbo frame settings unless the MAC actually is Jumbo capable as otherwise the PHY might not have the corresponding registers implemented. This is also in line with what the Linux tg3 driver does. PR: 165032 Submitted by: Alexander Milanov Obtained from: OpenBSD MFC after: 3 days
2012-02-19 12:09:17 +00:00
}
break;
- Remove attempts to implement setting of BMCR_LOOP/MIIF_NOLOOP (reporting IFM_LOOP based on BMCR_LOOP is left in place though as it might provide useful for debugging). For most mii(4) drivers it was unclear whether the PHYs driven by them actually support loopback or not. Moreover, typically loopback mode also needs to be activated on the MAC, which none of the Ethernet drivers using mii(4) implements. Given that loopback media has no real use (and obviously hardly had a chance to actually work) besides for driver development (which just loopback mode should be sufficient for though, i.e one doesn't necessary need support for loopback media) support for it is just dropped as both NetBSD and OpenBSD already did quite some time ago. - Let mii_phy_add_media() also announce the support of IFM_NONE. - 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 lxtphy(4) which disables MII interrupts (as is done for a few other PHYs we have drivers for). This includes changing NIC drivers which previously just called the generic mii_phy_reset() to now actually call the PHY-specific reset routine, which might be crucial in some cases. While at it, the redundant checks in these NIC drivers for mii->mii_instance not being zero before calling the reset routines were removed because as soon as one PHY driver attaches mii->mii_instance is incremented and we hardly can end up in their media change callbacks etc if no PHY driver has attached as mii_attach() would have failed in that case and not attach a miibus(4) instance. Consequently, NIC drivers now no longer should call mii_phy_reset() directly, so it was removed from EXPORT_SYMS. - Add a mii_phy_dev_attach() as a companion helper to mii_phy_dev_probe(). The purpose of that function is to perform the common steps to attach a PHY driver instance and to hook it up to the miibus(4) instance and to optionally also handle the probing, addition and initialization of the supported media. So all a PHY driver without any special requirements has to do in its bus attach method is to call mii_phy_dev_attach() along with PHY-specific MIIF_* flags, a pointer to its PHY functions and the add_media set to one. All PHY drivers were updated to take advantage of mii_phy_dev_attach() as appropriate. Along with these changes the capability mask was added to the mii_softc structure so PHY drivers taking advantage of mii_phy_dev_attach() but still handling media on their own do not need to fiddle with the MII attach arguments anyway. - Keep track of the PHY offset in the mii_softc structure. This is done for compatibility with NetBSD/OpenBSD. - Keep track of the PHY's OUI, model and revision in the mii_softc structure. Several PHY drivers require this information also after attaching and previously had to wrap their own softc around mii_softc. NetBSD/OpenBSD also keep track of the model and revision on their mii_softc structure. All PHY drivers were updated to take advantage as appropriate. - Convert the mebers of the MII data structure to unsigned where appropriate. This is partly inspired by NetBSD/OpenBSD. - According to IEEE 802.3-2002 the bits actually have to be reversed when mapping an OUI to the MII ID registers. All PHY drivers and miidevs where changed as necessary. Actually this now again allows to largely share miidevs with NetBSD, which fixed this problem already 9 years ago. Consequently miidevs was synced as far as possible. - Add MIIF_NOMANPAUSE and mii_phy_flowstatus() calls to drivers that weren't explicitly converted to support flow control before. It's unclear whether flow control actually works with these but typically it should and their net behavior should be more correct with these changes in place than without if the MAC driver sets MIIF_DOPAUSE. Obtained from: NetBSD (partially) Reviewed by: yongari (earlier version), silence on arch@ and net@
2011-05-03 19:51:29 +00:00
case MII_OUI_BROADCOM2:
switch (sc->mii_mpd_model) {
case MII_MODEL_BROADCOM2_BCM5708S:
- Added support for 5709S/5716S PHYs. Submitted by: pyunyh MFC after: 2 weeks
2010-03-18 20:57:57 +00:00
bsc->serdes_flags |= BRGPHY_5708S;
sc->mii_flags |= MIIF_HAVEFIBER;
break;
- Remove attempts to implement setting of BMCR_LOOP/MIIF_NOLOOP (reporting IFM_LOOP based on BMCR_LOOP is left in place though as it might provide useful for debugging). For most mii(4) drivers it was unclear whether the PHYs driven by them actually support loopback or not. Moreover, typically loopback mode also needs to be activated on the MAC, which none of the Ethernet drivers using mii(4) implements. Given that loopback media has no real use (and obviously hardly had a chance to actually work) besides for driver development (which just loopback mode should be sufficient for though, i.e one doesn't necessary need support for loopback media) support for it is just dropped as both NetBSD and OpenBSD already did quite some time ago. - Let mii_phy_add_media() also announce the support of IFM_NONE. - 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 lxtphy(4) which disables MII interrupts (as is done for a few other PHYs we have drivers for). This includes changing NIC drivers which previously just called the generic mii_phy_reset() to now actually call the PHY-specific reset routine, which might be crucial in some cases. While at it, the redundant checks in these NIC drivers for mii->mii_instance not being zero before calling the reset routines were removed because as soon as one PHY driver attaches mii->mii_instance is incremented and we hardly can end up in their media change callbacks etc if no PHY driver has attached as mii_attach() would have failed in that case and not attach a miibus(4) instance. Consequently, NIC drivers now no longer should call mii_phy_reset() directly, so it was removed from EXPORT_SYMS. - Add a mii_phy_dev_attach() as a companion helper to mii_phy_dev_probe(). The purpose of that function is to perform the common steps to attach a PHY driver instance and to hook it up to the miibus(4) instance and to optionally also handle the probing, addition and initialization of the supported media. So all a PHY driver without any special requirements has to do in its bus attach method is to call mii_phy_dev_attach() along with PHY-specific MIIF_* flags, a pointer to its PHY functions and the add_media set to one. All PHY drivers were updated to take advantage of mii_phy_dev_attach() as appropriate. Along with these changes the capability mask was added to the mii_softc structure so PHY drivers taking advantage of mii_phy_dev_attach() but still handling media on their own do not need to fiddle with the MII attach arguments anyway. - Keep track of the PHY offset in the mii_softc structure. This is done for compatibility with NetBSD/OpenBSD. - Keep track of the PHY's OUI, model and revision in the mii_softc structure. Several PHY drivers require this information also after attaching and previously had to wrap their own softc around mii_softc. NetBSD/OpenBSD also keep track of the model and revision on their mii_softc structure. All PHY drivers were updated to take advantage as appropriate. - Convert the mebers of the MII data structure to unsigned where appropriate. This is partly inspired by NetBSD/OpenBSD. - According to IEEE 802.3-2002 the bits actually have to be reversed when mapping an OUI to the MII ID registers. All PHY drivers and miidevs where changed as necessary. Actually this now again allows to largely share miidevs with NetBSD, which fixed this problem already 9 years ago. Consequently miidevs was synced as far as possible. - Add MIIF_NOMANPAUSE and mii_phy_flowstatus() calls to drivers that weren't explicitly converted to support flow control before. It's unclear whether flow control actually works with these but typically it should and their net behavior should be more correct with these changes in place than without if the MAC driver sets MIIF_DOPAUSE. Obtained from: NetBSD (partially) Reviewed by: yongari (earlier version), silence on arch@ and net@
2011-05-03 19:51:29 +00:00
case MII_MODEL_BROADCOM2_BCM5709S:
Recognize 5720S PHY and treat it as 5708S PHY. Unfortunately 5720S uses 5709S PHY id so add a hack to detect 5720S PHY by checking parent device name. 5720S PHY does not support 2500SX. Tested by: Geans Pin < geanspin <> broadcom dot com >
2012-12-20 05:02:12 +00:00
/*
* XXX
* 5720S and 5709S shares the same PHY id.
* Assume 5720S PHY if parent device is bge(4).
*/
if (bge_sc != NULL)
bsc->serdes_flags |= BRGPHY_5708S;
else
bsc->serdes_flags |= BRGPHY_5709S;
Consistently use tab characters instead of tab + space characters. No functional changes.
2010-09-07 23:08:38 +00:00
sc->mii_flags |= MIIF_HAVEFIBER;
break;
}
break;
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
}
- Remove attempts to implement setting of BMCR_LOOP/MIIF_NOLOOP (reporting IFM_LOOP based on BMCR_LOOP is left in place though as it might provide useful for debugging). For most mii(4) drivers it was unclear whether the PHYs driven by them actually support loopback or not. Moreover, typically loopback mode also needs to be activated on the MAC, which none of the Ethernet drivers using mii(4) implements. Given that loopback media has no real use (and obviously hardly had a chance to actually work) besides for driver development (which just loopback mode should be sufficient for though, i.e one doesn't necessary need support for loopback media) support for it is just dropped as both NetBSD and OpenBSD already did quite some time ago. - Let mii_phy_add_media() also announce the support of IFM_NONE. - 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 lxtphy(4) which disables MII interrupts (as is done for a few other PHYs we have drivers for). This includes changing NIC drivers which previously just called the generic mii_phy_reset() to now actually call the PHY-specific reset routine, which might be crucial in some cases. While at it, the redundant checks in these NIC drivers for mii->mii_instance not being zero before calling the reset routines were removed because as soon as one PHY driver attaches mii->mii_instance is incremented and we hardly can end up in their media change callbacks etc if no PHY driver has attached as mii_attach() would have failed in that case and not attach a miibus(4) instance. Consequently, NIC drivers now no longer should call mii_phy_reset() directly, so it was removed from EXPORT_SYMS. - Add a mii_phy_dev_attach() as a companion helper to mii_phy_dev_probe(). The purpose of that function is to perform the common steps to attach a PHY driver instance and to hook it up to the miibus(4) instance and to optionally also handle the probing, addition and initialization of the supported media. So all a PHY driver without any special requirements has to do in its bus attach method is to call mii_phy_dev_attach() along with PHY-specific MIIF_* flags, a pointer to its PHY functions and the add_media set to one. All PHY drivers were updated to take advantage of mii_phy_dev_attach() as appropriate. Along with these changes the capability mask was added to the mii_softc structure so PHY drivers taking advantage of mii_phy_dev_attach() but still handling media on their own do not need to fiddle with the MII attach arguments anyway. - Keep track of the PHY offset in the mii_softc structure. This is done for compatibility with NetBSD/OpenBSD. - Keep track of the PHY's OUI, model and revision in the mii_softc structure. Several PHY drivers require this information also after attaching and previously had to wrap their own softc around mii_softc. NetBSD/OpenBSD also keep track of the model and revision on their mii_softc structure. All PHY drivers were updated to take advantage as appropriate. - Convert the mebers of the MII data structure to unsigned where appropriate. This is partly inspired by NetBSD/OpenBSD. - According to IEEE 802.3-2002 the bits actually have to be reversed when mapping an OUI to the MII ID registers. All PHY drivers and miidevs where changed as necessary. Actually this now again allows to largely share miidevs with NetBSD, which fixed this problem already 9 years ago. Consequently miidevs was synced as far as possible. - Add MIIF_NOMANPAUSE and mii_phy_flowstatus() calls to drivers that weren't explicitly converted to support flow control before. It's unclear whether flow control actually works with these but typically it should and their net behavior should be more correct with these changes in place than without if the MAC driver sets MIIF_DOPAUSE. Obtained from: NetBSD (partially) Reviewed by: yongari (earlier version), silence on arch@ and net@
2011-05-03 19:51:29 +00:00
PHY_RESET(sc);
Add PHY drivers for the XMAC II's internal PHY and the Broadcom BCM5400 1000baseTX PHY. These will be used by the SysKonnect gigabit ethernet driver shortly.
2000-04-22 01:58:18 +00:00
Revert r213867; while this driver really doesn't use any of the generic subroutines, at least mii_capabilities is used within itself.
2010-10-18 08:36:03 +00:00
/* Read the PHY's capabilities. */
- Remove attempts to implement setting of BMCR_LOOP/MIIF_NOLOOP (reporting IFM_LOOP based on BMCR_LOOP is left in place though as it might provide useful for debugging). For most mii(4) drivers it was unclear whether the PHYs driven by them actually support loopback or not. Moreover, typically loopback mode also needs to be activated on the MAC, which none of the Ethernet drivers using mii(4) implements. Given that loopback media has no real use (and obviously hardly had a chance to actually work) besides for driver development (which just loopback mode should be sufficient for though, i.e one doesn't necessary need support for loopback media) support for it is just dropped as both NetBSD and OpenBSD already did quite some time ago. - Let mii_phy_add_media() also announce the support of IFM_NONE. - 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 lxtphy(4) which disables MII interrupts (as is done for a few other PHYs we have drivers for). This includes changing NIC drivers which previously just called the generic mii_phy_reset() to now actually call the PHY-specific reset routine, which might be crucial in some cases. While at it, the redundant checks in these NIC drivers for mii->mii_instance not being zero before calling the reset routines were removed because as soon as one PHY driver attaches mii->mii_instance is incremented and we hardly can end up in their media change callbacks etc if no PHY driver has attached as mii_attach() would have failed in that case and not attach a miibus(4) instance. Consequently, NIC drivers now no longer should call mii_phy_reset() directly, so it was removed from EXPORT_SYMS. - Add a mii_phy_dev_attach() as a companion helper to mii_phy_dev_probe(). The purpose of that function is to perform the common steps to attach a PHY driver instance and to hook it up to the miibus(4) instance and to optionally also handle the probing, addition and initialization of the supported media. So all a PHY driver without any special requirements has to do in its bus attach method is to call mii_phy_dev_attach() along with PHY-specific MIIF_* flags, a pointer to its PHY functions and the add_media set to one. All PHY drivers were updated to take advantage of mii_phy_dev_attach() as appropriate. Along with these changes the capability mask was added to the mii_softc structure so PHY drivers taking advantage of mii_phy_dev_attach() but still handling media on their own do not need to fiddle with the MII attach arguments anyway. - Keep track of the PHY offset in the mii_softc structure. This is done for compatibility with NetBSD/OpenBSD. - Keep track of the PHY's OUI, model and revision in the mii_softc structure. Several PHY drivers require this information also after attaching and previously had to wrap their own softc around mii_softc. NetBSD/OpenBSD also keep track of the model and revision on their mii_softc structure. All PHY drivers were updated to take advantage as appropriate. - Convert the mebers of the MII data structure to unsigned where appropriate. This is partly inspired by NetBSD/OpenBSD. - According to IEEE 802.3-2002 the bits actually have to be reversed when mapping an OUI to the MII ID registers. All PHY drivers and miidevs where changed as necessary. Actually this now again allows to largely share miidevs with NetBSD, which fixed this problem already 9 years ago. Consequently miidevs was synced as far as possible. - Add MIIF_NOMANPAUSE and mii_phy_flowstatus() calls to drivers that weren't explicitly converted to support flow control before. It's unclear whether flow control actually works with these but typically it should and their net behavior should be more correct with these changes in place than without if the MAC driver sets MIIF_DOPAUSE. Obtained from: NetBSD (partially) Reviewed by: yongari (earlier version), silence on arch@ and net@
2011-05-03 19:51:29 +00:00
sc->mii_capabilities = PHY_READ(sc, MII_BMSR) & sc->mii_capmask;
Revert r213867; while this driver really doesn't use any of the generic subroutines, at least mii_capabilities is used within itself.
2010-10-18 08:36:03 +00:00
if (sc->mii_capabilities & BMSR_EXTSTAT)
sc->mii_extcapabilities = PHY_READ(sc, MII_EXTSR);
Add PHY drivers for the XMAC II's internal PHY and the Broadcom BCM5400 1000baseTX PHY. These will be used by the SysKonnect gigabit ethernet driver shortly.
2000-04-22 01:58:18 +00:00
device_printf(dev, " ");
- Remove attempts to implement setting of BMCR_LOOP/MIIF_NOLOOP (reporting IFM_LOOP based on BMCR_LOOP is left in place though as it might provide useful for debugging). For most mii(4) drivers it was unclear whether the PHYs driven by them actually support loopback or not. Moreover, typically loopback mode also needs to be activated on the MAC, which none of the Ethernet drivers using mii(4) implements. Given that loopback media has no real use (and obviously hardly had a chance to actually work) besides for driver development (which just loopback mode should be sufficient for though, i.e one doesn't necessary need support for loopback media) support for it is just dropped as both NetBSD and OpenBSD already did quite some time ago. - Let mii_phy_add_media() also announce the support of IFM_NONE. - 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 lxtphy(4) which disables MII interrupts (as is done for a few other PHYs we have drivers for). This includes changing NIC drivers which previously just called the generic mii_phy_reset() to now actually call the PHY-specific reset routine, which might be crucial in some cases. While at it, the redundant checks in these NIC drivers for mii->mii_instance not being zero before calling the reset routines were removed because as soon as one PHY driver attaches mii->mii_instance is incremented and we hardly can end up in their media change callbacks etc if no PHY driver has attached as mii_attach() would have failed in that case and not attach a miibus(4) instance. Consequently, NIC drivers now no longer should call mii_phy_reset() directly, so it was removed from EXPORT_SYMS. - Add a mii_phy_dev_attach() as a companion helper to mii_phy_dev_probe(). The purpose of that function is to perform the common steps to attach a PHY driver instance and to hook it up to the miibus(4) instance and to optionally also handle the probing, addition and initialization of the supported media. So all a PHY driver without any special requirements has to do in its bus attach method is to call mii_phy_dev_attach() along with PHY-specific MIIF_* flags, a pointer to its PHY functions and the add_media set to one. All PHY drivers were updated to take advantage of mii_phy_dev_attach() as appropriate. Along with these changes the capability mask was added to the mii_softc structure so PHY drivers taking advantage of mii_phy_dev_attach() but still handling media on their own do not need to fiddle with the MII attach arguments anyway. - Keep track of the PHY offset in the mii_softc structure. This is done for compatibility with NetBSD/OpenBSD. - Keep track of the PHY's OUI, model and revision in the mii_softc structure. Several PHY drivers require this information also after attaching and previously had to wrap their own softc around mii_softc. NetBSD/OpenBSD also keep track of the model and revision on their mii_softc structure. All PHY drivers were updated to take advantage as appropriate. - Convert the mebers of the MII data structure to unsigned where appropriate. This is partly inspired by NetBSD/OpenBSD. - According to IEEE 802.3-2002 the bits actually have to be reversed when mapping an OUI to the MII ID registers. All PHY drivers and miidevs where changed as necessary. Actually this now again allows to largely share miidevs with NetBSD, which fixed this problem already 9 years ago. Consequently miidevs was synced as far as possible. - Add MIIF_NOMANPAUSE and mii_phy_flowstatus() calls to drivers that weren't explicitly converted to support flow control before. It's unclear whether flow control actually works with these but typically it should and their net behavior should be more correct with these changes in place than without if the MAC driver sets MIIF_DOPAUSE. Obtained from: NetBSD (partially) Reviewed by: yongari (earlier version), silence on arch@ and net@
2011-05-03 19:51:29 +00:00
#define ADD(m, c) ifmedia_add(&sc->mii_pdata->mii_media, (m), (c), NULL)
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
/* Add the supported media types */
if ((sc->mii_flags & MIIF_HAVEFIBER) == 0) {
o Flesh out the generic IEEE 802.3 annex 31B full duplex flow control support in mii(4): - Merge generic flow control advertisement (which can be enabled by passing by MIIF_DOPAUSE to mii_attach(9)) and parsing support from NetBSD into mii_physubr.c and ukphy_subr.c. Unlike as in NetBSD, IFM_FLOW isn't implemented as a global option via the "don't care mask" but instead as a media specific option this. This has the following advantages: o allows flow control advertisement with autonegotiation to be turned on and off via ifconfig(8) with the default typically being off (though MIIF_FORCEPAUSE has been added causing flow control to be always advertised, allowing to easily MFC this changes for drivers that previously used home-grown support for flow control that behaved that way without breaking POLA) o allows to deal with PHY drivers where flow control advertisement with manual selection doesn't work or at least isn't implemented, like it's the case with brgphy(4), e1000phy(4) and ip1000phy(4), by setting MIIF_NOMANPAUSE o the available combinations of media options are readily available from the `ifconfig -m` output - Add IFM_FLOW to IFM_SHARED_OPTION_DESCRIPTIONS and IFM_ETH_RXPAUSE and IFM_ETH_TXPAUSE to IFM_SUBTYPE_ETHERNET_OPTION_DESCRIPTIONS so these are understood by ifconfig(8). o Make the master/slave support in mii(4) actually usable: - Change IFM_ETH_MASTER from being implemented as a global option via the "don't care mask" to a media specific one as it actually is only applicable to IFM_1000_T to date. - Let mii_phy_setmedia() set GTCR_MAN_MS in IFM_1000_T slave mode to actually configure manually selected slave mode (like we also do in the PHY specific implementations). - Add IFM_ETH_MASTER to IFM_SUBTYPE_ETHERNET_OPTION_DESCRIPTIONS so it is understood by ifconfig(8). o Switch bge(4), bce(4), msk(4), nfe(4) and stge(4) along with brgphy(4), e1000phy(4) and ip1000phy(4) to use the generic flow control support instead of home-grown solutions via IFM_FLAGs. This includes changing these PHY drivers and smcphy(4) to no longer unconditionally advertise support for flow control but only if the selected media has IFM_FLOW set (or MIIF_FORCEPAUSE is set) and implemented for these media variants, i.e. typically only for copper. o Switch brgphy(4), ciphy(4), e1000phy(4) and ip1000phy(4) to report and set IFM_1000_T master mode via IFM_ETH_MASTER instead of via IFF_LINK0 and some IFM_FLAGn. o Switch brgphy(4) to add at least the the supported copper media based on the contents of the BMSR via mii_phy_add_media() instead of hardcoding them. The latter approach seems to have developed historically, besides causing unnecessary code duplication it was also undesirable because brgphy_mii_phy_auto() already based the capability advertisement on the contents of the BMSR though. o Let brgphy(4) set IFM_1000_T master mode on all supported PHY and not just BCM5701. Apparently this was a misinterpretation of a workaround in the Linux tg3 driver; BCM5701 seem to require RGPHY_1000CTL_MSE and BRGPHY_1000CTL_MSC to be set when configuring autonegotiation but this doesn't mean we can't set these as well on other PHYs for manual media selection. o Let ukphy_status() report IFM_1000_T master mode via IFM_ETH_MASTER so IFM_1000_T master mode support now is generally available with all PHY drivers. o Don't let e1000phy(4) set master/slave bits for IFM_1000_SX as it's not applicable there. Reviewed by: yongari (plus additional testing) Obtained from: NetBSD (partially), OpenBSD (partially) MFC after: 2 weeks
2010-11-14 13:26:10 +00:00
mii_phy_add_media(sc);
printf("\n");
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
} else {
o Flesh out the generic IEEE 802.3 annex 31B full duplex flow control support in mii(4): - Merge generic flow control advertisement (which can be enabled by passing by MIIF_DOPAUSE to mii_attach(9)) and parsing support from NetBSD into mii_physubr.c and ukphy_subr.c. Unlike as in NetBSD, IFM_FLOW isn't implemented as a global option via the "don't care mask" but instead as a media specific option this. This has the following advantages: o allows flow control advertisement with autonegotiation to be turned on and off via ifconfig(8) with the default typically being off (though MIIF_FORCEPAUSE has been added causing flow control to be always advertised, allowing to easily MFC this changes for drivers that previously used home-grown support for flow control that behaved that way without breaking POLA) o allows to deal with PHY drivers where flow control advertisement with manual selection doesn't work or at least isn't implemented, like it's the case with brgphy(4), e1000phy(4) and ip1000phy(4), by setting MIIF_NOMANPAUSE o the available combinations of media options are readily available from the `ifconfig -m` output - Add IFM_FLOW to IFM_SHARED_OPTION_DESCRIPTIONS and IFM_ETH_RXPAUSE and IFM_ETH_TXPAUSE to IFM_SUBTYPE_ETHERNET_OPTION_DESCRIPTIONS so these are understood by ifconfig(8). o Make the master/slave support in mii(4) actually usable: - Change IFM_ETH_MASTER from being implemented as a global option via the "don't care mask" to a media specific one as it actually is only applicable to IFM_1000_T to date. - Let mii_phy_setmedia() set GTCR_MAN_MS in IFM_1000_T slave mode to actually configure manually selected slave mode (like we also do in the PHY specific implementations). - Add IFM_ETH_MASTER to IFM_SUBTYPE_ETHERNET_OPTION_DESCRIPTIONS so it is understood by ifconfig(8). o Switch bge(4), bce(4), msk(4), nfe(4) and stge(4) along with brgphy(4), e1000phy(4) and ip1000phy(4) to use the generic flow control support instead of home-grown solutions via IFM_FLAGs. This includes changing these PHY drivers and smcphy(4) to no longer unconditionally advertise support for flow control but only if the selected media has IFM_FLOW set (or MIIF_FORCEPAUSE is set) and implemented for these media variants, i.e. typically only for copper. o Switch brgphy(4), ciphy(4), e1000phy(4) and ip1000phy(4) to report and set IFM_1000_T master mode via IFM_ETH_MASTER instead of via IFF_LINK0 and some IFM_FLAGn. o Switch brgphy(4) to add at least the the supported copper media based on the contents of the BMSR via mii_phy_add_media() instead of hardcoding them. The latter approach seems to have developed historically, besides causing unnecessary code duplication it was also undesirable because brgphy_mii_phy_auto() already based the capability advertisement on the contents of the BMSR though. o Let brgphy(4) set IFM_1000_T master mode on all supported PHY and not just BCM5701. Apparently this was a misinterpretation of a workaround in the Linux tg3 driver; BCM5701 seem to require RGPHY_1000CTL_MSE and BRGPHY_1000CTL_MSC to be set when configuring autonegotiation but this doesn't mean we can't set these as well on other PHYs for manual media selection. o Let ukphy_status() report IFM_1000_T master mode via IFM_ETH_MASTER so IFM_1000_T master mode support now is generally available with all PHY drivers. o Don't let e1000phy(4) set master/slave bits for IFM_1000_SX as it's not applicable there. Reviewed by: yongari (plus additional testing) Obtained from: NetBSD (partially), OpenBSD (partially) MFC after: 2 weeks
2010-11-14 13:26:10 +00:00
sc->mii_anegticks = MII_ANEGTICKS_GIGE;
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
ADD(IFM_MAKEWORD(IFM_ETHER, IFM_1000_SX, IFM_FDX, sc->mii_inst),
BRGPHY_S1000 | BRGPHY_BMCR_FDX);
printf("1000baseSX-FDX, ");
- Added support for BCM5709 and BCM5716. MFC after: 2 weeks
2008-06-13 01:20:29 +00:00
/* 2.5G support is a software enabled feature on the 5708S and 5709S. */
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
if (bce_sc && (bce_sc->bce_phy_flags & BCE_PHY_2_5G_CAPABLE_FLAG)) {
ADD(IFM_MAKEWORD(IFM_ETHER, IFM_2500_SX, IFM_FDX, sc->mii_inst), 0);
printf("2500baseSX-FDX, ");
Provide workaround for the ages old bug affecting certain BCM5708S chip revision often found in the blades and resulting in interfaces not sensing carrier signal. Looking at all problem reports it appears that it only affects some very specific silicon revision (ASIC (0x57081021); Rev (B2)) and version of the PHY that supports 1000baseSX-FDX media only. Therefore, narrow the scope of workaround to combination of that revision and media type. Given that the first report on this issue is dated back to 2007, there is not much hope that this issue will ever be properly resolved. Among affected systems are IBM HS21, Intel SBXD132 and HP BL460c. PR: 118238, 122551, 140970 MFC after: 1 month
2010-03-10 05:19:14 +00:00
} else if ((bsc->serdes_flags & BRGPHY_5708S) && bce_sc &&
further narrow down no carrier workaround, since it appears to only affect very specific IBM hardware and other machines with the same BCM ASIC chip id 0x57081021 are just fine. MFC after: 1 month
2010-03-10 23:00:15 +00:00
(detect_hs21(bce_sc) != 0)) {
Provide workaround for the ages old bug affecting certain BCM5708S chip revision often found in the blades and resulting in interfaces not sensing carrier signal. Looking at all problem reports it appears that it only affects some very specific silicon revision (ASIC (0x57081021); Rev (B2)) and version of the PHY that supports 1000baseSX-FDX media only. Therefore, narrow the scope of workaround to combination of that revision and media type. Given that the first report on this issue is dated back to 2007, there is not much hope that this issue will ever be properly resolved. Among affected systems are IBM HS21, Intel SBXD132 and HP BL460c. PR: 118238, 122551, 140970 MFC after: 1 month
2010-03-10 05:19:14 +00:00
/*
* There appears to be certain silicon revision
further narrow down no carrier workaround, since it appears to only affect very specific IBM hardware and other machines with the same BCM ASIC chip id 0x57081021 are just fine. MFC after: 1 month
2010-03-10 23:00:15 +00:00
* in IBM HS21 blades that is having issues with
Provide workaround for the ages old bug affecting certain BCM5708S chip revision often found in the blades and resulting in interfaces not sensing carrier signal. Looking at all problem reports it appears that it only affects some very specific silicon revision (ASIC (0x57081021); Rev (B2)) and version of the PHY that supports 1000baseSX-FDX media only. Therefore, narrow the scope of workaround to combination of that revision and media type. Given that the first report on this issue is dated back to 2007, there is not much hope that this issue will ever be properly resolved. Among affected systems are IBM HS21, Intel SBXD132 and HP BL460c. PR: 118238, 122551, 140970 MFC after: 1 month
2010-03-10 05:19:14 +00:00
* this driver wating for the auto-negotiation to
* complete. This happens with a specific chip id
* only and when the 1000baseSX-FDX is the only
* mode. Workaround this issue since it's unlikely
* to be ever addressed.
*/
printf("auto-neg workaround, ");
bsc->serdes_flags |= BRGPHY_NOANWAIT;
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
}
o Flesh out the generic IEEE 802.3 annex 31B full duplex flow control support in mii(4): - Merge generic flow control advertisement (which can be enabled by passing by MIIF_DOPAUSE to mii_attach(9)) and parsing support from NetBSD into mii_physubr.c and ukphy_subr.c. Unlike as in NetBSD, IFM_FLOW isn't implemented as a global option via the "don't care mask" but instead as a media specific option this. This has the following advantages: o allows flow control advertisement with autonegotiation to be turned on and off via ifconfig(8) with the default typically being off (though MIIF_FORCEPAUSE has been added causing flow control to be always advertised, allowing to easily MFC this changes for drivers that previously used home-grown support for flow control that behaved that way without breaking POLA) o allows to deal with PHY drivers where flow control advertisement with manual selection doesn't work or at least isn't implemented, like it's the case with brgphy(4), e1000phy(4) and ip1000phy(4), by setting MIIF_NOMANPAUSE o the available combinations of media options are readily available from the `ifconfig -m` output - Add IFM_FLOW to IFM_SHARED_OPTION_DESCRIPTIONS and IFM_ETH_RXPAUSE and IFM_ETH_TXPAUSE to IFM_SUBTYPE_ETHERNET_OPTION_DESCRIPTIONS so these are understood by ifconfig(8). o Make the master/slave support in mii(4) actually usable: - Change IFM_ETH_MASTER from being implemented as a global option via the "don't care mask" to a media specific one as it actually is only applicable to IFM_1000_T to date. - Let mii_phy_setmedia() set GTCR_MAN_MS in IFM_1000_T slave mode to actually configure manually selected slave mode (like we also do in the PHY specific implementations). - Add IFM_ETH_MASTER to IFM_SUBTYPE_ETHERNET_OPTION_DESCRIPTIONS so it is understood by ifconfig(8). o Switch bge(4), bce(4), msk(4), nfe(4) and stge(4) along with brgphy(4), e1000phy(4) and ip1000phy(4) to use the generic flow control support instead of home-grown solutions via IFM_FLAGs. This includes changing these PHY drivers and smcphy(4) to no longer unconditionally advertise support for flow control but only if the selected media has IFM_FLOW set (or MIIF_FORCEPAUSE is set) and implemented for these media variants, i.e. typically only for copper. o Switch brgphy(4), ciphy(4), e1000phy(4) and ip1000phy(4) to report and set IFM_1000_T master mode via IFM_ETH_MASTER instead of via IFF_LINK0 and some IFM_FLAGn. o Switch brgphy(4) to add at least the the supported copper media based on the contents of the BMSR via mii_phy_add_media() instead of hardcoding them. The latter approach seems to have developed historically, besides causing unnecessary code duplication it was also undesirable because brgphy_mii_phy_auto() already based the capability advertisement on the contents of the BMSR though. o Let brgphy(4) set IFM_1000_T master mode on all supported PHY and not just BCM5701. Apparently this was a misinterpretation of a workaround in the Linux tg3 driver; BCM5701 seem to require RGPHY_1000CTL_MSE and BRGPHY_1000CTL_MSC to be set when configuring autonegotiation but this doesn't mean we can't set these as well on other PHYs for manual media selection. o Let ukphy_status() report IFM_1000_T master mode via IFM_ETH_MASTER so IFM_1000_T master mode support now is generally available with all PHY drivers. o Don't let e1000phy(4) set master/slave bits for IFM_1000_SX as it's not applicable there. Reviewed by: yongari (plus additional testing) Obtained from: NetBSD (partially), OpenBSD (partially) MFC after: 2 weeks
2010-11-14 13:26:10 +00:00
ADD(IFM_MAKEWORD(IFM_ETHER, IFM_AUTO, 0, sc->mii_inst), 0);
printf("auto\n");
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
}
#undef ADD
Add PHY drivers for the XMAC II's internal PHY and the Broadcom BCM5400 1000baseTX PHY. These will be used by the SysKonnect gigabit ethernet driver shortly.
2000-04-22 01:58:18 +00:00
MIIBUS_MEDIAINIT(sc->mii_dev);
Some style changes to a couple of PHY drivers: - Fix some whitespace nits. - Fix some spelling in comments. - Use MII_ANEGTICKS instead of 5. - Don't define variables in nested scope. - Remove superfluous returns at the end of void functions. - Remove unused static global rgephy_mii_model. - Remove dupe $Id$ in tdkphy(4). - Sort brgphys table. MFC after: 2 weeks
2006-12-02 19:36:25 +00:00
return (0);
Add PHY drivers for the XMAC II's internal PHY and the Broadcom BCM5400 1000baseTX PHY. These will be used by the SysKonnect gigabit ethernet driver shortly.
2000-04-22 01:58:18 +00:00
}
Cleanup pass for mii drivers. . Make internal service routines static. . Use a consistent ordering of checks in MII_TICK. Do the work in the mii_phy_tick() subroutine if appropriate. . Call mii_phy_update() to trigger the callbacks.
2001-09-29 19:18:52 +00:00
static int
Use ansi function definitions in preference to K&R to reduce diffs with NetBSD (and cause it looks cooler).
2005-09-30 19:39:27 +00:00
brgphy_service(struct mii_softc *sc, struct mii_data *mii, int cmd)
Add PHY drivers for the XMAC II's internal PHY and the Broadcom BCM5400 1000baseTX PHY. These will be used by the SysKonnect gigabit ethernet driver shortly.
2000-04-22 01:58:18 +00:00
{
struct ifmedia_entry *ife = mii->mii_media.ifm_cur;
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
int val;
Add PHY drivers for the XMAC II's internal PHY and the Broadcom BCM5400 1000baseTX PHY. These will be used by the SysKonnect gigabit ethernet driver shortly.
2000-04-22 01:58:18 +00:00
switch (cmd) {
case MII_POLLSTAT:
break;
case MII_MEDIACHG:
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
/* Todo: Why is this here? Is it really needed? */
- Remove attempts to implement setting of BMCR_LOOP/MIIF_NOLOOP (reporting IFM_LOOP based on BMCR_LOOP is left in place though as it might provide useful for debugging). For most mii(4) drivers it was unclear whether the PHYs driven by them actually support loopback or not. Moreover, typically loopback mode also needs to be activated on the MAC, which none of the Ethernet drivers using mii(4) implements. Given that loopback media has no real use (and obviously hardly had a chance to actually work) besides for driver development (which just loopback mode should be sufficient for though, i.e one doesn't necessary need support for loopback media) support for it is just dropped as both NetBSD and OpenBSD already did quite some time ago. - Let mii_phy_add_media() also announce the support of IFM_NONE. - 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 lxtphy(4) which disables MII interrupts (as is done for a few other PHYs we have drivers for). This includes changing NIC drivers which previously just called the generic mii_phy_reset() to now actually call the PHY-specific reset routine, which might be crucial in some cases. While at it, the redundant checks in these NIC drivers for mii->mii_instance not being zero before calling the reset routines were removed because as soon as one PHY driver attaches mii->mii_instance is incremented and we hardly can end up in their media change callbacks etc if no PHY driver has attached as mii_attach() would have failed in that case and not attach a miibus(4) instance. Consequently, NIC drivers now no longer should call mii_phy_reset() directly, so it was removed from EXPORT_SYMS. - Add a mii_phy_dev_attach() as a companion helper to mii_phy_dev_probe(). The purpose of that function is to perform the common steps to attach a PHY driver instance and to hook it up to the miibus(4) instance and to optionally also handle the probing, addition and initialization of the supported media. So all a PHY driver without any special requirements has to do in its bus attach method is to call mii_phy_dev_attach() along with PHY-specific MIIF_* flags, a pointer to its PHY functions and the add_media set to one. All PHY drivers were updated to take advantage of mii_phy_dev_attach() as appropriate. Along with these changes the capability mask was added to the mii_softc structure so PHY drivers taking advantage of mii_phy_dev_attach() but still handling media on their own do not need to fiddle with the MII attach arguments anyway. - Keep track of the PHY offset in the mii_softc structure. This is done for compatibility with NetBSD/OpenBSD. - Keep track of the PHY's OUI, model and revision in the mii_softc structure. Several PHY drivers require this information also after attaching and previously had to wrap their own softc around mii_softc. NetBSD/OpenBSD also keep track of the model and revision on their mii_softc structure. All PHY drivers were updated to take advantage as appropriate. - Convert the mebers of the MII data structure to unsigned where appropriate. This is partly inspired by NetBSD/OpenBSD. - According to IEEE 802.3-2002 the bits actually have to be reversed when mapping an OUI to the MII ID registers. All PHY drivers and miidevs where changed as necessary. Actually this now again allows to largely share miidevs with NetBSD, which fixed this problem already 9 years ago. Consequently miidevs was synced as far as possible. - Add MIIF_NOMANPAUSE and mii_phy_flowstatus() calls to drivers that weren't explicitly converted to support flow control before. It's unclear whether flow control actually works with these but typically it should and their net behavior should be more correct with these changes in place than without if the MAC driver sets MIIF_DOPAUSE. Obtained from: NetBSD (partially) Reviewed by: yongari (earlier version), silence on arch@ and net@
2011-05-03 19:51:29 +00:00
PHY_RESET(sc); /* XXX hardware bug work-around */
Add PHY drivers for the XMAC II's internal PHY and the Broadcom BCM5400 1000baseTX PHY. These will be used by the SysKonnect gigabit ethernet driver shortly.
2000-04-22 01:58:18 +00:00
switch (IFM_SUBTYPE(ife->ifm_media)) {
case IFM_AUTO:
o Flesh out the generic IEEE 802.3 annex 31B full duplex flow control support in mii(4): - Merge generic flow control advertisement (which can be enabled by passing by MIIF_DOPAUSE to mii_attach(9)) and parsing support from NetBSD into mii_physubr.c and ukphy_subr.c. Unlike as in NetBSD, IFM_FLOW isn't implemented as a global option via the "don't care mask" but instead as a media specific option this. This has the following advantages: o allows flow control advertisement with autonegotiation to be turned on and off via ifconfig(8) with the default typically being off (though MIIF_FORCEPAUSE has been added causing flow control to be always advertised, allowing to easily MFC this changes for drivers that previously used home-grown support for flow control that behaved that way without breaking POLA) o allows to deal with PHY drivers where flow control advertisement with manual selection doesn't work or at least isn't implemented, like it's the case with brgphy(4), e1000phy(4) and ip1000phy(4), by setting MIIF_NOMANPAUSE o the available combinations of media options are readily available from the `ifconfig -m` output - Add IFM_FLOW to IFM_SHARED_OPTION_DESCRIPTIONS and IFM_ETH_RXPAUSE and IFM_ETH_TXPAUSE to IFM_SUBTYPE_ETHERNET_OPTION_DESCRIPTIONS so these are understood by ifconfig(8). o Make the master/slave support in mii(4) actually usable: - Change IFM_ETH_MASTER from being implemented as a global option via the "don't care mask" to a media specific one as it actually is only applicable to IFM_1000_T to date. - Let mii_phy_setmedia() set GTCR_MAN_MS in IFM_1000_T slave mode to actually configure manually selected slave mode (like we also do in the PHY specific implementations). - Add IFM_ETH_MASTER to IFM_SUBTYPE_ETHERNET_OPTION_DESCRIPTIONS so it is understood by ifconfig(8). o Switch bge(4), bce(4), msk(4), nfe(4) and stge(4) along with brgphy(4), e1000phy(4) and ip1000phy(4) to use the generic flow control support instead of home-grown solutions via IFM_FLAGs. This includes changing these PHY drivers and smcphy(4) to no longer unconditionally advertise support for flow control but only if the selected media has IFM_FLOW set (or MIIF_FORCEPAUSE is set) and implemented for these media variants, i.e. typically only for copper. o Switch brgphy(4), ciphy(4), e1000phy(4) and ip1000phy(4) to report and set IFM_1000_T master mode via IFM_ETH_MASTER instead of via IFF_LINK0 and some IFM_FLAGn. o Switch brgphy(4) to add at least the the supported copper media based on the contents of the BMSR via mii_phy_add_media() instead of hardcoding them. The latter approach seems to have developed historically, besides causing unnecessary code duplication it was also undesirable because brgphy_mii_phy_auto() already based the capability advertisement on the contents of the BMSR though. o Let brgphy(4) set IFM_1000_T master mode on all supported PHY and not just BCM5701. Apparently this was a misinterpretation of a workaround in the Linux tg3 driver; BCM5701 seem to require RGPHY_1000CTL_MSE and BRGPHY_1000CTL_MSC to be set when configuring autonegotiation but this doesn't mean we can't set these as well on other PHYs for manual media selection. o Let ukphy_status() report IFM_1000_T master mode via IFM_ETH_MASTER so IFM_1000_T master mode support now is generally available with all PHY drivers. o Don't let e1000phy(4) set master/slave bits for IFM_1000_SX as it's not applicable there. Reviewed by: yongari (plus additional testing) Obtained from: NetBSD (partially), OpenBSD (partially) MFC after: 2 weeks
2010-11-14 13:26:10 +00:00
brgphy_mii_phy_auto(sc, ife->ifm_media);
Add PHY drivers for the XMAC II's internal PHY and the Broadcom BCM5400 1000baseTX PHY. These will be used by the SysKonnect gigabit ethernet driver shortly.
2000-04-22 01:58:18 +00:00
break;
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
case IFM_2500_SX:
case IFM_1000_SX:
Follow NetBSD and s/IFM_1000_TX/IFM_1000_T/
2002-04-28 20:34:20 +00:00
case IFM_1000_T:
Add support for the BCM5401 and BCM5411 10/100/1000Mbps copper gigE PHYs. This basically updates the brgphy driver to support 10/100 modes in addition to 1000Mbps modes.
2001-09-04 22:00:33 +00:00
case IFM_100_TX:
case IFM_10_T:
o Flesh out the generic IEEE 802.3 annex 31B full duplex flow control support in mii(4): - Merge generic flow control advertisement (which can be enabled by passing by MIIF_DOPAUSE to mii_attach(9)) and parsing support from NetBSD into mii_physubr.c and ukphy_subr.c. Unlike as in NetBSD, IFM_FLOW isn't implemented as a global option via the "don't care mask" but instead as a media specific option this. This has the following advantages: o allows flow control advertisement with autonegotiation to be turned on and off via ifconfig(8) with the default typically being off (though MIIF_FORCEPAUSE has been added causing flow control to be always advertised, allowing to easily MFC this changes for drivers that previously used home-grown support for flow control that behaved that way without breaking POLA) o allows to deal with PHY drivers where flow control advertisement with manual selection doesn't work or at least isn't implemented, like it's the case with brgphy(4), e1000phy(4) and ip1000phy(4), by setting MIIF_NOMANPAUSE o the available combinations of media options are readily available from the `ifconfig -m` output - Add IFM_FLOW to IFM_SHARED_OPTION_DESCRIPTIONS and IFM_ETH_RXPAUSE and IFM_ETH_TXPAUSE to IFM_SUBTYPE_ETHERNET_OPTION_DESCRIPTIONS so these are understood by ifconfig(8). o Make the master/slave support in mii(4) actually usable: - Change IFM_ETH_MASTER from being implemented as a global option via the "don't care mask" to a media specific one as it actually is only applicable to IFM_1000_T to date. - Let mii_phy_setmedia() set GTCR_MAN_MS in IFM_1000_T slave mode to actually configure manually selected slave mode (like we also do in the PHY specific implementations). - Add IFM_ETH_MASTER to IFM_SUBTYPE_ETHERNET_OPTION_DESCRIPTIONS so it is understood by ifconfig(8). o Switch bge(4), bce(4), msk(4), nfe(4) and stge(4) along with brgphy(4), e1000phy(4) and ip1000phy(4) to use the generic flow control support instead of home-grown solutions via IFM_FLAGs. This includes changing these PHY drivers and smcphy(4) to no longer unconditionally advertise support for flow control but only if the selected media has IFM_FLOW set (or MIIF_FORCEPAUSE is set) and implemented for these media variants, i.e. typically only for copper. o Switch brgphy(4), ciphy(4), e1000phy(4) and ip1000phy(4) to report and set IFM_1000_T master mode via IFM_ETH_MASTER instead of via IFF_LINK0 and some IFM_FLAGn. o Switch brgphy(4) to add at least the the supported copper media based on the contents of the BMSR via mii_phy_add_media() instead of hardcoding them. The latter approach seems to have developed historically, besides causing unnecessary code duplication it was also undesirable because brgphy_mii_phy_auto() already based the capability advertisement on the contents of the BMSR though. o Let brgphy(4) set IFM_1000_T master mode on all supported PHY and not just BCM5701. Apparently this was a misinterpretation of a workaround in the Linux tg3 driver; BCM5701 seem to require RGPHY_1000CTL_MSE and BRGPHY_1000CTL_MSC to be set when configuring autonegotiation but this doesn't mean we can't set these as well on other PHYs for manual media selection. o Let ukphy_status() report IFM_1000_T master mode via IFM_ETH_MASTER so IFM_1000_T master mode support now is generally available with all PHY drivers. o Don't let e1000phy(4) set master/slave bits for IFM_1000_SX as it's not applicable there. Reviewed by: yongari (plus additional testing) Obtained from: NetBSD (partially), OpenBSD (partially) MFC after: 2 weeks
2010-11-14 13:26:10 +00:00
brgphy_setmedia(sc, ife->ifm_media);
Add PHY drivers for the XMAC II's internal PHY and the Broadcom BCM5400 1000baseTX PHY. These will be used by the SysKonnect gigabit ethernet driver shortly.
2000-04-22 01:58:18 +00:00
break;
default:
- In the spirit of previous simplifications factor out the checks for a different PHY instance being selected and isolation out into the wrappers around the service methods rather than duplicating them over and over again (besides, a PHY driver shouldn't need to care about which instance it actually is). - Centralize the check for the need to isolate a non-zero PHY instance not supporting isolation in mii_mediachg() and just ignore it rather than panicing, which should sufficient given that a) things are likely to just work anyway if one doesn't plug in more than one port at a time and b) refusing to attach in this case just leaves us in a unknown but most likely also not exactly correct configuration (besides several drivers setting MIIF_NOISOLATE didn't care about these anyway, probably due to setting this flag for no real reason). - Minor fixes like removing unnecessary setting of sc->mii_anegticks, using sc->mii_anegticks instead of hardcoded values etc.
2010-10-02 18:53:12 +00:00
return (EINVAL);
Add PHY drivers for the XMAC II's internal PHY and the Broadcom BCM5400 1000baseTX PHY. These will be used by the SysKonnect gigabit ethernet driver shortly.
2000-04-22 01:58:18 +00:00
}
break;
case MII_TICK:
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
/* Bail if autoneg isn't in process. */
- Fix autonegotiation timer. - Use MII_ANEGTICKS/MII_ANEGTICKS_GIGE defines instead of hardcoded value. - Fix some comments. - style(9) MFC after: 2 month
2006-12-19 08:41:48 +00:00
if (IFM_SUBTYPE(ife->ifm_media) != IFM_AUTO) {
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
sc->mii_ticks = 0;
Cleanup pass for mii drivers. . Make internal service routines static. . Use a consistent ordering of checks in MII_TICK. Do the work in the mii_phy_tick() subroutine if appropriate. . Call mii_phy_update() to trigger the callbacks.
2001-09-29 19:18:52 +00:00
break;
- Fix autonegotiation timer. - Use MII_ANEGTICKS/MII_ANEGTICKS_GIGE defines instead of hardcoded value. - Fix some comments. - style(9) MFC after: 2 month
2006-12-19 08:41:48 +00:00
}
Add PHY drivers for the XMAC II's internal PHY and the Broadcom BCM5400 1000baseTX PHY. These will be used by the SysKonnect gigabit ethernet driver shortly.
2000-04-22 01:58:18 +00:00
/*
* Check to see if we have link. If we do, we don't
- Fix autonegotiation timer. - Use MII_ANEGTICKS/MII_ANEGTICKS_GIGE defines instead of hardcoded value. - Fix some comments. - style(9) MFC after: 2 month
2006-12-19 08:41:48 +00:00
* need to restart the autonegotiation process.
Add PHY drivers for the XMAC II's internal PHY and the Broadcom BCM5400 1000baseTX PHY. These will be used by the SysKonnect gigabit ethernet driver shortly.
2000-04-22 01:58:18 +00:00
*/
o Flesh out the generic IEEE 802.3 annex 31B full duplex flow control support in mii(4): - Merge generic flow control advertisement (which can be enabled by passing by MIIF_DOPAUSE to mii_attach(9)) and parsing support from NetBSD into mii_physubr.c and ukphy_subr.c. Unlike as in NetBSD, IFM_FLOW isn't implemented as a global option via the "don't care mask" but instead as a media specific option this. This has the following advantages: o allows flow control advertisement with autonegotiation to be turned on and off via ifconfig(8) with the default typically being off (though MIIF_FORCEPAUSE has been added causing flow control to be always advertised, allowing to easily MFC this changes for drivers that previously used home-grown support for flow control that behaved that way without breaking POLA) o allows to deal with PHY drivers where flow control advertisement with manual selection doesn't work or at least isn't implemented, like it's the case with brgphy(4), e1000phy(4) and ip1000phy(4), by setting MIIF_NOMANPAUSE o the available combinations of media options are readily available from the `ifconfig -m` output - Add IFM_FLOW to IFM_SHARED_OPTION_DESCRIPTIONS and IFM_ETH_RXPAUSE and IFM_ETH_TXPAUSE to IFM_SUBTYPE_ETHERNET_OPTION_DESCRIPTIONS so these are understood by ifconfig(8). o Make the master/slave support in mii(4) actually usable: - Change IFM_ETH_MASTER from being implemented as a global option via the "don't care mask" to a media specific one as it actually is only applicable to IFM_1000_T to date. - Let mii_phy_setmedia() set GTCR_MAN_MS in IFM_1000_T slave mode to actually configure manually selected slave mode (like we also do in the PHY specific implementations). - Add IFM_ETH_MASTER to IFM_SUBTYPE_ETHERNET_OPTION_DESCRIPTIONS so it is understood by ifconfig(8). o Switch bge(4), bce(4), msk(4), nfe(4) and stge(4) along with brgphy(4), e1000phy(4) and ip1000phy(4) to use the generic flow control support instead of home-grown solutions via IFM_FLAGs. This includes changing these PHY drivers and smcphy(4) to no longer unconditionally advertise support for flow control but only if the selected media has IFM_FLOW set (or MIIF_FORCEPAUSE is set) and implemented for these media variants, i.e. typically only for copper. o Switch brgphy(4), ciphy(4), e1000phy(4) and ip1000phy(4) to report and set IFM_1000_T master mode via IFM_ETH_MASTER instead of via IFF_LINK0 and some IFM_FLAGn. o Switch brgphy(4) to add at least the the supported copper media based on the contents of the BMSR via mii_phy_add_media() instead of hardcoding them. The latter approach seems to have developed historically, besides causing unnecessary code duplication it was also undesirable because brgphy_mii_phy_auto() already based the capability advertisement on the contents of the BMSR though. o Let brgphy(4) set IFM_1000_T master mode on all supported PHY and not just BCM5701. Apparently this was a misinterpretation of a workaround in the Linux tg3 driver; BCM5701 seem to require RGPHY_1000CTL_MSE and BRGPHY_1000CTL_MSC to be set when configuring autonegotiation but this doesn't mean we can't set these as well on other PHYs for manual media selection. o Let ukphy_status() report IFM_1000_T master mode via IFM_ETH_MASTER so IFM_1000_T master mode support now is generally available with all PHY drivers. o Don't let e1000phy(4) set master/slave bits for IFM_1000_SX as it's not applicable there. Reviewed by: yongari (plus additional testing) Obtained from: NetBSD (partially), OpenBSD (partially) MFC after: 2 weeks
2010-11-14 13:26:10 +00:00
val = PHY_READ(sc, MII_BMSR) | PHY_READ(sc, MII_BMSR);
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
if (val & BMSR_LINK) {
- Fix autonegotiation timer. - Use MII_ANEGTICKS/MII_ANEGTICKS_GIGE defines instead of hardcoded value. - Fix some comments. - style(9) MFC after: 2 month
2006-12-19 08:41:48 +00:00
sc->mii_ticks = 0; /* Reset autoneg timer. */
Add PHY drivers for the XMAC II's internal PHY and the Broadcom BCM5400 1000baseTX PHY. These will be used by the SysKonnect gigabit ethernet driver shortly.
2000-04-22 01:58:18 +00:00
break;
- Fix autonegotiation timer. - Use MII_ANEGTICKS/MII_ANEGTICKS_GIGE defines instead of hardcoded value. - Fix some comments. - style(9) MFC after: 2 month
2006-12-19 08:41:48 +00:00
}
Add PHY drivers for the XMAC II's internal PHY and the Broadcom BCM5400 1000baseTX PHY. These will be used by the SysKonnect gigabit ethernet driver shortly.
2000-04-22 01:58:18 +00:00
- Fix autonegotiation timer. - Use MII_ANEGTICKS/MII_ANEGTICKS_GIGE defines instead of hardcoded value. - Fix some comments. - style(9) MFC after: 2 month
2006-12-19 08:41:48 +00:00
/* Announce link loss right after it happens. */
if (sc->mii_ticks++ == 0)
Correct the phy_service() routine case MII_TICK to correctly track the falling edge of a media state change. This is in preparation for media state change notification to the routing socket. No objections by: sam, wpaul, ru, bms Brucification by: bde
2004-05-03 13:01:34 +00:00
break;
Some style changes to a couple of PHY drivers: - Fix some whitespace nits. - Fix some spelling in comments. - Use MII_ANEGTICKS instead of 5. - Don't define variables in nested scope. - Remove superfluous returns at the end of void functions. - Remove unused static global rgephy_mii_model. - Remove dupe $Id$ in tdkphy(4). - Sort brgphys table. MFC after: 2 weeks
2006-12-02 19:36:25 +00:00
- Fix autonegotiation timer. - Use MII_ANEGTICKS/MII_ANEGTICKS_GIGE defines instead of hardcoded value. - Fix some comments. - style(9) MFC after: 2 month
2006-12-19 08:41:48 +00:00
/* Only retry autonegotiation every mii_anegticks seconds. */
if (sc->mii_ticks <= sc->mii_anegticks)
Restore link state handling which was broken in rev 1.69. Also report current link state while auto-negotiation is in progress. With this change link loss should be reported within a second and drivers that rely on link state should work. Reported by: Pete French < petefrench at ticketswitch dot com > Tested by: Pete French < petefrench at ticketswitch dot com > MFC after: 1 week
2008-08-12 00:57:39 +00:00
break;
- Fix autonegotiation timer. - Use MII_ANEGTICKS/MII_ANEGTICKS_GIGE defines instead of hardcoded value. - Fix some comments. - style(9) MFC after: 2 month
2006-12-19 08:41:48 +00:00
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
/* Retry autonegotiation */
Cleanup pass for mii drivers. . Make internal service routines static. . Use a consistent ordering of checks in MII_TICK. Do the work in the mii_phy_tick() subroutine if appropriate. . Call mii_phy_update() to trigger the callbacks.
2001-09-29 19:18:52 +00:00
sc->mii_ticks = 0;
o Flesh out the generic IEEE 802.3 annex 31B full duplex flow control support in mii(4): - Merge generic flow control advertisement (which can be enabled by passing by MIIF_DOPAUSE to mii_attach(9)) and parsing support from NetBSD into mii_physubr.c and ukphy_subr.c. Unlike as in NetBSD, IFM_FLOW isn't implemented as a global option via the "don't care mask" but instead as a media specific option this. This has the following advantages: o allows flow control advertisement with autonegotiation to be turned on and off via ifconfig(8) with the default typically being off (though MIIF_FORCEPAUSE has been added causing flow control to be always advertised, allowing to easily MFC this changes for drivers that previously used home-grown support for flow control that behaved that way without breaking POLA) o allows to deal with PHY drivers where flow control advertisement with manual selection doesn't work or at least isn't implemented, like it's the case with brgphy(4), e1000phy(4) and ip1000phy(4), by setting MIIF_NOMANPAUSE o the available combinations of media options are readily available from the `ifconfig -m` output - Add IFM_FLOW to IFM_SHARED_OPTION_DESCRIPTIONS and IFM_ETH_RXPAUSE and IFM_ETH_TXPAUSE to IFM_SUBTYPE_ETHERNET_OPTION_DESCRIPTIONS so these are understood by ifconfig(8). o Make the master/slave support in mii(4) actually usable: - Change IFM_ETH_MASTER from being implemented as a global option via the "don't care mask" to a media specific one as it actually is only applicable to IFM_1000_T to date. - Let mii_phy_setmedia() set GTCR_MAN_MS in IFM_1000_T slave mode to actually configure manually selected slave mode (like we also do in the PHY specific implementations). - Add IFM_ETH_MASTER to IFM_SUBTYPE_ETHERNET_OPTION_DESCRIPTIONS so it is understood by ifconfig(8). o Switch bge(4), bce(4), msk(4), nfe(4) and stge(4) along with brgphy(4), e1000phy(4) and ip1000phy(4) to use the generic flow control support instead of home-grown solutions via IFM_FLAGs. This includes changing these PHY drivers and smcphy(4) to no longer unconditionally advertise support for flow control but only if the selected media has IFM_FLOW set (or MIIF_FORCEPAUSE is set) and implemented for these media variants, i.e. typically only for copper. o Switch brgphy(4), ciphy(4), e1000phy(4) and ip1000phy(4) to report and set IFM_1000_T master mode via IFM_ETH_MASTER instead of via IFF_LINK0 and some IFM_FLAGn. o Switch brgphy(4) to add at least the the supported copper media based on the contents of the BMSR via mii_phy_add_media() instead of hardcoding them. The latter approach seems to have developed historically, besides causing unnecessary code duplication it was also undesirable because brgphy_mii_phy_auto() already based the capability advertisement on the contents of the BMSR though. o Let brgphy(4) set IFM_1000_T master mode on all supported PHY and not just BCM5701. Apparently this was a misinterpretation of a workaround in the Linux tg3 driver; BCM5701 seem to require RGPHY_1000CTL_MSE and BRGPHY_1000CTL_MSC to be set when configuring autonegotiation but this doesn't mean we can't set these as well on other PHYs for manual media selection. o Let ukphy_status() report IFM_1000_T master mode via IFM_ETH_MASTER so IFM_1000_T master mode support now is generally available with all PHY drivers. o Don't let e1000phy(4) set master/slave bits for IFM_1000_SX as it's not applicable there. Reviewed by: yongari (plus additional testing) Obtained from: NetBSD (partially), OpenBSD (partially) MFC after: 2 weeks
2010-11-14 13:26:10 +00:00
brgphy_mii_phy_auto(sc, ife->ifm_media);
1) fix tiny bug in bge_start_locked() 2) rework link state detection code & use it in POLLING mode 3) fix 2 bugs in link state detection code: a) driver unable to detect link loss on bcm5721 b) on bcm570x chips (tested on bcm5700 bcm5701 bcm5702) driver fails to detect link loss with probability 1/6 (solved in brgphy.c) Devices working in TBI mode should not be affected by this change. Approved by: glebius (mentor) MFC after: 1 month
2005-12-08 13:31:52 +00:00
break;
Add PHY drivers for the XMAC II's internal PHY and the Broadcom BCM5400 1000baseTX PHY. These will be used by the SysKonnect gigabit ethernet driver shortly.
2000-04-22 01:58:18 +00:00
}
/* Update the media status. */
- Remove attempts to implement setting of BMCR_LOOP/MIIF_NOLOOP (reporting IFM_LOOP based on BMCR_LOOP is left in place though as it might provide useful for debugging). For most mii(4) drivers it was unclear whether the PHYs driven by them actually support loopback or not. Moreover, typically loopback mode also needs to be activated on the MAC, which none of the Ethernet drivers using mii(4) implements. Given that loopback media has no real use (and obviously hardly had a chance to actually work) besides for driver development (which just loopback mode should be sufficient for though, i.e one doesn't necessary need support for loopback media) support for it is just dropped as both NetBSD and OpenBSD already did quite some time ago. - Let mii_phy_add_media() also announce the support of IFM_NONE. - 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 lxtphy(4) which disables MII interrupts (as is done for a few other PHYs we have drivers for). This includes changing NIC drivers which previously just called the generic mii_phy_reset() to now actually call the PHY-specific reset routine, which might be crucial in some cases. While at it, the redundant checks in these NIC drivers for mii->mii_instance not being zero before calling the reset routines were removed because as soon as one PHY driver attaches mii->mii_instance is incremented and we hardly can end up in their media change callbacks etc if no PHY driver has attached as mii_attach() would have failed in that case and not attach a miibus(4) instance. Consequently, NIC drivers now no longer should call mii_phy_reset() directly, so it was removed from EXPORT_SYMS. - Add a mii_phy_dev_attach() as a companion helper to mii_phy_dev_probe(). The purpose of that function is to perform the common steps to attach a PHY driver instance and to hook it up to the miibus(4) instance and to optionally also handle the probing, addition and initialization of the supported media. So all a PHY driver without any special requirements has to do in its bus attach method is to call mii_phy_dev_attach() along with PHY-specific MIIF_* flags, a pointer to its PHY functions and the add_media set to one. All PHY drivers were updated to take advantage of mii_phy_dev_attach() as appropriate. Along with these changes the capability mask was added to the mii_softc structure so PHY drivers taking advantage of mii_phy_dev_attach() but still handling media on their own do not need to fiddle with the MII attach arguments anyway. - Keep track of the PHY offset in the mii_softc structure. This is done for compatibility with NetBSD/OpenBSD. - Keep track of the PHY's OUI, model and revision in the mii_softc structure. Several PHY drivers require this information also after attaching and previously had to wrap their own softc around mii_softc. NetBSD/OpenBSD also keep track of the model and revision on their mii_softc structure. All PHY drivers were updated to take advantage as appropriate. - Convert the mebers of the MII data structure to unsigned where appropriate. This is partly inspired by NetBSD/OpenBSD. - According to IEEE 802.3-2002 the bits actually have to be reversed when mapping an OUI to the MII ID registers. All PHY drivers and miidevs where changed as necessary. Actually this now again allows to largely share miidevs with NetBSD, which fixed this problem already 9 years ago. Consequently miidevs was synced as far as possible. - Add MIIF_NOMANPAUSE and mii_phy_flowstatus() calls to drivers that weren't explicitly converted to support flow control before. It's unclear whether flow control actually works with these but typically it should and their net behavior should be more correct with these changes in place than without if the MAC driver sets MIIF_DOPAUSE. Obtained from: NetBSD (partially) Reviewed by: yongari (earlier version), silence on arch@ and net@
2011-05-03 19:51:29 +00:00
PHY_STATUS(sc);
Add PHY drivers for the XMAC II's internal PHY and the Broadcom BCM5400 1000baseTX PHY. These will be used by the SysKonnect gigabit ethernet driver shortly.
2000-04-22 01:58:18 +00:00
Something I missed in my tree. If the media changes, poke the DSP on the 5401/5411 Broadcom PHY's. Obtained from: NetBSD
2003-05-04 02:03:20 +00:00
/*
* Callback if something changed. Note that we need to poke
* the DSP on the Broadcom PHYs if the media changes.
*/
Some style changes to a couple of PHY drivers: - Fix some whitespace nits. - Fix some spelling in comments. - Use MII_ANEGTICKS instead of 5. - Don't define variables in nested scope. - Remove superfluous returns at the end of void functions. - Remove unused static global rgephy_mii_model. - Remove dupe $Id$ in tdkphy(4). - Sort brgphys table. MFC after: 2 weeks
2006-12-02 19:36:25 +00:00
if (sc->mii_media_active != mii->mii_media_active ||
Something I missed in my tree. If the media changes, poke the DSP on the 5401/5411 Broadcom PHY's. Obtained from: NetBSD
2003-05-04 02:03:20 +00:00
sc->mii_media_status != mii->mii_media_status ||
cmd == MII_MEDIACHG) {
- Remove attempts to implement setting of BMCR_LOOP/MIIF_NOLOOP (reporting IFM_LOOP based on BMCR_LOOP is left in place though as it might provide useful for debugging). For most mii(4) drivers it was unclear whether the PHYs driven by them actually support loopback or not. Moreover, typically loopback mode also needs to be activated on the MAC, which none of the Ethernet drivers using mii(4) implements. Given that loopback media has no real use (and obviously hardly had a chance to actually work) besides for driver development (which just loopback mode should be sufficient for though, i.e one doesn't necessary need support for loopback media) support for it is just dropped as both NetBSD and OpenBSD already did quite some time ago. - Let mii_phy_add_media() also announce the support of IFM_NONE. - 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 lxtphy(4) which disables MII interrupts (as is done for a few other PHYs we have drivers for). This includes changing NIC drivers which previously just called the generic mii_phy_reset() to now actually call the PHY-specific reset routine, which might be crucial in some cases. While at it, the redundant checks in these NIC drivers for mii->mii_instance not being zero before calling the reset routines were removed because as soon as one PHY driver attaches mii->mii_instance is incremented and we hardly can end up in their media change callbacks etc if no PHY driver has attached as mii_attach() would have failed in that case and not attach a miibus(4) instance. Consequently, NIC drivers now no longer should call mii_phy_reset() directly, so it was removed from EXPORT_SYMS. - Add a mii_phy_dev_attach() as a companion helper to mii_phy_dev_probe(). The purpose of that function is to perform the common steps to attach a PHY driver instance and to hook it up to the miibus(4) instance and to optionally also handle the probing, addition and initialization of the supported media. So all a PHY driver without any special requirements has to do in its bus attach method is to call mii_phy_dev_attach() along with PHY-specific MIIF_* flags, a pointer to its PHY functions and the add_media set to one. All PHY drivers were updated to take advantage of mii_phy_dev_attach() as appropriate. Along with these changes the capability mask was added to the mii_softc structure so PHY drivers taking advantage of mii_phy_dev_attach() but still handling media on their own do not need to fiddle with the MII attach arguments anyway. - Keep track of the PHY offset in the mii_softc structure. This is done for compatibility with NetBSD/OpenBSD. - Keep track of the PHY's OUI, model and revision in the mii_softc structure. Several PHY drivers require this information also after attaching and previously had to wrap their own softc around mii_softc. NetBSD/OpenBSD also keep track of the model and revision on their mii_softc structure. All PHY drivers were updated to take advantage as appropriate. - Convert the mebers of the MII data structure to unsigned where appropriate. This is partly inspired by NetBSD/OpenBSD. - According to IEEE 802.3-2002 the bits actually have to be reversed when mapping an OUI to the MII ID registers. All PHY drivers and miidevs where changed as necessary. Actually this now again allows to largely share miidevs with NetBSD, which fixed this problem already 9 years ago. Consequently miidevs was synced as far as possible. - Add MIIF_NOMANPAUSE and mii_phy_flowstatus() calls to drivers that weren't explicitly converted to support flow control before. It's unclear whether flow control actually works with these but typically it should and their net behavior should be more correct with these changes in place than without if the MAC driver sets MIIF_DOPAUSE. Obtained from: NetBSD (partially) Reviewed by: yongari (earlier version), silence on arch@ and net@
2011-05-03 19:51:29 +00:00
switch (sc->mii_mpd_oui) {
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
case MII_OUI_BROADCOM:
- Remove attempts to implement setting of BMCR_LOOP/MIIF_NOLOOP (reporting IFM_LOOP based on BMCR_LOOP is left in place though as it might provide useful for debugging). For most mii(4) drivers it was unclear whether the PHYs driven by them actually support loopback or not. Moreover, typically loopback mode also needs to be activated on the MAC, which none of the Ethernet drivers using mii(4) implements. Given that loopback media has no real use (and obviously hardly had a chance to actually work) besides for driver development (which just loopback mode should be sufficient for though, i.e one doesn't necessary need support for loopback media) support for it is just dropped as both NetBSD and OpenBSD already did quite some time ago. - Let mii_phy_add_media() also announce the support of IFM_NONE. - 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 lxtphy(4) which disables MII interrupts (as is done for a few other PHYs we have drivers for). This includes changing NIC drivers which previously just called the generic mii_phy_reset() to now actually call the PHY-specific reset routine, which might be crucial in some cases. While at it, the redundant checks in these NIC drivers for mii->mii_instance not being zero before calling the reset routines were removed because as soon as one PHY driver attaches mii->mii_instance is incremented and we hardly can end up in their media change callbacks etc if no PHY driver has attached as mii_attach() would have failed in that case and not attach a miibus(4) instance. Consequently, NIC drivers now no longer should call mii_phy_reset() directly, so it was removed from EXPORT_SYMS. - Add a mii_phy_dev_attach() as a companion helper to mii_phy_dev_probe(). The purpose of that function is to perform the common steps to attach a PHY driver instance and to hook it up to the miibus(4) instance and to optionally also handle the probing, addition and initialization of the supported media. So all a PHY driver without any special requirements has to do in its bus attach method is to call mii_phy_dev_attach() along with PHY-specific MIIF_* flags, a pointer to its PHY functions and the add_media set to one. All PHY drivers were updated to take advantage of mii_phy_dev_attach() as appropriate. Along with these changes the capability mask was added to the mii_softc structure so PHY drivers taking advantage of mii_phy_dev_attach() but still handling media on their own do not need to fiddle with the MII attach arguments anyway. - Keep track of the PHY offset in the mii_softc structure. This is done for compatibility with NetBSD/OpenBSD. - Keep track of the PHY's OUI, model and revision in the mii_softc structure. Several PHY drivers require this information also after attaching and previously had to wrap their own softc around mii_softc. NetBSD/OpenBSD also keep track of the model and revision on their mii_softc structure. All PHY drivers were updated to take advantage as appropriate. - Convert the mebers of the MII data structure to unsigned where appropriate. This is partly inspired by NetBSD/OpenBSD. - According to IEEE 802.3-2002 the bits actually have to be reversed when mapping an OUI to the MII ID registers. All PHY drivers and miidevs where changed as necessary. Actually this now again allows to largely share miidevs with NetBSD, which fixed this problem already 9 years ago. Consequently miidevs was synced as far as possible. - Add MIIF_NOMANPAUSE and mii_phy_flowstatus() calls to drivers that weren't explicitly converted to support flow control before. It's unclear whether flow control actually works with these but typically it should and their net behavior should be more correct with these changes in place than without if the MAC driver sets MIIF_DOPAUSE. Obtained from: NetBSD (partially) Reviewed by: yongari (earlier version), silence on arch@ and net@
2011-05-03 19:51:29 +00:00
switch (sc->mii_mpd_model) {
case MII_MODEL_BROADCOM_BCM5400:
- Fix BCM5754 support found in Dell PowerEdge SC440. - Move some PHY bug detections from brgphy.c to if_bge.c. - Do not penalize working PHYs. - Re-arrange bge_flags roughly by their categories. - Fix minor style(9) nits. PR: kern/107257 Obtained from: OpenBSD Tested by: Mike Hibler <mike at flux dot utah dot edu>
2007-01-15 21:43:43 +00:00
bcm5401_load_dspcode(sc);
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
break;
- Remove attempts to implement setting of BMCR_LOOP/MIIF_NOLOOP (reporting IFM_LOOP based on BMCR_LOOP is left in place though as it might provide useful for debugging). For most mii(4) drivers it was unclear whether the PHYs driven by them actually support loopback or not. Moreover, typically loopback mode also needs to be activated on the MAC, which none of the Ethernet drivers using mii(4) implements. Given that loopback media has no real use (and obviously hardly had a chance to actually work) besides for driver development (which just loopback mode should be sufficient for though, i.e one doesn't necessary need support for loopback media) support for it is just dropped as both NetBSD and OpenBSD already did quite some time ago. - Let mii_phy_add_media() also announce the support of IFM_NONE. - 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 lxtphy(4) which disables MII interrupts (as is done for a few other PHYs we have drivers for). This includes changing NIC drivers which previously just called the generic mii_phy_reset() to now actually call the PHY-specific reset routine, which might be crucial in some cases. While at it, the redundant checks in these NIC drivers for mii->mii_instance not being zero before calling the reset routines were removed because as soon as one PHY driver attaches mii->mii_instance is incremented and we hardly can end up in their media change callbacks etc if no PHY driver has attached as mii_attach() would have failed in that case and not attach a miibus(4) instance. Consequently, NIC drivers now no longer should call mii_phy_reset() directly, so it was removed from EXPORT_SYMS. - Add a mii_phy_dev_attach() as a companion helper to mii_phy_dev_probe(). The purpose of that function is to perform the common steps to attach a PHY driver instance and to hook it up to the miibus(4) instance and to optionally also handle the probing, addition and initialization of the supported media. So all a PHY driver without any special requirements has to do in its bus attach method is to call mii_phy_dev_attach() along with PHY-specific MIIF_* flags, a pointer to its PHY functions and the add_media set to one. All PHY drivers were updated to take advantage of mii_phy_dev_attach() as appropriate. Along with these changes the capability mask was added to the mii_softc structure so PHY drivers taking advantage of mii_phy_dev_attach() but still handling media on their own do not need to fiddle with the MII attach arguments anyway. - Keep track of the PHY offset in the mii_softc structure. This is done for compatibility with NetBSD/OpenBSD. - Keep track of the PHY's OUI, model and revision in the mii_softc structure. Several PHY drivers require this information also after attaching and previously had to wrap their own softc around mii_softc. NetBSD/OpenBSD also keep track of the model and revision on their mii_softc structure. All PHY drivers were updated to take advantage as appropriate. - Convert the mebers of the MII data structure to unsigned where appropriate. This is partly inspired by NetBSD/OpenBSD. - According to IEEE 802.3-2002 the bits actually have to be reversed when mapping an OUI to the MII ID registers. All PHY drivers and miidevs where changed as necessary. Actually this now again allows to largely share miidevs with NetBSD, which fixed this problem already 9 years ago. Consequently miidevs was synced as far as possible. - Add MIIF_NOMANPAUSE and mii_phy_flowstatus() calls to drivers that weren't explicitly converted to support flow control before. It's unclear whether flow control actually works with these but typically it should and their net behavior should be more correct with these changes in place than without if the MAC driver sets MIIF_DOPAUSE. Obtained from: NetBSD (partially) Reviewed by: yongari (earlier version), silence on arch@ and net@
2011-05-03 19:51:29 +00:00
case MII_MODEL_BROADCOM_BCM5401:
if (sc->mii_mpd_rev == 1 || sc->mii_mpd_rev == 3)
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
bcm5401_load_dspcode(sc);
break;
- Remove attempts to implement setting of BMCR_LOOP/MIIF_NOLOOP (reporting IFM_LOOP based on BMCR_LOOP is left in place though as it might provide useful for debugging). For most mii(4) drivers it was unclear whether the PHYs driven by them actually support loopback or not. Moreover, typically loopback mode also needs to be activated on the MAC, which none of the Ethernet drivers using mii(4) implements. Given that loopback media has no real use (and obviously hardly had a chance to actually work) besides for driver development (which just loopback mode should be sufficient for though, i.e one doesn't necessary need support for loopback media) support for it is just dropped as both NetBSD and OpenBSD already did quite some time ago. - Let mii_phy_add_media() also announce the support of IFM_NONE. - 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 lxtphy(4) which disables MII interrupts (as is done for a few other PHYs we have drivers for). This includes changing NIC drivers which previously just called the generic mii_phy_reset() to now actually call the PHY-specific reset routine, which might be crucial in some cases. While at it, the redundant checks in these NIC drivers for mii->mii_instance not being zero before calling the reset routines were removed because as soon as one PHY driver attaches mii->mii_instance is incremented and we hardly can end up in their media change callbacks etc if no PHY driver has attached as mii_attach() would have failed in that case and not attach a miibus(4) instance. Consequently, NIC drivers now no longer should call mii_phy_reset() directly, so it was removed from EXPORT_SYMS. - Add a mii_phy_dev_attach() as a companion helper to mii_phy_dev_probe(). The purpose of that function is to perform the common steps to attach a PHY driver instance and to hook it up to the miibus(4) instance and to optionally also handle the probing, addition and initialization of the supported media. So all a PHY driver without any special requirements has to do in its bus attach method is to call mii_phy_dev_attach() along with PHY-specific MIIF_* flags, a pointer to its PHY functions and the add_media set to one. All PHY drivers were updated to take advantage of mii_phy_dev_attach() as appropriate. Along with these changes the capability mask was added to the mii_softc structure so PHY drivers taking advantage of mii_phy_dev_attach() but still handling media on their own do not need to fiddle with the MII attach arguments anyway. - Keep track of the PHY offset in the mii_softc structure. This is done for compatibility with NetBSD/OpenBSD. - Keep track of the PHY's OUI, model and revision in the mii_softc structure. Several PHY drivers require this information also after attaching and previously had to wrap their own softc around mii_softc. NetBSD/OpenBSD also keep track of the model and revision on their mii_softc structure. All PHY drivers were updated to take advantage as appropriate. - Convert the mebers of the MII data structure to unsigned where appropriate. This is partly inspired by NetBSD/OpenBSD. - According to IEEE 802.3-2002 the bits actually have to be reversed when mapping an OUI to the MII ID registers. All PHY drivers and miidevs where changed as necessary. Actually this now again allows to largely share miidevs with NetBSD, which fixed this problem already 9 years ago. Consequently miidevs was synced as far as possible. - Add MIIF_NOMANPAUSE and mii_phy_flowstatus() calls to drivers that weren't explicitly converted to support flow control before. It's unclear whether flow control actually works with these but typically it should and their net behavior should be more correct with these changes in place than without if the MAC driver sets MIIF_DOPAUSE. Obtained from: NetBSD (partially) Reviewed by: yongari (earlier version), silence on arch@ and net@
2011-05-03 19:51:29 +00:00
case MII_MODEL_BROADCOM_BCM5411:
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
bcm5411_load_dspcode(sc);
break;
- Remove attempts to implement setting of BMCR_LOOP/MIIF_NOLOOP (reporting IFM_LOOP based on BMCR_LOOP is left in place though as it might provide useful for debugging). For most mii(4) drivers it was unclear whether the PHYs driven by them actually support loopback or not. Moreover, typically loopback mode also needs to be activated on the MAC, which none of the Ethernet drivers using mii(4) implements. Given that loopback media has no real use (and obviously hardly had a chance to actually work) besides for driver development (which just loopback mode should be sufficient for though, i.e one doesn't necessary need support for loopback media) support for it is just dropped as both NetBSD and OpenBSD already did quite some time ago. - Let mii_phy_add_media() also announce the support of IFM_NONE. - 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 lxtphy(4) which disables MII interrupts (as is done for a few other PHYs we have drivers for). This includes changing NIC drivers which previously just called the generic mii_phy_reset() to now actually call the PHY-specific reset routine, which might be crucial in some cases. While at it, the redundant checks in these NIC drivers for mii->mii_instance not being zero before calling the reset routines were removed because as soon as one PHY driver attaches mii->mii_instance is incremented and we hardly can end up in their media change callbacks etc if no PHY driver has attached as mii_attach() would have failed in that case and not attach a miibus(4) instance. Consequently, NIC drivers now no longer should call mii_phy_reset() directly, so it was removed from EXPORT_SYMS. - Add a mii_phy_dev_attach() as a companion helper to mii_phy_dev_probe(). The purpose of that function is to perform the common steps to attach a PHY driver instance and to hook it up to the miibus(4) instance and to optionally also handle the probing, addition and initialization of the supported media. So all a PHY driver without any special requirements has to do in its bus attach method is to call mii_phy_dev_attach() along with PHY-specific MIIF_* flags, a pointer to its PHY functions and the add_media set to one. All PHY drivers were updated to take advantage of mii_phy_dev_attach() as appropriate. Along with these changes the capability mask was added to the mii_softc structure so PHY drivers taking advantage of mii_phy_dev_attach() but still handling media on their own do not need to fiddle with the MII attach arguments anyway. - Keep track of the PHY offset in the mii_softc structure. This is done for compatibility with NetBSD/OpenBSD. - Keep track of the PHY's OUI, model and revision in the mii_softc structure. Several PHY drivers require this information also after attaching and previously had to wrap their own softc around mii_softc. NetBSD/OpenBSD also keep track of the model and revision on their mii_softc structure. All PHY drivers were updated to take advantage as appropriate. - Convert the mebers of the MII data structure to unsigned where appropriate. This is partly inspired by NetBSD/OpenBSD. - According to IEEE 802.3-2002 the bits actually have to be reversed when mapping an OUI to the MII ID registers. All PHY drivers and miidevs where changed as necessary. Actually this now again allows to largely share miidevs with NetBSD, which fixed this problem already 9 years ago. Consequently miidevs was synced as far as possible. - Add MIIF_NOMANPAUSE and mii_phy_flowstatus() calls to drivers that weren't explicitly converted to support flow control before. It's unclear whether flow control actually works with these but typically it should and their net behavior should be more correct with these changes in place than without if the MAC driver sets MIIF_DOPAUSE. Obtained from: NetBSD (partially) Reviewed by: yongari (earlier version), silence on arch@ and net@
2011-05-03 19:51:29 +00:00
case MII_MODEL_BROADCOM_BCM54K2:
Add support for BCM54K2 found in combination with Apple K2 GMAC. Submitted by: Andreas Tobler Obtained from: OpenBSD MFC after: 1 week
2010-02-20 22:01:24 +00:00
bcm54k2_load_dspcode(sc);
break;
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
}
- Fix BCM5754 support found in Dell PowerEdge SC440. - Move some PHY bug detections from brgphy.c to if_bge.c. - Do not penalize working PHYs. - Re-arrange bge_flags roughly by their categories. - Fix minor style(9) nits. PR: kern/107257 Obtained from: OpenBSD Tested by: Mike Hibler <mike at flux dot utah dot edu>
2007-01-15 21:43:43 +00:00
break;
Something I missed in my tree. If the media changes, poke the DSP on the 5401/5411 Broadcom PHY's. Obtained from: NetBSD
2003-05-04 02:03:20 +00:00
}
}
Correct the phy_service() routine case MII_TICK to correctly track the falling edge of a media state change. This is in preparation for media state change notification to the routing socket. No objections by: sam, wpaul, ru, bms Brucification by: bde
2004-05-03 13:01:34 +00:00
mii_phy_update(sc, cmd);
- In the spirit of previous simplifications factor out the checks for a different PHY instance being selected and isolation out into the wrappers around the service methods rather than duplicating them over and over again (besides, a PHY driver shouldn't need to care about which instance it actually is). - Centralize the check for the need to isolate a non-zero PHY instance not supporting isolation in mii_mediachg() and just ignore it rather than panicing, which should sufficient given that a) things are likely to just work anyway if one doesn't plug in more than one port at a time and b) refusing to attach in this case just leaves us in a unknown but most likely also not exactly correct configuration (besides several drivers setting MIIF_NOISOLATE didn't care about these anyway, probably due to setting this flag for no real reason). - Minor fixes like removing unnecessary setting of sc->mii_anegticks, using sc->mii_anegticks instead of hardcoded values etc.
2010-10-02 18:53:12 +00:00
return (0);
Add PHY drivers for the XMAC II's internal PHY and the Broadcom BCM5400 1000baseTX PHY. These will be used by the SysKonnect gigabit ethernet driver shortly.
2000-04-22 01:58:18 +00:00
}
- Added support for BCM5709 and BCM5716. MFC after: 2 weeks
2008-06-13 01:20:29 +00:00
/****************************************************************************/
/* Sets the PHY link speed. */
/* */
/* Returns: */
/* None */
/****************************************************************************/
- Do not depend on auto negotiation for link speed/duplex status. - Read link status from BMSR instead of auxilary status register. - Clean up style(9) nits.
2006-12-20 00:08:47 +00:00
static void
o Flesh out the generic IEEE 802.3 annex 31B full duplex flow control support in mii(4): - Merge generic flow control advertisement (which can be enabled by passing by MIIF_DOPAUSE to mii_attach(9)) and parsing support from NetBSD into mii_physubr.c and ukphy_subr.c. Unlike as in NetBSD, IFM_FLOW isn't implemented as a global option via the "don't care mask" but instead as a media specific option this. This has the following advantages: o allows flow control advertisement with autonegotiation to be turned on and off via ifconfig(8) with the default typically being off (though MIIF_FORCEPAUSE has been added causing flow control to be always advertised, allowing to easily MFC this changes for drivers that previously used home-grown support for flow control that behaved that way without breaking POLA) o allows to deal with PHY drivers where flow control advertisement with manual selection doesn't work or at least isn't implemented, like it's the case with brgphy(4), e1000phy(4) and ip1000phy(4), by setting MIIF_NOMANPAUSE o the available combinations of media options are readily available from the `ifconfig -m` output - Add IFM_FLOW to IFM_SHARED_OPTION_DESCRIPTIONS and IFM_ETH_RXPAUSE and IFM_ETH_TXPAUSE to IFM_SUBTYPE_ETHERNET_OPTION_DESCRIPTIONS so these are understood by ifconfig(8). o Make the master/slave support in mii(4) actually usable: - Change IFM_ETH_MASTER from being implemented as a global option via the "don't care mask" to a media specific one as it actually is only applicable to IFM_1000_T to date. - Let mii_phy_setmedia() set GTCR_MAN_MS in IFM_1000_T slave mode to actually configure manually selected slave mode (like we also do in the PHY specific implementations). - Add IFM_ETH_MASTER to IFM_SUBTYPE_ETHERNET_OPTION_DESCRIPTIONS so it is understood by ifconfig(8). o Switch bge(4), bce(4), msk(4), nfe(4) and stge(4) along with brgphy(4), e1000phy(4) and ip1000phy(4) to use the generic flow control support instead of home-grown solutions via IFM_FLAGs. This includes changing these PHY drivers and smcphy(4) to no longer unconditionally advertise support for flow control but only if the selected media has IFM_FLOW set (or MIIF_FORCEPAUSE is set) and implemented for these media variants, i.e. typically only for copper. o Switch brgphy(4), ciphy(4), e1000phy(4) and ip1000phy(4) to report and set IFM_1000_T master mode via IFM_ETH_MASTER instead of via IFF_LINK0 and some IFM_FLAGn. o Switch brgphy(4) to add at least the the supported copper media based on the contents of the BMSR via mii_phy_add_media() instead of hardcoding them. The latter approach seems to have developed historically, besides causing unnecessary code duplication it was also undesirable because brgphy_mii_phy_auto() already based the capability advertisement on the contents of the BMSR though. o Let brgphy(4) set IFM_1000_T master mode on all supported PHY and not just BCM5701. Apparently this was a misinterpretation of a workaround in the Linux tg3 driver; BCM5701 seem to require RGPHY_1000CTL_MSE and BRGPHY_1000CTL_MSC to be set when configuring autonegotiation but this doesn't mean we can't set these as well on other PHYs for manual media selection. o Let ukphy_status() report IFM_1000_T master mode via IFM_ETH_MASTER so IFM_1000_T master mode support now is generally available with all PHY drivers. o Don't let e1000phy(4) set master/slave bits for IFM_1000_SX as it's not applicable there. Reviewed by: yongari (plus additional testing) Obtained from: NetBSD (partially), OpenBSD (partially) MFC after: 2 weeks
2010-11-14 13:26:10 +00:00
brgphy_setmedia(struct mii_softc *sc, int media)
- Do not depend on auto negotiation for link speed/duplex status. - Read link status from BMSR instead of auxilary status register. - Clean up style(9) nits.
2006-12-20 00:08:47 +00:00
{
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
int bmcr = 0, gig;
- Do not depend on auto negotiation for link speed/duplex status. - Read link status from BMSR instead of auxilary status register. - Clean up style(9) nits.
2006-12-20 00:08:47 +00:00
switch (IFM_SUBTYPE(media)) {
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
case IFM_2500_SX:
break;
case IFM_1000_SX:
- Do not depend on auto negotiation for link speed/duplex status. - Read link status from BMSR instead of auxilary status register. - Clean up style(9) nits.
2006-12-20 00:08:47 +00:00
case IFM_1000_T:
bmcr = BRGPHY_S1000;
break;
case IFM_100_TX:
bmcr = BRGPHY_S100;
break;
case IFM_10_T:
default:
bmcr = BRGPHY_S10;
break;
}
- Added support for BCM5709 and BCM5716. MFC after: 2 weeks
2008-06-13 01:20:29 +00:00
- Even after masking the media with IFM_GMASK the result may have bits besides the duplex ones set so just comparing it with IFM_FDX may lead to false negatives. - Simplify ciphy_service() to only set the manual configuration bits once after we have figured them all out. This also means we no longer unnecessarily update the hardware along the road. MFC after: 1 week
2011-01-14 19:29:53 +00:00
if ((media & IFM_FDX) != 0) {
- Do not depend on auto negotiation for link speed/duplex status. - Read link status from BMSR instead of auxilary status register. - Clean up style(9) nits.
2006-12-20 00:08:47 +00:00
bmcr |= BRGPHY_BMCR_FDX;
gig = BRGPHY_1000CTL_AFD;
} else {
gig = BRGPHY_1000CTL_AHD;
}
o Flesh out the generic IEEE 802.3 annex 31B full duplex flow control support in mii(4): - Merge generic flow control advertisement (which can be enabled by passing by MIIF_DOPAUSE to mii_attach(9)) and parsing support from NetBSD into mii_physubr.c and ukphy_subr.c. Unlike as in NetBSD, IFM_FLOW isn't implemented as a global option via the "don't care mask" but instead as a media specific option this. This has the following advantages: o allows flow control advertisement with autonegotiation to be turned on and off via ifconfig(8) with the default typically being off (though MIIF_FORCEPAUSE has been added causing flow control to be always advertised, allowing to easily MFC this changes for drivers that previously used home-grown support for flow control that behaved that way without breaking POLA) o allows to deal with PHY drivers where flow control advertisement with manual selection doesn't work or at least isn't implemented, like it's the case with brgphy(4), e1000phy(4) and ip1000phy(4), by setting MIIF_NOMANPAUSE o the available combinations of media options are readily available from the `ifconfig -m` output - Add IFM_FLOW to IFM_SHARED_OPTION_DESCRIPTIONS and IFM_ETH_RXPAUSE and IFM_ETH_TXPAUSE to IFM_SUBTYPE_ETHERNET_OPTION_DESCRIPTIONS so these are understood by ifconfig(8). o Make the master/slave support in mii(4) actually usable: - Change IFM_ETH_MASTER from being implemented as a global option via the "don't care mask" to a media specific one as it actually is only applicable to IFM_1000_T to date. - Let mii_phy_setmedia() set GTCR_MAN_MS in IFM_1000_T slave mode to actually configure manually selected slave mode (like we also do in the PHY specific implementations). - Add IFM_ETH_MASTER to IFM_SUBTYPE_ETHERNET_OPTION_DESCRIPTIONS so it is understood by ifconfig(8). o Switch bge(4), bce(4), msk(4), nfe(4) and stge(4) along with brgphy(4), e1000phy(4) and ip1000phy(4) to use the generic flow control support instead of home-grown solutions via IFM_FLAGs. This includes changing these PHY drivers and smcphy(4) to no longer unconditionally advertise support for flow control but only if the selected media has IFM_FLOW set (or MIIF_FORCEPAUSE is set) and implemented for these media variants, i.e. typically only for copper. o Switch brgphy(4), ciphy(4), e1000phy(4) and ip1000phy(4) to report and set IFM_1000_T master mode via IFM_ETH_MASTER instead of via IFF_LINK0 and some IFM_FLAGn. o Switch brgphy(4) to add at least the the supported copper media based on the contents of the BMSR via mii_phy_add_media() instead of hardcoding them. The latter approach seems to have developed historically, besides causing unnecessary code duplication it was also undesirable because brgphy_mii_phy_auto() already based the capability advertisement on the contents of the BMSR though. o Let brgphy(4) set IFM_1000_T master mode on all supported PHY and not just BCM5701. Apparently this was a misinterpretation of a workaround in the Linux tg3 driver; BCM5701 seem to require RGPHY_1000CTL_MSE and BRGPHY_1000CTL_MSC to be set when configuring autonegotiation but this doesn't mean we can't set these as well on other PHYs for manual media selection. o Let ukphy_status() report IFM_1000_T master mode via IFM_ETH_MASTER so IFM_1000_T master mode support now is generally available with all PHY drivers. o Don't let e1000phy(4) set master/slave bits for IFM_1000_SX as it's not applicable there. Reviewed by: yongari (plus additional testing) Obtained from: NetBSD (partially), OpenBSD (partially) MFC after: 2 weeks
2010-11-14 13:26:10 +00:00
/* Force loopback to disconnect PHY from Ethernet medium. */
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
brgphy_enable_loopback(sc);
- Added support for BCM5709 and BCM5716. MFC after: 2 weeks
2008-06-13 01:20:29 +00:00
- Do not depend on auto negotiation for link speed/duplex status. - Read link status from BMSR instead of auxilary status register. - Clean up style(9) nits.
2006-12-20 00:08:47 +00:00
PHY_WRITE(sc, BRGPHY_MII_1000CTL, 0);
PHY_WRITE(sc, BRGPHY_MII_ANAR, BRGPHY_SEL_TYPE);
o Flesh out the generic IEEE 802.3 annex 31B full duplex flow control support in mii(4): - Merge generic flow control advertisement (which can be enabled by passing by MIIF_DOPAUSE to mii_attach(9)) and parsing support from NetBSD into mii_physubr.c and ukphy_subr.c. Unlike as in NetBSD, IFM_FLOW isn't implemented as a global option via the "don't care mask" but instead as a media specific option this. This has the following advantages: o allows flow control advertisement with autonegotiation to be turned on and off via ifconfig(8) with the default typically being off (though MIIF_FORCEPAUSE has been added causing flow control to be always advertised, allowing to easily MFC this changes for drivers that previously used home-grown support for flow control that behaved that way without breaking POLA) o allows to deal with PHY drivers where flow control advertisement with manual selection doesn't work or at least isn't implemented, like it's the case with brgphy(4), e1000phy(4) and ip1000phy(4), by setting MIIF_NOMANPAUSE o the available combinations of media options are readily available from the `ifconfig -m` output - Add IFM_FLOW to IFM_SHARED_OPTION_DESCRIPTIONS and IFM_ETH_RXPAUSE and IFM_ETH_TXPAUSE to IFM_SUBTYPE_ETHERNET_OPTION_DESCRIPTIONS so these are understood by ifconfig(8). o Make the master/slave support in mii(4) actually usable: - Change IFM_ETH_MASTER from being implemented as a global option via the "don't care mask" to a media specific one as it actually is only applicable to IFM_1000_T to date. - Let mii_phy_setmedia() set GTCR_MAN_MS in IFM_1000_T slave mode to actually configure manually selected slave mode (like we also do in the PHY specific implementations). - Add IFM_ETH_MASTER to IFM_SUBTYPE_ETHERNET_OPTION_DESCRIPTIONS so it is understood by ifconfig(8). o Switch bge(4), bce(4), msk(4), nfe(4) and stge(4) along with brgphy(4), e1000phy(4) and ip1000phy(4) to use the generic flow control support instead of home-grown solutions via IFM_FLAGs. This includes changing these PHY drivers and smcphy(4) to no longer unconditionally advertise support for flow control but only if the selected media has IFM_FLOW set (or MIIF_FORCEPAUSE is set) and implemented for these media variants, i.e. typically only for copper. o Switch brgphy(4), ciphy(4), e1000phy(4) and ip1000phy(4) to report and set IFM_1000_T master mode via IFM_ETH_MASTER instead of via IFF_LINK0 and some IFM_FLAGn. o Switch brgphy(4) to add at least the the supported copper media based on the contents of the BMSR via mii_phy_add_media() instead of hardcoding them. The latter approach seems to have developed historically, besides causing unnecessary code duplication it was also undesirable because brgphy_mii_phy_auto() already based the capability advertisement on the contents of the BMSR though. o Let brgphy(4) set IFM_1000_T master mode on all supported PHY and not just BCM5701. Apparently this was a misinterpretation of a workaround in the Linux tg3 driver; BCM5701 seem to require RGPHY_1000CTL_MSE and BRGPHY_1000CTL_MSC to be set when configuring autonegotiation but this doesn't mean we can't set these as well on other PHYs for manual media selection. o Let ukphy_status() report IFM_1000_T master mode via IFM_ETH_MASTER so IFM_1000_T master mode support now is generally available with all PHY drivers. o Don't let e1000phy(4) set master/slave bits for IFM_1000_SX as it's not applicable there. Reviewed by: yongari (plus additional testing) Obtained from: NetBSD (partially), OpenBSD (partially) MFC after: 2 weeks
2010-11-14 13:26:10 +00:00
if (IFM_SUBTYPE(media) != IFM_1000_T &&
IFM_SUBTYPE(media) != IFM_1000_SX) {
PHY_WRITE(sc, BRGPHY_MII_BMCR, bmcr);
return;
}
- Do not depend on auto negotiation for link speed/duplex status. - Read link status from BMSR instead of auxilary status register. - Clean up style(9) nits.
2006-12-20 00:08:47 +00:00
o Flesh out the generic IEEE 802.3 annex 31B full duplex flow control support in mii(4): - Merge generic flow control advertisement (which can be enabled by passing by MIIF_DOPAUSE to mii_attach(9)) and parsing support from NetBSD into mii_physubr.c and ukphy_subr.c. Unlike as in NetBSD, IFM_FLOW isn't implemented as a global option via the "don't care mask" but instead as a media specific option this. This has the following advantages: o allows flow control advertisement with autonegotiation to be turned on and off via ifconfig(8) with the default typically being off (though MIIF_FORCEPAUSE has been added causing flow control to be always advertised, allowing to easily MFC this changes for drivers that previously used home-grown support for flow control that behaved that way without breaking POLA) o allows to deal with PHY drivers where flow control advertisement with manual selection doesn't work or at least isn't implemented, like it's the case with brgphy(4), e1000phy(4) and ip1000phy(4), by setting MIIF_NOMANPAUSE o the available combinations of media options are readily available from the `ifconfig -m` output - Add IFM_FLOW to IFM_SHARED_OPTION_DESCRIPTIONS and IFM_ETH_RXPAUSE and IFM_ETH_TXPAUSE to IFM_SUBTYPE_ETHERNET_OPTION_DESCRIPTIONS so these are understood by ifconfig(8). o Make the master/slave support in mii(4) actually usable: - Change IFM_ETH_MASTER from being implemented as a global option via the "don't care mask" to a media specific one as it actually is only applicable to IFM_1000_T to date. - Let mii_phy_setmedia() set GTCR_MAN_MS in IFM_1000_T slave mode to actually configure manually selected slave mode (like we also do in the PHY specific implementations). - Add IFM_ETH_MASTER to IFM_SUBTYPE_ETHERNET_OPTION_DESCRIPTIONS so it is understood by ifconfig(8). o Switch bge(4), bce(4), msk(4), nfe(4) and stge(4) along with brgphy(4), e1000phy(4) and ip1000phy(4) to use the generic flow control support instead of home-grown solutions via IFM_FLAGs. This includes changing these PHY drivers and smcphy(4) to no longer unconditionally advertise support for flow control but only if the selected media has IFM_FLOW set (or MIIF_FORCEPAUSE is set) and implemented for these media variants, i.e. typically only for copper. o Switch brgphy(4), ciphy(4), e1000phy(4) and ip1000phy(4) to report and set IFM_1000_T master mode via IFM_ETH_MASTER instead of via IFF_LINK0 and some IFM_FLAGn. o Switch brgphy(4) to add at least the the supported copper media based on the contents of the BMSR via mii_phy_add_media() instead of hardcoding them. The latter approach seems to have developed historically, besides causing unnecessary code duplication it was also undesirable because brgphy_mii_phy_auto() already based the capability advertisement on the contents of the BMSR though. o Let brgphy(4) set IFM_1000_T master mode on all supported PHY and not just BCM5701. Apparently this was a misinterpretation of a workaround in the Linux tg3 driver; BCM5701 seem to require RGPHY_1000CTL_MSE and BRGPHY_1000CTL_MSC to be set when configuring autonegotiation but this doesn't mean we can't set these as well on other PHYs for manual media selection. o Let ukphy_status() report IFM_1000_T master mode via IFM_ETH_MASTER so IFM_1000_T master mode support now is generally available with all PHY drivers. o Don't let e1000phy(4) set master/slave bits for IFM_1000_SX as it's not applicable there. Reviewed by: yongari (plus additional testing) Obtained from: NetBSD (partially), OpenBSD (partially) MFC after: 2 weeks
2010-11-14 13:26:10 +00:00
if (IFM_SUBTYPE(media) == IFM_1000_T) {
gig |= BRGPHY_1000CTL_MSE;
if ((media & IFM_ETH_MASTER) != 0)
gig |= BRGPHY_1000CTL_MSC;
}
- Do not depend on auto negotiation for link speed/duplex status. - Read link status from BMSR instead of auxilary status register. - Clean up style(9) nits.
2006-12-20 00:08:47 +00:00
PHY_WRITE(sc, BRGPHY_MII_1000CTL, gig);
PHY_WRITE(sc, BRGPHY_MII_BMCR,
bmcr | BRGPHY_BMCR_AUTOEN | BRGPHY_BMCR_STARTNEG);
}
- Added support for BCM5709 and BCM5716. MFC after: 2 weeks
2008-06-13 01:20:29 +00:00
/****************************************************************************/
/* Set the media status based on the PHY settings. */
/* */
/* Returns: */
/* None */
/****************************************************************************/
Cleanup pass for mii drivers. . Make internal service routines static. . Use a consistent ordering of checks in MII_TICK. Do the work in the mii_phy_tick() subroutine if appropriate. . Call mii_phy_update() to trigger the callbacks.
2001-09-29 19:18:52 +00:00
static void
Use ansi function definitions in preference to K&R to reduce diffs with NetBSD (and cause it looks cooler).
2005-09-30 19:39:27 +00:00
brgphy_status(struct mii_softc *sc)
Add PHY drivers for the XMAC II's internal PHY and the Broadcom BCM5400 1000baseTX PHY. These will be used by the SysKonnect gigabit ethernet driver shortly.
2000-04-22 01:58:18 +00:00
{
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
struct brgphy_softc *bsc = (struct brgphy_softc *)sc;
Add PHY drivers for the XMAC II's internal PHY and the Broadcom BCM5400 1000baseTX PHY. These will be used by the SysKonnect gigabit ethernet driver shortly.
2000-04-22 01:58:18 +00:00
struct mii_data *mii = sc->mii_pdata;
o Flesh out the generic IEEE 802.3 annex 31B full duplex flow control support in mii(4): - Merge generic flow control advertisement (which can be enabled by passing by MIIF_DOPAUSE to mii_attach(9)) and parsing support from NetBSD into mii_physubr.c and ukphy_subr.c. Unlike as in NetBSD, IFM_FLOW isn't implemented as a global option via the "don't care mask" but instead as a media specific option this. This has the following advantages: o allows flow control advertisement with autonegotiation to be turned on and off via ifconfig(8) with the default typically being off (though MIIF_FORCEPAUSE has been added causing flow control to be always advertised, allowing to easily MFC this changes for drivers that previously used home-grown support for flow control that behaved that way without breaking POLA) o allows to deal with PHY drivers where flow control advertisement with manual selection doesn't work or at least isn't implemented, like it's the case with brgphy(4), e1000phy(4) and ip1000phy(4), by setting MIIF_NOMANPAUSE o the available combinations of media options are readily available from the `ifconfig -m` output - Add IFM_FLOW to IFM_SHARED_OPTION_DESCRIPTIONS and IFM_ETH_RXPAUSE and IFM_ETH_TXPAUSE to IFM_SUBTYPE_ETHERNET_OPTION_DESCRIPTIONS so these are understood by ifconfig(8). o Make the master/slave support in mii(4) actually usable: - Change IFM_ETH_MASTER from being implemented as a global option via the "don't care mask" to a media specific one as it actually is only applicable to IFM_1000_T to date. - Let mii_phy_setmedia() set GTCR_MAN_MS in IFM_1000_T slave mode to actually configure manually selected slave mode (like we also do in the PHY specific implementations). - Add IFM_ETH_MASTER to IFM_SUBTYPE_ETHERNET_OPTION_DESCRIPTIONS so it is understood by ifconfig(8). o Switch bge(4), bce(4), msk(4), nfe(4) and stge(4) along with brgphy(4), e1000phy(4) and ip1000phy(4) to use the generic flow control support instead of home-grown solutions via IFM_FLAGs. This includes changing these PHY drivers and smcphy(4) to no longer unconditionally advertise support for flow control but only if the selected media has IFM_FLOW set (or MIIF_FORCEPAUSE is set) and implemented for these media variants, i.e. typically only for copper. o Switch brgphy(4), ciphy(4), e1000phy(4) and ip1000phy(4) to report and set IFM_1000_T master mode via IFM_ETH_MASTER instead of via IFF_LINK0 and some IFM_FLAGn. o Switch brgphy(4) to add at least the the supported copper media based on the contents of the BMSR via mii_phy_add_media() instead of hardcoding them. The latter approach seems to have developed historically, besides causing unnecessary code duplication it was also undesirable because brgphy_mii_phy_auto() already based the capability advertisement on the contents of the BMSR though. o Let brgphy(4) set IFM_1000_T master mode on all supported PHY and not just BCM5701. Apparently this was a misinterpretation of a workaround in the Linux tg3 driver; BCM5701 seem to require RGPHY_1000CTL_MSE and BRGPHY_1000CTL_MSC to be set when configuring autonegotiation but this doesn't mean we can't set these as well on other PHYs for manual media selection. o Let ukphy_status() report IFM_1000_T master mode via IFM_ETH_MASTER so IFM_1000_T master mode support now is generally available with all PHY drivers. o Don't let e1000phy(4) set master/slave bits for IFM_1000_SX as it's not applicable there. Reviewed by: yongari (plus additional testing) Obtained from: NetBSD (partially), OpenBSD (partially) MFC after: 2 weeks
2010-11-14 13:26:10 +00:00
int aux, bmcr, bmsr, val, xstat;
u_int flowstat;
Add PHY drivers for the XMAC II's internal PHY and the Broadcom BCM5400 1000baseTX PHY. These will be used by the SysKonnect gigabit ethernet driver shortly.
2000-04-22 01:58:18 +00:00
mii->mii_media_status = IFM_AVALID;
mii->mii_media_active = IFM_ETHER;
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
bmsr = PHY_READ(sc, BRGPHY_MII_BMSR) | PHY_READ(sc, BRGPHY_MII_BMSR);
Remove 'cr' at the end of line.
2008-08-12 00:55:03 +00:00
bmcr = PHY_READ(sc, BRGPHY_MII_BMCR);
Add PHY drivers for the XMAC II's internal PHY and the Broadcom BCM5400 1000baseTX PHY. These will be used by the SysKonnect gigabit ethernet driver shortly.
2000-04-22 01:58:18 +00:00
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
if (bmcr & BRGPHY_BMCR_LOOP) {
Add PHY drivers for the XMAC II's internal PHY and the Broadcom BCM5400 1000baseTX PHY. These will be used by the SysKonnect gigabit ethernet driver shortly.
2000-04-22 01:58:18 +00:00
mii->mii_media_active |= IFM_LOOP;
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
}
Add PHY drivers for the XMAC II's internal PHY and the Broadcom BCM5400 1000baseTX PHY. These will be used by the SysKonnect gigabit ethernet driver shortly.
2000-04-22 01:58:18 +00:00
Revert couple of changes from 1.51 and 1.52. Reading link status with BMSR is okay for most of the chipsets but BCM5701 PHY does not seem to like it. Set media to IFM_NONE if link is not up instead of the previous value. Reported by: Goran Lowkrantz (goran dot lowkrantz at ismobile dot com)
2007-03-19 23:17:39 +00:00
if ((bmcr & BRGPHY_BMCR_AUTOEN) &&
Provide workaround for the ages old bug affecting certain BCM5708S chip revision often found in the blades and resulting in interfaces not sensing carrier signal. Looking at all problem reports it appears that it only affects some very specific silicon revision (ASIC (0x57081021); Rev (B2)) and version of the PHY that supports 1000baseSX-FDX media only. Therefore, narrow the scope of workaround to combination of that revision and media type. Given that the first report on this issue is dated back to 2007, there is not much hope that this issue will ever be properly resolved. Among affected systems are IBM HS21, Intel SBXD132 and HP BL460c. PR: 118238, 122551, 140970 MFC after: 1 month
2010-03-10 05:19:14 +00:00
(bmsr & BRGPHY_BMSR_ACOMP) == 0 &&
(bsc->serdes_flags & BRGPHY_NOANWAIT) == 0) {
Revert couple of changes from 1.51 and 1.52. Reading link status with BMSR is okay for most of the chipsets but BCM5701 PHY does not seem to like it. Set media to IFM_NONE if link is not up instead of the previous value. Reported by: Goran Lowkrantz (goran dot lowkrantz at ismobile dot com)
2007-03-19 23:17:39 +00:00
/* Erg, still trying, I guess... */
mii->mii_media_active |= IFM_NONE;
o Flesh out the generic IEEE 802.3 annex 31B full duplex flow control support in mii(4): - Merge generic flow control advertisement (which can be enabled by passing by MIIF_DOPAUSE to mii_attach(9)) and parsing support from NetBSD into mii_physubr.c and ukphy_subr.c. Unlike as in NetBSD, IFM_FLOW isn't implemented as a global option via the "don't care mask" but instead as a media specific option this. This has the following advantages: o allows flow control advertisement with autonegotiation to be turned on and off via ifconfig(8) with the default typically being off (though MIIF_FORCEPAUSE has been added causing flow control to be always advertised, allowing to easily MFC this changes for drivers that previously used home-grown support for flow control that behaved that way without breaking POLA) o allows to deal with PHY drivers where flow control advertisement with manual selection doesn't work or at least isn't implemented, like it's the case with brgphy(4), e1000phy(4) and ip1000phy(4), by setting MIIF_NOMANPAUSE o the available combinations of media options are readily available from the `ifconfig -m` output - Add IFM_FLOW to IFM_SHARED_OPTION_DESCRIPTIONS and IFM_ETH_RXPAUSE and IFM_ETH_TXPAUSE to IFM_SUBTYPE_ETHERNET_OPTION_DESCRIPTIONS so these are understood by ifconfig(8). o Make the master/slave support in mii(4) actually usable: - Change IFM_ETH_MASTER from being implemented as a global option via the "don't care mask" to a media specific one as it actually is only applicable to IFM_1000_T to date. - Let mii_phy_setmedia() set GTCR_MAN_MS in IFM_1000_T slave mode to actually configure manually selected slave mode (like we also do in the PHY specific implementations). - Add IFM_ETH_MASTER to IFM_SUBTYPE_ETHERNET_OPTION_DESCRIPTIONS so it is understood by ifconfig(8). o Switch bge(4), bce(4), msk(4), nfe(4) and stge(4) along with brgphy(4), e1000phy(4) and ip1000phy(4) to use the generic flow control support instead of home-grown solutions via IFM_FLAGs. This includes changing these PHY drivers and smcphy(4) to no longer unconditionally advertise support for flow control but only if the selected media has IFM_FLOW set (or MIIF_FORCEPAUSE is set) and implemented for these media variants, i.e. typically only for copper. o Switch brgphy(4), ciphy(4), e1000phy(4) and ip1000phy(4) to report and set IFM_1000_T master mode via IFM_ETH_MASTER instead of via IFF_LINK0 and some IFM_FLAGn. o Switch brgphy(4) to add at least the the supported copper media based on the contents of the BMSR via mii_phy_add_media() instead of hardcoding them. The latter approach seems to have developed historically, besides causing unnecessary code duplication it was also undesirable because brgphy_mii_phy_auto() already based the capability advertisement on the contents of the BMSR though. o Let brgphy(4) set IFM_1000_T master mode on all supported PHY and not just BCM5701. Apparently this was a misinterpretation of a workaround in the Linux tg3 driver; BCM5701 seem to require RGPHY_1000CTL_MSE and BRGPHY_1000CTL_MSC to be set when configuring autonegotiation but this doesn't mean we can't set these as well on other PHYs for manual media selection. o Let ukphy_status() report IFM_1000_T master mode via IFM_ETH_MASTER so IFM_1000_T master mode support now is generally available with all PHY drivers. o Don't let e1000phy(4) set master/slave bits for IFM_1000_SX as it's not applicable there. Reviewed by: yongari (plus additional testing) Obtained from: NetBSD (partially), OpenBSD (partially) MFC after: 2 weeks
2010-11-14 13:26:10 +00:00
return;
- Do not depend on auto negotiation for link speed/duplex status. - Read link status from BMSR instead of auxilary status register. - Clean up style(9) nits.
2006-12-20 00:08:47 +00:00
}
Add PHY drivers for the XMAC II's internal PHY and the Broadcom BCM5400 1000baseTX PHY. These will be used by the SysKonnect gigabit ethernet driver shortly.
2000-04-22 01:58:18 +00:00
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
if ((sc->mii_flags & MIIF_HAVEFIBER) == 0) {
o Flesh out the generic IEEE 802.3 annex 31B full duplex flow control support in mii(4): - Merge generic flow control advertisement (which can be enabled by passing by MIIF_DOPAUSE to mii_attach(9)) and parsing support from NetBSD into mii_physubr.c and ukphy_subr.c. Unlike as in NetBSD, IFM_FLOW isn't implemented as a global option via the "don't care mask" but instead as a media specific option this. This has the following advantages: o allows flow control advertisement with autonegotiation to be turned on and off via ifconfig(8) with the default typically being off (though MIIF_FORCEPAUSE has been added causing flow control to be always advertised, allowing to easily MFC this changes for drivers that previously used home-grown support for flow control that behaved that way without breaking POLA) o allows to deal with PHY drivers where flow control advertisement with manual selection doesn't work or at least isn't implemented, like it's the case with brgphy(4), e1000phy(4) and ip1000phy(4), by setting MIIF_NOMANPAUSE o the available combinations of media options are readily available from the `ifconfig -m` output - Add IFM_FLOW to IFM_SHARED_OPTION_DESCRIPTIONS and IFM_ETH_RXPAUSE and IFM_ETH_TXPAUSE to IFM_SUBTYPE_ETHERNET_OPTION_DESCRIPTIONS so these are understood by ifconfig(8). o Make the master/slave support in mii(4) actually usable: - Change IFM_ETH_MASTER from being implemented as a global option via the "don't care mask" to a media specific one as it actually is only applicable to IFM_1000_T to date. - Let mii_phy_setmedia() set GTCR_MAN_MS in IFM_1000_T slave mode to actually configure manually selected slave mode (like we also do in the PHY specific implementations). - Add IFM_ETH_MASTER to IFM_SUBTYPE_ETHERNET_OPTION_DESCRIPTIONS so it is understood by ifconfig(8). o Switch bge(4), bce(4), msk(4), nfe(4) and stge(4) along with brgphy(4), e1000phy(4) and ip1000phy(4) to use the generic flow control support instead of home-grown solutions via IFM_FLAGs. This includes changing these PHY drivers and smcphy(4) to no longer unconditionally advertise support for flow control but only if the selected media has IFM_FLOW set (or MIIF_FORCEPAUSE is set) and implemented for these media variants, i.e. typically only for copper. o Switch brgphy(4), ciphy(4), e1000phy(4) and ip1000phy(4) to report and set IFM_1000_T master mode via IFM_ETH_MASTER instead of via IFF_LINK0 and some IFM_FLAGn. o Switch brgphy(4) to add at least the the supported copper media based on the contents of the BMSR via mii_phy_add_media() instead of hardcoding them. The latter approach seems to have developed historically, besides causing unnecessary code duplication it was also undesirable because brgphy_mii_phy_auto() already based the capability advertisement on the contents of the BMSR though. o Let brgphy(4) set IFM_1000_T master mode on all supported PHY and not just BCM5701. Apparently this was a misinterpretation of a workaround in the Linux tg3 driver; BCM5701 seem to require RGPHY_1000CTL_MSE and BRGPHY_1000CTL_MSC to be set when configuring autonegotiation but this doesn't mean we can't set these as well on other PHYs for manual media selection. o Let ukphy_status() report IFM_1000_T master mode via IFM_ETH_MASTER so IFM_1000_T master mode support now is generally available with all PHY drivers. o Don't let e1000phy(4) set master/slave bits for IFM_1000_SX as it's not applicable there. Reviewed by: yongari (plus additional testing) Obtained from: NetBSD (partially), OpenBSD (partially) MFC after: 2 weeks
2010-11-14 13:26:10 +00:00
/*
* NB: reading the ANAR, ANLPAR or 1000STS after the AUXSTS
* wedges at least the PHY of BCM5704 (but not others).
*/
flowstat = mii_phy_flowstatus(sc);
xstat = PHY_READ(sc, BRGPHY_MII_1000STS);
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
aux = PHY_READ(sc, BRGPHY_MII_AUXSTS);
/* If copper link is up, get the negotiated speed/duplex. */
if (aux & BRGPHY_AUXSTS_LINK) {
mii->mii_media_status |= IFM_ACTIVE;
switch (aux & BRGPHY_AUXSTS_AN_RES) {
case BRGPHY_RES_1000FD:
mii->mii_media_active |= IFM_1000_T | IFM_FDX; break;
case BRGPHY_RES_1000HD:
mii->mii_media_active |= IFM_1000_T | IFM_HDX; break;
case BRGPHY_RES_100FD:
mii->mii_media_active |= IFM_100_TX | IFM_FDX; break;
case BRGPHY_RES_100T4:
mii->mii_media_active |= IFM_100_T4; break;
case BRGPHY_RES_100HD:
mii->mii_media_active |= IFM_100_TX | IFM_HDX; break;
case BRGPHY_RES_10FD:
mii->mii_media_active |= IFM_10_T | IFM_FDX; break;
case BRGPHY_RES_10HD:
mii->mii_media_active |= IFM_10_T | IFM_HDX; break;
default:
mii->mii_media_active |= IFM_NONE; break;
}
o Flesh out the generic IEEE 802.3 annex 31B full duplex flow control support in mii(4): - Merge generic flow control advertisement (which can be enabled by passing by MIIF_DOPAUSE to mii_attach(9)) and parsing support from NetBSD into mii_physubr.c and ukphy_subr.c. Unlike as in NetBSD, IFM_FLOW isn't implemented as a global option via the "don't care mask" but instead as a media specific option this. This has the following advantages: o allows flow control advertisement with autonegotiation to be turned on and off via ifconfig(8) with the default typically being off (though MIIF_FORCEPAUSE has been added causing flow control to be always advertised, allowing to easily MFC this changes for drivers that previously used home-grown support for flow control that behaved that way without breaking POLA) o allows to deal with PHY drivers where flow control advertisement with manual selection doesn't work or at least isn't implemented, like it's the case with brgphy(4), e1000phy(4) and ip1000phy(4), by setting MIIF_NOMANPAUSE o the available combinations of media options are readily available from the `ifconfig -m` output - Add IFM_FLOW to IFM_SHARED_OPTION_DESCRIPTIONS and IFM_ETH_RXPAUSE and IFM_ETH_TXPAUSE to IFM_SUBTYPE_ETHERNET_OPTION_DESCRIPTIONS so these are understood by ifconfig(8). o Make the master/slave support in mii(4) actually usable: - Change IFM_ETH_MASTER from being implemented as a global option via the "don't care mask" to a media specific one as it actually is only applicable to IFM_1000_T to date. - Let mii_phy_setmedia() set GTCR_MAN_MS in IFM_1000_T slave mode to actually configure manually selected slave mode (like we also do in the PHY specific implementations). - Add IFM_ETH_MASTER to IFM_SUBTYPE_ETHERNET_OPTION_DESCRIPTIONS so it is understood by ifconfig(8). o Switch bge(4), bce(4), msk(4), nfe(4) and stge(4) along with brgphy(4), e1000phy(4) and ip1000phy(4) to use the generic flow control support instead of home-grown solutions via IFM_FLAGs. This includes changing these PHY drivers and smcphy(4) to no longer unconditionally advertise support for flow control but only if the selected media has IFM_FLOW set (or MIIF_FORCEPAUSE is set) and implemented for these media variants, i.e. typically only for copper. o Switch brgphy(4), ciphy(4), e1000phy(4) and ip1000phy(4) to report and set IFM_1000_T master mode via IFM_ETH_MASTER instead of via IFF_LINK0 and some IFM_FLAGn. o Switch brgphy(4) to add at least the the supported copper media based on the contents of the BMSR via mii_phy_add_media() instead of hardcoding them. The latter approach seems to have developed historically, besides causing unnecessary code duplication it was also undesirable because brgphy_mii_phy_auto() already based the capability advertisement on the contents of the BMSR though. o Let brgphy(4) set IFM_1000_T master mode on all supported PHY and not just BCM5701. Apparently this was a misinterpretation of a workaround in the Linux tg3 driver; BCM5701 seem to require RGPHY_1000CTL_MSE and BRGPHY_1000CTL_MSC to be set when configuring autonegotiation but this doesn't mean we can't set these as well on other PHYs for manual media selection. o Let ukphy_status() report IFM_1000_T master mode via IFM_ETH_MASTER so IFM_1000_T master mode support now is generally available with all PHY drivers. o Don't let e1000phy(4) set master/slave bits for IFM_1000_SX as it's not applicable there. Reviewed by: yongari (plus additional testing) Obtained from: NetBSD (partially), OpenBSD (partially) MFC after: 2 weeks
2010-11-14 13:26:10 +00:00
if ((mii->mii_media_active & IFM_FDX) != 0)
mii->mii_media_active |= flowstat;
if (IFM_SUBTYPE(mii->mii_media_active) == IFM_1000_T &&
(xstat & BRGPHY_1000STS_MSR) != 0)
mii->mii_media_active |= IFM_ETH_MASTER;
Add support for the BCM5401 and BCM5411 10/100/1000Mbps copper gigE PHYs. This basically updates the brgphy driver to support 10/100 modes in addition to 1000Mbps modes.
2001-09-04 22:00:33 +00:00
}
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
} else {
o Flesh out the generic IEEE 802.3 annex 31B full duplex flow control support in mii(4): - Merge generic flow control advertisement (which can be enabled by passing by MIIF_DOPAUSE to mii_attach(9)) and parsing support from NetBSD into mii_physubr.c and ukphy_subr.c. Unlike as in NetBSD, IFM_FLOW isn't implemented as a global option via the "don't care mask" but instead as a media specific option this. This has the following advantages: o allows flow control advertisement with autonegotiation to be turned on and off via ifconfig(8) with the default typically being off (though MIIF_FORCEPAUSE has been added causing flow control to be always advertised, allowing to easily MFC this changes for drivers that previously used home-grown support for flow control that behaved that way without breaking POLA) o allows to deal with PHY drivers where flow control advertisement with manual selection doesn't work or at least isn't implemented, like it's the case with brgphy(4), e1000phy(4) and ip1000phy(4), by setting MIIF_NOMANPAUSE o the available combinations of media options are readily available from the `ifconfig -m` output - Add IFM_FLOW to IFM_SHARED_OPTION_DESCRIPTIONS and IFM_ETH_RXPAUSE and IFM_ETH_TXPAUSE to IFM_SUBTYPE_ETHERNET_OPTION_DESCRIPTIONS so these are understood by ifconfig(8). o Make the master/slave support in mii(4) actually usable: - Change IFM_ETH_MASTER from being implemented as a global option via the "don't care mask" to a media specific one as it actually is only applicable to IFM_1000_T to date. - Let mii_phy_setmedia() set GTCR_MAN_MS in IFM_1000_T slave mode to actually configure manually selected slave mode (like we also do in the PHY specific implementations). - Add IFM_ETH_MASTER to IFM_SUBTYPE_ETHERNET_OPTION_DESCRIPTIONS so it is understood by ifconfig(8). o Switch bge(4), bce(4), msk(4), nfe(4) and stge(4) along with brgphy(4), e1000phy(4) and ip1000phy(4) to use the generic flow control support instead of home-grown solutions via IFM_FLAGs. This includes changing these PHY drivers and smcphy(4) to no longer unconditionally advertise support for flow control but only if the selected media has IFM_FLOW set (or MIIF_FORCEPAUSE is set) and implemented for these media variants, i.e. typically only for copper. o Switch brgphy(4), ciphy(4), e1000phy(4) and ip1000phy(4) to report and set IFM_1000_T master mode via IFM_ETH_MASTER instead of via IFF_LINK0 and some IFM_FLAGn. o Switch brgphy(4) to add at least the the supported copper media based on the contents of the BMSR via mii_phy_add_media() instead of hardcoding them. The latter approach seems to have developed historically, besides causing unnecessary code duplication it was also undesirable because brgphy_mii_phy_auto() already based the capability advertisement on the contents of the BMSR though. o Let brgphy(4) set IFM_1000_T master mode on all supported PHY and not just BCM5701. Apparently this was a misinterpretation of a workaround in the Linux tg3 driver; BCM5701 seem to require RGPHY_1000CTL_MSE and BRGPHY_1000CTL_MSC to be set when configuring autonegotiation but this doesn't mean we can't set these as well on other PHYs for manual media selection. o Let ukphy_status() report IFM_1000_T master mode via IFM_ETH_MASTER so IFM_1000_T master mode support now is generally available with all PHY drivers. o Don't let e1000phy(4) set master/slave bits for IFM_1000_SX as it's not applicable there. Reviewed by: yongari (plus additional testing) Obtained from: NetBSD (partially), OpenBSD (partially) MFC after: 2 weeks
2010-11-14 13:26:10 +00:00
/* Todo: Add support for flow control. */
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
/* If serdes link is up, get the negotiated speed/duplex. */
if (bmsr & BRGPHY_BMSR_LINK) {
mii->mii_media_status |= IFM_ACTIVE;
}
/* Check the link speed/duplex based on the PHY type. */
if (bsc->serdes_flags & BRGPHY_5706S) {
mii->mii_media_active |= IFM_1000_SX;
/* If autoneg enabled, read negotiated duplex settings */
if (bmcr & BRGPHY_BMCR_AUTOEN) {
val = PHY_READ(sc, BRGPHY_SERDES_ANAR) & PHY_READ(sc, BRGPHY_SERDES_ANLPAR);
if (val & BRGPHY_SERDES_ANAR_FDX)
mii->mii_media_active |= IFM_FDX;
else
mii->mii_media_active |= IFM_HDX;
}
} else if (bsc->serdes_flags & BRGPHY_5708S) {
PHY_WRITE(sc, BRGPHY_5708S_BLOCK_ADDR, BRGPHY_5708S_DIG_PG0);
xstat = PHY_READ(sc, BRGPHY_5708S_PG0_1000X_STAT1);
Consistently use tab characters instead of tab + space characters. No functional changes.
2010-09-07 23:08:38 +00:00
/* Check for MRBE auto-negotiated speed results. */
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
switch (xstat & BRGPHY_5708S_PG0_1000X_STAT1_SPEED_MASK) {
Remove whitespace at the end of line.
2008-08-12 00:52:10 +00:00
case BRGPHY_5708S_PG0_1000X_STAT1_SPEED_10:
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
mii->mii_media_active |= IFM_10_FL; break;
Remove whitespace at the end of line.
2008-08-12 00:52:10 +00:00
case BRGPHY_5708S_PG0_1000X_STAT1_SPEED_100:
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
mii->mii_media_active |= IFM_100_FX; break;
Remove whitespace at the end of line.
2008-08-12 00:52:10 +00:00
case BRGPHY_5708S_PG0_1000X_STAT1_SPEED_1G:
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
mii->mii_media_active |= IFM_1000_SX; break;
Remove whitespace at the end of line.
2008-08-12 00:52:10 +00:00
case BRGPHY_5708S_PG0_1000X_STAT1_SPEED_25G:
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
mii->mii_media_active |= IFM_2500_SX; break;
}
Consistently use tab characters instead of tab + space characters. No functional changes.
2010-09-07 23:08:38 +00:00
/* Check for MRBE auto-negotiated duplex results. */
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
if (xstat & BRGPHY_5708S_PG0_1000X_STAT1_FDX)
mii->mii_media_active |= IFM_FDX;
else
mii->mii_media_active |= IFM_HDX;
Consistently use tab characters instead of tab + space characters. No functional changes.
2010-09-07 23:08:38 +00:00
} else if (bsc->serdes_flags & BRGPHY_5709S) {
/* Select GP Status Block of the AN MMD, get autoneg results. */
PHY_WRITE(sc, BRGPHY_BLOCK_ADDR, BRGPHY_BLOCK_ADDR_GP_STATUS);
- Added support for 5709S/5716S PHYs. Submitted by: pyunyh MFC after: 2 weeks
2010-03-18 20:57:57 +00:00
xstat = PHY_READ(sc, BRGPHY_GP_STATUS_TOP_ANEG_STATUS);
Consistently use tab characters instead of tab + space characters. No functional changes.
2010-09-07 23:08:38 +00:00
/* Restore IEEE0 block (assumed in all brgphy(4) code). */
PHY_WRITE(sc, BRGPHY_BLOCK_ADDR, BRGPHY_BLOCK_ADDR_COMBO_IEEE0);
/* Check for MRBE auto-negotiated speed results. */
switch (xstat & BRGPHY_GP_STATUS_TOP_ANEG_SPEED_MASK) {
case BRGPHY_GP_STATUS_TOP_ANEG_SPEED_10:
mii->mii_media_active |= IFM_10_FL; break;
case BRGPHY_GP_STATUS_TOP_ANEG_SPEED_100:
mii->mii_media_active |= IFM_100_FX; break;
case BRGPHY_GP_STATUS_TOP_ANEG_SPEED_1G:
mii->mii_media_active |= IFM_1000_SX; break;
case BRGPHY_GP_STATUS_TOP_ANEG_SPEED_25G:
mii->mii_media_active |= IFM_2500_SX; break;
- Added support for 5709S/5716S PHYs. Submitted by: pyunyh MFC after: 2 weeks
2010-03-18 20:57:57 +00:00
}
Consistently use tab characters instead of tab + space characters. No functional changes.
2010-09-07 23:08:38 +00:00
/* Check for MRBE auto-negotiated duplex results. */
- Added support for 5709S/5716S PHYs. Submitted by: pyunyh MFC after: 2 weeks
2010-03-18 20:57:57 +00:00
if (xstat & BRGPHY_GP_STATUS_TOP_ANEG_FDX)
mii->mii_media_active |= IFM_FDX;
else
mii->mii_media_active |= IFM_HDX;
Consistently use tab characters instead of tab + space characters. No functional changes.
2010-09-07 23:08:38 +00:00
}
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
}
Add PHY drivers for the XMAC II's internal PHY and the Broadcom BCM5400 1000baseTX PHY. These will be used by the SysKonnect gigabit ethernet driver shortly.
2000-04-22 01:58:18 +00:00
}
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
static void
o Flesh out the generic IEEE 802.3 annex 31B full duplex flow control support in mii(4): - Merge generic flow control advertisement (which can be enabled by passing by MIIF_DOPAUSE to mii_attach(9)) and parsing support from NetBSD into mii_physubr.c and ukphy_subr.c. Unlike as in NetBSD, IFM_FLOW isn't implemented as a global option via the "don't care mask" but instead as a media specific option this. This has the following advantages: o allows flow control advertisement with autonegotiation to be turned on and off via ifconfig(8) with the default typically being off (though MIIF_FORCEPAUSE has been added causing flow control to be always advertised, allowing to easily MFC this changes for drivers that previously used home-grown support for flow control that behaved that way without breaking POLA) o allows to deal with PHY drivers where flow control advertisement with manual selection doesn't work or at least isn't implemented, like it's the case with brgphy(4), e1000phy(4) and ip1000phy(4), by setting MIIF_NOMANPAUSE o the available combinations of media options are readily available from the `ifconfig -m` output - Add IFM_FLOW to IFM_SHARED_OPTION_DESCRIPTIONS and IFM_ETH_RXPAUSE and IFM_ETH_TXPAUSE to IFM_SUBTYPE_ETHERNET_OPTION_DESCRIPTIONS so these are understood by ifconfig(8). o Make the master/slave support in mii(4) actually usable: - Change IFM_ETH_MASTER from being implemented as a global option via the "don't care mask" to a media specific one as it actually is only applicable to IFM_1000_T to date. - Let mii_phy_setmedia() set GTCR_MAN_MS in IFM_1000_T slave mode to actually configure manually selected slave mode (like we also do in the PHY specific implementations). - Add IFM_ETH_MASTER to IFM_SUBTYPE_ETHERNET_OPTION_DESCRIPTIONS so it is understood by ifconfig(8). o Switch bge(4), bce(4), msk(4), nfe(4) and stge(4) along with brgphy(4), e1000phy(4) and ip1000phy(4) to use the generic flow control support instead of home-grown solutions via IFM_FLAGs. This includes changing these PHY drivers and smcphy(4) to no longer unconditionally advertise support for flow control but only if the selected media has IFM_FLOW set (or MIIF_FORCEPAUSE is set) and implemented for these media variants, i.e. typically only for copper. o Switch brgphy(4), ciphy(4), e1000phy(4) and ip1000phy(4) to report and set IFM_1000_T master mode via IFM_ETH_MASTER instead of via IFF_LINK0 and some IFM_FLAGn. o Switch brgphy(4) to add at least the the supported copper media based on the contents of the BMSR via mii_phy_add_media() instead of hardcoding them. The latter approach seems to have developed historically, besides causing unnecessary code duplication it was also undesirable because brgphy_mii_phy_auto() already based the capability advertisement on the contents of the BMSR though. o Let brgphy(4) set IFM_1000_T master mode on all supported PHY and not just BCM5701. Apparently this was a misinterpretation of a workaround in the Linux tg3 driver; BCM5701 seem to require RGPHY_1000CTL_MSE and BRGPHY_1000CTL_MSC to be set when configuring autonegotiation but this doesn't mean we can't set these as well on other PHYs for manual media selection. o Let ukphy_status() report IFM_1000_T master mode via IFM_ETH_MASTER so IFM_1000_T master mode support now is generally available with all PHY drivers. o Don't let e1000phy(4) set master/slave bits for IFM_1000_SX as it's not applicable there. Reviewed by: yongari (plus additional testing) Obtained from: NetBSD (partially), OpenBSD (partially) MFC after: 2 weeks
2010-11-14 13:26:10 +00:00
brgphy_mii_phy_auto(struct mii_softc *sc, int media)
Add PHY drivers for the XMAC II's internal PHY and the Broadcom BCM5400 1000baseTX PHY. These will be used by the SysKonnect gigabit ethernet driver shortly.
2000-04-22 01:58:18 +00:00
{
o Flesh out the generic IEEE 802.3 annex 31B full duplex flow control support in mii(4): - Merge generic flow control advertisement (which can be enabled by passing by MIIF_DOPAUSE to mii_attach(9)) and parsing support from NetBSD into mii_physubr.c and ukphy_subr.c. Unlike as in NetBSD, IFM_FLOW isn't implemented as a global option via the "don't care mask" but instead as a media specific option this. This has the following advantages: o allows flow control advertisement with autonegotiation to be turned on and off via ifconfig(8) with the default typically being off (though MIIF_FORCEPAUSE has been added causing flow control to be always advertised, allowing to easily MFC this changes for drivers that previously used home-grown support for flow control that behaved that way without breaking POLA) o allows to deal with PHY drivers where flow control advertisement with manual selection doesn't work or at least isn't implemented, like it's the case with brgphy(4), e1000phy(4) and ip1000phy(4), by setting MIIF_NOMANPAUSE o the available combinations of media options are readily available from the `ifconfig -m` output - Add IFM_FLOW to IFM_SHARED_OPTION_DESCRIPTIONS and IFM_ETH_RXPAUSE and IFM_ETH_TXPAUSE to IFM_SUBTYPE_ETHERNET_OPTION_DESCRIPTIONS so these are understood by ifconfig(8). o Make the master/slave support in mii(4) actually usable: - Change IFM_ETH_MASTER from being implemented as a global option via the "don't care mask" to a media specific one as it actually is only applicable to IFM_1000_T to date. - Let mii_phy_setmedia() set GTCR_MAN_MS in IFM_1000_T slave mode to actually configure manually selected slave mode (like we also do in the PHY specific implementations). - Add IFM_ETH_MASTER to IFM_SUBTYPE_ETHERNET_OPTION_DESCRIPTIONS so it is understood by ifconfig(8). o Switch bge(4), bce(4), msk(4), nfe(4) and stge(4) along with brgphy(4), e1000phy(4) and ip1000phy(4) to use the generic flow control support instead of home-grown solutions via IFM_FLAGs. This includes changing these PHY drivers and smcphy(4) to no longer unconditionally advertise support for flow control but only if the selected media has IFM_FLOW set (or MIIF_FORCEPAUSE is set) and implemented for these media variants, i.e. typically only for copper. o Switch brgphy(4), ciphy(4), e1000phy(4) and ip1000phy(4) to report and set IFM_1000_T master mode via IFM_ETH_MASTER instead of via IFF_LINK0 and some IFM_FLAGn. o Switch brgphy(4) to add at least the the supported copper media based on the contents of the BMSR via mii_phy_add_media() instead of hardcoding them. The latter approach seems to have developed historically, besides causing unnecessary code duplication it was also undesirable because brgphy_mii_phy_auto() already based the capability advertisement on the contents of the BMSR though. o Let brgphy(4) set IFM_1000_T master mode on all supported PHY and not just BCM5701. Apparently this was a misinterpretation of a workaround in the Linux tg3 driver; BCM5701 seem to require RGPHY_1000CTL_MSE and BRGPHY_1000CTL_MSC to be set when configuring autonegotiation but this doesn't mean we can't set these as well on other PHYs for manual media selection. o Let ukphy_status() report IFM_1000_T master mode via IFM_ETH_MASTER so IFM_1000_T master mode support now is generally available with all PHY drivers. o Don't let e1000phy(4) set master/slave bits for IFM_1000_SX as it's not applicable there. Reviewed by: yongari (plus additional testing) Obtained from: NetBSD (partially), OpenBSD (partially) MFC after: 2 weeks
2010-11-14 13:26:10 +00:00
int anar, ktcr = 0;
Clean up mii/phy drivers: Remove the MIIF_DOINGAUTO which doesn't really do anything at the end of the day except bloat the drivers which has copy&pasted it.
2002-05-04 11:00:30 +00:00
- Remove attempts to implement setting of BMCR_LOOP/MIIF_NOLOOP (reporting IFM_LOOP based on BMCR_LOOP is left in place though as it might provide useful for debugging). For most mii(4) drivers it was unclear whether the PHYs driven by them actually support loopback or not. Moreover, typically loopback mode also needs to be activated on the MAC, which none of the Ethernet drivers using mii(4) implements. Given that loopback media has no real use (and obviously hardly had a chance to actually work) besides for driver development (which just loopback mode should be sufficient for though, i.e one doesn't necessary need support for loopback media) support for it is just dropped as both NetBSD and OpenBSD already did quite some time ago. - Let mii_phy_add_media() also announce the support of IFM_NONE. - 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 lxtphy(4) which disables MII interrupts (as is done for a few other PHYs we have drivers for). This includes changing NIC drivers which previously just called the generic mii_phy_reset() to now actually call the PHY-specific reset routine, which might be crucial in some cases. While at it, the redundant checks in these NIC drivers for mii->mii_instance not being zero before calling the reset routines were removed because as soon as one PHY driver attaches mii->mii_instance is incremented and we hardly can end up in their media change callbacks etc if no PHY driver has attached as mii_attach() would have failed in that case and not attach a miibus(4) instance. Consequently, NIC drivers now no longer should call mii_phy_reset() directly, so it was removed from EXPORT_SYMS. - Add a mii_phy_dev_attach() as a companion helper to mii_phy_dev_probe(). The purpose of that function is to perform the common steps to attach a PHY driver instance and to hook it up to the miibus(4) instance and to optionally also handle the probing, addition and initialization of the supported media. So all a PHY driver without any special requirements has to do in its bus attach method is to call mii_phy_dev_attach() along with PHY-specific MIIF_* flags, a pointer to its PHY functions and the add_media set to one. All PHY drivers were updated to take advantage of mii_phy_dev_attach() as appropriate. Along with these changes the capability mask was added to the mii_softc structure so PHY drivers taking advantage of mii_phy_dev_attach() but still handling media on their own do not need to fiddle with the MII attach arguments anyway. - Keep track of the PHY offset in the mii_softc structure. This is done for compatibility with NetBSD/OpenBSD. - Keep track of the PHY's OUI, model and revision in the mii_softc structure. Several PHY drivers require this information also after attaching and previously had to wrap their own softc around mii_softc. NetBSD/OpenBSD also keep track of the model and revision on their mii_softc structure. All PHY drivers were updated to take advantage as appropriate. - Convert the mebers of the MII data structure to unsigned where appropriate. This is partly inspired by NetBSD/OpenBSD. - According to IEEE 802.3-2002 the bits actually have to be reversed when mapping an OUI to the MII ID registers. All PHY drivers and miidevs where changed as necessary. Actually this now again allows to largely share miidevs with NetBSD, which fixed this problem already 9 years ago. Consequently miidevs was synced as far as possible. - Add MIIF_NOMANPAUSE and mii_phy_flowstatus() calls to drivers that weren't explicitly converted to support flow control before. It's unclear whether flow control actually works with these but typically it should and their net behavior should be more correct with these changes in place than without if the MAC driver sets MIIF_DOPAUSE. Obtained from: NetBSD (partially) Reviewed by: yongari (earlier version), silence on arch@ and net@
2011-05-03 19:51:29 +00:00
PHY_RESET(sc);
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
if ((sc->mii_flags & MIIF_HAVEFIBER) == 0) {
o Flesh out the generic IEEE 802.3 annex 31B full duplex flow control support in mii(4): - Merge generic flow control advertisement (which can be enabled by passing by MIIF_DOPAUSE to mii_attach(9)) and parsing support from NetBSD into mii_physubr.c and ukphy_subr.c. Unlike as in NetBSD, IFM_FLOW isn't implemented as a global option via the "don't care mask" but instead as a media specific option this. This has the following advantages: o allows flow control advertisement with autonegotiation to be turned on and off via ifconfig(8) with the default typically being off (though MIIF_FORCEPAUSE has been added causing flow control to be always advertised, allowing to easily MFC this changes for drivers that previously used home-grown support for flow control that behaved that way without breaking POLA) o allows to deal with PHY drivers where flow control advertisement with manual selection doesn't work or at least isn't implemented, like it's the case with brgphy(4), e1000phy(4) and ip1000phy(4), by setting MIIF_NOMANPAUSE o the available combinations of media options are readily available from the `ifconfig -m` output - Add IFM_FLOW to IFM_SHARED_OPTION_DESCRIPTIONS and IFM_ETH_RXPAUSE and IFM_ETH_TXPAUSE to IFM_SUBTYPE_ETHERNET_OPTION_DESCRIPTIONS so these are understood by ifconfig(8). o Make the master/slave support in mii(4) actually usable: - Change IFM_ETH_MASTER from being implemented as a global option via the "don't care mask" to a media specific one as it actually is only applicable to IFM_1000_T to date. - Let mii_phy_setmedia() set GTCR_MAN_MS in IFM_1000_T slave mode to actually configure manually selected slave mode (like we also do in the PHY specific implementations). - Add IFM_ETH_MASTER to IFM_SUBTYPE_ETHERNET_OPTION_DESCRIPTIONS so it is understood by ifconfig(8). o Switch bge(4), bce(4), msk(4), nfe(4) and stge(4) along with brgphy(4), e1000phy(4) and ip1000phy(4) to use the generic flow control support instead of home-grown solutions via IFM_FLAGs. This includes changing these PHY drivers and smcphy(4) to no longer unconditionally advertise support for flow control but only if the selected media has IFM_FLOW set (or MIIF_FORCEPAUSE is set) and implemented for these media variants, i.e. typically only for copper. o Switch brgphy(4), ciphy(4), e1000phy(4) and ip1000phy(4) to report and set IFM_1000_T master mode via IFM_ETH_MASTER instead of via IFF_LINK0 and some IFM_FLAGn. o Switch brgphy(4) to add at least the the supported copper media based on the contents of the BMSR via mii_phy_add_media() instead of hardcoding them. The latter approach seems to have developed historically, besides causing unnecessary code duplication it was also undesirable because brgphy_mii_phy_auto() already based the capability advertisement on the contents of the BMSR though. o Let brgphy(4) set IFM_1000_T master mode on all supported PHY and not just BCM5701. Apparently this was a misinterpretation of a workaround in the Linux tg3 driver; BCM5701 seem to require RGPHY_1000CTL_MSE and BRGPHY_1000CTL_MSC to be set when configuring autonegotiation but this doesn't mean we can't set these as well on other PHYs for manual media selection. o Let ukphy_status() report IFM_1000_T master mode via IFM_ETH_MASTER so IFM_1000_T master mode support now is generally available with all PHY drivers. o Don't let e1000phy(4) set master/slave bits for IFM_1000_SX as it's not applicable there. Reviewed by: yongari (plus additional testing) Obtained from: NetBSD (partially), OpenBSD (partially) MFC after: 2 weeks
2010-11-14 13:26:10 +00:00
anar = BMSR_MEDIA_TO_ANAR(sc->mii_capabilities) | ANAR_CSMA;
if ((media & IFM_FLOW) != 0 ||
(sc->mii_flags & MIIF_FORCEPAUSE) != 0)
anar |= BRGPHY_ANAR_PC | BRGPHY_ANAR_ASP;
PHY_WRITE(sc, BRGPHY_MII_ANAR, anar);
For fiber PHYs, BRGPHY_MII_1000CTL register is not defined at all so do not touch it.
2012-12-20 04:47:31 +00:00
ktcr = BRGPHY_1000CTL_AFD | BRGPHY_1000CTL_AHD;
if (sc->mii_mpd_model == MII_MODEL_BROADCOM_BCM5701)
ktcr |= BRGPHY_1000CTL_MSE | BRGPHY_1000CTL_MSC;
PHY_WRITE(sc, BRGPHY_MII_1000CTL, ktcr);
PHY_READ(sc, BRGPHY_MII_1000CTL);
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
} else {
o Flesh out the generic IEEE 802.3 annex 31B full duplex flow control support in mii(4): - Merge generic flow control advertisement (which can be enabled by passing by MIIF_DOPAUSE to mii_attach(9)) and parsing support from NetBSD into mii_physubr.c and ukphy_subr.c. Unlike as in NetBSD, IFM_FLOW isn't implemented as a global option via the "don't care mask" but instead as a media specific option this. This has the following advantages: o allows flow control advertisement with autonegotiation to be turned on and off via ifconfig(8) with the default typically being off (though MIIF_FORCEPAUSE has been added causing flow control to be always advertised, allowing to easily MFC this changes for drivers that previously used home-grown support for flow control that behaved that way without breaking POLA) o allows to deal with PHY drivers where flow control advertisement with manual selection doesn't work or at least isn't implemented, like it's the case with brgphy(4), e1000phy(4) and ip1000phy(4), by setting MIIF_NOMANPAUSE o the available combinations of media options are readily available from the `ifconfig -m` output - Add IFM_FLOW to IFM_SHARED_OPTION_DESCRIPTIONS and IFM_ETH_RXPAUSE and IFM_ETH_TXPAUSE to IFM_SUBTYPE_ETHERNET_OPTION_DESCRIPTIONS so these are understood by ifconfig(8). o Make the master/slave support in mii(4) actually usable: - Change IFM_ETH_MASTER from being implemented as a global option via the "don't care mask" to a media specific one as it actually is only applicable to IFM_1000_T to date. - Let mii_phy_setmedia() set GTCR_MAN_MS in IFM_1000_T slave mode to actually configure manually selected slave mode (like we also do in the PHY specific implementations). - Add IFM_ETH_MASTER to IFM_SUBTYPE_ETHERNET_OPTION_DESCRIPTIONS so it is understood by ifconfig(8). o Switch bge(4), bce(4), msk(4), nfe(4) and stge(4) along with brgphy(4), e1000phy(4) and ip1000phy(4) to use the generic flow control support instead of home-grown solutions via IFM_FLAGs. This includes changing these PHY drivers and smcphy(4) to no longer unconditionally advertise support for flow control but only if the selected media has IFM_FLOW set (or MIIF_FORCEPAUSE is set) and implemented for these media variants, i.e. typically only for copper. o Switch brgphy(4), ciphy(4), e1000phy(4) and ip1000phy(4) to report and set IFM_1000_T master mode via IFM_ETH_MASTER instead of via IFF_LINK0 and some IFM_FLAGn. o Switch brgphy(4) to add at least the the supported copper media based on the contents of the BMSR via mii_phy_add_media() instead of hardcoding them. The latter approach seems to have developed historically, besides causing unnecessary code duplication it was also undesirable because brgphy_mii_phy_auto() already based the capability advertisement on the contents of the BMSR though. o Let brgphy(4) set IFM_1000_T master mode on all supported PHY and not just BCM5701. Apparently this was a misinterpretation of a workaround in the Linux tg3 driver; BCM5701 seem to require RGPHY_1000CTL_MSE and BRGPHY_1000CTL_MSC to be set when configuring autonegotiation but this doesn't mean we can't set these as well on other PHYs for manual media selection. o Let ukphy_status() report IFM_1000_T master mode via IFM_ETH_MASTER so IFM_1000_T master mode support now is generally available with all PHY drivers. o Don't let e1000phy(4) set master/slave bits for IFM_1000_SX as it's not applicable there. Reviewed by: yongari (plus additional testing) Obtained from: NetBSD (partially), OpenBSD (partially) MFC after: 2 weeks
2010-11-14 13:26:10 +00:00
anar = BRGPHY_SERDES_ANAR_FDX | BRGPHY_SERDES_ANAR_HDX;
if ((media & IFM_FLOW) != 0 ||
(sc->mii_flags & MIIF_FORCEPAUSE) != 0)
anar |= BRGPHY_SERDES_ANAR_BOTH_PAUSE;
PHY_WRITE(sc, BRGPHY_SERDES_ANAR, anar);
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
}
o Flesh out the generic IEEE 802.3 annex 31B full duplex flow control support in mii(4): - Merge generic flow control advertisement (which can be enabled by passing by MIIF_DOPAUSE to mii_attach(9)) and parsing support from NetBSD into mii_physubr.c and ukphy_subr.c. Unlike as in NetBSD, IFM_FLOW isn't implemented as a global option via the "don't care mask" but instead as a media specific option this. This has the following advantages: o allows flow control advertisement with autonegotiation to be turned on and off via ifconfig(8) with the default typically being off (though MIIF_FORCEPAUSE has been added causing flow control to be always advertised, allowing to easily MFC this changes for drivers that previously used home-grown support for flow control that behaved that way without breaking POLA) o allows to deal with PHY drivers where flow control advertisement with manual selection doesn't work or at least isn't implemented, like it's the case with brgphy(4), e1000phy(4) and ip1000phy(4), by setting MIIF_NOMANPAUSE o the available combinations of media options are readily available from the `ifconfig -m` output - Add IFM_FLOW to IFM_SHARED_OPTION_DESCRIPTIONS and IFM_ETH_RXPAUSE and IFM_ETH_TXPAUSE to IFM_SUBTYPE_ETHERNET_OPTION_DESCRIPTIONS so these are understood by ifconfig(8). o Make the master/slave support in mii(4) actually usable: - Change IFM_ETH_MASTER from being implemented as a global option via the "don't care mask" to a media specific one as it actually is only applicable to IFM_1000_T to date. - Let mii_phy_setmedia() set GTCR_MAN_MS in IFM_1000_T slave mode to actually configure manually selected slave mode (like we also do in the PHY specific implementations). - Add IFM_ETH_MASTER to IFM_SUBTYPE_ETHERNET_OPTION_DESCRIPTIONS so it is understood by ifconfig(8). o Switch bge(4), bce(4), msk(4), nfe(4) and stge(4) along with brgphy(4), e1000phy(4) and ip1000phy(4) to use the generic flow control support instead of home-grown solutions via IFM_FLAGs. This includes changing these PHY drivers and smcphy(4) to no longer unconditionally advertise support for flow control but only if the selected media has IFM_FLOW set (or MIIF_FORCEPAUSE is set) and implemented for these media variants, i.e. typically only for copper. o Switch brgphy(4), ciphy(4), e1000phy(4) and ip1000phy(4) to report and set IFM_1000_T master mode via IFM_ETH_MASTER instead of via IFF_LINK0 and some IFM_FLAGn. o Switch brgphy(4) to add at least the the supported copper media based on the contents of the BMSR via mii_phy_add_media() instead of hardcoding them. The latter approach seems to have developed historically, besides causing unnecessary code duplication it was also undesirable because brgphy_mii_phy_auto() already based the capability advertisement on the contents of the BMSR though. o Let brgphy(4) set IFM_1000_T master mode on all supported PHY and not just BCM5701. Apparently this was a misinterpretation of a workaround in the Linux tg3 driver; BCM5701 seem to require RGPHY_1000CTL_MSE and BRGPHY_1000CTL_MSC to be set when configuring autonegotiation but this doesn't mean we can't set these as well on other PHYs for manual media selection. o Let ukphy_status() report IFM_1000_T master mode via IFM_ETH_MASTER so IFM_1000_T master mode support now is generally available with all PHY drivers. o Don't let e1000phy(4) set master/slave bits for IFM_1000_SX as it's not applicable there. Reviewed by: yongari (plus additional testing) Obtained from: NetBSD (partially), OpenBSD (partially) MFC after: 2 weeks
2010-11-14 13:26:10 +00:00
PHY_WRITE(sc, BRGPHY_MII_BMCR, BRGPHY_BMCR_AUTOEN |
BRGPHY_BMCR_STARTNEG);
- Do not depend on auto negotiation for link speed/duplex status. - Read link status from BMSR instead of auxilary status register. - Clean up style(9) nits.
2006-12-20 00:08:47 +00:00
PHY_WRITE(sc, BRGPHY_MII_IMR, 0xFF00);
Add PHY drivers for the XMAC II's internal PHY and the Broadcom BCM5400 1000baseTX PHY. These will be used by the SysKonnect gigabit ethernet driver shortly.
2000-04-22 01:58:18 +00:00
}
- Move bge_phy_hack into the phy code and implement the various DSP patch workarounds for each phy revision. Obtained from: NetBSD & Broadcom Linux driver - Disable AUTOPOLL when accessing the PHY as it may cause PCI errors. Obtained from: NetBSD - Check the UPDATED bit in the status block so the driver knows that the status block as indeed changed since the last access. Broadcom documentation states drivers should unset the UPDATED/CHANGED bits after reading them. - When changing media types, first loop the phy then set the media. Broadcom documentation and Linux drivers do this and I observed much better handling of link after this change. - Broadcom documentation states that for 1000BaseT operation, autonegotiation must be enabled. Fix hard coding of media so that the driver only advertises 1000BaseT as the supported media type and enable autonegotition. - Only set Master/Slave on the 5701. Obtained from Broadcom Linux driver.
2003-05-03 19:06:50 +00:00
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
/* Enable loopback to force the link down. */
- Move bge_phy_hack into the phy code and implement the various DSP patch workarounds for each phy revision. Obtained from: NetBSD & Broadcom Linux driver - Disable AUTOPOLL when accessing the PHY as it may cause PCI errors. Obtained from: NetBSD - Check the UPDATED bit in the status block so the driver knows that the status block as indeed changed since the last access. Broadcom documentation states drivers should unset the UPDATED/CHANGED bits after reading them. - When changing media types, first loop the phy then set the media. Broadcom documentation and Linux drivers do this and I observed much better handling of link after this change. - Broadcom documentation states that for 1000BaseT operation, autonegotiation must be enabled. Fix hard coding of media so that the driver only advertises 1000BaseT as the supported media type and enable autonegotition. - Only set Master/Slave on the 5701. Obtained from Broadcom Linux driver.
2003-05-03 19:06:50 +00:00
static void
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
brgphy_enable_loopback(struct mii_softc *sc)
- Move bge_phy_hack into the phy code and implement the various DSP patch workarounds for each phy revision. Obtained from: NetBSD & Broadcom Linux driver - Disable AUTOPOLL when accessing the PHY as it may cause PCI errors. Obtained from: NetBSD - Check the UPDATED bit in the status block so the driver knows that the status block as indeed changed since the last access. Broadcom documentation states drivers should unset the UPDATED/CHANGED bits after reading them. - When changing media types, first loop the phy then set the media. Broadcom documentation and Linux drivers do this and I observed much better handling of link after this change. - Broadcom documentation states that for 1000BaseT operation, autonegotiation must be enabled. Fix hard coding of media so that the driver only advertises 1000BaseT as the supported media type and enable autonegotition. - Only set Master/Slave on the 5701. Obtained from Broadcom Linux driver.
2003-05-03 19:06:50 +00:00
{
int i;
PHY_WRITE(sc, BRGPHY_MII_BMCR, BRGPHY_BMCR_LOOP);
for (i = 0; i < 15000; i++) {
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
if (!(PHY_READ(sc, BRGPHY_MII_BMSR) & BRGPHY_BMSR_LINK))
- Move bge_phy_hack into the phy code and implement the various DSP patch workarounds for each phy revision. Obtained from: NetBSD & Broadcom Linux driver - Disable AUTOPOLL when accessing the PHY as it may cause PCI errors. Obtained from: NetBSD - Check the UPDATED bit in the status block so the driver knows that the status block as indeed changed since the last access. Broadcom documentation states drivers should unset the UPDATED/CHANGED bits after reading them. - When changing media types, first loop the phy then set the media. Broadcom documentation and Linux drivers do this and I observed much better handling of link after this change. - Broadcom documentation states that for 1000BaseT operation, autonegotiation must be enabled. Fix hard coding of media so that the driver only advertises 1000BaseT as the supported media type and enable autonegotition. - Only set Master/Slave on the 5701. Obtained from Broadcom Linux driver.
2003-05-03 19:06:50 +00:00
break;
DELAY(10);
}
}
/* Turn off tap power management on 5401. */
static void
bcm5401_load_dspcode(struct mii_softc *sc)
{
static const struct {
int reg;
uint16_t val;
} dspcode[] = {
{ BRGPHY_MII_AUXCTL, 0x0c20 },
{ BRGPHY_MII_DSP_ADDR_REG, 0x0012 },
{ BRGPHY_MII_DSP_RW_PORT, 0x1804 },
{ BRGPHY_MII_DSP_ADDR_REG, 0x0013 },
{ BRGPHY_MII_DSP_RW_PORT, 0x1204 },
{ BRGPHY_MII_DSP_ADDR_REG, 0x8006 },
{ BRGPHY_MII_DSP_RW_PORT, 0x0132 },
{ BRGPHY_MII_DSP_ADDR_REG, 0x8006 },
{ BRGPHY_MII_DSP_RW_PORT, 0x0232 },
{ BRGPHY_MII_DSP_ADDR_REG, 0x201f },
{ BRGPHY_MII_DSP_RW_PORT, 0x0a20 },
{ 0, 0 },
};
int i;
for (i = 0; dspcode[i].reg != 0; i++)
PHY_WRITE(sc, dspcode[i].reg, dspcode[i].val);
DELAY(40);
}
static void
bcm5411_load_dspcode(struct mii_softc *sc)
{
static const struct {
int reg;
uint16_t val;
} dspcode[] = {
{ 0x1c, 0x8c23 },
{ 0x1c, 0x8ca3 },
{ 0x1c, 0x8c23 },
{ 0, 0 },
};
int i;
for (i = 0; dspcode[i].reg != 0; i++)
PHY_WRITE(sc, dspcode[i].reg, dspcode[i].val);
}
Add support for BCM54K2 found in combination with Apple K2 GMAC. Submitted by: Andreas Tobler Obtained from: OpenBSD MFC after: 1 week
2010-02-20 22:01:24 +00:00
void
bcm54k2_load_dspcode(struct mii_softc *sc)
{
static const struct {
int reg;
uint16_t val;
} dspcode[] = {
{ 4, 0x01e1 },
{ 9, 0x0300 },
{ 0, 0 },
};
int i;
for (i = 0; dspcode[i].reg != 0; i++)
PHY_WRITE(sc, dspcode[i].reg, dspcode[i].val);
}
- Move bge_phy_hack into the phy code and implement the various DSP patch workarounds for each phy revision. Obtained from: NetBSD & Broadcom Linux driver - Disable AUTOPOLL when accessing the PHY as it may cause PCI errors. Obtained from: NetBSD - Check the UPDATED bit in the status block so the driver knows that the status block as indeed changed since the last access. Broadcom documentation states drivers should unset the UPDATED/CHANGED bits after reading them. - When changing media types, first loop the phy then set the media. Broadcom documentation and Linux drivers do this and I observed much better handling of link after this change. - Broadcom documentation states that for 1000BaseT operation, autonegotiation must be enabled. Fix hard coding of media so that the driver only advertises 1000BaseT as the supported media type and enable autonegotition. - Only set Master/Slave on the 5701. Obtained from Broadcom Linux driver.
2003-05-03 19:06:50 +00:00
static void
Fix style(9). Pointed out by: many
2007-02-12 23:33:05 +00:00
brgphy_fixup_5704_a0_bug(struct mii_softc *sc)
- Move bge_phy_hack into the phy code and implement the various DSP patch workarounds for each phy revision. Obtained from: NetBSD & Broadcom Linux driver - Disable AUTOPOLL when accessing the PHY as it may cause PCI errors. Obtained from: NetBSD - Check the UPDATED bit in the status block so the driver knows that the status block as indeed changed since the last access. Broadcom documentation states drivers should unset the UPDATED/CHANGED bits after reading them. - When changing media types, first loop the phy then set the media. Broadcom documentation and Linux drivers do this and I observed much better handling of link after this change. - Broadcom documentation states that for 1000BaseT operation, autonegotiation must be enabled. Fix hard coding of media so that the driver only advertises 1000BaseT as the supported media type and enable autonegotition. - Only set Master/Slave on the 5701. Obtained from Broadcom Linux driver.
2003-05-03 19:06:50 +00:00
{
static const struct {
int reg;
uint16_t val;
} dspcode[] = {
Fix style(9). Pointed out by: many
2007-02-12 23:33:05 +00:00
{ 0x1c, 0x8d68 },
{ 0x1c, 0x8d68 },
- Move bge_phy_hack into the phy code and implement the various DSP patch workarounds for each phy revision. Obtained from: NetBSD & Broadcom Linux driver - Disable AUTOPOLL when accessing the PHY as it may cause PCI errors. Obtained from: NetBSD - Check the UPDATED bit in the status block so the driver knows that the status block as indeed changed since the last access. Broadcom documentation states drivers should unset the UPDATED/CHANGED bits after reading them. - When changing media types, first loop the phy then set the media. Broadcom documentation and Linux drivers do this and I observed much better handling of link after this change. - Broadcom documentation states that for 1000BaseT operation, autonegotiation must be enabled. Fix hard coding of media so that the driver only advertises 1000BaseT as the supported media type and enable autonegotition. - Only set Master/Slave on the 5701. Obtained from Broadcom Linux driver.
2003-05-03 19:06:50 +00:00
{ 0, 0 },
};
int i;
for (i = 0; dspcode[i].reg != 0; i++)
PHY_WRITE(sc, dspcode[i].reg, dspcode[i].val);
}
static void
Fix style(9). Pointed out by: many
2007-02-12 23:33:05 +00:00
brgphy_fixup_adc_bug(struct mii_softc *sc)
- Move bge_phy_hack into the phy code and implement the various DSP patch workarounds for each phy revision. Obtained from: NetBSD & Broadcom Linux driver - Disable AUTOPOLL when accessing the PHY as it may cause PCI errors. Obtained from: NetBSD - Check the UPDATED bit in the status block so the driver knows that the status block as indeed changed since the last access. Broadcom documentation states drivers should unset the UPDATED/CHANGED bits after reading them. - When changing media types, first loop the phy then set the media. Broadcom documentation and Linux drivers do this and I observed much better handling of link after this change. - Broadcom documentation states that for 1000BaseT operation, autonegotiation must be enabled. Fix hard coding of media so that the driver only advertises 1000BaseT as the supported media type and enable autonegotition. - Only set Master/Slave on the 5701. Obtained from Broadcom Linux driver.
2003-05-03 19:06:50 +00:00
{
static const struct {
int reg;
Fix style(9). Pointed out by: many
2007-02-12 23:33:05 +00:00
uint16_t val;
- Move bge_phy_hack into the phy code and implement the various DSP patch workarounds for each phy revision. Obtained from: NetBSD & Broadcom Linux driver - Disable AUTOPOLL when accessing the PHY as it may cause PCI errors. Obtained from: NetBSD - Check the UPDATED bit in the status block so the driver knows that the status block as indeed changed since the last access. Broadcom documentation states drivers should unset the UPDATED/CHANGED bits after reading them. - When changing media types, first loop the phy then set the media. Broadcom documentation and Linux drivers do this and I observed much better handling of link after this change. - Broadcom documentation states that for 1000BaseT operation, autonegotiation must be enabled. Fix hard coding of media so that the driver only advertises 1000BaseT as the supported media type and enable autonegotition. - Only set Master/Slave on the 5701. Obtained from Broadcom Linux driver.
2003-05-03 19:06:50 +00:00
} dspcode[] = {
Fix style(9). Pointed out by: many
2007-02-12 23:33:05 +00:00
{ BRGPHY_MII_AUXCTL, 0x0c00 },
{ BRGPHY_MII_DSP_ADDR_REG, 0x201f },
{ BRGPHY_MII_DSP_RW_PORT, 0x2aaa },
- Move bge_phy_hack into the phy code and implement the various DSP patch workarounds for each phy revision. Obtained from: NetBSD & Broadcom Linux driver - Disable AUTOPOLL when accessing the PHY as it may cause PCI errors. Obtained from: NetBSD - Check the UPDATED bit in the status block so the driver knows that the status block as indeed changed since the last access. Broadcom documentation states drivers should unset the UPDATED/CHANGED bits after reading them. - When changing media types, first loop the phy then set the media. Broadcom documentation and Linux drivers do this and I observed much better handling of link after this change. - Broadcom documentation states that for 1000BaseT operation, autonegotiation must be enabled. Fix hard coding of media so that the driver only advertises 1000BaseT as the supported media type and enable autonegotition. - Only set Master/Slave on the 5701. Obtained from Broadcom Linux driver.
2003-05-03 19:06:50 +00:00
{ 0, 0 },
};
int i;
for (i = 0; dspcode[i].reg != 0; i++)
PHY_WRITE(sc, dspcode[i].reg, dspcode[i].val);
}
Add PHY DSP code for BCM5755M. Obtained from: OpenBSD
2007-02-12 22:51:25 +00:00
static void
brgphy_fixup_adjust_trim(struct mii_softc *sc)
{
static const struct {
int reg;
uint16_t val;
} dspcode[] = {
{ BRGPHY_MII_AUXCTL, 0x0c00 },
{ BRGPHY_MII_DSP_ADDR_REG, 0x000a },
{ BRGPHY_MII_DSP_RW_PORT, 0x110b },
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
{ BRGPHY_MII_TEST1, 0x0014 },
Add PHY DSP code for BCM5755M. Obtained from: OpenBSD
2007-02-12 22:51:25 +00:00
{ BRGPHY_MII_AUXCTL, 0x0400 },
{ 0, 0 },
};
int i;
for (i = 0; dspcode[i].reg != 0; i++)
PHY_WRITE(sc, dspcode[i].reg, dspcode[i].val);
}
Add support for the BCM5750/5751. Unfortunately the documentation I have from Broadcom does not give much information on these devices, so the Broadcom Linux driver was used for clues to what these chips support. It turns out they are similar to the 5705 with the 5751 being the PCI-Express version and needing special work-arounds and settings.
2004-09-24 22:24:33 +00:00
static void
- Fix BCM5754 support found in Dell PowerEdge SC440. - Move some PHY bug detections from brgphy.c to if_bge.c. - Do not penalize working PHYs. - Re-arrange bge_flags roughly by their categories. - Fix minor style(9) nits. PR: kern/107257 Obtained from: OpenBSD Tested by: Mike Hibler <mike at flux dot utah dot edu>
2007-01-15 21:43:43 +00:00
brgphy_fixup_ber_bug(struct mii_softc *sc)
Add support for the BCM5750/5751. Unfortunately the documentation I have from Broadcom does not give much information on these devices, so the Broadcom Linux driver was used for clues to what these chips support. It turns out they are similar to the 5705 with the 5751 being the PCI-Express version and needing special work-arounds and settings.
2004-09-24 22:24:33 +00:00
{
static const struct {
int reg;
Fix style(9). Pointed out by: many
2007-02-12 23:33:05 +00:00
uint16_t val;
Add support for the BCM5750/5751. Unfortunately the documentation I have from Broadcom does not give much information on these devices, so the Broadcom Linux driver was used for clues to what these chips support. It turns out they are similar to the 5705 with the 5751 being the PCI-Express version and needing special work-arounds and settings.
2004-09-24 22:24:33 +00:00
} dspcode[] = {
Replace magic numbers with corresponding definitions.
2007-02-12 19:33:22 +00:00
{ BRGPHY_MII_AUXCTL, 0x0c00 },
{ BRGPHY_MII_DSP_ADDR_REG, 0x000a },
{ BRGPHY_MII_DSP_RW_PORT, 0x310b },
{ BRGPHY_MII_DSP_ADDR_REG, 0x201f },
{ BRGPHY_MII_DSP_RW_PORT, 0x9506 },
{ BRGPHY_MII_DSP_ADDR_REG, 0x401f },
{ BRGPHY_MII_DSP_RW_PORT, 0x14e2 },
{ BRGPHY_MII_AUXCTL, 0x0400 },
Add support for the BCM5750/5751. Unfortunately the documentation I have from Broadcom does not give much information on these devices, so the Broadcom Linux driver was used for clues to what these chips support. It turns out they are similar to the 5705 with the 5751 being the PCI-Express version and needing special work-arounds and settings.
2004-09-24 22:24:33 +00:00
{ 0, 0 },
};
int i;
for (i = 0; dspcode[i].reg != 0; i++)
PHY_WRITE(sc, dspcode[i].reg, dspcode[i].val);
}
Add BCM5701 A0/B0 CRC bug workaround. Magic values taken from Linux driver.
2007-02-12 23:58:52 +00:00
static void
brgphy_fixup_crc_bug(struct mii_softc *sc)
{
static const struct {
int reg;
uint16_t val;
} dspcode[] = {
Fix a typo from the previous commit. Pointed out by: brad@openbsd.org
2007-02-14 18:21:32 +00:00
{ BRGPHY_MII_DSP_RW_PORT, 0x0a75 },
Add BCM5701 A0/B0 CRC bug workaround. Magic values taken from Linux driver.
2007-02-12 23:58:52 +00:00
{ 0x1c, 0x8c68 },
{ 0x1c, 0x8d68 },
{ 0x1c, 0x8c68 },
{ 0, 0 },
};
int i;
for (i = 0; dspcode[i].reg != 0; i++)
PHY_WRITE(sc, dspcode[i].reg, dspcode[i].val);
}
- Fix BCM5754 support found in Dell PowerEdge SC440. - Move some PHY bug detections from brgphy.c to if_bge.c. - Do not penalize working PHYs. - Re-arrange bge_flags roughly by their categories. - Fix minor style(9) nits. PR: kern/107257 Obtained from: OpenBSD Tested by: Mike Hibler <mike at flux dot utah dot edu>
2007-01-15 21:43:43 +00:00
static void
brgphy_fixup_jitter_bug(struct mii_softc *sc)
{
static const struct {
int reg;
uint16_t val;
} dspcode[] = {
{ BRGPHY_MII_AUXCTL, 0x0c00 },
{ BRGPHY_MII_DSP_ADDR_REG, 0x000a },
{ BRGPHY_MII_DSP_RW_PORT, 0x010b },
{ BRGPHY_MII_AUXCTL, 0x0400 },
{ 0, 0 },
};
int i;
for (i = 0; dspcode[i].reg != 0; i++)
PHY_WRITE(sc, dspcode[i].reg, dspcode[i].val);
}
- Added support for BCM5709 and BCM5716. MFC after: 2 weeks
2008-06-13 01:20:29 +00:00
static void
brgphy_fixup_disable_early_dac(struct mii_softc *sc)
{
uint32_t val;
PHY_WRITE(sc, BRGPHY_MII_DSP_ADDR_REG, 0x0f08);
val = PHY_READ(sc, BRGPHY_MII_DSP_RW_PORT);
val &= ~(1 << 8);
PHY_WRITE(sc, BRGPHY_MII_DSP_RW_PORT, val);
}
- Move bge_phy_hack into the phy code and implement the various DSP patch workarounds for each phy revision. Obtained from: NetBSD & Broadcom Linux driver - Disable AUTOPOLL when accessing the PHY as it may cause PCI errors. Obtained from: NetBSD - Check the UPDATED bit in the status block so the driver knows that the status block as indeed changed since the last access. Broadcom documentation states drivers should unset the UPDATED/CHANGED bits after reading them. - When changing media types, first loop the phy then set the media. Broadcom documentation and Linux drivers do this and I observed much better handling of link after this change. - Broadcom documentation states that for 1000BaseT operation, autonegotiation must be enabled. Fix hard coding of media so that the driver only advertises 1000BaseT as the supported media type and enable autonegotition. - Only set Master/Slave on the 5701. Obtained from Broadcom Linux driver.
2003-05-03 19:06:50 +00:00
static void
- Move Ethernet@WireSpeed and jumbo frame configurations to separate functions. The idea is taken from OpenBSD. - Set/clear jumbo frame configurations for bge(4). - Re-add BCM5750 PHY workaround for bce(4), which was mistakenly removed from the previous commit.
2007-01-15 22:21:44 +00:00
brgphy_ethernet_wirespeed(struct mii_softc *sc)
{
Fix style(9). Pointed out by: many
2007-02-12 23:33:05 +00:00
uint32_t val;
- Move Ethernet@WireSpeed and jumbo frame configurations to separate functions. The idea is taken from OpenBSD. - Set/clear jumbo frame configurations for bge(4). - Re-add BCM5750 PHY workaround for bce(4), which was mistakenly removed from the previous commit.
2007-01-15 22:21:44 +00:00
/* Enable Ethernet@WireSpeed. */
PHY_WRITE(sc, BRGPHY_MII_AUXCTL, 0x7007);
val = PHY_READ(sc, BRGPHY_MII_AUXCTL);
PHY_WRITE(sc, BRGPHY_MII_AUXCTL, val | (1 << 15) | (1 << 4));
}
static void
brgphy_jumbo_settings(struct mii_softc *sc, u_long mtu)
- Move bge_phy_hack into the phy code and implement the various DSP patch workarounds for each phy revision. Obtained from: NetBSD & Broadcom Linux driver - Disable AUTOPOLL when accessing the PHY as it may cause PCI errors. Obtained from: NetBSD - Check the UPDATED bit in the status block so the driver knows that the status block as indeed changed since the last access. Broadcom documentation states drivers should unset the UPDATED/CHANGED bits after reading them. - When changing media types, first loop the phy then set the media. Broadcom documentation and Linux drivers do this and I observed much better handling of link after this change. - Broadcom documentation states that for 1000BaseT operation, autonegotiation must be enabled. Fix hard coding of media so that the driver only advertises 1000BaseT as the supported media type and enable autonegotition. - Only set Master/Slave on the 5701. Obtained from Broadcom Linux driver.
2003-05-03 19:06:50 +00:00
{
Fix style(9). Pointed out by: many
2007-02-12 23:33:05 +00:00
uint32_t val;
- Move Ethernet@WireSpeed and jumbo frame configurations to separate functions. The idea is taken from OpenBSD. - Set/clear jumbo frame configurations for bge(4). - Re-add BCM5750 PHY workaround for bce(4), which was mistakenly removed from the previous commit.
2007-01-15 22:21:44 +00:00
/* Set or clear jumbo frame settings in the PHY. */
if (mtu > ETHER_MAX_LEN) {
- Remove attempts to implement setting of BMCR_LOOP/MIIF_NOLOOP (reporting IFM_LOOP based on BMCR_LOOP is left in place though as it might provide useful for debugging). For most mii(4) drivers it was unclear whether the PHYs driven by them actually support loopback or not. Moreover, typically loopback mode also needs to be activated on the MAC, which none of the Ethernet drivers using mii(4) implements. Given that loopback media has no real use (and obviously hardly had a chance to actually work) besides for driver development (which just loopback mode should be sufficient for though, i.e one doesn't necessary need support for loopback media) support for it is just dropped as both NetBSD and OpenBSD already did quite some time ago. - Let mii_phy_add_media() also announce the support of IFM_NONE. - 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 lxtphy(4) which disables MII interrupts (as is done for a few other PHYs we have drivers for). This includes changing NIC drivers which previously just called the generic mii_phy_reset() to now actually call the PHY-specific reset routine, which might be crucial in some cases. While at it, the redundant checks in these NIC drivers for mii->mii_instance not being zero before calling the reset routines were removed because as soon as one PHY driver attaches mii->mii_instance is incremented and we hardly can end up in their media change callbacks etc if no PHY driver has attached as mii_attach() would have failed in that case and not attach a miibus(4) instance. Consequently, NIC drivers now no longer should call mii_phy_reset() directly, so it was removed from EXPORT_SYMS. - Add a mii_phy_dev_attach() as a companion helper to mii_phy_dev_probe(). The purpose of that function is to perform the common steps to attach a PHY driver instance and to hook it up to the miibus(4) instance and to optionally also handle the probing, addition and initialization of the supported media. So all a PHY driver without any special requirements has to do in its bus attach method is to call mii_phy_dev_attach() along with PHY-specific MIIF_* flags, a pointer to its PHY functions and the add_media set to one. All PHY drivers were updated to take advantage of mii_phy_dev_attach() as appropriate. Along with these changes the capability mask was added to the mii_softc structure so PHY drivers taking advantage of mii_phy_dev_attach() but still handling media on their own do not need to fiddle with the MII attach arguments anyway. - Keep track of the PHY offset in the mii_softc structure. This is done for compatibility with NetBSD/OpenBSD. - Keep track of the PHY's OUI, model and revision in the mii_softc structure. Several PHY drivers require this information also after attaching and previously had to wrap their own softc around mii_softc. NetBSD/OpenBSD also keep track of the model and revision on their mii_softc structure. All PHY drivers were updated to take advantage as appropriate. - Convert the mebers of the MII data structure to unsigned where appropriate. This is partly inspired by NetBSD/OpenBSD. - According to IEEE 802.3-2002 the bits actually have to be reversed when mapping an OUI to the MII ID registers. All PHY drivers and miidevs where changed as necessary. Actually this now again allows to largely share miidevs with NetBSD, which fixed this problem already 9 years ago. Consequently miidevs was synced as far as possible. - Add MIIF_NOMANPAUSE and mii_phy_flowstatus() calls to drivers that weren't explicitly converted to support flow control before. It's unclear whether flow control actually works with these but typically it should and their net behavior should be more correct with these changes in place than without if the MAC driver sets MIIF_DOPAUSE. Obtained from: NetBSD (partially) Reviewed by: yongari (earlier version), silence on arch@ and net@
2011-05-03 19:51:29 +00:00
if (sc->mii_mpd_model == MII_MODEL_BROADCOM_BCM5401) {
Fix typos in comments while I am here.
2007-02-13 00:34:32 +00:00
/* BCM5401 PHY cannot read-modify-write. */
BCM5701 PHY cannot read-modify-write. Just re-use the magic number from DSP init code.
2007-02-12 20:26:56 +00:00
PHY_WRITE(sc, BRGPHY_MII_AUXCTL, 0x4c20);
Fix typos in comments while I am here.
2007-02-13 00:34:32 +00:00
} else {
BCM5701 PHY cannot read-modify-write. Just re-use the magic number from DSP init code.
2007-02-12 20:26:56 +00:00
PHY_WRITE(sc, BRGPHY_MII_AUXCTL, 0x7);
val = PHY_READ(sc, BRGPHY_MII_AUXCTL);
PHY_WRITE(sc, BRGPHY_MII_AUXCTL,
val | BRGPHY_AUXCTL_LONG_PKT);
}
- Move Ethernet@WireSpeed and jumbo frame configurations to separate functions. The idea is taken from OpenBSD. - Set/clear jumbo frame configurations for bge(4). - Re-add BCM5750 PHY workaround for bce(4), which was mistakenly removed from the previous commit.
2007-01-15 22:21:44 +00:00
val = PHY_READ(sc, BRGPHY_MII_PHY_EXTCTL);
PHY_WRITE(sc, BRGPHY_MII_PHY_EXTCTL,
val | BRGPHY_PHY_EXTCTL_HIGH_LA);
} else {
PHY_WRITE(sc, BRGPHY_MII_AUXCTL, 0x7);
val = PHY_READ(sc, BRGPHY_MII_AUXCTL);
PHY_WRITE(sc, BRGPHY_MII_AUXCTL,
val & ~(BRGPHY_AUXCTL_LONG_PKT | 0x7));
val = PHY_READ(sc, BRGPHY_MII_PHY_EXTCTL);
Remove whitespace at the end of line.
2008-08-12 00:52:10 +00:00
PHY_WRITE(sc, BRGPHY_MII_PHY_EXTCTL,
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
val & ~BRGPHY_PHY_EXTCTL_HIGH_LA);
- Move Ethernet@WireSpeed and jumbo frame configurations to separate functions. The idea is taken from OpenBSD. - Set/clear jumbo frame configurations for bge(4). - Re-add BCM5750 PHY workaround for bce(4), which was mistakenly removed from the previous commit.
2007-01-15 22:21:44 +00:00
}
}
static void
brgphy_reset(struct mii_softc *sc)
{
Move MII model and revision into softc.
2007-01-16 00:52:26 +00:00
struct bge_softc *bge_sc = NULL;
struct bce_softc *bce_sc = NULL;
struct ifnet *ifp;
r221812 reveals that at least some Broadcom PHYs default to being not only isolated but also powered down after a reset and while they just work fine [sic] when both is the case they don't if they are only deisolate but still powered down. So in order to put PHYs in an overall normal operation mode for the common case, ensure in mii_phy_reset() that they are not powered down after a reset. Unfortunately, this only helps in case of BCM5421, while BCM5709S apparently only work when they remain isolated and powered down after a reset. So don't call mii_phy_reset() in brgphy_reset() and implement the reset locally leaving the problematic bits alone. Effectively this bypasses r221812 for brgphy(4). Thanks to Justin Hibbits for doing a binary search in order to identify the problematic commit. PR: 157405, 158156 Reviewed by: yongari (mii_phy_reset() part) Approved by: re (kib) MFC after: 3 days
2011-08-19 19:12:58 +00:00
int i, val;
/*
* Perform a reset. Note that at least some Broadcom PHYs default to
* being powered down as well as isolated after a reset but don't work
* if one or both of these bits are cleared. However, they just work
* fine if both bits remain set, so we don't use mii_phy_reset() here.
*/
PHY_WRITE(sc, BRGPHY_MII_BMCR, BRGPHY_BMCR_RESET);
- Move bge_phy_hack into the phy code and implement the various DSP patch workarounds for each phy revision. Obtained from: NetBSD & Broadcom Linux driver - Disable AUTOPOLL when accessing the PHY as it may cause PCI errors. Obtained from: NetBSD - Check the UPDATED bit in the status block so the driver knows that the status block as indeed changed since the last access. Broadcom documentation states drivers should unset the UPDATED/CHANGED bits after reading them. - When changing media types, first loop the phy then set the media. Broadcom documentation and Linux drivers do this and I observed much better handling of link after this change. - Broadcom documentation states that for 1000BaseT operation, autonegotiation must be enabled. Fix hard coding of media so that the driver only advertises 1000BaseT as the supported media type and enable autonegotition. - Only set Master/Slave on the 5701. Obtained from Broadcom Linux driver.
2003-05-03 19:06:50 +00:00
r221812 reveals that at least some Broadcom PHYs default to being not only isolated but also powered down after a reset and while they just work fine [sic] when both is the case they don't if they are only deisolate but still powered down. So in order to put PHYs in an overall normal operation mode for the common case, ensure in mii_phy_reset() that they are not powered down after a reset. Unfortunately, this only helps in case of BCM5421, while BCM5709S apparently only work when they remain isolated and powered down after a reset. So don't call mii_phy_reset() in brgphy_reset() and implement the reset locally leaving the problematic bits alone. Effectively this bypasses r221812 for brgphy(4). Thanks to Justin Hibbits for doing a binary search in order to identify the problematic commit. PR: 157405, 158156 Reviewed by: yongari (mii_phy_reset() part) Approved by: re (kib) MFC after: 3 days
2011-08-19 19:12:58 +00:00
/* Wait 100ms for it to complete. */
for (i = 0; i < 100; i++) {
if ((PHY_READ(sc, BRGPHY_MII_BMCR) & BRGPHY_BMCR_RESET) == 0)
break;
DELAY(1000);
}
- Move bge_phy_hack into the phy code and implement the various DSP patch workarounds for each phy revision. Obtained from: NetBSD & Broadcom Linux driver - Disable AUTOPOLL when accessing the PHY as it may cause PCI errors. Obtained from: NetBSD - Check the UPDATED bit in the status block so the driver knows that the status block as indeed changed since the last access. Broadcom documentation states drivers should unset the UPDATED/CHANGED bits after reading them. - When changing media types, first loop the phy then set the media. Broadcom documentation and Linux drivers do this and I observed much better handling of link after this change. - Broadcom documentation states that for 1000BaseT operation, autonegotiation must be enabled. Fix hard coding of media so that the driver only advertises 1000BaseT as the supported media type and enable autonegotition. - Only set Master/Slave on the 5701. Obtained from Broadcom Linux driver.
2003-05-03 19:06:50 +00:00
- Added support for BCM5709 and BCM5716. MFC after: 2 weeks
2008-06-13 01:20:29 +00:00
/* Handle any PHY specific procedures following the reset. */
- Remove attempts to implement setting of BMCR_LOOP/MIIF_NOLOOP (reporting IFM_LOOP based on BMCR_LOOP is left in place though as it might provide useful for debugging). For most mii(4) drivers it was unclear whether the PHYs driven by them actually support loopback or not. Moreover, typically loopback mode also needs to be activated on the MAC, which none of the Ethernet drivers using mii(4) implements. Given that loopback media has no real use (and obviously hardly had a chance to actually work) besides for driver development (which just loopback mode should be sufficient for though, i.e one doesn't necessary need support for loopback media) support for it is just dropped as both NetBSD and OpenBSD already did quite some time ago. - Let mii_phy_add_media() also announce the support of IFM_NONE. - 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 lxtphy(4) which disables MII interrupts (as is done for a few other PHYs we have drivers for). This includes changing NIC drivers which previously just called the generic mii_phy_reset() to now actually call the PHY-specific reset routine, which might be crucial in some cases. While at it, the redundant checks in these NIC drivers for mii->mii_instance not being zero before calling the reset routines were removed because as soon as one PHY driver attaches mii->mii_instance is incremented and we hardly can end up in their media change callbacks etc if no PHY driver has attached as mii_attach() would have failed in that case and not attach a miibus(4) instance. Consequently, NIC drivers now no longer should call mii_phy_reset() directly, so it was removed from EXPORT_SYMS. - Add a mii_phy_dev_attach() as a companion helper to mii_phy_dev_probe(). The purpose of that function is to perform the common steps to attach a PHY driver instance and to hook it up to the miibus(4) instance and to optionally also handle the probing, addition and initialization of the supported media. So all a PHY driver without any special requirements has to do in its bus attach method is to call mii_phy_dev_attach() along with PHY-specific MIIF_* flags, a pointer to its PHY functions and the add_media set to one. All PHY drivers were updated to take advantage of mii_phy_dev_attach() as appropriate. Along with these changes the capability mask was added to the mii_softc structure so PHY drivers taking advantage of mii_phy_dev_attach() but still handling media on their own do not need to fiddle with the MII attach arguments anyway. - Keep track of the PHY offset in the mii_softc structure. This is done for compatibility with NetBSD/OpenBSD. - Keep track of the PHY's OUI, model and revision in the mii_softc structure. Several PHY drivers require this information also after attaching and previously had to wrap their own softc around mii_softc. NetBSD/OpenBSD also keep track of the model and revision on their mii_softc structure. All PHY drivers were updated to take advantage as appropriate. - Convert the mebers of the MII data structure to unsigned where appropriate. This is partly inspired by NetBSD/OpenBSD. - According to IEEE 802.3-2002 the bits actually have to be reversed when mapping an OUI to the MII ID registers. All PHY drivers and miidevs where changed as necessary. Actually this now again allows to largely share miidevs with NetBSD, which fixed this problem already 9 years ago. Consequently miidevs was synced as far as possible. - Add MIIF_NOMANPAUSE and mii_phy_flowstatus() calls to drivers that weren't explicitly converted to support flow control before. It's unclear whether flow control actually works with these but typically it should and their net behavior should be more correct with these changes in place than without if the MAC driver sets MIIF_DOPAUSE. Obtained from: NetBSD (partially) Reviewed by: yongari (earlier version), silence on arch@ and net@
2011-05-03 19:51:29 +00:00
switch (sc->mii_mpd_oui) {
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
case MII_OUI_BROADCOM:
- Remove attempts to implement setting of BMCR_LOOP/MIIF_NOLOOP (reporting IFM_LOOP based on BMCR_LOOP is left in place though as it might provide useful for debugging). For most mii(4) drivers it was unclear whether the PHYs driven by them actually support loopback or not. Moreover, typically loopback mode also needs to be activated on the MAC, which none of the Ethernet drivers using mii(4) implements. Given that loopback media has no real use (and obviously hardly had a chance to actually work) besides for driver development (which just loopback mode should be sufficient for though, i.e one doesn't necessary need support for loopback media) support for it is just dropped as both NetBSD and OpenBSD already did quite some time ago. - Let mii_phy_add_media() also announce the support of IFM_NONE. - 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 lxtphy(4) which disables MII interrupts (as is done for a few other PHYs we have drivers for). This includes changing NIC drivers which previously just called the generic mii_phy_reset() to now actually call the PHY-specific reset routine, which might be crucial in some cases. While at it, the redundant checks in these NIC drivers for mii->mii_instance not being zero before calling the reset routines were removed because as soon as one PHY driver attaches mii->mii_instance is incremented and we hardly can end up in their media change callbacks etc if no PHY driver has attached as mii_attach() would have failed in that case and not attach a miibus(4) instance. Consequently, NIC drivers now no longer should call mii_phy_reset() directly, so it was removed from EXPORT_SYMS. - Add a mii_phy_dev_attach() as a companion helper to mii_phy_dev_probe(). The purpose of that function is to perform the common steps to attach a PHY driver instance and to hook it up to the miibus(4) instance and to optionally also handle the probing, addition and initialization of the supported media. So all a PHY driver without any special requirements has to do in its bus attach method is to call mii_phy_dev_attach() along with PHY-specific MIIF_* flags, a pointer to its PHY functions and the add_media set to one. All PHY drivers were updated to take advantage of mii_phy_dev_attach() as appropriate. Along with these changes the capability mask was added to the mii_softc structure so PHY drivers taking advantage of mii_phy_dev_attach() but still handling media on their own do not need to fiddle with the MII attach arguments anyway. - Keep track of the PHY offset in the mii_softc structure. This is done for compatibility with NetBSD/OpenBSD. - Keep track of the PHY's OUI, model and revision in the mii_softc structure. Several PHY drivers require this information also after attaching and previously had to wrap their own softc around mii_softc. NetBSD/OpenBSD also keep track of the model and revision on their mii_softc structure. All PHY drivers were updated to take advantage as appropriate. - Convert the mebers of the MII data structure to unsigned where appropriate. This is partly inspired by NetBSD/OpenBSD. - According to IEEE 802.3-2002 the bits actually have to be reversed when mapping an OUI to the MII ID registers. All PHY drivers and miidevs where changed as necessary. Actually this now again allows to largely share miidevs with NetBSD, which fixed this problem already 9 years ago. Consequently miidevs was synced as far as possible. - Add MIIF_NOMANPAUSE and mii_phy_flowstatus() calls to drivers that weren't explicitly converted to support flow control before. It's unclear whether flow control actually works with these but typically it should and their net behavior should be more correct with these changes in place than without if the MAC driver sets MIIF_DOPAUSE. Obtained from: NetBSD (partially) Reviewed by: yongari (earlier version), silence on arch@ and net@
2011-05-03 19:51:29 +00:00
switch (sc->mii_mpd_model) {
case MII_MODEL_BROADCOM_BCM5400:
- Fix BCM5754 support found in Dell PowerEdge SC440. - Move some PHY bug detections from brgphy.c to if_bge.c. - Do not penalize working PHYs. - Re-arrange bge_flags roughly by their categories. - Fix minor style(9) nits. PR: kern/107257 Obtained from: OpenBSD Tested by: Mike Hibler <mike at flux dot utah dot edu>
2007-01-15 21:43:43 +00:00
bcm5401_load_dspcode(sc);
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
break;
- Remove attempts to implement setting of BMCR_LOOP/MIIF_NOLOOP (reporting IFM_LOOP based on BMCR_LOOP is left in place though as it might provide useful for debugging). For most mii(4) drivers it was unclear whether the PHYs driven by them actually support loopback or not. Moreover, typically loopback mode also needs to be activated on the MAC, which none of the Ethernet drivers using mii(4) implements. Given that loopback media has no real use (and obviously hardly had a chance to actually work) besides for driver development (which just loopback mode should be sufficient for though, i.e one doesn't necessary need support for loopback media) support for it is just dropped as both NetBSD and OpenBSD already did quite some time ago. - Let mii_phy_add_media() also announce the support of IFM_NONE. - 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 lxtphy(4) which disables MII interrupts (as is done for a few other PHYs we have drivers for). This includes changing NIC drivers which previously just called the generic mii_phy_reset() to now actually call the PHY-specific reset routine, which might be crucial in some cases. While at it, the redundant checks in these NIC drivers for mii->mii_instance not being zero before calling the reset routines were removed because as soon as one PHY driver attaches mii->mii_instance is incremented and we hardly can end up in their media change callbacks etc if no PHY driver has attached as mii_attach() would have failed in that case and not attach a miibus(4) instance. Consequently, NIC drivers now no longer should call mii_phy_reset() directly, so it was removed from EXPORT_SYMS. - Add a mii_phy_dev_attach() as a companion helper to mii_phy_dev_probe(). The purpose of that function is to perform the common steps to attach a PHY driver instance and to hook it up to the miibus(4) instance and to optionally also handle the probing, addition and initialization of the supported media. So all a PHY driver without any special requirements has to do in its bus attach method is to call mii_phy_dev_attach() along with PHY-specific MIIF_* flags, a pointer to its PHY functions and the add_media set to one. All PHY drivers were updated to take advantage of mii_phy_dev_attach() as appropriate. Along with these changes the capability mask was added to the mii_softc structure so PHY drivers taking advantage of mii_phy_dev_attach() but still handling media on their own do not need to fiddle with the MII attach arguments anyway. - Keep track of the PHY offset in the mii_softc structure. This is done for compatibility with NetBSD/OpenBSD. - Keep track of the PHY's OUI, model and revision in the mii_softc structure. Several PHY drivers require this information also after attaching and previously had to wrap their own softc around mii_softc. NetBSD/OpenBSD also keep track of the model and revision on their mii_softc structure. All PHY drivers were updated to take advantage as appropriate. - Convert the mebers of the MII data structure to unsigned where appropriate. This is partly inspired by NetBSD/OpenBSD. - According to IEEE 802.3-2002 the bits actually have to be reversed when mapping an OUI to the MII ID registers. All PHY drivers and miidevs where changed as necessary. Actually this now again allows to largely share miidevs with NetBSD, which fixed this problem already 9 years ago. Consequently miidevs was synced as far as possible. - Add MIIF_NOMANPAUSE and mii_phy_flowstatus() calls to drivers that weren't explicitly converted to support flow control before. It's unclear whether flow control actually works with these but typically it should and their net behavior should be more correct with these changes in place than without if the MAC driver sets MIIF_DOPAUSE. Obtained from: NetBSD (partially) Reviewed by: yongari (earlier version), silence on arch@ and net@
2011-05-03 19:51:29 +00:00
case MII_MODEL_BROADCOM_BCM5401:
if (sc->mii_mpd_rev == 1 || sc->mii_mpd_rev == 3)
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
bcm5401_load_dspcode(sc);
break;
- Remove attempts to implement setting of BMCR_LOOP/MIIF_NOLOOP (reporting IFM_LOOP based on BMCR_LOOP is left in place though as it might provide useful for debugging). For most mii(4) drivers it was unclear whether the PHYs driven by them actually support loopback or not. Moreover, typically loopback mode also needs to be activated on the MAC, which none of the Ethernet drivers using mii(4) implements. Given that loopback media has no real use (and obviously hardly had a chance to actually work) besides for driver development (which just loopback mode should be sufficient for though, i.e one doesn't necessary need support for loopback media) support for it is just dropped as both NetBSD and OpenBSD already did quite some time ago. - Let mii_phy_add_media() also announce the support of IFM_NONE. - 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 lxtphy(4) which disables MII interrupts (as is done for a few other PHYs we have drivers for). This includes changing NIC drivers which previously just called the generic mii_phy_reset() to now actually call the PHY-specific reset routine, which might be crucial in some cases. While at it, the redundant checks in these NIC drivers for mii->mii_instance not being zero before calling the reset routines were removed because as soon as one PHY driver attaches mii->mii_instance is incremented and we hardly can end up in their media change callbacks etc if no PHY driver has attached as mii_attach() would have failed in that case and not attach a miibus(4) instance. Consequently, NIC drivers now no longer should call mii_phy_reset() directly, so it was removed from EXPORT_SYMS. - Add a mii_phy_dev_attach() as a companion helper to mii_phy_dev_probe(). The purpose of that function is to perform the common steps to attach a PHY driver instance and to hook it up to the miibus(4) instance and to optionally also handle the probing, addition and initialization of the supported media. So all a PHY driver without any special requirements has to do in its bus attach method is to call mii_phy_dev_attach() along with PHY-specific MIIF_* flags, a pointer to its PHY functions and the add_media set to one. All PHY drivers were updated to take advantage of mii_phy_dev_attach() as appropriate. Along with these changes the capability mask was added to the mii_softc structure so PHY drivers taking advantage of mii_phy_dev_attach() but still handling media on their own do not need to fiddle with the MII attach arguments anyway. - Keep track of the PHY offset in the mii_softc structure. This is done for compatibility with NetBSD/OpenBSD. - Keep track of the PHY's OUI, model and revision in the mii_softc structure. Several PHY drivers require this information also after attaching and previously had to wrap their own softc around mii_softc. NetBSD/OpenBSD also keep track of the model and revision on their mii_softc structure. All PHY drivers were updated to take advantage as appropriate. - Convert the mebers of the MII data structure to unsigned where appropriate. This is partly inspired by NetBSD/OpenBSD. - According to IEEE 802.3-2002 the bits actually have to be reversed when mapping an OUI to the MII ID registers. All PHY drivers and miidevs where changed as necessary. Actually this now again allows to largely share miidevs with NetBSD, which fixed this problem already 9 years ago. Consequently miidevs was synced as far as possible. - Add MIIF_NOMANPAUSE and mii_phy_flowstatus() calls to drivers that weren't explicitly converted to support flow control before. It's unclear whether flow control actually works with these but typically it should and their net behavior should be more correct with these changes in place than without if the MAC driver sets MIIF_DOPAUSE. Obtained from: NetBSD (partially) Reviewed by: yongari (earlier version), silence on arch@ and net@
2011-05-03 19:51:29 +00:00
case MII_MODEL_BROADCOM_BCM5411:
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
bcm5411_load_dspcode(sc);
break;
- Remove attempts to implement setting of BMCR_LOOP/MIIF_NOLOOP (reporting IFM_LOOP based on BMCR_LOOP is left in place though as it might provide useful for debugging). For most mii(4) drivers it was unclear whether the PHYs driven by them actually support loopback or not. Moreover, typically loopback mode also needs to be activated on the MAC, which none of the Ethernet drivers using mii(4) implements. Given that loopback media has no real use (and obviously hardly had a chance to actually work) besides for driver development (which just loopback mode should be sufficient for though, i.e one doesn't necessary need support for loopback media) support for it is just dropped as both NetBSD and OpenBSD already did quite some time ago. - Let mii_phy_add_media() also announce the support of IFM_NONE. - 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 lxtphy(4) which disables MII interrupts (as is done for a few other PHYs we have drivers for). This includes changing NIC drivers which previously just called the generic mii_phy_reset() to now actually call the PHY-specific reset routine, which might be crucial in some cases. While at it, the redundant checks in these NIC drivers for mii->mii_instance not being zero before calling the reset routines were removed because as soon as one PHY driver attaches mii->mii_instance is incremented and we hardly can end up in their media change callbacks etc if no PHY driver has attached as mii_attach() would have failed in that case and not attach a miibus(4) instance. Consequently, NIC drivers now no longer should call mii_phy_reset() directly, so it was removed from EXPORT_SYMS. - Add a mii_phy_dev_attach() as a companion helper to mii_phy_dev_probe(). The purpose of that function is to perform the common steps to attach a PHY driver instance and to hook it up to the miibus(4) instance and to optionally also handle the probing, addition and initialization of the supported media. So all a PHY driver without any special requirements has to do in its bus attach method is to call mii_phy_dev_attach() along with PHY-specific MIIF_* flags, a pointer to its PHY functions and the add_media set to one. All PHY drivers were updated to take advantage of mii_phy_dev_attach() as appropriate. Along with these changes the capability mask was added to the mii_softc structure so PHY drivers taking advantage of mii_phy_dev_attach() but still handling media on their own do not need to fiddle with the MII attach arguments anyway. - Keep track of the PHY offset in the mii_softc structure. This is done for compatibility with NetBSD/OpenBSD. - Keep track of the PHY's OUI, model and revision in the mii_softc structure. Several PHY drivers require this information also after attaching and previously had to wrap their own softc around mii_softc. NetBSD/OpenBSD also keep track of the model and revision on their mii_softc structure. All PHY drivers were updated to take advantage as appropriate. - Convert the mebers of the MII data structure to unsigned where appropriate. This is partly inspired by NetBSD/OpenBSD. - According to IEEE 802.3-2002 the bits actually have to be reversed when mapping an OUI to the MII ID registers. All PHY drivers and miidevs where changed as necessary. Actually this now again allows to largely share miidevs with NetBSD, which fixed this problem already 9 years ago. Consequently miidevs was synced as far as possible. - Add MIIF_NOMANPAUSE and mii_phy_flowstatus() calls to drivers that weren't explicitly converted to support flow control before. It's unclear whether flow control actually works with these but typically it should and their net behavior should be more correct with these changes in place than without if the MAC driver sets MIIF_DOPAUSE. Obtained from: NetBSD (partially) Reviewed by: yongari (earlier version), silence on arch@ and net@
2011-05-03 19:51:29 +00:00
case MII_MODEL_BROADCOM_BCM54K2:
Add support for BCM54K2 found in combination with Apple K2 GMAC. Submitted by: Andreas Tobler Obtained from: OpenBSD MFC after: 1 week
2010-02-20 22:01:24 +00:00
bcm54k2_load_dspcode(sc);
break;
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
}
- Fix BCM5754 support found in Dell PowerEdge SC440. - Move some PHY bug detections from brgphy.c to if_bge.c. - Do not penalize working PHYs. - Re-arrange bge_flags roughly by their categories. - Fix minor style(9) nits. PR: kern/107257 Obtained from: OpenBSD Tested by: Mike Hibler <mike at flux dot utah dot edu>
2007-01-15 21:43:43 +00:00
break;
For 5717C/5719C/5720C and 57765 PHYs, do not perform any special handling(jumbo, wire speed etc) in brgphy_reset(). Touching BRGPHY_MII_AUXCTL register seems to confuse APE firmware such that it couldn't establish a link.
2012-10-11 06:07:48 +00:00
case MII_OUI_BROADCOM3:
switch (sc->mii_mpd_model) {
case MII_MODEL_BROADCOM3_BCM5717C:
case MII_MODEL_BROADCOM3_BCM5719C:
case MII_MODEL_BROADCOM3_BCM5720C:
case MII_MODEL_BROADCOM3_BCM57765:
return;
}
break;
Recognize BCM5725C PHY.
2013-07-20 07:24:01 +00:00
case MII_OUI_BROADCOM4:
return;
- Move bge_phy_hack into the phy code and implement the various DSP patch workarounds for each phy revision. Obtained from: NetBSD & Broadcom Linux driver - Disable AUTOPOLL when accessing the PHY as it may cause PCI errors. Obtained from: NetBSD - Check the UPDATED bit in the status block so the driver knows that the status block as indeed changed since the last access. Broadcom documentation states drivers should unset the UPDATED/CHANGED bits after reading them. - When changing media types, first loop the phy then set the media. Broadcom documentation and Linux drivers do this and I observed much better handling of link after this change. - Broadcom documentation states that for 1000BaseT operation, autonegotiation must be enabled. Fix hard coding of media so that the driver only advertises 1000BaseT as the supported media type and enable autonegotition. - Only set Master/Slave on the 5701. Obtained from Broadcom Linux driver.
2003-05-03 19:06:50 +00:00
}
Add support for the BCM5705 and its ilk. Changes: - 5705 doesn't support jumbo frames - Statistics must be read from registers - RX return ring must be capped at 512 entries - Omit initialization of certain device blocks - Acknowledge link change interrupts by setting the 'link changed' bit in the status register (used to have no effect) - Remember to toggle the MI completion bit too - Set the mbuf low watermark differently (on-chip memory buffers, not BSD mbufs) - Don't enable Ethernet@WireSpeed feature for certain 5705 chip revs - Add additional PCI IDs for 5705 and 5782 parts - Add a forgotten 5704 PCI ID Most changes ripped kicking and screaming from the Broadcom linux driver. Thanks to Paul Saab for sanity testing. (My lack of sanity has been confirmed.)
2003-07-16 00:09:56 +00:00
ifp = sc->mii_pdata->mii_ifp;
Add a driver for the Broadcom NetXtreme II (BCM5706/BCM5708) PCI/PCIe Gigabit Ethernet adapeter. Submitted by: David Christensen
2006-04-10 19:55:23 +00:00
/* Find the driver associated with this PHY. */
if (strcmp(ifp->if_dname, "bge") == 0) {
Use BMSR for link status in one more place and clean up more.
2006-12-20 00:34:12 +00:00
bge_sc = ifp->if_softc;
Add a driver for the Broadcom NetXtreme II (BCM5706/BCM5708) PCI/PCIe Gigabit Ethernet adapeter. Submitted by: David Christensen
2006-04-10 19:55:23 +00:00
} else if (strcmp(ifp->if_dname, "bce") == 0) {
bce_sc = ifp->if_softc;
}
if (bge_sc) {
- Fix BCM5754 support found in Dell PowerEdge SC440. - Move some PHY bug detections from brgphy.c to if_bge.c. - Do not penalize working PHYs. - Re-arrange bge_flags roughly by their categories. - Fix minor style(9) nits. PR: kern/107257 Obtained from: OpenBSD Tested by: Mike Hibler <mike at flux dot utah dot edu>
2007-01-15 21:43:43 +00:00
/* Fix up various bugs */
Separate common flags into controller specific and PHY related flags. There should be no functional changes. This change will make it easy to add more quirk/flags in future. Reviewed by: davidch
2010-10-05 23:03:48 +00:00
if (bge_sc->bge_phy_flags & BGE_PHY_5704_A0_BUG)
- Fix BCM5754 support found in Dell PowerEdge SC440. - Move some PHY bug detections from brgphy.c to if_bge.c. - Do not penalize working PHYs. - Re-arrange bge_flags roughly by their categories. - Fix minor style(9) nits. PR: kern/107257 Obtained from: OpenBSD Tested by: Mike Hibler <mike at flux dot utah dot edu>
2007-01-15 21:43:43 +00:00
brgphy_fixup_5704_a0_bug(sc);
Separate common flags into controller specific and PHY related flags. There should be no functional changes. This change will make it easy to add more quirk/flags in future. Reviewed by: davidch
2010-10-05 23:03:48 +00:00
if (bge_sc->bge_phy_flags & BGE_PHY_ADC_BUG)
Fix style(9). Pointed out by: many
2007-02-12 23:33:05 +00:00
brgphy_fixup_adc_bug(sc);
Separate common flags into controller specific and PHY related flags. There should be no functional changes. This change will make it easy to add more quirk/flags in future. Reviewed by: davidch
2010-10-05 23:03:48 +00:00
if (bge_sc->bge_phy_flags & BGE_PHY_ADJUST_TRIM)
Add PHY DSP code for BCM5755M. Obtained from: OpenBSD
2007-02-12 22:51:25 +00:00
brgphy_fixup_adjust_trim(sc);
Separate common flags into controller specific and PHY related flags. There should be no functional changes. This change will make it easy to add more quirk/flags in future. Reviewed by: davidch
2010-10-05 23:03:48 +00:00
if (bge_sc->bge_phy_flags & BGE_PHY_BER_BUG)
- Fix BCM5754 support found in Dell PowerEdge SC440. - Move some PHY bug detections from brgphy.c to if_bge.c. - Do not penalize working PHYs. - Re-arrange bge_flags roughly by their categories. - Fix minor style(9) nits. PR: kern/107257 Obtained from: OpenBSD Tested by: Mike Hibler <mike at flux dot utah dot edu>
2007-01-15 21:43:43 +00:00
brgphy_fixup_ber_bug(sc);
Separate common flags into controller specific and PHY related flags. There should be no functional changes. This change will make it easy to add more quirk/flags in future. Reviewed by: davidch
2010-10-05 23:03:48 +00:00
if (bge_sc->bge_phy_flags & BGE_PHY_CRC_BUG)
Add BCM5701 A0/B0 CRC bug workaround. Magic values taken from Linux driver.
2007-02-12 23:58:52 +00:00
brgphy_fixup_crc_bug(sc);
Separate common flags into controller specific and PHY related flags. There should be no functional changes. This change will make it easy to add more quirk/flags in future. Reviewed by: davidch
2010-10-05 23:03:48 +00:00
if (bge_sc->bge_phy_flags & BGE_PHY_JITTER_BUG)
- Fix BCM5754 support found in Dell PowerEdge SC440. - Move some PHY bug detections from brgphy.c to if_bge.c. - Do not penalize working PHYs. - Re-arrange bge_flags roughly by their categories. - Fix minor style(9) nits. PR: kern/107257 Obtained from: OpenBSD Tested by: Mike Hibler <mike at flux dot utah dot edu>
2007-01-15 21:43:43 +00:00
brgphy_fixup_jitter_bug(sc);
- Move Ethernet@WireSpeed and jumbo frame configurations to separate functions. The idea is taken from OpenBSD. - Set/clear jumbo frame configurations for bge(4). - Re-add BCM5750 PHY workaround for bce(4), which was mistakenly removed from the previous commit.
2007-01-15 22:21:44 +00:00
- Probe BCM57780. - In case the parent is bge(4), don't set the Jumbo frame settings unless the MAC actually is Jumbo capable as otherwise the PHY might not have the corresponding registers implemented. This is also in line with what the Linux tg3 driver does. PR: 165032 Submitted by: Alexander Milanov Obtained from: OpenBSD MFC after: 3 days
2012-02-19 12:09:17 +00:00
if (bge_sc->bge_flags & BGE_FLAG_JUMBO)
brgphy_jumbo_settings(sc, ifp->if_mtu);
- Move Ethernet@WireSpeed and jumbo frame configurations to separate functions. The idea is taken from OpenBSD. - Set/clear jumbo frame configurations for bge(4). - Re-add BCM5750 PHY workaround for bce(4), which was mistakenly removed from the previous commit.
2007-01-15 22:21:44 +00:00
Enable Ethernet@WireSpeed for BCM5718/BCM57765 family. While I'm here inverse meaning of PHY flag as Ethernet@WireSpeed is enabled for most PHYs.
2011-05-05 00:43:40 +00:00
if ((bge_sc->bge_phy_flags & BGE_PHY_NO_WIRESPEED) == 0)
- Move Ethernet@WireSpeed and jumbo frame configurations to separate functions. The idea is taken from OpenBSD. - Set/clear jumbo frame configurations for bge(4). - Re-add BCM5750 PHY workaround for bce(4), which was mistakenly removed from the previous commit.
2007-01-15 22:21:44 +00:00
brgphy_ethernet_wirespeed(sc);
/* Enable Link LED on Dell boxes */
Separate common flags into controller specific and PHY related flags. There should be no functional changes. This change will make it easy to add more quirk/flags in future. Reviewed by: davidch
2010-10-05 23:03:48 +00:00
if (bge_sc->bge_phy_flags & BGE_PHY_NO_3LED) {
Some style changes to a couple of PHY drivers: - Fix some whitespace nits. - Fix some spelling in comments. - Use MII_ANEGTICKS instead of 5. - Don't define variables in nested scope. - Remove superfluous returns at the end of void functions. - Remove unused static global rgephy_mii_model. - Remove dupe $Id$ in tdkphy(4). - Sort brgphys table. MFC after: 2 weeks
2006-12-02 19:36:25 +00:00
PHY_WRITE(sc, BRGPHY_MII_PHY_EXTCTL,
- Move Ethernet@WireSpeed and jumbo frame configurations to separate functions. The idea is taken from OpenBSD. - Set/clear jumbo frame configurations for bge(4). - Re-add BCM5750 PHY workaround for bce(4), which was mistakenly removed from the previous commit.
2007-01-15 22:21:44 +00:00
PHY_READ(sc, BRGPHY_MII_PHY_EXTCTL) &
~BRGPHY_PHY_EXTCTL_3_LED);
Add a driver for the Broadcom NetXtreme II (BCM5706/BCM5708) PCI/PCIe Gigabit Ethernet adapeter. Submitted by: David Christensen
2006-04-10 19:55:23 +00:00
}
Add support for the BCM5906[M] adapters. These adapters only support 10/100 operation and place the mailbox registers at a different offset. They also do not have an EEPROM, so the MAC address must be read from NVRAM instead. MFC after: 1 month PR: kern/118975 Submitted by: benjsc, Thomas Nyström thn at saeab dot se Submitted by: sephe (original patch for DragonflyBSD)
2008-04-29 19:47:13 +00:00
/* Adjust output voltage (From Linux driver) */
if (bge_sc->bge_asicrev == BGE_ASICREV_BCM5906)
PHY_WRITE(sc, BRGPHY_MII_EPHY_PTEST, 0x12);
- Move Ethernet@WireSpeed and jumbo frame configurations to separate functions. The idea is taken from OpenBSD. - Set/clear jumbo frame configurations for bge(4). - Re-add BCM5750 PHY workaround for bce(4), which was mistakenly removed from the previous commit.
2007-01-15 22:21:44 +00:00
} else if (bce_sc) {
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
if (BCE_CHIP_NUM(bce_sc) == BCE_CHIP_NUM_5708 &&
- Add PHY ID for BCM5709C 1000Base-T controllers. MFC after: 1 week
2008-03-05 22:58:02 +00:00
(bce_sc->bce_phy_flags & BCE_PHY_SERDES_FLAG)) {
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
/* Store autoneg capabilities/results in digital block (Page 0) */
PHY_WRITE(sc, BRGPHY_5708S_BLOCK_ADDR, BRGPHY_5708S_DIG3_PG2);
Remove 'cr' at the end of line.
2008-08-12 00:55:03 +00:00
PHY_WRITE(sc, BRGPHY_5708S_PG2_DIGCTL_3_0,
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
BRGPHY_5708S_PG2_DIGCTL_3_0_USE_IEEE);
PHY_WRITE(sc, BRGPHY_5708S_BLOCK_ADDR, BRGPHY_5708S_DIG_PG0);
/* Enable fiber mode and autodetection */
Remove 'cr' at the end of line.
2008-08-12 00:55:03 +00:00
PHY_WRITE(sc, BRGPHY_5708S_PG0_1000X_CTL1,
PHY_READ(sc, BRGPHY_5708S_PG0_1000X_CTL1) |
BRGPHY_5708S_PG0_1000X_CTL1_AUTODET_EN |
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
BRGPHY_5708S_PG0_1000X_CTL1_FIBER_MODE);
/* Enable parallel detection */
Remove 'cr' at the end of line.
2008-08-12 00:55:03 +00:00
PHY_WRITE(sc, BRGPHY_5708S_PG0_1000X_CTL2,
PHY_READ(sc, BRGPHY_5708S_PG0_1000X_CTL2) |
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
BRGPHY_5708S_PG0_1000X_CTL2_PAR_DET_EN);
/* Advertise 2.5G support through next page during autoneg */
if (bce_sc->bce_phy_flags & BCE_PHY_2_5G_CAPABLE_FLAG)
Remove 'cr' at the end of line.
2008-08-12 00:55:03 +00:00
PHY_WRITE(sc, BRGPHY_5708S_ANEG_NXT_PG_XMIT1,
PHY_READ(sc, BRGPHY_5708S_ANEG_NXT_PG_XMIT1) |
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
BRGPHY_5708S_ANEG_NXT_PG_XMIT1_25G);
/* Increase TX signal amplitude */
if ((BCE_CHIP_ID(bce_sc) == BCE_CHIP_ID_5708_A0) ||
(BCE_CHIP_ID(bce_sc) == BCE_CHIP_ID_5708_B0) ||
(BCE_CHIP_ID(bce_sc) == BCE_CHIP_ID_5708_B1)) {
Remove 'cr' at the end of line.
2008-08-12 00:55:03 +00:00
PHY_WRITE(sc, BRGPHY_5708S_BLOCK_ADDR,
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
BRGPHY_5708S_TX_MISC_PG5);
Remove 'cr' at the end of line.
2008-08-12 00:55:03 +00:00
PHY_WRITE(sc, BRGPHY_5708S_PG5_TXACTL1,
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
PHY_READ(sc, BRGPHY_5708S_PG5_TXACTL1) & ~0x30);
Remove 'cr' at the end of line.
2008-08-12 00:55:03 +00:00
PHY_WRITE(sc, BRGPHY_5708S_BLOCK_ADDR,
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
BRGPHY_5708S_DIG_PG0);
}
Remove 'cr' at the end of line.
2008-08-12 00:55:03 +00:00
/* Backplanes use special driver/pre-driver/pre-emphasis values. */
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
if ((bce_sc->bce_shared_hw_cfg & BCE_SHARED_HW_CFG_PHY_BACKPLANE) &&
(bce_sc->bce_port_hw_cfg & BCE_PORT_HW_CFG_CFG_TXCTL3_MASK)) {
Remove 'cr' at the end of line.
2008-08-12 00:55:03 +00:00
PHY_WRITE(sc, BRGPHY_5708S_BLOCK_ADDR,
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
BRGPHY_5708S_TX_MISC_PG5);
Remove 'cr' at the end of line.
2008-08-12 00:55:03 +00:00
PHY_WRITE(sc, BRGPHY_5708S_PG5_TXACTL3,
bce_sc->bce_port_hw_cfg &
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
BCE_PORT_HW_CFG_CFG_TXCTL3_MASK);
Remove 'cr' at the end of line.
2008-08-12 00:55:03 +00:00
PHY_WRITE(sc, BRGPHY_5708S_BLOCK_ADDR,
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
BRGPHY_5708S_DIG_PG0);
}
- Added support for 5709S/5716S PHYs. Submitted by: pyunyh MFC after: 2 weeks
2010-03-18 20:57:57 +00:00
} else if (BCE_CHIP_NUM(bce_sc) == BCE_CHIP_NUM_5709 &&
(bce_sc->bce_phy_flags & BCE_PHY_SERDES_FLAG)) {
Consistently use tab characters instead of tab + space characters. No functional changes.
2010-09-07 23:08:38 +00:00
/* Select the SerDes Digital block of the AN MMD. */
PHY_WRITE(sc, BRGPHY_BLOCK_ADDR, BRGPHY_BLOCK_ADDR_SERDES_DIG);
- Added support for 5709S/5716S PHYs. Submitted by: pyunyh MFC after: 2 weeks
2010-03-18 20:57:57 +00:00
val = PHY_READ(sc, BRGPHY_SERDES_DIG_1000X_CTL1);
val &= ~BRGPHY_SD_DIG_1000X_CTL1_AUTODET;
val |= BRGPHY_SD_DIG_1000X_CTL1_FIBER;
PHY_WRITE(sc, BRGPHY_SERDES_DIG_1000X_CTL1, val);
Consistently use tab characters instead of tab + space characters. No functional changes.
2010-09-07 23:08:38 +00:00
/* Select the Over 1G block of the AN MMD. */
- Added support for 5709S/5716S PHYs. Submitted by: pyunyh MFC after: 2 weeks
2010-03-18 20:57:57 +00:00
PHY_WRITE(sc, BRGPHY_BLOCK_ADDR, BRGPHY_BLOCK_ADDR_OVER_1G);
Consistently use tab characters instead of tab + space characters. No functional changes.
2010-09-07 23:08:38 +00:00
/* Enable autoneg "Next Page" to advertise 2.5G support. */
val = PHY_READ(sc, BRGPHY_OVER_1G_UNFORMAT_PG1);
- Added support for 5709S/5716S PHYs. Submitted by: pyunyh MFC after: 2 weeks
2010-03-18 20:57:57 +00:00
if (bce_sc->bce_phy_flags & BCE_PHY_2_5G_CAPABLE_FLAG)
val |= BRGPHY_5708S_ANEG_NXT_PG_XMIT1_25G;
else
val &= ~BRGPHY_5708S_ANEG_NXT_PG_XMIT1_25G;
PHY_WRITE(sc, BRGPHY_OVER_1G_UNFORMAT_PG1, val);
Consistently use tab characters instead of tab + space characters. No functional changes.
2010-09-07 23:08:38 +00:00
/* Select the Multi-Rate Backplane Ethernet block of the AN MMD. */
- Added support for 5709S/5716S PHYs. Submitted by: pyunyh MFC after: 2 weeks
2010-03-18 20:57:57 +00:00
PHY_WRITE(sc, BRGPHY_BLOCK_ADDR, BRGPHY_BLOCK_ADDR_MRBE);
Consistently use tab characters instead of tab + space characters. No functional changes.
2010-09-07 23:08:38 +00:00
/* Enable MRBE speed autoneg. */
val = PHY_READ(sc, BRGPHY_MRBE_MSG_PG5_NP);
- Added support for 5709S/5716S PHYs. Submitted by: pyunyh MFC after: 2 weeks
2010-03-18 20:57:57 +00:00
val |= BRGPHY_MRBE_MSG_PG5_NP_MBRE |
BRGPHY_MRBE_MSG_PG5_NP_T2;
PHY_WRITE(sc, BRGPHY_MRBE_MSG_PG5_NP, val);
Consistently use tab characters instead of tab + space characters. No functional changes.
2010-09-07 23:08:38 +00:00
/* Select the Clause 73 User B0 block of the AN MMD. */
PHY_WRITE(sc, BRGPHY_BLOCK_ADDR, BRGPHY_BLOCK_ADDR_CL73_USER_B0);
- Added support for 5709S/5716S PHYs. Submitted by: pyunyh MFC after: 2 weeks
2010-03-18 20:57:57 +00:00
Consistently use tab characters instead of tab + space characters. No functional changes.
2010-09-07 23:08:38 +00:00
/* Enable MRBE speed autoneg. */
- Added support for 5709S/5716S PHYs. Submitted by: pyunyh MFC after: 2 weeks
2010-03-18 20:57:57 +00:00
PHY_WRITE(sc, BRGPHY_CL73_USER_B0_MBRE_CTL1,
BRGPHY_CL73_USER_B0_MBRE_CTL1_NP_AFT_BP |
BRGPHY_CL73_USER_B0_MBRE_CTL1_STA_MGR |
BRGPHY_CL73_USER_B0_MBRE_CTL1_ANEG);
Consistently use tab characters instead of tab + space characters. No functional changes.
2010-09-07 23:08:38 +00:00
/* Restore IEEE0 block (assumed in all brgphy(4) code). */
PHY_WRITE(sc, BRGPHY_BLOCK_ADDR, BRGPHY_BLOCK_ADDR_COMBO_IEEE0);
- Added support for 5709S/5716S PHYs. Submitted by: pyunyh MFC after: 2 weeks
2010-03-18 20:57:57 +00:00
} else if (BCE_CHIP_NUM(bce_sc) == BCE_CHIP_NUM_5709) {
- Added support for BCM5709 and BCM5716. MFC after: 2 weeks
2008-06-13 01:20:29 +00:00
if ((BCE_CHIP_REV(bce_sc) == BCE_CHIP_REV_Ax) ||
(BCE_CHIP_REV(bce_sc) == BCE_CHIP_REV_Bx))
brgphy_fixup_disable_early_dac(sc);
Consistently use tab characters instead of tab + space characters. No functional changes.
2010-09-07 23:08:38 +00:00
- Added support for BCM5709 and BCM5716. MFC after: 2 weeks
2008-06-13 01:20:29 +00:00
brgphy_jumbo_settings(sc, ifp->if_mtu);
brgphy_ethernet_wirespeed(sc);
New features: - Moved BCM5706S/5708S SerDes support to brgphy (since they are not technically TBI interfaces) - Added 2.5G support for BCM5708S Comments: Since this driver is shared with bge I tested several available controllers supported by bge and all worked as expected, however the list was not exhaustive. Need wider testing. MFC after: 4 weeks
2007-06-07 02:21:38 +00:00
} else {
brgphy_fixup_ber_bug(sc);
brgphy_jumbo_settings(sc, ifp->if_mtu);
brgphy_ethernet_wirespeed(sc);
}
On Dell boxes such as the PE2650 we need to disable 3 led mode for the link LED to work. Reviewed by: wpaul Obtained from: Linux driver
2003-08-20 04:06:00 +00:00
}
- Move bge_phy_hack into the phy code and implement the various DSP patch workarounds for each phy revision. Obtained from: NetBSD & Broadcom Linux driver - Disable AUTOPOLL when accessing the PHY as it may cause PCI errors. Obtained from: NetBSD - Check the UPDATED bit in the status block so the driver knows that the status block as indeed changed since the last access. Broadcom documentation states drivers should unset the UPDATED/CHANGED bits after reading them. - When changing media types, first loop the phy then set the media. Broadcom documentation and Linux drivers do this and I observed much better handling of link after this change. - Broadcom documentation states that for 1000BaseT operation, autonegotiation must be enabled. Fix hard coding of media so that the driver only advertises 1000BaseT as the supported media type and enable autonegotition. - Only set Master/Slave on the 5701. Obtained from Broadcom Linux driver.
2003-05-03 19:06:50 +00:00
}
Reference in New Issue Copy Permalink