freebsd-dev/sys/dev/mii/brgphy.c

1094 lines
31 KiB
C
Raw Normal View History

/*-
* 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.
*/
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
/*
2007-02-13 00:34:32 +00:00
* Driver for the Broadcom BCM54xx/57xx 1000baseTX PHY.
*/
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/kernel.h>
#include <sys/module.h>
#include <sys/socket.h>
#include <sys/bus.h>
Introduce a procedural interface to the ifnet structure. The new interface allows the ifnet structure to be defined as an opaque type in NIC drivers. This then allows the ifnet structure to be changed without a need to change or recompile NIC drivers. Put differently, NIC drivers can be written and compiled once and be used with different network stack implementations, provided of course that those network stack implementations have an API and ABI compatible interface. This commit introduces the 'if_t' type to replace 'struct ifnet *' as the type of a network interface. The 'if_t' type is defined as 'void *' to enable the compiler to perform type conversion to 'struct ifnet *' and vice versa where needed and without warnings. The functions that implement the API are the only functions that need to have an explicit cast. The MII code has been converted to use the driver API to avoid unnecessary code churn. Code churn comes from having to work with both converted and unconverted drivers in correlation with having callback functions that take an interface. By converting the MII code first, the callback functions can be defined so that the compiler will perform the typecasts automatically. As soon as all drivers have been converted, the if_t type can be redefined as needed and the API functions can be fix to not need an explicit cast. The immediate benefactors of this change are: 1. Juniper Networks - The network stack implementation in Junos is entirely different from FreeBSD's one and this change allows Juniper to build "stock" NIC drivers that can be used in combination with both the FreeBSD and Junos stacks. 2. FreeBSD - This change opens the door towards changing ifnet and implementing new features and optimizations in the network stack without it requiring a change in the many NIC drivers FreeBSD has. Submitted by: Anuranjan Shukla <anshukla@juniper.net> Reviewed by: glebius@ Obtained from: Juniper Networks, Inc.
2014-06-02 17:54:39 +00:00
#include <sys/taskqueue.h>
#include <net/if.h>
#include <net/if_var.h>
#include <net/ethernet.h>
#include <net/if_media.h>
#include <dev/mii/mii.h>
#include <dev/mii/miivar.h>
#include "miidevs.h"
#include <dev/mii/brgphyreg.h>
#include <net/if_arp.h>
#include <machine/bus.h>
#include <dev/bge/if_bgereg.h>
#include <dev/bce/if_bcereg.h>
#include <dev/pci/pcireg.h>
#include <dev/pci/pcivar.h>
#include "miibus_if.h"
static int brgphy_probe(device_t);
static int brgphy_attach(device_t);
struct brgphy_softc {
struct mii_softc mii_sc;
int serdes_flags; /* Keeps track of the serdes type used */
#define BRGPHY_5706S 0x0001
#define BRGPHY_5708S 0x0002
#define BRGPHY_NOANWAIT 0x0004
#define BRGPHY_5709S 0x0008
int bce_phy_flags; /* PHY flags transferred from the MAC driver */
};
static device_method_t brgphy_methods[] = {
/* device interface */
DEVMETHOD(device_probe, brgphy_probe),
DEVMETHOD(device_attach, brgphy_attach),
DEVMETHOD(device_detach, mii_phy_detach),
DEVMETHOD(device_shutdown, bus_generic_shutdown),
2011-11-23 20:27:26 +00:00
DEVMETHOD_END
};
static devclass_t brgphy_devclass;
static driver_t brgphy_driver = {
"brgphy",
brgphy_methods,
sizeof(struct brgphy_softc)
};
DRIVER_MODULE(brgphy, miibus, brgphy_driver, brgphy_devclass, 0, 0);
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);
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);
static void brgphy_reset(struct mii_softc *);
static void brgphy_enable_loopback(struct mii_softc *);
static void bcm5401_load_dspcode(struct mii_softc *);
static void bcm5411_load_dspcode(struct mii_softc *);
static void bcm54k2_load_dspcode(struct mii_softc *);
static void brgphy_fixup_5704_a0_bug(struct mii_softc *);
2007-02-12 23:33:05 +00:00
static void brgphy_fixup_adc_bug(struct mii_softc *);
static void brgphy_fixup_adjust_trim(struct mii_softc *);
static void brgphy_fixup_ber_bug(struct mii_softc *);
static void brgphy_fixup_crc_bug(struct mii_softc *);
static void brgphy_fixup_jitter_bug(struct mii_softc *);
static void brgphy_ethernet_wirespeed(struct mii_softc *);
static void brgphy_jumbo_settings(struct mii_softc *, u_long);
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, BCM5402),
- 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, BCM5411),
MII_PHY_DESC(BROADCOM, BCM5404),
MII_PHY_DESC(BROADCOM, BCM5424),
- 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, 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(BROADCOM, BCM5466),
MII_PHY_DESC(BROADCOM2, BCM5478),
MII_PHY_DESC(BROADCOM2, BCM5488),
- 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(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),
#ifdef notyet /* better handled by ukphy(4) until WARs are implemented */
MII_PHY_DESC(BROADCOM2, BCM5785),
#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),
MII_PHY_DESC(BROADCOM3, BCM5719C),
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),
MII_PHY_DESC(BROADCOM3, BCM57780),
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),
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
};
static const struct hs21_type {
const uint32_t id;
const char *prod;
} hs21_type_lists[] = {
{ 0x57081021, "IBM eServer BladeCenter HS21" },
{ 0x57081011, "IBM eServer BladeCenter HS21 -[8853PAU]-" },
};
static int
detect_hs21(struct bce_softc *bce_sc)
{
char *sysenv;
int found, i;
found = 0;
sysenv = kern_getenv("smbios.system.product");
if (sysenv == NULL)
return (found);
for (i = 0; i < nitems(hs21_type_lists); i++) {
if (bce_sc->bce_chipid == hs21_type_lists[i].id &&
strncmp(sysenv, hs21_type_lists[i].prod,
strlen(hs21_type_lists[i].prod)) == 0) {
found++;
break;
}
}
freeenv(sysenv);
return (found);
}
/* Search for our PHY in the list of known PHYs */
static int
brgphy_probe(device_t dev)
{
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 (mii_phy_dev_probe(dev, brgphys, BUS_PROBE_DEFAULT));
}
/* Attach the PHY to the MII bus */
static int
brgphy_attach(device_t dev)
{
struct brgphy_softc *bsc;
struct bge_softc *bge_sc = NULL;
struct bce_softc *bce_sc = NULL;
struct mii_softc *sc;
bsc = device_get_softc(dev);
sc = &bsc->mii_sc;
- 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);
bsc->serdes_flags = 0;
/* Find the MAC driver associated with this PHY. */
if (mii_dev_mac_match(dev, "bge"))
bge_sc = mii_dev_mac_softc(dev);
else if (mii_dev_mac_match(dev, "bce"))
bce_sc = mii_dev_mac_softc(dev);
/* 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) {
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:
/*
* 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;
}
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:
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:
/*
* 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;
sc->mii_flags |= MIIF_HAVEFIBER;
break;
}
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
PHY_RESET(sc);
/* 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;
if (sc->mii_capabilities & BMSR_EXTSTAT)
sc->mii_extcapabilities = PHY_READ(sc, MII_EXTSR);
device_printf(dev, " ");
/* 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");
} 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;
ifmedia_add(&sc->mii_pdata->mii_media,
IFM_MAKEWORD(IFM_ETHER, IFM_1000_SX, IFM_FDX, sc->mii_inst),
0, NULL);
printf("1000baseSX-FDX, ");
/*
* 2.5G support is a software enabled feature
* on the 5708S and 5709S.
*/
if (bce_sc && (bce_sc->bce_phy_flags &
BCE_PHY_2_5G_CAPABLE_FLAG)) {
ifmedia_add(&sc->mii_pdata->mii_media,
IFM_MAKEWORD(IFM_ETHER, IFM_2500_SX, IFM_FDX,
sc->mii_inst), 0, NULL);
printf("2500baseSX-FDX, ");
} else if ((bsc->serdes_flags & BRGPHY_5708S) && bce_sc &&
(detect_hs21(bce_sc) != 0)) {
/*
* There appears to be certain silicon revision
* in IBM HS21 blades that is having issues with
* 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;
}
ifmedia_add(&sc->mii_pdata->mii_media, IFM_MAKEWORD(IFM_ETHER,
IFM_AUTO, 0, sc->mii_inst), 0, NULL);
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
printf("auto\n");
}
MIIBUS_MEDIAINIT(sc->mii_dev);
return (0);
}
static int
brgphy_service(struct mii_softc *sc, struct mii_data *mii, int cmd)
{
struct ifmedia_entry *ife = mii->mii_media.ifm_cur;
int val;
switch (cmd) {
case MII_POLLSTAT:
break;
case MII_MEDIACHG:
/* 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 */
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);
break;
case IFM_2500_SX:
case IFM_1000_SX:
case IFM_1000_T:
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);
break;
default:
return (EINVAL);
}
break;
case MII_TICK:
/* Bail if autoneg isn't in process. */
if (IFM_SUBTYPE(ife->ifm_media) != IFM_AUTO) {
sc->mii_ticks = 0;
break;
}
/*
* Check to see if we have link. If we do, we don't
* need to restart the autonegotiation process.
*/
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);
if (val & BMSR_LINK) {
sc->mii_ticks = 0; /* Reset autoneg timer. */
break;
}
/* Announce link loss right after it happens. */
if (sc->mii_ticks++ == 0)
break;
/* Only retry autonegotiation every mii_anegticks seconds. */
if (sc->mii_ticks <= sc->mii_anegticks)
break;
/* Retry autonegotiation */
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);
break;
}
/* 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);
/*
* Callback if something changed. Note that we need to poke
* the DSP on the Broadcom PHYs if the media changes.
*/
if (sc->mii_media_active != mii->mii_media_active ||
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) {
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:
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_BCM5401:
if (sc->mii_mpd_rev == 1 || sc->mii_mpd_rev == 3)
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:
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:
bcm54k2_load_dspcode(sc);
break;
}
break;
}
}
mii_phy_update(sc, cmd);
return (0);
}
/****************************************************************************/
/* Sets the PHY link speed. */
/* */
/* Returns: */
/* None */
/****************************************************************************/
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)
{
int bmcr = 0, gig;
switch (IFM_SUBTYPE(media)) {
case IFM_2500_SX:
break;
case IFM_1000_SX:
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;
}
if ((media & IFM_FDX) != 0) {
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. */
brgphy_enable_loopback(sc);
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;
}
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;
}
PHY_WRITE(sc, BRGPHY_MII_1000CTL, gig);
PHY_WRITE(sc, BRGPHY_MII_BMCR,
bmcr | BRGPHY_BMCR_AUTOEN | BRGPHY_BMCR_STARTNEG);
}
/****************************************************************************/
/* Set the media status based on the PHY settings. */
/* */
/* Returns: */
/* None */
/****************************************************************************/
static void
brgphy_status(struct mii_softc *sc)
{
struct brgphy_softc *bsc = (struct brgphy_softc *)sc;
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;
mii->mii_media_status = IFM_AVALID;
mii->mii_media_active = IFM_ETHER;
bmsr = PHY_READ(sc, BRGPHY_MII_BMSR) | PHY_READ(sc, BRGPHY_MII_BMSR);
2008-08-12 00:55:03 +00:00
bmcr = PHY_READ(sc, BRGPHY_MII_BMCR);
if (bmcr & BRGPHY_BMCR_LOOP) {
mii->mii_media_active |= IFM_LOOP;
}
if ((bmcr & BRGPHY_BMCR_AUTOEN) &&
(bmsr & BRGPHY_BMSR_ACOMP) == 0 &&
(bsc->serdes_flags & BRGPHY_NOANWAIT) == 0) {
/* 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;
}
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);
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;
}
} 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. */
/* 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);
/* Check for MRBE auto-negotiated speed results. */
switch (xstat & BRGPHY_5708S_PG0_1000X_STAT1_SPEED_MASK) {
2008-08-12 00:52:10 +00:00
case BRGPHY_5708S_PG0_1000X_STAT1_SPEED_10:
mii->mii_media_active |= IFM_10_FL; break;
2008-08-12 00:52:10 +00:00
case BRGPHY_5708S_PG0_1000X_STAT1_SPEED_100:
mii->mii_media_active |= IFM_100_FX; break;
2008-08-12 00:52:10 +00:00
case BRGPHY_5708S_PG0_1000X_STAT1_SPEED_1G:
mii->mii_media_active |= IFM_1000_SX; break;
2008-08-12 00:52:10 +00:00
case BRGPHY_5708S_PG0_1000X_STAT1_SPEED_25G:
mii->mii_media_active |= IFM_2500_SX; break;
}
/* Check for MRBE auto-negotiated duplex results. */
if (xstat & BRGPHY_5708S_PG0_1000X_STAT1_FDX)
mii->mii_media_active |= IFM_FDX;
else
mii->mii_media_active |= IFM_HDX;
} 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);
xstat = PHY_READ(sc, BRGPHY_GP_STATUS_TOP_ANEG_STATUS);
/* 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;
}
/* Check for MRBE auto-negotiated duplex results. */
if (xstat & BRGPHY_GP_STATUS_TOP_ANEG_FDX)
mii->mii_media_active |= IFM_FDX;
else
mii->mii_media_active |= IFM_HDX;
}
}
}
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)
{
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;
- 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);
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);
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);
} 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);
}
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);
PHY_WRITE(sc, BRGPHY_MII_IMR, 0xFF00);
}
/* Enable loopback to force the link down. */
static void
brgphy_enable_loopback(struct mii_softc *sc)
{
int i;
PHY_WRITE(sc, BRGPHY_MII_BMCR, BRGPHY_BMCR_LOOP);
for (i = 0; i < 15000; i++) {
if (!(PHY_READ(sc, BRGPHY_MII_BMSR) & BRGPHY_BMSR_LINK))
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);
}
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);
}
static void
2007-02-12 23:33:05 +00:00
brgphy_fixup_5704_a0_bug(struct mii_softc *sc)
{
static const struct {
int reg;
uint16_t val;
} dspcode[] = {
2007-02-12 23:33:05 +00:00
{ 0x1c, 0x8d68 },
{ 0x1c, 0x8d68 },
{ 0, 0 },
};
int i;
for (i = 0; dspcode[i].reg != 0; i++)
PHY_WRITE(sc, dspcode[i].reg, dspcode[i].val);
}
static void
2007-02-12 23:33:05 +00:00
brgphy_fixup_adc_bug(struct mii_softc *sc)
{
static const struct {
int reg;
2007-02-12 23:33:05 +00:00
uint16_t val;
} dspcode[] = {
2007-02-12 23:33:05 +00:00
{ BRGPHY_MII_AUXCTL, 0x0c00 },
{ BRGPHY_MII_DSP_ADDR_REG, 0x201f },
{ BRGPHY_MII_DSP_RW_PORT, 0x2aaa },
{ 0, 0 },
};
int i;
for (i = 0; dspcode[i].reg != 0; i++)
PHY_WRITE(sc, dspcode[i].reg, dspcode[i].val);
}
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 },
{ BRGPHY_MII_TEST1, 0x0014 },
{ 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);
}
static void
brgphy_fixup_ber_bug(struct mii_softc *sc)
{
static const struct {
int reg;
2007-02-12 23:33:05 +00:00
uint16_t val;
} dspcode[] = {
{ 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 },
{ 0, 0 },
};
int i;
for (i = 0; dspcode[i].reg != 0; i++)
PHY_WRITE(sc, dspcode[i].reg, dspcode[i].val);
}
static void
brgphy_fixup_crc_bug(struct mii_softc *sc)
{
static const struct {
int reg;
uint16_t val;
} dspcode[] = {
{ BRGPHY_MII_DSP_RW_PORT, 0x0a75 },
{ 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);
}
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);
}
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);
}
static void
brgphy_ethernet_wirespeed(struct mii_softc *sc)
{
2007-02-12 23:33:05 +00:00
uint32_t val;
/* 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)
{
2007-02-12 23:33:05 +00:00
uint32_t val;
/* 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) {
2007-02-13 00:34:32 +00:00
/* BCM5401 PHY cannot read-modify-write. */
PHY_WRITE(sc, BRGPHY_MII_AUXCTL, 0x4c20);
2007-02-13 00:34:32 +00:00
} 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);
}
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);
2008-08-12 00:52:10 +00:00
PHY_WRITE(sc, BRGPHY_MII_PHY_EXTCTL,
val & ~BRGPHY_PHY_EXTCTL_HIGH_LA);
}
}
static void
brgphy_reset(struct mii_softc *sc)
{
struct bge_softc *bge_sc = NULL;
struct bce_softc *bce_sc = NULL;
Introduce a procedural interface to the ifnet structure. The new interface allows the ifnet structure to be defined as an opaque type in NIC drivers. This then allows the ifnet structure to be changed without a need to change or recompile NIC drivers. Put differently, NIC drivers can be written and compiled once and be used with different network stack implementations, provided of course that those network stack implementations have an API and ABI compatible interface. This commit introduces the 'if_t' type to replace 'struct ifnet *' as the type of a network interface. The 'if_t' type is defined as 'void *' to enable the compiler to perform type conversion to 'struct ifnet *' and vice versa where needed and without warnings. The functions that implement the API are the only functions that need to have an explicit cast. The MII code has been converted to use the driver API to avoid unnecessary code churn. Code churn comes from having to work with both converted and unconverted drivers in correlation with having callback functions that take an interface. By converting the MII code first, the callback functions can be defined so that the compiler will perform the typecasts automatically. As soon as all drivers have been converted, the if_t type can be redefined as needed and the API functions can be fix to not need an explicit cast. The immediate benefactors of this change are: 1. Juniper Networks - The network stack implementation in Junos is entirely different from FreeBSD's one and this change allows Juniper to build "stock" NIC drivers that can be used in combination with both the FreeBSD and Junos stacks. 2. FreeBSD - This change opens the door towards changing ifnet and implementing new features and optimizations in the network stack without it requiring a change in the many NIC drivers FreeBSD has. Submitted by: Anuranjan Shukla <anshukla@juniper.net> Reviewed by: glebius@ Obtained from: Juniper Networks, Inc.
2014-06-02 17:54:39 +00:00
if_t ifp;
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);
/* 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);
}
/* 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) {
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:
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_BCM5401:
if (sc->mii_mpd_rev == 1 || sc->mii_mpd_rev == 3)
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:
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:
bcm54k2_load_dspcode(sc);
break;
}
break;
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;
2013-07-20 07:24:01 +00:00
case MII_OUI_BROADCOM4:
return;
}
ifp = sc->mii_pdata->mii_ifp;
/* Find the driver associated with this PHY. */
if (mii_phy_mac_match(sc, "bge"))
bge_sc = mii_phy_mac_softc(sc);
else if (mii_phy_mac_match(sc, "bce"))
bce_sc = mii_phy_mac_softc(sc);
if (bge_sc) {
/* Fix up various bugs */
if (bge_sc->bge_phy_flags & BGE_PHY_5704_A0_BUG)
brgphy_fixup_5704_a0_bug(sc);
if (bge_sc->bge_phy_flags & BGE_PHY_ADC_BUG)
2007-02-12 23:33:05 +00:00
brgphy_fixup_adc_bug(sc);
if (bge_sc->bge_phy_flags & BGE_PHY_ADJUST_TRIM)
brgphy_fixup_adjust_trim(sc);
if (bge_sc->bge_phy_flags & BGE_PHY_BER_BUG)
brgphy_fixup_ber_bug(sc);
if (bge_sc->bge_phy_flags & BGE_PHY_CRC_BUG)
brgphy_fixup_crc_bug(sc);
if (bge_sc->bge_phy_flags & BGE_PHY_JITTER_BUG)
brgphy_fixup_jitter_bug(sc);
if (bge_sc->bge_flags & BGE_FLAG_JUMBO)
Introduce a procedural interface to the ifnet structure. The new interface allows the ifnet structure to be defined as an opaque type in NIC drivers. This then allows the ifnet structure to be changed without a need to change or recompile NIC drivers. Put differently, NIC drivers can be written and compiled once and be used with different network stack implementations, provided of course that those network stack implementations have an API and ABI compatible interface. This commit introduces the 'if_t' type to replace 'struct ifnet *' as the type of a network interface. The 'if_t' type is defined as 'void *' to enable the compiler to perform type conversion to 'struct ifnet *' and vice versa where needed and without warnings. The functions that implement the API are the only functions that need to have an explicit cast. The MII code has been converted to use the driver API to avoid unnecessary code churn. Code churn comes from having to work with both converted and unconverted drivers in correlation with having callback functions that take an interface. By converting the MII code first, the callback functions can be defined so that the compiler will perform the typecasts automatically. As soon as all drivers have been converted, the if_t type can be redefined as needed and the API functions can be fix to not need an explicit cast. The immediate benefactors of this change are: 1. Juniper Networks - The network stack implementation in Junos is entirely different from FreeBSD's one and this change allows Juniper to build "stock" NIC drivers that can be used in combination with both the FreeBSD and Junos stacks. 2. FreeBSD - This change opens the door towards changing ifnet and implementing new features and optimizations in the network stack without it requiring a change in the many NIC drivers FreeBSD has. Submitted by: Anuranjan Shukla <anshukla@juniper.net> Reviewed by: glebius@ Obtained from: Juniper Networks, Inc.
2014-06-02 17:54:39 +00:00
brgphy_jumbo_settings(sc, if_getmtu(ifp));
if ((bge_sc->bge_phy_flags & BGE_PHY_NO_WIRESPEED) == 0)
brgphy_ethernet_wirespeed(sc);
/* Enable Link LED on Dell boxes */
if (bge_sc->bge_phy_flags & BGE_PHY_NO_3LED) {
PHY_WRITE(sc, BRGPHY_MII_PHY_EXTCTL,
PHY_READ(sc, BRGPHY_MII_PHY_EXTCTL) &
~BRGPHY_PHY_EXTCTL_3_LED);
}
/* Adjust output voltage (From Linux driver) */
if (bge_sc->bge_asicrev == BGE_ASICREV_BCM5906)
PHY_WRITE(sc, BRGPHY_MII_EPHY_PTEST, 0x12);
} else if (bce_sc) {
if (BCE_CHIP_NUM(bce_sc) == BCE_CHIP_NUM_5708 &&
(bce_sc->bce_phy_flags & BCE_PHY_SERDES_FLAG)) {
/* Store autoneg capabilities/results in digital block (Page 0) */
PHY_WRITE(sc, BRGPHY_5708S_BLOCK_ADDR, BRGPHY_5708S_DIG3_PG2);
2008-08-12 00:55:03 +00:00
PHY_WRITE(sc, BRGPHY_5708S_PG2_DIGCTL_3_0,
BRGPHY_5708S_PG2_DIGCTL_3_0_USE_IEEE);
PHY_WRITE(sc, BRGPHY_5708S_BLOCK_ADDR, BRGPHY_5708S_DIG_PG0);
/* Enable fiber mode and autodetection */
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 |
BRGPHY_5708S_PG0_1000X_CTL1_FIBER_MODE);
/* Enable parallel detection */
2008-08-12 00:55:03 +00:00
PHY_WRITE(sc, BRGPHY_5708S_PG0_1000X_CTL2,
PHY_READ(sc, BRGPHY_5708S_PG0_1000X_CTL2) |
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)
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) |
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)) {
2008-08-12 00:55:03 +00:00
PHY_WRITE(sc, BRGPHY_5708S_BLOCK_ADDR,
BRGPHY_5708S_TX_MISC_PG5);
2008-08-12 00:55:03 +00:00
PHY_WRITE(sc, BRGPHY_5708S_PG5_TXACTL1,
PHY_READ(sc, BRGPHY_5708S_PG5_TXACTL1) & ~0x30);
2008-08-12 00:55:03 +00:00
PHY_WRITE(sc, BRGPHY_5708S_BLOCK_ADDR,
BRGPHY_5708S_DIG_PG0);
}
2008-08-12 00:55:03 +00:00
/* Backplanes use special driver/pre-driver/pre-emphasis values. */
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)) {
2008-08-12 00:55:03 +00:00
PHY_WRITE(sc, BRGPHY_5708S_BLOCK_ADDR,
BRGPHY_5708S_TX_MISC_PG5);
2008-08-12 00:55:03 +00:00
PHY_WRITE(sc, BRGPHY_5708S_PG5_TXACTL3,
bce_sc->bce_port_hw_cfg &
BCE_PORT_HW_CFG_CFG_TXCTL3_MASK);
2008-08-12 00:55:03 +00:00
PHY_WRITE(sc, BRGPHY_5708S_BLOCK_ADDR,
BRGPHY_5708S_DIG_PG0);
}
} else if (BCE_CHIP_NUM(bce_sc) == BCE_CHIP_NUM_5709 &&
(bce_sc->bce_phy_flags & BCE_PHY_SERDES_FLAG)) {
/* Select the SerDes Digital block of the AN MMD. */
PHY_WRITE(sc, BRGPHY_BLOCK_ADDR, BRGPHY_BLOCK_ADDR_SERDES_DIG);
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);
/* Select the Over 1G block of the AN MMD. */
PHY_WRITE(sc, BRGPHY_BLOCK_ADDR, BRGPHY_BLOCK_ADDR_OVER_1G);
/* Enable autoneg "Next Page" to advertise 2.5G support. */
val = PHY_READ(sc, BRGPHY_OVER_1G_UNFORMAT_PG1);
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);
/* Select the Multi-Rate Backplane Ethernet block of the AN MMD. */
PHY_WRITE(sc, BRGPHY_BLOCK_ADDR, BRGPHY_BLOCK_ADDR_MRBE);
/* Enable MRBE speed autoneg. */
val = PHY_READ(sc, BRGPHY_MRBE_MSG_PG5_NP);
val |= BRGPHY_MRBE_MSG_PG5_NP_MBRE |
BRGPHY_MRBE_MSG_PG5_NP_T2;
PHY_WRITE(sc, BRGPHY_MRBE_MSG_PG5_NP, val);
/* Select the Clause 73 User B0 block of the AN MMD. */
PHY_WRITE(sc, BRGPHY_BLOCK_ADDR, BRGPHY_BLOCK_ADDR_CL73_USER_B0);
/* Enable MRBE speed autoneg. */
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);
/* Restore IEEE0 block (assumed in all brgphy(4) code). */
PHY_WRITE(sc, BRGPHY_BLOCK_ADDR, BRGPHY_BLOCK_ADDR_COMBO_IEEE0);
} else if (BCE_CHIP_NUM(bce_sc) == BCE_CHIP_NUM_5709) {
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);
Introduce a procedural interface to the ifnet structure. The new interface allows the ifnet structure to be defined as an opaque type in NIC drivers. This then allows the ifnet structure to be changed without a need to change or recompile NIC drivers. Put differently, NIC drivers can be written and compiled once and be used with different network stack implementations, provided of course that those network stack implementations have an API and ABI compatible interface. This commit introduces the 'if_t' type to replace 'struct ifnet *' as the type of a network interface. The 'if_t' type is defined as 'void *' to enable the compiler to perform type conversion to 'struct ifnet *' and vice versa where needed and without warnings. The functions that implement the API are the only functions that need to have an explicit cast. The MII code has been converted to use the driver API to avoid unnecessary code churn. Code churn comes from having to work with both converted and unconverted drivers in correlation with having callback functions that take an interface. By converting the MII code first, the callback functions can be defined so that the compiler will perform the typecasts automatically. As soon as all drivers have been converted, the if_t type can be redefined as needed and the API functions can be fix to not need an explicit cast. The immediate benefactors of this change are: 1. Juniper Networks - The network stack implementation in Junos is entirely different from FreeBSD's one and this change allows Juniper to build "stock" NIC drivers that can be used in combination with both the FreeBSD and Junos stacks. 2. FreeBSD - This change opens the door towards changing ifnet and implementing new features and optimizations in the network stack without it requiring a change in the many NIC drivers FreeBSD has. Submitted by: Anuranjan Shukla <anshukla@juniper.net> Reviewed by: glebius@ Obtained from: Juniper Networks, Inc.
2014-06-02 17:54:39 +00:00
brgphy_jumbo_settings(sc, if_getmtu(ifp));
brgphy_ethernet_wirespeed(sc);
} else {
brgphy_fixup_ber_bug(sc);
Introduce a procedural interface to the ifnet structure. The new interface allows the ifnet structure to be defined as an opaque type in NIC drivers. This then allows the ifnet structure to be changed without a need to change or recompile NIC drivers. Put differently, NIC drivers can be written and compiled once and be used with different network stack implementations, provided of course that those network stack implementations have an API and ABI compatible interface. This commit introduces the 'if_t' type to replace 'struct ifnet *' as the type of a network interface. The 'if_t' type is defined as 'void *' to enable the compiler to perform type conversion to 'struct ifnet *' and vice versa where needed and without warnings. The functions that implement the API are the only functions that need to have an explicit cast. The MII code has been converted to use the driver API to avoid unnecessary code churn. Code churn comes from having to work with both converted and unconverted drivers in correlation with having callback functions that take an interface. By converting the MII code first, the callback functions can be defined so that the compiler will perform the typecasts automatically. As soon as all drivers have been converted, the if_t type can be redefined as needed and the API functions can be fix to not need an explicit cast. The immediate benefactors of this change are: 1. Juniper Networks - The network stack implementation in Junos is entirely different from FreeBSD's one and this change allows Juniper to build "stock" NIC drivers that can be used in combination with both the FreeBSD and Junos stacks. 2. FreeBSD - This change opens the door towards changing ifnet and implementing new features and optimizations in the network stack without it requiring a change in the many NIC drivers FreeBSD has. Submitted by: Anuranjan Shukla <anshukla@juniper.net> Reviewed by: glebius@ Obtained from: Juniper Networks, Inc.
2014-06-02 17:54:39 +00:00
brgphy_jumbo_settings(sc, if_getmtu(ifp));
brgphy_ethernet_wirespeed(sc);
}
}
}