mirror/freebsd
1
0
Fork 0
mirror of https://git.FreeBSD.org/src.git synced 2024-12-29 12:03:03 +00:00
Code Issues Releases Activity
801bb689ca
freebsd/sys/pci/if_dc.c

3772 lines
91 KiB
C
Raw Normal View History

Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
/*
* Copyright (c) 1997, 1998, 1999
* 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.
*
* $FreeBSD$
*/
/*
* DEC "tulip" clone ethernet driver. Supports the DEC/Intel 21143
* series chips and several workalikes including the following:
*
Add the PCI IDs for the Macronix 98727 and 98732 parts. These are 3.3volt PCI/cardbus chipsets similar to the 98715 (and they have 512-bit hash tables). Also update the man page to mention the 98727/98732 and the SOHOware SFA110A Rev B4 card with the 98715AEC-C chip.
2000-07-17 19:27:41 +00:00
* Macronix 98713/98715/98725/98727/98732 PMAC (www.macronix.com)
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
* Macronix/Lite-On 82c115 PNIC II (www.macronix.com)
* Lite-On 82c168/82c169 PNIC (www.litecom.com)
* ASIX Electronics AX88140A (www.asix.com.tw)
* ASIX Electronics AX88141 (www.asix.com.tw)
* ADMtek AL981 (www.admtek.com.tw)
* ADMtek AN985 (www.admtek.com.tw)
Add support for the Davicom DM9102A 10/100 ethernet controller chip. This is just to make sure we initialize the chip correctly: we need to make the sure the port select bit in CSR6 is set properly so that we use the internal PHY for 10/100 support. (The eval boards I have also include an external HomePNA PHY, but I need to play with that more before I can support it.)
2000-01-19 19:03:08 +00:00
* Davicom DM9100, DM9102, DM9102A (www.davicom8.com)
Add support for the Accton EN1217. PR: 18735 Submitted by: Adoal Xu <adoal@iname.com>
2000-06-11 11:54:52 +00:00
* Accton EN1217 (www.accton.com)
NEWCARD/Cardbus - This commit adds support for Xircom X3201 based cardbus cards. Support for the TDK 78Q2120 MII is also added. IBM Etherjet, Intel and Xircom cards uses these chips. Note that as a result of this commit, some Intel/DEC 21143 based cardbus cards will also attach, but not get link. That is being looked at.
2000-10-19 08:34:32 +00:00
* Xircom X3201 (www.xircom.com)
Add PCI IDs for some additional cardbus cards. Yes, there really is a RealTek 8139 cardbus device. Unfortunately it doesn't quite work yet because the CIS parser barfs on it. Submitted by msmith, with some small tweaks by me.
2000-10-28 09:00:20 +00:00
* Abocom FE2500
Add support for Conexant LANfinity miniPCI controllers. People who have laptops with this chip should test this and report back as I don't have access to this hardware myself. People with -stable systems should try the patch at: http://www.freebsd.org/~wpaul/conexant.patch.gz Submitted by: Phil Kernick <Phil@Kernick.org>
2001-09-04 17:10:11 +00:00
* Conexant LANfinity (www.conexant.com)
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
*
* Datasheets for the 21143 are available at developer.intel.com.
* Datasheets for the clone parts can be found at their respective sites.
* (Except for the PNIC; see www.freebsd.org/~wpaul/PNIC/pnic.ps.gz.)
* The PNIC II is essentially a Macronix 98715A chip; the only difference
* worth noting is that its multicast hash table is only 128 bits wide
* instead of 512.
*
* Written by Bill Paul <wpaul@ee.columbia.edu>
* Electrical Engineering Department
* Columbia University, New York City
*/
/*
* The Intel 21143 is the successor to the DEC 21140. It is basically
* the same as the 21140 but with a few new features. The 21143 supports
* three kinds of media attachments:
*
* o MII port, for 10Mbps and 100Mbps support and NWAY
* autonegotiation provided by an external PHY.
* o SYM port, for symbol mode 100Mbps support.
* o 10baseT port.
* o AUI/BNC port.
*
* The 100Mbps SYM port and 10baseT port can be used together in
* combination with the internal NWAY support to create a 10/100
* autosensing configuration.
*
* Note that not all tulip workalikes are handled in this driver: we only
* deal with those which are relatively well behaved. The Winbond is
* handled separately due to its different register offsets and the
* special handling needed for its various bugs. The PNIC is handled
* here, but I'm not thrilled about it.
*
* All of the workalike chips use some form of MII transceiver support
* with the exception of the Macronix chips, which also have a SYM port.
* The ASIX AX88140A is also documented to have a SYM port, but all
* the cards I've seen use an MII transceiver, probably because the
* AX88140A doesn't support internal NWAY.
*/
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/sockio.h>
#include <sys/mbuf.h>
#include <sys/malloc.h>
#include <sys/kernel.h>
#include <sys/socket.h>
For i386 architecture, remove an expensive m_devget() (and the underlying unaligned bcopy) on incoming packets that are already available (albeit unaligned) in a buffer. The performance improvement varies, depending on CPU and memory speed, but can be quite large especially on slow CPUs. I have seen over 50% increase on forwarding speed on the sis driver for the 486/133 (embedded systems), which does exactly the same thing. The behaviour is controlled by a sysctl variable, hw.dc_quick which defaults to 1. Set it to 0 to restore the old behaviour. After running a few experiments (in userland, though) I am convinced that doing the m_devget() is detrimental to performance in almost all cases. Even if your CPU has degraded performance with misaligned data, the bcopy() in the driver has the same overhead due to misaligment as the one that you save in the uiomove(), plus you do one extra copy and pollute the cache. But more often than not, you do not even have to touch the payload, e.g. when you are forwarding packets, and even in the often-cited case of NFS, you often end up passing a pointer to the payload to the disk controller. In any case, you can play with the sysctl variable to toggle between the two behaviours, and see if it makes a difference. MFC-after: 3 days
2001-11-29 22:46:48 +00:00
#include <sys/sysctl.h>
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
#include <net/if.h>
#include <net/if_arp.h>
#include <net/ethernet.h>
#include <net/if_dl.h>
#include <net/if_media.h>
Add VLAN for the dc(4) driver (ie long frame). The patch is 2 parts. One to notify the system that the MTU for VLAN can be 1500 so the vlan will automatically be configured with a 1500 MTU the other is to ignore the error case if the received frame is to long. The frame size notification came from code in the SIS driver, and the support for long frames derived from the NetBSD Tulip driver. Tested on: 4 port D-Link adapter DFE-570TX 4 Intel 21143 Netgear card with 82c169 PNIC 10/100BaseTX Reviewed by: ru (manpage), wpaul (not objected to), archie Approved by: imp Obtained from: NetBSD
2002-01-16 16:33:58 +00:00
#include <net/if_types.h>
#include <net/if_vlan_var.h>
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
#include <net/bpf.h>
#include <vm/vm.h> /* for vtophys */
#include <vm/pmap.h> /* for vtophys */
#include <machine/bus_pio.h>
#include <machine/bus_memio.h>
#include <machine/bus.h>
#include <machine/resource.h>
#include <sys/bus.h>
#include <sys/rman.h>
#include <dev/mii/mii.h>
#include <dev/mii/miivar.h>
#include <pci/pcireg.h>
#include <pci/pcivar.h>
#define DC_USEIOSPACE
Add support for parsing the media blocks from the SROM on 21143 adapters. This is necessary in order to make this driver work with the built-in ethernet on the alpha Miata machines. These systems have a 21143-PC chip on-board and optional daughtercards with either a 10/100 MII transceiver or a 10baseT/10base2 transceiver. In both cases, you need to twiddle the GPIO bits on the controller in order to turn the transceivers on, and you have to read the media info from the SROM in order to find out what bits to twiddle.
2000-10-05 17:36:14 +00:00
#ifdef __alpha__
#define SRM_MEDIA
#endif
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
#include <pci/if_dcreg.h>
Depend on miibus. Note that if_aue doesn't strictly depend on usb because it uses the method interface for calls rather than using internal symbols, and because it's a child driver of usb and therefore will not try and do anything unless the parent usb code is loaded at some point. if_aue does strictly depend on miibus as it will fail to link if it is missing.
2000-04-29 13:41:57 +00:00
MODULE_DEPEND(dc, miibus, 1, 1, 1);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
/* "controller miibus0" required. See GENERIC if you get errors here. */
#include "miibus_if.h"
#ifndef lint
static const char rcsid[] =
"$FreeBSD$";
#endif
/*
* Various supported device vendors/types and their names.
*/
static struct dc_type dc_devs[] = {
{ DC_VENDORID_DEC, DC_DEVICEID_21143,
"Intel 21143 10/100BaseTX" },
Add support for the Davicom DM9009 chipset. PR: 46859 Submitted by: Boaz Haberman <boaz@ool-182f8b09.dyn.optonline.net> Approved by: rwatson
2003-01-09 20:51:39 +00:00
{ DC_VENDORID_DAVICOM, DC_DEVICEID_DM9009,
"Davicom DM9009 10/100BaseTX" },
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
{ DC_VENDORID_DAVICOM, DC_DEVICEID_DM9100,
"Davicom DM9100 10/100BaseTX" },
{ DC_VENDORID_DAVICOM, DC_DEVICEID_DM9102,
"Davicom DM9102 10/100BaseTX" },
Add support for the Davicom DM9102A 10/100 ethernet controller chip. This is just to make sure we initialize the chip correctly: we need to make the sure the port select bit in CSR6 is set properly so that we use the internal PHY for 10/100 support. (The eval boards I have also include an external HomePNA PHY, but I need to play with that more before I can support it.)
2000-01-19 19:03:08 +00:00
{ DC_VENDORID_DAVICOM, DC_DEVICEID_DM9102,
"Davicom DM9102A 10/100BaseTX" },
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
{ DC_VENDORID_ADMTEK, DC_DEVICEID_AL981,
"ADMtek AL981 10/100BaseTX" },
{ DC_VENDORID_ADMTEK, DC_DEVICEID_AN985,
"ADMtek AN985 10/100BaseTX" },
{ DC_VENDORID_ASIX, DC_DEVICEID_AX88140A,
"ASIX AX88140A 10/100BaseTX" },
{ DC_VENDORID_ASIX, DC_DEVICEID_AX88140A,
"ASIX AX88141 10/100BaseTX" },
{ DC_VENDORID_MX, DC_DEVICEID_98713,
"Macronix 98713 10/100BaseTX" },
{ DC_VENDORID_MX, DC_DEVICEID_98713,
"Macronix 98713A 10/100BaseTX" },
{ DC_VENDORID_CP, DC_DEVICEID_98713_CP,
"Compex RL100-TX 10/100BaseTX" },
{ DC_VENDORID_CP, DC_DEVICEID_98713_CP,
"Compex RL100-TX 10/100BaseTX" },
{ DC_VENDORID_MX, DC_DEVICEID_987x5,
"Macronix 98715/98715A 10/100BaseTX" },
Apply patch to the dc driver to handle Macronix MX98715AEC-C/D/E chips, which differ slightly from the Macronix MX98715AEC chip on the sample adapter that I have in that the multicast hash table is only 128 bits wide instead of 512. New adapters are popping up with this chip, and due to improper handling of the smaller hash table, broadcast packets were not being received correctly.
2000-07-15 17:54:30 +00:00
{ DC_VENDORID_MX, DC_DEVICEID_987x5,
"Macronix 98715AEC-C 10/100BaseTX" },
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
{ DC_VENDORID_MX, DC_DEVICEID_987x5,
"Macronix 98725 10/100BaseTX" },
Add the PCI IDs for the Macronix 98727 and 98732 parts. These are 3.3volt PCI/cardbus chipsets similar to the 98715 (and they have 512-bit hash tables). Also update the man page to mention the 98727/98732 and the SOHOware SFA110A Rev B4 card with the 98715AEC-C chip.
2000-07-17 19:27:41 +00:00
{ DC_VENDORID_MX, DC_DEVICEID_98727,
"Macronix 98727/98732 10/100BaseTX" },
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
{ DC_VENDORID_LO, DC_DEVICEID_82C115,
"LC82C115 PNIC II 10/100BaseTX" },
{ DC_VENDORID_LO, DC_DEVICEID_82C168,
"82c168 PNIC 10/100BaseTX" },
{ DC_VENDORID_LO, DC_DEVICEID_82C168,
"82c169 PNIC 10/100BaseTX" },
Add support for the Accton EN1217. PR: 18735 Submitted by: Adoal Xu <adoal@iname.com>
2000-06-11 11:54:52 +00:00
{ DC_VENDORID_ACCTON, DC_DEVICEID_EN1217,
"Accton EN1217 10/100BaseTX" },
Add support for the Accton EN2242 MiniPCI adapter. This is just an ADMtek Centaur chip, so all we need is the PCI ID. Submitted by: Scott Lang <scottl@FreeBSD.org>
2000-11-14 19:35:22 +00:00
{ DC_VENDORID_ACCTON, DC_DEVICEID_EN2242,
"Accton EN2242 MiniPCI 10/100BaseTX" },
Make consistent; turn spaces into tabs where there is a mixture.
2002-09-04 18:14:17 +00:00
{ DC_VENDORID_XIRCOM, DC_DEVICEID_X3201,
NEWCARD/Cardbus - This commit adds support for Xircom X3201 based cardbus cards. Support for the TDK 78Q2120 MII is also added. IBM Etherjet, Intel and Xircom cards uses these chips. Note that as a result of this commit, some Intel/DEC 21143 based cardbus cards will also attach, but not get link. That is being looked at.
2000-10-19 08:34:32 +00:00
"Xircom X3201 10/100BaseTX" },
Add PCI IDs for some additional cardbus cards. Yes, there really is a RealTek 8139 cardbus device. Unfortunately it doesn't quite work yet because the CIS parser barfs on it. Submitted by msmith, with some small tweaks by me.
2000-10-28 09:00:20 +00:00
{ DC_VENDORID_ABOCOM, DC_DEVICEID_FE2500,
"Abocom FE2500 10/100BaseTX" },
Add support for Conexant LANfinity miniPCI controllers. People who have laptops with this chip should test this and report back as I don't have access to this hardware myself. People with -stable systems should try the patch at: http://www.freebsd.org/~wpaul/conexant.patch.gz Submitted by: Phil Kernick <Phil@Kernick.org>
2001-09-04 17:10:11 +00:00
{ DC_VENDORID_CONEXANT, DC_DEVICEID_RS7112,
"Conexant LANfinity MiniPCI 10/100BaseTX" },
Add preliminary support for the Hawking PN672TX CardBus cards. # Preliminary because there are some subtle things the NetBSD driver does # that we don't do yet. My card works for me w/o them.
2003-01-10 05:37:15 +00:00
{ DC_VENDORID_HAWKING, DC_DEVICEID_HAWKING_PN672TX,
"Hawking CB102 CardBus 10/100" },
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
{ 0, 0, NULL }
};
Remove __P.
2002-03-20 02:08:01 +00:00
static int dc_probe (device_t);
static int dc_attach (device_t);
static int dc_detach (device_t);
Add suspend/resume code mostly merged from fxp/rl driver. This is temporary hack, better and generalized solution probably should be implemented at lower layer(MII or PCI?). Tested by: shoko.araki@soliton.co.jp MFC after: 1 week
2002-05-06 10:55:42 +00:00
static int dc_suspend (device_t);
static int dc_resume (device_t);
Remove __P.
2002-03-20 02:08:01 +00:00
static void dc_acpi (device_t);
static struct dc_type *dc_devtype (device_t);
static int dc_newbuf (struct dc_softc *, int, struct mbuf *);
static int dc_encap (struct dc_softc *, struct mbuf *, u_int32_t *);
static int dc_coal (struct dc_softc *, struct mbuf **);
static void dc_pnic_rx_bug_war (struct dc_softc *, int);
static int dc_rx_resync (struct dc_softc *);
static void dc_rxeof (struct dc_softc *);
static void dc_txeof (struct dc_softc *);
static void dc_tick (void *);
static void dc_tx_underrun (struct dc_softc *);
static void dc_intr (void *);
static void dc_start (struct ifnet *);
static int dc_ioctl (struct ifnet *, u_long, caddr_t);
static void dc_init (void *);
static void dc_stop (struct dc_softc *);
static void dc_watchdog (struct ifnet *);
static void dc_shutdown (device_t);
static int dc_ifmedia_upd (struct ifnet *);
static void dc_ifmedia_sts (struct ifnet *, struct ifmediareq *);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
Remove __P.
2002-03-20 02:08:01 +00:00
static void dc_delay (struct dc_softc *);
static void dc_eeprom_idle (struct dc_softc *);
static void dc_eeprom_putbyte (struct dc_softc *, int);
static void dc_eeprom_getword (struct dc_softc *, int, u_int16_t *);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
static void dc_eeprom_getword_pnic
Remove __P.
2002-03-20 02:08:01 +00:00
(struct dc_softc *, int, u_int16_t *);
NEWCARD/Cardbus - This commit adds support for Xircom X3201 based cardbus cards. Support for the TDK 78Q2120 MII is also added. IBM Etherjet, Intel and Xircom cards uses these chips. Note that as a result of this commit, some Intel/DEC 21143 based cardbus cards will also attach, but not get link. That is being looked at.
2000-10-19 08:34:32 +00:00
static void dc_eeprom_getword_xircom
Remove __P.
2002-03-20 02:08:01 +00:00
(struct dc_softc *, int, u_int16_t *);
Dynamically configure the width of the srom. This code comes from OpenBSD who got the code (or the idea) from the NetBSD tlp driver. This gets some cardbus dc cards working (either completely or nearly so). It also appears to get additional pci cards working, without breaking working ones. # Maybe some additional work is needed here. Also, the cardbus attachment # might need to match on the CIS rather than on the vendor/device so we have # a finer level of detail as to what the card is. Technically, the # vendor/device fields are undefined for CardBus (even though most cards are # using common silicon with pci models).
2002-10-07 05:26:35 +00:00
static void dc_eeprom_width (struct dc_softc *);
Remove __P.
2002-03-20 02:08:01 +00:00
static void dc_read_eeprom (struct dc_softc *, caddr_t, int, int, int);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
Remove __P.
2002-03-20 02:08:01 +00:00
static void dc_mii_writebit (struct dc_softc *, int);
static int dc_mii_readbit (struct dc_softc *);
static void dc_mii_sync (struct dc_softc *);
static void dc_mii_send (struct dc_softc *, u_int32_t, int);
static int dc_mii_readreg (struct dc_softc *, struct dc_mii_frame *);
static int dc_mii_writereg (struct dc_softc *, struct dc_mii_frame *);
static int dc_miibus_readreg (device_t, int, int);
static int dc_miibus_writereg (device_t, int, int, int);
static void dc_miibus_statchg (device_t);
static void dc_miibus_mediainit (device_t);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
Remove __P.
2002-03-20 02:08:01 +00:00
static void dc_setcfg (struct dc_softc *, int);
static u_int32_t dc_crc_le (struct dc_softc *, caddr_t);
static u_int32_t dc_crc_be (caddr_t);
static void dc_setfilt_21143 (struct dc_softc *);
static void dc_setfilt_asix (struct dc_softc *);
static void dc_setfilt_admtek (struct dc_softc *);
static void dc_setfilt_xircom (struct dc_softc *);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
Remove __P.
2002-03-20 02:08:01 +00:00
static void dc_setfilt (struct dc_softc *);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
Remove __P.
2002-03-20 02:08:01 +00:00
static void dc_reset (struct dc_softc *);
static int dc_list_rx_init (struct dc_softc *);
static int dc_list_tx_init (struct dc_softc *);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
Dynamically configure the width of the srom. This code comes from OpenBSD who got the code (or the idea) from the NetBSD tlp driver. This gets some cardbus dc cards working (either completely or nearly so). It also appears to get additional pci cards working, without breaking working ones. # Maybe some additional work is needed here. Also, the cardbus attachment # might need to match on the CIS rather than on the vendor/device so we have # a finer level of detail as to what the card is. Technically, the # vendor/device fields are undefined for CardBus (even though most cards are # using common silicon with pci models).
2002-10-07 05:26:35 +00:00
static void dc_read_srom (struct dc_softc *, int);
Remove __P.
2002-03-20 02:08:01 +00:00
static void dc_parse_21143_srom (struct dc_softc *);
static void dc_decode_leaf_sia (struct dc_softc *, struct dc_eblock_sia *);
static void dc_decode_leaf_mii (struct dc_softc *, struct dc_eblock_mii *);
static void dc_decode_leaf_sym (struct dc_softc *, struct dc_eblock_sym *);
static void dc_apply_fixup (struct dc_softc *, int);
Add support for parsing the media blocks from the SROM on 21143 adapters. This is necessary in order to make this driver work with the built-in ethernet on the alpha Miata machines. These systems have a 21143-PC chip on-board and optional daughtercards with either a 10/100 MII transceiver or a 10baseT/10base2 transceiver. In both cases, you need to twiddle the GPIO bits on the controller in order to turn the transceivers on, and you have to read the media info from the SROM in order to find out what bits to twiddle.
2000-10-05 17:36:14 +00:00
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
#ifdef DC_USEIOSPACE
#define DC_RES SYS_RES_IOPORT
#define DC_RID DC_PCI_CFBIO
#else
#define DC_RES SYS_RES_MEMORY
#define DC_RID DC_PCI_CFBMA
#endif
static device_method_t dc_methods[] = {
/* Device interface */
DEVMETHOD(device_probe, dc_probe),
DEVMETHOD(device_attach, dc_attach),
DEVMETHOD(device_detach, dc_detach),
Add suspend/resume code mostly merged from fxp/rl driver. This is temporary hack, better and generalized solution probably should be implemented at lower layer(MII or PCI?). Tested by: shoko.araki@soliton.co.jp MFC after: 1 week
2002-05-06 10:55:42 +00:00
DEVMETHOD(device_suspend, dc_suspend),
DEVMETHOD(device_resume, dc_resume),
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
DEVMETHOD(device_shutdown, dc_shutdown),
/* bus interface */
DEVMETHOD(bus_print_child, bus_generic_print_child),
DEVMETHOD(bus_driver_added, bus_generic_driver_added),
/* MII interface */
DEVMETHOD(miibus_readreg, dc_miibus_readreg),
DEVMETHOD(miibus_writereg, dc_miibus_writereg),
DEVMETHOD(miibus_statchg, dc_miibus_statchg),
Add support for DM9102A boards with Davicom DM9801 HomePNA PHYs.
2000-01-24 17:19:37 +00:00
DEVMETHOD(miibus_mediainit, dc_miibus_mediainit),
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
{ 0, 0 }
};
static driver_t dc_driver = {
"dc",
dc_methods,
sizeof(struct dc_softc)
};
static devclass_t dc_devclass;
For i386 architecture, remove an expensive m_devget() (and the underlying unaligned bcopy) on incoming packets that are already available (albeit unaligned) in a buffer. The performance improvement varies, depending on CPU and memory speed, but can be quite large especially on slow CPUs. I have seen over 50% increase on forwarding speed on the sis driver for the 486/133 (embedded systems), which does exactly the same thing. The behaviour is controlled by a sysctl variable, hw.dc_quick which defaults to 1. Set it to 0 to restore the old behaviour. After running a few experiments (in userland, though) I am convinced that doing the m_devget() is detrimental to performance in almost all cases. Even if your CPU has degraded performance with misaligned data, the bcopy() in the driver has the same overhead due to misaligment as the one that you save in the uiomove(), plus you do one extra copy and pollute the cache. But more often than not, you do not even have to touch the payload, e.g. when you are forwarding packets, and even in the often-cited case of NFS, you often end up passing a pointer to the payload to the disk controller. In any case, you can play with the sysctl variable to toggle between the two behaviours, and see if it makes a difference. MFC-after: 3 days
2001-11-29 22:46:48 +00:00
#ifdef __i386__
static int dc_quick=1;
SYSCTL_INT(_hw, OID_AUTO, dc_quick, CTLFLAG_RW,
&dc_quick,0,"do not mdevget in dc driver");
#endif
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
NEWCARD/Cardbus - This commit adds support for Xircom X3201 based cardbus cards. Support for the TDK 78Q2120 MII is also added. IBM Etherjet, Intel and Xircom cards uses these chips. Note that as a result of this commit, some Intel/DEC 21143 based cardbus cards will also attach, but not get link. That is being looked at.
2000-10-19 08:34:32 +00:00
DRIVER_MODULE(if_dc, cardbus, dc_driver, dc_devclass, 0, 0);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
DRIVER_MODULE(if_dc, pci, dc_driver, dc_devclass, 0, 0);
DRIVER_MODULE(miibus, dc, miibus_driver, miibus_devclass, 0, 0);
#define DC_SETBIT(sc, reg, x) \
CSR_WRITE_4(sc, reg, CSR_READ_4(sc, reg) | (x))
#define DC_CLRBIT(sc, reg, x) \
CSR_WRITE_4(sc, reg, CSR_READ_4(sc, reg) & ~(x))
#define SIO_SET(x) DC_SETBIT(sc, DC_SIO, (x))
#define SIO_CLR(x) DC_CLRBIT(sc, DC_SIO, (x))
Change the driver to allocate its own callout structure, and modify the interface to use callout_* instead of timeout(). Also add an IS_MPSAFE #define (currently off) which will mark the driver as mpsafe to the upper layers.
2000-11-25 08:00:10 +00:00
#define IS_MPSAFE 0
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static void
dc_delay(sc)
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
struct dc_softc *sc;
{
int idx;
for (idx = (300 / 33) + 1; idx > 0; idx--)
CSR_READ_4(sc, DC_BUSCTL);
}
Be consistent about functions being static. Spotted by: FlexeLint.
2002-10-16 09:04:52 +00:00
static void
dc_eeprom_width(sc)
Dynamically configure the width of the srom. This code comes from OpenBSD who got the code (or the idea) from the NetBSD tlp driver. This gets some cardbus dc cards working (either completely or nearly so). It also appears to get additional pci cards working, without breaking working ones. # Maybe some additional work is needed here. Also, the cardbus attachment # might need to match on the CIS rather than on the vendor/device so we have # a finer level of detail as to what the card is. Technically, the # vendor/device fields are undefined for CardBus (even though most cards are # using common silicon with pci models).
2002-10-07 05:26:35 +00:00
struct dc_softc *sc;
{
int i;
/* Force EEPROM to idle state. */
dc_eeprom_idle(sc);
/* Enter EEPROM access mode. */
CSR_WRITE_4(sc, DC_SIO, DC_SIO_EESEL);
dc_delay(sc);
DC_SETBIT(sc, DC_SIO, DC_SIO_ROMCTL_READ);
dc_delay(sc);
DC_CLRBIT(sc, DC_SIO, DC_SIO_EE_CLK);
dc_delay(sc);
DC_SETBIT(sc, DC_SIO, DC_SIO_EE_CS);
dc_delay(sc);
for (i = 3; i--;) {
if (6 & (1 << i))
DC_SETBIT(sc, DC_SIO, DC_SIO_EE_DATAIN);
else
DC_CLRBIT(sc, DC_SIO, DC_SIO_EE_DATAIN);
dc_delay(sc);
DC_SETBIT(sc, DC_SIO, DC_SIO_EE_CLK);
dc_delay(sc);
DC_CLRBIT(sc, DC_SIO, DC_SIO_EE_CLK);
dc_delay(sc);
}
for (i = 1; i <= 12; i++) {
DC_SETBIT(sc, DC_SIO, DC_SIO_EE_CLK);
dc_delay(sc);
if (!(CSR_READ_4(sc, DC_SIO) & DC_SIO_EE_DATAOUT)) {
DC_CLRBIT(sc, DC_SIO, DC_SIO_EE_CLK);
dc_delay(sc);
break;
}
DC_CLRBIT(sc, DC_SIO, DC_SIO_EE_CLK);
dc_delay(sc);
}
/* Turn off EEPROM access mode. */
dc_eeprom_idle(sc);
if (i < 4 || i > 12)
sc->dc_romwidth = 6;
else
sc->dc_romwidth = i;
/* Enter EEPROM access mode. */
CSR_WRITE_4(sc, DC_SIO, DC_SIO_EESEL);
dc_delay(sc);
DC_SETBIT(sc, DC_SIO, DC_SIO_ROMCTL_READ);
dc_delay(sc);
DC_CLRBIT(sc, DC_SIO, DC_SIO_EE_CLK);
dc_delay(sc);
DC_SETBIT(sc, DC_SIO, DC_SIO_EE_CS);
dc_delay(sc);
/* Turn off EEPROM access mode. */
dc_eeprom_idle(sc);
}
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static void
dc_eeprom_idle(sc)
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
struct dc_softc *sc;
{
register int i;
CSR_WRITE_4(sc, DC_SIO, DC_SIO_EESEL);
dc_delay(sc);
DC_SETBIT(sc, DC_SIO, DC_SIO_ROMCTL_READ);
dc_delay(sc);
DC_CLRBIT(sc, DC_SIO, DC_SIO_EE_CLK);
dc_delay(sc);
DC_SETBIT(sc, DC_SIO, DC_SIO_EE_CS);
dc_delay(sc);
for (i = 0; i < 25; i++) {
DC_CLRBIT(sc, DC_SIO, DC_SIO_EE_CLK);
dc_delay(sc);
DC_SETBIT(sc, DC_SIO, DC_SIO_EE_CLK);
dc_delay(sc);
}
DC_CLRBIT(sc, DC_SIO, DC_SIO_EE_CLK);
dc_delay(sc);
DC_CLRBIT(sc, DC_SIO, DC_SIO_EE_CS);
dc_delay(sc);
CSR_WRITE_4(sc, DC_SIO, 0x00000000);
return;
}
/*
* Send a read command and address to the EEPROM, check for ACK.
*/
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static void
dc_eeprom_putbyte(sc, addr)
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
struct dc_softc *sc;
int addr;
{
register int d, i;
Dynamically configure the width of the srom. This code comes from OpenBSD who got the code (or the idea) from the NetBSD tlp driver. This gets some cardbus dc cards working (either completely or nearly so). It also appears to get additional pci cards working, without breaking working ones. # Maybe some additional work is needed here. Also, the cardbus attachment # might need to match on the CIS rather than on the vendor/device so we have # a finer level of detail as to what the card is. Technically, the # vendor/device fields are undefined for CardBus (even though most cards are # using common silicon with pci models).
2002-10-07 05:26:35 +00:00
d = DC_EECMD_READ >> 6;
for (i = 3; i--; ) {
if (d & (1 << i))
DC_SETBIT(sc, DC_SIO, DC_SIO_EE_DATAIN);
else
DC_CLRBIT(sc, DC_SIO, DC_SIO_EE_DATAIN);
dc_delay(sc);
DC_SETBIT(sc, DC_SIO, DC_SIO_EE_CLK);
dc_delay(sc);
DC_CLRBIT(sc, DC_SIO, DC_SIO_EE_CLK);
dc_delay(sc);
}
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
/*
* Feed in each bit and strobe the clock.
*/
Dynamically configure the width of the srom. This code comes from OpenBSD who got the code (or the idea) from the NetBSD tlp driver. This gets some cardbus dc cards working (either completely or nearly so). It also appears to get additional pci cards working, without breaking working ones. # Maybe some additional work is needed here. Also, the cardbus attachment # might need to match on the CIS rather than on the vendor/device so we have # a finer level of detail as to what the card is. Technically, the # vendor/device fields are undefined for CardBus (even though most cards are # using common silicon with pci models).
2002-10-07 05:26:35 +00:00
for (i = sc->dc_romwidth; i--;) {
if (addr & (1 << i)) {
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
SIO_SET(DC_SIO_EE_DATAIN);
} else {
SIO_CLR(DC_SIO_EE_DATAIN);
}
dc_delay(sc);
SIO_SET(DC_SIO_EE_CLK);
dc_delay(sc);
SIO_CLR(DC_SIO_EE_CLK);
dc_delay(sc);
}
return;
}
/*
* Read a word of data stored in the EEPROM at address 'addr.'
* The PNIC 82c168/82c169 has its own non-standard way to read
* the EEPROM.
*/
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static void
dc_eeprom_getword_pnic(sc, addr, dest)
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
struct dc_softc *sc;
int addr;
u_int16_t *dest;
{
register int i;
u_int32_t r;
CSR_WRITE_4(sc, DC_PN_SIOCTL, DC_PN_EEOPCODE_READ|addr);
for (i = 0; i < DC_TIMEOUT; i++) {
DELAY(1);
r = CSR_READ_4(sc, DC_SIO);
if (!(r & DC_PN_SIOCTL_BUSY)) {
*dest = (u_int16_t)(r & 0xFFFF);
return;
}
}
return;
}
NEWCARD/Cardbus - This commit adds support for Xircom X3201 based cardbus cards. Support for the TDK 78Q2120 MII is also added. IBM Etherjet, Intel and Xircom cards uses these chips. Note that as a result of this commit, some Intel/DEC 21143 based cardbus cards will also attach, but not get link. That is being looked at.
2000-10-19 08:34:32 +00:00
/*
* Read a word of data stored in the EEPROM at address 'addr.'
* The Xircom X3201 has its own non-standard way to read
* the EEPROM, too.
*/
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static void
dc_eeprom_getword_xircom(sc, addr, dest)
NEWCARD/Cardbus - This commit adds support for Xircom X3201 based cardbus cards. Support for the TDK 78Q2120 MII is also added. IBM Etherjet, Intel and Xircom cards uses these chips. Note that as a result of this commit, some Intel/DEC 21143 based cardbus cards will also attach, but not get link. That is being looked at.
2000-10-19 08:34:32 +00:00
struct dc_softc *sc;
int addr;
u_int16_t *dest;
{
SIO_SET(DC_SIO_ROMSEL | DC_SIO_ROMCTL_READ);
addr *= 2;
CSR_WRITE_4(sc, DC_ROM, addr | 0x160);
*dest = (u_int16_t)CSR_READ_4(sc, DC_SIO)&0xff;
addr += 1;
CSR_WRITE_4(sc, DC_ROM, addr | 0x160);
*dest |= ((u_int16_t)CSR_READ_4(sc, DC_SIO)&0xff) << 8;
SIO_CLR(DC_SIO_ROMSEL | DC_SIO_ROMCTL_READ);
return;
}
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
/*
* Read a word of data stored in the EEPROM at address 'addr.'
*/
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static void
dc_eeprom_getword(sc, addr, dest)
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
struct dc_softc *sc;
int addr;
u_int16_t *dest;
{
register int i;
u_int16_t word = 0;
/* Force EEPROM to idle state. */
dc_eeprom_idle(sc);
/* Enter EEPROM access mode. */
CSR_WRITE_4(sc, DC_SIO, DC_SIO_EESEL);
dc_delay(sc);
DC_SETBIT(sc, DC_SIO, DC_SIO_ROMCTL_READ);
dc_delay(sc);
DC_CLRBIT(sc, DC_SIO, DC_SIO_EE_CLK);
dc_delay(sc);
DC_SETBIT(sc, DC_SIO, DC_SIO_EE_CS);
dc_delay(sc);
/*
* Send address of word we want to read.
*/
dc_eeprom_putbyte(sc, addr);
/*
* Start reading bits from EEPROM.
*/
for (i = 0x8000; i; i >>= 1) {
SIO_SET(DC_SIO_EE_CLK);
dc_delay(sc);
if (CSR_READ_4(sc, DC_SIO) & DC_SIO_EE_DATAOUT)
word |= i;
dc_delay(sc);
SIO_CLR(DC_SIO_EE_CLK);
dc_delay(sc);
}
/* Turn off EEPROM access mode. */
dc_eeprom_idle(sc);
*dest = word;
return;
}
/*
* Read a sequence of words from the EEPROM.
*/
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static void
dc_read_eeprom(sc, dest, off, cnt, swap)
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
struct dc_softc *sc;
caddr_t dest;
int off;
int cnt;
int swap;
{
int i;
u_int16_t word = 0, *ptr;
for (i = 0; i < cnt; i++) {
if (DC_IS_PNIC(sc))
dc_eeprom_getword_pnic(sc, off + i, &word);
NEWCARD/Cardbus - This commit adds support for Xircom X3201 based cardbus cards. Support for the TDK 78Q2120 MII is also added. IBM Etherjet, Intel and Xircom cards uses these chips. Note that as a result of this commit, some Intel/DEC 21143 based cardbus cards will also attach, but not get link. That is being looked at.
2000-10-19 08:34:32 +00:00
else if (DC_IS_XIRCOM(sc))
dc_eeprom_getword_xircom(sc, off + i, &word);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
else
dc_eeprom_getword(sc, off + i, &word);
ptr = (u_int16_t *)(dest + (i * 2));
if (swap)
*ptr = ntohs(word);
else
*ptr = word;
}
return;
}
/*
* The following two routines are taken from the Macronix 98713
* Application Notes pp.19-21.
*/
/*
* Write a bit to the MII bus.
*/
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static void
dc_mii_writebit(sc, bit)
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
struct dc_softc *sc;
int bit;
{
if (bit)
CSR_WRITE_4(sc, DC_SIO,
DC_SIO_ROMCTL_WRITE|DC_SIO_MII_DATAOUT);
else
CSR_WRITE_4(sc, DC_SIO, DC_SIO_ROMCTL_WRITE);
DC_SETBIT(sc, DC_SIO, DC_SIO_MII_CLK);
DC_CLRBIT(sc, DC_SIO, DC_SIO_MII_CLK);
return;
}
/*
* Read a bit from the MII bus.
*/
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static int
dc_mii_readbit(sc)
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
struct dc_softc *sc;
{
CSR_WRITE_4(sc, DC_SIO, DC_SIO_ROMCTL_READ|DC_SIO_MII_DIR);
CSR_READ_4(sc, DC_SIO);
DC_SETBIT(sc, DC_SIO, DC_SIO_MII_CLK);
DC_CLRBIT(sc, DC_SIO, DC_SIO_MII_CLK);
if (CSR_READ_4(sc, DC_SIO) & DC_SIO_MII_DATAIN)
return(1);
return(0);
}
/*
* Sync the PHYs by setting data bit and strobing the clock 32 times.
*/
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static void
dc_mii_sync(sc)
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
struct dc_softc *sc;
{
register int i;
CSR_WRITE_4(sc, DC_SIO, DC_SIO_ROMCTL_WRITE);
for (i = 0; i < 32; i++)
dc_mii_writebit(sc, 1);
return;
}
/*
* Clock a series of bits through the MII.
*/
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static void
dc_mii_send(sc, bits, cnt)
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
struct dc_softc *sc;
u_int32_t bits;
int cnt;
{
int i;
for (i = (0x1 << (cnt - 1)); i; i >>= 1)
dc_mii_writebit(sc, bits & i);
}
/*
* Read an PHY register through the MII.
*/
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static int
dc_mii_readreg(sc, frame)
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
struct dc_softc *sc;
struct dc_mii_frame *frame;
{
First round of converting network drivers from spls to mutexes. This takes care of all the 10/100 and gigE PCI drivers that I've done. Next will be the wireless drivers, then the USB ones. I may pick up some stragglers along the way. I'm sort of playing this by ear: if anyone spots any places where I've screwed up horribly, please let me know.
2000-10-13 17:54:19 +00:00
int i, ack;
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
First round of converting network drivers from spls to mutexes. This takes care of all the 10/100 and gigE PCI drivers that I've done. Next will be the wireless drivers, then the USB ones. I may pick up some stragglers along the way. I'm sort of playing this by ear: if anyone spots any places where I've screwed up horribly, please let me know.
2000-10-13 17:54:19 +00:00
DC_LOCK(sc);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
/*
* Set up frame for RX.
*/
frame->mii_stdelim = DC_MII_STARTDELIM;
frame->mii_opcode = DC_MII_READOP;
frame->mii_turnaround = 0;
frame->mii_data = 0;
/*
* Sync the PHYs.
*/
dc_mii_sync(sc);
/*
* Send command/address info.
*/
dc_mii_send(sc, frame->mii_stdelim, 2);
dc_mii_send(sc, frame->mii_opcode, 2);
dc_mii_send(sc, frame->mii_phyaddr, 5);
dc_mii_send(sc, frame->mii_regaddr, 5);
#ifdef notdef
/* Idle bit */
dc_mii_writebit(sc, 1);
dc_mii_writebit(sc, 0);
#endif
/* Check for ack */
ack = dc_mii_readbit(sc);
/*
* Now try reading data bits. If the ack failed, we still
* need to clock through 16 cycles to keep the PHY(s) in sync.
*/
if (ack) {
for(i = 0; i < 16; i++) {
dc_mii_readbit(sc);
}
goto fail;
}
for (i = 0x8000; i; i >>= 1) {
if (!ack) {
if (dc_mii_readbit(sc))
frame->mii_data |= i;
}
}
fail:
dc_mii_writebit(sc, 0);
dc_mii_writebit(sc, 0);
First round of converting network drivers from spls to mutexes. This takes care of all the 10/100 and gigE PCI drivers that I've done. Next will be the wireless drivers, then the USB ones. I may pick up some stragglers along the way. I'm sort of playing this by ear: if anyone spots any places where I've screwed up horribly, please let me know.
2000-10-13 17:54:19 +00:00
DC_UNLOCK(sc);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
if (ack)
return(1);
return(0);
}
/*
* Write to a PHY register through the MII.
*/
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static int
dc_mii_writereg(sc, frame)
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
struct dc_softc *sc;
struct dc_mii_frame *frame;
{
First round of converting network drivers from spls to mutexes. This takes care of all the 10/100 and gigE PCI drivers that I've done. Next will be the wireless drivers, then the USB ones. I may pick up some stragglers along the way. I'm sort of playing this by ear: if anyone spots any places where I've screwed up horribly, please let me know.
2000-10-13 17:54:19 +00:00
DC_LOCK(sc);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
/*
* Set up frame for TX.
*/
frame->mii_stdelim = DC_MII_STARTDELIM;
frame->mii_opcode = DC_MII_WRITEOP;
frame->mii_turnaround = DC_MII_TURNAROUND;
/*
* Sync the PHYs.
*/
dc_mii_sync(sc);
dc_mii_send(sc, frame->mii_stdelim, 2);
dc_mii_send(sc, frame->mii_opcode, 2);
dc_mii_send(sc, frame->mii_phyaddr, 5);
dc_mii_send(sc, frame->mii_regaddr, 5);
dc_mii_send(sc, frame->mii_turnaround, 2);
dc_mii_send(sc, frame->mii_data, 16);
/* Idle bit. */
dc_mii_writebit(sc, 0);
dc_mii_writebit(sc, 0);
First round of converting network drivers from spls to mutexes. This takes care of all the 10/100 and gigE PCI drivers that I've done. Next will be the wireless drivers, then the USB ones. I may pick up some stragglers along the way. I'm sort of playing this by ear: if anyone spots any places where I've screwed up horribly, please let me know.
2000-10-13 17:54:19 +00:00
DC_UNLOCK(sc);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
return(0);
}
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static int
dc_miibus_readreg(dev, phy, reg)
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
device_t dev;
int phy, reg;
{
struct dc_mii_frame frame;
struct dc_softc *sc;
Silence compiler warnings.
2001-01-20 00:07:49 +00:00
int i, rval, phy_reg = 0;
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
sc = device_get_softc(dev);
bzero((char *)&frame, sizeof(frame));
/*
* Note: both the AL981 and AN985 have internal PHYs,
* however the AL981 provides direct access to the PHY
* registers while the AN985 uses a serial MII interface.
* The AN985's MII interface is also buggy in that you
* can read from any MII address (0 to 31), but only address 1
* behaves normally. To deal with both cases, we pretend
* that the PHY is at MII address 1.
*/
if (DC_IS_ADMTEK(sc) && phy != DC_ADMTEK_PHYADDR)
return(0);
Add support for Conexant LANfinity miniPCI controllers. People who have laptops with this chip should test this and report back as I don't have access to this hardware myself. People with -stable systems should try the patch at: http://www.freebsd.org/~wpaul/conexant.patch.gz Submitted by: Phil Kernick <Phil@Kernick.org>
2001-09-04 17:10:11 +00:00
/*
* Note: the ukphy probes of the RS7112 report a PHY at
* MII address 0 (possibly HomePNA?) and 1 (ethernet)
* so we only respond to correct one.
*/
if (DC_IS_CONEXANT(sc) && phy != DC_CONEXANT_PHYADDR)
return(0);
Add support for parsing the media blocks from the SROM on 21143 adapters. This is necessary in order to make this driver work with the built-in ethernet on the alpha Miata machines. These systems have a 21143-PC chip on-board and optional daughtercards with either a 10/100 MII transceiver or a 10baseT/10base2 transceiver. In both cases, you need to twiddle the GPIO bits on the controller in order to turn the transceivers on, and you have to read the media info from the SROM in order to find out what bits to twiddle.
2000-10-05 17:36:14 +00:00
if (sc->dc_pmode != DC_PMODE_MII) {
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
if (phy == (MII_NPHY - 1)) {
switch(reg) {
case MII_BMSR:
/*
* Fake something to make the probe
* code think there's a PHY here.
*/
return(BMSR_MEDIAMASK);
break;
case MII_PHYIDR1:
if (DC_IS_PNIC(sc))
return(DC_VENDORID_LO);
return(DC_VENDORID_DEC);
break;
case MII_PHYIDR2:
if (DC_IS_PNIC(sc))
return(DC_DEVICEID_82C168);
return(DC_DEVICEID_21143);
break;
default:
return(0);
break;
}
} else
return(0);
}
if (DC_IS_PNIC(sc)) {
CSR_WRITE_4(sc, DC_PN_MII, DC_PN_MIIOPCODE_READ |
(phy << 23) | (reg << 18));
for (i = 0; i < DC_TIMEOUT; i++) {
DELAY(1);
rval = CSR_READ_4(sc, DC_PN_MII);
if (!(rval & DC_PN_MII_BUSY)) {
rval &= 0xFFFF;
return(rval == 0xFFFF ? 0 : rval);
}
}
return(0);
}
if (DC_IS_COMET(sc)) {
switch(reg) {
case MII_BMCR:
phy_reg = DC_AL_BMCR;
break;
case MII_BMSR:
phy_reg = DC_AL_BMSR;
break;
case MII_PHYIDR1:
phy_reg = DC_AL_VENID;
break;
case MII_PHYIDR2:
phy_reg = DC_AL_DEVID;
break;
case MII_ANAR:
phy_reg = DC_AL_ANAR;
break;
case MII_ANLPAR:
phy_reg = DC_AL_LPAR;
break;
case MII_ANER:
phy_reg = DC_AL_ANER;
break;
default:
printf("dc%d: phy_read: bad phy register %x\n",
sc->dc_unit, reg);
return(0);
break;
}
rval = CSR_READ_4(sc, phy_reg) & 0x0000FFFF;
if (rval == 0xFFFF)
return(0);
return(rval);
}
frame.mii_phyaddr = phy;
frame.mii_regaddr = reg;
Bug fixes that I've put together while working on a project in the office: if_vr: handle the case where vr_encap() returns failure: bust out of the packet sending loop instead of panicking. Also add some missing newlines to some printf()s. if_dc: The miibus_read and miibus_write methods keep swapping in and out of MII mode by fiddling with CSR6 for cards with MII PHYs. This is a hack to support the original Macronix 98713 card which has built-in NWAY that uses an MII-like management interface even though it uses serial transceivers. Conditionalize this so that we only do this on 98713 chips, since it does bad things to genuine tulip chips (and maybe other clones).
2001-01-19 23:55:07 +00:00
if (sc->dc_type == DC_TYPE_98713) {
phy_reg = CSR_READ_4(sc, DC_NETCFG);
CSR_WRITE_4(sc, DC_NETCFG, phy_reg & ~DC_NETCFG_PORTSEL);
}
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
dc_mii_readreg(sc, &frame);
Bug fixes that I've put together while working on a project in the office: if_vr: handle the case where vr_encap() returns failure: bust out of the packet sending loop instead of panicking. Also add some missing newlines to some printf()s. if_dc: The miibus_read and miibus_write methods keep swapping in and out of MII mode by fiddling with CSR6 for cards with MII PHYs. This is a hack to support the original Macronix 98713 card which has built-in NWAY that uses an MII-like management interface even though it uses serial transceivers. Conditionalize this so that we only do this on 98713 chips, since it does bad things to genuine tulip chips (and maybe other clones).
2001-01-19 23:55:07 +00:00
if (sc->dc_type == DC_TYPE_98713)
CSR_WRITE_4(sc, DC_NETCFG, phy_reg);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
return(frame.mii_data);
}
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static int
dc_miibus_writereg(dev, phy, reg, data)
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
device_t dev;
int phy, reg, data;
{
struct dc_softc *sc;
struct dc_mii_frame frame;
Silence compiler warnings.
2001-01-20 00:07:49 +00:00
int i, phy_reg = 0;
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
sc = device_get_softc(dev);
bzero((char *)&frame, sizeof(frame));
if (DC_IS_ADMTEK(sc) && phy != DC_ADMTEK_PHYADDR)
return(0);
Add support for Conexant LANfinity miniPCI controllers. People who have laptops with this chip should test this and report back as I don't have access to this hardware myself. People with -stable systems should try the patch at: http://www.freebsd.org/~wpaul/conexant.patch.gz Submitted by: Phil Kernick <Phil@Kernick.org>
2001-09-04 17:10:11 +00:00
if (DC_IS_CONEXANT(sc) && phy != DC_CONEXANT_PHYADDR)
return(0);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
if (DC_IS_PNIC(sc)) {
CSR_WRITE_4(sc, DC_PN_MII, DC_PN_MIIOPCODE_WRITE |
(phy << 23) | (reg << 10) | data);
for (i = 0; i < DC_TIMEOUT; i++) {
if (!(CSR_READ_4(sc, DC_PN_MII) & DC_PN_MII_BUSY))
break;
}
return(0);
}
if (DC_IS_COMET(sc)) {
switch(reg) {
case MII_BMCR:
phy_reg = DC_AL_BMCR;
break;
case MII_BMSR:
phy_reg = DC_AL_BMSR;
break;
case MII_PHYIDR1:
phy_reg = DC_AL_VENID;
break;
case MII_PHYIDR2:
phy_reg = DC_AL_DEVID;
break;
case MII_ANAR:
phy_reg = DC_AL_ANAR;
break;
case MII_ANLPAR:
phy_reg = DC_AL_LPAR;
break;
case MII_ANER:
phy_reg = DC_AL_ANER;
break;
default:
printf("dc%d: phy_write: bad phy register %x\n",
sc->dc_unit, reg);
return(0);
break;
}
CSR_WRITE_4(sc, phy_reg, data);
return(0);
}
frame.mii_phyaddr = phy;
frame.mii_regaddr = reg;
frame.mii_data = data;
Bug fixes that I've put together while working on a project in the office: if_vr: handle the case where vr_encap() returns failure: bust out of the packet sending loop instead of panicking. Also add some missing newlines to some printf()s. if_dc: The miibus_read and miibus_write methods keep swapping in and out of MII mode by fiddling with CSR6 for cards with MII PHYs. This is a hack to support the original Macronix 98713 card which has built-in NWAY that uses an MII-like management interface even though it uses serial transceivers. Conditionalize this so that we only do this on 98713 chips, since it does bad things to genuine tulip chips (and maybe other clones).
2001-01-19 23:55:07 +00:00
if (sc->dc_type == DC_TYPE_98713) {
phy_reg = CSR_READ_4(sc, DC_NETCFG);
CSR_WRITE_4(sc, DC_NETCFG, phy_reg & ~DC_NETCFG_PORTSEL);
}
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
dc_mii_writereg(sc, &frame);
Bug fixes that I've put together while working on a project in the office: if_vr: handle the case where vr_encap() returns failure: bust out of the packet sending loop instead of panicking. Also add some missing newlines to some printf()s. if_dc: The miibus_read and miibus_write methods keep swapping in and out of MII mode by fiddling with CSR6 for cards with MII PHYs. This is a hack to support the original Macronix 98713 card which has built-in NWAY that uses an MII-like management interface even though it uses serial transceivers. Conditionalize this so that we only do this on 98713 chips, since it does bad things to genuine tulip chips (and maybe other clones).
2001-01-19 23:55:07 +00:00
if (sc->dc_type == DC_TYPE_98713)
CSR_WRITE_4(sc, DC_NETCFG, phy_reg);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
return(0);
}
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static void
dc_miibus_statchg(dev)
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
device_t dev;
{
struct dc_softc *sc;
struct mii_data *mii;
Add support for DM9102A boards with Davicom DM9801 HomePNA PHYs.
2000-01-24 17:19:37 +00:00
struct ifmedia *ifm;
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
sc = device_get_softc(dev);
if (DC_IS_ADMTEK(sc))
return;
Add support for parsing the media blocks from the SROM on 21143 adapters. This is necessary in order to make this driver work with the built-in ethernet on the alpha Miata machines. These systems have a 21143-PC chip on-board and optional daughtercards with either a 10/100 MII transceiver or a 10baseT/10base2 transceiver. In both cases, you need to twiddle the GPIO bits on the controller in order to turn the transceivers on, and you have to read the media info from the SROM in order to find out what bits to twiddle.
2000-10-05 17:36:14 +00:00
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
mii = device_get_softc(sc->dc_miibus);
Add support for DM9102A boards with Davicom DM9801 HomePNA PHYs.
2000-01-24 17:19:37 +00:00
ifm = &mii->mii_media;
if (DC_IS_DAVICOM(sc) &&
Move us yet closer to IFM_* definitions in NetBSD.
2002-04-29 05:32:44 +00:00
IFM_SUBTYPE(ifm->ifm_media) == IFM_HPNA_1) {
Add support for DM9102A boards with Davicom DM9801 HomePNA PHYs.
2000-01-24 17:19:37 +00:00
dc_setcfg(sc, ifm->ifm_media);
sc->dc_if_media = ifm->ifm_media;
} else {
dc_setcfg(sc, mii->mii_media_active);
sc->dc_if_media = mii->mii_media_active;
}
return;
}
/*
* Special support for DM9102A cards with HomePNA PHYs. Note:
* with the Davicom DM9102A/DM9801 eval board that I have, it seems
* to be impossible to talk to the management interface of the DM9801
* PHY (its MDIO pin is not connected to anything). Consequently,
* the driver has to just 'know' about the additional mode and deal
* with it itself. *sigh*
*/
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static void
dc_miibus_mediainit(dev)
Add support for DM9102A boards with Davicom DM9801 HomePNA PHYs.
2000-01-24 17:19:37 +00:00
device_t dev;
{
struct dc_softc *sc;
struct mii_data *mii;
struct ifmedia *ifm;
int rev;
rev = pci_read_config(dev, DC_PCI_CFRV, 4) & 0xFF;
sc = device_get_softc(dev);
mii = device_get_softc(sc->dc_miibus);
ifm = &mii->mii_media;
if (DC_IS_DAVICOM(sc) && rev >= DC_REVISION_DM9102A)
Move us yet closer to IFM_* definitions in NetBSD.
2002-04-29 05:32:44 +00:00
ifmedia_add(ifm, IFM_ETHER|IFM_HPNA_1, 0, NULL);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
return;
}
#define DC_POLY 0xEDB88320
Apply patch to the dc driver to handle Macronix MX98715AEC-C/D/E chips, which differ slightly from the Macronix MX98715AEC chip on the sample adapter that I have in that the multicast hash table is only 128 bits wide instead of 512. New adapters are popping up with this chip, and due to improper handling of the smaller hash table, broadcast packets were not being received correctly.
2000-07-15 17:54:30 +00:00
#define DC_BITS_512 9
#define DC_BITS_128 7
#define DC_BITS_64 6
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static u_int32_t
dc_crc_le(sc, addr)
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
struct dc_softc *sc;
caddr_t addr;
{
u_int32_t idx, bit, data, crc;
/* Compute CRC for the address value. */
crc = 0xFFFFFFFF; /* initial value */
for (idx = 0; idx < 6; idx++) {
for (data = *addr++, bit = 0; bit < 8; bit++, data >>= 1)
crc = (crc >> 1) ^ (((crc ^ data) & 1) ? DC_POLY : 0);
}
Apply patch to the dc driver to handle Macronix MX98715AEC-C/D/E chips, which differ slightly from the Macronix MX98715AEC chip on the sample adapter that I have in that the multicast hash table is only 128 bits wide instead of 512. New adapters are popping up with this chip, and due to improper handling of the smaller hash table, broadcast packets were not being received correctly.
2000-07-15 17:54:30 +00:00
/*
* The hash table on the PNIC II and the MX98715AEC-C/D/E
* chips is only 128 bits wide.
*/
if (sc->dc_flags & DC_128BIT_HASH)
return (crc & ((1 << DC_BITS_128) - 1));
/* The hash table on the MX98715BEC is only 64 bits wide. */
if (sc->dc_flags & DC_64BIT_HASH)
return (crc & ((1 << DC_BITS_64) - 1));
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
NEWCARD/Cardbus - This commit adds support for Xircom X3201 based cardbus cards. Support for the TDK 78Q2120 MII is also added. IBM Etherjet, Intel and Xircom cards uses these chips. Note that as a result of this commit, some Intel/DEC 21143 based cardbus cards will also attach, but not get link. That is being looked at.
2000-10-19 08:34:32 +00:00
/* Xircom's hash filtering table is different (read: weird) */
/* Xircom uses the LEAST significant bits */
if (DC_IS_XIRCOM(sc)) {
if ((crc & 0x180) == 0x180)
return (crc & 0x0F) + (crc & 0x70)*3 + (14 << 4);
else
return (crc & 0x1F) + ((crc>>1) & 0xF0)*3 + (12 << 4);
}
Apply patch to the dc driver to handle Macronix MX98715AEC-C/D/E chips, which differ slightly from the Macronix MX98715AEC chip on the sample adapter that I have in that the multicast hash table is only 128 bits wide instead of 512. New adapters are popping up with this chip, and due to improper handling of the smaller hash table, broadcast packets were not being received correctly.
2000-07-15 17:54:30 +00:00
return (crc & ((1 << DC_BITS_512) - 1));
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
}
/*
* Calculate CRC of a multicast group address, return the lower 6 bits.
*/
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static u_int32_t
dc_crc_be(addr)
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
caddr_t addr;
{
u_int32_t crc, carry;
int i, j;
u_int8_t c;
/* Compute CRC for the address value. */
crc = 0xFFFFFFFF; /* initial value */
for (i = 0; i < 6; i++) {
c = *(addr + i);
for (j = 0; j < 8; j++) {
carry = ((crc & 0x80000000) ? 1 : 0) ^ (c & 0x01);
crc <<= 1;
c >>= 1;
if (carry)
crc = (crc ^ 0x04c11db6) | carry;
}
}
/* return the filter bit position */
return((crc >> 26) & 0x0000003F);
}
/*
* 21143-style RX filter setup routine. Filter programming is done by
* downloading a special setup frame into the TX engine. 21143, Macronix,
* PNIC, PNIC II and Davicom chips are programmed this way.
*
* We always program the chip using 'hash perfect' mode, i.e. one perfect
* address (our node address) and a 512-bit hash filter for multicast
* frames. We also sneak the broadcast address into the hash filter since
* we need that too.
*/
Be consistent about functions being static. Spotted by: FlexeLint.
2002-10-16 09:04:52 +00:00
static void
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
dc_setfilt_21143(sc)
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
struct dc_softc *sc;
{
struct dc_desc *sframe;
u_int32_t h, *sp;
struct ifmultiaddr *ifma;
struct ifnet *ifp;
int i;
ifp = &sc->arpcom.ac_if;
i = sc->dc_cdata.dc_tx_prod;
DC_INC(sc->dc_cdata.dc_tx_prod, DC_TX_LIST_CNT);
sc->dc_cdata.dc_tx_cnt++;
sframe = &sc->dc_ldata->dc_tx_list[i];
sp = (u_int32_t *)&sc->dc_cdata.dc_sbuf;
bzero((char *)sp, DC_SFRAME_LEN);
sframe->dc_data = vtophys(&sc->dc_cdata.dc_sbuf);
sframe->dc_ctl = DC_SFRAME_LEN | DC_TXCTL_SETUP | DC_TXCTL_TLINK |
DC_FILTER_HASHPERF | DC_TXCTL_FINT;
sc->dc_cdata.dc_tx_chain[i] = (struct mbuf *)&sc->dc_cdata.dc_sbuf;
/* If we want promiscuous mode, set the allframes bit. */
if (ifp->if_flags & IFF_PROMISC)
DC_SETBIT(sc, DC_NETCFG, DC_NETCFG_RX_PROMISC);
else
DC_CLRBIT(sc, DC_NETCFG, DC_NETCFG_RX_PROMISC);
if (ifp->if_flags & IFF_ALLMULTI)
DC_SETBIT(sc, DC_NETCFG, DC_NETCFG_RX_ALLMULTI);
else
DC_CLRBIT(sc, DC_NETCFG, DC_NETCFG_RX_ALLMULTI);
Convert if_multiaddrs from LIST to TAILQ so that it can be traversed backwards in the three drivers which want to do that. Reviewed by: mikeh
2001-02-06 10:12:15 +00:00
TAILQ_FOREACH(ifma, &ifp->if_multiaddrs, ifma_link) {
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
if (ifma->ifma_addr->sa_family != AF_LINK)
continue;
h = dc_crc_le(sc,
LLADDR((struct sockaddr_dl *)ifma->ifma_addr));
sp[h >> 4] |= 1 << (h & 0xF);
}
if (ifp->if_flags & IFF_BROADCAST) {
h = dc_crc_le(sc, (caddr_t)&etherbroadcastaddr);
sp[h >> 4] |= 1 << (h & 0xF);
}
/* Set our MAC address */
sp[39] = ((u_int16_t *)sc->arpcom.ac_enaddr)[0];
sp[40] = ((u_int16_t *)sc->arpcom.ac_enaddr)[1];
sp[41] = ((u_int16_t *)sc->arpcom.ac_enaddr)[2];
sframe->dc_status = DC_TXSTAT_OWN;
CSR_WRITE_4(sc, DC_TXSTART, 0xFFFFFFFF);
/*
* The PNIC takes an exceedingly long time to process its
* setup frame; wait 10ms after posting the setup frame
* before proceeding, just so it has time to swallow its
* medicine.
*/
DELAY(10000);
ifp->if_timer = 5;
return;
}
Be consistent about functions being static. Spotted by: FlexeLint.
2002-10-16 09:04:52 +00:00
static void
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
dc_setfilt_admtek(sc)
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
struct dc_softc *sc;
{
struct ifnet *ifp;
int h = 0;
u_int32_t hashes[2] = { 0, 0 };
struct ifmultiaddr *ifma;
ifp = &sc->arpcom.ac_if;
/* Init our MAC address */
CSR_WRITE_4(sc, DC_AL_PAR0, *(u_int32_t *)(&sc->arpcom.ac_enaddr[0]));
CSR_WRITE_4(sc, DC_AL_PAR1, *(u_int32_t *)(&sc->arpcom.ac_enaddr[4]));
/* If we want promiscuous mode, set the allframes bit. */
if (ifp->if_flags & IFF_PROMISC)
DC_SETBIT(sc, DC_NETCFG, DC_NETCFG_RX_PROMISC);
else
DC_CLRBIT(sc, DC_NETCFG, DC_NETCFG_RX_PROMISC);
if (ifp->if_flags & IFF_ALLMULTI)
DC_SETBIT(sc, DC_NETCFG, DC_NETCFG_RX_ALLMULTI);
else
DC_CLRBIT(sc, DC_NETCFG, DC_NETCFG_RX_ALLMULTI);
/* first, zot all the existing hash bits */
CSR_WRITE_4(sc, DC_AL_MAR0, 0);
CSR_WRITE_4(sc, DC_AL_MAR1, 0);
/*
* If we're already in promisc or allmulti mode, we
* don't have to bother programming the multicast filter.
*/
if (ifp->if_flags & (IFF_PROMISC|IFF_ALLMULTI))
return;
/* now program new ones */
Convert if_multiaddrs from LIST to TAILQ so that it can be traversed backwards in the three drivers which want to do that. Reviewed by: mikeh
2001-02-06 10:12:15 +00:00
TAILQ_FOREACH(ifma, &ifp->if_multiaddrs, ifma_link) {
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
if (ifma->ifma_addr->sa_family != AF_LINK)
continue;
h = dc_crc_be(LLADDR((struct sockaddr_dl *)ifma->ifma_addr));
if (h < 32)
hashes[0] |= (1 << h);
else
hashes[1] |= (1 << (h - 32));
}
CSR_WRITE_4(sc, DC_AL_MAR0, hashes[0]);
CSR_WRITE_4(sc, DC_AL_MAR1, hashes[1]);
return;
}
Be consistent about functions being static. Spotted by: FlexeLint.
2002-10-16 09:04:52 +00:00
static void
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
dc_setfilt_asix(sc)
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
struct dc_softc *sc;
{
struct ifnet *ifp;
int h = 0;
u_int32_t hashes[2] = { 0, 0 };
struct ifmultiaddr *ifma;
ifp = &sc->arpcom.ac_if;
Make consistent; turn spaces into tabs where there is a mixture.
2002-09-04 18:14:17 +00:00
/* Init our MAC address */
CSR_WRITE_4(sc, DC_AX_FILTIDX, DC_AX_FILTIDX_PAR0);
CSR_WRITE_4(sc, DC_AX_FILTDATA,
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
*(u_int32_t *)(&sc->arpcom.ac_enaddr[0]));
Make consistent; turn spaces into tabs where there is a mixture.
2002-09-04 18:14:17 +00:00
CSR_WRITE_4(sc, DC_AX_FILTIDX, DC_AX_FILTIDX_PAR1);
CSR_WRITE_4(sc, DC_AX_FILTDATA,
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
*(u_int32_t *)(&sc->arpcom.ac_enaddr[4]));
/* If we want promiscuous mode, set the allframes bit. */
if (ifp->if_flags & IFF_PROMISC)
DC_SETBIT(sc, DC_NETCFG, DC_NETCFG_RX_PROMISC);
else
DC_CLRBIT(sc, DC_NETCFG, DC_NETCFG_RX_PROMISC);
if (ifp->if_flags & IFF_ALLMULTI)
DC_SETBIT(sc, DC_NETCFG, DC_NETCFG_RX_ALLMULTI);
else
DC_CLRBIT(sc, DC_NETCFG, DC_NETCFG_RX_ALLMULTI);
/*
* The ASIX chip has a special bit to enable reception
* of broadcast frames.
*/
if (ifp->if_flags & IFF_BROADCAST)
DC_SETBIT(sc, DC_NETCFG, DC_AX_NETCFG_RX_BROAD);
else
DC_CLRBIT(sc, DC_NETCFG, DC_AX_NETCFG_RX_BROAD);
/* first, zot all the existing hash bits */
CSR_WRITE_4(sc, DC_AX_FILTIDX, DC_AX_FILTIDX_MAR0);
CSR_WRITE_4(sc, DC_AX_FILTDATA, 0);
CSR_WRITE_4(sc, DC_AX_FILTIDX, DC_AX_FILTIDX_MAR1);
CSR_WRITE_4(sc, DC_AX_FILTDATA, 0);
/*
* If we're already in promisc or allmulti mode, we
* don't have to bother programming the multicast filter.
*/
if (ifp->if_flags & (IFF_PROMISC|IFF_ALLMULTI))
return;
/* now program new ones */
Convert if_multiaddrs from LIST to TAILQ so that it can be traversed backwards in the three drivers which want to do that. Reviewed by: mikeh
2001-02-06 10:12:15 +00:00
TAILQ_FOREACH(ifma, &ifp->if_multiaddrs, ifma_link) {
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
if (ifma->ifma_addr->sa_family != AF_LINK)
continue;
h = dc_crc_be(LLADDR((struct sockaddr_dl *)ifma->ifma_addr));
if (h < 32)
hashes[0] |= (1 << h);
else
hashes[1] |= (1 << (h - 32));
}
CSR_WRITE_4(sc, DC_AX_FILTIDX, DC_AX_FILTIDX_MAR0);
CSR_WRITE_4(sc, DC_AX_FILTDATA, hashes[0]);
CSR_WRITE_4(sc, DC_AX_FILTIDX, DC_AX_FILTIDX_MAR1);
CSR_WRITE_4(sc, DC_AX_FILTDATA, hashes[1]);
return;
}
Be consistent about functions being static. Spotted by: FlexeLint.
2002-10-16 09:04:52 +00:00
static void
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
dc_setfilt_xircom(sc)
NEWCARD/Cardbus - This commit adds support for Xircom X3201 based cardbus cards. Support for the TDK 78Q2120 MII is also added. IBM Etherjet, Intel and Xircom cards uses these chips. Note that as a result of this commit, some Intel/DEC 21143 based cardbus cards will also attach, but not get link. That is being looked at.
2000-10-19 08:34:32 +00:00
struct dc_softc *sc;
{
struct dc_desc *sframe;
u_int32_t h, *sp;
struct ifmultiaddr *ifma;
struct ifnet *ifp;
int i;
ifp = &sc->arpcom.ac_if;
DC_CLRBIT(sc, DC_NETCFG, (DC_NETCFG_TX_ON|DC_NETCFG_RX_ON));
i = sc->dc_cdata.dc_tx_prod;
DC_INC(sc->dc_cdata.dc_tx_prod, DC_TX_LIST_CNT);
sc->dc_cdata.dc_tx_cnt++;
sframe = &sc->dc_ldata->dc_tx_list[i];
sp = (u_int32_t *)&sc->dc_cdata.dc_sbuf;
bzero((char *)sp, DC_SFRAME_LEN);
sframe->dc_data = vtophys(&sc->dc_cdata.dc_sbuf);
sframe->dc_ctl = DC_SFRAME_LEN | DC_TXCTL_SETUP | DC_TXCTL_TLINK |
DC_FILTER_HASHPERF | DC_TXCTL_FINT;
sc->dc_cdata.dc_tx_chain[i] = (struct mbuf *)&sc->dc_cdata.dc_sbuf;
/* If we want promiscuous mode, set the allframes bit. */
if (ifp->if_flags & IFF_PROMISC)
DC_SETBIT(sc, DC_NETCFG, DC_NETCFG_RX_PROMISC);
else
DC_CLRBIT(sc, DC_NETCFG, DC_NETCFG_RX_PROMISC);
if (ifp->if_flags & IFF_ALLMULTI)
Make consistent; turn spaces into tabs where there is a mixture.
2002-09-04 18:14:17 +00:00
DC_SETBIT(sc, DC_NETCFG, DC_NETCFG_RX_ALLMULTI);
NEWCARD/Cardbus - This commit adds support for Xircom X3201 based cardbus cards. Support for the TDK 78Q2120 MII is also added. IBM Etherjet, Intel and Xircom cards uses these chips. Note that as a result of this commit, some Intel/DEC 21143 based cardbus cards will also attach, but not get link. That is being looked at.
2000-10-19 08:34:32 +00:00
else
DC_CLRBIT(sc, DC_NETCFG, DC_NETCFG_RX_ALLMULTI);
Convert if_multiaddrs from LIST to TAILQ so that it can be traversed backwards in the three drivers which want to do that. Reviewed by: mikeh
2001-02-06 10:12:15 +00:00
TAILQ_FOREACH(ifma, &ifp->if_multiaddrs, ifma_link) {
NEWCARD/Cardbus - This commit adds support for Xircom X3201 based cardbus cards. Support for the TDK 78Q2120 MII is also added. IBM Etherjet, Intel and Xircom cards uses these chips. Note that as a result of this commit, some Intel/DEC 21143 based cardbus cards will also attach, but not get link. That is being looked at.
2000-10-19 08:34:32 +00:00
if (ifma->ifma_addr->sa_family != AF_LINK)
continue;
Add PCI IDs for some additional cardbus cards. Yes, there really is a RealTek 8139 cardbus device. Unfortunately it doesn't quite work yet because the CIS parser barfs on it. Submitted by msmith, with some small tweaks by me.
2000-10-28 09:00:20 +00:00
h = dc_crc_le(sc,
LLADDR((struct sockaddr_dl *)ifma->ifma_addr));
NEWCARD/Cardbus - This commit adds support for Xircom X3201 based cardbus cards. Support for the TDK 78Q2120 MII is also added. IBM Etherjet, Intel and Xircom cards uses these chips. Note that as a result of this commit, some Intel/DEC 21143 based cardbus cards will also attach, but not get link. That is being looked at.
2000-10-19 08:34:32 +00:00
sp[h >> 4] |= 1 << (h & 0xF);
}
if (ifp->if_flags & IFF_BROADCAST) {
h = dc_crc_le(sc, (caddr_t)&etherbroadcastaddr);
sp[h >> 4] |= 1 << (h & 0xF);
}
/* Set our MAC address */
sp[0] = ((u_int16_t *)sc->arpcom.ac_enaddr)[0];
sp[1] = ((u_int16_t *)sc->arpcom.ac_enaddr)[1];
sp[2] = ((u_int16_t *)sc->arpcom.ac_enaddr)[2];
DC_SETBIT(sc, DC_NETCFG, DC_NETCFG_TX_ON);
DC_SETBIT(sc, DC_NETCFG, DC_NETCFG_RX_ON);
ifp->if_flags |= IFF_RUNNING;
sframe->dc_status = DC_TXSTAT_OWN;
CSR_WRITE_4(sc, DC_TXSTART, 0xFFFFFFFF);
/*
* wait some time...
*/
DELAY(1000);
ifp->if_timer = 5;
return;
}
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static void
dc_setfilt(sc)
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
struct dc_softc *sc;
{
if (DC_IS_INTEL(sc) || DC_IS_MACRONIX(sc) || DC_IS_PNIC(sc) ||
Add support for Conexant LANfinity miniPCI controllers. People who have laptops with this chip should test this and report back as I don't have access to this hardware myself. People with -stable systems should try the patch at: http://www.freebsd.org/~wpaul/conexant.patch.gz Submitted by: Phil Kernick <Phil@Kernick.org>
2001-09-04 17:10:11 +00:00
DC_IS_PNICII(sc) || DC_IS_DAVICOM(sc) || DC_IS_CONEXANT(sc))
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
dc_setfilt_21143(sc);
if (DC_IS_ASIX(sc))
dc_setfilt_asix(sc);
if (DC_IS_ADMTEK(sc))
dc_setfilt_admtek(sc);
NEWCARD/Cardbus - This commit adds support for Xircom X3201 based cardbus cards. Support for the TDK 78Q2120 MII is also added. IBM Etherjet, Intel and Xircom cards uses these chips. Note that as a result of this commit, some Intel/DEC 21143 based cardbus cards will also attach, but not get link. That is being looked at.
2000-10-19 08:34:32 +00:00
if (DC_IS_XIRCOM(sc))
dc_setfilt_xircom(sc);
Make consistent; turn spaces into tabs where there is a mixture.
2002-09-04 18:14:17 +00:00
return;
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
}
/*
* In order to fiddle with the
* 'full-duplex' and '100Mbps' bits in the netconfig register, we
* first have to put the transmit and/or receive logic in the idle state.
*/
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static void
dc_setcfg(sc, media)
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
struct dc_softc *sc;
int media;
{
int i, restart = 0;
u_int32_t isr;
if (IFM_SUBTYPE(media) == IFM_NONE)
return;
if (CSR_READ_4(sc, DC_NETCFG) & (DC_NETCFG_TX_ON|DC_NETCFG_RX_ON)) {
restart = 1;
DC_CLRBIT(sc, DC_NETCFG, (DC_NETCFG_TX_ON|DC_NETCFG_RX_ON));
for (i = 0; i < DC_TIMEOUT; i++) {
isr = CSR_READ_4(sc, DC_ISR);
Apply patch to allow TX underrun handling without issuing a complete chip reset. Just temporarily turn off the transmitter instead. Submitted by: Stephen McKay <mckay@freebsd.org>
2001-02-22 19:26:55 +00:00
if (isr & DC_ISR_TX_IDLE &&
Fix the support for the AN985/983 chips, which do not set the RXSTATE to STOPPED, but to WAIT. This should fix hangs which could only be solved by replugging the cable. Submitted by: jhb Reviewed by: phk MFC after: 2 weeks
2002-09-20 15:18:13 +00:00
((isr & DC_ISR_RX_STATE) == DC_RXSTATE_STOPPED ||
(isr & DC_ISR_RX_STATE) == DC_RXSTATE_WAIT))
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
break;
Apply patch to allow TX underrun handling without issuing a complete chip reset. Just temporarily turn off the transmitter instead. Submitted by: Stephen McKay <mckay@freebsd.org>
2001-02-22 19:26:55 +00:00
DELAY(10);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
}
if (i == DC_TIMEOUT)
printf("dc%d: failed to force tx and "
"rx to idle state\n", sc->dc_unit);
}
if (IFM_SUBTYPE(media) == IFM_100_TX) {
When I tweaked if_dc.c to alter the polling interval for non-MII 21143 chips, I accidentally removed the DC_MII_REDUCED_POLL flag for all 21143 cards. This caused problems with timer-instigated TCP retransmits, which happened to occur at the same time as an MII poll tick on MII-based cards (e.g. D-Link DFE-570TX). Fixed this, plus made some other cleanups. The autoneg fixes for the non-MII cards still work. Also tested the PNIC II now that I have one again.
2000-06-07 17:07:44 +00:00
DC_CLRBIT(sc, DC_NETCFG, DC_NETCFG_SPEEDSEL);
DC_SETBIT(sc, DC_NETCFG, DC_NETCFG_HEARTBEAT);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
if (sc->dc_pmode == DC_PMODE_MII) {
Apply patch supplied by John Hood <jhood@sitaranetworks.com> to fix problems with LEDs on some cards being stomped on when clearing the "jabber disable" bit. Using DC_SETBIT() has an unwanted side effect of setting a write enable bit in the watchdog timer register which we really want to be cleared when we do a write.
2000-08-01 19:34:13 +00:00
int watchdogreg;
Close PR 20438. Make fix for preserving LED settings conditional on presence Intel 21143 chip.
2000-08-07 17:03:20 +00:00
if (DC_IS_INTEL(sc)) {
Apply patch supplied by John Hood <jhood@sitaranetworks.com> to fix problems with LEDs on some cards being stomped on when clearing the "jabber disable" bit. Using DC_SETBIT() has an unwanted side effect of setting a write enable bit in the watchdog timer register which we really want to be cleared when we do a write.
2000-08-01 19:34:13 +00:00
/* there's a write enable bit here that reads as 1 */
Close PR 20438. Make fix for preserving LED settings conditional on presence Intel 21143 chip.
2000-08-07 17:03:20 +00:00
watchdogreg = CSR_READ_4(sc, DC_WATCHDOG);
watchdogreg &= ~DC_WDOG_CTLWREN;
watchdogreg |= DC_WDOG_JABBERDIS;
Clean up a few things in dc_setcfg() pointed out to be me by aaron@openbsd.com on IRC earlier today.
2000-10-14 00:40:14 +00:00
CSR_WRITE_4(sc, DC_WATCHDOG, watchdogreg);
Close PR 20438. Make fix for preserving LED settings conditional on presence Intel 21143 chip.
2000-08-07 17:03:20 +00:00
} else {
DC_SETBIT(sc, DC_WATCHDOG, DC_WDOG_JABBERDIS);
}
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
DC_CLRBIT(sc, DC_NETCFG, (DC_NETCFG_PCS|
DC_NETCFG_PORTSEL|DC_NETCFG_SCRAMBLER));
if (sc->dc_type == DC_TYPE_98713)
DC_SETBIT(sc, DC_NETCFG, (DC_NETCFG_PCS|
DC_NETCFG_SCRAMBLER));
Add support for the Davicom DM9102A 10/100 ethernet controller chip. This is just to make sure we initialize the chip correctly: we need to make the sure the port select bit in CSR6 is set properly so that we use the internal PHY for 10/100 support. (The eval boards I have also include an external HomePNA PHY, but I need to play with that more before I can support it.)
2000-01-19 19:03:08 +00:00
if (!DC_IS_DAVICOM(sc))
DC_SETBIT(sc, DC_NETCFG, DC_NETCFG_PORTSEL);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
DC_CLRBIT(sc, DC_10BTCTRL, 0xFFFF);
Call dc_apply_fixup() in dc_setcfg() for the MII case.
2000-10-31 00:06:39 +00:00
if (DC_IS_INTEL(sc))
dc_apply_fixup(sc, IFM_AUTO);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
} else {
if (DC_IS_PNIC(sc)) {
DC_PN_GPIO_SETBIT(sc, DC_PN_GPIO_SPEEDSEL);
DC_PN_GPIO_SETBIT(sc, DC_PN_GPIO_100TX_LOOP);
DC_SETBIT(sc, DC_PN_NWAY, DC_PN_NWAY_SPEEDSEL);
}
Rework the support for the internal autonegotiation on the 21143 and workalike chips (Macronix 98713A/98715 and PNIC II). Timing is somewhat critical: you need to bring the link as soon as possible after NWAY is done, and the old one second polling interval was too long. Now we poll every 10th of a second until NWAY completes (at which point we return to the 1 second interval again to keep an eye on the link state). I tested all the other cards I had on hand to make sure I didn't bust any of them and they seem to work (including the MII-based 21143 card). This should fix some autoneg problems with DE500-BA cards and the built-in 10/100 ethernet on some alpha systems. (Now before anyone asks why I never noticed this before, the old code worked just find with the Intel swich I used for testing back in NY. Apparently not all switches are as picky about the timing.)
2000-05-31 05:40:53 +00:00
DC_SETBIT(sc, DC_NETCFG, DC_NETCFG_PORTSEL);
DC_SETBIT(sc, DC_NETCFG, DC_NETCFG_PCS);
DC_SETBIT(sc, DC_NETCFG, DC_NETCFG_SCRAMBLER);
Add support for parsing the media blocks from the SROM on 21143 adapters. This is necessary in order to make this driver work with the built-in ethernet on the alpha Miata machines. These systems have a 21143-PC chip on-board and optional daughtercards with either a 10/100 MII transceiver or a 10baseT/10base2 transceiver. In both cases, you need to twiddle the GPIO bits on the controller in order to turn the transceivers on, and you have to read the media info from the SROM in order to find out what bits to twiddle.
2000-10-05 17:36:14 +00:00
if (DC_IS_INTEL(sc))
dc_apply_fixup(sc,
(media & IFM_GMASK) == IFM_FDX ?
IFM_100_TX|IFM_FDX : IFM_100_TX);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
}
}
if (IFM_SUBTYPE(media) == IFM_10_T) {
When I tweaked if_dc.c to alter the polling interval for non-MII 21143 chips, I accidentally removed the DC_MII_REDUCED_POLL flag for all 21143 cards. This caused problems with timer-instigated TCP retransmits, which happened to occur at the same time as an MII poll tick on MII-based cards (e.g. D-Link DFE-570TX). Fixed this, plus made some other cleanups. The autoneg fixes for the non-MII cards still work. Also tested the PNIC II now that I have one again.
2000-06-07 17:07:44 +00:00
DC_SETBIT(sc, DC_NETCFG, DC_NETCFG_SPEEDSEL);
DC_CLRBIT(sc, DC_NETCFG, DC_NETCFG_HEARTBEAT);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
if (sc->dc_pmode == DC_PMODE_MII) {
Apply patch supplied by John Hood <jhood@sitaranetworks.com> to fix problems with LEDs on some cards being stomped on when clearing the "jabber disable" bit. Using DC_SETBIT() has an unwanted side effect of setting a write enable bit in the watchdog timer register which we really want to be cleared when we do a write.
2000-08-01 19:34:13 +00:00
int watchdogreg;
/* there's a write enable bit here that reads as 1 */
Clean up a few things in dc_setcfg() pointed out to be me by aaron@openbsd.com on IRC earlier today.
2000-10-14 00:40:14 +00:00
if (DC_IS_INTEL(sc)) {
watchdogreg = CSR_READ_4(sc, DC_WATCHDOG);
watchdogreg &= ~DC_WDOG_CTLWREN;
watchdogreg |= DC_WDOG_JABBERDIS;
CSR_WRITE_4(sc, DC_WATCHDOG, watchdogreg);
} else {
DC_SETBIT(sc, DC_WATCHDOG, DC_WDOG_JABBERDIS);
}
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
DC_CLRBIT(sc, DC_NETCFG, (DC_NETCFG_PCS|
DC_NETCFG_PORTSEL|DC_NETCFG_SCRAMBLER));
if (sc->dc_type == DC_TYPE_98713)
DC_SETBIT(sc, DC_NETCFG, DC_NETCFG_PCS);
Add support for the Davicom DM9102A 10/100 ethernet controller chip. This is just to make sure we initialize the chip correctly: we need to make the sure the port select bit in CSR6 is set properly so that we use the internal PHY for 10/100 support. (The eval boards I have also include an external HomePNA PHY, but I need to play with that more before I can support it.)
2000-01-19 19:03:08 +00:00
if (!DC_IS_DAVICOM(sc))
DC_SETBIT(sc, DC_NETCFG, DC_NETCFG_PORTSEL);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
DC_CLRBIT(sc, DC_10BTCTRL, 0xFFFF);
Call dc_apply_fixup() in dc_setcfg() for the MII case.
2000-10-31 00:06:39 +00:00
if (DC_IS_INTEL(sc))
dc_apply_fixup(sc, IFM_AUTO);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
} else {
if (DC_IS_PNIC(sc)) {
DC_PN_GPIO_CLRBIT(sc, DC_PN_GPIO_SPEEDSEL);
DC_PN_GPIO_SETBIT(sc, DC_PN_GPIO_100TX_LOOP);
DC_CLRBIT(sc, DC_PN_NWAY, DC_PN_NWAY_SPEEDSEL);
}
DC_CLRBIT(sc, DC_NETCFG, DC_NETCFG_PORTSEL);
Rework the support for the internal autonegotiation on the 21143 and workalike chips (Macronix 98713A/98715 and PNIC II). Timing is somewhat critical: you need to bring the link as soon as possible after NWAY is done, and the old one second polling interval was too long. Now we poll every 10th of a second until NWAY completes (at which point we return to the 1 second interval again to keep an eye on the link state). I tested all the other cards I had on hand to make sure I didn't bust any of them and they seem to work (including the MII-based 21143 card). This should fix some autoneg problems with DE500-BA cards and the built-in 10/100 ethernet on some alpha systems. (Now before anyone asks why I never noticed this before, the old code worked just find with the Intel swich I used for testing back in NY. Apparently not all switches are as picky about the timing.)
2000-05-31 05:40:53 +00:00
DC_CLRBIT(sc, DC_NETCFG, DC_NETCFG_PCS);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
DC_CLRBIT(sc, DC_NETCFG, DC_NETCFG_SCRAMBLER);
Add support for parsing the media blocks from the SROM on 21143 adapters. This is necessary in order to make this driver work with the built-in ethernet on the alpha Miata machines. These systems have a 21143-PC chip on-board and optional daughtercards with either a 10/100 MII transceiver or a 10baseT/10base2 transceiver. In both cases, you need to twiddle the GPIO bits on the controller in order to turn the transceivers on, and you have to read the media info from the SROM in order to find out what bits to twiddle.
2000-10-05 17:36:14 +00:00
if (DC_IS_INTEL(sc)) {
DC_CLRBIT(sc, DC_SIARESET, DC_SIA_RESET);
DC_CLRBIT(sc, DC_10BTCTRL, 0xFFFF);
if ((media & IFM_GMASK) == IFM_FDX)
DC_SETBIT(sc, DC_10BTCTRL, 0x7F3D);
else
DC_SETBIT(sc, DC_10BTCTRL, 0x7F3F);
DC_SETBIT(sc, DC_SIARESET, DC_SIA_RESET);
DC_CLRBIT(sc, DC_10BTCTRL,
DC_TCTL_AUTONEGENBL);
dc_apply_fixup(sc,
(media & IFM_GMASK) == IFM_FDX ?
IFM_10_T|IFM_FDX : IFM_10_T);
DELAY(20000);
}
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
}
}
Add support for DM9102A boards with Davicom DM9801 HomePNA PHYs.
2000-01-24 17:19:37 +00:00
/*
* If this is a Davicom DM9102A card with a DM9801 HomePNA
* PHY and we want HomePNA mode, set the portsel bit to turn
* on the external MII port.
*/
if (DC_IS_DAVICOM(sc)) {
Move us yet closer to IFM_* definitions in NetBSD.
2002-04-29 05:32:44 +00:00
if (IFM_SUBTYPE(media) == IFM_HPNA_1) {
Add support for DM9102A boards with Davicom DM9801 HomePNA PHYs.
2000-01-24 17:19:37 +00:00
DC_SETBIT(sc, DC_NETCFG, DC_NETCFG_PORTSEL);
sc->dc_link = 1;
} else {
DC_CLRBIT(sc, DC_NETCFG, DC_NETCFG_PORTSEL);
}
}
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
if ((media & IFM_GMASK) == IFM_FDX) {
DC_SETBIT(sc, DC_NETCFG, DC_NETCFG_FULLDUPLEX);
if (sc->dc_pmode == DC_PMODE_SYM && DC_IS_PNIC(sc))
DC_SETBIT(sc, DC_PN_NWAY, DC_PN_NWAY_DUPLEX);
} else {
DC_CLRBIT(sc, DC_NETCFG, DC_NETCFG_FULLDUPLEX);
if (sc->dc_pmode == DC_PMODE_SYM && DC_IS_PNIC(sc))
DC_CLRBIT(sc, DC_PN_NWAY, DC_PN_NWAY_DUPLEX);
}
if (restart)
DC_SETBIT(sc, DC_NETCFG, DC_NETCFG_TX_ON|DC_NETCFG_RX_ON);
return;
}
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static void
dc_reset(sc)
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
struct dc_softc *sc;
{
register int i;
DC_SETBIT(sc, DC_BUSCTL, DC_BUSCTL_RESET);
for (i = 0; i < DC_TIMEOUT; i++) {
DELAY(10);
if (!(CSR_READ_4(sc, DC_BUSCTL) & DC_BUSCTL_RESET))
break;
}
Add support for Conexant LANfinity miniPCI controllers. People who have laptops with this chip should test this and report back as I don't have access to this hardware myself. People with -stable systems should try the patch at: http://www.freebsd.org/~wpaul/conexant.patch.gz Submitted by: Phil Kernick <Phil@Kernick.org>
2001-09-04 17:10:11 +00:00
if (DC_IS_ASIX(sc) || DC_IS_ADMTEK(sc) || DC_IS_CONEXANT(sc) ||
Add PCI IDs for some additional cardbus cards. Yes, there really is a RealTek 8139 cardbus device. Unfortunately it doesn't quite work yet because the CIS parser barfs on it. Submitted by msmith, with some small tweaks by me.
2000-10-28 09:00:20 +00:00
DC_IS_XIRCOM(sc) || DC_IS_INTEL(sc)) {
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
DELAY(10000);
DC_CLRBIT(sc, DC_BUSCTL, DC_BUSCTL_RESET);
i = 0;
}
if (i == DC_TIMEOUT)
printf("dc%d: reset never completed!\n", sc->dc_unit);
/* Wait a little while for the chip to get its brains in order. */
DELAY(1000);
CSR_WRITE_4(sc, DC_IMR, 0x00000000);
CSR_WRITE_4(sc, DC_BUSCTL, 0x00000000);
CSR_WRITE_4(sc, DC_NETCFG, 0x00000000);
Fix some problems reported by Mike Pritchard: - Add a flag DC_TX_INTR_ALWAYS which causes the transmit code to request a TX done interrupt for every packet. The PNIC seems to need this to insure that the sent TX buffers get reaped in a timely fashion. - Try to unreset the SIA as soon as possible after resetting the whole chip. - Change dcphy to support either 10/100 or 10Mbps only NICs. The built-in 21143 ethernet in Compaq Presario machines is 10Mbps only and it doesn't work right if we try to advertise 100Mbps modes during autoneg. When restricted to only 10mbps modes, it works fine. Note that for now, I detect this condition by checking the PCI subsystem ID on this NIC (which has a Compaq vendor/device ID). Yes, I know that's what the SROM is supposed to be for. I'm deliberately ignoring the SROM wherever possible. Sue me. The latter two fixes allow if_dc to work correctly with the built-in ethernet on certain Compaq Presario boxes. There are liable to be quite a few people using these as their home systems who might want to try FreeBSD; may as well be nice to them. Now if anybody out there has an Alpha miata with 10Mbps ethernet and can show me the output from pciconf -l on their system, I'd be grateful.
1999-12-13 21:45:13 +00:00
/*
* Bring the SIA out of reset. In some cases, it looks
* like failing to unreset the SIA soon enough gets it
* into a state where it will never come out of reset
* until we reset the whole chip again.
*/
Add support for parsing the media blocks from the SROM on 21143 adapters. This is necessary in order to make this driver work with the built-in ethernet on the alpha Miata machines. These systems have a 21143-PC chip on-board and optional daughtercards with either a 10/100 MII transceiver or a 10baseT/10base2 transceiver. In both cases, you need to twiddle the GPIO bits on the controller in order to turn the transceivers on, and you have to read the media info from the SROM in order to find out what bits to twiddle.
2000-10-05 17:36:14 +00:00
if (DC_IS_INTEL(sc)) {
Fix some problems reported by Mike Pritchard: - Add a flag DC_TX_INTR_ALWAYS which causes the transmit code to request a TX done interrupt for every packet. The PNIC seems to need this to insure that the sent TX buffers get reaped in a timely fashion. - Try to unreset the SIA as soon as possible after resetting the whole chip. - Change dcphy to support either 10/100 or 10Mbps only NICs. The built-in 21143 ethernet in Compaq Presario machines is 10Mbps only and it doesn't work right if we try to advertise 100Mbps modes during autoneg. When restricted to only 10mbps modes, it works fine. Note that for now, I detect this condition by checking the PCI subsystem ID on this NIC (which has a Compaq vendor/device ID). Yes, I know that's what the SROM is supposed to be for. I'm deliberately ignoring the SROM wherever possible. Sue me. The latter two fixes allow if_dc to work correctly with the built-in ethernet on certain Compaq Presario boxes. There are liable to be quite a few people using these as their home systems who might want to try FreeBSD; may as well be nice to them. Now if anybody out there has an Alpha miata with 10Mbps ethernet and can show me the output from pciconf -l on their system, I'd be grateful.
1999-12-13 21:45:13 +00:00
DC_SETBIT(sc, DC_SIARESET, DC_SIA_RESET);
Add support for parsing the media blocks from the SROM on 21143 adapters. This is necessary in order to make this driver work with the built-in ethernet on the alpha Miata machines. These systems have a 21143-PC chip on-board and optional daughtercards with either a 10/100 MII transceiver or a 10baseT/10base2 transceiver. In both cases, you need to twiddle the GPIO bits on the controller in order to turn the transceivers on, and you have to read the media info from the SROM in order to find out what bits to twiddle.
2000-10-05 17:36:14 +00:00
CSR_WRITE_4(sc, DC_10BTCTRL, 0);
CSR_WRITE_4(sc, DC_WATCHDOG, 0);
}
Fix some problems reported by Mike Pritchard: - Add a flag DC_TX_INTR_ALWAYS which causes the transmit code to request a TX done interrupt for every packet. The PNIC seems to need this to insure that the sent TX buffers get reaped in a timely fashion. - Try to unreset the SIA as soon as possible after resetting the whole chip. - Change dcphy to support either 10/100 or 10Mbps only NICs. The built-in 21143 ethernet in Compaq Presario machines is 10Mbps only and it doesn't work right if we try to advertise 100Mbps modes during autoneg. When restricted to only 10mbps modes, it works fine. Note that for now, I detect this condition by checking the PCI subsystem ID on this NIC (which has a Compaq vendor/device ID). Yes, I know that's what the SROM is supposed to be for. I'm deliberately ignoring the SROM wherever possible. Sue me. The latter two fixes allow if_dc to work correctly with the built-in ethernet on certain Compaq Presario boxes. There are liable to be quite a few people using these as their home systems who might want to try FreeBSD; may as well be nice to them. Now if anybody out there has an Alpha miata with 10Mbps ethernet and can show me the output from pciconf -l on their system, I'd be grateful.
1999-12-13 21:45:13 +00:00
Make consistent; turn spaces into tabs where there is a mixture.
2002-09-04 18:14:17 +00:00
return;
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
}
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static struct dc_type *
dc_devtype(dev)
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
device_t dev;
{
struct dc_type *t;
u_int32_t rev;
t = dc_devs;
while(t->dc_name != NULL) {
if ((pci_get_vendor(dev) == t->dc_vid) &&
(pci_get_device(dev) == t->dc_did)) {
/* Check the PCI revision */
rev = pci_read_config(dev, DC_PCI_CFRV, 4) & 0xFF;
if (t->dc_did == DC_DEVICEID_98713 &&
rev >= DC_REVISION_98713A)
t++;
if (t->dc_did == DC_DEVICEID_98713_CP &&
rev >= DC_REVISION_98713A)
t++;
Apply patch to the dc driver to handle Macronix MX98715AEC-C/D/E chips, which differ slightly from the Macronix MX98715AEC chip on the sample adapter that I have in that the multicast hash table is only 128 bits wide instead of 512. New adapters are popping up with this chip, and due to improper handling of the smaller hash table, broadcast packets were not being received correctly.
2000-07-15 17:54:30 +00:00
if (t->dc_did == DC_DEVICEID_987x5 &&
rev >= DC_REVISION_98715AEC_C)
t++;
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
if (t->dc_did == DC_DEVICEID_987x5 &&
rev >= DC_REVISION_98725)
t++;
if (t->dc_did == DC_DEVICEID_AX88140A &&
rev >= DC_REVISION_88141)
t++;
if (t->dc_did == DC_DEVICEID_82C168 &&
rev >= DC_REVISION_82C169)
t++;
Add support for the Davicom DM9102A 10/100 ethernet controller chip. This is just to make sure we initialize the chip correctly: we need to make the sure the port select bit in CSR6 is set properly so that we use the internal PHY for 10/100 support. (The eval boards I have also include an external HomePNA PHY, but I need to play with that more before I can support it.)
2000-01-19 19:03:08 +00:00
if (t->dc_did == DC_DEVICEID_DM9102 &&
rev >= DC_REVISION_DM9102A)
t++;
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
return(t);
}
t++;
}
return(NULL);
}
/*
* Probe for a 21143 or clone chip. Check the PCI vendor and device
* IDs against our list and return a device name if we find a match.
* We do a little bit of extra work to identify the exact type of
* chip. The MX98713 and MX98713A have the same PCI vendor/device ID,
* but different revision IDs. The same is true for 98715/98715A
* chips and the 98725, as well as the ASIX and ADMtek chips. In some
* cases, the exact chip revision affects driver behavior.
*/
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static int
dc_probe(dev)
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
device_t dev;
{
struct dc_type *t;
t = dc_devtype(dev);
if (t != NULL) {
device_set_desc(dev, t->dc_name);
return(0);
}
return(ENXIO);
}
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static void
dc_acpi(dev)
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
device_t dev;
{
int unit;
unit = device_get_unit(dev);
Use pci_get_powerstate()/pci_set_powerstate() which now exists in the PCI code. This saves each driver from having to grovel around looking for the right registers to twiddle. I should eventually convert the other PCI drivers to do this; for now, these three are ones which I know need power state handling.
2000-12-18 21:53:05 +00:00
if (pci_get_powerstate(dev) != PCI_POWERSTATE_D0) {
u_int32_t iobase, membase, irq;
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
Use pci_get_powerstate()/pci_set_powerstate() which now exists in the PCI code. This saves each driver from having to grovel around looking for the right registers to twiddle. I should eventually convert the other PCI drivers to do this; for now, these three are ones which I know need power state handling.
2000-12-18 21:53:05 +00:00
/* Save important PCI config data. */
iobase = pci_read_config(dev, DC_PCI_CFBIO, 4);
membase = pci_read_config(dev, DC_PCI_CFBMA, 4);
irq = pci_read_config(dev, DC_PCI_CFIT, 4);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
Use pci_get_powerstate()/pci_set_powerstate() which now exists in the PCI code. This saves each driver from having to grovel around looking for the right registers to twiddle. I should eventually convert the other PCI drivers to do this; for now, these three are ones which I know need power state handling.
2000-12-18 21:53:05 +00:00
/* Reset the power state. */
printf("dc%d: chip is in D%d power mode "
"-- setting to D0\n", unit,
pci_get_powerstate(dev));
pci_set_powerstate(dev, PCI_POWERSTATE_D0);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
Use pci_get_powerstate()/pci_set_powerstate() which now exists in the PCI code. This saves each driver from having to grovel around looking for the right registers to twiddle. I should eventually convert the other PCI drivers to do this; for now, these three are ones which I know need power state handling.
2000-12-18 21:53:05 +00:00
/* Restore PCI config data. */
pci_write_config(dev, DC_PCI_CFBIO, iobase, 4);
pci_write_config(dev, DC_PCI_CFBMA, membase, 4);
pci_write_config(dev, DC_PCI_CFIT, irq, 4);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
}
Use pci_get_powerstate()/pci_set_powerstate() which now exists in the PCI code. This saves each driver from having to grovel around looking for the right registers to twiddle. I should eventually convert the other PCI drivers to do this; for now, these three are ones which I know need power state handling.
2000-12-18 21:53:05 +00:00
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
return;
}
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static void
dc_apply_fixup(sc, media)
Add support for parsing the media blocks from the SROM on 21143 adapters. This is necessary in order to make this driver work with the built-in ethernet on the alpha Miata machines. These systems have a 21143-PC chip on-board and optional daughtercards with either a 10/100 MII transceiver or a 10baseT/10base2 transceiver. In both cases, you need to twiddle the GPIO bits on the controller in order to turn the transceivers on, and you have to read the media info from the SROM in order to find out what bits to twiddle.
2000-10-05 17:36:14 +00:00
struct dc_softc *sc;
int media;
{
struct dc_mediainfo *m;
u_int8_t *p;
int i;
Grrr. The 'reg' variable in dc_apply_fixup() needs to be a u_int32_t, not a u_int8_t. Pass the conical hat. This should fix certain cardbus 21143 cards that require SROM h0h0magic in order to enable their transceivers.
2000-10-30 23:51:39 +00:00
u_int32_t reg;
Add support for parsing the media blocks from the SROM on 21143 adapters. This is necessary in order to make this driver work with the built-in ethernet on the alpha Miata machines. These systems have a 21143-PC chip on-board and optional daughtercards with either a 10/100 MII transceiver or a 10baseT/10base2 transceiver. In both cases, you need to twiddle the GPIO bits on the controller in order to turn the transceivers on, and you have to read the media info from the SROM in order to find out what bits to twiddle.
2000-10-05 17:36:14 +00:00
m = sc->dc_mi;
while (m != NULL) {
if (m->dc_media == media)
break;
m = m->dc_next;
}
if (m == NULL)
return;
for (i = 0, p = m->dc_reset_ptr; i < m->dc_reset_len; i++, p += 2) {
reg = (p[0] | (p[1] << 8)) << 16;
CSR_WRITE_4(sc, DC_WATCHDOG, reg);
}
for (i = 0, p = m->dc_gp_ptr; i < m->dc_gp_len; i++, p += 2) {
reg = (p[0] | (p[1] << 8)) << 16;
CSR_WRITE_4(sc, DC_WATCHDOG, reg);
}
return;
}
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static void
dc_decode_leaf_sia(sc, l)
Add support for parsing the media blocks from the SROM on 21143 adapters. This is necessary in order to make this driver work with the built-in ethernet on the alpha Miata machines. These systems have a 21143-PC chip on-board and optional daughtercards with either a 10/100 MII transceiver or a 10baseT/10base2 transceiver. In both cases, you need to twiddle the GPIO bits on the controller in order to turn the transceivers on, and you have to read the media info from the SROM in order to find out what bits to twiddle.
2000-10-05 17:36:14 +00:00
struct dc_softc *sc;
struct dc_eblock_sia *l;
{
struct dc_mediainfo *m;
m = malloc(sizeof(struct dc_mediainfo), M_DEVBUF, M_NOWAIT);
Grrrr. Remember to bzero() the mediainfo structures after we allocate them. If we leave garbage in them, the dc_apply_fixup() routine may try to follow bogus pointers when applying the reset fixup. Noticed by: Andrew Gallatin
2000-11-03 00:03:03 +00:00
bzero(m, sizeof(struct dc_mediainfo));
Add support for parsing the media blocks from the SROM on 21143 adapters. This is necessary in order to make this driver work with the built-in ethernet on the alpha Miata machines. These systems have a 21143-PC chip on-board and optional daughtercards with either a 10/100 MII transceiver or a 10baseT/10base2 transceiver. In both cases, you need to twiddle the GPIO bits on the controller in order to turn the transceivers on, and you have to read the media info from the SROM in order to find out what bits to twiddle.
2000-10-05 17:36:14 +00:00
if (l->dc_sia_code == DC_SIA_CODE_10BT)
m->dc_media = IFM_10_T;
if (l->dc_sia_code == DC_SIA_CODE_10BT_FDX)
m->dc_media = IFM_10_T|IFM_FDX;
if (l->dc_sia_code == DC_SIA_CODE_10B2)
m->dc_media = IFM_10_2;
if (l->dc_sia_code == DC_SIA_CODE_10B5)
m->dc_media = IFM_10_5;
m->dc_gp_len = 2;
m->dc_gp_ptr = (u_int8_t *)&l->dc_sia_gpio_ctl;
m->dc_next = sc->dc_mi;
sc->dc_mi = m;
sc->dc_pmode = DC_PMODE_SIA;
return;
}
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static void
dc_decode_leaf_sym(sc, l)
Add support for parsing the media blocks from the SROM on 21143 adapters. This is necessary in order to make this driver work with the built-in ethernet on the alpha Miata machines. These systems have a 21143-PC chip on-board and optional daughtercards with either a 10/100 MII transceiver or a 10baseT/10base2 transceiver. In both cases, you need to twiddle the GPIO bits on the controller in order to turn the transceivers on, and you have to read the media info from the SROM in order to find out what bits to twiddle.
2000-10-05 17:36:14 +00:00
struct dc_softc *sc;
struct dc_eblock_sym *l;
{
struct dc_mediainfo *m;
m = malloc(sizeof(struct dc_mediainfo), M_DEVBUF, M_NOWAIT);
Grrrr. Remember to bzero() the mediainfo structures after we allocate them. If we leave garbage in them, the dc_apply_fixup() routine may try to follow bogus pointers when applying the reset fixup. Noticed by: Andrew Gallatin
2000-11-03 00:03:03 +00:00
bzero(m, sizeof(struct dc_mediainfo));
Add support for parsing the media blocks from the SROM on 21143 adapters. This is necessary in order to make this driver work with the built-in ethernet on the alpha Miata machines. These systems have a 21143-PC chip on-board and optional daughtercards with either a 10/100 MII transceiver or a 10baseT/10base2 transceiver. In both cases, you need to twiddle the GPIO bits on the controller in order to turn the transceivers on, and you have to read the media info from the SROM in order to find out what bits to twiddle.
2000-10-05 17:36:14 +00:00
if (l->dc_sym_code == DC_SYM_CODE_100BT)
m->dc_media = IFM_100_TX;
if (l->dc_sym_code == DC_SYM_CODE_100BT_FDX)
m->dc_media = IFM_100_TX|IFM_FDX;
m->dc_gp_len = 2;
m->dc_gp_ptr = (u_int8_t *)&l->dc_sym_gpio_ctl;
m->dc_next = sc->dc_mi;
sc->dc_mi = m;
sc->dc_pmode = DC_PMODE_SYM;
return;
}
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static void
dc_decode_leaf_mii(sc, l)
Add support for parsing the media blocks from the SROM on 21143 adapters. This is necessary in order to make this driver work with the built-in ethernet on the alpha Miata machines. These systems have a 21143-PC chip on-board and optional daughtercards with either a 10/100 MII transceiver or a 10baseT/10base2 transceiver. In both cases, you need to twiddle the GPIO bits on the controller in order to turn the transceivers on, and you have to read the media info from the SROM in order to find out what bits to twiddle.
2000-10-05 17:36:14 +00:00
struct dc_softc *sc;
struct dc_eblock_mii *l;
{
u_int8_t *p;
struct dc_mediainfo *m;
m = malloc(sizeof(struct dc_mediainfo), M_DEVBUF, M_NOWAIT);
Grrrr. Remember to bzero() the mediainfo structures after we allocate them. If we leave garbage in them, the dc_apply_fixup() routine may try to follow bogus pointers when applying the reset fixup. Noticed by: Andrew Gallatin
2000-11-03 00:03:03 +00:00
bzero(m, sizeof(struct dc_mediainfo));
Add support for parsing the media blocks from the SROM on 21143 adapters. This is necessary in order to make this driver work with the built-in ethernet on the alpha Miata machines. These systems have a 21143-PC chip on-board and optional daughtercards with either a 10/100 MII transceiver or a 10baseT/10base2 transceiver. In both cases, you need to twiddle the GPIO bits on the controller in order to turn the transceivers on, and you have to read the media info from the SROM in order to find out what bits to twiddle.
2000-10-05 17:36:14 +00:00
/* We abuse IFM_AUTO to represent MII. */
m->dc_media = IFM_AUTO;
m->dc_gp_len = l->dc_gpr_len;
p = (u_int8_t *)l;
p += sizeof(struct dc_eblock_mii);
m->dc_gp_ptr = p;
p += 2 * l->dc_gpr_len;
m->dc_reset_len = *p;
p++;
m->dc_reset_ptr = p;
m->dc_next = sc->dc_mi;
sc->dc_mi = m;
return;
}
Be consistent about functions being static. Spotted by: FlexeLint.
2002-10-16 09:04:52 +00:00
static void
dc_read_srom(sc, bits)
Dynamically configure the width of the srom. This code comes from OpenBSD who got the code (or the idea) from the NetBSD tlp driver. This gets some cardbus dc cards working (either completely or nearly so). It also appears to get additional pci cards working, without breaking working ones. # Maybe some additional work is needed here. Also, the cardbus attachment # might need to match on the CIS rather than on the vendor/device so we have # a finer level of detail as to what the card is. Technically, the # vendor/device fields are undefined for CardBus (even though most cards are # using common silicon with pci models).
2002-10-07 05:26:35 +00:00
struct dc_softc *sc;
int bits;
{
int size;
size = 2 << bits;
sc->dc_srom = malloc(size, M_DEVBUF, M_NOWAIT);
dc_read_eeprom(sc, (caddr_t)sc->dc_srom, 0, (size / 2), 0);
}
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static void
dc_parse_21143_srom(sc)
Add support for parsing the media blocks from the SROM on 21143 adapters. This is necessary in order to make this driver work with the built-in ethernet on the alpha Miata machines. These systems have a 21143-PC chip on-board and optional daughtercards with either a 10/100 MII transceiver or a 10baseT/10base2 transceiver. In both cases, you need to twiddle the GPIO bits on the controller in order to turn the transceivers on, and you have to read the media info from the SROM in order to find out what bits to twiddle.
2000-10-05 17:36:14 +00:00
struct dc_softc *sc;
{
struct dc_leaf_hdr *lhdr;
struct dc_eblock_hdr *hdr;
int i, loff;
char *ptr;
loff = sc->dc_srom[27];
lhdr = (struct dc_leaf_hdr *)&(sc->dc_srom[loff]);
ptr = (char *)lhdr;
ptr += sizeof(struct dc_leaf_hdr) - 1;
for (i = 0; i < lhdr->dc_mcnt; i++) {
hdr = (struct dc_eblock_hdr *)ptr;
switch(hdr->dc_type) {
case DC_EBLOCK_MII:
dc_decode_leaf_mii(sc, (struct dc_eblock_mii *)hdr);
break;
case DC_EBLOCK_SIA:
dc_decode_leaf_sia(sc, (struct dc_eblock_sia *)hdr);
break;
case DC_EBLOCK_SYM:
dc_decode_leaf_sym(sc, (struct dc_eblock_sym *)hdr);
break;
default:
/* Don't care. Yet. */
break;
}
ptr += (hdr->dc_len & 0x7F);
ptr++;
}
return;
}
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
/*
* Attach the interface. Allocate softc structures, do ifmedia
* setup and ethernet/BPF attach.
*/
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static int
dc_attach(dev)
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
device_t dev;
{
First round of converting network drivers from spls to mutexes. This takes care of all the 10/100 and gigE PCI drivers that I've done. Next will be the wireless drivers, then the USB ones. I may pick up some stragglers along the way. I'm sort of playing this by ear: if anyone spots any places where I've screwed up horribly, please let me know.
2000-10-13 17:54:19 +00:00
int tmp = 0;
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
u_char eaddr[ETHER_ADDR_LEN];
u_int32_t command;
struct dc_softc *sc;
struct ifnet *ifp;
u_int32_t revision;
int unit, error = 0, rid, mac_offset;
Xircom cards store the MAC address in the CIS, so get it from the new pci_get_ether accesor, which gets it from the CIS for cardbus cards (and from other pci-like buses via whatever mechanism is used there). Submitted by: sam Approved by: re (blanket)
2002-11-27 07:04:10 +00:00
u_int8_t *mac;
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
sc = device_get_softc(dev);
unit = device_get_unit(dev);
bzero(sc, sizeof(struct dc_softc));
Change callers of mtx_init() to pass in an appropriate lock type name. In most cases NULL is passed, but in some cases such as network driver locks (which use the MTX_NETWORK_LOCK macro) and UMA zone locks, a name is used. Tested on: i386, alpha, sparc64
2002-04-04 21:03:38 +00:00
mtx_init(&sc->dc_mtx, device_get_nameunit(dev), MTX_NETWORK_LOCK,
MTX_DEF | MTX_RECURSE);
Initialize/grab the mutex earlier in the attach phase, so that bailing out to the fail: label where we release/destroy the mutex will work without exploding.
2000-12-04 22:46:50 +00:00
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
/*
* Handle power management nonsense.
*/
dc_acpi(dev);
/*
* Map control/status registers.
*/
Big round of minor updates: - Use pci_get_powerstate()/pci_set_powerstate() in all the other drivers that need them so we don't have to fiddle with the PCI power management registers directly. - Use pci_enable_busmaster()/pci_enable_io() to turn on busmastering and PIO/memory mapped accesses. - Add support to the RealTek driver for the D-Link DFE-530TX+ which has a RealTek 8139 with its own PCI ID. (Submitted by Jason Wright) - Have the SiS 900/National DP83815 driver be sure to disable PME mode in sis_reset(). This apparently fixes a problem on some motherboards where the DP83815 chip fails to receive packets. (Submitted by Chuck McCrobie <mccrobie@cablespeed.com>)
2001-02-21 20:54:22 +00:00
pci_enable_busmaster(dev);
Apply patch supplied by Jonathan Chen: use the correct arguments to pci_enable_io(). We need to use SYS_RES_IOPORT/SYS_RES_MEMORY instead of PCIM_CMD_PORTEN/PCIM_CMD_MEMEN.
2001-07-09 17:58:42 +00:00
pci_enable_io(dev, SYS_RES_IOPORT);
pci_enable_io(dev, SYS_RES_MEMORY);
Use the correct register names. s/PCI_COMMAND_STATUS_REG/PCIR_COMMAND/
2000-05-28 16:02:05 +00:00
command = pci_read_config(dev, PCIR_COMMAND, 4);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
#ifdef DC_USEIOSPACE
if (!(command & PCIM_CMD_PORTEN)) {
printf("dc%d: failed to enable I/O ports!\n", unit);
error = ENXIO;
Remove bogus locking from dc(4). Instead, move interrupt allocation and ether_ifattach() to end. This fixes a "could sleep" case and simplifies error exit cases as well. Also be sure to set errno and clean up resources in !mac error case. Tested by: Ryan Beasley
2003-01-14 19:31:27 +00:00
goto fail;
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
}
#else
if (!(command & PCIM_CMD_MEMEN)) {
printf("dc%d: failed to enable memory mapping!\n", unit);
error = ENXIO;
Remove bogus locking from dc(4). Instead, move interrupt allocation and ether_ifattach() to end. This fixes a "could sleep" case and simplifies error exit cases as well. Also be sure to set errno and clean up resources in !mac error case. Tested by: Ryan Beasley
2003-01-14 19:31:27 +00:00
goto fail;
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
}
#endif
rid = DC_RID;
sc->dc_res = bus_alloc_resource(dev, DC_RES, &rid,
0, ~0, 1, RF_ACTIVE);
if (sc->dc_res == NULL) {
printf("dc%d: couldn't map ports/memory\n", unit);
error = ENXIO;
Remove bogus locking from dc(4). Instead, move interrupt allocation and ether_ifattach() to end. This fixes a "could sleep" case and simplifies error exit cases as well. Also be sure to set errno and clean up resources in !mac error case. Tested by: Ryan Beasley
2003-01-14 19:31:27 +00:00
goto fail;
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
}
sc->dc_btag = rman_get_bustag(sc->dc_res);
sc->dc_bhandle = rman_get_bushandle(sc->dc_res);
/* Need this info to decide on a chip type. */
sc->dc_info = dc_devtype(dev);
revision = pci_read_config(dev, DC_PCI_CFRV, 4) & 0x000000FF;
switch(sc->dc_info->dc_did) {
case DC_DEVICEID_21143:
sc->dc_type = DC_TYPE_21143;
sc->dc_flags |= DC_TX_POLL|DC_TX_USE_TX_INTR;
When I tweaked if_dc.c to alter the polling interval for non-MII 21143 chips, I accidentally removed the DC_MII_REDUCED_POLL flag for all 21143 cards. This caused problems with timer-instigated TCP retransmits, which happened to occur at the same time as an MII poll tick on MII-based cards (e.g. D-Link DFE-570TX). Fixed this, plus made some other cleanups. The autoneg fixes for the non-MII cards still work. Also tested the PNIC II now that I have one again.
2000-06-07 17:07:44 +00:00
sc->dc_flags |= DC_REDUCED_MII_POLL;
Add support for parsing the media blocks from the SROM on 21143 adapters. This is necessary in order to make this driver work with the built-in ethernet on the alpha Miata machines. These systems have a 21143-PC chip on-board and optional daughtercards with either a 10/100 MII transceiver or a 10baseT/10base2 transceiver. In both cases, you need to twiddle the GPIO bits on the controller in order to turn the transceivers on, and you have to read the media info from the SROM in order to find out what bits to twiddle.
2000-10-05 17:36:14 +00:00
/* Save EEPROM contents so we can parse them later. */
Dynamically configure the width of the srom. This code comes from OpenBSD who got the code (or the idea) from the NetBSD tlp driver. This gets some cardbus dc cards working (either completely or nearly so). It also appears to get additional pci cards working, without breaking working ones. # Maybe some additional work is needed here. Also, the cardbus attachment # might need to match on the CIS rather than on the vendor/device so we have # a finer level of detail as to what the card is. Technically, the # vendor/device fields are undefined for CardBus (even though most cards are # using common silicon with pci models).
2002-10-07 05:26:35 +00:00
dc_eeprom_width(sc);
dc_read_srom(sc, sc->dc_romwidth);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
break;
Add support for the Davicom DM9009 chipset. PR: 46859 Submitted by: Boaz Haberman <boaz@ool-182f8b09.dyn.optonline.net> Approved by: rwatson
2003-01-09 20:51:39 +00:00
case DC_DEVICEID_DM9009:
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
case DC_DEVICEID_DM9100:
case DC_DEVICEID_DM9102:
sc->dc_type = DC_TYPE_DM9102;
Set the DC_TX_INTR_ALWAYS and DC_TX_STORENFWD flags for the Davicom DM9100/DM9102 chips. Do not set DC_TX_ONE. The DC_TX_USE_TX_INTR flag causes dc_encap() to set the 'interrupt on TX completion' bit only once every 64 packets. This is an attempt to reduce the number of interrupts generated by the chip. You're supposed to get a 'no more TX buffers left' interrupt once you hit the last packet whether you ask for one or not, however it seems the Davicom chip doesn't generate this interrupt, or at least it doesn't generate it under the same circumstances. The result is that if you transmit n packets, where n is less than 64, and then wait 5 seconds, you'll get a watchdog timeout whether you want one or not. The DC_TX_INTR_ALWAYS causes dc_encap() to request an interrupt for every frame. I'm still waiting on confirmation from a couple of users to see if this fixes their problems with the Davicom DM9102 before I merge this into -stable, but this fixed the problem for me in my own testing so I'm willing to make the change to -current right away.
2000-10-25 23:46:31 +00:00
sc->dc_flags |= DC_TX_COALESCE|DC_TX_INTR_ALWAYS;
sc->dc_flags |= DC_REDUCED_MII_POLL|DC_TX_STORENFWD;
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
sc->dc_pmode = DC_PMODE_MII;
Yet another bug fix/optimization for the Davicom DM9100/9102: increase the PCI latency timer value to 0x80. Davicom's Linux driver does this, and it drastically reduces the number of TX underruns in my tests. (Note: this is done only for the Davicom chips. I'm not sure it's a good idea to do it for all of them.) Again, still waiting on confirmation before merging to stable.
2000-10-27 00:15:04 +00:00
/* Increase the latency timer value. */
command = pci_read_config(dev, DC_PCI_CFLT, 4);
command &= 0xFFFF00FF;
command |= 0x00008000;
pci_write_config(dev, DC_PCI_CFLT, command, 4);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
break;
case DC_DEVICEID_AL981:
sc->dc_type = DC_TYPE_AL981;
sc->dc_flags |= DC_TX_USE_TX_INTR;
sc->dc_flags |= DC_TX_ADMTEK_WAR;
sc->dc_pmode = DC_PMODE_MII;
Dynamically configure the width of the srom. This code comes from OpenBSD who got the code (or the idea) from the NetBSD tlp driver. This gets some cardbus dc cards working (either completely or nearly so). It also appears to get additional pci cards working, without breaking working ones. # Maybe some additional work is needed here. Also, the cardbus attachment # might need to match on the CIS rather than on the vendor/device so we have # a finer level of detail as to what the card is. Technically, the # vendor/device fields are undefined for CardBus (even though most cards are # using common silicon with pci models).
2002-10-07 05:26:35 +00:00
dc_eeprom_width(sc);
dc_read_srom(sc, sc->dc_romwidth);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
break;
case DC_DEVICEID_AN985:
Fix typo s/DE_DEVICEID_FE2500/DC_DEVICEID_FE2500/
2000-10-28 10:03:54 +00:00
case DC_DEVICEID_FE2500:
Add support for the Accton EN2242 MiniPCI adapter. This is just an ADMtek Centaur chip, so all we need is the PCI ID. Submitted by: Scott Lang <scottl@FreeBSD.org>
2000-11-14 19:35:22 +00:00
case DC_DEVICEID_EN2242:
Add preliminary support for the Hawking PN672TX CardBus cards. # Preliminary because there are some subtle things the NetBSD driver does # that we don't do yet. My card works for me w/o them.
2003-01-10 05:37:15 +00:00
case DC_DEVICEID_HAWKING_PN672TX:
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
sc->dc_type = DC_TYPE_AN985;
sc->dc_flags |= DC_TX_USE_TX_INTR;
sc->dc_flags |= DC_TX_ADMTEK_WAR;
sc->dc_pmode = DC_PMODE_MII;
Dynamically configure the width of the srom. This code comes from OpenBSD who got the code (or the idea) from the NetBSD tlp driver. This gets some cardbus dc cards working (either completely or nearly so). It also appears to get additional pci cards working, without breaking working ones. # Maybe some additional work is needed here. Also, the cardbus attachment # might need to match on the CIS rather than on the vendor/device so we have # a finer level of detail as to what the card is. Technically, the # vendor/device fields are undefined for CardBus (even though most cards are # using common silicon with pci models).
2002-10-07 05:26:35 +00:00
dc_eeprom_width(sc);
dc_read_srom(sc, sc->dc_romwidth);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
break;
case DC_DEVICEID_98713:
case DC_DEVICEID_98713_CP:
if (revision < DC_REVISION_98713A) {
sc->dc_type = DC_TYPE_98713;
}
Rework the support for the internal autonegotiation on the 21143 and workalike chips (Macronix 98713A/98715 and PNIC II). Timing is somewhat critical: you need to bring the link as soon as possible after NWAY is done, and the old one second polling interval was too long. Now we poll every 10th of a second until NWAY completes (at which point we return to the 1 second interval again to keep an eye on the link state). I tested all the other cards I had on hand to make sure I didn't bust any of them and they seem to work (including the MII-based 21143 card). This should fix some autoneg problems with DE500-BA cards and the built-in 10/100 ethernet on some alpha systems. (Now before anyone asks why I never noticed this before, the old code worked just find with the Intel swich I used for testing back in NY. Apparently not all switches are as picky about the timing.)
2000-05-31 05:40:53 +00:00
if (revision >= DC_REVISION_98713A) {
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
sc->dc_type = DC_TYPE_98713A;
Rework the support for the internal autonegotiation on the 21143 and workalike chips (Macronix 98713A/98715 and PNIC II). Timing is somewhat critical: you need to bring the link as soon as possible after NWAY is done, and the old one second polling interval was too long. Now we poll every 10th of a second until NWAY completes (at which point we return to the 1 second interval again to keep an eye on the link state). I tested all the other cards I had on hand to make sure I didn't bust any of them and they seem to work (including the MII-based 21143 card). This should fix some autoneg problems with DE500-BA cards and the built-in 10/100 ethernet on some alpha systems. (Now before anyone asks why I never noticed this before, the old code worked just find with the Intel swich I used for testing back in NY. Apparently not all switches are as picky about the timing.)
2000-05-31 05:40:53 +00:00
sc->dc_flags |= DC_21143_NWAY;
}
sc->dc_flags |= DC_REDUCED_MII_POLL;
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
sc->dc_flags |= DC_TX_POLL|DC_TX_USE_TX_INTR;
break;
case DC_DEVICEID_987x5:
Add support for the Accton EN1217. PR: 18735 Submitted by: Adoal Xu <adoal@iname.com>
2000-06-11 11:54:52 +00:00
case DC_DEVICEID_EN1217:
Apply patch to the dc driver to handle Macronix MX98715AEC-C/D/E chips, which differ slightly from the Macronix MX98715AEC chip on the sample adapter that I have in that the multicast hash table is only 128 bits wide instead of 512. New adapters are popping up with this chip, and due to improper handling of the smaller hash table, broadcast packets were not being received correctly.
2000-07-15 17:54:30 +00:00
/*
* Macronix MX98715AEC-C/D/E parts have only a
* 128-bit hash table. We need to deal with these
* in the same manner as the PNIC II so that we
* get the right number of bits out of the
* CRC routine.
*/
if (revision >= DC_REVISION_98715AEC_C &&
revision < DC_REVISION_98725)
sc->dc_flags |= DC_128BIT_HASH;
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
sc->dc_type = DC_TYPE_987x5;
sc->dc_flags |= DC_TX_POLL|DC_TX_USE_TX_INTR;
Rework the support for the internal autonegotiation on the 21143 and workalike chips (Macronix 98713A/98715 and PNIC II). Timing is somewhat critical: you need to bring the link as soon as possible after NWAY is done, and the old one second polling interval was too long. Now we poll every 10th of a second until NWAY completes (at which point we return to the 1 second interval again to keep an eye on the link state). I tested all the other cards I had on hand to make sure I didn't bust any of them and they seem to work (including the MII-based 21143 card). This should fix some autoneg problems with DE500-BA cards and the built-in 10/100 ethernet on some alpha systems. (Now before anyone asks why I never noticed this before, the old code worked just find with the Intel swich I used for testing back in NY. Apparently not all switches are as picky about the timing.)
2000-05-31 05:40:53 +00:00
sc->dc_flags |= DC_REDUCED_MII_POLL|DC_21143_NWAY;
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
break;
Add the PCI IDs for the Macronix 98727 and 98732 parts. These are 3.3volt PCI/cardbus chipsets similar to the 98715 (and they have 512-bit hash tables). Also update the man page to mention the 98727/98732 and the SOHOware SFA110A Rev B4 card with the 98715AEC-C chip.
2000-07-17 19:27:41 +00:00
case DC_DEVICEID_98727:
sc->dc_type = DC_TYPE_987x5;
sc->dc_flags |= DC_TX_POLL|DC_TX_USE_TX_INTR;
sc->dc_flags |= DC_REDUCED_MII_POLL|DC_21143_NWAY;
break;
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
case DC_DEVICEID_82C115:
sc->dc_type = DC_TYPE_PNICII;
Apply patch to the dc driver to handle Macronix MX98715AEC-C/D/E chips, which differ slightly from the Macronix MX98715AEC chip on the sample adapter that I have in that the multicast hash table is only 128 bits wide instead of 512. New adapters are popping up with this chip, and due to improper handling of the smaller hash table, broadcast packets were not being received correctly.
2000-07-15 17:54:30 +00:00
sc->dc_flags |= DC_TX_POLL|DC_TX_USE_TX_INTR|DC_128BIT_HASH;
Rework the support for the internal autonegotiation on the 21143 and workalike chips (Macronix 98713A/98715 and PNIC II). Timing is somewhat critical: you need to bring the link as soon as possible after NWAY is done, and the old one second polling interval was too long. Now we poll every 10th of a second until NWAY completes (at which point we return to the 1 second interval again to keep an eye on the link state). I tested all the other cards I had on hand to make sure I didn't bust any of them and they seem to work (including the MII-based 21143 card). This should fix some autoneg problems with DE500-BA cards and the built-in 10/100 ethernet on some alpha systems. (Now before anyone asks why I never noticed this before, the old code worked just find with the Intel swich I used for testing back in NY. Apparently not all switches are as picky about the timing.)
2000-05-31 05:40:53 +00:00
sc->dc_flags |= DC_REDUCED_MII_POLL|DC_21143_NWAY;
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
break;
case DC_DEVICEID_82C168:
sc->dc_type = DC_TYPE_PNIC;
Fix some problems reported by Mike Pritchard: - Add a flag DC_TX_INTR_ALWAYS which causes the transmit code to request a TX done interrupt for every packet. The PNIC seems to need this to insure that the sent TX buffers get reaped in a timely fashion. - Try to unreset the SIA as soon as possible after resetting the whole chip. - Change dcphy to support either 10/100 or 10Mbps only NICs. The built-in 21143 ethernet in Compaq Presario machines is 10Mbps only and it doesn't work right if we try to advertise 100Mbps modes during autoneg. When restricted to only 10mbps modes, it works fine. Note that for now, I detect this condition by checking the PCI subsystem ID on this NIC (which has a Compaq vendor/device ID). Yes, I know that's what the SROM is supposed to be for. I'm deliberately ignoring the SROM wherever possible. Sue me. The latter two fixes allow if_dc to work correctly with the built-in ethernet on certain Compaq Presario boxes. There are liable to be quite a few people using these as their home systems who might want to try FreeBSD; may as well be nice to them. Now if anybody out there has an Alpha miata with 10Mbps ethernet and can show me the output from pciconf -l on their system, I'd be grateful.
1999-12-13 21:45:13 +00:00
sc->dc_flags |= DC_TX_STORENFWD|DC_TX_INTR_ALWAYS;
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
sc->dc_flags |= DC_PNIC_RX_BUG_WAR;
sc->dc_pnic_rx_buf = malloc(DC_RXLEN * 5, M_DEVBUF, M_NOWAIT);
if (revision < DC_REVISION_82C169)
sc->dc_pmode = DC_PMODE_SYM;
break;
case DC_DEVICEID_AX88140A:
sc->dc_type = DC_TYPE_ASIX;
sc->dc_flags |= DC_TX_USE_TX_INTR|DC_TX_INTR_FIRSTFRAG;
sc->dc_flags |= DC_REDUCED_MII_POLL;
sc->dc_pmode = DC_PMODE_MII;
break;
NEWCARD/Cardbus - This commit adds support for Xircom X3201 based cardbus cards. Support for the TDK 78Q2120 MII is also added. IBM Etherjet, Intel and Xircom cards uses these chips. Note that as a result of this commit, some Intel/DEC 21143 based cardbus cards will also attach, but not get link. That is being looked at.
2000-10-19 08:34:32 +00:00
case DC_DEVICEID_X3201:
sc->dc_type = DC_TYPE_XIRCOM;
Avoid an unnecessary copy of a packet if it is already in a single mbuf. Introduce an additional device flag for those NICs which require the transmit buffers to be aligned to 32-bit boundaries. (the equivalen fix for STABLE is slightly simpler because there are no supported chips which require this alignment there.)
2001-12-11 02:47:53 +00:00
sc->dc_flags |= DC_TX_INTR_ALWAYS | DC_TX_COALESCE |
Patch up some existing style bugs and some that crept in with the DEVICE_POLLING stuff.
2001-12-15 02:51:21 +00:00
DC_TX_ALIGN;
NEWCARD/Cardbus - This commit adds support for Xircom X3201 based cardbus cards. Support for the TDK 78Q2120 MII is also added. IBM Etherjet, Intel and Xircom cards uses these chips. Note that as a result of this commit, some Intel/DEC 21143 based cardbus cards will also attach, but not get link. That is being looked at.
2000-10-19 08:34:32 +00:00
/*
* We don't actually need to coalesce, but we're doing
* it to obtain a double word aligned buffer.
Avoid an unnecessary copy of a packet if it is already in a single mbuf. Introduce an additional device flag for those NICs which require the transmit buffers to be aligned to 32-bit boundaries. (the equivalen fix for STABLE is slightly simpler because there are no supported chips which require this alignment there.)
2001-12-11 02:47:53 +00:00
* The DC_TX_COALESCE flag is required.
NEWCARD/Cardbus - This commit adds support for Xircom X3201 based cardbus cards. Support for the TDK 78Q2120 MII is also added. IBM Etherjet, Intel and Xircom cards uses these chips. Note that as a result of this commit, some Intel/DEC 21143 based cardbus cards will also attach, but not get link. That is being looked at.
2000-10-19 08:34:32 +00:00
*/
Dynamically configure the width of the srom. This code comes from OpenBSD who got the code (or the idea) from the NetBSD tlp driver. This gets some cardbus dc cards working (either completely or nearly so). It also appears to get additional pci cards working, without breaking working ones. # Maybe some additional work is needed here. Also, the cardbus attachment # might need to match on the CIS rather than on the vendor/device so we have # a finer level of detail as to what the card is. Technically, the # vendor/device fields are undefined for CardBus (even though most cards are # using common silicon with pci models).
2002-10-07 05:26:35 +00:00
sc->dc_pmode = DC_PMODE_MII;
NEWCARD/Cardbus - This commit adds support for Xircom X3201 based cardbus cards. Support for the TDK 78Q2120 MII is also added. IBM Etherjet, Intel and Xircom cards uses these chips. Note that as a result of this commit, some Intel/DEC 21143 based cardbus cards will also attach, but not get link. That is being looked at.
2000-10-19 08:34:32 +00:00
break;
Add support for Conexant LANfinity miniPCI controllers. People who have laptops with this chip should test this and report back as I don't have access to this hardware myself. People with -stable systems should try the patch at: http://www.freebsd.org/~wpaul/conexant.patch.gz Submitted by: Phil Kernick <Phil@Kernick.org>
2001-09-04 17:10:11 +00:00
case DC_DEVICEID_RS7112:
sc->dc_type = DC_TYPE_CONEXANT;
sc->dc_flags |= DC_TX_INTR_ALWAYS;
sc->dc_flags |= DC_REDUCED_MII_POLL;
sc->dc_pmode = DC_PMODE_MII;
Dynamically configure the width of the srom. This code comes from OpenBSD who got the code (or the idea) from the NetBSD tlp driver. This gets some cardbus dc cards working (either completely or nearly so). It also appears to get additional pci cards working, without breaking working ones. # Maybe some additional work is needed here. Also, the cardbus attachment # might need to match on the CIS rather than on the vendor/device so we have # a finer level of detail as to what the card is. Technically, the # vendor/device fields are undefined for CardBus (even though most cards are # using common silicon with pci models).
2002-10-07 05:26:35 +00:00
dc_eeprom_width(sc);
dc_read_srom(sc, sc->dc_romwidth);
Add support for Conexant LANfinity miniPCI controllers. People who have laptops with this chip should test this and report back as I don't have access to this hardware myself. People with -stable systems should try the patch at: http://www.freebsd.org/~wpaul/conexant.patch.gz Submitted by: Phil Kernick <Phil@Kernick.org>
2001-09-04 17:10:11 +00:00
break;
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
default:
printf("dc%d: unknown device: %x\n", sc->dc_unit,
sc->dc_info->dc_did);
break;
}
/* Save the cache line size. */
Add support for the Davicom DM9102A 10/100 ethernet controller chip. This is just to make sure we initialize the chip correctly: we need to make the sure the port select bit in CSR6 is set properly so that we use the internal PHY for 10/100 support. (The eval boards I have also include an external HomePNA PHY, but I need to play with that more before I can support it.)
2000-01-19 19:03:08 +00:00
if (DC_IS_DAVICOM(sc))
sc->dc_cachesize = 0;
else
sc->dc_cachesize = pci_read_config(dev,
DC_PCI_CFLT, 4) & 0xFF;
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
/* Reset the adapter. */
dc_reset(sc);
/* Take 21143 out of snooze mode */
NEWCARD/Cardbus - This commit adds support for Xircom X3201 based cardbus cards. Support for the TDK 78Q2120 MII is also added. IBM Etherjet, Intel and Xircom cards uses these chips. Note that as a result of this commit, some Intel/DEC 21143 based cardbus cards will also attach, but not get link. That is being looked at.
2000-10-19 08:34:32 +00:00
if (DC_IS_INTEL(sc) || DC_IS_XIRCOM(sc)) {
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
command = pci_read_config(dev, DC_PCI_CFDD, 4);
command &= ~(DC_CFDD_SNOOZE_MODE|DC_CFDD_SLEEP_MODE);
pci_write_config(dev, DC_PCI_CFDD, command, 4);
}
/*
* Try to learn something about the supported media.
* We know that ASIX and ADMtek and Davicom devices
* will *always* be using MII media, so that's a no-brainer.
* The tricky ones are the Macronix/PNIC II and the
* Intel 21143.
*/
Add support for parsing the media blocks from the SROM on 21143 adapters. This is necessary in order to make this driver work with the built-in ethernet on the alpha Miata machines. These systems have a 21143-PC chip on-board and optional daughtercards with either a 10/100 MII transceiver or a 10baseT/10base2 transceiver. In both cases, you need to twiddle the GPIO bits on the controller in order to turn the transceivers on, and you have to read the media info from the SROM in order to find out what bits to twiddle.
2000-10-05 17:36:14 +00:00
if (DC_IS_INTEL(sc))
dc_parse_21143_srom(sc);
else if (DC_IS_MACRONIX(sc) || DC_IS_PNICII(sc)) {
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
if (sc->dc_type == DC_TYPE_98713)
sc->dc_pmode = DC_PMODE_MII;
else
sc->dc_pmode = DC_PMODE_SYM;
} else if (!sc->dc_pmode)
sc->dc_pmode = DC_PMODE_MII;
/*
* Get station address from the EEPROM.
*/
switch(sc->dc_type) {
case DC_TYPE_98713:
case DC_TYPE_98713A:
case DC_TYPE_987x5:
case DC_TYPE_PNICII:
dc_read_eeprom(sc, (caddr_t)&mac_offset,
(DC_EE_NODEADDR_OFFSET / 2), 1, 0);
dc_read_eeprom(sc, (caddr_t)&eaddr, (mac_offset / 2), 3, 0);
break;
case DC_TYPE_PNIC:
dc_read_eeprom(sc, (caddr_t)&eaddr, 0, 3, 1);
break;
case DC_TYPE_DM9102:
case DC_TYPE_21143:
case DC_TYPE_ASIX:
dc_read_eeprom(sc, (caddr_t)&eaddr, DC_EE_NODEADDR, 3, 0);
break;
case DC_TYPE_AL981:
case DC_TYPE_AN985:
Dynamically configure the width of the srom. This code comes from OpenBSD who got the code (or the idea) from the NetBSD tlp driver. This gets some cardbus dc cards working (either completely or nearly so). It also appears to get additional pci cards working, without breaking working ones. # Maybe some additional work is needed here. Also, the cardbus attachment # might need to match on the CIS rather than on the vendor/device so we have # a finer level of detail as to what the card is. Technically, the # vendor/device fields are undefined for CardBus (even though most cards are # using common silicon with pci models).
2002-10-07 05:26:35 +00:00
bcopy(&sc->dc_srom[DC_AL_EE_NODEADDR], (caddr_t)&eaddr,
ETHER_ADDR_LEN);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
dc_read_eeprom(sc, (caddr_t)&eaddr, DC_AL_EE_NODEADDR, 3, 0);
break;
Add support for Conexant LANfinity miniPCI controllers. People who have laptops with this chip should test this and report back as I don't have access to this hardware myself. People with -stable systems should try the patch at: http://www.freebsd.org/~wpaul/conexant.patch.gz Submitted by: Phil Kernick <Phil@Kernick.org>
2001-09-04 17:10:11 +00:00
case DC_TYPE_CONEXANT:
bcopy(sc->dc_srom + DC_CONEXANT_EE_NODEADDR, &eaddr, 6);
break;
NEWCARD/Cardbus - This commit adds support for Xircom X3201 based cardbus cards. Support for the TDK 78Q2120 MII is also added. IBM Etherjet, Intel and Xircom cards uses these chips. Note that as a result of this commit, some Intel/DEC 21143 based cardbus cards will also attach, but not get link. That is being looked at.
2000-10-19 08:34:32 +00:00
case DC_TYPE_XIRCOM:
Xircom cards store the MAC address in the CIS, so get it from the new pci_get_ether accesor, which gets it from the CIS for cardbus cards (and from other pci-like buses via whatever mechanism is used there). Submitted by: sam Approved by: re (blanket)
2002-11-27 07:04:10 +00:00
/* The MAC comes from the CIS */
mac = pci_get_ether(dev);
if (!mac) {
device_printf(dev, "No station address in CIS!\n");
Remove bogus locking from dc(4). Instead, move interrupt allocation and ether_ifattach() to end. This fixes a "could sleep" case and simplifies error exit cases as well. Also be sure to set errno and clean up resources in !mac error case. Tested by: Ryan Beasley
2003-01-14 19:31:27 +00:00
bus_release_resource(dev, DC_RES, DC_RID, sc->dc_res);
error = ENXIO;
Xircom cards store the MAC address in the CIS, so get it from the new pci_get_ether accesor, which gets it from the CIS for cardbus cards (and from other pci-like buses via whatever mechanism is used there). Submitted by: sam Approved by: re (blanket)
2002-11-27 07:04:10 +00:00
goto fail;
}
bcopy(mac, eaddr, ETHER_ADDR_LEN);
NEWCARD/Cardbus - This commit adds support for Xircom X3201 based cardbus cards. Support for the TDK 78Q2120 MII is also added. IBM Etherjet, Intel and Xircom cards uses these chips. Note that as a result of this commit, some Intel/DEC 21143 based cardbus cards will also attach, but not get link. That is being looked at.
2000-10-19 08:34:32 +00:00
break;
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
default:
dc_read_eeprom(sc, (caddr_t)&eaddr, DC_EE_NODEADDR, 3, 0);
break;
}
/*
* A 21143 or clone chip was detected. Inform the world.
*/
printf("dc%d: Ethernet address: %6D\n", unit, eaddr, ":");
sc->dc_unit = unit;
bcopy(eaddr, (char *)&sc->arpcom.ac_enaddr, ETHER_ADDR_LEN);
sc->dc_ldata = contigmalloc(sizeof(struct dc_list_data), M_DEVBUF,
M_NOWAIT, 0, 0xffffffff, PAGE_SIZE, 0);
if (sc->dc_ldata == NULL) {
printf("dc%d: no memory for list buffers!\n", unit);
bus_release_resource(dev, DC_RES, DC_RID, sc->dc_res);
error = ENXIO;
goto fail;
}
bzero(sc->dc_ldata, sizeof(struct dc_list_data));
ifp = &sc->arpcom.ac_if;
ifp->if_softc = sc;
ifp->if_unit = unit;
ifp->if_name = "dc";
NEWCARD/Cardbus - This commit adds support for Xircom X3201 based cardbus cards. Support for the TDK 78Q2120 MII is also added. IBM Etherjet, Intel and Xircom cards uses these chips. Note that as a result of this commit, some Intel/DEC 21143 based cardbus cards will also attach, but not get link. That is being looked at.
2000-10-19 08:34:32 +00:00
/* XXX: bleah, MTU gets overwritten in ether_ifattach() */
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
ifp->if_mtu = ETHERMTU;
ifp->if_flags = IFF_BROADCAST | IFF_SIMPLEX | IFF_MULTICAST;
ifp->if_ioctl = dc_ioctl;
ifp->if_output = ether_output;
ifp->if_start = dc_start;
ifp->if_watchdog = dc_watchdog;
ifp->if_init = dc_init;
ifp->if_baudrate = 10000000;
ifp->if_snd.ifq_maxlen = DC_TX_LIST_CNT - 1;
/*
Add support for parsing the media blocks from the SROM on 21143 adapters. This is necessary in order to make this driver work with the built-in ethernet on the alpha Miata machines. These systems have a 21143-PC chip on-board and optional daughtercards with either a 10/100 MII transceiver or a 10baseT/10base2 transceiver. In both cases, you need to twiddle the GPIO bits on the controller in order to turn the transceivers on, and you have to read the media info from the SROM in order to find out what bits to twiddle.
2000-10-05 17:36:14 +00:00
* Do MII setup. If this is a 21143, check for a PHY on the
* MII bus after applying any necessary fixups to twiddle the
* GPIO bits. If we don't end up finding a PHY, restore the
* old selection (SIA only or SIA/SYM) and attach the dcphy
* driver instead.
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
*/
Add support for parsing the media blocks from the SROM on 21143 adapters. This is necessary in order to make this driver work with the built-in ethernet on the alpha Miata machines. These systems have a 21143-PC chip on-board and optional daughtercards with either a 10/100 MII transceiver or a 10baseT/10base2 transceiver. In both cases, you need to twiddle the GPIO bits on the controller in order to turn the transceivers on, and you have to read the media info from the SROM in order to find out what bits to twiddle.
2000-10-05 17:36:14 +00:00
if (DC_IS_INTEL(sc)) {
dc_apply_fixup(sc, IFM_AUTO);
tmp = sc->dc_pmode;
sc->dc_pmode = DC_PMODE_MII;
}
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
error = mii_phy_probe(dev, &sc->dc_miibus,
dc_ifmedia_upd, dc_ifmedia_sts);
if (error && DC_IS_INTEL(sc)) {
Add support for parsing the media blocks from the SROM on 21143 adapters. This is necessary in order to make this driver work with the built-in ethernet on the alpha Miata machines. These systems have a 21143-PC chip on-board and optional daughtercards with either a 10/100 MII transceiver or a 10baseT/10base2 transceiver. In both cases, you need to twiddle the GPIO bits on the controller in order to turn the transceivers on, and you have to read the media info from the SROM in order to find out what bits to twiddle.
2000-10-05 17:36:14 +00:00
sc->dc_pmode = tmp;
if (sc->dc_pmode != DC_PMODE_SIA)
sc->dc_pmode = DC_PMODE_SYM;
When I tweaked if_dc.c to alter the polling interval for non-MII 21143 chips, I accidentally removed the DC_MII_REDUCED_POLL flag for all 21143 cards. This caused problems with timer-instigated TCP retransmits, which happened to occur at the same time as an MII poll tick on MII-based cards (e.g. D-Link DFE-570TX). Fixed this, plus made some other cleanups. The autoneg fixes for the non-MII cards still work. Also tested the PNIC II now that I have one again.
2000-06-07 17:07:44 +00:00
sc->dc_flags |= DC_21143_NWAY;
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
mii_phy_probe(dev, &sc->dc_miibus,
dc_ifmedia_upd, dc_ifmedia_sts);
Special-case the LED twiddling code so that it doesn't do anything on the NEC VersaPro NoteBook PC. This 21143 implementation has no LEDs, and flipping the LED control bits somehow stops it from establishing a link. We check the subsystem ID and don't flip the LED control bits for the NEC NIC.
2000-09-07 18:51:04 +00:00
/*
* For non-MII cards, we need to have the 21143
* drive the LEDs. Except there are some systems
* like the NEC VersaPro NoteBook PC which have no
* LEDs, and twiddling these bits has adverse effects
* on them. (I.e. you suddenly can't get a link.)
*/
if (pci_read_config(dev, DC_PCI_CSID, 4) != 0x80281033)
sc->dc_flags |= DC_TULIP_LEDS;
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
error = 0;
}
if (error) {
printf("dc%d: MII without any PHY!\n", sc->dc_unit);
bus_release_resource(dev, DC_RES, DC_RID, sc->dc_res);
error = ENXIO;
goto fail;
}
NEWCARD/Cardbus - This commit adds support for Xircom X3201 based cardbus cards. Support for the TDK 78Q2120 MII is also added. IBM Etherjet, Intel and Xircom cards uses these chips. Note that as a result of this commit, some Intel/DEC 21143 based cardbus cards will also attach, but not get link. That is being looked at.
2000-10-19 08:34:32 +00:00
if (DC_IS_XIRCOM(sc)) {
/*
* setup General Purpose Port mode and data so the tulip
* can talk to the MII.
*/
CSR_WRITE_4(sc, DC_SIAGP, DC_SIAGP_WRITE_EN | DC_SIAGP_INT1_EN |
DC_SIAGP_MD_GP2_OUTPUT | DC_SIAGP_MD_GP0_OUTPUT);
DELAY(10);
CSR_WRITE_4(sc, DC_SIAGP, DC_SIAGP_INT1_EN |
DC_SIAGP_MD_GP2_OUTPUT | DC_SIAGP_MD_GP0_OUTPUT);
DELAY(10);
}
Move the DC_SETBIT to the attach function. Reviewed by: phk
2003-02-05 21:21:26 +00:00
if (DC_IS_ADMTEK(sc)) {
/*
* Set automatic TX underrun recovery for the ADMtek chips
*/
DC_SETBIT(sc, DC_AL_CR, DC_AL_CR_ATUR);
}
Add VLAN for the dc(4) driver (ie long frame). The patch is 2 parts. One to notify the system that the MTU for VLAN can be 1500 so the vlan will automatically be configured with a 1500 MTU the other is to ignore the error case if the received frame is to long. The frame size notification came from code in the SIS driver, and the support for long frames derived from the NetBSD Tulip driver. Tested on: 4 port D-Link adapter DFE-570TX 4 Intel 21143 Netgear card with 82c169 PNIC 10/100BaseTX Reviewed by: ru (manpage), wpaul (not objected to), archie Approved by: imp Obtained from: NetBSD
2002-01-16 16:33:58 +00:00
/*
* Tell the upper layer(s) we support long frames.
*/
ifp->if_data.ifi_hdrlen = sizeof(struct ether_vlan_header);
o track either_ifattach/ether_ifdetach API changes o use if_input for input packet processing o don't strip the Ethernet header for input packets o use BPF_* macros bpf tapping o call ether_ioctl to handle default ioctl case o track vlan changes Reviewed by: many Approved by: re
2002-11-14 23:49:09 +00:00
ifp->if_capabilities |= IFCAP_VLAN_MTU;
Add VLAN for the dc(4) driver (ie long frame). The patch is 2 parts. One to notify the system that the MTU for VLAN can be 1500 so the vlan will automatically be configured with a 1500 MTU the other is to ignore the error case if the received frame is to long. The frame size notification came from code in the SIS driver, and the support for long frames derived from the NetBSD Tulip driver. Tested on: 4 port D-Link adapter DFE-570TX 4 Intel 21143 Netgear card with 82c169 PNIC 10/100BaseTX Reviewed by: ru (manpage), wpaul (not objected to), archie Approved by: imp Obtained from: NetBSD
2002-01-16 16:33:58 +00:00
Change the driver to allocate its own callout structure, and modify the interface to use callout_* instead of timeout(). Also add an IS_MPSAFE #define (currently off) which will mark the driver as mpsafe to the upper layers.
2000-11-25 08:00:10 +00:00
callout_init(&sc->dc_stat_ch, IS_MPSAFE);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
Add support for parsing the media blocks from the SROM on 21143 adapters. This is necessary in order to make this driver work with the built-in ethernet on the alpha Miata machines. These systems have a 21143-PC chip on-board and optional daughtercards with either a 10/100 MII transceiver or a 10baseT/10base2 transceiver. In both cases, you need to twiddle the GPIO bits on the controller in order to turn the transceivers on, and you have to read the media info from the SROM in order to find out what bits to twiddle.
2000-10-05 17:36:14 +00:00
#ifdef SRM_MEDIA
Make consistent; turn spaces into tabs where there is a mixture.
2002-09-04 18:14:17 +00:00
sc->dc_srm_media = 0;
Teach the 'dc' driver how to pick up settings left over by the SRM on alpha systems. This is an expedient if not entirely elegant solution to the problem. Submitted by: gallatin Approved by: jkh
2000-03-11 05:20:56 +00:00
/* Remember the SRM console media setting */
if (DC_IS_INTEL(sc)) {
command = pci_read_config(dev, DC_PCI_CFDD, 4);
command &= ~(DC_CFDD_SNOOZE_MODE|DC_CFDD_SLEEP_MODE);
switch ((command >> 8) & 0xff) {
case 3:
sc->dc_srm_media = IFM_10_T;
break;
case 4:
sc->dc_srm_media = IFM_10_T | IFM_FDX;
break;
case 5:
sc->dc_srm_media = IFM_100_TX;
break;
case 6:
sc->dc_srm_media = IFM_100_TX | IFM_FDX;
break;
}
if (sc->dc_srm_media)
sc->dc_srm_media |= IFM_ACTIVE | IFM_ETHER;
}
#endif
Remove bogus locking from dc(4). Instead, move interrupt allocation and ether_ifattach() to end. This fixes a "could sleep" case and simplifies error exit cases as well. Also be sure to set errno and clean up resources in !mac error case. Tested by: Ryan Beasley
2003-01-14 19:31:27 +00:00
/*
* Call MI attach routine.
*/
ether_ifattach(ifp, eaddr);
/* Allocate interrupt */
rid = 0;
sc->dc_irq = bus_alloc_resource(dev, SYS_RES_IRQ, &rid, 0, ~0, 1,
RF_SHAREABLE | RF_ACTIVE);
if (sc->dc_irq == NULL) {
printf("dc%d: couldn't map interrupt\n", unit);
bus_release_resource(dev, DC_RES, DC_RID, sc->dc_res);
error = ENXIO;
goto fail;
}
error = bus_setup_intr(dev, sc->dc_irq, INTR_TYPE_NET |
(IS_MPSAFE ? INTR_MPSAFE : 0),
dc_intr, sc, &sc->dc_intrhand);
if (error) {
bus_release_resource(dev, SYS_RES_IRQ, 0, sc->dc_irq);
bus_release_resource(dev, DC_RES, DC_RID, sc->dc_res);
printf("dc%d: couldn't set up irq\n", unit);
}
Teach the 'dc' driver how to pick up settings left over by the SRM on alpha systems. This is an expedient if not entirely elegant solution to the problem. Submitted by: gallatin Approved by: jkh
2000-03-11 05:20:56 +00:00
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
fail:
Remove bogus locking from dc(4). Instead, move interrupt allocation and ether_ifattach() to end. This fixes a "could sleep" case and simplifies error exit cases as well. Also be sure to set errno and clean up resources in !mac error case. Tested by: Ryan Beasley
2003-01-14 19:31:27 +00:00
if (error != 0)
mtx_destroy(&sc->dc_mtx);
return (error);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
}
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static int
dc_detach(dev)
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
device_t dev;
{
struct dc_softc *sc;
struct ifnet *ifp;
Add support for parsing the media blocks from the SROM on 21143 adapters. This is necessary in order to make this driver work with the built-in ethernet on the alpha Miata machines. These systems have a 21143-PC chip on-board and optional daughtercards with either a 10/100 MII transceiver or a 10baseT/10base2 transceiver. In both cases, you need to twiddle the GPIO bits on the controller in order to turn the transceivers on, and you have to read the media info from the SROM in order to find out what bits to twiddle.
2000-10-05 17:36:14 +00:00
struct dc_mediainfo *m;
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
sc = device_get_softc(dev);
First round of converting network drivers from spls to mutexes. This takes care of all the 10/100 and gigE PCI drivers that I've done. Next will be the wireless drivers, then the USB ones. I may pick up some stragglers along the way. I'm sort of playing this by ear: if anyone spots any places where I've screwed up horribly, please let me know.
2000-10-13 17:54:19 +00:00
DC_LOCK(sc);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
ifp = &sc->arpcom.ac_if;
dc_stop(sc);
o track either_ifattach/ether_ifdetach API changes o use if_input for input packet processing o don't strip the Ethernet header for input packets o use BPF_* macros bpf tapping o call ether_ioctl to handle default ioctl case o track vlan changes Reviewed by: many Approved by: re
2002-11-14 23:49:09 +00:00
ether_ifdetach(ifp);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
bus_generic_detach(dev);
device_delete_child(dev, sc->dc_miibus);
bus_teardown_intr(dev, sc->dc_irq, sc->dc_intrhand);
bus_release_resource(dev, SYS_RES_IRQ, 0, sc->dc_irq);
bus_release_resource(dev, DC_RES, DC_RID, sc->dc_res);
contigfree(sc->dc_ldata, sizeof(struct dc_list_data), M_DEVBUF);
if (sc->dc_pnic_rx_buf != NULL)
free(sc->dc_pnic_rx_buf, M_DEVBUF);
Add support for parsing the media blocks from the SROM on 21143 adapters. This is necessary in order to make this driver work with the built-in ethernet on the alpha Miata machines. These systems have a 21143-PC chip on-board and optional daughtercards with either a 10/100 MII transceiver or a 10baseT/10base2 transceiver. In both cases, you need to twiddle the GPIO bits on the controller in order to turn the transceivers on, and you have to read the media info from the SROM in order to find out what bits to twiddle.
2000-10-05 17:36:14 +00:00
while(sc->dc_mi != NULL) {
m = sc->dc_mi->dc_next;
free(sc->dc_mi, M_DEVBUF);
sc->dc_mi = m;
}
Ooops. Need to free dc_srom on detach to not leak memory. Pointy Hat to: The Mad Redhead of Niwot
2002-10-07 05:29:13 +00:00
free(sc->dc_srom, M_DEVBUF);
Add support for parsing the media blocks from the SROM on 21143 adapters. This is necessary in order to make this driver work with the built-in ethernet on the alpha Miata machines. These systems have a 21143-PC chip on-board and optional daughtercards with either a 10/100 MII transceiver or a 10baseT/10base2 transceiver. In both cases, you need to twiddle the GPIO bits on the controller in order to turn the transceivers on, and you have to read the media info from the SROM in order to find out what bits to twiddle.
2000-10-05 17:36:14 +00:00
First round of converting network drivers from spls to mutexes. This takes care of all the 10/100 and gigE PCI drivers that I've done. Next will be the wireless drivers, then the USB ones. I may pick up some stragglers along the way. I'm sort of playing this by ear: if anyone spots any places where I've screwed up horribly, please let me know.
2000-10-13 17:54:19 +00:00
DC_UNLOCK(sc);
mtx_destroy(&sc->dc_mtx);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
return(0);
}
/*
* Initialize the transmit descriptors.
*/
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static int
dc_list_tx_init(sc)
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
struct dc_softc *sc;
{
struct dc_chain_data *cd;
struct dc_list_data *ld;
For i386 architecture, remove an expensive m_devget() (and the underlying unaligned bcopy) on incoming packets that are already available (albeit unaligned) in a buffer. The performance improvement varies, depending on CPU and memory speed, but can be quite large especially on slow CPUs. I have seen over 50% increase on forwarding speed on the sis driver for the 486/133 (embedded systems), which does exactly the same thing. The behaviour is controlled by a sysctl variable, hw.dc_quick which defaults to 1. Set it to 0 to restore the old behaviour. After running a few experiments (in userland, though) I am convinced that doing the m_devget() is detrimental to performance in almost all cases. Even if your CPU has degraded performance with misaligned data, the bcopy() in the driver has the same overhead due to misaligment as the one that you save in the uiomove(), plus you do one extra copy and pollute the cache. But more often than not, you do not even have to touch the payload, e.g. when you are forwarding packets, and even in the often-cited case of NFS, you often end up passing a pointer to the payload to the disk controller. In any case, you can play with the sysctl variable to toggle between the two behaviours, and see if it makes a difference. MFC-after: 3 days
2001-11-29 22:46:48 +00:00
int i, nexti;
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
cd = &sc->dc_cdata;
ld = sc->dc_ldata;
for (i = 0; i < DC_TX_LIST_CNT; i++) {
Per jlemon request, reintroduce some printf() when an mbuf allocation fails, and fix (i hope) a couple of style bugs. I believe these printf() are extremely dangerous because now they can occur on every incoming packet and are not rate limited. They were meant to warn the sysadmin about lack of resources, but now they can become a nice way to panic your system under load. Other drivers (e.g. the fxp driver) have nothing like this. There is a pending discussion on putting this kind of warnings elsewhere, and I hope we can fix this soon.
2001-11-29 23:47:47 +00:00
nexti = (i == (DC_TX_LIST_CNT - 1)) ? 0 : i+1;
For i386 architecture, remove an expensive m_devget() (and the underlying unaligned bcopy) on incoming packets that are already available (albeit unaligned) in a buffer. The performance improvement varies, depending on CPU and memory speed, but can be quite large especially on slow CPUs. I have seen over 50% increase on forwarding speed on the sis driver for the 486/133 (embedded systems), which does exactly the same thing. The behaviour is controlled by a sysctl variable, hw.dc_quick which defaults to 1. Set it to 0 to restore the old behaviour. After running a few experiments (in userland, though) I am convinced that doing the m_devget() is detrimental to performance in almost all cases. Even if your CPU has degraded performance with misaligned data, the bcopy() in the driver has the same overhead due to misaligment as the one that you save in the uiomove(), plus you do one extra copy and pollute the cache. But more often than not, you do not even have to touch the payload, e.g. when you are forwarding packets, and even in the often-cited case of NFS, you often end up passing a pointer to the payload to the disk controller. In any case, you can play with the sysctl variable to toggle between the two behaviours, and see if it makes a difference. MFC-after: 3 days
2001-11-29 22:46:48 +00:00
ld->dc_tx_list[i].dc_next = vtophys(&ld->dc_tx_list[nexti]);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
cd->dc_tx_chain[i] = NULL;
ld->dc_tx_list[i].dc_data = 0;
ld->dc_tx_list[i].dc_ctl = 0;
}
cd->dc_tx_prod = cd->dc_tx_cons = cd->dc_tx_cnt = 0;
return(0);
}
/*
* Initialize the RX descriptors and allocate mbufs for them. Note that
* we arrange the descriptors in a closed ring, so that the last descriptor
* points back to the first.
*/
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static int
dc_list_rx_init(sc)
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
struct dc_softc *sc;
{
struct dc_chain_data *cd;
struct dc_list_data *ld;
For i386 architecture, remove an expensive m_devget() (and the underlying unaligned bcopy) on incoming packets that are already available (albeit unaligned) in a buffer. The performance improvement varies, depending on CPU and memory speed, but can be quite large especially on slow CPUs. I have seen over 50% increase on forwarding speed on the sis driver for the 486/133 (embedded systems), which does exactly the same thing. The behaviour is controlled by a sysctl variable, hw.dc_quick which defaults to 1. Set it to 0 to restore the old behaviour. After running a few experiments (in userland, though) I am convinced that doing the m_devget() is detrimental to performance in almost all cases. Even if your CPU has degraded performance with misaligned data, the bcopy() in the driver has the same overhead due to misaligment as the one that you save in the uiomove(), plus you do one extra copy and pollute the cache. But more often than not, you do not even have to touch the payload, e.g. when you are forwarding packets, and even in the often-cited case of NFS, you often end up passing a pointer to the payload to the disk controller. In any case, you can play with the sysctl variable to toggle between the two behaviours, and see if it makes a difference. MFC-after: 3 days
2001-11-29 22:46:48 +00:00
int i, nexti;
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
cd = &sc->dc_cdata;
ld = sc->dc_ldata;
for (i = 0; i < DC_RX_LIST_CNT; i++) {
if (dc_newbuf(sc, i, NULL) == ENOBUFS)
return(ENOBUFS);
Per jlemon request, reintroduce some printf() when an mbuf allocation fails, and fix (i hope) a couple of style bugs. I believe these printf() are extremely dangerous because now they can occur on every incoming packet and are not rate limited. They were meant to warn the sysadmin about lack of resources, but now they can become a nice way to panic your system under load. Other drivers (e.g. the fxp driver) have nothing like this. There is a pending discussion on putting this kind of warnings elsewhere, and I hope we can fix this soon.
2001-11-29 23:47:47 +00:00
nexti = (i == (DC_RX_LIST_CNT - 1)) ? 0 : i+1;
For i386 architecture, remove an expensive m_devget() (and the underlying unaligned bcopy) on incoming packets that are already available (albeit unaligned) in a buffer. The performance improvement varies, depending on CPU and memory speed, but can be quite large especially on slow CPUs. I have seen over 50% increase on forwarding speed on the sis driver for the 486/133 (embedded systems), which does exactly the same thing. The behaviour is controlled by a sysctl variable, hw.dc_quick which defaults to 1. Set it to 0 to restore the old behaviour. After running a few experiments (in userland, though) I am convinced that doing the m_devget() is detrimental to performance in almost all cases. Even if your CPU has degraded performance with misaligned data, the bcopy() in the driver has the same overhead due to misaligment as the one that you save in the uiomove(), plus you do one extra copy and pollute the cache. But more often than not, you do not even have to touch the payload, e.g. when you are forwarding packets, and even in the often-cited case of NFS, you often end up passing a pointer to the payload to the disk controller. In any case, you can play with the sysctl variable to toggle between the two behaviours, and see if it makes a difference. MFC-after: 3 days
2001-11-29 22:46:48 +00:00
ld->dc_rx_list[i].dc_next = vtophys(&ld->dc_rx_list[nexti]);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
}
cd->dc_rx_prod = 0;
return(0);
}
/*
* Initialize an RX descriptor and attach an MBUF cluster.
*/
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static int
dc_newbuf(sc, i, m)
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
struct dc_softc *sc;
int i;
struct mbuf *m;
{
struct mbuf *m_new = NULL;
struct dc_desc *c;
c = &sc->dc_ldata->dc_rx_list[i];
if (m == NULL) {
Remove M_TRYWAIT/M_WAITOK/M_WAIT. Callers should use 0. Merge M_NOWAIT/M_DONTWAIT into a single flag M_NOWAIT.
2003-01-21 08:56:16 +00:00
MGETHDR(m_new, M_NOWAIT, MT_DATA);
Remove error messages on mbuf allocation failures, now this is done more safely in kern/subr_mbuf.c Two-days'-delay-thanks-to: @home shutting down service
2001-12-04 02:30:53 +00:00
if (m_new == NULL)
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
return(ENOBUFS);
Remove M_TRYWAIT/M_WAITOK/M_WAIT. Callers should use 0. Merge M_NOWAIT/M_DONTWAIT into a single flag M_NOWAIT.
2003-01-21 08:56:16 +00:00
MCLGET(m_new, M_NOWAIT);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
if (!(m_new->m_flags & M_EXT)) {
m_freem(m_new);
return(ENOBUFS);
}
m_new->m_len = m_new->m_pkthdr.len = MCLBYTES;
} else {
m_new = m;
m_new->m_len = m_new->m_pkthdr.len = MCLBYTES;
m_new->m_data = m_new->m_ext.ext_buf;
}
m_adj(m_new, sizeof(u_int64_t));
/*
* If this is a PNIC chip, zero the buffer. This is part
* of the workaround for the receive bug in the 82c168 and
* 82c169 chips.
*/
if (sc->dc_flags & DC_PNIC_RX_BUG_WAR)
bzero((char *)mtod(m_new, char *), m_new->m_len);
sc->dc_cdata.dc_rx_chain[i] = m_new;
c->dc_data = vtophys(mtod(m_new, caddr_t));
c->dc_ctl = DC_RXCTL_RLINK | DC_RXLEN;
c->dc_status = DC_RXSTAT_OWN;
return(0);
}
/*
* Grrrrr.
* The PNIC chip has a terrible bug in it that manifests itself during
* periods of heavy activity. The exact mode of failure if difficult to
* pinpoint: sometimes it only happens in promiscuous mode, sometimes it
* will happen on slow machines. The bug is that sometimes instead of
* uploading one complete frame during reception, it uploads what looks
* like the entire contents of its FIFO memory. The frame we want is at
* the end of the whole mess, but we never know exactly how much data has
* been uploaded, so salvaging the frame is hard.
*
* There is only one way to do it reliably, and it's disgusting.
* Here's what we know:
*
* - We know there will always be somewhere between one and three extra
* descriptors uploaded.
*
* - We know the desired received frame will always be at the end of the
* total data upload.
*
* - We know the size of the desired received frame because it will be
* provided in the length field of the status word in the last descriptor.
*
* Here's what we do:
*
* - When we allocate buffers for the receive ring, we bzero() them.
* This means that we know that the buffer contents should be all
* zeros, except for data uploaded by the chip.
*
* - We also force the PNIC chip to upload frames that include the
* ethernet CRC at the end.
*
* - We gather all of the bogus frame data into a single buffer.
*
* - We then position a pointer at the end of this buffer and scan
* backwards until we encounter the first non-zero byte of data.
* This is the end of the received frame. We know we will encounter
* some data at the end of the frame because the CRC will always be
* there, so even if the sender transmits a packet of all zeros,
* we won't be fooled.
*
* - We know the size of the actual received frame, so we subtract
* that value from the current pointer location. This brings us
* to the start of the actual received packet.
*
* - We copy this into an mbuf and pass it on, along with the actual
* frame length.
*
* The performance hit is tremendous, but it beats dropping frames all
* the time.
*/
#define DC_WHOLEFRAME (DC_RXSTAT_FIRSTFRAG|DC_RXSTAT_LASTFRAG)
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static void
dc_pnic_rx_bug_war(sc, idx)
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
struct dc_softc *sc;
int idx;
{
struct dc_desc *cur_rx;
struct dc_desc *c = NULL;
struct mbuf *m = NULL;
unsigned char *ptr;
int i, total_len;
u_int32_t rxstat = 0;
i = sc->dc_pnic_rx_bug_save;
cur_rx = &sc->dc_ldata->dc_rx_list[idx];
ptr = sc->dc_pnic_rx_buf;
bzero(ptr, sizeof(DC_RXLEN * 5));
/* Copy all the bytes from the bogus buffers. */
while (1) {
c = &sc->dc_ldata->dc_rx_list[i];
rxstat = c->dc_status;
m = sc->dc_cdata.dc_rx_chain[i];
bcopy(mtod(m, char *), ptr, DC_RXLEN);
ptr += DC_RXLEN;
/* If this is the last buffer, break out. */
if (i == idx || rxstat & DC_RXSTAT_LASTFRAG)
break;
dc_newbuf(sc, i, m);
DC_INC(i, DC_RX_LIST_CNT);
}
/* Find the length of the actual receive frame. */
total_len = DC_RXBYTES(rxstat);
/* Scan backwards until we hit a non-zero byte. */
while(*ptr == 0x00)
ptr--;
/* Round off. */
if ((uintptr_t)(ptr) & 0x3)
ptr -= 1;
/* Now find the start of the frame. */
ptr -= total_len;
if (ptr < sc->dc_pnic_rx_buf)
ptr = sc->dc_pnic_rx_buf;
/*
* Now copy the salvaged frame to the last mbuf and fake up
* the status word to make it look like a successful
Make consistent; turn spaces into tabs where there is a mixture.
2002-09-04 18:14:17 +00:00
* frame reception.
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
*/
dc_newbuf(sc, i, m);
bcopy(ptr, mtod(m, char *), total_len);
cur_rx->dc_status = rxstat | DC_RXSTAT_FIRSTFRAG;
return;
}
It appears that under certain circumstances that I still can't quite pin down, the dc driver and receiver can fall out of sync with one another, resulting in a condition where the chip continues to receive packets but the driver never notices. Normally, the receive handler checks each descriptor starting from the current producer index to see if the chip has relinquished ownership, indicating that a packet has been received. The driver hands the packet off to ether_input() and then prepares the descriptor to receive another frame before moving on to the next descriptor in the ring. But sometimes, the chip appears to skip a descriptor. This leaves the driver testing the status word in a descriptor that never gets updated. The driver still gets "RX done" interrupts but never advances further into the RX ring, until the ring fills up and the chip interrupts again to signal an error condition. Sometimes, the driver will remain in this desynchronized state, resulting in spotty performance until the interface is reset. Fortunately, it's fairly simple to detect this condition: if we call the rxeof routine but the number of received packets doesn't increase, we suspect that there could be a problem. In this case, we call a new routine called dc_rx_resync(), which scans ahead in the RX ring to see if there's a frame waiting for us somewhere beyond that the driver thinks is the current producer index. If it finds one, it bumps up the index and calls the rxeof handler again to snarf up the packet and bring the driver back in sync with the chip. (It may actually do this several times in the event that there's more than one "hole" in the ring.) So far the only card supported by if_dc which has exhibited this problem is a LinkSys LNE100TX v2.0 (82c115 PNIC II), and it only seems to happen on one particular system, however the fix is general enough and has low enough overhead that we may as well apply it for all supported chipsets. I also implemented the same fix for the 3Com xl driver, which is apparently vulnerable to the same problem. Problem originally noted and patch tested by: Matt Dillon
2000-01-03 15:28:47 +00:00
/*
* This routine searches the RX ring for dirty descriptors in the
* event that the rxeof routine falls out of sync with the chip's
* current descriptor pointer. This may happen sometimes as a result
* of a "no RX buffer available" condition that happens when the chip
* consumes all of the RX buffers before the driver has a chance to
* process the RX ring. This routine may need to be called more than
* once to bring the driver back in sync with the chip, however we
* should still be getting RX DONE interrupts to drive the search
* for new packets in the RX ring, so we should catch up eventually.
*/
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static int
dc_rx_resync(sc)
It appears that under certain circumstances that I still can't quite pin down, the dc driver and receiver can fall out of sync with one another, resulting in a condition where the chip continues to receive packets but the driver never notices. Normally, the receive handler checks each descriptor starting from the current producer index to see if the chip has relinquished ownership, indicating that a packet has been received. The driver hands the packet off to ether_input() and then prepares the descriptor to receive another frame before moving on to the next descriptor in the ring. But sometimes, the chip appears to skip a descriptor. This leaves the driver testing the status word in a descriptor that never gets updated. The driver still gets "RX done" interrupts but never advances further into the RX ring, until the ring fills up and the chip interrupts again to signal an error condition. Sometimes, the driver will remain in this desynchronized state, resulting in spotty performance until the interface is reset. Fortunately, it's fairly simple to detect this condition: if we call the rxeof routine but the number of received packets doesn't increase, we suspect that there could be a problem. In this case, we call a new routine called dc_rx_resync(), which scans ahead in the RX ring to see if there's a frame waiting for us somewhere beyond that the driver thinks is the current producer index. If it finds one, it bumps up the index and calls the rxeof handler again to snarf up the packet and bring the driver back in sync with the chip. (It may actually do this several times in the event that there's more than one "hole" in the ring.) So far the only card supported by if_dc which has exhibited this problem is a LinkSys LNE100TX v2.0 (82c115 PNIC II), and it only seems to happen on one particular system, however the fix is general enough and has low enough overhead that we may as well apply it for all supported chipsets. I also implemented the same fix for the 3Com xl driver, which is apparently vulnerable to the same problem. Problem originally noted and patch tested by: Matt Dillon
2000-01-03 15:28:47 +00:00
struct dc_softc *sc;
{
int i, pos;
struct dc_desc *cur_rx;
pos = sc->dc_cdata.dc_rx_prod;
for (i = 0; i < DC_RX_LIST_CNT; i++) {
cur_rx = &sc->dc_ldata->dc_rx_list[pos];
if (!(cur_rx->dc_status & DC_RXSTAT_OWN))
break;
DC_INC(pos, DC_RX_LIST_CNT);
}
/* If the ring really is empty, then just return. */
if (i == DC_RX_LIST_CNT)
return(0);
/* We've fallen behing the chip: catch it. */
sc->dc_cdata.dc_rx_prod = pos;
return(EAGAIN);
}
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
/*
* A frame has been uploaded: pass the resulting mbuf chain up to
* the higher level protocols.
*/
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static void
dc_rxeof(sc)
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
struct dc_softc *sc;
{
Make consistent; turn spaces into tabs where there is a mixture.
2002-09-04 18:14:17 +00:00
struct mbuf *m;
struct ifnet *ifp;
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
struct dc_desc *cur_rx;
int i, total_len = 0;
u_int32_t rxstat;
ifp = &sc->arpcom.ac_if;
i = sc->dc_cdata.dc_rx_prod;
while(!(sc->dc_ldata->dc_rx_list[i].dc_status & DC_RXSTAT_OWN)) {
Device Polling code for -current. Non-SMP, i386-only, no polling in the idle loop at the moment. To use this code you must compile a kernel with options DEVICE_POLLING and at runtime enable polling with sysctl kern.polling.enable=1 The percentage of CPU reserved to userland can be set with sysctl kern.polling.user_frac=NN (default is 50) while the remainder is used by polling device drivers and netisr's. These are the only two variables that you should need to touch. There are a few more parameters in kern.polling but the default values are adequate for all purposes. See the code in kern_poll.c for more details on them. Polling in the idle loop will be implemented shortly by introducing a kernel thread which does the job. Until then, the amount of CPU dedicated to polling will never exceed (100-user_frac). The equivalent (actually, better) code for -stable is at http://info.iet.unipi.it/~luigi/polling/ and also supports polling in the idle loop. NOTE to Alpha developers: There is really nothing in this code that is i386-specific. If you move the 2 lines supporting the new option from sys/conf/{files,options}.i386 to sys/conf/{files,options} I am pretty sure that this should work on the Alpha as well, just that I do not have a suitable test box to try it. If someone feels like trying it, I would appreciate it. NOTE to other developers: sure some things could be done better, and as always I am open to constructive criticism, which a few of you have already given and I greatly appreciated. However, before proposing radical architectural changes, please take some time to possibly try out this code, or at the very least read the comments in kern_poll.c, especially re. the reason why I am using a soft netisr and cannot (I believe) replace it with a simple timeout. Quick description of files touched by this commit: sys/conf/files.i386 new file kern/kern_poll.c sys/conf/options.i386 new option sys/i386/i386/trap.c poll in trap (disabled by default) sys/kern/kern_clock.c initialization and hardclock hooks. sys/kern/kern_intr.c minor swi_net changes sys/kern/kern_poll.c the bulk of the code. sys/net/if.h new flag sys/net/if_var.h declaration for functions used in device drivers. sys/net/netisr.h NETISR_POLL sys/dev/fxp/if_fxp.c sys/dev/fxp/if_fxpvar.h sys/pci/if_dc.c sys/pci/if_dcreg.h sys/pci/if_sis.c sys/pci/if_sisreg.h device driver modifications
2001-12-14 17:56:12 +00:00
#ifdef DEVICE_POLLING
Increase size of ifnet.if_flags from 16 bits (short) to 32 bits (int). To avoid breaking application ABI use unused ifreq.ifru_flags[1] for upper 16 bits in SIOCSIFFLAGS and SIOCGIFFLAGS ioctl's. Reviewed by: -hackers, -net
2002-08-18 07:05:00 +00:00
if (ifp->if_flags & IFF_POLLING) {
Device Polling code for -current. Non-SMP, i386-only, no polling in the idle loop at the moment. To use this code you must compile a kernel with options DEVICE_POLLING and at runtime enable polling with sysctl kern.polling.enable=1 The percentage of CPU reserved to userland can be set with sysctl kern.polling.user_frac=NN (default is 50) while the remainder is used by polling device drivers and netisr's. These are the only two variables that you should need to touch. There are a few more parameters in kern.polling but the default values are adequate for all purposes. See the code in kern_poll.c for more details on them. Polling in the idle loop will be implemented shortly by introducing a kernel thread which does the job. Until then, the amount of CPU dedicated to polling will never exceed (100-user_frac). The equivalent (actually, better) code for -stable is at http://info.iet.unipi.it/~luigi/polling/ and also supports polling in the idle loop. NOTE to Alpha developers: There is really nothing in this code that is i386-specific. If you move the 2 lines supporting the new option from sys/conf/{files,options}.i386 to sys/conf/{files,options} I am pretty sure that this should work on the Alpha as well, just that I do not have a suitable test box to try it. If someone feels like trying it, I would appreciate it. NOTE to other developers: sure some things could be done better, and as always I am open to constructive criticism, which a few of you have already given and I greatly appreciated. However, before proposing radical architectural changes, please take some time to possibly try out this code, or at the very least read the comments in kern_poll.c, especially re. the reason why I am using a soft netisr and cannot (I believe) replace it with a simple timeout. Quick description of files touched by this commit: sys/conf/files.i386 new file kern/kern_poll.c sys/conf/options.i386 new option sys/i386/i386/trap.c poll in trap (disabled by default) sys/kern/kern_clock.c initialization and hardclock hooks. sys/kern/kern_intr.c minor swi_net changes sys/kern/kern_poll.c the bulk of the code. sys/net/if.h new flag sys/net/if_var.h declaration for functions used in device drivers. sys/net/netisr.h NETISR_POLL sys/dev/fxp/if_fxp.c sys/dev/fxp/if_fxpvar.h sys/pci/if_dc.c sys/pci/if_dcreg.h sys/pci/if_sis.c sys/pci/if_sisreg.h device driver modifications
2001-12-14 17:56:12 +00:00
if (sc->rxcycles <= 0)
break;
sc->rxcycles--;
}
#endif /* DEVICE_POLLING */
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
cur_rx = &sc->dc_ldata->dc_rx_list[i];
rxstat = cur_rx->dc_status;
m = sc->dc_cdata.dc_rx_chain[i];
total_len = DC_RXBYTES(rxstat);
if (sc->dc_flags & DC_PNIC_RX_BUG_WAR) {
if ((rxstat & DC_WHOLEFRAME) != DC_WHOLEFRAME) {
if (rxstat & DC_RXSTAT_FIRSTFRAG)
sc->dc_pnic_rx_bug_save = i;
if ((rxstat & DC_RXSTAT_LASTFRAG) == 0) {
DC_INC(i, DC_RX_LIST_CNT);
continue;
}
dc_pnic_rx_bug_war(sc, i);
rxstat = cur_rx->dc_status;
total_len = DC_RXBYTES(rxstat);
}
}
sc->dc_cdata.dc_rx_chain[i] = NULL;
/*
* If an error occurs, update stats, clear the
* status word and leave the mbuf cluster in place:
* it should simply get re-used next time this descriptor
Add VLAN for the dc(4) driver (ie long frame). The patch is 2 parts. One to notify the system that the MTU for VLAN can be 1500 so the vlan will automatically be configured with a 1500 MTU the other is to ignore the error case if the received frame is to long. The frame size notification came from code in the SIS driver, and the support for long frames derived from the NetBSD Tulip driver. Tested on: 4 port D-Link adapter DFE-570TX 4 Intel 21143 Netgear card with 82c169 PNIC 10/100BaseTX Reviewed by: ru (manpage), wpaul (not objected to), archie Approved by: imp Obtained from: NetBSD
2002-01-16 16:33:58 +00:00
* comes up in the ring. However, don't report long
* frames as errors since they could be vlans
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
*/
Add VLAN for the dc(4) driver (ie long frame). The patch is 2 parts. One to notify the system that the MTU for VLAN can be 1500 so the vlan will automatically be configured with a 1500 MTU the other is to ignore the error case if the received frame is to long. The frame size notification came from code in the SIS driver, and the support for long frames derived from the NetBSD Tulip driver. Tested on: 4 port D-Link adapter DFE-570TX 4 Intel 21143 Netgear card with 82c169 PNIC 10/100BaseTX Reviewed by: ru (manpage), wpaul (not objected to), archie Approved by: imp Obtained from: NetBSD
2002-01-16 16:33:58 +00:00
if ((rxstat & DC_RXSTAT_RXERR)){
if (!(rxstat & DC_RXSTAT_GIANT) ||
(rxstat & (DC_RXSTAT_CRCERR | DC_RXSTAT_DRIBBLE |
DC_RXSTAT_MIIERE | DC_RXSTAT_COLLSEEN |
DC_RXSTAT_RUNT | DC_RXSTAT_DE))) {
ifp->if_ierrors++;
if (rxstat & DC_RXSTAT_COLLSEEN)
ifp->if_collisions++;
dc_newbuf(sc, i, m);
if (rxstat & DC_RXSTAT_CRCERR) {
DC_INC(i, DC_RX_LIST_CNT);
continue;
} else {
dc_init(sc);
return;
}
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
}
}
/* No errors; receive the packet. */
total_len -= ETHER_CRC_LEN;
For i386 architecture, remove an expensive m_devget() (and the underlying unaligned bcopy) on incoming packets that are already available (albeit unaligned) in a buffer. The performance improvement varies, depending on CPU and memory speed, but can be quite large especially on slow CPUs. I have seen over 50% increase on forwarding speed on the sis driver for the 486/133 (embedded systems), which does exactly the same thing. The behaviour is controlled by a sysctl variable, hw.dc_quick which defaults to 1. Set it to 0 to restore the old behaviour. After running a few experiments (in userland, though) I am convinced that doing the m_devget() is detrimental to performance in almost all cases. Even if your CPU has degraded performance with misaligned data, the bcopy() in the driver has the same overhead due to misaligment as the one that you save in the uiomove(), plus you do one extra copy and pollute the cache. But more often than not, you do not even have to touch the payload, e.g. when you are forwarding packets, and even in the often-cited case of NFS, you often end up passing a pointer to the payload to the disk controller. In any case, you can play with the sysctl variable to toggle between the two behaviours, and see if it makes a difference. MFC-after: 3 days
2001-11-29 22:46:48 +00:00
#ifdef __i386__
/*
* On the x86 we do not have alignment problems, so try to
* allocate a new buffer for the receive ring, and pass up
* the one where the packet is already, saving the expensive
* copy done in m_devget().
* If we are on an architecture with alignment problems, or
* if the allocation fails, then use m_devget and leave the
* existing buffer in the receive ring.
*/
if (dc_quick && dc_newbuf(sc, i, NULL) == 0) {
m->m_pkthdr.rcvif = ifp;
m->m_pkthdr.len = m->m_len = total_len;
DC_INC(i, DC_RX_LIST_CNT);
} else
#endif
{
struct mbuf *m0;
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
For i386 architecture, remove an expensive m_devget() (and the underlying unaligned bcopy) on incoming packets that are already available (albeit unaligned) in a buffer. The performance improvement varies, depending on CPU and memory speed, but can be quite large especially on slow CPUs. I have seen over 50% increase on forwarding speed on the sis driver for the 486/133 (embedded systems), which does exactly the same thing. The behaviour is controlled by a sysctl variable, hw.dc_quick which defaults to 1. Set it to 0 to restore the old behaviour. After running a few experiments (in userland, though) I am convinced that doing the m_devget() is detrimental to performance in almost all cases. Even if your CPU has degraded performance with misaligned data, the bcopy() in the driver has the same overhead due to misaligment as the one that you save in the uiomove(), plus you do one extra copy and pollute the cache. But more often than not, you do not even have to touch the payload, e.g. when you are forwarding packets, and even in the often-cited case of NFS, you often end up passing a pointer to the payload to the disk controller. In any case, you can play with the sysctl variable to toggle between the two behaviours, and see if it makes a difference. MFC-after: 3 days
2001-11-29 22:46:48 +00:00
m0 = m_devget(mtod(m, char *), total_len,
ETHER_ALIGN, ifp, NULL);
dc_newbuf(sc, i, m);
DC_INC(i, DC_RX_LIST_CNT);
if (m0 == NULL) {
ifp->if_ierrors++;
continue;
}
m = m0;
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
}
ifp->if_ipackets++;
o track either_ifattach/ether_ifdetach API changes o use if_input for input packet processing o don't strip the Ethernet header for input packets o use BPF_* macros bpf tapping o call ether_ioctl to handle default ioctl case o track vlan changes Reviewed by: many Approved by: re
2002-11-14 23:49:09 +00:00
(*ifp->if_input)(ifp, m);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
}
sc->dc_cdata.dc_rx_prod = i;
}
/*
* A frame was downloaded to the chip. It's safe for us to clean up
* the list buffers.
*/
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static void
dc_txeof(sc)
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
struct dc_softc *sc;
{
struct dc_desc *cur_tx = NULL;
struct ifnet *ifp;
int idx;
ifp = &sc->arpcom.ac_if;
/*
* Go through our tx list and free mbufs for those
* frames that have been transmitted.
*/
idx = sc->dc_cdata.dc_tx_cons;
while(idx != sc->dc_cdata.dc_tx_prod) {
u_int32_t txstat;
cur_tx = &sc->dc_ldata->dc_tx_list[idx];
txstat = cur_tx->dc_status;
if (txstat & DC_TXSTAT_OWN)
break;
if (!(cur_tx->dc_ctl & DC_TXCTL_LASTFRAG) ||
cur_tx->dc_ctl & DC_TXCTL_SETUP) {
if (cur_tx->dc_ctl & DC_TXCTL_SETUP) {
/*
* Yes, the PNIC is so brain damaged
* that it will sometimes generate a TX
* underrun error while DMAing the RX
* filter setup frame. If we detect this,
* we have to send the setup frame again,
* or else the filter won't be programmed
* correctly.
*/
if (DC_IS_PNIC(sc)) {
if (txstat & DC_TXSTAT_ERRSUM)
dc_setfilt(sc);
}
sc->dc_cdata.dc_tx_chain[idx] = NULL;
}
Fix if_timer logic to make sure that there is always a timeout pending if there are packets queued for transmission. MFC after: 3 days
2002-06-30 22:05:46 +00:00
sc->dc_cdata.dc_tx_cnt--;
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
DC_INC(idx, DC_TX_LIST_CNT);
continue;
}
Fix the "conexant chips don't work in full duplexmode" problem. According to Phil Kernick: "The problem is that in full duplex mode, the Conexant chip always reports a carrier lost error, even when the frame is successfully sent. So, if we have a Conexant chip, then ignore carrier lost when in full duplex mode." Since the Xircom chips seem to have the same issue and since we already have a workaround for this, just expand the workaround test to also check for DC_IS_CONEXANT().
2001-12-19 18:23:45 +00:00
if (DC_IS_XIRCOM(sc) || DC_IS_CONEXANT(sc)) {
NEWCARD/Cardbus - This commit adds support for Xircom X3201 based cardbus cards. Support for the TDK 78Q2120 MII is also added. IBM Etherjet, Intel and Xircom cards uses these chips. Note that as a result of this commit, some Intel/DEC 21143 based cardbus cards will also attach, but not get link. That is being looked at.
2000-10-19 08:34:32 +00:00
/*
* XXX: Why does my Xircom taunt me so?
* For some reason it likes setting the CARRLOST flag
Fix the "conexant chips don't work in full duplexmode" problem. According to Phil Kernick: "The problem is that in full duplex mode, the Conexant chip always reports a carrier lost error, even when the frame is successfully sent. So, if we have a Conexant chip, then ignore carrier lost when in full duplex mode." Since the Xircom chips seem to have the same issue and since we already have a workaround for this, just expand the workaround test to also check for DC_IS_CONEXANT().
2001-12-19 18:23:45 +00:00
* even when the carrier is there. wtf?!?
* Who knows, but Conexant chips have the
* same problem. Maybe they took lessons
* from Xircom.
*/
NEWCARD/Cardbus - This commit adds support for Xircom X3201 based cardbus cards. Support for the TDK 78Q2120 MII is also added. IBM Etherjet, Intel and Xircom cards uses these chips. Note that as a result of this commit, some Intel/DEC 21143 based cardbus cards will also attach, but not get link. That is being looked at.
2000-10-19 08:34:32 +00:00
if (/*sc->dc_type == DC_TYPE_21143 &&*/
sc->dc_pmode == DC_PMODE_MII &&
((txstat & 0xFFFF) & ~(DC_TXSTAT_ERRSUM|
Fix the "conexant chips don't work in full duplexmode" problem. According to Phil Kernick: "The problem is that in full duplex mode, the Conexant chip always reports a carrier lost error, even when the frame is successfully sent. So, if we have a Conexant chip, then ignore carrier lost when in full duplex mode." Since the Xircom chips seem to have the same issue and since we already have a workaround for this, just expand the workaround test to also check for DC_IS_CONEXANT().
2001-12-19 18:23:45 +00:00
DC_TXSTAT_NOCARRIER)))
NEWCARD/Cardbus - This commit adds support for Xircom X3201 based cardbus cards. Support for the TDK 78Q2120 MII is also added. IBM Etherjet, Intel and Xircom cards uses these chips. Note that as a result of this commit, some Intel/DEC 21143 based cardbus cards will also attach, but not get link. That is being looked at.
2000-10-19 08:34:32 +00:00
txstat &= ~DC_TXSTAT_ERRSUM;
} else {
if (/*sc->dc_type == DC_TYPE_21143 &&*/
sc->dc_pmode == DC_PMODE_MII &&
((txstat & 0xFFFF) & ~(DC_TXSTAT_ERRSUM|
Fix the "conexant chips don't work in full duplexmode" problem. According to Phil Kernick: "The problem is that in full duplex mode, the Conexant chip always reports a carrier lost error, even when the frame is successfully sent. So, if we have a Conexant chip, then ignore carrier lost when in full duplex mode." Since the Xircom chips seem to have the same issue and since we already have a workaround for this, just expand the workaround test to also check for DC_IS_CONEXANT().
2001-12-19 18:23:45 +00:00
DC_TXSTAT_NOCARRIER|DC_TXSTAT_CARRLOST)))
NEWCARD/Cardbus - This commit adds support for Xircom X3201 based cardbus cards. Support for the TDK 78Q2120 MII is also added. IBM Etherjet, Intel and Xircom cards uses these chips. Note that as a result of this commit, some Intel/DEC 21143 based cardbus cards will also attach, but not get link. That is being looked at.
2000-10-19 08:34:32 +00:00
txstat &= ~DC_TXSTAT_ERRSUM;
}
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
if (txstat & DC_TXSTAT_ERRSUM) {
ifp->if_oerrors++;
if (txstat & DC_TXSTAT_EXCESSCOLL)
ifp->if_collisions++;
if (txstat & DC_TXSTAT_LATECOLL)
ifp->if_collisions++;
if (!(txstat & DC_TXSTAT_UNDERRUN)) {
dc_init(sc);
return;
}
}
ifp->if_collisions += (txstat & DC_TXSTAT_COLLCNT) >> 3;
ifp->if_opackets++;
if (sc->dc_cdata.dc_tx_chain[idx] != NULL) {
m_freem(sc->dc_cdata.dc_tx_chain[idx]);
sc->dc_cdata.dc_tx_chain[idx] = NULL;
}
sc->dc_cdata.dc_tx_cnt--;
DC_INC(idx, DC_TX_LIST_CNT);
}
Fix if_timer logic to make sure that there is always a timeout pending if there are packets queued for transmission. MFC after: 3 days
2002-06-30 22:05:46 +00:00
if (idx != sc->dc_cdata.dc_tx_cons) {
/* some buffers have been freed */
sc->dc_cdata.dc_tx_cons = idx;
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
ifp->if_flags &= ~IFF_OACTIVE;
Fix if_timer logic to make sure that there is always a timeout pending if there are packets queued for transmission. MFC after: 3 days
2002-06-30 22:05:46 +00:00
}
ifp->if_timer = (sc->dc_cdata.dc_tx_cnt == 0) ? 0 : 5;
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
return;
}
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static void
dc_tick(xsc)
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
void *xsc;
{
struct dc_softc *sc;
struct mii_data *mii;
struct ifnet *ifp;
u_int32_t r;
sc = xsc;
First round of converting network drivers from spls to mutexes. This takes care of all the 10/100 and gigE PCI drivers that I've done. Next will be the wireless drivers, then the USB ones. I may pick up some stragglers along the way. I'm sort of playing this by ear: if anyone spots any places where I've screwed up horribly, please let me know.
2000-10-13 17:54:19 +00:00
DC_LOCK(sc);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
ifp = &sc->arpcom.ac_if;
mii = device_get_softc(sc->dc_miibus);
if (sc->dc_flags & DC_REDUCED_MII_POLL) {
Rework the support for the internal autonegotiation on the 21143 and workalike chips (Macronix 98713A/98715 and PNIC II). Timing is somewhat critical: you need to bring the link as soon as possible after NWAY is done, and the old one second polling interval was too long. Now we poll every 10th of a second until NWAY completes (at which point we return to the 1 second interval again to keep an eye on the link state). I tested all the other cards I had on hand to make sure I didn't bust any of them and they seem to work (including the MII-based 21143 card). This should fix some autoneg problems with DE500-BA cards and the built-in 10/100 ethernet on some alpha systems. (Now before anyone asks why I never noticed this before, the old code worked just find with the Intel swich I used for testing back in NY. Apparently not all switches are as picky about the timing.)
2000-05-31 05:40:53 +00:00
if (sc->dc_flags & DC_21143_NWAY) {
r = CSR_READ_4(sc, DC_10BTSTAT);
if (IFM_SUBTYPE(mii->mii_media_active) ==
IFM_100_TX && (r & DC_TSTAT_LS100)) {
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
sc->dc_link = 0;
Rework the support for the internal autonegotiation on the 21143 and workalike chips (Macronix 98713A/98715 and PNIC II). Timing is somewhat critical: you need to bring the link as soon as possible after NWAY is done, and the old one second polling interval was too long. Now we poll every 10th of a second until NWAY completes (at which point we return to the 1 second interval again to keep an eye on the link state). I tested all the other cards I had on hand to make sure I didn't bust any of them and they seem to work (including the MII-based 21143 card). This should fix some autoneg problems with DE500-BA cards and the built-in 10/100 ethernet on some alpha systems. (Now before anyone asks why I never noticed this before, the old code worked just find with the Intel swich I used for testing back in NY. Apparently not all switches are as picky about the timing.)
2000-05-31 05:40:53 +00:00
mii_mediachg(mii);
}
if (IFM_SUBTYPE(mii->mii_media_active) ==
IFM_10_T && (r & DC_TSTAT_LS10)) {
sc->dc_link = 0;
mii_mediachg(mii);
}
Tweak the DC_REDUCED_MII_POLL code in dc_mii_tick() for the DC_IS_INTEL() case. The idea is to reduce how often we call mii_tick(), however currently it may not be called often enough, which prevents autonegotiation from being driven correctly. This should improve the chances of successfully autonegotiating media settings on non-MII 21143 NICs. (Still waiting for confirmation from some testers, but the code is clearly wrong in any case.)
1999-12-07 19:18:41 +00:00
if (sc->dc_link == 0)
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
mii_tick(mii);
} else {
Rework the support for the internal autonegotiation on the 21143 and workalike chips (Macronix 98713A/98715 and PNIC II). Timing is somewhat critical: you need to bring the link as soon as possible after NWAY is done, and the old one second polling interval was too long. Now we poll every 10th of a second until NWAY completes (at which point we return to the 1 second interval again to keep an eye on the link state). I tested all the other cards I had on hand to make sure I didn't bust any of them and they seem to work (including the MII-based 21143 card). This should fix some autoneg problems with DE500-BA cards and the built-in 10/100 ethernet on some alpha systems. (Now before anyone asks why I never noticed this before, the old code worked just find with the Intel swich I used for testing back in NY. Apparently not all switches are as picky about the timing.)
2000-05-31 05:40:53 +00:00
r = CSR_READ_4(sc, DC_ISR);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
if ((r & DC_ISR_RX_STATE) == DC_RXSTATE_WAIT &&
Add missing braces. Found by: FlexeLint (phk) Reviewed by: wpaul, phk
2003-02-07 13:55:02 +00:00
sc->dc_cdata.dc_tx_cnt == 0) {
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
mii_tick(mii);
When I tweaked if_dc.c to alter the polling interval for non-MII 21143 chips, I accidentally removed the DC_MII_REDUCED_POLL flag for all 21143 cards. This caused problems with timer-instigated TCP retransmits, which happened to occur at the same time as an MII poll tick on MII-based cards (e.g. D-Link DFE-570TX). Fixed this, plus made some other cleanups. The autoneg fixes for the non-MII cards still work. Also tested the PNIC II now that I have one again.
2000-06-07 17:07:44 +00:00
if (!(mii->mii_media_status & IFM_ACTIVE))
sc->dc_link = 0;
Add missing braces. Found by: FlexeLint (phk) Reviewed by: wpaul, phk
2003-02-07 13:55:02 +00:00
}
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
}
} else
mii_tick(mii);
/*
* When the init routine completes, we expect to be able to send
* packets right away, and in fact the network code will send a
* gratuitous ARP the moment the init routine marks the interface
* as running. However, even though the MAC may have been initialized,
* there may be a delay of a few seconds before the PHY completes
* autonegotiation and the link is brought up. Any transmissions
* made during that delay will be lost. Dealing with this is tricky:
* we can't just pause in the init routine while waiting for the
* PHY to come ready since that would bring the whole system to
* a screeching halt for several seconds.
*
* What we do here is prevent the TX start routine from sending
* any packets until a link has been established. After the
* interface has been initialized, the tick routine will poll
* the state of the PHY until the IFM_ACTIVE flag is set. Until
* that time, packets will stay in the send queue, and once the
* link comes up, they will be flushed out to the wire.
*/
Do not call mii_pollstat() from within device tick routines; the status information is updated by mii_tick(). Pointed out by: wpaul (a while back)
2001-09-29 19:28:31 +00:00
if (!sc->dc_link && mii->mii_media_status & IFM_ACTIVE &&
IFM_SUBTYPE(mii->mii_media_active) != IFM_NONE) {
sc->dc_link++;
if (ifp->if_snd.ifq_head != NULL)
dc_start(ifp);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
}
Rework the support for the internal autonegotiation on the 21143 and workalike chips (Macronix 98713A/98715 and PNIC II). Timing is somewhat critical: you need to bring the link as soon as possible after NWAY is done, and the old one second polling interval was too long. Now we poll every 10th of a second until NWAY completes (at which point we return to the 1 second interval again to keep an eye on the link state). I tested all the other cards I had on hand to make sure I didn't bust any of them and they seem to work (including the MII-based 21143 card). This should fix some autoneg problems with DE500-BA cards and the built-in 10/100 ethernet on some alpha systems. (Now before anyone asks why I never noticed this before, the old code worked just find with the Intel swich I used for testing back in NY. Apparently not all switches are as picky about the timing.)
2000-05-31 05:40:53 +00:00
if (sc->dc_flags & DC_21143_NWAY && !sc->dc_link)
Change the driver to allocate its own callout structure, and modify the interface to use callout_* instead of timeout(). Also add an IS_MPSAFE #define (currently off) which will mark the driver as mpsafe to the upper layers.
2000-11-25 08:00:10 +00:00
callout_reset(&sc->dc_stat_ch, hz/10, dc_tick, sc);
Rework the support for the internal autonegotiation on the 21143 and workalike chips (Macronix 98713A/98715 and PNIC II). Timing is somewhat critical: you need to bring the link as soon as possible after NWAY is done, and the old one second polling interval was too long. Now we poll every 10th of a second until NWAY completes (at which point we return to the 1 second interval again to keep an eye on the link state). I tested all the other cards I had on hand to make sure I didn't bust any of them and they seem to work (including the MII-based 21143 card). This should fix some autoneg problems with DE500-BA cards and the built-in 10/100 ethernet on some alpha systems. (Now before anyone asks why I never noticed this before, the old code worked just find with the Intel swich I used for testing back in NY. Apparently not all switches are as picky about the timing.)
2000-05-31 05:40:53 +00:00
else
Change the driver to allocate its own callout structure, and modify the interface to use callout_* instead of timeout(). Also add an IS_MPSAFE #define (currently off) which will mark the driver as mpsafe to the upper layers.
2000-11-25 08:00:10 +00:00
callout_reset(&sc->dc_stat_ch, hz, dc_tick, sc);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
First round of converting network drivers from spls to mutexes. This takes care of all the 10/100 and gigE PCI drivers that I've done. Next will be the wireless drivers, then the USB ones. I may pick up some stragglers along the way. I'm sort of playing this by ear: if anyone spots any places where I've screwed up horribly, please let me know.
2000-10-13 17:54:19 +00:00
DC_UNLOCK(sc);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
return;
}
Apply patch to allow TX underrun handling without issuing a complete chip reset. Just temporarily turn off the transmitter instead. Submitted by: Stephen McKay <mckay@freebsd.org>
2001-02-22 19:26:55 +00:00
/*
* A transmit underrun has occurred. Back off the transmit threshold,
* or switch to store and forward mode if we have to.
*/
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static void
dc_tx_underrun(sc)
Apply patch to allow TX underrun handling without issuing a complete chip reset. Just temporarily turn off the transmitter instead. Submitted by: Stephen McKay <mckay@freebsd.org>
2001-02-22 19:26:55 +00:00
struct dc_softc *sc;
{
u_int32_t isr;
int i;
if (DC_IS_DAVICOM(sc))
dc_init(sc);
if (DC_IS_INTEL(sc)) {
/*
* The real 21143 requires that the transmitter be idle
* in order to change the transmit threshold or store
* and forward state.
*/
DC_CLRBIT(sc, DC_NETCFG, DC_NETCFG_TX_ON);
for (i = 0; i < DC_TIMEOUT; i++) {
isr = CSR_READ_4(sc, DC_ISR);
if (isr & DC_ISR_TX_IDLE)
break;
DELAY(10);
}
if (i == DC_TIMEOUT) {
printf("dc%d: failed to force tx to idle state\n",
sc->dc_unit);
dc_init(sc);
}
}
printf("dc%d: TX underrun -- ", sc->dc_unit);
sc->dc_txthresh += DC_TXTHRESH_INC;
if (sc->dc_txthresh > DC_TXTHRESH_MAX) {
printf("using store and forward mode\n");
DC_SETBIT(sc, DC_NETCFG, DC_NETCFG_STORENFWD);
} else {
printf("increasing TX threshold\n");
DC_CLRBIT(sc, DC_NETCFG, DC_NETCFG_TX_THRESH);
DC_SETBIT(sc, DC_NETCFG, sc->dc_txthresh);
}
if (DC_IS_INTEL(sc))
DC_SETBIT(sc, DC_NETCFG, DC_NETCFG_TX_ON);
return;
}
Device Polling code for -current. Non-SMP, i386-only, no polling in the idle loop at the moment. To use this code you must compile a kernel with options DEVICE_POLLING and at runtime enable polling with sysctl kern.polling.enable=1 The percentage of CPU reserved to userland can be set with sysctl kern.polling.user_frac=NN (default is 50) while the remainder is used by polling device drivers and netisr's. These are the only two variables that you should need to touch. There are a few more parameters in kern.polling but the default values are adequate for all purposes. See the code in kern_poll.c for more details on them. Polling in the idle loop will be implemented shortly by introducing a kernel thread which does the job. Until then, the amount of CPU dedicated to polling will never exceed (100-user_frac). The equivalent (actually, better) code for -stable is at http://info.iet.unipi.it/~luigi/polling/ and also supports polling in the idle loop. NOTE to Alpha developers: There is really nothing in this code that is i386-specific. If you move the 2 lines supporting the new option from sys/conf/{files,options}.i386 to sys/conf/{files,options} I am pretty sure that this should work on the Alpha as well, just that I do not have a suitable test box to try it. If someone feels like trying it, I would appreciate it. NOTE to other developers: sure some things could be done better, and as always I am open to constructive criticism, which a few of you have already given and I greatly appreciated. However, before proposing radical architectural changes, please take some time to possibly try out this code, or at the very least read the comments in kern_poll.c, especially re. the reason why I am using a soft netisr and cannot (I believe) replace it with a simple timeout. Quick description of files touched by this commit: sys/conf/files.i386 new file kern/kern_poll.c sys/conf/options.i386 new option sys/i386/i386/trap.c poll in trap (disabled by default) sys/kern/kern_clock.c initialization and hardclock hooks. sys/kern/kern_intr.c minor swi_net changes sys/kern/kern_poll.c the bulk of the code. sys/net/if.h new flag sys/net/if_var.h declaration for functions used in device drivers. sys/net/netisr.h NETISR_POLL sys/dev/fxp/if_fxp.c sys/dev/fxp/if_fxpvar.h sys/pci/if_dc.c sys/pci/if_dcreg.h sys/pci/if_sis.c sys/pci/if_sisreg.h device driver modifications
2001-12-14 17:56:12 +00:00
#ifdef DEVICE_POLLING
static poll_handler_t dc_poll;
static void
dc_poll(struct ifnet *ifp, enum poll_cmd cmd, int count)
{
Patch up some existing style bugs and some that crept in with the DEVICE_POLLING stuff.
2001-12-15 02:51:21 +00:00
struct dc_softc *sc = ifp->if_softc;
Device Polling code for -current. Non-SMP, i386-only, no polling in the idle loop at the moment. To use this code you must compile a kernel with options DEVICE_POLLING and at runtime enable polling with sysctl kern.polling.enable=1 The percentage of CPU reserved to userland can be set with sysctl kern.polling.user_frac=NN (default is 50) while the remainder is used by polling device drivers and netisr's. These are the only two variables that you should need to touch. There are a few more parameters in kern.polling but the default values are adequate for all purposes. See the code in kern_poll.c for more details on them. Polling in the idle loop will be implemented shortly by introducing a kernel thread which does the job. Until then, the amount of CPU dedicated to polling will never exceed (100-user_frac). The equivalent (actually, better) code for -stable is at http://info.iet.unipi.it/~luigi/polling/ and also supports polling in the idle loop. NOTE to Alpha developers: There is really nothing in this code that is i386-specific. If you move the 2 lines supporting the new option from sys/conf/{files,options}.i386 to sys/conf/{files,options} I am pretty sure that this should work on the Alpha as well, just that I do not have a suitable test box to try it. If someone feels like trying it, I would appreciate it. NOTE to other developers: sure some things could be done better, and as always I am open to constructive criticism, which a few of you have already given and I greatly appreciated. However, before proposing radical architectural changes, please take some time to possibly try out this code, or at the very least read the comments in kern_poll.c, especially re. the reason why I am using a soft netisr and cannot (I believe) replace it with a simple timeout. Quick description of files touched by this commit: sys/conf/files.i386 new file kern/kern_poll.c sys/conf/options.i386 new option sys/i386/i386/trap.c poll in trap (disabled by default) sys/kern/kern_clock.c initialization and hardclock hooks. sys/kern/kern_intr.c minor swi_net changes sys/kern/kern_poll.c the bulk of the code. sys/net/if.h new flag sys/net/if_var.h declaration for functions used in device drivers. sys/net/netisr.h NETISR_POLL sys/dev/fxp/if_fxp.c sys/dev/fxp/if_fxpvar.h sys/pci/if_dc.c sys/pci/if_dcreg.h sys/pci/if_sis.c sys/pci/if_sisreg.h device driver modifications
2001-12-14 17:56:12 +00:00
if (cmd == POLL_DEREGISTER) { /* final call, enable interrupts */
/* Re-enable interrupts. */
CSR_WRITE_4(sc, DC_IMR, DC_INTRS);
return;
}
sc->rxcycles = count;
dc_rxeof(sc);
dc_txeof(sc);
if (ifp->if_snd.ifq_head != NULL && !(ifp->if_flags & IFF_OACTIVE))
dc_start(ifp);
if (cmd == POLL_AND_CHECK_STATUS) { /* also check status register */
u_int32_t status;
status = CSR_READ_4(sc, DC_ISR);
status &= (DC_ISR_RX_WATDOGTIMEO|DC_ISR_RX_NOBUF|
DC_ISR_TX_NOBUF|DC_ISR_TX_IDLE|DC_ISR_TX_UNDERRUN|
DC_ISR_BUS_ERR);
if (!status)
Patch up some existing style bugs and some that crept in with the DEVICE_POLLING stuff.
2001-12-15 02:51:21 +00:00
return;
Device Polling code for -current. Non-SMP, i386-only, no polling in the idle loop at the moment. To use this code you must compile a kernel with options DEVICE_POLLING and at runtime enable polling with sysctl kern.polling.enable=1 The percentage of CPU reserved to userland can be set with sysctl kern.polling.user_frac=NN (default is 50) while the remainder is used by polling device drivers and netisr's. These are the only two variables that you should need to touch. There are a few more parameters in kern.polling but the default values are adequate for all purposes. See the code in kern_poll.c for more details on them. Polling in the idle loop will be implemented shortly by introducing a kernel thread which does the job. Until then, the amount of CPU dedicated to polling will never exceed (100-user_frac). The equivalent (actually, better) code for -stable is at http://info.iet.unipi.it/~luigi/polling/ and also supports polling in the idle loop. NOTE to Alpha developers: There is really nothing in this code that is i386-specific. If you move the 2 lines supporting the new option from sys/conf/{files,options}.i386 to sys/conf/{files,options} I am pretty sure that this should work on the Alpha as well, just that I do not have a suitable test box to try it. If someone feels like trying it, I would appreciate it. NOTE to other developers: sure some things could be done better, and as always I am open to constructive criticism, which a few of you have already given and I greatly appreciated. However, before proposing radical architectural changes, please take some time to possibly try out this code, or at the very least read the comments in kern_poll.c, especially re. the reason why I am using a soft netisr and cannot (I believe) replace it with a simple timeout. Quick description of files touched by this commit: sys/conf/files.i386 new file kern/kern_poll.c sys/conf/options.i386 new option sys/i386/i386/trap.c poll in trap (disabled by default) sys/kern/kern_clock.c initialization and hardclock hooks. sys/kern/kern_intr.c minor swi_net changes sys/kern/kern_poll.c the bulk of the code. sys/net/if.h new flag sys/net/if_var.h declaration for functions used in device drivers. sys/net/netisr.h NETISR_POLL sys/dev/fxp/if_fxp.c sys/dev/fxp/if_fxpvar.h sys/pci/if_dc.c sys/pci/if_dcreg.h sys/pci/if_sis.c sys/pci/if_sisreg.h device driver modifications
2001-12-14 17:56:12 +00:00
/* ack what we have */
CSR_WRITE_4(sc, DC_ISR, status);
Patch up some existing style bugs and some that crept in with the DEVICE_POLLING stuff.
2001-12-15 02:51:21 +00:00
if (status & (DC_ISR_RX_WATDOGTIMEO|DC_ISR_RX_NOBUF)) {
Device Polling code for -current. Non-SMP, i386-only, no polling in the idle loop at the moment. To use this code you must compile a kernel with options DEVICE_POLLING and at runtime enable polling with sysctl kern.polling.enable=1 The percentage of CPU reserved to userland can be set with sysctl kern.polling.user_frac=NN (default is 50) while the remainder is used by polling device drivers and netisr's. These are the only two variables that you should need to touch. There are a few more parameters in kern.polling but the default values are adequate for all purposes. See the code in kern_poll.c for more details on them. Polling in the idle loop will be implemented shortly by introducing a kernel thread which does the job. Until then, the amount of CPU dedicated to polling will never exceed (100-user_frac). The equivalent (actually, better) code for -stable is at http://info.iet.unipi.it/~luigi/polling/ and also supports polling in the idle loop. NOTE to Alpha developers: There is really nothing in this code that is i386-specific. If you move the 2 lines supporting the new option from sys/conf/{files,options}.i386 to sys/conf/{files,options} I am pretty sure that this should work on the Alpha as well, just that I do not have a suitable test box to try it. If someone feels like trying it, I would appreciate it. NOTE to other developers: sure some things could be done better, and as always I am open to constructive criticism, which a few of you have already given and I greatly appreciated. However, before proposing radical architectural changes, please take some time to possibly try out this code, or at the very least read the comments in kern_poll.c, especially re. the reason why I am using a soft netisr and cannot (I believe) replace it with a simple timeout. Quick description of files touched by this commit: sys/conf/files.i386 new file kern/kern_poll.c sys/conf/options.i386 new option sys/i386/i386/trap.c poll in trap (disabled by default) sys/kern/kern_clock.c initialization and hardclock hooks. sys/kern/kern_intr.c minor swi_net changes sys/kern/kern_poll.c the bulk of the code. sys/net/if.h new flag sys/net/if_var.h declaration for functions used in device drivers. sys/net/netisr.h NETISR_POLL sys/dev/fxp/if_fxp.c sys/dev/fxp/if_fxpvar.h sys/pci/if_dc.c sys/pci/if_dcreg.h sys/pci/if_sis.c sys/pci/if_sisreg.h device driver modifications
2001-12-14 17:56:12 +00:00
u_int32_t r = CSR_READ_4(sc, DC_FRAMESDISCARDED);
ifp->if_ierrors += (r & 0xffff) + ((r >> 17) & 0x7ff);
if (dc_rx_resync(sc))
dc_rxeof(sc);
}
/* restart transmit unit if necessary */
if (status & DC_ISR_TX_IDLE && sc->dc_cdata.dc_tx_cnt)
CSR_WRITE_4(sc, DC_TXSTART, 0xFFFFFFFF);
if (status & DC_ISR_TX_UNDERRUN)
dc_tx_underrun(sc);
if (status & DC_ISR_BUS_ERR) {
printf("dc_poll: dc%d bus error\n", sc->dc_unit);
dc_reset(sc);
dc_init(sc);
}
}
}
#endif /* DEVICE_POLLING */
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static void
dc_intr(arg)
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
void *arg;
{
struct dc_softc *sc;
struct ifnet *ifp;
u_int32_t status;
sc = arg;
Add suspend/resume code mostly merged from fxp/rl driver. This is temporary hack, better and generalized solution probably should be implemented at lower layer(MII or PCI?). Tested by: shoko.araki@soliton.co.jp MFC after: 1 week
2002-05-06 10:55:42 +00:00
if (sc->suspended) {
return;
}
Patch up some existing style bugs and some that crept in with the DEVICE_POLLING stuff.
2001-12-15 02:51:21 +00:00
if ((CSR_READ_4(sc, DC_ISR) & DC_INTRS) == 0)
return;
Take luigi's suggestion and move the check for nothing to do to before the lock so we don't have lots of null lock/release pairs.
2001-02-20 04:43:59 +00:00
DC_LOCK(sc);
ifp = &sc->arpcom.ac_if;
Device Polling code for -current. Non-SMP, i386-only, no polling in the idle loop at the moment. To use this code you must compile a kernel with options DEVICE_POLLING and at runtime enable polling with sysctl kern.polling.enable=1 The percentage of CPU reserved to userland can be set with sysctl kern.polling.user_frac=NN (default is 50) while the remainder is used by polling device drivers and netisr's. These are the only two variables that you should need to touch. There are a few more parameters in kern.polling but the default values are adequate for all purposes. See the code in kern_poll.c for more details on them. Polling in the idle loop will be implemented shortly by introducing a kernel thread which does the job. Until then, the amount of CPU dedicated to polling will never exceed (100-user_frac). The equivalent (actually, better) code for -stable is at http://info.iet.unipi.it/~luigi/polling/ and also supports polling in the idle loop. NOTE to Alpha developers: There is really nothing in this code that is i386-specific. If you move the 2 lines supporting the new option from sys/conf/{files,options}.i386 to sys/conf/{files,options} I am pretty sure that this should work on the Alpha as well, just that I do not have a suitable test box to try it. If someone feels like trying it, I would appreciate it. NOTE to other developers: sure some things could be done better, and as always I am open to constructive criticism, which a few of you have already given and I greatly appreciated. However, before proposing radical architectural changes, please take some time to possibly try out this code, or at the very least read the comments in kern_poll.c, especially re. the reason why I am using a soft netisr and cannot (I believe) replace it with a simple timeout. Quick description of files touched by this commit: sys/conf/files.i386 new file kern/kern_poll.c sys/conf/options.i386 new option sys/i386/i386/trap.c poll in trap (disabled by default) sys/kern/kern_clock.c initialization and hardclock hooks. sys/kern/kern_intr.c minor swi_net changes sys/kern/kern_poll.c the bulk of the code. sys/net/if.h new flag sys/net/if_var.h declaration for functions used in device drivers. sys/net/netisr.h NETISR_POLL sys/dev/fxp/if_fxp.c sys/dev/fxp/if_fxpvar.h sys/pci/if_dc.c sys/pci/if_dcreg.h sys/pci/if_sis.c sys/pci/if_sisreg.h device driver modifications
2001-12-14 17:56:12 +00:00
#ifdef DEVICE_POLLING
Increase size of ifnet.if_flags from 16 bits (short) to 32 bits (int). To avoid breaking application ABI use unused ifreq.ifru_flags[1] for upper 16 bits in SIOCSIFFLAGS and SIOCGIFFLAGS ioctl's. Reviewed by: -hackers, -net
2002-08-18 07:05:00 +00:00
if (ifp->if_flags & IFF_POLLING)
Device Polling code for -current. Non-SMP, i386-only, no polling in the idle loop at the moment. To use this code you must compile a kernel with options DEVICE_POLLING and at runtime enable polling with sysctl kern.polling.enable=1 The percentage of CPU reserved to userland can be set with sysctl kern.polling.user_frac=NN (default is 50) while the remainder is used by polling device drivers and netisr's. These are the only two variables that you should need to touch. There are a few more parameters in kern.polling but the default values are adequate for all purposes. See the code in kern_poll.c for more details on them. Polling in the idle loop will be implemented shortly by introducing a kernel thread which does the job. Until then, the amount of CPU dedicated to polling will never exceed (100-user_frac). The equivalent (actually, better) code for -stable is at http://info.iet.unipi.it/~luigi/polling/ and also supports polling in the idle loop. NOTE to Alpha developers: There is really nothing in this code that is i386-specific. If you move the 2 lines supporting the new option from sys/conf/{files,options}.i386 to sys/conf/{files,options} I am pretty sure that this should work on the Alpha as well, just that I do not have a suitable test box to try it. If someone feels like trying it, I would appreciate it. NOTE to other developers: sure some things could be done better, and as always I am open to constructive criticism, which a few of you have already given and I greatly appreciated. However, before proposing radical architectural changes, please take some time to possibly try out this code, or at the very least read the comments in kern_poll.c, especially re. the reason why I am using a soft netisr and cannot (I believe) replace it with a simple timeout. Quick description of files touched by this commit: sys/conf/files.i386 new file kern/kern_poll.c sys/conf/options.i386 new option sys/i386/i386/trap.c poll in trap (disabled by default) sys/kern/kern_clock.c initialization and hardclock hooks. sys/kern/kern_intr.c minor swi_net changes sys/kern/kern_poll.c the bulk of the code. sys/net/if.h new flag sys/net/if_var.h declaration for functions used in device drivers. sys/net/netisr.h NETISR_POLL sys/dev/fxp/if_fxp.c sys/dev/fxp/if_fxpvar.h sys/pci/if_dc.c sys/pci/if_dcreg.h sys/pci/if_sis.c sys/pci/if_sisreg.h device driver modifications
2001-12-14 17:56:12 +00:00
goto done;
if (ether_poll_register(dc_poll, ifp)) { /* ok, disable interrupts */
CSR_WRITE_4(sc, DC_IMR, 0x00000000);
goto done;
}
#endif /* DEVICE_POLLING */
Add a check in the interrupt service routine to return quickly in case there is nothing to do. This happens normally when the card shares the interrupt line with other devices. This code saves a couple of microseconds per interrupt even on a fast CPU. You normally would not care, except under heavy tinygram traffic where you can have some 50-100.000 interrupts per second... On passing, correct a spelling error.
2001-02-18 07:21:28 +00:00
/* Suppress unwanted interrupts */
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
if (!(ifp->if_flags & IFF_UP)) {
if (CSR_READ_4(sc, DC_ISR) & DC_INTRS)
dc_stop(sc);
First round of converting network drivers from spls to mutexes. This takes care of all the 10/100 and gigE PCI drivers that I've done. Next will be the wireless drivers, then the USB ones. I may pick up some stragglers along the way. I'm sort of playing this by ear: if anyone spots any places where I've screwed up horribly, please let me know.
2000-10-13 17:54:19 +00:00
DC_UNLOCK(sc);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
return;
}
/* Disable interrupts. */
CSR_WRITE_4(sc, DC_IMR, 0x00000000);
NEWCARD/Cardbus - This commit adds support for Xircom X3201 based cardbus cards. Support for the TDK 78Q2120 MII is also added. IBM Etherjet, Intel and Xircom cards uses these chips. Note that as a result of this commit, some Intel/DEC 21143 based cardbus cards will also attach, but not get link. That is being looked at.
2000-10-19 08:34:32 +00:00
while(((status = CSR_READ_4(sc, DC_ISR)) & DC_INTRS)
&& status != 0xFFFFFFFF) {
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
CSR_WRITE_4(sc, DC_ISR, status);
It appears that under certain circumstances that I still can't quite pin down, the dc driver and receiver can fall out of sync with one another, resulting in a condition where the chip continues to receive packets but the driver never notices. Normally, the receive handler checks each descriptor starting from the current producer index to see if the chip has relinquished ownership, indicating that a packet has been received. The driver hands the packet off to ether_input() and then prepares the descriptor to receive another frame before moving on to the next descriptor in the ring. But sometimes, the chip appears to skip a descriptor. This leaves the driver testing the status word in a descriptor that never gets updated. The driver still gets "RX done" interrupts but never advances further into the RX ring, until the ring fills up and the chip interrupts again to signal an error condition. Sometimes, the driver will remain in this desynchronized state, resulting in spotty performance until the interface is reset. Fortunately, it's fairly simple to detect this condition: if we call the rxeof routine but the number of received packets doesn't increase, we suspect that there could be a problem. In this case, we call a new routine called dc_rx_resync(), which scans ahead in the RX ring to see if there's a frame waiting for us somewhere beyond that the driver thinks is the current producer index. If it finds one, it bumps up the index and calls the rxeof handler again to snarf up the packet and bring the driver back in sync with the chip. (It may actually do this several times in the event that there's more than one "hole" in the ring.) So far the only card supported by if_dc which has exhibited this problem is a LinkSys LNE100TX v2.0 (82c115 PNIC II), and it only seems to happen on one particular system, however the fix is general enough and has low enough overhead that we may as well apply it for all supported chipsets. I also implemented the same fix for the 3Com xl driver, which is apparently vulnerable to the same problem. Problem originally noted and patch tested by: Matt Dillon
2000-01-03 15:28:47 +00:00
if (status & DC_ISR_RX_OK) {
int curpkts;
curpkts = ifp->if_ipackets;
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
dc_rxeof(sc);
It appears that under certain circumstances that I still can't quite pin down, the dc driver and receiver can fall out of sync with one another, resulting in a condition where the chip continues to receive packets but the driver never notices. Normally, the receive handler checks each descriptor starting from the current producer index to see if the chip has relinquished ownership, indicating that a packet has been received. The driver hands the packet off to ether_input() and then prepares the descriptor to receive another frame before moving on to the next descriptor in the ring. But sometimes, the chip appears to skip a descriptor. This leaves the driver testing the status word in a descriptor that never gets updated. The driver still gets "RX done" interrupts but never advances further into the RX ring, until the ring fills up and the chip interrupts again to signal an error condition. Sometimes, the driver will remain in this desynchronized state, resulting in spotty performance until the interface is reset. Fortunately, it's fairly simple to detect this condition: if we call the rxeof routine but the number of received packets doesn't increase, we suspect that there could be a problem. In this case, we call a new routine called dc_rx_resync(), which scans ahead in the RX ring to see if there's a frame waiting for us somewhere beyond that the driver thinks is the current producer index. If it finds one, it bumps up the index and calls the rxeof handler again to snarf up the packet and bring the driver back in sync with the chip. (It may actually do this several times in the event that there's more than one "hole" in the ring.) So far the only card supported by if_dc which has exhibited this problem is a LinkSys LNE100TX v2.0 (82c115 PNIC II), and it only seems to happen on one particular system, however the fix is general enough and has low enough overhead that we may as well apply it for all supported chipsets. I also implemented the same fix for the 3Com xl driver, which is apparently vulnerable to the same problem. Problem originally noted and patch tested by: Matt Dillon
2000-01-03 15:28:47 +00:00
if (curpkts == ifp->if_ipackets) {
while(dc_rx_resync(sc))
dc_rxeof(sc);
}
}
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
if (status & (DC_ISR_TX_OK|DC_ISR_TX_NOBUF))
dc_txeof(sc);
if (status & DC_ISR_TX_IDLE) {
dc_txeof(sc);
if (sc->dc_cdata.dc_tx_cnt) {
DC_SETBIT(sc, DC_NETCFG, DC_NETCFG_TX_ON);
CSR_WRITE_4(sc, DC_TXSTART, 0xFFFFFFFF);
}
}
Apply patch to allow TX underrun handling without issuing a complete chip reset. Just temporarily turn off the transmitter instead. Submitted by: Stephen McKay <mckay@freebsd.org>
2001-02-22 19:26:55 +00:00
if (status & DC_ISR_TX_UNDERRUN)
dc_tx_underrun(sc);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
if ((status & DC_ISR_RX_WATDOGTIMEO)
It appears that under certain circumstances that I still can't quite pin down, the dc driver and receiver can fall out of sync with one another, resulting in a condition where the chip continues to receive packets but the driver never notices. Normally, the receive handler checks each descriptor starting from the current producer index to see if the chip has relinquished ownership, indicating that a packet has been received. The driver hands the packet off to ether_input() and then prepares the descriptor to receive another frame before moving on to the next descriptor in the ring. But sometimes, the chip appears to skip a descriptor. This leaves the driver testing the status word in a descriptor that never gets updated. The driver still gets "RX done" interrupts but never advances further into the RX ring, until the ring fills up and the chip interrupts again to signal an error condition. Sometimes, the driver will remain in this desynchronized state, resulting in spotty performance until the interface is reset. Fortunately, it's fairly simple to detect this condition: if we call the rxeof routine but the number of received packets doesn't increase, we suspect that there could be a problem. In this case, we call a new routine called dc_rx_resync(), which scans ahead in the RX ring to see if there's a frame waiting for us somewhere beyond that the driver thinks is the current producer index. If it finds one, it bumps up the index and calls the rxeof handler again to snarf up the packet and bring the driver back in sync with the chip. (It may actually do this several times in the event that there's more than one "hole" in the ring.) So far the only card supported by if_dc which has exhibited this problem is a LinkSys LNE100TX v2.0 (82c115 PNIC II), and it only seems to happen on one particular system, however the fix is general enough and has low enough overhead that we may as well apply it for all supported chipsets. I also implemented the same fix for the 3Com xl driver, which is apparently vulnerable to the same problem. Problem originally noted and patch tested by: Matt Dillon
2000-01-03 15:28:47 +00:00
|| (status & DC_ISR_RX_NOBUF)) {
int curpkts;
curpkts = ifp->if_ipackets;
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
dc_rxeof(sc);
It appears that under certain circumstances that I still can't quite pin down, the dc driver and receiver can fall out of sync with one another, resulting in a condition where the chip continues to receive packets but the driver never notices. Normally, the receive handler checks each descriptor starting from the current producer index to see if the chip has relinquished ownership, indicating that a packet has been received. The driver hands the packet off to ether_input() and then prepares the descriptor to receive another frame before moving on to the next descriptor in the ring. But sometimes, the chip appears to skip a descriptor. This leaves the driver testing the status word in a descriptor that never gets updated. The driver still gets "RX done" interrupts but never advances further into the RX ring, until the ring fills up and the chip interrupts again to signal an error condition. Sometimes, the driver will remain in this desynchronized state, resulting in spotty performance until the interface is reset. Fortunately, it's fairly simple to detect this condition: if we call the rxeof routine but the number of received packets doesn't increase, we suspect that there could be a problem. In this case, we call a new routine called dc_rx_resync(), which scans ahead in the RX ring to see if there's a frame waiting for us somewhere beyond that the driver thinks is the current producer index. If it finds one, it bumps up the index and calls the rxeof handler again to snarf up the packet and bring the driver back in sync with the chip. (It may actually do this several times in the event that there's more than one "hole" in the ring.) So far the only card supported by if_dc which has exhibited this problem is a LinkSys LNE100TX v2.0 (82c115 PNIC II), and it only seems to happen on one particular system, however the fix is general enough and has low enough overhead that we may as well apply it for all supported chipsets. I also implemented the same fix for the 3Com xl driver, which is apparently vulnerable to the same problem. Problem originally noted and patch tested by: Matt Dillon
2000-01-03 15:28:47 +00:00
if (curpkts == ifp->if_ipackets) {
while(dc_rx_resync(sc))
dc_rxeof(sc);
}
}
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
if (status & DC_ISR_BUS_ERR) {
dc_reset(sc);
dc_init(sc);
}
}
/* Re-enable interrupts. */
CSR_WRITE_4(sc, DC_IMR, DC_INTRS);
if (ifp->if_snd.ifq_head != NULL)
dc_start(ifp);
Fix compiler warning in dc_intr(): if the only code that does a "goto" to a label is inside an #ifdef block, then the label should *also* be inside an #ifdef block. Hide the "done:" label which is only used if DEVICE_POLLING is enabled under #ifdef DEVICE_POLLING.
2001-12-19 18:13:44 +00:00
#ifdef DEVICE_POLLING
Device Polling code for -current. Non-SMP, i386-only, no polling in the idle loop at the moment. To use this code you must compile a kernel with options DEVICE_POLLING and at runtime enable polling with sysctl kern.polling.enable=1 The percentage of CPU reserved to userland can be set with sysctl kern.polling.user_frac=NN (default is 50) while the remainder is used by polling device drivers and netisr's. These are the only two variables that you should need to touch. There are a few more parameters in kern.polling but the default values are adequate for all purposes. See the code in kern_poll.c for more details on them. Polling in the idle loop will be implemented shortly by introducing a kernel thread which does the job. Until then, the amount of CPU dedicated to polling will never exceed (100-user_frac). The equivalent (actually, better) code for -stable is at http://info.iet.unipi.it/~luigi/polling/ and also supports polling in the idle loop. NOTE to Alpha developers: There is really nothing in this code that is i386-specific. If you move the 2 lines supporting the new option from sys/conf/{files,options}.i386 to sys/conf/{files,options} I am pretty sure that this should work on the Alpha as well, just that I do not have a suitable test box to try it. If someone feels like trying it, I would appreciate it. NOTE to other developers: sure some things could be done better, and as always I am open to constructive criticism, which a few of you have already given and I greatly appreciated. However, before proposing radical architectural changes, please take some time to possibly try out this code, or at the very least read the comments in kern_poll.c, especially re. the reason why I am using a soft netisr and cannot (I believe) replace it with a simple timeout. Quick description of files touched by this commit: sys/conf/files.i386 new file kern/kern_poll.c sys/conf/options.i386 new option sys/i386/i386/trap.c poll in trap (disabled by default) sys/kern/kern_clock.c initialization and hardclock hooks. sys/kern/kern_intr.c minor swi_net changes sys/kern/kern_poll.c the bulk of the code. sys/net/if.h new flag sys/net/if_var.h declaration for functions used in device drivers. sys/net/netisr.h NETISR_POLL sys/dev/fxp/if_fxp.c sys/dev/fxp/if_fxpvar.h sys/pci/if_dc.c sys/pci/if_dcreg.h sys/pci/if_sis.c sys/pci/if_sisreg.h device driver modifications
2001-12-14 17:56:12 +00:00
done:
Fix compiler warning in dc_intr(): if the only code that does a "goto" to a label is inside an #ifdef block, then the label should *also* be inside an #ifdef block. Hide the "done:" label which is only used if DEVICE_POLLING is enabled under #ifdef DEVICE_POLLING.
2001-12-19 18:13:44 +00:00
#endif /* DEVICE_POLLING */
First round of converting network drivers from spls to mutexes. This takes care of all the 10/100 and gigE PCI drivers that I've done. Next will be the wireless drivers, then the USB ones. I may pick up some stragglers along the way. I'm sort of playing this by ear: if anyone spots any places where I've screwed up horribly, please let me know.
2000-10-13 17:54:19 +00:00
DC_UNLOCK(sc);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
return;
}
/*
* Encapsulate an mbuf chain in a descriptor by coupling the mbuf data
* pointers to the fragment pointers.
*/
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static int
dc_encap(sc, m_head, txidx)
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
struct dc_softc *sc;
struct mbuf *m_head;
u_int32_t *txidx;
{
struct dc_desc *f = NULL;
struct mbuf *m;
int frag, cur, cnt = 0;
/*
Make consistent; turn spaces into tabs where there is a mixture.
2002-09-04 18:14:17 +00:00
* Start packing the mbufs in this chain into
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
* the fragment pointers. Stop when we run out
Make consistent; turn spaces into tabs where there is a mixture.
2002-09-04 18:14:17 +00:00
* of fragments or hit the end of the mbuf chain.
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
*/
m = m_head;
cur = frag = *txidx;
for (m = m_head; m != NULL; m = m->m_next) {
if (m->m_len != 0) {
if (sc->dc_flags & DC_TX_ADMTEK_WAR) {
if (*txidx != sc->dc_cdata.dc_tx_prod &&
frag == (DC_TX_LIST_CNT - 1))
return(ENOBUFS);
}
if ((DC_TX_LIST_CNT -
(sc->dc_cdata.dc_tx_cnt + cnt)) < 5)
return(ENOBUFS);
f = &sc->dc_ldata->dc_tx_list[frag];
f->dc_ctl = DC_TXCTL_TLINK | m->m_len;
if (cnt == 0) {
f->dc_status = 0;
f->dc_ctl |= DC_TXCTL_FIRSTFRAG;
} else
f->dc_status = DC_TXSTAT_OWN;
f->dc_data = vtophys(mtod(m, vm_offset_t));
cur = frag;
DC_INC(frag, DC_TX_LIST_CNT);
cnt++;
}
}
if (m != NULL)
return(ENOBUFS);
sc->dc_cdata.dc_tx_cnt += cnt;
sc->dc_cdata.dc_tx_chain[cur] = m_head;
sc->dc_ldata->dc_tx_list[cur].dc_ctl |= DC_TXCTL_LASTFRAG;
if (sc->dc_flags & DC_TX_INTR_FIRSTFRAG)
sc->dc_ldata->dc_tx_list[*txidx].dc_ctl |= DC_TXCTL_FINT;
Fix some problems reported by Mike Pritchard: - Add a flag DC_TX_INTR_ALWAYS which causes the transmit code to request a TX done interrupt for every packet. The PNIC seems to need this to insure that the sent TX buffers get reaped in a timely fashion. - Try to unreset the SIA as soon as possible after resetting the whole chip. - Change dcphy to support either 10/100 or 10Mbps only NICs. The built-in 21143 ethernet in Compaq Presario machines is 10Mbps only and it doesn't work right if we try to advertise 100Mbps modes during autoneg. When restricted to only 10mbps modes, it works fine. Note that for now, I detect this condition by checking the PCI subsystem ID on this NIC (which has a Compaq vendor/device ID). Yes, I know that's what the SROM is supposed to be for. I'm deliberately ignoring the SROM wherever possible. Sue me. The latter two fixes allow if_dc to work correctly with the built-in ethernet on certain Compaq Presario boxes. There are liable to be quite a few people using these as their home systems who might want to try FreeBSD; may as well be nice to them. Now if anybody out there has an Alpha miata with 10Mbps ethernet and can show me the output from pciconf -l on their system, I'd be grateful.
1999-12-13 21:45:13 +00:00
if (sc->dc_flags & DC_TX_INTR_ALWAYS)
sc->dc_ldata->dc_tx_list[cur].dc_ctl |= DC_TXCTL_FINT;
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
if (sc->dc_flags & DC_TX_USE_TX_INTR && sc->dc_cdata.dc_tx_cnt > 64)
sc->dc_ldata->dc_tx_list[cur].dc_ctl |= DC_TXCTL_FINT;
sc->dc_ldata->dc_tx_list[*txidx].dc_status = DC_TXSTAT_OWN;
*txidx = frag;
return(0);
}
Reintroduce the dc_coal() workaround routine for coalescing outbound packets into a single buffer, and set the DC_TX_COALESCE flag for the Davicom DM9102 chip. I thought I had escaped this problem, but... This chip appears to silently corrupt or discard transmitted frames when using scatter/gather DMA (i.e. DMAing each packet fragment in place with a separate descriptor). The only way to insure reliable transmission is to coalesce transmitted packets into a single cluster buffer. (There may also be an alignment constraint here, but mbuf cluster buffers are naturally aligned on 2K boundaries, which seems to be good enough.) The DM9102 driver for Linux written by Davicom also uses this workaround. Unfortunately, the Davicom datasheet has no errata section describing this or any other apparently known defect. Problem noted by: allan_chou@davicom.com.tw
2000-01-12 22:24:05 +00:00
/*
* Coalesce an mbuf chain into a single mbuf cluster buffer.
* Needed for some really badly behaved chips that just can't
* do scatter/gather correctly.
*/
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static int
dc_coal(sc, m_head)
Reintroduce the dc_coal() workaround routine for coalescing outbound packets into a single buffer, and set the DC_TX_COALESCE flag for the Davicom DM9102 chip. I thought I had escaped this problem, but... This chip appears to silently corrupt or discard transmitted frames when using scatter/gather DMA (i.e. DMAing each packet fragment in place with a separate descriptor). The only way to insure reliable transmission is to coalesce transmitted packets into a single cluster buffer. (There may also be an alignment constraint here, but mbuf cluster buffers are naturally aligned on 2K boundaries, which seems to be good enough.) The DM9102 driver for Linux written by Davicom also uses this workaround. Unfortunately, the Davicom datasheet has no errata section describing this or any other apparently known defect. Problem noted by: allan_chou@davicom.com.tw
2000-01-12 22:24:05 +00:00
struct dc_softc *sc;
struct mbuf **m_head;
{
Make consistent; turn spaces into tabs where there is a mixture.
2002-09-04 18:14:17 +00:00
struct mbuf *m_new, *m;
Reintroduce the dc_coal() workaround routine for coalescing outbound packets into a single buffer, and set the DC_TX_COALESCE flag for the Davicom DM9102 chip. I thought I had escaped this problem, but... This chip appears to silently corrupt or discard transmitted frames when using scatter/gather DMA (i.e. DMAing each packet fragment in place with a separate descriptor). The only way to insure reliable transmission is to coalesce transmitted packets into a single cluster buffer. (There may also be an alignment constraint here, but mbuf cluster buffers are naturally aligned on 2K boundaries, which seems to be good enough.) The DM9102 driver for Linux written by Davicom also uses this workaround. Unfortunately, the Davicom datasheet has no errata section describing this or any other apparently known defect. Problem noted by: allan_chou@davicom.com.tw
2000-01-12 22:24:05 +00:00
m = *m_head;
Remove M_TRYWAIT/M_WAITOK/M_WAIT. Callers should use 0. Merge M_NOWAIT/M_DONTWAIT into a single flag M_NOWAIT.
2003-01-21 08:56:16 +00:00
MGETHDR(m_new, M_NOWAIT, MT_DATA);
Remove error messages on mbuf allocation failures, now this is done more safely in kern/subr_mbuf.c Two-days'-delay-thanks-to: @home shutting down service
2001-12-04 02:30:53 +00:00
if (m_new == NULL)
Reintroduce the dc_coal() workaround routine for coalescing outbound packets into a single buffer, and set the DC_TX_COALESCE flag for the Davicom DM9102 chip. I thought I had escaped this problem, but... This chip appears to silently corrupt or discard transmitted frames when using scatter/gather DMA (i.e. DMAing each packet fragment in place with a separate descriptor). The only way to insure reliable transmission is to coalesce transmitted packets into a single cluster buffer. (There may also be an alignment constraint here, but mbuf cluster buffers are naturally aligned on 2K boundaries, which seems to be good enough.) The DM9102 driver for Linux written by Davicom also uses this workaround. Unfortunately, the Davicom datasheet has no errata section describing this or any other apparently known defect. Problem noted by: allan_chou@davicom.com.tw
2000-01-12 22:24:05 +00:00
return(ENOBUFS);
if (m->m_pkthdr.len > MHLEN) {
Remove M_TRYWAIT/M_WAITOK/M_WAIT. Callers should use 0. Merge M_NOWAIT/M_DONTWAIT into a single flag M_NOWAIT.
2003-01-21 08:56:16 +00:00
MCLGET(m_new, M_NOWAIT);
Reintroduce the dc_coal() workaround routine for coalescing outbound packets into a single buffer, and set the DC_TX_COALESCE flag for the Davicom DM9102 chip. I thought I had escaped this problem, but... This chip appears to silently corrupt or discard transmitted frames when using scatter/gather DMA (i.e. DMAing each packet fragment in place with a separate descriptor). The only way to insure reliable transmission is to coalesce transmitted packets into a single cluster buffer. (There may also be an alignment constraint here, but mbuf cluster buffers are naturally aligned on 2K boundaries, which seems to be good enough.) The DM9102 driver for Linux written by Davicom also uses this workaround. Unfortunately, the Davicom datasheet has no errata section describing this or any other apparently known defect. Problem noted by: allan_chou@davicom.com.tw
2000-01-12 22:24:05 +00:00
if (!(m_new->m_flags & M_EXT)) {
m_freem(m_new);
return(ENOBUFS);
}
}
m_copydata(m, 0, m->m_pkthdr.len, mtod(m_new, caddr_t));
m_new->m_pkthdr.len = m_new->m_len = m->m_pkthdr.len;
m_freem(m);
*m_head = m_new;
return(0);
}
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
/*
* Main transmit routine. To avoid having to do mbuf copies, we put pointers
* to the mbuf data regions directly in the transmit lists. We also save a
* copy of the pointers since the transmit list fragment pointers are
* physical addresses.
*/
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static void
dc_start(ifp)
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
struct ifnet *ifp;
{
struct dc_softc *sc;
struct mbuf *m_head = NULL;
int idx;
sc = ifp->if_softc;
First round of converting network drivers from spls to mutexes. This takes care of all the 10/100 and gigE PCI drivers that I've done. Next will be the wireless drivers, then the USB ones. I may pick up some stragglers along the way. I'm sort of playing this by ear: if anyone spots any places where I've screwed up horribly, please let me know.
2000-10-13 17:54:19 +00:00
DC_LOCK(sc);
Deal with the condition where we lose link in the middle of transmitting a bunch of frames. In this case, the dc_link flag is cleared, and dc_start() stops draining the if_snd send queue, which results in lots of 'no buffers available' errors being reported to applications. The whole idea behind not draining the send queue until the link comes up was to avoid having the gratuitous ARP being lost while we're waiting for autoneg to complete after the interface is first brought up. As an optimization, change the test in dc_start() so that we only bail if dc_link is not set _and_ there are less than 10 packets in the send queue. If the queue has many frames in it, we need to drain them. If the queue has a small number of frames in it, we can hold off on sending them until the link comes up. MFC after: 1 week
2001-07-12 22:51:59 +00:00
if (!sc->dc_link && ifp->if_snd.ifq_len < 10) {
First round of converting network drivers from spls to mutexes. This takes care of all the 10/100 and gigE PCI drivers that I've done. Next will be the wireless drivers, then the USB ones. I may pick up some stragglers along the way. I'm sort of playing this by ear: if anyone spots any places where I've screwed up horribly, please let me know.
2000-10-13 17:54:19 +00:00
DC_UNLOCK(sc);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
return;
First round of converting network drivers from spls to mutexes. This takes care of all the 10/100 and gigE PCI drivers that I've done. Next will be the wireless drivers, then the USB ones. I may pick up some stragglers along the way. I'm sort of playing this by ear: if anyone spots any places where I've screwed up horribly, please let me know.
2000-10-13 17:54:19 +00:00
}
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
First round of converting network drivers from spls to mutexes. This takes care of all the 10/100 and gigE PCI drivers that I've done. Next will be the wireless drivers, then the USB ones. I may pick up some stragglers along the way. I'm sort of playing this by ear: if anyone spots any places where I've screwed up horribly, please let me know.
2000-10-13 17:54:19 +00:00
if (ifp->if_flags & IFF_OACTIVE) {
DC_UNLOCK(sc);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
return;
First round of converting network drivers from spls to mutexes. This takes care of all the 10/100 and gigE PCI drivers that I've done. Next will be the wireless drivers, then the USB ones. I may pick up some stragglers along the way. I'm sort of playing this by ear: if anyone spots any places where I've screwed up horribly, please let me know.
2000-10-13 17:54:19 +00:00
}
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
idx = sc->dc_cdata.dc_tx_prod;
while(sc->dc_cdata.dc_tx_chain[idx] == NULL) {
IF_DEQUEUE(&ifp->if_snd, m_head);
if (m_head == NULL)
break;
Avoid an unnecessary copy of a packet if it is already in a single mbuf. Introduce an additional device flag for those NICs which require the transmit buffers to be aligned to 32-bit boundaries. (the equivalen fix for STABLE is slightly simpler because there are no supported chips which require this alignment there.)
2001-12-11 02:47:53 +00:00
if (sc->dc_flags & DC_TX_COALESCE &&
(m_head->m_next != NULL ||
Patch up some existing style bugs and some that crept in with the DEVICE_POLLING stuff.
2001-12-15 02:51:21 +00:00
sc->dc_flags & DC_TX_ALIGN)) {
Reintroduce the dc_coal() workaround routine for coalescing outbound packets into a single buffer, and set the DC_TX_COALESCE flag for the Davicom DM9102 chip. I thought I had escaped this problem, but... This chip appears to silently corrupt or discard transmitted frames when using scatter/gather DMA (i.e. DMAing each packet fragment in place with a separate descriptor). The only way to insure reliable transmission is to coalesce transmitted packets into a single cluster buffer. (There may also be an alignment constraint here, but mbuf cluster buffers are naturally aligned on 2K boundaries, which seems to be good enough.) The DM9102 driver for Linux written by Davicom also uses this workaround. Unfortunately, the Davicom datasheet has no errata section describing this or any other apparently known defect. Problem noted by: allan_chou@davicom.com.tw
2000-01-12 22:24:05 +00:00
if (dc_coal(sc, &m_head)) {
IF_PREPEND(&ifp->if_snd, m_head);
ifp->if_flags |= IFF_OACTIVE;
break;
}
}
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
if (dc_encap(sc, m_head, &idx)) {
IF_PREPEND(&ifp->if_snd, m_head);
ifp->if_flags |= IFF_OACTIVE;
break;
}
/*
* If there's a BPF listener, bounce a copy of this frame
* to him.
*/
o track either_ifattach/ether_ifdetach API changes o use if_input for input packet processing o don't strip the Ethernet header for input packets o use BPF_* macros bpf tapping o call ether_ioctl to handle default ioctl case o track vlan changes Reviewed by: many Approved by: re
2002-11-14 23:49:09 +00:00
BPF_MTAP(ifp, m_head);
Add support for parsing the media blocks from the SROM on 21143 adapters. This is necessary in order to make this driver work with the built-in ethernet on the alpha Miata machines. These systems have a 21143-PC chip on-board and optional daughtercards with either a 10/100 MII transceiver or a 10baseT/10base2 transceiver. In both cases, you need to twiddle the GPIO bits on the controller in order to turn the transceivers on, and you have to read the media info from the SROM in order to find out what bits to twiddle.
2000-10-05 17:36:14 +00:00
if (sc->dc_flags & DC_TX_ONE) {
ifp->if_flags |= IFF_OACTIVE;
break;
}
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
}
/* Transmit */
sc->dc_cdata.dc_tx_prod = idx;
if (!(sc->dc_flags & DC_TX_POLL))
CSR_WRITE_4(sc, DC_TXSTART, 0xFFFFFFFF);
/*
* Set a timeout in case the chip goes out to lunch.
*/
ifp->if_timer = 5;
First round of converting network drivers from spls to mutexes. This takes care of all the 10/100 and gigE PCI drivers that I've done. Next will be the wireless drivers, then the USB ones. I may pick up some stragglers along the way. I'm sort of playing this by ear: if anyone spots any places where I've screwed up horribly, please let me know.
2000-10-13 17:54:19 +00:00
DC_UNLOCK(sc);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
return;
}
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static void
dc_init(xsc)
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
void *xsc;
{
struct dc_softc *sc = xsc;
struct ifnet *ifp = &sc->arpcom.ac_if;
struct mii_data *mii;
First round of converting network drivers from spls to mutexes. This takes care of all the 10/100 and gigE PCI drivers that I've done. Next will be the wireless drivers, then the USB ones. I may pick up some stragglers along the way. I'm sort of playing this by ear: if anyone spots any places where I've screwed up horribly, please let me know.
2000-10-13 17:54:19 +00:00
DC_LOCK(sc);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
mii = device_get_softc(sc->dc_miibus);
/*
* Cancel pending I/O and free all RX/TX buffers.
*/
dc_stop(sc);
dc_reset(sc);
/*
* Set cache alignment and burst length.
*/
Add support for the Davicom DM9102A 10/100 ethernet controller chip. This is just to make sure we initialize the chip correctly: we need to make the sure the port select bit in CSR6 is set properly so that we use the internal PHY for 10/100 support. (The eval boards I have also include an external HomePNA PHY, but I need to play with that more before I can support it.)
2000-01-19 19:03:08 +00:00
if (DC_IS_ASIX(sc) || DC_IS_DAVICOM(sc))
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
CSR_WRITE_4(sc, DC_BUSCTL, 0);
else
CSR_WRITE_4(sc, DC_BUSCTL, DC_BUSCTL_MRME|DC_BUSCTL_MRLE);
Enable round-robin arbitration between transmit and receive unit in the 21143, instead of giving priority to the receive unit. This gives a 10-15% performance improvement in the forwarding rate under heavy load. Reviewed-by: Bill Paul
2001-10-27 00:59:17 +00:00
/*
* Evenly share the bus between receive and transmit process.
*/
if (DC_IS_INTEL(sc))
DC_SETBIT(sc, DC_BUSCTL, DC_BUSCTL_ARBITRATION);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
if (DC_IS_DAVICOM(sc) || DC_IS_INTEL(sc)) {
DC_SETBIT(sc, DC_BUSCTL, DC_BURSTLEN_USECA);
} else {
DC_SETBIT(sc, DC_BUSCTL, DC_BURSTLEN_16LONG);
}
if (sc->dc_flags & DC_TX_POLL)
DC_SETBIT(sc, DC_BUSCTL, DC_TXPOLL_1);
switch(sc->dc_cachesize) {
case 32:
DC_SETBIT(sc, DC_BUSCTL, DC_CACHEALIGN_32LONG);
break;
case 16:
DC_SETBIT(sc, DC_BUSCTL, DC_CACHEALIGN_16LONG);
break;
case 8:
DC_SETBIT(sc, DC_BUSCTL, DC_CACHEALIGN_8LONG);
break;
case 0:
default:
DC_SETBIT(sc, DC_BUSCTL, DC_CACHEALIGN_NONE);
break;
}
if (sc->dc_flags & DC_TX_STORENFWD)
DC_SETBIT(sc, DC_NETCFG, DC_NETCFG_STORENFWD);
else {
Apply patch to allow TX underrun handling without issuing a complete chip reset. Just temporarily turn off the transmitter instead. Submitted by: Stephen McKay <mckay@freebsd.org>
2001-02-22 19:26:55 +00:00
if (sc->dc_txthresh > DC_TXTHRESH_MAX) {
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
DC_SETBIT(sc, DC_NETCFG, DC_NETCFG_STORENFWD);
} else {
DC_CLRBIT(sc, DC_NETCFG, DC_NETCFG_STORENFWD);
DC_SETBIT(sc, DC_NETCFG, sc->dc_txthresh);
}
}
DC_SETBIT(sc, DC_NETCFG, DC_NETCFG_NO_RXCRC);
DC_CLRBIT(sc, DC_NETCFG, DC_NETCFG_TX_BACKOFF);
if (DC_IS_MACRONIX(sc) || DC_IS_PNICII(sc)) {
/*
* The app notes for the 98713 and 98715A say that
* in order to have the chips operate properly, a magic
* number must be written to CSR16. Macronix does not
* document the meaning of these bits so there's no way
* to know exactly what they do. The 98713 has a magic
* number all its own; the rest all use a different one.
*/
DC_CLRBIT(sc, DC_MX_MAGICPACKET, 0xFFFF0000);
if (sc->dc_type == DC_TYPE_98713)
DC_SETBIT(sc, DC_MX_MAGICPACKET, DC_MX_MAGIC_98713);
else
DC_SETBIT(sc, DC_MX_MAGICPACKET, DC_MX_MAGIC_98715);
}
NEWCARD/Cardbus - This commit adds support for Xircom X3201 based cardbus cards. Support for the TDK 78Q2120 MII is also added. IBM Etherjet, Intel and Xircom cards uses these chips. Note that as a result of this commit, some Intel/DEC 21143 based cardbus cards will also attach, but not get link. That is being looked at.
2000-10-19 08:34:32 +00:00
if (DC_IS_XIRCOM(sc)) {
/*
* setup General Purpose Port mode and data so the tulip
* can talk to the MII.
*/
CSR_WRITE_4(sc, DC_SIAGP, DC_SIAGP_WRITE_EN | DC_SIAGP_INT1_EN |
DC_SIAGP_MD_GP2_OUTPUT | DC_SIAGP_MD_GP0_OUTPUT);
DELAY(10);
CSR_WRITE_4(sc, DC_SIAGP, DC_SIAGP_INT1_EN |
DC_SIAGP_MD_GP2_OUTPUT | DC_SIAGP_MD_GP0_OUTPUT);
DELAY(10);
}
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
DC_CLRBIT(sc, DC_NETCFG, DC_NETCFG_TX_THRESH);
Apply patch to allow TX underrun handling without issuing a complete chip reset. Just temporarily turn off the transmitter instead. Submitted by: Stephen McKay <mckay@freebsd.org>
2001-02-22 19:26:55 +00:00
DC_SETBIT(sc, DC_NETCFG, DC_TXTHRESH_MIN);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
/* Init circular RX list. */
if (dc_list_rx_init(sc) == ENOBUFS) {
printf("dc%d: initialization failed: no "
"memory for rx buffers\n", sc->dc_unit);
dc_stop(sc);
First round of converting network drivers from spls to mutexes. This takes care of all the 10/100 and gigE PCI drivers that I've done. Next will be the wireless drivers, then the USB ones. I may pick up some stragglers along the way. I'm sort of playing this by ear: if anyone spots any places where I've screwed up horribly, please let me know.
2000-10-13 17:54:19 +00:00
DC_UNLOCK(sc);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
return;
}
/*
* Init tx descriptors.
*/
dc_list_tx_init(sc);
/*
* Load the address of the RX list.
*/
CSR_WRITE_4(sc, DC_RXADDR, vtophys(&sc->dc_ldata->dc_rx_list[0]));
CSR_WRITE_4(sc, DC_TXADDR, vtophys(&sc->dc_ldata->dc_tx_list[0]));
/*
* Enable interrupts.
*/
Device Polling code for -current. Non-SMP, i386-only, no polling in the idle loop at the moment. To use this code you must compile a kernel with options DEVICE_POLLING and at runtime enable polling with sysctl kern.polling.enable=1 The percentage of CPU reserved to userland can be set with sysctl kern.polling.user_frac=NN (default is 50) while the remainder is used by polling device drivers and netisr's. These are the only two variables that you should need to touch. There are a few more parameters in kern.polling but the default values are adequate for all purposes. See the code in kern_poll.c for more details on them. Polling in the idle loop will be implemented shortly by introducing a kernel thread which does the job. Until then, the amount of CPU dedicated to polling will never exceed (100-user_frac). The equivalent (actually, better) code for -stable is at http://info.iet.unipi.it/~luigi/polling/ and also supports polling in the idle loop. NOTE to Alpha developers: There is really nothing in this code that is i386-specific. If you move the 2 lines supporting the new option from sys/conf/{files,options}.i386 to sys/conf/{files,options} I am pretty sure that this should work on the Alpha as well, just that I do not have a suitable test box to try it. If someone feels like trying it, I would appreciate it. NOTE to other developers: sure some things could be done better, and as always I am open to constructive criticism, which a few of you have already given and I greatly appreciated. However, before proposing radical architectural changes, please take some time to possibly try out this code, or at the very least read the comments in kern_poll.c, especially re. the reason why I am using a soft netisr and cannot (I believe) replace it with a simple timeout. Quick description of files touched by this commit: sys/conf/files.i386 new file kern/kern_poll.c sys/conf/options.i386 new option sys/i386/i386/trap.c poll in trap (disabled by default) sys/kern/kern_clock.c initialization and hardclock hooks. sys/kern/kern_intr.c minor swi_net changes sys/kern/kern_poll.c the bulk of the code. sys/net/if.h new flag sys/net/if_var.h declaration for functions used in device drivers. sys/net/netisr.h NETISR_POLL sys/dev/fxp/if_fxp.c sys/dev/fxp/if_fxpvar.h sys/pci/if_dc.c sys/pci/if_dcreg.h sys/pci/if_sis.c sys/pci/if_sisreg.h device driver modifications
2001-12-14 17:56:12 +00:00
#ifdef DEVICE_POLLING
/*
* ... but only if we are not polling, and make sure they are off in
* the case of polling. Some cards (e.g. fxp) turn interrupts on
* after a reset.
*/
Increase size of ifnet.if_flags from 16 bits (short) to 32 bits (int). To avoid breaking application ABI use unused ifreq.ifru_flags[1] for upper 16 bits in SIOCSIFFLAGS and SIOCGIFFLAGS ioctl's. Reviewed by: -hackers, -net
2002-08-18 07:05:00 +00:00
if (ifp->if_flags & IFF_POLLING)
Device Polling code for -current. Non-SMP, i386-only, no polling in the idle loop at the moment. To use this code you must compile a kernel with options DEVICE_POLLING and at runtime enable polling with sysctl kern.polling.enable=1 The percentage of CPU reserved to userland can be set with sysctl kern.polling.user_frac=NN (default is 50) while the remainder is used by polling device drivers and netisr's. These are the only two variables that you should need to touch. There are a few more parameters in kern.polling but the default values are adequate for all purposes. See the code in kern_poll.c for more details on them. Polling in the idle loop will be implemented shortly by introducing a kernel thread which does the job. Until then, the amount of CPU dedicated to polling will never exceed (100-user_frac). The equivalent (actually, better) code for -stable is at http://info.iet.unipi.it/~luigi/polling/ and also supports polling in the idle loop. NOTE to Alpha developers: There is really nothing in this code that is i386-specific. If you move the 2 lines supporting the new option from sys/conf/{files,options}.i386 to sys/conf/{files,options} I am pretty sure that this should work on the Alpha as well, just that I do not have a suitable test box to try it. If someone feels like trying it, I would appreciate it. NOTE to other developers: sure some things could be done better, and as always I am open to constructive criticism, which a few of you have already given and I greatly appreciated. However, before proposing radical architectural changes, please take some time to possibly try out this code, or at the very least read the comments in kern_poll.c, especially re. the reason why I am using a soft netisr and cannot (I believe) replace it with a simple timeout. Quick description of files touched by this commit: sys/conf/files.i386 new file kern/kern_poll.c sys/conf/options.i386 new option sys/i386/i386/trap.c poll in trap (disabled by default) sys/kern/kern_clock.c initialization and hardclock hooks. sys/kern/kern_intr.c minor swi_net changes sys/kern/kern_poll.c the bulk of the code. sys/net/if.h new flag sys/net/if_var.h declaration for functions used in device drivers. sys/net/netisr.h NETISR_POLL sys/dev/fxp/if_fxp.c sys/dev/fxp/if_fxpvar.h sys/pci/if_dc.c sys/pci/if_dcreg.h sys/pci/if_sis.c sys/pci/if_sisreg.h device driver modifications
2001-12-14 17:56:12 +00:00
CSR_WRITE_4(sc, DC_IMR, 0x00000000);
else
#endif
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
CSR_WRITE_4(sc, DC_IMR, DC_INTRS);
CSR_WRITE_4(sc, DC_ISR, 0xFFFFFFFF);
/* Enable transmitter. */
DC_SETBIT(sc, DC_NETCFG, DC_NETCFG_TX_ON);
Make the blinkylights on non-MII 21143 cards work. We need to enable the link and activity LED control bits in CSR15 in order for the controller to drive the LEDs correctly. This was largely done for the ZNYX multiport cards, but should also work with the DEC DE500-BA and other non-MII cards.
2000-09-01 23:59:54 +00:00
/*
* If this is an Intel 21143 and we're not using the
* MII port, program the LED control pins so we get
* link and activity indications.
*/
Special-case the LED twiddling code so that it doesn't do anything on the NEC VersaPro NoteBook PC. This 21143 implementation has no LEDs, and flipping the LED control bits somehow stops it from establishing a link. We check the subsystem ID and don't flip the LED control bits for the NEC NIC.
2000-09-07 18:51:04 +00:00
if (sc->dc_flags & DC_TULIP_LEDS) {
Make the blinkylights on non-MII 21143 cards work. We need to enable the link and activity LED control bits in CSR15 in order for the controller to drive the LEDs correctly. This was largely done for the ZNYX multiport cards, but should also work with the DEC DE500-BA and other non-MII cards.
2000-09-01 23:59:54 +00:00
CSR_WRITE_4(sc, DC_WATCHDOG,
DC_WDOG_CTLWREN|DC_WDOG_LINK|DC_WDOG_ACTIVITY);
Special-case the LED twiddling code so that it doesn't do anything on the NEC VersaPro NoteBook PC. This 21143 implementation has no LEDs, and flipping the LED control bits somehow stops it from establishing a link. We check the subsystem ID and don't flip the LED control bits for the NEC NIC.
2000-09-07 18:51:04 +00:00
CSR_WRITE_4(sc, DC_WATCHDOG, 0);
Make the blinkylights on non-MII 21143 cards work. We need to enable the link and activity LED control bits in CSR15 in order for the controller to drive the LEDs correctly. This was largely done for the ZNYX multiport cards, but should also work with the DEC DE500-BA and other non-MII cards.
2000-09-01 23:59:54 +00:00
}
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
/*
* Load the RX/multicast filter. We do this sort of late
* because the filter programming scheme on the 21143 and
* some clones requires DMAing a setup frame via the TX
* engine, and we need the transmitter enabled for that.
*/
dc_setfilt(sc);
/* Enable receiver. */
DC_SETBIT(sc, DC_NETCFG, DC_NETCFG_RX_ON);
CSR_WRITE_4(sc, DC_RXSTART, 0xFFFFFFFF);
mii_mediachg(mii);
dc_setcfg(sc, sc->dc_if_media);
ifp->if_flags |= IFF_RUNNING;
ifp->if_flags &= ~IFF_OACTIVE;
If this is a Davicom DM9102A and we're enabling the homePNA link, force dc_link to 1 and don't activate the tick routine. Without this, dc_start() always thinks the link is down and never transmits in homePNA mode.
2000-09-20 00:59:17 +00:00
/* Don't start the ticker if this is a homePNA link. */
Move us yet closer to IFM_* definitions in NetBSD.
2002-04-29 05:32:44 +00:00
if (IFM_SUBTYPE(mii->mii_media.ifm_media) == IFM_HPNA_1)
If this is a Davicom DM9102A and we're enabling the homePNA link, force dc_link to 1 and don't activate the tick routine. Without this, dc_start() always thinks the link is down and never transmits in homePNA mode.
2000-09-20 00:59:17 +00:00
sc->dc_link = 1;
else {
if (sc->dc_flags & DC_21143_NWAY)
Change the driver to allocate its own callout structure, and modify the interface to use callout_* instead of timeout(). Also add an IS_MPSAFE #define (currently off) which will mark the driver as mpsafe to the upper layers.
2000-11-25 08:00:10 +00:00
callout_reset(&sc->dc_stat_ch, hz/10, dc_tick, sc);
If this is a Davicom DM9102A and we're enabling the homePNA link, force dc_link to 1 and don't activate the tick routine. Without this, dc_start() always thinks the link is down and never transmits in homePNA mode.
2000-09-20 00:59:17 +00:00
else
Change the driver to allocate its own callout structure, and modify the interface to use callout_* instead of timeout(). Also add an IS_MPSAFE #define (currently off) which will mark the driver as mpsafe to the upper layers.
2000-11-25 08:00:10 +00:00
callout_reset(&sc->dc_stat_ch, hz, dc_tick, sc);
If this is a Davicom DM9102A and we're enabling the homePNA link, force dc_link to 1 and don't activate the tick routine. Without this, dc_start() always thinks the link is down and never transmits in homePNA mode.
2000-09-20 00:59:17 +00:00
}
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
Add support for parsing the media blocks from the SROM on 21143 adapters. This is necessary in order to make this driver work with the built-in ethernet on the alpha Miata machines. These systems have a 21143-PC chip on-board and optional daughtercards with either a 10/100 MII transceiver or a 10baseT/10base2 transceiver. In both cases, you need to twiddle the GPIO bits on the controller in order to turn the transceivers on, and you have to read the media info from the SROM in order to find out what bits to twiddle.
2000-10-05 17:36:14 +00:00
#ifdef SRM_MEDIA
Make consistent; turn spaces into tabs where there is a mixture.
2002-09-04 18:14:17 +00:00
if(sc->dc_srm_media) {
Teach the 'dc' driver how to pick up settings left over by the SRM on alpha systems. This is an expedient if not entirely elegant solution to the problem. Submitted by: gallatin Approved by: jkh
2000-03-11 05:20:56 +00:00
struct ifreq ifr;
ifr.ifr_media = sc->dc_srm_media;
ifmedia_ioctl(ifp, &ifr, &mii->mii_media, SIOCSIFMEDIA);
sc->dc_srm_media = 0;
}
#endif
First round of converting network drivers from spls to mutexes. This takes care of all the 10/100 and gigE PCI drivers that I've done. Next will be the wireless drivers, then the USB ones. I may pick up some stragglers along the way. I'm sort of playing this by ear: if anyone spots any places where I've screwed up horribly, please let me know.
2000-10-13 17:54:19 +00:00
DC_UNLOCK(sc);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
return;
}
/*
* Set media options.
*/
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static int
dc_ifmedia_upd(ifp)
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
struct ifnet *ifp;
{
struct dc_softc *sc;
struct mii_data *mii;
Add support for DM9102A boards with Davicom DM9801 HomePNA PHYs.
2000-01-24 17:19:37 +00:00
struct ifmedia *ifm;
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
sc = ifp->if_softc;
mii = device_get_softc(sc->dc_miibus);
mii_mediachg(mii);
Add support for DM9102A boards with Davicom DM9801 HomePNA PHYs.
2000-01-24 17:19:37 +00:00
ifm = &mii->mii_media;
if (DC_IS_DAVICOM(sc) &&
Move us yet closer to IFM_* definitions in NetBSD.
2002-04-29 05:32:44 +00:00
IFM_SUBTYPE(ifm->ifm_media) == IFM_HPNA_1)
Add support for DM9102A boards with Davicom DM9801 HomePNA PHYs.
2000-01-24 17:19:37 +00:00
dc_setcfg(sc, ifm->ifm_media);
else
sc->dc_link = 0;
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
return(0);
}
/*
* Report current media status.
*/
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static void
dc_ifmedia_sts(ifp, ifmr)
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
struct ifnet *ifp;
struct ifmediareq *ifmr;
{
struct dc_softc *sc;
struct mii_data *mii;
Add support for DM9102A boards with Davicom DM9801 HomePNA PHYs.
2000-01-24 17:19:37 +00:00
struct ifmedia *ifm;
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
sc = ifp->if_softc;
mii = device_get_softc(sc->dc_miibus);
mii_pollstat(mii);
Add support for DM9102A boards with Davicom DM9801 HomePNA PHYs.
2000-01-24 17:19:37 +00:00
ifm = &mii->mii_media;
if (DC_IS_DAVICOM(sc)) {
Move us yet closer to IFM_* definitions in NetBSD.
2002-04-29 05:32:44 +00:00
if (IFM_SUBTYPE(ifm->ifm_media) == IFM_HPNA_1) {
Add support for DM9102A boards with Davicom DM9801 HomePNA PHYs.
2000-01-24 17:19:37 +00:00
ifmr->ifm_active = ifm->ifm_media;
ifmr->ifm_status = 0;
return;
}
}
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
ifmr->ifm_active = mii->mii_media_active;
ifmr->ifm_status = mii->mii_media_status;
return;
}
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static int
dc_ioctl(ifp, command, data)
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
struct ifnet *ifp;
u_long command;
caddr_t data;
{
struct dc_softc *sc = ifp->if_softc;
struct ifreq *ifr = (struct ifreq *) data;
struct mii_data *mii;
First round of converting network drivers from spls to mutexes. This takes care of all the 10/100 and gigE PCI drivers that I've done. Next will be the wireless drivers, then the USB ones. I may pick up some stragglers along the way. I'm sort of playing this by ear: if anyone spots any places where I've screwed up horribly, please let me know.
2000-10-13 17:54:19 +00:00
int error = 0;
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
First round of converting network drivers from spls to mutexes. This takes care of all the 10/100 and gigE PCI drivers that I've done. Next will be the wireless drivers, then the USB ones. I may pick up some stragglers along the way. I'm sort of playing this by ear: if anyone spots any places where I've screwed up horribly, please let me know.
2000-10-13 17:54:19 +00:00
DC_LOCK(sc);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
switch(command) {
case SIOCSIFFLAGS:
if (ifp->if_flags & IFF_UP) {
Fix handling of IFF_ALLMULTI. The same bug in various forms affects the following drivers: dc mn sf sk ste ti tl xl an bge em gem gx ie lge sr aue cue kue wi xe Approved by: re
2002-11-25 19:28:01 +00:00
int need_setfilt = (ifp->if_flags ^ sc->dc_if_flags) &
(IFF_PROMISC | IFF_ALLMULTI);
if (ifp->if_flags & IFF_RUNNING) {
if (need_setfilt)
dc_setfilt(sc);
} else {
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
sc->dc_txthresh = 0;
dc_init(sc);
}
} else {
if (ifp->if_flags & IFF_RUNNING)
dc_stop(sc);
}
sc->dc_if_flags = ifp->if_flags;
error = 0;
break;
case SIOCADDMULTI:
case SIOCDELMULTI:
dc_setfilt(sc);
error = 0;
break;
case SIOCGIFMEDIA:
case SIOCSIFMEDIA:
mii = device_get_softc(sc->dc_miibus);
error = ifmedia_ioctl(ifp, ifr, &mii->mii_media, command);
Add support for parsing the media blocks from the SROM on 21143 adapters. This is necessary in order to make this driver work with the built-in ethernet on the alpha Miata machines. These systems have a 21143-PC chip on-board and optional daughtercards with either a 10/100 MII transceiver or a 10baseT/10base2 transceiver. In both cases, you need to twiddle the GPIO bits on the controller in order to turn the transceivers on, and you have to read the media info from the SROM in order to find out what bits to twiddle.
2000-10-05 17:36:14 +00:00
#ifdef SRM_MEDIA
Teach the 'dc' driver how to pick up settings left over by the SRM on alpha systems. This is an expedient if not entirely elegant solution to the problem. Submitted by: gallatin Approved by: jkh
2000-03-11 05:20:56 +00:00
if (sc->dc_srm_media)
sc->dc_srm_media = 0;
#endif
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
break;
default:
o track either_ifattach/ether_ifdetach API changes o use if_input for input packet processing o don't strip the Ethernet header for input packets o use BPF_* macros bpf tapping o call ether_ioctl to handle default ioctl case o track vlan changes Reviewed by: many Approved by: re
2002-11-14 23:49:09 +00:00
error = ether_ioctl(ifp, command, data);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
break;
}
First round of converting network drivers from spls to mutexes. This takes care of all the 10/100 and gigE PCI drivers that I've done. Next will be the wireless drivers, then the USB ones. I may pick up some stragglers along the way. I'm sort of playing this by ear: if anyone spots any places where I've screwed up horribly, please let me know.
2000-10-13 17:54:19 +00:00
DC_UNLOCK(sc);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
return(error);
}
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static void
dc_watchdog(ifp)
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
struct ifnet *ifp;
{
struct dc_softc *sc;
sc = ifp->if_softc;
First round of converting network drivers from spls to mutexes. This takes care of all the 10/100 and gigE PCI drivers that I've done. Next will be the wireless drivers, then the USB ones. I may pick up some stragglers along the way. I'm sort of playing this by ear: if anyone spots any places where I've screwed up horribly, please let me know.
2000-10-13 17:54:19 +00:00
DC_LOCK(sc);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
ifp->if_oerrors++;
printf("dc%d: watchdog timeout\n", sc->dc_unit);
dc_stop(sc);
dc_reset(sc);
dc_init(sc);
if (ifp->if_snd.ifq_head != NULL)
dc_start(ifp);
First round of converting network drivers from spls to mutexes. This takes care of all the 10/100 and gigE PCI drivers that I've done. Next will be the wireless drivers, then the USB ones. I may pick up some stragglers along the way. I'm sort of playing this by ear: if anyone spots any places where I've screwed up horribly, please let me know.
2000-10-13 17:54:19 +00:00
DC_UNLOCK(sc);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
return;
}
/*
* Stop the adapter and free any mbufs allocated to the
* RX and TX lists.
*/
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static void
dc_stop(sc)
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
struct dc_softc *sc;
{
register int i;
struct ifnet *ifp;
First round of converting network drivers from spls to mutexes. This takes care of all the 10/100 and gigE PCI drivers that I've done. Next will be the wireless drivers, then the USB ones. I may pick up some stragglers along the way. I'm sort of playing this by ear: if anyone spots any places where I've screwed up horribly, please let me know.
2000-10-13 17:54:19 +00:00
DC_LOCK(sc);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
ifp = &sc->arpcom.ac_if;
ifp->if_timer = 0;
Change the driver to allocate its own callout structure, and modify the interface to use callout_* instead of timeout(). Also add an IS_MPSAFE #define (currently off) which will mark the driver as mpsafe to the upper layers.
2000-11-25 08:00:10 +00:00
callout_stop(&sc->dc_stat_ch);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
MFS (merge from stable): rev 1.9.2.28, fix ordering of IFF_RUNNING mods. The reason we are required to commit to -current first is so that later MFC's do not risk the loss of existing bug fixes. Even if this was not strictly required in -current, it should still be fixed there too.
2001-12-07 00:57:57 +00:00
ifp->if_flags &= ~(IFF_RUNNING | IFF_OACTIVE);
Device Polling code for -current. Non-SMP, i386-only, no polling in the idle loop at the moment. To use this code you must compile a kernel with options DEVICE_POLLING and at runtime enable polling with sysctl kern.polling.enable=1 The percentage of CPU reserved to userland can be set with sysctl kern.polling.user_frac=NN (default is 50) while the remainder is used by polling device drivers and netisr's. These are the only two variables that you should need to touch. There are a few more parameters in kern.polling but the default values are adequate for all purposes. See the code in kern_poll.c for more details on them. Polling in the idle loop will be implemented shortly by introducing a kernel thread which does the job. Until then, the amount of CPU dedicated to polling will never exceed (100-user_frac). The equivalent (actually, better) code for -stable is at http://info.iet.unipi.it/~luigi/polling/ and also supports polling in the idle loop. NOTE to Alpha developers: There is really nothing in this code that is i386-specific. If you move the 2 lines supporting the new option from sys/conf/{files,options}.i386 to sys/conf/{files,options} I am pretty sure that this should work on the Alpha as well, just that I do not have a suitable test box to try it. If someone feels like trying it, I would appreciate it. NOTE to other developers: sure some things could be done better, and as always I am open to constructive criticism, which a few of you have already given and I greatly appreciated. However, before proposing radical architectural changes, please take some time to possibly try out this code, or at the very least read the comments in kern_poll.c, especially re. the reason why I am using a soft netisr and cannot (I believe) replace it with a simple timeout. Quick description of files touched by this commit: sys/conf/files.i386 new file kern/kern_poll.c sys/conf/options.i386 new option sys/i386/i386/trap.c poll in trap (disabled by default) sys/kern/kern_clock.c initialization and hardclock hooks. sys/kern/kern_intr.c minor swi_net changes sys/kern/kern_poll.c the bulk of the code. sys/net/if.h new flag sys/net/if_var.h declaration for functions used in device drivers. sys/net/netisr.h NETISR_POLL sys/dev/fxp/if_fxp.c sys/dev/fxp/if_fxpvar.h sys/pci/if_dc.c sys/pci/if_dcreg.h sys/pci/if_sis.c sys/pci/if_sisreg.h device driver modifications
2001-12-14 17:56:12 +00:00
#ifdef DEVICE_POLLING
ether_poll_deregister(ifp);
#endif
MFS (merge from stable): rev 1.9.2.28, fix ordering of IFF_RUNNING mods. The reason we are required to commit to -current first is so that later MFC's do not risk the loss of existing bug fixes. Even if this was not strictly required in -current, it should still be fixed there too.
2001-12-07 00:57:57 +00:00
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
DC_CLRBIT(sc, DC_NETCFG, (DC_NETCFG_RX_ON|DC_NETCFG_TX_ON));
CSR_WRITE_4(sc, DC_IMR, 0x00000000);
CSR_WRITE_4(sc, DC_TXADDR, 0x00000000);
CSR_WRITE_4(sc, DC_RXADDR, 0x00000000);
sc->dc_link = 0;
/*
* Free data in the RX lists.
*/
for (i = 0; i < DC_RX_LIST_CNT; i++) {
if (sc->dc_cdata.dc_rx_chain[i] != NULL) {
m_freem(sc->dc_cdata.dc_rx_chain[i]);
sc->dc_cdata.dc_rx_chain[i] = NULL;
}
}
bzero((char *)&sc->dc_ldata->dc_rx_list,
sizeof(sc->dc_ldata->dc_rx_list));
/*
* Free the TX list buffers.
*/
for (i = 0; i < DC_TX_LIST_CNT; i++) {
if (sc->dc_cdata.dc_tx_chain[i] != NULL) {
if (sc->dc_ldata->dc_tx_list[i].dc_ctl &
DC_TXCTL_SETUP) {
sc->dc_cdata.dc_tx_chain[i] = NULL;
continue;
}
m_freem(sc->dc_cdata.dc_tx_chain[i]);
sc->dc_cdata.dc_tx_chain[i] = NULL;
}
}
bzero((char *)&sc->dc_ldata->dc_tx_list,
sizeof(sc->dc_ldata->dc_tx_list));
First round of converting network drivers from spls to mutexes. This takes care of all the 10/100 and gigE PCI drivers that I've done. Next will be the wireless drivers, then the USB ones. I may pick up some stragglers along the way. I'm sort of playing this by ear: if anyone spots any places where I've screwed up horribly, please let me know.
2000-10-13 17:54:19 +00:00
DC_UNLOCK(sc);
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
return;
}
Add suspend/resume code mostly merged from fxp/rl driver. This is temporary hack, better and generalized solution probably should be implemented at lower layer(MII or PCI?). Tested by: shoko.araki@soliton.co.jp MFC after: 1 week
2002-05-06 10:55:42 +00:00
/*
* Device suspend routine. Stop the interface and save some PCI
* settings in case the BIOS doesn't restore them properly on
* resume.
*/
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static int
dc_suspend(dev)
Add suspend/resume code mostly merged from fxp/rl driver. This is temporary hack, better and generalized solution probably should be implemented at lower layer(MII or PCI?). Tested by: shoko.araki@soliton.co.jp MFC after: 1 week
2002-05-06 10:55:42 +00:00
device_t dev;
{
register int i;
int s;
struct dc_softc *sc;
s = splimp();
sc = device_get_softc(dev);
dc_stop(sc);
for (i = 0; i < 5; i++)
sc->saved_maps[i] = pci_read_config(dev, PCIR_MAPS + i * 4, 4);
sc->saved_biosaddr = pci_read_config(dev, PCIR_BIOS, 4);
sc->saved_intline = pci_read_config(dev, PCIR_INTLINE, 1);
sc->saved_cachelnsz = pci_read_config(dev, PCIR_CACHELNSZ, 1);
sc->saved_lattimer = pci_read_config(dev, PCIR_LATTIMER, 1);
sc->suspended = 1;
splx(s);
return (0);
}
/*
* Device resume routine. Restore some PCI settings in case the BIOS
* doesn't, re-enable busmastering, and restart the interface if
* appropriate.
*/
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static int
dc_resume(dev)
Add suspend/resume code mostly merged from fxp/rl driver. This is temporary hack, better and generalized solution probably should be implemented at lower layer(MII or PCI?). Tested by: shoko.araki@soliton.co.jp MFC after: 1 week
2002-05-06 10:55:42 +00:00
device_t dev;
{
register int i;
int s;
struct dc_softc *sc;
struct ifnet *ifp;
s = splimp();
sc = device_get_softc(dev);
ifp = &sc->arpcom.ac_if;
dc_acpi(dev);
/* better way to do this? */
for (i = 0; i < 5; i++)
pci_write_config(dev, PCIR_MAPS + i * 4, sc->saved_maps[i], 4);
pci_write_config(dev, PCIR_BIOS, sc->saved_biosaddr, 4);
pci_write_config(dev, PCIR_INTLINE, sc->saved_intline, 1);
pci_write_config(dev, PCIR_CACHELNSZ, sc->saved_cachelnsz, 1);
pci_write_config(dev, PCIR_LATTIMER, sc->saved_lattimer, 1);
/* reenable busmastering */
pci_enable_busmaster(dev);
pci_enable_io(dev, DC_RES);
Make consistent; turn spaces into tabs where there is a mixture.
2002-09-04 18:14:17 +00:00
/* reinitialize interface if necessary */
if (ifp->if_flags & IFF_UP)
dc_init(sc);
Add suspend/resume code mostly merged from fxp/rl driver. This is temporary hack, better and generalized solution probably should be implemented at lower layer(MII or PCI?). Tested by: shoko.araki@soliton.co.jp MFC after: 1 week
2002-05-06 10:55:42 +00:00
sc->suspended = 0;
splx(s);
return (0);
}
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
/*
* Stop all chip I/O so that the kernel's probe routines don't
* get confused by errant DMAs when rebooting.
*/
style: put return values on a line by themselves. fix some paste issues where whitespace was used instead of tabs.
2002-08-23 23:49:02 +00:00
static void
dc_shutdown(dev)
Add the if_dc driver and remove all of the al, ax, dm, pn and mx drivers which it replaces. The new driver supports all of the chips supported by the ones it replaces, as well as many DEC/Intel 21143 10/100 cards. This also completes my quest to convert things to miibus and add Alpha support.
1999-12-04 17:41:31 +00:00
device_t dev;
{
struct dc_softc *sc;
sc = device_get_softc(dev);
dc_stop(sc);
return;
}
Reference in New Issue Copy Permalink