d/freebsd-nq
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
4a431baa58
freebsd-nq/sys/dev/ed/if_ed_pccard.c

538 lines
15 KiB
C
Raw Normal View History

Break out the bus front end code into separate files. Note that these haven't been turned on nor has the old code been removed from if_ed.c. The next commit will address that.
1999-10-15 01:28:04 +00:00
/*
* Copyright (c) 1995, David Greenman
* 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 unmodified, 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.
*
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR 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 THE AUTHOR OR CONTRIBUTORS 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.
*
* $FreeBSD$
*/
Add an option ED_NO_MIIBUS, which causes the `ed' driver to be built without support for miibus PHYs. Most ed cards don't need miibus support, so it's useful to be able to avoid the bloat of all the mii devices for small fixed-purpose kernels.
2001-09-29 22:32:03 +00:00
#include "opt_ed.h"
Break out the bus front end code into separate files. Note that these haven't been turned on nor has the old code been removed from if_ed.c. The next commit will address that.
1999-10-15 01:28:04 +00:00
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/socket.h>
Add pccard attachment. This won't work with memory mapped ed devices yet, but that should be resolved shortly. Non memory mapped ed devices should work, but I cannot test this since my only ed card is memory mapped. Submitted by: Matt Dodd <mdodd@freebsd.org>
1999-10-25 02:55:14 +00:00
#include <sys/kernel.h>
MFPAO: Add support for AX88190, equipped in MELCO LPC3-TX.
2000-07-23 15:15:43 +00:00
#include <sys/conf.h>
#include <sys/uio.h>
Break out the bus front end code into separate files. Note that these haven't been turned on nor has the old code been removed from if_ed.c. The next commit will address that.
1999-10-15 01:28:04 +00:00
#include <sys/module.h>
Add pccard attachment. This won't work with memory mapped ed devices yet, but that should be resolved shortly. Non memory mapped ed devices should work, but I cannot test this since my only ed card is memory mapped. Submitted by: Matt Dodd <mdodd@freebsd.org>
1999-10-25 02:55:14 +00:00
#include <sys/bus.h>
#include <machine/bus.h>
Allocate memory resource to access the attribute memory of a card. Reviewed by: imp
2000-09-25 06:01:42 +00:00
#include <sys/rman.h>
#include <machine/resource.h>
Add pccard attachment. This won't work with memory mapped ed devices yet, but that should be resolved shortly. Non memory mapped ed devices should work, but I cannot test this since my only ed card is memory mapped. Submitted by: Matt Dodd <mdodd@freebsd.org>
1999-10-25 02:55:14 +00:00
#include <net/ethernet.h>
#include <net/if.h>
#include <net/if_arp.h>
#include <net/if_mib.h>
Add support for Dlink DL10022 to the ed driver. This is a mii part bolted to a ne-2000 chip. This is necessary for the NetGear FA-410TX and other cards. This also requires you add mii to your kernel if you have an ed driver configured. This code will result in a couple of timeout messages for ed on the impacted cards. Additional work will be needed, but this does work right now, and many people need these cards. Submitted by: Ian Dowse <iedowse@maths.tcd.ie>
2001-03-03 08:31:37 +00:00
#include <net/if_media.h>
Add pccard attachment. This won't work with memory mapped ed devices yet, but that should be resolved shortly. Non memory mapped ed devices should work, but I cannot test this since my only ed card is memory mapped. Submitted by: Matt Dodd <mdodd@freebsd.org>
1999-10-25 02:55:14 +00:00
MFPAO: Add support for AX88190, equipped in MELCO LPC3-TX.
2000-07-23 15:15:43 +00:00
#include <dev/ed/if_edreg.h>
Break out the bus front end code into separate files. Note that these haven't been turned on nor has the old code been removed from if_ed.c. The next commit will address that.
1999-10-15 01:28:04 +00:00
#include <dev/ed/if_edvar.h>
Propagate MAC address in CIS tupple to dev/ed/ driver. Many ed-based Ethernet PC-cards can't get correct MAC address without this patch. Submitted by: Takanori Watanabe <takawata@shidahara1.planet.sci.kobe-u.ac.jp> Reviewed by: Warner Losh <imp@village.org>
2000-01-21 03:08:46 +00:00
#include <dev/pccard/pccardvar.h>
Update the ed driver to probe and attach under a NEWCARD kernel (I was using a cardbus based system with pccbb providing the pcic interface). Something isn't quite right.. when the driver allocates and activates its resources, the IO space that was requested reads as all zeros (versus the original 0xff's as it normally is when there is no device responding). Also, deactivate the resources before releasing them. OLDCARD doesn't seem to care but NEWCARD/CARDBUS get rather unhappy if you release a resource that hasn't been deactivated yet. Make pcic_p.c only compile with oldcard kernels.
2000-11-25 03:36:09 +00:00
#include <dev/pccard/pccarddevs.h>
Add an option ED_NO_MIIBUS, which causes the `ed' driver to be built without support for miibus PHYs. Most ed cards don't need miibus support, so it's useful to be able to avoid the bloat of all the mii devices for small fixed-purpose kernels.
2001-09-29 22:32:03 +00:00
#ifndef ED_NO_MIIBUS
Add support for Dlink DL10022 to the ed driver. This is a mii part bolted to a ne-2000 chip. This is necessary for the NetGear FA-410TX and other cards. This also requires you add mii to your kernel if you have an ed driver configured. This code will result in a couple of timeout messages for ed on the impacted cards. Additional work will be needed, but this does work right now, and many people need these cards. Submitted by: Ian Dowse <iedowse@maths.tcd.ie>
2001-03-03 08:31:37 +00:00
#include <dev/mii/mii.h>
#include <dev/mii/miivar.h>
Add an option ED_NO_MIIBUS, which causes the `ed' driver to be built without support for miibus PHYs. Most ed cards don't need miibus support, so it's useful to be able to avoid the bloat of all the mii devices for small fixed-purpose kernels.
2001-09-29 22:32:03 +00:00
#endif
Update the ed driver to probe and attach under a NEWCARD kernel (I was using a cardbus based system with pccbb providing the pcic interface). Something isn't quite right.. when the driver allocates and activates its resources, the IO space that was requested reads as all zeros (versus the original 0xff's as it normally is when there is no device responding). Also, deactivate the resources before releasing them. OLDCARD doesn't seem to care but NEWCARD/CARDBUS get rather unhappy if you release a resource that hasn't been deactivated yet. Make pcic_p.c only compile with oldcard kernels.
2000-11-25 03:36:09 +00:00
Allocate memory resource to access the attribute memory of a card. Reviewed by: imp
2000-09-25 06:01:42 +00:00
#include "card_if.h"
Add an option ED_NO_MIIBUS, which causes the `ed' driver to be built without support for miibus PHYs. Most ed cards don't need miibus support, so it's useful to be able to avoid the bloat of all the mii devices for small fixed-purpose kernels.
2001-09-29 22:32:03 +00:00
#ifndef ED_NO_MIIBUS
Add support for Dlink DL10022 to the ed driver. This is a mii part bolted to a ne-2000 chip. This is necessary for the NetGear FA-410TX and other cards. This also requires you add mii to your kernel if you have an ed driver configured. This code will result in a couple of timeout messages for ed on the impacted cards. Additional work will be needed, but this does work right now, and many people need these cards. Submitted by: Ian Dowse <iedowse@maths.tcd.ie>
2001-03-03 08:31:37 +00:00
/* "device miibus" required. See GENERIC if you get errors here. */
#include "miibus_if.h"
MODULE_DEPEND(ed, miibus, 1, 1, 1);
Add an option ED_NO_MIIBUS, which causes the `ed' driver to be built without support for miibus PHYs. Most ed cards don't need miibus support, so it's useful to be able to avoid the bloat of all the mii devices for small fixed-purpose kernels.
2001-09-29 22:32:03 +00:00
#endif
- Express hard dependencies on bus (pci, isa, pccard) and network layer (ether). - Don't abuse module names to facilitate ifconfig module loading; such abuse isn't really needed. (And if we do need type information associated with a module then we should make it explicit and not use hacks.)
2003-04-15 06:37:30 +00:00
MODULE_DEPEND(ed, ether, 1, 1, 1);
MODULE_DEPEND(ed, pccard, 1, 1, 1);
Break out the bus front end code into separate files. Note that these haven't been turned on nor has the old code been removed from if_ed.c. The next commit will address that.
1999-10-15 01:28:04 +00:00
/*
* PC-Card (PCMCIA) specific code.
*/
Update the ed driver to probe and attach under a NEWCARD kernel (I was using a cardbus based system with pccbb providing the pcic interface). Something isn't quite right.. when the driver allocates and activates its resources, the IO space that was requested reads as all zeros (versus the original 0xff's as it normally is when there is no device responding). Also, deactivate the resources before releasing them. OLDCARD doesn't seem to care but NEWCARD/CARDBUS get rather unhappy if you release a resource that hasn't been deactivated yet. Make pcic_p.c only compile with oldcard kernels.
2000-11-25 03:36:09 +00:00
static int ed_pccard_match(device_t);
Add pccard attachment. This won't work with memory mapped ed devices yet, but that should be resolved shortly. Non memory mapped ed devices should work, but I cannot test this since my only ed card is memory mapped. Submitted by: Matt Dodd <mdodd@freebsd.org>
1999-10-25 02:55:14 +00:00
static int ed_pccard_probe(device_t);
static int ed_pccard_attach(device_t);
Fix the hang on card eject problem and maybe the hang on suspend problem. o Create new timeout routine so we don't detach the card inside a ISR but instead drop back to spl0 via a timeout of 0. o Actually delete the child of the pccard device rather than just faking it badly. o Fix sio, ed and ep to have pccard detach routines that are int rather than void. o Fix ep and ed pccard detach routines to use if_detach rather than just if_down. if_detach destroys the device, while if_down just marks it down. In this incarnation of the pccard things, we map the disable the slot action to detach the driver, which removes the driver from the device tree. When that is done, a panic would soon follow as the ifconfig tried to down the device. Didn't fix: o Should cache the pccard dev child's pointer in struct slot o remove now unused parts of struct slot o Any driver using softc after detach has been called. sio's softc used to be statically allocated, so you could check sc->gone, but that is now gone. o Didn't remove gone from softc of drivers that use the old pccard method. Didn't test: o ed driver changes o sio driver changes on pccards o suspend (no laptop or apm support on my desktop)
1999-12-01 07:38:54 +00:00
static int ed_pccard_detach(device_t);
Break out the bus front end code into separate files. Note that these haven't been turned on nor has the old code been removed from if_ed.c. The next commit will address that.
1999-10-15 01:28:04 +00:00
Minor newbus/style(9) cleanups. o Move the ax88190 code to its own function. o Move all device_method_t, driver_t and DRIVER_MODULE definitions to the end of files. o Wrap a few lines > 80 characters. o Use the same devclass for all ed drivers. This allows machines with multiple types of cards to have their cards numbered correctly. Before, you could wind up with two ed0's. o Protect if_edvar.h from multiple includes because I was there.
2000-12-26 06:38:04 +00:00
static int ed_pccard_Linksys(device_t dev);
static int ed_pccard_ax88190(device_t dev);
- Fixed the conversion to bus_space interface. - Added PC-98 Cbus devices support. The original patch is submitted by chi@bd.mbn.or.jp (Chiharu Shibata) - Removed old ed driver.
2000-08-17 12:15:45 +00:00
static void ax88190_geteprom(struct ed_softc *);
MFPAO: Add support for AX88190, equipped in MELCO LPC3-TX.
2000-07-23 15:15:43 +00:00
static int ed_pccard_memwrite(device_t dev, off_t offset, u_char byte);
Add an option ED_NO_MIIBUS, which causes the `ed' driver to be built without support for miibus PHYs. Most ed cards don't need miibus support, so it's useful to be able to avoid the bloat of all the mii devices for small fixed-purpose kernels.
2001-09-29 22:32:03 +00:00
#ifndef ED_NO_MIIBUS
Add support for Dlink DL10022 to the ed driver. This is a mii part bolted to a ne-2000 chip. This is necessary for the NetGear FA-410TX and other cards. This also requires you add mii to your kernel if you have an ed driver configured. This code will result in a couple of timeout messages for ed on the impacted cards. Additional work will be needed, but this does work right now, and many people need these cards. Submitted by: Ian Dowse <iedowse@maths.tcd.ie>
2001-03-03 08:31:37 +00:00
static void ed_pccard_dlink_mii_reset(struct ed_softc *sc);
static u_int ed_pccard_dlink_mii_readbits(struct ed_softc *sc, int nbits);
static void ed_pccard_dlink_mii_writebits(struct ed_softc *sc, u_int val,
int nbits);
Add an option ED_NO_MIIBUS, which causes the `ed' driver to be built without support for miibus PHYs. Most ed cards don't need miibus support, so it's useful to be able to avoid the bloat of all the mii devices for small fixed-purpose kernels.
2001-09-29 22:32:03 +00:00
#endif
MFPAO: Add support for AX88190, equipped in MELCO LPC3-TX.
2000-07-23 15:15:43 +00:00
Break out the bus front end code into separate files. Note that these haven't been turned on nor has the old code been removed from if_ed.c. The next commit will address that.
1999-10-15 01:28:04 +00:00
/*
Add pccard attachment. This won't work with memory mapped ed devices yet, but that should be resolved shortly. Non memory mapped ed devices should work, but I cannot test this since my only ed card is memory mapped. Submitted by: Matt Dodd <mdodd@freebsd.org>
1999-10-25 02:55:14 +00:00
* ed_pccard_detach - unload the driver and clear the table.
Break out the bus front end code into separate files. Note that these haven't been turned on nor has the old code been removed from if_ed.c. The next commit will address that.
1999-10-15 01:28:04 +00:00
* XXX TODO:
* This is usually called when the card is ejected, but
* can be caused by a modunload of a controller driver.
* The idea is to reset the driver's view of the device
* and ensure that any driver entry points such as
* read and write do not hang.
*/
Fix the hang on card eject problem and maybe the hang on suspend problem. o Create new timeout routine so we don't detach the card inside a ISR but instead drop back to spl0 via a timeout of 0. o Actually delete the child of the pccard device rather than just faking it badly. o Fix sio, ed and ep to have pccard detach routines that are int rather than void. o Fix ep and ed pccard detach routines to use if_detach rather than just if_down. if_detach destroys the device, while if_down just marks it down. In this incarnation of the pccard things, we map the disable the slot action to detach the driver, which removes the driver from the device tree. When that is done, a panic would soon follow as the ifconfig tried to down the device. Didn't fix: o Should cache the pccard dev child's pointer in struct slot o remove now unused parts of struct slot o Any driver using softc after detach has been called. sio's softc used to be statically allocated, so you could check sc->gone, but that is now gone. o Didn't remove gone from softc of drivers that use the old pccard method. Didn't test: o ed driver changes o sio driver changes on pccards o suspend (no laptop or apm support on my desktop)
1999-12-01 07:38:54 +00:00
static int
Add pccard attachment. This won't work with memory mapped ed devices yet, but that should be resolved shortly. Non memory mapped ed devices should work, but I cannot test this since my only ed card is memory mapped. Submitted by: Matt Dodd <mdodd@freebsd.org>
1999-10-25 02:55:14 +00:00
ed_pccard_detach(device_t dev)
Break out the bus front end code into separate files. Note that these haven't been turned on nor has the old code been removed from if_ed.c. The next commit will address that.
1999-10-15 01:28:04 +00:00
{
Add pccard attachment. This won't work with memory mapped ed devices yet, but that should be resolved shortly. Non memory mapped ed devices should work, but I cannot test this since my only ed card is memory mapped. Submitted by: Matt Dodd <mdodd@freebsd.org>
1999-10-25 02:55:14 +00:00
struct ed_softc *sc = device_get_softc(dev);
Break out the bus front end code into separate files. Note that these haven't been turned on nor has the old code been removed from if_ed.c. The next commit will address that.
1999-10-15 01:28:04 +00:00
struct ifnet *ifp = &sc->arpcom.ac_if;
if (sc->gone) {
Add pccard attachment. This won't work with memory mapped ed devices yet, but that should be resolved shortly. Non memory mapped ed devices should work, but I cannot test this since my only ed card is memory mapped. Submitted by: Matt Dodd <mdodd@freebsd.org>
1999-10-25 02:55:14 +00:00
device_printf(dev, "already unloaded\n");
Fix the hang on card eject problem and maybe the hang on suspend problem. o Create new timeout routine so we don't detach the card inside a ISR but instead drop back to spl0 via a timeout of 0. o Actually delete the child of the pccard device rather than just faking it badly. o Fix sio, ed and ep to have pccard detach routines that are int rather than void. o Fix ep and ed pccard detach routines to use if_detach rather than just if_down. if_detach destroys the device, while if_down just marks it down. In this incarnation of the pccard things, we map the disable the slot action to detach the driver, which removes the driver from the device tree. When that is done, a panic would soon follow as the ifconfig tried to down the device. Didn't fix: o Should cache the pccard dev child's pointer in struct slot o remove now unused parts of struct slot o Any driver using softc after detach has been called. sio's softc used to be statically allocated, so you could check sc->gone, but that is now gone. o Didn't remove gone from softc of drivers that use the old pccard method. Didn't test: o ed driver changes o sio driver changes on pccards o suspend (no laptop or apm support on my desktop)
1999-12-01 07:38:54 +00:00
return (0);
Break out the bus front end code into separate files. Note that these haven't been turned on nor has the old code been removed from if_ed.c. The next commit will address that.
1999-10-15 01:28:04 +00:00
}
Fix pccard ed driver, I think. o Expose ed_stop and call it early to shutdown the hardware. o When releasing the interrupt, pass the cookie for the irq, not a pointer to the cookie (this is the base problem). o Release other resources used, just like the ep driver
1999-12-10 07:22:53 +00:00
ed_stop(sc);
Break out the bus front end code into separate files. Note that these haven't been turned on nor has the old code been removed from if_ed.c. The next commit will address that.
1999-10-15 01:28:04 +00:00
ifp->if_flags &= ~IFF_RUNNING;
network interface driver changes: o don't strip the Ethernet header from inbound packets; pass packets up the stack intact (required significant changes to some drivers) o reference common definitions in net/ethernet.h (e.g. ETHER_ALIGN) o track ether_ifattach/ether_ifdetach API changes o track bpf changes (use BPF_TAP and BPF_MTAP) o track vlan changes (ifnet capabilities, revised processing scheme, etc.) o use if_input to pass packets "up" o call ether_ioctl for default handling of ioctls Reviewed by: many Approved by: re
2002-11-14 23:54:55 +00:00
ether_ifdetach(ifp);
Break out the bus front end code into separate files. Note that these haven't been turned on nor has the old code been removed from if_ed.c. The next commit will address that.
1999-10-15 01:28:04 +00:00
sc->gone = 1;
Fix pccard ed driver, I think. o Expose ed_stop and call it early to shutdown the hardware. o When releasing the interrupt, pass the cookie for the irq, not a pointer to the cookie (this is the base problem). o Release other resources used, just like the ep driver
1999-12-10 07:22:53 +00:00
bus_teardown_intr(dev, sc->irq_res, sc->irq_handle);
ed_release_resources(dev);
Fix the hang on card eject problem and maybe the hang on suspend problem. o Create new timeout routine so we don't detach the card inside a ISR but instead drop back to spl0 via a timeout of 0. o Actually delete the child of the pccard device rather than just faking it badly. o Fix sio, ed and ep to have pccard detach routines that are int rather than void. o Fix ep and ed pccard detach routines to use if_detach rather than just if_down. if_detach destroys the device, while if_down just marks it down. In this incarnation of the pccard things, we map the disable the slot action to detach the driver, which removes the driver from the device tree. When that is done, a panic would soon follow as the ifconfig tried to down the device. Didn't fix: o Should cache the pccard dev child's pointer in struct slot o remove now unused parts of struct slot o Any driver using softc after detach has been called. sio's softc used to be statically allocated, so you could check sc->gone, but that is now gone. o Didn't remove gone from softc of drivers that use the old pccard method. Didn't test: o ed driver changes o sio driver changes on pccards o suspend (no laptop or apm support on my desktop)
1999-12-01 07:38:54 +00:00
return (0);
Break out the bus front end code into separate files. Note that these haven't been turned on nor has the old code been removed from if_ed.c. The next commit will address that.
1999-10-15 01:28:04 +00:00
}
Add a bunch of entries from NetBSD's if_ne_pcmcia.c v1.67. We still need to verify that all the entries in pccard.conf are present in this table.
2001-01-21 08:20:23 +00:00
static const struct ed_product {
struct pccard_product prod;
int flags;
#define NE2000DVF_DL10019 0x0001 /* chip is D-Link DL10019 */
#define NE2000DVF_AX88190 0x0002 /* chip is ASIX AX88190 */
} ed_pccard_products[] = {
reorganize the supported card table. We don't need the enet_vendor or enet_mcast fields, so remove them. Sort. Eliminate now duplicate entries. This reorg saves about 500 bytes in the binary. I've tested this only with a couple of cards, so please let me know if I've broken anything.
2003-04-10 07:04:39 +00:00
{ PCMCIA_CARD(ACCTON, EN2212, 0), 0},
{ PCMCIA_CARD(ALLIEDTELESIS, LA_PCM, 0), 0},
{ PCMCIA_CARD(AMBICOM, AMB8002T, 0), 0},
{ PCMCIA_CARD(BILLIONTON, LNT10TN, 0), 0},
Add Billionton CFLT10N and SMC8041 (although I'm not 100% sure the latter is needed). Remove duplicate kingston kne2 entry. Info obtained from NetBSD
2003-04-10 07:24:18 +00:00
{ PCMCIA_CARD(BILLIONTON, CFLT10N, 0), 0},
reorganize the supported card table. We don't need the enet_vendor or enet_mcast fields, so remove them. Sort. Eliminate now duplicate entries. This reorg saves about 500 bytes in the binary. I've tested this only with a couple of cards, so please let me know if I've broken anything.
2003-04-10 07:04:39 +00:00
{ PCMCIA_CARD(BUFFALO, LPC3_CLT, 0), 0},
Add support for BUFFALO LPC3-CLX 10/100Base-T PC-Card PR: 47786 Submitted by: IHA, genta -san
2003-07-19 05:36:09 +00:00
{ PCMCIA_CARD(BUFFALO, LPC3_CLX, 0), NE2000DVF_AX88190},
reorganize the supported card table. We don't need the enet_vendor or enet_mcast fields, so remove them. Sort. Eliminate now duplicate entries. This reorg saves about 500 bytes in the binary. I've tested this only with a couple of cards, so please let me know if I've broken anything.
2003-04-10 07:04:39 +00:00
{ PCMCIA_CARD(BUFFALO, LPC_CF_CLT, 0), 0},
{ PCMCIA_CARD(CNET, NE2000, 0), 0},
{ PCMCIA_CARD(COMPEX, LINKPORT_ENET_B, 0), 0},
{ PCMCIA_CARD(COREGA, ETHER_II_PCC_T, 0), 0},
{ PCMCIA_CARD(COREGA, ETHER_II_PCC_TD, 0), 0},
{ PCMCIA_CARD(COREGA, ETHER_PCC_T, 0), 0},
{ PCMCIA_CARD(COREGA, ETHER_PCC_TD, 0), 0},
{ PCMCIA_CARD(COREGA, FAST_ETHER_PCC_TX, 0), NE2000DVF_DL10019 },
{ PCMCIA_CARD(COREGA, FETHER_PCC_TXD, 0), NE2000DVF_AX88190 },
{ PCMCIA_CARD(COREGA, FETHER_PCC_TXF, 0), NE2000DVF_DL10019 },
{ PCMCIA_CARD(DAYNA, COMMUNICARD_E_1, 0), 0},
{ PCMCIA_CARD(DAYNA, COMMUNICARD_E_2, 0), 0},
{ PCMCIA_CARD(DIGITAL, DEPCMXX, 0), 0 },
{ PCMCIA_CARD(DLINK, DE650, 0), 0},
{ PCMCIA_CARD(DLINK, DE660, 0), 0 },
{ PCMCIA_CARD(DLINK, DE660PLUS, 0), 0},
{ PCMCIA_CARD(DLINK, DFE670TXD, 0), NE2000DVF_DL10019},
{ PCMCIA_CARD(DYNALINK, L10C, 0), 0},
{ PCMCIA_CARD(EDIMAX, EP4000A, 0), 0},
{ PCMCIA_CARD(EPSON, EEN10B, 0), 0},
EXPsys PCMCIA Ethernet Combo Submitted by: Ludwig Pummer
2003-09-20 18:03:07 +00:00
{ PCMCIA_CARD(EXP, THINLANCOMBO, 0), 0},
reorganize the supported card table. We don't need the enet_vendor or enet_mcast fields, so remove them. Sort. Eliminate now duplicate entries. This reorg saves about 500 bytes in the binary. I've tested this only with a couple of cards, so please let me know if I've broken anything.
2003-04-10 07:04:39 +00:00
{ PCMCIA_CARD(IBM, INFOMOVER, 0), 0},
{ PCMCIA_CARD(IODATA, PCLAT, 0), 0},
{ PCMCIA_CARD(IODATA, PCLATE, 0), 0},
{ PCMCIA_CARD(KINGSTON, KNE2, 0), 0},
{ PCMCIA_CARD(LANTECH, FASTNETTX, 0),NE2000DVF_AX88190 },
{ PCMCIA_CARD(LINKSYS, COMBO_ECARD, 0), NE2000DVF_DL10019 },
{ PCMCIA_CARD(LINKSYS, ECARD_1, 0), 0},
{ PCMCIA_CARD(LINKSYS, ECARD_2, 0), 0},
{ PCMCIA_CARD(LINKSYS, ETHERFAST, 0), NE2000DVF_DL10019 },
{ PCMCIA_CARD(LINKSYS, PCM100, 0), 0},
{ PCMCIA_CARD(LINKSYS, TRUST_COMBO_ECARD, 0), 0},
{ PCMCIA_CARD(LINKSYS, ETHERFAST, 0), NE2000DVF_DL10019 },
{ PCMCIA_CARD(MACNICA, ME1_JEIDA, 0), 0},
Add MELCO LPC3_CTX. This is the same as Buffalo card of the same name with different vendor id. Submitted by: Shizuka Kudo-san
2003-08-07 19:40:05 +00:00
{ PCMCIA_CARD(MELCO, LPC3_CLX, 0), NE2000DVF_AX88190},
reorganize the supported card table. We don't need the enet_vendor or enet_mcast fields, so remove them. Sort. Eliminate now duplicate entries. This reorg saves about 500 bytes in the binary. I've tested this only with a couple of cards, so please let me know if I've broken anything.
2003-04-10 07:04:39 +00:00
{ PCMCIA_CARD(MELCO, LPC3_TX, 0), NE2000DVF_AX88190 },
{ PCMCIA_CARD(NDC, ND5100_E, 0), 0},
{ PCMCIA_CARD(NETGEAR, FA410TXC, 0), NE2000DVF_DL10019},
{ PCMCIA_CARD(NETGEAR, FA411, 0), NE2000DVF_AX88190},
{ PCMCIA_CARD(PLANET, SMARTCOM2000, 0), 0 },
{ PCMCIA_CARD(PREMAX, PE200, 0), 0},
{ PCMCIA_CARD(RPTI, EP400, 0), 0},
{ PCMCIA_CARD(RPTI, EP401, 0), 0},
{ PCMCIA_CARD(SMC, EZCARD, 0), 0},
{ PCMCIA_CARD(SOCKET, EA_ETHER, 0), 0},
{ PCMCIA_CARD(SOCKET, LP_ETHER, 0), 0},
{ PCMCIA_CARD(SOCKET, LP_ETHER_CF, 0), 0},
The LP_ETH_10_100_CF entry needs to be tagged as a DL100019. Submitted by: Scott Renfro
2003-07-28 00:07:58 +00:00
{ PCMCIA_CARD(SOCKET, LP_ETH_10_100_CF, 0), NE2000DVF_DL10019},
reorganize the supported card table. We don't need the enet_vendor or enet_mcast fields, so remove them. Sort. Eliminate now duplicate entries. This reorg saves about 500 bytes in the binary. I've tested this only with a couple of cards, so please let me know if I've broken anything.
2003-04-10 07:04:39 +00:00
{ PCMCIA_CARD(SVEC, COMBOCARD, 0), 0},
{ PCMCIA_CARD(SVEC, LANCARD, 0), 0},
{ PCMCIA_CARD(SYNERGY21, S21810, 0), 0},
{ PCMCIA_CARD(TELECOMDEVICE, TCD_HPC100, 0), NE2000DVF_AX88190 },
{ PCMCIA_CARD(XIRCOM, CFE_10, 0), 0},
{ PCMCIA_CARD(ZONET, ZEN, 0), 0},
Migrate to PCMCIA_CARD() macros
2001-11-15 07:52:49 +00:00
{ { NULL } }
Update the ed driver to probe and attach under a NEWCARD kernel (I was using a cardbus based system with pccbb providing the pcic interface). Something isn't quite right.. when the driver allocates and activates its resources, the IO space that was requested reads as all zeros (versus the original 0xff's as it normally is when there is no device responding). Also, deactivate the resources before releasing them. OLDCARD doesn't seem to care but NEWCARD/CARDBUS get rather unhappy if you release a resource that hasn't been deactivated yet. Make pcic_p.c only compile with oldcard kernels.
2000-11-25 03:36:09 +00:00
};
static int
ed_pccard_match(device_t dev)
{
integrate from p4 repo: o Honor NE2000DVF_{AX88190,DL10019} flags by setting the ED_FLAGS_xxxx flag. o Mark linksys combo_ecard as ax88190 o Set the type_str to AX88190 for the ax88190 cards. This fixes ax88190 based cards, for the most part, but doesn't seem to fix the mii based dl10019 cards (aka linksys cards).
2002-07-22 21:28:46 +00:00
const struct ed_product *pp;
Update the ed driver to probe and attach under a NEWCARD kernel (I was using a cardbus based system with pccbb providing the pcic interface). Something isn't quite right.. when the driver allocates and activates its resources, the IO space that was requested reads as all zeros (versus the original 0xff's as it normally is when there is no device responding). Also, deactivate the resources before releasing them. OLDCARD doesn't seem to care but NEWCARD/CARDBUS get rather unhappy if you release a resource that hasn't been deactivated yet. Make pcic_p.c only compile with oldcard kernels.
2000-11-25 03:36:09 +00:00
integrate from p4 repo: o Honor NE2000DVF_{AX88190,DL10019} flags by setting the ED_FLAGS_xxxx flag. o Mark linksys combo_ecard as ax88190 o Set the type_str to AX88190 for the ax88190 cards. This fixes ax88190 based cards, for the most part, but doesn't seem to fix the mii based dl10019 cards (aka linksys cards).
2002-07-22 21:28:46 +00:00
if ((pp = (const struct ed_product *) pccard_product_lookup(dev,
Add a bunch of entries from NetBSD's if_ne_pcmcia.c v1.67. We still need to verify that all the entries in pccard.conf are present in this table.
2001-01-21 08:20:23 +00:00
(const struct pccard_product *) ed_pccard_products,
Argh, I have fixed this cut/paste error twice before. I must have committed the wrong patch. :-( sn_pccard_products[] should have been static anyway.
2000-11-25 07:25:08 +00:00
sizeof(ed_pccard_products[0]), NULL)) != NULL) {
Don't set the description if NULL.
2003-04-10 06:08:27 +00:00
if (pp->prod.pp_name != NULL)
device_set_desc(dev, pp->prod.pp_name);
integrate from p4 repo: o Honor NE2000DVF_{AX88190,DL10019} flags by setting the ED_FLAGS_xxxx flag. o Mark linksys combo_ecard as ax88190 o Set the type_str to AX88190 for the ax88190 cards. This fixes ax88190 based cards, for the most part, but doesn't seem to fix the mii based dl10019 cards (aka linksys cards).
2002-07-22 21:28:46 +00:00
if (pp->flags & NE2000DVF_DL10019)
device_set_flags(dev, ED_FLAGS_LINKSYS);
else if (pp->flags & NE2000DVF_AX88190)
device_set_flags(dev, ED_FLAGS_AX88190);
return (0);
Update the ed driver to probe and attach under a NEWCARD kernel (I was using a cardbus based system with pccbb providing the pcic interface). Something isn't quite right.. when the driver allocates and activates its resources, the IO space that was requested reads as all zeros (versus the original 0xff's as it normally is when there is no device responding). Also, deactivate the resources before releasing them. OLDCARD doesn't seem to care but NEWCARD/CARDBUS get rather unhappy if you release a resource that hasn't been deactivated yet. Make pcic_p.c only compile with oldcard kernels.
2000-11-25 03:36:09 +00:00
}
integrate from p4 repo: o Honor NE2000DVF_{AX88190,DL10019} flags by setting the ED_FLAGS_xxxx flag. o Mark linksys combo_ecard as ax88190 o Set the type_str to AX88190 for the ax88190 cards. This fixes ax88190 based cards, for the most part, but doesn't seem to fix the mii based dl10019 cards (aka linksys cards).
2002-07-22 21:28:46 +00:00
return (EIO);
Update the ed driver to probe and attach under a NEWCARD kernel (I was using a cardbus based system with pccbb providing the pcic interface). Something isn't quite right.. when the driver allocates and activates its resources, the IO space that was requested reads as all zeros (versus the original 0xff's as it normally is when there is no device responding). Also, deactivate the resources before releasing them. OLDCARD doesn't seem to care but NEWCARD/CARDBUS get rather unhappy if you release a resource that hasn't been deactivated yet. Make pcic_p.c only compile with oldcard kernels.
2000-11-25 03:36:09 +00:00
}
Break out the bus front end code into separate files. Note that these haven't been turned on nor has the old code been removed from if_ed.c. The next commit will address that.
1999-10-15 01:28:04 +00:00
/*
* Probe framework for pccards. Replicates the standard framework,
* minus the pccard driver registration and ignores the ether address
* supplied (from the CIS), relying on the probe to find it instead.
*/
static int
Add pccard attachment. This won't work with memory mapped ed devices yet, but that should be resolved shortly. Non memory mapped ed devices should work, but I cannot test this since my only ed card is memory mapped. Submitted by: Matt Dodd <mdodd@freebsd.org>
1999-10-25 02:55:14 +00:00
ed_pccard_probe(device_t dev)
Break out the bus front end code into separate files. Note that these haven't been turned on nor has the old code been removed from if_ed.c. The next commit will address that.
1999-10-15 01:28:04 +00:00
{
- Fixed the conversion to bus_space interface. - Added PC-98 Cbus devices support. The original patch is submitted by chi@bd.mbn.or.jp (Chiharu Shibata) - Removed old ed driver.
2000-08-17 12:15:45 +00:00
int error;
Minor newbus/style(9) cleanups. o Move the ax88190 code to its own function. o Move all device_method_t, driver_t and DRIVER_MODULE definitions to the end of files. o Wrap a few lines > 80 characters. o Use the same devclass for all ed drivers. This allows machines with multiple types of cards to have their cards numbered correctly. Before, you could wind up with two ed0's. o Protect if_edvar.h from multiple includes because I was there.
2000-12-26 06:38:04 +00:00
int flags = device_get_flags(dev);
MFPAO: Add support for AX88190, equipped in MELCO LPC3-TX.
2000-07-23 15:15:43 +00:00
if (ED_FLAGS_GETTYPE(flags) == ED_FLAGS_AX88190) {
Minor newbus/style(9) cleanups. o Move the ax88190 code to its own function. o Move all device_method_t, driver_t and DRIVER_MODULE definitions to the end of files. o Wrap a few lines > 80 characters. o Use the same devclass for all ed drivers. This allows machines with multiple types of cards to have their cards numbered correctly. Before, you could wind up with two ed0's. o Protect if_edvar.h from multiple includes because I was there.
2000-12-26 06:38:04 +00:00
error = ed_pccard_ax88190(dev);
Fix problem with ax88190 based cards trying to probe further after matching the ax88190.
2000-12-13 06:27:23 +00:00
goto end2;
MFPAO: Add support for AX88190, equipped in MELCO LPC3-TX.
2000-07-23 15:15:43 +00:00
}
Add pccard attachment. This won't work with memory mapped ed devices yet, but that should be resolved shortly. Non memory mapped ed devices should work, but I cannot test this since my only ed card is memory mapped. Submitted by: Matt Dodd <mdodd@freebsd.org>
1999-10-25 02:55:14 +00:00
- Fixed the conversion to bus_space interface. - Added PC-98 Cbus devices support. The original patch is submitted by chi@bd.mbn.or.jp (Chiharu Shibata) - Removed old ed driver.
2000-08-17 12:15:45 +00:00
error = ed_probe_Novell(dev, 0, flags);
Break out the bus front end code into separate files. Note that these haven't been turned on nor has the old code been removed from if_ed.c. The next commit will address that.
1999-10-15 01:28:04 +00:00
if (error == 0)
goto end;
ed_release_resources(dev);
- Fixed the conversion to bus_space interface. - Added PC-98 Cbus devices support. The original patch is submitted by chi@bd.mbn.or.jp (Chiharu Shibata) - Removed old ed driver.
2000-08-17 12:15:45 +00:00
error = ed_probe_WD80x3(dev, 0, flags);
Break out the bus front end code into separate files. Note that these haven't been turned on nor has the old code been removed from if_ed.c. The next commit will address that.
1999-10-15 01:28:04 +00:00
if (error == 0)
goto end;
ed_release_resources(dev);
Allocate memory resource to access the attribute memory of a card. Reviewed by: imp
2000-09-25 06:01:42 +00:00
goto end2;
Break out the bus front end code into separate files. Note that these haven't been turned on nor has the old code been removed from if_ed.c. The next commit will address that.
1999-10-15 01:28:04 +00:00
end:
Remove an old hack that remembered if the card type was Linksys by storing a flag in the global variable 'linksys' during the probe routine and reading it during the attach routine. We now have the ED_VENDOR_LINKSYS vendor code stored in sc->vendor, so check that instead.
2001-07-22 21:31:44 +00:00
if (ED_FLAGS_GETTYPE(flags) & ED_FLAGS_LINKSYS)
ed_pccard_Linksys(dev);
Allocate memory resource to access the attribute memory of a card. Reviewed by: imp
2000-09-25 06:01:42 +00:00
end2:
Break out the bus front end code into separate files. Note that these haven't been turned on nor has the old code been removed from if_ed.c. The next commit will address that.
1999-10-15 01:28:04 +00:00
if (error == 0)
error = ed_alloc_irq(dev, 0, 0);
ed_release_resources(dev);
return (error);
}
static int
Add pccard attachment. This won't work with memory mapped ed devices yet, but that should be resolved shortly. Non memory mapped ed devices should work, but I cannot test this since my only ed card is memory mapped. Submitted by: Matt Dodd <mdodd@freebsd.org>
1999-10-25 02:55:14 +00:00
ed_pccard_attach(device_t dev)
Break out the bus front end code into separate files. Note that these haven't been turned on nor has the old code been removed from if_ed.c. The next commit will address that.
1999-10-15 01:28:04 +00:00
{
Add pccard attachment. This won't work with memory mapped ed devices yet, but that should be resolved shortly. Non memory mapped ed devices should work, but I cannot test this since my only ed card is memory mapped. Submitted by: Matt Dodd <mdodd@freebsd.org>
1999-10-25 02:55:14 +00:00
int error;
Propagate MAC address in CIS tupple to dev/ed/ driver. Many ed-based Ethernet PC-cards can't get correct MAC address without this patch. Submitted by: Takanori Watanabe <takawata@shidahara1.planet.sci.kobe-u.ac.jp> Reviewed by: Warner Losh <imp@village.org>
2000-01-21 03:08:46 +00:00
int i;
Minor newbus/style(9) cleanups. o Move the ax88190 code to its own function. o Move all device_method_t, driver_t and DRIVER_MODULE definitions to the end of files. o Wrap a few lines > 80 characters. o Use the same devclass for all ed drivers. This allows machines with multiple types of cards to have their cards numbered correctly. Before, you could wind up with two ed0's. o Protect if_edvar.h from multiple includes because I was there.
2000-12-26 06:38:04 +00:00
struct ed_softc *sc = device_get_softc(dev);
Propagate MAC address in CIS tupple to dev/ed/ driver. Many ed-based Ethernet PC-cards can't get correct MAC address without this patch. Submitted by: Takanori Watanabe <takawata@shidahara1.planet.sci.kobe-u.ac.jp> Reviewed by: Warner Losh <imp@village.org>
2000-01-21 03:08:46 +00:00
u_char sum;
u_char ether_addr[ETHER_ADDR_LEN];
Add pccard attachment. This won't work with memory mapped ed devices yet, but that should be resolved shortly. Non memory mapped ed devices should work, but I cannot test this since my only ed card is memory mapped. Submitted by: Matt Dodd <mdodd@freebsd.org>
1999-10-25 02:55:14 +00:00
if (sc->port_used > 0)
Set port_used and mem_used to the resource size instead of 1. Submitted by: chi@bd.mbn.or.jp (Chiharu Shibata)
1999-11-18 14:18:19 +00:00
ed_alloc_port(dev, sc->port_rid, sc->port_used);
Add pccard attachment. This won't work with memory mapped ed devices yet, but that should be resolved shortly. Non memory mapped ed devices should work, but I cannot test this since my only ed card is memory mapped. Submitted by: Matt Dodd <mdodd@freebsd.org>
1999-10-25 02:55:14 +00:00
if (sc->mem_used)
Set port_used and mem_used to the resource size instead of 1. Submitted by: chi@bd.mbn.or.jp (Chiharu Shibata)
1999-11-18 14:18:19 +00:00
ed_alloc_memory(dev, sc->mem_rid, sc->mem_used);
Add pccard attachment. This won't work with memory mapped ed devices yet, but that should be resolved shortly. Non memory mapped ed devices should work, but I cannot test this since my only ed card is memory mapped. Submitted by: Matt Dodd <mdodd@freebsd.org>
1999-10-25 02:55:14 +00:00
ed_alloc_irq(dev, sc->irq_rid, 0);
error = bus_setup_intr(dev, sc->irq_res, INTR_TYPE_NET,
edintr, sc, &sc->irq_handle);
if (error) {
printf("setup intr failed %d \n", error);
ed_release_resources(dev);
return (error);
}
Remove an old hack that remembered if the card type was Linksys by storing a flag in the global variable 'linksys' during the probe routine and reading it during the attach routine. We now have the ED_VENDOR_LINKSYS vendor code stored in sc->vendor, so check that instead.
2001-07-22 21:31:44 +00:00
if (sc->vendor != ED_VENDOR_LINKSYS) {
Only try to detect Linksys PCMCIA cards when we are in a pccard environment. This fixes the breakage to ISA ethernet cards. Reviewed by: peter
2000-06-29 07:31:37 +00:00
pccard_get_ether(dev, ether_addr);
for (i = 0, sum = 0; i < ETHER_ADDR_LEN; i++)
sum |= ether_addr[i];
if (sum)
bcopy(ether_addr, sc->arpcom.ac_enaddr, ETHER_ADDR_LEN);
}
Propagate MAC address in CIS tupple to dev/ed/ driver. Many ed-based Ethernet PC-cards can't get correct MAC address without this patch. Submitted by: Takanori Watanabe <takawata@shidahara1.planet.sci.kobe-u.ac.jp> Reviewed by: Warner Losh <imp@village.org>
2000-01-21 03:08:46 +00:00
Replace the if_name and if_unit members of struct ifnet with new members if_xname, if_dname, and if_dunit. if_xname is the name of the interface and if_dname/unit are the driver name and instance. This change paves the way for interface renaming and enhanced pseudo device creation and configuration symantics. Approved By: re (in principle) Reviewed By: njl, imp Tested On: i386, amd64, sparc64 Obtained From: NetBSD (if_xname)
2003-10-31 18:32:15 +00:00
error = ed_attach(dev);
Add an option ED_NO_MIIBUS, which causes the `ed' driver to be built without support for miibus PHYs. Most ed cards don't need miibus support, so it's useful to be able to avoid the bloat of all the mii devices for small fixed-purpose kernels.
2001-09-29 22:32:03 +00:00
#ifndef ED_NO_MIIBUS
Attach the miibus for Linksys/Dlink cards from the attach routine, not from the probe routine. This was an oversight when I originally ported the miibus support to -current, though it was mostly harmless. We now set the vendor code to the new value ED_VENDOR_LINKSYS in ed_pccard_Linksys() at probe time. Then ed_pccard_attach() checks the vendor code, and sets up the miibus if appropriate. Reviewed by: imp
2001-07-22 18:58:44 +00:00
if (error == 0 && sc->vendor == ED_VENDOR_LINKSYS) {
/* Probe for an MII bus, but ignore errors. */
ed_pccard_dlink_mii_reset(sc);
sc->mii_readbits = ed_pccard_dlink_mii_readbits;
sc->mii_writebits = ed_pccard_dlink_mii_writebits;
mii_phy_probe(dev, &sc->miibus, ed_ifmedia_upd,
ed_ifmedia_sts);
}
Add an option ED_NO_MIIBUS, which causes the `ed' driver to be built without support for miibus PHYs. Most ed cards don't need miibus support, so it's useful to be able to avoid the bloat of all the mii devices for small fixed-purpose kernels.
2001-09-29 22:32:03 +00:00
#endif
Attach the miibus for Linksys/Dlink cards from the attach routine, not from the probe routine. This was an oversight when I originally ported the miibus support to -current, though it was mostly harmless. We now set the vendor code to the new value ED_VENDOR_LINKSYS in ed_pccard_Linksys() at probe time. Then ed_pccard_attach() checks the vendor code, and sets up the miibus if appropriate. Reviewed by: imp
2001-07-22 18:58:44 +00:00
Have only one attach routine, rather than two. Remove debug printfs. This should allow the ed driver to work when memory mapping works again.
1999-10-25 06:15:45 +00:00
return (error);
- Fixed the conversion to bus_space interface. - Added PC-98 Cbus devices support. The original patch is submitted by chi@bd.mbn.or.jp (Chiharu Shibata) - Removed old ed driver.
2000-08-17 12:15:45 +00:00
}
static void
ax88190_geteprom(struct ed_softc *sc)
{
int prom[16],i;
u_char tmp;
struct {
unsigned char offset, value;
} pg_seq[] = {
Minor newbus/style(9) cleanups. o Move the ax88190 code to its own function. o Move all device_method_t, driver_t and DRIVER_MODULE definitions to the end of files. o Wrap a few lines > 80 characters. o Use the same devclass for all ed drivers. This allows machines with multiple types of cards to have their cards numbered correctly. Before, you could wind up with two ed0's. o Protect if_edvar.h from multiple includes because I was there.
2000-12-26 06:38:04 +00:00
{ED_P0_CR, ED_CR_RD2|ED_CR_STP},/* Select Page0 */
- Fixed the conversion to bus_space interface. - Added PC-98 Cbus devices support. The original patch is submitted by chi@bd.mbn.or.jp (Chiharu Shibata) - Removed old ed driver.
2000-08-17 12:15:45 +00:00
{ED_P0_DCR, 0x01},
Minor newbus/style(9) cleanups. o Move the ax88190 code to its own function. o Move all device_method_t, driver_t and DRIVER_MODULE definitions to the end of files. o Wrap a few lines > 80 characters. o Use the same devclass for all ed drivers. This allows machines with multiple types of cards to have their cards numbered correctly. Before, you could wind up with two ed0's. o Protect if_edvar.h from multiple includes because I was there.
2000-12-26 06:38:04 +00:00
{ED_P0_RBCR0, 0x00}, /* Clear the count regs. */
- Fixed the conversion to bus_space interface. - Added PC-98 Cbus devices support. The original patch is submitted by chi@bd.mbn.or.jp (Chiharu Shibata) - Removed old ed driver.
2000-08-17 12:15:45 +00:00
{ED_P0_RBCR1, 0x00},
Minor newbus/style(9) cleanups. o Move the ax88190 code to its own function. o Move all device_method_t, driver_t and DRIVER_MODULE definitions to the end of files. o Wrap a few lines > 80 characters. o Use the same devclass for all ed drivers. This allows machines with multiple types of cards to have their cards numbered correctly. Before, you could wind up with two ed0's. o Protect if_edvar.h from multiple includes because I was there.
2000-12-26 06:38:04 +00:00
{ED_P0_IMR, 0x00}, /* Mask completion irq. */
- Fixed the conversion to bus_space interface. - Added PC-98 Cbus devices support. The original patch is submitted by chi@bd.mbn.or.jp (Chiharu Shibata) - Removed old ed driver.
2000-08-17 12:15:45 +00:00
{ED_P0_ISR, 0xff},
Minor newbus/style(9) cleanups. o Move the ax88190 code to its own function. o Move all device_method_t, driver_t and DRIVER_MODULE definitions to the end of files. o Wrap a few lines > 80 characters. o Use the same devclass for all ed drivers. This allows machines with multiple types of cards to have their cards numbered correctly. Before, you could wind up with two ed0's. o Protect if_edvar.h from multiple includes because I was there.
2000-12-26 06:38:04 +00:00
{ED_P0_RCR, ED_RCR_MON | ED_RCR_INTT}, /* Set To Monitor */
- Fixed the conversion to bus_space interface. - Added PC-98 Cbus devices support. The original patch is submitted by chi@bd.mbn.or.jp (Chiharu Shibata) - Removed old ed driver.
2000-08-17 12:15:45 +00:00
{ED_P0_TCR, ED_TCR_LB0}, /* loopback mode. */
{ED_P0_RBCR0, 32},
{ED_P0_RBCR1, 0x00},
{ED_P0_RSAR0, 0x00},
{ED_P0_RSAR1, 0x04},
{ED_P0_CR ,ED_CR_RD0 | ED_CR_STA},
};
/* Reset Card */
tmp = ed_asic_inb(sc, ED_NOVELL_RESET);
ed_asic_outb(sc, ED_NOVELL_RESET, tmp);
DELAY(5000);
ed_asic_outb(sc, ED_P0_CR, ED_CR_RD2 | ED_CR_STP);
DELAY(5000);
/* Card Settings */
for (i = 0; i < sizeof(pg_seq) / sizeof(pg_seq[0]); i++)
ed_nic_outb(sc, pg_seq[i].offset, pg_seq[i].value);
/* Get Data */
for (i = 0; i < 16; i++)
prom[i] = ed_asic_inb(sc, 0);
sc->arpcom.ac_enaddr[0] = prom[0] & 0xff;
sc->arpcom.ac_enaddr[1] = prom[0] >> 8;
sc->arpcom.ac_enaddr[2] = prom[1] & 0xff;
sc->arpcom.ac_enaddr[3] = prom[1] >> 8;
sc->arpcom.ac_enaddr[4] = prom[2] & 0xff;
sc->arpcom.ac_enaddr[5] = prom[2] >> 8;
}
MFPAO: Add support for AX88190, equipped in MELCO LPC3-TX.
2000-07-23 15:15:43 +00:00
static int
ed_pccard_memwrite(device_t dev, off_t offset, u_char byte)
{
Allocate memory resource to access the attribute memory of a card. Reviewed by: imp
2000-09-25 06:01:42 +00:00
int cis_rid;
struct resource *cis;
MFPAO: Add support for AX88190, equipped in MELCO LPC3-TX.
2000-07-23 15:15:43 +00:00
Allocate memory resource to access the attribute memory of a card. Reviewed by: imp
2000-09-25 06:01:42 +00:00
cis_rid = 0;
Minor newbus/style(9) cleanups. o Move the ax88190 code to its own function. o Move all device_method_t, driver_t and DRIVER_MODULE definitions to the end of files. o Wrap a few lines > 80 characters. o Use the same devclass for all ed drivers. This allows machines with multiple types of cards to have their cards numbered correctly. Before, you could wind up with two ed0's. o Protect if_edvar.h from multiple includes because I was there.
2000-12-26 06:38:04 +00:00
cis = bus_alloc_resource(dev, SYS_RES_MEMORY, &cis_rid, 0, ~0,
4 << 10, RF_ACTIVE | RF_SHAREABLE);
Allocate memory resource to access the attribute memory of a card. Reviewed by: imp
2000-09-25 06:01:42 +00:00
if (cis == NULL)
return (ENXIO);
Minor newbus/style(9) cleanups. o Move the ax88190 code to its own function. o Move all device_method_t, driver_t and DRIVER_MODULE definitions to the end of files. o Wrap a few lines > 80 characters. o Use the same devclass for all ed drivers. This allows machines with multiple types of cards to have their cards numbered correctly. Before, you could wind up with two ed0's. o Protect if_edvar.h from multiple includes because I was there.
2000-12-26 06:38:04 +00:00
CARD_SET_RES_FLAGS(device_get_parent(dev), dev, SYS_RES_MEMORY,
cis_rid, PCCARD_A_MEM_ATTR);
Allocate memory resource to access the attribute memory of a card. Reviewed by: imp
2000-09-25 06:01:42 +00:00
Minor newbus/style(9) cleanups. o Move the ax88190 code to its own function. o Move all device_method_t, driver_t and DRIVER_MODULE definitions to the end of files. o Wrap a few lines > 80 characters. o Use the same devclass for all ed drivers. This allows machines with multiple types of cards to have their cards numbered correctly. Before, you could wind up with two ed0's. o Protect if_edvar.h from multiple includes because I was there.
2000-12-26 06:38:04 +00:00
bus_space_write_1(rman_get_bustag(cis), rman_get_bushandle(cis),
offset, byte);
Allocate memory resource to access the attribute memory of a card. Reviewed by: imp
2000-09-25 06:01:42 +00:00
Update the ed driver to probe and attach under a NEWCARD kernel (I was using a cardbus based system with pccbb providing the pcic interface). Something isn't quite right.. when the driver allocates and activates its resources, the IO space that was requested reads as all zeros (versus the original 0xff's as it normally is when there is no device responding). Also, deactivate the resources before releasing them. OLDCARD doesn't seem to care but NEWCARD/CARDBUS get rather unhappy if you release a resource that hasn't been deactivated yet. Make pcic_p.c only compile with oldcard kernels.
2000-11-25 03:36:09 +00:00
bus_deactivate_resource(dev, SYS_RES_MEMORY, cis_rid, cis);
Allocate memory resource to access the attribute memory of a card. Reviewed by: imp
2000-09-25 06:01:42 +00:00
bus_release_resource(dev, SYS_RES_MEMORY, cis_rid, cis);
return (0);
MFPAO: Add support for AX88190, equipped in MELCO LPC3-TX.
2000-07-23 15:15:43 +00:00
}
Linksys Fast Ethernet PCCARD cards supported by the ed driver now require the addition of flag 0x80000 to their config line in pccard.conf(5). This flag is not optional. These Linksys cards will not be recognized without it. Reviewed by: imp, iwasaki
2000-12-18 15:28:53 +00:00
/*
* Probe the Ethernet MAC addrees for PCMCIA Linksys EtherFast 10/100
* and compatible cards (DL10019C Ethernet controller).
*
* Note: The PAO patches try to use more memory for the card, but that
* seems to fail for my card. A future optimization would add this back
* conditionally.
*/
static int
ed_pccard_Linksys(device_t dev)
{
struct ed_softc *sc = device_get_softc(dev);
u_char sum;
int i;
/*
* Linksys registers(offset from ASIC base)
*
* 0x04-0x09 : Physical Address Register 0-5 (PAR0-PAR5)
* 0x0A : Card ID Register (CIR)
* 0x0B : Check Sum Register (SR)
*/
for (sum = 0, i = 0x04; i < 0x0c; i++)
sum += ed_asic_inb(sc, i);
if (sum != 0xff)
return (0); /* invalid DL10019C */
for (i = 0; i < ETHER_ADDR_LEN; i++) {
sc->arpcom.ac_enaddr[i] = ed_asic_inb(sc, 0x04 + i);
}
ed_nic_outb(sc, ED_P0_DCR, ED_DCR_WTS | ED_DCR_FT1 | ED_DCR_LS);
sc->isa16bit = 1;
Attach the miibus for Linksys/Dlink cards from the attach routine, not from the probe routine. This was an oversight when I originally ported the miibus support to -current, though it was mostly harmless. We now set the vendor code to the new value ED_VENDOR_LINKSYS in ed_pccard_Linksys() at probe time. Then ed_pccard_attach() checks the vendor code, and sets up the miibus if appropriate. Reviewed by: imp
2001-07-22 18:58:44 +00:00
sc->vendor = ED_VENDOR_LINKSYS;
Linksys Fast Ethernet PCCARD cards supported by the ed driver now require the addition of flag 0x80000 to their config line in pccard.conf(5). This flag is not optional. These Linksys cards will not be recognized without it. Reviewed by: imp, iwasaki
2000-12-18 15:28:53 +00:00
sc->type = ED_TYPE_NE2000;
sc->type_str = "Linksys";
Add support for Dlink DL10022 to the ed driver. This is a mii part bolted to a ne-2000 chip. This is necessary for the NetGear FA-410TX and other cards. This also requires you add mii to your kernel if you have an ed driver configured. This code will result in a couple of timeout messages for ed on the impacted cards. Additional work will be needed, but this does work right now, and many people need these cards. Submitted by: Ian Dowse <iedowse@maths.tcd.ie>
2001-03-03 08:31:37 +00:00
Linksys Fast Ethernet PCCARD cards supported by the ed driver now require the addition of flag 0x80000 to their config line in pccard.conf(5). This flag is not optional. These Linksys cards will not be recognized without it. Reviewed by: imp, iwasaki
2000-12-18 15:28:53 +00:00
return (1);
}
Minor newbus/style(9) cleanups. o Move the ax88190 code to its own function. o Move all device_method_t, driver_t and DRIVER_MODULE definitions to the end of files. o Wrap a few lines > 80 characters. o Use the same devclass for all ed drivers. This allows machines with multiple types of cards to have their cards numbered correctly. Before, you could wind up with two ed0's. o Protect if_edvar.h from multiple includes because I was there.
2000-12-26 06:38:04 +00:00
/*
* Special setup for AX88190
*/
static int
ed_pccard_ax88190(device_t dev)
{
int error;
int flags = device_get_flags(dev);
int iobase;
struct ed_softc *sc = device_get_softc(dev);
/* Allocate the port resource during setup. */
error = ed_alloc_port(dev, 0, ED_NOVELL_IO_PORTS);
if (error)
return (error);
sc->asic_offset = ED_NOVELL_ASIC_OFFSET;
sc->nic_offset = ED_NOVELL_NIC_OFFSET;
sc->chip_type = ED_CHIP_TYPE_AX88190;
/*
EXPsys PCMCIA Ethernet Combo Submitted by: Ludwig Pummer
2003-09-20 18:03:07 +00:00
* Set Attribute Memory IOBASE Register. Is this a deficiency in
* the PC Card layer, or an ax88190 specific issue? xxx
Minor newbus/style(9) cleanups. o Move the ax88190 code to its own function. o Move all device_method_t, driver_t and DRIVER_MODULE definitions to the end of files. o Wrap a few lines > 80 characters. o Use the same devclass for all ed drivers. This allows machines with multiple types of cards to have their cards numbered correctly. Before, you could wind up with two ed0's. o Protect if_edvar.h from multiple includes because I was there.
2000-12-26 06:38:04 +00:00
*/
iobase = rman_get_start(sc->port_res);
ed_pccard_memwrite(dev, ED_AX88190_IOBASE0, iobase & 0xff);
ed_pccard_memwrite(dev, ED_AX88190_IOBASE1, (iobase >> 8) & 0xff);
ax88190_geteprom(sc);
ed_release_resources(dev);
error = ed_probe_Novell(dev, 0, flags);
integrate from p4 repo: o Honor NE2000DVF_{AX88190,DL10019} flags by setting the ED_FLAGS_xxxx flag. o Mark linksys combo_ecard as ax88190 o Set the type_str to AX88190 for the ax88190 cards. This fixes ax88190 based cards, for the most part, but doesn't seem to fix the mii based dl10019 cards (aka linksys cards).
2002-07-22 21:28:46 +00:00
if (error == 0) {
sc->vendor = ED_VENDOR_PCCARD;
sc->type = ED_TYPE_NE2000;
sc->type_str = "AX88190";
}
Minor newbus/style(9) cleanups. o Move the ax88190 code to its own function. o Move all device_method_t, driver_t and DRIVER_MODULE definitions to the end of files. o Wrap a few lines > 80 characters. o Use the same devclass for all ed drivers. This allows machines with multiple types of cards to have their cards numbered correctly. Before, you could wind up with two ed0's. o Protect if_edvar.h from multiple includes because I was there.
2000-12-26 06:38:04 +00:00
return (error);
}
Add an option ED_NO_MIIBUS, which causes the `ed' driver to be built without support for miibus PHYs. Most ed cards don't need miibus support, so it's useful to be able to avoid the bloat of all the mii devices for small fixed-purpose kernels.
2001-09-29 22:32:03 +00:00
#ifndef ED_NO_MIIBUS
Add support for Dlink DL10022 to the ed driver. This is a mii part bolted to a ne-2000 chip. This is necessary for the NetGear FA-410TX and other cards. This also requires you add mii to your kernel if you have an ed driver configured. This code will result in a couple of timeout messages for ed on the impacted cards. Additional work will be needed, but this does work right now, and many people need these cards. Submitted by: Ian Dowse <iedowse@maths.tcd.ie>
2001-03-03 08:31:37 +00:00
/* MII bit-twiddling routines for cards using Dlink chipset */
#define DLINK_MIISET(sc, x) ed_asic_outb(sc, ED_DLINK_MIIBUS, \
ed_asic_inb(sc, ED_DLINK_MIIBUS) | (x))
#define DLINK_MIICLR(sc, x) ed_asic_outb(sc, ED_DLINK_MIIBUS, \
ed_asic_inb(sc, ED_DLINK_MIIBUS) & ~(x))
static void
ed_pccard_dlink_mii_reset(sc)
struct ed_softc *sc;
{
ed_asic_outb(sc, ED_DLINK_MIIBUS, 0);
DELAY(10);
DLINK_MIISET(sc, ED_DLINK_MII_RESET2);
DELAY(10);
DLINK_MIISET(sc, ED_DLINK_MII_RESET1);
DELAY(10);
DLINK_MIICLR(sc, ED_DLINK_MII_RESET1);
DELAY(10);
DLINK_MIICLR(sc, ED_DLINK_MII_RESET2);
DELAY(10);
}
static void
ed_pccard_dlink_mii_writebits(sc, val, nbits)
struct ed_softc *sc;
u_int val;
int nbits;
{
int i;
DLINK_MIISET(sc, ED_DLINK_MII_DIROUT);
for (i = nbits - 1; i >= 0; i--) {
if ((val >> i) & 1)
DLINK_MIISET(sc, ED_DLINK_MII_DATAOUT);
else
DLINK_MIICLR(sc, ED_DLINK_MII_DATAOUT);
DELAY(10);
DLINK_MIISET(sc, ED_DLINK_MII_CLK);
DELAY(10);
DLINK_MIICLR(sc, ED_DLINK_MII_CLK);
DELAY(10);
}
}
static u_int
ed_pccard_dlink_mii_readbits(sc, nbits)
struct ed_softc *sc;
int nbits;
{
int i;
u_int val = 0;
DLINK_MIICLR(sc, ED_DLINK_MII_DIROUT);
for (i = nbits - 1; i >= 0; i--) {
DLINK_MIISET(sc, ED_DLINK_MII_CLK);
DELAY(10);
val <<= 1;
if (ed_asic_inb(sc, ED_DLINK_MIIBUS) & ED_DLINK_MII_DATATIN)
val++;
DLINK_MIICLR(sc, ED_DLINK_MII_CLK);
DELAY(10);
}
return val;
}
Add an option ED_NO_MIIBUS, which causes the `ed' driver to be built without support for miibus PHYs. Most ed cards don't need miibus support, so it's useful to be able to avoid the bloat of all the mii devices for small fixed-purpose kernels.
2001-09-29 22:32:03 +00:00
#endif
Add support for Dlink DL10022 to the ed driver. This is a mii part bolted to a ne-2000 chip. This is necessary for the NetGear FA-410TX and other cards. This also requires you add mii to your kernel if you have an ed driver configured. This code will result in a couple of timeout messages for ed on the impacted cards. Additional work will be needed, but this does work right now, and many people need these cards. Submitted by: Ian Dowse <iedowse@maths.tcd.ie>
2001-03-03 08:31:37 +00:00
Minor newbus/style(9) cleanups. o Move the ax88190 code to its own function. o Move all device_method_t, driver_t and DRIVER_MODULE definitions to the end of files. o Wrap a few lines > 80 characters. o Use the same devclass for all ed drivers. This allows machines with multiple types of cards to have their cards numbered correctly. Before, you could wind up with two ed0's. o Protect if_edvar.h from multiple includes because I was there.
2000-12-26 06:38:04 +00:00
static device_method_t ed_pccard_methods[] = {
/* Device interface */
DEVMETHOD(device_probe, pccard_compat_probe),
DEVMETHOD(device_attach, pccard_compat_attach),
DEVMETHOD(device_detach, ed_pccard_detach),
Add an option ED_NO_MIIBUS, which causes the `ed' driver to be built without support for miibus PHYs. Most ed cards don't need miibus support, so it's useful to be able to avoid the bloat of all the mii devices for small fixed-purpose kernels.
2001-09-29 22:32:03 +00:00
#ifndef ED_NO_MIIBUS
Add support for Dlink DL10022 to the ed driver. This is a mii part bolted to a ne-2000 chip. This is necessary for the NetGear FA-410TX and other cards. This also requires you add mii to your kernel if you have an ed driver configured. This code will result in a couple of timeout messages for ed on the impacted cards. Additional work will be needed, but this does work right now, and many people need these cards. Submitted by: Ian Dowse <iedowse@maths.tcd.ie>
2001-03-03 08:31:37 +00:00
/* Bus interface */
DEVMETHOD(bus_child_detached, ed_child_detached),
/* MII interface */
DEVMETHOD(miibus_readreg, ed_miibus_readreg),
DEVMETHOD(miibus_writereg, ed_miibus_writereg),
Add an option ED_NO_MIIBUS, which causes the `ed' driver to be built without support for miibus PHYs. Most ed cards don't need miibus support, so it's useful to be able to avoid the bloat of all the mii devices for small fixed-purpose kernels.
2001-09-29 22:32:03 +00:00
#endif
Add support for Dlink DL10022 to the ed driver. This is a mii part bolted to a ne-2000 chip. This is necessary for the NetGear FA-410TX and other cards. This also requires you add mii to your kernel if you have an ed driver configured. This code will result in a couple of timeout messages for ed on the impacted cards. Additional work will be needed, but this does work right now, and many people need these cards. Submitted by: Ian Dowse <iedowse@maths.tcd.ie>
2001-03-03 08:31:37 +00:00
Minor newbus/style(9) cleanups. o Move the ax88190 code to its own function. o Move all device_method_t, driver_t and DRIVER_MODULE definitions to the end of files. o Wrap a few lines > 80 characters. o Use the same devclass for all ed drivers. This allows machines with multiple types of cards to have their cards numbered correctly. Before, you could wind up with two ed0's. o Protect if_edvar.h from multiple includes because I was there.
2000-12-26 06:38:04 +00:00
/* Card interface */
DEVMETHOD(card_compat_match, ed_pccard_match),
DEVMETHOD(card_compat_probe, ed_pccard_probe),
DEVMETHOD(card_compat_attach, ed_pccard_attach),
{ 0, 0 }
};
static driver_t ed_pccard_driver = {
"ed",
ed_pccard_methods,
sizeof(struct ed_softc)
};
- Express hard dependencies on bus (pci, isa, pccard) and network layer (ether). - Don't abuse module names to facilitate ifconfig module loading; such abuse isn't really needed. (And if we do need type information associated with a module then we should make it explicit and not use hacks.)
2003-04-15 06:37:30 +00:00
DRIVER_MODULE(ed, pccard, ed_pccard_driver, ed_devclass, 0, 0);
Add an option ED_NO_MIIBUS, which causes the `ed' driver to be built without support for miibus PHYs. Most ed cards don't need miibus support, so it's useful to be able to avoid the bloat of all the mii devices for small fixed-purpose kernels.
2001-09-29 22:32:03 +00:00
#ifndef ED_NO_MIIBUS
Add support for Dlink DL10022 to the ed driver. This is a mii part bolted to a ne-2000 chip. This is necessary for the NetGear FA-410TX and other cards. This also requires you add mii to your kernel if you have an ed driver configured. This code will result in a couple of timeout messages for ed on the impacted cards. Additional work will be needed, but this does work right now, and many people need these cards. Submitted by: Ian Dowse <iedowse@maths.tcd.ie>
2001-03-03 08:31:37 +00:00
DRIVER_MODULE(miibus, ed, miibus_driver, miibus_devclass, 0, 0);
Add an option ED_NO_MIIBUS, which causes the `ed' driver to be built without support for miibus PHYs. Most ed cards don't need miibus support, so it's useful to be able to avoid the bloat of all the mii devices for small fixed-purpose kernels.
2001-09-29 22:32:03 +00:00
#endif
Reference in New Issue Copy Permalink