freebsd-dev/sys/dev/pccbb/pccbb.c

1938 lines
56 KiB
C
Raw Normal View History

/*
* Copyright (c) 2000,2001 Jonathan Chen.
* 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,
* without modification, immediately at the beginning of the file.
* 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.
*
* $FreeBSD$
*/
/*
* Driver for PCI to Cardbus Bridge chips
*
* References:
* TI Datasheets:
* http://www-s.ti.com/cgi-bin/sc/generic2.cgi?family=PCI+CARDBUS+CONTROLLERS
* Much of the 16-bit PC Card compatibility code stolen from dev/pcic/i82365.c
* XXX and should be cleaned up to share as much as possible.
*
* Written by Jonathan Chen <jon@freebsd.org>
* The author would like to acknowledge:
* * HAYAKAWA Koichi: Author of the NetBSD code for the same thing
* * Warner Losh: Newbus/newcard guru and author of the pccard side of things
* * YAMAMOTO Shigeru: Author of another FreeBSD cardbus driver
* * David Cross: Author of the initial ugly hack for a specific cardbus card
*/
#define CBB_DEBUG
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/errno.h>
#include <sys/kernel.h>
#include <sys/kthread.h>
#include <sys/lock.h>
#include <sys/malloc.h>
#include <sys/mutex.h>
#include <sys/bus.h>
#include <machine/bus.h>
#include <sys/rman.h>
#include <machine/resource.h>
#include <pci/pcireg.h>
#include <pci/pcivar.h>
#include <machine/clock.h>
#include <dev/pccard/pccardreg.h>
#include <dev/pccard/pccardvar.h>
#include <dev/pcic/i82365reg.h>
#include <dev/pccbb/pccbbreg.h>
#include <dev/pccbb/pccbbvar.h>
#include "power_if.h"
#include "card_if.h"
#include "pcib_if.h"
#define CBB_DEBUG
#if defined CBB_DEBUG
#define DPRINTF(x) printf x
#define DEVPRINTF(x) device_printf x
#else
#define DPRINTF(x)
#define DEVPRINTF(x)
#endif
#define PCI_MASK_CONFIG(DEV,REG,MASK,SIZE) \
pci_write_config(DEV, REG, pci_read_config(DEV, REG, SIZE) MASK, SIZE)
#define PCI_MASK2_CONFIG(DEV,REG,MASK1,MASK2,SIZE) \
pci_write_config(DEV, REG, ( \
pci_read_config(DEV, REG, SIZE) MASK1) MASK2, SIZE)
/*
* XXX all the pcic code really doesn't belong here and needs to be
* XXX migrated to its own file, shared with the 16-bit code
*/
#define PCIC_READ(SC,REG) \
(((u_int8_t*)((SC)->sc_socketreg))[0x800+(REG)])
#define PCIC_WRITE(SC,REG,val) \
(((u_int8_t*)((SC)->sc_socketreg))[0x800+(REG)]) = (val)
#define PCIC_MASK(SC,REG,MASK) \
PCIC_WRITE(SC,REG,PCIC_READ(SC,REG) MASK)
#define PCIC_MASK2(SC,REG,MASK,MASK2) \
PCIC_WRITE(SC,REG,(PCIC_READ(SC,REG) MASK) MASK2)
struct pccbb_sclist {
struct pccbb_softc *sc;
STAILQ_ENTRY(pccbb_sclist) entries;
};
static STAILQ_HEAD(, pccbb_sclist) softcs;
static int softcs_init = 0;
struct yenta_chipinfo {
u_int32_t yc_id;
const char *yc_name;
int yc_chiptype;
int yc_flags;
} yc_chipsets[] = {
/* Texas Instruments chips */
{PCI_DEVICE_ID_PCIC_TI1130, "TI1130 PCI-CardBus Bridge", CB_TI113X,
PCCBB_PCIC_IO_RELOC | PCCBB_PCIC_MEM_32},
{PCI_DEVICE_ID_PCIC_TI1131, "TI1131 PCI-CardBus Bridge", CB_TI113X,
PCCBB_PCIC_IO_RELOC | PCCBB_PCIC_MEM_32},
{PCI_DEVICE_ID_PCIC_TI1211, "TI1211 PCI-CardBus Bridge", CB_TI12XX,
PCCBB_PCIC_IO_RELOC | PCCBB_PCIC_MEM_32},
{PCI_DEVICE_ID_PCIC_TI1220, "TI1220 PCI-CardBus Bridge", CB_TI12XX,
PCCBB_PCIC_IO_RELOC | PCCBB_PCIC_MEM_32},
{PCI_DEVICE_ID_PCIC_TI1221, "TI1221 PCI-CardBus Bridge", CB_TI12XX,
PCCBB_PCIC_IO_RELOC | PCCBB_PCIC_MEM_32},
{PCI_DEVICE_ID_PCIC_TI1225, "TI1225 PCI-CardBus Bridge", CB_TI12XX,
PCCBB_PCIC_IO_RELOC | PCCBB_PCIC_MEM_32},
{PCI_DEVICE_ID_PCIC_TI1250, "TI1250 PCI-CardBus Bridge", CB_TI12XX,
PCCBB_PCIC_IO_RELOC | PCCBB_PCIC_MEM_32},
{PCI_DEVICE_ID_PCIC_TI1251, "TI1251 PCI-CardBus Bridge", CB_TI12XX,
PCCBB_PCIC_IO_RELOC | PCCBB_PCIC_MEM_32},
{PCI_DEVICE_ID_PCIC_TI1251B,"TI1251B PCI-CardBus Bridge",CB_TI12XX,
PCCBB_PCIC_IO_RELOC | PCCBB_PCIC_MEM_32},
{PCI_DEVICE_ID_PCIC_TI1410, "TI1410 PCI-CardBus Bridge", CB_TI12XX,
PCCBB_PCIC_IO_RELOC | PCCBB_PCIC_MEM_32},
{PCI_DEVICE_ID_PCIC_TI1420, "TI1420 PCI-CardBus Bridge", CB_TI12XX,
PCCBB_PCIC_IO_RELOC | PCCBB_PCIC_MEM_32},
{PCI_DEVICE_ID_PCIC_TI1450, "TI1450 PCI-CardBus Bridge", CB_TI12XX,
PCCBB_PCIC_IO_RELOC | PCCBB_PCIC_MEM_32},
{PCI_DEVICE_ID_PCIC_TI1451, "TI1451 PCI-CardBus Bridge", CB_TI12XX,
PCCBB_PCIC_IO_RELOC | PCCBB_PCIC_MEM_32},
{PCI_DEVICE_ID_PCIC_TI4451, "TI4451 PCI-CardBus Bridge", CB_TI12XX,
PCCBB_PCIC_IO_RELOC | PCCBB_PCIC_MEM_32},
/* Ricoh chips */
{PCI_DEVICE_ID_RICOH_RL5C465, "RF5C465 PCI-CardBus Bridge",
CB_RF5C46X, PCCBB_PCIC_MEM_32},
{PCI_DEVICE_ID_RICOH_RL5C466, "RF5C466 PCI-CardBus Bridge",
CB_RF5C46X, PCCBB_PCIC_MEM_32},
{PCI_DEVICE_ID_RICOH_RL5C475, "RF5C475 PCI-CardBus Bridge",
CB_RF5C47X, PCCBB_PCIC_MEM_32},
{PCI_DEVICE_ID_RICOH_RL5C476, "RF5C476 PCI-CardBus Bridge",
CB_RF5C47X, PCCBB_PCIC_MEM_32},
{PCI_DEVICE_ID_RICOH_RL5C478, "RF5C478 PCI-CardBus Bridge",
CB_RF5C47X, PCCBB_PCIC_MEM_32},
/* Toshiba products */
{PCI_DEVICE_ID_TOSHIBA_TOPIC95, "ToPIC95 PCI-CardBus Bridge",
CB_TOPIC95, PCCBB_PCIC_MEM_32},
{PCI_DEVICE_ID_TOSHIBA_TOPIC95B, "ToPIC95B PCI-CardBus Bridge",
CB_TOPIC95B, PCCBB_PCIC_MEM_32},
{PCI_DEVICE_ID_TOSHIBA_TOPIC97, "ToPIC97 PCI-CardBus Bridge",
CB_TOPIC97, PCCBB_PCIC_MEM_32},
{PCI_DEVICE_ID_TOSHIBA_TOPIC100, "ToPIC100 PCI-CardBus Bridge",
CB_TOPIC97, PCCBB_PCIC_MEM_32},
/* Cirrus Logic */
{PCI_DEVICE_ID_PCIC_CLPD6832, "CLPD6832 PCI-CardBus Bridge",
CB_CIRRUS, PCCBB_PCIC_MEM_32},
{PCI_DEVICE_ID_PCIC_CLPD6833, "CLPD6833 PCI-CardBus Bridge",
CB_CIRRUS, PCCBB_PCIC_MEM_32},
/* 02Micro */
{PCI_DEVICE_ID_PCIC_OZ6832, "O2Mirco OZ6832/6833 PCI-CardBus Bridge",
CB_CIRRUS, PCCBB_PCIC_MEM_32},
{PCI_DEVICE_ID_PCIC_OZ6860, "O2Mirco OZ6836/6860 PCI-CardBus Bridge",
CB_CIRRUS, PCCBB_PCIC_MEM_32},
{PCI_DEVICE_ID_PCIC_OZ6872, "O2Mirco OZ6812/6872 PCI-CardBus Bridge",
CB_CIRRUS, PCCBB_PCIC_MEM_32},
/* sentinel */
{0 /* null id */, "unknown",
CB_UNKNOWN, 0},
};
static int cb_chipset(u_int32_t pci_id, const char** namep, int* flagp);
static int pccbb_probe(device_t dev);
static void pccbb_chipinit(struct pccbb_softc* sc);
static int pccbb_attach(device_t dev);
static int pccbb_detach(device_t dev);
static void pccbb_driver_added(device_t dev, driver_t *driver);
static void pccbb_child_detached(device_t dev, device_t child);
static void pccbb_event_thread (void *arg);
static void pccbb_create_event_thread (struct pccbb_softc *sc);
static void pccbb_start_threads(void *arg);
static void pccbb_insert (struct pccbb_softc *sc);
static void pccbb_removal (struct pccbb_softc *sc);
static void pccbb_intr(void* arg);
static int pccbb_detect_voltage(device_t dev);
static int pccbb_power(device_t dev, int volts);
static void pccbb_cardbus_reset(device_t dev);
static int pccbb_cardbus_power_enable_socket(device_t self, device_t child);
static void pccbb_cardbus_power_disable_socket(device_t self, device_t child);
static int pccbb_cardbus_io_open(device_t dev, int win,
u_int32_t start, u_int32_t end);
static int pccbb_cardbus_mem_open(device_t dev, int win,
u_int32_t start, u_int32_t end);
static void pccbb_cardbus_auto_open(struct pccbb_softc *sc, int type);
static int pccbb_cardbus_activate_resource(device_t self, device_t child,
int type, int rid,
struct resource *r);
static int pccbb_cardbus_deactivate_resource(device_t self, device_t child,
int type, int rid,
struct resource *r);
static struct resource* pccbb_cardbus_alloc_resource(device_t self,
device_t child, int type, int* rid,
u_long start, u_long end, u_long count,
u_int flags);
static int pccbb_cardbus_release_resource(device_t self, device_t child,
int type,int rid,
struct resource *r);
static int pccbb_pcic_power_enable_socket(device_t self, device_t child);
static void pccbb_pcic_power_disable_socket(device_t self, device_t child);
static void pccbb_pcic_wait_ready(struct pccbb_softc *sc);
static void pccbb_pcic_do_mem_map(struct pccbb_softc *sc, int win);
static int pccbb_pcic_mem_map(struct pccbb_softc *sc, int kind,
struct resource *r, bus_addr_t card_addr,
int *win);
static void pccbb_pcic_mem_unmap(struct pccbb_softc *sc, int window);
static void pccbb_pcic_do_io_map(struct pccbb_softc *sc, int win);
static int pccbb_pcic_io_map(struct pccbb_softc *sc, int width,
struct resource *r, bus_addr_t card_addr,
int *win);
static void pccbb_pcic_io_unmap(struct pccbb_softc *sc, int window);
static int pccbb_pcic_activate_resource(device_t self, device_t child,
int type, int rid, struct resource *r);
static int pccbb_pcic_deactivate_resource(device_t self, device_t child,
int type,int rid, struct resource *r);
static struct resource* pccbb_pcic_alloc_resource(device_t self,device_t child,
int type, int* rid, u_long start,
u_long end, u_long count, u_int flags);
static int pccbb_pcic_release_resource(device_t self, device_t child, int type,
int rid, struct resource *res);
static int pccbb_pcic_set_res_flags(device_t self, device_t child, int type,
int rid, u_int32_t flags);
static int pccbb_pcic_set_memory_offset(device_t self, device_t child, int rid,
u_int32_t offset, u_int32_t *deltap);
static int pccbb_power_enable_socket(device_t self, device_t child);
static void pccbb_power_disable_socket(device_t self, device_t child);
static int pccbb_activate_resource(device_t self, device_t child, int type,
int rid, struct resource *r);
static int pccbb_deactivate_resource(device_t self, device_t child, int type,
int rid, struct resource *r);
static struct resource* pccbb_alloc_resource(device_t self, device_t child,
int type, int* rid, u_long start,
u_long end, u_long count,
u_int flags);
static int pccbb_release_resource(device_t self, device_t child,
int type, int rid, struct resource *r);
static int pccbb_maxslots(device_t dev);
static u_int32_t pccbb_read_config(device_t dev, int b, int s, int f,
int reg, int width);
static void pccbb_write_config(device_t dev, int b, int s, int f, int reg,
u_int32_t val, int width);
/************************************************************************/
/* Probe/Attach */
/************************************************************************/
static int
cb_chipset(u_int32_t pci_id, const char** namep, int* flagp)
{
int loopend = sizeof(yc_chipsets)/sizeof(yc_chipsets[0]);
struct yenta_chipinfo *ycp, *ycend;
ycend = yc_chipsets + loopend;
for (ycp = yc_chipsets; ycp < ycend && pci_id != ycp->yc_id; ++ycp);
if (ycp == ycend) {
/* not found */
ycp = yc_chipsets + loopend - 1; /* to point the sentinel */
}
if (namep != NULL) {
*namep = ycp->yc_name;
}
if (flagp != NULL) {
*flagp = ycp->yc_flags;
}
return ycp->yc_chiptype;
}
static int
pccbb_probe(device_t dev)
{
const char *name;
if (cb_chipset(pci_get_devid(dev), &name, NULL) == CB_UNKNOWN)
return ENXIO;
device_set_desc(dev, name);
return 0;
}
static void
pccbb_chipinit(struct pccbb_softc* sc)
{
/* Set CardBus latency timer */
if (pci_read_config(sc->sc_dev, PCIR_SECLAT_1, 1) < 0x20)
pci_write_config(sc->sc_dev, PCIR_SECLAT_1, 0x20, 1);
/* Set PCI latency timer */
if (pci_read_config(sc->sc_dev, PCIR_LATTIMER, 1) < 0x20)
pci_write_config(sc->sc_dev, PCIR_LATTIMER, 0x20, 1);
/* Enable memory access */
PCI_MASK_CONFIG(sc->sc_dev, PCIR_COMMAND,
| PCIM_CMD_MEMEN
| PCIM_CMD_PORTEN
| PCIM_CMD_BUSMASTEREN, 2);
/* disable Legacy IO */
switch (sc->sc_chipset) {
case CB_RF5C46X:
case CB_RF5C47X:
PCI_MASK_CONFIG(sc->sc_dev, PCCBBR_BRIDGECTRL,
& ~(PCCBBM_BRIDGECTRL_RL_3E0_EN|
PCCBBM_BRIDGECTRL_RL_3E2_EN), 2);
break;
default:
pci_write_config(sc->sc_dev, PCCBBR_LEGACY, 0x0, 4);
break;
}
/* Use PCI interrupt for interrupt routing */
PCI_MASK2_CONFIG(sc->sc_dev, PCCBBR_BRIDGECTRL,
& ~(PCCBBM_BRIDGECTRL_MASTER_ABORT |
PCCBBM_BRIDGECTRL_INTR_IREQ_EN),
| PCCBBM_BRIDGECTRL_WRITE_POST_EN,
2);
switch (sc->sc_chipset) {
case CB_TI113X:
PCI_MASK2_CONFIG(sc->sc_dev, PCCBBR_CBCTRL,
& ~PCCBBM_CBCTRL_113X_PCI_INTR,
| PCCBBM_CBCTRL_113X_PCI_CSC
| PCCBBM_CBCTRL_113X_PCI_IRQ_EN, 1);
PCI_MASK_CONFIG(sc->sc_dev, PCCBBR_DEVCTRL,
& ~(PCCBBM_DEVCTRL_INT_SERIAL|
PCCBBM_DEVCTRL_INT_PCI), 1);
PCIC_WRITE(sc, PCIC_INTR, PCIC_INTR_ENABLE);
PCIC_WRITE(sc, PCIC_CSC_INTR, 0);
break;
case CB_TI12XX:
PCIC_WRITE(sc, PCIC_INTR, PCIC_INTR_ENABLE);
PCIC_WRITE(sc, PCIC_CSC_INTR, 0);
break;
case CB_TOPIC95B:
PCI_MASK_CONFIG(sc->sc_dev, PCCBBR_TOPIC_SOCKETCTRL,
| PCCBBM_TOPIC_SOCKETCTRL_SCR_IRQSEL, 4);
PCI_MASK2_CONFIG(sc->sc_dev, PCCBBR_TOPIC_SLOTCTRL,
| PCCBBM_TOPIC_SLOTCTRL_SLOTON
| PCCBBM_TOPIC_SLOTCTRL_SLOTEN
| PCCBBM_TOPIC_SLOTCTRL_ID_LOCK
| PCCBBM_TOPIC_SLOTCTRL_CARDBUS,
& ~PCCBBM_TOPIC_SLOTCTRL_SWDETECT, 4);
break;
}
/* close all memory and io windows */
pci_write_config(sc->sc_dev, PCCBBR_MEMBASE0, 0xffffffff, 4);
pci_write_config(sc->sc_dev, PCCBBR_MEMLIMIT0, 0, 4);
pci_write_config(sc->sc_dev, PCCBBR_MEMBASE1, 0xffffffff, 4);
pci_write_config(sc->sc_dev, PCCBBR_MEMLIMIT1, 0, 4);
pci_write_config(sc->sc_dev, PCCBBR_IOBASE0, 0xffffffff, 4);
pci_write_config(sc->sc_dev, PCCBBR_IOLIMIT0, 0, 4);
pci_write_config(sc->sc_dev, PCCBBR_IOBASE1, 0xffffffff, 4);
pci_write_config(sc->sc_dev, PCCBBR_IOLIMIT1, 0, 4);
}
static int
pccbb_attach(device_t dev)
{
struct pccbb_softc *sc = (struct pccbb_softc *)device_get_softc(dev);
int rid;
struct pccbb_sclist *sclist;
u_int32_t sockbase;
u_int32_t tmp;
if (!softcs_init) {
softcs_init = 1;
STAILQ_INIT(&softcs);
}
mtx_init(&sc->sc_mtx, device_get_nameunit(dev), MTX_DEF);
sc->sc_chipset = cb_chipset(pci_get_devid(dev), NULL, &sc->sc_flags);
sc->sc_dev = dev;
sc->sc_cbdev = NULL;
sc->sc_pccarddev = NULL;
sc->sc_secbus = pci_read_config(dev, PCIR_SECBUS_2, 1);
sc->sc_subbus = pci_read_config(dev, PCIR_SUBBUS_2, 1);
sc->memalloc = 0;
sc->ioalloc = 0;
SLIST_INIT(&sc->rl);
/* Ths PCI bus should have given me memory... right? */
rid = PCCBBR_SOCKBASE;
sc->sc_base_res=bus_alloc_resource(dev, SYS_RES_MEMORY, &rid,
0, ~0, 1, RF_ACTIVE);
if (sc->sc_base_res == NULL) {
/*
* XXX eVILE HACK BAD THING! XXX
* The pci bus device should do this for us.
* Some BIOSes doesn't assign a memory space properly.
* So we try to manually put one in...
*/
sockbase = pci_read_config(dev, rid, 4);
if (sockbase < 0x100000 || sockbase >= 0xfffffff0) {
pci_write_config(dev, rid, 0xffffffff, 4);
sockbase = pci_read_config(dev, rid, 4);
sockbase = (sockbase & 0xfffffff0) &
-(sockbase & 0xfffffff0);
sc->sc_base_res = bus_alloc_resource(dev,
SYS_RES_MEMORY, &rid, CARDBUS_SYS_RES_MEMORY_START,
CARDBUS_SYS_RES_MEMORY_END, sockbase,
RF_ACTIVE | rman_make_alignment_flags(sockbase));
if (sc->sc_base_res == NULL) {
device_printf(dev,
"Could not grab register memory\n");
mtx_destroy(&sc->sc_mtx);
return ENOMEM;
}
pci_write_config(dev, PCCBBR_SOCKBASE,
rman_get_start(sc->sc_base_res), 4);
DEVPRINTF((dev, "PCI Memory allocated: %08lx\n",
rman_get_start(sc->sc_base_res)));
} else {
device_printf(dev, "Could not map register memory\n");
mtx_destroy(&sc->sc_mtx);
return ENOMEM;
}
}
sc->sc_socketreg =
(struct pccbb_socketreg *) rman_get_virtual(sc->sc_base_res);
pccbb_chipinit(sc);
/* CSC Interrupt: Card detect interrupt on */
sc->sc_socketreg->socket_mask |= PCCBB_SOCKET_MASK_CD;
/* reset interrupt */
tmp = sc->sc_socketreg->socket_event;
sc->sc_socketreg->socket_event = tmp;
/* Map and establish the interrupt. */
rid = 0;
sc->sc_irq_res=bus_alloc_resource(dev, SYS_RES_IRQ, &rid, 0, ~0, 1,
RF_SHAREABLE | RF_ACTIVE);
if (sc->sc_irq_res == NULL) {
printf("pccbb: Unable to map IRQ...\n");
bus_release_resource(dev, SYS_RES_MEMORY, PCCBBR_SOCKBASE,
sc->sc_base_res);
mtx_destroy(&sc->sc_mtx);
return ENOMEM;
}
if (bus_setup_intr(dev, sc->sc_irq_res, INTR_TYPE_BIO, pccbb_intr, sc,
&sc->sc_intrhand)) {
device_printf(dev, "couldn't establish interrupt");
bus_release_resource(dev, SYS_RES_IRQ, 0, sc->sc_irq_res);
bus_release_resource(dev, SYS_RES_MEMORY, PCCBBR_SOCKBASE,
sc->sc_base_res);
mtx_destroy(&sc->sc_mtx);
return ENOMEM;
}
/* attach children */
sc->sc_cbdev = device_add_child(dev, "cardbus", -1);
if (sc->sc_cbdev == NULL) {
DEVPRINTF((dev, "WARNING: cannot add cardbus bus.\n"));
} else if (device_probe_and_attach(sc->sc_cbdev) != 0) {
DEVPRINTF((dev, "WARNING: cannot attach cardbus bus!\n"));
sc->sc_cbdev = NULL;
}
sc->sc_pccarddev = device_add_child(dev, "pccard", -1);
if (sc->sc_pccarddev == NULL) {
DEVPRINTF((dev, "WARNING: cannot add pccard bus.\n"));
} else if (device_probe_and_attach(sc->sc_pccarddev) != 0) {
DEVPRINTF((dev, "WARNING: cannot attach pccard bus.\n"));
sc->sc_pccarddev = NULL;
}
if (sc->sc_cbdev == NULL && sc->sc_pccarddev == NULL) {
device_printf(dev,
"ERROR: Failed to attach cardbus/pccard bus!\n");
bus_teardown_intr(dev, sc->sc_irq_res, sc->sc_intrhand);
bus_release_resource(dev, SYS_RES_IRQ, 0, sc->sc_irq_res);
bus_release_resource(dev, SYS_RES_MEMORY, PCCBBR_SOCKBASE,
sc->sc_base_res);
mtx_destroy(&sc->sc_mtx);
return ENOMEM;
}
sclist = malloc(sizeof(struct pccbb_sclist), M_DEVBUF, M_WAITOK);
sclist->sc = sc;
STAILQ_INSERT_TAIL(&softcs, sclist, entries);
return 0;
}
static int
pccbb_detach(device_t dev)
{
struct pccbb_softc *sc = device_get_softc(dev);
int numdevs;
device_t *devlist;
int tmp;
int error;
device_get_children(dev, &devlist, &numdevs);
error = 0;
for (tmp = 0; tmp < numdevs; tmp++) {
if (device_detach(devlist[tmp]) == 0)
device_delete_child(dev, devlist[tmp]);
else
error++;
}
free(devlist, M_TEMP);
if (error > 0)
return ENXIO;
Change and clean the mutex lock interface. mtx_enter(lock, type) becomes: mtx_lock(lock) for sleep locks (MTX_DEF-initialized locks) mtx_lock_spin(lock) for spin locks (MTX_SPIN-initialized) similarily, for releasing a lock, we now have: mtx_unlock(lock) for MTX_DEF and mtx_unlock_spin(lock) for MTX_SPIN. We change the caller interface for the two different types of locks because the semantics are entirely different for each case, and this makes it explicitly clear and, at the same time, it rids us of the extra `type' argument. The enter->lock and exit->unlock change has been made with the idea that we're "locking data" and not "entering locked code" in mind. Further, remove all additional "flags" previously passed to the lock acquire/release routines with the exception of two: MTX_QUIET and MTX_NOSWITCH The functionality of these flags is preserved and they can be passed to the lock/unlock routines by calling the corresponding wrappers: mtx_{lock, unlock}_flags(lock, flag(s)) and mtx_{lock, unlock}_spin_flags(lock, flag(s)) for MTX_DEF and MTX_SPIN locks, respectively. Re-inline some lock acq/rel code; in the sleep lock case, we only inline the _obtain_lock()s in order to ensure that the inlined code fits into a cache line. In the spin lock case, we inline recursion and actually only perform a function call if we need to spin. This change has been made with the idea that we generally tend to avoid spin locks and that also the spin locks that we do have and are heavily used (i.e. sched_lock) do recurse, and therefore in an effort to reduce function call overhead for some architectures (such as alpha), we inline recursion for this case. Create a new malloc type for the witness code and retire from using the M_DEV type. The new type is called M_WITNESS and is only declared if WITNESS is enabled. Begin cleaning up some machdep/mutex.h code - specifically updated the "optimized" inlined code in alpha/mutex.h and wrote MTX_LOCK_SPIN and MTX_UNLOCK_SPIN asm macros for the i386/mutex.h as we presently need those. Finally, caught up to the interface changes in all sys code. Contributors: jake, jhb, jasone (in no particular order)
2001-02-09 06:11:45 +00:00
mtx_lock(&sc->sc_mtx);
bus_teardown_intr(dev, sc->sc_irq_res, sc->sc_intrhand);
sc->sc_flags |= PCCBB_KTHREAD_DONE;
if (sc->sc_flags & PCCBB_KTHREAD_RUNNING) {
wakeup(sc);
Change and clean the mutex lock interface. mtx_enter(lock, type) becomes: mtx_lock(lock) for sleep locks (MTX_DEF-initialized locks) mtx_lock_spin(lock) for spin locks (MTX_SPIN-initialized) similarily, for releasing a lock, we now have: mtx_unlock(lock) for MTX_DEF and mtx_unlock_spin(lock) for MTX_SPIN. We change the caller interface for the two different types of locks because the semantics are entirely different for each case, and this makes it explicitly clear and, at the same time, it rids us of the extra `type' argument. The enter->lock and exit->unlock change has been made with the idea that we're "locking data" and not "entering locked code" in mind. Further, remove all additional "flags" previously passed to the lock acquire/release routines with the exception of two: MTX_QUIET and MTX_NOSWITCH The functionality of these flags is preserved and they can be passed to the lock/unlock routines by calling the corresponding wrappers: mtx_{lock, unlock}_flags(lock, flag(s)) and mtx_{lock, unlock}_spin_flags(lock, flag(s)) for MTX_DEF and MTX_SPIN locks, respectively. Re-inline some lock acq/rel code; in the sleep lock case, we only inline the _obtain_lock()s in order to ensure that the inlined code fits into a cache line. In the spin lock case, we inline recursion and actually only perform a function call if we need to spin. This change has been made with the idea that we generally tend to avoid spin locks and that also the spin locks that we do have and are heavily used (i.e. sched_lock) do recurse, and therefore in an effort to reduce function call overhead for some architectures (such as alpha), we inline recursion for this case. Create a new malloc type for the witness code and retire from using the M_DEV type. The new type is called M_WITNESS and is only declared if WITNESS is enabled. Begin cleaning up some machdep/mutex.h code - specifically updated the "optimized" inlined code in alpha/mutex.h and wrote MTX_LOCK_SPIN and MTX_UNLOCK_SPIN asm macros for the i386/mutex.h as we presently need those. Finally, caught up to the interface changes in all sys code. Contributors: jake, jhb, jasone (in no particular order)
2001-02-09 06:11:45 +00:00
mtx_unlock(&sc->sc_mtx);
DEVPRINTF((dev, "waiting for kthread exit..."));
error = tsleep(sc, PWAIT, "pccbb-detach-wait", 60 * hz);
if (error)
DPRINTF(("timeout\n"));
else
DPRINTF(("done\n"));
} else
Change and clean the mutex lock interface. mtx_enter(lock, type) becomes: mtx_lock(lock) for sleep locks (MTX_DEF-initialized locks) mtx_lock_spin(lock) for spin locks (MTX_SPIN-initialized) similarily, for releasing a lock, we now have: mtx_unlock(lock) for MTX_DEF and mtx_unlock_spin(lock) for MTX_SPIN. We change the caller interface for the two different types of locks because the semantics are entirely different for each case, and this makes it explicitly clear and, at the same time, it rids us of the extra `type' argument. The enter->lock and exit->unlock change has been made with the idea that we're "locking data" and not "entering locked code" in mind. Further, remove all additional "flags" previously passed to the lock acquire/release routines with the exception of two: MTX_QUIET and MTX_NOSWITCH The functionality of these flags is preserved and they can be passed to the lock/unlock routines by calling the corresponding wrappers: mtx_{lock, unlock}_flags(lock, flag(s)) and mtx_{lock, unlock}_spin_flags(lock, flag(s)) for MTX_DEF and MTX_SPIN locks, respectively. Re-inline some lock acq/rel code; in the sleep lock case, we only inline the _obtain_lock()s in order to ensure that the inlined code fits into a cache line. In the spin lock case, we inline recursion and actually only perform a function call if we need to spin. This change has been made with the idea that we generally tend to avoid spin locks and that also the spin locks that we do have and are heavily used (i.e. sched_lock) do recurse, and therefore in an effort to reduce function call overhead for some architectures (such as alpha), we inline recursion for this case. Create a new malloc type for the witness code and retire from using the M_DEV type. The new type is called M_WITNESS and is only declared if WITNESS is enabled. Begin cleaning up some machdep/mutex.h code - specifically updated the "optimized" inlined code in alpha/mutex.h and wrote MTX_LOCK_SPIN and MTX_UNLOCK_SPIN asm macros for the i386/mutex.h as we presently need those. Finally, caught up to the interface changes in all sys code. Contributors: jake, jhb, jasone (in no particular order)
2001-02-09 06:11:45 +00:00
mtx_unlock(&sc->sc_mtx);
bus_release_resource(dev, SYS_RES_IRQ, 0, sc->sc_irq_res);
bus_release_resource(dev, SYS_RES_MEMORY, PCCBBR_SOCKBASE,
sc->sc_base_res);
mtx_destroy(&sc->sc_mtx);
return 0;
}
static void
pccbb_driver_added(device_t dev, driver_t *driver)
{
struct pccbb_softc *sc = device_get_softc(dev);
device_t *devlist;
int tmp;
int numdevs;
DEVICE_IDENTIFY(driver, dev);
device_get_children(dev, &devlist, &numdevs);
for (tmp = 0; tmp < numdevs; tmp++) {
if (device_get_state(devlist[tmp]) == DS_NOTPRESENT &&
device_probe_and_attach(devlist[tmp]) == 0) {
if (devlist[tmp] == NULL)
/* NOTHING */;
else if (strcmp(driver->name, "cardbus") == 0) {
sc->sc_cbdev = devlist[tmp];
if ((sc->sc_socketreg->socket_state
& PCCBB_SOCKET_STAT_CD) == 0) {
Change and clean the mutex lock interface. mtx_enter(lock, type) becomes: mtx_lock(lock) for sleep locks (MTX_DEF-initialized locks) mtx_lock_spin(lock) for spin locks (MTX_SPIN-initialized) similarily, for releasing a lock, we now have: mtx_unlock(lock) for MTX_DEF and mtx_unlock_spin(lock) for MTX_SPIN. We change the caller interface for the two different types of locks because the semantics are entirely different for each case, and this makes it explicitly clear and, at the same time, it rids us of the extra `type' argument. The enter->lock and exit->unlock change has been made with the idea that we're "locking data" and not "entering locked code" in mind. Further, remove all additional "flags" previously passed to the lock acquire/release routines with the exception of two: MTX_QUIET and MTX_NOSWITCH The functionality of these flags is preserved and they can be passed to the lock/unlock routines by calling the corresponding wrappers: mtx_{lock, unlock}_flags(lock, flag(s)) and mtx_{lock, unlock}_spin_flags(lock, flag(s)) for MTX_DEF and MTX_SPIN locks, respectively. Re-inline some lock acq/rel code; in the sleep lock case, we only inline the _obtain_lock()s in order to ensure that the inlined code fits into a cache line. In the spin lock case, we inline recursion and actually only perform a function call if we need to spin. This change has been made with the idea that we generally tend to avoid spin locks and that also the spin locks that we do have and are heavily used (i.e. sched_lock) do recurse, and therefore in an effort to reduce function call overhead for some architectures (such as alpha), we inline recursion for this case. Create a new malloc type for the witness code and retire from using the M_DEV type. The new type is called M_WITNESS and is only declared if WITNESS is enabled. Begin cleaning up some machdep/mutex.h code - specifically updated the "optimized" inlined code in alpha/mutex.h and wrote MTX_LOCK_SPIN and MTX_UNLOCK_SPIN asm macros for the i386/mutex.h as we presently need those. Finally, caught up to the interface changes in all sys code. Contributors: jake, jhb, jasone (in no particular order)
2001-02-09 06:11:45 +00:00
mtx_lock(&sc->sc_mtx);
wakeup(sc);
Change and clean the mutex lock interface. mtx_enter(lock, type) becomes: mtx_lock(lock) for sleep locks (MTX_DEF-initialized locks) mtx_lock_spin(lock) for spin locks (MTX_SPIN-initialized) similarily, for releasing a lock, we now have: mtx_unlock(lock) for MTX_DEF and mtx_unlock_spin(lock) for MTX_SPIN. We change the caller interface for the two different types of locks because the semantics are entirely different for each case, and this makes it explicitly clear and, at the same time, it rids us of the extra `type' argument. The enter->lock and exit->unlock change has been made with the idea that we're "locking data" and not "entering locked code" in mind. Further, remove all additional "flags" previously passed to the lock acquire/release routines with the exception of two: MTX_QUIET and MTX_NOSWITCH The functionality of these flags is preserved and they can be passed to the lock/unlock routines by calling the corresponding wrappers: mtx_{lock, unlock}_flags(lock, flag(s)) and mtx_{lock, unlock}_spin_flags(lock, flag(s)) for MTX_DEF and MTX_SPIN locks, respectively. Re-inline some lock acq/rel code; in the sleep lock case, we only inline the _obtain_lock()s in order to ensure that the inlined code fits into a cache line. In the spin lock case, we inline recursion and actually only perform a function call if we need to spin. This change has been made with the idea that we generally tend to avoid spin locks and that also the spin locks that we do have and are heavily used (i.e. sched_lock) do recurse, and therefore in an effort to reduce function call overhead for some architectures (such as alpha), we inline recursion for this case. Create a new malloc type for the witness code and retire from using the M_DEV type. The new type is called M_WITNESS and is only declared if WITNESS is enabled. Begin cleaning up some machdep/mutex.h code - specifically updated the "optimized" inlined code in alpha/mutex.h and wrote MTX_LOCK_SPIN and MTX_UNLOCK_SPIN asm macros for the i386/mutex.h as we presently need those. Finally, caught up to the interface changes in all sys code. Contributors: jake, jhb, jasone (in no particular order)
2001-02-09 06:11:45 +00:00
mtx_unlock(&sc->sc_mtx);
}
} else if (strcmp(driver->name, "pccard") == 0) {
sc->sc_pccarddev = devlist[tmp];
if ((sc->sc_socketreg->socket_state
& PCCBB_SOCKET_STAT_CD) == 0) {
Change and clean the mutex lock interface. mtx_enter(lock, type) becomes: mtx_lock(lock) for sleep locks (MTX_DEF-initialized locks) mtx_lock_spin(lock) for spin locks (MTX_SPIN-initialized) similarily, for releasing a lock, we now have: mtx_unlock(lock) for MTX_DEF and mtx_unlock_spin(lock) for MTX_SPIN. We change the caller interface for the two different types of locks because the semantics are entirely different for each case, and this makes it explicitly clear and, at the same time, it rids us of the extra `type' argument. The enter->lock and exit->unlock change has been made with the idea that we're "locking data" and not "entering locked code" in mind. Further, remove all additional "flags" previously passed to the lock acquire/release routines with the exception of two: MTX_QUIET and MTX_NOSWITCH The functionality of these flags is preserved and they can be passed to the lock/unlock routines by calling the corresponding wrappers: mtx_{lock, unlock}_flags(lock, flag(s)) and mtx_{lock, unlock}_spin_flags(lock, flag(s)) for MTX_DEF and MTX_SPIN locks, respectively. Re-inline some lock acq/rel code; in the sleep lock case, we only inline the _obtain_lock()s in order to ensure that the inlined code fits into a cache line. In the spin lock case, we inline recursion and actually only perform a function call if we need to spin. This change has been made with the idea that we generally tend to avoid spin locks and that also the spin locks that we do have and are heavily used (i.e. sched_lock) do recurse, and therefore in an effort to reduce function call overhead for some architectures (such as alpha), we inline recursion for this case. Create a new malloc type for the witness code and retire from using the M_DEV type. The new type is called M_WITNESS and is only declared if WITNESS is enabled. Begin cleaning up some machdep/mutex.h code - specifically updated the "optimized" inlined code in alpha/mutex.h and wrote MTX_LOCK_SPIN and MTX_UNLOCK_SPIN asm macros for the i386/mutex.h as we presently need those. Finally, caught up to the interface changes in all sys code. Contributors: jake, jhb, jasone (in no particular order)
2001-02-09 06:11:45 +00:00
mtx_lock(&sc->sc_mtx);
wakeup(sc);
Change and clean the mutex lock interface. mtx_enter(lock, type) becomes: mtx_lock(lock) for sleep locks (MTX_DEF-initialized locks) mtx_lock_spin(lock) for spin locks (MTX_SPIN-initialized) similarily, for releasing a lock, we now have: mtx_unlock(lock) for MTX_DEF and mtx_unlock_spin(lock) for MTX_SPIN. We change the caller interface for the two different types of locks because the semantics are entirely different for each case, and this makes it explicitly clear and, at the same time, it rids us of the extra `type' argument. The enter->lock and exit->unlock change has been made with the idea that we're "locking data" and not "entering locked code" in mind. Further, remove all additional "flags" previously passed to the lock acquire/release routines with the exception of two: MTX_QUIET and MTX_NOSWITCH The functionality of these flags is preserved and they can be passed to the lock/unlock routines by calling the corresponding wrappers: mtx_{lock, unlock}_flags(lock, flag(s)) and mtx_{lock, unlock}_spin_flags(lock, flag(s)) for MTX_DEF and MTX_SPIN locks, respectively. Re-inline some lock acq/rel code; in the sleep lock case, we only inline the _obtain_lock()s in order to ensure that the inlined code fits into a cache line. In the spin lock case, we inline recursion and actually only perform a function call if we need to spin. This change has been made with the idea that we generally tend to avoid spin locks and that also the spin locks that we do have and are heavily used (i.e. sched_lock) do recurse, and therefore in an effort to reduce function call overhead for some architectures (such as alpha), we inline recursion for this case. Create a new malloc type for the witness code and retire from using the M_DEV type. The new type is called M_WITNESS and is only declared if WITNESS is enabled. Begin cleaning up some machdep/mutex.h code - specifically updated the "optimized" inlined code in alpha/mutex.h and wrote MTX_LOCK_SPIN and MTX_UNLOCK_SPIN asm macros for the i386/mutex.h as we presently need those. Finally, caught up to the interface changes in all sys code. Contributors: jake, jhb, jasone (in no particular order)
2001-02-09 06:11:45 +00:00
mtx_unlock(&sc->sc_mtx);
}
} else
device_printf(dev,
"Unsupported child bus: %s\n",
driver->name);
}
}
free(devlist, M_TEMP);
}
static void
pccbb_child_detached(device_t dev, device_t child)
{
struct pccbb_softc *sc = device_get_softc(dev);
if (child == sc->sc_cbdev)
sc->sc_cbdev = NULL;
else if (child == sc->sc_pccarddev)
sc->sc_pccarddev = NULL;
else
device_printf(dev, "Unknown child detached: %s %p/%p\n",
device_get_nameunit(child), sc->sc_cbdev, sc->sc_pccarddev);
}
/************************************************************************/
/* Kthreads */
/************************************************************************/
static void
pccbb_event_thread (void *arg)
{
struct pccbb_softc *sc = arg;
u_int32_t status;
Change and clean the mutex lock interface. mtx_enter(lock, type) becomes: mtx_lock(lock) for sleep locks (MTX_DEF-initialized locks) mtx_lock_spin(lock) for spin locks (MTX_SPIN-initialized) similarily, for releasing a lock, we now have: mtx_unlock(lock) for MTX_DEF and mtx_unlock_spin(lock) for MTX_SPIN. We change the caller interface for the two different types of locks because the semantics are entirely different for each case, and this makes it explicitly clear and, at the same time, it rids us of the extra `type' argument. The enter->lock and exit->unlock change has been made with the idea that we're "locking data" and not "entering locked code" in mind. Further, remove all additional "flags" previously passed to the lock acquire/release routines with the exception of two: MTX_QUIET and MTX_NOSWITCH The functionality of these flags is preserved and they can be passed to the lock/unlock routines by calling the corresponding wrappers: mtx_{lock, unlock}_flags(lock, flag(s)) and mtx_{lock, unlock}_spin_flags(lock, flag(s)) for MTX_DEF and MTX_SPIN locks, respectively. Re-inline some lock acq/rel code; in the sleep lock case, we only inline the _obtain_lock()s in order to ensure that the inlined code fits into a cache line. In the spin lock case, we inline recursion and actually only perform a function call if we need to spin. This change has been made with the idea that we generally tend to avoid spin locks and that also the spin locks that we do have and are heavily used (i.e. sched_lock) do recurse, and therefore in an effort to reduce function call overhead for some architectures (such as alpha), we inline recursion for this case. Create a new malloc type for the witness code and retire from using the M_DEV type. The new type is called M_WITNESS and is only declared if WITNESS is enabled. Begin cleaning up some machdep/mutex.h code - specifically updated the "optimized" inlined code in alpha/mutex.h and wrote MTX_LOCK_SPIN and MTX_UNLOCK_SPIN asm macros for the i386/mutex.h as we presently need those. Finally, caught up to the interface changes in all sys code. Contributors: jake, jhb, jasone (in no particular order)
2001-02-09 06:11:45 +00:00
mtx_lock(&Giant);
for(;;) {
if (!(sc->sc_flags & PCCBB_KTHREAD_RUNNING))
sc->sc_flags |= PCCBB_KTHREAD_RUNNING;
else {
tsleep (sc, PWAIT, "pccbbev", 0);
/*
* Delay 1 second, make sure the user is done with
* whatever he is doing. We tsleep on sc->sc_flags,
* which should never be woken up.
*/
tsleep (&sc->sc_flags, PWAIT, "pccbbev", 1*hz);
}
Change and clean the mutex lock interface. mtx_enter(lock, type) becomes: mtx_lock(lock) for sleep locks (MTX_DEF-initialized locks) mtx_lock_spin(lock) for spin locks (MTX_SPIN-initialized) similarily, for releasing a lock, we now have: mtx_unlock(lock) for MTX_DEF and mtx_unlock_spin(lock) for MTX_SPIN. We change the caller interface for the two different types of locks because the semantics are entirely different for each case, and this makes it explicitly clear and, at the same time, it rids us of the extra `type' argument. The enter->lock and exit->unlock change has been made with the idea that we're "locking data" and not "entering locked code" in mind. Further, remove all additional "flags" previously passed to the lock acquire/release routines with the exception of two: MTX_QUIET and MTX_NOSWITCH The functionality of these flags is preserved and they can be passed to the lock/unlock routines by calling the corresponding wrappers: mtx_{lock, unlock}_flags(lock, flag(s)) and mtx_{lock, unlock}_spin_flags(lock, flag(s)) for MTX_DEF and MTX_SPIN locks, respectively. Re-inline some lock acq/rel code; in the sleep lock case, we only inline the _obtain_lock()s in order to ensure that the inlined code fits into a cache line. In the spin lock case, we inline recursion and actually only perform a function call if we need to spin. This change has been made with the idea that we generally tend to avoid spin locks and that also the spin locks that we do have and are heavily used (i.e. sched_lock) do recurse, and therefore in an effort to reduce function call overhead for some architectures (such as alpha), we inline recursion for this case. Create a new malloc type for the witness code and retire from using the M_DEV type. The new type is called M_WITNESS and is only declared if WITNESS is enabled. Begin cleaning up some machdep/mutex.h code - specifically updated the "optimized" inlined code in alpha/mutex.h and wrote MTX_LOCK_SPIN and MTX_UNLOCK_SPIN asm macros for the i386/mutex.h as we presently need those. Finally, caught up to the interface changes in all sys code. Contributors: jake, jhb, jasone (in no particular order)
2001-02-09 06:11:45 +00:00
mtx_lock(&sc->sc_mtx);
if (sc->sc_flags & PCCBB_KTHREAD_DONE)
break;
status = sc->sc_socketreg->socket_state;
if ((status & PCCBB_SOCKET_STAT_CD) == 0) {
pccbb_insert(sc);
} else {
pccbb_removal(sc);
}
Change and clean the mutex lock interface. mtx_enter(lock, type) becomes: mtx_lock(lock) for sleep locks (MTX_DEF-initialized locks) mtx_lock_spin(lock) for spin locks (MTX_SPIN-initialized) similarily, for releasing a lock, we now have: mtx_unlock(lock) for MTX_DEF and mtx_unlock_spin(lock) for MTX_SPIN. We change the caller interface for the two different types of locks because the semantics are entirely different for each case, and this makes it explicitly clear and, at the same time, it rids us of the extra `type' argument. The enter->lock and exit->unlock change has been made with the idea that we're "locking data" and not "entering locked code" in mind. Further, remove all additional "flags" previously passed to the lock acquire/release routines with the exception of two: MTX_QUIET and MTX_NOSWITCH The functionality of these flags is preserved and they can be passed to the lock/unlock routines by calling the corresponding wrappers: mtx_{lock, unlock}_flags(lock, flag(s)) and mtx_{lock, unlock}_spin_flags(lock, flag(s)) for MTX_DEF and MTX_SPIN locks, respectively. Re-inline some lock acq/rel code; in the sleep lock case, we only inline the _obtain_lock()s in order to ensure that the inlined code fits into a cache line. In the spin lock case, we inline recursion and actually only perform a function call if we need to spin. This change has been made with the idea that we generally tend to avoid spin locks and that also the spin locks that we do have and are heavily used (i.e. sched_lock) do recurse, and therefore in an effort to reduce function call overhead for some architectures (such as alpha), we inline recursion for this case. Create a new malloc type for the witness code and retire from using the M_DEV type. The new type is called M_WITNESS and is only declared if WITNESS is enabled. Begin cleaning up some machdep/mutex.h code - specifically updated the "optimized" inlined code in alpha/mutex.h and wrote MTX_LOCK_SPIN and MTX_UNLOCK_SPIN asm macros for the i386/mutex.h as we presently need those. Finally, caught up to the interface changes in all sys code. Contributors: jake, jhb, jasone (in no particular order)
2001-02-09 06:11:45 +00:00
mtx_unlock(&sc->sc_mtx);
}
Change and clean the mutex lock interface. mtx_enter(lock, type) becomes: mtx_lock(lock) for sleep locks (MTX_DEF-initialized locks) mtx_lock_spin(lock) for spin locks (MTX_SPIN-initialized) similarily, for releasing a lock, we now have: mtx_unlock(lock) for MTX_DEF and mtx_unlock_spin(lock) for MTX_SPIN. We change the caller interface for the two different types of locks because the semantics are entirely different for each case, and this makes it explicitly clear and, at the same time, it rids us of the extra `type' argument. The enter->lock and exit->unlock change has been made with the idea that we're "locking data" and not "entering locked code" in mind. Further, remove all additional "flags" previously passed to the lock acquire/release routines with the exception of two: MTX_QUIET and MTX_NOSWITCH The functionality of these flags is preserved and they can be passed to the lock/unlock routines by calling the corresponding wrappers: mtx_{lock, unlock}_flags(lock, flag(s)) and mtx_{lock, unlock}_spin_flags(lock, flag(s)) for MTX_DEF and MTX_SPIN locks, respectively. Re-inline some lock acq/rel code; in the sleep lock case, we only inline the _obtain_lock()s in order to ensure that the inlined code fits into a cache line. In the spin lock case, we inline recursion and actually only perform a function call if we need to spin. This change has been made with the idea that we generally tend to avoid spin locks and that also the spin locks that we do have and are heavily used (i.e. sched_lock) do recurse, and therefore in an effort to reduce function call overhead for some architectures (such as alpha), we inline recursion for this case. Create a new malloc type for the witness code and retire from using the M_DEV type. The new type is called M_WITNESS and is only declared if WITNESS is enabled. Begin cleaning up some machdep/mutex.h code - specifically updated the "optimized" inlined code in alpha/mutex.h and wrote MTX_LOCK_SPIN and MTX_UNLOCK_SPIN asm macros for the i386/mutex.h as we presently need those. Finally, caught up to the interface changes in all sys code. Contributors: jake, jhb, jasone (in no particular order)
2001-02-09 06:11:45 +00:00
mtx_unlock(&sc->sc_mtx);
sc->sc_flags &= ~PCCBB_KTHREAD_RUNNING;
wakeup(sc);
kthread_exit(0);
}
static void
pccbb_create_event_thread (struct pccbb_softc *sc)
{
if (kthread_create(pccbb_event_thread, sc, &sc->event_thread,
0, "%s%d", device_get_name(sc->sc_dev),
device_get_unit(sc->sc_dev))) {
device_printf (sc->sc_dev, "unable to create event thread.\n");
panic ("pccbb_create_event_thread");
}
}
static void
pccbb_start_threads(void *arg)
{
struct pccbb_sclist *sclist;
STAILQ_FOREACH(sclist, &softcs, entries) {
pccbb_create_event_thread(sclist->sc);
}
}
/************************************************************************/
/* Insert/removal */
/************************************************************************/
static void
pccbb_insert (struct pccbb_softc *sc)
{
u_int32_t sockevent, sockstate;
int timeout = 30;
do {
sockevent = sc->sc_socketreg->socket_event;
sockstate = sc->sc_socketreg->socket_state;
} while (sockstate & PCCBB_SOCKET_STAT_CD && --timeout > 0);
if (timeout < 0) {
device_printf (sc->sc_dev, "insert timeout");
return;
}
DEVPRINTF((sc->sc_dev, "card inserted: event=0x%08x, state=%08x\n",
sockevent, sockstate));
if (sockstate & PCCBB_SOCKET_STAT_16BIT && sc->sc_pccarddev != NULL) {
sc->sc_flags |= PCCBB_16BIT_CARD;
if (CARD_ATTACH_CARD(sc->sc_pccarddev) != 0)
device_printf(sc->sc_dev, "card activation failed\n");
} else if (sockstate & PCCBB_SOCKET_STAT_CB && sc->sc_cbdev != NULL) {
sc->sc_flags &= ~PCCBB_16BIT_CARD;
if (CARD_ATTACH_CARD(sc->sc_cbdev) != 0)
device_printf(sc->sc_dev, "card activation failed\n");
} else {
device_printf (sc->sc_dev, "Unsupported card type detected\n");
}
}
static void
pccbb_removal (struct pccbb_softc *sc)
{
u_int32_t sockstate;
struct pccbb_reslist *rle;
sockstate = sc->sc_socketreg->socket_state;
if (sockstate & PCCBB_16BIT_CARD && sc->sc_pccarddev != NULL)
CARD_DETACH_CARD(sc->sc_pccarddev, DETACH_FORCE);
else if ((!(sockstate & PCCBB_16BIT_CARD)) && sc->sc_cbdev != NULL)
CARD_DETACH_CARD(sc->sc_cbdev, DETACH_FORCE);
while (NULL != (rle = SLIST_FIRST(&sc->rl))) {
device_printf(sc->sc_dev, "WARNING: Resource left allocated! "
"This is a bug... (rid=%x, type=%d, addr=%x)\n",
rle->rid, rle->type, rle->start);
SLIST_REMOVE_HEAD(&sc->rl, entries);
}
}
/************************************************************************/
/* Interrupt Handler */
/************************************************************************/
static void
pccbb_intr(void* arg)
{
struct pccbb_softc *sc = arg;
u_int32_t sockevent;
if (!(sockevent = sc->sc_socketreg->socket_event)) {
/* not for me. */
return;
}
/* reset bit */
sc->sc_socketreg->socket_event = sockevent | 0x01;
if (sockevent & PCCBB_SOCKET_EVENT_CD) {
Change and clean the mutex lock interface. mtx_enter(lock, type) becomes: mtx_lock(lock) for sleep locks (MTX_DEF-initialized locks) mtx_lock_spin(lock) for spin locks (MTX_SPIN-initialized) similarily, for releasing a lock, we now have: mtx_unlock(lock) for MTX_DEF and mtx_unlock_spin(lock) for MTX_SPIN. We change the caller interface for the two different types of locks because the semantics are entirely different for each case, and this makes it explicitly clear and, at the same time, it rids us of the extra `type' argument. The enter->lock and exit->unlock change has been made with the idea that we're "locking data" and not "entering locked code" in mind. Further, remove all additional "flags" previously passed to the lock acquire/release routines with the exception of two: MTX_QUIET and MTX_NOSWITCH The functionality of these flags is preserved and they can be passed to the lock/unlock routines by calling the corresponding wrappers: mtx_{lock, unlock}_flags(lock, flag(s)) and mtx_{lock, unlock}_spin_flags(lock, flag(s)) for MTX_DEF and MTX_SPIN locks, respectively. Re-inline some lock acq/rel code; in the sleep lock case, we only inline the _obtain_lock()s in order to ensure that the inlined code fits into a cache line. In the spin lock case, we inline recursion and actually only perform a function call if we need to spin. This change has been made with the idea that we generally tend to avoid spin locks and that also the spin locks that we do have and are heavily used (i.e. sched_lock) do recurse, and therefore in an effort to reduce function call overhead for some architectures (such as alpha), we inline recursion for this case. Create a new malloc type for the witness code and retire from using the M_DEV type. The new type is called M_WITNESS and is only declared if WITNESS is enabled. Begin cleaning up some machdep/mutex.h code - specifically updated the "optimized" inlined code in alpha/mutex.h and wrote MTX_LOCK_SPIN and MTX_UNLOCK_SPIN asm macros for the i386/mutex.h as we presently need those. Finally, caught up to the interface changes in all sys code. Contributors: jake, jhb, jasone (in no particular order)
2001-02-09 06:11:45 +00:00
mtx_lock(&sc->sc_mtx);
wakeup(sc);
Change and clean the mutex lock interface. mtx_enter(lock, type) becomes: mtx_lock(lock) for sleep locks (MTX_DEF-initialized locks) mtx_lock_spin(lock) for spin locks (MTX_SPIN-initialized) similarily, for releasing a lock, we now have: mtx_unlock(lock) for MTX_DEF and mtx_unlock_spin(lock) for MTX_SPIN. We change the caller interface for the two different types of locks because the semantics are entirely different for each case, and this makes it explicitly clear and, at the same time, it rids us of the extra `type' argument. The enter->lock and exit->unlock change has been made with the idea that we're "locking data" and not "entering locked code" in mind. Further, remove all additional "flags" previously passed to the lock acquire/release routines with the exception of two: MTX_QUIET and MTX_NOSWITCH The functionality of these flags is preserved and they can be passed to the lock/unlock routines by calling the corresponding wrappers: mtx_{lock, unlock}_flags(lock, flag(s)) and mtx_{lock, unlock}_spin_flags(lock, flag(s)) for MTX_DEF and MTX_SPIN locks, respectively. Re-inline some lock acq/rel code; in the sleep lock case, we only inline the _obtain_lock()s in order to ensure that the inlined code fits into a cache line. In the spin lock case, we inline recursion and actually only perform a function call if we need to spin. This change has been made with the idea that we generally tend to avoid spin locks and that also the spin locks that we do have and are heavily used (i.e. sched_lock) do recurse, and therefore in an effort to reduce function call overhead for some architectures (such as alpha), we inline recursion for this case. Create a new malloc type for the witness code and retire from using the M_DEV type. The new type is called M_WITNESS and is only declared if WITNESS is enabled. Begin cleaning up some machdep/mutex.h code - specifically updated the "optimized" inlined code in alpha/mutex.h and wrote MTX_LOCK_SPIN and MTX_UNLOCK_SPIN asm macros for the i386/mutex.h as we presently need those. Finally, caught up to the interface changes in all sys code. Contributors: jake, jhb, jasone (in no particular order)
2001-02-09 06:11:45 +00:00
mtx_unlock(&sc->sc_mtx);
} else {
if (sockevent & PCCBB_SOCKET_EVENT_CSTS) {
DPRINTF((" cstsevent occures, 0x%08x\n",
sc->sc_socketreg->socket_state));
}
if (sockevent & PCCBB_SOCKET_EVENT_POWER) {
DPRINTF((" pwrevent occures, 0x%08x\n",
sc->sc_socketreg->socket_state));
}
}
return;
}
/************************************************************************/
/* Generic Power functions */
/************************************************************************/
static int
pccbb_detect_voltage(device_t dev)
{
struct pccbb_softc *sc = device_get_softc(dev);
u_int32_t psr;
int vol = CARD_UKN_CARD;
psr = sc->sc_socketreg->socket_state;
if (psr & PCCBB_SOCKET_STAT_5VCARD) {
vol |= CARD_5V_CARD;
}
if (psr & PCCBB_SOCKET_STAT_3VCARD) {
vol |= CARD_3V_CARD;
}
if (psr & PCCBB_SOCKET_STAT_XVCARD) {
vol |= CARD_XV_CARD;
}
if (psr & PCCBB_SOCKET_STAT_YVCARD) {
vol |= CARD_YV_CARD;
}
return vol;
}
static int
pccbb_power(device_t dev, int volts)
{
u_int32_t status, sock_ctrl;
struct pccbb_softc *sc = device_get_softc(dev);
DEVPRINTF((sc->sc_dev, "pccbb_power: %s and %s [%x]\n",
(volts & CARD_VCCMASK) == CARD_VCC_UC ? "CARD_VCC_UC" :
(volts & CARD_VCCMASK) == CARD_VCC_5V ? "CARD_VCC_5V" :
(volts & CARD_VCCMASK) == CARD_VCC_3V ? "CARD_VCC_3V" :
(volts & CARD_VCCMASK) == CARD_VCC_XV ? "CARD_VCC_XV" :
(volts & CARD_VCCMASK) == CARD_VCC_YV ? "CARD_VCC_YV" :
(volts & CARD_VCCMASK) == CARD_VCC_0V ? "CARD_VCC_0V" :
"VCC-UNKNOWN",
(volts & CARD_VPPMASK) == CARD_VPP_UC ? "CARD_VPP_UC" :
(volts & CARD_VPPMASK) == CARD_VPP_12V ? "CARD_VPP_12V" :
(volts & CARD_VPPMASK) == CARD_VPP_VCC ? "CARD_VPP_VCC" :
(volts & CARD_VPPMASK) == CARD_VPP_0V ? "CARD_VPP_0V" :
"VPP-UNKNOWN",
volts));
status = sc->sc_socketreg->socket_state;
sock_ctrl = sc->sc_socketreg->socket_control;
switch (volts & CARD_VCCMASK) {
case CARD_VCC_UC:
break;
case CARD_VCC_5V:
if (PCCBB_SOCKET_STAT_5VCARD & status) { /* check 5 V card */
sock_ctrl &= ~PCCBB_SOCKET_CTRL_VCCMASK;
sock_ctrl |= PCCBB_SOCKET_CTRL_VCC_5V;
} else {
device_printf(sc->sc_dev,
"BAD voltage request: no 5 V card\n");
}
break;
case CARD_VCC_3V:
if (PCCBB_SOCKET_STAT_3VCARD & status) {
sock_ctrl &= ~PCCBB_SOCKET_CTRL_VCCMASK;
sock_ctrl |= PCCBB_SOCKET_CTRL_VCC_3V;
} else {
device_printf(sc->sc_dev,
"BAD voltage request: no 3.3 V card\n");
}
break;
case CARD_VCC_0V:
sock_ctrl &= ~PCCBB_SOCKET_CTRL_VCCMASK;
break;
default:
return 0; /* power NEVER changed */
break;
}
switch (volts & CARD_VPPMASK) {
case CARD_VPP_UC:
break;
case CARD_VPP_0V:
sock_ctrl &= ~PCCBB_SOCKET_CTRL_VPPMASK;
break;
case CARD_VPP_VCC:
sock_ctrl &= ~PCCBB_SOCKET_CTRL_VPPMASK;
sock_ctrl |= ((sock_ctrl >> 4) & 0x07);
break;
case CARD_VPP_12V:
sock_ctrl &= ~PCCBB_SOCKET_CTRL_VPPMASK;
sock_ctrl |= PCCBB_SOCKET_CTRL_VPP_12V;
break;
}
if (sc->sc_socketreg->socket_control == sock_ctrl)
return 1; /* no change necessary */
sc->sc_socketreg->socket_control = sock_ctrl;
status = sc->sc_socketreg->socket_state;
{
int timeout = 20;
u_int32_t sockevent;
do {
DELAY(20*1000);
sockevent = sc->sc_socketreg->socket_event;
} while (!(sockevent & PCCBB_SOCKET_EVENT_POWER) &&
--timeout > 0);
/* reset event status */
sc->sc_socketreg->socket_event = sockevent;
if ( timeout < 0 ) {
printf ("VCC supply failed.\n");
return 0;
}
}
/* XXX
* delay 400 ms: thgough the standard defines that the Vcc set-up time
* is 20 ms, some PC-Card bridge requires longer duration.
*/
DELAY(400*1000);
if (status & PCCBB_SOCKET_STAT_BADVCC) {
device_printf(sc->sc_dev,
"bad Vcc request. ctrl=0x%x, status=0x%x\n",
sock_ctrl ,status);
printf("pccbb_power: %s and %s [%x]\n",
(volts & CARD_VCCMASK) == CARD_VCC_UC ? "CARD_VCC_UC" :
(volts & CARD_VCCMASK) == CARD_VCC_5V ? "CARD_VCC_5V" :
(volts & CARD_VCCMASK) == CARD_VCC_3V ? "CARD_VCC_3V" :
(volts & CARD_VCCMASK) == CARD_VCC_XV ? "CARD_VCC_XV" :
(volts & CARD_VCCMASK) == CARD_VCC_YV ? "CARD_VCC_YV" :
(volts & CARD_VCCMASK) == CARD_VCC_0V ? "CARD_VCC_0V" :
"VCC-UNKNOWN",
(volts & CARD_VPPMASK) == CARD_VPP_UC ? "CARD_VPP_UC" :
(volts & CARD_VPPMASK) == CARD_VPP_12V ? "CARD_VPP_12V":
(volts & CARD_VPPMASK) == CARD_VPP_VCC ? "CARD_VPP_VCC":
(volts & CARD_VPPMASK) == CARD_VPP_0V ? "CARD_VPP_0V" :
"VPP-UNKNOWN",
volts);
return 0;
}
return 1; /* power changed correctly */
}
/************************************************************************/
/* Cardbus power functions */
/************************************************************************/
static void
pccbb_cardbus_reset(device_t dev)
{
struct pccbb_softc *sc = device_get_softc(dev);
int delay_us;
delay_us = sc->sc_chipset == CB_RF5C47X ? 400*1000 : 20*1000;
PCI_MASK_CONFIG(dev, PCCBBR_BRIDGECTRL, |PCCBBM_BRIDGECTRL_RESET, 2);
DELAY(delay_us);
/* If a card exists, unreset it! */
if ((sc->sc_socketreg->socket_state & PCCBB_SOCKET_STAT_CD) == 0) {
PCI_MASK_CONFIG(dev, PCCBBR_BRIDGECTRL,
&~PCCBBM_BRIDGECTRL_RESET, 2);
DELAY(delay_us);
}
}
static int
pccbb_cardbus_power_enable_socket(device_t self, device_t child)
{
struct pccbb_softc *sc = device_get_softc(self);
int voltage;
if ((sc->sc_socketreg->socket_state & PCCBB_SOCKET_STAT_CD)
== PCCBB_SOCKET_STAT_CD)
return ENODEV;
voltage = pccbb_detect_voltage(self);
pccbb_power(self, CARD_VCC_0V | CARD_VPP_0V);
if (voltage & CARD_5V_CARD)
pccbb_power(self, CARD_VCC_5V | CARD_VPP_VCC);
else if (voltage & CARD_3V_CARD)
pccbb_power(self, CARD_VCC_3V | CARD_VPP_VCC);
else {
device_printf(self, "Unknown card voltage\n");
return ENXIO;
}
pccbb_cardbus_reset(self);
return 0;
}
static void
pccbb_cardbus_power_disable_socket(device_t self, device_t child)
{
pccbb_power(self, CARD_VCC_0V | CARD_VPP_0V);
pccbb_cardbus_reset(self);
}
/************************************************************************/
/* Cardbus Resource */
/************************************************************************/
static int
pccbb_cardbus_io_open(device_t dev, int win, u_int32_t start, u_int32_t end)
{
int basereg;
int limitreg;
if ((win < 0) || (win > 1)) {
DEVPRINTF((dev,
"pccbb_cardbus_io_open: window out of range %d\n",
win));
return EINVAL;
}
basereg = win*8 + PCCBBR_IOBASE0;
limitreg = win*8 + PCCBBR_IOLIMIT0;
pci_write_config(dev, basereg, start, 4);
pci_write_config(dev, limitreg, end, 4);
return 0;
}
static int
pccbb_cardbus_mem_open(device_t dev, int win, u_int32_t start, u_int32_t end)
{
int basereg;
int limitreg;
if ((win < 0) || (win > 1)) {
DEVPRINTF((dev,
"pccbb_cardbus_mem_open: window out of range %d\n",
win));
return EINVAL;
}
basereg = win*8 + PCCBBR_MEMBASE0;
limitreg = win*8 + PCCBBR_MEMLIMIT0;
pci_write_config(dev, basereg, start, 4);
pci_write_config(dev, limitreg, end, 4);
return 0;
}
static void
pccbb_cardbus_auto_open(struct pccbb_softc *sc, int type)
{
u_int32_t starts[2];
u_int32_t ends[2];
struct pccbb_reslist *rle;
int align;
starts[0] = starts[1] = 0xffffffff;
ends[0] = ends[1] = 0;
SLIST_FOREACH(rle, &sc->rl, entries) {
if (rle->type != type)
;
else if (starts[0] == 0xffffffff) {
starts[0] = rle->start;
ends[0] = rle->end;
rle->win = 0;
} else if (rle->end > ends[0] &&
rle->start - ends[0] < PCCBB_AUTO_OPEN_SMALLHOLE) {
ends[0] = rle->end;
rle->win = 0;
} else if (rle->start < starts[0] &&
starts[0] - rle->end < PCCBB_AUTO_OPEN_SMALLHOLE) {
starts[0] = rle->start;
rle->win = 0;
} else if (starts[1] == 0xffffffff) {
starts[1] = rle->start;
ends[1] = rle->end;
rle->win = 1;
} else if (rle->end > ends[1] &&
rle->start - ends[1] < PCCBB_AUTO_OPEN_SMALLHOLE) {
ends[1] = rle->end;
rle->win = 1;
} else if (rle->start < starts[1] &&
starts[1] - rle->end < PCCBB_AUTO_OPEN_SMALLHOLE) {
starts[1] = rle->start;
rle->win = 1;
} else {
u_int32_t diffs[2];
diffs[0] = diffs[1] = 0xffffffff;
if (rle->start > ends[0])
diffs[0] = rle->start - ends[0];
else if (rle->end < starts[0])
diffs[0] = starts[0] - rle->end;
if (rle->start > ends[1])
diffs[1] = rle->start - ends[1];
else if (rle->end < starts[1])
diffs[1] = starts[1] - rle->end;
rle->win = (diffs[0] <= diffs[1])?0:1;
if (rle->start > ends[rle->win])
ends[rle->win] = rle->end;
else if (rle->end < starts[rle->win])
starts[rle->win] = rle->start;
}
}
if (type == SYS_RES_MEMORY)
align = PCCBB_MEMALIGN;
else if (type == SYS_RES_IOPORT)
align = PCCBB_IOALIGN;
else
align = 1;
if (starts[0] != 0xffffffff)
starts[0] -= starts[0] % align;
if (starts[1] != 0xffffffff)
starts[1] -= starts[1] % align;
if (ends[0] % align != 0)
ends[0] += align - ends[0]%align - 1;
if (ends[1] % align != 0)
ends[1] += align - ends[1]%align - 1;
if (type == SYS_RES_MEMORY) {
pccbb_cardbus_mem_open(sc->sc_dev, 0, starts[0], ends[0]);
pccbb_cardbus_mem_open(sc->sc_dev, 1, starts[1], ends[1]);
} else if (type == SYS_RES_IOPORT) {
pccbb_cardbus_io_open(sc->sc_dev, 0, starts[0], ends[0]);
pccbb_cardbus_io_open(sc->sc_dev, 1, starts[1], ends[1]);
}
}
static int
pccbb_cardbus_activate_resource(device_t self, device_t child, int type,
int rid, struct resource *r)
{
struct pccbb_softc *sc = device_get_softc(self);
struct pccbb_reslist *rle;
if (type == SYS_RES_MEMORY || type == SYS_RES_IOPORT) {
SLIST_FOREACH(rle, &sc->rl, entries) {
if (type == rle->type && rid == rle->rid &&
child == rle->odev)
return bus_generic_activate_resource(
self, child, type, rid, r);
}
rle = malloc(sizeof(struct pccbb_reslist), M_DEVBUF, M_WAITOK);
rle->type = type;
rle->rid = rid;
rle->start = rman_get_start(r);
rle->end = rman_get_end(r);
rle->odev = child;
rle->win = -1;
SLIST_INSERT_HEAD(&sc->rl, rle, entries);
pccbb_cardbus_auto_open(sc, type);
}
return bus_generic_activate_resource(self, child, type, rid, r);
}
static int
pccbb_cardbus_deactivate_resource(device_t self, device_t child, int type,
int rid, struct resource *r)
{
struct pccbb_softc *sc = device_get_softc(self);
struct pccbb_reslist *rle;
SLIST_FOREACH(rle, &sc->rl, entries) {
if (type == rle->type && rid == rle->rid &&
child == rle->odev) {
SLIST_REMOVE(&sc->rl, rle, pccbb_reslist, entries);
if (type == SYS_RES_IOPORT ||
type == SYS_RES_MEMORY)
pccbb_cardbus_auto_open(sc, type);
free(rle, M_DEVBUF);
break;
}
}
return bus_generic_deactivate_resource(self, child, type, rid, r);
}
static struct resource*
pccbb_cardbus_alloc_resource(device_t self, device_t child, int type, int* rid,
u_long start, u_long end, u_long count,
u_int flags)
{
if (type == SYS_RES_IRQ) {
struct pccbb_softc *sc = device_get_softc(self);
if (start == 0) {
start = end = rman_get_start(sc->sc_irq_res);
}
return bus_generic_alloc_resource(self, child, type, rid,
start, end, count, flags);
} else {
if (type == SYS_RES_MEMORY && start == 0 && end == ~0) {
start = CARDBUS_SYS_RES_MEMORY_START;
end = CARDBUS_SYS_RES_MEMORY_END;
} else if (type == SYS_RES_IOPORT && start == 0 && end == ~0) {
start = CARDBUS_SYS_RES_IOPORT_START;
end = CARDBUS_SYS_RES_IOPORT_END;
}
return bus_generic_alloc_resource(self, child, type, rid,
start, end, count, flags);
}
}
static int
pccbb_cardbus_release_resource(device_t self, device_t child, int type,
int rid, struct resource *r)
{
return bus_generic_release_resource(self, child, type, rid, r);
}
/************************************************************************/
/* PC Card Power Functions */
/************************************************************************/
static int
pccbb_pcic_power_enable_socket(device_t self, device_t child)
{
struct pccbb_softc *sc = device_get_softc(self);
DPRINTF(("pccbb_pcic_socket_enable:\n"));
/* power down/up the socket to reset */
{
int voltage = pccbb_detect_voltage(self);
pccbb_power(self, CARD_VCC_0V | CARD_VPP_0V);
if (voltage & CARD_5V_CARD)
pccbb_power(self, CARD_VCC_5V | CARD_VPP_VCC);
else if (voltage & CARD_3V_CARD)
pccbb_power(self, CARD_VCC_3V | CARD_VPP_VCC);
else {
device_printf(self, "Unknown card voltage\n");
return ENXIO;
}
}
/* enable socket i/o */
PCIC_MASK(sc, PCIC_PWRCTL, | PCIC_PWRCTL_OE);
PCIC_WRITE(sc, PCIC_INTR, PCIC_INTR_ENABLE);
/* hold reset for 30ms */
DELAY(30*1000);
/* clear the reset flag */
PCIC_MASK(sc, PCIC_INTR, | PCIC_INTR_RESET);
/* wait 20ms as per pc card standard (r2.01) section 4.3.6 */
DELAY(20*1000);
pccbb_pcic_wait_ready(sc);
/* disable all address windows */
PCIC_WRITE(sc, PCIC_ADDRWIN_ENABLE, 0);
{
int cardtype;
CARD_GET_TYPE(child, &cardtype);
PCIC_MASK(sc, PCIC_INTR, | ((cardtype == PCCARD_IFTYPE_IO) ?
PCIC_INTR_CARDTYPE_IO :
PCIC_INTR_CARDTYPE_MEM));
DEVPRINTF((sc->sc_dev, "card type is %s\n",
(cardtype == PCCARD_IFTYPE_IO) ? "io" : "mem"));
}
/* reinstall all the memory and io mappings */
{
int win;
for (win = 0; win < PCIC_MEM_WINS; ++win) {
if (sc->memalloc & (1 << win)) {
pccbb_pcic_do_mem_map(sc, win);
}
}
for (win = 0; win < PCIC_IO_WINS; ++win) {
if (sc->ioalloc & (1 << win)) {
pccbb_pcic_do_io_map(sc, win);
}
}
}
return 0;
}
static void
pccbb_pcic_power_disable_socket(device_t self, device_t child)
{
struct pccbb_softc *sc = device_get_softc(self);
DPRINTF(("pccbb_pcic_socket_disable\n"));
/* reset signal asserting... */
PCIC_MASK(sc, PCIC_INTR, & ~PCIC_INTR_RESET);
DELAY(2*1000);
/* power down the socket */
PCIC_MASK(sc, PCIC_PWRCTL, &~PCIC_PWRCTL_OE);
pccbb_power(self, CARD_VCC_0V | CARD_VPP_0V);
/* wait 300ms until power fails (Tpf). */
DELAY(300 * 1000);
}
/************************************************************************/
/* PC Card Resource Functions */
/************************************************************************/
static void
pccbb_pcic_wait_ready(struct pccbb_softc *sc)
{
int i;
DEVPRINTF((sc->sc_dev, "pccbb_pcic_wait_ready: status 0x%02x\n",
PCIC_READ(sc, PCIC_IF_STATUS)));
for (i = 0; i < 10000; i++) {
if (PCIC_READ(sc, PCIC_IF_STATUS) & PCIC_IF_STATUS_READY) {
return;
}
DELAY(500);
}
device_printf(sc->sc_dev, "ready never happened, status = %02x\n",
PCIC_READ(sc, PCIC_IF_STATUS));
}
#define PCIC_MEMINFO(NUM) { \
PCIC_SYSMEM_ADDR ## NUM ## _START_LSB, \
PCIC_SYSMEM_ADDR ## NUM ## _START_MSB, \
PCIC_SYSMEM_ADDR ## NUM ## _STOP_LSB, \
PCIC_SYSMEM_ADDR ## NUM ## _STOP_MSB, \
PCIC_SYSMEM_ADDR ## NUM ## _WIN, \
PCIC_CARDMEM_ADDR ## NUM ## _LSB, \
PCIC_CARDMEM_ADDR ## NUM ## _MSB, \
PCIC_ADDRWIN_ENABLE_MEM ## NUM ## , \
}
static struct mem_map_index_st {
int sysmem_start_lsb;
int sysmem_start_msb;
int sysmem_stop_lsb;
int sysmem_stop_msb;
int sysmem_win;
int cardmem_lsb;
int cardmem_msb;
int memenable;
} mem_map_index[] = {
PCIC_MEMINFO(0),
PCIC_MEMINFO(1),
PCIC_MEMINFO(2),
PCIC_MEMINFO(3),
PCIC_MEMINFO(4),
};
#undef PCIC_MEMINFO
static void
pccbb_pcic_do_mem_map(struct pccbb_softc *sc, int win)
{
PCIC_WRITE(sc, mem_map_index[win].sysmem_start_lsb,
(sc->mem[win].addr >> PCIC_SYSMEM_ADDRX_SHIFT) & 0xff);
PCIC_WRITE(sc, mem_map_index[win].sysmem_start_msb,
((sc->mem[win].addr >> (PCIC_SYSMEM_ADDRX_SHIFT + 8)) &
PCIC_SYSMEM_ADDRX_START_MSB_ADDR_MASK) | 0x80);
PCIC_WRITE(sc, mem_map_index[win].sysmem_stop_lsb,
((sc->mem[win].addr + sc->mem[win].realsize - 1) >>
PCIC_SYSMEM_ADDRX_SHIFT) & 0xff);
PCIC_WRITE(sc, mem_map_index[win].sysmem_stop_msb,
(((sc->mem[win].addr + sc->mem[win].realsize - 1) >>
(PCIC_SYSMEM_ADDRX_SHIFT + 8)) &
PCIC_SYSMEM_ADDRX_STOP_MSB_ADDR_MASK) |
PCIC_SYSMEM_ADDRX_STOP_MSB_WAIT2);
PCIC_WRITE(sc, mem_map_index[win].sysmem_win,
(sc->mem[win].addr >> PCIC_MEMREG_WIN_SHIFT) & 0xff);
PCIC_WRITE(sc, mem_map_index[win].cardmem_lsb,
(sc->mem[win].offset >> PCIC_CARDMEM_ADDRX_SHIFT) & 0xff);
PCIC_WRITE(sc, mem_map_index[win].cardmem_msb,
((sc->mem[win].offset >> (PCIC_CARDMEM_ADDRX_SHIFT + 8)) &
PCIC_CARDMEM_ADDRX_MSB_ADDR_MASK) |
((sc->mem[win].kind == PCCARD_MEM_ATTR) ?
PCIC_CARDMEM_ADDRX_MSB_REGACTIVE_ATTR : 0));
PCIC_MASK(sc, PCIC_ADDRWIN_ENABLE, | PCIC_ADDRWIN_ENABLE_MEMCS16
| mem_map_index[win].memenable);
DELAY(100);
#ifdef CBB_DEBUG
{
int r1, r2, r3, r4, r5, r6, r7;
r1 = PCIC_READ(sc, mem_map_index[win].sysmem_start_msb);
r2 = PCIC_READ(sc, mem_map_index[win].sysmem_start_lsb);
r3 = PCIC_READ(sc, mem_map_index[win].sysmem_stop_msb);
r4 = PCIC_READ(sc, mem_map_index[win].sysmem_stop_lsb);
r5 = PCIC_READ(sc, mem_map_index[win].cardmem_msb);
r6 = PCIC_READ(sc, mem_map_index[win].cardmem_lsb);
r7 = PCIC_READ(sc, mem_map_index[win].sysmem_win);
DPRINTF(("pccbb_pcic_do_mem_map window %d: %02x%02x %02x%02x "
"%02x%02x %02x (%08x+%08x.%08x*%08lx)\n", win, r1, r2, r3, r4, r5, r6, r7,
sc->mem[win].addr, sc->mem[win].size, sc->mem[win].realsize, sc->mem[win].offset));
}
#endif
}
static int
pccbb_pcic_mem_map(struct pccbb_softc *sc, int kind,
struct resource *r, bus_addr_t card_addr, int *win)
{
int i;
*win = -1;
for (i = 0; i < PCIC_MEM_WINS; i++) {
if ((sc->memalloc & (1 << i)) == 0) {
*win = i;
sc->memalloc |= (1 << i);
break;
}
}
if (*win == -1)
return (1);
card_addr = card_addr - card_addr % PCIC_MEM_PAGESIZE;
sc->mem[*win].memt = rman_get_bustag(r);
sc->mem[*win].memh = rman_get_bushandle(r);
sc->mem[*win].addr = rman_get_start(r);
sc->mem[*win].size = rman_get_end(r) - sc->mem[*win].addr + 1;
sc->mem[*win].realsize = sc->mem[*win].size + PCIC_MEM_PAGESIZE - 1;
sc->mem[*win].realsize = sc->mem[*win].realsize -
(sc->mem[*win].realsize % PCIC_MEM_PAGESIZE);
sc->mem[*win].offset = ((long)card_addr) -
((long)(sc->mem[*win].addr));
sc->mem[*win].kind = kind;
DPRINTF(("pccbb_pcic_mem_map window %d bus %x+%x+%lx card addr %x\n",
*win, sc->mem[*win].addr, sc->mem[*win].size,
sc->mem[*win].offset, card_addr));
pccbb_pcic_do_mem_map(sc, *win);
return (0);
}
static void
pccbb_pcic_mem_unmap(struct pccbb_softc *sc, int window)
{
if (window >= PCIC_MEM_WINS)
panic("pccbb_pcic_mem_unmap: window out of range");
PCIC_MASK(sc, PCIC_ADDRWIN_ENABLE, & ~mem_map_index[window].memenable);
sc->memalloc &= ~(1 << window);
}
#define PCIC_IOINFO(NUM) { \
PCIC_IOADDR ## NUM ## _START_LSB, \
PCIC_IOADDR ## NUM ## _START_MSB, \
PCIC_IOADDR ## NUM ## _STOP_LSB, \
PCIC_IOADDR ## NUM ## _STOP_MSB, \
PCIC_ADDRWIN_ENABLE_IO ## NUM ## , \
PCIC_IOCTL_IO ## NUM ## _WAITSTATE \
| PCIC_IOCTL_IO ## NUM ## _ZEROWAIT \
| PCIC_IOCTL_IO ## NUM ## _IOCS16SRC_MASK \
| PCIC_IOCTL_IO ## NUM ## _DATASIZE_MASK, \
{ \
PCIC_IOCTL_IO ## NUM ## _IOCS16SRC_CARD, \
PCIC_IOCTL_IO ## NUM ## _IOCS16SRC_DATASIZE \
| PCIC_IOCTL_IO ## NUM ## _DATASIZE_8BIT, \
PCIC_IOCTL_IO ## NUM ## _IOCS16SRC_DATASIZE \
| PCIC_IOCTL_IO ## NUM ## _DATASIZE_16BIT, \
} \
}
static struct io_map_index_st {
int start_lsb;
int start_msb;
int stop_lsb;
int stop_msb;
int ioenable;
int ioctlmask;
int ioctlbits[3]; /* indexed by PCCARD_WIDTH_* */
} io_map_index[] = {
PCIC_IOINFO(0),
PCIC_IOINFO(1),
};
#undef PCIC_IOINFO
static void pccbb_pcic_do_io_map(struct pccbb_softc *sc, int win)
{
PCIC_WRITE(sc, io_map_index[win].start_lsb, sc->io[win].addr & 0xff);
PCIC_WRITE(sc, io_map_index[win].start_msb,
(sc->io[win].addr >> 8) & 0xff);
PCIC_WRITE(sc, io_map_index[win].stop_lsb,
(sc->io[win].addr + sc->io[win].size - 1) & 0xff);
PCIC_WRITE(sc, io_map_index[win].stop_msb,
((sc->io[win].addr + sc->io[win].size - 1) >> 8) & 0xff);
PCIC_MASK2(sc, PCIC_IOCTL,
& ~io_map_index[win].ioctlmask,
| io_map_index[win].ioctlbits[sc->io[win].width]);
PCIC_MASK(sc, PCIC_ADDRWIN_ENABLE, | io_map_index[win].ioenable);
#ifdef CBB_DEBUG
{
int r1, r2, r3, r4;
r1 = PCIC_READ(sc, io_map_index[win].start_msb);
r2 = PCIC_READ(sc, io_map_index[win].start_lsb);
r3 = PCIC_READ(sc, io_map_index[win].stop_msb);
r4 = PCIC_READ(sc, io_map_index[win].stop_lsb);
DPRINTF(("pccbb_pcic_do_io_map window %d: %02x%02x %02x%02x "
"(%08x+%08x)\n", win, r1, r2, r3, r4,
sc->io[win].addr, sc->io[win].size));
}
#endif
}
static int
pccbb_pcic_io_map(struct pccbb_softc *sc, int width,
struct resource *r, bus_addr_t card_addr, int *win)
{
int i;
#ifdef CBB_DEBUG
static char *width_names[] = { "auto", "io8", "io16"};
#endif
*win = -1;
for (i=0; i < PCIC_IO_WINS; i++) {
if ((sc->ioalloc & (1 << i)) == 0) {
*win = i;
sc->ioalloc |= (1 << i);
break;
}
}
if (*win == -1)
return (1);
sc->io[*win].iot = rman_get_bustag(r);
sc->io[*win].ioh = rman_get_bushandle(r);
sc->io[*win].addr = rman_get_start(r);
sc->io[*win].size = rman_get_end(r) - sc->io[*win].addr + 1;
sc->io[*win].flags = 0;
sc->io[*win].width = width;
DPRINTF(("pccbb_pcic_io_map window %d %s port %x+%x\n",
*win, width_names[width], sc->io[*win].addr,
sc->io[*win].size));
pccbb_pcic_do_io_map(sc, *win);
return (0);
}
static void
pccbb_pcic_io_unmap(struct pccbb_softc *sc, int window)
{
if (window >= PCIC_IO_WINS)
panic("pccbb_pcic_io_unmap: window out of range");
PCIC_MASK(sc, PCIC_ADDRWIN_ENABLE, & ~io_map_index[window].ioenable);
sc->ioalloc &= ~(1 << window);
sc->io[window].iot = 0;
sc->io[window].ioh = 0;
sc->io[window].addr = 0;
sc->io[window].size = 0;
sc->io[window].flags = 0;
sc->io[window].width = 0;
}
static int
pccbb_pcic_activate_resource(device_t self, device_t child, int type, int rid,
struct resource *r)
{
int err;
int win;
struct pccbb_reslist *rle;
struct pccbb_softc *sc = device_get_softc(self);
if (rman_get_flags(r) & RF_ACTIVE)
return 0;
switch (type) {
case SYS_RES_IOPORT:
err = pccbb_pcic_io_map(sc, 0, r, 0, &win);
if (err)
return err;
break;
case SYS_RES_MEMORY:
err = pccbb_pcic_mem_map(sc, 0, r, 0, &win);
if (err)
return err;
break;
default:
break;
}
SLIST_FOREACH(rle, &sc->rl, entries) {
if (type == rle->type && rid == rle->rid &&
child == rle->odev) {
rle->win = win;
break;
}
}
err = bus_generic_activate_resource(self, child, type, rid, r);
return (err);
}
static int
pccbb_pcic_deactivate_resource(device_t self, device_t child, int type,
int rid, struct resource *r)
{
struct pccbb_softc *sc = device_get_softc(self);
int win;
struct pccbb_reslist *rle;
win = -1;
SLIST_FOREACH(rle, &sc->rl, entries) {
if (type == rle->type && rid == rle->rid &&
child == rle->odev) {
win = rle->win;
break;
}
}
if (win == -1) {
panic("pccbb_pcic: deactivating bogus resoure");
return 1;
}
switch (type) {
case SYS_RES_IOPORT:
pccbb_pcic_io_unmap(sc, win);
break;
case SYS_RES_MEMORY:
pccbb_pcic_mem_unmap(sc, win);
break;
default:
break;
}
return bus_generic_deactivate_resource(self, child, type, rid, r);
}
static struct resource*
pccbb_pcic_alloc_resource(device_t self, device_t child, int type, int* rid,
u_long start, u_long end, u_long count, u_int flags)
{
struct resource *r = NULL;
struct pccbb_softc *sc = device_get_softc(self);
struct pccbb_reslist *rle;
if ((sc->sc_flags & PCCBB_PCIC_MEM_32) == 0) {
panic("PCCBB bridge cannot handle non MEM_32 bridges\n");
}
switch (type) {
case SYS_RES_MEMORY:
/* Nearly default */
if (start == 0 && end == ~0 && count != 1) {
start = CARDBUS_SYS_RES_MEMORY_START; /* XXX -- should be tweakable*/
end = CARDBUS_SYS_RES_MEMORY_END;
}
flags = (flags & ~RF_ALIGNMENT_MASK)
| rman_make_alignment_flags(PCCBB_MEMALIGN);
break;
case SYS_RES_IOPORT:
if (start < 0x100)
start = 0x100; /* XXX tweakable? */
if (end < start)
end = start;
break;
case SYS_RES_IRQ:
flags |= RF_SHAREABLE;
start = end = rman_get_start(sc->sc_irq_res);
break;
}
r = bus_generic_alloc_resource(self, child, type, rid, start, end,
count, flags & ~RF_ACTIVE);
if (r == NULL)
return NULL;
rle = malloc(sizeof(struct pccbb_reslist), M_DEVBUF, M_WAITOK);
rle->type = type;
rle->rid = *rid;
rle->start = rman_get_start(r);
rle->end = rman_get_end(r);
rle->odev = child;
rle->win = -1;
SLIST_INSERT_HEAD(&sc->rl, rle, entries);
if (flags & RF_ACTIVE) {
if (bus_activate_resource(child, type, *rid, r) != 0) {
BUS_RELEASE_RESOURCE(self, child, type, *rid, r);
return NULL;
}
}
return r;
}
static int
pccbb_pcic_release_resource(device_t self, device_t child, int type,
int rid, struct resource *res)
{
struct pccbb_softc *sc = device_get_softc(self);
struct pccbb_reslist *rle;
int count = 0;
if (rman_get_flags(res) & RF_ACTIVE) {
int error;
error = bus_deactivate_resource(child, type, rid, res);
if (error != 0)
return error;
}
SLIST_FOREACH(rle, &sc->rl, entries) {
if (type == rle->type && rid == rle->rid &&
child == rle->odev) {
SLIST_REMOVE(&sc->rl, rle, pccbb_reslist, entries);
free(rle, M_DEVBUF);
count++;
break;
}
}
if (count == 0) {
panic("pccbb_pcic: releasing bogus resource");
}
return bus_generic_release_resource(self, child, type, rid, res);
}
/************************************************************************/
/* PC Card methods */
/************************************************************************/
static int
pccbb_pcic_set_res_flags(device_t self, device_t child, int type, int rid,
u_int32_t flags)
{
struct pccbb_softc *sc = device_get_softc(self);
if (type != SYS_RES_MEMORY)
return (EINVAL);
sc->mem[rid].kind = flags;
pccbb_pcic_do_mem_map(sc, rid);
return 0;
}
static int
pccbb_pcic_set_memory_offset(device_t self, device_t child, int rid,
u_int32_t cardaddr, u_int32_t *deltap)
{
struct pccbb_softc *sc = device_get_softc(self);
int win;
struct pccbb_reslist *rle;
u_int32_t delta;
win = -1;
SLIST_FOREACH(rle, &sc->rl, entries) {
if (SYS_RES_MEMORY == rle->type && rid == rle->rid &&
child == rle->odev) {
win = rle->win;
break;
}
}
if (win == -1) {
panic("pccbb_pcic: setting memory offset of bogus resource");
return 1;
}
delta = cardaddr % PCIC_MEM_PAGESIZE;
if (deltap)
*deltap = delta;
cardaddr -= delta;
sc->mem[win].realsize = sc->mem[win].size + delta +
PCIC_MEM_PAGESIZE - 1;
sc->mem[win].realsize = sc->mem[win].realsize -
(sc->mem[win].realsize % PCIC_MEM_PAGESIZE);
sc->mem[win].offset = cardaddr - sc->mem[win].addr;
pccbb_pcic_do_mem_map(sc, win);
return 0;
}
/************************************************************************/
/* POWER methods */
/************************************************************************/
static int
pccbb_power_enable_socket(device_t self, device_t child)
{
struct pccbb_softc *sc = device_get_softc(self);
if (sc->sc_flags & PCCBB_16BIT_CARD)
return pccbb_pcic_power_enable_socket(self, child);
else
return pccbb_cardbus_power_enable_socket(self, child);
}
static void
pccbb_power_disable_socket(device_t self, device_t child)
{
struct pccbb_softc *sc = device_get_softc(self);
if (sc->sc_flags & PCCBB_16BIT_CARD)
pccbb_pcic_power_disable_socket(self, child);
else
pccbb_cardbus_power_disable_socket(self, child);
}
/************************************************************************/
/* BUS Methods */
/************************************************************************/
static int
pccbb_activate_resource(device_t self, device_t child, int type, int rid,
struct resource *r)
{
struct pccbb_softc *sc = device_get_softc(self);
if (sc->sc_flags & PCCBB_16BIT_CARD)
return pccbb_pcic_activate_resource(self, child, type, rid, r);
else
return pccbb_cardbus_activate_resource(self, child, type, rid,
r);
}
static int
pccbb_deactivate_resource(device_t self, device_t child, int type,
int rid, struct resource *r)
{
struct pccbb_softc *sc = device_get_softc(self);
if (sc->sc_flags & PCCBB_16BIT_CARD)
return pccbb_pcic_deactivate_resource(self, child, type,
rid, r);
else
return pccbb_cardbus_deactivate_resource(self, child, type,
rid, r);
}
static struct resource*
pccbb_alloc_resource(device_t self, device_t child, int type, int* rid,
u_long start, u_long end, u_long count, u_int flags)
{
struct pccbb_softc *sc = device_get_softc(self);
if (sc->sc_flags & PCCBB_16BIT_CARD)
return pccbb_pcic_alloc_resource(self, child, type, rid,
start, end, count, flags);
else
return pccbb_cardbus_alloc_resource(self, child, type, rid,
start, end, count, flags);
}
static int
pccbb_release_resource(device_t self, device_t child, int type, int rid,
struct resource *r)
{
struct pccbb_softc *sc = device_get_softc(self);
if (sc->sc_flags & PCCBB_16BIT_CARD)
return pccbb_pcic_release_resource(self, child, type,
rid, r);
else
return pccbb_cardbus_release_resource(self, child, type,
rid, r);
}
static int
pccbb_read_ivar(device_t dev, device_t child, int which, uintptr_t *result)
{
struct pccbb_softc *sc = device_get_softc(dev);
switch (which) {
case PCIB_IVAR_BUS:
*result = sc->sc_secbus;
return(0);
}
return(ENOENT);
}
static int
pccbb_write_ivar(device_t dev, device_t child, int which, uintptr_t value)
{
struct pccbb_softc *sc = device_get_softc(dev);
switch (which) {
case PCIB_IVAR_BUS:
sc->sc_secbus = value;
break;
}
return(ENOENT);
}
/************************************************************************/
/* PCI compat methods */
/************************************************************************/
static int
pccbb_maxslots(device_t dev)
{
return 0;
}
static u_int32_t
pccbb_read_config(device_t dev, int b, int s, int f, int reg, int width)
{
/*
* Pass through to the next ppb up the chain (i.e. our grandparent).
*/
return PCIB_READ_CONFIG(device_get_parent(device_get_parent(dev)),
b, s, f, reg, width);
}
static void
pccbb_write_config(device_t dev, int b, int s, int f, int reg, u_int32_t val,
int width)
{
/*
* Pass through to the next ppb up the chain (i.e. our grandparent).
*/
PCIB_WRITE_CONFIG(device_get_parent(device_get_parent(dev)),
b, s, f, reg, val, width);
}
static device_method_t pccbb_methods[] = {
/* Device interface */
DEVMETHOD(device_probe, pccbb_probe),
DEVMETHOD(device_attach, pccbb_attach),
DEVMETHOD(device_detach, pccbb_detach),
DEVMETHOD(device_suspend, bus_generic_suspend),
DEVMETHOD(device_resume, bus_generic_resume),
/* bus methods */
DEVMETHOD(bus_print_child, bus_generic_print_child),
DEVMETHOD(bus_read_ivar, pccbb_read_ivar),
DEVMETHOD(bus_write_ivar, pccbb_write_ivar),
DEVMETHOD(bus_alloc_resource, pccbb_alloc_resource),
DEVMETHOD(bus_release_resource, pccbb_release_resource),
DEVMETHOD(bus_activate_resource, pccbb_activate_resource),
DEVMETHOD(bus_deactivate_resource, pccbb_deactivate_resource),
DEVMETHOD(bus_driver_added, pccbb_driver_added),
DEVMETHOD(bus_child_detached, pccbb_child_detached),
DEVMETHOD(bus_setup_intr, bus_generic_setup_intr),
DEVMETHOD(bus_teardown_intr, bus_generic_teardown_intr),
/* 16-bit card interface */
DEVMETHOD(card_set_res_flags, pccbb_pcic_set_res_flags),
DEVMETHOD(card_set_memory_offset, pccbb_pcic_set_memory_offset),
/* power interface */
DEVMETHOD(power_enable_socket, pccbb_power_enable_socket),
DEVMETHOD(power_disable_socket, pccbb_power_disable_socket),
/* pcib compatibility interface */
DEVMETHOD(pcib_maxslots, pccbb_maxslots),
DEVMETHOD(pcib_read_config, pccbb_read_config),
DEVMETHOD(pcib_write_config, pccbb_write_config),
{0,0}
};
static driver_t pccbb_driver = {
"pccbb",
pccbb_methods,
sizeof(struct pccbb_softc)
};
static devclass_t pccbb_devclass;
DRIVER_MODULE(pccbb, pci, pccbb_driver, pccbb_devclass, 0, 0);
SYSINIT(pccbb, SI_SUB_KTHREAD_IDLE, SI_ORDER_ANY, pccbb_start_threads, 0);