freebsd-nq/sys/dev/ppc/ppc_isa.c

275 lines
6.8 KiB
C
Raw Normal View History

/*-
* Copyright (c) 1997-2000 Nicolas Souchu
* Copyright (c) 2001 Alcove - Nicolas Souchu
* Copyright (c) 2006 Marcel Moolenaar
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
*
* 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.
*/
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/kernel.h>
Add locking to ppc and ppbus and mark the whole lot MPSAFE: - To avoid having a bunch of locks that end up always getting acquired as a group, give each ppc(4) device a mutex which it shares with all the child devices including ppbus(4), lpt(4), plip(4), etc. This mutex is then used for all the locking. - Rework the interrupt handling stuff yet again. Now ppbus drivers setup their interrupt handler during attach and tear it down during detach like most other drivers. ppbus(4) only invokes the interrupt handler of the device that currently owns the bus (if any) when an interrupt occurs, however. Also, interrupt handlers in general now accept their softc pointers as their argument rather than the device_t. Another feature of the ppbus interrupt handlers is that they are called with the parent ppc device's lock already held. This minimizes the number of lock operations during an interrupt. - Mark plip(4), lpt(4), pcfclock(4), ppi(4), vpo(4) MPSAFE. - lpbb(4) uses the ppc lock instead of Giant. - Other plip(4) changes: - Add a mutex to protect the global tables in plip(4) and free them on module unload. - Add a detach routine. - Split out the init/stop code from the ioctl routine into separate functions. - Other lpt(4) changes: - Use device_printf(). - Use a dedicated callout for the lptout timer. - Allocate the I/O buffers at attach and detach rather than during open and close as this simplifies the locking at the cost of 1024+32 bytes when the driver is attached. - Other ppi(4) changes: - Use an sx lock to serialize open and close. - Remove unused HADBUS flag. - Add a detach routine. - Use a malloc'd buffer for each read and write to avoid races with concurrent read/write. - Other pps(4) changes: - Use a callout rather than a callout handle with timeout(). - Conform to the new ppbus requirements (regular mutex, non-filter interrupt handler). pps(4) is probably going to have to become a standalone driver that doesn't use ppbus(4) to satisfy it's requirements for low latency as a result. - Use an sx lock to serialize open and close. - Other vpo(4) changes: - Use the parent ppc device's lock to create the CAM sim instead of Giant. - Other ppc(4) changes: - Fix ppc_isa's detach method to detach instead of calling attach. Tested by: no one :-(
2009-01-21 23:10:06 +00:00
#include <sys/lock.h>
#include <sys/module.h>
Add locking to ppc and ppbus and mark the whole lot MPSAFE: - To avoid having a bunch of locks that end up always getting acquired as a group, give each ppc(4) device a mutex which it shares with all the child devices including ppbus(4), lpt(4), plip(4), etc. This mutex is then used for all the locking. - Rework the interrupt handling stuff yet again. Now ppbus drivers setup their interrupt handler during attach and tear it down during detach like most other drivers. ppbus(4) only invokes the interrupt handler of the device that currently owns the bus (if any) when an interrupt occurs, however. Also, interrupt handlers in general now accept their softc pointers as their argument rather than the device_t. Another feature of the ppbus interrupt handlers is that they are called with the parent ppc device's lock already held. This minimizes the number of lock operations during an interrupt. - Mark plip(4), lpt(4), pcfclock(4), ppi(4), vpo(4) MPSAFE. - lpbb(4) uses the ppc lock instead of Giant. - Other plip(4) changes: - Add a mutex to protect the global tables in plip(4) and free them on module unload. - Add a detach routine. - Split out the init/stop code from the ioctl routine into separate functions. - Other lpt(4) changes: - Use device_printf(). - Use a dedicated callout for the lptout timer. - Allocate the I/O buffers at attach and detach rather than during open and close as this simplifies the locking at the cost of 1024+32 bytes when the driver is attached. - Other ppi(4) changes: - Use an sx lock to serialize open and close. - Remove unused HADBUS flag. - Add a detach routine. - Use a malloc'd buffer for each read and write to avoid races with concurrent read/write. - Other pps(4) changes: - Use a callout rather than a callout handle with timeout(). - Conform to the new ppbus requirements (regular mutex, non-filter interrupt handler). pps(4) is probably going to have to become a standalone driver that doesn't use ppbus(4) to satisfy it's requirements for low latency as a result. - Use an sx lock to serialize open and close. - Other vpo(4) changes: - Use the parent ppc device's lock to create the CAM sim instead of Giant. - Other ppc(4) changes: - Fix ppc_isa's detach method to detach instead of calling attach. Tested by: no one :-(
2009-01-21 23:10:06 +00:00
#include <sys/mutex.h>
#include <sys/bus.h>
#include <machine/bus.h>
#include <sys/malloc.h>
#include <sys/rman.h>
#include <isa/isavar.h>
#include <dev/ppbus/ppbconf.h>
#include <dev/ppbus/ppb_msq.h>
#include <dev/ppc/ppcvar.h>
#include <dev/ppc/ppcreg.h>
#include "ppbus_if.h"
static int ppc_isa_probe(device_t dev);
int ppc_isa_attach(device_t dev);
int ppc_isa_write(device_t, char *, int, int);
static device_method_t ppc_isa_methods[] = {
/* device interface */
DEVMETHOD(device_probe, ppc_isa_probe),
DEVMETHOD(device_attach, ppc_isa_attach),
Add locking to ppc and ppbus and mark the whole lot MPSAFE: - To avoid having a bunch of locks that end up always getting acquired as a group, give each ppc(4) device a mutex which it shares with all the child devices including ppbus(4), lpt(4), plip(4), etc. This mutex is then used for all the locking. - Rework the interrupt handling stuff yet again. Now ppbus drivers setup their interrupt handler during attach and tear it down during detach like most other drivers. ppbus(4) only invokes the interrupt handler of the device that currently owns the bus (if any) when an interrupt occurs, however. Also, interrupt handlers in general now accept their softc pointers as their argument rather than the device_t. Another feature of the ppbus interrupt handlers is that they are called with the parent ppc device's lock already held. This minimizes the number of lock operations during an interrupt. - Mark plip(4), lpt(4), pcfclock(4), ppi(4), vpo(4) MPSAFE. - lpbb(4) uses the ppc lock instead of Giant. - Other plip(4) changes: - Add a mutex to protect the global tables in plip(4) and free them on module unload. - Add a detach routine. - Split out the init/stop code from the ioctl routine into separate functions. - Other lpt(4) changes: - Use device_printf(). - Use a dedicated callout for the lptout timer. - Allocate the I/O buffers at attach and detach rather than during open and close as this simplifies the locking at the cost of 1024+32 bytes when the driver is attached. - Other ppi(4) changes: - Use an sx lock to serialize open and close. - Remove unused HADBUS flag. - Add a detach routine. - Use a malloc'd buffer for each read and write to avoid races with concurrent read/write. - Other pps(4) changes: - Use a callout rather than a callout handle with timeout(). - Conform to the new ppbus requirements (regular mutex, non-filter interrupt handler). pps(4) is probably going to have to become a standalone driver that doesn't use ppbus(4) to satisfy it's requirements for low latency as a result. - Use an sx lock to serialize open and close. - Other vpo(4) changes: - Use the parent ppc device's lock to create the CAM sim instead of Giant. - Other ppc(4) changes: - Fix ppc_isa's detach method to detach instead of calling attach. Tested by: no one :-(
2009-01-21 23:10:06 +00:00
DEVMETHOD(device_detach, ppc_detach),
/* bus interface */
DEVMETHOD(bus_read_ivar, ppc_read_ivar),
Add locking to ppc and ppbus and mark the whole lot MPSAFE: - To avoid having a bunch of locks that end up always getting acquired as a group, give each ppc(4) device a mutex which it shares with all the child devices including ppbus(4), lpt(4), plip(4), etc. This mutex is then used for all the locking. - Rework the interrupt handling stuff yet again. Now ppbus drivers setup their interrupt handler during attach and tear it down during detach like most other drivers. ppbus(4) only invokes the interrupt handler of the device that currently owns the bus (if any) when an interrupt occurs, however. Also, interrupt handlers in general now accept their softc pointers as their argument rather than the device_t. Another feature of the ppbus interrupt handlers is that they are called with the parent ppc device's lock already held. This minimizes the number of lock operations during an interrupt. - Mark plip(4), lpt(4), pcfclock(4), ppi(4), vpo(4) MPSAFE. - lpbb(4) uses the ppc lock instead of Giant. - Other plip(4) changes: - Add a mutex to protect the global tables in plip(4) and free them on module unload. - Add a detach routine. - Split out the init/stop code from the ioctl routine into separate functions. - Other lpt(4) changes: - Use device_printf(). - Use a dedicated callout for the lptout timer. - Allocate the I/O buffers at attach and detach rather than during open and close as this simplifies the locking at the cost of 1024+32 bytes when the driver is attached. - Other ppi(4) changes: - Use an sx lock to serialize open and close. - Remove unused HADBUS flag. - Add a detach routine. - Use a malloc'd buffer for each read and write to avoid races with concurrent read/write. - Other pps(4) changes: - Use a callout rather than a callout handle with timeout(). - Conform to the new ppbus requirements (regular mutex, non-filter interrupt handler). pps(4) is probably going to have to become a standalone driver that doesn't use ppbus(4) to satisfy it's requirements for low latency as a result. - Use an sx lock to serialize open and close. - Other vpo(4) changes: - Use the parent ppc device's lock to create the CAM sim instead of Giant. - Other ppc(4) changes: - Fix ppc_isa's detach method to detach instead of calling attach. Tested by: no one :-(
2009-01-21 23:10:06 +00:00
DEVMETHOD(bus_write_ivar, ppc_write_ivar),
DEVMETHOD(bus_alloc_resource, ppc_alloc_resource),
DEVMETHOD(bus_release_resource, ppc_release_resource),
/* ppbus interface */
DEVMETHOD(ppbus_io, ppc_io),
DEVMETHOD(ppbus_exec_microseq, ppc_exec_microseq),
DEVMETHOD(ppbus_reset_epp, ppc_reset_epp),
DEVMETHOD(ppbus_setmode, ppc_setmode),
DEVMETHOD(ppbus_ecp_sync, ppc_ecp_sync),
DEVMETHOD(ppbus_read, ppc_read),
DEVMETHOD(ppbus_write, ppc_isa_write),
{ 0, 0 }
};
static driver_t ppc_isa_driver = {
ppc_driver_name,
ppc_isa_methods,
sizeof(struct ppc_data),
};
static struct isa_pnp_id lpc_ids[] = {
{ 0x0004d041, "Standard parallel printer port" }, /* PNP0400 */
{ 0x0104d041, "ECP parallel printer port" }, /* PNP0401 */
{ 0 }
};
static void
ppc_isa_dmadone(struct ppc_data *ppc)
{
isa_dmadone(ppc->ppc_dmaflags, ppc->ppc_dmaddr, ppc->ppc_dmacnt,
ppc->ppc_dmachan);
}
int
ppc_isa_attach(device_t dev)
{
struct ppc_data *ppc = device_get_softc(dev);
if ((ppc->ppc_avm & PPB_ECP) && (ppc->ppc_dmachan > 0)) {
/* acquire the DMA channel forever */ /* XXX */
isa_dma_acquire(ppc->ppc_dmachan);
isa_dmainit(ppc->ppc_dmachan, 1024); /* nlpt.BUFSIZE */
ppc->ppc_dmadone = ppc_isa_dmadone;
}
return (ppc_attach(dev));
}
static int
ppc_isa_probe(device_t dev)
{
device_t parent;
int error;
parent = device_get_parent(dev);
error = ISA_PNP_PROBE(parent, dev, lpc_ids);
if (error == ENXIO)
return (ENXIO);
if (error != 0) /* XXX shall be set after detection */
device_set_desc(dev, "Parallel port");
return (ppc_probe(dev, 0));
}
/*
* Call this function if you want to send data in any advanced mode
* of your parallel port: FIFO, DMA
*
* If what you want is not possible (no ECP, no DMA...),
* EINVAL is returned
*/
int
ppc_isa_write(device_t dev, char *buf, int len, int how)
{
struct ppc_data *ppc = device_get_softc(dev);
char ecr, ecr_sav, ctr, ctr_sav;
int error = 0;
int spin;
Add locking to ppc and ppbus and mark the whole lot MPSAFE: - To avoid having a bunch of locks that end up always getting acquired as a group, give each ppc(4) device a mutex which it shares with all the child devices including ppbus(4), lpt(4), plip(4), etc. This mutex is then used for all the locking. - Rework the interrupt handling stuff yet again. Now ppbus drivers setup their interrupt handler during attach and tear it down during detach like most other drivers. ppbus(4) only invokes the interrupt handler of the device that currently owns the bus (if any) when an interrupt occurs, however. Also, interrupt handlers in general now accept their softc pointers as their argument rather than the device_t. Another feature of the ppbus interrupt handlers is that they are called with the parent ppc device's lock already held. This minimizes the number of lock operations during an interrupt. - Mark plip(4), lpt(4), pcfclock(4), ppi(4), vpo(4) MPSAFE. - lpbb(4) uses the ppc lock instead of Giant. - Other plip(4) changes: - Add a mutex to protect the global tables in plip(4) and free them on module unload. - Add a detach routine. - Split out the init/stop code from the ioctl routine into separate functions. - Other lpt(4) changes: - Use device_printf(). - Use a dedicated callout for the lptout timer. - Allocate the I/O buffers at attach and detach rather than during open and close as this simplifies the locking at the cost of 1024+32 bytes when the driver is attached. - Other ppi(4) changes: - Use an sx lock to serialize open and close. - Remove unused HADBUS flag. - Add a detach routine. - Use a malloc'd buffer for each read and write to avoid races with concurrent read/write. - Other pps(4) changes: - Use a callout rather than a callout handle with timeout(). - Conform to the new ppbus requirements (regular mutex, non-filter interrupt handler). pps(4) is probably going to have to become a standalone driver that doesn't use ppbus(4) to satisfy it's requirements for low latency as a result. - Use an sx lock to serialize open and close. - Other vpo(4) changes: - Use the parent ppc device's lock to create the CAM sim instead of Giant. - Other ppc(4) changes: - Fix ppc_isa's detach method to detach instead of calling attach. Tested by: no one :-(
2009-01-21 23:10:06 +00:00
PPC_ASSERT_LOCKED(ppc);
if (!(ppc->ppc_avm & PPB_ECP))
return (EINVAL);
if (ppc->ppc_dmachan == 0)
return (EINVAL);
#ifdef PPC_DEBUG
printf("w");
#endif
ecr_sav = r_ecr(ppc);
ctr_sav = r_ctr(ppc);
/*
* Send buffer with DMA, FIFO and interrupts
*/
/* byte mode, no intr, no DMA, dir=0, flush fifo */
ecr = PPC_ECR_STD | PPC_DISABLE_INTR;
w_ecr(ppc, ecr);
/* disable nAck interrupts */
ctr = r_ctr(ppc);
ctr &= ~IRQENABLE;
w_ctr(ppc, ctr);
ppc->ppc_dmaflags = 0;
ppc->ppc_dmaddr = (caddr_t)buf;
ppc->ppc_dmacnt = (u_int)len;
switch (ppc->ppc_mode) {
case PPB_COMPATIBLE:
/* compatible mode with FIFO, no intr, DMA, dir=0 */
ecr = PPC_ECR_FIFO | PPC_DISABLE_INTR | PPC_ENABLE_DMA;
break;
case PPB_ECP:
ecr = PPC_ECR_ECP | PPC_DISABLE_INTR | PPC_ENABLE_DMA;
break;
default:
error = EINVAL;
goto error;
}
w_ecr(ppc, ecr);
ecr = r_ecr(ppc);
ppc->ppc_dmastat = PPC_DMA_INIT;
/* enable interrupts */
ecr &= ~PPC_SERVICE_INTR;
ppc->ppc_irqstat = PPC_IRQ_DMA;
w_ecr(ppc, ecr);
isa_dmastart(ppc->ppc_dmaflags, ppc->ppc_dmaddr, ppc->ppc_dmacnt,
ppc->ppc_dmachan);
ppc->ppc_dmastat = PPC_DMA_STARTED;
#ifdef PPC_DEBUG
printf("s%d", ppc->ppc_dmacnt);
#endif
/* Wait for the DMA completed interrupt. We hope we won't
* miss it, otherwise a signal will be necessary to unlock the
* process.
*/
do {
/* release CPU */
Add locking to ppc and ppbus and mark the whole lot MPSAFE: - To avoid having a bunch of locks that end up always getting acquired as a group, give each ppc(4) device a mutex which it shares with all the child devices including ppbus(4), lpt(4), plip(4), etc. This mutex is then used for all the locking. - Rework the interrupt handling stuff yet again. Now ppbus drivers setup their interrupt handler during attach and tear it down during detach like most other drivers. ppbus(4) only invokes the interrupt handler of the device that currently owns the bus (if any) when an interrupt occurs, however. Also, interrupt handlers in general now accept their softc pointers as their argument rather than the device_t. Another feature of the ppbus interrupt handlers is that they are called with the parent ppc device's lock already held. This minimizes the number of lock operations during an interrupt. - Mark plip(4), lpt(4), pcfclock(4), ppi(4), vpo(4) MPSAFE. - lpbb(4) uses the ppc lock instead of Giant. - Other plip(4) changes: - Add a mutex to protect the global tables in plip(4) and free them on module unload. - Add a detach routine. - Split out the init/stop code from the ioctl routine into separate functions. - Other lpt(4) changes: - Use device_printf(). - Use a dedicated callout for the lptout timer. - Allocate the I/O buffers at attach and detach rather than during open and close as this simplifies the locking at the cost of 1024+32 bytes when the driver is attached. - Other ppi(4) changes: - Use an sx lock to serialize open and close. - Remove unused HADBUS flag. - Add a detach routine. - Use a malloc'd buffer for each read and write to avoid races with concurrent read/write. - Other pps(4) changes: - Use a callout rather than a callout handle with timeout(). - Conform to the new ppbus requirements (regular mutex, non-filter interrupt handler). pps(4) is probably going to have to become a standalone driver that doesn't use ppbus(4) to satisfy it's requirements for low latency as a result. - Use an sx lock to serialize open and close. - Other vpo(4) changes: - Use the parent ppc device's lock to create the CAM sim instead of Giant. - Other ppc(4) changes: - Fix ppc_isa's detach method to detach instead of calling attach. Tested by: no one :-(
2009-01-21 23:10:06 +00:00
error = mtx_sleep(ppc, &ppc->ppc_lock, PPBPRI | PCATCH,
"ppcdma", 0);
} while (error == EWOULDBLOCK);
if (error) {
#ifdef PPC_DEBUG
printf("i");
#endif
/* stop DMA */
isa_dmadone(ppc->ppc_dmaflags, ppc->ppc_dmaddr,
ppc->ppc_dmacnt, ppc->ppc_dmachan);
/* no dma, no interrupt, flush the fifo */
w_ecr(ppc, PPC_ECR_RESET);
ppc->ppc_dmastat = PPC_DMA_INTERRUPTED;
goto error;
}
/* wait for an empty fifo */
while (!(r_ecr(ppc) & PPC_FIFO_EMPTY)) {
for (spin=100; spin; spin--)
if (r_ecr(ppc) & PPC_FIFO_EMPTY)
goto fifo_empty;
#ifdef PPC_DEBUG
printf("Z");
#endif
Add locking to ppc and ppbus and mark the whole lot MPSAFE: - To avoid having a bunch of locks that end up always getting acquired as a group, give each ppc(4) device a mutex which it shares with all the child devices including ppbus(4), lpt(4), plip(4), etc. This mutex is then used for all the locking. - Rework the interrupt handling stuff yet again. Now ppbus drivers setup their interrupt handler during attach and tear it down during detach like most other drivers. ppbus(4) only invokes the interrupt handler of the device that currently owns the bus (if any) when an interrupt occurs, however. Also, interrupt handlers in general now accept their softc pointers as their argument rather than the device_t. Another feature of the ppbus interrupt handlers is that they are called with the parent ppc device's lock already held. This minimizes the number of lock operations during an interrupt. - Mark plip(4), lpt(4), pcfclock(4), ppi(4), vpo(4) MPSAFE. - lpbb(4) uses the ppc lock instead of Giant. - Other plip(4) changes: - Add a mutex to protect the global tables in plip(4) and free them on module unload. - Add a detach routine. - Split out the init/stop code from the ioctl routine into separate functions. - Other lpt(4) changes: - Use device_printf(). - Use a dedicated callout for the lptout timer. - Allocate the I/O buffers at attach and detach rather than during open and close as this simplifies the locking at the cost of 1024+32 bytes when the driver is attached. - Other ppi(4) changes: - Use an sx lock to serialize open and close. - Remove unused HADBUS flag. - Add a detach routine. - Use a malloc'd buffer for each read and write to avoid races with concurrent read/write. - Other pps(4) changes: - Use a callout rather than a callout handle with timeout(). - Conform to the new ppbus requirements (regular mutex, non-filter interrupt handler). pps(4) is probably going to have to become a standalone driver that doesn't use ppbus(4) to satisfy it's requirements for low latency as a result. - Use an sx lock to serialize open and close. - Other vpo(4) changes: - Use the parent ppc device's lock to create the CAM sim instead of Giant. - Other ppc(4) changes: - Fix ppc_isa's detach method to detach instead of calling attach. Tested by: no one :-(
2009-01-21 23:10:06 +00:00
error = mtx_sleep(ppc, &ppc->ppc_lock, PPBPRI | PCATCH,
"ppcfifo", hz / 100);
if (error != EWOULDBLOCK) {
#ifdef PPC_DEBUG
printf("I");
#endif
/* no dma, no interrupt, flush the fifo */
w_ecr(ppc, PPC_ECR_RESET);
ppc->ppc_dmastat = PPC_DMA_INTERRUPTED;
error = EINTR;
goto error;
}
}
fifo_empty:
/* no dma, no interrupt, flush the fifo */
w_ecr(ppc, PPC_ECR_RESET);
error:
/* PDRQ must be kept unasserted until nPDACK is
* deasserted for a minimum of 350ns (SMC datasheet)
*
* Consequence may be a FIFO that never empty
*/
DELAY(1);
w_ecr(ppc, ecr_sav);
w_ctr(ppc, ctr_sav);
return (error);
}
DRIVER_MODULE(ppc, isa, ppc_isa_driver, ppc_devclass, 0, 0);