linux/drivers/pcmcia/soc_common.c

894 lines
23 KiB
C
Raw Normal View History

/*======================================================================
Common support code for the PCMCIA control functionality of
integrated SOCs like the SA-11x0 and PXA2xx microprocessors.
The contents of this file are subject to the Mozilla Public
License Version 1.1 (the "License"); you may not use this file
except in compliance with the License. You may obtain a copy of
the License at http://www.mozilla.org/MPL/
Software distributed under the License is distributed on an "AS
IS" basis, WITHOUT WARRANTY OF ANY KIND, either express or
implied. See the License for the specific language governing
rights and limitations under the License.
The initial developer of the original code is John G. Dorsey
<john+@cs.cmu.edu>. Portions created by John G. Dorsey are
Copyright (C) 1999 John G. Dorsey. All Rights Reserved.
Alternatively, the contents of this file may be used under the
terms of the GNU Public License version 2 (the "GPL"), in which
case the provisions of the GPL are applicable instead of the
above. If you wish to allow the use of your version of this file
only under the terms of the GPL and not to allow others to use
your version of this file under the MPL, indicate your decision
by deleting the provisions above and replace them with the notice
and other provisions required by the GPL. If you do not delete
the provisions above, a recipient may use your version of this
file under either the MPL or the GPL.
======================================================================*/
#include <linux/cpufreq.h>
#include <linux/gpio.h>
#include <linux/gpio/consumer.h>
#include <linux/init.h>
#include <linux/interrupt.h>
#include <linux/io.h>
#include <linux/irq.h>
#include <linux/kernel.h>
#include <linux/mm.h>
#include <linux/module.h>
#include <linux/moduleparam.h>
#include <linux/mutex.h>
#include <linux/regulator/consumer.h>
#include <linux/spinlock.h>
#include <linux/timer.h>
ARM: pxa/sa1100: move I/O space to PCI_IOBASE PXA and StrongARM1100 traditionally map their I/O space 1:1 into virtual memory, using a per-bus io_offset that matches the base address of the ioremap mapping. In order for PXA to work in a multiplatform config, this needs to change so I/O space starts at PCI_IOBASE (0xfee00000). Since the pcmcia soc_common support is shared with StrongARM1100, both have to change at the same time. The affected machines are: - Anything with a PCMCIA slot now uses pci_remap_iospace, which is made available to PCMCIA configurations as well, rather than just PCI. The first PCMCIA slot now starts at port number 0x10000. - The Zeus and Viper platforms have PC/104-style ISA buses, which have a static mapping for both I/O and memory space at 0xf1000000, which can no longer work. It does not appear to have any in-tree users, so moving it to port number 0 makes them behave like a traditional PC. - SA1100 does support ISA slots in theory, but all machines that originally enabled this appear to have been removed from the tree ages ago, and the I/O space is never mapped anywhere. - The Nanoengine machine has support for PCI slots, but looks like this never included I/O space, the resources only define the location for memory and config space. With this, the definitions of __io() and IO_SPACE_LIMIT can be simplified, as the only remaining cases are the generic PCI_IOBASE and the custom inb()/outb() macros on RiscPC. S3C24xx still has a custom inb()/outb() in this here, but this is already removed in another branch. Signed-off-by: Arnd Bergmann <arnd@arndb.de>
2022-04-24 20:36:35 +08:00
#include <linux/pci.h>
#include "soc_common.h"
static irqreturn_t soc_common_pcmcia_interrupt(int irq, void *dev);
#ifdef CONFIG_PCMCIA_DEBUG
static int pc_debug;
module_param(pc_debug, int, 0644);
void soc_pcmcia_debug(struct soc_pcmcia_socket *skt, const char *func,
int lvl, const char *fmt, ...)
{
struct va_format vaf;
va_list args;
if (pc_debug > lvl) {
va_start(args, fmt);
vaf.fmt = fmt;
vaf.va = &args;
printk(KERN_DEBUG "skt%u: %s: %pV", skt->nr, func, &vaf);
va_end(args);
}
}
EXPORT_SYMBOL(soc_pcmcia_debug);
#endif
#define to_soc_pcmcia_socket(x) \
container_of(x, struct soc_pcmcia_socket, socket)
int soc_pcmcia_regulator_set(struct soc_pcmcia_socket *skt,
struct soc_pcmcia_regulator *r, int v)
{
bool on;
int ret;
if (!r->reg)
return 0;
on = v != 0;
if (r->on == on)
return 0;
if (on) {
ret = regulator_set_voltage(r->reg, v * 100000, v * 100000);
if (ret) {
int vout = regulator_get_voltage(r->reg) / 100000;
dev_warn(&skt->socket.dev,
"CS requested %s=%u.%uV, applying %u.%uV\n",
r == &skt->vcc ? "Vcc" : "Vpp",
v / 10, v % 10, vout / 10, vout % 10);
}
ret = regulator_enable(r->reg);
} else {
ret = regulator_disable(r->reg);
}
if (ret == 0)
r->on = on;
return ret;
}
EXPORT_SYMBOL_GPL(soc_pcmcia_regulator_set);
static unsigned short
calc_speed(unsigned short *spds, int num, unsigned short dflt)
{
unsigned short speed = 0;
int i;
for (i = 0; i < num; i++)
if (speed < spds[i])
speed = spds[i];
if (speed == 0)
speed = dflt;
return speed;
}
void soc_common_pcmcia_get_timing(struct soc_pcmcia_socket *skt,
struct soc_pcmcia_timing *timing)
{
timing->io =
calc_speed(skt->spd_io, MAX_IO_WIN, SOC_PCMCIA_IO_ACCESS);
timing->mem =
calc_speed(skt->spd_mem, MAX_WIN, SOC_PCMCIA_3V_MEM_ACCESS);
timing->attr =
calc_speed(skt->spd_attr, MAX_WIN, SOC_PCMCIA_3V_MEM_ACCESS);
}
EXPORT_SYMBOL(soc_common_pcmcia_get_timing);
static void __soc_pcmcia_hw_shutdown(struct soc_pcmcia_socket *skt,
unsigned int nr)
{
unsigned int i;
for (i = 0; i < nr; i++)
if (skt->stat[i].irq)
free_irq(skt->stat[i].irq, skt);
if (skt->ops->hw_shutdown)
skt->ops->hw_shutdown(skt);
clk_disable_unprepare(skt->clk);
}
static void soc_pcmcia_hw_shutdown(struct soc_pcmcia_socket *skt)
{
__soc_pcmcia_hw_shutdown(skt, ARRAY_SIZE(skt->stat));
}
int soc_pcmcia_request_gpiods(struct soc_pcmcia_socket *skt)
{
struct device *dev = skt->socket.dev.parent;
struct gpio_desc *desc;
int i;
for (i = 0; i < ARRAY_SIZE(skt->stat); i++) {
if (!skt->stat[i].name)
continue;
desc = devm_gpiod_get(dev, skt->stat[i].name, GPIOD_IN);
if (IS_ERR(desc)) {
dev_err(dev, "Failed to get GPIO for %s: %ld\n",
skt->stat[i].name, PTR_ERR(desc));
return PTR_ERR(desc);
}
skt->stat[i].desc = desc;
}
return 0;
}
EXPORT_SYMBOL_GPL(soc_pcmcia_request_gpiods);
static int soc_pcmcia_hw_init(struct soc_pcmcia_socket *skt)
{
int ret = 0, i;
ret = clk_prepare_enable(skt->clk);
if (ret)
return ret;
if (skt->ops->hw_init) {
ret = skt->ops->hw_init(skt);
if (ret) {
clk_disable_unprepare(skt->clk);
return ret;
}
}
for (i = 0; i < ARRAY_SIZE(skt->stat); i++) {
if (gpio_is_valid(skt->stat[i].gpio)) {
unsigned long flags = GPIOF_IN;
/* CD is active low by default */
if (i == SOC_STAT_CD)
flags |= GPIOF_ACTIVE_LOW;
ret = devm_gpio_request_one(skt->socket.dev.parent,
skt->stat[i].gpio, flags,
skt->stat[i].name);
if (ret) {
__soc_pcmcia_hw_shutdown(skt, i);
return ret;
}
skt->stat[i].desc = gpio_to_desc(skt->stat[i].gpio);
}
if (i < SOC_STAT_VS1 && skt->stat[i].desc) {
int irq = gpiod_to_irq(skt->stat[i].desc);
if (irq > 0) {
if (i == SOC_STAT_RDY)
skt->socket.pci_irq = irq;
else
skt->stat[i].irq = irq;
}
}
if (skt->stat[i].irq) {
ret = request_irq(skt->stat[i].irq,
soc_common_pcmcia_interrupt,
IRQF_TRIGGER_NONE,
skt->stat[i].name, skt);
if (ret) {
__soc_pcmcia_hw_shutdown(skt, i);
return ret;
}
}
}
return ret;
}
static void soc_pcmcia_hw_enable(struct soc_pcmcia_socket *skt)
{
int i;
for (i = 0; i < ARRAY_SIZE(skt->stat); i++)
if (skt->stat[i].irq) {
irq_set_irq_type(skt->stat[i].irq, IRQ_TYPE_EDGE_RISING);
irq_set_irq_type(skt->stat[i].irq, IRQ_TYPE_EDGE_BOTH);
}
}
static void soc_pcmcia_hw_disable(struct soc_pcmcia_socket *skt)
{
int i;
for (i = 0; i < ARRAY_SIZE(skt->stat); i++)
if (skt->stat[i].irq)
irq_set_irq_type(skt->stat[i].irq, IRQ_TYPE_NONE);
}
/*
* The CF 3.0 specification says that cards tie VS1 to ground and leave
* VS2 open. Many implementations do not wire up the VS signals, so we
* provide hard-coded values as per the CF 3.0 spec.
*/
void soc_common_cf_socket_state(struct soc_pcmcia_socket *skt,
struct pcmcia_state *state)
{
state->vs_3v = 1;
}
EXPORT_SYMBOL_GPL(soc_common_cf_socket_state);
static unsigned int soc_common_pcmcia_skt_state(struct soc_pcmcia_socket *skt)
{
struct pcmcia_state state;
unsigned int stat;
memset(&state, 0, sizeof(struct pcmcia_state));
/* Make battery voltage state report 'good' */
state.bvd1 = 1;
state.bvd2 = 1;
if (skt->stat[SOC_STAT_CD].desc)
state.detect = !!gpiod_get_value(skt->stat[SOC_STAT_CD].desc);
if (skt->stat[SOC_STAT_RDY].desc)
state.ready = !!gpiod_get_value(skt->stat[SOC_STAT_RDY].desc);
if (skt->stat[SOC_STAT_BVD1].desc)
state.bvd1 = !!gpiod_get_value(skt->stat[SOC_STAT_BVD1].desc);
if (skt->stat[SOC_STAT_BVD2].desc)
state.bvd2 = !!gpiod_get_value(skt->stat[SOC_STAT_BVD2].desc);
if (skt->stat[SOC_STAT_VS1].desc)
state.vs_3v = !!gpiod_get_value(skt->stat[SOC_STAT_VS1].desc);
if (skt->stat[SOC_STAT_VS2].desc)
state.vs_Xv = !!gpiod_get_value(skt->stat[SOC_STAT_VS2].desc);
skt->ops->socket_state(skt, &state);
stat = state.detect ? SS_DETECT : 0;
stat |= state.ready ? SS_READY : 0;
stat |= state.wrprot ? SS_WRPROT : 0;
stat |= state.vs_3v ? SS_3VCARD : 0;
stat |= state.vs_Xv ? SS_XVCARD : 0;
/* The power status of individual sockets is not available
* explicitly from the hardware, so we just remember the state
* and regurgitate it upon request:
*/
stat |= skt->cs_state.Vcc ? SS_POWERON : 0;
if (skt->cs_state.flags & SS_IOCARD)
stat |= state.bvd1 ? 0 : SS_STSCHG;
else {
if (state.bvd1 == 0)
stat |= SS_BATDEAD;
else if (state.bvd2 == 0)
stat |= SS_BATWARN;
}
return stat;
}
/*
* soc_common_pcmcia_config_skt
* ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
*
* Convert PCMCIA socket state to our socket configure structure.
*/
static int soc_common_pcmcia_config_skt(
struct soc_pcmcia_socket *skt, socket_state_t *state)
{
int ret;
ret = skt->ops->configure_socket(skt, state);
if (ret < 0) {
pr_err("soc_common_pcmcia: unable to configure socket %d\n",
skt->nr);
/* restore the previous state */
WARN_ON(skt->ops->configure_socket(skt, &skt->cs_state));
return ret;
}
if (ret == 0) {
struct gpio_desc *descs[2];
gpiolib: Pass bitmaps, not integer arrays, to get/set array Most users of get/set array functions iterate consecutive bits of data, usually a single integer, while processing array of results obtained from, or building an array of values to be passed to those functions. Save time wasted on those iterations by changing the functions' API to accept bitmaps. All current users are updated as well. More benefits from the change are expected as soon as planned support for accepting/passing those bitmaps directly from/to respective GPIO chip callbacks if applicable is implemented. Cc: Jonathan Corbet <corbet@lwn.net> Cc: Miguel Ojeda Sandonis <miguel.ojeda.sandonis@gmail.com> Cc: Sebastien Bourdelin <sebastien.bourdelin@savoirfairelinux.com> Cc: Lukas Wunner <lukas@wunner.de> Cc: Peter Korsgaard <peter.korsgaard@barco.com> Cc: Peter Rosin <peda@axentia.se> Cc: Andrew Lunn <andrew@lunn.ch> Cc: Florian Fainelli <f.fainelli@gmail.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Rojhalat Ibrahim <imr@rtschenk.de> Cc: Dominik Brodowski <linux@dominikbrodowski.net> Cc: Russell King <rmk+kernel@armlinux.org.uk> Cc: Kishon Vijay Abraham I <kishon@ti.com> Cc: Tony Lindgren <tony@atomide.com> Cc: Lars-Peter Clausen <lars@metafoo.de> Cc: Michael Hennerich <Michael.Hennerich@analog.com> Cc: Jonathan Cameron <jic23@kernel.org> Cc: Hartmut Knaack <knaack.h@gmx.de> Cc: Peter Meerwald-Stadler <pmeerw@pmeerw.net> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Cc: Jiri Slaby <jslaby@suse.com> Cc: Yegor Yefremov <yegorslists@googlemail.com> Cc: Uwe Kleine-König <u.kleine-koenig@pengutronix.de> Signed-off-by: Janusz Krzysztofik <jmkrzyszt@gmail.com> Acked-by: Ulf Hansson <ulf.hansson@linaro.org> Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be> Tested-by: Geert Uytterhoeven <geert+renesas@glider.be> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2018-09-06 05:50:05 +08:00
DECLARE_BITMAP(values, 2);
int n = 0;
if (skt->gpio_reset) {
descs[n] = skt->gpio_reset;
gpiolib: Pass bitmaps, not integer arrays, to get/set array Most users of get/set array functions iterate consecutive bits of data, usually a single integer, while processing array of results obtained from, or building an array of values to be passed to those functions. Save time wasted on those iterations by changing the functions' API to accept bitmaps. All current users are updated as well. More benefits from the change are expected as soon as planned support for accepting/passing those bitmaps directly from/to respective GPIO chip callbacks if applicable is implemented. Cc: Jonathan Corbet <corbet@lwn.net> Cc: Miguel Ojeda Sandonis <miguel.ojeda.sandonis@gmail.com> Cc: Sebastien Bourdelin <sebastien.bourdelin@savoirfairelinux.com> Cc: Lukas Wunner <lukas@wunner.de> Cc: Peter Korsgaard <peter.korsgaard@barco.com> Cc: Peter Rosin <peda@axentia.se> Cc: Andrew Lunn <andrew@lunn.ch> Cc: Florian Fainelli <f.fainelli@gmail.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Rojhalat Ibrahim <imr@rtschenk.de> Cc: Dominik Brodowski <linux@dominikbrodowski.net> Cc: Russell King <rmk+kernel@armlinux.org.uk> Cc: Kishon Vijay Abraham I <kishon@ti.com> Cc: Tony Lindgren <tony@atomide.com> Cc: Lars-Peter Clausen <lars@metafoo.de> Cc: Michael Hennerich <Michael.Hennerich@analog.com> Cc: Jonathan Cameron <jic23@kernel.org> Cc: Hartmut Knaack <knaack.h@gmx.de> Cc: Peter Meerwald-Stadler <pmeerw@pmeerw.net> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Cc: Jiri Slaby <jslaby@suse.com> Cc: Yegor Yefremov <yegorslists@googlemail.com> Cc: Uwe Kleine-König <u.kleine-koenig@pengutronix.de> Signed-off-by: Janusz Krzysztofik <jmkrzyszt@gmail.com> Acked-by: Ulf Hansson <ulf.hansson@linaro.org> Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be> Tested-by: Geert Uytterhoeven <geert+renesas@glider.be> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2018-09-06 05:50:05 +08:00
__assign_bit(n++, values, state->flags & SS_RESET);
}
if (skt->gpio_bus_enable) {
descs[n] = skt->gpio_bus_enable;
gpiolib: Pass bitmaps, not integer arrays, to get/set array Most users of get/set array functions iterate consecutive bits of data, usually a single integer, while processing array of results obtained from, or building an array of values to be passed to those functions. Save time wasted on those iterations by changing the functions' API to accept bitmaps. All current users are updated as well. More benefits from the change are expected as soon as planned support for accepting/passing those bitmaps directly from/to respective GPIO chip callbacks if applicable is implemented. Cc: Jonathan Corbet <corbet@lwn.net> Cc: Miguel Ojeda Sandonis <miguel.ojeda.sandonis@gmail.com> Cc: Sebastien Bourdelin <sebastien.bourdelin@savoirfairelinux.com> Cc: Lukas Wunner <lukas@wunner.de> Cc: Peter Korsgaard <peter.korsgaard@barco.com> Cc: Peter Rosin <peda@axentia.se> Cc: Andrew Lunn <andrew@lunn.ch> Cc: Florian Fainelli <f.fainelli@gmail.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Rojhalat Ibrahim <imr@rtschenk.de> Cc: Dominik Brodowski <linux@dominikbrodowski.net> Cc: Russell King <rmk+kernel@armlinux.org.uk> Cc: Kishon Vijay Abraham I <kishon@ti.com> Cc: Tony Lindgren <tony@atomide.com> Cc: Lars-Peter Clausen <lars@metafoo.de> Cc: Michael Hennerich <Michael.Hennerich@analog.com> Cc: Jonathan Cameron <jic23@kernel.org> Cc: Hartmut Knaack <knaack.h@gmx.de> Cc: Peter Meerwald-Stadler <pmeerw@pmeerw.net> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Cc: Jiri Slaby <jslaby@suse.com> Cc: Yegor Yefremov <yegorslists@googlemail.com> Cc: Uwe Kleine-König <u.kleine-koenig@pengutronix.de> Signed-off-by: Janusz Krzysztofik <jmkrzyszt@gmail.com> Acked-by: Ulf Hansson <ulf.hansson@linaro.org> Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be> Tested-by: Geert Uytterhoeven <geert+renesas@glider.be> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2018-09-06 05:50:05 +08:00
__assign_bit(n++, values, state->flags & SS_OUTPUT_ENA);
}
if (n)
gpiolib: Pass array info to get/set array functions In order to make use of array info obtained from gpiod_get_array() and speed up processing of arrays matching single GPIO chip layout, that information must be passed to get/set array functions. Extend the functions' API with that additional parameter and update all users. Pass NULL if a user builds an array itself from single GPIOs. Cc: Jonathan Corbet <corbet@lwn.net> Cc: Miguel Ojeda Sandonis <miguel.ojeda.sandonis@gmail.com> Cc: Geert Uytterhoeven <geert@linux-m68k.org> Cc: Sebastien Bourdelin <sebastien.bourdelin@savoirfairelinux.com> Cc: Lukas Wunner <lukas@wunner.de> Cc: Peter Korsgaard <peter.korsgaard@barco.com> Cc: Peter Rosin <peda@axentia.se> Cc: Andrew Lunn <andrew@lunn.ch> Cc: Florian Fainelli <f.fainelli@gmail.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Rojhalat Ibrahim <imr@rtschenk.de> Cc: Dominik Brodowski <linux@dominikbrodowski.net> Cc: Russell King <rmk+kernel@armlinux.org.uk> Cc: Kishon Vijay Abraham I <kishon@ti.com> Cc: Tony Lindgren <tony@atomide.com> Cc: Lars-Peter Clausen <lars@metafoo.de> Cc: Michael Hennerich <Michael.Hennerich@analog.com> Cc: Jonathan Cameron <jic23@kernel.org> Cc: Hartmut Knaack <knaack.h@gmx.de> Cc: Peter Meerwald-Stadler <pmeerw@pmeerw.net> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Cc: Jiri Slaby <jslaby@suse.com> Cc: Yegor Yefremov <yegorslists@googlemail.com> Cc: Uwe Kleine-König <u.kleine-koenig@pengutronix.de> Signed-off-by: Janusz Krzysztofik <jmkrzyszt@gmail.com> Acked-by: Ulf Hansson <ulf.hansson@linaro.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
2018-09-06 05:50:07 +08:00
gpiod_set_array_value_cansleep(n, descs, NULL, values);
/*
* This really needs a better solution. The IRQ
* may or may not be claimed by the driver.
*/
if (skt->irq_state != 1 && state->io_irq) {
skt->irq_state = 1;
irq_set_irq_type(skt->socket.pci_irq,
IRQ_TYPE_EDGE_FALLING);
} else if (skt->irq_state == 1 && state->io_irq == 0) {
skt->irq_state = 0;
irq_set_irq_type(skt->socket.pci_irq, IRQ_TYPE_NONE);
}
skt->cs_state = *state;
}
return ret;
}
/* soc_common_pcmcia_sock_init()
* ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
*
* (Re-)Initialise the socket, turning on status interrupts
* and PCMCIA bus. This must wait for power to stabilise
* so that the card status signals report correctly.
*
* Returns: 0
*/
static int soc_common_pcmcia_sock_init(struct pcmcia_socket *sock)
{
struct soc_pcmcia_socket *skt = to_soc_pcmcia_socket(sock);
debug(skt, 2, "initializing socket\n");
if (skt->ops->socket_init)
skt->ops->socket_init(skt);
soc_pcmcia_hw_enable(skt);
return 0;
}
/*
* soc_common_pcmcia_suspend()
* ^^^^^^^^^^^^^^^^^^^^^^^^^^^
*
* Remove power on the socket, disable IRQs from the card.
* Turn off status interrupts, and disable the PCMCIA bus.
*
* Returns: 0
*/
static int soc_common_pcmcia_suspend(struct pcmcia_socket *sock)
{
struct soc_pcmcia_socket *skt = to_soc_pcmcia_socket(sock);
debug(skt, 2, "suspending socket\n");
soc_pcmcia_hw_disable(skt);
if (skt->ops->socket_suspend)
skt->ops->socket_suspend(skt);
return 0;
}
static DEFINE_SPINLOCK(status_lock);
static void soc_common_check_status(struct soc_pcmcia_socket *skt)
{
unsigned int events;
debug(skt, 4, "entering PCMCIA monitoring thread\n");
do {
unsigned int status;
unsigned long flags;
status = soc_common_pcmcia_skt_state(skt);
spin_lock_irqsave(&status_lock, flags);
events = (status ^ skt->status) & skt->cs_state.csc_mask;
skt->status = status;
spin_unlock_irqrestore(&status_lock, flags);
debug(skt, 4, "events: %s%s%s%s%s%s\n",
events == 0 ? "<NONE>" : "",
events & SS_DETECT ? "DETECT " : "",
events & SS_READY ? "READY " : "",
events & SS_BATDEAD ? "BATDEAD " : "",
events & SS_BATWARN ? "BATWARN " : "",
events & SS_STSCHG ? "STSCHG " : "");
if (events)
pcmcia_parse_events(&skt->socket, events);
} while (events);
}
/* Let's poll for events in addition to IRQs since IRQ only is unreliable... */
static void soc_common_pcmcia_poll_event(struct timer_list *t)
{
struct soc_pcmcia_socket *skt = from_timer(skt, t, poll_timer);
debug(skt, 4, "polling for events\n");
mod_timer(&skt->poll_timer, jiffies + SOC_PCMCIA_POLL_PERIOD);
soc_common_check_status(skt);
}
/*
* Service routine for socket driver interrupts (requested by the
* low-level PCMCIA init() operation via soc_common_pcmcia_thread()).
* The actual interrupt-servicing work is performed by
* soc_common_pcmcia_thread(), largely because the Card Services event-
* handling code performs scheduling operations which cannot be
* executed from within an interrupt context.
*/
IRQ: Maintain regs pointer globally rather than passing to IRQ handlers Maintain a per-CPU global "struct pt_regs *" variable which can be used instead of passing regs around manually through all ~1800 interrupt handlers in the Linux kernel. The regs pointer is used in few places, but it potentially costs both stack space and code to pass it around. On the FRV arch, removing the regs parameter from all the genirq function results in a 20% speed up of the IRQ exit path (ie: from leaving timer_interrupt() to leaving do_IRQ()). Where appropriate, an arch may override the generic storage facility and do something different with the variable. On FRV, for instance, the address is maintained in GR28 at all times inside the kernel as part of general exception handling. Having looked over the code, it appears that the parameter may be handed down through up to twenty or so layers of functions. Consider a USB character device attached to a USB hub, attached to a USB controller that posts its interrupts through a cascaded auxiliary interrupt controller. A character device driver may want to pass regs to the sysrq handler through the input layer which adds another few layers of parameter passing. I've build this code with allyesconfig for x86_64 and i386. I've runtested the main part of the code on FRV and i386, though I can't test most of the drivers. I've also done partial conversion for powerpc and MIPS - these at least compile with minimal configurations. This will affect all archs. Mostly the changes should be relatively easy. Take do_IRQ(), store the regs pointer at the beginning, saving the old one: struct pt_regs *old_regs = set_irq_regs(regs); And put the old one back at the end: set_irq_regs(old_regs); Don't pass regs through to generic_handle_irq() or __do_IRQ(). In timer_interrupt(), this sort of change will be necessary: - update_process_times(user_mode(regs)); - profile_tick(CPU_PROFILING, regs); + update_process_times(user_mode(get_irq_regs())); + profile_tick(CPU_PROFILING); I'd like to move update_process_times()'s use of get_irq_regs() into itself, except that i386, alone of the archs, uses something other than user_mode(). Some notes on the interrupt handling in the drivers: (*) input_dev() is now gone entirely. The regs pointer is no longer stored in the input_dev struct. (*) finish_unlinks() in drivers/usb/host/ohci-q.c needs checking. It does something different depending on whether it's been supplied with a regs pointer or not. (*) Various IRQ handler function pointers have been moved to type irq_handler_t. Signed-Off-By: David Howells <dhowells@redhat.com> (cherry picked from 1b16e7ac850969f38b375e511e3fa2f474a33867 commit)
2006-10-05 21:55:46 +08:00
static irqreturn_t soc_common_pcmcia_interrupt(int irq, void *dev)
{
struct soc_pcmcia_socket *skt = dev;
debug(skt, 3, "servicing IRQ %d\n", irq);
soc_common_check_status(skt);
return IRQ_HANDLED;
}
/*
* Implements the get_status() operation for the in-kernel PCMCIA
* service (formerly SS_GetStatus in Card Services). Essentially just
* fills in bits in `status' according to internal driver state or
* the value of the voltage detect chipselect register.
*
* As a debugging note, during card startup, the PCMCIA core issues
* three set_socket() commands in a row the first with RESET deasserted,
* the second with RESET asserted, and the last with RESET deasserted
* again. Following the third set_socket(), a get_status() command will
* be issued. The kernel is looking for the SS_READY flag (see
* setup_socket(), reset_socket(), and unreset_socket() in cs.c).
*
* Returns: 0
*/
static int
soc_common_pcmcia_get_status(struct pcmcia_socket *sock, unsigned int *status)
{
struct soc_pcmcia_socket *skt = to_soc_pcmcia_socket(sock);
skt->status = soc_common_pcmcia_skt_state(skt);
*status = skt->status;
return 0;
}
/*
* Implements the set_socket() operation for the in-kernel PCMCIA
* service (formerly SS_SetSocket in Card Services). We more or
* less punt all of this work and let the kernel handle the details
* of power configuration, reset, &c. We also record the value of
* `state' in order to regurgitate it to the PCMCIA core later.
*/
static int soc_common_pcmcia_set_socket(
struct pcmcia_socket *sock, socket_state_t *state)
{
struct soc_pcmcia_socket *skt = to_soc_pcmcia_socket(sock);
debug(skt, 2, "mask: %s%s%s%s%s%s flags: %s%s%s%s%s%s Vcc %d Vpp %d irq %d\n",
(state->csc_mask == 0) ? "<NONE> " : "",
(state->csc_mask & SS_DETECT) ? "DETECT " : "",
(state->csc_mask & SS_READY) ? "READY " : "",
(state->csc_mask & SS_BATDEAD) ? "BATDEAD " : "",
(state->csc_mask & SS_BATWARN) ? "BATWARN " : "",
(state->csc_mask & SS_STSCHG) ? "STSCHG " : "",
(state->flags == 0) ? "<NONE> " : "",
(state->flags & SS_PWR_AUTO) ? "PWR_AUTO " : "",
(state->flags & SS_IOCARD) ? "IOCARD " : "",
(state->flags & SS_RESET) ? "RESET " : "",
(state->flags & SS_SPKR_ENA) ? "SPKR_ENA " : "",
(state->flags & SS_OUTPUT_ENA) ? "OUTPUT_ENA " : "",
state->Vcc, state->Vpp, state->io_irq);
return soc_common_pcmcia_config_skt(skt, state);
}
/*
* Implements the set_io_map() operation for the in-kernel PCMCIA
* service (formerly SS_SetIOMap in Card Services). We configure
* the map speed as requested, but override the address ranges
* supplied by Card Services.
*
* Returns: 0 on success, -1 on error
*/
static int soc_common_pcmcia_set_io_map(
struct pcmcia_socket *sock, struct pccard_io_map *map)
{
struct soc_pcmcia_socket *skt = to_soc_pcmcia_socket(sock);
unsigned short speed = map->speed;
debug(skt, 2, "map %u speed %u start 0x%08llx stop 0x%08llx\n",
map->map, map->speed, (unsigned long long)map->start,
(unsigned long long)map->stop);
debug(skt, 2, "flags: %s%s%s%s%s%s%s%s\n",
(map->flags == 0) ? "<NONE>" : "",
(map->flags & MAP_ACTIVE) ? "ACTIVE " : "",
(map->flags & MAP_16BIT) ? "16BIT " : "",
(map->flags & MAP_AUTOSZ) ? "AUTOSZ " : "",
(map->flags & MAP_0WS) ? "0WS " : "",
(map->flags & MAP_WRPROT) ? "WRPROT " : "",
(map->flags & MAP_USE_WAIT) ? "USE_WAIT " : "",
(map->flags & MAP_PREFETCH) ? "PREFETCH " : "");
if (map->map >= MAX_IO_WIN) {
printk(KERN_ERR "%s(): map (%d) out of range\n", __func__,
map->map);
return -1;
}
if (map->flags & MAP_ACTIVE) {
if (speed == 0)
speed = SOC_PCMCIA_IO_ACCESS;
} else {
speed = 0;
}
skt->spd_io[map->map] = speed;
skt->ops->set_timing(skt);
if (map->stop == 1)
map->stop = PAGE_SIZE-1;
map->stop -= map->start;
map->stop += skt->socket.io_offset;
map->start = skt->socket.io_offset;
return 0;
}
/*
* Implements the set_mem_map() operation for the in-kernel PCMCIA
* service (formerly SS_SetMemMap in Card Services). We configure
* the map speed as requested, but override the address ranges
* supplied by Card Services.
*
* Returns: 0 on success, -ERRNO on error
*/
static int soc_common_pcmcia_set_mem_map(
struct pcmcia_socket *sock, struct pccard_mem_map *map)
{
struct soc_pcmcia_socket *skt = to_soc_pcmcia_socket(sock);
struct resource *res;
unsigned short speed = map->speed;
debug(skt, 2, "map %u speed %u card_start %08x\n",
map->map, map->speed, map->card_start);
debug(skt, 2, "flags: %s%s%s%s%s%s%s%s\n",
(map->flags == 0) ? "<NONE>" : "",
(map->flags & MAP_ACTIVE) ? "ACTIVE " : "",
(map->flags & MAP_16BIT) ? "16BIT " : "",
(map->flags & MAP_AUTOSZ) ? "AUTOSZ " : "",
(map->flags & MAP_0WS) ? "0WS " : "",
(map->flags & MAP_WRPROT) ? "WRPROT " : "",
(map->flags & MAP_ATTRIB) ? "ATTRIB " : "",
(map->flags & MAP_USE_WAIT) ? "USE_WAIT " : "");
if (map->map >= MAX_WIN)
return -EINVAL;
if (map->flags & MAP_ACTIVE) {
if (speed == 0)
speed = 300;
} else {
speed = 0;
}
if (map->flags & MAP_ATTRIB) {
res = &skt->res_attr;
skt->spd_attr[map->map] = speed;
skt->spd_mem[map->map] = 0;
} else {
res = &skt->res_mem;
skt->spd_attr[map->map] = 0;
skt->spd_mem[map->map] = speed;
}
skt->ops->set_timing(skt);
map->static_start = res->start + map->card_start;
return 0;
}
struct bittbl {
unsigned int mask;
const char *name;
};
static struct bittbl status_bits[] = {
{ SS_WRPROT, "SS_WRPROT" },
{ SS_BATDEAD, "SS_BATDEAD" },
{ SS_BATWARN, "SS_BATWARN" },
{ SS_READY, "SS_READY" },
{ SS_DETECT, "SS_DETECT" },
{ SS_POWERON, "SS_POWERON" },
{ SS_STSCHG, "SS_STSCHG" },
{ SS_3VCARD, "SS_3VCARD" },
{ SS_XVCARD, "SS_XVCARD" },
};
static struct bittbl conf_bits[] = {
{ SS_PWR_AUTO, "SS_PWR_AUTO" },
{ SS_IOCARD, "SS_IOCARD" },
{ SS_RESET, "SS_RESET" },
{ SS_DMA_MODE, "SS_DMA_MODE" },
{ SS_SPKR_ENA, "SS_SPKR_ENA" },
{ SS_OUTPUT_ENA, "SS_OUTPUT_ENA" },
};
static void dump_bits(char **p, const char *prefix,
unsigned int val, struct bittbl *bits, int sz)
{
char *b = *p;
int i;
b += sprintf(b, "%-9s:", prefix);
for (i = 0; i < sz; i++)
if (val & bits[i].mask)
b += sprintf(b, " %s", bits[i].name);
*b++ = '\n';
*p = b;
}
/*
* Implements the /sys/class/pcmcia_socket/??/status file.
*
* Returns: the number of characters added to the buffer
*/
static ssize_t show_status(
struct device *dev, struct device_attribute *attr, char *buf)
{
struct soc_pcmcia_socket *skt =
container_of(dev, struct soc_pcmcia_socket, socket.dev);
char *p = buf;
p += sprintf(p, "slot : %d\n", skt->nr);
dump_bits(&p, "status", skt->status,
status_bits, ARRAY_SIZE(status_bits));
dump_bits(&p, "csc_mask", skt->cs_state.csc_mask,
status_bits, ARRAY_SIZE(status_bits));
dump_bits(&p, "cs_flags", skt->cs_state.flags,
conf_bits, ARRAY_SIZE(conf_bits));
p += sprintf(p, "Vcc : %d\n", skt->cs_state.Vcc);
p += sprintf(p, "Vpp : %d\n", skt->cs_state.Vpp);
p += sprintf(p, "IRQ : %d (%d)\n", skt->cs_state.io_irq,
skt->socket.pci_irq);
if (skt->ops->show_timing)
p += skt->ops->show_timing(skt, p);
return p-buf;
}
static DEVICE_ATTR(status, S_IRUGO, show_status, NULL);
static struct pccard_operations soc_common_pcmcia_operations = {
.init = soc_common_pcmcia_sock_init,
.suspend = soc_common_pcmcia_suspend,
.get_status = soc_common_pcmcia_get_status,
.set_socket = soc_common_pcmcia_set_socket,
.set_io_map = soc_common_pcmcia_set_io_map,
.set_mem_map = soc_common_pcmcia_set_mem_map,
};
#ifdef CONFIG_CPU_FREQ
static int soc_common_pcmcia_cpufreq_nb(struct notifier_block *nb,
unsigned long val, void *data)
{
struct soc_pcmcia_socket *skt = container_of(nb, struct soc_pcmcia_socket, cpufreq_nb);
struct cpufreq_freqs *freqs = data;
return skt->ops->frequency_change(skt, val, freqs);
}
#endif
void soc_pcmcia_init_one(struct soc_pcmcia_socket *skt,
const struct pcmcia_low_level *ops, struct device *dev)
{
int i;
skt->ops = ops;
skt->socket.owner = ops->owner;
skt->socket.dev.parent = dev;
skt->socket.pci_irq = NO_IRQ;
for (i = 0; i < ARRAY_SIZE(skt->stat); i++)
skt->stat[i].gpio = -EINVAL;
}
EXPORT_SYMBOL(soc_pcmcia_init_one);
void soc_pcmcia_remove_one(struct soc_pcmcia_socket *skt)
{
del_timer_sync(&skt->poll_timer);
pcmcia_unregister_socket(&skt->socket);
#ifdef CONFIG_CPU_FREQ
if (skt->ops->frequency_change)
cpufreq_unregister_notifier(&skt->cpufreq_nb,
CPUFREQ_TRANSITION_NOTIFIER);
#endif
soc_pcmcia_hw_shutdown(skt);
/* should not be required; violates some lowlevel drivers */
soc_common_pcmcia_config_skt(skt, &dead_socket);
ARM: pxa/sa1100: move I/O space to PCI_IOBASE PXA and StrongARM1100 traditionally map their I/O space 1:1 into virtual memory, using a per-bus io_offset that matches the base address of the ioremap mapping. In order for PXA to work in a multiplatform config, this needs to change so I/O space starts at PCI_IOBASE (0xfee00000). Since the pcmcia soc_common support is shared with StrongARM1100, both have to change at the same time. The affected machines are: - Anything with a PCMCIA slot now uses pci_remap_iospace, which is made available to PCMCIA configurations as well, rather than just PCI. The first PCMCIA slot now starts at port number 0x10000. - The Zeus and Viper platforms have PC/104-style ISA buses, which have a static mapping for both I/O and memory space at 0xf1000000, which can no longer work. It does not appear to have any in-tree users, so moving it to port number 0 makes them behave like a traditional PC. - SA1100 does support ISA slots in theory, but all machines that originally enabled this appear to have been removed from the tree ages ago, and the I/O space is never mapped anywhere. - The Nanoengine machine has support for PCI slots, but looks like this never included I/O space, the resources only define the location for memory and config space. With this, the definitions of __io() and IO_SPACE_LIMIT can be simplified, as the only remaining cases are the generic PCI_IOBASE and the custom inb()/outb() macros on RiscPC. S3C24xx still has a custom inb()/outb() in this here, but this is already removed in another branch. Signed-off-by: Arnd Bergmann <arnd@arndb.de>
2022-04-24 20:36:35 +08:00
iounmap(PCI_IOBASE + skt->res_io_io.start);
release_resource(&skt->res_attr);
release_resource(&skt->res_mem);
release_resource(&skt->res_io);
release_resource(&skt->res_skt);
}
EXPORT_SYMBOL(soc_pcmcia_remove_one);
int soc_pcmcia_add_one(struct soc_pcmcia_socket *skt)
{
int ret;
skt->cs_state = dead_socket;
timer_setup(&skt->poll_timer, soc_common_pcmcia_poll_event, 0);
skt->poll_timer.expires = jiffies + SOC_PCMCIA_POLL_PERIOD;
ret = request_resource(&iomem_resource, &skt->res_skt);
if (ret)
goto out_err_1;
ret = request_resource(&skt->res_skt, &skt->res_io);
if (ret)
goto out_err_2;
ret = request_resource(&skt->res_skt, &skt->res_mem);
if (ret)
goto out_err_3;
ret = request_resource(&skt->res_skt, &skt->res_attr);
if (ret)
goto out_err_4;
ARM: pxa/sa1100: move I/O space to PCI_IOBASE PXA and StrongARM1100 traditionally map their I/O space 1:1 into virtual memory, using a per-bus io_offset that matches the base address of the ioremap mapping. In order for PXA to work in a multiplatform config, this needs to change so I/O space starts at PCI_IOBASE (0xfee00000). Since the pcmcia soc_common support is shared with StrongARM1100, both have to change at the same time. The affected machines are: - Anything with a PCMCIA slot now uses pci_remap_iospace, which is made available to PCMCIA configurations as well, rather than just PCI. The first PCMCIA slot now starts at port number 0x10000. - The Zeus and Viper platforms have PC/104-style ISA buses, which have a static mapping for both I/O and memory space at 0xf1000000, which can no longer work. It does not appear to have any in-tree users, so moving it to port number 0 makes them behave like a traditional PC. - SA1100 does support ISA slots in theory, but all machines that originally enabled this appear to have been removed from the tree ages ago, and the I/O space is never mapped anywhere. - The Nanoengine machine has support for PCI slots, but looks like this never included I/O space, the resources only define the location for memory and config space. With this, the definitions of __io() and IO_SPACE_LIMIT can be simplified, as the only remaining cases are the generic PCI_IOBASE and the custom inb()/outb() macros on RiscPC. S3C24xx still has a custom inb()/outb() in this here, but this is already removed in another branch. Signed-off-by: Arnd Bergmann <arnd@arndb.de>
2022-04-24 20:36:35 +08:00
skt->res_io_io = (struct resource)
DEFINE_RES_IO_NAMED(skt->nr * 0x1000 + 0x10000, 0x1000,
"PCMCIA I/O");
ret = pci_remap_iospace(&skt->res_io_io, skt->res_io.start);
if (ret)
goto out_err_5;
/*
* We initialize default socket timing here, because
* we are not guaranteed to see a SetIOMap operation at
* runtime.
*/
skt->ops->set_timing(skt);
ret = soc_pcmcia_hw_init(skt);
if (ret)
goto out_err_6;
skt->socket.ops = &soc_common_pcmcia_operations;
skt->socket.features = SS_CAP_STATIC_MAP|SS_CAP_PCCARD;
skt->socket.resource_ops = &pccard_static_ops;
skt->socket.irq_mask = 0;
skt->socket.map_size = PAGE_SIZE;
ARM: pxa/sa1100: move I/O space to PCI_IOBASE PXA and StrongARM1100 traditionally map their I/O space 1:1 into virtual memory, using a per-bus io_offset that matches the base address of the ioremap mapping. In order for PXA to work in a multiplatform config, this needs to change so I/O space starts at PCI_IOBASE (0xfee00000). Since the pcmcia soc_common support is shared with StrongARM1100, both have to change at the same time. The affected machines are: - Anything with a PCMCIA slot now uses pci_remap_iospace, which is made available to PCMCIA configurations as well, rather than just PCI. The first PCMCIA slot now starts at port number 0x10000. - The Zeus and Viper platforms have PC/104-style ISA buses, which have a static mapping for both I/O and memory space at 0xf1000000, which can no longer work. It does not appear to have any in-tree users, so moving it to port number 0 makes them behave like a traditional PC. - SA1100 does support ISA slots in theory, but all machines that originally enabled this appear to have been removed from the tree ages ago, and the I/O space is never mapped anywhere. - The Nanoengine machine has support for PCI slots, but looks like this never included I/O space, the resources only define the location for memory and config space. With this, the definitions of __io() and IO_SPACE_LIMIT can be simplified, as the only remaining cases are the generic PCI_IOBASE and the custom inb()/outb() macros on RiscPC. S3C24xx still has a custom inb()/outb() in this here, but this is already removed in another branch. Signed-off-by: Arnd Bergmann <arnd@arndb.de>
2022-04-24 20:36:35 +08:00
skt->socket.io_offset = (unsigned long)skt->res_io_io.start;
skt->status = soc_common_pcmcia_skt_state(skt);
#ifdef CONFIG_CPU_FREQ
if (skt->ops->frequency_change) {
skt->cpufreq_nb.notifier_call = soc_common_pcmcia_cpufreq_nb;
ret = cpufreq_register_notifier(&skt->cpufreq_nb,
CPUFREQ_TRANSITION_NOTIFIER);
if (ret < 0)
dev_err(skt->socket.dev.parent,
"unable to register CPU frequency change notifier for PCMCIA (%d)\n",
ret);
}
#endif
ret = pcmcia_register_socket(&skt->socket);
if (ret)
goto out_err_7;
ret = device_create_file(&skt->socket.dev, &dev_attr_status);
if (ret)
goto out_err_8;
return ret;
out_err_8:
del_timer_sync(&skt->poll_timer);
pcmcia_unregister_socket(&skt->socket);
out_err_7:
soc_pcmcia_hw_shutdown(skt);
out_err_6:
ARM: pxa/sa1100: move I/O space to PCI_IOBASE PXA and StrongARM1100 traditionally map their I/O space 1:1 into virtual memory, using a per-bus io_offset that matches the base address of the ioremap mapping. In order for PXA to work in a multiplatform config, this needs to change so I/O space starts at PCI_IOBASE (0xfee00000). Since the pcmcia soc_common support is shared with StrongARM1100, both have to change at the same time. The affected machines are: - Anything with a PCMCIA slot now uses pci_remap_iospace, which is made available to PCMCIA configurations as well, rather than just PCI. The first PCMCIA slot now starts at port number 0x10000. - The Zeus and Viper platforms have PC/104-style ISA buses, which have a static mapping for both I/O and memory space at 0xf1000000, which can no longer work. It does not appear to have any in-tree users, so moving it to port number 0 makes them behave like a traditional PC. - SA1100 does support ISA slots in theory, but all machines that originally enabled this appear to have been removed from the tree ages ago, and the I/O space is never mapped anywhere. - The Nanoengine machine has support for PCI slots, but looks like this never included I/O space, the resources only define the location for memory and config space. With this, the definitions of __io() and IO_SPACE_LIMIT can be simplified, as the only remaining cases are the generic PCI_IOBASE and the custom inb()/outb() macros on RiscPC. S3C24xx still has a custom inb()/outb() in this here, but this is already removed in another branch. Signed-off-by: Arnd Bergmann <arnd@arndb.de>
2022-04-24 20:36:35 +08:00
iounmap(PCI_IOBASE + skt->res_io_io.start);
out_err_5:
release_resource(&skt->res_attr);
out_err_4:
release_resource(&skt->res_mem);
out_err_3:
release_resource(&skt->res_io);
out_err_2:
release_resource(&skt->res_skt);
out_err_1:
return ret;
}
EXPORT_SYMBOL(soc_pcmcia_add_one);
MODULE_AUTHOR("John Dorsey <john+@cs.cmu.edu>");
MODULE_DESCRIPTION("Linux PCMCIA Card Services: Common SoC support");
MODULE_LICENSE("Dual MPL/GPL");