1
linux/drivers/macintosh/via-pmu68k.c

1063 lines
25 KiB
C
Raw Normal View History

/*
* Device driver for the PMU on 68K-based Apple PowerBooks
*
* The VIA (versatile interface adapter) interfaces to the PMU,
* a 6805 microprocessor core whose primary function is to control
* battery charging and system power on the PowerBooks.
* The PMU also controls the ADB (Apple Desktop Bus) which connects
* to the keyboard and mouse, as well as the non-volatile RAM
* and the RTC (real time clock) chip.
*
* Adapted for 68K PMU by Joshua M. Thompson
*
* Based largely on the PowerMac PMU code by Paul Mackerras and
* Fabio Riccardi.
*
* Also based on the PMU driver from MkLinux by Apple Computer, Inc.
* and the Open Software Foundation, Inc.
*/
#include <stdarg.h>
#include <linux/types.h>
#include <linux/errno.h>
#include <linux/kernel.h>
#include <linux/delay.h>
#include <linux/sched.h>
#include <linux/miscdevice.h>
#include <linux/blkdev.h>
#include <linux/pci.h>
#include <linux/slab.h>
#include <linux/init.h>
#include <linux/interrupt.h>
#include <linux/adb.h>
#include <linux/pmu.h>
#include <linux/cuda.h>
#include <asm/macintosh.h>
#include <asm/macints.h>
#include <asm/machw.h>
#include <asm/mac_via.h>
#include <asm/pgtable.h>
#include <asm/system.h>
#include <asm/irq.h>
#include <asm/uaccess.h>
/* Misc minor number allocated for /dev/pmu */
#define PMU_MINOR 154
/* VIA registers - spaced 0x200 bytes apart */
#define RS 0x200 /* skip between registers */
#define B 0 /* B-side data */
#define A RS /* A-side data */
#define DIRB (2*RS) /* B-side direction (1=output) */
#define DIRA (3*RS) /* A-side direction (1=output) */
#define T1CL (4*RS) /* Timer 1 ctr/latch (low 8 bits) */
#define T1CH (5*RS) /* Timer 1 counter (high 8 bits) */
#define T1LL (6*RS) /* Timer 1 latch (low 8 bits) */
#define T1LH (7*RS) /* Timer 1 latch (high 8 bits) */
#define T2CL (8*RS) /* Timer 2 ctr/latch (low 8 bits) */
#define T2CH (9*RS) /* Timer 2 counter (high 8 bits) */
#define SR (10*RS) /* Shift register */
#define ACR (11*RS) /* Auxiliary control register */
#define PCR (12*RS) /* Peripheral control register */
#define IFR (13*RS) /* Interrupt flag register */
#define IER (14*RS) /* Interrupt enable register */
#define ANH (15*RS) /* A-side data, no handshake */
/* Bits in B data register: both active low */
#define TACK 0x02 /* Transfer acknowledge (input) */
#define TREQ 0x04 /* Transfer request (output) */
/* Bits in ACR */
#define SR_CTRL 0x1c /* Shift register control bits */
#define SR_EXT 0x0c /* Shift on external clock */
#define SR_OUT 0x10 /* Shift out if 1 */
/* Bits in IFR and IER */
#define SR_INT 0x04 /* Shift register full/empty */
#define CB1_INT 0x10 /* transition on CB1 input */
static enum pmu_state {
idle,
sending,
intack,
reading,
reading_intr,
} pmu_state;
static struct adb_request *current_req;
static struct adb_request *last_req;
static struct adb_request *req_awaiting_reply;
static unsigned char interrupt_data[32];
static unsigned char *reply_ptr;
static int data_index;
static int data_len;
static int adb_int_pending;
static int pmu_adb_flags;
static int adb_dev_map = 0;
static struct adb_request bright_req_1, bright_req_2, bright_req_3;
static int pmu_kind = PMU_UNKNOWN;
static int pmu_fully_inited = 0;
int asleep;
[PATCH] Notifier chain update: API changes The kernel's implementation of notifier chains is unsafe. There is no protection against entries being added to or removed from a chain while the chain is in use. The issues were discussed in this thread: http://marc.theaimsgroup.com/?l=linux-kernel&m=113018709002036&w=2 We noticed that notifier chains in the kernel fall into two basic usage classes: "Blocking" chains are always called from a process context and the callout routines are allowed to sleep; "Atomic" chains can be called from an atomic context and the callout routines are not allowed to sleep. We decided to codify this distinction and make it part of the API. Therefore this set of patches introduces three new, parallel APIs: one for blocking notifiers, one for atomic notifiers, and one for "raw" notifiers (which is really just the old API under a new name). New kinds of data structures are used for the heads of the chains, and new routines are defined for registration, unregistration, and calling a chain. The three APIs are explained in include/linux/notifier.h and their implementation is in kernel/sys.c. With atomic and blocking chains, the implementation guarantees that the chain links will not be corrupted and that chain callers will not get messed up by entries being added or removed. For raw chains the implementation provides no guarantees at all; users of this API must provide their own protections. (The idea was that situations may come up where the assumptions of the atomic and blocking APIs are not appropriate, so it should be possible for users to handle these things in their own way.) There are some limitations, which should not be too hard to live with. For atomic/blocking chains, registration and unregistration must always be done in a process context since the chain is protected by a mutex/rwsem. Also, a callout routine for a non-raw chain must not try to register or unregister entries on its own chain. (This did happen in a couple of places and the code had to be changed to avoid it.) Since atomic chains may be called from within an NMI handler, they cannot use spinlocks for synchronization. Instead we use RCU. The overhead falls almost entirely in the unregister routine, which is okay since unregistration is much less frequent that calling a chain. Here is the list of chains that we adjusted and their classifications. None of them use the raw API, so for the moment it is only a placeholder. ATOMIC CHAINS ------------- arch/i386/kernel/traps.c: i386die_chain arch/ia64/kernel/traps.c: ia64die_chain arch/powerpc/kernel/traps.c: powerpc_die_chain arch/sparc64/kernel/traps.c: sparc64die_chain arch/x86_64/kernel/traps.c: die_chain drivers/char/ipmi/ipmi_si_intf.c: xaction_notifier_list kernel/panic.c: panic_notifier_list kernel/profile.c: task_free_notifier net/bluetooth/hci_core.c: hci_notifier net/ipv4/netfilter/ip_conntrack_core.c: ip_conntrack_chain net/ipv4/netfilter/ip_conntrack_core.c: ip_conntrack_expect_chain net/ipv6/addrconf.c: inet6addr_chain net/netfilter/nf_conntrack_core.c: nf_conntrack_chain net/netfilter/nf_conntrack_core.c: nf_conntrack_expect_chain net/netlink/af_netlink.c: netlink_chain BLOCKING CHAINS --------------- arch/powerpc/platforms/pseries/reconfig.c: pSeries_reconfig_chain arch/s390/kernel/process.c: idle_chain arch/x86_64/kernel/process.c idle_notifier drivers/base/memory.c: memory_chain drivers/cpufreq/cpufreq.c cpufreq_policy_notifier_list drivers/cpufreq/cpufreq.c cpufreq_transition_notifier_list drivers/macintosh/adb.c: adb_client_list drivers/macintosh/via-pmu.c sleep_notifier_list drivers/macintosh/via-pmu68k.c sleep_notifier_list drivers/macintosh/windfarm_core.c wf_client_list drivers/usb/core/notify.c usb_notifier_list drivers/video/fbmem.c fb_notifier_list kernel/cpu.c cpu_chain kernel/module.c module_notify_list kernel/profile.c munmap_notifier kernel/profile.c task_exit_notifier kernel/sys.c reboot_notifier_list net/core/dev.c netdev_chain net/decnet/dn_dev.c: dnaddr_chain net/ipv4/devinet.c: inetaddr_chain It's possible that some of these classifications are wrong. If they are, please let us know or submit a patch to fix them. Note that any chain that gets called very frequently should be atomic, because the rwsem read-locking used for blocking chains is very likely to incur cache misses on SMP systems. (However, if the chain's callout routines may sleep then the chain cannot be atomic.) The patch set was written by Alan Stern and Chandra Seetharaman, incorporating material written by Keith Owens and suggestions from Paul McKenney and Andrew Morton. [jes@sgi.com: restructure the notifier chain initialization macros] Signed-off-by: Alan Stern <stern@rowland.harvard.edu> Signed-off-by: Chandra Seetharaman <sekharan@us.ibm.com> Signed-off-by: Jes Sorensen <jes@sgi.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-03-27 02:16:30 -07:00
BLOCKING_NOTIFIER_HEAD(sleep_notifier_list);
static int pmu_probe(void);
static int pmu_init(void);
static void pmu_start(void);
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 06:55:46 -07:00
static irqreturn_t pmu_interrupt(int irq, void *arg);
static int pmu_send_request(struct adb_request *req, int sync);
static int pmu_autopoll(int devs);
void pmu_poll(void);
static int pmu_reset_bus(void);
static int pmu_queue_request(struct adb_request *req);
static void pmu_start(void);
static void send_byte(int x);
static void recv_byte(void);
static void pmu_done(struct adb_request *req);
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 06:55:46 -07:00
static void pmu_handle_data(unsigned char *data, int len);
static void set_volume(int level);
static void pmu_enable_backlight(int on);
static void pmu_set_brightness(int level);
struct adb_driver via_pmu_driver = {
"68K PMU",
pmu_probe,
pmu_init,
pmu_send_request,
pmu_autopoll,
pmu_poll,
pmu_reset_bus
};
/*
* This table indicates for each PMU opcode:
* - the number of data bytes to be sent with the command, or -1
* if a length byte should be sent,
* - the number of response bytes which the PMU will return, or
* -1 if it will send a length byte.
*/
static s8 pmu_data_len[256][2] = {
/* 0 1 2 3 4 5 6 7 */
/*00*/ {-1, 0},{-1, 0},{-1, 0},{-1, 0},{-1, 0},{-1, 0},{-1, 0},{-1, 0},
/*08*/ {-1,-1},{-1,-1},{-1,-1},{-1,-1},{-1,-1},{-1,-1},{-1,-1},{-1,-1},
/*10*/ { 1, 0},{ 1, 0},{-1, 0},{-1, 0},{-1, 0},{-1, 0},{-1, 0},{-1, 0},
/*18*/ { 0, 1},{ 0, 1},{-1,-1},{-1,-1},{-1,-1},{-1,-1},{-1,-1},{ 0, 0},
/*20*/ {-1, 0},{ 0, 0},{ 2, 0},{ 1, 0},{ 1, 0},{-1, 0},{-1, 0},{-1, 0},
/*28*/ { 0,-1},{ 0,-1},{-1,-1},{-1,-1},{-1,-1},{-1,-1},{-1,-1},{ 0,-1},
/*30*/ { 4, 0},{20, 0},{-1, 0},{ 3, 0},{-1, 0},{-1, 0},{-1, 0},{-1, 0},
/*38*/ { 0, 4},{ 0,20},{ 2,-1},{ 2, 1},{ 3,-1},{-1,-1},{-1,-1},{ 4, 0},
/*40*/ { 1, 0},{ 1, 0},{-1, 0},{-1, 0},{-1, 0},{-1, 0},{-1, 0},{-1, 0},
/*48*/ { 0, 1},{ 0, 1},{-1,-1},{ 1, 0},{ 1, 0},{-1,-1},{-1,-1},{-1,-1},
/*50*/ { 1, 0},{ 0, 0},{ 2, 0},{ 2, 0},{-1, 0},{ 1, 0},{ 3, 0},{ 1, 0},
/*58*/ { 0, 1},{ 1, 0},{ 0, 2},{ 0, 2},{ 0,-1},{-1,-1},{-1,-1},{-1,-1},
/*60*/ { 2, 0},{-1, 0},{-1, 0},{-1, 0},{-1, 0},{-1, 0},{-1, 0},{-1, 0},
/*68*/ { 0, 3},{ 0, 3},{ 0, 2},{ 0, 8},{ 0,-1},{ 0,-1},{-1,-1},{-1,-1},
/*70*/ { 1, 0},{ 1, 0},{ 1, 0},{-1, 0},{-1, 0},{-1, 0},{-1, 0},{-1, 0},
/*78*/ { 0,-1},{ 0,-1},{-1,-1},{-1,-1},{-1,-1},{ 5, 1},{ 4, 1},{ 4, 1},
/*80*/ { 4, 0},{-1, 0},{ 0, 0},{-1, 0},{-1, 0},{-1, 0},{-1, 0},{-1, 0},
/*88*/ { 0, 5},{-1,-1},{-1,-1},{-1,-1},{-1,-1},{-1,-1},{-1,-1},{-1,-1},
/*90*/ { 1, 0},{ 2, 0},{-1, 0},{-1, 0},{-1, 0},{-1, 0},{-1, 0},{-1, 0},
/*98*/ { 0, 1},{ 0, 1},{-1,-1},{-1,-1},{-1,-1},{-1,-1},{-1,-1},{-1,-1},
/*a0*/ { 2, 0},{ 2, 0},{ 2, 0},{ 4, 0},{-1, 0},{ 0, 0},{-1, 0},{-1, 0},
/*a8*/ { 1, 1},{ 1, 0},{ 3, 0},{ 2, 0},{-1,-1},{-1,-1},{-1,-1},{-1,-1},
/*b0*/ {-1, 0},{-1, 0},{-1, 0},{-1, 0},{-1, 0},{-1, 0},{-1, 0},{-1, 0},
/*b8*/ {-1,-1},{-1,-1},{-1,-1},{-1,-1},{-1,-1},{-1,-1},{-1,-1},{-1,-1},
/*c0*/ {-1, 0},{-1, 0},{-1, 0},{-1, 0},{-1, 0},{-1, 0},{-1, 0},{-1, 0},
/*c8*/ {-1,-1},{-1,-1},{-1,-1},{-1,-1},{-1,-1},{-1,-1},{-1,-1},{-1,-1},
/*d0*/ { 0, 0},{-1, 0},{-1, 0},{-1, 0},{-1, 0},{-1, 0},{-1, 0},{-1, 0},
/*d8*/ { 1, 1},{ 1, 1},{-1,-1},{-1,-1},{ 0, 1},{ 0,-1},{-1,-1},{-1,-1},
/*e0*/ {-1, 0},{ 4, 0},{ 0, 1},{-1, 0},{-1, 0},{ 4, 0},{-1, 0},{-1, 0},
/*e8*/ { 3,-1},{-1,-1},{ 0, 1},{-1,-1},{ 0,-1},{-1,-1},{-1,-1},{ 0, 0},
/*f0*/ {-1, 0},{-1, 0},{-1, 0},{-1, 0},{-1, 0},{-1, 0},{-1, 0},{-1, 0},
/*f8*/ {-1,-1},{-1,-1},{-1,-1},{-1,-1},{-1,-1},{-1,-1},{-1,-1},{-1,-1},
};
int pmu_probe(void)
{
if (macintosh_config->adb_type == MAC_ADB_PB1) {
pmu_kind = PMU_68K_V1;
} else if (macintosh_config->adb_type == MAC_ADB_PB2) {
pmu_kind = PMU_68K_V2;
} else {
return -ENODEV;
}
pmu_state = idle;
return 0;
}
static int
pmu_init(void)
{
int timeout;
volatile struct adb_request req;
via2[B] |= TREQ; /* negate TREQ */
via2[DIRB] = (via2[DIRB] | TREQ) & ~TACK; /* TACK in, TREQ out */
pmu_request((struct adb_request *) &req, NULL, 2, PMU_SET_INTR_MASK, PMU_INT_ADB);
timeout = 100000;
while (!req.complete) {
if (--timeout < 0) {
printk(KERN_ERR "pmu_init: no response from PMU\n");
return -EAGAIN;
}
udelay(10);
pmu_poll();
}
/* ack all pending interrupts */
timeout = 100000;
interrupt_data[0] = 1;
while (interrupt_data[0] || pmu_state != idle) {
if (--timeout < 0) {
printk(KERN_ERR "pmu_init: timed out acking intrs\n");
return -EAGAIN;
}
if (pmu_state == idle) {
adb_int_pending = 1;
pmu_interrupt(0, NULL);
}
pmu_poll();
udelay(10);
}
pmu_request((struct adb_request *) &req, NULL, 2, PMU_SET_INTR_MASK,
PMU_INT_ADB_AUTO|PMU_INT_SNDBRT|PMU_INT_ADB);
timeout = 100000;
while (!req.complete) {
if (--timeout < 0) {
printk(KERN_ERR "pmu_init: no response from PMU\n");
return -EAGAIN;
}
udelay(10);
pmu_poll();
}
bright_req_1.complete = 1;
bright_req_2.complete = 1;
bright_req_3.complete = 1;
if (request_irq(IRQ_MAC_ADB_SR, pmu_interrupt, 0, "pmu-shift",
pmu_interrupt)) {
printk(KERN_ERR "pmu_init: can't get irq %d\n",
IRQ_MAC_ADB_SR);
return -EAGAIN;
}
if (request_irq(IRQ_MAC_ADB_CL, pmu_interrupt, 0, "pmu-clock",
pmu_interrupt)) {
printk(KERN_ERR "pmu_init: can't get irq %d\n",
IRQ_MAC_ADB_CL);
free_irq(IRQ_MAC_ADB_SR, pmu_interrupt);
return -EAGAIN;
}
pmu_fully_inited = 1;
/* Enable backlight */
pmu_enable_backlight(1);
printk("adb: PMU 68K driver v0.5 for Unified ADB.\n");
return 0;
}
int
pmu_get_model(void)
{
return pmu_kind;
}
/* Send an ADB command */
static int
pmu_send_request(struct adb_request *req, int sync)
{
int i, ret;
if (!pmu_fully_inited)
{
req->complete = 1;
return -ENXIO;
}
ret = -EINVAL;
switch (req->data[0]) {
case PMU_PACKET:
for (i = 0; i < req->nbytes - 1; ++i)
req->data[i] = req->data[i+1];
--req->nbytes;
if (pmu_data_len[req->data[0]][1] != 0) {
req->reply[0] = ADB_RET_OK;
req->reply_len = 1;
} else
req->reply_len = 0;
ret = pmu_queue_request(req);
break;
case CUDA_PACKET:
switch (req->data[1]) {
case CUDA_GET_TIME:
if (req->nbytes != 2)
break;
req->data[0] = PMU_READ_RTC;
req->nbytes = 1;
req->reply_len = 3;
req->reply[0] = CUDA_PACKET;
req->reply[1] = 0;
req->reply[2] = CUDA_GET_TIME;
ret = pmu_queue_request(req);
break;
case CUDA_SET_TIME:
if (req->nbytes != 6)
break;
req->data[0] = PMU_SET_RTC;
req->nbytes = 5;
for (i = 1; i <= 4; ++i)
req->data[i] = req->data[i+1];
req->reply_len = 3;
req->reply[0] = CUDA_PACKET;
req->reply[1] = 0;
req->reply[2] = CUDA_SET_TIME;
ret = pmu_queue_request(req);
break;
case CUDA_GET_PRAM:
if (req->nbytes != 4)
break;
req->data[0] = PMU_READ_NVRAM;
req->data[1] = req->data[2];
req->data[2] = req->data[3];
req->nbytes = 3;
req->reply_len = 3;
req->reply[0] = CUDA_PACKET;
req->reply[1] = 0;
req->reply[2] = CUDA_GET_PRAM;
ret = pmu_queue_request(req);
break;
case CUDA_SET_PRAM:
if (req->nbytes != 5)
break;
req->data[0] = PMU_WRITE_NVRAM;
req->data[1] = req->data[2];
req->data[2] = req->data[3];
req->data[3] = req->data[4];
req->nbytes = 4;
req->reply_len = 3;
req->reply[0] = CUDA_PACKET;
req->reply[1] = 0;
req->reply[2] = CUDA_SET_PRAM;
ret = pmu_queue_request(req);
break;
}
break;
case ADB_PACKET:
for (i = req->nbytes - 1; i > 1; --i)
req->data[i+2] = req->data[i];
req->data[3] = req->nbytes - 2;
req->data[2] = pmu_adb_flags;
/*req->data[1] = req->data[1];*/
req->data[0] = PMU_ADB_CMD;
req->nbytes += 2;
req->reply_expected = 1;
req->reply_len = 0;
ret = pmu_queue_request(req);
break;
}
if (ret)
{
req->complete = 1;
return ret;
}
if (sync) {
while (!req->complete)
pmu_poll();
}
return 0;
}
/* Enable/disable autopolling */
static int
pmu_autopoll(int devs)
{
struct adb_request req;
if (!pmu_fully_inited) return -ENXIO;
if (devs) {
adb_dev_map = devs;
pmu_request(&req, NULL, 5, PMU_ADB_CMD, 0, 0x86,
adb_dev_map >> 8, adb_dev_map);
pmu_adb_flags = 2;
} else {
pmu_request(&req, NULL, 1, PMU_ADB_POLL_OFF);
pmu_adb_flags = 0;
}
while (!req.complete)
pmu_poll();
return 0;
}
/* Reset the ADB bus */
static int
pmu_reset_bus(void)
{
struct adb_request req;
long timeout;
int save_autopoll = adb_dev_map;
if (!pmu_fully_inited) return -ENXIO;
/* anyone got a better idea?? */
pmu_autopoll(0);
req.nbytes = 5;
req.done = NULL;
req.data[0] = PMU_ADB_CMD;
req.data[1] = 0;
req.data[2] = 3; /* ADB_BUSRESET ??? */
req.data[3] = 0;
req.data[4] = 0;
req.reply_len = 0;
req.reply_expected = 1;
if (pmu_queue_request(&req) != 0)
{
printk(KERN_ERR "pmu_adb_reset_bus: pmu_queue_request failed\n");
return -EIO;
}
while (!req.complete)
pmu_poll();
timeout = 100000;
while (!req.complete) {
if (--timeout < 0) {
printk(KERN_ERR "pmu_adb_reset_bus (reset): no response from PMU\n");
return -EIO;
}
udelay(10);
pmu_poll();
}
if (save_autopoll != 0)
pmu_autopoll(save_autopoll);
return 0;
}
/* Construct and send a pmu request */
int
pmu_request(struct adb_request *req, void (*done)(struct adb_request *),
int nbytes, ...)
{
va_list list;
int i;
if (nbytes < 0 || nbytes > 32) {
printk(KERN_ERR "pmu_request: bad nbytes (%d)\n", nbytes);
req->complete = 1;
return -EINVAL;
}
req->nbytes = nbytes;
req->done = done;
va_start(list, nbytes);
for (i = 0; i < nbytes; ++i)
req->data[i] = va_arg(list, int);
va_end(list);
if (pmu_data_len[req->data[0]][1] != 0) {
req->reply[0] = ADB_RET_OK;
req->reply_len = 1;
} else
req->reply_len = 0;
req->reply_expected = 0;
return pmu_queue_request(req);
}
static int
pmu_queue_request(struct adb_request *req)
{
unsigned long flags;
int nsend;
if (req->nbytes <= 0) {
req->complete = 1;
return 0;
}
nsend = pmu_data_len[req->data[0]][0];
if (nsend >= 0 && req->nbytes != nsend + 1) {
req->complete = 1;
return -EINVAL;
}
req->next = NULL;
req->sent = 0;
req->complete = 0;
local_irq_save(flags);
if (current_req != 0) {
last_req->next = req;
last_req = req;
} else {
current_req = req;
last_req = req;
if (pmu_state == idle)
pmu_start();
}
local_irq_restore(flags);
return 0;
}
static void
send_byte(int x)
{
via1[ACR] |= SR_CTRL;
via1[SR] = x;
via2[B] &= ~TREQ; /* assert TREQ */
}
static void
recv_byte(void)
{
char c;
via1[ACR] = (via1[ACR] | SR_EXT) & ~SR_OUT;
c = via1[SR]; /* resets SR */
via2[B] &= ~TREQ;
}
static void
pmu_start(void)
{
unsigned long flags;
struct adb_request *req;
/* assert pmu_state == idle */
/* get the packet to send */
local_irq_save(flags);
req = current_req;
if (req == 0 || pmu_state != idle
|| (req->reply_expected && req_awaiting_reply))
goto out;
pmu_state = sending;
data_index = 1;
data_len = pmu_data_len[req->data[0]][0];
/* set the shift register to shift out and send a byte */
send_byte(req->data[0]);
out:
local_irq_restore(flags);
}
void
pmu_poll(void)
{
unsigned long flags;
local_irq_save(flags);
if (via1[IFR] & SR_INT) {
via1[IFR] = SR_INT;
pmu_interrupt(IRQ_MAC_ADB_SR, NULL);
}
if (via1[IFR] & CB1_INT) {
via1[IFR] = CB1_INT;
pmu_interrupt(IRQ_MAC_ADB_CL, NULL);
}
local_irq_restore(flags);
}
static irqreturn_t
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 06:55:46 -07:00
pmu_interrupt(int irq, void *dev_id)
{
struct adb_request *req;
int timeout, bite = 0; /* to prevent compiler warning */
#if 0
printk("pmu_interrupt: irq %d state %d acr %02X, b %02X data_index %d/%d adb_int_pending %d\n",
irq, pmu_state, (uint) via1[ACR], (uint) via2[B], data_index, data_len, adb_int_pending);
#endif
if (irq == IRQ_MAC_ADB_CL) { /* CB1 interrupt */
adb_int_pending = 1;
} else if (irq == IRQ_MAC_ADB_SR) { /* SR interrupt */
if (via2[B] & TACK) {
printk(KERN_DEBUG "PMU: SR_INT but ack still high! (%x)\n", via2[B]);
}
/* if reading grab the byte */
if ((via1[ACR] & SR_OUT) == 0) bite = via1[SR];
/* reset TREQ and wait for TACK to go high */
via2[B] |= TREQ;
timeout = 3200;
while (!(via2[B] & TACK)) {
if (--timeout < 0) {
printk(KERN_ERR "PMU not responding (!ack)\n");
goto finish;
}
udelay(10);
}
switch (pmu_state) {
case sending:
req = current_req;
if (data_len < 0) {
data_len = req->nbytes - 1;
send_byte(data_len);
break;
}
if (data_index <= data_len) {
send_byte(req->data[data_index++]);
break;
}
req->sent = 1;
data_len = pmu_data_len[req->data[0]][1];
if (data_len == 0) {
pmu_state = idle;
current_req = req->next;
if (req->reply_expected)
req_awaiting_reply = req;
else
pmu_done(req);
} else {
pmu_state = reading;
data_index = 0;
reply_ptr = req->reply + req->reply_len;
recv_byte();
}
break;
case intack:
data_index = 0;
data_len = -1;
pmu_state = reading_intr;
reply_ptr = interrupt_data;
recv_byte();
break;
case reading:
case reading_intr:
if (data_len == -1) {
data_len = bite;
if (bite > 32)
printk(KERN_ERR "PMU: bad reply len %d\n",
bite);
} else {
reply_ptr[data_index++] = bite;
}
if (data_index < data_len) {
recv_byte();
break;
}
if (pmu_state == reading_intr) {
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 06:55:46 -07:00
pmu_handle_data(interrupt_data, data_index);
} else {
req = current_req;
current_req = req->next;
req->reply_len += data_index;
pmu_done(req);
}
pmu_state = idle;
break;
default:
printk(KERN_ERR "pmu_interrupt: unknown state %d?\n",
pmu_state);
}
}
finish:
if (pmu_state == idle) {
if (adb_int_pending) {
pmu_state = intack;
send_byte(PMU_INT_ACK);
adb_int_pending = 0;
} else if (current_req) {
pmu_start();
}
}
#if 0
printk("pmu_interrupt: exit state %d acr %02X, b %02X data_index %d/%d adb_int_pending %d\n",
pmu_state, (uint) via1[ACR], (uint) via2[B], data_index, data_len, adb_int_pending);
#endif
return IRQ_HANDLED;
}
static void
pmu_done(struct adb_request *req)
{
req->complete = 1;
if (req->done)
(*req->done)(req);
}
/* Interrupt data could be the result data from an ADB cmd */
static void
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 06:55:46 -07:00
pmu_handle_data(unsigned char *data, int len)
{
static int show_pmu_ints = 1;
asleep = 0;
if (len < 1) {
adb_int_pending = 0;
return;
}
if (data[0] & PMU_INT_ADB) {
if ((data[0] & PMU_INT_ADB_AUTO) == 0) {
struct adb_request *req = req_awaiting_reply;
if (req == 0) {
printk(KERN_ERR "PMU: extra ADB reply\n");
return;
}
req_awaiting_reply = NULL;
if (len <= 2)
req->reply_len = 0;
else {
memcpy(req->reply, data + 1, len - 1);
req->reply_len = len - 1;
}
pmu_done(req);
} else {
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 06:55:46 -07:00
adb_input(data+1, len-1, 1);
}
} else {
if (data[0] == 0x08 && len == 3) {
/* sound/brightness buttons pressed */
pmu_set_brightness(data[1] >> 3);
set_volume(data[2]);
} else if (show_pmu_ints
&& !(data[0] == PMU_INT_TICK && len == 1)) {
int i;
printk(KERN_DEBUG "pmu intr");
for (i = 0; i < len; ++i)
printk(" %.2x", data[i]);
printk("\n");
}
}
}
int backlight_level = -1;
int backlight_enabled = 0;
#define LEVEL_TO_BRIGHT(lev) ((lev) < 1? 0x7f: 0x4a - ((lev) << 1))
static void
pmu_enable_backlight(int on)
{
struct adb_request req;
if (on) {
/* first call: get current backlight value */
if (backlight_level < 0) {
switch(pmu_kind) {
case PMU_68K_V1:
case PMU_68K_V2:
pmu_request(&req, NULL, 3, PMU_READ_NVRAM, 0x14, 0xe);
while (!req.complete)
pmu_poll();
printk(KERN_DEBUG "pmu: nvram returned bright: %d\n", (int)req.reply[1]);
backlight_level = req.reply[1];
break;
default:
backlight_enabled = 0;
return;
}
}
pmu_request(&req, NULL, 2, PMU_BACKLIGHT_BRIGHT,
LEVEL_TO_BRIGHT(backlight_level));
while (!req.complete)
pmu_poll();
}
pmu_request(&req, NULL, 2, PMU_POWER_CTRL,
PMU_POW_BACKLIGHT | (on ? PMU_POW_ON : PMU_POW_OFF));
while (!req.complete)
pmu_poll();
backlight_enabled = on;
}
static void
pmu_set_brightness(int level)
{
int bright;
backlight_level = level;
bright = LEVEL_TO_BRIGHT(level);
if (!backlight_enabled)
return;
if (bright_req_1.complete)
pmu_request(&bright_req_1, NULL, 2, PMU_BACKLIGHT_BRIGHT,
bright);
if (bright_req_2.complete)
pmu_request(&bright_req_2, NULL, 2, PMU_POWER_CTRL,
PMU_POW_BACKLIGHT | (bright < 0x7f ? PMU_POW_ON : PMU_POW_OFF));
}
void
pmu_enable_irled(int on)
{
struct adb_request req;
pmu_request(&req, NULL, 2, PMU_POWER_CTRL, PMU_POW_IRLED |
(on ? PMU_POW_ON : PMU_POW_OFF));
while (!req.complete)
pmu_poll();
}
static void
set_volume(int level)
{
}
int
pmu_present(void)
{
return (pmu_kind != PMU_UNKNOWN);
}
#if 0 /* needs some work for 68K */
/*
* This struct is used to store config register values for
* PCI devices which may get powered off when we sleep.
*/
static struct pci_save {
u16 command;
u16 cache_lat;
u16 intr;
} *pbook_pci_saves;
static int n_pbook_pci_saves;
static inline void
pbook_pci_save(void)
{
int npci;
struct pci_dev *pd = NULL;
struct pci_save *ps;
npci = 0;
while ((pd = pci_get_device(PCI_ANY_ID, PCI_ANY_ID, pd)) != NULL)
++npci;
n_pbook_pci_saves = npci;
if (npci == 0)
return;
[PATCH] getting rid of all casts of k[cmz]alloc() calls Run this: #!/bin/sh for f in $(grep -Erl "\([^\)]*\) *k[cmz]alloc" *) ; do echo "De-casting $f..." perl -pi -e "s/ ?= ?\([^\)]*\) *(k[cmz]alloc) *\(/ = \1\(/" $f done And then go through and reinstate those cases where code is casting pointers to non-pointers. And then drop a few hunks which conflicted with outstanding work. Cc: Russell King <rmk@arm.linux.org.uk>, Ian Molton <spyro@f2s.com> Cc: Mikael Starvik <starvik@axis.com> Cc: Yoshinori Sato <ysato@users.sourceforge.jp> Cc: Roman Zippel <zippel@linux-m68k.org> Cc: Geert Uytterhoeven <geert@linux-m68k.org> Cc: Ralf Baechle <ralf@linux-mips.org> Cc: Paul Mackerras <paulus@samba.org> Cc: Kyle McMartin <kyle@mcmartin.ca> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Jeff Dike <jdike@addtoit.com> Cc: Greg KH <greg@kroah.com> Cc: Jens Axboe <jens.axboe@oracle.com> Cc: Paul Fulghum <paulkf@microgate.com> Cc: Alan Cox <alan@lxorguk.ukuu.org.uk> Cc: Karsten Keil <kkeil@suse.de> Cc: Mauro Carvalho Chehab <mchehab@infradead.org> Cc: Jeff Garzik <jeff@garzik.org> Cc: James Bottomley <James.Bottomley@steeleye.com> Cc: Ian Kent <raven@themaw.net> Cc: Steven French <sfrench@us.ibm.com> Cc: David Woodhouse <dwmw2@infradead.org> Cc: Neil Brown <neilb@cse.unsw.edu.au> Cc: Jaroslav Kysela <perex@suse.cz> Cc: Takashi Iwai <tiwai@suse.de> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-12-13 01:35:56 -07:00
ps = kmalloc(npci * sizeof(*ps), GFP_KERNEL);
pbook_pci_saves = ps;
if (ps == NULL)
return;
pd = NULL;
while ((pd = pci_get_device(PCI_ANY_ID, PCI_ANY_ID, pd)) != NULL) {
pci_read_config_word(pd, PCI_COMMAND, &ps->command);
pci_read_config_word(pd, PCI_CACHE_LINE_SIZE, &ps->cache_lat);
pci_read_config_word(pd, PCI_INTERRUPT_LINE, &ps->intr);
++ps;
--npci;
}
}
static inline void
pbook_pci_restore(void)
{
u16 cmd;
struct pci_save *ps = pbook_pci_saves;
struct pci_dev *pd = NULL;
int j;
while ((pd = pci_get_device(PCI_ANY_ID, PCI_ANY_ID, pd)) != NULL) {
if (ps->command == 0)
continue;
pci_read_config_word(pd, PCI_COMMAND, &cmd);
if ((ps->command & ~cmd) == 0)
continue;
switch (pd->hdr_type) {
case PCI_HEADER_TYPE_NORMAL:
for (j = 0; j < 6; ++j)
pci_write_config_dword(pd,
PCI_BASE_ADDRESS_0 + j*4,
pd->resource[j].start);
pci_write_config_dword(pd, PCI_ROM_ADDRESS,
pd->resource[PCI_ROM_RESOURCE].start);
pci_write_config_word(pd, PCI_CACHE_LINE_SIZE,
ps->cache_lat);
pci_write_config_word(pd, PCI_INTERRUPT_LINE,
ps->intr);
pci_write_config_word(pd, PCI_COMMAND, ps->command);
break;
/* other header types not restored at present */
}
}
}
/*
* Put the powerbook to sleep.
*/
#define IRQ_ENABLE ((unsigned int *)0xf3000024)
#define MEM_CTRL ((unsigned int *)0xf8000070)
int powerbook_sleep(void)
{
int ret, i, x;
static int save_backlight;
static unsigned int save_irqen;
unsigned long msr;
unsigned int hid0;
unsigned long p, wait;
struct adb_request sleep_req;
/* Notify device drivers */
[PATCH] Notifier chain update: API changes The kernel's implementation of notifier chains is unsafe. There is no protection against entries being added to or removed from a chain while the chain is in use. The issues were discussed in this thread: http://marc.theaimsgroup.com/?l=linux-kernel&m=113018709002036&w=2 We noticed that notifier chains in the kernel fall into two basic usage classes: "Blocking" chains are always called from a process context and the callout routines are allowed to sleep; "Atomic" chains can be called from an atomic context and the callout routines are not allowed to sleep. We decided to codify this distinction and make it part of the API. Therefore this set of patches introduces three new, parallel APIs: one for blocking notifiers, one for atomic notifiers, and one for "raw" notifiers (which is really just the old API under a new name). New kinds of data structures are used for the heads of the chains, and new routines are defined for registration, unregistration, and calling a chain. The three APIs are explained in include/linux/notifier.h and their implementation is in kernel/sys.c. With atomic and blocking chains, the implementation guarantees that the chain links will not be corrupted and that chain callers will not get messed up by entries being added or removed. For raw chains the implementation provides no guarantees at all; users of this API must provide their own protections. (The idea was that situations may come up where the assumptions of the atomic and blocking APIs are not appropriate, so it should be possible for users to handle these things in their own way.) There are some limitations, which should not be too hard to live with. For atomic/blocking chains, registration and unregistration must always be done in a process context since the chain is protected by a mutex/rwsem. Also, a callout routine for a non-raw chain must not try to register or unregister entries on its own chain. (This did happen in a couple of places and the code had to be changed to avoid it.) Since atomic chains may be called from within an NMI handler, they cannot use spinlocks for synchronization. Instead we use RCU. The overhead falls almost entirely in the unregister routine, which is okay since unregistration is much less frequent that calling a chain. Here is the list of chains that we adjusted and their classifications. None of them use the raw API, so for the moment it is only a placeholder. ATOMIC CHAINS ------------- arch/i386/kernel/traps.c: i386die_chain arch/ia64/kernel/traps.c: ia64die_chain arch/powerpc/kernel/traps.c: powerpc_die_chain arch/sparc64/kernel/traps.c: sparc64die_chain arch/x86_64/kernel/traps.c: die_chain drivers/char/ipmi/ipmi_si_intf.c: xaction_notifier_list kernel/panic.c: panic_notifier_list kernel/profile.c: task_free_notifier net/bluetooth/hci_core.c: hci_notifier net/ipv4/netfilter/ip_conntrack_core.c: ip_conntrack_chain net/ipv4/netfilter/ip_conntrack_core.c: ip_conntrack_expect_chain net/ipv6/addrconf.c: inet6addr_chain net/netfilter/nf_conntrack_core.c: nf_conntrack_chain net/netfilter/nf_conntrack_core.c: nf_conntrack_expect_chain net/netlink/af_netlink.c: netlink_chain BLOCKING CHAINS --------------- arch/powerpc/platforms/pseries/reconfig.c: pSeries_reconfig_chain arch/s390/kernel/process.c: idle_chain arch/x86_64/kernel/process.c idle_notifier drivers/base/memory.c: memory_chain drivers/cpufreq/cpufreq.c cpufreq_policy_notifier_list drivers/cpufreq/cpufreq.c cpufreq_transition_notifier_list drivers/macintosh/adb.c: adb_client_list drivers/macintosh/via-pmu.c sleep_notifier_list drivers/macintosh/via-pmu68k.c sleep_notifier_list drivers/macintosh/windfarm_core.c wf_client_list drivers/usb/core/notify.c usb_notifier_list drivers/video/fbmem.c fb_notifier_list kernel/cpu.c cpu_chain kernel/module.c module_notify_list kernel/profile.c munmap_notifier kernel/profile.c task_exit_notifier kernel/sys.c reboot_notifier_list net/core/dev.c netdev_chain net/decnet/dn_dev.c: dnaddr_chain net/ipv4/devinet.c: inetaddr_chain It's possible that some of these classifications are wrong. If they are, please let us know or submit a patch to fix them. Note that any chain that gets called very frequently should be atomic, because the rwsem read-locking used for blocking chains is very likely to incur cache misses on SMP systems. (However, if the chain's callout routines may sleep then the chain cannot be atomic.) The patch set was written by Alan Stern and Chandra Seetharaman, incorporating material written by Keith Owens and suggestions from Paul McKenney and Andrew Morton. [jes@sgi.com: restructure the notifier chain initialization macros] Signed-off-by: Alan Stern <stern@rowland.harvard.edu> Signed-off-by: Chandra Seetharaman <sekharan@us.ibm.com> Signed-off-by: Jes Sorensen <jes@sgi.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-03-27 02:16:30 -07:00
ret = blocking_notifier_call_chain(&sleep_notifier_list,
PBOOK_SLEEP, NULL);
if (ret & NOTIFY_STOP_MASK)
return -EBUSY;
/* Sync the disks. */
/* XXX It would be nice to have some way to ensure that
* nobody is dirtying any new buffers while we wait. */
sys_sync();
/* Turn off the display backlight */
save_backlight = backlight_enabled;
if (save_backlight)
pmu_enable_backlight(0);
/* Give the disks a little time to actually finish writing */
for (wait = jiffies + (HZ/4); time_before(jiffies, wait); )
mb();
/* Disable all interrupts except pmu */
save_irqen = in_le32(IRQ_ENABLE);
for (i = 0; i < 32; ++i)
if (i != vias->intrs[0].line && (save_irqen & (1 << i)))
disable_irq(i);
asm volatile("mtdec %0" : : "r" (0x7fffffff));
/* Save the state of PCI config space for some slots */
pbook_pci_save();
/* Set the memory controller to keep the memory refreshed
while we're asleep */
for (i = 0x403f; i >= 0x4000; --i) {
out_be32(MEM_CTRL, i);
do {
x = (in_be32(MEM_CTRL) >> 16) & 0x3ff;
} while (x == 0);
if (x >= 0x100)
break;
}
/* Ask the PMU to put us to sleep */
pmu_request(&sleep_req, NULL, 5, PMU_SLEEP, 'M', 'A', 'T', 'T');
while (!sleep_req.complete)
mb();
/* displacement-flush the L2 cache - necessary? */
for (p = KERNELBASE; p < KERNELBASE + 0x100000; p += 0x1000)
i = *(volatile int *)p;
asleep = 1;
/* Put the CPU into sleep mode */
asm volatile("mfspr %0,1008" : "=r" (hid0) :);
hid0 = (hid0 & ~(HID0_NAP | HID0_DOZE)) | HID0_SLEEP;
asm volatile("mtspr 1008,%0" : : "r" (hid0));
local_save_flags(msr);
msr |= MSR_POW | MSR_EE;
local_irq_restore(msr);
udelay(10);
/* OK, we're awake again, start restoring things */
out_be32(MEM_CTRL, 0x3f);
pbook_pci_restore();
/* wait for the PMU interrupt sequence to complete */
while (asleep)
mb();
/* reenable interrupts */
for (i = 0; i < 32; ++i)
if (i != vias->intrs[0].line && (save_irqen & (1 << i)))
enable_irq(i);
/* Notify drivers */
[PATCH] Notifier chain update: API changes The kernel's implementation of notifier chains is unsafe. There is no protection against entries being added to or removed from a chain while the chain is in use. The issues were discussed in this thread: http://marc.theaimsgroup.com/?l=linux-kernel&m=113018709002036&w=2 We noticed that notifier chains in the kernel fall into two basic usage classes: "Blocking" chains are always called from a process context and the callout routines are allowed to sleep; "Atomic" chains can be called from an atomic context and the callout routines are not allowed to sleep. We decided to codify this distinction and make it part of the API. Therefore this set of patches introduces three new, parallel APIs: one for blocking notifiers, one for atomic notifiers, and one for "raw" notifiers (which is really just the old API under a new name). New kinds of data structures are used for the heads of the chains, and new routines are defined for registration, unregistration, and calling a chain. The three APIs are explained in include/linux/notifier.h and their implementation is in kernel/sys.c. With atomic and blocking chains, the implementation guarantees that the chain links will not be corrupted and that chain callers will not get messed up by entries being added or removed. For raw chains the implementation provides no guarantees at all; users of this API must provide their own protections. (The idea was that situations may come up where the assumptions of the atomic and blocking APIs are not appropriate, so it should be possible for users to handle these things in their own way.) There are some limitations, which should not be too hard to live with. For atomic/blocking chains, registration and unregistration must always be done in a process context since the chain is protected by a mutex/rwsem. Also, a callout routine for a non-raw chain must not try to register or unregister entries on its own chain. (This did happen in a couple of places and the code had to be changed to avoid it.) Since atomic chains may be called from within an NMI handler, they cannot use spinlocks for synchronization. Instead we use RCU. The overhead falls almost entirely in the unregister routine, which is okay since unregistration is much less frequent that calling a chain. Here is the list of chains that we adjusted and their classifications. None of them use the raw API, so for the moment it is only a placeholder. ATOMIC CHAINS ------------- arch/i386/kernel/traps.c: i386die_chain arch/ia64/kernel/traps.c: ia64die_chain arch/powerpc/kernel/traps.c: powerpc_die_chain arch/sparc64/kernel/traps.c: sparc64die_chain arch/x86_64/kernel/traps.c: die_chain drivers/char/ipmi/ipmi_si_intf.c: xaction_notifier_list kernel/panic.c: panic_notifier_list kernel/profile.c: task_free_notifier net/bluetooth/hci_core.c: hci_notifier net/ipv4/netfilter/ip_conntrack_core.c: ip_conntrack_chain net/ipv4/netfilter/ip_conntrack_core.c: ip_conntrack_expect_chain net/ipv6/addrconf.c: inet6addr_chain net/netfilter/nf_conntrack_core.c: nf_conntrack_chain net/netfilter/nf_conntrack_core.c: nf_conntrack_expect_chain net/netlink/af_netlink.c: netlink_chain BLOCKING CHAINS --------------- arch/powerpc/platforms/pseries/reconfig.c: pSeries_reconfig_chain arch/s390/kernel/process.c: idle_chain arch/x86_64/kernel/process.c idle_notifier drivers/base/memory.c: memory_chain drivers/cpufreq/cpufreq.c cpufreq_policy_notifier_list drivers/cpufreq/cpufreq.c cpufreq_transition_notifier_list drivers/macintosh/adb.c: adb_client_list drivers/macintosh/via-pmu.c sleep_notifier_list drivers/macintosh/via-pmu68k.c sleep_notifier_list drivers/macintosh/windfarm_core.c wf_client_list drivers/usb/core/notify.c usb_notifier_list drivers/video/fbmem.c fb_notifier_list kernel/cpu.c cpu_chain kernel/module.c module_notify_list kernel/profile.c munmap_notifier kernel/profile.c task_exit_notifier kernel/sys.c reboot_notifier_list net/core/dev.c netdev_chain net/decnet/dn_dev.c: dnaddr_chain net/ipv4/devinet.c: inetaddr_chain It's possible that some of these classifications are wrong. If they are, please let us know or submit a patch to fix them. Note that any chain that gets called very frequently should be atomic, because the rwsem read-locking used for blocking chains is very likely to incur cache misses on SMP systems. (However, if the chain's callout routines may sleep then the chain cannot be atomic.) The patch set was written by Alan Stern and Chandra Seetharaman, incorporating material written by Keith Owens and suggestions from Paul McKenney and Andrew Morton. [jes@sgi.com: restructure the notifier chain initialization macros] Signed-off-by: Alan Stern <stern@rowland.harvard.edu> Signed-off-by: Chandra Seetharaman <sekharan@us.ibm.com> Signed-off-by: Jes Sorensen <jes@sgi.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-03-27 02:16:30 -07:00
blocking_notifier_call_chain(&sleep_notifier_list, PBOOK_WAKE, NULL);
/* reenable ADB autopoll */
pmu_adb_autopoll(adb_dev_map);
/* Turn on the screen backlight, if it was on before */
if (save_backlight)
pmu_enable_backlight(1);
/* Wait for the hard disk to spin up */
return 0;
}
/*
* Support for /dev/pmu device
*/
static int pmu_open(struct inode *inode, struct file *file)
{
return 0;
}
static ssize_t pmu_read(struct file *file, char *buf,
size_t count, loff_t *ppos)
{
return 0;
}
static ssize_t pmu_write(struct file *file, const char *buf,
size_t count, loff_t *ppos)
{
return 0;
}
static int pmu_ioctl(struct inode * inode, struct file *filp,
u_int cmd, u_long arg)
{
int error;
__u32 value;
switch (cmd) {
case PMU_IOC_SLEEP:
return -ENOSYS;
case PMU_IOC_GET_BACKLIGHT:
return put_user(backlight_level, (__u32 *)arg);
case PMU_IOC_SET_BACKLIGHT:
error = get_user(value, (__u32 *)arg);
if (!error)
pmu_set_brightness(value);
return error;
case PMU_IOC_GET_MODEL:
return put_user(pmu_kind, (__u32 *)arg);
}
return -EINVAL;
}
static struct file_operations pmu_device_fops = {
.read = pmu_read,
.write = pmu_write,
.ioctl = pmu_ioctl,
.open = pmu_open,
};
static struct miscdevice pmu_device = {
PMU_MINOR, "pmu", &pmu_device_fops
};
void pmu_device_init(void)
{
if (!via)
return;
if (misc_register(&pmu_device) < 0)
printk(KERN_ERR "via-pmu68k: cannot register misc device.\n");
}
#endif /* CONFIG_PMAC_PBOOK */