d/freebsd-dev
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Rewrite OsdSynch.c to reflect the latest ACPICA more closely:

Browse Source 
- Implement ACPI semaphore (ACPI_SEMAPHORE) with condvar(9) and mutex(9).
- Implement ACPI mutex (ACPI_MUTEX) with mutex(9).
- Implement ACPI lock (ACPI_SPINLOCK) with spin mutex(9).
This commit is contained in:
Jung-uk Kim 2009-06-08 20:07:16 +00:00
parent 5c40c5e989
commit 230bb4d90d
Notes: svn2git 2020-12-20 02:59:44 +00:00 
svn path=/head/; revision=193750
5 changed files with 457 additions and 285 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
sys/amd64/conf/NOTES
View File

@ -237,12 +237,9 @@ device tdfx # Enable 3Dfx Voodoo support
# kernel environment variables to select initial debugging levels for the
# Intel ACPICA code. (Note that the Intel code must also have USE_DEBUGGER
# defined when it is built).
#
# ACPI_NO_SEMAPHORES makes the AcpiOs*Semaphore routines a no-op.
device acpi
options ACPI_DEBUG
#!options ACPI_NO_SEMAPHORES
# The cpufreq(4) driver provides support for non-ACPI CPU frequency control
device cpufreq

1
sys/conf/options
View File

@ -664,7 +664,6 @@ WITNESS_SKIPSPIN opt_witness.h
# options for ACPI support
ACPI_DEBUG opt_acpi.h
ACPI_MAX_THREADS opt_acpi.h
ACPI_NO_SEMAPHORES opt_acpi.h
# ISA support
DEV_ISA opt_isa.h

1
sys/contrib/dev/acpica/include/platform/acfreebsd.h
View File

@ -140,6 +140,7 @@
#include "opt_acpi.h"
#define ACPI_THREAD_ID lwpid_t
#define ACPI_MUTEX_TYPE ACPI_OSL_MUTEX
#ifdef ACPI_DEBUG
#define ACPI_DEBUG_OUTPUT /* for backward compatibility */

734
sys/dev/acpica/Osd/OsdSynch.c
View File

@ -1,6 +1,7 @@
/*-
* Copyright (c) 2000 Michael Smith
* Copyright (c) 2000 BSDi
* Copyright (c) 2007-2009 Jung-uk Kim <jkim@FreeBSD.org>
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
@ -35,366 +36,543 @@ __FBSDID("$FreeBSD$");
#include <contrib/dev/acpica/include/acpi.h>
#include <contrib/dev/acpica/include/accommon.h>
#include "opt_acpi.h"
#include <sys/condvar.h>
#include <sys/kernel.h>
#include <sys/malloc.h>
#include <sys/sysctl.h>
#include <sys/lock.h>
#include <sys/malloc.h>
#include <sys/mutex.h>
#define _COMPONENT ACPI_OS_SERVICES
#define _COMPONENT ACPI_OS_SERVICES
ACPI_MODULE_NAME("SYNCH")
MALLOC_DEFINE(M_ACPISEM, "acpisem", "ACPI semaphore");
#define AS_LOCK(as) mtx_lock(&(as)->as_mtx)
#define AS_UNLOCK(as) mtx_unlock(&(as)->as_mtx)
/*
* Convert milliseconds to ticks.
*/
static int
timeout2hz(UINT16 Timeout)
{
struct timeval tv;
tv.tv_sec = (time_t)(Timeout / 1000);
tv.tv_usec = (suseconds_t)(Timeout % 1000) * 1000;
return (tvtohz(&tv));
}
/*
* Simple counting semaphore implemented using a mutex. (Subsequently used
* in the OSI code to implement a mutex. Go figure.)
* ACPI_SEMAPHORE
*/
struct acpi_semaphore {
struct mtx as_mtx;
UINT32 as_units;
UINT32 as_maxunits;
UINT32 as_pendings;
UINT32 as_resetting;
UINT32 as_timeouts;
struct acpi_sema {
struct mtx as_lock;
char as_name[32];
struct cv as_cv;
UINT32 as_maxunits;
UINT32 as_units;
int as_waiters;
int as_reset;
};
/* Default number of maximum pending threads. */
#ifndef ACPI_NO_SEMAPHORES
#ifndef ACPI_SEMAPHORES_MAX_PENDING
#define ACPI_SEMAPHORES_MAX_PENDING 4
#endif
static int acpi_semaphore_debug = 0;
TUNABLE_INT("debug.acpi_semaphore_debug", &acpi_semaphore_debug);
SYSCTL_DECL(_debug_acpi);
SYSCTL_INT(_debug_acpi, OID_AUTO, semaphore_debug, CTLFLAG_RW,
&acpi_semaphore_debug, 0, "Enable ACPI semaphore debug messages");
#endif /* !ACPI_NO_SEMAPHORES */
ACPI_STATUS
AcpiOsCreateSemaphore(UINT32 MaxUnits, UINT32 InitialUnits,
ACPI_SEMAPHORE *OutHandle)
{
#ifndef ACPI_NO_SEMAPHORES
struct acpi_semaphore *as;
struct acpi_sema *as;
ACPI_FUNCTION_TRACE((char *)(uintptr_t)__func__);
ACPI_FUNCTION_TRACE((char *)(uintptr_t)__func__);
if (OutHandle == NULL)
return_ACPI_STATUS (AE_BAD_PARAMETER);
if (InitialUnits > MaxUnits)
return_ACPI_STATUS (AE_BAD_PARAMETER);
if (OutHandle == NULL || MaxUnits == 0 || InitialUnits > MaxUnits)
return_ACPI_STATUS (AE_BAD_PARAMETER);
if ((as = malloc(sizeof(*as), M_ACPISEM, M_NOWAIT | M_ZERO)) == NULL)
return_ACPI_STATUS (AE_NO_MEMORY);
if ((as = malloc(sizeof(*as), M_ACPISEM, M_NOWAIT | M_ZERO)) == NULL)
return_ACPI_STATUS (AE_NO_MEMORY);
mtx_init(&as->as_mtx, "ACPI semaphore", NULL, MTX_DEF);
as->as_units = InitialUnits;
as->as_maxunits = MaxUnits;
as->as_pendings = as->as_resetting = as->as_timeouts = 0;
snprintf(as->as_name, sizeof(as->as_name), "ACPI sema (%p)", as);
mtx_init(&as->as_lock, as->as_name, NULL, MTX_DEF);
cv_init(&as->as_cv, as->as_name);
as->as_maxunits = MaxUnits;
as->as_units = InitialUnits;
ACPI_DEBUG_PRINT((ACPI_DB_MUTEX,
"created semaphore %p max %d, initial %d\n",
as, InitialUnits, MaxUnits));
*OutHandle = (ACPI_SEMAPHORE)as;
*OutHandle = (ACPI_HANDLE)as;
#else
*OutHandle = (ACPI_HANDLE)OutHandle;
#endif /* !ACPI_NO_SEMAPHORES */
ACPI_DEBUG_PRINT((ACPI_DB_MUTEX, "created %s, max %u, initial %u\n",
as->as_name, MaxUnits, InitialUnits));
return_ACPI_STATUS (AE_OK);
return_ACPI_STATUS (AE_OK);
}
ACPI_STATUS
AcpiOsDeleteSemaphore(ACPI_SEMAPHORE Handle)
{
#ifndef ACPI_NO_SEMAPHORES
struct acpi_semaphore *as = (struct acpi_semaphore *)Handle;
struct acpi_sema *as = (struct acpi_sema *)Handle;
ACPI_FUNCTION_TRACE((char *)(uintptr_t)__func__);
ACPI_FUNCTION_TRACE((char *)(uintptr_t)__func__);
ACPI_DEBUG_PRINT((ACPI_DB_MUTEX, "destroyed semaphore %p\n", as));
mtx_destroy(&as->as_mtx);
free(Handle, M_ACPISEM);
#endif /* !ACPI_NO_SEMAPHORES */
if (as == NULL)
return_ACPI_STATUS (AE_BAD_PARAMETER);
return_ACPI_STATUS (AE_OK);
mtx_lock(&as->as_lock);
ACPI_DEBUG_PRINT((ACPI_DB_MUTEX, "delete %s\n", as->as_name));
if (as->as_waiters > 0) {
ACPI_DEBUG_PRINT((ACPI_DB_MUTEX,
"reset %s, units %u, waiters %d\n",
as->as_name, as->as_units, as->as_waiters));
as->as_reset = 1;
cv_broadcast(&as->as_cv);
while (as->as_waiters > 0) {
if (mtx_sleep(&as->as_reset, &as->as_lock,
PCATCH, "acsrst", hz) == EINTR) {
ACPI_DEBUG_PRINT((ACPI_DB_MUTEX,
"failed to reset %s, waiters %d\n",
as->as_name, as->as_waiters));
mtx_unlock(&as->as_lock);
return_ACPI_STATUS (AE_ERROR);
}
ACPI_DEBUG_PRINT((ACPI_DB_MUTEX,
"wait %s, units %u, waiters %d\n",
as->as_name, as->as_units, as->as_waiters));
}
}
mtx_unlock(&as->as_lock);
mtx_destroy(&as->as_lock);
cv_destroy(&as->as_cv);
free(as, M_ACPISEM);
return_ACPI_STATUS (AE_OK);
}
/*
* This implementation has a bug, in that it has to stall for the entire
* timeout before it will return AE_TIME. A better implementation would
* use getmicrotime() to correctly adjust the timeout after being woken up.
*/
#define ACPISEM_AVAIL(s, u) ((s)->as_units >= (u))
ACPI_STATUS
AcpiOsWaitSemaphore(ACPI_SEMAPHORE Handle, UINT32 Units, UINT16 Timeout)
{
#ifndef ACPI_NO_SEMAPHORES
ACPI_STATUS result;
struct acpi_semaphore *as = (struct acpi_semaphore *)Handle;
int rv, tmo;
struct timeval timeouttv, currenttv, timelefttv;
struct acpi_sema *as = (struct acpi_sema *)Handle;
int error, prevtick, slptick, tmo;
ACPI_STATUS status = AE_OK;
ACPI_FUNCTION_TRACE((char *)(uintptr_t)__func__);
ACPI_FUNCTION_TRACE((char *)(uintptr_t)__func__);
if (as == NULL)
return_ACPI_STATUS (AE_BAD_PARAMETER);
if (as == NULL || Units == 0)
return_ACPI_STATUS (AE_BAD_PARAMETER);
if (cold)
return_ACPI_STATUS (AE_OK);
#if 0
if (as->as_units < Units && as->as_timeouts > 10) {
printf("%s: semaphore %p too many timeouts, resetting\n", __func__, as);
AS_LOCK(as);
as->as_units = as->as_maxunits;
if (as->as_pendings)
as->as_resetting = 1;
as->as_timeouts = 0;
wakeup(as);
AS_UNLOCK(as);
return_ACPI_STATUS (AE_TIME);
}
if (as->as_resetting)
return_ACPI_STATUS (AE_TIME);
#endif
/* a timeout of ACPI_WAIT_FOREVER means "forever" */
if (Timeout == ACPI_WAIT_FOREVER) {
tmo = 0;
timeouttv.tv_sec = ((0xffff/1000) + 1); /* cf. ACPI spec */
timeouttv.tv_usec = 0;
} else {
/* compute timeout using microseconds per tick */
tmo = (Timeout * 1000) / (1000000 / hz);
if (tmo <= 0)
tmo = 1;
timeouttv.tv_sec = Timeout / 1000;
timeouttv.tv_usec = (Timeout % 1000) * 1000;
}
/* calculate timeout value in timeval */
getmicrotime(&currenttv);
timevaladd(&timeouttv, &currenttv);
AS_LOCK(as);
ACPI_DEBUG_PRINT((ACPI_DB_MUTEX,
"get %d units from semaphore %p (has %d), timeout %d\n",
Units, as, as->as_units, Timeout));
for (;;) {
if (as->as_maxunits == ACPI_NO_UNIT_LIMIT) {
result = AE_OK;
break;
}
if (as->as_units >= Units) {
as->as_units -= Units;
result = AE_OK;
break;
}
/* limit number of pending threads */
if (as->as_pendings >= ACPI_SEMAPHORES_MAX_PENDING) {
result = AE_TIME;
break;
}
/* if timeout values of zero is specified, return immediately */
if (Timeout == 0) {
result = AE_TIME;
break;
}
mtx_lock(&as->as_lock);
ACPI_DEBUG_PRINT((ACPI_DB_MUTEX,
"semaphore blocked, calling msleep(%p, %p, %d, \"acsem\", %d)\n",
as, &as->as_mtx, PCATCH, tmo));
"get %u unit(s) from %s, units %u, waiters %d, timeout %u\n",
Units, as->as_name, as->as_units, as->as_waiters, Timeout));
as->as_pendings++;
if (acpi_semaphore_debug) {
printf("%s: Sleep %d, pending %d, semaphore %p, thread %d\n",
__func__, Timeout, as->as_pendings, as, AcpiOsGetThreadId());
if (as->as_maxunits != ACPI_NO_UNIT_LIMIT && as->as_maxunits < Units) {
mtx_unlock(&as->as_lock);
return_ACPI_STATUS (AE_LIMIT);
}
rv = msleep(as, &as->as_mtx, PCATCH, "acsem", tmo);
as->as_pendings--;
#if 0
if (as->as_resetting) {
/* semaphore reset, return immediately */
if (as->as_pendings == 0) {
as->as_resetting = 0;
}
result = AE_TIME;
break;
switch (Timeout) {
case ACPI_DO_NOT_WAIT:
if (!ACPISEM_AVAIL(as, Units))
status = AE_TIME;
break;
case ACPI_WAIT_FOREVER:
while (!ACPISEM_AVAIL(as, Units)) {
as->as_waiters++;
error = cv_wait_sig(&as->as_cv, &as->as_lock);
as->as_waiters--;
if (error == EINTR || as->as_reset) {
status = AE_ERROR;
break;
}
if (ACPISEM_AVAIL(as, Units))
break;
}
break;
default:
tmo = timeout2hz(Timeout);
while (!ACPISEM_AVAIL(as, Units)) {
prevtick = ticks;
as->as_waiters++;
error = cv_timedwait_sig(&as->as_cv, &as->as_lock, tmo);
as->as_waiters--;
if (error == EINTR || as->as_reset) {
status = AE_ERROR;
break;
}
if (ACPISEM_AVAIL(as, Units))
break;
slptick = ticks - prevtick;
if (slptick >= tmo || slptick < 0) {
status = AE_TIME;
break;
}
tmo -= slptick;
}
}
#endif
if (status == AE_OK)
as->as_units -= Units;
ACPI_DEBUG_PRINT((ACPI_DB_MUTEX, "msleep(%d) returned %d\n", tmo, rv));
if (rv == EWOULDBLOCK) {
result = AE_TIME;
break;
}
mtx_unlock(&as->as_lock);
/* check if we already awaited enough */
timelefttv = timeouttv;
getmicrotime(&currenttv);
timevalsub(&timelefttv, &currenttv);
if (timelefttv.tv_sec < 0) {
ACPI_DEBUG_PRINT((ACPI_DB_MUTEX, "await semaphore %p timeout\n",
as));
result = AE_TIME;
break;
}
/* adjust timeout for the next sleep */
tmo = (timelefttv.tv_sec * 1000000 + timelefttv.tv_usec) /
(1000000 / hz);
if (tmo <= 0)
tmo = 1;
if (acpi_semaphore_debug) {
printf("%s: Wakeup timeleft(%jd, %lu), tmo %u, sem %p, thread %d\n",
__func__, (intmax_t)timelefttv.tv_sec, timelefttv.tv_usec, tmo, as,
AcpiOsGetThreadId());
}
}
if (acpi_semaphore_debug) {
if (result == AE_TIME && Timeout > 0) {
printf("%s: Timeout %d, pending %d, semaphore %p\n",
__func__, Timeout, as->as_pendings, as);
}
if (result == AE_OK && (as->as_timeouts > 0 || as->as_pendings > 0)) {
printf("%s: Acquire %d, units %d, pending %d, sem %p, thread %d\n",
__func__, Units, as->as_units, as->as_pendings, as,
AcpiOsGetThreadId());
}
}
if (result == AE_TIME)
as->as_timeouts++;
else
as->as_timeouts = 0;
AS_UNLOCK(as);
return_ACPI_STATUS (result);
#else
return_ACPI_STATUS (AE_OK);
#endif /* !ACPI_NO_SEMAPHORES */
return_ACPI_STATUS (status);
}
ACPI_STATUS
AcpiOsSignalSemaphore(ACPI_SEMAPHORE Handle, UINT32 Units)
{
#ifndef ACPI_NO_SEMAPHORES
struct acpi_semaphore *as = (struct acpi_semaphore *)Handle;
struct acpi_sema *as = (struct acpi_sema *)Handle;
UINT32 i;
ACPI_FUNCTION_TRACE((char *)(uintptr_t)__func__);
ACPI_FUNCTION_TRACE((char *)(uintptr_t)__func__);
if (as == NULL)
return_ACPI_STATUS(AE_BAD_PARAMETER);
if (as == NULL || Units == 0)
return_ACPI_STATUS (AE_BAD_PARAMETER);
mtx_lock(&as->as_lock);
ACPI_DEBUG_PRINT((ACPI_DB_MUTEX,
"return %u units to %s, units %u, waiters %d\n",
Units, as->as_name, as->as_units, as->as_waiters));
if (as->as_maxunits != ACPI_NO_UNIT_LIMIT &&
(as->as_maxunits < Units ||
as->as_maxunits - Units < as->as_units)) {
ACPI_DEBUG_PRINT((ACPI_DB_MUTEX,
"exceeded max units %u\n", as->as_maxunits));
mtx_unlock(&as->as_lock);
return_ACPI_STATUS (AE_LIMIT);
}
AS_LOCK(as);
ACPI_DEBUG_PRINT((ACPI_DB_MUTEX,
"return %d units to semaphore %p (has %d)\n",
Units, as, as->as_units));
if (as->as_maxunits != ACPI_NO_UNIT_LIMIT) {
as->as_units += Units;
if (as->as_units > as->as_maxunits)
as->as_units = as->as_maxunits;
}
if (as->as_waiters > 0 && ACPISEM_AVAIL(as, Units))
for (i = 0; i < Units; i++)
cv_signal(&as->as_cv);
if (acpi_semaphore_debug && (as->as_timeouts > 0 || as->as_pendings > 0)) {
printf("%s: Release %d, units %d, pending %d, semaphore %p, thread %d\n",
__func__, Units, as->as_units, as->as_pendings, as, AcpiOsGetThreadId());
}
mtx_unlock(&as->as_lock);
wakeup(as);
AS_UNLOCK(as);
#endif /* !ACPI_NO_SEMAPHORES */
return_ACPI_STATUS (AE_OK);
return_ACPI_STATUS (AE_OK);
}
/* Combined mutex + mutex name storage since the latter must persist. */
struct acpi_spinlock {
struct mtx lock;
char name[32];
#undef ACPISEM_AVAIL
/*
* ACPI_MUTEX
*/
struct acpi_mutex {
struct mtx am_lock;
char am_name[32];
struct thread *am_owner;
int am_nested;
int am_waiters;
int am_reset;
};
ACPI_STATUS
AcpiOsCreateLock (ACPI_SPINLOCK *OutHandle)
AcpiOsCreateMutex(ACPI_MUTEX *OutHandle)
{
struct acpi_spinlock *h;
struct acpi_mutex *am;
if (OutHandle == NULL)
return (AE_BAD_PARAMETER);
h = malloc(sizeof(*h), M_ACPISEM, M_NOWAIT | M_ZERO);
if (h == NULL)
return (AE_NO_MEMORY);
ACPI_FUNCTION_TRACE((char *)(uintptr_t)__func__);
/* Build a unique name based on the address of the handle. */
if (OutHandle == &AcpiGbl_GpeLock)
snprintf(h->name, sizeof(h->name), "acpi subsystem GPE lock");
else if (OutHandle == &AcpiGbl_HardwareLock)
snprintf(h->name, sizeof(h->name), "acpi subsystem HW lock");
else
snprintf(h->name, sizeof(h->name), "acpi subsys %p", OutHandle);
mtx_init(&h->lock, h->name, NULL, MTX_DEF|MTX_RECURSE);
*OutHandle = (ACPI_SPINLOCK)h;
return (AE_OK);
if (OutHandle == NULL)
return_ACPI_STATUS (AE_BAD_PARAMETER);
if ((am = malloc(sizeof(*am), M_ACPISEM, M_NOWAIT | M_ZERO)) == NULL)
return_ACPI_STATUS (AE_NO_MEMORY);
snprintf(am->am_name, sizeof(am->am_name), "ACPI mutex (%p)", am);
mtx_init(&am->am_lock, am->am_name, NULL, MTX_DEF);
ACPI_DEBUG_PRINT((ACPI_DB_MUTEX, "created %s\n", am->am_name));
*OutHandle = (ACPI_MUTEX)am;
return_ACPI_STATUS (AE_OK);
}
#define ACPIMTX_AVAIL(m) ((m)->am_owner == NULL)
#define ACPIMTX_OWNED(m) ((m)->am_owner == curthread)
void
AcpiOsDeleteMutex(ACPI_MUTEX Handle)
{
struct acpi_mutex *am = (struct acpi_mutex *)Handle;
ACPI_FUNCTION_TRACE((char *)(uintptr_t)__func__);
if (am == NULL) {
ACPI_DEBUG_PRINT((ACPI_DB_MUTEX, "cannot delete null mutex\n"));
return_VOID;
}
mtx_lock(&am->am_lock);
ACPI_DEBUG_PRINT((ACPI_DB_MUTEX, "delete %s\n", am->am_name));
if (am->am_waiters > 0) {
ACPI_DEBUG_PRINT((ACPI_DB_MUTEX,
"reset %s, owner %p\n", am->am_name, am->am_owner));
am->am_reset = 1;
wakeup(am);
while (am->am_waiters > 0) {
if (mtx_sleep(&am->am_reset, &am->am_lock,
PCATCH, "acmrst", hz) == EINTR) {
ACPI_DEBUG_PRINT((ACPI_DB_MUTEX,
"failed to reset %s, waiters %d\n",
am->am_name, am->am_waiters));
mtx_unlock(&am->am_lock);
return_VOID;
}
if (ACPIMTX_AVAIL(am))
ACPI_DEBUG_PRINT((ACPI_DB_MUTEX,
"wait %s, waiters %d\n",
am->am_name, am->am_waiters));
else
ACPI_DEBUG_PRINT((ACPI_DB_MUTEX,
"wait %s, owner %p, waiters %d\n",
am->am_name, am->am_owner, am->am_waiters));
}
}
mtx_unlock(&am->am_lock);
mtx_destroy(&am->am_lock);
free(am, M_ACPISEM);
}
ACPI_STATUS
AcpiOsAcquireMutex(ACPI_MUTEX Handle, UINT16 Timeout)
{
struct acpi_mutex *am = (struct acpi_mutex *)Handle;
int error, prevtick, slptick, tmo;
ACPI_STATUS status = AE_OK;
ACPI_FUNCTION_TRACE((char *)(uintptr_t)__func__);
if (am == NULL)
return_ACPI_STATUS (AE_BAD_PARAMETER);
mtx_lock(&am->am_lock);
ACPI_DEBUG_PRINT((ACPI_DB_MUTEX, "acquire %s\n", am->am_name));
if (ACPIMTX_OWNED(am)) {
am->am_nested++;
ACPI_DEBUG_PRINT((ACPI_DB_MUTEX,
"acquire nested %s, depth %d\n",
am->am_name, am->am_nested));
mtx_unlock(&am->am_lock);
return_ACPI_STATUS (AE_OK);
}
switch (Timeout) {
case ACPI_DO_NOT_WAIT:
if (!ACPIMTX_AVAIL(am))
status = AE_TIME;
break;
case ACPI_WAIT_FOREVER:
while (!ACPIMTX_AVAIL(am)) {
am->am_waiters++;
error = mtx_sleep(am, &am->am_lock, PCATCH, "acmtx", 0);
am->am_waiters--;
if (error == EINTR || am->am_reset) {
status = AE_ERROR;
break;
}
if (ACPIMTX_AVAIL(am))
break;
}
break;
default:
tmo = timeout2hz(Timeout);
while (!ACPIMTX_AVAIL(am)) {
prevtick = ticks;
am->am_waiters++;
error = mtx_sleep(am, &am->am_lock, PCATCH,
"acmtx", tmo);
am->am_waiters--;
if (error == EINTR || am->am_reset) {
status = AE_ERROR;
break;
}
if (ACPIMTX_AVAIL(am))
break;
slptick = ticks - prevtick;
if (slptick >= tmo || slptick < 0) {
status = AE_TIME;
break;
}
tmo -= slptick;
}
}
if (status == AE_OK)
am->am_owner = curthread;
mtx_unlock(&am->am_lock);
return_ACPI_STATUS (status);
}
void
AcpiOsDeleteLock (ACPI_SPINLOCK Handle)
AcpiOsReleaseMutex(ACPI_MUTEX Handle)
{
struct acpi_spinlock *h = (struct acpi_spinlock *)Handle;
struct acpi_mutex *am = (struct acpi_mutex *)Handle;
if (Handle == NULL)
return;
mtx_destroy(&h->lock);
free(h, M_ACPISEM);
ACPI_FUNCTION_TRACE((char *)(uintptr_t)__func__);
if (am == NULL)
ACPI_DEBUG_PRINT((ACPI_DB_MUTEX,
"cannot release null mutex\n"));
mtx_lock(&am->am_lock);
ACPI_DEBUG_PRINT((ACPI_DB_MUTEX, "release %s\n", am->am_name));
if (ACPIMTX_OWNED(am)) {
if (am->am_nested > 0) {
ACPI_DEBUG_PRINT((ACPI_DB_MUTEX,
"release nested %s, depth %d\n",
am->am_name, am->am_nested));
am->am_nested--;
} else
am->am_owner = NULL;
} else {
if (ACPIMTX_AVAIL(am))
ACPI_DEBUG_PRINT((ACPI_DB_MUTEX,
"release already available %s\n", am->am_name));
else
ACPI_DEBUG_PRINT((ACPI_DB_MUTEX,
"release unowned %s from %p, depth %d\n",
am->am_name, am->am_owner, am->am_nested));
}
if (am->am_waiters > 0 && ACPIMTX_AVAIL(am))
wakeup_one(am);
mtx_unlock(&am->am_lock);
}
#undef ACPIMTX_AVAIL
#undef ACPIMTX_OWNED
/*
* The Flags parameter seems to state whether or not caller is an ISR
* (and thus can't block) but since we have ithreads, we don't worry
* about potentially blocking.
* ACPI_SPINLOCK
*/
ACPI_CPU_FLAGS
AcpiOsAcquireLock (ACPI_SPINLOCK Handle)
{
struct acpi_spinlock *h = (struct acpi_spinlock *)Handle;
struct acpi_spinlock {
struct mtx al_lock;
char al_name[32];
int al_nested;
};
if (Handle == NULL)
return (0);
mtx_lock(&h->lock);
return (0);
ACPI_STATUS
AcpiOsCreateLock(ACPI_SPINLOCK *OutHandle)
{
struct acpi_spinlock *al;
ACPI_FUNCTION_TRACE((char *)(uintptr_t)__func__);
if (OutHandle == NULL)
return_ACPI_STATUS (AE_BAD_PARAMETER);
if ((al = malloc(sizeof(*al), M_ACPISEM, M_NOWAIT | M_ZERO)) == NULL)
return_ACPI_STATUS (AE_NO_MEMORY);
#ifdef ACPI_DEBUG
if (OutHandle == &AcpiGbl_GpeLock)
snprintf(al->al_name, sizeof(al->al_name), "ACPI lock (GPE)");
else if (OutHandle == &AcpiGbl_HardwareLock)
snprintf(al->al_name, sizeof(al->al_name), "ACPI lock (HW)");
else
#endif
snprintf(al->al_name, sizeof(al->al_name), "ACPI lock (%p)", al);
mtx_init(&al->al_lock, al->al_name, NULL, MTX_SPIN);
ACPI_DEBUG_PRINT((ACPI_DB_MUTEX, "created %s\n", al->al_name));
*OutHandle = (ACPI_SPINLOCK)al;
return_ACPI_STATUS (AE_OK);
}
void
AcpiOsReleaseLock (ACPI_SPINLOCK Handle, ACPI_CPU_FLAGS Flags)
AcpiOsDeleteLock(ACPI_SPINLOCK Handle)
{
struct acpi_spinlock *h = (struct acpi_spinlock *)Handle;
struct acpi_spinlock *al = (struct acpi_spinlock *)Handle;
if (Handle == NULL)
return;
mtx_unlock(&h->lock);
ACPI_FUNCTION_TRACE((char *)(uintptr_t)__func__);
if (al == NULL) {
ACPI_DEBUG_PRINT((ACPI_DB_MUTEX,
"cannot delete null spinlock\n"));
return_VOID;
}
ACPI_DEBUG_PRINT((ACPI_DB_MUTEX, "delete %s\n", al->al_name));
mtx_destroy(&al->al_lock);
free(al, M_ACPISEM);
}
/* Section 5.2.9.1: global lock acquire/release functions */
#define GL_ACQUIRED (-1)
#define GL_BUSY 0
#define GL_BIT_PENDING 0x1
#define GL_BIT_OWNED 0x2
#define GL_BIT_MASK (GL_BIT_PENDING | GL_BIT_OWNED)
ACPI_CPU_FLAGS
AcpiOsAcquireLock(ACPI_SPINLOCK Handle)
{
struct acpi_spinlock *al = (struct acpi_spinlock *)Handle;
ACPI_FUNCTION_TRACE((char *)(uintptr_t)__func__);
if (al == NULL) {
ACPI_DEBUG_PRINT((ACPI_DB_MUTEX,
"cannot acquire null spinlock\n"));
return (0);
}
ACPI_DEBUG_PRINT((ACPI_DB_MUTEX, "acquire %s\n", al->al_name));
if (mtx_owned(&al->al_lock)) {
al->al_nested++;
ACPI_DEBUG_PRINT((ACPI_DB_MUTEX,
"acquire nested %s, depth %d\n",
al->al_name, al->al_nested));
} else
mtx_lock_spin(&al->al_lock);
return (0);
}
void
AcpiOsReleaseLock(ACPI_SPINLOCK Handle, ACPI_CPU_FLAGS Flags)
{
struct acpi_spinlock *al = (struct acpi_spinlock *)Handle;
ACPI_FUNCTION_TRACE((char *)(uintptr_t)__func__);
if (al == NULL) {
ACPI_DEBUG_PRINT((ACPI_DB_MUTEX,
"cannot release null spinlock\n"));
return_VOID;
}
ACPI_DEBUG_PRINT((ACPI_DB_MUTEX, "release %s\n", al->al_name));
if (mtx_owned(&al->al_lock)) {
if (al->al_nested > 0) {
ACPI_DEBUG_PRINT((ACPI_DB_MUTEX,
"release nested %s, depth %d\n",
al->al_name, al->al_nested));
al->al_nested--;
} else
mtx_unlock_spin(&al->al_lock);
} else
ACPI_DEBUG_PRINT((ACPI_DB_MUTEX,
"cannot release unowned %s\n", al->al_name));
}
/* Section 5.2.10.1: global lock acquire/release functions */
#define GL_ACQUIRED (-1)
#define GL_BUSY 0
#define GL_BIT_PENDING 0x01
#define GL_BIT_OWNED 0x02
#define GL_BIT_MASK (GL_BIT_PENDING | GL_BIT_OWNED)
/*
* Acquire the global lock. If busy, set the pending bit. The caller
@ -404,7 +582,7 @@ AcpiOsReleaseLock (ACPI_SPINLOCK Handle, ACPI_CPU_FLAGS Flags)
int
acpi_acquire_global_lock(uint32_t *lock)
{
uint32_t new, old;
uint32_t new, old;
do {
old = *lock;
@ -423,7 +601,7 @@ acpi_acquire_global_lock(uint32_t *lock)
int
acpi_release_global_lock(uint32_t *lock)
{
uint32_t new, old;
uint32_t new, old;
do {
old = *lock;

3
sys/i386/conf/NOTES
View File

@ -465,14 +465,11 @@ device tdfx_linux # Enable Linuxulator support
# Intel ACPICA code. (Note that the Intel code must also have USE_DEBUGGER
# defined when it is built).
#
# ACPI_NO_SEMAPHORES makes the AcpiOs*Semaphore routines a no-op.
#
# Note that building ACPI into the kernel is deprecated; the module is
# normally loaded automatically by the loader.
device acpi
options ACPI_DEBUG
#!options ACPI_NO_SEMAPHORES
# ACPI Asus Desktop Extras. (voltage, temp, fan)
device acpi_aiboost