freebsd-dev/sys/dev/acpica/acpi_throttle.c

444 lines
13 KiB
C
Raw Normal View History

/*-
* Copyright (c) 2003-2005 Nate Lawson (SDG)
* Copyright (c) 2001 Michael Smith
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
*
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*/
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
#include "opt_acpi.h"
#include <sys/param.h>
#include <sys/bus.h>
#include <sys/cpu.h>
#include <sys/kernel.h>
#include <sys/malloc.h>
#include <sys/module.h>
#include <sys/rman.h>
#include <machine/bus.h>
2009-06-05 18:44:36 +00:00
#include <contrib/dev/acpica/include/acpi.h>
#include <dev/acpica/acpivar.h>
#include <dev/pci/pcivar.h>
#include "cpufreq_if.h"
/*
* Throttling provides relative frequency control. It involves modulating
* the clock so that the CPU is active for only a fraction of the normal
* clock cycle. It does not change voltage and so is less efficient than
* other mechanisms. Since it is relative, it can be used in addition to
* absolute cpufreq drivers. We support the ACPI 2.0 specification.
*/
struct acpi_throttle_softc {
device_t cpu_dev;
ACPI_HANDLE cpu_handle;
uint32_t cpu_p_blk; /* ACPI P_BLK location */
uint32_t cpu_p_blk_len; /* P_BLK length (must be 6). */
struct resource *cpu_p_cnt; /* Throttling control register */
int cpu_p_type; /* Resource type for cpu_p_cnt. */
uint32_t cpu_thr_state; /* Current throttle setting. */
};
#define THR_GET_REG(reg) \
(bus_space_read_4(rman_get_bustag((reg)), \
rman_get_bushandle((reg)), 0))
#define THR_SET_REG(reg, val) \
(bus_space_write_4(rman_get_bustag((reg)), \
rman_get_bushandle((reg)), 0, (val)))
/*
* Speeds are stored in counts, from 1 to CPU_MAX_SPEED, and
* reported to the user in hundredths of a percent.
*/
#define CPU_MAX_SPEED (1 << cpu_duty_width)
#define CPU_SPEED_PERCENT(x) ((10000 * (x)) / CPU_MAX_SPEED)
#define CPU_SPEED_PRINTABLE(x) (CPU_SPEED_PERCENT(x) / 10), \
(CPU_SPEED_PERCENT(x) % 10)
#define CPU_P_CNT_THT_EN (1<<4)
#define CPU_QUIRK_NO_THROTTLE (1<<1) /* Throttling is not usable. */
#define PCI_VENDOR_INTEL 0x8086
#define PCI_DEVICE_82371AB_3 0x7113 /* PIIX4 chipset for quirks. */
#define PCI_REVISION_A_STEP 0
#define PCI_REVISION_B_STEP 1
static uint32_t cpu_duty_offset; /* Offset in P_CNT of throttle val. */
static uint32_t cpu_duty_width; /* Bit width of throttle value. */
static int thr_rid; /* Driver-wide resource id. */
static int thr_quirks; /* Indicate any hardware bugs. */
static void acpi_throttle_identify(driver_t *driver, device_t parent);
static int acpi_throttle_probe(device_t dev);
static int acpi_throttle_attach(device_t dev);
static int acpi_throttle_evaluate(struct acpi_throttle_softc *sc);
static int acpi_throttle_quirks(struct acpi_throttle_softc *sc);
static int acpi_thr_settings(device_t dev, struct cf_setting *sets,
int *count);
static int acpi_thr_set(device_t dev, const struct cf_setting *set);
static int acpi_thr_get(device_t dev, struct cf_setting *set);
static int acpi_thr_type(device_t dev, int *type);
static device_method_t acpi_throttle_methods[] = {
/* Device interface */
DEVMETHOD(device_identify, acpi_throttle_identify),
DEVMETHOD(device_probe, acpi_throttle_probe),
DEVMETHOD(device_attach, acpi_throttle_attach),
/* cpufreq interface */
DEVMETHOD(cpufreq_drv_set, acpi_thr_set),
DEVMETHOD(cpufreq_drv_get, acpi_thr_get),
DEVMETHOD(cpufreq_drv_type, acpi_thr_type),
DEVMETHOD(cpufreq_drv_settings, acpi_thr_settings),
DEVMETHOD_END
};
static driver_t acpi_throttle_driver = {
"acpi_throttle",
acpi_throttle_methods,
sizeof(struct acpi_throttle_softc),
};
static devclass_t acpi_throttle_devclass;
DRIVER_MODULE(acpi_throttle, cpu, acpi_throttle_driver, acpi_throttle_devclass,
0, 0);
static void
acpi_throttle_identify(driver_t *driver, device_t parent)
{
ACPI_BUFFER buf;
ACPI_HANDLE handle;
ACPI_OBJECT *obj;
/* Make sure we're not being doubly invoked. */
if (device_find_child(parent, "acpi_throttle", -1))
return;
/* Check for a valid duty width and parent CPU type. */
handle = acpi_get_handle(parent);
if (handle == NULL)
return;
2007-03-22 18:16:43 +00:00
if (AcpiGbl_FADT.DutyWidth == 0 ||
acpi_get_type(parent) != ACPI_TYPE_PROCESSOR)
return;
/*
* Add a child if there's a non-NULL P_BLK and correct length, or
* if the _PTC method is present.
*/
buf.Pointer = NULL;
buf.Length = ACPI_ALLOCATE_BUFFER;
if (ACPI_FAILURE(AcpiEvaluateObject(handle, NULL, NULL, &buf)))
return;
obj = (ACPI_OBJECT *)buf.Pointer;
if ((obj->Processor.PblkAddress && obj->Processor.PblkLength >= 4) ||
ACPI_SUCCESS(AcpiEvaluateObject(handle, "_PTC", NULL, NULL))) {
if (BUS_ADD_CHILD(parent, 0, "acpi_throttle", -1) == NULL)
device_printf(parent, "add throttle child failed\n");
}
AcpiOsFree(obj);
}
static int
acpi_throttle_probe(device_t dev)
{
if (resource_disabled("acpi_throttle", 0))
return (ENXIO);
/*
* On i386 platforms at least, ACPI throttling is accomplished by
* the chipset modulating the STPCLK# pin based on the duty cycle.
* Since p4tcc uses the same mechanism (but internal to the CPU),
* we disable acpi_throttle when p4tcc is also present.
*/
if (device_find_child(device_get_parent(dev), "p4tcc", -1) &&
!resource_disabled("p4tcc", 0))
return (ENXIO);
device_set_desc(dev, "ACPI CPU Throttling");
return (0);
}
static int
acpi_throttle_attach(device_t dev)
{
struct acpi_throttle_softc *sc;
struct cf_setting set;
ACPI_BUFFER buf;
ACPI_OBJECT *obj;
ACPI_STATUS status;
int error;
sc = device_get_softc(dev);
sc->cpu_dev = dev;
sc->cpu_handle = acpi_get_handle(dev);
buf.Pointer = NULL;
buf.Length = ACPI_ALLOCATE_BUFFER;
status = AcpiEvaluateObject(sc->cpu_handle, NULL, NULL, &buf);
if (ACPI_FAILURE(status)) {
device_printf(dev, "attach failed to get Processor obj - %s\n",
AcpiFormatException(status));
return (ENXIO);
}
obj = (ACPI_OBJECT *)buf.Pointer;
sc->cpu_p_blk = obj->Processor.PblkAddress;
sc->cpu_p_blk_len = obj->Processor.PblkLength;
AcpiOsFree(obj);
/* If this is the first device probed, check for quirks. */
if (device_get_unit(dev) == 0)
acpi_throttle_quirks(sc);
/* Attempt to attach the actual throttling register. */
error = acpi_throttle_evaluate(sc);
if (error)
return (error);
/*
* Set our initial frequency to the highest since some systems
* seem to boot with this at the lowest setting.
*/
set.freq = 10000;
acpi_thr_set(dev, &set);
/* Everything went ok, register with cpufreq(4). */
cpufreq_register(dev);
return (0);
}
static int
acpi_throttle_evaluate(struct acpi_throttle_softc *sc)
{
uint32_t duty_end;
ACPI_BUFFER buf;
ACPI_OBJECT obj;
ACPI_GENERIC_ADDRESS gas;
ACPI_STATUS status;
/* Get throttling parameters from the FADT. 0 means not supported. */
if (device_get_unit(sc->cpu_dev) == 0) {
2007-03-22 18:16:43 +00:00
cpu_duty_offset = AcpiGbl_FADT.DutyOffset;
cpu_duty_width = AcpiGbl_FADT.DutyWidth;
}
if (cpu_duty_width == 0 || (thr_quirks & CPU_QUIRK_NO_THROTTLE) != 0)
return (ENXIO);
/* Validate the duty offset/width. */
duty_end = cpu_duty_offset + cpu_duty_width - 1;
if (duty_end > 31) {
device_printf(sc->cpu_dev,
"CLK_VAL field overflows P_CNT register\n");
return (ENXIO);
}
if (cpu_duty_offset <= 4 && duty_end >= 4) {
device_printf(sc->cpu_dev,
"CLK_VAL field overlaps THT_EN bit\n");
return (ENXIO);
}
/*
* If not present, fall back to using the processor's P_BLK to find
* the P_CNT register.
*
* Note that some systems seem to duplicate the P_BLK pointer
* across multiple CPUs, so not getting the resource is not fatal.
*/
buf.Pointer = &obj;
buf.Length = sizeof(obj);
status = AcpiEvaluateObject(sc->cpu_handle, "_PTC", NULL, &buf);
if (ACPI_SUCCESS(status)) {
if (obj.Buffer.Length < sizeof(ACPI_GENERIC_ADDRESS) + 3) {
device_printf(sc->cpu_dev, "_PTC buffer too small\n");
return (ENXIO);
}
memcpy(&gas, obj.Buffer.Pointer + 3, sizeof(gas));
acpi_bus_alloc_gas(sc->cpu_dev, &sc->cpu_p_type, &thr_rid,
Re-work Cx handling to be per-cpu and asymmetrical, fixing support on modern dual-core systems as well. - Parse the _CST packages for each cpu and track all the states individually, on a per-cpu basis. - Revert to generic FADT/P_BLK based Cx control if the _CST package is not present on all cpus. In that case, the new driver will still support per-cpu Cx state handling. The driver will determine the highest Cx level that can be supported by all the cpus and configure the available Cx state based on that. - Fixed the case where multiple cpus in the system share the same registers for Cx state handling. To do that, added a new flag parameter to the acpi_PkgGas and acpi_bus_alloc_gas functions that enable the caller to add the RF_SHAREABLE flag. This flag could also be useful to other callers (acpi_throttle?) in the tree but this change is not yet made. - For Core Duo cpus, both cores seems to be taken out of C3 state when any one of the cores need to transition out. This broke the short sleep detection logic. It is disabled now if there is more than one cpu in the system for now as it fixed it in my case. This quirk may need to be re-enabled later differently. - Added support to control cx_lowest on a per-cpu basis. There is still a generic cx_lowest to enable changing cx_lowest for all cpus with a single sysctl and for ease of use. Sample output for the new sysctl: dev.cpu.0.cx_supported: C1/1 C2/1 C3/57 dev.cpu.0.cx_lowest: C3 dev.cpu.0.cx_usage: 0.00% 43.16% 56.83% dev.cpu.1.cx_supported: C1/1 C2/1 C3/57 dev.cpu.1.cx_lowest: C3 dev.cpu.1.cx_usage: 0.00% 45.65% 54.34% hw.acpi.cpu.cx_lowest: C3 This work was done by Stephane E. Potvin with some simple reworking by myself. Thank you. Submitted by: Stephane E. Potvin <sepotvin / videotron.ca> MFC after: 2 weeks
2007-01-07 21:53:42 +00:00
&gas, &sc->cpu_p_cnt, 0);
if (sc->cpu_p_cnt != NULL && bootverbose) {
device_printf(sc->cpu_dev, "P_CNT from _PTC %#jx\n",
gas.Address);
}
}
/* If _PTC not present or other failure, try the P_BLK. */
if (sc->cpu_p_cnt == NULL) {
/*
* The spec says P_BLK must be 6 bytes long. However, some
* systems use it to indicate a fractional set of features
* present so we take anything >= 4.
*/
if (sc->cpu_p_blk_len < 4)
return (ENXIO);
gas.Address = sc->cpu_p_blk;
2007-03-22 18:16:43 +00:00
gas.SpaceId = ACPI_ADR_SPACE_SYSTEM_IO;
gas.BitWidth = 32;
acpi_bus_alloc_gas(sc->cpu_dev, &sc->cpu_p_type, &thr_rid,
Re-work Cx handling to be per-cpu and asymmetrical, fixing support on modern dual-core systems as well. - Parse the _CST packages for each cpu and track all the states individually, on a per-cpu basis. - Revert to generic FADT/P_BLK based Cx control if the _CST package is not present on all cpus. In that case, the new driver will still support per-cpu Cx state handling. The driver will determine the highest Cx level that can be supported by all the cpus and configure the available Cx state based on that. - Fixed the case where multiple cpus in the system share the same registers for Cx state handling. To do that, added a new flag parameter to the acpi_PkgGas and acpi_bus_alloc_gas functions that enable the caller to add the RF_SHAREABLE flag. This flag could also be useful to other callers (acpi_throttle?) in the tree but this change is not yet made. - For Core Duo cpus, both cores seems to be taken out of C3 state when any one of the cores need to transition out. This broke the short sleep detection logic. It is disabled now if there is more than one cpu in the system for now as it fixed it in my case. This quirk may need to be re-enabled later differently. - Added support to control cx_lowest on a per-cpu basis. There is still a generic cx_lowest to enable changing cx_lowest for all cpus with a single sysctl and for ease of use. Sample output for the new sysctl: dev.cpu.0.cx_supported: C1/1 C2/1 C3/57 dev.cpu.0.cx_lowest: C3 dev.cpu.0.cx_usage: 0.00% 43.16% 56.83% dev.cpu.1.cx_supported: C1/1 C2/1 C3/57 dev.cpu.1.cx_lowest: C3 dev.cpu.1.cx_usage: 0.00% 45.65% 54.34% hw.acpi.cpu.cx_lowest: C3 This work was done by Stephane E. Potvin with some simple reworking by myself. Thank you. Submitted by: Stephane E. Potvin <sepotvin / videotron.ca> MFC after: 2 weeks
2007-01-07 21:53:42 +00:00
&gas, &sc->cpu_p_cnt, 0);
if (sc->cpu_p_cnt != NULL) {
if (bootverbose)
device_printf(sc->cpu_dev,
"P_CNT from P_BLK %#x\n", sc->cpu_p_blk);
} else {
device_printf(sc->cpu_dev, "failed to attach P_CNT\n");
return (ENXIO);
}
}
thr_rid++;
return (0);
}
static int
acpi_throttle_quirks(struct acpi_throttle_softc *sc)
{
device_t acpi_dev;
/* Look for various quirks of the PIIX4 part. */
acpi_dev = pci_find_device(PCI_VENDOR_INTEL, PCI_DEVICE_82371AB_3);
if (acpi_dev) {
switch (pci_get_revid(acpi_dev)) {
/*
* Disable throttling control on PIIX4 A and B-step.
* See specification changes #13 ("Manual Throttle Duty Cycle")
* and #14 ("Enabling and Disabling Manual Throttle"), plus
* erratum #5 ("STPCLK# Deassertion Time") from the January
* 2002 PIIX4 specification update. Note that few (if any)
* mobile systems ever used this part.
*/
case PCI_REVISION_A_STEP:
case PCI_REVISION_B_STEP:
thr_quirks |= CPU_QUIRK_NO_THROTTLE;
break;
default:
break;
}
}
return (0);
}
static int
acpi_thr_settings(device_t dev, struct cf_setting *sets, int *count)
{
int i, speed;
if (sets == NULL || count == NULL)
return (EINVAL);
if (*count < CPU_MAX_SPEED)
return (E2BIG);
/* Return a list of valid settings for this driver. */
memset(sets, CPUFREQ_VAL_UNKNOWN, sizeof(*sets) * CPU_MAX_SPEED);
for (i = 0, speed = CPU_MAX_SPEED; speed != 0; i++, speed--) {
sets[i].freq = CPU_SPEED_PERCENT(speed);
sets[i].dev = dev;
}
*count = CPU_MAX_SPEED;
return (0);
}
static int
acpi_thr_set(device_t dev, const struct cf_setting *set)
{
struct acpi_throttle_softc *sc;
uint32_t clk_val, p_cnt, speed;
if (set == NULL)
return (EINVAL);
sc = device_get_softc(dev);
/*
* Validate requested state converts to a duty cycle that is an
* integer from [1 .. CPU_MAX_SPEED].
*/
speed = set->freq * CPU_MAX_SPEED / 10000;
if (speed * 10000 != set->freq * CPU_MAX_SPEED ||
speed < 1 || speed > CPU_MAX_SPEED)
return (EINVAL);
/* If we're at this setting, don't bother applying it again. */
if (speed == sc->cpu_thr_state)
return (0);
/* Get the current P_CNT value and disable throttling */
p_cnt = THR_GET_REG(sc->cpu_p_cnt);
p_cnt &= ~CPU_P_CNT_THT_EN;
THR_SET_REG(sc->cpu_p_cnt, p_cnt);
/* If we're at maximum speed, that's all */
if (speed < CPU_MAX_SPEED) {
/* Mask the old CLK_VAL off and OR in the new value */
clk_val = (CPU_MAX_SPEED - 1) << cpu_duty_offset;
p_cnt &= ~clk_val;
p_cnt |= (speed << cpu_duty_offset);
/* Write the new P_CNT value and then enable throttling */
THR_SET_REG(sc->cpu_p_cnt, p_cnt);
p_cnt |= CPU_P_CNT_THT_EN;
THR_SET_REG(sc->cpu_p_cnt, p_cnt);
}
sc->cpu_thr_state = speed;
return (0);
}
static int
acpi_thr_get(device_t dev, struct cf_setting *set)
{
struct acpi_throttle_softc *sc;
uint32_t p_cnt, clk_val;
if (set == NULL)
return (EINVAL);
sc = device_get_softc(dev);
/* Get the current throttling setting from P_CNT. */
p_cnt = THR_GET_REG(sc->cpu_p_cnt);
clk_val = (p_cnt >> cpu_duty_offset) & (CPU_MAX_SPEED - 1);
sc->cpu_thr_state = clk_val;
memset(set, CPUFREQ_VAL_UNKNOWN, sizeof(*set));
set->freq = CPU_SPEED_PERCENT(clk_val);
set->dev = dev;
return (0);
}
static int
acpi_thr_type(device_t dev, int *type)
{
if (type == NULL)
return (EINVAL);
*type = CPUFREQ_TYPE_RELATIVE;
return (0);
}