linux_dsm_epyc7002/arch/arm/mach-omap2/pm24xx.c

588 lines
15 KiB
C
Raw Normal View History

/*
* OMAP2 Power Management Routines
*
* Copyright (C) 2005 Texas Instruments, Inc.
* Copyright (C) 2006-2008 Nokia Corporation
*
* Written by:
* Richard Woodruff <r-woodruff2@ti.com>
* Tony Lindgren
* Juha Yrjola
* Amit Kucheria <amit.kucheria@nokia.com>
* Igor Stoppa <igor.stoppa@nokia.com>
*
* Based on pm.c for omap1
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License version 2 as
* published by the Free Software Foundation.
*/
#include <linux/suspend.h>
#include <linux/sched.h>
#include <linux/proc_fs.h>
#include <linux/interrupt.h>
#include <linux/sysfs.h>
#include <linux/module.h>
#include <linux/delay.h>
#include <linux/clk.h>
#include <linux/io.h>
#include <linux/irq.h>
#include <linux/time.h>
#include <linux/gpio.h>
OMAP2+: PM/serial: hold console semaphore while OMAP UARTs are disabled The console semaphore must be held while the OMAP UART devices are disabled, lest a console write cause an ARM abort (and a kernel crash) when the underlying console device is inaccessible. These crashes only occur when the console is on one of the OMAP internal serial ports. While this problem has been latent in the PM idle loop for some time, the crash was not triggerable with an unmodified kernel until commit 6f251e9db1093c187addc309b5f2f7fe3efd2995 ("OMAP: UART: omap_device conversions, remove implicit 8520 assumptions"). After this patch, a console write often occurs after the console UART has been disabled in the idle loop, crashing the system. Several users have encountered this bug: http://www.mail-archive.com/linux-omap@vger.kernel.org/msg38396.html http://www.mail-archive.com/linux-omap@vger.kernel.org/msg36602.html The same commit also introduced new code that disabled the UARTs during init, in omap_serial_init_port(). The kernel will also crash in this code when earlyconsole and extra debugging is enabled: http://www.mail-archive.com/linux-omap@vger.kernel.org/msg36411.html The minimal fix for the -rc series is to hold the console semaphore while the OMAP UARTs are disabled. This is a somewhat overbroad fix, since the console may not be located on an OMAP UART, as is the case with the GPMC UART on Zoom3. While it is technically possible to determine which devices the console or earlyconsole is actually running on, it is not a trivial problem to solve, and the code to do so is not really appropriate for the -rc series. The right long-term fix is to ensure that no code outside of the OMAP serial driver can disable an OMAP UART. As I understand it, code to implement this is under development by TI. This patch is a collaboration between Paul Walmsley <paul@pwsan.com> and Tony Lindgren <tony@atomide.com>. Thanks to Ming Lei <tom.leiming@gmail.com> and Pramod <pramod.gurav@ti.com> for their feedback on earlier versions of this patch. Signed-off-by: Paul Walmsley <paul@pwsan.com> Signed-off-by: Tony Lindgren <tony@atomide.com> Acked-by: Kevin Hilman <khilman@deeprootsystems.com> Cc: Ming Lei <tom.leiming@gmail.com> Cc: Pramod <pramod.gurav@ti.com> Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com> Cc: Jean Pihet <jean.pihet@newoldbits.com> Cc: Govindraj.R <govindraj.raja@ti.com>
2010-11-25 06:49:05 +07:00
#include <linux/console.h>
#include <asm/mach/time.h>
#include <asm/mach/irq.h>
#include <asm/mach-types.h>
#include <mach/irqs.h>
#include <plat/clock.h>
#include <plat/sram.h>
#include <plat/dma.h>
#include <plat/board.h>
#include "prm2xxx_3xxx.h"
#include "prm-regbits-24xx.h"
#include "cm2xxx_3xxx.h"
#include "cm-regbits-24xx.h"
#include "sdrc.h"
#include "pm.h"
#include "control.h"
#include "powerdomain.h"
#include "clockdomain.h"
#ifdef CONFIG_SUSPEND
static suspend_state_t suspend_state = PM_SUSPEND_ON;
static inline bool is_suspending(void)
{
return (suspend_state != PM_SUSPEND_ON);
}
#else
static inline bool is_suspending(void)
{
return false;
}
#endif
static void (*omap2_sram_idle)(void);
static void (*omap2_sram_suspend)(u32 dllctrl, void __iomem *sdrc_dlla_ctrl,
void __iomem *sdrc_power);
OMAP clockdomains: add usecounting for wakeup and sleep dependencies Add usecounting for wakeup and sleep dependencies. In the current situation, if several functions add dependencies on the same clockdomains, when the first dependency removal function is called, the dependency will be incorrectly removed from the hardware. Add clkdm_clear_all_wkdeps() and clkdm_clear_all_sleepdeps(), which provide a fast and usecounting-consistent way to clear all hardware clockdomain dependencies, since accesses to these registers can be quite slow. pm{2,3}4xx.c has been updated to use these new functions. The original version of this patch did not touch these files, which previously wrote directly to the wkdep registers, and thus confused the usecounting code. This problem was found by Kevin Hilman <khilman@deeprootsystems.com>. N.B.: This patch introduces one significant functional difference over the previous pm34xx.c code: sleepdeps are now cleared during clockdomain initialization, whereas previously they were left untouched. This has been tested by Kevin and confirmed to work. The original version of this patch also did not take into consideration that some clockdomains do not have sleep or wakeup dependency sources, which caused NULL pointer dereferences. This problem was debugged and fixed by Kevin Hilman <khilman@deeprootsystems.com>. Signed-off-by: Paul Walmsley <paul@pwsan.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Cc: Jouni Högander <jouni.hogander@nokia.com>
2010-01-27 10:13:01 +07:00
static struct powerdomain *mpu_pwrdm, *core_pwrdm;
static struct clockdomain *dsp_clkdm, *mpu_clkdm, *wkup_clkdm, *gfx_clkdm;
static struct clk *osc_ck, *emul_ck;
static int omap2_fclks_active(void)
{
u32 f1, f2;
f1 = omap2_cm_read_mod_reg(CORE_MOD, CM_FCLKEN1);
f2 = omap2_cm_read_mod_reg(CORE_MOD, OMAP24XX_CM_FCLKEN2);
/* Ignore UART clocks. These are handled by UART core (serial.c) */
f1 &= ~(OMAP24XX_EN_UART1_MASK | OMAP24XX_EN_UART2_MASK);
f2 &= ~OMAP24XX_EN_UART3_MASK;
if (f1 | f2)
return 1;
return 0;
}
static void omap2_enter_full_retention(void)
{
u32 l;
struct timespec ts_preidle, ts_postidle, ts_idle;
/* There is 1 reference hold for all children of the oscillator
* clock, the following will remove it. If no one else uses the
* oscillator itself it will be disabled if/when we enter retention
* mode.
*/
clk_disable(osc_ck);
/* Clear old wake-up events */
/* REVISIT: These write to reserved bits? */
omap2_prm_write_mod_reg(0xffffffff, CORE_MOD, PM_WKST1);
omap2_prm_write_mod_reg(0xffffffff, CORE_MOD, OMAP24XX_PM_WKST2);
omap2_prm_write_mod_reg(0xffffffff, WKUP_MOD, PM_WKST);
/*
* Set MPU powerdomain's next power state to RETENTION;
* preserve logic state during retention
*/
pwrdm_set_logic_retst(mpu_pwrdm, PWRDM_POWER_RET);
pwrdm_set_next_pwrst(mpu_pwrdm, PWRDM_POWER_RET);
/* Workaround to kill USB */
l = omap_ctrl_readl(OMAP2_CONTROL_DEVCONF0) | OMAP24XX_USBSTANDBYCTRL;
omap_ctrl_writel(l, OMAP2_CONTROL_DEVCONF0);
omap2_gpio_prepare_for_idle(0);
if (omap2_pm_debug) {
omap2_pm_dump(0, 0, 0);
getnstimeofday(&ts_preidle);
}
/* One last check for pending IRQs to avoid extra latency due
* to sleeping unnecessarily. */
if (omap_irq_pending())
goto no_sleep;
OMAP2+: PM/serial: hold console semaphore while OMAP UARTs are disabled The console semaphore must be held while the OMAP UART devices are disabled, lest a console write cause an ARM abort (and a kernel crash) when the underlying console device is inaccessible. These crashes only occur when the console is on one of the OMAP internal serial ports. While this problem has been latent in the PM idle loop for some time, the crash was not triggerable with an unmodified kernel until commit 6f251e9db1093c187addc309b5f2f7fe3efd2995 ("OMAP: UART: omap_device conversions, remove implicit 8520 assumptions"). After this patch, a console write often occurs after the console UART has been disabled in the idle loop, crashing the system. Several users have encountered this bug: http://www.mail-archive.com/linux-omap@vger.kernel.org/msg38396.html http://www.mail-archive.com/linux-omap@vger.kernel.org/msg36602.html The same commit also introduced new code that disabled the UARTs during init, in omap_serial_init_port(). The kernel will also crash in this code when earlyconsole and extra debugging is enabled: http://www.mail-archive.com/linux-omap@vger.kernel.org/msg36411.html The minimal fix for the -rc series is to hold the console semaphore while the OMAP UARTs are disabled. This is a somewhat overbroad fix, since the console may not be located on an OMAP UART, as is the case with the GPMC UART on Zoom3. While it is technically possible to determine which devices the console or earlyconsole is actually running on, it is not a trivial problem to solve, and the code to do so is not really appropriate for the -rc series. The right long-term fix is to ensure that no code outside of the OMAP serial driver can disable an OMAP UART. As I understand it, code to implement this is under development by TI. This patch is a collaboration between Paul Walmsley <paul@pwsan.com> and Tony Lindgren <tony@atomide.com>. Thanks to Ming Lei <tom.leiming@gmail.com> and Pramod <pramod.gurav@ti.com> for their feedback on earlier versions of this patch. Signed-off-by: Paul Walmsley <paul@pwsan.com> Signed-off-by: Tony Lindgren <tony@atomide.com> Acked-by: Kevin Hilman <khilman@deeprootsystems.com> Cc: Ming Lei <tom.leiming@gmail.com> Cc: Pramod <pramod.gurav@ti.com> Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com> Cc: Jean Pihet <jean.pihet@newoldbits.com> Cc: Govindraj.R <govindraj.raja@ti.com>
2010-11-25 06:49:05 +07:00
/* Block console output in case it is on one of the OMAP UARTs */
if (!is_suspending())
if (try_acquire_console_sem())
goto no_sleep;
OMAP2+: PM/serial: hold console semaphore while OMAP UARTs are disabled The console semaphore must be held while the OMAP UART devices are disabled, lest a console write cause an ARM abort (and a kernel crash) when the underlying console device is inaccessible. These crashes only occur when the console is on one of the OMAP internal serial ports. While this problem has been latent in the PM idle loop for some time, the crash was not triggerable with an unmodified kernel until commit 6f251e9db1093c187addc309b5f2f7fe3efd2995 ("OMAP: UART: omap_device conversions, remove implicit 8520 assumptions"). After this patch, a console write often occurs after the console UART has been disabled in the idle loop, crashing the system. Several users have encountered this bug: http://www.mail-archive.com/linux-omap@vger.kernel.org/msg38396.html http://www.mail-archive.com/linux-omap@vger.kernel.org/msg36602.html The same commit also introduced new code that disabled the UARTs during init, in omap_serial_init_port(). The kernel will also crash in this code when earlyconsole and extra debugging is enabled: http://www.mail-archive.com/linux-omap@vger.kernel.org/msg36411.html The minimal fix for the -rc series is to hold the console semaphore while the OMAP UARTs are disabled. This is a somewhat overbroad fix, since the console may not be located on an OMAP UART, as is the case with the GPMC UART on Zoom3. While it is technically possible to determine which devices the console or earlyconsole is actually running on, it is not a trivial problem to solve, and the code to do so is not really appropriate for the -rc series. The right long-term fix is to ensure that no code outside of the OMAP serial driver can disable an OMAP UART. As I understand it, code to implement this is under development by TI. This patch is a collaboration between Paul Walmsley <paul@pwsan.com> and Tony Lindgren <tony@atomide.com>. Thanks to Ming Lei <tom.leiming@gmail.com> and Pramod <pramod.gurav@ti.com> for their feedback on earlier versions of this patch. Signed-off-by: Paul Walmsley <paul@pwsan.com> Signed-off-by: Tony Lindgren <tony@atomide.com> Acked-by: Kevin Hilman <khilman@deeprootsystems.com> Cc: Ming Lei <tom.leiming@gmail.com> Cc: Pramod <pramod.gurav@ti.com> Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com> Cc: Jean Pihet <jean.pihet@newoldbits.com> Cc: Govindraj.R <govindraj.raja@ti.com>
2010-11-25 06:49:05 +07:00
omap_uart_prepare_idle(0);
omap_uart_prepare_idle(1);
omap_uart_prepare_idle(2);
/* Jump to SRAM suspend code */
omap2_sram_suspend(sdrc_read_reg(SDRC_DLLA_CTRL),
OMAP_SDRC_REGADDR(SDRC_DLLA_CTRL),
OMAP_SDRC_REGADDR(SDRC_POWER));
omap_uart_resume_idle(2);
omap_uart_resume_idle(1);
omap_uart_resume_idle(0);
if (!is_suspending())
release_console_sem();
OMAP2+: PM/serial: hold console semaphore while OMAP UARTs are disabled The console semaphore must be held while the OMAP UART devices are disabled, lest a console write cause an ARM abort (and a kernel crash) when the underlying console device is inaccessible. These crashes only occur when the console is on one of the OMAP internal serial ports. While this problem has been latent in the PM idle loop for some time, the crash was not triggerable with an unmodified kernel until commit 6f251e9db1093c187addc309b5f2f7fe3efd2995 ("OMAP: UART: omap_device conversions, remove implicit 8520 assumptions"). After this patch, a console write often occurs after the console UART has been disabled in the idle loop, crashing the system. Several users have encountered this bug: http://www.mail-archive.com/linux-omap@vger.kernel.org/msg38396.html http://www.mail-archive.com/linux-omap@vger.kernel.org/msg36602.html The same commit also introduced new code that disabled the UARTs during init, in omap_serial_init_port(). The kernel will also crash in this code when earlyconsole and extra debugging is enabled: http://www.mail-archive.com/linux-omap@vger.kernel.org/msg36411.html The minimal fix for the -rc series is to hold the console semaphore while the OMAP UARTs are disabled. This is a somewhat overbroad fix, since the console may not be located on an OMAP UART, as is the case with the GPMC UART on Zoom3. While it is technically possible to determine which devices the console or earlyconsole is actually running on, it is not a trivial problem to solve, and the code to do so is not really appropriate for the -rc series. The right long-term fix is to ensure that no code outside of the OMAP serial driver can disable an OMAP UART. As I understand it, code to implement this is under development by TI. This patch is a collaboration between Paul Walmsley <paul@pwsan.com> and Tony Lindgren <tony@atomide.com>. Thanks to Ming Lei <tom.leiming@gmail.com> and Pramod <pramod.gurav@ti.com> for their feedback on earlier versions of this patch. Signed-off-by: Paul Walmsley <paul@pwsan.com> Signed-off-by: Tony Lindgren <tony@atomide.com> Acked-by: Kevin Hilman <khilman@deeprootsystems.com> Cc: Ming Lei <tom.leiming@gmail.com> Cc: Pramod <pramod.gurav@ti.com> Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com> Cc: Jean Pihet <jean.pihet@newoldbits.com> Cc: Govindraj.R <govindraj.raja@ti.com>
2010-11-25 06:49:05 +07:00
no_sleep:
if (omap2_pm_debug) {
unsigned long long tmp;
getnstimeofday(&ts_postidle);
ts_idle = timespec_sub(ts_postidle, ts_preidle);
tmp = timespec_to_ns(&ts_idle) * NSEC_PER_USEC;
omap2_pm_dump(0, 1, tmp);
}
omap2_gpio_resume_after_idle();
clk_enable(osc_ck);
/* clear CORE wake-up events */
omap2_prm_write_mod_reg(0xffffffff, CORE_MOD, PM_WKST1);
omap2_prm_write_mod_reg(0xffffffff, CORE_MOD, OMAP24XX_PM_WKST2);
/* wakeup domain events - bit 1: GPT1, bit5 GPIO */
omap2_prm_clear_mod_reg_bits(0x4 | 0x1, WKUP_MOD, PM_WKST);
/* MPU domain wake events */
l = omap2_prm_read_mod_reg(OCP_MOD, OMAP2_PRCM_IRQSTATUS_MPU_OFFSET);
if (l & 0x01)
omap2_prm_write_mod_reg(0x01, OCP_MOD,
OMAP2_PRCM_IRQSTATUS_MPU_OFFSET);
if (l & 0x20)
omap2_prm_write_mod_reg(0x20, OCP_MOD,
OMAP2_PRCM_IRQSTATUS_MPU_OFFSET);
/* Mask future PRCM-to-MPU interrupts */
omap2_prm_write_mod_reg(0x0, OCP_MOD, OMAP2_PRCM_IRQSTATUS_MPU_OFFSET);
}
static int omap2_i2c_active(void)
{
u32 l;
l = omap2_cm_read_mod_reg(CORE_MOD, CM_FCLKEN1);
return l & (OMAP2420_EN_I2C2_MASK | OMAP2420_EN_I2C1_MASK);
}
static int sti_console_enabled;
static int omap2_allow_mpu_retention(void)
{
u32 l;
/* Check for MMC, UART2, UART1, McSPI2, McSPI1 and DSS1. */
l = omap2_cm_read_mod_reg(CORE_MOD, CM_FCLKEN1);
if (l & (OMAP2420_EN_MMC_MASK | OMAP24XX_EN_UART2_MASK |
OMAP24XX_EN_UART1_MASK | OMAP24XX_EN_MCSPI2_MASK |
OMAP24XX_EN_MCSPI1_MASK | OMAP24XX_EN_DSS1_MASK))
return 0;
/* Check for UART3. */
l = omap2_cm_read_mod_reg(CORE_MOD, OMAP24XX_CM_FCLKEN2);
if (l & OMAP24XX_EN_UART3_MASK)
return 0;
if (sti_console_enabled)
return 0;
return 1;
}
static void omap2_enter_mpu_retention(void)
{
int only_idle = 0;
struct timespec ts_preidle, ts_postidle, ts_idle;
/* Putting MPU into the WFI state while a transfer is active
* seems to cause the I2C block to timeout. Why? Good question. */
if (omap2_i2c_active())
return;
/* The peripherals seem not to be able to wake up the MPU when
* it is in retention mode. */
if (omap2_allow_mpu_retention()) {
/* REVISIT: These write to reserved bits? */
omap2_prm_write_mod_reg(0xffffffff, CORE_MOD, PM_WKST1);
omap2_prm_write_mod_reg(0xffffffff, CORE_MOD, OMAP24XX_PM_WKST2);
omap2_prm_write_mod_reg(0xffffffff, WKUP_MOD, PM_WKST);
/* Try to enter MPU retention */
omap2_prm_write_mod_reg((0x01 << OMAP_POWERSTATE_SHIFT) |
OMAP_LOGICRETSTATE_MASK,
MPU_MOD, OMAP2_PM_PWSTCTRL);
} else {
/* Block MPU retention */
omap2_prm_write_mod_reg(OMAP_LOGICRETSTATE_MASK, MPU_MOD,
OMAP2_PM_PWSTCTRL);
only_idle = 1;
}
if (omap2_pm_debug) {
omap2_pm_dump(only_idle ? 2 : 1, 0, 0);
getnstimeofday(&ts_preidle);
}
omap2_sram_idle();
if (omap2_pm_debug) {
unsigned long long tmp;
getnstimeofday(&ts_postidle);
ts_idle = timespec_sub(ts_postidle, ts_preidle);
tmp = timespec_to_ns(&ts_idle) * NSEC_PER_USEC;
omap2_pm_dump(only_idle ? 2 : 1, 1, tmp);
}
}
static int omap2_can_sleep(void)
{
if (omap2_fclks_active())
return 0;
if (!omap_uart_can_sleep())
return 0;
if (osc_ck->usecount > 1)
return 0;
if (omap_dma_running())
return 0;
return 1;
}
static void omap2_pm_idle(void)
{
local_irq_disable();
local_fiq_disable();
if (!omap2_can_sleep()) {
if (omap_irq_pending())
goto out;
omap2_enter_mpu_retention();
goto out;
}
if (omap_irq_pending())
goto out;
omap2_enter_full_retention();
out:
local_fiq_enable();
local_irq_enable();
}
static int omap2_pm_begin(suspend_state_t state)
{
disable_hlt();
suspend_state = state;
return 0;
}
static int omap2_pm_suspend(void)
{
u32 wken_wkup, mir1;
wken_wkup = omap2_prm_read_mod_reg(WKUP_MOD, PM_WKEN);
wken_wkup &= ~OMAP24XX_EN_GPT1_MASK;
omap2_prm_write_mod_reg(wken_wkup, WKUP_MOD, PM_WKEN);
/* Mask GPT1 */
mir1 = omap_readl(0x480fe0a4);
omap_writel(1 << 5, 0x480fe0ac);
omap_uart_prepare_suspend();
omap2_enter_full_retention();
omap_writel(mir1, 0x480fe0a4);
omap2_prm_write_mod_reg(wken_wkup, WKUP_MOD, PM_WKEN);
return 0;
}
static int omap2_pm_enter(suspend_state_t state)
{
int ret = 0;
switch (state) {
case PM_SUSPEND_STANDBY:
case PM_SUSPEND_MEM:
ret = omap2_pm_suspend();
break;
default:
ret = -EINVAL;
}
return ret;
}
static void omap2_pm_end(void)
{
suspend_state = PM_SUSPEND_ON;
enable_hlt();
}
static struct platform_suspend_ops omap_pm_ops = {
.begin = omap2_pm_begin,
.enter = omap2_pm_enter,
.end = omap2_pm_end,
.valid = suspend_valid_only_mem,
};
OMAP clockdomains: add usecounting for wakeup and sleep dependencies Add usecounting for wakeup and sleep dependencies. In the current situation, if several functions add dependencies on the same clockdomains, when the first dependency removal function is called, the dependency will be incorrectly removed from the hardware. Add clkdm_clear_all_wkdeps() and clkdm_clear_all_sleepdeps(), which provide a fast and usecounting-consistent way to clear all hardware clockdomain dependencies, since accesses to these registers can be quite slow. pm{2,3}4xx.c has been updated to use these new functions. The original version of this patch did not touch these files, which previously wrote directly to the wkdep registers, and thus confused the usecounting code. This problem was found by Kevin Hilman <khilman@deeprootsystems.com>. N.B.: This patch introduces one significant functional difference over the previous pm34xx.c code: sleepdeps are now cleared during clockdomain initialization, whereas previously they were left untouched. This has been tested by Kevin and confirmed to work. The original version of this patch also did not take into consideration that some clockdomains do not have sleep or wakeup dependency sources, which caused NULL pointer dereferences. This problem was debugged and fixed by Kevin Hilman <khilman@deeprootsystems.com>. Signed-off-by: Paul Walmsley <paul@pwsan.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Cc: Jouni Högander <jouni.hogander@nokia.com>
2010-01-27 10:13:01 +07:00
/* XXX This function should be shareable between OMAP2xxx and OMAP3 */
static int __init clkdms_setup(struct clockdomain *clkdm, void *unused)
{
OMAP clockdomains: add usecounting for wakeup and sleep dependencies Add usecounting for wakeup and sleep dependencies. In the current situation, if several functions add dependencies on the same clockdomains, when the first dependency removal function is called, the dependency will be incorrectly removed from the hardware. Add clkdm_clear_all_wkdeps() and clkdm_clear_all_sleepdeps(), which provide a fast and usecounting-consistent way to clear all hardware clockdomain dependencies, since accesses to these registers can be quite slow. pm{2,3}4xx.c has been updated to use these new functions. The original version of this patch did not touch these files, which previously wrote directly to the wkdep registers, and thus confused the usecounting code. This problem was found by Kevin Hilman <khilman@deeprootsystems.com>. N.B.: This patch introduces one significant functional difference over the previous pm34xx.c code: sleepdeps are now cleared during clockdomain initialization, whereas previously they were left untouched. This has been tested by Kevin and confirmed to work. The original version of this patch also did not take into consideration that some clockdomains do not have sleep or wakeup dependency sources, which caused NULL pointer dereferences. This problem was debugged and fixed by Kevin Hilman <khilman@deeprootsystems.com>. Signed-off-by: Paul Walmsley <paul@pwsan.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Cc: Jouni Högander <jouni.hogander@nokia.com>
2010-01-27 10:13:01 +07:00
clkdm_clear_all_wkdeps(clkdm);
clkdm_clear_all_sleepdeps(clkdm);
if (clkdm->flags & CLKDM_CAN_ENABLE_AUTO)
omap2_clkdm_allow_idle(clkdm);
else if (clkdm->flags & CLKDM_CAN_FORCE_SLEEP &&
atomic_read(&clkdm->usecount) == 0)
omap2_clkdm_sleep(clkdm);
return 0;
}
static void __init prcm_setup_regs(void)
{
int i, num_mem_banks;
struct powerdomain *pwrdm;
/* Enable autoidle */
omap2_prm_write_mod_reg(OMAP24XX_AUTOIDLE_MASK, OCP_MOD,
OMAP2_PRCM_SYSCONFIG_OFFSET);
/*
* Set CORE powerdomain memory banks to retain their contents
* during RETENTION
*/
num_mem_banks = pwrdm_get_mem_bank_count(core_pwrdm);
for (i = 0; i < num_mem_banks; i++)
pwrdm_set_mem_retst(core_pwrdm, i, PWRDM_POWER_RET);
/* Set CORE powerdomain's next power state to RETENTION */
pwrdm_set_next_pwrst(core_pwrdm, PWRDM_POWER_RET);
/*
* Set MPU powerdomain's next power state to RETENTION;
* preserve logic state during retention
*/
pwrdm_set_logic_retst(mpu_pwrdm, PWRDM_POWER_RET);
pwrdm_set_next_pwrst(mpu_pwrdm, PWRDM_POWER_RET);
/* Force-power down DSP, GFX powerdomains */
pwrdm = clkdm_get_pwrdm(dsp_clkdm);
pwrdm_set_next_pwrst(pwrdm, PWRDM_POWER_OFF);
omap2_clkdm_sleep(dsp_clkdm);
pwrdm = clkdm_get_pwrdm(gfx_clkdm);
pwrdm_set_next_pwrst(pwrdm, PWRDM_POWER_OFF);
omap2_clkdm_sleep(gfx_clkdm);
OMAP clockdomains: add usecounting for wakeup and sleep dependencies Add usecounting for wakeup and sleep dependencies. In the current situation, if several functions add dependencies on the same clockdomains, when the first dependency removal function is called, the dependency will be incorrectly removed from the hardware. Add clkdm_clear_all_wkdeps() and clkdm_clear_all_sleepdeps(), which provide a fast and usecounting-consistent way to clear all hardware clockdomain dependencies, since accesses to these registers can be quite slow. pm{2,3}4xx.c has been updated to use these new functions. The original version of this patch did not touch these files, which previously wrote directly to the wkdep registers, and thus confused the usecounting code. This problem was found by Kevin Hilman <khilman@deeprootsystems.com>. N.B.: This patch introduces one significant functional difference over the previous pm34xx.c code: sleepdeps are now cleared during clockdomain initialization, whereas previously they were left untouched. This has been tested by Kevin and confirmed to work. The original version of this patch also did not take into consideration that some clockdomains do not have sleep or wakeup dependency sources, which caused NULL pointer dereferences. This problem was debugged and fixed by Kevin Hilman <khilman@deeprootsystems.com>. Signed-off-by: Paul Walmsley <paul@pwsan.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Cc: Jouni Högander <jouni.hogander@nokia.com>
2010-01-27 10:13:01 +07:00
/*
* Clear clockdomain wakeup dependencies and enable
* hardware-supervised idle for all clkdms
*/
clkdm_for_each(clkdms_setup, NULL);
clkdm_add_wkdep(mpu_clkdm, wkup_clkdm);
/* Enable clock autoidle for all domains */
omap2_cm_write_mod_reg(OMAP24XX_AUTO_CAM_MASK |
OMAP24XX_AUTO_MAILBOXES_MASK |
OMAP24XX_AUTO_WDT4_MASK |
OMAP2420_AUTO_WDT3_MASK |
OMAP24XX_AUTO_MSPRO_MASK |
OMAP2420_AUTO_MMC_MASK |
OMAP24XX_AUTO_FAC_MASK |
OMAP2420_AUTO_EAC_MASK |
OMAP24XX_AUTO_HDQ_MASK |
OMAP24XX_AUTO_UART2_MASK |
OMAP24XX_AUTO_UART1_MASK |
OMAP24XX_AUTO_I2C2_MASK |
OMAP24XX_AUTO_I2C1_MASK |
OMAP24XX_AUTO_MCSPI2_MASK |
OMAP24XX_AUTO_MCSPI1_MASK |
OMAP24XX_AUTO_MCBSP2_MASK |
OMAP24XX_AUTO_MCBSP1_MASK |
OMAP24XX_AUTO_GPT12_MASK |
OMAP24XX_AUTO_GPT11_MASK |
OMAP24XX_AUTO_GPT10_MASK |
OMAP24XX_AUTO_GPT9_MASK |
OMAP24XX_AUTO_GPT8_MASK |
OMAP24XX_AUTO_GPT7_MASK |
OMAP24XX_AUTO_GPT6_MASK |
OMAP24XX_AUTO_GPT5_MASK |
OMAP24XX_AUTO_GPT4_MASK |
OMAP24XX_AUTO_GPT3_MASK |
OMAP24XX_AUTO_GPT2_MASK |
OMAP2420_AUTO_VLYNQ_MASK |
OMAP24XX_AUTO_DSS_MASK,
CORE_MOD, CM_AUTOIDLE1);
omap2_cm_write_mod_reg(OMAP24XX_AUTO_UART3_MASK |
OMAP24XX_AUTO_SSI_MASK |
OMAP24XX_AUTO_USB_MASK,
CORE_MOD, CM_AUTOIDLE2);
omap2_cm_write_mod_reg(OMAP24XX_AUTO_SDRC_MASK |
OMAP24XX_AUTO_GPMC_MASK |
OMAP24XX_AUTO_SDMA_MASK,
CORE_MOD, CM_AUTOIDLE3);
omap2_cm_write_mod_reg(OMAP24XX_AUTO_PKA_MASK |
OMAP24XX_AUTO_AES_MASK |
OMAP24XX_AUTO_RNG_MASK |
OMAP24XX_AUTO_SHA_MASK |
OMAP24XX_AUTO_DES_MASK,
CORE_MOD, OMAP24XX_CM_AUTOIDLE4);
omap2_cm_write_mod_reg(OMAP2420_AUTO_DSP_IPI_MASK, OMAP24XX_DSP_MOD,
CM_AUTOIDLE);
/* Put DPLL and both APLLs into autoidle mode */
omap2_cm_write_mod_reg((0x03 << OMAP24XX_AUTO_DPLL_SHIFT) |
(0x03 << OMAP24XX_AUTO_96M_SHIFT) |
(0x03 << OMAP24XX_AUTO_54M_SHIFT),
PLL_MOD, CM_AUTOIDLE);
omap2_cm_write_mod_reg(OMAP24XX_AUTO_OMAPCTRL_MASK |
OMAP24XX_AUTO_WDT1_MASK |
OMAP24XX_AUTO_MPU_WDT_MASK |
OMAP24XX_AUTO_GPIOS_MASK |
OMAP24XX_AUTO_32KSYNC_MASK |
OMAP24XX_AUTO_GPT1_MASK,
WKUP_MOD, CM_AUTOIDLE);
/* REVISIT: Configure number of 32 kHz clock cycles for sys_clk
* stabilisation */
omap2_prm_write_mod_reg(15 << OMAP_SETUP_TIME_SHIFT, OMAP24XX_GR_MOD,
OMAP2_PRCM_CLKSSETUP_OFFSET);
/* Configure automatic voltage transition */
omap2_prm_write_mod_reg(2 << OMAP_SETUP_TIME_SHIFT, OMAP24XX_GR_MOD,
OMAP2_PRCM_VOLTSETUP_OFFSET);
omap2_prm_write_mod_reg(OMAP24XX_AUTO_EXTVOLT_MASK |
(0x1 << OMAP24XX_SETOFF_LEVEL_SHIFT) |
OMAP24XX_MEMRETCTRL_MASK |
(0x1 << OMAP24XX_SETRET_LEVEL_SHIFT) |
(0x0 << OMAP24XX_VOLT_LEVEL_SHIFT),
OMAP24XX_GR_MOD, OMAP2_PRCM_VOLTCTRL_OFFSET);
/* Enable wake-up events */
omap2_prm_write_mod_reg(OMAP24XX_EN_GPIOS_MASK | OMAP24XX_EN_GPT1_MASK,
WKUP_MOD, PM_WKEN);
}
static int __init omap2_pm_init(void)
{
u32 l;
if (!cpu_is_omap24xx())
return -ENODEV;
printk(KERN_INFO "Power Management for OMAP2 initializing\n");
l = omap2_prm_read_mod_reg(OCP_MOD, OMAP2_PRCM_REVISION_OFFSET);
printk(KERN_INFO "PRCM revision %d.%d\n", (l >> 4) & 0x0f, l & 0x0f);
OMAP clockdomains: add usecounting for wakeup and sleep dependencies Add usecounting for wakeup and sleep dependencies. In the current situation, if several functions add dependencies on the same clockdomains, when the first dependency removal function is called, the dependency will be incorrectly removed from the hardware. Add clkdm_clear_all_wkdeps() and clkdm_clear_all_sleepdeps(), which provide a fast and usecounting-consistent way to clear all hardware clockdomain dependencies, since accesses to these registers can be quite slow. pm{2,3}4xx.c has been updated to use these new functions. The original version of this patch did not touch these files, which previously wrote directly to the wkdep registers, and thus confused the usecounting code. This problem was found by Kevin Hilman <khilman@deeprootsystems.com>. N.B.: This patch introduces one significant functional difference over the previous pm34xx.c code: sleepdeps are now cleared during clockdomain initialization, whereas previously they were left untouched. This has been tested by Kevin and confirmed to work. The original version of this patch also did not take into consideration that some clockdomains do not have sleep or wakeup dependency sources, which caused NULL pointer dereferences. This problem was debugged and fixed by Kevin Hilman <khilman@deeprootsystems.com>. Signed-off-by: Paul Walmsley <paul@pwsan.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Cc: Jouni Högander <jouni.hogander@nokia.com>
2010-01-27 10:13:01 +07:00
/* Look up important powerdomains */
mpu_pwrdm = pwrdm_lookup("mpu_pwrdm");
if (!mpu_pwrdm)
pr_err("PM: mpu_pwrdm not found\n");
core_pwrdm = pwrdm_lookup("core_pwrdm");
if (!core_pwrdm)
pr_err("PM: core_pwrdm not found\n");
OMAP clockdomains: add usecounting for wakeup and sleep dependencies Add usecounting for wakeup and sleep dependencies. In the current situation, if several functions add dependencies on the same clockdomains, when the first dependency removal function is called, the dependency will be incorrectly removed from the hardware. Add clkdm_clear_all_wkdeps() and clkdm_clear_all_sleepdeps(), which provide a fast and usecounting-consistent way to clear all hardware clockdomain dependencies, since accesses to these registers can be quite slow. pm{2,3}4xx.c has been updated to use these new functions. The original version of this patch did not touch these files, which previously wrote directly to the wkdep registers, and thus confused the usecounting code. This problem was found by Kevin Hilman <khilman@deeprootsystems.com>. N.B.: This patch introduces one significant functional difference over the previous pm34xx.c code: sleepdeps are now cleared during clockdomain initialization, whereas previously they were left untouched. This has been tested by Kevin and confirmed to work. The original version of this patch also did not take into consideration that some clockdomains do not have sleep or wakeup dependency sources, which caused NULL pointer dereferences. This problem was debugged and fixed by Kevin Hilman <khilman@deeprootsystems.com>. Signed-off-by: Paul Walmsley <paul@pwsan.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Cc: Jouni Högander <jouni.hogander@nokia.com>
2010-01-27 10:13:01 +07:00
/* Look up important clockdomains */
mpu_clkdm = clkdm_lookup("mpu_clkdm");
if (!mpu_clkdm)
pr_err("PM: mpu_clkdm not found\n");
wkup_clkdm = clkdm_lookup("wkup_clkdm");
if (!wkup_clkdm)
pr_err("PM: wkup_clkdm not found\n");
dsp_clkdm = clkdm_lookup("dsp_clkdm");
if (!dsp_clkdm)
OMAP clockdomains: add usecounting for wakeup and sleep dependencies Add usecounting for wakeup and sleep dependencies. In the current situation, if several functions add dependencies on the same clockdomains, when the first dependency removal function is called, the dependency will be incorrectly removed from the hardware. Add clkdm_clear_all_wkdeps() and clkdm_clear_all_sleepdeps(), which provide a fast and usecounting-consistent way to clear all hardware clockdomain dependencies, since accesses to these registers can be quite slow. pm{2,3}4xx.c has been updated to use these new functions. The original version of this patch did not touch these files, which previously wrote directly to the wkdep registers, and thus confused the usecounting code. This problem was found by Kevin Hilman <khilman@deeprootsystems.com>. N.B.: This patch introduces one significant functional difference over the previous pm34xx.c code: sleepdeps are now cleared during clockdomain initialization, whereas previously they were left untouched. This has been tested by Kevin and confirmed to work. The original version of this patch also did not take into consideration that some clockdomains do not have sleep or wakeup dependency sources, which caused NULL pointer dereferences. This problem was debugged and fixed by Kevin Hilman <khilman@deeprootsystems.com>. Signed-off-by: Paul Walmsley <paul@pwsan.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Cc: Jouni Högander <jouni.hogander@nokia.com>
2010-01-27 10:13:01 +07:00
pr_err("PM: dsp_clkdm not found\n");
gfx_clkdm = clkdm_lookup("gfx_clkdm");
if (!gfx_clkdm)
pr_err("PM: gfx_clkdm not found\n");
osc_ck = clk_get(NULL, "osc_ck");
if (IS_ERR(osc_ck)) {
printk(KERN_ERR "could not get osc_ck\n");
return -ENODEV;
}
if (cpu_is_omap242x()) {
emul_ck = clk_get(NULL, "emul_ck");
if (IS_ERR(emul_ck)) {
printk(KERN_ERR "could not get emul_ck\n");
clk_put(osc_ck);
return -ENODEV;
}
}
prcm_setup_regs();
/* Hack to prevent MPU retention when STI console is enabled. */
{
const struct omap_sti_console_config *sti;
sti = omap_get_config(OMAP_TAG_STI_CONSOLE,
struct omap_sti_console_config);
if (sti != NULL && sti->enable)
sti_console_enabled = 1;
}
/*
* We copy the assembler sleep/wakeup routines to SRAM.
* These routines need to be in SRAM as that's the only
* memory the MPU can see when it wakes up.
*/
if (cpu_is_omap24xx()) {
omap2_sram_idle = omap_sram_push(omap24xx_idle_loop_suspend,
omap24xx_idle_loop_suspend_sz);
omap2_sram_suspend = omap_sram_push(omap24xx_cpu_suspend,
omap24xx_cpu_suspend_sz);
}
suspend_set_ops(&omap_pm_ops);
pm_idle = omap2_pm_idle;
return 0;
}
late_initcall(omap2_pm_init);