1
0
Fork 0
alistair23-linux/arch/arm/mach-tegra/sleep-tegra30.S

835 lines
21 KiB
ArmAsm
Raw Normal View History

/* SPDX-License-Identifier: GPL-2.0-only */
/*
* Copyright (c) 2012, NVIDIA Corporation. All rights reserved.
*/
#include <linux/linkage.h>
#include <soc/tegra/flowctrl.h>
#include <soc/tegra/fuse.h>
#include <asm/asm-offsets.h>
#include <asm/assembler.h>
ARM: tegra: add LP1 suspend support for Tegra30 The LP1 suspend mode will power off the CPU, clock gated the PLLs and put SDRAM to self-refresh mode. Any interrupt can wake up device from LP1. The sequence when LP1 suspending: * tunning off L1 data cache and the MMU * storing some EMC registers, DPD (deep power down) status, clk source of mselect and SCLK burst policy * putting SDRAM into self-refresh * switching CPU to CLK_M (12MHz OSC) * tunning off PLLM, PLLP, PLLA, PLLC and PLLX * switching SCLK to CLK_S (32KHz OSC) * shutting off the CPU rail The sequence of LP1 resuming: * re-enabling PLLM, PLLP, PLLA, PLLC and PLLX * restoring the clk source of mselect and SCLK burst policy * setting up CCLK burst policy to PLLX * restoring DPD status and some EMC registers * resuming SDRAM to normal mode * jumping to the "tegra_resume" from PMC_SCRATCH41 Due to the SDRAM will be put into self-refresh mode, the low level procedures of LP1 suspending and resuming should be copied to TEGRA_IRAM_CODE_AREA (TEGRA_IRAM_BASE + SZ_4K) when suspending. Before restoring the CPU context when resuming, the SDRAM needs to be switched back to normal mode. And the PLLs need to be re-enabled, SCLK burst policy be restored, CCLK burst policy be set in PLLX. Then jumping to "tegra_resume" that was expected to be stored in PMC_SCRATCH41 to restore CPU context and back to kernel. Based on the work by: Scott Williams <scwilliams@nvidia.com> Signed-off-by: Joseph Lo <josephl@nvidia.com> Signed-off-by: Stephen Warren <swarren@nvidia.com>
2013-08-12 03:40:04 -06:00
#include <asm/cache.h>
#include "irammap.h"
#include "sleep.h"
ARM: tegra: add LP1 suspend support for Tegra30 The LP1 suspend mode will power off the CPU, clock gated the PLLs and put SDRAM to self-refresh mode. Any interrupt can wake up device from LP1. The sequence when LP1 suspending: * tunning off L1 data cache and the MMU * storing some EMC registers, DPD (deep power down) status, clk source of mselect and SCLK burst policy * putting SDRAM into self-refresh * switching CPU to CLK_M (12MHz OSC) * tunning off PLLM, PLLP, PLLA, PLLC and PLLX * switching SCLK to CLK_S (32KHz OSC) * shutting off the CPU rail The sequence of LP1 resuming: * re-enabling PLLM, PLLP, PLLA, PLLC and PLLX * restoring the clk source of mselect and SCLK burst policy * setting up CCLK burst policy to PLLX * restoring DPD status and some EMC registers * resuming SDRAM to normal mode * jumping to the "tegra_resume" from PMC_SCRATCH41 Due to the SDRAM will be put into self-refresh mode, the low level procedures of LP1 suspending and resuming should be copied to TEGRA_IRAM_CODE_AREA (TEGRA_IRAM_BASE + SZ_4K) when suspending. Before restoring the CPU context when resuming, the SDRAM needs to be switched back to normal mode. And the PLLs need to be re-enabled, SCLK burst policy be restored, CCLK burst policy be set in PLLX. Then jumping to "tegra_resume" that was expected to be stored in PMC_SCRATCH41 to restore CPU context and back to kernel. Based on the work by: Scott Williams <scwilliams@nvidia.com> Signed-off-by: Joseph Lo <josephl@nvidia.com> Signed-off-by: Stephen Warren <swarren@nvidia.com>
2013-08-12 03:40:04 -06:00
#define EMC_CFG 0xc
#define EMC_ADR_CFG 0x10
#define EMC_TIMING_CONTROL 0x28
#define EMC_NOP 0xdc
#define EMC_SELF_REF 0xe0
#define EMC_MRW 0xe8
#define EMC_FBIO_CFG5 0x104
#define EMC_AUTO_CAL_CONFIG 0x2a4
#define EMC_AUTO_CAL_INTERVAL 0x2a8
#define EMC_AUTO_CAL_STATUS 0x2ac
#define EMC_REQ_CTRL 0x2b0
#define EMC_CFG_DIG_DLL 0x2bc
#define EMC_EMC_STATUS 0x2b4
#define EMC_ZCAL_INTERVAL 0x2e0
#define EMC_ZQ_CAL 0x2ec
#define EMC_XM2VTTGENPADCTRL 0x310
#define EMC_XM2VTTGENPADCTRL2 0x314
#define PMC_CTRL 0x0
#define PMC_CTRL_SIDE_EFFECT_LP0 (1 << 14) /* enter LP0 when CPU pwr gated */
#define PMC_PLLP_WB0_OVERRIDE 0xf8
#define PMC_IO_DPD_REQ 0x1b8
#define PMC_IO_DPD_STATUS 0x1bc
#define CLK_RESET_CCLK_BURST 0x20
#define CLK_RESET_CCLK_DIVIDER 0x24
#define CLK_RESET_SCLK_BURST 0x28
#define CLK_RESET_SCLK_DIVIDER 0x2c
#define CLK_RESET_PLLC_BASE 0x80
#define CLK_RESET_PLLC_MISC 0x8c
#define CLK_RESET_PLLM_BASE 0x90
#define CLK_RESET_PLLM_MISC 0x9c
#define CLK_RESET_PLLP_BASE 0xa0
#define CLK_RESET_PLLP_MISC 0xac
#define CLK_RESET_PLLA_BASE 0xb0
#define CLK_RESET_PLLA_MISC 0xbc
#define CLK_RESET_PLLX_BASE 0xe0
#define CLK_RESET_PLLX_MISC 0xe4
2013-08-12 03:40:06 -06:00
#define CLK_RESET_PLLX_MISC3 0x518
#define CLK_RESET_PLLX_MISC3_IDDQ 3
#define CLK_RESET_PLLM_MISC_IDDQ 5
#define CLK_RESET_PLLC_MISC_IDDQ 26
ARM: tegra: add LP1 suspend support for Tegra30 The LP1 suspend mode will power off the CPU, clock gated the PLLs and put SDRAM to self-refresh mode. Any interrupt can wake up device from LP1. The sequence when LP1 suspending: * tunning off L1 data cache and the MMU * storing some EMC registers, DPD (deep power down) status, clk source of mselect and SCLK burst policy * putting SDRAM into self-refresh * switching CPU to CLK_M (12MHz OSC) * tunning off PLLM, PLLP, PLLA, PLLC and PLLX * switching SCLK to CLK_S (32KHz OSC) * shutting off the CPU rail The sequence of LP1 resuming: * re-enabling PLLM, PLLP, PLLA, PLLC and PLLX * restoring the clk source of mselect and SCLK burst policy * setting up CCLK burst policy to PLLX * restoring DPD status and some EMC registers * resuming SDRAM to normal mode * jumping to the "tegra_resume" from PMC_SCRATCH41 Due to the SDRAM will be put into self-refresh mode, the low level procedures of LP1 suspending and resuming should be copied to TEGRA_IRAM_CODE_AREA (TEGRA_IRAM_BASE + SZ_4K) when suspending. Before restoring the CPU context when resuming, the SDRAM needs to be switched back to normal mode. And the PLLs need to be re-enabled, SCLK burst policy be restored, CCLK burst policy be set in PLLX. Then jumping to "tegra_resume" that was expected to be stored in PMC_SCRATCH41 to restore CPU context and back to kernel. Based on the work by: Scott Williams <scwilliams@nvidia.com> Signed-off-by: Joseph Lo <josephl@nvidia.com> Signed-off-by: Stephen Warren <swarren@nvidia.com>
2013-08-12 03:40:04 -06:00
#define CLK_RESET_CLK_SOURCE_MSELECT 0x3b4
#define MSELECT_CLKM (0x3 << 30)
#define LOCK_DELAY 50 /* safety delay after lock is detected */
#define TEGRA30_POWER_HOTPLUG_SHUTDOWN (1 << 27) /* Hotplug shutdown */
ARM: tegra: add LP1 suspend support for Tegra30 The LP1 suspend mode will power off the CPU, clock gated the PLLs and put SDRAM to self-refresh mode. Any interrupt can wake up device from LP1. The sequence when LP1 suspending: * tunning off L1 data cache and the MMU * storing some EMC registers, DPD (deep power down) status, clk source of mselect and SCLK burst policy * putting SDRAM into self-refresh * switching CPU to CLK_M (12MHz OSC) * tunning off PLLM, PLLP, PLLA, PLLC and PLLX * switching SCLK to CLK_S (32KHz OSC) * shutting off the CPU rail The sequence of LP1 resuming: * re-enabling PLLM, PLLP, PLLA, PLLC and PLLX * restoring the clk source of mselect and SCLK burst policy * setting up CCLK burst policy to PLLX * restoring DPD status and some EMC registers * resuming SDRAM to normal mode * jumping to the "tegra_resume" from PMC_SCRATCH41 Due to the SDRAM will be put into self-refresh mode, the low level procedures of LP1 suspending and resuming should be copied to TEGRA_IRAM_CODE_AREA (TEGRA_IRAM_BASE + SZ_4K) when suspending. Before restoring the CPU context when resuming, the SDRAM needs to be switched back to normal mode. And the PLLs need to be re-enabled, SCLK burst policy be restored, CCLK burst policy be set in PLLX. Then jumping to "tegra_resume" that was expected to be stored in PMC_SCRATCH41 to restore CPU context and back to kernel. Based on the work by: Scott Williams <scwilliams@nvidia.com> Signed-off-by: Joseph Lo <josephl@nvidia.com> Signed-off-by: Stephen Warren <swarren@nvidia.com>
2013-08-12 03:40:04 -06:00
.macro emc_device_mask, rd, base
ldr \rd, [\base, #EMC_ADR_CFG]
tst \rd, #0x1
moveq \rd, #(0x1 << 8) @ just 1 device
movne \rd, #(0x3 << 8) @ 2 devices
.endm
.macro emc_timing_update, rd, base
mov \rd, #1
str \rd, [\base, #EMC_TIMING_CONTROL]
1001:
ldr \rd, [\base, #EMC_EMC_STATUS]
tst \rd, #(0x1<<23) @ wait EMC_STATUS_TIMING_UPDATE_STALLED is clear
bne 1001b
.endm
.macro pll_enable, rd, r_car_base, pll_base, pll_misc
ldr \rd, [\r_car_base, #\pll_base]
tst \rd, #(1 << 30)
orreq \rd, \rd, #(1 << 30)
streq \rd, [\r_car_base, #\pll_base]
/* Enable lock detector */
.if \pll_misc
ldr \rd, [\r_car_base, #\pll_misc]
bic \rd, \rd, #(1 << 18)
str \rd, [\r_car_base, #\pll_misc]
ldr \rd, [\r_car_base, #\pll_misc]
ldr \rd, [\r_car_base, #\pll_misc]
orr \rd, \rd, #(1 << 18)
str \rd, [\r_car_base, #\pll_misc]
.endif
.endm
.macro pll_locked, rd, r_car_base, pll_base
1:
ldr \rd, [\r_car_base, #\pll_base]
tst \rd, #(1 << 27)
beq 1b
.endm
2013-08-12 03:40:06 -06:00
.macro pll_iddq_exit, rd, car, iddq, iddq_bit
ldr \rd, [\car, #\iddq]
bic \rd, \rd, #(1<<\iddq_bit)
str \rd, [\car, #\iddq]
.endm
.macro pll_iddq_entry, rd, car, iddq, iddq_bit
ldr \rd, [\car, #\iddq]
orr \rd, \rd, #(1<<\iddq_bit)
str \rd, [\car, #\iddq]
.endm
#if defined(CONFIG_HOTPLUG_CPU) || defined(CONFIG_PM_SLEEP)
/*
* tegra30_hotplug_shutdown(void)
*
* Powergates the current CPU.
* Should never return.
*/
ENTRY(tegra30_hotplug_shutdown)
/* Powergate this CPU */
mov r0, #TEGRA30_POWER_HOTPLUG_SHUTDOWN
bl tegra30_cpu_shutdown
ARM: convert all "mov.* pc, reg" to "bx reg" for ARMv6+ ARMv6 and greater introduced a new instruction ("bx") which can be used to return from function calls. Recent CPUs perform better when the "bx lr" instruction is used rather than the "mov pc, lr" instruction, and this sequence is strongly recommended to be used by the ARM architecture manual (section A.4.1.1). We provide a new macro "ret" with all its variants for the condition code which will resolve to the appropriate instruction. Rather than doing this piecemeal, and miss some instances, change all the "mov pc" instances to use the new macro, with the exception of the "movs" instruction and the kprobes code. This allows us to detect the "mov pc, lr" case and fix it up - and also gives us the possibility of deploying this for other registers depending on the CPU selection. Reported-by: Will Deacon <will.deacon@arm.com> Tested-by: Stephen Warren <swarren@nvidia.com> # Tegra Jetson TK1 Tested-by: Robert Jarzmik <robert.jarzmik@free.fr> # mioa701_bootresume.S Tested-by: Andrew Lunn <andrew@lunn.ch> # Kirkwood Tested-by: Shawn Guo <shawn.guo@freescale.com> Tested-by: Tony Lindgren <tony@atomide.com> # OMAPs Tested-by: Gregory CLEMENT <gregory.clement@free-electrons.com> # Armada XP, 375, 385 Acked-by: Sekhar Nori <nsekhar@ti.com> # DaVinci Acked-by: Christoffer Dall <christoffer.dall@linaro.org> # kvm/hyp Acked-by: Haojian Zhuang <haojian.zhuang@gmail.com> # PXA3xx Acked-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com> # Xen Tested-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de> # ARMv7M Tested-by: Simon Horman <horms+renesas@verge.net.au> # Shmobile Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
2014-06-30 09:29:12 -06:00
ret lr @ should never get here
ENDPROC(tegra30_hotplug_shutdown)
/*
* tegra30_cpu_shutdown(unsigned long flags)
*
* Puts the current CPU in wait-for-event mode on the flow controller
* and powergates it -- flags (in R0) indicate the request type.
*
* r10 = SoC ID
* corrupts r0-r4, r10-r12
*/
ENTRY(tegra30_cpu_shutdown)
cpu_id r3
tegra_get_soc_id TEGRA_APB_MISC_VIRT, r10
cmp r10, #TEGRA30
bne _no_cpu0_chk @ It's not Tegra30
cmp r3, #0
ARM: convert all "mov.* pc, reg" to "bx reg" for ARMv6+ ARMv6 and greater introduced a new instruction ("bx") which can be used to return from function calls. Recent CPUs perform better when the "bx lr" instruction is used rather than the "mov pc, lr" instruction, and this sequence is strongly recommended to be used by the ARM architecture manual (section A.4.1.1). We provide a new macro "ret" with all its variants for the condition code which will resolve to the appropriate instruction. Rather than doing this piecemeal, and miss some instances, change all the "mov pc" instances to use the new macro, with the exception of the "movs" instruction and the kprobes code. This allows us to detect the "mov pc, lr" case and fix it up - and also gives us the possibility of deploying this for other registers depending on the CPU selection. Reported-by: Will Deacon <will.deacon@arm.com> Tested-by: Stephen Warren <swarren@nvidia.com> # Tegra Jetson TK1 Tested-by: Robert Jarzmik <robert.jarzmik@free.fr> # mioa701_bootresume.S Tested-by: Andrew Lunn <andrew@lunn.ch> # Kirkwood Tested-by: Shawn Guo <shawn.guo@freescale.com> Tested-by: Tony Lindgren <tony@atomide.com> # OMAPs Tested-by: Gregory CLEMENT <gregory.clement@free-electrons.com> # Armada XP, 375, 385 Acked-by: Sekhar Nori <nsekhar@ti.com> # DaVinci Acked-by: Christoffer Dall <christoffer.dall@linaro.org> # kvm/hyp Acked-by: Haojian Zhuang <haojian.zhuang@gmail.com> # PXA3xx Acked-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com> # Xen Tested-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de> # ARMv7M Tested-by: Simon Horman <horms+renesas@verge.net.au> # Shmobile Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
2014-06-30 09:29:12 -06:00
reteq lr @ Must never be called for CPU 0
_no_cpu0_chk:
ldr r12, =TEGRA_FLOW_CTRL_VIRT
cpu_to_csr_reg r1, r3
add r1, r1, r12 @ virtual CSR address for this CPU
cpu_to_halt_reg r2, r3
add r2, r2, r12 @ virtual HALT_EVENTS address for this CPU
/*
* Clear this CPU's "event" and "interrupt" flags and power gate
* it when halting but not before it is in the "WFE" state.
*/
movw r12, \
FLOW_CTRL_CSR_INTR_FLAG | FLOW_CTRL_CSR_EVENT_FLAG | \
FLOW_CTRL_CSR_ENABLE
cmp r10, #TEGRA30
moveq r4, #(1 << 4) @ wfe bitmap
movne r4, #(1 << 8) @ wfi bitmap
ARM( orr r12, r12, r4, lsl r3 )
THUMB( lsl r4, r4, r3 )
THUMB( orr r12, r12, r4 )
str r12, [r1]
/* Halt this CPU. */
mov r3, #0x400
delay_1:
subs r3, r3, #1 @ delay as a part of wfe war.
bge delay_1;
cpsid a @ disable imprecise aborts.
ldr r3, [r1] @ read CSR
str r3, [r1] @ clear CSR
tst r0, #TEGRA30_POWER_HOTPLUG_SHUTDOWN
beq flow_ctrl_setting_for_lp2
/* flow controller set up for hotplug */
mov r3, #FLOW_CTRL_WAITEVENT @ For hotplug
b flow_ctrl_done
flow_ctrl_setting_for_lp2:
/* flow controller set up for LP2 */
cmp r10, #TEGRA30
moveq r3, #FLOW_CTRL_WAIT_FOR_INTERRUPT @ For LP2
movne r3, #FLOW_CTRL_WAITEVENT
orrne r3, r3, #FLOW_CTRL_HALT_GIC_IRQ
orrne r3, r3, #FLOW_CTRL_HALT_GIC_FIQ
flow_ctrl_done:
cmp r10, #TEGRA30
str r3, [r2]
ldr r0, [r2]
b wfe_war
__cpu_reset_again:
dsb
.align 5
wfeeq @ CPU should be power gated here
wfine
wfe_war:
b __cpu_reset_again
/*
* 38 nop's, which fills rest of wfe cache line and
* 4 more cachelines with nop
*/
.rept 38
nop
.endr
b . @ should never get here
ENDPROC(tegra30_cpu_shutdown)
#endif
#ifdef CONFIG_PM_SLEEP
ARM: tegra: add LP1 suspend support for Tegra30 The LP1 suspend mode will power off the CPU, clock gated the PLLs and put SDRAM to self-refresh mode. Any interrupt can wake up device from LP1. The sequence when LP1 suspending: * tunning off L1 data cache and the MMU * storing some EMC registers, DPD (deep power down) status, clk source of mselect and SCLK burst policy * putting SDRAM into self-refresh * switching CPU to CLK_M (12MHz OSC) * tunning off PLLM, PLLP, PLLA, PLLC and PLLX * switching SCLK to CLK_S (32KHz OSC) * shutting off the CPU rail The sequence of LP1 resuming: * re-enabling PLLM, PLLP, PLLA, PLLC and PLLX * restoring the clk source of mselect and SCLK burst policy * setting up CCLK burst policy to PLLX * restoring DPD status and some EMC registers * resuming SDRAM to normal mode * jumping to the "tegra_resume" from PMC_SCRATCH41 Due to the SDRAM will be put into self-refresh mode, the low level procedures of LP1 suspending and resuming should be copied to TEGRA_IRAM_CODE_AREA (TEGRA_IRAM_BASE + SZ_4K) when suspending. Before restoring the CPU context when resuming, the SDRAM needs to be switched back to normal mode. And the PLLs need to be re-enabled, SCLK burst policy be restored, CCLK burst policy be set in PLLX. Then jumping to "tegra_resume" that was expected to be stored in PMC_SCRATCH41 to restore CPU context and back to kernel. Based on the work by: Scott Williams <scwilliams@nvidia.com> Signed-off-by: Joseph Lo <josephl@nvidia.com> Signed-off-by: Stephen Warren <swarren@nvidia.com>
2013-08-12 03:40:04 -06:00
/*
* tegra30_sleep_core_finish(unsigned long v2p)
*
* Enters suspend in LP0 or LP1 by turning off the MMU and jumping to
* tegra30_tear_down_core in IRAM
*/
ENTRY(tegra30_sleep_core_finish)
ARM: tegra: Ensure entire dcache is flushed on entering LP0/1 Tegra support several low-power (LPx) states, which are: - LP0: CPU + Core voltage off and DRAM in self-refresh - LP1: CPU voltage off and DRAM in self-refresh - LP2: CPU voltage off When entering any of the above states the tegra_disable_clean_inv_dcache() function is called to flush the dcache. The function tegra_disable_clean_inv_dcache() will either flush the entire data cache or up to the Level of Unification Inner Shareable (LoUIS) depending on the value in r0. When tegra_disable_clean_inv_dcache() is called by tegra20_sleep_core_finish() or tegra30_sleep_core_finish(), to enter LP0 and LP1 power state, the r0 register contains a physical memory address which will not be equal to TEGRA_FLUSH_CACHE_ALL (1) and so the data cache will be only flushed to the LoUIS. However, when tegra_disable_clean_inv_dcache() called by tegra_sleep_cpu_finish() to enter to LP2 power state, r0 is set to TEGRA_FLUSH_CACHE_ALL to flush the entire dcache. Please note that tegra20_sleep_core_finish(), tegra30_sleep_core_finish() and tegra_sleep_cpu_finish() are called by the boot CPU once all other CPUs have been disabled and so it seems appropriate to flush the entire cache at this stage. Therefore, ensure that r0 is set to TEGRA_FLUSH_CACHE_ALL when calling tegra_disable_clean_inv_dcache() from tegra20_sleep_core_finish() and tegra30_sleep_core_finish(). Signed-off-by: Jon Hunter <jonathanh@nvidia.com> Reviewed-by: Joseph Lo <josephl@nvidia.com> Signed-off-by: Thierry Reding <treding@nvidia.com>
2015-11-19 07:19:47 -07:00
mov r4, r0
ARM: tegra: add LP1 suspend support for Tegra30 The LP1 suspend mode will power off the CPU, clock gated the PLLs and put SDRAM to self-refresh mode. Any interrupt can wake up device from LP1. The sequence when LP1 suspending: * tunning off L1 data cache and the MMU * storing some EMC registers, DPD (deep power down) status, clk source of mselect and SCLK burst policy * putting SDRAM into self-refresh * switching CPU to CLK_M (12MHz OSC) * tunning off PLLM, PLLP, PLLA, PLLC and PLLX * switching SCLK to CLK_S (32KHz OSC) * shutting off the CPU rail The sequence of LP1 resuming: * re-enabling PLLM, PLLP, PLLA, PLLC and PLLX * restoring the clk source of mselect and SCLK burst policy * setting up CCLK burst policy to PLLX * restoring DPD status and some EMC registers * resuming SDRAM to normal mode * jumping to the "tegra_resume" from PMC_SCRATCH41 Due to the SDRAM will be put into self-refresh mode, the low level procedures of LP1 suspending and resuming should be copied to TEGRA_IRAM_CODE_AREA (TEGRA_IRAM_BASE + SZ_4K) when suspending. Before restoring the CPU context when resuming, the SDRAM needs to be switched back to normal mode. And the PLLs need to be re-enabled, SCLK burst policy be restored, CCLK burst policy be set in PLLX. Then jumping to "tegra_resume" that was expected to be stored in PMC_SCRATCH41 to restore CPU context and back to kernel. Based on the work by: Scott Williams <scwilliams@nvidia.com> Signed-off-by: Joseph Lo <josephl@nvidia.com> Signed-off-by: Stephen Warren <swarren@nvidia.com>
2013-08-12 03:40:04 -06:00
/* Flush, disable the L1 data cache and exit SMP */
ARM: tegra: Ensure entire dcache is flushed on entering LP0/1 Tegra support several low-power (LPx) states, which are: - LP0: CPU + Core voltage off and DRAM in self-refresh - LP1: CPU voltage off and DRAM in self-refresh - LP2: CPU voltage off When entering any of the above states the tegra_disable_clean_inv_dcache() function is called to flush the dcache. The function tegra_disable_clean_inv_dcache() will either flush the entire data cache or up to the Level of Unification Inner Shareable (LoUIS) depending on the value in r0. When tegra_disable_clean_inv_dcache() is called by tegra20_sleep_core_finish() or tegra30_sleep_core_finish(), to enter LP0 and LP1 power state, the r0 register contains a physical memory address which will not be equal to TEGRA_FLUSH_CACHE_ALL (1) and so the data cache will be only flushed to the LoUIS. However, when tegra_disable_clean_inv_dcache() called by tegra_sleep_cpu_finish() to enter to LP2 power state, r0 is set to TEGRA_FLUSH_CACHE_ALL to flush the entire dcache. Please note that tegra20_sleep_core_finish(), tegra30_sleep_core_finish() and tegra_sleep_cpu_finish() are called by the boot CPU once all other CPUs have been disabled and so it seems appropriate to flush the entire cache at this stage. Therefore, ensure that r0 is set to TEGRA_FLUSH_CACHE_ALL when calling tegra_disable_clean_inv_dcache() from tegra20_sleep_core_finish() and tegra30_sleep_core_finish(). Signed-off-by: Jon Hunter <jonathanh@nvidia.com> Reviewed-by: Joseph Lo <josephl@nvidia.com> Signed-off-by: Thierry Reding <treding@nvidia.com>
2015-11-19 07:19:47 -07:00
mov r0, #TEGRA_FLUSH_CACHE_ALL
ARM: tegra: add LP1 suspend support for Tegra30 The LP1 suspend mode will power off the CPU, clock gated the PLLs and put SDRAM to self-refresh mode. Any interrupt can wake up device from LP1. The sequence when LP1 suspending: * tunning off L1 data cache and the MMU * storing some EMC registers, DPD (deep power down) status, clk source of mselect and SCLK burst policy * putting SDRAM into self-refresh * switching CPU to CLK_M (12MHz OSC) * tunning off PLLM, PLLP, PLLA, PLLC and PLLX * switching SCLK to CLK_S (32KHz OSC) * shutting off the CPU rail The sequence of LP1 resuming: * re-enabling PLLM, PLLP, PLLA, PLLC and PLLX * restoring the clk source of mselect and SCLK burst policy * setting up CCLK burst policy to PLLX * restoring DPD status and some EMC registers * resuming SDRAM to normal mode * jumping to the "tegra_resume" from PMC_SCRATCH41 Due to the SDRAM will be put into self-refresh mode, the low level procedures of LP1 suspending and resuming should be copied to TEGRA_IRAM_CODE_AREA (TEGRA_IRAM_BASE + SZ_4K) when suspending. Before restoring the CPU context when resuming, the SDRAM needs to be switched back to normal mode. And the PLLs need to be re-enabled, SCLK burst policy be restored, CCLK burst policy be set in PLLX. Then jumping to "tegra_resume" that was expected to be stored in PMC_SCRATCH41 to restore CPU context and back to kernel. Based on the work by: Scott Williams <scwilliams@nvidia.com> Signed-off-by: Joseph Lo <josephl@nvidia.com> Signed-off-by: Stephen Warren <swarren@nvidia.com>
2013-08-12 03:40:04 -06:00
bl tegra_disable_clean_inv_dcache
ARM: tegra: Ensure entire dcache is flushed on entering LP0/1 Tegra support several low-power (LPx) states, which are: - LP0: CPU + Core voltage off and DRAM in self-refresh - LP1: CPU voltage off and DRAM in self-refresh - LP2: CPU voltage off When entering any of the above states the tegra_disable_clean_inv_dcache() function is called to flush the dcache. The function tegra_disable_clean_inv_dcache() will either flush the entire data cache or up to the Level of Unification Inner Shareable (LoUIS) depending on the value in r0. When tegra_disable_clean_inv_dcache() is called by tegra20_sleep_core_finish() or tegra30_sleep_core_finish(), to enter LP0 and LP1 power state, the r0 register contains a physical memory address which will not be equal to TEGRA_FLUSH_CACHE_ALL (1) and so the data cache will be only flushed to the LoUIS. However, when tegra_disable_clean_inv_dcache() called by tegra_sleep_cpu_finish() to enter to LP2 power state, r0 is set to TEGRA_FLUSH_CACHE_ALL to flush the entire dcache. Please note that tegra20_sleep_core_finish(), tegra30_sleep_core_finish() and tegra_sleep_cpu_finish() are called by the boot CPU once all other CPUs have been disabled and so it seems appropriate to flush the entire cache at this stage. Therefore, ensure that r0 is set to TEGRA_FLUSH_CACHE_ALL when calling tegra_disable_clean_inv_dcache() from tegra20_sleep_core_finish() and tegra30_sleep_core_finish(). Signed-off-by: Jon Hunter <jonathanh@nvidia.com> Reviewed-by: Joseph Lo <josephl@nvidia.com> Signed-off-by: Thierry Reding <treding@nvidia.com>
2015-11-19 07:19:47 -07:00
mov r0, r4
ARM: tegra: add LP1 suspend support for Tegra30 The LP1 suspend mode will power off the CPU, clock gated the PLLs and put SDRAM to self-refresh mode. Any interrupt can wake up device from LP1. The sequence when LP1 suspending: * tunning off L1 data cache and the MMU * storing some EMC registers, DPD (deep power down) status, clk source of mselect and SCLK burst policy * putting SDRAM into self-refresh * switching CPU to CLK_M (12MHz OSC) * tunning off PLLM, PLLP, PLLA, PLLC and PLLX * switching SCLK to CLK_S (32KHz OSC) * shutting off the CPU rail The sequence of LP1 resuming: * re-enabling PLLM, PLLP, PLLA, PLLC and PLLX * restoring the clk source of mselect and SCLK burst policy * setting up CCLK burst policy to PLLX * restoring DPD status and some EMC registers * resuming SDRAM to normal mode * jumping to the "tegra_resume" from PMC_SCRATCH41 Due to the SDRAM will be put into self-refresh mode, the low level procedures of LP1 suspending and resuming should be copied to TEGRA_IRAM_CODE_AREA (TEGRA_IRAM_BASE + SZ_4K) when suspending. Before restoring the CPU context when resuming, the SDRAM needs to be switched back to normal mode. And the PLLs need to be re-enabled, SCLK burst policy be restored, CCLK burst policy be set in PLLX. Then jumping to "tegra_resume" that was expected to be stored in PMC_SCRATCH41 to restore CPU context and back to kernel. Based on the work by: Scott Williams <scwilliams@nvidia.com> Signed-off-by: Joseph Lo <josephl@nvidia.com> Signed-off-by: Stephen Warren <swarren@nvidia.com>
2013-08-12 03:40:04 -06:00
/*
* Preload all the address literals that are needed for the
* CPU power-gating process, to avoid loading from SDRAM which
* are not supported once SDRAM is put into self-refresh.
* LP0 / LP1 use physical address, since the MMU needs to be
* disabled before putting SDRAM into self-refresh to avoid
* memory access due to page table walks.
*/
mov32 r4, TEGRA_PMC_BASE
mov32 r5, TEGRA_CLK_RESET_BASE
mov32 r6, TEGRA_FLOW_CTRL_BASE
mov32 r7, TEGRA_TMRUS_BASE
mov32 r3, tegra_shut_off_mmu
add r3, r3, r0
mov32 r0, tegra30_tear_down_core
mov32 r1, tegra30_iram_start
sub r0, r0, r1
mov32 r1, TEGRA_IRAM_LPx_RESUME_AREA
ARM: tegra: add LP1 suspend support for Tegra30 The LP1 suspend mode will power off the CPU, clock gated the PLLs and put SDRAM to self-refresh mode. Any interrupt can wake up device from LP1. The sequence when LP1 suspending: * tunning off L1 data cache and the MMU * storing some EMC registers, DPD (deep power down) status, clk source of mselect and SCLK burst policy * putting SDRAM into self-refresh * switching CPU to CLK_M (12MHz OSC) * tunning off PLLM, PLLP, PLLA, PLLC and PLLX * switching SCLK to CLK_S (32KHz OSC) * shutting off the CPU rail The sequence of LP1 resuming: * re-enabling PLLM, PLLP, PLLA, PLLC and PLLX * restoring the clk source of mselect and SCLK burst policy * setting up CCLK burst policy to PLLX * restoring DPD status and some EMC registers * resuming SDRAM to normal mode * jumping to the "tegra_resume" from PMC_SCRATCH41 Due to the SDRAM will be put into self-refresh mode, the low level procedures of LP1 suspending and resuming should be copied to TEGRA_IRAM_CODE_AREA (TEGRA_IRAM_BASE + SZ_4K) when suspending. Before restoring the CPU context when resuming, the SDRAM needs to be switched back to normal mode. And the PLLs need to be re-enabled, SCLK burst policy be restored, CCLK burst policy be set in PLLX. Then jumping to "tegra_resume" that was expected to be stored in PMC_SCRATCH41 to restore CPU context and back to kernel. Based on the work by: Scott Williams <scwilliams@nvidia.com> Signed-off-by: Joseph Lo <josephl@nvidia.com> Signed-off-by: Stephen Warren <swarren@nvidia.com>
2013-08-12 03:40:04 -06:00
add r0, r0, r1
ARM: convert all "mov.* pc, reg" to "bx reg" for ARMv6+ ARMv6 and greater introduced a new instruction ("bx") which can be used to return from function calls. Recent CPUs perform better when the "bx lr" instruction is used rather than the "mov pc, lr" instruction, and this sequence is strongly recommended to be used by the ARM architecture manual (section A.4.1.1). We provide a new macro "ret" with all its variants for the condition code which will resolve to the appropriate instruction. Rather than doing this piecemeal, and miss some instances, change all the "mov pc" instances to use the new macro, with the exception of the "movs" instruction and the kprobes code. This allows us to detect the "mov pc, lr" case and fix it up - and also gives us the possibility of deploying this for other registers depending on the CPU selection. Reported-by: Will Deacon <will.deacon@arm.com> Tested-by: Stephen Warren <swarren@nvidia.com> # Tegra Jetson TK1 Tested-by: Robert Jarzmik <robert.jarzmik@free.fr> # mioa701_bootresume.S Tested-by: Andrew Lunn <andrew@lunn.ch> # Kirkwood Tested-by: Shawn Guo <shawn.guo@freescale.com> Tested-by: Tony Lindgren <tony@atomide.com> # OMAPs Tested-by: Gregory CLEMENT <gregory.clement@free-electrons.com> # Armada XP, 375, 385 Acked-by: Sekhar Nori <nsekhar@ti.com> # DaVinci Acked-by: Christoffer Dall <christoffer.dall@linaro.org> # kvm/hyp Acked-by: Haojian Zhuang <haojian.zhuang@gmail.com> # PXA3xx Acked-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com> # Xen Tested-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de> # ARMv7M Tested-by: Simon Horman <horms+renesas@verge.net.au> # Shmobile Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
2014-06-30 09:29:12 -06:00
ret r3
ARM: tegra: add LP1 suspend support for Tegra30 The LP1 suspend mode will power off the CPU, clock gated the PLLs and put SDRAM to self-refresh mode. Any interrupt can wake up device from LP1. The sequence when LP1 suspending: * tunning off L1 data cache and the MMU * storing some EMC registers, DPD (deep power down) status, clk source of mselect and SCLK burst policy * putting SDRAM into self-refresh * switching CPU to CLK_M (12MHz OSC) * tunning off PLLM, PLLP, PLLA, PLLC and PLLX * switching SCLK to CLK_S (32KHz OSC) * shutting off the CPU rail The sequence of LP1 resuming: * re-enabling PLLM, PLLP, PLLA, PLLC and PLLX * restoring the clk source of mselect and SCLK burst policy * setting up CCLK burst policy to PLLX * restoring DPD status and some EMC registers * resuming SDRAM to normal mode * jumping to the "tegra_resume" from PMC_SCRATCH41 Due to the SDRAM will be put into self-refresh mode, the low level procedures of LP1 suspending and resuming should be copied to TEGRA_IRAM_CODE_AREA (TEGRA_IRAM_BASE + SZ_4K) when suspending. Before restoring the CPU context when resuming, the SDRAM needs to be switched back to normal mode. And the PLLs need to be re-enabled, SCLK burst policy be restored, CCLK burst policy be set in PLLX. Then jumping to "tegra_resume" that was expected to be stored in PMC_SCRATCH41 to restore CPU context and back to kernel. Based on the work by: Scott Williams <scwilliams@nvidia.com> Signed-off-by: Joseph Lo <josephl@nvidia.com> Signed-off-by: Stephen Warren <swarren@nvidia.com>
2013-08-12 03:40:04 -06:00
ENDPROC(tegra30_sleep_core_finish)
/*
* tegra30_sleep_cpu_secondary_finish(unsigned long v2p)
*
* Enters LP2 on secondary CPU by exiting coherency and powergating the CPU.
*/
ENTRY(tegra30_sleep_cpu_secondary_finish)
mov r7, lr
/* Flush and disable the L1 data cache */
mov r0, #TEGRA_FLUSH_CACHE_LOUIS
bl tegra_disable_clean_inv_dcache
/* Powergate this CPU. */
mov r0, #0 @ power mode flags (!hotplug)
bl tegra30_cpu_shutdown
mov r0, #1 @ never return here
ARM: convert all "mov.* pc, reg" to "bx reg" for ARMv6+ ARMv6 and greater introduced a new instruction ("bx") which can be used to return from function calls. Recent CPUs perform better when the "bx lr" instruction is used rather than the "mov pc, lr" instruction, and this sequence is strongly recommended to be used by the ARM architecture manual (section A.4.1.1). We provide a new macro "ret" with all its variants for the condition code which will resolve to the appropriate instruction. Rather than doing this piecemeal, and miss some instances, change all the "mov pc" instances to use the new macro, with the exception of the "movs" instruction and the kprobes code. This allows us to detect the "mov pc, lr" case and fix it up - and also gives us the possibility of deploying this for other registers depending on the CPU selection. Reported-by: Will Deacon <will.deacon@arm.com> Tested-by: Stephen Warren <swarren@nvidia.com> # Tegra Jetson TK1 Tested-by: Robert Jarzmik <robert.jarzmik@free.fr> # mioa701_bootresume.S Tested-by: Andrew Lunn <andrew@lunn.ch> # Kirkwood Tested-by: Shawn Guo <shawn.guo@freescale.com> Tested-by: Tony Lindgren <tony@atomide.com> # OMAPs Tested-by: Gregory CLEMENT <gregory.clement@free-electrons.com> # Armada XP, 375, 385 Acked-by: Sekhar Nori <nsekhar@ti.com> # DaVinci Acked-by: Christoffer Dall <christoffer.dall@linaro.org> # kvm/hyp Acked-by: Haojian Zhuang <haojian.zhuang@gmail.com> # PXA3xx Acked-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com> # Xen Tested-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de> # ARMv7M Tested-by: Simon Horman <horms+renesas@verge.net.au> # Shmobile Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
2014-06-30 09:29:12 -06:00
ret r7
ENDPROC(tegra30_sleep_cpu_secondary_finish)
/*
* tegra30_tear_down_cpu
*
* Switches the CPU to enter sleep.
*/
ENTRY(tegra30_tear_down_cpu)
mov32 r6, TEGRA_FLOW_CTRL_BASE
b tegra30_enter_sleep
ENDPROC(tegra30_tear_down_cpu)
ARM: tegra: add LP1 suspend support for Tegra30 The LP1 suspend mode will power off the CPU, clock gated the PLLs and put SDRAM to self-refresh mode. Any interrupt can wake up device from LP1. The sequence when LP1 suspending: * tunning off L1 data cache and the MMU * storing some EMC registers, DPD (deep power down) status, clk source of mselect and SCLK burst policy * putting SDRAM into self-refresh * switching CPU to CLK_M (12MHz OSC) * tunning off PLLM, PLLP, PLLA, PLLC and PLLX * switching SCLK to CLK_S (32KHz OSC) * shutting off the CPU rail The sequence of LP1 resuming: * re-enabling PLLM, PLLP, PLLA, PLLC and PLLX * restoring the clk source of mselect and SCLK burst policy * setting up CCLK burst policy to PLLX * restoring DPD status and some EMC registers * resuming SDRAM to normal mode * jumping to the "tegra_resume" from PMC_SCRATCH41 Due to the SDRAM will be put into self-refresh mode, the low level procedures of LP1 suspending and resuming should be copied to TEGRA_IRAM_CODE_AREA (TEGRA_IRAM_BASE + SZ_4K) when suspending. Before restoring the CPU context when resuming, the SDRAM needs to be switched back to normal mode. And the PLLs need to be re-enabled, SCLK burst policy be restored, CCLK burst policy be set in PLLX. Then jumping to "tegra_resume" that was expected to be stored in PMC_SCRATCH41 to restore CPU context and back to kernel. Based on the work by: Scott Williams <scwilliams@nvidia.com> Signed-off-by: Joseph Lo <josephl@nvidia.com> Signed-off-by: Stephen Warren <swarren@nvidia.com>
2013-08-12 03:40:04 -06:00
/* START OF ROUTINES COPIED TO IRAM */
.align L1_CACHE_SHIFT
.globl tegra30_iram_start
tegra30_iram_start:
/*
* tegra30_lp1_reset
*
* reset vector for LP1 restore; copied into IRAM during suspend.
* Brings the system back up to a safe staring point (SDRAM out of
* self-refresh, PLLC, PLLM and PLLP reenabled, CPU running on PLLX,
* system clock running on the same PLL that it suspended at), and
* jumps to tegra_resume to restore virtual addressing.
* The physical address of tegra_resume expected to be stored in
* PMC_SCRATCH41.
*
* NOTE: THIS *MUST* BE RELOCATED TO TEGRA_IRAM_LPx_RESUME_AREA.
ARM: tegra: add LP1 suspend support for Tegra30 The LP1 suspend mode will power off the CPU, clock gated the PLLs and put SDRAM to self-refresh mode. Any interrupt can wake up device from LP1. The sequence when LP1 suspending: * tunning off L1 data cache and the MMU * storing some EMC registers, DPD (deep power down) status, clk source of mselect and SCLK burst policy * putting SDRAM into self-refresh * switching CPU to CLK_M (12MHz OSC) * tunning off PLLM, PLLP, PLLA, PLLC and PLLX * switching SCLK to CLK_S (32KHz OSC) * shutting off the CPU rail The sequence of LP1 resuming: * re-enabling PLLM, PLLP, PLLA, PLLC and PLLX * restoring the clk source of mselect and SCLK burst policy * setting up CCLK burst policy to PLLX * restoring DPD status and some EMC registers * resuming SDRAM to normal mode * jumping to the "tegra_resume" from PMC_SCRATCH41 Due to the SDRAM will be put into self-refresh mode, the low level procedures of LP1 suspending and resuming should be copied to TEGRA_IRAM_CODE_AREA (TEGRA_IRAM_BASE + SZ_4K) when suspending. Before restoring the CPU context when resuming, the SDRAM needs to be switched back to normal mode. And the PLLs need to be re-enabled, SCLK burst policy be restored, CCLK burst policy be set in PLLX. Then jumping to "tegra_resume" that was expected to be stored in PMC_SCRATCH41 to restore CPU context and back to kernel. Based on the work by: Scott Williams <scwilliams@nvidia.com> Signed-off-by: Joseph Lo <josephl@nvidia.com> Signed-off-by: Stephen Warren <swarren@nvidia.com>
2013-08-12 03:40:04 -06:00
*/
ENTRY(tegra30_lp1_reset)
/*
* The CPU and system bus are running at 32KHz and executing from
* IRAM when this code is executed; immediately switch to CLKM and
* enable PLLP, PLLM, PLLC, PLLA and PLLX.
*/
mov32 r0, TEGRA_CLK_RESET_BASE
mov r1, #(1 << 28)
str r1, [r0, #CLK_RESET_SCLK_BURST]
str r1, [r0, #CLK_RESET_CCLK_BURST]
mov r1, #0
str r1, [r0, #CLK_RESET_CCLK_DIVIDER]
str r1, [r0, #CLK_RESET_SCLK_DIVIDER]
2013-08-12 03:40:06 -06:00
tegra_get_soc_id TEGRA_APB_MISC_BASE, r10
cmp r10, #TEGRA30
beq _no_pll_iddq_exit
pll_iddq_exit r1, r0, CLK_RESET_PLLM_MISC, CLK_RESET_PLLM_MISC_IDDQ
pll_iddq_exit r1, r0, CLK_RESET_PLLC_MISC, CLK_RESET_PLLC_MISC_IDDQ
pll_iddq_exit r1, r0, CLK_RESET_PLLX_MISC3, CLK_RESET_PLLX_MISC3_IDDQ
mov32 r7, TEGRA_TMRUS_BASE
ldr r1, [r7]
add r1, r1, #2
wait_until r1, r7, r3
/* enable PLLM via PMC */
mov32 r2, TEGRA_PMC_BASE
ldr r1, [r2, #PMC_PLLP_WB0_OVERRIDE]
orr r1, r1, #(1 << 12)
str r1, [r2, #PMC_PLLP_WB0_OVERRIDE]
pll_enable r1, r0, CLK_RESET_PLLM_BASE, 0
pll_enable r1, r0, CLK_RESET_PLLC_BASE, 0
pll_enable r1, r0, CLK_RESET_PLLX_BASE, 0
b _pll_m_c_x_done
_no_pll_iddq_exit:
ARM: tegra: add LP1 suspend support for Tegra30 The LP1 suspend mode will power off the CPU, clock gated the PLLs and put SDRAM to self-refresh mode. Any interrupt can wake up device from LP1. The sequence when LP1 suspending: * tunning off L1 data cache and the MMU * storing some EMC registers, DPD (deep power down) status, clk source of mselect and SCLK burst policy * putting SDRAM into self-refresh * switching CPU to CLK_M (12MHz OSC) * tunning off PLLM, PLLP, PLLA, PLLC and PLLX * switching SCLK to CLK_S (32KHz OSC) * shutting off the CPU rail The sequence of LP1 resuming: * re-enabling PLLM, PLLP, PLLA, PLLC and PLLX * restoring the clk source of mselect and SCLK burst policy * setting up CCLK burst policy to PLLX * restoring DPD status and some EMC registers * resuming SDRAM to normal mode * jumping to the "tegra_resume" from PMC_SCRATCH41 Due to the SDRAM will be put into self-refresh mode, the low level procedures of LP1 suspending and resuming should be copied to TEGRA_IRAM_CODE_AREA (TEGRA_IRAM_BASE + SZ_4K) when suspending. Before restoring the CPU context when resuming, the SDRAM needs to be switched back to normal mode. And the PLLs need to be re-enabled, SCLK burst policy be restored, CCLK burst policy be set in PLLX. Then jumping to "tegra_resume" that was expected to be stored in PMC_SCRATCH41 to restore CPU context and back to kernel. Based on the work by: Scott Williams <scwilliams@nvidia.com> Signed-off-by: Joseph Lo <josephl@nvidia.com> Signed-off-by: Stephen Warren <swarren@nvidia.com>
2013-08-12 03:40:04 -06:00
/* enable PLLM via PMC */
mov32 r2, TEGRA_PMC_BASE
ldr r1, [r2, #PMC_PLLP_WB0_OVERRIDE]
orr r1, r1, #(1 << 12)
str r1, [r2, #PMC_PLLP_WB0_OVERRIDE]
pll_enable r1, r0, CLK_RESET_PLLM_BASE, CLK_RESET_PLLM_MISC
pll_enable r1, r0, CLK_RESET_PLLC_BASE, CLK_RESET_PLLC_MISC
pll_enable r1, r0, CLK_RESET_PLLX_BASE, CLK_RESET_PLLX_MISC
2013-08-12 03:40:06 -06:00
_pll_m_c_x_done:
pll_enable r1, r0, CLK_RESET_PLLP_BASE, CLK_RESET_PLLP_MISC
pll_enable r1, r0, CLK_RESET_PLLA_BASE, CLK_RESET_PLLA_MISC
ARM: tegra: add LP1 suspend support for Tegra30 The LP1 suspend mode will power off the CPU, clock gated the PLLs and put SDRAM to self-refresh mode. Any interrupt can wake up device from LP1. The sequence when LP1 suspending: * tunning off L1 data cache and the MMU * storing some EMC registers, DPD (deep power down) status, clk source of mselect and SCLK burst policy * putting SDRAM into self-refresh * switching CPU to CLK_M (12MHz OSC) * tunning off PLLM, PLLP, PLLA, PLLC and PLLX * switching SCLK to CLK_S (32KHz OSC) * shutting off the CPU rail The sequence of LP1 resuming: * re-enabling PLLM, PLLP, PLLA, PLLC and PLLX * restoring the clk source of mselect and SCLK burst policy * setting up CCLK burst policy to PLLX * restoring DPD status and some EMC registers * resuming SDRAM to normal mode * jumping to the "tegra_resume" from PMC_SCRATCH41 Due to the SDRAM will be put into self-refresh mode, the low level procedures of LP1 suspending and resuming should be copied to TEGRA_IRAM_CODE_AREA (TEGRA_IRAM_BASE + SZ_4K) when suspending. Before restoring the CPU context when resuming, the SDRAM needs to be switched back to normal mode. And the PLLs need to be re-enabled, SCLK burst policy be restored, CCLK burst policy be set in PLLX. Then jumping to "tegra_resume" that was expected to be stored in PMC_SCRATCH41 to restore CPU context and back to kernel. Based on the work by: Scott Williams <scwilliams@nvidia.com> Signed-off-by: Joseph Lo <josephl@nvidia.com> Signed-off-by: Stephen Warren <swarren@nvidia.com>
2013-08-12 03:40:04 -06:00
pll_locked r1, r0, CLK_RESET_PLLM_BASE
pll_locked r1, r0, CLK_RESET_PLLP_BASE
pll_locked r1, r0, CLK_RESET_PLLA_BASE
pll_locked r1, r0, CLK_RESET_PLLC_BASE
pll_locked r1, r0, CLK_RESET_PLLX_BASE
ARM: tegra: Enable PLLP bypass during Tegra124 LP1 commit 1a3388d506bf5b45bb283e6a4c4706cfb4897333 upstream. For a little over a year, U-Boot has configured the flow controller to perform automatic RAM re-repair on off->on power transitions of the CPU rail[1]. This is mandatory for correct operation of Tegra124. However, RAM re-repair relies on certain clocks, which the kernel must enable and leave running. PLLP is one of those clocks. This clock is shut down during LP1 in order to save power. Enable bypass (which I believe routes osc_div_clk, essentially the crystal clock, to the PLL output) so that this clock signal toggles even though the PLL is not active. This is required so that LP1 power mode (system suspend) operates correctly. The bypass configuration must then be undone when resuming from LP1, so that all peripheral clocks run at the expected rate. Without this, many peripherals won't work correctly; for example, the UART baud rate would be incorrect. NVIDIA's downstream kernel code only does this if not compiled for Tegra30, so the added code is made conditional upon the chip ID. NVIDIA's downstream code makes this change conditional upon the active CPU cluster. The upstream kernel currently doesn't support cluster switching, so this patch doesn't test the active CPU cluster ID. [1] 3cc7942a4ae5 ARM: tegra: implement RAM repair Reported-by: Jonathan Hunter <jonathanh@nvidia.com> Cc: stable@vger.kernel.org Signed-off-by: Stephen Warren <swarren@nvidia.com> Signed-off-by: Thierry Reding <treding@nvidia.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2019-10-03 14:50:31 -06:00
tegra_get_soc_id TEGRA_APB_MISC_BASE, r1
cmp r1, #TEGRA30
beq 1f
ldr r1, [r0, #CLK_RESET_PLLP_BASE]
bic r1, r1, #(1<<31) @ disable PllP bypass
str r1, [r0, #CLK_RESET_PLLP_BASE]
1:
ARM: tegra: add LP1 suspend support for Tegra30 The LP1 suspend mode will power off the CPU, clock gated the PLLs and put SDRAM to self-refresh mode. Any interrupt can wake up device from LP1. The sequence when LP1 suspending: * tunning off L1 data cache and the MMU * storing some EMC registers, DPD (deep power down) status, clk source of mselect and SCLK burst policy * putting SDRAM into self-refresh * switching CPU to CLK_M (12MHz OSC) * tunning off PLLM, PLLP, PLLA, PLLC and PLLX * switching SCLK to CLK_S (32KHz OSC) * shutting off the CPU rail The sequence of LP1 resuming: * re-enabling PLLM, PLLP, PLLA, PLLC and PLLX * restoring the clk source of mselect and SCLK burst policy * setting up CCLK burst policy to PLLX * restoring DPD status and some EMC registers * resuming SDRAM to normal mode * jumping to the "tegra_resume" from PMC_SCRATCH41 Due to the SDRAM will be put into self-refresh mode, the low level procedures of LP1 suspending and resuming should be copied to TEGRA_IRAM_CODE_AREA (TEGRA_IRAM_BASE + SZ_4K) when suspending. Before restoring the CPU context when resuming, the SDRAM needs to be switched back to normal mode. And the PLLs need to be re-enabled, SCLK burst policy be restored, CCLK burst policy be set in PLLX. Then jumping to "tegra_resume" that was expected to be stored in PMC_SCRATCH41 to restore CPU context and back to kernel. Based on the work by: Scott Williams <scwilliams@nvidia.com> Signed-off-by: Joseph Lo <josephl@nvidia.com> Signed-off-by: Stephen Warren <swarren@nvidia.com>
2013-08-12 03:40:04 -06:00
mov32 r7, TEGRA_TMRUS_BASE
ldr r1, [r7]
add r1, r1, #LOCK_DELAY
wait_until r1, r7, r3
adr r5, tegra_sdram_pad_save
ARM: tegra: add LP1 suspend support for Tegra30 The LP1 suspend mode will power off the CPU, clock gated the PLLs and put SDRAM to self-refresh mode. Any interrupt can wake up device from LP1. The sequence when LP1 suspending: * tunning off L1 data cache and the MMU * storing some EMC registers, DPD (deep power down) status, clk source of mselect and SCLK burst policy * putting SDRAM into self-refresh * switching CPU to CLK_M (12MHz OSC) * tunning off PLLM, PLLP, PLLA, PLLC and PLLX * switching SCLK to CLK_S (32KHz OSC) * shutting off the CPU rail The sequence of LP1 resuming: * re-enabling PLLM, PLLP, PLLA, PLLC and PLLX * restoring the clk source of mselect and SCLK burst policy * setting up CCLK burst policy to PLLX * restoring DPD status and some EMC registers * resuming SDRAM to normal mode * jumping to the "tegra_resume" from PMC_SCRATCH41 Due to the SDRAM will be put into self-refresh mode, the low level procedures of LP1 suspending and resuming should be copied to TEGRA_IRAM_CODE_AREA (TEGRA_IRAM_BASE + SZ_4K) when suspending. Before restoring the CPU context when resuming, the SDRAM needs to be switched back to normal mode. And the PLLs need to be re-enabled, SCLK burst policy be restored, CCLK burst policy be set in PLLX. Then jumping to "tegra_resume" that was expected to be stored in PMC_SCRATCH41 to restore CPU context and back to kernel. Based on the work by: Scott Williams <scwilliams@nvidia.com> Signed-off-by: Joseph Lo <josephl@nvidia.com> Signed-off-by: Stephen Warren <swarren@nvidia.com>
2013-08-12 03:40:04 -06:00
ldr r4, [r5, #0x18] @ restore CLK_SOURCE_MSELECT
str r4, [r0, #CLK_RESET_CLK_SOURCE_MSELECT]
ldr r4, [r5, #0x1C] @ restore SCLK_BURST
str r4, [r0, #CLK_RESET_SCLK_BURST]
2013-08-12 03:40:06 -06:00
cmp r10, #TEGRA30
movweq r4, #:lower16:((1 << 28) | (0x8)) @ burst policy is PLLX
movteq r4, #:upper16:((1 << 28) | (0x8))
movwne r4, #:lower16:((1 << 28) | (0xe))
movtne r4, #:upper16:((1 << 28) | (0xe))
ARM: tegra: add LP1 suspend support for Tegra30 The LP1 suspend mode will power off the CPU, clock gated the PLLs and put SDRAM to self-refresh mode. Any interrupt can wake up device from LP1. The sequence when LP1 suspending: * tunning off L1 data cache and the MMU * storing some EMC registers, DPD (deep power down) status, clk source of mselect and SCLK burst policy * putting SDRAM into self-refresh * switching CPU to CLK_M (12MHz OSC) * tunning off PLLM, PLLP, PLLA, PLLC and PLLX * switching SCLK to CLK_S (32KHz OSC) * shutting off the CPU rail The sequence of LP1 resuming: * re-enabling PLLM, PLLP, PLLA, PLLC and PLLX * restoring the clk source of mselect and SCLK burst policy * setting up CCLK burst policy to PLLX * restoring DPD status and some EMC registers * resuming SDRAM to normal mode * jumping to the "tegra_resume" from PMC_SCRATCH41 Due to the SDRAM will be put into self-refresh mode, the low level procedures of LP1 suspending and resuming should be copied to TEGRA_IRAM_CODE_AREA (TEGRA_IRAM_BASE + SZ_4K) when suspending. Before restoring the CPU context when resuming, the SDRAM needs to be switched back to normal mode. And the PLLs need to be re-enabled, SCLK burst policy be restored, CCLK burst policy be set in PLLX. Then jumping to "tegra_resume" that was expected to be stored in PMC_SCRATCH41 to restore CPU context and back to kernel. Based on the work by: Scott Williams <scwilliams@nvidia.com> Signed-off-by: Joseph Lo <josephl@nvidia.com> Signed-off-by: Stephen Warren <swarren@nvidia.com>
2013-08-12 03:40:04 -06:00
str r4, [r0, #CLK_RESET_CCLK_BURST]
/* Restore pad power state to normal */
ldr r1, [r5, #0x14] @ PMC_IO_DPD_STATUS
mvn r1, r1
bic r1, r1, #(1 << 31)
orr r1, r1, #(1 << 30)
str r1, [r2, #PMC_IO_DPD_REQ] @ DPD_OFF
2013-08-12 03:40:06 -06:00
cmp r10, #TEGRA30
movweq r0, #:lower16:TEGRA_EMC_BASE @ r0 reserved for emc base
movteq r0, #:upper16:TEGRA_EMC_BASE
cmp r10, #TEGRA114
movweq r0, #:lower16:TEGRA_EMC0_BASE
movteq r0, #:upper16:TEGRA_EMC0_BASE
cmp r10, #TEGRA124
movweq r0, #:lower16:TEGRA124_EMC_BASE
movteq r0, #:upper16:TEGRA124_EMC_BASE
ARM: tegra: add LP1 suspend support for Tegra30 The LP1 suspend mode will power off the CPU, clock gated the PLLs and put SDRAM to self-refresh mode. Any interrupt can wake up device from LP1. The sequence when LP1 suspending: * tunning off L1 data cache and the MMU * storing some EMC registers, DPD (deep power down) status, clk source of mselect and SCLK burst policy * putting SDRAM into self-refresh * switching CPU to CLK_M (12MHz OSC) * tunning off PLLM, PLLP, PLLA, PLLC and PLLX * switching SCLK to CLK_S (32KHz OSC) * shutting off the CPU rail The sequence of LP1 resuming: * re-enabling PLLM, PLLP, PLLA, PLLC and PLLX * restoring the clk source of mselect and SCLK burst policy * setting up CCLK burst policy to PLLX * restoring DPD status and some EMC registers * resuming SDRAM to normal mode * jumping to the "tegra_resume" from PMC_SCRATCH41 Due to the SDRAM will be put into self-refresh mode, the low level procedures of LP1 suspending and resuming should be copied to TEGRA_IRAM_CODE_AREA (TEGRA_IRAM_BASE + SZ_4K) when suspending. Before restoring the CPU context when resuming, the SDRAM needs to be switched back to normal mode. And the PLLs need to be re-enabled, SCLK burst policy be restored, CCLK burst policy be set in PLLX. Then jumping to "tegra_resume" that was expected to be stored in PMC_SCRATCH41 to restore CPU context and back to kernel. Based on the work by: Scott Williams <scwilliams@nvidia.com> Signed-off-by: Joseph Lo <josephl@nvidia.com> Signed-off-by: Stephen Warren <swarren@nvidia.com>
2013-08-12 03:40:04 -06:00
2013-08-12 03:40:06 -06:00
exit_self_refresh:
ARM: tegra: add LP1 suspend support for Tegra30 The LP1 suspend mode will power off the CPU, clock gated the PLLs and put SDRAM to self-refresh mode. Any interrupt can wake up device from LP1. The sequence when LP1 suspending: * tunning off L1 data cache and the MMU * storing some EMC registers, DPD (deep power down) status, clk source of mselect and SCLK burst policy * putting SDRAM into self-refresh * switching CPU to CLK_M (12MHz OSC) * tunning off PLLM, PLLP, PLLA, PLLC and PLLX * switching SCLK to CLK_S (32KHz OSC) * shutting off the CPU rail The sequence of LP1 resuming: * re-enabling PLLM, PLLP, PLLA, PLLC and PLLX * restoring the clk source of mselect and SCLK burst policy * setting up CCLK burst policy to PLLX * restoring DPD status and some EMC registers * resuming SDRAM to normal mode * jumping to the "tegra_resume" from PMC_SCRATCH41 Due to the SDRAM will be put into self-refresh mode, the low level procedures of LP1 suspending and resuming should be copied to TEGRA_IRAM_CODE_AREA (TEGRA_IRAM_BASE + SZ_4K) when suspending. Before restoring the CPU context when resuming, the SDRAM needs to be switched back to normal mode. And the PLLs need to be re-enabled, SCLK burst policy be restored, CCLK burst policy be set in PLLX. Then jumping to "tegra_resume" that was expected to be stored in PMC_SCRATCH41 to restore CPU context and back to kernel. Based on the work by: Scott Williams <scwilliams@nvidia.com> Signed-off-by: Joseph Lo <josephl@nvidia.com> Signed-off-by: Stephen Warren <swarren@nvidia.com>
2013-08-12 03:40:04 -06:00
ldr r1, [r5, #0xC] @ restore EMC_XM2VTTGENPADCTRL
str r1, [r0, #EMC_XM2VTTGENPADCTRL]
ldr r1, [r5, #0x10] @ restore EMC_XM2VTTGENPADCTRL2
str r1, [r0, #EMC_XM2VTTGENPADCTRL2]
ldr r1, [r5, #0x8] @ restore EMC_AUTO_CAL_INTERVAL
str r1, [r0, #EMC_AUTO_CAL_INTERVAL]
/* Relock DLL */
ldr r1, [r0, #EMC_CFG_DIG_DLL]
orr r1, r1, #(1 << 30) @ set DLL_RESET
str r1, [r0, #EMC_CFG_DIG_DLL]
emc_timing_update r1, r0
2013-08-12 03:40:06 -06:00
cmp r10, #TEGRA114
movweq r1, #:lower16:TEGRA_EMC1_BASE
movteq r1, #:upper16:TEGRA_EMC1_BASE
cmpeq r0, r1
ARM: tegra: add LP1 suspend support for Tegra30 The LP1 suspend mode will power off the CPU, clock gated the PLLs and put SDRAM to self-refresh mode. Any interrupt can wake up device from LP1. The sequence when LP1 suspending: * tunning off L1 data cache and the MMU * storing some EMC registers, DPD (deep power down) status, clk source of mselect and SCLK burst policy * putting SDRAM into self-refresh * switching CPU to CLK_M (12MHz OSC) * tunning off PLLM, PLLP, PLLA, PLLC and PLLX * switching SCLK to CLK_S (32KHz OSC) * shutting off the CPU rail The sequence of LP1 resuming: * re-enabling PLLM, PLLP, PLLA, PLLC and PLLX * restoring the clk source of mselect and SCLK burst policy * setting up CCLK burst policy to PLLX * restoring DPD status and some EMC registers * resuming SDRAM to normal mode * jumping to the "tegra_resume" from PMC_SCRATCH41 Due to the SDRAM will be put into self-refresh mode, the low level procedures of LP1 suspending and resuming should be copied to TEGRA_IRAM_CODE_AREA (TEGRA_IRAM_BASE + SZ_4K) when suspending. Before restoring the CPU context when resuming, the SDRAM needs to be switched back to normal mode. And the PLLs need to be re-enabled, SCLK burst policy be restored, CCLK burst policy be set in PLLX. Then jumping to "tegra_resume" that was expected to be stored in PMC_SCRATCH41 to restore CPU context and back to kernel. Based on the work by: Scott Williams <scwilliams@nvidia.com> Signed-off-by: Joseph Lo <josephl@nvidia.com> Signed-off-by: Stephen Warren <swarren@nvidia.com>
2013-08-12 03:40:04 -06:00
ldr r1, [r0, #EMC_AUTO_CAL_CONFIG]
orr r1, r1, #(1 << 31) @ set AUTO_CAL_ACTIVE
2013-08-12 03:40:06 -06:00
orreq r1, r1, #(1 << 27) @ set slave mode for channel 1
ARM: tegra: add LP1 suspend support for Tegra30 The LP1 suspend mode will power off the CPU, clock gated the PLLs and put SDRAM to self-refresh mode. Any interrupt can wake up device from LP1. The sequence when LP1 suspending: * tunning off L1 data cache and the MMU * storing some EMC registers, DPD (deep power down) status, clk source of mselect and SCLK burst policy * putting SDRAM into self-refresh * switching CPU to CLK_M (12MHz OSC) * tunning off PLLM, PLLP, PLLA, PLLC and PLLX * switching SCLK to CLK_S (32KHz OSC) * shutting off the CPU rail The sequence of LP1 resuming: * re-enabling PLLM, PLLP, PLLA, PLLC and PLLX * restoring the clk source of mselect and SCLK burst policy * setting up CCLK burst policy to PLLX * restoring DPD status and some EMC registers * resuming SDRAM to normal mode * jumping to the "tegra_resume" from PMC_SCRATCH41 Due to the SDRAM will be put into self-refresh mode, the low level procedures of LP1 suspending and resuming should be copied to TEGRA_IRAM_CODE_AREA (TEGRA_IRAM_BASE + SZ_4K) when suspending. Before restoring the CPU context when resuming, the SDRAM needs to be switched back to normal mode. And the PLLs need to be re-enabled, SCLK burst policy be restored, CCLK burst policy be set in PLLX. Then jumping to "tegra_resume" that was expected to be stored in PMC_SCRATCH41 to restore CPU context and back to kernel. Based on the work by: Scott Williams <scwilliams@nvidia.com> Signed-off-by: Joseph Lo <josephl@nvidia.com> Signed-off-by: Stephen Warren <swarren@nvidia.com>
2013-08-12 03:40:04 -06:00
str r1, [r0, #EMC_AUTO_CAL_CONFIG]
emc_wait_auto_cal_onetime:
ldr r1, [r0, #EMC_AUTO_CAL_STATUS]
tst r1, #(1 << 31) @ wait until AUTO_CAL_ACTIVE is cleared
bne emc_wait_auto_cal_onetime
ldr r1, [r0, #EMC_CFG]
bic r1, r1, #(1 << 31) @ disable DRAM_CLK_STOP_PD
str r1, [r0, #EMC_CFG]
mov r1, #0
str r1, [r0, #EMC_SELF_REF] @ take DRAM out of self refresh
mov r1, #1
2013-08-12 03:40:06 -06:00
cmp r10, #TEGRA30
streq r1, [r0, #EMC_NOP]
streq r1, [r0, #EMC_NOP]
ARM: tegra: add LP1 suspend support for Tegra30 The LP1 suspend mode will power off the CPU, clock gated the PLLs and put SDRAM to self-refresh mode. Any interrupt can wake up device from LP1. The sequence when LP1 suspending: * tunning off L1 data cache and the MMU * storing some EMC registers, DPD (deep power down) status, clk source of mselect and SCLK burst policy * putting SDRAM into self-refresh * switching CPU to CLK_M (12MHz OSC) * tunning off PLLM, PLLP, PLLA, PLLC and PLLX * switching SCLK to CLK_S (32KHz OSC) * shutting off the CPU rail The sequence of LP1 resuming: * re-enabling PLLM, PLLP, PLLA, PLLC and PLLX * restoring the clk source of mselect and SCLK burst policy * setting up CCLK burst policy to PLLX * restoring DPD status and some EMC registers * resuming SDRAM to normal mode * jumping to the "tegra_resume" from PMC_SCRATCH41 Due to the SDRAM will be put into self-refresh mode, the low level procedures of LP1 suspending and resuming should be copied to TEGRA_IRAM_CODE_AREA (TEGRA_IRAM_BASE + SZ_4K) when suspending. Before restoring the CPU context when resuming, the SDRAM needs to be switched back to normal mode. And the PLLs need to be re-enabled, SCLK burst policy be restored, CCLK burst policy be set in PLLX. Then jumping to "tegra_resume" that was expected to be stored in PMC_SCRATCH41 to restore CPU context and back to kernel. Based on the work by: Scott Williams <scwilliams@nvidia.com> Signed-off-by: Joseph Lo <josephl@nvidia.com> Signed-off-by: Stephen Warren <swarren@nvidia.com>
2013-08-12 03:40:04 -06:00
emc_device_mask r1, r0
exit_selfrefresh_loop:
ldr r2, [r0, #EMC_EMC_STATUS]
ands r2, r2, r1
bne exit_selfrefresh_loop
lsr r1, r1, #8 @ devSel, bit0:dev0, bit1:dev1
mov32 r7, TEGRA_TMRUS_BASE
ldr r2, [r0, #EMC_FBIO_CFG5]
and r2, r2, #3 @ check DRAM_TYPE
cmp r2, #2
beq emc_lpddr2
/* Issue a ZQ_CAL for dev0 - DDR3 */
mov32 r2, 0x80000011 @ DEV_SELECTION=2, LENGTH=LONG, CMD=1
str r2, [r0, #EMC_ZQ_CAL]
ldr r2, [r7]
add r2, r2, #10
wait_until r2, r7, r3
tst r1, #2
beq zcal_done
/* Issue a ZQ_CAL for dev1 - DDR3 */
mov32 r2, 0x40000011 @ DEV_SELECTION=1, LENGTH=LONG, CMD=1
str r2, [r0, #EMC_ZQ_CAL]
ldr r2, [r7]
add r2, r2, #10
wait_until r2, r7, r3
b zcal_done
emc_lpddr2:
/* Issue a ZQ_CAL for dev0 - LPDDR2 */
mov32 r2, 0x800A00AB @ DEV_SELECTION=2, MA=10, OP=0xAB
str r2, [r0, #EMC_MRW]
ldr r2, [r7]
add r2, r2, #1
wait_until r2, r7, r3
tst r1, #2
beq zcal_done
/* Issue a ZQ_CAL for dev0 - LPDDR2 */
mov32 r2, 0x400A00AB @ DEV_SELECTION=1, MA=10, OP=0xAB
str r2, [r0, #EMC_MRW]
ldr r2, [r7]
add r2, r2, #1
wait_until r2, r7, r3
zcal_done:
mov r1, #0 @ unstall all transactions
str r1, [r0, #EMC_REQ_CTRL]
ldr r1, [r5, #0x4] @ restore EMC_ZCAL_INTERVAL
str r1, [r0, #EMC_ZCAL_INTERVAL]
ldr r1, [r5, #0x0] @ restore EMC_CFG
str r1, [r0, #EMC_CFG]
emc_timing_update r1, r0
2013-08-12 03:40:06 -06:00
/* Tegra114 had dual EMC channel, now config the other one */
cmp r10, #TEGRA114
bne __no_dual_emc_chanl
mov32 r1, TEGRA_EMC1_BASE
cmp r0, r1
movne r0, r1
addne r5, r5, #0x20
bne exit_self_refresh
__no_dual_emc_chanl:
ARM: tegra: add LP1 suspend support for Tegra30 The LP1 suspend mode will power off the CPU, clock gated the PLLs and put SDRAM to self-refresh mode. Any interrupt can wake up device from LP1. The sequence when LP1 suspending: * tunning off L1 data cache and the MMU * storing some EMC registers, DPD (deep power down) status, clk source of mselect and SCLK burst policy * putting SDRAM into self-refresh * switching CPU to CLK_M (12MHz OSC) * tunning off PLLM, PLLP, PLLA, PLLC and PLLX * switching SCLK to CLK_S (32KHz OSC) * shutting off the CPU rail The sequence of LP1 resuming: * re-enabling PLLM, PLLP, PLLA, PLLC and PLLX * restoring the clk source of mselect and SCLK burst policy * setting up CCLK burst policy to PLLX * restoring DPD status and some EMC registers * resuming SDRAM to normal mode * jumping to the "tegra_resume" from PMC_SCRATCH41 Due to the SDRAM will be put into self-refresh mode, the low level procedures of LP1 suspending and resuming should be copied to TEGRA_IRAM_CODE_AREA (TEGRA_IRAM_BASE + SZ_4K) when suspending. Before restoring the CPU context when resuming, the SDRAM needs to be switched back to normal mode. And the PLLs need to be re-enabled, SCLK burst policy be restored, CCLK burst policy be set in PLLX. Then jumping to "tegra_resume" that was expected to be stored in PMC_SCRATCH41 to restore CPU context and back to kernel. Based on the work by: Scott Williams <scwilliams@nvidia.com> Signed-off-by: Joseph Lo <josephl@nvidia.com> Signed-off-by: Stephen Warren <swarren@nvidia.com>
2013-08-12 03:40:04 -06:00
mov32 r0, TEGRA_PMC_BASE
ldr r0, [r0, #PMC_SCRATCH41]
ARM: convert all "mov.* pc, reg" to "bx reg" for ARMv6+ ARMv6 and greater introduced a new instruction ("bx") which can be used to return from function calls. Recent CPUs perform better when the "bx lr" instruction is used rather than the "mov pc, lr" instruction, and this sequence is strongly recommended to be used by the ARM architecture manual (section A.4.1.1). We provide a new macro "ret" with all its variants for the condition code which will resolve to the appropriate instruction. Rather than doing this piecemeal, and miss some instances, change all the "mov pc" instances to use the new macro, with the exception of the "movs" instruction and the kprobes code. This allows us to detect the "mov pc, lr" case and fix it up - and also gives us the possibility of deploying this for other registers depending on the CPU selection. Reported-by: Will Deacon <will.deacon@arm.com> Tested-by: Stephen Warren <swarren@nvidia.com> # Tegra Jetson TK1 Tested-by: Robert Jarzmik <robert.jarzmik@free.fr> # mioa701_bootresume.S Tested-by: Andrew Lunn <andrew@lunn.ch> # Kirkwood Tested-by: Shawn Guo <shawn.guo@freescale.com> Tested-by: Tony Lindgren <tony@atomide.com> # OMAPs Tested-by: Gregory CLEMENT <gregory.clement@free-electrons.com> # Armada XP, 375, 385 Acked-by: Sekhar Nori <nsekhar@ti.com> # DaVinci Acked-by: Christoffer Dall <christoffer.dall@linaro.org> # kvm/hyp Acked-by: Haojian Zhuang <haojian.zhuang@gmail.com> # PXA3xx Acked-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com> # Xen Tested-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de> # ARMv7M Tested-by: Simon Horman <horms+renesas@verge.net.au> # Shmobile Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
2014-06-30 09:29:12 -06:00
ret r0 @ jump to tegra_resume
ARM: tegra: add LP1 suspend support for Tegra30 The LP1 suspend mode will power off the CPU, clock gated the PLLs and put SDRAM to self-refresh mode. Any interrupt can wake up device from LP1. The sequence when LP1 suspending: * tunning off L1 data cache and the MMU * storing some EMC registers, DPD (deep power down) status, clk source of mselect and SCLK burst policy * putting SDRAM into self-refresh * switching CPU to CLK_M (12MHz OSC) * tunning off PLLM, PLLP, PLLA, PLLC and PLLX * switching SCLK to CLK_S (32KHz OSC) * shutting off the CPU rail The sequence of LP1 resuming: * re-enabling PLLM, PLLP, PLLA, PLLC and PLLX * restoring the clk source of mselect and SCLK burst policy * setting up CCLK burst policy to PLLX * restoring DPD status and some EMC registers * resuming SDRAM to normal mode * jumping to the "tegra_resume" from PMC_SCRATCH41 Due to the SDRAM will be put into self-refresh mode, the low level procedures of LP1 suspending and resuming should be copied to TEGRA_IRAM_CODE_AREA (TEGRA_IRAM_BASE + SZ_4K) when suspending. Before restoring the CPU context when resuming, the SDRAM needs to be switched back to normal mode. And the PLLs need to be re-enabled, SCLK burst policy be restored, CCLK burst policy be set in PLLX. Then jumping to "tegra_resume" that was expected to be stored in PMC_SCRATCH41 to restore CPU context and back to kernel. Based on the work by: Scott Williams <scwilliams@nvidia.com> Signed-off-by: Joseph Lo <josephl@nvidia.com> Signed-off-by: Stephen Warren <swarren@nvidia.com>
2013-08-12 03:40:04 -06:00
ENDPROC(tegra30_lp1_reset)
.align L1_CACHE_SHIFT
tegra30_sdram_pad_address:
.word TEGRA_EMC_BASE + EMC_CFG @0x0
.word TEGRA_EMC_BASE + EMC_ZCAL_INTERVAL @0x4
.word TEGRA_EMC_BASE + EMC_AUTO_CAL_INTERVAL @0x8
.word TEGRA_EMC_BASE + EMC_XM2VTTGENPADCTRL @0xc
.word TEGRA_EMC_BASE + EMC_XM2VTTGENPADCTRL2 @0x10
.word TEGRA_PMC_BASE + PMC_IO_DPD_STATUS @0x14
.word TEGRA_CLK_RESET_BASE + CLK_RESET_CLK_SOURCE_MSELECT @0x18
.word TEGRA_CLK_RESET_BASE + CLK_RESET_SCLK_BURST @0x1c
tegra30_sdram_pad_address_end:
ARM: tegra: add LP1 suspend support for Tegra30 The LP1 suspend mode will power off the CPU, clock gated the PLLs and put SDRAM to self-refresh mode. Any interrupt can wake up device from LP1. The sequence when LP1 suspending: * tunning off L1 data cache and the MMU * storing some EMC registers, DPD (deep power down) status, clk source of mselect and SCLK burst policy * putting SDRAM into self-refresh * switching CPU to CLK_M (12MHz OSC) * tunning off PLLM, PLLP, PLLA, PLLC and PLLX * switching SCLK to CLK_S (32KHz OSC) * shutting off the CPU rail The sequence of LP1 resuming: * re-enabling PLLM, PLLP, PLLA, PLLC and PLLX * restoring the clk source of mselect and SCLK burst policy * setting up CCLK burst policy to PLLX * restoring DPD status and some EMC registers * resuming SDRAM to normal mode * jumping to the "tegra_resume" from PMC_SCRATCH41 Due to the SDRAM will be put into self-refresh mode, the low level procedures of LP1 suspending and resuming should be copied to TEGRA_IRAM_CODE_AREA (TEGRA_IRAM_BASE + SZ_4K) when suspending. Before restoring the CPU context when resuming, the SDRAM needs to be switched back to normal mode. And the PLLs need to be re-enabled, SCLK burst policy be restored, CCLK burst policy be set in PLLX. Then jumping to "tegra_resume" that was expected to be stored in PMC_SCRATCH41 to restore CPU context and back to kernel. Based on the work by: Scott Williams <scwilliams@nvidia.com> Signed-off-by: Joseph Lo <josephl@nvidia.com> Signed-off-by: Stephen Warren <swarren@nvidia.com>
2013-08-12 03:40:04 -06:00
2013-08-12 03:40:06 -06:00
tegra114_sdram_pad_address:
.word TEGRA_EMC0_BASE + EMC_CFG @0x0
.word TEGRA_EMC0_BASE + EMC_ZCAL_INTERVAL @0x4
.word TEGRA_EMC0_BASE + EMC_AUTO_CAL_INTERVAL @0x8
.word TEGRA_EMC0_BASE + EMC_XM2VTTGENPADCTRL @0xc
.word TEGRA_EMC0_BASE + EMC_XM2VTTGENPADCTRL2 @0x10
.word TEGRA_PMC_BASE + PMC_IO_DPD_STATUS @0x14
.word TEGRA_CLK_RESET_BASE + CLK_RESET_CLK_SOURCE_MSELECT @0x18
.word TEGRA_CLK_RESET_BASE + CLK_RESET_SCLK_BURST @0x1c
.word TEGRA_EMC1_BASE + EMC_CFG @0x20
.word TEGRA_EMC1_BASE + EMC_ZCAL_INTERVAL @0x24
.word TEGRA_EMC1_BASE + EMC_AUTO_CAL_INTERVAL @0x28
.word TEGRA_EMC1_BASE + EMC_XM2VTTGENPADCTRL @0x2c
.word TEGRA_EMC1_BASE + EMC_XM2VTTGENPADCTRL2 @0x30
tegra114_sdram_pad_adress_end:
2013-08-12 03:40:06 -06:00
tegra124_sdram_pad_address:
.word TEGRA124_EMC_BASE + EMC_CFG @0x0
.word TEGRA124_EMC_BASE + EMC_ZCAL_INTERVAL @0x4
.word TEGRA124_EMC_BASE + EMC_AUTO_CAL_INTERVAL @0x8
.word TEGRA124_EMC_BASE + EMC_XM2VTTGENPADCTRL @0xc
.word TEGRA124_EMC_BASE + EMC_XM2VTTGENPADCTRL2 @0x10
.word TEGRA_PMC_BASE + PMC_IO_DPD_STATUS @0x14
.word TEGRA_CLK_RESET_BASE + CLK_RESET_CLK_SOURCE_MSELECT @0x18
.word TEGRA_CLK_RESET_BASE + CLK_RESET_SCLK_BURST @0x1c
tegra124_sdram_pad_address_end:
ARM: tegra: add LP1 suspend support for Tegra30 The LP1 suspend mode will power off the CPU, clock gated the PLLs and put SDRAM to self-refresh mode. Any interrupt can wake up device from LP1. The sequence when LP1 suspending: * tunning off L1 data cache and the MMU * storing some EMC registers, DPD (deep power down) status, clk source of mselect and SCLK burst policy * putting SDRAM into self-refresh * switching CPU to CLK_M (12MHz OSC) * tunning off PLLM, PLLP, PLLA, PLLC and PLLX * switching SCLK to CLK_S (32KHz OSC) * shutting off the CPU rail The sequence of LP1 resuming: * re-enabling PLLM, PLLP, PLLA, PLLC and PLLX * restoring the clk source of mselect and SCLK burst policy * setting up CCLK burst policy to PLLX * restoring DPD status and some EMC registers * resuming SDRAM to normal mode * jumping to the "tegra_resume" from PMC_SCRATCH41 Due to the SDRAM will be put into self-refresh mode, the low level procedures of LP1 suspending and resuming should be copied to TEGRA_IRAM_CODE_AREA (TEGRA_IRAM_BASE + SZ_4K) when suspending. Before restoring the CPU context when resuming, the SDRAM needs to be switched back to normal mode. And the PLLs need to be re-enabled, SCLK burst policy be restored, CCLK burst policy be set in PLLX. Then jumping to "tegra_resume" that was expected to be stored in PMC_SCRATCH41 to restore CPU context and back to kernel. Based on the work by: Scott Williams <scwilliams@nvidia.com> Signed-off-by: Joseph Lo <josephl@nvidia.com> Signed-off-by: Stephen Warren <swarren@nvidia.com>
2013-08-12 03:40:04 -06:00
tegra30_sdram_pad_size:
.word tegra30_sdram_pad_address_end - tegra30_sdram_pad_address
2013-08-12 03:40:06 -06:00
tegra114_sdram_pad_size:
.word tegra114_sdram_pad_adress_end - tegra114_sdram_pad_address
ARM: tegra: add LP1 suspend support for Tegra30 The LP1 suspend mode will power off the CPU, clock gated the PLLs and put SDRAM to self-refresh mode. Any interrupt can wake up device from LP1. The sequence when LP1 suspending: * tunning off L1 data cache and the MMU * storing some EMC registers, DPD (deep power down) status, clk source of mselect and SCLK burst policy * putting SDRAM into self-refresh * switching CPU to CLK_M (12MHz OSC) * tunning off PLLM, PLLP, PLLA, PLLC and PLLX * switching SCLK to CLK_S (32KHz OSC) * shutting off the CPU rail The sequence of LP1 resuming: * re-enabling PLLM, PLLP, PLLA, PLLC and PLLX * restoring the clk source of mselect and SCLK burst policy * setting up CCLK burst policy to PLLX * restoring DPD status and some EMC registers * resuming SDRAM to normal mode * jumping to the "tegra_resume" from PMC_SCRATCH41 Due to the SDRAM will be put into self-refresh mode, the low level procedures of LP1 suspending and resuming should be copied to TEGRA_IRAM_CODE_AREA (TEGRA_IRAM_BASE + SZ_4K) when suspending. Before restoring the CPU context when resuming, the SDRAM needs to be switched back to normal mode. And the PLLs need to be re-enabled, SCLK burst policy be restored, CCLK burst policy be set in PLLX. Then jumping to "tegra_resume" that was expected to be stored in PMC_SCRATCH41 to restore CPU context and back to kernel. Based on the work by: Scott Williams <scwilliams@nvidia.com> Signed-off-by: Joseph Lo <josephl@nvidia.com> Signed-off-by: Stephen Warren <swarren@nvidia.com>
2013-08-12 03:40:04 -06:00
.type tegra_sdram_pad_save, %object
tegra_sdram_pad_save:
.rept (tegra114_sdram_pad_adress_end - tegra114_sdram_pad_address) / 4
ARM: tegra: add LP1 suspend support for Tegra30 The LP1 suspend mode will power off the CPU, clock gated the PLLs and put SDRAM to self-refresh mode. Any interrupt can wake up device from LP1. The sequence when LP1 suspending: * tunning off L1 data cache and the MMU * storing some EMC registers, DPD (deep power down) status, clk source of mselect and SCLK burst policy * putting SDRAM into self-refresh * switching CPU to CLK_M (12MHz OSC) * tunning off PLLM, PLLP, PLLA, PLLC and PLLX * switching SCLK to CLK_S (32KHz OSC) * shutting off the CPU rail The sequence of LP1 resuming: * re-enabling PLLM, PLLP, PLLA, PLLC and PLLX * restoring the clk source of mselect and SCLK burst policy * setting up CCLK burst policy to PLLX * restoring DPD status and some EMC registers * resuming SDRAM to normal mode * jumping to the "tegra_resume" from PMC_SCRATCH41 Due to the SDRAM will be put into self-refresh mode, the low level procedures of LP1 suspending and resuming should be copied to TEGRA_IRAM_CODE_AREA (TEGRA_IRAM_BASE + SZ_4K) when suspending. Before restoring the CPU context when resuming, the SDRAM needs to be switched back to normal mode. And the PLLs need to be re-enabled, SCLK burst policy be restored, CCLK burst policy be set in PLLX. Then jumping to "tegra_resume" that was expected to be stored in PMC_SCRATCH41 to restore CPU context and back to kernel. Based on the work by: Scott Williams <scwilliams@nvidia.com> Signed-off-by: Joseph Lo <josephl@nvidia.com> Signed-off-by: Stephen Warren <swarren@nvidia.com>
2013-08-12 03:40:04 -06:00
.long 0
.endr
/*
* tegra30_tear_down_core
*
* copied into and executed from IRAM
* puts memory in self-refresh for LP0 and LP1
*/
tegra30_tear_down_core:
bl tegra30_sdram_self_refresh
bl tegra30_switch_cpu_to_clk32k
b tegra30_enter_sleep
/*
* tegra30_switch_cpu_to_clk32k
*
* In LP0 and LP1 all PLLs will be turned off. Switching the CPU and System CLK
* to the 32KHz clock.
* r4 = TEGRA_PMC_BASE
* r5 = TEGRA_CLK_RESET_BASE
* r6 = TEGRA_FLOW_CTRL_BASE
* r7 = TEGRA_TMRUS_BASE
2013-08-12 03:40:06 -06:00
* r10= SoC ID
ARM: tegra: add LP1 suspend support for Tegra30 The LP1 suspend mode will power off the CPU, clock gated the PLLs and put SDRAM to self-refresh mode. Any interrupt can wake up device from LP1. The sequence when LP1 suspending: * tunning off L1 data cache and the MMU * storing some EMC registers, DPD (deep power down) status, clk source of mselect and SCLK burst policy * putting SDRAM into self-refresh * switching CPU to CLK_M (12MHz OSC) * tunning off PLLM, PLLP, PLLA, PLLC and PLLX * switching SCLK to CLK_S (32KHz OSC) * shutting off the CPU rail The sequence of LP1 resuming: * re-enabling PLLM, PLLP, PLLA, PLLC and PLLX * restoring the clk source of mselect and SCLK burst policy * setting up CCLK burst policy to PLLX * restoring DPD status and some EMC registers * resuming SDRAM to normal mode * jumping to the "tegra_resume" from PMC_SCRATCH41 Due to the SDRAM will be put into self-refresh mode, the low level procedures of LP1 suspending and resuming should be copied to TEGRA_IRAM_CODE_AREA (TEGRA_IRAM_BASE + SZ_4K) when suspending. Before restoring the CPU context when resuming, the SDRAM needs to be switched back to normal mode. And the PLLs need to be re-enabled, SCLK burst policy be restored, CCLK burst policy be set in PLLX. Then jumping to "tegra_resume" that was expected to be stored in PMC_SCRATCH41 to restore CPU context and back to kernel. Based on the work by: Scott Williams <scwilliams@nvidia.com> Signed-off-by: Joseph Lo <josephl@nvidia.com> Signed-off-by: Stephen Warren <swarren@nvidia.com>
2013-08-12 03:40:04 -06:00
*/
tegra30_switch_cpu_to_clk32k:
/*
* start by jumping to CLKM to safely disable PLLs, then jump to
* CLKS.
*/
mov r0, #(1 << 28)
str r0, [r5, #CLK_RESET_SCLK_BURST]
/* 2uS delay delay between changing SCLK and CCLK */
ldr r1, [r7]
add r1, r1, #2
wait_until r1, r7, r9
str r0, [r5, #CLK_RESET_CCLK_BURST]
mov r0, #0
str r0, [r5, #CLK_RESET_CCLK_DIVIDER]
str r0, [r5, #CLK_RESET_SCLK_DIVIDER]
/* switch the clock source of mselect to be CLK_M */
ldr r0, [r5, #CLK_RESET_CLK_SOURCE_MSELECT]
orr r0, r0, #MSELECT_CLKM
str r0, [r5, #CLK_RESET_CLK_SOURCE_MSELECT]
/* 2uS delay delay between changing SCLK and disabling PLLs */
ldr r1, [r7]
add r1, r1, #2
wait_until r1, r7, r9
/* disable PLLM via PMC in LP1 */
ldr r0, [r4, #PMC_PLLP_WB0_OVERRIDE]
bic r0, r0, #(1 << 12)
str r0, [r4, #PMC_PLLP_WB0_OVERRIDE]
/* disable PLLP, PLLA, PLLC and PLLX */
ARM: tegra: Enable PLLP bypass during Tegra124 LP1 commit 1a3388d506bf5b45bb283e6a4c4706cfb4897333 upstream. For a little over a year, U-Boot has configured the flow controller to perform automatic RAM re-repair on off->on power transitions of the CPU rail[1]. This is mandatory for correct operation of Tegra124. However, RAM re-repair relies on certain clocks, which the kernel must enable and leave running. PLLP is one of those clocks. This clock is shut down during LP1 in order to save power. Enable bypass (which I believe routes osc_div_clk, essentially the crystal clock, to the PLL output) so that this clock signal toggles even though the PLL is not active. This is required so that LP1 power mode (system suspend) operates correctly. The bypass configuration must then be undone when resuming from LP1, so that all peripheral clocks run at the expected rate. Without this, many peripherals won't work correctly; for example, the UART baud rate would be incorrect. NVIDIA's downstream kernel code only does this if not compiled for Tegra30, so the added code is made conditional upon the chip ID. NVIDIA's downstream code makes this change conditional upon the active CPU cluster. The upstream kernel currently doesn't support cluster switching, so this patch doesn't test the active CPU cluster ID. [1] 3cc7942a4ae5 ARM: tegra: implement RAM repair Reported-by: Jonathan Hunter <jonathanh@nvidia.com> Cc: stable@vger.kernel.org Signed-off-by: Stephen Warren <swarren@nvidia.com> Signed-off-by: Thierry Reding <treding@nvidia.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2019-10-03 14:50:31 -06:00
tegra_get_soc_id TEGRA_APB_MISC_BASE, r1
cmp r1, #TEGRA30
ARM: tegra: add LP1 suspend support for Tegra30 The LP1 suspend mode will power off the CPU, clock gated the PLLs and put SDRAM to self-refresh mode. Any interrupt can wake up device from LP1. The sequence when LP1 suspending: * tunning off L1 data cache and the MMU * storing some EMC registers, DPD (deep power down) status, clk source of mselect and SCLK burst policy * putting SDRAM into self-refresh * switching CPU to CLK_M (12MHz OSC) * tunning off PLLM, PLLP, PLLA, PLLC and PLLX * switching SCLK to CLK_S (32KHz OSC) * shutting off the CPU rail The sequence of LP1 resuming: * re-enabling PLLM, PLLP, PLLA, PLLC and PLLX * restoring the clk source of mselect and SCLK burst policy * setting up CCLK burst policy to PLLX * restoring DPD status and some EMC registers * resuming SDRAM to normal mode * jumping to the "tegra_resume" from PMC_SCRATCH41 Due to the SDRAM will be put into self-refresh mode, the low level procedures of LP1 suspending and resuming should be copied to TEGRA_IRAM_CODE_AREA (TEGRA_IRAM_BASE + SZ_4K) when suspending. Before restoring the CPU context when resuming, the SDRAM needs to be switched back to normal mode. And the PLLs need to be re-enabled, SCLK burst policy be restored, CCLK burst policy be set in PLLX. Then jumping to "tegra_resume" that was expected to be stored in PMC_SCRATCH41 to restore CPU context and back to kernel. Based on the work by: Scott Williams <scwilliams@nvidia.com> Signed-off-by: Joseph Lo <josephl@nvidia.com> Signed-off-by: Stephen Warren <swarren@nvidia.com>
2013-08-12 03:40:04 -06:00
ldr r0, [r5, #CLK_RESET_PLLP_BASE]
ARM: tegra: Enable PLLP bypass during Tegra124 LP1 commit 1a3388d506bf5b45bb283e6a4c4706cfb4897333 upstream. For a little over a year, U-Boot has configured the flow controller to perform automatic RAM re-repair on off->on power transitions of the CPU rail[1]. This is mandatory for correct operation of Tegra124. However, RAM re-repair relies on certain clocks, which the kernel must enable and leave running. PLLP is one of those clocks. This clock is shut down during LP1 in order to save power. Enable bypass (which I believe routes osc_div_clk, essentially the crystal clock, to the PLL output) so that this clock signal toggles even though the PLL is not active. This is required so that LP1 power mode (system suspend) operates correctly. The bypass configuration must then be undone when resuming from LP1, so that all peripheral clocks run at the expected rate. Without this, many peripherals won't work correctly; for example, the UART baud rate would be incorrect. NVIDIA's downstream kernel code only does this if not compiled for Tegra30, so the added code is made conditional upon the chip ID. NVIDIA's downstream code makes this change conditional upon the active CPU cluster. The upstream kernel currently doesn't support cluster switching, so this patch doesn't test the active CPU cluster ID. [1] 3cc7942a4ae5 ARM: tegra: implement RAM repair Reported-by: Jonathan Hunter <jonathanh@nvidia.com> Cc: stable@vger.kernel.org Signed-off-by: Stephen Warren <swarren@nvidia.com> Signed-off-by: Thierry Reding <treding@nvidia.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2019-10-03 14:50:31 -06:00
orrne r0, r0, #(1 << 31) @ enable PllP bypass on fast cluster
ARM: tegra: add LP1 suspend support for Tegra30 The LP1 suspend mode will power off the CPU, clock gated the PLLs and put SDRAM to self-refresh mode. Any interrupt can wake up device from LP1. The sequence when LP1 suspending: * tunning off L1 data cache and the MMU * storing some EMC registers, DPD (deep power down) status, clk source of mselect and SCLK burst policy * putting SDRAM into self-refresh * switching CPU to CLK_M (12MHz OSC) * tunning off PLLM, PLLP, PLLA, PLLC and PLLX * switching SCLK to CLK_S (32KHz OSC) * shutting off the CPU rail The sequence of LP1 resuming: * re-enabling PLLM, PLLP, PLLA, PLLC and PLLX * restoring the clk source of mselect and SCLK burst policy * setting up CCLK burst policy to PLLX * restoring DPD status and some EMC registers * resuming SDRAM to normal mode * jumping to the "tegra_resume" from PMC_SCRATCH41 Due to the SDRAM will be put into self-refresh mode, the low level procedures of LP1 suspending and resuming should be copied to TEGRA_IRAM_CODE_AREA (TEGRA_IRAM_BASE + SZ_4K) when suspending. Before restoring the CPU context when resuming, the SDRAM needs to be switched back to normal mode. And the PLLs need to be re-enabled, SCLK burst policy be restored, CCLK burst policy be set in PLLX. Then jumping to "tegra_resume" that was expected to be stored in PMC_SCRATCH41 to restore CPU context and back to kernel. Based on the work by: Scott Williams <scwilliams@nvidia.com> Signed-off-by: Joseph Lo <josephl@nvidia.com> Signed-off-by: Stephen Warren <swarren@nvidia.com>
2013-08-12 03:40:04 -06:00
bic r0, r0, #(1 << 30)
str r0, [r5, #CLK_RESET_PLLP_BASE]
ldr r0, [r5, #CLK_RESET_PLLA_BASE]
bic r0, r0, #(1 << 30)
str r0, [r5, #CLK_RESET_PLLA_BASE]
ldr r0, [r5, #CLK_RESET_PLLC_BASE]
bic r0, r0, #(1 << 30)
str r0, [r5, #CLK_RESET_PLLC_BASE]
ldr r0, [r5, #CLK_RESET_PLLX_BASE]
bic r0, r0, #(1 << 30)
str r0, [r5, #CLK_RESET_PLLX_BASE]
2013-08-12 03:40:06 -06:00
cmp r10, #TEGRA30
beq _no_pll_in_iddq
pll_iddq_entry r1, r5, CLK_RESET_PLLX_MISC3, CLK_RESET_PLLX_MISC3_IDDQ
_no_pll_in_iddq:
ARM: tegra: add LP1 suspend support for Tegra30 The LP1 suspend mode will power off the CPU, clock gated the PLLs and put SDRAM to self-refresh mode. Any interrupt can wake up device from LP1. The sequence when LP1 suspending: * tunning off L1 data cache and the MMU * storing some EMC registers, DPD (deep power down) status, clk source of mselect and SCLK burst policy * putting SDRAM into self-refresh * switching CPU to CLK_M (12MHz OSC) * tunning off PLLM, PLLP, PLLA, PLLC and PLLX * switching SCLK to CLK_S (32KHz OSC) * shutting off the CPU rail The sequence of LP1 resuming: * re-enabling PLLM, PLLP, PLLA, PLLC and PLLX * restoring the clk source of mselect and SCLK burst policy * setting up CCLK burst policy to PLLX * restoring DPD status and some EMC registers * resuming SDRAM to normal mode * jumping to the "tegra_resume" from PMC_SCRATCH41 Due to the SDRAM will be put into self-refresh mode, the low level procedures of LP1 suspending and resuming should be copied to TEGRA_IRAM_CODE_AREA (TEGRA_IRAM_BASE + SZ_4K) when suspending. Before restoring the CPU context when resuming, the SDRAM needs to be switched back to normal mode. And the PLLs need to be re-enabled, SCLK burst policy be restored, CCLK burst policy be set in PLLX. Then jumping to "tegra_resume" that was expected to be stored in PMC_SCRATCH41 to restore CPU context and back to kernel. Based on the work by: Scott Williams <scwilliams@nvidia.com> Signed-off-by: Joseph Lo <josephl@nvidia.com> Signed-off-by: Stephen Warren <swarren@nvidia.com>
2013-08-12 03:40:04 -06:00
/* switch to CLKS */
mov r0, #0 /* brust policy = 32KHz */
str r0, [r5, #CLK_RESET_SCLK_BURST]
ARM: convert all "mov.* pc, reg" to "bx reg" for ARMv6+ ARMv6 and greater introduced a new instruction ("bx") which can be used to return from function calls. Recent CPUs perform better when the "bx lr" instruction is used rather than the "mov pc, lr" instruction, and this sequence is strongly recommended to be used by the ARM architecture manual (section A.4.1.1). We provide a new macro "ret" with all its variants for the condition code which will resolve to the appropriate instruction. Rather than doing this piecemeal, and miss some instances, change all the "mov pc" instances to use the new macro, with the exception of the "movs" instruction and the kprobes code. This allows us to detect the "mov pc, lr" case and fix it up - and also gives us the possibility of deploying this for other registers depending on the CPU selection. Reported-by: Will Deacon <will.deacon@arm.com> Tested-by: Stephen Warren <swarren@nvidia.com> # Tegra Jetson TK1 Tested-by: Robert Jarzmik <robert.jarzmik@free.fr> # mioa701_bootresume.S Tested-by: Andrew Lunn <andrew@lunn.ch> # Kirkwood Tested-by: Shawn Guo <shawn.guo@freescale.com> Tested-by: Tony Lindgren <tony@atomide.com> # OMAPs Tested-by: Gregory CLEMENT <gregory.clement@free-electrons.com> # Armada XP, 375, 385 Acked-by: Sekhar Nori <nsekhar@ti.com> # DaVinci Acked-by: Christoffer Dall <christoffer.dall@linaro.org> # kvm/hyp Acked-by: Haojian Zhuang <haojian.zhuang@gmail.com> # PXA3xx Acked-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com> # Xen Tested-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de> # ARMv7M Tested-by: Simon Horman <horms+renesas@verge.net.au> # Shmobile Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
2014-06-30 09:29:12 -06:00
ret lr
ARM: tegra: add LP1 suspend support for Tegra30 The LP1 suspend mode will power off the CPU, clock gated the PLLs and put SDRAM to self-refresh mode. Any interrupt can wake up device from LP1. The sequence when LP1 suspending: * tunning off L1 data cache and the MMU * storing some EMC registers, DPD (deep power down) status, clk source of mselect and SCLK burst policy * putting SDRAM into self-refresh * switching CPU to CLK_M (12MHz OSC) * tunning off PLLM, PLLP, PLLA, PLLC and PLLX * switching SCLK to CLK_S (32KHz OSC) * shutting off the CPU rail The sequence of LP1 resuming: * re-enabling PLLM, PLLP, PLLA, PLLC and PLLX * restoring the clk source of mselect and SCLK burst policy * setting up CCLK burst policy to PLLX * restoring DPD status and some EMC registers * resuming SDRAM to normal mode * jumping to the "tegra_resume" from PMC_SCRATCH41 Due to the SDRAM will be put into self-refresh mode, the low level procedures of LP1 suspending and resuming should be copied to TEGRA_IRAM_CODE_AREA (TEGRA_IRAM_BASE + SZ_4K) when suspending. Before restoring the CPU context when resuming, the SDRAM needs to be switched back to normal mode. And the PLLs need to be re-enabled, SCLK burst policy be restored, CCLK burst policy be set in PLLX. Then jumping to "tegra_resume" that was expected to be stored in PMC_SCRATCH41 to restore CPU context and back to kernel. Based on the work by: Scott Williams <scwilliams@nvidia.com> Signed-off-by: Joseph Lo <josephl@nvidia.com> Signed-off-by: Stephen Warren <swarren@nvidia.com>
2013-08-12 03:40:04 -06:00
/*
* tegra30_enter_sleep
*
* uses flow controller to enter sleep state
* executes from IRAM with SDRAM in selfrefresh when target state is LP0 or LP1
* executes from SDRAM with target state is LP2
* r6 = TEGRA_FLOW_CTRL_BASE
*/
tegra30_enter_sleep:
cpu_id r1
cpu_to_csr_reg r2, r1
ldr r0, [r6, r2]
orr r0, r0, #FLOW_CTRL_CSR_INTR_FLAG | FLOW_CTRL_CSR_EVENT_FLAG
orr r0, r0, #FLOW_CTRL_CSR_ENABLE
str r0, [r6, r2]
tegra_get_soc_id TEGRA_APB_MISC_BASE, r10
cmp r10, #TEGRA30
mov r0, #FLOW_CTRL_WAIT_FOR_INTERRUPT
orreq r0, r0, #FLOW_CTRL_HALT_CPU_IRQ | FLOW_CTRL_HALT_CPU_FIQ
orrne r0, r0, #FLOW_CTRL_HALT_LIC_IRQ | FLOW_CTRL_HALT_LIC_FIQ
cpu_to_halt_reg r2, r1
str r0, [r6, r2]
dsb
ldr r0, [r6, r2] /* memory barrier */
halted:
isb
dsb
wfi /* CPU should be power gated here */
/* !!!FIXME!!! Implement halt failure handler */
b halted
ARM: tegra: add LP1 suspend support for Tegra30 The LP1 suspend mode will power off the CPU, clock gated the PLLs and put SDRAM to self-refresh mode. Any interrupt can wake up device from LP1. The sequence when LP1 suspending: * tunning off L1 data cache and the MMU * storing some EMC registers, DPD (deep power down) status, clk source of mselect and SCLK burst policy * putting SDRAM into self-refresh * switching CPU to CLK_M (12MHz OSC) * tunning off PLLM, PLLP, PLLA, PLLC and PLLX * switching SCLK to CLK_S (32KHz OSC) * shutting off the CPU rail The sequence of LP1 resuming: * re-enabling PLLM, PLLP, PLLA, PLLC and PLLX * restoring the clk source of mselect and SCLK burst policy * setting up CCLK burst policy to PLLX * restoring DPD status and some EMC registers * resuming SDRAM to normal mode * jumping to the "tegra_resume" from PMC_SCRATCH41 Due to the SDRAM will be put into self-refresh mode, the low level procedures of LP1 suspending and resuming should be copied to TEGRA_IRAM_CODE_AREA (TEGRA_IRAM_BASE + SZ_4K) when suspending. Before restoring the CPU context when resuming, the SDRAM needs to be switched back to normal mode. And the PLLs need to be re-enabled, SCLK burst policy be restored, CCLK burst policy be set in PLLX. Then jumping to "tegra_resume" that was expected to be stored in PMC_SCRATCH41 to restore CPU context and back to kernel. Based on the work by: Scott Williams <scwilliams@nvidia.com> Signed-off-by: Joseph Lo <josephl@nvidia.com> Signed-off-by: Stephen Warren <swarren@nvidia.com>
2013-08-12 03:40:04 -06:00
/*
* tegra30_sdram_self_refresh
*
* called with MMU off and caches disabled
* must be executed from IRAM
* r4 = TEGRA_PMC_BASE
* r5 = TEGRA_CLK_RESET_BASE
* r6 = TEGRA_FLOW_CTRL_BASE
* r7 = TEGRA_TMRUS_BASE
2013-08-12 03:40:06 -06:00
* r10= SoC ID
ARM: tegra: add LP1 suspend support for Tegra30 The LP1 suspend mode will power off the CPU, clock gated the PLLs and put SDRAM to self-refresh mode. Any interrupt can wake up device from LP1. The sequence when LP1 suspending: * tunning off L1 data cache and the MMU * storing some EMC registers, DPD (deep power down) status, clk source of mselect and SCLK burst policy * putting SDRAM into self-refresh * switching CPU to CLK_M (12MHz OSC) * tunning off PLLM, PLLP, PLLA, PLLC and PLLX * switching SCLK to CLK_S (32KHz OSC) * shutting off the CPU rail The sequence of LP1 resuming: * re-enabling PLLM, PLLP, PLLA, PLLC and PLLX * restoring the clk source of mselect and SCLK burst policy * setting up CCLK burst policy to PLLX * restoring DPD status and some EMC registers * resuming SDRAM to normal mode * jumping to the "tegra_resume" from PMC_SCRATCH41 Due to the SDRAM will be put into self-refresh mode, the low level procedures of LP1 suspending and resuming should be copied to TEGRA_IRAM_CODE_AREA (TEGRA_IRAM_BASE + SZ_4K) when suspending. Before restoring the CPU context when resuming, the SDRAM needs to be switched back to normal mode. And the PLLs need to be re-enabled, SCLK burst policy be restored, CCLK burst policy be set in PLLX. Then jumping to "tegra_resume" that was expected to be stored in PMC_SCRATCH41 to restore CPU context and back to kernel. Based on the work by: Scott Williams <scwilliams@nvidia.com> Signed-off-by: Joseph Lo <josephl@nvidia.com> Signed-off-by: Stephen Warren <swarren@nvidia.com>
2013-08-12 03:40:04 -06:00
*/
tegra30_sdram_self_refresh:
adr r8, tegra_sdram_pad_save
2013-08-12 03:40:06 -06:00
tegra_get_soc_id TEGRA_APB_MISC_BASE, r10
cmp r10, #TEGRA30
adreq r2, tegra30_sdram_pad_address
ldreq r3, tegra30_sdram_pad_size
cmp r10, #TEGRA114
adreq r2, tegra114_sdram_pad_address
ldreq r3, tegra114_sdram_pad_size
cmp r10, #TEGRA124
adreq r2, tegra124_sdram_pad_address
ldreq r3, tegra30_sdram_pad_size
ARM: tegra: add LP1 suspend support for Tegra30 The LP1 suspend mode will power off the CPU, clock gated the PLLs and put SDRAM to self-refresh mode. Any interrupt can wake up device from LP1. The sequence when LP1 suspending: * tunning off L1 data cache and the MMU * storing some EMC registers, DPD (deep power down) status, clk source of mselect and SCLK burst policy * putting SDRAM into self-refresh * switching CPU to CLK_M (12MHz OSC) * tunning off PLLM, PLLP, PLLA, PLLC and PLLX * switching SCLK to CLK_S (32KHz OSC) * shutting off the CPU rail The sequence of LP1 resuming: * re-enabling PLLM, PLLP, PLLA, PLLC and PLLX * restoring the clk source of mselect and SCLK burst policy * setting up CCLK burst policy to PLLX * restoring DPD status and some EMC registers * resuming SDRAM to normal mode * jumping to the "tegra_resume" from PMC_SCRATCH41 Due to the SDRAM will be put into self-refresh mode, the low level procedures of LP1 suspending and resuming should be copied to TEGRA_IRAM_CODE_AREA (TEGRA_IRAM_BASE + SZ_4K) when suspending. Before restoring the CPU context when resuming, the SDRAM needs to be switched back to normal mode. And the PLLs need to be re-enabled, SCLK burst policy be restored, CCLK burst policy be set in PLLX. Then jumping to "tegra_resume" that was expected to be stored in PMC_SCRATCH41 to restore CPU context and back to kernel. Based on the work by: Scott Williams <scwilliams@nvidia.com> Signed-off-by: Joseph Lo <josephl@nvidia.com> Signed-off-by: Stephen Warren <swarren@nvidia.com>
2013-08-12 03:40:04 -06:00
mov r9, #0
padsave:
ldr r0, [r2, r9] @ r0 is the addr in the pad_address
ldr r1, [r0]
str r1, [r8, r9] @ save the content of the addr
add r9, r9, #4
cmp r3, r9
bne padsave
padsave_done:
dsb
2013-08-12 03:40:06 -06:00
cmp r10, #TEGRA30
ldreq r0, =TEGRA_EMC_BASE @ r0 reserved for emc base addr
cmp r10, #TEGRA114
ldreq r0, =TEGRA_EMC0_BASE
cmp r10, #TEGRA124
ldreq r0, =TEGRA124_EMC_BASE
ARM: tegra: add LP1 suspend support for Tegra30 The LP1 suspend mode will power off the CPU, clock gated the PLLs and put SDRAM to self-refresh mode. Any interrupt can wake up device from LP1. The sequence when LP1 suspending: * tunning off L1 data cache and the MMU * storing some EMC registers, DPD (deep power down) status, clk source of mselect and SCLK burst policy * putting SDRAM into self-refresh * switching CPU to CLK_M (12MHz OSC) * tunning off PLLM, PLLP, PLLA, PLLC and PLLX * switching SCLK to CLK_S (32KHz OSC) * shutting off the CPU rail The sequence of LP1 resuming: * re-enabling PLLM, PLLP, PLLA, PLLC and PLLX * restoring the clk source of mselect and SCLK burst policy * setting up CCLK burst policy to PLLX * restoring DPD status and some EMC registers * resuming SDRAM to normal mode * jumping to the "tegra_resume" from PMC_SCRATCH41 Due to the SDRAM will be put into self-refresh mode, the low level procedures of LP1 suspending and resuming should be copied to TEGRA_IRAM_CODE_AREA (TEGRA_IRAM_BASE + SZ_4K) when suspending. Before restoring the CPU context when resuming, the SDRAM needs to be switched back to normal mode. And the PLLs need to be re-enabled, SCLK burst policy be restored, CCLK burst policy be set in PLLX. Then jumping to "tegra_resume" that was expected to be stored in PMC_SCRATCH41 to restore CPU context and back to kernel. Based on the work by: Scott Williams <scwilliams@nvidia.com> Signed-off-by: Joseph Lo <josephl@nvidia.com> Signed-off-by: Stephen Warren <swarren@nvidia.com>
2013-08-12 03:40:04 -06:00
2013-08-12 03:40:06 -06:00
enter_self_refresh:
cmp r10, #TEGRA30
ARM: tegra: add LP1 suspend support for Tegra30 The LP1 suspend mode will power off the CPU, clock gated the PLLs and put SDRAM to self-refresh mode. Any interrupt can wake up device from LP1. The sequence when LP1 suspending: * tunning off L1 data cache and the MMU * storing some EMC registers, DPD (deep power down) status, clk source of mselect and SCLK burst policy * putting SDRAM into self-refresh * switching CPU to CLK_M (12MHz OSC) * tunning off PLLM, PLLP, PLLA, PLLC and PLLX * switching SCLK to CLK_S (32KHz OSC) * shutting off the CPU rail The sequence of LP1 resuming: * re-enabling PLLM, PLLP, PLLA, PLLC and PLLX * restoring the clk source of mselect and SCLK burst policy * setting up CCLK burst policy to PLLX * restoring DPD status and some EMC registers * resuming SDRAM to normal mode * jumping to the "tegra_resume" from PMC_SCRATCH41 Due to the SDRAM will be put into self-refresh mode, the low level procedures of LP1 suspending and resuming should be copied to TEGRA_IRAM_CODE_AREA (TEGRA_IRAM_BASE + SZ_4K) when suspending. Before restoring the CPU context when resuming, the SDRAM needs to be switched back to normal mode. And the PLLs need to be re-enabled, SCLK burst policy be restored, CCLK burst policy be set in PLLX. Then jumping to "tegra_resume" that was expected to be stored in PMC_SCRATCH41 to restore CPU context and back to kernel. Based on the work by: Scott Williams <scwilliams@nvidia.com> Signed-off-by: Joseph Lo <josephl@nvidia.com> Signed-off-by: Stephen Warren <swarren@nvidia.com>
2013-08-12 03:40:04 -06:00
mov r1, #0
str r1, [r0, #EMC_ZCAL_INTERVAL]
str r1, [r0, #EMC_AUTO_CAL_INTERVAL]
ldr r1, [r0, #EMC_CFG]
bic r1, r1, #(1 << 28)
2013-08-12 03:40:06 -06:00
bicne r1, r1, #(1 << 29)
ARM: tegra: add LP1 suspend support for Tegra30 The LP1 suspend mode will power off the CPU, clock gated the PLLs and put SDRAM to self-refresh mode. Any interrupt can wake up device from LP1. The sequence when LP1 suspending: * tunning off L1 data cache and the MMU * storing some EMC registers, DPD (deep power down) status, clk source of mselect and SCLK burst policy * putting SDRAM into self-refresh * switching CPU to CLK_M (12MHz OSC) * tunning off PLLM, PLLP, PLLA, PLLC and PLLX * switching SCLK to CLK_S (32KHz OSC) * shutting off the CPU rail The sequence of LP1 resuming: * re-enabling PLLM, PLLP, PLLA, PLLC and PLLX * restoring the clk source of mselect and SCLK burst policy * setting up CCLK burst policy to PLLX * restoring DPD status and some EMC registers * resuming SDRAM to normal mode * jumping to the "tegra_resume" from PMC_SCRATCH41 Due to the SDRAM will be put into self-refresh mode, the low level procedures of LP1 suspending and resuming should be copied to TEGRA_IRAM_CODE_AREA (TEGRA_IRAM_BASE + SZ_4K) when suspending. Before restoring the CPU context when resuming, the SDRAM needs to be switched back to normal mode. And the PLLs need to be re-enabled, SCLK burst policy be restored, CCLK burst policy be set in PLLX. Then jumping to "tegra_resume" that was expected to be stored in PMC_SCRATCH41 to restore CPU context and back to kernel. Based on the work by: Scott Williams <scwilliams@nvidia.com> Signed-off-by: Joseph Lo <josephl@nvidia.com> Signed-off-by: Stephen Warren <swarren@nvidia.com>
2013-08-12 03:40:04 -06:00
str r1, [r0, #EMC_CFG] @ disable DYN_SELF_REF
emc_timing_update r1, r0
ldr r1, [r7]
add r1, r1, #5
wait_until r1, r7, r2
emc_wait_auto_cal:
ldr r1, [r0, #EMC_AUTO_CAL_STATUS]
tst r1, #(1 << 31) @ wait until AUTO_CAL_ACTIVE is cleared
bne emc_wait_auto_cal
mov r1, #3
str r1, [r0, #EMC_REQ_CTRL] @ stall incoming DRAM requests
emcidle:
ldr r1, [r0, #EMC_EMC_STATUS]
tst r1, #4
beq emcidle
mov r1, #1
str r1, [r0, #EMC_SELF_REF]
emc_device_mask r1, r0
emcself:
ldr r2, [r0, #EMC_EMC_STATUS]
and r2, r2, r1
cmp r2, r1
bne emcself @ loop until DDR in self-refresh
/* Put VTTGEN in the lowest power mode */
ldr r1, [r0, #EMC_XM2VTTGENPADCTRL]
mov32 r2, 0xF8F8FFFF @ clear XM2VTTGEN_DRVUP and XM2VTTGEN_DRVDN
and r1, r1, r2
str r1, [r0, #EMC_XM2VTTGENPADCTRL]
ldr r1, [r0, #EMC_XM2VTTGENPADCTRL2]
2013-08-12 03:40:06 -06:00
cmp r10, #TEGRA30
orreq r1, r1, #7 @ set E_NO_VTTGEN
orrne r1, r1, #0x3f
ARM: tegra: add LP1 suspend support for Tegra30 The LP1 suspend mode will power off the CPU, clock gated the PLLs and put SDRAM to self-refresh mode. Any interrupt can wake up device from LP1. The sequence when LP1 suspending: * tunning off L1 data cache and the MMU * storing some EMC registers, DPD (deep power down) status, clk source of mselect and SCLK burst policy * putting SDRAM into self-refresh * switching CPU to CLK_M (12MHz OSC) * tunning off PLLM, PLLP, PLLA, PLLC and PLLX * switching SCLK to CLK_S (32KHz OSC) * shutting off the CPU rail The sequence of LP1 resuming: * re-enabling PLLM, PLLP, PLLA, PLLC and PLLX * restoring the clk source of mselect and SCLK burst policy * setting up CCLK burst policy to PLLX * restoring DPD status and some EMC registers * resuming SDRAM to normal mode * jumping to the "tegra_resume" from PMC_SCRATCH41 Due to the SDRAM will be put into self-refresh mode, the low level procedures of LP1 suspending and resuming should be copied to TEGRA_IRAM_CODE_AREA (TEGRA_IRAM_BASE + SZ_4K) when suspending. Before restoring the CPU context when resuming, the SDRAM needs to be switched back to normal mode. And the PLLs need to be re-enabled, SCLK burst policy be restored, CCLK burst policy be set in PLLX. Then jumping to "tegra_resume" that was expected to be stored in PMC_SCRATCH41 to restore CPU context and back to kernel. Based on the work by: Scott Williams <scwilliams@nvidia.com> Signed-off-by: Joseph Lo <josephl@nvidia.com> Signed-off-by: Stephen Warren <swarren@nvidia.com>
2013-08-12 03:40:04 -06:00
str r1, [r0, #EMC_XM2VTTGENPADCTRL2]
emc_timing_update r1, r0
2013-08-12 03:40:06 -06:00
/* Tegra114 had dual EMC channel, now config the other one */
cmp r10, #TEGRA114
bne no_dual_emc_chanl
mov32 r1, TEGRA_EMC1_BASE
cmp r0, r1
movne r0, r1
bne enter_self_refresh
no_dual_emc_chanl:
ARM: tegra: add LP1 suspend support for Tegra30 The LP1 suspend mode will power off the CPU, clock gated the PLLs and put SDRAM to self-refresh mode. Any interrupt can wake up device from LP1. The sequence when LP1 suspending: * tunning off L1 data cache and the MMU * storing some EMC registers, DPD (deep power down) status, clk source of mselect and SCLK burst policy * putting SDRAM into self-refresh * switching CPU to CLK_M (12MHz OSC) * tunning off PLLM, PLLP, PLLA, PLLC and PLLX * switching SCLK to CLK_S (32KHz OSC) * shutting off the CPU rail The sequence of LP1 resuming: * re-enabling PLLM, PLLP, PLLA, PLLC and PLLX * restoring the clk source of mselect and SCLK burst policy * setting up CCLK burst policy to PLLX * restoring DPD status and some EMC registers * resuming SDRAM to normal mode * jumping to the "tegra_resume" from PMC_SCRATCH41 Due to the SDRAM will be put into self-refresh mode, the low level procedures of LP1 suspending and resuming should be copied to TEGRA_IRAM_CODE_AREA (TEGRA_IRAM_BASE + SZ_4K) when suspending. Before restoring the CPU context when resuming, the SDRAM needs to be switched back to normal mode. And the PLLs need to be re-enabled, SCLK burst policy be restored, CCLK burst policy be set in PLLX. Then jumping to "tegra_resume" that was expected to be stored in PMC_SCRATCH41 to restore CPU context and back to kernel. Based on the work by: Scott Williams <scwilliams@nvidia.com> Signed-off-by: Joseph Lo <josephl@nvidia.com> Signed-off-by: Stephen Warren <swarren@nvidia.com>
2013-08-12 03:40:04 -06:00
ldr r1, [r4, #PMC_CTRL]
tst r1, #PMC_CTRL_SIDE_EFFECT_LP0
bne pmc_io_dpd_skip
/*
* Put DDR_DATA, DISC_ADDR_CMD, DDR_ADDR_CMD, POP_ADDR_CMD, POP_CLK
* and COMP in the lowest power mode when LP1.
*/
mov32 r1, 0x8EC00000
str r1, [r4, #PMC_IO_DPD_REQ]
pmc_io_dpd_skip:
dsb
ARM: convert all "mov.* pc, reg" to "bx reg" for ARMv6+ ARMv6 and greater introduced a new instruction ("bx") which can be used to return from function calls. Recent CPUs perform better when the "bx lr" instruction is used rather than the "mov pc, lr" instruction, and this sequence is strongly recommended to be used by the ARM architecture manual (section A.4.1.1). We provide a new macro "ret" with all its variants for the condition code which will resolve to the appropriate instruction. Rather than doing this piecemeal, and miss some instances, change all the "mov pc" instances to use the new macro, with the exception of the "movs" instruction and the kprobes code. This allows us to detect the "mov pc, lr" case and fix it up - and also gives us the possibility of deploying this for other registers depending on the CPU selection. Reported-by: Will Deacon <will.deacon@arm.com> Tested-by: Stephen Warren <swarren@nvidia.com> # Tegra Jetson TK1 Tested-by: Robert Jarzmik <robert.jarzmik@free.fr> # mioa701_bootresume.S Tested-by: Andrew Lunn <andrew@lunn.ch> # Kirkwood Tested-by: Shawn Guo <shawn.guo@freescale.com> Tested-by: Tony Lindgren <tony@atomide.com> # OMAPs Tested-by: Gregory CLEMENT <gregory.clement@free-electrons.com> # Armada XP, 375, 385 Acked-by: Sekhar Nori <nsekhar@ti.com> # DaVinci Acked-by: Christoffer Dall <christoffer.dall@linaro.org> # kvm/hyp Acked-by: Haojian Zhuang <haojian.zhuang@gmail.com> # PXA3xx Acked-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com> # Xen Tested-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de> # ARMv7M Tested-by: Simon Horman <horms+renesas@verge.net.au> # Shmobile Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
2014-06-30 09:29:12 -06:00
ret lr
ARM: tegra: add LP1 suspend support for Tegra30 The LP1 suspend mode will power off the CPU, clock gated the PLLs and put SDRAM to self-refresh mode. Any interrupt can wake up device from LP1. The sequence when LP1 suspending: * tunning off L1 data cache and the MMU * storing some EMC registers, DPD (deep power down) status, clk source of mselect and SCLK burst policy * putting SDRAM into self-refresh * switching CPU to CLK_M (12MHz OSC) * tunning off PLLM, PLLP, PLLA, PLLC and PLLX * switching SCLK to CLK_S (32KHz OSC) * shutting off the CPU rail The sequence of LP1 resuming: * re-enabling PLLM, PLLP, PLLA, PLLC and PLLX * restoring the clk source of mselect and SCLK burst policy * setting up CCLK burst policy to PLLX * restoring DPD status and some EMC registers * resuming SDRAM to normal mode * jumping to the "tegra_resume" from PMC_SCRATCH41 Due to the SDRAM will be put into self-refresh mode, the low level procedures of LP1 suspending and resuming should be copied to TEGRA_IRAM_CODE_AREA (TEGRA_IRAM_BASE + SZ_4K) when suspending. Before restoring the CPU context when resuming, the SDRAM needs to be switched back to normal mode. And the PLLs need to be re-enabled, SCLK burst policy be restored, CCLK burst policy be set in PLLX. Then jumping to "tegra_resume" that was expected to be stored in PMC_SCRATCH41 to restore CPU context and back to kernel. Based on the work by: Scott Williams <scwilliams@nvidia.com> Signed-off-by: Joseph Lo <josephl@nvidia.com> Signed-off-by: Stephen Warren <swarren@nvidia.com>
2013-08-12 03:40:04 -06:00
.ltorg
/* dummy symbol for end of IRAM */
.align L1_CACHE_SHIFT
.global tegra30_iram_end
tegra30_iram_end:
b .
#endif