u-boot-brain/arch/x86/cpu/ivybridge/model_206ax.c
Simon Glass 45b5a37836 x86: Add multi-processor init
Most modern x86 CPUs include more than one CPU core. The OS normally requires
that these 'Application Processors' (APs) be brought up by the boot loader.
Add the required support to U-Boot to init additional APs.

Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
2015-04-30 16:13:38 -06:00

515 lines
12 KiB
C

/*
* From Coreboot file of same name
*
* Copyright (C) 2007-2009 coresystems GmbH
* Copyright (C) 2011 The Chromium Authors
*
* SPDX-License-Identifier: GPL-2.0
*/
#include <common.h>
#include <fdtdec.h>
#include <malloc.h>
#include <asm/acpi.h>
#include <asm/cpu.h>
#include <asm/lapic.h>
#include <asm/lapic_def.h>
#include <asm/msr.h>
#include <asm/mtrr.h>
#include <asm/processor.h>
#include <asm/speedstep.h>
#include <asm/turbo.h>
#include <asm/arch/model_206ax.h>
static void enable_vmx(void)
{
struct cpuid_result regs;
#ifdef CONFIG_ENABLE_VMX
int enable = true;
#else
int enable = false;
#endif
msr_t msr;
regs = cpuid(1);
/* Check that the VMX is supported before reading or writing the MSR. */
if (!((regs.ecx & CPUID_VMX) || (regs.ecx & CPUID_SMX)))
return;
msr = msr_read(MSR_IA32_FEATURE_CONTROL);
if (msr.lo & (1 << 0)) {
debug("VMX is locked, so %s will do nothing\n", __func__);
/* VMX locked. If we set it again we get an illegal
* instruction
*/
return;
}
/* The IA32_FEATURE_CONTROL MSR may initialize with random values.
* It must be cleared regardless of VMX config setting.
*/
msr.hi = 0;
msr.lo = 0;
debug("%s VMX\n", enable ? "Enabling" : "Disabling");
/*
* Even though the Intel manual says you must set the lock bit in
* addition to the VMX bit in order for VMX to work, it is incorrect.
* Thus we leave it unlocked for the OS to manage things itself.
* This is good for a few reasons:
* - No need to reflash the bios just to toggle the lock bit.
* - The VMX bits really really should match each other across cores,
* so hard locking it on one while another has the opposite setting
* can easily lead to crashes as code using VMX migrates between
* them.
* - Vendors that want to "upsell" from a bios that disables+locks to
* one that doesn't is sleazy.
* By leaving this to the OS (e.g. Linux), people can do exactly what
* they want on the fly, and do it correctly (e.g. across multiple
* cores).
*/
if (enable) {
msr.lo |= (1 << 2);
if (regs.ecx & CPUID_SMX)
msr.lo |= (1 << 1);
}
msr_write(MSR_IA32_FEATURE_CONTROL, msr);
}
/* Convert time in seconds to POWER_LIMIT_1_TIME MSR value */
static const u8 power_limit_time_sec_to_msr[] = {
[0] = 0x00,
[1] = 0x0a,
[2] = 0x0b,
[3] = 0x4b,
[4] = 0x0c,
[5] = 0x2c,
[6] = 0x4c,
[7] = 0x6c,
[8] = 0x0d,
[10] = 0x2d,
[12] = 0x4d,
[14] = 0x6d,
[16] = 0x0e,
[20] = 0x2e,
[24] = 0x4e,
[28] = 0x6e,
[32] = 0x0f,
[40] = 0x2f,
[48] = 0x4f,
[56] = 0x6f,
[64] = 0x10,
[80] = 0x30,
[96] = 0x50,
[112] = 0x70,
[128] = 0x11,
};
/* Convert POWER_LIMIT_1_TIME MSR value to seconds */
static const u8 power_limit_time_msr_to_sec[] = {
[0x00] = 0,
[0x0a] = 1,
[0x0b] = 2,
[0x4b] = 3,
[0x0c] = 4,
[0x2c] = 5,
[0x4c] = 6,
[0x6c] = 7,
[0x0d] = 8,
[0x2d] = 10,
[0x4d] = 12,
[0x6d] = 14,
[0x0e] = 16,
[0x2e] = 20,
[0x4e] = 24,
[0x6e] = 28,
[0x0f] = 32,
[0x2f] = 40,
[0x4f] = 48,
[0x6f] = 56,
[0x10] = 64,
[0x30] = 80,
[0x50] = 96,
[0x70] = 112,
[0x11] = 128,
};
int cpu_config_tdp_levels(void)
{
struct cpuid_result result;
msr_t platform_info;
/* Minimum CPU revision */
result = cpuid(1);
if (result.eax < IVB_CONFIG_TDP_MIN_CPUID)
return 0;
/* Bits 34:33 indicate how many levels supported */
platform_info = msr_read(MSR_PLATFORM_INFO);
return (platform_info.hi >> 1) & 3;
}
/*
* Configure processor power limits if possible
* This must be done AFTER set of BIOS_RESET_CPL
*/
void set_power_limits(u8 power_limit_1_time)
{
msr_t msr = msr_read(MSR_PLATFORM_INFO);
msr_t limit;
unsigned power_unit;
unsigned tdp, min_power, max_power, max_time;
u8 power_limit_1_val;
if (power_limit_1_time > ARRAY_SIZE(power_limit_time_sec_to_msr))
return;
if (!(msr.lo & PLATFORM_INFO_SET_TDP))
return;
/* Get units */
msr = msr_read(MSR_PKG_POWER_SKU_UNIT);
power_unit = 2 << ((msr.lo & 0xf) - 1);
/* Get power defaults for this SKU */
msr = msr_read(MSR_PKG_POWER_SKU);
tdp = msr.lo & 0x7fff;
min_power = (msr.lo >> 16) & 0x7fff;
max_power = msr.hi & 0x7fff;
max_time = (msr.hi >> 16) & 0x7f;
debug("CPU TDP: %u Watts\n", tdp / power_unit);
if (power_limit_time_msr_to_sec[max_time] > power_limit_1_time)
power_limit_1_time = power_limit_time_msr_to_sec[max_time];
if (min_power > 0 && tdp < min_power)
tdp = min_power;
if (max_power > 0 && tdp > max_power)
tdp = max_power;
power_limit_1_val = power_limit_time_sec_to_msr[power_limit_1_time];
/* Set long term power limit to TDP */
limit.lo = 0;
limit.lo |= tdp & PKG_POWER_LIMIT_MASK;
limit.lo |= PKG_POWER_LIMIT_EN;
limit.lo |= (power_limit_1_val & PKG_POWER_LIMIT_TIME_MASK) <<
PKG_POWER_LIMIT_TIME_SHIFT;
/* Set short term power limit to 1.25 * TDP */
limit.hi = 0;
limit.hi |= ((tdp * 125) / 100) & PKG_POWER_LIMIT_MASK;
limit.hi |= PKG_POWER_LIMIT_EN;
/* Power limit 2 time is only programmable on SNB EP/EX */
msr_write(MSR_PKG_POWER_LIMIT, limit);
/* Use nominal TDP values for CPUs with configurable TDP */
if (cpu_config_tdp_levels()) {
msr = msr_read(MSR_CONFIG_TDP_NOMINAL);
limit.hi = 0;
limit.lo = msr.lo & 0xff;
msr_write(MSR_TURBO_ACTIVATION_RATIO, limit);
}
}
static void configure_c_states(void)
{
struct cpuid_result result;
msr_t msr;
msr = msr_read(MSR_PMG_CST_CONFIG_CTL);
msr.lo |= (1 << 28); /* C1 Auto Undemotion Enable */
msr.lo |= (1 << 27); /* C3 Auto Undemotion Enable */
msr.lo |= (1 << 26); /* C1 Auto Demotion Enable */
msr.lo |= (1 << 25); /* C3 Auto Demotion Enable */
msr.lo &= ~(1 << 10); /* Disable IO MWAIT redirection */
msr.lo |= 7; /* No package C-state limit */
msr_write(MSR_PMG_CST_CONFIG_CTL, msr);
msr = msr_read(MSR_PMG_IO_CAPTURE_ADR);
msr.lo &= ~0x7ffff;
msr.lo |= (PMB0_BASE + 4); /* LVL_2 base address */
msr.lo |= (2 << 16); /* CST Range: C7 is max C-state */
msr_write(MSR_PMG_IO_CAPTURE_ADR, msr);
msr = msr_read(MSR_MISC_PWR_MGMT);
msr.lo &= ~(1 << 0); /* Enable P-state HW_ALL coordination */
msr_write(MSR_MISC_PWR_MGMT, msr);
msr = msr_read(MSR_POWER_CTL);
msr.lo |= (1 << 18); /* Enable Energy Perf Bias MSR 0x1b0 */
msr.lo |= (1 << 1); /* C1E Enable */
msr.lo |= (1 << 0); /* Bi-directional PROCHOT# */
msr_write(MSR_POWER_CTL, msr);
/* C3 Interrupt Response Time Limit */
msr.hi = 0;
msr.lo = IRTL_VALID | IRTL_1024_NS | 0x50;
msr_write(MSR_PKGC3_IRTL, msr);
/* C6 Interrupt Response Time Limit */
msr.hi = 0;
msr.lo = IRTL_VALID | IRTL_1024_NS | 0x68;
msr_write(MSR_PKGC6_IRTL, msr);
/* C7 Interrupt Response Time Limit */
msr.hi = 0;
msr.lo = IRTL_VALID | IRTL_1024_NS | 0x6D;
msr_write(MSR_PKGC7_IRTL, msr);
/* Primary Plane Current Limit */
msr = msr_read(MSR_PP0_CURRENT_CONFIG);
msr.lo &= ~0x1fff;
msr.lo |= PP0_CURRENT_LIMIT;
msr_write(MSR_PP0_CURRENT_CONFIG, msr);
/* Secondary Plane Current Limit */
msr = msr_read(MSR_PP1_CURRENT_CONFIG);
msr.lo &= ~0x1fff;
result = cpuid(1);
if (result.eax >= 0x30600)
msr.lo |= PP1_CURRENT_LIMIT_IVB;
else
msr.lo |= PP1_CURRENT_LIMIT_SNB;
msr_write(MSR_PP1_CURRENT_CONFIG, msr);
}
static int configure_thermal_target(void)
{
int tcc_offset;
msr_t msr;
int node;
/* Find pointer to CPU configuration */
node = fdtdec_next_compatible(gd->fdt_blob, 0,
COMPAT_INTEL_MODEL_206AX);
if (node < 0)
return -ENOENT;
tcc_offset = fdtdec_get_int(gd->fdt_blob, node, "tcc-offset", 0);
/* Set TCC activaiton offset if supported */
msr = msr_read(MSR_PLATFORM_INFO);
if ((msr.lo & (1 << 30)) && tcc_offset) {
msr = msr_read(MSR_TEMPERATURE_TARGET);
msr.lo &= ~(0xf << 24); /* Bits 27:24 */
msr.lo |= (tcc_offset & 0xf) << 24;
msr_write(MSR_TEMPERATURE_TARGET, msr);
}
return 0;
}
static void configure_misc(void)
{
msr_t msr;
msr = msr_read(IA32_MISC_ENABLE);
msr.lo |= (1 << 0); /* Fast String enable */
msr.lo |= (1 << 3); /* TM1/TM2/EMTTM enable */
msr.lo |= (1 << 16); /* Enhanced SpeedStep Enable */
msr_write(IA32_MISC_ENABLE, msr);
/* Disable Thermal interrupts */
msr.lo = 0;
msr.hi = 0;
msr_write(IA32_THERM_INTERRUPT, msr);
/* Enable package critical interrupt only */
msr.lo = 1 << 4;
msr.hi = 0;
msr_write(IA32_PACKAGE_THERM_INTERRUPT, msr);
}
static void enable_lapic_tpr(void)
{
msr_t msr;
msr = msr_read(MSR_PIC_MSG_CONTROL);
msr.lo &= ~(1 << 10); /* Enable APIC TPR updates */
msr_write(MSR_PIC_MSG_CONTROL, msr);
}
static void configure_dca_cap(void)
{
struct cpuid_result cpuid_regs;
msr_t msr;
/* Check feature flag in CPUID.(EAX=1):ECX[18]==1 */
cpuid_regs = cpuid(1);
if (cpuid_regs.ecx & (1 << 18)) {
msr = msr_read(IA32_PLATFORM_DCA_CAP);
msr.lo |= 1;
msr_write(IA32_PLATFORM_DCA_CAP, msr);
}
}
static void set_max_ratio(void)
{
msr_t msr, perf_ctl;
perf_ctl.hi = 0;
/* Check for configurable TDP option */
if (cpu_config_tdp_levels()) {
/* Set to nominal TDP ratio */
msr = msr_read(MSR_CONFIG_TDP_NOMINAL);
perf_ctl.lo = (msr.lo & 0xff) << 8;
} else {
/* Platform Info bits 15:8 give max ratio */
msr = msr_read(MSR_PLATFORM_INFO);
perf_ctl.lo = msr.lo & 0xff00;
}
msr_write(IA32_PERF_CTL, perf_ctl);
debug("model_x06ax: frequency set to %d\n",
((perf_ctl.lo >> 8) & 0xff) * SANDYBRIDGE_BCLK);
}
static void set_energy_perf_bias(u8 policy)
{
msr_t msr;
/* Energy Policy is bits 3:0 */
msr = msr_read(IA32_ENERGY_PERFORMANCE_BIAS);
msr.lo &= ~0xf;
msr.lo |= policy & 0xf;
msr_write(IA32_ENERGY_PERFORMANCE_BIAS, msr);
debug("model_x06ax: energy policy set to %u\n", policy);
}
static void configure_mca(void)
{
msr_t msr;
int i;
msr.lo = 0;
msr.hi = 0;
/* This should only be done on a cold boot */
for (i = 0; i < 7; i++)
msr_write(IA32_MC0_STATUS + (i * 4), msr);
}
#if CONFIG_USBDEBUG
static unsigned ehci_debug_addr;
#endif
/*
* Initialize any extra cores/threads in this package.
*/
static int intel_cores_init(struct x86_cpu_priv *cpu)
{
struct cpuid_result result;
unsigned threads_per_package, threads_per_core, i;
/* Logical processors (threads) per core */
result = cpuid_ext(0xb, 0);
threads_per_core = result.ebx & 0xffff;
/* Logical processors (threads) per package */
result = cpuid_ext(0xb, 1);
threads_per_package = result.ebx & 0xffff;
debug("CPU: %u has %u cores, %u threads per core\n",
cpu->apic_id, threads_per_package / threads_per_core,
threads_per_core);
for (i = 1; i < threads_per_package; ++i) {
struct x86_cpu_priv *new_cpu;
new_cpu = calloc(1, sizeof(*new_cpu));
if (!new_cpu)
return -ENOMEM;
new_cpu->apic_id = cpu->apic_id + i;
/* Update APIC ID if no hyperthreading */
if (threads_per_core == 1)
new_cpu->apic_id <<= 1;
debug("CPU: %u has core %u\n", cpu->apic_id, new_cpu->apic_id);
#if 0 && CONFIG_SMP && CONFIG_MAX_CPUS > 1
/* TODO(sjg@chromium.org): Start the new cpu */
if (!start_cpu(new_cpu)) {
/* Record the error in cpu? */
printk(BIOS_ERR, "CPU %u would not start!\n",
new_cpu->apic_id);
new_cpu->start_err = 1;
}
#endif
}
return 0;
}
int model_206ax_init(struct x86_cpu_priv *cpu)
{
int ret;
/* Clear out pending MCEs */
configure_mca();
#if CONFIG_USBDEBUG
/* Is this caution really needed? */
if (!ehci_debug_addr)
ehci_debug_addr = get_ehci_debug();
set_ehci_debug(0);
#endif
/* Setup MTRRs based on physical address size */
#if 0 /* TODO: Implement this */
struct cpuid_result cpuid_regs;
cpuid_regs = cpuid(0x80000008);
x86_setup_fixed_mtrrs();
x86_setup_var_mtrrs(cpuid_regs.eax & 0xff, 2);
x86_mtrr_check();
#endif
#if CONFIG_USBDEBUG
set_ehci_debug(ehci_debug_addr);
#endif
/* Enable the local cpu apics */
enable_lapic_tpr();
lapic_setup();
/* Enable virtualization if enabled in CMOS */
enable_vmx();
/* Configure C States */
configure_c_states();
/* Configure Enhanced SpeedStep and Thermal Sensors */
configure_misc();
/* Thermal throttle activation offset */
ret = configure_thermal_target();
if (ret)
return ret;
/* Enable Direct Cache Access */
configure_dca_cap();
/* Set energy policy */
set_energy_perf_bias(ENERGY_POLICY_NORMAL);
/* Set Max Ratio */
set_max_ratio();
/* Enable Turbo */
turbo_enable();
/* Start up extra cores */
intel_cores_init(cpu);
return 0;
}