xref: /src/sys/arch/x86/x86/cpu.c

/*	$NetBSD: cpu.c,v 1.136 2017/09/28 17:48:20 maxv Exp $	*/

/*
 * Copyright (c) 2000-2012 NetBSD Foundation, Inc.
 * All rights reserved.
 *
 * This code is derived from software contributed to The NetBSD Foundation
 * by Bill Sommerfeld of RedBack Networks Inc, and by Andrew Doran.
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions
 * are met:
 * 1. Redistributions of source code must retain the above copyright
 *    notice, this list of conditions and the following disclaimer.
 * 2. Redistributions in binary form must reproduce the above copyright
 *    notice, this list of conditions and the following disclaimer in the
 *    documentation and/or other materials provided with the distribution.
 *
 * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS
 * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED
 * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
 * PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS
 * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
 * POSSIBILITY OF SUCH DAMAGE.
 */

/*
 * Copyright (c) 1999 Stefan Grefen
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions
 * are met:
 * 1. Redistributions of source code must retain the above copyright
 *    notice, this list of conditions and the following disclaimer.
 * 2. Redistributions in binary form must reproduce the above copyright
 *    notice, this list of conditions and the following disclaimer in the
 *    documentation and/or other materials provided with the distribution.
 * 3. All advertising materials mentioning features or use of this software
 *    must display the following acknowledgement:
 *      This product includes software developed by the NetBSD
 *      Foundation, Inc. and its contributors.
 * 4. Neither the name of The NetBSD Foundation nor the names of its
 *    contributors may be used to endorse or promote products derived
 *    from this software without specific prior written permission.
 *
 * THIS SOFTWARE IS PROVIDED BY AUTHOR AND CONTRIBUTORS ``AS IS'' AND ANY
 * EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
 * ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR AND CONTRIBUTORS BE LIABLE
 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
 * SUCH DAMAGE.
 */

#include <sys/cdefs.h>
__KERNEL_RCSID(0, "$NetBSD: cpu.c,v 1.136 2017/09/28 17:48:20 maxv Exp $");

#include "opt_ddb.h"
#include "opt_mpbios.h"		/* for MPDEBUG */
#include "opt_mtrr.h"
#include "opt_multiprocessor.h"

#include "lapic.h"
#include "ioapic.h"

#include <sys/param.h>
#include <sys/proc.h>
#include <sys/systm.h>
#include <sys/device.h>
#include <sys/kmem.h>
#include <sys/cpu.h>
#include <sys/cpufreq.h>
#include <sys/idle.h>
#include <sys/atomic.h>
#include <sys/reboot.h>

#include <uvm/uvm.h>

#include "acpica.h"		/* for NACPICA, for mp_verbose */

#include <machine/cpufunc.h>
#include <machine/cpuvar.h>
#include <machine/pmap.h>
#include <machine/vmparam.h>
#if defined(MULTIPROCESSOR)
#include <machine/mpbiosvar.h>
#endif
#include <machine/mpconfig.h>		/* for mp_verbose */
#include <machine/pcb.h>
#include <machine/specialreg.h>
#include <machine/segments.h>
#include <machine/gdt.h>
#include <machine/mtrr.h>
#include <machine/pio.h>
#include <machine/cpu_counter.h>

#include <x86/fpu.h>

#if NLAPIC > 0
#include <machine/apicvar.h>
#include <machine/i82489reg.h>
#include <machine/i82489var.h>
#endif

#include <dev/ic/mc146818reg.h>
#include <i386/isa/nvram.h>
#include <dev/isa/isareg.h>

#include "tsc.h"

static int	cpu_match(device_t, cfdata_t, void *);
static void	cpu_attach(device_t, device_t, void *);
static void	cpu_defer(device_t);
static int	cpu_rescan(device_t, const char *, const int *);
static void	cpu_childdetached(device_t, device_t);
static bool	cpu_stop(device_t);
static bool	cpu_suspend(device_t, const pmf_qual_t *);
static bool	cpu_resume(device_t, const pmf_qual_t *);
static bool	cpu_shutdown(device_t, int);

struct cpu_softc {
	device_t sc_dev;		/* device tree glue */
	struct cpu_info *sc_info;	/* pointer to CPU info */
	bool sc_wasonline;
};

#ifdef MULTIPROCESSOR
int mp_cpu_start(struct cpu_info *, paddr_t);
void mp_cpu_start_cleanup(struct cpu_info *);
const struct cpu_functions mp_cpu_funcs = { mp_cpu_start, NULL,
					    mp_cpu_start_cleanup };
#endif


CFATTACH_DECL2_NEW(cpu, sizeof(struct cpu_softc),
    cpu_match, cpu_attach, NULL, NULL, cpu_rescan, cpu_childdetached);

/*
 * Statically-allocated CPU info for the primary CPU (or the only
 * CPU, on uniprocessors).  The CPU info list is initialized to
 * point at it.
 */
struct cpu_info cpu_info_primary __aligned(CACHE_LINE_SIZE) = {
	.ci_dev = 0,
	.ci_self = &cpu_info_primary,
	.ci_idepth = -1,
	.ci_curlwp = &lwp0,
	.ci_curldt = -1,
};

struct cpu_info *cpu_info_list = &cpu_info_primary;

#ifdef i386
void		cpu_set_tss_gates(struct cpu_info *);
#endif

static void	cpu_init_idle_lwp(struct cpu_info *);

uint32_t cpu_feature[7] __read_mostly; /* X86 CPUID feature bits */
			/* [0] basic features cpuid.1:%edx
			 * [1] basic features cpuid.1:%ecx (CPUID2_xxx bits)
			 * [2] extended features cpuid:80000001:%edx
			 * [3] extended features cpuid:80000001:%ecx
			 * [4] VIA padlock features
			 * [5] structured extended features cpuid.7:%ebx
			 * [6] structured extended features cpuid.7:%ecx
			 */

#ifdef MULTIPROCESSOR
bool x86_mp_online;
paddr_t mp_trampoline_paddr = MP_TRAMPOLINE;
#endif
#if NLAPIC > 0
static vaddr_t cmos_data_mapping;
#endif
struct cpu_info *cpu_starting;

#ifdef MULTIPROCESSOR
void    	cpu_hatch(void *);
static void    	cpu_boot_secondary(struct cpu_info *ci);
static void    	cpu_start_secondary(struct cpu_info *ci);
#endif
#if NLAPIC > 0
static void	cpu_copy_trampoline(paddr_t);
#endif

/*
 * Runs once per boot once multiprocessor goo has been detected and
 * the local APIC on the boot processor has been mapped.
 *
 * Called from lapic_boot_init() (from mpbios_scan()).
 */
#if NLAPIC > 0
void
cpu_init_first(void)
{

	cpu_info_primary.ci_cpuid = lapic_cpu_number();

	cmos_data_mapping = uvm_km_alloc(kernel_map, PAGE_SIZE, 0, UVM_KMF_VAONLY);
	if (cmos_data_mapping == 0)
		panic("No KVA for page 0");
	pmap_kenter_pa(cmos_data_mapping, 0, VM_PROT_READ|VM_PROT_WRITE, 0);
	pmap_update(pmap_kernel());
}
#endif

static int
cpu_match(device_t parent, cfdata_t match, void *aux)
{

	return 1;
}

static void
cpu_vm_init(struct cpu_info *ci)
{
	int ncolors = 2, i;

	for (i = CAI_ICACHE; i <= CAI_L2CACHE; i++) {
		struct x86_cache_info *cai;
		int tcolors;

		cai = &ci->ci_cinfo[i];

		tcolors = atop(cai->cai_totalsize);
		switch(cai->cai_associativity) {
		case 0xff:
			tcolors = 1; /* fully associative */
			break;
		case 0:
		case 1:
			break;
		default:
			tcolors /= cai->cai_associativity;
		}
		ncolors = max(ncolors, tcolors);
		/*
		 * If the desired number of colors is not a power of
		 * two, it won't be good.  Find the greatest power of
		 * two which is an even divisor of the number of colors,
		 * to preserve even coloring of pages.
		 */
		if (ncolors & (ncolors - 1) ) {
			int try, picked = 1;
			for (try = 1; try < ncolors; try *= 2) {
				if (ncolors % try == 0) picked = try;
			}
			if (picked == 1) {
				panic("desired number of cache colors %d is "
			      	" > 1, but not even!", ncolors);
			}
			ncolors = picked;
		}
	}

	/*
	 * Knowing the size of the largest cache on this CPU, potentially
	 * re-color our pages.
	 */
	aprint_debug_dev(ci->ci_dev, "%d page colors\n", ncolors);
	uvm_page_recolor(ncolors);

	pmap_tlb_cpu_init(ci);
#ifndef __HAVE_DIRECT_MAP
	pmap_vpage_cpu_init(ci);
#endif
}

static void
cpu_attach(device_t parent, device_t self, void *aux)
{
	struct cpu_softc *sc = device_private(self);
	struct cpu_attach_args *caa = aux;
	struct cpu_info *ci;
	uintptr_t ptr;
#if NLAPIC > 0
	int cpunum = caa->cpu_number;
#endif
	static bool again;

	sc->sc_dev = self;

	if (ncpu == maxcpus) {
#ifndef _LP64
		aprint_error(": too many CPUs, please use NetBSD/amd64\n");
#else
		aprint_error(": too many CPUs\n");
#endif
		return;
	}

	/*
	 * If we're an Application Processor, allocate a cpu_info
	 * structure, otherwise use the primary's.
	 */
	if (caa->cpu_role == CPU_ROLE_AP) {
		if ((boothowto & RB_MD1) != 0) {
			aprint_error(": multiprocessor boot disabled\n");
			if (!pmf_device_register(self, NULL, NULL))
				aprint_error_dev(self,
				    "couldn't establish power handler\n");
			return;
		}
		aprint_naive(": Application Processor\n");
		ptr = (uintptr_t)kmem_zalloc(sizeof(*ci) + CACHE_LINE_SIZE - 1,
		    KM_SLEEP);
		ci = (struct cpu_info *)roundup2(ptr, CACHE_LINE_SIZE);
		ci->ci_curldt = -1;
	} else {
		aprint_naive(": %s Processor\n",
		    caa->cpu_role == CPU_ROLE_SP ? "Single" : "Boot");
		ci = &cpu_info_primary;
#if NLAPIC > 0
		if (cpunum != lapic_cpu_number()) {
			/* XXX should be done earlier. */
			uint32_t reg;
			aprint_verbose("\n");
			aprint_verbose_dev(self, "running CPU at apic %d"
			    " instead of at expected %d", lapic_cpu_number(),
			    cpunum);
			reg = lapic_readreg(LAPIC_ID);
			lapic_writereg(LAPIC_ID, (reg & ~LAPIC_ID_MASK) |
			    (cpunum << LAPIC_ID_SHIFT));
		}
		if (cpunum != lapic_cpu_number()) {
			aprint_error_dev(self, "unable to reset apic id\n");
		}
#endif
	}

	ci->ci_self = ci;
	sc->sc_info = ci;
	ci->ci_dev = self;
	ci->ci_acpiid = caa->cpu_id;
	ci->ci_cpuid = caa->cpu_number;
	ci->ci_func = caa->cpu_func;
	aprint_normal("\n");

	/* Must be before mi_cpu_attach(). */
	cpu_vm_init(ci);

	if (caa->cpu_role == CPU_ROLE_AP) {
		int error;

		error = mi_cpu_attach(ci);
		if (error != 0) {
			aprint_error_dev(self,
			    "mi_cpu_attach failed with %d\n", error);
			return;
		}
		cpu_init_tss(ci);
	} else {
		KASSERT(ci->ci_data.cpu_idlelwp != NULL);
	}

	pmap_reference(pmap_kernel());
	ci->ci_pmap = pmap_kernel();
	ci->ci_tlbstate = TLBSTATE_STALE;

	/*
	 * Boot processor may not be attached first, but the below
	 * must be done to allow booting other processors.
	 */
	if (!again) {
		atomic_or_32(&ci->ci_flags, CPUF_PRESENT | CPUF_PRIMARY);
		/* Basic init. */
		cpu_intr_init(ci);
		cpu_get_tsc_freq(ci);
		cpu_init(ci);
#ifdef i386
		cpu_set_tss_gates(ci);
#endif
		pmap_cpu_init_late(ci);
#if NLAPIC > 0
		if (caa->cpu_role != CPU_ROLE_SP) {
			/* Enable lapic. */
			lapic_enable();
			lapic_set_lvt();
			lapic_calibrate_timer(ci);
		}
#endif
		/* Make sure DELAY() is initialized. */
		DELAY(1);
		again = true;
	}

	/* further PCB init done later. */

	switch (caa->cpu_role) {
	case CPU_ROLE_SP:
		atomic_or_32(&ci->ci_flags, CPUF_SP);
		cpu_identify(ci);
		x86_errata();
		x86_cpu_idle_init();
		break;

	case CPU_ROLE_BP:
		atomic_or_32(&ci->ci_flags, CPUF_BSP);
		cpu_identify(ci);
		x86_errata();
		x86_cpu_idle_init();
		break;

#ifdef MULTIPROCESSOR
	case CPU_ROLE_AP:
		/*
		 * report on an AP
		 */
		cpu_intr_init(ci);
		gdt_alloc_cpu(ci);
#ifdef i386
		cpu_set_tss_gates(ci);
#endif
		pmap_cpu_init_late(ci);
		cpu_start_secondary(ci);
		if (ci->ci_flags & CPUF_PRESENT) {
			struct cpu_info *tmp;

			cpu_identify(ci);
			tmp = cpu_info_list;
			while (tmp->ci_next)
				tmp = tmp->ci_next;

			tmp->ci_next = ci;
		}
		break;
#endif

	default:
		panic("unknown processor type??\n");
	}

	pat_init(ci);

	if (!pmf_device_register1(self, cpu_suspend, cpu_resume, cpu_shutdown))
		aprint_error_dev(self, "couldn't establish power handler\n");

#ifdef MULTIPROCESSOR
	if (mp_verbose) {
		struct lwp *l = ci->ci_data.cpu_idlelwp;
		struct pcb *pcb = lwp_getpcb(l);

		aprint_verbose_dev(self,
		    "idle lwp at %p, idle sp at %p\n",
		    l,
#ifdef i386
		    (void *)pcb->pcb_esp
#else
		    (void *)pcb->pcb_rsp
#endif
		);
	}
#endif

	/*
	 * Postpone the "cpufeaturebus" scan.
	 * It is safe to scan the pseudo-bus
	 * only after all CPUs have attached.
	 */
	(void)config_defer(self, cpu_defer);
}

static void
cpu_defer(device_t self)
{
	cpu_rescan(self, NULL, NULL);
}

static int
cpu_rescan(device_t self, const char *ifattr, const int *locators)
{
	struct cpu_softc *sc = device_private(self);
	struct cpufeature_attach_args cfaa;
	struct cpu_info *ci = sc->sc_info;

	memset(&cfaa, 0, sizeof(cfaa));
	cfaa.ci = ci;

	if (ifattr_match(ifattr, "cpufeaturebus")) {
		if (ci->ci_frequency == NULL) {
			cfaa.name = "frequency";
			ci->ci_frequency = config_found_ia(self,
			    "cpufeaturebus", &cfaa, NULL);
		}

		if (ci->ci_padlock == NULL) {
			cfaa.name = "padlock";
			ci->ci_padlock = config_found_ia(self,
			    "cpufeaturebus", &cfaa, NULL);
		}

		if (ci->ci_temperature == NULL) {
			cfaa.name = "temperature";
			ci->ci_temperature = config_found_ia(self,
			    "cpufeaturebus", &cfaa, NULL);
		}

		if (ci->ci_vm == NULL) {
			cfaa.name = "vm";
			ci->ci_vm = config_found_ia(self,
			    "cpufeaturebus", &cfaa, NULL);
		}
	}

	return 0;
}

static void
cpu_childdetached(device_t self, device_t child)
{
	struct cpu_softc *sc = device_private(self);
	struct cpu_info *ci = sc->sc_info;

	if (ci->ci_frequency == child)
		ci->ci_frequency = NULL;

	if (ci->ci_padlock == child)
		ci->ci_padlock = NULL;

	if (ci->ci_temperature == child)
		ci->ci_temperature = NULL;

	if (ci->ci_vm == child)
		ci->ci_vm = NULL;
}

/*
 * Initialize the processor appropriately.
 */

void
cpu_init(struct cpu_info *ci)
{
	uint32_t cr4 = 0;

	lcr0(rcr0() | CR0_WP);

	/*
	 * On a P6 or above, enable global TLB caching if the
	 * hardware supports it.
	 */
	if (cpu_feature[0] & CPUID_PGE)
		cr4 |= CR4_PGE;	/* enable global TLB caching */

	/*
	 * If we have FXSAVE/FXRESTOR, use them.
	 */
	if (cpu_feature[0] & CPUID_FXSR) {
		cr4 |= CR4_OSFXSR;

		/*
		 * If we have SSE/SSE2, enable XMM exceptions.
		 */
		if (cpu_feature[0] & (CPUID_SSE|CPUID_SSE2))
			cr4 |= CR4_OSXMMEXCPT;
	}

	/* If xsave is supported, enable it */
	if (cpu_feature[1] & CPUID2_XSAVE)
		cr4 |= CR4_OSXSAVE;

	/* If SMEP is supported, enable it */
	if (cpu_feature[5] & CPUID_SEF_SMEP)
		cr4 |= CR4_SMEP;

	if (cr4) {
		cr4 |= rcr4();
		lcr4(cr4);
	}

	/* If xsave is enabled, enable all fpu features */
	if (cr4 & CR4_OSXSAVE)
		wrxcr(0, x86_xsave_features & XCR0_FPU);

#ifdef MTRR
	/*
	 * On a P6 or above, initialize MTRR's if the hardware supports them.
	 */
	if (cpu_feature[0] & CPUID_MTRR) {
		if ((ci->ci_flags & CPUF_AP) == 0)
			i686_mtrr_init_first();
		mtrr_init_cpu(ci);
	}

#ifdef i386
	if (strcmp((char *)(ci->ci_vendor), "AuthenticAMD") == 0) {
		/*
		 * Must be a K6-2 Step >= 7 or a K6-III.
		 */
		if (CPUID_TO_FAMILY(ci->ci_signature) == 5) {
			if (CPUID_TO_MODEL(ci->ci_signature) > 8 ||
			    (CPUID_TO_MODEL(ci->ci_signature) == 8 &&
			     CPUID_TO_STEPPING(ci->ci_signature) >= 7)) {
				mtrr_funcs = &k6_mtrr_funcs;
				k6_mtrr_init_first();
				mtrr_init_cpu(ci);
			}
		}
	}
#endif	/* i386 */
#endif /* MTRR */

	if (ci != &cpu_info_primary) {
		/* Synchronize TSC again, and check for drift. */
		wbinvd();
		atomic_or_32(&ci->ci_flags, CPUF_RUNNING);
		tsc_sync_ap(ci);
	} else {
		atomic_or_32(&ci->ci_flags, CPUF_RUNNING);
	}
}

#ifdef MULTIPROCESSOR
void
cpu_boot_secondary_processors(void)
{
	struct cpu_info *ci;
	kcpuset_t *cpus;
	u_long i;

	/* Now that we know the number of CPUs, patch the text segment. */
	x86_patch(false);

	kcpuset_create(&cpus, true);
	kcpuset_set(cpus, cpu_index(curcpu()));
	for (i = 0; i < maxcpus; i++) {
		ci = cpu_lookup(i);
		if (ci == NULL)
			continue;
		if (ci->ci_data.cpu_idlelwp == NULL)
			continue;
		if ((ci->ci_flags & CPUF_PRESENT) == 0)
			continue;
		if (ci->ci_flags & (CPUF_BSP|CPUF_SP|CPUF_PRIMARY))
			continue;
		cpu_boot_secondary(ci);
		kcpuset_set(cpus, cpu_index(ci));
	}
	while (!kcpuset_match(cpus, kcpuset_running))
		;
	kcpuset_destroy(cpus);

	x86_mp_online = true;

	/* Now that we know about the TSC, attach the timecounter. */
	tsc_tc_init();

	/* Enable zeroing of pages in the idle loop if we have SSE2. */
	vm_page_zero_enable = ((cpu_feature[0] & CPUID_SSE2) != 0);
}
#endif

static void
cpu_init_idle_lwp(struct cpu_info *ci)
{
	struct lwp *l = ci->ci_data.cpu_idlelwp;
	struct pcb *pcb = lwp_getpcb(l);

	pcb->pcb_cr0 = rcr0();
}

void
cpu_init_idle_lwps(void)
{
	struct cpu_info *ci;
	u_long i;

	for (i = 0; i < maxcpus; i++) {
		ci = cpu_lookup(i);
		if (ci == NULL)
			continue;
		if (ci->ci_data.cpu_idlelwp == NULL)
			continue;
		if ((ci->ci_flags & CPUF_PRESENT) == 0)
			continue;
		cpu_init_idle_lwp(ci);
	}
}

#ifdef MULTIPROCESSOR
void
cpu_start_secondary(struct cpu_info *ci)
{
	paddr_t mp_pdirpa;
	u_long psl;
	int i;

	mp_pdirpa = pmap_init_tmp_pgtbl(mp_trampoline_paddr);
	cpu_copy_trampoline(mp_pdirpa);

	atomic_or_32(&ci->ci_flags, CPUF_AP);
	ci->ci_curlwp = ci->ci_data.cpu_idlelwp;
	if (CPU_STARTUP(ci, mp_trampoline_paddr) != 0) {
		return;
	}

	/*
	 * Wait for it to become ready.   Setting cpu_starting opens the
	 * initial gate and allows the AP to start soft initialization.
	 */
	KASSERT(cpu_starting == NULL);
	cpu_starting = ci;
	for (i = 100000; (!(ci->ci_flags & CPUF_PRESENT)) && i > 0; i--) {
		i8254_delay(10);
	}

	if ((ci->ci_flags & CPUF_PRESENT) == 0) {
		aprint_error_dev(ci->ci_dev, "failed to become ready\n");
#if defined(MPDEBUG) && defined(DDB)
		printf("dropping into debugger; continue from here to resume boot\n");
		Debugger();
#endif
	} else {
		/*
		 * Synchronize time stamp counters. Invalidate cache and do
		 * twice to try and minimize possible cache effects. Disable
		 * interrupts to try and rule out any external interference.
		 */
		psl = x86_read_psl();
		x86_disable_intr();
		wbinvd();
		tsc_sync_bp(ci);
		x86_write_psl(psl);
	}

	CPU_START_CLEANUP(ci);
	cpu_starting = NULL;
}

void
cpu_boot_secondary(struct cpu_info *ci)
{
	int64_t drift;
	u_long psl;
	int i;

	atomic_or_32(&ci->ci_flags, CPUF_GO);
	for (i = 100000; (!(ci->ci_flags & CPUF_RUNNING)) && i > 0; i--) {
		i8254_delay(10);
	}
	if ((ci->ci_flags & CPUF_RUNNING) == 0) {
		aprint_error_dev(ci->ci_dev, "failed to start\n");
#if defined(MPDEBUG) && defined(DDB)
		printf("dropping into debugger; continue from here to resume boot\n");
		Debugger();
#endif
	} else {
		/* Synchronize TSC again, check for drift. */
		drift = ci->ci_data.cpu_cc_skew;
		psl = x86_read_psl();
		x86_disable_intr();
		wbinvd();
		tsc_sync_bp(ci);
		x86_write_psl(psl);
		drift -= ci->ci_data.cpu_cc_skew;
		aprint_debug_dev(ci->ci_dev, "TSC skew=%lld drift=%lld\n",
		    (long long)ci->ci_data.cpu_cc_skew, (long long)drift);
		tsc_sync_drift(drift);
	}
}

/*
 * The CPU ends up here when it's ready to run.
 * This is called from code in mptramp.s; at this point, we are running
 * in the idle pcb/idle stack of the new CPU.  When this function returns,
 * this processor will enter the idle loop and start looking for work.
 */
void
cpu_hatch(void *v)
{
	struct cpu_info *ci = (struct cpu_info *)v;
	struct pcb *pcb;
	int s, i;

	cpu_init_msrs(ci, true);
	cpu_probe(ci);

	ci->ci_data.cpu_cc_freq = cpu_info_primary.ci_data.cpu_cc_freq;
	/* cpu_get_tsc_freq(ci); */

	KDASSERT((ci->ci_flags & CPUF_PRESENT) == 0);

	/*
	 * Synchronize time stamp counters.  Invalidate cache and do twice
	 * to try and minimize possible cache effects.  Note that interrupts
	 * are off at this point.
	 */
	wbinvd();
	atomic_or_32(&ci->ci_flags, CPUF_PRESENT);
	tsc_sync_ap(ci);

	/*
	 * Wait to be brought online.  Use 'monitor/mwait' if available,
	 * in order to make the TSC drift as much as possible. so that
	 * we can detect it later.  If not available, try 'pause'.
	 * We'd like to use 'hlt', but we have interrupts off.
	 */
	while ((ci->ci_flags & CPUF_GO) == 0) {
		if ((cpu_feature[1] & CPUID2_MONITOR) != 0) {
			x86_monitor(&ci->ci_flags, 0, 0);
			if ((ci->ci_flags & CPUF_GO) != 0) {
				continue;
			}
			x86_mwait(0, 0);
		} else {
	/*
	 * XXX The loop repetition count could be a lot higher, but
	 * XXX currently qemu emulator takes a _very_long_time_ to
	 * XXX execute the pause instruction.  So for now, use a low
	 * XXX value to allow the cpu to hatch before timing out.
	 */
			for (i = 50; i != 0; i--) {
				x86_pause();
			}
		}
	}

	/* Because the text may have been patched in x86_patch(). */
	wbinvd();
	x86_flush();
	tlbflushg();

	KASSERT((ci->ci_flags & CPUF_RUNNING) == 0);

#ifdef PAE
	pd_entry_t * l3_pd = ci->ci_pae_l3_pdir;
	for (i = 0 ; i < PDP_SIZE; i++) {
		l3_pd[i] = pmap_kernel()->pm_pdirpa[i] | PG_V;
	}
	lcr3(ci->ci_pae_l3_pdirpa);
#else
	lcr3(pmap_pdirpa(pmap_kernel(), 0));
#endif

	pcb = lwp_getpcb(curlwp);
	pcb->pcb_cr3 = rcr3();
	pcb = lwp_getpcb(ci->ci_data.cpu_idlelwp);
	lcr0(pcb->pcb_cr0);

	cpu_init_idt();
	gdt_init_cpu(ci);
#if NLAPIC > 0
	lapic_enable();
	lapic_set_lvt();
	lapic_initclocks();
#endif

	fpuinit(ci);
	lldt(GSYSSEL(GLDT_SEL, SEL_KPL));
	ltr(ci->ci_tss_sel);

	cpu_init(ci);
	cpu_get_tsc_freq(ci);

	s = splhigh();
	lapic_write_tpri(0);
	x86_enable_intr();
	splx(s);
	x86_errata();

	aprint_debug_dev(ci->ci_dev, "running\n");

	idle_loop(NULL);
	KASSERT(false);
}
#endif

#if defined(DDB)

#include <ddb/db_output.h>
#include <machine/db_machdep.h>

/*
 * Dump CPU information from ddb.
 */
void
cpu_debug_dump(void)
{
	struct cpu_info *ci;
	CPU_INFO_ITERATOR cii;

	db_printf("addr		dev	id	flags	ipis	curlwp 		fpcurlwp\n");
	for (CPU_INFO_FOREACH(cii, ci)) {
		db_printf("%p	%s	%ld	%x	%x	%10p	%10p\n",
		    ci,
		    ci->ci_dev == NULL ? "BOOT" : device_xname(ci->ci_dev),
		    (long)ci->ci_cpuid,
		    ci->ci_flags, ci->ci_ipis,
		    ci->ci_curlwp,
		    ci->ci_fpcurlwp);
	}
}
#endif

#if NLAPIC > 0
static void
cpu_copy_trampoline(paddr_t pdir_pa)
{
	extern uint32_t nox_flag;
	extern u_char cpu_spinup_trampoline[];
	extern u_char cpu_spinup_trampoline_end[];
	vaddr_t mp_trampoline_vaddr;
	struct {
		uint32_t large;
		uint32_t nox;
		uint32_t pdir;
	} smp_data;
	CTASSERT(sizeof(smp_data) == 3 * 4);

	smp_data.large = (pmap_largepages != 0);
	smp_data.nox = nox_flag;
	smp_data.pdir = (uint32_t)(pdir_pa & 0xFFFFFFFF);

	/* Enter the physical address */
	mp_trampoline_vaddr = uvm_km_alloc(kernel_map, PAGE_SIZE, 0,
	    UVM_KMF_VAONLY);
	pmap_kenter_pa(mp_trampoline_vaddr, mp_trampoline_paddr,
	    VM_PROT_READ | VM_PROT_WRITE, 0);
	pmap_update(pmap_kernel());

	/* Copy boot code */
	memcpy((void *)mp_trampoline_vaddr,
	    cpu_spinup_trampoline,
	    cpu_spinup_trampoline_end - cpu_spinup_trampoline);

	/* Copy smp_data at the end */
	memcpy((void *)(mp_trampoline_vaddr + PAGE_SIZE - sizeof(smp_data)),
	    &smp_data, sizeof(smp_data));

	pmap_kremove(mp_trampoline_vaddr, PAGE_SIZE);
	pmap_update(pmap_kernel());
	uvm_km_free(kernel_map, mp_trampoline_vaddr, PAGE_SIZE, UVM_KMF_VAONLY);
}
#endif

#ifdef MULTIPROCESSOR
int
mp_cpu_start(struct cpu_info *ci, paddr_t target)
{
	unsigned short dwordptr[2];
	int error;

	/*
	 * Bootstrap code must be addressable in real mode
	 * and it must be page aligned.
	 */
	KASSERT(target < 0x10000 && target % PAGE_SIZE == 0);

	/*
	 * "The BSP must initialize CMOS shutdown code to 0Ah ..."
	 */

	outb(IO_RTC, NVRAM_RESET);
	outb(IO_RTC+1, NVRAM_RESET_JUMP);

	/*
	 * "and the warm reset vector (DWORD based at 40:67) to point
	 * to the AP startup code ..."
	 */

	dwordptr[0] = 0;
	dwordptr[1] = target >> 4;

#if NLAPIC > 0
	memcpy((uint8_t *)cmos_data_mapping + 0x467, dwordptr, 4);
#endif

	if ((cpu_feature[0] & CPUID_APIC) == 0) {
		aprint_error("mp_cpu_start: CPU does not have APIC\n");
		return ENODEV;
	}

	/*
	 * ... prior to executing the following sequence:".  We'll also add in
	 * local cache flush, in case the BIOS has left the AP with its cache
	 * disabled.  It may not be able to cope with MP coherency.
	 */
	wbinvd();

	if (ci->ci_flags & CPUF_AP) {
		error = x86_ipi_init(ci->ci_cpuid);
		if (error != 0) {
			aprint_error_dev(ci->ci_dev, "%s: IPI not taken (1)\n",
			    __func__);
			return error;
		}
		i8254_delay(10000);

		error = x86_ipi_startup(ci->ci_cpuid, target / PAGE_SIZE);
		if (error != 0) {
			aprint_error_dev(ci->ci_dev, "%s: IPI not taken (2)\n",
			    __func__);
			return error;
		}
		i8254_delay(200);

		error = x86_ipi_startup(ci->ci_cpuid, target / PAGE_SIZE);
		if (error != 0) {
			aprint_error_dev(ci->ci_dev, "%s: IPI not taken (3)\n",
			    __func__);
			return error;
		}
		i8254_delay(200);
	}

	return 0;
}

void
mp_cpu_start_cleanup(struct cpu_info *ci)
{
	/*
	 * Ensure the NVRAM reset byte contains something vaguely sane.
	 */

	outb(IO_RTC, NVRAM_RESET);
	outb(IO_RTC+1, NVRAM_RESET_RST);
}
#endif

#ifdef __x86_64__
typedef void (vector)(void);
extern vector Xsyscall, Xsyscall32;
#endif

void
cpu_init_msrs(struct cpu_info *ci, bool full)
{
#ifdef __x86_64__
	wrmsr(MSR_STAR,
	    ((uint64_t)GSEL(GCODE_SEL, SEL_KPL) << 32) |
	    ((uint64_t)LSEL(LSYSRETBASE_SEL, SEL_UPL) << 48));
	wrmsr(MSR_LSTAR, (uint64_t)Xsyscall);
	wrmsr(MSR_CSTAR, (uint64_t)Xsyscall32);
	wrmsr(MSR_SFMASK, PSL_NT|PSL_T|PSL_I|PSL_C);

	if (full) {
		wrmsr(MSR_FSBASE, 0);
		wrmsr(MSR_GSBASE, (uint64_t)ci);
		wrmsr(MSR_KERNELGSBASE, 0);
	}
#endif	/* __x86_64__ */

	if (cpu_feature[2] & CPUID_NOX)
		wrmsr(MSR_EFER, rdmsr(MSR_EFER) | EFER_NXE);
}

void
cpu_offline_md(void)
{
	int s;

	s = splhigh();
	fpusave_cpu(true);
	splx(s);
}

/* XXX joerg restructure and restart CPUs individually */
static bool
cpu_stop(device_t dv)
{
	struct cpu_softc *sc = device_private(dv);
	struct cpu_info *ci = sc->sc_info;
	int err;

	KASSERT((ci->ci_flags & CPUF_PRESENT) != 0);

	if ((ci->ci_flags & CPUF_PRIMARY) != 0)
		return true;

	if (ci->ci_data.cpu_idlelwp == NULL)
		return true;

	sc->sc_wasonline = !(ci->ci_schedstate.spc_flags & SPCF_OFFLINE);

	if (sc->sc_wasonline) {
		mutex_enter(&cpu_lock);
		err = cpu_setstate(ci, false);
		mutex_exit(&cpu_lock);

		if (err != 0)
			return false;
	}

	return true;
}

static bool
cpu_suspend(device_t dv, const pmf_qual_t *qual)
{
	struct cpu_softc *sc = device_private(dv);
	struct cpu_info *ci = sc->sc_info;

	if ((ci->ci_flags & CPUF_PRESENT) == 0)
		return true;
	else {
		cpufreq_suspend(ci);
	}

	return cpu_stop(dv);
}

static bool
cpu_resume(device_t dv, const pmf_qual_t *qual)
{
	struct cpu_softc *sc = device_private(dv);
	struct cpu_info *ci = sc->sc_info;
	int err = 0;

	if ((ci->ci_flags & CPUF_PRESENT) == 0)
		return true;

	if ((ci->ci_flags & CPUF_PRIMARY) != 0)
		goto out;

	if (ci->ci_data.cpu_idlelwp == NULL)
		goto out;

	if (sc->sc_wasonline) {
		mutex_enter(&cpu_lock);
		err = cpu_setstate(ci, true);
		mutex_exit(&cpu_lock);
	}

out:
	if (err != 0)
		return false;

	cpufreq_resume(ci);

	return true;
}

static bool
cpu_shutdown(device_t dv, int how)
{
	struct cpu_softc *sc = device_private(dv);
	struct cpu_info *ci = sc->sc_info;

	if ((ci->ci_flags & CPUF_BSP) != 0)
		return false;

	if ((ci->ci_flags & CPUF_PRESENT) == 0)
		return true;

	return cpu_stop(dv);
}

void
cpu_get_tsc_freq(struct cpu_info *ci)
{
	uint64_t last_tsc;

	if (cpu_hascounter()) {
		last_tsc = cpu_counter_serializing();
		i8254_delay(100000);
		ci->ci_data.cpu_cc_freq =
		    (cpu_counter_serializing() - last_tsc) * 10;
	}
}

void
x86_cpu_idle_mwait(void)
{
	struct cpu_info *ci = curcpu();

	KASSERT(ci->ci_ilevel == IPL_NONE);

	x86_monitor(&ci->ci_want_resched, 0, 0);
	if (__predict_false(ci->ci_want_resched)) {
		return;
	}
	x86_mwait(0, 0);
}

void
x86_cpu_idle_halt(void)
{
	struct cpu_info *ci = curcpu();

	KASSERT(ci->ci_ilevel == IPL_NONE);

	x86_disable_intr();
	if (!__predict_false(ci->ci_want_resched)) {
		x86_stihlt();
	} else {
		x86_enable_intr();
	}
}

/*
 * Loads pmap for the current CPU.
 */
void
cpu_load_pmap(struct pmap *pmap, struct pmap *oldpmap)
{
#ifdef PAE
	struct cpu_info *ci = curcpu();
	bool interrupts_enabled;
	pd_entry_t *l3_pd = ci->ci_pae_l3_pdir;
	int i;

	/*
	 * disable interrupts to block TLB shootdowns, which can reload cr3.
	 * while this doesn't block NMIs, it's probably ok as NMIs unlikely
	 * reload cr3.
	 */
	interrupts_enabled = (x86_read_flags() & PSL_I) != 0;
	if (interrupts_enabled)
		x86_disable_intr();

	for (i = 0 ; i < PDP_SIZE; i++) {
		l3_pd[i] = pmap->pm_pdirpa[i] | PG_V;
	}

	if (interrupts_enabled)
		x86_enable_intr();
	tlbflush();
#else /* PAE */
	lcr3(pmap_pdirpa(pmap, 0));
#endif /* PAE */
}

/*
 * Notify all other cpus to halt.
 */

void
cpu_broadcast_halt(void)
{
	x86_broadcast_ipi(X86_IPI_HALT);
}

/*
 * Send a dummy ipi to a cpu to force it to run splraise()/spllower()
 */

void
cpu_kick(struct cpu_info *ci)
{
	x86_send_ipi(ci, 0);
}