traps.c

xen虚拟机源代码安装包

#include <xen/config.h>
#include <xen/version.h>
#include <xen/domain_page.h>
#include <xen/init.h>
#include <xen/sched.h>
#include <xen/lib.h>
#include <xen/console.h>
#include <xen/mm.h>
#include <xen/irq.h>
#include <xen/symbols.h>
#include <xen/shutdown.h>
#include <xen/nmi.h>
#include <asm/current.h>
#include <asm/flushtlb.h>
#include <asm/hvm/hvm.h>
#include <asm/hvm/support.h>

#include <public/callback.h>

static void print_xen_info(void)
{
    char taint_str[TAINT_STRING_MAX_LEN];
    char debug = 'n', *arch = "x86_32p";

#ifndef NDEBUG
    debug = 'y';
#endif

    printk("----[ Xen-%d.%d%s  %s  debug=%c  %s ]----\n",
           xen_major_version(), xen_minor_version(), xen_extra_version(),
           arch, debug, print_tainted(taint_str));
}

enum context { CTXT_hypervisor, CTXT_pv_guest, CTXT_hvm_guest };

static void _show_registers(
    const struct cpu_user_regs *regs, unsigned long crs[8],
    enum context context, const struct vcpu *v)
{
    const static char *context_names[] = {
        [CTXT_hypervisor] = "hypervisor",
        [CTXT_pv_guest]   = "pv guest",
        [CTXT_hvm_guest]  = "hvm guest"
    };

    printk("EIP:    %04x:[<%08x>]", regs->cs, regs->eip);
    if ( context == CTXT_hypervisor )
        print_symbol(" %s", regs->eip);
    printk("\nEFLAGS: %08x   ", regs->eflags);
    if ( (context == CTXT_pv_guest) && v && v->vcpu_info )
        printk("EM: %d   ", !!v->vcpu_info->evtchn_upcall_mask);
    printk("CONTEXT: %s\n", context_names[context]);

    printk("eax: %08x   ebx: %08x   ecx: %08x   edx: %08x\n",
           regs->eax, regs->ebx, regs->ecx, regs->edx);
    printk("esi: %08x   edi: %08x   ebp: %08x   esp: %08x\n",
           regs->esi, regs->edi, regs->ebp, regs->esp);
    printk("cr0: %08lx   cr4: %08lx   cr3: %08lx   cr2: %08lx\n",
           crs[0], crs[4], crs[3], crs[2]);
    printk("ds: %04x   es: %04x   fs: %04x   gs: %04x   "
           "ss: %04x   cs: %04x\n",
           regs->ds, regs->es, regs->fs,
           regs->gs, regs->ss, regs->cs);
}

void show_registers(struct cpu_user_regs *regs)
{
    struct cpu_user_regs fault_regs = *regs;
    unsigned long fault_crs[8];
    enum context context;
    struct vcpu *v = current;

    if ( is_hvm_vcpu(v) && guest_mode(regs) )
    {
        struct segment_register sreg;
        context = CTXT_hvm_guest;
        fault_crs[0] = v->arch.hvm_vcpu.guest_cr[0];
        fault_crs[2] = v->arch.hvm_vcpu.guest_cr[2];
        fault_crs[3] = v->arch.hvm_vcpu.guest_cr[3];
        fault_crs[4] = v->arch.hvm_vcpu.guest_cr[4];
        hvm_get_segment_register(v, x86_seg_cs, &sreg);
        fault_regs.cs = sreg.sel;
        hvm_get_segment_register(v, x86_seg_ds, &sreg);
        fault_regs.ds = sreg.sel;
        hvm_get_segment_register(v, x86_seg_es, &sreg);
        fault_regs.es = sreg.sel;
        hvm_get_segment_register(v, x86_seg_fs, &sreg);
        fault_regs.fs = sreg.sel;
        hvm_get_segment_register(v, x86_seg_gs, &sreg);
        fault_regs.gs = sreg.sel;
        hvm_get_segment_register(v, x86_seg_ss, &sreg);
        fault_regs.ss = sreg.sel;
    }
    else
    {
        if ( !guest_mode(regs) )
        {
            context = CTXT_hypervisor;
            fault_regs.esp = (unsigned long)&regs->esp;
            fault_regs.ss = read_segment_register(ss);
            fault_regs.ds = read_segment_register(ds);
            fault_regs.es = read_segment_register(es);
            fault_regs.fs = read_segment_register(fs);
            fault_regs.gs = read_segment_register(gs);
            fault_crs[2] = read_cr2();
        }
        else
        {
            context = CTXT_pv_guest;
            fault_crs[2] = v->vcpu_info->arch.cr2;
        }

        fault_crs[0] = read_cr0();
        fault_crs[3] = read_cr3();
        fault_crs[4] = read_cr4();
    }

    print_xen_info();
    printk("CPU:    %d\n", smp_processor_id());
    _show_registers(&fault_regs, fault_crs, context, v);

    if ( this_cpu(ler_msr) && !guest_mode(regs) )
    {
        u32 from, to, hi;
        rdmsr(this_cpu(ler_msr), from, hi);
        rdmsr(this_cpu(ler_msr) + 1, to, hi);
        printk("ler: %08x -> %08x\n", from, to);
    }
}

void vcpu_show_registers(const struct vcpu *v)
{
    unsigned long crs[8];

    /* No need to handle HVM for now. */
    if ( is_hvm_vcpu(v) )
        return;

    crs[0] = v->arch.guest_context.ctrlreg[0];
    crs[2] = v->vcpu_info->arch.cr2;
    crs[3] = pagetable_get_paddr(v->arch.guest_table);
    crs[4] = v->arch.guest_context.ctrlreg[4];

    _show_registers(&v->arch.guest_context.user_regs, crs, CTXT_pv_guest, v);
}

void show_page_walk(unsigned long addr)
{
    unsigned long pfn, mfn, cr3 = read_cr3();
    l3_pgentry_t l3e, *l3t;
    l2_pgentry_t l2e, *l2t;
    l1_pgentry_t l1e, *l1t;

    printk("Pagetable walk from %08lx:\n", addr);

    mfn = cr3 >> PAGE_SHIFT;

    l3t  = map_domain_page(mfn);
    l3t += (cr3 & 0xFE0UL) >> 3;
    l3e = l3t[l3_table_offset(addr)];
    mfn = l3e_get_pfn(l3e);
    pfn = mfn_valid(mfn) ? get_gpfn_from_mfn(mfn) : INVALID_M2P_ENTRY;
    printk(" L3[0x%03lx] = %"PRIpte" %08lx\n",
           l3_table_offset(addr), l3e_get_intpte(l3e), pfn);
    unmap_domain_page(l3t);
    if ( !(l3e_get_flags(l3e) & _PAGE_PRESENT) )
        return;

    l2t = map_domain_page(mfn);
    l2e = l2t[l2_table_offset(addr)];
    mfn = l2e_get_pfn(l2e);
    pfn = mfn_valid(mfn) ? get_gpfn_from_mfn(mfn) : INVALID_M2P_ENTRY;
    printk(" L2[0x%03lx] = %"PRIpte" %08lx %s\n",
           l2_table_offset(addr), l2e_get_intpte(l2e), pfn,
           (l2e_get_flags(l2e) & _PAGE_PSE) ? "(PSE)" : "");
    unmap_domain_page(l2t);
    if ( !(l2e_get_flags(l2e) & _PAGE_PRESENT) ||
         (l2e_get_flags(l2e) & _PAGE_PSE) )
        return;

    l1t = map_domain_page(mfn);
    l1e = l1t[l1_table_offset(addr)];
    mfn = l1e_get_pfn(l1e);
    pfn = mfn_valid(mfn) ? get_gpfn_from_mfn(mfn) : INVALID_M2P_ENTRY;
    printk(" L1[0x%03lx] = %"PRIpte" %08lx\n",
           l1_table_offset(addr), l1e_get_intpte(l1e), pfn);
    unmap_domain_page(l1t);
}

#define DOUBLEFAULT_STACK_SIZE 2048
static struct tss_struct doublefault_tss;
static unsigned char doublefault_stack[DOUBLEFAULT_STACK_SIZE];

asmlinkage void do_double_fault(void)
{
    struct tss_struct *tss = &doublefault_tss;
    unsigned int cpu = ((tss->back_link>>3)-__FIRST_TSS_ENTRY)>>1;

    watchdog_disable();

    console_force_unlock();

    /* Find information saved during fault and dump it to the console. */
    tss = &init_tss[cpu];
    printk("*** DOUBLE FAULT ***\n");
    print_xen_info();
    printk("CPU:    %d\nEIP:    %04x:[<%08x>]",
           cpu, tss->cs, tss->eip);
    print_symbol(" %s\n", tss->eip);
    printk("EFLAGS: %08x\n", tss->eflags);
    printk("CR3:    %08x\n", tss->__cr3);
    printk("eax: %08x   ebx: %08x   ecx: %08x   edx: %08x\n",
           tss->eax, tss->ebx, tss->ecx, tss->edx);
    printk("esi: %08x   edi: %08x   ebp: %08x   esp: %08x\n",
           tss->esi, tss->edi, tss->ebp, tss->esp);
    printk("ds: %04x   es: %04x   fs: %04x   gs: %04x   ss: %04x\n",
           tss->ds, tss->es, tss->fs, tss->gs, tss->ss);
    show_stack_overflow(cpu, tss->esp);

    panic("DOUBLE FAULT -- system shutdown\n");
}

unsigned long do_iret(void)
{
    struct cpu_user_regs *regs = guest_cpu_user_regs();
    struct vcpu *v = current;
    u32 eflags;

    /* Check worst-case stack frame for overlap with Xen protected area. */
    if ( unlikely(!access_ok(regs->esp, 40)) )
        goto exit_and_crash;

    /* Pop and restore EAX (clobbered by hypercall). */
    if ( unlikely(__copy_from_user(&regs->eax, (void *)regs->esp, 4)) )
        goto exit_and_crash;
    regs->esp += 4;

    /* Pop and restore CS and EIP. */
    if ( unlikely(__copy_from_user(&regs->eip, (void *)regs->esp, 8)) )
        goto exit_and_crash;
    regs->esp += 8;

    /*
     * Pop, fix up and restore EFLAGS. We fix up in a local staging area
     * to avoid firing the BUG_ON(IOPL) check in arch_get_info_guest.
     */
    if ( unlikely(__copy_from_user(&eflags, (void *)regs->esp, 4)) )
        goto exit_and_crash;
    regs->esp += 4;
    regs->eflags = (eflags & ~X86_EFLAGS_IOPL) | X86_EFLAGS_IF;

    if ( vm86_mode(regs) )
    {
        /* Return to VM86 mode: pop and restore ESP,SS,ES,DS,FS and GS. */
        if ( __copy_from_user(&regs->esp, (void *)regs->esp, 24) )
            goto exit_and_crash;
    }
    else if ( unlikely(ring_0(regs)) )
    {
        goto exit_and_crash;
    }
    else if ( !ring_1(regs) )
    {
        /* Return to ring 2/3: pop and restore ESP and SS. */
        if ( __copy_from_user(&regs->esp, (void *)regs->esp, 8) )
            goto exit_and_crash;
    }

    /* Restore affinity.  */
    if ((v->trap_priority >= VCPU_TRAP_NMI)
       && !cpus_equal(v->cpu_affinity_tmp, v->cpu_affinity))
        vcpu_set_affinity(v, &v->cpu_affinity_tmp);

    /* Restore previous trap priority */
    v->trap_priority = v->old_trap_priority;

    /* Restore upcall mask from supplied EFLAGS.IF. */
    vcpu_info(v, evtchn_upcall_mask) = !(eflags & X86_EFLAGS_IF);

    /*
     * The hypercall exit path will overwrite EAX with this return
     * value.
     */
    return regs->eax;

 exit_and_crash:
    gdprintk(XENLOG_ERR, "Fatal error\n");
    domain_crash(v->domain);
    return 0;
}

static void set_task_gate(unsigned int n, unsigned int sel)
{
    idt_table[n].b = 0;
    wmb(); /* disable gate /then/ rewrite */
    idt_table[n].a = sel << 16;
    wmb(); /* rewrite /then/ enable gate */
    idt_table[n].b = 0x8500;