diff options
Diffstat (limited to 'arch/x86/xen/enlighten_pv.c')
| -rw-r--r-- | arch/x86/xen/enlighten_pv.c | 786 |
1 files changed, 480 insertions, 306 deletions
diff --git a/arch/x86/xen/enlighten_pv.c b/arch/x86/xen/enlighten_pv.c index c54a493e139a..4806cc28d7ca 100644 --- a/arch/x86/xen/enlighten_pv.c +++ b/arch/x86/xen/enlighten_pv.c @@ -23,16 +23,17 @@ #include <linux/start_kernel.h> #include <linux/sched.h> #include <linux/kprobes.h> +#include <linux/kstrtox.h> #include <linux/memblock.h> #include <linux/export.h> #include <linux/mm.h> #include <linux/page-flags.h> -#include <linux/highmem.h> -#include <linux/console.h> #include <linux/pci.h> #include <linux/gfp.h> #include <linux/edd.h> -#include <linux/frame.h> +#include <linux/reboot.h> +#include <linux/virtio_anchor.h> +#include <linux/stackprotector.h> #include <xen/xen.h> #include <xen/events.h> @@ -48,6 +49,7 @@ #include <xen/hvc-console.h> #include <xen/acpi.h> +#include <asm/cpuid/api.h> #include <asm/paravirt.h> #include <asm/apic.h> #include <asm/page.h> @@ -59,33 +61,33 @@ #include <asm/processor.h> #include <asm/proto.h> #include <asm/msr-index.h> +#include <asm/msr.h> #include <asm/traps.h> #include <asm/setup.h> #include <asm/desc.h> #include <asm/pgalloc.h> -#include <asm/pgtable.h> #include <asm/tlbflush.h> #include <asm/reboot.h> -#include <asm/stackprotector.h> #include <asm/hypervisor.h> #include <asm/mach_traps.h> +#include <asm/mtrr.h> #include <asm/mwait.h> #include <asm/pci_x86.h> #include <asm/cpu.h> +#include <asm/irq_stack.h> +#ifdef CONFIG_X86_IOPL_IOPERM +#include <asm/io_bitmap.h> +#endif #ifdef CONFIG_ACPI #include <linux/acpi.h> #include <asm/acpi.h> -#include <acpi/pdc_intel.h> +#include <acpi/proc_cap_intel.h> #include <acpi/processor.h> #include <xen/interface/platform.h> #endif #include "xen-ops.h" -#include "mmu.h" -#include "smp.h" -#include "multicalls.h" -#include "pmu.h" #include "../kernel/cpu/cpu.h" /* get_cpu_cap() */ @@ -94,10 +96,58 @@ void *xen_initial_gdt; static int xen_cpu_up_prepare_pv(unsigned int cpu); static int xen_cpu_dead_pv(unsigned int cpu); +#ifndef CONFIG_PREEMPTION +/* + * Some hypercalls issued by the toolstack can take many 10s of + * seconds. Allow tasks running hypercalls via the privcmd driver to + * be voluntarily preempted even if full kernel preemption is + * disabled. + * + * Such preemptible hypercalls are bracketed by + * xen_preemptible_hcall_begin() and xen_preemptible_hcall_end() + * calls. + */ +DEFINE_PER_CPU(bool, xen_in_preemptible_hcall); +EXPORT_SYMBOL_GPL(xen_in_preemptible_hcall); + +/* + * In case of scheduling the flag must be cleared and restored after + * returning from schedule as the task might move to a different CPU. + */ +static __always_inline bool get_and_clear_inhcall(void) +{ + bool inhcall = __this_cpu_read(xen_in_preemptible_hcall); + + __this_cpu_write(xen_in_preemptible_hcall, false); + return inhcall; +} + +static __always_inline void restore_inhcall(bool inhcall) +{ + __this_cpu_write(xen_in_preemptible_hcall, inhcall); +} + +#else + +static __always_inline bool get_and_clear_inhcall(void) { return false; } +static __always_inline void restore_inhcall(bool inhcall) { } + +#endif + struct tls_descs { struct desc_struct desc[3]; }; +DEFINE_PER_CPU(enum xen_lazy_mode, xen_lazy_mode) = XEN_LAZY_NONE; + +enum xen_lazy_mode xen_get_lazy_mode(void) +{ + if (in_interrupt()) + return XEN_LAZY_NONE; + + return this_cpu_read(xen_lazy_mode); +} + /* * Updating the 3 TLS descriptors in the GDT on every task switch is * surprisingly expensive so we avoid updating them if they haven't @@ -107,20 +157,70 @@ struct tls_descs { */ static DEFINE_PER_CPU(struct tls_descs, shadow_tls_desc); -static void __init xen_banner(void) +static __read_mostly bool xen_msr_safe = IS_ENABLED(CONFIG_XEN_PV_MSR_SAFE); + +static int __init parse_xen_msr_safe(char *str) { - unsigned version = HYPERVISOR_xen_version(XENVER_version, NULL); - struct xen_extraversion extra; - HYPERVISOR_xen_version(XENVER_extraversion, &extra); + if (str) + return kstrtobool(str, &xen_msr_safe); + return -EINVAL; +} +early_param("xen_msr_safe", parse_xen_msr_safe); - pr_info("Booting paravirtualized kernel on %s\n", pv_info.name); - printk(KERN_INFO "Xen version: %d.%d%s%s\n", - version >> 16, version & 0xffff, extra.extraversion, - xen_feature(XENFEAT_mmu_pt_update_preserve_ad) ? " (preserve-AD)" : ""); +/* Get MTRR settings from Xen and put them into mtrr_state. */ +static void __init xen_set_mtrr_data(void) +{ +#ifdef CONFIG_MTRR + struct xen_platform_op op = { + .cmd = XENPF_read_memtype, + .interface_version = XENPF_INTERFACE_VERSION, + }; + unsigned int reg; + unsigned long mask; + uint32_t eax, width; + static struct mtrr_var_range var[MTRR_MAX_VAR_RANGES] __initdata; + + /* Get physical address width (only 64-bit cpus supported). */ + width = 36; + eax = cpuid_eax(0x80000000); + if ((eax >> 16) == 0x8000 && eax >= 0x80000008) { + eax = cpuid_eax(0x80000008); + width = eax & 0xff; + } + + for (reg = 0; reg < MTRR_MAX_VAR_RANGES; reg++) { + op.u.read_memtype.reg = reg; + if (HYPERVISOR_platform_op(&op)) + break; + + /* + * Only called in dom0, which has all RAM PFNs mapped at + * RAM MFNs, and all PCI space etc. is identity mapped. + * This means we can treat MFN == PFN regarding MTRR settings. + */ + var[reg].base_lo = op.u.read_memtype.type; + var[reg].base_lo |= op.u.read_memtype.mfn << PAGE_SHIFT; + var[reg].base_hi = op.u.read_memtype.mfn >> (32 - PAGE_SHIFT); + mask = ~((op.u.read_memtype.nr_mfns << PAGE_SHIFT) - 1); + mask &= (1UL << width) - 1; + if (mask) + mask |= MTRR_PHYSMASK_V; + var[reg].mask_lo = mask; + var[reg].mask_hi = mask >> 32; + } + + /* Only overwrite MTRR state if any MTRR could be got from Xen. */ + if (reg) + guest_force_mtrr_state(var, reg, MTRR_TYPE_UNCACHABLE); +#endif } static void __init xen_pv_init_platform(void) { + /* PV guests can't operate virtio devices without grants. */ + if (IS_ENABLED(CONFIG_XEN_VIRTIO)) + virtio_set_mem_acc_cb(xen_virtio_restricted_mem_acc); + populate_extra_pte(fix_to_virt(FIX_PARAVIRT_BOOTMAP)); set_fixmap(FIX_PARAVIRT_BOOTMAP, xen_start_info->shared_info); @@ -131,6 +231,14 @@ static void __init xen_pv_init_platform(void) /* pvclock is in shared info area */ xen_init_time_ops(); + + if (xen_initial_domain()) + xen_set_mtrr_data(); + else + guest_force_mtrr_state(NULL, 0, MTRR_TYPE_WRBACK); + + /* Adjust nr_cpu_ids before "enumeration" happens */ + xen_smp_count_cpus(); } static void __init xen_pv_guest_late_init(void) @@ -141,36 +249,28 @@ static void __init xen_pv_guest_late_init(void) #endif } -/* Check if running on Xen version (major, minor) or later */ -bool -xen_running_on_version_or_later(unsigned int major, unsigned int minor) -{ - unsigned int version; - - if (!xen_domain()) - return false; - - version = HYPERVISOR_xen_version(XENVER_version, NULL); - if ((((version >> 16) == major) && ((version & 0xffff) >= minor)) || - ((version >> 16) > major)) - return true; - return false; -} - static __read_mostly unsigned int cpuid_leaf5_ecx_val; static __read_mostly unsigned int cpuid_leaf5_edx_val; static void xen_cpuid(unsigned int *ax, unsigned int *bx, unsigned int *cx, unsigned int *dx) { - unsigned maskebx = ~0; + unsigned int maskebx = ~0; + unsigned int or_ebx = 0; /* * Mask out inconvenient features, to try and disable as many * unsupported kernel subsystems as possible. */ switch (*ax) { - case CPUID_MWAIT_LEAF: + case 0x1: + /* Replace initial APIC ID in bits 24-31 of EBX. */ + /* See xen_pv_smp_config() for related topology preparations. */ + maskebx = 0x00ffffff; + or_ebx = smp_processor_id() << 24; + break; + + case CPUID_LEAF_MWAIT: /* Synthesize the values.. */ *ax = 0; *bx = 0; @@ -192,8 +292,8 @@ static void xen_cpuid(unsigned int *ax, unsigned int *bx, : "0" (*ax), "2" (*cx)); *bx &= maskebx; + *bx |= or_ebx; } -STACK_FRAME_NON_STANDARD(xen_cpuid); /* XEN_EMULATE_PREFIX */ static bool __init xen_check_mwait(void) { @@ -240,24 +340,24 @@ static bool __init xen_check_mwait(void) * ecx and edx. The hypercall provides only partial information. */ - ax = CPUID_MWAIT_LEAF; + ax = CPUID_LEAF_MWAIT; bx = 0; cx = 0; dx = 0; native_cpuid(&ax, &bx, &cx, &dx); - /* Ask the Hypervisor whether to clear ACPI_PDC_C_C2C3_FFH. If so, + /* Ask the Hypervisor whether to clear ACPI_PROC_CAP_C_C2C3_FFH. If so, * don't expose MWAIT_LEAF and let ACPI pick the IOPORT version of C3. */ buf[0] = ACPI_PDC_REVISION_ID; buf[1] = 1; - buf[2] = (ACPI_PDC_C_CAPABILITY_SMP | ACPI_PDC_EST_CAPABILITY_SWSMP); + buf[2] = (ACPI_PROC_CAP_C_CAPABILITY_SMP | ACPI_PROC_CAP_EST_CAPABILITY_SWSMP); set_xen_guest_handle(op.u.set_pminfo.pdc, buf); if ((HYPERVISOR_platform_op(&op) == 0) && - (buf[2] & (ACPI_PDC_C_C1_FFH | ACPI_PDC_C_C2C3_FFH))) { + (buf[2] & (ACPI_PROC_CAP_C_C1_FFH | ACPI_PROC_CAP_C_C2C3_FFH))) { cpuid_leaf5_ecx_val = cx; cpuid_leaf5_edx_val = dx; } @@ -282,13 +382,13 @@ static bool __init xen_check_xsave(void) static void __init xen_init_capabilities(void) { - setup_force_cpu_cap(X86_FEATURE_XENPV); setup_clear_cpu_cap(X86_FEATURE_DCA); setup_clear_cpu_cap(X86_FEATURE_APERFMPERF); setup_clear_cpu_cap(X86_FEATURE_MTRR); setup_clear_cpu_cap(X86_FEATURE_ACC); setup_clear_cpu_cap(X86_FEATURE_X2APIC); setup_clear_cpu_cap(X86_FEATURE_SME); + setup_clear_cpu_cap(X86_FEATURE_LKGS); /* * Xen PV would need some work to support PCID: CR3 handling as well @@ -310,20 +410,35 @@ static void __init xen_init_capabilities(void) } } -static void xen_set_debugreg(int reg, unsigned long val) +static noinstr void xen_set_debugreg(int reg, unsigned long val) { HYPERVISOR_set_debugreg(reg, val); } -static unsigned long xen_get_debugreg(int reg) +static noinstr unsigned long xen_get_debugreg(int reg) { return HYPERVISOR_get_debugreg(reg); } +static void xen_start_context_switch(struct task_struct *prev) +{ + BUG_ON(preemptible()); + + if (this_cpu_read(xen_lazy_mode) == XEN_LAZY_MMU) { + arch_leave_lazy_mmu_mode(); + set_ti_thread_flag(task_thread_info(prev), TIF_LAZY_MMU_UPDATES); + } + enter_lazy(XEN_LAZY_CPU); +} + static void xen_end_context_switch(struct task_struct *next) { + BUG_ON(preemptible()); + xen_mc_flush(); - paravirt_end_context_switch(next); + leave_lazy(XEN_LAZY_CPU); + if (test_and_clear_ti_thread_flag(task_thread_info(next), TIF_LAZY_MMU_UPDATES)) + arch_enter_lazy_mmu_mode(); } static unsigned long xen_store_tr(void) @@ -343,15 +458,13 @@ static void set_aliased_prot(void *v, pgprot_t prot) pte_t *ptep; pte_t pte; unsigned long pfn; - struct page *page; unsigned char dummy; + void *va; ptep = lookup_address((unsigned long)v, &level); BUG_ON(ptep == NULL); pfn = pte_pfn(*ptep); - page = pfn_to_page(pfn); - pte = pfn_pte(pfn, prot); /* @@ -376,19 +489,15 @@ static void set_aliased_prot(void *v, pgprot_t prot) preempt_disable(); - probe_kernel_read(&dummy, v, 1); + copy_from_kernel_nofault(&dummy, v, 1); if (HYPERVISOR_update_va_mapping((unsigned long)v, pte, 0)) BUG(); - if (!PageHighMem(page)) { - void *av = __va(PFN_PHYS(pfn)); + va = __va(PFN_PHYS(pfn)); - if (av != v) - if (HYPERVISOR_update_va_mapping((unsigned long)av, pte, 0)) - BUG(); - } else - kmap_flush_unused(); + if (va != v && HYPERVISOR_update_va_mapping((unsigned long)va, pte, 0)) + BUG(); preempt_enable(); } @@ -436,7 +545,7 @@ static void xen_set_ldt(const void *addr, unsigned entries) MULTI_mmuext_op(mcs.mc, op, 1, NULL, DOMID_SELF); - xen_mc_issue(PARAVIRT_LAZY_CPU); + xen_mc_issue(XEN_LAZY_CPU); } static void xen_load_gdt(const struct desc_ptr *dtr) @@ -487,7 +596,7 @@ static void __init xen_load_gdt_boot(const struct desc_ptr *dtr) BUG_ON(size > PAGE_SIZE); BUG_ON(va & ~PAGE_MASK); - pfn = virt_to_pfn(va); + pfn = virt_to_pfn((void *)va); mfn = pfn_to_mfn(pfn); pte = pfn_pte(pfn, PAGE_KERNEL_RO); @@ -528,30 +637,12 @@ static void load_TLS_descriptor(struct thread_struct *t, static void xen_load_tls(struct thread_struct *t, unsigned int cpu) { /* - * XXX sleazy hack: If we're being called in a lazy-cpu zone - * and lazy gs handling is enabled, it means we're in a - * context switch, and %gs has just been saved. This means we - * can zero it out to prevent faults on exit from the - * hypervisor if the next process has no %gs. Either way, it - * has been saved, and the new value will get loaded properly. - * This will go away as soon as Xen has been modified to not - * save/restore %gs for normal hypercalls. - * - * On x86_64, this hack is not used for %gs, because gs points - * to KERNEL_GS_BASE (and uses it for PDA references), so we - * must not zero %gs on x86_64 - * - * For x86_64, we need to zero %fs, otherwise we may get an + * In lazy mode we need to zero %fs, otherwise we may get an * exception between the new %fs descriptor being loaded and * %fs being effectively cleared at __switch_to(). */ - if (paravirt_get_lazy_mode() == PARAVIRT_LAZY_CPU) { -#ifdef CONFIG_X86_32 - lazy_load_gs(0); -#else + if (xen_get_lazy_mode() == XEN_LAZY_CPU) loadsegment(fs, 0); -#endif - } xen_mc_batch(); @@ -559,16 +650,14 @@ static void xen_load_tls(struct thread_struct *t, unsigned int cpu) load_TLS_descriptor(t, cpu, 1); load_TLS_descriptor(t, cpu, 2); - xen_mc_issue(PARAVIRT_LAZY_CPU); + xen_mc_issue(XEN_LAZY_CPU); } -#ifdef CONFIG_X86_64 static void xen_load_gs_index(unsigned int idx) { if (HYPERVISOR_set_segment_base(SEGBASE_GS_USER_SEL, idx)) BUG(); } -#endif static void xen_write_ldt_entry(struct desc_struct *dt, int entrynum, const void *ptr) @@ -587,51 +676,143 @@ static void xen_write_ldt_entry(struct desc_struct *dt, int entrynum, preempt_enable(); } -#ifdef CONFIG_X86_64 +void noist_exc_debug(struct pt_regs *regs); + +DEFINE_IDTENTRY_RAW(xenpv_exc_nmi) +{ + /* On Xen PV, NMI doesn't use IST. The C part is the same as native. */ + exc_nmi(regs); +} + +DEFINE_IDTENTRY_RAW_ERRORCODE(xenpv_exc_double_fault) +{ + /* On Xen PV, DF doesn't use IST. The C part is the same as native. */ + exc_double_fault(regs, error_code); +} + +DEFINE_IDTENTRY_RAW(xenpv_exc_debug) +{ + /* + * There's no IST on Xen PV, but we still need to dispatch + * to the correct handler. + */ + if (user_mode(regs)) + noist_exc_debug(regs); + else + exc_debug(regs); +} + +DEFINE_IDTENTRY_RAW(exc_xen_unknown_trap) +{ + /* This should never happen and there is no way to handle it. */ + instrumentation_begin(); + pr_err("Unknown trap in Xen PV mode."); + BUG(); + instrumentation_end(); +} + +#ifdef CONFIG_X86_MCE +DEFINE_IDTENTRY_RAW(xenpv_exc_machine_check) +{ + /* + * There's no IST on Xen PV, but we still need to dispatch + * to the correct handler. + */ + if (user_mode(regs)) + noist_exc_machine_check(regs); + else + exc_machine_check(regs); +} +#endif + +static void __xen_pv_evtchn_do_upcall(struct pt_regs *regs) +{ + struct pt_regs *old_regs = set_irq_regs(regs); + + inc_irq_stat(irq_hv_callback_count); + + xen_evtchn_do_upcall(); + + set_irq_regs(old_regs); +} + +__visible noinstr void xen_pv_evtchn_do_upcall(struct pt_regs *regs) +{ + irqentry_state_t state = irqentry_enter(regs); + bool inhcall; + + instrumentation_begin(); + run_sysvec_on_irqstack_cond(__xen_pv_evtchn_do_upcall, regs); + + inhcall = get_and_clear_inhcall(); + if (inhcall && !WARN_ON_ONCE(state.exit_rcu)) { + irqentry_exit_cond_resched(); + instrumentation_end(); + restore_inhcall(inhcall); + } else { + instrumentation_end(); + irqentry_exit(regs, state); + } +} + struct trap_array_entry { void (*orig)(void); void (*xen)(void); bool ist_okay; }; +#define TRAP_ENTRY(func, ist_ok) { \ + .orig = asm_##func, \ + .xen = xen_asm_##func, \ + .ist_okay = ist_ok } + +#define TRAP_ENTRY_REDIR(func, ist_ok) { \ + .orig = asm_##func, \ + .xen = xen_asm_xenpv_##func, \ + .ist_okay = ist_ok } + static struct trap_array_entry trap_array[] = { - { debug, xen_xendebug, true }, - { int3, xen_xenint3, true }, - { double_fault, xen_double_fault, true }, + TRAP_ENTRY_REDIR(exc_debug, true ), + TRAP_ENTRY_REDIR(exc_double_fault, true ), #ifdef CONFIG_X86_MCE - { machine_check, xen_machine_check, true }, + TRAP_ENTRY_REDIR(exc_machine_check, true ), #endif - { nmi, xen_xennmi, true }, - { overflow, xen_overflow, false }, + TRAP_ENTRY_REDIR(exc_nmi, true ), + TRAP_ENTRY(exc_int3, false ), + TRAP_ENTRY(exc_overflow, false ), #ifdef CONFIG_IA32_EMULATION - { entry_INT80_compat, xen_entry_INT80_compat, false }, + TRAP_ENTRY(int80_emulation, false ), +#endif + TRAP_ENTRY(exc_page_fault, false ), + TRAP_ENTRY(exc_divide_error, false ), + TRAP_ENTRY(exc_bounds, false ), + TRAP_ENTRY(exc_invalid_op, false ), + TRAP_ENTRY(exc_device_not_available, false ), + TRAP_ENTRY(exc_coproc_segment_overrun, false ), + TRAP_ENTRY(exc_invalid_tss, false ), + TRAP_ENTRY(exc_segment_not_present, false ), + TRAP_ENTRY(exc_stack_segment, false ), + TRAP_ENTRY(exc_general_protection, false ), + TRAP_ENTRY(exc_spurious_interrupt_bug, false ), + TRAP_ENTRY(exc_coprocessor_error, false ), + TRAP_ENTRY(exc_alignment_check, false ), + TRAP_ENTRY(exc_simd_coprocessor_error, false ), +#ifdef CONFIG_X86_CET + TRAP_ENTRY(exc_control_protection, false ), #endif - { page_fault, xen_page_fault, false }, - { divide_error, xen_divide_error, false }, - { bounds, xen_bounds, false }, - { invalid_op, xen_invalid_op, false }, - { device_not_available, xen_device_not_available, false }, - { coprocessor_segment_overrun, xen_coprocessor_segment_overrun, false }, - { invalid_TSS, xen_invalid_TSS, false }, - { segment_not_present, xen_segment_not_present, false }, - { stack_segment, xen_stack_segment, false }, - { general_protection, xen_general_protection, false }, - { spurious_interrupt_bug, xen_spurious_interrupt_bug, false }, - { coprocessor_error, xen_coprocessor_error, false }, - { alignment_check, xen_alignment_check, false }, - { simd_coprocessor_error, xen_simd_coprocessor_error, false }, }; static bool __ref get_trap_addr(void **addr, unsigned int ist) { unsigned int nr; bool ist_okay = false; + bool found = false; /* * Replace trap handler addresses by Xen specific ones. * Check for known traps using IST and whitelist them. * The debugger ones are the only ones we care about. - * Xen will handle faults like double_fault, * so we should never see + * Xen will handle faults like double_fault, so we should never see * them. Warn if there's an unexpected IST-using fault handler. */ for (nr = 0; nr < ARRAY_SIZE(trap_array); nr++) { @@ -640,6 +821,7 @@ static bool __ref get_trap_addr(void **addr, unsigned int ist) if (*addr == entry->orig) { *addr = entry->xen; ist_okay = entry->ist_okay; + found = true; break; } } @@ -650,14 +832,17 @@ static bool __ref get_trap_addr(void **addr, unsigned int ist) nr = (*addr - (void *)early_idt_handler_array[0]) / EARLY_IDT_HANDLER_SIZE; *addr = (void *)xen_early_idt_handler_array[nr]; + found = true; } - if (WARN_ON(ist != 0 && !ist_okay)) + if (!found) + *addr = (void *)xen_asm_exc_xen_unknown_trap; + + if (WARN_ON(found && ist != 0 && !ist_okay)) return false; return true; } -#endif static int cvt_gate_to_trap(int vector, const gate_desc *val, struct trap_info *info) @@ -670,10 +855,8 @@ static int cvt_gate_to_trap(int vector, const gate_desc *val, info->vector = vector; addr = gate_offset(val); -#ifdef CONFIG_X86_64 if (!get_trap_addr((void **)&addr, val->bits.ist)) return 0; -#endif /* CONFIG_X86_64 */ info->address = addr; info->cs = gate_segment(val); @@ -719,8 +902,8 @@ static void xen_write_idt_entry(gate_desc *dt, int entrynum, const gate_desc *g) preempt_enable(); } -static void xen_convert_trap_info(const struct desc_ptr *desc, - struct trap_info *traps) +static unsigned xen_convert_trap_info(const struct desc_ptr *desc, + struct trap_info *traps, bool full) { unsigned in, out, count; @@ -730,17 +913,18 @@ static void xen_convert_trap_info(const struct desc_ptr *desc, for (in = out = 0; in < count; in++) { gate_desc *entry = (gate_desc *)(desc->address) + in; - if (cvt_gate_to_trap(in, entry, &traps[out])) + if (cvt_gate_to_trap(in, entry, &traps[out]) || full) out++; } - traps[out].address = 0; + + return out; } void xen_copy_trap_info(struct trap_info *traps) { const struct desc_ptr *desc = this_cpu_ptr(&idt_desc); - xen_convert_trap_info(desc, traps); + xen_convert_trap_info(desc, traps, true); } /* Load a new IDT into Xen. In principle this can be per-CPU, so we @@ -750,6 +934,8 @@ static void xen_load_idt(const struct desc_ptr *desc) { static DEFINE_SPINLOCK(lock); static struct trap_info traps[257]; + static const struct trap_info zero = { }; + unsigned out; trace_xen_cpu_load_idt(desc); @@ -757,7 +943,8 @@ static void xen_load_idt(const struct desc_ptr *desc) memcpy(this_cpu_ptr(&idt_desc), desc, sizeof(idt_desc)); - xen_convert_trap_info(desc, traps); + out = xen_convert_trap_info(desc, traps, false); + traps[out] = zero; xen_mc_flush(); if (HYPERVISOR_set_trap_table(traps)) @@ -825,18 +1012,39 @@ static void xen_load_sp0(unsigned long sp0) mcs = xen_mc_entry(0); MULTI_stack_switch(mcs.mc, __KERNEL_DS, sp0); - xen_mc_issue(PARAVIRT_LAZY_CPU); + xen_mc_issue(XEN_LAZY_CPU); this_cpu_write(cpu_tss_rw.x86_tss.sp0, sp0); } -void xen_set_iopl_mask(unsigned mask) +#ifdef CONFIG_X86_IOPL_IOPERM +static void xen_invalidate_io_bitmap(void) { - struct physdev_set_iopl set_iopl; + struct physdev_set_iobitmap iobitmap = { + .bitmap = NULL, + .nr_ports = 0, + }; + + native_tss_invalidate_io_bitmap(); + HYPERVISOR_physdev_op(PHYSDEVOP_set_iobitmap, &iobitmap); +} + +static void xen_update_io_bitmap(void) +{ + struct physdev_set_iobitmap iobitmap; + struct tss_struct *tss = this_cpu_ptr(&cpu_tss_rw); + + native_tss_update_io_bitmap(); + + iobitmap.bitmap = (uint8_t *)(&tss->x86_tss) + + tss->x86_tss.io_bitmap_base; + if (tss->x86_tss.io_bitmap_base == IO_BITMAP_OFFSET_INVALID) + iobitmap.nr_ports = 0; + else + iobitmap.nr_ports = IO_BITMAP_BITS; - /* Force the change at ring 0. */ - set_iopl.iopl = (mask == 0) ? 1 : (mask >> 12) & 3; - HYPERVISOR_physdev_op(PHYSDEVOP_set_iopl, &set_iopl); + HYPERVISOR_physdev_op(PHYSDEVOP_set_iobitmap, &iobitmap); } +#endif static void xen_io_delay(void) { @@ -868,7 +1076,7 @@ static void xen_write_cr0(unsigned long cr0) MULTI_fpu_taskswitch(mcs.mc, (cr0 & X86_CR0_TS) != 0); - xen_mc_issue(PARAVIRT_LAZY_CPU); + xen_mc_issue(XEN_LAZY_CPU); } static void xen_write_cr4(unsigned long cr4) @@ -877,54 +1085,56 @@ static void xen_write_cr4(unsigned long cr4) native_write_cr4(cr4); } -#ifdef CONFIG_X86_64 -static inline unsigned long xen_read_cr8(void) -{ - return 0; -} -static inline void xen_write_cr8(unsigned long val) -{ - BUG_ON(val); -} -#endif -static u64 xen_read_msr_safe(unsigned int msr, int *err) +static u64 xen_do_read_msr(u32 msr, int *err) { - u64 val; + u64 val = 0; /* Avoid uninitialized value for safe variant. */ - if (pmu_msr_read(msr, &val, err)) + if (pmu_msr_chk_emulated(msr, &val, true)) return val; - val = native_read_msr_safe(msr, err); + if (err) + *err = native_read_msr_safe(msr, &val); + else + val = native_read_msr(msr); + switch (msr) { case MSR_IA32_APICBASE: val &= ~X2APIC_ENABLE; + if (smp_processor_id() == 0) + val |= MSR_IA32_APICBASE_BSP; + else + val &= ~MSR_IA32_APICBASE_BSP; break; } return val; } -static int xen_write_msr_safe(unsigned int msr, unsigned low, unsigned high) +static void set_seg(u32 which, u64 base) { - int ret; - - ret = 0; + if (HYPERVISOR_set_segment_base(which, base)) + WARN(1, "Xen set_segment_base(%u, %llx) failed\n", which, base); +} +/* + * Support write_msr_safe() and write_msr() semantics. + * With err == NULL write_msr() semantics are selected. + * Supplying an err pointer requires err to be pre-initialized with 0. + */ +static void xen_do_write_msr(u32 msr, u64 val, int *err) +{ switch (msr) { -#ifdef CONFIG_X86_64 - unsigned which; - u64 base; - - case MSR_FS_BASE: which = SEGBASE_FS; goto set; - case MSR_KERNEL_GS_BASE: which = SEGBASE_GS_USER; goto set; - case MSR_GS_BASE: which = SEGBASE_GS_KERNEL; goto set; - - set: - base = ((u64)high << 32) | low; - if (HYPERVISOR_set_segment_base(which, base) != 0) - ret = -EIO; + case MSR_FS_BASE: + set_seg(SEGBASE_FS, val); + break; + + case MSR_KERNEL_GS_BASE: + set_seg(SEGBASE_GS_USER, val); + break; + + case MSR_GS_BASE: + set_seg(SEGBASE_GS_KERNEL, val); break; -#endif case MSR_STAR: case MSR_CSTAR: @@ -939,31 +1149,45 @@ static int xen_write_msr_safe(unsigned int msr, unsigned low, unsigned high) break; default: - if (!pmu_msr_write(msr, low, high, &ret)) - ret = native_write_msr_safe(msr, low, high); + if (pmu_msr_chk_emulated(msr, &val, false)) + return; + + if (err) + *err = native_write_msr_safe(msr, val); + else + native_write_msr(msr, val); } +} - return ret; +static int xen_read_msr_safe(u32 msr, u64 *val) +{ + int err = 0; + + *val = xen_do_read_msr(msr, &err); + return err; } -static u64 xen_read_msr(unsigned int msr) +static int xen_write_msr_safe(u32 msr, u64 val) { - /* - * This will silently swallow a #GP from RDMSR. It may be worth - * changing that. - */ - int err; + int err = 0; - return xen_read_msr_safe(msr, &err); + xen_do_write_msr(msr, val, &err); + + return err; } -static void xen_write_msr(unsigned int msr, unsigned low, unsigned high) +static u64 xen_read_msr(u32 msr) { - /* - * This will silently swallow a #GP from WRMSR. It may be worth - * changing that. - */ - xen_write_msr_safe(msr, low, high); + int err = 0; + + return xen_do_read_msr(msr, xen_msr_safe ? &err : NULL); +} + +static void xen_write_msr(u32 msr, u64 val) +{ + int err; + + xen_do_write_msr(msr, val, xen_msr_safe ? &err : NULL); } /* This is called once we have the cpu_possible_mask */ @@ -974,101 +1198,66 @@ void __init xen_setup_vcpu_info_placement(void) for_each_possible_cpu(cpu) { /* Set up direct vCPU id mapping for PV guests. */ per_cpu(xen_vcpu_id, cpu) = cpu; - - /* - * xen_vcpu_setup(cpu) can fail -- in which case it - * falls back to the shared_info version for cpus - * where xen_vcpu_nr(cpu) < MAX_VIRT_CPUS. - * - * xen_cpu_up_prepare_pv() handles the rest by failing - * them in hotplug. - */ - (void) xen_vcpu_setup(cpu); + xen_vcpu_setup(cpu); } - /* - * xen_vcpu_setup managed to place the vcpu_info within the - * percpu area for all cpus, so make use of it. - */ - if (xen_have_vcpu_info_placement) { - pv_ops.irq.save_fl = __PV_IS_CALLEE_SAVE(xen_save_fl_direct); - pv_ops.irq.restore_fl = - __PV_IS_CALLEE_SAVE(xen_restore_fl_direct); - pv_ops.irq.irq_disable = - __PV_IS_CALLEE_SAVE(xen_irq_disable_direct); - pv_ops.irq.irq_enable = - __PV_IS_CALLEE_SAVE(xen_irq_enable_direct); - pv_ops.mmu.read_cr2 = xen_read_cr2_direct; - } + pv_ops.irq.save_fl = __PV_IS_CALLEE_SAVE(xen_save_fl_direct); + pv_ops.irq.irq_disable = __PV_IS_CALLEE_SAVE(xen_irq_disable_direct); + pv_ops.irq.irq_enable = __PV_IS_CALLEE_SAVE(xen_irq_enable_direct); + pv_ops.mmu.read_cr2 = __PV_IS_CALLEE_SAVE(xen_read_cr2_direct); } static const struct pv_info xen_info __initconst = { - .shared_kernel_pmd = 0, - -#ifdef CONFIG_X86_64 .extra_user_64bit_cs = FLAT_USER_CS64, -#endif .name = "Xen", }; -static const struct pv_cpu_ops xen_cpu_ops __initconst = { - .cpuid = xen_cpuid, +static const typeof(pv_ops) xen_cpu_ops __initconst = { + .cpu = { + .cpuid = xen_cpuid, - .set_debugreg = xen_set_debugreg, - .get_debugreg = xen_get_debugreg, + .set_debugreg = xen_set_debugreg, + .get_debugreg = xen_get_debugreg, - .read_cr0 = xen_read_cr0, - .write_cr0 = xen_write_cr0, + .read_cr0 = xen_read_cr0, + .write_cr0 = xen_write_cr0, - .write_cr4 = xen_write_cr4, + .write_cr4 = xen_write_cr4, -#ifdef CONFIG_X86_64 - .read_cr8 = xen_read_cr8, - .write_cr8 = xen_write_cr8, -#endif + .read_msr = xen_read_msr, + .write_msr = xen_write_msr, - .wbinvd = native_wbinvd, + .read_msr_safe = xen_read_msr_safe, + .write_msr_safe = xen_write_msr_safe, - .read_msr = xen_read_msr, - .write_msr = xen_write_msr, - - .read_msr_safe = xen_read_msr_safe, - .write_msr_safe = xen_write_msr_safe, - - .read_pmc = xen_read_pmc, - - .iret = xen_iret, -#ifdef CONFIG_X86_64 - .usergs_sysret64 = xen_sysret64, -#endif - - .load_tr_desc = paravirt_nop, - .set_ldt = xen_set_ldt, - .load_gdt = xen_load_gdt, - .load_idt = xen_load_idt, - .load_tls = xen_load_tls, -#ifdef CONFIG_X86_64 - .load_gs_index = xen_load_gs_index, -#endif + .read_pmc = xen_read_pmc, - .alloc_ldt = xen_alloc_ldt, - .free_ldt = xen_free_ldt, + .load_tr_desc = paravirt_nop, + .set_ldt = xen_set_ldt, + .load_gdt = xen_load_gdt, + .load_idt = xen_load_idt, + .load_tls = xen_load_tls, + .load_gs_index = xen_load_gs_index, - .store_tr = xen_store_tr, + .alloc_ldt = xen_alloc_ldt, + .free_ldt = xen_free_ldt, - .write_ldt_entry = xen_write_ldt_entry, - .write_gdt_entry = xen_write_gdt_entry, - .write_idt_entry = xen_write_idt_entry, - .load_sp0 = xen_load_sp0, + .store_tr = xen_store_tr, - .set_iopl_mask = xen_set_iopl_mask, - .io_delay = xen_io_delay, + .write_ldt_entry = xen_write_ldt_entry, + .write_gdt_entry = xen_write_gdt_entry, + .write_idt_entry = xen_write_idt_entry, + .load_sp0 = xen_load_sp0, - /* Xen takes care of %gs when switching to usermode for us */ - .swapgs = paravirt_nop, +#ifdef CONFIG_X86_IOPL_IOPERM + .invalidate_io_bitmap = xen_invalidate_io_bitmap, + .update_io_bitmap = xen_update_io_bitmap, +#endif + .io_delay = xen_io_delay, - .start_context_switch = paravirt_start_context_switch, - .end_context_switch = xen_end_context_switch, + .start_context_switch = xen_start_context_switch, + .end_context_switch = xen_end_context_switch, + }, }; static void xen_restart(char *msg) @@ -1083,8 +1272,7 @@ static void xen_machine_halt(void) static void xen_machine_power_off(void) { - if (pm_power_off) - pm_power_off(); + do_kernel_power_off(); xen_reboot(SHUTDOWN_poweroff); } @@ -1176,8 +1364,7 @@ static void __init xen_setup_gdt(int cpu) pv_ops.cpu.write_gdt_entry = xen_write_gdt_entry_boot; pv_ops.cpu.load_gdt = xen_load_gdt_boot; - setup_stack_canary_segment(cpu); - switch_to_new_gdt(cpu); + switch_gdt_and_percpu_base(cpu); pv_ops.cpu.write_gdt_entry = xen_write_gdt_entry; pv_ops.cpu.load_gdt = xen_load_gdt; @@ -1188,25 +1375,43 @@ static void __init xen_dom0_set_legacy_features(void) x86_platform.legacy.rtc = 1; } +static void __init xen_domu_set_legacy_features(void) +{ + x86_platform.legacy.rtc = 0; +} + +extern void early_xen_iret_patch(void); + /* First C function to be called on Xen boot */ -asmlinkage __visible void __init xen_start_kernel(void) +asmlinkage __visible void __init xen_start_kernel(struct start_info *si) { struct physdev_set_iopl set_iopl; unsigned long initrd_start = 0; int rc; - if (!xen_start_info) + if (!si) return; + clear_bss(); + + xen_start_info = si; + + __text_gen_insn(&early_xen_iret_patch, + JMP32_INSN_OPCODE, &early_xen_iret_patch, &xen_iret, + JMP32_INSN_SIZE); + xen_domain_type = XEN_PV_DOMAIN; + setup_force_cpu_cap(X86_FEATURE_XENPV); xen_start_flags = xen_start_info->flags; + /* Interrupts are guaranteed to be off initially. */ + early_boot_irqs_disabled = true; + static_call_update_early(xen_hypercall, xen_hypercall_pv); xen_setup_features(); /* Install Xen paravirt ops */ pv_info = xen_info; - pv_ops.init.patch = paravirt_patch_default; - pv_ops.cpu = xen_cpu_ops; + pv_ops.cpu = xen_cpu_ops.cpu; xen_init_irq_ops(); /* @@ -1219,9 +1424,12 @@ asmlinkage __visible void __init xen_start_kernel(void) xen_vcpu_info_reset(0); x86_platform.get_nmi_reason = xen_get_nmi_reason; + x86_platform.realmode_reserve = x86_init_noop; + x86_platform.realmode_init = x86_init_noop; x86_init.resources.memory_setup = xen_memory_setup; - x86_init.irqs.intr_mode_init = x86_init_noop; + x86_init.irqs.intr_mode_select = x86_init_noop; + x86_init.irqs.intr_mode_init = x86_64_probe_apic; x86_init.oem.arch_setup = xen_arch_setup; x86_init.oem.banner = xen_banner; x86_init.hyper.init_platform = xen_pv_init_platform; @@ -1238,25 +1446,19 @@ asmlinkage __visible void __init xen_start_kernel(void) __supported_pte_mask &= ~_PAGE_GLOBAL; __default_kernel_pte_mask &= ~_PAGE_GLOBAL; - /* - * Prevent page tables from being allocated in highmem, even - * if CONFIG_HIGHPTE is enabled. - */ - __userpte_alloc_gfp &= ~__GFP_HIGHMEM; - /* Get mfn list */ xen_build_dynamic_phys_to_machine(); + /* Work out if we support NX */ + get_cpu_cap(&boot_cpu_data); + x86_configure_nx(); + /* * Set up kernel GDT and segment registers, mainly so that * -fstack-protector code can be executed. */ xen_setup_gdt(0); - /* Work out if we support NX */ - get_cpu_cap(&boot_cpu_data); - x86_configure_nx(); - /* Determine virtual and physical address sizes */ get_cpu_address_sizes(&boot_cpu_data); @@ -1267,19 +1469,10 @@ asmlinkage __visible void __init xen_start_kernel(void) xen_init_capabilities(); -#ifdef CONFIG_X86_LOCAL_APIC /* * set up the basic apic ops. */ xen_init_apic(); -#endif - - if (xen_feature(XENFEAT_mmu_pt_update_preserve_ad)) { - pv_ops.mmu.ptep_modify_prot_start = - xen_ptep_modify_prot_start; - pv_ops.mmu.ptep_modify_prot_commit = - xen_ptep_modify_prot_commit; - } machine_ops = xen_machine_ops; @@ -1298,30 +1491,17 @@ asmlinkage __visible void __init xen_start_kernel(void) * any NUMA information the kernel tries to get from ACPI will * be meaningless. Prevent it from trying. */ - acpi_numa = -1; + disable_srat(); #endif WARN_ON(xen_cpuhp_setup(xen_cpu_up_prepare_pv, xen_cpu_dead_pv)); local_irq_disable(); - early_boot_irqs_disabled = true; xen_raw_console_write("mapping kernel into physical memory\n"); xen_setup_kernel_pagetable((pgd_t *)xen_start_info->pt_base, xen_start_info->nr_pages); xen_reserve_special_pages(); - /* keep using Xen gdt for now; no urgent need to change it */ - -#ifdef CONFIG_X86_32 - pv_info.kernel_rpl = 1; - if (xen_feature(XENFEAT_supervisor_mode_kernel)) - pv_info.kernel_rpl = 0; -#else - pv_info.kernel_rpl = 0; -#endif - /* set the limit of our address space */ - xen_reserve_top(); - /* * We used to do this in xen_arch_setup, but that is too late * on AMD were early_cpu_init (run before ->arch_setup()) calls @@ -1332,12 +1512,6 @@ asmlinkage __visible void __init xen_start_kernel(void) if (rc != 0) xen_raw_printk("physdev_op failed %d\n", rc); -#ifdef CONFIG_X86_32 - /* set up basic CPUID stuff */ - cpu_detect(&new_cpu_data); - set_cpu_cap(&new_cpu_data, X86_FEATURE_FPU); - new_cpu_data.x86_capability[CPUID_1_EDX] = cpuid_edx(1); -#endif if (xen_start_info->mod_start) { if (xen_start_info->flags & SIF_MOD_START_PFN) @@ -1354,9 +1528,10 @@ asmlinkage __visible void __init xen_start_kernel(void) boot_params.hdr.hardware_subarch = X86_SUBARCH_XEN; if (!xen_initial_domain()) { - add_preferred_console("xenboot", 0, NULL); if (pci_xen) x86_init.pci.arch_init = pci_xen_init; + x86_platform.set_legacy_features = + xen_domu_set_legacy_features; } else { const struct dom0_vga_console_info *info = (void *)((char *)xen_start_info + @@ -1369,7 +1544,8 @@ asmlinkage __visible void __init xen_start_kernel(void) x86_platform.set_legacy_features = xen_dom0_set_legacy_features; - xen_init_vga(info, xen_start_info->console.dom0.info_size); + xen_init_vga(info, xen_start_info->console.dom0.info_size, + &boot_params.screen_info); xen_start_info->console.domU.mfn = 0; xen_start_info->console.domU.evtchn = 0; @@ -1381,18 +1557,19 @@ asmlinkage __visible void __init xen_start_kernel(void) xen_acpi_sleep_register(); - /* Avoid searching for BIOS MP tables */ - x86_init.mpparse.find_smp_config = x86_init_noop; - x86_init.mpparse.get_smp_config = x86_init_uint_noop; - xen_boot_params_init_edd(); + +#ifdef CONFIG_ACPI + /* + * Disable selecting "Firmware First mode" for correctable + * memory errors, as this is the duty of the hypervisor to + * decide. + */ + acpi_disable_cmcff = 1; +#endif } - if (!boot_params.screen_info.orig_video_isVGA) - add_preferred_console("tty", 0, NULL); - add_preferred_console("hvc", 0, NULL); - if (boot_params.screen_info.orig_video_isVGA) - add_preferred_console("tty", 0, NULL); + xen_add_preferred_consoles(); #ifdef CONFIG_PCI /* PCI BIOS service won't work from a PV guest. */ @@ -1403,15 +1580,11 @@ asmlinkage __visible void __init xen_start_kernel(void) /* We need this for printk timestamps */ xen_setup_runstate_info(0); - xen_efi_init(); + xen_efi_init(&boot_params); /* Start the world */ -#ifdef CONFIG_X86_32 - i386_start_kernel(); -#else cr4_init_shadow(); /* 32b kernel does this in i386_start_kernel() */ x86_64_start_reservations((char *)__pa_symbol(&boot_params)); -#endif } static int xen_cpu_up_prepare_pv(unsigned int cpu) @@ -1463,4 +1636,5 @@ const __initconst struct hypervisor_x86 x86_hyper_xen_pv = { .detect = xen_platform_pv, .type = X86_HYPER_XEN_PV, .runtime.pin_vcpu = xen_pin_vcpu, + .ignore_nopv = true, }; |