diff options
Diffstat (limited to 'arch/x86/xen/xen-asm.S')
| -rw-r--r-- | arch/x86/xen/xen-asm.S | 367 |
1 files changed, 287 insertions, 80 deletions
diff --git a/arch/x86/xen/xen-asm.S b/arch/x86/xen/xen-asm.S index 8019edd0125c..461bb1526502 100644 --- a/arch/x86/xen/xen-asm.S +++ b/arch/x86/xen/xen-asm.S @@ -6,22 +6,95 @@ * operations here; the indirect forms are better handled in C. */ +#include <asm/errno.h> #include <asm/asm-offsets.h> #include <asm/percpu.h> #include <asm/processor-flags.h> +#include <asm/segment.h> +#include <asm/thread_info.h> +#include <asm/asm.h> #include <asm/frame.h> +#include <asm/unwind_hints.h> +#include <xen/interface/xen.h> + +#include <linux/init.h> #include <linux/linkage.h> +#include <linux/objtool.h> +#include <../entry/calling.h> + +.pushsection .noinstr.text, "ax" +/* + * PV hypercall interface to the hypervisor. + * + * Called via inline asm(), so better preserve %rcx and %r11. + * + * Input: + * %eax: hypercall number + * %rdi, %rsi, %rdx, %r10, %r8: args 1..5 for the hypercall + * Output: %rax + */ +SYM_FUNC_START(xen_hypercall_pv) + ANNOTATE_NOENDBR + push %rcx + push %r11 + UNWIND_HINT_SAVE + syscall + UNWIND_HINT_RESTORE + pop %r11 + pop %rcx + RET +SYM_FUNC_END(xen_hypercall_pv) + +/* + * Disabling events is simply a matter of making the event mask + * non-zero. + */ +SYM_FUNC_START(xen_irq_disable_direct) + ENDBR + movb $1, PER_CPU_VAR(xen_vcpu_info + XEN_vcpu_info_mask) + RET +SYM_FUNC_END(xen_irq_disable_direct) + +/* + * Force an event check by making a hypercall, but preserve regs + * before making the call. + */ +SYM_FUNC_START(check_events) + FRAME_BEGIN + push %rax + push %rcx + push %rdx + push %rsi + push %rdi + push %r8 + push %r9 + push %r10 + push %r11 + call xen_force_evtchn_callback + pop %r11 + pop %r10 + pop %r9 + pop %r8 + pop %rdi + pop %rsi + pop %rdx + pop %rcx + pop %rax + FRAME_END + RET +SYM_FUNC_END(check_events) /* * Enable events. This clears the event mask and tests the pending * event status with one and operation. If there are pending events, * then enter the hypervisor to get them handled. */ -ENTRY(xen_irq_enable_direct) +SYM_FUNC_START(xen_irq_enable_direct) + ENDBR FRAME_BEGIN /* Unmask events */ - movb $0, PER_CPU_VAR(xen_vcpu_info) + XEN_vcpu_info_mask + movb $0, PER_CPU_VAR(xen_vcpu_info + XEN_vcpu_info_mask) /* * Preempt here doesn't matter because that will deal with any @@ -30,24 +103,14 @@ ENTRY(xen_irq_enable_direct) */ /* Test for pending */ - testb $0xff, PER_CPU_VAR(xen_vcpu_info) + XEN_vcpu_info_pending + testb $0xff, PER_CPU_VAR(xen_vcpu_info + XEN_vcpu_info_pending) jz 1f call check_events 1: FRAME_END - ret - ENDPROC(xen_irq_enable_direct) - - -/* - * Disabling events is simply a matter of making the event mask - * non-zero. - */ -ENTRY(xen_irq_disable_direct) - movb $1, PER_CPU_VAR(xen_vcpu_info) + XEN_vcpu_info_mask - ret -ENDPROC(xen_irq_disable_direct) + RET +SYM_FUNC_END(xen_irq_enable_direct) /* * (xen_)save_fl is used to get the current interrupt enable status. @@ -58,80 +121,224 @@ ENDPROC(xen_irq_disable_direct) * undefined. We need to toggle the state of the bit, because Xen and * x86 use opposite senses (mask vs enable). */ -ENTRY(xen_save_fl_direct) - testb $0xff, PER_CPU_VAR(xen_vcpu_info) + XEN_vcpu_info_mask +SYM_FUNC_START(xen_save_fl_direct) + ENDBR + testb $0xff, PER_CPU_VAR(xen_vcpu_info + XEN_vcpu_info_mask) setz %ah addb %ah, %ah - ret - ENDPROC(xen_save_fl_direct) - + RET +SYM_FUNC_END(xen_save_fl_direct) -/* - * In principle the caller should be passing us a value return from - * xen_save_fl_direct, but for robustness sake we test only the - * X86_EFLAGS_IF flag rather than the whole byte. After setting the - * interrupt mask state, it checks for unmasked pending events and - * enters the hypervisor to get them delivered if so. - */ -ENTRY(xen_restore_fl_direct) +SYM_FUNC_START(xen_read_cr2) + ENDBR FRAME_BEGIN -#ifdef CONFIG_X86_64 - testw $X86_EFLAGS_IF, %di -#else - testb $X86_EFLAGS_IF>>8, %ah -#endif - setz PER_CPU_VAR(xen_vcpu_info) + XEN_vcpu_info_mask - /* - * Preempt here doesn't matter because that will deal with any - * pending interrupts. The pending check may end up being run - * on the wrong CPU, but that doesn't hurt. - */ + _ASM_MOV PER_CPU_VAR(xen_vcpu), %_ASM_AX + _ASM_MOV XEN_vcpu_info_arch_cr2(%_ASM_AX), %_ASM_AX + FRAME_END + RET +SYM_FUNC_END(xen_read_cr2); - /* check for unmasked and pending */ - cmpw $0x0001, PER_CPU_VAR(xen_vcpu_info) + XEN_vcpu_info_pending - jnz 1f - call check_events -1: +SYM_FUNC_START(xen_read_cr2_direct) + ENDBR + FRAME_BEGIN + _ASM_MOV PER_CPU_VAR(xen_vcpu_info + XEN_vcpu_info_arch_cr2), %_ASM_AX FRAME_END - ret - ENDPROC(xen_restore_fl_direct) + RET +SYM_FUNC_END(xen_read_cr2_direct); +.popsection +.macro xen_pv_trap name +SYM_CODE_START(xen_\name) + UNWIND_HINT_ENTRY + ENDBR + pop %rcx + pop %r11 + jmp \name +SYM_CODE_END(xen_\name) +_ASM_NOKPROBE(xen_\name) +.endm + +xen_pv_trap asm_exc_divide_error +xen_pv_trap asm_xenpv_exc_debug +xen_pv_trap asm_exc_int3 +xen_pv_trap asm_xenpv_exc_nmi +xen_pv_trap asm_exc_overflow +xen_pv_trap asm_exc_bounds +xen_pv_trap asm_exc_invalid_op +xen_pv_trap asm_exc_device_not_available +xen_pv_trap asm_xenpv_exc_double_fault +xen_pv_trap asm_exc_coproc_segment_overrun +xen_pv_trap asm_exc_invalid_tss +xen_pv_trap asm_exc_segment_not_present +xen_pv_trap asm_exc_stack_segment +xen_pv_trap asm_exc_general_protection +xen_pv_trap asm_exc_page_fault +xen_pv_trap asm_exc_spurious_interrupt_bug +xen_pv_trap asm_exc_coprocessor_error +xen_pv_trap asm_exc_alignment_check +#ifdef CONFIG_X86_CET +xen_pv_trap asm_exc_control_protection +#endif +#ifdef CONFIG_X86_MCE +xen_pv_trap asm_xenpv_exc_machine_check +#endif /* CONFIG_X86_MCE */ +xen_pv_trap asm_exc_simd_coprocessor_error +#ifdef CONFIG_IA32_EMULATION +xen_pv_trap asm_int80_emulation +#endif +xen_pv_trap asm_exc_xen_unknown_trap +xen_pv_trap asm_exc_xen_hypervisor_callback + + __INIT +SYM_CODE_START(xen_early_idt_handler_array) + i = 0 + .rept NUM_EXCEPTION_VECTORS + UNWIND_HINT_UNDEFINED + ENDBR + pop %rcx + pop %r11 + jmp early_idt_handler_array + i*EARLY_IDT_HANDLER_SIZE + i = i + 1 + .fill xen_early_idt_handler_array + i*XEN_EARLY_IDT_HANDLER_SIZE - ., 1, 0xcc + .endr +SYM_CODE_END(xen_early_idt_handler_array) + __FINIT /* - * Force an event check by making a hypercall, but preserve regs - * before making the call. + * Xen64 iret frame: + * + * ss + * rsp + * rflags + * cs + * rip <-- standard iret frame + * + * flags <-- xen_iret must push from here on + * + * rcx + * r11 + * rsp->rax */ -ENTRY(check_events) - FRAME_BEGIN -#ifdef CONFIG_X86_32 - push %eax - push %ecx - push %edx - call xen_force_evtchn_callback - pop %edx - pop %ecx - pop %eax -#else - push %rax +.macro xen_hypercall_iret + pushq $0 /* Flags */ push %rcx - push %rdx - push %rsi - push %rdi - push %r8 - push %r9 - push %r10 push %r11 - call xen_force_evtchn_callback - pop %r11 - pop %r10 - pop %r9 - pop %r8 - pop %rdi - pop %rsi - pop %rdx - pop %rcx - pop %rax -#endif - FRAME_END - ret -ENDPROC(check_events) + push %rax + mov $__HYPERVISOR_iret, %eax + syscall /* Do the IRET. */ + ud2 /* The SYSCALL should never return. */ +.endm + +SYM_CODE_START(xen_iret) + UNWIND_HINT_UNDEFINED + ANNOTATE_NOENDBR + xen_hypercall_iret +SYM_CODE_END(xen_iret) + +/* + * XEN pv doesn't use trampoline stack, PER_CPU_VAR(cpu_tss_rw + TSS_sp0) is + * also the kernel stack. Reusing swapgs_restore_regs_and_return_to_usermode() + * in XEN pv would cause %rsp to move up to the top of the kernel stack and + * leave the IRET frame below %rsp, which is dangerous to be corrupted if #NMI + * interrupts. And swapgs_restore_regs_and_return_to_usermode() pushing the IRET + * frame at the same address is useless. + */ +SYM_CODE_START(xenpv_restore_regs_and_return_to_usermode) + UNWIND_HINT_REGS + POP_REGS + + /* stackleak_erase() can work safely on the kernel stack. */ + STACKLEAK_ERASE_NOCLOBBER + + addq $8, %rsp /* skip regs->orig_ax */ + jmp xen_iret +SYM_CODE_END(xenpv_restore_regs_and_return_to_usermode) + +/* + * Xen handles syscall callbacks much like ordinary exceptions, which + * means we have: + * - kernel gs + * - kernel rsp + * - an iret-like stack frame on the stack (including rcx and r11): + * ss + * rsp + * rflags + * cs + * rip + * r11 + * rsp->rcx + */ + +/* Normal 64-bit system call target */ +SYM_CODE_START(xen_entry_SYSCALL_64) + UNWIND_HINT_ENTRY + ENDBR + popq %rcx + popq %r11 + + /* + * Neither Xen nor the kernel really knows what the old SS and + * CS were. The kernel expects __USER_DS and __USER_CS, so + * report those values even though Xen will guess its own values. + */ + movq $__USER_DS, 4*8(%rsp) + movq $__USER_CS, 1*8(%rsp) + + jmp entry_SYSCALL_64_after_hwframe +SYM_CODE_END(xen_entry_SYSCALL_64) + +#ifdef CONFIG_IA32_EMULATION + +/* 32-bit compat syscall target */ +SYM_CODE_START(xen_entry_SYSCALL_compat) + UNWIND_HINT_ENTRY + ENDBR + popq %rcx + popq %r11 + + /* + * Neither Xen nor the kernel really knows what the old SS and + * CS were. The kernel expects __USER_DS and __USER32_CS, so + * report those values even though Xen will guess its own values. + */ + movq $__USER_DS, 4*8(%rsp) + movq $__USER32_CS, 1*8(%rsp) + + jmp entry_SYSCALL_compat_after_hwframe +SYM_CODE_END(xen_entry_SYSCALL_compat) + +/* 32-bit compat sysenter target */ +SYM_CODE_START(xen_entry_SYSENTER_compat) + UNWIND_HINT_ENTRY + ENDBR + /* + * NB: Xen is polite and clears TF from EFLAGS for us. This means + * that we don't need to guard against single step exceptions here. + */ + popq %rcx + popq %r11 + + /* + * Neither Xen nor the kernel really knows what the old SS and + * CS were. The kernel expects __USER_DS and __USER32_CS, so + * report those values even though Xen will guess its own values. + */ + movq $__USER_DS, 4*8(%rsp) + movq $__USER32_CS, 1*8(%rsp) + + jmp entry_SYSENTER_compat_after_hwframe +SYM_CODE_END(xen_entry_SYSENTER_compat) + +#else /* !CONFIG_IA32_EMULATION */ + +SYM_CODE_START(xen_entry_SYSCALL_compat) +SYM_CODE_START(xen_entry_SYSENTER_compat) + UNWIND_HINT_ENTRY + ENDBR + lea 16(%rsp), %rsp /* strip %rcx, %r11 */ + mov $-ENOSYS, %rax + xen_hypercall_iret +SYM_CODE_END(xen_entry_SYSENTER_compat) +SYM_CODE_END(xen_entry_SYSCALL_compat) + +#endif /* CONFIG_IA32_EMULATION */ |