freebsd-skq/sys/amd64/vmm/vmm.c

1828 lines
40 KiB
C
Raw Normal View History

/*-
* Copyright (c) 2011 NetApp, Inc.
* All rights reserved.
*
* 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 NETAPP, INC ``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 NETAPP, INC 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.
*
* $FreeBSD$
*/
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/kernel.h>
#include <sys/module.h>
#include <sys/sysctl.h>
#include <sys/malloc.h>
#include <sys/pcpu.h>
#include <sys/lock.h>
#include <sys/mutex.h>
#include <sys/proc.h>
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
#include <sys/rwlock.h>
#include <sys/sched.h>
#include <sys/smp.h>
#include <sys/systm.h>
#include <vm/vm.h>
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
#include <vm/vm_object.h>
#include <vm/vm_page.h>
#include <vm/pmap.h>
#include <vm/vm_map.h>
#include <vm/vm_extern.h>
#include <vm/vm_param.h>
#include <machine/cpu.h>
#include <machine/vm.h>
#include <machine/pcb.h>
#include <machine/smp.h>
#include <x86/psl.h>
#include <x86/apicreg.h>
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
#include <machine/vmparam.h>
#include <machine/vmm.h>
#include <machine/vmm_dev.h>
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
#include "vmm_ktr.h"
#include "vmm_host.h"
#include "vmm_mem.h"
#include "vmm_util.h"
#include "vatpic.h"
#include "vatpit.h"
#include "vhpet.h"
#include "vioapic.h"
#include "vlapic.h"
#include "vmm_msr.h"
#include "vmm_ipi.h"
#include "vmm_stat.h"
#include "vmm_lapic.h"
#include "io/ppt.h"
#include "io/iommu.h"
struct vlapic;
struct vcpu {
int flags;
enum vcpu_state state;
struct mtx mtx;
int hostcpu; /* host cpuid this vcpu last ran on */
uint64_t guest_msrs[VMM_MSR_NUM];
struct vlapic *vlapic;
int vcpuid;
struct savefpu *guestfpu; /* guest fpu state */
uint64_t guest_xcr0;
void *stats;
struct vm_exit exitinfo;
enum x2apic_state x2apic_state;
int nmi_pending;
int extint_pending;
struct vm_exception exception;
int exception_pending;
};
#define vcpu_lock_init(v) mtx_init(&((v)->mtx), "vcpu lock", 0, MTX_SPIN)
#define vcpu_lock(v) mtx_lock_spin(&((v)->mtx))
#define vcpu_unlock(v) mtx_unlock_spin(&((v)->mtx))
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
#define vcpu_assert_locked(v) mtx_assert(&((v)->mtx), MA_OWNED)
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
struct mem_seg {
vm_paddr_t gpa;
size_t len;
boolean_t wired;
vm_object_t object;
};
#define VM_MAX_MEMORY_SEGMENTS 2
struct vm {
void *cookie; /* processor-specific data */
void *iommu; /* iommu-specific data */
struct vhpet *vhpet; /* virtual HPET */
struct vioapic *vioapic; /* virtual ioapic */
struct vatpic *vatpic; /* virtual atpic */
struct vatpit *vatpit; /* virtual atpit */
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
struct vmspace *vmspace; /* guest's address space */
struct vcpu vcpu[VM_MAXCPU];
int num_mem_segs;
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
struct mem_seg mem_segs[VM_MAX_MEMORY_SEGMENTS];
char name[VM_MAX_NAMELEN];
/*
2011-06-28 06:26:03 +00:00
* Set of active vcpus.
* An active vcpu is one that has been started implicitly (BSP) or
* explicitly (AP) by sending it a startup ipi.
*/
volatile cpuset_t active_cpus;
struct mtx rendezvous_mtx;
cpuset_t rendezvous_req_cpus;
cpuset_t rendezvous_done_cpus;
void *rendezvous_arg;
vm_rendezvous_func_t rendezvous_func;
int suspend;
volatile cpuset_t suspended_cpus;
};
static int vmm_initialized;
static struct vmm_ops *ops;
2014-01-09 03:25:54 +00:00
#define VMM_INIT(num) (ops != NULL ? (*ops->init)(num) : 0)
#define VMM_CLEANUP() (ops != NULL ? (*ops->cleanup)() : 0)
#define VMM_RESUME() (ops != NULL ? (*ops->resume)() : 0)
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
#define VMINIT(vm, pmap) (ops != NULL ? (*ops->vminit)(vm, pmap): NULL)
#define VMRUN(vmi, vcpu, rip, pmap, rptr, sptr) \
(ops != NULL ? (*ops->vmrun)(vmi, vcpu, rip, pmap, rptr, sptr) : ENXIO)
#define VMCLEANUP(vmi) (ops != NULL ? (*ops->vmcleanup)(vmi) : NULL)
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
#define VMSPACE_ALLOC(min, max) \
(ops != NULL ? (*ops->vmspace_alloc)(min, max) : NULL)
#define VMSPACE_FREE(vmspace) \
(ops != NULL ? (*ops->vmspace_free)(vmspace) : ENXIO)
#define VMGETREG(vmi, vcpu, num, retval) \
(ops != NULL ? (*ops->vmgetreg)(vmi, vcpu, num, retval) : ENXIO)
#define VMSETREG(vmi, vcpu, num, val) \
(ops != NULL ? (*ops->vmsetreg)(vmi, vcpu, num, val) : ENXIO)
#define VMGETDESC(vmi, vcpu, num, desc) \
(ops != NULL ? (*ops->vmgetdesc)(vmi, vcpu, num, desc) : ENXIO)
#define VMSETDESC(vmi, vcpu, num, desc) \
(ops != NULL ? (*ops->vmsetdesc)(vmi, vcpu, num, desc) : ENXIO)
#define VMGETCAP(vmi, vcpu, num, retval) \
(ops != NULL ? (*ops->vmgetcap)(vmi, vcpu, num, retval) : ENXIO)
#define VMSETCAP(vmi, vcpu, num, val) \
(ops != NULL ? (*ops->vmsetcap)(vmi, vcpu, num, val) : ENXIO)
#define VLAPIC_INIT(vmi, vcpu) \
(ops != NULL ? (*ops->vlapic_init)(vmi, vcpu) : NULL)
#define VLAPIC_CLEANUP(vmi, vlapic) \
(ops != NULL ? (*ops->vlapic_cleanup)(vmi, vlapic) : NULL)
#define fpu_start_emulating() load_cr0(rcr0() | CR0_TS)
#define fpu_stop_emulating() clts()
static MALLOC_DEFINE(M_VM, "vm", "vm");
CTASSERT(VMM_MSR_NUM <= 64); /* msr_mask can keep track of up to 64 msrs */
/* statistics */
static VMM_STAT(VCPU_TOTAL_RUNTIME, "vcpu total runtime");
2014-01-09 03:25:54 +00:00
SYSCTL_NODE(_hw, OID_AUTO, vmm, CTLFLAG_RW, NULL, NULL);
static int vmm_ipinum;
SYSCTL_INT(_hw_vmm, OID_AUTO, ipinum, CTLFLAG_RD, &vmm_ipinum, 0,
"IPI vector used for vcpu notifications");
static void vm_deactivate_cpu(struct vm *vm, int vcpuid);
static void
vcpu_cleanup(struct vm *vm, int i)
{
struct vcpu *vcpu = &vm->vcpu[i];
VLAPIC_CLEANUP(vm->cookie, vcpu->vlapic);
vmm_stat_free(vcpu->stats);
fpu_save_area_free(vcpu->guestfpu);
}
static void
vcpu_init(struct vm *vm, uint32_t vcpu_id)
{
struct vcpu *vcpu;
vcpu = &vm->vcpu[vcpu_id];
vcpu_lock_init(vcpu);
vcpu->hostcpu = NOCPU;
vcpu->vcpuid = vcpu_id;
vcpu->vlapic = VLAPIC_INIT(vm->cookie, vcpu_id);
vm_set_x2apic_state(vm, vcpu_id, X2APIC_DISABLED);
vcpu->guest_xcr0 = XFEATURE_ENABLED_X87;
vcpu->guestfpu = fpu_save_area_alloc();
fpu_save_area_reset(vcpu->guestfpu);
vcpu->stats = vmm_stat_alloc();
}
struct vm_exit *
vm_exitinfo(struct vm *vm, int cpuid)
{
struct vcpu *vcpu;
if (cpuid < 0 || cpuid >= VM_MAXCPU)
panic("vm_exitinfo: invalid cpuid %d", cpuid);
vcpu = &vm->vcpu[cpuid];
return (&vcpu->exitinfo);
}
static void
vmm_resume(void)
{
VMM_RESUME();
}
static int
vmm_init(void)
{
int error;
vmm_host_state_init();
2014-01-09 03:25:54 +00:00
vmm_ipinum = vmm_ipi_alloc();
if (vmm_ipinum == 0)
vmm_ipinum = IPI_AST;
error = vmm_mem_init();
if (error)
return (error);
if (vmm_is_intel())
ops = &vmm_ops_intel;
else if (vmm_is_amd())
ops = &vmm_ops_amd;
else
return (ENXIO);
vmm_msr_init();
vmm_resume_p = vmm_resume;
2014-01-09 03:25:54 +00:00
return (VMM_INIT(vmm_ipinum));
}
static int
vmm_handler(module_t mod, int what, void *arg)
{
int error;
switch (what) {
case MOD_LOAD:
vmmdev_init();
if (ppt_avail_devices() > 0)
iommu_init();
error = vmm_init();
if (error == 0)
vmm_initialized = 1;
break;
case MOD_UNLOAD:
error = vmmdev_cleanup();
if (error == 0) {
vmm_resume_p = NULL;
iommu_cleanup();
2014-01-09 03:25:54 +00:00
if (vmm_ipinum != IPI_AST)
vmm_ipi_free(vmm_ipinum);
error = VMM_CLEANUP();
/*
* Something bad happened - prevent new
* VMs from being created
*/
if (error)
vmm_initialized = 0;
}
break;
default:
error = 0;
break;
}
return (error);
}
static moduledata_t vmm_kmod = {
"vmm",
vmm_handler,
NULL
};
/*
* vmm initialization has the following dependencies:
*
* - iommu initialization must happen after the pci passthru driver has had
* a chance to attach to any passthru devices (after SI_SUB_CONFIGURE).
*
* - VT-x initialization requires smp_rendezvous() and therefore must happen
* after SMP is fully functional (after SI_SUB_SMP).
*/
DECLARE_MODULE(vmm, vmm_kmod, SI_SUB_SMP + 1, SI_ORDER_ANY);
MODULE_VERSION(vmm, 1);
int
vm_create(const char *name, struct vm **retvm)
{
int i;
struct vm *vm;
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
struct vmspace *vmspace;
const int BSP = 0;
/*
* If vmm.ko could not be successfully initialized then don't attempt
* to create the virtual machine.
*/
if (!vmm_initialized)
return (ENXIO);
if (name == NULL || strlen(name) >= VM_MAX_NAMELEN)
return (EINVAL);
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
vmspace = VMSPACE_ALLOC(VM_MIN_ADDRESS, VM_MAXUSER_ADDRESS);
if (vmspace == NULL)
return (ENOMEM);
vm = malloc(sizeof(struct vm), M_VM, M_WAITOK | M_ZERO);
strcpy(vm->name, name);
vm->vmspace = vmspace;
mtx_init(&vm->rendezvous_mtx, "vm rendezvous lock", 0, MTX_DEF);
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
vm->cookie = VMINIT(vm, vmspace_pmap(vmspace));
vm->vioapic = vioapic_init(vm);
vm->vhpet = vhpet_init(vm);
vm->vatpic = vatpic_init(vm);
vm->vatpit = vatpit_init(vm);
for (i = 0; i < VM_MAXCPU; i++) {
vcpu_init(vm, i);
guest_msrs_init(vm, i);
}
vm_activate_cpu(vm, BSP);
*retvm = vm;
return (0);
}
static void
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
vm_free_mem_seg(struct vm *vm, struct mem_seg *seg)
{
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
if (seg->object != NULL)
vmm_mem_free(vm->vmspace, seg->gpa, seg->len);
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
bzero(seg, sizeof(*seg));
}
void
vm_destroy(struct vm *vm)
{
int i;
ppt_unassign_all(vm);
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
if (vm->iommu != NULL)
iommu_destroy_domain(vm->iommu);
vatpit_cleanup(vm->vatpit);
vhpet_cleanup(vm->vhpet);
vatpic_cleanup(vm->vatpic);
vioapic_cleanup(vm->vioapic);
for (i = 0; i < vm->num_mem_segs; i++)
vm_free_mem_seg(vm, &vm->mem_segs[i]);
vm->num_mem_segs = 0;
for (i = 0; i < VM_MAXCPU; i++)
vcpu_cleanup(vm, i);
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
VMSPACE_FREE(vm->vmspace);
VMCLEANUP(vm->cookie);
free(vm, M_VM);
}
const char *
vm_name(struct vm *vm)
{
return (vm->name);
}
int
vm_map_mmio(struct vm *vm, vm_paddr_t gpa, size_t len, vm_paddr_t hpa)
{
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
vm_object_t obj;
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
if ((obj = vmm_mmio_alloc(vm->vmspace, gpa, len, hpa)) == NULL)
return (ENOMEM);
else
return (0);
}
int
vm_unmap_mmio(struct vm *vm, vm_paddr_t gpa, size_t len)
{
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
vmm_mmio_free(vm->vmspace, gpa, len);
return (0);
}
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
boolean_t
vm_mem_allocated(struct vm *vm, vm_paddr_t gpa)
{
int i;
vm_paddr_t gpabase, gpalimit;
for (i = 0; i < vm->num_mem_segs; i++) {
gpabase = vm->mem_segs[i].gpa;
gpalimit = gpabase + vm->mem_segs[i].len;
if (gpa >= gpabase && gpa < gpalimit)
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
return (TRUE); /* 'gpa' is regular memory */
}
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
if (ppt_is_mmio(vm, gpa))
return (TRUE); /* 'gpa' is pci passthru mmio */
return (FALSE);
}
int
vm_malloc(struct vm *vm, vm_paddr_t gpa, size_t len)
{
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
int available, allocated;
struct mem_seg *seg;
vm_object_t object;
vm_paddr_t g;
if ((gpa & PAGE_MASK) || (len & PAGE_MASK) || len == 0)
return (EINVAL);
available = allocated = 0;
g = gpa;
while (g < gpa + len) {
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
if (vm_mem_allocated(vm, g))
allocated++;
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
else
available++;
g += PAGE_SIZE;
}
/*
* If there are some allocated and some available pages in the address
* range then it is an error.
*/
if (allocated && available)
return (EINVAL);
/*
* If the entire address range being requested has already been
* allocated then there isn't anything more to do.
*/
if (allocated && available == 0)
return (0);
if (vm->num_mem_segs >= VM_MAX_MEMORY_SEGMENTS)
return (E2BIG);
seg = &vm->mem_segs[vm->num_mem_segs];
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
if ((object = vmm_mem_alloc(vm->vmspace, gpa, len)) == NULL)
return (ENOMEM);
seg->gpa = gpa;
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
seg->len = len;
seg->object = object;
seg->wired = FALSE;
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
vm->num_mem_segs++;
return (0);
}
static void
vm_gpa_unwire(struct vm *vm)
{
int i, rv;
struct mem_seg *seg;
for (i = 0; i < vm->num_mem_segs; i++) {
seg = &vm->mem_segs[i];
if (!seg->wired)
continue;
rv = vm_map_unwire(&vm->vmspace->vm_map,
seg->gpa, seg->gpa + seg->len,
VM_MAP_WIRE_USER | VM_MAP_WIRE_NOHOLES);
KASSERT(rv == KERN_SUCCESS, ("vm(%s) memory segment "
"%#lx/%ld could not be unwired: %d",
vm_name(vm), seg->gpa, seg->len, rv));
seg->wired = FALSE;
}
}
static int
vm_gpa_wire(struct vm *vm)
{
int i, rv;
struct mem_seg *seg;
for (i = 0; i < vm->num_mem_segs; i++) {
seg = &vm->mem_segs[i];
if (seg->wired)
continue;
/* XXX rlimits? */
rv = vm_map_wire(&vm->vmspace->vm_map,
seg->gpa, seg->gpa + seg->len,
VM_MAP_WIRE_USER | VM_MAP_WIRE_NOHOLES);
if (rv != KERN_SUCCESS)
break;
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
seg->wired = TRUE;
}
if (i < vm->num_mem_segs) {
/*
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
* Undo the wiring before returning an error.
*/
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
vm_gpa_unwire(vm);
return (EAGAIN);
}
return (0);
}
static void
vm_iommu_modify(struct vm *vm, boolean_t map)
{
int i, sz;
vm_paddr_t gpa, hpa;
struct mem_seg *seg;
void *vp, *cookie, *host_domain;
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
sz = PAGE_SIZE;
host_domain = iommu_host_domain();
for (i = 0; i < vm->num_mem_segs; i++) {
seg = &vm->mem_segs[i];
KASSERT(seg->wired, ("vm(%s) memory segment %#lx/%ld not wired",
vm_name(vm), seg->gpa, seg->len));
gpa = seg->gpa;
while (gpa < seg->gpa + seg->len) {
vp = vm_gpa_hold(vm, gpa, PAGE_SIZE, VM_PROT_WRITE,
&cookie);
KASSERT(vp != NULL, ("vm(%s) could not map gpa %#lx",
vm_name(vm), gpa));
vm_gpa_release(cookie);
hpa = DMAP_TO_PHYS((uintptr_t)vp);
if (map) {
iommu_create_mapping(vm->iommu, gpa, hpa, sz);
iommu_remove_mapping(host_domain, hpa, sz);
} else {
iommu_remove_mapping(vm->iommu, gpa, sz);
iommu_create_mapping(host_domain, hpa, hpa, sz);
}
gpa += PAGE_SIZE;
}
}
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
/*
* Invalidate the cached translations associated with the domain
* from which pages were removed.
*/
if (map)
iommu_invalidate_tlb(host_domain);
else
iommu_invalidate_tlb(vm->iommu);
}
#define vm_iommu_unmap(vm) vm_iommu_modify((vm), FALSE)
#define vm_iommu_map(vm) vm_iommu_modify((vm), TRUE)
int
vm_unassign_pptdev(struct vm *vm, int bus, int slot, int func)
{
int error;
error = ppt_unassign_device(vm, bus, slot, func);
if (error)
return (error);
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
if (ppt_assigned_devices(vm) == 0) {
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
vm_iommu_unmap(vm);
vm_gpa_unwire(vm);
}
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
return (0);
}
int
vm_assign_pptdev(struct vm *vm, int bus, int slot, int func)
{
int error;
vm_paddr_t maxaddr;
/*
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
* Virtual machines with pci passthru devices get special treatment:
* - the guest physical memory is wired
* - the iommu is programmed to do the 'gpa' to 'hpa' translation
*
* We need to do this before the first pci passthru device is attached.
*/
if (ppt_assigned_devices(vm) == 0) {
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
KASSERT(vm->iommu == NULL,
("vm_assign_pptdev: iommu must be NULL"));
maxaddr = vmm_mem_maxaddr();
vm->iommu = iommu_create_domain(maxaddr);
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
error = vm_gpa_wire(vm);
if (error)
return (error);
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
vm_iommu_map(vm);
}
error = ppt_assign_device(vm, bus, slot, func);
return (error);
}
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
void *
vm_gpa_hold(struct vm *vm, vm_paddr_t gpa, size_t len, int reqprot,
void **cookie)
{
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
int count, pageoff;
vm_page_t m;
pageoff = gpa & PAGE_MASK;
if (len > PAGE_SIZE - pageoff)
panic("vm_gpa_hold: invalid gpa/len: 0x%016lx/%lu", gpa, len);
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
count = vm_fault_quick_hold_pages(&vm->vmspace->vm_map,
trunc_page(gpa), PAGE_SIZE, reqprot, &m, 1);
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
if (count == 1) {
*cookie = m;
return ((void *)(PHYS_TO_DMAP(VM_PAGE_TO_PHYS(m)) + pageoff));
} else {
*cookie = NULL;
return (NULL);
}
}
void
vm_gpa_release(void *cookie)
{
vm_page_t m = cookie;
vm_page_lock(m);
vm_page_unhold(m);
vm_page_unlock(m);
}
int
vm_gpabase2memseg(struct vm *vm, vm_paddr_t gpabase,
struct vm_memory_segment *seg)
{
int i;
for (i = 0; i < vm->num_mem_segs; i++) {
if (gpabase == vm->mem_segs[i].gpa) {
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
seg->gpa = vm->mem_segs[i].gpa;
seg->len = vm->mem_segs[i].len;
seg->wired = vm->mem_segs[i].wired;
return (0);
}
}
return (-1);
}
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
int
vm_get_memobj(struct vm *vm, vm_paddr_t gpa, size_t len,
vm_offset_t *offset, struct vm_object **object)
{
int i;
size_t seg_len;
vm_paddr_t seg_gpa;
vm_object_t seg_obj;
for (i = 0; i < vm->num_mem_segs; i++) {
if ((seg_obj = vm->mem_segs[i].object) == NULL)
continue;
seg_gpa = vm->mem_segs[i].gpa;
seg_len = vm->mem_segs[i].len;
if (gpa >= seg_gpa && gpa < seg_gpa + seg_len) {
*offset = gpa - seg_gpa;
*object = seg_obj;
vm_object_reference(seg_obj);
return (0);
}
}
return (EINVAL);
}
int
vm_get_register(struct vm *vm, int vcpu, int reg, uint64_t *retval)
{
if (vcpu < 0 || vcpu >= VM_MAXCPU)
return (EINVAL);
if (reg >= VM_REG_LAST)
return (EINVAL);
return (VMGETREG(vm->cookie, vcpu, reg, retval));
}
int
vm_set_register(struct vm *vm, int vcpu, int reg, uint64_t val)
{
if (vcpu < 0 || vcpu >= VM_MAXCPU)
return (EINVAL);
if (reg >= VM_REG_LAST)
return (EINVAL);
return (VMSETREG(vm->cookie, vcpu, reg, val));
}
static boolean_t
is_descriptor_table(int reg)
{
switch (reg) {
case VM_REG_GUEST_IDTR:
case VM_REG_GUEST_GDTR:
return (TRUE);
default:
return (FALSE);
}
}
static boolean_t
is_segment_register(int reg)
{
switch (reg) {
case VM_REG_GUEST_ES:
case VM_REG_GUEST_CS:
case VM_REG_GUEST_SS:
case VM_REG_GUEST_DS:
case VM_REG_GUEST_FS:
case VM_REG_GUEST_GS:
case VM_REG_GUEST_TR:
case VM_REG_GUEST_LDTR:
return (TRUE);
default:
return (FALSE);
}
}
int
vm_get_seg_desc(struct vm *vm, int vcpu, int reg,
struct seg_desc *desc)
{
if (vcpu < 0 || vcpu >= VM_MAXCPU)
return (EINVAL);
if (!is_segment_register(reg) && !is_descriptor_table(reg))
return (EINVAL);
return (VMGETDESC(vm->cookie, vcpu, reg, desc));
}
int
vm_set_seg_desc(struct vm *vm, int vcpu, int reg,
struct seg_desc *desc)
{
if (vcpu < 0 || vcpu >= VM_MAXCPU)
return (EINVAL);
if (!is_segment_register(reg) && !is_descriptor_table(reg))
return (EINVAL);
return (VMSETDESC(vm->cookie, vcpu, reg, desc));
}
static void
restore_guest_fpustate(struct vcpu *vcpu)
{
/* flush host state to the pcb */
fpuexit(curthread);
/* restore guest FPU state */
fpu_stop_emulating();
fpurestore(vcpu->guestfpu);
/* restore guest XCR0 if XSAVE is enabled in the host */
if (rcr4() & CR4_XSAVE)
load_xcr(0, vcpu->guest_xcr0);
/*
* The FPU is now "dirty" with the guest's state so turn on emulation
* to trap any access to the FPU by the host.
*/
fpu_start_emulating();
}
static void
save_guest_fpustate(struct vcpu *vcpu)
{
if ((rcr0() & CR0_TS) == 0)
panic("fpu emulation not enabled in host!");
/* save guest XCR0 and restore host XCR0 */
if (rcr4() & CR4_XSAVE) {
vcpu->guest_xcr0 = rxcr(0);
load_xcr(0, vmm_get_host_xcr0());
}
/* save guest FPU state */
fpu_stop_emulating();
fpusave(vcpu->guestfpu);
fpu_start_emulating();
}
static VMM_STAT(VCPU_IDLE_TICKS, "number of ticks vcpu was idle");
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
static int
vcpu_set_state_locked(struct vcpu *vcpu, enum vcpu_state newstate,
bool from_idle)
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
{
int error;
vcpu_assert_locked(vcpu);
/*
* State transitions from the vmmdev_ioctl() must always begin from
* the VCPU_IDLE state. This guarantees that there is only a single
* ioctl() operating on a vcpu at any point.
*/
if (from_idle) {
while (vcpu->state != VCPU_IDLE)
msleep_spin(&vcpu->state, &vcpu->mtx, "vmstat", hz);
} else {
KASSERT(vcpu->state != VCPU_IDLE, ("invalid transition from "
"vcpu idle state"));
}
if (vcpu->state == VCPU_RUNNING) {
KASSERT(vcpu->hostcpu == curcpu, ("curcpu %d and hostcpu %d "
"mismatch for running vcpu", curcpu, vcpu->hostcpu));
} else {
KASSERT(vcpu->hostcpu == NOCPU, ("Invalid hostcpu %d for a "
"vcpu that is not running", vcpu->hostcpu));
}
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
/*
* The following state transitions are allowed:
* IDLE -> FROZEN -> IDLE
* FROZEN -> RUNNING -> FROZEN
* FROZEN -> SLEEPING -> FROZEN
*/
switch (vcpu->state) {
case VCPU_IDLE:
case VCPU_RUNNING:
case VCPU_SLEEPING:
error = (newstate != VCPU_FROZEN);
break;
case VCPU_FROZEN:
error = (newstate == VCPU_FROZEN);
break;
default:
error = 1;
break;
}
if (error)
return (EBUSY);
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
vcpu->state = newstate;
if (newstate == VCPU_RUNNING)
vcpu->hostcpu = curcpu;
else
vcpu->hostcpu = NOCPU;
if (newstate == VCPU_IDLE)
wakeup(&vcpu->state);
return (0);
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
}
static void
vcpu_require_state(struct vm *vm, int vcpuid, enum vcpu_state newstate)
{
int error;
if ((error = vcpu_set_state(vm, vcpuid, newstate, false)) != 0)
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
panic("Error %d setting state to %d\n", error, newstate);
}
static void
vcpu_require_state_locked(struct vcpu *vcpu, enum vcpu_state newstate)
{
int error;
if ((error = vcpu_set_state_locked(vcpu, newstate, false)) != 0)
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
panic("Error %d setting state to %d", error, newstate);
}
static void
vm_set_rendezvous_func(struct vm *vm, vm_rendezvous_func_t func)
{
KASSERT(mtx_owned(&vm->rendezvous_mtx), ("rendezvous_mtx not locked"));
/*
* Update 'rendezvous_func' and execute a write memory barrier to
* ensure that it is visible across all host cpus. This is not needed
* for correctness but it does ensure that all the vcpus will notice
* that the rendezvous is requested immediately.
*/
vm->rendezvous_func = func;
wmb();
}
#define RENDEZVOUS_CTR0(vm, vcpuid, fmt) \
do { \
if (vcpuid >= 0) \
VCPU_CTR0(vm, vcpuid, fmt); \
else \
VM_CTR0(vm, fmt); \
} while (0)
static void
vm_handle_rendezvous(struct vm *vm, int vcpuid)
{
KASSERT(vcpuid == -1 || (vcpuid >= 0 && vcpuid < VM_MAXCPU),
("vm_handle_rendezvous: invalid vcpuid %d", vcpuid));
mtx_lock(&vm->rendezvous_mtx);
while (vm->rendezvous_func != NULL) {
/* 'rendezvous_req_cpus' must be a subset of 'active_cpus' */
CPU_AND(&vm->rendezvous_req_cpus, &vm->active_cpus);
if (vcpuid != -1 &&
CPU_ISSET(vcpuid, &vm->rendezvous_req_cpus) &&
!CPU_ISSET(vcpuid, &vm->rendezvous_done_cpus)) {
VCPU_CTR0(vm, vcpuid, "Calling rendezvous func");
(*vm->rendezvous_func)(vm, vcpuid, vm->rendezvous_arg);
CPU_SET(vcpuid, &vm->rendezvous_done_cpus);
}
if (CPU_CMP(&vm->rendezvous_req_cpus,
&vm->rendezvous_done_cpus) == 0) {
VCPU_CTR0(vm, vcpuid, "Rendezvous completed");
vm_set_rendezvous_func(vm, NULL);
wakeup(&vm->rendezvous_func);
break;
}
RENDEZVOUS_CTR0(vm, vcpuid, "Wait for rendezvous completion");
mtx_sleep(&vm->rendezvous_func, &vm->rendezvous_mtx, 0,
"vmrndv", 0);
}
mtx_unlock(&vm->rendezvous_mtx);
}
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
/*
* Emulate a guest 'hlt' by sleeping until the vcpu is ready to run.
*/
static int
vm_handle_hlt(struct vm *vm, int vcpuid, bool intr_disabled, bool *retu)
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
{
struct vm_exit *vmexit;
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
struct vcpu *vcpu;
int t, timo, spindown;
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
vcpu = &vm->vcpu[vcpuid];
spindown = 0;
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
vcpu_lock(vcpu);
/*
* Do a final check for pending NMI or interrupts before
* really putting this thread to sleep.
*
* These interrupts could have happened any time after we
* returned from VMRUN() and before we grabbed the vcpu lock.
*/
if (vm->rendezvous_func == NULL &&
!vm_nmi_pending(vm, vcpuid) &&
(intr_disabled || !vlapic_pending_intr(vcpu->vlapic, NULL))) {
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
t = ticks;
vcpu_require_state_locked(vcpu, VCPU_SLEEPING);
if (vlapic_enabled(vcpu->vlapic)) {
/*
* XXX msleep_spin() is not interruptible so use the
* 'timo' to put an upper bound on the sleep time.
*/
timo = hz;
msleep_spin(vcpu, &vcpu->mtx, "vmidle", timo);
} else {
/*
* Spindown the vcpu if the apic is disabled and it
* had entered the halted state.
*/
spindown = 1;
}
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
vcpu_require_state_locked(vcpu, VCPU_FROZEN);
vmm_stat_incr(vm, vcpuid, VCPU_IDLE_TICKS, ticks - t);
}
vcpu_unlock(vcpu);
/*
* Since 'vm_deactivate_cpu()' grabs a sleep mutex we must call it
* outside the confines of the vcpu spinlock.
*/
if (spindown) {
*retu = true;
vmexit = vm_exitinfo(vm, vcpuid);
vmexit->exitcode = VM_EXITCODE_SPINDOWN_CPU;
vm_deactivate_cpu(vm, vcpuid);
VCPU_CTR0(vm, vcpuid, "spinning down cpu");
}
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
return (0);
}
static int
vm_handle_paging(struct vm *vm, int vcpuid, bool *retu)
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
{
int rv, ftype;
struct vm_map *map;
struct vcpu *vcpu;
struct vm_exit *vme;
vcpu = &vm->vcpu[vcpuid];
vme = &vcpu->exitinfo;
ftype = vme->u.paging.fault_type;
KASSERT(ftype == VM_PROT_READ ||
ftype == VM_PROT_WRITE || ftype == VM_PROT_EXECUTE,
("vm_handle_paging: invalid fault_type %d", ftype));
if (ftype == VM_PROT_READ || ftype == VM_PROT_WRITE) {
rv = pmap_emulate_accessed_dirty(vmspace_pmap(vm->vmspace),
vme->u.paging.gpa, ftype);
if (rv == 0)
goto done;
}
map = &vm->vmspace->vm_map;
rv = vm_fault(map, vme->u.paging.gpa, ftype, VM_FAULT_NORMAL);
VCPU_CTR3(vm, vcpuid, "vm_handle_paging rv = %d, gpa = %#lx, "
"ftype = %d", rv, vme->u.paging.gpa, ftype);
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
if (rv != KERN_SUCCESS)
return (EFAULT);
done:
/* restart execution at the faulting instruction */
vme->inst_length = 0;
return (0);
}
static int
vm_handle_inst_emul(struct vm *vm, int vcpuid, bool *retu)
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
{
struct vie *vie;
struct vcpu *vcpu;
struct vm_exit *vme;
int error, inst_length;
uint64_t rip, gla, gpa, cr3;
Add support for FreeBSD/i386 guests under bhyve. - Similar to the hack for bootinfo32.c in userboot, define _MACHINE_ELF_WANT_32BIT in the load_elf32 file handlers in userboot. This allows userboot to load 32-bit kernels and modules. - Copy the SMAP generation code out of bootinfo64.c and into its own file so it can be shared with bootinfo32.c to pass an SMAP to the i386 kernel. - Use uint32_t instead of u_long when aligning module metadata in bootinfo32.c in userboot, as otherwise the metadata used 64-bit alignment which corrupted the layout. - Populate the basemem and extmem members of the bootinfo struct passed to 32-bit kernels. - Fix the 32-bit stack in userboot to start at the top of the stack instead of the bottom so that there is room to grow before the kernel switches to its own stack. - Push a fake return address onto the 32-bit stack in addition to the arguments normally passed to exec() in the loader. This return address is needed to convince recover_bootinfo() in the 32-bit locore code that it is being invoked from a "new" boot block. - Add a routine to libvmmapi to setup a 32-bit flat mode register state including a GDT and TSS that is able to start the i386 kernel and update bhyveload to use it when booting an i386 kernel. - Use the guest register state to determine the CPU's current instruction mode (32-bit vs 64-bit) and paging mode (flat, 32-bit, PAE, or long mode) in the instruction emulation code. Update the gla2gpa() routine used when fetching instructions to handle flat mode, 32-bit paging, and PAE paging in addition to long mode paging. Don't look for a REX prefix when the CPU is in 32-bit mode, and use the detected mode to enable the existing 32-bit mode code when decoding the mod r/m byte. Reviewed by: grehan, neel MFC after: 1 month
2014-02-05 04:39:03 +00:00
enum vie_cpu_mode cpu_mode;
enum vie_paging_mode paging_mode;
mem_region_read_t mread;
mem_region_write_t mwrite;
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
vcpu = &vm->vcpu[vcpuid];
vme = &vcpu->exitinfo;
rip = vme->rip;
inst_length = vme->inst_length;
gla = vme->u.inst_emul.gla;
gpa = vme->u.inst_emul.gpa;
cr3 = vme->u.inst_emul.cr3;
Add support for FreeBSD/i386 guests under bhyve. - Similar to the hack for bootinfo32.c in userboot, define _MACHINE_ELF_WANT_32BIT in the load_elf32 file handlers in userboot. This allows userboot to load 32-bit kernels and modules. - Copy the SMAP generation code out of bootinfo64.c and into its own file so it can be shared with bootinfo32.c to pass an SMAP to the i386 kernel. - Use uint32_t instead of u_long when aligning module metadata in bootinfo32.c in userboot, as otherwise the metadata used 64-bit alignment which corrupted the layout. - Populate the basemem and extmem members of the bootinfo struct passed to 32-bit kernels. - Fix the 32-bit stack in userboot to start at the top of the stack instead of the bottom so that there is room to grow before the kernel switches to its own stack. - Push a fake return address onto the 32-bit stack in addition to the arguments normally passed to exec() in the loader. This return address is needed to convince recover_bootinfo() in the 32-bit locore code that it is being invoked from a "new" boot block. - Add a routine to libvmmapi to setup a 32-bit flat mode register state including a GDT and TSS that is able to start the i386 kernel and update bhyveload to use it when booting an i386 kernel. - Use the guest register state to determine the CPU's current instruction mode (32-bit vs 64-bit) and paging mode (flat, 32-bit, PAE, or long mode) in the instruction emulation code. Update the gla2gpa() routine used when fetching instructions to handle flat mode, 32-bit paging, and PAE paging in addition to long mode paging. Don't look for a REX prefix when the CPU is in 32-bit mode, and use the detected mode to enable the existing 32-bit mode code when decoding the mod r/m byte. Reviewed by: grehan, neel MFC after: 1 month
2014-02-05 04:39:03 +00:00
cpu_mode = vme->u.inst_emul.cpu_mode;
paging_mode = vme->u.inst_emul.paging_mode;
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
vie = &vme->u.inst_emul.vie;
vie_init(vie);
/* Fetch, decode and emulate the faulting instruction */
Add support for FreeBSD/i386 guests under bhyve. - Similar to the hack for bootinfo32.c in userboot, define _MACHINE_ELF_WANT_32BIT in the load_elf32 file handlers in userboot. This allows userboot to load 32-bit kernels and modules. - Copy the SMAP generation code out of bootinfo64.c and into its own file so it can be shared with bootinfo32.c to pass an SMAP to the i386 kernel. - Use uint32_t instead of u_long when aligning module metadata in bootinfo32.c in userboot, as otherwise the metadata used 64-bit alignment which corrupted the layout. - Populate the basemem and extmem members of the bootinfo struct passed to 32-bit kernels. - Fix the 32-bit stack in userboot to start at the top of the stack instead of the bottom so that there is room to grow before the kernel switches to its own stack. - Push a fake return address onto the 32-bit stack in addition to the arguments normally passed to exec() in the loader. This return address is needed to convince recover_bootinfo() in the 32-bit locore code that it is being invoked from a "new" boot block. - Add a routine to libvmmapi to setup a 32-bit flat mode register state including a GDT and TSS that is able to start the i386 kernel and update bhyveload to use it when booting an i386 kernel. - Use the guest register state to determine the CPU's current instruction mode (32-bit vs 64-bit) and paging mode (flat, 32-bit, PAE, or long mode) in the instruction emulation code. Update the gla2gpa() routine used when fetching instructions to handle flat mode, 32-bit paging, and PAE paging in addition to long mode paging. Don't look for a REX prefix when the CPU is in 32-bit mode, and use the detected mode to enable the existing 32-bit mode code when decoding the mod r/m byte. Reviewed by: grehan, neel MFC after: 1 month
2014-02-05 04:39:03 +00:00
if (vmm_fetch_instruction(vm, vcpuid, rip, inst_length, cr3,
paging_mode, vie) != 0)
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
return (EFAULT);
Add support for FreeBSD/i386 guests under bhyve. - Similar to the hack for bootinfo32.c in userboot, define _MACHINE_ELF_WANT_32BIT in the load_elf32 file handlers in userboot. This allows userboot to load 32-bit kernels and modules. - Copy the SMAP generation code out of bootinfo64.c and into its own file so it can be shared with bootinfo32.c to pass an SMAP to the i386 kernel. - Use uint32_t instead of u_long when aligning module metadata in bootinfo32.c in userboot, as otherwise the metadata used 64-bit alignment which corrupted the layout. - Populate the basemem and extmem members of the bootinfo struct passed to 32-bit kernels. - Fix the 32-bit stack in userboot to start at the top of the stack instead of the bottom so that there is room to grow before the kernel switches to its own stack. - Push a fake return address onto the 32-bit stack in addition to the arguments normally passed to exec() in the loader. This return address is needed to convince recover_bootinfo() in the 32-bit locore code that it is being invoked from a "new" boot block. - Add a routine to libvmmapi to setup a 32-bit flat mode register state including a GDT and TSS that is able to start the i386 kernel and update bhyveload to use it when booting an i386 kernel. - Use the guest register state to determine the CPU's current instruction mode (32-bit vs 64-bit) and paging mode (flat, 32-bit, PAE, or long mode) in the instruction emulation code. Update the gla2gpa() routine used when fetching instructions to handle flat mode, 32-bit paging, and PAE paging in addition to long mode paging. Don't look for a REX prefix when the CPU is in 32-bit mode, and use the detected mode to enable the existing 32-bit mode code when decoding the mod r/m byte. Reviewed by: grehan, neel MFC after: 1 month
2014-02-05 04:39:03 +00:00
if (vmm_decode_instruction(vm, vcpuid, gla, cpu_mode, vie) != 0)
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
return (EFAULT);
/* return to userland unless this is an in-kernel emulated device */
if (gpa >= DEFAULT_APIC_BASE && gpa < DEFAULT_APIC_BASE + PAGE_SIZE) {
mread = lapic_mmio_read;
mwrite = lapic_mmio_write;
} else if (gpa >= VIOAPIC_BASE && gpa < VIOAPIC_BASE + VIOAPIC_SIZE) {
mread = vioapic_mmio_read;
mwrite = vioapic_mmio_write;
} else if (gpa >= VHPET_BASE && gpa < VHPET_BASE + VHPET_SIZE) {
mread = vhpet_mmio_read;
mwrite = vhpet_mmio_write;
} else {
*retu = true;
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
return (0);
}
error = vmm_emulate_instruction(vm, vcpuid, gpa, vie, mread, mwrite,
retu);
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
return (error);
}
static int
vm_handle_suspend(struct vm *vm, int vcpuid, bool *retu)
{
int i, done;
struct vcpu *vcpu;
done = 0;
vcpu = &vm->vcpu[vcpuid];
CPU_SET_ATOMIC(vcpuid, &vm->suspended_cpus);
/*
* Wait until all 'active_cpus' have suspended themselves.
*
* Since a VM may be suspended at any time including when one or
* more vcpus are doing a rendezvous we need to call the rendezvous
* handler while we are waiting to prevent a deadlock.
*/
vcpu_lock(vcpu);
while (1) {
if (CPU_CMP(&vm->suspended_cpus, &vm->active_cpus) == 0) {
VCPU_CTR0(vm, vcpuid, "All vcpus suspended");
break;
}
if (vm->rendezvous_func == NULL) {
VCPU_CTR0(vm, vcpuid, "Sleeping during suspend");
vcpu_require_state_locked(vcpu, VCPU_SLEEPING);
msleep_spin(vcpu, &vcpu->mtx, "vmsusp", hz);
vcpu_require_state_locked(vcpu, VCPU_FROZEN);
} else {
VCPU_CTR0(vm, vcpuid, "Rendezvous during suspend");
vcpu_unlock(vcpu);
vm_handle_rendezvous(vm, vcpuid);
vcpu_lock(vcpu);
}
}
vcpu_unlock(vcpu);
/*
* Wakeup the other sleeping vcpus and return to userspace.
*/
for (i = 0; i < VM_MAXCPU; i++) {
if (CPU_ISSET(i, &vm->suspended_cpus)) {
vcpu_notify_event(vm, i, false);
}
}
*retu = true;
return (0);
}
int
vm_suspend(struct vm *vm)
{
if (atomic_cmpset_int(&vm->suspend, 0, 1)) {
VM_CTR0(vm, "virtual machine suspended");
return (0);
} else {
VM_CTR0(vm, "virtual machine already suspended");
return (EALREADY);
}
}
int
vm_run(struct vm *vm, struct vm_run *vmrun)
{
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
int error, vcpuid;
struct vcpu *vcpu;
struct pcb *pcb;
uint64_t tscval, rip;
struct vm_exit *vme;
bool retu, intr_disabled;
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
pmap_t pmap;
void *rptr, *sptr;
vcpuid = vmrun->cpuid;
if (vcpuid < 0 || vcpuid >= VM_MAXCPU)
return (EINVAL);
rptr = &vm->rendezvous_func;
sptr = &vm->suspend;
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
pmap = vmspace_pmap(vm->vmspace);
vcpu = &vm->vcpu[vcpuid];
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
vme = &vcpu->exitinfo;
rip = vmrun->rip;
restart:
critical_enter();
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
KASSERT(!CPU_ISSET(curcpu, &pmap->pm_active),
("vm_run: absurd pm_active"));
tscval = rdtsc();
pcb = PCPU_GET(curpcb);
set_pcb_flags(pcb, PCB_FULL_IRET);
restore_guest_msrs(vm, vcpuid);
restore_guest_fpustate(vcpu);
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
vcpu_require_state(vm, vcpuid, VCPU_RUNNING);
error = VMRUN(vm->cookie, vcpuid, rip, pmap, rptr, sptr);
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
vcpu_require_state(vm, vcpuid, VCPU_FROZEN);
save_guest_fpustate(vcpu);
restore_host_msrs(vm, vcpuid);
vmm_stat_incr(vm, vcpuid, VCPU_TOTAL_RUNTIME, rdtsc() - tscval);
critical_exit();
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
if (error == 0) {
retu = false;
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
switch (vme->exitcode) {
case VM_EXITCODE_SUSPENDED:
error = vm_handle_suspend(vm, vcpuid, &retu);
break;
case VM_EXITCODE_IOAPIC_EOI:
vioapic_process_eoi(vm, vcpuid,
vme->u.ioapic_eoi.vector);
break;
case VM_EXITCODE_RENDEZVOUS:
vm_handle_rendezvous(vm, vcpuid);
error = 0;
break;
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
case VM_EXITCODE_HLT:
intr_disabled = ((vme->u.hlt.rflags & PSL_I) == 0);
error = vm_handle_hlt(vm, vcpuid, intr_disabled, &retu);
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
break;
case VM_EXITCODE_PAGING:
error = vm_handle_paging(vm, vcpuid, &retu);
break;
case VM_EXITCODE_INST_EMUL:
error = vm_handle_inst_emul(vm, vcpuid, &retu);
break;
default:
retu = true; /* handled in userland */
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
break;
}
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
}
if (error == 0 && retu == false) {
rip = vme->rip + vme->inst_length;
goto restart;
}
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
/* copy the exit information */
bcopy(vme, &vmrun->vm_exit, sizeof(struct vm_exit));
return (error);
}
int
vm_inject_exception(struct vm *vm, int vcpuid, struct vm_exception *exception)
{
struct vcpu *vcpu;
if (vcpuid < 0 || vcpuid >= VM_MAXCPU)
return (EINVAL);
if (exception->vector < 0 || exception->vector >= 32)
return (EINVAL);
vcpu = &vm->vcpu[vcpuid];
if (vcpu->exception_pending) {
VCPU_CTR2(vm, vcpuid, "Unable to inject exception %d due to "
"pending exception %d", exception->vector,
vcpu->exception.vector);
return (EBUSY);
}
vcpu->exception_pending = 1;
vcpu->exception = *exception;
VCPU_CTR1(vm, vcpuid, "Exception %d pending", exception->vector);
return (0);
}
int
vm_exception_pending(struct vm *vm, int vcpuid, struct vm_exception *exception)
{
struct vcpu *vcpu;
int pending;
KASSERT(vcpuid >= 0 && vcpuid < VM_MAXCPU, ("invalid vcpu %d", vcpuid));
vcpu = &vm->vcpu[vcpuid];
pending = vcpu->exception_pending;
if (pending) {
vcpu->exception_pending = 0;
*exception = vcpu->exception;
VCPU_CTR1(vm, vcpuid, "Exception %d delivered",
exception->vector);
}
return (pending);
}
static void
vm_inject_fault(struct vm *vm, int vcpuid, struct vm_exception *exception)
{
struct vm_exit *vmexit;
int error;
error = vm_inject_exception(vm, vcpuid, exception);
KASSERT(error == 0, ("vm_inject_exception error %d", error));
/*
* A fault-like exception allows the instruction to be restarted
* after the exception handler returns.
*
* By setting the inst_length to 0 we ensure that the instruction
* pointer remains at the faulting instruction.
*/
vmexit = vm_exitinfo(vm, vcpuid);
vmexit->inst_length = 0;
}
void
vm_inject_gp(struct vm *vm, int vcpuid)
{
struct vm_exception gpf = {
.vector = IDT_GP,
.error_code_valid = 1,
.error_code = 0
};
vm_inject_fault(vm, vcpuid, &gpf);
}
void
vm_inject_ud(struct vm *vm, int vcpuid)
{
struct vm_exception udf = {
.vector = IDT_UD,
.error_code_valid = 0
};
vm_inject_fault(vm, vcpuid, &udf);
}
static VMM_STAT(VCPU_NMI_COUNT, "number of NMIs delivered to vcpu");
int
vm_inject_nmi(struct vm *vm, int vcpuid)
{
struct vcpu *vcpu;
if (vcpuid < 0 || vcpuid >= VM_MAXCPU)
return (EINVAL);
vcpu = &vm->vcpu[vcpuid];
vcpu->nmi_pending = 1;
vcpu_notify_event(vm, vcpuid, false);
return (0);
}
int
vm_nmi_pending(struct vm *vm, int vcpuid)
{
struct vcpu *vcpu;
if (vcpuid < 0 || vcpuid >= VM_MAXCPU)
panic("vm_nmi_pending: invalid vcpuid %d", vcpuid);
vcpu = &vm->vcpu[vcpuid];
return (vcpu->nmi_pending);
}
void
vm_nmi_clear(struct vm *vm, int vcpuid)
{
struct vcpu *vcpu;
if (vcpuid < 0 || vcpuid >= VM_MAXCPU)
panic("vm_nmi_pending: invalid vcpuid %d", vcpuid);
vcpu = &vm->vcpu[vcpuid];
if (vcpu->nmi_pending == 0)
panic("vm_nmi_clear: inconsistent nmi_pending state");
vcpu->nmi_pending = 0;
vmm_stat_incr(vm, vcpuid, VCPU_NMI_COUNT, 1);
}
static VMM_STAT(VCPU_EXTINT_COUNT, "number of ExtINTs delivered to vcpu");
int
vm_inject_extint(struct vm *vm, int vcpuid)
{
struct vcpu *vcpu;
if (vcpuid < 0 || vcpuid >= VM_MAXCPU)
return (EINVAL);
vcpu = &vm->vcpu[vcpuid];
vcpu->extint_pending = 1;
vcpu_notify_event(vm, vcpuid, false);
return (0);
}
int
vm_extint_pending(struct vm *vm, int vcpuid)
{
struct vcpu *vcpu;
if (vcpuid < 0 || vcpuid >= VM_MAXCPU)
panic("vm_extint_pending: invalid vcpuid %d", vcpuid);
vcpu = &vm->vcpu[vcpuid];
return (vcpu->extint_pending);
}
void
vm_extint_clear(struct vm *vm, int vcpuid)
{
struct vcpu *vcpu;
if (vcpuid < 0 || vcpuid >= VM_MAXCPU)
panic("vm_extint_pending: invalid vcpuid %d", vcpuid);
vcpu = &vm->vcpu[vcpuid];
if (vcpu->extint_pending == 0)
panic("vm_extint_clear: inconsistent extint_pending state");
vcpu->extint_pending = 0;
vmm_stat_incr(vm, vcpuid, VCPU_EXTINT_COUNT, 1);
}
int
vm_get_capability(struct vm *vm, int vcpu, int type, int *retval)
{
if (vcpu < 0 || vcpu >= VM_MAXCPU)
return (EINVAL);
if (type < 0 || type >= VM_CAP_MAX)
return (EINVAL);
return (VMGETCAP(vm->cookie, vcpu, type, retval));
}
int
vm_set_capability(struct vm *vm, int vcpu, int type, int val)
{
if (vcpu < 0 || vcpu >= VM_MAXCPU)
return (EINVAL);
if (type < 0 || type >= VM_CAP_MAX)
return (EINVAL);
return (VMSETCAP(vm->cookie, vcpu, type, val));
}
uint64_t *
vm_guest_msrs(struct vm *vm, int cpu)
{
return (vm->vcpu[cpu].guest_msrs);
}
struct vlapic *
vm_lapic(struct vm *vm, int cpu)
{
return (vm->vcpu[cpu].vlapic);
}
struct vioapic *
vm_ioapic(struct vm *vm)
{
return (vm->vioapic);
}
struct vhpet *
vm_hpet(struct vm *vm)
{
return (vm->vhpet);
}
boolean_t
vmm_is_pptdev(int bus, int slot, int func)
{
int found, i, n;
int b, s, f;
char *val, *cp, *cp2;
/*
* XXX
* The length of an environment variable is limited to 128 bytes which
* puts an upper limit on the number of passthru devices that may be
* specified using a single environment variable.
*
* Work around this by scanning multiple environment variable
* names instead of a single one - yuck!
*/
const char *names[] = { "pptdevs", "pptdevs2", "pptdevs3", NULL };
/* set pptdevs="1/2/3 4/5/6 7/8/9 10/11/12" */
found = 0;
for (i = 0; names[i] != NULL && !found; i++) {
cp = val = getenv(names[i]);
while (cp != NULL && *cp != '\0') {
if ((cp2 = strchr(cp, ' ')) != NULL)
*cp2 = '\0';
n = sscanf(cp, "%d/%d/%d", &b, &s, &f);
if (n == 3 && bus == b && slot == s && func == f) {
found = 1;
break;
}
if (cp2 != NULL)
*cp2++ = ' ';
cp = cp2;
}
freeenv(val);
}
return (found);
}
void *
vm_iommu_domain(struct vm *vm)
{
return (vm->iommu);
}
int
vcpu_set_state(struct vm *vm, int vcpuid, enum vcpu_state newstate,
bool from_idle)
{
int error;
struct vcpu *vcpu;
if (vcpuid < 0 || vcpuid >= VM_MAXCPU)
panic("vm_set_run_state: invalid vcpuid %d", vcpuid);
vcpu = &vm->vcpu[vcpuid];
vcpu_lock(vcpu);
error = vcpu_set_state_locked(vcpu, newstate, from_idle);
vcpu_unlock(vcpu);
return (error);
}
enum vcpu_state
vcpu_get_state(struct vm *vm, int vcpuid, int *hostcpu)
{
struct vcpu *vcpu;
enum vcpu_state state;
if (vcpuid < 0 || vcpuid >= VM_MAXCPU)
panic("vm_get_run_state: invalid vcpuid %d", vcpuid);
vcpu = &vm->vcpu[vcpuid];
vcpu_lock(vcpu);
state = vcpu->state;
if (hostcpu != NULL)
*hostcpu = vcpu->hostcpu;
vcpu_unlock(vcpu);
return (state);
}
void
vm_activate_cpu(struct vm *vm, int vcpuid)
{
KASSERT(vcpuid >= 0 && vcpuid < VM_MAXCPU,
("vm_activate_cpu: invalid vcpuid %d", vcpuid));
KASSERT(!CPU_ISSET(vcpuid, &vm->active_cpus),
("vm_activate_cpu: vcpuid %d is already active", vcpuid));
VCPU_CTR0(vm, vcpuid, "activated");
CPU_SET_ATOMIC(vcpuid, &vm->active_cpus);
}
static void
vm_deactivate_cpu(struct vm *vm, int vcpuid)
{
KASSERT(vcpuid >= 0 && vcpuid < VM_MAXCPU,
("vm_deactivate_cpu: invalid vcpuid %d", vcpuid));
KASSERT(CPU_ISSET(vcpuid, &vm->active_cpus),
("vm_deactivate_cpu: vcpuid %d is not active", vcpuid));
VCPU_CTR0(vm, vcpuid, "deactivated");
CPU_CLR_ATOMIC(vcpuid, &vm->active_cpus);
/*
* If a vcpu rendezvous is in progress then it could be blocked
* on 'vcpuid' - unblock it before disappearing forever.
*/
mtx_lock(&vm->rendezvous_mtx);
if (vm->rendezvous_func != NULL) {
VCPU_CTR0(vm, vcpuid, "unblock rendezvous after deactivation");
wakeup(&vm->rendezvous_func);
}
mtx_unlock(&vm->rendezvous_mtx);
}
2011-06-28 06:26:03 +00:00
cpuset_t
vm_active_cpus(struct vm *vm)
{
return (vm->active_cpus);
}
void *
vcpu_stats(struct vm *vm, int vcpuid)
{
return (vm->vcpu[vcpuid].stats);
}
int
vm_get_x2apic_state(struct vm *vm, int vcpuid, enum x2apic_state *state)
{
if (vcpuid < 0 || vcpuid >= VM_MAXCPU)
return (EINVAL);
*state = vm->vcpu[vcpuid].x2apic_state;
return (0);
}
int
vm_set_x2apic_state(struct vm *vm, int vcpuid, enum x2apic_state state)
{
if (vcpuid < 0 || vcpuid >= VM_MAXCPU)
return (EINVAL);
if (state >= X2APIC_STATE_LAST)
return (EINVAL);
vm->vcpu[vcpuid].x2apic_state = state;
vlapic_set_x2apic_state(vm, vcpuid, state);
return (0);
}
/*
* This function is called to ensure that a vcpu "sees" a pending event
* as soon as possible:
* - If the vcpu thread is sleeping then it is woken up.
* - If the vcpu is running on a different host_cpu then an IPI will be directed
* to the host_cpu to cause the vcpu to trap into the hypervisor.
*/
void
vcpu_notify_event(struct vm *vm, int vcpuid, bool lapic_intr)
{
int hostcpu;
struct vcpu *vcpu;
vcpu = &vm->vcpu[vcpuid];
vcpu_lock(vcpu);
hostcpu = vcpu->hostcpu;
if (vcpu->state == VCPU_RUNNING) {
KASSERT(hostcpu != NOCPU, ("vcpu running on invalid hostcpu"));
if (hostcpu != curcpu) {
if (lapic_intr) {
2014-01-09 03:25:54 +00:00
vlapic_post_intr(vcpu->vlapic, hostcpu,
vmm_ipinum);
} else {
ipi_cpu(hostcpu, vmm_ipinum);
}
} else {
/*
* If the 'vcpu' is running on 'curcpu' then it must
* be sending a notification to itself (e.g. SELF_IPI).
* The pending event will be picked up when the vcpu
* transitions back to guest context.
*/
}
} else {
KASSERT(hostcpu == NOCPU, ("vcpu state %d not consistent "
"with hostcpu %d", vcpu->state, hostcpu));
if (vcpu->state == VCPU_SLEEPING)
wakeup_one(vcpu);
}
vcpu_unlock(vcpu);
}
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
struct vmspace *
vm_get_vmspace(struct vm *vm)
{
return (vm->vmspace);
}
int
vm_apicid2vcpuid(struct vm *vm, int apicid)
{
/*
* XXX apic id is assumed to be numerically identical to vcpu id
*/
return (apicid);
}
void
vm_smp_rendezvous(struct vm *vm, int vcpuid, cpuset_t dest,
vm_rendezvous_func_t func, void *arg)
{
int i;
/*
* Enforce that this function is called without any locks
*/
WITNESS_WARN(WARN_PANIC, NULL, "vm_smp_rendezvous");
KASSERT(vcpuid == -1 || (vcpuid >= 0 && vcpuid < VM_MAXCPU),
("vm_smp_rendezvous: invalid vcpuid %d", vcpuid));
restart:
mtx_lock(&vm->rendezvous_mtx);
if (vm->rendezvous_func != NULL) {
/*
* If a rendezvous is already in progress then we need to
* call the rendezvous handler in case this 'vcpuid' is one
* of the targets of the rendezvous.
*/
RENDEZVOUS_CTR0(vm, vcpuid, "Rendezvous already in progress");
mtx_unlock(&vm->rendezvous_mtx);
vm_handle_rendezvous(vm, vcpuid);
goto restart;
}
KASSERT(vm->rendezvous_func == NULL, ("vm_smp_rendezvous: previous "
"rendezvous is still in progress"));
RENDEZVOUS_CTR0(vm, vcpuid, "Initiating rendezvous");
vm->rendezvous_req_cpus = dest;
CPU_ZERO(&vm->rendezvous_done_cpus);
vm->rendezvous_arg = arg;
vm_set_rendezvous_func(vm, func);
mtx_unlock(&vm->rendezvous_mtx);
/*
* Wake up any sleeping vcpus and trigger a VM-exit in any running
* vcpus so they handle the rendezvous as soon as possible.
*/
for (i = 0; i < VM_MAXCPU; i++) {
if (CPU_ISSET(i, &dest))
vcpu_notify_event(vm, i, false);
}
vm_handle_rendezvous(vm, vcpuid);
}
struct vatpic *
vm_atpic(struct vm *vm)
{
return (vm->vatpic);
}
struct vatpit *
vm_atpit(struct vm *vm)
{
return (vm->vatpit);
}