2008-04-17 04:28:09 +00:00
|
|
|
/*
|
|
|
|
* This program is free software; you can redistribute it and/or modify
|
|
|
|
* it under the terms of the GNU General Public License, version 2, as
|
|
|
|
* published by the Free Software Foundation.
|
|
|
|
*
|
|
|
|
* This program is distributed in the hope that it will be useful,
|
|
|
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
|
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|
|
|
* GNU General Public License for more details.
|
|
|
|
*
|
|
|
|
* You should have received a copy of the GNU General Public License
|
|
|
|
* along with this program; if not, write to the Free Software
|
|
|
|
* Foundation, 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
|
|
|
|
*
|
|
|
|
* Copyright IBM Corp. 2007
|
|
|
|
*
|
|
|
|
* Authors: Hollis Blanchard <hollisb@us.ibm.com>
|
|
|
|
*/
|
|
|
|
|
|
|
|
#ifndef __POWERPC_KVM_HOST_H__
|
|
|
|
#define __POWERPC_KVM_HOST_H__
|
|
|
|
|
|
|
|
#include <linux/mutex.h>
|
2009-11-02 12:02:31 +00:00
|
|
|
#include <linux/hrtimer.h>
|
|
|
|
#include <linux/interrupt.h>
|
2008-04-17 04:28:09 +00:00
|
|
|
#include <linux/types.h>
|
|
|
|
#include <linux/kvm_types.h>
|
2010-07-29 12:47:42 +00:00
|
|
|
#include <linux/kvm_para.h>
|
2008-04-17 04:28:09 +00:00
|
|
|
#include <asm/kvm_asm.h>
|
|
|
|
|
|
|
|
#define KVM_MAX_VCPUS 1
|
|
|
|
#define KVM_MEMORY_SLOTS 32
|
|
|
|
/* memory slots that does not exposed to userspace */
|
|
|
|
#define KVM_PRIVATE_MEM_SLOTS 4
|
|
|
|
|
KVM: PPC: Add support for Book3S processors in hypervisor mode
This adds support for KVM running on 64-bit Book 3S processors,
specifically POWER7, in hypervisor mode. Using hypervisor mode means
that the guest can use the processor's supervisor mode. That means
that the guest can execute privileged instructions and access privileged
registers itself without trapping to the host. This gives excellent
performance, but does mean that KVM cannot emulate a processor
architecture other than the one that the hardware implements.
This code assumes that the guest is running paravirtualized using the
PAPR (Power Architecture Platform Requirements) interface, which is the
interface that IBM's PowerVM hypervisor uses. That means that existing
Linux distributions that run on IBM pSeries machines will also run
under KVM without modification. In order to communicate the PAPR
hypercalls to qemu, this adds a new KVM_EXIT_PAPR_HCALL exit code
to include/linux/kvm.h.
Currently the choice between book3s_hv support and book3s_pr support
(i.e. the existing code, which runs the guest in user mode) has to be
made at kernel configuration time, so a given kernel binary can only
do one or the other.
This new book3s_hv code doesn't support MMIO emulation at present.
Since we are running paravirtualized guests, this isn't a serious
restriction.
With the guest running in supervisor mode, most exceptions go straight
to the guest. We will never get data or instruction storage or segment
interrupts, alignment interrupts, decrementer interrupts, program
interrupts, single-step interrupts, etc., coming to the hypervisor from
the guest. Therefore this introduces a new KVMTEST_NONHV macro for the
exception entry path so that we don't have to do the KVM test on entry
to those exception handlers.
We do however get hypervisor decrementer, hypervisor data storage,
hypervisor instruction storage, and hypervisor emulation assist
interrupts, so we have to handle those.
In hypervisor mode, real-mode accesses can access all of RAM, not just
a limited amount. Therefore we put all the guest state in the vcpu.arch
and use the shadow_vcpu in the PACA only for temporary scratch space.
We allocate the vcpu with kzalloc rather than vzalloc, and we don't use
anything in the kvmppc_vcpu_book3s struct, so we don't allocate it.
We don't have a shared page with the guest, but we still need a
kvm_vcpu_arch_shared struct to store the values of various registers,
so we include one in the vcpu_arch struct.
The POWER7 processor has a restriction that all threads in a core have
to be in the same partition. MMU-on kernel code counts as a partition
(partition 0), so we have to do a partition switch on every entry to and
exit from the guest. At present we require the host and guest to run
in single-thread mode because of this hardware restriction.
This code allocates a hashed page table for the guest and initializes
it with HPTEs for the guest's Virtual Real Memory Area (VRMA). We
require that the guest memory is allocated using 16MB huge pages, in
order to simplify the low-level memory management. This also means that
we can get away without tracking paging activity in the host for now,
since huge pages can't be paged or swapped.
This also adds a few new exports needed by the book3s_hv code.
Signed-off-by: Paul Mackerras <paulus@samba.org>
Signed-off-by: Alexander Graf <agraf@suse.de>
2011-06-29 00:21:34 +00:00
|
|
|
#ifdef CONFIG_KVM_MMIO
|
2008-05-30 14:05:56 +00:00
|
|
|
#define KVM_COALESCED_MMIO_PAGE_OFFSET 1
|
KVM: PPC: Add support for Book3S processors in hypervisor mode
This adds support for KVM running on 64-bit Book 3S processors,
specifically POWER7, in hypervisor mode. Using hypervisor mode means
that the guest can use the processor's supervisor mode. That means
that the guest can execute privileged instructions and access privileged
registers itself without trapping to the host. This gives excellent
performance, but does mean that KVM cannot emulate a processor
architecture other than the one that the hardware implements.
This code assumes that the guest is running paravirtualized using the
PAPR (Power Architecture Platform Requirements) interface, which is the
interface that IBM's PowerVM hypervisor uses. That means that existing
Linux distributions that run on IBM pSeries machines will also run
under KVM without modification. In order to communicate the PAPR
hypercalls to qemu, this adds a new KVM_EXIT_PAPR_HCALL exit code
to include/linux/kvm.h.
Currently the choice between book3s_hv support and book3s_pr support
(i.e. the existing code, which runs the guest in user mode) has to be
made at kernel configuration time, so a given kernel binary can only
do one or the other.
This new book3s_hv code doesn't support MMIO emulation at present.
Since we are running paravirtualized guests, this isn't a serious
restriction.
With the guest running in supervisor mode, most exceptions go straight
to the guest. We will never get data or instruction storage or segment
interrupts, alignment interrupts, decrementer interrupts, program
interrupts, single-step interrupts, etc., coming to the hypervisor from
the guest. Therefore this introduces a new KVMTEST_NONHV macro for the
exception entry path so that we don't have to do the KVM test on entry
to those exception handlers.
We do however get hypervisor decrementer, hypervisor data storage,
hypervisor instruction storage, and hypervisor emulation assist
interrupts, so we have to handle those.
In hypervisor mode, real-mode accesses can access all of RAM, not just
a limited amount. Therefore we put all the guest state in the vcpu.arch
and use the shadow_vcpu in the PACA only for temporary scratch space.
We allocate the vcpu with kzalloc rather than vzalloc, and we don't use
anything in the kvmppc_vcpu_book3s struct, so we don't allocate it.
We don't have a shared page with the guest, but we still need a
kvm_vcpu_arch_shared struct to store the values of various registers,
so we include one in the vcpu_arch struct.
The POWER7 processor has a restriction that all threads in a core have
to be in the same partition. MMU-on kernel code counts as a partition
(partition 0), so we have to do a partition switch on every entry to and
exit from the guest. At present we require the host and guest to run
in single-thread mode because of this hardware restriction.
This code allocates a hashed page table for the guest and initializes
it with HPTEs for the guest's Virtual Real Memory Area (VRMA). We
require that the guest memory is allocated using 16MB huge pages, in
order to simplify the low-level memory management. This also means that
we can get away without tracking paging activity in the host for now,
since huge pages can't be paged or swapped.
This also adds a few new exports needed by the book3s_hv code.
Signed-off-by: Paul Mackerras <paulus@samba.org>
Signed-off-by: Alexander Graf <agraf@suse.de>
2011-06-29 00:21:34 +00:00
|
|
|
#endif
|
2008-05-30 14:05:56 +00:00
|
|
|
|
2008-04-17 04:28:09 +00:00
|
|
|
/* We don't currently support large pages. */
|
2010-07-01 14:00:11 +00:00
|
|
|
#define KVM_HPAGE_GFN_SHIFT(x) 0
|
2009-06-19 13:16:23 +00:00
|
|
|
#define KVM_NR_PAGE_SIZES 1
|
|
|
|
#define KVM_PAGES_PER_HPAGE(x) (1UL<<31)
|
2008-04-17 04:28:09 +00:00
|
|
|
|
2010-06-30 13:18:46 +00:00
|
|
|
#define HPTEG_CACHE_NUM (1 << 15)
|
|
|
|
#define HPTEG_HASH_BITS_PTE 13
|
2010-07-29 13:04:19 +00:00
|
|
|
#define HPTEG_HASH_BITS_PTE_LONG 12
|
2010-06-30 13:18:46 +00:00
|
|
|
#define HPTEG_HASH_BITS_VPTE 13
|
|
|
|
#define HPTEG_HASH_BITS_VPTE_LONG 5
|
|
|
|
#define HPTEG_HASH_NUM_PTE (1 << HPTEG_HASH_BITS_PTE)
|
2010-07-29 13:04:19 +00:00
|
|
|
#define HPTEG_HASH_NUM_PTE_LONG (1 << HPTEG_HASH_BITS_PTE_LONG)
|
2010-06-30 13:18:46 +00:00
|
|
|
#define HPTEG_HASH_NUM_VPTE (1 << HPTEG_HASH_BITS_VPTE)
|
|
|
|
#define HPTEG_HASH_NUM_VPTE_LONG (1 << HPTEG_HASH_BITS_VPTE_LONG)
|
2009-10-30 05:47:04 +00:00
|
|
|
|
2010-07-29 12:47:52 +00:00
|
|
|
/* Physical Address Mask - allowed range of real mode RAM access */
|
|
|
|
#define KVM_PAM 0x0fffffffffffffffULL
|
|
|
|
|
2008-04-17 04:28:09 +00:00
|
|
|
struct kvm;
|
|
|
|
struct kvm_run;
|
|
|
|
struct kvm_vcpu;
|
|
|
|
|
2011-06-29 00:22:05 +00:00
|
|
|
struct lppaca;
|
|
|
|
struct slb_shadow;
|
|
|
|
struct dtl;
|
|
|
|
|
2008-04-17 04:28:09 +00:00
|
|
|
struct kvm_vm_stat {
|
|
|
|
u32 remote_tlb_flush;
|
|
|
|
};
|
|
|
|
|
|
|
|
struct kvm_vcpu_stat {
|
|
|
|
u32 sum_exits;
|
|
|
|
u32 mmio_exits;
|
|
|
|
u32 dcr_exits;
|
|
|
|
u32 signal_exits;
|
|
|
|
u32 light_exits;
|
|
|
|
/* Account for special types of light exits: */
|
|
|
|
u32 itlb_real_miss_exits;
|
|
|
|
u32 itlb_virt_miss_exits;
|
|
|
|
u32 dtlb_real_miss_exits;
|
|
|
|
u32 dtlb_virt_miss_exits;
|
|
|
|
u32 syscall_exits;
|
|
|
|
u32 isi_exits;
|
|
|
|
u32 dsi_exits;
|
|
|
|
u32 emulated_inst_exits;
|
|
|
|
u32 dec_exits;
|
|
|
|
u32 ext_intr_exits;
|
2008-04-25 22:55:49 +00:00
|
|
|
u32 halt_wakeup;
|
2010-04-15 22:11:42 +00:00
|
|
|
#ifdef CONFIG_PPC_BOOK3S
|
2009-10-30 05:47:04 +00:00
|
|
|
u32 pf_storage;
|
|
|
|
u32 pf_instruc;
|
|
|
|
u32 sp_storage;
|
|
|
|
u32 sp_instruc;
|
|
|
|
u32 queue_intr;
|
|
|
|
u32 ld;
|
|
|
|
u32 ld_slow;
|
|
|
|
u32 st;
|
|
|
|
u32 st_slow;
|
|
|
|
#endif
|
2008-04-17 04:28:09 +00:00
|
|
|
};
|
|
|
|
|
2008-12-02 21:51:57 +00:00
|
|
|
enum kvm_exit_types {
|
|
|
|
MMIO_EXITS,
|
|
|
|
DCR_EXITS,
|
|
|
|
SIGNAL_EXITS,
|
|
|
|
ITLB_REAL_MISS_EXITS,
|
|
|
|
ITLB_VIRT_MISS_EXITS,
|
|
|
|
DTLB_REAL_MISS_EXITS,
|
|
|
|
DTLB_VIRT_MISS_EXITS,
|
|
|
|
SYSCALL_EXITS,
|
|
|
|
ISI_EXITS,
|
|
|
|
DSI_EXITS,
|
|
|
|
EMULATED_INST_EXITS,
|
|
|
|
EMULATED_MTMSRWE_EXITS,
|
|
|
|
EMULATED_WRTEE_EXITS,
|
|
|
|
EMULATED_MTSPR_EXITS,
|
|
|
|
EMULATED_MFSPR_EXITS,
|
|
|
|
EMULATED_MTMSR_EXITS,
|
|
|
|
EMULATED_MFMSR_EXITS,
|
|
|
|
EMULATED_TLBSX_EXITS,
|
|
|
|
EMULATED_TLBWE_EXITS,
|
|
|
|
EMULATED_RFI_EXITS,
|
|
|
|
DEC_EXITS,
|
|
|
|
EXT_INTR_EXITS,
|
|
|
|
HALT_WAKEUP,
|
|
|
|
USR_PR_INST,
|
|
|
|
FP_UNAVAIL,
|
|
|
|
DEBUG_EXITS,
|
|
|
|
TIMEINGUEST,
|
|
|
|
__NUMBER_OF_KVM_EXIT_TYPES
|
|
|
|
};
|
|
|
|
|
|
|
|
/* allow access to big endian 32bit upper/lower parts and 64bit var */
|
2008-12-02 21:51:58 +00:00
|
|
|
struct kvmppc_exit_timing {
|
2008-12-02 21:51:57 +00:00
|
|
|
union {
|
|
|
|
u64 tv64;
|
|
|
|
struct {
|
|
|
|
u32 tbu, tbl;
|
|
|
|
} tv32;
|
|
|
|
};
|
|
|
|
};
|
|
|
|
|
KVM: PPC: Add support for Book3S processors in hypervisor mode
This adds support for KVM running on 64-bit Book 3S processors,
specifically POWER7, in hypervisor mode. Using hypervisor mode means
that the guest can use the processor's supervisor mode. That means
that the guest can execute privileged instructions and access privileged
registers itself without trapping to the host. This gives excellent
performance, but does mean that KVM cannot emulate a processor
architecture other than the one that the hardware implements.
This code assumes that the guest is running paravirtualized using the
PAPR (Power Architecture Platform Requirements) interface, which is the
interface that IBM's PowerVM hypervisor uses. That means that existing
Linux distributions that run on IBM pSeries machines will also run
under KVM without modification. In order to communicate the PAPR
hypercalls to qemu, this adds a new KVM_EXIT_PAPR_HCALL exit code
to include/linux/kvm.h.
Currently the choice between book3s_hv support and book3s_pr support
(i.e. the existing code, which runs the guest in user mode) has to be
made at kernel configuration time, so a given kernel binary can only
do one or the other.
This new book3s_hv code doesn't support MMIO emulation at present.
Since we are running paravirtualized guests, this isn't a serious
restriction.
With the guest running in supervisor mode, most exceptions go straight
to the guest. We will never get data or instruction storage or segment
interrupts, alignment interrupts, decrementer interrupts, program
interrupts, single-step interrupts, etc., coming to the hypervisor from
the guest. Therefore this introduces a new KVMTEST_NONHV macro for the
exception entry path so that we don't have to do the KVM test on entry
to those exception handlers.
We do however get hypervisor decrementer, hypervisor data storage,
hypervisor instruction storage, and hypervisor emulation assist
interrupts, so we have to handle those.
In hypervisor mode, real-mode accesses can access all of RAM, not just
a limited amount. Therefore we put all the guest state in the vcpu.arch
and use the shadow_vcpu in the PACA only for temporary scratch space.
We allocate the vcpu with kzalloc rather than vzalloc, and we don't use
anything in the kvmppc_vcpu_book3s struct, so we don't allocate it.
We don't have a shared page with the guest, but we still need a
kvm_vcpu_arch_shared struct to store the values of various registers,
so we include one in the vcpu_arch struct.
The POWER7 processor has a restriction that all threads in a core have
to be in the same partition. MMU-on kernel code counts as a partition
(partition 0), so we have to do a partition switch on every entry to and
exit from the guest. At present we require the host and guest to run
in single-thread mode because of this hardware restriction.
This code allocates a hashed page table for the guest and initializes
it with HPTEs for the guest's Virtual Real Memory Area (VRMA). We
require that the guest memory is allocated using 16MB huge pages, in
order to simplify the low-level memory management. This also means that
we can get away without tracking paging activity in the host for now,
since huge pages can't be paged or swapped.
This also adds a few new exports needed by the book3s_hv code.
Signed-off-by: Paul Mackerras <paulus@samba.org>
Signed-off-by: Alexander Graf <agraf@suse.de>
2011-06-29 00:21:34 +00:00
|
|
|
struct kvmppc_pginfo {
|
|
|
|
unsigned long pfn;
|
|
|
|
atomic_t refcnt;
|
|
|
|
};
|
|
|
|
|
2008-04-17 04:28:09 +00:00
|
|
|
struct kvm_arch {
|
KVM: PPC: Add support for Book3S processors in hypervisor mode
This adds support for KVM running on 64-bit Book 3S processors,
specifically POWER7, in hypervisor mode. Using hypervisor mode means
that the guest can use the processor's supervisor mode. That means
that the guest can execute privileged instructions and access privileged
registers itself without trapping to the host. This gives excellent
performance, but does mean that KVM cannot emulate a processor
architecture other than the one that the hardware implements.
This code assumes that the guest is running paravirtualized using the
PAPR (Power Architecture Platform Requirements) interface, which is the
interface that IBM's PowerVM hypervisor uses. That means that existing
Linux distributions that run on IBM pSeries machines will also run
under KVM without modification. In order to communicate the PAPR
hypercalls to qemu, this adds a new KVM_EXIT_PAPR_HCALL exit code
to include/linux/kvm.h.
Currently the choice between book3s_hv support and book3s_pr support
(i.e. the existing code, which runs the guest in user mode) has to be
made at kernel configuration time, so a given kernel binary can only
do one or the other.
This new book3s_hv code doesn't support MMIO emulation at present.
Since we are running paravirtualized guests, this isn't a serious
restriction.
With the guest running in supervisor mode, most exceptions go straight
to the guest. We will never get data or instruction storage or segment
interrupts, alignment interrupts, decrementer interrupts, program
interrupts, single-step interrupts, etc., coming to the hypervisor from
the guest. Therefore this introduces a new KVMTEST_NONHV macro for the
exception entry path so that we don't have to do the KVM test on entry
to those exception handlers.
We do however get hypervisor decrementer, hypervisor data storage,
hypervisor instruction storage, and hypervisor emulation assist
interrupts, so we have to handle those.
In hypervisor mode, real-mode accesses can access all of RAM, not just
a limited amount. Therefore we put all the guest state in the vcpu.arch
and use the shadow_vcpu in the PACA only for temporary scratch space.
We allocate the vcpu with kzalloc rather than vzalloc, and we don't use
anything in the kvmppc_vcpu_book3s struct, so we don't allocate it.
We don't have a shared page with the guest, but we still need a
kvm_vcpu_arch_shared struct to store the values of various registers,
so we include one in the vcpu_arch struct.
The POWER7 processor has a restriction that all threads in a core have
to be in the same partition. MMU-on kernel code counts as a partition
(partition 0), so we have to do a partition switch on every entry to and
exit from the guest. At present we require the host and guest to run
in single-thread mode because of this hardware restriction.
This code allocates a hashed page table for the guest and initializes
it with HPTEs for the guest's Virtual Real Memory Area (VRMA). We
require that the guest memory is allocated using 16MB huge pages, in
order to simplify the low-level memory management. This also means that
we can get away without tracking paging activity in the host for now,
since huge pages can't be paged or swapped.
This also adds a few new exports needed by the book3s_hv code.
Signed-off-by: Paul Mackerras <paulus@samba.org>
Signed-off-by: Alexander Graf <agraf@suse.de>
2011-06-29 00:21:34 +00:00
|
|
|
#ifdef CONFIG_KVM_BOOK3S_64_HV
|
|
|
|
unsigned long hpt_virt;
|
|
|
|
unsigned long ram_npages;
|
|
|
|
unsigned long ram_psize;
|
|
|
|
unsigned long ram_porder;
|
|
|
|
struct kvmppc_pginfo *ram_pginfo;
|
|
|
|
unsigned int lpid;
|
|
|
|
unsigned int host_lpid;
|
|
|
|
unsigned long host_lpcr;
|
|
|
|
unsigned long sdr1;
|
|
|
|
unsigned long host_sdr1;
|
|
|
|
int tlbie_lock;
|
|
|
|
unsigned short last_vcpu[NR_CPUS];
|
|
|
|
#endif /* CONFIG_KVM_BOOK3S_64_HV */
|
2008-04-17 04:28:09 +00:00
|
|
|
};
|
|
|
|
|
2009-10-30 05:47:04 +00:00
|
|
|
struct kvmppc_pte {
|
2010-04-20 00:49:46 +00:00
|
|
|
ulong eaddr;
|
2009-10-30 05:47:04 +00:00
|
|
|
u64 vpage;
|
2010-04-20 00:49:46 +00:00
|
|
|
ulong raddr;
|
2010-03-24 20:48:36 +00:00
|
|
|
bool may_read : 1;
|
|
|
|
bool may_write : 1;
|
|
|
|
bool may_execute : 1;
|
2009-10-30 05:47:04 +00:00
|
|
|
};
|
|
|
|
|
|
|
|
struct kvmppc_mmu {
|
|
|
|
/* book3s_64 only */
|
|
|
|
void (*slbmte)(struct kvm_vcpu *vcpu, u64 rb, u64 rs);
|
|
|
|
u64 (*slbmfee)(struct kvm_vcpu *vcpu, u64 slb_nr);
|
|
|
|
u64 (*slbmfev)(struct kvm_vcpu *vcpu, u64 slb_nr);
|
|
|
|
void (*slbie)(struct kvm_vcpu *vcpu, u64 slb_nr);
|
|
|
|
void (*slbia)(struct kvm_vcpu *vcpu);
|
|
|
|
/* book3s */
|
|
|
|
void (*mtsrin)(struct kvm_vcpu *vcpu, u32 srnum, ulong value);
|
|
|
|
u32 (*mfsrin)(struct kvm_vcpu *vcpu, u32 srnum);
|
|
|
|
int (*xlate)(struct kvm_vcpu *vcpu, gva_t eaddr, struct kvmppc_pte *pte, bool data);
|
|
|
|
void (*reset_msr)(struct kvm_vcpu *vcpu);
|
|
|
|
void (*tlbie)(struct kvm_vcpu *vcpu, ulong addr, bool large);
|
2010-04-20 00:49:46 +00:00
|
|
|
int (*esid_to_vsid)(struct kvm_vcpu *vcpu, ulong esid, u64 *vsid);
|
2009-10-30 05:47:04 +00:00
|
|
|
u64 (*ea_to_vp)(struct kvm_vcpu *vcpu, gva_t eaddr, bool data);
|
|
|
|
bool (*is_dcbz32)(struct kvm_vcpu *vcpu);
|
|
|
|
};
|
|
|
|
|
2011-06-29 00:17:33 +00:00
|
|
|
struct kvmppc_slb {
|
|
|
|
u64 esid;
|
|
|
|
u64 vsid;
|
|
|
|
u64 orige;
|
|
|
|
u64 origv;
|
|
|
|
bool valid : 1;
|
|
|
|
bool Ks : 1;
|
|
|
|
bool Kp : 1;
|
|
|
|
bool nx : 1;
|
|
|
|
bool large : 1; /* PTEs are 16MB */
|
|
|
|
bool tb : 1; /* 1TB segment */
|
|
|
|
bool class : 1;
|
2009-10-30 05:47:04 +00:00
|
|
|
};
|
|
|
|
|
2008-04-17 04:28:09 +00:00
|
|
|
struct kvm_vcpu_arch {
|
2009-10-30 05:47:04 +00:00
|
|
|
ulong host_stack;
|
2008-04-17 04:28:09 +00:00
|
|
|
u32 host_pid;
|
2010-04-15 22:11:42 +00:00
|
|
|
#ifdef CONFIG_PPC_BOOK3S
|
2009-10-30 05:47:04 +00:00
|
|
|
ulong host_msr;
|
|
|
|
ulong host_r2;
|
|
|
|
void *host_retip;
|
|
|
|
ulong trampoline_lowmem;
|
|
|
|
ulong trampoline_enter;
|
|
|
|
ulong highmem_handler;
|
2010-01-08 01:58:06 +00:00
|
|
|
ulong rmcall;
|
2009-10-30 05:47:04 +00:00
|
|
|
ulong host_paca_phys;
|
2011-06-29 00:17:33 +00:00
|
|
|
struct kvmppc_slb slb[64];
|
KVM: PPC: Add support for Book3S processors in hypervisor mode
This adds support for KVM running on 64-bit Book 3S processors,
specifically POWER7, in hypervisor mode. Using hypervisor mode means
that the guest can use the processor's supervisor mode. That means
that the guest can execute privileged instructions and access privileged
registers itself without trapping to the host. This gives excellent
performance, but does mean that KVM cannot emulate a processor
architecture other than the one that the hardware implements.
This code assumes that the guest is running paravirtualized using the
PAPR (Power Architecture Platform Requirements) interface, which is the
interface that IBM's PowerVM hypervisor uses. That means that existing
Linux distributions that run on IBM pSeries machines will also run
under KVM without modification. In order to communicate the PAPR
hypercalls to qemu, this adds a new KVM_EXIT_PAPR_HCALL exit code
to include/linux/kvm.h.
Currently the choice between book3s_hv support and book3s_pr support
(i.e. the existing code, which runs the guest in user mode) has to be
made at kernel configuration time, so a given kernel binary can only
do one or the other.
This new book3s_hv code doesn't support MMIO emulation at present.
Since we are running paravirtualized guests, this isn't a serious
restriction.
With the guest running in supervisor mode, most exceptions go straight
to the guest. We will never get data or instruction storage or segment
interrupts, alignment interrupts, decrementer interrupts, program
interrupts, single-step interrupts, etc., coming to the hypervisor from
the guest. Therefore this introduces a new KVMTEST_NONHV macro for the
exception entry path so that we don't have to do the KVM test on entry
to those exception handlers.
We do however get hypervisor decrementer, hypervisor data storage,
hypervisor instruction storage, and hypervisor emulation assist
interrupts, so we have to handle those.
In hypervisor mode, real-mode accesses can access all of RAM, not just
a limited amount. Therefore we put all the guest state in the vcpu.arch
and use the shadow_vcpu in the PACA only for temporary scratch space.
We allocate the vcpu with kzalloc rather than vzalloc, and we don't use
anything in the kvmppc_vcpu_book3s struct, so we don't allocate it.
We don't have a shared page with the guest, but we still need a
kvm_vcpu_arch_shared struct to store the values of various registers,
so we include one in the vcpu_arch struct.
The POWER7 processor has a restriction that all threads in a core have
to be in the same partition. MMU-on kernel code counts as a partition
(partition 0), so we have to do a partition switch on every entry to and
exit from the guest. At present we require the host and guest to run
in single-thread mode because of this hardware restriction.
This code allocates a hashed page table for the guest and initializes
it with HPTEs for the guest's Virtual Real Memory Area (VRMA). We
require that the guest memory is allocated using 16MB huge pages, in
order to simplify the low-level memory management. This also means that
we can get away without tracking paging activity in the host for now,
since huge pages can't be paged or swapped.
This also adds a few new exports needed by the book3s_hv code.
Signed-off-by: Paul Mackerras <paulus@samba.org>
Signed-off-by: Alexander Graf <agraf@suse.de>
2011-06-29 00:21:34 +00:00
|
|
|
int slb_max; /* 1 + index of last valid entry in slb[] */
|
2011-06-29 00:17:33 +00:00
|
|
|
int slb_nr; /* total number of entries in SLB */
|
2009-10-30 05:47:04 +00:00
|
|
|
struct kvmppc_mmu mmu;
|
|
|
|
#endif
|
2008-04-17 04:28:09 +00:00
|
|
|
|
2008-11-05 15:36:19 +00:00
|
|
|
ulong gpr[32];
|
2008-04-17 04:28:09 +00:00
|
|
|
|
2010-01-15 13:49:11 +00:00
|
|
|
u64 fpr[32];
|
2010-02-19 10:00:28 +00:00
|
|
|
u64 fpscr;
|
2010-01-15 13:49:11 +00:00
|
|
|
|
2011-06-14 23:34:31 +00:00
|
|
|
#ifdef CONFIG_SPE
|
|
|
|
ulong evr[32];
|
|
|
|
ulong spefscr;
|
|
|
|
ulong host_spefscr;
|
|
|
|
u64 acc;
|
|
|
|
#endif
|
2010-01-15 13:49:11 +00:00
|
|
|
#ifdef CONFIG_ALTIVEC
|
|
|
|
vector128 vr[32];
|
|
|
|
vector128 vscr;
|
|
|
|
#endif
|
|
|
|
|
|
|
|
#ifdef CONFIG_VSX
|
KVM: PPC: Add support for Book3S processors in hypervisor mode
This adds support for KVM running on 64-bit Book 3S processors,
specifically POWER7, in hypervisor mode. Using hypervisor mode means
that the guest can use the processor's supervisor mode. That means
that the guest can execute privileged instructions and access privileged
registers itself without trapping to the host. This gives excellent
performance, but does mean that KVM cannot emulate a processor
architecture other than the one that the hardware implements.
This code assumes that the guest is running paravirtualized using the
PAPR (Power Architecture Platform Requirements) interface, which is the
interface that IBM's PowerVM hypervisor uses. That means that existing
Linux distributions that run on IBM pSeries machines will also run
under KVM without modification. In order to communicate the PAPR
hypercalls to qemu, this adds a new KVM_EXIT_PAPR_HCALL exit code
to include/linux/kvm.h.
Currently the choice between book3s_hv support and book3s_pr support
(i.e. the existing code, which runs the guest in user mode) has to be
made at kernel configuration time, so a given kernel binary can only
do one or the other.
This new book3s_hv code doesn't support MMIO emulation at present.
Since we are running paravirtualized guests, this isn't a serious
restriction.
With the guest running in supervisor mode, most exceptions go straight
to the guest. We will never get data or instruction storage or segment
interrupts, alignment interrupts, decrementer interrupts, program
interrupts, single-step interrupts, etc., coming to the hypervisor from
the guest. Therefore this introduces a new KVMTEST_NONHV macro for the
exception entry path so that we don't have to do the KVM test on entry
to those exception handlers.
We do however get hypervisor decrementer, hypervisor data storage,
hypervisor instruction storage, and hypervisor emulation assist
interrupts, so we have to handle those.
In hypervisor mode, real-mode accesses can access all of RAM, not just
a limited amount. Therefore we put all the guest state in the vcpu.arch
and use the shadow_vcpu in the PACA only for temporary scratch space.
We allocate the vcpu with kzalloc rather than vzalloc, and we don't use
anything in the kvmppc_vcpu_book3s struct, so we don't allocate it.
We don't have a shared page with the guest, but we still need a
kvm_vcpu_arch_shared struct to store the values of various registers,
so we include one in the vcpu_arch struct.
The POWER7 processor has a restriction that all threads in a core have
to be in the same partition. MMU-on kernel code counts as a partition
(partition 0), so we have to do a partition switch on every entry to and
exit from the guest. At present we require the host and guest to run
in single-thread mode because of this hardware restriction.
This code allocates a hashed page table for the guest and initializes
it with HPTEs for the guest's Virtual Real Memory Area (VRMA). We
require that the guest memory is allocated using 16MB huge pages, in
order to simplify the low-level memory management. This also means that
we can get away without tracking paging activity in the host for now,
since huge pages can't be paged or swapped.
This also adds a few new exports needed by the book3s_hv code.
Signed-off-by: Paul Mackerras <paulus@samba.org>
Signed-off-by: Alexander Graf <agraf@suse.de>
2011-06-29 00:21:34 +00:00
|
|
|
u64 vsr[64];
|
2010-01-15 13:49:11 +00:00
|
|
|
#endif
|
|
|
|
|
2010-02-19 10:00:27 +00:00
|
|
|
#ifdef CONFIG_PPC_BOOK3S
|
|
|
|
/* For Gekko paired singles */
|
|
|
|
u32 qpr[32];
|
|
|
|
#endif
|
|
|
|
|
2008-11-05 15:36:19 +00:00
|
|
|
ulong pc;
|
|
|
|
ulong ctr;
|
|
|
|
ulong lr;
|
2010-01-08 01:58:03 +00:00
|
|
|
|
2008-11-05 15:36:19 +00:00
|
|
|
ulong xer;
|
2010-01-08 01:58:03 +00:00
|
|
|
u32 cr;
|
2008-04-17 04:28:09 +00:00
|
|
|
|
2010-04-15 22:11:42 +00:00
|
|
|
#ifdef CONFIG_PPC_BOOK3S
|
2009-10-30 05:47:04 +00:00
|
|
|
ulong hflags;
|
2010-01-15 13:49:11 +00:00
|
|
|
ulong guest_owned_ext;
|
KVM: PPC: Add support for Book3S processors in hypervisor mode
This adds support for KVM running on 64-bit Book 3S processors,
specifically POWER7, in hypervisor mode. Using hypervisor mode means
that the guest can use the processor's supervisor mode. That means
that the guest can execute privileged instructions and access privileged
registers itself without trapping to the host. This gives excellent
performance, but does mean that KVM cannot emulate a processor
architecture other than the one that the hardware implements.
This code assumes that the guest is running paravirtualized using the
PAPR (Power Architecture Platform Requirements) interface, which is the
interface that IBM's PowerVM hypervisor uses. That means that existing
Linux distributions that run on IBM pSeries machines will also run
under KVM without modification. In order to communicate the PAPR
hypercalls to qemu, this adds a new KVM_EXIT_PAPR_HCALL exit code
to include/linux/kvm.h.
Currently the choice between book3s_hv support and book3s_pr support
(i.e. the existing code, which runs the guest in user mode) has to be
made at kernel configuration time, so a given kernel binary can only
do one or the other.
This new book3s_hv code doesn't support MMIO emulation at present.
Since we are running paravirtualized guests, this isn't a serious
restriction.
With the guest running in supervisor mode, most exceptions go straight
to the guest. We will never get data or instruction storage or segment
interrupts, alignment interrupts, decrementer interrupts, program
interrupts, single-step interrupts, etc., coming to the hypervisor from
the guest. Therefore this introduces a new KVMTEST_NONHV macro for the
exception entry path so that we don't have to do the KVM test on entry
to those exception handlers.
We do however get hypervisor decrementer, hypervisor data storage,
hypervisor instruction storage, and hypervisor emulation assist
interrupts, so we have to handle those.
In hypervisor mode, real-mode accesses can access all of RAM, not just
a limited amount. Therefore we put all the guest state in the vcpu.arch
and use the shadow_vcpu in the PACA only for temporary scratch space.
We allocate the vcpu with kzalloc rather than vzalloc, and we don't use
anything in the kvmppc_vcpu_book3s struct, so we don't allocate it.
We don't have a shared page with the guest, but we still need a
kvm_vcpu_arch_shared struct to store the values of various registers,
so we include one in the vcpu_arch struct.
The POWER7 processor has a restriction that all threads in a core have
to be in the same partition. MMU-on kernel code counts as a partition
(partition 0), so we have to do a partition switch on every entry to and
exit from the guest. At present we require the host and guest to run
in single-thread mode because of this hardware restriction.
This code allocates a hashed page table for the guest and initializes
it with HPTEs for the guest's Virtual Real Memory Area (VRMA). We
require that the guest memory is allocated using 16MB huge pages, in
order to simplify the low-level memory management. This also means that
we can get away without tracking paging activity in the host for now,
since huge pages can't be paged or swapped.
This also adds a few new exports needed by the book3s_hv code.
Signed-off-by: Paul Mackerras <paulus@samba.org>
Signed-off-by: Alexander Graf <agraf@suse.de>
2011-06-29 00:21:34 +00:00
|
|
|
ulong purr;
|
|
|
|
ulong spurr;
|
|
|
|
ulong lpcr;
|
|
|
|
ulong dscr;
|
|
|
|
ulong amr;
|
|
|
|
ulong uamor;
|
|
|
|
u32 ctrl;
|
|
|
|
ulong dabr;
|
2009-10-30 05:47:04 +00:00
|
|
|
#endif
|
2011-04-27 22:24:10 +00:00
|
|
|
u32 vrsave; /* also USPRG0 */
|
2008-04-17 04:28:09 +00:00
|
|
|
u32 mmucr;
|
2011-06-14 23:34:29 +00:00
|
|
|
ulong shadow_msr;
|
2008-11-05 15:36:19 +00:00
|
|
|
ulong sprg4;
|
|
|
|
ulong sprg5;
|
|
|
|
ulong sprg6;
|
|
|
|
ulong sprg7;
|
|
|
|
ulong csrr0;
|
|
|
|
ulong csrr1;
|
|
|
|
ulong dsrr0;
|
|
|
|
ulong dsrr1;
|
2011-04-27 22:24:21 +00:00
|
|
|
ulong mcsrr0;
|
|
|
|
ulong mcsrr1;
|
|
|
|
ulong mcsr;
|
2008-11-05 15:36:19 +00:00
|
|
|
ulong esr;
|
2008-04-17 04:28:09 +00:00
|
|
|
u32 dec;
|
|
|
|
u32 decar;
|
|
|
|
u32 tbl;
|
|
|
|
u32 tbu;
|
|
|
|
u32 tcr;
|
|
|
|
u32 tsr;
|
2009-01-03 22:23:13 +00:00
|
|
|
u32 ivor[64];
|
2008-11-05 15:36:19 +00:00
|
|
|
ulong ivpr;
|
2008-04-17 04:28:09 +00:00
|
|
|
u32 pir;
|
2009-10-30 05:47:04 +00:00
|
|
|
u32 pvr;
|
2008-07-25 18:54:53 +00:00
|
|
|
|
|
|
|
u32 shadow_pid;
|
2011-06-14 23:35:14 +00:00
|
|
|
u32 shadow_pid1;
|
2008-04-17 04:28:09 +00:00
|
|
|
u32 pid;
|
2008-07-25 18:54:53 +00:00
|
|
|
u32 swap_pid;
|
|
|
|
|
2008-04-17 04:28:09 +00:00
|
|
|
u32 ccr0;
|
|
|
|
u32 ccr1;
|
|
|
|
u32 dbcr0;
|
|
|
|
u32 dbcr1;
|
2009-01-03 22:23:07 +00:00
|
|
|
u32 dbsr;
|
2008-04-17 04:28:09 +00:00
|
|
|
|
KVM: PPC: Add support for Book3S processors in hypervisor mode
This adds support for KVM running on 64-bit Book 3S processors,
specifically POWER7, in hypervisor mode. Using hypervisor mode means
that the guest can use the processor's supervisor mode. That means
that the guest can execute privileged instructions and access privileged
registers itself without trapping to the host. This gives excellent
performance, but does mean that KVM cannot emulate a processor
architecture other than the one that the hardware implements.
This code assumes that the guest is running paravirtualized using the
PAPR (Power Architecture Platform Requirements) interface, which is the
interface that IBM's PowerVM hypervisor uses. That means that existing
Linux distributions that run on IBM pSeries machines will also run
under KVM without modification. In order to communicate the PAPR
hypercalls to qemu, this adds a new KVM_EXIT_PAPR_HCALL exit code
to include/linux/kvm.h.
Currently the choice between book3s_hv support and book3s_pr support
(i.e. the existing code, which runs the guest in user mode) has to be
made at kernel configuration time, so a given kernel binary can only
do one or the other.
This new book3s_hv code doesn't support MMIO emulation at present.
Since we are running paravirtualized guests, this isn't a serious
restriction.
With the guest running in supervisor mode, most exceptions go straight
to the guest. We will never get data or instruction storage or segment
interrupts, alignment interrupts, decrementer interrupts, program
interrupts, single-step interrupts, etc., coming to the hypervisor from
the guest. Therefore this introduces a new KVMTEST_NONHV macro for the
exception entry path so that we don't have to do the KVM test on entry
to those exception handlers.
We do however get hypervisor decrementer, hypervisor data storage,
hypervisor instruction storage, and hypervisor emulation assist
interrupts, so we have to handle those.
In hypervisor mode, real-mode accesses can access all of RAM, not just
a limited amount. Therefore we put all the guest state in the vcpu.arch
and use the shadow_vcpu in the PACA only for temporary scratch space.
We allocate the vcpu with kzalloc rather than vzalloc, and we don't use
anything in the kvmppc_vcpu_book3s struct, so we don't allocate it.
We don't have a shared page with the guest, but we still need a
kvm_vcpu_arch_shared struct to store the values of various registers,
so we include one in the vcpu_arch struct.
The POWER7 processor has a restriction that all threads in a core have
to be in the same partition. MMU-on kernel code counts as a partition
(partition 0), so we have to do a partition switch on every entry to and
exit from the guest. At present we require the host and guest to run
in single-thread mode because of this hardware restriction.
This code allocates a hashed page table for the guest and initializes
it with HPTEs for the guest's Virtual Real Memory Area (VRMA). We
require that the guest memory is allocated using 16MB huge pages, in
order to simplify the low-level memory management. This also means that
we can get away without tracking paging activity in the host for now,
since huge pages can't be paged or swapped.
This also adds a few new exports needed by the book3s_hv code.
Signed-off-by: Paul Mackerras <paulus@samba.org>
Signed-off-by: Alexander Graf <agraf@suse.de>
2011-06-29 00:21:34 +00:00
|
|
|
u64 mmcr[3];
|
|
|
|
u32 pmc[6];
|
|
|
|
|
2008-12-02 21:51:57 +00:00
|
|
|
#ifdef CONFIG_KVM_EXIT_TIMING
|
2011-03-25 05:02:13 +00:00
|
|
|
struct mutex exit_timing_lock;
|
2008-12-02 21:51:58 +00:00
|
|
|
struct kvmppc_exit_timing timing_exit;
|
|
|
|
struct kvmppc_exit_timing timing_last_enter;
|
2008-12-02 21:51:57 +00:00
|
|
|
u32 last_exit_type;
|
|
|
|
u32 timing_count_type[__NUMBER_OF_KVM_EXIT_TYPES];
|
|
|
|
u64 timing_sum_duration[__NUMBER_OF_KVM_EXIT_TYPES];
|
|
|
|
u64 timing_sum_quad_duration[__NUMBER_OF_KVM_EXIT_TYPES];
|
|
|
|
u64 timing_min_duration[__NUMBER_OF_KVM_EXIT_TYPES];
|
|
|
|
u64 timing_max_duration[__NUMBER_OF_KVM_EXIT_TYPES];
|
|
|
|
u64 timing_last_exit;
|
|
|
|
struct dentry *debugfs_exit_timing;
|
|
|
|
#endif
|
|
|
|
|
KVM: PPC: Add support for Book3S processors in hypervisor mode
This adds support for KVM running on 64-bit Book 3S processors,
specifically POWER7, in hypervisor mode. Using hypervisor mode means
that the guest can use the processor's supervisor mode. That means
that the guest can execute privileged instructions and access privileged
registers itself without trapping to the host. This gives excellent
performance, but does mean that KVM cannot emulate a processor
architecture other than the one that the hardware implements.
This code assumes that the guest is running paravirtualized using the
PAPR (Power Architecture Platform Requirements) interface, which is the
interface that IBM's PowerVM hypervisor uses. That means that existing
Linux distributions that run on IBM pSeries machines will also run
under KVM without modification. In order to communicate the PAPR
hypercalls to qemu, this adds a new KVM_EXIT_PAPR_HCALL exit code
to include/linux/kvm.h.
Currently the choice between book3s_hv support and book3s_pr support
(i.e. the existing code, which runs the guest in user mode) has to be
made at kernel configuration time, so a given kernel binary can only
do one or the other.
This new book3s_hv code doesn't support MMIO emulation at present.
Since we are running paravirtualized guests, this isn't a serious
restriction.
With the guest running in supervisor mode, most exceptions go straight
to the guest. We will never get data or instruction storage or segment
interrupts, alignment interrupts, decrementer interrupts, program
interrupts, single-step interrupts, etc., coming to the hypervisor from
the guest. Therefore this introduces a new KVMTEST_NONHV macro for the
exception entry path so that we don't have to do the KVM test on entry
to those exception handlers.
We do however get hypervisor decrementer, hypervisor data storage,
hypervisor instruction storage, and hypervisor emulation assist
interrupts, so we have to handle those.
In hypervisor mode, real-mode accesses can access all of RAM, not just
a limited amount. Therefore we put all the guest state in the vcpu.arch
and use the shadow_vcpu in the PACA only for temporary scratch space.
We allocate the vcpu with kzalloc rather than vzalloc, and we don't use
anything in the kvmppc_vcpu_book3s struct, so we don't allocate it.
We don't have a shared page with the guest, but we still need a
kvm_vcpu_arch_shared struct to store the values of various registers,
so we include one in the vcpu_arch struct.
The POWER7 processor has a restriction that all threads in a core have
to be in the same partition. MMU-on kernel code counts as a partition
(partition 0), so we have to do a partition switch on every entry to and
exit from the guest. At present we require the host and guest to run
in single-thread mode because of this hardware restriction.
This code allocates a hashed page table for the guest and initializes
it with HPTEs for the guest's Virtual Real Memory Area (VRMA). We
require that the guest memory is allocated using 16MB huge pages, in
order to simplify the low-level memory management. This also means that
we can get away without tracking paging activity in the host for now,
since huge pages can't be paged or swapped.
This also adds a few new exports needed by the book3s_hv code.
Signed-off-by: Paul Mackerras <paulus@samba.org>
Signed-off-by: Alexander Graf <agraf@suse.de>
2011-06-29 00:21:34 +00:00
|
|
|
#ifdef CONFIG_PPC_BOOK3S
|
|
|
|
ulong fault_dar;
|
|
|
|
u32 fault_dsisr;
|
|
|
|
#endif
|
|
|
|
|
2010-04-15 22:11:44 +00:00
|
|
|
#ifdef CONFIG_BOOKE
|
2008-11-05 15:36:19 +00:00
|
|
|
ulong fault_dear;
|
|
|
|
ulong fault_esr;
|
2010-02-02 11:44:35 +00:00
|
|
|
ulong queued_dear;
|
|
|
|
ulong queued_esr;
|
2010-04-15 22:11:44 +00:00
|
|
|
#endif
|
2008-04-17 04:28:09 +00:00
|
|
|
gpa_t paddr_accessed;
|
|
|
|
|
|
|
|
u8 io_gpr; /* GPR used as IO source/target */
|
|
|
|
u8 mmio_is_bigendian;
|
2010-02-19 10:00:30 +00:00
|
|
|
u8 mmio_sign_extend;
|
2008-04-17 04:28:09 +00:00
|
|
|
u8 dcr_needed;
|
|
|
|
u8 dcr_is_write;
|
2010-03-24 20:48:30 +00:00
|
|
|
u8 osi_needed;
|
|
|
|
u8 osi_enabled;
|
KVM: PPC: Add support for Book3S processors in hypervisor mode
This adds support for KVM running on 64-bit Book 3S processors,
specifically POWER7, in hypervisor mode. Using hypervisor mode means
that the guest can use the processor's supervisor mode. That means
that the guest can execute privileged instructions and access privileged
registers itself without trapping to the host. This gives excellent
performance, but does mean that KVM cannot emulate a processor
architecture other than the one that the hardware implements.
This code assumes that the guest is running paravirtualized using the
PAPR (Power Architecture Platform Requirements) interface, which is the
interface that IBM's PowerVM hypervisor uses. That means that existing
Linux distributions that run on IBM pSeries machines will also run
under KVM without modification. In order to communicate the PAPR
hypercalls to qemu, this adds a new KVM_EXIT_PAPR_HCALL exit code
to include/linux/kvm.h.
Currently the choice between book3s_hv support and book3s_pr support
(i.e. the existing code, which runs the guest in user mode) has to be
made at kernel configuration time, so a given kernel binary can only
do one or the other.
This new book3s_hv code doesn't support MMIO emulation at present.
Since we are running paravirtualized guests, this isn't a serious
restriction.
With the guest running in supervisor mode, most exceptions go straight
to the guest. We will never get data or instruction storage or segment
interrupts, alignment interrupts, decrementer interrupts, program
interrupts, single-step interrupts, etc., coming to the hypervisor from
the guest. Therefore this introduces a new KVMTEST_NONHV macro for the
exception entry path so that we don't have to do the KVM test on entry
to those exception handlers.
We do however get hypervisor decrementer, hypervisor data storage,
hypervisor instruction storage, and hypervisor emulation assist
interrupts, so we have to handle those.
In hypervisor mode, real-mode accesses can access all of RAM, not just
a limited amount. Therefore we put all the guest state in the vcpu.arch
and use the shadow_vcpu in the PACA only for temporary scratch space.
We allocate the vcpu with kzalloc rather than vzalloc, and we don't use
anything in the kvmppc_vcpu_book3s struct, so we don't allocate it.
We don't have a shared page with the guest, but we still need a
kvm_vcpu_arch_shared struct to store the values of various registers,
so we include one in the vcpu_arch struct.
The POWER7 processor has a restriction that all threads in a core have
to be in the same partition. MMU-on kernel code counts as a partition
(partition 0), so we have to do a partition switch on every entry to and
exit from the guest. At present we require the host and guest to run
in single-thread mode because of this hardware restriction.
This code allocates a hashed page table for the guest and initializes
it with HPTEs for the guest's Virtual Real Memory Area (VRMA). We
require that the guest memory is allocated using 16MB huge pages, in
order to simplify the low-level memory management. This also means that
we can get away without tracking paging activity in the host for now,
since huge pages can't be paged or swapped.
This also adds a few new exports needed by the book3s_hv code.
Signed-off-by: Paul Mackerras <paulus@samba.org>
Signed-off-by: Alexander Graf <agraf@suse.de>
2011-06-29 00:21:34 +00:00
|
|
|
u8 hcall_needed;
|
2008-04-17 04:28:09 +00:00
|
|
|
|
|
|
|
u32 cpr0_cfgaddr; /* holds the last set cpr0_cfgaddr */
|
|
|
|
|
2009-11-02 12:02:31 +00:00
|
|
|
struct hrtimer dec_timer;
|
|
|
|
struct tasklet_struct tasklet;
|
2009-10-30 05:47:04 +00:00
|
|
|
u64 dec_jiffies;
|
KVM: PPC: Add support for Book3S processors in hypervisor mode
This adds support for KVM running on 64-bit Book 3S processors,
specifically POWER7, in hypervisor mode. Using hypervisor mode means
that the guest can use the processor's supervisor mode. That means
that the guest can execute privileged instructions and access privileged
registers itself without trapping to the host. This gives excellent
performance, but does mean that KVM cannot emulate a processor
architecture other than the one that the hardware implements.
This code assumes that the guest is running paravirtualized using the
PAPR (Power Architecture Platform Requirements) interface, which is the
interface that IBM's PowerVM hypervisor uses. That means that existing
Linux distributions that run on IBM pSeries machines will also run
under KVM without modification. In order to communicate the PAPR
hypercalls to qemu, this adds a new KVM_EXIT_PAPR_HCALL exit code
to include/linux/kvm.h.
Currently the choice between book3s_hv support and book3s_pr support
(i.e. the existing code, which runs the guest in user mode) has to be
made at kernel configuration time, so a given kernel binary can only
do one or the other.
This new book3s_hv code doesn't support MMIO emulation at present.
Since we are running paravirtualized guests, this isn't a serious
restriction.
With the guest running in supervisor mode, most exceptions go straight
to the guest. We will never get data or instruction storage or segment
interrupts, alignment interrupts, decrementer interrupts, program
interrupts, single-step interrupts, etc., coming to the hypervisor from
the guest. Therefore this introduces a new KVMTEST_NONHV macro for the
exception entry path so that we don't have to do the KVM test on entry
to those exception handlers.
We do however get hypervisor decrementer, hypervisor data storage,
hypervisor instruction storage, and hypervisor emulation assist
interrupts, so we have to handle those.
In hypervisor mode, real-mode accesses can access all of RAM, not just
a limited amount. Therefore we put all the guest state in the vcpu.arch
and use the shadow_vcpu in the PACA only for temporary scratch space.
We allocate the vcpu with kzalloc rather than vzalloc, and we don't use
anything in the kvmppc_vcpu_book3s struct, so we don't allocate it.
We don't have a shared page with the guest, but we still need a
kvm_vcpu_arch_shared struct to store the values of various registers,
so we include one in the vcpu_arch struct.
The POWER7 processor has a restriction that all threads in a core have
to be in the same partition. MMU-on kernel code counts as a partition
(partition 0), so we have to do a partition switch on every entry to and
exit from the guest. At present we require the host and guest to run
in single-thread mode because of this hardware restriction.
This code allocates a hashed page table for the guest and initializes
it with HPTEs for the guest's Virtual Real Memory Area (VRMA). We
require that the guest memory is allocated using 16MB huge pages, in
order to simplify the low-level memory management. This also means that
we can get away without tracking paging activity in the host for now,
since huge pages can't be paged or swapped.
This also adds a few new exports needed by the book3s_hv code.
Signed-off-by: Paul Mackerras <paulus@samba.org>
Signed-off-by: Alexander Graf <agraf@suse.de>
2011-06-29 00:21:34 +00:00
|
|
|
u64 dec_expires;
|
2008-04-17 04:28:09 +00:00
|
|
|
unsigned long pending_exceptions;
|
KVM: PPC: Add support for Book3S processors in hypervisor mode
This adds support for KVM running on 64-bit Book 3S processors,
specifically POWER7, in hypervisor mode. Using hypervisor mode means
that the guest can use the processor's supervisor mode. That means
that the guest can execute privileged instructions and access privileged
registers itself without trapping to the host. This gives excellent
performance, but does mean that KVM cannot emulate a processor
architecture other than the one that the hardware implements.
This code assumes that the guest is running paravirtualized using the
PAPR (Power Architecture Platform Requirements) interface, which is the
interface that IBM's PowerVM hypervisor uses. That means that existing
Linux distributions that run on IBM pSeries machines will also run
under KVM without modification. In order to communicate the PAPR
hypercalls to qemu, this adds a new KVM_EXIT_PAPR_HCALL exit code
to include/linux/kvm.h.
Currently the choice between book3s_hv support and book3s_pr support
(i.e. the existing code, which runs the guest in user mode) has to be
made at kernel configuration time, so a given kernel binary can only
do one or the other.
This new book3s_hv code doesn't support MMIO emulation at present.
Since we are running paravirtualized guests, this isn't a serious
restriction.
With the guest running in supervisor mode, most exceptions go straight
to the guest. We will never get data or instruction storage or segment
interrupts, alignment interrupts, decrementer interrupts, program
interrupts, single-step interrupts, etc., coming to the hypervisor from
the guest. Therefore this introduces a new KVMTEST_NONHV macro for the
exception entry path so that we don't have to do the KVM test on entry
to those exception handlers.
We do however get hypervisor decrementer, hypervisor data storage,
hypervisor instruction storage, and hypervisor emulation assist
interrupts, so we have to handle those.
In hypervisor mode, real-mode accesses can access all of RAM, not just
a limited amount. Therefore we put all the guest state in the vcpu.arch
and use the shadow_vcpu in the PACA only for temporary scratch space.
We allocate the vcpu with kzalloc rather than vzalloc, and we don't use
anything in the kvmppc_vcpu_book3s struct, so we don't allocate it.
We don't have a shared page with the guest, but we still need a
kvm_vcpu_arch_shared struct to store the values of various registers,
so we include one in the vcpu_arch struct.
The POWER7 processor has a restriction that all threads in a core have
to be in the same partition. MMU-on kernel code counts as a partition
(partition 0), so we have to do a partition switch on every entry to and
exit from the guest. At present we require the host and guest to run
in single-thread mode because of this hardware restriction.
This code allocates a hashed page table for the guest and initializes
it with HPTEs for the guest's Virtual Real Memory Area (VRMA). We
require that the guest memory is allocated using 16MB huge pages, in
order to simplify the low-level memory management. This also means that
we can get away without tracking paging activity in the host for now,
since huge pages can't be paged or swapped.
This also adds a few new exports needed by the book3s_hv code.
Signed-off-by: Paul Mackerras <paulus@samba.org>
Signed-off-by: Alexander Graf <agraf@suse.de>
2011-06-29 00:21:34 +00:00
|
|
|
u16 last_cpu;
|
2011-06-29 00:22:05 +00:00
|
|
|
u8 ceded;
|
|
|
|
u8 prodded;
|
KVM: PPC: Add support for Book3S processors in hypervisor mode
This adds support for KVM running on 64-bit Book 3S processors,
specifically POWER7, in hypervisor mode. Using hypervisor mode means
that the guest can use the processor's supervisor mode. That means
that the guest can execute privileged instructions and access privileged
registers itself without trapping to the host. This gives excellent
performance, but does mean that KVM cannot emulate a processor
architecture other than the one that the hardware implements.
This code assumes that the guest is running paravirtualized using the
PAPR (Power Architecture Platform Requirements) interface, which is the
interface that IBM's PowerVM hypervisor uses. That means that existing
Linux distributions that run on IBM pSeries machines will also run
under KVM without modification. In order to communicate the PAPR
hypercalls to qemu, this adds a new KVM_EXIT_PAPR_HCALL exit code
to include/linux/kvm.h.
Currently the choice between book3s_hv support and book3s_pr support
(i.e. the existing code, which runs the guest in user mode) has to be
made at kernel configuration time, so a given kernel binary can only
do one or the other.
This new book3s_hv code doesn't support MMIO emulation at present.
Since we are running paravirtualized guests, this isn't a serious
restriction.
With the guest running in supervisor mode, most exceptions go straight
to the guest. We will never get data or instruction storage or segment
interrupts, alignment interrupts, decrementer interrupts, program
interrupts, single-step interrupts, etc., coming to the hypervisor from
the guest. Therefore this introduces a new KVMTEST_NONHV macro for the
exception entry path so that we don't have to do the KVM test on entry
to those exception handlers.
We do however get hypervisor decrementer, hypervisor data storage,
hypervisor instruction storage, and hypervisor emulation assist
interrupts, so we have to handle those.
In hypervisor mode, real-mode accesses can access all of RAM, not just
a limited amount. Therefore we put all the guest state in the vcpu.arch
and use the shadow_vcpu in the PACA only for temporary scratch space.
We allocate the vcpu with kzalloc rather than vzalloc, and we don't use
anything in the kvmppc_vcpu_book3s struct, so we don't allocate it.
We don't have a shared page with the guest, but we still need a
kvm_vcpu_arch_shared struct to store the values of various registers,
so we include one in the vcpu_arch struct.
The POWER7 processor has a restriction that all threads in a core have
to be in the same partition. MMU-on kernel code counts as a partition
(partition 0), so we have to do a partition switch on every entry to and
exit from the guest. At present we require the host and guest to run
in single-thread mode because of this hardware restriction.
This code allocates a hashed page table for the guest and initializes
it with HPTEs for the guest's Virtual Real Memory Area (VRMA). We
require that the guest memory is allocated using 16MB huge pages, in
order to simplify the low-level memory management. This also means that
we can get away without tracking paging activity in the host for now,
since huge pages can't be paged or swapped.
This also adds a few new exports needed by the book3s_hv code.
Signed-off-by: Paul Mackerras <paulus@samba.org>
Signed-off-by: Alexander Graf <agraf@suse.de>
2011-06-29 00:21:34 +00:00
|
|
|
u32 last_inst;
|
2011-06-29 00:22:05 +00:00
|
|
|
|
|
|
|
struct lppaca *vpa;
|
|
|
|
struct slb_shadow *slb_shadow;
|
|
|
|
struct dtl *dtl;
|
|
|
|
struct dtl *dtl_end;
|
KVM: PPC: Add support for Book3S processors in hypervisor mode
This adds support for KVM running on 64-bit Book 3S processors,
specifically POWER7, in hypervisor mode. Using hypervisor mode means
that the guest can use the processor's supervisor mode. That means
that the guest can execute privileged instructions and access privileged
registers itself without trapping to the host. This gives excellent
performance, but does mean that KVM cannot emulate a processor
architecture other than the one that the hardware implements.
This code assumes that the guest is running paravirtualized using the
PAPR (Power Architecture Platform Requirements) interface, which is the
interface that IBM's PowerVM hypervisor uses. That means that existing
Linux distributions that run on IBM pSeries machines will also run
under KVM without modification. In order to communicate the PAPR
hypercalls to qemu, this adds a new KVM_EXIT_PAPR_HCALL exit code
to include/linux/kvm.h.
Currently the choice between book3s_hv support and book3s_pr support
(i.e. the existing code, which runs the guest in user mode) has to be
made at kernel configuration time, so a given kernel binary can only
do one or the other.
This new book3s_hv code doesn't support MMIO emulation at present.
Since we are running paravirtualized guests, this isn't a serious
restriction.
With the guest running in supervisor mode, most exceptions go straight
to the guest. We will never get data or instruction storage or segment
interrupts, alignment interrupts, decrementer interrupts, program
interrupts, single-step interrupts, etc., coming to the hypervisor from
the guest. Therefore this introduces a new KVMTEST_NONHV macro for the
exception entry path so that we don't have to do the KVM test on entry
to those exception handlers.
We do however get hypervisor decrementer, hypervisor data storage,
hypervisor instruction storage, and hypervisor emulation assist
interrupts, so we have to handle those.
In hypervisor mode, real-mode accesses can access all of RAM, not just
a limited amount. Therefore we put all the guest state in the vcpu.arch
and use the shadow_vcpu in the PACA only for temporary scratch space.
We allocate the vcpu with kzalloc rather than vzalloc, and we don't use
anything in the kvmppc_vcpu_book3s struct, so we don't allocate it.
We don't have a shared page with the guest, but we still need a
kvm_vcpu_arch_shared struct to store the values of various registers,
so we include one in the vcpu_arch struct.
The POWER7 processor has a restriction that all threads in a core have
to be in the same partition. MMU-on kernel code counts as a partition
(partition 0), so we have to do a partition switch on every entry to and
exit from the guest. At present we require the host and guest to run
in single-thread mode because of this hardware restriction.
This code allocates a hashed page table for the guest and initializes
it with HPTEs for the guest's Virtual Real Memory Area (VRMA). We
require that the guest memory is allocated using 16MB huge pages, in
order to simplify the low-level memory management. This also means that
we can get away without tracking paging activity in the host for now,
since huge pages can't be paged or swapped.
This also adds a few new exports needed by the book3s_hv code.
Signed-off-by: Paul Mackerras <paulus@samba.org>
Signed-off-by: Alexander Graf <agraf@suse.de>
2011-06-29 00:21:34 +00:00
|
|
|
int trap;
|
2010-07-29 12:47:42 +00:00
|
|
|
struct kvm_vcpu_arch_shared *shared;
|
2010-07-29 12:47:53 +00:00
|
|
|
unsigned long magic_page_pa; /* phys addr to map the magic page to */
|
|
|
|
unsigned long magic_page_ea; /* effect. addr to map the magic page to */
|
KVM: PPC: Add support for Book3S processors in hypervisor mode
This adds support for KVM running on 64-bit Book 3S processors,
specifically POWER7, in hypervisor mode. Using hypervisor mode means
that the guest can use the processor's supervisor mode. That means
that the guest can execute privileged instructions and access privileged
registers itself without trapping to the host. This gives excellent
performance, but does mean that KVM cannot emulate a processor
architecture other than the one that the hardware implements.
This code assumes that the guest is running paravirtualized using the
PAPR (Power Architecture Platform Requirements) interface, which is the
interface that IBM's PowerVM hypervisor uses. That means that existing
Linux distributions that run on IBM pSeries machines will also run
under KVM without modification. In order to communicate the PAPR
hypercalls to qemu, this adds a new KVM_EXIT_PAPR_HCALL exit code
to include/linux/kvm.h.
Currently the choice between book3s_hv support and book3s_pr support
(i.e. the existing code, which runs the guest in user mode) has to be
made at kernel configuration time, so a given kernel binary can only
do one or the other.
This new book3s_hv code doesn't support MMIO emulation at present.
Since we are running paravirtualized guests, this isn't a serious
restriction.
With the guest running in supervisor mode, most exceptions go straight
to the guest. We will never get data or instruction storage or segment
interrupts, alignment interrupts, decrementer interrupts, program
interrupts, single-step interrupts, etc., coming to the hypervisor from
the guest. Therefore this introduces a new KVMTEST_NONHV macro for the
exception entry path so that we don't have to do the KVM test on entry
to those exception handlers.
We do however get hypervisor decrementer, hypervisor data storage,
hypervisor instruction storage, and hypervisor emulation assist
interrupts, so we have to handle those.
In hypervisor mode, real-mode accesses can access all of RAM, not just
a limited amount. Therefore we put all the guest state in the vcpu.arch
and use the shadow_vcpu in the PACA only for temporary scratch space.
We allocate the vcpu with kzalloc rather than vzalloc, and we don't use
anything in the kvmppc_vcpu_book3s struct, so we don't allocate it.
We don't have a shared page with the guest, but we still need a
kvm_vcpu_arch_shared struct to store the values of various registers,
so we include one in the vcpu_arch struct.
The POWER7 processor has a restriction that all threads in a core have
to be in the same partition. MMU-on kernel code counts as a partition
(partition 0), so we have to do a partition switch on every entry to and
exit from the guest. At present we require the host and guest to run
in single-thread mode because of this hardware restriction.
This code allocates a hashed page table for the guest and initializes
it with HPTEs for the guest's Virtual Real Memory Area (VRMA). We
require that the guest memory is allocated using 16MB huge pages, in
order to simplify the low-level memory management. This also means that
we can get away without tracking paging activity in the host for now,
since huge pages can't be paged or swapped.
This also adds a few new exports needed by the book3s_hv code.
Signed-off-by: Paul Mackerras <paulus@samba.org>
Signed-off-by: Alexander Graf <agraf@suse.de>
2011-06-29 00:21:34 +00:00
|
|
|
|
|
|
|
#ifdef CONFIG_KVM_BOOK3S_64_HV
|
|
|
|
struct kvm_vcpu_arch_shared shregs;
|
|
|
|
#endif
|
2008-04-17 04:28:09 +00:00
|
|
|
};
|
|
|
|
|
|
|
|
#endif /* __POWERPC_KVM_HOST_H__ */
|