iv/linux
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
16aaaff684
linux/arch/powerpc/kvm/bookehv_interrupts.S

598 lines
17 KiB
ArmAsm
Raw Normal View History

KVM: PPC: booke: category E.HV (GS-mode) support Chips such as e500mc that implement category E.HV in Power ISA 2.06 provide hardware virtualization features, including a new MSR mode for guest state. The guest OS can perform many operations without trapping into the hypervisor, including transitions to and from guest userspace. Since we can use SRR1[GS] to reliably tell whether an exception came from guest state, instead of messing around with IVPR, we use DO_KVM similarly to book3s. Current issues include: - Machine checks from guest state are not routed to the host handler. - The guest can cause a host oops by executing an emulated instruction in a page that lacks read permission. Existing e500/4xx support has the same problem. Includes work by Ashish Kalra <Ashish.Kalra@freescale.com>, Varun Sethi <Varun.Sethi@freescale.com>, and Liu Yu <yu.liu@freescale.com>. Signed-off-by: Scott Wood <scottwood@freescale.com> [agraf: remove pt_regs usage] Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Avi Kivity <avi@redhat.com>
2011-12-20 19:34:43 +04:00
/*
* 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 (C) 2010-2011 Freescale Semiconductor, Inc.
*
* Author: Varun Sethi <varun.sethi@freescale.com>
* Author: Scott Wood <scotwood@freescale.com>
*
* This file is derived from arch/powerpc/kvm/booke_interrupts.S
*/
#include <asm/ppc_asm.h>
#include <asm/kvm_asm.h>
#include <asm/reg.h>
#include <asm/mmu-44x.h>
#include <asm/page.h>
#include <asm/asm-compat.h>
#include <asm/asm-offsets.h>
#include <asm/bitsperlong.h>
KVM: PPC: e500mc: add load inst fixup There's always a chance we're unable to read a guest instruction. The guest could have its TLB mapped execute-, but not readable, something odd happens and our TLB gets flushed. So it's a good idea to be prepared for that case and have a fallback that allows us to fix things up in that case. Add fixup code that keeps guest code from potentially crashing our host kernel. Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Avi Kivity <avi@redhat.com>
2012-02-16 03:24:28 +04:00
#include <asm/thread_info.h>
KVM: PPC: booke: category E.HV (GS-mode) support Chips such as e500mc that implement category E.HV in Power ISA 2.06 provide hardware virtualization features, including a new MSR mode for guest state. The guest OS can perform many operations without trapping into the hypervisor, including transitions to and from guest userspace. Since we can use SRR1[GS] to reliably tell whether an exception came from guest state, instead of messing around with IVPR, we use DO_KVM similarly to book3s. Current issues include: - Machine checks from guest state are not routed to the host handler. - The guest can cause a host oops by executing an emulated instruction in a page that lacks read permission. Existing e500/4xx support has the same problem. Includes work by Ashish Kalra <Ashish.Kalra@freescale.com>, Varun Sethi <Varun.Sethi@freescale.com>, and Liu Yu <yu.liu@freescale.com>. Signed-off-by: Scott Wood <scottwood@freescale.com> [agraf: remove pt_regs usage] Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Avi Kivity <avi@redhat.com>
2011-12-20 19:34:43 +04:00
#include "../kernel/head_booke.h" /* for THREAD_NORMSAVE() */
#define GET_VCPU(vcpu, thread) \
PPC_LL vcpu, THREAD_KVM_VCPU(thread)
#define LONGBYTES (BITS_PER_LONG / 8)
#define VCPU_GPR(n) (VCPU_GPRS + (n * LONGBYTES))
#define VCPU_GUEST_SPRG(n) (VCPU_GUEST_SPRGS + (n * LONGBYTES))
/* The host stack layout: */
#define HOST_R1 (0 * LONGBYTES) /* Implied by stwu. */
#define HOST_CALLEE_LR (1 * LONGBYTES)
#define HOST_RUN (2 * LONGBYTES) /* struct kvm_run */
/*
* r2 is special: it holds 'current', and it made nonvolatile in the
* kernel with the -ffixed-r2 gcc option.
*/
#define HOST_R2 (3 * LONGBYTES)
KVM: PPC: Save/Restore CR over vcpu_run On PPC, CR2-CR4 are nonvolatile, thus have to be saved across function calls. We didn't respect that for any architecture until Paul spotted it in his patch for Book3S-HV. This patch saves/restores CR for all KVM capable PPC hosts. Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Avi Kivity <avi@redhat.com>
2012-03-05 19:00:28 +04:00
#define HOST_CR (4 * LONGBYTES)
#define HOST_NV_GPRS (5 * LONGBYTES)
KVM: PPC: booke: category E.HV (GS-mode) support Chips such as e500mc that implement category E.HV in Power ISA 2.06 provide hardware virtualization features, including a new MSR mode for guest state. The guest OS can perform many operations without trapping into the hypervisor, including transitions to and from guest userspace. Since we can use SRR1[GS] to reliably tell whether an exception came from guest state, instead of messing around with IVPR, we use DO_KVM similarly to book3s. Current issues include: - Machine checks from guest state are not routed to the host handler. - The guest can cause a host oops by executing an emulated instruction in a page that lacks read permission. Existing e500/4xx support has the same problem. Includes work by Ashish Kalra <Ashish.Kalra@freescale.com>, Varun Sethi <Varun.Sethi@freescale.com>, and Liu Yu <yu.liu@freescale.com>. Signed-off-by: Scott Wood <scottwood@freescale.com> [agraf: remove pt_regs usage] Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Avi Kivity <avi@redhat.com>
2011-12-20 19:34:43 +04:00
#define HOST_NV_GPR(n) (HOST_NV_GPRS + ((n - 14) * LONGBYTES))
#define HOST_MIN_STACK_SIZE (HOST_NV_GPR(31) + LONGBYTES)
#define HOST_STACK_SIZE ((HOST_MIN_STACK_SIZE + 15) & ~15) /* Align. */
#define HOST_STACK_LR (HOST_STACK_SIZE + LONGBYTES) /* In caller stack frame. */
#define NEED_EMU 0x00000001 /* emulation -- save nv regs */
#define NEED_DEAR 0x00000002 /* save faulting DEAR */
#define NEED_ESR 0x00000004 /* save faulting ESR */
/*
* On entry:
* r4 = vcpu, r5 = srr0, r6 = srr1
* saved in vcpu: cr, ctr, r3-r13
*/
.macro kvm_handler_common intno, srr0, flags
KVM: PPC: e500mc: Move r1/r2 restoration very early If we hit any exception whatsoever in the restore path and r1/r2 aren't the host registers, we don't get a working oops. So it's always a good idea to restore them as early as possible. This time, it actually has practical reasons to do so too, since we need to have the host page fault handler fix up our guest instruction read code. And for that to work we need r1/r2 restored. Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Avi Kivity <avi@redhat.com>
2012-02-16 03:06:24 +04:00
/* Restore host stack pointer */
PPC_STL r1, VCPU_GPR(r1)(r4)
PPC_STL r2, VCPU_GPR(r2)(r4)
PPC_LL r1, VCPU_HOST_STACK(r4)
PPC_LL r2, HOST_R2(r1)
KVM: PPC: booke: category E.HV (GS-mode) support Chips such as e500mc that implement category E.HV in Power ISA 2.06 provide hardware virtualization features, including a new MSR mode for guest state. The guest OS can perform many operations without trapping into the hypervisor, including transitions to and from guest userspace. Since we can use SRR1[GS] to reliably tell whether an exception came from guest state, instead of messing around with IVPR, we use DO_KVM similarly to book3s. Current issues include: - Machine checks from guest state are not routed to the host handler. - The guest can cause a host oops by executing an emulated instruction in a page that lacks read permission. Existing e500/4xx support has the same problem. Includes work by Ashish Kalra <Ashish.Kalra@freescale.com>, Varun Sethi <Varun.Sethi@freescale.com>, and Liu Yu <yu.liu@freescale.com>. Signed-off-by: Scott Wood <scottwood@freescale.com> [agraf: remove pt_regs usage] Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Avi Kivity <avi@redhat.com>
2011-12-20 19:34:43 +04:00
mfspr r10, SPRN_PID
lwz r8, VCPU_HOST_PID(r4)
PPC_LL r11, VCPU_SHARED(r4)
PPC_STL r14, VCPU_GPR(r14)(r4) /* We need a non-volatile GPR. */
li r14, \intno
stw r10, VCPU_GUEST_PID(r4)
mtspr SPRN_PID, r8
#ifdef CONFIG_KVM_EXIT_TIMING
/* save exit time */
1: mfspr r7, SPRN_TBRU
mfspr r8, SPRN_TBRL
mfspr r9, SPRN_TBRU
cmpw r9, r7
KVM: PPC: bookehv: Use lwz/stw instead of PPC_LL/PPC_STL for 32-bit fields Interrupt code used PPC_LL/PPC_STL macros to load/store some of u32 fields which led to memory overflow on 64-bit. Use lwz/stw instead. Signed-off-by: Mihai Caraman <mihai.caraman@freescale.com> Signed-off-by: Alexander Graf <agraf@suse.de>
2012-04-16 08:08:54 +04:00
stw r8, VCPU_TIMING_EXIT_TBL(r4)
KVM: PPC: booke: category E.HV (GS-mode) support Chips such as e500mc that implement category E.HV in Power ISA 2.06 provide hardware virtualization features, including a new MSR mode for guest state. The guest OS can perform many operations without trapping into the hypervisor, including transitions to and from guest userspace. Since we can use SRR1[GS] to reliably tell whether an exception came from guest state, instead of messing around with IVPR, we use DO_KVM similarly to book3s. Current issues include: - Machine checks from guest state are not routed to the host handler. - The guest can cause a host oops by executing an emulated instruction in a page that lacks read permission. Existing e500/4xx support has the same problem. Includes work by Ashish Kalra <Ashish.Kalra@freescale.com>, Varun Sethi <Varun.Sethi@freescale.com>, and Liu Yu <yu.liu@freescale.com>. Signed-off-by: Scott Wood <scottwood@freescale.com> [agraf: remove pt_regs usage] Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Avi Kivity <avi@redhat.com>
2011-12-20 19:34:43 +04:00
bne- 1b
KVM: PPC: bookehv: Use lwz/stw instead of PPC_LL/PPC_STL for 32-bit fields Interrupt code used PPC_LL/PPC_STL macros to load/store some of u32 fields which led to memory overflow on 64-bit. Use lwz/stw instead. Signed-off-by: Mihai Caraman <mihai.caraman@freescale.com> Signed-off-by: Alexander Graf <agraf@suse.de>
2012-04-16 08:08:54 +04:00
stw r9, VCPU_TIMING_EXIT_TBU(r4)
KVM: PPC: booke: category E.HV (GS-mode) support Chips such as e500mc that implement category E.HV in Power ISA 2.06 provide hardware virtualization features, including a new MSR mode for guest state. The guest OS can perform many operations without trapping into the hypervisor, including transitions to and from guest userspace. Since we can use SRR1[GS] to reliably tell whether an exception came from guest state, instead of messing around with IVPR, we use DO_KVM similarly to book3s. Current issues include: - Machine checks from guest state are not routed to the host handler. - The guest can cause a host oops by executing an emulated instruction in a page that lacks read permission. Existing e500/4xx support has the same problem. Includes work by Ashish Kalra <Ashish.Kalra@freescale.com>, Varun Sethi <Varun.Sethi@freescale.com>, and Liu Yu <yu.liu@freescale.com>. Signed-off-by: Scott Wood <scottwood@freescale.com> [agraf: remove pt_regs usage] Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Avi Kivity <avi@redhat.com>
2011-12-20 19:34:43 +04:00
#endif
oris r8, r6, MSR_CE@h
KVM: PPC: bookehv: Use a Macro for saving/restoring guest registers to/from their 64 bit copies. Introduced PPC_STD/PPC_LD macros for saving/restoring guest registers to/from their 64 bit copies. Signed-off-by: Varun Sethi <Varun.Sethi@freescale.com> Signed-off-by: Alexander Graf <agraf@suse.de>
2012-04-25 05:26:43 +04:00
PPC_STD(r6, VCPU_SHARED_MSR, r11)
KVM: PPC: booke: category E.HV (GS-mode) support Chips such as e500mc that implement category E.HV in Power ISA 2.06 provide hardware virtualization features, including a new MSR mode for guest state. The guest OS can perform many operations without trapping into the hypervisor, including transitions to and from guest userspace. Since we can use SRR1[GS] to reliably tell whether an exception came from guest state, instead of messing around with IVPR, we use DO_KVM similarly to book3s. Current issues include: - Machine checks from guest state are not routed to the host handler. - The guest can cause a host oops by executing an emulated instruction in a page that lacks read permission. Existing e500/4xx support has the same problem. Includes work by Ashish Kalra <Ashish.Kalra@freescale.com>, Varun Sethi <Varun.Sethi@freescale.com>, and Liu Yu <yu.liu@freescale.com>. Signed-off-by: Scott Wood <scottwood@freescale.com> [agraf: remove pt_regs usage] Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Avi Kivity <avi@redhat.com>
2011-12-20 19:34:43 +04:00
ori r8, r8, MSR_ME | MSR_RI
PPC_STL r5, VCPU_PC(r4)
/*
* Make sure CE/ME/RI are set (if appropriate for exception type)
* whether or not the guest had it set. Since mfmsr/mtmsr are
* somewhat expensive, skip in the common case where the guest
* had all these bits set (and thus they're still set if
* appropriate for the exception type).
*/
cmpw r6, r8
beq 1f
mfmsr r7
.if \srr0 != SPRN_MCSRR0 && \srr0 != SPRN_CSRR0
oris r7, r7, MSR_CE@h
.endif
.if \srr0 != SPRN_MCSRR0
ori r7, r7, MSR_ME | MSR_RI
.endif
mtmsr r7
1:
.if \flags & NEED_EMU
/*
* This assumes you have external PID support.
* To support a bookehv CPU without external PID, you'll
* need to look up the TLB entry and create a temporary mapping.
*
* FIXME: we don't currently handle if the lwepx faults. PR-mode
* booke doesn't handle it either. Since Linux doesn't use
* broadcast tlbivax anymore, the only way this should happen is
* if the guest maps its memory execute-but-not-read, or if we
* somehow take a TLB miss in the middle of this entry code and
* evict the relevant entry. On e500mc, all kernel lowmem is
* bolted into TLB1 large page mappings, and we don't use
* broadcast invalidates, so we should not take a TLB miss here.
*
* Later we'll need to deal with faults here. Disallowing guest
* mappings that are execute-but-not-read could be an option on
* e500mc, but not on chips with an LRAT if it is used.
*/
mfspr r3, SPRN_EPLC /* will already have correct ELPID and EGS */
PPC_STL r15, VCPU_GPR(r15)(r4)
PPC_STL r16, VCPU_GPR(r16)(r4)
PPC_STL r17, VCPU_GPR(r17)(r4)
PPC_STL r18, VCPU_GPR(r18)(r4)
PPC_STL r19, VCPU_GPR(r19)(r4)
mr r8, r3
PPC_STL r20, VCPU_GPR(r20)(r4)
rlwimi r8, r6, EPC_EAS_SHIFT - MSR_IR_LG, EPC_EAS
PPC_STL r21, VCPU_GPR(r21)(r4)
rlwimi r8, r6, EPC_EPR_SHIFT - MSR_PR_LG, EPC_EPR
PPC_STL r22, VCPU_GPR(r22)(r4)
rlwimi r8, r10, EPC_EPID_SHIFT, EPC_EPID
PPC_STL r23, VCPU_GPR(r23)(r4)
PPC_STL r24, VCPU_GPR(r24)(r4)
PPC_STL r25, VCPU_GPR(r25)(r4)
PPC_STL r26, VCPU_GPR(r26)(r4)
PPC_STL r27, VCPU_GPR(r27)(r4)
PPC_STL r28, VCPU_GPR(r28)(r4)
PPC_STL r29, VCPU_GPR(r29)(r4)
PPC_STL r30, VCPU_GPR(r30)(r4)
PPC_STL r31, VCPU_GPR(r31)(r4)
mtspr SPRN_EPLC, r8
KVM: PPC: e500mc: add load inst fixup There's always a chance we're unable to read a guest instruction. The guest could have its TLB mapped execute-, but not readable, something odd happens and our TLB gets flushed. So it's a good idea to be prepared for that case and have a fallback that allows us to fix things up in that case. Add fixup code that keeps guest code from potentially crashing our host kernel. Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Avi Kivity <avi@redhat.com>
2012-02-16 03:24:28 +04:00
/* disable preemption, so we are sure we hit the fixup handler */
#ifdef CONFIG_PPC64
clrrdi r8,r1,THREAD_SHIFT
#else
rlwinm r8,r1,0,0,31-THREAD_SHIFT /* current thread_info */
#endif
li r7, 1
stw r7, TI_PREEMPT(r8)
KVM: PPC: booke: category E.HV (GS-mode) support Chips such as e500mc that implement category E.HV in Power ISA 2.06 provide hardware virtualization features, including a new MSR mode for guest state. The guest OS can perform many operations without trapping into the hypervisor, including transitions to and from guest userspace. Since we can use SRR1[GS] to reliably tell whether an exception came from guest state, instead of messing around with IVPR, we use DO_KVM similarly to book3s. Current issues include: - Machine checks from guest state are not routed to the host handler. - The guest can cause a host oops by executing an emulated instruction in a page that lacks read permission. Existing e500/4xx support has the same problem. Includes work by Ashish Kalra <Ashish.Kalra@freescale.com>, Varun Sethi <Varun.Sethi@freescale.com>, and Liu Yu <yu.liu@freescale.com>. Signed-off-by: Scott Wood <scottwood@freescale.com> [agraf: remove pt_regs usage] Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Avi Kivity <avi@redhat.com>
2011-12-20 19:34:43 +04:00
isync
KVM: PPC: e500mc: add load inst fixup There's always a chance we're unable to read a guest instruction. The guest could have its TLB mapped execute-, but not readable, something odd happens and our TLB gets flushed. So it's a good idea to be prepared for that case and have a fallback that allows us to fix things up in that case. Add fixup code that keeps guest code from potentially crashing our host kernel. Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Avi Kivity <avi@redhat.com>
2012-02-16 03:24:28 +04:00
/*
* In case the read goes wrong, we catch it and write an invalid value
* in LAST_INST instead.
*/
1: lwepx r9, 0, r5
2:
.section .fixup, "ax"
3: li r9, KVM_INST_FETCH_FAILED
b 2b
.previous
.section __ex_table,"a"
PPC_LONG_ALIGN
PPC_LONG 1b,3b
.previous
KVM: PPC: booke: category E.HV (GS-mode) support Chips such as e500mc that implement category E.HV in Power ISA 2.06 provide hardware virtualization features, including a new MSR mode for guest state. The guest OS can perform many operations without trapping into the hypervisor, including transitions to and from guest userspace. Since we can use SRR1[GS] to reliably tell whether an exception came from guest state, instead of messing around with IVPR, we use DO_KVM similarly to book3s. Current issues include: - Machine checks from guest state are not routed to the host handler. - The guest can cause a host oops by executing an emulated instruction in a page that lacks read permission. Existing e500/4xx support has the same problem. Includes work by Ashish Kalra <Ashish.Kalra@freescale.com>, Varun Sethi <Varun.Sethi@freescale.com>, and Liu Yu <yu.liu@freescale.com>. Signed-off-by: Scott Wood <scottwood@freescale.com> [agraf: remove pt_regs usage] Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Avi Kivity <avi@redhat.com>
2011-12-20 19:34:43 +04:00
mtspr SPRN_EPLC, r3
KVM: PPC: e500mc: add load inst fixup There's always a chance we're unable to read a guest instruction. The guest could have its TLB mapped execute-, but not readable, something odd happens and our TLB gets flushed. So it's a good idea to be prepared for that case and have a fallback that allows us to fix things up in that case. Add fixup code that keeps guest code from potentially crashing our host kernel. Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Avi Kivity <avi@redhat.com>
2012-02-16 03:24:28 +04:00
li r7, 0
stw r7, TI_PREEMPT(r8)
KVM: PPC: booke: category E.HV (GS-mode) support Chips such as e500mc that implement category E.HV in Power ISA 2.06 provide hardware virtualization features, including a new MSR mode for guest state. The guest OS can perform many operations without trapping into the hypervisor, including transitions to and from guest userspace. Since we can use SRR1[GS] to reliably tell whether an exception came from guest state, instead of messing around with IVPR, we use DO_KVM similarly to book3s. Current issues include: - Machine checks from guest state are not routed to the host handler. - The guest can cause a host oops by executing an emulated instruction in a page that lacks read permission. Existing e500/4xx support has the same problem. Includes work by Ashish Kalra <Ashish.Kalra@freescale.com>, Varun Sethi <Varun.Sethi@freescale.com>, and Liu Yu <yu.liu@freescale.com>. Signed-off-by: Scott Wood <scottwood@freescale.com> [agraf: remove pt_regs usage] Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Avi Kivity <avi@redhat.com>
2011-12-20 19:34:43 +04:00
stw r9, VCPU_LAST_INST(r4)
.endif
.if \flags & NEED_ESR
mfspr r8, SPRN_ESR
PPC_STL r8, VCPU_FAULT_ESR(r4)
.endif
.if \flags & NEED_DEAR
mfspr r9, SPRN_DEAR
PPC_STL r9, VCPU_FAULT_DEAR(r4)
.endif
b kvmppc_resume_host
.endm
/*
* For input register values, see arch/powerpc/include/asm/kvm_booke_hv_asm.h
*/
.macro kvm_handler intno srr0, srr1, flags
_GLOBAL(kvmppc_handler_\intno\()_\srr1)
GET_VCPU(r11, r10)
PPC_STL r3, VCPU_GPR(r3)(r11)
mfspr r3, SPRN_SPRG_RSCRATCH0
PPC_STL r4, VCPU_GPR(r4)(r11)
PPC_LL r4, THREAD_NORMSAVE(0)(r10)
PPC_STL r5, VCPU_GPR(r5)(r11)
KVM: PPC: bookehv: Use lwz/stw instead of PPC_LL/PPC_STL for 32-bit fields Interrupt code used PPC_LL/PPC_STL macros to load/store some of u32 fields which led to memory overflow on 64-bit. Use lwz/stw instead. Signed-off-by: Mihai Caraman <mihai.caraman@freescale.com> Signed-off-by: Alexander Graf <agraf@suse.de>
2012-04-16 08:08:54 +04:00
stw r13, VCPU_CR(r11)
KVM: PPC: booke: category E.HV (GS-mode) support Chips such as e500mc that implement category E.HV in Power ISA 2.06 provide hardware virtualization features, including a new MSR mode for guest state. The guest OS can perform many operations without trapping into the hypervisor, including transitions to and from guest userspace. Since we can use SRR1[GS] to reliably tell whether an exception came from guest state, instead of messing around with IVPR, we use DO_KVM similarly to book3s. Current issues include: - Machine checks from guest state are not routed to the host handler. - The guest can cause a host oops by executing an emulated instruction in a page that lacks read permission. Existing e500/4xx support has the same problem. Includes work by Ashish Kalra <Ashish.Kalra@freescale.com>, Varun Sethi <Varun.Sethi@freescale.com>, and Liu Yu <yu.liu@freescale.com>. Signed-off-by: Scott Wood <scottwood@freescale.com> [agraf: remove pt_regs usage] Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Avi Kivity <avi@redhat.com>
2011-12-20 19:34:43 +04:00
mfspr r5, \srr0
PPC_STL r3, VCPU_GPR(r10)(r11)
PPC_LL r3, THREAD_NORMSAVE(2)(r10)
PPC_STL r6, VCPU_GPR(r6)(r11)
PPC_STL r4, VCPU_GPR(r11)(r11)
mfspr r6, \srr1
PPC_STL r7, VCPU_GPR(r7)(r11)
PPC_STL r8, VCPU_GPR(r8)(r11)
PPC_STL r9, VCPU_GPR(r9)(r11)
PPC_STL r3, VCPU_GPR(r13)(r11)
mfctr r7
PPC_STL r12, VCPU_GPR(r12)(r11)
PPC_STL r7, VCPU_CTR(r11)
mr r4, r11
kvm_handler_common \intno, \srr0, \flags
.endm
.macro kvm_lvl_handler intno scratch srr0, srr1, flags
_GLOBAL(kvmppc_handler_\intno\()_\srr1)
mfspr r10, SPRN_SPRG_THREAD
GET_VCPU(r11, r10)
PPC_STL r3, VCPU_GPR(r3)(r11)
mfspr r3, \scratch
PPC_STL r4, VCPU_GPR(r4)(r11)
PPC_LL r4, GPR9(r8)
PPC_STL r5, VCPU_GPR(r5)(r11)
KVM: PPC: bookehv: Use lwz/stw instead of PPC_LL/PPC_STL for 32-bit fields Interrupt code used PPC_LL/PPC_STL macros to load/store some of u32 fields which led to memory overflow on 64-bit. Use lwz/stw instead. Signed-off-by: Mihai Caraman <mihai.caraman@freescale.com> Signed-off-by: Alexander Graf <agraf@suse.de>
2012-04-16 08:08:54 +04:00
stw r9, VCPU_CR(r11)
KVM: PPC: booke: category E.HV (GS-mode) support Chips such as e500mc that implement category E.HV in Power ISA 2.06 provide hardware virtualization features, including a new MSR mode for guest state. The guest OS can perform many operations without trapping into the hypervisor, including transitions to and from guest userspace. Since we can use SRR1[GS] to reliably tell whether an exception came from guest state, instead of messing around with IVPR, we use DO_KVM similarly to book3s. Current issues include: - Machine checks from guest state are not routed to the host handler. - The guest can cause a host oops by executing an emulated instruction in a page that lacks read permission. Existing e500/4xx support has the same problem. Includes work by Ashish Kalra <Ashish.Kalra@freescale.com>, Varun Sethi <Varun.Sethi@freescale.com>, and Liu Yu <yu.liu@freescale.com>. Signed-off-by: Scott Wood <scottwood@freescale.com> [agraf: remove pt_regs usage] Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Avi Kivity <avi@redhat.com>
2011-12-20 19:34:43 +04:00
mfspr r5, \srr0
PPC_STL r3, VCPU_GPR(r8)(r11)
PPC_LL r3, GPR10(r8)
PPC_STL r6, VCPU_GPR(r6)(r11)
PPC_STL r4, VCPU_GPR(r9)(r11)
mfspr r6, \srr1
PPC_LL r4, GPR11(r8)
PPC_STL r7, VCPU_GPR(r7)(r11)
PPC_STL r3, VCPU_GPR(r10)(r11)
mfctr r7
PPC_STL r12, VCPU_GPR(r12)(r11)
KVM: PPC: bookehv: Fix r8/r13 storing in level exception handler Guest r8 register is held in the scratch register and stored correctly, so remove the instruction that clobbers it. Guest r13 was missing from vcpu, store it there. Signed-off-by: Mihai Caraman <mihai.caraman@freescale.com> Signed-off-by: Alexander Graf <agraf@suse.de>
2012-04-16 08:08:53 +04:00
PPC_STL r13, VCPU_GPR(r13)(r11)
KVM: PPC: booke: category E.HV (GS-mode) support Chips such as e500mc that implement category E.HV in Power ISA 2.06 provide hardware virtualization features, including a new MSR mode for guest state. The guest OS can perform many operations without trapping into the hypervisor, including transitions to and from guest userspace. Since we can use SRR1[GS] to reliably tell whether an exception came from guest state, instead of messing around with IVPR, we use DO_KVM similarly to book3s. Current issues include: - Machine checks from guest state are not routed to the host handler. - The guest can cause a host oops by executing an emulated instruction in a page that lacks read permission. Existing e500/4xx support has the same problem. Includes work by Ashish Kalra <Ashish.Kalra@freescale.com>, Varun Sethi <Varun.Sethi@freescale.com>, and Liu Yu <yu.liu@freescale.com>. Signed-off-by: Scott Wood <scottwood@freescale.com> [agraf: remove pt_regs usage] Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Avi Kivity <avi@redhat.com>
2011-12-20 19:34:43 +04:00
PPC_STL r4, VCPU_GPR(r11)(r11)
PPC_STL r7, VCPU_CTR(r11)
mr r4, r11
kvm_handler_common \intno, \srr0, \flags
.endm
kvm_lvl_handler BOOKE_INTERRUPT_CRITICAL, \
SPRN_SPRG_RSCRATCH_CRIT, SPRN_CSRR0, SPRN_CSRR1, 0
kvm_lvl_handler BOOKE_INTERRUPT_MACHINE_CHECK, \
SPRN_SPRG_RSCRATCH_MC, SPRN_MCSRR0, SPRN_MCSRR1, 0
kvm_handler BOOKE_INTERRUPT_DATA_STORAGE, \
SPRN_SRR0, SPRN_SRR1, (NEED_EMU | NEED_DEAR)
kvm_handler BOOKE_INTERRUPT_INST_STORAGE, SPRN_SRR0, SPRN_SRR1, NEED_ESR
kvm_handler BOOKE_INTERRUPT_EXTERNAL, SPRN_SRR0, SPRN_SRR1, 0
kvm_handler BOOKE_INTERRUPT_ALIGNMENT, \
SPRN_SRR0, SPRN_SRR1, (NEED_DEAR | NEED_ESR)
kvm_handler BOOKE_INTERRUPT_PROGRAM, SPRN_SRR0, SPRN_SRR1, NEED_ESR
kvm_handler BOOKE_INTERRUPT_FP_UNAVAIL, SPRN_SRR0, SPRN_SRR1, 0
kvm_handler BOOKE_INTERRUPT_SYSCALL, SPRN_SRR0, SPRN_SRR1, 0
kvm_handler BOOKE_INTERRUPT_AP_UNAVAIL, SPRN_SRR0, SPRN_SRR1, 0
kvm_handler BOOKE_INTERRUPT_DECREMENTER, SPRN_SRR0, SPRN_SRR1, 0
kvm_handler BOOKE_INTERRUPT_FIT, SPRN_SRR0, SPRN_SRR1, 0
kvm_lvl_handler BOOKE_INTERRUPT_WATCHDOG, \
SPRN_SPRG_RSCRATCH_CRIT, SPRN_CSRR0, SPRN_CSRR1, 0
kvm_handler BOOKE_INTERRUPT_DTLB_MISS, \
SPRN_SRR0, SPRN_SRR1, (NEED_EMU | NEED_DEAR | NEED_ESR)
kvm_handler BOOKE_INTERRUPT_ITLB_MISS, SPRN_SRR0, SPRN_SRR1, 0
kvm_handler BOOKE_INTERRUPT_SPE_UNAVAIL, SPRN_SRR0, SPRN_SRR1, 0
kvm_handler BOOKE_INTERRUPT_SPE_FP_DATA, SPRN_SRR0, SPRN_SRR1, 0
kvm_handler BOOKE_INTERRUPT_SPE_FP_ROUND, SPRN_SRR0, SPRN_SRR1, 0
kvm_handler BOOKE_INTERRUPT_PERFORMANCE_MONITOR, SPRN_SRR0, SPRN_SRR1, 0
kvm_handler BOOKE_INTERRUPT_DOORBELL, SPRN_SRR0, SPRN_SRR1, 0
kvm_lvl_handler BOOKE_INTERRUPT_DOORBELL_CRITICAL, \
SPRN_SPRG_RSCRATCH_CRIT, SPRN_CSRR0, SPRN_CSRR1, 0
kvm_handler BOOKE_INTERRUPT_HV_PRIV, SPRN_SRR0, SPRN_SRR1, NEED_EMU
kvm_handler BOOKE_INTERRUPT_HV_SYSCALL, SPRN_SRR0, SPRN_SRR1, 0
kvm_handler BOOKE_INTERRUPT_GUEST_DBELL, SPRN_GSRR0, SPRN_GSRR1, 0
kvm_lvl_handler BOOKE_INTERRUPT_GUEST_DBELL_CRIT, \
SPRN_SPRG_RSCRATCH_CRIT, SPRN_CSRR0, SPRN_CSRR1, 0
kvm_lvl_handler BOOKE_INTERRUPT_DEBUG, \
SPRN_SPRG_RSCRATCH_CRIT, SPRN_CSRR0, SPRN_CSRR1, 0
kvm_lvl_handler BOOKE_INTERRUPT_DEBUG, \
SPRN_SPRG_RSCRATCH_DBG, SPRN_DSRR0, SPRN_DSRR1, 0
/* Registers:
* SPRG_SCRATCH0: guest r10
* r4: vcpu pointer
* r11: vcpu->arch.shared
* r14: KVM exit number
*/
_GLOBAL(kvmppc_resume_host)
/* Save remaining volatile guest register state to vcpu. */
mfspr r3, SPRN_VRSAVE
PPC_STL r0, VCPU_GPR(r0)(r4)
mflr r5
mfspr r6, SPRN_SPRG4
PPC_STL r5, VCPU_LR(r4)
mfspr r7, SPRN_SPRG5
KVM: PPC: bookehv: Use lwz/stw instead of PPC_LL/PPC_STL for 32-bit fields Interrupt code used PPC_LL/PPC_STL macros to load/store some of u32 fields which led to memory overflow on 64-bit. Use lwz/stw instead. Signed-off-by: Mihai Caraman <mihai.caraman@freescale.com> Signed-off-by: Alexander Graf <agraf@suse.de>
2012-04-16 08:08:54 +04:00
stw r3, VCPU_VRSAVE(r4)
KVM: PPC: booke(hv): Fix save/restore of guest accessible SPRGs. For Guest accessible SPRGs 4-7, save/restore must be handled differently for 64bit and non-64 bit case. Use the PPC_STD/PPC_LD macros for saving/restoring to/from these registers. Signed-off-by: Varun Sethi <Varun.Sethi@freescale.com> Signed-off-by: Alexander Graf <agraf@suse.de>
2012-04-25 05:27:34 +04:00
PPC_STD(r6, VCPU_SHARED_SPRG4, r11)
KVM: PPC: booke: category E.HV (GS-mode) support Chips such as e500mc that implement category E.HV in Power ISA 2.06 provide hardware virtualization features, including a new MSR mode for guest state. The guest OS can perform many operations without trapping into the hypervisor, including transitions to and from guest userspace. Since we can use SRR1[GS] to reliably tell whether an exception came from guest state, instead of messing around with IVPR, we use DO_KVM similarly to book3s. Current issues include: - Machine checks from guest state are not routed to the host handler. - The guest can cause a host oops by executing an emulated instruction in a page that lacks read permission. Existing e500/4xx support has the same problem. Includes work by Ashish Kalra <Ashish.Kalra@freescale.com>, Varun Sethi <Varun.Sethi@freescale.com>, and Liu Yu <yu.liu@freescale.com>. Signed-off-by: Scott Wood <scottwood@freescale.com> [agraf: remove pt_regs usage] Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Avi Kivity <avi@redhat.com>
2011-12-20 19:34:43 +04:00
mfspr r8, SPRN_SPRG6
KVM: PPC: booke(hv): Fix save/restore of guest accessible SPRGs. For Guest accessible SPRGs 4-7, save/restore must be handled differently for 64bit and non-64 bit case. Use the PPC_STD/PPC_LD macros for saving/restoring to/from these registers. Signed-off-by: Varun Sethi <Varun.Sethi@freescale.com> Signed-off-by: Alexander Graf <agraf@suse.de>
2012-04-25 05:27:34 +04:00
PPC_STD(r7, VCPU_SHARED_SPRG5, r11)
KVM: PPC: booke: category E.HV (GS-mode) support Chips such as e500mc that implement category E.HV in Power ISA 2.06 provide hardware virtualization features, including a new MSR mode for guest state. The guest OS can perform many operations without trapping into the hypervisor, including transitions to and from guest userspace. Since we can use SRR1[GS] to reliably tell whether an exception came from guest state, instead of messing around with IVPR, we use DO_KVM similarly to book3s. Current issues include: - Machine checks from guest state are not routed to the host handler. - The guest can cause a host oops by executing an emulated instruction in a page that lacks read permission. Existing e500/4xx support has the same problem. Includes work by Ashish Kalra <Ashish.Kalra@freescale.com>, Varun Sethi <Varun.Sethi@freescale.com>, and Liu Yu <yu.liu@freescale.com>. Signed-off-by: Scott Wood <scottwood@freescale.com> [agraf: remove pt_regs usage] Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Avi Kivity <avi@redhat.com>
2011-12-20 19:34:43 +04:00
mfspr r9, SPRN_SPRG7
KVM: PPC: booke(hv): Fix save/restore of guest accessible SPRGs. For Guest accessible SPRGs 4-7, save/restore must be handled differently for 64bit and non-64 bit case. Use the PPC_STD/PPC_LD macros for saving/restoring to/from these registers. Signed-off-by: Varun Sethi <Varun.Sethi@freescale.com> Signed-off-by: Alexander Graf <agraf@suse.de>
2012-04-25 05:27:34 +04:00
PPC_STD(r8, VCPU_SHARED_SPRG6, r11)
KVM: PPC: booke: category E.HV (GS-mode) support Chips such as e500mc that implement category E.HV in Power ISA 2.06 provide hardware virtualization features, including a new MSR mode for guest state. The guest OS can perform many operations without trapping into the hypervisor, including transitions to and from guest userspace. Since we can use SRR1[GS] to reliably tell whether an exception came from guest state, instead of messing around with IVPR, we use DO_KVM similarly to book3s. Current issues include: - Machine checks from guest state are not routed to the host handler. - The guest can cause a host oops by executing an emulated instruction in a page that lacks read permission. Existing e500/4xx support has the same problem. Includes work by Ashish Kalra <Ashish.Kalra@freescale.com>, Varun Sethi <Varun.Sethi@freescale.com>, and Liu Yu <yu.liu@freescale.com>. Signed-off-by: Scott Wood <scottwood@freescale.com> [agraf: remove pt_regs usage] Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Avi Kivity <avi@redhat.com>
2011-12-20 19:34:43 +04:00
mfxer r3
KVM: PPC: booke(hv): Fix save/restore of guest accessible SPRGs. For Guest accessible SPRGs 4-7, save/restore must be handled differently for 64bit and non-64 bit case. Use the PPC_STD/PPC_LD macros for saving/restoring to/from these registers. Signed-off-by: Varun Sethi <Varun.Sethi@freescale.com> Signed-off-by: Alexander Graf <agraf@suse.de>
2012-04-25 05:27:34 +04:00
PPC_STD(r9, VCPU_SHARED_SPRG7, r11)
KVM: PPC: booke: category E.HV (GS-mode) support Chips such as e500mc that implement category E.HV in Power ISA 2.06 provide hardware virtualization features, including a new MSR mode for guest state. The guest OS can perform many operations without trapping into the hypervisor, including transitions to and from guest userspace. Since we can use SRR1[GS] to reliably tell whether an exception came from guest state, instead of messing around with IVPR, we use DO_KVM similarly to book3s. Current issues include: - Machine checks from guest state are not routed to the host handler. - The guest can cause a host oops by executing an emulated instruction in a page that lacks read permission. Existing e500/4xx support has the same problem. Includes work by Ashish Kalra <Ashish.Kalra@freescale.com>, Varun Sethi <Varun.Sethi@freescale.com>, and Liu Yu <yu.liu@freescale.com>. Signed-off-by: Scott Wood <scottwood@freescale.com> [agraf: remove pt_regs usage] Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Avi Kivity <avi@redhat.com>
2011-12-20 19:34:43 +04:00
/* save guest MAS registers and restore host mas4 & mas6 */
mfspr r5, SPRN_MAS0
PPC_STL r3, VCPU_XER(r4)
mfspr r6, SPRN_MAS1
stw r5, VCPU_SHARED_MAS0(r11)
mfspr r7, SPRN_MAS2
stw r6, VCPU_SHARED_MAS1(r11)
KVM: PPC: bookehv: Use a Macro for saving/restoring guest registers to/from their 64 bit copies. Introduced PPC_STD/PPC_LD macros for saving/restoring guest registers to/from their 64 bit copies. Signed-off-by: Varun Sethi <Varun.Sethi@freescale.com> Signed-off-by: Alexander Graf <agraf@suse.de>
2012-04-25 05:26:43 +04:00
PPC_STD(r7, VCPU_SHARED_MAS2, r11)
KVM: PPC: booke: category E.HV (GS-mode) support Chips such as e500mc that implement category E.HV in Power ISA 2.06 provide hardware virtualization features, including a new MSR mode for guest state. The guest OS can perform many operations without trapping into the hypervisor, including transitions to and from guest userspace. Since we can use SRR1[GS] to reliably tell whether an exception came from guest state, instead of messing around with IVPR, we use DO_KVM similarly to book3s. Current issues include: - Machine checks from guest state are not routed to the host handler. - The guest can cause a host oops by executing an emulated instruction in a page that lacks read permission. Existing e500/4xx support has the same problem. Includes work by Ashish Kalra <Ashish.Kalra@freescale.com>, Varun Sethi <Varun.Sethi@freescale.com>, and Liu Yu <yu.liu@freescale.com>. Signed-off-by: Scott Wood <scottwood@freescale.com> [agraf: remove pt_regs usage] Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Avi Kivity <avi@redhat.com>
2011-12-20 19:34:43 +04:00
mfspr r5, SPRN_MAS3
mfspr r6, SPRN_MAS4
stw r5, VCPU_SHARED_MAS7_3+4(r11)
mfspr r7, SPRN_MAS6
stw r6, VCPU_SHARED_MAS4(r11)
mfspr r5, SPRN_MAS7
lwz r6, VCPU_HOST_MAS4(r4)
stw r7, VCPU_SHARED_MAS6(r11)
lwz r8, VCPU_HOST_MAS6(r4)
mtspr SPRN_MAS4, r6
stw r5, VCPU_SHARED_MAS7_3+0(r11)
mtspr SPRN_MAS6, r8
KVM: PPC: bookehv: disable MAS register updates early We need to make sure that no MAS updates happen automatically while we have the guest MAS registers loaded. So move the disabling code a bit higher up so that it covers the full time we have guest values in MAS registers. The race this patch fixes should never occur, but it makes the code a bit more logical to do it this way around. Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Avi Kivity <avi@redhat.com>
2012-02-16 18:53:04 +04:00
/* Enable MAS register updates via exception */
KVM: PPC: booke: category E.HV (GS-mode) support Chips such as e500mc that implement category E.HV in Power ISA 2.06 provide hardware virtualization features, including a new MSR mode for guest state. The guest OS can perform many operations without trapping into the hypervisor, including transitions to and from guest userspace. Since we can use SRR1[GS] to reliably tell whether an exception came from guest state, instead of messing around with IVPR, we use DO_KVM similarly to book3s. Current issues include: - Machine checks from guest state are not routed to the host handler. - The guest can cause a host oops by executing an emulated instruction in a page that lacks read permission. Existing e500/4xx support has the same problem. Includes work by Ashish Kalra <Ashish.Kalra@freescale.com>, Varun Sethi <Varun.Sethi@freescale.com>, and Liu Yu <yu.liu@freescale.com>. Signed-off-by: Scott Wood <scottwood@freescale.com> [agraf: remove pt_regs usage] Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Avi Kivity <avi@redhat.com>
2011-12-20 19:34:43 +04:00
mfspr r3, SPRN_EPCR
rlwinm r3, r3, 0, ~SPRN_EPCR_DMIUH
mtspr SPRN_EPCR, r3
isync
/* Switch to kernel stack and jump to handler. */
PPC_LL r3, HOST_RUN(r1)
mr r5, r14 /* intno */
mr r14, r4 /* Save vcpu pointer. */
bl kvmppc_handle_exit
/* Restore vcpu pointer and the nonvolatiles we used. */
mr r4, r14
PPC_LL r14, VCPU_GPR(r14)(r4)
andi. r5, r3, RESUME_FLAG_NV
beq skip_nv_load
PPC_LL r15, VCPU_GPR(r15)(r4)
PPC_LL r16, VCPU_GPR(r16)(r4)
PPC_LL r17, VCPU_GPR(r17)(r4)
PPC_LL r18, VCPU_GPR(r18)(r4)
PPC_LL r19, VCPU_GPR(r19)(r4)
PPC_LL r20, VCPU_GPR(r20)(r4)
PPC_LL r21, VCPU_GPR(r21)(r4)
PPC_LL r22, VCPU_GPR(r22)(r4)
PPC_LL r23, VCPU_GPR(r23)(r4)
PPC_LL r24, VCPU_GPR(r24)(r4)
PPC_LL r25, VCPU_GPR(r25)(r4)
PPC_LL r26, VCPU_GPR(r26)(r4)
PPC_LL r27, VCPU_GPR(r27)(r4)
PPC_LL r28, VCPU_GPR(r28)(r4)
PPC_LL r29, VCPU_GPR(r29)(r4)
PPC_LL r30, VCPU_GPR(r30)(r4)
PPC_LL r31, VCPU_GPR(r31)(r4)
skip_nv_load:
/* Should we return to the guest? */
andi. r5, r3, RESUME_FLAG_HOST
beq lightweight_exit
srawi r3, r3, 2 /* Shift -ERR back down. */
heavyweight_exit:
/* Not returning to guest. */
PPC_LL r5, HOST_STACK_LR(r1)
KVM: PPC: Save/Restore CR over vcpu_run On PPC, CR2-CR4 are nonvolatile, thus have to be saved across function calls. We didn't respect that for any architecture until Paul spotted it in his patch for Book3S-HV. This patch saves/restores CR for all KVM capable PPC hosts. Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Avi Kivity <avi@redhat.com>
2012-03-05 19:00:28 +04:00
lwz r6, HOST_CR(r1)
KVM: PPC: booke: category E.HV (GS-mode) support Chips such as e500mc that implement category E.HV in Power ISA 2.06 provide hardware virtualization features, including a new MSR mode for guest state. The guest OS can perform many operations without trapping into the hypervisor, including transitions to and from guest userspace. Since we can use SRR1[GS] to reliably tell whether an exception came from guest state, instead of messing around with IVPR, we use DO_KVM similarly to book3s. Current issues include: - Machine checks from guest state are not routed to the host handler. - The guest can cause a host oops by executing an emulated instruction in a page that lacks read permission. Existing e500/4xx support has the same problem. Includes work by Ashish Kalra <Ashish.Kalra@freescale.com>, Varun Sethi <Varun.Sethi@freescale.com>, and Liu Yu <yu.liu@freescale.com>. Signed-off-by: Scott Wood <scottwood@freescale.com> [agraf: remove pt_regs usage] Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Avi Kivity <avi@redhat.com>
2011-12-20 19:34:43 +04:00
/*
* We already saved guest volatile register state; now save the
* non-volatiles.
*/
PPC_STL r15, VCPU_GPR(r15)(r4)
PPC_STL r16, VCPU_GPR(r16)(r4)
PPC_STL r17, VCPU_GPR(r17)(r4)
PPC_STL r18, VCPU_GPR(r18)(r4)
PPC_STL r19, VCPU_GPR(r19)(r4)
PPC_STL r20, VCPU_GPR(r20)(r4)
PPC_STL r21, VCPU_GPR(r21)(r4)
PPC_STL r22, VCPU_GPR(r22)(r4)
PPC_STL r23, VCPU_GPR(r23)(r4)
PPC_STL r24, VCPU_GPR(r24)(r4)
PPC_STL r25, VCPU_GPR(r25)(r4)
PPC_STL r26, VCPU_GPR(r26)(r4)
PPC_STL r27, VCPU_GPR(r27)(r4)
PPC_STL r28, VCPU_GPR(r28)(r4)
PPC_STL r29, VCPU_GPR(r29)(r4)
PPC_STL r30, VCPU_GPR(r30)(r4)
PPC_STL r31, VCPU_GPR(r31)(r4)
/* Load host non-volatile register state from host stack. */
PPC_LL r14, HOST_NV_GPR(r14)(r1)
PPC_LL r15, HOST_NV_GPR(r15)(r1)
PPC_LL r16, HOST_NV_GPR(r16)(r1)
PPC_LL r17, HOST_NV_GPR(r17)(r1)
PPC_LL r18, HOST_NV_GPR(r18)(r1)
PPC_LL r19, HOST_NV_GPR(r19)(r1)
PPC_LL r20, HOST_NV_GPR(r20)(r1)
PPC_LL r21, HOST_NV_GPR(r21)(r1)
PPC_LL r22, HOST_NV_GPR(r22)(r1)
PPC_LL r23, HOST_NV_GPR(r23)(r1)
PPC_LL r24, HOST_NV_GPR(r24)(r1)
PPC_LL r25, HOST_NV_GPR(r25)(r1)
PPC_LL r26, HOST_NV_GPR(r26)(r1)
PPC_LL r27, HOST_NV_GPR(r27)(r1)
PPC_LL r28, HOST_NV_GPR(r28)(r1)
PPC_LL r29, HOST_NV_GPR(r29)(r1)
PPC_LL r30, HOST_NV_GPR(r30)(r1)
PPC_LL r31, HOST_NV_GPR(r31)(r1)
/* Return to kvm_vcpu_run(). */
mtlr r5
KVM: PPC: Save/Restore CR over vcpu_run On PPC, CR2-CR4 are nonvolatile, thus have to be saved across function calls. We didn't respect that for any architecture until Paul spotted it in his patch for Book3S-HV. This patch saves/restores CR for all KVM capable PPC hosts. Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Avi Kivity <avi@redhat.com>
2012-03-05 19:00:28 +04:00
mtcr r6
KVM: PPC: booke: category E.HV (GS-mode) support Chips such as e500mc that implement category E.HV in Power ISA 2.06 provide hardware virtualization features, including a new MSR mode for guest state. The guest OS can perform many operations without trapping into the hypervisor, including transitions to and from guest userspace. Since we can use SRR1[GS] to reliably tell whether an exception came from guest state, instead of messing around with IVPR, we use DO_KVM similarly to book3s. Current issues include: - Machine checks from guest state are not routed to the host handler. - The guest can cause a host oops by executing an emulated instruction in a page that lacks read permission. Existing e500/4xx support has the same problem. Includes work by Ashish Kalra <Ashish.Kalra@freescale.com>, Varun Sethi <Varun.Sethi@freescale.com>, and Liu Yu <yu.liu@freescale.com>. Signed-off-by: Scott Wood <scottwood@freescale.com> [agraf: remove pt_regs usage] Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Avi Kivity <avi@redhat.com>
2011-12-20 19:34:43 +04:00
addi r1, r1, HOST_STACK_SIZE
/* r3 still contains the return code from kvmppc_handle_exit(). */
blr
/* Registers:
* r3: kvm_run pointer
* r4: vcpu pointer
*/
_GLOBAL(__kvmppc_vcpu_run)
stwu r1, -HOST_STACK_SIZE(r1)
PPC_STL r1, VCPU_HOST_STACK(r4) /* Save stack pointer to vcpu. */
/* Save host state to stack. */
PPC_STL r3, HOST_RUN(r1)
mflr r3
KVM: PPC: Save/Restore CR over vcpu_run On PPC, CR2-CR4 are nonvolatile, thus have to be saved across function calls. We didn't respect that for any architecture until Paul spotted it in his patch for Book3S-HV. This patch saves/restores CR for all KVM capable PPC hosts. Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Avi Kivity <avi@redhat.com>
2012-03-05 19:00:28 +04:00
mfcr r5
KVM: PPC: booke: category E.HV (GS-mode) support Chips such as e500mc that implement category E.HV in Power ISA 2.06 provide hardware virtualization features, including a new MSR mode for guest state. The guest OS can perform many operations without trapping into the hypervisor, including transitions to and from guest userspace. Since we can use SRR1[GS] to reliably tell whether an exception came from guest state, instead of messing around with IVPR, we use DO_KVM similarly to book3s. Current issues include: - Machine checks from guest state are not routed to the host handler. - The guest can cause a host oops by executing an emulated instruction in a page that lacks read permission. Existing e500/4xx support has the same problem. Includes work by Ashish Kalra <Ashish.Kalra@freescale.com>, Varun Sethi <Varun.Sethi@freescale.com>, and Liu Yu <yu.liu@freescale.com>. Signed-off-by: Scott Wood <scottwood@freescale.com> [agraf: remove pt_regs usage] Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Avi Kivity <avi@redhat.com>
2011-12-20 19:34:43 +04:00
PPC_STL r3, HOST_STACK_LR(r1)
KVM: PPC: Save/Restore CR over vcpu_run On PPC, CR2-CR4 are nonvolatile, thus have to be saved across function calls. We didn't respect that for any architecture until Paul spotted it in his patch for Book3S-HV. This patch saves/restores CR for all KVM capable PPC hosts. Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Avi Kivity <avi@redhat.com>
2012-03-05 19:00:28 +04:00
stw r5, HOST_CR(r1)
KVM: PPC: booke: category E.HV (GS-mode) support Chips such as e500mc that implement category E.HV in Power ISA 2.06 provide hardware virtualization features, including a new MSR mode for guest state. The guest OS can perform many operations without trapping into the hypervisor, including transitions to and from guest userspace. Since we can use SRR1[GS] to reliably tell whether an exception came from guest state, instead of messing around with IVPR, we use DO_KVM similarly to book3s. Current issues include: - Machine checks from guest state are not routed to the host handler. - The guest can cause a host oops by executing an emulated instruction in a page that lacks read permission. Existing e500/4xx support has the same problem. Includes work by Ashish Kalra <Ashish.Kalra@freescale.com>, Varun Sethi <Varun.Sethi@freescale.com>, and Liu Yu <yu.liu@freescale.com>. Signed-off-by: Scott Wood <scottwood@freescale.com> [agraf: remove pt_regs usage] Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Avi Kivity <avi@redhat.com>
2011-12-20 19:34:43 +04:00
/* Save host non-volatile register state to stack. */
PPC_STL r14, HOST_NV_GPR(r14)(r1)
PPC_STL r15, HOST_NV_GPR(r15)(r1)
PPC_STL r16, HOST_NV_GPR(r16)(r1)
PPC_STL r17, HOST_NV_GPR(r17)(r1)
PPC_STL r18, HOST_NV_GPR(r18)(r1)
PPC_STL r19, HOST_NV_GPR(r19)(r1)
PPC_STL r20, HOST_NV_GPR(r20)(r1)
PPC_STL r21, HOST_NV_GPR(r21)(r1)
PPC_STL r22, HOST_NV_GPR(r22)(r1)
PPC_STL r23, HOST_NV_GPR(r23)(r1)
PPC_STL r24, HOST_NV_GPR(r24)(r1)
PPC_STL r25, HOST_NV_GPR(r25)(r1)
PPC_STL r26, HOST_NV_GPR(r26)(r1)
PPC_STL r27, HOST_NV_GPR(r27)(r1)
PPC_STL r28, HOST_NV_GPR(r28)(r1)
PPC_STL r29, HOST_NV_GPR(r29)(r1)
PPC_STL r30, HOST_NV_GPR(r30)(r1)
PPC_STL r31, HOST_NV_GPR(r31)(r1)
/* Load guest non-volatiles. */
PPC_LL r14, VCPU_GPR(r14)(r4)
PPC_LL r15, VCPU_GPR(r15)(r4)
PPC_LL r16, VCPU_GPR(r16)(r4)
PPC_LL r17, VCPU_GPR(r17)(r4)
PPC_LL r18, VCPU_GPR(r18)(r4)
PPC_LL r19, VCPU_GPR(r19)(r4)
PPC_LL r20, VCPU_GPR(r20)(r4)
PPC_LL r21, VCPU_GPR(r21)(r4)
PPC_LL r22, VCPU_GPR(r22)(r4)
PPC_LL r23, VCPU_GPR(r23)(r4)
PPC_LL r24, VCPU_GPR(r24)(r4)
PPC_LL r25, VCPU_GPR(r25)(r4)
PPC_LL r26, VCPU_GPR(r26)(r4)
PPC_LL r27, VCPU_GPR(r27)(r4)
PPC_LL r28, VCPU_GPR(r28)(r4)
PPC_LL r29, VCPU_GPR(r29)(r4)
PPC_LL r30, VCPU_GPR(r30)(r4)
PPC_LL r31, VCPU_GPR(r31)(r4)
lightweight_exit:
PPC_STL r2, HOST_R2(r1)
mfspr r3, SPRN_PID
stw r3, VCPU_HOST_PID(r4)
lwz r3, VCPU_GUEST_PID(r4)
mtspr SPRN_PID, r3
PPC_LL r11, VCPU_SHARED(r4)
KVM: PPC: bookehv: disable MAS register updates early We need to make sure that no MAS updates happen automatically while we have the guest MAS registers loaded. So move the disabling code a bit higher up so that it covers the full time we have guest values in MAS registers. The race this patch fixes should never occur, but it makes the code a bit more logical to do it this way around. Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Avi Kivity <avi@redhat.com>
2012-02-16 18:53:04 +04:00
/* Disable MAS register updates via exception */
mfspr r3, SPRN_EPCR
oris r3, r3, SPRN_EPCR_DMIUH@h
mtspr SPRN_EPCR, r3
isync
KVM: PPC: booke: category E.HV (GS-mode) support Chips such as e500mc that implement category E.HV in Power ISA 2.06 provide hardware virtualization features, including a new MSR mode for guest state. The guest OS can perform many operations without trapping into the hypervisor, including transitions to and from guest userspace. Since we can use SRR1[GS] to reliably tell whether an exception came from guest state, instead of messing around with IVPR, we use DO_KVM similarly to book3s. Current issues include: - Machine checks from guest state are not routed to the host handler. - The guest can cause a host oops by executing an emulated instruction in a page that lacks read permission. Existing e500/4xx support has the same problem. Includes work by Ashish Kalra <Ashish.Kalra@freescale.com>, Varun Sethi <Varun.Sethi@freescale.com>, and Liu Yu <yu.liu@freescale.com>. Signed-off-by: Scott Wood <scottwood@freescale.com> [agraf: remove pt_regs usage] Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Avi Kivity <avi@redhat.com>
2011-12-20 19:34:43 +04:00
/* Save host mas4 and mas6 and load guest MAS registers */
mfspr r3, SPRN_MAS4
stw r3, VCPU_HOST_MAS4(r4)
mfspr r3, SPRN_MAS6
stw r3, VCPU_HOST_MAS6(r4)
lwz r3, VCPU_SHARED_MAS0(r11)
lwz r5, VCPU_SHARED_MAS1(r11)
KVM: PPC: bookehv: Use a Macro for saving/restoring guest registers to/from their 64 bit copies. Introduced PPC_STD/PPC_LD macros for saving/restoring guest registers to/from their 64 bit copies. Signed-off-by: Varun Sethi <Varun.Sethi@freescale.com> Signed-off-by: Alexander Graf <agraf@suse.de>
2012-04-25 05:26:43 +04:00
PPC_LD(r6, VCPU_SHARED_MAS2, r11)
KVM: PPC: booke: category E.HV (GS-mode) support Chips such as e500mc that implement category E.HV in Power ISA 2.06 provide hardware virtualization features, including a new MSR mode for guest state. The guest OS can perform many operations without trapping into the hypervisor, including transitions to and from guest userspace. Since we can use SRR1[GS] to reliably tell whether an exception came from guest state, instead of messing around with IVPR, we use DO_KVM similarly to book3s. Current issues include: - Machine checks from guest state are not routed to the host handler. - The guest can cause a host oops by executing an emulated instruction in a page that lacks read permission. Existing e500/4xx support has the same problem. Includes work by Ashish Kalra <Ashish.Kalra@freescale.com>, Varun Sethi <Varun.Sethi@freescale.com>, and Liu Yu <yu.liu@freescale.com>. Signed-off-by: Scott Wood <scottwood@freescale.com> [agraf: remove pt_regs usage] Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Avi Kivity <avi@redhat.com>
2011-12-20 19:34:43 +04:00
lwz r7, VCPU_SHARED_MAS7_3+4(r11)
lwz r8, VCPU_SHARED_MAS4(r11)
mtspr SPRN_MAS0, r3
mtspr SPRN_MAS1, r5
mtspr SPRN_MAS2, r6
mtspr SPRN_MAS3, r7
mtspr SPRN_MAS4, r8
lwz r3, VCPU_SHARED_MAS6(r11)
lwz r5, VCPU_SHARED_MAS7_3+0(r11)
mtspr SPRN_MAS6, r3
mtspr SPRN_MAS7, r5
/*
* Host interrupt handlers may have clobbered these guest-readable
* SPRGs, so we need to reload them here with the guest's values.
*/
lwz r3, VCPU_VRSAVE(r4)
KVM: PPC: booke(hv): Fix save/restore of guest accessible SPRGs. For Guest accessible SPRGs 4-7, save/restore must be handled differently for 64bit and non-64 bit case. Use the PPC_STD/PPC_LD macros for saving/restoring to/from these registers. Signed-off-by: Varun Sethi <Varun.Sethi@freescale.com> Signed-off-by: Alexander Graf <agraf@suse.de>
2012-04-25 05:27:34 +04:00
PPC_LD(r5, VCPU_SHARED_SPRG4, r11)
KVM: PPC: booke: category E.HV (GS-mode) support Chips such as e500mc that implement category E.HV in Power ISA 2.06 provide hardware virtualization features, including a new MSR mode for guest state. The guest OS can perform many operations without trapping into the hypervisor, including transitions to and from guest userspace. Since we can use SRR1[GS] to reliably tell whether an exception came from guest state, instead of messing around with IVPR, we use DO_KVM similarly to book3s. Current issues include: - Machine checks from guest state are not routed to the host handler. - The guest can cause a host oops by executing an emulated instruction in a page that lacks read permission. Existing e500/4xx support has the same problem. Includes work by Ashish Kalra <Ashish.Kalra@freescale.com>, Varun Sethi <Varun.Sethi@freescale.com>, and Liu Yu <yu.liu@freescale.com>. Signed-off-by: Scott Wood <scottwood@freescale.com> [agraf: remove pt_regs usage] Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Avi Kivity <avi@redhat.com>
2011-12-20 19:34:43 +04:00
mtspr SPRN_VRSAVE, r3
KVM: PPC: booke(hv): Fix save/restore of guest accessible SPRGs. For Guest accessible SPRGs 4-7, save/restore must be handled differently for 64bit and non-64 bit case. Use the PPC_STD/PPC_LD macros for saving/restoring to/from these registers. Signed-off-by: Varun Sethi <Varun.Sethi@freescale.com> Signed-off-by: Alexander Graf <agraf@suse.de>
2012-04-25 05:27:34 +04:00
PPC_LD(r6, VCPU_SHARED_SPRG5, r11)
KVM: PPC: booke: category E.HV (GS-mode) support Chips such as e500mc that implement category E.HV in Power ISA 2.06 provide hardware virtualization features, including a new MSR mode for guest state. The guest OS can perform many operations without trapping into the hypervisor, including transitions to and from guest userspace. Since we can use SRR1[GS] to reliably tell whether an exception came from guest state, instead of messing around with IVPR, we use DO_KVM similarly to book3s. Current issues include: - Machine checks from guest state are not routed to the host handler. - The guest can cause a host oops by executing an emulated instruction in a page that lacks read permission. Existing e500/4xx support has the same problem. Includes work by Ashish Kalra <Ashish.Kalra@freescale.com>, Varun Sethi <Varun.Sethi@freescale.com>, and Liu Yu <yu.liu@freescale.com>. Signed-off-by: Scott Wood <scottwood@freescale.com> [agraf: remove pt_regs usage] Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Avi Kivity <avi@redhat.com>
2011-12-20 19:34:43 +04:00
mtspr SPRN_SPRG4W, r5
KVM: PPC: booke(hv): Fix save/restore of guest accessible SPRGs. For Guest accessible SPRGs 4-7, save/restore must be handled differently for 64bit and non-64 bit case. Use the PPC_STD/PPC_LD macros for saving/restoring to/from these registers. Signed-off-by: Varun Sethi <Varun.Sethi@freescale.com> Signed-off-by: Alexander Graf <agraf@suse.de>
2012-04-25 05:27:34 +04:00
PPC_LD(r7, VCPU_SHARED_SPRG6, r11)
KVM: PPC: booke: category E.HV (GS-mode) support Chips such as e500mc that implement category E.HV in Power ISA 2.06 provide hardware virtualization features, including a new MSR mode for guest state. The guest OS can perform many operations without trapping into the hypervisor, including transitions to and from guest userspace. Since we can use SRR1[GS] to reliably tell whether an exception came from guest state, instead of messing around with IVPR, we use DO_KVM similarly to book3s. Current issues include: - Machine checks from guest state are not routed to the host handler. - The guest can cause a host oops by executing an emulated instruction in a page that lacks read permission. Existing e500/4xx support has the same problem. Includes work by Ashish Kalra <Ashish.Kalra@freescale.com>, Varun Sethi <Varun.Sethi@freescale.com>, and Liu Yu <yu.liu@freescale.com>. Signed-off-by: Scott Wood <scottwood@freescale.com> [agraf: remove pt_regs usage] Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Avi Kivity <avi@redhat.com>
2011-12-20 19:34:43 +04:00
mtspr SPRN_SPRG5W, r6
KVM: PPC: booke(hv): Fix save/restore of guest accessible SPRGs. For Guest accessible SPRGs 4-7, save/restore must be handled differently for 64bit and non-64 bit case. Use the PPC_STD/PPC_LD macros for saving/restoring to/from these registers. Signed-off-by: Varun Sethi <Varun.Sethi@freescale.com> Signed-off-by: Alexander Graf <agraf@suse.de>
2012-04-25 05:27:34 +04:00
PPC_LD(r8, VCPU_SHARED_SPRG7, r11)
KVM: PPC: booke: category E.HV (GS-mode) support Chips such as e500mc that implement category E.HV in Power ISA 2.06 provide hardware virtualization features, including a new MSR mode for guest state. The guest OS can perform many operations without trapping into the hypervisor, including transitions to and from guest userspace. Since we can use SRR1[GS] to reliably tell whether an exception came from guest state, instead of messing around with IVPR, we use DO_KVM similarly to book3s. Current issues include: - Machine checks from guest state are not routed to the host handler. - The guest can cause a host oops by executing an emulated instruction in a page that lacks read permission. Existing e500/4xx support has the same problem. Includes work by Ashish Kalra <Ashish.Kalra@freescale.com>, Varun Sethi <Varun.Sethi@freescale.com>, and Liu Yu <yu.liu@freescale.com>. Signed-off-by: Scott Wood <scottwood@freescale.com> [agraf: remove pt_regs usage] Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Avi Kivity <avi@redhat.com>
2011-12-20 19:34:43 +04:00
mtspr SPRN_SPRG6W, r7
mtspr SPRN_SPRG7W, r8
/* Load some guest volatiles. */
PPC_LL r3, VCPU_LR(r4)
PPC_LL r5, VCPU_XER(r4)
PPC_LL r6, VCPU_CTR(r4)
KVM: PPC: bookehv: Use lwz/stw instead of PPC_LL/PPC_STL for 32-bit fields Interrupt code used PPC_LL/PPC_STL macros to load/store some of u32 fields which led to memory overflow on 64-bit. Use lwz/stw instead. Signed-off-by: Mihai Caraman <mihai.caraman@freescale.com> Signed-off-by: Alexander Graf <agraf@suse.de>
2012-04-16 08:08:54 +04:00
lwz r7, VCPU_CR(r4)
KVM: PPC: booke: category E.HV (GS-mode) support Chips such as e500mc that implement category E.HV in Power ISA 2.06 provide hardware virtualization features, including a new MSR mode for guest state. The guest OS can perform many operations without trapping into the hypervisor, including transitions to and from guest userspace. Since we can use SRR1[GS] to reliably tell whether an exception came from guest state, instead of messing around with IVPR, we use DO_KVM similarly to book3s. Current issues include: - Machine checks from guest state are not routed to the host handler. - The guest can cause a host oops by executing an emulated instruction in a page that lacks read permission. Existing e500/4xx support has the same problem. Includes work by Ashish Kalra <Ashish.Kalra@freescale.com>, Varun Sethi <Varun.Sethi@freescale.com>, and Liu Yu <yu.liu@freescale.com>. Signed-off-by: Scott Wood <scottwood@freescale.com> [agraf: remove pt_regs usage] Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Avi Kivity <avi@redhat.com>
2011-12-20 19:34:43 +04:00
PPC_LL r8, VCPU_PC(r4)
KVM: PPC: bookehv: Use a Macro for saving/restoring guest registers to/from their 64 bit copies. Introduced PPC_STD/PPC_LD macros for saving/restoring guest registers to/from their 64 bit copies. Signed-off-by: Varun Sethi <Varun.Sethi@freescale.com> Signed-off-by: Alexander Graf <agraf@suse.de>
2012-04-25 05:26:43 +04:00
PPC_LD(r9, VCPU_SHARED_MSR, r11)
KVM: PPC: booke: category E.HV (GS-mode) support Chips such as e500mc that implement category E.HV in Power ISA 2.06 provide hardware virtualization features, including a new MSR mode for guest state. The guest OS can perform many operations without trapping into the hypervisor, including transitions to and from guest userspace. Since we can use SRR1[GS] to reliably tell whether an exception came from guest state, instead of messing around with IVPR, we use DO_KVM similarly to book3s. Current issues include: - Machine checks from guest state are not routed to the host handler. - The guest can cause a host oops by executing an emulated instruction in a page that lacks read permission. Existing e500/4xx support has the same problem. Includes work by Ashish Kalra <Ashish.Kalra@freescale.com>, Varun Sethi <Varun.Sethi@freescale.com>, and Liu Yu <yu.liu@freescale.com>. Signed-off-by: Scott Wood <scottwood@freescale.com> [agraf: remove pt_regs usage] Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Avi Kivity <avi@redhat.com>
2011-12-20 19:34:43 +04:00
PPC_LL r0, VCPU_GPR(r0)(r4)
PPC_LL r1, VCPU_GPR(r1)(r4)
PPC_LL r2, VCPU_GPR(r2)(r4)
PPC_LL r10, VCPU_GPR(r10)(r4)
PPC_LL r11, VCPU_GPR(r11)(r4)
PPC_LL r12, VCPU_GPR(r12)(r4)
PPC_LL r13, VCPU_GPR(r13)(r4)
mtlr r3
mtxer r5
mtctr r6
mtsrr0 r8
mtsrr1 r9
#ifdef CONFIG_KVM_EXIT_TIMING
/* save enter time */
1:
mfspr r6, SPRN_TBRU
Restore guest CR after exit timing calculation No instruction which can change Condition Register (CR) should be executed after Guest CR is loaded. So the guest CR is restored after the Exit Timing in lightweight_exit executes cmpw, which can clobber CR. Signed-off-by: Bharat Bhushan <bharat.bhushan@freescale.com> Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Avi Kivity <avi@redhat.com>
2012-03-05 05:34:08 +04:00
mfspr r9, SPRN_TBRL
KVM: PPC: booke: category E.HV (GS-mode) support Chips such as e500mc that implement category E.HV in Power ISA 2.06 provide hardware virtualization features, including a new MSR mode for guest state. The guest OS can perform many operations without trapping into the hypervisor, including transitions to and from guest userspace. Since we can use SRR1[GS] to reliably tell whether an exception came from guest state, instead of messing around with IVPR, we use DO_KVM similarly to book3s. Current issues include: - Machine checks from guest state are not routed to the host handler. - The guest can cause a host oops by executing an emulated instruction in a page that lacks read permission. Existing e500/4xx support has the same problem. Includes work by Ashish Kalra <Ashish.Kalra@freescale.com>, Varun Sethi <Varun.Sethi@freescale.com>, and Liu Yu <yu.liu@freescale.com>. Signed-off-by: Scott Wood <scottwood@freescale.com> [agraf: remove pt_regs usage] Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Avi Kivity <avi@redhat.com>
2011-12-20 19:34:43 +04:00
mfspr r8, SPRN_TBRU
cmpw r8, r6
KVM: PPC: bookehv: Use lwz/stw instead of PPC_LL/PPC_STL for 32-bit fields Interrupt code used PPC_LL/PPC_STL macros to load/store some of u32 fields which led to memory overflow on 64-bit. Use lwz/stw instead. Signed-off-by: Mihai Caraman <mihai.caraman@freescale.com> Signed-off-by: Alexander Graf <agraf@suse.de>
2012-04-16 08:08:54 +04:00
stw r9, VCPU_TIMING_LAST_ENTER_TBL(r4)
KVM: PPC: booke: category E.HV (GS-mode) support Chips such as e500mc that implement category E.HV in Power ISA 2.06 provide hardware virtualization features, including a new MSR mode for guest state. The guest OS can perform many operations without trapping into the hypervisor, including transitions to and from guest userspace. Since we can use SRR1[GS] to reliably tell whether an exception came from guest state, instead of messing around with IVPR, we use DO_KVM similarly to book3s. Current issues include: - Machine checks from guest state are not routed to the host handler. - The guest can cause a host oops by executing an emulated instruction in a page that lacks read permission. Existing e500/4xx support has the same problem. Includes work by Ashish Kalra <Ashish.Kalra@freescale.com>, Varun Sethi <Varun.Sethi@freescale.com>, and Liu Yu <yu.liu@freescale.com>. Signed-off-by: Scott Wood <scottwood@freescale.com> [agraf: remove pt_regs usage] Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Avi Kivity <avi@redhat.com>
2011-12-20 19:34:43 +04:00
bne 1b
KVM: PPC: bookehv: Use lwz/stw instead of PPC_LL/PPC_STL for 32-bit fields Interrupt code used PPC_LL/PPC_STL macros to load/store some of u32 fields which led to memory overflow on 64-bit. Use lwz/stw instead. Signed-off-by: Mihai Caraman <mihai.caraman@freescale.com> Signed-off-by: Alexander Graf <agraf@suse.de>
2012-04-16 08:08:54 +04:00
stw r8, VCPU_TIMING_LAST_ENTER_TBU(r4)
KVM: PPC: booke: category E.HV (GS-mode) support Chips such as e500mc that implement category E.HV in Power ISA 2.06 provide hardware virtualization features, including a new MSR mode for guest state. The guest OS can perform many operations without trapping into the hypervisor, including transitions to and from guest userspace. Since we can use SRR1[GS] to reliably tell whether an exception came from guest state, instead of messing around with IVPR, we use DO_KVM similarly to book3s. Current issues include: - Machine checks from guest state are not routed to the host handler. - The guest can cause a host oops by executing an emulated instruction in a page that lacks read permission. Existing e500/4xx support has the same problem. Includes work by Ashish Kalra <Ashish.Kalra@freescale.com>, Varun Sethi <Varun.Sethi@freescale.com>, and Liu Yu <yu.liu@freescale.com>. Signed-off-by: Scott Wood <scottwood@freescale.com> [agraf: remove pt_regs usage] Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Avi Kivity <avi@redhat.com>
2011-12-20 19:34:43 +04:00
#endif
Restore guest CR after exit timing calculation No instruction which can change Condition Register (CR) should be executed after Guest CR is loaded. So the guest CR is restored after the Exit Timing in lightweight_exit executes cmpw, which can clobber CR. Signed-off-by: Bharat Bhushan <bharat.bhushan@freescale.com> Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Avi Kivity <avi@redhat.com>
2012-03-05 05:34:08 +04:00
/*
* Don't execute any instruction which can change CR after
* below instruction.
*/
mtcr r7
KVM: PPC: booke: category E.HV (GS-mode) support Chips such as e500mc that implement category E.HV in Power ISA 2.06 provide hardware virtualization features, including a new MSR mode for guest state. The guest OS can perform many operations without trapping into the hypervisor, including transitions to and from guest userspace. Since we can use SRR1[GS] to reliably tell whether an exception came from guest state, instead of messing around with IVPR, we use DO_KVM similarly to book3s. Current issues include: - Machine checks from guest state are not routed to the host handler. - The guest can cause a host oops by executing an emulated instruction in a page that lacks read permission. Existing e500/4xx support has the same problem. Includes work by Ashish Kalra <Ashish.Kalra@freescale.com>, Varun Sethi <Varun.Sethi@freescale.com>, and Liu Yu <yu.liu@freescale.com>. Signed-off-by: Scott Wood <scottwood@freescale.com> [agraf: remove pt_regs usage] Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Avi Kivity <avi@redhat.com>
2011-12-20 19:34:43 +04:00
/* Finish loading guest volatiles and jump to guest. */
PPC_LL r5, VCPU_GPR(r5)(r4)
PPC_LL r6, VCPU_GPR(r6)(r4)
PPC_LL r7, VCPU_GPR(r7)(r4)
PPC_LL r8, VCPU_GPR(r8)(r4)
PPC_LL r9, VCPU_GPR(r9)(r4)
PPC_LL r3, VCPU_GPR(r3)(r4)
PPC_LL r4, VCPU_GPR(r4)(r4)
rfi
Reference in New Issue Copy Permalink