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>

arch/powerpc/include/asm/dbell.h

@@ -19,6 +19,7 @@
 
 #define PPC_DBELL_MSG_BRDCAST	(0x04000000)
 #define PPC_DBELL_TYPE(x)	(((x) & 0xf) << (63-36))
+#define PPC_DBELL_LPID(x)	((x) << (63 - 49))
 enum ppc_dbell {
 	PPC_DBELL = 0,		/* doorbell */
 	PPC_DBELL_CRIT = 1,	/* critical doorbell */

arch/powerpc/include/asm/kvm_asm.h

@@ -48,6 +48,14 @@
 #define BOOKE_INTERRUPT_SPE_FP_DATA 33
 #define BOOKE_INTERRUPT_SPE_FP_ROUND 34
 #define BOOKE_INTERRUPT_PERFORMANCE_MONITOR 35
+#define BOOKE_INTERRUPT_DOORBELL 36
+#define BOOKE_INTERRUPT_DOORBELL_CRITICAL 37
+
+/* booke_hv */
+#define BOOKE_INTERRUPT_GUEST_DBELL 38
+#define BOOKE_INTERRUPT_GUEST_DBELL_CRIT 39
+#define BOOKE_INTERRUPT_HV_SYSCALL 40
+#define BOOKE_INTERRUPT_HV_PRIV 41
 
 /* book3s */
 

arch/powerpc/include/asm/kvm_booke_hv_asm.h

@@ -0,0 +1,49 @@
+/*
+ * Copyright 2010-2011 Freescale Semiconductor, Inc.
+ *
+ * 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.
+ */
+
+#ifndef ASM_KVM_BOOKE_HV_ASM_H
+#define ASM_KVM_BOOKE_HV_ASM_H
+
+#ifdef __ASSEMBLY__
+
+/*
+ * All exceptions from guest state must go through KVM
+ * (except for those which are delivered directly to the guest) --
+ * there are no exceptions for which we fall through directly to
+ * the normal host handler.
+ *
+ * Expected inputs (normal exceptions):
+ *   SCRATCH0 = saved r10
+ *   r10 = thread struct
+ *   r11 = appropriate SRR1 variant (currently used as scratch)
+ *   r13 = saved CR
+ *   *(r10 + THREAD_NORMSAVE(0)) = saved r11
+ *   *(r10 + THREAD_NORMSAVE(2)) = saved r13
+ *
+ * Expected inputs (crit/mcheck/debug exceptions):
+ *   appropriate SCRATCH = saved r8
+ *   r8 = exception level stack frame
+ *   r9 = *(r8 + _CCR) = saved CR
+ *   r11 = appropriate SRR1 variant (currently used as scratch)
+ *   *(r8 + GPR9) = saved r9
+ *   *(r8 + GPR10) = saved r10 (r10 not yet clobbered)
+ *   *(r8 + GPR11) = saved r11
+ */
+.macro DO_KVM intno srr1
+#ifdef CONFIG_KVM_BOOKE_HV
+BEGIN_FTR_SECTION
+	mtocrf	0x80, r11	/* check MSR[GS] without clobbering reg */
+	bf	3, kvmppc_resume_\intno\()_\srr1
+	b	kvmppc_handler_\intno\()_\srr1
+kvmppc_resume_\intno\()_\srr1:
+END_FTR_SECTION_IFSET(CPU_FTR_EMB_HV)
+#endif
+.endm
+
+#endif /*__ASSEMBLY__ */
+#endif /* ASM_KVM_BOOKE_HV_ASM_H */

arch/powerpc/include/asm/kvm_host.h

@@ -106,6 +106,8 @@ struct kvm_vcpu_stat {
 	u32 dec_exits;
 	u32 ext_intr_exits;
 	u32 halt_wakeup;
+	u32 dbell_exits;
+	u32 gdbell_exits;
 #ifdef CONFIG_PPC_BOOK3S
 	u32 pf_storage;
 	u32 pf_instruc;
@@ -140,13 +142,16 @@ enum kvm_exit_types {
 	EMULATED_TLBSX_EXITS,
 	EMULATED_TLBWE_EXITS,
 	EMULATED_RFI_EXITS,
+	EMULATED_RFCI_EXITS,
 	DEC_EXITS,
 	EXT_INTR_EXITS,
 	HALT_WAKEUP,
 	USR_PR_INST,
 	FP_UNAVAIL,
 	DEBUG_EXITS,
 	TIMEINGUEST,
+	DBELL_EXITS,
+	GDBELL_EXITS,
 	__NUMBER_OF_KVM_EXIT_TYPES
 };
 
@@ -217,10 +222,10 @@ struct kvm_arch_memory_slot {
 };
 
 struct kvm_arch {
+	unsigned int lpid;
 #ifdef CONFIG_KVM_BOOK3S_64_HV
 	unsigned long hpt_virt;
 	struct revmap_entry *revmap;
-	unsigned int lpid;
 	unsigned int host_lpid;
 	unsigned long host_lpcr;
 	unsigned long sdr1;
@@ -345,6 +350,17 @@ struct kvm_vcpu_arch {
 	u64 vsr[64];
 #endif
 
+#ifdef CONFIG_KVM_BOOKE_HV
+	u32 host_mas4;
+	u32 host_mas6;
+	u32 shadow_epcr;
+	u32 epcr;
+	u32 shadow_msrp;
+	u32 eplc;
+	u32 epsc;
+	u32 oldpir;
+#endif
+
 #ifdef CONFIG_PPC_BOOK3S
 	/* For Gekko paired singles */
 	u32 qpr[32];
@@ -428,6 +444,7 @@ struct kvm_vcpu_arch {
 	ulong queued_esr;
 	u32 tlbcfg[4];
 	u32 mmucfg;
+	u32 epr;
 #endif
 	gpa_t paddr_accessed;
 

arch/powerpc/include/asm/kvm_ppc.h

@@ -139,6 +139,9 @@ extern int kvmppc_core_prepare_memory_region(struct kvm *kvm,
 extern void kvmppc_core_commit_memory_region(struct kvm *kvm,
 				struct kvm_userspace_memory_region *mem);
 
+extern int kvmppc_bookehv_init(void);
+extern void kvmppc_bookehv_exit(void);
+
 /*
  * Cuts out inst bits with ordering according to spec.
  * That means the leftmost bit is zero. All given bits are included.

arch/powerpc/include/asm/mmu-book3e.h

@@ -104,6 +104,8 @@
 #define MAS4_TSIZED_MASK	0x00000f80	/* Default TSIZE */
 #define MAS4_TSIZED_SHIFT	7
 
+#define MAS5_SGS		0x80000000
+
 #define MAS6_SPID0		0x3FFF0000
 #define MAS6_SPID1		0x00007FFE
 #define MAS6_ISIZE(x)		MAS1_TSIZE(x)
@@ -118,6 +120,10 @@
 
 #define MAS7_RPN		0xFFFFFFFF
 
+#define MAS8_TGS		0x80000000 /* Guest space */
+#define MAS8_VF			0x40000000 /* Virtualization Fault */
+#define MAS8_TLPID		0x000000ff
+
 /* Bit definitions for MMUCFG */
 #define MMUCFG_MAVN	0x00000003	/* MMU Architecture Version Number */
 #define MMUCFG_MAVN_V1	0x00000000	/* v1.0 */

arch/powerpc/include/asm/processor.h

@@ -243,6 +243,9 @@ struct thread_struct {
 #ifdef CONFIG_KVM_BOOK3S_32_HANDLER
 	void*		kvm_shadow_vcpu; /* KVM internal data */
 #endif /* CONFIG_KVM_BOOK3S_32_HANDLER */
+#if defined(CONFIG_KVM) && defined(CONFIG_BOOKE)
+	struct kvm_vcpu	*kvm_vcpu;
+#endif
 #ifdef CONFIG_PPC64
 	unsigned long	dscr;
 	int		dscr_inherit;

arch/powerpc/include/asm/reg.h

@@ -257,7 +257,9 @@
 #define   LPCR_LPES_SH	2
 #define   LPCR_RMI     0x00000002      /* real mode is cache inhibit */
 #define   LPCR_HDICE   0x00000001      /* Hyp Decr enable (HV,PR,EE) */
+#ifndef SPRN_LPID
 #define SPRN_LPID	0x13F	/* Logical Partition Identifier */
+#endif
 #define   LPID_RSVD	0x3ff		/* Reserved LPID for partn switching */
 #define	SPRN_HMER	0x150	/* Hardware m? error recovery */
 #define	SPRN_HMEER	0x151	/* Hardware m? enable error recovery */

arch/powerpc/include/asm/reg_booke.h

@@ -61,18 +61,30 @@ extern u32 booke_wdt_period;
 #define SPRN_SPRG7W	0x117	/* Special Purpose Register General 7 Write */
 #define SPRN_EPCR	0x133	/* Embedded Processor Control Register */
 #define SPRN_DBCR2	0x136	/* Debug Control Register 2 */
+#define SPRN_MSRP	0x137	/* MSR Protect Register */
 #define SPRN_IAC3	0x13A	/* Instruction Address Compare 3 */
 #define SPRN_IAC4	0x13B	/* Instruction Address Compare 4 */
 #define SPRN_DVC1	0x13E	/* Data Value Compare Register 1 */
 #define SPRN_DVC2	0x13F	/* Data Value Compare Register 2 */
+#define SPRN_LPID	0x152	/* Logical Partition ID */
 #define SPRN_MAS8	0x155	/* MMU Assist Register 8 */
 #define SPRN_TLB0PS	0x158	/* TLB 0 Page Size Register */
 #define SPRN_TLB1PS	0x159	/* TLB 1 Page Size Register */
 #define SPRN_MAS5_MAS6	0x15c	/* MMU Assist Register 5 || 6 */
 #define SPRN_MAS8_MAS1	0x15d	/* MMU Assist Register 8 || 1 */
 #define SPRN_EPTCFG	0x15e	/* Embedded Page Table Config */
+#define SPRN_GSPRG0	0x170	/* Guest SPRG0 */
+#define SPRN_GSPRG1	0x171	/* Guest SPRG1 */
+#define SPRN_GSPRG2	0x172	/* Guest SPRG2 */
+#define SPRN_GSPRG3	0x173	/* Guest SPRG3 */
 #define SPRN_MAS7_MAS3	0x174	/* MMU Assist Register 7 || 3 */
 #define SPRN_MAS0_MAS1	0x175	/* MMU Assist Register 0 || 1 */
+#define SPRN_GSRR0	0x17A	/* Guest SRR0 */
+#define SPRN_GSRR1	0x17B	/* Guest SRR1 */
+#define SPRN_GEPR	0x17C	/* Guest EPR */
+#define SPRN_GDEAR	0x17D	/* Guest DEAR */
+#define SPRN_GPIR	0x17E	/* Guest PIR */
+#define SPRN_GESR	0x17F	/* Guest Exception Syndrome Register */
 #define SPRN_IVOR0	0x190	/* Interrupt Vector Offset Register 0 */
 #define SPRN_IVOR1	0x191	/* Interrupt Vector Offset Register 1 */
 #define SPRN_IVOR2	0x192	/* Interrupt Vector Offset Register 2 */
@@ -93,6 +105,13 @@ extern u32 booke_wdt_period;
 #define SPRN_IVOR39	0x1B1	/* Interrupt Vector Offset Register 39 */
 #define SPRN_IVOR40	0x1B2	/* Interrupt Vector Offset Register 40 */
 #define SPRN_IVOR41	0x1B3	/* Interrupt Vector Offset Register 41 */
+#define SPRN_GIVOR2	0x1B8	/* Guest IVOR2 */
+#define SPRN_GIVOR3	0x1B9	/* Guest IVOR3 */
+#define SPRN_GIVOR4	0x1BA	/* Guest IVOR4 */
+#define SPRN_GIVOR8	0x1BB	/* Guest IVOR8 */
+#define SPRN_GIVOR13	0x1BC	/* Guest IVOR13 */
+#define SPRN_GIVOR14	0x1BD	/* Guest IVOR14 */
+#define SPRN_GIVPR	0x1BF	/* Guest IVPR */
 #define SPRN_SPEFSCR	0x200	/* SPE & Embedded FP Status & Control */
 #define SPRN_BBEAR	0x201	/* Branch Buffer Entry Address Register */
 #define SPRN_BBTAR	0x202	/* Branch Buffer Target Address Register */
@@ -245,6 +264,10 @@ extern u32 booke_wdt_period;
 #define MCSR_LDG	0x00002000UL /* Guarded Load */
 #define MCSR_TLBSYNC	0x00000002UL /* Multiple tlbsyncs detected */
 #define MCSR_BSL2_ERR	0x00000001UL /* Backside L2 cache error */
+
+#define MSRP_UCLEP	0x04000000 /* Protect MSR[UCLE] */
+#define MSRP_DEP	0x00000200 /* Protect MSR[DE] */
+#define MSRP_PMMP	0x00000004 /* Protect MSR[PMM] */
 #endif
 
 #ifdef CONFIG_E200
@@ -599,6 +622,17 @@ extern u32 booke_wdt_period;
 #define SPRN_EPCR_DMIUH		0x00400000	/* Disable MAS Interrupt updates
 						 * for hypervisor */
 
+/* Bit definitions for EPLC/EPSC */
+#define EPC_EPR		0x80000000 /* 1 = user, 0 = kernel */
+#define EPC_EPR_SHIFT	31
+#define EPC_EAS		0x40000000 /* Address Space */
+#define EPC_EAS_SHIFT	30
+#define EPC_EGS		0x20000000 /* 1 = guest, 0 = hypervisor */
+#define EPC_EGS_SHIFT	29
+#define EPC_ELPID	0x00ff0000
+#define EPC_ELPID_SHIFT	16
+#define EPC_EPID	0x00003fff
+#define EPC_EPID_SHIFT	0
 
 /*
  * The IBM-403 is an even more odd special case, as it is much

arch/powerpc/kernel/asm-offsets.c

@@ -116,6 +116,9 @@ int main(void)
 #ifdef CONFIG_KVM_BOOK3S_32_HANDLER
 	DEFINE(THREAD_KVM_SVCPU, offsetof(struct thread_struct, kvm_shadow_vcpu));
 #endif
+#ifdef CONFIG_KVM_BOOKE_HV
+	DEFINE(THREAD_KVM_VCPU, offsetof(struct thread_struct, kvm_vcpu));
+#endif
 
 	DEFINE(TI_FLAGS, offsetof(struct thread_info, flags));
 	DEFINE(TI_LOCAL_FLAGS, offsetof(struct thread_info, local_flags));
@@ -387,6 +390,7 @@ int main(void)
 #ifdef CONFIG_KVM
 	DEFINE(VCPU_HOST_STACK, offsetof(struct kvm_vcpu, arch.host_stack));
 	DEFINE(VCPU_HOST_PID, offsetof(struct kvm_vcpu, arch.host_pid));
+	DEFINE(VCPU_GUEST_PID, offsetof(struct kvm_vcpu, arch.pid));
 	DEFINE(VCPU_GPRS, offsetof(struct kvm_vcpu, arch.gpr));
 	DEFINE(VCPU_VRSAVE, offsetof(struct kvm_vcpu, arch.vrsave));
 	DEFINE(VCPU_FPRS, offsetof(struct kvm_vcpu, arch.fpr));
@@ -429,9 +433,11 @@ int main(void)
 	DEFINE(VCPU_SHARED_MAS4, offsetof(struct kvm_vcpu_arch_shared, mas4));
 	DEFINE(VCPU_SHARED_MAS6, offsetof(struct kvm_vcpu_arch_shared, mas6));
 
+	DEFINE(VCPU_KVM, offsetof(struct kvm_vcpu, kvm));
+	DEFINE(KVM_LPID, offsetof(struct kvm, arch.lpid));
+
 	/* book3s */
 #ifdef CONFIG_KVM_BOOK3S_64_HV
-	DEFINE(KVM_LPID, offsetof(struct kvm, arch.lpid));
 	DEFINE(KVM_SDR1, offsetof(struct kvm, arch.sdr1));
 	DEFINE(KVM_HOST_LPID, offsetof(struct kvm, arch.host_lpid));
 	DEFINE(KVM_HOST_LPCR, offsetof(struct kvm, arch.host_lpcr));
@@ -446,7 +452,6 @@ int main(void)
 	DEFINE(VCPU_DAR, offsetof(struct kvm_vcpu, arch.shregs.dar));
 #endif
 #ifdef CONFIG_PPC_BOOK3S
-	DEFINE(VCPU_KVM, offsetof(struct kvm_vcpu, kvm));
 	DEFINE(VCPU_VCPUID, offsetof(struct kvm_vcpu, vcpu_id));
 	DEFINE(VCPU_PURR, offsetof(struct kvm_vcpu, arch.purr));
 	DEFINE(VCPU_SPURR, offsetof(struct kvm_vcpu, arch.spurr));
@@ -597,6 +602,12 @@ int main(void)
 	DEFINE(VCPU_HOST_SPEFSCR, offsetof(struct kvm_vcpu, arch.host_spefscr));
 #endif
 
+#ifdef CONFIG_KVM_BOOKE_HV
+	DEFINE(VCPU_HOST_MAS4, offsetof(struct kvm_vcpu, arch.host_mas4));
+	DEFINE(VCPU_HOST_MAS6, offsetof(struct kvm_vcpu, arch.host_mas6));
+	DEFINE(VCPU_EPLC, offsetof(struct kvm_vcpu, arch.eplc));
+#endif
+
 #ifdef CONFIG_KVM_EXIT_TIMING
 	DEFINE(VCPU_TIMING_EXIT_TBU, offsetof(struct kvm_vcpu,
 						arch.timing_exit.tv32.tbu));

arch/powerpc/kernel/head_booke.h

@@ -3,6 +3,7 @@
 
 #include <asm/ptrace.h>	/* for STACK_FRAME_REGS_MARKER */
 #include <asm/kvm_asm.h>
+#include <asm/kvm_booke_hv_asm.h>
 
 /*
  * Macros used for common Book-e exception handling
@@ -36,8 +37,9 @@
 	stw	r11, THREAD_NORMSAVE(0)(r10);				     \
 	stw	r13, THREAD_NORMSAVE(2)(r10);				     \
 	mfcr	r13;			/* save CR in r13 for now	   */\
-	mfspr	r11,SPRN_SRR1;		/* check whether user or kernel    */\
-	andi.	r11,r11,MSR_PR;						     \
+	mfspr	r11, SPRN_SRR1;		                                     \
+	DO_KVM	BOOKE_INTERRUPT_##intno SPRN_SRR1;			     \
+	andi.	r11, r11, MSR_PR;	/* check whether user or kernel    */\
 	mr	r11, r1;						     \
 	beq	1f;							     \
 	/* if from user, start at top of this thread's kernel stack */       \
@@ -123,8 +125,9 @@
 	stw	r10,GPR10(r8);						     \
 	stw	r11,GPR11(r8);						     \
 	stw	r9,_CCR(r8);		/* save CR on stack		   */\
-	mfspr	r10,exc_level_srr1;	/* check whether user or kernel    */\
-	andi.	r10,r10,MSR_PR;						     \
+	mfspr	r11,exc_level_srr1;	/* check whether user or kernel    */\
+	DO_KVM	BOOKE_INTERRUPT_##intno exc_level_srr1;		             \
+	andi.	r11,r11,MSR_PR;						     \
 	mfspr	r11,SPRN_SPRG_THREAD;	/* if from user, start at top of   */\
 	lwz	r11,THREAD_INFO-THREAD(r11); /* this thread's kernel stack */\
 	addi	r11,r11,EXC_LVL_FRAME_OVERHEAD;	/* allocate stack frame    */\
@@ -172,6 +175,23 @@
 		EXC_LEVEL_EXCEPTION_PROLOG(MC, MACHINE_CHECK, \
 			SPRN_MCSRR0, SPRN_MCSRR1)
 
+/*
+ * Guest Doorbell -- this is a bit odd in that uses GSRR0/1 despite
+ * being delivered to the host.  This exception can only happen
+ * inside a KVM guest -- so we just handle up to the DO_KVM rather
+ * than try to fit this into one of the existing prolog macros.
+ */
+#define GUEST_DOORBELL_EXCEPTION \
+	START_EXCEPTION(GuestDoorbell);					     \
+	mtspr	SPRN_SPRG_WSCRATCH0, r10;	/* save one register */	     \
+	mfspr	r10, SPRN_SPRG_THREAD;					     \
+	stw	r11, THREAD_NORMSAVE(0)(r10);				     \
+	mfspr	r11, SPRN_SRR1;		                                     \
+	stw	r13, THREAD_NORMSAVE(2)(r10);				     \
+	mfcr	r13;			/* save CR in r13 for now	   */\
+	DO_KVM	BOOKE_INTERRUPT_GUEST_DBELL SPRN_GSRR1;			     \
+	trap
+
 /*
  * Exception vectors.
  */

arch/powerpc/kvm/Kconfig

@@ -90,6 +90,9 @@ config KVM_BOOK3S_64_PR
 	depends on KVM_BOOK3S_64 && !KVM_BOOK3S_64_HV
 	select KVM_BOOK3S_PR
 
+config KVM_BOOKE_HV
+	bool
+
 config KVM_440
 	bool "KVM support for PowerPC 440 processors"
 	depends on EXPERIMENTAL && 44x