Skip to contentSkip to conversation

Support POWER8 in HV KVM

RFC PATCH In review Paul Mackerras 2 participants 16 messages Sep 6, 2013 – Nov 6, 2013

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

cover Support POWER8 in HV KVM

This series adds support for the POWER8 CPU in HV KVM. POWER8 adds several new guest-accessible instructions, special-purpose registers, and other features such as doorbell interrupts and hardware transactional memory. It also adds new hypervisor-controlled features such as relocation-on interrupts, and replaces the DABR/DABRX registers with the new DAWR/DAWRX registers with expanded capabilities. The POWER8 CPU has compatibility modes for architecture 2.06 (POWER7/POWER7+) and 2.05 (POWER6).

This series does not yet handle the checkpointed register state of the guest.

 arch/powerpc/include/asm/kvm_asm.h        |   2 +
 arch/powerpc/include/asm/kvm_book3s_asm.h |   1 +
 arch/powerpc/include/asm/kvm_host.h       |  29 +-
 arch/powerpc/include/asm/reg.h            |  25 +-
 arch/powerpc/include/uapi/asm/kvm.h       |   1 +
 arch/powerpc/kernel/asm-offsets.c         |  28 +-
 arch/powerpc/kvm/book3s_hv.c              | 234 +++++++++++--
 arch/powerpc/kvm/book3s_hv_interrupts.S   |   8 +-
 arch/powerpc/kvm/book3s_hv_rmhandlers.S   | 538 ++++++++++++++++++++++--------
 9 files changed, 694 insertions(+), 172 deletions(-)

Paul.

#1 lore raw

RFC PATCH 1/10 KVM: PPC: Book3S HV: Align physical CPU thread numbers with virtual

On a threaded processor such as POWER7, we group VCPUs into virtual cores and arrange that the VCPUs in a virtual core run on the same physical core. Currently we don't enforce any correspondence between virtual thread numbers within a virtual core and physical thread numbers. Physical threads are allocated starting at 0 on a first-come first-served basis to runnable virtual threads (VCPUs).

POWER8 implements a new "msgsndp" instruction which guest kernels can use to interrupt other threads in the same core or sub-core. Since the instruction takes the destination physical thread ID as a parameter, it becomes necessary to align the physical thread IDs with the virtual thread IDs, that is, to make sure virtual thread N within a virtual core always runs on physical thread N.

This means that it's possible that thread 0, which is where we call __kvmppc_vcore_entry, may end up running some other vcpu than the one whose task called kvmppc_run_core(), or it may end up running no vcpu at all, if for example thread 0 of the virtual core is currently executing in userspace. Thus we can't rely on thread 0 to be the master responsible for switching the MMU. Instead we now have an explicit 'is_master' flag which is set for the vcpu whose task called kvmppc_run_core(). The master then has to wait for thread 0 to enter real mode before switching the MMU. Also, we no longer pass the vcpu pointer to __kvmppc_vcore_entry, but instead let the assembly code load it from the PACA.

Since the assembly code will need to know the kvm pointer and the thread ID for threads which don't have a vcpu, we move the thread ID into the PACA and we add a kvm pointer to the virtual core structure.

In the case where thread 0 has no vcpu to run, we arrange for it to go to nap mode, using a new flag value in the PACA 'napping' field so we can differentiate it when it wakes from the other nap cases. We set the bit for the thread in the vcore 'napping_threads' field so that when other threads come out of the guest they will send an IPI to thread 0 to wake it up. When it does wake up, we clear that bit, see what caused the wakeup, and either exit back to the kernel, or start running virtual thread 0 in the case where it now wants to enter the guest and the other threads are still in the guest.

---
 arch/powerpc/include/asm/kvm_book3s_asm.h |   1 +
 arch/powerpc/include/asm/kvm_host.h       |   4 +
 arch/powerpc/kernel/asm-offsets.c         |   5 +-
 arch/powerpc/kvm/book3s_hv.c              |  49 ++++----
 arch/powerpc/kvm/book3s_hv_interrupts.S   |   6 +-
 arch/powerpc/kvm/book3s_hv_rmhandlers.S   | 188 +++++++++++++++++++++++++-----
 6 files changed, 192 insertions(+), 61 deletions(-)
arch/powerpc/include/asm/kvm_book3s_asm.h +1 -0
index 22f4606..8669ac0 100644
@@ -87,6 +87,7 @@ struct kvmppc_host_state {
 	u8 hwthread_req;
 	u8 hwthread_state;
 	u8 host_ipi;
+	u8 ptid;
 	struct kvm_vcpu *kvm_vcpu;
 	struct kvmppc_vcore *kvm_vcore;
 	unsigned long xics_phys;
arch/powerpc/include/asm/kvm_host.h +4 -0
index 5a40270..6a43455 100644
@@ -284,6 +284,8 @@ struct kvmppc_vcore {
 	int n_woken;
 	int nap_count;
 	int napping_threads;
+	int real_mode_threads;
+	int first_vcpuid;
 	u16 pcpu;
 	u16 last_cpu;
 	u8 vcore_state;
@@ -294,6 +296,7 @@ struct kvmppc_vcore {
 	u64 stolen_tb;
 	u64 preempt_tb;
 	struct kvm_vcpu *runner;
+	struct kvm *kvm;
 	u64 tb_offset;		/* guest timebase - host timebase */
 	u32 arch_compat;
 	ulong pcr;
@@ -575,6 +578,7 @@ struct kvm_vcpu_arch {
 	int state;
 	int ptid;
 	bool timer_running;
+	u8 is_master;
 	wait_queue_head_t cpu_run;
 
 	struct kvm_vcpu_arch_shared *shared;
arch/powerpc/kernel/asm-offsets.c +4 -1
index 115dd64..51d7eed 100644
@@ -514,13 +514,15 @@ int main(void)
 	DEFINE(VCPU_FAULT_DAR, offsetof(struct kvm_vcpu, arch.fault_dar));
 	DEFINE(VCPU_LAST_INST, offsetof(struct kvm_vcpu, arch.last_inst));
 	DEFINE(VCPU_TRAP, offsetof(struct kvm_vcpu, arch.trap));
-	DEFINE(VCPU_PTID, offsetof(struct kvm_vcpu, arch.ptid));
+	DEFINE(VCPU_IS_MASTER, offsetof(struct kvm_vcpu, arch.is_master));
 	DEFINE(VCPU_CFAR, offsetof(struct kvm_vcpu, arch.cfar));
 	DEFINE(VCPU_PPR, offsetof(struct kvm_vcpu, arch.ppr));
 	DEFINE(VCORE_ENTRY_EXIT, offsetof(struct kvmppc_vcore, entry_exit_count));
 	DEFINE(VCORE_NAP_COUNT, offsetof(struct kvmppc_vcore, nap_count));
 	DEFINE(VCORE_IN_GUEST, offsetof(struct kvmppc_vcore, in_guest));
 	DEFINE(VCORE_NAPPING_THREADS, offsetof(struct kvmppc_vcore, napping_threads));
+	DEFINE(VCORE_RM_THREADS, offsetof(struct kvmppc_vcore, real_mode_threads));
+	DEFINE(VCORE_KVM, offsetof(struct kvmppc_vcore, kvm));
 	DEFINE(VCORE_TB_OFFSET, offsetof(struct kvmppc_vcore, tb_offset));
 	DEFINE(VCORE_PCR, offsetof(struct kvmppc_vcore, pcr));
 	DEFINE(VCPU_SVCPU, offsetof(struct kvmppc_vcpu_book3s, shadow_vcpu) -
@@ -590,6 +592,7 @@ int main(void)
 	HSTATE_FIELD(HSTATE_XICS_PHYS, xics_phys);
 	HSTATE_FIELD(HSTATE_SAVED_XIRR, saved_xirr);
 	HSTATE_FIELD(HSTATE_HOST_IPI, host_ipi);
+	HSTATE_FIELD(HSTATE_PTID, ptid);
 	HSTATE_FIELD(HSTATE_MMCR, host_mmcr);
 	HSTATE_FIELD(HSTATE_PMC, host_pmc);
 	HSTATE_FIELD(HSTATE_PURR, host_purr);
arch/powerpc/kvm/book3s_hv.c +20 -29

134 lines not on this page · raw message

arch/powerpc/kvm/book3s_hv_interrupts.S +2 -4
index 37f1cc4..2a85bf6 100644
@@ -35,7 +35,7 @@
  ****************************************************************************/
 
 /* Registers:
- *  r4: vcpu pointer
+ *  none
  */
 _GLOBAL(__kvmppc_vcore_entry)
 
@@ -69,7 +69,6 @@ END_FTR_SECTION_IFSET(CPU_FTR_ARCH_206)
 	mtmsrd  r10,1
 
 	/* Save host PMU registers */
-	/* R4 is live here (vcpu pointer) but not r3 or r5 */
 	li	r3, 1
 	sldi	r3, r3, 31		/* MMCR0_FC (freeze counters) bit */
 	mfspr	r7, SPRN_MMCR0		/* save MMCR0 */
@@ -134,16 +133,15 @@ END_FTR_SECTION_IFSET(CPU_FTR_ARCH_201)
 	 * enters the guest with interrupts enabled.
 	 */
 BEGIN_FTR_SECTION
+	ld	r4, HSTATE_KVM_VCPU(r13)
 	ld	r0, VCPU_PENDING_EXC(r4)
 	li	r7, (1 << BOOK3S_IRQPRIO_EXTERNAL)
 	oris	r7, r7, (1 << BOOK3S_IRQPRIO_EXTERNAL_LEVEL)@h
 	and.	r0, r0, r7
 	beq	32f
-	mr	r31, r4
 	lhz	r3, PACAPACAINDEX(r13)
 	bl	smp_send_reschedule
 	nop
-	mr	r4, r31
 32:
 END_FTR_SECTION_IFSET(CPU_FTR_ARCH_201)
 #endif /* CONFIG_SMP */
arch/powerpc/kvm/book3s_hv_rmhandlers.S +161 -27

294 lines not on this page · raw message

Signature · 1 line
1.8.4.rc3
#2 lore raw

RFC PATCH 2/10 KVM: PPC: Book3S HV: Don't set DABR on POWER8

In reply to Paul Mackerras

From: Michael Neuling <[email protected]>

POWER8 doesn't have the DABR and DABRX registers; instead it has new DAWR/DAWRX registers, which will be handled in a later patch.

---
 arch/powerpc/kvm/book3s_hv_interrupts.S |  2 ++
 arch/powerpc/kvm/book3s_hv_rmhandlers.S | 13 ++++++++++---
 2 files changed, 12 insertions(+), 3 deletions(-)
arch/powerpc/kvm/book3s_hv_interrupts.S +2 -0
index 2a85bf6..0be3404 100644
@@ -57,9 +57,11 @@ BEGIN_FTR_SECTION
 	std	r3, HSTATE_DSCR(r13)
 END_FTR_SECTION_IFSET(CPU_FTR_ARCH_206)
 
+BEGIN_FTR_SECTION
 	/* Save host DABR */
 	mfspr	r3, SPRN_DABR
 	std	r3, HSTATE_DABR(r13)
+END_FTR_SECTION_IFCLR(CPU_FTR_ARCH_207S)
 
 	/* Hard-disable interrupts */
 	mfmsr   r10
arch/powerpc/kvm/book3s_hv_rmhandlers.S +10 -3
index 0721d2a..d4f6f5f 100644
@@ -95,11 +95,13 @@ kvmppc_got_guest:
 
 	/* Back from guest - restore host state and return to caller */
 
+BEGIN_FTR_SECTION
 	/* Restore host DABR and DABRX */
 	ld	r5,HSTATE_DABR(r13)
 	li	r6,7
 	mtspr	SPRN_DABR,r5
 	mtspr	SPRN_DABRX,r6
+END_FTR_SECTION_IFCLR(CPU_FTR_ARCH_207S)
 
 	/* Restore SPRG3 */
 	ld	r3,PACA_SPRG3(r13)
@@ -423,15 +425,17 @@ kvmppc_hv_entry:
 	std	r0, PPC_LR_STKOFF(r1)
 	stdu	r1, -112(r1)
 
+BEGIN_FTR_SECTION
 	/* Set partition DABR */
 	/* Do this before re-enabling PMU to avoid P7 DABR corruption bug */
 	li	r5,3
 	ld	r6,VCPU_DABR(r4)
 	mtspr	SPRN_DABRX,r5
 	mtspr	SPRN_DABR,r6
-BEGIN_FTR_SECTION
+ BEGIN_FTR_SECTION_NESTED(89)
 	isync
-END_FTR_SECTION_IFSET(CPU_FTR_ARCH_206)
+ END_FTR_SECTION_NESTED(CPU_FTR_ARCH_206, CPU_FTR_ARCH_206, 89)
+END_FTR_SECTION_IFCLR(CPU_FTR_ARCH_207S)
 
 	/* Load guest PMU registers */
 	/* R4 is live here (vcpu pointer) */
@@ -1716,6 +1720,9 @@ ignore_hdec:
 	b	fast_guest_return
 
 _GLOBAL(kvmppc_h_set_dabr)
+BEGIN_FTR_SECTION
+	b	2f
+END_FTR_SECTION_IFSET(CPU_FTR_ARCH_207S)
 	std	r4,VCPU_DABR(r3)
 	/* Work around P7 bug where DABR can get corrupted on mtspr */
 1:	mtspr	SPRN_DABR,r4
@@ -1723,7 +1730,7 @@ _GLOBAL(kvmppc_h_set_dabr)
 	cmpd	r4, r5
 	bne	1b
 	isync
-	li	r3,0
+2:	li	r3,0
 	blr
 
 _GLOBAL(kvmppc_h_cede)
Signature · 1 line
1.8.4.rc3
#3 lore raw

RFC PATCH 3/10 KVM: PPC: Book3S HV: Context-switch new POWER8 SPRs

In reply to Paul Mackerras

From: Michael Neuling <[email protected]>

This adds fields to the struct kvm_vcpu_arch to store the new guest-accessible SPRs on POWER8, adds code to the get/set_one_reg functions to allow userspace to access this state, and adds code to the guest entry and exit to context-switch these SPRs between host and guest.

Note that DPDES (Directed Privileged Doorbell Exception State) is shared between threads on a core; hence we store it in struct kvmppc_vcore and have the master thread save and restore it.

---
 arch/powerpc/include/asm/kvm_host.h     |  25 +++++-
 arch/powerpc/include/asm/reg.h          |  17 ++++
 arch/powerpc/include/uapi/asm/kvm.h     |   1 +
 arch/powerpc/kernel/asm-offsets.c       |  23 +++++
 arch/powerpc/kvm/book3s_hv.c            | 153 +++++++++++++++++++++++++++++++-
 arch/powerpc/kvm/book3s_hv_rmhandlers.S | 144 ++++++++++++++++++++++++++++++
 6 files changed, 360 insertions(+), 3 deletions(-)
arch/powerpc/include/asm/kvm_host.h +24 -1
index 6a43455..4786bb0 100644
@@ -300,6 +300,7 @@ struct kvmppc_vcore {
 	u64 tb_offset;		/* guest timebase - host timebase */
 	u32 arch_compat;
 	ulong pcr;
+	ulong dpdes;		/* doorbell state (POWER8) */
 };
 
 #define VCORE_ENTRY_COUNT(vc)	((vc)->entry_exit_count & 0xff)
@@ -454,6 +455,7 @@ struct kvm_vcpu_arch {
 	ulong pc;
 	ulong ctr;
 	ulong lr;
+	ulong tar;
 
 	ulong xer;
 	u32 cr;
@@ -463,13 +465,32 @@ struct kvm_vcpu_arch {
 	ulong guest_owned_ext;
 	ulong purr;
 	ulong spurr;
+	ulong ic;
+	ulong vtb;
 	ulong dscr;
 	ulong amr;
 	ulong uamor;
+	ulong iamr;
 	u32 ctrl;
 	ulong dabr;
+	ulong dawr;
+	ulong dawrx;
+	ulong ciabr;
 	ulong cfar;
 	ulong ppr;
+	ulong pspb;
+	ulong fscr;
+	ulong tfhar;
+	ulong tfiar;
+	ulong texasr;
+	ulong ebbhr;
+	ulong ebbrr;
+	ulong bescr;
+	ulong csigr;
+	ulong tacr;
+	ulong tcscr;
+	ulong acop;
+	ulong wort;
 #endif
 	u32 vrsave; /* also USPRG0 */
 	u32 mmucr;
@@ -503,10 +524,12 @@ struct kvm_vcpu_arch {
 	u32 ccr1;
 	u32 dbsr;
 
-	u64 mmcr[3];
+	u64 mmcr[5];
 	u32 pmc[8];
+	u32 spmc[2];
 	u64 siar;
 	u64 sdar;
+	u64 sier;
 
 #ifdef CONFIG_KVM_EXIT_TIMING
 	struct mutex exit_timing_lock;
arch/powerpc/include/asm/reg.h +17 -0
index 52ff962..4ca8b85 100644
@@ -218,6 +218,11 @@
 #define   CTRL_TE	0x00c00000	/* thread enable */
 #define   CTRL_RUNLATCH	0x1
 #define SPRN_DAWR	0xB4
+#define SPRN_CIABR	0xBB
+#define   CIABR_PRIV		0x3
+#define   CIABR_PRIV_USER	1
+#define   CIABR_PRIV_SUPER	2
+#define   CIABR_PRIV_HYPER	3
 #define SPRN_DAWRX	0xBC
 #define   DAWRX_USER	(1UL << 0)
 #define   DAWRX_KERNEL	(1UL << 1)
@@ -255,6 +260,8 @@
 #define SPRN_HRMOR	0x139	/* Real mode offset register */
 #define SPRN_HSRR0	0x13A	/* Hypervisor Save/Restore 0 */
 #define SPRN_HSRR1	0x13B	/* Hypervisor Save/Restore 1 */
+#define SPRN_IC		0x350	/* Virtual Instruction Count */
+#define SPRN_VTB	0x351	/* Virtual Time Base */
 #define SPRN_FSCR	0x099	/* Facility Status & Control Register */
 #define   FSCR_TAR	(1 << (63-55)) /* Enable Target Address Register */
 #define   FSCR_EBB	(1 << (63-56)) /* Enable Event Based Branching */
@@ -354,6 +361,8 @@
 #define DER_EBRKE	0x00000002	/* External Breakpoint Interrupt */
 #define DER_DPIE	0x00000001	/* Dev. Port Nonmaskable Request */
 #define SPRN_DMISS	0x3D0		/* Data TLB Miss Register */
+#define SPRN_DHDES	0x0B1		/* Directed Hyp. Doorbell Exc. State */
+#define SPRN_DPDES	0x0B0		/* Directed Priv. Doorbell Exc. State */
 #define SPRN_EAR	0x11A		/* External Address Register */
 #define SPRN_HASH1	0x3D2		/* Primary Hash Address Register */
 #define SPRN_HASH2	0x3D3		/* Secondary Hash Address Resgister */
@@ -413,6 +422,7 @@
 #define SPRN_IABR	0x3F2	/* Instruction Address Breakpoint Register */
 #define SPRN_IABR2	0x3FA		/* 83xx */
 #define SPRN_IBCR	0x135		/* 83xx Insn Breakpoint Control Reg */
+#define SPRN_IAMR	0x03D		/* Instr. Authority Mask Reg */
 #define SPRN_HID4	0x3F4		/* 970 HID4 */
 #define  HID4_LPES0	 (1ul << (63-0)) /* LPAR env. sel. bit 0 */
 #define	 HID4_RMLS2_SH	 (63 - 2)	/* Real mode limit bottom 2 bits */
@@ -526,6 +536,7 @@
 #define SPRN_PIR	0x3FF	/* Processor Identification Register */
 #endif
 #define SPRN_TIR	0x1BE	/* Thread Identification Register */
+#define SPRN_PSPB	0x09F	/* Problem State Priority Boost reg */
 #define SPRN_PTEHI	0x3D5	/* 981 7450 PTE HI word (S/W TLB load) */
 #define SPRN_PTELO	0x3D6	/* 982 7450 PTE LO word (S/W TLB load) */
 #define SPRN_PURR	0x135	/* Processor Utilization of Resources Reg */
@@ -667,6 +678,7 @@
 #define SPRN_EBBHR	804	/* Event based branch handler register */
 #define SPRN_EBBRR	805	/* Event based branch return register */
 #define SPRN_BESCR	806	/* Branch event status and control register */
+#define SPRN_WORT	895	/* Workload optimization register - thread */
 
 #define SPRN_PMC1	787
 #define SPRN_PMC2	788
@@ -683,6 +695,11 @@
 #define   SIER_SIHV		0x1000000	/* Sampled MSR_HV */
 #define   SIER_SIAR_VALID	0x0400000	/* SIAR contents valid */
 #define   SIER_SDAR_VALID	0x0200000	/* SDAR contents valid */
+#define SPRN_TACR	888
+#define SPRN_TCSCR	889
+#define SPRN_CSIGR	890
+#define SPRN_SPMC1	892
+#define SPRN_SPMC2	893
 
 /* When EBB is enabled, some of MMCR0/MMCR2/SIER are user accessible */
 #define MMCR0_USER_MASK	(MMCR0_FC | MMCR0_PMXE | MMCR0_PMAO)
arch/powerpc/include/uapi/asm/kvm.h +1 -0
index e420d46..26d3fd6 100644
@@ -531,6 +531,7 @@ struct kvm_get_htab_header {
 #define KVM_REG_PPC_TCSCR	(KVM_REG_PPC | KVM_REG_SIZE_U64 | 0xb1)
 #define KVM_REG_PPC_PID		(KVM_REG_PPC | KVM_REG_SIZE_U64 | 0xb2)
 #define KVM_REG_PPC_ACOP	(KVM_REG_PPC | KVM_REG_SIZE_U64 | 0xb3)
+#define KVM_REG_PPC_WORT	(KVM_REG_PPC | KVM_REG_SIZE_U64 | 0xb4)
 
 #define KVM_REG_PPC_VRSAVE	(KVM_REG_PPC | KVM_REG_SIZE_U32 | 0xb4)
 #define KVM_REG_PPC_LPCR	(KVM_REG_PPC | KVM_REG_SIZE_U32 | 0xb5)
arch/powerpc/kernel/asm-offsets.c +23 -0
index 51d7eed..e05f42a 100644
@@ -440,6 +440,7 @@ int main(void)
 	DEFINE(VCPU_XER, offsetof(struct kvm_vcpu, arch.xer));
 	DEFINE(VCPU_CTR, offsetof(struct kvm_vcpu, arch.ctr));
 	DEFINE(VCPU_LR, offsetof(struct kvm_vcpu, arch.lr));
+	DEFINE(VCPU_TAR, offsetof(struct kvm_vcpu, arch.tar));
 	DEFINE(VCPU_CR, offsetof(struct kvm_vcpu, arch.cr));
 	DEFINE(VCPU_PC, offsetof(struct kvm_vcpu, arch.pc));
 #ifdef CONFIG_KVM_BOOK3S_64_HV
@@ -492,11 +493,17 @@ int main(void)
 	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));
+	DEFINE(VCPU_IC, offsetof(struct kvm_vcpu, arch.ic));
+	DEFINE(VCPU_VTB, offsetof(struct kvm_vcpu, arch.vtb));
 	DEFINE(VCPU_DSCR, offsetof(struct kvm_vcpu, arch.dscr));
 	DEFINE(VCPU_AMR, offsetof(struct kvm_vcpu, arch.amr));
 	DEFINE(VCPU_UAMOR, offsetof(struct kvm_vcpu, arch.uamor));
+	DEFINE(VCPU_IAMR, offsetof(struct kvm_vcpu, arch.iamr));
 	DEFINE(VCPU_CTRL, offsetof(struct kvm_vcpu, arch.ctrl));
 	DEFINE(VCPU_DABR, offsetof(struct kvm_vcpu, arch.dabr));
+	DEFINE(VCPU_DAWR, offsetof(struct kvm_vcpu, arch.dawr));
+	DEFINE(VCPU_DAWRX, offsetof(struct kvm_vcpu, arch.dawrx));
+	DEFINE(VCPU_CIABR, offsetof(struct kvm_vcpu, arch.ciabr));
 	DEFINE(VCPU_HFLAGS, offsetof(struct kvm_vcpu, arch.hflags));
 	DEFINE(VCPU_DEC, offsetof(struct kvm_vcpu, arch.dec));
 	DEFINE(VCPU_DEC_EXPIRES, offsetof(struct kvm_vcpu, arch.dec_expires));
@@ -505,8 +512,10 @@ int main(void)
 	DEFINE(VCPU_PRODDED, offsetof(struct kvm_vcpu, arch.prodded));
 	DEFINE(VCPU_MMCR, offsetof(struct kvm_vcpu, arch.mmcr));
 	DEFINE(VCPU_PMC, offsetof(struct kvm_vcpu, arch.pmc));
+	DEFINE(VCPU_SPMC, offsetof(struct kvm_vcpu, arch.spmc));
 	DEFINE(VCPU_SIAR, offsetof(struct kvm_vcpu, arch.siar));
 	DEFINE(VCPU_SDAR, offsetof(struct kvm_vcpu, arch.sdar));
+	DEFINE(VCPU_SIER, offsetof(struct kvm_vcpu, arch.sier));
 	DEFINE(VCPU_SLB, offsetof(struct kvm_vcpu, arch.slb));
 	DEFINE(VCPU_SLB_MAX, offsetof(struct kvm_vcpu, arch.slb_max));
 	DEFINE(VCPU_SLB_NR, offsetof(struct kvm_vcpu, arch.slb_nr));
@@ -517,6 +526,19 @@ int main(void)
 	DEFINE(VCPU_IS_MASTER, offsetof(struct kvm_vcpu, arch.is_master));
 	DEFINE(VCPU_CFAR, offsetof(struct kvm_vcpu, arch.cfar));
 	DEFINE(VCPU_PPR, offsetof(struct kvm_vcpu, arch.ppr));
+	DEFINE(VCPU_FSCR, offsetof(struct kvm_vcpu, arch.fscr));
+	DEFINE(VCPU_PSPB, offsetof(struct kvm_vcpu, arch.pspb));
+	DEFINE(VCPU_TFHAR, offsetof(struct kvm_vcpu, arch.tfhar));
+	DEFINE(VCPU_TFIAR, offsetof(struct kvm_vcpu, arch.tfiar));
+	DEFINE(VCPU_TEXASR, offsetof(struct kvm_vcpu, arch.texasr));
+	DEFINE(VCPU_EBBHR, offsetof(struct kvm_vcpu, arch.ebbhr));
+	DEFINE(VCPU_EBBRR, offsetof(struct kvm_vcpu, arch.ebbrr));
+	DEFINE(VCPU_BESCR, offsetof(struct kvm_vcpu, arch.bescr));
+	DEFINE(VCPU_CSIGR, offsetof(struct kvm_vcpu, arch.csigr));
+	DEFINE(VCPU_TACR, offsetof(struct kvm_vcpu, arch.tacr));
+	DEFINE(VCPU_TCSCR, offsetof(struct kvm_vcpu, arch.tcscr));
+	DEFINE(VCPU_ACOP, offsetof(struct kvm_vcpu, arch.acop));
+	DEFINE(VCPU_WORT, offsetof(struct kvm_vcpu, arch.wort));
 	DEFINE(VCORE_ENTRY_EXIT, offsetof(struct kvmppc_vcore, entry_exit_count));
 	DEFINE(VCORE_NAP_COUNT, offsetof(struct kvmppc_vcore, nap_count));
 	DEFINE(VCORE_IN_GUEST, offsetof(struct kvmppc_vcore, in_guest));
@@ -525,6 +547,7 @@ int main(void)
 	DEFINE(VCORE_KVM, offsetof(struct kvmppc_vcore, kvm));
 	DEFINE(VCORE_TB_OFFSET, offsetof(struct kvmppc_vcore, tb_offset));
 	DEFINE(VCORE_PCR, offsetof(struct kvmppc_vcore, pcr));
+	DEFINE(VCORE_DPDES, offsetof(struct kvmppc_vcore, dpdes));
 	DEFINE(VCPU_SVCPU, offsetof(struct kvmppc_vcpu_book3s, shadow_vcpu) -
 			   offsetof(struct kvmppc_vcpu_book3s, vcpu));
 	DEFINE(VCPU_SLB_E, offsetof(struct kvmppc_slb, orige));
arch/powerpc/kvm/book3s_hv.c +151 -2

189 lines not on this page · raw message

arch/powerpc/kvm/book3s_hv_rmhandlers.S +144 -0

180 lines not on this page · raw message

Signature · 1 line
1.8.4.rc3
#4 lore raw

RFC PATCH 4/10 KVM: PPC: Book3S HV: Flush the correct number of TLB sets on POWER8

In reply to Paul Mackerras

POWER8 has 512 sets in the TLB, compared to 128 for POWER7, so we need to do more tlbiel instructions when flushing the TLB on POWER8.

---
 arch/powerpc/kvm/book3s_hv_rmhandlers.S | 8 +++++++-
 1 file changed, 7 insertions(+), 1 deletion(-)
arch/powerpc/kvm/book3s_hv_rmhandlers.S +7 -1
index 1de0b65..612c9c8 100644
@@ -667,7 +667,13 @@ END_FTR_SECTION_IFSET(CPU_FTR_ARCH_201)
 	andc	r7,r7,r0
 	stdcx.	r7,0,r6
 	bne	23b
-	li	r6,128			/* and flush the TLB */
+	/* Flush the TLB of any entries for this LPID */
+	/* use arch 2.07S as a proxy for POWER8 */
+BEGIN_FTR_SECTION
+	li	r6,512			/* POWER8 has 512 sets */
+FTR_SECTION_ELSE
+	li	r6,128			/* POWER7 has 128 sets */
+ALT_FTR_SECTION_END_IFSET(CPU_FTR_ARCH_207S)
 	mtctr	r6
 	li	r7,0x800		/* IS field = 0b10 */
 	ptesync
Signature · 1 line
1.8.4.rc3
#5 lore raw

RFC PATCH 5/10 KVM: PPC: Book3S HV: Add handler for HV facility unavailable

In reply to Paul Mackerras

From: Michael Ellerman <[email protected]>

At present this should never happen, since the host kernel sets HFSCR to allow access to all facilities. It's better to be prepared to handle it cleanly if it does ever happen, though.

---
 arch/powerpc/include/asm/kvm_asm.h | 1 +
 arch/powerpc/kvm/book3s_hv.c       | 9 +++++++++
 2 files changed, 10 insertions(+)
arch/powerpc/include/asm/kvm_asm.h +1 -0
index 851bac7..f6401eb 100644
@@ -99,6 +99,7 @@
 #define BOOK3S_INTERRUPT_PERFMON	0xf00
 #define BOOK3S_INTERRUPT_ALTIVEC	0xf20
 #define BOOK3S_INTERRUPT_VSX		0xf40
+#define BOOK3S_INTERRUPT_H_FAC_UNAVAIL	0xf80
 
 #define BOOK3S_IRQPRIO_SYSTEM_RESET		0
 #define BOOK3S_IRQPRIO_DATA_SEGMENT		1
arch/powerpc/kvm/book3s_hv.c +9 -0
index 724daa5..da8619e 100644
@@ -704,6 +704,15 @@ static int kvmppc_handle_exit(struct kvm_run *run, struct kvm_vcpu *vcpu,
 		kvmppc_core_queue_program(vcpu, 0x80000);
 		r = RESUME_GUEST;
 		break;
+	/*
+	 * This occurs if the guest (kernel or userspace), does something that
+	 * is prohibited by HFSCR.  We just generate a program interrupt to
+	 * the guest.
+	 */
+	case BOOK3S_INTERRUPT_H_FAC_UNAVAIL:
+		kvmppc_core_queue_program(vcpu, 0x80000);
+		r = RESUME_GUEST;
+		break;
 	default:
 		kvmppc_dump_regs(vcpu);
 		printk(KERN_EMERG "trap=0x%x | pc=0x%lx | msr=0x%llx\n",
Signature · 1 line
1.8.4.rc3
#6 lore raw

Re: [RFC PATCH 05/10] KVM: PPC: Book3S HV: Add handler for HV facility unavailable

Quoted 30 lines · Paul Mackerras

From: Michael Ellerman <[email protected]>

At present this should never happen, since the host kernel sets HFSCR to allow access to all facilities. It's better to be prepared to handle it cleanly if it does ever happen, though.

---
arch/powerpc/include/asm/kvm_asm.h | 1 +
arch/powerpc/kvm/book3s_hv.c | 9 +++++++++

2 files changed, 10 insertions(+)
arch/powerpc/include/asm/kvm_asm.h +1 -0
index 851bac7..f6401eb 100644
@@ -99,6 +99,7 @@
 #define BOOK3S_INTERRUPT_PERFMON	0xf00
 #define BOOK3S_INTERRUPT_ALTIVEC	0xf20
 #define BOOK3S_INTERRUPT_VSX		0xf40
+#define BOOK3S_INTERRUPT_H_FAC_UNAVAIL	0xf80
 
 #define BOOK3S_IRQPRIO_SYSTEM_RESET		0
 #define BOOK3S_IRQPRIO_DATA_SEGMENT		1
arch/powerpc/kvm/book3s_hv.c +0 -0
index 724daa5..da8619e 100644

arch/powerpc/kvm/book3s_hv.c static int kvmppc_handle_exit(struct kvm_run *run, struct kvm_vcpu *vcpu, · line 713

Quoted 10 lines
arch/powerpc/kvm/book3s_hv.c +7 -0
@@ -704,6 +704,15 @@ static int kvmppc_handle_exit(struct kvm_run *run, struct kvm_vcpu *vcpu,
 		kvmppc_core_queue_program(vcpu, 0x80000);
 		r = RESUME_GUEST;
 		break;
+	/*
+	 * This occurs if the guest (kernel or userspace), does something that
+	 * is prohibited by HFSCR.  We just generate a program interrupt to
+	 * the guest.
+	 */
+	case BOOK3S_INTERRUPT_H_FAC_UNAVAIL:
+		kvmppc_core_queue_program(vcpu, 0x80000);

Would be nice to have a name for that bit.

Alex

Quoted 8 lines
arch/powerpc/kvm/book3s_hv.c +2 -0
+		r = RESUME_GUEST;
+		break;
 	default:
 		kvmppc_dump_regs(vcpu);
 		printk(KERN_EMERG "trap=0x%x | pc=0x%lx | msr=0x%llx\n",
Signature · 1 line
1.8.4.rc3
#7 lore raw

RFC PATCH 6/10 KVM: PPC: Book3S HV: Implement architecture compatibility modes for POWER8

In reply to Paul Mackerras

This allows us to select architecture 2.05 (POWER6) or 2.06 (POWER7) compatibility modes on a POWER8 processor.

---
 arch/powerpc/include/asm/reg.h |  2 ++
 arch/powerpc/kvm/book3s_hv.c   | 16 +++++++++++++++-
 2 files changed, 17 insertions(+), 1 deletion(-)
arch/powerpc/include/asm/reg.h +2 -0
index 4ca8b85..483e0a2 100644
@@ -315,6 +315,8 @@
 #define SPRN_PCR	0x152	/* Processor compatibility register */
 #define   PCR_VEC_DIS	(1ul << (63-0))	/* Vec. disable (pre POWER8) */
 #define   PCR_VSX_DIS	(1ul << (63-1))	/* VSX disable (pre POWER8) */
+#define   PCR_TM_DIS	(1ul << (63-2))	/* Trans. memory disable (POWER8) */
+#define   PCR_ARCH_206	0x4		/* Architecture 2.06 */
 #define   PCR_ARCH_205	0x2		/* Architecture 2.05 */
 #define	SPRN_HEIR	0x153	/* Hypervisor Emulated Instruction Register */
 #define SPRN_TLBINDEXR	0x154	/* P7 TLB control register */
arch/powerpc/kvm/book3s_hv.c +15 -1
index da8619e..217041f 100644
@@ -177,14 +177,28 @@ int kvmppc_set_arch_compat(struct kvm_vcpu *vcpu, u32 arch_compat)
 
 		switch (arch_compat) {
 		case PVR_ARCH_205:
-			pcr = PCR_ARCH_205;
+			/*
+			 * If an arch bit is set in PCR, all the defined
+			 * higher-order arch bits also have to be set.
+			 */
+			pcr = PCR_ARCH_206 | PCR_ARCH_205;
 			break;
 		case PVR_ARCH_206:
 		case PVR_ARCH_206p:
+			pcr = PCR_ARCH_206;
+			break;
+		case PVR_ARCH_207:
 			break;
 		default:
 			return -EINVAL;
 		}
+
+		if (!cpu_has_feature(CPU_FTR_ARCH_207S)) {
+			/* POWER7 can't emulate POWER8 */
+			if (!(pcr & PCR_ARCH_206))
+				return -EINVAL;
+			pcr &= ~PCR_ARCH_206;
+		}
 	}
 
 	spin_lock(&vc->lock);
Signature · 1 line
1.8.4.rc3
#8 lore raw

Re: [RFC PATCH 06/10] KVM: PPC: Book3S HV: Implement architecture compatibility modes for POWER8

Quoted 14 lines · Paul Mackerras

This allows us to select architecture 2.05 (POWER6) or 2.06 (POWER7) compatibility modes on a POWER8 processor.

---
arch/powerpc/include/asm/reg.h | 2 ++
arch/powerpc/kvm/book3s_hv.c | 16 +++++++++++++++-

2 files changed, 17 insertions(+), 1 deletion(-)
arch/powerpc/include/asm/reg.h +0 -0
index 4ca8b85..483e0a2 100644

arch/powerpc/include/asm/reg.h line 318

Quoted 4 lines
arch/powerpc/include/asm/reg.h +1 -0
@@ -315,6 +315,8 @@
 #define SPRN_PCR	0x152	/* Processor compatibility register */
 #define   PCR_VEC_DIS	(1ul << (63-0))	/* Vec. disable (pre POWER8) */
 #define   PCR_VSX_DIS	(1ul << (63-1))	/* VSX disable (pre POWER8) */
+#define   PCR_TM_DIS	(1ul << (63-2))	/* Trans. memory disable (POWER8) */

Is this going to get used?

Alex

#9 lore raw

Re: [RFC PATCH 06/10] KVM: PPC: Book3S HV: Implement architecture compatibility modes for POWER8

Quoted 12 lines · Alexander Graf
Quoted 9 lines · Paul Mackerras

This allows us to select architecture 2.05 (POWER6) or 2.06 (POWER7) compatibility modes on a POWER8 processor.

---
arch/powerpc/include/asm/reg.h | 2 ++
arch/powerpc/kvm/book3s_hv.c | 16 +++++++++++++++-

2 files changed, 17 insertions(+), 1 deletion(-)
Quoted 11 lines

arch/powerpc/include/asm/reg.h line 318

Quoted 9 lines
arch/powerpc/include/asm/reg.h +1 -0
index 4ca8b85..483e0a2 100644
@@ -315,6 +315,8 @@
 #define SPRN_PCR	0x152	/* Processor compatibility register */
 #define   PCR_VEC_DIS	(1ul << (63-0))	/* Vec. disable (pre POWER8) */
 #define   PCR_VSX_DIS	(1ul << (63-1))	/* VSX disable (pre POWER8) */
+#define   PCR_TM_DIS	(1ul << (63-2))	/* Trans. memory disable (POWER8) */

Is this going to get used?

Perhaps not, but I thought it worthwhile to document that the bit exists.

Paul.

#10 lore raw

Re: [RFC PATCH 06/10] KVM: PPC: Book3S HV: Implement architecture compatibility modes for POWER8

Am 05.11.2013 um 04:53 schrieb Paul Mackerras <[email protected]>:

Quoted 25 lines · Alexander Graf
Quoted 12 lines · Paul Mackerras
Quoted 10 lines

This allows us to select architecture 2.05 (POWER6) or 2.06 (POWER7) compatibility modes on a POWER8 processor.

---
arch/powerpc/include/asm/reg.h | 2 ++
arch/powerpc/kvm/book3s_hv.c | 16 +++++++++++++++-

2 files changed, 17 insertions(+), 1 deletion(-)
Quoted 11 lines

arch/powerpc/include/asm/reg.h line 318

Quoted 9 lines
arch/powerpc/include/asm/reg.h +1 -0
index 4ca8b85..483e0a2 100644
@@ -315,6 +315,8 @@
 #define SPRN_PCR    0x152    /* Processor compatibility register */
 #define   PCR_VEC_DIS    (1ul << (63-0))    /* Vec. disable (pre POWER8) */
 #define   PCR_VSX_DIS    (1ul << (63-1))    /* VSX disable (pre POWER8) */
+#define   PCR_TM_DIS    (1ul << (63-2))    /* Trans. memory disable (POWER8) */

Is this going to get used?

Quoted 2 lines

Perhaps not, but I thought it worthwhile to document that the bit exists.

But why not? Does that mean we allow TM to be used in p7 compat mode?

Alex

Quoted 2 lines

Paul.

#11 lore raw

Re: [RFC PATCH 06/10] KVM: PPC: Book3S HV: Implement architecture compatibility modes for POWER8

Quoted 29 lines · Alexander Graf

Am 05.11.2013 um 04:53 schrieb Paul Mackerras <[email protected]>:

Quoted 25 lines · Alexander Graf
Quoted 23 lines · Paul Mackerras
Quoted 15 lines

This allows us to select architecture 2.05 (POWER6) or 2.06 (POWER7) compatibility modes on a POWER8 processor.

---
arch/powerpc/include/asm/reg.h | 2 ++
arch/powerpc/kvm/book3s_hv.c | 16 +++++++++++++++-

2 files changed, 17 insertions(+), 1 deletion(-)
arch/powerpc/include/asm/reg.h +0 -0
index 4ca8b85..483e0a2 100644

arch/powerpc/include/asm/reg.h line 318

Quoted 4 lines
arch/powerpc/include/asm/reg.h +1 -0
@@ -315,6 +315,8 @@
 #define SPRN_PCR    0x152    /* Processor compatibility register */
 #define   PCR_VEC_DIS    (1ul << (63-0))    /* Vec. disable (pre POWER8) */
 #define   PCR_VSX_DIS    (1ul << (63-1))    /* VSX disable (pre POWER8) */
+#define   PCR_TM_DIS    (1ul << (63-2))    /* Trans. memory disable (POWER8) */

Is this going to get used?

Quoted 4 lines
Quoted 2 lines

Perhaps not, but I thought it worthwhile to document that the bit exists.

But why not? Does that mean we allow TM to be used in p7 compat mode?

No; TM is disabled if either or both of the PCR_TM_DIS and PCR_ARCH_206 bits are set.

Paul.

#12 lore raw

RFC PATCH 7/10 KVM: PPC: Book3S HV: Consolidate code that checks reason for wake from nap

In reply to Paul Mackerras

Currently in book3s_hv_rmhandlers.S we have three places where we have woken up from nap mode and we check the reason field in SRR1 to see what event woke us up. This consolidates them into a new function, kvmppc_check_wake_reason. It looks at the wake reason field in SRR1, and if it indicates that an external interrupt caused the wakeup, calls kvmppc_read_intr to check what sort of interrupt it was.

This also consolidates the two places where we synthesize an external interrupt (0x500 vector) for the guest. Now, if the guest exit code finds that there was an external interrupt which has been handled (i.e. it was an IPI indicating that there is now an interrupt pending for the guest), it jumps to deliver_guest_interrupt, which is in the last part of the guest entry code, where we synthesize guest external and decrementer interrupts. That code has been streamlined a little and now clears LPCR[MER] when appropriate as well as setting it.

The extra clearing of any pending IPI on a secondary, offline CPU thread before going back to nap mode has been removed. It is no longer necessary now that we have code to read and acknowledge IPIs in the guest exit path.

This fixes a minor bug in the H_CEDE real-mode handling - previously, if we found that other threads were already exiting the guest when we were about to go to nap mode, we would branch to the cede wakeup path and end up looking in SRR1 for a wakeup reason. Now we branch to a point after we have checked the wakeup reason.

This also fixes a minor bug in kvmppc_read_intr - previously it could return 0xff rather than 1, in the case where we find that a host IPI is pending after we have cleared the IPI. Now it returns 1.

---
 arch/powerpc/kvm/book3s_hv_rmhandlers.S | 193 +++++++++++++-------------------
 1 file changed, 78 insertions(+), 115 deletions(-)
arch/powerpc/kvm/book3s_hv_rmhandlers.S +78 -115
index 612c9c8..6351ce2 100644
@@ -234,8 +234,10 @@ kvm_novcpu_wakeup:
 	stb	r0, HSTATE_NAPPING(r13)
 	stb	r0, HSTATE_HWTHREAD_REQ(r13)
 
+	/* check the wake reason */
+	bl	kvmppc_check_wake_reason
+	
 	/* see if any other thread is already exiting */
-	li	r12, 0
 	lwz	r0, VCORE_ENTRY_EXIT(r5)
 	cmpwi	r0, 0x100
 	bge	kvm_novcpu_exit
@@ -245,23 +247,14 @@ kvm_novcpu_wakeup:
 	li	r0, 1
 	sld	r0, r0, r7
 	addi	r6, r5, VCORE_NAPPING_THREADS
-4:	lwarx	r3, 0, r6
-	andc	r3, r3, r0
-	stwcx.	r3, 0, r6
+4:	lwarx	r7, 0, r6
+	andc	r7, r7, r0
+	stwcx.	r7, 0, r6
 	bne	4b
 
-	/* Check the wake reason in SRR1 to see why we got here */
-	mfspr	r3, SPRN_SRR1
-	rlwinm	r3, r3, 44-31, 0x7	/* extract wake reason field */
-	cmpwi	r3, 4			/* was it an external interrupt? */
-	bne	kvm_novcpu_exit		/* if not, exit the guest */
-
-	/* extern interrupt - read and handle it */
-	li	r12, BOOK3S_INTERRUPT_EXTERNAL
-	bl	kvmppc_read_intr
+	/* See if the wake reason means we need to exit */
 	cmpdi	r3, 0
 	bge	kvm_novcpu_exit
-	li	r12, 0
 
 	/* Got an IPI but other vcpus aren't yet exiting, must be a latecomer */
 	ld	r4, HSTATE_KVM_VCPU(r13)
@@ -325,58 +318,23 @@ kvm_start_guest:
 	 */
 
 	/* Check the wake reason in SRR1 to see why we got here */
-	mfspr	r3,SPRN_SRR1
-	rlwinm	r3,r3,44-31,0x7		/* extract wake reason field */
-	cmpwi	r3,4			/* was it an external interrupt? */
-	bne	27f			/* if not */
-	ld	r5,HSTATE_XICS_PHYS(r13)
-	li	r7,XICS_XIRR		/* if it was an external interrupt, */
-	lwzcix	r8,r5,r7		/* get and ack the interrupt */
-	sync
-	clrldi.	r9,r8,40		/* get interrupt source ID. */
-	beq	28f			/* none there? */
-	cmpwi	r9,XICS_IPI		/* was it an IPI? */
-	bne	29f
-	li	r0,0xff
-	li	r6,XICS_MFRR
-	stbcix	r0,r5,r6		/* clear IPI */
-	stwcix	r8,r5,r7		/* EOI the interrupt */
-	sync				/* order loading of vcpu after that */
+	bl	kvmppc_check_wake_reason
+	cmpdi	r3, 0
+	bge	kvm_no_guest
 
 	/* get vcpu pointer, NULL if we have no vcpu to run */
 	ld	r4,HSTATE_KVM_VCPU(r13)
 	cmpdi	r4,0
 	/* if we have no vcpu to run, go back to sleep */
 	beq	kvm_no_guest
-	b	30f
 
-27:	/* XXX should handle hypervisor maintenance interrupts etc. here */
-	b	kvm_no_guest
-28:	/* SRR1 said external but ICP said nope?? */
-	b	kvm_no_guest
-29:	/* External non-IPI interrupt to offline secondary thread? help?? */
-	stw	r8,HSTATE_SAVED_XIRR(r13)
-	b	kvm_no_guest
-
-30:	bl	kvmppc_hv_entry
+	bl	kvmppc_hv_entry
 
 	/* Back from the guest, go back to nap */
 	/* Clear our vcpu pointer so we don't come back in early */
 	li	r0, 0
 	std	r0, HSTATE_KVM_VCPU(r13)
 	lwsync
-	/* Clear any pending IPI - we're an offline thread */
-	ld	r5, HSTATE_XICS_PHYS(r13)
-	li	r7, XICS_XIRR
-	lwzcix	r3, r5, r7		/* ack any pending interrupt */
-	rlwinm.	r0, r3, 0, 0xffffff	/* any pending? */
-	beq	37f
-	sync
-	li	r0, 0xff
-	li	r6, XICS_MFRR
-	stbcix	r0, r5, r6		/* clear the IPI */
-	stwcix	r3, r5, r7		/* EOI it */
-37:	sync
 
 	/* increment the nap count and then go to nap mode */
 	ld	r4, HSTATE_KVM_VCORE(r13)
@@ -862,47 +820,46 @@ toc_tlbie_lock:
 	mtctr	r6
 	mtxer	r7
 
+kvmppc_cede_reentry:		/* r4 = vcpu, r13 = paca */
 	ld	r10, VCPU_PC(r4)
 	ld	r11, VCPU_MSR(r4)
-kvmppc_cede_reentry:		/* r4 = vcpu, r13 = paca */
 	ld	r6, VCPU_SRR0(r4)
 	ld	r7, VCPU_SRR1(r4)
+	mtspr	SPRN_SRR0, r6
+	mtspr	SPRN_SRR1, r7
 
+deliver_guest_interrupt:
 	/* r11 = vcpu->arch.msr & ~MSR_HV */
 	rldicl	r11, r11, 63 - MSR_HV_LG, 1
 	rotldi	r11, r11, 1 + MSR_HV_LG
 	ori	r11, r11, MSR_ME
 
 	/* Check if we can deliver an external or decrementer interrupt now */
-	ld	r0,VCPU_PENDING_EXC(r4)
-	lis	r8,(1 << BOOK3S_IRQPRIO_EXTERNAL_LEVEL)@h
-	and	r0,r0,r8
-	cmpdi	cr1,r0,0
-	andi.	r0,r11,MSR_EE
-	beq	cr1,11f
+	ld	r0, VCPU_PENDING_EXC(r4)
+	rldicl	r0, r0, 64 - BOOK3S_IRQPRIO_EXTERNAL_LEVEL, 63
+	cmpdi	cr1, r0, 0
+	andi.	r8, r11, MSR_EE
 BEGIN_FTR_SECTION
-	mfspr	r8,SPRN_LPCR
-	ori	r8,r8,LPCR_MER
-	mtspr	SPRN_LPCR,r8
+	mfspr	r8, SPRN_LPCR
+	/* Insert EXTERNAL_LEVEL bit into LPCR at the MER bit position */
+	rldimi	r8, r0, LPCR_MER_SH, 63 - LPCR_MER_SH
+	mtspr	SPRN_LPCR, r8
 	isync
 END_FTR_SECTION_IFSET(CPU_FTR_ARCH_206)
 	beq	5f
-	li	r0,BOOK3S_INTERRUPT_EXTERNAL
-12:	mr	r6,r10
+	li	r0, BOOK3S_INTERRUPT_EXTERNAL
+	bne	cr1, 12f
+	mfspr	r0, SPRN_DEC
+	cmpwi	r0, 0
+	li	r0, BOOK3S_INTERRUPT_DECREMENTER
+	bge	5f
+
+12:	mtspr	SPRN_SRR0, r10
 	mr	r10,r0
-	mr	r7,r11
+	mtspr	SPRN_SRR1, r11
 	li	r11,(MSR_ME << 1) | 1	/* synthesize MSR_SF | MSR_ME */
 	rotldi	r11,r11,63
-	b	5f
-11:	beq	5f
-	mfspr	r0,SPRN_DEC
-	cmpwi	r0,0
-	li	r0,BOOK3S_INTERRUPT_DECREMENTER
-	blt	12b
-
-	/* Move SRR0 and SRR1 into the respective regs */
-5:	mtspr	SPRN_SRR0, r6
-	mtspr	SPRN_SRR1, r7
+5:
 
 fast_guest_return:
 	li	r0,0
@@ -1075,39 +1032,19 @@ END_FTR_SECTION_IFCLR(CPU_FTR_ARCH_206)
 	/* External interrupt, first check for host_ipi. If this is
 	 * set, we know the host wants us out so let's do it now
 	 */
-do_ext_interrupt:
 	bl	kvmppc_read_intr
 	cmpdi	r3, 0
 	bgt	ext_interrupt_to_host
 
-	/* Allright, looks like an IPI for the guest, we need to set MER */
 	/* Check if any CPU is heading out to the host, if so head out too */
 	ld	r5, HSTATE_KVM_VCORE(r13)
 	lwz	r0, VCORE_ENTRY_EXIT(r5)
 	cmpwi	r0, 0x100
 	bge	ext_interrupt_to_host
 
-	/* See if there is a pending interrupt for the guest */
-	mfspr	r8, SPRN_LPCR
-	ld	r0, VCPU_PENDING_EXC(r9)
-	/* Insert EXTERNAL_LEVEL bit into LPCR at the MER bit position */
-	rldicl.	r0, r0, 64 - BOOK3S_IRQPRIO_EXTERNAL_LEVEL, 63
-	rldimi	r8, r0, LPCR_MER_SH, 63 - LPCR_MER_SH
-	beq	2f
-
-	/* And if the guest EE is set, we can deliver immediately, else
-	 * we return to the guest with MER set
-	 */
-	andi.	r0, r11, MSR_EE
-	beq	2f
-	mtspr	SPRN_SRR0, r10
-	mtspr	SPRN_SRR1, r11
-	li	r10, BOOK3S_INTERRUPT_EXTERNAL
-	li	r11, (MSR_ME << 1) | 1	/* synthesize MSR_SF | MSR_ME */
-	rotldi	r11, r11, 63
-2:	mr	r4, r9
-	mtspr	SPRN_LPCR, r8
-	b	fast_guest_return
+	/* Return to guest after delivering any pending interrupt */
+	mr	r4, r9
+	b	deliver_guest_interrupt
 
 ext_interrupt_to_host:
 
@@ -1923,7 +1860,6 @@ END_FTR_SECTION_IFCLR(CPU_FTR_ARCH_206)
 	stb	r0,HSTATE_NAPPING(r13)
 	/* order napping_threads update vs testing entry_exit_count */
 	lwsync
-	mr	r4,r3
 	lwz	r7,VCORE_ENTRY_EXIT(r5)
 	cmpwi	r7,0x100
 	bge	33f		/* another thread already exiting */
@@ -1976,6 +1912,11 @@ END_FTR_SECTION_IFCLR(CPU_FTR_ARCH_206)
 	nap
 	b	.
 
+33:	mr	r4, r3
+	li	r3, 0
+	li	r12, 0
+	b	34f
+
 kvm_end_cede:
 	/* get vcpu pointer */
 	ld	r4, HSTATE_KVM_VCPU(r13)
@@ -2005,12 +1946,15 @@ kvm_end_cede:
 	ld	r29, VCPU_GPR(R29)(r4)
 	ld	r30, VCPU_GPR(R30)(r4)
 	ld	r31, VCPU_GPR(R31)(r4)
+ 
+	/* Check the wake reason in SRR1 to see why we got here */
+	bl	kvmppc_check_wake_reason
 
 	/* clear our bit in vcore->napping_threads */
-33:	ld	r5,HSTATE_KVM_VCORE(r13)
-	lbz	r3,HSTATE_PTID(r13)
+34:	ld	r5,HSTATE_KVM_VCORE(r13)
+	lbz	r7,HSTATE_PTID(r13)
 	li	r0,1
-	sld	r0,r0,r3
+	sld	r0,r0,r7
 	addi	r6,r5,VCORE_NAPPING_THREADS
 32:	lwarx	r7,0,r6
 	andc	r7,r7,r0
@@ -2019,23 +1963,18 @@ kvm_end_cede:
 	li	r0,0
 	stb	r0,HSTATE_NAPPING(r13)
 
-	/* Check the wake reason in SRR1 to see why we got here */
-	mfspr	r3, SPRN_SRR1
-	rlwinm	r3, r3, 44-31, 0x7	/* extract wake reason field */
-	cmpwi	r3, 4			/* was it an external interrupt? */
-	li	r12, BOOK3S_INTERRUPT_EXTERNAL
+	/* See if the wake reason means we need to exit */
+	stw	r12, VCPU_TRAP(r4)
 	mr	r9, r4
-	ld	r10, VCPU_PC(r9)
-	ld	r11, VCPU_MSR(r9)
-	beq	do_ext_interrupt	/* if so */
+	cmpdi	r3, 0
+	bgt	guest_exit_cont
 
 	/* see if any other thread is already exiting */
 	lwz	r0,VCORE_ENTRY_EXIT(r5)
 	cmpwi	r0,0x100
-	blt	kvmppc_cede_reentry	/* if not go back to guest */
+	bge	guest_exit_cont
 
-	/* some threads are exiting, so go to the guest exit path */
-	b	hcall_real_fallback
+	b	kvmppc_cede_reentry	/* if not go back to guest */
 
 	/* cede when already previously prodded case */
 kvm_cede_prodded:
@@ -2066,6 +2005,29 @@ machine_check_realmode:
 	b	fast_interrupt_c_return
 
 /*
+ * Check the reason we woke from nap, and take appropriate action.
+ * Returns:
+ *	0 if nothing needs to be done
+ *	1 if something happened that needs to be handled by the host
+ *	-1 if there was a guest wakeup (IPI)
+ *
+ * Also sets r12 to the interrupt vector for any interrupt that needs
+ * to be handled now by the host (0x500 for external interrupt), or zero.
+ */
+kvmppc_check_wake_reason:
+	mfspr	r6, SPRN_SRR1
+	rlwinm	r6, r6, 44-31, 0x7	/* extract wake reason field */
+	cmpwi	r6, 4			/* was it an external interrupt? */
+	li	r12, BOOK3S_INTERRUPT_EXTERNAL
+	beq	kvmppc_read_intr	/* if so, see what it was */
+	li	r3, 0
+	li	r12, 0
+	cmpwi	r6, 6			/* was it the decrementer? */
+	beq	0f
+	li	r3, 1			/* anything else, return 1 */
+0:	blr
+
+/*
  * Determine what sort of external interrupt is pending (if any).
  * Returns:
  *	0 if no interrupt is pending
@@ -2096,7 +2058,6 @@ kvmppc_read_intr:
 	 * interrupts directly to the guest
 	 */
 	cmpwi	r3, XICS_IPI		/* if there is, is it an IPI? */
-	li	r3, 1
 	bne	42f
 
 	/* It's an IPI, clear the MFRR and EOI it */
@@ -2122,12 +2083,14 @@ kvmppc_read_intr:
 	 * before exit, it will be picked up by the host ICP driver
 	 */
 	stw	r0, HSTATE_SAVED_XIRR(r13)
+	li	r3, 1
 	b	1b
 
 43:	/* We raced with the host, we need to resend that IPI, bummer */
 	li	r0, IPI_PRIORITY
 	stbcix	r0, r6, r8		/* set the IPI */
 	sync
+	li	r3, 1
 	b	1b
 
 /*
Signature · 1 line
1.8.4.rc3
#13 lore raw

RFC PATCH 8/10 KVM: PPC: Book3S HV: Handle guest using doorbells for IPIs

In reply to Paul Mackerras

* SRR1 wake reason field for system reset interrupt on wakeup from nap is now a 4-bit field on P8, compared to 3 bits on P7.

* Set PECEDP in LPCR when napping because of H_CEDE so guest doorbells will wake us up.

* Waking up from nap because of a guest doorbell interrupt is not a reason to exit the guest.

---
 arch/powerpc/include/asm/reg.h          |  4 +++-
 arch/powerpc/kvm/book3s_hv_rmhandlers.S | 19 +++++++++++++++----
 2 files changed, 18 insertions(+), 5 deletions(-)
arch/powerpc/include/asm/reg.h +3 -1
index 483e0a2..73fdd62 100644
@@ -293,7 +293,9 @@
 #define   LPCR_ILE     0x02000000      /* !HV irqs set MSR:LE */
 #define   LPCR_AIL_0	0x00000000	/* MMU off exception offset 0x0 */
 #define   LPCR_AIL_3	0x01800000	/* MMU on exception offset 0xc00...4xxx */
-#define   LPCR_PECE	0x00007000	/* powersave exit cause enable */
+#define   LPCR_PECE	0x0001f000	/* powersave exit cause enable */
+#define     LPCR_PECEDP	0x00010000	/* directed priv dbells cause exit */
+#define     LPCR_PECEDH	0x00008000	/* directed hyp dbells cause exit */
 #define     LPCR_PECE0	0x00004000	/* ext. exceptions can cause exit */
 #define     LPCR_PECE1	0x00002000	/* decrementer can cause exit */
 #define     LPCR_PECE2	0x00001000	/* machine check etc can cause exit */
arch/powerpc/kvm/book3s_hv_rmhandlers.S +15 -4
index 6351ce2..4b3a10e 100644
@@ -1894,13 +1894,16 @@ END_FTR_SECTION_IFCLR(CPU_FTR_ARCH_206)
 	bl	.kvmppc_save_fp
 
 	/*
-	 * Take a nap until a decrementer or external interrupt occurs,
-	 * with PECE1 (wake on decr) and PECE0 (wake on external) set in LPCR
+	 * Take a nap until a decrementer or external or doobell interrupt
+	 * occurs, with PECE1, PECE0 and PECEDP set in LPCR
 	 */
 	li	r0,1
 	stb	r0,HSTATE_HWTHREAD_REQ(r13)
 	mfspr	r5,SPRN_LPCR
 	ori	r5,r5,LPCR_PECE0 | LPCR_PECE1
+BEGIN_FTR_SECTION
+	oris	r5,r5,LPCR_PECEDP@h
+END_FTR_SECTION_IFSET(CPU_FTR_ARCH_207S)
 	mtspr	SPRN_LPCR,r5
 	isync
 	li	r0, 0
@@ -2016,14 +2019,22 @@ machine_check_realmode:
  */
 kvmppc_check_wake_reason:
 	mfspr	r6, SPRN_SRR1
-	rlwinm	r6, r6, 44-31, 0x7	/* extract wake reason field */
-	cmpwi	r6, 4			/* was it an external interrupt? */
+BEGIN_FTR_SECTION
+	rlwinm	r6, r6, 45-31, 0xf	/* extract wake reason field (P8) */
+FTR_SECTION_ELSE
+	rlwinm	r6, r6, 45-31, 0xe	/* P7 wake reason field is 3 bits */
+ALT_FTR_SECTION_END_IFSET(CPU_FTR_ARCH_207S)
+	cmpwi	r6, 8			/* was it an external interrupt? */
 	li	r12, BOOK3S_INTERRUPT_EXTERNAL
 	beq	kvmppc_read_intr	/* if so, see what it was */
 	li	r3, 0
 	li	r12, 0
 	cmpwi	r6, 6			/* was it the decrementer? */
 	beq	0f
+BEGIN_FTR_SECTION
+	cmpwi	r6, 5			/* privileged doorbell? */
+	beq	0f
+END_FTR_SECTION_IFSET(CPU_FTR_ARCH_207S)
 	li	r3, 1			/* anything else, return 1 */
 0:	blr
 
Signature · 1 line
1.8.4.rc3
#14 lore raw

RFC PATCH 9/10 KVM: PPC: Book3S HV: Handle new LPCR bits on POWER8

In reply to Paul Mackerras

POWER8 has a bit in the LPCR to enable or disable the PURR and SPURR registers to count when in the guest. Set this bit.

POWER8 has a field in the LPCR called AIL (Alternate Interrupt Location) which is used to enable relocation-on interrupts. Allow userspace to set this field.

---
 arch/powerpc/include/asm/reg.h | 2 ++
 arch/powerpc/kvm/book3s_hv.c   | 6 ++++++
 2 files changed, 8 insertions(+)
arch/powerpc/include/asm/reg.h +2 -0
index 73fdd62..60c2dd8 100644
@@ -291,8 +291,10 @@
 #define   LPCR_RMLS    0x1C000000      /* impl dependent rmo limit sel */
 #define	  LPCR_RMLS_SH	(63-37)
 #define   LPCR_ILE     0x02000000      /* !HV irqs set MSR:LE */
+#define   LPCR_AIL	0x01800000	/* Alternate interrupt location */
 #define   LPCR_AIL_0	0x00000000	/* MMU off exception offset 0x0 */
 #define   LPCR_AIL_3	0x01800000	/* MMU on exception offset 0xc00...4xxx */
+#define   LPCR_ONL	0x00040000	/* online - PURR/SPURR count */
 #define   LPCR_PECE	0x0001f000	/* powersave exit cause enable */
 #define     LPCR_PECEDP	0x00010000	/* directed priv dbells cause exit */
 #define     LPCR_PECEDH	0x00008000	/* directed hyp dbells cause exit */
arch/powerpc/kvm/book3s_hv.c +6 -0
index 217041f..95a635d 100644
@@ -783,8 +783,11 @@ static void kvmppc_set_lpcr(struct kvm_vcpu *vcpu, u64 new_lpcr)
 	/*
 	 * Userspace can only modify DPFD (default prefetch depth),
 	 * ILE (interrupt little-endian) and TC (translation control).
+	 * On POWER8 userspace can also modify AIL (alt. interrupt loc.)
 	 */
 	mask = LPCR_DPFD | LPCR_ILE | LPCR_TC;
+	if (cpu_has_feature(CPU_FTR_ARCH_207S))
+		mask |= LPCR_AIL;
 	kvm->arch.lpcr = (kvm->arch.lpcr & ~mask) | (new_lpcr & mask);
 	mutex_unlock(&kvm->lock);
 }
@@ -2169,6 +2172,9 @@ int kvmppc_core_init_vm(struct kvm *kvm)
 			LPCR_VPM0 | LPCR_VPM1;
 		kvm->arch.vrma_slb_v = SLB_VSID_B_1T |
 			(VRMA_VSID << SLB_VSID_SHIFT_1T);
+		/* On POWER8 turn on online bit to enable PURR/SPURR */
+		if (cpu_has_feature(CPU_FTR_ARCH_207S))
+			lpcr |= LPCR_ONL;
 	}
 	kvm->arch.lpcr = lpcr;
 
Signature · 1 line
1.8.4.rc3
#15 lore raw

RFC PATCH 10/10 KVM: PPC: Book3S HV: Prepare for host using hypervisor doorbells

In reply to Paul Mackerras

POWER8 has support for hypervisor doorbell interrupts. Though the kernel doesn't use them for IPIs on the powernv platform yet, it probably will in future, so this makes KVM cope gracefully if a hypervisor doorbell interrupt arrives while in a guest.

---
 arch/powerpc/include/asm/kvm_asm.h      | 1 +
 arch/powerpc/kvm/book3s_hv.c            | 1 +
 arch/powerpc/kvm/book3s_hv_rmhandlers.S | 7 +++++++
 3 files changed, 9 insertions(+)
arch/powerpc/include/asm/kvm_asm.h +1 -0
index f6401eb..4c2040a 100644
@@ -96,6 +96,7 @@
 #define BOOK3S_INTERRUPT_H_DATA_STORAGE	0xe00
 #define BOOK3S_INTERRUPT_H_INST_STORAGE	0xe20
 #define BOOK3S_INTERRUPT_H_EMUL_ASSIST	0xe40
+#define BOOK3S_INTERRUPT_H_DOORBELL	0xe80
 #define BOOK3S_INTERRUPT_PERFMON	0xf00
 #define BOOK3S_INTERRUPT_ALTIVEC	0xf20
 #define BOOK3S_INTERRUPT_VSX		0xf40
arch/powerpc/kvm/book3s_hv.c +1 -0
index 95a635d..dc5ce77 100644
@@ -644,6 +644,7 @@ static int kvmppc_handle_exit(struct kvm_run *run, struct kvm_vcpu *vcpu,
 		r = RESUME_GUEST;
 		break;
 	case BOOK3S_INTERRUPT_EXTERNAL:
+	case BOOK3S_INTERRUPT_H_DOORBELL:
 		vcpu->stat.ext_intr_exits++;
 		r = RESUME_GUEST;
 		break;
arch/powerpc/kvm/book3s_hv_rmhandlers.S +7 -0
index 4b3a10e..018513a 100644
@@ -2034,10 +2034,17 @@ ALT_FTR_SECTION_END_IFSET(CPU_FTR_ARCH_207S)
 BEGIN_FTR_SECTION
 	cmpwi	r6, 5			/* privileged doorbell? */
 	beq	0f
+	cmpwi	r6, 3			/* hypervisor doorbell? */
+	beq	3f
 END_FTR_SECTION_IFSET(CPU_FTR_ARCH_207S)
 	li	r3, 1			/* anything else, return 1 */
 0:	blr
 
+	/* hypervisor doorbell */
+3:	li	r12, BOOK3S_INTERRUPT_H_DOORBELL
+	li	r3, 1
+	blr
+
 /*
  * Determine what sort of external interrupt is pending (if any).
  * Returns:
Signature · 1 line
1.8.4.rc3