Merge branch 'topic/ppc-kvm' into next

Merge the topic branch we share with kvm-ppc, this brings in two xive
commits, one from Paul to rework HMI handling, and a minor cleanup to
drop an unused flag.

11 files changed, 206 insertions, 78 deletions

diff --git a/arch/powerpc/include/asm/hmi.h b/arch/powerpc/include/asm/hmi.h
index 85b7a1a21e22..9c14f7b5c46c 100644
--- a/arch/powerpc/include/asm/hmi.h
+++ b/arch/powerpc/include/asm/hmi.h
@@ -42,4 +42,8 @@ extern void wait_for_tb_resync(void);
 static inline void wait_for_subcore_guest_exit(void) { }
 static inline void wait_for_tb_resync(void) { }
 #endif
+
+struct pt_regs;
+extern long hmi_handle_debugtrig(struct pt_regs *regs);
+
 #endif /* __ASM_PPC64_HMI_H__ */
diff --git a/arch/powerpc/include/asm/hvcall.h b/arch/powerpc/include/asm/hvcall.h
index a409177be8bd..f0461618bf7b 100644
--- a/arch/powerpc/include/asm/hvcall.h
+++ b/arch/powerpc/include/asm/hvcall.h
@@ -241,6 +241,7 @@
 #define H_GET_HCA_INFO          0x1B8
 #define H_GET_PERF_COUNT        0x1BC
 #define H_MANAGE_TRACE          0x1C0
+#define H_GET_CPU_CHARACTERISTICS 0x1C8
 #define H_FREE_LOGICAL_LAN_BUFFER 0x1D4
 #define H_QUERY_INT_STATE       0x1E4
 #define H_POLL_PENDING		0x1D8
@@ -330,6 +331,17 @@
 #define H_SIGNAL_SYS_RESET_ALL_OTHERS		-2
 /* >= 0 values are CPU number */
 
+/* H_GET_CPU_CHARACTERISTICS return values */
+#define H_CPU_CHAR_SPEC_BAR_ORI31	(1ull << 63) // IBM bit 0
+#define H_CPU_CHAR_BCCTRL_SERIALISED	(1ull << 62) // IBM bit 1
+#define H_CPU_CHAR_L1D_FLUSH_ORI30	(1ull << 61) // IBM bit 2
+#define H_CPU_CHAR_L1D_FLUSH_TRIG2	(1ull << 60) // IBM bit 3
+#define H_CPU_CHAR_L1D_THREAD_PRIV	(1ull << 59) // IBM bit 4
+
+#define H_CPU_BEHAV_FAVOUR_SECURITY	(1ull << 63) // IBM bit 0
+#define H_CPU_BEHAV_L1D_FLUSH_PR	(1ull << 62) // IBM bit 1
+#define H_CPU_BEHAV_BNDS_CHK_SPEC_BAR	(1ull << 61) // IBM bit 2
+
 /* Flag values used in H_REGISTER_PROC_TBL hcall */
 #define PROC_TABLE_OP_MASK	0x18
 #define PROC_TABLE_DEREG	0x10
@@ -436,6 +448,11 @@ static inline unsigned int get_longbusy_msecs(int longbusy_rc)
 	}
 }
 
+struct h_cpu_char_result {
+	u64 character;
+	u64 behaviour;
+};
+
 #endif /* __ASSEMBLY__ */
 #endif /* __KERNEL__ */
 #endif /* _ASM_POWERPC_HVCALL_H */
diff --git a/arch/powerpc/include/asm/plpar_wrappers.h b/arch/powerpc/include/asm/plpar_wrappers.h
index 7f01b22fa6cb..55eddf50d149 100644
--- a/arch/powerpc/include/asm/plpar_wrappers.h
+++ b/arch/powerpc/include/asm/plpar_wrappers.h
@@ -326,4 +326,18 @@ static inline long plapr_signal_sys_reset(long cpu)
 	return plpar_hcall_norets(H_SIGNAL_SYS_RESET, cpu);
 }
 
+static inline long plpar_get_cpu_characteristics(struct h_cpu_char_result *p)
+{
+	unsigned long retbuf[PLPAR_HCALL_BUFSIZE];
+	long rc;
+
+	rc = plpar_hcall(H_GET_CPU_CHARACTERISTICS, retbuf);
+	if (rc == H_SUCCESS) {
+		p->character = retbuf[0];
+		p->behaviour = retbuf[1];
+	}
+
+	return rc;
+}
+
 #endif /* _ASM_POWERPC_PLPAR_WRAPPERS_H */
diff --git a/arch/powerpc/include/asm/reg.h b/arch/powerpc/include/asm/reg.h
index ffc9990e8040..e6c7eadf6bce 100644
--- a/arch/powerpc/include/asm/reg.h
+++ b/arch/powerpc/include/asm/reg.h
@@ -431,8 +431,9 @@
 #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 */
+#define	SPRN_HMER	0x150	/* Hypervisor maintenance exception reg */
+#define   HMER_DEBUG_TRIG	(1ul << (63 - 17)) /* Debug trigger */
+#define	SPRN_HMEER	0x151	/* Hyp maintenance exception enable reg */
 #define SPRN_PCR	0x152	/* Processor compatibility register */
 #define   PCR_VEC_DIS	(1ul << (63-0))	/* Vec. disable (bit NA since POWER8) */
 #define   PCR_VSX_DIS	(1ul << (63-1))	/* VSX disable (bit NA since POWER8) */
diff --git a/arch/powerpc/include/asm/xive-regs.h b/arch/powerpc/include/asm/xive-regs.h
index 1d3f2be5ae39..fa4288822b68 100644
--- a/arch/powerpc/include/asm/xive-regs.h
+++ b/arch/powerpc/include/asm/xive-regs.h
@@ -10,6 +10,41 @@
 #define _ASM_POWERPC_XIVE_REGS_H
 
 /*
+ * "magic" Event State Buffer (ESB) MMIO offsets.
+ *
+ * Each interrupt source has a 2-bit state machine called ESB
+ * which can be controlled by MMIO. It's made of 2 bits, P and
+ * Q. P indicates that an interrupt is pending (has been sent
+ * to a queue and is waiting for an EOI). Q indicates that the
+ * interrupt has been triggered while pending.
+ *
+ * This acts as a coalescing mechanism in order to guarantee
+ * that a given interrupt only occurs at most once in a queue.
+ *
+ * When doing an EOI, the Q bit will indicate if the interrupt
+ * needs to be re-triggered.
+ *
+ * The following offsets into the ESB MMIO allow to read or
+ * manipulate the PQ bits. They must be used with an 8-bytes
+ * load instruction. They all return the previous state of the
+ * interrupt (atomically).
+ *
+ * Additionally, some ESB pages support doing an EOI via a
+ * store at 0 and some ESBs support doing a trigger via a
+ * separate trigger page.
+ */
+#define XIVE_ESB_STORE_EOI	0x400 /* Store */
+#define XIVE_ESB_LOAD_EOI	0x000 /* Load */
+#define XIVE_ESB_GET		0x800 /* Load */
+#define XIVE_ESB_SET_PQ_00	0xc00 /* Load */
+#define XIVE_ESB_SET_PQ_01	0xd00 /* Load */
+#define XIVE_ESB_SET_PQ_10	0xe00 /* Load */
+#define XIVE_ESB_SET_PQ_11	0xf00 /* Load */
+
+#define XIVE_ESB_VAL_P		0x2
+#define XIVE_ESB_VAL_Q		0x1
+
+/*
  * Thread Management (aka "TM") registers
  */
 
diff --git a/arch/powerpc/include/asm/xive.h b/arch/powerpc/include/asm/xive.h
index 371fbebf1ec9..b619a5585cd6 100644
--- a/arch/powerpc/include/asm/xive.h
+++ b/arch/powerpc/include/asm/xive.h
@@ -58,6 +58,9 @@ struct xive_irq_data {
 #define XIVE_IRQ_FLAG_EOI_FW	0x10
 #define XIVE_IRQ_FLAG_H_INT_ESB	0x20
 
+/* Special flag set by KVM for excalation interrupts */
+#define XIVE_IRQ_NO_EOI		0x80
+
 #define XIVE_INVALID_CHIP_ID	-1
 
 /* A queue tracking structure in a CPU */
@@ -72,41 +75,6 @@ struct xive_q {
 	atomic_t		pending_count;
 };
 
-/*
- * "magic" Event State Buffer (ESB) MMIO offsets.
- *
- * Each interrupt source has a 2-bit state machine called ESB
- * which can be controlled by MMIO. It's made of 2 bits, P and
- * Q. P indicates that an interrupt is pending (has been sent
- * to a queue and is waiting for an EOI). Q indicates that the
- * interrupt has been triggered while pending.
- *
- * This acts as a coalescing mechanism in order to guarantee
- * that a given interrupt only occurs at most once in a queue.
- *
- * When doing an EOI, the Q bit will indicate if the interrupt
- * needs to be re-triggered.
- *
- * The following offsets into the ESB MMIO allow to read or
- * manipulate the PQ bits. They must be used with an 8-bytes
- * load instruction. They all return the previous state of the
- * interrupt (atomically).
- *
- * Additionally, some ESB pages support doing an EOI via a
- * store at 0 and some ESBs support doing a trigger via a
- * separate trigger page.
- */
-#define XIVE_ESB_STORE_EOI	0x400 /* Store */
-#define XIVE_ESB_LOAD_EOI	0x000 /* Load */
-#define XIVE_ESB_GET		0x800 /* Load */
-#define XIVE_ESB_SET_PQ_00	0xc00 /* Load */
-#define XIVE_ESB_SET_PQ_01	0xd00 /* Load */
-#define XIVE_ESB_SET_PQ_10	0xe00 /* Load */
-#define XIVE_ESB_SET_PQ_11	0xf00 /* Load */
-
-#define XIVE_ESB_VAL_P		0x2
-#define XIVE_ESB_VAL_Q		0x1
-
 /* Global enable flags for the XIVE support */
 extern bool __xive_enabled;
 
diff --git a/arch/powerpc/kernel/mce.c b/arch/powerpc/kernel/mce.c
index 742e4658c5dc..d2fecaec4fec 100644
--- a/arch/powerpc/kernel/mce.c
+++ b/arch/powerpc/kernel/mce.c
@@ -495,37 +495,123 @@ long machine_check_early(struct pt_regs *regs)
 	return handled;
 }
 
-long hmi_exception_realmode(struct pt_regs *regs)
+/* Possible meanings for HMER_DEBUG_TRIG bit being set on POWER9 */
+static enum {
+	DTRIG_UNKNOWN,
+	DTRIG_VECTOR_CI,	/* need to emulate vector CI load instr */
+	DTRIG_SUSPEND_ESCAPE,	/* need to escape from TM suspend mode */
+} hmer_debug_trig_function;
+
+static int init_debug_trig_function(void)
 {
-	__this_cpu_inc(irq_stat.hmi_exceptions);
-
-#ifdef CONFIG_PPC_BOOK3S_64
-	/* Workaround for P9 vector CI loads (see p9_hmi_special_emu) */
-	if (pvr_version_is(PVR_POWER9)) {
-		unsigned long hmer = mfspr(SPRN_HMER);
-
-		/* Do we have the debug bit set */
-		if (hmer & PPC_BIT(17)) {
-			hmer &= ~PPC_BIT(17);
-			mtspr(SPRN_HMER, hmer);
-
-			/*
-			 * Now to avoid problems with soft-disable we
-			 * only do the emulation if we are coming from
-			 * user space
-			 */
-			if (user_mode(regs))
-				local_paca->hmi_p9_special_emu = 1;
-
-			/*
-			 * Don't bother going to OPAL if that's the
-			 * only relevant bit.
-			 */
-			if (!(hmer & mfspr(SPRN_HMEER)))
-				return local_paca->hmi_p9_special_emu;
+	int pvr;
+	struct device_node *cpun;
+	struct property *prop = NULL;
+	const char *str;
+
+	/* First look in the device tree */
+	preempt_disable();
+	cpun = of_get_cpu_node(smp_processor_id(), NULL);
+	if (cpun) {
+		of_property_for_each_string(cpun, "ibm,hmi-special-triggers",
+					    prop, str) {
+			if (strcmp(str, "bit17-vector-ci-load") == 0)
+				hmer_debug_trig_function = DTRIG_VECTOR_CI;
+			else if (strcmp(str, "bit17-tm-suspend-escape") == 0)
+				hmer_debug_trig_function = DTRIG_SUSPEND_ESCAPE;
 		}
+		of_node_put(cpun);
+	}
+	preempt_enable();
+
+	/* If we found the property, don't look at PVR */
+	if (prop)
+		goto out;
+
+	pvr = mfspr(SPRN_PVR);
+	/* Check for POWER9 Nimbus (scale-out) */
+	if ((PVR_VER(pvr) == PVR_POWER9) && (pvr & 0xe000) == 0) {
+		/* DD2.2 and later */
+		if ((pvr & 0xfff) >= 0x202)
+			hmer_debug_trig_function = DTRIG_SUSPEND_ESCAPE;
+		/* DD2.0 and DD2.1 - used for vector CI load emulation */
+		else if ((pvr & 0xfff) >= 0x200)
+			hmer_debug_trig_function = DTRIG_VECTOR_CI;
+	}
+
+ out:
+	switch (hmer_debug_trig_function) {
+	case DTRIG_VECTOR_CI:
+		pr_debug("HMI debug trigger used for vector CI load\n");
+		break;
+	case DTRIG_SUSPEND_ESCAPE:
+		pr_debug("HMI debug trigger used for TM suspend escape\n");
+		break;
+	default:
+		break;
 	}
-#endif /* CONFIG_PPC_BOOK3S_64 */
+	return 0;
+}
+__initcall(init_debug_trig_function);
+
+/*
+ * Handle HMIs that occur as a result of a debug trigger.
+ * Return values:
+ * -1 means this is not a HMI cause that we know about
+ *  0 means no further handling is required
+ *  1 means further handling is required
+ */
+long hmi_handle_debugtrig(struct pt_regs *regs)
+{
+	unsigned long hmer = mfspr(SPRN_HMER);
+	long ret = 0;
+
+	/* HMER_DEBUG_TRIG bit is used for various workarounds on P9 */
+	if (!((hmer & HMER_DEBUG_TRIG)
+	      && hmer_debug_trig_function != DTRIG_UNKNOWN))
+		return -1;
+		
+	hmer &= ~HMER_DEBUG_TRIG;
+	/* HMER is a write-AND register */
+	mtspr(SPRN_HMER, ~HMER_DEBUG_TRIG);
+
+	switch (hmer_debug_trig_function) {
+	case DTRIG_VECTOR_CI:
+		/*
+		 * Now to avoid problems with soft-disable we
+		 * only do the emulation if we are coming from
+		 * host user space
+		 */
+		if (regs && user_mode(regs))
+			ret = local_paca->hmi_p9_special_emu = 1;
+
+		break;
+
+	default:
+		break;
+	}
+
+	/*
+	 * See if any other HMI causes remain to be handled
+	 */
+	if (hmer & mfspr(SPRN_HMEER))
+		return -1;
+
+	return ret;
+}
+
+/*
+ * Return values:
+ */
+long hmi_exception_realmode(struct pt_regs *regs)
+{	
+	int ret;
+
+	__this_cpu_inc(irq_stat.hmi_exceptions);
+
+	ret = hmi_handle_debugtrig(regs);
+	if (ret >= 0)
+		return ret;
 
 	wait_for_subcore_guest_exit();
 
diff --git a/arch/powerpc/kvm/book3s_hv_ras.c b/arch/powerpc/kvm/book3s_hv_ras.c
index e61066bb6725..b11043b23c18 100644
--- a/arch/powerpc/kvm/book3s_hv_ras.c
+++ b/arch/powerpc/kvm/book3s_hv_ras.c
@@ -266,17 +266,19 @@ static void kvmppc_tb_resync_done(void)
  *   secondary threads to proceed.
  * - All secondary threads will eventually call opal hmi handler on
  *   their exit path.
+ *
+ * Returns 1 if the timebase offset should be applied, 0 if not.
  */
 
 long kvmppc_realmode_hmi_handler(void)
 {
-	int ptid = local_paca->kvm_hstate.ptid;
 	bool resync_req;
 
-	/* This is only called on primary thread. */
-	BUG_ON(ptid != 0);
 	__this_cpu_inc(irq_stat.hmi_exceptions);
 
+	if (hmi_handle_debugtrig(NULL) >= 0)
+		return 1;
+
 	/*
 	 * By now primary thread has already completed guest->host
 	 * partition switch but haven't signaled secondaries yet.
diff --git a/arch/powerpc/kvm/book3s_hv_rm_mmu.c b/arch/powerpc/kvm/book3s_hv_rm_mmu.c
index 26c11f678fbf..8888e625a999 100644
--- a/arch/powerpc/kvm/book3s_hv_rm_mmu.c
+++ b/arch/powerpc/kvm/book3s_hv_rm_mmu.c
@@ -42,7 +42,7 @@ static void *real_vmalloc_addr(void *x)
 }
 
 /* Return 1 if we need to do a global tlbie, 0 if we can use tlbiel */
-static int global_invalidates(struct kvm *kvm, unsigned long flags)
+static int global_invalidates(struct kvm *kvm)
 {
 	int global;
 	int cpu;
@@ -522,7 +522,7 @@ long kvmppc_do_h_remove(struct kvm *kvm, unsigned long flags,
 	if (v & HPTE_V_VALID) {
 		hpte[0] &= ~cpu_to_be64(HPTE_V_VALID);
 		rb = compute_tlbie_rb(v, pte_r, pte_index);
-		do_tlbies(kvm, &rb, 1, global_invalidates(kvm, flags), true);
+		do_tlbies(kvm, &rb, 1, global_invalidates(kvm), true);
 		/*
 		 * The reference (R) and change (C) bits in a HPT
 		 * entry can be set by hardware at any time up until
@@ -572,7 +572,7 @@ long kvmppc_h_bulk_remove(struct kvm_vcpu *vcpu)
 
 	if (kvm_is_radix(kvm))
 		return H_FUNCTION;
-	global = global_invalidates(kvm, 0);
+	global = global_invalidates(kvm);
 	for (i = 0; i < 4 && ret == H_SUCCESS; ) {
 		n = 0;
 		for (; i < 4; ++i) {
@@ -732,8 +732,7 @@ long kvmppc_h_protect(struct kvm_vcpu *vcpu, unsigned long flags,
 			rb = compute_tlbie_rb(v, r, pte_index);
 			hpte[0] = cpu_to_be64((pte_v & ~HPTE_V_VALID) |
 					      HPTE_V_ABSENT);
-			do_tlbies(kvm, &rb, 1, global_invalidates(kvm, flags),
-				  true);
+			do_tlbies(kvm, &rb, 1, global_invalidates(kvm), true);
 			/* Don't lose R/C bit updates done by hardware */
 			r |= be64_to_cpu(hpte[1]) & (HPTE_R_R | HPTE_R_C);
 			hpte[1] = cpu_to_be64(r);
diff --git a/arch/powerpc/kvm/book3s_hv_rmhandlers.S b/arch/powerpc/kvm/book3s_hv_rmhandlers.S
index a92ad8500917..b484c4750677 100644
--- a/arch/powerpc/kvm/book3s_hv_rmhandlers.S
+++ b/arch/powerpc/kvm/book3s_hv_rmhandlers.S
@@ -1909,16 +1909,17 @@ END_FTR_SECTION_IFSET(CPU_FTR_ARCH_207S)
 	bne	27f
 	bl	kvmppc_realmode_hmi_handler
 	nop
+	cmpdi	r3, 0
 	li	r12, BOOK3S_INTERRUPT_HMI
 	/*
-	 * At this point kvmppc_realmode_hmi_handler would have resync-ed
-	 * the TB. Hence it is not required to subtract guest timebase
-	 * offset from timebase. So, skip it.
+	 * At this point kvmppc_realmode_hmi_handler may have resync-ed
+	 * the TB, and if it has, we must not subtract the guest timebase
+	 * offset from the timebase. So, skip it.
 	 *
 	 * Also, do not call kvmppc_subcore_exit_guest() because it has
 	 * been invoked as part of kvmppc_realmode_hmi_handler().
 	 */
-	b	30f
+	beq	30f
 
 27:
 	/* Subtract timebase offset from timebase */
diff --git a/arch/powerpc/sysdev/xive/common.c b/arch/powerpc/sysdev/xive/common.c
index 838ebdbfe4c5..40c06110821c 100644
--- a/arch/powerpc/sysdev/xive/common.c
+++ b/arch/powerpc/sysdev/xive/common.c
@@ -367,7 +367,8 @@ static void xive_irq_eoi(struct irq_data *d)
 	 * EOI the source if it hasn't been disabled and hasn't
 	 * been passed-through to a KVM guest
 	 */
-	if (!irqd_irq_disabled(d) && !irqd_is_forwarded_to_vcpu(d))
+	if (!irqd_irq_disabled(d) && !irqd_is_forwarded_to_vcpu(d) &&
+	    !(xd->flags & XIVE_IRQ_NO_EOI))
 		xive_do_source_eoi(irqd_to_hwirq(d), xd);
 
 	/*