diff --git a/Documentation/admin-guide/kernel-parameters.txt b/Documentation/admin-guide/kernel-parameters.txt
index 243444711129..1bd225eca34c 100644
--- a/Documentation/admin-guide/kernel-parameters.txt
+++ b/Documentation/admin-guide/kernel-parameters.txt
@@ -7420,17 +7420,18 @@
 			Crash from Xen panic notifier, without executing late
 			panic() code such as dumping handler.
 
+	xen_mc_debug	[X86,XEN,EARLY]
+			Enable multicall debugging when running as a Xen PV guest.
+			Enabling this feature will reduce performance a little
+			bit, so it should only be enabled for obtaining extended
+			debug data in case of multicall errors.
+
 	xen_msr_safe=	[X86,XEN,EARLY]
 			Format: <bool>
 			Select whether to always use non-faulting (safe) MSR
 			access functions when running as Xen PV guest. The
 			default value is controlled by CONFIG_XEN_PV_MSR_SAFE.
 
-	xen_nopvspin	[X86,XEN,EARLY]
-			Disables the qspinlock slowpath using Xen PV optimizations.
-			This parameter is obsoleted by "nopvspin" parameter, which
-			has equivalent effect for XEN platform.
-
 	xen_nopv	[X86]
 			Disables the PV optimizations forcing the HVM guest to
 			run as generic HVM guest with no PV drivers.
diff --git a/arch/arm/xen/p2m.c b/arch/arm/xen/p2m.c
index 309648c17f48..9da57a5b81c7 100644
--- a/arch/arm/xen/p2m.c
+++ b/arch/arm/xen/p2m.c
@@ -109,7 +109,7 @@ int set_foreign_p2m_mapping(struct gnttab_map_grant_ref *map_ops,
 		 * immediate unmapping.
 		 */
 		map_ops[i].status = GNTST_general_error;
-		unmap.host_addr = map_ops[i].host_addr,
+		unmap.host_addr = map_ops[i].host_addr;
 		unmap.handle = map_ops[i].handle;
 		map_ops[i].handle = INVALID_GRANT_HANDLE;
 		if (map_ops[i].flags & GNTMAP_device_map)
diff --git a/arch/x86/xen/apic.c b/arch/x86/xen/apic.c
index 8b045dd25196..bb0f3f368446 100644
--- a/arch/x86/xen/apic.c
+++ b/arch/x86/xen/apic.c
@@ -10,8 +10,6 @@
 #include <xen/xen.h>
 #include <xen/interface/physdev.h>
 #include "xen-ops.h"
-#include "pmu.h"
-#include "smp.h"
 
 static unsigned int xen_io_apic_read(unsigned apic, unsigned reg)
 {
diff --git a/arch/x86/xen/debugfs.c b/arch/x86/xen/debugfs.c
index 532410998684..b8c9f2a7d9b6 100644
--- a/arch/x86/xen/debugfs.c
+++ b/arch/x86/xen/debugfs.c
@@ -3,7 +3,7 @@
 #include <linux/debugfs.h>
 #include <linux/slab.h>
 
-#include "debugfs.h"
+#include "xen-ops.h"
 
 static struct dentry *d_xen_debug;
 
diff --git a/arch/x86/xen/debugfs.h b/arch/x86/xen/debugfs.h
deleted file mode 100644
index 6b813ad1091c..000000000000
--- a/arch/x86/xen/debugfs.h
+++ /dev/null
@@ -1,7 +0,0 @@
-/* SPDX-License-Identifier: GPL-2.0 */
-#ifndef _XEN_DEBUGFS_H
-#define _XEN_DEBUGFS_H
-
-struct dentry * __init xen_init_debugfs(void);
-
-#endif /* _XEN_DEBUGFS_H */
diff --git a/arch/x86/xen/enlighten.c b/arch/x86/xen/enlighten.c
index 0305485edcd3..84e5adbd0925 100644
--- a/arch/x86/xen/enlighten.c
+++ b/arch/x86/xen/enlighten.c
@@ -20,8 +20,6 @@
 #include <asm/setup.h>
 
 #include "xen-ops.h"
-#include "smp.h"
-#include "pmu.h"
 
 EXPORT_SYMBOL_GPL(hypercall_page);
 
diff --git a/arch/x86/xen/enlighten_hvm.c b/arch/x86/xen/enlighten_hvm.c
index c001a2296582..24d2957a4726 100644
--- a/arch/x86/xen/enlighten_hvm.c
+++ b/arch/x86/xen/enlighten_hvm.c
@@ -28,8 +28,6 @@
 #include <asm/xen/page.h>
 
 #include "xen-ops.h"
-#include "mmu.h"
-#include "smp.h"
 
 static unsigned long shared_info_pfn;
 
diff --git a/arch/x86/xen/enlighten_pv.c b/arch/x86/xen/enlighten_pv.c
index 9ba53814ed6a..2c12ae42dc8b 100644
--- a/arch/x86/xen/enlighten_pv.c
+++ b/arch/x86/xen/enlighten_pv.c
@@ -85,10 +85,6 @@
 #endif
 
 #include "xen-ops.h"
-#include "mmu.h"
-#include "smp.h"
-#include "multicalls.h"
-#include "pmu.h"
 
 #include "../kernel/cpu/cpu.h" /* get_cpu_cap() */
 
diff --git a/arch/x86/xen/mmu.c b/arch/x86/xen/mmu.c
index 60e9c37fd79f..c4c479373249 100644
--- a/arch/x86/xen/mmu.c
+++ b/arch/x86/xen/mmu.c
@@ -5,8 +5,7 @@
 #include <asm/xen/hypercall.h>
 #include <xen/interface/memory.h>
 
-#include "multicalls.h"
-#include "mmu.h"
+#include "xen-ops.h"
 
 unsigned long arbitrary_virt_to_mfn(void *vaddr)
 {
diff --git a/arch/x86/xen/mmu.h b/arch/x86/xen/mmu.h
deleted file mode 100644
index 6e4c6bd62203..000000000000
--- a/arch/x86/xen/mmu.h
+++ /dev/null
@@ -1,28 +0,0 @@
-/* SPDX-License-Identifier: GPL-2.0 */
-#ifndef _XEN_MMU_H
-
-#include <linux/linkage.h>
-#include <asm/page.h>
-
-enum pt_level {
-	PT_PGD,
-	PT_P4D,
-	PT_PUD,
-	PT_PMD,
-	PT_PTE
-};
-
-
-bool __set_phys_to_machine(unsigned long pfn, unsigned long mfn);
-
-void set_pte_mfn(unsigned long vaddr, unsigned long pfn, pgprot_t flags);
-
-pte_t xen_ptep_modify_prot_start(struct vm_area_struct *vma, unsigned long addr, pte_t *ptep);
-void  xen_ptep_modify_prot_commit(struct vm_area_struct *vma, unsigned long addr,
-				  pte_t *ptep, pte_t pte);
-
-unsigned long xen_read_cr2_direct(void);
-
-extern void xen_init_mmu_ops(void);
-extern void xen_hvm_init_mmu_ops(void);
-#endif	/* _XEN_MMU_H */
diff --git a/arch/x86/xen/mmu_hvm.c b/arch/x86/xen/mmu_hvm.c
index 509bdee3ab90..337955652202 100644
--- a/arch/x86/xen/mmu_hvm.c
+++ b/arch/x86/xen/mmu_hvm.c
@@ -5,7 +5,7 @@
 #include <xen/interface/xen.h>
 #include <xen/hvm.h>
 
-#include "mmu.h"
+#include "xen-ops.h"
 
 #ifdef CONFIG_PROC_VMCORE
 /*
diff --git a/arch/x86/xen/mmu_pv.c b/arch/x86/xen/mmu_pv.c
index 54e0d311dcc9..f1ce39d6d32c 100644
--- a/arch/x86/xen/mmu_pv.c
+++ b/arch/x86/xen/mmu_pv.c
@@ -82,9 +82,7 @@
 #include <xen/hvc-console.h>
 #include <xen/swiotlb-xen.h>
 
-#include "multicalls.h"
-#include "mmu.h"
-#include "debugfs.h"
+#include "xen-ops.h"
 
 /*
  * Prototypes for functions called via PV_CALLEE_SAVE_REGS_THUNK() in order
@@ -128,7 +126,7 @@ static DEFINE_SPINLOCK(xen_reservation_lock);
  * looking at another vcpu's cr3 value, it should use this variable.
  */
 DEFINE_PER_CPU(unsigned long, xen_cr3);	 /* cr3 stored as physaddr */
-DEFINE_PER_CPU(unsigned long, xen_current_cr3);	 /* actual vcpu cr3 */
+static DEFINE_PER_CPU(unsigned long, xen_current_cr3);	/* actual vcpu cr3 */
 
 static phys_addr_t xen_pt_base, xen_pt_size __initdata;
 
@@ -305,16 +303,17 @@ static void xen_set_pte(pte_t *ptep, pte_t pteval)
 	__xen_set_pte(ptep, pteval);
 }
 
-pte_t xen_ptep_modify_prot_start(struct vm_area_struct *vma,
-				 unsigned long addr, pte_t *ptep)
+static pte_t xen_ptep_modify_prot_start(struct vm_area_struct *vma,
+					unsigned long addr, pte_t *ptep)
 {
 	/* Just return the pte as-is.  We preserve the bits on commit */
 	trace_xen_mmu_ptep_modify_prot_start(vma->vm_mm, addr, ptep, *ptep);
 	return *ptep;
 }
 
-void xen_ptep_modify_prot_commit(struct vm_area_struct *vma, unsigned long addr,
-				 pte_t *ptep, pte_t pte)
+static void xen_ptep_modify_prot_commit(struct vm_area_struct *vma,
+					unsigned long addr,
+					pte_t *ptep, pte_t pte)
 {
 	struct mmu_update u;
 
diff --git a/arch/x86/xen/multicalls.c b/arch/x86/xen/multicalls.c
index 07054572297f..d4cefd8a9af4 100644
--- a/arch/x86/xen/multicalls.c
+++ b/arch/x86/xen/multicalls.c
@@ -23,26 +23,21 @@
 #include <linux/percpu.h>
 #include <linux/hardirq.h>
 #include <linux/debugfs.h>
+#include <linux/jump_label.h>
+#include <linux/printk.h>
 
 #include <asm/xen/hypercall.h>
 
-#include "multicalls.h"
-#include "debugfs.h"
+#include "xen-ops.h"
 
 #define MC_BATCH	32
 
-#define MC_DEBUG	0
-
 #define MC_ARGS		(MC_BATCH * 16)
 
 
 struct mc_buffer {
 	unsigned mcidx, argidx, cbidx;
 	struct multicall_entry entries[MC_BATCH];
-#if MC_DEBUG
-	struct multicall_entry debug[MC_BATCH];
-	void *caller[MC_BATCH];
-#endif
 	unsigned char args[MC_ARGS];
 	struct callback {
 		void (*fn)(void *);
@@ -50,13 +45,98 @@ struct mc_buffer {
 	} callbacks[MC_BATCH];
 };
 
+struct mc_debug_data {
+	struct multicall_entry entries[MC_BATCH];
+	void *caller[MC_BATCH];
+	size_t argsz[MC_BATCH];
+	unsigned long *args[MC_BATCH];
+};
+
 static DEFINE_PER_CPU(struct mc_buffer, mc_buffer);
+static struct mc_debug_data mc_debug_data_early __initdata;
+static struct mc_debug_data __percpu *mc_debug_data __refdata =
+	&mc_debug_data_early;
 DEFINE_PER_CPU(unsigned long, xen_mc_irq_flags);
 
+static struct static_key mc_debug __ro_after_init;
+static bool mc_debug_enabled __initdata;
+
+static int __init xen_parse_mc_debug(char *arg)
+{
+	mc_debug_enabled = true;
+	static_key_slow_inc(&mc_debug);
+
+	return 0;
+}
+early_param("xen_mc_debug", xen_parse_mc_debug);
+
+static int __init mc_debug_enable(void)
+{
+	struct mc_debug_data __percpu *mcdb;
+	unsigned long flags;
+
+	if (!mc_debug_enabled)
+		return 0;
+
+	mcdb = alloc_percpu(struct mc_debug_data);
+	if (!mcdb) {
+		pr_err("xen_mc_debug inactive\n");
+		static_key_slow_dec(&mc_debug);
+		return -ENOMEM;
+	}
+
+	/* Be careful when switching to percpu debug data. */
+	local_irq_save(flags);
+	xen_mc_flush();
+	mc_debug_data = mcdb;
+	local_irq_restore(flags);
+
+	pr_info("xen_mc_debug active\n");
+
+	return 0;
+}
+early_initcall(mc_debug_enable);
+
+/* Number of parameters of hypercalls used via multicalls. */
+static const uint8_t hpcpars[] = {
+	[__HYPERVISOR_mmu_update] = 4,
+	[__HYPERVISOR_stack_switch] = 2,
+	[__HYPERVISOR_fpu_taskswitch] = 1,
+	[__HYPERVISOR_update_descriptor] = 2,
+	[__HYPERVISOR_update_va_mapping] = 3,
+	[__HYPERVISOR_mmuext_op] = 4,
+};
+
+static void print_debug_data(struct mc_buffer *b, struct mc_debug_data *mcdb,
+			     int idx)
+{
+	unsigned int arg;
+	unsigned int opidx = mcdb->entries[idx].op & 0xff;
+	unsigned int pars = 0;
+
+	pr_err("  call %2d: op=%lu result=%ld caller=%pS ", idx + 1,
+	       mcdb->entries[idx].op, b->entries[idx].result,
+	       mcdb->caller[idx]);
+	if (opidx < ARRAY_SIZE(hpcpars))
+		pars = hpcpars[opidx];
+	if (pars) {
+		pr_cont("pars=");
+		for (arg = 0; arg < pars; arg++)
+			pr_cont("%lx ", mcdb->entries[idx].args[arg]);
+	}
+	if (mcdb->argsz[idx]) {
+		pr_cont("args=");
+		for (arg = 0; arg < mcdb->argsz[idx] / 8; arg++)
+			pr_cont("%lx ", mcdb->args[idx][arg]);
+	}
+	pr_cont("\n");
+}
+
 void xen_mc_flush(void)
 {
 	struct mc_buffer *b = this_cpu_ptr(&mc_buffer);
 	struct multicall_entry *mc;
+	struct mc_debug_data *mcdb = NULL;
 	int ret = 0;
 	unsigned long flags;
 	int i;
@@ -69,10 +149,11 @@ void xen_mc_flush(void)
 
 	trace_xen_mc_flush(b->mcidx, b->argidx, b->cbidx);
 
-#if MC_DEBUG
-	memcpy(b->debug, b->entries,
-	       b->mcidx * sizeof(struct multicall_entry));
-#endif
+	if (static_key_false(&mc_debug)) {
+		mcdb = this_cpu_ptr(mc_debug_data);
+		memcpy(mcdb->entries, b->entries,
+		       b->mcidx * sizeof(struct multicall_entry));
+	}
 
 	switch (b->mcidx) {
 	case 0:
@@ -103,21 +184,14 @@ void xen_mc_flush(void)
 		pr_err("%d of %d multicall(s) failed: cpu %d\n",
 		       ret, b->mcidx, smp_processor_id());
 		for (i = 0; i < b->mcidx; i++) {
-			if (b->entries[i].result < 0) {
-#if MC_DEBUG
-				pr_err("  call %2d: op=%lu arg=[%lx] result=%ld\t%pS\n",
-				       i + 1,
-				       b->debug[i].op,
-				       b->debug[i].args[0],
-				       b->entries[i].result,
-				       b->caller[i]);
-#else
+			if (static_key_false(&mc_debug)) {
+				print_debug_data(b, mcdb, i);
+			} else if (b->entries[i].result < 0) {
 				pr_err("  call %2d: op=%lu arg=[%lx] result=%ld\n",
 				       i + 1,
 				       b->entries[i].op,
 				       b->entries[i].args[0],
 				       b->entries[i].result);
-#endif
 			}
 		}
 	}
@@ -155,9 +229,13 @@ struct multicall_space __xen_mc_entry(size_t args)
 	}
 
 	ret.mc = &b->entries[b->mcidx];
-#if MC_DEBUG
-	b->caller[b->mcidx] = __builtin_return_address(0);
-#endif
+	if (static_key_false(&mc_debug)) {
+		struct mc_debug_data *mcdb = this_cpu_ptr(mc_debug_data);
+
+		mcdb->caller[b->mcidx] = __builtin_return_address(0);
+		mcdb->argsz[b->mcidx] = args;
+		mcdb->args[b->mcidx] = (unsigned long *)(&b->args[argidx]);
+	}
 	b->mcidx++;
 	ret.args = &b->args[argidx];
 	b->argidx = argidx + args;
diff --git a/arch/x86/xen/multicalls.h b/arch/x86/xen/multicalls.h
deleted file mode 100644
index c3867b585e0d..000000000000
--- a/arch/x86/xen/multicalls.h
+++ /dev/null
@@ -1,69 +0,0 @@
-/* SPDX-License-Identifier: GPL-2.0 */
-#ifndef _XEN_MULTICALLS_H
-#define _XEN_MULTICALLS_H
-
-#include <trace/events/xen.h>
-
-#include "xen-ops.h"
-
-/* Multicalls */
-struct multicall_space
-{
-	struct multicall_entry *mc;
-	void *args;
-};
-
-/* Allocate room for a multicall and its args */
-struct multicall_space __xen_mc_entry(size_t args);
-
-DECLARE_PER_CPU(unsigned long, xen_mc_irq_flags);
-
-/* Call to start a batch of multiple __xen_mc_entry()s.  Must be
-   paired with xen_mc_issue() */
-static inline void xen_mc_batch(void)
-{
-	unsigned long flags;
-
-	/* need to disable interrupts until this entry is complete */
-	local_irq_save(flags);
-	trace_xen_mc_batch(xen_get_lazy_mode());
-	__this_cpu_write(xen_mc_irq_flags, flags);
-}
-
-static inline struct multicall_space xen_mc_entry(size_t args)
-{
-	xen_mc_batch();
-	return __xen_mc_entry(args);
-}
-
-/* Flush all pending multicalls */
-void xen_mc_flush(void);
-
-/* Issue a multicall if we're not in a lazy mode */
-static inline void xen_mc_issue(unsigned mode)
-{
-	trace_xen_mc_issue(mode);
-
-	if ((xen_get_lazy_mode() & mode) == 0)
-		xen_mc_flush();
-
-	/* restore flags saved in xen_mc_batch */
-	local_irq_restore(this_cpu_read(xen_mc_irq_flags));
-}
-
-/* Set up a callback to be called when the current batch is flushed */
-void xen_mc_callback(void (*fn)(void *), void *data);
-
-/*
- * Try to extend the arguments of the previous multicall command.  The
- * previous command's op must match.  If it does, then it attempts to
- * extend the argument space allocated to the multicall entry by
- * arg_size bytes.
- *
- * The returned multicall_space will return with mc pointing to the
- * command on success, or NULL on failure, and args pointing to the
- * newly allocated space.
- */
-struct multicall_space xen_mc_extend_args(unsigned long op, size_t arg_size);
-
-#endif /* _XEN_MULTICALLS_H */
diff --git a/arch/x86/xen/p2m.c b/arch/x86/xen/p2m.c
index 99918beccd80..7c735b730acd 100644
--- a/arch/x86/xen/p2m.c
+++ b/arch/x86/xen/p2m.c
@@ -81,7 +81,6 @@
 #include <xen/balloon.h>
 #include <xen/grant_table.h>
 
-#include "multicalls.h"
 #include "xen-ops.h"
 
 #define P2M_MID_PER_PAGE	(PAGE_SIZE / sizeof(unsigned long *))
@@ -730,7 +729,7 @@ int set_foreign_p2m_mapping(struct gnttab_map_grant_ref *map_ops,
 		 * immediate unmapping.
 		 */
 		map_ops[i].status = GNTST_general_error;
-		unmap[0].host_addr = map_ops[i].host_addr,
+		unmap[0].host_addr = map_ops[i].host_addr;
 		unmap[0].handle = map_ops[i].handle;
 		map_ops[i].handle = INVALID_GRANT_HANDLE;
 		if (map_ops[i].flags & GNTMAP_device_map)
@@ -740,7 +739,7 @@ int set_foreign_p2m_mapping(struct gnttab_map_grant_ref *map_ops,
 
 		if (kmap_ops) {
 			kmap_ops[i].status = GNTST_general_error;
-			unmap[1].host_addr = kmap_ops[i].host_addr,
+			unmap[1].host_addr = kmap_ops[i].host_addr;
 			unmap[1].handle = kmap_ops[i].handle;
 			kmap_ops[i].handle = INVALID_GRANT_HANDLE;
 			if (kmap_ops[i].flags & GNTMAP_device_map)
@@ -795,7 +794,6 @@ int clear_foreign_p2m_mapping(struct gnttab_unmap_grant_ref *unmap_ops,
 
 #ifdef CONFIG_XEN_DEBUG_FS
 #include <linux/debugfs.h>
-#include "debugfs.h"
 static int p2m_dump_show(struct seq_file *m, void *v)
 {
 	static const char * const type_name[] = {
diff --git a/arch/x86/xen/pmu.c b/arch/x86/xen/pmu.c
index 246d67dab510..f06987b0efc3 100644
--- a/arch/x86/xen/pmu.c
+++ b/arch/x86/xen/pmu.c
@@ -10,7 +10,6 @@
 #include <xen/interface/xenpmu.h>
 
 #include "xen-ops.h"
-#include "pmu.h"
 
 /* x86_pmu.handle_irq definition */
 #include "../events/perf_event.h"
diff --git a/arch/x86/xen/pmu.h b/arch/x86/xen/pmu.h
deleted file mode 100644
index 65c58894fc79..000000000000
--- a/arch/x86/xen/pmu.h
+++ /dev/null
@@ -1,22 +0,0 @@
-/* SPDX-License-Identifier: GPL-2.0 */
-#ifndef __XEN_PMU_H
-#define __XEN_PMU_H
-
-#include <xen/interface/xenpmu.h>
-
-extern bool is_xen_pmu;
-
-irqreturn_t xen_pmu_irq_handler(int irq, void *dev_id);
-#ifdef CONFIG_XEN_HAVE_VPMU
-void xen_pmu_init(int cpu);
-void xen_pmu_finish(int cpu);
-#else
-static inline void xen_pmu_init(int cpu) {}
-static inline void xen_pmu_finish(int cpu) {}
-#endif
-bool pmu_msr_read(unsigned int msr, uint64_t *val, int *err);
-bool pmu_msr_write(unsigned int msr, uint32_t low, uint32_t high, int *err);
-int pmu_apic_update(uint32_t reg);
-unsigned long long xen_read_pmc(int counter);
-
-#endif /* __XEN_PMU_H */
diff --git a/arch/x86/xen/setup.c b/arch/x86/xen/setup.c
index 380591028cb8..a0c3e77e3d5b 100644
--- a/arch/x86/xen/setup.c
+++ b/arch/x86/xen/setup.c
@@ -34,7 +34,6 @@
 #include <xen/features.h>
 #include <xen/hvc-console.h>
 #include "xen-ops.h"
-#include "mmu.h"
 
 #define GB(x) ((uint64_t)(x) * 1024 * 1024 * 1024)
 
diff --git a/arch/x86/xen/smp.c b/arch/x86/xen/smp.c
index 935771726f9c..05f92c812ac8 100644
--- a/arch/x86/xen/smp.c
+++ b/arch/x86/xen/smp.c
@@ -9,7 +9,6 @@
 
 #include <xen/hvc-console.h>
 #include "xen-ops.h"
-#include "smp.h"
 
 static DEFINE_PER_CPU(struct xen_common_irq, xen_resched_irq) = { .irq = -1 };
 static DEFINE_PER_CPU(struct xen_common_irq, xen_callfunc_irq) = { .irq = -1 };
diff --git a/arch/x86/xen/smp.h b/arch/x86/xen/smp.h
deleted file mode 100644
index b8efdbc693f7..000000000000
--- a/arch/x86/xen/smp.h
+++ /dev/null
@@ -1,51 +0,0 @@
-/* SPDX-License-Identifier: GPL-2.0 */
-#ifndef _XEN_SMP_H
-
-#ifdef CONFIG_SMP
-
-void asm_cpu_bringup_and_idle(void);
-asmlinkage void cpu_bringup_and_idle(void);
-
-extern void xen_send_IPI_mask(const struct cpumask *mask,
-			      int vector);
-extern void xen_send_IPI_mask_allbutself(const struct cpumask *mask,
-				int vector);
-extern void xen_send_IPI_allbutself(int vector);
-extern void xen_send_IPI_all(int vector);
-extern void xen_send_IPI_self(int vector);
-
-extern int xen_smp_intr_init(unsigned int cpu);
-extern void xen_smp_intr_free(unsigned int cpu);
-int xen_smp_intr_init_pv(unsigned int cpu);
-void xen_smp_intr_free_pv(unsigned int cpu);
-
-void xen_smp_count_cpus(void);
-void xen_smp_cpus_done(unsigned int max_cpus);
-
-void xen_smp_send_reschedule(int cpu);
-void xen_smp_send_call_function_ipi(const struct cpumask *mask);
-void xen_smp_send_call_function_single_ipi(int cpu);
-
-void __noreturn xen_cpu_bringup_again(unsigned long stack);
-
-struct xen_common_irq {
-	int irq;
-	char *name;
-};
-#else /* CONFIG_SMP */
-
-static inline int xen_smp_intr_init(unsigned int cpu)
-{
-	return 0;
-}
-static inline void xen_smp_intr_free(unsigned int cpu) {}
-
-static inline int xen_smp_intr_init_pv(unsigned int cpu)
-{
-	return 0;
-}
-static inline void xen_smp_intr_free_pv(unsigned int cpu) {}
-static inline void xen_smp_count_cpus(void) { }
-#endif /* CONFIG_SMP */
-
-#endif
diff --git a/arch/x86/xen/smp_hvm.c b/arch/x86/xen/smp_hvm.c
index ac95d1981cc0..485c1d8804f7 100644
--- a/arch/x86/xen/smp_hvm.c
+++ b/arch/x86/xen/smp_hvm.c
@@ -5,8 +5,6 @@
 #include <xen/events.h>
 
 #include "xen-ops.h"
-#include "smp.h"
-
 
 static void __init xen_hvm_smp_prepare_boot_cpu(void)
 {
diff --git a/arch/x86/xen/smp_pv.c b/arch/x86/xen/smp_pv.c
index ac41d83b38d3..7ea57f728b89 100644
--- a/arch/x86/xen/smp_pv.c
+++ b/arch/x86/xen/smp_pv.c
@@ -46,9 +46,6 @@
 
 #include <xen/hvc-console.h>
 #include "xen-ops.h"
-#include "mmu.h"
-#include "smp.h"
-#include "pmu.h"
 
 cpumask_var_t xen_cpu_initialized_map;
 
diff --git a/arch/x86/xen/spinlock.c b/arch/x86/xen/spinlock.c
index 5c6fc16e4b92..8e4efe0fb6f9 100644
--- a/arch/x86/xen/spinlock.c
+++ b/arch/x86/xen/spinlock.c
@@ -18,7 +18,6 @@
 static DEFINE_PER_CPU(int, lock_kicker_irq) = -1;
 static DEFINE_PER_CPU(char *, irq_name);
 static DEFINE_PER_CPU(atomic_t, xen_qlock_wait_nest);
-static bool xen_pvspin = true;
 
 static void xen_qlock_kick(int cpu)
 {
@@ -68,7 +67,7 @@ void xen_init_lock_cpu(int cpu)
 	int irq;
 	char *name;
 
-	if (!xen_pvspin)
+	if (nopvspin)
 		return;
 
 	WARN(per_cpu(lock_kicker_irq, cpu) >= 0, "spinlock on CPU%d exists on IRQ%d!\n",
@@ -95,7 +94,7 @@ void xen_uninit_lock_cpu(int cpu)
 {
 	int irq;
 
-	if (!xen_pvspin)
+	if (nopvspin)
 		return;
 
 	kfree(per_cpu(irq_name, cpu));
@@ -125,10 +124,10 @@ PV_CALLEE_SAVE_REGS_THUNK(xen_vcpu_stolen);
 void __init xen_init_spinlocks(void)
 {
 	/*  Don't need to use pvqspinlock code if there is only 1 vCPU. */
-	if (num_possible_cpus() == 1 || nopvspin)
-		xen_pvspin = false;
+	if (num_possible_cpus() == 1)
+		nopvspin = true;
 
-	if (!xen_pvspin) {
+	if (nopvspin) {
 		printk(KERN_DEBUG "xen: PV spinlocks disabled\n");
 		static_branch_disable(&virt_spin_lock_key);
 		return;
@@ -143,12 +142,3 @@ void __init xen_init_spinlocks(void)
 	pv_ops.lock.kick = xen_qlock_kick;
 	pv_ops.lock.vcpu_is_preempted = PV_CALLEE_SAVE(xen_vcpu_stolen);
 }
-
-static __init int xen_parse_nopvspin(char *arg)
-{
-	pr_notice("\"xen_nopvspin\" is deprecated, please use \"nopvspin\" instead\n");
-	xen_pvspin = false;
-	return 0;
-}
-early_param("xen_nopvspin", xen_parse_nopvspin);
-
diff --git a/arch/x86/xen/suspend.c b/arch/x86/xen/suspend.c
index 1d83152c761b..77a6ea1c60e4 100644
--- a/arch/x86/xen/suspend.c
+++ b/arch/x86/xen/suspend.c
@@ -15,8 +15,6 @@
 #include <asm/fixmap.h>
 
 #include "xen-ops.h"
-#include "mmu.h"
-#include "pmu.h"
 
 static DEFINE_PER_CPU(u64, spec_ctrl);
 
diff --git a/arch/x86/xen/time.c b/arch/x86/xen/time.c
index 52fa5609b7f6..96521b1874ac 100644
--- a/arch/x86/xen/time.c
+++ b/arch/x86/xen/time.c
@@ -30,7 +30,7 @@
 #include "xen-ops.h"
 
 /* Minimum amount of time until next clock event fires */
-#define TIMER_SLOP	100000
+#define TIMER_SLOP	1
 
 static u64 xen_sched_clock_offset __read_mostly;
 
diff --git a/arch/x86/xen/xen-ops.h b/arch/x86/xen/xen-ops.h
index 79cf93f2c92f..e7775dff9452 100644
--- a/arch/x86/xen/xen-ops.h
+++ b/arch/x86/xen/xen-ops.h
@@ -5,8 +5,15 @@
 #include <linux/init.h>
 #include <linux/clocksource.h>
 #include <linux/irqreturn.h>
+#include <linux/linkage.h>
+
+#include <xen/interface/xenpmu.h>
 #include <xen/xen-ops.h>
 
+#include <asm/page.h>
+
+#include <trace/events/xen.h>
+
 /* These are code, but not functions.  Defined in entry.S */
 extern const char xen_failsafe_callback[];
 
@@ -23,14 +30,11 @@ void xen_copy_trap_info(struct trap_info *traps);
 
 DECLARE_PER_CPU_ALIGNED(struct vcpu_info, xen_vcpu_info);
 DECLARE_PER_CPU(unsigned long, xen_cr3);
-DECLARE_PER_CPU(unsigned long, xen_current_cr3);
 
 extern struct start_info *xen_start_info;
 extern struct shared_info xen_dummy_shared_info;
 extern struct shared_info *HYPERVISOR_shared_info;
 
-extern bool xen_fifo_events;
-
 void xen_setup_mfn_list_list(void);
 void xen_build_mfn_list_list(void);
 void xen_setup_machphys_mapping(void);
@@ -177,4 +181,142 @@ static inline void xen_hvm_post_suspend(int suspend_cancelled) {}
 
 void xen_add_extra_mem(unsigned long start_pfn, unsigned long n_pfns);
 
+struct dentry * __init xen_init_debugfs(void);
+
+enum pt_level {
+	PT_PGD,
+	PT_P4D,
+	PT_PUD,
+	PT_PMD,
+	PT_PTE
+};
+
+bool __set_phys_to_machine(unsigned long pfn, unsigned long mfn);
+void set_pte_mfn(unsigned long vaddr, unsigned long pfn, pgprot_t flags);
+unsigned long xen_read_cr2_direct(void);
+void xen_init_mmu_ops(void);
+void xen_hvm_init_mmu_ops(void);
+
+/* Multicalls */
+struct multicall_space
+{
+	struct multicall_entry *mc;
+	void *args;
+};
+
+/* Allocate room for a multicall and its args */
+struct multicall_space __xen_mc_entry(size_t args);
+
+DECLARE_PER_CPU(unsigned long, xen_mc_irq_flags);
+
+/* Call to start a batch of multiple __xen_mc_entry()s.  Must be
+   paired with xen_mc_issue() */
+static inline void xen_mc_batch(void)
+{
+	unsigned long flags;
+
+	/* need to disable interrupts until this entry is complete */
+	local_irq_save(flags);
+	trace_xen_mc_batch(xen_get_lazy_mode());
+	__this_cpu_write(xen_mc_irq_flags, flags);
+}
+
+static inline struct multicall_space xen_mc_entry(size_t args)
+{
+	xen_mc_batch();
+	return __xen_mc_entry(args);
+}
+
+/* Flush all pending multicalls */
+void xen_mc_flush(void);
+
+/* Issue a multicall if we're not in a lazy mode */
+static inline void xen_mc_issue(unsigned mode)
+{
+	trace_xen_mc_issue(mode);
+
+	if ((xen_get_lazy_mode() & mode) == 0)
+		xen_mc_flush();
+
+	/* restore flags saved in xen_mc_batch */
+	local_irq_restore(this_cpu_read(xen_mc_irq_flags));
+}
+
+/* Set up a callback to be called when the current batch is flushed */
+void xen_mc_callback(void (*fn)(void *), void *data);
+
+/*
+ * Try to extend the arguments of the previous multicall command.  The
+ * previous command's op must match.  If it does, then it attempts to
+ * extend the argument space allocated to the multicall entry by
+ * arg_size bytes.
+ *
+ * The returned multicall_space will return with mc pointing to the
+ * command on success, or NULL on failure, and args pointing to the
+ * newly allocated space.
+ */
+struct multicall_space xen_mc_extend_args(unsigned long op, size_t arg_size);
+
+extern bool is_xen_pmu;
+
+irqreturn_t xen_pmu_irq_handler(int irq, void *dev_id);
+#ifdef CONFIG_XEN_HAVE_VPMU
+void xen_pmu_init(int cpu);
+void xen_pmu_finish(int cpu);
+#else
+static inline void xen_pmu_init(int cpu) {}
+static inline void xen_pmu_finish(int cpu) {}
+#endif
+bool pmu_msr_read(unsigned int msr, uint64_t *val, int *err);
+bool pmu_msr_write(unsigned int msr, uint32_t low, uint32_t high, int *err);
+int pmu_apic_update(uint32_t reg);
+unsigned long long xen_read_pmc(int counter);
+
+#ifdef CONFIG_SMP
+
+void asm_cpu_bringup_and_idle(void);
+asmlinkage void cpu_bringup_and_idle(void);
+
+extern void xen_send_IPI_mask(const struct cpumask *mask,
+			      int vector);
+extern void xen_send_IPI_mask_allbutself(const struct cpumask *mask,
+				int vector);
+extern void xen_send_IPI_allbutself(int vector);
+extern void xen_send_IPI_all(int vector);
+extern void xen_send_IPI_self(int vector);
+
+extern int xen_smp_intr_init(unsigned int cpu);
+extern void xen_smp_intr_free(unsigned int cpu);
+int xen_smp_intr_init_pv(unsigned int cpu);
+void xen_smp_intr_free_pv(unsigned int cpu);
+
+void xen_smp_count_cpus(void);
+void xen_smp_cpus_done(unsigned int max_cpus);
+
+void xen_smp_send_reschedule(int cpu);
+void xen_smp_send_call_function_ipi(const struct cpumask *mask);
+void xen_smp_send_call_function_single_ipi(int cpu);
+
+void __noreturn xen_cpu_bringup_again(unsigned long stack);
+
+struct xen_common_irq {
+	int irq;
+	char *name;
+};
+#else /* CONFIG_SMP */
+
+static inline int xen_smp_intr_init(unsigned int cpu)
+{
+	return 0;
+}
+static inline void xen_smp_intr_free(unsigned int cpu) {}
+
+static inline int xen_smp_intr_init_pv(unsigned int cpu)
+{
+	return 0;
+}
+static inline void xen_smp_intr_free_pv(unsigned int cpu) {}
+static inline void xen_smp_count_cpus(void) { }
+#endif /* CONFIG_SMP */
+
 #endif /* XEN_OPS_H */
diff --git a/drivers/xen/evtchn.c b/drivers/xen/evtchn.c
index f6a2216c2c87..9b7fcc7dbb38 100644
--- a/drivers/xen/evtchn.c
+++ b/drivers/xen/evtchn.c
@@ -729,4 +729,5 @@ static void __exit evtchn_cleanup(void)
 module_init(evtchn_init);
 module_exit(evtchn_cleanup);
 
+MODULE_DESCRIPTION("Xen /dev/xen/evtchn device driver");
 MODULE_LICENSE("GPL");
diff --git a/drivers/xen/manage.c b/drivers/xen/manage.c
index c16df629907e..b4b4ebed68da 100644
--- a/drivers/xen/manage.c
+++ b/drivers/xen/manage.c
@@ -208,7 +208,7 @@ static void do_reboot(void)
 	orderly_reboot();
 }
 
-static struct shutdown_handler shutdown_handlers[] = {
+static const struct shutdown_handler shutdown_handlers[] = {
 	{ "poweroff",	true,	do_poweroff },
 	{ "halt",	false,	do_poweroff },
 	{ "reboot",	true,	do_reboot   },
diff --git a/drivers/xen/privcmd-buf.c b/drivers/xen/privcmd-buf.c
index 2fa10ca5be14..0f0dad427d7e 100644
--- a/drivers/xen/privcmd-buf.c
+++ b/drivers/xen/privcmd-buf.c
@@ -19,6 +19,7 @@
 
 #include "privcmd.h"
 
+MODULE_DESCRIPTION("Xen Mmap of hypercall buffers");
 MODULE_LICENSE("GPL");
 
 struct privcmd_buf_private {
diff --git a/drivers/xen/privcmd.c b/drivers/xen/privcmd.c
index 67dfa4778864..9563650dfbaf 100644
--- a/drivers/xen/privcmd.c
+++ b/drivers/xen/privcmd.c
@@ -17,6 +17,7 @@
 #include <linux/poll.h>
 #include <linux/sched.h>
 #include <linux/slab.h>
+#include <linux/srcu.h>
 #include <linux/string.h>
 #include <linux/workqueue.h>
 #include <linux/errno.h>
@@ -48,6 +49,7 @@
 
 #include "privcmd.h"
 
+MODULE_DESCRIPTION("Xen hypercall passthrough driver");
 MODULE_LICENSE("GPL");
 
 #define PRIV_VMA_LOCKED ((void *)1)
@@ -845,7 +847,8 @@ out:
 #ifdef CONFIG_XEN_PRIVCMD_EVENTFD
 /* Irqfd support */
 static struct workqueue_struct *irqfd_cleanup_wq;
-static DEFINE_MUTEX(irqfds_lock);
+static DEFINE_SPINLOCK(irqfds_lock);
+DEFINE_STATIC_SRCU(irqfds_srcu);
 static LIST_HEAD(irqfds_list);
 
 struct privcmd_kernel_irqfd {
@@ -873,6 +876,9 @@ static void irqfd_shutdown(struct work_struct *work)
 		container_of(work, struct privcmd_kernel_irqfd, shutdown);
 	u64 cnt;
 
+	/* Make sure irqfd has been initialized in assign path */
+	synchronize_srcu(&irqfds_srcu);
+
 	eventfd_ctx_remove_wait_queue(kirqfd->eventfd, &kirqfd->wait, &cnt);
 	eventfd_ctx_put(kirqfd->eventfd);
 	kfree(kirqfd);
@@ -909,9 +915,11 @@ irqfd_wakeup(wait_queue_entry_t *wait, unsigned int mode, int sync, void *key)
 		irqfd_inject(kirqfd);
 
 	if (flags & EPOLLHUP) {
-		mutex_lock(&irqfds_lock);
+		unsigned long flags;
+
+		spin_lock_irqsave(&irqfds_lock, flags);
 		irqfd_deactivate(kirqfd);
-		mutex_unlock(&irqfds_lock);
+		spin_unlock_irqrestore(&irqfds_lock, flags);
 	}
 
 	return 0;
@@ -929,10 +937,11 @@ irqfd_poll_func(struct file *file, wait_queue_head_t *wqh, poll_table *pt)
 static int privcmd_irqfd_assign(struct privcmd_irqfd *irqfd)
 {
 	struct privcmd_kernel_irqfd *kirqfd, *tmp;
+	unsigned long flags;
 	__poll_t events;
 	struct fd f;
 	void *dm_op;
-	int ret;
+	int ret, idx;
 
 	kirqfd = kzalloc(sizeof(*kirqfd) + irqfd->size, GFP_KERNEL);
 	if (!kirqfd)
@@ -968,18 +977,19 @@ static int privcmd_irqfd_assign(struct privcmd_irqfd *irqfd)
 	init_waitqueue_func_entry(&kirqfd->wait, irqfd_wakeup);
 	init_poll_funcptr(&kirqfd->pt, irqfd_poll_func);
 
-	mutex_lock(&irqfds_lock);
+	spin_lock_irqsave(&irqfds_lock, flags);
 
 	list_for_each_entry(tmp, &irqfds_list, list) {
 		if (kirqfd->eventfd == tmp->eventfd) {
 			ret = -EBUSY;
-			mutex_unlock(&irqfds_lock);
+			spin_unlock_irqrestore(&irqfds_lock, flags);
 			goto error_eventfd;
 		}
 	}
 
+	idx = srcu_read_lock(&irqfds_srcu);
 	list_add_tail(&kirqfd->list, &irqfds_list);
-	mutex_unlock(&irqfds_lock);
+	spin_unlock_irqrestore(&irqfds_lock, flags);
 
 	/*
 	 * Check if there was an event already pending on the eventfd before we
@@ -989,6 +999,8 @@ static int privcmd_irqfd_assign(struct privcmd_irqfd *irqfd)
 	if (events & EPOLLIN)
 		irqfd_inject(kirqfd);
 
+	srcu_read_unlock(&irqfds_srcu, idx);
+
 	/*
 	 * Do not drop the file until the kirqfd is fully initialized, otherwise
 	 * we might race against the EPOLLHUP.
@@ -1011,12 +1023,13 @@ static int privcmd_irqfd_deassign(struct privcmd_irqfd *irqfd)
 {
 	struct privcmd_kernel_irqfd *kirqfd;
 	struct eventfd_ctx *eventfd;
+	unsigned long flags;
 
 	eventfd = eventfd_ctx_fdget(irqfd->fd);
 	if (IS_ERR(eventfd))
 		return PTR_ERR(eventfd);
 
-	mutex_lock(&irqfds_lock);
+	spin_lock_irqsave(&irqfds_lock, flags);
 
 	list_for_each_entry(kirqfd, &irqfds_list, list) {
 		if (kirqfd->eventfd == eventfd) {
@@ -1025,7 +1038,7 @@ static int privcmd_irqfd_deassign(struct privcmd_irqfd *irqfd)
 		}
 	}
 
-	mutex_unlock(&irqfds_lock);
+	spin_unlock_irqrestore(&irqfds_lock, flags);
 
 	eventfd_ctx_put(eventfd);
 
@@ -1073,13 +1086,14 @@ static int privcmd_irqfd_init(void)
 static void privcmd_irqfd_exit(void)
 {
 	struct privcmd_kernel_irqfd *kirqfd, *tmp;
+	unsigned long flags;
 
-	mutex_lock(&irqfds_lock);
+	spin_lock_irqsave(&irqfds_lock, flags);
 
 	list_for_each_entry_safe(kirqfd, tmp, &irqfds_list, list)
 		irqfd_deactivate(kirqfd);
 
-	mutex_unlock(&irqfds_lock);
+	spin_unlock_irqrestore(&irqfds_lock, flags);
 
 	destroy_workqueue(irqfd_cleanup_wq);
 }
diff --git a/drivers/xen/xen-pciback/pci_stub.c b/drivers/xen/xen-pciback/pci_stub.c
index e34b623e4b41..4faebbb84999 100644
--- a/drivers/xen/xen-pciback/pci_stub.c
+++ b/drivers/xen/xen-pciback/pci_stub.c
@@ -1708,5 +1708,6 @@ static void __exit xen_pcibk_cleanup(void)
 module_init(xen_pcibk_init);
 module_exit(xen_pcibk_cleanup);
 
+MODULE_DESCRIPTION("Xen PCI-device stub driver");
 MODULE_LICENSE("Dual BSD/GPL");
 MODULE_ALIAS("xen-backend:pci");
diff --git a/include/xen/events.h b/include/xen/events.h
index 3b07409f8032..de5da58a0205 100644
--- a/include/xen/events.h
+++ b/include/xen/events.h
@@ -144,4 +144,6 @@ static inline void xen_evtchn_close(evtchn_port_t port)
 		BUG();
 }
 
+extern bool xen_fifo_events;
+
 #endif	/* _XEN_EVENTS_H */
diff --git a/kernel/locking/qspinlock.c b/kernel/locking/qspinlock.c
index 1df5fef8a656..7d96bed718e4 100644
--- a/kernel/locking/qspinlock.c
+++ b/kernel/locking/qspinlock.c
@@ -583,7 +583,7 @@ EXPORT_SYMBOL(queued_spin_lock_slowpath);
 #include "qspinlock_paravirt.h"
 #include "qspinlock.c"
 
-bool nopvspin __initdata;
+bool nopvspin;
 static __init int parse_nopvspin(char *arg)
 {
 	nopvspin = true;