xen: features and fixes for 4.13-rc1

-----BEGIN PGP SIGNATURE-----
 Version: GnuPG v2
 
 iQEcBAABAgAGBQJZXdVXAAoJELDendYovxMvVA0IAITmvH21SDTFiilKCOrxhCv0
 W3q3cOhZA4D+UtTqqIm/os/et08n72864s0mUFoY4PxETaUsb1jBav7z7Tod2c6B
 wh26UgIAhVO3ZewFSmpdPYoW0l3elC5JUMkVMfwSvHkROaU+YDEYUsLWGuIHZiiy
 V/kIskcKe08HLObU//BMjfFusmMHmQSg+TruyqRWodlWj4Rwm7q5fNZ/xaap1UCM
 O7GcHyq1k699w5YYTlIEkLWsX/pGM+auGSlT1xdjJEc2bpjH8ps0xbvAn6dsAKsE
 yoDyxQWtX2wBUXCqF0hXYAB2r1iFx2aFfLQjwc7p+V6BvxpWwSsC7Ur4QIDnm3E=
 =OLb7
 -----END PGP SIGNATURE-----

Merge tag 'for-linus-4.13-rc1-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/xen/tip

Pull xen updates from Juergen Gross:
 "Other than fixes and cleanups it contains:

   - support > 32 VCPUs at domain restore

   - support for new sysfs nodes related to Xen

   - some performance tuning for Linux running as Xen guest"

* tag 'for-linus-4.13-rc1-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/xen/tip:
  x86/xen: allow userspace access during hypercalls
  x86: xen: remove unnecessary variable in xen_foreach_remap_area()
  xen: allocate page for shared info page from low memory
  xen: avoid deadlock in xenbus driver
  xen: add sysfs node for hypervisor build id
  xen: sync include/xen/interface/version.h
  xen: add sysfs node for guest type
  doc,xen: document hypervisor sysfs nodes for xen
  xen/vcpu: Handle xen_vcpu_setup() failure at boot
  xen/vcpu: Handle xen_vcpu_setup() failure in hotplug
  xen/pv: Fix OOPS on restore for a PV, !SMP domain
  xen/pvh*: Support > 32 VCPUs at domain restore
  xen/vcpu: Simplify xen_vcpu related code
  xen-evtchn: Bind dyn evtchn:qemu-dm interrupt to next online VCPU
  xen: avoid type warning in xchg_xen_ulong
  xen: fix HYPERVISOR_dm_op() prototype
  xen: don't print error message in case of missing Xenstore entry
  arm/xen: Adjust one function call together with a variable assignment
  arm/xen: Delete an error message for a failed memory allocation in __set_phys_to_machine_multi()
  arm/xen: Improve a size determination in __set_phys_to_machine_multi()

Documentation/ABI/stable/sysfs-hypervisor-xen

@@ -0,0 +1,119 @@
+What:		/sys/hypervisor/compilation/compile_date
+Date:		March 2009
+KernelVersion:	2.6.30
+Contact:	xen-devel@lists.xenproject.org
+Description:	If running under Xen:
+		Contains the build time stamp of the Xen hypervisor
+		Might return "<denied>" in case of special security settings
+		in the hypervisor.
+
+What:		/sys/hypervisor/compilation/compiled_by
+Date:		March 2009
+KernelVersion:	2.6.30
+Contact:	xen-devel@lists.xenproject.org
+Description:	If running under Xen:
+		Contains information who built the Xen hypervisor
+		Might return "<denied>" in case of special security settings
+		in the hypervisor.
+
+What:		/sys/hypervisor/compilation/compiler
+Date:		March 2009
+KernelVersion:	2.6.30
+Contact:	xen-devel@lists.xenproject.org
+Description:	If running under Xen:
+		Compiler which was used to build the Xen hypervisor
+		Might return "<denied>" in case of special security settings
+		in the hypervisor.
+
+What:		/sys/hypervisor/properties/capabilities
+Date:		March 2009
+KernelVersion:	2.6.30
+Contact:	xen-devel@lists.xenproject.org
+Description:	If running under Xen:
+		Space separated list of supported guest system types. Each type
+		is in the format: <class>-<major>.<minor>-<arch>
+		With:
+			<class>: "xen" -- x86: paravirtualized, arm: standard
+				 "hvm" -- x86 only: fully virtualized
+			<major>: major guest interface version
+			<minor>: minor guest interface version
+			<arch>:  architecture, e.g.:
+				 "x86_32": 32 bit x86 guest without PAE
+				 "x86_32p": 32 bit x86 guest with PAE
+				 "x86_64": 64 bit x86 guest
+				 "armv7l": 32 bit arm guest
+				 "aarch64": 64 bit arm guest
+
+What:		/sys/hypervisor/properties/changeset
+Date:		March 2009
+KernelVersion:	2.6.30
+Contact:	xen-devel@lists.xenproject.org
+Description:	If running under Xen:
+		Changeset of the hypervisor (git commit)
+		Might return "<denied>" in case of special security settings
+		in the hypervisor.
+
+What:		/sys/hypervisor/properties/features
+Date:		March 2009
+KernelVersion:	2.6.30
+Contact:	xen-devel@lists.xenproject.org
+Description:	If running under Xen:
+		Features the Xen hypervisor supports for the guest as defined
+		in include/xen/interface/features.h printed as a hex value.
+
+What:		/sys/hypervisor/properties/pagesize
+Date:		March 2009
+KernelVersion:	2.6.30
+Contact:	xen-devel@lists.xenproject.org
+Description:	If running under Xen:
+		Default page size of the hypervisor printed as a hex value.
+		Might return "0" in case of special security settings
+		in the hypervisor.
+
+What:		/sys/hypervisor/properties/virtual_start
+Date:		March 2009
+KernelVersion:	2.6.30
+Contact:	xen-devel@lists.xenproject.org
+Description:	If running under Xen:
+		Virtual address of the hypervisor as a hex value.
+
+What:		/sys/hypervisor/type
+Date:		March 2009
+KernelVersion:	2.6.30
+Contact:	xen-devel@lists.xenproject.org
+Description:	If running under Xen:
+		Type of hypervisor:
+		"xen": Xen hypervisor
+
+What:		/sys/hypervisor/uuid
+Date:		March 2009
+KernelVersion:	2.6.30
+Contact:	xen-devel@lists.xenproject.org
+Description:	If running under Xen:
+		UUID of the guest as known to the Xen hypervisor.
+
+What:		/sys/hypervisor/version/extra
+Date:		March 2009
+KernelVersion:	2.6.30
+Contact:	xen-devel@lists.xenproject.org
+Description:	If running under Xen:
+		The Xen version is in the format <major>.<minor><extra>
+		This is the <extra> part of it.
+		Might return "<denied>" in case of special security settings
+		in the hypervisor.
+
+What:		/sys/hypervisor/version/major
+Date:		March 2009
+KernelVersion:	2.6.30
+Contact:	xen-devel@lists.xenproject.org
+Description:	If running under Xen:
+		The Xen version is in the format <major>.<minor><extra>
+		This is the <major> part of it.
+
+What:		/sys/hypervisor/version/minor
+Date:		March 2009
+KernelVersion:	2.6.30
+Contact:	xen-devel@lists.xenproject.org
+Description:	If running under Xen:
+		The Xen version is in the format <major>.<minor><extra>
+		This is the <minor> part of it.

Documentation/ABI/testing/sysfs-hypervisor-xen

@@ -1,8 +1,19 @@
+What:		/sys/hypervisor/guest_type
+Date:		June 2017
+KernelVersion:	4.13
+Contact:	xen-devel@lists.xenproject.org
+Description:	If running under Xen:
+		Type of guest:
+		"Xen": standard guest type on arm
+		"HVM": fully virtualized guest (x86)
+		"PV": paravirtualized guest (x86)
+		"PVH": fully virtualized guest without legacy emulation (x86)
+
 What:		/sys/hypervisor/pmu/pmu_mode
 Date:		August 2015
 KernelVersion:	4.3
 Contact:	Boris Ostrovsky <boris.ostrovsky@oracle.com>
-Description:
+Description:	If running under Xen:
 		Describes mode that Xen's performance-monitoring unit (PMU)
 		uses. Accepted values are
 			"off"  -- PMU is disabled
@@ -17,7 +28,16 @@ What:           /sys/hypervisor/pmu/pmu_features
 Date:           August 2015
 KernelVersion:  4.3
 Contact:        Boris Ostrovsky <boris.ostrovsky@oracle.com>
-Description:
+Description:	If running under Xen:
 		Describes Xen PMU features (as an integer). A set bit indicates
 		that the corresponding feature is enabled. See
 		include/xen/interface/xenpmu.h for available features
+
+What:		/sys/hypervisor/properties/buildid
+Date:		June 2017
+KernelVersion:	4.13
+Contact:	xen-devel@lists.xenproject.org
+Description:	If running under Xen:
+		Build id of the hypervisor, needed for hypervisor live patching.
+		Might return "<denied>" in case of special security settings
+		in the hypervisor.

MAINTAINERS

@@ -14228,6 +14228,8 @@ F:	drivers/xen/
 F:	arch/x86/include/asm/xen/
 F:	include/xen/
 F:	include/uapi/xen/
+F:	Documentation/ABI/stable/sysfs-hypervisor-xen
+F:	Documentation/ABI/testing/sysfs-hypervisor-xen
 
 XEN HYPERVISOR ARM
 M:	Stefano Stabellini <sstabellini@kernel.org>

arch/arm/include/asm/xen/events.h

@@ -16,7 +16,7 @@ static inline int xen_irqs_disabled(struct pt_regs *regs)
 	return raw_irqs_disabled_flags(regs->ARM_cpsr);
 }
 
-#define xchg_xen_ulong(ptr, val) atomic64_xchg(container_of((ptr),	\
+#define xchg_xen_ulong(ptr, val) atomic64_xchg(container_of((long long*)(ptr),\
 							    atomic64_t,	\
 							    counter), (val))
 

arch/arm/xen/p2m.c

@@ -144,17 +144,17 @@ bool __set_phys_to_machine_multi(unsigned long pfn,
 		return true;
 	}
 
-	p2m_entry = kzalloc(sizeof(struct xen_p2m_entry), GFP_NOWAIT);
+	p2m_entry = kzalloc(sizeof(*p2m_entry), GFP_NOWAIT);
-	if (!p2m_entry) {
+	if (!p2m_entry)
-		pr_warn("cannot allocate xen_p2m_entry\n");
 		return false;
-	}
+
 	p2m_entry->pfn = pfn;
 	p2m_entry->nr_pages = nr_pages;
 	p2m_entry->mfn = mfn;
 
 	write_lock_irqsave(&p2m_lock, irqflags);
-	if ((rc = xen_add_phys_to_mach_entry(p2m_entry)) < 0) {
+	rc = xen_add_phys_to_mach_entry(p2m_entry);
+	if (rc < 0) {
 		write_unlock_irqrestore(&p2m_lock, irqflags);
 		return false;
 	}

arch/x86/include/asm/xen/hypercall.h

@@ -43,6 +43,7 @@
 
 #include <asm/page.h>
 #include <asm/pgtable.h>
+#include <asm/smap.h>
 
 #include <xen/interface/xen.h>
 #include <xen/interface/sched.h>
@@ -50,6 +51,8 @@
 #include <xen/interface/platform.h>
 #include <xen/interface/xen-mca.h>
 
+struct xen_dm_op_buf;
+
 /*
  * The hypercall asms have to meet several constraints:
  * - Work on 32- and 64-bit.
@@ -214,10 +217,12 @@ privcmd_call(unsigned call,
 	__HYPERCALL_DECLS;
 	__HYPERCALL_5ARG(a1, a2, a3, a4, a5);
 
+	stac();
 	asm volatile("call *%[call]"
 		     : __HYPERCALL_5PARAM
 		     : [call] "a" (&hypercall_page[call])
 		     : __HYPERCALL_CLOBBER5);
+	clac();
 
 	return (long)__res;
 }
@@ -474,9 +479,13 @@ HYPERVISOR_xenpmu_op(unsigned int op, void *arg)
 
 static inline int
 HYPERVISOR_dm_op(
-	domid_t dom, unsigned int nr_bufs, void *bufs)
+	domid_t dom, unsigned int nr_bufs, struct xen_dm_op_buf *bufs)
 {
-	return _hypercall3(int, dm_op, dom, nr_bufs, bufs);
+	int ret;
+	stac();
+	ret = _hypercall3(int, dm_op, dom, nr_bufs, bufs);
+	clac();
+	return ret;
 }
 
 static inline void

arch/x86/xen/enlighten.c

@@ -106,15 +106,83 @@ int xen_cpuhp_setup(int (*cpu_up_prepare_cb)(unsigned int),
 	return rc >= 0 ? 0 : rc;
 }
 
-static void clamp_max_cpus(void)
+static int xen_vcpu_setup_restore(int cpu)
 {
-#ifdef CONFIG_SMP
+	int rc = 0;
-	if (setup_max_cpus > MAX_VIRT_CPUS)
+
-		setup_max_cpus = MAX_VIRT_CPUS;
+	/* Any per_cpu(xen_vcpu) is stale, so reset it */
-#endif
+	xen_vcpu_info_reset(cpu);
+
+	/*
+	 * For PVH and PVHVM, setup online VCPUs only. The rest will
+	 * be handled by hotplug.
+	 */
+	if (xen_pv_domain() ||
+	    (xen_hvm_domain() && cpu_online(cpu))) {
+		rc = xen_vcpu_setup(cpu);
+	}
+
+	return rc;
 }
 
-void xen_vcpu_setup(int cpu)
+/*
+ * On restore, set the vcpu placement up again.
+ * If it fails, then we're in a bad state, since
+ * we can't back out from using it...
+ */
+void xen_vcpu_restore(void)
+{
+	int cpu, rc;
+
+	for_each_possible_cpu(cpu) {
+		bool other_cpu = (cpu != smp_processor_id());
+		bool is_up;
+
+		if (xen_vcpu_nr(cpu) == XEN_VCPU_ID_INVALID)
+			continue;
+
+		/* Only Xen 4.5 and higher support this. */
+		is_up = HYPERVISOR_vcpu_op(VCPUOP_is_up,
+					   xen_vcpu_nr(cpu), NULL) > 0;
+
+		if (other_cpu && is_up &&
+		    HYPERVISOR_vcpu_op(VCPUOP_down, xen_vcpu_nr(cpu), NULL))
+			BUG();
+
+		if (xen_pv_domain() || xen_feature(XENFEAT_hvm_safe_pvclock))
+			xen_setup_runstate_info(cpu);
+
+		rc = xen_vcpu_setup_restore(cpu);
+		if (rc)
+			pr_emerg_once("vcpu restore failed for cpu=%d err=%d. "
+					"System will hang.\n", cpu, rc);
+		/*
+		 * In case xen_vcpu_setup_restore() fails, do not bring up the
+		 * VCPU. This helps us avoid the resulting OOPS when the VCPU
+		 * accesses pvclock_vcpu_time via xen_vcpu (which is NULL.)
+		 * Note that this does not improve the situation much -- now the
+		 * VM hangs instead of OOPSing -- with the VCPUs that did not
+		 * fail, spinning in stop_machine(), waiting for the failed
+		 * VCPUs to come up.
+		 */
+		if (other_cpu && is_up && (rc == 0) &&
+		    HYPERVISOR_vcpu_op(VCPUOP_up, xen_vcpu_nr(cpu), NULL))
+			BUG();
+	}
+}
+
+void xen_vcpu_info_reset(int cpu)
+{
+	if (xen_vcpu_nr(cpu) < MAX_VIRT_CPUS) {
+		per_cpu(xen_vcpu, cpu) =
+			&HYPERVISOR_shared_info->vcpu_info[xen_vcpu_nr(cpu)];
+	} else {
+		/* Set to NULL so that if somebody accesses it we get an OOPS */
+		per_cpu(xen_vcpu, cpu) = NULL;
+	}
+}
+
+int xen_vcpu_setup(int cpu)
 {
 	struct vcpu_register_vcpu_info info;
 	int err;
@@ -123,11 +191,11 @@ void xen_vcpu_setup(int cpu)
 	BUG_ON(HYPERVISOR_shared_info == &xen_dummy_shared_info);
 
 	/*
-	 * This path is called twice on PVHVM - first during bootup via
+	 * This path is called on PVHVM at bootup (xen_hvm_smp_prepare_boot_cpu)
-	 * smp_init -> xen_hvm_cpu_notify, and then if the VCPU is being
+	 * and at restore (xen_vcpu_restore). Also called for hotplugged
-	 * hotplugged: cpu_up -> xen_hvm_cpu_notify.
+	 * VCPUs (cpu_init -> xen_hvm_cpu_prepare_hvm).
-	 * As we can only do the VCPUOP_register_vcpu_info once lets
+	 * However, the hypercall can only be done once (see below) so if a VCPU
-	 * not over-write its result.
+	 * is offlined and comes back online then let's not redo the hypercall.
 	 *
 	 * For PV it is called during restore (xen_vcpu_restore) and bootup
 	 * (xen_setup_vcpu_info_placement). The hotplug mechanism does not
@@ -135,42 +203,44 @@ void xen_vcpu_setup(int cpu)
 	 */
 	if (xen_hvm_domain()) {
 		if (per_cpu(xen_vcpu, cpu) == &per_cpu(xen_vcpu_info, cpu))
-			return;
+			return 0;
-	}
-	if (xen_vcpu_nr(cpu) < MAX_VIRT_CPUS)
-		per_cpu(xen_vcpu, cpu) =
-			&HYPERVISOR_shared_info->vcpu_info[xen_vcpu_nr(cpu)];
-
-	if (!xen_have_vcpu_info_placement) {
-		if (cpu >= MAX_VIRT_CPUS)
-			clamp_max_cpus();
-		return;
 	}
 
-	vcpup = &per_cpu(xen_vcpu_info, cpu);
+	if (xen_have_vcpu_info_placement) {
-	info.mfn = arbitrary_virt_to_mfn(vcpup);
+		vcpup = &per_cpu(xen_vcpu_info, cpu);
-	info.offset = offset_in_page(vcpup);
+		info.mfn = arbitrary_virt_to_mfn(vcpup);
+		info.offset = offset_in_page(vcpup);
 
-	/* Check to see if the hypervisor will put the vcpu_info
+		/*
-	   structure where we want it, which allows direct access via
+		 * Check to see if the hypervisor will put the vcpu_info
-	   a percpu-variable.
+		 * structure where we want it, which allows direct access via
-	   N.B. This hypercall can _only_ be called once per CPU. Subsequent
+		 * a percpu-variable.
-	   calls will error out with -EINVAL. This is due to the fact that
+		 * N.B. This hypercall can _only_ be called once per CPU.
-	   hypervisor has no unregister variant and this hypercall does not
+		 * Subsequent calls will error out with -EINVAL. This is due to
-	   allow to over-write info.mfn and info.offset.
+		 * the fact that hypervisor has no unregister variant and this
-	 */
+		 * hypercall does not allow to over-write info.mfn and
-	err = HYPERVISOR_vcpu_op(VCPUOP_register_vcpu_info, xen_vcpu_nr(cpu),
+		 * info.offset.
-				 &info);
+		 */
+		err = HYPERVISOR_vcpu_op(VCPUOP_register_vcpu_info,
+					 xen_vcpu_nr(cpu), &info);
 
-	if (err) {
+		if (err) {
-		printk(KERN_DEBUG "register_vcpu_info failed: err=%d\n", err);
+			pr_warn_once("register_vcpu_info failed: cpu=%d err=%d\n",
-		xen_have_vcpu_info_placement = 0;
+				     cpu, err);
-		clamp_max_cpus();
+			xen_have_vcpu_info_placement = 0;
-	} else {
+		} else {
-		/* This cpu is using the registered vcpu info, even if
+			/*
-		   later ones fail to. */
+			 * This cpu is using the registered vcpu info, even if
-		per_cpu(xen_vcpu, cpu) = vcpup;
+			 * later ones fail to.
+			 */
+			per_cpu(xen_vcpu, cpu) = vcpup;
+		}
 	}
+
+	if (!xen_have_vcpu_info_placement)
+		xen_vcpu_info_reset(cpu);
+
+	return ((per_cpu(xen_vcpu, cpu) == NULL) ? -ENODEV : 0);
 }
 
 void xen_reboot(int reason)

arch/x86/xen/enlighten_hvm.c

@@ -1,5 +1,6 @@
 #include <linux/cpu.h>
 #include <linux/kexec.h>
+#include <linux/memblock.h>
 
 #include <xen/features.h>
 #include <xen/events.h>
@@ -10,9 +11,11 @@
 #include <asm/reboot.h>
 #include <asm/setup.h>
 #include <asm/hypervisor.h>
+#include <asm/e820/api.h>
 
 #include <asm/xen/cpuid.h>
 #include <asm/xen/hypervisor.h>
+#include <asm/xen/page.h>
 
 #include "xen-ops.h"
 #include "mmu.h"
@@ -20,37 +23,34 @@
 
 void __ref xen_hvm_init_shared_info(void)
 {
-	int cpu;
 	struct xen_add_to_physmap xatp;
-	static struct shared_info *shared_info_page;
+	u64 pa;
+
+	if (HYPERVISOR_shared_info == &xen_dummy_shared_info) {
+		/*
+		 * Search for a free page starting at 4kB physical address.
+		 * Low memory is preferred to avoid an EPT large page split up
+		 * by the mapping.
+		 * Starting below X86_RESERVE_LOW (usually 64kB) is fine as
+		 * the BIOS used for HVM guests is well behaved and won't
+		 * clobber memory other than the first 4kB.
+		 */
+		for (pa = PAGE_SIZE;
+		     !e820__mapped_all(pa, pa + PAGE_SIZE, E820_TYPE_RAM) ||
+		     memblock_is_reserved(pa);
+		     pa += PAGE_SIZE)
+			;
+
+		memblock_reserve(pa, PAGE_SIZE);
+		HYPERVISOR_shared_info = __va(pa);
+	}
 
-	if (!shared_info_page)
-		shared_info_page = (struct shared_info *)
-			extend_brk(PAGE_SIZE, PAGE_SIZE);
 	xatp.domid = DOMID_SELF;
 	xatp.idx = 0;
 	xatp.space = XENMAPSPACE_shared_info;
-	xatp.gpfn = __pa(shared_info_page) >> PAGE_SHIFT;
+	xatp.gpfn = virt_to_pfn(HYPERVISOR_shared_info);
 	if (HYPERVISOR_memory_op(XENMEM_add_to_physmap, &xatp))
 		BUG();
-
-	HYPERVISOR_shared_info = (struct shared_info *)shared_info_page;
-
-	/* xen_vcpu is a pointer to the vcpu_info struct in the shared_info
-	 * page, we use it in the event channel upcall and in some pvclock
-	 * related functions. We don't need the vcpu_info placement
-	 * optimizations because we don't use any pv_mmu or pv_irq op on
-	 * HVM.
-	 * When xen_hvm_init_shared_info is run at boot time only vcpu 0 is
-	 * online but xen_hvm_init_shared_info is run at resume time too and
-	 * in that case multiple vcpus might be online. */
-	for_each_online_cpu(cpu) {
-		/* Leave it to be NULL. */
-		if (xen_vcpu_nr(cpu) >= MAX_VIRT_CPUS)
-			continue;
-		per_cpu(xen_vcpu, cpu) =
-			&HYPERVISOR_shared_info->vcpu_info[xen_vcpu_nr(cpu)];
-	}
 }
 
 static void __init init_hvm_pv_info(void)
@@ -106,7 +106,7 @@ static void xen_hvm_crash_shutdown(struct pt_regs *regs)
 
 static int xen_cpu_up_prepare_hvm(unsigned int cpu)
 {
-	int rc;
+	int rc = 0;
 
 	/*
 	 * This can happen if CPU was offlined earlier and
@@ -121,7 +121,9 @@ static int xen_cpu_up_prepare_hvm(unsigned int cpu)
 		per_cpu(xen_vcpu_id, cpu) = cpu_acpi_id(cpu);
 	else
 		per_cpu(xen_vcpu_id, cpu) = cpu;
-	xen_vcpu_setup(cpu);
+	rc = xen_vcpu_setup(cpu);
+	if (rc)
+		return rc;
 
 	if (xen_have_vector_callback && xen_feature(XENFEAT_hvm_safe_pvclock))
 		xen_setup_timer(cpu);
@@ -130,9 +132,8 @@ static int xen_cpu_up_prepare_hvm(unsigned int cpu)
 	if (rc) {
 		WARN(1, "xen_smp_intr_init() for CPU %d failed: %d\n",
 		     cpu, rc);
-		return rc;
 	}
-	return 0;
+	return rc;
 }
 
 static int xen_cpu_dead_hvm(unsigned int cpu)
@@ -154,6 +155,13 @@ static void __init xen_hvm_guest_init(void)
 
 	xen_hvm_init_shared_info();
 
+	/*
+	 * xen_vcpu is a pointer to the vcpu_info struct in the shared_info
+	 * page, we use it in the event channel upcall and in some pvclock
+	 * related functions.
+	 */
+	xen_vcpu_info_reset(0);
+
 	xen_panic_handler_init();
 
 	if (xen_feature(XENFEAT_hvm_callback_vector))

arch/x86/xen/enlighten_pv.c

@@ -89,8 +89,6 @@
 
 void *xen_initial_gdt;
 
-RESERVE_BRK(shared_info_page_brk, PAGE_SIZE);
-
 static int xen_cpu_up_prepare_pv(unsigned int cpu);
 static int xen_cpu_dead_pv(unsigned int cpu);
 
@@ -107,35 +105,6 @@ struct tls_descs {
  */
 static DEFINE_PER_CPU(struct tls_descs, shadow_tls_desc);
 
-/*
- * On restore, set the vcpu placement up again.
- * If it fails, then we're in a bad state, since
- * we can't back out from using it...
- */
-void xen_vcpu_restore(void)
-{
-	int cpu;
-
-	for_each_possible_cpu(cpu) {
-		bool other_cpu = (cpu != smp_processor_id());
-		bool is_up = HYPERVISOR_vcpu_op(VCPUOP_is_up, xen_vcpu_nr(cpu),
-						NULL);
-
-		if (other_cpu && is_up &&
-		    HYPERVISOR_vcpu_op(VCPUOP_down, xen_vcpu_nr(cpu), NULL))
-			BUG();
-
-		xen_setup_runstate_info(cpu);
-
-		if (xen_have_vcpu_info_placement)
-			xen_vcpu_setup(cpu);
-
-		if (other_cpu && is_up &&
-		    HYPERVISOR_vcpu_op(VCPUOP_up, xen_vcpu_nr(cpu), NULL))
-			BUG();
-	}
-}
-
 static void __init xen_banner(void)
 {
 	unsigned version = HYPERVISOR_xen_version(XENVER_version, NULL);
@@ -960,30 +929,43 @@ void xen_setup_shared_info(void)
 	HYPERVISOR_shared_info =
 		(struct shared_info *)fix_to_virt(FIX_PARAVIRT_BOOTMAP);
 
-#ifndef CONFIG_SMP
-	/* In UP this is as good a place as any to set up shared info */
-	xen_setup_vcpu_info_placement();
-#endif
-
 	xen_setup_mfn_list_list();
 
-	/*
+	if (system_state == SYSTEM_BOOTING) {
-	 * Now that shared info is set up we can start using routines that
+#ifndef CONFIG_SMP
-	 * point to pvclock area.
+		/*
-	 */
+		 * In UP this is as good a place as any to set up shared info.
-	if (system_state == SYSTEM_BOOTING)
+		 * Limit this to boot only, at restore vcpu setup is done via
+		 * xen_vcpu_restore().
+		 */
+		xen_setup_vcpu_info_placement();
+#endif
+		/*
+		 * Now that shared info is set up we can start using routines
+		 * that point to pvclock area.
+		 */
 		xen_init_time_ops();
+	}
 }
 
 /* This is called once we have the cpu_possible_mask */
-void xen_setup_vcpu_info_placement(void)
+void __ref xen_setup_vcpu_info_placement(void)
 {
 	int cpu;
 
 	for_each_possible_cpu(cpu) {
 		/* Set up direct vCPU id mapping for PV guests. */
 		per_cpu(xen_vcpu_id, cpu) = cpu;
-		xen_vcpu_setup(cpu);
+
+		/*
+		 * xen_vcpu_setup(cpu) can fail  -- in which case it
+		 * falls back to the shared_info version for cpus
+		 * where xen_vcpu_nr(cpu) < MAX_VIRT_CPUS.
+		 *
+		 * xen_cpu_up_prepare_pv() handles the rest by failing
+		 * them in hotplug.
+		 */
+		(void) xen_vcpu_setup(cpu);
 	}
 
 	/*
@@ -1332,9 +1314,17 @@ asmlinkage __visible void __init xen_start_kernel(void)
 	 */
 	acpi_numa = -1;
 #endif
-	/* Don't do the full vcpu_info placement stuff until we have a
+	/* Let's presume PV guests always boot on vCPU with id 0. */
-	   possible map and a non-dummy shared_info. */
+	per_cpu(xen_vcpu_id, 0) = 0;
-	per_cpu(xen_vcpu, 0) = &HYPERVISOR_shared_info->vcpu_info[0];
+
+	/*
+	 * Setup xen_vcpu early because start_kernel needs it for
+	 * local_irq_disable(), irqs_disabled().
+	 *
+	 * Don't do the full vcpu_info placement stuff until we have
+	 * the cpu_possible_mask and a non-dummy shared_info.
+	 */
+	xen_vcpu_info_reset(0);
 
 	WARN_ON(xen_cpuhp_setup(xen_cpu_up_prepare_pv, xen_cpu_dead_pv));
 
@@ -1431,9 +1421,7 @@ asmlinkage __visible void __init xen_start_kernel(void)
 #endif
 	xen_raw_console_write("about to get started...\n");
 
-	/* Let's presume PV guests always boot on vCPU with id 0. */
+	/* We need this for printk timestamps */
-	per_cpu(xen_vcpu_id, 0) = 0;
-
 	xen_setup_runstate_info(0);
 
 	xen_efi_init();
@@ -1451,6 +1439,9 @@ static int xen_cpu_up_prepare_pv(unsigned int cpu)
 {
 	int rc;
 
+	if (per_cpu(xen_vcpu, cpu) == NULL)
+		return -ENODEV;
+
 	xen_setup_timer(cpu);
 
 	rc = xen_smp_intr_init(cpu);

arch/x86/xen/setup.c

@@ -499,7 +499,7 @@ static unsigned long __init xen_foreach_remap_area(unsigned long nr_pages,
 void __init xen_remap_memory(void)
 {
 	unsigned long buf = (unsigned long)&xen_remap_buf;
-	unsigned long mfn_save, mfn, pfn;
+	unsigned long mfn_save, pfn;
 	unsigned long remapped = 0;
 	unsigned int i;
 	unsigned long pfn_s = ~0UL;
@@ -515,8 +515,7 @@ void __init xen_remap_memory(void)
 
 		pfn = xen_remap_buf.target_pfn;
 		for (i = 0; i < xen_remap_buf.size; i++) {
-			mfn = xen_remap_buf.mfns[i];
+			xen_update_mem_tables(pfn, xen_remap_buf.mfns[i]);
-			xen_update_mem_tables(pfn, mfn);
 			remapped++;
 			pfn++;
 		}
@@ -530,8 +529,6 @@ void __init xen_remap_memory(void)
 			pfn_s = xen_remap_buf.target_pfn;
 			len = xen_remap_buf.size;
 		}
-
-		mfn = xen_remap_mfn;
 		xen_remap_mfn = xen_remap_buf.next_area_mfn;
 	}
 

arch/x86/xen/smp.c

@@ -1,4 +1,5 @@
 #include <linux/smp.h>
+#include <linux/cpu.h>
 #include <linux/slab.h>
 #include <linux/cpumask.h>
 #include <linux/percpu.h>
@@ -114,6 +115,36 @@ int xen_smp_intr_init(unsigned int cpu)
 	return rc;
 }
 
+void __init xen_smp_cpus_done(unsigned int max_cpus)
+{
+	int cpu, rc, count = 0;
+
+	if (xen_hvm_domain())
+		native_smp_cpus_done(max_cpus);
+
+	if (xen_have_vcpu_info_placement)
+		return;
+
+	for_each_online_cpu(cpu) {
+		if (xen_vcpu_nr(cpu) < MAX_VIRT_CPUS)
+			continue;
+
+		rc = cpu_down(cpu);
+
+		if (rc == 0) {
+			/*
+			 * Reset vcpu_info so this cpu cannot be onlined again.
+			 */
+			xen_vcpu_info_reset(cpu);
+			count++;
+		} else {
+			pr_warn("%s: failed to bring CPU %d down, error %d\n",
+				__func__, cpu, rc);
+		}
+	}
+	WARN(count, "%s: brought %d CPUs offline\n", __func__, count);
+}
+
 void xen_smp_send_reschedule(int cpu)
 {
 	xen_send_IPI_one(cpu, XEN_RESCHEDULE_VECTOR);

arch/x86/xen/smp.h

@@ -14,6 +14,8 @@ 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_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);

arch/x86/xen/smp_hvm.c

@@ -12,7 +12,8 @@ static void __init xen_hvm_smp_prepare_boot_cpu(void)
 	native_smp_prepare_boot_cpu();
 
 	/*
-	 * Setup vcpu_info for boot CPU.
+	 * Setup vcpu_info for boot CPU. Secondary CPUs get their vcpu_info
+	 * in xen_cpu_up_prepare_hvm().
 	 */
 	xen_vcpu_setup(0);
 
@@ -27,10 +28,20 @@ static void __init xen_hvm_smp_prepare_boot_cpu(void)
 
 static void __init xen_hvm_smp_prepare_cpus(unsigned int max_cpus)
 {
+	int cpu;
+
 	native_smp_prepare_cpus(max_cpus);
 	WARN_ON(xen_smp_intr_init(0));
 
 	xen_init_lock_cpu(0);
+
+	for_each_possible_cpu(cpu) {
+		if (cpu == 0)
+			continue;
+
+		/* Set default vcpu_id to make sure that we don't use cpu-0's */
+		per_cpu(xen_vcpu_id, cpu) = XEN_VCPU_ID_INVALID;
+	}
 }
 
 #ifdef CONFIG_HOTPLUG_CPU
@@ -60,4 +71,5 @@ void __init xen_hvm_smp_init(void)
 	smp_ops.send_call_func_ipi = xen_smp_send_call_function_ipi;
 	smp_ops.send_call_func_single_ipi = xen_smp_send_call_function_single_ipi;
 	smp_ops.smp_prepare_boot_cpu = xen_hvm_smp_prepare_boot_cpu;
+	smp_ops.smp_cpus_done = xen_smp_cpus_done;
 }

arch/x86/xen/smp_pv.c

@@ -371,10 +371,6 @@ static int xen_pv_cpu_up(unsigned int cpu, struct task_struct *idle)
 	return 0;
 }
 
-static void xen_pv_smp_cpus_done(unsigned int max_cpus)
-{
-}
-
 #ifdef CONFIG_HOTPLUG_CPU
 static int xen_pv_cpu_disable(void)
 {
@@ -469,7 +465,7 @@ static irqreturn_t xen_irq_work_interrupt(int irq, void *dev_id)
 static const struct smp_ops xen_smp_ops __initconst = {
 	.smp_prepare_boot_cpu = xen_pv_smp_prepare_boot_cpu,
 	.smp_prepare_cpus = xen_pv_smp_prepare_cpus,
-	.smp_cpus_done = xen_pv_smp_cpus_done,
+	.smp_cpus_done = xen_smp_cpus_done,
 
 	.cpu_up = xen_pv_cpu_up,
 	.cpu_die = xen_pv_cpu_die,

arch/x86/xen/suspend_hvm.c

@@ -8,15 +8,10 @@
 
 void xen_hvm_post_suspend(int suspend_cancelled)
 {
-	int cpu;
+	if (!suspend_cancelled) {
-
-	if (!suspend_cancelled)
 		xen_hvm_init_shared_info();
+		xen_vcpu_restore();
+	}
 	xen_callback_vector();
 	xen_unplug_emulated_devices();
-	if (xen_feature(XENFEAT_hvm_safe_pvclock)) {
-		for_each_online_cpu(cpu) {
-			xen_setup_runstate_info(cpu);
-		}
-	}
 }

arch/x86/xen/xen-ops.h

@@ -78,7 +78,8 @@ bool xen_vcpu_stolen(int vcpu);
 
 extern int xen_have_vcpu_info_placement;
 
-void xen_vcpu_setup(int cpu);
+int xen_vcpu_setup(int cpu);
+void xen_vcpu_info_reset(int cpu);
 void xen_setup_vcpu_info_placement(void);
 
 #ifdef CONFIG_SMP

drivers/xen/events/events_base.c

@@ -1303,10 +1303,9 @@ void rebind_evtchn_irq(int evtchn, int irq)
 }
 
 /* Rebind an evtchn so that it gets delivered to a specific cpu */
-static int rebind_irq_to_cpu(unsigned irq, unsigned tcpu)
+int xen_rebind_evtchn_to_cpu(int evtchn, unsigned tcpu)
 {
 	struct evtchn_bind_vcpu bind_vcpu;
-	int evtchn = evtchn_from_irq(irq);
 	int masked;
 
 	if (!VALID_EVTCHN(evtchn))
@@ -1338,12 +1337,13 @@ static int rebind_irq_to_cpu(unsigned irq, unsigned tcpu)
 
 	return 0;
 }
+EXPORT_SYMBOL_GPL(xen_rebind_evtchn_to_cpu);
 
 static int set_affinity_irq(struct irq_data *data, const struct cpumask *dest,
 			    bool force)
 {
 	unsigned tcpu = cpumask_first_and(dest, cpu_online_mask);
-	int ret = rebind_irq_to_cpu(data->irq, tcpu);
+	int ret = xen_rebind_evtchn_to_cpu(evtchn_from_irq(data->irq), tcpu);
 
 	if (!ret)
 		irq_data_update_effective_affinity(data, cpumask_of(tcpu));

drivers/xen/evtchn.c

@@ -421,6 +421,36 @@ static void evtchn_unbind_from_user(struct per_user_data *u,
 	del_evtchn(u, evtchn);
 }
 
+static DEFINE_PER_CPU(int, bind_last_selected_cpu);
+
+static void evtchn_bind_interdom_next_vcpu(int evtchn)
+{
+	unsigned int selected_cpu, irq;
+	struct irq_desc *desc;
+	unsigned long flags;
+
+	irq = irq_from_evtchn(evtchn);
+	desc = irq_to_desc(irq);
+
+	if (!desc)
+		return;
+
+	raw_spin_lock_irqsave(&desc->lock, flags);
+	selected_cpu = this_cpu_read(bind_last_selected_cpu);
+	selected_cpu = cpumask_next_and(selected_cpu,
+			desc->irq_common_data.affinity, cpu_online_mask);
+
+	if (unlikely(selected_cpu >= nr_cpu_ids))
+		selected_cpu = cpumask_first_and(desc->irq_common_data.affinity,
+				cpu_online_mask);
+
+	this_cpu_write(bind_last_selected_cpu, selected_cpu);
+
+	/* unmask expects irqs to be disabled */
+	xen_rebind_evtchn_to_cpu(evtchn, selected_cpu);
+	raw_spin_unlock_irqrestore(&desc->lock, flags);
+}
+
 static long evtchn_ioctl(struct file *file,
 			 unsigned int cmd, unsigned long arg)
 {
@@ -478,8 +508,10 @@ static long evtchn_ioctl(struct file *file,
 			break;
 
 		rc = evtchn_bind_to_user(u, bind_interdomain.local_port);
-		if (rc == 0)
+		if (rc == 0) {
 			rc = bind_interdomain.local_port;
+			evtchn_bind_interdom_next_vcpu(rc);
+		}
 		break;
 	}
 

drivers/xen/manage.c

@@ -278,8 +278,16 @@ static void sysrq_handler(struct xenbus_watch *watch, const char *path,
 	err = xenbus_transaction_start(&xbt);
 	if (err)
 		return;
-	if (xenbus_scanf(xbt, "control", "sysrq", "%c", &sysrq_key) < 0) {
+	err = xenbus_scanf(xbt, "control", "sysrq", "%c", &sysrq_key);
-		pr_err("Unable to read sysrq code in control/sysrq\n");
+	if (err < 0) {
+		/*
+		 * The Xenstore watch fires directly after registering it and
+		 * after a suspend/resume cycle. So ENOENT is no error but
+		 * might happen in those cases.
+		 */
+		if (err != -ENOENT)
+			pr_err("Error %d reading sysrq code in control/sysrq\n",
+			       err);
 		xenbus_transaction_end(xbt, 1);
 		return;
 	}

drivers/xen/sys-hypervisor.c

@@ -50,6 +50,35 @@ static int __init xen_sysfs_type_init(void)
 	return sysfs_create_file(hypervisor_kobj, &type_attr.attr);
 }
 
+static ssize_t guest_type_show(struct hyp_sysfs_attr *attr, char *buffer)
+{
+	const char *type;
+
+	switch (xen_domain_type) {
+	case XEN_NATIVE:
+		/* ARM only. */
+		type = "Xen";
+		break;
+	case XEN_PV_DOMAIN:
+		type = "PV";
+		break;
+	case XEN_HVM_DOMAIN:
+		type = xen_pvh_domain() ? "PVH" : "HVM";
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	return sprintf(buffer, "%s\n", type);
+}
+
+HYPERVISOR_ATTR_RO(guest_type);
+
+static int __init xen_sysfs_guest_type_init(void)
+{
+	return sysfs_create_file(hypervisor_kobj, &guest_type_attr.attr);
+}
+
 /* xen version attributes */
 static ssize_t major_show(struct hyp_sysfs_attr *attr, char *buffer)
 {
@@ -327,12 +356,40 @@ static ssize_t features_show(struct hyp_sysfs_attr *attr, char *buffer)
 
 HYPERVISOR_ATTR_RO(features);
 
+static ssize_t buildid_show(struct hyp_sysfs_attr *attr, char *buffer)
+{
+	ssize_t ret;
+	struct xen_build_id *buildid;
+
+	ret = HYPERVISOR_xen_version(XENVER_build_id, NULL);
+	if (ret < 0) {
+		if (ret == -EPERM)
+			ret = sprintf(buffer, "<denied>");
+		return ret;
+	}
+
+	buildid = kmalloc(sizeof(*buildid) + ret, GFP_KERNEL);
+	if (!buildid)
+		return -ENOMEM;
+
+	buildid->len = ret;
+	ret = HYPERVISOR_xen_version(XENVER_build_id, buildid);
+	if (ret > 0)
+		ret = sprintf(buffer, "%s", buildid->buf);
+	kfree(buildid);
+
+	return ret;
+}
+
+HYPERVISOR_ATTR_RO(buildid);
+
 static struct attribute *xen_properties_attrs[] = {
 	&capabilities_attr.attr,
 	&changeset_attr.attr,
 	&virtual_start_attr.attr,
 	&pagesize_attr.attr,
 	&features_attr.attr,
+	&buildid_attr.attr,
 	NULL
 };
 
@@ -471,6 +528,9 @@ static int __init hyper_sysfs_init(void)
 	ret = xen_sysfs_type_init();
 	if (ret)
 		goto out;
+	ret = xen_sysfs_guest_type_init();
+	if (ret)
+		goto guest_type_out;
 	ret = xen_sysfs_version_init();
 	if (ret)
 		goto version_out;
@@ -502,6 +562,8 @@ uuid_out:
 comp_out:
 	sysfs_remove_group(hypervisor_kobj, &version_group);
 version_out:
+	sysfs_remove_file(hypervisor_kobj, &guest_type_attr.attr);
+guest_type_out:
 	sysfs_remove_file(hypervisor_kobj, &type_attr.attr);
 out:
 	return ret;

drivers/xen/xenbus/xenbus_comms.c

@@ -299,17 +299,7 @@ static int process_msg(void)
 		mutex_lock(&xb_write_mutex);
 		list_for_each_entry(req, &xs_reply_list, list) {
 			if (req->msg.req_id == state.msg.req_id) {
-				if (req->state == xb_req_state_wait_reply) {
+				list_del(&req->list);
-					req->msg.type = state.msg.type;
-					req->msg.len = state.msg.len;
-					req->body = state.body;
-					req->state = xb_req_state_got_reply;
-					list_del(&req->list);
-					req->cb(req);
-				} else {
-					list_del(&req->list);
-					kfree(req);
-				}
 				err = 0;
 				break;
 			}
@@ -317,6 +307,15 @@ static int process_msg(void)
 		mutex_unlock(&xb_write_mutex);
 		if (err)
 			goto out;
+
+		if (req->state == xb_req_state_wait_reply) {
+			req->msg.type = state.msg.type;
+			req->msg.len = state.msg.len;
+			req->body = state.body;
+			req->state = xb_req_state_got_reply;
+			req->cb(req);
+		} else
+			kfree(req);
 	}
 
 	mutex_unlock(&xs_response_mutex);

include/xen/arm/hypercall.h

@@ -39,6 +39,8 @@
 #include <xen/interface/sched.h>
 #include <xen/interface/platform.h>
 
+struct xen_dm_op_buf;
+
 long privcmd_call(unsigned call, unsigned long a1,
 		unsigned long a2, unsigned long a3,
 		unsigned long a4, unsigned long a5);
@@ -53,7 +55,8 @@ int HYPERVISOR_physdev_op(int cmd, void *arg);
 int HYPERVISOR_vcpu_op(int cmd, int vcpuid, void *extra_args);
 int HYPERVISOR_tmem_op(void *arg);
 int HYPERVISOR_vm_assist(unsigned int cmd, unsigned int type);
-int HYPERVISOR_dm_op(domid_t domid, unsigned int nr_bufs, void *bufs);
+int HYPERVISOR_dm_op(domid_t domid, unsigned int nr_bufs,
+		     struct xen_dm_op_buf *bufs);
 int HYPERVISOR_platform_op_raw(void *arg);
 static inline int HYPERVISOR_platform_op(struct xen_platform_op *op)
 {

include/xen/events.h

@@ -58,6 +58,7 @@ void evtchn_put(unsigned int evtchn);
 
 void xen_send_IPI_one(unsigned int cpu, enum ipi_vector vector);
 void rebind_evtchn_irq(int evtchn, int irq);
+int xen_rebind_evtchn_to_cpu(int evtchn, unsigned tcpu);
 
 static inline void notify_remote_via_evtchn(int port)
 {

include/xen/interface/version.h

@@ -63,4 +63,19 @@ struct xen_feature_info {
 /* arg == xen_domain_handle_t. */
 #define XENVER_guest_handle 8
 
+#define XENVER_commandline 9
+struct xen_commandline {
+	char buf[1024];
+};
+
+/*
+ * Return value is the number of bytes written, or XEN_Exx on error.
+ * Calling with empty parameter returns the size of build_id.
+ */
+#define XENVER_build_id 10
+struct xen_build_id {
+	uint32_t	len; /* IN: size of buf[]. */
+	unsigned char	buf[];
+};
+
 #endif /* __XEN_PUBLIC_VERSION_H__ */

include/xen/xen-ops.h

@@ -15,6 +15,8 @@ static inline uint32_t xen_vcpu_nr(int cpu)
 	return per_cpu(xen_vcpu_id, cpu);
 }
 
+#define XEN_VCPU_ID_INVALID U32_MAX
+
 void xen_arch_pre_suspend(void);
 void xen_arch_post_suspend(int suspend_cancelled);