Merge tag 'stable/for-linus-3.6-rc0-tag' of...

Merge tag 'stable/for-linus-3.6-rc0-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/konrad/xen

Pull Xen update from Konrad Rzeszutek Wilk:
 "Features:
   * Performance improvement to lower the amount of traps the hypervisor
     has to do 32-bit guests.  Mainly for setting PTE entries and
     updating TLS descriptors.
   * MCE polling driver to collect hypervisor MCE buffer and present
     them to /dev/mcelog.
   * Physical CPU online/offline support.  When an privileged guest is
     booted it is present with virtual CPUs, which might have an 1:1 to
     physical CPUs but usually don't.  This provides mechanism to
     offline/online physical CPUs.
  Bug-fixes for:
   * Coverity found fixes in the console and ACPI processor driver.
   * PVonHVM kexec fixes along with some cleanups.
   * Pages that fall within E820 gaps and non-RAM regions (and had been
     released to hypervisor) would be populated back, but potentially in
     non-RAM regions."

* tag 'stable/for-linus-3.6-rc0-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/konrad/xen:
  xen: populate correct number of pages when across mem boundary (v2)
  xen PVonHVM: move shared_info to MMIO before kexec
  xen: simplify init_hvm_pv_info
  xen: remove cast from HYPERVISOR_shared_info assignment
  xen: enable platform-pci only in a Xen guest
  xen/pv-on-hvm kexec: shutdown watches from old kernel
  xen/x86: avoid updating TLS descriptors if they haven't changed
  xen/x86: add desc_equal() to compare GDT descriptors
  xen/mm: zero PTEs for non-present MFNs in the initial page table
  xen/mm: do direct hypercall in xen_set_pte() if batching is unavailable
  xen/hvc: Fix up checks when the info is allocated.
  xen/acpi: Fix potential memory leak.
  xen/mce: add .poll method for mcelog device driver
  xen/mce: schedule a workqueue to avoid sleep in atomic context
  xen/pcpu: Xen physical cpus online/offline sys interface
  xen/mce: Register native mce handler as vMCE bounce back point
  x86, MCE, AMD: Adjust initcall sequence for xen
  xen/mce: Add mcelog support for Xen platform

Documentation/ABI/testing/sysfs-devices-system-xen_cpu

+What:		/sys/devices/system/xen_cpu/
+Date:		May 2012
+Contact:	Liu, Jinsong <jinsong.liu@intel.com>
+Description:
+		A collection of global/individual Xen physical cpu attributes
+
+		Individual physical cpu attributes are contained in
+		subdirectories named by the Xen's logical cpu number, e.g.:
+		/sys/devices/system/xen_cpu/xen_cpu#/
+
+
+What:		/sys/devices/system/xen_cpu/xen_cpu#/online
+Date:		May 2012
+Contact:	Liu, Jinsong <jinsong.liu@intel.com>
+Description:
+		Interface to online/offline Xen physical cpus
+
+		When running under Xen platform, it provide user interface
+		to online/offline physical cpus, except cpu0 due to several
+		logic restrictions and assumptions.

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

@@ -48,6 +48,7 @@
 #include <xen/interface/sched.h>
 #include <xen/interface/physdev.h>
 #include <xen/interface/platform.h>
+#include <xen/interface/xen-mca.h>
 
 /*
  * The hypercall asms have to meet several constraints:
@@ -301,6 +302,13 @@ HYPERVISOR_set_timer_op(u64 timeout)
 	return _hypercall2(long, set_timer_op, timeout_lo, timeout_hi);
 }
 
+static inline int
+HYPERVISOR_mca(struct xen_mc *mc_op)
+{
+	mc_op->interface_version = XEN_MCA_INTERFACE_VERSION;
+	return _hypercall1(int, mca, mc_op);
+}
+
 static inline int
 HYPERVISOR_dom0_op(struct xen_platform_op *platform_op)
 {

arch/x86/kernel/cpu/mcheck/mce.c

@@ -60,8 +60,6 @@ static DEFINE_MUTEX(mce_chrdev_read_mutex);
 
 int mce_disabled __read_mostly;
 
-#define MISC_MCELOG_MINOR	227
-
 #define SPINUNIT 100	/* 100ns */
 
 atomic_t mce_entry;
@@ -2346,7 +2344,7 @@ static __init int mcheck_init_device(void)
 
 	return err;
 }
-device_initcall(mcheck_init_device);
+device_initcall_sync(mcheck_init_device);
 
 /*
  * Old style boot options parsing. Only for compatibility.

arch/x86/kernel/cpu/mcheck/mce_amd.c

@@ -759,4 +759,24 @@ static __init int threshold_init_device(void)
 
 	return 0;
 }
-device_initcall(threshold_init_device);
+/*
+ * there are 3 funcs which need to be _initcalled in a logic sequence:
+ * 1. xen_late_init_mcelog
+ * 2. mcheck_init_device
+ * 3. threshold_init_device
+ *
+ * xen_late_init_mcelog must register xen_mce_chrdev_device before
+ * native mce_chrdev_device registration if running under xen platform;
+ *
+ * mcheck_init_device should be inited before threshold_init_device to
+ * initialize mce_device, otherwise a NULL ptr dereference will cause panic.
+ *
+ * so we use following _initcalls
+ * 1. device_initcall(xen_late_init_mcelog);
+ * 2. device_initcall_sync(mcheck_init_device);
+ * 3. late_initcall(threshold_init_device);
+ *
+ * when running under xen, the initcall order is 1,2,3;
+ * on baremetal, we skip 1 and we do only 2 and 3.
+ */
+late_initcall(threshold_init_device);

arch/x86/xen/enlighten.c

@@ -31,6 +31,7 @@
 #include <linux/pci.h>
 #include <linux/gfp.h>
 #include <linux/memblock.h>
+#include <linux/syscore_ops.h>
 
 #include <xen/xen.h>
 #include <xen/interface/xen.h>
@@ -38,6 +39,7 @@
 #include <xen/interface/physdev.h>
 #include <xen/interface/vcpu.h>
 #include <xen/interface/memory.h>
+#include <xen/interface/xen-mca.h>
 #include <xen/features.h>
 #include <xen/page.h>
 #include <xen/hvm.h>
@@ -107,7 +109,7 @@ EXPORT_SYMBOL_GPL(xen_have_vector_callback);
  * Point at some empty memory to start with. We map the real shared_info
  * page as soon as fixmap is up and running.
  */
-struct shared_info *HYPERVISOR_shared_info = (void *)&xen_dummy_shared_info;
+struct shared_info *HYPERVISOR_shared_info = &xen_dummy_shared_info;
 
 /*
  * Flag to determine whether vcpu info placement is available on all
@@ -124,6 +126,19 @@ struct shared_info *HYPERVISOR_shared_info = (void *)&xen_dummy_shared_info;
  */
 static int have_vcpu_info_placement = 1;
 
+struct tls_descs {
+	struct desc_struct desc[3];
+};
+
+/*
+ * Updating the 3 TLS descriptors in the GDT on every task switch is
+ * surprisingly expensive so we avoid updating them if they haven't
+ * changed.  Since Xen writes different descriptors than the one
+ * passed in the update_descriptor hypercall we keep shadow copies to
+ * compare against.
+ */
+static DEFINE_PER_CPU(struct tls_descs, shadow_tls_desc);
+
 static void clamp_max_cpus(void)
 {
 #ifdef CONFIG_SMP
@@ -341,9 +356,7 @@ static void __init xen_init_cpuid_mask(void)
 	unsigned int xsave_mask;
 
 	cpuid_leaf1_edx_mask =
-		~((1 << X86_FEATURE_MCE)  |  /* disable MCE */
-		  (1 << X86_FEATURE_MCA)  |  /* disable MCA */
-		  (1 << X86_FEATURE_MTRR) |  /* disable MTRR */
+		~((1 << X86_FEATURE_MTRR) |  /* disable MTRR */
 		  (1 << X86_FEATURE_ACC));   /* thermal monitoring */
 
 	if (!xen_initial_domain())
@@ -540,12 +553,28 @@ static void __init xen_load_gdt_boot(const struct desc_ptr *dtr)
 		BUG();
 }
 
+static inline bool desc_equal(const struct desc_struct *d1,
+			      const struct desc_struct *d2)
+{
+	return d1->a == d2->a && d1->b == d2->b;
+}
+
 static void load_TLS_descriptor(struct thread_struct *t,
 				unsigned int cpu, unsigned int i)
 {
-	struct desc_struct *gdt = get_cpu_gdt_table(cpu);
-	xmaddr_t maddr = arbitrary_virt_to_machine(&gdt[GDT_ENTRY_TLS_MIN+i]);
-	struct multicall_space mc = __xen_mc_entry(0);
+	struct desc_struct *shadow = &per_cpu(shadow_tls_desc, cpu).desc[i];
+	struct desc_struct *gdt;
+	xmaddr_t maddr;
+	struct multicall_space mc;
+
+	if (desc_equal(shadow, &t->tls_array[i]))
+		return;
+
+	*shadow = t->tls_array[i];
+
+	gdt = get_cpu_gdt_table(cpu);
+	maddr = arbitrary_virt_to_machine(&gdt[GDT_ENTRY_TLS_MIN+i]);
+	mc = __xen_mc_entry(0);
 
 	MULTI_update_descriptor(mc.mc, maddr.maddr, t->tls_array[i]);
 }
@@ -627,8 +656,8 @@ static int cvt_gate_to_trap(int vector, const gate_desc *val,
 	/*
 	 * Look for known traps using IST, and substitute them
 	 * appropriately.  The debugger ones are the only ones we care
-	 * about.  Xen will handle faults like double_fault and
-	 * machine_check, so we should never see them.  Warn if
+	 * about.  Xen will handle faults like double_fault,
+	 * so we should never see them.  Warn if
 	 * there's an unexpected IST-using fault handler.
 	 */
 	if (addr == (unsigned long)debug)
@@ -643,7 +672,11 @@ static int cvt_gate_to_trap(int vector, const gate_desc *val,
 		return 0;
 #ifdef CONFIG_X86_MCE
 	} else if (addr == (unsigned long)machine_check) {
-		return 0;
+		/*
+		 * when xen hypervisor inject vMCE to guest,
+		 * use native mce handler to handle it
+		 */
+		;
 #endif
 	} else {
 		/* Some other trap using IST? */
@@ -1437,64 +1470,155 @@ asmlinkage void __init xen_start_kernel(void)
 #endif
 }
 
-static int init_hvm_pv_info(int *major, int *minor)
-{
-	uint32_t eax, ebx, ecx, edx, pages, msr, base;
-	u64 pfn;
-
-	base = xen_cpuid_base();
-	cpuid(base + 1, &eax, &ebx, &ecx, &edx);
-
-	*major = eax >> 16;
-	*minor = eax & 0xffff;
-	printk(KERN_INFO "Xen version %d.%d.\n", *major, *minor);
-
-	cpuid(base + 2, &pages, &msr, &ecx, &edx);
-
-	pfn = __pa(hypercall_page);
-	wrmsr_safe(msr, (u32)pfn, (u32)(pfn >> 32));
-
-	xen_setup_features();
-
-	pv_info.name = "Xen HVM";
-
-	xen_domain_type = XEN_HVM_DOMAIN;
+#ifdef CONFIG_XEN_PVHVM
+/*
+ * The pfn containing the shared_info is located somewhere in RAM. This
+ * will cause trouble if the current kernel is doing a kexec boot into a
+ * new kernel. The new kernel (and its startup code) can not know where
+ * the pfn is, so it can not reserve the page. The hypervisor will
+ * continue to update the pfn, and as a result memory corruption occours
+ * in the new kernel.
+ *
+ * One way to work around this issue is to allocate a page in the
+ * xen-platform pci device's BAR memory range. But pci init is done very
+ * late and the shared_info page is already in use very early to read
+ * the pvclock. So moving the pfn from RAM to MMIO is racy because some
+ * code paths on other vcpus could access the pfn during the small
+ * window when the old pfn is moved to the new pfn. There is even a
+ * small window were the old pfn is not backed by a mfn, and during that
+ * time all reads return -1.
+ *
+ * Because it is not known upfront where the MMIO region is located it
+ * can not be used right from the start in xen_hvm_init_shared_info.
+ *
+ * To minimise trouble the move of the pfn is done shortly before kexec.
+ * This does not eliminate the race because all vcpus are still online
+ * when the syscore_ops will be called. But hopefully there is no work
+ * pending at this point in time. Also the syscore_op is run last which
+ * reduces the risk further.
+ */
 
-	return 0;
-}
+static struct shared_info *xen_hvm_shared_info;
 
-void __ref xen_hvm_init_shared_info(void)
+static void xen_hvm_connect_shared_info(unsigned long pfn)
 {
-	int cpu;
 	struct xen_add_to_physmap xatp;
-	static struct shared_info *shared_info_page = 0;
 
-	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 = pfn;
 	if (HYPERVISOR_memory_op(XENMEM_add_to_physmap, &xatp))
 		BUG();
 
-	HYPERVISOR_shared_info = (struct shared_info *)shared_info_page;
+}
+static void xen_hvm_set_shared_info(struct shared_info *sip)
+{
+	int cpu;
+
+	HYPERVISOR_shared_info = sip;
 
 	/* 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
+	 * When xen_hvm_set_shared_info is run at boot time only vcpu 0 is
+	 * online but xen_hvm_set_shared_info is run at resume time too and
 	 * in that case multiple vcpus might be online. */
 	for_each_online_cpu(cpu) {
 		per_cpu(xen_vcpu, cpu) = &HYPERVISOR_shared_info->vcpu_info[cpu];
 	}
 }
 
-#ifdef CONFIG_XEN_PVHVM
+/* Reconnect the shared_info pfn to a mfn */
+void xen_hvm_resume_shared_info(void)
+{
+	xen_hvm_connect_shared_info(__pa(xen_hvm_shared_info) >> PAGE_SHIFT);
+}
+
+#ifdef CONFIG_KEXEC
+static struct shared_info *xen_hvm_shared_info_kexec;
+static unsigned long xen_hvm_shared_info_pfn_kexec;
+
+/* Remember a pfn in MMIO space for kexec reboot */
+void __devinit xen_hvm_prepare_kexec(struct shared_info *sip, unsigned long pfn)
+{
+	xen_hvm_shared_info_kexec = sip;
+	xen_hvm_shared_info_pfn_kexec = pfn;
+}
+
+static void xen_hvm_syscore_shutdown(void)
+{
+	struct xen_memory_reservation reservation = {
+		.domid = DOMID_SELF,
+		.nr_extents = 1,
+	};
+	unsigned long prev_pfn;
+	int rc;
+
+	if (!xen_hvm_shared_info_kexec)
+		return;
+
+	prev_pfn = __pa(xen_hvm_shared_info) >> PAGE_SHIFT;
+	set_xen_guest_handle(reservation.extent_start, &prev_pfn);
+
+	/* Move pfn to MMIO, disconnects previous pfn from mfn */
+	xen_hvm_connect_shared_info(xen_hvm_shared_info_pfn_kexec);
+
+	/* Update pointers, following hypercall is also a memory barrier */
+	xen_hvm_set_shared_info(xen_hvm_shared_info_kexec);
+
+	/* Allocate new mfn for previous pfn */
+	do {
+		rc = HYPERVISOR_memory_op(XENMEM_populate_physmap, &reservation);
+		if (rc == 0)
+			msleep(123);
+	} while (rc == 0);
+
+	/* Make sure the previous pfn is really connected to a (new) mfn */
+	BUG_ON(rc != 1);
+}
+
+static struct syscore_ops xen_hvm_syscore_ops = {
+	.shutdown = xen_hvm_syscore_shutdown,
+};
+#endif
+
+/* Use a pfn in RAM, may move to MMIO before kexec. */
+static void __init xen_hvm_init_shared_info(void)
+{
+	/* Remember pointer for resume */
+	xen_hvm_shared_info = extend_brk(PAGE_SIZE, PAGE_SIZE);
+	xen_hvm_connect_shared_info(__pa(xen_hvm_shared_info) >> PAGE_SHIFT);
+	xen_hvm_set_shared_info(xen_hvm_shared_info);
+}
+
+static void __init init_hvm_pv_info(void)
+{
+	int major, minor;
+	uint32_t eax, ebx, ecx, edx, pages, msr, base;
+	u64 pfn;
+
+	base = xen_cpuid_base();
+	cpuid(base + 1, &eax, &ebx, &ecx, &edx);
+
+	major = eax >> 16;
+	minor = eax & 0xffff;
+	printk(KERN_INFO "Xen version %d.%d.\n", major, minor);
+
+	cpuid(base + 2, &pages, &msr, &ecx, &edx);
+
+	pfn = __pa(hypercall_page);
+	wrmsr_safe(msr, (u32)pfn, (u32)(pfn >> 32));
+
+	xen_setup_features();
+
+	pv_info.name = "Xen HVM";
+
+	xen_domain_type = XEN_HVM_DOMAIN;
+}
+
 static int __cpuinit xen_hvm_cpu_notify(struct notifier_block *self,
 				    unsigned long action, void *hcpu)
 {
@@ -1517,14 +1641,12 @@ static struct notifier_block xen_hvm_cpu_notifier __cpuinitdata = {
 
 static void __init xen_hvm_guest_init(void)
 {
-	int r;
-	int major, minor;
-
-	r = init_hvm_pv_info(&major, &minor);
-	if (r < 0)
-		return;
+	init_hvm_pv_info();
 
 	xen_hvm_init_shared_info();
+#ifdef CONFIG_KEXEC
+	register_syscore_ops(&xen_hvm_syscore_ops);
+#endif
 
 	if (xen_feature(XENFEAT_hvm_callback_vector))
 		xen_have_vector_callback = 1;

arch/x86/xen/mmu.c

@@ -308,8 +308,20 @@ static bool xen_batched_set_pte(pte_t *ptep, pte_t pteval)
 
 static inline void __xen_set_pte(pte_t *ptep, pte_t pteval)
 {
-	if (!xen_batched_set_pte(ptep, pteval))
-		native_set_pte(ptep, pteval);
+	if (!xen_batched_set_pte(ptep, pteval)) {
+		/*
+		 * Could call native_set_pte() here and trap and
+		 * emulate the PTE write but with 32-bit guests this
+		 * needs two traps (one for each of the two 32-bit
+		 * words in the PTE) so do one hypercall directly
+		 * instead.
+		 */
+		struct mmu_update u;
+
+		u.ptr = virt_to_machine(ptep).maddr | MMU_NORMAL_PT_UPDATE;
+		u.val = pte_val_ma(pteval);
+		HYPERVISOR_mmu_update(&u, 1, NULL, DOMID_SELF);
+	}
 }
 
 static void xen_set_pte(pte_t *ptep, pte_t pteval)
@@ -1416,13 +1428,28 @@ static pte_t __init mask_rw_pte(pte_t *ptep, pte_t pte)
 }
 #endif /* CONFIG_X86_64 */
 
-/* Init-time set_pte while constructing initial pagetables, which
-   doesn't allow RO pagetable pages to be remapped RW */
+/*
+ * Init-time set_pte while constructing initial pagetables, which
+ * doesn't allow RO page table pages to be remapped RW.
+ *
+ * If there is no MFN for this PFN then this page is initially
+ * ballooned out so clear the PTE (as in decrease_reservation() in
+ * drivers/xen/balloon.c).
+ *
+ * Many of these PTE updates are done on unpinned and writable pages
+ * and doing a hypercall for these is unnecessary and expensive.  At
+ * this point it is not possible to tell if a page is pinned or not,
+ * so always write the PTE directly and rely on Xen trapping and
+ * emulating any updates as necessary.
+ */
 static void __init xen_set_pte_init(pte_t *ptep, pte_t pte)
 {
+	if (pte_mfn(pte) != INVALID_P2M_ENTRY)
 		pte = mask_rw_pte(ptep, pte);
+	else
+		pte = __pte_ma(0);
 
-	xen_set_pte(ptep, pte);
+	native_set_pte(ptep, pte);
 }
 
 static void pin_pagetable_pfn(unsigned cmd, unsigned long pfn)

arch/x86/xen/setup.c

@@ -157,25 +157,24 @@ static unsigned long __init xen_populate_chunk(
 	unsigned long dest_pfn;
 
 	for (i = 0, entry = list; i < map_size; i++, entry++) {
-		unsigned long credits = credits_left;
 		unsigned long s_pfn;
 		unsigned long e_pfn;
 		unsigned long pfns;
 		long capacity;
 
-		if (credits <= 0)
+		if (credits_left <= 0)
 			break;
 
 		if (entry->type != E820_RAM)
 			continue;
 
-		e_pfn = PFN_UP(entry->addr + entry->size);
+		e_pfn = PFN_DOWN(entry->addr + entry->size);
 
 		/* We only care about E820 after the xen_start_info->nr_pages */
 		if (e_pfn <= max_pfn)
 			continue;
 
-		s_pfn = PFN_DOWN(entry->addr);
+		s_pfn = PFN_UP(entry->addr);
 		/* If the E820 falls within the nr_pages, we want to start
 		 * at the nr_pages PFN.
 		 * If that would mean going past the E820 entry, skip it
@@ -184,23 +183,19 @@ static unsigned long __init xen_populate_chunk(
 			capacity = e_pfn - max_pfn;
 			dest_pfn = max_pfn;
 		} else {
-			/* last_pfn MUST be within E820_RAM regions */
-			if (*last_pfn && e_pfn >= *last_pfn)
-				s_pfn = *last_pfn;
 			capacity = e_pfn - s_pfn;
 			dest_pfn = s_pfn;
 		}
-		/* If we had filled this E820_RAM entry, go to the next one. */
-		if (capacity <= 0)
-			continue;
 
-		if (credits > capacity)
-			credits = capacity;
+		if (credits_left < capacity)
+			capacity = credits_left;
 
-		pfns = xen_do_chunk(dest_pfn, dest_pfn + credits, false);
+		pfns = xen_do_chunk(dest_pfn, dest_pfn + capacity, false);
 		done += pfns;
-		credits_left -= pfns;
 		*last_pfn = (dest_pfn + pfns);
+		if (pfns < capacity)
+			break;
+		credits_left -= pfns;
 	}
 	return done;
 }

arch/x86/xen/suspend.c

@@ -30,7 +30,7 @@ void xen_arch_hvm_post_suspend(int suspend_cancelled)
 {
 #ifdef CONFIG_XEN_PVHVM
 	int cpu;
-	xen_hvm_init_shared_info();
+	xen_hvm_resume_shared_info();
 	xen_callback_vector();
 	xen_unplug_emulated_devices();
 	if (xen_feature(XENFEAT_hvm_safe_pvclock)) {

arch/x86/xen/xen-ops.h

@@ -41,7 +41,7 @@ void xen_enable_syscall(void);
 void xen_vcpu_restore(void);
 
 void xen_callback_vector(void);
-void xen_hvm_init_shared_info(void);
+void xen_hvm_resume_shared_info(void);
 void xen_unplug_emulated_devices(void);
 
 void __init xen_build_dynamic_phys_to_machine(void);

drivers/tty/hvc/hvc_xen.c

@@ -209,11 +209,10 @@ static int xen_hvm_console_init(void)
 		info = kzalloc(sizeof(struct xencons_info), GFP_KERNEL | __GFP_ZERO);
 		if (!info)
 			return -ENOMEM;
-	}
-
+	} else if (info->intf != NULL) {
 		/* already configured */
-	if (info->intf != NULL)
 		return 0;
+	}
 	/*
 	 * If the toolstack (or the hypervisor) hasn't set these values, the
 	 * default value is 0. Even though mfn = 0 and evtchn = 0 are
@@ -259,12 +258,10 @@ static int xen_pv_console_init(void)
 		info = kzalloc(sizeof(struct xencons_info), GFP_KERNEL | __GFP_ZERO);
 		if (!info)
 			return -ENOMEM;
-	}
-
+	} else if (info->intf != NULL) {
 		/* already configured */
-	if (info->intf != NULL)
 		return 0;
-
+	}
 	info->evtchn = xen_start_info->console.domU.evtchn;
 	info->intf = mfn_to_virt(xen_start_info->console.domU.mfn);
 	info->vtermno = HVC_COOKIE;

drivers/xen/Kconfig

@@ -196,4 +196,12 @@ config XEN_ACPI_PROCESSOR
 	  called xen_acpi_processor  If you do not know what to choose, select
 	  M here. If the CPUFREQ drivers are built in, select Y here.
 
+config XEN_MCE_LOG
+	bool "Xen platform mcelog"
+	depends on XEN_DOM0 && X86_64 && X86_MCE
+	default n
+	help
+	  Allow kernel fetching MCE error from Xen platform and
+	  converting it into Linux mcelog format for mcelog tools
+
 endmenu

drivers/xen/Makefile

@@ -17,7 +17,9 @@ obj-$(CONFIG_XEN_SYS_HYPERVISOR)	+= sys-hypervisor.o
 obj-$(CONFIG_XEN_PVHVM)			+= platform-pci.o
 obj-$(CONFIG_XEN_TMEM)			+= tmem.o
 obj-$(CONFIG_SWIOTLB_XEN)		+= swiotlb-xen.o
+obj-$(CONFIG_XEN_DOM0)			+= pcpu.o
 obj-$(CONFIG_XEN_DOM0)			+= pci.o acpi.o
+obj-$(CONFIG_XEN_MCE_LOG)		+= mcelog.o
 obj-$(CONFIG_XEN_PCIDEV_BACKEND)	+= xen-pciback/
 obj-$(CONFIG_XEN_PRIVCMD)		+= xen-privcmd.o
 obj-$(CONFIG_XEN_ACPI_PROCESSOR)	+= xen-acpi-processor.o

drivers/xen/pcpu.c

+/******************************************************************************
+ * pcpu.c
+ * Management physical cpu in dom0, get pcpu info and provide sys interface
+ *
+ * Copyright (c) 2012 Intel Corporation
+ * Author: Liu, Jinsong <jinsong.liu@intel.com>
+ * Author: Jiang, Yunhong <yunhong.jiang@intel.com>
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License version 2
+ * as published by the Free Software Foundation; or, when distributed
+ * separately from the Linux kernel or incorporated into other
+ * software packages, subject to the following license:
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this source file (the "Software"), to deal in the Software without
+ * restriction, including without limitation the rights to use, copy, modify,
+ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
+ * and to permit persons to whom the Software is furnished to do so, subject to
+ * the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
+ * IN THE SOFTWARE.
+ */
+
+#include <linux/interrupt.h>
+#include <linux/spinlock.h>
+#include <linux/cpu.h>
+#include <linux/stat.h>
+#include <linux/capability.h>
+
+#include <xen/xen.h>
+#include <xen/xenbus.h>
+#include <xen/events.h>
+#include <xen/interface/platform.h>
+#include <asm/xen/hypervisor.h>
+#include <asm/xen/hypercall.h>
+
+#define XEN_PCPU "xen_cpu: "
+
+/*
+ * @cpu_id: Xen physical cpu logic number
+ * @flags: Xen physical cpu status flag
+ * - XEN_PCPU_FLAGS_ONLINE: cpu is online
+ * - XEN_PCPU_FLAGS_INVALID: cpu is not present
+ */
+struct pcpu {
+	struct list_head list;
+	struct device dev;
+	uint32_t cpu_id;
+	uint32_t flags;
+};
+
+static struct bus_type xen_pcpu_subsys = {
+	.name = "xen_cpu",
+	.dev_name = "xen_cpu",
+};
+
+static DEFINE_MUTEX(xen_pcpu_lock);
+
+static LIST_HEAD(xen_pcpus);
+
+static int xen_pcpu_down(uint32_t cpu_id)
+{
+	struct xen_platform_op op = {
+		.cmd			= XENPF_cpu_offline,
+		.interface_version	= XENPF_INTERFACE_VERSION,
+		.u.cpu_ol.cpuid		= cpu_id,
+	};
+
+	return HYPERVISOR_dom0_op(&op);
+}
+
+static int xen_pcpu_up(uint32_t cpu_id)
+{
+	struct xen_platform_op op = {
+		.cmd			= XENPF_cpu_online,
+		.interface_version	= XENPF_INTERFACE_VERSION,
+		.u.cpu_ol.cpuid		= cpu_id,
+	};
+
+	return HYPERVISOR_dom0_op(&op);
+}
+
+static ssize_t show_online(struct device *dev,
+			   struct device_attribute *attr,
+			   char *buf)
+{
+	struct pcpu *cpu = container_of(dev, struct pcpu, dev);
+
+	return sprintf(buf, "%u\n", !!(cpu->flags & XEN_PCPU_FLAGS_ONLINE));
+}
+
+static ssize_t __ref store_online(struct device *dev,
+				  struct device_attribute *attr,
+				  const char *buf, size_t count)
+{
+	struct pcpu *pcpu = container_of(dev, struct pcpu, dev);
+	unsigned long long val;
+	ssize_t ret;
+
+	if (!capable(CAP_SYS_ADMIN))
+		return -EPERM;
+
+	if (kstrtoull(buf, 0, &val) < 0)
+		return -EINVAL;
+
+	switch (val) {
+	case 0:
+		ret = xen_pcpu_down(pcpu->cpu_id);
+		break;
+	case 1:
+		ret = xen_pcpu_up(pcpu->cpu_id);
+		break;
+	default:
+		ret = -EINVAL;
+	}
+
+	if (ret >= 0)
+		ret = count;
+	return ret;
+}
+static DEVICE_ATTR(online, S_IRUGO | S_IWUSR, show_online, store_online);
+
+static bool xen_pcpu_online(uint32_t flags)
+{
+	return !!(flags & XEN_PCPU_FLAGS_ONLINE);
+}
+
+static void pcpu_online_status(struct xenpf_pcpuinfo *info,
+			       struct pcpu *pcpu)
+{
+	if (xen_pcpu_online(info->flags) &&
+	   !xen_pcpu_online(pcpu->flags)) {
+		/* the pcpu is onlined */
+		pcpu->flags |= XEN_PCPU_FLAGS_ONLINE;
+		kobject_uevent(&pcpu->dev.kobj, KOBJ_ONLINE);
+	} else if (!xen_pcpu_online(info->flags) &&
+		    xen_pcpu_online(pcpu->flags)) {
+		/* The pcpu is offlined */
+		pcpu->flags &= ~XEN_PCPU_FLAGS_ONLINE;
+		kobject_uevent(&pcpu->dev.kobj, KOBJ_OFFLINE);
+	}
+}
+
+static struct pcpu *get_pcpu(uint32_t cpu_id)
+{
+	struct pcpu *pcpu;
+
+	list_for_each_entry(pcpu, &xen_pcpus, list) {
+		if (pcpu->cpu_id == cpu_id)
+			return pcpu;
+	}
+
+	return NULL;
+}
+
+static void pcpu_release(struct device *dev)
+{
+	struct pcpu *pcpu = container_of(dev, struct pcpu, dev);
+
+	list_del(&pcpu->list);
+	kfree(pcpu);
+}
+
+static void unregister_and_remove_pcpu(struct pcpu *pcpu)
+{
+	struct device *dev;
+
+	if (!pcpu)
+		return;
+
+	dev = &pcpu->dev;
+	if (dev->id)
+		device_remove_file(dev, &dev_attr_online);
+
+	/* pcpu remove would be implicitly done */
+	device_unregister(dev);
+}
+
+static int register_pcpu(struct pcpu *pcpu)
+{
+	struct device *dev;
+	int err = -EINVAL;
+
+	if (!pcpu)
+		return err;
+
+	dev = &pcpu->dev;
+	dev->bus = &xen_pcpu_subsys;
+	dev->id = pcpu->cpu_id;
+	dev->release = pcpu_release;
+
+	err = device_register(dev);
+	if (err) {
+		pcpu_release(dev);
+		return err;
+	}
+
+	/*
+	 * Xen never offline cpu0 due to several restrictions
+	 * and assumptions. This basically doesn't add a sys control
+	 * to user, one cannot attempt to offline BSP.
+	 */
+	if (dev->id) {
+		err = device_create_file(dev, &dev_attr_online);
+		if (err) {
+			device_unregister(dev);
+			return err;
+		}
+	}
+
+	return 0;
+}
+
+static struct pcpu *create_and_register_pcpu(struct xenpf_pcpuinfo *info)
+{
+	struct pcpu *pcpu;
+	int err;
+
+	if (info->flags & XEN_PCPU_FLAGS_INVALID)
+		return ERR_PTR(-ENODEV);
+
+	pcpu = kzalloc(sizeof(struct pcpu), GFP_KERNEL);
+	if (!pcpu)
+		return ERR_PTR(-ENOMEM);
+
+	INIT_LIST_HEAD(&pcpu->list);
+	pcpu->cpu_id = info->xen_cpuid;
+	pcpu->flags = info->flags;
+
+	/* Need hold on xen_pcpu_lock before pcpu list manipulations */
+	list_add_tail(&pcpu->list, &xen_pcpus);
+
+	err = register_pcpu(pcpu);
+	if (err) {
+		pr_warning(XEN_PCPU "Failed to register pcpu%u\n",
+			   info->xen_cpuid);
+		return ERR_PTR(-ENOENT);
+	}
+
+	return pcpu;
+}
+
+/*
+ * Caller should hold the xen_pcpu_lock
+ */
+static int sync_pcpu(uint32_t cpu, uint32_t *max_cpu)
+{
+	int ret;
+	struct pcpu *pcpu = NULL;
+	struct xenpf_pcpuinfo *info;
+	struct xen_platform_op op = {
+		.cmd                   = XENPF_get_cpuinfo,
+		.interface_version     = XENPF_INTERFACE_VERSION,
+		.u.pcpu_info.xen_cpuid = cpu,
+	};
+
+	ret = HYPERVISOR_dom0_op(&op);
+	if (ret)
+		return ret;
+
+	info = &op.u.pcpu_info;
+	if (max_cpu)
+		*max_cpu = info->max_present;
+
+	pcpu = get_pcpu(cpu);
+
+	/*
+	 * Only those at cpu present map has its sys interface.
+	 */
+	if (info->flags & XEN_PCPU_FLAGS_INVALID) {
+		if (pcpu)
+			unregister_and_remove_pcpu(pcpu);
+		return 0;
+	}
+
+	if (!pcpu) {
+		pcpu = create_and_register_pcpu(info);
+		if (IS_ERR_OR_NULL(pcpu))
+			return -ENODEV;
+	} else
+		pcpu_online_status(info, pcpu);
+
+	return 0;
+}
+
+/*
+ * Sync dom0's pcpu information with xen hypervisor's
+ */
+static int xen_sync_pcpus(void)
+{
+	/*
+	 * Boot cpu always have cpu_id 0 in xen
+	 */
+	uint32_t cpu = 0, max_cpu = 0;
+	int err = 0;
+	struct pcpu *pcpu, *tmp;
+
+	mutex_lock(&xen_pcpu_lock);
+
+	while (!err && (cpu <= max_cpu)) {
+		err = sync_pcpu(cpu, &max_cpu);
+		cpu++;
+	}
+
+	if (err)
+		list_for_each_entry_safe(pcpu, tmp, &xen_pcpus, list)
+			unregister_and_remove_pcpu(pcpu);
+
+	mutex_unlock(&xen_pcpu_lock);
+
+	return err;
+}
+
+static void xen_pcpu_work_fn(struct work_struct *work)
+{
+	xen_sync_pcpus();
+}
+static DECLARE_WORK(xen_pcpu_work, xen_pcpu_work_fn);
+
+static irqreturn_t xen_pcpu_interrupt(int irq, void *dev_id)
+{
+	schedule_work(&xen_pcpu_work);
+	return IRQ_HANDLED;
+}
+
+static int __init xen_pcpu_init(void)
+{
+	int irq, ret;
+
+	if (!xen_initial_domain())
+		return -ENODEV;
+
+	irq = bind_virq_to_irqhandler(VIRQ_PCPU_STATE, 0,
+				      xen_pcpu_interrupt, 0,
+				      "xen-pcpu", NULL);
+	if (irq < 0) {
+		pr_warning(XEN_PCPU "Failed to bind pcpu virq\n");
+		return irq;
+	}
+
+	ret = subsys_system_register(&xen_pcpu_subsys, NULL);
+	if (ret) {
+		pr_warning(XEN_PCPU "Failed to register pcpu subsys\n");
+		goto err1;
+	}
+
+	ret = xen_sync_pcpus();
+	if (ret) {
+		pr_warning(XEN_PCPU "Failed to sync pcpu info\n");
+		goto err2;
+	}
+
+	return 0;
+
+err2:
+	bus_unregister(&xen_pcpu_subsys);
+err1:
+	unbind_from_irqhandler(irq, NULL);
+	return ret;
+}
+arch_initcall(xen_pcpu_init);

drivers/xen/platform-pci.c

@@ -101,6 +101,19 @@ static int platform_pci_resume(struct pci_dev *pdev)
 	return 0;
 }
 
+static void __devinit prepare_shared_info(void)
+{
+#ifdef CONFIG_KEXEC
+	unsigned long addr;
+	struct shared_info *hvm_shared_info;
+
+	addr = alloc_xen_mmio(PAGE_SIZE);
+	hvm_shared_info = ioremap(addr, PAGE_SIZE);
+	memset(hvm_shared_info, 0, PAGE_SIZE);
+	xen_hvm_prepare_kexec(hvm_shared_info, addr >> PAGE_SHIFT);
+#endif
+}
+
 static int __devinit platform_pci_init(struct pci_dev *pdev,
 				       const struct pci_device_id *ent)
 {
@@ -109,6 +122,9 @@ static int __devinit platform_pci_init(struct pci_dev *pdev,
 	long mmio_addr, mmio_len;
 	unsigned int max_nr_gframes;
 
+	if (!xen_domain())
+		return -ENODEV;
+
 	i = pci_enable_device(pdev);
 	if (i)
 		return i;
@@ -135,6 +151,8 @@ static int __devinit platform_pci_init(struct pci_dev *pdev,
 	platform_mmio = mmio_addr;
 	platform_mmiolen = mmio_len;
 
+	prepare_shared_info();
+
 	if (!xen_have_vector_callback) {
 		ret = xen_allocate_irq(pdev);
 		if (ret) {

drivers/xen/xen-acpi-processor.c

@@ -520,15 +520,18 @@ static int __init xen_acpi_processor_init(void)
 
 		if (!pr_backup) {
 			pr_backup = kzalloc(sizeof(struct acpi_processor), GFP_KERNEL);
+			if (pr_backup)
 				memcpy(pr_backup, _pr, sizeof(struct acpi_processor));
 		}
 		(void)upload_pm_data(_pr);
 	}
 	rc = check_acpi_ids(pr_backup);
-	if (rc)
-		goto err_unregister;
 
 	kfree(pr_backup);
+	pr_backup = NULL;
+
+	if (rc)
+		goto err_unregister;
 
 	return 0;
 err_unregister:

drivers/xen/xenbus/xenbus_xs.c

@@ -618,6 +618,23 @@ static struct xenbus_watch *find_watch(const char *token)
 	return NULL;
 }
 
+static void xs_reset_watches(void)
+{
+	int err, supported = 0;
+
+	if (!xen_hvm_domain())
+		return;
+
+	err = xenbus_scanf(XBT_NIL, "control",
+			"platform-feature-xs_reset_watches", "%d", &supported);
+	if (err != 1 || !supported)
+		return;
+
+	err = xs_error(xs_single(XBT_NIL, XS_RESET_WATCHES, "", NULL));
+	if (err && err != -EEXIST)
+		printk(KERN_WARNING "xs_reset_watches failed: %d\n", err);
+}
+
 /* Register callback to watch this node. */
 int register_xenbus_watch(struct xenbus_watch *watch)
 {
@@ -900,5 +917,8 @@ int xs_init(void)
 	if (IS_ERR(task))
 		return PTR_ERR(task);
 
+	/* shutdown watches for kexec boot */
+	xs_reset_watches();
+
 	return 0;
 }

include/linux/miscdevice.h

@@ -35,6 +35,7 @@
 #define MPT_MINOR		220
 #define MPT2SAS_MINOR		221
 #define UINPUT_MINOR		223
+#define MISC_MCELOG_MINOR	227
 #define HPET_MINOR		228
 #define FUSE_MINOR		229
 #define KVM_MINOR		232

include/xen/events.h

@@ -58,6 +58,8 @@ void notify_remote_via_irq(int irq);
 
 void xen_irq_resume(void);
 
+void xen_hvm_prepare_kexec(struct shared_info *sip, unsigned long pfn);
+
 /* Clear an irq's pending state, in preparation for polling on it */
 void xen_clear_irq_pending(int irq);
 void xen_set_irq_pending(int irq);

include/xen/interface/io/xs_wire.h

@@ -29,7 +29,8 @@ enum xsd_sockmsg_type
     XS_IS_DOMAIN_INTRODUCED,
     XS_RESUME,
     XS_SET_TARGET,
-    XS_RESTRICT
+    XS_RESTRICT,
+    XS_RESET_WATCHES,
 };
 
 #define XS_WRITE_NONE "NONE"

include/xen/interface/platform.h

@@ -314,6 +314,13 @@ struct xenpf_pcpuinfo {
 };
 DEFINE_GUEST_HANDLE_STRUCT(xenpf_pcpuinfo);
 
+#define XENPF_cpu_online	56
+#define XENPF_cpu_offline	57
+struct xenpf_cpu_ol {
+	uint32_t cpuid;
+};
+DEFINE_GUEST_HANDLE_STRUCT(xenpf_cpu_ol);
+
 struct xen_platform_op {
 	uint32_t cmd;
 	uint32_t interface_version; /* XENPF_INTERFACE_VERSION */
@@ -330,6 +337,7 @@ struct xen_platform_op {
 		struct xenpf_getidletime       getidletime;
 		struct xenpf_set_processor_pminfo set_pminfo;
 		struct xenpf_pcpuinfo          pcpu_info;
+		struct xenpf_cpu_ol            cpu_ol;
 		uint8_t                        pad[128];
 	} u;
 };

include/xen/interface/xen.h

@@ -80,6 +80,7 @@
 #define VIRQ_CONSOLE    2  /* (DOM0) Bytes received on emergency console. */
 #define VIRQ_DOM_EXC    3  /* (DOM0) Exceptional event for some domain.   */
 #define VIRQ_DEBUGGER   6  /* (DOM0) A domain has paused for debugging.   */
+#define VIRQ_PCPU_STATE 9  /* (DOM0) PCPU state changed                   */
 
 /* Architecture-specific VIRQ definitions. */
 #define VIRQ_ARCH_0    16