Merge branch 'perf-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip

Pull perf updates from Ingo Molnar:
 "Main kernel side changes:

   - Big reorganization of the x86 perf support code.  The old code grew
     organically deep inside arch/x86/kernel/cpu/perf* and its naming
     became somewhat messy.

     The new location is under arch/x86/events/, using the following
     cleaner hierarchy of source code files:

       perf/x86: Move perf_event.c .................. => x86/events/core.c
       perf/x86: Move perf_event_amd.c .............. => x86/events/amd/core.c
       perf/x86: Move perf_event_amd_ibs.c .......... => x86/events/amd/ibs.c
       perf/x86: Move perf_event_amd_iommu.[ch] ..... => x86/events/amd/iommu.[ch]
       perf/x86: Move perf_event_amd_uncore.c ....... => x86/events/amd/uncore.c
       perf/x86: Move perf_event_intel_bts.c ........ => x86/events/intel/bts.c
       perf/x86: Move perf_event_intel.c ............ => x86/events/intel/core.c
       perf/x86: Move perf_event_intel_cqm.c ........ => x86/events/intel/cqm.c
       perf/x86: Move perf_event_intel_cstate.c ..... => x86/events/intel/cstate.c
       perf/x86: Move perf_event_intel_ds.c ......... => x86/events/intel/ds.c
       perf/x86: Move perf_event_intel_lbr.c ........ => x86/events/intel/lbr.c
       perf/x86: Move perf_event_intel_pt.[ch] ...... => x86/events/intel/pt.[ch]
       perf/x86: Move perf_event_intel_rapl.c ....... => x86/events/intel/rapl.c
       perf/x86: Move perf_event_intel_uncore.[ch] .. => x86/events/intel/uncore.[ch]
       perf/x86: Move perf_event_intel_uncore_nhmex.c => x86/events/intel/uncore_nmhex.c
       perf/x86: Move perf_event_intel_uncore_snb.c   => x86/events/intel/uncore_snb.c
       perf/x86: Move perf_event_intel_uncore_snbep.c => x86/events/intel/uncore_snbep.c
       perf/x86: Move perf_event_knc.c .............. => x86/events/intel/knc.c
       perf/x86: Move perf_event_p4.c ............... => x86/events/intel/p4.c
       perf/x86: Move perf_event_p6.c ............... => x86/events/intel/p6.c
       perf/x86: Move perf_event_msr.c .............. => x86/events/msr.c

     (Borislav Petkov)

   - Update various x86 PMU constraint and hw support details (Stephane
     Eranian)

   - Optimize kprobes for BPF execution (Martin KaFai Lau)

   - Rewrite, refactor and fix the Intel uncore PMU driver code (Thomas
     Gleixner)

   - Rewrite, refactor and fix the Intel RAPL PMU code (Thomas Gleixner)

   - Various fixes and smaller cleanups.

  There are lots of perf tooling updates as well.  A few highlights:

  perf report/top:

     - Hierarchy histogram mode for 'perf top' and 'perf report',
       showing multiple levels, one per --sort entry: (Namhyung Kim)

       On a mostly idle system:

         # perf top --hierarchy -s comm,dso

       Then expand some levels and use 'P' to take a snapshot:

         # cat perf.hist.0
         -  92.32%         perf
               58.20%         perf
               22.29%         libc-2.22.so
                5.97%         [kernel]
                4.18%         libelf-0.165.so
                1.69%         [unknown]
         -   4.71%         qemu-system-x86
                3.10%         [kernel]
                1.60%         qemu-system-x86_64 (deleted)
         +   2.97%         swapper
         #

     - Add 'L' hotkey to dynamicly set the percent threshold for
       histogram entries and callchains, i.e.  dynamicly do what the
       --percent-limit command line option to 'top' and 'report' does.
       (Namhyung Kim)

  perf mem:

     - Allow specifying events via -e in 'perf mem record', also listing
       what events can be specified via 'perf mem record -e list' (Jiri
       Olsa)

  perf record:

     - Add 'perf record' --all-user/--all-kernel options, so that one
       can tell that all the events in the command line should be
       restricted to the user or kernel levels (Jiri Olsa), i.e.:

         perf record -e cycles:u,instructions:u

       is equivalent to:

         perf record --all-user -e cycles,instructions

     - Make 'perf record' collect CPU cache info in the perf.data file header:

         $ perf record usleep 1
         [ perf record: Woken up 1 times to write data ]
         [ perf record: Captured and wrote 0.017 MB perf.data (7 samples) ]
         $ perf report --header-only -I | tail -10 | head -8
         # CPU cache info:
         #  L1 Data                 32K [0-1]
         #  L1 Instruction          32K [0-1]
         #  L1 Data                 32K [2-3]
         #  L1 Instruction          32K [2-3]
         #  L2 Unified             256K [0-1]
         #  L2 Unified             256K [2-3]
         #  L3 Unified            4096K [0-3]

       Will be used in 'perf c2c' and eventually in 'perf diff' to
       allow, for instance running the same workload in multiple
       machines and then when using 'diff' show the hardware difference.
       (Jiri Olsa)

     - Improved support for Java, using the JVMTI agent library to do
       jitdumps that then will be inserted in synthesized
       PERF_RECORD_MMAP2 events via 'perf inject' pointed to synthesized
       ELF files stored in ~/.debug and keyed with build-ids, to allow
       symbol resolution and even annotation with source line info, see
       the changeset comments to see how to use it (Stephane Eranian)

  perf script/trace:

     - Decode data_src values (e.g.  perf.data files generated by 'perf
       mem record') in 'perf script': (Jiri Olsa)

         # perf script
           perf 693 [1] 4.088652: 1 cpu/mem-loads,ldlat=30/P: ffff88007d0b0f40 68100142 L1 hit|SNP None|TLB L1 or L2 hit|LCK No <SNIP>
                                                                              ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
     - Improve support to 'data_src', 'weight' and 'addr' fields in
       'perf script' (Jiri Olsa)

     - Handle empty print fmts in 'perf script -s' i.e. when running
       python or perl scripts (Taeung Song)

  perf stat:

     - 'perf stat' now shows shadow metrics (insn per cycle, etc) in
       interval mode too.  E.g:

         # perf stat -I 1000 -e instructions,cycles sleep 1
         #         time   counts unit events
            1.000215928  519,620      instructions     #  0.69 insn per cycle
            1.000215928  752,003      cycles
         <SNIP>

     - Port 'perf kvm stat' to PowerPC (Hemant Kumar)

     - Implement CSV metrics output in 'perf stat' (Andi Kleen)

  perf BPF support:

     - Support converting data from bpf events in 'perf data' (Wang Nan)

     - Print bpf-output events in 'perf script': (Wang Nan).

         # perf record -e bpf-output/no-inherit,name=evt/ -e ./test_bpf_output_3.c/map:channel.event=evt/ usleep 1000
         # perf script
            usleep  4882 21384.532523:   evt:  ffffffff810e97d1 sys_nanosleep ([kernel.kallsyms])
             BPF output: 0000: 52 61 69 73 65 20 61 20  Raise a
                         0008: 42 50 46 20 65 76 65 6e  BPF even
                         0010: 74 21 00 00              t!..
             BPF string: "Raise a BPF event!"
         #

     - Add API to set values of map entries in a BPF object, be it
       individual map slots or ranges (Wang Nan)

     - Introduce support for the 'bpf-output' event (Wang Nan)

     - Add glue to read perf events in a BPF program (Wang Nan)

     - Improve support for bpf-output events in 'perf trace' (Wang Nan)

  ... and tons of other changes as well - see the shortlog and git log
  for details!"

* 'perf-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (342 commits)
  perf stat: Add --metric-only support for -A
  perf stat: Implement --metric-only mode
  perf stat: Document CSV format in manpage
  perf hists browser: Check sort keys before hot key actions
  perf hists browser: Allow thread filtering for comm sort key
  perf tools: Add sort__has_comm variable
  perf tools: Recalc total periods using top-level entries in hierarchy
  perf tools: Remove nr_sort_keys field
  perf hists browser: Cleanup hist_browser__fprintf_hierarchy_entry()
  perf tools: Remove hist_entry->fmt field
  perf tools: Fix command line filters in hierarchy mode
  perf tools: Add more sort entry check functions
  perf tools: Fix hist_entry__filter() for hierarchy
  perf jitdump: Build only on supported archs
  tools lib traceevent: Add '~' operation within arg_num_eval()
  perf tools: Omit unnecessary cast in perf_pmu__parse_scale
  perf tools: Pass perf_hpp_list all the way through setup_sort_list
  perf tools: Fix perf script python database export crash
  perf jitdump: DWARF is also needed
  perf bench mem: Prepare the x86-64 build for upstream memcpy_mcsafe() changes
  ...

62 files changed, 1047 insertions, 804 deletions

diff --git a/arch/x86/Kbuild b/arch/x86/Kbuild
index 1538562..eb3abf8 100644
--- a/arch/x86/Kbuild
+++ b/arch/x86/Kbuild
@@ -1,6 +1,7 @@
-
 obj-y += entry/
 
+obj-$(CONFIG_PERF_EVENTS) += events/
+
 obj-$(CONFIG_KVM) += kvm/
 
 # Xen paravirtualization support
diff --git a/arch/x86/events/Makefile b/arch/x86/events/Makefile
new file mode 100644
index 0000000..fdfea15
--- /dev/null
+++ b/arch/x86/events/Makefile
@@ -0,0 +1,13 @@
+obj-y			+= core.o
+
+obj-$(CONFIG_CPU_SUP_AMD)               += amd/core.o amd/uncore.o
+obj-$(CONFIG_X86_LOCAL_APIC)            += amd/ibs.o msr.o
+ifdef CONFIG_AMD_IOMMU
+obj-$(CONFIG_CPU_SUP_AMD)               += amd/iommu.o
+endif
+obj-$(CONFIG_CPU_SUP_INTEL)		+= intel/core.o intel/bts.o intel/cqm.o
+obj-$(CONFIG_CPU_SUP_INTEL)		+= intel/cstate.o intel/ds.o intel/knc.o 
+obj-$(CONFIG_CPU_SUP_INTEL)		+= intel/lbr.o intel/p4.o intel/p6.o intel/pt.o
+obj-$(CONFIG_CPU_SUP_INTEL)		+= intel/rapl.o msr.o
+obj-$(CONFIG_PERF_EVENTS_INTEL_UNCORE)	+= intel/uncore.o intel/uncore_nhmex.o
+obj-$(CONFIG_PERF_EVENTS_INTEL_UNCORE)	+= intel/uncore_snb.o intel/uncore_snbep.o
diff --git a/arch/x86/kernel/cpu/perf_event_amd.c b/arch/x86/events/amd/core.c
index 5861053..049ada8d 100644
--- a/arch/x86/kernel/cpu/perf_event_amd.c
+++ b/arch/x86/events/amd/core.c
@@ -5,7 +5,7 @@
 #include <linux/slab.h>
 #include <asm/apicdef.h>
 
-#include "perf_event.h"
+#include "../perf_event.h"
 
 static __initconst const u64 amd_hw_cache_event_ids
 				[PERF_COUNT_HW_CACHE_MAX]
diff --git a/arch/x86/kernel/cpu/perf_event_amd_ibs.c b/arch/x86/events/amd/ibs.c
index 989d3c2..51087c2 100644
--- a/arch/x86/kernel/cpu/perf_event_amd_ibs.c
+++ b/arch/x86/events/amd/ibs.c
@@ -14,7 +14,7 @@
 
 #include <asm/apic.h>
 
-#include "perf_event.h"
+#include "../perf_event.h"
 
 static u32 ibs_caps;
 
@@ -670,7 +670,7 @@ static __init int perf_event_ibs_init(void)
 	perf_ibs_pmu_init(&perf_ibs_op, "ibs_op");
 
 	register_nmi_handler(NMI_LOCAL, perf_ibs_nmi_handler, 0, "perf_ibs");
-	printk(KERN_INFO "perf: AMD IBS detected (0x%08x)\n", ibs_caps);
+	pr_info("perf: AMD IBS detected (0x%08x)\n", ibs_caps);
 
 	return 0;
 }
@@ -774,14 +774,14 @@ static int setup_ibs_ctl(int ibs_eilvt_off)
 		pci_read_config_dword(cpu_cfg, IBSCTL, &value);
 		if (value != (ibs_eilvt_off | IBSCTL_LVT_OFFSET_VALID)) {
 			pci_dev_put(cpu_cfg);
-			printk(KERN_DEBUG "Failed to setup IBS LVT offset, "
-			       "IBSCTL = 0x%08x\n", value);
+			pr_debug("Failed to setup IBS LVT offset, IBSCTL = 0x%08x\n",
+				 value);
 			return -EINVAL;
 		}
 	} while (1);
 
 	if (!nodes) {
-		printk(KERN_DEBUG "No CPU node configured for IBS\n");
+		pr_debug("No CPU node configured for IBS\n");
 		return -ENODEV;
 	}
 
@@ -810,7 +810,7 @@ static void force_ibs_eilvt_setup(void)
 	preempt_enable();
 
 	if (offset == APIC_EILVT_NR_MAX) {
-		printk(KERN_DEBUG "No EILVT entry available\n");
+		pr_debug("No EILVT entry available\n");
 		return;
 	}
 
diff --git a/arch/x86/kernel/cpu/perf_event_amd_iommu.c b/arch/x86/events/amd/iommu.c
index 97242a9..635e5eb 100644
--- a/arch/x86/kernel/cpu/perf_event_amd_iommu.c
+++ b/arch/x86/events/amd/iommu.c
@@ -16,8 +16,8 @@
 #include <linux/cpumask.h>
 #include <linux/slab.h>
 
-#include "perf_event.h"
-#include "perf_event_amd_iommu.h"
+#include "../perf_event.h"
+#include "iommu.h"
 
 #define COUNTER_SHIFT		16
 
diff --git a/arch/x86/kernel/cpu/perf_event_amd_iommu.h b/arch/x86/events/amd/iommu.h
index 845d173..845d173 100644
--- a/arch/x86/kernel/cpu/perf_event_amd_iommu.h
+++ b/arch/x86/events/amd/iommu.h
diff --git a/arch/x86/kernel/cpu/perf_event_amd_uncore.c b/arch/x86/events/amd/uncore.c
index 8836fc9..3db9569 100644
--- a/arch/x86/kernel/cpu/perf_event_amd_uncore.c
+++ b/arch/x86/events/amd/uncore.c
@@ -538,7 +538,7 @@ static int __init amd_uncore_init(void)
 		if (ret)
 			goto fail_nb;
 
-		printk(KERN_INFO "perf: AMD NB counters detected\n");
+		pr_info("perf: AMD NB counters detected\n");
 		ret = 0;
 	}
 
@@ -552,7 +552,7 @@ static int __init amd_uncore_init(void)
 		if (ret)
 			goto fail_l2;
 
-		printk(KERN_INFO "perf: AMD L2I counters detected\n");
+		pr_info("perf: AMD L2I counters detected\n");
 		ret = 0;
 	}
 
diff --git a/arch/x86/kernel/cpu/perf_event.c b/arch/x86/events/core.c
index 1b443db..5e830d0 100644
--- a/arch/x86/kernel/cpu/perf_event.c
+++ b/arch/x86/events/core.c
@@ -254,15 +254,16 @@ static bool check_hw_exists(void)
 	 * We still allow the PMU driver to operate:
 	 */
 	if (bios_fail) {
-		printk(KERN_CONT "Broken BIOS detected, complain to your hardware vendor.\n");
-		printk(KERN_ERR FW_BUG "the BIOS has corrupted hw-PMU resources (MSR %x is %Lx)\n", reg_fail, val_fail);
+		pr_cont("Broken BIOS detected, complain to your hardware vendor.\n");
+		pr_err(FW_BUG "the BIOS has corrupted hw-PMU resources (MSR %x is %Lx)\n",
+			      reg_fail, val_fail);
 	}
 
 	return true;
 
 msr_fail:
-	printk(KERN_CONT "Broken PMU hardware detected, using software events only.\n");
-	printk("%sFailed to access perfctr msr (MSR %x is %Lx)\n",
+	pr_cont("Broken PMU hardware detected, using software events only.\n");
+	pr_info("%sFailed to access perfctr msr (MSR %x is %Lx)\n",
 		boot_cpu_has(X86_FEATURE_HYPERVISOR) ? KERN_INFO : KERN_ERR,
 		reg, val_new);
 
@@ -596,6 +597,19 @@ void x86_pmu_disable_all(void)
 	}
 }
 
+/*
+ * There may be PMI landing after enabled=0. The PMI hitting could be before or
+ * after disable_all.
+ *
+ * If PMI hits before disable_all, the PMU will be disabled in the NMI handler.
+ * It will not be re-enabled in the NMI handler again, because enabled=0. After
+ * handling the NMI, disable_all will be called, which will not change the
+ * state either. If PMI hits after disable_all, the PMU is already disabled
+ * before entering NMI handler. The NMI handler will not change the state
+ * either.
+ *
+ * So either situation is harmless.
+ */
 static void x86_pmu_disable(struct pmu *pmu)
 {
 	struct cpu_hw_events *cpuc = this_cpu_ptr(&cpu_hw_events);
diff --git a/arch/x86/kernel/cpu/perf_event_intel_bts.c b/arch/x86/events/intel/bts.c
index 2cad71d..b99dc92 100644
--- a/arch/x86/kernel/cpu/perf_event_intel_bts.c
+++ b/arch/x86/events/intel/bts.c
@@ -26,7 +26,7 @@
 #include <asm-generic/sizes.h>
 #include <asm/perf_event.h>
 
-#include "perf_event.h"
+#include "../perf_event.h"
 
 struct bts_ctx {
 	struct perf_output_handle	handle;
diff --git a/arch/x86/kernel/cpu/perf_event_intel.c b/arch/x86/events/intel/core.c
index fed2ab1..68fa55b 100644
--- a/arch/x86/kernel/cpu/perf_event_intel.c
+++ b/arch/x86/events/intel/core.c
@@ -18,7 +18,7 @@
 #include <asm/hardirq.h>
 #include <asm/apic.h>
 
-#include "perf_event.h"
+#include "../perf_event.h"
 
 /*
  * Intel PerfMon, used on Core and later.
@@ -1502,7 +1502,15 @@ static __initconst const u64 knl_hw_cache_extra_regs
 };
 
 /*
- * Use from PMIs where the LBRs are already disabled.
+ * Used from PMIs where the LBRs are already disabled.
+ *
+ * This function could be called consecutively. It is required to remain in
+ * disabled state if called consecutively.
+ *
+ * During consecutive calls, the same disable value will be written to related
+ * registers, so the PMU state remains unchanged. hw.state in
+ * intel_bts_disable_local will remain PERF_HES_STOPPED too in consecutive
+ * calls.
  */
 static void __intel_pmu_disable_all(void)
 {
@@ -1884,6 +1892,16 @@ again:
 	if (__test_and_clear_bit(62, (unsigned long *)&status)) {
 		handled++;
 		x86_pmu.drain_pebs(regs);
+		/*
+		 * There are cases where, even though, the PEBS ovfl bit is set
+		 * in GLOBAL_OVF_STATUS, the PEBS events may also have their
+		 * overflow bits set for their counters. We must clear them
+		 * here because they have been processed as exact samples in
+		 * the drain_pebs() routine. They must not be processed again
+		 * in the for_each_bit_set() loop for regular samples below.
+		 */
+		status &= ~cpuc->pebs_enabled;
+		status &= x86_pmu.intel_ctrl | GLOBAL_STATUS_TRACE_TOPAPMI;
 	}
 
 	/*
@@ -1929,7 +1947,10 @@ again:
 		goto again;
 
 done:
-	__intel_pmu_enable_all(0, true);
+	/* Only restore PMU state when it's active. See x86_pmu_disable(). */
+	if (cpuc->enabled)
+		__intel_pmu_enable_all(0, true);
+
 	/*
 	 * Only unmask the NMI after the overflow counters
 	 * have been reset. This avoids spurious NMIs on
@@ -3396,6 +3417,7 @@ __init int intel_pmu_init(void)
 		intel_perfmon_event_map[PERF_COUNT_HW_STALLED_CYCLES_BACKEND] =
 			X86_CONFIG(.event=0xb1, .umask=0x3f, .inv=1, .cmask=1);
 
+		intel_pmu_pebs_data_source_nhm();
 		x86_add_quirk(intel_nehalem_quirk);
 
 		pr_cont("Nehalem events, ");
@@ -3459,6 +3481,7 @@ __init int intel_pmu_init(void)
 		intel_perfmon_event_map[PERF_COUNT_HW_STALLED_CYCLES_BACKEND] =
 			X86_CONFIG(.event=0xb1, .umask=0x3f, .inv=1, .cmask=1);
 
+		intel_pmu_pebs_data_source_nhm();
 		pr_cont("Westmere events, ");
 		break;
 
@@ -3581,7 +3604,7 @@ __init int intel_pmu_init(void)
 		intel_pmu_lbr_init_hsw();
 
 		x86_pmu.event_constraints = intel_bdw_event_constraints;
-		x86_pmu.pebs_constraints = intel_hsw_pebs_event_constraints;
+		x86_pmu.pebs_constraints = intel_bdw_pebs_event_constraints;
 		x86_pmu.extra_regs = intel_snbep_extra_regs;
 		x86_pmu.pebs_aliases = intel_pebs_aliases_ivb;
 		x86_pmu.pebs_prec_dist = true;
diff --git a/arch/x86/kernel/cpu/perf_event_intel_cqm.c b/arch/x86/events/intel/cqm.c
index a316ca9..93cb412 100644
--- a/arch/x86/kernel/cpu/perf_event_intel_cqm.c
+++ b/arch/x86/events/intel/cqm.c
@@ -7,7 +7,7 @@
 #include <linux/perf_event.h>
 #include <linux/slab.h>
 #include <asm/cpu_device_id.h>
-#include "perf_event.h"
+#include "../perf_event.h"
 
 #define MSR_IA32_PQR_ASSOC	0x0c8f
 #define MSR_IA32_QM_CTR		0x0c8e
@@ -1244,15 +1244,12 @@ static struct pmu intel_cqm_pmu = {
 
 static inline void cqm_pick_event_reader(int cpu)
 {
-	int phys_id = topology_physical_package_id(cpu);
-	int i;
+	int reader;
 
-	for_each_cpu(i, &cqm_cpumask) {
-		if (phys_id == topology_physical_package_id(i))
-			return;	/* already got reader for this socket */
-	}
-
-	cpumask_set_cpu(cpu, &cqm_cpumask);
+	/* First online cpu in package becomes the reader */
+	reader = cpumask_any_and(&cqm_cpumask, topology_core_cpumask(cpu));
+	if (reader >= nr_cpu_ids)
+		cpumask_set_cpu(cpu, &cqm_cpumask);
 }
 
 static void intel_cqm_cpu_starting(unsigned int cpu)
@@ -1270,24 +1267,17 @@ static void intel_cqm_cpu_starting(unsigned int cpu)
 
 static void intel_cqm_cpu_exit(unsigned int cpu)
 {
-	int phys_id = topology_physical_package_id(cpu);
-	int i;
+	int target;
 
-	/*
-	 * Is @cpu a designated cqm reader?
-	 */
+	/* Is @cpu the current cqm reader for this package ? */
 	if (!cpumask_test_and_clear_cpu(cpu, &cqm_cpumask))
 		return;
 
-	for_each_online_cpu(i) {
-		if (i == cpu)
-			continue;
+	/* Find another online reader in this package */
+	target = cpumask_any_but(topology_core_cpumask(cpu), cpu);
 
-		if (phys_id == topology_physical_package_id(i)) {
-			cpumask_set_cpu(i, &cqm_cpumask);
-			break;
-		}
-	}
+	if (target < nr_cpu_ids)
+		cpumask_set_cpu(target, &cqm_cpumask);
 }
 
 static int intel_cqm_cpu_notifier(struct notifier_block *nb,
diff --git a/arch/x86/kernel/cpu/perf_event_intel_cstate.c b/arch/x86/events/intel/cstate.c
index 75a38b5..7946c42 100644
--- a/arch/x86/kernel/cpu/perf_event_intel_cstate.c
+++ b/arch/x86/events/intel/cstate.c
@@ -89,7 +89,7 @@
 #include <linux/slab.h>
 #include <linux/perf_event.h>
 #include <asm/cpu_device_id.h>
-#include "perf_event.h"
+#include "../perf_event.h"
 
 #define DEFINE_CSTATE_FORMAT_ATTR(_var, _name, _format)		\
 static ssize_t __cstate_##_var##_show(struct kobject *kobj,	\
diff --git a/arch/x86/kernel/cpu/perf_event_intel_ds.c b/arch/x86/events/intel/ds.c
index 10602f0..ce7211a 100644
--- a/arch/x86/kernel/cpu/perf_event_intel_ds.c
+++ b/arch/x86/events/intel/ds.c
@@ -5,7 +5,7 @@
 #include <asm/perf_event.h>
 #include <asm/insn.h>
 
-#include "perf_event.h"
+#include "../perf_event.h"
 
 /* The size of a BTS record in bytes: */
 #define BTS_RECORD_SIZE		24
@@ -51,7 +51,8 @@ union intel_x86_pebs_dse {
 #define OP_LH (P(OP, LOAD) | P(LVL, HIT))
 #define SNOOP_NONE_MISS (P(SNOOP, NONE) | P(SNOOP, MISS))
 
-static const u64 pebs_data_source[] = {
+/* Version for Sandy Bridge and later */
+static u64 pebs_data_source[] = {
 	P(OP, LOAD) | P(LVL, MISS) | P(LVL, L3) | P(SNOOP, NA),/* 0x00:ukn L3 */
 	OP_LH | P(LVL, L1)  | P(SNOOP, NONE),	/* 0x01: L1 local */
 	OP_LH | P(LVL, LFB) | P(SNOOP, NONE),	/* 0x02: LFB hit */
@@ -70,6 +71,14 @@ static const u64 pebs_data_source[] = {
 	OP_LH | P(LVL, UNC) | P(SNOOP, NONE), /* 0x0f: uncached */
 };
 
+/* Patch up minor differences in the bits */
+void __init intel_pmu_pebs_data_source_nhm(void)
+{
+	pebs_data_source[0x05] = OP_LH | P(LVL, L3)  | P(SNOOP, HIT);
+	pebs_data_source[0x06] = OP_LH | P(LVL, L3)  | P(SNOOP, HITM);
+	pebs_data_source[0x07] = OP_LH | P(LVL, L3)  | P(SNOOP, HITM);
+}
+
 static u64 precise_store_data(u64 status)
 {
 	union intel_x86_pebs_dse dse;
@@ -269,7 +278,7 @@ static int alloc_pebs_buffer(int cpu)
 	if (!x86_pmu.pebs)
 		return 0;
 
-	buffer = kzalloc_node(PEBS_BUFFER_SIZE, GFP_KERNEL, node);
+	buffer = kzalloc_node(x86_pmu.pebs_buffer_size, GFP_KERNEL, node);
 	if (unlikely(!buffer))
 		return -ENOMEM;
 
@@ -286,7 +295,7 @@ static int alloc_pebs_buffer(int cpu)
 		per_cpu(insn_buffer, cpu) = ibuffer;
 	}
 
-	max = PEBS_BUFFER_SIZE / x86_pmu.pebs_record_size;
+	max = x86_pmu.pebs_buffer_size / x86_pmu.pebs_record_size;
 
 	ds->pebs_buffer_base = (u64)(unsigned long)buffer;
 	ds->pebs_index = ds->pebs_buffer_base;
@@ -722,6 +731,30 @@ struct event_constraint intel_hsw_pebs_event_constraints[] = {
 	EVENT_CONSTRAINT_END
 };
 
+struct event_constraint intel_bdw_pebs_event_constraints[] = {
+	INTEL_FLAGS_UEVENT_CONSTRAINT(0x01c0, 0x2), /* INST_RETIRED.PRECDIST */
+	INTEL_PLD_CONSTRAINT(0x01cd, 0xf),    /* MEM_TRANS_RETIRED.* */
+	/* UOPS_RETIRED.ALL, inv=1, cmask=16 (cycles:p). */
+	INTEL_FLAGS_EVENT_CONSTRAINT(0x108001c2, 0xf),
+	/* INST_RETIRED.PREC_DIST, inv=1, cmask=16 (cycles:ppp). */
+	INTEL_FLAGS_EVENT_CONSTRAINT(0x108001c0, 0x2),
+	INTEL_FLAGS_UEVENT_CONSTRAINT_DATALA_NA(0x01c2, 0xf), /* UOPS_RETIRED.ALL */
+	INTEL_FLAGS_UEVENT_CONSTRAINT_DATALA_LD(0x11d0, 0xf), /* MEM_UOPS_RETIRED.STLB_MISS_LOADS */
+	INTEL_FLAGS_UEVENT_CONSTRAINT_DATALA_LD(0x21d0, 0xf), /* MEM_UOPS_RETIRED.LOCK_LOADS */
+	INTEL_FLAGS_UEVENT_CONSTRAINT_DATALA_LD(0x41d0, 0xf), /* MEM_UOPS_RETIRED.SPLIT_LOADS */
+	INTEL_FLAGS_UEVENT_CONSTRAINT_DATALA_LD(0x81d0, 0xf), /* MEM_UOPS_RETIRED.ALL_LOADS */
+	INTEL_FLAGS_UEVENT_CONSTRAINT_DATALA_ST(0x12d0, 0xf), /* MEM_UOPS_RETIRED.STLB_MISS_STORES */
+	INTEL_FLAGS_UEVENT_CONSTRAINT_DATALA_ST(0x42d0, 0xf), /* MEM_UOPS_RETIRED.SPLIT_STORES */
+	INTEL_FLAGS_UEVENT_CONSTRAINT_DATALA_ST(0x82d0, 0xf), /* MEM_UOPS_RETIRED.ALL_STORES */
+	INTEL_FLAGS_EVENT_CONSTRAINT_DATALA_LD(0xd1, 0xf),    /* MEM_LOAD_UOPS_RETIRED.* */
+	INTEL_FLAGS_EVENT_CONSTRAINT_DATALA_LD(0xd2, 0xf),    /* MEM_LOAD_UOPS_L3_HIT_RETIRED.* */
+	INTEL_FLAGS_EVENT_CONSTRAINT_DATALA_LD(0xd3, 0xf),    /* MEM_LOAD_UOPS_L3_MISS_RETIRED.* */
+	/* Allow all events as PEBS with no flags */
+	INTEL_ALL_EVENT_CONSTRAINT(0, 0xf),
+	EVENT_CONSTRAINT_END
+};
+
+
 struct event_constraint intel_skl_pebs_event_constraints[] = {
 	INTEL_FLAGS_UEVENT_CONSTRAINT(0x1c0, 0x2),	/* INST_RETIRED.PREC_DIST */
 	/* INST_RETIRED.PREC_DIST, inv=1, cmask=16 (cycles:ppp). */
@@ -1319,19 +1352,28 @@ void __init intel_ds_init(void)
 
 	x86_pmu.bts  = boot_cpu_has(X86_FEATURE_BTS);
 	x86_pmu.pebs = boot_cpu_has(X86_FEATURE_PEBS);
+	x86_pmu.pebs_buffer_size = PEBS_BUFFER_SIZE;
 	if (x86_pmu.pebs) {
 		char pebs_type = x86_pmu.intel_cap.pebs_trap ?  '+' : '-';
 		int format = x86_pmu.intel_cap.pebs_format;
 
 		switch (format) {
 		case 0:
-			printk(KERN_CONT "PEBS fmt0%c, ", pebs_type);
+			pr_cont("PEBS fmt0%c, ", pebs_type);
 			x86_pmu.pebs_record_size = sizeof(struct pebs_record_core);
+			/*
+			 * Using >PAGE_SIZE buffers makes the WRMSR to
+			 * PERF_GLOBAL_CTRL in intel_pmu_enable_all()
+			 * mysteriously hang on Core2.
+			 *
+			 * As a workaround, we don't do this.
+			 */
+			x86_pmu.pebs_buffer_size = PAGE_SIZE;
 			x86_pmu.drain_pebs = intel_pmu_drain_pebs_core;
 			break;
 
 		case 1:
-			printk(KERN_CONT "PEBS fmt1%c, ", pebs_type);
+			pr_cont("PEBS fmt1%c, ", pebs_type);
 			x86_pmu.pebs_record_size = sizeof(struct pebs_record_nhm);
 			x86_pmu.drain_pebs = intel_pmu_drain_pebs_nhm;
 			break;
@@ -1351,7 +1393,7 @@ void __init intel_ds_init(void)
 			break;
 
 		default:
-			printk(KERN_CONT "no PEBS fmt%d%c, ", format, pebs_type);
+			pr_cont("no PEBS fmt%d%c, ", format, pebs_type);
 			x86_pmu.pebs = 0;
 		}
 	}
diff --git a/arch/x86/kernel/cpu/perf_event_knc.c b/arch/x86/events/intel/knc.c
index 5b0c232..548d5f7 100644
--- a/arch/x86/kernel/cpu/perf_event_knc.c
+++ b/arch/x86/events/intel/knc.c
@@ -5,7 +5,7 @@
 
 #include <asm/hardirq.h>
 
-#include "perf_event.h"
+#include "../perf_event.h"
 
 static const u64 knc_perfmon_event_map[] =
 {
@@ -263,7 +263,9 @@ again:
 		goto again;
 
 done:
-	knc_pmu_enable_all(0);
+	/* Only restore PMU state when it's active. See x86_pmu_disable(). */
+	if (cpuc->enabled)
+		knc_pmu_enable_all(0);
 
 	return handled;
 }
diff --git a/arch/x86/kernel/cpu/perf_event_intel_lbr.c b/arch/x86/events/intel/lbr.c
index 653f88d..69dd118 100644
--- a/arch/x86/kernel/cpu/perf_event_intel_lbr.c
+++ b/arch/x86/events/intel/lbr.c
@@ -5,7 +5,7 @@
 #include <asm/msr.h>
 #include <asm/insn.h>
 
-#include "perf_event.h"
+#include "../perf_event.h"
 
 enum {
 	LBR_FORMAT_32		= 0x00,
diff --git a/arch/x86/kernel/cpu/perf_event_p4.c b/arch/x86/events/intel/p4.c
index f2e5678..0a5ede1 100644
--- a/arch/x86/kernel/cpu/perf_event_p4.c
+++ b/arch/x86/events/intel/p4.c
@@ -13,7 +13,7 @@
 #include <asm/hardirq.h>
 #include <asm/apic.h>
 
-#include "perf_event.h"
+#include "../perf_event.h"
 
 #define P4_CNTR_LIMIT 3
 /*
diff --git a/arch/x86/kernel/cpu/perf_event_p6.c b/arch/x86/events/intel/p6.c
index 7c1a0c0..1f5c47a 100644
--- a/arch/x86/kernel/cpu/perf_event_p6.c
+++ b/arch/x86/events/intel/p6.c
@@ -1,7 +1,7 @@
 #include <linux/perf_event.h>
 #include <linux/types.h>
 
-#include "perf_event.h"
+#include "../perf_event.h"
 
 /*
  * Not sure about some of these
diff --git a/arch/x86/kernel/cpu/perf_event_intel_pt.c b/arch/x86/events/intel/pt.c
index c0bbd10..6af7cf7 100644
--- a/arch/x86/kernel/cpu/perf_event_intel_pt.c
+++ b/arch/x86/events/intel/pt.c
@@ -29,8 +29,8 @@
 #include <asm/io.h>
 #include <asm/intel_pt.h>
 
-#include "perf_event.h"
-#include "intel_pt.h"
+#include "../perf_event.h"
+#include "pt.h"
 
 static DEFINE_PER_CPU(struct pt, pt_ctx);
 
diff --git a/arch/x86/kernel/cpu/intel_pt.h b/arch/x86/events/intel/pt.h
index 336878a..336878a 100644
--- a/arch/x86/kernel/cpu/intel_pt.h
+++ b/arch/x86/events/intel/pt.h
diff --git a/arch/x86/kernel/cpu/perf_event_intel_rapl.c b/arch/x86/events/intel/rapl.c
index 24a351a..b834a3f 100644
--- a/arch/x86/kernel/cpu/perf_event_intel_rapl.c
+++ b/arch/x86/events/intel/rapl.c
@@ -44,11 +44,14 @@
  * the duration of the measurement. Tools may use a function such as
  * ldexp(raw_count, -32);
  */
+
+#define pr_fmt(fmt) "RAPL PMU: " fmt
+
 #include <linux/module.h>
 #include <linux/slab.h>
 #include <linux/perf_event.h>
 #include <asm/cpu_device_id.h>
-#include "perf_event.h"
+#include "../perf_event.h"
 
 /*
  * RAPL energy status counters
@@ -107,7 +110,7 @@ static ssize_t __rapl_##_var##_show(struct kobject *kobj,	\
 static struct kobj_attribute format_attr_##_var =		\
 	__ATTR(_name, 0444, __rapl_##_var##_show, NULL)
 
-#define RAPL_CNTR_WIDTH 32 /* 32-bit rapl counters */
+#define RAPL_CNTR_WIDTH 32
 
 #define RAPL_EVENT_ATTR_STR(_name, v, str)					\
 static struct perf_pmu_events_attr event_attr_##v = {				\
@@ -117,23 +120,33 @@ static struct perf_pmu_events_attr event_attr_##v = {				\
 };
 
 struct rapl_pmu {
-	spinlock_t	 lock;
-	int		 n_active; /* number of active events */
-	struct list_head active_list;
-	struct pmu	 *pmu; /* pointer to rapl_pmu_class */
-	ktime_t		 timer_interval; /* in ktime_t unit */
-	struct hrtimer   hrtimer;
+	raw_spinlock_t		lock;
+	int			n_active;
+	int			cpu;
+	struct list_head	active_list;
+	struct pmu		*pmu;
+	ktime_t			timer_interval;
+	struct hrtimer		hrtimer;
 };
 
-static int rapl_hw_unit[NR_RAPL_DOMAINS] __read_mostly;  /* 1/2^hw_unit Joule */
-static struct pmu rapl_pmu_class;
+struct rapl_pmus {
+	struct pmu		pmu;
+	unsigned int		maxpkg;
+	struct rapl_pmu		*pmus[];
+};
+
+ /* 1/2^hw_unit Joule */
+static int rapl_hw_unit[NR_RAPL_DOMAINS] __read_mostly;
+static struct rapl_pmus *rapl_pmus;
 static cpumask_t rapl_cpu_mask;
-static int rapl_cntr_mask;
+static unsigned int rapl_cntr_mask;
+static u64 rapl_timer_ms;
 
-static DEFINE_PER_CPU(struct rapl_pmu *, rapl_pmu);
-static DEFINE_PER_CPU(struct rapl_pmu *, rapl_pmu_to_free);
+static inline struct rapl_pmu *cpu_to_rapl_pmu(unsigned int cpu)
+{
+	return rapl_pmus->pmus[topology_logical_package_id(cpu)];
+}
 
-static struct x86_pmu_quirk *rapl_quirks;
 static inline u64 rapl_read_counter(struct perf_event *event)
 {
 	u64 raw;
@@ -141,19 +154,10 @@ static inline u64 rapl_read_counter(struct perf_event *event)
 	return raw;
 }
 
-#define rapl_add_quirk(func_)						\
-do {									\
-	static struct x86_pmu_quirk __quirk __initdata = {		\
-		.func = func_,						\
-	};								\
-	__quirk.next = rapl_quirks;					\
-	rapl_quirks = &__quirk;						\
-} while (0)
-
 static inline u64 rapl_scale(u64 v, int cfg)
 {
 	if (cfg > NR_RAPL_DOMAINS) {
-		pr_warn("invalid domain %d, failed to scale data\n", cfg);
+		pr_warn("Invalid domain %d, failed to scale data\n", cfg);
 		return v;
 	}
 	/*
@@ -206,27 +210,21 @@ static void rapl_start_hrtimer(struct rapl_pmu *pmu)
 		     HRTIMER_MODE_REL_PINNED);
 }
 
-static void rapl_stop_hrtimer(struct rapl_pmu *pmu)
-{
-	hrtimer_cancel(&pmu->hrtimer);
-}
-
 static enum hrtimer_restart rapl_hrtimer_handle(struct hrtimer *hrtimer)
 {
-	struct rapl_pmu *pmu = __this_cpu_read(rapl_pmu);
+	struct rapl_pmu *pmu = container_of(hrtimer, struct rapl_pmu, hrtimer);
 	struct perf_event *event;
 	unsigned long flags;
 
 	if (!pmu->n_active)
 		return HRTIMER_NORESTART;
 
-	spin_lock_irqsave(&pmu->lock, flags);
+	raw_spin_lock_irqsave(&pmu->lock, flags);
 
-	list_for_each_entry(event, &pmu->active_list, active_entry) {
+	list_for_each_entry(event, &pmu->active_list, active_entry)
 		rapl_event_update(event);
-	}
 
-	spin_unlock_irqrestore(&pmu->lock, flags);
+	raw_spin_unlock_irqrestore(&pmu->lock, flags);
 
 	hrtimer_forward_now(hrtimer, pmu->timer_interval);
 
@@ -260,28 +258,28 @@ static void __rapl_pmu_event_start(struct rapl_pmu *pmu,
 
 static void rapl_pmu_event_start(struct perf_event *event, int mode)
 {
-	struct rapl_pmu *pmu = __this_cpu_read(rapl_pmu);
+	struct rapl_pmu *pmu = event->pmu_private;
 	unsigned long flags;
 
-	spin_lock_irqsave(&pmu->lock, flags);
+	raw_spin_lock_irqsave(&pmu->lock, flags);
 	__rapl_pmu_event_start(pmu, event);
-	spin_unlock_irqrestore(&pmu->lock, flags);
+	raw_spin_unlock_irqrestore(&pmu->lock, flags);
 }
 
 static void rapl_pmu_event_stop(struct perf_event *event, int mode)
 {
-	struct rapl_pmu *pmu = __this_cpu_read(rapl_pmu);
+	struct rapl_pmu *pmu = event->pmu_private;
 	struct hw_perf_event *hwc = &event->hw;
 	unsigned long flags;
 
-	spin_lock_irqsave(&pmu->lock, flags);
+	raw_spin_lock_irqsave(&pmu->lock, flags);
 
 	/* mark event as deactivated and stopped */
 	if (!(hwc->state & PERF_HES_STOPPED)) {
 		WARN_ON_ONCE(pmu->n_active <= 0);
 		pmu->n_active--;
 		if (pmu->n_active == 0)
-			rapl_stop_hrtimer(pmu);
+			hrtimer_cancel(&pmu->hrtimer);
 
 		list_del(&event->active_entry);
 
@@ -299,23 +297,23 @@ static void rapl_pmu_event_stop(struct perf_event *event, int mode)
 		hwc->state |= PERF_HES_UPTODATE;
 	}
 
-	spin_unlock_irqrestore(&pmu->lock, flags);
+	raw_spin_unlock_irqrestore(&pmu->lock, flags);
 }
 
 static int rapl_pmu_event_add(struct perf_event *event, int mode)
 {
-	struct rapl_pmu *pmu = __this_cpu_read(rapl_pmu);
+	struct rapl_pmu *pmu = event->pmu_private;
 	struct hw_perf_event *hwc = &event->hw;
 	unsigned long flags;
 
-	spin_lock_irqsave(&pmu->lock, flags);
+	raw_spin_lock_irqsave(&pmu->lock, flags);
 
 	hwc->state = PERF_HES_UPTODATE | PERF_HES_STOPPED;
 
 	if (mode & PERF_EF_START)
 		__rapl_pmu_event_start(pmu, event);
 
-	spin_unlock_irqrestore(&pmu->lock, flags);
+	raw_spin_unlock_irqrestore(&pmu->lock, flags);
 
 	return 0;
 }
@@ -329,15 +327,19 @@ static int rapl_pmu_event_init(struct perf_event *event)
 {
 	u64 cfg = event->attr.config & RAPL_EVENT_MASK;
 	int bit, msr, ret = 0;
+	struct rapl_pmu *pmu;
 
 	/* only look at RAPL events */
-	if (event->attr.type != rapl_pmu_class.type)
+	if (event->attr.type != rapl_pmus->pmu.type)
 		return -ENOENT;
 
 	/* check only supported bits are set */
 	if (event->attr.config & ~RAPL_EVENT_MASK)
 		return -EINVAL;
 
+	if (event->cpu < 0)
+		return -EINVAL;
+
 	/*
 	 * check event is known (determines counter)
 	 */
@@ -376,6 +378,9 @@ static int rapl_pmu_event_init(struct perf_event *event)
 		return -EINVAL;
 
 	/* must be done before validate_group */
+	pmu = cpu_to_rapl_pmu(event->cpu);
+	event->cpu = pmu->cpu;
+	event->pmu_private = pmu;
 	event->hw.event_base = msr;
 	event->hw.config = cfg;
 	event->hw.idx = bit;
@@ -506,139 +511,62 @@ const struct attribute_group *rapl_attr_groups[] = {
 	NULL,
 };
 
-static struct pmu rapl_pmu_class = {
-	.attr_groups	= rapl_attr_groups,
-	.task_ctx_nr	= perf_invalid_context, /* system-wide only */
-	.event_init	= rapl_pmu_event_init,
-	.add		= rapl_pmu_event_add, /* must have */
-	.del		= rapl_pmu_event_del, /* must have */
-	.start		= rapl_pmu_event_start,
-	.stop		= rapl_pmu_event_stop,
-	.read		= rapl_pmu_event_read,
-};
-
 static void rapl_cpu_exit(int cpu)
 {
-	struct rapl_pmu *pmu = per_cpu(rapl_pmu, cpu);
-	int i, phys_id = topology_physical_package_id(cpu);
-	int target = -1;
+	struct rapl_pmu *pmu = cpu_to_rapl_pmu(cpu);
+	int target;
 
-	/* find a new cpu on same package */
-	for_each_online_cpu(i) {
-		if (i == cpu)
-			continue;
-		if (phys_id == topology_physical_package_id(i)) {
-			target = i;
-			break;
-		}
-	}
-	/*
-	 * clear cpu from cpumask
-	 * if was set in cpumask and still some cpu on package,
-	 * then move to new cpu
-	 */
-	if (cpumask_test_and_clear_cpu(cpu, &rapl_cpu_mask) && target >= 0)
-		cpumask_set_cpu(target, &rapl_cpu_mask);
+	/* Check if exiting cpu is used for collecting rapl events */
+	if (!cpumask_test_and_clear_cpu(cpu, &rapl_cpu_mask))
+		return;
 
-	WARN_ON(cpumask_empty(&rapl_cpu_mask));
-	/*
-	 * migrate events and context to new cpu
-	 */
-	if (target >= 0)
-		perf_pmu_migrate_context(pmu->pmu, cpu, target);
+	pmu->cpu = -1;
+	/* Find a new cpu to collect rapl events */
+	target = cpumask_any_but(topology_core_cpumask(cpu), cpu);
 
-	/* cancel overflow polling timer for CPU */
-	rapl_stop_hrtimer(pmu);
+	/* Migrate rapl events to the new target */
+	if (target < nr_cpu_ids) {
+		cpumask_set_cpu(target, &rapl_cpu_mask);
+		pmu->cpu = target;
+		perf_pmu_migrate_context(pmu->pmu, cpu, target);
+	}
 }
 
 static void rapl_cpu_init(int cpu)
 {
-	int i, phys_id = topology_physical_package_id(cpu);
+	struct rapl_pmu *pmu = cpu_to_rapl_pmu(cpu);
+	int target;
 
-	/* check if phys_is is already covered */
-	for_each_cpu(i, &rapl_cpu_mask) {
-		if (phys_id == topology_physical_package_id(i))
-			return;
-	}
-	/* was not found, so add it */
-	cpumask_set_cpu(cpu, &rapl_cpu_mask);
-}
-
-static __init void rapl_hsw_server_quirk(void)
-{
 	/*
-	 * DRAM domain on HSW server has fixed energy unit which can be
-	 * different than the unit from power unit MSR.
-	 * "Intel Xeon Processor E5-1600 and E5-2600 v3 Product Families, V2
-	 * of 2. Datasheet, September 2014, Reference Number: 330784-001 "
+	 * Check if there is an online cpu in the package which collects rapl
+	 * events already.
 	 */
-	rapl_hw_unit[RAPL_IDX_RAM_NRG_STAT] = 16;
+	target = cpumask_any_and(&rapl_cpu_mask, topology_core_cpumask(cpu));
+	if (target < nr_cpu_ids)
+		return;
+
+	cpumask_set_cpu(cpu, &rapl_cpu_mask);
+	pmu->cpu = cpu;
 }
 
 static int rapl_cpu_prepare(int cpu)
 {
-	struct rapl_pmu *pmu = per_cpu(rapl_pmu, cpu);
-	int phys_id = topology_physical_package_id(cpu);
-	u64 ms;
+	struct rapl_pmu *pmu = cpu_to_rapl_pmu(cpu);
 
 	if (pmu)
 		return 0;
 
-	if (phys_id < 0)
-		return -1;
-
 	pmu = kzalloc_node(sizeof(*pmu), GFP_KERNEL, cpu_to_node(cpu));
 	if (!pmu)
-		return -1;
-	spin_lock_init(&pmu->lock);
+		return -ENOMEM;
 
+	raw_spin_lock_init(&pmu->lock);
 	INIT_LIST_HEAD(&pmu->active_list);
-
-	pmu->pmu = &rapl_pmu_class;
-
-	/*
-	 * use reference of 200W for scaling the timeout
-	 * to avoid missing counter overflows.
-	 * 200W = 200 Joules/sec
-	 * divide interval by 2 to avoid lockstep (2 * 100)
-	 * if hw unit is 32, then we use 2 ms 1/200/2
-	 */
-	if (rapl_hw_unit[0] < 32)
-		ms = (1000 / (2 * 100)) * (1ULL << (32 - rapl_hw_unit[0] - 1));
-	else
-		ms = 2;
-
-	pmu->timer_interval = ms_to_ktime(ms);
-
+	pmu->pmu = &rapl_pmus->pmu;
+	pmu->timer_interval = ms_to_ktime(rapl_timer_ms);
+	pmu->cpu = -1;
 	rapl_hrtimer_init(pmu);
-
-	/* set RAPL pmu for this cpu for now */
-	per_cpu(rapl_pmu, cpu) = pmu;
-	per_cpu(rapl_pmu_to_free, cpu) = NULL;
-
-	return 0;
-}
-
-static void rapl_cpu_kfree(int cpu)
-{
-	struct rapl_pmu *pmu = per_cpu(rapl_pmu_to_free, cpu);
-
-	kfree(pmu);
-
-	per_cpu(rapl_pmu_to_free, cpu) = NULL;
-}
-
-static int rapl_cpu_dying(int cpu)
-{
-	struct rapl_pmu *pmu = per_cpu(rapl_pmu, cpu);
-
-	if (!pmu)
-		return 0;
-
-	per_cpu(rapl_pmu, cpu) = NULL;
-
-	per_cpu(rapl_pmu_to_free, cpu) = pmu;
-
+	rapl_pmus->pmus[topology_logical_package_id(cpu)] = pmu;
 	return 0;
 }
 
@@ -651,28 +579,20 @@ static int rapl_cpu_notifier(struct notifier_block *self,
 	case CPU_UP_PREPARE:
 		rapl_cpu_prepare(cpu);
 		break;
-	case CPU_STARTING:
-		rapl_cpu_init(cpu);
-		break;
-	case CPU_UP_CANCELED:
-	case CPU_DYING:
-		rapl_cpu_dying(cpu);
-		break;
+
+	case CPU_DOWN_FAILED:
 	case CPU_ONLINE:
-	case CPU_DEAD:
-		rapl_cpu_kfree(cpu);
+		rapl_cpu_init(cpu);
 		break;
+
 	case CPU_DOWN_PREPARE:
 		rapl_cpu_exit(cpu);
 		break;
-	default:
-		break;
 	}
-
 	return NOTIFY_OK;
 }
 
-static int rapl_check_hw_unit(void)
+static int rapl_check_hw_unit(bool apply_quirk)
 {
 	u64 msr_rapl_power_unit_bits;
 	int i;
@@ -683,28 +603,107 @@ static int rapl_check_hw_unit(void)
 	for (i = 0; i < NR_RAPL_DOMAINS; i++)
 		rapl_hw_unit[i] = (msr_rapl_power_unit_bits >> 8) & 0x1FULL;
 
+	/*
+	 * DRAM domain on HSW server and KNL has fixed energy unit which can be
+	 * different than the unit from power unit MSR. See
+	 * "Intel Xeon Processor E5-1600 and E5-2600 v3 Product Families, V2
+	 * of 2. Datasheet, September 2014, Reference Number: 330784-001 "
+	 */
+	if (apply_quirk)
+		rapl_hw_unit[RAPL_IDX_RAM_NRG_STAT] = 16;
+
+	/*
+	 * Calculate the timer rate:
+	 * Use reference of 200W for scaling the timeout to avoid counter
+	 * overflows. 200W = 200 Joules/sec
+	 * Divide interval by 2 to avoid lockstep (2 * 100)
+	 * if hw unit is 32, then we use 2 ms 1/200/2
+	 */
+	rapl_timer_ms = 2;
+	if (rapl_hw_unit[0] < 32) {
+		rapl_timer_ms = (1000 / (2 * 100));
+		rapl_timer_ms *= (1ULL << (32 - rapl_hw_unit[0] - 1));
+	}
+	return 0;
+}
+
+static void __init rapl_advertise(void)
+{
+	int i;
+
+	pr_info("API unit is 2^-32 Joules, %d fixed counters, %llu ms ovfl timer\n",
+		hweight32(rapl_cntr_mask), rapl_timer_ms);
+
+	for (i = 0; i < NR_RAPL_DOMAINS; i++) {
+		if (rapl_cntr_mask & (1 << i)) {
+			pr_info("hw unit of domain %s 2^-%d Joules\n",
+				rapl_domain_names[i], rapl_hw_unit[i]);
+		}
+	}
+}
+
+static int __init rapl_prepare_cpus(void)
+{
+	unsigned int cpu, pkg;
+	int ret;
+
+	for_each_online_cpu(cpu) {
+		pkg = topology_logical_package_id(cpu);
+		if (rapl_pmus->pmus[pkg])
+			continue;
+
+		ret = rapl_cpu_prepare(cpu);
+		if (ret)
+			return ret;
+		rapl_cpu_init(cpu);
+	}
+	return 0;
+}
+
+static void __init cleanup_rapl_pmus(void)
+{
+	int i;
+
+	for (i = 0; i < rapl_pmus->maxpkg; i++)
+		kfree(rapl_pmus->pmus + i);
+	kfree(rapl_pmus);
+}
+
+static int __init init_rapl_pmus(void)
+{
+	int maxpkg = topology_max_packages();
+	size_t size;
+
+	size = sizeof(*rapl_pmus) + maxpkg * sizeof(struct rapl_pmu *);
+	rapl_pmus = kzalloc(size, GFP_KERNEL);
+	if (!rapl_pmus)
+		return -ENOMEM;
+
+	rapl_pmus->maxpkg		= maxpkg;
+	rapl_pmus->pmu.attr_groups	= rapl_attr_groups;
+	rapl_pmus->pmu.task_ctx_nr	= perf_invalid_context;
+	rapl_pmus->pmu.event_init	= rapl_pmu_event_init;
+	rapl_pmus->pmu.add		= rapl_pmu_event_add;
+	rapl_pmus->pmu.del		= rapl_pmu_event_del;
+	rapl_pmus->pmu.start		= rapl_pmu_event_start;
+	rapl_pmus->pmu.stop		= rapl_pmu_event_stop;
+	rapl_pmus->pmu.read		= rapl_pmu_event_read;
 	return 0;
 }
 
-static const struct x86_cpu_id rapl_cpu_match[] = {
+static const struct x86_cpu_id rapl_cpu_match[] __initconst = {
 	[0] = { .vendor = X86_VENDOR_INTEL, .family = 6 },
 	[1] = {},
 };
 
 static int __init rapl_pmu_init(void)
 {
-	struct rapl_pmu *pmu;
-	int cpu, ret;
-	struct x86_pmu_quirk *quirk;
-	int i;
+	bool apply_quirk = false;
+	int ret;
 
-	/*
-	 * check for Intel processor family 6
-	 */
 	if (!x86_match_cpu(rapl_cpu_match))
-		return 0;
+		return -ENODEV;
 
-	/* check supported CPU */
 	switch (boot_cpu_data.x86_model) {
 	case 42: /* Sandy Bridge */
 	case 58: /* Ivy Bridge */
@@ -712,7 +711,7 @@ static int __init rapl_pmu_init(void)
 		rapl_pmu_events_group.attrs = rapl_events_cln_attr;
 		break;
 	case 63: /* Haswell-Server */
-		rapl_add_quirk(rapl_hsw_server_quirk);
+		apply_quirk = true;
 		rapl_cntr_mask = RAPL_IDX_SRV;
 		rapl_pmu_events_group.attrs = rapl_events_srv_attr;
 		break;
@@ -728,56 +727,41 @@ static int __init rapl_pmu_init(void)
 		rapl_pmu_events_group.attrs = rapl_events_srv_attr;
 		break;
 	case 87: /* Knights Landing */
-		rapl_add_quirk(rapl_hsw_server_quirk);
+		apply_quirk = true;
 		rapl_cntr_mask = RAPL_IDX_KNL;
 		rapl_pmu_events_group.attrs = rapl_events_knl_attr;
-
+		break;
 	default:
-		/* unsupported */
-		return 0;
+		return -ENODEV;
 	}
-	ret = rapl_check_hw_unit();
+
+	ret = rapl_check_hw_unit(apply_quirk);
 	if (ret)
 		return ret;
 
-	/* run cpu model quirks */
-	for (quirk = rapl_quirks; quirk; quirk = quirk->next)
-		quirk->func();
-	cpu_notifier_register_begin();
+	ret = init_rapl_pmus();
+	if (ret)
+		return ret;
 
-	for_each_online_cpu(cpu) {
-		ret = rapl_cpu_prepare(cpu);
-		if (ret)
-			goto out;
-		rapl_cpu_init(cpu);
-	}
+	cpu_notifier_register_begin();
 
-	__perf_cpu_notifier(rapl_cpu_notifier);
+	ret = rapl_prepare_cpus();
+	if (ret)
+		goto out;
 
-	ret = perf_pmu_register(&rapl_pmu_class, "power", -1);
-	if (WARN_ON(ret)) {
-		pr_info("RAPL PMU detected, registration failed (%d), RAPL PMU disabled\n", ret);
-		cpu_notifier_register_done();
-		return -1;
-	}
+	ret = perf_pmu_register(&rapl_pmus->pmu, "power", -1);
+	if (ret)
+		goto out;
 
-	pmu = __this_cpu_read(rapl_pmu);
+	__perf_cpu_notifier(rapl_cpu_notifier);
+	cpu_notifier_register_done();
+	rapl_advertise();
+	return 0;
 
-	pr_info("RAPL PMU detected,"
-		" API unit is 2^-32 Joules,"
-		" %d fixed counters"
-		" %llu ms ovfl timer\n",
-		hweight32(rapl_cntr_mask),
-		ktime_to_ms(pmu->timer_interval));
-	for (i = 0; i < NR_RAPL_DOMAINS; i++) {
-		if (rapl_cntr_mask & (1 << i)) {
-			pr_info("hw unit of domain %s 2^-%d Joules\n",
-				rapl_domain_names[i], rapl_hw_unit[i]);
-		}
-	}
 out:
+	pr_warn("Initialization failed (%d), disabled\n", ret);
+	cleanup_rapl_pmus();
 	cpu_notifier_register_done();
-
-	return 0;
+	return ret;
 }
 device_initcall(rapl_pmu_init);
diff --git a/arch/x86/kernel/cpu/perf_event_intel_uncore.c b/arch/x86/events/intel/uncore.c
index 3bf41d4..7012d18 100644
--- a/arch/x86/kernel/cpu/perf_event_intel_uncore.c
+++ b/arch/x86/events/intel/uncore.c
@@ -1,4 +1,4 @@
-#include "perf_event_intel_uncore.h"
+#include "uncore.h"
 
 static struct intel_uncore_type *empty_uncore[] = { NULL, };
 struct intel_uncore_type **uncore_msr_uncores = empty_uncore;
@@ -9,9 +9,9 @@ struct pci_driver *uncore_pci_driver;
 /* pci bus to socket mapping */
 DEFINE_RAW_SPINLOCK(pci2phy_map_lock);
 struct list_head pci2phy_map_head = LIST_HEAD_INIT(pci2phy_map_head);
-struct pci_dev *uncore_extra_pci_dev[UNCORE_SOCKET_MAX][UNCORE_EXTRA_PCI_DEV_MAX];
+struct pci_extra_dev *uncore_extra_pci_dev;
+static int max_packages;
 
-static DEFINE_RAW_SPINLOCK(uncore_box_lock);
 /* mask of cpus that collect uncore events */
 static cpumask_t uncore_cpu_mask;
 
@@ -21,7 +21,7 @@ static struct event_constraint uncore_constraint_fixed =
 struct event_constraint uncore_constraint_empty =
 	EVENT_CONSTRAINT(0, 0, 0);
 
-int uncore_pcibus_to_physid(struct pci_bus *bus)
+static int uncore_pcibus_to_physid(struct pci_bus *bus)
 {
 	struct pci2phy_map *map;
 	int phys_id = -1;
@@ -38,6 +38,16 @@ int uncore_pcibus_to_physid(struct pci_bus *bus)
 	return phys_id;
 }
 
+static void uncore_free_pcibus_map(void)
+{
+	struct pci2phy_map *map, *tmp;
+
+	list_for_each_entry_safe(map, tmp, &pci2phy_map_head, list) {
+		list_del(&map->list);
+		kfree(map);
+	}
+}
+
 struct pci2phy_map *__find_pci2phy_map(int segment)
 {
 	struct pci2phy_map *map, *alloc = NULL;
@@ -82,43 +92,9 @@ ssize_t uncore_event_show(struct kobject *kobj,
 	return sprintf(buf, "%s", event->config);
 }
 
-struct intel_uncore_pmu *uncore_event_to_pmu(struct perf_event *event)
-{
-	return container_of(event->pmu, struct intel_uncore_pmu, pmu);
-}
-
 struct intel_uncore_box *uncore_pmu_to_box(struct intel_uncore_pmu *pmu, int cpu)
 {
-	struct intel_uncore_box *box;
-
-	box = *per_cpu_ptr(pmu->box, cpu);
-	if (box)
-		return box;
-
-	raw_spin_lock(&uncore_box_lock);
-	/* Recheck in lock to handle races. */
-	if (*per_cpu_ptr(pmu->box, cpu))
-		goto out;
-	list_for_each_entry(box, &pmu->box_list, list) {
-		if (box->phys_id == topology_physical_package_id(cpu)) {
-			atomic_inc(&box->refcnt);
-			*per_cpu_ptr(pmu->box, cpu) = box;
-			break;
-		}
-	}
-out:
-	raw_spin_unlock(&uncore_box_lock);
-
-	return *per_cpu_ptr(pmu->box, cpu);
-}
-
-struct intel_uncore_box *uncore_event_to_box(struct perf_event *event)
-{
-	/*
-	 * perf core schedules event on the basis of cpu, uncore events are
-	 * collected by one of the cpus inside a physical package.
-	 */
-	return uncore_pmu_to_box(uncore_event_to_pmu(event), smp_processor_id());
+	return pmu->boxes[topology_logical_package_id(cpu)];
 }
 
 u64 uncore_msr_read_counter(struct intel_uncore_box *box, struct perf_event *event)
@@ -207,7 +183,8 @@ u64 uncore_shared_reg_config(struct intel_uncore_box *box, int idx)
 	return config;
 }
 
-static void uncore_assign_hw_event(struct intel_uncore_box *box, struct perf_event *event, int idx)
+static void uncore_assign_hw_event(struct intel_uncore_box *box,
+				   struct perf_event *event, int idx)
 {
 	struct hw_perf_event *hwc = &event->hw;
 
@@ -302,24 +279,25 @@ static void uncore_pmu_init_hrtimer(struct intel_uncore_box *box)
 	box->hrtimer.function = uncore_pmu_hrtimer;
 }
 
-static struct intel_uncore_box *uncore_alloc_box(struct intel_uncore_type *type, int node)
+static struct intel_uncore_box *uncore_alloc_box(struct intel_uncore_type *type,
+						 int node)
 {
+	int i, size, numshared = type->num_shared_regs ;
 	struct intel_uncore_box *box;
-	int i, size;
 
-	size = sizeof(*box) + type->num_shared_regs * sizeof(struct intel_uncore_extra_reg);
+	size = sizeof(*box) + numshared * sizeof(struct intel_uncore_extra_reg);
 
 	box = kzalloc_node(size, GFP_KERNEL, node);
 	if (!box)
 		return NULL;
 
-	for (i = 0; i < type->num_shared_regs; i++)
+	for (i = 0; i < numshared; i++)
 		raw_spin_lock_init(&box->shared_regs[i].lock);
 
 	uncore_pmu_init_hrtimer(box);
-	atomic_set(&box->refcnt, 1);
 	box->cpu = -1;
-	box->phys_id = -1;
+	box->pci_phys_id = -1;
+	box->pkgid = -1;
 
 	/* set default hrtimer timeout */
 	box->hrtimer_duration = UNCORE_PMU_HRTIMER_INTERVAL;
@@ -341,7 +319,8 @@ static bool is_uncore_event(struct perf_event *event)
 }
 
 static int
-uncore_collect_events(struct intel_uncore_box *box, struct perf_event *leader, bool dogrp)
+uncore_collect_events(struct intel_uncore_box *box, struct perf_event *leader,
+		      bool dogrp)
 {
 	struct perf_event *event;
 	int n, max_count;
@@ -402,7 +381,8 @@ uncore_get_event_constraint(struct intel_uncore_box *box, struct perf_event *eve
 	return &type->unconstrainted;
 }
 
-static void uncore_put_event_constraint(struct intel_uncore_box *box, struct perf_event *event)
+static void uncore_put_event_constraint(struct intel_uncore_box *box,
+					struct perf_event *event)
 {
 	if (box->pmu->type->ops->put_constraint)
 		box->pmu->type->ops->put_constraint(box, event);
@@ -582,7 +562,7 @@ static void uncore_pmu_event_del(struct perf_event *event, int flags)
 		if (event == box->event_list[i]) {
 			uncore_put_event_constraint(box, event);
 
-			while (++i < box->n_events)
+			for (++i; i < box->n_events; i++)
 				box->event_list[i - 1] = box->event_list[i];
 
 			--box->n_events;
@@ -676,6 +656,7 @@ static int uncore_pmu_event_init(struct perf_event *event)
 	if (!box || box->cpu < 0)
 		return -EINVAL;
 	event->cpu = box->cpu;
+	event->pmu_private = box;
 
 	event->hw.idx = -1;
 	event->hw.last_tag = ~0ULL;
@@ -760,64 +741,110 @@ static int uncore_pmu_register(struct intel_uncore_pmu *pmu)
 	}
 
 	ret = perf_pmu_register(&pmu->pmu, pmu->name, -1);
+	if (!ret)
+		pmu->registered = true;
 	return ret;
 }
 
+static void uncore_pmu_unregister(struct intel_uncore_pmu *pmu)
+{
+	if (!pmu->registered)
+		return;
+	perf_pmu_unregister(&pmu->pmu);
+	pmu->registered = false;
+}
+
+static void __init __uncore_exit_boxes(struct intel_uncore_type *type, int cpu)
+{
+	struct intel_uncore_pmu *pmu = type->pmus;
+	struct intel_uncore_box *box;
+	int i, pkg;
+
+	if (pmu) {
+		pkg = topology_physical_package_id(cpu);
+		for (i = 0; i < type->num_boxes; i++, pmu++) {
+			box = pmu->boxes[pkg];
+			if (box)
+				uncore_box_exit(box);
+		}
+	}
+}
+
+static void __init uncore_exit_boxes(void *dummy)
+{
+	struct intel_uncore_type **types;
+
+	for (types = uncore_msr_uncores; *types; types++)
+		__uncore_exit_boxes(*types++, smp_processor_id());
+}
+
+static void uncore_free_boxes(struct intel_uncore_pmu *pmu)
+{
+	int pkg;
+
+	for (pkg = 0; pkg < max_packages; pkg++)
+		kfree(pmu->boxes[pkg]);
+	kfree(pmu->boxes);
+}
+
 static void __init uncore_type_exit(struct intel_uncore_type *type)
 {
+	struct intel_uncore_pmu *pmu = type->pmus;
 	int i;
 
-	for (i = 0; i < type->num_boxes; i++)
-		free_percpu(type->pmus[i].box);
-	kfree(type->pmus);
-	type->pmus = NULL;
+	if (pmu) {
+		for (i = 0; i < type->num_boxes; i++, pmu++) {
+			uncore_pmu_unregister(pmu);
+			uncore_free_boxes(pmu);
+		}
+		kfree(type->pmus);
+		type->pmus = NULL;
+	}
 	kfree(type->events_group);
 	type->events_group = NULL;
 }
 
 static void __init uncore_types_exit(struct intel_uncore_type **types)
 {
-	int i;
-	for (i = 0; types[i]; i++)
-		uncore_type_exit(types[i]);
+	for (; *types; types++)
+		uncore_type_exit(*types);
 }
 
-static int __init uncore_type_init(struct intel_uncore_type *type)
+static int __init uncore_type_init(struct intel_uncore_type *type, bool setid)
 {
 	struct intel_uncore_pmu *pmus;
 	struct attribute_group *attr_group;
 	struct attribute **attrs;
+	size_t size;
 	int i, j;
 
 	pmus = kzalloc(sizeof(*pmus) * type->num_boxes, GFP_KERNEL);
 	if (!pmus)
 		return -ENOMEM;
 
-	type->pmus = pmus;
+	size = max_packages * sizeof(struct intel_uncore_box *);
 
+	for (i = 0; i < type->num_boxes; i++) {
+		pmus[i].func_id	= setid ? i : -1;
+		pmus[i].pmu_idx	= i;
+		pmus[i].type	= type;
+		pmus[i].boxes	= kzalloc(size, GFP_KERNEL);
+		if (!pmus[i].boxes)
+			return -ENOMEM;
+	}
+
+	type->pmus = pmus;
 	type->unconstrainted = (struct event_constraint)
 		__EVENT_CONSTRAINT(0, (1ULL << type->num_counters) - 1,
 				0, type->num_counters, 0, 0);
 
-	for (i = 0; i < type->num_boxes; i++) {
-		pmus[i].func_id = -1;
-		pmus[i].pmu_idx = i;
-		pmus[i].type = type;
-		INIT_LIST_HEAD(&pmus[i].box_list);
-		pmus[i].box = alloc_percpu(struct intel_uncore_box *);
-		if (!pmus[i].box)
-			goto fail;
-	}
-
 	if (type->event_descs) {
-		i = 0;
-		while (type->event_descs[i].attr.attr.name)
-			i++;
+		for (i = 0; type->event_descs[i].attr.attr.name; i++);
 
 		attr_group = kzalloc(sizeof(struct attribute *) * (i + 1) +
 					sizeof(*attr_group), GFP_KERNEL);
 		if (!attr_group)
-			goto fail;
+			return -ENOMEM;
 
 		attrs = (struct attribute **)(attr_group + 1);
 		attr_group->name = "events";
@@ -831,25 +858,19 @@ static int __init uncore_type_init(struct intel_uncore_type *type)
 
 	type->pmu_group = &uncore_pmu_attr_group;
 	return 0;
-fail:
-	uncore_type_exit(type);
-	return -ENOMEM;
 }
 
-static int __init uncore_types_init(struct intel_uncore_type **types)
+static int __init
+uncore_types_init(struct intel_uncore_type **types, bool setid)
 {
-	int i, ret;
+	int ret;
 
-	for (i = 0; types[i]; i++) {
-		ret = uncore_type_init(types[i]);
+	for (; *types; types++) {
+		ret = uncore_type_init(*types, setid);
 		if (ret)
-			goto fail;
+			return ret;
 	}
 	return 0;
-fail:
-	while (--i >= 0)
-		uncore_type_exit(types[i]);
-	return ret;
 }
 
 /*
@@ -857,28 +878,28 @@ fail:
  */
 static int uncore_pci_probe(struct pci_dev *pdev, const struct pci_device_id *id)
 {
+	struct intel_uncore_type *type;
 	struct intel_uncore_pmu *pmu;
 	struct intel_uncore_box *box;
-	struct intel_uncore_type *type;
-	int phys_id;
-	bool first_box = false;
+	int phys_id, pkg, ret;
 
 	phys_id = uncore_pcibus_to_physid(pdev->bus);
 	if (phys_id < 0)
 		return -ENODEV;
 
+	pkg = topology_phys_to_logical_pkg(phys_id);
+	if (WARN_ON_ONCE(pkg < 0))
+		return -EINVAL;
+
 	if (UNCORE_PCI_DEV_TYPE(id->driver_data) == UNCORE_EXTRA_PCI_DEV) {
 		int idx = UNCORE_PCI_DEV_IDX(id->driver_data);
-		uncore_extra_pci_dev[phys_id][idx] = pdev;
+
+		uncore_extra_pci_dev[pkg].dev[idx] = pdev;
 		pci_set_drvdata(pdev, NULL);
 		return 0;
 	}
 
 	type = uncore_pci_uncores[UNCORE_PCI_DEV_TYPE(id->driver_data)];
-	box = uncore_alloc_box(type, NUMA_NO_NODE);
-	if (!box)
-		return -ENOMEM;
-
 	/*
 	 * for performance monitoring unit with multiple boxes,
 	 * each box has a different function id.
@@ -890,44 +911,60 @@ static int uncore_pci_probe(struct pci_dev *pdev, const struct pci_device_id *id
 	 * some device types. Hence PCI device idx would be 0 for all devices.
 	 * So increment pmu pointer to point to an unused array element.
 	 */
-	if (boot_cpu_data.x86_model == 87)
+	if (boot_cpu_data.x86_model == 87) {
 		while (pmu->func_id >= 0)
 			pmu++;
+	}
+
+	if (WARN_ON_ONCE(pmu->boxes[pkg] != NULL))
+		return -EINVAL;
+
+	box = uncore_alloc_box(type, NUMA_NO_NODE);
+	if (!box)
+		return -ENOMEM;
+
 	if (pmu->func_id < 0)
 		pmu->func_id = pdev->devfn;
 	else
 		WARN_ON_ONCE(pmu->func_id != pdev->devfn);
 
-	box->phys_id = phys_id;
+	atomic_inc(&box->refcnt);
+	box->pci_phys_id = phys_id;
+	box->pkgid = pkg;
 	box->pci_dev = pdev;
 	box->pmu = pmu;
 	uncore_box_init(box);
 	pci_set_drvdata(pdev, box);
 
-	raw_spin_lock(&uncore_box_lock);
-	if (list_empty(&pmu->box_list))
-		first_box = true;
-	list_add_tail(&box->list, &pmu->box_list);
-	raw_spin_unlock(&uncore_box_lock);
+	pmu->boxes[pkg] = box;
+	if (atomic_inc_return(&pmu->activeboxes) > 1)
+		return 0;
 
-	if (first_box)
-		uncore_pmu_register(pmu);
-	return 0;
+	/* First active box registers the pmu */
+	ret = uncore_pmu_register(pmu);
+	if (ret) {
+		pci_set_drvdata(pdev, NULL);
+		pmu->boxes[pkg] = NULL;
+		uncore_box_exit(box);
+		kfree(box);
+	}
+	return ret;
 }
 
 static void uncore_pci_remove(struct pci_dev *pdev)
 {
 	struct intel_uncore_box *box = pci_get_drvdata(pdev);
 	struct intel_uncore_pmu *pmu;
-	int i, cpu, phys_id;
-	bool last_box = false;
+	int i, phys_id, pkg;
 
 	phys_id = uncore_pcibus_to_physid(pdev->bus);
+	pkg = topology_phys_to_logical_pkg(phys_id);
+
 	box = pci_get_drvdata(pdev);
 	if (!box) {
 		for (i = 0; i < UNCORE_EXTRA_PCI_DEV_MAX; i++) {
-			if (uncore_extra_pci_dev[phys_id][i] == pdev) {
-				uncore_extra_pci_dev[phys_id][i] = NULL;
+			if (uncore_extra_pci_dev[pkg].dev[i] == pdev) {
+				uncore_extra_pci_dev[pkg].dev[i] = NULL;
 				break;
 			}
 		}
@@ -936,33 +973,20 @@ static void uncore_pci_remove(struct pci_dev *pdev)
 	}
 
 	pmu = box->pmu;
-	if (WARN_ON_ONCE(phys_id != box->phys_id))
+	if (WARN_ON_ONCE(phys_id != box->pci_phys_id))
 		return;
 
 	pci_set_drvdata(pdev, NULL);
-
-	raw_spin_lock(&uncore_box_lock);
-	list_del(&box->list);
-	if (list_empty(&pmu->box_list))
-		last_box = true;
-	raw_spin_unlock(&uncore_box_lock);
-
-	for_each_possible_cpu(cpu) {
-		if (*per_cpu_ptr(pmu->box, cpu) == box) {
-			*per_cpu_ptr(pmu->box, cpu) = NULL;
-			atomic_dec(&box->refcnt);
-		}
-	}
-
-	WARN_ON_ONCE(atomic_read(&box->refcnt) != 1);
+	pmu->boxes[pkg] = NULL;
+	if (atomic_dec_return(&pmu->activeboxes) == 0)
+		uncore_pmu_unregister(pmu);
+	uncore_box_exit(box);
 	kfree(box);
-
-	if (last_box)
-		perf_pmu_unregister(&pmu->pmu);
 }
 
 static int __init uncore_pci_init(void)
 {
+	size_t size;
 	int ret;
 
 	switch (boot_cpu_data.x86_model) {
@@ -999,25 +1023,40 @@ static int __init uncore_pci_init(void)
 		ret = skl_uncore_pci_init();
 		break;
 	default:
-		return 0;
+		return -ENODEV;
 	}
 
 	if (ret)
 		return ret;
 
-	ret = uncore_types_init(uncore_pci_uncores);
+	size = max_packages * sizeof(struct pci_extra_dev);
+	uncore_extra_pci_dev = kzalloc(size, GFP_KERNEL);
+	if (!uncore_extra_pci_dev) {
+		ret = -ENOMEM;
+		goto err;
+	}
+
+	ret = uncore_types_init(uncore_pci_uncores, false);
 	if (ret)
-		return ret;
+		goto errtype;
 
 	uncore_pci_driver->probe = uncore_pci_probe;
 	uncore_pci_driver->remove = uncore_pci_remove;
 
 	ret = pci_register_driver(uncore_pci_driver);
-	if (ret == 0)
-		pcidrv_registered = true;
-	else
-		uncore_types_exit(uncore_pci_uncores);
+	if (ret)
+		goto errtype;
+
+	pcidrv_registered = true;
+	return 0;
 
+errtype:
+	uncore_types_exit(uncore_pci_uncores);
+	kfree(uncore_extra_pci_dev);
+	uncore_extra_pci_dev = NULL;
+	uncore_free_pcibus_map();
+err:
+	uncore_pci_uncores = empty_uncore;
 	return ret;
 }
 
@@ -1027,173 +1066,139 @@ static void __init uncore_pci_exit(void)
 		pcidrv_registered = false;
 		pci_unregister_driver(uncore_pci_driver);
 		uncore_types_exit(uncore_pci_uncores);
-	}
-}
-
-/* CPU hot plug/unplug are serialized by cpu_add_remove_lock mutex */
-static LIST_HEAD(boxes_to_free);
-
-static void uncore_kfree_boxes(void)
-{
-	struct intel_uncore_box *box;
-
-	while (!list_empty(&boxes_to_free)) {
-		box = list_entry(boxes_to_free.next,
-				 struct intel_uncore_box, list);
-		list_del(&box->list);
-		kfree(box);
+		kfree(uncore_extra_pci_dev);
+		uncore_free_pcibus_map();
 	}
 }
 
 static void uncore_cpu_dying(int cpu)
 {
-	struct intel_uncore_type *type;
+	struct intel_uncore_type *type, **types = uncore_msr_uncores;
 	struct intel_uncore_pmu *pmu;
 	struct intel_uncore_box *box;
-	int i, j;
-
-	for (i = 0; uncore_msr_uncores[i]; i++) {
-		type = uncore_msr_uncores[i];
-		for (j = 0; j < type->num_boxes; j++) {
-			pmu = &type->pmus[j];
-			box = *per_cpu_ptr(pmu->box, cpu);
-			*per_cpu_ptr(pmu->box, cpu) = NULL;
-			if (box && atomic_dec_and_test(&box->refcnt))
-				list_add(&box->list, &boxes_to_free);
+	int i, pkg;
+
+	pkg = topology_logical_package_id(cpu);
+	for (; *types; types++) {
+		type = *types;
+		pmu = type->pmus;
+		for (i = 0; i < type->num_boxes; i++, pmu++) {
+			box = pmu->boxes[pkg];
+			if (box && atomic_dec_return(&box->refcnt) == 0)
+				uncore_box_exit(box);
 		}
 	}
 }
 
-static int uncore_cpu_starting(int cpu)
+static void uncore_cpu_starting(int cpu, bool init)
 {
-	struct intel_uncore_type *type;
+	struct intel_uncore_type *type, **types = uncore_msr_uncores;
 	struct intel_uncore_pmu *pmu;
-	struct intel_uncore_box *box, *exist;
-	int i, j, k, phys_id;
-
-	phys_id = topology_physical_package_id(cpu);
-
-	for (i = 0; uncore_msr_uncores[i]; i++) {
-		type = uncore_msr_uncores[i];
-		for (j = 0; j < type->num_boxes; j++) {
-			pmu = &type->pmus[j];
-			box = *per_cpu_ptr(pmu->box, cpu);
-			/* called by uncore_cpu_init? */
-			if (box && box->phys_id >= 0) {
-				uncore_box_init(box);
-				continue;
-			}
+	struct intel_uncore_box *box;
+	int i, pkg, ncpus = 1;
 
-			for_each_online_cpu(k) {
-				exist = *per_cpu_ptr(pmu->box, k);
-				if (exist && exist->phys_id == phys_id) {
-					atomic_inc(&exist->refcnt);
-					*per_cpu_ptr(pmu->box, cpu) = exist;
-					if (box) {
-						list_add(&box->list,
-							 &boxes_to_free);
-						box = NULL;
-					}
-					break;
-				}
-			}
+	if (init) {
+		/*
+		 * On init we get the number of online cpus in the package
+		 * and set refcount for all of them.
+		 */
+		ncpus = cpumask_weight(topology_core_cpumask(cpu));
+	}
 
-			if (box) {
-				box->phys_id = phys_id;
+	pkg = topology_logical_package_id(cpu);
+	for (; *types; types++) {
+		type = *types;
+		pmu = type->pmus;
+		for (i = 0; i < type->num_boxes; i++, pmu++) {
+			box = pmu->boxes[pkg];
+			if (!box)
+				continue;
+			/* The first cpu on a package activates the box */
+			if (atomic_add_return(ncpus, &box->refcnt) == ncpus)
 				uncore_box_init(box);
-			}
 		}
 	}
-	return 0;
 }
 
-static int uncore_cpu_prepare(int cpu, int phys_id)
+static int uncore_cpu_prepare(int cpu)
 {
-	struct intel_uncore_type *type;
+	struct intel_uncore_type *type, **types = uncore_msr_uncores;
 	struct intel_uncore_pmu *pmu;
 	struct intel_uncore_box *box;
-	int i, j;
-
-	for (i = 0; uncore_msr_uncores[i]; i++) {
-		type = uncore_msr_uncores[i];
-		for (j = 0; j < type->num_boxes; j++) {
-			pmu = &type->pmus[j];
-			if (pmu->func_id < 0)
-				pmu->func_id = j;
-
+	int i, pkg;
+
+	pkg = topology_logical_package_id(cpu);
+	for (; *types; types++) {
+		type = *types;
+		pmu = type->pmus;
+		for (i = 0; i < type->num_boxes; i++, pmu++) {
+			if (pmu->boxes[pkg])
+				continue;
+			/* First cpu of a package allocates the box */
 			box = uncore_alloc_box(type, cpu_to_node(cpu));
 			if (!box)
 				return -ENOMEM;
-
 			box->pmu = pmu;
-			box->phys_id = phys_id;
-			*per_cpu_ptr(pmu->box, cpu) = box;
+			box->pkgid = pkg;
+			pmu->boxes[pkg] = box;
 		}
 	}
 	return 0;
 }
 
-static void
-uncore_change_context(struct intel_uncore_type **uncores, int old_cpu, int new_cpu)
+static void uncore_change_type_ctx(struct intel_uncore_type *type, int old_cpu,
+				   int new_cpu)
 {
-	struct intel_uncore_type *type;
-	struct intel_uncore_pmu *pmu;
+	struct intel_uncore_pmu *pmu = type->pmus;
 	struct intel_uncore_box *box;
-	int i, j;
+	int i, pkg;
 
-	for (i = 0; uncores[i]; i++) {
-		type = uncores[i];
-		for (j = 0; j < type->num_boxes; j++) {
-			pmu = &type->pmus[j];
-			if (old_cpu < 0)
-				box = uncore_pmu_to_box(pmu, new_cpu);
-			else
-				box = uncore_pmu_to_box(pmu, old_cpu);
-			if (!box)
-				continue;
-
-			if (old_cpu < 0) {
-				WARN_ON_ONCE(box->cpu != -1);
-				box->cpu = new_cpu;
-				continue;
-			}
+	pkg = topology_logical_package_id(old_cpu < 0 ? new_cpu : old_cpu);
+	for (i = 0; i < type->num_boxes; i++, pmu++) {
+		box = pmu->boxes[pkg];
+		if (!box)
+			continue;
 
-			WARN_ON_ONCE(box->cpu != old_cpu);
-			if (new_cpu >= 0) {
-				uncore_pmu_cancel_hrtimer(box);
-				perf_pmu_migrate_context(&pmu->pmu,
-						old_cpu, new_cpu);
-				box->cpu = new_cpu;
-			} else {
-				box->cpu = -1;
-			}
+		if (old_cpu < 0) {
+			WARN_ON_ONCE(box->cpu != -1);
+			box->cpu = new_cpu;
+			continue;
 		}
+
+		WARN_ON_ONCE(box->cpu != old_cpu);
+		box->cpu = -1;
+		if (new_cpu < 0)
+			continue;
+
+		uncore_pmu_cancel_hrtimer(box);
+		perf_pmu_migrate_context(&pmu->pmu, old_cpu, new_cpu);
+		box->cpu = new_cpu;
 	}
 }
 
+static void uncore_change_context(struct intel_uncore_type **uncores,
+				  int old_cpu, int new_cpu)
+{
+	for (; *uncores; uncores++)
+		uncore_change_type_ctx(*uncores, old_cpu, new_cpu);
+}
+
 static void uncore_event_exit_cpu(int cpu)
 {
-	int i, phys_id, target;
+	int target;
 
-	/* if exiting cpu is used for collecting uncore events */
+	/* Check if exiting cpu is used for collecting uncore events */
 	if (!cpumask_test_and_clear_cpu(cpu, &uncore_cpu_mask))
 		return;
 
-	/* find a new cpu to collect uncore events */
-	phys_id = topology_physical_package_id(cpu);
-	target = -1;
-	for_each_online_cpu(i) {
-		if (i == cpu)
-			continue;
-		if (phys_id == topology_physical_package_id(i)) {
-			target = i;
-			break;
-		}
-	}
+	/* Find a new cpu to collect uncore events */
+	target = cpumask_any_but(topology_core_cpumask(cpu), cpu);
 
-	/* migrate uncore events to the new cpu */
-	if (target >= 0)
+	/* Migrate uncore events to the new target */
+	if (target < nr_cpu_ids)
 		cpumask_set_cpu(target, &uncore_cpu_mask);
+	else
+		target = -1;
 
 	uncore_change_context(uncore_msr_uncores, cpu, target);
 	uncore_change_context(uncore_pci_uncores, cpu, target);
@@ -1201,13 +1206,15 @@ static void uncore_event_exit_cpu(int cpu)
 
 static void uncore_event_init_cpu(int cpu)
 {
-	int i, phys_id;
+	int target;
 
-	phys_id = topology_physical_package_id(cpu);
-	for_each_cpu(i, &uncore_cpu_mask) {
-		if (phys_id == topology_physical_package_id(i))
-			return;
-	}
+	/*
+	 * Check if there is an online cpu in the package
+	 * which collects uncore events already.
+	 */
+	target = cpumask_any_and(&uncore_cpu_mask, topology_core_cpumask(cpu));
+	if (target < nr_cpu_ids)
+		return;
 
 	cpumask_set_cpu(cpu, &uncore_cpu_mask);
 
@@ -1220,39 +1227,25 @@ static int uncore_cpu_notifier(struct notifier_block *self,
 {
 	unsigned int cpu = (long)hcpu;
 
-	/* allocate/free data structure for uncore box */
 	switch (action & ~CPU_TASKS_FROZEN) {
 	case CPU_UP_PREPARE:
-		uncore_cpu_prepare(cpu, -1);
-		break;
+		return notifier_from_errno(uncore_cpu_prepare(cpu));
+
 	case CPU_STARTING:
-		uncore_cpu_starting(cpu);
+		uncore_cpu_starting(cpu, false);
+	case CPU_DOWN_FAILED:
+		uncore_event_init_cpu(cpu);
 		break;
+
 	case CPU_UP_CANCELED:
 	case CPU_DYING:
 		uncore_cpu_dying(cpu);
 		break;
-	case CPU_ONLINE:
-	case CPU_DEAD:
-		uncore_kfree_boxes();
-		break;
-	default:
-		break;
-	}
 
-	/* select the cpu that collects uncore events */
-	switch (action & ~CPU_TASKS_FROZEN) {
-	case CPU_DOWN_FAILED:
-	case CPU_STARTING:
-		uncore_event_init_cpu(cpu);
-		break;
 	case CPU_DOWN_PREPARE:
 		uncore_event_exit_cpu(cpu);
 		break;
-	default:
-		break;
 	}
-
 	return NOTIFY_OK;
 }
 
@@ -1265,9 +1258,29 @@ static struct notifier_block uncore_cpu_nb = {
 	.priority	= CPU_PRI_PERF + 1,
 };
 
-static void __init uncore_cpu_setup(void *dummy)
+static int __init type_pmu_register(struct intel_uncore_type *type)
 {
-	uncore_cpu_starting(smp_processor_id());
+	int i, ret;
+
+	for (i = 0; i < type->num_boxes; i++) {
+		ret = uncore_pmu_register(&type->pmus[i]);
+		if (ret)
+			return ret;
+	}
+	return 0;
+}
+
+static int __init uncore_msr_pmus_register(void)
+{
+	struct intel_uncore_type **types = uncore_msr_uncores;
+	int ret;
+
+	for (; *types; types++) {
+		ret = type_pmu_register(*types);
+		if (ret)
+			return ret;
+	}
+	return 0;
 }
 
 static int __init uncore_cpu_init(void)
@@ -1311,71 +1324,61 @@ static int __init uncore_cpu_init(void)
 		knl_uncore_cpu_init();
 		break;
 	default:
-		return 0;
+		return -ENODEV;
 	}
 
-	ret = uncore_types_init(uncore_msr_uncores);
+	ret = uncore_types_init(uncore_msr_uncores, true);
 	if (ret)
-		return ret;
+		goto err;
 
+	ret = uncore_msr_pmus_register();
+	if (ret)
+		goto err;
 	return 0;
+err:
+	uncore_types_exit(uncore_msr_uncores);
+	uncore_msr_uncores = empty_uncore;
+	return ret;
 }
 
-static int __init uncore_pmus_register(void)
+static void __init uncore_cpu_setup(void *dummy)
 {
-	struct intel_uncore_pmu *pmu;
-	struct intel_uncore_type *type;
-	int i, j;
-
-	for (i = 0; uncore_msr_uncores[i]; i++) {
-		type = uncore_msr_uncores[i];
-		for (j = 0; j < type->num_boxes; j++) {
-			pmu = &type->pmus[j];
-			uncore_pmu_register(pmu);
-		}
-	}
-
-	return 0;
+	uncore_cpu_starting(smp_processor_id(), true);
 }
 
-static void __init uncore_cpumask_init(void)
-{
-	int cpu;
-
-	/*
-	 * ony invoke once from msr or pci init code
-	 */
-	if (!cpumask_empty(&uncore_cpu_mask))
-		return;
+/* Lazy to avoid allocation of a few bytes for the normal case */
+static __initdata DECLARE_BITMAP(packages, MAX_LOCAL_APIC);
 
-	cpu_notifier_register_begin();
+static int __init uncore_cpumask_init(bool msr)
+{
+	unsigned int cpu;
 
 	for_each_online_cpu(cpu) {
-		int i, phys_id = topology_physical_package_id(cpu);
+		unsigned int pkg = topology_logical_package_id(cpu);
+		int ret;
 
-		for_each_cpu(i, &uncore_cpu_mask) {
-			if (phys_id == topology_physical_package_id(i)) {
-				phys_id = -1;
-				break;
-			}
-		}
-		if (phys_id < 0)
+		if (test_and_set_bit(pkg, packages))
 			continue;
-
-		uncore_cpu_prepare(cpu, phys_id);
+		/*
+		 * The first online cpu of each package allocates and takes
+		 * the refcounts for all other online cpus in that package.
+		 * If msrs are not enabled no allocation is required.
+		 */
+		if (msr) {
+			ret = uncore_cpu_prepare(cpu);
+			if (ret)
+				return ret;
+		}
 		uncore_event_init_cpu(cpu);
+		smp_call_function_single(cpu, uncore_cpu_setup, NULL, 1);
 	}
-	on_each_cpu(uncore_cpu_setup, NULL, 1);
-
 	__register_cpu_notifier(&uncore_cpu_nb);
-
-	cpu_notifier_register_done();
+	return 0;
 }
 
-
 static int __init intel_uncore_init(void)
 {
-	int ret;
+	int pret, cret, ret;
 
 	if (boot_cpu_data.x86_vendor != X86_VENDOR_INTEL)
 		return -ENODEV;
@@ -1383,19 +1386,27 @@ static int __init intel_uncore_init(void)
 	if (cpu_has_hypervisor)
 		return -ENODEV;
 
-	ret = uncore_pci_init();
-	if (ret)
-		goto fail;
-	ret = uncore_cpu_init();
-	if (ret) {
-		uncore_pci_exit();
-		goto fail;
-	}
-	uncore_cpumask_init();
+	max_packages = topology_max_packages();
+
+	pret = uncore_pci_init();
+	cret = uncore_cpu_init();
 
-	uncore_pmus_register();
+	if (cret && pret)
+		return -ENODEV;
+
+	cpu_notifier_register_begin();
+	ret = uncore_cpumask_init(!cret);
+	if (ret)
+		goto err;
+	cpu_notifier_register_done();
 	return 0;
-fail:
+
+err:
+	/* Undo box->init_box() */
+	on_each_cpu_mask(&uncore_cpu_mask, uncore_exit_boxes, NULL, 1);
+	uncore_types_exit(uncore_msr_uncores);
+	uncore_pci_exit();
+	cpu_notifier_register_done();
 	return ret;
 }
 device_initcall(intel_uncore_init);
diff --git a/arch/x86/kernel/cpu/perf_event_intel_uncore.h b/arch/x86/events/intel/uncore.h
index a7086b8..79766b9 100644
--- a/arch/x86/kernel/cpu/perf_event_intel_uncore.h
+++ b/arch/x86/events/intel/uncore.h
@@ -1,8 +1,10 @@
 #include <linux/module.h>
 #include <linux/slab.h>
 #include <linux/pci.h>
+#include <asm/apicdef.h>
+
 #include <linux/perf_event.h>
-#include "perf_event.h"
+#include "../perf_event.h"
 
 #define UNCORE_PMU_NAME_LEN		32
 #define UNCORE_PMU_HRTIMER_INTERVAL	(60LL * NSEC_PER_SEC)
@@ -19,11 +21,12 @@
 #define UNCORE_EXTRA_PCI_DEV		0xff
 #define UNCORE_EXTRA_PCI_DEV_MAX	3
 
-/* support up to 8 sockets */
-#define UNCORE_SOCKET_MAX		8
-
 #define UNCORE_EVENT_CONSTRAINT(c, n) EVENT_CONSTRAINT(c, n, 0xff)
 
+struct pci_extra_dev {
+	struct pci_dev *dev[UNCORE_EXTRA_PCI_DEV_MAX];
+};
+
 struct intel_uncore_ops;
 struct intel_uncore_pmu;
 struct intel_uncore_box;
@@ -61,6 +64,7 @@ struct intel_uncore_type {
 
 struct intel_uncore_ops {
 	void (*init_box)(struct intel_uncore_box *);
+	void (*exit_box)(struct intel_uncore_box *);
 	void (*disable_box)(struct intel_uncore_box *);
 	void (*enable_box)(struct intel_uncore_box *);
 	void (*disable_event)(struct intel_uncore_box *, struct perf_event *);
@@ -73,13 +77,14 @@ struct intel_uncore_ops {
 };
 
 struct intel_uncore_pmu {
-	struct pmu pmu;
-	char name[UNCORE_PMU_NAME_LEN];
-	int pmu_idx;
-	int func_id;
-	struct intel_uncore_type *type;
-	struct intel_uncore_box ** __percpu box;
-	struct list_head box_list;
+	struct pmu			pmu;
+	char				name[UNCORE_PMU_NAME_LEN];
+	int				pmu_idx;
+	int				func_id;
+	bool				registered;
+	atomic_t			activeboxes;
+	struct intel_uncore_type	*type;
+	struct intel_uncore_box		**boxes;
 };
 
 struct intel_uncore_extra_reg {
@@ -89,7 +94,8 @@ struct intel_uncore_extra_reg {
 };
 
 struct intel_uncore_box {
-	int phys_id;
+	int pci_phys_id;
+	int pkgid;
 	int n_active;	/* number of active events */
 	int n_events;
 	int cpu;	/* cpu to collect events */
@@ -123,7 +129,6 @@ struct pci2phy_map {
 	int pbus_to_physid[256];
 };
 
-int uncore_pcibus_to_physid(struct pci_bus *bus);
 struct pci2phy_map *__find_pci2phy_map(int segment);
 
 ssize_t uncore_event_show(struct kobject *kobj,
@@ -305,14 +310,30 @@ static inline void uncore_box_init(struct intel_uncore_box *box)
 	}
 }
 
+static inline void uncore_box_exit(struct intel_uncore_box *box)
+{
+	if (test_and_clear_bit(UNCORE_BOX_FLAG_INITIATED, &box->flags)) {
+		if (box->pmu->type->ops->exit_box)
+			box->pmu->type->ops->exit_box(box);
+	}
+}
+
 static inline bool uncore_box_is_fake(struct intel_uncore_box *box)
 {
-	return (box->phys_id < 0);
+	return (box->pkgid < 0);
+}
+
+static inline struct intel_uncore_pmu *uncore_event_to_pmu(struct perf_event *event)
+{
+	return container_of(event->pmu, struct intel_uncore_pmu, pmu);
+}
+
+static inline struct intel_uncore_box *uncore_event_to_box(struct perf_event *event)
+{
+	return event->pmu_private;
 }
 
-struct intel_uncore_pmu *uncore_event_to_pmu(struct perf_event *event);
 struct intel_uncore_box *uncore_pmu_to_box(struct intel_uncore_pmu *pmu, int cpu);
-struct intel_uncore_box *uncore_event_to_box(struct perf_event *event);
 u64 uncore_msr_read_counter(struct intel_uncore_box *box, struct perf_event *event);
 void uncore_pmu_start_hrtimer(struct intel_uncore_box *box);
 void uncore_pmu_cancel_hrtimer(struct intel_uncore_box *box);
@@ -328,7 +349,7 @@ extern struct intel_uncore_type **uncore_pci_uncores;
 extern struct pci_driver *uncore_pci_driver;
 extern raw_spinlock_t pci2phy_map_lock;
 extern struct list_head pci2phy_map_head;
-extern struct pci_dev *uncore_extra_pci_dev[UNCORE_SOCKET_MAX][UNCORE_EXTRA_PCI_DEV_MAX];
+extern struct pci_extra_dev *uncore_extra_pci_dev;
 extern struct event_constraint uncore_constraint_empty;
 
 /* perf_event_intel_uncore_snb.c */
diff --git a/arch/x86/kernel/cpu/perf_event_intel_uncore_nhmex.c b/arch/x86/events/intel/uncore_nhmex.c
index 2749965..cda5693 100644
--- a/arch/x86/kernel/cpu/perf_event_intel_uncore_nhmex.c
+++ b/arch/x86/events/intel/uncore_nhmex.c
@@ -1,5 +1,5 @@
 /* Nehalem-EX/Westmere-EX uncore support */
-#include "perf_event_intel_uncore.h"
+#include "uncore.h"
 
 /* NHM-EX event control */
 #define NHMEX_PMON_CTL_EV_SEL_MASK	0x000000ff
@@ -201,6 +201,11 @@ static void nhmex_uncore_msr_init_box(struct intel_uncore_box *box)
 	wrmsrl(NHMEX_U_MSR_PMON_GLOBAL_CTL, NHMEX_U_PMON_GLOBAL_EN_ALL);
 }
 
+static void nhmex_uncore_msr_exit_box(struct intel_uncore_box *box)
+{
+	wrmsrl(NHMEX_U_MSR_PMON_GLOBAL_CTL, 0);
+}
+
 static void nhmex_uncore_msr_disable_box(struct intel_uncore_box *box)
 {
 	unsigned msr = uncore_msr_box_ctl(box);
@@ -250,6 +255,7 @@ static void nhmex_uncore_msr_enable_event(struct intel_uncore_box *box, struct p
 
 #define NHMEX_UNCORE_OPS_COMMON_INIT()				\
 	.init_box	= nhmex_uncore_msr_init_box,		\
+	.exit_box	= nhmex_uncore_msr_exit_box,		\
 	.disable_box	= nhmex_uncore_msr_disable_box,		\
 	.enable_box	= nhmex_uncore_msr_enable_box,		\
 	.disable_event	= nhmex_uncore_msr_disable_event,	\
diff --git a/arch/x86/kernel/cpu/perf_event_intel_uncore_snb.c b/arch/x86/events/intel/uncore_snb.c
index 2bd030d..96531d2 100644
--- a/arch/x86/kernel/cpu/perf_event_intel_uncore_snb.c
+++ b/arch/x86/events/intel/uncore_snb.c
@@ -1,5 +1,5 @@
 /* Nehalem/SandBridge/Haswell uncore support */
-#include "perf_event_intel_uncore.h"
+#include "uncore.h"
 
 /* Uncore IMC PCI IDs */
 #define PCI_DEVICE_ID_INTEL_SNB_IMC	0x0100
@@ -95,6 +95,12 @@ static void snb_uncore_msr_init_box(struct intel_uncore_box *box)
 	}
 }
 
+static void snb_uncore_msr_exit_box(struct intel_uncore_box *box)
+{
+	if (box->pmu->pmu_idx == 0)
+		wrmsrl(SNB_UNC_PERF_GLOBAL_CTL, 0);
+}
+
 static struct uncore_event_desc snb_uncore_events[] = {
 	INTEL_UNCORE_EVENT_DESC(clockticks, "event=0xff,umask=0x00"),
 	{ /* end: all zeroes */ },
@@ -116,6 +122,7 @@ static struct attribute_group snb_uncore_format_group = {
 
 static struct intel_uncore_ops snb_uncore_msr_ops = {
 	.init_box	= snb_uncore_msr_init_box,
+	.exit_box	= snb_uncore_msr_exit_box,
 	.disable_event	= snb_uncore_msr_disable_event,
 	.enable_event	= snb_uncore_msr_enable_event,
 	.read_counter	= uncore_msr_read_counter,
@@ -231,6 +238,11 @@ static void snb_uncore_imc_init_box(struct intel_uncore_box *box)
 	box->hrtimer_duration = UNCORE_SNB_IMC_HRTIMER_INTERVAL;
 }
 
+static void snb_uncore_imc_exit_box(struct intel_uncore_box *box)
+{
+	iounmap(box->io_addr);
+}
+
 static void snb_uncore_imc_enable_box(struct intel_uncore_box *box)
 {}
 
@@ -301,6 +313,7 @@ static int snb_uncore_imc_event_init(struct perf_event *event)
 		return -EINVAL;
 
 	event->cpu = box->cpu;
+	event->pmu_private = box;
 
 	event->hw.idx = -1;
 	event->hw.last_tag = ~0ULL;
@@ -458,6 +471,7 @@ static struct pmu snb_uncore_imc_pmu = {
 
 static struct intel_uncore_ops snb_uncore_imc_ops = {
 	.init_box	= snb_uncore_imc_init_box,
+	.exit_box	= snb_uncore_imc_exit_box,
 	.enable_box	= snb_uncore_imc_enable_box,
 	.disable_box	= snb_uncore_imc_disable_box,
 	.disable_event	= snb_uncore_imc_disable_event,
diff --git a/arch/x86/kernel/cpu/perf_event_intel_uncore_snbep.c b/arch/x86/events/intel/uncore_snbep.c
index 33acb88..93f6bd9 100644
--- a/arch/x86/kernel/cpu/perf_event_intel_uncore_snbep.c
+++ b/arch/x86/events/intel/uncore_snbep.c
@@ -1,6 +1,5 @@
 /* SandyBridge-EP/IvyTown uncore support */
-#include "perf_event_intel_uncore.h"
-
+#include "uncore.h"
 
 /* SNB-EP Box level control */
 #define SNBEP_PMON_BOX_CTL_RST_CTRL	(1 << 0)
@@ -987,7 +986,9 @@ static void snbep_qpi_enable_event(struct intel_uncore_box *box, struct perf_eve
 
 	if (reg1->idx != EXTRA_REG_NONE) {
 		int idx = box->pmu->pmu_idx + SNBEP_PCI_QPI_PORT0_FILTER;
-		struct pci_dev *filter_pdev = uncore_extra_pci_dev[box->phys_id][idx];
+		int pkg = topology_phys_to_logical_pkg(box->pci_phys_id);
+		struct pci_dev *filter_pdev = uncore_extra_pci_dev[pkg].dev[idx];
+
 		if (filter_pdev) {
 			pci_write_config_dword(filter_pdev, reg1->reg,
 						(u32)reg1->config);
@@ -2521,14 +2522,16 @@ static struct intel_uncore_type *hswep_msr_uncores[] = {
 
 void hswep_uncore_cpu_init(void)
 {
+	int pkg = topology_phys_to_logical_pkg(0);
+
 	if (hswep_uncore_cbox.num_boxes > boot_cpu_data.x86_max_cores)
 		hswep_uncore_cbox.num_boxes = boot_cpu_data.x86_max_cores;
 
 	/* Detect 6-8 core systems with only two SBOXes */
-	if (uncore_extra_pci_dev[0][HSWEP_PCI_PCU_3]) {
+	if (uncore_extra_pci_dev[pkg].dev[HSWEP_PCI_PCU_3]) {
 		u32 capid4;
 
-		pci_read_config_dword(uncore_extra_pci_dev[0][HSWEP_PCI_PCU_3],
+		pci_read_config_dword(uncore_extra_pci_dev[pkg].dev[HSWEP_PCI_PCU_3],
 				      0x94, &capid4);
 		if (((capid4 >> 6) & 0x3) == 0)
 			hswep_uncore_sbox.num_boxes = 2;
@@ -2875,11 +2878,13 @@ static struct intel_uncore_type bdx_uncore_sbox = {
 	.format_group		= &hswep_uncore_sbox_format_group,
 };
 
+#define BDX_MSR_UNCORE_SBOX	3
+
 static struct intel_uncore_type *bdx_msr_uncores[] = {
 	&bdx_uncore_ubox,
 	&bdx_uncore_cbox,
-	&bdx_uncore_sbox,
 	&hswep_uncore_pcu,
+	&bdx_uncore_sbox,
 	NULL,
 };
 
@@ -2888,6 +2893,10 @@ void bdx_uncore_cpu_init(void)
 	if (bdx_uncore_cbox.num_boxes > boot_cpu_data.x86_max_cores)
 		bdx_uncore_cbox.num_boxes = boot_cpu_data.x86_max_cores;
 	uncore_msr_uncores = bdx_msr_uncores;
+
+	/* BDX-DE doesn't have SBOX */
+	if (boot_cpu_data.x86_model == 86)
+		uncore_msr_uncores[BDX_MSR_UNCORE_SBOX] = NULL;
 }
 
 static struct intel_uncore_type bdx_uncore_ha = {
diff --git a/arch/x86/kernel/cpu/perf_event_msr.c b/arch/x86/events/msr.c
index ec863b9..ec863b9 100644
--- a/arch/x86/kernel/cpu/perf_event_msr.c
+++ b/arch/x86/events/msr.c
diff --git a/arch/x86/kernel/cpu/perf_event.h b/arch/x86/events/perf_event.h
index 7bb61e3..68155ca 100644
--- a/arch/x86/kernel/cpu/perf_event.h
+++ b/arch/x86/events/perf_event.h
@@ -586,6 +586,7 @@ struct x86_pmu {
 			pebs_broken	:1,
 			pebs_prec_dist	:1;
 	int		pebs_record_size;
+	int		pebs_buffer_size;
 	void		(*drain_pebs)(struct pt_regs *regs);
 	struct event_constraint *pebs_constraints;
 	void		(*pebs_aliases)(struct perf_event *event);
@@ -860,6 +861,8 @@ extern struct event_constraint intel_ivb_pebs_event_constraints[];
 
 extern struct event_constraint intel_hsw_pebs_event_constraints[];
 
+extern struct event_constraint intel_bdw_pebs_event_constraints[];
+
 extern struct event_constraint intel_skl_pebs_event_constraints[];
 
 struct event_constraint *intel_pebs_constraints(struct perf_event *event);
@@ -904,6 +907,8 @@ void intel_pmu_lbr_init_skl(void);
 
 void intel_pmu_lbr_init_knl(void);
 
+void intel_pmu_pebs_data_source_nhm(void);
+
 int intel_pmu_setup_lbr_filter(struct perf_event *event);
 
 void intel_pt_interrupt(void);
diff --git a/arch/x86/include/asm/elf.h b/arch/x86/include/asm/elf.h
index 1514753..15340e3 100644
--- a/arch/x86/include/asm/elf.h
+++ b/arch/x86/include/asm/elf.h
@@ -256,7 +256,7 @@ extern int force_personality32;
    instruction set this CPU supports.  This could be done in user space,
    but it's not easy, and we've already done it here.  */
 
-#define ELF_HWCAP		(boot_cpu_data.x86_capability[0])
+#define ELF_HWCAP		(boot_cpu_data.x86_capability[CPUID_1_EDX])
 
 /* This yields a string that ld.so will use to load implementation
    specific libraries for optimization.  This is more specific in
diff --git a/arch/x86/include/asm/perf_event.h b/arch/x86/include/asm/perf_event.h
index 7bcb861..5a2ed3e 100644
--- a/arch/x86/include/asm/perf_event.h
+++ b/arch/x86/include/asm/perf_event.h
@@ -165,6 +165,7 @@ struct x86_pmu_capability {
 #define GLOBAL_STATUS_ASIF				BIT_ULL(60)
 #define GLOBAL_STATUS_COUNTERS_FROZEN			BIT_ULL(59)
 #define GLOBAL_STATUS_LBRS_FROZEN			BIT_ULL(58)
+#define GLOBAL_STATUS_TRACE_TOPAPMI			BIT_ULL(55)
 
 /*
  * IBS cpuid feature detection
diff --git a/arch/x86/include/asm/processor.h b/arch/x86/include/asm/processor.h
index 20c11d1..813384e 100644
--- a/arch/x86/include/asm/processor.h
+++ b/arch/x86/include/asm/processor.h
@@ -129,6 +129,8 @@ struct cpuinfo_x86 {
 	u16			booted_cores;
 	/* Physical processor id: */
 	u16			phys_proc_id;
+	/* Logical processor id: */
+	u16			logical_proc_id;
 	/* Core id: */
 	u16			cpu_core_id;
 	/* Compute unit id */
diff --git a/arch/x86/include/asm/topology.h b/arch/x86/include/asm/topology.h
index 0fb4648..7f991bd5 100644
--- a/arch/x86/include/asm/topology.h
+++ b/arch/x86/include/asm/topology.h
@@ -119,12 +119,23 @@ static inline void setup_node_to_cpumask_map(void) { }
 
 extern const struct cpumask *cpu_coregroup_mask(int cpu);
 
+#define topology_logical_package_id(cpu)	(cpu_data(cpu).logical_proc_id)
 #define topology_physical_package_id(cpu)	(cpu_data(cpu).phys_proc_id)
 #define topology_core_id(cpu)			(cpu_data(cpu).cpu_core_id)
 
 #ifdef ENABLE_TOPO_DEFINES
 #define topology_core_cpumask(cpu)		(per_cpu(cpu_core_map, cpu))
 #define topology_sibling_cpumask(cpu)		(per_cpu(cpu_sibling_map, cpu))
+
+extern unsigned int __max_logical_packages;
+#define topology_max_packages()			(__max_logical_packages)
+int topology_update_package_map(unsigned int apicid, unsigned int cpu);
+extern int topology_phys_to_logical_pkg(unsigned int pkg);
+#else
+#define topology_max_packages()			(1)
+static inline int
+topology_update_package_map(unsigned int apicid, unsigned int cpu) { return 0; }
+static inline int topology_phys_to_logical_pkg(unsigned int pkg) { return 0; }
 #endif
 
 static inline void arch_fix_phys_package_id(int num, u32 slot)
diff --git a/arch/x86/kernel/apic/apic.c b/arch/x86/kernel/apic/apic.c
index 8a5cdda..531b961 100644
--- a/arch/x86/kernel/apic/apic.c
+++ b/arch/x86/kernel/apic/apic.c
@@ -2078,6 +2078,20 @@ int generic_processor_info(int apicid, int version)
 		cpu = cpumask_next_zero(-1, cpu_present_mask);
 
 	/*
+	 * This can happen on physical hotplug. The sanity check at boot time
+	 * is done from native_smp_prepare_cpus() after num_possible_cpus() is
+	 * established.
+	 */
+	if (topology_update_package_map(apicid, cpu) < 0) {
+		int thiscpu = max + disabled_cpus;
+
+		pr_warning("ACPI: Package limit reached. Processor %d/0x%x ignored.\n",
+			   thiscpu, apicid);
+		disabled_cpus++;
+		return -ENOSPC;
+	}
+
+	/*
 	 * Validate version
 	 */
 	if (version == 0x0) {
diff --git a/arch/x86/kernel/cpu/Makefile b/arch/x86/kernel/cpu/Makefile
index 5803130..7a60424 100644
--- a/arch/x86/kernel/cpu/Makefile
+++ b/arch/x86/kernel/cpu/Makefile
@@ -30,33 +30,11 @@ obj-$(CONFIG_CPU_SUP_CENTAUR)		+= centaur.o
 obj-$(CONFIG_CPU_SUP_TRANSMETA_32)	+= transmeta.o
 obj-$(CONFIG_CPU_SUP_UMC_32)		+= umc.o
 
-obj-$(CONFIG_PERF_EVENTS)		+= perf_event.o
-
-ifdef CONFIG_PERF_EVENTS
-obj-$(CONFIG_CPU_SUP_AMD)		+= perf_event_amd.o perf_event_amd_uncore.o
-ifdef CONFIG_AMD_IOMMU
-obj-$(CONFIG_CPU_SUP_AMD)		+= perf_event_amd_iommu.o
-endif
-obj-$(CONFIG_CPU_SUP_INTEL)		+= perf_event_p6.o perf_event_knc.o perf_event_p4.o
-obj-$(CONFIG_CPU_SUP_INTEL)		+= perf_event_intel_lbr.o perf_event_intel_ds.o perf_event_intel.o
-obj-$(CONFIG_CPU_SUP_INTEL)		+= perf_event_intel_rapl.o perf_event_intel_cqm.o
-obj-$(CONFIG_CPU_SUP_INTEL)		+= perf_event_intel_pt.o perf_event_intel_bts.o
-obj-$(CONFIG_CPU_SUP_INTEL)		+= perf_event_intel_cstate.o
-
-obj-$(CONFIG_PERF_EVENTS_INTEL_UNCORE)	+= perf_event_intel_uncore.o \
-					   perf_event_intel_uncore_snb.o \
-					   perf_event_intel_uncore_snbep.o \
-					   perf_event_intel_uncore_nhmex.o
-obj-$(CONFIG_CPU_SUP_INTEL)		+= perf_event_msr.o
-obj-$(CONFIG_CPU_SUP_AMD)		+= perf_event_msr.o
-endif
-
-
 obj-$(CONFIG_X86_MCE)			+= mcheck/
 obj-$(CONFIG_MTRR)			+= mtrr/
 obj-$(CONFIG_MICROCODE)			+= microcode/
 
-obj-$(CONFIG_X86_LOCAL_APIC)		+= perfctr-watchdog.o perf_event_amd_ibs.o
+obj-$(CONFIG_X86_LOCAL_APIC)		+= perfctr-watchdog.o
 
 obj-$(CONFIG_HYPERVISOR_GUEST)		+= vmware.o hypervisor.o mshyperv.o
 
diff --git a/arch/x86/kernel/cpu/amd.c b/arch/x86/kernel/cpu/amd.c
index a07956a..97c59fd 100644
--- a/arch/x86/kernel/cpu/amd.c
+++ b/arch/x86/kernel/cpu/amd.c
@@ -117,7 +117,7 @@ static void init_amd_k6(struct cpuinfo_x86 *c)
 		void (*f_vide)(void);
 		u64 d, d2;
 
-		printk(KERN_INFO "AMD K6 stepping B detected - ");
+		pr_info("AMD K6 stepping B detected - ");
 
 		/*
 		 * It looks like AMD fixed the 2.6.2 bug and improved indirect
@@ -133,10 +133,9 @@ static void init_amd_k6(struct cpuinfo_x86 *c)
 		d = d2-d;
 
 		if (d > 20*K6_BUG_LOOP)
-			printk(KERN_CONT
-				"system stability may be impaired when more than 32 MB are used.\n");
+			pr_cont("system stability may be impaired when more than 32 MB are used.\n");
 		else
-			printk(KERN_CONT "probably OK (after B9730xxxx).\n");
+			pr_cont("probably OK (after B9730xxxx).\n");
 	}
 
 	/* K6 with old style WHCR */
@@ -154,7 +153,7 @@ static void init_amd_k6(struct cpuinfo_x86 *c)
 			wbinvd();
 			wrmsr(MSR_K6_WHCR, l, h);
 			local_irq_restore(flags);
-			printk(KERN_INFO "Enabling old style K6 write allocation for %d Mb\n",
+			pr_info("Enabling old style K6 write allocation for %d Mb\n",
 				mbytes);
 		}
 		return;
@@ -175,7 +174,7 @@ static void init_amd_k6(struct cpuinfo_x86 *c)
 			wbinvd();
 			wrmsr(MSR_K6_WHCR, l, h);
 			local_irq_restore(flags);
-			printk(KERN_INFO "Enabling new style K6 write allocation for %d Mb\n",
+			pr_info("Enabling new style K6 write allocation for %d Mb\n",
 				mbytes);
 		}
 
@@ -202,7 +201,7 @@ static void init_amd_k7(struct cpuinfo_x86 *c)
 	 */
 	if (c->x86_model >= 6 && c->x86_model <= 10) {
 		if (!cpu_has(c, X86_FEATURE_XMM)) {
-			printk(KERN_INFO "Enabling disabled K7/SSE Support.\n");
+			pr_info("Enabling disabled K7/SSE Support.\n");
 			msr_clear_bit(MSR_K7_HWCR, 15);
 			set_cpu_cap(c, X86_FEATURE_XMM);
 		}
@@ -216,9 +215,8 @@ static void init_amd_k7(struct cpuinfo_x86 *c)
 	if ((c->x86_model == 8 && c->x86_mask >= 1) || (c->x86_model > 8)) {
 		rdmsr(MSR_K7_CLK_CTL, l, h);
 		if ((l & 0xfff00000) != 0x20000000) {
-			printk(KERN_INFO
-			    "CPU: CLK_CTL MSR was %x. Reprogramming to %x\n",
-					l, ((l & 0x000fffff)|0x20000000));
+			pr_info("CPU: CLK_CTL MSR was %x. Reprogramming to %x\n",
+				l, ((l & 0x000fffff)|0x20000000));
 			wrmsr(MSR_K7_CLK_CTL, (l & 0x000fffff)|0x20000000, h);
 		}
 	}
@@ -485,7 +483,7 @@ static void bsp_init_amd(struct cpuinfo_x86 *c)
 		if (!rdmsrl_safe(MSR_K8_TSEG_ADDR, &tseg)) {
 			unsigned long pfn = tseg >> PAGE_SHIFT;
 
-			printk(KERN_DEBUG "tseg: %010llx\n", tseg);
+			pr_debug("tseg: %010llx\n", tseg);
 			if (pfn_range_is_mapped(pfn, pfn + 1))
 				set_memory_4k((unsigned long)__va(tseg), 1);
 		}
@@ -500,8 +498,7 @@ static void bsp_init_amd(struct cpuinfo_x86 *c)
 
 			rdmsrl(MSR_K7_HWCR, val);
 			if (!(val & BIT(24)))
-				printk(KERN_WARNING FW_BUG "TSC doesn't count "
-					"with P0 frequency!\n");
+				pr_warn(FW_BUG "TSC doesn't count with P0 frequency!\n");
 		}
 	}
 
diff --git a/arch/x86/kernel/cpu/bugs_64.c b/arch/x86/kernel/cpu/bugs_64.c
index 04f0fe5..a972ac4 100644
--- a/arch/x86/kernel/cpu/bugs_64.c
+++ b/arch/x86/kernel/cpu/bugs_64.c
@@ -15,7 +15,7 @@ void __init check_bugs(void)
 {
 	identify_boot_cpu();
 #if !defined(CONFIG_SMP)
-	printk(KERN_INFO "CPU: ");
+	pr_info("CPU: ");
 	print_cpu_info(&boot_cpu_data);
 #endif
 	alternative_instructions();
diff --git a/arch/x86/kernel/cpu/centaur.c b/arch/x86/kernel/cpu/centaur.c
index ae20be6..ce197bb 100644
--- a/arch/x86/kernel/cpu/centaur.c
+++ b/arch/x86/kernel/cpu/centaur.c
@@ -29,7 +29,7 @@ static void init_c3(struct cpuinfo_x86 *c)
 			rdmsr(MSR_VIA_FCR, lo, hi);
 			lo |= ACE_FCR;		/* enable ACE unit */
 			wrmsr(MSR_VIA_FCR, lo, hi);
-			printk(KERN_INFO "CPU: Enabled ACE h/w crypto\n");
+			pr_info("CPU: Enabled ACE h/w crypto\n");
 		}
 
 		/* enable RNG unit, if present and disabled */
@@ -37,7 +37,7 @@ static void init_c3(struct cpuinfo_x86 *c)
 			rdmsr(MSR_VIA_RNG, lo, hi);
 			lo |= RNG_ENABLE;	/* enable RNG unit */
 			wrmsr(MSR_VIA_RNG, lo, hi);
-			printk(KERN_INFO "CPU: Enabled h/w RNG\n");
+			pr_info("CPU: Enabled h/w RNG\n");
 		}
 
 		/* store Centaur Extended Feature Flags as
@@ -130,7 +130,7 @@ static void init_centaur(struct cpuinfo_x86 *c)
 			name = "C6";
 			fcr_set = ECX8|DSMC|EDCTLB|EMMX|ERETSTK;
 			fcr_clr = DPDC;
-			printk(KERN_NOTICE "Disabling bugged TSC.\n");
+			pr_notice("Disabling bugged TSC.\n");
 			clear_cpu_cap(c, X86_FEATURE_TSC);
 			break;
 		case 8:
@@ -163,11 +163,11 @@ static void init_centaur(struct cpuinfo_x86 *c)
 		newlo = (lo|fcr_set) & (~fcr_clr);
 
 		if (newlo != lo) {
-			printk(KERN_INFO "Centaur FCR was 0x%X now 0x%X\n",
+			pr_info("Centaur FCR was 0x%X now 0x%X\n",
 				lo, newlo);
 			wrmsr(MSR_IDT_FCR1, newlo, hi);
 		} else {
-			printk(KERN_INFO "Centaur FCR is 0x%X\n", lo);
+			pr_info("Centaur FCR is 0x%X\n", lo);
 		}
 		/* Emulate MTRRs using Centaur's MCR. */
 		set_cpu_cap(c, X86_FEATURE_CENTAUR_MCR);
diff --git a/arch/x86/kernel/cpu/common.c b/arch/x86/kernel/cpu/common.c
index 37830de..81cf716 100644
--- a/arch/x86/kernel/cpu/common.c
+++ b/arch/x86/kernel/cpu/common.c
@@ -228,7 +228,7 @@ static void squash_the_stupid_serial_number(struct cpuinfo_x86 *c)
 	lo |= 0x200000;
 	wrmsr(MSR_IA32_BBL_CR_CTL, lo, hi);
 
-	printk(KERN_NOTICE "CPU serial number disabled.\n");
+	pr_notice("CPU serial number disabled.\n");
 	clear_cpu_cap(c, X86_FEATURE_PN);
 
 	/* Disabling the serial number may affect the cpuid level */
@@ -329,9 +329,8 @@ static void filter_cpuid_features(struct cpuinfo_x86 *c, bool warn)
 		if (!warn)
 			continue;
 
-		printk(KERN_WARNING
-		       "CPU: CPU feature " X86_CAP_FMT " disabled, no CPUID level 0x%x\n",
-				x86_cap_flag(df->feature), df->level);
+		pr_warn("CPU: CPU feature " X86_CAP_FMT " disabled, no CPUID level 0x%x\n",
+			x86_cap_flag(df->feature), df->level);
 	}
 }
 
@@ -510,7 +509,7 @@ void detect_ht(struct cpuinfo_x86 *c)
 	smp_num_siblings = (ebx & 0xff0000) >> 16;
 
 	if (smp_num_siblings == 1) {
-		printk_once(KERN_INFO "CPU0: Hyper-Threading is disabled\n");
+		pr_info_once("CPU0: Hyper-Threading is disabled\n");
 		goto out;
 	}
 
@@ -531,10 +530,10 @@ void detect_ht(struct cpuinfo_x86 *c)
 
 out:
 	if (!printed && (c->x86_max_cores * smp_num_siblings) > 1) {
-		printk(KERN_INFO  "CPU: Physical Processor ID: %d\n",
-		       c->phys_proc_id);
-		printk(KERN_INFO  "CPU: Processor Core ID: %d\n",
-		       c->cpu_core_id);
+		pr_info("CPU: Physical Processor ID: %d\n",
+			c->phys_proc_id);
+		pr_info("CPU: Processor Core ID: %d\n",
+			c->cpu_core_id);
 		printed = 1;
 	}
 #endif
@@ -559,9 +558,8 @@ static void get_cpu_vendor(struct cpuinfo_x86 *c)
 		}
 	}
 
-	printk_once(KERN_ERR
-			"CPU: vendor_id '%s' unknown, using generic init.\n" \
-			"CPU: Your system may be unstable.\n", v);
+	pr_err_once("CPU: vendor_id '%s' unknown, using generic init.\n" \
+		    "CPU: Your system may be unstable.\n", v);
 
 	c->x86_vendor = X86_VENDOR_UNKNOWN;
 	this_cpu = &default_cpu;
@@ -760,7 +758,7 @@ void __init early_cpu_init(void)
 	int count = 0;
 
 #ifdef CONFIG_PROCESSOR_SELECT
-	printk(KERN_INFO "KERNEL supported cpus:\n");
+	pr_info("KERNEL supported cpus:\n");
 #endif
 
 	for (cdev = __x86_cpu_dev_start; cdev < __x86_cpu_dev_end; cdev++) {
@@ -778,7 +776,7 @@ void __init early_cpu_init(void)
 			for (j = 0; j < 2; j++) {
 				if (!cpudev->c_ident[j])
 					continue;
-				printk(KERN_INFO "  %s %s\n", cpudev->c_vendor,
+				pr_info("  %s %s\n", cpudev->c_vendor,
 					cpudev->c_ident[j]);
 			}
 		}
@@ -977,6 +975,8 @@ static void identify_cpu(struct cpuinfo_x86 *c)
 #ifdef CONFIG_NUMA
 	numa_add_cpu(smp_processor_id());
 #endif
+	/* The boot/hotplug time assigment got cleared, restore it */
+	c->logical_proc_id = topology_phys_to_logical_pkg(c->phys_proc_id);
 }
 
 /*
@@ -1061,7 +1061,7 @@ static void __print_cpu_msr(void)
 		for (index = index_min; index < index_max; index++) {
 			if (rdmsrl_safe(index, &val))
 				continue;
-			printk(KERN_INFO " MSR%08x: %016llx\n", index, val);
+			pr_info(" MSR%08x: %016llx\n", index, val);
 		}
 	}
 }
@@ -1100,19 +1100,19 @@ void print_cpu_info(struct cpuinfo_x86 *c)
 	}
 
 	if (vendor && !strstr(c->x86_model_id, vendor))
-		printk(KERN_CONT "%s ", vendor);
+		pr_cont("%s ", vendor);
 
 	if (c->x86_model_id[0])
-		printk(KERN_CONT "%s", c->x86_model_id);
+		pr_cont("%s", c->x86_model_id);
 	else
-		printk(KERN_CONT "%d86", c->x86);
+		pr_cont("%d86", c->x86);
 
-	printk(KERN_CONT " (family: 0x%x, model: 0x%x", c->x86, c->x86_model);
+	pr_cont(" (family: 0x%x, model: 0x%x", c->x86, c->x86_model);
 
 	if (c->x86_mask || c->cpuid_level >= 0)
-		printk(KERN_CONT ", stepping: 0x%x)\n", c->x86_mask);
+		pr_cont(", stepping: 0x%x)\n", c->x86_mask);
 	else
-		printk(KERN_CONT ")\n");
+		pr_cont(")\n");
 
 	print_cpu_msr(c);
 }
@@ -1438,7 +1438,7 @@ void cpu_init(void)
 
 	show_ucode_info_early();
 
-	printk(KERN_INFO "Initializing CPU#%d\n", cpu);
+	pr_info("Initializing CPU#%d\n", cpu);
 
 	if (cpu_feature_enabled(X86_FEATURE_VME) ||
 	    cpu_has_tsc ||
diff --git a/arch/x86/kernel/cpu/cyrix.c b/arch/x86/kernel/cpu/cyrix.c
index aaf152e..187bb58 100644
--- a/arch/x86/kernel/cpu/cyrix.c
+++ b/arch/x86/kernel/cpu/cyrix.c
@@ -103,7 +103,7 @@ static void check_cx686_slop(struct cpuinfo_x86 *c)
 		local_irq_restore(flags);
 
 		if (ccr5 & 2) { /* possible wrong calibration done */
-			printk(KERN_INFO "Recalibrating delay loop with SLOP bit reset\n");
+			pr_info("Recalibrating delay loop with SLOP bit reset\n");
 			calibrate_delay();
 			c->loops_per_jiffy = loops_per_jiffy;
 		}
@@ -115,7 +115,7 @@ static void set_cx86_reorder(void)
 {
 	u8 ccr3;
 
-	printk(KERN_INFO "Enable Memory access reorder on Cyrix/NSC processor.\n");
+	pr_info("Enable Memory access reorder on Cyrix/NSC processor.\n");
 	ccr3 = getCx86(CX86_CCR3);
 	setCx86(CX86_CCR3, (ccr3 & 0x0f) | 0x10); /* enable MAPEN */
 
@@ -128,7 +128,7 @@ static void set_cx86_reorder(void)
 
 static void set_cx86_memwb(void)
 {
-	printk(KERN_INFO "Enable Memory-Write-back mode on Cyrix/NSC processor.\n");
+	pr_info("Enable Memory-Write-back mode on Cyrix/NSC processor.\n");
 
 	/* CCR2 bit 2: unlock NW bit */
 	setCx86_old(CX86_CCR2, getCx86_old(CX86_CCR2) & ~0x04);
@@ -268,7 +268,7 @@ static void init_cyrix(struct cpuinfo_x86 *c)
 		 *  VSA1 we work around however.
 		 */
 
-		printk(KERN_INFO "Working around Cyrix MediaGX virtual DMA bugs.\n");
+		pr_info("Working around Cyrix MediaGX virtual DMA bugs.\n");
 		isa_dma_bridge_buggy = 2;
 
 		/* We do this before the PCI layer is running. However we
@@ -426,7 +426,7 @@ static void cyrix_identify(struct cpuinfo_x86 *c)
 		if (dir0 == 5 || dir0 == 3) {
 			unsigned char ccr3;
 			unsigned long flags;
-			printk(KERN_INFO "Enabling CPUID on Cyrix processor.\n");
+			pr_info("Enabling CPUID on Cyrix processor.\n");
 			local_irq_save(flags);
 			ccr3 = getCx86(CX86_CCR3);
 			/* enable MAPEN  */
diff --git a/arch/x86/kernel/cpu/hypervisor.c b/arch/x86/kernel/cpu/hypervisor.c
index d820d8e..73d391a 100644
--- a/arch/x86/kernel/cpu/hypervisor.c
+++ b/arch/x86/kernel/cpu/hypervisor.c
@@ -56,7 +56,7 @@ detect_hypervisor_vendor(void)
 	}
 
 	if (max_pri)
-		printk(KERN_INFO "Hypervisor detected: %s\n", x86_hyper->name);
+		pr_info("Hypervisor detected: %s\n", x86_hyper->name);
 }
 
 void init_hypervisor(struct cpuinfo_x86 *c)
diff --git a/arch/x86/kernel/cpu/intel.c b/arch/x86/kernel/cpu/intel.c
index 565648b..38766c2 100644
--- a/arch/x86/kernel/cpu/intel.c
+++ b/arch/x86/kernel/cpu/intel.c
@@ -61,7 +61,7 @@ static void early_init_intel(struct cpuinfo_x86 *c)
 	 */
 	if (c->x86 == 6 && c->x86_model == 0x1c && c->x86_mask <= 2 &&
 	    c->microcode < 0x20e) {
-		printk(KERN_WARNING "Atom PSE erratum detected, BIOS microcode update recommended\n");
+		pr_warn("Atom PSE erratum detected, BIOS microcode update recommended\n");
 		clear_cpu_cap(c, X86_FEATURE_PSE);
 	}
 
@@ -140,7 +140,7 @@ static void early_init_intel(struct cpuinfo_x86 *c)
 	if (c->x86 > 6 || (c->x86 == 6 && c->x86_model >= 0xd)) {
 		rdmsrl(MSR_IA32_MISC_ENABLE, misc_enable);
 		if (!(misc_enable & MSR_IA32_MISC_ENABLE_FAST_STRING)) {
-			printk(KERN_INFO "Disabled fast string operations\n");
+			pr_info("Disabled fast string operations\n");
 			setup_clear_cpu_cap(X86_FEATURE_REP_GOOD);
 			setup_clear_cpu_cap(X86_FEATURE_ERMS);
 		}
@@ -160,6 +160,19 @@ static void early_init_intel(struct cpuinfo_x86 *c)
 		pr_info("Disabling PGE capability bit\n");
 		setup_clear_cpu_cap(X86_FEATURE_PGE);
 	}
+
+	if (c->cpuid_level >= 0x00000001) {
+		u32 eax, ebx, ecx, edx;
+
+		cpuid(0x00000001, &eax, &ebx, &ecx, &edx);
+		/*
+		 * If HTT (EDX[28]) is set EBX[16:23] contain the number of
+		 * apicids which are reserved per package. Store the resulting
+		 * shift value for the package management code.
+		 */
+		if (edx & (1U << 28))
+			c->x86_coreid_bits = get_count_order((ebx >> 16) & 0xff);
+	}
 }
 
 #ifdef CONFIG_X86_32
@@ -176,7 +189,7 @@ int ppro_with_ram_bug(void)
 	    boot_cpu_data.x86 == 6 &&
 	    boot_cpu_data.x86_model == 1 &&
 	    boot_cpu_data.x86_mask < 8) {
-		printk(KERN_INFO "Pentium Pro with Errata#50 detected. Taking evasive action.\n");
+		pr_info("Pentium Pro with Errata#50 detected. Taking evasive action.\n");
 		return 1;
 	}
 	return 0;
@@ -225,7 +238,7 @@ static void intel_workarounds(struct cpuinfo_x86 *c)
 
 		set_cpu_bug(c, X86_BUG_F00F);
 		if (!f00f_workaround_enabled) {
-			printk(KERN_NOTICE "Intel Pentium with F0 0F bug - workaround enabled.\n");
+			pr_notice("Intel Pentium with F0 0F bug - workaround enabled.\n");
 			f00f_workaround_enabled = 1;
 		}
 	}
@@ -244,7 +257,7 @@ static void intel_workarounds(struct cpuinfo_x86 *c)
 	 * Forcefully enable PAE if kernel parameter "forcepae" is present.
 	 */
 	if (forcepae) {
-		printk(KERN_WARNING "PAE forced!\n");
+		pr_warn("PAE forced!\n");
 		set_cpu_cap(c, X86_FEATURE_PAE);
 		add_taint(TAINT_CPU_OUT_OF_SPEC, LOCKDEP_NOW_UNRELIABLE);
 	}
diff --git a/arch/x86/kernel/cpu/intel_cacheinfo.c b/arch/x86/kernel/cpu/intel_cacheinfo.c
index 0b6c523..6ed779e 100644
--- a/arch/x86/kernel/cpu/intel_cacheinfo.c
+++ b/arch/x86/kernel/cpu/intel_cacheinfo.c
@@ -444,7 +444,7 @@ static ssize_t store_cache_disable(struct cacheinfo *this_leaf,
 	err = amd_set_l3_disable_slot(nb, cpu, slot, val);
 	if (err) {
 		if (err == -EEXIST)
-			pr_warning("L3 slot %d in use/index already disabled!\n",
+			pr_warn("L3 slot %d in use/index already disabled!\n",
 				   slot);
 		return err;
 	}
diff --git a/arch/x86/kernel/cpu/mcheck/mce-inject.c b/arch/x86/kernel/cpu/mcheck/mce-inject.c
index 4cfba43..517619e 100644
--- a/arch/x86/kernel/cpu/mcheck/mce-inject.c
+++ b/arch/x86/kernel/cpu/mcheck/mce-inject.c
@@ -115,7 +115,7 @@ static int raise_local(void)
 	int cpu = m->extcpu;
 
 	if (m->inject_flags & MCJ_EXCEPTION) {
-		printk(KERN_INFO "Triggering MCE exception on CPU %d\n", cpu);
+		pr_info("Triggering MCE exception on CPU %d\n", cpu);
 		switch (context) {
 		case MCJ_CTX_IRQ:
 			/*
@@ -128,15 +128,15 @@ static int raise_local(void)
 			raise_exception(m, NULL);
 			break;
 		default:
-			printk(KERN_INFO "Invalid MCE context\n");
+			pr_info("Invalid MCE context\n");
 			ret = -EINVAL;
 		}
-		printk(KERN_INFO "MCE exception done on CPU %d\n", cpu);
+		pr_info("MCE exception done on CPU %d\n", cpu);
 	} else if (m->status) {
-		printk(KERN_INFO "Starting machine check poll CPU %d\n", cpu);
+		pr_info("Starting machine check poll CPU %d\n", cpu);
 		raise_poll(m);
 		mce_notify_irq();
-		printk(KERN_INFO "Machine check poll done on CPU %d\n", cpu);
+		pr_info("Machine check poll done on CPU %d\n", cpu);
 	} else
 		m->finished = 0;
 
@@ -183,8 +183,7 @@ static void raise_mce(struct mce *m)
 		start = jiffies;
 		while (!cpumask_empty(mce_inject_cpumask)) {
 			if (!time_before(jiffies, start + 2*HZ)) {
-				printk(KERN_ERR
-				"Timeout waiting for mce inject %lx\n",
+				pr_err("Timeout waiting for mce inject %lx\n",
 					*cpumask_bits(mce_inject_cpumask));
 				break;
 			}
@@ -241,7 +240,7 @@ static int inject_init(void)
 {
 	if (!alloc_cpumask_var(&mce_inject_cpumask, GFP_KERNEL))
 		return -ENOMEM;
-	printk(KERN_INFO "Machine check injector initialized\n");
+	pr_info("Machine check injector initialized\n");
 	register_mce_write_callback(mce_write);
 	register_nmi_handler(NMI_LOCAL, mce_raise_notify, 0,
 				"mce_notify");
diff --git a/arch/x86/kernel/cpu/mcheck/p5.c b/arch/x86/kernel/cpu/mcheck/p5.c
index 12402e1..2a0717b 100644
--- a/arch/x86/kernel/cpu/mcheck/p5.c
+++ b/arch/x86/kernel/cpu/mcheck/p5.c
@@ -26,14 +26,12 @@ static void pentium_machine_check(struct pt_regs *regs, long error_code)
 	rdmsr(MSR_IA32_P5_MC_ADDR, loaddr, hi);
 	rdmsr(MSR_IA32_P5_MC_TYPE, lotype, hi);
 
-	printk(KERN_EMERG
-		"CPU#%d: Machine Check Exception:  0x%8X (type 0x%8X).\n",
-		smp_processor_id(), loaddr, lotype);
+	pr_emerg("CPU#%d: Machine Check Exception:  0x%8X (type 0x%8X).\n",
+		 smp_processor_id(), loaddr, lotype);
 
 	if (lotype & (1<<5)) {
-		printk(KERN_EMERG
-			"CPU#%d: Possible thermal failure (CPU on fire ?).\n",
-			smp_processor_id());
+		pr_emerg("CPU#%d: Possible thermal failure (CPU on fire ?).\n",
+			 smp_processor_id());
 	}
 
 	add_taint(TAINT_MACHINE_CHECK, LOCKDEP_NOW_UNRELIABLE);
@@ -61,12 +59,10 @@ void intel_p5_mcheck_init(struct cpuinfo_x86 *c)
 	/* Read registers before enabling: */
 	rdmsr(MSR_IA32_P5_MC_ADDR, l, h);
 	rdmsr(MSR_IA32_P5_MC_TYPE, l, h);
-	printk(KERN_INFO
-	       "Intel old style machine check architecture supported.\n");
+	pr_info("Intel old style machine check architecture supported.\n");
 
 	/* Enable MCE: */
 	cr4_set_bits(X86_CR4_MCE);
-	printk(KERN_INFO
-	       "Intel old style machine check reporting enabled on CPU#%d.\n",
-	       smp_processor_id());
+	pr_info("Intel old style machine check reporting enabled on CPU#%d.\n",
+		smp_processor_id());
 }
diff --git a/arch/x86/kernel/cpu/mcheck/therm_throt.c b/arch/x86/kernel/cpu/mcheck/therm_throt.c
index 2c5aaf8..0b445c2 100644
--- a/arch/x86/kernel/cpu/mcheck/therm_throt.c
+++ b/arch/x86/kernel/cpu/mcheck/therm_throt.c
@@ -190,7 +190,7 @@ static int therm_throt_process(bool new_event, int event, int level)
 	/* if we just entered the thermal event */
 	if (new_event) {
 		if (event == THERMAL_THROTTLING_EVENT)
-			printk(KERN_CRIT "CPU%d: %s temperature above threshold, cpu clock throttled (total events = %lu)\n",
+			pr_crit("CPU%d: %s temperature above threshold, cpu clock throttled (total events = %lu)\n",
 				this_cpu,
 				level == CORE_LEVEL ? "Core" : "Package",
 				state->count);
@@ -198,8 +198,7 @@ static int therm_throt_process(bool new_event, int event, int level)
 	}
 	if (old_event) {
 		if (event == THERMAL_THROTTLING_EVENT)
-			printk(KERN_INFO "CPU%d: %s temperature/speed normal\n",
-				this_cpu,
+			pr_info("CPU%d: %s temperature/speed normal\n", this_cpu,
 				level == CORE_LEVEL ? "Core" : "Package");
 		return 1;
 	}
@@ -417,8 +416,8 @@ static void intel_thermal_interrupt(void)
 
 static void unexpected_thermal_interrupt(void)
 {
-	printk(KERN_ERR "CPU%d: Unexpected LVT thermal interrupt!\n",
-			smp_processor_id());
+	pr_err("CPU%d: Unexpected LVT thermal interrupt!\n",
+		smp_processor_id());
 }
 
 static void (*smp_thermal_vector)(void) = unexpected_thermal_interrupt;
@@ -499,7 +498,7 @@ void intel_init_thermal(struct cpuinfo_x86 *c)
 
 	if ((l & MSR_IA32_MISC_ENABLE_TM1) && (h & APIC_DM_SMI)) {
 		if (system_state == SYSTEM_BOOTING)
-			printk(KERN_DEBUG "CPU%d: Thermal monitoring handled by SMI\n", cpu);
+			pr_debug("CPU%d: Thermal monitoring handled by SMI\n", cpu);
 		return;
 	}
 
@@ -557,8 +556,8 @@ void intel_init_thermal(struct cpuinfo_x86 *c)
 	l = apic_read(APIC_LVTTHMR);
 	apic_write(APIC_LVTTHMR, l & ~APIC_LVT_MASKED);
 
-	printk_once(KERN_INFO "CPU0: Thermal monitoring enabled (%s)\n",
-		       tm2 ? "TM2" : "TM1");
+	pr_info_once("CPU0: Thermal monitoring enabled (%s)\n",
+		      tm2 ? "TM2" : "TM1");
 
 	/* enable thermal throttle processing */
 	atomic_set(&therm_throt_en, 1);
diff --git a/arch/x86/kernel/cpu/mcheck/threshold.c b/arch/x86/kernel/cpu/mcheck/threshold.c
index 7245980..fcf9ae9 100644
--- a/arch/x86/kernel/cpu/mcheck/threshold.c
+++ b/arch/x86/kernel/cpu/mcheck/threshold.c
@@ -12,8 +12,8 @@
 
 static void default_threshold_interrupt(void)
 {
-	printk(KERN_ERR "Unexpected threshold interrupt at vector %x\n",
-			 THRESHOLD_APIC_VECTOR);
+	pr_err("Unexpected threshold interrupt at vector %x\n",
+		THRESHOLD_APIC_VECTOR);
 }
 
 void (*mce_threshold_vector)(void) = default_threshold_interrupt;
diff --git a/arch/x86/kernel/cpu/mcheck/winchip.c b/arch/x86/kernel/cpu/mcheck/winchip.c
index 01dd870..c6a722e 100644
--- a/arch/x86/kernel/cpu/mcheck/winchip.c
+++ b/arch/x86/kernel/cpu/mcheck/winchip.c
@@ -17,7 +17,7 @@ static void winchip_machine_check(struct pt_regs *regs, long error_code)
 {
 	ist_enter(regs);
 
-	printk(KERN_EMERG "CPU0: Machine Check Exception.\n");
+	pr_emerg("CPU0: Machine Check Exception.\n");
 	add_taint(TAINT_MACHINE_CHECK, LOCKDEP_NOW_UNRELIABLE);
 
 	ist_exit(regs);
@@ -39,6 +39,5 @@ void winchip_mcheck_init(struct cpuinfo_x86 *c)
 
 	cr4_set_bits(X86_CR4_MCE);
 
-	printk(KERN_INFO
-	       "Winchip machine check reporting enabled on CPU#0.\n");
+	pr_info("Winchip machine check reporting enabled on CPU#0.\n");
 }
diff --git a/arch/x86/kernel/cpu/microcode/amd.c b/arch/x86/kernel/cpu/microcode/amd.c
index 2233f8a..75d3aab 100644
--- a/arch/x86/kernel/cpu/microcode/amd.c
+++ b/arch/x86/kernel/cpu/microcode/amd.c
@@ -953,7 +953,7 @@ struct microcode_ops * __init init_amd_microcode(void)
 	struct cpuinfo_x86 *c = &boot_cpu_data;
 
 	if (c->x86_vendor != X86_VENDOR_AMD || c->x86 < 0x10) {
-		pr_warning("AMD CPU family 0x%x not supported\n", c->x86);
+		pr_warn("AMD CPU family 0x%x not supported\n", c->x86);
 		return NULL;
 	}
 
diff --git a/arch/x86/kernel/cpu/mshyperv.c b/arch/x86/kernel/cpu/mshyperv.c
index 20e242e..4e7c693 100644
--- a/arch/x86/kernel/cpu/mshyperv.c
+++ b/arch/x86/kernel/cpu/mshyperv.c
@@ -161,8 +161,8 @@ static void __init ms_hyperv_init_platform(void)
 	ms_hyperv.misc_features = cpuid_edx(HYPERV_CPUID_FEATURES);
 	ms_hyperv.hints    = cpuid_eax(HYPERV_CPUID_ENLIGHTMENT_INFO);
 
-	printk(KERN_INFO "HyperV: features 0x%x, hints 0x%x\n",
-	       ms_hyperv.features, ms_hyperv.hints);
+	pr_info("HyperV: features 0x%x, hints 0x%x\n",
+		ms_hyperv.features, ms_hyperv.hints);
 
 #ifdef CONFIG_X86_LOCAL_APIC
 	if (ms_hyperv.features & HV_X64_MSR_APIC_FREQUENCY_AVAILABLE) {
@@ -174,8 +174,8 @@ static void __init ms_hyperv_init_platform(void)
 		rdmsrl(HV_X64_MSR_APIC_FREQUENCY, hv_lapic_frequency);
 		hv_lapic_frequency = div_u64(hv_lapic_frequency, HZ);
 		lapic_timer_frequency = hv_lapic_frequency;
-		printk(KERN_INFO "HyperV: LAPIC Timer Frequency: %#x\n",
-				lapic_timer_frequency);
+		pr_info("HyperV: LAPIC Timer Frequency: %#x\n",
+			lapic_timer_frequency);
 	}
 #endif
 
diff --git a/arch/x86/kernel/cpu/mtrr/centaur.c b/arch/x86/kernel/cpu/mtrr/centaur.c
index 316fe3e..3d68993 100644
--- a/arch/x86/kernel/cpu/mtrr/centaur.c
+++ b/arch/x86/kernel/cpu/mtrr/centaur.c
@@ -103,7 +103,7 @@ centaur_validate_add_page(unsigned long base, unsigned long size, unsigned int t
 	 */
 	if (type != MTRR_TYPE_WRCOMB &&
 	    (centaur_mcr_type == 0 || type != MTRR_TYPE_UNCACHABLE)) {
-		pr_warning("mtrr: only write-combining%s supported\n",
+		pr_warn("mtrr: only write-combining%s supported\n",
 			   centaur_mcr_type ? " and uncacheable are" : " is");
 		return -EINVAL;
 	}
diff --git a/arch/x86/kernel/cpu/mtrr/cleanup.c b/arch/x86/kernel/cpu/mtrr/cleanup.c
index 0d98503..31e951c 100644
--- a/arch/x86/kernel/cpu/mtrr/cleanup.c
+++ b/arch/x86/kernel/cpu/mtrr/cleanup.c
@@ -57,9 +57,9 @@ static int __initdata				nr_range;
 static struct var_mtrr_range_state __initdata	range_state[RANGE_NUM];
 
 static int __initdata debug_print;
-#define Dprintk(x...) do { if (debug_print) printk(KERN_DEBUG x); } while (0)
+#define Dprintk(x...) do { if (debug_print) pr_debug(x); } while (0)
 
-#define BIOS_BUG_MSG KERN_WARNING \
+#define BIOS_BUG_MSG \
 	"WARNING: BIOS bug: VAR MTRR %d contains strange UC entry under 1M, check with your system vendor!\n"
 
 static int __init
@@ -81,9 +81,9 @@ x86_get_mtrr_mem_range(struct range *range, int nr_range,
 						base, base + size);
 	}
 	if (debug_print) {
-		printk(KERN_DEBUG "After WB checking\n");
+		pr_debug("After WB checking\n");
 		for (i = 0; i < nr_range; i++)
-			printk(KERN_DEBUG "MTRR MAP PFN: %016llx - %016llx\n",
+			pr_debug("MTRR MAP PFN: %016llx - %016llx\n",
 				 range[i].start, range[i].end);
 	}
 
@@ -101,7 +101,7 @@ x86_get_mtrr_mem_range(struct range *range, int nr_range,
 		    (mtrr_state.enabled & MTRR_STATE_MTRR_ENABLED) &&
 		    (mtrr_state.enabled & MTRR_STATE_MTRR_FIXED_ENABLED)) {
 			/* Var MTRR contains UC entry below 1M? Skip it: */
-			printk(BIOS_BUG_MSG, i);
+			pr_warn(BIOS_BUG_MSG, i);
 			if (base + size <= (1<<(20-PAGE_SHIFT)))
 				continue;
 			size -= (1<<(20-PAGE_SHIFT)) - base;
@@ -114,11 +114,11 @@ x86_get_mtrr_mem_range(struct range *range, int nr_range,
 				 extra_remove_base + extra_remove_size);
 
 	if  (debug_print) {
-		printk(KERN_DEBUG "After UC checking\n");
+		pr_debug("After UC checking\n");
 		for (i = 0; i < RANGE_NUM; i++) {
 			if (!range[i].end)
 				continue;
-			printk(KERN_DEBUG "MTRR MAP PFN: %016llx - %016llx\n",
+			pr_debug("MTRR MAP PFN: %016llx - %016llx\n",
 				 range[i].start, range[i].end);
 		}
 	}
@@ -126,9 +126,9 @@ x86_get_mtrr_mem_range(struct range *range, int nr_range,
 	/* sort the ranges */
 	nr_range = clean_sort_range(range, RANGE_NUM);
 	if  (debug_print) {
-		printk(KERN_DEBUG "After sorting\n");
+		pr_debug("After sorting\n");
 		for (i = 0; i < nr_range; i++)
-			printk(KERN_DEBUG "MTRR MAP PFN: %016llx - %016llx\n",
+			pr_debug("MTRR MAP PFN: %016llx - %016llx\n",
 				 range[i].start, range[i].end);
 	}
 
@@ -544,7 +544,7 @@ static void __init print_out_mtrr_range_state(void)
 		start_base = to_size_factor(start_base, &start_factor),
 		type = range_state[i].type;
 
-		printk(KERN_DEBUG "reg %d, base: %ld%cB, range: %ld%cB, type %s\n",
+		pr_debug("reg %d, base: %ld%cB, range: %ld%cB, type %s\n",
 			i, start_base, start_factor,
 			size_base, size_factor,
 			(type == MTRR_TYPE_UNCACHABLE) ? "UC" :
@@ -713,7 +713,7 @@ int __init mtrr_cleanup(unsigned address_bits)
 		return 0;
 
 	/* Print original var MTRRs at first, for debugging: */
-	printk(KERN_DEBUG "original variable MTRRs\n");
+	pr_debug("original variable MTRRs\n");
 	print_out_mtrr_range_state();
 
 	memset(range, 0, sizeof(range));
@@ -733,7 +733,7 @@ int __init mtrr_cleanup(unsigned address_bits)
 					  x_remove_base, x_remove_size);
 
 	range_sums = sum_ranges(range, nr_range);
-	printk(KERN_INFO "total RAM covered: %ldM\n",
+	pr_info("total RAM covered: %ldM\n",
 	       range_sums >> (20 - PAGE_SHIFT));
 
 	if (mtrr_chunk_size && mtrr_gran_size) {
@@ -745,12 +745,11 @@ int __init mtrr_cleanup(unsigned address_bits)
 
 		if (!result[i].bad) {
 			set_var_mtrr_all(address_bits);
-			printk(KERN_DEBUG "New variable MTRRs\n");
+			pr_debug("New variable MTRRs\n");
 			print_out_mtrr_range_state();
 			return 1;
 		}
-		printk(KERN_INFO "invalid mtrr_gran_size or mtrr_chunk_size, "
-		       "will find optimal one\n");
+		pr_info("invalid mtrr_gran_size or mtrr_chunk_size, will find optimal one\n");
 	}
 
 	i = 0;
@@ -768,7 +767,7 @@ int __init mtrr_cleanup(unsigned address_bits)
 				      x_remove_base, x_remove_size, i);
 			if (debug_print) {
 				mtrr_print_out_one_result(i);
-				printk(KERN_INFO "\n");
+				pr_info("\n");
 			}
 
 			i++;
@@ -779,7 +778,7 @@ int __init mtrr_cleanup(unsigned address_bits)
 	index_good = mtrr_search_optimal_index();
 
 	if (index_good != -1) {
-		printk(KERN_INFO "Found optimal setting for mtrr clean up\n");
+		pr_info("Found optimal setting for mtrr clean up\n");
 		i = index_good;
 		mtrr_print_out_one_result(i);
 
@@ -790,7 +789,7 @@ int __init mtrr_cleanup(unsigned address_bits)
 		gran_size <<= 10;
 		x86_setup_var_mtrrs(range, nr_range, chunk_size, gran_size);
 		set_var_mtrr_all(address_bits);
-		printk(KERN_DEBUG "New variable MTRRs\n");
+		pr_debug("New variable MTRRs\n");
 		print_out_mtrr_range_state();
 		return 1;
 	} else {
@@ -799,8 +798,8 @@ int __init mtrr_cleanup(unsigned address_bits)
 			mtrr_print_out_one_result(i);
 	}
 
-	printk(KERN_INFO "mtrr_cleanup: can not find optimal value\n");
-	printk(KERN_INFO "please specify mtrr_gran_size/mtrr_chunk_size\n");
+	pr_info("mtrr_cleanup: can not find optimal value\n");
+	pr_info("please specify mtrr_gran_size/mtrr_chunk_size\n");
 
 	return 0;
 }
@@ -918,7 +917,7 @@ int __init mtrr_trim_uncached_memory(unsigned long end_pfn)
 
 	/* kvm/qemu doesn't have mtrr set right, don't trim them all: */
 	if (!highest_pfn) {
-		printk(KERN_INFO "CPU MTRRs all blank - virtualized system.\n");
+		pr_info("CPU MTRRs all blank - virtualized system.\n");
 		return 0;
 	}
 
@@ -973,7 +972,8 @@ int __init mtrr_trim_uncached_memory(unsigned long end_pfn)
 							 end_pfn);
 
 	if (total_trim_size) {
-		pr_warning("WARNING: BIOS bug: CPU MTRRs don't cover all of memory, losing %lluMB of RAM.\n", total_trim_size >> 20);
+		pr_warn("WARNING: BIOS bug: CPU MTRRs don't cover all of memory, losing %lluMB of RAM.\n",
+			total_trim_size >> 20);
 
 		if (!changed_by_mtrr_cleanup)
 			WARN_ON(1);
diff --git a/arch/x86/kernel/cpu/mtrr/generic.c b/arch/x86/kernel/cpu/mtrr/generic.c
index c870af1..fcbcb2f 100644
--- a/arch/x86/kernel/cpu/mtrr/generic.c
+++ b/arch/x86/kernel/cpu/mtrr/generic.c
@@ -55,7 +55,7 @@ static inline void k8_check_syscfg_dram_mod_en(void)
 
 	rdmsr(MSR_K8_SYSCFG, lo, hi);
 	if (lo & K8_MTRRFIXRANGE_DRAM_MODIFY) {
-		printk(KERN_ERR FW_WARN "MTRR: CPU %u: SYSCFG[MtrrFixDramModEn]"
+		pr_err(FW_WARN "MTRR: CPU %u: SYSCFG[MtrrFixDramModEn]"
 		       " not cleared by BIOS, clearing this bit\n",
 		       smp_processor_id());
 		lo &= ~K8_MTRRFIXRANGE_DRAM_MODIFY;
@@ -501,14 +501,14 @@ void __init mtrr_state_warn(void)
 	if (!mask)
 		return;
 	if (mask & MTRR_CHANGE_MASK_FIXED)
-		pr_warning("mtrr: your CPUs had inconsistent fixed MTRR settings\n");
+		pr_warn("mtrr: your CPUs had inconsistent fixed MTRR settings\n");
 	if (mask & MTRR_CHANGE_MASK_VARIABLE)
-		pr_warning("mtrr: your CPUs had inconsistent variable MTRR settings\n");
+		pr_warn("mtrr: your CPUs had inconsistent variable MTRR settings\n");
 	if (mask & MTRR_CHANGE_MASK_DEFTYPE)
-		pr_warning("mtrr: your CPUs had inconsistent MTRRdefType settings\n");
+		pr_warn("mtrr: your CPUs had inconsistent MTRRdefType settings\n");
 
-	printk(KERN_INFO "mtrr: probably your BIOS does not setup all CPUs.\n");
-	printk(KERN_INFO "mtrr: corrected configuration.\n");
+	pr_info("mtrr: probably your BIOS does not setup all CPUs.\n");
+	pr_info("mtrr: corrected configuration.\n");
 }
 
 /*
@@ -519,8 +519,7 @@ void __init mtrr_state_warn(void)
 void mtrr_wrmsr(unsigned msr, unsigned a, unsigned b)
 {
 	if (wrmsr_safe(msr, a, b) < 0) {
-		printk(KERN_ERR
-			"MTRR: CPU %u: Writing MSR %x to %x:%x failed\n",
+		pr_err("MTRR: CPU %u: Writing MSR %x to %x:%x failed\n",
 			smp_processor_id(), msr, a, b);
 	}
 }
@@ -607,7 +606,7 @@ static void generic_get_mtrr(unsigned int reg, unsigned long *base,
 		tmp |= ~((1ULL<<(hi - 1)) - 1);
 
 		if (tmp != mask) {
-			printk(KERN_WARNING "mtrr: your BIOS has configured an incorrect mask, fixing it.\n");
+			pr_warn("mtrr: your BIOS has configured an incorrect mask, fixing it.\n");
 			add_taint(TAINT_FIRMWARE_WORKAROUND, LOCKDEP_STILL_OK);
 			mask = tmp;
 		}
@@ -858,13 +857,13 @@ int generic_validate_add_page(unsigned long base, unsigned long size,
 	    boot_cpu_data.x86_model == 1 &&
 	    boot_cpu_data.x86_mask <= 7) {
 		if (base & ((1 << (22 - PAGE_SHIFT)) - 1)) {
-			pr_warning("mtrr: base(0x%lx000) is not 4 MiB aligned\n", base);
+			pr_warn("mtrr: base(0x%lx000) is not 4 MiB aligned\n", base);
 			return -EINVAL;
 		}
 		if (!(base + size < 0x70000 || base > 0x7003F) &&
 		    (type == MTRR_TYPE_WRCOMB
 		     || type == MTRR_TYPE_WRBACK)) {
-			pr_warning("mtrr: writable mtrr between 0x70000000 and 0x7003FFFF may hang the CPU.\n");
+			pr_warn("mtrr: writable mtrr between 0x70000000 and 0x7003FFFF may hang the CPU.\n");
 			return -EINVAL;
 		}
 	}
@@ -878,7 +877,7 @@ int generic_validate_add_page(unsigned long base, unsigned long size,
 	     lbase = lbase >> 1, last = last >> 1)
 		;
 	if (lbase != last) {
-		pr_warning("mtrr: base(0x%lx000) is not aligned on a size(0x%lx000) boundary\n", base, size);
+		pr_warn("mtrr: base(0x%lx000) is not aligned on a size(0x%lx000) boundary\n", base, size);
 		return -EINVAL;
 	}
 	return 0;
diff --git a/arch/x86/kernel/cpu/mtrr/main.c b/arch/x86/kernel/cpu/mtrr/main.c
index 5c3d149..ba80d68 100644
--- a/arch/x86/kernel/cpu/mtrr/main.c
+++ b/arch/x86/kernel/cpu/mtrr/main.c
@@ -300,24 +300,24 @@ int mtrr_add_page(unsigned long base, unsigned long size,
 		return error;
 
 	if (type >= MTRR_NUM_TYPES) {
-		pr_warning("mtrr: type: %u invalid\n", type);
+		pr_warn("mtrr: type: %u invalid\n", type);
 		return -EINVAL;
 	}
 
 	/* If the type is WC, check that this processor supports it */
 	if ((type == MTRR_TYPE_WRCOMB) && !have_wrcomb()) {
-		pr_warning("mtrr: your processor doesn't support write-combining\n");
+		pr_warn("mtrr: your processor doesn't support write-combining\n");
 		return -ENOSYS;
 	}
 
 	if (!size) {
-		pr_warning("mtrr: zero sized request\n");
+		pr_warn("mtrr: zero sized request\n");
 		return -EINVAL;
 	}
 
 	if ((base | (base + size - 1)) >>
 	    (boot_cpu_data.x86_phys_bits - PAGE_SHIFT)) {
-		pr_warning("mtrr: base or size exceeds the MTRR width\n");
+		pr_warn("mtrr: base or size exceeds the MTRR width\n");
 		return -EINVAL;
 	}
 
@@ -348,7 +348,7 @@ int mtrr_add_page(unsigned long base, unsigned long size,
 				} else if (types_compatible(type, ltype))
 					continue;
 			}
-			pr_warning("mtrr: 0x%lx000,0x%lx000 overlaps existing"
+			pr_warn("mtrr: 0x%lx000,0x%lx000 overlaps existing"
 				" 0x%lx000,0x%lx000\n", base, size, lbase,
 				lsize);
 			goto out;
@@ -357,7 +357,7 @@ int mtrr_add_page(unsigned long base, unsigned long size,
 		if (ltype != type) {
 			if (types_compatible(type, ltype))
 				continue;
-			pr_warning("mtrr: type mismatch for %lx000,%lx000 old: %s new: %s\n",
+			pr_warn("mtrr: type mismatch for %lx000,%lx000 old: %s new: %s\n",
 				base, size, mtrr_attrib_to_str(ltype),
 				mtrr_attrib_to_str(type));
 			goto out;
@@ -395,7 +395,7 @@ int mtrr_add_page(unsigned long base, unsigned long size,
 static int mtrr_check(unsigned long base, unsigned long size)
 {
 	if ((base & (PAGE_SIZE - 1)) || (size & (PAGE_SIZE - 1))) {
-		pr_warning("mtrr: size and base must be multiples of 4 kiB\n");
+		pr_warn("mtrr: size and base must be multiples of 4 kiB\n");
 		pr_debug("mtrr: size: 0x%lx  base: 0x%lx\n", size, base);
 		dump_stack();
 		return -1;
@@ -493,16 +493,16 @@ int mtrr_del_page(int reg, unsigned long base, unsigned long size)
 		}
 	}
 	if (reg >= max) {
-		pr_warning("mtrr: register: %d too big\n", reg);
+		pr_warn("mtrr: register: %d too big\n", reg);
 		goto out;
 	}
 	mtrr_if->get(reg, &lbase, &lsize, &ltype);
 	if (lsize < 1) {
-		pr_warning("mtrr: MTRR %d not used\n", reg);
+		pr_warn("mtrr: MTRR %d not used\n", reg);
 		goto out;
 	}
 	if (mtrr_usage_table[reg] < 1) {
-		pr_warning("mtrr: reg: %d has count=0\n", reg);
+		pr_warn("mtrr: reg: %d has count=0\n", reg);
 		goto out;
 	}
 	if (--mtrr_usage_table[reg] < 1)
diff --git a/arch/x86/kernel/cpu/rdrand.c b/arch/x86/kernel/cpu/rdrand.c
index 819d949..f6f50c4 100644
--- a/arch/x86/kernel/cpu/rdrand.c
+++ b/arch/x86/kernel/cpu/rdrand.c
@@ -51,7 +51,7 @@ void x86_init_rdrand(struct cpuinfo_x86 *c)
 	for (i = 0; i < SANITY_CHECK_LOOPS; i++) {
 		if (!rdrand_long(&tmp)) {
 			clear_cpu_cap(c, X86_FEATURE_RDRAND);
-			printk_once(KERN_WARNING "rdrand: disabled\n");
+			pr_warn_once("rdrand: disabled\n");
 			return;
 		}
 	}
diff --git a/arch/x86/kernel/cpu/topology.c b/arch/x86/kernel/cpu/topology.c
index 4c60eaf..cd53135 100644
--- a/arch/x86/kernel/cpu/topology.c
+++ b/arch/x86/kernel/cpu/topology.c
@@ -87,10 +87,10 @@ void detect_extended_topology(struct cpuinfo_x86 *c)
 	c->x86_max_cores = (core_level_siblings / smp_num_siblings);
 
 	if (!printed) {
-		printk(KERN_INFO  "CPU: Physical Processor ID: %d\n",
+		pr_info("CPU: Physical Processor ID: %d\n",
 		       c->phys_proc_id);
 		if (c->x86_max_cores > 1)
-			printk(KERN_INFO  "CPU: Processor Core ID: %d\n",
+			pr_info("CPU: Processor Core ID: %d\n",
 			       c->cpu_core_id);
 		printed = 1;
 	}
diff --git a/arch/x86/kernel/cpu/transmeta.c b/arch/x86/kernel/cpu/transmeta.c
index 252da7a..e3b4d18 100644
--- a/arch/x86/kernel/cpu/transmeta.c
+++ b/arch/x86/kernel/cpu/transmeta.c
@@ -33,7 +33,7 @@ static void init_transmeta(struct cpuinfo_x86 *c)
 	if (max >= 0x80860001) {
 		cpuid(0x80860001, &dummy, &cpu_rev, &cpu_freq, &cpu_flags);
 		if (cpu_rev != 0x02000000) {
-			printk(KERN_INFO "CPU: Processor revision %u.%u.%u.%u, %u MHz\n",
+			pr_info("CPU: Processor revision %u.%u.%u.%u, %u MHz\n",
 				(cpu_rev >> 24) & 0xff,
 				(cpu_rev >> 16) & 0xff,
 				(cpu_rev >> 8) & 0xff,
@@ -44,10 +44,10 @@ static void init_transmeta(struct cpuinfo_x86 *c)
 	if (max >= 0x80860002) {
 		cpuid(0x80860002, &new_cpu_rev, &cms_rev1, &cms_rev2, &dummy);
 		if (cpu_rev == 0x02000000) {
-			printk(KERN_INFO "CPU: Processor revision %08X, %u MHz\n",
+			pr_info("CPU: Processor revision %08X, %u MHz\n",
 				new_cpu_rev, cpu_freq);
 		}
-		printk(KERN_INFO "CPU: Code Morphing Software revision %u.%u.%u-%u-%u\n",
+		pr_info("CPU: Code Morphing Software revision %u.%u.%u-%u-%u\n",
 		       (cms_rev1 >> 24) & 0xff,
 		       (cms_rev1 >> 16) & 0xff,
 		       (cms_rev1 >> 8) & 0xff,
@@ -76,7 +76,7 @@ static void init_transmeta(struct cpuinfo_x86 *c)
 		      (void *)&cpu_info[56],
 		      (void *)&cpu_info[60]);
 		cpu_info[64] = '\0';
-		printk(KERN_INFO "CPU: %s\n", cpu_info);
+		pr_info("CPU: %s\n", cpu_info);
 	}
 
 	/* Unhide possibly hidden capability flags */
diff --git a/arch/x86/kernel/cpu/vmware.c b/arch/x86/kernel/cpu/vmware.c
index 628a059..364e583 100644
--- a/arch/x86/kernel/cpu/vmware.c
+++ b/arch/x86/kernel/cpu/vmware.c
@@ -62,7 +62,7 @@ static unsigned long vmware_get_tsc_khz(void)
 	tsc_hz = eax | (((uint64_t)ebx) << 32);
 	do_div(tsc_hz, 1000);
 	BUG_ON(tsc_hz >> 32);
-	printk(KERN_INFO "TSC freq read from hypervisor : %lu.%03lu MHz\n",
+	pr_info("TSC freq read from hypervisor : %lu.%03lu MHz\n",
 			 (unsigned long) tsc_hz / 1000,
 			 (unsigned long) tsc_hz % 1000);
 
@@ -84,8 +84,7 @@ static void __init vmware_platform_setup(void)
 	if (ebx != UINT_MAX)
 		x86_platform.calibrate_tsc = vmware_get_tsc_khz;
 	else
-		printk(KERN_WARNING
-		       "Failed to get TSC freq from the hypervisor\n");
+		pr_warn("Failed to get TSC freq from the hypervisor\n");
 }
 
 /*
diff --git a/arch/x86/kernel/mpparse.c b/arch/x86/kernel/mpparse.c
index 30ca760..97340f2 100644
--- a/arch/x86/kernel/mpparse.c
+++ b/arch/x86/kernel/mpparse.c
@@ -408,7 +408,7 @@ static inline void __init construct_default_ISA_mptable(int mpc_default_type)
 	processor.cpuflag = CPU_ENABLED;
 	processor.cpufeature = (boot_cpu_data.x86 << 8) |
 	    (boot_cpu_data.x86_model << 4) | boot_cpu_data.x86_mask;
-	processor.featureflag = boot_cpu_data.x86_capability[0];
+	processor.featureflag = boot_cpu_data.x86_capability[CPUID_1_EDX];
 	processor.reserved[0] = 0;
 	processor.reserved[1] = 0;
 	for (i = 0; i < 2; i++) {
diff --git a/arch/x86/kernel/nmi.c b/arch/x86/kernel/nmi.c
index 8a2cdd7..04b132a 100644
--- a/arch/x86/kernel/nmi.c
+++ b/arch/x86/kernel/nmi.c
@@ -30,6 +30,7 @@
 #include <asm/nmi.h>
 #include <asm/x86_init.h>
 #include <asm/reboot.h>
+#include <asm/cache.h>
 
 #define CREATE_TRACE_POINTS
 #include <trace/events/nmi.h>
@@ -69,7 +70,7 @@ struct nmi_stats {
 
 static DEFINE_PER_CPU(struct nmi_stats, nmi_stats);
 
-static int ignore_nmis;
+static int ignore_nmis __read_mostly;
 
 int unknown_nmi_panic;
 /*
diff --git a/arch/x86/kernel/smpboot.c b/arch/x86/kernel/smpboot.c
index 24d57f7..3bf1e0b 100644
--- a/arch/x86/kernel/smpboot.c
+++ b/arch/x86/kernel/smpboot.c
@@ -97,6 +97,14 @@ DEFINE_PER_CPU_READ_MOSTLY(cpumask_var_t, cpu_llc_shared_map);
 DEFINE_PER_CPU_READ_MOSTLY(struct cpuinfo_x86, cpu_info);
 EXPORT_PER_CPU_SYMBOL(cpu_info);
 
+/* Logical package management. We might want to allocate that dynamically */
+static int *physical_to_logical_pkg __read_mostly;
+static unsigned long *physical_package_map __read_mostly;;
+static unsigned long *logical_package_map  __read_mostly;
+static unsigned int max_physical_pkg_id __read_mostly;
+unsigned int __max_logical_packages __read_mostly;
+EXPORT_SYMBOL(__max_logical_packages);
+
 static inline void smpboot_setup_warm_reset_vector(unsigned long start_eip)
 {
 	unsigned long flags;
@@ -251,6 +259,97 @@ static void notrace start_secondary(void *unused)
 	cpu_startup_entry(CPUHP_ONLINE);
 }
 
+int topology_update_package_map(unsigned int apicid, unsigned int cpu)
+{
+	unsigned int new, pkg = apicid >> boot_cpu_data.x86_coreid_bits;
+
+	/* Called from early boot ? */
+	if (!physical_package_map)
+		return 0;
+
+	if (pkg >= max_physical_pkg_id)
+		return -EINVAL;
+
+	/* Set the logical package id */
+	if (test_and_set_bit(pkg, physical_package_map))
+		goto found;
+
+	if (pkg < __max_logical_packages) {
+		set_bit(pkg, logical_package_map);
+		physical_to_logical_pkg[pkg] = pkg;
+		goto found;
+	}
+	new = find_first_zero_bit(logical_package_map, __max_logical_packages);
+	if (new >= __max_logical_packages) {
+		physical_to_logical_pkg[pkg] = -1;
+		pr_warn("APIC(%x) Package %u exceeds logical package map\n",
+			apicid, pkg);
+		return -ENOSPC;
+	}
+	set_bit(new, logical_package_map);
+	pr_info("APIC(%x) Converting physical %u to logical package %u\n",
+		apicid, pkg, new);
+	physical_to_logical_pkg[pkg] = new;
+
+found:
+	cpu_data(cpu).logical_proc_id = physical_to_logical_pkg[pkg];
+	return 0;
+}
+
+/**
+ * topology_phys_to_logical_pkg - Map a physical package id to a logical
+ *
+ * Returns logical package id or -1 if not found
+ */
+int topology_phys_to_logical_pkg(unsigned int phys_pkg)
+{
+	if (phys_pkg >= max_physical_pkg_id)
+		return -1;
+	return physical_to_logical_pkg[phys_pkg];
+}
+EXPORT_SYMBOL(topology_phys_to_logical_pkg);
+
+static void __init smp_init_package_map(void)
+{
+	unsigned int ncpus, cpu;
+	size_t size;
+
+	/*
+	 * Today neither Intel nor AMD support heterogenous systems. That
+	 * might change in the future....
+	 */
+	ncpus = boot_cpu_data.x86_max_cores * smp_num_siblings;
+	__max_logical_packages = DIV_ROUND_UP(nr_cpu_ids, ncpus);
+
+	/*
+	 * Possibly larger than what we need as the number of apic ids per
+	 * package can be smaller than the actual used apic ids.
+	 */
+	max_physical_pkg_id = DIV_ROUND_UP(MAX_LOCAL_APIC, ncpus);
+	size = max_physical_pkg_id * sizeof(unsigned int);
+	physical_to_logical_pkg = kmalloc(size, GFP_KERNEL);
+	memset(physical_to_logical_pkg, 0xff, size);
+	size = BITS_TO_LONGS(max_physical_pkg_id) * sizeof(unsigned long);
+	physical_package_map = kzalloc(size, GFP_KERNEL);
+	size = BITS_TO_LONGS(__max_logical_packages) * sizeof(unsigned long);
+	logical_package_map = kzalloc(size, GFP_KERNEL);
+
+	pr_info("Max logical packages: %u\n", __max_logical_packages);
+
+	for_each_present_cpu(cpu) {
+		unsigned int apicid = apic->cpu_present_to_apicid(cpu);
+
+		if (apicid == BAD_APICID || !apic->apic_id_valid(apicid))
+			continue;
+		if (!topology_update_package_map(apicid, cpu))
+			continue;
+		pr_warn("CPU %u APICId %x disabled\n", cpu, apicid);
+		per_cpu(x86_bios_cpu_apicid, cpu) = BAD_APICID;
+		set_cpu_possible(cpu, false);
+		set_cpu_present(cpu, false);
+	}
+}
+
 void __init smp_store_boot_cpu_info(void)
 {
 	int id = 0; /* CPU 0 */
@@ -258,6 +357,7 @@ void __init smp_store_boot_cpu_info(void)
 
 	*c = boot_cpu_data;
 	c->cpu_index = id;
+	smp_init_package_map();
 }
 
 /*
diff --git a/arch/x86/lguest/boot.c b/arch/x86/lguest/boot.c
index f56cc41..fd57d3a 100644
--- a/arch/x86/lguest/boot.c
+++ b/arch/x86/lguest/boot.c
@@ -1529,7 +1529,7 @@ __init void lguest_init(void)
 	 */
 	cpu_detect(&new_cpu_data);
 	/* head.S usually sets up the first capability word, so do it here. */
-	new_cpu_data.x86_capability[0] = cpuid_edx(1);
+	new_cpu_data.x86_capability[CPUID_1_EDX] = cpuid_edx(1);
 
 	/* Math is always hard! */
 	set_cpu_cap(&new_cpu_data, X86_FEATURE_FPU);
diff --git a/arch/x86/xen/enlighten.c b/arch/x86/xen/enlighten.c
index d09e4c9..2c26108 100644
--- a/arch/x86/xen/enlighten.c
+++ b/arch/x86/xen/enlighten.c
@@ -1654,7 +1654,7 @@ asmlinkage __visible void __init xen_start_kernel(void)
 	cpu_detect(&new_cpu_data);
 	set_cpu_cap(&new_cpu_data, X86_FEATURE_FPU);
 	new_cpu_data.wp_works_ok = 1;
-	new_cpu_data.x86_capability[0] = cpuid_edx(1);
+	new_cpu_data.x86_capability[CPUID_1_EDX] = cpuid_edx(1);
 #endif
 
 	if (xen_start_info->mod_start) {
diff --git a/arch/x86/xen/pmu.c b/arch/x86/xen/pmu.c
index 724a087..9466354 100644
--- a/arch/x86/xen/pmu.c
+++ b/arch/x86/xen/pmu.c
@@ -11,7 +11,7 @@
 #include "pmu.h"
 
 /* x86_pmu.handle_irq definition */
-#include "../kernel/cpu/perf_event.h"
+#include "../events/perf_event.h"
 
 #define XENPMU_IRQ_PROCESSING    1
 struct xenpmu {