[PATCH 3/3] powerpc/64s: cpufeatures: add initial implementation for cpufeatures

Nicholas Piggin npiggin at gmail.com
Thu Apr 13 03:48:32 AEST 2017


The /cpus/features dt binding describes architected CPU features along
with some compatibility, privilege, and enablement properties that allow
flexibility with discovering and enabling capabilities.

Presence of this feature implies a base level of functionality, then
additional feature nodes advertise the presence of new features.

A given feature and its setup procedure is defined once and used by all
CPUs which are compatible by that feature. Features that follow a
supported "prescription" can be enabled by a hypervisor or OS that
does not understand them natively.

One difference with features after this patch is that PPC_FEATURE2_EBB
is set independent of PMU init. EBB facility is more general than PMU,
so I think this is reasonable.

Signed-off-by: Nicholas Piggin <npiggin at gmail.com>

---
Since last post:
- Looked at XIVE patches compatibility. There is a dependency there now
  on LPCR setup (HEIC, LPES, etc).
- Split vector-scalar feature from vector
- Get cpu_name from cputable if exists, for /proc/cpuinfo compatibility.


Testing under mambo shows a few differences with POWER8, but they all
seem to be mambo related:
- HFSCR bit 54 and 57 are now clear (mambo sets at init)
- PMAO_BUG is set. This is due to mambo setting architected POWER8 mode
  and POWER8E PVR. Current kernels lose PMAO_BUG bit.
- CI_LARGE_PAGE is now set (mambo boot does not set it for some reason,
  haven't looked at why).

Under POWER9 I haven't found differences.

 .../devicetree/bindings/powerpc/cpufeatures.txt    | 264 +++++++++
 arch/powerpc/Kconfig                               |  16 +
 arch/powerpc/include/asm/cpu_has_feature.h         |   4 +-
 arch/powerpc/include/asm/cpufeatures.h             |  57 ++
 arch/powerpc/include/asm/cputable.h                |   2 +
 arch/powerpc/include/asm/reg.h                     |   1 +
 arch/powerpc/kernel/Makefile                       |   1 +
 arch/powerpc/kernel/cpufeatures.c                  | 631 +++++++++++++++++++++
 arch/powerpc/kernel/cputable.c                     |  37 +-
 arch/powerpc/kernel/prom.c                         | 333 ++++++++++-
 arch/powerpc/kernel/setup-common.c                 |   2 +-
 arch/powerpc/kernel/setup_64.c                     |  15 +-
 12 files changed, 1346 insertions(+), 17 deletions(-)
 create mode 100644 Documentation/devicetree/bindings/powerpc/cpufeatures.txt
 create mode 100644 arch/powerpc/include/asm/cpufeatures.h
 create mode 100644 arch/powerpc/kernel/cpufeatures.c

diff --git a/Documentation/devicetree/bindings/powerpc/cpufeatures.txt b/Documentation/devicetree/bindings/powerpc/cpufeatures.txt
new file mode 100644
index 000000000000..325b263f4cdf
--- /dev/null
+++ b/Documentation/devicetree/bindings/powerpc/cpufeatures.txt
@@ -0,0 +1,264 @@
+powerpc cpu features binding
+============================
+
+The device tree describes supported CPU features as nodes containing
+compatibility and enablement information as properties.
+
+The binding specifies features common to all CPUs in the system.
+Heterogeneous CPU features are not supported at present (such could be added
+by providing nodes with additional features and linking those to particular
+CPUs).
+
+This binding is intended to provide fine grained control of CPU features at
+all levels of the stack (firmware, hypervisor, OS, userspace), with the
+ability for new CPU features to be used by some components without all
+components being upgraded (e.g., a new floating point instruction could be
+used by userspace math library without upgrading kernel and hypervisor).
+
+The binding is passed to the hypervisor by firmware. The hypervisor must
+remove any features that require hypervisor enablement but that it does not
+enable. It must remove any features that depend on removed features. It may
+pass remaining features usable to the OS and PR to guests, depending on
+configuration policy (not specified here).
+
+The modified binding is passed to the guest by hypervisor, with HV bit
+cleared from the usable-mask and the hv-support and hfscr-bit properties
+removed. The guest must similarly rmeove features that require OS enablement
+that it does not enable. The OS may pass PR usable features to userspace via
+ELF AUX vectors AT_HWCAP, AT_HWCAP2, AT_HWCAP3, etc., or use some other
+method (outside the scope of this specification).
+
+The binding will specify a "base" level of features that will be present
+when the cpu features binding exists. Additional features will be explicitly
+specified.
+
+/cpus/features node binding
+---------------------------
+
+Node: features
+
+Description: Container of CPU feature nodes.
+
+The node name must be "features" and it must be a child of the node "/cpus".
+
+The node is optional but should be provided by new firmware.
+
+Each child node of cpufeatures represents an architected CPU feature (e.g.,
+a new set of vector instructions) or an important CPU performance
+characteristic (e.g., fast unaligned memory operations). The specification
+of each feature (instructions, registers, exceptions, etc.) will be
+documented with device tree bindings.
+
+As a rough guide, features should be based on functional groups of changes,
+those that share common enablement requirements, that share particular
+registers or functionality. For example, "POWER9" would be too general, and
+a new feature for every instruction would be too specific. The "summary of
+changes" preface in Power ISA specification is a good guideline for the
+architected features.
+
+Features should be "positive" where possible. For example the presence of a
+feature node should indicate the presence of a new CPU feature (e.g., a
+new instruction or register). An errata workaround for example would then
+remove that feature from the device tree to disable it. "Negative" features
+may be unavoidable in some cases.
+
+Properties:
+
+- device_type
+  Usage: required
+  Value type: string
+  Definition: "cpu-features"
+
+- isa
+  Usage: required
+  Value type: <u32>
+  Definition:
+
+  isa that the CPU is currently running in. This provides instruction set
+  compatibility, less the individual feature nodes. For example, an ISA v3.0
+  implementation that lacks the "transactional-memory" cpufeature node
+  should not use transactional memory facilities.
+
+  Value corresponds to the "Power ISA Version" multiplied by 1000.
+  For example, <3000> corresponds to Version 3.0, <2070> to Version 2.07.
+  The minor digit is available for revisions.
+
+/cpus/features/feature node bindings
+------------------------------------
+
+Node: A string describing an architected CPU feature, e.g., "vsx".
+
+Description: An architected feature supported by the CPUs.
+
+The name of the node will follow a convention such that software will
+match known features by a string comparison with the node name. Presence
+of the node indicates the feature is available to use (XXX: could
+advertise all supported by hardware, with disabled/enabled status
+property).
+
+The name of the child node corresponds to the name of the feature.
+Software will detect known features by string matching.
+
+Properties:
+
+- isa
+  Usage: required
+  Value type: <u32>
+  Definition:
+
+  First level of the Power ISA that the feature appears in.
+  Software should filter out features when constraining the
+  environment to a particular ISA version.
+
+  Value is defined similarly to /cpus/features/isa
+
+- usable-mask
+  Usage: required
+  Value type: <u32> bit mask
+  Definition:
+              Bit numbers are LSB0
+              bit 0 - PR (problem state / user mode)
+              bit 1 - OS (privileged state)
+              bit 2 - HV (hypervisor state)
+              All other bits reserved and should be zero.
+
+  This property describes the privilege levels and/or software components
+  that can use the feature.
+
+  If bit 0 is set, then the hwcap-bit-nr property will exist.
+
+  Reserved bits must be treated as the feature being unsupported.
+
+- hv-support
+  Usage: optional
+  Value type: <u32>
+  Definition:
+              1 -  Custom
+              2 -  HFSCR
+              Other values reserved.
+
+  This property describes the HV privilege state support required to
+  enable the feature. If the property does not exist then no support is
+  required.
+
+  If the value of this property is 1, then the hypervisor must have
+  explicit support for this feature.
+
+  If the value of this property is 2, then the hfscr-bit-nr property
+  will exist.
+
+  Reserved values must be treated as the feature being unsupported.
+
+- os-support
+  Usage: optional
+  Value type: <u32>
+  Definition:
+              1 -  Custom
+              2 -  FSCR
+              Other values reserved.
+
+  This property describes the OS privilege state support required to
+  enable the feature. If the property does not exist then no support is
+  required.
+
+  If the value of this property is 1, then the operating system must
+  have explicit support for this feature.
+
+  If the value of this property is 2, then the fscr-bit-nr property will
+  exist.
+
+  Reserved values must be treated as the feature being unsupported.
+
+- hfscr-bit-nr
+  Usage: optional
+  Value type: <u32>
+  Definition: HFSCR bit position (LSB0)
+
+  This property exists when the hv-support property value is 2. This
+  property describes the bit number in the HFSCR register that the
+  hypervisor must set in order to enable this feature.
+
+- fscr-bit-nr
+  Usage: optional
+  Value type: <u32>
+  Definition: FSCR bit position (LSB0)
+
+  This property exists when the os-support property value is 2. This
+  property describes the bit number in the FSCR register that the
+  operating system must set in order to enable this feature.
+
+- hwcap-bit-nr
+  Usage: optional
+  Value type: <u32>
+  Definition: AUX vector bit position (LSB0)
+
+  This property may exist when the usable-mask property value has bit 0
+  (PR) set. This property describes the bit number that should be set in
+  the ELF AUX hardware capability vectors in order to advertise this
+  feature to userspace. Bits 0-31 correspond to bits 0-31 in AT_HWCAP
+  vector. Bits 32-63 correspond to 0-31 in AT_HWCAP2 vector, and so on.
+  Missing AT_HWCAPx vectors implies that the feature is not enabled or
+  can not be advertised. Operating systems may provide a number of
+  unassigned hardware capability bits to allow for new features to be
+  advertised.
+
+  Some properties representing features created before this binding are
+  advertised to userspace without a one-to-one hwcap bit number may not
+  specify this bit. Operating system will handle those bits specifically.
+  All new features usable by userspace will have a hwcap-bit-nr property.
+
+- dependencies
+  Usage: optional
+  Value type: <prop-encoded-array>
+  Definition:
+
+  If this property exists then it is a list of phandles to cpu feature
+  nodes that must be enabled for this feature to be enabled.
+
+
+Example
+-------
+
+	/cpus/features {
+		device_type = "cpu-features";
+
+		isa = <3020>;
+
+		darn {
+			isa = <3000>;
+			usable-mask = <1 | 2 | 4>;
+			hwcap-bit-nr = <xx>;
+		};
+
+		scv {
+			isa = <3000>;
+			usable-mask = <1 | 2>;
+			os-support = <1>;
+			hwcap-bit-nr = <xx>;
+		};
+
+		stop {
+			isa = <3000>;
+			usable-mask = <2 | 4>;
+			hv-support = <1>;
+			os-support = <1>;
+		};
+
+		vsx2 (hypothetical) {
+			isa = <3010>;
+			usable-mask = <1 | 2 | 4>;
+			hv-support = <1>;
+			os-support = <1>;
+			hwcap-bit-nr = <xx>;
+		};
+
+		vsx2-newinsns {
+			isa = <3020>;
+			usable-mask = <1 | 2 | 4>;
+			os-support = <2>;
+			fscr-bit-nr = <xx>;
+			hwcap-bit-nr = <xx>;
+			dependencies = <&vsx2>;
+		};
+
+	};
diff --git a/arch/powerpc/Kconfig b/arch/powerpc/Kconfig
index 97a8bc8a095c..c4804cd65b9a 100644
--- a/arch/powerpc/Kconfig
+++ b/arch/powerpc/Kconfig
@@ -330,6 +330,22 @@ source "arch/powerpc/platforms/Kconfig"
 
 menu "Kernel options"
 
+config PPC_CPUFEATURES_DT
+	bool "cpufeatures cpu setup and device tree support (UNSTABLE)"
+	depends on PPC_BOOK3S_64
+	default n
+	---help---
+	  cpufeatures is a proposed new device tree binding for describing
+	  CPU compatibility and features. Saying y here will attempt to use
+	  cpufeatures binding if the firmware provides it. This binding is
+	  not stable, so a kernel built with this option may not function
+	  properly with future firmware.
+
+config PPC_CPUFEATURES_ENABLE_UNKNOWN
+	bool "cpufeatures pass through unknown features to guest/userspace"
+	depends on PPC_CPUFEATURES_DT
+	default y
+
 config HIGHMEM
 	bool "High memory support"
 	depends on PPC32
diff --git a/arch/powerpc/include/asm/cpu_has_feature.h b/arch/powerpc/include/asm/cpu_has_feature.h
index 6e834caa3720..445495aa2bbf 100644
--- a/arch/powerpc/include/asm/cpu_has_feature.h
+++ b/arch/powerpc/include/asm/cpu_has_feature.h
@@ -1,5 +1,5 @@
-#ifndef __ASM_POWERPC_CPUFEATURES_H
-#define __ASM_POWERPC_CPUFEATURES_H
+#ifndef __ASM_POWERPC_CPU_HAS_FEATURE_H
+#define __ASM_POWERPC_CPU_HAS_FEATURE_H
 
 #ifndef __ASSEMBLY__
 
diff --git a/arch/powerpc/include/asm/cpufeatures.h b/arch/powerpc/include/asm/cpufeatures.h
new file mode 100644
index 000000000000..edd91b3051ce
--- /dev/null
+++ b/arch/powerpc/include/asm/cpufeatures.h
@@ -0,0 +1,57 @@
+#ifndef __ASM_POWERPC_CPUFEATURES_H
+#define __ASM_POWERPC_CPUFEATURES_H
+
+/*
+ *  Copyright 2017, IBM Corporation
+ *  cpufeatures is the new way to discover CPU features with /cpus/features
+ *  devicetree. This supersedes PVR based discovery ("cputable"), and older
+ *  device tree feature advertisement.
+ */
+
+#include <linux/types.h>
+#include <asm/asm-compat.h>
+#include <asm/feature-fixups.h>
+#include <uapi/asm/cputable.h>
+
+extern void cpufeatures_setup_cpu(void);
+
+/* Types for device tree parsing */
+#define USABLE_PR               (1U << 0)
+#define USABLE_OS               (1U << 1)
+#define USABLE_HV               (1U << 2)
+
+#define HV_SUPPORT_NONE         0
+#define HV_SUPPORT_CUSTOM       1
+#define HV_SUPPORT_HFSCR        2
+
+#define OS_SUPPORT_NONE         0
+#define OS_SUPPORT_CUSTOM       1
+#define OS_SUPPORT_FSCR         2
+
+#define ISA_BASE                0
+#define ISA_V207                2070
+#define ISA_V3                  3000
+
+struct dt_cpu_feature {
+	const char *name;
+	uint32_t isa;
+	uint32_t usable_mask;
+	uint32_t hv_support;
+	uint32_t os_support;
+	uint32_t hfscr_bit_nr;
+	uint32_t fscr_bit_nr;
+	uint32_t hwcap_bit_nr;
+	/* fdt parsing */
+	unsigned long node;
+	int enabled;
+	int disabled;
+};
+
+extern void cpufeatures_setup_start(u32 isa);
+extern int cpufeatures_process_feature(struct dt_cpu_feature *f);
+extern void cpufeatures_setup_finished(void);
+
+/* kernel/prom.c */
+extern int early_init_devtree_check_cpu_features_exists(void);
+
+#endif
diff --git a/arch/powerpc/include/asm/cputable.h b/arch/powerpc/include/asm/cputable.h
index ab68d0ee7725..4edbc2f7569a 100644
--- a/arch/powerpc/include/asm/cputable.h
+++ b/arch/powerpc/include/asm/cputable.h
@@ -118,7 +118,9 @@ extern struct cpu_spec		*cur_cpu_spec;
 
 extern unsigned int __start___ftr_fixup, __stop___ftr_fixup;
 
+extern void set_cur_cpu_spec(struct cpu_spec *s);
 extern struct cpu_spec *identify_cpu(unsigned long offset, unsigned int pvr);
+extern void identify_cpu_name(unsigned int pvr);
 extern void do_feature_fixups(unsigned long value, void *fixup_start,
 			      void *fixup_end);
 
diff --git a/arch/powerpc/include/asm/reg.h b/arch/powerpc/include/asm/reg.h
index fc879fd6bdae..a43f2ad27b5d 100644
--- a/arch/powerpc/include/asm/reg.h
+++ b/arch/powerpc/include/asm/reg.h
@@ -365,6 +365,7 @@
 #define   LPCR_MER_SH		11
 #define	  LPCR_GTSE		ASM_CONST(0x0000000000000400)  	/* Guest Translation Shootdown Enable */
 #define   LPCR_TC		ASM_CONST(0x0000000000000200)	/* Translation control */
+#define   LPCR_HEIC		ASM_CONST(0x0000000000000010)	/* Hypervisor External Interrupt Control */
 #define   LPCR_LPES		0x0000000c
 #define   LPCR_LPES0		ASM_CONST(0x0000000000000008)      /* LPAR Env selector 0 */
 #define   LPCR_LPES1		ASM_CONST(0x0000000000000004)      /* LPAR Env selector 1 */
diff --git a/arch/powerpc/kernel/Makefile b/arch/powerpc/kernel/Makefile
index 811f441a125f..9b264ebe6ac1 100644
--- a/arch/powerpc/kernel/Makefile
+++ b/arch/powerpc/kernel/Makefile
@@ -58,6 +58,7 @@ obj-$(CONFIG_PPC_RTAS)		+= rtas.o rtas-rtc.o $(rtaspci-y-y)
 obj-$(CONFIG_PPC_RTAS_DAEMON)	+= rtasd.o
 obj-$(CONFIG_RTAS_FLASH)	+= rtas_flash.o
 obj-$(CONFIG_RTAS_PROC)		+= rtas-proc.o
+obj-$(CONFIG_PPC_CPUFEATURES_DT)	+= cpufeatures.o
 obj-$(CONFIG_EEH)              += eeh.o eeh_pe.o eeh_dev.o eeh_cache.o \
 				  eeh_driver.o eeh_event.o eeh_sysfs.o
 obj-$(CONFIG_GENERIC_TBSYNC)	+= smp-tbsync.o
diff --git a/arch/powerpc/kernel/cpufeatures.c b/arch/powerpc/kernel/cpufeatures.c
new file mode 100644
index 000000000000..1f5fce1506bb
--- /dev/null
+++ b/arch/powerpc/kernel/cpufeatures.c
@@ -0,0 +1,631 @@
+/*
+ *  Copyright 2017, IBM Corporation
+ */
+
+#include <linux/string.h>
+#include <linux/sched.h>
+#include <linux/threads.h>
+#include <linux/init.h>
+#include <linux/export.h>
+#include <linux/jump_label.h>
+
+#include <asm/cpufeatures.h>
+#include <asm/cputable.h>
+#include <asm/prom.h>		/* for PTRRELOC on ARCH=ppc */
+#include <asm/oprofile_impl.h>
+#include <asm/mmu.h>
+#include <asm/setup.h>
+
+#ifdef DEBUG
+#define DBG(fmt...) pr_err(fmt)
+#else
+#define DBG(fmt...)
+#endif
+
+#define CPU_FTRS_BASE \
+	   (CPU_FTR_USE_TB | \
+	    CPU_FTR_LWSYNC | \
+	    CPU_FTR_FPU_UNAVAILABLE |\
+	    CPU_FTR_NODSISRALIGN |\
+	    CPU_FTR_NOEXECUTE |\
+	    CPU_FTR_COHERENT_ICACHE | \
+	    CPU_FTR_STCX_CHECKS_ADDRESS |\
+	    CPU_FTR_POPCNTB | CPU_FTR_POPCNTD | \
+	    CPU_FTR_DAWR | \
+	    CPU_FTR_ARCH_206 |\
+	    CPU_FTR_ARCH_207S)
+
+#define MMU_FTRS_HASH_BASE (MMU_FTRS_POWER8)
+
+#define COMMON_USER_BASE	(PPC_FEATURE_32 | PPC_FEATURE_64 | \
+				 PPC_FEATURE_ARCH_2_06 |\
+				 PPC_FEATURE_ICACHE_SNOOP)
+#define COMMON_USER2_BASE	(PPC_FEATURE2_ARCH_2_07 | \
+				 PPC_FEATURE2_ISEL)
+/*
+ * Set up the base CPU
+ */
+
+extern void __flush_tlb_power8(unsigned int action);
+extern void __flush_tlb_power9(unsigned int action);
+extern long __machine_check_early_realmode_p8(struct pt_regs *regs);
+extern long __machine_check_early_realmode_p9(struct pt_regs *regs);
+
+static int hv_mode;
+
+static struct {
+	u64	hfscr;
+	u64	fscr;
+	u64	lpcr;
+} system_registers;
+
+static void (*init_pmu_registers)(void);
+
+static void __restore_cpu_cpufeatures(void)
+{
+	if (hv_mode) {
+		mtspr(SPRN_LPID, 0);
+		mtspr(SPRN_HFSCR, system_registers.hfscr);
+	}
+	mtspr(SPRN_FSCR, system_registers.fscr);
+	mtspr(SPRN_LPCR, system_registers.lpcr);
+
+	if (init_pmu_registers)
+		init_pmu_registers();
+}
+
+void __init cpufeatures_setup_cpu(void)
+{
+	struct cpu_spec s = {
+		.cpu_name		= NULL,
+		.cpu_features		= CPU_FTRS_BASE,
+		.cpu_user_features	= COMMON_USER_BASE,
+		.cpu_user_features2	= COMMON_USER2_BASE,
+		.mmu_features		= 0,
+		.icache_bsize		= 32, /* minimum block size, fixed by */
+		.dcache_bsize		= 32, /* cache info init.             */
+		.num_pmcs		= 0,
+		.pmc_type		= PPC_PMC_DEFAULT,
+		.oprofile_cpu_type	= NULL,
+		.oprofile_type		= PPC_OPROFILE_INVALID,
+		.cpu_setup		= NULL,
+		.cpu_restore		= __restore_cpu_cpufeatures,
+		.flush_tlb		= NULL,
+		.machine_check_early	= NULL,
+		.platform		= NULL,
+	};
+
+	s.pvr_mask = s.pvr_value = mfspr(SPRN_PVR);
+
+	set_cur_cpu_spec(&s);
+
+	/* Initialize the base environment -- clear FSCR/HFSCR.  */
+	hv_mode = !!(mfmsr() & MSR_HV);
+	if (hv_mode)
+		mtspr(SPRN_HFSCR, 0);
+	mtspr(SPRN_FSCR, 0);
+}
+
+static int __init feat_try_enable_unknown(struct dt_cpu_feature *f)
+{
+	if (f->hv_support == HV_SUPPORT_NONE) {
+	} else if (f->hv_support == HV_SUPPORT_HFSCR) {
+		u64 hfscr = mfspr(SPRN_HFSCR);
+		hfscr |= 1UL << f->hfscr_bit_nr;
+		mtspr(SPRN_HFSCR, hfscr);
+	} else {
+		return 0;
+	}
+
+	if (f->os_support == OS_SUPPORT_NONE) {
+	} else if (f->os_support == OS_SUPPORT_FSCR) {
+		u64 fscr = mfspr(SPRN_FSCR);
+		fscr |= 1UL << f->fscr_bit_nr;
+		mtspr(SPRN_FSCR, fscr);
+	} else {
+		return 0;
+	}
+
+	if ((f->usable_mask & USABLE_PR) && (f->hwcap_bit_nr != -1)) {
+		uint32_t word = f->hwcap_bit_nr / 32;
+		uint32_t bit = f->hwcap_bit_nr % 32;
+
+		if (word == 0)
+			cur_cpu_spec->cpu_user_features |= 1U << bit;
+		else if (word == 1)
+			cur_cpu_spec->cpu_user_features2 |= 1U << bit;
+		else
+			pr_err("CPU feature: %s could not advertise to user (no hwcap bits)\n", f->name);
+	}
+
+	return 1;
+}
+
+static int __init feat_enable(struct dt_cpu_feature *f)
+{
+	if (f->hv_support) {
+		if (f->hfscr_bit_nr != -1) {
+			u64 hfscr = mfspr(SPRN_HFSCR);
+			hfscr |= 1UL << f->hfscr_bit_nr;
+			mtspr(SPRN_HFSCR, hfscr);
+		}
+	}
+
+	if (f->os_support) {
+		if (f->fscr_bit_nr != -1) {
+			u64 fscr = mfspr(SPRN_FSCR);
+			fscr |= 1UL << f->fscr_bit_nr;
+			mtspr(SPRN_FSCR, fscr);
+		}
+	}
+
+	if ((f->usable_mask & USABLE_PR) && (f->hwcap_bit_nr != -1)) {
+		uint32_t word = f->hwcap_bit_nr / 32;
+		uint32_t bit = f->hwcap_bit_nr % 32;
+
+		if (word == 0)
+			cur_cpu_spec->cpu_user_features |= 1U << bit;
+		else if (word == 1)
+			cur_cpu_spec->cpu_user_features2 |= 1U << bit;
+		else
+			pr_err("CPU feature: %s could not advertise to user (no hwcap bits)\n", f->name);
+	}
+
+	return 1;
+}
+
+static int __init feat_disable(struct dt_cpu_feature *f)
+{
+	return 0;
+}
+
+static int __init feat_enable_hv(struct dt_cpu_feature *f)
+{
+	u64 lpcr;
+
+	if (!hv_mode) {
+		pr_err("CPU feature hypervisor present in device tree but HV mode not enabled in the CPU. Ignoring.\n");
+		return 0;
+	}
+
+	mtspr(SPRN_LPID, 0);
+
+	lpcr = mfspr(SPRN_LPCR);
+	lpcr &=  ~LPCR_LPES0; /* HV external interrupts */
+	mtspr(SPRN_LPCR, lpcr);
+
+	cur_cpu_spec->cpu_features |= CPU_FTR_HVMODE;
+
+	return 1;
+}
+
+static int __init feat_enable_le(struct dt_cpu_feature *f)
+{
+	cur_cpu_spec->cpu_user_features |= PPC_FEATURE_TRUE_LE;
+	return 1;
+}
+
+static int __init feat_enable_smt(struct dt_cpu_feature *f)
+{
+	cur_cpu_spec->cpu_features |= CPU_FTR_SMT;
+	cur_cpu_spec->cpu_user_features |= PPC_FEATURE_SMT;
+	return 1;
+}
+
+static int __init feat_enable_idle_nap(struct dt_cpu_feature *f)
+{
+	u64 lpcr;
+
+	/* Set PECE wakeup modes for ISA 207 */
+	lpcr = mfspr(SPRN_LPCR);
+	lpcr |=  LPCR_PECE0;
+	lpcr |=  LPCR_PECE1;
+	lpcr |=  LPCR_PECE2;
+	mtspr(SPRN_LPCR, lpcr);
+
+	return 1;
+}
+
+static int __init feat_enable_align_dsisr(struct dt_cpu_feature *f)
+{
+	cur_cpu_spec->cpu_features &= ~CPU_FTR_NODSISRALIGN;
+
+	return 1;
+}
+
+static int __init feat_enable_idle_stop(struct dt_cpu_feature *f)
+{
+	u64 lpcr;
+
+	/* Set PECE wakeup modes for ISAv3.0B */
+	lpcr = mfspr(SPRN_LPCR);
+	lpcr |=  LPCR_PECE0;
+	lpcr |=  LPCR_PECE1;
+	lpcr |=  LPCR_PECE2;
+	mtspr(SPRN_LPCR, lpcr);
+
+	return 1;
+}
+
+static int __init feat_enable_mmu_hash(struct dt_cpu_feature *f)
+{
+	u64 lpcr;
+
+	lpcr = mfspr(SPRN_LPCR);
+	lpcr &= ~LPCR_ISL;
+
+	/* VRMASD */
+	lpcr |= LPCR_VPM0;
+	lpcr &= ~LPCR_VPM1;
+	lpcr |= 0x10UL << LPCR_VRMASD_SH; /* L=1 LP=00 */
+	mtspr(SPRN_LPCR, lpcr);
+
+	cur_cpu_spec->mmu_features |= MMU_FTRS_HASH_BASE;
+	cur_cpu_spec->cpu_user_features |= PPC_FEATURE_HAS_MMU;
+
+	return 1;
+}
+
+static int __init feat_enable_mmu_hash_v3(struct dt_cpu_feature *f)
+{
+	u64 lpcr;
+
+	lpcr = mfspr(SPRN_LPCR);
+	lpcr &= ~LPCR_ISL;
+	mtspr(SPRN_LPCR, lpcr);
+
+	cur_cpu_spec->mmu_features |= MMU_FTRS_HASH_BASE;
+	cur_cpu_spec->cpu_user_features |= PPC_FEATURE_HAS_MMU;
+
+	return 1;
+}
+
+
+static int __init feat_enable_mmu_radix(struct dt_cpu_feature *f)
+{
+#ifdef CONFIG_PPC_RADIX_MMU
+	cur_cpu_spec->mmu_features |= MMU_FTR_TYPE_RADIX;
+	cur_cpu_spec->mmu_features |= MMU_FTRS_HASH_BASE;
+	cur_cpu_spec->cpu_user_features |= PPC_FEATURE_HAS_MMU;
+
+	return 1;
+#endif
+	return 0;
+}
+
+static int __init feat_enable_dscr(struct dt_cpu_feature *f)
+{
+	u64 lpcr;
+
+	feat_enable(f);
+
+	lpcr = mfspr(SPRN_LPCR);
+	lpcr &= ~LPCR_DPFD;
+	lpcr |=  (4UL << LPCR_DPFD_SH);
+	mtspr(SPRN_LPCR, lpcr);
+
+	return 1;
+}
+
+static void hfscr_pmu_enable(void)
+{
+	u64 hfscr = mfspr(SPRN_HFSCR);
+	hfscr |= PPC_BIT(60);
+	mtspr(SPRN_HFSCR, hfscr);
+}
+
+static void init_pmu_power8(void)
+{
+	if (hv_mode) {
+		mtspr(SPRN_MMCRC, 0);
+		mtspr(SPRN_MMCRH, 0);
+	}
+
+	mtspr(SPRN_MMCRA, 0);
+	mtspr(SPRN_MMCR0, 0);
+	mtspr(SPRN_MMCR1, 0);
+	mtspr(SPRN_MMCR2, 0);
+	mtspr(SPRN_MMCRS, 0);
+}
+
+static int __init feat_enable_mce_power8(struct dt_cpu_feature *f)
+{
+	cur_cpu_spec->platform = "power8";
+	cur_cpu_spec->flush_tlb = __flush_tlb_power8;
+	cur_cpu_spec->machine_check_early = __machine_check_early_realmode_p8;
+
+	return 1;
+}
+
+static int __init feat_enable_pmu_power8(struct dt_cpu_feature *f)
+{
+	hfscr_pmu_enable();
+
+	init_pmu_power8();
+	init_pmu_registers = init_pmu_power8;
+
+	cur_cpu_spec->cpu_features |= CPU_FTR_MMCRA;
+	cur_cpu_spec->cpu_user_features |= PPC_FEATURE_PSERIES_PERFMON_COMPAT;
+	if (pvr_version_is(PVR_POWER8E))
+		cur_cpu_spec->cpu_features |= CPU_FTR_PMAO_BUG;
+
+	cur_cpu_spec->num_pmcs		= 6;
+	cur_cpu_spec->pmc_type		= PPC_PMC_IBM;
+	cur_cpu_spec->oprofile_cpu_type	= "ppc64/power8";
+
+	return 1;
+}
+
+static void init_pmu_power9(void)
+{
+	if (hv_mode)
+		mtspr(SPRN_MMCRC, 0);
+
+	mtspr(SPRN_MMCRA, 0);
+	mtspr(SPRN_MMCR0, 0);
+	mtspr(SPRN_MMCR1, 0);
+	mtspr(SPRN_MMCR2, 0);
+}
+
+static int __init feat_enable_mce_power9(struct dt_cpu_feature *f)
+{
+	cur_cpu_spec->platform = "power9";
+	cur_cpu_spec->flush_tlb = __flush_tlb_power9;
+	cur_cpu_spec->machine_check_early = __machine_check_early_realmode_p9;
+
+	return 1;
+}
+
+static int __init feat_enable_pmu_power9(struct dt_cpu_feature *f)
+{
+	hfscr_pmu_enable();
+
+	init_pmu_power9();
+	init_pmu_registers = init_pmu_power9;
+
+	cur_cpu_spec->cpu_features |= CPU_FTR_MMCRA;
+	cur_cpu_spec->cpu_user_features |= PPC_FEATURE_PSERIES_PERFMON_COMPAT;
+
+	cur_cpu_spec->num_pmcs		= 6;
+	cur_cpu_spec->pmc_type		= PPC_PMC_IBM;
+	cur_cpu_spec->oprofile_cpu_type	= "ppc64/power9";
+
+	return 1;
+}
+
+static int __init feat_enable_tm(struct dt_cpu_feature *f)
+{
+#ifdef CONFIG_PPC_TRANSACTIONAL_MEM
+	feat_enable(f);
+	cur_cpu_spec->cpu_user_features2 |= PPC_FEATURE2_HTM_NOSC;
+	return 1;
+#endif
+	return 0;
+}
+
+static int __init feat_enable_fp(struct dt_cpu_feature *f)
+{
+	feat_enable(f);
+	cur_cpu_spec->cpu_features &= ~CPU_FTR_FPU_UNAVAILABLE;
+
+	return 1;
+}
+
+static int __init feat_enable_vector(struct dt_cpu_feature *f)
+{
+#ifdef CONFIG_ALTIVEC
+	feat_enable(f);
+	cur_cpu_spec->cpu_features |= CPU_FTR_ALTIVEC;
+	cur_cpu_spec->cpu_features |= CPU_FTR_VMX_COPY;
+	cur_cpu_spec->cpu_user_features |= PPC_FEATURE_HAS_ALTIVEC;
+
+	return 1;
+#endif
+	return 0;
+}
+
+static int __init feat_enable_vsx(struct dt_cpu_feature *f)
+{
+#ifdef CONFIG_VSX
+	feat_enable(f);
+	cur_cpu_spec->cpu_features |= CPU_FTR_VSX;
+	cur_cpu_spec->cpu_user_features |= PPC_FEATURE_HAS_VSX;
+
+	return 1;
+#endif
+	return 0;
+}
+
+static int __init feat_enable_purr(struct dt_cpu_feature *f)
+{
+	cur_cpu_spec->cpu_features |= CPU_FTR_PURR | CPU_FTR_SPURR;
+
+	return 1;
+}
+
+static int __init feat_enable_dbell(struct dt_cpu_feature *f)
+{
+	u64 lpcr;
+
+	/* P9 has an HFSCR for privileged state */
+	feat_enable(f);
+
+	cur_cpu_spec->cpu_features |= CPU_FTR_DBELL;
+
+	lpcr = mfspr(SPRN_LPCR);
+	lpcr |=  LPCR_PECEDH; /* hyp doorbell wakeup */
+	mtspr(SPRN_LPCR, lpcr);
+
+	return 1;
+}
+
+static int __init feat_enable_hvi(struct dt_cpu_feature *f)
+{
+	u64 lpcr;
+
+	/*
+	 * POWER9 XIVE interrupts including in OPAL XICS compatibility
+	 * are always delivered as HVI interrupts rather than EE, so
+	 * LPES0 can be set here.
+	 */
+
+	lpcr = mfspr(SPRN_LPCR);
+	lpcr |= LPCR_HVICE;	/* enable hvi interrupts */
+	lpcr |= LPCR_PECE_HVEE; /* hvi can wake from stop */
+#if 0
+XXX: XIVE topic branch adds HEIC (and LPES0 but that may be reverted)
+	lpcr |= LPCR_HEIC;	/* disable external interrupts in hypervisor */
+#endif
+	mtspr(SPRN_LPCR, lpcr);
+
+	return 1;
+}
+
+static int __init feat_enable_large_ci(struct dt_cpu_feature *f)
+{
+	cur_cpu_spec->mmu_features |= MMU_FTR_CI_LARGE_PAGE;
+
+	return 1;
+}
+
+struct dt_cpu_feature_match {
+	const char *name;
+	int (*enable)(struct dt_cpu_feature *f);
+	u64 cpu_ftr_bit_mask;
+};
+
+static struct dt_cpu_feature_match __initdata
+		dt_cpu_feature_match_table[] = {
+	{"hypervisor", feat_enable_hv, 0},
+	{"big-endian", feat_enable, 0},
+	{"little-endian", feat_enable_le, CPU_FTR_REAL_LE},
+	{"smt", feat_enable_smt, 0},
+	{"interrupt-facilities", feat_enable, 0},
+	{"timer-facilities", feat_enable, 0},
+	{"timer-facilities-v3", feat_enable, 0},
+	{"debug-facilities", feat_enable, 0},
+	{"come-from-address-register", feat_enable, CPU_FTR_CFAR},
+	{"branch-tracing", feat_enable, 0},
+	{"floating-point", feat_enable_fp, 0},
+	{"vector", feat_enable_vector, 0},
+	{"vector-scalar", feat_enable_vsx, 0},
+	{"vector-scalar-v3", feat_enable, 0},
+	{"decimal-floating-point", feat_enable, 0},
+	{"decimal-integer", feat_enable, 0},
+	{"vector-crypto", feat_enable, 0},
+	{"mmu-hash", feat_enable_mmu_hash, 0},
+	{"mmu-radix", feat_enable_mmu_radix, 0},
+	{"mmu-hash-v3", feat_enable_mmu_hash_v3, 0},
+	{"virtual-page-class-key-protection", feat_enable, 0},
+	{"transactional-memory", feat_enable_tm, CPU_FTR_TM},
+	{"transactional-memory-v3", feat_enable_tm, 0},
+	{"idle-nap", feat_enable_idle_nap, 0},
+	{"alignment-interrupt-dsisr", feat_enable_align_dsisr, 0},
+	{"idle-stop", feat_enable_idle_stop, 0},
+	{"machine-check-power8", feat_enable_mce_power8, 0},
+	{"performance-monitor-power8", feat_enable_pmu_power8, 0},
+	{"data-stream-control-register", feat_enable_dscr, CPU_FTR_DSCR},
+	{"event-based-branch", feat_enable, 0},
+	{"target-address-register", feat_enable, 0},
+	{"branch-history-rolling-buffer", feat_enable, 0},
+	{"control-register", feat_enable, CPU_FTR_CTRL},
+	{"processor-control-facility", feat_enable_dbell, CPU_FTR_DBELL},
+	{"processor-control-facility-v3", feat_enable_dbell, CPU_FTR_DBELL},
+	{"processor-utilization-of-resources-register", feat_enable_purr, 0},
+	{"subcore", feat_enable, CPU_FTR_SUBCORE},
+	{"no-execute", feat_enable, 0},
+	{"strong-access-ordering", feat_enable, CPU_FTR_SAO},
+	{"cache-inhibited-large-page", feat_enable_large_ci, 0},
+	{"coprocessor-icswx", feat_enable, CPU_FTR_ICSWX},
+	{"hypervisor-virtualization-interrupt", feat_enable_hvi, 0},
+	{"program-priority-register", feat_enable, CPU_FTR_HAS_PPR},
+	{"wait", feat_enable, 0},
+	{"atomic-memory-operations", feat_enable, 0},
+	{"branch-v3", feat_enable, 0},
+	{"copy-paste", feat_enable, 0},
+	{"decimal-floating-point-v3", feat_enable, 0},
+	{"decimal-integer-v3", feat_enable, 0},
+	{"fixed-point-v3", feat_enable, 0},
+	{"floating-point-v3", feat_enable, 0},
+	{"group-start-register", feat_enable, 0},
+	{"pc-relative-addressing", feat_enable, 0},
+	{"machine-check-power9", feat_enable_mce_power9, 0},
+	{"performance-monitor-power9", feat_enable_pmu_power9, 0},
+	{"event-based-branch-v3", feat_enable, 0},
+	{"random-number-generator", feat_enable, 0},
+	{"system-call-vectored", feat_disable, 0},
+	{"trace-interrupt-v3", feat_enable, 0},
+	{"vector-v3", feat_enable, 0},
+	{"vector-binary128", feat_enable, 0},
+	{"vector-binary16", feat_enable, 0},
+	{"wait-v3", feat_enable, 0},
+};
+
+/* XXX: how to configure this? Default + boot time? */
+#ifdef CONFIG_PPC_CPUFEATURES_ENABLE_UNKNOWN
+#define CPU_FEATURE_ENABLE_UNKNOWN 1
+#else
+#define CPU_FEATURE_ENABLE_UNKNOWN 0
+#endif
+
+void __init cpufeatures_setup_start(u32 isa)
+{
+	DBG("CPUFEATURES setup for isa %d\n", isa);
+
+	if (isa >= 3000) {
+		cur_cpu_spec->cpu_features |= CPU_FTR_ARCH_300;
+		cur_cpu_spec->cpu_user_features2 |= PPC_FEATURE2_ARCH_3_00;
+	}
+}
+
+int __init cpufeatures_process_feature(struct dt_cpu_feature *f)
+{
+	const struct dt_cpu_feature_match *m;
+	int known = 0;
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(dt_cpu_feature_match_table); i++) {
+		m = &dt_cpu_feature_match_table[i];
+		if (!strcmp(f->name, m->name)) {
+			known = 1;
+			if (m->enable(f))
+				goto enabled;
+			goto not_enabled;
+		}
+	}
+
+	if (CPU_FEATURE_ENABLE_UNKNOWN) {
+		if (feat_try_enable_unknown(f))
+			goto enabled;
+	}
+
+not_enabled:
+	if (known)
+		DBG("CPU feature not enabling:%s (disabled or unsupported by kernel)\n", f->name);
+	else
+		DBG("CPU feature not enabling:%s (unknown and unsupported by kernel)\n", f->name);
+
+	return 0;
+
+enabled:
+	if (m->cpu_ftr_bit_mask)
+		cur_cpu_spec->cpu_features |= m->cpu_ftr_bit_mask;
+	if (known)
+		DBG("CPU feature enabling:%s\n", f->name);
+	else
+		DBG("CPU feature enabling:%s (unknown)\n", f->name);
+
+	return 1;
+}
+
+void __init cpufeatures_setup_finished(void)
+{
+	if (hv_mode && !(cur_cpu_spec->cpu_features & CPU_FTR_HVMODE)) {
+		pr_err("CPU feature hypervisor not present in device tree but HV mode is enabled in the CPU. Enabling.\n");
+		cur_cpu_spec->cpu_features |= CPU_FTR_HVMODE;
+	}
+
+	system_registers.hfscr = mfspr(SPRN_HFSCR);
+	system_registers.fscr = mfspr(SPRN_FSCR);
+	system_registers.lpcr = mfspr(SPRN_LPCR);
+}
diff --git a/arch/powerpc/kernel/cputable.c b/arch/powerpc/kernel/cputable.c
index e79b9daa873c..9b3e88b1a9c8 100644
--- a/arch/powerpc/kernel/cputable.c
+++ b/arch/powerpc/kernel/cputable.c
@@ -23,7 +23,9 @@
 #include <asm/mmu.h>
 #include <asm/setup.h>
 
-struct cpu_spec* cur_cpu_spec = NULL;
+static struct cpu_spec the_cpu_spec __read_mostly;
+
+struct cpu_spec* cur_cpu_spec __read_mostly = NULL;
 EXPORT_SYMBOL(cur_cpu_spec);
 
 /* The platform string corresponding to the real PVR */
@@ -2179,7 +2181,15 @@ static struct cpu_spec __initdata cpu_specs[] = {
 #endif /* CONFIG_E500 */
 };
 
-static struct cpu_spec the_cpu_spec;
+void __init set_cur_cpu_spec(struct cpu_spec *s)
+{
+	struct cpu_spec *t = &the_cpu_spec;
+
+	t = PTRRELOC(t);
+	*t = *s;
+
+	*PTRRELOC(&cur_cpu_spec) = &the_cpu_spec;
+}
 
 static struct cpu_spec * __init setup_cpu_spec(unsigned long offset,
 					       struct cpu_spec *s)
@@ -2266,6 +2276,29 @@ struct cpu_spec * __init identify_cpu(unsigned long offset, unsigned int pvr)
 	return NULL;
 }
 
+/*
+ * Used by cpufeatures to get the name for CPUs with a PVR table.
+ * If they don't hae a PVR table, cpufeatures gets the name from
+ * cpu device-tree node.
+ */
+void __init identify_cpu_name(unsigned int pvr)
+{
+	struct cpu_spec *s = cpu_specs;
+	struct cpu_spec *t = &the_cpu_spec;
+	int i;
+
+	s = PTRRELOC(s);
+	t = PTRRELOC(t);
+
+	for (i = 0; i < ARRAY_SIZE(cpu_specs); i++,s++) {
+		if ((pvr & s->pvr_mask) == s->pvr_value) {
+			t->cpu_name = s->cpu_name;
+			return;
+		}
+	}
+}
+
+
 #ifdef CONFIG_JUMP_LABEL_FEATURE_CHECKS
 struct static_key_true cpu_feature_keys[NUM_CPU_FTR_KEYS] = {
 			[0 ... NUM_CPU_FTR_KEYS - 1] = STATIC_KEY_TRUE_INIT
diff --git a/arch/powerpc/kernel/prom.c b/arch/powerpc/kernel/prom.c
index f5d399e46193..be8f8357a890 100644
--- a/arch/powerpc/kernel/prom.c
+++ b/arch/powerpc/kernel/prom.c
@@ -58,6 +58,7 @@
 #include <asm/debug.h>
 #include <asm/epapr_hcalls.h>
 #include <asm/firmware.h>
+#include <asm/cpufeatures.h>
 
 #include <mm/mmu_decl.h>
 
@@ -67,6 +68,12 @@
 #define DBG(fmt...)
 #endif
 
+#ifdef CONFIG_PPC_CPUFEATURES_DT
+int __initdata has_cpufeatures_node = 0;
+#else
+static const int has_cpufeatures_node = 0;
+#endif
+
 #ifdef CONFIG_PPC64
 int __initdata iommu_is_off;
 int __initdata iommu_force_on;
@@ -300,6 +307,268 @@ static void __init check_cpu_feature_properties(unsigned long node)
 	}
 }
 
+#ifdef CONFIG_PPC_CPUFEATURES_DT
+#define CPUFT_WARN(str, name) printk(KERN_WARNING "WARNING: /cpus/features/%s dt:%s\n", name, str)
+
+static int nr_dt_cpu_features;
+static struct dt_cpu_feature *dt_cpu_features;
+
+static int __init process_cpufeatures_node(unsigned long node,
+					  const char *uname, int i)
+{
+	const __be32 *prop;
+	struct dt_cpu_feature *f;
+	int len;
+
+	f = &dt_cpu_features[i];
+	memset(f, 0, sizeof(struct dt_cpu_feature));
+
+	f->node = node;
+
+	f->name = uname;
+
+	prop = of_get_flat_dt_prop(node, "isa", &len);
+	if (!prop) {
+		CPUFT_WARN("missing isa property", uname);
+		return 0;
+	}
+	f->isa = be32_to_cpup(prop);
+
+	prop = of_get_flat_dt_prop(node, "usable-mask", &len);
+	if (!prop) {
+		CPUFT_WARN("missing usable-mask property", uname);
+		return 0;
+	}
+	f->usable_mask = be32_to_cpup(prop);
+
+	prop = of_get_flat_dt_prop(node, "hv-support", &len);
+	if (prop)
+		f->hv_support = be32_to_cpup(prop);
+	prop = of_get_flat_dt_prop(node, "os-support", &len);
+	if (prop)
+		f->os_support = be32_to_cpup(prop);
+
+	prop = of_get_flat_dt_prop(node, "hfscr-bit-nr", &len);
+	if (prop)
+		f->hfscr_bit_nr = be32_to_cpup(prop);
+	else
+		f->hfscr_bit_nr = -1;
+	prop = of_get_flat_dt_prop(node, "fscr-bit-nr", &len);
+	if (prop)
+		f->fscr_bit_nr = be32_to_cpup(prop);
+	else
+		f->fscr_bit_nr = -1;
+	prop = of_get_flat_dt_prop(node, "hwcap-bit-nr", &len);
+	if (prop)
+		f->hwcap_bit_nr = be32_to_cpup(prop);
+	else
+		f->hwcap_bit_nr = -1;
+
+	if (f->usable_mask & USABLE_HV) {
+		if (!(mfmsr() & MSR_HV)) {
+			CPUFT_WARN("HV feature passed to guest\n", uname);
+			return 0;
+		}
+
+		if (!f->hv_support && f->hfscr_bit_nr != -1) {
+			CPUFT_WARN("unwanted hfscr_bit_nr\n", uname);
+			return 0;
+		}
+
+		if (f->hv_support == HV_SUPPORT_HFSCR) {
+			if (f->hfscr_bit_nr == -1) {
+				CPUFT_WARN("missing hfscr_bit_nr\n", uname);
+				return 0;
+			}
+		}
+	} else {
+		if (f->hv_support != HV_SUPPORT_NONE || f->hfscr_bit_nr != -1) {
+			CPUFT_WARN("unwanted hv_support/hfscr_bit_nr\n", uname);
+			return 0;
+		}
+	}
+
+	if (f->usable_mask & USABLE_OS) {
+		if (!f->os_support && f->fscr_bit_nr != -1) {
+			CPUFT_WARN("unwanted fscr_bit_nr\n", uname);
+			return 0;
+		}
+
+		if (f->os_support == OS_SUPPORT_FSCR) {
+			if (f->fscr_bit_nr == -1) {
+				CPUFT_WARN("missing fscr_bit_nr\n", uname);
+				return 0;
+			}
+		}
+	} else {
+		if (f->os_support != OS_SUPPORT_NONE || f->fscr_bit_nr != -1) {
+			CPUFT_WARN("unwanted os_support/fscr_bit_nr\n", uname);
+			return 0;
+		}
+	}
+
+	if (!(f->usable_mask & USABLE_PR)) {
+		if (f->hwcap_bit_nr != -1) {
+			CPUFT_WARN("unwanted hwcap_bit_nr\n", uname);
+			return 0;
+		}
+	}
+
+	/* Do all the independent features in the first pass */
+	if (!of_get_flat_dt_prop(node, "dependencies", &len)) {
+		if (cpufeatures_process_feature(f))
+			f->enabled = 1;
+		else
+			f->disabled = 1;
+	}
+
+	return 0;
+}
+
+static void __init cpufeatures_deps_enable(struct dt_cpu_feature *f)
+{
+	const __be32 *prop;
+	int len;
+	int nr_deps;
+	int i;
+
+	if (f->enabled || f->disabled)
+		return;
+
+	prop = of_get_flat_dt_prop(f->node, "dependencies", &len);
+	if (!prop) {
+		CPUFT_WARN("missing dependencies property", f->name);
+		return;
+	}
+
+	nr_deps = len / sizeof(int);
+
+	for (i = 0; i < nr_deps; i++) {
+		unsigned long phandle = be32_to_cpu(prop[i]);
+		int j;
+
+		for (j = 0; j < nr_dt_cpu_features; j++) {
+			struct dt_cpu_feature *d = &dt_cpu_features[j];
+
+			if (of_get_flat_dt_phandle(d->node) == phandle) {
+				cpufeatures_deps_enable(d);
+				if (d->disabled) {
+					f->disabled = 1;
+					return;
+				}
+			}
+		}
+	}
+
+	if (cpufeatures_process_feature(f))
+		f->enabled = 1;
+	else
+		f->disabled = 1;
+}
+
+static int __init scan_cpufeatures_subnodes(unsigned long node,
+					  const char *uname,
+					  void *data)
+{
+	int *count = data;
+
+	process_cpufeatures_node(node, uname, *count);
+
+	(*count)++;
+
+	return 0;
+}
+
+static int __init count_cpufeatures_subnodes(unsigned long node,
+					  const char *uname,
+					  void *data)
+{
+	int *count = data;
+
+	(*count)++;
+
+	return 0;
+}
+
+static int __init early_init_dt_scan_cpufeatures(unsigned long node,
+					  const char *uname, int depth,
+					  void *data)
+{
+	const char *type = of_get_flat_dt_prop(node, "device_type", NULL);
+	const __be32 *prop;
+	int count, i;
+	u32 isa;
+
+	/* We are scanning "features" nodes only */
+	if (type == NULL || strcmp(type, "cpu-features") != 0)
+		return 0;
+
+	prop = of_get_flat_dt_prop(node, "isa", NULL);
+	if (!prop) {
+		printk("cpu-features node has missing property \"isa\"\n");
+		return 0;
+	}
+
+	isa = be32_to_cpup(prop);
+
+	has_cpufeatures_node = 1;
+
+	/* Count and allocate space for cpu features */
+	of_scan_flat_dt_subnodes(node, count_cpufeatures_subnodes,
+						&nr_dt_cpu_features);
+	dt_cpu_features = __va(
+		memblock_alloc(sizeof(struct dt_cpu_feature)*
+				nr_dt_cpu_features, PAGE_SIZE));
+
+	cpufeatures_setup_start(isa);
+
+	/* Scan nodes into dt_cpu_features and enable those without deps  */
+	count = 0;
+	of_scan_flat_dt_subnodes(node, scan_cpufeatures_subnodes, &count);
+
+	/* Recursive enable remaining features with dependencies */
+	for (i = 0; i < nr_dt_cpu_features; i++) {
+		struct dt_cpu_feature *f = &dt_cpu_features[i];
+
+		cpufeatures_deps_enable(f);
+	}
+
+	cpufeatures_setup_finished();
+
+	memblock_free(__pa(dt_cpu_features),
+			sizeof(struct dt_cpu_feature)*nr_dt_cpu_features);
+
+	return 0;
+}
+
+static int __init early_init_dt_scan_cpufeatures_exists(unsigned long node,
+					  const char *uname, int depth,
+					  void *data)
+{
+	const char *type = of_get_flat_dt_prop(node, "device_type", NULL);
+	int *exists = data;
+
+	/* We are scanning "features" nodes only */
+	if (type == NULL || strcmp(type, "cpu-features") != 0)
+		return 0;
+
+	*exists = 1;
+
+	return 0;
+}
+
+/*
+ * Does the /cpus/features/ node exist?
+ */
+int __init early_init_devtree_check_cpu_features_exists(void)
+{
+	int exists = 0;
+	of_scan_flat_dt(early_init_dt_scan_cpufeatures_exists, &exists);
+	return exists;
+}
+
+#endif /* CONFIG_PPC_CPUFEATURES_DT */
+
 static int __init early_init_dt_scan_cpus(unsigned long node,
 					  const char *uname, int depth,
 					  void *data)
@@ -376,23 +645,61 @@ static int __init early_init_dt_scan_cpus(unsigned long node,
 	 * A POWER6 partition in "POWER6 architected" mode
 	 * uses the 0x0f000002 PVR value; in POWER5+ mode
 	 * it uses 0x0f000001.
+	 *
+	 * In the case of the cpufeatures node existing, it is the
+	 * responsibility of the firmware/hypervisor to provide the
+	 * correct feature set for the architecture level.
 	 */
-	prop = of_get_flat_dt_prop(node, "cpu-version", NULL);
-	if (prop && (be32_to_cpup(prop) & 0xff000000) == 0x0f000000)
-		identify_cpu(0, be32_to_cpup(prop));
+	if (!has_cpufeatures_node) {
+		prop = of_get_flat_dt_prop(node, "cpu-version", NULL);
+		if (prop && (be32_to_cpup(prop) & 0xff000000) == 0x0f000000)
+			identify_cpu(0, be32_to_cpup(prop));
+		identical_pvr_fixup(node);
 
-	identical_pvr_fixup(node);
+		check_cpu_feature_properties(node);
+		check_cpu_pa_features(node);
 
-	check_cpu_feature_properties(node);
-	check_cpu_pa_features(node);
-	init_mmu_slb_size(node);
+#ifdef CONFIG_PPC64
+		if (nthreads > 1)
+			cur_cpu_spec->cpu_features |= CPU_FTR_SMT;
+		else
+			cur_cpu_spec->cpu_features &= ~CPU_FTR_SMT;
+#endif
+	} else {
+		unsigned int pvr;
+
+		/*
+		 * Fill cpu_name with PVR table if it exists, for backward
+		 * compatibility.
+		 */
+		pvr = cur_cpu_spec->pvr_value;
+		prop = of_get_flat_dt_prop(node, "cpu-version", NULL);
+		if (prop && (be32_to_cpup(prop) & 0xff000000) == 0x0f000000)
+			pvr = be32_to_cpup(prop);
+		identify_cpu_name(pvr);
+
+		/* Otherwise use cpu node dt */
+		if (!cur_cpu_spec->cpu_name) {
+			static char cpu_name[32];
+			char *c;
+
+			/* Get cpu_name from device-tree cpu node */
+			strncpy(cpu_name, uname, 32);
+			c = strchr(cpu_name, '@');
+			if (c)
+				*c = '\0';
+			cur_cpu_spec->cpu_name = cpu_name;
+		}
 
 #ifdef CONFIG_PPC64
-	if (nthreads > 1)
-		cur_cpu_spec->cpu_features |= CPU_FTR_SMT;
-	else
-		cur_cpu_spec->cpu_features &= ~CPU_FTR_SMT;
+		if (nthreads == 1)
+			cur_cpu_spec->cpu_features &= ~CPU_FTR_SMT;
 #endif
+	}
+
+
+	init_mmu_slb_size(node);
+
 	return 0;
 }
 
@@ -722,6 +1029,10 @@ void __init early_init_devtree(void *params)
 
 	DBG("Scanning CPUs ...\n");
 
+#ifdef CONFIG_PPC_CPUFEATURES_DT
+	of_scan_flat_dt(early_init_dt_scan_cpufeatures, NULL);
+#endif
+
 	/* Retrieve CPU related informations from the flat tree
 	 * (altivec support, boot CPU ID, ...)
 	 */
diff --git a/arch/powerpc/kernel/setup-common.c b/arch/powerpc/kernel/setup-common.c
index 4697da895133..fe52229681da 100644
--- a/arch/powerpc/kernel/setup-common.c
+++ b/arch/powerpc/kernel/setup-common.c
@@ -256,7 +256,7 @@ static int show_cpuinfo(struct seq_file *m, void *v)
 	seq_printf(m, "processor\t: %lu\n", cpu_id);
 	seq_printf(m, "cpu\t\t: ");
 
-	if (cur_cpu_spec->pvr_mask)
+	if (cur_cpu_spec->pvr_mask && cur_cpu_spec->cpu_name)
 		seq_printf(m, "%s", cur_cpu_spec->cpu_name);
 	else
 		seq_printf(m, "unknown (%08x)", pvr);
diff --git a/arch/powerpc/kernel/setup_64.c b/arch/powerpc/kernel/setup_64.c
index f997154dfc41..f3e65de63ce8 100644
--- a/arch/powerpc/kernel/setup_64.c
+++ b/arch/powerpc/kernel/setup_64.c
@@ -49,6 +49,7 @@
 #include <asm/paca.h>
 #include <asm/time.h>
 #include <asm/cputable.h>
+#include <asm/cpufeatures.h>
 #include <asm/sections.h>
 #include <asm/btext.h>
 #include <asm/nvram.h>
@@ -274,8 +275,17 @@ void __init early_setup(unsigned long dt_ptr)
 
 	/* -------- printk is _NOT_ safe to use here ! ------- */
 
-	/* Identify CPU type */
+#ifdef CONFIG_PPC_CPUFEATURES_DT
+	/* Setup flat device-tree pointer */
+	initial_boot_params = __va(dt_ptr);
+	if (early_init_devtree_check_cpu_features_exists())
+		cpufeatures_setup_cpu();
+	else
+		/* Legacy table-based approach when /cpus/features is missing */
+		identify_cpu(0, mfspr(SPRN_PVR));
+#else
 	identify_cpu(0, mfspr(SPRN_PVR));
+#endif
 
 	/* Assume we're on cpu 0 for now. Don't write to the paca yet! */
 	initialise_paca(&boot_paca, 0);
@@ -541,6 +551,9 @@ void __init initialize_cache_info(void)
 	dcache_bsize = ppc64_caches.l1d.block_size;
 	icache_bsize = ppc64_caches.l1i.block_size;
 
+	cur_cpu_spec->dcache_bsize = dcache_bsize;
+	cur_cpu_spec->icache_bsize = icache_bsize;
+
 	DBG(" <- initialize_cache_info()\n");
 }
 
-- 
2.11.0



More information about the Linuxppc-dev mailing list