[PATCH v3 0/5]

[PATCH v3 0/5]

[Marc Zyngier]

PMUv3 has been introduced with ARMv8 and, while it has only been used
on 64bit systems so far, it would definitely be useful for 32bit
guests running under KVM/arm64, for example.

There is also the case of people natively running 32bit kernels on
64bit HW and trying to upstream unspeakable hacks, hoping that the
stars will align and that they'll win the lottery (see [1]).

So let's try again, and make the PMUv3 driver usable for everyone.

This is done in three steps:
(1) Move the driver from arch/arm64 to drivers/perf
(2) Add a handful of system register accessors so that we can reuse
    the driver on 32bit
(3) Provide the same accessors on 32bit, enable compilation, and
    make it the default selection for mach-virt.

Tested on a Seattle box with 32bit guests.

* From v1:
  - Fixed encodings for some CP15 accessors
  - Added a terse note saying that CPU_V7 also covers ARMv8
  - Rebased on v4.12-rc5

* From v2:
  - SPDX tags on new and moved files. Yeah!
  - Annual rebase on 4.17-rc5

[1] https://patchwork.kernel.org/patch/10406793/

Marc Zyngier (5):
  arm64: perf: Move PMUv3 driver to drivers/perf
  arm64: perf: Abstract system register accesses away
  ARM: Make CONFIG_CPU_V7 valid for 32bit ARMv8 implementations
  ARM: perf: Allow the use of the PMUv3 driver on 32bit ARM
  ARM: mach-virt: Select PMUv3 driver by default

 arch/arm/Kconfig                                   |   1 +
 arch/arm/include/asm/arm_pmuv3.h                   | 125 +++++++++++++++++++++
 arch/arm/mm/Kconfig                                |   2 +-
 arch/arm64/include/asm/arm_pmuv3.h                 | 111 ++++++++++++++++++
 arch/arm64/include/asm/perf_event.h                |  55 ---------
 arch/arm64/kernel/Makefile                         |   1 -
 drivers/perf/Kconfig                               |   8 ++
 drivers/perf/Makefile                              |   1 +
 .../perf_event.c => drivers/perf/arm_pmuv3.c       |  42 ++++---
 include/kvm/arm_pmu.h                              |   2 +-
 include/linux/perf/arm_pmuv3.h                     |  78 +++++++++++++
 11 files changed, 346 insertions(+), 80 deletions(-)
 create mode 100644 arch/arm/include/asm/arm_pmuv3.h
 create mode 100644 arch/arm64/include/asm/arm_pmuv3.h
 rename arch/arm64/kernel/perf_event.c => drivers/perf/arm_pmuv3.c (97%)
 create mode 100644 include/linux/perf/arm_pmuv3.h

-- 
2.14.2

[PATCH v3 0/5]

[Marc Zyngier]

PMUv3 has been introduced with ARMv8 and, while it has only been used
on 64bit systems so far, it would definitely be useful for 32bit
guests running under KVM/arm64, for example.

There is also the case of people natively running 32bit kernels on
64bit HW and trying to upstream unspeakable hacks, hoping that the
stars will align and that they'll win the lottery (see [1]).

So let's try again, and make the PMUv3 driver usable for everyone.

This is done in three steps:
(1) Move the driver from arch/arm64 to drivers/perf
(2) Add a handful of system register accessors so that we can reuse
    the driver on 32bit
(3) Provide the same accessors on 32bit, enable compilation, and
    make it the default selection for mach-virt.

Tested on a Seattle box with 32bit guests.

* From v1:
  - Fixed encodings for some CP15 accessors
  - Added a terse note saying that CPU_V7 also covers ARMv8
  - Rebased on v4.12-rc5

* From v2:
  - SPDX tags on new and moved files. Yeah!
  - Annual rebase on 4.17-rc5

[1] https://patchwork.kernel.org/patch/10406793/

Marc Zyngier (5):
  arm64: perf: Move PMUv3 driver to drivers/perf
  arm64: perf: Abstract system register accesses away
  ARM: Make CONFIG_CPU_V7 valid for 32bit ARMv8 implementations
  ARM: perf: Allow the use of the PMUv3 driver on 32bit ARM
  ARM: mach-virt: Select PMUv3 driver by default

 arch/arm/Kconfig                                   |   1 +
 arch/arm/include/asm/arm_pmuv3.h                   | 125 +++++++++++++++++++++
 arch/arm/mm/Kconfig                                |   2 +-
 arch/arm64/include/asm/arm_pmuv3.h                 | 111 ++++++++++++++++++
 arch/arm64/include/asm/perf_event.h                |  55 ---------
 arch/arm64/kernel/Makefile                         |   1 -
 drivers/perf/Kconfig                               |   8 ++
 drivers/perf/Makefile                              |   1 +
 .../perf_event.c => drivers/perf/arm_pmuv3.c       |  42 ++++---
 include/kvm/arm_pmu.h                              |   2 +-
 include/linux/perf/arm_pmuv3.h                     |  78 +++++++++++++
 11 files changed, 346 insertions(+), 80 deletions(-)
 create mode 100644 arch/arm/include/asm/arm_pmuv3.h
 create mode 100644 arch/arm64/include/asm/arm_pmuv3.h
 rename arch/arm64/kernel/perf_event.c => drivers/perf/arm_pmuv3.c (97%)
 create mode 100644 include/linux/perf/arm_pmuv3.h

-- 
2.14.2

[PATCH v3 1/5] arm64: perf: Move PMUv3 driver to drivers/perf

[Marc Zyngier]

Having the ARM PMUv3 driver sitting in arch/arm64/kernel is getting
in the way of being able to use perf on ARMv8 cores running a 32bit
kernel, such as 32bit KVM guests.

This patch moves it into drivers/perf/arm_pmuv3.c, with an include
file in include/linux/perf/arm_pmuv3.h. The only thing left in
arch/arm64 is some mundane perf stuff.

Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
---
 arch/arm64/include/asm/perf_event.h                | 55 ----------------
 arch/arm64/kernel/Makefile                         |  1 -
 drivers/perf/Kconfig                               |  8 +++
 drivers/perf/Makefile                              |  1 +
 .../perf_event.c => drivers/perf/arm_pmuv3.c       |  2 +
 include/kvm/arm_pmu.h                              |  2 +-
 include/linux/perf/arm_pmuv3.h                     | 76 ++++++++++++++++++++++
 7 files changed, 88 insertions(+), 57 deletions(-)
 rename arch/arm64/kernel/perf_event.c => drivers/perf/arm_pmuv3.c (99%)
 create mode 100644 include/linux/perf/arm_pmuv3.h

diff --git a/arch/arm64/include/asm/perf_event.h b/arch/arm64/include/asm/perf_event.h
index f9ccc36d3dc3..5b33efeebabf 100644
--- a/arch/arm64/include/asm/perf_event.h
+++ b/arch/arm64/include/asm/perf_event.h
@@ -20,61 +20,6 @@
 #include <asm/stack_pointer.h>
 #include <asm/ptrace.h>
 
-#define	ARMV8_PMU_MAX_COUNTERS	32
-#define	ARMV8_PMU_COUNTER_MASK	(ARMV8_PMU_MAX_COUNTERS - 1)
-
-/*
- * Per-CPU PMCR: config reg
- */
-#define ARMV8_PMU_PMCR_E	(1 << 0) /* Enable all counters */
-#define ARMV8_PMU_PMCR_P	(1 << 1) /* Reset all counters */
-#define ARMV8_PMU_PMCR_C	(1 << 2) /* Cycle counter reset */
-#define ARMV8_PMU_PMCR_D	(1 << 3) /* CCNT counts every 64th cpu cycle */
-#define ARMV8_PMU_PMCR_X	(1 << 4) /* Export to ETM */
-#define ARMV8_PMU_PMCR_DP	(1 << 5) /* Disable CCNT if non-invasive debug*/
-#define ARMV8_PMU_PMCR_LC	(1 << 6) /* Overflow on 64 bit cycle counter */
-#define	ARMV8_PMU_PMCR_N_SHIFT	11	 /* Number of counters supported */
-#define	ARMV8_PMU_PMCR_N_MASK	0x1f
-#define	ARMV8_PMU_PMCR_MASK	0x7f	 /* Mask for writable bits */
-
-/*
- * PMOVSR: counters overflow flag status reg
- */
-#define	ARMV8_PMU_OVSR_MASK		0xffffffff	/* Mask for writable bits */
-#define	ARMV8_PMU_OVERFLOWED_MASK	ARMV8_PMU_OVSR_MASK
-
-/*
- * PMXEVTYPER: Event selection reg
- */
-#define	ARMV8_PMU_EVTYPE_MASK	0xc800ffff	/* Mask for writable bits */
-#define	ARMV8_PMU_EVTYPE_EVENT	0xffff		/* Mask for EVENT bits */
-
-/*
- * PMUv3 event types: required events
- */
-#define ARMV8_PMUV3_PERFCTR_SW_INCR				0x00
-#define ARMV8_PMUV3_PERFCTR_L1D_CACHE_REFILL			0x03
-#define ARMV8_PMUV3_PERFCTR_L1D_CACHE				0x04
-#define ARMV8_PMUV3_PERFCTR_BR_MIS_PRED				0x10
-#define ARMV8_PMUV3_PERFCTR_CPU_CYCLES				0x11
-#define ARMV8_PMUV3_PERFCTR_BR_PRED				0x12
-
-/*
- * Event filters for PMUv3
- */
-#define	ARMV8_PMU_EXCLUDE_EL1	(1 << 31)
-#define	ARMV8_PMU_EXCLUDE_EL0	(1 << 30)
-#define	ARMV8_PMU_INCLUDE_EL2	(1 << 27)
-
-/*
- * PMUSERENR: user enable reg
- */
-#define ARMV8_PMU_USERENR_MASK	0xf		/* Mask for writable bits */
-#define ARMV8_PMU_USERENR_EN	(1 << 0) /* PMU regs can be accessed at EL0 */
-#define ARMV8_PMU_USERENR_SW	(1 << 1) /* PMSWINC can be written at EL0 */
-#define ARMV8_PMU_USERENR_CR	(1 << 2) /* Cycle counter can be read at EL0 */
-#define ARMV8_PMU_USERENR_ER	(1 << 3) /* Event counter can be read at EL0 */
-
 #ifdef CONFIG_PERF_EVENTS
 struct pt_regs;
 extern unsigned long perf_instruction_pointer(struct pt_regs *regs);
diff --git a/arch/arm64/kernel/Makefile b/arch/arm64/kernel/Makefile
index bf825f38d206..aba9344a72ca 100644
--- a/arch/arm64/kernel/Makefile
+++ b/arch/arm64/kernel/Makefile
@@ -32,7 +32,6 @@ arm64-obj-$(CONFIG_FUNCTION_TRACER)	+= ftrace.o entry-ftrace.o
 arm64-obj-$(CONFIG_MODULES)		+= arm64ksyms.o module.o
 arm64-obj-$(CONFIG_ARM64_MODULE_PLTS)	+= module-plts.o
 arm64-obj-$(CONFIG_PERF_EVENTS)		+= perf_regs.o perf_callchain.o
-arm64-obj-$(CONFIG_HW_PERF_EVENTS)	+= perf_event.o
 arm64-obj-$(CONFIG_HAVE_HW_BREAKPOINT)	+= hw_breakpoint.o
 arm64-obj-$(CONFIG_CPU_PM)		+= sleep.o suspend.o
 arm64-obj-$(CONFIG_CPU_IDLE)		+= cpuidle.o
diff --git a/drivers/perf/Kconfig b/drivers/perf/Kconfig
index 28bb5a029558..39808b86b346 100644
--- a/drivers/perf/Kconfig
+++ b/drivers/perf/Kconfig
@@ -50,6 +50,14 @@ config ARM_PMU_ACPI
 	depends on ARM_PMU && ACPI
 	def_bool y
 
+config ARM_PMUV3
+	depends on HW_PERF_EVENTS && ARM64
+	bool "ARM PMUv3 support" if !ARM64
+	default y
+	help
+	  Say y if you want to use CPU performance monitors on ARMv8
+	  systems that implement the PMUv3 architecture.
+
 config ARM_DSU_PMU
 	tristate "ARM DynamIQ Shared Unit (DSU) PMU"
 	depends on ARM64
diff --git a/drivers/perf/Makefile b/drivers/perf/Makefile
index b3902bd37d53..a1a2f64e0c8f 100644
--- a/drivers/perf/Makefile
+++ b/drivers/perf/Makefile
@@ -4,6 +4,7 @@ obj-$(CONFIG_ARM_CCN) += arm-ccn.o
 obj-$(CONFIG_ARM_DSU_PMU) += arm_dsu_pmu.o
 obj-$(CONFIG_ARM_PMU) += arm_pmu.o arm_pmu_platform.o
 obj-$(CONFIG_ARM_PMU_ACPI) += arm_pmu_acpi.o
+obj-$(CONFIG_ARM_PMUV3) += arm_pmuv3.o
 obj-$(CONFIG_HISI_PMU) += hisilicon/
 obj-$(CONFIG_QCOM_L2_PMU)	+= qcom_l2_pmu.o
 obj-$(CONFIG_QCOM_L3_PMU) += qcom_l3_pmu.o
diff --git a/arch/arm64/kernel/perf_event.c b/drivers/perf/arm_pmuv3.c
similarity index 99%
rename from arch/arm64/kernel/perf_event.c
rename to drivers/perf/arm_pmuv3.c
index 85a251b6dfa8..bd19b16c44eb 100644
--- a/arch/arm64/kernel/perf_event.c
+++ b/drivers/perf/arm_pmuv3.c
@@ -1,3 +1,4 @@
+// SPDX-License-Identifier: GPL-2.0
 /*
  * PMU support
  *
@@ -27,6 +28,7 @@
 #include <linux/acpi.h>
 #include <linux/of.h>
 #include <linux/perf/arm_pmu.h>
+#include <linux/perf/arm_pmuv3.h>
 #include <linux/platform_device.h>
 
 /*
diff --git a/include/kvm/arm_pmu.h b/include/kvm/arm_pmu.h
index f87fe20fcb05..d16ce92cb2c0 100644
--- a/include/kvm/arm_pmu.h
+++ b/include/kvm/arm_pmu.h
@@ -19,7 +19,7 @@
 #define __ASM_ARM_KVM_PMU_H
 
 #include <linux/perf_event.h>
-#include <asm/perf_event.h>
+#include <linux/perf/arm_pmuv3.h>
 
 #define ARMV8_PMU_CYCLE_IDX		(ARMV8_PMU_MAX_COUNTERS - 1)
 
diff --git a/include/linux/perf/arm_pmuv3.h b/include/linux/perf/arm_pmuv3.h
new file mode 100644
index 000000000000..131f486643bc
--- /dev/null
+++ b/include/linux/perf/arm_pmuv3.h
@@ -0,0 +1,76 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright (C) 2012 ARM Ltd.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#ifndef __PERF_ARM_PMUV3_H
+#define __PERF_ARM_PMUV3_H
+
+#define	ARMV8_PMU_MAX_COUNTERS	32
+#define	ARMV8_PMU_COUNTER_MASK	(ARMV8_PMU_MAX_COUNTERS - 1)
+
+/*
+ * Per-CPU PMCR: config reg
+ */
+#define ARMV8_PMU_PMCR_E	(1 << 0) /* Enable all counters */
+#define ARMV8_PMU_PMCR_P	(1 << 1) /* Reset all counters */
+#define ARMV8_PMU_PMCR_C	(1 << 2) /* Cycle counter reset */
+#define ARMV8_PMU_PMCR_D	(1 << 3) /* CCNT counts every 64th cpu cycle */
+#define ARMV8_PMU_PMCR_X	(1 << 4) /* Export to ETM */
+#define ARMV8_PMU_PMCR_DP	(1 << 5) /* Disable CCNT if non-invasive debug*/
+#define ARMV8_PMU_PMCR_LC	(1 << 6) /* Overflow on 64 bit cycle counter */
+#define	ARMV8_PMU_PMCR_N_SHIFT	11	 /* Number of counters supported */
+#define	ARMV8_PMU_PMCR_N_MASK	0x1f
+#define	ARMV8_PMU_PMCR_MASK	0x7f	 /* Mask for writable bits */
+
+/*
+ * PMOVSR: counters overflow flag status reg
+ */
+#define	ARMV8_PMU_OVSR_MASK		0xffffffff	/* Mask for writable bits */
+#define	ARMV8_PMU_OVERFLOWED_MASK	ARMV8_PMU_OVSR_MASK
+
+/*
+ * PMXEVTYPER: Event selection reg
+ */
+#define	ARMV8_PMU_EVTYPE_MASK	0xc800ffff	/* Mask for writable bits */
+#define	ARMV8_PMU_EVTYPE_EVENT	0xffff		/* Mask for EVENT bits */
+
+/*
+ * PMUv3 event types: required events
+ */
+#define ARMV8_PMUV3_PERFCTR_SW_INCR				0x00
+#define ARMV8_PMUV3_PERFCTR_L1D_CACHE_REFILL			0x03
+#define ARMV8_PMUV3_PERFCTR_L1D_CACHE				0x04
+#define ARMV8_PMUV3_PERFCTR_BR_MIS_PRED				0x10
+#define ARMV8_PMUV3_PERFCTR_CPU_CYCLES				0x11
+#define ARMV8_PMUV3_PERFCTR_BR_PRED				0x12
+
+/*
+ * Event filters for PMUv3
+ */
+#define	ARMV8_PMU_EXCLUDE_EL1	(1 << 31)
+#define	ARMV8_PMU_EXCLUDE_EL0	(1 << 30)
+#define	ARMV8_PMU_INCLUDE_EL2	(1 << 27)
+
+/*
+ * PMUSERENR: user enable reg
+ */
+#define ARMV8_PMU_USERENR_MASK	0xf		/* Mask for writable bits */
+#define ARMV8_PMU_USERENR_EN	(1 << 0) /* PMU regs can be accessed at EL0 */
+#define ARMV8_PMU_USERENR_SW	(1 << 1) /* PMSWINC can be written at EL0 */
+#define ARMV8_PMU_USERENR_CR	(1 << 2) /* Cycle counter can be read at EL0 */
+#define ARMV8_PMU_USERENR_ER	(1 << 3) /* Event counter can be read at EL0 */
+
+#endif
-- 
2.14.2

[PATCH v3 1/5] arm64: perf: Move PMUv3 driver to drivers/perf

[Marc Zyngier]

Having the ARM PMUv3 driver sitting in arch/arm64/kernel is getting
in the way of being able to use perf on ARMv8 cores running a 32bit
kernel, such as 32bit KVM guests.

This patch moves it into drivers/perf/arm_pmuv3.c, with an include
file in include/linux/perf/arm_pmuv3.h. The only thing left in
arch/arm64 is some mundane perf stuff.

Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
---
 arch/arm64/include/asm/perf_event.h                | 55 ----------------
 arch/arm64/kernel/Makefile                         |  1 -
 drivers/perf/Kconfig                               |  8 +++
 drivers/perf/Makefile                              |  1 +
 .../perf_event.c => drivers/perf/arm_pmuv3.c       |  2 +
 include/kvm/arm_pmu.h                              |  2 +-
 include/linux/perf/arm_pmuv3.h                     | 76 ++++++++++++++++++++++
 7 files changed, 88 insertions(+), 57 deletions(-)
 rename arch/arm64/kernel/perf_event.c => drivers/perf/arm_pmuv3.c (99%)
 create mode 100644 include/linux/perf/arm_pmuv3.h

diff --git a/arch/arm64/include/asm/perf_event.h b/arch/arm64/include/asm/perf_event.h
index f9ccc36d3dc3..5b33efeebabf 100644
--- a/arch/arm64/include/asm/perf_event.h
+++ b/arch/arm64/include/asm/perf_event.h
@@ -20,61 +20,6 @@
 #include <asm/stack_pointer.h>
 #include <asm/ptrace.h>
 
-#define	ARMV8_PMU_MAX_COUNTERS	32
-#define	ARMV8_PMU_COUNTER_MASK	(ARMV8_PMU_MAX_COUNTERS - 1)
-
-/*
- * Per-CPU PMCR: config reg
- */
-#define ARMV8_PMU_PMCR_E	(1 << 0) /* Enable all counters */
-#define ARMV8_PMU_PMCR_P	(1 << 1) /* Reset all counters */
-#define ARMV8_PMU_PMCR_C	(1 << 2) /* Cycle counter reset */
-#define ARMV8_PMU_PMCR_D	(1 << 3) /* CCNT counts every 64th cpu cycle */
-#define ARMV8_PMU_PMCR_X	(1 << 4) /* Export to ETM */
-#define ARMV8_PMU_PMCR_DP	(1 << 5) /* Disable CCNT if non-invasive debug*/
-#define ARMV8_PMU_PMCR_LC	(1 << 6) /* Overflow on 64 bit cycle counter */
-#define	ARMV8_PMU_PMCR_N_SHIFT	11	 /* Number of counters supported */
-#define	ARMV8_PMU_PMCR_N_MASK	0x1f
-#define	ARMV8_PMU_PMCR_MASK	0x7f	 /* Mask for writable bits */
-
-/*
- * PMOVSR: counters overflow flag status reg
- */
-#define	ARMV8_PMU_OVSR_MASK		0xffffffff	/* Mask for writable bits */
-#define	ARMV8_PMU_OVERFLOWED_MASK	ARMV8_PMU_OVSR_MASK
-
-/*
- * PMXEVTYPER: Event selection reg
- */
-#define	ARMV8_PMU_EVTYPE_MASK	0xc800ffff	/* Mask for writable bits */
-#define	ARMV8_PMU_EVTYPE_EVENT	0xffff		/* Mask for EVENT bits */
-
-/*
- * PMUv3 event types: required events
- */
-#define ARMV8_PMUV3_PERFCTR_SW_INCR				0x00
-#define ARMV8_PMUV3_PERFCTR_L1D_CACHE_REFILL			0x03
-#define ARMV8_PMUV3_PERFCTR_L1D_CACHE				0x04
-#define ARMV8_PMUV3_PERFCTR_BR_MIS_PRED				0x10
-#define ARMV8_PMUV3_PERFCTR_CPU_CYCLES				0x11
-#define ARMV8_PMUV3_PERFCTR_BR_PRED				0x12
-
-/*
- * Event filters for PMUv3
- */
-#define	ARMV8_PMU_EXCLUDE_EL1	(1 << 31)
-#define	ARMV8_PMU_EXCLUDE_EL0	(1 << 30)
-#define	ARMV8_PMU_INCLUDE_EL2	(1 << 27)
-
-/*
- * PMUSERENR: user enable reg
- */
-#define ARMV8_PMU_USERENR_MASK	0xf		/* Mask for writable bits */
-#define ARMV8_PMU_USERENR_EN	(1 << 0) /* PMU regs can be accessed at EL0 */
-#define ARMV8_PMU_USERENR_SW	(1 << 1) /* PMSWINC can be written at EL0 */
-#define ARMV8_PMU_USERENR_CR	(1 << 2) /* Cycle counter can be read at EL0 */
-#define ARMV8_PMU_USERENR_ER	(1 << 3) /* Event counter can be read@EL0 */
-
 #ifdef CONFIG_PERF_EVENTS
 struct pt_regs;
 extern unsigned long perf_instruction_pointer(struct pt_regs *regs);
diff --git a/arch/arm64/kernel/Makefile b/arch/arm64/kernel/Makefile
index bf825f38d206..aba9344a72ca 100644
--- a/arch/arm64/kernel/Makefile
+++ b/arch/arm64/kernel/Makefile
@@ -32,7 +32,6 @@ arm64-obj-$(CONFIG_FUNCTION_TRACER)	+= ftrace.o entry-ftrace.o
 arm64-obj-$(CONFIG_MODULES)		+= arm64ksyms.o module.o
 arm64-obj-$(CONFIG_ARM64_MODULE_PLTS)	+= module-plts.o
 arm64-obj-$(CONFIG_PERF_EVENTS)		+= perf_regs.o perf_callchain.o
-arm64-obj-$(CONFIG_HW_PERF_EVENTS)	+= perf_event.o
 arm64-obj-$(CONFIG_HAVE_HW_BREAKPOINT)	+= hw_breakpoint.o
 arm64-obj-$(CONFIG_CPU_PM)		+= sleep.o suspend.o
 arm64-obj-$(CONFIG_CPU_IDLE)		+= cpuidle.o
diff --git a/drivers/perf/Kconfig b/drivers/perf/Kconfig
index 28bb5a029558..39808b86b346 100644
--- a/drivers/perf/Kconfig
+++ b/drivers/perf/Kconfig
@@ -50,6 +50,14 @@ config ARM_PMU_ACPI
 	depends on ARM_PMU && ACPI
 	def_bool y
 
+config ARM_PMUV3
+	depends on HW_PERF_EVENTS && ARM64
+	bool "ARM PMUv3 support" if !ARM64
+	default y
+	help
+	  Say y if you want to use CPU performance monitors on ARMv8
+	  systems that implement the PMUv3 architecture.
+
 config ARM_DSU_PMU
 	tristate "ARM DynamIQ Shared Unit (DSU) PMU"
 	depends on ARM64
diff --git a/drivers/perf/Makefile b/drivers/perf/Makefile
index b3902bd37d53..a1a2f64e0c8f 100644
--- a/drivers/perf/Makefile
+++ b/drivers/perf/Makefile
@@ -4,6 +4,7 @@ obj-$(CONFIG_ARM_CCN) += arm-ccn.o
 obj-$(CONFIG_ARM_DSU_PMU) += arm_dsu_pmu.o
 obj-$(CONFIG_ARM_PMU) += arm_pmu.o arm_pmu_platform.o
 obj-$(CONFIG_ARM_PMU_ACPI) += arm_pmu_acpi.o
+obj-$(CONFIG_ARM_PMUV3) += arm_pmuv3.o
 obj-$(CONFIG_HISI_PMU) += hisilicon/
 obj-$(CONFIG_QCOM_L2_PMU)	+= qcom_l2_pmu.o
 obj-$(CONFIG_QCOM_L3_PMU) += qcom_l3_pmu.o
diff --git a/arch/arm64/kernel/perf_event.c b/drivers/perf/arm_pmuv3.c
similarity index 99%
rename from arch/arm64/kernel/perf_event.c
rename to drivers/perf/arm_pmuv3.c
index 85a251b6dfa8..bd19b16c44eb 100644
--- a/arch/arm64/kernel/perf_event.c
+++ b/drivers/perf/arm_pmuv3.c
@@ -1,3 +1,4 @@
+// SPDX-License-Identifier: GPL-2.0
 /*
  * PMU support
  *
@@ -27,6 +28,7 @@
 #include <linux/acpi.h>
 #include <linux/of.h>
 #include <linux/perf/arm_pmu.h>
+#include <linux/perf/arm_pmuv3.h>
 #include <linux/platform_device.h>
 
 /*
diff --git a/include/kvm/arm_pmu.h b/include/kvm/arm_pmu.h
index f87fe20fcb05..d16ce92cb2c0 100644
--- a/include/kvm/arm_pmu.h
+++ b/include/kvm/arm_pmu.h
@@ -19,7 +19,7 @@
 #define __ASM_ARM_KVM_PMU_H
 
 #include <linux/perf_event.h>
-#include <asm/perf_event.h>
+#include <linux/perf/arm_pmuv3.h>
 
 #define ARMV8_PMU_CYCLE_IDX		(ARMV8_PMU_MAX_COUNTERS - 1)
 
diff --git a/include/linux/perf/arm_pmuv3.h b/include/linux/perf/arm_pmuv3.h
new file mode 100644
index 000000000000..131f486643bc
--- /dev/null
+++ b/include/linux/perf/arm_pmuv3.h
@@ -0,0 +1,76 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright (C) 2012 ARM Ltd.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#ifndef __PERF_ARM_PMUV3_H
+#define __PERF_ARM_PMUV3_H
+
+#define	ARMV8_PMU_MAX_COUNTERS	32
+#define	ARMV8_PMU_COUNTER_MASK	(ARMV8_PMU_MAX_COUNTERS - 1)
+
+/*
+ * Per-CPU PMCR: config reg
+ */
+#define ARMV8_PMU_PMCR_E	(1 << 0) /* Enable all counters */
+#define ARMV8_PMU_PMCR_P	(1 << 1) /* Reset all counters */
+#define ARMV8_PMU_PMCR_C	(1 << 2) /* Cycle counter reset */
+#define ARMV8_PMU_PMCR_D	(1 << 3) /* CCNT counts every 64th cpu cycle */
+#define ARMV8_PMU_PMCR_X	(1 << 4) /* Export to ETM */
+#define ARMV8_PMU_PMCR_DP	(1 << 5) /* Disable CCNT if non-invasive debug*/
+#define ARMV8_PMU_PMCR_LC	(1 << 6) /* Overflow on 64 bit cycle counter */
+#define	ARMV8_PMU_PMCR_N_SHIFT	11	 /* Number of counters supported */
+#define	ARMV8_PMU_PMCR_N_MASK	0x1f
+#define	ARMV8_PMU_PMCR_MASK	0x7f	 /* Mask for writable bits */
+
+/*
+ * PMOVSR: counters overflow flag status reg
+ */
+#define	ARMV8_PMU_OVSR_MASK		0xffffffff	/* Mask for writable bits */
+#define	ARMV8_PMU_OVERFLOWED_MASK	ARMV8_PMU_OVSR_MASK
+
+/*
+ * PMXEVTYPER: Event selection reg
+ */
+#define	ARMV8_PMU_EVTYPE_MASK	0xc800ffff	/* Mask for writable bits */
+#define	ARMV8_PMU_EVTYPE_EVENT	0xffff		/* Mask for EVENT bits */
+
+/*
+ * PMUv3 event types: required events
+ */
+#define ARMV8_PMUV3_PERFCTR_SW_INCR				0x00
+#define ARMV8_PMUV3_PERFCTR_L1D_CACHE_REFILL			0x03
+#define ARMV8_PMUV3_PERFCTR_L1D_CACHE				0x04
+#define ARMV8_PMUV3_PERFCTR_BR_MIS_PRED				0x10
+#define ARMV8_PMUV3_PERFCTR_CPU_CYCLES				0x11
+#define ARMV8_PMUV3_PERFCTR_BR_PRED				0x12
+
+/*
+ * Event filters for PMUv3
+ */
+#define	ARMV8_PMU_EXCLUDE_EL1	(1 << 31)
+#define	ARMV8_PMU_EXCLUDE_EL0	(1 << 30)
+#define	ARMV8_PMU_INCLUDE_EL2	(1 << 27)
+
+/*
+ * PMUSERENR: user enable reg
+ */
+#define ARMV8_PMU_USERENR_MASK	0xf		/* Mask for writable bits */
+#define ARMV8_PMU_USERENR_EN	(1 << 0) /* PMU regs can be accessed at EL0 */
+#define ARMV8_PMU_USERENR_SW	(1 << 1) /* PMSWINC can be written at EL0 */
+#define ARMV8_PMU_USERENR_CR	(1 << 2) /* Cycle counter can be read at EL0 */
+#define ARMV8_PMU_USERENR_ER	(1 << 3) /* Event counter can be read@EL0 */
+
+#endif
-- 
2.14.2

[PATCH v3 2/5] arm64: perf: Abstract system register accesses away

[Marc Zyngier]

As we want to enable 32bit support, we need to distanciate the
PMUv3 driver from the AArch64 system register names.

This patch moves all system register accesses to an architecture
specific include file, allowing the 32bit counterpart to be
slotted in at a later time.

Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
---
 arch/arm64/include/asm/arm_pmuv3.h | 111 +++++++++++++++++++++++++++++++++++++
 drivers/perf/arm_pmuv3.c           |  40 ++++++-------
 include/linux/perf/arm_pmuv3.h     |   2 +
 3 files changed, 131 insertions(+), 22 deletions(-)
 create mode 100644 arch/arm64/include/asm/arm_pmuv3.h

diff --git a/arch/arm64/include/asm/arm_pmuv3.h b/arch/arm64/include/asm/arm_pmuv3.h
new file mode 100644
index 000000000000..558ed16b0f70
--- /dev/null
+++ b/arch/arm64/include/asm/arm_pmuv3.h
@@ -0,0 +1,111 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright (C) 2012 ARM Ltd.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#ifndef __ASM_PMUV3_H
+#define __ASM_PMUV3_H
+
+#include <asm/cpufeature.h>
+#include <asm/sysreg.h>
+
+static inline int read_pmuver(void)
+{
+	u64 dfr0 = read_sysreg(id_aa64dfr0_el1);
+	return cpuid_feature_extract_unsigned_field(dfr0,
+						    ID_AA64DFR0_PMUVER_SHIFT);
+}
+
+static inline void write_pmcr(u32 val)
+{
+	write_sysreg(val, pmcr_el0);
+}
+
+static inline u32 read_pmcr(void)
+{
+	return read_sysreg(pmcr_el0);
+}
+
+static inline void write_pmselr(u32 val)
+{
+	write_sysreg(val, pmselr_el0);
+}
+
+static inline void write_pmccntr(u64 val)
+{
+	write_sysreg(val, pmccntr_el0);
+}
+
+static inline u64 read_pmccntr(void)
+{
+	return read_sysreg(pmccntr_el0);
+}
+
+static inline void write_pmxevcntr(u32 val)
+{
+	write_sysreg(val, pmxevcntr_el0);
+}
+
+static inline u32 read_pmxevcntr(void)
+{
+	return read_sysreg(pmxevcntr_el0);
+}
+
+static inline void write_pmxevtyper(u32 val)
+{
+	write_sysreg(val, pmxevtyper_el0);
+}
+
+static inline void write_pmcntenset(u32 val)
+{
+	write_sysreg(val, pmcntenset_el0);
+}
+
+static inline void write_pmcntenclr(u32 val)
+{
+	write_sysreg(val, pmcntenclr_el0);
+}
+
+static inline void write_pmintenset(u32 val)
+{
+	write_sysreg(val, pmintenset_el1);
+}
+
+static inline void write_pmintenclr(u32 val)
+{
+	write_sysreg(val, pmintenclr_el1);
+}
+
+static inline void write_pmovsclr(u32 val)
+{
+	write_sysreg(val, pmovsclr_el0);
+}
+
+static inline u32 read_pmovsclr(void)
+{
+	return read_sysreg(pmovsclr_el0);
+}
+
+static inline u32 read_pmceid0(void)
+{
+	return read_sysreg(pmceid0_el0);
+}
+
+static inline u32 read_pmceid1(void)
+{
+	return read_sysreg(pmceid1_el0);
+}
+
+#endif
diff --git a/drivers/perf/arm_pmuv3.c b/drivers/perf/arm_pmuv3.c
index bd19b16c44eb..9c7b2c10b52e 100644
--- a/drivers/perf/arm_pmuv3.c
+++ b/drivers/perf/arm_pmuv3.c
@@ -22,7 +22,6 @@
 
 #include <asm/irq_regs.h>
 #include <asm/perf_event.h>
-#include <asm/sysreg.h>
 #include <asm/virt.h>
 
 #include <linux/acpi.h>
@@ -479,14 +478,14 @@ static struct attribute_group armv8_pmuv3_format_attr_group = {
 
 static inline u32 armv8pmu_pmcr_read(void)
 {
-	return read_sysreg(pmcr_el0);
+	return read_pmcr();
 }
 
 static inline void armv8pmu_pmcr_write(u32 val)
 {
 	val &= ARMV8_PMU_PMCR_MASK;
 	isb();
-	write_sysreg(val, pmcr_el0);
+	write_pmcr(val);
 }
 
 static inline int armv8pmu_has_overflowed(u32 pmovsr)
@@ -508,7 +507,7 @@ static inline int armv8pmu_counter_has_overflowed(u32 pmnc, int idx)
 static inline int armv8pmu_select_counter(int idx)
 {
 	u32 counter = ARMV8_IDX_TO_COUNTER(idx);
-	write_sysreg(counter, pmselr_el0);
+	write_pmselr(counter);
 	isb();
 
 	return idx;
@@ -525,9 +524,9 @@ static inline u32 armv8pmu_read_counter(struct perf_event *event)
 		pr_err("CPU%u reading wrong counter %d\n",
 			smp_processor_id(), idx);
 	else if (idx == ARMV8_IDX_CYCLE_COUNTER)
-		value = read_sysreg(pmccntr_el0);
+		value = read_pmccntr();
 	else if (armv8pmu_select_counter(idx) == idx)
-		value = read_sysreg(pmxevcntr_el0);
+		value = read_pmxevcntr();
 
 	return value;
 }
@@ -549,47 +548,47 @@ static inline void armv8pmu_write_counter(struct perf_event *event, u32 value)
 		 */
 		u64 value64 = 0xffffffff00000000ULL | value;
 
-		write_sysreg(value64, pmccntr_el0);
+		write_pmccntr(value64);
 	} else if (armv8pmu_select_counter(idx) == idx)
-		write_sysreg(value, pmxevcntr_el0);
+		write_pmxevcntr(value);
 }
 
 static inline void armv8pmu_write_evtype(int idx, u32 val)
 {
 	if (armv8pmu_select_counter(idx) == idx) {
 		val &= ARMV8_PMU_EVTYPE_MASK;
-		write_sysreg(val, pmxevtyper_el0);
+		write_pmxevtyper(val);
 	}
 }
 
 static inline int armv8pmu_enable_counter(int idx)
 {
 	u32 counter = ARMV8_IDX_TO_COUNTER(idx);
-	write_sysreg(BIT(counter), pmcntenset_el0);
+	write_pmcntenset(BIT(counter));
 	return idx;
 }
 
 static inline int armv8pmu_disable_counter(int idx)
 {
 	u32 counter = ARMV8_IDX_TO_COUNTER(idx);
-	write_sysreg(BIT(counter), pmcntenclr_el0);
+	write_pmcntenclr(BIT(counter));
 	return idx;
 }
 
 static inline int armv8pmu_enable_intens(int idx)
 {
 	u32 counter = ARMV8_IDX_TO_COUNTER(idx);
-	write_sysreg(BIT(counter), pmintenset_el1);
+	write_pmintenset(BIT(counter));
 	return idx;
 }
 
 static inline int armv8pmu_disable_intens(int idx)
 {
 	u32 counter = ARMV8_IDX_TO_COUNTER(idx);
-	write_sysreg(BIT(counter), pmintenclr_el1);
+	write_pmintenclr(BIT(counter));
 	isb();
 	/* Clear the overflow flag in case an interrupt is pending. */
-	write_sysreg(BIT(counter), pmovsclr_el0);
+	write_pmovsclr(BIT(counter));
 	isb();
 
 	return idx;
@@ -600,11 +599,11 @@ static inline u32 armv8pmu_getreset_flags(void)
 	u32 value;
 
 	/* Read */
-	value = read_sysreg(pmovsclr_el0);
+	value = read_pmovsclr();
 
 	/* Write to clear flags */
 	value &= ARMV8_PMU_OVSR_MASK;
-	write_sysreg(value, pmovsclr_el0);
+	write_pmovsclr(value);
 
 	return value;
 }
@@ -905,13 +904,10 @@ static void __armv8pmu_probe_pmu(void *info)
 {
 	struct armv8pmu_probe_info *probe = info;
 	struct arm_pmu *cpu_pmu = probe->pmu;
-	u64 dfr0;
 	u32 pmceid[2];
 	int pmuver;
 
-	dfr0 = read_sysreg(id_aa64dfr0_el1);
-	pmuver = cpuid_feature_extract_unsigned_field(dfr0,
-			ID_AA64DFR0_PMUVER_SHIFT);
+	pmuver = read_pmuver();
 	if (pmuver == 0xf || pmuver == 0)
 		return;
 
@@ -924,8 +920,8 @@ static void __armv8pmu_probe_pmu(void *info)
 	/* Add the CPU cycles counter */
 	cpu_pmu->num_events += 1;
 
-	pmceid[0] = read_sysreg(pmceid0_el0);
-	pmceid[1] = read_sysreg(pmceid1_el0);
+	pmceid[0] = read_pmceid0();
+	pmceid[1] = read_pmceid1();
 
 	bitmap_from_arr32(cpu_pmu->pmceid_bitmap,
 			     pmceid, ARMV8_PMUV3_MAX_COMMON_EVENTS);
diff --git a/include/linux/perf/arm_pmuv3.h b/include/linux/perf/arm_pmuv3.h
index 131f486643bc..909188bd7db0 100644
--- a/include/linux/perf/arm_pmuv3.h
+++ b/include/linux/perf/arm_pmuv3.h
@@ -18,6 +18,8 @@
 #ifndef __PERF_ARM_PMUV3_H
 #define __PERF_ARM_PMUV3_H
 
+#include <asm/arm_pmuv3.h>
+
 #define	ARMV8_PMU_MAX_COUNTERS	32
 #define	ARMV8_PMU_COUNTER_MASK	(ARMV8_PMU_MAX_COUNTERS - 1)
 
-- 
2.14.2

[PATCH v3 2/5] arm64: perf: Abstract system register accesses away

[Marc Zyngier]

As we want to enable 32bit support, we need to distanciate the
PMUv3 driver from the AArch64 system register names.

This patch moves all system register accesses to an architecture
specific include file, allowing the 32bit counterpart to be
slotted in at a later time.

Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
---
 arch/arm64/include/asm/arm_pmuv3.h | 111 +++++++++++++++++++++++++++++++++++++
 drivers/perf/arm_pmuv3.c           |  40 ++++++-------
 include/linux/perf/arm_pmuv3.h     |   2 +
 3 files changed, 131 insertions(+), 22 deletions(-)
 create mode 100644 arch/arm64/include/asm/arm_pmuv3.h

diff --git a/arch/arm64/include/asm/arm_pmuv3.h b/arch/arm64/include/asm/arm_pmuv3.h
new file mode 100644
index 000000000000..558ed16b0f70
--- /dev/null
+++ b/arch/arm64/include/asm/arm_pmuv3.h
@@ -0,0 +1,111 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright (C) 2012 ARM Ltd.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#ifndef __ASM_PMUV3_H
+#define __ASM_PMUV3_H
+
+#include <asm/cpufeature.h>
+#include <asm/sysreg.h>
+
+static inline int read_pmuver(void)
+{
+	u64 dfr0 = read_sysreg(id_aa64dfr0_el1);
+	return cpuid_feature_extract_unsigned_field(dfr0,
+						    ID_AA64DFR0_PMUVER_SHIFT);
+}
+
+static inline void write_pmcr(u32 val)
+{
+	write_sysreg(val, pmcr_el0);
+}
+
+static inline u32 read_pmcr(void)
+{
+	return read_sysreg(pmcr_el0);
+}
+
+static inline void write_pmselr(u32 val)
+{
+	write_sysreg(val, pmselr_el0);
+}
+
+static inline void write_pmccntr(u64 val)
+{
+	write_sysreg(val, pmccntr_el0);
+}
+
+static inline u64 read_pmccntr(void)
+{
+	return read_sysreg(pmccntr_el0);
+}
+
+static inline void write_pmxevcntr(u32 val)
+{
+	write_sysreg(val, pmxevcntr_el0);
+}
+
+static inline u32 read_pmxevcntr(void)
+{
+	return read_sysreg(pmxevcntr_el0);
+}
+
+static inline void write_pmxevtyper(u32 val)
+{
+	write_sysreg(val, pmxevtyper_el0);
+}
+
+static inline void write_pmcntenset(u32 val)
+{
+	write_sysreg(val, pmcntenset_el0);
+}
+
+static inline void write_pmcntenclr(u32 val)
+{
+	write_sysreg(val, pmcntenclr_el0);
+}
+
+static inline void write_pmintenset(u32 val)
+{
+	write_sysreg(val, pmintenset_el1);
+}
+
+static inline void write_pmintenclr(u32 val)
+{
+	write_sysreg(val, pmintenclr_el1);
+}
+
+static inline void write_pmovsclr(u32 val)
+{
+	write_sysreg(val, pmovsclr_el0);
+}
+
+static inline u32 read_pmovsclr(void)
+{
+	return read_sysreg(pmovsclr_el0);
+}
+
+static inline u32 read_pmceid0(void)
+{
+	return read_sysreg(pmceid0_el0);
+}
+
+static inline u32 read_pmceid1(void)
+{
+	return read_sysreg(pmceid1_el0);
+}
+
+#endif
diff --git a/drivers/perf/arm_pmuv3.c b/drivers/perf/arm_pmuv3.c
index bd19b16c44eb..9c7b2c10b52e 100644
--- a/drivers/perf/arm_pmuv3.c
+++ b/drivers/perf/arm_pmuv3.c
@@ -22,7 +22,6 @@
 
 #include <asm/irq_regs.h>
 #include <asm/perf_event.h>
-#include <asm/sysreg.h>
 #include <asm/virt.h>
 
 #include <linux/acpi.h>
@@ -479,14 +478,14 @@ static struct attribute_group armv8_pmuv3_format_attr_group = {
 
 static inline u32 armv8pmu_pmcr_read(void)
 {
-	return read_sysreg(pmcr_el0);
+	return read_pmcr();
 }
 
 static inline void armv8pmu_pmcr_write(u32 val)
 {
 	val &= ARMV8_PMU_PMCR_MASK;
 	isb();
-	write_sysreg(val, pmcr_el0);
+	write_pmcr(val);
 }
 
 static inline int armv8pmu_has_overflowed(u32 pmovsr)
@@ -508,7 +507,7 @@ static inline int armv8pmu_counter_has_overflowed(u32 pmnc, int idx)
 static inline int armv8pmu_select_counter(int idx)
 {
 	u32 counter = ARMV8_IDX_TO_COUNTER(idx);
-	write_sysreg(counter, pmselr_el0);
+	write_pmselr(counter);
 	isb();
 
 	return idx;
@@ -525,9 +524,9 @@ static inline u32 armv8pmu_read_counter(struct perf_event *event)
 		pr_err("CPU%u reading wrong counter %d\n",
 			smp_processor_id(), idx);
 	else if (idx == ARMV8_IDX_CYCLE_COUNTER)
-		value = read_sysreg(pmccntr_el0);
+		value = read_pmccntr();
 	else if (armv8pmu_select_counter(idx) == idx)
-		value = read_sysreg(pmxevcntr_el0);
+		value = read_pmxevcntr();
 
 	return value;
 }
@@ -549,47 +548,47 @@ static inline void armv8pmu_write_counter(struct perf_event *event, u32 value)
 		 */
 		u64 value64 = 0xffffffff00000000ULL | value;
 
-		write_sysreg(value64, pmccntr_el0);
+		write_pmccntr(value64);
 	} else if (armv8pmu_select_counter(idx) == idx)
-		write_sysreg(value, pmxevcntr_el0);
+		write_pmxevcntr(value);
 }
 
 static inline void armv8pmu_write_evtype(int idx, u32 val)
 {
 	if (armv8pmu_select_counter(idx) == idx) {
 		val &= ARMV8_PMU_EVTYPE_MASK;
-		write_sysreg(val, pmxevtyper_el0);
+		write_pmxevtyper(val);
 	}
 }
 
 static inline int armv8pmu_enable_counter(int idx)
 {
 	u32 counter = ARMV8_IDX_TO_COUNTER(idx);
-	write_sysreg(BIT(counter), pmcntenset_el0);
+	write_pmcntenset(BIT(counter));
 	return idx;
 }
 
 static inline int armv8pmu_disable_counter(int idx)
 {
 	u32 counter = ARMV8_IDX_TO_COUNTER(idx);
-	write_sysreg(BIT(counter), pmcntenclr_el0);
+	write_pmcntenclr(BIT(counter));
 	return idx;
 }
 
 static inline int armv8pmu_enable_intens(int idx)
 {
 	u32 counter = ARMV8_IDX_TO_COUNTER(idx);
-	write_sysreg(BIT(counter), pmintenset_el1);
+	write_pmintenset(BIT(counter));
 	return idx;
 }
 
 static inline int armv8pmu_disable_intens(int idx)
 {
 	u32 counter = ARMV8_IDX_TO_COUNTER(idx);
-	write_sysreg(BIT(counter), pmintenclr_el1);
+	write_pmintenclr(BIT(counter));
 	isb();
 	/* Clear the overflow flag in case an interrupt is pending. */
-	write_sysreg(BIT(counter), pmovsclr_el0);
+	write_pmovsclr(BIT(counter));
 	isb();
 
 	return idx;
@@ -600,11 +599,11 @@ static inline u32 armv8pmu_getreset_flags(void)
 	u32 value;
 
 	/* Read */
-	value = read_sysreg(pmovsclr_el0);
+	value = read_pmovsclr();
 
 	/* Write to clear flags */
 	value &= ARMV8_PMU_OVSR_MASK;
-	write_sysreg(value, pmovsclr_el0);
+	write_pmovsclr(value);
 
 	return value;
 }
@@ -905,13 +904,10 @@ static void __armv8pmu_probe_pmu(void *info)
 {
 	struct armv8pmu_probe_info *probe = info;
 	struct arm_pmu *cpu_pmu = probe->pmu;
-	u64 dfr0;
 	u32 pmceid[2];
 	int pmuver;
 
-	dfr0 = read_sysreg(id_aa64dfr0_el1);
-	pmuver = cpuid_feature_extract_unsigned_field(dfr0,
-			ID_AA64DFR0_PMUVER_SHIFT);
+	pmuver = read_pmuver();
 	if (pmuver == 0xf || pmuver == 0)
 		return;
 
@@ -924,8 +920,8 @@ static void __armv8pmu_probe_pmu(void *info)
 	/* Add the CPU cycles counter */
 	cpu_pmu->num_events += 1;
 
-	pmceid[0] = read_sysreg(pmceid0_el0);
-	pmceid[1] = read_sysreg(pmceid1_el0);
+	pmceid[0] = read_pmceid0();
+	pmceid[1] = read_pmceid1();
 
 	bitmap_from_arr32(cpu_pmu->pmceid_bitmap,
 			     pmceid, ARMV8_PMUV3_MAX_COMMON_EVENTS);
diff --git a/include/linux/perf/arm_pmuv3.h b/include/linux/perf/arm_pmuv3.h
index 131f486643bc..909188bd7db0 100644
--- a/include/linux/perf/arm_pmuv3.h
+++ b/include/linux/perf/arm_pmuv3.h
@@ -18,6 +18,8 @@
 #ifndef __PERF_ARM_PMUV3_H
 #define __PERF_ARM_PMUV3_H
 
+#include <asm/arm_pmuv3.h>
+
 #define	ARMV8_PMU_MAX_COUNTERS	32
 #define	ARMV8_PMU_COUNTER_MASK	(ARMV8_PMU_MAX_COUNTERS - 1)
 
-- 
2.14.2

[PATCH v3 3/5] ARM: Make CONFIG_CPU_V7 valid for 32bit ARMv8 implementations

[Marc Zyngier]

ARMv8 is a superset of ARMv7, and all the ARMv8 features are
discoverable with a set of ID registers. It means that we can
use CPU_V7 to guard ARMv8 features at compile time.

This commit simply amends the CPU_V7 configuration symbol comment
to reflect that CPU_V7 also covers ARMv8.

Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
---
 arch/arm/mm/Kconfig | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/arch/arm/mm/Kconfig b/arch/arm/mm/Kconfig
index 7f14acf67caf..86550040de19 100644
--- a/arch/arm/mm/Kconfig
+++ b/arch/arm/mm/Kconfig
@@ -402,7 +402,7 @@ config CPU_V6K
 	select CPU_THUMB_CAPABLE
 	select CPU_TLB_V6 if MMU
 
-# ARMv7
+# ARMv7 and ARMv8 architectures
 config CPU_V7
 	bool
 	select CPU_32v6K
-- 
2.14.2

[PATCH v3 3/5] ARM: Make CONFIG_CPU_V7 valid for 32bit ARMv8 implementations

[Marc Zyngier]

ARMv8 is a superset of ARMv7, and all the ARMv8 features are
discoverable with a set of ID registers. It means that we can
use CPU_V7 to guard ARMv8 features at compile time.

This commit simply amends the CPU_V7 configuration symbol comment
to reflect that CPU_V7 also covers ARMv8.

Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
---
 arch/arm/mm/Kconfig | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/arch/arm/mm/Kconfig b/arch/arm/mm/Kconfig
index 7f14acf67caf..86550040de19 100644
--- a/arch/arm/mm/Kconfig
+++ b/arch/arm/mm/Kconfig
@@ -402,7 +402,7 @@ config CPU_V6K
 	select CPU_THUMB_CAPABLE
 	select CPU_TLB_V6 if MMU
 
-# ARMv7
+# ARMv7 and ARMv8 architectures
 config CPU_V7
 	bool
 	select CPU_32v6K
-- 
2.14.2

[PATCH v3 4/5] ARM: perf: Allow the use of the PMUv3 driver on 32bit ARM

[Marc Zyngier]

The only thing stopping the PMUv3 driver from compiling on 32bit
is the lack of defined system registers names. This is easily
solved by providing the sysreg accessors and updating the Kconfig entry.

Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
---
 arch/arm/include/asm/arm_pmuv3.h | 125 +++++++++++++++++++++++++++++++++++++++
 drivers/perf/Kconfig             |   4 +-
 2 files changed, 127 insertions(+), 2 deletions(-)
 create mode 100644 arch/arm/include/asm/arm_pmuv3.h

diff --git a/arch/arm/include/asm/arm_pmuv3.h b/arch/arm/include/asm/arm_pmuv3.h
new file mode 100644
index 000000000000..e0f66c1d42b4
--- /dev/null
+++ b/arch/arm/include/asm/arm_pmuv3.h
@@ -0,0 +1,125 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright (C) 2012 ARM Ltd.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#ifndef __ASM_PMUV3_H
+#define __ASM_PMUV3_H
+
+#include <asm/cp15.h>
+#include <asm/cputype.h>
+
+#define PMCCNTR		__ACCESS_CP15_64(0, c9)
+
+#define PMCR		__ACCESS_CP15(c9, 0, c12, 0)
+#define PMCNTENSET	__ACCESS_CP15(c9, 0, c12, 1)
+#define PMCNTENCLR	__ACCESS_CP15(c9, 0, c12, 2)
+#define PMOVSR		__ACCESS_CP15(c9, 0, c12, 3)
+#define PMSELR		__ACCESS_CP15(c9, 0, c12, 5)
+#define PMCEID0		__ACCESS_CP15(c9, 0, c12, 6)
+#define PMCEID1		__ACCESS_CP15(c9, 0, c12, 7)
+#define PMXEVTYPER	__ACCESS_CP15(c9, 0, c13, 1)
+#define PMXEVCNTR	__ACCESS_CP15(c9, 0, c13, 2)
+#define PMINTENSET	__ACCESS_CP15(c9, 0, c14, 1)
+#define PMINTENCLR	__ACCESS_CP15(c9, 0, c14, 2)
+
+static inline int read_pmuver(void)
+{
+	/* PMUVers is not a signed field */
+	u32 dfr0 = read_cpuid_ext(CPUID_EXT_DFR0);
+	return (dfr0 >> 24) & 0xf;
+}
+
+static inline void write_pmcr(u32 val)
+{
+	write_sysreg(val, PMCR);
+}
+
+static inline u32 read_pmcr(void)
+{
+	return read_sysreg(PMCR);
+}
+
+static inline void write_pmselr(u32 val)
+{
+	write_sysreg(val, PMSELR);
+}
+
+static inline void write_pmccntr(u64 val)
+{
+	write_sysreg(val, PMCCNTR);
+}
+
+static inline u64 read_pmccntr(void)
+{
+	return read_sysreg(PMCCNTR);
+}
+
+static inline void write_pmxevcntr(u32 val)
+{
+	write_sysreg(val, PMXEVCNTR);
+}
+
+static inline u32 read_pmxevcntr(void)
+{
+	return read_sysreg(PMXEVCNTR);
+}
+
+static inline void write_pmxevtyper(u32 val)
+{
+	write_sysreg(val, PMXEVTYPER);
+}
+
+static inline void write_pmcntenset(u32 val)
+{
+	write_sysreg(val, PMCNTENSET);
+}
+
+static inline void write_pmcntenclr(u32 val)
+{
+	write_sysreg(val, PMCNTENCLR);
+}
+
+static inline void write_pmintenset(u32 val)
+{
+	write_sysreg(val, PMINTENSET);
+}
+
+static inline void write_pmintenclr(u32 val)
+{
+	write_sysreg(val, PMINTENCLR);
+}
+
+static inline void write_pmovsclr(u32 val)
+{
+	write_sysreg(val, PMOVSR);
+}
+
+static inline u32 read_pmovsclr(void)
+{
+	return read_sysreg(PMOVSR);
+}
+
+static inline u32 read_pmceid0(void)
+{
+	return read_sysreg(PMCEID0);
+}
+
+static inline u32 read_pmceid1(void)
+{
+	return read_sysreg(PMCEID1);
+}
+
+#endif
diff --git a/drivers/perf/Kconfig b/drivers/perf/Kconfig
index 39808b86b346..bf01dc1414c8 100644
--- a/drivers/perf/Kconfig
+++ b/drivers/perf/Kconfig
@@ -51,9 +51,9 @@ config ARM_PMU_ACPI
 	def_bool y
 
 config ARM_PMUV3
-	depends on HW_PERF_EVENTS && ARM64
+	depends on HW_PERF_EVENTS && ((ARM && CPU_V7) || ARM64)
 	bool "ARM PMUv3 support" if !ARM64
-	default y
+	default ARM64
 	help
 	  Say y if you want to use CPU performance monitors on ARMv8
 	  systems that implement the PMUv3 architecture.
-- 
2.14.2

[PATCH v3 4/5] ARM: perf: Allow the use of the PMUv3 driver on 32bit ARM

[Marc Zyngier]

The only thing stopping the PMUv3 driver from compiling on 32bit
is the lack of defined system registers names. This is easily
solved by providing the sysreg accessors and updating the Kconfig entry.

Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
---
 arch/arm/include/asm/arm_pmuv3.h | 125 +++++++++++++++++++++++++++++++++++++++
 drivers/perf/Kconfig             |   4 +-
 2 files changed, 127 insertions(+), 2 deletions(-)
 create mode 100644 arch/arm/include/asm/arm_pmuv3.h

diff --git a/arch/arm/include/asm/arm_pmuv3.h b/arch/arm/include/asm/arm_pmuv3.h
new file mode 100644
index 000000000000..e0f66c1d42b4
--- /dev/null
+++ b/arch/arm/include/asm/arm_pmuv3.h
@@ -0,0 +1,125 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright (C) 2012 ARM Ltd.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#ifndef __ASM_PMUV3_H
+#define __ASM_PMUV3_H
+
+#include <asm/cp15.h>
+#include <asm/cputype.h>
+
+#define PMCCNTR		__ACCESS_CP15_64(0, c9)
+
+#define PMCR		__ACCESS_CP15(c9, 0, c12, 0)
+#define PMCNTENSET	__ACCESS_CP15(c9, 0, c12, 1)
+#define PMCNTENCLR	__ACCESS_CP15(c9, 0, c12, 2)
+#define PMOVSR		__ACCESS_CP15(c9, 0, c12, 3)
+#define PMSELR		__ACCESS_CP15(c9, 0, c12, 5)
+#define PMCEID0		__ACCESS_CP15(c9, 0, c12, 6)
+#define PMCEID1		__ACCESS_CP15(c9, 0, c12, 7)
+#define PMXEVTYPER	__ACCESS_CP15(c9, 0, c13, 1)
+#define PMXEVCNTR	__ACCESS_CP15(c9, 0, c13, 2)
+#define PMINTENSET	__ACCESS_CP15(c9, 0, c14, 1)
+#define PMINTENCLR	__ACCESS_CP15(c9, 0, c14, 2)
+
+static inline int read_pmuver(void)
+{
+	/* PMUVers is not a signed field */
+	u32 dfr0 = read_cpuid_ext(CPUID_EXT_DFR0);
+	return (dfr0 >> 24) & 0xf;
+}
+
+static inline void write_pmcr(u32 val)
+{
+	write_sysreg(val, PMCR);
+}
+
+static inline u32 read_pmcr(void)
+{
+	return read_sysreg(PMCR);
+}
+
+static inline void write_pmselr(u32 val)
+{
+	write_sysreg(val, PMSELR);
+}
+
+static inline void write_pmccntr(u64 val)
+{
+	write_sysreg(val, PMCCNTR);
+}
+
+static inline u64 read_pmccntr(void)
+{
+	return read_sysreg(PMCCNTR);
+}
+
+static inline void write_pmxevcntr(u32 val)
+{
+	write_sysreg(val, PMXEVCNTR);
+}
+
+static inline u32 read_pmxevcntr(void)
+{
+	return read_sysreg(PMXEVCNTR);
+}
+
+static inline void write_pmxevtyper(u32 val)
+{
+	write_sysreg(val, PMXEVTYPER);
+}
+
+static inline void write_pmcntenset(u32 val)
+{
+	write_sysreg(val, PMCNTENSET);
+}
+
+static inline void write_pmcntenclr(u32 val)
+{
+	write_sysreg(val, PMCNTENCLR);
+}
+
+static inline void write_pmintenset(u32 val)
+{
+	write_sysreg(val, PMINTENSET);
+}
+
+static inline void write_pmintenclr(u32 val)
+{
+	write_sysreg(val, PMINTENCLR);
+}
+
+static inline void write_pmovsclr(u32 val)
+{
+	write_sysreg(val, PMOVSR);
+}
+
+static inline u32 read_pmovsclr(void)
+{
+	return read_sysreg(PMOVSR);
+}
+
+static inline u32 read_pmceid0(void)
+{
+	return read_sysreg(PMCEID0);
+}
+
+static inline u32 read_pmceid1(void)
+{
+	return read_sysreg(PMCEID1);
+}
+
+#endif
diff --git a/drivers/perf/Kconfig b/drivers/perf/Kconfig
index 39808b86b346..bf01dc1414c8 100644
--- a/drivers/perf/Kconfig
+++ b/drivers/perf/Kconfig
@@ -51,9 +51,9 @@ config ARM_PMU_ACPI
 	def_bool y
 
 config ARM_PMUV3
-	depends on HW_PERF_EVENTS && ARM64
+	depends on HW_PERF_EVENTS && ((ARM && CPU_V7) || ARM64)
 	bool "ARM PMUv3 support" if !ARM64
-	default y
+	default ARM64
 	help
 	  Say y if you want to use CPU performance monitors on ARMv8
 	  systems that implement the PMUv3 architecture.
-- 
2.14.2

[PATCH v3 5/5] ARM: mach-virt: Select PMUv3 driver by default

[Marc Zyngier]

Since 32bit guests are not unlikely to run on an ARMv8 host,
let's select the PMUv3 driver, which allows the PMU to be used
on such systems.

Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
---
 arch/arm/Kconfig | 1 +
 1 file changed, 1 insertion(+)

diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig
index a7f8e7f4b88f..5dc5d5f15560 100644
--- a/arch/arm/Kconfig
+++ b/arch/arm/Kconfig
@@ -703,6 +703,7 @@ config ARCH_VIRT
 	select ARM_GIC_V3
 	select ARM_GIC_V3_ITS if PCI
 	select ARM_PSCI
+	select ARM_PMUV3 if PERF_EVENTS
 	select HAVE_ARM_ARCH_TIMER
 
 #
-- 
2.14.2

[PATCH v3 5/5] ARM: mach-virt: Select PMUv3 driver by default

[Marc Zyngier]

Since 32bit guests are not unlikely to run on an ARMv8 host,
let's select the PMUv3 driver, which allows the PMU to be used
on such systems.

Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
---
 arch/arm/Kconfig | 1 +
 1 file changed, 1 insertion(+)

diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig
index a7f8e7f4b88f..5dc5d5f15560 100644
--- a/arch/arm/Kconfig
+++ b/arch/arm/Kconfig
@@ -703,6 +703,7 @@ config ARCH_VIRT
 	select ARM_GIC_V3
 	select ARM_GIC_V3_ITS if PCI
 	select ARM_PSCI
+	select ARM_PMUV3 if PERF_EVENTS
 	select HAVE_ARM_ARCH_TIMER
 
 #
-- 
2.14.2

Re: [PATCH v3 0/5]

[Vince Weaver]

On Fri, 18 May 2018, Marc Zyngier wrote:

> There is also the case of people natively running 32bit kernels on
> 64bit HW and trying to upstream unspeakable hacks, hoping that the
> stars will align and that they'll win the lottery (see [1]).

I've tested these patches on a Raspberry Pi 3B running a 32-bit upstream 
(4.17-rc5-git) kernel and they work.

[    0.472906] hw perfevents: enabled with armv8_cortex_a53 PMU driver, 7 counters available

I only needed to add this to the devicetree

	arm-pmu {
		compatible = "arm,cortex-a53-pmu";
		interrupt-parent = <&local_intc>;
		interrupts = <9 IRQ_TYPE_LEVEL_HIGH>;
	};


Tested-by: Vince Weaver <vincent.weaver@maine.edu>

Vince

[PATCH v3 0/5]

[Vince Weaver]

On Fri, 18 May 2018, Marc Zyngier wrote:

> There is also the case of people natively running 32bit kernels on
> 64bit HW and trying to upstream unspeakable hacks, hoping that the
> stars will align and that they'll win the lottery (see [1]).

I've tested these patches on a Raspberry Pi 3B running a 32-bit upstream 
(4.17-rc5-git) kernel and they work.

[    0.472906] hw perfevents: enabled with armv8_cortex_a53 PMU driver, 7 counters available

I only needed to add this to the devicetree

	arm-pmu {
		compatible = "arm,cortex-a53-pmu";
		interrupt-parent = <&local_intc>;
		interrupts = <9 IRQ_TYPE_LEVEL_HIGH>;
	};


Tested-by: Vince Weaver <vincent.weaver@maine.edu>

Vince

Re: [PATCH v3 0/5]

[Marc Zyngier]

[/me beats himself for not writing a subject line...]

On 18/05/18 17:29, Vince Weaver wrote:
> On Fri, 18 May 2018, Marc Zyngier wrote:
> 
>> There is also the case of people natively running 32bit kernels on
>> 64bit HW and trying to upstream unspeakable hacks, hoping that the
>> stars will align and that they'll win the lottery (see [1]).
> 
> I've tested these patches on a Raspberry Pi 3B running a 32-bit upstream 
> (4.17-rc5-git) kernel and they work.
> 
> [    0.472906] hw perfevents: enabled with armv8_cortex_a53 PMU driver, 7 counters available
> 
> I only needed to add this to the devicetree
> 
> 	arm-pmu {
> 		compatible = "arm,cortex-a53-pmu";
> 		interrupt-parent = <&local_intc>;
> 		interrupts = <9 IRQ_TYPE_LEVEL_HIGH>;
> 	};

That's definitely the sensible thing to have on such hardware. Why isn't
it in the upstream DT already, irrespective of the state of the kernel
support?

> Tested-by: Vince Weaver <vincent.weaver@maine.edu>

Thanks a lot for testing.

	M.
-- 
Jazz is not dead. It just smells funny...

[PATCH v3 0/5]

[Marc Zyngier]

[/me beats himself for not writing a subject line...]

On 18/05/18 17:29, Vince Weaver wrote:
> On Fri, 18 May 2018, Marc Zyngier wrote:
> 
>> There is also the case of people natively running 32bit kernels on
>> 64bit HW and trying to upstream unspeakable hacks, hoping that the
>> stars will align and that they'll win the lottery (see [1]).
> 
> I've tested these patches on a Raspberry Pi 3B running a 32-bit upstream 
> (4.17-rc5-git) kernel and they work.
> 
> [    0.472906] hw perfevents: enabled with armv8_cortex_a53 PMU driver, 7 counters available
> 
> I only needed to add this to the devicetree
> 
> 	arm-pmu {
> 		compatible = "arm,cortex-a53-pmu";
> 		interrupt-parent = <&local_intc>;
> 		interrupts = <9 IRQ_TYPE_LEVEL_HIGH>;
> 	};

That's definitely the sensible thing to have on such hardware. Why isn't
it in the upstream DT already, irrespective of the state of the kernel
support?

> Tested-by: Vince Weaver <vincent.weaver@maine.edu>

Thanks a lot for testing.

	M.
-- 
Jazz is not dead. It just smells funny...

Re: [PATCH v3 0/5]

[Stefan Wahren]

> Marc Zyngier <marc.zyngier@arm.com> hat am 18. Mai 2018 um 18:41 geschrieben:
> 
> 
> [/me beats himself for not writing a subject line...]
> 
> On 18/05/18 17:29, Vince Weaver wrote:
> > On Fri, 18 May 2018, Marc Zyngier wrote:
> > 
> >> There is also the case of people natively running 32bit kernels on
> >> 64bit HW and trying to upstream unspeakable hacks, hoping that the
> >> stars will align and that they'll win the lottery (see [1]).
> > 
> > I've tested these patches on a Raspberry Pi 3B running a 32-bit upstream 
> > (4.17-rc5-git) kernel and they work.
> > 
> > [    0.472906] hw perfevents: enabled with armv8_cortex_a53 PMU driver, 7 counters available
> > 
> > I only needed to add this to the devicetree
> > 
> > 	arm-pmu {
> > 		compatible = "arm,cortex-a53-pmu";
> > 		interrupt-parent = <&local_intc>;
> > 		interrupts = <9 IRQ_TYPE_LEVEL_HIGH>;
> > 	};
> 
> That's definitely the sensible thing to have on such hardware. Why isn't
> it in the upstream DT already, irrespective of the state of the kernel
> support?

I remember that Vince point out the absence. He asked about how to implement it and i wasn't sure about it. At this time we hadn't IRQ polarity support. So we wanted to get this puzzle piece before. In march i put it on my TODO list, but then RPI 3 B+ support had higher prio to get into 4.18.

In general we have the problem that most of the users take the downstream kernel and don't know about the differences. Luckily more distributions switch to the upstream kernel, which increases the feedback.

> 
> > Tested-by: Vince Weaver <vincent.weaver@maine.edu>

Thanks again
Stefan

> 
> Thanks a lot for testing.
> 
> 	M.
> -- 
> Jazz is not dead. It just smells funny...

[PATCH v3 0/5]

[Stefan Wahren]

> Marc Zyngier <marc.zyngier@arm.com> hat am 18. Mai 2018 um 18:41 geschrieben:
> 
> 
> [/me beats himself for not writing a subject line...]
> 
> On 18/05/18 17:29, Vince Weaver wrote:
> > On Fri, 18 May 2018, Marc Zyngier wrote:
> > 
> >> There is also the case of people natively running 32bit kernels on
> >> 64bit HW and trying to upstream unspeakable hacks, hoping that the
> >> stars will align and that they'll win the lottery (see [1]).
> > 
> > I've tested these patches on a Raspberry Pi 3B running a 32-bit upstream 
> > (4.17-rc5-git) kernel and they work.
> > 
> > [    0.472906] hw perfevents: enabled with armv8_cortex_a53 PMU driver, 7 counters available
> > 
> > I only needed to add this to the devicetree
> > 
> > 	arm-pmu {
> > 		compatible = "arm,cortex-a53-pmu";
> > 		interrupt-parent = <&local_intc>;
> > 		interrupts = <9 IRQ_TYPE_LEVEL_HIGH>;
> > 	};
> 
> That's definitely the sensible thing to have on such hardware. Why isn't
> it in the upstream DT already, irrespective of the state of the kernel
> support?

I remember that Vince point out the absence. He asked about how to implement it and i wasn't sure about it. At this time we hadn't IRQ polarity support. So we wanted to get this puzzle piece before. In march i put it on my TODO list, but then RPI 3 B+ support had higher prio to get into 4.18.

In general we have the problem that most of the users take the downstream kernel and don't know about the differences. Luckily more distributions switch to the upstream kernel, which increases the feedback.

> 
> > Tested-by: Vince Weaver <vincent.weaver@maine.edu>

Thanks again
Stefan

> 
> Thanks a lot for testing.
> 
> 	M.
> -- 
> Jazz is not dead. It just smells funny...

Re: [PATCH v3 4/5] ARM: perf: Allow the use of the PMUv3 driver on 32bit ARM

[Vladimir Murzin]

On 18/05/18 15:39, Marc Zyngier wrote:
> +static inline int read_pmuver(void)
> +{
> +	/* PMUVers is not a signed field */
> +	u32 dfr0 = read_cpuid_ext(CPUID_EXT_DFR0);
> +	return (dfr0 >> 24) & 0xf;
> +}

Should we rule out versions prior v3 here or in __armv8pmu_probe_pmu()?

Thanks
Vladimir

[PATCH v3 4/5] ARM: perf: Allow the use of the PMUv3 driver on 32bit ARM

[Vladimir Murzin]

On 18/05/18 15:39, Marc Zyngier wrote:
> +static inline int read_pmuver(void)
> +{
> +	/* PMUVers is not a signed field */
> +	u32 dfr0 = read_cpuid_ext(CPUID_EXT_DFR0);
> +	return (dfr0 >> 24) & 0xf;
> +}

Should we rule out versions prior v3 here or in __armv8pmu_probe_pmu()?

Thanks
Vladimir

Re: [PATCH v3 4/5] ARM: perf: Allow the use of the PMUv3 driver on 32bit ARM

[Marc Zyngier]

On 21/05/18 10:34, Vladimir Murzin wrote:
> On 18/05/18 15:39, Marc Zyngier wrote:
>> +static inline int read_pmuver(void)
>> +{
>> +	/* PMUVers is not a signed field */
>> +	u32 dfr0 = read_cpuid_ext(CPUID_EXT_DFR0);
>> +	return (dfr0 >> 24) & 0xf;
>> +}
> 
> Should we rule out versions prior v3 here or in __armv8pmu_probe_pmu()?

I'm in two minds about it: The ARM ARM is quite clear about the fact
that this is not legal ("In any ARMv8 implementation the values 0001 and
0010 are not permitted."), and DT clearly lied to us in that case.

If we want to consistently handle that case, it should probably be done
in __armv8pmu_probe_pmu, bailing out if the version is our of scope for
the driver.

Thanks,

	M.
-- 
Jazz is not dead. It just smells funny...

[PATCH v3 4/5] ARM: perf: Allow the use of the PMUv3 driver on 32bit ARM

[Marc Zyngier]

On 21/05/18 10:34, Vladimir Murzin wrote:
> On 18/05/18 15:39, Marc Zyngier wrote:
>> +static inline int read_pmuver(void)
>> +{
>> +	/* PMUVers is not a signed field */
>> +	u32 dfr0 = read_cpuid_ext(CPUID_EXT_DFR0);
>> +	return (dfr0 >> 24) & 0xf;
>> +}
> 
> Should we rule out versions prior v3 here or in __armv8pmu_probe_pmu()?

I'm in two minds about it: The ARM ARM is quite clear about the fact
that this is not legal ("In any ARMv8 implementation the values 0001 and
0010 are not permitted."), and DT clearly lied to us in that case.

If we want to consistently handle that case, it should probably be done
in __armv8pmu_probe_pmu, bailing out if the version is our of scope for
the driver.

Thanks,

	M.
-- 
Jazz is not dead. It just smells funny...

Re: [PATCH v3 0/5]

[Will Deacon]

Hi Marc,

Thanks for this.

On Fri, May 18, 2018 at 03:39:08PM +0100, Marc Zyngier wrote:
> PMUv3 has been introduced with ARMv8 and, while it has only been used
> on 64bit systems so far, it would definitely be useful for 32bit
> guests running under KVM/arm64, for example.
> 
> There is also the case of people natively running 32bit kernels on
> 64bit HW and trying to upstream unspeakable hacks, hoping that the
> stars will align and that they'll win the lottery (see [1]).
> 
> So let's try again, and make the PMUv3 driver usable for everyone.
> 
> This is done in three steps:
> (1) Move the driver from arch/arm64 to drivers/perf
> (2) Add a handful of system register accessors so that we can reuse
>     the driver on 32bit
> (3) Provide the same accessors on 32bit, enable compilation, and
>     make it the default selection for mach-virt.
> 
> Tested on a Seattle box with 32bit guests.

I think we should go ahead with something like this, but I don't think
we're quite there with these patches. If we're going to move the arch code
out into drivers, let's do that for the perf_event* files under arch/arm/
as well. Then we could have a structure along the lines of:


  drivers/perf/arm_pmu.c			- As it is today
  drivers/perf/arm_cpu/xscale_pmu.c		- Only builds for 32-bit
  drivers/perf/arm_cpu/armv6_pmu.c		- Only builds for 32-bit
  drivers/perf/arm_cpu/arch_pmu.c		- Works for v7/v8 on
                                                  both 32-bit and 64-bit

The latter can then pull in whatever accessors it needs from the arch
code headers.

I know it's more of an invasive change, but this way we always end up
running the same code on the two architectures and it will be much easier
to maintain.

Will

[PATCH v3 0/5]

[Will Deacon]

Hi Marc,

Thanks for this.

On Fri, May 18, 2018 at 03:39:08PM +0100, Marc Zyngier wrote:
> PMUv3 has been introduced with ARMv8 and, while it has only been used
> on 64bit systems so far, it would definitely be useful for 32bit
> guests running under KVM/arm64, for example.
> 
> There is also the case of people natively running 32bit kernels on
> 64bit HW and trying to upstream unspeakable hacks, hoping that the
> stars will align and that they'll win the lottery (see [1]).
> 
> So let's try again, and make the PMUv3 driver usable for everyone.
> 
> This is done in three steps:
> (1) Move the driver from arch/arm64 to drivers/perf
> (2) Add a handful of system register accessors so that we can reuse
>     the driver on 32bit
> (3) Provide the same accessors on 32bit, enable compilation, and
>     make it the default selection for mach-virt.
> 
> Tested on a Seattle box with 32bit guests.

I think we should go ahead with something like this, but I don't think
we're quite there with these patches. If we're going to move the arch code
out into drivers, let's do that for the perf_event* files under arch/arm/
as well. Then we could have a structure along the lines of:


  drivers/perf/arm_pmu.c			- As it is today
  drivers/perf/arm_cpu/xscale_pmu.c		- Only builds for 32-bit
  drivers/perf/arm_cpu/armv6_pmu.c		- Only builds for 32-bit
  drivers/perf/arm_cpu/arch_pmu.c		- Works for v7/v8 on
                                                  both 32-bit and 64-bit

The latter can then pull in whatever accessors it needs from the arch
code headers.

I know it's more of an invasive change, but this way we always end up
running the same code on the two architectures and it will be much easier
to maintain.

Will

Re: [PATCH v3 0/5]

[Florian Fainelli]

Hi Will, Marc,

On 5/21/2018 11:19 AM, Will Deacon wrote:
> Hi Marc,
> 
> Thanks for this.
> 
> On Fri, May 18, 2018 at 03:39:08PM +0100, Marc Zyngier wrote:
>> PMUv3 has been introduced with ARMv8 and, while it has only been used
>> on 64bit systems so far, it would definitely be useful for 32bit
>> guests running under KVM/arm64, for example.
>>
>> There is also the case of people natively running 32bit kernels on
>> 64bit HW and trying to upstream unspeakable hacks, hoping that the
>> stars will align and that they'll win the lottery (see [1]).
>>
>> So let's try again, and make the PMUv3 driver usable for everyone.
>>
>> This is done in three steps:
>> (1) Move the driver from arch/arm64 to drivers/perf
>> (2) Add a handful of system register accessors so that we can reuse
>>      the driver on 32bit
>> (3) Provide the same accessors on 32bit, enable compilation, and
>>      make it the default selection for mach-virt.
>>
>> Tested on a Seattle box with 32bit guests.
> 
> I think we should go ahead with something like this, but I don't think
> we're quite there with these patches. If we're going to move the arch code
> out into drivers, let's do that for the perf_event* files under arch/arm/
> as well. Then we could have a structure along the lines of:
> 
> 
>    drivers/perf/arm_pmu.c			- As it is today
>    drivers/perf/arm_cpu/xscale_pmu.c		- Only builds for 32-bit
>    drivers/perf/arm_cpu/armv6_pmu.c		- Only builds for 32-bit
>    drivers/perf/arm_cpu/arch_pmu.c		- Works for v7/v8 on
>                                                    both 32-bit and 64-bit

There has been a recent renewed interest in this patch series from my 
side as well as other users, and I have picked up where Marc left and 
started moving the Xscale and ARMv6 PMU drivers as Will had asked the 
current result can be seen here:

https://github.com/ffainelli/linux/commits/arm-pmuv3

It is not entirely clear to me what you had in mind when you want to 
have a common driver for v7 and v8 under 
drivers/perf/arm_cpu/arch_pmu.c. How much sharing or consolidation would 
you expect to be done there?

Thanks!
-- 
Florian

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

Re: [PATCH v3 0/5]

[Marc Zyngier]

Hi Florian,

On Sun, 19 Dec 2021 23:46:32 +0000,
Florian Fainelli <f.fainelli@gmail.com> wrote:
> 
> Hi Will, Marc,

+Mark

> 
> On 5/21/2018 11:19 AM, Will Deacon wrote:
> > Hi Marc,
> > 
> > Thanks for this.
> > 
> > On Fri, May 18, 2018 at 03:39:08PM +0100, Marc Zyngier wrote:
> >> PMUv3 has been introduced with ARMv8 and, while it has only been used
> >> on 64bit systems so far, it would definitely be useful for 32bit
> >> guests running under KVM/arm64, for example.
> >> 
> >> There is also the case of people natively running 32bit kernels on
> >> 64bit HW and trying to upstream unspeakable hacks, hoping that the
> >> stars will align and that they'll win the lottery (see [1]).
> >> 
> >> So let's try again, and make the PMUv3 driver usable for everyone.
> >> 
> >> This is done in three steps:
> >> (1) Move the driver from arch/arm64 to drivers/perf
> >> (2) Add a handful of system register accessors so that we can reuse
> >>      the driver on 32bit
> >> (3) Provide the same accessors on 32bit, enable compilation, and
> >>      make it the default selection for mach-virt.
> >> 
> >> Tested on a Seattle box with 32bit guests.
> > 
> > I think we should go ahead with something like this, but I don't think
> > we're quite there with these patches. If we're going to move the arch code
> > out into drivers, let's do that for the perf_event* files under arch/arm/
> > as well. Then we could have a structure along the lines of:
> > 
> > 
> >    drivers/perf/arm_pmu.c			- As it is today
> >    drivers/perf/arm_cpu/xscale_pmu.c		- Only builds for 32-bit
> >    drivers/perf/arm_cpu/armv6_pmu.c		- Only builds for 32-bit
> >    drivers/perf/arm_cpu/arch_pmu.c		- Works for v7/v8 on
> >                                                    both 32-bit and 64-bit
> 
> There has been a recent renewed interest in this patch series from my
> side as well as other users, and I have picked up where Marc left and
> started moving the Xscale and ARMv6 PMU drivers as Will had asked the
> current result can be seen here:
> 
> https://github.com/ffainelli/linux/commits/arm-pmuv3

Blast from the past!

> 
> It is not entirely clear to me what you had in mind when you want to
> have a common driver for v7 and v8 under
> drivers/perf/arm_cpu/arch_pmu.c. How much sharing or consolidation
> would you expect to be done there?

The v7/v8 models are very similar (they share most of the registers,
and there is only a few bits and pieces that differ between PMUv3 and
its ancestors). It should be possible to build a single driver

However, the 32bit PMU code is bundled together with the
Krait/Scorpion code that isn't quite the same. Maybe this should be
spun as a different driver if this can be disentangled neatly enough?

	M.

-- 
Without deviation from the norm, progress is not possible.

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

Re: [PATCH v3 0/5]

[Matthias Brugger]

Hi Bernhard,

Please resend with a subject line for the cover letter.

Regards,
Matthias

On 29/05/2023 17:35, Bernhard Rosenkränzer wrote:
> From: Balsam CHIHI <bchihi@baylibre.com>
> 
> Add full LVTS support (MCU thermal domain + AP thermal domain) to MediaTek MT8192 SoC.
> Also, add Suspend and Resume support to LVTS Driver (all SoCs),
> and update the documentation that describes the Calibration Data Offsets.
> 
> Changelog:
>      v3 :
>          - Rebased :
>              base-commit: 6a3d37b4d885129561e1cef361216f00472f7d2e
>          - Fix issues in v2 pointed out by Nícolas F. R. A. Prado <nfraprado@collabora.com>:
>            Use filtered mode to make sure threshold interrupts are triggered,
>            protocol documentation, cosmetics
>          - I (bero@baylibre.com) will be taking care of this patchset
>            from now on, since Balsam has left BayLibre. Thanks for
>            getting it almost ready, Balsam!
> 
>      v2 :
>          - Based on top of thermal/linux-next :
>              base-commit: 7ac82227ee046f8234471de4c12a40b8c2d3ddcc
>          - Squash "add thermal zones and thermal nodes" and
>              "add temperature mitigation threshold" commits together to form
>              "arm64: dts: mediatek: mt8192: Add thermal nodes and thermal zones" commit.
>          - Add Suspend and Resume support to LVTS Driver.
>          - Update Calibration Data documentation.
>          - Fix calibration data offsets for mt8192
>              (Thanks to "Chen-Yu Tsai" and "Nícolas F. R. A. Prado").
>          https://lore.kernel.org/all/20230425133052.199767-1-bchihi@baylibre.com/
>          Tested-by: Chen-Yu Tsai <wenst@chromium.org>
> 
>      v1 :
>          - The initial series "Add LVTS support for mt8192" :
>              "https://lore.kernel.org/all/20230307163413.143334-1-bchihi@baylibre.com/".
> 
> Balsam CHIHI (5):
>    dt-bindings: thermal: mediatek: Add LVTS thermal controller definition
>      for mt8192
>    thermal/drivers/mediatek/lvts_thermal: Add suspend and resume
>    thermal/drivers/mediatek/lvts_thermal: Add mt8192 support
>    arm64: dts: mediatek: mt8192: Add thermal nodes and thermal zones
>    thermal/drivers/mediatek/lvts_thermal: Update calibration data
>      documentation
> 
>   arch/arm64/boot/dts/mediatek/mt8192.dtsi      | 454 ++++++++++++++++++
>   drivers/thermal/mediatek/lvts_thermal.c       | 160 +++++-
>   .../thermal/mediatek,lvts-thermal.h           |  19 +
>   3 files changed, 631 insertions(+), 2 deletions(-)
> 
> base-commit: 6a3d37b4d885129561e1cef361216f00472f7d2e

Re: [PATCH v3 0/5]

[Matthias Brugger]

Hi Bernhard,

Please resend with a subject line for the cover letter.

Regards,
Matthias

On 29/05/2023 17:35, Bernhard Rosenkränzer wrote:
> From: Balsam CHIHI <bchihi@baylibre.com>
> 
> Add full LVTS support (MCU thermal domain + AP thermal domain) to MediaTek MT8192 SoC.
> Also, add Suspend and Resume support to LVTS Driver (all SoCs),
> and update the documentation that describes the Calibration Data Offsets.
> 
> Changelog:
>      v3 :
>          - Rebased :
>              base-commit: 6a3d37b4d885129561e1cef361216f00472f7d2e
>          - Fix issues in v2 pointed out by Nícolas F. R. A. Prado <nfraprado@collabora.com>:
>            Use filtered mode to make sure threshold interrupts are triggered,
>            protocol documentation, cosmetics
>          - I (bero@baylibre.com) will be taking care of this patchset
>            from now on, since Balsam has left BayLibre. Thanks for
>            getting it almost ready, Balsam!
> 
>      v2 :
>          - Based on top of thermal/linux-next :
>              base-commit: 7ac82227ee046f8234471de4c12a40b8c2d3ddcc
>          - Squash "add thermal zones and thermal nodes" and
>              "add temperature mitigation threshold" commits together to form
>              "arm64: dts: mediatek: mt8192: Add thermal nodes and thermal zones" commit.
>          - Add Suspend and Resume support to LVTS Driver.
>          - Update Calibration Data documentation.
>          - Fix calibration data offsets for mt8192
>              (Thanks to "Chen-Yu Tsai" and "Nícolas F. R. A. Prado").
>          https://lore.kernel.org/all/20230425133052.199767-1-bchihi@baylibre.com/
>          Tested-by: Chen-Yu Tsai <wenst@chromium.org>
> 
>      v1 :
>          - The initial series "Add LVTS support for mt8192" :
>              "https://lore.kernel.org/all/20230307163413.143334-1-bchihi@baylibre.com/".
> 
> Balsam CHIHI (5):
>    dt-bindings: thermal: mediatek: Add LVTS thermal controller definition
>      for mt8192
>    thermal/drivers/mediatek/lvts_thermal: Add suspend and resume
>    thermal/drivers/mediatek/lvts_thermal: Add mt8192 support
>    arm64: dts: mediatek: mt8192: Add thermal nodes and thermal zones
>    thermal/drivers/mediatek/lvts_thermal: Update calibration data
>      documentation
> 
>   arch/arm64/boot/dts/mediatek/mt8192.dtsi      | 454 ++++++++++++++++++
>   drivers/thermal/mediatek/lvts_thermal.c       | 160 +++++-
>   .../thermal/mediatek,lvts-thermal.h           |  19 +
>   3 files changed, 631 insertions(+), 2 deletions(-)
> 
> base-commit: 6a3d37b4d885129561e1cef361216f00472f7d2e

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH v3 0/5]

[Bernhard Rosenkränzer]

From: Balsam CHIHI <bchihi@baylibre.com>

Add full LVTS support (MCU thermal domain + AP thermal domain) to MediaTek MT8192 SoC.
Also, add Suspend and Resume support to LVTS Driver (all SoCs),
and update the documentation that describes the Calibration Data Offsets.

Changelog:
    v3 : 
        - Rebased :
            base-commit: 6a3d37b4d885129561e1cef361216f00472f7d2e
        - Fix issues in v2 pointed out by Nícolas F. R. A. Prado <nfraprado@collabora.com>:
          Use filtered mode to make sure threshold interrupts are triggered,
          protocol documentation, cosmetics
        - I (bero@baylibre.com) will be taking care of this patchset
          from now on, since Balsam has left BayLibre. Thanks for
          getting it almost ready, Balsam!

    v2 :
        - Based on top of thermal/linux-next :
            base-commit: 7ac82227ee046f8234471de4c12a40b8c2d3ddcc
        - Squash "add thermal zones and thermal nodes" and
            "add temperature mitigation threshold" commits together to form
            "arm64: dts: mediatek: mt8192: Add thermal nodes and thermal zones" commit.
        - Add Suspend and Resume support to LVTS Driver.
        - Update Calibration Data documentation.
        - Fix calibration data offsets for mt8192
            (Thanks to "Chen-Yu Tsai" and "Nícolas F. R. A. Prado").
        https://lore.kernel.org/all/20230425133052.199767-1-bchihi@baylibre.com/
        Tested-by: Chen-Yu Tsai <wenst@chromium.org>

    v1 :
        - The initial series "Add LVTS support for mt8192" :
            "https://lore.kernel.org/all/20230307163413.143334-1-bchihi@baylibre.com/".

Balsam CHIHI (5):
  dt-bindings: thermal: mediatek: Add LVTS thermal controller definition
    for mt8192
  thermal/drivers/mediatek/lvts_thermal: Add suspend and resume
  thermal/drivers/mediatek/lvts_thermal: Add mt8192 support
  arm64: dts: mediatek: mt8192: Add thermal nodes and thermal zones
  thermal/drivers/mediatek/lvts_thermal: Update calibration data
    documentation

 arch/arm64/boot/dts/mediatek/mt8192.dtsi      | 454 ++++++++++++++++++
 drivers/thermal/mediatek/lvts_thermal.c       | 160 +++++-
 .../thermal/mediatek,lvts-thermal.h           |  19 +
 3 files changed, 631 insertions(+), 2 deletions(-)

base-commit: 6a3d37b4d885129561e1cef361216f00472f7d2e
-- 
2.41.0.rc2

[PATCH v3 0/5]

[Bernhard Rosenkränzer]

From: Balsam CHIHI <bchihi@baylibre.com>

Add full LVTS support (MCU thermal domain + AP thermal domain) to MediaTek MT8192 SoC.
Also, add Suspend and Resume support to LVTS Driver (all SoCs),
and update the documentation that describes the Calibration Data Offsets.

Changelog:
    v3 : 
        - Rebased :
            base-commit: 6a3d37b4d885129561e1cef361216f00472f7d2e
        - Fix issues in v2 pointed out by Nícolas F. R. A. Prado <nfraprado@collabora.com>:
          Use filtered mode to make sure threshold interrupts are triggered,
          protocol documentation, cosmetics
        - I (bero@baylibre.com) will be taking care of this patchset
          from now on, since Balsam has left BayLibre. Thanks for
          getting it almost ready, Balsam!

    v2 :
        - Based on top of thermal/linux-next :
            base-commit: 7ac82227ee046f8234471de4c12a40b8c2d3ddcc
        - Squash "add thermal zones and thermal nodes" and
            "add temperature mitigation threshold" commits together to form
            "arm64: dts: mediatek: mt8192: Add thermal nodes and thermal zones" commit.
        - Add Suspend and Resume support to LVTS Driver.
        - Update Calibration Data documentation.
        - Fix calibration data offsets for mt8192
            (Thanks to "Chen-Yu Tsai" and "Nícolas F. R. A. Prado").
        https://lore.kernel.org/all/20230425133052.199767-1-bchihi@baylibre.com/
        Tested-by: Chen-Yu Tsai <wenst@chromium.org>

    v1 :
        - The initial series "Add LVTS support for mt8192" :
            "https://lore.kernel.org/all/20230307163413.143334-1-bchihi@baylibre.com/".

Balsam CHIHI (5):
  dt-bindings: thermal: mediatek: Add LVTS thermal controller definition
    for mt8192
  thermal/drivers/mediatek/lvts_thermal: Add suspend and resume
  thermal/drivers/mediatek/lvts_thermal: Add mt8192 support
  arm64: dts: mediatek: mt8192: Add thermal nodes and thermal zones
  thermal/drivers/mediatek/lvts_thermal: Update calibration data
    documentation

 arch/arm64/boot/dts/mediatek/mt8192.dtsi      | 454 ++++++++++++++++++
 drivers/thermal/mediatek/lvts_thermal.c       | 160 +++++-
 .../thermal/mediatek,lvts-thermal.h           |  19 +
 3 files changed, 631 insertions(+), 2 deletions(-)

base-commit: 6a3d37b4d885129561e1cef361216f00472f7d2e
-- 
2.41.0.rc2


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH v3 0/5]

[Teng Long]

Sorry for the late reply (in emergency project).

I reread the replies in patch v2 and fixed/optimized
in patch v3 in respective, I hope I haven't missed
anything.

Thanks.

Teng Long (5):
  pack-bitmap.c: continue looping when first MIDX bitmap is found
  pack-bitmap.c: rename "idx_name" to "bitmap_name"
  pack-bitmap.c: make warnings support i18N when opening bitmap
  pack-bitmap.c: using error() instead of silently returning -1
  bitmap: add trace2 outputs during open "bitmap" file

 pack-bitmap.c   | 58 +++++++++++++++++++++++++++++++------------------
 repo-settings.c |  1 +
 2 files changed, 38 insertions(+), 21 deletions(-)

Range-diff against v2:
1:  1bfd2fb6ab ! 1:  589e3f4075 pack-bitmap.c: continue looping when first MIDX bitmap is found
    @@ Commit message
         the first one has been found, then will break out by a "return"
         directly.
     
    -    But actually, it's better to don't stop the loop until we have visited
    +    But actually, it's better to continue the loop until we have visited
         both the MIDX in our repository, as well as any alternates (along with
         _their_ alternates, recursively).
     
    +    The reason for this is, there may exist more than one MIDX file in
    +    a repo. The "multi_pack_index" struct is actually designed as a singly
    +    linked list, and if a MIDX file has been already opened successfully,
    +    then the other MIDX files will be skipped and left with a warning
    +    "ignoring extra bitmap file." to the output.
    +
         The discussion link of community:
     
           https://public-inbox.org/git/YjzCTLLDCby+kJrZ@nand.local/
2:  1fff3b3ca7 ! 2:  b6b30047fc pack-bitmap.c: rename "idx_name" to "bitmap_name"
    @@ Commit message
     
         There may bring some confusion in this "idx_name" naming, which
         might lead us to think of ".idx "or" multi-pack-index" files,    -    although bitmap is essentially can be understood as a kind of index
    -    , let's define this name a little more accurate here.
    +    although bitmap is essentially can be understood as a kind of index,
    +    let's define this name a little more accurate here.
     
         Signed-off-by: Teng Long <dyroneteng@gmail.com>
     
3:  5a8f5afccf ! 3:  d8dfe53dd4 pack-bitmap.c: make warnings more detailed when opening bitmap
    @@ Metadata
     Author: Teng Long <dyroneteng@gmail.com>
     
      ## Commit message ##
    -    pack-bitmap.c: make warnings more detailed when opening bitmap
    +    pack-bitmap.c: make warnings support i18N when opening bitmap
     
         When calling the "open_midx_bitmap_1()" or "open_pack_bitmap_1()", there
         will be a warning if a normal bitmap or MIDX bitmap already has been
    -    opened, then let's make the warning information more detailed. For
    -    example, it makes the error clearer in case of an accidental
    -    regression where we start looking for single-pack bitmaps after
    -    successfully opening a multi-pack one.
    -
    -    At the same time, we made the previous and new warning texts support
    -    translation.
    +    opened, then let's make the warning text supporting for translation.
     
         Discussion in community:
     
    @@ pack-bitmap.c: static int open_midx_bitmap_1(struct bitmap_index *bitmap_git,
      		get_midx_filename(&buf, midx->object_dir);
      		/* ignore extra bitmap file; we can only handle one */
     -		warning("ignoring extra bitmap file: %s", buf.buf);
    -+		warning(_("a normal or midx bitmap already has been opened"));
     +		warning(_("ignoring extra bitmap file: %s"), buf.buf);
      		close(fd);
      		strbuf_release(&buf);
    @@ pack-bitmap.c: static int open_pack_bitmap_1(struct bitmap_index *bitmap_git, st
      	if (bitmap_git->pack || bitmap_git->midx) {
      		/* ignore extra bitmap file; we can only handle one */
     -		warning("ignoring extra bitmap file: %s", packfile->pack_name);
    -+		warning(_("a normal or midx bitmap already has been opened "));
     +		warning(_("ignoring extra bitmap file: %s"), packfile->pack_name);
      		close(fd);
      		return -1;
      	}
    -
    - ## t/t5310-pack-bitmaps.sh ##
    -@@ t/t5310-pack-bitmaps.sh: test_expect_success 'complains about multiple pack bitmaps' '
    - 		test_line_count = 2 bitmaps &&
    - 
    - 		git rev-list --use-bitmap-index HEAD 2>err &&
    -+		grep "a bitmap has been opened" err &&
    - 		grep "ignoring extra bitmap file" err
    - 	)
    - '
5:  1a169d7b5e ! 4:  72da3b5844 pack-bitmap.c: using error() instead of silently returning -1
    @@ Commit message
     
      ## pack-bitmap.c ##
     @@ pack-bitmap.c: static int open_midx_bitmap_1(struct bitmap_index *bitmap_git,
    - 		trace2_data_string("midx", the_repository, "stat bitmap file",
    - 				   "failed");
    + 
    + 	if (fstat(fd, &st)) {
      		close(fd);
     -		return -1;
    -+		return error("cannot stat bitmap file");
    ++		return error(_("cannot stat bitmap file"));
      	}
      
      	if (bitmap_git->pack || bitmap_git->midx) {
    @@ pack-bitmap.c: static int open_midx_bitmap_1(struct bitmap_index *bitmap_git,
      
      static int open_pack_bitmap_1(struct bitmap_index *bitmap_git, struct packed_git *packfile)
     @@ pack-bitmap.c: static int open_pack_bitmap_1(struct bitmap_index *bitmap_git, struct packed_git
    - 		trace2_data_string("bitmap", the_repository, "stat bitmap file",
    - 				   "failed");
    + 
    + 	if (fstat(fd, &st)) {
      		close(fd);
     -		return -1;
    -+		return error("cannot stat bitmap file");
    ++		return error(_("cannot stat bitmap file"));
      	}
      
      	if (bitmap_git->pack || bitmap_git->midx) {
     @@ pack-bitmap.c: static int open_pack_bitmap_1(struct bitmap_index *bitmap_git, struct packed_git
    + 
      	if (!is_pack_valid(packfile)) {
    - 		trace2_data_string("bitmap", the_repository, "packfile", "invalid");
      		close(fd);
     -		return -1;
    -+		return error("packfile is invalid");
    ++		return error(_("packfile is invalid"));
      	}
      
      	bitmap_git->pack = packfile;
    @@ pack-bitmap.c: static int open_pack_bitmap_1(struct bitmap_index *bitmap_git, st
      		bitmap_git->map_pos = 0;
      		bitmap_git->pack = NULL;
     -		return -1;
    -+		return error("bitmap header is invalid");
    ++		return error(_("bitmap header is invalid"));
      	}
      
      	return 0;
4:  2016ef2e34 ! 5:  e118758d1d bitmap: add trace2 outputs during open "bitmap" file
    @@ Metadata
      ## Commit message ##
         bitmap: add trace2 outputs during open "bitmap" file
     
    -    It's supported for a Git repo to use bitmap in both normal bitmap way
    -    or a multi-pack-index bitmap.
    -
    -    Sometimes the debug path is not obvious, for example, when executing:
    +    It's supported for a repo to use bitmap in both "NORMAL" bitmap way
    +    or a MIDX (multi-pack-index) bitmap. Either of two bitmap kinds can
    +    exist in the repository, or both can be stored but let the config
    +    controls which kind of bitmap is used (like "core.multipackIndex",
    +    etc.). Because of this, sometimes the bitmap debug path is not
    +    obvious enough, for example, when executing:
     
          $git rev-list  --test-bitmap  HEAD
          fatal: failed to load bitmap indexes
     
    -    If we see the output like this, we are not sure about what's happened,
    -    because the cause should be :
    +    If we see the output like this, It's not sure for us to know
    +    what's happened concretely, because the cause should be :
     
    -      1. neither normal nor midx bitmap exists
    -      2. only midx bitmap exists but core.multipackIndex="false"
    -      3. core.multipackIndex="true" but midx bitmap file is currupt
    -      4. core.multipackIndex="true" and no midx bitmap exists but
    -         normal bitmap file is currupt
    +      1. Neither normal nor MIDX bitmap exists.
    +      2. Only MIDX bitmap exists but core.multipackIndex="false".
    +      3. Config core.multipackIndex set to "true" but MIDX  bitmap is
    +         corrupted.
    +      4. Config core.multipackIndex set to "true" and no MIDX bitmap
    +         exists but normal bitmap file is corrupted.
           ....
     
         These are some of the scenarios I briefly tested, but maybe there are
    -    others (some scenarios is produced manually like "currupt bitmap file",
    +    others (some scenarios is produced manually like "corrupted bitmap file",
         but it's not represent it's an existed bug.).
     
         Therefore, we added some TRACE2 code so that when we read the bitmap
         we can be more clear about the decision path, such as whether it is
    -    working on midx bitmap or normal bitmap, or is it simply because the
    -    related configuration is disabled. This may help with logging, user
    -    troubleshooting, and development debugging.
    -
    -    Here are some output examples when executing
    -    "$GIT_TRACE2_PERF=1 git rev-list --test-bitmap HEAD" under different
    -    situations:
    -
    -    1. _HAVE_ MIDX bitmap and NORMAL bitmap file, but the
    -    "core.multipackIndex" configures as "false":
    -
    -      19:38:43.007840 common-main.c:49             | d0 | main                     | version      |     |           |           |              | 2.35.1.582.g8e9092487a
    -      19:38:43.007874 common-main.c:50             | d0 | main                     | start        |     |  0.000305 |           |              | /opt/git/master/bin/git rev-list --test-bitmap HEAD
    -      19:38:43.007950 compat/linux/procinfo.c:170  | d0 | main                     | cmd_ancestry |     |           |           |              | ancestry:[bash sshd sshd sshd systemd]
    -      19:38:43.008091 git.c:460                    | d0 | main                     | cmd_name     |     |           |           |              | rev-list (rev-list)
    -      19:38:43.008284 repo-settings.c:11           | d0 | main                     | data         | r0  |  0.000720 |  0.000720 | config       | feature.manyfiles:false
    -      19:38:43.008297 repo-settings.c:11           | d0 | main                     | data         | r0  |  0.000734 |  0.000734 | config       | feature.experimental:false
    -      19:38:43.008305 repo-settings.c:11           | d0 | main                     | data         | r0  |  0.000742 |  0.000742 | config       | core.commitgraph:true
    -      19:38:43.008313 repo-settings.c:11           | d0 | main                     | data         | r0  |  0.000749 |  0.000749 | config       | commitgraph.readchangedpaths:true
    -      19:38:43.008320 repo-settings.c:11           | d0 | main                     | data         | r0  |  0.000756 |  0.000756 | config       | gc.writecommitgraph:true
    -      19:38:43.008327 repo-settings.c:11           | d0 | main                     | data         | r0  |  0.000764 |  0.000764 | config       | fetch.writecommitgraph:false
    -      19:38:43.008334 repo-settings.c:11           | d0 | main                     | data         | r0  |  0.000770 |  0.000770 | config       | pack.usesparse:true
    -      19:38:43.008341 repo-settings.c:11           | d0 | main                     | data         | r0  |  0.000777 |  0.000777 | config       | core.multipackindex:false
    -      19:38:43.008348 repo-settings.c:11           | d0 | main                     | data         | r0  |  0.000784 |  0.000784 | config       | index.sparse:false
    -      19:38:43.008724 pack-bitmap.c:536            | d0 | main                     | region_enter | r0  |  0.001160 |           | pack-bitmap  | label:open_bitmap
    -      19:38:43.008738 pack-bitmap.c:390            | d0 | main                     | data         | r0  |  0.001173 |  0.000013 | bitmap       | ..try to open bitmap:/home/tenglong.tl/test/dyrone/.git/objects/pack/pack-2d1b3f749fb859b874710e33263d0847ef009e03.bitmap
    -      19:38:43.008754 pack-bitmap.c:543            | d0 | main                     | region_leave | r0  |  0.001191 |  0.000031 | pack-bitmap  | label:open_bitmap
    -      Bitmap v1 test (8 entries loaded)
    -      Found bitmap for 0a7df7ae92f8d8ed879c240f8ae9cdd33d18085e. 64 bits / 1801edc6 checksum
    -      19:38:43.009099 progress.c:268               | d0 | main                     | region_enter | r0  |  0.001535 |           | progress     | label:Verifying bitmap entries
    -      Verifying bitmap entries: 100% (27/27), done.
    -      19:38:43.009294 progress.c:339               | d0 | main                     | data         | r0  |  0.001730 |  0.000195 | progress     | ..total_objects:27
    -      19:38:43.009302 progress.c:346               | d0 | main                     | region_leave | r0  |  0.001739 |  0.000204 | progress     | label:Verifying bitmap entries
    -      OK!
    -      19:38:43.009321 git.c:718                    | d0 | main                     | exit         |     |  0.001757 |           |              | code:0
    -      19:38:43.009329 trace2/tr2_tgt_perf.c:215    | d0 | main                     | atexit       |     |  0.001766 |           |              | code:0
    -
    -    2. _HAVE_ MIDX bitmap and NORMAL bitmap file, but the
    -    "core.multipackIndex" configures as "true":
    -
    -      (omit duplicate outputs)
    -      ...
    -      20:02:31.288797 repo-settings.c:11           | d0 | main                     | data         | r0  |  0.000811 |  0.000811 | config       | core.multipackindex:true
    -      20:02:31.288806 repo-settings.c:11           | d0 | main                     | data         | r0  |  0.000819 |  0.000819 | config       | index.sparse:false
    -      20:02:31.288836 midx.c:185                   | d0 | main                     | data         | r0  |  0.000849 |  0.000849 | midx         | load/num_packs:1
    -      20:02:31.288843 midx.c:186                   | d0 | main                     | data         | r0  |  0.000857 |  0.000857 | midx         | load/num_objects:27
    -      20:02:31.289217 pack-bitmap.c:536            | d0 | main                     | region_enter | r0  |  0.001229 |           | pack-bitmap  | label:open_bitmap
    -      20:02:31.289230 pack-bitmap.c:318            | d0 | main                     | data         | r0  |  0.001242 |  0.000013 | midx         | ..try to open bitmap:/home/tenglong.tl/test/dyrone/.git/objects/pack/multi-pack-index-b6b04fbe053bd500d9ca13354466d3249dc275ac.bitmap
    -      20:02:31.289252 pack-revindex.c:315          | d0 | main                     | data         | r0  |  0.001265 |  0.000036 | load_midx_re | ..source:midx
    -      20:02:31.289261 pack-bitmap.c:543            | d0 | main                     | region_leave | r0  |  0.001274 |  0.000045 | pack-bitmap  | label:open_bitmap
    -      Bitmap v1 test (8 entries loaded)
    -      Found bitmap for 0a7df7ae92f8d8ed879c240f8ae9cdd33d18085e. 64 bits / 1801edc6 checksum
    -      20:02:31.289594 progress.c:268               | d0 | main                     | region_enter | r0  |  0.001607 |           | progress     | label:Verifying bitmap entries
    -      Verifying bitmap entries: 100% (27/27), done.
    -      20:02:31.289810 progress.c:339               | d0 | main                     | data         | r0  |  0.001823 |  0.000216 | progress     | ..total_objects:27
    -      20:02:31.289824 progress.c:346               | d0 | main                     | region_leave | r0  |  0.001837 |  0.000230 | progress     | label:Verifying bitmap entries
    -      OK!
    -      20:02:31.289843 git.c:718                    | d0 | main                     | exit         |     |  0.001856 |           |              | code:0
    -      20:02:31.289860 trace2/tr2_tgt_perf.c:215    | d0 | main                     | atexit       |     |  0.001873 |           |              | code:0
    -
    -    3. _HAVE_ MIDX bitmap and a corrupt NORMAL bitmap file, but the
    -    "core.multipackIndex" configures as "false" :
    -
    -      (omit duplicate outputs)
    -      ...
    -      20:14:06.539305 repo-settings.c:11           | d0 | main                     | data         | r0  |  0.000794 |  0.000794 | config       | core.multipackindex:false
    -      20:14:06.539310 repo-settings.c:11           | d0 | main                     | data         | r0  |  0.000799 |  0.000799 | config       | index.sparse:false
    -      20:14:06.539658 pack-bitmap.c:536            | d0 | main                     | region_enter | r0  |  0.001146 |           | pack-bitmap  | label:open_bitmap
    -      20:14:06.539671 pack-bitmap.c:390            | d0 | main                     | data         | r0  |  0.001160 |  0.000014 | bitmap       | ..try to open bitmap:/home/tenglong.tl/test/dyrone/.git/objects/pack/pack-2d1b3f749fb859b874710e33263d0847ef009e03.bitmap
    -      20:14:06.539686 usage.c:79                   | d0 | main                     | error        |     |           |           |              | ..Corrupted bitmap index file (wrong header)
    -      error: Corrupted bitmap index file (wrong header)
    -      20:14:06.539696 pack-bitmap.c:426            | d0 | main                     | data         | r0  |  0.001185 |  0.000039 | bitmap       | ..load bitmap header:failed
    -      20:14:06.539709 usage.c:79                   | d0 | main                     | error        |     |           |           |              | ..bitmap header is invalid
    -      error: bitmap header is invalid
    -      20:14:06.539719 pack-bitmap.c:543            | d0 | main                     | region_leave | r0  |  0.001208 |  0.000062 | pack-bitmap  | label:open_bitmap
    -      20:14:06.539726 usage.c:60                   | d0 | main                     | error        |     |           |           |              | failed to load bitmap indexes
    -      fatal: failed to load bitmap indexes
    -      20:14:06.539735 usage.c:74                   | d0 | main                     | exit         |     |  0.001224 |           |              | code:128
    -      20:14:06.539744 trace2/tr2_tgt_perf.c:215    | d0 | main                     | atexit       |     |  0.001233 |           |              | code:128
    +    working on MIDX or NORMAL bitmap at present, or the related config is
    +    enabled or not. This may help with logging, user troubleshooting, and
    +    development debugging.
    +
    +    Here are some brief output examples on different scenarios when
    +    executing:
    +
    +      $GIT_TRACE2_PERF=1 git rev-list --test-bitmap HEAD
    +
    +    Scenario 1: core.multipackIndex [false], midx bitmap exists [Y],
    +    normal bitmap exists [N]
    +
    +           19:21:56.580349 repo-settings.c:11           | d0 | main                     | data         | r1  |  0.000827 |  0.000827 | config       | core.multipackindex:false
    +           19:21:56.580356 repo-settings.c:11           | d0 | main                     | data         | r1  |  0.000834 |  0.000834 | config       | index.sparse:false
    +           19:21:56.580706 pack-bitmap.c:525            | d0 | main                     | region_enter | r1  |  0.001183 |           | pack-bitmap  | label:open_bitmap
    +           19:21:56.580719 pack-bitmap.c:386            | d0 | main                     | data         | r1  |  0.001196 |  0.000013 | bitmap       | ..path:.git/objects/pack/pack-e9eb18e6a423057f4424a762069e13804a75d01e.bitmap
    +           19:21:56.580729 pack-bitmap.c:530            | d0 | main                     | region_leave | r1  |  0.001207 |  0.000024 | pack-bitmap  | label:open_bitmap
    +           19:21:56.580737 usage.c:60                   | d0 | main                     | error        |     |           |           |              | failed to load bitmap indexes
    +           fatal: failed to load bitmap indexes
    +           19:21:56.580746 usage.c:74                   | d0 | main                     | exit         |     |  0.001224 |           |              | code:128
    +           19:21:56.580754 trace2/tr2_tgt_perf.c:215    | d0 | main                     | atexit       |     |  0.001232 |           |              | code:128
    +
    +    Scenario 2: core.multipackIndex [false], midx bitmap exists [Y],
    +    normal bitmap exists [Y]
    +
    +            19:23:44.692384 repo-settings.c:11           | d0 | main                     | data         | r0  |  0.000765 |  0.000765 | config       | core.multipackindex:false
    +            19:23:44.692755 pack-bitmap.c:525            | d0 | main                     | region_enter | r0  |  0.001135 |           | pack-bitmap  | label:open_bitmap
    +            19:23:44.692768 pack-bitmap.c:386            | d0 | main                     | data         | r0  |  0.001149 |  0.000014 | bitmap       | ..path:/home/tenglong.tl/test/dyrone_bitmap/.git/objects/pack/pack-e9eb18e6a423057f4424a762069e13804a75d01e.bitmap
    +            19:23:44.692790 pack-bitmap.c:530            | d0 | main                     | region_leave | r0  |  0.001171 |  0.000036 | pack-bitmap  | label:open_bitmap
    +            Bitmap v1 test (1 entries loaded)
    +            Found bitmap for d864fefa87415d6cd289c72aa9ffd45b4a8ffd84. 64 bits / 11030517 checksum
    +            19:23:44.693119 progress.c:268               | d0 | main                     | region_enter | r0  |  0.001500 |           | progress     | label:Verifying bitmap entries
    +            Verifying bitmap entries: 100% (3/3), done.
    +            19:23:44.693208 progress.c:339               | d0 | main                     | data         | r0  |  0.001589 |  0.000089 | progress     | ..total_objects:3
    +            19:23:44.693216 progress.c:346               | d0 | main                     | region_leave | r0  |  0.001597 |  0.000097 | progress     | label:Verifying bitmap entries
    +            OK!
    +            19:23:44.693234 git.c:718                    | d0 | main                     | exit         |     |  0.001615 |           |              | code:0
    +            19:23:44.693244 trace2/tr2_tgt_perf.c:215    | d0 | main                     | atexit       |     |  0.001625 |           |              | code:0
    +
    +    Scenario 3: core.multipackIndex [true], midx bitmap exists [Y],
    +    normal bitmap exists [Y]
    +
    +            19:26:03.625055 repo-settings.c:11           | d0 | main                     | data         | r0  |  0.000760 |  0.000760 | config       | core.multipackindex:true
    +            19:26:03.625090 midx.c:185                   | d0 | main                     | data         | r0  |  0.000795 |  0.000795 | midx         | load/num_packs:1
    +            19:26:03.625097 midx.c:186                   | d0 | main                     | data         | r0  |  0.000803 |  0.000803 | midx         | load/num_objects:3
    +            19:26:03.625455 pack-bitmap.c:525            | d0 | main                     | region_enter | r0  |  0.001160 |           | pack-bitmap  | label:open_bitmap
    +            19:26:03.625470 pack-bitmap.c:318            | d0 | main                     | data         | r0  |  0.001175 |  0.000015 | midx         | ..path:/home/tenglong.tl/test/dyrone_bitmap/.git/objects/pack/multi-pack-index-fe8e96790bd34926423bdf3efd762dbbea9f3213.bitmap
    +            19:26:03.625489 pack-revindex.c:315          | d0 | main                     | data         | r0  |  0.001194 |  0.000034 | load_midx_re | ..source:midx
    +            19:26:03.625496 pack-bitmap.c:530            | d0 | main                     | region_leave | r0  |  0.001202 |  0.000042 | pack-bitmap  | label:open_bitmap
    +            Bitmap v1 test (1 entries loaded)
    +            Found bitmap for d864fefa87415d6cd289c72aa9ffd45b4a8ffd84. 64 bits / 11030517 checksum
    +            19:26:03.625818 progress.c:268               | d0 | main                     | region_enter | r0  |  0.001523 |           | progress     | label:Verifying bitmap entries
    +            Verifying bitmap entries: 100% (3/3), done.
    +            19:26:03.625916 progress.c:339               | d0 | main                     | data         | r0  |  0.001622 |  0.000099 | progress     | ..total_objects:3
    +            19:26:03.625925 progress.c:346               | d0 | main                     | region_leave | r0  |  0.001630 |  0.000107 | progress     | label:Verifying bitmap entries
    +            OK!
    +            19:26:03.625943 git.c:718                    | d0 | main                     | exit         |     |  0.001648 |           |              | code:0
    +            19:26:03.625952 trace2/tr2_tgt_perf.c:215    | d0 | main                     | atexit       |     |  0.001658 |           |              | code:0
    +
    +    Situation 4: core.multipackIndex [false], midx bitmap exists [N],
    +    normal bitmap exists [Y].
    +
    +            19:27:15.383037 repo-settings.c:11           | d0 | main                     | data         | r0  |  0.000746 |  0.000746 | config       | core.multipackindex:true
    +            19:27:15.383397 pack-bitmap.c:525            | d0 | main                     | region_enter | r0  |  0.001105 |           | pack-bitmap  | label:open_bitmap
    +            19:27:15.383408 pack-bitmap.c:386            | d0 | main                     | data         | r0  |  0.001116 |  0.000011 | bitmap       | ..path:/home/tenglong.tl/test/dyrone_bitmap/.git/objects/pack/pack-e9eb18e6a423057f4424a762069e13804a75d01e.bitmap
    +            19:27:15.383419 pack-bitmap.c:530            | d0 | main                     | region_leave | r0  |  0.001128 |  0.000023 | pack-bitmap  | label:open_bitmap
    +            Bitmap v1 test (1 entries loaded)
    +            Found bitmap for d864fefa87415d6cd289c72aa9ffd45b4a8ffd84. 64 bits / 11030517 checksum
    +            19:27:15.383730 progress.c:268               | d0 | main                     | region_enter | r0  |  0.001439 |           | progress     | label:Verifying bitmap entries
    +            Verifying bitmap entries: 100% (3/3), done.
    +            19:27:15.383822 progress.c:339               | d0 | main                     | data         | r0  |  0.001531 |  0.000092 | progress     | ..total_objects:3
    +            19:27:15.383830 progress.c:346               | d0 | main                     | region_leave | r0  |  0.001539 |  0.000100 | progress     | label:Verifying bitmap entries
    +            OK!
    +            19:27:15.383848 git.c:718                    | d0 | main                     | exit         |     |  0.001557 |           |              | code:0
    +            19:27:15.383867 trace2/tr2_tgt_perf.c:215    | d0 | main                     | atexit       |     |  0.001576 |           |              | code:0
    +
    +    Scenario 5: core.multipackIndex [true], midx bitmap exists [Y] but corrupted,
    +    normal bitmap exists [Y]
    +
    +            19:29:25.888233 repo-settings.c:11           | d0 | main                     | data         | r0  |  0.000794 |  0.000794 | config       | core.multipackindex:true
    +            19:29:25.888591 pack-bitmap.c:525            | d0 | main                     | region_enter | r0  |  0.001152 |           | pack-bitmap  | label:open_bitmap
    +            19:29:25.888603 pack-bitmap.c:386            | d0 | main                     | data         | r0  |  0.001163 |  0.000011 | bitmap       | ..path:/home/tenglong.tl/test/dyrone_bitmap/.git/objects/pack/pack-e9eb18e6a423057f4424a762069e13804a75d01e.bitmap
    +            19:29:25.888622 usage.c:79                   | d0 | main                     | error        |     |           |           |              | ..Corrupted bitmap index file (wrong header)
    +            error: Corrupted bitmap index file (wrong header)
    +            19:29:25.888638 usage.c:79                   | d0 | main                     | error        |     |           |           |              | ..bitmap header is invalid
    +            error: bitmap header is invalid
    +            19:29:25.888650 pack-bitmap.c:530            | d0 | main                     | region_leave | r0  |  0.001211 |  0.000059 | pack-bitmap  | label:open_bitmap
    +            19:29:25.888659 usage.c:60                   | d0 | main                     | error        |     |           |           |              | failed to load bitmap indexes
    +            fatal: failed to load bitmap indexes
    +            19:29:25.888670 usage.c:74                   | d0 | main                     | exit         |     |  0.001231 |           |              | code:128
    +            19:29:25.888680 trace2/tr2_tgt_perf.c:215    | d0 | main                     | atexit       |     |  0.001241 |           |              | code:128
     
         Signed-off-by: Teng Long <dyroneteng@gmail.com>
     
    @@ pack-bitmap.c: char *pack_bitmap_filename(struct packed_git *p)
      	struct stat st;
      	char *bitmap_name = midx_bitmap_filename(midx);
     -	int fd = git_open(bitmap_name);
    -+	trace2_data_string("midx", the_repository, "try to open bitmap",
    -+			   bitmap_name);
    ++	trace2_data_string("midx", the_repository, "path", bitmap_name);
     +	fd = git_open(bitmap_name);
      
      	free(bitmap_name);
      
    -@@ pack-bitmap.c: static int open_midx_bitmap_1(struct bitmap_index *bitmap_git,
    - 		return -1;
    - 
    - 	if (fstat(fd, &st)) {
    -+		trace2_data_string("midx", the_repository, "stat bitmap file",
    -+				   "failed");
    - 		close(fd);
    - 		return -1;
    - 	}
     @@ pack-bitmap.c: static int open_midx_bitmap_1(struct bitmap_index *bitmap_git,
      				MAP_PRIVATE, fd, 0);
      	close(fd);
    @@ pack-bitmap.c: static int open_pack_bitmap_1(struct bitmap_index *bitmap_git, st
      		return -1;
      
      	bitmap_name = pack_bitmap_filename(packfile);
    -+	trace2_data_string("bitmap", the_repository, "try to open bitmap",
    -+			   bitmap_name);
    ++	trace2_data_string("bitmap", the_repository, "path", bitmap_name);
      	fd = git_open(bitmap_name);
      	free(bitmap_name);
      
    -@@ pack-bitmap.c: static int open_pack_bitmap_1(struct bitmap_index *bitmap_git, struct packed_git
    - 		return -1;
    - 
    - 	if (fstat(fd, &st)) {
    -+		trace2_data_string("bitmap", the_repository, "stat bitmap file",
    -+				   "failed");
    - 		close(fd);
    - 		return -1;
    - 	}
    -@@ pack-bitmap.c: static int open_pack_bitmap_1(struct bitmap_index *bitmap_git, struct packed_git
    - 	}
    - 
    - 	if (!is_pack_valid(packfile)) {
    -+		trace2_data_string("bitmap", the_repository, "packfile", "invalid");
    - 		close(fd);
    - 		return -1;
    - 	}
    -@@ pack-bitmap.c: static int open_pack_bitmap_1(struct bitmap_index *bitmap_git, struct packed_git
    - 	close(fd);
    - 
    - 	if (load_bitmap_header(bitmap_git) < 0) {
    -+		trace2_data_string("bitmap", the_repository,
    -+				   "load bitmap header", "failed");
    - 		munmap(bitmap_git->map, bitmap_git->map_size);
    - 		bitmap_git->map = NULL;
    - 		bitmap_git->map_size = 0;
     @@ pack-bitmap.c: static int open_midx_bitmap(struct repository *r,
      static int open_bitmap(struct repository *r,
      		       struct bitmap_index *bitmap_git)
      {
     -	assert(!bitmap_git->map);
    -+	int ret = -1;
    ++	int ret = 0;
      
    --	if (!open_midx_bitmap(r, bitmap_git))
    --		return 0;
    --	return open_pack_bitmap(r, bitmap_git);
     +	assert(!bitmap_git->map);
     +	trace2_region_enter("pack-bitmap", "open_bitmap", r);
    -+	if (!open_midx_bitmap(r, bitmap_git)) {
    -+		ret = 0;
    + 	if (!open_midx_bitmap(r, bitmap_git))
    +-		return 0;
    +-	return open_pack_bitmap(r, bitmap_git);
     +		goto done;
    -+	}
     +	ret = open_pack_bitmap(r, bitmap_git);
     +done:
     +	trace2_region_leave("pack-bitmap", "open_bitmap", r);
    @@ repo-settings.c: static void repo_cfg_bool(struct repository *r, const char *key
      }
      
      void prepare_repo_settings(struct repository *r)
    -
    - ## t/t5310-pack-bitmaps.sh ##
    -@@ t/t5310-pack-bitmaps.sh: test_expect_success 'complains about multiple pack bitmaps' '
    - 		test_line_count = 2 bitmaps &&
    - 
    - 		git rev-list --use-bitmap-index HEAD 2>err &&
    --		grep "a bitmap has been opened" err &&
    --		grep "ignoring extra bitmap file" err
    -+		grep "warning: a normal or midx bitmap already has been opened" err &&
    -+		grep "warning: ignoring extra bitmap file" err
    - 	)
    - '
    - 
-- 
2.35.1.582.g320e881567

Re: [PATCH v3 0/5]

[Andrew Lunn]

On Fri, Feb 08, 2019 at 05:11:56PM +0100, Federico Vaga wrote:
> This patch set provides improvements to the i2c-ocore driver.

Hi Federico

Please could you fixup the SPDX patch, add my review/tested by tags,
and i think we are good to go.

Thanks
	Andrew

[PATCH v3 0/5]

[Federico Vaga]

This patch set provides improvements to the i2c-ocore driver.

[V2 -> V3]
- fix error condition on platform_get_irq(). Copied from
  https://patchwork.ozlabs.org/patch/1038409/

[V1 -> V2]
- replaced usleep_range() with udelay() so that the polling version can be
  used in atomic context.
- added dedicated patch for minor style issues
- fixed delay computation
- use spin_lock_irqsave(), instead of spin_trylock_irqsave(). IACK is always
  necessary and a trylock would generate an extra interrupt for nothing
- make the driver ready for an eventual master_xfer_irqless()

[PATCH v3 0/5]

[Federico Vaga]

This patch set provides improvements to the i2c-ocore driver.

[V2 -> V3]
- fix error condition on platform_get_irq(). Copied from
  https://patchwork.ozlabs.org/patch/1038409/

[V1 -> V2]
- replaced usleep_range() with udelay() so that the polling version can be
  used in atomic context.
- added dedicated patch for minor style issues
- fixed delay computation
- use spin_lock_irqsave(), instead of spin_trylock_irqsave(). IACK is always
  necessary and a trylock would generate an extra interrupt for nothing
- make the driver ready for an eventual master_xfer_irqless()

[PATCH v3 0/5]

[Taeung Song]

Changes in v3:
	- builtin-config.c: Add a config variable ’kmem.default’ with a default value into ‘struct default_configset’
	which has default config variables and values.
	- cmd_config(): Add a option ‘—global’ and ‘—local’ to enable  config file location to  be selected

Changes in v2:
	- Renaming variables a more suitable name
	  1. ’—list-all' instead of '--all'
	  2. ’name' instead of 'subkey'
	  3. 'section, name, value' instead of 'given_section,subkey,value'
	- Documentation/perf-config.txt: Correct small infelicities or typing errors in a perf-config documention.
	- Documentation/perf-config.txt: Remove a part description of report.children because it was duplicated
	in Documentation/callchain-overhead-calculation.txt
	- builtin-config.c: Use a variable ’int actions’ instead of struct params which has ‘bool list_action’,
	‘bool get_action’ and etc. , to simplify a branching statement for perf-config options 
	- builtin-config.c: Declaration a global variable ‘static struct default_configsets’ has config variables
	with default values instead of using a 'util/PERFCONFIG-DEFAULT' file and remove functions merge()
	and perse_key() to get perf config default values.
	- normalize_value(): Add a function to normalize a value and check data type of it.
	- cmd_config(): Simplify parsing arguments as arguments is just divided by '=' and then in front of '.' is a section,
	between '.' and '=' is a name, and behind '=' is a value.
	- show_all_config(): Print config variables ‘struct default_configsets’ haven't 
	- cmd_config(): Make a command ’perf config' without a option work as with a option ’—list’ instead of ‘—list-all’.

Taeung Song (5):
  perf tools: Add 'perf-config' command
  perf config: Add functions which can get or set perf config variables.
  perf config: Add a option 'list-all' to perf-config.
  perf config: Add a option 'remove' to perf-config.
  perf config: Add ‘—system’ and ‘—global’ options to be able to select
    which config file to be used.

 tools/perf/Build                            |   1 +
 tools/perf/Documentation/perf-config.txt    | 401 ++++++++++++++++
 tools/perf/Documentation/perfconfig.example |  70 ++-
 tools/perf/builtin-config.c                 | 708 ++++++++++++++++++++++++++++
 tools/perf/builtin.h                        |   1 +
 tools/perf/command-list.txt                 |   1 +
 tools/perf/perf.c                           |   1 +
 tools/perf/util/cache.h                     |  20 +
 tools/perf/util/config.c                    |  84 +++-
 9 files changed, 1251 insertions(+), 36 deletions(-)
 create mode 100644 tools/perf/Documentation/perf-config.txt
 create mode 100644 tools/perf/builtin-config.c

-- 
1.9.1

[PATCH v3 0/5]

[Yunzhi Li]

Patches to add support for Rockchip usb phys.Add a new Rockchip
usb phy driver and modify dwc2 controller driver to make dwc2
platform devices support a generic PHY framework driver. This
patch set has been tested on my rk3288-evb and power off the usb
phys would reduce about 60mW power budget in total during sustem
suspend.

Changes in v3:
- Use BIT macro instead of bit shift ops.
- Rename the config entry to PHY_ROCKCHIP_USB.
- Fix coding style: both branches of the if() which only one
  branch of the conditional statement is a single statement should
  have braces.
- No need to test dwc2->phy for NULL before calling generic phy
  APIs.
- Add more context about the changes in the long description.

Yunzhi Li (5):
  phy: add a driver for the Rockchip SoC internal USB2.0 PHY
  Documentation: bindings: add doc for the Rockchip usb PHY
  usb: dwc2: add generic PHY framework support for dwc2 usb    
    controler platform driver.
  ARM: dts: add rk3288 usb PHY
  ARM: dts: Enable usb PHY on rk3288-evb board

 .../devicetree/bindings/phy/rockchip-usb-phy.txt   |  22 +++
 arch/arm/boot/dts/rk3288-evb.dtsi                  |   4 +
 arch/arm/boot/dts/rk3288.dtsi                      |  13 ++
 drivers/phy/Kconfig                                |   7 +
 drivers/phy/Makefile                               |   1 +
 drivers/phy/phy-rockchip-usb.c                     | 179 +++++++++++++++++++++
 drivers/usb/dwc2/gadget.c                          |  33 ++--
 drivers/usb/dwc2/platform.c                        |  36 ++++-
 8 files changed, 272 insertions(+), 23 deletions(-)
 create mode 100644 Documentation/devicetree/bindings/phy/rockchip-usb-phy.txt
 create mode 100644 drivers/phy/phy-rockchip-usb.c

-- 
2.0.0

[PATCH v3 0/5]

[Yunzhi Li]

Patches to add support for Rockchip usb phys.Add a new Rockchip
usb phy driver and modify dwc2 controller driver to make dwc2
platform devices support a generic PHY framework driver. This
patch set has been tested on my rk3288-evb and power off the usb
phys would reduce about 60mW power budget in total during sustem
suspend.

Changes in v3:
- Use BIT macro instead of bit shift ops.
- Rename the config entry to PHY_ROCKCHIP_USB.
- Fix coding style: both branches of the if() which only one
  branch of the conditional statement is a single statement should
  have braces.
- No need to test dwc2->phy for NULL before calling generic phy
  APIs.
- Add more context about the changes in the long description.

Yunzhi Li (5):
  phy: add a driver for the Rockchip SoC internal USB2.0 PHY
  Documentation: bindings: add doc for the Rockchip usb PHY
  usb: dwc2: add generic PHY framework support for dwc2 usb    
    controler platform driver.
  ARM: dts: add rk3288 usb PHY
  ARM: dts: Enable usb PHY on rk3288-evb board

 .../devicetree/bindings/phy/rockchip-usb-phy.txt   |  22 +++
 arch/arm/boot/dts/rk3288-evb.dtsi                  |   4 +
 arch/arm/boot/dts/rk3288.dtsi                      |  13 ++
 drivers/phy/Kconfig                                |   7 +
 drivers/phy/Makefile                               |   1 +
 drivers/phy/phy-rockchip-usb.c                     | 179 +++++++++++++++++++++
 drivers/usb/dwc2/gadget.c                          |  33 ++--
 drivers/usb/dwc2/platform.c                        |  36 ++++-
 8 files changed, 272 insertions(+), 23 deletions(-)
 create mode 100644 Documentation/devicetree/bindings/phy/rockchip-usb-phy.txt
 create mode 100644 drivers/phy/phy-rockchip-usb.c

-- 
2.0.0

[PATCH v3 0/5]

[Linus Walleij]

On Thu, Jun 5, 2014 at 3:25 PM, Maxime Ripard
<maxime.ripard@free-electrons.com> wrote:

> This is an attempt at making the external interrupts on the A31 and
> the A23 working.

I've applied these 5 patches to my devel branch, but you won't
see this on linux-next until after the merge window is closed.

Yours,
Linus Walleij

[PATCH v3 0/5]

[Maxime Ripard]

Hi everyone,

This is an attempt at making the external interrupts on the A31 and
the A23 working.

The main difference with the previous code is that there's now several
banks and several parent interrupts, instead of a single bank with a
single parent interrupt in the older SoCs.

Most of the diffstat is to register the new irq pin functions. The
real code additions are those in the core pinctrl driver.

Maxime

Changes from v2:
   - Fixed the bogus the irq_cfg_reg function that was not taking into
     account the fact that we have several registers for this one.
   - Remove the SUNXI_IRQ_NUMBER, since it was not really a good name
     anymore, and replaced it for IRQ_PER_BANK.
   - Removed meaningless and broken sunxi_irq_cfg_reg_from_bank function.

Changes from v1:
   - Bail out of the interrupt handler in case of a spurious interrupt
   - Fix the wrong register offset in sunxi_irq_*_reg functions.

Maxime Ripard (5):
  pinctrl: sunxi: Remove irq_mask_ack and use irq_ack instead
  pinctrl: sunxi: Add macro definition for pinctrl with more than one
    interrupt
  pinctrl: sunxi: Declare the number of interrupt banks in the
    descriptor
  pinctrl: sunxi: Declare the interrupt function for the A31
  pinctrl: sunxi: Implement multiple interrupt banks support

 drivers/pinctrl/sunxi/pinctrl-sun4i-a10.c   |   1 +
 drivers/pinctrl/sunxi/pinctrl-sun5i-a10s.c  |   1 +
 drivers/pinctrl/sunxi/pinctrl-sun5i-a13.c   |   1 +
 drivers/pinctrl/sunxi/pinctrl-sun6i-a31-r.c |   1 +
 drivers/pinctrl/sunxi/pinctrl-sun6i-a31.c   | 217 +++++++++++++++++++---------
 drivers/pinctrl/sunxi/pinctrl-sun7i-a20.c   |   1 +
 drivers/pinctrl/sunxi/pinctrl-sunxi.c       |  81 +++++++----
 drivers/pinctrl/sunxi/pinctrl-sunxi.h       |  44 ++++--
 8 files changed, 238 insertions(+), 109 deletions(-)

-- 
2.0.0

[PATCH v3 0/5]

[Alex Chiang]

This is v3 of the series.

I based it off of Linus's latest tree. 

I did not include David Rientjes's "mm: slab allocate memory section nodemask
for large systems" patch in my series, since it's not necessarily related.

Please consider for inclusion for the next merge window (v2.6.33).

Thanks,
/ac

v2 -> v3:
	- rebased to Linus's latest tree (799dd75b)
	- Added David Rientjes's Acked-by: flags
	- dropped S390 cc's, since they are unaffected by this series

v1 -> v2: http://thread.gmane.org/gmane.linux.kernel.mm/40084/
        Address David Rientjes's comments
        - check return value of sysfs_create_link in register_cpu_under_node
        - do /not/ convert [un]register_cpu_under_node to return void, since
          sparse starts whinging if you ignore sysfs_create_link()'s return
          value and working around sparse makes the code ugly
        - adjust documentation

        Added S390 maintainers to cc: for patch [1/5] as per Kame-san's
        suggestion. S390 may map a memory section to more than one node,
        causing this series to break.


---

Alex Chiang (5):
      mm: add numa node symlink for memory section in sysfs
      mm: refactor register_cpu_under_node()
      mm: refactor unregister_cpu_under_node()
      mm: add numa node symlink for cpu devices in sysfs
      Documentation: ABI: /sys/devices/system/cpu/cpu#/node


 Documentation/ABI/testing/sysfs-devices-memory     |   14 ++++-
 Documentation/ABI/testing/sysfs-devices-system-cpu |   14 +++++
 Documentation/memory-hotplug.txt                   |   11 ++--
 drivers/base/node.c                                |   58 ++++++++++++++------
 4 files changed, 76 insertions(+), 21 deletions(-)

[PATCH v3 0/5]

[Alex Chiang]

This is v3 of the series.

I based it off of Linus's latest tree. 

I did not include David Rientjes's "mm: slab allocate memory section nodemask
for large systems" patch in my series, since it's not necessarily related.

Please consider for inclusion for the next merge window (v2.6.33).

Thanks,
/ac

v2 -> v3:
	- rebased to Linus's latest tree (799dd75b)
	- Added David Rientjes's Acked-by: flags
	- dropped S390 cc's, since they are unaffected by this series

v1 -> v2: http://thread.gmane.org/gmane.linux.kernel.mm/40084/
        Address David Rientjes's comments
        - check return value of sysfs_create_link in register_cpu_under_node
        - do /not/ convert [un]register_cpu_under_node to return void, since
          sparse starts whinging if you ignore sysfs_create_link()'s return
          value and working around sparse makes the code ugly
        - adjust documentation

        Added S390 maintainers to cc: for patch [1/5] as per Kame-san's
        suggestion. S390 may map a memory section to more than one node,
        causing this series to break.


---

Alex Chiang (5):
      mm: add numa node symlink for memory section in sysfs
      mm: refactor register_cpu_under_node()
      mm: refactor unregister_cpu_under_node()
      mm: add numa node symlink for cpu devices in sysfs
      Documentation: ABI: /sys/devices/system/cpu/cpu#/node


 Documentation/ABI/testing/sysfs-devices-memory     |   14 ++++-
 Documentation/ABI/testing/sysfs-devices-system-cpu |   14 +++++
 Documentation/memory-hotplug.txt                   |   11 ++--
 drivers/base/node.c                                |   58 ++++++++++++++------
 4 files changed, 76 insertions(+), 21 deletions(-)

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>