diff mbox

[v6,02/21] KVM: ARM64: Define PMU data structure for each vcpu

Message ID 1449578860-15808-3-git-send-email-zhaoshenglong@huawei.com (mailing list archive)
State New, archived
Headers show

Commit Message

Shannon Zhao Dec. 8, 2015, 12:47 p.m. UTC
From: Shannon Zhao <shannon.zhao@linaro.org>

Here we plan to support virtual PMU for guest by full software
emulation, so define some basic structs and functions preparing for
futher steps. Define struct kvm_pmc for performance monitor counter and
struct kvm_pmu for performance monitor unit for each vcpu. According to
ARMv8 spec, the PMU contains at most 32(ARMV8_MAX_COUNTERS) counters.

Since this only supports ARM64 (or PMUv3), add a separate config symbol
for it.

Signed-off-by: Shannon Zhao <shannon.zhao@linaro.org>
---
 arch/arm64/include/asm/kvm_host.h |  2 ++
 arch/arm64/kvm/Kconfig            |  8 ++++++++
 include/kvm/arm_pmu.h             | 40 +++++++++++++++++++++++++++++++++++++++
 3 files changed, 50 insertions(+)
 create mode 100644 include/kvm/arm_pmu.h

Comments

Marc Zyngier Dec. 8, 2015, 1:37 p.m. UTC | #1
On 08/12/15 12:47, Shannon Zhao wrote:
> From: Shannon Zhao <shannon.zhao@linaro.org>
> 
> Here we plan to support virtual PMU for guest by full software
> emulation, so define some basic structs and functions preparing for
> futher steps. Define struct kvm_pmc for performance monitor counter and
> struct kvm_pmu for performance monitor unit for each vcpu. According to
> ARMv8 spec, the PMU contains at most 32(ARMV8_MAX_COUNTERS) counters.
> 
> Since this only supports ARM64 (or PMUv3), add a separate config symbol
> for it.
> 
> Signed-off-by: Shannon Zhao <shannon.zhao@linaro.org>
> ---
>  arch/arm64/include/asm/kvm_host.h |  2 ++
>  arch/arm64/kvm/Kconfig            |  8 ++++++++
>  include/kvm/arm_pmu.h             | 40 +++++++++++++++++++++++++++++++++++++++
>  3 files changed, 50 insertions(+)
>  create mode 100644 include/kvm/arm_pmu.h
> 
> diff --git a/arch/arm64/include/asm/kvm_host.h b/arch/arm64/include/asm/kvm_host.h
> index a35ce72..42e15bb 100644
> --- a/arch/arm64/include/asm/kvm_host.h
> +++ b/arch/arm64/include/asm/kvm_host.h
> @@ -37,6 +37,7 @@
>  
>  #include <kvm/arm_vgic.h>
>  #include <kvm/arm_arch_timer.h>
> +#include <kvm/arm_pmu.h>
>  
>  #define KVM_MAX_VCPUS VGIC_V3_MAX_CPUS
>  
> @@ -132,6 +133,7 @@ struct kvm_vcpu_arch {
>  	/* VGIC state */
>  	struct vgic_cpu vgic_cpu;
>  	struct arch_timer_cpu timer_cpu;
> +	struct kvm_pmu pmu;
>  
>  	/*
>  	 * Anything that is not used directly from assembly code goes
> diff --git a/arch/arm64/kvm/Kconfig b/arch/arm64/kvm/Kconfig
> index a5272c0..66da9a2 100644
> --- a/arch/arm64/kvm/Kconfig
> +++ b/arch/arm64/kvm/Kconfig
> @@ -36,6 +36,7 @@ config KVM
>  	select HAVE_KVM_EVENTFD
>  	select HAVE_KVM_IRQFD
>  	select KVM_ARM_VGIC_V3
> +	select KVM_ARM_PMU

What if HW_PERF_EVENTS is not selected? Also, selecting HW_PERF_EVENTS
is not enough, and you probably need PERF_EVENTS as well, So this should
probably read:

	select KVM_ARM_PMU if (HW_PERF_EVENTS && PERF_EVENTS)

>  	---help---
>  	  Support hosting virtualized guest machines.
>  	  We don't support KVM with 16K page tables yet, due to the multiple
> @@ -48,6 +49,13 @@ config KVM_ARM_HOST
>  	---help---
>  	  Provides host support for ARM processors.
>  
> +config KVM_ARM_PMU
> +	bool
> +	depends on KVM_ARM_HOST && HW_PERF_EVENTS

and this line should be dropped.

> +	---help---
> +	  Adds support for a virtual Performance Monitoring Unit (PMU) in
> +	  virtual machines.
> +
>  source drivers/vhost/Kconfig
>  
>  endif # VIRTUALIZATION
> diff --git a/include/kvm/arm_pmu.h b/include/kvm/arm_pmu.h
> new file mode 100644
> index 0000000..dea78f8
> --- /dev/null
> +++ b/include/kvm/arm_pmu.h
> @@ -0,0 +1,40 @@
> +/*
> + * Copyright (C) 2015 Linaro Ltd.
> + * Author: Shannon Zhao <shannon.zhao@linaro.org>
> + *
> + * 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_ARM_KVM_PMU_H
> +#define __ASM_ARM_KVM_PMU_H
> +
> +#include <linux/perf_event.h>
> +#ifdef CONFIG_KVM_ARM_PMU
> +#include <asm/pmu.h>
> +#endif
> +
> +struct kvm_pmc {
> +	u8 idx;/* index into the pmu->pmc array */
> +	struct perf_event *perf_event;
> +	u64 bitmask;
> +};
> +
> +struct kvm_pmu {
> +#ifdef CONFIG_KVM_ARM_PMU
> +	/* PMU IRQ Number per VCPU */
> +	int irq_num;
> +	struct kvm_pmc pmc[ARMV8_MAX_COUNTERS];
> +#endif
> +};
> +
> +#endif
> 

The repetition of #ifdef CONFIG_KVM_ARM_PMU is a bit ugly. How about
something like this instead:

#ifndef __ASM_ARM_KVM_PMU_H
#define __ASM_ARM_KVM_PMU_H

#ifdef CONFIG_KVM_ARM_PMU

#include <linux/perf_event.h>
#include <asm/pmu.h>

struct kvm_pmc {
	u8 idx;/* index into the pmu->pmc array */
	struct perf_event *perf_event;
	u64 bitmask;
};

struct kvm_pmu {
	/* PMU IRQ Number per VCPU */
	int irq_num;
	struct kvm_pmc pmc[ARMV8_MAX_COUNTERS];
};

#else

struct kvm_pmu {
};

#endif

#endif

and you can then populate the rest of the function prototype and stubs
where needed.

Thanks,

	M.
Will Deacon Dec. 8, 2015, 1:53 p.m. UTC | #2
On Tue, Dec 08, 2015 at 01:37:14PM +0000, Marc Zyngier wrote:
> On 08/12/15 12:47, Shannon Zhao wrote:
> > From: Shannon Zhao <shannon.zhao@linaro.org>
> > 
> > Here we plan to support virtual PMU for guest by full software
> > emulation, so define some basic structs and functions preparing for
> > futher steps. Define struct kvm_pmc for performance monitor counter and
> > struct kvm_pmu for performance monitor unit for each vcpu. According to
> > ARMv8 spec, the PMU contains at most 32(ARMV8_MAX_COUNTERS) counters.
> > 
> > Since this only supports ARM64 (or PMUv3), add a separate config symbol
> > for it.
> > 
> > Signed-off-by: Shannon Zhao <shannon.zhao@linaro.org>
> > ---
> >  arch/arm64/include/asm/kvm_host.h |  2 ++
> >  arch/arm64/kvm/Kconfig            |  8 ++++++++
> >  include/kvm/arm_pmu.h             | 40 +++++++++++++++++++++++++++++++++++++++
> >  3 files changed, 50 insertions(+)
> >  create mode 100644 include/kvm/arm_pmu.h

[...]

> > diff --git a/arch/arm64/kvm/Kconfig b/arch/arm64/kvm/Kconfig
> > index a5272c0..66da9a2 100644
> > --- a/arch/arm64/kvm/Kconfig
> > +++ b/arch/arm64/kvm/Kconfig
> > @@ -36,6 +36,7 @@ config KVM
> >  	select HAVE_KVM_EVENTFD
> >  	select HAVE_KVM_IRQFD
> >  	select KVM_ARM_VGIC_V3
> > +	select KVM_ARM_PMU
> 
> What if HW_PERF_EVENTS is not selected? Also, selecting HW_PERF_EVENTS
> is not enough, and you probably need PERF_EVENTS as well, So this should
> probably read:
> 
> 	select KVM_ARM_PMU if (HW_PERF_EVENTS && PERF_EVENTS)

HW_PERF_EVENTS depends on ARM_PMU which in turn depends on PERF_EVENTS.

Will
Marc Zyngier Dec. 8, 2015, 2:10 p.m. UTC | #3
On 08/12/15 13:53, Will Deacon wrote:
> On Tue, Dec 08, 2015 at 01:37:14PM +0000, Marc Zyngier wrote:
>> On 08/12/15 12:47, Shannon Zhao wrote:
>>> From: Shannon Zhao <shannon.zhao@linaro.org>
>>>
>>> Here we plan to support virtual PMU for guest by full software
>>> emulation, so define some basic structs and functions preparing for
>>> futher steps. Define struct kvm_pmc for performance monitor counter and
>>> struct kvm_pmu for performance monitor unit for each vcpu. According to
>>> ARMv8 spec, the PMU contains at most 32(ARMV8_MAX_COUNTERS) counters.
>>>
>>> Since this only supports ARM64 (or PMUv3), add a separate config symbol
>>> for it.
>>>
>>> Signed-off-by: Shannon Zhao <shannon.zhao@linaro.org>
>>> ---
>>>  arch/arm64/include/asm/kvm_host.h |  2 ++
>>>  arch/arm64/kvm/Kconfig            |  8 ++++++++
>>>  include/kvm/arm_pmu.h             | 40 +++++++++++++++++++++++++++++++++++++++
>>>  3 files changed, 50 insertions(+)
>>>  create mode 100644 include/kvm/arm_pmu.h
> 
> [...]
> 
>>> diff --git a/arch/arm64/kvm/Kconfig b/arch/arm64/kvm/Kconfig
>>> index a5272c0..66da9a2 100644
>>> --- a/arch/arm64/kvm/Kconfig
>>> +++ b/arch/arm64/kvm/Kconfig
>>> @@ -36,6 +36,7 @@ config KVM
>>>  	select HAVE_KVM_EVENTFD
>>>  	select HAVE_KVM_IRQFD
>>>  	select KVM_ARM_VGIC_V3
>>> +	select KVM_ARM_PMU
>>
>> What if HW_PERF_EVENTS is not selected? Also, selecting HW_PERF_EVENTS
>> is not enough, and you probably need PERF_EVENTS as well, So this should
>> probably read:
>>
>> 	select KVM_ARM_PMU if (HW_PERF_EVENTS && PERF_EVENTS)
> 
> HW_PERF_EVENTS depends on ARM_PMU which in turn depends on PERF_EVENTS.

in which case, let's make it:

	select KVM_ARM_PMU if HW_PERF_EVENTS

which should give us the minimal chain. I hate the kernel config
language! ;-)

	M.
Shannon Zhao Dec. 8, 2015, 2:14 p.m. UTC | #4
On 2015/12/8 22:10, Marc Zyngier wrote:
> On 08/12/15 13:53, Will Deacon wrote:
>> On Tue, Dec 08, 2015 at 01:37:14PM +0000, Marc Zyngier wrote:
>>> On 08/12/15 12:47, Shannon Zhao wrote:
>>>> From: Shannon Zhao <shannon.zhao@linaro.org>
>>>>
>>>> Here we plan to support virtual PMU for guest by full software
>>>> emulation, so define some basic structs and functions preparing for
>>>> futher steps. Define struct kvm_pmc for performance monitor counter and
>>>> struct kvm_pmu for performance monitor unit for each vcpu. According to
>>>> ARMv8 spec, the PMU contains at most 32(ARMV8_MAX_COUNTERS) counters.
>>>>
>>>> Since this only supports ARM64 (or PMUv3), add a separate config symbol
>>>> for it.
>>>>
>>>> Signed-off-by: Shannon Zhao <shannon.zhao@linaro.org>
>>>> ---
>>>>   arch/arm64/include/asm/kvm_host.h |  2 ++
>>>>   arch/arm64/kvm/Kconfig            |  8 ++++++++
>>>>   include/kvm/arm_pmu.h             | 40 +++++++++++++++++++++++++++++++++++++++
>>>>   3 files changed, 50 insertions(+)
>>>>   create mode 100644 include/kvm/arm_pmu.h
>>
>> [...]
>>
>>>> diff --git a/arch/arm64/kvm/Kconfig b/arch/arm64/kvm/Kconfig
>>>> index a5272c0..66da9a2 100644
>>>> --- a/arch/arm64/kvm/Kconfig
>>>> +++ b/arch/arm64/kvm/Kconfig
>>>> @@ -36,6 +36,7 @@ config KVM
>>>>   	select HAVE_KVM_EVENTFD
>>>>   	select HAVE_KVM_IRQFD
>>>>   	select KVM_ARM_VGIC_V3
>>>> +	select KVM_ARM_PMU
>>>
>>> What if HW_PERF_EVENTS is not selected? Also, selecting HW_PERF_EVENTS
>>> is not enough, and you probably need PERF_EVENTS as well, So this should
>>> probably read:
>>>
>>> 	select KVM_ARM_PMU if (HW_PERF_EVENTS && PERF_EVENTS)
>>
>> HW_PERF_EVENTS depends on ARM_PMU which in turn depends on PERF_EVENTS.
>
Yeah, this is the reason why I choose HW_PERF_EVENTS.

> in which case, let's make it:
>
> 	select KVM_ARM_PMU if HW_PERF_EVENTS
>
Sure, will do.

> which should give us the minimal chain. I hate the kernel config
> language! ;-)
>
> 	M.
>
diff mbox

Patch

diff --git a/arch/arm64/include/asm/kvm_host.h b/arch/arm64/include/asm/kvm_host.h
index a35ce72..42e15bb 100644
--- a/arch/arm64/include/asm/kvm_host.h
+++ b/arch/arm64/include/asm/kvm_host.h
@@ -37,6 +37,7 @@ 
 
 #include <kvm/arm_vgic.h>
 #include <kvm/arm_arch_timer.h>
+#include <kvm/arm_pmu.h>
 
 #define KVM_MAX_VCPUS VGIC_V3_MAX_CPUS
 
@@ -132,6 +133,7 @@  struct kvm_vcpu_arch {
 	/* VGIC state */
 	struct vgic_cpu vgic_cpu;
 	struct arch_timer_cpu timer_cpu;
+	struct kvm_pmu pmu;
 
 	/*
 	 * Anything that is not used directly from assembly code goes
diff --git a/arch/arm64/kvm/Kconfig b/arch/arm64/kvm/Kconfig
index a5272c0..66da9a2 100644
--- a/arch/arm64/kvm/Kconfig
+++ b/arch/arm64/kvm/Kconfig
@@ -36,6 +36,7 @@  config KVM
 	select HAVE_KVM_EVENTFD
 	select HAVE_KVM_IRQFD
 	select KVM_ARM_VGIC_V3
+	select KVM_ARM_PMU
 	---help---
 	  Support hosting virtualized guest machines.
 	  We don't support KVM with 16K page tables yet, due to the multiple
@@ -48,6 +49,13 @@  config KVM_ARM_HOST
 	---help---
 	  Provides host support for ARM processors.
 
+config KVM_ARM_PMU
+	bool
+	depends on KVM_ARM_HOST && HW_PERF_EVENTS
+	---help---
+	  Adds support for a virtual Performance Monitoring Unit (PMU) in
+	  virtual machines.
+
 source drivers/vhost/Kconfig
 
 endif # VIRTUALIZATION
diff --git a/include/kvm/arm_pmu.h b/include/kvm/arm_pmu.h
new file mode 100644
index 0000000..dea78f8
--- /dev/null
+++ b/include/kvm/arm_pmu.h
@@ -0,0 +1,40 @@ 
+/*
+ * Copyright (C) 2015 Linaro Ltd.
+ * Author: Shannon Zhao <shannon.zhao@linaro.org>
+ *
+ * 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_ARM_KVM_PMU_H
+#define __ASM_ARM_KVM_PMU_H
+
+#include <linux/perf_event.h>
+#ifdef CONFIG_KVM_ARM_PMU
+#include <asm/pmu.h>
+#endif
+
+struct kvm_pmc {
+	u8 idx;/* index into the pmu->pmc array */
+	struct perf_event *perf_event;
+	u64 bitmask;
+};
+
+struct kvm_pmu {
+#ifdef CONFIG_KVM_ARM_PMU
+	/* PMU IRQ Number per VCPU */
+	int irq_num;
+	struct kvm_pmc pmc[ARMV8_MAX_COUNTERS];
+#endif
+};
+
+#endif