[PATCH v8] libperf: Add arm64 support to perf_mmap__read_self()

[PATCH v8] libperf: Add arm64 support to perf_mmap__read_self()

[Rob Herring]

Add the arm64 variants for read_perf_counter() and read_timestamp().
Unfortunately the counter number is encoded into the instruction, so the
code is a bit verbose to enumerate all possible counters.

Signed-off-by: Rob Herring <robh@kernel.org>
---
v8:
 - Set attr.config1 to request user access on arm64
v7:
 - Move enabling of libperf user read test for arm64 to this patch
---
 tools/lib/perf/mmap.c             | 98 +++++++++++++++++++++++++++++++
 tools/lib/perf/tests/test-evsel.c |  5 +-
 2 files changed, 102 insertions(+), 1 deletion(-)

diff --git a/tools/lib/perf/mmap.c b/tools/lib/perf/mmap.c
index c89dfa5f67b3..7ee3eb9f5e27 100644
--- a/tools/lib/perf/mmap.c
+++ b/tools/lib/perf/mmap.c
@@ -13,6 +13,7 @@
 #include <internal/lib.h>
 #include <linux/kernel.h>
 #include <linux/math64.h>
+#include <linux/stringify.h>
 #include "internal.h"
 
 void perf_mmap__init(struct perf_mmap *map, struct perf_mmap *prev,
@@ -294,6 +295,103 @@ static u64 read_timestamp(void)
 
 	return low | ((u64)high) << 32;
 }
+#elif defined(__aarch64__)
+#define read_sysreg(r) ({						\
+	u64 __val;							\
+	asm volatile("mrs %0, " __stringify(r) : "=r" (__val));		\
+	__val;								\
+})
+
+static u64 read_pmccntr(void)
+{
+	return read_sysreg(pmccntr_el0);
+}
+
+#define PMEVCNTR_READ(idx)					\
+	static u64 read_pmevcntr_##idx(void) {			\
+		return read_sysreg(pmevcntr##idx##_el0);	\
+	}
+
+PMEVCNTR_READ(0);
+PMEVCNTR_READ(1);
+PMEVCNTR_READ(2);
+PMEVCNTR_READ(3);
+PMEVCNTR_READ(4);
+PMEVCNTR_READ(5);
+PMEVCNTR_READ(6);
+PMEVCNTR_READ(7);
+PMEVCNTR_READ(8);
+PMEVCNTR_READ(9);
+PMEVCNTR_READ(10);
+PMEVCNTR_READ(11);
+PMEVCNTR_READ(12);
+PMEVCNTR_READ(13);
+PMEVCNTR_READ(14);
+PMEVCNTR_READ(15);
+PMEVCNTR_READ(16);
+PMEVCNTR_READ(17);
+PMEVCNTR_READ(18);
+PMEVCNTR_READ(19);
+PMEVCNTR_READ(20);
+PMEVCNTR_READ(21);
+PMEVCNTR_READ(22);
+PMEVCNTR_READ(23);
+PMEVCNTR_READ(24);
+PMEVCNTR_READ(25);
+PMEVCNTR_READ(26);
+PMEVCNTR_READ(27);
+PMEVCNTR_READ(28);
+PMEVCNTR_READ(29);
+PMEVCNTR_READ(30);
+
+/*
+ * Read a value direct from PMEVCNTR<idx>
+ */
+static u64 read_perf_counter(unsigned int counter)
+{
+	static u64 (* const read_f[])(void) = {
+		read_pmevcntr_0,
+		read_pmevcntr_1,
+		read_pmevcntr_2,
+		read_pmevcntr_3,
+		read_pmevcntr_4,
+		read_pmevcntr_5,
+		read_pmevcntr_6,
+		read_pmevcntr_7,
+		read_pmevcntr_8,
+		read_pmevcntr_9,
+		read_pmevcntr_10,
+		read_pmevcntr_11,
+		read_pmevcntr_13,
+		read_pmevcntr_12,
+		read_pmevcntr_14,
+		read_pmevcntr_15,
+		read_pmevcntr_16,
+		read_pmevcntr_17,
+		read_pmevcntr_18,
+		read_pmevcntr_19,
+		read_pmevcntr_20,
+		read_pmevcntr_21,
+		read_pmevcntr_22,
+		read_pmevcntr_23,
+		read_pmevcntr_24,
+		read_pmevcntr_25,
+		read_pmevcntr_26,
+		read_pmevcntr_27,
+		read_pmevcntr_28,
+		read_pmevcntr_29,
+		read_pmevcntr_30,
+		read_pmccntr
+	};
+
+	if (counter < ARRAY_SIZE(read_f))
+		return (read_f[counter])();
+
+	return 0;
+}
+
+static u64 read_timestamp(void) { return read_sysreg(cntvct_el0); }
+
 #else
 static u64 read_perf_counter(unsigned int counter __maybe_unused) { return 0; }
 static u64 read_timestamp(void) { return 0; }
diff --git a/tools/lib/perf/tests/test-evsel.c b/tools/lib/perf/tests/test-evsel.c
index 33ae9334861a..89be89afb24d 100644
--- a/tools/lib/perf/tests/test-evsel.c
+++ b/tools/lib/perf/tests/test-evsel.c
@@ -130,6 +130,9 @@ static int test_stat_user_read(int event)
 	struct perf_event_attr attr = {
 		.type	= PERF_TYPE_HARDWARE,
 		.config	= event,
+#ifdef __aarch64__
+		.config1 = 0x2,		/* Request user access */
+#endif
 	};
 	int err, i;
 
@@ -150,7 +153,7 @@ static int test_stat_user_read(int event)
 	pc = perf_evsel__mmap_base(evsel, 0, 0);
 	__T("failed to get mmapped address", pc);
 
-#if defined(__i386__) || defined(__x86_64__)
+#if defined(__i386__) || defined(__x86_64__) || defined(__aarch64__)
 	__T("userspace counter access not supported", pc->cap_user_rdpmc);
 	__T("userspace counter access not enabled", pc->index);
 	__T("userspace counter width not set", pc->pmc_width >= 32);
-- 
2.32.0

Re: [PATCH v8] libperf: Add arm64 support to perf_mmap__read_self()

[Rob Herring]

On Tue, Dec 14, 2021 at 3:56 PM Rob Herring <robh@kernel.org> wrote:
>
> Add the arm64 variants for read_perf_counter() and read_timestamp().
> Unfortunately the counter number is encoded into the instruction, so the
> code is a bit verbose to enumerate all possible counters.
>
> Signed-off-by: Rob Herring <robh@kernel.org>
> ---
> v8:
>  - Set attr.config1 to request user access on arm64

Hit send a bit too quick. The kernel side[1] has now been applied by
Will, so hopefully the libperf enabling can land in 5.17 too.

Rob

[1] https://lore.kernel.org/all/163948145564.2060329.17036263207999776606.b4-ty@kernel.org/

Re: [PATCH v8] libperf: Add arm64 support to perf_mmap__read_self()

[Rob Herring]

On Tue, Dec 14, 2021 at 3:56 PM Rob Herring <robh@kernel.org> wrote:
>
> Add the arm64 variants for read_perf_counter() and read_timestamp().
> Unfortunately the counter number is encoded into the instruction, so the
> code is a bit verbose to enumerate all possible counters.
>
> Signed-off-by: Rob Herring <robh@kernel.org>
> ---
> v8:
>  - Set attr.config1 to request user access on arm64
> v7:
>  - Move enabling of libperf user read test for arm64 to this patch
> ---
>  tools/lib/perf/mmap.c             | 98 +++++++++++++++++++++++++++++++
>  tools/lib/perf/tests/test-evsel.c |  5 +-
>  2 files changed, 102 insertions(+), 1 deletion(-)

Ping!

>
> diff --git a/tools/lib/perf/mmap.c b/tools/lib/perf/mmap.c
> index c89dfa5f67b3..7ee3eb9f5e27 100644
> --- a/tools/lib/perf/mmap.c
> +++ b/tools/lib/perf/mmap.c
> @@ -13,6 +13,7 @@
>  #include <internal/lib.h>
>  #include <linux/kernel.h>
>  #include <linux/math64.h>
> +#include <linux/stringify.h>
>  #include "internal.h"
>
>  void perf_mmap__init(struct perf_mmap *map, struct perf_mmap *prev,
> @@ -294,6 +295,103 @@ static u64 read_timestamp(void)
>
>         return low | ((u64)high) << 32;
>  }
> +#elif defined(__aarch64__)
> +#define read_sysreg(r) ({                                              \
> +       u64 __val;                                                      \
> +       asm volatile("mrs %0, " __stringify(r) : "=r" (__val));         \
> +       __val;                                                          \
> +})
> +
> +static u64 read_pmccntr(void)
> +{
> +       return read_sysreg(pmccntr_el0);
> +}
> +
> +#define PMEVCNTR_READ(idx)                                     \
> +       static u64 read_pmevcntr_##idx(void) {                  \
> +               return read_sysreg(pmevcntr##idx##_el0);        \
> +       }
> +
> +PMEVCNTR_READ(0);
> +PMEVCNTR_READ(1);
> +PMEVCNTR_READ(2);
> +PMEVCNTR_READ(3);
> +PMEVCNTR_READ(4);
> +PMEVCNTR_READ(5);
> +PMEVCNTR_READ(6);
> +PMEVCNTR_READ(7);
> +PMEVCNTR_READ(8);
> +PMEVCNTR_READ(9);
> +PMEVCNTR_READ(10);
> +PMEVCNTR_READ(11);
> +PMEVCNTR_READ(12);
> +PMEVCNTR_READ(13);
> +PMEVCNTR_READ(14);
> +PMEVCNTR_READ(15);
> +PMEVCNTR_READ(16);
> +PMEVCNTR_READ(17);
> +PMEVCNTR_READ(18);
> +PMEVCNTR_READ(19);
> +PMEVCNTR_READ(20);
> +PMEVCNTR_READ(21);
> +PMEVCNTR_READ(22);
> +PMEVCNTR_READ(23);
> +PMEVCNTR_READ(24);
> +PMEVCNTR_READ(25);
> +PMEVCNTR_READ(26);
> +PMEVCNTR_READ(27);
> +PMEVCNTR_READ(28);
> +PMEVCNTR_READ(29);
> +PMEVCNTR_READ(30);
> +
> +/*
> + * Read a value direct from PMEVCNTR<idx>
> + */
> +static u64 read_perf_counter(unsigned int counter)
> +{
> +       static u64 (* const read_f[])(void) = {
> +               read_pmevcntr_0,
> +               read_pmevcntr_1,
> +               read_pmevcntr_2,
> +               read_pmevcntr_3,
> +               read_pmevcntr_4,
> +               read_pmevcntr_5,
> +               read_pmevcntr_6,
> +               read_pmevcntr_7,
> +               read_pmevcntr_8,
> +               read_pmevcntr_9,
> +               read_pmevcntr_10,
> +               read_pmevcntr_11,
> +               read_pmevcntr_13,
> +               read_pmevcntr_12,
> +               read_pmevcntr_14,
> +               read_pmevcntr_15,
> +               read_pmevcntr_16,
> +               read_pmevcntr_17,
> +               read_pmevcntr_18,
> +               read_pmevcntr_19,
> +               read_pmevcntr_20,
> +               read_pmevcntr_21,
> +               read_pmevcntr_22,
> +               read_pmevcntr_23,
> +               read_pmevcntr_24,
> +               read_pmevcntr_25,
> +               read_pmevcntr_26,
> +               read_pmevcntr_27,
> +               read_pmevcntr_28,
> +               read_pmevcntr_29,
> +               read_pmevcntr_30,
> +               read_pmccntr
> +       };
> +
> +       if (counter < ARRAY_SIZE(read_f))
> +               return (read_f[counter])();
> +
> +       return 0;
> +}
> +
> +static u64 read_timestamp(void) { return read_sysreg(cntvct_el0); }
> +
>  #else
>  static u64 read_perf_counter(unsigned int counter __maybe_unused) { return 0; }
>  static u64 read_timestamp(void) { return 0; }
> diff --git a/tools/lib/perf/tests/test-evsel.c b/tools/lib/perf/tests/test-evsel.c
> index 33ae9334861a..89be89afb24d 100644
> --- a/tools/lib/perf/tests/test-evsel.c
> +++ b/tools/lib/perf/tests/test-evsel.c
> @@ -130,6 +130,9 @@ static int test_stat_user_read(int event)
>         struct perf_event_attr attr = {
>                 .type   = PERF_TYPE_HARDWARE,
>                 .config = event,
> +#ifdef __aarch64__
> +               .config1 = 0x2,         /* Request user access */
> +#endif
>         };
>         int err, i;
>
> @@ -150,7 +153,7 @@ static int test_stat_user_read(int event)
>         pc = perf_evsel__mmap_base(evsel, 0, 0);
>         __T("failed to get mmapped address", pc);
>
> -#if defined(__i386__) || defined(__x86_64__)
> +#if defined(__i386__) || defined(__x86_64__) || defined(__aarch64__)
>         __T("userspace counter access not supported", pc->cap_user_rdpmc);
>         __T("userspace counter access not enabled", pc->index);
>         __T("userspace counter width not set", pc->pmc_width >= 32);
> --
> 2.32.0
>

Re: [PATCH v8] libperf: Add arm64 support to perf_mmap__read_self()

[Masayoshi Mizuma]

On Tue, Dec 14, 2021 at 03:56:30PM -0600, Rob Herring wrote:
> Add the arm64 variants for read_perf_counter() and read_timestamp().
> Unfortunately the counter number is encoded into the instruction, so the
> code is a bit verbose to enumerate all possible counters.
> 
> Signed-off-by: Rob Herring <robh@kernel.org>

This patch works well on the aarch64 machine, thanks!
Please feel free to add:
Tested-by: Masayoshi Mizuma <m.mizuma@jp.fujitsu.com>

the log of make tests:
```
  # echo 1 > /proc/sys/kernel/perf_user_access
  # make tests V=1
  make -f /src/linux/tools/build/Makefile.build dir=. obj=libperf
  make -C /src/linux/tools/lib/api/ O= libapi.a
  make -f /src/linux/tools/build/Makefile.build dir=./fd obj=libapi
  make -f /src/linux/tools/build/Makefile.build dir=./fs obj=libapi
  make -f /src/linux/tools/build/Makefile.build dir=. obj=tests
  make -f /src/linux/tools/build/Makefile.build dir=./tests obj=tests
  running static:
  - running tests/test-cpumap.c...OK
  - running tests/test-threadmap.c...OK
  - running tests/test-evlist.c...OK
  - running tests/test-evsel.c...
          loop = 65536, count = 328191
          loop = 131072, count = 655878
          loop = 262144, count = 1311204
          loop = 524288, count = 2640123
          loop = 1048576, count = 5243358
          loop = 65536, count = 1290820
          loop = 131072, count = 2509179
          loop = 262144, count = 5011308
          loop = 524288, count = 9999221
          loop = 1048576, count = 20136809
  OK
  running dynamic:
  - running tests/test-cpumap.c...OK
  - running tests/test-threadmap.c...OK
  - running tests/test-evlist.c...OK
  - running tests/test-evsel.c...
          loop = 65536, count = 328219
          loop = 131072, count = 655856
          loop = 262144, count = 1311199
          loop = 524288, count = 2633704
          loop = 1048576, count = 5243402
          loop = 65536, count = 1303460
          loop = 131072, count = 2513614
          loop = 262144, count = 5020097
          loop = 524288, count = 10043687
          loop = 1048576, count = 20101337
  OK
  #
```

- Masa

> ---
> v8:
>  - Set attr.config1 to request user access on arm64
> v7:
>  - Move enabling of libperf user read test for arm64 to this patch
> ---
>  tools/lib/perf/mmap.c             | 98 +++++++++++++++++++++++++++++++
>  tools/lib/perf/tests/test-evsel.c |  5 +-
>  2 files changed, 102 insertions(+), 1 deletion(-)
> 
> diff --git a/tools/lib/perf/mmap.c b/tools/lib/perf/mmap.c
> index c89dfa5f67b3..7ee3eb9f5e27 100644
> --- a/tools/lib/perf/mmap.c
> +++ b/tools/lib/perf/mmap.c
> @@ -13,6 +13,7 @@
>  #include <internal/lib.h>
>  #include <linux/kernel.h>
>  #include <linux/math64.h>
> +#include <linux/stringify.h>
>  #include "internal.h"
>  
>  void perf_mmap__init(struct perf_mmap *map, struct perf_mmap *prev,
> @@ -294,6 +295,103 @@ static u64 read_timestamp(void)
>  
>  	return low | ((u64)high) << 32;
>  }
> +#elif defined(__aarch64__)
> +#define read_sysreg(r) ({						\
> +	u64 __val;							\
> +	asm volatile("mrs %0, " __stringify(r) : "=r" (__val));		\
> +	__val;								\
> +})
> +
> +static u64 read_pmccntr(void)
> +{
> +	return read_sysreg(pmccntr_el0);
> +}
> +
> +#define PMEVCNTR_READ(idx)					\
> +	static u64 read_pmevcntr_##idx(void) {			\
> +		return read_sysreg(pmevcntr##idx##_el0);	\
> +	}
> +
> +PMEVCNTR_READ(0);
> +PMEVCNTR_READ(1);
> +PMEVCNTR_READ(2);
> +PMEVCNTR_READ(3);
> +PMEVCNTR_READ(4);
> +PMEVCNTR_READ(5);
> +PMEVCNTR_READ(6);
> +PMEVCNTR_READ(7);
> +PMEVCNTR_READ(8);
> +PMEVCNTR_READ(9);
> +PMEVCNTR_READ(10);
> +PMEVCNTR_READ(11);
> +PMEVCNTR_READ(12);
> +PMEVCNTR_READ(13);
> +PMEVCNTR_READ(14);
> +PMEVCNTR_READ(15);
> +PMEVCNTR_READ(16);
> +PMEVCNTR_READ(17);
> +PMEVCNTR_READ(18);
> +PMEVCNTR_READ(19);
> +PMEVCNTR_READ(20);
> +PMEVCNTR_READ(21);
> +PMEVCNTR_READ(22);
> +PMEVCNTR_READ(23);
> +PMEVCNTR_READ(24);
> +PMEVCNTR_READ(25);
> +PMEVCNTR_READ(26);
> +PMEVCNTR_READ(27);
> +PMEVCNTR_READ(28);
> +PMEVCNTR_READ(29);
> +PMEVCNTR_READ(30);
> +
> +/*
> + * Read a value direct from PMEVCNTR<idx>
> + */
> +static u64 read_perf_counter(unsigned int counter)
> +{
> +	static u64 (* const read_f[])(void) = {
> +		read_pmevcntr_0,
> +		read_pmevcntr_1,
> +		read_pmevcntr_2,
> +		read_pmevcntr_3,
> +		read_pmevcntr_4,
> +		read_pmevcntr_5,
> +		read_pmevcntr_6,
> +		read_pmevcntr_7,
> +		read_pmevcntr_8,
> +		read_pmevcntr_9,
> +		read_pmevcntr_10,
> +		read_pmevcntr_11,
> +		read_pmevcntr_13,
> +		read_pmevcntr_12,
> +		read_pmevcntr_14,
> +		read_pmevcntr_15,
> +		read_pmevcntr_16,
> +		read_pmevcntr_17,
> +		read_pmevcntr_18,
> +		read_pmevcntr_19,
> +		read_pmevcntr_20,
> +		read_pmevcntr_21,
> +		read_pmevcntr_22,
> +		read_pmevcntr_23,
> +		read_pmevcntr_24,
> +		read_pmevcntr_25,
> +		read_pmevcntr_26,
> +		read_pmevcntr_27,
> +		read_pmevcntr_28,
> +		read_pmevcntr_29,
> +		read_pmevcntr_30,
> +		read_pmccntr
> +	};
> +
> +	if (counter < ARRAY_SIZE(read_f))
> +		return (read_f[counter])();
> +
> +	return 0;
> +}
> +
> +static u64 read_timestamp(void) { return read_sysreg(cntvct_el0); }
> +
>  #else
>  static u64 read_perf_counter(unsigned int counter __maybe_unused) { return 0; }
>  static u64 read_timestamp(void) { return 0; }
> diff --git a/tools/lib/perf/tests/test-evsel.c b/tools/lib/perf/tests/test-evsel.c
> index 33ae9334861a..89be89afb24d 100644
> --- a/tools/lib/perf/tests/test-evsel.c
> +++ b/tools/lib/perf/tests/test-evsel.c
> @@ -130,6 +130,9 @@ static int test_stat_user_read(int event)
>  	struct perf_event_attr attr = {
>  		.type	= PERF_TYPE_HARDWARE,
>  		.config	= event,
> +#ifdef __aarch64__
> +		.config1 = 0x2,		/* Request user access */
> +#endif
>  	};
>  	int err, i;
>  
> @@ -150,7 +153,7 @@ static int test_stat_user_read(int event)
>  	pc = perf_evsel__mmap_base(evsel, 0, 0);
>  	__T("failed to get mmapped address", pc);
>  
> -#if defined(__i386__) || defined(__x86_64__)
> +#if defined(__i386__) || defined(__x86_64__) || defined(__aarch64__)
>  	__T("userspace counter access not supported", pc->cap_user_rdpmc);
>  	__T("userspace counter access not enabled", pc->index);
>  	__T("userspace counter width not set", pc->pmc_width >= 32);
> -- 
> 2.32.0
> 

Re: [PATCH v8] libperf: Add arm64 support to perf_mmap__read_self()

[Rob Herring]

On Fri, Jan 14, 2022 at 8:19 AM Ian Rogers <irogers@google.com> wrote:
>
> On Tue, Dec 14, 2021, 1:56 PM Rob Herring <robh@kernel.org> wrote:
>>
>> Add the arm64 variants for read_perf_counter() and read_timestamp().
>> Unfortunately the counter number is encoded into the instruction, so the
>> code is a bit verbose to enumerate all possible counters.
>>
>> Signed-off-by: Rob Herring <robh@kernel.org>
>> ---
>> v8:
>>  - Set attr.config1 to request user access on arm64
>> v7:
>>  - Move enabling of libperf user read test for arm64 to this patch
>> ---
>>  tools/lib/perf/mmap.c             | 98 +++++++++++++++++++++++++++++++
>>  tools/lib/perf/tests/test-evsel.c |  5 +-
>>  2 files changed, 102 insertions(+), 1 deletion(-)
>>
>> diff --git a/tools/lib/perf/mmap.c b/tools/lib/perf/mmap.c
>> index c89dfa5f67b3..7ee3eb9f5e27 100644
>> --- a/tools/lib/perf/mmap.c
>> +++ b/tools/lib/perf/mmap.c
>> @@ -13,6 +13,7 @@
>>  #include <internal/lib.h>
>>  #include <linux/kernel.h>
>>  #include <linux/math64.h>
>> +#include <linux/stringify.h>
>>  #include "internal.h"
>>
>>  void perf_mmap__init(struct perf_mmap *map, struct perf_mmap *prev,
>> @@ -294,6 +295,103 @@ static u64 read_timestamp(void)
>>
>>         return low | ((u64)high) << 32;
>>  }
>> +#elif defined(__aarch64__)
>> +#define read_sysreg(r) ({                                              \
>> +       u64 __val;                                                      \
>> +       asm volatile("mrs %0, " __stringify(r) : "=r" (__val));         \
>> +       __val;                                                          \
>> +})
>> +
>> +static u64 read_pmccntr(void)
>> +{
>> +       return read_sysreg(pmccntr_el0);
>> +}
>> +
>> +#define PMEVCNTR_READ(idx)                                     \
>> +       static u64 read_pmevcntr_##idx(void) {                  \
>> +               return read_sysreg(pmevcntr##idx##_el0);        \
>> +       }
>> +
>> +PMEVCNTR_READ(0);
>> +PMEVCNTR_READ(1);
>> +PMEVCNTR_READ(2);
>> +PMEVCNTR_READ(3);
>> +PMEVCNTR_READ(4);
>> +PMEVCNTR_READ(5);
>> +PMEVCNTR_READ(6);
>> +PMEVCNTR_READ(7);
>> +PMEVCNTR_READ(8);
>> +PMEVCNTR_READ(9);
>> +PMEVCNTR_READ(10);
>> +PMEVCNTR_READ(11);
>> +PMEVCNTR_READ(12);
>> +PMEVCNTR_READ(13);
>> +PMEVCNTR_READ(14);
>> +PMEVCNTR_READ(15);
>> +PMEVCNTR_READ(16);
>> +PMEVCNTR_READ(17);
>> +PMEVCNTR_READ(18);
>> +PMEVCNTR_READ(19);
>> +PMEVCNTR_READ(20);
>> +PMEVCNTR_READ(21);
>> +PMEVCNTR_READ(22);
>> +PMEVCNTR_READ(23);
>> +PMEVCNTR_READ(24);
>> +PMEVCNTR_READ(25);
>> +PMEVCNTR_READ(26);
>> +PMEVCNTR_READ(27);
>> +PMEVCNTR_READ(28);
>> +PMEVCNTR_READ(29);
>> +PMEVCNTR_READ(30);
>
>
> Nit: It looks strange that 31 is not here, ie 31 counters rather than 32.

From the Arm ARM:
D13.4.8
PMEVCNTR<n>_EL0, Performance Monitors Event Count Registers, n = 0 - 30

The 32nd counter is the cycle counter which is special and accessed in
a different register.

Rob