Atomic instructions are both memory-reading and memory-writing
instructions and so should be counted by both LD_RETIRED and ST_RETIRED
performance monitoring events. However LD_RETIRED does not count atomic
instructions.

Signed-off-by: Ilkka Koskinen <ilkka@os.amperecomputing.com>
---
 tools/perf/pmu-events/arch/arm64/ampere/ampereone/memory.json | 4 +++-
 .../perf/pmu-events/arch/arm64/ampere/ampereonex/memory.json  | 4 +++-
 2 files changed, 6 insertions(+), 2 deletions(-)

diff --git a/tools/perf/pmu-events/arch/arm64/ampere/ampereone/memory.json b/tools/perf/pmu-events/arch/arm64/ampere/ampereone/memory.json
index 0711782bfa6b..13382d29b25f 100644
--- a/tools/perf/pmu-events/arch/arm64/ampere/ampereone/memory.json
+++ b/tools/perf/pmu-events/arch/arm64/ampere/ampereone/memory.json
@@ -1,6 +1,8 @@
 [
     {
-        "ArchStdEvent": "LD_RETIRED"
+        "ArchStdEvent": "LD_RETIRED",
+        "Errata": "Errata AC03_CPU_52",
+        "BriefDescription": "Instruction architecturally executed, condition code check pass, load. Impacted by errata -"
     },
     {
         "ArchStdEvent": "MEM_ACCESS_RD"
diff --git a/tools/perf/pmu-events/arch/arm64/ampere/ampereonex/memory.json b/tools/perf/pmu-events/arch/arm64/ampere/ampereonex/memory.json
index a211d94aacde..6c06bc928415 100644
--- a/tools/perf/pmu-events/arch/arm64/ampere/ampereonex/memory.json
+++ b/tools/perf/pmu-events/arch/arm64/ampere/ampereonex/memory.json
@@ -1,6 +1,8 @@
 [
     {
-        "ArchStdEvent": "LD_RETIRED"
+        "ArchStdEvent": "LD_RETIRED",
+        "Errata": "Errata AC04_CPU_21",
+        "BriefDescription": "Instruction architecturally executed, condition code check pass, load. Impacted by errata -"
     },
     {
         "ArchStdEvent": "MEM_ACCESS_RD"
-- 
2.47.0

On 13/03/2025 8:15 pm, Ilkka Koskinen wrote:
> Atomic instructions are both memory-reading and memory-writing
> instructions and so should be counted by both LD_RETIRED and ST_RETIRED
> performance monitoring events. However LD_RETIRED does not count atomic
> instructions.
> 
> Signed-off-by: Ilkka Koskinen <ilkka@os.amperecomputing.com>
> ---
>   tools/perf/pmu-events/arch/arm64/ampere/ampereone/memory.json | 4 +++-
>   .../perf/pmu-events/arch/arm64/ampere/ampereonex/memory.json  | 4 +++-
>   2 files changed, 6 insertions(+), 2 deletions(-)
> 
> diff --git a/tools/perf/pmu-events/arch/arm64/ampere/ampereone/memory.json b/tools/perf/pmu-events/arch/arm64/ampere/ampereone/memory.json
> index 0711782bfa6b..13382d29b25f 100644
> --- a/tools/perf/pmu-events/arch/arm64/ampere/ampereone/memory.json
> +++ b/tools/perf/pmu-events/arch/arm64/ampere/ampereone/memory.json
> @@ -1,6 +1,8 @@
>   [
>       {
> -        "ArchStdEvent": "LD_RETIRED"
> +        "ArchStdEvent": "LD_RETIRED",
> +        "Errata": "Errata AC03_CPU_52",
> +        "BriefDescription": "Instruction architecturally executed, condition code check pass, load. Impacted by errata -"

I think this could also have a 'Fixes:' tag, either way:

Reviewed-by: James Clark <james.clark@linaro.org>

Hi James,

On Tue, 18 Mar 2025, James Clark wrote:
> On 13/03/2025 8:15 pm, Ilkka Koskinen wrote:
>> Atomic instructions are both memory-reading and memory-writing
>> instructions and so should be counted by both LD_RETIRED and ST_RETIRED
>> performance monitoring events. However LD_RETIRED does not count atomic
>> instructions.
>> 
>> Signed-off-by: Ilkka Koskinen <ilkka@os.amperecomputing.com>
>> ---
>>   tools/perf/pmu-events/arch/arm64/ampere/ampereone/memory.json | 4 +++-
>>   .../perf/pmu-events/arch/arm64/ampere/ampereonex/memory.json  | 4 +++-
>>   2 files changed, 6 insertions(+), 2 deletions(-)
>> 
>> diff --git a/tools/perf/pmu-events/arch/arm64/ampere/ampereone/memory.json 
>> b/tools/perf/pmu-events/arch/arm64/ampere/ampereone/memory.json
>> index 0711782bfa6b..13382d29b25f 100644
>> --- a/tools/perf/pmu-events/arch/arm64/ampere/ampereone/memory.json
>> +++ b/tools/perf/pmu-events/arch/arm64/ampere/ampereone/memory.json
>> @@ -1,6 +1,8 @@
>>   [
>>       {
>> -        "ArchStdEvent": "LD_RETIRED"
>> +        "ArchStdEvent": "LD_RETIRED",
>> +        "Errata": "Errata AC03_CPU_52",
>> +        "BriefDescription": "Instruction architecturally executed, 
>> condition code check pass, load. Impacted by errata -"
>
> I think this could also have a 'Fixes:' tag, either way:
>
> Reviewed-by: James Clark <james.clark@linaro.org>

Yeah, that's true. I guess, I was thinking about using 'Fixes' more on 
broken patches rather than when adding a patch dealing with a new hw 
errata. Well, given you're find with this, I don't think I'll resubmit it
with a Fixes tag this time.

Cheers, Ilkka