1,062 questions
1
vote
0
answers
27
views
Raspberry Pi 5 PMUs return 0 when configured manually from EL1
I am working on a Raspberry Pi 5 (aarch64).
uname -a:
Linux rpi5 6.8.0-1036-raspi #40-Ubuntu SMP PREEMPT_DYNAMIC Mon Aug 18 09:50:42 UTC 2025 aarch64 aarch64 aarch64 GNU/Linux
I load a kernel module I ...
1
vote
0
answers
41
views
rdpmc: Getting LLC references and misses efficiently
I use Linux (ubuntu) and C (gcc). I try to get the last level cache references and misses by using rdpmc. For me it's important that it's efficient and due to that I cannot use syscalls. It's okay to ...
1
vote
1
answer
227
views
perf report: Understanding the first line of output
I was looking for an answer but found nothing definitive. How do I interpret the first line in perf report output. It goes like this:
Samples: 173M of event 'cache-misses', Event count (approx.): ...
3
votes
0
answers
329
views
How to understand intel performance counters for evaluating cpu stalls due to memory subsystem?
I played with various Intel performance counters by directly accessing the Model Specific Registers (MSRs) on a Xeon Skylake chip, and I found some readings about the cache-miss caused stalls like ...
-2
votes
1
answer
829
views
time.perf_counter() or time.perf_counter_ns() giving weird and incorrect values
I have written this simple while I was working with multi-processing.
from multiprocessing import Process, cpu_count
import time
def counter(num):
count = 0
while count < num:
...
1
vote
0
answers
58
views
Windows OS: Win32_PerfRawData_PerfProc_Process Virtual Memory is not matching with resource monitor's Commit Memory
I am trying to pull the Process metrics using Win32_PerfRawData_PerfProc_Process as documented at https://learn.microsoft.com/en-us/previous-versions/aa394323(v=vs.85)
From the document, VirtualBytes ...
1
vote
0
answers
98
views
Memory Level Parallelism (MLP) measurement
I need to find the Memory Level Paralelism (MLP) or the number of memory request that are concurrently held by Miss Status Handling registers (MSHRs) / Fill buffers at each cache level during a ...
0
votes
2
answers
851
views
Issue with PDH Counter returning incorrect values after upgrading to Windows 11
I've encountered an issue with the PDH counter after updating my system from Windows 10 to Windows 11. I have a C++ program that monitors CPU usage and it was working perfectly on the previous version ...
0
votes
0
answers
216
views
What does the event `stall_slot_backend` represent?
I am trying to diagnose low cycles-per-instruction in a small program. Using perf stat -e cycles,stalled-cycles-backend ./myprogram, I see that around 90% backend cycles are idle.
Performance counter ...
1
vote
0
answers
105
views
Perf: how to display cache miss percentage from raw counters
I am using the perf tool on Linux with an Intel Xeon E5-1650 v4 processor (Broadwell architecture).
For several cache load/miss indicators, perf stat indicates that these are <not supported>:
$ ...
1
vote
0
answers
323
views
CPU performance counters in C++ (Mac/PC, Intel)
I want to measure specific parts of my code to understand how well they perform. I already found the following:
How to get the CPU cycle count in x86_64 from C++?
...which allows me to easily ...
1
vote
0
answers
213
views
No PAPI counters available on Ubuntu 20.04
I am trying to install the PAPI library in order to measure the cache misses of other programs. However, apparently there aren't any events available as can be seen in the output of papi_avail :
$ ./...
2
votes
0
answers
76
views
Can rdpmc be used to read the fixed-function counters on AMD?
On Intel the fixed-function performance counters can be read by setting bit 30 of ecx as well the index of the counter to read (0-4) in the bottom bits of that same register.
Is something similar ...
3
votes
1
answer
614
views
Memory Leak in a Rust Program
I wrote a Rust program using windows-sys crate and collecting the performance counter values for query that is being supplied. I use the sample program provided by Microsoft team (here).
The program ...
0
votes
1
answer
147
views
PdhAddCounterW - no rules expected this token in macro call
I am writing a program to collect Performance Counter. The program is written in Rust and uses the windows crate.
Below is the relevant portion of the code.
let mut query = 0;
PdhOpenQueryW(None, 0, &...
1
vote
1
answer
604
views
PerformanceCounter sometimes throws The Counter layout for the Category specified is invalid
We are monitoring a couple of PerformanceCounters every 5 minutes in a Windows-service. Sometimes our code throws:
"Error Message: The Counter layout for the Category specified is invalid, a ...
5
votes
1
answer
234
views
In Windows, what is an auxiliary performance counter?
Windows provides a number of documented APIs relating to the "auxiliary counter" aka "auxiliary performance counter".
In user-mode, QueryAuxiliaryCounterFrequency, ...
1
vote
1
answer
255
views
In c# how to get given disk's realtime "Active time" in%, "Average response time"?
Been trying to get this work but I just couldn't get the exact solution
PerformanceCounter disk = new PerformanceCounter("LogicalDisk", "% Disk Time", @"F:");
//disk....
4
votes
0
answers
416
views
What could be the causes of this performance regression, and how to investigate it?
Context
I'm writing some high-performance code for ARM64 using NEON SIMD instructions, which I am trying to further optimize. I only use integer operations, no floating-point. This code is fully CPU- ...
1
vote
1
answer
384
views
Is GC stats still accessible on perfmon/PerformanceCounter.dll for a .NET6 desktop app?
We have a WPF app (x64) running on .NET4.8 and we used to print GC/memory related stats onto log periodically using System.Diagnostics.PerformanceCounter. (Using category: ".NET CLR Memory")
...
0
votes
1
answer
177
views
C# WPF MVVM | Use Performancecounter with Timer
I try to set up a very simple application.
The application should display the current CPU Usage.
My view:
<Grid>
<Label Content="{Binding CpuUsage}" />
</Grid&...
0
votes
0
answers
159
views
Measure load stalls using Intel performance monitoring counters for RESOURCE_STALLS
I am trying to understand meaning of various Intel performance monitoring counters and also want to measure load stalls using Intel performance monitoring counters available for RESOURCE_STALLS.
The ...
0
votes
1
answer
143
views
Where can I find the Performance Metrics generated by SageMaker Debugger/Profiler?
Where can I look for the performance metrics generated by Amazon SageMaker Debugger/Profiler?
1
vote
1
answer
288
views
Get disk time in percentage in windows
I have the following scenario, I want to get the utilization (Disk Time) per disk in percent.
Unfortunately it always returns the value 0 when I switch through the instances, why?
Here is my code:
...
0
votes
1
answer
274
views
Accidentally deleted performance counter on a performance monitor report
On Performance Monitor, I was watching a Report of a test I had run, using a Custom Data Collector set. I use Highligh to see the counter I want, but I accidentally clicked on the button next to it, ...