profiling | thin abstraction over instrumented profiling crates | Monitoring library

I must have fat-fingered something to turn it off, but when I try to run any c# project (.net 6 windows wpf, or even just a commandline application) in debug mode, I no longer see the Diagnostic Tools window. This is the thing that displays a real-time chart of memory consumption and other things.

I've been trying to re-enable Diagnostic Tools but so far haven't been able to do it. I can run in debug mode, everything else seems normal, I can add breakpoints etc. What could have gone wrong?

This is so perplexing!

Looking at the docs (Measure Performance While Debugging), it says...

The Diagnostic Tools window appears automatically unless you have turned it off. To bring up the window, click Debug / Windows / Show Diagnostic Tools (or press Ctrl + Alt + F2).

Here's what I tried:

• ctrl + alt + f2 Does nothing.

• In visual studio going to menu bar: Debug/Windows, there's no item for "Show Diagnostic Tools".

• When I search for Diagnostic Tools in the Visual studio search bar, it gives me a link to Debugging > General, but there is no checkbox for "Enable Diagnostic Tools while debugging" in the options dialog.

• Of course, I also tried closing/re-opening VS. Also, performed an update and did a "repair" from the VS installer. Still the same behavior.

• I also tried to "Reset Windows Layout" and "Reset All Settings" following the directions from this msdn forum post. Still stuck.

• According to Mohsyn's suggestion, Looked at Tools >> Options >> Debugging. There didn't seem to be anything checked relating to "Managed or Native Compatibility Mode".

• Blue Shell's answer had the solution!!! It was a matter of launching the visual studio installer, clicking Modify for my installation, going to "Individual Components" and checking ".NET profiling tools"

The next step WOULD HAVE BEEN to uninstall VS 2022 and re-install it.

I'd like to confirm the battery usage of an app I am developing on iOS, specifically on Xcode 13 and iOS 15. (Note: This app previously showed no issues with battery usage on previous versions of iOS.)

Previously, it seems that there were two ways to gather energy usage information:

#1. On the device under Settings > Developer > Logging

• As per Apple's documentation described in the section titled "Log Energy Usage Directly on an iOS Device".
• However, on iOS15, I can't find any options for logging under Developer or anywhere under settings even when searching.

#2. Profiling via Instruments using the "Energy Log" template

• As per the same documentation from Apple described in the section "Use the Energy Diagnostics Profiling Template".
• While it is still available in Xcode 12, this template is missing in Xcode 13. Naturally, it's also not possible to profile an iOS15 device with Xcode 12.

Digging through the Xcode 13 release notes, I found the following:

Instruments no longer includes the Energy template; use metrics reporting in the Xcode Organizer instead. (74161279)

When I access the Organizer in Xcode (12 or 13), select an app and click "Energy" for all versions of the app, it shows the following:

Apple's documentation for "Analyzing the Performance of Your Shipping App" says:

"In some cases the pane shows “Insufficient usage data available,” because there may not be enough anonymized data reported by participating user devices. When this happens, try checking back in a few days."

Well over a year into production and having sufficient install numbers, I have a feeling that waiting a few days might not do much.

I would like to determine if this is a bug in my app or a bug in iOS15. How can energy usage data be gathered using Xcode 13 on iOS 15?

When I open Android Studio I receive a notification saying that an update is available:

I am sorry but I am really confused and leery now, so I am resorting to SO to get some clarity.

I am running Android Studio Bumblebee and saw a notification about a major new release wit the following text:

My application consists of calling dozens of functions millions of times. In each of those functions, one or a few temporary std::vector containers of POD (plain old data) types are initialized, used, and then destructed. By profiling my code, I find the allocations and deallocations lead to a huge overhead.

A lazy solution is to rewrite all the functions as functors containing those temporary buffer containers as class members. However this would blow up the memory consumption as the functions are many and the buffer sizes are not trivial.

A better way is to analyze the code, gather all the buffers, premeditate how to maximally reuse them, and feed a minimal set of shared buffer containers to the functions as arguments. But this can be too much work.

I want to solve this problem once for all my future development during which temporary POD buffers become necessary, without having to have much premeditation. My idea is to implement a container port, and take the reference to it as an argument for every function that may need temporary buffers. Inside those functions, one should be able to fetch containers of any POD type from the port, and the port should also auto-recall the containers before the functions return.

I have this output :

[Pandas-profiling] ImportError: cannot import name 'ABCIndexClass' from 'pandas.core.dtypes.generic'

when trying to import pandas-profiling in this fashion :

I have a data table of 20M rows and 20 columns, to which I apply vectorized operations that return lists, themselves assigned by reference to additional columns in the data table.

The memory usage increases predictably and modestly throughout those operations, until I apply the (presumably highly efficient) frollmean() function to a column that contains lists of length 10 using an adaptive window. Running even the much smaller RepRex in R 4.1.2 on Windows 10 x64, with package data.table 1.14.2, the memory usage spikes by ~17GB when executing frollmean(), before coming back down, as seen in Windows' Task Manager (Performance tab) and measured in the Rprof memory profiling report.

I understand that frollmean() uses parallelism where possible, so I did set setDTthreads(threads = 1L) to make sure the memory spike is not an artifact of making copies of the data table for additional cores.

My question: why is frollmean() using so much memory relative to other operations, and can I avoid that?

I'm using perf for profiling on Ubuntu 20.04 (though I can use any other free tool). It allows to pass a delay in CLI, so that event collection starts after a certain time since program launch. However, this time varies a lot (by 20 seconds out of 1000) and there are tail computations which I am not interested in either.

So it would be great to call some API from my program to start perf event collection for the fragment of code I'm interested in, and then stop collection after the code finishes.

It's not really an option to run the code in a loop because there is a ~30 seconds initialization phase and 10 seconds measurement phase and I'm only interested in the latter.

In this programming problem, the input is an n×m integer matrix. Typically, n≈ 10⁵ and m ≈ 10. The official solution (1606D, Tutorial) is quite imperative: it involves some matrix manipulation, precomputation and aggregation. For fun, I took it as an STUArray implementation exercise.

I have managed to implement it using STUArray, but still the program takes way more memory than permitted (256MB). Even when run locally, the maximum resident set size is >400 MB. On profiling, reading from stdin seems to be dominating the memory footprint:

Functions readv and readv.readInt, responsible for parsing integers and saving them into a 2D list, are taking around 50-70 MB, as opposed to around 16 MB = (10⁶ integers) × (8 bytes per integer + 8 bytes per link).

Is there a hope I can get the total memory below 256 MB? I'm already using Text package for input. Maybe I should avoid lists altogether and directly read integers from stdin to the array. How can we do that? Or, is the issue elsewhere?