|
I opened up my laptop and noticed that it was responding super slowly.
It's AMD R7 8-core (4 core with threading-tech) with 16GB RAM running Win10 and it always performs very well.
Internet?? or was it the laptop itself??
8 Cores All At Same Level of Utilization? Unlikely!
I checked the CPU Utilization and saw this (snapshot)[^].
Isn't that interesting?? I've never seen all 8 cores report that they are exactly utilized the same.
I closed CPU monitor and opened it again and saw the same thing.
After Reboot: Looks Normal Again
Here's a snapshot that seems more realistic[^].
Then I Noticed...
Hey, check out the first graphic and you'll see that the Speed was at .38GHz -- even though these chips are really at 2.0GHz
Not sure what happened there, so we are filing it under Weird!! 
modified 25-Jan-23 14:37pm.
|
|
|
|
|
Fake processor!
|
|
|
|
|
That flat 19% utilization of all cores is rather strange.
All modern general chips are prepared to reduce the clock speed when load is low, to reduce heating and battery drain. I have seen my 3.3 GHz CPU going down to about 1000 MHz, never as low as 380 MHz. Maybe I am good at keeping it busy  . Maybe the OS routine that should increase your clock speed when needed crashed, and 380 MHz is some sort of minimum value (for your processor), so it never was raised, and that is why your PC was sluggish.
Your CPU was stuck at approx. 1/5 its max clock speed. The CPU utilization was approx. 1/5 of the total capacity. If 19% utilization refers to the capacity at full clock speed rather than the current one (I don't know which is the case!), then the flat 19% can be read as '100% utilization at the current clock speed'. If there is a great backlog of CPU tasks, it isn't surprising that all cores are saturated. When I run Handbrake, I see all cores at flat 100% utilization for minutes at a time (at the nominal 3.3 MHz clock frequency). Trying to use the PC for other tasks in parallel with Handbrake is futile.
If you immediately after boot up turn on the Resource Monitor CPU view, you will loads of processes coming and going, for completing the boot up. When you see the login screen, there is still a lot to be done in the background. It takes several minutes before activity calms down, even if you do not touch the keyboard, activate no user task. If the CPU is strangled to 1/5 the normal speed, it may take five times as long. Maybe this startup activity is what you saw saturating all cores (given the clock speed reduction).
If this theory is correct, the one remaining problem is: What caused your PC to refuse to raise the clock frequency? If it had done so, that backlog of startup tasks could have been processed before you even noticed it. I have no idea about what could cause the clock adjustment to fail. (I have never looked into Windows source code at all - never had access to it, never cared to.)
|
|
|
|
|
I noticed from your first screenshot, your last boot-up was three days ago, so those 19% load isn't startup tasks!
However, the PC spends almost 8 GiBytes of RAM on something. 18% disk activity on a fast SSD, is remarkable. For a good WiFi unit, 104 Kbps isn't that much, but some process is using it. I am not familiar with AMD GPUs, but if your are just displaying the Task Manager, 27% utilization is somewhat on the high side. There is something keeping your CPU busy (and RAM, and disk, and GPU), even if it isn't startup tasks!
Compare that to the second screen dump: Half the memory load, 1% disk, half the Wi-Fi load, less than half GPU load (still surprising high, though!), half the handle count thread count. 1:34 after boot up, some of the load could come from start up tasks.
Before rebooting your PC, you could have produced screen dumps of process lists and thread counts and I/O-activity, to see which might be suspected for preventing the clock speed adjustment (it could still be a long way to the answer, though). If you experience the same again, you may want to take dumps of all the Resource Monitor's displays.
In the DOS days, any software problem could be handled by Ctrl-Alt-Delete. This is the first time for many years that I hear of that 
|
|
|
|
|
Fantastic info! Thanks for replying. 
trønderen wrote: then the flat 19% can be read as '100% utilization at the current clock speed'. If there is a great backlog of CPU tasks, it isn't surprising that all cores are saturated.
Yes, that was exactly what I was thinking too.
trønderen wrote: immediately after boot up turn on the Resource Monitor CPU view, you will loads of processes coming and going
I've definitely noticed this.
In my case my laptop had been sitting idle overnight (plugged in). I removed the plug then opened the lid. I think this is what caused the problem. It was probably in low-power mode then I removed the plug and I know from past that this laptop (and Microsoft) try to idle some cores to save power.
I think it attempted to do that but then I was attempting to activate the laptop also and something went screwy.
Also, notice that when my laptop is removed from power, it tries to idle the last 4 cores (to conserve power) even though I've forced my laptop to have the setting "prefer speed over power saving".
I've talked at length with people about this and many say "there is a BIOS setting!" But there actually isn't on my (quite new) HP laptop.
What I've Learned, What I will Try Next Time
In future I will first activate my laptop (open the lid) then after I see the desktop and only after will I pull the plug. Should'a been doing this all along but the hardware & software shouldn't do that anyways.
Thanks again for the discussion, really great info.
|
|
|
|
|
Hmm, if this happens a lot I would check the CPU temperature. Probably just a bug though.
|
|
|
|
|
As MS achieves wonderfulness by adding rounded icons and advertisements, one of their most used tools - windows explorer - continues to be trash. It's like it's just ignored with all the updates. One of my favorite "poorly implemented" features is search. Yes, I know there are better tools out there. So, I'm searching my dev drive for a particular file.... search is zipping along and displays file.zip. While the search is still running, I right mouse click on the file and select "open file location".
And the search aborts - going to that file. Why not launch another WE and keep the search going? NAh.
Fire up a search and change the sort order after completion - re-search.
Charlie Gilley
“They who can give up essential liberty to obtain a little temporary safety deserve neither liberty nor safety.” BF, 1759
Has never been more appropriate.
|
|
|
|
|
I have just received a new craft knife from Amazon. As expected, they needed to verify my age as it is an age-restricted item in the UK.
My wife got to the delivery driver first - he asked her what her birth year was and then asked her to confirm her age was x - it isn't, it is x-1.
Then he tried me - he asked me what my birth year was and then asked me to confirm my age was y - it isn't, it is y-1.
I guess that their hand-held devices calculate age = this_year - birth_year.
Surely, every student learning computing is told that this only works correctly on 31st December and for the rest of the year it is just luck if it guesses correctly. At this time of year, it is wrong for 96.2% of the population. A company the size of Amazon that owns its own cloud platforms should have sufficient skilled staff not to allow such a simple Computing 101 error through their vigorous testing procedures.
I might mention this if I get a 'Rate the quality of your experience with the delivery of this item' request.
|
|
|
|
|
Don't blame the driver - he's just following the on-screen instructions. Blame the idiot who wrote the app he's forced to use.
(Of course, if he couldn't see the problem with the question and just tick "yes", that's another story...)
"These people looked deep within my soul and assigned me a number based on the order in which I joined."
- Homer
|
|
|
|
|
Richard Deeming wrote: Don't blame the driver
I didn't! He accepted that we were both over 18 (min age required for buying a knife) as the calculated (albeit wrong) ages were much greater than 18 and than was the only verification that he needed. The problem would have been if we were 17 year olds as the calculation would have erroneously said we could legally accept the knife as it said we were 18.Richard Deeming wrote: (Of course, if he couldn't see the problem with the question and just tick "yes", that's another story...)
I didn't even attempt to try to explain why the answer was wrong - delivery drivers have very tight schedules that do not include timeouts for lessons in basic algorithms. However, if Amazon send me a request to write a review, I will include this flaw and that there is the potential for a legal transgression in the review.
|
|
|
|
|
And this is why we write unit tests...
Jeremy Falcon
|
|
|
|
|
yes, I've heard the UK has a knife control problem 
But, just how big was the knife? You can die from injuries from a sharp pen knife. Just curious.
Charlie Gilley
“They who can give up essential liberty to obtain a little temporary safety deserve neither liberty nor safety.” BF, 1759
Has never been more appropriate.
|
|
|
|
|
|
Today, a tester reported that they were unable to launch a couple of newer executables from one of our desktop apps. By 'unable to launch' I mean that instead of the called executable starting, it was opening a web browser stating that it couldn't find the file. (also some wonderful advice about possible mispelling and refresh) Every other executable in the menu worked correctly. 
My first thought was, 'silly tester, it works on my machine!' 
My next thought was, security/av interference? Nope, they start on a double-click. It's only a problem when started as a process.
So, two exe's that were failing to launch have one thing in common...the first part of the name which happens to be 'Mosaic'.
I changed the first letter in the filename, adjusted the calling app accordingly and sure enough, it worked.  
This just happened, so I'm still trying to figure out what exactly is going on. My guess is that 'Mosaic' in the filename is triggering a browser add-on to fire. I have customers using one of these 'modules' (the other is new) and have not had issues like this, though I don't know for sure if they were using the same OS environment as my tester. (Win11 w/latest patches) I don't have a Win11 system around (still on Win10) so I can't verify if this is a new behavior by design. Googling on the key terms isn't yielding anything useful. 
I'm already considering what it will take to change a filename that's been in production for a few years and installed on at least a dozen customer systems. Sure, I've got a process for it, but it's not fun. 
Edit: I've just had the same behaviour confirmed on another colleague's Win11 system. 'They hate cans!' (or just files starting with the letters mosaic! Damnit! 
Edit 2: After hands-on experimentation on my wife's Win11 laptop, I can confirm that there is an issue using ShellExecute to launch an executable with a filename starting with 'mosaic'. Doing so incorrectly starts a web browser with the message 'Hmmm, I can't find that file...Typo?'. It only happens on Win11 which is just now starting to be adopted by my users. Definitely weird.
"Go forth into the source" - Neal Morse
"Hope is contagious"
modified 3-Jan-23 9:43am.
|
|
|
|
|
Depending on how the image for the new process is identified, it could depend on the CWD and $PATH.
|
|
|
|
|
trønderen wrote: Depending on how the image for the new process is identified, it could depend on the CWD and $PATH.
It should be identified as a win32 executable with a handful of parameters added by the launcher which is calling with the full path and filename.
I've only been able to reproduce the issue on Win11. Changing the filename resolves the issue. I'm waiting for my wife to bring home her Win11 laptop so I can experiment with the issue more.
Again, the issue as I currently see it: with Win11 launching an executable as a process is doing something weird (starting a web browser) when the file name starts with 'Mosaic'. FWIW, the working directory is specified in the command and the executable is signed.
I could believe that some text in a filename might trigger an a/v action, but that doesn't seem to be the case. The file is where it's supposed to be, and runs fine manually. I'll know more later when I get my hands on a Win11 system and can test more thoroughly. I hate to change the filename as it accurately describes the product it interfaces with and what it does. 
I absolutely hate problems like this...no logical reason why changing the filename resolves the issue.
Thanks for the suggestions and reply!
"Go forth into the source" - Neal Morse
"Hope is contagious"
|
|
|
|
|
kmoorevs wrote: So, two exe's that were failing to launch have one thing in common...the first part of the name which happens to be 'Mosaic'. 
Theere's a list of exceptions in Windows for some executables; famously some sim that would not work in W95.
..but no. Just naming it won't trigger that.
kmoorevs wrote: I'm already considering what it will take to change a filename that's been in production for a few years and installed on at least a dozen customer systems
To a developer, that costs no time.
If it is that expensive to you, then hire a developer. Let me say that again; you're not fit for your position by your own words.
Bastard Programmer from Hell 
"If you just follow the bacon Eddy, wherever it leads you, then you won't have to think about politics." -- Some Bell.
|
|
|
|
|
Eddy Vluggen wrote: To a developer, that costs no time.
If it is that expensive to you, then hire a developer. Let me say that again; you're not fit for your position by your own words.
Modifying the program is simple. Distributing it (and ensuring that everyone installed the fix) is less so, especially if no provision for online updates was built into the program.
Freedom is the freedom to say that two plus two make four. If that is granted, all else follows.
-- 6079 Smith W.
|
|
|
|
|
Daniel Pfeffer wrote: Modifying the program is simple. Distributing it (and ensuring that everyone installed the fix) is less so, especially if no provision for online updates was built into the program.
You don't play Warcraft?
If the client isn't up to date, you can't log in. It starts updating as soon as there's a connection, and only presents a login screen if it is up to date.
I disabled auto-updating on Windows ofcourse, but you can't do that with Warcraft. The game is more secure than the platform I'm running it on.
Bastard Programmer from Hell
"If you just follow the bacon Eddy, wherever it leads you, then you won't have to think about politics." -- Some Bell.
|
|
|
|
|
Your final comment looks pretty spiteful. Is there a need for your nastiness?
|
|
|
|
|
Harsh, yes. No spite.
haughtonomous wrote: Is there a need for your nastiness?
Ofcourse there's no need for it.
..and there's no "off" switch either. I'm not going to explain the concept of a "launcher" executable. It's not like we're distributing software on floppy's.
Happy new year BTW.
Bastard Programmer from Hell
"If you just follow the bacon Eddy, wherever it leads you, then you won't have to think about politics." -- Some Bell.
|
|
|
|
|
Eddy Vluggen wrote: you're not fit for your position by your own words.
That was not fair.
Go install World of Warcraft. It will show you an easy fix.
Bastard Programmer from Hell
"If you just follow the bacon Eddy, wherever it leads you, then you won't have to think about politics." -- Some Bell.
|
|
|
|
|
First, thank you for your constructive input! I really appreciate it! 
Eddy Vluggen wrote: ..but no. Just naming it won't trigger that.
and yet, it does...consistently. Changing the filename changes the behavior.
Second, regarding changing a filename, I never said it was expensive however to a developer (me) it does cost time and requires some planning to make it a seamless transition for my customers.
Third, your unfounded insult means nothing to me. My colleagues and paying clients would disagree with you.
Fourth, I might be hiring for a jr. position. Are you available?
"Go forth into the source" - Neal Morse
"Hope is contagious"
|
|
|
|
|
Long, long ago, a Microsoft operating system only looked at, I think, the first 8 characters of the filename to find a file. But that was long, long ago.
|
|
|
|
|
Good reference to The Jerk!
|
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
