3.3 Crash (How Embarrassing for me see last post)

jdow
Posts: 803
Joined: Mon Aug 10, 2020 8:17 pm

Re: 3.3 Crash

#11

Unread post by jdow »

@Lappydesktop Take a picture before it has crashed showing the settings. Maybe somebody will notice an awkward setting.

Can you disable automatic throttling? That is noted to cause problems in some configurations. If not try looping a video in a background browser window.

{o.o}

User avatar
lappydesktop
Posts: 70
Joined: Sat Jan 21, 2023 11:22 am
Location: DownUnder

Re: 3.3 Crash

#12

Unread post by lappydesktop »

jdow wrote: Thu Feb 16, 2023 11:07 am @Lappydesktop Take a picture before it has crashed showing the settings. Maybe somebody will notice an awkward setting.
Can you disable automatic throttling? That is noted to cause problems in some configurations. If not try looping a video in a background browser window.
{o.o}
Hi , what's to throttle ?
10.jpg
and that's with that there under this , not much happening , and one can forget about all this until i get back on 2Mhz BW at a later date , thanks .
Screenshot-2023-02-16-225009.jpeg

jdow
Posts: 803
Joined: Mon Aug 10, 2020 8:17 pm

Re: 3.3 Crash

#13

Unread post by jdow »

ll I can see that might be different from what I have done here in the past is the precise 2 MHz bandwidth and the NR3 selection. Am I correct to rashly presume it behaves the same with no NR enabled? If not we have something narrowed down for Simon to troubleshoot.

That looks like a nice computer. In theory nothing should crash. About the only other test that comes to mind is running a looping video with audio off in the background to give some load so that any CPU throttling would not happen. Usually that manifests as choppy audio. But, who knows what evil lies in newly released AMD or Intel or NVIDIA hardware? (Until ASUS revised the BIOS on my machine it was a little flaky.) Of course it's not unlikely that the RSPdx code and Simon's code have some exotic form of disagreement. I agree with you that the immediate work around is "Don't do that!" I just would be worrying about some other hidden problem waiting to erupt due to the same root problem.

{^_^}

User avatar
lappydesktop
Posts: 70
Joined: Sat Jan 21, 2023 11:22 am
Location: DownUnder

Re: 3.3 Crash

#14

Unread post by lappydesktop »

jdow wrote: Thu Feb 16, 2023 10:56 pm ll I can see that might be different from what I have done here in the past is the precise 2 MHz bandwidth and the NR3 selection. Am I correct to rashly presume it behaves the same with no NR enabled? If not we have something narrowed down for Simon to troubleshoot.

That looks like a nice computer. In theory nothing should crash. About the only other test that comes to mind is running a looping video with audio off in the background to give some load so that any CPU throttling would not happen. Usually that manifests as choppy audio. But, who knows what evil lies in newly released AMD or Intel or NVIDIA hardware? (Until ASUS revised the BIOS on my machine it was a little flaky.) Of course it's not unlikely that the RSPdx code and Simon's code have some exotic form of disagreement. I agree with you that the immediate work around is "Don't do that!" I just would be worrying about some other hidden problem waiting to erupt due to the same root problem.

{^_^}
Yeh ok on any suspicions i'm with you now and assume there is something obviously going on , something sneaky and not constant or reliable .

Regarding the NR3 , i would not have a clue as it all happens very erratic and not constant and i could not be bothered waiting to find out (it would be many hours or days or weeks)

These series of crash's are the very 1st ones i have had on this win10 and computer since built up so i was a bit surprised as leaving w7 was a risk as i hardly ever got a operating system crash on that one for years (apart from my fault or an update fault) , so it seemed coincidental that it has happened only on sdrconsole so far (i'm not saying that is the fault as there might be something else going on) .

Last night for instance i noticed sdrconsole was in amd Adrenalin as a game so i deleted that profile (who knows?) and has remained deleted in there , also there is the issue of 1 monitor (the main) being a new amd freesync one and the other not and smaller , an amd 32" curved compared to older hp 27" flat , one on DVI the other HDMI , the hp is the external one where sdrconsole sits on atm and yet to even try sdrconsole on the main just to see as i don't have the time or inclination to do any research/experiments and i'm Lazy! .

So , also i took a look about last night on the net for specific reports of crashing for AMD w10 in the past and saw all those from a few years ago regarding cpu cache hierarchy , plenty about back in those days intel or amd but mostly amd , and reading about what a joke it all was too , like plugging into a different mains power point fixed it then did not then did then did not and all those other "fixes" , seems like it went on for a few years up to about 2021 and i really did not get to the "what really caused it all" scenario as i became disinterested and bored with it all , so as i am really just on about enjoying the sdr and software and using it and not really interested about diagnosing until it becomes a real drama (i can't use the sdr at all) i'll just carry on .

But i'll guess and say it's a cpu cache hierarchy thing going by the past , of course i'll find out on the next time one when it or if it gets logged , i mentioned i cleared all the event logs a few days back and still on 3Mhz BW now .

Cu

jdow
Posts: 803
Joined: Mon Aug 10, 2020 8:17 pm

Re: 3.3 Crash

#15

Unread post by jdow »

Gee, I'd not considered freesync. That would likely break some SDR playing I've done here, or at least provide a suboptimal experience. (heh, in one case SDRPlaying was almost the internal name of a project.)

If you can lock the monitor's refresh rate ideally to match the other monitor that MIGHT save problems, especially when SDRC elements overlap screen boundaries. Peculiar things like that might be noted, if you remember, the next time it crashes.

As an aside I note that Intel has just released some XEON SKUs for consumer use with simply huge numbers of available PCIe 5 lanes. (64 and 112). That would be nice for running LOTS of high end USB 3 ports. That would save connector wear. I could leave everything connected all the time.

{^_^}

User avatar
lappydesktop
Posts: 70
Joined: Sat Jan 21, 2023 11:22 am
Location: DownUnder

Re: 3.3 Crash

#16

Unread post by lappydesktop »

AMD Freesync only works on a Freesync monitor in which is the main monitor in this config , sdrconsole sits on the other monitor (no freesync available) .

No crash as yet btw and its been awhile now (i guess rx time 60 hours plus at least) , tonight looking about i noticed cuda (Nvidia) was enabled for some reason which i did not enable but i remember seeing it when i 1st looked ages ago , so i disabled that , that's all for now with nothing else to report .

User avatar
lappydesktop
Posts: 70
Joined: Sat Jan 21, 2023 11:22 am
Location: DownUnder

Re: 3.3 Crash (How Embarrassing for me see last post)

#17

Unread post by lappydesktop »

Last night remembered i liked that s-meter circle that was on the Needle but i could not find the setting for it , so i'm wondering what's going on!

I looked and found the post here on the forums in where 3.3 did crash and was reported in the BUGS area where when one right clicked the top bar which also happened to this computer and then noticed Simeon fixed it a short while later and put out the 3.3 2870 , i then realized that at that time i went back to 3.2 and did not keep up with what was going on and forgot to install 3.3 2870 .

Then last night after giving up on finding the s-meter needle pointer i just happened to check the saved installer i used and yep it said 3.2 duh .

So i downloaded the 3.3 2870 version and installed , found the nice pointer and now about to start on the 2Mhz BW and hopefully it will all be well , so i humbly apologize especially to Simeon for my mistake , and others .

Post Reply