Jump to content

Runtime error and crash


Mustang

Recommended Posts

But Mustang... see you got Physical Memory usage at 94%? I don't think you can have 100% anyway as you need some space to read/write (same as with HDD, it can't be utilised 100%) so maybe that's why it crashes? RAM usage limit?

No longer active in DCS...

Link to comment
Share on other sites

  • Replies 134
  • Created
  • Last Reply

Top Posters In This Topic

But Mustang... see you got Physical Memory usage at 94%? I don't think you can have 100% anyway as you need some space to read/write (same as with HDD, it can't be utilised 100%) so maybe that's why it crashes? RAM usage limit?

 

Yes good thinking, but ok then, shouldn't it start to use virtual memory at this point? i wished it would, i prefer HDD thrashing over constant CTD's.

Link to comment
Share on other sites

hm... I don't think you can use virtual memory for programs but I don't know, I'm just guessing here, but that would be PageFile right? if so it should only be used for System itself for backup as it works through RAM etc...

No longer active in DCS...

Link to comment
Share on other sites

Yes good thinking, but ok then, shouldn't it start to use virtual memory at this point? i wished it would, i prefer HDD thrashing over constant CTD's.

 

Have you set/limit your paging file to anything specific in Windows 7?

 

On my XP I have paging file size set to 2 GB on my secondary drive (been running this way for years) but it would be an easy test to change it and see if things changes, will do so later tonight and see.

 

/KC

Link to comment
Share on other sites

If you are really curious, I did some similar poking around regarding Win7 and memory issues wiuth DCS, a back while before I joined the tester-team.

http://forums.eagle.ru/showthread.php?t=41695

There's a little bit of discussing what the causes and limitations are - and c0ff did some changes to change the issues (debug log writer).

 

But there may be some useful tips there that you guys may try to see if it changes the behaviour of your crashes.

The mind is like a parachute. It only works when it's open | The important thing is not to stop questioning

Link to comment
Share on other sites

Ok i think I know what is happening. It is not a memory leak, but rather a lack of

memory clearing. Old stuff is simply not cleared out. Video memory is not being

overrun but normal ram IS being overrun.

 

This picture is on my 4 GB system and it is about to run out of memory. (although

I never got the crash I am basically out of memory and the game is stuttering.)

 

oom.png

 

The reason why it is not a leak but just a lack of memory clearing is that once I

"fly through an area (just f11+spin wasnt enough for me, I also used F11 view to

fly all around the map in the 104th server), the second time I enter an area no new

memory is allocated. SO it seems very clear to me, that there is

 

  • No memory cleanup of old potentially unused resources
  • A lot new materials to load with time
  • You will run out of memory on a 2 GB system, potentially also on 4 GB systems

 

Because this is a design flaw/choice/result of the current code, I do not think we can

expect any change to this behaviour or any "patch" to change it. It would require too

much work. So in the meantime all I can suggest is that you turn down the textures

scenes to medium and get more RAM, at least 4 GB as a workaround. If you can, get

6 or 8GB of ram.

 

I will ask the developers.

 

Update: I have made a very long post in testers forums asking the developers. Will relay the answer

 

 

-->Mustang do you have any external app to read and plot VRAM usage with during the same test?<--

 

.


Edited by =RvE=Yoda

S = SPARSE(m,n) abbreviates SPARSE([],[],[],m,n,0). This generates the ultimate sparse matrix, an m-by-n all zero matrix. - Matlab help on 'sparse'

Link to comment
Share on other sites

So in the meantime all I can suggest is that you turn down the textures

scenes to medium and get more RAM, at least 4 GB as a workaround. If you can, get

6 or 8GB of ram.

wont help.(the memory thing)

Simulator.exe is an 32bit application and is bound to 2GB virtual memory max. Once you reach that, the application will crash.

I'm really in doubt if additing aditional 2GB will solve the problem. In theory of memory virtualization - it shouldn't.

I would try to manualy rise the pagefile to around 4GB. It maybe that win just ran out of virtual memory and starts refusing memory allocation.


Edited by winz
Link to comment
Share on other sites

wont help.

Simulator.exe is an 32bit application and is bound to 2GB virtual memory max. Once you reach that, the application will crash.

I'm really in doubt if additing aditional 2GB will solve the problem. In theory of memory virtualization - it shouldn't.

I would try to manualy rise the pagefile to around 4GB. It maybe that win just ran out of virtual memory and starts refusing memory allocation.

 

Yes 32 bit is bound to 2 GB max, but if a system only has 4gigs maximum it won't always allow

applications to take 2GB, that is why i rec 6 or more. (My own system doesn't like above 1.4

GB or so for one app)

S = SPARSE(m,n) abbreviates SPARSE([],[],[],m,n,0). This generates the ultimate sparse matrix, an m-by-n all zero matrix. - Matlab help on 'sparse'

Link to comment
Share on other sites

It will, as long as you don't exceed the page file limitation, it will allow you to allocate full 2gigs of ram. But, the computer will switch to crawl mode.

 

Mustang may be running without page file I guess?

 

Myself I've entered crawl mode so far :). Also consider memory fragmentation.

My friend wrote a paper on that, however I've never read it and don't know how good xp is at it.

S = SPARSE(m,n) abbreviates SPARSE([],[],[],m,n,0). This generates the ultimate sparse matrix, an m-by-n all zero matrix. - Matlab help on 'sparse'

Link to comment
Share on other sites

simulator.exe is /LARGEADDRESSAWARE and can eat up to 3GB (if os allows).

BUt it's only a workaround.

 

Very good to know. Does this mean that the game currently does not perform cleanup?

(Future feature perhaps?)


Edited by =RvE=Yoda

S = SPARSE(m,n) abbreviates SPARSE([],[],[],m,n,0). This generates the ultimate sparse matrix, an m-by-n all zero matrix. - Matlab help on 'sparse'

Link to comment
Share on other sites

To everyone with this problem:

 

http://forums.eagle.ru/showthread.php?p=710319#post710319

This is apparently already investigated for DCS BS. The best workaround so far seems to be

get 6 gigs of ram and run 64bit OS

 

Thank you c0ff and Panzer for your explanations! :)

Thx everyone else also for your input.


Edited by =RvE=Yoda

S = SPARSE(m,n) abbreviates SPARSE([],[],[],m,n,0). This generates the ultimate sparse matrix, an m-by-n all zero matrix. - Matlab help on 'sparse'

Link to comment
Share on other sites

Wow, now this is nice, i believe i've stopped my crashes at 1.2gb!

 

I've just flashed my motherboard BIOS to version 2.20a (going to need this if i want to run 4gb of RAM on my board) and i've also changed my virtualram size to this (beforehand it was set to system managed size):

 

virtualram.jpg

 

and now look!

 

fixedx.jpg

 

RAM usage shot past that dreaded 1.2gb mark, infact it got upto around 1.6 just before i quit, also no textures were disappearing and no menu screens were turning white.

 

I'll test this out some more right now to solidly confirm, but man, i think this is it :D


Edited by Mustang
Link to comment
Share on other sites

Keep us posted. Glad its showing improvments.

[sigpic]http://forums.eagle.ru/signaturepics/sigpic4448_29.gif[/sigpic]

My PC specs below:

Case: Corsair 400C

PSU: SEASONIC SS-760XP2 760W Platinum

CPU: AMD RYZEN 3900X (12C/24T)

RAM: 32 GB 4266Mhz (two 2x8 kits) of trident Z RGB @3600Mhz CL 14 CR=1T

MOBO: ASUS CROSSHAIR HERO VI AM4

GFX: GTX 1080Ti MSI Gaming X

Cooler: NXZT Kraken X62 280mm AIO

Storage: Samsung 960 EVO 1TB M.2+6GB WD 6Gb red

HOTAS: Thrustmaster Warthog + CH pro pedals

Monitor: Gigabyte AORUS AD27QD Freesync HDR400 1440P

 

Link to comment
Share on other sites

It's still all good :)

 

Man, i had to crash out lastnight and get some sleep, this PC has had me up for too long and stressing me out, so damn happy it's finally resolved :D

 

Have to say big thankyou to everyone who helped me out, no doubt the discussions we've all been having on here will help out others who are still encountering runtime errors or crashes.

 

Now i'm wondering, was it the BIOS update i did, or the virtual memorysize change that corrected my issues......

 

edit: changed VirtualRAM/pagefile back down to system managed, and i haven't encountered any crashes, so i believe it was something in my BIOS update that fixed it for me.


Edited by Mustang
Link to comment
Share on other sites

I'm glad to hear you did find something that finally did make a change, Mustang.

I hope others with this horrible problem is also able to find a working solution, like you did. :thumbup:

The mind is like a parachute. It only works when it's open | The important thing is not to stop questioning

Link to comment
Share on other sites

This reminds me I have to start building my new computer and get more ram

S = SPARSE(m,n) abbreviates SPARSE([],[],[],m,n,0). This generates the ultimate sparse matrix, an m-by-n all zero matrix. - Matlab help on 'sparse'

Link to comment
Share on other sites

if it was set to system managed size before it would 1.5-2x your available ram. I would chalk it up more to the bios update maybe having something to do with your memory management. Don't know your processor but AMD's can be finicky between boards sometimes with on die memory controllers and sometimes bios's contain microcode updates.

Link to comment
Share on other sites

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...