[RESOLVED] DCS crash when typing in landing with coordinates for road landings

[nsksleeper]

hi i tried to type in the coords for a road to land on, but as i press L-Mål DCS is crashing every time, didn't fine a report on this so i do.

 

log file is attached

 

crash message tells problem AJS37.dll

dcs.log

Edited September 20, 2021 by IronMike

[Flappie]

I can't reproduce your crash, at least over Caucasus. This is what I've tested:

• coordinates of a road segment of the map: OK (video below)
  
• coordinates off the map (Atlantic ocean): OK
  
• coordinates off the map (North Pole): OK
  

 

 

The function that crashes your DCS is called "FlightComputer::FindClosestRunway" by the way. Can you explain the steps to reproduce the crash?

A track would be welcome too.

 

2020-11-10 18:18:18.735 INFO EDCORE: try to write dump information

2020-11-10 18:18:18.741 INFO EDCORE: # -------------- 20201110-181819 --------------

2020-11-10 18:18:18.748 INFO EDCORE: DCS/2.5.6.57530 (x86_64; Windows NT 10.0.18363)

2020-11-10 18:18:18.754 INFO EDCORE: C:WINDOWSSystem32ucrtbase.dll

2020-11-10 18:18:18.759 INFO EDCORE: # 80000003 BREAKPOINT at B13EE9AC 00:00000000

2020-11-10 18:18:18.767 INFO EDCORE: SymInit: Symbol-SearchPath: '.;C:SpieleDCS Worldin', symOptions: 534, UserName: 'Ekorre'

2020-11-10 18:18:18.770 INFO EDCORE: OS-Version: 10.0.18363 () 0x300-0x1

2020-11-10 18:18:19.549 INFO EDCORE: 0x000000000006E9AC (ucrtbase): get_wpgmptr + 0xE2C

2020-11-10 18:18:19.552 INFO EDCORE: 0x0000000000065AE4 (AJS37): FlightComputer::FindClosestRunway + 0x7504

2020-11-10 18:18:19.557 INFO EDCORE: 0x00000000000201A0 (AJS37): ed_fm_set_draw_args + 0x45D0

2020-11-10 18:18:19.564 INFO EDCORE: 0x0000000000019AA3 (AJS37): ed_fm_simulate + 0x73

2020-11-10 18:18:19.564 INFO EDCORE: 0x00000000005E8058 (DCS): CoreUtils::TempFilesManager::operator= + 0x1C41A8

2020-11-10 18:18:19.564 INFO EDCORE: 0x00000000000BA3EB (FMBase): EagleFM::AerodyneFM::simulate + 0x43B

2020-11-10 18:18:19.565 INFO EDCORE: 0x0000000000746DA9 (DCS): CoreUtils::TempFilesManager::operator= + 0x322EF9

2020-11-10 18:18:19.565 INFO EDCORE: 0x00000000005D9565 (DCS): CoreUtils::TempFilesManager::operator= + 0x1B56B5

2020-11-10 18:18:19.565 INFO EDCORE: 0x00000000005D906F (DCS): CoreUtils::TempFilesManager::operator= + 0x1B51BF

2020-11-10 18:18:19.565 INFO EDCORE: 0x0000000000003B94 (World): wSimTrace::CommandsTraceDiscreteIsOn + 0x504

2020-11-10 18:18:19.565 INFO EDCORE: 0x000000000000403D (World): wSimCalendar::DoActionsUntil + 0x1FD

2020-11-10 18:18:19.565 INFO EDCORE: 0x00000000007DC31F (DCS): CoreUtils::TempFilesManager::operator= + 0x3B846F

2020-11-10 18:18:19.566 INFO EDCORE: 0x00000000007EEF23 (DCS): CoreUtils::TempFilesManager::operator= + 0x3CB073

2020-11-10 18:18:19.566 INFO EDCORE: 0x00000000007C0D84 (DCS): CoreUtils::TempFilesManager::operator= + 0x39CED4

2020-11-10 18:18:19.566 INFO EDCORE: 0x00000000007C1149 (DCS): CoreUtils::TempFilesManager::operator= + 0x39D299

2020-11-10 18:18:19.566 INFO EDCORE: 0x00000000017C4F14 (DCS): AmdPowerXpressRequestHighPerformance + 0xB1CF10

2020-11-10 18:18:19.566 INFO EDCORE: 0x00000000009ADCCE (DCS): CoreUtils::TempFilesManager::operator= + 0x589E1E

2020-11-10 18:18:19.566 INFO EDCORE: 0x0000000000017C24 (KERNEL32): BaseThreadInitThunk + 0x14

2020-11-10 18:18:19.567 INFO EDCORE: 0x000000000006CEA1 (ntdll): RtlUserThreadStart + 0x21

2020-11-10 18:18:19.979 INFO EDCORE: Minidump created.

[Flappie]

@Choke

Can you explain the steps to reproduce the crash?

A track would be welcome too.

[Choke]

@Flappie My track files aren't loading correctly, but I will save the track next time it happens. It happened consistenly on PGAW, and I never managed to get off the ground. While taxiing I was cycling waypoints as a sanity check, and to make sure the coordinates were correct. The first thing I noticed was the first digit on the data panel wasn't showing the first digit of the lattitude when cyling between Long and Lat. The other numbers were showing so I could read them, but the first digit was blank as the other ones cycled. I tried to input the coordinates again because I was worried they were incorrect, so I went REFLOLA, INPUT, and started to type the coordinates. This is when the crash happened both times. I will do some experiments and try to reproduce it, but fiddling with the data panel too quickly and whilst moving seems to be the killer for me.

 

Another issue that may or may not have been flagged is the formation lights on the left wing never turning off, even though the switch is off.

[Flappie]

Thank you. What map was it on? Syria?

 

The formation lights bug has been acknowledged by Heatblur already.

[Choke]

Persian Gulf, on the Hoggit server PGAW.

[Flappie]

I don't have the Hoggit PGAW password so I've tried it on an a Through the Inferno server, but it didn't crash.

If you can 100% reproduce it, it would be interesting to take a video of the procedure which leads to the crash.

[nsksleeper]

this should be the track, i tried it right now and didn't got in single player, now trying it in multiplayer. what i found out is that the landingstrip is close to the airports, so it shows the 9008 instead of coordinates, which would make some sence to me. keep you informed

Persian_Gulf_New_2020_targets-20201110-194907.trk

[Flappie]

It gets clearer: here's the link to the "INFO EDCORE: 0x0000000000065AE4 (AJS37): FlightComputer::FindClosestRunway + 0x7504" error message.

I'll try on my side with a road very close to a runway.

[Flappie]

I'll try on my side with a road very close to a runway.

I've tried it in PG on a multiplayer server. I thought maybe the bug was happening when there were 2 airports close to the selected coordinates. I've tried a road in Bandar Ababs, but so far I'm unable to crash my DCS doing so.

[Flappie]

Folks, it might not be a Viggen-only bug: https://forums.eagle.ru/forum/englis...68#post7139168

 

Maybe this is a PGAW script issue with current open beta.

[Flappie]

It seems this crash also happens when entering coordinates during cold start.

 

Weidlich crash

Pitbull1979 crash

[Pitbull1979]

But I have this problem in Caucas map

[Flappie]

@Pitbull1979I'm not sure this bug is limited to the Persian Gulf map. Are you able to reproduce your crash or has it happened only once?

 

[Pitbull1979]

No, I have it twice or more only in Caucas after last Stable update

[Flappie]

OK. I will try to reproduce the crash. I need some details. You said "When I created navigation data before cold start game go crashed".

 

Was the engine already ON or were you using the ground generator?

Which airfield were you at?

Were you entering coordinates or a 90xx code (9099, 9001, ...) ?

If it crashed when you recorded the waypoint, which waypoint was it ? (LS, B1, B2, Bx1, L-MAL...)

Edited January 2, 2021 by Flappie

[Pitbull1979]

AD 1. Engine ON, eject seat ARMED, canopy closed, all external lights ON.

AD 2. After activate cartrige I put 90990 and confirm by LS. DCS crash when I make mistake and put code seckond time

AD 3. Noworosyisk. 

Ps. After update Noworosyisk has more bugs. For exemple when I coldstarting with wingmen in SP mission, he Rolling out runway and go to the trees and... not explode:)))

Edited January 2, 2021 by Pitbull1979

[Flappie]

1 hour ago, Pitbull1979 said:

AD 1. Engine ON, eject seat ARMED, canopy closed, all external lights ON.

AD 2. After activate catrige I put 90990 and confirm by LS. DCS crash when I make mistake and put code seckond time

AD 3. Noworosyisk. 

Ps. After update Noworosyisk has more bugs. For exemple when I coldstarting with wingmen in SP mission, he Rolling out runway and go to the trees and... not explode:)))

I've tried this but couldn't reproduce your DCS crash.

I've also tried to replay the track in your zip file but can't because I don't have the VIRPIL airfield mode.

 

I couldn't reproduce wingmen crashing at Novo, either.

[Pitbull1979]

Tomorrow I will delete VIRPIL mode and check if game go crash. I let You know about results

[Flappie]

OK. And if it still crashes, please attach the resulting zip, so I can try to reproduce the issue.

Edited January 2, 2021 by Flappie

[Pitbull1979]

Ok, It's crashe's with VIRPIL and with clear tech folder. DCS report dcs.log-20210103-220931 is with VIRPIL and dcs.log-20210103-222412 is cleare (no mods). 

dcs.log-20210103-222412.zip dcs.log-20210103-220931.zip

[Flappie]

OK, thanks. I'm able to reproduce your crash thanks to the "222412" track. I'll look for an explanation this evening.

[Flappie]

Here is what happens in your track (starting from the initial AKT POS out mode).

1. REF LOLA out
2. Typing "9912"
3. AKT POS out
4. AKT POS in (entry "9912" gets erased from memory thanks to the out/in switch)
5. REF LOLA in
6. Typing "9921"
7. AKT POS in
8. REF LOLA in
9. Typing "9"
10. Typing "0" -> Display goes "000000" as if LS button had been pressed and code "992190" validated. Shouldn't it display "900000" instead?
11. Typing "9099"
12. Pressing LS -> Leads to an instant crash

I think the CK37 computer is not acting as it should during step 10. I'm assuming this malfunction is what leads to the crash as soon as the tape is loaded.

 

I've simplified the steps that lead to the crash:

1. REF LOLA in
2. Typing "990000" -> Display goes "000000" (it seems one can type any 4 digits code after "99", it will crash as well)
3. Typing "9099"
4. Pressing LS -> Leads to an instant crash

@RagnarDaCan you please try this? It's 100% reproducible on my side.

 

@Pitbull1979In the meantime, when you type something wrong, I suggest you use out/in switching to reset entry memory rather than AKT POS/REF LOLA switching.

[Pitbull1979]

Thank You Flappie. Working

[Flappie]

I don't know when, but it seems this bug has been fixed for good. No more crashes.