LateralGM don't like me :(

[fervi]

http://gmclan.org/uploader/6184/error.txt

Fervi

[faissaloo]

Other people seem to be having this issue but noone seems able to reproduce it.

[fervi]

Maybe it's something with newest version of Java

openjdk version "1.8.0_102"
OpenJDK Runtime Environment (build 1.8.0_102-8u102-b14.1-2-b14)
OpenJDK 64-Bit Server VM (build 25.102-b14, mixed mode)

Fervi

EDIT: It's not a LateralGM bug, because i can run it (on 32 bit java, it give me a enigma error, but lateralGM works [but cannot compile]). It's something between Enigma and LateralGM

[Goombert]

Hi Fervi! Please try this fix by Josh, this is a known bug now:

http://enigma-dev.org/forums/index.php?topic=2669.msg25839#msg25839

[fervi]

Hi Rob

http://87.204.44.63:92/logi/
http://pastebin.com/vbGqGZc9

I am on holidays, but in this computer (not mine) i have errors like this.

9108 is before Josh patch
9663 is after Josh patch

Fervi

[Goombert]

Yeah it's becoming super common on Linux now fervi and we are really having a hard time diagnosing it.
https://github.com/enigma-dev/enigma-dev/issues/944

Code: [Select]


Stack: [0x00007f1bb44a6000,0x00007f1bb45a7000],  sp=0x00007f1bb45a2a20,  free space=1010k
Native frames: (J=compiled Java code, j=interpreted, Vv=VM code, C=native code)
C  [libcompileEGMf.so+0xaccd6]  jdip::read_qualified_definition(jdi::lexer*, jdi::definition_scope*, jdip::token_t&, jdip::context_parser*, jdi::error_handler*)+0x66
C  [libcompileEGMf.so+0x97fb9]  jdip::context_parser::handle_scope(jdi::definition_scope*, jdip::token_t&, unsigned int)+0x13b9
C  [libcompileEGMf.so+0x9ab20]  jdip::context_parser::handle_class(jdi::definition_scope*, jdip::token_t&, int)+0x920
C  [libcompileEGMf.so+0xa9a27]  jdip::read_type(jdi::lexer*, jdip::token_t&, jdi::definition_scope*, jdip::context_parser*, jdi::error_handler*)+0x87

Java frames: (J=compiled Java code, j=interpreted, Vv=VM code)
j  com.sun.jna.Native.invokePointer(JI[Ljava/lang/Object;)J+0
j  com.sun.jna.Function.invokePointer(I[Ljava/lang/Object;)Lcom/sun/jna/Pointer;+6
j  com.sun.jna.Function.invoke([Ljava/lang/Object;Ljava/lang/Class;Z)Ljava/lang/Object;+486
j  com.sun.jna.Function.invoke(Ljava/lang/Class;[Ljava/lang/Object;Ljava/util/Map;)Ljava/lang/Object;+262
j  com.sun.jna.Library$Handler.invoke(Ljava/lang/Object;Ljava/lang/reflect/Method;[Ljava/lang/Object;)Ljava/lang/Object;+316
j  com.sun.proxy.$Proxy0.definitionsModified(Ljava/lang/String;Ljava/lang/String;)Lorg/enigma/backend/EnigmaDriver$SyntaxError;+20
j  org.enigma.backend.EnigmaSettings.commitToDriver(Lorg/enigma/backend/EnigmaDriver;)Lorg/enigma/backend/EnigmaDriver$SyntaxError;+30
j  org.enigma.EnigmaRunner$3.run()V+175
v  ~StubRoutines::call_stub

The stack trace to me shows its always occurring in syntax checking. Which is weird because I don't think syntax checking is even run at startup. Did you happen to change any definitions in the ENIGMA Game Settings frame?

Can you try running with the following flag on java -jar lateralgm.jar?

Code: [Select]

-Djna.nosys=true

[fervi]

1)Did you happen to change any definitions in the ENIGMA Game Settings frame?

No, it breaks down immediately after compiling their data (first run). I cannot change anything

2) -Djna.nosys=true

Yes, I lateralGM it very well, but I have no way to compile the project (can be disabled).

When you start getting a form Yes / No from the titlebar - "File not found" and the question "Would you like to choose a different file?"

Fervi

[Goombert]

No you do it this way fervi:

Code: [Select]

java -jar -Djna.nosys=true lateralgm.jar
Also please try reverting one of my commits that I am suspicious of:
https://github.com/enigma-dev/enigma-dev/commit/5b99151173b1dad9acb059f9e99a75974801d42b

Code: [Select]

git revert -m 1 5b99151173b1dad9acb059f9e99a75974801d42b
Another thing to try is making a backup of "enigma-dev/ENIGMAsystem/SHELL/Universal_System/Extensions" and then deleting everything in that "Extensions" folder and starting ENIGMA with no extensions.

[fervi]

1) Logdjna
2)
error: Mainline was specified but commit 5b99151173b1dad9acb059f9e99a75974801d42b is not a merge.
fatal: revert failed

So i download
https://github.com/enigma-dev/enigma-dev/archive/5b99151173b1dad9acb059f9e99a75974801d42b.zip

And also this
https://github.com/enigma-dev/enigma-dev/archive/7fb87c34d89cebeec545016ff2c5036c838c93c1.zip

Log5b99 and Log7fb8

3) Without extensions also crash
lognoextensions

http://87.204.44.63:92/logi/

Fervi

[Goombert]

Yeah those were pull requests I gave you I am sorry. I would have to go use the command line to dig up the commits that were used for the actual merge on GitHub.

Anyway, do you have another platform to try ENIGMA on for the time being fervi? I really don't know how to work around this issue. The reason it crept up is because JDI has poor support for templates and other things and we keep using them in various parts of ENIGMA and something must have changed in newer GCC's or another and it's bringing out crashes in the compiler. Even worse is that only people on Linux are having the problem and nobody has reported it on Windows. I am going to try to nag Josh enough into tracing down the bug this weekend.

[fervi]

I do not know if this post will help, but ...
1) It's certainly not the fault of Java, because the upload older version does not solve the problem
2) It's certainly not the fault of the compiler, because the conversion of it to clang (link clang++ asg ++) library compiles correctly, but it crashes lateralgm.

Maybe it will help you xD
On debian 8 works very well, so maybe some library used something breaks down (Debian 8 in Debian 9 by chroot)

Fervi

[Dragonite]

Getting the same problem in Fedora 25. I can't even start the LateralGM with 64-bits Java, and I'm unable to use ENIGMA if I run it with 32-bits Java.