[0.9.2] crash when loading save with "Test mode" mod

[sparr]

I can't attach a save because it's too large, but here's the crash:

Thread 3 Crashed:
0 com.factorio.factorio 0x00000001000a439e CustomGui::setElementPosition(CustomGuiElement*, CustomGui::Position) + 286
1 com.factorio.factorio 0x00000001000a46f9 CustomGui::updateRootElementsPosition() + 217
2 com.factorio.factorio 0x00000001000a3f98 CustomGuiElement::buildWidgetRecursively() + 328
3 com.factorio.factorio 0x00000001000a3f8f CustomGuiElement::buildWidgetRecursively() + 319
4 com.factorio.factorio 0x00000001000a3f8f CustomGuiElement::buildWidgetRecursively() + 319
5 com.factorio.factorio 0x00000001000a3d59 CustomGui::loadWidgets() + 73
6 com.factorio.factorio 0x000000010006d58c Scenario::setupAfterGameWasCreated() + 76
7 com.factorio.factorio 0x00000001000696ab AppManager::processLoadGameDialog() + 347
8 com.factorio.factorio 0x000000010006634f AppManager::process(bool&) + 511
9 com.factorio.factorio 0x00000001002b8363 tickStep(boost::cpu_timer&, boost::cpu_timer&, bool) + 83
10 com.factorio.factorio 0x00000001002b90b1 run(char*, bool) + 305
11 com.factorio.factorio 0x00000001002ba857 _al_mangled_main + 4551
12 com.factorio.factorio 0x0000000100670747 +[AllegroAppDelegate app_main:] + 23
13 com.apple.Foundation 0x00007fff9093170b __NSThread__main__ + 1318
14 libsystem_pthread.dylib 0x00007fff92421899 _pthread_body + 138
15 libsystem_pthread.dylib 0x00007fff9242172a _pthread_start + 137
16 libsystem_pthread.dylib 0x00007fff92425fc9 thread_start + 13

[kovarex]

Thank you for the report, it is caused by our fix of the repositioning the script based gui elements correctly.

It is now fixed in 0.9.3

[MF-]

How come he has such nice backtraces with names, while
all linux backtraces I've seen / made are only numbers?

EDIT: I guess that's a factorio-produced crash log instead of a segfault?

[sparr]

How come he has such nice backtraces with names, while
all linux backtraces I've seen / made are only numbers?

EDIT: I guess that's a factorio-produced crash log instead of a segfault?

That's an OSX crash log. Unless they are building with fewer debugging symbols on linux you should be able to get just as nice of a report from a core dump or a debugger on linux.

[MF-]

I guess that could be the case.

[DRBLN]

Please, could you share a hotfix?

[N3X15]

Please, could you share a hotfix?

This error appears to be coming from the C++ code, so I don't think a lua-based workaround ("hotfix") would be possible.

[rk84]

"Activate" command can be used to toggle gui off, before saving.

Code: Select all

remote.call("testmode","activate")