glx wrote:

KERNELBASE.dll!RaiseException() + 0x3d octets
openttd.exe!raise(int signum=0x00000001) Ligne 597 C
openttd.exe!abort() Ligne 78 C
openttd.exe!error(const char * s=0x000000013f5bb540, ...) Ligne 131 C++
openttd.exe!DoCommandPInternal(unsigned int tile=0xffffffff, unsigned int p1=0x00000004, unsigned int p2=0x00000000, unsigned int cmd=0x00000017, void (const CommandCost &, unsigned int, unsigned int, unsigned int)* callback=0x0000000000000000, const char * text=0x0000000000000000, bool my_cmd=false, bool estimate_only=false) Ligne 659 + 0x34 octets C++
openttd.exe!AIObject::DoCommand(unsigned int tile=0x00000002, unsigned int p1=0x00006f9f, unsigned int p2=0x0457fee0, unsigned int cmd=0x00000004, const char * text=0x0000000000000000, void (AIInstance ) callback=0x000000013f3918b4) Ligne 216 C++
openttd.exe!DoCommandReturnBuildTunnel1(AIInstance * instance=0x000000000457fee0) Ligne 62 + 0xf5 octets C++
openttd.exe!AIInstance::GameLoop() Ligne 330 + 0x6 octets C++
openttd.exe!StateGameLoop() Ligne 1203 + 0x58 octets C++
openttd.exe!GameLoop() Ligne 1285 C++
openttd.exe!VideoDriver_Win32::MainLoop() Ligne 889 C++
openttd.exe!ttd_main(int argc=0x3f200000, char * * argv=0x000000013f5fd208) Ligne 758 C++
openttd.exe!WinMain(HINSTANCE_ * hInstance=0x0000000000000000, HINSTANCE__ * hPrevInstance=0x0000000000000000, char * lpCmdLine=0x000000013f200000, int nCmdShow=0x3f200000) Ligne 430 C++
openttd.exe!__tmainCRTStartup() Ligne 263 + 0x21 octets C
kernel32.dll!BaseThreadInitThunk() + 0xd octets
ntdll.dll!RtlUserThreadStart() + 0x21 octets

This comment was imported from FlySpray: https://bugs.openttd.org/task/3578#comment7483

Rubidium wrote:

Very hard to reproduce reliably though because it's something the AI causes, but don't know exactly what circumstances are needed.

This comment was imported from FlySpray: https://bugs.openttd.org/task/3578#comment7487

Rubidium wrote:

In r18961; loading the last autosave doesn't reproduce it.

# 0 0x00002aaaadad6f55 in raise () from /lib/libc.so.6
No symbol table info available.
# 1 0x00002aaaadad9d90 in abort () from /lib/libc.so.6
No symbol table info available.
# 2 0x00002aaaadad007a in _assert_fail () from /lib/libc.so.6
No symbol table info available.
# 3 0x000000000054ef6a in DoCommandPInternal (tile=41325, p1=, p2=,
cmd=23, callback=, text=0x0, my_cmd=false, estimate_only=false)
at /home/rubidium/openttd/clean/src/command.cpp:659
proc = 0x684e20 <CmdBuildRoad(unsigned int, DoCommandFlag, unsigned int, unsigned int, char const*)>
cmd_flags =
flags = DC_NONE
skip_test = false
res = {expense_type = EXPENSES_CONSTRUCTION, cost = {m_value = 4620}, message = 65535, success = true}
res2 = {expense_type = EXPENSES_CONSTRUCTION, cost = {m_value = 770}, message = 65535, success = true}
_PRETTY_FUNCTION = "CommandCost DoCommandPInternal(TileIndex, uint32, uint32, uint32, void ()(const CommandCost&, TileIndex, uint32, uint32), const char, bool, bool)"
cmd_id = 23 '\027'
test_and_exec_can_differ = false
# 4 0x0000000000518776 in AIObject::DoCommand (tile=41325, p1=1, p2=0, cmd=23, text=,
callback=0x521e40 <_DoCommandReturnBuildTunnel2>) at /home/rubidium/openttd/clean/src/ai/api/ai_object.cpp:213
estimate_only = false
res = {expense_type = EXPENSES_CONSTRUCTION, cost = {m_value = 7531690}, message = 0, success = false}
# 5 0x0000000000521f89 in _DoCommandReturnBuildTunnel1 (instance=0xd93)
at /home/rubidium/openttd/clean/src/ai/api/ai_tunnel.cpp:62
No locals.
# 6 0x00000000004e2775 in AIInstance::GameLoop (this=0x13335a0)
at /home/rubidium/openttd/clean/src/ai/ai_instance.cpp:330
No locals.

This comment was imported from FlySpray: https://bugs.openttd.org/task/3578#comment7488

stickedy wrote:

Yeah, playing the last savegame doesn't lead to another crash! I'll play this savegame further to look if it will happen again...

This comment was imported from FlySpray: https://bugs.openttd.org/task/3578#comment7493

stickedy wrote:

I have discovered something new to this case: Yesterday I played the savegame again and the game crashes many years after the first crash and it happen shortly after autosaving! I guess it has something to do with saving the game.

This comment was imported from FlySpray: https://bugs.openttd.org/task/3578#comment7496

Rubidium wrote:

The first crash is in the middle of the month (16th), so it's unlikely it's related to autosaving. Unless autosaving takes half a month for you.

This comment was imported from FlySpray: https://bugs.openttd.org/task/3578#comment7497

frosch wrote:

Just for the record. Attached diff causes the regression to trigger the issue.

Attachments

• reproducefs3578.diff (0.53 KiB)

This comment was imported from FlySpray: https://bugs.openttd.org/task/3578#comment7500

frosch closed the ticket.

Reason for closing: Fixed

in r18974

This comment was imported from FlySpray: https://bugs.openttd.org/task/3578