Page 1 of 1

linux x64 0.13.0 segfaults

Posted: Tue Jun 28, 2016 11:37 pm
by golfmiketango
Got like three of these now in early game, no mods. They're rare-ish and proceeded by glitchy pauses. I'd bet money it's a factorio-specific bug, not a heat crisis or similar issue. I don't know exactly what I'm doing when they happen but at least once I was mucking around with stuff in a chest. Each time I've been "busy" (it never happened when I was just sitting there thinking, so the backtrace's suggestion that it's something to do with input could be right).

vital statistics:

radeon open-source drivers
kde frameworks 5
multimonitor
xrender compositor w/ full-screen apps (ie factorio) non-composited
up-to-date gentoo ~amd64
kernel 4.5.2
everything compiled by gcc 5.4 (a few 5.3-compiled remnants may have slipped through)
amd piledriver
a lot of ram

Code: Select all

   0.000 2016-06-28 14:57:19; Factorio 0.13.0 (Build 22782, linux64, alpha)
   0.024 Operating system: Linux (Gentoo 2.2)
   0.024 Program arguments: "./bin/x64/factorio" 
   0.024 Read data path: /home/greg/factorio13/data
   0.024 Write data path: /home/greg/factorio13
   0.024 Binaries path: /home/greg/factorio13/bin
   0.044 Graphics options: [FullScreen: true] [VSync: true] [UIScale: 100%] [MultiSampling: OFF] [Graphics quality: normal] [Video memory usage: high] [Light scale: 30%] [Screen: 0] [DXT: auto]
   0.187 Available display adapters: 3
   0.187  [0]: resolution 1680x1050px at [2560,108]
   0.187  [1]: resolution 1280x1024px at [4240,108]
   0.187  [2]: resolution 2560x1440px at [0,0]
   0.187 Create display on adapter 0. Size 1280x720 at position [190, 147].
   0.684 Initialised OpenGL:[0] Gallium 0.4 on AMD PITCAIRN (DRM 2.43.0, LLVM 3.8.0); driver: 3.0 Mesa 11.2.2
   0.685     Video memory size (dedicated/total available/current available): 2048/4091/2032 MB
   0.685     Video memory size (free/largest block/aux. free/aux. largest block): 2032/2032/2042/2042 MB
   0.856 Loading mod core 0.0.0 (data.lua)
   0.859 Loading mod base 0.13.0 (data.lua)
   0.948 Checksum for core: 1399339455
   0.948 Checksum for mod base: 2039900801
   1.422 Initial atlas bitmap size is 16384
   1.424 Created atlas bitmap 16384x4618
   1.425 Created atlas bitmap 4096x4088
   1.426 Created atlas bitmap 4096x4076
   1.426 Created atlas bitmap 4096x3988
   1.427 Created atlas bitmap 4096x1104
  15.678 Sprites loaded
  15.678 Convert atlas 4096x4088 to: compressed 
  17.072 Convert atlas 4096x4076 to: compressed 
  18.388 Convert atlas 4096x3988 to: compressed 
  19.591 Convert atlas 4096x1104 to: trilinear-filtering 
  19.644 Loading sounds...
  21.118 Custom inputs active: 0
  21.118 Info Updater.cpp:751: Downloading https://www.factorio.com/updater/get-available-versions?username=golfmiketango&token=<private>&apiVersion=2
  22.083 0 packages available to download (experimental updates enabled).
  22.121 Factorio initialised
  27.089 Loading map /home/greg/factorio13/saves/_autosave2.zip
  27.121 Info Scenario.cpp:129: Map version 0.13.0-93
  27.799 Checksum for script /home/greg/factorio13/temp/currently-playing/control.lua: 900459546
2206.198 Loading map /home/greg/factorio13/saves/_autosave3.zip
2206.230 Info Scenario.cpp:129: Map version 0.13.0-93
2206.947 Checksum for script /home/greg/factorio13/temp/currently-playing/control.lua: 900459546
2537.446 Error CrashHandler.cpp:183: Received SIGSEGV
Factorio crashed. Generating symbolized stacktrace, please wait ...
2538.257 Warning Logger.cpp:360: Symbols.size() == 13, usedSize == 11
#0  0x8f2d7d in Logger::logStacktrace(StackTraceInfo*) at /tmp/factorio-v6JSYj/src/Util/Logger.cpp:382
#1  0x800a6b in CrashHandler::writeStackTrace() at /tmp/factorio-v6JSYj/src/Util/CrashHandler.cpp:84
#2  0x800ce4 in CrashHandler::SignalHandler(int) at /tmp/factorio-v6JSYj/src/Util/CrashHandler.cpp:188
#3  0x3691632e60 in ?? at ??:0
#4  0xb8bf63 in InputHandlerAllegro::processGuiAction(GuiAction const&, InputState const&) at /tmp/factorio-v6JSYj/src/Input/InputHandlerAllegro.cpp:450 (discriminator 1)
#5  0x6b6d76 in QueuedGuiActionHandler::processGuiActions() at /tmp/factorio-v6JSYj/src/Input/GuiActionHandler.cpp:24
#6  0xbd433c in MainLoop::processAllegroEvents(bool) at /tmp/factorio-v6JSYj/src/MainLoop.cpp:185
#7  0xc12915 in processAllegroEvents at /tmp/factorio-v6JSYj/src/MainLoop.cpp:205
#8  0xc8313e in MainLoop::run(boost::filesystem::path const&, boost::filesystem::path const&, bool, bool, bool, std::string const&, boost::filesystem::path const&, MainLoop::HeavyMode, unsigned int) [clone .constprop.68090] at /tmp/factorio-v6JSYj/src/MainLoop.cpp:547
#9  0x3691620790 in run at /tmp/factorio-v6JSYj/src/MainLoop.cpp:691
#10 0x40a459 in main at /tmp/factorio-v6JSYj/src/Main.cpp:733
2539.207 Warning Logger.cpp:360: Symbols.size() == 12, usedSize == 10
2539.207 Error CrashHandler.cpp:85: Map tick at moment of crash: 1005160
2539.207 Error Util.cpp:77: Unexpected error occurred. You can help us to solve the problem by posting the contents of the log file on the Factorio forums.

Re: linux x64 0.13.0 segfaults

Posted: Fri Jul 01, 2016 2:59 pm
by kovarex
That looks confusing to me, can you provide logs when it happens some other time, so I can see if it crashes on the same place?

Re: linux x64 0.13.0 segfaults

Posted: Fri Jul 01, 2016 4:58 pm
by TruePikachu
Frame 3 looks very suspicious to me, since it makes me think a foreign call is responsible. My next course of action would be to run under gdb, and get a backtrace from there if/when this happens again, so we can tell what module (and hopefully function) is responsible.

Devs: It might be worth it to allow Factorio to dump core with a command line parameter; this would allow people who know how to operate a debugger to get more information from the dump than CrashHandler.cpp can from the live executable, without having to keep a debugger attached all the time, which can be very useful for those bugs which are random.

Re: linux x64 0.13.0 segfaults

Posted: Fri Jul 01, 2016 8:39 pm
by posila
It is 0.13.0, so my money are on bug causing crash when shift-clicking item stacks in GUI, which was fixed in 0.13.1.

Re: linux x64 0.13.0 segfaults

Posted: Fri Jul 01, 2016 11:46 pm
by Rseding91
Most likely this was fixed in 0.13.1.

If you can reproduce it -not in 0.13.0- please make a new bug report with the crash log :)