2025-06-14 04:42 CEST
  

View Issue Details [ Jump to Notes ] [ Issue History ] [ Print ]
IDProjectCategoryView StatusDate SubmittedLast Update
0002585Spring engineGeneralpublic2011-07-31 17:552011-08-01 12:20
Reporterabma 
Assigned ToKloot 
PrioritynormalSeveritycrashReproducibilityalways
StatusresolvedResolutionfixed 
Product Version 
Target Version83.0Fixed in Version0.82.7+git 
Summary0002585: (master) assert in UnitHandler.cpp:273
Descriptionhappened in a KAIKvsKAIK testgame (testmonkey found it! :)

(see attachment for more logs if needed...)
Additional Informationspring: /home/abma/dev/spring/master/rts/Sim/Units/UnitHandler.cpp:273: void CUnitHandler::Update(): Assertion `unit->pos.z >= -(float3::maxzpos * 16.0f)' failed.

Program received signal SIGABRT, Aborted.
#0 0x00007ffff43fed05 in raise (sig=6) at ../nptl/sysdeps/unix/sysv/linux/raise.c:64
        resultvar = 0
        pid = <value optimized out>
        selftid = 7659
#1 0x00007ffff4402ab6 in abort () at abort.c:92
        save_stage = 2
        act = {__sigaction_handler = {sa_handler = 0x111, sa_sigaction = 0x111}, sa_mask = {__val = {140737292396174, 23028384, 140737292388058, 140737488345296, 140737291425890, 206158430232, 140737488345312, 140737488345088, 140737291337076, 206158430256, 140737488345336, 140736817887280, 140736817266720, 5, 4193823639842088488, 140737488349684}}, sa_flags = -195967270, sa_restorer = 0x15f5698}
        sigs = {__val = {32, 0 <repeats 15 times>}}
#2 0x00007ffff43f77c5 in __assert_fail (assertion=0x15f5fa0 "unit->pos.z >= -(float3::maxzpos * 16.0f)", file=<value optimized out>, line=273, function=<value optimized out>) at assert.c:81
        buf = 0x7fffd8097830 "spring: /home/abma/dev/spring/master/rts/Sim/Units/UnitHandler.cpp:273: void CUnitHandler::Update(): Assertion `unit->pos.z >= -(float3::maxzpos * 16.0f)' failed.\n"
0000003 0x0000000000f78837 in CUnitHandler::Update (this=0x5910ec0) at /home/abma/dev/spring/master/rts/Sim/Units/UnitHandler.cpp:273
        unit = 0x7fffd81db9e0
        moveType = 0x7fffd80e1c00
        myScopedTimerFromMakro = {<BasicTimer> = {<boost::noncopyable_::noncopyable> = {<No data fields>}, name = {static npos = <optimized out>, _M_dataplus = {<std::allocator<char>> = {<__gnu_cxx::new_allocator<char>> = {<No data fields>}, <No data fields>}, _M_p = 0x7fffd8158728 "Unit::MoveType::Update"}}, starttime = 292988}, autoShowGraph = false}
        usi = {_M_node = 0x7fffd81964e0}
        __PRETTY_FUNCTION__ = "void CUnitHandler::Update()"
0000004 0x000000000095c077 in CGame::SimFrame (this=0x7fffe41316a0) at /home/abma/dev/spring/master/rts/Game/Game.cpp:1510
        cputimer = {<BasicTimer> = {<boost::noncopyable_::noncopyable> = {<No data fields>}, name = {static npos = <optimized out>, _M_dataplus = {<std::allocator<char>> = {<__gnu_cxx::new_allocator<char>> = {<No data fields>}, <No data fields>}, _M_p = 0x7fffd81ba0c8 "Game::SimFrame"}}, starttime = 292988}, autoShowGraph = true}
        forced = {<BasicTimer> = {<boost::noncopyable_::noncopyable> = {<No data fields>}, name = {static npos = <optimized out>, _M_dataplus = {<std::allocator<char>> = {<__gnu_cxx::new_allocator<char>> = {<No data fields>}, <No data fields>}, _M_p = 0x7fffd80f9478 "Game::SimFrame (Update)"}}, starttime = 292988}, autoShowGraph = false}
0000005 0x00000000009df41b in CGame::ClientReadNet (this=0x7fffe41316a0) at /home/abma/dev/spring/master/rts/Game/NetCommands.cpp:364
        inbuf = 0x7fffe5068090 "\002{\276\344\377\177"
        dataLength = 1
        packetCode = 2 '\002'
        packet = {px = 0x7fffe53f1870, pn = {pi_ = 0x7fffe4be7b40}}
        procstarttime = 292975
        __PRETTY_FUNCTION__ = "void CGame::ClientReadNet()"
#6 0x0000000000958a8e in CGame::Update (this=0x7fffe41316a0) at /home/abma/dev/spring/master/rts/Game/Game.cpp:906
        timeNow = 292975
        difTime = 66
        dif = 0.0659999996
#7 0x00000000010261cd in SpringApp::UpdateSim (this=0x7fffffffe530, ac=0x7fffe41316a0) at /home/abma/dev/spring/master/rts/System/SpringApp.cpp:980
        ret = false
#8 0x0000000001026243 in SpringApp::Update (this=0x7fffffffe530) at /home/abma/dev/spring/master/rts/System/SpringApp.cpp:1058
        ret = 1
#9 0x0000000001026504 in SpringApp::Run (this=0x7fffffffe530, argc=6, argv=0x7fffffffe6f8) at /home/abma/dev/spring/master/rts/System/SpringApp.cpp:1162
No locals.
0000010 0x00000000010012e6 in MainFunc (argc=6, argv=0x7fffffffe6f8, ret=0x7fffffffe5dc) at /home/abma/dev/spring/master/rts/System/Main.cpp:54
        app = {cmdline = 0x1de44a0, lastRequiredDraw = 23272, static ogc = 0x0}
#11 0x00000000010015d4 in Run (argc=6, argv=0x7fffffffe6f8) at /home/abma/dev/spring/master/rts/System/Main.cpp:65
        ret = -1
        err = 0x214
0000012 0x0000000001001674 in main (argc=6, argv=0x7fffffffe6f8) at /home/abma/dev/spring/master/rts/System/Main.cpp:95
No locals.
TagsNo tags attached.
Checked infolog.txt for Errors
Attached Files

-Relationships
has duplicate 0002517resolvedKloot (master) desync+SIGSEGV in PathEstimator.cpp:494 
+Relationships

-Notes

~0007156

abma (administrator)

2011-07-31 17:55

 2011.07.31_16-34-13.log: raise (sig=6) at ../nptl/sysdeps/unix/sysv/linux/raise.c:64
2011.07.31_16-39-20.log: raise (sig=6) at ../nptl/sysdeps/unix/sysv/linux/raise.c:64
2011.07.31_16-42-53.log: raise (sig=6) at ../nptl/sysdeps/unix/sysv/linux/raise.c:64
2011.07.31_16-43-41.log: raise (sig=6) at ../nptl/sysdeps/unix/sysv/linux/raise.c:64
2011.07.31_16-47-38.log: raise (sig=6) at ../nptl/sysdeps/unix/sysv/linux/raise.c:64
2011.07.31_16-50-10.log: raise (sig=6) at ../nptl/sysdeps/unix/sysv/linux/raise.c:64
2011.07.31_16-51-07.log: raise (sig=6) at ../nptl/sysdeps/unix/sysv/linux/raise.c:64
2011.07.31_16-59-59.log: CalculateCofactor<float> (m=0x7fffffffcd10, ei=3, ej=0) at /home/abma/dev/spring/master/rts/System/Matrix44f.cpp:358
2011.07.31_17-26-17.log: raise (sig=6) at ../nptl/sysdeps/unix/sysv/linux/raise.c:64
2011.07.31_17-31-53.log: raise (sig=6) at ../nptl/sysdeps/unix/sysv/linux/raise.c:64
2011.07.31_17-33-43.log: raise (sig=6) at ../nptl/sysdeps/unix/sysv/linux/raise.c:64
2011.07.31_17-39-44.log: raise (sig=6) at ../nptl/sysdeps/unix/sysv/linux/raise.c:64
2011.07.31_17-41-59.log: raise (sig=6) at ../nptl/sysdeps/unix/sysv/linux/raise.c:64
2011.07.31_17-45-54.log: raise (sig=6) at ../nptl/sysdeps/unix/sysv/linux/raise.c:64

~0007157

abma (administrator)

2011-07-31 17:57

 Note: Matrix44f.cpp:358 is from Kaik, see 0002586

~0007159

abma (administrator)

2011-07-31 18:14

Last edited: 2011-07-31 18:16

 an example pos:
(gdb) print unit->pos.x
$2 = 65880.7266
(gdb) print unit->pos.y
$3 = 181.863312
(gdb) print unit->pos.z
$4 = -39539.9219

static maxxpos = 4095, static maxzpos = 4095


(gdb) print unit->speed
$7 = {static binder = {class_ = 0x1de4d00, base = 0x0, flags = creg::CF_None, memberRegistrator = 0x1d85ba0, name = 0x1609a4a "float3", size = 12,
    constructor = 0x104525a <float3::_ConstructInstance(void*)>, destructor = 0x104528a <float3::_DestructInstance(void*)>, nextBinder = 0x1d85b40},
  static memberRegistrator = 0x1d85c30, static CMP_EPS = 9.99999975e-05, static NORMALIZE_EPS = 9.99999996e-13, static maxxpos = 4095, static maxzpos = 4095,
  x = 1886.7876, y = 0.0755235404, z = -1187.33984}


more/different vars needed?

~0007161

Kloot (developer)

2011-07-31 18:31

 The crashes in CMatrix are again symptoms, now I need demos from games that triggered UnitHandler asserts but did not crash elsewhere (fex. 2011.07.31_17-45-54.log).

~0007162

abma (administrator)

2011-07-31 20:07

Last edited: 2011-07-31 20:31

 attached hopefully the correct file (demofilename != name in log)

update: aaah, this is because the file is renamed on exit, and this isn't done because of the assert...

~0007163

Kloot (developer)

2011-07-31 20:44

 Sorry, playing the demo is of no help (it does not reproduce the assertion).

Can you check if any of those other demos even crash deterministically?

~0007164

abma (administrator)

2011-07-31 20:46

Last edited: 2011-07-31 20:47

 end of demo is at f=0012307 (~6:48). hm... i guess to reproduce, you need to get the same frame renderd as me and have the same viewport, when it triggered the assertion. (?)

i hope i already said: when i run my test at 30x speed, i very rarely got crashes, this is why i added the widget, that controls speed, to keep the frame-rate high, see https://github.com/spring/testmonkey/blob/master/LuaUI/Widgets/test.lua

edit: yes, will try!

~0007165

abma (administrator)

2011-07-31 21:07

Last edited: 2011-07-31 21:14

 i didn't test all, i tested only the first five. only the last one crashes in about 1/5 tries or so (20110731_164348_unnamed_0.82+.4.sdf )... i ran it at 20x speed

anything that could help here: http://pastebin.com/XK2E0aYz ?

~0007166

abma (administrator)

2011-07-31 21:29

Last edited: 2011-07-31 21:34

 replaying the same demo i got the same assert, but with a different unit: http://pastebin.com/SkJmE70F

(gdb) print unit->unitDef->name
$3 = {static npos = <optimized out>,
  _M_dataplus = {<std::allocator<char>> = {<__gnu_cxx::new_allocator<char>> = {<No data fields>}, <No data fields>}, _M_p = 0x5b900f8 "armwar"}}
(gdb) print unit->pos
$4 = {static binder = {class_ = 0x1de4d00, base = 0x0, flags = creg::CF_None,
    memberRegistrator = 0x1d85ba0, name = 0x1609a4a "float3", size = 12,
    constructor = 0x104525a <float3::_ConstructInstance(void*)>,
    destructor = 0x104528a <float3::_DestructInstance(void*)>, nextBinder = 0x1d85b40},
  static memberRegistrator = 0x1d85c30, static CMP_EPS = 9.99999975e-05,
  static NORMALIZE_EPS = 9.99999996e-13, static maxxpos = 4095, static maxzpos = 4095,
  x = 9880.31055, y = 75952.7734, z = 65819.8984}
(gdb) print unit->speed
$5 = {static binder = {class_ = 0x1de4d00, base = 0x0, flags = creg::CF_None,
    memberRegistrator = 0x1d85ba0, name = 0x1609a4a "float3", size = 12,
    constructor = 0x104525a <float3::_ConstructInstance(void*)>,
    destructor = 0x104528a <float3::_DestructInstance(void*)>, nextBinder = 0x1d85b40},
  static memberRegistrator = 0x1d85c30, static CMP_EPS = 9.99999975e-05,
  static NORMALIZE_EPS = 9.99999996e-13, static maxxpos = 4095, static maxzpos = 4095,
  x = 266.033691, y = 2704.60474, z = 2292.21777}

so i think i can answer my question myself: not useful this info :-/

in the demo, the units seems to bump each other, see screenshot.jpg, maybe that could be the cause?

(just a guess, i try only as much info as i can)

screenshot is taken after the crash (luckily, the new opengl desktops keeps the window-contents)

~0007167

Kloot (developer)

2011-07-31 21:35

 "hm... i guess to reproduce, you need to get the same frame renderd as me and have the same viewport, when it triggered the assertion."

no, the UnitHandler asserts are checked before and after each individual MoveType::Update call, and the one that triggers is on line 273 (ie. the post-update assert) so it should be synced.

"anything that could help here: http://pastebin.com/XK2E0aYz [^] ?"

yes, look closely for a specific armjanus flying outside the map :)

~0007168

abma (administrator)

2011-07-31 21:36

 hu? spring stacktrace is different?

(gdb) cont
Continuing.
[f=0024452] Error: Aborted (SIGABRT) in spring 0.82+.4.0 (Debug)
[f=0024452] Error: Stacktrace:
[f=0024452] Error: <0> /lib/x86_64-linux-gnu/libc.so.6(+0x33d80) [0x7ffff43fed80]
[f=0024452] Error: <1> /lib/x86_64-linux-gnu/libc.so.6(gsignal+0x35) [0x7ffff43fed05]
[f=0024452] Error: <2> /lib/x86_64-linux-gnu/libc.so.6(abort+0x186) [0x7ffff4402ab6]
[f=0024452] Error: <3> /lib/x86_64-linux-gnu/libc.so.6(__assert_fail+0xf5) [0x7ffff43f77c5]
[f=0024452] Error: <4> /home/abma/dev/spring/master/rts/Lua/LuaSyncedRead.cpp:3854
[f=0024452] Error: <5> /var/tmp/local/spring/master/bin/spring() [0x95c077]
[f=0024452] Error: <6> /var/tmp/local/spring/master/bin/spring() [0x9df41b]
[f=0024452] Error: <7> /var/tmp/local/spring/master/bin/spring() [0x958a8e]
[f=0024452] Error: <8> /home/abma/dev/spring/master/rts/Map/SM3/terrain/TerrainUtil.cpp:209
[f=0024452] Error: <9> /home/abma/dev/spring/master/rts/Map/SM3/terrain/TerrainUtil.cpp:217

(looks a bit broken...)

~0007169

abma (administrator)

2011-07-31 22:03

Last edited: 2011-07-31 22:36

 aahhh! found an easy way to reproduce:


/cheat
/give 100 armjanus

let them all move to an unit in the middle of the block!

http://pastebin.com/9e0GFBGj

/cheat
/give all
/give all 1 seems to cause this:
http://pastebin.com/XDzZxQPn or a SIGFPE: http://pastebin.com/GX9RJyTj

we are back in business! ^^

~0007170

abma (administrator)

2011-07-31 22:45

Last edited: 2011-07-31 22:47

 /give 100 armjanus seems to be fixed! :)

git describe
0.82.3-2989-ge6fd066

/cheat /give all 0+1 crash:
http://pastebin.com/Gz2sYXUj or the SIGFPE: http://pastebin.com/dQNzzZg0

~0007171

abma (administrator)

2011-07-31 23:58

Last edited: 2011-08-01 00:00

 git describe
0.82.3-2991-g2233071

spring: /home/abma/dev/spring/master/rts/Sim/MoveTypes/GroundMoveType.cpp:638: virtual void CGroundMoveType::ImpulseAdded(const float3&): Assertion `speed.SqLength() < 1e6f' failed.


http://pastebin.com/2VadHJuq

~0007172

Kloot (developer)

2011-08-01 00:21

Last edited: 2011-08-01 00:23

 Thanks, seems the problem lies in how excessive impulse values are handled (see "impulseStrength = 1000000" in the trace, that is _after_ clamping it in one my earlier commits and not even a sane value for nukes).

~0007176

Kloot (developer)

2011-08-01 12:20

 Units are now killed ASAP if they exit the map with insane speeds, so the UnitHandler assertions should never trigger anymore. As a result, position coordinates can never become so large as to cause SIGFPE's in float3::SqLength etc.
+Notes

-Issue History
Date Modified Username Field Change
2011-07-31 17:55 abma New Issue
2011-07-31 17:55 abma File Added: logs5.7z
2011-07-31 17:55 abma Note Added: 0007156
2011-07-31 17:57 abma Note Added: 0007157
2011-07-31 17:59 abma Target Version => 0.83.0
2011-07-31 18:01 abma Relationship added has duplicate 0002517
2011-07-31 18:14 abma Note Added: 0007159
2011-07-31 18:15 abma Note Edited: 0007159
2011-07-31 18:16 abma Note Edited: 0007159
2011-07-31 18:19 abma Severity minor => crash
2011-07-31 18:31 Kloot Note Added: 0007161
2011-07-31 20:07 abma File Added: 20110731_174600_unnamed_0.82+.4.sdf
2011-07-31 20:07 abma Note Added: 0007162
2011-07-31 20:31 abma Note Edited: 0007162
2011-07-31 20:44 Kloot Note Added: 0007163
2011-07-31 20:46 abma Note Added: 0007164
2011-07-31 20:47 abma Note Edited: 0007164
2011-07-31 21:05 abma File Added: 20110731_164348_unnamed_0.82+.4.sdf
2011-07-31 21:07 abma Note Added: 0007165
2011-07-31 21:07 abma Note Edited: 0007165
2011-07-31 21:08 abma Note Edited: 0007165
2011-07-31 21:09 abma Note Edited: 0007165
2011-07-31 21:14 abma Note Edited: 0007165
2011-07-31 21:29 abma Note Added: 0007166
2011-07-31 21:30 abma File Added: screenshot.jpg
2011-07-31 21:31 abma Note Edited: 0007166
2011-07-31 21:32 abma Note Edited: 0007166
2011-07-31 21:34 abma Note Edited: 0007166
2011-07-31 21:35 Kloot Note Added: 0007167
2011-07-31 21:36 abma Note Added: 0007168
2011-07-31 22:03 abma Note Added: 0007169
2011-07-31 22:09 abma Note Edited: 0007169
2011-07-31 22:09 abma Note Edited: 0007169
2011-07-31 22:29 abma Note Edited: 0007169
2011-07-31 22:32 abma Note Edited: 0007169
2011-07-31 22:33 abma Note Edited: 0007169
2011-07-31 22:36 abma Note Edited: 0007169
2011-07-31 22:45 abma Note Added: 0007170
2011-07-31 22:47 abma Note Edited: 0007170
2011-07-31 23:58 abma Note Added: 0007171
2011-08-01 00:00 abma Note Edited: 0007171
2011-08-01 00:21 Kloot Note Added: 0007172
2011-08-01 00:23 Kloot Note Edited: 0007172
2011-08-01 12:20 Kloot Note Added: 0007176
2011-08-01 12:20 Kloot Status new => resolved
2011-08-01 12:20 Kloot Fixed in Version => 0.82.7+git
2011-08-01 12:20 Kloot Resolution open => fixed
2011-08-01 12:20 Kloot Assigned To => Kloot
+Issue History