View Issue Details
| ID | Project | Category | View Status | Date Submitted | Last Update |
|---|---|---|---|---|---|
| 0003911 | Spring engine | General | public | 2013-08-02 10:44 | 2013-08-05 00:45 |
| Reporter | cleanrock | Assigned To | jK | ||
| Priority | normal | Severity | crash | Reproducibility | sometimes |
| Status | resolved | Resolution | fixed | ||
| Product Version | 94.1.1+git | ||||
| Summary | 0003911: boost::optional Assertion `this->is_initialized()' failed | ||||
| Description | Pretty rare problem, I have seen this twice now: spring: /usr/include/boost/optional/optional.hpp:631: boost::optional<T>::reference_type boost::optional<T>::get() [with T = long unsigned int; boost::optional<T>::reference_type = long unsigned int&]: Assertion `this->is_initialized()' failed. Full backtrace attached. The attached infolog contain a different stack trace which was generated after i did continue in gdb to get the infolog file. | ||||
| Tags | No tags attached. | ||||
| Attached Files | |||||
| Checked infolog.txt for Errors | |||||
|
|
this is how test_ThreadPool fails for me: test/test_ThreadPool Running 1 test case... test/engine/System/testThreadPool.cpp(20): info: passed test/engine/System/testThreadPool.cpp(20): info: passed test/engine/System/testThreadPool.cpp(20): info: passed test/engine/System/testThreadPool.cpp(20): info: passed test/engine/System/testThreadPool.cpp(20): info: passed test/engine/System/testThreadPool.cpp(29): error in "ThreadPoolTest": check nums[i] == i failed test/engine/System/testThreadPool.cpp(29): error in "ThreadPoolTest": check nums[i] == i failed test/engine/System/testThreadPool.cpp(29): error in "ThreadPoolTest": check nums[i] == i failed test/engine/System/testThreadPool.cpp(29): error in "ThreadPoolTest": check nums[i] == i failed test/engine/System/testThreadPool.cpp(29): error in "ThreadPoolTest": check nums[i] == i failed test/engine/System/testThreadPool.cpp(29): error in "ThreadPoolTest": check nums[i] == i failed test/engine/System/testThreadPool.cpp(29): error in "ThreadPoolTest": check nums[i] == i failed test/engine/System/testThreadPool.cpp(29): error in "ThreadPoolTest": check nums[i] == i failed test/engine/System/testThreadPool.cpp(29): error in "ThreadPoolTest": check nums[i] == i failed test/engine/System/testThreadPool.cpp(29): error in "ThreadPoolTest": check nums[i] == i failed test/engine/System/testThreadPool.cpp(29): error in "ThreadPoolTest": check nums[i] == i failed test/engine/System/testThreadPool.cpp(29): error in "ThreadPoolTest": check nums[i] == i failed test/engine/System/testThreadPool.cpp(29): error in "ThreadPoolTest": check nums[i] == i failed test/engine/System/testThreadPool.cpp(29): error in "ThreadPoolTest": check nums[i] == i failed test/engine/System/testThreadPool.cpp(29): error in "ThreadPoolTest": check nums[i] == i failed test/engine/System/testThreadPool.cpp(29): error in "ThreadPoolTest": check nums[i] == i failed test/engine/System/testThreadPool.cpp(29): error in "ThreadPoolTest": check nums[i] == i failed Thread 0 executed 17 times Thread 1 executed 131 times Thread 2 executed 26 times Thread 3 executed 301 times Thread 4 executed 223 times Thread 5 executed 37 times Thread 6 executed 100 times Thread 7 executed 25 times Thread 8 executed 118 times Thread 9 executed 5 times *** 17 failures detected in test suite "ThreadPool" test_ThreadPool: /usr/include/boost/thread/pthread/condition_variable_fwd.hpp:81: boost::condition_variable::~condition_variable(): Assertion `!ret' failed. Abgebrochen (Speicherabzug geschrieben) it looks like its some environment specific. maybe boost / gcc is broken. on the gentoo buildslave the test seems to always succeed, on the validation slave it fails (Fedora Linux), on my machine it fails, too (Ubuntu), on cleanrocks it seems to fail to (arch linux). why does it work on gentoo? BOOST_THREAD_PROVIDES_INTERRUPTIONS maybe?! edit: no, BOOST_THREAD_DONT_PROVIDE_INTERRUPTIONS doesn't change the failed assertion, just line number changes. |
|
|
fixed "boost::condition_variable::~condition_variable(): Assertion `!ret' failed." never saw cleanrock's one |
|
|
I seen this 2 times now again in spring 94.1.1-810-g1498ea1 develop (Debug): [f=0000000] Loaded gadget: Start Point Remover Gadget <init_start_point_remover_gadget.lua> spring: /usr/include/boost/optional/optional.hpp:631: boost::optional<T>::reference_type boost::optional<T>::get() [with T = long unsigned int; boost::optional<T>::reference_type = long unsigned int&]: Assertion `this->is_initialized()' failed. Happens about one in ten for me. I start a BA 7.78 vs KAIK on Altair_Crossing-V1. |
|
|
Saw this again now: spring: /usr/include/boost/optional/optional.hpp:631: boost::optional<T>::reference_type boost::optional<T>::get() [with T = long unsigned int; boost::optional<T>::reference_type = long unsigned int&]: Assertion `this->is_initialized()' failed. (gdb) bt #0 0x00007ffff39ba1c9 in raise () from /usr/lib/libc.so.6 #1 0x00007ffff39bb5c8 in abort () from /usr/lib/libc.so.6 #2 0x00007ffff39b3356 in __assert_fail_base () from /usr/lib/libc.so.6 0000003 0x00007ffff39b3402 in __assert_fail () from /usr/lib/libc.so.6 0000004 0x0000000000c91bfd in boost::optional<unsigned long>::get (this=0x1b62220 <Threading::simThreadID>) at /usr/include/boost/optional/optional.hpp:631 0000005 0x0000000000c916be in boost::optional<unsigned long>::operator* (this=0x1b62220 <Threading::simThreadID>) at /usr/include/boost/optional/optional.hpp:647 #6 0x0000000000c8e5a5 in Threading::IsSimThread () at /home/johanr/my_projects/spring/rts/System/Platform/Threading.cpp:326 #7 0x000000000073cc68 in CLuaHandle::CallOutUnsyncedUpdateCallIn (L=0x7fffa1f76550) at /home/johanr/my_projects/spring/rts/Lua/LuaHandle.cpp:3043 (gdb) f 6 #6 0x0000000000c8e5a5 in Threading::IsSimThread () at /home/johanr/my_projects/spring/rts/System/Platform/Threading.cpp:326 326 return !simThreadID ? false : NativeThreadIdsEqual(Threading::GetCurrentThreadId(), *simThreadID); (gdb) f 5 0000005 0x0000000000c916be in boost::optional<unsigned long>::operator* (this=0x1b62220 <Threading::simThreadID>) at /usr/include/boost/optional/optional.hpp:647 647 reference_type operator *() { return this->get() ; } (gdb) f 4 0000004 0x0000000000c91bfd in boost::optional<unsigned long>::get (this=0x1b62220 <Threading::simThreadID>) at /usr/include/boost/optional/optional.hpp:631 631 reference_type get() { BOOST_ASSERT(this->is_initialized()) ; return this->get_impl(); } Looks like the !operator doesn't bite. I will take a look at it since it looks like only i can reproduce it and its not very critical. |
|
|
https://github.com/spring/spring/commit/296bcb38d3323a1eb2319b71b10b29ea8e625190 seems to have fixed it. @jk: any idea why it worked before? the code seems to have not changed since the switch to threadpools. (imo this report can be closed) |
| Date Modified | Username | Field | Change |
|---|---|---|---|
| 2013-08-02 10:44 | cleanrock | New Issue | |
| 2013-08-02 10:44 | cleanrock | File Added: boost_optional_assert_backtrace.txt | |
| 2013-08-02 10:45 | cleanrock | File Added: infolog.txt | |
| 2013-08-02 14:58 | abma | Note Added: 0011142 | |
| 2013-08-02 15:02 | abma | Note Edited: 0011142 | |
| 2013-08-02 15:05 | abma | Note Edited: 0011142 | |
| 2013-08-02 17:29 | abma | Relationship added | related to 0003907 |
| 2013-08-03 00:29 | jK | Note Added: 0011145 | |
| 2013-08-03 05:09 | cleanrock | Note Added: 0011146 | |
| 2013-08-03 06:50 | jK | Changeset attached | => spring develop 37a02809 |
| 2013-08-03 06:50 | jK | Assigned To | => jK |
| 2013-08-03 06:50 | jK | Status | new => resolved |
| 2013-08-03 06:50 | jK | Resolution | open => fixed |
| 2013-08-03 13:09 | cleanrock | Assigned To | jK => cleanrock |
| 2013-08-03 13:09 | cleanrock | Note Added: 0011148 | |
| 2013-08-03 13:09 | cleanrock | Status | resolved => feedback |
| 2013-08-03 13:09 | cleanrock | Resolution | fixed => reopened |
| 2013-08-03 13:10 | cleanrock | Status | feedback => assigned |
| 2013-08-03 15:55 | abma | Note Added: 0011150 | |
| 2013-08-03 15:55 | abma | Assigned To | cleanrock => jK |
| 2013-08-03 15:55 | abma | Status | assigned => feedback |
| 2013-08-05 00:45 | abma | Status | feedback => resolved |
| 2013-08-05 00:45 | abma | Resolution | reopened => fixed |
