Re: Release 3.8.1 on ubuntu 18.04 qa_qtgui (Failed) with Segmentation fault

[Marcus Müller]

Hi Joe,

I'm not ruling out this is a GNU Radio bug, but I find it rather
unlikely; your complete call trace doesn't ever wander into any GNU
Radio libraries, but:

> ==22123==    by 0x13B1D604: QPrinterInfoPrivate (qprinterinfo_p.h:71)

Sounds like you've either found a bug in Qt (unlikely), or you're mixing
multiple versions of QT somehow on your machine. Is that possible?

Best regards,
Marcus
On 26.04.20 14:13, Joe D wrote:
> 
> Thanks Vasil for correcting my approach
> 
> below the relevant excerpt from the new valgrind log
> 
> Attached the full one
> 
> Regarding gdb , is there any additional steps to load the symbols or the
> instruction on https://wiki.gnuradio.org/index.php/TutorialsDebugging
> will suffice
> 
> ==22123== Process terminating with default action of signal 11
> (SIGSEGV): dumping core
> ==22123==  Bad permissions for mapped region at address 0x104C66C0
> ==22123==    at 0x13B1D604: ref (qlist.h:121)
> ==22123==    by 0x13B1D604: QList (qlist.h:121)
> ==22123==    by 0x13B1D604: QPrinterInfoPrivate (qprinterinfo_p.h:71)
> ==22123==    by 0x13B1D604: __static_initialization_and_destruction_0
> (qprinterinfo.cpp:35)
> ==22123==    by 0x13B1D604: _GLOBAL__sub_I_qprinterinfo.cpp
> (qprinterinfo.cpp:163)
> ==22123==    by 0x4010732: call_init (dl-init.c:72)
> ==22123==    by 0x4010732: _dl_init (dl-init.c:119)
> ==22123==    by 0x40151FE: dl_open_worker (dl-open.c:522)
> ==22123==    by 0x4FA32DE: _dl_catch_exception (dl-error-skeleton.c:196)
> ==22123==    by 0x40147C9: _dl_open (dl-open.c:605)
> ==22123==    by 0x544CF95: dlopen_doit (dlopen.c:66)
> ==22123==    by 0x4FA32DE: _dl_catch_exception (dl-error-skeleton.c:196)
> ==22123==    by 0x4FA336E: _dl_catch_error (dl-error-skeleton.c:215)
> ==22123==    by 0x544D734: _dlerror_run (dlerror.c:162)
> ==22123==    by 0x544D050: dlopen@@GLIBC_2.2.5 (dlopen.c:87)
> ==22123==    by 0x5F7DBC: _PyImport_FindSharedFuncptr (in
> /usr/bin/python3.6)
> ==22123==    by 0x5FB027: _PyImport_LoadDynamicModuleWithSpec (in
> /usr/bin/python3.6)
> ==22123==
> ==22123== HEAP SUMMARY:
> ==22123==     in use at exit: 5,615,552 bytes in 7,555 blocks
> ==22123==   total heap usage: 25,234 allocs, 17,679 frees, 48,677,061
> bytes allocated
> ==22123==
> ==22123== LEAK SUMMARY:
> ==22123==    definitely lost: 157,672 bytes in 86 blocks
> ==22123==    indirectly lost: 0 bytes in 0 blocks
> ==22123==      possibly lost: 230,559 bytes in 220 blocks
> ==22123==    still reachable: 5,227,321 bytes in 7,249 blocks
> ==22123==                       of which reachable via heuristic:
> ==22123==                         newarray           : 1,536 bytes in 16
> blocks
> ==22123==         suppressed: 0 bytes in 0 blocks
> ==22123== Rerun with --leak-check=full to see details of leaked memory
> ==22123==
> ==22123== For counts of detected and suppressed errors, rerun with: -v
> ==22123== Use --track-origins=yes to see where uninitialised values come
> from
> ==22123== ERROR SUMMARY: 11565 errors from 209 contexts (suppressed: 0
> from 0)
> Segmentation fault
> 
> On Sun, Apr 26, 2020 at 9:27 PM Vasil Velichkov <address@hidden
> <mailto:address@hidden>> wrote:
> 
>     Hi Joe,
> 
>     On 26/04/2020 14.01, Joe D wrote:
>     > Attached a valgrind log
> 
>     >>
>     root@ubuntu-srv-10:/Applications/gnuradio/build/gr-qtgui/python/qtgui#
>     valgrind -v ctest -V -R qa_qtgui
> 
>     You can't run this test under valgrind like this. You need to open
>     ./build/gr-qtgui/python/qtgui/qa_qtgui_test.sh and on the last line
>     add valgrind in front of python3. Something like:
> 
>         valgrind /usr/bin/python3 -B
>     /Applications/gnuradio/gr-qtgui/python/qtgui/qa_qtgui.py
> 
>     Similarly you can run it under gdb in order to get a full backtrace
>     of this segfault.
> 
>     Regards,
>     Vasil
>