[packman] Bug in HandBrakeGUI on Tumbleweed openSUSE

Stefan Seyfried stefan.seyfried at googlemail.com
Sun Mar 22 09:26:24 CET 2020


Hi Dave,

Am 22.03.20 um 07:08 schrieb Dave Plater:
> On 3/22/20, Freek de Kruijf <freek at opensuse.org> wrote:
>> Op zaterdag 21 maart 2020 13:34:12 CET schreef Dave Plater:
>>> Unfortunately after updating to the same snapshot handbrake still
>>> works. Force install handbrake-gtk handbrake and the debug packages,
>>> run gdb HandBrakeGUI again and post the result of bt full after the
>>> crash.
>>>
>>> Regards
>>> Dave
>>
>> After the latest upgrade, later today, the versions for i586 and x86_64 are
>> the same. So
>>
>> Output of bt full:
>>
>> #0  0x00007ffff3ece6d1 in std::vector<std::__cxx11::basic_string<char,
>> std::char_traits<char>, std::allocator<char> >,
>> std::allocator<std::__cxx11::basic_string<char, std::char_traits<char>,
>> std::allocator<char> > > >::_M_realloc_insert<char const (&)
>> [16]>(__gnu_cxx::__normal_iterator<std::__cxx11::basic_string<char,
>> std::char_traits<char>, std::allocator<char> >*,
>> std::vector<std::__cxx11::basic_string<char, std::char_traits<char>,
>> std::allocator<char> >, std::allocator<std::__cxx11::basic_string<char,
>> std::char_traits<char>, std::allocator<char> > > > >, char const (&) [16])
>>     (this=0x7fffffffd4e0, __position=__position at entry=non-dereferenceable
>> iterator for std::vector)
>>     at /usr/include/c++/9/ext/new_allocator.h:119
>>         __len = <optimized out>
>>         __old_start = <optimized out>
>>         __old_finish = <optimized out>
>>         __elems_before = <optimized out>
>>         __new_start = 0x5555563f8c60
>>         __new_finish = 0x5555563f8c80


> this was fixed. The problem appears to be coming from libmfx1 which is
> an intel hardware acceleration library. This means that your problem

I don't think so. It looks like the illegal instruction is happening in
/usr/include/c++/9/ext/new_allocator.h:119 (I'm not really familiar with
c++ internals and might be wrong, though).

Anyway, this "illegal instruction" might point to a miscompiled, or
wrongly cpu-optimized stanard library.

I'd file a bug with this information in the openSUSE bugzilla, so that
the toolchain team can have a look.
-- 
Stefan Seyfried

"For a successful technology, reality must take precedence over
 public relations, for nature cannot be fooled." -- Richard Feynman



More information about the Packman mailing list