Re: boost::asio::deadline_timer segmentation in version 1.71

From: Richard Hodges (hodges.r_at_[hidden])
Date: 2020-10-14 07:43:23

There's not a lot of information to go on here. It looks as if you might
have a segfault in the scheduler rather than the timer itself. Is the
io_context still in scope?

Happy to chat in real time on slack if you have access to it. I generally
hang out in the #beast channel as I maintain this library.

https://cppalliance.org/slack/

On Wed, 14 Oct 2020 at 02:30, Jupiter via Boost <boost_at_[hidden]>
wrote:

> Hi,
>
> My application is running one thread under boost::asio::io_service, it
> is an embedded system running on an ARM processor with Linux, so no
> gdb installed. The application is also using a
> boost::asio::deadline_timer to call a function every second, but I got
> following segmentation stack trace:
>
>
> _ZN5boost4asio6detail9scheduler10do_run_oneERNS1_27conditionally_enabled_mutex11scoped_lockERNS1_21scheduler_thread_infoERKNS_6system10error_codeE+0xdd
>
> _ZN5boost4asio6detail9scheduler3runERNS_6system10error_codeE+0x83
>
> I thought I could run c++filt to demangling the library name, not sure
> why it didn't demangle the name today.
>
> I used the same boost::asio::deadline_timer version 1.69 for another
> application running on Ubuntu which was working fine, what I could be
> missing?
>
> Thank you.
>
> Kind regards,
>
> Jupiter
>
> _______________________________________________
> Unsubscribe & other changes:
> http://lists.boost.org/mailman/listinfo.cgi/boost
>

-- 
Richard Hodges
hodges.r_at_[hidden]
office: [phone]
home: [phone]
mobile: [phone]