I for one really support this idea.
Having immediate access to all the debuginfo for all the installed software really useful. The practical benefit is not entirely obvious. For example a week or so ago I needed to track down a problem with libboost and where it was storing a variable. So I ended up doing:
$ ~/bin/objdump --dwarf=loc,follow-links /lib64/libboost_system.so.1.69.0
Also being able to use userspace systemtap quickly and easily really is helpful
For example listing the functions in a library and what parameters they take is helpful.
$ stap -L 'process("/lib64/libm.so.6").function("*").call'
It is really useful when you are looking at some unusual library that you aren't familiar with. All of these functions can become systemtap tracepoints that you tap into and monitor.
Having online access really makes it much more convenient. Yes you can debuginfo install things as needed and most of us have plenty of disk space and we don't mind installing the debuginfo in an ad hoc matter. But when you are working on minimal software sets like containers, you don't want to have to be installing all sorts of debuginfo in your containers while you are tracking down problems.
I think that overall providing online access to debuginfo through a debuginfod service is one of those things which may seem like redundant to something more manual but it is one of those things which is of creeping utility as you find that the convenience of it leads to you using it more (mostly indirectly) than you ever thought that you would just because it is always there.
-ben
_______________________________________________
devel mailing list -- devel@xxxxxxxxxxxxxxxxxxxxxxx
To unsubscribe send an email to devel-leave@xxxxxxxxxxxxxxxxxxxxxxx
Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: https://lists.fedoraproject.org/archives/list/devel@xxxxxxxxxxxxxxxxxxxxxxx
