On 24.08.2017 21:42, Jan Kratochvil wrote:
On Thu, 24 Aug 2017 21:36:29 +0200, Sandro Mani wrote:
> While I'm at it, my current technique for interpreting mingw stacktraces
> produced without debuginfos is parsing the text and calling addr2line for
> each stack frame. Is there a neater technique?
Unaware of. But then why you do not have the debuginfos?
You deploy the
application without debuginfos to save space, and then
have a crash-handler which attaches gdb to the dying process, produces a
stacktrace, and sends it to the developer if the user wants.
And also why ABRT
does not process it? ABRT has its retrace server infrastructure to do exactly
that but I guess ABRT cannot handle PE32.
May well be mistaken, but doesn't
ABRT work with coredumps, which you
can't get on Windows systems?
> when one has to do live-debugging on without debuginfos.
Why? debuginfos are easily available from standard repos.
One example: you are at
a customers to debug some misbehaviour on a
crippled down Windows system where you can't place the *.debug files
next to the binaries, and perhaps don't event have internet access...