On Thu, 27 Nov 2014 12:05:51 +0100, Mark Wielaard wrote:
> We have -o,--output support in other utilities.

Those are ld, strip and unstrip.  That is not for text output but for the
resulting ELF files which is IMO a different case.

BTW I am not against it, I just find it redundant.  If thiAny program could
have -o|--output but in UNIX it is solved by the more general shell '>'.


> But maybe it isn't
> useful in general and people will always use shell redirection?

Currently it uses fopen "wx" which requires unique filename for each crash
dump unwind.  One could use %p in the filename but that is also not completely
unique.  Besides that one should IMO provide more information from /proc/PID/
(like fd/ ) than just the backtrace eu-stack can provide.  One is also
interested in /proc/PID/exe which is not obvious from -l or -m (at least for
PIEs); although one could possibly use %E for core_pattern.

I just find it improbable one would not need a single other kind of
information than what eu-stack can provide.


> But if --core-pattern isn't useful without other code, then we should
> include that code too.

Yes, there could be provided some "contrib"-like core dumping shell script.


> Could you post what you have?

Nothing interesting, only for debugging eu-stack --core-pattern, when you ask:
	#! /bin/bash
	exec >>/tmp/out 2>>/tmp/out
	date --iso=seconds
	echo "$*"
	cd /home/jkratoch/redhat/elfutils
	. /home/jkratoch/t/elfutils-boot
	ldd ./src/stack
	# It locks up as strace holds fd 0:
	#strace -s200 -o /tmp/out.strace -q ./src/stack --core-pattern -1 --pid=$1
	exec ./src/stack --core-pattern -1 -l -m --pid=$1


Jan