This version of krb5 landed in the buildroot and caused builds to start failing: https://koji.fedoraproject.org/koji/taskinfo?taskID=28626689 is one example. From coredumpctl on the builder: PID: 7141 (curl) UID: 1000 (kojibuilder) GID: 425 (mock) Signal: 11 (SEGV) Timestamp: Thu 2018-07-26 19:56:24 UTC (1h 8min ago) Command Line: curl -L -H Pragma: -o ./seesaw-0.10.0.tar.gz -R -S --fail https://src.fedoraproject.org/repo/pkgs/python-seesaw/seesaw-0.10.0.tar.gz/sha512/21c705bb99443008d677c9869d2144bcb03960b92eb9a55a48491ebcd70541042ffdc0779b0e21f8859e0bba990e86ff1ce643eb9eebb8a1f67053f3abf4fe86/seesaw-0.10.0.tar.gz Executable: /var/lib/mock/f29-build-13207084-962551/root/usr/bin/curl Control Group: /system.slice/kojid.service Unit: kojid.service Slice: system.slice Boot ID: 15441ca5f2b44562ab11054b5c94c5e1 Machine ID: 44a8c998799141cfbc8fcec28e881874 Hostname: buildvm-27.phx2.fedoraproject.org Storage: /var/lib/systemd/coredump/core.curl.1000.15441ca5f2b44562ab11054b5c94c5e1.7141.1532634984000000.lz4 Message: Process 7141 (curl) of user 1000 dumped core. Stack trace of thread 7141: #0 0x00000000f7f17c7b n/a (/usr/lib/ld-2.27.9000.so) I've untagged this build for now.
I'll investigate further as I can, but you should know: this version of krb5 is VERY unlikely to be the actual culprit - the only thing changed in this build is in the test suite.
you are of course right... it seems to be a binutils bug that just happened to hit this build. ;( Either 1609069 or 1609182 not sure which to dupe it on...
Since everything seemed to link okay, I think it more closely resembles 1609182. Thanks! *** This bug has been marked as a duplicate of bug 1609182 ***