there have been some failures in the last two runs of the slow sessions:
(Relevant changesets are listed there.)
Unfortunately there have been no emails because for some reason, they
ran into the "hard" timeout as imposed by Jenkins. I will investigate
why they haven't been aborted the proper way.
What seems interesting are those two consecutive lines of log:
11:07:25.458 Iptables_Semantics_Examples: theory
19:01:48.518 Run out of store - interrupting threads
(timestamps are elapsed time)
So, nothing happens for a while and then we get an out of memory error.
Note that the "slow" sessions are executed on 64 bit with an abundance
isabelle-dev mailing list