richton@nbcs.rutgers.edu wrote:

I saw this three times in testing of HEAD. One backtrace and -d -1:

https://www.nbcs.rutgers.edu/~richton/test039-bt-200906200640.txt https://www.nbcs.rutgers.edu/~richton/test039-slapdlog-200906200640.txt

These are under libumem with transaction logging enabled. All malloc/free for a given block can be retrieved, in theory...

Are people still running into this? Seems like not. Time to close this issue?