I've noticed some odd test failures from the buildbots recently. e.g. last
night I saw the following, all on the 6.10 branch:
on x86_64-Linux:
conc023(threaded2)
signals004(threaded2)
on x86-Linux:
conc023(threaded2)
on x86-Windows:
list001(ghci)
There are several other failures that are less suspicious, but the above
ones have me stumped. I've been trying in vain to reproduce them today, on
the same machines, in the same builds. I've been running conc023
repeatedly on both the Linux machines all morning, without a single
failure. I've tried varying various parameters, using more cores etc., but
no luck.
The list001 failure on Windows initially seemed to be reproducible, but on
closer inspection it went away. Now I can't reproduce it at all.
If anyone sees a crash with +RTS -N2 or higher, please let me know!
Cheers,
Simon
_______________________________________________
Cvs-ghc mailing list
[email protected]
http://www.haskell.org/mailman/listinfo/cvs-ghc
