[isabelle-dev] Fwd: status (AFP)

[Gerwin Klein]

Is anyone else observing intermittent problems like this?

Building Jinja ...
poly: scanaddrs.cpp:107: PolyWord ScanAddress::ScanStackAddress(PolyWord, StackObject*, bool): Assertion `val.IsDataPtr()' failed.
/home/kleing/volatile/isadist/Isabelle_13-Oct-2011/lib/scripts/run-polyml: line 77: 20095 Aborted                 "$POLY" -q $ML_OPTIONS
Jinja FAILED

We're investigating if possibly something is wrong with the test server's memory, but it seems unlikely (our L4.verified sessions are larger and stable).

The problem does not seem to be Jinja itself, because Jinja does get built fine in one of the other sessions that depend on it later.

This is poly 5.4.0 64 bit on Linux, I might try updating it to 5.4.1 or the current svn. The same setup has been running without such problems for about a year, though. The only thing that I can think of that has changed in the last few weeks might be server load which has gone up, but is nowhere near saturated. And of course Isabelle is always changing, so the allocation of something might have changed as well.

Cheers,
Gerwin

Begin forwarded message:

> From: Gerwin Klein <kleing at ertos.nicta.com.au>
> Date: 14 October 2011 8:25:51 AM AEDT
> To: <kleing at cse.unsw.edu.au>
> Subject: status (AFP)
> 
> The status of the following AFP entries changed or remains FAIL: 
> [Jinja] is still on FAIL.
> 
> Full entry status at http://afp.sourceforge.net/status.shtml
> 
> AFP version: development -- hg id ceb9698de567
> Isabelle version: devel -- hg id 2214ba5bdfff
> Test ended on: lemma, Fri Oct 14 08:25:51 EST 2011.
> 
> Have a nice day,
>  isatest