[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Gcl-devel] Re: still have "Bad plist" problem

From: Matt Kaufmann
Subject: [Gcl-devel] Re: still have "Bad plist" problem
Date: Tue, 19 May 2009 07:57:37 -0500

Hi, Camm --

By the way, I tried again (after rebuilding ACL2) and this time, I
didn't get an error.  The problem really is intermittent (and
reasonably rare).  Actually, I tried with SGC off as well as in the
normal way (SGC on).  In case you're interested, here are times (on an
8-core machine).

SGC off:
18483.339u 855.773s 54:49.55 587.8%     0+0k 0+0io 88pf+0w

SGC on (normal):
16096.517u 815.982s 48:03.46 586.5%     0+0k 0+0io 1pf+0w

-- Matt
   Cc: address@hidden, address@hidden
   From: Camm Maguire <address@hidden>
   Date: Mon, 18 May 2009 13:19:26 -0400
   X-SpamAssassin-Status: No, hits=-2.6 required=5.0
   X-UTCS-Spam-Status: No, hits=-252 required=165


   Matt Kaufmann <address@hidden> writes:

   > Hi, Camm --
   > By the way, I'm still seeing a problem with GCL 2.6.8pre that we've
   > discussed over the years:  A "Bad plist" error that seems to occur
   > randomly.  I don't have any suggestions -- we've kind of exhausted
   > what I can offer, and unfortunately, the error seems intermittent.
   > Anyhow, I thought  you might want to know it's still there.  At UT CS,
   > the version I used is /lusr/opt/gcl-2.6.8pre/bin/gcl, and my latest
   > error looks like this:
   >   Error: Bad plist ((< (EXPO X) K) (IF (< # A) (IF # # ...) ...) (QUOTE 
   > But "Bad plist" has taken many different forms over time.

   Does this mean you get a different error message on occasion? 

   Does the gc mode (i.e. sgc) affect this?

   Any other list corruption?  (I'm assuming you can examine and verify
   the bad plist after this error.)

   > I've run the ACL2 regression on Clozure CL (CCL) with safety 3 not
   > long ago (and ACL2 has changed little since then), so there's some
   > evidence at least that this isn't ACL2's fault.
   > My only suggestion is that you run a bunch of regression runs and see
   > if you can make this happen on your own system.  But maybe that's not
   > helpful.

   Is it possible to reactivate my account at ut so that I might try to
   reproduce this?

   Take care,

   > Regards,
   > Matt

   Camm Maguire                                     address@hidden
   "The earth is but one country, and mankind its citizens."  --  Baha'u'llah

reply via email to

[Prev in Thread] Current Thread [Next in Thread]