Home / Browse / Steel Bank Common Lisp / Mail / Archive

Steel Bank Common Lisp

Email Archive: sbcl-help (read-only)

[Sbcl-help] SB-INT:BUG: failed AVER: "(NULL CURRENT)"

From: Robert Uhl <eadmund42@gm...> - 2006-08-16 05:13

I'm getting a weird behaviour--when executing a macro while compiling a
file using ASDF, SBCL generates the following output:

  debugger invoked on a SB-INT:BUG in thread #<THREAD "initial thread"
  {A68B5F9}>:    failed AVER: "(NULL CURRENT)"

    This is probably a bug in SBCL itself. (Alternatively, SBCL might
    have been corrupted by bad user code, e.g. by an undefined Lisp
    operation like (FMAKUNBOUND 'COMPILE), or by stray pointers from
    alien code or from unsafe Lisp code; or there might be a bug in the
    OS or hardware that SBCL is running on.) If it seems to be a bug in
    SBCL itself, the maintainers would like to know about it. Bug
    reports are welcome on the SBCL mailing lists, which you can find at
    <http://sbcl.sourceforge.net/>.

Oddly, when compiling by hand the bug doesn't appear to rear its ugly
head--and of course loading from the FASL is no problem.

I've instrumented my macro (both the macro itself and the code
generated), and the crash happens at the very end, when the generated
code is returned.  As I noted, this only happens when compiling through
ASDF--it compiles cleanly with COMPILE-FILE, so it's possible it's a
strange ASDF bug.

My macros returns a bunch of PROGNs containing various definitions,
including a bunch of DEF-CLP-FUNCTIONs, a DEFMETHOD and a
DEF-VIEW-CLASS, as well as SETFing a couple of hash table (which I made
sure are created at compile time using EVAL-WHEN).

This is with SBCL 0.9.14.

-- 
Robert Uhl <http://public.xdi.org/=ruhl>
Every cloud has a silver lining but every year lightning kills hundreds
of people who are looking for it.                       --Oliver Barker

Thread View

Thread	Author	Date
[Sbcl-help] SB-INT:BUG: failed AVER: "(NULL CURRENT)"	Robert Uhl <eadmund42@gm...>

Re: [Sbcl-help] SB-INT:BUG: failed AVER: "(NULL CURRENT)"

From: Robert Uhl <eadmund42@gm...> - 2006-08-18 05:22

I figured out the root cause of the problem: referring to a global
(special) variable in one file before the file it's defined in was
loaded; ASDF loaded the one file before the other, whereas loading by
hand was generally in the proper order.  Once I changed the file
dependencies in my DEFSYSTEM it appears that the files compile without
the error.

Is this the proper behaviour in such an instance?

-- 
Robert Uhl <http://public.xdi.org/=ruhl>
That's how you know you're hooked on something; when it makes you forget
to drink beer.                     --Paul Mather, commenting on The Sims

Re: [Sbcl-help] SB-INT:BUG: failed AVER: "(NULL CURRENT)"

From: Juho Snellman <jsnell@ik...> - 2006-08-18 05:38

Robert Uhl <eadmund42@...> writes:

> I figured out the root cause of the problem: referring to a global
> (special) variable in one file before the file it's defined in was
> loaded; ASDF loaded the one file before the other, whereas loading by
> hand was generally in the proper order.  Once I changed the file
> dependencies in my DEFSYSTEM it appears that the files compile without
> the error.
> 
> Is this the proper behaviour in such an instance?

Probably not, can you send a test case of this to the list?

-- 
Juho Snellman

Re: [Sbcl-help] SB-INT:BUG: failed AVER: "(NULL CURRENT)"

From: Robert Uhl <eadmund42@gm...> - 2006-08-19 03:52

Juho Snellman <jsnell@...> writes:
>
> Probably not, can you send a test case of this to the list?

I've been unable to narrow it down to a model test case--just a
repeatable version with my source code.  I've put a tarball up at
<http://latakia.dyndns.org/~ruhl/lisp/crash.tar.gz>.  Simply untar and
run:

   sbcl --load run.lisp

I believe that the code is safe to run and will not do any harm--but it
does demonstrate a crash after all, and the code could do something
nasty.  So I'd recommend running it as a user with no rights, and
examining the code first.

To run it, you'll need portableaserve and CLSQL.

-- 
Robert Uhl <http://public.xdi.org/=ruhl>
They say `the early bird gets the worm.'  What they often fail to
consider, however, is the early cat.