Re: A reproducible Rosetta2 crash
Ron Garret <[email protected]> Sat, 17 Feb 2024 10:59:05 -0800
| Newsgroups | gmane.lisp.openmcl.devel |
|---|---|
| Message-ID | <[email protected]> |
Well, yeah, it's a bug, but a relatively minor one. It's something most users will never encounter, and if they do, it's trivial to fix or work around. This is really the least of our worries at the moment.
Just for the record, here's the fix:
Clozure Common Lisp Version 1.12.1 (v1.12.1-10-gca107b94) DarwinX8664
? (defmethod finalize-inheritance ((class ccl::std-class))
(unless (class-finalized-p class) (ccl::update-class class t)))
#<STANDARD-METHOD FINALIZE-INHERITANCE (CCL::STD-CLASS)>
? (finalize-inheritance (find-class 'standard-class))
NIL
?
> On Feb 17, 2024, at 8:48 AM, Bruce O'Neel <[email protected]> wrote:
>
> Hi,
>
> Thanks - this still seems like a bug though, right?
>
> As you said it also crashes on x86-64.
>
> Clozure Common Lisp Version 1.12.2 (v1.12.2-15-g79629c9d) LinuxX8664
>
> For more information about CCL, please see http://ccl.clozure.com <http://ccl.clozure.com/>.
>
> CCL is free software. It is distributed under the terms of the Apache
> Licence, Version 2.0.
> ? (finalize-inheritance (find-class 'standard-class))
> Unrecoverable stack overflow.
> ? for help
> [31448] Clozure CL kernel debugger:
>
> Clozure Common Lisp Version 1.12.2 (v1.12.2-15-g79629c9d) DarwinX8664
>
> For more information about CCL, please see http://ccl.clozure.com <http://ccl.clozure.com/>.
>
> CCL is free software. It is distributed under the terms of the Apache
> Licence, Version 2.0.
> ? (finalize-inheritance (find-class 'standard-class))
> Unrecoverable stack overflow.
> ? for help
> [92664] Clozure CL kernel debugger:
>
> Thanks.
>
> bruce
>
>
> On 2024-02-17T01:32:02.000+01:00, Ron Garret <[email protected]> wrote:
> Arrgh, please ignore this entire thread. It turns out that this crashes on x86 too, it's just that I had a different version of the code over on my x86 machine which didn't contain the call to finalize-inheritance.
>
> My apologies for the noise.
>
> rg