Lem hangs when redefining struct as class

[anlsh]

Steps to reproduce

Put this some file

;; foo.lisp
(defstruct foo)
(defclass foo () ())

Then open it with latest main (I tested with 35e010f4de35d7713f632dadbf82be31176bcf32), and M-x lisp-compile-and-load-file, or evaluate both forms independently, or whatever. Lem will hang.

---

To be clear, sbcl doesn't think that this is a valid sequence of forms, and breaks accordingly

$ sbcl --load /tmp/foo.lisp 
This is SBCL 2.1.11.debian, an implementation of ANSI Common Lisp.
More information about SBCL is available at <http://www.sbcl.org/>.

SBCL is free software, provided as is, with absolutely no warranty.
It is mostly in the public domain; some portions are provided under
BSD-style licenses.  See the CREDITS and COPYING files in the
distribution for more information.
While evaluating the form starting at line 3, column 0
  of #P"/tmp/foo.lisp":

debugger invoked on a SB-PCL::METAOBJECT-INITIALIZATION-VIOLATION in thread
#<THREAD "main thread" RUNNING {1001834103}>:
  Cannot CHANGE-CLASS objects into CLASS metaobjects.
See also:
  AMOP, Initialization of Class Metaobjects

Type HELP for debugger help, or (SB-EXT:EXIT) to exit from SBCL.

restarts (invokable by number or by possibly-abbreviated name):
  0: [RETRY   ] Retry EVAL of current toplevel form.
  1: [CONTINUE] Ignore error and continue loading file "/tmp/foo.lisp".
  2: [ABORT   ] Abort loading file "/tmp/foo.lisp".
  3:            Ignore runtime option --load "/tmp/foo.lisp".
  4:            Skip rest of --eval and --load options.
  5:            Skip to toplevel READ/EVAL/PRINT loop.
  6: [EXIT    ] Exit SBCL (calling #'EXIT, killing the process).

(SB-PCL::CHANGE-CLASS-TO-METAOBJECT-VIOLATION CLASS NIL ((:AMOP :INITIALIZATION CLASS)))
0]

But it offers appropriate restarts so it's obiously recoverable. Hanging lem is bad: how can we fix this? I'm willing to go dig around given some initial pointers

[johnlorentzson]

It appears sending a SIGTERM (via C-c in the terminal) after it freezes causes it to display a memory fault error and it took a while after that to reach ldb. Strange.

[cxxxr]

Although I haven't investigated anything yet, I suspect that the cause of the hang-up is the communication part with micros.

[anlsh]

@cxxxr Where do I start digging on this? If you can give me some pointers on the basics (streaming micros logs to a console, say) then I can give debugging a shot

[cxxxr]

@anlsh I often print debug to a specific file (e.g. /tmp/foo.log).

(defun dbg (fmt &rest args)
  (with-open-file (stream "/tmp/foo.log" :direction :output :if-exists :append)
    (apply #'format stream fmt args)
    (terpri stream)))

I also think the problem lies in the part where it is compiled and loaded, so I will start the print debug from there.

https://github.com/lem-project/micros/blob/9fc7f1e5b0dbf1b9218a3f0aca7ed46e90aa86fd/micros.lisp#L2537