bug#48252: Crashes continue ...

[andrei . elkin]

Eli Zaretskii <eliz@gnu.org> writes:

>> Date: Wed, 21 Dec 2022 15:18:39 +0200
>> From: andrei.elkin--- via "Bug reports for GNU Emacs,
>>  the Swiss army knife of text editors" <bug-gnu-emacs@gnu.org>
>> 
>> Hello again!
>> 
>> I have a question below.
>> 
>> Meanwhile ...
>> the stack persistently comes back for me, and yesterday in the
>> up-to-date emacs-28 branch (version 28.2.50).
>
> And in Emacs 29 what happens?

I'll relocate to it after any next crash.

>
>> It takes some days, maybe weeks until I start feeling some slowness
>> in responses to commands.
>> I must have learned that the stack appears in correlation with
>> C-k or just delete-char commands.
>> 
>> I previously thought it might have something to do with `lsp`. Can't yet
>> dismiss that feeling.
>> 
>> What if I skip (comment out)
>>   #1 print_object
>> if I know somehow that  SYMBOL_NAME() will segfault?
>> If that would deter emacs from crashing even for few more weeks :-), I'd
>> be happier than I am now.
>
> I'm sorry, but the only way forward is either to have a reproducible
> recipe of sorts.  We must understand how this invalid object comes to
> life.
>
> The segfault happens when Emacs tries to dereference an invalid
> pointer.  If someone knows how to detect that, please speak up.

My hope obviously has been that an invalid object gets exposed so far
in my case only via that stack, and might be harmless otherwise.