Re: Patch to remove a bit of duplicated code in eval.c

[Federico Tedin <federicotedin@gmail.com>]

>> My concerns now are:
>> 1) Could I have broken anything without realizing it, since this is such
>> a central function in Lisp code evaluation? Everything seems to be
>> compiling fine (without warnings) and so far I haven't had any crashes.
>
> I tried to compare the old and the new code, and their differences are
> not trivial to audit.  For example, you remove this part:
>
>> -      else if (XSUBR (fun)->max_args == UNEVALLED)
>> -	val = (XSUBR (fun)->function.aUNEVALLED) (args_left);
>
> but I don't immediately see how you have something equivalent and at
> the right opportunity in the new code.

This part in particular is handled by the else-case in
'apply_subr'. Functionally it should be the same because I'm just taking
the value of 'args_left' and passing it to 'funcall_subr' (through the
'arg_vector' array of Lisp_Objects, with fixed size 1). Then
'funcall_subr' just calls 'aUNEVALLED' with that single Lisp_Object,
which corresponds to the section you pointed out.

> I'm not saying I already see something you have broken, I'm saying
> these changes are not easy to audit for correctness, at least not for
> me.  They are not a mechanical move of code from one place to another,
> they really change the flow of control and processing.
>
> So I think we'd like to have tests for each of the cases supported by
> the code to make sure nothing is broken.  Is it feasible to write such
> tests?

On this I am not very sure. There are some tests for this file on
eval-tests.el, however they don't seem to cover subroutines
stuff. However it seems like some stuff could be tested; for example, I
have realized that my patch evaluates all arguments before doing an
arguments length check, which is already different to the current
implementation (but would be easy to fix also). I will look into adding
relevant tests to other parts I feel I might have affected as well.

>> 2) I removed a comment that made reference to Bug#21245, but it seems
>> like it makes sense since the variable it refers to is no longer needed.
>
> If the reason for the comment is gone, no need to keep the comment.
>
>> 3) Have I maybe made Emacs slower by always using SAFE_ALLOCA_LISP for
>> the subroutine arguments (instead of only for 'max_args=MANY')?
>
> This should be measured.

Sounds good. I also thought about it again and realized that if
max_args!=MANY, then I can use simply use a Lisp_Object argvals[8]...
like in the code I removed. So in that case the behavior should be
exactly the same as before, and the change was a mistake on my side.

> In any case, let's delay installing this patch until after the
> emacs-28 release branch is cut, so as not to destabilize Emacs 28
> unnecessarily.

Yes good point! I will re-visit this patch on the following days to see
how I can improve it, alongside with Stefan's feedback.

Thanks for your feedback!