[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: scratch/comp-static-data 5aa3db2f11: comp: Add support for compiling

From: Eli Zaretskii
Subject: Re: scratch/comp-static-data 5aa3db2f11: comp: Add support for compiling elisp constants into static data.
Date: Fri, 18 Nov 2022 10:28:26 +0200

> From: Richard Stallman <rms@gnu.org>
> Cc: akrl@sdf.org, luangruo@yahoo.com, emacs-devel@gnu.org
> Date: Fri, 18 Nov 2022 00:07:00 -0500
>   > Perhaps "constant" is a misnomer here. If you are refering to defconst-
>   > defined variables, that is not what the patch is about. I'm refering to
>   > self-evaluating forms/objects referenced by elisp code, i.e. the ones
>   > that the constant vector in a bytecode object contains. The elisp
>   > reference manual states that mutating self-evaluating forms results in
>   > undefined behaviour, so marking them as const should be safe IMO.
> Your change "ought to be" harmless in principle, but in practice it
> may be harmful.  There may be cases that are currently "incorrect" in
> principle but in practice do no harm.  Your change might cause them to
> get faults for trying to write read-only memory.  Why take that risk?
> I'd rather avoid risk by not making this change.

I'm not aware of any aspect of the change which would cause crashes in
these cases.  Vibhav, if this is true, please show the relevant code
in your branch, and explain how it works and in what cases it causes
an abort.


reply via email to

[Prev in Thread] Current Thread [Next in Thread]