Zsh Mailing List Archive
Messages sorted by:
Reverse Date,
Date,
Thread,
Author
Re: PATCH: remove ancient STOUC() workaround
- X-seq: zsh-workers 51213
- From: Peter Stephenson <p.w.stephenson@xxxxxxxxxxxx>
- To: Zsh workers <zsh-workers@xxxxxxx>
- Subject: Re: PATCH: remove ancient STOUC() workaround
- Date: Wed, 14 Dec 2022 17:07:17 +0000 (GMT)
- Archived-at: <https://zsh.org/workers/51213>
- Importance: Normal
- In-reply-to: <36532-1671036747.322209@Psau.rCcr.8m-L>
- List-id: <zsh-workers.zsh.org>
- References: <36532-1671036747.322209@Psau.rCcr.8m-L>
> On 14/12/2022 16:52 Oliver Kiddle <opk@xxxxxxx> wrote:
> The STOUC() macro is a wrapper for (unsigned char) casts that works
> around a compiler bug by doing ((unsigned char)(unsigned short)(X)) if
> autoconf detects the bug. This has been there since before the earliest
> (1995) list archives. I've never come across the need for this in any
> other C codebase. Surely whatever compiler needed this has long since
> ceased to be able to compile zsh for a myriad other reasons and been
> confined to a museum. A web search for the autoconf output of "signed to
> unsigned casting is broken..." finds matches with the word "no" but none
> with "yes". Or is there some further need for this that I'm missing?
I would imagine you're right and it's well past its shelf life. The only
way to know with this ancient stuff is to suck it and see.
pws
Messages sorted by:
Reverse Date,
Date,
Thread,
Author