[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <CAHk-=wikCh5JSR7=2mH0_d9hdAEtcG=Xz7-Md7xzp1GLfJFFKQ@mail.gmail.com>
Date: Mon, 18 Nov 2024 09:07:05 -0800
From: Linus Torvalds <torvalds@...ux-foundation.org>
To: Rasmus Villemoes <linux@...musvillemoes.dk>
Cc: Al Viro <viro@...iv.linux.org.uk>, Nir Lichtman <nir@...htman.org>, ebiederm@...ssion.com,
kees@...nel.org, brauner@...nel.org, jack@...e.cz,
linux-kernel@...r.kernel.org
Subject: Re: [PATCH v2] exec: make printable macro more concise
On Mon, 18 Nov 2024 at 03:45, Rasmus Villemoes <linux@...musvillemoes.dk> wrote:
>
> Exactly. But you said otherwise two years ago:
> https://lore.kernel.org/lkml/3a2fa7c1-2e31-0479-761f-9c189f8ed8c3@rasmusvillemoes.dk/
I do tend to stand by the "anybody who uses ctype deserves to get what
they asked for" thing.
But yeah, I think the time has come to admit that Latin1 isn't what we
want. And I'm looking at some of the users we _do_ have, and (ignoring
tools/ and script/) the most common ones seem to be
14 isascii
32 isalpha
46 isxdigit
56 isprint
59 isalnum
158 isspace
187 isdigit
and none of them would really care if we just limited it to ASCII
again. In fact, the isprint() ones would seem to generally be much
improved (looking at the ACPI uses).
I suspect we could make them be range-based instead of based on that
table lookup while at it.
So if somebody sends me a tested patch...
Linus
Powered by blists - more mailing lists