[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <dcdf1abc-7b8e-1f42-a955-0438b90fe9dc@rasmusvillemoes.dk>
Date: Wed, 4 Sep 2019 10:07:58 +0200
From: Rasmus Villemoes <linux@...musvillemoes.dk>
To: Jani Nikula <jani.nikula@...el.com>, linux-kernel@...r.kernel.org
Cc: Joonas Lahtinen <joonas.lahtinen@...ux.intel.com>,
Rodrigo Vivi <rodrigo.vivi@...el.com>,
intel-gfx@...ts.freedesktop.org,
Vishal Kulkarni <vishal@...lsio.com>, netdev@...r.kernel.org,
Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
linux-usb@...r.kernel.org,
Andrew Morton <akpm@...ux-foundation.org>,
Julia Lawall <julia.lawall@...6.fr>
Subject: Re: [PATCH 1/2] linux/kernel.h: add yesno(), onoff(),
enableddisabled(), plural() helpers
On 03/09/2019 15.37, Jani Nikula wrote:
> While the main goal here is to abstract recurring patterns, and slightly
> clean up the code base by not open coding the ternary operators, there
> are also some space savings to be had via better string constant
> pooling.
Eh, no? The linker does that across translation units anyway - moreover,
given that you make them static inlines, "yes" and "no" will still live
in .rodata.strX.Y in each individual TU that uses the yesno() helper.
The enableddisabled() is a mouthful, perhaps the helpers should have an
underscore between the choices
yes_no()
enabled_disabled()
on_off()
?
> drivers/gpu/drm/i915/i915_utils.h | 15 -------------
> .../ethernet/chelsio/cxgb4/cxgb4_debugfs.c | 11 ----------
> drivers/usb/core/config.c | 5 -----
> drivers/usb/core/generic.c | 5 -----
> include/linux/kernel.h | 21 +++++++++++++++++++
Pet peeve: Can we please stop using linux/kernel.h as a dumping ground
for every little utility/helper? That makes each and every translation
unit in the kernel slightly larger, hence slower to compile. Please make
a linux/string-choice.h and put them there.
Rasmus
Powered by blists - more mailing lists