[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <1428701143.17822.72.camel@x220>
Date: Fri, 10 Apr 2015 23:25:43 +0200
From: Paul Bolle <pebolle@...cali.nl>
To: Gregory Fong <gregory.0xf0@...il.com>
Cc: Michal Marek <mmarek@...e.cz>,
Valentin Rothberg <valentinrothberg@...il.com>,
Stefan Hengelein <stefan.hengelein@....de>,
Andreas Ruprecht <rupran@...server.de>,
Martin Walch <walch.martin@....de>,
linux-kbuild@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH 1/2] kconfig: Print full defined and depends for
multiply-defined symbols
[Removed Yann. Added the people that I hope might actually understand
what this is all about.]
On Wed, 2015-04-08 at 16:56 -0700, Gregory Fong wrote:
> get_symbol_str() was assuming that symbols would only have a single
> property for the purpose of printing define and depends information.
> This is not true, and one current example is FRAME_POINTER which is
> both in lib/Kconfig.debug and arch/arm/Kconfig.debug.
>
> In order to print out the correct Defined and Depends info, iterate
> over all properties associated with the given symbol, similarly to was
> done for selects. And for depends, rather than iterating over the
> property, just use the direct dependency expression.
>
> CONFIG_FRAME_POINTER text, before:
> Defined at lib/Kconfig.debug:323
> Depends on: DEBUG_KERNEL [=y] && (ARM [=y] || CRIS || M68K || FRV || UML || AVR32 || SUPERH || BLACKFIN || MN10300 || METAG) || ARCH_WANT_FRAME_POINTERS [=n]
>
> After:
> Defined at lib/Kconfig.debug:323, arch/arm/Kconfig.debug:35
> Depends on: DEBUG_KERNEL [=y] && (ARM [=y] || CRIS || M68K || FRV || UML || AVR32 || SUPERH || BLACKFIN || MN10300 || METAG) || ARCH_WANT_FRAME_POINTERS [=n] || !THUMB2_KERNEL [=n]
Note that there are currently six places where FRAME_POINTER is defined:
arch/arm/Kconfig.debug:35:config FRAME_POINTER
arch/arm64/Kconfig.debug:5:config FRAME_POINTER
arch/hexagon/Kconfig:40:config FRAME_POINTER
arch/m32r/Kconfig.debug:13:config FRAME_POINTER
arch/sparc/Kconfig.debug:19:config FRAME_POINTER
lib/Kconfig.debug:323:config FRAME_POINTER
Anyhow, does anyone dare to state that the After: line above describes
FRAME_POINTER for arm correctly? (Of course, I could dive in the semi
documented kconfig code myself. But that might mean that Gregory will be
waiting for feedback for quite some time.)
> Removes now-unused function get_symbol_prop().
It might be better to do that in a separate patch.
> Signed-off-by: Gregory Fong <gregory.0xf0@...il.com>
> ---
> scripts/kconfig/menu.c | 35 +++++++++++++++--------------------
> 1 file changed, 15 insertions(+), 20 deletions(-)
>
> diff --git a/scripts/kconfig/menu.c b/scripts/kconfig/menu.c
> index 72c9dba..da482ff 100644
> --- a/scripts/kconfig/menu.c
> +++ b/scripts/kconfig/menu.c
> @@ -601,18 +601,6 @@ static void get_prompt_str(struct gstr *r, struct property *prop,
> }
>
> /*
> - * get property of type P_SYMBOL
> - */
> -static struct property *get_symbol_prop(struct symbol *sym)
> -{
> - struct property *prop = NULL;
> -
> - for_all_properties(sym, prop, P_SYMBOL)
> - break;
> - return prop;
> -}
> -
> -/*
> * head is optional and may be NULL
> */
> void get_symbol_str(struct gstr *r, struct symbol *sym,
> @@ -637,15 +625,22 @@ void get_symbol_str(struct gstr *r, struct symbol *sym,
> for_all_prompts(sym, prop)
> get_prompt_str(r, prop, head);
>
> - prop = get_symbol_prop(sym);
> - if (prop) {
> - str_printf(r, _(" Defined at %s:%d\n"), prop->menu->file->name,
> + hit = false;
> + for_all_properties(sym, prop, P_SYMBOL) {
> + if (!hit) {
> + str_append(r, " Defined at ");
> + hit = true;
> + } else
> + str_append(r, ", ");
> + str_printf(r, _("%s:%d"), prop->menu->file->name,
> prop->menu->lineno);
> - if (!expr_is_yes(prop->visible.expr)) {
> - str_append(r, _(" Depends on: "));
> - expr_gstr_print(prop->visible.expr, r);
> - str_append(r, "\n");
> - }
> + }
> + if (hit)
> + str_append(r, "\n");
> + if (!expr_is_yes(sym->dir_dep.expr)) {
> + str_append(r, _(" Depends on: "));
> + expr_gstr_print(sym->dir_dep.expr, r);
> + str_append(r, "\n");
> }
>
> hit = false;
Thanks,
Paul Bolle
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists