linux-kernel - Re: kbuild: fixing the select problem

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-Id: <20100506135952.eeb0e4c5.randy.dunlap@oracle.com>
Date:	Thu, 6 May 2010 13:59:52 -0700
From:	Randy Dunlap <randy.dunlap@...cle.com>
To:	James Bottomley <James.Bottomley@...senPartnership.com>
Cc:	Michal Marek <mmarek@...e.cz>, linux-arch@...r.kernel.org,
	linux-kernel <linux-kernel@...r.kernel.org>,
	Sam Ravnborg <sam@...nborg.org>, linux-kbuild@...r.kernel.org,
	Vegard Nossum <vegard.nossum@...il.com>
Subject: Re: kbuild: fixing the select problem

On Thu, 06 May 2010 15:48:35 -0500 James Bottomley wrote:

> On Thu, 2010-05-06 at 09:52 -0500, James Bottomley wrote:
> > The list is
> > 
> > USB_ARCH_HAS_HCD has 4 defaults
> > DEFCONFIG_LIST has 5 defaults
> > MAC80211_RC_DEFAULT has 2 defaults
> > X86_L1_CACHE_SHIFT has 2 defaults
> > SPLIT_PTLOCK_CPUS has 2 defaults
> > X86_MINIMUM_CPU_FAMILY has 3 defaults
> > DEFAULT_TCP_CONG has 2 defaults
> > DEFCONFIG_LIST has 5 defaults
> > USB_ARCH_HAS_HCD has 4 defaults
> > X86_L1_CACHE_SHIFT has 2 defaults
> > X86_MINIMUM_CPU_FAMILY has 3 defaults
> > SPLIT_PTLOCK_CPUS has 2 defaults
> > DEFAULT_TCP_CONG has 2 defaults
> > MAC80211_RC_DEFAULT has 2 defaults
> 
> Here's a patch that alters the default processing to the needed form and
> checks the old vs new values.  I've been running randconfigs but I can't
> get the warning to trip ... have at it.
> 

I don't see the warning, but I took a problem config from linux-next 2010-0506
(today) [attached].  In it, CONFIG_TCG_BIOS_LOG=y and that selects ACPI,
so now ACPI is enabled, but ACPI depends on PM and PCI, but they are still not
selected...

or is this only step 1 of a multi-step implementation?

> 
> ---
> 
> diff --git a/scripts/kconfig/symbol.c b/scripts/kconfig/symbol.c
> index 6c8fbbb..722bc4e 100644
> --- a/scripts/kconfig/symbol.c
> +++ b/scripts/kconfig/symbol.c
> @@ -115,14 +115,31 @@ struct property *sym_get_env_prop(struct symbol *sym)
>  
>  struct property *sym_get_default_prop(struct symbol *sym)
>  {
> -	struct property *prop;
> +	struct property *prop, *ret = NULL;
> +	tristate old_val = no, val = no;
>  
>  	for_all_defaults(sym, prop) {
>  		prop->visible.tri = expr_calc_value(prop->visible.expr);
> -		if (prop->visible.tri != no)
> -			return prop;
> +		if (prop->visible.tri != no) {
> +			tristate v = expr_calc_value(prop->expr);
> +			if (!ret)
> +				old_val = v;
> +			if (v >= val) {
> +				val = v;
> +				ret = prop;
> +			}
> +		}
>  	}
> -	return NULL;
> +	/*
> +	 * Previously, we took the first valid default we found (this
> +	 * is now old_val).  In the new scheme, the value is the or of
> +	 * all the defaults.
> +	 */
> +	if (old_val != val)
> +		menu_warn(ret->menu, "ERROR: new parser has inconsistent "
> +			  "values for %s (%d != %d)\n", sym->name,
> +			  old_val, val);
> +	return ret;
>  }
>  
>  static struct property *sym_get_range_prop(struct symbol *sym)

---
~Randy
*** Remember to use Documentation/SubmitChecklist when testing your code ***

Download attachment "config-tpm" of type "application/octet-stream" (58581 bytes)