linux-kernel - Re: [PATCH] Unhide DEBUG_BUGVERBOSE when EXPERT=y, even if DEBUG

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: <20110605173143.GA1297@elte.hu>
Date:	Sun, 5 Jun 2011 19:31:43 +0200
From:	Ingo Molnar <mingo@...e.hu>
To:	Josh Triplett <josh@...htriplett.org>
Cc:	Sam Ravnborg <sam@...nborg.org>,
	Jeremy Fitzhardinge <jeremy@...p.org>,
	Andrew Morton <akpm@...ux-foundation.org>,
	Linus Torvalds <torvalds@...ux-foundation.org>,
	linux-kernel@...r.kernel.org
Subject: Re: [PATCH] Unhide DEBUG_BUGVERBOSE when EXPERT=y, even if
 DEBUG_KERNEL=n


* Josh Triplett <josh@...htriplett.org> wrote:

> On Sun, Jun 05, 2011 at 11:34:45AM +0200, Ingo Molnar wrote:
> > 
> > * Josh Triplett <josh@...htriplett.org> wrote:
> > 
> > >  config DEBUG_BUGVERBOSE
> > > -	bool "Verbose BUG() reporting (adds 70K)" if DEBUG_KERNEL && EXPERT
> > > +	bool "Verbose BUG() reporting (adds 70K)" if DEBUG_KERNEL || EXPERT
> > 
> > Well, DEBUG_KERNEL really means two things:
> > 
> >   - make more debugging options available
> >   - allow the *disabling* of existing (default-enabled) debug options
> 
> As well as one more: a quick "git grep DEBUG_KERNEL" turns up a few uses
> in actual kernel source code, to control debugging features.  Those
> should likely use separately selectable debug options, but they
> currently don't.
> 
> ~/src/linux-2.6$ find * -not -name 'Kconfig*' -not -name '*defconfig' | xargs grep -n DEBUG_KERNEL
> arch/parisc/mm/init.c:653:#ifdef CONFIG_DEBUG_KERNEL /* double-sanity-check paranoia */
> arch/powerpc/kernel/sysfs.c:229:#ifdef CONFIG_DEBUG_KERNEL
> arch/powerpc/kernel/sysfs.c:258:#endif /* CONFIG_DEBUG_KERNEL */
> arch/powerpc/kernel/sysfs.c:299:#ifdef CONFIG_DEBUG_KERNEL
> arch/powerpc/kernel/sysfs.c:328:#endif /* CONFIG_DEBUG_KERNEL */
> arch/blackfin/include/asm/entry.h:53:/* As a debugging aid - we save IPEND when DEBUG_KERNEL is on,
> arch/blackfin/include/asm/entry.h:56:# ifndef CONFIG_DEBUG_KERNEL
> arch/blackfin/include/asm/entry.h:65:# else /* CONFIG_DEBUG_KERNEL */
> arch/blackfin/include/asm/entry.h:77:# endif /* CONFIG_DEBUG_KERNEL */
> arch/blackfin/include/asm/context.S:208:#ifdef CONFIG_DEBUG_KERNEL
> drivers/usb/musb/musb_core.c:557:                * REVISIT:  do delays from lots of DEBUG_KERNEL checks

These are basically just 4 cases out of thousands of drivers, zero in 
essence. Also, none seems significant in terms of code size.

> > So i think the right solution would be to select DEBUG_KERNEL if 
> > EXPERT is enabled - this would simplify things and would allow the 
> > removal of a lot of EXPERT conditions from the debug options.
> > 
> > Ok?
> 
> I could live with that, as long as CONFIG_DEBUG_KERNEL never 
> directly enables any debugging code like it does above, and just 
> acts like CONFIG_EXPERT in hiding a pile of unnecessary options.  
> DEBUG_KERNEL should probably also have some text saying it doesn't 
> actually enable any kernel debugging on its own, once that becomes 
> true.

Yeah. I'd suggest you try that route instead of cluttering all of the 
debugging Kconfigs with || EXPERT conditions, which doesnt really 
look acceptable.

Thanks,

	Ingo
--
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/