Date: Mon, 1 Nov 2004 07:04:20 -0700 (MST)
From: Zwane Mwaikambo <zwane@linuxpower.ca>
To: carlos@tarkus.se
cc: linux-kernel@vger.kernel.org
Subject: Re: spinlock_t typedef visibility and uninitialized spinlock
In-Reply-To: <cb62342b04110103057dc7dff0@mail.gmail.com>
Message-ID: <Pine.LNX.4.61.0411010702370.19123@musoma.fsmlabs.com>
References: <cb62342b04110103057dc7dff0@mail.gmail.com>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1315
Lines: 34

On Mon, 1 Nov 2004, Carlos Vidal wrote:

> I'm porting CIPE 1.6.0 kernel module to Kernel 2.6.8 and had problems
> with "spin_is_locked on uninitialized spinlock".
> 
> After tracing the problem I found that the spinlock_t structure is not
> visible to the module code. A 'gcc -E' yields:
>      typedef struct { } spinlock_t;
> 
> In spinlock.h, this declaration is inside a #ifdef
> CONFIG_DEBUG_SPINLOCK block, so it becomes visible only
> CONFIG_DEBUG_SPINLOCK is 'y'.
> 
> If I turn CONFIG_DEBUG_SPINLOCK on, the module loads nicely. Otherwise
> I get a nasty error in syslog and sometimes a system crash, as if in
> CIPE the struct was not allocated (what is the case if the compiler
> uses the typedef as it is above).
> 
> The question is: is this a bug or a feature? ;-)
> 
> Should the declaration be out of the #ifdef CONFIG_DEBUG_SPINLOCK? Or
> should I use a special compiler flag?

Try including <linux/config.h> at the top of the compilation unit and run 
through the preprocessor again. By the way, spinlock code got shuffled 
around in 2.6.9...

	Zwane

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/