mtd/maps/ceiva.c: possibly uninitialised local variable nr

mtd/maps/ceiva.c: possibly uninitialised local variable nr

[d binderman]

Hello there,

I just ran the sourceforge tool cppcheck over the source code of the
new Linux kernel 2.6.33-rc6

It said

[./mtd/maps/ceiva.c:269]: (error) Uninitialized variable: nr

The source code is

static int __init clps_setup_flash(void)
{
        int nr;

#ifdef CONFIG_ARCH_CEIVA
        if (machine_is_ceiva()) {
                info[0].base = CS0_PHYS_BASE;
                info[0].size = SZ_32M;
                info[0].width = CEIVA_FLASH_WIDTH;
                info[1].base = CS1_PHYS_BASE;
                info[1].size = SZ_32M;
                info[1].width = CEIVA_FLASH_WIDTH;
                nr = 2;
        }
#endif
        return nr;
}

There is no code to cope with the case that function machine_is_ceiva() returns 0
or that CONFIG_ARCH_CEIVA is undefined.

Suggest new belt'n'braces code

static int __init clps_setup_flash(void)
{
        int nr = 0;

#ifdef CONFIG_ARCH_CEIVA
        if (machine_is_ceiva()) {
                info[0].base = CS0_PHYS_BASE;
                info[0].size = SZ_32M;
                info[0].width = CEIVA_FLASH_WIDTH;
                info[1].base = CS1_PHYS_BASE;
                info[1].size = SZ_32M;
                info[1].width = CEIVA_FLASH_WIDTH;
                nr = 2;
        }
#endif
        return nr;
}

Regards

David Binderman

 		 	   		  
_________________________________________________________________
Tell us your greatest, weirdest and funniest Hotmail stories
http://clk.atdmt.com/UKM/go/195013117/direct/01/

Re: mtd/maps/ceiva.c: possibly uninitialised local variable nr

[Artem Bityutskiy]

On Mon, 2010-02-01 at 10:26 +0000, d binderman wrote:

[snip]

> static int __init clps_setup_flash(void)
> {
>         int nr = 0;
> 
> #ifdef CONFIG_ARCH_CEIVA
>         if (machine_is_ceiva()) {
>                 info[0].base = CS0_PHYS_BASE;
>                 info[0].size = SZ_32M;
>                 info[0].width = CEIVA_FLASH_WIDTH;
>                 info[1].base = CS1_PHYS_BASE;
>                 info[1].size = SZ_32M;
>                 info[1].width = CEIVA_FLASH_WIDTH;
>                 nr = 2;
>         }
> #endif
>         return nr;
> }

I've made a patch for this in my l2-mtd-2.6 / master, thanks. Here is
the URL, but it is not persistent:

http://git.infradead.org/users/dedekind/l2-mtd-2.6.git/commit/331d53a3eca96f7ae0a216aa9eb3addffccc9e6f

-- 
Best Regards,
Artem Bityutskiy (Артём Битюцкий)

Re: mtd/maps/ceiva.c: possibly uninitialised local variable nr

[David Woodhouse]

On Mon, 2010-02-01 at 10:26 +0000, d binderman wrote:
> There is no code to cope with the case that function
> machine_is_ceiva() returns 0 or that CONFIG_ARCH_CEIVA is undefined.

CONFIG_ARCH_CEIVA can never be undefined -- this driver depends on
ARCH_CEIVA in Kconfig. I don't quite understand why it's littered with
the ifdefs. Russell?

The error you spot could conceivably happen in a multiplatform build
though (if such a build {is,becomes} possible for ARM platforms).

-- 
David Woodhouse                            Open Source Technology Centre
David.Woodhouse@intel.com                              Intel Corporation

Re: mtd/maps/ceiva.c: possibly uninitialised local variable nr

[Russell King]

On Thu, Feb 25, 2010 at 12:50:24PM +0000, David Woodhouse wrote:
> On Mon, 2010-02-01 at 10:26 +0000, d binderman wrote:
> > There is no code to cope with the case that function
> > machine_is_ceiva() returns 0 or that CONFIG_ARCH_CEIVA is undefined.
> 
> CONFIG_ARCH_CEIVA can never be undefined -- this driver depends on
> ARCH_CEIVA in Kconfig. I don't quite understand why it's littered with
> the ifdefs. Russell?

Me neither.  Better ask 'Rob Scott' who appears to be the person
who created this in the first place.  Whether he's still interested
in it, or even doing Linux stuff I've no idea.

-- 
Russell King
 Linux kernel    2.6 ARM Linux   - http://www.arm.linux.org.uk/
 maintainer of: