Re: [PATCH] phy: fix compiler array bounds warning on settings[]

From: Bjorn Helgaas
Date: Wed Mar 05 2014 - 15:54:41 EST


[+cc Florian]

On Wed, Mar 5, 2014 at 2:10 AM, David Laight <David.Laight@xxxxxxxxxx> wrote:
> From: Bjorn Helgaas
>> With -Werror=array-bounds, gcc v4.7.x warns that in phy_find_valid(), the
>> settings[] "array subscript is above array bounds", I think because idx is
>> a signed integer and if the caller supplied idx < 0, we pass the guard but
>> still reference out of bounds.
>
> Not rejecting the patch but...
>
> Just indexing an array with 'int' shouldn't cause this warning,
> so somewhere a caller must actually be passing an idx < 0.
>
> While changing the type to unsigned will make the comparison
> against the array bound reject the -1, I suspect that the
> specific call path didn't really intend passing a hard-coded -1.
>
> It might be worth trying to locate the call site that passes -1.

I'm stumped. phy_find_valid() is static and only called from one
place. The 'idx' argument is always the result of phy_find_setting(),
which should always return something between 0 and
ARRAY_SIZE(settings), so I don't see any way idx can be < 0.

I stripped this down as far as I could; the resulting test code is at
http://pastebin.com/pp1zMEWu if anybody else wants to look at it. I'm
using gcc 4.8.x 20131105 (prerelease), with "-Warray-bounds -O2"
flags.

I hesitate to suspect a compiler bug, but it is very strange. For
example, in my test code, replacing "MAX_NUM_SETTINGS" with "2" gets
rid of the warnings. MAX_NUM_SETTINGS is known to be 2 at
compile-time, so I don't know why this should make a difference.

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