[PATCH] [RFC] kexec: Fix off-by-one errors in locate_hole()

[Simon Horman]

On Mon, Sep 23, 2013 at 10:18:12AM +0200, Geert Uytterhoeven wrote:
> When calling locate_hole() with "hole_size" equal to the size of an
> available memory block, it fails to use that memory block.
> 
> "end" and "hole_max" point to the last byte within the range, hence
>   - "size = end - start" is one less than "hole_size",
>   - "hole_base + hole_size" is one more than "hole_max".
> 
> Subtract one from "hole_size" when doing the comparison (adding 1 to "size"
> could overflow in case of one big range covering the whole address space).
> 
> Signed-off-by: Geert Uytterhoeven <geert at linux-m68k.org>
> ---
> Question:
>     The code no longer handles the case where "hole_size" is zero.
>     Should this be rejected (like is done for a zero "hole_end" at the top
>     of the function), or accepted?

Either way would be fine by me though I lean towards accepting it.

> 
>  kexec/kexec.c |    4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
> 
> diff --git a/kexec/kexec.c b/kexec/kexec.c
> index 2b98ef0..eafd6c2 100644
> --- a/kexec/kexec.c
> +++ b/kexec/kexec.c
> @@ -270,7 +270,7 @@ unsigned long locate_hole(struct kexec_info *info,
>  		}
>  		/* Is there enough space left so we can use it? */
>  		size = end - start;
> -		if (size >= hole_size) {
> +		if (size >= hole_size - 1) {
>  			if (hole_end > 0) {
>  				hole_base = start;
>  				break;
> @@ -286,7 +286,7 @@ unsigned long locate_hole(struct kexec_info *info,
>  			"0x%lx bytes...\n", hole_size);
>  		return ULONG_MAX;
>  	}
> -	if ((hole_base + hole_size)  > hole_max) {
> +	if ((hole_base + hole_size - 1)  > hole_max) {
>  		fprintf(stderr, "Could not find a free area of memory below: "
>  			"0x%lx...\n", hole_max);
>  		return ULONG_MAX;
> -- 
> 1.7.9.5
>