Re: [PATCH] iommu: remove unnecessary code

[David Woodhouse]

On Wed, 2017-05-10 at 22:49 -0500, Gustavo A. R. Silva wrote:
> did_old is an unsigned variable and, greater-than-or-equal-to-zero
> comparison of an unsigned variable is always true.
> 
> Addresses-Coverity-ID: 1398477
> Signed-off-by: Gustavo A. R. Silva <garsilva@xxxxxxxxxxxxxx>

So... why do you think that check was there? Do you think it's possible
that someone mistakenly *thought* it could be negative? What were they
actually checking for? Have you actually *fixed* a bug here, or have
you just masked it?

Even if you've done all that analysis and it *is* correct just to drop
the comparison rather than fixing it, you need to put verbiage to that
effect into the commit comment.

Never write patches to "fix warnings". Always to fix bugs.

> ---
> Âdrivers/iommu/intel-iommu.c | 2 +-
> Â1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/drivers/iommu/intel-iommu.c b/drivers/iommu/intel-iommu.c
> index d412a31..98daf4a 100644
> --- a/drivers/iommu/intel-iommu.c
> +++ b/drivers/iommu/intel-iommu.c
> @@ -2050,7 +2050,7 @@ static int domain_context_mapping_one(struct dmar_domain *domain,
> Â	if (context_copied(context)) {
> Â		u16 did_old = context_domain_id(context);
> Â
> -		if (did_old >= 0 && did_old < cap_ndoms(iommu->cap))
> +		if (did_old < cap_ndoms(iommu->cap))
> Â			iommu->flush.flush_context(iommu, did_old,
> Â						ÂÂÂ(((u16)bus) << 8) | devfn,
> Â						ÂÂÂDMA_CCMD_MASK_NOBIT,

Attachment: smime.p7s
Description: S/MIME cryptographic signature