Re: [Qemu-devel] a64 simd decode in handle_vec_simd_shli()

[Philippe Mathieu-Daudé]

On 29/10/18 13:06, Peter Maydell wrote:
> Hi; Coverity is complaining (in CID 1396476) about a problem in the
> handle_vec_simd_shli() function, where we might dereference sli_op[]
> with a size that's greater than 3. It thinks size might be > 3 because
> we do a check
>       if (size > 3 && !is_q) {
>           unallocated_encoding(s);
>           return;
>       }
> suggesting that we could have is_q and size > 3.
>
> I'm having difficulty figuring out where this check has come from;
> it doesn't seem to match up with the pseudocode and in any case
> I don't think size can ever be > 3. We calculate:
>
>      int size = 32 - clz32(immh) - 1;
> where immh is a 4 bit field which we know cannot be all-zeroes.
> So the clz32() return must be in {28,29,30,31} and the resulting
> size is in {0,1,2,3}, so the check above can't ever fire.

Clang was emitting the same warning in 
memory_region_oldmmio_read_accessor() before you remove it:

    tmp = mr->ops->old_mmio.read[ctz32(size)](mr->opaque, addr);

> Am I missing something? As far as I can see we should simply delete
> the can't-happen condition, which will probably satisfy coverity.
>
> thanks
> -- PMM