Re: [Qemu-devel] [PATCH v4 09/17] dirty-bitmap: Change bdrv_dirty_iter_next() to report byte offset

[Eric Blake]

On 07/03/2017 10:10 AM, Eric Blake wrote:
> Thanks to recent cleanups, most callers were scaling a return value
> of sectors into bytes (the exception, in qcow2-bitmap, will be
> converted to byte-based iteration later).  Update the interface to
> do the scaling internally instead.
> 
> Signed-off-by: Eric Blake <address@hidden>

> @@ -506,7 +506,7 @@ void bdrv_dirty_iter_free(BdrvDirtyBitmapIter *iter)
> 
>  int64_t bdrv_dirty_iter_next(BdrvDirtyBitmapIter *iter)
>  {
> -    return hbitmap_iter_next(&iter->hbi);
> +    return hbitmap_iter_next(&iter->hbi) * BDRV_SECTOR_SIZE;
>  }
> 

> +++ b/block/qcow2-bitmap.c
> @@ -1105,7 +1105,7 @@ static uint64_t *store_bitmap_data(BlockDriverState *bs,
>      sbc = limit >> BDRV_SECTOR_BITS;
>      assert(DIV_ROUND_UP(bm_size, limit) == tb_size);
> 
> -    while ((sector = bdrv_dirty_iter_next(dbi)) != -1) {
> +    while ((sector = bdrv_dirty_iter_next(dbi) >> BDRV_SECTOR_BITS) != -1) {

Does clang's sanitizer complain about right shift of a negative number,
or is that one of the things that we can rely on even though C says it
is not strictly portable?  If it is a problem, I can always split the
check for bdrv_dirty_iter_next()<0 to occur separately from the while()
conditional; but I'd rather not go through with the churn unless it
actually chokes a build-bot as written.

-- 
Eric Blake, Principal Software Engineer
Red Hat, Inc.           +1-919-301-3266
Virtualization:  qemu.org | libvirt.org

signature.asc
Description: OpenPGP digital signature