Re: [Qemu-block] [PATCH 05/12] hbitmap: enable merging across granularities

[Vladimir Sementsov-Ogievskiy]

20.06.2019 4:03, John Snow wrote:
> Signed-off-by: John Snow <address@hidden>
> ---
>   util/hbitmap.c | 22 +++++++++++++++++++++-
>   1 file changed, 21 insertions(+), 1 deletion(-)
> 
> diff --git a/util/hbitmap.c b/util/hbitmap.c
> index 45d1725daf..0d6724b7bc 100644
> --- a/util/hbitmap.c
> +++ b/util/hbitmap.c
> @@ -777,7 +777,17 @@ void hbitmap_truncate(HBitmap *hb, uint64_t size)
>   
>   bool hbitmap_can_merge(const HBitmap *a, const HBitmap *b)
>   {
> -    return (a->size == b->size) && (a->granularity == b->granularity);
> +    return (a->size == b->size);
> +}
> +
> +static void hbitmap_sparse_merge(HBitmap *dst, const HBitmap *src)
> +{
> +    uint64_t offset = 0;
> +    uint64_t count = src->orig_size;
> +
> +    while (hbitmap_next_dirty_area(src, &offset, &count)) {
> +        hbitmap_set(dst, offset, count);

This will not work, as hbitmap_next_dirty_area will return the same answer all 
the time I think.
You need to update offset and count in a loop, like it is done in 
backup_incremental_init_copy_bitmap.


> +    }
>   }
>   
>   /**
> @@ -804,6 +814,16 @@ bool hbitmap_merge(const HBitmap *a, const HBitmap *b, 
> HBitmap *result)
>           return true;
>       }
>   
> +    if (a->size != b->size) {
> +        if (a != result) {
> +            hbitmap_sparse_merge(result, a);
> +        }
> +        if (b != result) {
> +            hbitmap_sparse_merge(result, b);
> +        }
> +        return true;
> +    }
> +
>       /* This merge is O(size), as BITS_PER_LONG and HBITMAP_LEVELS are 
> constant.
>        * It may be possible to improve running times for sparsely populated 
> maps
>        * by using hbitmap_iter_next, but this is suboptimal for dense maps.
> 


-- 
Best regards,
Vladimir