Re: [Qemu-devel] [PATCH v15 08/14] block: Support dropping active in bdrv_drop_intermediate

[Jeff Cody <jcody@redhat.com>]

On Sun, Feb 23, 2014 at 09:54:49AM +0800, Fam Zheng wrote:
> Dropping intermediate could be useful both for commit and stream, and
> BDS refcnt plus bdrv_swap could do most of the job nicely. It also needs
> to work with op blockers.
> 
> Signed-off-by: Fam Zheng <famz@redhat.com>
> ---
>  block.c        | 137 +++++++++++++++++++++++----------------------------------
>  block/commit.c |   2 +-
>  2 files changed, 56 insertions(+), 83 deletions(-)
> 
> diff --git a/block.c b/block.c
> index a8d1012..13cbdbf 100644
> --- a/block.c
> +++ b/block.c
> @@ -2495,115 +2495,88 @@ BlockDriverState *bdrv_find_overlay(BlockDriverState *active,
>      return overlay;
>  }
>  
> -typedef struct BlkIntermediateStates {
> -    BlockDriverState *bs;
> -    QSIMPLEQ_ENTRY(BlkIntermediateStates) entry;
> -} BlkIntermediateStates;
> -
> -
>  /*
> - * Drops images above 'base' up to and including 'top', and sets the image
> - * above 'top' to have base as its backing file.
> + * Drops images above 'base' up to and including 'top', and sets new 'base' as
> + * backing_hd of top's overlay (the image orignally has 'top' as backing file).
> + * top's overlay may be NULL if 'top' is active, no such update needed.
> + * Requires that the top's overlay to 'top' is opened r/w.
> + *
> + * 1) This will convert the following chain:
> + *
> + *     ... <- base <- ... <- top <- overlay <-... <- active
> + *
> + * to
> + *
> + *     ... <- base <- overlay <- active
>   *
> - * Requires that the overlay to 'top' is opened r/w, so that the backing file
> - * information in 'bs' can be properly updated.
> + * 2) It is allowed for bottom==base, in which case it converts:
>   *
> - * E.g., this will convert the following chain:
> - * bottom <- base <- intermediate <- top <- active
> + *     base <- ... <- top <- overlay <- ... <- active
>   *
>   * to
>   *
> - * bottom <- base <- active
> + *     base <- overlay <- active
>   *
> - * It is allowed for bottom==base, in which case it converts:
> + * 2) It also allows active==top, in which case it converts:
>   *
> - * base <- intermediate <- top <- active
> + *     ... <- base <- ... <- top (active)
>   *
>   * to
>   *
> - * base <- active
> + *     ... <- base == active == top
> + *
> + * i.e. only base and lower remains: *top == *base when return.
> + *
> + * 3) If base==NULL, it will drop all the BDS below overlay and set its
> + * backing_hd to NULL. I.e.:
> + *
> + *     base(NULL) <- ... <- overlay <- ... <- active
> + *
> + * to
>   *
> - * Error conditions:
> - *  if active == top, that is considered an error
> + *     overlay <- ... <- active
>   *
>   */
>  int bdrv_drop_intermediate(BlockDriverState *active, BlockDriverState *top,
>                             BlockDriverState *base)
>  {
> -    BlockDriverState *intermediate;
> -    BlockDriverState *base_bs = NULL;
> -    BlockDriverState *new_top_bs = NULL;
> -    BlkIntermediateStates *intermediate_state, *next;
> -    int ret = -EIO;
> -
> -    QSIMPLEQ_HEAD(states_to_delete, BlkIntermediateStates) states_to_delete;
> -    QSIMPLEQ_INIT(&states_to_delete);
> -
> -    if (!top->drv || !base->drv) {
> -        goto exit;
> -    }
> +    BlockDriverState *drop_start, *overlay;
> +    int ret = -EINVAL;
>  
> -    new_top_bs = bdrv_find_overlay(active, top);
> -
> -    if (new_top_bs == NULL) {
> -        /* we could not find the image above 'top', this is an error */
> +    if (!top->drv || (base && !base->drv)) {
>          goto exit;
>      }
> -
> -    /* special case of new_top_bs->backing_hd already pointing to base - nothing
> -     * to do, no intermediate images */
> -    if (new_top_bs->backing_hd == base) {
> +    if (top == base) {
>          ret = 0;
> -        goto exit;
> -    }
> -
> -    intermediate = top;
> -
> -    /* now we will go down through the list, and add each BDS we find
> -     * into our deletion queue, until we hit the 'base'
> -     */
> -    while (intermediate) {
> -        intermediate_state = g_malloc0(sizeof(BlkIntermediateStates));
> -        intermediate_state->bs = intermediate;
> -        QSIMPLEQ_INSERT_TAIL(&states_to_delete, intermediate_state, entry);
> -
> -        if (intermediate->backing_hd == base) {
> -            base_bs = intermediate->backing_hd;
> -            break;
> +    } else if (top == active) {
> +        assert(base);

Hmm.  I'm trying to think of any edge cases here, if something could
fail.  We know base != NULL, we know that top = active, and that
top != base.  That is good.  But we don't have any verification at
this point that base is in the same chain as top / active.

Is this worth verifying?  I think it probably is...  a simple way
would be to just add here something like this:

   drop_start = active;
   while (drop_start && drop_start != base) {
       drop_start = drop_start->backing_hd;
   } 
   if (drop_start == NULL) {
       goto exit;
   }

(or use a new BDS pointer for better clarity, rather than reusing
drop_start, if desired)

> +        drop_start = active->backing_hd;
> +        bdrv_swap(active, base);
> +        base->backing_hd = NULL;
> +        bdrv_unref(drop_start);
> +        ret = 0;
> +    } else if (base) {
> +        overlay = bdrv_find_overlay(active, top);

Error check needed - overlay could be NULL.

> +        bdrv_set_backing_hd(overlay, base);
> +        ret = 0;

This breaks block-commit here, because the backing file is not changed
in the overlay image file.

> +    } else {
> +        /* If there's an overlay, its backing_hd points to top's BDS now,
> +         * the top image is dropped but this BDS structure is kept and swapped
> +         * with base, this way we keep the pointers valid after dropping top */
> +        overlay = bdrv_find_overlay(active, top);
> +        if (!overlay) {
> +            goto exit;
>          }
> -        intermediate = intermediate->backing_hd;
> -    }
> -    if (base_bs == NULL) {
> -        /* something went wrong, we did not end at the base. safely
> -         * unravel everything, and exit with error */
> -        goto exit;
> -    }
> -
> -    /* success - we can delete the intermediate states, and link top->base */
> -    ret = bdrv_change_backing_file(new_top_bs, base_bs->filename,
> -                                   base_bs->drv ? base_bs->drv->format_name : "");
> -    if (ret) {
> -        goto exit;
> -    }
> -    new_top_bs->backing_hd = base_bs;
> -
> -    bdrv_refresh_limits(new_top_bs);
> -
> -    QSIMPLEQ_FOREACH_SAFE(intermediate_state, &states_to_delete, entry, next) {
> -        /* so that bdrv_close() does not recursively close the chain */
> -        intermediate_state->bs->backing_hd = NULL;
> -        bdrv_unref(intermediate_state->bs);
> +        ret = bdrv_change_backing_file(overlay, NULL, NULL);
> +        if (ret) {
> +            goto exit;
> +        }

As mentioned above, the backing file needs to be changed in each case
(except the top == base case), not just when base == NULL.

> +        bdrv_set_backing_hd(overlay, NULL);
>      }
> -    ret = 0;
> -
>  exit:
> -    QSIMPLEQ_FOREACH_SAFE(intermediate_state, &states_to_delete, entry, next) {
> -        g_free(intermediate_state);
> -    }
>      return ret;
>  }
>  
> -
>  static int bdrv_check_byte_request(BlockDriverState *bs, int64_t offset,
>                                     size_t size)
>  {
> diff --git a/block/commit.c b/block/commit.c
> index acec4ac..ac598d6 100644
> --- a/block/commit.c
> +++ b/block/commit.c
> @@ -73,6 +73,7 @@ static void coroutine_fn commit_run(void *opaque)
>      int bytes_written = 0;
>      int64_t base_len;
>  
> +    overlay_bs = bdrv_find_overlay(active, top);
>      ret = s->common.len = bdrv_getlength(top);
>  
>  
> @@ -154,7 +155,6 @@ exit_restore_reopen:
>      if (s->base_flags != bdrv_get_flags(base)) {
>          bdrv_reopen(base, s->base_flags, NULL);
>      }
> -    overlay_bs = bdrv_find_overlay(active, top);
>      if (overlay_bs && s->orig_overlay_flags != bdrv_get_flags(overlay_bs)) {
>          bdrv_reopen(overlay_bs, s->orig_overlay_flags, NULL);
>      }
> -- 
> 1.8.5.4
> 
>