Re: [PATCH v2 01/33] block: Add BlockDriver.is_format

[Eric Blake]

On 2/4/20 11:08 AM, Max Reitz wrote:
> We want to unify child_format and child_file at some point.  One of the
> important things that set format drivers apart from other drivers is
> that they do not expect other format nodes under them (except in the
> backing chain).  That means we need something on which to distinguish
> format drivers from others, and hence this flag.

It _is_ possible to set up a format node on top of another; in fact, our 
testsuite does that in at least iotest 072.  I agree that setups like 
'qcow2 - qcow2 - file' are uncommon, but the setup 'qcow2 - raw - file' 
may be useful for extracting a partition of a raw disk when it is known 
that the partition in that disk itself contains qcow2 data.

> Signed-off-by: Max Reitz <address@hidden>
> ---

> +++ b/include/block/block_int.h
> @@ -94,6 +94,13 @@ struct BlockDriver {
>        * must implement them and return -ENOTSUP.
>        */
>       bool is_filter;
> +    /*
> +     * Set to true if the BlockDriver is a format driver.  Format nodes
> +     * generally do not expect their children to be other format nodes
> +     * (except for backing files), and so format probing is disabled
> +     * on those children.

Aha - nested formats ARE still allowed when you explicitly request it 
(which is what iotest 72 does) - what you are stating here is that 
implicit probing of is forbidden for a parent declared as a format 
driver.  That makes more sense.

I'm not sure if the commit message needs a tweak, but the patch itself 
is sane as-is.

Reviewed-by: Eric Blake <address@hidden>

> +     */
> +    bool is_format;
>       /*
>        * Return true if @to_replace can be replaced by a BDS with the
>        * same data as @bs without it affecting @bs's behavior (that is,

--
Eric Blake, Principal Software Engineer
Red Hat, Inc.           +1-919-301-3226
Virtualization:  qemu.org | libvirt.org