qemu-block
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Qemu-block] [RFC PATCH 13/41] block: Request real permissions in bd


From: Max Reitz
Subject: Re: [Qemu-block] [RFC PATCH 13/41] block: Request real permissions in bdrv_attach_child()
Date: Wed, 15 Feb 2017 20:23:00 +0100
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:45.0) Gecko/20100101 Thunderbird/45.7.0

On 13.02.2017 18:22, Kevin Wolf wrote:
> Now that all block drivers with children tell us what permissions they
> need from each of their children, bdrv_attach_child() can use this
> information and make the right requirements while trying to attach new
> children.
> 
> Signed-off-by: Kevin Wolf <address@hidden>
> ---
>  block.c | 17 +++++++++++++++--
>  1 file changed, 15 insertions(+), 2 deletions(-)
> 
> diff --git a/block.c b/block.c
> index 682548a..fd06549 100644
> --- a/block.c
> +++ b/block.c
> @@ -1558,10 +1558,21 @@ BdrvChild *bdrv_attach_child(BlockDriverState 
> *parent_bs,
>                               Error **errp)
>  {
>      BdrvChild *child;
> +    uint64_t perm, shared_perm;
> +
> +    /* XXX This check is a workaround for bdrv_open_inherit(), which attaches
> +     * bs->file before bs is opened for format probing. The 'else' branch
> +     * really shouldn't exist. */
> +    if (parent_bs->drv) {
> +        parent_bs->drv->bdrv_child_perm(parent_bs, NULL, child_role,
> +                                        0, BLK_PERM_ALL, &perm, 
> &shared_perm);
> +    } else {
> +        perm = 0;
> +        shared_perm = BLK_PERM_ALL;
> +    }
>  
> -    /* FIXME Use real permissions */
>      child = bdrv_root_attach_child(child_bs, child_name, child_role,
> -                                   0, BLK_PERM_ALL, parent_bs, errp);
> +                                   perm, shared_perm, parent_bs, errp);
>      if (child == NULL) {
>          return NULL;
>      }
> @@ -2041,6 +2052,8 @@ static BlockDriverState *bdrv_open_inherit(const char 
> *filename,
>       * probing, the block drivers will do their own bdrv_open_child() for the
>       * same BDS, which is why we put the node name back into options. */
>      if ((flags & BDRV_O_PROTOCOL) == 0) {
> +        /* FIXME Really shouldn't try to open a child for a still closed bs,
> +         * see bdrv_attach_child() */

That gives me hope for my remark on patch 1. :-)

Max

>          file = bdrv_open_child(filename, options, "file", bs,
>                                 &child_file, true, &local_err);
>          if (local_err) {
> 


Attachment: signature.asc
Description: OpenPGP digital signature


reply via email to

[Prev in Thread] Current Thread [Next in Thread]