Re: [PATCH] doc: Clarify NBD_CMD_READ payload on error

[Wouter Verhelst <w@uter.be>]

On Mon, Nov 13, 2017 at 10:49:39AM -0600, Eric Blake wrote:
> Commit e6c6fb39 clarified that a payload for the reply to NBD_CMD_READ
> must not be sent if the error field was set (which matches existing
> nbd-server and qemu-nbd behavior; I'm also in the process of patching
> nbdkit to comply).  But it missed another spot in the document, which
> if read in isolation, results in the wrong behavior.

Good catch, thanks. Please apply.

> Signed-off-by: Eric Blake <eblake@redhat.com>
> ---
>  doc/proto.md | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/doc/proto.md b/doc/proto.md
> index f6a9063..88154f1 100644
> --- a/doc/proto.md
> +++ b/doc/proto.md
> @@ -340,7 +340,8 @@ follows:
>  S: 32 bits, 0x67446698, magic (`NBD_REPLY_MAGIC`)  
>  S: 32 bits, error (MAY be zero)  
>  S: 64 bits, handle  
> -S: (*length* bytes of data if the request is of type `NBD_CMD_READ`)  
> +S: (*length* bytes of data if the request is of type `NBD_CMD_READ` and
> +    *error* is zero)  
> 
>  #### Terminating the transmission phase
> 
> -- 
> 2.13.6
> 
> 

-- 
Could you people please use IRC like normal people?!?

  -- Amaya Rodrigo Sastre, trying to quiet down the buzz in the DebConf 2008
     Hacklab