--signoff option is misleading when using --pull-request

[ehabkost]

Quoting @jnsnow , at https://lore.kernel.org/qemu-devel/13430122-c3b2-b68e-8cc8-59e2cb864add@redhat.com/

On 10/31/19 11:02 AM, Peter Maydell wrote:

Hi -- this passed the merge tests but it looks like you forgot
to add your signed-off by line as the submaintainer to Sam's
patches. Could you fix that up and resend, please?

thanks
-- PMM

I bit myself twice with this now: adding --signoff to a pull request
signs the messages that get sent to list, but not the ones that get staged.

Could always be a bug in my local copy, but I'm documenting it on the
list, in case I don't get time to look at this in the next 24h.