Re: [Xen-devel] [PATCH 07/10 V7] libxl: use the API to setup/teardown network buffering

[Ian Jackson <Ian.Jackson@xxxxxxxxxxxxx>]

Lai Jiangshan writes ("[PATCH 07/10 V7] libxl: use the API to setup/teardown 
network buffering"):
> From: Shriram Rajagopalan <rshriram@xxxxxxxxx>
> 
> If there is network buffering hotplug scripts, call
> libxl__remus_netbuf_setup() to setup the network
> buffering and libxl__remus_netbuf_teardown() to
> teardown network buffering.

> +    if (dss->remus_state) {
> +        /*
> +         * With Remus, if we reach this point, it means either
> +         * backup died or some network error occurred preventing us
> +         * from sending checkpoints. Teardown the network buffers and
> +         * release netlink resources.  This is an async op.
> +         */
> +        libxl__remus_teardown_initiate(egc, dss, rc);
> +        return;
> +    }

This patch seems plausible.  But I wonder if it might not be better to
provide a firmer interface between the remus code and the rest of the
save/restore machinery.  That is, have an explicit callback function
recorded by the save/restore code which is called back by the remus
machinery when it has done its work.  What do you think ?

I think having the flow of control spring off into libxl_remus.c and
magically come back by libxl_remus.c knowing to call
domain_suspend_done is rather opaque.

Thanks,
Ian.

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel