Re: [PATCH v8 11/11] (lib)xl: soft reset support

[Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>]

On Tue, Jun 23, 2015 at 06:11:53PM +0200, Vitaly Kuznetsov wrote:
> Use existing create/restore path to perform 'soft reset' for HVM domains.
> Tear everything down, e.g. destroy domain's device model, remove the domain
> from xenstore, save toolstack record and start over.
> 
> Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
> ---
> Changes since 'v7'
> - 'Reset everything approach': XEN_DOMCTL_soft_reset doesn't destroy the
>   original domain any more.
> - libxl__domain_soft_reset_state introduced [Ian Campbell]
> - Separate do_domain_soft_reset to not intertwine with do_domain_create
>   [Ian Campbell]
> 
> Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
> ---
>  docs/man/xl.cfg.pod.5        |  10 +++
>  tools/libxl/libxl.c          |  16 +++-
>  tools/libxl/libxl.h          |  12 +++
>  tools/libxl/libxl_create.c   | 181 +++++++++++++++++++++++++++++++++++++------
>  tools/libxl/libxl_internal.h |   4 +
>  tools/libxl/libxl_types.idl  |   3 +
>  tools/libxl/xl.h             |   1 +
>  tools/libxl/xl_cmdimpl.c     |  33 +++++++-
>  8 files changed, 232 insertions(+), 28 deletions(-)
> 
> diff --git a/docs/man/xl.cfg.pod.5 b/docs/man/xl.cfg.pod.5
> index a3e0e2e..27f4c2a 100644
> --- a/docs/man/xl.cfg.pod.5
> +++ b/docs/man/xl.cfg.pod.5
> @@ -368,6 +368,11 @@ destroy the domain.
>  write a "coredump" of the domain to F</var/lib/xen/dump/NAME> and then
>  restart the domain.
>  
> +=item B<soft-reset>
> +
> +cleanup the domain without destroying it, restart the device model. This action
> +is supported for HVM guests only.
> +
>  =back
>  
>  The default for C<on_poweroff> is C<destroy>.
> @@ -386,6 +391,11 @@ Default is C<destroy>.
>  
>  Action to take if the domain crashes.  Default is C<destroy>.
>  
> +=item B<on_soft_reset="ACTION">
> +
> +Action to take if the domain performs 'soft reset' (e.g. does kexec).
> +Default is C<soft-reset>.
> +
>  =back
>  
>  =head3 Direct Kernel Boot
> diff --git a/tools/libxl/libxl.c b/tools/libxl/libxl.c
> index d86ea62..a6743a7 100644
> --- a/tools/libxl/libxl.c
> +++ b/tools/libxl/libxl.c
> @@ -1492,6 +1492,7 @@ void libxl__domain_destroy(libxl__egc *egc, libxl__domain_destroy_state *dds)
>          dds->stubdom.ao = ao;
>          dds->stubdom.domid = stubdomid;
>          dds->stubdom.callback = stubdom_destroy_callback;
> +        dds->stubdom.soft_reset = false;
>          libxl__destroy_domid(egc, &dds->stubdom);
>      } else {
>          dds->stubdom_finished = 1;
> @@ -1500,6 +1501,7 @@ void libxl__domain_destroy(libxl__egc *egc, libxl__domain_destroy_state *dds)
>      dds->domain.ao = ao;
>      dds->domain.domid = dds->domid;
>      dds->domain.callback = domain_destroy_callback;
> +    dds->domain.soft_reset = dds->soft_reset;
>      libxl__destroy_domid(egc, &dds->domain);
>  }
>  
> @@ -1680,10 +1682,13 @@ static void devices_destroy_cb(libxl__egc *egc,
>  
>      /* Clean up qemu-save and qemu-resume files. They are
>       * intermediate files created by libxc. Unfortunately they
> -     * don't fit in existing userdata scheme very well.
> +     * don't fit in existing userdata scheme very well. In soft reset
> +     * case we need to keep the file.
>       */
> -    rc = libxl__remove_file(gc, libxl__device_model_savefile(gc, domid));
> -    if (rc < 0) goto out;
> +    if (!dis->soft_reset) {
> +        rc = libxl__remove_file(gc, libxl__device_model_savefile(gc, domid));
> +        if (rc < 0) goto out;
> +    }
>      rc = libxl__remove_file(gc,
>               GCSPRINTF(XC_DEVICE_MODEL_RESTORE_FILE".%u", domid));
>      if (rc < 0) goto out;
> @@ -1694,7 +1699,10 @@ static void devices_destroy_cb(libxl__egc *egc,
>          ctx->xch = xc_interface_open(ctx->lg,0,0);
>          if (!ctx->xch) goto badchild;
>  
> -        rc = xc_domain_destroy(ctx->xch, domid);
> +        if (!dis->soft_reset)
> +            rc = xc_domain_destroy(ctx->xch, domid);
> +        else

I would also pause the guest before doing this.

I didn't look at the rest very much, sorry to say.