All of lore.kernel.org
 help / color / mirror / Atom feed
From: Ian Campbell <ian.campbell@citrix.com>
To: Wei Liu <wei.liu2@citrix.com>
Cc: anthony.perard@citrix.com, stefano.stabellini@eu.citrix.com,
	ian.jackson@eu.citrix.com, xen-devel@lists.xen.org
Subject: Re: [PATCH 4/5] libxl: wait for stubdom to be ready
Date: Wed, 18 Mar 2015 13:35:15 +0000	[thread overview]
Message-ID: <1426685715.14291.36.camel@citrix.com> (raw)
In-Reply-To: <1426242875-24446-5-git-send-email-wei.liu2@citrix.com>

On Fri, 2015-03-13 at 10:34 +0000, Wei Liu wrote:
> Watch /local/domain/$dm_domid/device-model/$domid/state, wait until
> state turns "running" then unpause guest.
> 
> LIBXL_STUBDOM_START_TIMEOUT is the timeout used wait for stubdom to be
> ready. My test on a very old machine (Core 2 6400) showed that it might
> need more than 20s before the stubdom is ready to serve DomU.
> 
> Signed-off-by: Wei Liu <wei.liu2@citrix.com>
> ---
>  tools/libxl/libxl_dm.c       | 39 ++++++++++++++++++++++++++++++++++++++-
>  tools/libxl/libxl_internal.h |  2 ++
>  2 files changed, 40 insertions(+), 1 deletion(-)
> 
> diff --git a/tools/libxl/libxl_dm.c b/tools/libxl/libxl_dm.c
> index 4a38455..ad2ef41 100644
> --- a/tools/libxl/libxl_dm.c
> +++ b/tools/libxl/libxl_dm.c
> @@ -984,6 +984,8 @@ static void stubdom_pvqemu_cb(libxl__egc *egc,
>  static void spawn_stubdom_pvqemu_destroy_cb(libxl__egc *egc,
>                                              libxl__destroy_domid_state *dis,
>                                              int rc);
> +static void stubdom_xswait_cb(libxl__egc *egc, libxl__xswait_state *xswait,
> +                              int rc, const char *p);
>  
>  char *libxl__stub_dm_name(libxl__gc *gc, const char *guest_name)
>  {
> @@ -1273,16 +1275,51 @@ static void stubdom_pvqemu_cb(libxl__egc *egc,
>      rc = libxl_domain_unpause(CTX, dm_domid);
>      if (rc) goto out;
>  
> +    libxl__xswait_init(&sdss->xswait);
> +    sdss->xswait.ao = ao;
> +    sdss->xswait.what = GCSPRINTF("Stubdom %d startup", dm_domid);

Maybe include the domid too, e.g. "Stubdom %d for dom %d startup"?

(And probably %u throughout?)

> +static void stubdom_xswait_cb(libxl__egc *egc, libxl__xswait_state *xswait,
> +                              int rc, const char *p)
> +{
> +    EGC_GC;
> +    libxl__stub_dm_spawn_state *sdss = CONTAINER_OF(xswait, *sdss, xswait);
> +    uint32_t dm_domid = sdss->pvqemu.guest_domid;
> +
> +    if (rc) {
> +        if (rc == ERROR_TIMEDOUT)
> +            LOG(ERROR, "%s: startup timed out", xswait->what);
> +        goto out;
> +    }
> +
> +    if (!p) return;
> +
> +    if (strcmp(p, "running"))
> +        return;
>   out:
>      if (rc) {
>          if (dm_domid) {
> -            sdss->dis.ao = ao;
> +            sdss->dis.ao = sdss->dm.spawn.ao;
>              sdss->dis.domid = dm_domid;
>              sdss->dis.callback = spawn_stubdom_pvqemu_destroy_cb;
>              libxl__destroy_domid(egc, &sdss->dis);
> +            libxl__xswait_stop(gc, xswait);

Can you do this once befire the if (rc)?

Ian.

  parent reply	other threads:[~2015-03-18 13:35 UTC|newest]

Thread overview: 25+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-03-13 10:34 [PATCH 0/5] Fix QEMU startup protocol Wei Liu
2015-03-13 10:34 ` [PATCH 1/5] libxl: remove DM path in libxl__device_model_destroy Wei Liu
2015-03-18 13:21   ` Ian Campbell
2015-03-18 13:23     ` Ian Campbell
2015-03-13 10:34 ` [PATCH 2/5] libxl: use LIBXL_TOOLSTACK_DOMID Wei Liu
2015-03-18 13:21   ` Ian Campbell
2015-03-13 10:34 ` [PATCH 3/5] libxl: use new QEMU xenstore protocol Wei Liu
2015-03-18 13:30   ` Ian Campbell
2015-03-19 10:36     ` Wei Liu
2015-03-19 10:55       ` Ian Campbell
2015-03-19 11:04         ` Wei Liu
2015-03-13 10:34 ` [PATCH 4/5] libxl: wait for stubdom to be ready Wei Liu
2015-03-16 17:55   ` Wei Liu
2015-03-18 13:35   ` Ian Campbell [this message]
2015-03-19 10:46     ` Wei Liu
2015-03-13 10:34 ` [PATCH 5/5] Revert "x86/hvm: wait for at least one ioreq server to be enabled" Wei Liu
2015-03-13 11:32   ` Paul Durrant
2015-03-13 10:38 ` [PATCH] xenstore: use relative path for device-model node Wei Liu
2015-03-13 10:46   ` Ian Campbell
2015-03-13 11:01     ` Wei Liu
2015-03-13 11:04   ` Stefano Stabellini
2015-03-18 13:24   ` Ian Campbell
2015-03-19 10:54     ` Wei Liu
2015-03-19 11:04       ` Ian Campbell
2015-03-24 17:16   ` Ian Jackson

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1426685715.14291.36.camel@citrix.com \
    --to=ian.campbell@citrix.com \
    --cc=anthony.perard@citrix.com \
    --cc=ian.jackson@eu.citrix.com \
    --cc=stefano.stabellini@eu.citrix.com \
    --cc=wei.liu2@citrix.com \
    --cc=xen-devel@lists.xen.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.