From: Ian Campbell <ian.campbell@citrix.com>
To: Wei Liu <wei.liu2@citrix.com>
Cc: wency@cn.fujitsu.com, andrew.cooper3@citrix.com,
yunhong.jiang@intel.com, eddie.dong@intel.com,
xen-devel@lists.xen.org, guijianfeng@cn.fujitsu.com,
rshriram@cs.ubc.ca, Yang Hongyang <yanghy@cn.fujitsu.com>,
Ian Jackson <Ian.Jackson@eu.citrix.com>
Subject: Re: [PATCH v4 --for 4.6 COLOPre 15/25] tools/libxl: check QEMU state before resume dm
Date: Wed, 15 Jul 2015 14:48:53 +0100 [thread overview]
Message-ID: <1436968133.32371.76.camel@citrix.com> (raw)
In-Reply-To: <20150715130056.GG29717@zion.uk.xensource.com>
On Wed, 2015-07-15 at 14:00 +0100, Wei Liu wrote:
> On Wed, Jul 15, 2015 at 01:54:12PM +0100, Ian Campbell wrote:
> > On Wed, 2015-07-15 at 13:48 +0100, Ian Campbell wrote:
> > > > switch (libxl__device_model_version_running(gc, domid)) {
> > > > case LIBXL_DEVICE_MODEL_VERSION_QEMU_XEN_TRADITIONAL: {
> > > > - libxl__qemu_traditional_cmd(gc, domid, "continue");
> > > > - libxl__wait_for_device_model_deprecated(gc, domid, "running", NULL, NULL, NULL);
> > > > + uint32_t dm_domid = libxl_get_stubdom_id(CTX, domid);
> > > > +
> > > > + path = libxl__device_model_xs_path(gc, dm_domid, domid, "/state");
> > > > + state = libxl__xs_read(gc, XBT_NULL, path);
> > > > + if (state != NULL && !strcmp(state, "paused")) {
> > > > + libxl__qemu_traditional_cmd(gc, domid, "continue");
> > >
> > > Please can you explain the apparent discrepancy between the use of
> > > dm_domid and domid here?
> >
> > I see from the next patch that this pattern came from the existing
> > libxl_domain_unpause, which hopes to use this helper in the future.
> >
> > Looking at git annotate:
> > 83cc69fa (Ian Jackson 2012-06-28 18:43:28 +0100 1045) if (type == LIBXL_DOMAIN_TYPE_HVM) {
> > 1fc3aeb3 ( Wei Liu 2015-04-09 19:49:25 +0100 1046) uint32_t dm_domid = libxl_get_stubdom_id(ctx, domid);
> > 1fc3aeb3 ( Wei Liu 2015-04-09 19:49:25 +0100 1047)
> > 1fc3aeb3 ( Wei Liu 2015-04-09 19:49:25 +0100 1048) path = libxl__device_model_xs_path(gc, dm_domid, domid, "/state");
> > bdf07e8e (Ian Jackson 2011-12-12 17:48:42 +0000 1049) state = libxl__xs_read(gc, XBT_NULL, path);
> > d1c7c3ef (Keir Fraser 2009-11-30 10:53:39 +0000 1050) if (state != NULL && !strcmp(state, "paused")) {
> > 0cb90b31 (Shriram Rajagopalan 2012-02-09 18:07:48 +0000 1051) libxl__qemu_traditional_cmd(gc, domid, "continue");
> > 47cb2273 (Ian Jackson 2013-10-14 17:26:01 +0100 1052) libxl__wait_for_device_model_deprecated(gc, domid, "running",
> > 3b6eaa3e (Ian Campbell 2011-05-24 15:57:24 +0100 1053) NULL, NULL, NULL);
> > d1c7c3ef (Keir Fraser 2009-11-30 10:53:39 +0000 1054) }
> >
> > It seems this came from Wei in 1fc3aeb3aa26 "libxl: use new QEMU
> > xenstore protocol". I suspect it was a mistake. Wei?
> >
>
> No, it's not.
>
> libxl__qemu_traditional_cmd accepts domid and then it calls
> libxl_get_stubdom_id to extract dm_domid.
How... exciting.
Some sort of helper to get the DM state would help to hide this sort of
wrinkle.
Anyway, I shall go see if this means I can ack the COLO patches which
moved this code.
Ian.
next prev parent reply other threads:[~2015-07-15 13:48 UTC|newest]
Thread overview: 101+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-07-15 7:45 [PATCH v4 --for 4.6 COLOPre 00/25] Prerequisite patches for COLO Yang Hongyang
2015-07-15 7:45 ` [PATCH v4 --for 4.6 COLOPre 01/25] tools/libxl: rename libxl__domain_suspend to libxl__domain_save Yang Hongyang
2015-07-15 11:16 ` Ian Campbell
2015-07-15 7:45 ` [PATCH v4 --for 4.6 COLOPre 02/25] tools/libxl: move domain suspend code into libxl_dom_suspend.c Yang Hongyang
2015-07-15 7:45 ` [PATCH v4 --for 4.6 COLOPre 03/25] tools/libxl: move domain resume " Yang Hongyang
2015-07-15 7:45 ` [PATCH v4 --for 4.6 COLOPre 04/25] tools/libxl: rename remus checkpoint callbacks Yang Hongyang
2015-07-15 11:17 ` Ian Campbell
2015-07-16 1:43 ` Yang Hongyang
2015-07-15 7:45 ` [PATCH v4 --for 4.6 COLOPre 05/25] libxl/remus: introduce libxl__remus_setup Yang Hongyang
2015-07-15 11:26 ` Ian Campbell
2015-07-16 5:32 ` Yang Hongyang
2015-07-16 10:40 ` Ian Campbell
2015-07-15 7:45 ` [PATCH v4 --for 4.6 COLOPre 06/25] libxl/remus: introduce libxl__remus_teardown Yang Hongyang
2015-07-15 11:59 ` Ian Campbell
2015-07-16 1:43 ` Yang Hongyang
2015-07-15 7:45 ` [PATCH v4 --for 4.6 COLOPre 07/25] libxl/remus: init checkpoint_callback in Remus checkpoint callback Yang Hongyang
2015-07-15 12:02 ` Ian Campbell
2015-07-15 12:35 ` Yang Hongyang
2015-07-16 10:32 ` Ian Campbell
2015-07-16 11:00 ` Yang Hongyang
2015-07-16 11:16 ` Ian Campbell
2015-07-15 7:45 ` [PATCH v4 --for 4.6 COLOPre 08/25] tools/libxl: move remus code into libxl_remus.c Yang Hongyang
2015-07-15 12:05 ` Ian Campbell
2015-07-15 7:45 ` [PATCH v4 --for 4.6 COLOPre 09/25] tools/libxl: move save/restore code into libxl_dom_save.c Yang Hongyang
2015-07-15 7:45 ` [PATCH v4 --for 4.6 COLOPre 10/25] libxl/save: Refactor libxl__domain_suspend_state Yang Hongyang
2015-07-15 12:10 ` Ian Campbell
2015-07-15 7:45 ` [PATCH v4 --for 4.6 COLOPre 11/25] tools/libxc: support to resume uncooperative HVM guests Yang Hongyang
2015-07-15 12:26 ` Ian Campbell
2015-07-16 5:57 ` Yang Hongyang
2015-07-16 15:40 ` Ian Jackson
2015-07-16 16:15 ` Yang Hongyang
2015-07-16 16:27 ` Ian Jackson
2015-12-15 2:05 ` Wen Congyang
2016-01-04 16:33 ` Ian Jackson
2015-07-15 7:45 ` [PATCH v4 --for 4.6 COLOPre 12/25] tools/libxl: introduce enum type libxl_checkpointed_stream Yang Hongyang
2015-07-15 12:34 ` Ian Campbell
2015-07-15 13:58 ` Yang Hongyang
2015-07-16 10:34 ` Ian Campbell
2015-07-16 10:47 ` Yang Hongyang
2015-07-15 7:45 ` [PATCH v4 --for 4.6 COLOPre 13/25] migration/save: pass checkpointed_stream from libxl to libxc Yang Hongyang
2015-07-15 12:38 ` Ian Campbell
2015-07-16 6:05 ` Yang Hongyang
2015-07-16 10:47 ` Ian Campbell
2015-07-16 16:13 ` Wei Liu
2015-07-16 16:21 ` Yang Hongyang
2015-07-16 16:39 ` Wei Liu
2015-07-16 16:10 ` Wei Liu
2015-07-16 16:24 ` Yang Hongyang
2015-07-16 16:37 ` Wei Liu
2015-07-15 7:45 ` [PATCH v4 --for 4.6 COLOPre 14/25] tools/libxl: introduce libxl__domain_restore_device_model to load qemu state Yang Hongyang
2015-07-15 12:45 ` Ian Campbell
2015-07-15 13:42 ` Yang Hongyang
2015-07-15 7:45 ` [PATCH v4 --for 4.6 COLOPre 15/25] tools/libxl: check QEMU state before resume dm Yang Hongyang
2015-07-15 12:48 ` Ian Campbell
2015-07-15 12:54 ` Ian Campbell
2015-07-15 13:00 ` Wei Liu
2015-07-15 13:48 ` Ian Campbell [this message]
2015-07-15 13:49 ` Ian Campbell
2015-07-16 14:43 ` Wei Liu
2015-07-16 15:43 ` Yang Hongyang
2015-07-15 7:45 ` [PATCH v4 --for 4.6 COLOPre 16/25] tools/libxl: Update libxl_domain_unpause() to support qemu-xen Yang Hongyang
2015-07-15 12:50 ` Ian Campbell
2015-07-16 3:49 ` Yang Hongyang
2015-07-16 10:39 ` Ian Campbell
2015-07-16 10:51 ` Yang Hongyang
2015-07-16 16:26 ` Wei Liu
2015-07-15 7:45 ` [PATCH v4 --for 4.6 COLOPre 17/25] tools/libxl: introduce libxl__domain_common_switch_qemu_logdirty() Yang Hongyang
2015-07-15 7:45 ` [PATCH v4 --for 4.6 COLOPre 18/25] tools/libxl: export logdirty_init Yang Hongyang
2015-07-15 7:45 ` [PATCH v4 --for 4.6 COLOPre 19/25] tools/libxl: Add back channel to allow migration target send data back Yang Hongyang
2015-07-15 7:45 ` [PATCH v4 --for 4.6 COLOPre 20/25] tools/libx{l, c}: add back channel to libxc Yang Hongyang
2015-07-15 13:13 ` Ian Campbell
2015-07-16 6:29 ` Yang Hongyang
2015-07-16 11:01 ` Ian Campbell
2015-07-15 13:21 ` Andrew Cooper
2015-07-16 6:07 ` Yang Hongyang
2015-07-15 7:45 ` [PATCH v4 --for 4.6 COLOPre 21/25] tools/libxl: rename remus device to checkpoint device Yang Hongyang
2015-07-15 13:15 ` Ian Campbell
2015-07-15 13:34 ` Yang Hongyang
2015-07-16 9:26 ` Andrew Cooper
2015-07-16 9:29 ` Yang Hongyang
2015-07-15 13:32 ` Ian Campbell
2015-07-15 13:38 ` Yang Hongyang
2015-07-16 9:23 ` Yang Hongyang
2015-07-16 9:31 ` Ian Campbell
2015-07-16 9:36 ` Yang Hongyang
2015-07-16 10:14 ` Ian Campbell
2015-07-16 10:22 ` Yang Hongyang
2015-07-15 7:45 ` [PATCH v4 --for 4.6 COLOPre 22/25] tools/libxl: adjust the indentation Yang Hongyang
2015-07-15 7:45 ` [PATCH v4 --for 4.6 COLOPre 23/25] tools/libxl: store remus_ops in checkpoint device state Yang Hongyang
2015-07-15 13:21 ` Ian Campbell
2015-07-15 7:45 ` [PATCH v4 --for 4.6 COLOPre 24/25] tools/libxl: move remus state into a seperate structure Yang Hongyang
2015-07-15 13:28 ` Ian Campbell
2015-07-15 13:50 ` Yang Hongyang
2015-07-16 10:37 ` Ian Campbell
2015-07-16 11:10 ` Ian Jackson
2015-07-16 11:19 ` Ian Campbell
2015-07-15 15:08 ` Ian Jackson
2015-07-15 15:18 ` Yang Hongyang
2015-07-15 7:45 ` [PATCH v4 --for 4.6 COLOPre 25/25] tools/libxl: seperate device init/cleanup from checkpoint device layer Yang Hongyang
2015-07-15 13:37 ` Ian Campbell
2015-07-16 1:37 ` [PATCH v4 --for 4.6 COLOPre 00/25] Prerequisite patches for COLO Yang Hongyang
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=1436968133.32371.76.camel@citrix.com \
--to=ian.campbell@citrix.com \
--cc=Ian.Jackson@eu.citrix.com \
--cc=andrew.cooper3@citrix.com \
--cc=eddie.dong@intel.com \
--cc=guijianfeng@cn.fujitsu.com \
--cc=rshriram@cs.ubc.ca \
--cc=wei.liu2@citrix.com \
--cc=wency@cn.fujitsu.com \
--cc=xen-devel@lists.xen.org \
--cc=yanghy@cn.fujitsu.com \
--cc=yunhong.jiang@intel.com \
/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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).