# HG changeset patch
# User Ian Campbell <ian.campbell@xxxxxxxxxx>
# Date 1283519033 -3600
# Node ID 684cfeffdb1b4bacd736bc05ae26211cb91833df
# Parent de66ba50997cd8670fceffa7b01b50704f82edc7
libxc: provide notification of final checkpoint to restore end
When the restore code sees this notification it will restore the
currently in-progress checkpoint when it completes.
This allows the restore end to finish up without waiting for a
spurious timeout on the receive fd and thereby avoids unnecessary
error logging in the case of a successful migration or restore.
In the normal migration or restore case the first checkpoint is always
the last. For a rolling checkpoint (such as Remus) the notification is
currently unused but could be used in the future for example to
provide a controlled failover for reasons other than error
Signed-off-by: Ian Campbell <ian.campbell@xxxxxxxxxx>
diff -r de66ba50997c -r 684cfeffdb1b tools/libxc/xc_domain_restore.c
--- a/tools/libxc/xc_domain_restore.c Fri Sep 03 14:03:53 2010 +0100
+++ b/tools/libxc/xc_domain_restore.c Fri Sep 03 14:03:53 2010 +0100
@@ -42,6 +42,7 @@ struct restore_ctx {
xen_pfn_t *p2m; /* A table mapping each PFN to its new MFN. */
xen_pfn_t *p2m_batch; /* A table of P2M mappings in the current region. */
int completed; /* Set when a consistent image is available */
+ int last_checkpoint; /* Set when we should commit to the current
checkpoint when it completes. */
struct domain_info_context dinfo;
};
@@ -765,6 +766,11 @@ static int pagebuf_get_one(xc_interface
// DPRINTF("console pfn location: %llx\n", buf->console_pfn);
return pagebuf_get_one(xch, ctx, buf, fd, dom);
+ case XC_SAVE_ID_LAST_CHECKPOINT:
+ ctx->last_checkpoint = 1;
+ // DPRINTF("last checkpoint indication received");
+ return pagebuf_get_one(xch, ctx, buf, fd, dom);
+
default:
if ( (count > MAX_BATCH_SIZE) || (count < 0) ) {
ERROR("Max batch size exceeded (%d). Giving up.", count);
@@ -1296,10 +1302,23 @@ int xc_domain_restore(xc_interface *xch,
goto out;
}
ctx->completed = 1;
- /* shift into nonblocking mode for the remainder */
- if ( (flags = fcntl(io_fd, F_GETFL,0)) < 0 )
- flags = 0;
- fcntl(io_fd, F_SETFL, flags | O_NONBLOCK);
+
+ /*
+ * If more checkpoints are expected then shift into
+ * nonblocking mode for the remainder.
+ */
+ if ( !ctx->last_checkpoint )
+ {
+ if ( (flags = fcntl(io_fd, F_GETFL,0)) < 0 )
+ flags = 0;
+ fcntl(io_fd, F_SETFL, flags | O_NONBLOCK);
+ }
+ }
+
+ if ( ctx->last_checkpoint )
+ {
+ // DPRINTF("Last checkpoint, finishing\n");
+ goto finish;
}
// DPRINTF("Buffered checkpoint\n");
diff -r de66ba50997c -r 684cfeffdb1b tools/libxc/xc_domain_save.c
--- a/tools/libxc/xc_domain_save.c Fri Sep 03 14:03:53 2010 +0100
+++ b/tools/libxc/xc_domain_save.c Fri Sep 03 14:03:53 2010 +0100
@@ -1616,6 +1616,20 @@ int xc_domain_save(xc_interface *xch, in
}
}
+ if ( !callbacks->checkpoint )
+ {
+ /*
+ * If this is not a checkpointed save then this must be the first and
+ * last checkpoint.
+ */
+ i = XC_SAVE_ID_LAST_CHECKPOINT;
+ if ( wrexact(io_fd, &i, sizeof(int)) )
+ {
+ PERROR("Error when writing last checkpoint chunk");
+ goto out;
+ }
+ }
+
/* Zero terminate */
i = 0;
if ( wrexact(io_fd, &i, sizeof(int)) )
diff -r de66ba50997c -r 684cfeffdb1b tools/libxc/xg_save_restore.h
--- a/tools/libxc/xg_save_restore.h Fri Sep 03 14:03:53 2010 +0100
+++ b/tools/libxc/xg_save_restore.h Fri Sep 03 14:03:53 2010 +0100
@@ -131,6 +131,7 @@
#define XC_SAVE_ID_TMEM_EXTRA -6
#define XC_SAVE_ID_TSC_INFO -7
#define XC_SAVE_ID_HVM_CONSOLE_PFN -8 /* (HVM-only) */
+#define XC_SAVE_ID_LAST_CHECKPOINT -9 /* Commit to restoring after
completion of current iteration. */
/*
** We process save/restore/migrate in batches of pages; the below
_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-devel
|