On Thu, Nov 12, 2015 at 19:37:21 +0100, Jiri Denemark wrote:
Once qemuProcessInit was called, qemuProcessLaunch will launch a new QEMU process with stopped virtual CPUs.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com> --- src/qemu/qemu_process.c | 162 ++++++++++++++++++++++++++++++++---------------- src/qemu/qemu_process.h | 9 +++ 2 files changed, 118 insertions(+), 53 deletions(-)
[...]
@@ -5051,20 +5119,15 @@ int qemuProcessStart(virConnectPtr conn, VIR_DOMAIN_PAUSED_USER); }
- if (flags & VIR_QEMU_PROCESS_START_AUTODESTROY && - qemuProcessAutoDestroyAdd(driver, vm, conn) < 0) - goto error; - - VIR_DEBUG("Writing domain status to disk"); - if (virDomainSaveStatus(driver->xmlopt, cfg->stateDir, vm) < 0) - goto error; - - /* finally we can call the 'started' hook script if any */ if (qemuProcessStartHook(driver, vm, VIR_HOOK_QEMU_OP_STARTED, VIR_HOOK_SUBOP_BEGIN) < 0) goto error;
+ VIR_DEBUG("Writing domain status to disk"); + if (virDomainSaveStatus(driver->xmlopt, cfg->stateDir, vm) < 0) + goto error;
I think that the order of calling the VIR_HOOK_QEMU_OP_STARTED hook should not be changed with status XML saving since that may still jump to the path where the VM will be killed off.
+ /* Keep watching qemu log for errors during incoming migration, otherwise * unset reporting errors from qemu log. */ if (!incoming)
ACK with ^^ fixed.