On 9/20/18 1:44 PM, Marc Hartmayer wrote:
...although priv->qemuCaps will be NULL in almost every case when the post parse callback has failed. That may change in the future.
Signed-off-by: Marc Hartmayer <mhartmay@linux.ibm.com> Reviewed-by: Boris Fiuczynski <fiuczy@linux.ibm.com> --- src/qemu/qemu_process.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/src/qemu/qemu_process.c b/src/qemu/qemu_process.c index 44c63c42d618..6c5a6472d8cd 100644 --- a/src/qemu/qemu_process.c +++ b/src/qemu/qemu_process.c @@ -5282,7 +5282,7 @@ qemuProcessInit(virQEMUDriverPtr driver,
The comment just above here could use a tweak for grammar ;-): /* in case when the post parse callback failed we need to re-run it on the * old config prior we start the VM */
if (vm->def->postParseFailed) { VIR_DEBUG("re-running the post parse callback");
- if (virDomainDefPostParse(vm->def, caps, 0, driver->xmlopt, NULL) < 0) + if (virDomainDefPostParse(vm->def, caps, 0, driver->xmlopt, priv->qemuCaps) < 0)
Searching through history of this line finds Peter's original commit in this area - 7726d158, which seems to indicate a very specific reason for providing a NULL capabilities value here. I think from this patch on is something Peter has worked on a lot, so I would prefer to defer to Peter on them since I'm sure he understands all the various PostParse and Validate special conditions and various flags usage better than I do. I'll still provide a few comments along the way. John I did CC Peter just to bring his attention to the series...
goto cleanup; }