On Tue, 2019-02-19 at 17:01 +0000, Daniel P. Berrangé wrote:
On Tue, Feb 19, 2019 at 05:52:29PM +0100, Andrea Bolognani wrote: [...]
@@ -3231,6 +3231,7 @@ qemuDomainSaveMemory(virQEMUDriverPtr driver,
cleanup: VIR_FORCE_CLOSE(fd); + qemuFileWrapperFDClose(vm, wrapperFd);
Don't we need to check & propagate the return status of this, otherwise callers would mistakenly think qemuDomainSaveMemory has succeeeded, despite qemuFileWrapperFDClose having raised an error. Likewise all the other places below.
In cases where qemuFileWrapperFDClose() returning an error was not considered an overall failure in the existing code, I have preserved that behavior. Then again, we're ultimately going to call virReportError() from it instead of just logging a message with VIR_WARN(), so perhaps not returning an overall failure would be confusing... -- Andrea Bolognani / Red Hat / Virtualization