On Wed, Dec 17, 2008 at 04:25:55PM +0000, Daniel P. Berrange wrote:
On Wed, Dec 17, 2008 at 04:19:47PM +0100, Daniel Veillard wrote:
Basically having the debug information in one place and the errors in another place is not very helpful, so this plugs the virterror handling into the logging routine. mostly a bit of refactoring, and calling virLogMessage. Just that we pass an extra flag to virLogMessage stating it comes from the error handling level to avoid outputting the error twice on stderr in case where there is no output defined.
Yep, makes sense.
@@ -392,6 +411,12 @@ }
/* + * Hook up the error or warning to the logging facility + */ + virLogMessage(virErrorDomainName(domain), virErrorLevelPriority(level), + 1, "%s", str); + + /* * Save the information about the error */ virResetError(to);
Hmm, but later on down the file we still call virDefaultErrorFunc() which would log the message a second time surely ?
To the application using libvirt, but not to the logs file. The potential annoyance is when you get both going to stderr for example if using virsh. Daniel -- Daniel Veillard | libxml Gnome XML XSLT toolkit http://xmlsoft.org/ daniel@veillard.com | Rpmfind RPM search engine http://rpmfind.net/ http://veillard.com/ | virtualization library http://libvirt.org/