On Thu, May 20, 2010 at 10:12:35AM -0400, Chris Lalancette wrote: > On 05/20/2010 01:06 AM, Charles Duffy wrote: > > On 05/19/2010 11:54 PM, Charles Duffy wrote: > >> A revised patch is attached. This lifts its logic from its kernel > >> counterpart, and is updated only to permit forward slashes (which, while > >> disallowed for interface names with the kernel, are required for > >> *device* names -- for which the ESX driver happens to overload this > >> field. Ugh). > > Hm. You know, I'm wondering if we should have "isValidIfname" at all in > domain_conf.c. The thing is that if you think about different platforms > that this could be running on (Windows, Linux, Solaris, etc), and the > number of different drivers that go through here (Qemu, Xen, ESX, Virtualbox, > Phyp, openvz, ONE, UML), it seems like we have to allow everything to > cover all bases. Indeed, there is very little else in domain_conf.c that > does validity checking, probably for just this reason. Yep, if the check is platform/hypervisor specific, then it shouldn't be in the XML parsers. The parsers should only attempt to validate syntactic correctness, not semantic correctness (with exception of semantics that can be validated based on the capabilities data which accurately reflects the hypervisor specific rules in that case) Regards, Daniel -- |: Red Hat, Engineering, London -o- http://people.redhat.com/berrange/ :| |: http://libvirt.org -o- http://virt-manager.org -o- http://deltacloud.org :| |: http://autobuild.org -o- http://search.cpan.org/~danberr/ :| |: GnuPG: 7D3B9505 -o- F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 :| -- libvir-list mailing list libvir-list@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/libvir-list