On 10/09/2013 03:27 AM, Daniel P. Berrange wrote: > On Tue, Oct 08, 2013 at 11:24:57AM -0600, Eric Blake wrote: >> Mostly cleanup, but at least one bug fix (xen was instantiating >> a memballoon device too early, during the post-parse callback >> of the first device rather than after all devices were known). >> >> Spawned by the conversation here: >> https://www.redhat.com/archives/libvir-list/2013-September/thread.html#01311 > > If we're going down the route of removing 'Ptr' suffixes in favour > of normal '*' where we need to have const, how about just doing > away with this 'Ptr' suffix *everywhere* (well except public header > where we'd need to keep the typedef for back compat). Internally > we could completely remove it though. The 'Ptr' typedefs actually > may the code longer, for no obvious benefit, and create problems > with const-ness. > > The main problem is that the huge change would create pain for people > backporting patches. Indeed, which is why I limited things to JUST the case where the syntax checker proved we were using the wrong type. A bigger cleanup to remove Ptr may be entirely mechanical, but it does make backports interesting. -- Eric Blake eblake redhat com +1-919-301-3266 Libvirt virtualization library http://libvirt.org
Attachment:
signature.asc
Description: OpenPGP digital signature
-- libvir-list mailing list libvir-list@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/libvir-list