>>> On 23.09.15 at 17:45, <roger....@citrix.com> wrote: > El 16/09/15 a les 12.05, Jan Beulich ha escrit: >>>>> On 04.09.15 at 14:08, <roger....@citrix.com> wrote: >> Also - aren't all the changes to this file (and perhaps othersfurther >> down) bug fixes in their own right? > > Whether they should be considered bugs or not it's hard to tell. There > was no code that executed this paths before with this configuration, and > probably nobody considered running HVM guests without an emulated lapic > a possibility, so I would argue that they are merely omissions.
Whether these were active or latent bugs doesn't really matter. What I'd prefer is for the code adjustments not directly related to the feature suppression you work on to be in their own patch, so that the two steps taken can be viewed as two steps. Particularly if it later turns out that one or more of those apparent latent bugs are found to be actively harming some special case, backporting that adjustment without the feature suppression parts would become a straightforward option. Jan _______________________________________________ Xen-devel mailing list Xen-devel@lists.xen.org http://lists.xen.org/xen-devel
