> -----Original Message----- > From: Ian Jackson [mailto:ian.jack...@eu.citrix.com] > Sent: 12 December 2016 18:04 > To: Paul Durrant <paul.durr...@citrix.com> > Cc: Konrad Rzeszutek Wilk <konrad.w...@oracle.com>; Zhang Chen > <zhangchen.f...@cn.fujitsu.com>; Changlong Xie > <xiecl.f...@cn.fujitsu.com>; Wei Liu <wei.l...@citrix.com>; Eddie Dong > <eddie.d...@intel.com>; Andrew Cooper <andrew.coop...@citrix.com>; > Ian Jackson <ian.jack...@citrix.com>; Wen Congyang > <wencongy...@gmail.com>; Yang Hongyang <imhy.y...@gmail.com>; Xen > devel <xen-de...@lists.xenproject.org> > Subject: RE: [Xen-devel] [PATCH 1/3] Don't create default ioreq server > > Paul Durrant writes ("RE: [Xen-devel] [PATCH 1/3] Don't create default ioreq > server"): > > The read side effects are indeed because of the need to support the > > old qemu interface. If trad were patched then we could at least > > deprecate the default ioreq server but I'm not sure how long we'd > > need to leave it in place after that before it was removed. > > (I see further discussion has obviated the need to deal with this, > but:) > > You mean to patch trad to make a few more hypercalls to explicitly > create an ioreq server ? Would it have to register all of its memory > regions ? If it wouldn't, then the change to qemu-trad wouldn't be > too hard. >
No, it would have to explicitly register all IO regions too... and understand new-fangled PCI config space ioreqs. Paul > qemu-trad is fairly closely tied to the Xen releases so we wouldn't > have to support the old way for very long. > > Ian. _______________________________________________ Xen-devel mailing list Xen-devel@lists.xen.org https://lists.xen.org/xen-devel
