| Message ID | 20250403214608.152954-6-jason.andryuk@amd.com (mailing list archive) |
|---|---|
| State | Superseded |
| Headers | show |
| Series | ARM split hardware and control domains | expand |
On Thu, 3 Apr 2025, Jason Andryuk wrote: > The hardware domain is unable to seed a control domain, but we want the > control domain to use xenstore. Rely on the hypervisor to seed dom0less > grant table entries for Xenstore, so this seeding is unnecessary. > > However, that only works for the new xenstore late init. The legacy > protocol which uses init-dom0less to populate the page still needs to > seed the grant. > > Signed-off-by: Jason Andryuk <jason.andryuk@amd.com> I was thinking whether it is an issue that we change behavior of init-dom0less for the new protocol. I convinced myself it is not an issue because init-dom0less is supposed to be a matched set with the Xen hypervisor version. I also thought whether we could check if the call to xc_dom_gnttab_seed is needed and only do it if it is, so that init-dom0less.c could be less depedent on the specific Xen behavior. Overall I thought it is not worth it, so: Reviewed-by: Stefano Stabellini <sstabellini@kernel.org> > --- > tools/helpers/init-dom0less.c | 10 +++++----- > 1 file changed, 5 insertions(+), 5 deletions(-) > > diff --git a/tools/helpers/init-dom0less.c b/tools/helpers/init-dom0less.c > index 17579fe2e8..91edf17b6c 100644 > --- a/tools/helpers/init-dom0less.c > +++ b/tools/helpers/init-dom0less.c > @@ -286,12 +286,12 @@ static int init_domain(struct xs_handle *xsh, > xenstore_pfn); > if (rc < 0) > return rc; > - } > > - rc = xc_dom_gnttab_seed(xch, info->domid, true, > - (xen_pfn_t)-1, xenstore_pfn, 0, 0); > - if (rc) > - err(1, "xc_dom_gnttab_seed"); > + rc = xc_dom_gnttab_seed(xch, info->domid, true, > + (xen_pfn_t)-1, xenstore_pfn, 0, 0); > + if (rc) > + err(1, "xc_dom_gnttab_seed"); > + } > > libxl_uuid_generate(&uuid); > xc_domain_sethandle(xch, info->domid, libxl_uuid_bytearray(&uuid)); > -- > 2.49.0 > >
diff --git a/tools/helpers/init-dom0less.c b/tools/helpers/init-dom0less.c index 17579fe2e8..91edf17b6c 100644 --- a/tools/helpers/init-dom0less.c +++ b/tools/helpers/init-dom0less.c @@ -286,12 +286,12 @@ static int init_domain(struct xs_handle *xsh, xenstore_pfn); if (rc < 0) return rc; - } - rc = xc_dom_gnttab_seed(xch, info->domid, true, - (xen_pfn_t)-1, xenstore_pfn, 0, 0); - if (rc) - err(1, "xc_dom_gnttab_seed"); + rc = xc_dom_gnttab_seed(xch, info->domid, true, + (xen_pfn_t)-1, xenstore_pfn, 0, 0); + if (rc) + err(1, "xc_dom_gnttab_seed"); + } libxl_uuid_generate(&uuid); xc_domain_sethandle(xch, info->domid, libxl_uuid_bytearray(&uuid));
The hardware domain is unable to seed a control domain, but we want the control domain to use xenstore. Rely on the hypervisor to seed dom0less grant table entries for Xenstore, so this seeding is unnecessary. However, that only works for the new xenstore late init. The legacy protocol which uses init-dom0less to populate the page still needs to seed the grant. Signed-off-by: Jason Andryuk <jason.andryuk@amd.com> --- tools/helpers/init-dom0less.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-)