Re: [Xen-devel] [PATCH] libxl: do not start dom0 qemu for stubdomain when not needed
On Fri, Jul 28, 2017 at 05:17:12PM +0100, Ian Jackson wrote: > Wei Liu writes ("Re: [PATCH] libxl: do not start dom0 qemu for stubdomain > when not needed"): > > On Fri, Jul 28, 2017 at 06:05:13PM +0200, Marek Marczykowski-Górecki wrote: > > > On Fri, Jul 28, 2017 at 04:17:51PM +0100, Wei Liu wrote: > > > > On Thu, Jul 27, 2017 at 12:44:25AM +0200, Marek Marczykowski-Górecki > > > > wrote: > > > > > +if (libxl_defbool_val(guest_config->b_info.u.hvm.vnc.enable) || > > > > > + > > > > > libxl_defbool_val(guest_config->b_info.u.hvm.spice.enable) || > > > > > + > > > > > libxl_defbool_val(guest_config->b_info.u.hvm.sdl.enable)) { > > > > > > > > Indentation. > > > > > > Should it really be indented at the same level as the code inside this > > > block? Looks misleading (you need to look for ") {" to see where > > > condition ends). > > > > That's how existing code in libxl is like, isn't it? > > Yes. Sadly. > > If you like we could have a bunfight about changing the indent level. > I like 2, personally :-P. > > A compromise might be to move the || to the start of the next line, so > >if (cond1 >|| cond2 >|| cond3) { >code1; >code2; > > I have done that sometimes in libxl and no-one has objected. Ok, it would be much better, thanks. -- Best Regards, Marek Marczykowski-Górecki Invisible Things Lab A: Because it messes up the order in which people normally read text. Q: Why is top-posting such a bad thing? signature.asc Description: PGP signature ___ Xen-devel mailing list Xen-devel@lists.xen.org https://lists.xen.org/xen-devel
Re: [Xen-devel] [PATCH] libxl: do not start dom0 qemu for stubdomain when not needed
Wei Liu writes ("Re: [PATCH] libxl: do not start dom0 qemu for stubdomain when not needed"): > On Fri, Jul 28, 2017 at 06:05:13PM +0200, Marek Marczykowski-Górecki wrote: > > On Fri, Jul 28, 2017 at 04:17:51PM +0100, Wei Liu wrote: > > > On Thu, Jul 27, 2017 at 12:44:25AM +0200, Marek Marczykowski-Górecki > > > wrote: > > > > +if (libxl_defbool_val(guest_config->b_info.u.hvm.vnc.enable) || > > > > +libxl_defbool_val(guest_config->b_info.u.hvm.spice.enable) > > > > || > > > > +libxl_defbool_val(guest_config->b_info.u.hvm.sdl.enable)) { > > > > > > Indentation. > > > > Should it really be indented at the same level as the code inside this > > block? Looks misleading (you need to look for ") {" to see where > > condition ends). > > That's how existing code in libxl is like, isn't it? Yes. Sadly. If you like we could have a bunfight about changing the indent level. I like 2, personally :-P. A compromise might be to move the || to the start of the next line, so if (cond1 || cond2 || cond3) { code1; code2; I have done that sometimes in libxl and no-one has objected. Ian. ___ Xen-devel mailing list Xen-devel@lists.xen.org https://lists.xen.org/xen-devel
Re: [Xen-devel] [PATCH] libxl: do not start dom0 qemu for stubdomain when not needed
On Fri, Jul 28, 2017 at 06:05:13PM +0200, Marek Marczykowski-Górecki wrote: > On Fri, Jul 28, 2017 at 04:17:51PM +0100, Wei Liu wrote: > > On Thu, Jul 27, 2017 at 12:44:25AM +0200, Marek Marczykowski-Górecki wrote: > > > diff --git a/tools/libxl/libxl_dm.c b/tools/libxl/libxl_dm.c > > > index 44ebd70..c9aefe15 100644 > > > --- a/tools/libxl/libxl_dm.c > > > +++ b/tools/libxl/libxl_dm.c > > > @@ -1868,13 +1868,17 @@ void libxl__spawn_stub_dm(libxl__egc *egc, > > > libxl__stub_dm_spawn_state *sdss) > > > ret = libxl__domain_build_info_setdefault(gc, _config->b_info); > > > if (ret) goto out; > > > > > > -GCNEW(vfb); > > > -GCNEW(vkb); > > > -libxl__vfb_and_vkb_from_hvm_guest_config(gc, guest_config, vfb, vkb); > > > -dm_config->vfbs = vfb; > > > -dm_config->num_vfbs = 1; > > > -dm_config->vkbs = vkb; > > > -dm_config->num_vkbs = 1; > > > +if (libxl_defbool_val(guest_config->b_info.u.hvm.vnc.enable) || > > > +libxl_defbool_val(guest_config->b_info.u.hvm.spice.enable) || > > > +libxl_defbool_val(guest_config->b_info.u.hvm.sdl.enable)) { > > > > Indentation. > > Should it really be indented at the same level as the code inside this > block? Looks misleading (you need to look for ") {" to see where > condition ends). That's how existing code in libxl is like, isn't it? ___ Xen-devel mailing list Xen-devel@lists.xen.org https://lists.xen.org/xen-devel
Re: [Xen-devel] [PATCH] libxl: do not start dom0 qemu for stubdomain when not needed
On Fri, Jul 28, 2017 at 04:17:51PM +0100, Wei Liu wrote: > On Thu, Jul 27, 2017 at 12:44:25AM +0200, Marek Marczykowski-Górecki wrote: > > diff --git a/tools/libxl/libxl_dm.c b/tools/libxl/libxl_dm.c > > index 44ebd70..c9aefe15 100644 > > --- a/tools/libxl/libxl_dm.c > > +++ b/tools/libxl/libxl_dm.c > > @@ -1868,13 +1868,17 @@ void libxl__spawn_stub_dm(libxl__egc *egc, > > libxl__stub_dm_spawn_state *sdss) > > ret = libxl__domain_build_info_setdefault(gc, _config->b_info); > > if (ret) goto out; > > > > -GCNEW(vfb); > > -GCNEW(vkb); > > -libxl__vfb_and_vkb_from_hvm_guest_config(gc, guest_config, vfb, vkb); > > -dm_config->vfbs = vfb; > > -dm_config->num_vfbs = 1; > > -dm_config->vkbs = vkb; > > -dm_config->num_vkbs = 1; > > +if (libxl_defbool_val(guest_config->b_info.u.hvm.vnc.enable) || > > +libxl_defbool_val(guest_config->b_info.u.hvm.spice.enable) || > > +libxl_defbool_val(guest_config->b_info.u.hvm.sdl.enable)) { > > Indentation. Should it really be indented at the same level as the code inside this block? Looks misleading (you need to look for ") {" to see where condition ends). > > +GCNEW(vfb); > > +GCNEW(vkb); > > +libxl__vfb_and_vkb_from_hvm_guest_config(gc, guest_config, vfb, > > vkb); > > +dm_config->vfbs = vfb; > > +dm_config->num_vfbs = 1; > > +dm_config->vkbs = vkb; > > +dm_config->num_vkbs = 1; > > +} -- Best Regards, Marek Marczykowski-Górecki Invisible Things Lab A: Because it messes up the order in which people normally read text. Q: Why is top-posting such a bad thing? signature.asc Description: PGP signature ___ Xen-devel mailing list Xen-devel@lists.xen.org https://lists.xen.org/xen-devel
Re: [Xen-devel] [PATCH] libxl: do not start dom0 qemu for stubdomain when not needed
On Fri, Jul 28, 2017 at 05:34:42PM +0200, Marek Marczykowski-Górecki wrote: [...] > This patch itself does not break it (yet). Such guest will have all > consoles needed for migration by default. This also means that need_qemu > will always be true in practice. I imagine the next step would be either > adding some option to disable those consoles (maybe name it > "break_migration"? ;) ). Or adding support for multiple consoles to > xenconsoled. > The latter is preferably and a developer is already working on it (for a different reason). ___ Xen-devel mailing list Xen-devel@lists.xen.org https://lists.xen.org/xen-devel
Re: [Xen-devel] [PATCH] libxl: do not start dom0 qemu for stubdomain when not needed
On Fri, Jul 28, 2017 at 04:17:51PM +0100, Wei Liu wrote: > On Thu, Jul 27, 2017 at 12:44:25AM +0200, Marek Marczykowski-Górecki wrote: > > Use xen-blkback for 'vbd' disk types by default and do not setup vfb+vkb > > when no access method was configured. Then check if qemu is really > > needed. > > > > The only not configurable thing forcing qemu running in dom0 after this > > change are consoles used to save/restore. But even in that case, there > > is much smaller part of qemu exposed. > > > > Signed-off-by: Marek Marczykowski-Górecki> > --- > > tools/libxl/libxl_disk.c | 4 +++- > > tools/libxl/libxl_dm.c | 52 > > ++-- > > 2 files changed, 40 insertions(+), 16 deletions(-) > > > > diff --git a/tools/libxl/libxl_disk.c b/tools/libxl/libxl_disk.c > > index 63de75c..7842d9b 100644 > > --- a/tools/libxl/libxl_disk.c > > +++ b/tools/libxl/libxl_disk.c > > @@ -56,10 +56,12 @@ static void disk_eject_xswatch_callback(libxl__egc > > *egc, libxl__ev_xswatch *w, > > "/local/domain/%d/backend/%" TOSTRING(BACKEND_STRING_SIZE) > > "[a-z]/%*d/%*d", > > >backend_domid, backend_type); > > -if (!strcmp(backend_type, "tap") || !strcmp(backend_type, "vbd")) { > > +if (!strcmp(backend_type, "tap")) { > > disk->backend = LIBXL_DISK_BACKEND_TAP; > > } else if (!strcmp(backend_type, "qdisk")) { > > disk->backend = LIBXL_DISK_BACKEND_QDISK; > > +} else if (!strcmp(backend_type, "vbd")) { > > +disk->backend = LIBXL_DISK_BACKEND_PHY; > > This should be split into a separate patch. > > > } else { > > disk->backend = LIBXL_DISK_BACKEND_UNKNOWN; > > } > > diff --git a/tools/libxl/libxl_dm.c b/tools/libxl/libxl_dm.c > > index 44ebd70..c9aefe15 100644 > > --- a/tools/libxl/libxl_dm.c > > +++ b/tools/libxl/libxl_dm.c > > @@ -1868,13 +1868,17 @@ void libxl__spawn_stub_dm(libxl__egc *egc, > > libxl__stub_dm_spawn_state *sdss) > > ret = libxl__domain_build_info_setdefault(gc, _config->b_info); > > if (ret) goto out; > > > > -GCNEW(vfb); > > -GCNEW(vkb); > > -libxl__vfb_and_vkb_from_hvm_guest_config(gc, guest_config, vfb, vkb); > > -dm_config->vfbs = vfb; > > -dm_config->num_vfbs = 1; > > -dm_config->vkbs = vkb; > > -dm_config->num_vkbs = 1; > > +if (libxl_defbool_val(guest_config->b_info.u.hvm.vnc.enable) || > > +libxl_defbool_val(guest_config->b_info.u.hvm.spice.enable) || > > +libxl_defbool_val(guest_config->b_info.u.hvm.sdl.enable)) { > > Indentation. > > > +GCNEW(vfb); > > +GCNEW(vkb); > > +libxl__vfb_and_vkb_from_hvm_guest_config(gc, guest_config, vfb, > > vkb); > > +dm_config->vfbs = vfb; > > +dm_config->num_vfbs = 1; > > +dm_config->vkbs = vkb; > > +dm_config->num_vkbs = 1; > > +} > > > > stubdom_state->pv_kernel.path > > = libxl__abs_path(gc, "ioemu-stubdom.gz", > > libxl__xenfirmwaredir_path()); > > @@ -1959,6 +1963,7 @@ static void spawn_stub_launch_dm(libxl__egc *egc, > > libxl__domain_build_state *const d_state = sdss->dm.build_state; > > libxl__domain_build_state *const stubdom_state = >dm_state; > > uint32_t dm_domid = sdss->pvqemu.guest_domid; > > +int need_qemu; > > > > if (ret) { > > LOGD(ERROR, guest_domid, "error connecting disk devices"); > > @@ -1975,12 +1980,16 @@ static void spawn_stub_launch_dm(libxl__egc *egc, > > if (ret) > > goto out; > > } > > -ret = libxl__device_vfb_add(gc, dm_domid, _config->vfbs[0]); > > -if (ret) > > -goto out; > > -ret = libxl__device_vkb_add(gc, dm_domid, _config->vkbs[0]); > > -if (ret) > > -goto out; > > +if (dm_config->num_vfbs) { > > +ret = libxl__device_vfb_add(gc, dm_domid, _config->vfbs[0]); > > +if (ret) > > +goto out; > > +} > > +if (dm_config->num_vkbs) { > > +ret = libxl__device_vkb_add(gc, dm_domid, _config->vkbs[0]); > > +if (ret) > > +goto out; > > +} > > > > if (guest_config->b_info.u.hvm.serial) > > num_console++; > > @@ -1988,7 +1997,6 @@ static void spawn_stub_launch_dm(libxl__egc *egc, > > console = libxl__calloc(gc, num_console, > > sizeof(libxl__device_console)); > > > > for (i = 0; i < num_console; i++) { > > -libxl__device device; > > console[i].devid = i; > > console[i].consback = LIBXL__CONSOLE_BACKEND_IOEMU; > > /* STUBDOM_CONSOLE_LOGGING (console 0) is for minios logging > > @@ -2005,6 +2013,9 @@ static void spawn_stub_launch_dm(libxl__egc *egc, > > if (ret) goto out; > > console[i].output = GCSPRINTF("file:%s", filename); > > free(filename); > > +/* will be changed back to LIBXL__CONSOLE_BACKEND_IOEMU if > > qemu > > +
Re: [Xen-devel] [PATCH] libxl: do not start dom0 qemu for stubdomain when not needed
On Thu, Jul 27, 2017 at 12:44:25AM +0200, Marek Marczykowski-Górecki wrote: > Use xen-blkback for 'vbd' disk types by default and do not setup vfb+vkb > when no access method was configured. Then check if qemu is really > needed. > > The only not configurable thing forcing qemu running in dom0 after this > change are consoles used to save/restore. But even in that case, there > is much smaller part of qemu exposed. > > Signed-off-by: Marek Marczykowski-Górecki> --- > tools/libxl/libxl_disk.c | 4 +++- > tools/libxl/libxl_dm.c | 52 > ++-- > 2 files changed, 40 insertions(+), 16 deletions(-) > > diff --git a/tools/libxl/libxl_disk.c b/tools/libxl/libxl_disk.c > index 63de75c..7842d9b 100644 > --- a/tools/libxl/libxl_disk.c > +++ b/tools/libxl/libxl_disk.c > @@ -56,10 +56,12 @@ static void disk_eject_xswatch_callback(libxl__egc *egc, > libxl__ev_xswatch *w, > "/local/domain/%d/backend/%" TOSTRING(BACKEND_STRING_SIZE) > "[a-z]/%*d/%*d", > >backend_domid, backend_type); > -if (!strcmp(backend_type, "tap") || !strcmp(backend_type, "vbd")) { > +if (!strcmp(backend_type, "tap")) { > disk->backend = LIBXL_DISK_BACKEND_TAP; > } else if (!strcmp(backend_type, "qdisk")) { > disk->backend = LIBXL_DISK_BACKEND_QDISK; > +} else if (!strcmp(backend_type, "vbd")) { > +disk->backend = LIBXL_DISK_BACKEND_PHY; This should be split into a separate patch. > } else { > disk->backend = LIBXL_DISK_BACKEND_UNKNOWN; > } > diff --git a/tools/libxl/libxl_dm.c b/tools/libxl/libxl_dm.c > index 44ebd70..c9aefe15 100644 > --- a/tools/libxl/libxl_dm.c > +++ b/tools/libxl/libxl_dm.c > @@ -1868,13 +1868,17 @@ void libxl__spawn_stub_dm(libxl__egc *egc, > libxl__stub_dm_spawn_state *sdss) > ret = libxl__domain_build_info_setdefault(gc, _config->b_info); > if (ret) goto out; > > -GCNEW(vfb); > -GCNEW(vkb); > -libxl__vfb_and_vkb_from_hvm_guest_config(gc, guest_config, vfb, vkb); > -dm_config->vfbs = vfb; > -dm_config->num_vfbs = 1; > -dm_config->vkbs = vkb; > -dm_config->num_vkbs = 1; > +if (libxl_defbool_val(guest_config->b_info.u.hvm.vnc.enable) || > +libxl_defbool_val(guest_config->b_info.u.hvm.spice.enable) || > +libxl_defbool_val(guest_config->b_info.u.hvm.sdl.enable)) { Indentation. > +GCNEW(vfb); > +GCNEW(vkb); > +libxl__vfb_and_vkb_from_hvm_guest_config(gc, guest_config, vfb, vkb); > +dm_config->vfbs = vfb; > +dm_config->num_vfbs = 1; > +dm_config->vkbs = vkb; > +dm_config->num_vkbs = 1; > +} > > stubdom_state->pv_kernel.path > = libxl__abs_path(gc, "ioemu-stubdom.gz", > libxl__xenfirmwaredir_path()); > @@ -1959,6 +1963,7 @@ static void spawn_stub_launch_dm(libxl__egc *egc, > libxl__domain_build_state *const d_state = sdss->dm.build_state; > libxl__domain_build_state *const stubdom_state = >dm_state; > uint32_t dm_domid = sdss->pvqemu.guest_domid; > +int need_qemu; > > if (ret) { > LOGD(ERROR, guest_domid, "error connecting disk devices"); > @@ -1975,12 +1980,16 @@ static void spawn_stub_launch_dm(libxl__egc *egc, > if (ret) > goto out; > } > -ret = libxl__device_vfb_add(gc, dm_domid, _config->vfbs[0]); > -if (ret) > -goto out; > -ret = libxl__device_vkb_add(gc, dm_domid, _config->vkbs[0]); > -if (ret) > -goto out; > +if (dm_config->num_vfbs) { > +ret = libxl__device_vfb_add(gc, dm_domid, _config->vfbs[0]); > +if (ret) > +goto out; > +} > +if (dm_config->num_vkbs) { > +ret = libxl__device_vkb_add(gc, dm_domid, _config->vkbs[0]); > +if (ret) > +goto out; > +} > > if (guest_config->b_info.u.hvm.serial) > num_console++; > @@ -1988,7 +1997,6 @@ static void spawn_stub_launch_dm(libxl__egc *egc, > console = libxl__calloc(gc, num_console, sizeof(libxl__device_console)); > > for (i = 0; i < num_console; i++) { > -libxl__device device; > console[i].devid = i; > console[i].consback = LIBXL__CONSOLE_BACKEND_IOEMU; > /* STUBDOM_CONSOLE_LOGGING (console 0) is for minios logging > @@ -2005,6 +2013,9 @@ static void spawn_stub_launch_dm(libxl__egc *egc, > if (ret) goto out; > console[i].output = GCSPRINTF("file:%s", filename); > free(filename); > +/* will be changed back to LIBXL__CONSOLE_BACKEND_IOEMU if > qemu > + * will be in use */ > +console[i].consback = LIBXL__CONSOLE_BACKEND_XENCONSOLED; Where is the code to change it back? > break; > case STUBDOM_CONSOLE_SAVE: > console[i].output = GCSPRINTF("file:%s", > @@ -2019,6 +2030,12 @@ static void
[Xen-devel] [PATCH] libxl: do not start dom0 qemu for stubdomain when not needed
Use xen-blkback for 'vbd' disk types by default and do not setup vfb+vkb when no access method was configured. Then check if qemu is really needed. The only not configurable thing forcing qemu running in dom0 after this change are consoles used to save/restore. But even in that case, there is much smaller part of qemu exposed. Signed-off-by: Marek Marczykowski-Górecki--- tools/libxl/libxl_disk.c | 4 +++- tools/libxl/libxl_dm.c | 52 ++-- 2 files changed, 40 insertions(+), 16 deletions(-) diff --git a/tools/libxl/libxl_disk.c b/tools/libxl/libxl_disk.c index 63de75c..7842d9b 100644 --- a/tools/libxl/libxl_disk.c +++ b/tools/libxl/libxl_disk.c @@ -56,10 +56,12 @@ static void disk_eject_xswatch_callback(libxl__egc *egc, libxl__ev_xswatch *w, "/local/domain/%d/backend/%" TOSTRING(BACKEND_STRING_SIZE) "[a-z]/%*d/%*d", >backend_domid, backend_type); -if (!strcmp(backend_type, "tap") || !strcmp(backend_type, "vbd")) { +if (!strcmp(backend_type, "tap")) { disk->backend = LIBXL_DISK_BACKEND_TAP; } else if (!strcmp(backend_type, "qdisk")) { disk->backend = LIBXL_DISK_BACKEND_QDISK; +} else if (!strcmp(backend_type, "vbd")) { +disk->backend = LIBXL_DISK_BACKEND_PHY; } else { disk->backend = LIBXL_DISK_BACKEND_UNKNOWN; } diff --git a/tools/libxl/libxl_dm.c b/tools/libxl/libxl_dm.c index 44ebd70..c9aefe15 100644 --- a/tools/libxl/libxl_dm.c +++ b/tools/libxl/libxl_dm.c @@ -1868,13 +1868,17 @@ void libxl__spawn_stub_dm(libxl__egc *egc, libxl__stub_dm_spawn_state *sdss) ret = libxl__domain_build_info_setdefault(gc, _config->b_info); if (ret) goto out; -GCNEW(vfb); -GCNEW(vkb); -libxl__vfb_and_vkb_from_hvm_guest_config(gc, guest_config, vfb, vkb); -dm_config->vfbs = vfb; -dm_config->num_vfbs = 1; -dm_config->vkbs = vkb; -dm_config->num_vkbs = 1; +if (libxl_defbool_val(guest_config->b_info.u.hvm.vnc.enable) || +libxl_defbool_val(guest_config->b_info.u.hvm.spice.enable) || +libxl_defbool_val(guest_config->b_info.u.hvm.sdl.enable)) { +GCNEW(vfb); +GCNEW(vkb); +libxl__vfb_and_vkb_from_hvm_guest_config(gc, guest_config, vfb, vkb); +dm_config->vfbs = vfb; +dm_config->num_vfbs = 1; +dm_config->vkbs = vkb; +dm_config->num_vkbs = 1; +} stubdom_state->pv_kernel.path = libxl__abs_path(gc, "ioemu-stubdom.gz", libxl__xenfirmwaredir_path()); @@ -1959,6 +1963,7 @@ static void spawn_stub_launch_dm(libxl__egc *egc, libxl__domain_build_state *const d_state = sdss->dm.build_state; libxl__domain_build_state *const stubdom_state = >dm_state; uint32_t dm_domid = sdss->pvqemu.guest_domid; +int need_qemu; if (ret) { LOGD(ERROR, guest_domid, "error connecting disk devices"); @@ -1975,12 +1980,16 @@ static void spawn_stub_launch_dm(libxl__egc *egc, if (ret) goto out; } -ret = libxl__device_vfb_add(gc, dm_domid, _config->vfbs[0]); -if (ret) -goto out; -ret = libxl__device_vkb_add(gc, dm_domid, _config->vkbs[0]); -if (ret) -goto out; +if (dm_config->num_vfbs) { +ret = libxl__device_vfb_add(gc, dm_domid, _config->vfbs[0]); +if (ret) +goto out; +} +if (dm_config->num_vkbs) { +ret = libxl__device_vkb_add(gc, dm_domid, _config->vkbs[0]); +if (ret) +goto out; +} if (guest_config->b_info.u.hvm.serial) num_console++; @@ -1988,7 +1997,6 @@ static void spawn_stub_launch_dm(libxl__egc *egc, console = libxl__calloc(gc, num_console, sizeof(libxl__device_console)); for (i = 0; i < num_console; i++) { -libxl__device device; console[i].devid = i; console[i].consback = LIBXL__CONSOLE_BACKEND_IOEMU; /* STUBDOM_CONSOLE_LOGGING (console 0) is for minios logging @@ -2005,6 +2013,9 @@ static void spawn_stub_launch_dm(libxl__egc *egc, if (ret) goto out; console[i].output = GCSPRINTF("file:%s", filename); free(filename); +/* will be changed back to LIBXL__CONSOLE_BACKEND_IOEMU if qemu + * will be in use */ +console[i].consback = LIBXL__CONSOLE_BACKEND_XENCONSOLED; break; case STUBDOM_CONSOLE_SAVE: console[i].output = GCSPRINTF("file:%s", @@ -2019,6 +2030,12 @@ static void spawn_stub_launch_dm(libxl__egc *egc, console[i].output = "pty"; break; } +} + +need_qemu = libxl__need_xenpv_qemu(gc, dm_config); + +for (i = 0; i < num_console; i++) { +libxl__device device; ret = libxl__device_console_add(gc, dm_domid, [i], i == STUBDOM_CONSOLE_LOGGING ? stubdom_state : NULL,