|
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] Re: [Xen-devel] [PATCH v4 31/31] libxl: allow the creation of HVM domains without a device model.
On Fri, Aug 07, 2015 at 12:18:08PM +0200, Roger Pau Monne wrote:
> Replace the firmware loaded into HVM guests with an OS kernel. Since the HVM
> builder now uses the PV xc_dom_* set of functions this kernel will be parsed
> and loaded inside the guest like on PV, but the container is a pure HVM
> guest.
>
> Also, if device_model_version is set to none or a device model for the
> specified domain is not present unconditinally set the nic type to
> LIBXL_NIC_TYPE_VIF.
>
> Signed-off-by: Roger Pau Monné <roger.pau@xxxxxxxxxx>
> Cc: Ian Jackson <ian.jackson@xxxxxxxxxxxxx>
> Cc: Stefano Stabellini <stefano.stabellini@xxxxxxxxxxxxx>
> Cc: Ian Campbell <ian.campbell@xxxxxxxxxx>
> Cc: Wei Liu <wei.liu2@xxxxxxxxxx>
> ---
> Changes since v3:
> - Add explicit /* fall through */ comments.
> - Expand libxl__device_nic_setdefault so that it sets the right nic type
> for HVMlite guests.
> - Remove stray space in hvm_build_set_params.
> - Fix the error paths of libxl__domain_firmware.
> ---
> docs/man/xl.cfg.pod.5 | 5 ++++
> tools/libxc/xc_dom_x86.c | 7 +++++
> tools/libxl/libxl.c | 39 ++++++++++++++-----------
> tools/libxl/libxl_create.c | 16 ++++++++++-
> tools/libxl/libxl_dm.c | 13 ++++++++-
> tools/libxl/libxl_dom.c | 68
> ++++++++++++++++++++++++++++++--------------
> tools/libxl/libxl_internal.h | 5 +++-
> tools/libxl/libxl_types.idl | 1 +
> tools/libxl/libxl_x86.c | 4 ++-
> tools/libxl/xl_cmdimpl.c | 2 ++
> 10 files changed, 118 insertions(+), 42 deletions(-)
>
> diff --git a/docs/man/xl.cfg.pod.5 b/docs/man/xl.cfg.pod.5
> index 80e51bb..8cd7726 100644
> --- a/docs/man/xl.cfg.pod.5
> +++ b/docs/man/xl.cfg.pod.5
> @@ -1741,6 +1741,11 @@ This device-model is the default for Linux dom0.
> Use the device-model based upon the historical Xen fork of Qemu.
> This device-model is still the default for NetBSD dom0.
>
> +=item B<none>
> +
> +Don't use any device model. This requires a kernel capable of booting
> +in this mode.
booting without emulated devices?
> +
> =back
>
> It is recommended to accept the default value for new guests. If
> diff --git a/tools/libxc/xc_dom_x86.c b/tools/libxc/xc_dom_x86.c
> index 1599de4..d67feb0 100644
> --- a/tools/libxc/xc_dom_x86.c
> +++ b/tools/libxc/xc_dom_x86.c
> @@ -1269,6 +1269,13 @@ static int meminit_hvm(struct xc_dom_image *dom)
> if ( nr_pages > target_pages )
> memflags |= XENMEMF_populate_on_demand;
>
> + /* Make sure there's a MMIO hole for the special pages. */
> + if ( dom->mmio_size == 0 )
> + {
> + dom->mmio_size = NR_SPECIAL_PAGES << PAGE_SHIFT;
> + dom->mmio_start = special_pfn(0);
> + }
> +
Better to just assert(dom->mmio_size != 0);
It's really libxl's responsibility to generate memory layout for guest.
Libxc doesn't have all information to make the decision.
> if ( dom->nr_vmemranges == 0 )
> {
> /* Build dummy vnode information
> diff --git a/tools/libxl/libxl.c b/tools/libxl/libxl.c
> index 083f099..a01868a 100644
> --- a/tools/libxl/libxl.c
> +++ b/tools/libxl/libxl.c
> @@ -1033,11 +1033,13 @@ int libxl_domain_unpause(libxl_ctx *ctx, uint32_t
> domid)
> }
>
> if (type == LIBXL_DOMAIN_TYPE_HVM) {
> - rc = libxl__domain_resume_device_model(gc, domid);
> - if (rc < 0) {
> - LOG(ERROR, "failed to unpause device model for domain %u:%d",
> - domid, rc);
> - goto out;
> + if (libxl__domain_has_device_model(gc, domid)) {
Checking for device model version is not enough?
> + rc = libxl__domain_resume_device_model(gc, domid);
> + if (rc < 0) {
> + LOG(ERROR, "failed to unpause device model for domain %u:%d",
> + domid, rc);
> + goto out;
> + }
> }
> }
> ret = xc_domain_unpause(ctx->xch, domid);
> @@ -1567,7 +1569,6 @@ void libxl__destroy_domid(libxl__egc *egc,
> libxl__destroy_domid_state *dis)
> libxl_ctx *ctx = CTX;
> uint32_t domid = dis->domid;
> char *dom_path;
> - char *pid;
> int rc, dm_present;
>
> libxl__ev_child_init(&dis->destroyer);
> @@ -1584,14 +1585,13 @@ void libxl__destroy_domid(libxl__egc *egc,
> libxl__destroy_domid_state *dis)
>
> switch (libxl__domain_type(gc, domid)) {
> case LIBXL_DOMAIN_TYPE_HVM:
> - if (!libxl_get_stubdom_id(CTX, domid))
> - dm_present = 1;
> - else
> + if (libxl_get_stubdom_id(CTX, domid)) {
> dm_present = 0;
> - break;
> + break;
> + }
> + /* fall through */
> case LIBXL_DOMAIN_TYPE_PV:
> - pid = libxl__xs_read(gc, XBT_NULL, libxl__sprintf(gc,
> "/local/domain/%d/image/device-model-pid", domid));
> - dm_present = (pid != NULL);
> + dm_present = libxl__domain_has_device_model(gc, domid);
> break;
> case LIBXL_DOMAIN_TYPE_INVALID:
> rc = ERROR_FAIL;
> @@ -3203,7 +3203,7 @@ out:
>
> /******************************************************************************/
>
> int libxl__device_nic_setdefault(libxl__gc *gc, libxl_device_nic *nic,
> - uint32_t domid)
> + uint32_t domid, libxl_domain_build_info
> *info)
> {
> int rc;
>
> @@ -3240,8 +3240,15 @@ int libxl__device_nic_setdefault(libxl__gc *gc,
> libxl_device_nic *nic,
>
> switch (libxl__domain_type(gc, domid)) {
> case LIBXL_DOMAIN_TYPE_HVM:
> - if (!nic->nictype)
> - nic->nictype = LIBXL_NIC_TYPE_VIF_IOEMU;
> + if (!nic->nictype) {
> + if (info != NULL &&
> + info->device_model_version !=
> LIBXL_DEVICE_MODEL_VERSION_NONE)
> + nic->nictype = LIBXL_NIC_TYPE_VIF_IOEMU;
> + else if (libxl__domain_has_device_model(gc, domid))
> + nic->nictype = LIBXL_NIC_TYPE_VIF_IOEMU;
> + else
> + nic->nictype = LIBXL_NIC_TYPE_VIF;
I think all you need is to pass in device model version and
if version != none
nictype = ioemu
else
nictype = vif
?
Otherwise the code suggests that there can be case you have specified a
device model when creating a domain but it somehow disappears when
domain is running?
Wei.
_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel
|
 |
Lists.xenproject.org is hosted with RackSpace, monitoring our |