[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Xen-devel] [PATCH] libxl/arm: fix guest type conversion



Hi Wei,

On 01/11/2018 08:59, Wei Liu wrote:
On Wed, Oct 31, 2018 at 08:16:52PM +0000, Julien Grall wrote:
Hi Wei,

On 10/31/18 2:25 PM, Wei Liu wrote:
Commit 359970fd8b ("tools/libxl: Switch Arm guest type to PVH") missed
changing the type field in c_info. This issue didn't surface until
ef72c93df9 which made creating PV guest on Arm unusable.

Signed-off-by: Wei Liu <wei.liu2@xxxxxxxxxx>
---
Cc: Stefano Stabellini <sstabellini@xxxxxxxxxx>
Cc: Julien Grall <julien.grall@xxxxxxx>

Julien, can you give this a quick test? Put type='pv' in your Arm
guest xl cfg and note the difference before and after this patch.

I tested on arm64 with and without the patch:

Tested-by: Julien Grall <julien.grall@xxxxxxx>


This should fix libvirt breakage.
---
   tools/libxl/libxl_create.c | 10 ++++++++++
   1 file changed, 10 insertions(+)

diff --git a/tools/libxl/libxl_create.c b/tools/libxl/libxl_create.c
index 4bb750e951..311957f87c 100644
--- a/tools/libxl/libxl_create.c
+++ b/tools/libxl/libxl_create.c
@@ -35,6 +35,16 @@ int libxl__domain_create_info_setdefault(libxl__gc *gc,
           return ERROR_INVAL;
       }
+#if defined(__arm__) || defined(__aarch64__)
+    if (c_info->type == LIBXL_DOMAIN_TYPE_PV) {
+        LOG(WARN, "Converting PV guest to PVH.");
+        LOG(WARN, "Arm guest are now PVH.");
+        LOG(WARN, "Please fix your configuration file/toolstack.");

We now end up to have this message printed twice.

I thought we need some logging so I copied them here.

I think we can remove this logging if we rely on c_info->type == b_info->type.


Do you see a use case
where c_info->type is different from b_info->type? Do we need a check for
that somewhere in the code?

No, I don't think there will be cases where c_info->type is different
from b_info->type. The type in b_info is used as a key to the keyed
union. The normal way of using it is to call
libxl_domain_build_info_init_type with c_info->type (see
xl_parse.c:L1316).  I tried to follow that path as well, but the
plumbing became unwieldy.

Regarding adding a check, that's probably a good idea, but it is out of
scope of this patch.

I wasn't suggested it for this patch :).

Cheers,

--
Julien Grall

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxxx
https://lists.xenproject.org/mailman/listinfo/xen-devel

 


Rackspace

Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.