[PATCHv1 0/5] tools: rework VM Generation ID

[PATCHv1 0/5] tools: rework VM Generation ID

[David Vrabel]

This series reworks the VM Generation ID to a) conform to the
published spec from Microsoft; b) simplify the save/restore code; and
c) extend the libxl API to allow toolstacks to use this feature.

The VM Generation ID must be regenerated with a new random ID after
certain VM operations. For xl, a new ID is required when restoring
from a saved image (but not after a migration, reboot, pause/unpause).

Can someone advise on the best way to implement this in xl?

We could put a vm-generation-id variable in the domain configuration
file.  This would end up in the saved configuration file and would be
reused or regenerated on restore depending on whether the domain was
migrated or restored from a file.

You can download the spec from:

  http://www.microsoft.com/en-us/download/details.aspx?id=30707

David

[PATCH 1/5] docs: update docs for the ~/platform/generation-id key

[David Vrabel]

Signed-off-by: David Vrabel <david.vrabel@citrix.com>
---
 docs/misc/xenstore-paths.markdown |   24 ++++++++++++++++--------
 1 file changed, 16 insertions(+), 8 deletions(-)

diff --git a/docs/misc/xenstore-paths.markdown b/docs/misc/xenstore-paths.markdown
index 70ab7f4..41d7d6d 100644
--- a/docs/misc/xenstore-paths.markdown
+++ b/docs/misc/xenstore-paths.markdown
@@ -166,11 +166,6 @@ use the xenstore-based protocol instead (see ~/control/shutdown,
 below) even if the guest has advertised support for the event channel
 protocol.
 
-#### ~/hvmloader/generation-id-address = ADDRESS [r,HVM,INTERNAL]
-
-The hexadecimal representation of the address of the domain's
-"generation id".
-
 #### ~/hvmloader/allow-memory-relocate = ("1"|"0") [HVM,INTERNAL]
 
 If the default low MMIO hole (below 4GiB) is not big enough for all
@@ -193,9 +188,22 @@ Various platform properties.
 
 #### ~/platform/generation-id = INTEGER ":" INTEGER [HVM,INTERNAL]
 
-Two 64 bit values that represent the Windows Generation ID.
-Is used by the BIOS initializer to get this value.
-If not present or "0:0" (all zeroes) device will not be present to the machine.
+The upper and lower 64-bit words of the 128-bit VM Generation ID.
+
+This key is used by hvmloader to create the ACPI VM Generation ID
+device.  It initialises a 16 octet region of guest memory with this
+value.  The guest physical address of this region is saved in the
+HVM_PARAM_VM_GENERATION_ID_ADDR HVM parameter.
+
+If this key is not present, is empty, or is all-zeros ("0:0") then the
+ACPI device is not created.
+
+The toolstack should, before unpausing a created or restored HVM
+domain, set this key and write the same ID to the guest memory
+location in HVM_PARAM_VM_GENERATION_ID_ADDR (if this address is
+non-zero).
+
+See Microsoft's "Virtual Machine Generation ID" specification.
 
 ### Frontend device paths
 
-- 
1.7.10.4

[PATCH 2/5] hvm: add HVM_PARAM_VM_GENERATION_ID_ADDR

[David Vrabel]

HVM_PARAM_VM_GENERATION_ID_ADDR is the guest physical address of the
VM Generation ID.  This parameter will be written by hvmloader and read
by the toolstack when updating the gen. ID (e.g., after restoring from a
snapshot).

A HVM parameter is easier for the save/restore code to work with (than
a XenStore key).

Signed-off-by: David Vrabel <david.vrabel@citrix.com>
---
 xen/include/public/hvm/params.h |    5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/xen/include/public/hvm/params.h b/xen/include/public/hvm/params.h
index 517a184..c972f9e 100644
--- a/xen/include/public/hvm/params.h
+++ b/xen/include/public/hvm/params.h
@@ -145,6 +145,9 @@
 /* SHUTDOWN_* action in case of a triple fault */
 #define HVM_PARAM_TRIPLE_FAULT_REASON 31
 
-#define HVM_NR_PARAMS          32
+/* Location of the VM Generation ID in guest address space. */
+#define HVM_PARAM_VM_GENERATION_ID_ADDR 32
+
+#define HVM_NR_PARAMS          33
 
 #endif /* __XEN_PUBLIC_HVM_PARAMS_H__ */
-- 
1.7.10.4

Re: [PATCH 2/5] hvm: add HVM_PARAM_VM_GENERATION_ID_ADDR

[Andrew Cooper]

On 21/05/14 14:21, David Vrabel wrote:
> HVM_PARAM_VM_GENERATION_ID_ADDR is the guest physical address of the
> VM Generation ID.  This parameter will be written by hvmloader and read
> by the toolstack when updating the gen. ID (e.g., after restoring from a
> snapshot).
>
> A HVM parameter is easier for the save/restore code to work with (than
> a XenStore key).
>
> Signed-off-by: David Vrabel <david.vrabel@citrix.com>
> ---
>  xen/include/public/hvm/params.h |    5 ++++-
>  1 file changed, 4 insertions(+), 1 deletion(-)
>
> diff --git a/xen/include/public/hvm/params.h b/xen/include/public/hvm/params.h
> index 517a184..c972f9e 100644
> --- a/xen/include/public/hvm/params.h
> +++ b/xen/include/public/hvm/params.h
> @@ -145,6 +145,9 @@
>  /* SHUTDOWN_* action in case of a triple fault */
>  #define HVM_PARAM_TRIPLE_FAULT_REASON 31
>  
> -#define HVM_NR_PARAMS          32
> +/* Location of the VM Generation ID in guest address space. */

Might be more clear to say "guest physical address space" ?

> +#define HVM_PARAM_VM_GENERATION_ID_ADDR 32
> +
> +#define HVM_NR_PARAMS          33
>  
>  #endif /* __XEN_PUBLIC_HVM_PARAMS_H__ */

[PATCH 3/5] tools/hvmloader: add helper functions to get/set HVM params

[David Vrabel]

Signed-off-by: David Vrabel <david.vrabel@citrix.com>
---
 tools/firmware/hvmloader/Makefile    |    1 +
 tools/firmware/hvmloader/hvm_param.c |   36 ++++++++++++++++++++++++++++++++++
 tools/firmware/hvmloader/hvmloader.c |   14 ++-----------
 tools/firmware/hvmloader/util.h      |    9 +++++++++
 tools/firmware/hvmloader/xenbus.c    |   14 +++++--------
 5 files changed, 53 insertions(+), 21 deletions(-)
 create mode 100644 tools/firmware/hvmloader/hvm_param.c

diff --git a/tools/firmware/hvmloader/Makefile b/tools/firmware/hvmloader/Makefile
index 00ee952..46a79c5 100644
--- a/tools/firmware/hvmloader/Makefile
+++ b/tools/firmware/hvmloader/Makefile
@@ -31,6 +31,7 @@ CFLAGS += $(CFLAGS_xeninclude)
 OBJS  = hvmloader.o mp_tables.o util.o smbios.o 
 OBJS += smp.o cacheattr.o xenbus.o
 OBJS += e820.o pci.o pir.o ctype.o
+OBJS += hvm_param.o
 ifeq ($(debug),y)
 OBJS += tests.o
 endif
diff --git a/tools/firmware/hvmloader/hvm_param.c b/tools/firmware/hvmloader/hvm_param.c
new file mode 100644
index 0000000..f7d8720
--- /dev/null
+++ b/tools/firmware/hvmloader/hvm_param.c
@@ -0,0 +1,36 @@
+/*
+ * hvm_param.c: get/set HVM params.
+ *
+ * Copyright (C) 2014 Citrix Systems R&D Ltd.
+ */
+#include "util.h"
+#include "config.h"
+#include "hypercall.h"
+
+#include <xen/hvm/params.h>
+
+int hvm_param_get(uint32_t index, uint64_t *value)
+{
+    struct xen_hvm_param p;
+    int ret;
+
+    p.domid = DOMID_SELF;
+    p.index = index;
+
+    ret = hypercall_hvm_op(HVMOP_get_param, &p);
+    if (ret == 0)
+        *value = p.value;
+
+    return ret;
+}
+
+int hvm_param_set(uint32_t index, uint64_t value)
+{
+    struct xen_hvm_param p;
+
+    p.domid = DOMID_SELF;
+    p.index = index;
+    p.value = value;
+
+    return hypercall_hvm_op(HVMOP_set_param, &p);
+}
diff --git a/tools/firmware/hvmloader/hvmloader.c b/tools/firmware/hvmloader/hvmloader.c
index 1cc8cf2..7b0da38 100644
--- a/tools/firmware/hvmloader/hvmloader.c
+++ b/tools/firmware/hvmloader/hvmloader.c
@@ -176,14 +176,10 @@ static void cmos_write_memory_size(void)
 static void init_vm86_tss(void)
 {
     void *tss;
-    struct xen_hvm_param p;
 
     tss = mem_alloc(128, 128);
     memset(tss, 0, 128);
-    p.domid = DOMID_SELF;
-    p.index = HVM_PARAM_VM86_TSS;
-    p.value = virt_to_phys(tss);
-    hypercall_hvm_op(HVMOP_set_param, &p);
+    hvm_param_set(HVM_PARAM_VM86_TSS, virt_to_phys(tss));
     printf("vm86 TSS at %08lx\n", virt_to_phys(tss));
 }
 
@@ -314,12 +310,6 @@ int main(void)
 
     if ( acpi_enabled )
     {
-        struct xen_hvm_param p = {
-            .domid = DOMID_SELF,
-            .index = HVM_PARAM_ACPI_IOPORTS_LOCATION,
-            .value = 1,
-        };
-
         if ( bios->acpi_build_tables )
         {
             printf("Loading ACPI ...\n");
@@ -328,7 +318,7 @@ int main(void)
 
         acpi_enable_sci();
 
-        hypercall_hvm_op(HVMOP_set_param, &p);
+        hvm_param_set(HVM_PARAM_ACPI_IOPORTS_LOCATION, 1);
     }
 
     init_vm86_tss();
diff --git a/tools/firmware/hvmloader/util.h b/tools/firmware/hvmloader/util.h
index 9ccb905..e4b6be4 100644
--- a/tools/firmware/hvmloader/util.h
+++ b/tools/firmware/hvmloader/util.h
@@ -210,6 +210,15 @@ const char *xenstore_read(const char *path, const char *default_resp);
  */
 int xenstore_write(const char *path, const char *value);
 
+
+/* Get a HVM param.
+ */
+int hvm_param_get(uint32_t index, uint64_t *value);
+
+/* Set a HVM param.
+ */
+int hvm_param_set(uint32_t index, uint64_t value);
+
 /* Setup PCI bus */
 void pci_setup(void);
 
diff --git a/tools/firmware/hvmloader/xenbus.c b/tools/firmware/hvmloader/xenbus.c
index fe72e97..64c2176 100644
--- a/tools/firmware/hvmloader/xenbus.c
+++ b/tools/firmware/hvmloader/xenbus.c
@@ -41,21 +41,17 @@ static char payload[XENSTORE_PAYLOAD_MAX + 1];  /* Unmarshalling area */
  * Call once, before any other xenbus actions. */
 void xenbus_setup(void)
 {
-    xen_hvm_param_t param;
+    uint64_t val;
 
     /* Ask Xen where the xenbus shared page is. */
-    param.domid = DOMID_SELF;
-    param.index = HVM_PARAM_STORE_PFN;
-    if ( hypercall_hvm_op(HVMOP_get_param, &param) )
+    if ( hvm_param_get(HVM_PARAM_STORE_PFN, &val) )
         BUG();
-    rings = (void *) (unsigned long) (param.value << PAGE_SHIFT);
+    rings = (void *) (unsigned long) (val << PAGE_SHIFT);
 
     /* Ask Xen where the xenbus event channel is. */
-    param.domid = DOMID_SELF;
-    param.index = HVM_PARAM_STORE_EVTCHN;
-    if ( hypercall_hvm_op(HVMOP_get_param, &param) )
+    if ( hvm_param_get(HVM_PARAM_STORE_EVTCHN, &val) )
         BUG();
-    event = param.value;
+    event = val;
 
     printf("Xenbus rings @0x%lx, event channel %lu\n",
            (unsigned long) rings, (unsigned long) event);
-- 
1.7.10.4

[PATCH 4/5] libxc, libxl, hvmloader: strip out outdated VM generation ID implementation

[David Vrabel]

The VM generation ID support in libxc/libxl was based on a draft
specification which subsequently changed considerably.  Remove much of
the code in anticipation of introducing something simpler that
conforms to the current specification from Microsoft.

Switch to using a HVM param (HVM_PARAM_VM_GENERATION_ID_ADDR) instead
of the hvmloader/generation-id-address XenStore key.  This simplifies
save/restore since it only needs to transfer the value of this param.

Signed-off-by: David Vrabel <david.vrabel@citrix.com>
---
 tools/firmware/hvmloader/acpi/build.c |    9 +++----
 tools/libxc/xc_domain_restore.c       |   44 +++------------------------------
 tools/libxc/xc_domain_save.c          |    5 ++--
 tools/libxc/xenguest.h                |    8 ++----
 tools/libxl/libxl_create.c            |   12 +++------
 tools/libxl/libxl_dom.c               |   25 ++-----------------
 tools/libxl/libxl_internal.h          |    8 ++----
 tools/libxl/libxl_save_callout.c      |   10 +++-----
 tools/libxl/libxl_save_helper.c       |    9 +++----
 tools/libxl/libxl_save_msgs_gen.pl    |    3 +--
 10 files changed, 26 insertions(+), 107 deletions(-)

diff --git a/tools/firmware/hvmloader/acpi/build.c b/tools/firmware/hvmloader/acpi/build.c
index f1dd3f0..edaafd7 100644
--- a/tools/firmware/hvmloader/acpi/build.c
+++ b/tools/firmware/hvmloader/acpi/build.c
@@ -24,6 +24,7 @@
 #include "../config.h"
 #include "../util.h"
 #include <xen/hvm/hvm_xs_strings.h>
+#include <xen/hvm/params.h>
 
 #define ACPI_MAX_SECONDARY_TABLES 16
 
@@ -361,7 +362,6 @@ static int construct_secondary_tables(unsigned long *table_ptrs,
 static int new_vm_gid(struct acpi_info *acpi_info)
 {
     uint64_t vm_gid[2], *buf;
-    char addr[12];
     const char * s;
     char *end;
 
@@ -382,12 +382,9 @@ static int new_vm_gid(struct acpi_info *acpi_info)
         return 0;
     memcpy(buf, vm_gid, sizeof(vm_gid));
 
-    /* set into ACPI table and XenStore the address */
+    /* set into ACPI table and HVM param the address */
     acpi_info->vm_gid_addr = virt_to_phys(buf);
-    if ( snprintf(addr, sizeof(addr), "0x%lx", virt_to_phys(buf))
-         >= sizeof(addr) )
-        return 0;
-    xenstore_write("hvmloader/generation-id-address", addr);
+    hvm_param_set(HVM_PARAM_VM_GENERATION_ID_ADDR, acpi_info->vm_gid_addr);
 
     return 1;
 }
diff --git a/tools/libxc/xc_domain_restore.c b/tools/libxc/xc_domain_restore.c
index bcb0ae0..cf1c489 100644
--- a/tools/libxc/xc_domain_restore.c
+++ b/tools/libxc/xc_domain_restore.c
@@ -1408,8 +1408,7 @@ int xc_domain_restore(xc_interface *xch, int io_fd, uint32_t dom,
                       domid_t store_domid, unsigned int console_evtchn,
                       unsigned long *console_mfn, domid_t console_domid,
                       unsigned int hvm, unsigned int pae, int superpages,
-                      int no_incr_generationid, int checkpointed_stream,
-                      unsigned long *vm_generationid_addr,
+                      int checkpointed_stream,
                       struct restore_callbacks *callbacks)
 {
     DECLARE_DOMCTL;
@@ -1634,44 +1633,9 @@ int xc_domain_restore(xc_interface *xch, int io_fd, uint32_t dom,
                 xc_set_hvm_param(xch, dom, HVM_PARAM_VM86_TSS, pagebuf.vm86_tss);
             if ( pagebuf.console_pfn )
                 console_pfn = pagebuf.console_pfn;
-            if ( pagebuf.vm_generationid_addr ) {
-                if ( !no_incr_generationid ) {
-                    unsigned int offset;
-                    unsigned char *buf;
-                    unsigned long long generationid;
-
-                    /*
-                     * Map the VM generation id buffer and inject the new value.
-                     */
-
-                    pfn = pagebuf.vm_generationid_addr >> PAGE_SHIFT;
-                    offset = pagebuf.vm_generationid_addr & (PAGE_SIZE - 1);
-                
-                    if ( (pfn >= dinfo->p2m_size) ||
-                         (pfn_type[pfn] != XEN_DOMCTL_PFINFO_NOTAB) )
-                    {
-                        ERROR("generation id buffer frame is bad");
-                        goto out;
-                    }
-
-                    mfn = ctx->p2m[pfn];
-                    buf = xc_map_foreign_range(xch, dom, PAGE_SIZE,
-                                               PROT_READ | PROT_WRITE, mfn);
-                    if ( buf == NULL )
-                    {
-                        ERROR("xc_map_foreign_range for generation id"
-                              " buffer failed");
-                        goto out;
-                    }
-
-                    generationid = *(unsigned long long *)(buf + offset);
-                    *(unsigned long long *)(buf + offset) = generationid + 1;
-
-                    munmap(buf, PAGE_SIZE);
-                }
-
-                *vm_generationid_addr = pagebuf.vm_generationid_addr;
-            }
+            if ( pagebuf.vm_generationid_addr )
+                xc_set_hvm_param(xch, dom, HVM_PARAM_VM_GENERATION_ID_ADDR,
+                                 pagebuf.vm_generationid_addr);
 
             break;  /* our work here is done */
         }
diff --git a/tools/libxc/xc_domain_save.c b/tools/libxc/xc_domain_save.c
index 71f9b59..42094e8 100644
--- a/tools/libxc/xc_domain_save.c
+++ b/tools/libxc/xc_domain_save.c
@@ -802,8 +802,7 @@ static int save_tsc_info(xc_interface *xch, uint32_t dom, int io_fd)
 
 int xc_domain_save(xc_interface *xch, int io_fd, uint32_t dom, uint32_t max_iters,
                    uint32_t max_factor, uint32_t flags,
-                   struct save_callbacks* callbacks, int hvm,
-                   unsigned long vm_generationid_addr)
+                   struct save_callbacks* callbacks, int hvm)
 {
     xc_dominfo_t info;
     DECLARE_DOMCTL;
@@ -1634,7 +1633,7 @@ int xc_domain_save(xc_interface *xch, int io_fd, uint32_t dom, uint32_t max_iter
         } chunk = { 0, };
 
         chunk.id = XC_SAVE_ID_HVM_GENERATION_ID_ADDR;
-        chunk.data = vm_generationid_addr;
+        xc_get_hvm_param(xch, dom, HVM_PARAM_VM_GENERATION_ID_ADDR, &chunk.data);
 
         if ( (chunk.data != 0) &&
              wrexact(io_fd, &chunk, sizeof(chunk)) )
diff --git a/tools/libxc/xenguest.h b/tools/libxc/xenguest.h
index 1f216cd..40bbac8 100644
--- a/tools/libxc/xenguest.h
+++ b/tools/libxc/xenguest.h
@@ -86,8 +86,7 @@ struct save_callbacks {
  */
 int xc_domain_save(xc_interface *xch, int io_fd, uint32_t dom, uint32_t max_iters,
                    uint32_t max_factor, uint32_t flags /* XCFLAGS_xxx */,
-                   struct save_callbacks* callbacks, int hvm,
-                   unsigned long vm_generationid_addr);
+                   struct save_callbacks* callbacks, int hvm);
 
 
 /* callbacks provided by xc_domain_restore */
@@ -113,9 +112,7 @@ struct restore_callbacks {
  * @parm hvm non-zero if this is a HVM restore
  * @parm pae non-zero if this HVM domain has PAE support enabled
  * @parm superpages non-zero to allocate guest memory with superpages
- * @parm no_incr_generationid non-zero if generation id is NOT to be incremented
  * @parm checkpointed_stream non-zero if the far end of the stream is using checkpointing
- * @parm vm_generationid_addr returned with the address of the generation id buffer
  * @parm callbacks non-NULL to receive a callback to restore toolstack
  *       specific data
  * @return 0 on success, -1 on failure
@@ -125,8 +122,7 @@ int xc_domain_restore(xc_interface *xch, int io_fd, uint32_t dom,
                       domid_t store_domid, unsigned int console_evtchn,
                       unsigned long *console_mfn, domid_t console_domid,
                       unsigned int hvm, unsigned int pae, int superpages,
-                      int no_incr_generationid, int checkpointed_stream,
-                      unsigned long *vm_generationid_addr,
+                      int checkpointed_stream,
                       struct restore_callbacks *callbacks);
 /**
  * xc_domain_restore writes a file to disk that contains the device
diff --git a/tools/libxl/libxl_create.c b/tools/libxl/libxl_create.c
index d015cf4..cae6f53 100644
--- a/tools/libxl/libxl_create.c
+++ b/tools/libxl/libxl_create.c
@@ -584,12 +584,7 @@ retry_transaction:
                         ARRAY_SIZE(rwperm));
     }
 
-    if (info->type == LIBXL_DOMAIN_TYPE_HVM)
-        libxl__xs_mkdir(gc, t,
-            libxl__sprintf(gc, "%s/hvmloader/generation-id-address", dom_path),
-                        rwperm, ARRAY_SIZE(rwperm));
-
-                    vm_list = libxl_list_vm(ctx, &nb_vm);
+    vm_list = libxl_list_vm(ctx, &nb_vm);
     if (!vm_list) {
         LOG(ERROR, "cannot get number of running guests");
         rc = ERROR_FAIL;
@@ -903,7 +898,7 @@ static void domcreate_bootloader_done(libxl__egc *egc,
         goto out;
     }
     libxl__xc_domain_restore(egc, dcs,
-                             hvm, pae, superpages, 1);
+                             hvm, pae, superpages);
     return;
 
  out:
@@ -911,7 +906,7 @@ static void domcreate_bootloader_done(libxl__egc *egc,
 }
 
 void libxl__srm_callout_callback_restore_results(unsigned long store_mfn,
-          unsigned long console_mfn, unsigned long genidad, void *user)
+          unsigned long console_mfn, void *user)
 {
     libxl__save_helper_state *shs = user;
     libxl__domain_create_state *dcs = CONTAINER_OF(shs, *dcs, shs);
@@ -920,7 +915,6 @@ void libxl__srm_callout_callback_restore_results(unsigned long store_mfn,
 
     state->store_mfn =            store_mfn;
     state->console_mfn =          console_mfn;
-    state->vm_generationid_addr = genidad;
     shs->need_results =           0;
 }
 
diff --git a/tools/libxl/libxl_dom.c b/tools/libxl/libxl_dom.c
index 661999c..4804fdf 100644
--- a/tools/libxl/libxl_dom.c
+++ b/tools/libxl/libxl_dom.c
@@ -279,7 +279,6 @@ int libxl__build_pre(libxl__gc *gc, uint32_t domid,
 
     state->store_port = xc_evtchn_alloc_unbound(ctx->xch, domid, state->store_domid);
     state->console_port = xc_evtchn_alloc_unbound(ctx->xch, domid, state->console_domid);
-    state->vm_generationid_addr = 0;
 
     if (info->type == LIBXL_DOMAIN_TYPE_HVM)
         hvm_set_conf_params(ctx->xch, domid, info);
@@ -297,7 +296,7 @@ int libxl__build_post(libxl__gc *gc, uint32_t domid,
     libxl_ctx *ctx = libxl__gc_owner(gc);
     char *dom_path, *vm_path;
     xs_transaction_t t;
-    char **ents, **hvm_ents;
+    char **ents;
     int i, rc;
 
     rc = libxl_domain_sched_params_set(CTX, domid, &info->sched_params);
@@ -334,13 +333,6 @@ int libxl__build_post(libxl__gc *gc, uint32_t domid,
                             ? "online" : "offline";
     }
 
-    hvm_ents = NULL;
-    if (info->type == LIBXL_DOMAIN_TYPE_HVM) {
-        hvm_ents = libxl__calloc(gc, 3, sizeof(char *));
-        hvm_ents[0] = "hvmloader/generation-id-address";
-        hvm_ents[1] = GCSPRINTF("0x%lx", state->vm_generationid_addr);
-    }
-
     dom_path = libxl__xs_get_dompath(gc, domid);
     if (!dom_path) {
         return ERROR_FAIL;
@@ -351,9 +343,6 @@ retry_transaction:
     t = xs_transaction_start(ctx->xsh);
 
     libxl__xs_writev(gc, t, dom_path, ents);
-    if (info->type == LIBXL_DOMAIN_TYPE_HVM)
-        libxl__xs_writev(gc, t, dom_path, hvm_ents);
-
     libxl__xs_writev(gc, t, dom_path, local_ents);
     libxl__xs_writev(gc, t, vm_path, vms_ents);
 
@@ -1502,7 +1491,6 @@ void libxl__domain_suspend(libxl__egc *egc, libxl__domain_suspend_state *dss)
     STATE_AO_GC(dss->ao);
     int port;
     int rc = ERROR_FAIL;
-    unsigned long vm_generationid_addr;
 
     /* Convenience aliases */
     const uint32_t domid = dss->domid;
@@ -1521,19 +1509,10 @@ void libxl__domain_suspend(libxl__egc *egc, libxl__domain_suspend_state *dss)
 
     switch (type) {
     case LIBXL_DOMAIN_TYPE_HVM: {
-        char *path;
-        char *addr;
-
-        path = GCSPRINTF("%s/hvmloader/generation-id-address",
-                              libxl__xs_get_dompath(gc, domid));
-        addr = libxl__xs_read(gc, XBT_NULL, path);
-
-        vm_generationid_addr = (addr) ? strtoul(addr, NULL, 0) : 0;
         dss->hvm = 1;
         break;
     }
     case LIBXL_DOMAIN_TYPE_PV:
-        vm_generationid_addr = 0;
         dss->hvm = 0;
         break;
     default:
@@ -1580,7 +1559,7 @@ void libxl__domain_suspend(libxl__egc *egc, libxl__domain_suspend_state *dss)
     callbacks->switch_qemu_logdirty = libxl__domain_suspend_common_switch_qemu_logdirty;
     dss->shs.callbacks.save.toolstack_save = libxl__toolstack_save;
 
-    libxl__xc_domain_save(egc, dss, vm_generationid_addr);
+    libxl__xc_domain_save(egc, dss);
     return;
 
  out:
diff --git a/tools/libxl/libxl_internal.h b/tools/libxl/libxl_internal.h
index c2b73c4..7ac38fe 100644
--- a/tools/libxl/libxl_internal.h
+++ b/tools/libxl/libxl_internal.h
@@ -960,8 +960,6 @@ typedef struct {
     uint32_t console_domid;
     unsigned long console_mfn;
 
-    unsigned long vm_generationid_addr;
-
     char *saved_state;
 
     libxl__file_reference pv_kernel;
@@ -2737,8 +2735,7 @@ _hidden void libxl__domain_suspend(libxl__egc *egc,
 
 
 /* calls libxl__xc_domain_suspend_done when done */
-_hidden void libxl__xc_domain_save(libxl__egc*, libxl__domain_suspend_state*,
-                                   unsigned long vm_generationid_addr);
+_hidden void libxl__xc_domain_save(libxl__egc*, libxl__domain_suspend_state*);
 /* If rc==0 then retval is the return value from xc_domain_save
  * and errnoval is the errno value it provided.
  * If rc!=0, retval and errnoval are undefined. */
@@ -2762,8 +2759,7 @@ _hidden int libxl__toolstack_save(uint32_t domid, uint8_t **buf,
 /* calls libxl__xc_domain_restore_done when done */
 _hidden void libxl__xc_domain_restore(libxl__egc *egc,
                                       libxl__domain_create_state *dcs,
-                                      int hvm, int pae, int superpages,
-                                      int no_incr_generationid);
+                                      int hvm, int pae, int superpages);
 /* If rc==0 then retval is the return value from xc_domain_save
  * and errnoval is the errno value it provided.
  * If rc!=0, retval and errnoval are undefined. */
diff --git a/tools/libxl/libxl_save_callout.c b/tools/libxl/libxl_save_callout.c
index e3bda8f..b37ab3b 100644
--- a/tools/libxl/libxl_save_callout.c
+++ b/tools/libxl/libxl_save_callout.c
@@ -41,8 +41,7 @@ static void helper_done(libxl__egc *egc, libxl__save_helper_state *shs);
 /*----- entrypoints -----*/
 
 void libxl__xc_domain_restore(libxl__egc *egc, libxl__domain_create_state *dcs,
-                              int hvm, int pae, int superpages,
-                              int no_incr_generationid)
+                              int hvm, int pae, int superpages)
 {
     STATE_AO_GC(dcs->ao);
 
@@ -59,7 +58,7 @@ void libxl__xc_domain_restore(libxl__egc *egc, libxl__domain_create_state *dcs,
         state->store_port,
         state->store_domid, state->console_port,
         state->console_domid,
-        hvm, pae, superpages, no_incr_generationid,
+        hvm, pae, superpages,
         cbflags, dcs->checkpointed_stream,
     };
 
@@ -75,8 +74,7 @@ void libxl__xc_domain_restore(libxl__egc *egc, libxl__domain_create_state *dcs,
                argnums, ARRAY_SIZE(argnums));
 }
 
-void libxl__xc_domain_save(libxl__egc *egc, libxl__domain_suspend_state *dss,
-                           unsigned long vm_generationid_addr)
+void libxl__xc_domain_save(libxl__egc *egc, libxl__domain_suspend_state *dss)
 {
     STATE_AO_GC(dss->ao);
     int r, rc, toolstack_data_fd = -1;
@@ -108,7 +106,7 @@ void libxl__xc_domain_save(libxl__egc *egc, libxl__domain_suspend_state *dss,
     }
 
     const unsigned long argnums[] = {
-        dss->domid, 0, 0, dss->xcflags, dss->hvm, vm_generationid_addr,
+        dss->domid, 0, 0, dss->xcflags, dss->hvm,
         toolstack_data_fd, toolstack_data_len,
         cbflags,
     };
diff --git a/tools/libxl/libxl_save_helper.c b/tools/libxl/libxl_save_helper.c
index c36314c..a94a8e9 100644
--- a/tools/libxl/libxl_save_helper.c
+++ b/tools/libxl/libxl_save_helper.c
@@ -212,7 +212,6 @@ int main(int argc, char **argv)
         uint32_t max_factor =      strtoul(NEXTARG,0,10);
         uint32_t flags =           strtoul(NEXTARG,0,10);
         int hvm =                  atoi(NEXTARG);
-        unsigned long genidad =    strtoul(NEXTARG,0,10);
         toolstack_save_fd  =       atoi(NEXTARG);
         toolstack_save_len =       strtoul(NEXTARG,0,10);
         unsigned cbflags =         strtoul(NEXTARG,0,10);
@@ -225,7 +224,7 @@ int main(int argc, char **argv)
 
         startup("save");
         r = xc_domain_save(xch, io_fd, dom, max_iters, max_factor, flags,
-                           &helper_save_callbacks, hvm, genidad);
+                           &helper_save_callbacks, hvm);
         complete(r);
 
     } else if (!strcmp(mode,"--restore-domain")) {
@@ -239,7 +238,6 @@ int main(int argc, char **argv)
         unsigned int hvm =         strtoul(NEXTARG,0,10);
         unsigned int pae =         strtoul(NEXTARG,0,10);
         int superpages =           strtoul(NEXTARG,0,10);
-        int no_incr_genidad =      strtoul(NEXTARG,0,10);
         unsigned cbflags =         strtoul(NEXTARG,0,10);
         int checkpointed =         strtoul(NEXTARG,0,10);
         assert(!*++argv);
@@ -248,15 +246,14 @@ int main(int argc, char **argv)
 
         unsigned long store_mfn = 0;
         unsigned long console_mfn = 0;
-        unsigned long genidad = 0;
 
         startup("restore");
         r = xc_domain_restore(xch, io_fd, dom, store_evtchn, &store_mfn,
                               store_domid, console_evtchn, &console_mfn,
                               console_domid, hvm, pae, superpages,
-                              no_incr_genidad, checkpointed, &genidad,
+                              checkpointed,
                               &helper_restore_callbacks);
-        helper_stub_restore_results(store_mfn,console_mfn,genidad,0);
+        helper_stub_restore_results(store_mfn,console_mfn,0);
         complete(r);
 
     } else {
diff --git a/tools/libxl/libxl_save_msgs_gen.pl b/tools/libxl/libxl_save_msgs_gen.pl
index 745e2ac..88f4921 100755
--- a/tools/libxl/libxl_save_msgs_gen.pl
+++ b/tools/libxl/libxl_save_msgs_gen.pl
@@ -32,8 +32,7 @@ our @msgs = (
     [  7, 'rcxW',   "toolstack_restore",     [qw(uint32_t domid
                                                 BLOCK tsdata)] ],
     [  8, 'r',      "restore_results",       ['unsigned long', 'store_mfn',
-                                              'unsigned long', 'console_mfn',
-                                              'unsigned long', 'genidad'] ],
+                                              'unsigned long', 'console_mfn'] ],
     [  9, 'srW',    "complete",              [qw(int retval
                                                  int errnoval)] ],
 );
-- 
1.7.10.4

[PATCH 5/5] libxl: add libxl_vm_generation_id_set()

[David Vrabel]

Add a function to allow a toolstack to set a domain's VM generation
ID.

Although the specification requires that a ACPI Notify event is raised
if the generation ID is changed, in practice this appears not to be
necessary if the generation ID is changed while the Windows VM is
suspended.

Signed-off-by: David Vrabel <david.vrabel@citrix.com>
---
 tools/libxl/Makefile         |    1 +
 tools/libxl/libxl.h          |    3 ++
 tools/libxl/libxl_vm_genid.c |   82 ++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 86 insertions(+)
 create mode 100644 tools/libxl/libxl_vm_genid.c

diff --git a/tools/libxl/Makefile b/tools/libxl/Makefile
index 4cfa275..5d0e2ec 100644
--- a/tools/libxl/Makefile
+++ b/tools/libxl/Makefile
@@ -77,6 +77,7 @@ LIBXL_OBJS = flexarray.o libxl.o libxl_create.o libxl_dm.o libxl_pci.o \
 			libxl_json.o libxl_aoutils.o libxl_numa.o \
 			libxl_save_callout.o _libxl_save_msgs_callout.o \
 			libxl_qmp.o libxl_event.o libxl_fork.o $(LIBXL_OBJS-y)
+LIBXL_OBJS += libxl_vm_genid.o
 LIBXL_OBJS += _libxl_types.o libxl_flask.o _libxl_types_internal.o
 
 LIBXL_TESTS += timedereg
diff --git a/tools/libxl/libxl.h b/tools/libxl/libxl.h
index 84f9c0e..dace880 100644
--- a/tools/libxl/libxl.h
+++ b/tools/libxl/libxl.h
@@ -1158,6 +1158,9 @@ int libxl_flask_getenforce(libxl_ctx *ctx);
 int libxl_flask_setenforce(libxl_ctx *ctx, int mode);
 int libxl_flask_loadpolicy(libxl_ctx *ctx, void *policy, uint32_t size);
 
+int libxl_vm_generation_id_set(libxl_ctx *ctx, uint32_t domid,
+                               const uint64_t *genid);
+
 /* misc */
 
 /* Each of these sets or clears the flag according to whether the
diff --git a/tools/libxl/libxl_vm_genid.c b/tools/libxl/libxl_vm_genid.c
new file mode 100644
index 0000000..0de24f8
--- /dev/null
+++ b/tools/libxl/libxl_vm_genid.c
@@ -0,0 +1,82 @@
+/*
+ * Copyright (C) 2014 Citrix Systems R&D Ltd.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU Lesser General Public License as published
+ * by the Free Software Foundation; version 2.1 only. with the special
+ * exception on linking described in file LICENSE.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU Lesser General Public License for more details.
+ */
+
+#include "libxl_osdeps.h" /* must come before any other headers */
+
+#include "libxl_internal.h"
+
+#include <xenctrl.h>
+#include <xen/hvm/params.h>
+
+/*
+ * Set a HVM domain's VM Generation ID.
+ *
+ * Toolstacks should call this after building or restoring the domain,
+ * but before unpausing it.
+ *
+ * For further details, refer to the "Virtual Machine Generation ID"
+ * document from Microsoft.
+ */
+int libxl_vm_generation_id_set(libxl_ctx *ctx, uint32_t domid,
+                               const uint64_t genid[2])
+{
+    GC_INIT(ctx);
+    char *dom_path;
+    char *key;
+    uint64_t paddr = 0;
+    int rc;
+
+    /*
+     * Set the "platform/generation-id" XenStore key to pass the ID to
+     * hvmloader.
+     */
+    dom_path = libxl__xs_get_dompath(gc, domid);
+    if (!dom_path) {
+        rc = ERROR_FAIL;
+        goto out;
+    }
+    key = libxl__sprintf(gc, "%s/platform/generation-id", dom_path);
+    rc = libxl__xs_write(gc, XBT_NULL, key, "%"PRIx64 ":%" PRIx64,
+                         genid[0], genid[1]);
+    if (rc < 0)
+        goto out;
+
+    /*
+     * Update the ID in guest memory (if available).
+     */
+    xc_get_hvm_param(ctx->xch, domid, HVM_PARAM_VM_GENERATION_ID_ADDR, &paddr);
+    if (paddr) {
+        void *vaddr;
+
+        vaddr = xc_map_foreign_range(ctx->xch, domid, XC_PAGE_SIZE,
+                                     PROT_READ | PROT_WRITE,
+                                     paddr >> XC_PAGE_SHIFT);
+        if (vaddr == NULL) {
+            rc = ERROR_FAIL;
+            goto out;
+        }
+        memcpy(vaddr + (paddr & ~XC_PAGE_MASK), genid, 2 * sizeof(*genid));
+        munmap(vaddr, XC_PAGE_SIZE);
+
+        /*
+         * FIXME: Inject ACPI Notify event.
+         */
+    }
+
+    rc = 0;
+
+  out:
+    GC_FREE;
+    return rc;
+}
-- 
1.7.10.4

Re: [PATCH 5/5] libxl: add libxl_vm_generation_id_set()

[Ian Campbell]

On Wed, 2014-05-21 at 14:21 +0100, David Vrabel wrote:
> Add a function to allow a toolstack to set a domain's VM generation
> ID.
> 
> Although the specification requires that a ACPI Notify event is raised
> if the generation ID is changed, in practice this appears not to be
> necessary if the generation ID is changed while the Windows VM is
> suspended.

Er, ok, that sounds suspicious though, can it be relied on for Windows
2015? Is generating the event so difficult? (I thought we had a libxl
function to do it...)

> 
> +/*
> + * Set a HVM domain's VM Generation ID.
> + *
> + * Toolstacks should call this after building or restoring the domain,
> + * but before unpausing it.

This would suggest that libxl_domain_config is the right place for it
then, and it should be handled within the library. But I'll await
completion of the discussion in 0/5 before committing 100% to that...

> + *
> + * For further details, refer to the "Virtual Machine Generation ID"
> + * document from Microsoft.

You linked to this in 0/5 -- perhaps include that in wherever this
comment ends up too?

> + */

Re: [PATCH 5/5] libxl: add libxl_vm_generation_id_set()

[Paul Durrant]

> -----Original Message-----
> From: xen-devel-bounces@lists.xen.org [mailto:xen-devel-
> bounces@lists.xen.org] On Behalf Of Ian Campbell
> Sent: 21 May 2014 17:10
> To: David Vrabel
> Cc: xen-devel@lists.xenproject.org; Andrew Cooper; Ian Jackson; Stefano
> Stabellini
> Subject: Re: [Xen-devel] [PATCH 5/5] libxl: add libxl_vm_generation_id_set()
> 
> On Wed, 2014-05-21 at 14:21 +0100, David Vrabel wrote:
> > Add a function to allow a toolstack to set a domain's VM generation
> > ID.
> >
> > Although the specification requires that a ACPI Notify event is raised
> > if the generation ID is changed, in practice this appears not to be
> > necessary if the generation ID is changed while the Windows VM is
> > suspended.
> 
> Er, ok, that sounds suspicious though, can it be relied on for Windows
> 2015? Is generating the event so difficult? (I thought we had a libxl
> function to do it...)
> 

I don't believe so. All that logic is still in QEMU AFAICT; it was part of what I was planning to pull into Xen with my PCI hotplug controller implementation.

  Paul

> >
> > +/*
> > + * Set a HVM domain's VM Generation ID.
> > + *
> > + * Toolstacks should call this after building or restoring the domain,
> > + * but before unpausing it.
> 
> This would suggest that libxl_domain_config is the right place for it
> then, and it should be handled within the library. But I'll await
> completion of the discussion in 0/5 before committing 100% to that...
> 
> > + *
> > + * For further details, refer to the "Virtual Machine Generation ID"
> > + * document from Microsoft.
> 
> You linked to this in 0/5 -- perhaps include that in wherever this
> comment ends up too?
> 
> > + */
> 
> 
> 
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@lists.xen.org
> http://lists.xen.org/xen-devel

Re: [PATCHv1 0/5] tools: rework VM Generation ID

[Ian Campbell]

On Wed, 2014-05-21 at 14:21 +0100, David Vrabel wrote:
> This series reworks the VM Generation ID to a) conform to the
> published spec from Microsoft; b) simplify the save/restore code; and
> c) extend the libxl API to allow toolstacks to use this feature.
> 
> The VM Generation ID must be regenerated with a new random ID after
> certain VM operations. For xl, a new ID is required when restoring
> from a saved image (but not after a migration, reboot, pause/unpause).

Does reboot include shutdown and recreate? i.e. if I do
 xl shutdown domain
 xl create domain.cfg

what should happen?

If we have to preserve then that is going to be problematic for xl,
since:
 xl create domain.cfg
 xl save domain saved
 xl restore saved # now the genid has changed
 xl shutdown domain
 xl cr domain.cfg

has nowhere to store the new genid, even if domain.cfg were to contain
the original one.

> Can someone advise on the best way to implement this in xl?

Would a new bool field in libxl_domain_restore_params to indicate
whether to preserve or regenerate the ID be sufficient? xl would set it
appropriately when restoring or on migrate receive.

Hrm, that doesn't work for reboot does it.

> We could put a vm-generation-id variable in the domain configuration
> file.  This would end up in the saved configuration file and would be
> reused or regenerated on restore depending on whether the domain was
> migrated or restored from a file.

Do you really mean the domain configuration file or did you just mean
libxl_domain_config? (which one I think you mean depends on the answer
to the shutdown/create question I think).

If libxl_domain_config then Wei's work to preserve domain config on
save/restore/migrate etc would then cause the right thing to happen.

Ian.

> 
> You can download the spec from:
> 
>   http://www.microsoft.com/en-us/download/details.aspx?id=30707
> 
> David
> 

Re: [PATCHv1 0/5] tools: rework VM Generation ID

[David Vrabel]

On 21/05/14 17:07, Ian Campbell wrote:
> On Wed, 2014-05-21 at 14:21 +0100, David Vrabel wrote:
>> This series reworks the VM Generation ID to a) conform to the
>> published spec from Microsoft; b) simplify the save/restore code; and
>> c) extend the libxl API to allow toolstacks to use this feature.
>>
>> The VM Generation ID must be regenerated with a new random ID after
>> certain VM operations. For xl, a new ID is required when restoring
>> from a saved image (but not after a migration, reboot, pause/unpause).
> 
> Does reboot include shutdown and recreate? i.e. if I do
>  xl shutdown domain
>  xl create domain.cfg
> 
> what should happen?

It may use the same ID.

> If we have to preserve then that is going to be problematic for xl,
> since:
>  xl create domain.cfg

ID = A

>  xl save domain saved
>  xl restore saved # now the genid has changed

ID = B

>  xl shutdown domain
>  xl cr domain.cfg

ID = B

> has nowhere to store the new genid, even if domain.cfg were to contain
> the original one.

Yes.

I think I might just get xl to generate a new ID on every domain create
and restore/migrate.  This will give the correct behaviour where it
matter, even if it is sub-optimal.

I doubt anyone is running a Windows 2012 domain controller with xl anyway.

David

Re: [PATCHv1 0/5] tools: rework VM Generation ID

[Ian Jackson]

David Vrabel writes ("[PATCHv1 0/5] tools: rework VM Generation ID"):
> This series reworks the VM Generation ID to a) conform to the
> published spec from Microsoft; b) simplify the save/restore code; and
> c) extend the libxl API to allow toolstacks to use this feature.
> 
> The VM Generation ID must be regenerated with a new random ID after
> certain VM operations. For xl, a new ID is required when restoring
> from a saved image (but not after a migration, reboot, pause/unpause).

At first blush, I think this specification is incoherent.

What is the difference between save/restore and migration ?  Or
between pause/unpause and save/restore ?  AFAICT the only high-level
semantic difference is whether the domain state is saved in a disk
file.  There are performance differences of course.

Are there even any guest-visible differences between migration and
save/restore ?

Ian.

Re: [PATCHv1 0/5] tools: rework VM Generation ID

[Ian Jackson]

David Vrabel writes ("[PATCHv1 0/5] tools: rework VM Generation ID"):
> You can download the spec from:
>   http://www.microsoft.com/en-us/download/details.aspx?id=30707

The spec is clearer about what's really needed.  What they want is
something to allow the guest to discover if it has been rewound.

The summary table which you have essentially reproduced in your
commentary should be read with that in mind.

Ian.