merge with xen-unstable.hg

extras/mini-os/Makefile

@@ -123,3 +123,7 @@ cscope:
 tags:
 	$(all_sources) | xargs ctags
 
+.PHONY: TAGS
+TAGS:
+	$(all_sources) | xargs etags
+

extras/mini-os/arch/x86/Makefile

@@ -3,6 +3,9 @@
 # It's is used for x86_32, x86_32y and x86_64
 #
 
+XEN_ROOT = ../../../..
+include $(XEN_ROOT)/Config.mk
+
 # include arch.mk has to be before mini-os.mk!
 
 include arch.mk

extras/mini-os/arch/x86/mm.c

@@ -405,9 +405,9 @@ void *map_frames(unsigned long *f, unsigned long n)
 
 void arch_init_p2m(unsigned long max_pfn)
 {
-#define L1_P2M_SHIFT    10
-#define L2_P2M_SHIFT    20    
-#define L3_P2M_SHIFT    30    
+#define L1_P2M_SHIFT    9
+#define L2_P2M_SHIFT    18    
+#define L3_P2M_SHIFT    27    
 #define L1_P2M_ENTRIES  (1 << L1_P2M_SHIFT)    
 #define L2_P2M_ENTRIES  (1 << (L2_P2M_SHIFT - L1_P2M_SHIFT))    
 #define L3_P2M_ENTRIES  (1 << (L3_P2M_SHIFT - L2_P2M_SHIFT))    

tools/firmware/etherboot/README

@@ -6,3 +6,6 @@ and for stricter DHCP spec compliance.
 Rom-o-matic.net will provide this image at the following URL:
 
 http://rom-o-matic.net/5.4.3/build.php?version=5.4.3&F=ignore&nic=rtl8139%3Artl8139+--+%5B0x10ec%2C0x8139%5D&ofmt=Binary+ROM+Image%28.zrom%29&arch=i386&ASK_BOOT=-1&BOOT_FIRST=BOOT_NIC&BOOT_SECOND=BOOT_NOTHING&BOOT_THIRD=BOOT_NOTHING&BOOT_INDEX=0&STATIC_CLIENT_IP=&STATIC_SUBNET_MASK=&STATIC_SERVER_IP=&STATIC_GATEWAY_IP=&STATIC_BOOTFILE=&EXIT_ON_FILE_LOAD_ERROR=on&DHCP_CLIENT_ID=&DHCP_CLIENT_ID_LEN=&DHCP_CLIENT_ID_TYPE=&DHCP_USER_CLASS=&DHCP_USER_CLASS_LEN=&ALLOW_ONLY_ENCAPSULATED=on&DEFAULT_BOOTFILE=&CONGESTED=on&BACKOFF_LIMIT=7&TIMEOUT=180&TRY_FLOPPY_FIRST=0&EXIT_IF_NO_OFFER=on&TAGGED_IMAGE=on&ELF_IMAGE=on&PXE_IMAGE=on&DOWNLOAD_PROTO_TFTP=on&COMCONSOLE=0x3F8&CONSPEED=9600&COMPARM=0x03&PXE_EXPORT=on&CONFIG_PCI=on&CONFIG_ISA=on&BUILD_ID=&PCBIOS=on&PXE_DHCP_STRICT=on&A=Get+ROM
+
+and the mkhex script in tools/firmware/hvmloader will make the header 
+file from the downloaded image.

tools/firmware/hvmloader/Makefile

@@ -51,12 +51,12 @@ hvmloader: roms.h acpi/acpi.a $(SRCS)
 acpi/acpi.a:
 	$(MAKE) -C acpi
 
-roms.h:	../rombios/BIOS-bochs-latest ../vgabios/VGABIOS-lgpl-latest.bin ../vgabios/VGABIOS-lgpl-latest.cirrus.bin ../vmxassist/vmxassist.bin ../etherboot/eb-rtl8139.zrom
+roms.h:	../rombios/BIOS-bochs-latest ../vgabios/VGABIOS-lgpl-latest.bin ../vgabios/VGABIOS-lgpl-latest.cirrus.bin ../vmxassist/vmxassist.bin ../etherboot/eb-rtl8139.zrom.h
 	sh ./mkhex rombios ../rombios/BIOS-bochs-latest > roms.h
 	sh ./mkhex vgabios_stdvga ../vgabios/VGABIOS-lgpl-latest.bin >> roms.h
 	sh ./mkhex vgabios_cirrusvga ../vgabios/VGABIOS-lgpl-latest.cirrus.bin >> roms.h
 	sh ./mkhex vmxassist ../vmxassist/vmxassist.bin >> roms.h
-	sh ./mkhex etherboot ../etherboot/eb-rtl8139.zrom >> roms.h
+	cat ../etherboot/eb-rtl8139.zrom.h >> roms.h
 
 .PHONY: clean
 clean:

tools/libxc/xc_dom_x86.c

@@ -417,6 +417,7 @@ static int start_info_x86_32(struct xc_dom_image *dom)
 
     xc_dom_printf("%s: called\n", __FUNCTION__);
 
+    memset(start_info, 0, sizeof(*start_info));
     sprintf(start_info->magic, dom->guest_type);
     start_info->nr_pages = dom->total_pages;
     start_info->shared_info = shinfo << PAGE_SHIFT_X86;
@@ -455,6 +456,7 @@ static int start_info_x86_64(struct xc_dom_image *dom)
 
     xc_dom_printf("%s: called\n", __FUNCTION__);
 
+    memset(start_info, 0, sizeof(*start_info));
     sprintf(start_info->magic, dom->guest_type);
     start_info->nr_pages = dom->total_pages;
     start_info->shared_info = shinfo << PAGE_SHIFT_X86;

tools/pygrub/src/LiloConf.py

@@ -18,12 +18,13 @@ class LiloImage(object):
                 "  initrd: %s\n" %(self.title, self.root, self.kernel,
                                    self.args, self.initrd))
     def reset(self, lines, path):
-        self._root = self._initrd = self._kernel = self._args = None
+        self._initrd = self._kernel = self._readonly = None
+        self._args = ""
         self.title = ""
         self.lines = []
         self.path = path
+        self.root = ""
         map(self.set_from_line, lines)
-        self.root = "" # dummy
 
     def set_from_line(self, line, replace = None):
         (com, arg) = GrubConf.grub_exact_split(line, 2)
@@ -55,6 +56,23 @@ class LiloImage(object):
         return self._initrd
     initrd = property(get_initrd, set_initrd)
 
+    def set_args(self, val):
+        self._args = val
+    def get_args(self):
+        args = self._args
+        if self.root:
+            args += " root=" + self.root
+        if self.readonly:
+            args += " ro"
+        return args
+    args = property(get_args, set_args)
+
+    def set_readonly(self, val):
+        self._readonly = 1
+    def get_readonly(self):
+        return self._readonly
+    readonly = property(get_readonly, set_readonly)
+
     # set up command handlers
     commands = { "label": "self.title",
                  "root": "self.root",
@@ -62,7 +80,7 @@ class LiloImage(object):
                  "image": "self.kernel",
                  "initrd": "self.initrd",
                  "append": "self.args",
-                 "read-only": None,
+                 "read-only": "self.readonly",
                  "chainloader": None,
                  "module": None}
 

tools/python/xen/util/acmpolicy.py

@@ -47,6 +47,9 @@ ACM_POLICY_UNDEFINED = 15
 
 ACM_SCHEMA_FILE = "/etc/xen/acm-security/policies/security_policy.xsd"
 
+ACM_LABEL_UNLABELED = "__UNLABELED__"
+ACM_LABEL_UNLABELED_DISPLAY = "unlabeled"
+
 class ACMPolicy(XSPolicy):
     """
      ACMPolicy class. Implements methods for getting information from
@@ -139,6 +142,21 @@ class ACMPolicy(XSPolicy):
         return xsconstants.XSERR_SUCCESS
 
 
+    def is_default_policy(self):
+        """
+           Determine whether this is the default policy
+        """
+        default = ['SystemManagement']
+        if self.policy_get_virtualmachinelabel_names() == default and \
+           self.policy_get_bootstrap_vmlabel() == default[0] and \
+           self.policy_get_stetypes_types() == default and \
+           self.policy_get_stes_of_vmlabel(default[0]) == default and \
+           self.policy_get_resourcelabel_names() == [] and \
+           self.policy_get_chwall_types() == default and \
+           self.get_name() == "DEFAULT":
+            return True
+        return False
+
     def update(self, xml_new):
         """
             Update the policy with the new XML. The hypervisor decides
@@ -153,27 +171,18 @@ class ACMPolicy(XSPolicy):
             return -xsconstants.XSERR_XML_PROCESSING, errors
 
         vmlabel_map = acmpol_new.policy_get_vmlabel_translation_map()
+
         # An update requires version information in the current
         # and new policy. The version number of the current policy
         # must be the same as what is in the FromPolicy/Version node
         # in the new one and the current policy's name must be the
         # same as in FromPolicy/PolicyName
-
-        now_vers    = acmpol_old.policy_dom_get_hdr_item("Version")
-        now_name    = acmpol_old.policy_dom_get_hdr_item("PolicyName")
-        req_oldvers = acmpol_new.policy_dom_get_frompol_item("Version")
-        req_oldname = acmpol_new.policy_dom_get_frompol_item("PolicyName")
-
-        if now_vers == "" or \
-           now_vers != req_oldvers or \
-           now_name != req_oldname:
-            log.info("Policy rejected: %s != %s or %s != %s" % \
-                     (now_vers,req_oldvers,now_name,req_oldname))
-            return -xsconstants.XSERR_VERSION_PREVENTS_UPDATE, errors
-
-        if not self.isVersionUpdate(acmpol_new):
-            log.info("Policy rejected since new version is not an update.")
-            return -xsconstants.XSERR_VERSION_PREVENTS_UPDATE, errors
+        # The default policy when it is set skips this step.
+        if not acmpol_new.is_default_policy() and \
+           not acmpol_old.is_default_policy():
+            irc = self.__do_update_version_check(acmpol_new)
+            if irc != xsconstants.XSERR_SUCCESS:
+                return irc, errors
 
         if self.isloaded():
             newvmnames = \
@@ -255,6 +264,29 @@ class ACMPolicy(XSPolicy):
             self.compile()
         return rc, errors
 
+
+    def __do_update_version_check(self, acmpol_new):
+        acmpol_old = self
+
+        now_vers    = acmpol_old.policy_dom_get_hdr_item("Version")
+        now_name    = acmpol_old.policy_dom_get_hdr_item("PolicyName")
+        req_oldvers = acmpol_new.policy_dom_get_frompol_item("Version")
+        req_oldname = acmpol_new.policy_dom_get_frompol_item("PolicyName")
+
+        if now_vers == "" or \
+           now_vers != req_oldvers or \
+           now_name != req_oldname:
+            log.info("Policy rejected: %s != %s or %s != %s" % \
+                     (now_vers,req_oldvers,now_name,req_oldname))
+            return -xsconstants.XSERR_VERSION_PREVENTS_UPDATE
+
+        if not self.isVersionUpdate(acmpol_new):
+            log.info("Policy rejected since new version is not an update.")
+            return -xsconstants.XSERR_VERSION_PREVENTS_UPDATE
+
+        return xsconstants.XSERR_SUCCESS
+
+
     def compareVersions(self, v1, v2):
         """
             Compare two policy versions given their tuples of major and
@@ -845,8 +877,7 @@ class ACMPolicy(XSPolicy):
         if self.isloaded():
             return -xsconstants.XSERR_POLICY_LOADED
         files = [ self.get_filename(".map",""),
-                  self.get_filename(".bin",""),
-                  self.path_from_policy_name(self.get_name())]
+                  self.get_filename(".bin","") ]
         for f in files:
             try:
                 os.unlink(f)
@@ -925,11 +956,13 @@ class ACMPolicy(XSPolicy):
             return -xsconstants.XSERR_POLICY_INCONSISTENT, "", ""
 
         vms_with_chws = []
-        chws_by_vm = {}
+        chws_by_vm = { ACM_LABEL_UNLABELED : [] }
         for v in vms:
             if v.has_key("chws"):
                 vms_with_chws.append(v["name"])
                 chws_by_vm[v["name"]] = v["chws"]
+
+
         if bootstrap in vms_with_chws:
             vms_with_chws.remove(bootstrap)
             vms_with_chws.sort()
@@ -937,12 +970,16 @@ class ACMPolicy(XSPolicy):
         else:
             vms_with_chws.sort()
 
+        if ACM_LABEL_UNLABELED in vms_with_chws:
+            vms_with_chws.remove(ACM_LABEL_UNLABELED) ; # @1
+
         vms_with_stes = []
-        stes_by_vm = {}
+        stes_by_vm = { ACM_LABEL_UNLABELED : [] }
         for v in vms:
             if v.has_key("stes"):
                 vms_with_stes.append(v["name"])
                 stes_by_vm[v["name"]] = v["stes"]
+
         if bootstrap in vms_with_stes:
             vms_with_stes.remove(bootstrap)
             vms_with_stes.sort()
@@ -950,6 +987,9 @@ class ACMPolicy(XSPolicy):
         else:
             vms_with_stes.sort()
 
+        if ACM_LABEL_UNLABELED in vms_with_stes:
+            vms_with_stes.remove(ACM_LABEL_UNLABELED) ; # @2
+
         resnames = self.policy_get_resourcelabel_names()
         resnames.sort()
         stes_by_res = {}
@@ -958,6 +998,9 @@ class ACMPolicy(XSPolicy):
             if r.has_key("stes"):
                 stes_by_res[r["name"]] = r["stes"]
 
+        if ACM_LABEL_UNLABELED in resnames:
+            resnames.remove(ACM_LABEL_UNLABELED)
+
         max_chw_ssids = 1 + len(vms_with_chws)
         max_chw_types = 1 + len(vms_with_chws)
         max_ste_ssids = 1 + len(vms_with_stes) + len(resnames)
@@ -1083,6 +1126,8 @@ class ACMPolicy(XSPolicy):
              pr_bin += "\x00"
 
         # Build chinese wall part
+        vms_with_chws.insert(0, ACM_LABEL_UNLABELED)
+
         cfses_names = self.policy_get_chwall_cfses_names_sorted()
         cfses = self.policy_get_chwall_cfses()
 
@@ -1105,9 +1150,7 @@ class ACMPolicy(XSPolicy):
                               chw_running_types_offset,
                               chw_conf_agg_offset)
         chw_bin_body = ""
-        # simulate __NULL_LABEL__
-        for c in chws:
-            chw_bin_body += struct.pack("!h",0)
+
         # VMs that are listed and their chinese walls
         for v in vms_with_chws:
             for c in chws:
@@ -1143,6 +1186,8 @@ class ACMPolicy(XSPolicy):
             chw_bin += "\x00"
 
         # Build STE part
+        vms_with_stes.insert(0, ACM_LABEL_UNLABELED) # Took out in @2
+
         steformat="!iiiii"
         ste_bin = struct.pack(steformat,
                               ACM_STE_VERSION,
@@ -1152,10 +1197,7 @@ class ACMPolicy(XSPolicy):
                               struct.calcsize(steformat))
         ste_bin_body = ""
         if stes:
-            # Simulate __NULL_LABEL__
-            for s in stes:
-                ste_bin_body += struct.pack("!h",0)
-            # VMs that are listed and their chinese walls
+            # VMs that are listed and their STE types
             for v in vms_with_stes:
                 unknown_ste |= (set(stes_by_vm[v]) - set(stes))
                 for s in stes:

tools/python/xen/util/security.py

@@ -146,7 +146,7 @@ def calc_dom_ssidref_from_info(info):
                 raise VmError("VM label '%s' in wrong format." % seclab)
             typ, policyname, vmlabel = seclab.split(":")
             if typ != xsconstants.ACM_POLICY_ID:
-                raise VmError("Policy type '%s' not supported." % typ)
+                raise VmError("Policy type '%s' must be changed." % typ)
             refresh_security_policy()
             if active_policy != policyname:
                 raise VmError("Active policy '%s' different than "
@@ -155,7 +155,7 @@ def calc_dom_ssidref_from_info(info):
             ssidref = label2ssidref(vmlabel, policyname, "dom")
             return ssidref
         else:
-            return 0
+            return 0x0
     raise VmError("security.calc_dom_ssidref_from_info: info of type '%s'"
                   "not supported." % type(info))
 
@@ -232,6 +232,10 @@ def ssidref2label(ssidref_var):
     else:
         err("Instance type of ssidref not supported (must be of type 'str' or 'int')")
 
+    if ssidref == 0:
+        from xen.util.acmpolicy import ACM_LABEL_UNLABELED
+        return ACM_LABEL_UNLABELED
+
     try:
         mapfile_lock()
 
@@ -498,7 +502,7 @@ def hv_chg_policy(bin_pol, del_array, chg_array):
         rc, errors = acm.chgpolicy(bin_pol, del_array, chg_array)
     except Exception, e:
         pass
-    if (len(errors) > 0):
+    if len(errors) > 0:
         rc = -xsconstants.XSERR_HV_OP_FAILED
     return rc, errors
 
@@ -781,6 +785,24 @@ def res_security_check_xapi(rlabel, rssidref, rpolicy, xapi_dom_label):
     return rtnval
 
 
+def validate_label(label, policyref):
+    """
+       Make sure that this label is part of the currently enforced policy
+       and that it reference the current policy.
+    """
+    rc = xsconstants.XSERR_SUCCESS
+    from xen.xend.XendXSPolicyAdmin import XSPolicyAdminInstance
+    curpol = XSPolicyAdminInstance().get_loaded_policy()
+    if not curpol or curpol.get_name() != policyref:
+        rc = -xsconstants.XSERR_BAD_LABEL
+    else:
+        try:
+            label2ssidref(label, curpol.get_name() , 'res')
+        except:
+            rc = -xsconstants.XSERR_BAD_LABEL
+    return rc
+
+
 def set_resource_label_xapi(resource, reslabel_xapi, oldlabel_xapi):
     """Assign a resource label to a resource
     @param resource: The name of a resource, i.e., "phy:/dev/hda", or
@@ -805,10 +827,16 @@ def set_resource_label_xapi(resource, reslabel_xapi, oldlabel_xapi):
             return -xsconstants.XSERR_BAD_LABEL_FORMAT
         otyp, opolicyref, olabel = tmp
         # Only ACM is supported
-        if otyp != xsconstants.ACM_POLICY_ID:
+        if otyp != xsconstants.ACM_POLICY_ID  and \
+           otyp != xsconstants.INVALID_POLICY_PREFIX + \
+                   xsconstants.ACM_POLICY_ID:
             return -xsconstants.XSERR_WRONG_POLICY_TYPE
+    rc = validate_label(label, policyref)
+    if rc != xsconstants.XSERR_SUCCESS:
+        return rc
     return set_resource_label(resource, typ, policyref, label, olabel)
 
+
 def is_resource_in_use(resource):
     """ Investigate all running domains whether they use this device """
     from xen.xend import XendDomain
@@ -867,7 +895,7 @@ def get_domain_resources(dominfo):
                 resources[typ].append("%s:%s:%s" %
                                       (xsconstants.ACM_POLICY_ID,
                                        active_policy,
-                                       "unlabeled"))
+                                       ACM_LABEL_UNLABELED))
 
     return resources
 
@@ -1224,7 +1252,7 @@ def change_acm_policy(bin_pol, del_array, chg_array,
                 sec_lab, new_seclab = labels
                 if sec_lab != new_seclab:
                     log.info("Updating domain %s to new label '%s'." % \
-                             (new_seclab, sec_lab))
+                             (sec_lab, new_seclab))
                     # This better be working!
                     dominfo.set_security_label(new_seclab,
                                                sec_lab,

tools/python/xen/xend/XendDomainInfo.py

@@ -559,18 +559,8 @@ class XendDomainInfo:
             self.getDeviceController(devclass).waitForDevices()
 
     def destroyDevice(self, deviceClass, devid, force = False):
-        try:
-            dev = int(devid)
-        except ValueError:
-            # devid is not a number but a string containing either device
-            # name (e.g. xvda) or device_type/device_id (e.g. vbd/51728)
-            dev = type(devid) is str and devid.split('/')[-1] or None
-            if dev == None:
-                log.debug("Could not find the device %s", devid)
-                return None
-
-        log.debug("dev = %s", dev)
-        return self.getDeviceController(deviceClass).destroyDevice(dev, force)
+        log.debug("dev = %s", devid)
+        return self.getDeviceController(deviceClass).destroyDevice(devid, force)
 
     def getDeviceSxprs(self, deviceClass):
         if self._stateGet() in (DOM_STATE_RUNNING, DOM_STATE_PAUSED):
@@ -1463,8 +1453,6 @@ class XendDomainInfo:
         ssidref = 0
         if security.on():
             ssidref = security.calc_dom_ssidref_from_info(self.info)
-            if ssidref == 0:
-                raise VmError('VM is not properly labeled.')
             if security.has_authorization(ssidref) == False:
                 raise VmError("VM is not authorized to run.")
 

tools/python/xen/xend/server/DevController.py

@@ -203,27 +203,32 @@ class DevController:
 
         The implementation here simply deletes the appropriate paths from the
         store.  This may be overridden by subclasses who need to perform other
-        tasks on destruction.  Further, the implementation here can only
-        accept integer device IDs, or values that can be converted to
-        integers.  Subclasses may accept other values and convert them to
-        integers before passing them here.
+        tasks on destruction. The implementation here accepts integer device
+        IDs or paths containg integer deviceIDs, e.g. vfb/0.  Subclasses may
+        accept other values and convert them to integers before passing them
+        here.
         """
 
-        devid = int(devid)
+        try:
+            dev = int(devid)
+        except ValueError:
+            # Does devid contain devicetype/deviceid?
+            # Propogate exception if unable to find an integer devid
+            dev = int(type(devid) is str and devid.split('/')[-1] or None)
 
         # Modify online status /before/ updating state (latter is watched by
         # drivers, so this ordering avoids a race).
-        self.writeBackend(devid, 'online', "0")
-        self.writeBackend(devid, 'state', str(xenbusState['Closing']))
+        self.writeBackend(dev, 'online', "0")
+        self.writeBackend(dev, 'state', str(xenbusState['Closing']))
 
         if force:
-            frontpath = self.frontendPath(devid)
+            frontpath = self.frontendPath(dev)
             backpath = xstransact.Read(frontpath, "backend")
             if backpath:
                 xstransact.Remove(backpath)
             xstransact.Remove(frontpath)
 
-        self.vm._removeVm("device/%s/%d" % (self.deviceClass, devid))
+        self.vm._removeVm("device/%s/%d" % (self.deviceClass, dev))
 
     def configurations(self):
         return map(self.configuration, self.deviceIDs())

tools/python/xen/xend/server/blkif.py

@@ -73,17 +73,7 @@ class BlkifController(DevController):
             back['uuid'] = uuid
 
         if security.on():
-            (label, ssidref, policy) = \
-                                 security.get_res_security_details(uname)
-            domain_label = self.vm.get_security_label()
-            if domain_label:
-                rc = security.res_security_check_xapi(label, ssidref, policy,
-                                                      domain_label)
-                if rc == 0:
-                    raise VmError("VM's access to block device '%s' denied." %
-                                  uname)
-            else:
-                raise VmError("VM must have a security label.")
+            self.do_access_control(config, uname)
 
         devid = blkif.blkdev_name_to_number(dev)
         if devid is None:
@@ -95,6 +85,21 @@ class BlkifController(DevController):
 
         return (devid, back, front)
 
+    def do_access_control(self, config, uname):
+        (label, ssidref, policy) = \
+                             security.get_res_security_details(uname)
+        domain_label = self.vm.get_security_label()
+        if domain_label:
+            rc = security.res_security_check_xapi(label, ssidref, policy,
+                                                  domain_label)
+            if rc == 0:
+                raise VmError("VM's access to block device '%s' denied" %
+                              uname)
+        else:
+            from xen.util.acmpolicy import ACM_LABEL_UNLABELED
+            if label != ACM_LABEL_UNLABELED:
+                raise VmError("VM must have a security label to access "
+                              "block device '%s'" % uname)
 
     def reconfigureDevice(self, _, config):
         """@see DevController.reconfigureDevice"""
@@ -149,13 +154,16 @@ class BlkifController(DevController):
     def destroyDevice(self, devid, force):
         """@see DevController.destroyDevice"""
 
-        # If we are given a device name, then look up the device ID from it,
-        # and destroy that ID instead.  If what we are given is an integer,
-        # then assume it's a device ID and pass it straight through to our
-        # superclass's method.
-
+        # vbd device IDs can be either string or integer.  Further, the
+        # following string values are possible:
+        #    - devicetype/deviceid (vbd/51728)
+        #    - devicetype/devicename (/dev/xvdb)
+        #    - devicename (xvdb)
+        # Let our superclass handle integer or devicetype/deviceid forms.
+        # If we are given a device name form, then look up the device ID
+        # from it, and destroy that ID instead.
         try:
-            DevController.destroyDevice(self, int(devid), force)
+            DevController.destroyDevice(self, devid, force)
         except ValueError:
             devid_end = type(devid) is str and devid.split('/')[-1] or None
 

tools/python/xen/xm/activatepolicy.py

@@ -23,7 +23,7 @@ import sys
 from xen.util import xsconstants
 from xml.dom import minidom
 from xen.xm.opts import OptionError
-from xen.xm import getpolicy
+from xen.xm import getpolicy, setpolicy
 from xen.xm import main as xm_main
 from xen.xm.main import server
 
@@ -38,6 +38,9 @@ def help():
       --boot     Have the system boot with the policy. Changes the default
                  title in grub.conf.
       --noboot   Remove the policy from the default entry in grub.conf.
+      --remove   Attempt to remove the current policy by installing the
+                 default policy; this works only if no domains are
+                 running.
     """
 
 def activate_policy(flags):
@@ -56,6 +59,25 @@ def activate_policy(flags):
 def remove_bootpolicy():
     server.xenapi.XSPolicy.rm_xsbootpolicy()
 
+def install_default_policy():
+    if xm_main.serverType != xm_main.SERVER_XEN_API:
+        raise OptionError('xm needs to be configured to use the xen-api.')
+    xs_type = int(server.xenapi.XSPolicy.get_xstype())
+    if xs_type & xsconstants.XS_POLICY_ACM == 0:
+        raise OptionError('ACM policy type not supported on system.')
+    policystate = server.xenapi.XSPolicy.get_xspolicy()
+    if int(policystate['type']) == 0:
+        print 'No policy is installed.'
+        return
+    if int(policystate['type']) != xsconstants.XS_POLICY_ACM:
+        print "Unknown policy type '%s'." % policystate['type']
+    flags = int(policystate['flags'])
+    if flags & xsconstants.XS_INST_LOAD == 0:
+        print "Default policy is already loaded."
+        return
+    setpolicy.setpolicy(xsconstants.ACM_POLICY_ID, 'default', flags, True,
+                        False)
+
 def main(argv):
     if xm_main.serverType != xm_main.SERVER_XEN_API:
         raise OptionError('xm needs to be configured to use the xen-api.')
@@ -69,6 +91,9 @@ def main(argv):
             flags |= xsconstants.XS_INST_LOAD
         elif '--noboot' == argv[c]:
             remove_bootpolicy()
+        elif '--remove' == argv[c]:
+            install_default_policy()
+            return
         else:
             raise OptionError("Unknown command line option '%s'" % argv[c])
         c += 1

tools/python/xen/xm/main.py

@@ -25,7 +25,6 @@ import atexit
 import cmd
 import os
 import pprint
-import readline
 import shlex
 import sys
 import re
@@ -51,6 +50,7 @@ from xen.xm.opts import OptionError, Opts, wrap, set_true
 from xen.xm import console
 from xen.util.xmlrpcclient import ServerProxy
 from xen.util.security import ACMError
+from xen.util.acmpolicy import ACM_LABEL_UNLABELED_DISPLAY
 
 import XenAPI
 
@@ -615,6 +615,7 @@ class Shell(cmd.Cmd):
 
     def preloop(self):
         cmd.Cmd.preloop(self)
+        import readline
         readline.set_completer_delims(' ')
 
     def default(self, line):
@@ -947,7 +948,7 @@ def xm_label_list(doms):
         d = parse_doms_info(dom)
         if security.active_policy not in ['INACTIVE', 'NULL', 'DEFAULT']:
             if not d['seclabel']:
-                d['seclabel'] = 'ERROR'
+                d['seclabel'] = ACM_LABEL_UNLABELED_DISPLAY
         elif security.active_policy in ['DEFAULT']:
             d['seclabel'] = 'DEFAULT'
         else:

tools/python/xen/xm/new.py

@@ -75,7 +75,8 @@ def main(argv):
         doc = sxp2xml_inst.convert_sxp_to_xml(config) 
         
         xenapi_create_inst = xenapi_create()
-        vm_refs = xenapi_create_inst.create(document = doc)
+        vm_refs = xenapi_create_inst.create(document = doc,
+                                            skipdtd=opts.vals.skipdtd)
     else:       
         make_unstarted_domain(opts, config)
         

tools/python/xen/xm/resources.py

@@ -58,7 +58,6 @@ def main (argv):
         try:
             filename = security.res_label_filename
             access_control = dictio.dict_read("resources", filename)
-            print access_control
         except:
             raise OptionError("Resource file not found")
 

tools/python/xen/xm/setpolicy.py

@@ -24,6 +24,7 @@ import struct
 import sys
 import string
 from xen.util import xsconstants
+from xen.util.acmpolicy import ACMPolicy
 from xen.xm.opts import OptionError
 from xen.util.security import policy_dir_prefix
 from xen.xm import main as xm_main
@@ -40,9 +41,38 @@ def help():
     The following options are defined
       --load     Load the policy immediately
       --boot     Have the system load the policy during boot
+      --update   Automatically adapt the policy so that it will be
+                 treated as an update to the current policy
     """
 
-def setpolicy(policytype, policy_name, flags, overwrite):
+def create_update_xml(xml):
+    """
+        Adapt the new policy's xml header to be a simple type of an
+        update to the currently enforce policy on the remote system.
+        Increases the minor number by '1'.
+    """
+    policystate = server.xenapi.XSPolicy.get_xspolicy()
+    if int(policystate['type']) == 0:
+        return xml
+    curpol = ACMPolicy(xml = policystate['repr'])
+    curpol_version = curpol.get_version()
+    tmp = curpol_version.split('.')
+    if len(tmp) == 2:
+        maj = int(tmp[0])
+        min = int(tmp[1])
+    else:
+        maj = int(tmp)
+        min = 0
+    min += 1
+    newpol_version = ""+str(maj)+"."+str(min)
+
+    newpol = ACMPolicy(xml = xml)
+    newpol.set_frompolicy_name(curpol.get_name())
+    newpol.set_frompolicy_version(curpol.get_version())
+    newpol.set_policy_version(newpol_version)
+    return newpol.toxml()
+
+def setpolicy(policytype, policy_name, flags, overwrite, is_update=False):
     if xm_main.serverType != xm_main.SERVER_XEN_API:
         raise OptionError('xm needs to be configured to use the xen-api.')
     if policytype != xsconstants.ACM_POLICY_ID:
@@ -61,6 +91,9 @@ def setpolicy(policytype, policy_name, flags, overwrite):
         except:
             raise OptionError("Not a valid policy file")
 
+        if is_update:
+            xml = create_update_xml(xml)
+
         try:
             policystate = server.xenapi.XSPolicy.set_xspolicy(xs_type,
                                                               xml,
@@ -96,18 +129,21 @@ def main(argv):
 
     policytype  = argv[1]
     policy_name = argv[2]
+    is_update = False
 
     flags = 0
     if '--load' in argv:
         flags |= xsconstants.XS_INST_LOAD
     if '--boot' in argv:
         flags |= xsconstants.XS_INST_BOOT
+    if '--update' in argv:
+        is_update = True
 
     overwrite = True
     if '--nooverwrite' in argv:
         overwrite = False
 
-    setpolicy(policytype, policy_name, flags, overwrite)
+    setpolicy(policytype, policy_name, flags, overwrite, is_update)
 
 if __name__ == '__main__':
     try:

tools/security/Makefile

@@ -13,13 +13,6 @@ CFLAGS     += $(shell xml2-config --cflags )
 CFLAGS     += $(shell if [[ $(XML2VERSION) < 2.6.20 ]]; then echo ""; else echo "-DVALIDATE_SCHEMA"; fi )
 LDFLAGS    += $(shell xml2-config --libs ) # if this does not work, try -L/usr/lib -lxml2 -lz -lpthread -lm
 
-ifeq ($(ACM_DEFAULT_SECURITY_POLICY),ACM_NULL_POLICY)
-POLICY=null
-endif
-ifeq ($(ACM_DEFAULT_SECURITY_POLICY),ACM_CHINESE_WALL_AND_SIMPLE_TYPE_ENFORCEMENT_POLICY)
-POLICY=chwall_ste
-endif
-
 SRCS_TOOL     = secpol_tool.c
 OBJS_TOOL    := $(patsubst %.c,%.o,$(filter %.c,$(SRCS_TOOL)))
 SRCS_XML2BIN  = secpol_xml2bin.c secpol_xml2bin.h
@@ -41,6 +34,7 @@ ACM_SECGEN_CGIDIR = $(ACM_SECGEN_HTMLDIR)/cgi-bin
 
 ACM_SCHEMA        = security_policy.xsd
 ACM_EXAMPLES      = client_v1 test
+ACM_DEF_POLICIES  = default default-ul
 ACM_POLICY_SUFFIX = security_policy.xml
 
 ifeq ($(ACM_SECURITY),y)
@@ -62,6 +56,9 @@ install: all $(ACM_CONFIG_FILE)
 	for i in $(ACM_EXAMPLES); do \
 		$(INSTALL_DATA) policies/example/$$i-$(ACM_POLICY_SUFFIX) $(DESTDIR)$(ACM_POLICY_DIR)/example/; \
 	done
+	for i in $(ACM_DEF_POLICIES); do \
+		$(INSTALL_DATA) policies/$$i-$(ACM_POLICY_SUFFIX) $(DESTDIR)$(ACM_POLICY_DIR); \
+	done
 	$(INSTALL_DIR) $(DESTDIR)$(ACM_SCRIPT_DIR)
 	$(INSTALL_PROG) $(ACM_SCRIPTS) $(DESTDIR)$(ACM_SCRIPT_DIR)
 	$(INSTALL_DIR) $(DESTDIR)$(ACM_SECGEN_HTMLDIR)

tools/security/policies/default-security_policy.xml

@@ -0,0 +1,30 @@
+<?xml version="1.0" ?>
+<SecurityPolicyDefinition xmlns="http://www.ibm.com" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://www.ibm.com ../../security_policy.xsd">
+  <PolicyHeader>
+    <PolicyName>DEFAULT</PolicyName>
+    <Version>1.0</Version>
+  </PolicyHeader>
+  <SimpleTypeEnforcement>
+    <SimpleTypeEnforcementTypes>
+      <Type>SystemManagement</Type>
+    </SimpleTypeEnforcementTypes>
+  </SimpleTypeEnforcement>
+  <ChineseWall>
+    <ChineseWallTypes>
+      <Type>SystemManagement</Type>
+    </ChineseWallTypes>
+  </ChineseWall>
+  <SecurityLabelTemplate>
+    <SubjectLabels bootstrap="SystemManagement">
+      <VirtualMachineLabel>
+        <Name>SystemManagement</Name>
+        <SimpleTypeEnforcementTypes>
+          <Type>SystemManagement</Type>
+        </SimpleTypeEnforcementTypes>
+        <ChineseWallTypes>
+          <Type/>
+        </ChineseWallTypes>
+      </VirtualMachineLabel>
+    </SubjectLabels>
+  </SecurityLabelTemplate>
+</SecurityPolicyDefinition>

tools/security/policies/default-ul-security_policy.xml

@@ -0,0 +1,41 @@
+<?xml version="1.0" ?>
+<SecurityPolicyDefinition xmlns="http://www.ibm.com" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://www.ibm.com ../../security_policy.xsd">
+  <PolicyHeader>
+    <PolicyName>DEFAULT-UL</PolicyName>
+    <Version>1.0</Version>
+  </PolicyHeader>
+  <SimpleTypeEnforcement>
+    <SimpleTypeEnforcementTypes>
+      <Type>SystemManagement</Type>
+      <Type>__UNLABELED__</Type>
+    </SimpleTypeEnforcementTypes>
+  </SimpleTypeEnforcement>
+  <ChineseWall>
+    <ChineseWallTypes>
+      <Type>SystemManagement</Type>
+    </ChineseWallTypes>
+  </ChineseWall>
+  <SecurityLabelTemplate>
+    <SubjectLabels bootstrap="SystemManagement">
+      <VirtualMachineLabel>
+        <Name>SystemManagement</Name>
+        <SimpleTypeEnforcementTypes>
+          <Type>SystemManagement</Type>
+          <Type>__UNLABELED__</Type>
+        </SimpleTypeEnforcementTypes>
+        <ChineseWallTypes>
+          <Type/>
+        </ChineseWallTypes>
+      </VirtualMachineLabel>
+      <VirtualMachineLabel>
+        <Name>__UNLABELED__</Name>
+        <SimpleTypeEnforcementTypes>
+          <Type>__UNLABELED__</Type>
+        </SimpleTypeEnforcementTypes>
+        <ChineseWallTypes>
+          <Type/>
+        </ChineseWallTypes>
+      </VirtualMachineLabel>
+    </SubjectLabels>
+  </SecurityLabelTemplate>
+</SecurityPolicyDefinition>

tools/security/xensec_ezpolicy

@@ -36,6 +36,8 @@ conflict_bmp = None
 realm_icon = None
 workload_icon = None
 
+ACM_LABEL_UNLABELED = '__UNLABELED__'
+
 class orgTreeCtrl(wx.TreeCtrl):
 
     event = None
@@ -870,7 +872,8 @@ class ezFrame(wx.Frame):
             self.realm_menu.Enable(self.ID_ORGDEL, True)
             self.realm_menu.Enable(self.ID_ORGEDT, True)
             self.realm_menu.Enable(self.ID_ORGADD, True)
-            if len(self.orgs.GetSelections()) > 1:
+            if len(self.orgs.GetSelections()) > 1 or \
+               ACM_LABEL_UNLABELED == self.orgs.GetItemText(item):
                 self.realm_menu.Enable(self.ID_ORGEDT, False)
                 self.realm_menu.Enable(self.ID_ORGADD, False)
             self.PopupMenu(self.realm_menu)
@@ -1622,6 +1625,8 @@ def main():
     app = ezApp(0)
     if len(sys.argv) in [2]:
         app.Load(sys.argv[1])
+    else:
+        dict2org({'orgs' : [[ACM_LABEL_UNLABELED,[]]], 'cons': []})
     app.MainLoop()
     print "Goodbye"
 

tools/xenstat/libxenstat/src/xenstat.c

@@ -677,7 +677,7 @@ static void xenstat_prune_domain(xenstat_node *node, unsigned int entry)
 	/* shift entries following specified entry up by one */
 	if (entry < node->num_domains) {
 		xenstat_domain *domain = &node->domains[entry];
-		memmove(domain,domain+1,node->num_domains-entry);
+		memmove(domain,domain+1,(node->num_domains - entry) * sizeof(xenstat_domain) );
 	}
 
 	/* zero out original last entry from node -- not

tools/xenstore/utils.c

@@ -10,19 +10,18 @@
 #include <signal.h>
 #include "utils.h"
 
-void xprintf(const char *fmt, ...)
+static void default_xprintf(const char *fmt, ...)
 {
 	va_list args;
 
-	if (!stderr)
-		return; /* could trace()? */
-
 	va_start(args, fmt);
 	vfprintf(stderr, fmt, args);
 	va_end(args);
 	fflush(stderr);
 }
 
+void (*xprintf)(const char *fmt, ...) = default_xprintf;
+
 void barf(const char *fmt, ...)
 {
 	char *str;

tools/xenstore/utils.h

@@ -24,7 +24,7 @@ static inline bool strends(const char *a, const char *b)
 void barf(const char *fmt, ...) __attribute__((noreturn));
 void barf_perror(const char *fmt, ...) __attribute__((noreturn));
 
-void xprintf(const char *fmt, ...);
+void (*xprintf)(const char *fmt, ...);
 
 #define eprintf(_fmt, _args...) xprintf("[ERR] %s" _fmt, __FUNCTION__, ##_args)
 

tools/xenstore/xenstored_core.c

@@ -1880,14 +1880,14 @@ int main(int argc, char *argv[])
 
 	/* close stdin/stdout now we're ready to accept connections */
 	if (dofork) {
-		close(STDIN_FILENO);
-		close(STDOUT_FILENO);
-		close(STDERR_FILENO);
-
-		/* Get ourselves a nice xenstored crash if these are used. */
-		stdin = NULL;
-		stdout = NULL;
-		stderr = NULL;
+		int devnull = open("/dev/null", O_RDWR);
+		if (devnull == -1)
+			barf_perror("Could not open /dev/null\n");
+		close(STDIN_FILENO);  dup2(STDIN_FILENO, devnull);
+		close(STDOUT_FILENO); dup2(STDOUT_FILENO, devnull);
+		close(STDERR_FILENO); dup2(STDERR_FILENO, devnull);
+		close(devnull);
+		xprintf = trace;
 	}
 
 	signal(SIGHUP, trigger_reopen_log);

unmodified_drivers/linux-2.6/platform-pci/machine_reboot.c

@@ -11,6 +11,12 @@ struct ap_suspend_info {
 	atomic_t nr_spinning;
 };
 
+/*
+ * Use a rwlock to protect the hypercall page from being executed in AP context
+ * while the BSP is re-initializing it after restore.
+ */
+static DEFINE_RWLOCK(suspend_lock);
+
 /*
  * Spinning prevents, for example, APs touching grant table entries while
  * the shared grant table is not mapped into the address space imemdiately
@@ -27,7 +33,9 @@ static void ap_suspend(void *_info)
 
 	while (info->do_spin) {
 		cpu_relax();
+		read_lock(&suspend_lock);
 		HYPERVISOR_yield();
+		read_unlock(&suspend_lock);
 	}
 
 	mb();
@@ -43,7 +51,9 @@ static int bp_suspend(void)
 	suspend_cancelled = HYPERVISOR_shutdown(SHUTDOWN_suspend);
 
 	if (!suspend_cancelled) {
+		write_lock(&suspend_lock);
 		platform_pci_resume();
+		write_unlock(&suspend_lock);
 		gnttab_resume();
 		irq_resume();
 	}

xen/acm/acm_chinesewall_hooks.c

@@ -93,6 +93,7 @@ int acm_init_chwall_policy(void)
     return ACM_OK;
 }
 
+
 static int chwall_init_domain_ssid(void **chwall_ssid, ssidref_t ssidref)
 {
     struct chwall_ssid *chwall_ssidp = xmalloc(struct chwall_ssid);
@@ -104,10 +105,10 @@ static int chwall_init_domain_ssid(void **chwall_ssid, ssidref_t ssidref)
     chwall_ssidp->chwall_ssidref =
         GET_SSIDREF(ACM_CHINESE_WALL_POLICY, ssidref);
 
-    if ( (chwall_ssidp->chwall_ssidref >= chwall_bin_pol.max_ssidrefs)
-        || (chwall_ssidp->chwall_ssidref == ACM_DEFAULT_LOCAL_SSID) )
+    if ( chwall_ssidp->chwall_ssidref >= chwall_bin_pol.max_ssidrefs )
     {
-        printkd("%s: ERROR chwall_ssidref(%x) undefined (>max) or unset (0).\n",
+        printkd("%s: ERROR chwall_ssidref(%x) undefined (>max) or unset "
+                "(0).\n",
                 __func__, chwall_ssidp->chwall_ssidref);
         xfree(chwall_ssidp);
         return ACM_INIT_SSID_ERROR;
@@ -118,6 +119,7 @@ static int chwall_init_domain_ssid(void **chwall_ssid, ssidref_t ssidref)
     return ACM_OK;
 }
 
+
 static void chwall_free_domain_ssid(void *chwall_ssid)
 {
     xfree(chwall_ssid);
@@ -205,7 +207,9 @@ chwall_init_state(struct acm_chwall_policy_buffer *chwall_buf,
 
     read_lock(&ssid_list_rwlock);
 
-    /* go through all domains and adjust policy as if this domain was started now */
+    /* go through all domains and adjust policy as if this domain was
+     * started now
+     */
     for_each_acmssid( rawssid )
     {
         chwall_ssid =
@@ -220,8 +224,8 @@ chwall_init_state(struct acm_chwall_policy_buffer *chwall_buf,
 
         /* b) check for conflict */
         for ( i = 0; i < chwall_buf->chwall_max_types; i++ )
-            if (conflict_aggregate_set[i] &&
-                ssidrefs[chwall_ssidref * chwall_buf->chwall_max_types + i])
+            if ( conflict_aggregate_set[i] &&
+                 ssidrefs[chwall_ssidref * chwall_buf->chwall_max_types + i] )
             {
                 printk("%s: CHINESE WALL CONFLICT in type %02x.\n",
                        __func__, i);
@@ -231,37 +235,46 @@ chwall_init_state(struct acm_chwall_policy_buffer *chwall_buf,
 
                 goto out;
             }
+
         /* set violation and break out of the loop */
-        /* c) adapt conflict aggregate set for this domain (notice conflicts) */
+        /* c) adapt conflict aggregate set for this domain
+         *    (notice conflicts)
+         */
         for ( i = 0; i < chwall_buf->chwall_max_conflictsets; i++ )
         {
             int common = 0;
             /* check if conflict_set_i and ssidref have common types */
             for ( j = 0; j < chwall_buf->chwall_max_types; j++ )
-                if (conflict_sets[i * chwall_buf->chwall_max_types + j] &&
-                    ssidrefs[chwall_ssidref *
-                            chwall_buf->chwall_max_types + j])
+                if ( conflict_sets[i * chwall_buf->chwall_max_types + j] &&
+                     ssidrefs[chwall_ssidref *
+                              chwall_buf->chwall_max_types + j] )
                 {
                     common = 1;
                     break;
                 }
-            if (common == 0)
+
+            if ( common == 0 )
                 continue;       /* try next conflict set */
-            /* now add types of the conflict set to conflict_aggregate_set (except types in chwall_ssidref) */
+
+            /* now add types of the conflict set to conflict_aggregate_set
+             * (except types in chwall_ssidref)
+             */
             for ( j = 0; j < chwall_buf->chwall_max_types; j++ )
-                if (conflict_sets[i * chwall_buf->chwall_max_types + j] &&
-                    !ssidrefs[chwall_ssidref *
-                             chwall_buf->chwall_max_types + j])
+                if ( conflict_sets[i * chwall_buf->chwall_max_types + j] &&
+                     !ssidrefs[chwall_ssidref *
+                               chwall_buf->chwall_max_types + j] )
                     conflict_aggregate_set[j]++;
         }
     }
  out:
     read_unlock(&ssid_list_rwlock);
     return violation;
-    /* returning "violation != 0" means that the currently running set of domains would
-     * not be possible if the new policy had been enforced before starting them; for chinese
-     * wall, this means that the new policy includes at least one conflict set of which
-     * more than one type is currently running */
+    /* returning "violation != 0" means that the currently running set of
+     * domains would not be possible if the new policy had been enforced
+     * before starting them; for chinese wall, this means that the new
+     * policy includes at least one conflict set of which more than one
+     * type is currently running
+     */
 }
 
 
@@ -348,8 +361,10 @@ static int _chwall_update_policy(u8 *buf, u32 buf_size, int test_only,
     memset(conflict_aggregate_set, 0,
            sizeof(domaintype_t) * chwall_buf->chwall_max_types);
 
-    /* 3. now re-calculate the state for the new policy based on running domains;
-     *    this can fail if new policy is conflicting with running domains */
+    /* 3. now re-calculate the state for the new policy based on
+     *    running domains; this can fail if new policy is conflicting
+     *    with running domains
+     */
     if ( chwall_init_state(chwall_buf, ssids,
                            conflict_sets, running_types,
                            conflict_aggregate_set,
@@ -483,31 +498,27 @@ static int _chwall_pre_domain_create(void *subject_ssid, ssidref_t ssidref)
 
     chwall_ssidref = GET_SSIDREF(ACM_CHINESE_WALL_POLICY, ssidref);
 
-    if (chwall_ssidref == ACM_DEFAULT_LOCAL_SSID)
-    {
-        printk("%s: ERROR CHWALL SSID is NOT SET but policy enforced.\n",
-               __func__);
-        return ACM_ACCESS_DENIED;       /* catching and indicating config error */
-    }
-
-    if (chwall_ssidref >= chwall_bin_pol.max_ssidrefs)
+    if ( chwall_ssidref >= chwall_bin_pol.max_ssidrefs )
     {
         printk("%s: ERROR chwall_ssidref > max(%x).\n",
                __func__, chwall_bin_pol.max_ssidrefs - 1);
         return ACM_ACCESS_DENIED;
     }
+
     /* A: chinese wall check for conflicts */
-    for (i = 0; i < chwall_bin_pol.max_types; i++)
-        if (chwall_bin_pol.conflict_aggregate_set[i] &&
-            chwall_bin_pol.ssidrefs[chwall_ssidref *
-                                   chwall_bin_pol.max_types + i])
+    for ( i = 0; i < chwall_bin_pol.max_types; i++ )
+        if ( chwall_bin_pol.conflict_aggregate_set[i] &&
+             chwall_bin_pol.ssidrefs[chwall_ssidref *
+                                     chwall_bin_pol.max_types + i] )
         {
             printk("%s: CHINESE WALL CONFLICT in type %02x.\n", __func__, i);
             return ACM_ACCESS_DENIED;
         }
 
     /* B: chinese wall conflict set adjustment (so that other
-     *      other domains simultaneously created are evaluated against this new set)*/
+     *    other domains simultaneously created are evaluated against
+     *    this new set)
+     */
     for ( i = 0; i < chwall_bin_pol.max_conflictsets; i++ )
     {
         int common = 0;
@@ -521,7 +532,7 @@ static int _chwall_pre_domain_create(void *subject_ssid, ssidref_t ssidref)
                 common = 1;
                 break;
             }
-        if (common == 0)
+        if ( common == 0 )
             continue;           /* try next conflict set */
         /* now add types of the conflict set to conflict_aggregate_set (except types in chwall_ssidref) */
         for ( j = 0; j < chwall_bin_pol.max_types; j++ )
@@ -571,9 +582,15 @@ static void _chwall_post_domain_create(domid_t domid, ssidref_t ssidref)
                 common = 1;
                 break;
             }
+
         if ( common == 0 )
-            continue;           /* try next conflict set */
-        /* now add types of the conflict set to conflict_aggregate_set (except types in chwall_ssidref) */
+        {
+            /* try next conflict set */
+            continue;
+        }
+
+        /* now add types of the conflict set to conflict_aggregate_set
+           (except types in chwall_ssidref) */
         for ( j = 0; j < chwall_bin_pol.max_types; j++ )
             if ( chwall_bin_pol.
                  conflict_sets[i * chwall_bin_pol.max_types + j]
@@ -638,9 +655,15 @@ static void chwall_domain_destroy(void *object_ssid, struct domain *d)
                 common = 1;
                 break;
             }
-        if (common == 0)
-            continue;           /* try next conflict set, this one does not include any type of chwall_ssidref */
-        /* now add types of the conflict set to conflict_aggregate_set (except types in chwall_ssidref) */
+        if ( common == 0 )
+        {
+            /* try next conflict set, this one does not include
+               any type of chwall_ssidref */
+            continue;
+        }
+
+        /* now add types of the conflict set to conflict_aggregate_set
+           (except types in chwall_ssidref) */
         for ( j = 0; j < chwall_bin_pol.max_types; j++ )
             if ( chwall_bin_pol.
                  conflict_sets[i * chwall_bin_pol.max_types + j]

xen/acm/acm_policy.c

@@ -710,10 +710,10 @@ acm_change_policy(struct acm_change_policy *chgpolicy)
         goto acm_chg_policy_exit;
     }
 
-    if ( copy_from_guest(dels.array,
+    if ( copy_from_guest((u8 *)dels.array,
                          chgpolicy->del_array,
                          chgpolicy->delarray_size) ||
-         copy_from_guest(ssidmap.array,
+         copy_from_guest((u8 *)ssidmap.array,
                          chgpolicy->chg_array,
                          chgpolicy->chgarray_size) ||
          copy_from_guest(binpolicy,
@@ -844,7 +844,7 @@ acm_relabel_domains(struct acm_relabel_doms *relabel)
         memset(errors.array, 0x0, sizeof(uint32_t) * errors.num_items);
     }
 
-    if ( copy_from_guest(relabels.array,
+    if ( copy_from_guest((u8 *)relabels.array,
                          relabel->relabel_map,
                          relabel->relabel_map_size) )
     {

xen/acm/acm_simple_type_enforcement_hooks.c

@@ -408,7 +408,7 @@ _ste_update_policy(u8 *buf, u32 buf_size, int test_only,
         ste_bin_pol.max_ssidrefs = ste_buf->ste_max_ssidrefs;
         ste_bin_pol.ssidrefs = (domaintype_t *)ssidrefsbuf;
 
-        if ( ste_init_state(NULL) )
+        if ( ste_init_state(errors) )
         {
             /* new policy conflicts with sharing of running domains */
             printk("%s: New policy conflicts with running domains. "

xen/arch/powerpc/multiboot2.c

@@ -0,0 +1,67 @@
+/*
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; either version 2 of the
+ * License, or (at your option) any later version.
+ *
+ * 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 General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
+ *
+ * Copyright IBM Corp. 2006, 2007
+ *
+ * Authors: Hollis Blanchard <hollisb@us.ibm.com>
+ */
+
+#include <xen/config.h>
+#include <xen/lib.h>
+#include <xen/multiboot2.h>
+#include <asm/boot.h>
+#include <asm/init.h>
+
+static struct mb2_tag_module *mb2_tag_mod_find(struct mb2_tag_header *tags,
+                                                const char *type)
+{
+    struct mb2_tag_header *tag;
+
+    for_each_tag(tag, tags) {
+        if (tag->key == MB2_TAG_MODULE) {
+            struct mb2_tag_module *mod = (struct mb2_tag_module *)tag;
+            if (!strcmp((char *)mod->type, type))
+                return mod;
+        }
+    }
+    return NULL;
+}
+
+void parse_multiboot(ulong tags_addr)
+{
+    struct mb2_tag_header *tags = (struct mb2_tag_header *)tags_addr;
+    struct mb2_tag_module *mod;
+
+    if (tags->key != MB2_TAG_START)
+        return;
+
+    mod = mb2_tag_mod_find(tags, "kernel");
+    if (mod) {
+        xen_cmdline = (char *)mod->cmdline;
+    }
+
+    mod = mb2_tag_mod_find(tags, "dom0");
+    if (mod) {
+        dom0_addr = mod->addr;
+        dom0_len = mod->size;
+        dom0_cmdline = (char *)mod->cmdline;
+    }
+
+    mod = mb2_tag_mod_find(tags, "initrd");
+    if (mod) {
+        initrd_start = mod->addr;
+        initrd_len = mod->size;
+    }
+}

xen/arch/x86/domain_build.c

@@ -316,9 +316,9 @@ int __init construct_dom0(
            parms.pae       ? ", PAE"  : "",
            elf_msb(&elf)   ? "msb"    : "lsb",
            elf.pstart, elf.pend);
-    if ( parms.bsd_symtab )
+    if ( elf.bsd_symtab_pstart )
         printk(" Dom0 symbol map 0x%" PRIx64 " -> 0x%" PRIx64 "\n",
-               elf.sstart, elf.send);
+               elf.bsd_symtab_pstart, elf.bsd_symtab_pend);
 
     if ( !compatible )
     {
@@ -388,7 +388,7 @@ int __init construct_dom0(
     v_start          = parms.virt_base;
     vkern_start      = parms.virt_kstart;
     vkern_end        = parms.virt_kend;
-    vinitrd_start    = round_pgup(parms.virt_end);
+    vinitrd_start    = round_pgup(vkern_end);
     vinitrd_end      = vinitrd_start + initrd_len;
     vphysmap_start   = round_pgup(vinitrd_end);
     vphysmap_end     = vphysmap_start + (nr_pages * (!is_pv_32on64_domain(d) ?
@@ -798,7 +798,7 @@ int __init construct_dom0(
 
     /* Copy the OS image and free temporary buffer. */
     elf.dest = (void*)vkern_start;
-    elf_xen_dom_load_binary(&elf, &parms);
+    elf_load_binary(&elf);
 
     if ( UNSET_ADDR != parms.virt_hypercall )
     {

xen/arch/x86/genapic/Makefile

@@ -1,7 +1,5 @@
 obj-y += bigsmp.o
 obj-y += default.o
 obj-y += delivery.o
-obj-y += es7000.o
-obj-y += es7000plat.o
 obj-y += probe.o
 obj-y += summit.o

xen/arch/x86/genapic/bigsmp.c

@@ -13,7 +13,7 @@
 
 static int dmi_bigsmp; /* can be set by dmi scanners */
 
-static __init int hp_ht_bigsmp(struct dmi_system_id *d)
+static __init int force_bigsmp(struct dmi_system_id *d)
 {
 	printk(KERN_NOTICE "%s detected: force use of apic=bigsmp\n", d->ident);
 	dmi_bigsmp = 1;
@@ -22,15 +22,19 @@ static __init int hp_ht_bigsmp(struct dmi_system_id *d)
 
 
 static struct dmi_system_id __initdata bigsmp_dmi_table[] = {
-	{ hp_ht_bigsmp, "HP ProLiant DL760 G2", {
+	{ force_bigsmp, "HP ProLiant DL760 G2", {
 		DMI_MATCH(DMI_BIOS_VENDOR, "HP"),
 		DMI_MATCH(DMI_BIOS_VERSION, "P44-"),
 	}},
 
-	{ hp_ht_bigsmp, "HP ProLiant DL740", {
+	{ force_bigsmp, "HP ProLiant DL740", {
 		DMI_MATCH(DMI_BIOS_VENDOR, "HP"),
 		DMI_MATCH(DMI_BIOS_VERSION, "P47-"),
 	 }},
+	{ force_bigsmp, "UNISYS ES7000-ONE", {
+		DMI_MATCH(DMI_PRODUCT_NAME, "ES7000-ONE")
+	 }},
+	
 	 { }
 };
 

xen/arch/x86/genapic/es7000.c

@@ -1,27 +0,0 @@
-/*
- * APIC driver for the Unisys ES7000 chipset.
- */
-#include <xen/config.h>
-#include <xen/cpumask.h>
-#include <asm/current.h>
-#include <asm/mpspec.h>
-#include <asm/genapic.h>
-#include <asm/fixmap.h>
-#include <asm/apicdef.h>
-#include <asm/atomic.h>
-#include <xen/kernel.h>
-#include <xen/string.h>
-#include <xen/smp.h>
-#include <xen/init.h>
-#include <asm/mach-es7000/mach_mpparse.h>
-
-static __init int probe_es7000(void)
-{
-	/* probed later in mptable/ACPI hooks */
-	return 0;
-}
-
-struct genapic apic_es7000 = {
-	APIC_INIT("es7000", probe_es7000),
-	GENAPIC_PHYS
-};

xen/arch/x86/genapic/es7000plat.c

@@ -1,141 +0,0 @@
-/*
- * Written by: Garry Forsgren, Unisys Corporation
- *             Natalie Protasevich, Unisys Corporation
- * Modified by: Raj Subrahmanian <raj.subrahmanian@unisys.com> Unisys Corp.
- * This file contains the code to configure and interface
- * with Unisys ES7000 series hardware system manager.
- *
- * Copyright (c) 2003 Unisys Corporation.  All Rights Reserved.
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of version 2 of the GNU General Public License as
- * published by the Free Software Foundation.
- *
- * This program is distributed in the hope that it would be useful, but
- * WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
- *
- * You should have received a copy of the GNU General Public License along
- * with this program; if not, write the Free Software Foundation, Inc., 59
- * Temple Place - Suite 330, Boston MA 02111-1307, USA.
- *
- * Contact information: Unisys Corporation, Township Line & Union Meeting
- * Roads-A, Unisys Way, Blue Bell, Pennsylvania, 19424, or:
- *
- * http://www.unisys.com
- */
-
-#include <xen/config.h>
-#include <xen/types.h>
-#include <xen/kernel.h>
-#include <xen/smp.h>
-#include <xen/string.h>
-#include <xen/spinlock.h>
-#include <xen/errno.h>
-#include <xen/init.h>
-#include <xen/acpi.h>
-#include <asm/io.h>
-#include <asm/smp.h>
-#include <asm/apicdef.h>
-
-#define	MIP_REG			1
-#define	MIP_PSAI_REG		4
-
-struct acpi_table_sdt {
-	unsigned long pa;
-	unsigned long count;
-	struct {
-		unsigned long pa;
-		enum acpi_table_id id;
-		unsigned long size;
-	}	entry[50];
-};
-
-struct oem_table {
-	struct acpi_table_header Header;
-	u32 OEMTableAddr;
-	u32 OEMTableSize;
-};
-
-int __init
-parse_unisys_oem(char *oemptr)
-{
-	int                     i;
-	int 			success = 0;
-	unsigned char           type, size;
-	char                    *tp = NULL;
-
-	tp = oemptr;
-
-	tp += 8;
-
-	for (i=0; i <= 6; i++) {
-		type = *tp++;
-		size = *tp++;
-		tp -= 2;
-		switch (type) {
-		case MIP_REG:
-		case MIP_PSAI_REG:
-			success++;
-			break;
-		default:
-			break;
-		}
-		tp += size;
-	}
-
-	return (success >= 2);
-}
-
-int __init
-find_unisys_acpi_oem_table(unsigned long *oem_addr)
-{
-	struct acpi_table_rsdp		*rsdp = NULL;
-	unsigned long			rsdp_phys = 0;
-	struct acpi_table_header 	*header = NULL;
-	int				i;
-	struct acpi_table_sdt		sdt = { 0 }; /* initialise sdt.count */
-
-	rsdp_phys = acpi_find_rsdp();
-	rsdp = __va(rsdp_phys);
-	if (rsdp->rsdt_address) {
-		struct acpi_table_rsdt	*mapped_rsdt = NULL;
-		sdt.pa = rsdp->rsdt_address;
-
-		header = (struct acpi_table_header *)
-			__acpi_map_table(sdt.pa, sizeof(struct acpi_table_header));
-		if (!header)
-			return -ENODEV;
-
-		sdt.count = (header->length - sizeof(struct acpi_table_header)) >> 3;
-		mapped_rsdt = (struct acpi_table_rsdt *)
-			__acpi_map_table(sdt.pa, header->length);
-		if (!mapped_rsdt)
-			return -ENODEV;
-
-		header = &mapped_rsdt->header;
-
-		for (i = 0; i < sdt.count; i++)
-			sdt.entry[i].pa = (unsigned long) mapped_rsdt->entry[i];
-	};
-	for (i = 0; i < sdt.count; i++) {
-
-		header = (struct acpi_table_header *)
-			__acpi_map_table(sdt.entry[i].pa,
-				sizeof(struct acpi_table_header));
-		if (!header)
-			continue;
-		if (!strncmp((char *) &header->signature, "OEM1", 4)) {
-			if (!strncmp((char *) &header->oem_id, "UNISYS", 6)) {
-				void *addr;
-				struct oem_table *t;
-				acpi_table_print(header, sdt.entry[i].pa);
-				t = (struct oem_table *) __acpi_map_table(sdt.entry[i].pa, header->length);
-				addr = (void *) __acpi_map_table(t->OEMTableAddr, t->OEMTableSize);
-				*oem_addr = (unsigned long) addr;
-				return 0;
-			}
-		}
-	}
-	return -1;
-}

xen/arch/x86/genapic/probe.c

@@ -16,7 +16,6 @@
 
 extern struct genapic apic_summit;
 extern struct genapic apic_bigsmp;
-extern struct genapic apic_es7000;
 extern struct genapic apic_default;
 
 struct genapic *genapic;
@@ -24,7 +23,6 @@ struct genapic *genapic;
 struct genapic *apic_probe[] __initdata = { 
 	&apic_summit,
 	&apic_bigsmp, 
-	&apic_es7000,
 	&apic_default,	/* must be last */
 	NULL,
 };

xen/arch/x86/hvm/irq.c

@@ -395,9 +395,33 @@ static void irq_dump(struct domain *d)
 static int irq_save_pci(struct domain *d, hvm_domain_context_t *h)
 {
     struct hvm_irq *hvm_irq = &d->arch.hvm_domain.irq;
+    unsigned int asserted, pdev, pintx;
+    int rc;
+
+    spin_lock(&d->arch.hvm_domain.irq_lock);
+
+    pdev  = hvm_irq->callback_via.pci.dev;
+    pintx = hvm_irq->callback_via.pci.intx;
+    asserted = (hvm_irq->callback_via_asserted &&
+                (hvm_irq->callback_via_type == HVMIRQ_callback_pci_intx));
+
+    /*
+     * Deassert virtual interrupt via PCI INTx line. The virtual interrupt
+     * status is not save/restored, so the INTx line must be deasserted in
+     * the restore context.
+     */
+    if ( asserted )
+        __hvm_pci_intx_deassert(d, pdev, pintx);
 
     /* Save PCI IRQ lines */
-    return ( hvm_save_entry(PCI_IRQ, 0, h, &hvm_irq->pci_intx) );
+    rc = hvm_save_entry(PCI_IRQ, 0, h, &hvm_irq->pci_intx);
+
+    if ( asserted )
+        __hvm_pci_intx_assert(d, pdev, pintx);    
+
+    spin_unlock(&d->arch.hvm_domain.irq_lock);
+
+    return rc;
 }
 
 static int irq_save_isa(struct domain *d, hvm_domain_context_t *h)

xen/arch/x86/hvm/svm/svm.c

@@ -2159,6 +2159,7 @@ static void svm_do_msr_access(
         case MSR_K8_MC2_STATUS:
         case MSR_K8_MC3_STATUS:
         case MSR_K8_MC4_STATUS:
+        case MSR_K8_MC5_STATUS:
             /* No point in letting the guest see real MCEs */
             msr_content = 0;
             break;

xen/arch/x86/hvm/vmx/vmx.c

@@ -2596,6 +2596,7 @@ static int vmx_do_msr_read(struct cpu_user_regs *regs)
     case MSR_K8_MC2_STATUS:
     case MSR_K8_MC3_STATUS:
     case MSR_K8_MC4_STATUS:
+    case MSR_K8_MC5_STATUS:
         /* No point in letting the guest see real MCEs */
         msr_content = 0;
         break;

xen/arch/x86/mm/hap/guest_walk.c

@@ -84,7 +84,7 @@ unsigned long hap_gva_to_gfn(GUEST_PAGING_LEVELS)(
         mfn = get_mfn_from_gpfn(gpfn);
         if ( mfn == INVALID_MFN )
         {
-            HAP_PRINTK("bad pfn=0x%lx from gva=0x%lx at lev%d\n", gpfn, gva, 
+            HAP_PRINTK("bad pfn=0x%lx from gva=0x%lx at lev%d\n", gpfn, gva,
                        lev);
             success = 0;
             break;

xen/arch/x86/mm/hap/hap.c

@@ -73,7 +73,7 @@ int hap_disable_log_dirty(struct domain *d)
     hap_unlock(d);
 
     /* set l1e entries of P2M table with normal mode */
-    p2m_set_flags_global(d, __PAGE_HYPERVISOR|_PAGE_USER);    
+    p2m_set_flags_global(d, __PAGE_HYPERVISOR|_PAGE_USER);
     return 0;
 }
 
@@ -111,7 +111,7 @@ static struct page_info *hap_alloc(struct domain *d)
 
 static void hap_free(struct domain *d, mfn_t mfn)
 {
-    struct page_info *pg = mfn_to_page(mfn); 
+    struct page_info *pg = mfn_to_page(mfn);
 
     ASSERT(hap_locked_by_me(d));
 
@@ -128,7 +128,7 @@ static struct page_info *hap_alloc_p2m_page(struct domain *d)
 
 #if CONFIG_PAGING_LEVELS == 3
     /* Under PAE mode, top-level P2M table should be allocated below 4GB space
-     * because the size of h_cr3 is only 32-bit. We use alloc_domheap_pages to 
+     * because the size of h_cr3 is only 32-bit. We use alloc_domheap_pages to
      * force this requirement, and exchange the guaranteed 32-bit-clean
      * page for the one we just hap_alloc()ed. */
     if ( d->arch.paging.hap.p2m_pages == 0
@@ -166,9 +166,9 @@ void hap_free_p2m_page(struct domain *d, struct page_info *pg)
         HAP_ERROR("Odd p2m page count c=%#x t=%"PRtype_info"\n",
                   pg->count_info, pg->u.inuse.type_info);
     pg->count_info = 0;
-    /* Free should not decrement domain's total allocation, since 
+    /* Free should not decrement domain's total allocation, since
      * these pages were allocated without an owner. */
-    page_set_owner(pg, NULL); 
+    page_set_owner(pg, NULL);
     free_domheap_page(pg);
     d->arch.paging.hap.p2m_pages--;
     ASSERT(d->arch.paging.hap.p2m_pages >= 0);
@@ -221,7 +221,7 @@ hap_set_allocation(struct domain *d, unsigned int pages, int *preempted)
             pg->count_info = 0;
             free_domheap_page(pg);
         }
-        
+
         /* Check to see if we need to yield and try again */
         if ( preempted && hypercall_preempt_check() )
         {
@@ -275,7 +275,7 @@ static void hap_install_xen_entries_in_l2h(struct vcpu *v, mfn_t l2hmfn)
 
     l2e = hap_map_domain_page(l2hmfn);
     ASSERT(l2e != NULL);
-    
+
     /* Copy the common Xen mappings from the idle domain */
     memcpy(&l2e[L2_PAGETABLE_FIRST_XEN_SLOT & (L2_PAGETABLE_ENTRIES-1)],
            &idle_pg_table_l2[L2_PAGETABLE_FIRST_XEN_SLOT],
@@ -318,7 +318,7 @@ static void hap_install_xen_entries_in_l2(struct vcpu *v, mfn_t l2mfn)
 
     l2e = hap_map_domain_page(l2mfn);
     ASSERT(l2e != NULL);
-    
+
     /* Copy the common Xen mappings from the idle domain */
     memcpy(&l2e[L2_PAGETABLE_FIRST_XEN_SLOT],
            &idle_pg_table[L2_PAGETABLE_FIRST_XEN_SLOT],
@@ -362,7 +362,7 @@ static mfn_t hap_make_monitor_table(struct vcpu *v)
     }
 #elif CONFIG_PAGING_LEVELS == 3
     {
-        mfn_t m3mfn, m2mfn; 
+        mfn_t m3mfn, m2mfn;
         l3_pgentry_t *l3e;
         l2_pgentry_t *l2e;
         int i;
@@ -384,8 +384,8 @@ static mfn_t hap_make_monitor_table(struct vcpu *v)
         l2e = hap_map_domain_page(m2mfn);
         for ( i = 0; i < L3_PAGETABLE_ENTRIES; i++ )
             l2e[l2_table_offset(LINEAR_PT_VIRT_START) + i] =
-                (l3e_get_flags(l3e[i]) & _PAGE_PRESENT) 
-                ? l2e_from_pfn(l3e_get_pfn(l3e[i]), __PAGE_HYPERVISOR) 
+                (l3e_get_flags(l3e[i]) & _PAGE_PRESENT)
+                ? l2e_from_pfn(l3e_get_pfn(l3e[i]), __PAGE_HYPERVISOR)
                 : l2e_empty();
         hap_unmap_domain_page(l2e);
         hap_unmap_domain_page(l3e);
@@ -536,7 +536,7 @@ void hap_teardown(struct domain *d)
                       d->arch.paging.hap.p2m_pages);
         ASSERT(d->arch.paging.hap.total_pages == 0);
     }
-    
+
     d->arch.paging.mode &= ~PG_log_dirty;
 
     hap_unlock(d);
@@ -555,7 +555,7 @@ int hap_domctl(struct domain *d, xen_domctl_shadow_op_t *sc,
         hap_unlock(d);
         if ( preempted )
             /* Not finished.  Set up to re-run the call. */
-            rc = hypercall_create_continuation(__HYPERVISOR_domctl, "h", 
+            rc = hypercall_create_continuation(__HYPERVISOR_domctl, "h",
                                                u_domctl);
         else
             /* Finished.  Return the new allocation */
@@ -578,11 +578,11 @@ void hap_vcpu_init(struct vcpu *v)
 /************************************************/
 /*          HAP PAGING MODE FUNCTIONS           */
 /************************************************/
-/* 
+/*
  * HAP guests can handle page faults (in the guest page tables) without
  * needing any action from Xen, so we should not be intercepting them.
  */
-static int hap_page_fault(struct vcpu *v, unsigned long va, 
+static int hap_page_fault(struct vcpu *v, unsigned long va,
                           struct cpu_user_regs *regs)
 {
     HAP_ERROR("Intercepted a guest #PF (%u:%u) with HAP enabled.\n",
@@ -591,9 +591,9 @@ static int hap_page_fault(struct vcpu *v, unsigned long va,
     return 0;
 }
 
-/* 
+/*
  * HAP guests can handle invlpg without needing any action from Xen, so
- * should not be intercepting it. 
+ * should not be intercepting it.
  */
 static int hap_invlpg(struct vcpu *v, unsigned long va)
 {
@@ -649,7 +649,7 @@ static void hap_update_paging_modes(struct vcpu *v)
 }
 
 #if CONFIG_PAGING_LEVELS == 3
-static void p2m_install_entry_in_monitors(struct domain *d, l3_pgentry_t *l3e) 
+static void p2m_install_entry_in_monitors(struct domain *d, l3_pgentry_t *l3e)
 /* Special case, only used for PAE hosts: update the mapping of the p2m
  * table.  This is trivial in other paging modes (one top-level entry
  * points to the top-level p2m, no maintenance needed), but PAE makes
@@ -660,13 +660,13 @@ static void p2m_install_entry_in_monitors(struct domain *d, l3_pgentry_t *l3e)
     l2_pgentry_t *ml2e;
     struct vcpu *v;
     unsigned int index;
-    
+
     index = ((unsigned long)l3e & ~PAGE_MASK) / sizeof(l3_pgentry_t);
     ASSERT(index < MACHPHYS_MBYTES>>1);
-    
+
     for_each_vcpu ( d, v )
     {
-        if ( pagetable_get_pfn(v->arch.monitor_table) == 0 ) 
+        if ( pagetable_get_pfn(v->arch.monitor_table) == 0 )
             continue;
 
         ASSERT(paging_mode_external(v->domain));
@@ -689,7 +689,7 @@ static void p2m_install_entry_in_monitors(struct domain *d, l3_pgentry_t *l3e)
 }
 #endif
 
-static void 
+static void
 hap_write_p2m_entry(struct vcpu *v, unsigned long gfn, l1_pgentry_t *p,
                     mfn_t table_mfn, l1_pgentry_t new, unsigned int level)
 {
@@ -698,12 +698,12 @@ hap_write_p2m_entry(struct vcpu *v, unsigned long gfn, l1_pgentry_t *p,
     safe_write_pte(p, new);
 #if CONFIG_PAGING_LEVELS == 3
     /* install P2M in monitor table for PAE Xen */
-    if ( level == 3 ) 
+    if ( level == 3 )
         /* We have written to the p2m l3: need to sync the per-vcpu
          * copies of it in the monitor tables */
         p2m_install_entry_in_monitors(v->domain, (l3_pgentry_t *)p);
 #endif
-    
+
     hap_unlock(v->domain);
 }
 
@@ -715,7 +715,7 @@ static unsigned long hap_gva_to_gfn_real_mode(
 
 /* Entry points into this mode of the hap code. */
 struct paging_mode hap_paging_real_mode = {
-    .page_fault             = hap_page_fault, 
+    .page_fault             = hap_page_fault,
     .invlpg                 = hap_invlpg,
     .gva_to_gfn             = hap_gva_to_gfn_real_mode,
     .update_cr3             = hap_update_cr3,
@@ -725,7 +725,7 @@ struct paging_mode hap_paging_real_mode = {
 };
 
 struct paging_mode hap_paging_protected_mode = {
-    .page_fault             = hap_page_fault, 
+    .page_fault             = hap_page_fault,
     .invlpg                 = hap_invlpg,
     .gva_to_gfn             = hap_gva_to_gfn_2level,
     .update_cr3             = hap_update_cr3,
@@ -735,7 +735,7 @@ struct paging_mode hap_paging_protected_mode = {
 };
 
 struct paging_mode hap_paging_pae_mode = {
-    .page_fault             = hap_page_fault, 
+    .page_fault             = hap_page_fault,
     .invlpg                 = hap_invlpg,
     .gva_to_gfn             = hap_gva_to_gfn_3level,
     .update_cr3             = hap_update_cr3,
@@ -745,7 +745,7 @@ struct paging_mode hap_paging_pae_mode = {
 };
 
 struct paging_mode hap_paging_long_mode = {
-    .page_fault             = hap_page_fault, 
+    .page_fault             = hap_page_fault,
     .invlpg                 = hap_invlpg,
     .gva_to_gfn             = hap_gva_to_gfn_4level,
     .update_cr3             = hap_update_cr3,

xen/arch/x86/mm/p2m.c

@@ -2,12 +2,12 @@
  * arch/x86/mm/p2m.c
  *
  * physical-to-machine mappings for automatically-translated domains.
- * 
+ *
  * Parts of this code are Copyright (c) 2007 by Advanced Micro Devices.
  * Parts of this code are Copyright (c) 2006 by XenSource Inc.
  * Parts of this code are Copyright (c) 2006 by Michael A Fetterman
  * Parts based on earlier work by Michael A Fetterman, Ian Pratt et al.
- * 
+ *
  * This program is free software; you can redistribute it and/or modify
  * it under the terms of the GNU General Public License as published by
  * the Free Software Foundation; either version 2 of the License, or
@@ -34,7 +34,7 @@
 
 /*
  * The P2M lock.  This protects all updates to the p2m table.
- * Updates are expected to be safe against concurrent reads, 
+ * Updates are expected to be safe against concurrent reads,
  * which do *not* require the lock.
  *
  * Locking discipline: always acquire this lock before the shadow or HAP one
@@ -80,7 +80,7 @@
 #define P2M_DEBUG(_f, _a...)                                 \
     debugtrace_printk("p2mdebug: %s(): " _f, __func__, ##_a)
 #else
-#define P2M_DEBUG(_f, _a...) do { (void)(_f); } while(0) 
+#define P2M_DEBUG(_f, _a...) do { (void)(_f); } while(0)
 #endif
 
 
@@ -119,8 +119,8 @@ p2m_find_entry(void *table, unsigned long *gfn_remainder,
 // Returns 0 on error.
 //
 static int
-p2m_next_level(struct domain *d, mfn_t *table_mfn, void **table, 
-               unsigned long *gfn_remainder, unsigned long gfn, u32 shift, 
+p2m_next_level(struct domain *d, mfn_t *table_mfn, void **table,
+               unsigned long *gfn_remainder, unsigned long gfn, u32 shift,
                u32 max, unsigned long type)
 {
     l1_pgentry_t *p2m_entry;
@@ -146,7 +146,7 @@ p2m_next_level(struct domain *d, mfn_t *table_mfn, void **table,
 
         switch ( type ) {
         case PGT_l3_page_table:
-            paging_write_p2m_entry(d, gfn, 
+            paging_write_p2m_entry(d, gfn,
                                    p2m_entry, *table_mfn, new_entry, 4);
             break;
         case PGT_l2_page_table:
@@ -154,11 +154,11 @@ p2m_next_level(struct domain *d, mfn_t *table_mfn, void **table,
             /* for PAE mode, PDPE only has PCD/PWT/P bits available */
             new_entry = l1e_from_pfn(mfn_x(page_to_mfn(pg)), _PAGE_PRESENT);
 #endif
-            paging_write_p2m_entry(d, gfn, 
+            paging_write_p2m_entry(d, gfn,
                                    p2m_entry, *table_mfn, new_entry, 3);
             break;
         case PGT_l1_page_table:
-            paging_write_p2m_entry(d, gfn, 
+            paging_write_p2m_entry(d, gfn,
                                    p2m_entry, *table_mfn, new_entry, 2);
             break;
         default:
@@ -216,7 +216,7 @@ set_p2m_entry(struct domain *d, unsigned long gfn, mfn_t mfn, u32 l1e_flags)
     ASSERT(p2m_entry);
 
     /* Track the highest gfn for which we have ever had a valid mapping */
-    if ( mfn_valid(mfn) && (gfn > d->arch.p2m.max_mapped_pfn) ) 
+    if ( mfn_valid(mfn) && (gfn > d->arch.p2m.max_mapped_pfn) )
         d->arch.p2m.max_mapped_pfn = gfn;
 
     if ( mfn_valid(mfn) )
@@ -229,7 +229,7 @@ set_p2m_entry(struct domain *d, unsigned long gfn, mfn_t mfn, u32 l1e_flags)
 
     /* Success */
     rv = 1;
- 
+
  out:
     unmap_domain_page(table);
     return rv;
@@ -250,7 +250,7 @@ void p2m_init(struct domain *d)
 // controlled by CONFIG_PAGING_LEVELS).
 //
 // The alloc_page and free_page functions will be used to get memory to
-// build the p2m, and to release it again at the end of day. 
+// build the p2m, and to release it again at the end of day.
 //
 // Returns 0 for success or -errno.
 //
@@ -264,7 +264,7 @@ int p2m_alloc_table(struct domain *d,
     struct page_info *page, *p2m_top;
     unsigned int page_count = 0;
     unsigned long gfn;
-    
+
     p2m_lock(d);
 
     if ( pagetable_get_pfn(d->arch.phys_table) != 0 )
@@ -288,7 +288,7 @@ int p2m_alloc_table(struct domain *d,
     list_add_tail(&p2m_top->list, &d->arch.p2m.pages);
 
     p2m_top->count_info = 1;
-    p2m_top->u.inuse.type_info = 
+    p2m_top->u.inuse.type_info =
 #if CONFIG_PAGING_LEVELS == 4
         PGT_l4_page_table
 #elif CONFIG_PAGING_LEVELS == 3
@@ -301,7 +301,7 @@ int p2m_alloc_table(struct domain *d,
     d->arch.phys_table = pagetable_from_mfn(page_to_mfn(p2m_top));
 
     P2M_PRINTK("populating p2m table\n");
- 
+
     /* Initialise physmap tables for slot zero. Other code assumes this. */
     gfn = 0;
     mfn = _mfn(INVALID_MFN);
@@ -365,17 +365,17 @@ gfn_to_mfn_foreign(struct domain *d, unsigned long gpfn)
     paddr_t addr = ((paddr_t)gpfn) << PAGE_SHIFT;
     l2_pgentry_t *l2e;
     l1_pgentry_t *l1e;
-    
+
     ASSERT(paging_mode_translate(d));
     mfn = pagetable_get_mfn(d->arch.phys_table);
 
 
-    if ( gpfn > d->arch.p2m.max_mapped_pfn ) 
+    if ( gpfn > d->arch.p2m.max_mapped_pfn )
         /* This pfn is higher than the highest the p2m map currently holds */
         return _mfn(INVALID_MFN);
 
 #if CONFIG_PAGING_LEVELS >= 4
-    { 
+    {
         l4_pgentry_t *l4e = map_domain_page(mfn_x(mfn));
         l4e += l4_table_offset(addr);
         if ( (l4e_get_flags(*l4e) & _PAGE_PRESENT) == 0 )
@@ -398,7 +398,7 @@ gfn_to_mfn_foreign(struct domain *d, unsigned long gpfn)
          * the bounds of the p2m. */
         l3e += (addr >> L3_PAGETABLE_SHIFT);
 #else
-        l3e += l3_table_offset(addr);        
+        l3e += l3_table_offset(addr);
 #endif
         if ( (l3e_get_flags(*l3e) & _PAGE_PRESENT) == 0 )
         {
@@ -443,18 +443,18 @@ static void audit_p2m(struct domain *d)
     mfn_t p2mfn;
     unsigned long orphans_d = 0, orphans_i = 0, mpbad = 0, pmbad = 0;
     int test_linear;
-    
+
     if ( !paging_mode_translate(d) )
         return;
 
     //P2M_PRINTK("p2m audit starts\n");
 
-    test_linear = ( (d == current->domain) 
+    test_linear = ( (d == current->domain)
                     && !pagetable_is_null(current->arch.monitor_table) );
     if ( test_linear )
-        local_flush_tlb(); 
+        local_flush_tlb();
 
-    /* Audit part one: walk the domain's page allocation list, checking 
+    /* Audit part one: walk the domain's page allocation list, checking
      * the m2p entries. */
     for ( entry = d->page_list.next;
           entry != &d->page_list;
@@ -463,11 +463,11 @@ static void audit_p2m(struct domain *d)
         page = list_entry(entry, struct page_info, list);
         mfn = mfn_x(page_to_mfn(page));
 
-        // P2M_PRINTK("auditing guest page, mfn=%#lx\n", mfn); 
+        // P2M_PRINTK("auditing guest page, mfn=%#lx\n", mfn);
 
         od = page_get_owner(page);
 
-        if ( od != d ) 
+        if ( od != d )
         {
             P2M_PRINTK("wrong owner %#lx -> %p(%u) != %p(%u)\n",
                        mfn, od, (od?od->domain_id:-1), d, d->domain_id);
@@ -475,19 +475,19 @@ static void audit_p2m(struct domain *d)
         }
 
         gfn = get_gpfn_from_mfn(mfn);
-        if ( gfn == INVALID_M2P_ENTRY ) 
+        if ( gfn == INVALID_M2P_ENTRY )
         {
             orphans_i++;
             //P2M_PRINTK("orphaned guest page: mfn=%#lx has invalid gfn\n",
-            //               mfn); 
+            //               mfn);
             continue;
         }
 
-        if ( gfn == 0x55555555 ) 
+        if ( gfn == 0x55555555 )
         {
             orphans_d++;
-            //P2M_PRINTK("orphaned guest page: mfn=%#lx has debug gfn\n", 
-            //               mfn); 
+            //P2M_PRINTK("orphaned guest page: mfn=%#lx has debug gfn\n",
+            //               mfn);
             continue;
         }
 
@@ -503,7 +503,7 @@ static void audit_p2m(struct domain *d)
                         : -1u));
             /* This m2p entry is stale: the domain has another frame in
              * this physical slot.  No great disaster, but for neatness,
-             * blow away the m2p entry. */ 
+             * blow away the m2p entry. */
             set_gpfn_from_mfn(mfn, INVALID_M2P_ENTRY, __PAGE_HYPERVISOR|_PAGE_USER);
         }
 
@@ -517,9 +517,9 @@ static void audit_p2m(struct domain *d)
             }
         }
 
-        // P2M_PRINTK("OK: mfn=%#lx, gfn=%#lx, p2mfn=%#lx, lp2mfn=%#lx\n", 
-        //                mfn, gfn, p2mfn, lp2mfn); 
-    }   
+        // P2M_PRINTK("OK: mfn=%#lx, gfn=%#lx, p2mfn=%#lx, lp2mfn=%#lx\n",
+        //                mfn, gfn, p2mfn, lp2mfn);
+    }
 
     /* Audit part two: walk the domain's p2m table, checking the entries. */
     if ( pagetable_get_pfn(d->arch.phys_table) != 0 )
@@ -527,7 +527,7 @@ static void audit_p2m(struct domain *d)
         l2_pgentry_t *l2e;
         l1_pgentry_t *l1e;
         int i1, i2;
-        
+
 #if CONFIG_PAGING_LEVELS == 4
         l4_pgentry_t *l4e;
         l3_pgentry_t *l3e;
@@ -553,8 +553,8 @@ static void audit_p2m(struct domain *d)
             }
             l3e = map_domain_page(mfn_x(_mfn(l4e_get_pfn(l4e[i4]))));
 #endif /* now at levels 3 or 4... */
-            for ( i3 = 0; 
-                  i3 < ((CONFIG_PAGING_LEVELS==4) ? L3_PAGETABLE_ENTRIES : 8); 
+            for ( i3 = 0;
+                  i3 < ((CONFIG_PAGING_LEVELS==4) ? L3_PAGETABLE_ENTRIES : 8);
                   i3++ )
             {
                 if ( !(l3e_get_flags(l3e[i3]) & _PAGE_PRESENT) )
@@ -572,7 +572,7 @@ static void audit_p2m(struct domain *d)
                         continue;
                     }
                     l1e = map_domain_page(mfn_x(_mfn(l2e_get_pfn(l2e[i2]))));
-                    
+
                     for ( i1 = 0; i1 < L1_PAGETABLE_ENTRIES; i1++, gfn++ )
                     {
                         if ( !(l1e_get_flags(l1e[i1]) & _PAGE_PRESENT) )
@@ -610,14 +610,14 @@ static void audit_p2m(struct domain *d)
     }
 
     //P2M_PRINTK("p2m audit complete\n");
-    //if ( orphans_i | orphans_d | mpbad | pmbad ) 
+    //if ( orphans_i | orphans_d | mpbad | pmbad )
     //    P2M_PRINTK("p2m audit found %lu orphans (%lu inval %lu debug)\n",
     //                   orphans_i + orphans_d, orphans_i, orphans_d,
-    if ( mpbad | pmbad ) 
+    if ( mpbad | pmbad )
         P2M_PRINTK("p2m audit found %lu odd p2m, %lu bad m2p entries\n",
                    pmbad, mpbad);
 }
-#else 
+#else
 #define audit_p2m(_d) do { (void)(_d); } while(0)
 #endif /* P2M_AUDIT */
 
@@ -645,7 +645,7 @@ guest_physmap_remove_page(struct domain *d, unsigned long gfn,
     audit_p2m(d);
     p2m_remove_page(d, gfn, mfn);
     audit_p2m(d);
-    p2m_unlock(d);    
+    p2m_unlock(d);
 }
 
 void
@@ -683,11 +683,11 @@ guest_physmap_add_page(struct domain *d, unsigned long gfn,
         /* This machine frame is already mapped at another physical address */
         P2M_DEBUG("aliased! mfn=%#lx, old gfn=%#lx, new gfn=%#lx\n",
                   mfn, ogfn, gfn);
-        if ( mfn_valid(omfn = gfn_to_mfn(d, ogfn)) ) 
+        if ( mfn_valid(omfn = gfn_to_mfn(d, ogfn)) )
         {
-            P2M_DEBUG("old gfn=%#lx -> mfn %#lx\n", 
+            P2M_DEBUG("old gfn=%#lx -> mfn %#lx\n",
                       ogfn , mfn_x(omfn));
-            if ( mfn_x(omfn) == mfn ) 
+            if ( mfn_x(omfn) == mfn )
                 p2m_remove_page(d, ogfn, mfn);
         }
     }
@@ -720,15 +720,15 @@ void p2m_set_flags_global(struct domain *d, u32 l1e_flags)
     int i4;
 #endif /* CONFIG_PAGING_LEVELS == 4 */
 #endif /* CONFIG_PAGING_LEVELS >= 3 */
-    
+
     if ( !paging_mode_translate(d) )
         return;
- 
+
     if ( pagetable_get_pfn(d->arch.phys_table) == 0 )
         return;
 
     p2m_lock(d);
-        
+
 #if CONFIG_PAGING_LEVELS == 4
     l4e = map_domain_page(mfn_x(pagetable_get_mfn(d->arch.phys_table)));
 #elif CONFIG_PAGING_LEVELS == 3
@@ -739,52 +739,52 @@ void p2m_set_flags_global(struct domain *d, u32 l1e_flags)
 
 #if CONFIG_PAGING_LEVELS >= 3
 #if CONFIG_PAGING_LEVELS >= 4
-    for ( i4 = 0; i4 < L4_PAGETABLE_ENTRIES; i4++ ) 
+    for ( i4 = 0; i4 < L4_PAGETABLE_ENTRIES; i4++ )
     {
-	if ( !(l4e_get_flags(l4e[i4]) & _PAGE_PRESENT) )
-	{
-	    continue;
-	}
-	l3e = map_domain_page(l4e_get_pfn(l4e[i4]));
+        if ( !(l4e_get_flags(l4e[i4]) & _PAGE_PRESENT) )
+        {
+            continue;
+        }
+        l3e = map_domain_page(l4e_get_pfn(l4e[i4]));
 #endif /* now at levels 3 or 4... */
-	for ( i3 = 0; 
-	      i3 < ((CONFIG_PAGING_LEVELS==4) ? L3_PAGETABLE_ENTRIES : 8); 
-	      i3++ )
-	{
-	    if ( !(l3e_get_flags(l3e[i3]) & _PAGE_PRESENT) )
-	    {
-		continue;
-	    }
-	    l2e = map_domain_page(l3e_get_pfn(l3e[i3]));
+        for ( i3 = 0;
+              i3 < ((CONFIG_PAGING_LEVELS==4) ? L3_PAGETABLE_ENTRIES : 8);
+              i3++ )
+        {
+            if ( !(l3e_get_flags(l3e[i3]) & _PAGE_PRESENT) )
+            {
+                continue;
+            }
+            l2e = map_domain_page(l3e_get_pfn(l3e[i3]));
 #endif /* all levels... */
-	    for ( i2 = 0; i2 < L2_PAGETABLE_ENTRIES; i2++ )
-	    {
-		if ( !(l2e_get_flags(l2e[i2]) & _PAGE_PRESENT) )
-		{
-		    continue;
-		}
+            for ( i2 = 0; i2 < L2_PAGETABLE_ENTRIES; i2++ )
+            {
+                if ( !(l2e_get_flags(l2e[i2]) & _PAGE_PRESENT) )
+                {
+                    continue;
+                }
 
                 l1mfn = _mfn(l2e_get_pfn(l2e[i2]));
-		l1e = map_domain_page(mfn_x(l1mfn));
-		
-		for ( i1 = 0; i1 < L1_PAGETABLE_ENTRIES; i1++, gfn++ )
-		{
-		    if ( !(l1e_get_flags(l1e[i1]) & _PAGE_PRESENT) )
-			continue;
-		    mfn = l1e_get_pfn(l1e[i1]);
-		    gfn = get_gpfn_from_mfn(mfn);
-		    /* create a new 1le entry using l1e_flags */
-		    l1e_content = l1e_from_pfn(mfn, l1e_flags);
-		    paging_write_p2m_entry(d, gfn, &l1e[i1], 
+                l1e = map_domain_page(mfn_x(l1mfn));
+
+                for ( i1 = 0; i1 < L1_PAGETABLE_ENTRIES; i1++, gfn++ )
+                {
+                    if ( !(l1e_get_flags(l1e[i1]) & _PAGE_PRESENT) )
+                        continue;
+                    mfn = l1e_get_pfn(l1e[i1]);
+                    gfn = get_gpfn_from_mfn(mfn);
+                    /* create a new 1le entry using l1e_flags */
+                    l1e_content = l1e_from_pfn(mfn, l1e_flags);
+                    paging_write_p2m_entry(d, gfn, &l1e[i1],
                                            l1mfn, l1e_content, 1);
-		}
-		unmap_domain_page(l1e);
-	    }
+                }
+                unmap_domain_page(l1e);
+            }
 #if CONFIG_PAGING_LEVELS >= 3
-	    unmap_domain_page(l2e);
-	}
+            unmap_domain_page(l2e);
+        }
 #if CONFIG_PAGING_LEVELS >= 4
-	unmap_domain_page(l3e);
+        unmap_domain_page(l3e);
     }
 #endif
 #endif
@@ -814,7 +814,7 @@ int p2m_set_flags(struct domain *d, paddr_t gpa, u32 l1e_flags)
     mfn = gfn_to_mfn(d, gfn);
     if ( mfn_valid(mfn) )
         set_p2m_entry(d, gfn, mfn, l1e_flags);
-    
+
     p2m_unlock(d);
 
     return 1;

xen/arch/x86/mm/paging.c

@@ -54,10 +54,10 @@ boolean_param("hap", opt_hap_enabled);
 #define page_to_mfn(_pg) (_mfn((_pg) - frame_table))
 
 /* The log-dirty lock.  This protects the log-dirty bitmap from
- * concurrent accesses (and teardowns, etc). 
- * 
+ * concurrent accesses (and teardowns, etc).
+ *
  * Locking discipline: always acquire shadow or HAP lock before this one.
- * 
+ *
  * Because mark_dirty is called from a lot of places, the log-dirty lock
  * may be acquired with the shadow or HAP locks already held.  When the
  * log-dirty code makes callbacks into HAP or shadow code to reset
@@ -105,7 +105,7 @@ int paging_alloc_log_dirty_bitmap(struct domain *d)
 
     d->arch.paging.log_dirty.bitmap_size =
         (domain_get_maximum_gpfn(d) + BITS_PER_LONG) & ~(BITS_PER_LONG - 1);
-    d->arch.paging.log_dirty.bitmap = 
+    d->arch.paging.log_dirty.bitmap =
         xmalloc_array(unsigned long,
                       d->arch.paging.log_dirty.bitmap_size / BITS_PER_LONG);
     if ( d->arch.paging.log_dirty.bitmap == NULL )
@@ -152,8 +152,8 @@ int paging_log_dirty_enable(struct domain *d)
 
     log_dirty_unlock(d);
 
-    /* Safe because the domain is paused. */    
-    ret = d->arch.paging.log_dirty.enable_log_dirty(d);    
+    /* Safe because the domain is paused. */
+    ret = d->arch.paging.log_dirty.enable_log_dirty(d);
 
     /* Possibility of leaving the bitmap allocated here but it'll be
      * tidied on domain teardown. */
@@ -202,7 +202,7 @@ void paging_mark_dirty(struct domain *d, unsigned long guest_mfn)
     pfn = get_gpfn_from_mfn(mfn_x(gmfn));
 
     /*
-     * Values with the MSB set denote MFNs that aren't really part of the 
+     * Values with the MSB set denote MFNs that aren't really part of the
      * domain's pseudo-physical memory map (e.g., the shared info frame).
      * Nothing to do here...
      */
@@ -212,11 +212,11 @@ void paging_mark_dirty(struct domain *d, unsigned long guest_mfn)
         return;
     }
 
-    if ( likely(pfn < d->arch.paging.log_dirty.bitmap_size) ) 
-    { 
+    if ( likely(pfn < d->arch.paging.log_dirty.bitmap_size) )
+    {
         if ( !__test_and_set_bit(pfn, d->arch.paging.log_dirty.bitmap) )
         {
-            PAGING_DEBUG(LOGDIRTY, 
+            PAGING_DEBUG(LOGDIRTY,
                          "marked mfn %" PRI_mfn " (pfn=%lx), dom %d\n",
                          mfn_x(gmfn), pfn, d->domain_id);
             d->arch.paging.log_dirty.dirty_count++;
@@ -227,21 +227,21 @@ void paging_mark_dirty(struct domain *d, unsigned long guest_mfn)
         PAGING_PRINTK("mark_dirty OOR! "
                       "mfn=%" PRI_mfn " pfn=%lx max=%x (dom %d)\n"
                       "owner=%d c=%08x t=%" PRtype_info "\n",
-                      mfn_x(gmfn), 
-                      pfn, 
+                      mfn_x(gmfn),
+                      pfn,
                       d->arch.paging.log_dirty.bitmap_size,
                       d->domain_id,
                       (page_get_owner(mfn_to_page(gmfn))
                        ? page_get_owner(mfn_to_page(gmfn))->domain_id
                        : -1),
-                      mfn_to_page(gmfn)->count_info, 
+                      mfn_to_page(gmfn)->count_info,
                       mfn_to_page(gmfn)->u.inuse.type_info);
     }
-    
+
     log_dirty_unlock(d);
 }
 
-/* Read a domain's log-dirty bitmap and stats.  If the operation is a CLEAN, 
+/* Read a domain's log-dirty bitmap and stats.  If the operation is a CLEAN,
  * clear the bitmap and stats as well. */
 int paging_log_dirty_op(struct domain *d, struct xen_domctl_shadow_op *sc)
 {
@@ -252,15 +252,15 @@ int paging_log_dirty_op(struct domain *d, struct xen_domctl_shadow_op *sc)
 
     clean = (sc->op == XEN_DOMCTL_SHADOW_OP_CLEAN);
 
-    PAGING_DEBUG(LOGDIRTY, "log-dirty %s: dom %u faults=%u dirty=%u\n", 
+    PAGING_DEBUG(LOGDIRTY, "log-dirty %s: dom %u faults=%u dirty=%u\n",
                  (clean) ? "clean" : "peek",
                  d->domain_id,
-                 d->arch.paging.log_dirty.fault_count, 
+                 d->arch.paging.log_dirty.fault_count,
                  d->arch.paging.log_dirty.dirty_count);
 
     sc->stats.fault_count = d->arch.paging.log_dirty.fault_count;
     sc->stats.dirty_count = d->arch.paging.log_dirty.dirty_count;
-    
+
     if ( clean )
     {
         d->arch.paging.log_dirty.fault_count = 0;
@@ -276,7 +276,7 @@ int paging_log_dirty_op(struct domain *d, struct xen_domctl_shadow_op *sc)
         rv = -EINVAL; /* perhaps should be ENOMEM? */
         goto out;
     }
- 
+
     if ( sc->pages > d->arch.paging.log_dirty.bitmap_size )
         sc->pages = d->arch.paging.log_dirty.bitmap_size;
 
@@ -322,11 +322,11 @@ int paging_log_dirty_op(struct domain *d, struct xen_domctl_shadow_op *sc)
 
 
 /* Note that this function takes three function pointers. Callers must supply
- * these functions for log dirty code to call. This function usually is 
- * invoked when paging is enabled. Check shadow_enable() and hap_enable() for 
+ * these functions for log dirty code to call. This function usually is
+ * invoked when paging is enabled. Check shadow_enable() and hap_enable() for
  * reference.
  *
- * These function pointers must not be followed with the log-dirty lock held. 
+ * These function pointers must not be followed with the log-dirty lock held.
  */
 void paging_log_dirty_init(struct domain *d,
                            int    (*enable_log_dirty)(struct domain *d),
@@ -335,7 +335,7 @@ void paging_log_dirty_init(struct domain *d,
 {
     /* We initialize log dirty lock first */
     log_dirty_lock_init(d);
-    
+
     d->arch.paging.log_dirty.enable_log_dirty = enable_log_dirty;
     d->arch.paging.log_dirty.disable_log_dirty = disable_log_dirty;
     d->arch.paging.log_dirty.clean_dirty_bitmap = clean_dirty_bitmap;
@@ -387,7 +387,7 @@ int paging_domctl(struct domain *d, xen_domctl_shadow_op_t *sc,
                  d->domain_id);
         return -EINVAL;
     }
-    
+
     if ( unlikely(d->is_dying) )
     {
         gdprintk(XENLOG_INFO, "Ignoring paging op on dying domain %u\n",
@@ -401,38 +401,38 @@ int paging_domctl(struct domain *d, xen_domctl_shadow_op_t *sc,
                      d->domain_id);
         return -EINVAL;
     }
-    
+
     /* Code to handle log-dirty. Note that some log dirty operations
-     * piggy-back on shadow operations. For example, when 
+     * piggy-back on shadow operations. For example, when
      * XEN_DOMCTL_SHADOW_OP_OFF is called, it first checks whether log dirty
-     * mode is enabled. If does, we disables log dirty and continues with 
-     * shadow code. For this reason, we need to further dispatch domctl 
+     * mode is enabled. If does, we disables log dirty and continues with
+     * shadow code. For this reason, we need to further dispatch domctl
      * to next-level paging code (shadow or hap).
      */
     switch ( sc->op )
     {
     case XEN_DOMCTL_SHADOW_OP_ENABLE_LOGDIRTY:
-        return paging_log_dirty_enable(d);	
-	
-    case XEN_DOMCTL_SHADOW_OP_ENABLE:	
+        return paging_log_dirty_enable(d);
+
+    case XEN_DOMCTL_SHADOW_OP_ENABLE:
         if ( sc->mode & XEN_DOMCTL_SHADOW_ENABLE_LOG_DIRTY )
             return paging_log_dirty_enable(d);
 
     case XEN_DOMCTL_SHADOW_OP_OFF:
         if ( paging_mode_log_dirty(d) )
-            if ( (rc = paging_log_dirty_disable(d)) != 0 ) 
+            if ( (rc = paging_log_dirty_disable(d)) != 0 )
                 return rc;
 
     case XEN_DOMCTL_SHADOW_OP_CLEAN:
     case XEN_DOMCTL_SHADOW_OP_PEEK:
-	return paging_log_dirty_op(d, sc);
+        return paging_log_dirty_op(d, sc);
     }
-	
+
     /* Here, dispatch domctl to the appropriate paging code */
     if ( opt_hap_enabled && is_hvm_domain(d) )
-	return hap_domctl(d, sc, u_domctl);
+        return hap_domctl(d, sc, u_domctl);
     else
-	return shadow_domctl(d, sc, u_domctl);
+        return shadow_domctl(d, sc, u_domctl);
 }
 
 /* Call when destroying a domain */
@@ -492,7 +492,7 @@ void paging_dump_vcpu_info(struct vcpu *v)
 {
     if ( paging_mode_enabled(v->domain) )
     {
-        printk("    paging assistance: ");        
+        printk("    paging assistance: ");
         if ( paging_mode_shadow(v->domain) )
         {
             if ( v->arch.paging.mode )
@@ -504,7 +504,7 @@ void paging_dump_vcpu_info(struct vcpu *v)
                 printk("not shadowed\n");
         }
         else if ( paging_mode_hap(v->domain) && v->arch.paging.mode )
-            printk("hap, %u levels\n", 
+            printk("hap, %u levels\n",
                    v->arch.paging.mode->guest_levels);
         else
             printk("none\n");

xen/common/compat/grant_table.c

@@ -161,7 +161,7 @@ int compat_grant_table_op(unsigned int cmd,
                 while ( n-- )
                 {
                     guest_handle_add_offset(xfer, -1);
-                    if ( __copy_field_to_guest(xfer, nat.xfer, status) )
+                    if ( __copy_field_to_guest(xfer, nat.xfer + n, status) )
                         rc = -EFAULT;
                 }
             }
@@ -199,7 +199,7 @@ int compat_grant_table_op(unsigned int cmd,
                 while ( n-- )
                 {
                     guest_handle_add_offset(copy, -1);
-                    if ( __copy_field_to_guest(copy, nat.copy, status) )
+                    if ( __copy_field_to_guest(copy, nat.copy + n, status) )
                         rc = -EFAULT;
                 }
             }

xen/common/libelf/libelf-dominfo.c

@@ -333,99 +333,6 @@ static int elf_xen_note_check(struct elf_binary *elf,
     return 0;
 }
 
-
-static void elf_xen_loadsymtab(struct elf_binary *elf,
-                               struct elf_dom_parms *parms)
-{
-    unsigned long maxva, len;
-
-    if ( !parms->bsd_symtab )
-        return;
-
-    /* Calculate the required additional kernel space for the elf image */
-
-    /* The absolute base address of the elf image */
-    maxva = elf_round_up(elf, parms->virt_kend);
-    maxva += sizeof(long); /* Space to store the size of the elf image */
-    /* Space for the elf and elf section headers */
-    maxva += (elf_uval(elf, elf->ehdr, e_ehsize) +
-              elf_shdr_count(elf) * elf_uval(elf, elf->ehdr, e_shentsize));
-    maxva = elf_round_up(elf, maxva);
-
-    /* Space for the symbol and string tabs */
-    len = (unsigned long)elf->send - (unsigned long)elf->sstart;
-    maxva = elf_round_up(elf, maxva + len);
-
-    /* The address the kernel must expanded to */
-    parms->virt_end = maxva;
-}
-
-int elf_xen_dom_load_binary(struct elf_binary *elf,
-                            struct elf_dom_parms *parms)
-{
-    elf_ehdr *sym_ehdr;
-    unsigned long shdr, symtab_addr;
-    unsigned long maxva, symbase;
-    uint8_t i;
-    char *p;
-
-    elf_load_binary(elf);
-
-    if ( !parms->bsd_symtab )
-        return 0;
-
-#define elf_hdr_elm(_elf, _hdr, _elm, _val)     \
-do {                                            \
-    if ( elf_64bit(_elf) )                      \
-        (_hdr)->e64._elm = _val;                \
-    else                                        \
-        (_hdr)->e32._elm = _val;                \
-} while ( 0 )
-
-    /* ehdr right after the kernel image (4 byte aligned) */
-    symbase = elf_round_up(elf, parms->virt_kend);
-    symtab_addr = maxva = symbase + sizeof(long);
-
-    /* Set up Elf header. */
-    sym_ehdr = (elf_ehdr *)symtab_addr;
-    maxva = elf_copy_ehdr(elf, sym_ehdr);
-
-    elf_hdr_elm(elf, sym_ehdr, e_phoff, 0);
-    elf_hdr_elm(elf, sym_ehdr, e_shoff, elf_uval(elf, elf->ehdr, e_ehsize));
-    elf_hdr_elm(elf, sym_ehdr, e_phentsize, 0);
-    elf_hdr_elm(elf, sym_ehdr, e_phnum, 0);
-
-    /* Copy Elf section headers. */
-    shdr = maxva;
-    maxva = elf_copy_shdr(elf, (elf_shdr *)shdr);
-
-    for ( i = 0; i < elf_shdr_count(elf); i++ )
-    {
-        uint8_t type;
-        unsigned long tmp;
-        type = elf_uval(elf, (elf_shdr *)shdr, sh_type);
-        if ( (type == SHT_STRTAB) || (type == SHT_SYMTAB) )
-        {
-             elf_msg(elf, "%s: shdr %i at 0x%p -> 0x%p\n", __func__, i,
-                     elf_section_start(elf, (elf_shdr *)shdr), (void *)maxva);
-             tmp = elf_copy_section(elf, (elf_shdr *)shdr, (void *)maxva);
-             /* Mangled to be based on ELF header location. */
-             elf_hdr_elm(elf, (elf_shdr *)shdr, sh_offset,
-                         maxva - symtab_addr);
-             maxva = tmp;
-        }
-        shdr += elf_uval(elf, elf->ehdr, e_shentsize);
-    }
-
-    /* Write down the actual sym size. */
-    p = (char *)symbase;
-    *(long *)p = maxva - symtab_addr; /* sym size */
-
-#undef elf_ehdr_elm
-
-    return 0;
-}
-
 static int elf_xen_addr_calc_check(struct elf_binary *elf,
                                    struct elf_dom_parms *parms)
 {
@@ -467,13 +374,16 @@ static int elf_xen_addr_calc_check(struct elf_binary *elf,
     parms->virt_offset = parms->virt_base - parms->elf_paddr_offset;
     parms->virt_kstart = elf->pstart + parms->virt_offset;
     parms->virt_kend   = elf->pend   + parms->virt_offset;
-    parms->virt_end    = parms->virt_kend;
 
     if ( parms->virt_entry == UNSET_ADDR )
         parms->virt_entry = elf_uval(elf, elf->ehdr, e_entry);
 
     if ( parms->bsd_symtab )
-        elf_xen_loadsymtab(elf, parms);
+    {
+        elf_parse_bsdsyms(elf, parms->virt_kend);
+        if ( elf->bsd_symtab_pend )
+            parms->virt_kend = elf->bsd_symtab_pend + parms->virt_offset;
+    }
 
     elf_msg(elf, "%s: addresses:\n", __FUNCTION__);
     elf_msg(elf, "    virt_base        = 0x%" PRIx64 "\n", parms->virt_base);
@@ -481,14 +391,12 @@ static int elf_xen_addr_calc_check(struct elf_binary *elf,
     elf_msg(elf, "    virt_offset      = 0x%" PRIx64 "\n", parms->virt_offset);
     elf_msg(elf, "    virt_kstart      = 0x%" PRIx64 "\n", parms->virt_kstart);
     elf_msg(elf, "    virt_kend        = 0x%" PRIx64 "\n", parms->virt_kend);
-    elf_msg(elf, "    virt_end         = 0x%" PRIx64 "\n", parms->virt_end);
     elf_msg(elf, "    virt_entry       = 0x%" PRIx64 "\n", parms->virt_entry);
 
     if ( (parms->virt_kstart > parms->virt_kend) ||
          (parms->virt_entry < parms->virt_kstart) ||
          (parms->virt_entry > parms->virt_kend) ||
-         (parms->virt_base > parms->virt_kstart) ||
-         (parms->virt_kend > parms->virt_end) )
+         (parms->virt_base > parms->virt_kstart) )
     {
         elf_err(elf, "%s: ERROR: ELF start or entries are out of bounds.\n",
                 __FUNCTION__);

xen/common/libelf/libelf-loader.c

@@ -10,8 +10,6 @@ int elf_init(struct elf_binary *elf, const char *image, size_t size)
 {
     const elf_shdr *shdr;
     uint64_t i, count, section, offset;
-    uint64_t low = -1;
-    uint64_t high = 0;
 
     if ( !elf_is_elfbinary(image) )
     {
@@ -22,14 +20,10 @@ int elf_init(struct elf_binary *elf, const char *image, size_t size)
     memset(elf, 0, sizeof(*elf));
     elf->image = image;
     elf->size = size;
-    elf->ehdr = (elf_ehdr *) image;
+    elf->ehdr = (elf_ehdr *)image;
     elf->class = elf->ehdr->e32.e_ident[EI_CLASS];
     elf->data = elf->ehdr->e32.e_ident[EI_DATA];
 
-#ifdef VERBOSE
-    elf_set_verbose(elf);
-#endif
-
     /* Sanity check phdr. */
     offset = elf_uval(elf, elf->ehdr, e_phoff) +
         elf_uval(elf, elf->ehdr, e_phentsize) * elf_phdr_count(elf);
@@ -60,43 +54,18 @@ int elf_init(struct elf_binary *elf, const char *image, size_t size)
     count = elf_shdr_count(elf);
     for ( i = 0; i < count; i++ )
     {
-        const char *sh_symend, *sh_strend;
-
         shdr = elf_shdr_by_index(elf, i);
         if ( elf_uval(elf, shdr, sh_type) != SHT_SYMTAB )
             continue;
         elf->sym_tab = shdr;
-        sh_symend = (const char *)elf_section_end(elf, shdr);
         shdr = elf_shdr_by_index(elf, elf_uval(elf, shdr, sh_link));
         if ( shdr == NULL )
         {
             elf->sym_tab = NULL;
-            sh_symend = 0;
             continue;
         }
         elf->sym_strtab = elf_section_start(elf, shdr);
-        sh_strend = (const char *)elf_section_end(elf, shdr);
-
-        if ( low > (unsigned long)elf->sym_tab )
-            low = (unsigned long)elf->sym_tab;
-        if ( low > (unsigned long)shdr )
-            low = (unsigned long)shdr;
-
-        if ( high < ((unsigned long)sh_symend) )
-            high = (unsigned long)sh_symend;
-        if ( high < ((unsigned long)sh_strend) )
-            high = (unsigned long)sh_strend;
-
-        elf_msg(elf, "%s: shdr: sym_tab=%p size=0x%" PRIx64 "\n",
-                __FUNCTION__, elf->sym_tab,
-                elf_uval(elf, elf->sym_tab, sh_size));
-        elf_msg(elf, "%s: shdr: str_tab=%p size=0x%" PRIx64 "\n",
-                __FUNCTION__, elf->sym_strtab, elf_uval(elf, shdr, sh_size));
-
-        elf->sstart = low;
-        elf->send = high;
-        elf_msg(elf, "%s: symbol map: 0x%" PRIx64 " -> 0x%" PRIx64 "\n",
-                __FUNCTION__, elf->sstart, elf->send);
+        break;
     }
 
     return 0;
@@ -115,6 +84,101 @@ void elf_set_verbose(struct elf_binary *elf)
 }
 #endif
 
+/* Calculate the required additional kernel space for the elf image */
+void elf_parse_bsdsyms(struct elf_binary *elf, uint64_t pstart)
+{
+    uint64_t sz;
+    const elf_shdr *shdr;
+    int i, type;
+
+    if ( !elf->sym_tab )
+        return;
+
+    pstart = elf_round_up(elf, pstart);
+
+    /* Space to store the size of the elf image */
+    sz = sizeof(uint32_t);
+
+    /* Space for the elf and elf section headers */
+    sz += (elf_uval(elf, elf->ehdr, e_ehsize) +
+           elf_shdr_count(elf) * elf_uval(elf, elf->ehdr, e_shentsize));
+    sz = elf_round_up(elf, sz);
+
+    /* Space for the symbol and string tables. */
+    for ( i = 0; i < elf_shdr_count(elf); i++ )
+    {
+        shdr = elf_shdr_by_index(elf, i);
+        type = elf_uval(elf, (elf_shdr *)shdr, sh_type);
+        if ( (type == SHT_STRTAB) || (type == SHT_SYMTAB) )
+            sz = elf_round_up(elf, sz + elf_uval(elf, shdr, sh_size));
+    }
+
+    elf->bsd_symtab_pstart = pstart;
+    elf->bsd_symtab_pend   = pstart + sz;
+}
+
+static void elf_load_bsdsyms(struct elf_binary *elf)
+{
+    elf_ehdr *sym_ehdr;
+    unsigned long sz;
+    char *maxva, *symbase, *symtab_addr;
+    elf_shdr *shdr;
+    int i, type;
+
+    if ( !elf->bsd_symtab_pstart )
+        return;
+
+#define elf_hdr_elm(_elf, _hdr, _elm, _val)     \
+do {                                            \
+    if ( elf_64bit(_elf) )                      \
+        (_hdr)->e64._elm = _val;                \
+    else                                        \
+        (_hdr)->e32._elm = _val;                \
+} while ( 0 )
+
+    symbase = elf_get_ptr(elf, elf->bsd_symtab_pstart);
+    symtab_addr = maxva = symbase + sizeof(uint32_t);
+
+    /* Set up Elf header. */
+    sym_ehdr = (elf_ehdr *)symtab_addr;
+    sz = elf_uval(elf, elf->ehdr, e_ehsize);
+    memcpy(sym_ehdr, elf->ehdr, sz);
+    maxva += sz; /* no round up */
+
+    elf_hdr_elm(elf, sym_ehdr, e_phoff, 0);
+    elf_hdr_elm(elf, sym_ehdr, e_shoff, elf_uval(elf, elf->ehdr, e_ehsize));
+    elf_hdr_elm(elf, sym_ehdr, e_phentsize, 0);
+    elf_hdr_elm(elf, sym_ehdr, e_phnum, 0);
+
+    /* Copy Elf section headers. */
+    shdr = (elf_shdr *)maxva;
+    sz = elf_shdr_count(elf) * elf_uval(elf, elf->ehdr, e_shentsize);
+    memcpy(shdr, elf->image + elf_uval(elf, elf->ehdr, e_shoff), sz);
+    maxva = (char *)(long)elf_round_up(elf, (long)maxva + sz);
+
+    for ( i = 0; i < elf_shdr_count(elf); i++ )
+    {
+        type = elf_uval(elf, shdr, sh_type);
+        if ( (type == SHT_STRTAB) || (type == SHT_SYMTAB) )
+        {
+             elf_msg(elf, "%s: shdr %i at 0x%p -> 0x%p\n", __func__, i,
+                     elf_section_start(elf, shdr), maxva);
+             sz = elf_uval(elf, shdr, sh_size);
+             memcpy(maxva, elf_section_start(elf, shdr), sz);
+             /* Mangled to be based on ELF header location. */
+             elf_hdr_elm(elf, shdr, sh_offset, maxva - symtab_addr);
+             maxva = (char *)(long)elf_round_up(elf, (long)maxva + sz);
+        }
+        shdr = (elf_shdr *)((long)shdr +
+                            (long)elf_uval(elf, elf->ehdr, e_shentsize));
+    }
+
+    /* Write down the actual sym size. */
+    *(uint32_t *)symbase = maxva - symtab_addr;
+
+#undef elf_ehdr_elm
+}
+
 void elf_parse_binary(struct elf_binary *elf)
 {
     const elf_phdr *phdr;
@@ -165,6 +229,8 @@ void elf_load_binary(struct elf_binary *elf)
         memcpy(dest, elf->image + offset, filesz);
         memset(dest + filesz, 0, memsz - filesz);
     }
+
+    elf_load_bsdsyms(elf);
 }
 
 void *elf_get_ptr(struct elf_binary *elf, unsigned long addr)

xen/common/libelf/libelf-tools.c

@@ -238,36 +238,6 @@ int elf_phdr_is_loadable(struct elf_binary *elf, const elf_phdr * phdr)
     return ((p_type == PT_LOAD) && (p_flags & (PF_W | PF_X)) != 0);
 }
 
-unsigned long
-elf_copy_ehdr(struct elf_binary *elf, void *dest)
-{
-    uint64_t size;
-
-    size = elf_uval(elf, elf->ehdr, e_ehsize);
-    memcpy(dest, elf->ehdr, size);
-    return elf_round_up(elf, (unsigned long)(dest) + size);
-}
-
-unsigned long
-elf_copy_shdr(struct elf_binary *elf, void *dest)
-{
-    uint64_t size;
-
-    size = elf_shdr_count(elf) * elf_uval(elf, elf->ehdr, e_shentsize);
-    memcpy(dest, elf->image + elf_uval(elf, elf->ehdr, e_shoff), size);
-    return elf_round_up(elf, (unsigned long)(dest) + size);
-}
-
-unsigned long
-elf_copy_section(struct elf_binary *elf, const elf_shdr *shdr, void *dest)
-{
-    uint64_t size;
-
-    size = elf_uval(elf, shdr, sh_size);
-    memcpy(dest, elf_section_start(elf, shdr), size);
-    return elf_round_up(elf, (unsigned long)(dest) + size);
-}
-
 /*
  * Local variables:
  * mode: C

xen/common/page_alloc.c

@@ -91,6 +91,14 @@ custom_param("dma_emergency_pool", parse_dma_emergency_pool);
 #define round_pgdown(_p)  ((_p)&PAGE_MASK)
 #define round_pgup(_p)    (((_p)+(PAGE_SIZE-1))&PAGE_MASK)
 
+#ifndef NDEBUG
+/* Avoid callers relying on allocations returning zeroed pages. */
+#define scrub_page(p) memset((p), 0xc2, PAGE_SIZE)
+#else
+/* For a production build, clear_page() is the fastest way to scrub. */
+#define scrub_page(p) clear_page(p)
+#endif
+
 static DEFINE_SPINLOCK(page_scrub_lock);
 LIST_HEAD(page_scrub_list);
 static unsigned long scrub_pages;
@@ -618,13 +626,13 @@ void __init scrub_heap_pages(void)
             {
                 p = page_to_virt(mfn_to_page(mfn));
                 memguard_unguard_range(p, PAGE_SIZE);
-                clear_page(p);
+                scrub_page(p);
                 memguard_guard_range(p, PAGE_SIZE);
             }
             else
             {
                 p = map_domain_page(mfn);
-                clear_page(p);
+                scrub_page(p);
                 unmap_domain_page(p);
             }
         }
@@ -1018,7 +1026,7 @@ static void page_scrub_softirq(void)
             pg = list_entry(ent, struct page_info, list);
             ent = ent->prev;
             p = map_domain_page(page_to_mfn(pg));
-            clear_page(p);
+            scrub_page(p);
             unmap_domain_page(p);
             free_heap_pages(pfn_dom_zone_type(page_to_mfn(pg)), pg, 0);
         }

xen/include/asm-powerpc/boot.h

@@ -0,0 +1,46 @@
+/*
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * 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 General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
+ *
+ * Copyright IBM Corp. 2007
+ *
+ * Authors: Hollis Blanchard <hollisb@us.ibm.com>
+ */
+
+#ifndef _ASM_BOOT_H
+#define _ASM_BOOT_H
+
+/* a collection of interfaces used during boot. */
+
+extern void boot_of_init(ulong, ulong);
+extern void *boot_of_devtree(void);
+extern void boot_of_serial(void *);
+extern void boot_of_finish(void);
+extern int boot_of_mem_avail(int pos, ulong *startpage, ulong *endpage);
+
+extern void parse_multiboot(ulong tags_addr);
+
+extern void memory_init(void);
+
+extern char *xen_cmdline;
+extern ulong dom0_addr;
+extern ulong dom0_len;
+extern char *dom0_cmdline;
+extern ulong initrd_start;
+extern ulong initrd_len;
+
+/* From linker script. */
+extern char builtin_cmdline[];
+
+#endif

xen/include/asm-x86/domain.h

@@ -77,10 +77,10 @@ struct shadow_domain {
     int               locker; /* processor which holds the lock */
     const char       *locker_function; /* Func that took it */
     unsigned int      opt_flags;    /* runtime tunable optimizations on/off */
-    struct list_head  pinned_shadows; 
+    struct list_head  pinned_shadows;
 
     /* Memory allocation */
-    struct list_head  freelists[SHADOW_MAX_ORDER + 1]; 
+    struct list_head  freelists[SHADOW_MAX_ORDER + 1];
     struct list_head  p2m_freelist;
     unsigned int      total_pages;  /* number of pages allocated */
     unsigned int      free_pages;   /* number of pages on freelists */
@@ -116,7 +116,7 @@ struct hap_domain {
     spinlock_t        lock;
     int               locker;
     const char       *locker_function;
-    
+
     struct list_head  freelist;
     unsigned int      total_pages;  /* number of pages allocated */
     unsigned int      free_pages;   /* number of pages on freelists */
@@ -131,13 +131,13 @@ struct p2m_domain {
     spinlock_t         lock;
     int                locker;   /* processor which holds the lock */
     const char        *locker_function; /* Func that took it */
-    
+
     /* Pages used to construct the p2m */
     struct list_head   pages;
 
     /* Functions to call to get or free pages for the p2m */
     struct page_info * (*alloc_page  )(struct domain *d);
-    void               (*free_page   )(struct domain *d, 
+    void               (*free_page   )(struct domain *d,
                                        struct page_info *pg);
 
     /* Highest guest frame that's ever been mapped in the p2m */
@@ -177,6 +177,7 @@ struct paging_domain {
     /* log dirty support */
     struct log_dirty_domain log_dirty;
 };
+
 struct paging_vcpu {
     /* Pointers to mode-specific entry points. */
     struct paging_mode *mode;
@@ -184,9 +185,9 @@ struct paging_vcpu {
     unsigned int translate_enabled:1;
     /* HVM guest: last emulate was to a pagetable */
     unsigned int last_write_was_pt:1;
-    /* Translated guest: virtual TLB */    
+    /* Translated guest: virtual TLB */
     struct shadow_vtlb *vtlb;
-    spinlock_t          vtlb_lock; 
+    spinlock_t          vtlb_lock;
 
     /* paging support extension */
     struct shadow_vcpu shadow;
@@ -303,7 +304,7 @@ struct arch_vcpu
      * shadow refcounts are in use */
     pagetable_t shadow_table[4];        /* (MFN) shadow(s) of guest */
     pagetable_t monitor_table;          /* (MFN) hypervisor PT (for HVM) */
-    unsigned long cr3;           	    /* (MA) value to install in HW CR3 */
+    unsigned long cr3;                  /* (MA) value to install in HW CR3 */
 
     /* Current LDT details. */
     unsigned long shadow_ldt_mapcnt;

xen/include/asm-x86/hap.h

@@ -3,7 +3,7 @@
  *
  * hardware-assisted paging
  * Copyright (c) 2007 Advanced Micro Devices (Wei Huang)
- * 
+ *
  * Parts of this code are Copyright (c) 2006 by XenSource Inc.
  * Parts of this code are Copyright (c) 2006 by Michael A Fetterman
  * Parts based on earlier work by Michael A Fetterman, Ian Pratt et al.

xen/include/asm-x86/mach-es7000/mach_mpparse.h

@@ -1,45 +0,0 @@
-#ifndef __ASM_MACH_MPPARSE_H
-#define __ASM_MACH_MPPARSE_H
-
-#include <xen/acpi.h>
-
-extern int parse_unisys_oem(char *oemptr);
-extern int find_unisys_acpi_oem_table(unsigned long *oem_addr);
-
-static inline int mps_oem_check(struct mp_config_table *mpc, char *oem,
-		char *productid)
-{
-	if (mpc->mpc_oemptr) {
-		struct mp_config_oemtable *oem_table = 
-			(struct mp_config_oemtable *)(long)mpc->mpc_oemptr;
-		if (!strncmp(oem, "UNISYS", 6))
-			return parse_unisys_oem((char *)oem_table);
-	}
-	return 0;
-}
-
-static inline int es7000_check_dsdt(void)
-{
-	struct acpi_table_header *header = NULL;
-	if(!acpi_get_table_header_early(ACPI_DSDT, &header))
-		acpi_table_print(header, 0);
-	if (!strncmp(header->oem_id, "UNISYS", 6))
-		return 1;
-	return 0;
-}
-
-/* Hook from generic ACPI tables.c */
-static inline int acpi_madt_oem_check(char *oem_id, char *oem_table_id)
-{
-	unsigned long oem_addr;
-	if (!find_unisys_acpi_oem_table(&oem_addr)) {
-		if (es7000_check_dsdt())
-			return parse_unisys_oem((char *)oem_addr);
-		else
-			return 1;
-	}
-	return 0;
-}
-
-
-#endif /* __ASM_MACH_MPPARSE_H */

xen/include/asm-x86/msr.h

@@ -240,6 +240,11 @@ static inline void write_efer(__u64 val)
 #define MSR_K8_MC4_ADDR			0x412
 #define MSR_K8_MC4_MISC			0x413
 
+#define MSR_K8_MC5_CTL			0x414
+#define MSR_K8_MC5_STATUS		0x415
+#define MSR_K8_MC5_ADDR			0x416
+#define MSR_K8_MC5_MISC			0x417
+
 /* Pentium IV performance counter MSRs */
 #define MSR_P4_BPU_PERFCTR0 		0x300
 #define MSR_P4_BPU_PERFCTR1 		0x301

xen/include/public/libelf.h

@@ -65,12 +65,13 @@ struct elf_binary {
 
     /* loaded to */
     char *dest;
-    uint64_t sstart;
-    uint64_t send;
     uint64_t pstart;
     uint64_t pend;
     uint64_t reloc_offset;
 
+    uint64_t bsd_symtab_pstart;
+    uint64_t bsd_symtab_pend;
+
 #ifndef __XEN__
     /* misc */
     FILE *log;
@@ -150,11 +151,6 @@ const elf_note *elf_note_next(struct elf_binary *elf, const elf_note * note);
 int elf_is_elfbinary(const void *image);
 int elf_phdr_is_loadable(struct elf_binary *elf, const elf_phdr * phdr);
 
-unsigned long elf_copy_ehdr(struct elf_binary *elf, void *dest);
-unsigned long elf_copy_shdr(struct elf_binary *elf, void *dest);
-unsigned long elf_copy_section(struct elf_binary *elf,
-                               const elf_shdr *shdr, void *dest);
-
 /* ------------------------------------------------------------------------ */
 /* xc_libelf_loader.c                                                       */
 
@@ -171,6 +167,8 @@ void elf_load_binary(struct elf_binary *elf);
 void *elf_get_ptr(struct elf_binary *elf, unsigned long addr);
 uint64_t elf_lookup_addr(struct elf_binary *elf, const char *symbol);
 
+void elf_parse_bsdsyms(struct elf_binary *elf, uint64_t pstart); /* private */
+
 /* ------------------------------------------------------------------------ */
 /* xc_libelf_relocate.c                                                     */
 
@@ -221,8 +219,7 @@ struct elf_dom_parms {
     /* calculated */
     uint64_t virt_offset;
     uint64_t virt_kstart;
-    uint64_t virt_kend; /* end of kernel image */
-    uint64_t virt_end;  /* end of kernel symtab (== virt_kend if none) */
+    uint64_t virt_kend;
 };
 
 static inline void elf_xen_feature_set(int nr, uint32_t * addr)
@@ -245,7 +242,4 @@ int elf_xen_parse_guest_info(struct elf_binary *elf,
 int elf_xen_parse(struct elf_binary *elf,
                   struct elf_dom_parms *parms);
 
-int elf_xen_dom_load_binary(struct elf_binary *elf,
-                            struct elf_dom_parms *parms);
-
 #endif /* __XC_LIBELF__ */

xen/include/xen/multiboot2.h

@@ -0,0 +1,99 @@
+/*
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; either version 2 of the
+ * License, or (at your option) any later version.
+ *
+ * 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 General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
+ *
+ * Copyright IBM Corp. 2006, 2007
+ *
+ * Authors: Hollis Blanchard <hollisb@us.ibm.com>
+ *          
+ */
+
+#ifndef _MULTIBOOT2_H_
+#define _MULTIBOOT2_H_
+
+/* How many bytes from the start of the file we search for the header.  */
+#define MB2_HEADER_SEARCH           8192
+
+/* The magic field should contain this.  */
+#define MB2_HEADER_MAGIC            0xe85250d6
+
+/* Passed from the bootloader to the kernel.  */
+#define MB2_BOOTLOADER_MAGIC        0x36d76289
+
+#include <stdint.h>
+
+#define for_each_tag(_tag, _tags) \
+    for ((_tag) = (_tags); \
+            ((_tag)->key != MB2_TAG_END && (_tag)->key != 0); \
+            (_tag) = (void *)(_tag) + (_tag)->len)
+
+typedef uint32_t mb2_word;
+
+struct mb2_header
+{
+  uint32_t magic;
+};
+
+struct mb2_tag_header
+{
+  uint32_t key;
+  uint32_t len;
+};
+
+#define MB2_TAG_START     1
+struct mb2_tag_start
+{
+  struct mb2_tag_header header;
+  mb2_word size; /* Total size of all mb2 tags. */
+};
+
+#define MB2_TAG_NAME      2
+struct mb2_tag_name
+{
+  struct mb2_tag_header header;
+  char name[1];
+};
+
+#define MB2_TAG_MODULE    3
+struct mb2_tag_module
+{
+  struct mb2_tag_header header;
+  mb2_word addr;
+  mb2_word size;
+  unsigned char type[36];
+  unsigned char cmdline[1];
+};
+
+#define MB2_TAG_MEMORY    4
+struct mb2_tag_memory
+{
+  struct mb2_tag_header header;
+  mb2_word addr;
+  mb2_word size;
+  mb2_word type;
+};
+
+#define MB2_TAG_UNUSED    5
+struct mb2_tag_unused
+{
+  struct mb2_tag_header header;
+};
+
+#define MB2_TAG_END       0xffff
+struct mb2_tag_end
+{
+  struct mb2_tag_header header;
+};
+
+#endif