# HG changeset patch
# User Isaku Yamahata <yamahata@xxxxxxxxxxxxx>
# Date 1235986016 -32400
# Node ID 6227bf629626105380f8f192da78a9e25adf4a27
# Parent c4c4ba857d8b4b045f169b33c7e36dcdc07af720
# Parent e5c696aaf2a6e8805231c0c0f1414560262e7005
merge with xen-unstable.hg
---
buildconfigs/enable-xen-config | 57 +++++-
buildconfigs/mk.linux-2.6-pvops | 14 +
docs/misc/xsm-flask.txt | 148 ++++++++++++++++
tools/console/daemon/io.c | 20 +-
tools/hotplug/Linux/init.d/xendomains | 63 ++++--
tools/libxc/xc_ptrace.c | 255 ++--------------------------
tools/misc/xenpm.c | 38 ++--
tools/pygrub/src/pygrub | 6
tools/python/xen/util/vscsi_util.py | 55 +++++-
tools/python/xen/xend/XendAPI.py | 9
tools/python/xen/xend/XendBootloader.py | 73 ++++++--
tools/python/xen/xend/XendConfig.py | 41 +++-
tools/python/xen/xend/XendConstants.py | 1
tools/python/xen/xend/XendDomainInfo.py | 17 +
tools/python/xen/xend/server/vscsiif.py | 15 +
tools/python/xen/xm/create.py | 115 +++++++-----
tools/python/xen/xm/main.py | 61 ++++--
tools/python/xen/xm/xenapi_create.py | 1
tools/xenstore/xenstored_core.c | 6
xen/arch/ia64/xen/domain.c | 10 -
xen/arch/x86/acpi/power.c | 2
xen/arch/x86/hvm/hvm.c | 16 +
xen/arch/x86/hvm/mtrr.c | 19 +-
xen/arch/x86/hvm/vpic.c | 2
xen/arch/x86/hvm/vpt.c | 26 ++
xen/arch/x86/io_apic.c | 71 ++-----
xen/arch/x86/irq.c | 18 +
xen/arch/x86/mm/shadow/multi.c | 19 +-
xen/arch/x86/msi.c | 111 ++++++++----
xen/arch/x86/shutdown.c | 1
xen/arch/x86/time.c | 7
xen/common/timer.c | 7
xen/drivers/passthrough/amd/iommu_init.c | 6
xen/drivers/passthrough/amd/iommu_intr.c | 14 -
xen/drivers/passthrough/amd/iommu_map.c | 31 +--
xen/drivers/passthrough/amd/pci_amd_iommu.c | 23 +-
xen/drivers/passthrough/pci.c | 1
xen/drivers/passthrough/vtd/ia64/vtd.c | 31 +++
xen/drivers/passthrough/vtd/iommu.c | 45 ++--
xen/drivers/passthrough/vtd/iommu.h | 2
xen/drivers/passthrough/vtd/qinval.c | 5
xen/drivers/passthrough/vtd/x86/vtd.c | 20 ++
xen/include/asm-ia64/msi.h | 20 ++
xen/include/asm-x86/fixmap.h | 2
xen/include/asm-x86/io_apic.h | 8
xen/include/asm-x86/msi.h | 8
xen/include/asm-x86/mtrr.h | 3
xen/include/public/hvm/params.h | 5
xen/include/xen/iommu.h | 2
xen/include/xen/pci.h | 6
xen/include/xen/timer.h | 3
xen/xsm/flask/ss/policydb.c | 4
52 files changed, 965 insertions(+), 578 deletions(-)
diff -r c4c4ba857d8b -r 6227bf629626 buildconfigs/enable-xen-config
--- a/buildconfigs/enable-xen-config Mon Mar 02 16:52:22 2009 +0900
+++ b/buildconfigs/enable-xen-config Mon Mar 02 18:26:56 2009 +0900
@@ -19,29 +19,70 @@ setopt()
# Then append the new value
case ${VALUE} in
- y|m) echo "${OPTION}=${VALUE}" >> "${CONFIG}" ;;
- n) echo "# ${OPTION} is not set" >> "${CONFIG}" ;;
- *) echo "Invalid value ${VALUE} for ${OPTION}" 1>&2 ; exit 1 ;;
+ n) echo "# ${OPTION} is not set" >> "${CONFIG}" ;;
+ y|m|*) echo "${OPTION}=${VALUE}" >> "${CONFIG}" ;;
esac
}
setopt CONFIG_PARAVIRT y
+setopt CONFIG_PARAVIRT_DEBUG y
setopt CONFIG_PARAVIRT_GUEST y
+
setopt CONFIG_XEN y
+setopt CONFIG_XEN_BLKDEV_FRONTEND y
+setopt CONFIG_XEN_NETDEV_FRONTEND y
+setopt CONFIG_XEN_KBDDEV_FRONTEND y
+setopt CONFIG_XEN_FBDEV_FRONTEND y
+setopt CONFIG_XEN_BALLOON y
+setopt CONFIG_XEN_SCRUB_PAGES y
+setopt CONFIG_XEN_DEV_EVTCHN y
+setopt CONFIG_XEN_BACKEND y
+setopt CONFIG_XEN_BLKDEV_BACKEND y
+setopt CONFIG_XEN_NETDEV_BACKEND y
+setopt CONFIG_XENFS y
+setopt CONFIG_XEN_COMPAT_XENFS y
+setopt CONFIG_HVC_XEN y
+setopt CONFIG_XEN_MAX_DOMAIN_MEMORY 32
+setopt CONFIG_XEN_DEBUG_FS y
+setopt CONFIG_XEN_DOM0 y
+
setopt CONFIG_VMI y
+
setopt CONFIG_KVM y
setopt CONFIG_KVM_INTEL y
setopt CONFIG_KVM_AMD y
+setopt CONFIG_KVM_CLOCK y
+setopt CONFIG_KVM_GUEST n
+setopt CONFIG_KVM_TRACE n
+
setopt CONFIG_LGUEST n
-setopt CONFIG_XEN_BLKDEV_FRONTEND y
-setopt CONFIG_XEN_NETDEV_FRONTEND y
-setopt CONFIG_HVC_XEN y
+
+setopt CONFIG_LOCALVERSION_AUTO n
+
+# Should all be set one way or another in defconfig but aren't
setopt CONFIG_NUMA n
-setopt CONFIG_LOCALVERSION_AUTO n
+setopt CONFIG_X86_VSMP n
+setopt CONFIG_X86_UV n
+setopt CONFIG_CALGARY_IOMMU n
+setopt CONFIG_AMD_IOMMU n
+setopt CONFIG_MAXSMP n
+setopt CONFIG_SPARSEMEM_VMEMMAP n
+setopt CONFIG_I7300_IDLE n
+setopt CONFIG_DMAR n
+setopt CONFIG_INTR_REMAP n
+setopt CONFIG_GFS2_FS n
+setopt CONFIG_IOMMU_DEBUG n
case ${XEN_TARGET_ARCH} in
x86_32) setopt CONFIG_64BIT n ;;
- x86_64) setopt CONFIG_64BIT y ;;
+ x86_64)
+ setopt CONFIG_64BIT y
+ setopt CONFIG_IA32_EMULATION y
+ setopt CONFIG_IA32_AOUT n
+ setopt CONFIG_CRYPTO_AES_X86_64 n
+ setopt CONFIG_CRYPTO_SALSA20_X86_64 n
+ setopt CONFIG_CRYPTO_TWOFISH_X86_64 n
+ ;;
*) ;;
esac
diff -r c4c4ba857d8b -r 6227bf629626 buildconfigs/mk.linux-2.6-pvops
--- /dev/null Thu Jan 01 00:00:00 1970 +0000
+++ b/buildconfigs/mk.linux-2.6-pvops Mon Mar 02 18:26:56 2009 +0900
@@ -0,0 +1,14 @@
+XEN_LINUX_SOURCE ?= git-clone
+LINUX_VER ?= 2.6-pvops
+
+IMAGE_TARGET ?= bzImage
+
+XEN_LINUX_CONFIG_UPDATE := buildconfigs/enable-xen-config
+
+XEN_LINUX_GIT_URL ?=
git://git.kernel.org/pub/scm/linux/kernel/git/jeremy/xen.git
+XEN_LINUX_GIT_REMOTENAME ?= xen
+XEN_LINUX_GIT_REMOTEBRANCH ?= xen/dom0/hackery
+
+EXTRAVERSION ?=
+
+include buildconfigs/mk.linux-2.6-common
diff -r c4c4ba857d8b -r 6227bf629626 docs/misc/xsm-flask.txt
--- /dev/null Thu Jan 01 00:00:00 1970 +0000
+++ b/docs/misc/xsm-flask.txt Mon Mar 02 18:26:56 2009 +0900
@@ -0,0 +1,148 @@
+These notes are compiled from xen-devel questions and postings that have
occured
+since the inclusion of XSM. These notes are not intended to be definitive
+documentation but should address many common problems that arrise when
+experimenting with XSM:FLASK.
+
+Xen XSM:FLASK configuration
+---------------------------
+
+1) cd xen-unstable.hg
+2) edit Config.mk in the toplevel xen directory as follows:
+
+ XSM_ENABLE ?= y
+ FLASK_ENABLE ?= y
+ ACM_SECURITY ?= n
+
+NB: Only one security module can be selected at a time. If no module is
+selected, then the default DUMMY module will be enforced. The DUMMY module
+only exercises the security framework and does not enforce any security
+policies. Changing the security module selection will require recompiling xen.
+These settings will also configure the corresponding toolchain support.
+
+3) make xen
+4) make tools
+
+
+Xen XSM:FLASK policy
+--------------------
+
+These instructions will enable the configuration and build of the sample
policy.
+The sample policy provides the MINIMUM policy necessary to boot a
+paravirtualized dom0 and create a paravirtualized domU. Many of the
+default capabilities and usages supported by dom0/domU are disallowed by the
+sample policy. Further, the policy is comprised of a limited number of types
and
+must be adjusted to meet the specific security goals of the installation.
+Modification of the policy is straightforward and is covered in a later
section.
+
+NB: The policy is not automatically built as part of the tool support because
+of an external dependancy on the checkpolicy compiler. The FLASK policy uses
+the same syntax and structure as SELinux and compiling the policy relies on
+the SELinux policy toolchain. This toolchain is available under many
+distributions as well as the following URL,
+
+
http://userspace.selinuxproject.org/releases/20080909/stable/checkpolicy-1.34.7.tar.gz
+
+1) cd xen-unstable.hg/tools/flask/policy
+2) make policy
+3) cp policy.20 /boot/xenpolicy.20
+4) edit /etc/grub.conf, add a module line to the xen entry,
+
+ module /xenpolicy.20
+
+5) reboot, and select the updated xen entry
+
+NB: The module entry can be inserted on any line after the xen kernel line.
Typical
+configurations use the last module entry or the module entry that immediately
+follows the xen kernel entry.
+
+Xen configuration of xend
+-------------------------
+
+1) cd /etc/xen
+2) edit xend-config.sxp
+3) uncomment the line containing the key:value pair entry,
+
+ #(xsm_module_name dummy)
+
+4) change the value entry to 'flask'
+
+ (xsm_module_name flask)
+
+5) restart xend
+
+Creating policy controlled domains
+----------------------------------
+
+2) Edit the domain config file and add the following entry,
+
+ access_control = ["policy=,label=system_u:object_r:domU_t"]
+
+NB: The 'policy' field is not used by XSM:FLASK. The 'label' must exist in
the
+loaded policy. 'system_u:object_r:domU_t' is one of the existing labels from
+the sample policy and shown for example purposes.
+
+2) Create the domain using the 'xm create' command.
+3) Use the 'xm list -l' command to list the running domains and their labels.
+
+Updating the XSM:FLASK policy
+-----------------------------
+
+It is recommended that the XSM:FLASK policy be tailored to meet the specific
+security goals of the platform. The policy is tailored by editing the xen.te
+file in the 'policy' subdirectory.
+
+1) cd xen-unstable.hg/tools/flask/policy
+2) edit policy/modules/xen/xen.te - make changes to support platform security
goals.
+3) make policy
+4) cp policy.20 /boot/xenpolicy.20
+5) reboot
+
+Alternatively, one may reload the policy using the 'flask_loadpolicy' tool
+installed by the xen tools.
+
+1) flask_loadpolicy policy.20
+
+NB: The sample policy permits policy reloads as well as general manipulation of
+the Flask security server only from dom0. The policy can be tailored further
to
+restrict policy reloads and other manipulations to boot-time only, by removing
+the corresponding statements from the policy.
+
+Enforcing the XSM:FLASK policy
+------------------------------
+
+By default, XSM:FLASK is compiled and installed in permissive mode. This
+configuration will allow an XSM:FLASK system to start in enforcing mode.
+
+1) edit /etc/grub.conf
+2) append the parameter 'flask_enforcing=1' to the xen kernel line.
+3) reboot, and select the updated xen entry
+
+
+Additional notes on XSM:FLASK
+-----------------------------
+
+1) xen command line parameters
+
+ a) flask_enforcing
+
+ The default value for flask_enforcing is '0'. This parameter causes
the
+ platform to boot in permissive mode which means that the policy is
loaded
+ but not enforced. This mode is often helpful for developing new
systems
+ and policies as the policy violations are reported on the xen console
and
+ may be viewed in dom0 through 'xm dmesg'.
+
+ To boot the platform into enforcing mode, which means that the policy is
+ loaded and enforced, append 'flask_enforcing=1' on the grub line.
+
+ This parameter may also be changed through the flask hyercall.
+
+ b) flask_enabled
+
+ The default value for flask_enabled is '1'. This parameter causes the
+ platform to enable the FLASK security module under the XSM framework.
+ The parameter may be enabled/disabled only once per boot. If the
parameter
+ is set to '0', only a reboot can re-enable flask. When flask_enabled
is '0'
+ the DUMMY module is enforced.
+
+ This parameter may also be changed through the flask hypercall. But may
+ only be performed once per boot.
diff -r c4c4ba857d8b -r 6227bf629626 tools/console/daemon/io.c
--- a/tools/console/daemon/io.c Mon Mar 02 16:52:22 2009 +0900
+++ b/tools/console/daemon/io.c Mon Mar 02 18:26:56 2009 +0900
@@ -402,14 +402,28 @@ static int domain_create_tty(struct doma
assert(dom->slave_fd == -1);
assert(dom->master_fd == -1);
- cfmakeraw(&term);
-
- if (openpty(&dom->master_fd, &dom->slave_fd, NULL, &term, NULL) < 0) {
+ if (openpty(&dom->master_fd, &dom->slave_fd, NULL, NULL, NULL) < 0) {
err = errno;
dolog(LOG_ERR, "Failed to create tty for domain-%d "
"(errno = %i, %s)",
dom->domid, err, strerror(err));
return 0;
+ }
+
+ if (tcgetattr(dom->slave_fd, &term) < 0) {
+ err = errno;
+ dolog(LOG_ERR, "Failed to get tty attributes for domain-%d "
+ "(errno = %i, %s)",
+ dom->domid, err, strerror(err));
+ goto out;
+ }
+ cfmakeraw(&term);
+ if (tcsetattr(dom->slave_fd, TCSANOW, &term) < 0) {
+ err = errno;
+ dolog(LOG_ERR, "Failed to set tty attributes for domain-%d "
+ "(errno = %i, %s)",
+ dom->domid, err, strerror(err));
+ goto out;
}
if ((slave = ptsname(dom->master_fd)) == NULL) {
diff -r c4c4ba857d8b -r 6227bf629626 tools/hotplug/Linux/init.d/xendomains
--- a/tools/hotplug/Linux/init.d/xendomains Mon Mar 02 16:52:22 2009 +0900
+++ b/tools/hotplug/Linux/init.d/xendomains Mon Mar 02 18:26:56 2009 +0900
@@ -213,7 +213,7 @@ start()
start()
{
if [ -f $LOCKFILE ]; then
- echo -n "xendomains already running (lockfile exists)"
+ echo -e "xendomains already running (lockfile exists)"
return;
fi
@@ -230,10 +230,12 @@ start()
HEADER=`head -c 16 $dom | head -n 1 2> /dev/null`
if [ $HEADER = "LinuxGuestRecord" ]; then
echo -n " ${dom##*/}"
- xm restore $dom
+ XMR=`xm restore $dom 2>&1 1>/dev/null`
+ #xm restore $dom
if [ $? -ne 0 ]; then
+ echo -e "\nAn error occurred while restoring domain
${dom##*/}:\n$XMR"
rc_failed $?
- echo -n '!'
+ echo -e '!'
else
# mv $dom ${dom%/*}/.${dom##*/}
rm $dom
@@ -241,7 +243,7 @@ start()
fi
fi
done
- echo .
+ echo -e
fi
if contains_something "$XENDOMAINS_AUTO"
@@ -264,16 +266,17 @@ start()
if [ $? -eq 0 ] || is_running $dom; then
echo -n "(skip)"
else
- xm create --quiet --defconfig $dom
+ XMC=`xm create --quiet --defconfig $dom`
if [ $? -ne 0 ]; then
+ echo -e "\nAn error occurred while creating domain
${dom##*/}: $XMC\n"
rc_failed $?
- echo -n '!'
+ echo -e '!'
else
usleep $XENDOMAINS_CREATE_USLEEP
fi
fi
done
- fi
+ fi
}
all_zombies()
@@ -293,18 +296,21 @@ all_zombies()
# if it has not exited by that time kill it, so the init script will
# succeed within a finite amount of time; if $2 is nonnull, it will
# kill the command as well as soon as no domain (except for zombies)
-# are left (used for shutdown --all).
+# are left (used for shutdown --all). Third parameter, if any, suppresses
+# output of dots per working state (formatting issues)
watchdog_xm()
{
if test -z "$XENDOMAINS_STOP_MAXWAIT" -o "$XENDOMAINS_STOP_MAXWAIT" = "0";
then
exit
fi
+
usleep 20000
for no in `seq 0 $XENDOMAINS_STOP_MAXWAIT`; do
# exit if xm save/migrate/shutdown is finished
PSAX=`ps axlw | grep "xm $1" | grep -v grep`
if test -z "$PSAX"; then exit; fi
- echo -n "."; sleep 1
+ if ! test -n "$3"; then echo -n '.'; fi
+ sleep 1
# go to kill immediately if there's only zombies left
if all_zombies && test -n "$2"; then break; fi
done
@@ -312,10 +318,14 @@ watchdog_xm()
read PSF PSUID PSPID PSPPID < <(echo "$PSAX")
# kill xm $1
kill $PSPID >/dev/null 2>&1
+
+ echo -e .
}
stop()
{
+ exec 3>&2 2> /dev/null
+
# Collect list of domains to shut down
if test "$XENDOMAINS_AUTO_ONLY" = "true"; then
rdnames
@@ -333,7 +343,7 @@ stop()
# nothing
;;
(*)
- echo -n '(skip)'
+ echo -e '(skip)'
continue
;;
esac
@@ -345,8 +355,9 @@ stop()
if test -n "$XENDOMAINS_SYSRQ"; then
for sysrq in $XENDOMAINS_SYSRQ; do
echo -n "(SR-$sysrq)"
- xm sysrq $id $sysrq
+ XMR=`xm sysrq $id $sysrq 2>&1 1>/dev/null`
if test $? -ne 0; then
+ echo -e "\nAn error occurred while doing sysrq on
domain:\n$XMR\n"
rc_failed $?
echo -n '!'
fi
@@ -362,13 +373,18 @@ stop()
echo -n "(migr)"
watchdog_xm migrate &
WDOG_PID=$!
- xm migrate $id $XENDOMAINS_MIGRATE
+ XMR=`xm migrate $id $XENDOMAINS_MIGRATE 2>&1 1>/dev/null`
if test $? -ne 0; then
+ echo -e "\nAn error occurred while migrating domain:\n$XMR\n"
rc_failed $?
- echo -n '!'
+ echo -e '!'
+
kill $WDOG_PID >/dev/null 2>&1
else
kill $WDOG_PID >/dev/null 2>&1
+
+ echo -e .
+ usleep 1000
continue
fi
fi
@@ -377,13 +393,16 @@ stop()
watchdog_xm save &
WDOG_PID=$!
mkdir -p "$XENDOMAINS_SAVE"
- xm save $id $XENDOMAINS_SAVE/$name
+ XMR=`xm save $id $XENDOMAINS_SAVE/$name 2>&1 1>/dev/null`
if test $? -ne 0; then
+ echo -e "\nAn error occurred while saving domain:\n$XMR\n"
rc_failed $?
- echo -n '!'
+ echo -e '!'
kill $WDOG_PID >/dev/null 2>&1
else
kill $WDOG_PID >/dev/null 2>&1
+ echo -e .
+ usleep 1000
continue
fi
fi
@@ -392,10 +411,11 @@ stop()
echo -n "(shut)"
watchdog_xm shutdown &
WDOG_PID=$!
- xm shutdown $id $XENDOMAINS_SHUTDOWN
+ XMR=`xm shutdown $id $XENDOMAINS_SHUTDOWN 2>&1 1>/dev/null`
if test $? -ne 0; then
+ echo -e "\nAn error occurred while shutting down
domain:\n$XMR\n"
rc_failed $?
- echo -n '!'
+ echo -e '!'
fi
kill $WDOG_PID >/dev/null 2>&1
fi
@@ -408,18 +428,21 @@ stop()
if ! all_zombies && test -n "$XENDOMAINS_SHUTDOWN_ALL"; then
# XENDOMAINS_SHUTDOWN_ALL should be "--all --halt --wait"
echo -n " SHUTDOWN_ALL "
- watchdog_xm shutdown 1 &
+ watchdog_xm shutdown 1 false &
WDOG_PID=$!
- xm shutdown $XENDOMAINS_SHUTDOWN_ALL
+ XMR=`xm shutdown $XENDOMAINS_SHUTDOWN_ALL 2>&1 1>/dev/null`
if test $? -ne 0; then
+ echo -e "\nAn error occurred while shutting down all domains:
$XMR\n"
rc_failed $?
- echo -n '!'
+ echo -e '!'
fi
kill $WDOG_PID >/dev/null 2>&1
fi
# Unconditionally delete lock file
rm -f $LOCKFILE
+
+ exec 2>&3
}
check_domain_up()
diff -r c4c4ba857d8b -r 6227bf629626 tools/libxc/xc_ptrace.c
--- a/tools/libxc/xc_ptrace.c Mon Mar 02 16:52:22 2009 +0900
+++ b/tools/libxc/xc_ptrace.c Mon Mar 02 18:26:56 2009 +0900
@@ -149,173 +149,6 @@ online_vcpus_changed(uint64_t cpumap)
}
-/* --------------------- */
-/* XXX application state */
-static long nr_pages = 0;
-static uint64_t *page_array = NULL;
-
-static uint64_t to_ma(int cpu, uint64_t maddr)
-{
- return maddr;
-}
-
-static void *
-map_domain_va_32(
- int xc_handle,
- int cpu,
- void *guest_va,
- int perm)
-{
- unsigned long l2e, l1e, l1p, p, va = (unsigned long)guest_va;
- uint32_t *l2, *l1;
- static void *v[MAX_VIRT_CPUS];
-
- l2 = xc_map_foreign_range(
- xc_handle, current_domid, PAGE_SIZE, PROT_READ,
- xen_cr3_to_pfn(ctxt[cpu].c.ctrlreg[3]));
- if ( l2 == NULL )
- return NULL;
-
- l2e = l2[l2_table_offset_i386(va)];
- munmap(l2, PAGE_SIZE);
- if ( !(l2e & _PAGE_PRESENT) )
- return NULL;
- l1p = to_ma(cpu, l2e);
- l1 = xc_map_foreign_range(xc_handle, current_domid, PAGE_SIZE, PROT_READ,
l1p >> PAGE_SHIFT);
- if ( l1 == NULL )
- return NULL;
-
- l1e = l1[l1_table_offset_i386(va)];
- munmap(l1, PAGE_SIZE);
- if ( !(l1e & _PAGE_PRESENT) )
- return NULL;
- p = to_ma(cpu, l1e);
- if ( v[cpu] != NULL )
- munmap(v[cpu], PAGE_SIZE);
- v[cpu] = xc_map_foreign_range(xc_handle, current_domid, PAGE_SIZE, perm, p
>> PAGE_SHIFT);
- if ( v[cpu] == NULL )
- return NULL;
-
- return (void *)((unsigned long)v[cpu] | (va & (PAGE_SIZE - 1)));
-}
-
-
-static void *
-map_domain_va_pae(
- int xc_handle,
- int cpu,
- void *guest_va,
- int perm)
-{
- uint64_t l3e, l2e, l1e, l2p, l1p, p;
- unsigned long va = (unsigned long)guest_va;
- uint64_t *l3, *l2, *l1;
- static void *v[MAX_VIRT_CPUS];
-
- l3 = xc_map_foreign_range(
- xc_handle, current_domid, PAGE_SIZE, PROT_READ,
- xen_cr3_to_pfn(ctxt[cpu].c.ctrlreg[3]));
- if ( l3 == NULL )
- return NULL;
-
- l3e = l3[l3_table_offset_pae(va)];
- munmap(l3, PAGE_SIZE);
- if ( !(l3e & _PAGE_PRESENT) )
- return NULL;
- l2p = to_ma(cpu, l3e);
- l2 = xc_map_foreign_range(xc_handle, current_domid, PAGE_SIZE, PROT_READ,
l2p >> PAGE_SHIFT);
- if ( l2 == NULL )
- return NULL;
-
- l2e = l2[l2_table_offset_pae(va)];
- munmap(l2, PAGE_SIZE);
- if ( !(l2e & _PAGE_PRESENT) )
- return NULL;
- l1p = to_ma(cpu, l2e);
- l1 = xc_map_foreign_range(xc_handle, current_domid, PAGE_SIZE, PROT_READ,
l1p >> PAGE_SHIFT);
- if ( l1 == NULL )
- return NULL;
-
- l1e = l1[l1_table_offset_pae(va)];
- munmap(l1, PAGE_SIZE);
- if ( !(l1e & _PAGE_PRESENT) )
- return NULL;
- p = to_ma(cpu, l1e);
- if ( v[cpu] != NULL )
- munmap(v[cpu], PAGE_SIZE);
- v[cpu] = xc_map_foreign_range(xc_handle, current_domid, PAGE_SIZE, perm, p
>> PAGE_SHIFT);
- if ( v[cpu] == NULL )
- return NULL;
-
- return (void *)((unsigned long)v[cpu] | (va & (PAGE_SIZE - 1)));
-}
-
-#ifdef __x86_64__
-static void *
-map_domain_va_64(
- int xc_handle,
- int cpu,
- void *guest_va,
- int perm)
-{
- unsigned long l4e, l3e, l2e, l1e, l3p, l2p, l1p, p, va = (unsigned
long)guest_va;
- uint64_t *l4, *l3, *l2, *l1;
- static void *v[MAX_VIRT_CPUS];
-
- if ((ctxt[cpu].c.ctrlreg[4] & 0x20) == 0 ) /* legacy ia32 mode */
- return map_domain_va_32(xc_handle, cpu, guest_va, perm);
-
- l4 = xc_map_foreign_range(
- xc_handle, current_domid, PAGE_SIZE, PROT_READ,
- xen_cr3_to_pfn(ctxt[cpu].c.ctrlreg[3]));
- if ( l4 == NULL )
- return NULL;
-
- l4e = l4[l4_table_offset(va)];
- munmap(l4, PAGE_SIZE);
- if ( !(l4e & _PAGE_PRESENT) )
- return NULL;
- l3p = to_ma(cpu, l4e);
- l3 = xc_map_foreign_range(xc_handle, current_domid, PAGE_SIZE, PROT_READ,
l3p >> PAGE_SHIFT);
- if ( l3 == NULL )
- return NULL;
-
- l3e = l3[l3_table_offset(va)];
- munmap(l3, PAGE_SIZE);
- if ( !(l3e & _PAGE_PRESENT) )
- return NULL;
- l2p = to_ma(cpu, l3e);
- l2 = xc_map_foreign_range(xc_handle, current_domid, PAGE_SIZE, PROT_READ,
l2p >> PAGE_SHIFT);
- if ( l2 == NULL )
- return NULL;
-
- l2e = l2[l2_table_offset(va)];
- munmap(l2, PAGE_SIZE);
- if ( !(l2e & _PAGE_PRESENT) )
- return NULL;
- l1p = to_ma(cpu, l2e);
- if (l2e & 0x80) { /* 2M pages */
- p = to_ma(cpu, l1p + (l1_table_offset(va) << PAGE_SHIFT));
- } else { /* 4K pages */
- l1 = xc_map_foreign_range(xc_handle, current_domid, PAGE_SIZE,
PROT_READ, l1p >> PAGE_SHIFT);
- if ( l1 == NULL )
- return NULL;
-
- l1e = l1[l1_table_offset(va)];
- munmap(l1, PAGE_SIZE);
- if ( !(l1e & _PAGE_PRESENT) )
- return NULL;
- p = to_ma(cpu, l1e);
- }
- if ( v[cpu] != NULL )
- munmap(v[cpu], PAGE_SIZE);
- v[cpu] = xc_map_foreign_range(xc_handle, current_domid, PAGE_SIZE, perm, p
>> PAGE_SHIFT);
- if ( v[cpu] == NULL )
- return NULL;
-
- return (void *)((unsigned long)v[cpu] | (va & (PAGE_SIZE - 1)));
-}
-#endif
static void *
map_domain_va(
@@ -324,68 +157,30 @@ map_domain_va(
void *guest_va,
int perm)
{
- unsigned long va = (unsigned long) guest_va;
- long npgs = xc_get_tot_pages(xc_handle, current_domid);
- static enum { MODE_UNKNOWN, MODE_64, MODE_32, MODE_PAE } mode;
-
- if ( mode == MODE_UNKNOWN )
- {
- xen_capabilities_info_t caps;
- (void)xc_version(xc_handle, XENVER_capabilities, caps);
- if ( strstr(caps, "-x86_64") )
- mode = MODE_64;
- else if ( strstr(caps, "-x86_32p") )
- mode = MODE_PAE;
- else if ( strstr(caps, "-x86_32") )
- mode = MODE_32;
- }
-
- if ( nr_pages != npgs )
- {
- if ( nr_pages > 0 )
- free(page_array);
- nr_pages = npgs;
- if ( (page_array = malloc(nr_pages * sizeof(*page_array))) == NULL )
- {
- IPRINTF("Could not allocate memory\n");
- return NULL;
- }
- if ( xc_get_pfn_list(xc_handle, current_domid,
- page_array, nr_pages) != nr_pages )
- {
- IPRINTF("Could not get the page frame list\n");
- return NULL;
- }
- }
-
- if (fetch_regs(xc_handle, cpu, NULL))
+ unsigned long va = (unsigned long)guest_va;
+ unsigned long mfn;
+ void *map;
+
+ /* cross page boundary */
+ if ( (va & ~PAGE_MASK) + sizeof(long) > PAGE_SIZE )
return NULL;
- if (!paging_enabled(&ctxt[cpu])) {
- static void * v;
- uint64_t page;
-
- if ( v != NULL )
- munmap(v, PAGE_SIZE);
-
- page = to_ma(cpu, va);
-
- v = xc_map_foreign_range( xc_handle, current_domid, PAGE_SIZE,
- perm, page >> PAGE_SHIFT);
-
- if ( v == NULL )
- return NULL;
-
- return (void *)(((unsigned long)v) | (va & BSD_PAGE_MASK));
- }
-#ifdef __x86_64__
- if ( mode == MODE_64 )
- return map_domain_va_64(xc_handle, cpu, guest_va, perm);
-#endif
- if ( mode == MODE_PAE )
- return map_domain_va_pae(xc_handle, cpu, guest_va, perm);
- /* else ( mode == MODE_32 ) */
- return map_domain_va_32(xc_handle, cpu, guest_va, perm);
+ mfn = xc_translate_foreign_address(xc_handle, current_domid, cpu, va);
+ if ( mfn == 0 )
+ return NULL;
+
+ map = xc_map_foreign_range(xc_handle, current_domid, PAGE_SIZE,
+ perm, mfn);
+ if (map == NULL)
+ return NULL;
+
+ return map + (va & ~PAGE_MASK);
+}
+
+static void
+unmap_domain_va(void *guest_va)
+{
+ munmap((void *)((unsigned long)guest_va & PAGE_MASK), PAGE_SIZE);
}
int control_c_pressed_flag = 0;
@@ -473,6 +268,8 @@ xc_ptrace(
if ( guest_va == NULL )
goto out_error;
retval = *guest_va;
+ if (!current_isfile)
+ unmap_domain_va(guest_va);
break;
case PTRACE_POKETEXT:
@@ -486,7 +283,9 @@ xc_ptrace(
xc_handle, cpu, addr, PROT_READ|PROT_WRITE);
if ( guest_va == NULL )
goto out_error;
- *guest_va = (unsigned long)data;
+ *guest_va = edata;
+ if (!current_isfile)
+ unmap_domain_va(guest_va);
break;
case PTRACE_GETREGS:
diff -r c4c4ba857d8b -r 6227bf629626 tools/misc/xenpm.c
--- a/tools/misc/xenpm.c Mon Mar 02 16:52:22 2009 +0900
+++ b/tools/misc/xenpm.c Mon Mar 02 18:26:56 2009 +0900
@@ -303,35 +303,34 @@ static void signal_int_handler(int signo
printf("Elapsed time (ms): %"PRIu64"\n", (usec_end - usec_start) / 1000UL);
for ( i = 0; i < max_cpu_nr; i++ )
{
- uint64_t temp;
- printf("CPU%d:\n\tresidency\tpercentage\n", i);
- if ( cx_cap )
+ uint64_t res, triggers;
+ double avg_res;
+
+ printf("\nCPU%d:\tResidency(ms)\t\tAvg Res(ms)\n",i);
+ if ( cx_cap && sum_cx[i] > 0 )
{
for ( j = 0; j < cxstat_end[i].nr; j++ )
{
- if ( sum_cx[i] > 0 )
- {
- temp = cxstat_end[i].residencies[j] -
- cxstat_start[i].residencies[j];
- printf(" C%d\t%"PRIu64" ms\t%.2f%%\n", j,
- temp / 1000000UL, 100UL * temp / (double)sum_cx[i]);
- }
+ res = cxstat_end[i].residencies[j] -
+ cxstat_start[i].residencies[j];
+ triggers = cxstat_end[i].triggers[j] -
+ cxstat_start[i].triggers[j];
+ avg_res = (triggers==0) ? 0: (double)res/triggers/1000000.0;
+ printf(" C%d\t%"PRIu64"\t(%5.2f%%)\t%.2f\n", j, res/1000000UL,
+ 100 * res / (double)sum_cx[i], avg_res );
}
+ printf("\n");
}
- if ( px_cap )
+ if ( px_cap && sum_px[i]>0 )
{
for ( j = 0; j < pxstat_end[i].total; j++ )
{
- if ( sum_px[i] > 0 )
- {
- temp = pxstat_end[i].pt[j].residency -
- pxstat_start[i].pt[j].residency;
- printf(" P%d\t%"PRIu64" ms\t%.2f%%\n", j,
- temp / 1000000UL, 100UL * temp / (double)sum_px[i]);
- }
+ res = pxstat_end[i].pt[j].residency -
+ pxstat_start[i].pt[j].residency;
+ printf(" P%d\t%"PRIu64"\t(%5.2f%%)\n", j,
+ res / 1000000UL, 100UL * res / (double)sum_px[i]);
}
}
- printf("\n");
}
/* some clean up and then exits */
@@ -408,6 +407,7 @@ void start_gather_func(int argc, char *a
free(cxstat);
return ;
}
+ printf("Start sampling, waiting for CTRL-C or SIGINT signal ...\n");
pause();
}
diff -r c4c4ba857d8b -r 6227bf629626 tools/pygrub/src/pygrub
--- a/tools/pygrub/src/pygrub Mon Mar 02 16:52:22 2009 +0900
+++ b/tools/pygrub/src/pygrub Mon Mar 02 18:26:56 2009 +0900
@@ -501,7 +501,7 @@ def get_entry_idx(cf, entry):
return None
-def run_grub(file, entry, fs):
+def run_grub(file, entry, fs, arg):
global g
global sel
@@ -534,7 +534,7 @@ def run_grub(file, entry, fs):
if img.initrd:
grubcfg["ramdisk"] = img.initrd[1]
if img.args:
- grubcfg["args"] = img.args
+ grubcfg["args"] = img.args + " " + arg
return grubcfg
@@ -659,7 +659,7 @@ if __name__ == "__main__":
chosencfg = sniff_solaris(fs, incfg)
if not chosencfg["kernel"]:
- chosencfg = run_grub(file, entry, fs)
+ chosencfg = run_grub(file, entry, fs, incfg["args"])
data = fs.open_file(chosencfg["kernel"]).read()
(tfd, bootcfg["kernel"]) = tempfile.mkstemp(prefix="boot_kernel.",
diff -r c4c4ba857d8b -r 6227bf629626 tools/python/xen/util/vscsi_util.py
--- a/tools/python/xen/util/vscsi_util.py Mon Mar 02 16:52:22 2009 +0900
+++ b/tools/python/xen/util/vscsi_util.py Mon Mar 02 18:26:56 2009 +0900
@@ -78,8 +78,38 @@ def _vscsi_get_hctl_by(phyname, scsi_dev
return (None, None)
-def vscsi_get_scsidevices():
- """ get all scsi devices"""
+def _vscsi_get_scsiid(sg):
+ scsi_id = os.popen('/sbin/scsi_id -gu -s /class/scsi_generic/' +
sg).read().split()
+ if len(scsi_id):
+ return scsi_id[0]
+ return None
+
+
+def _vscsi_get_scsidevices_by_lsscsi(option = ""):
+ """ get all scsi devices information by lsscsi """
+
+ devices = []
+
+ for scsiinfo in os.popen('lsscsi -g %s' % option).readlines():
+ s = scsiinfo.split()
+ hctl = s[0][1:-1]
+ try:
+ devname = s[-2].split('/dev/')[1]
+ except IndexError:
+ devname = None
+ try:
+ sg = s[-1].split('/dev/')[1]
+ scsi_id = _vscsi_get_scsiid(sg)
+ except IndexError:
+ sg = None
+ scsi_id = None
+ devices.append([hctl, devname, sg, scsi_id])
+
+ return devices
+
+
+def _vscsi_get_scsidevices_by_sysfs():
+ """ get all scsi devices information by sysfs """
devices = []
sysfs_mnt = utils.find_sysfs_mount()
@@ -100,18 +130,29 @@ def vscsi_get_scsidevices():
if re.match('^scsi_generic', f):
sg = os.path.basename(realpath)
- lines = os.popen('/sbin/scsi_id -gu -s
/class/scsi_generic/' + sg).read().split()
- if len(lines):
- scsi_id = lines[0]
-
+ scsi_id = _vscsi_get_scsiid(sg)
devices.append([hctl, devname, sg, scsi_id])
return devices
+
+
+def vscsi_get_scsidevices():
+ """ get all scsi devices information """
+
+ devices = _vscsi_get_scsidevices_by_lsscsi("")
+ if devices:
+ return devices
+ return _vscsi_get_scsidevices_by_sysfs()
def vscsi_get_hctl_and_devname_by(target, scsi_devices = None):
if scsi_devices is None:
- scsi_devices = vscsi_get_scsidevices()
+ if len(target.split(':')) == 4:
+ scsi_devices = _vscsi_get_scsidevices_by_lsscsi(target)
+ elif target.startswith('/dev/'):
+ scsi_devices = _vscsi_get_scsidevices_by_lsscsi("| grep %s" %
target)
+ else:
+ scsi_devices = vscsi_get_scsidevices()
if len(target.split(':')) == 4:
return _vscsi_get_devname_by(target, scsi_devices)
diff -r c4c4ba857d8b -r 6227bf629626 tools/python/xen/xend/XendAPI.py
--- a/tools/python/xen/xend/XendAPI.py Mon Mar 02 16:52:22 2009 +0900
+++ b/tools/python/xen/xend/XendAPI.py Mon Mar 02 18:26:56 2009 +0900
@@ -29,6 +29,7 @@ import XendDomain, XendDomainInfo, XendN
import XendDomain, XendDomainInfo, XendNode, XendDmesg
import XendLogging, XendTaskManager, XendAPIStore
+from xen.xend import uuid as genuuid
from XendAPIVersion import *
from XendAuthSessions import instance as auth_manager
from XendError import *
@@ -1867,7 +1868,7 @@ class XendAPI(object):
dom = xendom.get_vm_by_uuid(vbd_struct['VM'])
vdi = xennode.get_vdi_by_uuid(vbd_struct['VDI'])
if not vdi:
- return xen_api_error(['HANDLE_INVALID', 'VDI', vdi_ref])
+ return xen_api_error(['HANDLE_INVALID', 'VDI', vbd_struct['VDI']])
# new VBD via VDI/SR
vdi_image = vdi.get_location()
@@ -2392,7 +2393,7 @@ class XendAPI(object):
tpmif.destroy_vtpmstate(dom.getName())
return xen_api_success_void()
else:
- return xen_api_error(['HANDLE_INVALID', 'VM', vtpm_struct['VM']])
+ return xen_api_error(['HANDLE_INVALID', 'VTPM', vtpm_ref])
# class methods
def VTPM_create(self, session, vtpm_struct):
@@ -2614,7 +2615,7 @@ class XendAPI(object):
return xen_api_success_void()
def event_unregister(self, session, unreg_classes):
- event_unregister(session, reg_classes)
+ event_unregister(session, unreg_classes)
return xen_api_success_void()
def event_next(self, session):
@@ -2641,7 +2642,7 @@ class XendAPI(object):
return xen_api_error(['DEBUG_FAIL', session])
def debug_create(self, session):
- debug_uuid = uuid.createString()
+ debug_uuid = genuuid.createString()
self._debug[debug_uuid] = None
return xen_api_success(debug_uuid)
diff -r c4c4ba857d8b -r 6227bf629626 tools/python/xen/xend/XendBootloader.py
--- a/tools/python/xen/xend/XendBootloader.py Mon Mar 02 16:52:22 2009 +0900
+++ b/tools/python/xen/xend/XendBootloader.py Mon Mar 02 18:26:56 2009 +0900
@@ -67,9 +67,23 @@ def bootloader(blexec, disk, dom, quiet
# listening on the bootloader's fifo for the results.
(m1, s1) = pty.openpty()
- tty.setraw(m1);
- fcntl.fcntl(m1, fcntl.F_SETFL, os.O_NDELAY);
- os.close(s1)
+
+ # On Solaris, the pty master side will get cranky if we try
+ # to write to it while there is no slave. To work around this,
+ # keep the slave descriptor open until we're done. Set it
+ # to raw terminal parameters, otherwise it will echo back
+ # characters, which will confuse the I/O loop below.
+ # Furthermore, a raw master pty device has no terminal
+ # semantics on Solaris, so don't try to set any attributes
+ # for it.
+ if os.uname()[0] != 'SunOS' and os.uname()[0] != 'NetBSD':
+ tty.setraw(m1)
+ os.close(s1)
+ else:
+ tty.setraw(s1)
+
+ fcntl.fcntl(m1, fcntl.F_SETFL, os.O_NDELAY)
+
slavename = ptsname.ptsname(m1)
dom.storeDom("console/tty", slavename)
@@ -108,7 +122,11 @@ def bootloader(blexec, disk, dom, quiet
# record that this domain is bootloading
dom.bootloader_pid = child
- tty.setraw(m2);
+ # On Solaris, the master pty side does not have terminal semantics,
+ # so don't try to set any attributes, as it will fail.
+ if os.uname()[0] != 'SunOS':
+ tty.setraw(m2);
+
fcntl.fcntl(m2, fcntl.F_SETFL, os.O_NDELAY);
while True:
try:
@@ -117,32 +135,55 @@ def bootloader(blexec, disk, dom, quiet
if e.errno == errno.EINTR:
continue
break
+
+ fcntl.fcntl(r, fcntl.F_SETFL, os.O_NDELAY);
+
ret = ""
inbuf=""; outbuf="";
- while True:
- sel = select.select([r, m1, m2], [m1, m2], [])
+ # filedescriptors:
+ # r - input from the bootloader (bootstring output)
+ # m1 - input/output from/to xenconsole
+ # m2 - input/output from/to pty that controls the bootloader
+ # The filedescriptors are NDELAY, so it's ok to try to read
+ # bigger chunks than may be available, to keep e.g. curses
+ # screen redraws in the bootloader efficient. m1 is the side that
+ # gets xenconsole input, which will be keystrokes, so a small number
+ # is sufficient. m2 is pygrub output, which will be curses screen
+ # updates, so a larger number (1024) is appropriate there.
+ #
+ # For writeable descriptors, only include them in the set for select
+ # if there is actual data to write, otherwise this would loop too fast,
+ # eating up CPU time.
+
+ while True:
+ wsel = []
+ if len(outbuf) != 0:
+ wsel = wsel + [m1]
+ if len(inbuf) != 0:
+ wsel = wsel + [m2]
+ sel = select.select([r, m1, m2], wsel, [])
try:
if m1 in sel[0]:
- s = os.read(m1, 1)
+ s = os.read(m1, 16)
inbuf += s
- if m2 in sel[1] and len(inbuf) != 0:
- os.write(m2, inbuf[0])
- inbuf = inbuf[1:]
+ if m2 in sel[1]:
+ n = os.write(m2, inbuf)
+ inbuf = inbuf[n:]
except OSError, e:
if e.errno == errno.EIO:
pass
try:
if m2 in sel[0]:
- s = os.read(m2, 1)
+ s = os.read(m2, 1024)
outbuf += s
- if m1 in sel[1] and len(outbuf) != 0:
- os.write(m1, outbuf[0])
- outbuf = outbuf[1:]
+ if m1 in sel[1]:
+ n = os.write(m1, outbuf)
+ outbuf = outbuf[n:]
except OSError, e:
if e.errno == errno.EIO:
pass
if r in sel[0]:
- s = os.read(r, 1)
+ s = os.read(r, 128)
ret = ret + s
if len(s) == 0:
break
@@ -152,6 +193,8 @@ def bootloader(blexec, disk, dom, quiet
os.close(r)
os.close(m2)
os.close(m1)
+ if os.uname()[0] == 'SunOS' or os.uname()[0] == 'NetBSD':
+ os.close(s1)
os.unlink(fifo)
# Re-acquire the lock to cover the changes we're about to make
diff -r c4c4ba857d8b -r 6227bf629626 tools/python/xen/xend/XendConfig.py
--- a/tools/python/xen/xend/XendConfig.py Mon Mar 02 16:52:22 2009 +0900
+++ b/tools/python/xen/xend/XendConfig.py Mon Mar 02 18:26:56 2009 +0900
@@ -158,6 +158,7 @@ XENAPI_PLATFORM_CFG_TYPES = {
'vncdisplay': int,
'vnclisten': str,
'timer_mode': int,
+ 'vpt_align': int,
'viridian': int,
'vncpasswd': str,
'vncunused': int,
@@ -430,6 +431,8 @@ class XendConfig(dict):
def _vcpus_sanity_check(self):
if 'VCPUs_max' in self and 'vcpu_avail' not in self:
self['vcpu_avail'] = (1 << self['VCPUs_max']) - 1
+ if 'online_vcpus' in self:
+ self['VCPUs_live'] = self['online_vcpus']
def _uuid_sanity_check(self):
"""Make sure UUID is in proper string format with hyphens."""
@@ -459,6 +462,8 @@ class XendConfig(dict):
self['platform']['rtc_timeoffset'] = 0
if 'hpet' not in self['platform']:
self['platform']['hpet'] = 0
+ if 'vpt_align' not in self['platform']:
+ self['platform']['vpt_align'] = 1
if 'loader' not in self['platform']:
# Old configs may have hvmloader set as PV_kernel param
if self.has_key('PV_kernel') and self['PV_kernel'] != '':
@@ -1269,6 +1274,7 @@ class XendConfig(dict):
uuid.createString())
vscsi_dict = self.vscsi_convert_sxp_to_dict(config)
vscsi_devs = vscsi_dict['devs']
+ vscsi_mode = vscsi_dict['feature-host']
# create XenAPI DSCSI objects.
for vscsi_dev in vscsi_devs:
@@ -1283,9 +1289,14 @@ class XendConfig(dict):
}
XendDSCSI(dscsi_uuid, dscsi_record)
- target['devices'][vscsi_devs_uuid] = \
- (dev_type, {'devs': vscsi_devs, 'uuid': vscsi_devs_uuid} )
- log.debug("XendConfig: reading device: %s" % vscsi_devs)
+ vscsi_info = {
+ 'devs': vscsi_devs,
+ 'feature-host': vscsi_mode,
+ 'uuid': vscsi_devs_uuid
+ }
+ target['devices'][vscsi_devs_uuid] = (dev_type, vscsi_info)
+ log.debug("XendConfig: reading device: %s,%s" % \
+ (vscsi_devs, vscsi_mode))
return vscsi_devs_uuid
for opt_val in config[1:]:
@@ -1575,7 +1586,7 @@ class XendConfig(dict):
try:
opt, val = opt_val
pci_dev_info[opt] = val
- except TypeError:
+ except (TypeError, ValueError):
pass
# append uuid for each pci device.
dpci_uuid = pci_dev_info.get('uuid', uuid.createString())
@@ -1609,6 +1620,7 @@ class XendConfig(dict):
#
# [device,
# [vscsi,
+ # [feature-host, 0],
# [dev,
# [devid, 0], [p-devname, sdb], [p-dev, 1:0:0:1],
# [v-dev, 0:0:0:0], [state, 1]
@@ -1619,6 +1631,7 @@ class XendConfig(dict):
# ]
# ],
# [vscsi,
+ # [feature-host, 1],
# [dev,
# [devid, 1], [p-devname, sdg], [p-dev, 2:0:0:0],
# [v-dev, 1:0:0:0], [state, 1]
@@ -1639,6 +1652,7 @@ class XendConfig(dict):
#
# [device,
# [vscsi,
+ # [feature-host, 0],
# [dev,
# [devid, 0], [p-devname, sdd], [p-dev, 1:0:0:3],
# [v-dev, 0:0:0:2], [state, 1]
@@ -1653,7 +1667,8 @@ class XendConfig(dict):
# The Dict looks like this:
#
# { devs: [ {devid: 0, p-devname: sdd, p-dev: 1:0:0:3,
- # v-dev: 0:0:0:2, state: 1} ] }
+ # v-dev: 0:0:0:2, state: 1} ],
+ # feature-host: 1 }
dev_config = {}
@@ -1672,6 +1687,9 @@ class XendConfig(dict):
vscsi_devs.append(vscsi_dev_info)
dev_config['devs'] = vscsi_devs
+ vscsi_mode = sxp.children(dev_sxp, 'feature-host')[0]
+ dev_config['feature-host'] = vscsi_mode[1]
+
return dev_config
def console_add(self, protocol, location, other_config = {}):
@@ -1784,6 +1802,7 @@ class XendConfig(dict):
if dev_type == 'vscsi': # Special case for vscsi
vscsi_dict = self.vscsi_convert_sxp_to_dict(config)
vscsi_devs = vscsi_dict['devs']
+ vscsi_mode = vscsi_dict['feature-host']
# destroy existing XenAPI DSCSI objects
for dscsi_uuid in XendDSCSI.get_by_VM(self['uuid']):
@@ -1802,8 +1821,12 @@ class XendConfig(dict):
}
XendDSCSI(dscsi_uuid, dscsi_record)
- self['devices'][dev_uuid] = \
- (dev_type, {'devs': vscsi_devs, 'uuid': dev_uuid} )
+ vscsi_info = {
+ 'devs': vscsi_devs,
+ 'feature-host': vscsi_mode,
+ 'uuid': dev_uuid
+ }
+ self['devices'][dev_uuid] = (dev_type, vscsi_info)
return True
for opt_val in config[1:]:
@@ -1880,7 +1903,6 @@ class XendConfig(dict):
def all_devices_sxpr(self, target = None):
"""Returns the SXPR for all devices in the current configuration."""
sxprs = []
- pci_devs = []
if target == None:
target = self
@@ -1895,7 +1917,8 @@ class XendConfig(dict):
if dev_type == 'pci':
sxpr = ['pci', ['uuid', dev_info['uuid']]]
elif dev_type == 'vscsi':
- sxpr = ['vscsi', ['uuid', dev_info['uuid']]]
+ sxpr = ['vscsi', ['uuid', dev_info['uuid']],
+ ['feature-host',
dev_info['feature-host']]]
for pci_dev_info in dev_info['devs']:
pci_dev_sxpr = ['dev']
for opt, val in pci_dev_info.items():
diff -r c4c4ba857d8b -r 6227bf629626 tools/python/xen/xend/XendConstants.py
--- a/tools/python/xen/xend/XendConstants.py Mon Mar 02 16:52:22 2009 +0900
+++ b/tools/python/xen/xend/XendConstants.py Mon Mar 02 18:26:56 2009 +0900
@@ -50,6 +50,7 @@ HVM_PARAM_TIMER_MODE = 10
HVM_PARAM_TIMER_MODE = 10
HVM_PARAM_HPET_ENABLED = 11
HVM_PARAM_ACPI_S_STATE = 14
+HVM_PARAM_VPT_ALIGN = 16
restart_modes = [
"restart",
diff -r c4c4ba857d8b -r 6227bf629626 tools/python/xen/xend/XendDomainInfo.py
--- a/tools/python/xen/xend/XendDomainInfo.py Mon Mar 02 16:52:22 2009 +0900
+++ b/tools/python/xen/xend/XendDomainInfo.py Mon Mar 02 18:26:56 2009 +0900
@@ -898,15 +898,21 @@ class XendDomainInfo:
else:
cur_dev_sxp = self._getDeviceInfo_vscsi(req_devid, None)
new_dev_sxp = ['vscsi']
+ cur_mode = sxp.children(cur_dev_sxp, 'feature-host')[0]
+ new_dev_sxp.append(cur_mode)
+
for cur_dev in sxp.children(cur_dev_sxp, 'dev'):
if state == xenbusState['Closing']:
+ if int(cur_mode[1]) == 1:
+ continue
cur_dev_vdev = sxp.child_value(cur_dev, 'v-dev')
if cur_dev_vdev == dev['v-dev']:
continue
new_dev_sxp.append(cur_dev)
if state == xenbusState['Initialising']:
- new_dev_sxp.append(sxp.child0(dev_sxp, 'dev'))
+ for new_dev in sxp.children(dev_sxp, 'dev'):
+ new_dev_sxp.append(new_dev)
dev_uuid = sxp.child_value(cur_dev_sxp, 'uuid')
self.info.device_update(dev_uuid, new_dev_sxp)
@@ -1112,7 +1118,8 @@ class XendDomainInfo:
vscsi_dev.append(['frontstate', None])
vscsi_devs[1].append(vscsi_dev)
dev_num = int(sxp.child_value(vscsi_dev, 'devid'))
- sxprs.append([dev_num, [vscsi_devs]])
+ vscsi_mode = sxp.children(dev_info, 'feature-host')[0]
+ sxprs.append([dev_num, [vscsi_devs, vscsi_mode]])
elif deviceClass == 'vbd':
dev = sxp.child_value(dev_info, 'dev')
if 'ioemu:' in dev:
@@ -2236,6 +2243,12 @@ class XendDomainInfo:
if hvm and hpet is not None:
xc.hvm_set_param(self.domid, HVM_PARAM_HPET_ENABLED,
long(hpet))
+
+ # Optionally enable periodic vpt aligning
+ vpt_align = self.info["platform"].get("vpt_align")
+ if hvm and vpt_align is not None:
+ xc.hvm_set_param(self.domid, HVM_PARAM_VPT_ALIGN,
+ long(vpt_align))
# Set maximum number of vcpus in domain
xc.domain_max_vcpus(self.domid, int(self.info['VCPUs_max']))
diff -r c4c4ba857d8b -r 6227bf629626 tools/python/xen/xend/server/vscsiif.py
--- a/tools/python/xen/xend/server/vscsiif.py Mon Mar 02 16:52:22 2009 +0900
+++ b/tools/python/xen/xend/server/vscsiif.py Mon Mar 02 18:26:56 2009 +0900
@@ -68,6 +68,8 @@ class VSCSIController(DevController):
vscsi_config.append(['devs', devs])
state = self.readFrontend(devid, 'state')
vscsi_config.append(['state', state])
+ hostmode = self.readBackend(devid, 'feature-host')
+ vscsi_config.append(['feature-host', hostmode])
backid = self.readFrontend(devid, 'backend-id')
vscsi_config.append(['backend-id', backid])
backpath = self.readFrontend(devid, 'backend')
@@ -98,6 +100,8 @@ class VSCSIController(DevController):
devid = vscsi_config.get('devid', '')
back[devpath + '/devid'] = str(devid)
+ host_mode = config.get('feature-host','')
+ back['feature-host'] = str(host_mode)
back['uuid'] = config.get('uuid','')
devid = int(devid)
return (devid, back, {})
@@ -133,6 +137,7 @@ class VSCSIController(DevController):
vscsi_devs.append(dev_dict)
config['devs'] = vscsi_devs
+ config['feature-host'] = self.readBackend(devid, 'feature-host')
config['uuid'] = self.readBackend(devid, 'uuid')
return config
@@ -171,6 +176,7 @@ class VSCSIController(DevController):
vscsi_config = config['devs'][0]
state = vscsi_config.get('state', xenbusState['Unknown'])
driver_state = self.readBackend(devid, 'state')
+
if str(xenbusState['Connected']) != driver_state:
raise VmError("Driver status is not connected")
@@ -182,13 +188,20 @@ class VSCSIController(DevController):
elif state == xenbusState['Closing']:
found = False
devs = self.readBackendList(devid, "vscsi-devs")
+ hostmode = int(self.readBackend(devid, 'feature-host'))
vscsipath = "vscsi-devs/"
vdev = vscsi_config.get('v-dev', '')
for dev in devs:
devpath = vscsipath + dev
old_vdev = self.readBackend(devid, devpath + '/v-dev')
- if vdev == old_vdev:
+
+ if hostmode == 1:
+ #At hostmode, all v-dev that belongs to devid is deleted.
+ found = True
+ self.writeBackend(devid, devpath + '/state', \
+ str(xenbusState['Closing']))
+ elif vdev == old_vdev:
found = True
self.writeBackend(devid, devpath + '/state', \
str(xenbusState['Closing']))
diff -r c4c4ba857d8b -r 6227bf629626 tools/python/xen/xm/create.py
--- a/tools/python/xen/xm/create.py Mon Mar 02 16:52:22 2009 +0900
+++ b/tools/python/xen/xm/create.py Mon Mar 02 18:26:56 2009 +0900
@@ -218,6 +218,10 @@ gopts.var('timer_mode', val='TIMER_MODE'
fn=set_int, default=1,
use="""Timer mode (0=delay virtual time when ticks are missed;
1=virtual time is always wallclock time.""")
+
+gopts.var('vpt_align', val='VPT_ALIGN',
+ fn=set_int, default=1,
+ use="Enable aligning all periodic vpt to reduce timer interrupts.")
gopts.var('viridian', val='VIRIDIAN',
fn=set_int, default=0,
@@ -699,32 +703,19 @@ def configure_pci(config_devs, vals):
config_pci.insert(0, 'pci')
config_devs.append(['device', config_pci])
-def vscsi_convert_sxp_to_dict(dev_sxp):
- dev_dict = {}
- for opt_val in dev_sxp[1:]:
- try:
- opt, val = opt_val
- dev_dict[opt] = val
- except TypeError:
- pass
- return dev_dict
-
-def vscsi_lookup_devid(devlist, req_devid):
- if len(devlist) == 0:
- return 0
- else:
- for devid, backend in devlist:
- if devid == req_devid:
- return 1
- return 0
-
def configure_vscsis(config_devs, vals):
"""Create the config for vscsis (virtual scsi devices).
"""
- devidlist = []
- config_scsi = []
+
+ def get_devid(hctl):
+ return int(hctl.split(':')[0])
+
if len(vals.vscsi) == 0:
return 0
+
+ config_scsi = {}
+ pHCTL_list = []
+ vHCTL_list = []
scsi_devices = vscsi_util.vscsi_get_scsidevices()
for (p_dev, v_dev, backend) in vals.vscsi:
@@ -734,34 +725,55 @@ def configure_vscsis(config_devs, vals):
if p_hctl == None:
raise ValueError('Cannot find device "%s"' % p_dev)
- for config in config_scsi:
- dev = vscsi_convert_sxp_to_dict(config)
- if dev['v-dev'] == v_dev:
- raise ValueError('The virtual device "%s" is already defined'
% v_dev)
-
- v_hctl = v_dev.split(':')
- devid = int(v_hctl[0])
- config_scsi.append(['dev', \
- ['state', xenbusState['Initialising']], \
- ['devid', devid], \
- ['p-dev', p_hctl], \
- ['p-devname', devname], \
- ['v-dev', v_dev] ])
-
- if vscsi_lookup_devid(devidlist, devid) == 0:
- devidlist.append([devid, backend])
-
- for devid, backend in devidlist:
- tmp = []
- for config in config_scsi:
- dev = vscsi_convert_sxp_to_dict(config)
- if dev['devid'] == devid:
- tmp.append(config)
-
- tmp.insert(0, 'vscsi')
- if backend:
- tmp.append(['backend', backend])
- config_devs.append(['device', tmp])
+ feature_host = 0
+ if v_dev == 'host':
+ feature_host = 1
+ scsi_info = []
+ devid = get_devid(p_hctl)
+ for (pHCTL, devname, _, _) in scsi_devices:
+ if get_devid(pHCTL) == devid:
+ scsi_info.append([devid, pHCTL, devname, pHCTL])
+ else:
+ scsi_info = [[get_devid(v_dev), p_hctl, devname, v_dev]]
+
+ devid_key = scsi_info[0][0]
+ try:
+ config = config_scsi[devid_key]
+ except KeyError:
+ config = {'feature-host': feature_host, 'backend': backend,
'devs': []}
+
+ devs = config['devs']
+ for (devid, pHCTL, devname, vHCTL) in scsi_info:
+ if pHCTL in pHCTL_list:
+ raise ValueError('The physical device "%s" is already defined'
% pHCTL)
+ if vHCTL in vHCTL_list:
+ raise ValueError('The virtual device "%s" is already defined'
% vHCTL)
+ pHCTL_list.append(pHCTL)
+ vHCTL_list.append(vHCTL)
+ devs.append(['dev', \
+ ['state', xenbusState['Initialising']], \
+ ['devid', devid], \
+ ['p-dev', pHCTL], \
+ ['p-devname', devname], \
+ ['v-dev', vHCTL] ])
+
+ if config['feature-host'] != feature_host:
+ raise ValueError('The physical device "%s" cannot define '
+ 'because mode is different' % scsi_info[0][1])
+ if config['backend'] != backend:
+ raise ValueError('The physical device "%s" cannot define '
+ 'because backend is different' % scsi_info[0][1])
+
+ config['devs'] = devs
+ config_scsi[devid_key] = config
+
+ for config in config_scsi.values():
+ device = ['vscsi', ['feature-host', config['feature-host']]]
+ for dev in config['devs']:
+ device.append(dev)
+ if config['backend']:
+ device.append(['backend', config['backend']])
+ config_devs.append(['device', device])
def configure_ioports(config_devs, vals):
"""Create the config for legacy i/o ranges.
@@ -891,7 +903,8 @@ def configure_hvm(config_image, vals):
'sdl', 'display', 'xauthority', 'rtc_timeoffset', 'monitor',
'acpi', 'apic', 'usb', 'usbdevice', 'keymap', 'pci', 'hpet',
'guest_os_type', 'hap', 'opengl', 'cpuid', 'cpuid_check',
- 'viridian', 'xen_extended_power_mgmt', 'pci_msitranslate' ]
+ 'viridian', 'xen_extended_power_mgmt', 'pci_msitranslate',
+ 'vpt_align' ]
for a in args:
if a in vals.__dict__ and vals.__dict__[a] is not None:
@@ -1039,7 +1052,7 @@ def preprocess_vscsi(vals):
n = len(d)
if n == 2:
tmp = d[1].split(':')
- if len(tmp) != 4:
+ if d[1] != 'host' and len(tmp) != 4:
err('vscsi syntax error "%s"' % d[1])
else:
d.append(None)
diff -r c4c4ba857d8b -r 6227bf629626 tools/python/xen/xm/main.py
--- a/tools/python/xen/xm/main.py Mon Mar 02 16:52:22 2009 +0900
+++ b/tools/python/xen/xm/main.py Mon Mar 02 18:26:56 2009 +0900
@@ -2032,6 +2032,8 @@ def parse_dev_info(info):
'mac' : get_info('mac', str, '??'),
#block-device specific
'ring-ref' : get_info('ring-ref', int, -1),
+ #vscsi specific
+ 'feature-host' : get_info('feature-host', int, -1),
}
def arg_check_for_resource_list(args, name):
@@ -2275,14 +2277,14 @@ def xm_scsi_list(args):
hdr = 0
for x in devs:
if hdr == 0:
- print "%-3s %-3s %-5s %-10s %-5s %-10s %-4s" \
- % ('Idx', 'BE', 'state', 'phy-hctl', 'phy',
'vir-hctl', 'devstate')
+ print "%-3s %-3s %-5s %-4s %-10s %-5s %-10s %-4s" \
+ % ('Idx', 'BE', 'state', 'host', 'phy-hctl', 'phy',
'vir-hctl', 'devstate')
hdr = 1
ni = parse_dev_info(x[1])
ni['idx'] = int(x[0])
for dev in x[1][0][1]:
mi = vscsi_convert_sxp_to_dict(dev)
- print "%(idx)-3d %(backend-id)-3d %(state)-5d " % ni,
+ print "%(idx)-3d %(backend-id)-3d %(state)-5d
%(feature-host)-4d " % ni,
print "%(p-dev)-10s %(p-devname)-5s %(v-dev)-10s
%(frontstate)-4s" % mi
def parse_block_configuration(args):
@@ -2512,27 +2514,46 @@ def xm_pci_attach(args):
server.xend.domain.device_configure(dom, pci)
def parse_scsi_configuration(p_scsi, v_hctl, state):
- v = v_hctl.split(':')
- if len(v) != 4:
- raise OptionError("Invalid argument: %s" % v_hctl)
-
- p_hctl = None
- devname = None
+ def get_devid(hctl):
+ return int(hctl.split(':')[0])
+
+ host_mode = 0
+ scsi_devices = None
+
if p_scsi is not None:
+ # xm scsi-attach
+ if v_hctl == "host":
+ host_mode = 1
+ scsi_devices = vscsi_util.vscsi_get_scsidevices()
+ elif len(v_hctl.split(':')) != 4:
+ raise OptionError("Invalid argument: %s" % v_hctl)
(p_hctl, devname) = \
- vscsi_util.vscsi_get_hctl_and_devname_by(p_scsi)
+ vscsi_util.vscsi_get_hctl_and_devname_by(p_scsi, scsi_devices)
if p_hctl is None:
raise OptionError("Cannot find device '%s'" % p_scsi)
-
- scsi = ['vscsi']
- scsi.append(['dev', \
- ['state', state], \
- ['devid', int(v[0])], \
- ['p-dev', p_hctl], \
- ['p-devname', devname], \
- ['v-dev', v_hctl] \
- ])
-
+ if host_mode:
+ scsi_info = []
+ devid = get_devid(p_hctl)
+ for pHCTL, devname, _, _ in scsi_devices:
+ if get_devid(pHCTL) == devid:
+ scsi_info.append([devid, pHCTL, devname, pHCTL])
+ else:
+ scsi_info = [[get_devid(v_hctl), p_hctl, devname, v_hctl]]
+ else:
+ # xm scsi-detach
+ if len(v_hctl.split(':')) != 4:
+ raise OptionError("Invalid argument: %s" % v_hctl)
+ scsi_info = [[get_devid(v_hctl), None, None, v_hctl]]
+
+ scsi = ['vscsi', ['feature-host', host_mode]]
+ for devid, pHCTL, devname, vHCTL in scsi_info:
+ scsi.append(['dev', \
+ ['state', state], \
+ ['devid', devid], \
+ ['p-dev', pHCTL], \
+ ['p-devname', devname], \
+ ['v-dev', vHCTL] \
+ ])
return scsi
def xm_scsi_attach(args):
diff -r c4c4ba857d8b -r 6227bf629626 tools/python/xen/xm/xenapi_create.py
--- a/tools/python/xen/xm/xenapi_create.py Mon Mar 02 16:52:22 2009 +0900
+++ b/tools/python/xen/xm/xenapi_create.py Mon Mar 02 18:26:56 2009 +0900
@@ -1037,6 +1037,7 @@ class sxp2xml:
'usbdevice',
'hpet',
'timer_mode',
+ 'vpt_align',
'viridian',
'vhpt',
'guest_os_type',
diff -r c4c4ba857d8b -r 6227bf629626 tools/xenstore/xenstored_core.c
--- a/tools/xenstore/xenstored_core.c Mon Mar 02 16:52:22 2009 +0900
+++ b/tools/xenstore/xenstored_core.c Mon Mar 02 18:26:56 2009 +0900
@@ -1937,14 +1937,17 @@ int main(int argc, char *argv[])
handle_event();
next = list_entry(connections.next, typeof(*conn), list);
+ if (&next->list != &connections)
+ talloc_increase_ref_count(next);
while (&next->list != &connections) {
conn = next;
next = list_entry(conn->list.next,
typeof(*conn), list);
+ if (&next->list != &connections)
+ talloc_increase_ref_count(next);
if (conn->domain) {
- talloc_increase_ref_count(conn);
if (domain_can_read(conn))
handle_input(conn);
if (talloc_free(conn) == 0)
@@ -1957,7 +1960,6 @@ int main(int argc, char *argv[])
if (talloc_free(conn) == 0)
continue;
} else {
- talloc_increase_ref_count(conn);
if (FD_ISSET(conn->fd, &inset))
handle_input(conn);
if (talloc_free(conn) == 0)
diff -r c4c4ba857d8b -r 6227bf629626 xen/arch/ia64/xen/domain.c
--- a/xen/arch/ia64/xen/domain.c Mon Mar 02 16:52:22 2009 +0900
+++ b/xen/arch/ia64/xen/domain.c Mon Mar 02 18:26:56 2009 +0900
@@ -2023,6 +2023,7 @@ static void __init calc_dom0_size(void)
unsigned long p2m_pages;
unsigned long spare_hv_pages;
unsigned long max_dom0_size;
+ unsigned long iommu_pg_table_pages = 0;
/* Estimate maximum memory we can safely allocate for dom0
* by subtracting the p2m table allocation and a chunk of memory
@@ -2033,8 +2034,13 @@ static void __init calc_dom0_size(void)
domheap_pages = avail_domheap_pages();
p2m_pages = domheap_pages / PTRS_PER_PTE;
spare_hv_pages = 8192 + (domheap_pages / 4096);
- max_dom0_size = (domheap_pages - (p2m_pages + spare_hv_pages))
- * PAGE_SIZE;
+
+ if (iommu_enabled)
+ iommu_pg_table_pages = domheap_pages * 4 / 512;
+ /* There are 512 ptes in one 4K vtd page. */
+
+ max_dom0_size = (domheap_pages - (p2m_pages + spare_hv_pages) -
+ iommu_pg_table_pages) * PAGE_SIZE;
printk("Maximum permitted dom0 size: %luMB\n",
max_dom0_size / (1024*1024));
diff -r c4c4ba857d8b -r 6227bf629626 xen/arch/x86/acpi/power.c
--- a/xen/arch/x86/acpi/power.c Mon Mar 02 16:52:22 2009 +0900
+++ b/xen/arch/x86/acpi/power.c Mon Mar 02 18:26:56 2009 +0900
@@ -175,6 +175,7 @@ static int enter_state(u32 state)
printk("Entering ACPI S%d state.\n", state);
local_irq_save(flags);
+ spin_debug_disable();
if ( (error = device_power_down()) )
{
@@ -208,6 +209,7 @@ static int enter_state(u32 state)
printk(XENLOG_INFO "Finishing wakeup from ACPI S%d state.", state);
done:
+ spin_debug_enable();
local_irq_restore(flags);
console_end_sync();
acpi_sleep_post(state);
diff -r c4c4ba857d8b -r 6227bf629626 xen/arch/x86/hvm/hvm.c
--- a/xen/arch/x86/hvm/hvm.c Mon Mar 02 16:52:22 2009 +0900
+++ b/xen/arch/x86/hvm/hvm.c Mon Mar 02 18:26:56 2009 +0900
@@ -542,6 +542,22 @@ static int hvm_load_cpu_ctxt(struct doma
return -EINVAL;
}
+ /* Older Xen versions used to save the segment arbytes directly
+ * from the VMCS on Intel hosts. Detect this and rearrange them
+ * into the struct segment_register format. */
+#define UNFOLD_ARBYTES(_r) \
+ if ( (_r & 0xf000) && !(_r & 0x0f00) ) \
+ _r = ((_r & 0xff) | ((_r >> 4) & 0xf00))
+ UNFOLD_ARBYTES(ctxt.cs_arbytes);
+ UNFOLD_ARBYTES(ctxt.ds_arbytes);
+ UNFOLD_ARBYTES(ctxt.es_arbytes);
+ UNFOLD_ARBYTES(ctxt.fs_arbytes);
+ UNFOLD_ARBYTES(ctxt.gs_arbytes);
+ UNFOLD_ARBYTES(ctxt.ss_arbytes);
+ UNFOLD_ARBYTES(ctxt.tr_arbytes);
+ UNFOLD_ARBYTES(ctxt.ldtr_arbytes);
+#undef UNFOLD_ARBYTES
+
/* Architecture-specific vmcs/vmcb bits */
if ( hvm_funcs.load_cpu_ctxt(v, &ctxt) < 0 )
return -EINVAL;
diff -r c4c4ba857d8b -r 6227bf629626 xen/arch/x86/hvm/mtrr.c
--- a/xen/arch/x86/hvm/mtrr.c Mon Mar 02 16:52:22 2009 +0900
+++ b/xen/arch/x86/hvm/mtrr.c Mon Mar 02 18:26:56 2009 +0900
@@ -351,11 +351,18 @@ static uint8_t effective_mm_type(struct
static uint8_t effective_mm_type(struct mtrr_state *m,
uint64_t pat,
paddr_t gpa,
- uint32_t pte_flags)
+ uint32_t pte_flags,
+ uint8_t gmtrr_mtype)
{
uint8_t mtrr_mtype, pat_value, effective;
-
- mtrr_mtype = get_mtrr_type(m, gpa);
+
+ /* if get_pat_flags() gives a dedicated MTRR type,
+ * just use it
+ */
+ if ( gmtrr_mtype == NO_HARDCODE_MEM_TYPE )
+ mtrr_mtype = get_mtrr_type(m, gpa);
+ else
+ mtrr_mtype = gmtrr_mtype;
pat_value = page_pat_type(pat, pte_flags);
@@ -367,7 +374,8 @@ uint32_t get_pat_flags(struct vcpu *v,
uint32_t get_pat_flags(struct vcpu *v,
uint32_t gl1e_flags,
paddr_t gpaddr,
- paddr_t spaddr)
+ paddr_t spaddr,
+ uint8_t gmtrr_mtype)
{
uint8_t guest_eff_mm_type;
uint8_t shadow_mtrr_type;
@@ -378,7 +386,8 @@ uint32_t get_pat_flags(struct vcpu *v,
/* 1. Get the effective memory type of guest physical address,
* with the pair of guest MTRR and PAT
*/
- guest_eff_mm_type = effective_mm_type(g, pat, gpaddr, gl1e_flags);
+ guest_eff_mm_type = effective_mm_type(g, pat, gpaddr,
+ gl1e_flags, gmtrr_mtype);
/* 2. Get the memory type of host physical address, with MTRR */
shadow_mtrr_type = get_mtrr_type(&mtrr_state, spaddr);
diff -r c4c4ba857d8b -r 6227bf629626 xen/arch/x86/hvm/vpic.c
--- a/xen/arch/x86/hvm/vpic.c Mon Mar 02 16:52:22 2009 +0900
+++ b/xen/arch/x86/hvm/vpic.c Mon Mar 02 18:26:56 2009 +0900
@@ -56,7 +56,7 @@ static int vpic_get_priority(struct hvm_
/* prio = ffs(mask ROR vpic->priority_add); */
asm ( "ror %%cl,%b1 ; bsf %1,%0"
- : "=r" (prio) : "r" ((uint32_t)mask), "c" (vpic->priority_add) );
+ : "=r" (prio) : "q" ((uint32_t)mask), "c" (vpic->priority_add) );
return prio;
}
diff -r c4c4ba857d8b -r 6227bf629626 xen/arch/x86/hvm/vpt.c
--- a/xen/arch/x86/hvm/vpt.c Mon Mar 02 16:52:22 2009 +0900
+++ b/xen/arch/x86/hvm/vpt.c Mon Mar 02 18:26:56 2009 +0900
@@ -384,13 +384,25 @@ void create_periodic_time(
pt->period_cycles = (u64)period;
pt->one_shot = !period;
pt->scheduled = NOW() + delta;
- /*
- * Offset LAPIC ticks from other timer ticks. Otherwise guests which use
- * LAPIC ticks for process accounting can see long sequences of process
- * ticks incorrectly accounted to interrupt processing.
- */
- if ( !pt->one_shot && (pt->source == PTSRC_lapic) )
- pt->scheduled += delta >> 1;
+
+ if ( !pt->one_shot )
+ {
+ if ( v->domain->arch.hvm_domain.params[HVM_PARAM_VPT_ALIGN] )
+ {
+ pt->scheduled = align_timer(pt->scheduled, pt->period);
+ }
+ else if ( pt->source == PTSRC_lapic )
+ {
+ /*
+ * Offset LAPIC ticks from other timer ticks. Otherwise guests
+ * which use LAPIC ticks for process accounting can see long
+ * sequences of process ticks incorrectly accounted to interrupt
+ * processing (seen with RHEL3 guest).
+ */
+ pt->scheduled += delta >> 1;
+ }
+ }
+
pt->cb = cb;
pt->priv = data;
diff -r c4c4ba857d8b -r 6227bf629626 xen/arch/x86/io_apic.c
--- a/xen/arch/x86/io_apic.c Mon Mar 02 16:52:22 2009 +0900
+++ b/xen/arch/x86/io_apic.c Mon Mar 02 18:26:56 2009 +0900
@@ -1779,6 +1779,20 @@ static inline void check_timer(void)
*/
#define PIC_IRQS (1 << PIC_CASCADE_IR)
+static struct IO_APIC_route_entry *ioapic_pm_state;
+
+void ioapic_pm_state_alloc(void)
+{
+ int i, nr_entry = 0;
+
+ for (i = 0; i < nr_ioapics; i++)
+ nr_entry += nr_ioapic_registers[i];
+
+ ioapic_pm_state = _xmalloc(sizeof(struct IO_APIC_route_entry)*nr_entry,
+ sizeof(struct IO_APIC_route_entry));
+ BUG_ON(ioapic_pm_state == NULL);
+}
+
void __init setup_IO_APIC(void)
{
enable_IO_APIC();
@@ -1801,40 +1815,16 @@ void __init setup_IO_APIC(void)
init_IO_APIC_traps();
check_timer();
print_IO_APIC();
+ ioapic_pm_state_alloc();
register_keyhandler('z', print_IO_APIC_keyhandler, "print ioapic info");
}
-struct IO_APIC_route_entry *ioapic_pm_state=NULL;
-
-void ioapic_pm_state_alloc(void)
-{
- int i, nr_entry = 0;
-
- if (ioapic_pm_state != NULL)
- return;
-
- for (i = 0; i < nr_ioapics; i++)
- nr_entry += nr_ioapic_registers[i];
-
- ioapic_pm_state = _xmalloc(sizeof(struct IO_APIC_route_entry)*nr_entry,
- sizeof(struct IO_APIC_route_entry));
-}
-
-int ioapic_suspend(void)
-{
- struct IO_APIC_route_entry *entry;
+void ioapic_suspend(void)
+{
+ struct IO_APIC_route_entry *entry = ioapic_pm_state;
unsigned long flags;
- int apic,i;
-
- ioapic_pm_state_alloc();
-
- if (ioapic_pm_state == NULL) {
- printk("Cannot suspend ioapic due to lack of memory\n");
- return 1;
- }
-
- entry = ioapic_pm_state;
+ int apic, i;
spin_lock_irqsave(&ioapic_lock, flags);
for (apic = 0; apic < nr_ioapics; apic++) {
@@ -1844,23 +1834,14 @@ int ioapic_suspend(void)
}
}
spin_unlock_irqrestore(&ioapic_lock, flags);
-
- return 0;
-}
-
-int ioapic_resume(void)
-{
- struct IO_APIC_route_entry *entry;
+}
+
+void ioapic_resume(void)
+{
+ struct IO_APIC_route_entry *entry = ioapic_pm_state;
unsigned long flags;
union IO_APIC_reg_00 reg_00;
- int i,apic;
-
- if (ioapic_pm_state == NULL){
- printk("Cannot resume ioapic due to lack of memory\n");
- return 1;
- }
-
- entry = ioapic_pm_state;
+ int i, apic;
spin_lock_irqsave(&ioapic_lock, flags);
for (apic = 0; apic < nr_ioapics; apic++){
@@ -1875,8 +1856,6 @@ int ioapic_resume(void)
}
}
spin_unlock_irqrestore(&ioapic_lock, flags);
-
- return 0;
}
/* --------------------------------------------------------------------------
diff -r c4c4ba857d8b -r 6227bf629626 xen/arch/x86/irq.c
--- a/xen/arch/x86/irq.c Mon Mar 02 16:52:22 2009 +0900
+++ b/xen/arch/x86/irq.c Mon Mar 02 18:26:56 2009 +0900
@@ -1142,6 +1142,8 @@ void fixup_irqs(cpumask_t map)
static int warned;
irq_guest_action_t *action;
struct pending_eoi *peoi;
+ irq_desc_t *desc;
+ unsigned long flags;
/* Direct all future interrupts away from this CPU. */
for ( vector = 0; vector < NR_VECTORS; vector++ )
@@ -1150,18 +1152,24 @@ void fixup_irqs(cpumask_t map)
if ( vector_to_irq(vector) == 2 )
continue;
- cpus_and(mask, irq_desc[vector].affinity, map);
+ desc = &irq_desc[vector];
+
+ spin_lock_irqsave(&desc->lock, flags);
+
+ cpus_and(mask, desc->affinity, map);
if ( any_online_cpu(mask) == NR_CPUS )
{
printk("Breaking affinity for vector %u (irq %i)\n",
vector, vector_to_irq(vector));
mask = map;
}
- if ( irq_desc[vector].handler->set_affinity )
- irq_desc[vector].handler->set_affinity(vector, mask);
- else if ( irq_desc[vector].action && !(warned++) )
- printk("Cannot set affinity for irq %u (irq %i)\n",
+ if ( desc->handler->set_affinity )
+ desc->handler->set_affinity(vector, mask);
+ else if ( desc->action && !(warned++) )
+ printk("Cannot set affinity for vector %u (irq %i)\n",
vector, vector_to_irq(vector));
+
+ spin_unlock_irqrestore(&desc->lock, flags);
}
/* Service any interrupts that beat us in the re-direction race. */
diff -r c4c4ba857d8b -r 6227bf629626 xen/arch/x86/mm/shadow/multi.c
--- a/xen/arch/x86/mm/shadow/multi.c Mon Mar 02 16:52:22 2009 +0900
+++ b/xen/arch/x86/mm/shadow/multi.c Mon Mar 02 18:26:56 2009 +0900
@@ -546,15 +546,32 @@ _sh_propagate(struct vcpu *v,
!is_xen_heap_mfn(mfn_x(target_mfn)) )
{
unsigned int type;
+
+ /* compute the PAT index for shadow page entry when VT-d is enabled
+ * and device assigned.
+ * 1) direct MMIO: compute the PAT index with gMTRR=UC and gPAT.
+ * 2) if enables snoop control, compute the PAT index as WB.
+ * 3) if disables snoop control, compute the PAT index with
+ * gMTRR and gPAT.
+ */
if ( hvm_get_mem_pinned_cacheattr(d, gfn_x(target_gfn), &type) )
sflags |= pat_type_2_pte_flags(type);
else if ( d->arch.hvm_domain.is_in_uc_mode )
sflags |= pat_type_2_pte_flags(PAT_TYPE_UNCACHABLE);
+ else if ( p2mt == p2m_mmio_direct )
+ sflags |= get_pat_flags(v,
+ gflags,
+ gfn_to_paddr(target_gfn),
+ ((paddr_t)mfn_x(target_mfn)) << PAGE_SHIFT,
+ MTRR_TYPE_UNCACHABLE);
+ else if ( iommu_snoop )
+ sflags |= pat_type_2_pte_flags(PAT_TYPE_WRBACK);
else
sflags |= get_pat_flags(v,
gflags,
gfn_to_paddr(target_gfn),
- ((paddr_t)mfn_x(target_mfn)) <<
PAGE_SHIFT);
+ ((paddr_t)mfn_x(target_mfn)) << PAGE_SHIFT,
+ NO_HARDCODE_MEM_TYPE);
}
// Set the A&D bits for higher level shadows.
diff -r c4c4ba857d8b -r 6227bf629626 xen/arch/x86/msi.c
--- a/xen/arch/x86/msi.c Mon Mar 02 16:52:22 2009 +0900
+++ b/xen/arch/x86/msi.c Mon Mar 02 18:26:56 2009 +0900
@@ -29,17 +29,17 @@
/* bitmap indicate which fixed map is free */
DEFINE_SPINLOCK(msix_fixmap_lock);
-DECLARE_BITMAP(msix_fixmap_pages, MAX_MSIX_PAGES);
+DECLARE_BITMAP(msix_fixmap_pages, FIX_MSIX_MAX_PAGES);
static int msix_fixmap_alloc(void)
{
- int i, rc = -1;
+ int i, rc = -ENOMEM;
spin_lock(&msix_fixmap_lock);
- for ( i = 0; i < MAX_MSIX_PAGES; i++ )
+ for ( i = 0; i < FIX_MSIX_MAX_PAGES; i++ )
if ( !test_bit(i, &msix_fixmap_pages) )
break;
- if ( i == MAX_MSIX_PAGES )
+ if ( i == FIX_MSIX_MAX_PAGES )
goto out;
rc = FIX_MSIX_IO_RESERV_BASE + i;
set_bit(i, &msix_fixmap_pages);
@@ -51,8 +51,66 @@ static int msix_fixmap_alloc(void)
static void msix_fixmap_free(int idx)
{
+ spin_lock(&msix_fixmap_lock);
if ( idx >= FIX_MSIX_IO_RESERV_BASE )
clear_bit(idx - FIX_MSIX_IO_RESERV_BASE, &msix_fixmap_pages);
+ spin_unlock(&msix_fixmap_lock);
+}
+
+static int msix_get_fixmap(struct pci_dev *dev, unsigned long table_paddr,
+ unsigned long entry_paddr)
+{
+ int nr_page, idx;
+
+ nr_page = (entry_paddr >> PAGE_SHIFT) - (table_paddr >> PAGE_SHIFT);
+
+ if ( nr_page < 0 || nr_page >= MAX_MSIX_TABLE_PAGES )
+ return -EINVAL;
+
+ spin_lock(&dev->msix_table_lock);
+ if ( dev->msix_table_refcnt[nr_page]++ == 0 )
+ {
+ idx = msix_fixmap_alloc();
+ if ( idx < 0 )
+ {
+ dev->msix_table_refcnt[nr_page]--;
+ goto out;
+ }
+ set_fixmap_nocache(idx, entry_paddr);
+ dev->msix_table_idx[nr_page] = idx;
+ }
+ else
+ idx = dev->msix_table_idx[nr_page];
+
+ out:
+ spin_unlock(&dev->msix_table_lock);
+ return idx;
+}
+
+static void msix_put_fixmap(struct pci_dev *dev, int idx)
+{
+ int i;
+ unsigned long start;
+
+ spin_lock(&dev->msix_table_lock);
+ for ( i = 0; i < MAX_MSIX_TABLE_PAGES; i++ )
+ {
+ if ( dev->msix_table_idx[i] == idx )
+ break;
+ }
+ if ( i == MAX_MSIX_TABLE_PAGES )
+ goto out;
+
+ if ( --dev->msix_table_refcnt[i] == 0 )
+ {
+ start = fix_to_virt(idx);
+ destroy_xen_mappings(start, start + PAGE_SIZE);
+ msix_fixmap_free(idx);
+ dev->msix_table_idx[i] = 0;
+ }
+
+ out:
+ spin_unlock(&dev->msix_table_lock);
}
/*
@@ -122,8 +180,7 @@ static void read_msi_msg(struct msi_desc
case PCI_CAP_ID_MSIX:
{
void __iomem *base;
- base = entry->mask_base +
- entry->msi_attrib.entry_nr * PCI_MSIX_ENTRY_SIZE;
+ base = entry->mask_base;
msg->address_lo = readl(base + PCI_MSIX_ENTRY_LOWER_ADDR_OFFSET);
msg->address_hi = readl(base + PCI_MSIX_ENTRY_UPPER_ADDR_OFFSET);
@@ -199,8 +256,7 @@ static void write_msi_msg(struct msi_des
case PCI_CAP_ID_MSIX:
{
void __iomem *base;
- base = entry->mask_base +
- entry->msi_attrib.entry_nr * PCI_MSIX_ENTRY_SIZE;
+ base = entry->mask_base;
writel(msg->address_lo,
base + PCI_MSIX_ENTRY_LOWER_ADDR_OFFSET);
@@ -288,8 +344,7 @@ static void msix_flush_writes(unsigned i
break;
case PCI_CAP_ID_MSIX:
{
- int offset = entry->msi_attrib.entry_nr * PCI_MSIX_ENTRY_SIZE +
- PCI_MSIX_ENTRY_VECTOR_CTRL_OFFSET;
+ int offset = PCI_MSIX_ENTRY_VECTOR_CTRL_OFFSET;
readl(entry->mask_base + offset);
break;
}
@@ -330,8 +385,7 @@ static void msi_set_mask_bit(unsigned in
break;
case PCI_CAP_ID_MSIX:
{
- int offset = entry->msi_attrib.entry_nr * PCI_MSIX_ENTRY_SIZE +
- PCI_MSIX_ENTRY_VECTOR_CTRL_OFFSET;
+ int offset = PCI_MSIX_ENTRY_VECTOR_CTRL_OFFSET;
writel(flag, entry->mask_base + offset);
readl(entry->mask_base + offset);
break;
@@ -392,13 +446,10 @@ int msi_free_vector(struct msi_desc *ent
{
unsigned long start;
- writel(1, entry->mask_base + entry->msi_attrib.entry_nr
- * PCI_MSIX_ENTRY_SIZE
- + PCI_MSIX_ENTRY_VECTOR_CTRL_OFFSET);
+ writel(1, entry->mask_base + PCI_MSIX_ENTRY_VECTOR_CTRL_OFFSET);
start = (unsigned long)entry->mask_base & ~(PAGE_SIZE - 1);
- msix_fixmap_free(virt_to_fix(start));
- destroy_xen_mappings(start, start + PAGE_SIZE);
+ msix_put_fixmap(entry->dev, virt_to_fix(start));
}
list_del(&entry->list);
xfree(entry);
@@ -500,8 +551,8 @@ static int msix_capability_init(struct p
struct msi_desc *entry;
int pos;
u16 control;
- unsigned long phys_addr;
- u32 table_offset;
+ unsigned long table_paddr, entry_paddr;
+ u32 table_offset, entry_offset;
u8 bir;
void __iomem *base;
int idx;
@@ -525,15 +576,17 @@ static int msix_capability_init(struct p
table_offset = pci_conf_read32(bus, slot, func,
msix_table_offset_reg(pos));
bir = (u8)(table_offset & PCI_MSIX_FLAGS_BIRMASK);
table_offset &= ~PCI_MSIX_FLAGS_BIRMASK;
- phys_addr = msi->table_base + table_offset;
- idx = msix_fixmap_alloc();
+ entry_offset = msi->entry_nr * PCI_MSIX_ENTRY_SIZE;
+
+ table_paddr = msi->table_base + table_offset;
+ entry_paddr = table_paddr + entry_offset;
+ idx = msix_get_fixmap(dev, table_paddr, entry_paddr);
if ( idx < 0 )
{
xfree(entry);
- return -ENOMEM;
- }
- set_fixmap_nocache(idx, phys_addr);
- base = (void *)(fix_to_virt(idx) + (phys_addr & ((1UL << PAGE_SHIFT) -
1)));
+ return idx;
+ }
+ base = (void *)(fix_to_virt(idx) + (entry_paddr & ((1UL << PAGE_SHIFT) -
1)));
entry->msi_attrib.type = PCI_CAP_ID_MSIX;
entry->msi_attrib.is_64 = 1;
@@ -548,9 +601,7 @@ static int msix_capability_init(struct p
list_add_tail(&entry->list, &dev->msi_list);
/* Mask interrupt here */
- writel(1, entry->mask_base + entry->msi_attrib.entry_nr
- * PCI_MSIX_ENTRY_SIZE
- + PCI_MSIX_ENTRY_VECTOR_CTRL_OFFSET);
+ writel(1, entry->mask_base + PCI_MSIX_ENTRY_VECTOR_CTRL_OFFSET);
*desc = entry;
/* Restore MSI-X enabled bits */
@@ -675,9 +726,7 @@ static void __pci_disable_msix(struct ms
BUG_ON(list_empty(&dev->msi_list));
- writel(1, entry->mask_base + entry->msi_attrib.entry_nr
- * PCI_MSIX_ENTRY_SIZE
- + PCI_MSIX_ENTRY_VECTOR_CTRL_OFFSET);
+ writel(1, entry->mask_base + PCI_MSIX_ENTRY_VECTOR_CTRL_OFFSET);
pci_conf_write16(bus, slot, func, msix_control_reg(pos), control);
}
diff -r c4c4ba857d8b -r 6227bf629626 xen/arch/x86/shutdown.c
--- a/xen/arch/x86/shutdown.c Mon Mar 02 16:52:22 2009 +0900
+++ b/xen/arch/x86/shutdown.c Mon Mar 02 18:26:56 2009 +0900
@@ -302,6 +302,7 @@ void machine_restart(unsigned int delay_
watchdog_disable();
console_start_sync();
+ spin_debug_disable();
local_irq_enable();
diff -r c4c4ba857d8b -r 6227bf629626 xen/arch/x86/time.c
--- a/xen/arch/x86/time.c Mon Mar 02 16:52:22 2009 +0900
+++ b/xen/arch/x86/time.c Mon Mar 02 18:26:56 2009 +0900
@@ -607,13 +607,14 @@ static void platform_time_calibration(vo
{
u64 count;
s_time_t stamp;
-
- spin_lock_irq(&platform_timer_lock);
+ unsigned long flags;
+
+ spin_lock_irqsave(&platform_timer_lock, flags);
count = plt_stamp64 + ((plt_src.read_counter() - plt_stamp) & plt_mask);
stamp = __read_platform_stime(count);
stime_platform_stamp = stamp;
platform_timer_stamp = count;
- spin_unlock_irq(&platform_timer_lock);
+ spin_unlock_irqrestore(&platform_timer_lock, flags);
}
static void resume_platform_timer(void)
diff -r c4c4ba857d8b -r 6227bf629626 xen/common/timer.c
--- a/xen/common/timer.c Mon Mar 02 16:52:22 2009 +0900
+++ b/xen/common/timer.c Mon Mar 02 18:26:56 2009 +0900
@@ -473,6 +473,13 @@ void process_pending_timers(void)
timer_softirq_action();
}
+s_time_t align_timer(s_time_t firsttick, uint64_t period)
+{
+ if ( !period )
+ return firsttick;
+
+ return firsttick + (period - 1) - ((firsttick - 1) % period);
+}
static void dump_timerq(unsigned char key)
{
diff -r c4c4ba857d8b -r 6227bf629626 xen/drivers/passthrough/amd/iommu_init.c
--- a/xen/drivers/passthrough/amd/iommu_init.c Mon Mar 02 16:52:22 2009 +0900
+++ b/xen/drivers/passthrough/amd/iommu_init.c Mon Mar 02 18:26:56 2009 +0900
@@ -37,9 +37,6 @@ struct list_head amd_iommu_head;
struct list_head amd_iommu_head;
struct table_struct device_table;
-extern void *int_remap_table;
-extern spinlock_t int_remap_table_lock;
-
static int __init map_iommu_mmio_region(struct amd_iommu *iommu)
{
unsigned long mfn;
@@ -487,11 +484,13 @@ static int set_iommu_interrupt_handler(s
}
irq_desc[vector].handler = &iommu_msi_type;
+ vector_to_iommu[vector] = iommu;
ret = request_irq_vector(vector, amd_iommu_page_fault, 0,
"amd_iommu", iommu);
if ( ret )
{
irq_desc[vector].handler = &no_irq_type;
+ vector_to_iommu[vector] = NULL;
free_irq_vector(vector);
amd_iov_error("can't request irq\n");
return 0;
@@ -499,7 +498,6 @@ static int set_iommu_interrupt_handler(s
/* Make sure that vector is never re-used. */
vector_irq[vector] = NEVER_ASSIGN_IRQ;
- vector_to_iommu[vector] = iommu;
iommu->vector = vector;
return vector;
}
diff -r c4c4ba857d8b -r 6227bf629626 xen/drivers/passthrough/amd/iommu_intr.c
--- a/xen/drivers/passthrough/amd/iommu_intr.c Mon Mar 02 16:52:22 2009 +0900
+++ b/xen/drivers/passthrough/amd/iommu_intr.c Mon Mar 02 18:26:56 2009 +0900
@@ -23,7 +23,7 @@
#include <asm/hvm/svm/amd-iommu-proto.h>
#define INTREMAP_TABLE_ORDER 1
-DEFINE_SPINLOCK(int_remap_table_lock);
+static DEFINE_SPINLOCK(int_remap_table_lock);
void *int_remap_table = NULL;
static u8 *get_intremap_entry(u8 vector, u8 dm)
@@ -110,21 +110,13 @@ static void update_intremap_entry_from_i
int __init amd_iommu_setup_intremap_table(void)
{
- unsigned long flags;
-
- spin_lock_irqsave(&int_remap_table_lock, flags);
-
if ( int_remap_table == NULL )
{
int_remap_table = __alloc_amd_iommu_tables(INTREMAP_TABLE_ORDER);
if ( int_remap_table == NULL )
- {
- spin_unlock_irqrestore(&int_remap_table_lock, flags);
return -ENOMEM;
- }
memset(int_remap_table, 0, PAGE_SIZE * (1UL << INTREMAP_TABLE_ORDER));
}
- spin_unlock_irqrestore(&int_remap_table_lock, flags);
return 0;
}
@@ -210,15 +202,11 @@ void amd_iommu_msi_msg_update_ire(
int __init deallocate_intremap_table(void)
{
- unsigned long flags;
-
- spin_lock_irqsave(&int_remap_table_lock, flags);
if ( int_remap_table )
{
__free_amd_iommu_tables(int_remap_table, INTREMAP_TABLE_ORDER);
int_remap_table = NULL;
}
- spin_unlock_irqrestore(&int_remap_table_lock, flags);
return 0;
}
diff -r c4c4ba857d8b -r 6227bf629626 xen/drivers/passthrough/amd/iommu_map.c
--- a/xen/drivers/passthrough/amd/iommu_map.c Mon Mar 02 16:52:22 2009 +0900
+++ b/xen/drivers/passthrough/amd/iommu_map.c Mon Mar 02 18:26:56 2009 +0900
@@ -446,14 +446,13 @@ int amd_iommu_map_page(struct domain *d,
int amd_iommu_map_page(struct domain *d, unsigned long gfn, unsigned long mfn)
{
u64 iommu_l2e;
- unsigned long flags;
struct hvm_iommu *hd = domain_hvm_iommu(d);
int iw = IOMMU_IO_WRITE_ENABLED;
int ir = IOMMU_IO_READ_ENABLED;
BUG_ON( !hd->root_table );
- spin_lock_irqsave(&hd->mapping_lock, flags);
+ spin_lock(&hd->mapping_lock);
if ( is_hvm_domain(d) && !hd->p2m_synchronized )
goto out;
@@ -461,14 +460,14 @@ int amd_iommu_map_page(struct domain *d,
iommu_l2e = iommu_l2e_from_pfn(hd->root_table, hd->paging_mode, gfn);
if ( iommu_l2e == 0 )
{
- spin_unlock_irqrestore(&hd->mapping_lock, flags);
+ spin_unlock(&hd->mapping_lock);
amd_iov_error("Invalid IO pagetable entry gfn = %lx\n", gfn);
return -EFAULT;
}
set_iommu_l1e_present(iommu_l2e, gfn, (u64)mfn << PAGE_SHIFT, iw, ir);
out:
- spin_unlock_irqrestore(&hd->mapping_lock, flags);
+ spin_unlock(&hd->mapping_lock);
return 0;
}
@@ -481,11 +480,11 @@ int amd_iommu_unmap_page(struct domain *
BUG_ON( !hd->root_table );
- spin_lock_irqsave(&hd->mapping_lock, flags);
+ spin_lock(&hd->mapping_lock);
if ( is_hvm_domain(d) && !hd->p2m_synchronized )
{
- spin_unlock_irqrestore(&hd->mapping_lock, flags);
+ spin_unlock(&hd->mapping_lock);
return 0;
}
@@ -493,14 +492,14 @@ int amd_iommu_unmap_page(struct domain *
if ( iommu_l2e == 0 )
{
- spin_unlock_irqrestore(&hd->mapping_lock, flags);
+ spin_unlock(&hd->mapping_lock);
amd_iov_error("Invalid IO pagetable entry gfn = %lx\n", gfn);
return -EFAULT;
}
/* mark PTE as 'page not present' */
clear_iommu_l1e_present(iommu_l2e, gfn);
- spin_unlock_irqrestore(&hd->mapping_lock, flags);
+ spin_unlock(&hd->mapping_lock);
/* send INVALIDATE_IOMMU_PAGES command */
for_each_amd_iommu ( iommu )
@@ -520,12 +519,12 @@ int amd_iommu_reserve_domain_unity_map(
unsigned long size, int iw, int ir)
{
u64 iommu_l2e;
- unsigned long flags, npages, i;
+ unsigned long npages, i;
struct hvm_iommu *hd = domain_hvm_iommu(domain);
npages = region_to_pages(phys_addr, size);
- spin_lock_irqsave(&hd->mapping_lock, flags);
+ spin_lock(&hd->mapping_lock);
for ( i = 0; i < npages; ++i )
{
iommu_l2e = iommu_l2e_from_pfn(
@@ -533,7 +532,7 @@ int amd_iommu_reserve_domain_unity_map(
if ( iommu_l2e == 0 )
{
- spin_unlock_irqrestore(&hd->mapping_lock, flags);
+ spin_unlock(&hd->mapping_lock);
amd_iov_error("Invalid IO pagetable entry phys_addr = %lx\n",
phys_addr);
return -EFAULT;
@@ -544,13 +543,13 @@ int amd_iommu_reserve_domain_unity_map(
phys_addr += PAGE_SIZE;
}
- spin_unlock_irqrestore(&hd->mapping_lock, flags);
+ spin_unlock(&hd->mapping_lock);
return 0;
}
int amd_iommu_sync_p2m(struct domain *d)
{
- unsigned long mfn, gfn, flags;
+ unsigned long mfn, gfn;
u64 iommu_l2e;
struct page_info *page;
struct hvm_iommu *hd;
@@ -562,7 +561,7 @@ int amd_iommu_sync_p2m(struct domain *d)
hd = domain_hvm_iommu(d);
- spin_lock_irqsave(&hd->mapping_lock, flags);
+ spin_lock(&hd->mapping_lock);
if ( hd->p2m_synchronized )
goto out;
@@ -582,7 +581,7 @@ int amd_iommu_sync_p2m(struct domain *d)
if ( iommu_l2e == 0 )
{
spin_unlock(&d->page_alloc_lock);
- spin_unlock_irqrestore(&hd->mapping_lock, flags);
+ spin_unlock(&hd->mapping_lock);
amd_iov_error("Invalid IO pagetable entry gfn = %lx\n", gfn);
return -EFAULT;
}
@@ -595,7 +594,7 @@ int amd_iommu_sync_p2m(struct domain *d)
hd->p2m_synchronized = 1;
out:
- spin_unlock_irqrestore(&hd->mapping_lock, flags);
+ spin_unlock(&hd->mapping_lock);
return 0;
}
diff -r c4c4ba857d8b -r 6227bf629626 xen/drivers/passthrough/amd/pci_amd_iommu.c
--- a/xen/drivers/passthrough/amd/pci_amd_iommu.c Mon Mar 02 16:52:22
2009 +0900
+++ b/xen/drivers/passthrough/amd/pci_amd_iommu.c Mon Mar 02 18:26:56
2009 +0900
@@ -172,22 +172,18 @@ static int allocate_domain_resources(str
static int allocate_domain_resources(struct hvm_iommu *hd)
{
/* allocate root table */
- unsigned long flags;
-
- spin_lock_irqsave(&hd->mapping_lock, flags);
+ spin_lock(&hd->mapping_lock);
if ( !hd->root_table )
{
hd->root_table = alloc_amd_iommu_pgtable();
if ( !hd->root_table )
- goto error_out;
- }
- spin_unlock_irqrestore(&hd->mapping_lock, flags);
-
- return 0;
-
- error_out:
- spin_unlock_irqrestore(&hd->mapping_lock, flags);
- return -ENOMEM;
+ {
+ spin_unlock(&hd->mapping_lock);
+ return -ENOMEM;
+ }
+ }
+ spin_unlock(&hd->mapping_lock);
+ return 0;
}
static int get_paging_mode(unsigned long entries)
@@ -298,7 +294,6 @@ static int reassign_device( struct domai
bus, PCI_SLOT(devfn), PCI_FUNC(devfn),
source->domain_id, target->domain_id);
- spin_unlock(&pcidevs_lock);
return 0;
}
@@ -352,11 +347,13 @@ static void deallocate_iommu_page_tables
{
struct hvm_iommu *hd = domain_hvm_iommu(d);
+ spin_lock(&hd->mapping_lock);
if ( hd->root_table )
{
deallocate_next_page_table(hd->root_table, hd->paging_mode);
hd->root_table = NULL;
}
+ spin_unlock(&hd->mapping_lock);
}
diff -r c4c4ba857d8b -r 6227bf629626 xen/drivers/passthrough/pci.c
--- a/xen/drivers/passthrough/pci.c Mon Mar 02 16:52:22 2009 +0900
+++ b/xen/drivers/passthrough/pci.c Mon Mar 02 18:26:56 2009 +0900
@@ -48,6 +48,7 @@ struct pci_dev *alloc_pdev(u8 bus, u8 de
pdev->domain = NULL;
INIT_LIST_HEAD(&pdev->msi_list);
list_add(&pdev->alldevs_list, &alldevs_list);
+ spin_lock_init(&pdev->msix_table_lock);
return pdev;
}
diff -r c4c4ba857d8b -r 6227bf629626 xen/drivers/passthrough/vtd/ia64/vtd.c
--- a/xen/drivers/passthrough/vtd/ia64/vtd.c Mon Mar 02 16:52:22 2009 +0900
+++ b/xen/drivers/passthrough/vtd/ia64/vtd.c Mon Mar 02 18:26:56 2009 +0900
@@ -114,3 +114,34 @@ void hvm_dpci_isairq_eoi(struct domain *
{
/* dummy */
}
+
+static int do_dom0_iommu_mapping(unsigned long start, unsigned long end,
+ void *arg)
+{
+ unsigned long tmp, pfn, j, page_addr = start;
+ struct domain *d = (struct domain *)arg;
+
+ extern int xen_in_range(paddr_t start, paddr_t end);
+ /* Set up 1:1 page table for dom0 for all Ram except Xen bits.*/
+
+ while (page_addr < end)
+ {
+ if (xen_in_range(page_addr, page_addr + PAGE_SIZE))
+ continue;
+
+ pfn = page_addr >> PAGE_SHIFT;
+ tmp = 1 << (PAGE_SHIFT - PAGE_SHIFT_4K);
+ for ( j = 0; j < tmp; j++ )
+ iommu_map_page(d, (pfn*tmp+j), (pfn*tmp+j));
+
+ page_addr += PAGE_SIZE;
+ }
+ return 0;
+}
+
+void iommu_set_dom0_mapping(struct domain *d)
+{
+ if (dom0)
+ BUG_ON(d != dom0);
+ efi_memmap_walk(do_dom0_iommu_mapping, d);
+}
diff -r c4c4ba857d8b -r 6227bf629626 xen/drivers/passthrough/vtd/iommu.c
--- a/xen/drivers/passthrough/vtd/iommu.c Mon Mar 02 16:52:22 2009 +0900
+++ b/xen/drivers/passthrough/vtd/iommu.c Mon Mar 02 18:26:56 2009 +0900
@@ -30,6 +30,7 @@
#include <xen/pci.h>
#include <xen/pci_regs.h>
#include <xen/keyhandler.h>
+#include <asm/msi.h>
#include "iommu.h"
#include "dmar.h"
#include "extern.h"
@@ -40,6 +41,7 @@ static spinlock_t domid_bitmap_lock;
static spinlock_t domid_bitmap_lock; /* protect domain id bitmap */
static int domid_bitmap_size; /* domain id bitmap size in bits */
static unsigned long *domid_bitmap; /* iommu domain id bitmap */
+static bool_t rwbf_quirk;
static void setup_dom0_devices(struct domain *d);
static void setup_dom0_rmrr(struct domain *d);
@@ -230,7 +232,7 @@ static void iommu_flush_write_buffer(str
unsigned long flag;
s_time_t start_time;
- if ( !cap_rwbf(iommu->cap) )
+ if ( !rwbf_quirk && !cap_rwbf(iommu->cap) )
return;
val = iommu->gcmd | DMA_GCMD_WBF;
@@ -829,7 +831,6 @@ static void dma_msi_data_init(struct iom
spin_unlock_irqrestore(&iommu->register_lock, flags);
}
-#ifdef SUPPORT_MSI_REMAPPING
static void dma_msi_addr_init(struct iommu *iommu, int phy_cpu)
{
u64 msi_address;
@@ -846,12 +847,6 @@ static void dma_msi_addr_init(struct iom
dmar_writel(iommu->reg, DMAR_FEUADDR_REG, (u32)(msi_address >> 32));
spin_unlock_irqrestore(&iommu->register_lock, flags);
}
-#else
-static void dma_msi_addr_init(struct iommu *iommu, int phy_cpu)
-{
- /* ia64: TODO */
-}
-#endif
static void dma_msi_set_affinity(unsigned int vector, cpumask_t dest)
{
@@ -870,7 +865,7 @@ static struct hw_interrupt_type dma_msi_
.set_affinity = dma_msi_set_affinity,
};
-int iommu_set_interrupt(struct iommu *iommu)
+static int iommu_set_interrupt(struct iommu *iommu)
{
int vector, ret;
@@ -882,10 +877,12 @@ int iommu_set_interrupt(struct iommu *io
}
irq_desc[vector].handler = &dma_msi_type;
+ vector_to_iommu[vector] = iommu;
ret = request_irq_vector(vector, iommu_page_fault, 0, "dmar", iommu);
if ( ret )
{
irq_desc[vector].handler = &no_irq_type;
+ vector_to_iommu[vector] = NULL;
free_irq_vector(vector);
gdprintk(XENLOG_ERR VTDPREFIX, "IOMMU: can't request irq\n");
return ret;
@@ -893,7 +890,6 @@ int iommu_set_interrupt(struct iommu *io
/* Make sure that vector is never re-used. */
vector_irq[vector] = NEVER_ASSIGN_IRQ;
- vector_to_iommu[vector] = iommu;
return vector;
}
@@ -987,7 +983,6 @@ static int intel_iommu_domain_init(struc
{
struct hvm_iommu *hd = domain_hvm_iommu(d);
struct iommu *iommu = NULL;
- u64 i, j, tmp;
struct acpi_drhd_unit *drhd;
drhd = list_entry(acpi_drhd_units.next, typeof(*drhd), list);
@@ -999,17 +994,8 @@ static int intel_iommu_domain_init(struc
{
extern int xen_in_range(paddr_t start, paddr_t end);
- /* Set up 1:1 page table for dom0 for all RAM except Xen bits. */
- for ( i = 0; i < max_page; i++ )
- {
- if ( !page_is_conventional_ram(i) ||
- xen_in_range(i << PAGE_SHIFT, (i + 1) << PAGE_SHIFT) )
- continue;
-
- tmp = 1 << (PAGE_SHIFT - PAGE_SHIFT_4K);
- for ( j = 0; j < tmp; j++ )
- iommu_map_page(d, (i*tmp+j), (i*tmp+j));
- }
+ /* Set up 1:1 page table for dom0 */
+ iommu_set_dom0_mapping(d);
setup_dom0_devices(d);
setup_dom0_rmrr(d);
@@ -1734,6 +1720,19 @@ static void setup_dom0_rmrr(struct domai
spin_unlock(&pcidevs_lock);
}
+static void platform_quirks(void)
+{
+ u32 id;
+
+ /* Mobile 4 Series Chipset neglects to set RWBF capability. */
+ id = pci_conf_read32(0, 0, 0, 0);
+ if ( id == 0x2a408086 )
+ {
+ dprintk(XENLOG_INFO VTDPREFIX, "DMAR: Forcing write-buffer flush\n");
+ rwbf_quirk = 1;
+ }
+}
+
int intel_vtd_setup(void)
{
struct acpi_drhd_unit *drhd;
@@ -1741,6 +1740,8 @@ int intel_vtd_setup(void)
if ( !vtd_enabled )
return -ENODEV;
+
+ platform_quirks();
spin_lock_init(&domid_bitmap_lock);
clflush_size = get_cache_line_size();
diff -r c4c4ba857d8b -r 6227bf629626 xen/drivers/passthrough/vtd/iommu.h
--- a/xen/drivers/passthrough/vtd/iommu.h Mon Mar 02 16:52:22 2009 +0900
+++ b/xen/drivers/passthrough/vtd/iommu.h Mon Mar 02 18:26:56 2009 +0900
@@ -397,8 +397,8 @@ struct poll_info {
u32 udata;
};
-#define MAX_QINVAL_PAGES 8
#define NUM_QINVAL_PAGES 1
+#define IQA_REG_QS 0 // derived from NUM_QINVAL_PAGES per VT-d spec.
#define QINVAL_ENTRY_NR (PAGE_SIZE_4K*NUM_QINVAL_PAGES/sizeof(struct
qinval_entry))
#define qinval_present(v) ((v).lo & 1)
#define qinval_fault_disable(v) (((v).lo >> 1) & 1)
diff -r c4c4ba857d8b -r 6227bf629626 xen/drivers/passthrough/vtd/qinval.c
--- a/xen/drivers/passthrough/vtd/qinval.c Mon Mar 02 16:52:22 2009 +0900
+++ b/xen/drivers/passthrough/vtd/qinval.c Mon Mar 02 18:26:56 2009 +0900
@@ -317,9 +317,9 @@ int queue_invalidate_iec(struct iommu *i
return ret;
}
-u64 iec_cap;
int __iommu_flush_iec(struct iommu *iommu, u8 granu, u8 im, u16 iidx)
{
+ u64 iec_cap;
int ret;
ret = queue_invalidate_iec(iommu, granu, im, iidx);
ret |= invalidate_sync(iommu);
@@ -445,8 +445,7 @@ int qinval_setup(struct iommu *iommu)
* registers are automatically reset to 0 with write
* to IQA register.
*/
- if ( NUM_QINVAL_PAGES <= MAX_QINVAL_PAGES )
- qi_ctrl->qinval_maddr |= NUM_QINVAL_PAGES - 1;
+ qi_ctrl->qinval_maddr |= IQA_REG_QS;
dmar_writeq(iommu->reg, DMAR_IQA_REG, qi_ctrl->qinval_maddr);
/* enable queued invalidation hardware */
diff -r c4c4ba857d8b -r 6227bf629626 xen/drivers/passthrough/vtd/x86/vtd.c
--- a/xen/drivers/passthrough/vtd/x86/vtd.c Mon Mar 02 16:52:22 2009 +0900
+++ b/xen/drivers/passthrough/vtd/x86/vtd.c Mon Mar 02 18:26:56 2009 +0900
@@ -143,3 +143,23 @@ void hvm_dpci_isairq_eoi(struct domain *
}
spin_unlock(&d->event_lock);
}
+
+void iommu_set_dom0_mapping(struct domain *d)
+{
+ u64 i, j, tmp;
+ extern int xen_in_range(paddr_t start, paddr_t end);
+
+ BUG_ON(d->domain_id != 0);
+
+ for ( i = 0; i < max_page; i++ )
+ {
+ /* Set up 1:1 mapping for dom0 for all RAM except Xen bits. */
+ if ( !page_is_conventional_ram(i) ||
+ xen_in_range(i << PAGE_SHIFT, (i + 1) << PAGE_SHIFT) )
+ continue;
+
+ tmp = 1 << (PAGE_SHIFT - PAGE_SHIFT_4K);
+ for ( j = 0; j < tmp; j++ )
+ iommu_map_page(d, (i*tmp+j), (i*tmp+j));
+ }
+}
diff -r c4c4ba857d8b -r 6227bf629626 xen/include/asm-ia64/msi.h
--- /dev/null Thu Jan 01 00:00:00 1970 +0000
+++ b/xen/include/asm-ia64/msi.h Mon Mar 02 18:26:56 2009 +0900
@@ -0,0 +1,20 @@
+#ifndef __ASM_MSI_H
+#define __ASM_MSI_H
+
+/*
+ * MSI Defined Data Structures
+ */
+#define MSI_ADDRESS_HEADER 0xfee
+#define MSI_ADDRESS_HEADER_SHIFT 12
+#define MSI_ADDRESS_HEADER_MASK 0xfff000
+#define MSI_ADDRESS_DEST_ID_MASK 0xfff0000f
+#define MSI_TARGET_CPU_MASK 0xff
+#define MSI_TARGET_CPU_SHIFT 4
+#define MSI_DELIVERY_MODE 0
+#define MSI_LEVEL_MODE 1 /* Edge always assert */
+#define MSI_TRIGGER_MODE 0 /* MSI is edge sensitive */
+#define MSI_PHYSICAL_MODE 0
+#define MSI_LOGICAL_MODE 1
+#define MSI_REDIRECTION_HINT_MODE 0
+
+#endif /* __ASM_MSI_H */
diff -r c4c4ba857d8b -r 6227bf629626 xen/include/asm-x86/fixmap.h
--- a/xen/include/asm-x86/fixmap.h Mon Mar 02 16:52:22 2009 +0900
+++ b/xen/include/asm-x86/fixmap.h Mon Mar 02 18:26:56 2009 +0900
@@ -50,7 +50,7 @@ enum fixed_addresses {
FIX_IOMMU_MMIO_END = FIX_IOMMU_MMIO_BASE_0 + IOMMU_PAGES -1,
FIX_TBOOT_SHARED_BASE,
FIX_MSIX_IO_RESERV_BASE,
- FIX_MSIX_IO_RESERV_END = FIX_MSIX_IO_RESERV_BASE + MAX_MSIX_PAGES -1,
+ FIX_MSIX_IO_RESERV_END = FIX_MSIX_IO_RESERV_BASE + FIX_MSIX_MAX_PAGES -1,
__end_of_fixed_addresses
};
diff -r c4c4ba857d8b -r 6227bf629626 xen/include/asm-x86/io_apic.h
--- a/xen/include/asm-x86/io_apic.h Mon Mar 02 16:52:22 2009 +0900
+++ b/xen/include/asm-x86/io_apic.h Mon Mar 02 18:26:56 2009 +0900
@@ -182,13 +182,13 @@ extern int timer_uses_ioapic_pin_0;
#endif /*CONFIG_ACPI_BOOT*/
extern int (*ioapic_renumber_irq)(int ioapic, int irq);
-extern int ioapic_suspend(void);
-extern int ioapic_resume(void);
+extern void ioapic_suspend(void);
+extern void ioapic_resume(void);
#else /* !CONFIG_X86_IO_APIC */
#define io_apic_assign_pci_irqs 0
-static inline int ioapic_suspend(void) {return 0};
-static inline int ioapic_resume(void) {return 0};
+static inline void ioapic_suspend(void) {}
+static inline void ioapic_resume(void) {}
#endif
extern int assign_irq_vector(int irq);
diff -r c4c4ba857d8b -r 6227bf629626 xen/include/asm-x86/msi.h
--- a/xen/include/asm-x86/msi.h Mon Mar 02 16:52:22 2009 +0900
+++ b/xen/include/asm-x86/msi.h Mon Mar 02 18:26:56 2009 +0900
@@ -49,9 +49,9 @@
/* MAX fixed pages reserved for mapping MSIX tables. */
#if defined(__x86_64__)
-#define MAX_MSIX_PAGES 512
-#else
-#define MAX_MSIX_PAGES 32
+#define FIX_MSIX_MAX_PAGES 512
+#else
+#define FIX_MSIX_MAX_PAGES 32
#endif
struct msi_info {
@@ -93,7 +93,7 @@ struct msi_desc {
struct list_head list;
- void __iomem *mask_base;
+ void __iomem *mask_base; /* va for the entry in mask table */
struct pci_dev *dev;
int vector;
diff -r c4c4ba857d8b -r 6227bf629626 xen/include/asm-x86/mtrr.h
--- a/xen/include/asm-x86/mtrr.h Mon Mar 02 16:52:22 2009 +0900
+++ b/xen/include/asm-x86/mtrr.h Mon Mar 02 18:26:56 2009 +0900
@@ -11,6 +11,7 @@
#define MTRR_TYPE_WRBACK 6
#define MTRR_NUM_TYPES 7
#define MEMORY_NUM_TYPES MTRR_NUM_TYPES
+#define NO_HARDCODE_MEM_TYPE MTRR_NUM_TYPES
#define NORMAL_CACHE_MODE 0
#define NO_FILL_CACHE_MODE 2
@@ -63,7 +64,7 @@ extern int mtrr_del_page(int reg, unsign
extern int mtrr_del_page(int reg, unsigned long base, unsigned long size);
extern void mtrr_centaur_report_mcr(int mcr, u32 lo, u32 hi);
extern u32 get_pat_flags(struct vcpu *v, u32 gl1e_flags, paddr_t gpaddr,
- paddr_t spaddr);
+ paddr_t spaddr, uint8_t gmtrr_mtype);
extern uint8_t epte_get_entry_emt(
struct domain *d, unsigned long gfn, unsigned long mfn,
uint8_t *igmt, int direct_mmio);
diff -r c4c4ba857d8b -r 6227bf629626 xen/include/public/hvm/params.h
--- a/xen/include/public/hvm/params.h Mon Mar 02 16:52:22 2009 +0900
+++ b/xen/include/public/hvm/params.h Mon Mar 02 18:26:56 2009 +0900
@@ -103,6 +103,9 @@
/* TSS used on Intel when CR0.PE=0. */
#define HVM_PARAM_VM86_TSS 15
-#define HVM_NR_PARAMS 16
+/* Boolean: Enable aligning all periodic vpts to reduce interrupts */
+#define HVM_PARAM_VPT_ALIGN 16
+
+#define HVM_NR_PARAMS 17
#endif /* __XEN_PUBLIC_HVM_PARAMS_H__ */
diff -r c4c4ba857d8b -r 6227bf629626 xen/include/xen/iommu.h
--- a/xen/include/xen/iommu.h Mon Mar 02 16:52:22 2009 +0900
+++ b/xen/include/xen/iommu.h Mon Mar 02 18:26:56 2009 +0900
@@ -114,4 +114,6 @@ void iommu_suspend(void);
void iommu_suspend(void);
void iommu_resume(void);
+void iommu_set_dom0_mapping(struct domain *d);
+
#endif /* _IOMMU_H_ */
diff -r c4c4ba857d8b -r 6227bf629626 xen/include/xen/pci.h
--- a/xen/include/xen/pci.h Mon Mar 02 16:52:22 2009 +0900
+++ b/xen/include/xen/pci.h Mon Mar 02 18:26:56 2009 +0900
@@ -29,10 +29,16 @@
#define PCI_BDF(b,d,f) ((((b) & 0xff) << 8) | PCI_DEVFN(d,f))
#define PCI_BDF2(b,df) ((((b) & 0xff) << 8) | ((df) & 0xff))
+#define MAX_MSIX_TABLE_PAGES 8 /* 2048 entries */
struct pci_dev {
struct list_head alldevs_list;
struct list_head domain_list;
+
struct list_head msi_list;
+ int msix_table_refcnt[MAX_MSIX_TABLE_PAGES];
+ int msix_table_idx[MAX_MSIX_TABLE_PAGES];
+ spinlock_t msix_table_lock;
+
struct domain *domain;
const u8 bus;
const u8 devfn;
diff -r c4c4ba857d8b -r 6227bf629626 xen/include/xen/timer.h
--- a/xen/include/xen/timer.h Mon Mar 02 16:52:22 2009 +0900
+++ b/xen/include/xen/timer.h Mon Mar 02 18:26:56 2009 +0900
@@ -122,6 +122,9 @@ DECLARE_PER_CPU(s_time_t, timer_deadline
/* Arch-defined function to reprogram timer hardware for new deadline. */
extern int reprogram_timer(s_time_t timeout);
+/* calculate the aligned first tick time for a given periodic timer */
+extern s_time_t align_timer(s_time_t firsttick, uint64_t period);
+
#endif /* _TIMER_H_ */
/*
diff -r c4c4ba857d8b -r 6227bf629626 xen/xsm/flask/ss/policydb.c
--- a/xen/xsm/flask/ss/policydb.c Mon Mar 02 16:52:22 2009 +0900
+++ b/xen/xsm/flask/ss/policydb.c Mon Mar 02 18:26:56 2009 +0900
@@ -1515,8 +1515,8 @@ int policydb_read(struct policydb *p, vo
if ( len != strlen(POLICYDB_STRING) )
{
printk(KERN_ERR "security: policydb string length %d does not "
- "match expected length %Zu\n",
- len, (u32) strlen(POLICYDB_STRING));
+ "match expected length %lu\n",
+ len, strlen(POLICYDB_STRING));
goto bad;
}
policydb_str = xmalloc_array(char, len + 1);
_______________________________________________
Xen-changelog mailing list
Xen-changelog@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-changelog
|
Home