# HG changeset patch
# User Alex Williamson <alex.williamson@xxxxxx>
# Date 1206038140 21600
# Node ID 42f6c206c951568c8969c32f643235fe98f44410
# Parent 8c921adf4833a0359775c8f20f9549f6cb11df7b
# Parent c8c8b5c5f6d77a47bfa1869ac60e4975fa764ad6
merge with xen-unstable.hg
---
xen/arch/x86/hvm/iommu.c | 145 --
xen/drivers/passthrough/vtd/io.c | 296 ----
xen/include/acpi/actbl2.h | 232 ---
xen/include/asm-x86/hvm/iommu.h | 56
xen/include/asm-x86/hvm/vmx/intel-iommu.h | 460 -------
xen/include/asm-x86/iommu.h | 109 -
config/StdGNU.mk | 5
config/SunOS.mk | 5
docs/src/interface.tex | 6
extras/mini-os/Makefile | 7
extras/mini-os/README | 34
extras/mini-os/arch/ia64/time.c | 9
extras/mini-os/arch/x86/time.c | 6
extras/mini-os/blkfront.c | 19
extras/mini-os/console/xencons_ring.c | 1
extras/mini-os/daytime.c | 11
extras/mini-os/events.c | 6
extras/mini-os/fbfront.c | 28
extras/mini-os/fs-front.c | 1
extras/mini-os/include/blkfront.h | 2
extras/mini-os/include/fcntl.h | 6
extras/mini-os/include/netfront.h | 2
extras/mini-os/kernel.c | 13
extras/mini-os/lib/sys.c | 53
extras/mini-os/lwip-net.c | 17
extras/mini-os/main.c | 74 -
extras/mini-os/netfront.c | 29
extras/mini-os/xenbus/xenbus.c | 3
stubdom/Makefile | 14
stubdom/README | 2
stubdom/c/Makefile | 9
stubdom/c/main.c | 6
stubdom/stubdom-dm | 18
tools/blktap/drivers/Makefile | 26
tools/blktap/drivers/blktapctrl.c | 104 +
tools/blktap/drivers/block-qcow2.c | 5
tools/blktap/drivers/tapaio.c | 1
tools/blktap/drivers/tapdisk.h | 12
tools/blktap/lib/Makefile | 29
tools/blktap/lib/blktaplib.h | 10
tools/examples/xmexample.hvm | 5
tools/firmware/rombios/rombios.c | 2
tools/include/Makefile | 2
tools/include/xen-sys/Linux/gntdev.h | 14
tools/ioemu/Makefile.target | 4
tools/ioemu/block-vbd.c | 4
tools/ioemu/configure | 32
tools/ioemu/hw/ide.c | 24
tools/ioemu/hw/ne2000.c | 4
tools/ioemu/hw/scsi-disk.c | 5
tools/ioemu/hw/vga.c | 4
tools/ioemu/hw/xen_blktap.c | 686 +++++++++++
tools/ioemu/hw/xen_blktap.h | 57
tools/ioemu/hw/xen_machine_pv.c | 9
tools/ioemu/sdl.c | 171 ++
tools/ioemu/vl.c | 29
tools/ioemu/vl.h | 2
tools/ioemu/vnc.c | 23
tools/ioemu/xenstore.c | 4
tools/libxc/ia64/dom_fw_acpi.c | 1
tools/libxc/ia64/xc_dom_ia64_util.h | 4
tools/libxc/xc_domain_save.c | 40
tools/libxc/xc_linux.c | 15
tools/libxc/xc_minios.c | 14
tools/libxc/xenctrl.h | 16
tools/python/xen/xend/XendConfig.py | 101 +
tools/python/xen/xend/XendDomainInfo.py | 277 +++-
tools/python/xen/xend/image.py | 9
tools/python/xen/xend/server/BlktapController.py | 4
tools/python/xen/xend/server/DevController.py | 50
tools/python/xen/xend/server/SrvDomain.py | 16
tools/python/xen/xend/server/pciif.py | 231 +++
tools/python/xen/xend/server/vfbif.py | 2
tools/python/xen/xm/create.py | 16
tools/python/xen/xm/main.py | 60
tools/python/xen/xm/xenapi_create.py | 2
tools/xcutils/xc_save.c | 1
tools/xenstat/xentop/xentop.c | 1
tools/xenstore/Makefile | 32
xen/arch/ia64/linux-xen/setup.c | 15
xen/arch/ia64/xen/acpi.c | 47
xen/arch/ia64/xen/dom_fw_common.c | 107 -
xen/arch/ia64/xen/pcdp.c | 19
xen/arch/x86/acpi/boot.c | 125 +-
xen/arch/x86/acpi/power.c | 31
xen/arch/x86/domain.c | 11
xen/arch/x86/domctl.c | 2
xen/arch/x86/hvm/Makefile | 1
xen/arch/x86/hvm/hvm.c | 4
xen/arch/x86/hvm/intercept.c | 2
xen/arch/x86/hvm/svm/emulate.c | 4
xen/arch/x86/hvm/svm/svm.c | 10
xen/arch/x86/hvm/svm/vmcb.c | 14
xen/arch/x86/hvm/vlapic.c | 20
xen/arch/x86/hvm/vmx/vmcs.c | 14
xen/arch/x86/hvm/vmx/vmx.c | 14
xen/arch/x86/hvm/vmx/vpmu_core2.c | 16
xen/arch/x86/irq.c | 11
xen/arch/x86/mm.c | 12
xen/arch/x86/mm/p2m.c | 4
xen/arch/x86/mm/paging.c | 11
xen/arch/x86/mm/shadow/private.h | 2
xen/arch/x86/numa.c | 6
xen/arch/x86/srat.c | 4
xen/arch/x86/traps.c | 84 +
xen/arch/x86/x86_emulate.c | 27
xen/common/domain.c | 10
xen/common/event_channel.c | 14
xen/common/keyhandler.c | 6
xen/common/schedule.c | 2
xen/drivers/acpi/Makefile | 3
xen/drivers/acpi/hwregs.c | 385 ++++++
xen/drivers/acpi/numa.c | 2
xen/drivers/acpi/osl.c | 183 ++
xen/drivers/acpi/tables.c | 45
xen/drivers/passthrough/Makefile | 3
xen/drivers/passthrough/amd/iommu_acpi.c | 297 ++--
xen/drivers/passthrough/amd/iommu_detect.c | 168 +-
xen/drivers/passthrough/amd/iommu_init.c | 90 -
xen/drivers/passthrough/amd/iommu_map.c | 32
xen/drivers/passthrough/amd/pci_amd_iommu.c | 78 -
xen/drivers/passthrough/io.c | 272 ++++
xen/drivers/passthrough/iommu.c | 136 ++
xen/drivers/passthrough/vtd/Makefile | 1
xen/drivers/passthrough/vtd/dmar.c | 8
xen/drivers/passthrough/vtd/dmar.h | 3
xen/drivers/passthrough/vtd/extern.h | 2
xen/drivers/passthrough/vtd/intremap.c | 26
xen/drivers/passthrough/vtd/iommu.c | 145 +-
xen/drivers/passthrough/vtd/iommu.h | 454 +++++++
xen/drivers/passthrough/vtd/qinval.c | 10
xen/drivers/passthrough/vtd/utils.c | 20
xen/drivers/passthrough/vtd/vtd.h | 11
xen/drivers/video/vesa.c | 2
xen/include/acpi/acconfig.h | 85 -
xen/include/acpi/acexcep.h | 69 -
xen/include/acpi/acglobal.h | 404 +++---
xen/include/acpi/achware.h | 179 +-
xen/include/acpi/aclocal.h | 1225 +++++++++----------
xen/include/acpi/acmacros.h | 615 +++++-----
xen/include/acpi/acnames.h | 83 +
xen/include/acpi/acobject.h | 657 ++++------
xen/include/acpi/acoutput.h | 32
xen/include/acpi/acpi.h | 36
xen/include/acpi/acpi_bus.h | 381 +++---
xen/include/acpi/acpiosxf.h | 426 ++----
xen/include/acpi/acpixf.h | 679 ++++-------
xen/include/acpi/acstruct.h | 297 ++--
xen/include/acpi/actbl.h | 548 ++++----
xen/include/acpi/actbl1.h | 763 +++++++++++-
xen/include/acpi/actypes.h | 1414 ++++++++++-------------
xen/include/acpi/acutils.h | 1280 +++++++-------------
xen/include/acpi/pdc_intel.h | 33
xen/include/acpi/platform/acenv.h | 171 +-
xen/include/acpi/platform/acgcc.h | 18
xen/include/acpi/platform/aclinux.h | 63 -
xen/include/asm-x86/acpi.h | 134 --
xen/include/asm-x86/apic.h | 1
xen/include/asm-x86/bitops.h | 361 +++--
xen/include/asm-x86/cpufeature.h | 4
xen/include/asm-x86/event.h | 5
xen/include/asm-x86/fixmap.h | 2
xen/include/asm-x86/grant_table.h | 2
xen/include/asm-x86/hvm/domain.h | 4
xen/include/asm-x86/hvm/support.h | 2
xen/include/asm-x86/hvm/svm/vmcb.h | 2
xen/include/asm-x86/hvm/vmx/vmcs.h | 2
xen/include/asm-x86/io_apic.h | 2
xen/include/asm-x86/msr-index.h | 2
xen/include/asm-x86/numa.h | 2
xen/include/asm-x86/shared.h | 8
xen/include/public/arch-x86/hvm/save.h | 3
xen/include/public/arch-x86/xen-x86_64.h | 5
xen/include/public/vcpu.h | 10
xen/include/xen/acpi.h | 133 --
xen/include/xen/hvm/iommu.h | 56
xen/include/xen/iommu.h | 115 +
xen/include/xen/shared.h | 36
178 files changed, 9592 insertions(+), 7626 deletions(-)
diff -r 8c921adf4833 -r 42f6c206c951 config/StdGNU.mk
--- a/config/StdGNU.mk Fri Mar 14 15:07:45 2008 -0600
+++ b/config/StdGNU.mk Thu Mar 20 12:35:40 2008 -0600
@@ -41,9 +41,8 @@ SHLIB_CFLAGS = -shared
SHLIB_CFLAGS = -shared
ifneq ($(debug),y)
-# Optimisation flags are overridable
-CFLAGS ?= -O2 -fomit-frame-pointer
+CFLAGS += -O2 -fomit-frame-pointer
else
# Less than -O1 produces bad code and large stack frames
-CFLAGS ?= -O1 -fno-omit-frame-pointer
+CFLAGS += -O1 -fno-omit-frame-pointer -fno-optimize-sibling-calls
endif
diff -r 8c921adf4833 -r 42f6c206c951 config/SunOS.mk
--- a/config/SunOS.mk Fri Mar 14 15:07:45 2008 -0600
+++ b/config/SunOS.mk Thu Mar 20 12:35:40 2008 -0600
@@ -45,11 +45,10 @@ SHLIB_CFLAGS = -R $(SunOS_LIBDIR) -share
SHLIB_CFLAGS = -R $(SunOS_LIBDIR) -shared
ifneq ($(debug),y)
-# Optimisation flags are overridable
-CFLAGS ?= -O2 -fno-omit-frame-pointer
+CFLAGS += -O2 -fno-omit-frame-pointer
else
# Less than -O1 produces bad code and large stack frames
-CFLAGS ?= -O1 -fno-omit-frame-pointer
+CFLAGS += -O1 -fno-omit-frame-pointer
endif
CFLAGS += -Wa,--divide -D_POSIX_C_SOURCE=200112L -D__EXTENSIONS__
diff -r 8c921adf4833 -r 42f6c206c951 docs/src/interface.tex
--- a/docs/src/interface.tex Fri Mar 14 15:07:45 2008 -0600
+++ b/docs/src/interface.tex Thu Mar 20 12:35:40 2008 -0600
@@ -1,6 +1,6 @@
-\documentclass[11pt,twoside,final,openright]{report}
-\usepackage{a4,graphicx,html,setspace,times}
-\usepackage{comment,parskip}
+\documentclass[11pt,twoside,final,openright,a4paper]{report}
+\usepackage{graphicx,html,setspace,times}
+\usepackage{parskip}
\setstretch{1.15}
% LIBRARY FUNCTIONS
diff -r 8c921adf4833 -r 42f6c206c951 extras/mini-os/Makefile
--- a/extras/mini-os/Makefile Fri Mar 14 15:07:45 2008 -0600
+++ b/extras/mini-os/Makefile Thu Mar 20 12:35:40 2008 -0600
@@ -88,6 +88,11 @@ CFLAGS += -DCONFIG_QEMU
CFLAGS += -DCONFIG_QEMU
endif
+ifneq ($(CDIR),)
+OBJS += $(CDIR)/main.a
+LDLIBS +=
+endif
+
ifeq ($(libc),y)
LDLIBS += -L$(XEN_ROOT)/stubdom/libxc -lxenctrl -lxenguest
LDLIBS += -lpci
@@ -95,7 +100,7 @@ LDLIBS += -lc
LDLIBS += -lc
endif
-ifneq ($(caml)-$(qemu)-$(lwip),--y)
+ifneq ($(caml)-$(qemu)-$(CDIR)-$(lwip),---y)
OBJS := $(filter-out daytime.o, $(OBJS))
endif
diff -r 8c921adf4833 -r 42f6c206c951 extras/mini-os/README
--- a/extras/mini-os/README Fri Mar 14 15:07:45 2008 -0600
+++ b/extras/mini-os/README Thu Mar 20 12:35:40 2008 -0600
@@ -13,18 +13,34 @@ This includes:
* registering virtual interrupt handlers (for timer interrupts)
* a simple page and memory allocator
* minimal libc support
-
-Stuff it doesn't show:
-
- * modifying page tables
- * network code
- * block-device code
-
+ * minimal Copy-on-Write support
+ * network, block, framebuffer support
+ * transparent access to FileSystem exports (see tools/fs-back)
- to build it just type make.
+
+- to build it with TCP/IP support, download LWIP 1.3 source code and type
+
+ make LWIPDIR=/path/to/lwip/source
+
+- to build it with much better libc support, see the stubdom/ directory
- to start it do the following in domain0 (assuming xend is running)
# xm create domain_config
-this starts the kernel and prints out a bunch of stuff and then once
-every second the system time.
+This starts the kernel and prints out a bunch of stuff and then once every
+second the system time.
+
+If you have setup a disk in the config file (e.g.
+disk = [ 'file:/tmp/foo,hda,r' ] ), it will loop reading it. If that disk is
+writable (e.g. disk = [ 'file:/tmp/foo,hda,w' ] ), it will write data patterns
+and re-read them.
+
+If you have setup a network in the config file (e.g. vif = [''] ), it will
+print incoming packets.
+
+If you have setup a VFB in the config file (e.g. vfb = ['type=sdl'] ), it will
+show a mouse with which you can draw color squares.
+
+If you have compiled it with TCP/IP support, it will run a daytime server on
+TCP port 13.
diff -r 8c921adf4833 -r 42f6c206c951 extras/mini-os/arch/ia64/time.c
--- a/extras/mini-os/arch/ia64/time.c Fri Mar 14 15:07:45 2008 -0600
+++ b/extras/mini-os/arch/ia64/time.c Thu Mar 20 12:35:40 2008 -0600
@@ -246,7 +246,7 @@ init_time(void)
{
uint64_t new;
efi_time_t tm;
- int err = 0;
+ evtchn_port_t port = 0;
printk("Initialising time\n");
calculate_frequencies();
@@ -267,11 +267,12 @@ init_time(void)
} else
printk("efi_get_time() failed\n");
- err = bind_virq(VIRQ_ITC, timer_interrupt, NULL);
- if (err == -1) {
- printk("XEN timer request chn bind failed %i\n", err);
+ port = bind_virq(VIRQ_ITC, timer_interrupt, NULL);
+ if (port == -1) {
+ printk("XEN timer request chn bind failed %i\n", port);
return;
}
+ unmask_evtchn(port);
itc_alt = ia64_get_itc();
itc_at_boot = itc_alt;
new = ia64_get_itc() + itm_val;
diff -r 8c921adf4833 -r 42f6c206c951 extras/mini-os/arch/x86/time.c
--- a/extras/mini-os/arch/x86/time.c Fri Mar 14 15:07:45 2008 -0600
+++ b/extras/mini-os/arch/x86/time.c Thu Mar 20 12:35:40 2008 -0600
@@ -222,6 +222,8 @@ static void timer_handler(evtchn_port_t
void init_time(void)
{
+ evtchn_port_t port;
printk("Initialising timer interface\n");
- bind_virq(VIRQ_TIMER, &timer_handler, NULL);
-}
+ port = bind_virq(VIRQ_TIMER, &timer_handler, NULL);
+ unmask_evtchn(port);
+}
diff -r 8c921adf4833 -r 42f6c206c951 extras/mini-os/blkfront.c
--- a/extras/mini-os/blkfront.c Fri Mar 14 15:07:45 2008 -0600
+++ b/extras/mini-os/blkfront.c Thu Mar 20 12:35:40 2008 -0600
@@ -43,7 +43,7 @@ struct blkfront_dev {
struct blkif_front_ring ring;
grant_ref_t ring_ref;
- evtchn_port_t evtchn, local_port;
+ evtchn_port_t evtchn;
blkif_vdev_t handle;
char *nodename;
@@ -70,7 +70,7 @@ void blkfront_handler(evtchn_port_t port
wake_up(&blkfront_queue);
}
-struct blkfront_dev *init_blkfront(char *nodename, uint64_t *sectors, unsigned
*sector_size, int *mode)
+struct blkfront_dev *init_blkfront(char *nodename, uint64_t *sectors, unsigned
*sector_size, int *mode, int *info)
{
xenbus_transaction_t xbt;
char* err;
@@ -92,14 +92,9 @@ struct blkfront_dev *init_blkfront(char
dev = malloc(sizeof(*dev));
dev->nodename = strdup(nodename);
- evtchn_alloc_unbound_t op;
- op.dom = DOMID_SELF;
snprintf(path, sizeof(path), "%s/backend-id", nodename);
- dev->dom = op.remote_dom = xenbus_read_integer(path);
- HYPERVISOR_event_channel_op(EVTCHNOP_alloc_unbound, &op);
- clear_evtchn(op.port); /* Without, handler gets invoked now! */
- dev->local_port = bind_evtchn(op.port, blkfront_handler, dev);
- dev->evtchn=op.port;
+ dev->dom = xenbus_read_integer(path);
+ evtchn_alloc_unbound(dev->dom, blkfront_handler, dev, &dev->evtchn);
s = (struct blkif_sring*) alloc_page();
memset(s,0,PAGE_SIZE);
@@ -181,6 +176,9 @@ done:
xenbus_unwatch_path(XBT_NIL, path);
+ snprintf(path, sizeof(path), "%s/info", dev->backend);
+ *info = xenbus_read_integer(path);
+
snprintf(path, sizeof(path), "%s/sectors", dev->backend);
// FIXME: read_integer returns an int, so disk size limited to 1TB for
now
*sectors = dev->sectors = xenbus_read_integer(path);
@@ -194,6 +192,7 @@ done:
snprintf(path, sizeof(path), "%s/feature-flush-cache", dev->backend);
dev->flush = xenbus_read_integer(path);
}
+ unmask_evtchn(dev->evtchn);
printk("%u sectors of %u bytes\n", dev->sectors, dev->sector_size);
printk("**************************\n");
@@ -219,7 +218,7 @@ void shutdown_blkfront(struct blkfront_d
err = xenbus_printf(XBT_NIL, nodename, "state", "%u", 6);
xenbus_wait_for_value(path,"6");
- unbind_evtchn(dev->local_port);
+ unbind_evtchn(dev->evtchn);
free(nodename);
free(dev->backend);
diff -r 8c921adf4833 -r 42f6c206c951 extras/mini-os/console/xencons_ring.c
--- a/extras/mini-os/console/xencons_ring.c Fri Mar 14 15:07:45 2008 -0600
+++ b/extras/mini-os/console/xencons_ring.c Thu Mar 20 12:35:40 2008 -0600
@@ -86,6 +86,7 @@ int xencons_ring_init(void)
printk("XEN console request chn bind failed %i\n", err);
return err;
}
+ unmask_evtchn(start_info.console.domU.evtchn);
/* In case we have in-flight data after save/restore... */
notify_daemon();
diff -r 8c921adf4833 -r 42f6c206c951 extras/mini-os/daytime.c
--- a/extras/mini-os/daytime.c Fri Mar 14 15:07:45 2008 -0600
+++ b/extras/mini-os/daytime.c Thu Mar 20 12:35:40 2008 -0600
@@ -15,16 +15,19 @@ void run_server(void *p)
void run_server(void *p)
{
struct ip_addr listenaddr = { 0 };
- struct ip_addr ipaddr = { htonl(0x0a000001) };
- struct ip_addr netmask = { htonl(0xff000000) };
- struct ip_addr gw = { 0 };
struct netconn *listener;
struct netconn *session;
struct timeval tv;
err_t rc;
start_networking();
- networking_set_addr(&ipaddr, &netmask, &gw);
+
+ if (0) {
+ struct ip_addr ipaddr = { htonl(0x0a000001) };
+ struct ip_addr netmask = { htonl(0xff000000) };
+ struct ip_addr gw = { 0 };
+ networking_set_addr(&ipaddr, &netmask, &gw);
+ }
tprintk("Opening connection\n");
diff -r 8c921adf4833 -r 42f6c206c951 extras/mini-os/events.c
--- a/extras/mini-os/events.c Fri Mar 14 15:07:45 2008 -0600
+++ b/extras/mini-os/events.c Thu Mar 20 12:35:40 2008 -0600
@@ -86,9 +86,6 @@ evtchn_port_t bind_evtchn(evtchn_port_t
ev_actions[port].data = data;
wmb();
ev_actions[port].handler = handler;
-
- /* Finally unmask the port */
- unmask_evtchn(port);
return port;
}
@@ -191,8 +188,7 @@ int evtchn_bind_interdomain(domid_t pal,
if (err)
return err;
set_bit(op.local_port,bound_ports);
- evtchn_port_t port = op.local_port;
- clear_evtchn(port); /* Without, handler gets invoked now! */
+ evtchn_port_t port = op.local_port;
*local_port = bind_evtchn(port, handler, data);
return err;
}
diff -r 8c921adf4833 -r 42f6c206c951 extras/mini-os/fbfront.c
--- a/extras/mini-os/fbfront.c Fri Mar 14 15:07:45 2008 -0600
+++ b/extras/mini-os/fbfront.c Thu Mar 20 12:35:40 2008 -0600
@@ -26,7 +26,7 @@ struct kbdfront_dev {
domid_t dom;
struct xenkbd_page *page;
- evtchn_port_t evtchn, local_port;
+ evtchn_port_t evtchn;
char *nodename;
char *backend;
@@ -68,14 +68,9 @@ struct kbdfront_dev *init_kbdfront(char
dev = malloc(sizeof(*dev));
dev->nodename = strdup(nodename);
- evtchn_alloc_unbound_t op;
- op.dom = DOMID_SELF;
snprintf(path, sizeof(path), "%s/backend-id", nodename);
- dev->dom = op.remote_dom = xenbus_read_integer(path);
- HYPERVISOR_event_channel_op(EVTCHNOP_alloc_unbound, &op);
- clear_evtchn(op.port); /* Without, handler gets invoked now! */
- dev->local_port = bind_evtchn(op.port, kbdfront_handler, dev);
- dev->evtchn=op.port;
+ dev->dom = xenbus_read_integer(path);
+ evtchn_alloc_unbound(dev->dom, kbdfront_handler, dev, &dev->evtchn);
dev->page = s = (struct xenkbd_page*) alloc_page();
memset(s,0,PAGE_SIZE);
@@ -151,6 +146,7 @@ done:
err = xenbus_printf(XBT_NIL, nodename, "state", "%u", 4); /* connected
*/
}
+ unmask_evtchn(dev->evtchn);
printk("************************** KBDFRONT\n");
@@ -208,7 +204,7 @@ void shutdown_kbdfront(struct kbdfront_d
err = xenbus_printf(XBT_NIL, nodename, "state", "%u", 6);
xenbus_wait_for_value(path,"6");
- unbind_evtchn(dev->local_port);
+ unbind_evtchn(dev->evtchn);
free_pages(dev->page,0);
free(nodename);
@@ -241,7 +237,7 @@ struct fbfront_dev {
domid_t dom;
struct xenfb_page *page;
- evtchn_port_t evtchn, local_port;
+ evtchn_port_t evtchn;
char *nodename;
char *backend;
@@ -281,14 +277,9 @@ struct fbfront_dev *init_fbfront(char *n
dev = malloc(sizeof(*dev));
dev->nodename = strdup(nodename);
- evtchn_alloc_unbound_t op;
- op.dom = DOMID_SELF;
snprintf(path, sizeof(path), "%s/backend-id", nodename);
- dev->dom = op.remote_dom = xenbus_read_integer(path);
- HYPERVISOR_event_channel_op(EVTCHNOP_alloc_unbound, &op);
- clear_evtchn(op.port); /* Without, handler gets invoked now! */
- dev->local_port = bind_evtchn(op.port, fbfront_handler, dev);
- dev->evtchn=op.port;
+ dev->dom = xenbus_read_integer(path);
+ evtchn_alloc_unbound(dev->dom, fbfront_handler, dev, &dev->evtchn);
dev->page = s = (struct xenfb_page*) alloc_page();
memset(s,0,PAGE_SIZE);
@@ -397,6 +388,7 @@ done:
err = xenbus_printf(XBT_NIL, nodename, "state", "%u", 4); /* connected
*/
}
+ unmask_evtchn(dev->evtchn);
printk("************************** FBFRONT\n");
@@ -462,7 +454,7 @@ void shutdown_fbfront(struct fbfront_dev
err = xenbus_printf(XBT_NIL, nodename, "state", "%u", 6);
xenbus_wait_for_value(path,"6");
- unbind_evtchn(dev->local_port);
+ unbind_evtchn(dev->evtchn);
free_pages(dev->page,0);
free(nodename);
diff -r 8c921adf4833 -r 42f6c206c951 extras/mini-os/fs-front.c
--- a/extras/mini-os/fs-front.c Fri Mar 14 15:07:45 2008 -0600
+++ b/extras/mini-os/fs-front.c Thu Mar 20 12:35:40 2008 -0600
@@ -943,6 +943,7 @@ static int init_fs_import(struct fs_impo
//ANY_CPU,
import,
&import->local_port));
+ unmask_evtchn(import->local_port);
self_id = get_self_id();
diff -r 8c921adf4833 -r 42f6c206c951 extras/mini-os/include/blkfront.h
--- a/extras/mini-os/include/blkfront.h Fri Mar 14 15:07:45 2008 -0600
+++ b/extras/mini-os/include/blkfront.h Thu Mar 20 12:35:40 2008 -0600
@@ -15,7 +15,7 @@ struct blkfront_aiocb
void (*aio_cb)(struct blkfront_aiocb *aiocb, int ret);
};
-struct blkfront_dev *init_blkfront(char *nodename, uint64_t *sectors, unsigned
*sector_size, int *mode);
+struct blkfront_dev *init_blkfront(char *nodename, uint64_t *sectors, unsigned
*sector_size, int *mode, int *info);
#ifdef HAVE_LIBC
int blkfront_open(struct blkfront_dev *dev);
#endif
diff -r 8c921adf4833 -r 42f6c206c951 extras/mini-os/include/fcntl.h
--- a/extras/mini-os/include/fcntl.h Fri Mar 14 15:07:45 2008 -0600
+++ b/extras/mini-os/include/fcntl.h Thu Mar 20 12:35:40 2008 -0600
@@ -1,5 +1,9 @@
#ifndef _I386_FCNTL_H
#define _I386_FCNTL_H
+
+#ifdef HAVE_LIBC
+#include_next <fcntl.h>
+#else
/* open/fcntl - O_SYNC is only implemented on blocks devices and on files
located on an ext2 file system */
@@ -90,3 +94,5 @@ int open(const char *path, int flags, ..
int open(const char *path, int flags, ...);
int fcntl(int fd, int cmd, ...);
#endif
+
+#endif
diff -r 8c921adf4833 -r 42f6c206c951 extras/mini-os/include/netfront.h
--- a/extras/mini-os/include/netfront.h Fri Mar 14 15:07:45 2008 -0600
+++ b/extras/mini-os/include/netfront.h Thu Mar 20 12:35:40 2008 -0600
@@ -3,7 +3,7 @@
#include <lwip/netif.h>
#endif
struct netfront_dev;
-struct netfront_dev *init_netfront(char *nodename, void (*netif_rx)(unsigned
char *data, int len), unsigned char rawmac[6]);
+struct netfront_dev *init_netfront(char *nodename, void (*netif_rx)(unsigned
char *data, int len), unsigned char rawmac[6], char **ip);
void netfront_xmit(struct netfront_dev *dev, unsigned char* data,int len);
void shutdown_netfront(struct netfront_dev *dev);
#ifdef HAVE_LIBC
diff -r 8c921adf4833 -r 42f6c206c951 extras/mini-os/kernel.c
--- a/extras/mini-os/kernel.c Fri Mar 14 15:07:45 2008 -0600
+++ b/extras/mini-os/kernel.c Thu Mar 20 12:35:40 2008 -0600
@@ -87,7 +87,7 @@ static void periodic_thread(void *p)
static void netfront_thread(void *p)
{
- init_netfront(NULL, NULL, NULL);
+ init_netfront(NULL, NULL, NULL, NULL);
}
static struct blkfront_dev *blk_dev;
@@ -207,9 +207,18 @@ static void blkfront_thread(void *p)
static void blkfront_thread(void *p)
{
time_t lasttime = 0;
- blk_dev = init_blkfront(NULL, &blk_sectors, &blk_sector_size, &blk_mode);
+ int blk_info;
+
+ blk_dev = init_blkfront(NULL, &blk_sectors, &blk_sector_size, &blk_mode,
&blk_info);
if (!blk_dev)
return;
+
+ if (blk_info & VDISK_CDROM)
+ printk("Block device is a CDROM\n");
+ if (blk_info & VDISK_REMOVABLE)
+ printk("Block device is removable\n");
+ if (blk_info & VDISK_READONLY)
+ printk("Block device is read-only\n");
#ifdef BLKTEST_WRITE
if (blk_mode == O_RDWR) {
diff -r 8c921adf4833 -r 42f6c206c951 extras/mini-os/lib/sys.c
--- a/extras/mini-os/lib/sys.c Fri Mar 14 15:07:45 2008 -0600
+++ b/extras/mini-os/lib/sys.c Thu Mar 20 12:35:40 2008 -0600
@@ -171,8 +171,25 @@ int open(const char *pathname, int flags
printk("open(%s) -> %d\n", pathname, fd);
return fd;
}
- printk("open(%s)", pathname);
- fs_fd = fs_open(fs_import, (void *) pathname);
+ printk("open(%s, %x)", pathname, flags);
+ switch (flags & ~O_ACCMODE) {
+ case 0:
+ fs_fd = fs_open(fs_import, (void *) pathname);
+ break;
+ case O_CREAT|O_TRUNC:
+ {
+ va_list ap;
+ mode_t mode;
+ va_start(ap, flags);
+ mode = va_arg(ap, mode_t);
+ va_end(ap);
+ fs_fd = fs_create(fs_import, (void *) pathname, 0, mode);
+ break;
+ }
+ default:
+ printk(" unsupported flags\n");
+ do_exit();
+ }
if (fs_fd < 0) {
errno = EIO;
return -1;
@@ -819,9 +836,19 @@ int select(int nfds, fd_set *readfds, fd
add_waiter(w4, xenbus_watch_queue);
add_waiter(w5, kbdfront_queue);
- myread = *readfds;
- mywrite = *writefds;
- myexcept = *exceptfds;
+ if (readfds)
+ myread = *readfds;
+ else
+ FD_ZERO(&myread);
+ if (writefds)
+ mywrite = *writefds;
+ else
+ FD_ZERO(&mywrite);
+ if (exceptfds)
+ myexcept = *exceptfds;
+ else
+ FD_ZERO(&myexcept);
+
DEBUG("polling ");
dump_set(nfds, &myread, &mywrite, &myexcept, timeout);
DEBUG("\n");
@@ -860,9 +887,19 @@ int select(int nfds, fd_set *readfds, fd
thread->wakeup_time = stop;
schedule();
- myread = *readfds;
- mywrite = *writefds;
- myexcept = *exceptfds;
+ if (readfds)
+ myread = *readfds;
+ else
+ FD_ZERO(&myread);
+ if (writefds)
+ mywrite = *writefds;
+ else
+ FD_ZERO(&mywrite);
+ if (exceptfds)
+ myexcept = *exceptfds;
+ else
+ FD_ZERO(&myexcept);
+
n = select_poll(nfds, &myread, &mywrite, &myexcept);
if (n) {
diff -r 8c921adf4833 -r 42f6c206c951 extras/mini-os/lwip-net.c
--- a/extras/mini-os/lwip-net.c Fri Mar 14 15:07:45 2008 -0600
+++ b/extras/mini-os/lwip-net.c Thu Mar 20 12:35:40 2008 -0600
@@ -339,10 +339,25 @@ void start_networking(void)
struct ip_addr ipaddr = { htonl(IF_IPADDR) };
struct ip_addr netmask = { htonl(IF_NETMASK) };
struct ip_addr gw = { 0 };
+ char *ip;
tprintk("Waiting for network.\n");
- dev = init_netfront(NULL, NULL, rawmac);
+ dev = init_netfront(NULL, NULL, rawmac, &ip);
+
+ if (ip) {
+ ipaddr.addr = inet_addr(ip);
+ if (IN_CLASSA(ntohl(ipaddr.addr)))
+ netmask.addr = htonl(IN_CLASSA_NET);
+ else if (IN_CLASSB(ntohl(ipaddr.addr)))
+ netmask.addr = htonl(IN_CLASSB_NET);
+ else if (IN_CLASSC(ntohl(ipaddr.addr)))
+ netmask.addr = htonl(IN_CLASSC_NET);
+ else
+ tprintk("Strange IP %s, leaving netmask to 0.\n", ip);
+ }
+ tprintk("IP %x netmask %x gateway %x.\n",
+ ntohl(ipaddr.addr), ntohl(netmask.addr), ntohl(gw.addr));
tprintk("TCP/IP bringup begins.\n");
diff -r 8c921adf4833 -r 42f6c206c951 extras/mini-os/main.c
--- a/extras/mini-os/main.c Fri Mar 14 15:07:45 2008 -0600
+++ b/extras/mini-os/main.c Thu Mar 20 12:35:40 2008 -0600
@@ -42,6 +42,9 @@ static void call_main(void *p)
static void call_main(void *p)
{
char *args, /**path,*/ *msg, *c;
+#ifdef CONFIG_QEMU
+ char *domargs;
+#endif
int argc;
char **argv;
char *envp[] = { NULL };
@@ -63,14 +66,12 @@ static void call_main(void *p)
}
/* Fetch argc, argv from XenStore */
- char domid_s[10];
int domid;
domid = xenbus_read_integer("target");
if (domid == -1) {
printk("Couldn't read target\n");
do_exit();
}
- snprintf(domid_s, sizeof(domid_s), "%d", domid);
snprintf(path, sizeof(path), "/local/domain/%d/vm", domid);
msg = xenbus_read(XBT_NIL, path, &vm);
@@ -78,59 +79,64 @@ static void call_main(void *p)
printk("Couldn't read vm path\n");
do_exit();
}
- printk("vm is at %s\n", vm);
-#else
+ printk("dom vm is at %s\n", vm);
+
+ snprintf(path, sizeof(path), "%s/image/dmargs", vm);
+ free(vm);
+ msg = xenbus_read(XBT_NIL, path, &domargs);
+
+ if (msg) {
+ printk("Couldn't get stubdom args: %s\n", msg);
+ domargs = strdup("");
+ }
+#endif
+
msg = xenbus_read(XBT_NIL, "vm", &vm);
if (msg) {
printk("Couldn't read vm path\n");
do_exit();
}
-#endif
- snprintf(path, sizeof(path), "%s/image/dmargs", vm);
+ printk("my vm is at %s\n", vm);
+ snprintf(path, sizeof(path), "%s/image/cmdline", vm);
free(vm);
msg = xenbus_read(XBT_NIL, path, &args);
if (msg) {
- printk("Couldn't get stubdom args: %s\n", msg);
+ printk("Couldn't get my args: %s\n", msg);
args = strdup("");
}
argc = 1;
+
+#define PARSE_ARGS(ARGS,START,END) \
+ c = ARGS; \
+ while (*c) { \
+ if (*c != ' ') { \
+ START; \
+ while (*c && *c != ' ') \
+ c++; \
+ } else { \
+ END; \
+ while (*c == ' ') \
+ c++; \
+ } \
+ }
+
+ PARSE_ARGS(args, argc++, );
#ifdef CONFIG_QEMU
- argc += 2;
+ PARSE_ARGS(domargs, argc++, );
#endif
- c = args;
- while (*c) {
- if (*c != ' ') {
- argc++;
- while (*c && *c != ' ')
- c++;
- } else {
- while (*c == ' ')
- c++;
- }
- }
+
argv = alloca((argc + 1) * sizeof(char *));
argv[0] = "main";
argc = 1;
+
+ PARSE_ARGS(args, argv[argc++] = c, *c++ = 0)
#ifdef CONFIG_QEMU
- argv[1] = "-d";
- argv[2] = domid_s;
- argc += 2;
+ PARSE_ARGS(domargs, argv[argc++] = c, *c++ = 0)
#endif
- c = args;
- while (*c) {
- if (*c != ' ') {
- argv[argc++] = c;
- while (*c && *c != ' ')
- c++;
- } else {
- *c++ = 0;
- while (*c == ' ')
- c++;
- }
- }
+
argv[argc] = NULL;
for (i = 0; i < argc; i++)
diff -r 8c921adf4833 -r 42f6c206c951 extras/mini-os/netfront.c
--- a/extras/mini-os/netfront.c Fri Mar 14 15:07:45 2008 -0600
+++ b/extras/mini-os/netfront.c Thu Mar 20 12:35:40 2008 -0600
@@ -48,7 +48,7 @@ struct netfront_dev {
struct netif_rx_front_ring rx;
grant_ref_t tx_ring_ref;
grant_ref_t rx_ring_ref;
- evtchn_port_t evtchn, local_port;
+ evtchn_port_t evtchn;
char *nodename;
char *backend;
@@ -259,7 +259,7 @@ void netfront_select_handler(evtchn_port
}
#endif
-struct netfront_dev *init_netfront(char *nodename, void
(*thenetif_rx)(unsigned char* data, int len), unsigned char rawmac[6])
+struct netfront_dev *init_netfront(char *nodename, void
(*thenetif_rx)(unsigned char* data, int len), unsigned char rawmac[6], char
**ip)
{
xenbus_transaction_t xbt;
char* err;
@@ -301,19 +301,14 @@ struct netfront_dev *init_netfront(char
dev->rx_buffers[i].page = (char*)alloc_page();
}
- evtchn_alloc_unbound_t op;
- op.dom = DOMID_SELF;
snprintf(path, sizeof(path), "%s/backend-id", nodename);
- dev->dom = op.remote_dom = xenbus_read_integer(path);
- HYPERVISOR_event_channel_op(EVTCHNOP_alloc_unbound, &op);
- clear_evtchn(op.port); /* Without, handler gets invoked now! */
+ dev->dom = xenbus_read_integer(path);
#ifdef HAVE_LIBC
if (thenetif_rx == NETIF_SELECT_RX)
- dev->local_port = bind_evtchn(op.port, netfront_select_handler, dev);
+ evtchn_alloc_unbound(dev->dom, netfront_select_handler, dev,
&dev->evtchn);
else
#endif
- dev->local_port = bind_evtchn(op.port, netfront_handler, dev);
- dev->evtchn=op.port;
+ evtchn_alloc_unbound(dev->dom, netfront_handler, dev, &dev->evtchn);
txs = (struct netif_tx_sring*) alloc_page();
rxs = (struct netif_rx_sring *) alloc_page();
@@ -388,9 +383,9 @@ done:
msg = xenbus_read(XBT_NIL, path, &mac);
if ((dev->backend == NULL) || (mac == NULL)) {
- struct evtchn_close op = { dev->local_port };
+ struct evtchn_close op = { dev->evtchn };
printk("%s: backend/mac failed\n", __func__);
- unbind_evtchn(dev->local_port);
+ unbind_evtchn(dev->evtchn);
HYPERVISOR_event_channel_op(EVTCHNOP_close, &op);
return NULL;
}
@@ -407,11 +402,17 @@ done:
xenbus_wait_for_value(path,"4");
xenbus_unwatch_path(XBT_NIL, path);
+
+ if (ip) {
+ snprintf(path, sizeof(path), "%s/ip", dev->backend);
+ xenbus_read(XBT_NIL, path, ip);
+ }
}
printk("**************************\n");
init_rx_buffers(dev);
+ unmask_evtchn(dev->evtchn);
/* Special conversion specifier 'hh' needed for __ia64__. Without
this mini-os panics with 'Unaligned reference'. */
@@ -431,7 +432,7 @@ int netfront_tap_open(char *nodename) {
int netfront_tap_open(char *nodename) {
struct netfront_dev *dev;
- dev = init_netfront(nodename, NETIF_SELECT_RX, NULL);
+ dev = init_netfront(nodename, NETIF_SELECT_RX, NULL, NULL);
if (!dev) {
printk("TAP open failed\n");
errno = EIO;
@@ -460,7 +461,7 @@ void shutdown_netfront(struct netfront_d
err = xenbus_printf(XBT_NIL, nodename, "state", "%u", 6);
xenbus_wait_for_value(path,"6");
- unbind_evtchn(dev->local_port);
+ unbind_evtchn(dev->evtchn);
free(nodename);
free(dev->backend);
diff -r 8c921adf4833 -r 42f6c206c951 extras/mini-os/xenbus/xenbus.c
--- a/extras/mini-os/xenbus/xenbus.c Fri Mar 14 15:07:45 2008 -0600
+++ b/extras/mini-os/xenbus/xenbus.c Thu Mar 20 12:35:40 2008 -0600
@@ -257,6 +257,7 @@ void init_xenbus(void)
err = bind_evtchn(start_info.store_evtchn,
xenbus_evtchn_handler,
NULL);
+ unmask_evtchn(start_info.store_evtchn);
DEBUG("xenbus on irq %d\n", err);
}
@@ -452,7 +453,7 @@ char *xenbus_write(xenbus_transaction_t
{
struct write_req req[] = {
{path, strlen(path) + 1},
- {value, strlen(value) + 1},
+ {value, strlen(value)},
};
struct xsd_sockmsg *rep;
rep = xenbus_msg_reply(XS_WRITE, xbt, req, ARRAY_SIZE(req));
diff -r 8c921adf4833 -r 42f6c206c951 stubdom/Makefile
--- a/stubdom/Makefile Fri Mar 14 15:07:45 2008 -0600
+++ b/stubdom/Makefile Thu Mar 20 12:35:40 2008 -0600
@@ -201,6 +201,14 @@ caml:
caml:
$(MAKE) -C $@
+###
+# C
+###
+
+.PHONY: c
+c:
+ $(MAKE) -C $@
+
########
# minios
########
@@ -211,7 +219,11 @@ qemu-stubdom: mk-symlinks lwip-cvs libxc
.PHONY: caml-stubdom
caml-stubdom: mk-symlinks lwip-cvs libxc cross-libpci caml
- $(MAKE) -C mini-os LWIPDIR=$(CURDIR)/lwia-cvs CAMLDIR=$(CURDIR)/caml
+ $(MAKE) -C mini-os LWIPDIR=$(CURDIR)/lwip-cvs CAMLDIR=$(CURDIR)/caml
+
+.PHONY: c-stubdom
+c-stubdom: mk-symlinks lwip-cvs libxc cross-libpci c
+ $(MAKE) -C mini-os LWIPDIR=$(CURDIR)/lwip-cvs CDIR=$(CURDIR)/c
#########
# install
diff -r 8c921adf4833 -r 42f6c206c951 stubdom/README
--- a/stubdom/README Fri Mar 14 15:07:45 2008 -0600
+++ b/stubdom/README Thu Mar 20 12:35:40 2008 -0600
@@ -78,6 +78,8 @@ To run
mkdir -p /exports/usr/share/qemu
ln -s /usr/share/qemu/keymaps /exports/usr/share/qemu
+mkdir -p /exports/var/lib
+ln -s /var/lib/xen /exports/var/lib
/usr/sbin/fs-backend &
xm create hvmconfig
diff -r 8c921adf4833 -r 42f6c206c951 stubdom/c/Makefile
--- /dev/null Thu Jan 01 00:00:00 1970 +0000
+++ b/stubdom/c/Makefile Thu Mar 20 12:35:40 2008 -0600
@@ -0,0 +1,9 @@
+XEN_ROOT = ../..
+
+include $(XEN_ROOT)/Config.mk
+
+main.a: main.o
+ $(AR) cr $@ $^
+
+clean:
+ rm -f *.a *.o
diff -r 8c921adf4833 -r 42f6c206c951 stubdom/c/main.c
--- /dev/null Thu Jan 01 00:00:00 1970 +0000
+++ b/stubdom/c/main.c Thu Mar 20 12:35:40 2008 -0600
@@ -0,0 +1,6 @@
+#include <stdio.h>
+int main(void) {
+ sleep(2);
+ printf("Hello, world!\n");
+ return 0;
+}
diff -r 8c921adf4833 -r 42f6c206c951 stubdom/stubdom-dm
--- a/stubdom/stubdom-dm Fri Mar 14 15:07:45 2008 -0600
+++ b/stubdom/stubdom-dm Thu Mar 20 12:35:40 2008 -0600
@@ -14,18 +14,30 @@ domname=
domname=
vncviewer=0
vncpid=
+extra=
while [ "$#" -gt 0 ];
do
if [ "$#" -ge 2 ];
then
case "$1" in
- -d) domid=$2; shift ;;
- -domain-name) domname=$2; shift ;;
+ -d)
+ domid=$2;
+ extra="$extra -d $domid";
+ shift
+ ;;
+ -domain-name)
+ domname=$2;
+ shift
+ ;;
-vnc)
ip=${2%:*};
vnc_port=${2#*:};
shift
;;
+ -loadvm)
+ extra="$extra -loadvm $2";
+ shift
+ ;;
esac
fi
case "$1" in
@@ -60,7 +72,7 @@ do
sleep 1
done
-creation="xm create -c stubdom-$domname target=$domid memory=32"
+creation="xm create -c stubdom-$domname target=$domid memory=32
extra=\"$extra\""
(while true ; do sleep 60 ; done) | $creation &
#xterm -geometry +0+0 -e /bin/sh -c "$creation ; echo ; echo press ENTER to
shut down ; read" &
diff -r 8c921adf4833 -r 42f6c206c951 tools/blktap/drivers/Makefile
--- a/tools/blktap/drivers/Makefile Fri Mar 14 15:07:45 2008 -0600
+++ b/tools/blktap/drivers/Makefile Thu Mar 20 12:35:40 2008 -0600
@@ -1,7 +1,5 @@ XEN_ROOT = ../../..
XEN_ROOT = ../../..
include $(XEN_ROOT)/tools/Rules.mk
-
-INCLUDES += -I.. -I../lib -I$(XEN_INCLUDE)
IBIN = blktapctrl tapdisk
QCOW_UTIL = img2qcow qcow2raw qcow-create
@@ -9,18 +7,18 @@ LIBAIO_DIR = ../../libaio/src
CFLAGS += -Werror
CFLAGS += -Wno-unused
-CFLAGS += $(CFLAGS_libxenctrl) -I $(LIBAIO_DIR)
-CFLAGS += $(INCLUDES) -I. -I../../xenstore
+CFLAGS += -I../lib
+CFLAGS += $(CFLAGS_libxenctrl)
+CFLAGS += $(CFLAGS_libxenstore)
+CFLAGS += -I $(LIBAIO_DIR)
CFLAGS += -D_GNU_SOURCE
# Get gcc to generate the dependencies for us.
CFLAGS += -Wp,-MD,.$(@F).d
DEPS = .*.d
-LIBS := -L. -L.. -L../lib
-LIBS += -Wl,-rpath-link,$(XEN_XENSTORE)
-
-LIBS_IMG := $(LIBAIO_DIR)/libaio.a -lcrypto -lpthread -lz
+LDFLAGS_blktapctrl := $(LDFLAGS_libxenctrl) $(LDFLAGS_libxenstore) -L../lib
-lblktap
+LDFLAGS_img := $(LIBAIO_DIR)/libaio.a -lcrypto -lpthread -lz
BLK-OBJS-y := block-aio.o
BLK-OBJS-y += block-sync.o
@@ -34,17 +32,17 @@ BLK-OBJS-$(CONFIG_Linux) += blk_linux.c
all: $(IBIN) qcow-util
-blktapctrl: blktapctrl.c
- $(CC) $(CFLAGS) -o $@ $^ $(LDFLAGS) $(LIBS) -lblktap
$(LDFLAGS_libxenctrl)
+blktapctrl: blktapctrl.o
+ $(CC) $(CFLAGS) -o $@ $^ $(LDFLAGS) $(LDFLAGS_blktapctrl)
-tapdisk: tapdisk.c $(BLK-OBJS-y) tapdisk.c
- $(CC) $(CFLAGS) -o $@ $^ $(LDFLAGS) $(LIBS) $(LIBS_IMG)
+tapdisk: tapdisk.o $(BLK-OBJS-y)
+ $(CC) $(CFLAGS) -o $@ $^ $(LDFLAGS) $(LDFLAGS_img)
.PHONY: qcow-util
qcow-util: img2qcow qcow2raw qcow-create
-img2qcow qcow2raw qcow-create: %: $(BLK-OBJS-y)
- $(CC) $(CFLAGS) -o $* $(BLK-OBJS-y) $*.c $(LDFLAGS) $(LIBS_IMG)
+img2qcow qcow2raw qcow-create: %: %.o $(BLK-OBJS-y)
+ $(CC) $(CFLAGS) -o $* $^ $(LDFLAGS) $(LDFLAGS_img)
install: all
$(INSTALL_PROG) $(IBIN) $(QCOW_UTIL) $(VHD_UTIL) $(DESTDIR)$(SBINDIR)
diff -r 8c921adf4833 -r 42f6c206c951 tools/blktap/drivers/blktapctrl.c
--- a/tools/blktap/drivers/blktapctrl.c Fri Mar 14 15:07:45 2008 -0600
+++ b/tools/blktap/drivers/blktapctrl.c Thu Mar 20 12:35:40 2008 -0600
@@ -501,6 +501,80 @@ int launch_tapdisk(char *wrctldev, char
return 0;
}
+/* Connect to qemu-dm */
+static int connect_qemu(blkif_t *blkif)
+{
+ char *rdctldev, *wrctldev;
+
+ if (asprintf(&rdctldev, BLKTAP_CTRL_DIR "/qemu-read-%d",
+ blkif->domid) < 0)
+ return -1;
+
+ if (asprintf(&wrctldev, BLKTAP_CTRL_DIR "/qemu-write-%d",
+ blkif->domid) < 0) {
+ free(rdctldev);
+ return -1;
+ }
+
+ DPRINTF("Using qemu blktap pipe: %s\n", rdctldev);
+
+ blkif->fds[READ] = open_ctrl_socket(wrctldev);
+ blkif->fds[WRITE] = open_ctrl_socket(rdctldev);
+
+ free(rdctldev);
+ free(wrctldev);
+
+ if (blkif->fds[READ] == -1 || blkif->fds[WRITE] == -1)
+ return -1;
+
+ DPRINTF("Attached to qemu blktap pipes\n");
+ return 0;
+}
+
+/* Launch tapdisk instance */
+static int connect_tapdisk(blkif_t *blkif, int minor)
+{
+ char *rdctldev = NULL, *wrctldev = NULL;
+ int ret = -1;
+
+ DPRINTF("tapdisk process does not exist:\n");
+
+ if (asprintf(&rdctldev,
+ "%s/tapctrlread%d", BLKTAP_CTRL_DIR, minor) == -1)
+ goto fail;
+
+ if (asprintf(&wrctldev,
+ "%s/tapctrlwrite%d", BLKTAP_CTRL_DIR, minor) == -1)
+ goto fail;
+
+ blkif->fds[READ] = open_ctrl_socket(rdctldev);
+ blkif->fds[WRITE] = open_ctrl_socket(wrctldev);
+
+ if (blkif->fds[READ] == -1 || blkif->fds[WRITE] == -1)
+ goto fail;
+
+ /*launch the new process*/
+ DPRINTF("Launching process, CMDLINE [tapdisk %s %s]\n",
+ wrctldev, rdctldev);
+
+ if (launch_tapdisk(wrctldev, rdctldev) == -1) {
+ DPRINTF("Unable to fork, cmdline: [tapdisk %s %s]\n",
+ wrctldev, rdctldev);
+ goto fail;
+ }
+
+ ret = 0;
+
+fail:
+ if (rdctldev)
+ free(rdctldev);
+
+ if (wrctldev)
+ free(wrctldev);
+
+ return ret;
+}
+
int blktapctrl_new_blkif(blkif_t *blkif)
{
blkif_info_t *blk;
@@ -524,30 +598,14 @@ int blktapctrl_new_blkif(blkif_t *blkif)
blkif->cookie = next_cookie++;
if (!exist) {
- DPRINTF("Process does not exist:\n");
- if (asprintf(&rdctldev,
- "%s/tapctrlread%d", BLKTAP_CTRL_DIR,
minor) == -1)
- goto fail;
- if (asprintf(&wrctldev,
- "%s/tapctrlwrite%d", BLKTAP_CTRL_DIR,
minor) == -1) {
- free(rdctldev);
- goto fail;
+ if (type == DISK_TYPE_IOEMU) {
+ if (connect_qemu(blkif))
+ goto fail;
+ } else {
+ if (connect_tapdisk(blkif, minor))
+ goto fail;
}
- blkif->fds[READ] = open_ctrl_socket(rdctldev);
- blkif->fds[WRITE] = open_ctrl_socket(wrctldev);
-
- if (blkif->fds[READ] == -1 || blkif->fds[WRITE] == -1)
- goto fail;
-
- /*launch the new process*/
- DPRINTF("Launching process, CMDLINE [tapdisk %s
%s]\n",wrctldev, rdctldev);
- if (launch_tapdisk(wrctldev, rdctldev) == -1) {
- DPRINTF("Unable to fork, cmdline: [tapdisk %s
%s]\n",wrctldev, rdctldev);
- goto fail;
- }
-
- free(rdctldev);
- free(wrctldev);
+
} else {
DPRINTF("Process exists!\n");
blkif->fds[READ] = exist->fds[READ];
diff -r 8c921adf4833 -r 42f6c206c951 tools/blktap/drivers/block-qcow2.c
--- a/tools/blktap/drivers/block-qcow2.c Fri Mar 14 15:07:45 2008 -0600
+++ b/tools/blktap/drivers/block-qcow2.c Thu Mar 20 12:35:40 2008 -0600
@@ -43,6 +43,11 @@
#ifndef O_BINARY
#define O_BINARY 0
+#endif
+
+/* *BSD has no O_LARGEFILE */
+#ifndef O_LARGEFILE
+#define O_LARGEFILE 0
#endif
#define BLOCK_FLAG_ENCRYPT 1
diff -r 8c921adf4833 -r 42f6c206c951 tools/blktap/drivers/tapaio.c
--- a/tools/blktap/drivers/tapaio.c Fri Mar 14 15:07:45 2008 -0600
+++ b/tools/blktap/drivers/tapaio.c Thu Mar 20 12:35:40 2008 -0600
@@ -82,6 +82,7 @@ tap_aio_completion_thread(void *arg)
}
} while (!rc);
}
+ return NULL;
}
void
diff -r 8c921adf4833 -r 42f6c206c951 tools/blktap/drivers/tapdisk.h
--- a/tools/blktap/drivers/tapdisk.h Fri Mar 14 15:07:45 2008 -0600
+++ b/tools/blktap/drivers/tapdisk.h Thu Mar 20 12:35:40 2008 -0600
@@ -167,6 +167,7 @@ extern struct tap_disk tapdisk_qcow2;
#define DISK_TYPE_RAM 3
#define DISK_TYPE_QCOW 4
#define DISK_TYPE_QCOW2 5
+#define DISK_TYPE_IOEMU 6
/*Define Individual Disk Parameters here */
@@ -227,6 +228,16 @@ static disk_info_t qcow2_disk = {
0,
#ifdef TAPDISK
&tapdisk_qcow2,
+#endif
+};
+
+static disk_info_t ioemu_disk = {
+ DISK_TYPE_IOEMU,
+ "ioemu disk",
+ "ioemu",
+ 0,
+#ifdef TAPDISK
+ NULL
#endif
};
@@ -238,6 +249,7 @@ static disk_info_t *dtypes[] = {
&ram_disk,
&qcow_disk,
&qcow2_disk,
+ &ioemu_disk,
};
typedef struct driver_list_entry {
diff -r 8c921adf4833 -r 42f6c206c951 tools/blktap/lib/Makefile
--- a/tools/blktap/lib/Makefile Fri Mar 14 15:07:45 2008 -0600
+++ b/tools/blktap/lib/Makefile Thu Mar 20 12:35:40 2008 -0600
@@ -5,8 +5,9 @@ MINOR = 0
MINOR = 0
SONAME = libblktap.so.$(MAJOR)
-INCLUDES += -I. -I.. -I $(XEN_LIBXC) -I$(XEN_INCLUDE) $(CFLAGS_libxenstore)
-
+CFLAGS += -I.
+CFLAGS += $(CFLAGS_libxenctrl)
+CFLAGS += $(CFLAGS_libxenstore)
LDFLAGS += $(LDFLAGS_libxenstore)
SRCS :=
@@ -20,23 +21,18 @@ CFLAGS += -D _GNU_SOURCE
# Get gcc to generate the dependencies for us.
CFLAGS += -Wp,-MD,.$(@F).d
-CFLAGS += $(INCLUDES)
DEPS = .*.d
-OBJS = $(patsubst %.c,%.o,$(SRCS))
+OBJS = $(SRCS:.c=.o)
+OBJS_PIC = $(SRCS:.c=.opic)
IBINS :=
LIB = libblktap.a libblktap.so.$(MAJOR).$(MINOR)
.PHONY: all
-all: build
+all: $(LIB)
-.PHONY: build
-build: libblktap.a
-
-.PHONY: libblktap
-libblktap: libblktap.a
-
+.PHONY: install
install: all
$(INSTALL_DIR) $(DESTDIR)$(LIBDIR)
$(INSTALL_DIR) $(DESTDIR)$(INCLUDEDIR)
@@ -45,19 +41,20 @@ install: all
ln -sf libblktap.so.$(MAJOR) $(DESTDIR)$(LIBDIR)/libblktap.so
$(INSTALL_DATA) blktaplib.h $(DESTDIR)$(INCLUDEDIR)
+.PHONY: clean
clean:
rm -rf *.a *.so* *.o *.rpm $(LIB) *~ $(DEPS) xen TAGS
-libblktap.a: $(OBJS)
+libblktap.so.$(MAJOR).$(MINOR): $(OBJS_PIC)
$(CC) $(CFLAGS) -Wl,$(SONAME_LDFLAG) -Wl,$(SONAME) $(SHLIB_CFLAGS) \
- -L$(XEN_XENSTORE) $(LDFLAGS) \
- -o libblktap.so.$(MAJOR).$(MINOR) $^ $(LIBS)
+ $(LDFLAGS) -o $@ $^
ln -sf libblktap.so.$(MAJOR).$(MINOR) libblktap.so.$(MAJOR)
ln -sf libblktap.so.$(MAJOR) libblktap.so
- $(AR) rc $@ libblktap.so
-.PHONY: TAGS all build clean install libblktap
+libblktap.a: $(OBJS)
+ $(AR) rc $@ $^
+.PHONY: TAGS
TAGS:
etags -t $(SRCS) *.h
diff -r 8c921adf4833 -r 42f6c206c951 tools/blktap/lib/blktaplib.h
--- a/tools/blktap/lib/blktaplib.h Fri Mar 14 15:07:45 2008 -0600
+++ b/tools/blktap/lib/blktaplib.h Thu Mar 20 12:35:40 2008 -0600
@@ -221,15 +221,5 @@ int xs_fire_next_watch(struct xs_handle
((_req) * BLKIF_MAX_SEGMENTS_PER_REQUEST * getpagesize()) + \
((_seg) * getpagesize()))
-/* Defines that are only used by library clients */
-
-#ifndef __COMPILING_BLKTAP_LIB
-
-static char *blkif_op_name[] = {
- [BLKIF_OP_READ] = "READ",
- [BLKIF_OP_WRITE] = "WRITE",
-};
-
-#endif /* __COMPILING_BLKTAP_LIB */
#endif /* __BLKTAPLIB_H__ */
diff -r 8c921adf4833 -r 42f6c206c951 tools/examples/xmexample.hvm
--- a/tools/examples/xmexample.hvm Fri Mar 14 15:07:45 2008 -0600
+++ b/tools/examples/xmexample.hvm Thu Mar 20 12:35:40 2008 -0600
@@ -135,6 +135,11 @@ sdl=0
sdl=0
#----------------------------------------------------------------------------
+# enable OpenGL for texture rendering inside the SDL window, default = 1
+# valid only if sdl is enabled.
+opengl=1
+
+#----------------------------------------------------------------------------
# enable VNC library for graphics, default = 1
vnc=1
diff -r 8c921adf4833 -r 42f6c206c951 tools/firmware/rombios/rombios.c
--- a/tools/firmware/rombios/rombios.c Fri Mar 14 15:07:45 2008 -0600
+++ b/tools/firmware/rombios/rombios.c Thu Mar 20 12:35:40 2008 -0600
@@ -779,8 +779,8 @@ typedef struct {
Bit32u sector_count1;
Bit32u sector_count2;
Bit16u blksize;
+ Bit16u dpte_offset;
Bit16u dpte_segment;
- Bit16u dpte_offset;
Bit16u key;
Bit8u dpi_length;
Bit8u reserved1;
diff -r 8c921adf4833 -r 42f6c206c951 tools/include/Makefile
--- a/tools/include/Makefile Fri Mar 14 15:07:45 2008 -0600
+++ b/tools/include/Makefile Thu Mar 20 12:35:40 2008 -0600
@@ -22,6 +22,7 @@ install: all
install: all
$(INSTALL_DIR) $(DESTDIR)$(INCLUDEDIR)/xen/arch-ia64
$(INSTALL_DIR) $(DESTDIR)$(INCLUDEDIR)/xen/arch-x86
+ $(INSTALL_DIR) $(DESTDIR)$(INCLUDEDIR)/xen/arch-x86/hvm
$(INSTALL_DIR) $(DESTDIR)$(INCLUDEDIR)/xen/foreign
$(INSTALL_DIR) $(DESTDIR)$(INCLUDEDIR)/xen/hvm
$(INSTALL_DIR) $(DESTDIR)$(INCLUDEDIR)/xen/io
@@ -31,6 +32,7 @@ install: all
$(INSTALL_DATA) xen/*.h $(DESTDIR)$(INCLUDEDIR)/xen
$(INSTALL_DATA) xen/arch-ia64/*.h $(DESTDIR)$(INCLUDEDIR)/xen/arch-ia64
$(INSTALL_DATA) xen/arch-x86/*.h $(DESTDIR)$(INCLUDEDIR)/xen/arch-x86
+ $(INSTALL_DATA) xen/arch-x86/hvm/*.h
$(DESTDIR)$(INCLUDEDIR)/xen/arch-x86/hvm
$(INSTALL_DATA) xen/foreign/*.h $(DESTDIR)$(INCLUDEDIR)/xen/foreign
$(INSTALL_DATA) xen/hvm/*.h $(DESTDIR)$(INCLUDEDIR)/xen/hvm
$(INSTALL_DATA) xen/io/*.h $(DESTDIR)$(INCLUDEDIR)/xen/io
diff -r 8c921adf4833 -r 42f6c206c951 tools/include/xen-sys/Linux/gntdev.h
--- a/tools/include/xen-sys/Linux/gntdev.h Fri Mar 14 15:07:45 2008 -0600
+++ b/tools/include/xen-sys/Linux/gntdev.h Thu Mar 20 12:35:40 2008 -0600
@@ -102,4 +102,18 @@ struct ioctl_gntdev_get_offset_for_vaddr
uint32_t pad;
};
+/*
+ * Sets the maximum number of grants that may mapped at once by this gntdev
+ * instance.
+ *
+ * N.B. This must be called before any other ioctl is performed on the device.
+ */
+#define IOCTL_GNTDEV_SET_MAX_GRANTS \
+_IOC(_IOC_NONE, 'G', 3, sizeof(struct ioctl_gntdev_set_max_grants))
+struct ioctl_gntdev_set_max_grants {
+ /* IN parameter */
+ /* The maximum number of grants that may be mapped at once. */
+ uint32_t count;
+};
+
#endif /* __LINUX_PUBLIC_GNTDEV_H__ */
diff -r 8c921adf4833 -r 42f6c206c951 tools/ioemu/Makefile.target
--- a/tools/ioemu/Makefile.target Fri Mar 14 15:07:45 2008 -0600
+++ b/tools/ioemu/Makefile.target Thu Mar 20 12:35:40 2008 -0600
@@ -17,6 +17,7 @@ VPATH=$(SRC_PATH):$(TARGET_PATH):$(SRC_P
VPATH=$(SRC_PATH):$(TARGET_PATH):$(SRC_PATH)/hw:$(SRC_PATH)/audio
CPPFLAGS+=-I. -I.. -I$(TARGET_PATH) -I$(SRC_PATH)
CPPFLAGS+= -I$(XEN_ROOT)/tools/libxc
+CPPFLAGS+= -I$(XEN_ROOT)/tools/blktap/lib
CPPFLAGS+= -I$(XEN_ROOT)/tools/xenstore
CPPFLAGS+= -I$(XEN_ROOT)/tools/include
ifdef CONFIG_DARWIN_USER
@@ -429,6 +430,9 @@ VL_OBJS+= usb-uhci.o smbus_eeprom.o
VL_OBJS+= usb-uhci.o smbus_eeprom.o
VL_OBJS+= piix4acpi.o
VL_OBJS+= xenstore.o
+ifndef CONFIG_STUBDOM
+VL_OBJS+= xen_blktap.o
+endif
VL_OBJS+= xen_platform.o
VL_OBJS+= xen_machine_fv.o
VL_OBJS+= xen_machine_pv.o
diff -r 8c921adf4833 -r 42f6c206c951 tools/ioemu/block-vbd.c
--- a/tools/ioemu/block-vbd.c Fri Mar 14 15:07:45 2008 -0600
+++ b/tools/ioemu/block-vbd.c Thu Mar 20 12:35:40 2008 -0600
@@ -51,6 +51,7 @@ typedef struct BDRVVbdState {
int fd;
int type;
int mode;
+ int info;
uint64_t sectors;
unsigned sector_size;
QEMU_LIST_ENTRY(BDRVVbdState) list;
@@ -80,7 +81,7 @@ static int vbd_open(BlockDriverState *bs
//handy to test posix access
//return -EIO;
- s->dev = init_blkfront((char *) filename, &s->sectors, &s->sector_size,
&s->mode);
+ s->dev = init_blkfront((char *) filename, &s->sectors, &s->sector_size,
&s->mode, &s->info);
if (!s->dev)
return -EIO;
@@ -271,6 +272,7 @@ static void vbd_close(BlockDriverState *
BDRVVbdState *s = bs->opaque;
bs->total_sectors = 0;
if (s->fd >= 0) {
+ qemu_set_fd_handler(s->fd, NULL, NULL, NULL);
close(s->fd);
s->fd = -1;
}
diff -r 8c921adf4833 -r 42f6c206c951 tools/ioemu/configure
--- a/tools/ioemu/configure Fri Mar 14 15:07:45 2008 -0600
+++ b/tools/ioemu/configure Thu Mar 20 12:35:40 2008 -0600
@@ -189,6 +189,8 @@ for opt do
--static) static="yes"
;;
--disable-sdl) sdl="no"
+ ;;
+ --disable-opengl) opengl="no"
;;
--enable-coreaudio) coreaudio="yes"
;;
@@ -539,6 +541,26 @@ fi # -z $sdl
fi # -z $sdl
##########################################
+# OpenGL test
+
+if test -z "$opengl" && test "$sdl" = "yes"
+then
+cat > $TMPC << EOF
+#include <SDL_opengl.h>
+#ifndef GL_TEXTURE_RECTANGLE_ARB
+#error "Opengl doesn't support GL_TEXTURE_RECTANGLE_ARB"
+#endif
+int main( void ) { return (int) glGetString(GL_EXTENSIONS); }
+EOF
+if $cc -o $TMPE `$sdl_config --cflags --libs 2> /dev/null` -I/usr/include/GL
$TMPC -lXext -lGL 2> /dev/null
+then
+opengl="yes"
+else
+opengl="no"
+fi
+fi
+
+##########################################
# alsa sound support libraries
if test "$alsa" = "yes" ; then
@@ -612,6 +634,7 @@ if test "$sdl" != "no" ; then
if test "$sdl" != "no" ; then
echo "SDL static link $sdl_static"
fi
+echo "OpenGL support $opengl"
echo "mingw32 support $mingw32"
echo "Adlib support $adlib"
echo "CoreAudio support $coreaudio"
@@ -995,6 +1018,15 @@ if test "$target_user_only" = "no"; then
fi
fi
+if test $opengl = "yes"
+then
+ echo "#define CONFIG_OPENGL 1" >> $config_h
+ echo "CONFIG_OPENGL=yes" >> $config_mak
+ echo "SDL_CFLAGS+=-I/usr/include/GL" >> $config_mak
+ echo "SDL_LIBS+=-lXext" >> $config_mak
+ echo "SDL_LIBS+=-lGL" >> $config_mak
+fi
+
if test "$cocoa" = "yes" ; then
echo "#define CONFIG_COCOA 1" >> $config_h
echo "CONFIG_COCOA=yes" >> $config_mak
diff -r 8c921adf4833 -r 42f6c206c951 tools/ioemu/hw/ide.c
--- a/tools/ioemu/hw/ide.c Fri Mar 14 15:07:45 2008 -0600
+++ b/tools/ioemu/hw/ide.c Thu Mar 20 12:35:40 2008 -0600
@@ -189,6 +189,15 @@
/* set to 1 set disable mult support */
#define MAX_MULT_SECTORS 16
+#ifdef CONFIG_STUBDOM
+#include <xen/io/blkif.h>
+#define IDE_DMA_BUF_SIZE (BLKIF_MAX_SEGMENTS_PER_REQUEST * TARGET_PAGE_SIZE)
+#else
+#define IDE_DMA_BUF_SIZE 131072
+#endif
+#if (IDE_DMA_BUF_SIZE < MAX_MULT_SECTORS * 512)
+#error "IDE_DMA_BUF_SIZE must be bigger or equal to MAX_MULT_SECTORS * 512"
+#endif
/* ATAPI defines */
@@ -932,8 +941,8 @@ static void ide_read_dma_cb(void *opaque
/* launch next transfer */
n = s->nsector;
- if (n > MAX_MULT_SECTORS)
- n = MAX_MULT_SECTORS;
+ if (n > IDE_DMA_BUF_SIZE / 512)
+ n = IDE_DMA_BUF_SIZE / 512;
s->io_buffer_index = 0;
s->io_buffer_size = n * 512;
#ifdef DEBUG_AIO
@@ -1041,8 +1050,8 @@ static void ide_write_dma_cb(void *opaqu
/* launch next transfer */
n = s->nsector;
- if (n > MAX_MULT_SECTORS)
- n = MAX_MULT_SECTORS;
+ if (n > IDE_DMA_BUF_SIZE / 512)
+ n = IDE_DMA_BUF_SIZE / 512;
s->io_buffer_index = 0;
s->io_buffer_size = n * 512;
@@ -1336,8 +1345,8 @@ static void ide_atapi_cmd_read_dma_cb(vo
data_offset = 16;
} else {
n = s->packet_transfer_size >> 11;
- if (n > (MAX_MULT_SECTORS / 4))
- n = (MAX_MULT_SECTORS / 4);
+ if (n > (IDE_DMA_BUF_SIZE / 2048))
+ n = (IDE_DMA_BUF_SIZE / 2048);
s->io_buffer_size = n * 2048;
data_offset = 0;
}
@@ -1823,7 +1832,6 @@ static void ide_ioport_write(void *opaqu
break;
case WIN_SETMULT:
if (s->nsector > MAX_MULT_SECTORS ||
- s->nsector == 0 ||
(s->nsector & (s->nsector - 1)) != 0) {
ide_abort_command(s);
} else {
@@ -2306,7 +2314,7 @@ static void ide_init2(IDEState *ide_stat
for(i = 0; i < 2; i++) {
s = ide_state + i;
- s->io_buffer = qemu_memalign(getpagesize(), MAX_MULT_SECTORS*512 + 4);
+ s->io_buffer = qemu_memalign(getpagesize(), IDE_DMA_BUF_SIZE + 4);
if (i == 0)
s->bs = hd0;
else
diff -r 8c921adf4833 -r 42f6c206c951 tools/ioemu/hw/ne2000.c
--- a/tools/ioemu/hw/ne2000.c Fri Mar 14 15:07:45 2008 -0600
+++ b/tools/ioemu/hw/ne2000.c Thu Mar 20 12:35:40 2008 -0600
@@ -834,6 +834,10 @@ void pci_ne2000_init(PCIBus *bus, NICInf
pci_conf[0x0a] = 0x00; // ethernet network controller
pci_conf[0x0b] = 0x02;
pci_conf[0x0e] = 0x00; // header_type
+ pci_conf[0x2c] = 0x53; /* subsystem vendor: XenSource */
+ pci_conf[0x2d] = 0x58;
+ pci_conf[0x2e] = 0x01; /* subsystem device */
+ pci_conf[0x2f] = 0x00;
pci_conf[0x3d] = 1; // interrupt pin 0
pci_register_io_region(&d->dev, 0, 0x100,
diff -r 8c921adf4833 -r 42f6c206c951 tools/ioemu/hw/scsi-disk.c
--- a/tools/ioemu/hw/scsi-disk.c Fri Mar 14 15:07:45 2008 -0600
+++ b/tools/ioemu/hw/scsi-disk.c Thu Mar 20 12:35:40 2008 -0600
@@ -34,9 +34,10 @@ do { fprintf(stderr, "scsi-disk: " fmt ,
#define SENSE_ILLEGAL_REQUEST 5
#ifdef CONFIG_STUBDOM
-#define SCSI_DMA_BUF_SIZE 32768
+#include <xen/io/blkif.h>
+#define SCSI_DMA_BUF_SIZE (BLKIF_MAX_SEGMENTS_PER_REQUEST *
TARGET_PAGE_SIZE)
#else
-#define SCSI_DMA_BUF_SIZE 65536
+#define SCSI_DMA_BUF_SIZE 131072
#endif
typedef struct SCSIRequest {
diff -r 8c921adf4833 -r 42f6c206c951 tools/ioemu/hw/vga.c
--- a/tools/ioemu/hw/vga.c Fri Mar 14 15:07:45 2008 -0600
+++ b/tools/ioemu/hw/vga.c Thu Mar 20 12:35:40 2008 -0600
@@ -1148,7 +1148,7 @@ static void vga_draw_text(VGAState *s, i
cw != s->last_cw || cheight != s->last_ch) {
s->last_scr_width = width * cw;
s->last_scr_height = height * cheight;
- dpy_resize(s->ds, s->last_scr_width, s->last_scr_height,
s->last_scr_width * (depth / 8));
+ dpy_resize(s->ds, s->last_scr_width, s->last_scr_height,
s->last_scr_width * (s->ds->depth / 8));
s->last_width = width;
s->last_height = height;
s->last_ch = cheight;
@@ -1571,7 +1571,7 @@ static void vga_draw_graphic(VGAState *s
vga_draw_line = vga_draw_line_table[v * NB_DEPTHS +
get_depth_index(s->ds)];
if (disp_width != s->last_width ||
height != s->last_height) {
- dpy_resize(s->ds, disp_width, height, disp_width * (depth / 8));
+ dpy_resize(s->ds, disp_width, height, s->line_offset);
s->last_scr_width = disp_width;
s->last_scr_height = height;
s->last_width = disp_width;
diff -r 8c921adf4833 -r 42f6c206c951 tools/ioemu/hw/xen_blktap.c
--- /dev/null Thu Jan 01 00:00:00 1970 +0000
+++ b/tools/ioemu/hw/xen_blktap.c Thu Mar 20 12:35:40 2008 -0600
@@ -0,0 +1,686 @@
+/* xen_blktap.c
+ *
+ * Interface to blktapctrl to allow use of qemu block drivers with blktap.
+ * This file is based on tools/blktap/drivers/tapdisk.c
+ *
+ * Copyright (c) 2005 Julian Chesterfield and Andrew Warfield.
+ * Copyright (c) 2008 Kevin Wolf
+ */
+
+/*
+ * There are several communication channels which are used by this interface:
+ *
+ * - A pair of pipes for receiving and sending general control messages
+ * (qemu-read-N and qemu-writeN in /var/run/tap, where N is the domain ID).
+ * These control messages are handled by handle_blktap_ctrlmsg().
+ *
+ * - One file descriptor per attached disk (/dev/xen/blktapN) for disk
+ * specific control messages. A callback is triggered on this fd if there
+ * is a new IO request. The callback function is handle_blktap_iomsg().
+ *
+ * - A shared ring for each attached disk containing the actual IO requests
+ * and responses. Whenever handle_blktap_iomsg() is triggered it processes
+ * the requests on this ring.
+ */
+
+#include <sys/stat.h>
+#include <sys/types.h>
+#include <sys/mman.h>
+#include <sys/ioctl.h>
+#include <fcntl.h>
+#include <stdio.h>
+#include <errno.h>
+#include <stdlib.h>
+
+#include "vl.h"
+#include "blktaplib.h"
+#include "xen_blktap.h"
+#include "block_int.h"
+
+#define MSG_SIZE 4096
+
+#define BLKTAP_CTRL_DIR "/var/run/tap"
+
+/* If enabled, print debug messages to stderr */
+#if 1
+#define DPRINTF(_f, _a...) fprintf(stderr, __FILE__ ":%d: " _f, __LINE__, ##_a)
+#else
+#define DPRINTF(_f, _a...) ((void)0)
+#endif
+
+#if 1
+#define ASSERT(_p) \
+ if ( !(_p) ) { DPRINTF("Assertion '%s' failed, line %d, file %s\n", #_p , \
+ __LINE__, __FILE__); *(int*)0=0; }
+#else
+#define ASSERT(_p) ((void)0)
+#endif
+
+
+extern int domid;
+
+int read_fd;
+int write_fd;
+
+static pid_t process;
+fd_list_entry_t *fd_start = NULL;
+
+static void handle_blktap_iomsg(void* private);
+
+struct aiocb_info {
+ struct td_state *s;
+ uint64_t sector;
+ int nr_secs;
+ int idx;
+ long i;
+};
+
+static void unmap_disk(struct td_state *s)
+{
+ tapdev_info_t *info = s->ring_info;
+ fd_list_entry_t *entry;
+
+ bdrv_close(s->bs);
+
+ if (info != NULL && info->mem > 0)
+ munmap(info->mem, getpagesize() * BLKTAP_MMAP_REGION_SIZE);
+
+ entry = s->fd_entry;
+ *entry->pprev = entry->next;
+ if (entry->next)
+ entry->next->pprev = entry->pprev;
+
+ qemu_set_fd_handler2(info->fd, NULL, NULL, NULL, NULL);
+ close(info->fd);
+
+ free(s->fd_entry);
+ free(s->blkif);
+ free(s->ring_info);
+ free(s);
+
+ return;
+}
+
+static inline fd_list_entry_t *add_fd_entry(int tap_fd, struct td_state *s)
+{
+ fd_list_entry_t **pprev, *entry;
+
+ DPRINTF("Adding fd_list_entry\n");
+
+ /*Add to linked list*/
+ s->fd_entry = entry = malloc(sizeof(fd_list_entry_t));
+ entry->tap_fd = tap_fd;
+ entry->s = s;
+ entry->next = NULL;
+
+ pprev = &fd_start;
+ while (*pprev != NULL)
+ pprev = &(*pprev)->next;
+
+ *pprev = entry;
+ entry->pprev = pprev;
+
+ return entry;
+}
+
+static inline struct td_state *get_state(int cookie)
+{
+ fd_list_entry_t *ptr;
+
+ ptr = fd_start;
+ while (ptr != NULL) {
+ if (ptr->cookie == cookie) return ptr->s;
+ ptr = ptr->next;
+ }
+ return NULL;
+}
+
+static struct td_state *state_init(void)
+{
+ int i;
+ struct td_state *s;
+ blkif_t *blkif;
+
+ s = malloc(sizeof(struct td_state));
+ blkif = s->blkif = malloc(sizeof(blkif_t));
+ s->ring_info = calloc(1, sizeof(tapdev_info_t));
+
+ for (i = 0; i < MAX_REQUESTS; i++) {
+ blkif->pending_list[i].secs_pending = 0;
+ blkif->pending_list[i].submitting = 0;
+ }
+
+ return s;
+}
+
+static int map_new_dev(struct td_state *s, int minor)
+{
+ int tap_fd;
+ tapdev_info_t *info = s->ring_info;
+ char *devname;
+ fd_list_entry_t *ptr;
+ int page_size;
+
+ if (asprintf(&devname,"%s/%s%d", BLKTAP_DEV_DIR, BLKTAP_DEV_NAME,
minor) == -1)
+ return -1;
+ tap_fd = open(devname, O_RDWR);
+ if (tap_fd == -1)
+ {
+ DPRINTF("open failed on dev %s!\n",devname);
+ goto fail;
+ }
+ info->fd = tap_fd;
+
+ /*Map the shared memory*/
+ page_size = getpagesize();
+ info->mem = mmap(0, page_size * BLKTAP_MMAP_REGION_SIZE,
+ PROT_READ | PROT_WRITE, MAP_SHARED, info->fd, 0);
+ if ((long int)info->mem == -1)
+ {
+ DPRINTF("mmap failed on dev %s!\n",devname);
+ goto fail;
+ }
+
+ /* assign the rings to the mapped memory */
+ info->sring = (blkif_sring_t *)((unsigned long)info->mem);
+ BACK_RING_INIT(&info->fe_ring, info->sring, page_size);
+
+ info->vstart =
+ (unsigned long)info->mem + (BLKTAP_RING_PAGES * page_size);
+
+ ioctl(info->fd, BLKTAP_IOCTL_SENDPID, process );
+ ioctl(info->fd, BLKTAP_IOCTL_SETMODE, BLKTAP_MODE_INTERPOSE );
+ free(devname);
+
+ /*Update the fd entry*/
+ ptr = fd_start;
+ while (ptr != NULL) {
+ if (s == ptr->s) {
+ ptr->tap_fd = tap_fd;
+
+ /* Setup fd_handler for qemu main loop */
+ DPRINTF("set tap_fd = %d\n", tap_fd);
+ qemu_set_fd_handler2(tap_fd, NULL,
&handle_blktap_iomsg, NULL, s);
+
+ break;
+ }
+ ptr = ptr->next;
+ }
+
+
+ DPRINTF("map_new_dev = %d\n", minor);
+ return minor;
+
+ fail:
+ free(devname);
+ return -1;
+}
+
+static int open_disk(struct td_state *s, char *path, int readonly)
+{
+ struct disk_id id;
+ BlockDriverState* bs;
+
+ DPRINTF("Opening %s\n", path);
+ bs = calloc(1, sizeof(*bs));
+
+ memset(&id, 0, sizeof(struct disk_id));
+
+ if (bdrv_open(bs, path, 0) != 0) {
+ fprintf(stderr, "Could not open image file %s\n", path);
+ return -ENOMEM;
+ }
+
+ s->bs = bs;
+ s->flags = readonly ? TD_RDONLY : 0;
+ s->size = bs->total_sectors;
+ s->sector_size = 512;
+
+ s->info = ((s->flags & TD_RDONLY) ? VDISK_READONLY : 0);
+
+ return 0;
+}
+
+static inline void write_rsp_to_ring(struct td_state *s, blkif_response_t *rsp)
+{
+ tapdev_info_t *info = s->ring_info;
+ blkif_response_t *rsp_d;
+
+ rsp_d = RING_GET_RESPONSE(&info->fe_ring, info->fe_ring.rsp_prod_pvt);
+ memcpy(rsp_d, rsp, sizeof(blkif_response_t));
+ info->fe_ring.rsp_prod_pvt++;
+}
+
+static inline void kick_responses(struct td_state *s)
+{
+ tapdev_info_t *info = s->ring_info;
+
+ if (info->fe_ring.rsp_prod_pvt != info->fe_ring.sring->rsp_prod)
+ {
+ RING_PUSH_RESPONSES(&info->fe_ring);
+ ioctl(info->fd, BLKTAP_IOCTL_KICK_FE);
+ }
+}
+
+static int send_responses(struct td_state *s, int res,
+ uint64_t sector, int nr_secs, int idx, void *private)
+{
+ pending_req_t *preq;
+ blkif_request_t *req;
+ int responses_queued = 0;
+ blkif_t *blkif = s->blkif;
+ int secs_done = nr_secs;
+
+ if ( (idx > MAX_REQUESTS-1) )
+ {
+ DPRINTF("invalid index returned(%u)!\n", idx);
+ return 0;
+ }
+ preq = &blkif->pending_list[idx];
+ req = &preq->req;
+
+ preq->secs_pending -= secs_done;
+
+ if (res == -EBUSY && preq->submitting)
+ return -EBUSY; /* propagate -EBUSY back to higher layers */
+ if (res)
+ preq->status = BLKIF_RSP_ERROR;
+
+ if (!preq->submitting && preq->secs_pending == 0)
+ {
+ blkif_request_t tmp;
+ blkif_response_t *rsp;
+
+ tmp = preq->req;
+ rsp = (blkif_response_t *)req;
+
+ rsp->id = tmp.id;
+ rsp->operation = tmp.operation;
+ rsp->status = preq->status;
+
+ write_rsp_to_ring(s, rsp);
+ responses_queued++;
+
+ kick_responses(s);
+ }
+
+ return responses_queued;
+}
+
+static void qemu_send_responses(void* opaque, int ret)
+{
+ struct aiocb_info* info = opaque;
+
+ if (ret != 0) {
+ DPRINTF("ERROR: ret = %d (%s)\n", ret, strerror(-ret));
+ }
+
+ send_responses(info->s, ret, info->sector, info->nr_secs,
+ info->idx, (void*) info->i);
+ free(info);
+}
+
+/**
+ * Callback function for the IO message pipe. Reads requests from the ring
+ * and processes them (call qemu read/write functions).
+ *
+ * The private parameter points to the struct td_state representing the
+ * disk the request is targeted at.
+ */
+static void handle_blktap_iomsg(void* private)
+{
+ struct td_state* s = private;
+
+ RING_IDX rp, j, i;
+ blkif_request_t *req;
+ int idx, nsects, ret;
+ uint64_t sector_nr;
+ uint8_t *page;
+ blkif_t *blkif = s->blkif;
+ tapdev_info_t *info = s->ring_info;
+ int page_size = getpagesize();
+
+ struct aiocb_info *aiocb_info;
+
+ if (info->fe_ring.sring == NULL) {
+ DPRINTF(" sring == NULL, ignoring IO request\n");
+ return;
+ }
+
+ rp = info->fe_ring.sring->req_prod;
+ xen_rmb();
+
+ for (j = info->fe_ring.req_cons; j != rp; j++)
+ {
+ int start_seg = 0;
+
+ req = NULL;
+ req = RING_GET_REQUEST(&info->fe_ring, j);
+ ++info->fe_ring.req_cons;
+
+ if (req == NULL)
+ continue;
+
+ idx = req->id;
+
+ ASSERT(blkif->pending_list[idx].secs_pending == 0);
+ memcpy(&blkif->pending_list[idx].req, req, sizeof(*req));
+ blkif->pending_list[idx].status = BLKIF_RSP_OKAY;
+ blkif->pending_list[idx].submitting = 1;
+ sector_nr = req->sector_number;
+
+ /* Don't allow writes on readonly devices */
+ if ((s->flags & TD_RDONLY) &&
+ (req->operation == BLKIF_OP_WRITE)) {
+ blkif->pending_list[idx].status = BLKIF_RSP_ERROR;
+ goto send_response;
+ }
+
+ for (i = start_seg; i < req->nr_segments; i++) {
+ nsects = req->seg[i].last_sect -
+ req->seg[i].first_sect + 1;
+
+ if ((req->seg[i].last_sect >= page_size >> 9) ||
+ (nsects <= 0))
+ continue;
+
+ page = (uint8_t*) MMAP_VADDR(info->vstart,
+ (unsigned long)req->id, i);
+ page += (req->seg[i].first_sect << SECTOR_SHIFT);
+
+ if (sector_nr >= s->size) {
+ DPRINTF("Sector request failed:\n");
+ DPRINTF("%s request, idx [%d,%d] size [%llu], "
+ "sector [%llu,%llu]\n",
+ (req->operation == BLKIF_OP_WRITE ?
+ "WRITE" : "READ"),
+ idx,i,
+ (long long unsigned)
+ nsects<<SECTOR_SHIFT,
+ (long long unsigned)
+ sector_nr<<SECTOR_SHIFT,
+ (long long unsigned) sector_nr);
+ continue;
+ }
+
+ blkif->pending_list[idx].secs_pending += nsects;
+
+ switch (req->operation)
+ {
+ case BLKIF_OP_WRITE:
+ aiocb_info = malloc(sizeof(*aiocb_info));
+
+ aiocb_info->s = s;
+ aiocb_info->sector = sector_nr;
+ aiocb_info->nr_secs = nsects;
+ aiocb_info->idx = idx;
+ aiocb_info->i = i;
+
+ ret = (NULL == bdrv_aio_write(s->bs, sector_nr,
+ page, nsects,
+ qemu_send_responses,
+ aiocb_info));
+
+ if (ret) {
+ blkif->pending_list[idx].status =
BLKIF_RSP_ERROR;
+ DPRINTF("ERROR: bdrv_write() ==
NULL\n");
+ goto send_response;
+ }
+ break;
+
+ case BLKIF_OP_READ:
+ aiocb_info = malloc(sizeof(*aiocb_info));
+
+ aiocb_info->s = s;
+ aiocb_info->sector = sector_nr;
+ aiocb_info->nr_secs = nsects;
+ aiocb_info->idx = idx;
+ aiocb_info->i = i;
+
+ ret = (NULL == bdrv_aio_read(s->bs, sector_nr,
+ page, nsects,
+ qemu_send_responses,
+ aiocb_info));
+
+ if (ret) {
+ blkif->pending_list[idx].status =
BLKIF_RSP_ERROR;
+ DPRINTF("ERROR: bdrv_read() == NULL\n");
+ goto send_response;
+ }
+ break;
+
+ default:
+ DPRINTF("Unknown block operation\n");
+ break;
+ }
+ sector_nr += nsects;
+ }
+ send_response:
+ blkif->pending_list[idx].submitting = 0;
+
+ /* force write_rsp_to_ring for synchronous case */
+ if (blkif->pending_list[idx].secs_pending == 0)
+ send_responses(s, 0, 0, 0, idx, (void *)(long)0);
+ }
+}
+
+/**
+ * Callback function for the qemu-read pipe. Reads and processes control
+ * message from the pipe.
+ *
+ * The parameter private is unused.
+ */
+static void handle_blktap_ctrlmsg(void* private)
+{
+ int length, len, msglen;
+ char *ptr, *path;
+ image_t *img;
+ msg_hdr_t *msg;
+ msg_newdev_t *msg_dev;
+ msg_pid_t *msg_pid;
+ int ret = -1;
+ struct td_state *s = NULL;
+ fd_list_entry_t *entry;
+
+ char buf[MSG_SIZE];
+
+ length = read(read_fd, buf, MSG_SIZE);
+
+ if (length > 0 && length >= sizeof(msg_hdr_t))
+ {
+ msg = (msg_hdr_t *)buf;
+ DPRINTF("blktap: Received msg, len %d, type %d, UID %d\n",
+ length,msg->type,msg->cookie);
+
+ switch (msg->type) {
+ case CTLMSG_PARAMS:
+ ptr = buf + sizeof(msg_hdr_t);
+ len = (length - sizeof(msg_hdr_t));
+ path = calloc(1, len + 1);
+
+ memcpy(path, ptr, len);
+ DPRINTF("Received CTLMSG_PARAMS: [%s]\n", path);
+
+ /* Allocate the disk structs */
+ s = state_init();
+
+ /*Open file*/
+ if (s == NULL || open_disk(s, path, msg->readonly)) {
+ msglen = sizeof(msg_hdr_t);
+ msg->type = CTLMSG_IMG_FAIL;
+ msg->len = msglen;
+ } else {
+ entry = add_fd_entry(0, s);
+ entry->cookie = msg->cookie;
+ DPRINTF("Entered cookie %d\n", entry->cookie);
+
+ memset(buf, 0x00, MSG_SIZE);
+
+ msglen = sizeof(msg_hdr_t) + sizeof(image_t);
+ msg->type = CTLMSG_IMG;
+ img = (image_t *)(buf + sizeof(msg_hdr_t));
+ img->size = s->size;
+ img->secsize = s->sector_size;
+ img->info = s->info;
+ DPRINTF("Writing (size, secsize, info) = "
+ "(%#" PRIx64 ", %#" PRIx64 ", %d)\n",
+ s->size, s->sector_size, s->info);
+ }
+ len = write(write_fd, buf, msglen);
+ free(path);
+ break;
+
+ case CTLMSG_NEWDEV:
+ msg_dev = (msg_newdev_t *)(buf + sizeof(msg_hdr_t));
+
+ s = get_state(msg->cookie);
+ DPRINTF("Retrieving state, cookie %d.....[%s]\n",
+ msg->cookie, (s == NULL ? "FAIL":"OK"));
+ if (s != NULL) {
+ ret = ((map_new_dev(s, msg_dev->devnum)
+ == msg_dev->devnum ? 0: -1));
+ }
+
+ memset(buf, 0x00, MSG_SIZE);
+ msglen = sizeof(msg_hdr_t);
+ msg->type = (ret == 0 ? CTLMSG_NEWDEV_RSP
+ : CTLMSG_NEWDEV_FAIL);
+ msg->len = msglen;
+
+ len = write(write_fd, buf, msglen);
+ break;
+
+ case CTLMSG_CLOSE:
+ s = get_state(msg->cookie);
+ if (s) unmap_disk(s);
+ break;
+
+ case CTLMSG_PID:
+ memset(buf, 0x00, MSG_SIZE);
+ msglen = sizeof(msg_hdr_t) + sizeof(msg_pid_t);
+ msg->type = CTLMSG_PID_RSP;
+ msg->len = msglen;
+
+ msg_pid = (msg_pid_t *)(buf + sizeof(msg_hdr_t));
+ process = getpid();
+ msg_pid->pid = process;
+
+ len = write(write_fd, buf, msglen);
+ break;
+
+ default:
+ break;
+ }
+ }
+}
+
+/**
+ * Opens a control socket, i.e. a pipe to communicate with blktapctrl.
+ *
+ * Returns the file descriptor number for the pipe; -1 in error case
+ */
+static int open_ctrl_socket(char *devname)
+{
+ int ret;
+ int ipc_fd;
+
+ if (mkdir(BLKTAP_CTRL_DIR, 0755) == 0)
+ DPRINTF("Created %s directory\n", BLKTAP_CTRL_DIR);
+
+ ret = mkfifo(devname,S_IRWXU|S_IRWXG|S_IRWXO);
+ if ( (ret != 0) && (errno != EEXIST) ) {
+ DPRINTF("ERROR: pipe failed (%d)\n", errno);
+ return -1;
+ }
+
+ ipc_fd = open(devname,O_RDWR|O_NONBLOCK);
+
+ if (ipc_fd < 0) {
+ DPRINTF("FD open failed\n");
+ return -1;
+ }
+
+ return ipc_fd;
+}
+
+/**
+ * Unmaps all disks and closes their pipes
+ */
+void shutdown_blktap(void)
+{
+ fd_list_entry_t *ptr;
+ struct td_state *s;
+ char *devname;
+
+ DPRINTF("Shutdown blktap\n");
+
+ /* Unmap all disks */
+ ptr = fd_start;
+ while (ptr != NULL) {
+ s = ptr->s;
+ unmap_disk(s);
+ close(ptr->tap_fd);
+ ptr = ptr->next;
+ }
+
+ /* Delete control pipes */
+ if (asprintf(&devname, BLKTAP_CTRL_DIR "/qemu-read-%d", domid) >= 0) {
+ DPRINTF("Delete %s\n", devname);
+ if (unlink(devname))
+ DPRINTF("Could not delete: %s\n", strerror(errno));
+ free(devname);
+ }
+
+ if (asprintf(&devname, BLKTAP_CTRL_DIR "/qemu-write-%d", domid) >= 0) {
+ DPRINTF("Delete %s\n", devname);
+ if (unlink(devname))
+ DPRINTF("Could not delete: %s\n", strerror(errno));
+ free(devname);
+ }
+}
+
+/**
+ * Initialize the blktap interface, i.e. open a pair of pipes in /var/run/tap
+ * and register a fd handler.
+ *
+ * Returns 0 on success.
+ */
+int init_blktap(void)
+{
+ char* devname;
+
+ DPRINTF("Init blktap pipes\n");
+
+ /* Open the read pipe */
+ if (asprintf(&devname, BLKTAP_CTRL_DIR "/qemu-read-%d", domid) >= 0) {
+ read_fd = open_ctrl_socket(devname);
+ free(devname);
+
+ if (read_fd == -1) {
+ fprintf(stderr, "Could not open %s/qemu-read-%d\n",
+ BLKTAP_CTRL_DIR, domid);
+ return -1;
+ }
+ }
+
+ /* Open the write pipe */
+ if (asprintf(&devname, BLKTAP_CTRL_DIR "/qemu-write-%d", domid) >= 0) {
+ write_fd = open_ctrl_socket(devname);
+ free(devname);
+
+ if (write_fd == -1) {
+ fprintf(stderr, "Could not open %s/qemu-write-%d\n",
+ BLKTAP_CTRL_DIR, domid);
+ close(read_fd);
+ return -1;
+ }
+ }
+
+ /* Attach a handler to the read pipe (called from qemu main loop) */
+ qemu_set_fd_handler2(read_fd, NULL, &handle_blktap_ctrlmsg, NULL, NULL);
+
+ /* Register handler to clean up when the domain is destroyed */
+ atexit(&shutdown_blktap);
+
+ return 0;
+}
diff -r 8c921adf4833 -r 42f6c206c951 tools/ioemu/hw/xen_blktap.h
--- /dev/null Thu Jan 01 00:00:00 1970 +0000
+++ b/tools/ioemu/hw/xen_blktap.h Thu Mar 20 12:35:40 2008 -0600
@@ -0,0 +1,57 @@
+/* xen_blktap.h
+ *
+ * Generic disk interface for blktap-based image adapters.
+ *
+ * (c) 2006 Andrew Warfield and Julian Chesterfield
+ */
+
+#ifndef XEN_BLKTAP_H_
+#define XEN_BLKTAP_H_
+
+#include <stdint.h>
+#include <syslog.h>
+#include <stdio.h>
+
+#include "block_int.h"
+
+/* Things disks need to know about, these should probably be in a higher-level
+ * header. */
+#define MAX_SEGMENTS_PER_REQ 11
+#define SECTOR_SHIFT 9
+#define DEFAULT_SECTOR_SIZE 512
+
+#define MAX_IOFD 2
+
+#define BLK_NOT_ALLOCATED 99
+#define TD_NO_PARENT 1
+
+typedef uint32_t td_flag_t;
+
+#define TD_RDONLY 1
+
+struct disk_id {
+ char *name;
+ int drivertype;
+};
+
+/* This structure represents the state of an active virtual disk. */
+struct td_state {
+ BlockDriverState* bs;
+ td_flag_t flags;
+ void *blkif;
+ void *image;
+ void *ring_info;
+ void *fd_entry;
+ uint64_t sector_size;
+ uint64_t size;
+ unsigned int info;
+};
+
+typedef struct fd_list_entry {
+ int cookie;
+ int tap_fd;
+ struct td_state *s;
+ struct fd_list_entry **pprev, *next;
+} fd_list_entry_t;
+
+#endif /*XEN_BLKTAP_H_*/
diff -r 8c921adf4833 -r 42f6c206c951 tools/ioemu/hw/xen_machine_pv.c
--- a/tools/ioemu/hw/xen_machine_pv.c Fri Mar 14 15:07:45 2008 -0600
+++ b/tools/ioemu/hw/xen_machine_pv.c Thu Mar 20 12:35:40 2008 -0600
@@ -26,6 +26,9 @@
#include "xen_console.h"
#include "xenfb.h"
+extern void init_blktap(void);
+
+
/* The Xen PV machine currently provides
* - a virtual framebuffer
* - ....
@@ -40,6 +43,12 @@ static void xen_init_pv(uint64_t ram_siz
{
struct xenfb *xenfb;
extern int domid;
+
+
+#ifndef CONFIG_STUBDOM
+ /* Initialize tapdisk client */
+ init_blktap();
+#endif
/* Connect to text console */
if (serial_hds[0]) {
diff -r 8c921adf4833 -r 42f6c206c951 tools/ioemu/sdl.c
--- a/tools/ioemu/sdl.c Fri Mar 14 15:07:45 2008 -0600
+++ b/tools/ioemu/sdl.c Thu Mar 20 12:35:40 2008 -0600
@@ -29,6 +29,10 @@
#include <signal.h>
#endif
+#ifdef CONFIG_OPENGL
+#include <SDL_opengl.h>
+#endif
+
static SDL_Surface *screen;
static SDL_Surface *shared = NULL;
static int gui_grab; /* if true, all keyboard/mouse events are grabbed */
@@ -44,6 +48,99 @@ static SDL_Cursor *sdl_cursor_normal;
static SDL_Cursor *sdl_cursor_normal;
static SDL_Cursor *sdl_cursor_hidden;
static int absolute_enabled = 0;
+static int opengl_enabled;
+
+#ifdef CONFIG_OPENGL
+static GLint tex_format;
+static GLint tex_type;
+static GLuint texture_ref = 0;
+static GLint gl_format;
+
+static void opengl_setdata(DisplayState *ds, void *pixels)
+{
+ glEnable(GL_TEXTURE_RECTANGLE_ARB);
+ glTexEnvi(GL_TEXTURE_ENV, GL_TEXTURE_ENV_MODE, GL_REPLACE);
+ glClearColor(0, 0, 0, 0);
+ glDisable(GL_BLEND);
+ glDisable(GL_LIGHTING);
+ glDisable(GL_DEPTH_TEST);
+ glDepthMask(GL_FALSE);
+ glDisable(GL_CULL_FACE);
+ glViewport( 0, 0, screen->w, screen->h);
+ glMatrixMode(GL_PROJECTION);
+ glLoadIdentity();
+ glOrtho(0, screen->w, screen->h, 0, -1,1);
+ glMatrixMode(GL_MODELVIEW);
+ glLoadIdentity();
+ glClear(GL_COLOR_BUFFER_BIT);
+ ds->data = pixels;
+
+ if (texture_ref) {
+ glDeleteTextures(1, &texture_ref);
+ texture_ref = 0;
+ }
+
+ glGenTextures(1, &texture_ref);
+ glBindTexture(GL_TEXTURE_RECTANGLE_ARB, texture_ref);
+ glPixelStorei(GL_UNPACK_LSB_FIRST, 1);
+ switch (ds->depth) {
+ case 8:
+ tex_format = GL_RGB;
+ tex_type = GL_UNSIGNED_BYTE_3_3_2;
+ glPixelStorei (GL_UNPACK_ALIGNMENT, 1);
+ break;
+ case 16:
+ tex_format = GL_RGB;
+ tex_type = GL_UNSIGNED_SHORT_5_6_5;
+ glPixelStorei (GL_UNPACK_ALIGNMENT, 2);
+ break;
+ case 24:
+ tex_format = GL_BGR;
+ tex_type = GL_UNSIGNED_BYTE;
+ glPixelStorei (GL_UNPACK_ALIGNMENT, 1);
+ break;
+ case 32:
+ if (!ds->bgr) {
+ tex_format = GL_BGRA;
+ tex_type = GL_UNSIGNED_BYTE;
+ } else {
+ tex_format = GL_RGBA;
+ tex_type = GL_UNSIGNED_BYTE;
+ }
+ glPixelStorei (GL_UNPACK_ALIGNMENT, 4);
+ break;
+ }
+ glPixelStorei(GL_UNPACK_ROW_LENGTH, (ds->linesize * 8) / ds->depth);
+ glTexImage2D(GL_TEXTURE_RECTANGLE_ARB, 0, gl_format, ds->width,
ds->height, 0, tex_format, tex_type, pixels);
+ glTexParameterf(GL_TEXTURE_RECTANGLE_ARB, GL_TEXTURE_PRIORITY, 1.0);
+ glTexParameteri(GL_TEXTURE_RECTANGLE_ARB, GL_TEXTURE_MIN_FILTER,
GL_LINEAR);
+ glTexParameteri(GL_TEXTURE_RECTANGLE_ARB, GL_TEXTURE_MAG_FILTER,
GL_LINEAR);
+ glTexParameteri(GL_TEXTURE_RECTANGLE_ARB, GL_TEXTURE_WRAP_S,
GL_CLAMP_TO_EDGE);
+ glTexParameteri(GL_TEXTURE_RECTANGLE_ARB, GL_TEXTURE_WRAP_T,
GL_CLAMP_TO_EDGE);
+ glBindTexture(GL_TEXTURE_RECTANGLE_ARB, 0);
+}
+
+static void opengl_update(DisplayState *ds, int x, int y, int w, int h)
+{
+ int bpp = ds->depth / 8;
+ GLvoid *pixels = ds->data + y * ds->linesize + x * bpp;
+ glBindTexture(GL_TEXTURE_RECTANGLE_ARB, texture_ref);
+ glPixelStorei(GL_UNPACK_ROW_LENGTH, ds->linesize / bpp);
+ glTexSubImage2D(GL_TEXTURE_RECTANGLE_ARB, 0, x, y, w, h, tex_format,
tex_type, pixels);
+ glBegin(GL_QUADS);
+ glTexCoord2d(0, 0);
+ glVertex2d(0, 0);
+ glTexCoord2d(ds->width, 0);
+ glVertex2d(screen->w, 0);
+ glTexCoord2d(ds->width, ds->height);
+ glVertex2d(screen->w, screen->h);
+ glTexCoord2d(0, ds->height);
+ glVertex2d(0, screen->h);
+ glEnd();
+ glBindTexture(GL_TEXTURE_RECTANGLE_ARB, 0);
+ SDL_GL_SwapBuffers();
+}
+#endif
static void sdl_update(DisplayState *ds, int x, int y, int w, int h)
{
@@ -96,17 +193,26 @@ static void sdl_resize(DisplayState *ds,
// printf("resizing to %d %d\n", w, h);
- flags =
SDL_HWSURFACE|SDL_ASYNCBLIT|SDL_HWACCEL|SDL_DOUBLEBUF|SDL_HWPALETTE;
- if (gui_fullscreen)
+#ifdef CONFIG_OPENGL
+ if (ds->shared_buf && opengl_enabled)
+ flags = SDL_OPENGL|SDL_RESIZABLE;
+ else
+#endif
+ flags =
SDL_HWSURFACE|SDL_ASYNCBLIT|SDL_HWACCEL|SDL_DOUBLEBUF|SDL_HWPALETTE;
+
+ if (gui_fullscreen) {
flags |= SDL_FULLSCREEN;
-
+ flags &= ~SDL_RESIZABLE;
+ }
+
width = w;
height = h;
again:
screen = SDL_SetVideoMode(w, h, 0, flags);
+#ifndef CONFIG_OPENGL
if (!screen) {
- fprintf(stderr, "Could not open SDL display\n");
+ fprintf(stderr, "Could not open SDL display: %s\n", SDL_GetError());
exit(1);
}
if (!screen->pixels && (flags & SDL_HWSURFACE) && (flags &
SDL_FULLSCREEN)) {
@@ -115,9 +221,10 @@ static void sdl_resize(DisplayState *ds,
}
if (!screen->pixels) {
- fprintf(stderr, "Could not open SDL display\n");
+ fprintf(stderr, "Could not open SDL display: %s\n", SDL_GetError());
exit(1);
}
+#endif
ds->width = w;
ds->height = h;
if (!ds->shared_buf) {
@@ -131,6 +238,25 @@ static void sdl_resize(DisplayState *ds,
ds->linesize = screen->pitch;
} else {
ds->linesize = linesize;
+#ifdef CONFIG_OPENGL
+ switch(screen->format->BitsPerPixel) {
+ case 8:
+ gl_format = GL_RGB;
+ break;
+ case 16:
+ gl_format = GL_RGB;
+ break;
+ case 24:
+ gl_format = GL_RGB;
+ break;
+ case 32:
+ if (!screen->format->Rshift)
+ gl_format = GL_BGRA;
+ else
+ gl_format = GL_RGBA;
+ break;
+ };
+#endif
}
}
@@ -139,7 +265,13 @@ static void sdl_colourdepth(DisplayState
if (!depth || !ds->depth) return;
ds->shared_buf = 1;
ds->depth = depth;
- ds->linesize = width * depth / 8;
+ ds->linesize = width * depth / 8;
+#ifdef CONFIG_OPENGL
+ if (opengl_enabled) {
+ ds->dpy_update = opengl_update;
+ ds->dpy_setdata = opengl_setdata;
+ }
+#endif
}
/* generic keyboard conversion */
@@ -331,8 +463,8 @@ static void sdl_send_mouse_event(int dx,
}
SDL_GetMouseState(&dx, &dy);
- dx = dx * 0x7FFF / (width - 1);
- dy = dy * 0x7FFF / (height - 1);
+ dx = dx * 0x7FFF / (screen->w - 1);
+ dy = dy * 0x7FFF / (screen->h - 1);
} else if (absolute_enabled) {
sdl_show_cursor();
absolute_enabled = 0;
@@ -344,7 +476,8 @@ static void toggle_full_screen(DisplaySt
static void toggle_full_screen(DisplayState *ds)
{
gui_fullscreen = !gui_fullscreen;
- sdl_resize(ds, screen->w, screen->h, ds->linesize);
+ sdl_resize(ds, ds->width, ds->height, ds->linesize);
+ ds->dpy_setdata(ds, ds->data);
if (gui_fullscreen) {
gui_saved_grab = gui_grab;
sdl_grab_start();
@@ -371,7 +504,7 @@ static void sdl_refresh(DisplayState *ds
while (SDL_PollEvent(ev)) {
switch (ev->type) {
case SDL_VIDEOEXPOSE:
- sdl_update(ds, 0, 0, screen->w, screen->h);
+ ds->dpy_update(ds, 0, 0, ds->width, ds->height);
break;
case SDL_KEYDOWN:
case SDL_KEYUP:
@@ -528,6 +661,18 @@ static void sdl_refresh(DisplayState *ds
}
}
break;
+#ifdef CONFIG_OPENGL
+ case SDL_VIDEORESIZE:
+ {
+ if (ds->shared_buf && opengl_enabled) {
+ SDL_ResizeEvent *rev = &ev->resize;
+ screen = SDL_SetVideoMode(rev->w, rev->h, 0,
SDL_OPENGL|SDL_RESIZABLE);
+ opengl_setdata(ds, ds->data);
+ opengl_update(ds, 0, 0, ds->width, ds->height);
+ }
+ break;
+ }
+#endif
default:
break;
}
@@ -536,13 +681,17 @@ static void sdl_refresh(DisplayState *ds
static void sdl_cleanup(void)
{
+#ifdef CONFIG_OPENGL
+ if (texture_ref) glDeleteTextures(1, &texture_ref);
+#endif
SDL_Quit();
}
-void sdl_display_init(DisplayState *ds, int full_screen)
+void sdl_display_init(DisplayState *ds, int full_screen, int opengl)
{
int flags;
uint8_t data = 0;
+ opengl_enabled = opengl;
#if defined(__APPLE__)
/* always use generic keymaps */
diff -r 8c921adf4833 -r 42f6c206c951 tools/ioemu/vl.c
--- a/tools/ioemu/vl.c Fri Mar 14 15:07:45 2008 -0600
+++ b/tools/ioemu/vl.c Thu Mar 20 12:35:40 2008 -0600
@@ -174,6 +174,11 @@ int graphic_height = 600;
#endif
int graphic_depth = 15;
int full_screen = 0;
+#ifdef CONFIG_OPENGL
+int opengl_enabled = 1;
+#else
+int opengl_enabled = 0;
+#endif
int no_quit = 0;
CharDriverState *serial_hds[MAX_SERIAL_PORTS];
CharDriverState *parallel_hds[MAX_PARALLEL_PORTS];
@@ -276,9 +281,9 @@ void default_ioport_writel(void *opaque,
void init_ioports(void)
{
- ioport_opaque = malloc(MAX_IOPORTS * sizeof(*ioport_opaque));
- ioport_read_table = malloc(3 * MAX_IOPORTS * sizeof(**ioport_read_table));
- ioport_write_table = malloc(3 * MAX_IOPORTS *
sizeof(**ioport_write_table));
+ ioport_opaque = calloc(MAX_IOPORTS, sizeof(*ioport_opaque));
+ ioport_read_table = calloc(3 * MAX_IOPORTS, sizeof(**ioport_read_table));
+ ioport_write_table = calloc(3 * MAX_IOPORTS, sizeof(**ioport_write_table));
}
/* size is the word size in byte */
@@ -6272,6 +6277,12 @@ void qemu_system_powerdown_request(void)
cpu_interrupt(cpu_single_env, CPU_INTERRUPT_EXIT);
}
+static void qemu_sighup_handler(int signal)
+{
+ fprintf(stderr, "Received SIGHUP, terminating.\n");
+ exit(0);
+}
+
void main_loop_wait(int timeout)
{
IOHandlerRecord *ioh;
@@ -6482,6 +6493,9 @@ void help(void)
#ifdef CONFIG_SDL
"-no-quit disable SDL window close capability\n"
#endif
+#ifdef CONFIG_OPENGL
+ "-disable-opengl disable OpenGL rendering, using SDL"
+#endif
#ifdef TARGET_I386
"-no-fd-bootchk disable boot signature checking for floppy disks\n"
#endif
@@ -6660,6 +6674,7 @@ enum {
QEMU_OPTION_loadvm,
QEMU_OPTION_full_screen,
QEMU_OPTION_no_quit,
+ QEMU_OPTION_disable_opengl,
QEMU_OPTION_pidfile,
QEMU_OPTION_no_kqemu,
QEMU_OPTION_kernel_kqemu,
@@ -6757,6 +6772,7 @@ const QEMUOption qemu_options[] = {
#ifdef CONFIG_SDL
{ "no-quit", 0, QEMU_OPTION_no_quit },
#endif
+ { "disable-opengl", 0, QEMU_OPTION_disable_opengl },
{ "pidfile", HAS_ARG, QEMU_OPTION_pidfile },
{ "win2k-hack", 0, QEMU_OPTION_win2k_hack },
{ "usbdevice", HAS_ARG, QEMU_OPTION_usbdevice },
@@ -7528,6 +7544,9 @@ int main(int argc, char **argv)
no_quit = 1;
break;
#endif
+ case QEMU_OPTION_disable_opengl:
+ opengl_enabled = 0;
+ break;
case QEMU_OPTION_pidfile:
create_pidfile(optarg);
break;
@@ -7854,7 +7873,7 @@ int main(int argc, char **argv)
xenstore_write_vncport(vnc_display_port);
} else {
#if defined(CONFIG_SDL)
- sdl_display_init(ds, full_screen);
+ sdl_display_init(ds, full_screen, opengl_enabled);
#elif defined(CONFIG_COCOA)
cocoa_display_init(ds, full_screen);
#else
@@ -7980,7 +7999,7 @@ int main(int argc, char **argv)
#ifndef CONFIG_STUBDOM
/* Unblock SIGTERM and SIGHUP, which may have been blocked by the caller */
- signal(SIGHUP, SIG_DFL);
+ signal(SIGHUP, qemu_sighup_handler);
sigemptyset(&set);
sigaddset(&set, SIGTERM);
sigaddset(&set, SIGHUP);
diff -r 8c921adf4833 -r 42f6c206c951 tools/ioemu/vl.h
--- a/tools/ioemu/vl.h Fri Mar 14 15:07:45 2008 -0600
+++ b/tools/ioemu/vl.h Thu Mar 20 12:35:40 2008 -0600
@@ -982,7 +982,7 @@ void isa_cirrus_vga_init(DisplayState *d
unsigned long vga_ram_offset, int vga_ram_size);
/* sdl.c */
-void sdl_display_init(DisplayState *ds, int full_screen);
+void sdl_display_init(DisplayState *ds, int full_screen, int opengl_enable);
/* cocoa.m */
void cocoa_display_init(DisplayState *ds, int full_screen);
diff -r 8c921adf4833 -r 42f6c206c951 tools/ioemu/vnc.c
--- a/tools/ioemu/vnc.c Fri Mar 14 15:07:45 2008 -0600
+++ b/tools/ioemu/vnc.c Thu Mar 20 12:35:40 2008 -0600
@@ -369,20 +369,21 @@ static void vnc_dpy_resize(DisplayState
VncState *vs = ds->opaque;
int o;
- if (linesize != w * vs->depth)
- ds->shared_buf = 0;
-
if (!ds->shared_buf) {
+ ds->linesize = w * vs->depth;
if (allocated)
- ds->data = realloc(ds->data, w * h * vs->depth);
+ ds->data = realloc(ds->data, h * ds->linesize);
else
- ds->data = malloc(w * h * vs->depth);
+ ds->data = malloc(h * ds->linesize);
allocated = 1;
- } else if (allocated) {
- free(ds->data);
- allocated = 0;
- }
- vs->old_data = realloc(vs->old_data, w * h * vs->depth);
+ } else {
+ ds->linesize = linesize;
+ if (allocated) {
+ free(ds->data);
+ allocated = 0;
+ }
+ }
+ vs->old_data = realloc(vs->old_data, h * ds->linesize);
vs->dirty_row = realloc(vs->dirty_row, h * sizeof(vs->dirty_row[0]));
vs->update_row = realloc(vs->update_row, h * sizeof(vs->dirty_row[0]));
@@ -399,7 +400,6 @@ static void vnc_dpy_resize(DisplayState
size_changed = ds->width != w || ds->height != h;
ds->width = w;
ds->height = h;
- ds->linesize = w * vs->depth;
if (vs->csock != -1 && vs->has_resize && size_changed) {
vs->width = ds->width;
vs->height = ds->height;
@@ -2494,6 +2494,7 @@ void vnc_display_init(DisplayState *ds)
vs->ds->width = 640;
vs->ds->height = 400;
+ vs->ds->linesize = 640 * 4;
vnc_dpy_colourdepth(vs->ds, 24);
}
diff -r 8c921adf4833 -r 42f6c206c951 tools/ioemu/xenstore.c
--- a/tools/ioemu/xenstore.c Fri Mar 14 15:07:45 2008 -0600
+++ b/tools/ioemu/xenstore.c Thu Mar 20 12:35:40 2008 -0600
@@ -81,7 +81,7 @@ static void waitForDevice(char *fn)
#define DIRECT_PCI_STR_LEN 160
char direct_pci_str[DIRECT_PCI_STR_LEN];
-void xenstore_parse_domain_config(int domid)
+void xenstore_parse_domain_config(int hvm_domid)
{
char **e = NULL;
char *buf = NULL, *path;
@@ -100,7 +100,7 @@ void xenstore_parse_domain_config(int do
return;
}
- path = xs_get_domain_path(xsh, domid);
+ path = xs_get_domain_path(xsh, hvm_domid);
if (path == NULL) {
fprintf(logfile, "xs_get_domain_path() error\n");
goto out;
diff -r 8c921adf4833 -r 42f6c206c951 tools/libxc/ia64/dom_fw_acpi.c
--- a/tools/libxc/ia64/dom_fw_acpi.c Fri Mar 14 15:07:45 2008 -0600
+++ b/tools/libxc/ia64/dom_fw_acpi.c Thu Mar 20 12:35:40 2008 -0600
@@ -1,4 +1,5 @@
#include <inttypes.h>
+#include "xc_dom_ia64_util.h"
#include <xen/acpi.h>
uint8_t
diff -r 8c921adf4833 -r 42f6c206c951 tools/libxc/ia64/xc_dom_ia64_util.h
--- a/tools/libxc/ia64/xc_dom_ia64_util.h Fri Mar 14 15:07:45 2008 -0600
+++ b/tools/libxc/ia64/xc_dom_ia64_util.h Thu Mar 20 12:35:40 2008 -0600
@@ -23,4 +23,8 @@ extern int shared_info_ia64(struct xc_do
#define FW_MEM_BASE 0xff000000UL
#define FW_MEM_SIZE 0x01000000UL
+#ifdef __XEN_TOOLS__
+/* Necessary for including the acpi header chain when not in kernel context */
+typedef struct { } spinlock_t;
+#endif
#endif /* XC_IA64_DOM_IA64_UTIL_H */
diff -r 8c921adf4833 -r 42f6c206c951 tools/libxc/xc_domain_save.c
--- a/tools/libxc/xc_domain_save.c Fri Mar 14 15:07:45 2008 -0600
+++ b/tools/libxc/xc_domain_save.c Thu Mar 20 12:35:40 2008 -0600
@@ -123,36 +123,6 @@ static inline int count_bits ( int nr, v
for ( i = 0; i < (nr / (sizeof(unsigned long)*8)); i++, p++ )
count += hweight32(*p);
return count;
-}
-
-static inline int permute( int i, int nr, int order_nr )
-{
- /* Need a simple permutation function so that we scan pages in a
- pseudo random order, enabling us to get a better estimate of
- the domain's page dirtying rate as we go (there are often
- contiguous ranges of pfns that have similar behaviour, and we
- want to mix them up. */
-
- /* e.g. nr->oder 15->4 16->4 17->5 */
- /* 512MB domain, 128k pages, order 17 */
-
- /*
- QPONMLKJIHGFEDCBA
- QPONMLKJIH
- GFEDCBA
- */
-
- /*
- QPONMLKJIHGFEDCBA
- EDCBA
- QPONM
- LKJIHGF
- */
-
- do { i = ((i>>(order_nr-10)) | ( i<<10 ) ) & ((1<<order_nr)-1); }
- while ( i >= nr ); /* this won't ever loop if nr is a power of 2 */
-
- return i;
}
static uint64_t tv_to_us(struct timeval *new)
@@ -859,9 +829,6 @@ int xc_domain_save(int xc_handle, int io
/* base of the region in which domain memory is mapped */
unsigned char *region_base = NULL;
- /* power of 2 order of p2m_size */
- int order_nr;
-
/* bitmap of pages:
- that should be sent this iteration (unless later marked as skip);
- to skip this iteration because already dirty;
@@ -970,11 +937,6 @@ int xc_domain_save(int xc_handle, int io
/* pretend we sent all the pages last iteration */
sent_last_iter = p2m_size;
-
- /* calculate the power of 2 order of p2m_size, e.g.
- 15->4 16->4 17->5 */
- for ( i = p2m_size-1, order_nr = 0; i ; i >>= 1, order_nr++ )
- continue;
/* Setup to_send / to_fix and to_skip bitmaps */
to_send = malloc(BITMAP_SIZE);
@@ -1126,7 +1088,7 @@ int xc_domain_save(int xc_handle, int io
(batch < MAX_BATCH_SIZE) && (N < p2m_size);
N++ )
{
- int n = permute(N, p2m_size, order_nr);
+ int n = N;
if ( debug )
{
diff -r 8c921adf4833 -r 42f6c206c951 tools/libxc/xc_linux.c
--- a/tools/libxc/xc_linux.c Fri Mar 14 15:07:45 2008 -0600
+++ b/tools/libxc/xc_linux.c Thu Mar 20 12:35:40 2008 -0600
@@ -4,7 +4,7 @@
* Use is subject to license terms.
*
* xc_gnttab functions:
- * Copyright (c) 2007, D G Murray <Derek.Murray@xxxxxxxxxxxx>
+ * Copyright (c) 2007-2008, D G Murray <Derek.Murray@xxxxxxxxxxxx>
*
* This program is free software; you can redistribute it and/or
* modify it under the terms of the GNU General Public License as
@@ -505,6 +505,19 @@ int xc_gnttab_munmap(int xcg_handle,
return 0;
}
+int xc_gnttab_set_max_grants(int xcg_handle,
+ uint32_t count)
+{
+ struct ioctl_gntdev_set_max_grants set_max;
+ int rc;
+
+ set_max.count = count;
+ if ( (rc = ioctl(xcg_handle, IOCTL_GNTDEV_SET_MAX_GRANTS, &set_max)) )
+ return rc;
+
+ return 0;
+}
+
/*
* Local variables:
* mode: C
diff -r 8c921adf4833 -r 42f6c206c951 tools/libxc/xc_minios.c
--- a/tools/libxc/xc_minios.c Fri Mar 14 15:07:45 2008 -0600
+++ b/tools/libxc/xc_minios.c Thu Mar 20 12:35:40 2008 -0600
@@ -165,14 +165,6 @@ static int port_alloc(int xce_handle) {
return i;
}
-static void poke_port(int xce_handle, evtchn_port_t port)
-{
- shared_info_t *s = HYPERVISOR_shared_info;
- printk("poking port %d\n", port);
- synch_set_bit(port, &s->evtchn_pending[0]);
- xc_evtchn_unmask(xce_handle, port);
-}
-
static void evtchn_handler(evtchn_port_t port, struct pt_regs *regs, void
*data)
{
int xce_handle = (intptr_t) data;
@@ -211,6 +203,7 @@ evtchn_port_or_error_t xc_evtchn_bind_un
}
files[xce_handle].evtchn.ports[i].bound = 1;
files[xce_handle].evtchn.ports[i].port = port;
+ unmask_evtchn(port);
return port;
}
@@ -235,9 +228,7 @@ evtchn_port_or_error_t xc_evtchn_bind_in
}
files[xce_handle].evtchn.ports[i].bound = 1;
files[xce_handle].evtchn.ports[i].port = local_port;
-/* Poke port on start: HVM won't send an event for the very first request since
- * we were not ready yet */
- poke_port(xce_handle, local_port);
+ unmask_evtchn(local_port);
return local_port;
}
@@ -275,6 +266,7 @@ evtchn_port_or_error_t xc_evtchn_bind_vi
}
files[xce_handle].evtchn.ports[i].bound = 1;
files[xce_handle].evtchn.ports[i].port = port;
+ unmask_evtchn(port);
return port;
}
diff -r 8c921adf4833 -r 42f6c206c951 tools/libxc/xenctrl.h
--- a/tools/libxc/xenctrl.h Fri Mar 14 15:07:45 2008 -0600
+++ b/tools/libxc/xenctrl.h Thu Mar 20 12:35:40 2008 -0600
@@ -6,7 +6,7 @@
* Copyright (c) 2003-2004, K A Fraser.
*
* xc_gnttab functions:
- * Copyright (c) 2007, D G Murray <Derek.Murray@xxxxxxxxxxxx>
+ * Copyright (c) 2007-2008, D G Murray <Derek.Murray@xxxxxxxxxxxx>
*/
#ifndef XENCTRL_H
@@ -832,6 +832,20 @@ int xc_gnttab_munmap(int xcg_handle,
void *start_address,
uint32_t count);
+/*
+ * Sets the maximum number of grants that may be mapped by the given instance
+ * to @count.
+ *
+ * N.B. This function must be called after opening the handle, and before any
+ * other functions are invoked on it.
+ *
+ * N.B. When variable-length grants are mapped, fragmentation may be observed,
+ * and it may not be possible to satisfy requests up to the maximum number
+ * of grants.
+ */
+int xc_gnttab_set_max_grants(int xcg_handle,
+ uint32_t count);
+
int xc_hvm_set_pci_intx_level(
int xc_handle, domid_t dom,
uint8_t domain, uint8_t bus, uint8_t device, uint8_t intx,
diff -r 8c921adf4833 -r 42f6c206c951 tools/python/xen/xend/XendConfig.py
--- a/tools/python/xen/xend/XendConfig.py Fri Mar 14 15:07:45 2008 -0600
+++ b/tools/python/xen/xend/XendConfig.py Thu Mar 20 12:35:40 2008 -0600
@@ -123,19 +123,49 @@ XENAPI_CFG_TO_LEGACY_CFG = {
LEGACY_CFG_TO_XENAPI_CFG = reverse_dict(XENAPI_CFG_TO_LEGACY_CFG)
-# Platform configuration keys.
-XENAPI_PLATFORM_CFG = [ 'acpi', 'apic', 'boot', 'device_model', 'loader',
'display',
- 'fda', 'fdb', 'keymap', 'isa', 'localtime', 'monitor',
- 'nographic', 'pae', 'rtc_timeoffset', 'serial', 'sdl',
- 'soundhw','stdvga', 'usb', 'usbdevice', 'hpet', 'vnc',
- 'vncconsole', 'vncdisplay', 'vnclisten', 'timer_mode',
- 'vncpasswd', 'vncunused', 'xauthority', 'pci', 'vhpt',
- 'guest_os_type', 'hap']
+# Platform configuration keys and their types.
+XENAPI_PLATFORM_CFG_TYPES = {
+ 'acpi': int,
+ 'apic': int,
+ 'boot': str,
+ 'device_model': str,
+ 'loader': str,
+ 'display' : str,
+ 'fda': str,
+ 'fdb': str,
+ 'keymap': str,
+ 'isa' : int,
+ 'localtime': int,
+ 'monitor': int,
+ 'nographic': int,
+ 'pae' : int,
+ 'rtc_timeoffset': int,
+ 'serial': str,
+ 'sdl': int,
+ 'opengl': int,
+ 'soundhw': str,
+ 'stdvga': int,
+ 'usb': int,
+ 'usbdevice': str,
+ 'hpet': int,
+ 'vnc': int,
+ 'vncconsole': int,
+ 'vncdisplay': int,
+ 'vnclisten': str,
+ 'timer_mode': int,
+ 'vncpasswd': str,
+ 'vncunused': int,
+ 'xauthority': str,
+ 'pci': str,
+ 'vhpt': int,
+ 'guest_os_type': str,
+ 'hap': int,
+}
# Xen API console 'other_config' keys.
XENAPI_CONSOLE_OTHER_CFG = ['vncunused', 'vncdisplay', 'vnclisten',
'vncpasswd', 'type', 'display', 'xauthority',
- 'keymap']
+ 'keymap', 'opengl']
# List of XendConfig configuration keys that have no direct equivalent
# in the old world.
@@ -405,6 +435,12 @@ class XendConfig(dict):
self['platform']['device_model'] =
xen.util.auxbin.pathTo("qemu-dm")
if self.is_hvm():
+ if 'timer_mode' not in self['platform']:
+ self['platform']['timer_mode'] = 0
+ if 'rtc_timeoffset' not in self['platform']:
+ self['platform']['rtc_timeoffset'] = 0
+ if 'hpet' not in self['platform']:
+ self['platform']['hpet'] = 0
if 'loader' not in self['platform']:
# Old configs may have hvmloader set as PV_kernel param
if self.has_key('PV_kernel') and re.search('hvmloader',
self['PV_kernel']):
@@ -534,7 +570,7 @@ class XendConfig(dict):
cfg['platform']['localtime'] = localtime
# Compatibility hack -- can go soon.
- for key in XENAPI_PLATFORM_CFG:
+ for key in XENAPI_PLATFORM_CFG_TYPES.keys():
val = sxp.child_value(sxp_cfg, "platform_" + key, None)
if val is not None:
self['platform'][key] = val
@@ -713,7 +749,7 @@ class XendConfig(dict):
self.update_with_image_sxp(image_sxp)
# Convert Legacy HVM parameters to Xen API configuration
- for key in XENAPI_PLATFORM_CFG:
+ for key in XENAPI_PLATFORM_CFG_TYPES.keys():
if key in cfg:
self['platform'][key] = cfg[key]
@@ -763,7 +799,7 @@ class XendConfig(dict):
if image_type != 'hvm' and image_type != 'linux':
self['platform']['image_type'] = image_type
- for key in XENAPI_PLATFORM_CFG:
+ for key in XENAPI_PLATFORM_CFG_TYPES.keys():
val = sxp.child_value(image_sxp, key, None)
if val is not None and val != '':
self['platform'][key] = val
@@ -847,6 +883,19 @@ class XendConfig(dict):
self[key] = type_conv(val)
else:
self[key] = val
+
+ # XenAPI defines platform as a string-string map. If platform
+ # configuration exists, convert values to appropriate type.
+ if 'platform' in xapi:
+ for key, val in xapi['platform'].items():
+ type_conv = XENAPI_PLATFORM_CFG_TYPES.get(key)
+ if type_conv is None:
+ key = key.lower()
+ type_conv = XENAPI_PLATFORM_CFG_TYPES.get(key)
+ if callable(type_conv):
+ self['platform'][key] = type_conv(val)
+ else:
+ self['platform'][key] = val
self['vcpus_params']['weight'] = \
int(self['vcpus_params'].get('weight', 256))
@@ -942,6 +991,7 @@ class XendConfig(dict):
dev_type, dev_cfg =
self['devices'][dev_uuid]
is_bootable = dev_cfg.get('bootable', 0)
config.append(['bootable',
int(is_bootable)])
+ config.append(['VDI', dev_cfg.get('VDI',
'')])
sxpr.append(['device', config])
@@ -1276,6 +1326,12 @@ class XendConfig(dict):
target['devices'][dev_uuid] = ('vfb', dev_info)
target['console_refs'].append(dev_uuid)
+ # if console is rfb, set device_model ensuring qemu
+ # is invoked for pvfb services
+ if 'device_model' not in target['platform']:
+ target['platform']['device_model'] = \
+ xen.util.auxbin.pathTo("qemu-dm")
+
# Finally, if we are a pvfb, we need to make a vkbd
# as well that is not really exposed to Xen API
vkbd_uuid = uuid.createString()
@@ -1407,6 +1463,23 @@ class XendConfig(dict):
config = cfg_sxp
dev_type, dev_info = self['devices'][dev_uuid]
+
+ if dev_type == 'pci': # Special case for pci
+ pci_devs = []
+ for pci_dev in sxp.children(config, 'dev'):
+ pci_dev_info = {}
+ for opt_val in pci_dev[1:]:
+ try:
+ opt, val = opt_val
+ pci_dev_info[opt] = val
+ except TypeError:
+ pass
+ pci_devs.append(pci_dev_info)
+ self['devices'][dev_uuid] = (dev_type,
+ {'devs': pci_devs,
+ 'uuid': dev_uuid})
+ return True
+
for opt_val in config[1:]:
try:
opt, val = opt_val
@@ -1519,7 +1592,7 @@ class XendConfig(dict):
if self.has_key('PV_args') and self['PV_args']:
image.append(['args', self['PV_args']])
- for key in XENAPI_PLATFORM_CFG:
+ for key in XENAPI_PLATFORM_CFG_TYPES.keys():
if key in self['platform']:
image.append([key, self['platform'][key]])
@@ -1555,7 +1628,7 @@ class XendConfig(dict):
self['PV_ramdisk'] = sxp.child_value(image_sxp, 'ramdisk','')
self['PV_args'] = kernel_args
- for key in XENAPI_PLATFORM_CFG:
+ for key in XENAPI_PLATFORM_CFG_TYPES.keys():
val = sxp.child_value(image_sxp, key, None)
if val is not None and val != '':
self['platform'][key] = val
diff -r 8c921adf4833 -r 42f6c206c951 tools/python/xen/xend/XendDomainInfo.py
--- a/tools/python/xen/xend/XendDomainInfo.py Fri Mar 14 15:07:45 2008 -0600
+++ b/tools/python/xen/xend/XendDomainInfo.py Thu Mar 20 12:35:40 2008 -0600
@@ -558,18 +558,17 @@ class XendDomainInfo:
count += 1
- def pci_device_create(self, dev_config):
- log.debug("XendDomainInfo.pci_device_create: %s" %
scrub_password(dev_config))
+ def hvm_pci_device_create(self, dev_config):
+ log.debug("XendDomainInfo.hvm_pci_device_create: %s"
+ % scrub_password(dev_config))
if not self.info.is_hvm():
- raise VmError("only HVM guest support pci attach")
+ raise VmError("hvm_pci_device_create called on non-HVM guest")
#all the PCI devs share one conf node
devid = '0'
- dev_type = sxp.name(dev_config)
- new_devs = sxp.child_value(dev_config, 'devs')
- new_dev = new_devs[0]
+ new_dev = dev_config['devs'][0]
dev_info = self._getDeviceInfo_pci(devid)#from self.info['devices']
#check conflict before trigger hotplug event
@@ -611,35 +610,6 @@ class XendDomainInfo:
new_dev['vslt'])
self.image.signalDeviceModel('pci-ins', 'pci-inserted', bdf_str)
- # update the virtual pci slot
- vslt = xstransact.Read("/local/domain/0/device-model/%i/parameter"
- % self.getDomid())
- new_dev['vslt'] = vslt
-
- if dev_info is None:
- # create a new one from scrach
- dev_cfg_sxp = [dev_type,
- ['dev',
- ['domain', new_dev['domain']],
- ['bus', new_dev['bus']],
- ['slot', new_dev['slot']],
- ['func', new_dev['func']],
- ['vslt', new_dev['vslt']]
- ]]
- dev_uuid = self.info.device_add(dev_type, cfg_sxp = dev_cfg_sxp)
- dev_config_dict = self.info['devices'][dev_uuid][1]
- try:
- dev_config_dict['devid'] = devid = \
- self._createDevice(dev_type, dev_config_dict)
- self._waitForDevice(dev_type, devid)
- except VmError, ex:
- raise ex
- else:
- # update the pci config to add the new dev
- pci_devs.extend(new_devs)
- self._reconfigureDevice('pci', devid, pci_conf)
-
- return self.getDeviceController('pci').sxpr(devid)
def device_create(self, dev_config):
"""Create a new device.
@@ -649,11 +619,6 @@ class XendDomainInfo:
"""
log.debug("XendDomainInfo.device_create: %s" %
scrub_password(dev_config))
dev_type = sxp.name(dev_config)
-
- if dev_type == 'pci':
- rc = self.pci_device_create(dev_config)
- return rc
-
dev_uuid = self.info.device_add(dev_type, cfg_sxp = dev_config)
dev_config_dict = self.info['devices'][dev_uuid][1]
log.debug("XendDomainInfo.device_create: %s" %
scrub_password(dev_config_dict))
@@ -676,6 +641,151 @@ class XendDomainInfo:
xen.xend.XendDomain.instance().managed_config_save(self)
return self.getDeviceController(dev_type).sxpr(devid)
+ def pci_convert_sxp_to_dict(self, dev_sxp):
+ """Convert pci device sxp to dict
+ @param dev_sxp: device configuration
+ @type dev_sxp: SXP object (parsed config)
+ @return: dev_config
+ @rtype: dictionary
+ """
+ # In reconfigure phase, config of PCI device looks like below:
+ #
+ # sxp:
+ # [device, [pci, [dev, [domain, '0x0'], [bus, '0x0'], [slot, '0x0'],
+ # [func, '0x0'], [vslt, '0x0']],
+ # [state, 'Initialising']]]
+ #
+ # dict:
+ # {devs: [{domain: '0x0', bus: '0x0', slot: '0x0', func: '0x0',
+ # vslt: '0x0'}],
+ # states: ['Initialising']}
+ #
+ # state 'Initialising' means the device is being attached.
+ # state 'Closing' means the device is being detached.
+
+ dev_config = {}
+ pci_devs = []
+ for pci_dev in sxp.children(dev_sxp, 'dev'):
+ pci_dev_info = {}
+ for opt_val in pci_dev[1:]:
+ try:
+ opt, val = opt_val
+ pci_dev_info[opt] = val
+ except TypeError:
+ pass
+ pci_devs.append(pci_dev_info)
+ dev_config['devs'] = pci_devs
+ pci_states = []
+ for pci_state in sxp.children(dev_sxp, 'state'):
+ try:
+ pci_states.append(pci_state[1])
+ except IndexError:
+ raise XendError("Error reading state while parsing pci sxp")
+ dev_config['states'] = pci_states
+
+ return dev_config
+
+ def pci_device_configure(self, dev_sxp, devid = 0):
+ """Configure an existing pci device.
+
+ @param dev_sxp: device configuration
+ @type dev_sxp: SXP object (parsed config)
+ @param devid: device id
+ @type devid: int
+ @return: Returns True if successfully updated device
+ @rtype: boolean
+ """
+ log.debug("XendDomainInfo.pci_device_configure: %s"
+ % scrub_password(dev_sxp))
+
+ dev_class = sxp.name(dev_sxp)
+
+ if dev_class != 'pci':
+ return False
+
+ pci_state = sxp.child_value(dev_sxp, 'state')
+ existing_dev_info = self._getDeviceInfo_pci(devid)
+
+ if existing_dev_info is None and pci_state != 'Initialising':
+ raise XendError("Cannot detach when pci platform does not exist")
+
+ pci_dev = sxp.children(dev_sxp, 'dev')[0]
+ dev_config = self.pci_convert_sxp_to_dict(dev_sxp)
+ dev = dev_config['devs'][0]
+
+ # Do HVM specific processing
+ if self.info.is_hvm():
+ if pci_state == 'Initialising':
+ # HVM PCI device attachment
+ self.hvm_pci_device_create(dev_config)
+ # Update vslt
+ vslt =
xstransact.Read("/local/domain/0/device-model/%i/parameter"
+ % self.getDomid())
+ dev['vslt'] = vslt
+ for n in sxp.children(pci_dev):
+ if(n[0] == 'vslt'):
+ n[1] = vslt
+ else:
+ # HVM PCI device detachment
+ existing_dev_uuid = sxp.child_value(existing_dev_info, 'uuid')
+ existing_pci_conf = self.info['devices'][existing_dev_uuid][1]
+ existing_pci_devs = existing_pci_conf['devs']
+ vslt = '0x0'
+ for x in existing_pci_devs:
+ if ( int(x['domain'], 16) == int(dev['domain'], 16) and
+ int(x['bus'], 16) == int(dev['bus'], 16) and
+ int(x['slot'], 16) == int(dev['slot'], 16) and
+ int(x['func'], 16) == int(dev['func'], 16) ):
+ vslt = x['vslt']
+ break
+ if vslt == '0x0':
+ raise VmError("Device %04x:%02x:%02x.%02x is not connected"
+ % (int(dev['domain'],16), int(dev['bus'],16),
+ int(dev['slot'],16), int(dev['func'],16)))
+ self.hvm_destroyPCIDevice(int(vslt, 16))
+ # Update vslt
+ dev['vslt'] = vslt
+ for n in sxp.children(pci_dev):
+ if(n[0] == 'vslt'):
+ n[1] = vslt
+
+ # If pci platform does not exist, create and exit.
+ if existing_dev_info is None:
+ self.device_create(dev_sxp)
+ return True
+
+ # use DevController.reconfigureDevice to change device config
+ dev_control = self.getDeviceController(dev_class)
+ dev_uuid = dev_control.reconfigureDevice(devid, dev_config)
+ if not self.info.is_hvm():
+ # in PV case, wait until backend state becomes connected.
+ dev_control.waitForDevice_reconfigure(devid)
+ num_devs = dev_control.cleanupDevice(devid)
+
+ # update XendConfig with new device info
+ if dev_uuid:
+ new_dev_sxp = dev_control.configuration(devid)
+ self.info.device_update(dev_uuid, new_dev_sxp)
+
+ # If there is no device left, destroy pci and remove config.
+ if num_devs == 0:
+ if self.info.is_hvm():
+ self.destroyDevice('pci', devid, True)
+ del self.info['devices'][dev_uuid]
+ platform = self.info['platform']
+ orig_dev_num = len(platform['pci'])
+ # TODO: can use this to keep some info to ask high level
+ # management tools to hot insert a new passthrough dev
+ # after migration
+ if orig_dev_num != 0:
+ #platform['pci'] = ["%dDEVs" % orig_dev_num]
+ platform['pci'] = []
+ else:
+ self.destroyDevice('pci', devid)
+ del self.info['devices'][dev_uuid]
+
+ return True
+
def device_configure(self, dev_sxp, devid = None):
"""Configure an existing device.
@@ -690,6 +800,10 @@ class XendDomainInfo:
# convert device sxp to a dict
dev_class = sxp.name(dev_sxp)
dev_config = {}
+
+ if dev_class == 'pci':
+ return self.pci_device_configure(dev_sxp)
+
for opt_val in dev_sxp[1:]:
try:
dev_config[opt_val[0]] = opt_val[1]
@@ -714,11 +828,11 @@ class XendDomainInfo:
for devclass in XendDevices.valid_devices():
self.getDeviceController(devclass).waitForDevices()
- def destroyPCIDevice(self, vslot):
- log.debug("destroyPCIDevice called %s", vslot)
+ def hvm_destroyPCIDevice(self, vslot):
+ log.debug("hvm_destroyPCIDevice called %s", vslot)
if not self.info.is_hvm():
- raise VmError("only HVM guest support pci detach")
+ raise VmError("hvm_destroyPCIDevice called on non-HVM guest")
#all the PCI devs share one conf node
devid = '0'
@@ -744,34 +858,15 @@ class XendDomainInfo:
raise VmError("Device @ vslot 0x%x do not support hotplug." %
(vslot))
bdf_str = "%s:%s:%s.%s" % (x['domain'], x['bus'], x['slot'], x['func'])
- log.info("destroyPCIDevice:%s:%s!", x, bdf_str)
+ log.info("hvm_destroyPCIDevice:%s:%s!", x, bdf_str)
self.image.signalDeviceModel('pci-rem', 'pci-removed', bdf_str)
-
- if pci_len > 1:
- del pci_conf['devs'][devnum]
- self._reconfigureDevice('pci', devid, pci_conf)
- else:
- self.getDeviceController('pci').destroyDevice(devid, True)
- del self.info['devices'][dev_uuid]
- platform = self.info['platform']
- orig_dev_num = len(platform['pci'])
-
- #need remove the pci config
- #TODO:can use this to keep some info to ask high level management
tools to hot insert a new passthrough dev after migration
- if orig_dev_num != 0:
-# platform['pci'] = ["%dDEVs" % orig_dev_num]
- platform['pci'] = []
return 0
def destroyDevice(self, deviceClass, devid, force = False, rm_cfg = False):
log.debug("XendDomainInfo.destroyDevice: deviceClass = %s, device =
%s",
deviceClass, devid)
-
- if deviceClass == 'dpci':
- rc = self.destroyPCIDevice(devid)
- return rc
if rm_cfg:
# Convert devid to device number. A device number is
@@ -1967,36 +2062,44 @@ class XendDomainInfo:
for v in range(0, self.info['VCPUs_max']):
xc.vcpu_setaffinity(self.domid, v, self.info['cpus'])
else:
+ def find_relaxed_node(node_list):
+ import sys
+ if node_list is None:
+ node_list = range(0, info['nr_nodes'])
+ nodeload = [0]
+ nodeload = nodeload * info['nr_nodes']
+ from xen.xend import XendDomain
+ doms = XendDomain.instance().list('all')
+ for dom in doms:
+ cpuinfo = dom.getVCPUInfo()
+ for vcpu in sxp.children(cpuinfo, 'vcpu'):
+ def vinfo(n, t):
+ return t(sxp.child_value(vcpu, n))
+ cpumap = vinfo('cpumap', list)
+ for i in node_list:
+ node_cpumask = info['node_to_cpu'][i]
+ for j in node_cpumask:
+ if j in cpumap:
+ nodeload[i] += 1
+ break
+ for i in node_list:
+ if len(info['node_to_cpu'][i]) > 0:
+ nodeload[i] = int(nodeload[i] /
len(info['node_to_cpu'][i]))
+ else:
+ nodeload[i] = sys.maxint
+ index = nodeload.index( min(nodeload) )
+ return index
+
info = xc.physinfo()
if info['nr_nodes'] > 1:
node_memory_list = info['node_to_memory']
needmem =
self.image.getRequiredAvailableMemory(self.info['memory_dynamic_max']) / 1024
candidate_node_list = []
for i in range(0, info['nr_nodes']):
- if node_memory_list[i] >= needmem:
+ if node_memory_list[i] >= needmem and
len(info['node_to_cpu'][i]) > 0:
candidate_node_list.append(i)
- if candidate_node_list is None or len(candidate_node_list)
== 1:
- index = node_memory_list.index( max(node_memory_list) )
- cpumask = info['node_to_cpu'][index]
- else:
- nodeload = [0]
- nodeload = nodeload * info['nr_nodes']
- from xen.xend import XendDomain
- doms = XendDomain.instance().list('all')
- for dom in doms:
- cpuinfo = dom.getVCPUInfo()
- for vcpu in sxp.children(cpuinfo, 'vcpu'):
- def vinfo(n, t):
- return t(sxp.child_value(vcpu, n))
- cpumap = vinfo('cpumap', list)
- for i in candidate_node_list:
- node_cpumask = info['node_to_cpu'][i]
- for j in node_cpumask:
- if j in cpumap:
- nodeload[i] += 1
- break
- index = nodeload.index( min(nodeload) )
- cpumask = info['node_to_cpu'][index]
+ index = find_relaxed_node(candidate_node_list)
+ cpumask = info['node_to_cpu'][index]
for v in range(0, self.info['VCPUs_max']):
xc.vcpu_setaffinity(self.domid, v, cpumask)
@@ -2104,7 +2207,7 @@ class XendDomainInfo:
self.state_updated.acquire()
try:
while self._stateGet() in (DOM_STATE_RUNNING,DOM_STATE_PAUSED):
- self.state_updated.wait()
+ self.state_updated.wait(timeout=1.0)
finally:
self.state_updated.release()
diff -r 8c921adf4833 -r 42f6c206c951 tools/python/xen/xend/image.py
--- a/tools/python/xen/xend/image.py Fri Mar 14 15:07:45 2008 -0600
+++ b/tools/python/xen/xend/image.py Thu Mar 20 12:35:40 2008 -0600
@@ -90,6 +90,7 @@ class ImageHandler:
("image/kernel", self.kernel),
("image/cmdline", self.cmdline),
("image/ramdisk", self.ramdisk))
+ self.vm.permissionsVm("image/cmdline", { 'dom': self.vm.getDomid(),
'read': True } )
self.device_model = vmConfig['platform'].get('device_model')
@@ -201,6 +202,7 @@ class ImageHandler:
vnc_config = {}
has_vnc = int(vmConfig['platform'].get('vnc', 0)) != 0
has_sdl = int(vmConfig['platform'].get('sdl', 0)) != 0
+ opengl = 1
for dev_uuid in vmConfig['console_refs']:
dev_type, dev_info = vmConfig['devices'][dev_uuid]
if dev_type == 'vfb':
@@ -208,6 +210,7 @@ class ImageHandler:
if vfb_type == 'sdl':
self.display = dev_info.get('display', {})
self.xauthority = dev_info.get('xauthority', {})
+ opengl = int(dev_info.get('opengl', opengl))
has_sdl = True
else:
vnc_config = dev_info.get('other_config', {})
@@ -262,7 +265,8 @@ class ImageHandler:
elif has_sdl:
# SDL is default in QEMU.
- pass
+ if int(vmConfig['platform'].get('opengl', opengl)) != 1 :
+ ret.append('-disable-opengl')
else:
ret.append('-nographic')
@@ -580,7 +584,8 @@ class HVMImageHandler(ImageHandler):
ret.append("nic,vlan=%d,macaddr=%s,model=%s" %
(nics, mac, model))
ret.append("-net")
- ret.append("tap,vlan=%d,bridge=%s" % (nics, bridge))
+ ret.append("tap,vlan=%d,ifname=tap%d.%d,bridge=%s" %
+ (nics, self.vm.getDomid(), nics-1, bridge))
return ret
diff -r 8c921adf4833 -r 42f6c206c951
tools/python/xen/xend/server/BlktapController.py
--- a/tools/python/xen/xend/server/BlktapController.py Fri Mar 14 15:07:45
2008 -0600
+++ b/tools/python/xen/xend/server/BlktapController.py Thu Mar 20 12:35:40
2008 -0600
@@ -13,7 +13,9 @@ blktap_disk_types = [
'vmdk',
'ram',
'qcow',
- 'qcow2'
+ 'qcow2',
+
+ 'ioemu'
]
class BlktapController(BlkifController):
diff -r 8c921adf4833 -r 42f6c206c951
tools/python/xen/xend/server/DevController.py
--- a/tools/python/xen/xend/server/DevController.py Fri Mar 14 15:07:45
2008 -0600
+++ b/tools/python/xen/xend/server/DevController.py Thu Mar 20 12:35:40
2008 -0600
@@ -51,6 +51,8 @@ xenbusState = {
'Connected' : 4,
'Closing' : 5,
'Closed' : 6,
+ 'Reconfiguring': 7,
+ 'Reconfigured' : 8,
}
xoptions = XendOptions.instance()
@@ -88,6 +90,8 @@ class DevController:
(devid, back, front) = self.getDeviceDetails(config)
if devid is None:
return 0
+
+ self.setupDevice(config)
(backpath, frontpath) = self.addStoreEntries(config, devid, back,
front)
@@ -198,6 +202,15 @@ class DevController:
if status == Timeout:
raise VmError("Device %s (%s) could not be disconnected. " %
+ (devid, self.deviceClass))
+
+ def waitForDevice_reconfigure(self, devid):
+ log.debug("Waiting for %s - reconfigureDevice.", devid)
+
+ (status, err) = self.waitForBackend_reconfigure(devid)
+
+ if status == Timeout:
+ raise VmError("Device %s (%s) could not be reconfigured. " %
(devid, self.deviceClass))
@@ -325,6 +338,11 @@ class DevController:
"""
raise NotImplementedError()
+
+ def setupDevice(self, config):
+ """ Setup device from config.
+ """
+ return
def migrate(self, deviceConfig, network, dst, step, domName):
""" Migration of a device. The 'network' parameter indicates
@@ -569,6 +587,22 @@ class DevController:
return result['status']
+ def waitForBackend_reconfigure(self, devid):
+ frontpath = self.frontendPath(devid)
+ backpath = xstransact.Read(frontpath, "backend")
+ if backpath:
+ statusPath = backpath + '/' + "state"
+ ev = Event()
+ result = { 'status': Timeout }
+
+ xswatch(statusPath, xenbusStatusCallback, ev, result)
+
+ ev.wait(DEVICE_CREATE_TIMEOUT)
+
+ return (result['status'], None)
+ else:
+ return (Missing, None)
+
def backendPath(self, backdom, devid):
"""Construct backend path given the backend domain and device id.
@@ -634,3 +668,19 @@ def deviceDestroyCallback(statusPath, ev
ev.set()
return 0
+
+
+def xenbusStatusCallback(statusPath, ev, result):
+ log.debug("xenbusStatusCallback %s.", statusPath)
+
+ status = xstransact.Read(statusPath)
+
+ if status == str(xenbusState['Connected']):
+ result['status'] = Connected
+ else:
+ return 1
+
+ log.debug("xenbusStatusCallback %d.", result['status'])
+
+ ev.set()
+ return 0
diff -r 8c921adf4833 -r 42f6c206c951 tools/python/xen/xend/server/SrvDomain.py
--- a/tools/python/xen/xend/server/SrvDomain.py Fri Mar 14 15:07:45 2008 -0600
+++ b/tools/python/xen/xend/server/SrvDomain.py Thu Mar 20 12:35:40 2008 -0600
@@ -121,10 +121,10 @@ class SrvDomain(SrvDir):
def op_pincpu(self, _, req):
fn = FormFn(self.xd.domain_pincpu,
- [['dom', 'int'],
+ [['dom', 'str'],
['vcpu', 'str'],
['cpumap', 'str']])
- val = fn(req.args, {'dom': self.dom.domid})
+ val = fn(req.args, {'dom': self.dom.getName()})
return val
def op_cpu_sedf_get(self, _, req):
@@ -147,17 +147,17 @@ class SrvDomain(SrvDir):
def op_domain_sched_credit_get(self, _, req):
fn = FormFn(self.xd.domain_sched_credit_get,
- [['dom', 'int']])
- val = fn(req.args, {'dom': self.dom.domid})
+ [['dom', 'str']])
+ val = fn(req.args, {'dom': self.dom.getName()})
return val
def op_domain_sched_credit_set(self, _, req):
fn = FormFn(self.xd.domain_sched_credit_set,
- [['dom', 'int'],
+ [['dom', 'str'],
['weight', 'int'],
['cap', 'int']])
- val = fn(req.args, {'dom': self.dom.domid})
+ val = fn(req.args, {'dom': self.dom.getName()})
return val
def op_maxmem_set(self, _, req):
@@ -187,7 +187,9 @@ class SrvDomain(SrvDir):
def op_device_destroy(self, _, req):
return self.call(self.dom.destroyDevice,
[['type', 'str'],
- ['dev', 'str']],
+ ['dev', 'str'],
+ ['force', 'int'],
+ ['rm_cfg', 'int']],
req)
def op_device_configure(self, _, req):
diff -r 8c921adf4833 -r 42f6c206c951 tools/python/xen/xend/server/pciif.py
--- a/tools/python/xen/xend/server/pciif.py Fri Mar 14 15:07:45 2008 -0600
+++ b/tools/python/xen/xend/server/pciif.py Thu Mar 20 12:35:40 2008 -0600
@@ -24,7 +24,7 @@ from xen.xend.XendError import VmError
from xen.xend.XendError import VmError
from xen.xend.XendLogging import log
-from xen.xend.server.DevController import DevController
+from xen.xend.server.DevController import DevController, xenbusState
import xen.lowlevel.xc
@@ -44,6 +44,15 @@ while not (t&1):
t>>=1
PAGE_SHIFT+=1
+def parse_hex(val):
+ try:
+ if isinstance(val, types.StringTypes):
+ return int(val, 16)
+ else:
+ return val
+ except ValueError:
+ return None
+
class PciController(DevController):
def __init__(self, vm):
@@ -52,15 +61,6 @@ class PciController(DevController):
def getDeviceDetails(self, config):
"""@see DevController.getDeviceDetails"""
- def parse_hex(val):
- try:
- if isinstance(val, types.StringTypes):
- return int(val, 16)
- else:
- return val
- except ValueError:
- return None
-
back = {}
pcidevid = 0
vslots = ""
@@ -74,7 +74,6 @@ class PciController(DevController):
if vslt is not None:
vslots = vslots + vslt + ";"
- self.setupDevice(domain, bus, slot, func)
back['dev-%i' % pcidevid] = "%04x:%02x:%02x.%02x" % \
(domain, bus, slot, func)
pcidevid += 1
@@ -86,27 +85,80 @@ class PciController(DevController):
back['uuid'] = config.get('uuid','')
return (0, back, {})
+
def reconfigureDevice(self, _, config):
"""@see DevController.reconfigureDevice"""
- #currently only support config changes by hot insert/remove
pass-through dev
- #delete all the devices in xenstore
- (devid, new_back, new_front) = self.getDeviceDetails(config)
- num_devs = self.readBackend(devid, 'num_devs')
- for i in range(int(num_devs)):
- self.removeBackend(devid, 'dev-%d' % i)
- self.removeBackend(devid, 'num_devs')
-
- #create new devices config
- num_devs = new_back['num_devs']
- for i in range(int(num_devs)):
- dev_no = 'dev-%d' % i
- self.writeBackend(devid, dev_no, new_back[dev_no])
- self.writeBackend(devid, 'num_devs', num_devs)
-
- if new_back['vslots'] is not None:
- self.writeBackend(devid, 'vslots', new_back['vslots'])
-
- return new_back.get('uuid')
+ (devid, back, front) = self.getDeviceDetails(config)
+ num_devs = int(back['num_devs'])
+ states = config.get('states', [])
+
+ old_vslots = self.readBackend(devid, 'vslots')
+ if old_vslots is None:
+ old_vslots = ''
+ num_olddevs = int(self.readBackend(devid, 'num_devs'))
+
+ for i in range(num_devs):
+ try:
+ dev = back['dev-%i' % i]
+ state = states[i]
+ except:
+ raise XendError('Error reading config')
+
+ if state == 'Initialising':
+ # PCI device attachment
+ for j in range(num_olddevs):
+ if dev == self.readBackend(devid, 'dev-%i' % j):
+ raise XendError('Device %s is already connected.' %
dev)
+ log.debug('Attaching PCI device %s.' % dev)
+ (domain, bus, slotfunc) = dev.split(':')
+ (slot, func) = slotfunc.split('.')
+ domain = parse_hex(domain)
+ bus = parse_hex(bus)
+ slot = parse_hex(slot)
+ func = parse_hex(func)
+ self.setupOneDevice(domain, bus, slot, func)
+
+ self.writeBackend(devid, 'dev-%i' % (num_olddevs + i), dev)
+ self.writeBackend(devid, 'state-%i' % (num_olddevs + i),
+ str(xenbusState['Initialising']))
+ self.writeBackend(devid, 'num_devs', str(num_olddevs + i + 1))
+
+ # Update vslots
+ if back['vslots'] is not None:
+ vslots = old_vslots + back['vslots']
+ self.writeBackend(devid, 'vslots', vslots)
+
+ elif state == 'Closing':
+ # PCI device detachment
+ found = False
+ for j in range(num_olddevs):
+ if dev == self.readBackend(devid, 'dev-%i' % j):
+ found = True
+ log.debug('Detaching device %s' % dev)
+ self.writeBackend(devid, 'state-%i' % j,
+ str(xenbusState['Closing']))
+ if not found:
+ raise XendError('Device %s is not connected' % dev)
+
+ # Update vslots
+ if back['vslots'] is not None:
+ vslots = old_vslots
+ for vslt in back['vslots'].split(';'):
+ if vslt != '':
+ vslots = vslots.replace(vslt + ';', '', 1)
+ if vslots == '':
+ self.removeBackend(devid, 'vslots')
+ else:
+ self.writeBackend(devid, 'vslots', vslots)
+
+ else:
+ raise XendError('Error configuring device %s: invalid state %s'
+ % (dev,state))
+
+ self.writeBackend(devid, 'state', str(xenbusState['Reconfiguring']))
+
+ return self.readBackend(devid, 'uuid')
+
def getDeviceConfiguration(self, devid, transaction = None):
result = DevController.getDeviceConfiguration(self, devid, transaction)
@@ -125,7 +177,7 @@ class PciController(DevController):
pci_match = re.match(r"((?P<domain>[0-9a-fA-F]{1,4})[:,])?" +
r"(?P<bus>[0-9a-fA-F]{1,2})[:,]" +
r"(?P<slot>[0-9a-fA-F]{1,2})[.,]" +
- r"(?P<func>[0-9a-fA-F]{1,2})", dev_config)
+ r"(?P<func>[0-7]{1,2})$", dev_config)
if pci_match!=None:
pci_dev_info = pci_match.groupdict()
@@ -136,7 +188,10 @@ class PciController(DevController):
#append vslot info
if vslots is not None:
- dev_dict['vslt'] = slot_list[i]
+ try:
+ dev_dict['vslt'] = slot_list[i]
+ except IndexError:
+ dev_dict['vslt'] = '0x0'
pci_devs.append(dev_dict)
@@ -171,7 +226,7 @@ class PciController(DevController):
return sxpr
- def setupDevice(self, domain, bus, slot, func):
+ def setupOneDevice(self, domain, bus, slot, func):
""" Attach I/O resources for device to frontend domain
"""
fe_domid = self.getDomid()
@@ -225,6 +280,116 @@ class PciController(DevController):
raise VmError(('pci: failed to configure irq on device '+
'%s - errno=%d')%(dev.name,rc))
+ def setupDevice(self, config):
+ """Setup devices from config
+ """
+ for pci_config in config.get('devs', []):
+ domain = parse_hex(pci_config.get('domain', 0))
+ bus = parse_hex(pci_config.get('bus', 0))
+ slot = parse_hex(pci_config.get('slot', 0))
+ func = parse_hex(pci_config.get('func', 0))
+ self.setupOneDevice(domain, bus, slot, func)
+
+ return
+
+ def cleanupOneDevice(self, domain, bus, slot, func):
+ """ Detach I/O resources for device from frontend domain
+ """
+ fe_domid = self.getDomid()
+
+ try:
+ dev = PciDevice(domain, bus, slot, func)
+ except Exception, e:
+ raise VmError("pci: failed to locate device and "+
+ "parse it's resources - "+str(e))
+
+ if dev.driver!='pciback':
+ raise VmError(("pci: PCI Backend does not own device "+ \
+ "%s\n"+ \
+ "See the pciback.hide kernel "+ \
+ "command-line parameter or\n"+ \
+ "bind your slot/device to the PCI backend using sysfs" \
+ )%(dev.name))
+
+ for (start, size) in dev.ioports:
+ log.debug('pci: disabling ioport 0x%x/0x%x'%(start,size))
+ rc = xc.domain_ioport_permission(domid = fe_domid, first_port =
start,
+ nr_ports = size, allow_access = False)
+ if rc<0:
+ raise VmError(('pci: failed to configure I/O ports on device '+
+ '%s - errno=%d')%(dev.name,rc))
+
+ for (start, size) in dev.iomem:
+ # Convert start/size from bytes to page frame sizes
+ start_pfn = start>>PAGE_SHIFT
+ # Round number of pages up to nearest page boundary (if not on one)
+ nr_pfns = (size+(PAGE_SIZE-1))>>PAGE_SHIFT
+
+ log.debug('pci: disabling iomem 0x%x/0x%x pfn 0x%x/0x%x'% \
+ (start,size,start_pfn,nr_pfns))
+ rc = xc.domain_iomem_permission(domid = fe_domid,
+ first_pfn = start_pfn,
+ nr_pfns = nr_pfns,
+ allow_access = False)
+ if rc<0:
+ raise VmError(('pci: failed to configure I/O memory on device
'+
+ '%s - errno=%d')%(dev.name,rc))
+
+ if dev.irq>0:
+ log.debug('pci: disabling irq %d'%dev.irq)
+ rc = xc.domain_irq_permission(domid = fe_domid, pirq = dev.irq,
+ allow_access = False)
+ if rc<0:
+ raise VmError(('pci: failed to configure irq on device '+
+ '%s - errno=%d')%(dev.name,rc))
+
+ def cleanupDevice(self, devid):
+ """ Detach I/O resources for device and cleanup xenstore nodes
+ after reconfigure.
+
+ @param devid: The device ID
+ @type devid: int
+ @return: Return the number of devices connected
+ @rtype: int
+ """
+ num_devs = int(self.readBackend(devid, 'num_devs'))
+ new_num_devs = 0
+ for i in range(num_devs):
+ state = int(self.readBackend(devid, 'state-%i' % i))
+ if state == xenbusState['Closing']:
+ # Detach I/O resources.
+ dev = self.readBackend(devid, 'dev-%i' % i)
+ (domain, bus, slotfunc) = dev.split(':')
+ (slot, func) = slotfunc.split('.')
+ domain = parse_hex(domain)
+ bus = parse_hex(bus)
+ slot = parse_hex(slot)
+ func = parse_hex(func)
+ # In HVM case, I/O resources are disabled in ioemu.
+ self.cleanupOneDevice(domain, bus, slot, func)
+ # Remove xenstore nodes.
+ self.removeBackend(devid, 'dev-%i' % i)
+ self.removeBackend(devid, 'vdev-%i' % i)
+ self.removeBackend(devid, 'state-%i' % i)
+ else:
+ if new_num_devs != i:
+ tmpdev = self.readBackend(devid, 'dev-%i' % i)
+ self.writeBackend(devid, 'dev-%i' % new_num_devs, tmpdev)
+ self.removeBackend(devid, 'dev-%i' % i)
+ tmpvdev = self.readBackend(devid, 'vdev-%i' % i)
+ if tmpvdev is not None:
+ self.writeBackend(devid, 'vdev-%i' % new_num_devs,
+ tmpvdev)
+ self.removeBackend(devid, 'vdev-%i' % i)
+ tmpstate = self.readBackend(devid, 'state-%i' % i)
+ self.writeBackend(devid, 'state-%i' % new_num_devs,
tmpstate)
+ self.removeBackend(devid, 'state-%i' % i)
+ new_num_devs = new_num_devs + 1
+
+ self.writeBackend(devid, 'num_devs', str(new_num_devs))
+
+ return new_num_devs
+
def waitForBackend(self,devid):
return (0, "ok - no hotplug")
diff -r 8c921adf4833 -r 42f6c206c951 tools/python/xen/xend/server/vfbif.py
--- a/tools/python/xen/xend/server/vfbif.py Fri Mar 14 15:07:45 2008 -0600
+++ b/tools/python/xen/xend/server/vfbif.py Thu Mar 20 12:35:40 2008 -0600
@@ -7,7 +7,7 @@ import os
CONFIG_ENTRIES = ['type', 'vncdisplay', 'vnclisten', 'vncpasswd', 'vncunused',
'display', 'xauthority', 'keymap',
- 'uuid', 'location', 'protocol']
+ 'uuid', 'location', 'protocol', 'opengl']
class VfbifController(DevController):
"""Virtual frame buffer controller. Handles all vfb devices for a domain.
diff -r 8c921adf4833 -r 42f6c206c951 tools/python/xen/xm/create.py
--- a/tools/python/xen/xm/create.py Fri Mar 14 15:07:45 2008 -0600
+++ b/tools/python/xen/xm/create.py Thu Mar 20 12:35:40 2008 -0600
@@ -304,7 +304,7 @@ gopts.var('pci', val='BUS:DEV.FUNC',
gopts.var('pci', val='BUS:DEV.FUNC',
fn=append_value, default=[],
use="""Add a PCI device to a domain, using given params (in hex).
- For example 'pci=c0:02.1a'.
+ For example 'pci=c0:02.1'.
The option may be repeated to add more than one pci device.""")
gopts.var('ioports', val='FROM[-TO]',
@@ -319,7 +319,7 @@ gopts.var('irq', val='IRQ',
For example 'irq=7'.
This option may be repeated to add more than one IRQ.""")
-gopts.var('vfb',
val="type={vnc,sdl},vncunused=1,vncdisplay=N,vnclisten=ADDR,display=DISPLAY,xauthority=XAUTHORITY,vncpasswd=PASSWORD",
+gopts.var('vfb',
val="type={vnc,sdl},vncunused=1,vncdisplay=N,vnclisten=ADDR,display=DISPLAY,xauthority=XAUTHORITY,vncpasswd=PASSWORD,opengl=1",
fn=append_value, default=[],
use="""Make the domain a framebuffer backend.
The backend type should be either sdl or vnc.
@@ -330,7 +330,7 @@ gopts.var('vfb', val="type={vnc,sdl},vnc
default password.
For type=sdl, a viewer will be started automatically using the
given DISPLAY and XAUTHORITY, which default to the current user's
- ones.""")
+ ones. OpenGL will be used by default unless opengl is set to 0.""")
gopts.var('vif',
val="type=TYPE,mac=MAC,bridge=BRIDGE,ip=IPADDR,script=SCRIPT," + \
"backend=DOM,vifname=NAME,rate=RATE,model=MODEL,accel=ACCEL",
@@ -504,6 +504,10 @@ gopts.var('sdl', val='',
fn=set_value, default=None,
use="""Should the device model use SDL?""")
+gopts.var('opengl', val='',
+ fn=set_value, default=None,
+ use="""Enable\Disable OpenGL""")
+
gopts.var('display', val='DISPLAY',
fn=set_value, default=None,
use="X11 display to use")
@@ -641,7 +645,7 @@ def configure_vfbs(config_devs, vals):
d['type'] = 'sdl'
for (k,v) in d.iteritems():
if not k in [ 'vnclisten', 'vncunused', 'vncdisplay', 'display',
- 'xauthority', 'type', 'vncpasswd' ]:
+ 'xauthority', 'type', 'vncpasswd', 'opengl' ]:
err("configuration option %s unknown to vfbs" % k)
config.append([k,v])
if not d.has_key("keymap"):
@@ -745,7 +749,7 @@ def configure_hvm(config_image, vals):
'vnc', 'vncdisplay', 'vncunused', 'vncconsole', 'vnclisten',
'sdl', 'display', 'xauthority', 'rtc_timeoffset', 'monitor',
'acpi', 'apic', 'usb', 'usbdevice', 'keymap', 'pci', 'hpet',
- 'guest_os_type', 'hap']
+ 'guest_os_type', 'hap', 'opengl']
for a in args:
if a in vals.__dict__ and vals.__dict__[a] is not None:
@@ -840,7 +844,7 @@ def preprocess_pci(vals):
pci_match = re.match(r"((?P<domain>[0-9a-fA-F]{1,4})[:,])?" + \
r"(?P<bus>[0-9a-fA-F]{1,2})[:,]" + \
r"(?P<slot>[0-9a-fA-F]{1,2})[.,]" + \
- r"(?P<func>[0-9a-fA-F])", pci_dev_str)
+ r"(?P<func>[0-7])$", pci_dev_str)
if pci_match!=None:
pci_dev_info = pci_match.groupdict('0')
try:
diff -r 8c921adf4833 -r 42f6c206c951 tools/python/xen/xm/main.py
--- a/tools/python/xen/xm/main.py Fri Mar 14 15:07:45 2008 -0600
+++ b/tools/python/xen/xm/main.py Thu Mar 20 12:35:40 2008 -0600
@@ -175,11 +175,11 @@ SUBCOMMAND_HELP = {
'vnet-delete' : ('<VnetId>', 'Delete a Vnet.'),
'vnet-list' : ('[-l|--long]', 'List Vnets.'),
'vtpm-list' : ('<Domain> [--long]', 'List virtual TPM devices.'),
- 'pci-attach ' : ('<Domain> <dom> <bus> <slot> <func> [virtual slot]',
+ 'pci-attach' : ('<Domain> <domain:bus:slot.func> [virtual slot]',
'Insert a new pass-through pci device.'),
- 'pci-detach ' : ('<Domain> <virtual slot>',
+ 'pci-detach' : ('<Domain> <domain:bus:slot.func>',
'Remove a domain\'s pass-through pci device.'),
- 'pci-list' : ('<Domain>',
+ 'pci-list' : ('<Domain>',
'List pass-through pci devices for a domain.'),
# security
@@ -626,8 +626,11 @@ class Shell(cmd.Cmd):
def preloop(self):
cmd.Cmd.preloop(self)
- import readline
- readline.set_completer_delims(' ')
+ try:
+ import readline
+ readline.set_completer_delims(' ')
+ except ImportError:
+ pass
def default(self, line):
words = shlex.split(line)
@@ -2229,29 +2232,37 @@ def xm_network_attach(args):
vif.append(vif_param)
server.xend.domain.device_create(dom, vif)
-def parse_pci_configuration(args):
+def parse_pci_configuration(args, state):
dom = args[0]
-
- if len(args) == 6:
- vslt = args[5]
+ pci_dev_str = args[1]
+ if len(args) == 3:
+ vslt = args[2]
else:
vslt = '0x0' #chose a free virtual PCI slot
-
- pci = ['pci',
- ['devs',
- [{'domain': "0x%x" % int(args[1], 16),
- 'bus': "0x%x" % int(args[2], 16),
- 'slot': "0x%x" % int(args[3], 16),
- 'func': "0x%x" % int(args[4], 16),
- 'vslt': "0x%x" % int(vslt, 16)}]
- ]]
+ pci=['pci']
+ pci_match = re.match(r"((?P<domain>[0-9a-fA-F]{1,4})[:,])?" + \
+ r"(?P<bus>[0-9a-fA-F]{1,2})[:,]" + \
+ r"(?P<slot>[0-9a-fA-F]{1,2})[.,]" + \
+ r"(?P<func>[0-7])$", pci_dev_str)
+ if pci_match == None:
+ raise OptionError("Invalid argument: %s %s" % (pci_dev_str,vslt))
+ pci_dev_info = pci_match.groupdict('0')
+ try:
+ pci.append(['dev', ['domain', '0x'+ pci_dev_info['domain']], \
+ ['bus', '0x'+ pci_dev_info['bus']],
+ ['slot', '0x'+ pci_dev_info['slot']],
+ ['func', '0x'+ pci_dev_info['func']],
+ ['vslt', '0x%x' % int(vslt, 16)]])
+ except:
+ raise OptionError("Invalid argument: %s %s" % (pci_dev_str,vslt))
+ pci.append(['state', state])
return (dom, pci)
def xm_pci_attach(args):
- arg_check(args, 'pci-attach', 5, 6)
- (dom, pci) = parse_pci_configuration(args)
- server.xend.domain.device_create(dom, pci)
+ arg_check(args, 'pci-attach', 2, 3)
+ (dom, pci) = parse_pci_configuration(args, 'Initialising')
+ server.xend.domain.device_configure(dom, pci)
def detach(args, deviceClass):
rm_cfg = True
@@ -2316,12 +2327,11 @@ def xm_network_detach(args):
arg_check(args, 'network-detach', 2, 3)
detach(args, 'vif')
-
def xm_pci_detach(args):
arg_check(args, 'pci-detach', 2)
- dom = args[0]
- dev = args[1]
- server.xend.domain.destroyDevice(dom, 'dpci', dev)
+ (dom, pci) = parse_pci_configuration(args, 'Closing')
+ server.xend.domain.device_configure(dom, pci)
+
def xm_vnet_list(args):
xenapi_unsupported()
diff -r 8c921adf4833 -r 42f6c206c951 tools/python/xen/xm/xenapi_create.py
--- a/tools/python/xen/xm/xenapi_create.py Fri Mar 14 15:07:45 2008 -0600
+++ b/tools/python/xen/xm/xenapi_create.py Thu Mar 20 12:35:40 2008 -0600
@@ -810,7 +810,7 @@ class sxp2xml:
get_child_by_name(image, "vxauthority", "127.0.0.1"),
document))
console.appendChild(self.mk_other_config(
- "vncpasswd", get_child_by_name(image, "vncpasswd", ""),
+ "opengl", get_child_by_name(image, "opengl", "1"),
document))
consoles.append(console)
diff -r 8c921adf4833 -r 42f6c206c951 tools/xcutils/xc_save.c
--- a/tools/xcutils/xc_save.c Fri Mar 14 15:07:45 2008 -0600
+++ b/tools/xcutils/xc_save.c Thu Mar 20 12:35:40 2008 -0600
@@ -17,6 +17,7 @@
#include <sys/types.h>
#include <sys/stat.h>
#include <fcntl.h>
+#include <err.h>
#include <xs.h>
#include <xenctrl.h>
diff -r 8c921adf4833 -r 42f6c206c951 tools/xenstat/xentop/xentop.c
--- a/tools/xenstat/xentop/xentop.c Fri Mar 14 15:07:45 2008 -0600
+++ b/tools/xenstat/xentop/xentop.c Thu Mar 20 12:35:40 2008 -0600
@@ -1121,6 +1121,7 @@ int main(int argc, char **argv)
do {
gettimeofday(&curtime, NULL);
top();
+ fflush(stdout);
oldtime = curtime;
if ((!loop) && !(--iterations))
break;
diff -r 8c921adf4833 -r 42f6c206c951 tools/xenstore/Makefile
--- a/tools/xenstore/Makefile Fri Mar 14 15:07:45 2008 -0600
+++ b/tools/xenstore/Makefile Thu Mar 20 12:35:40 2008 -0600
@@ -1,21 +1,16 @@ XEN_ROOT=../..
XEN_ROOT=../..
include $(XEN_ROOT)/tools/Rules.mk
-XEN_LIBXC = $(XEN_ROOT)/tools/libxc
MAJOR = 3.0
MINOR = 0
-PROFILE=#-pg
-BASECFLAGS=-Werror
+CFLAGS += -Werror
+CFLAGS += -I.
+CFLAGS += $(CFLAGS_libxenctrl)
+
# Make gcc generate dependencies.
-BASECFLAGS += -Wp,-MD,.$(@F).d
-PROG_DEP = .*.d
-BASECFLAGS+= $(PROFILE)
-BASECFLAGS+= $(CFLAGS_libxenctrl)
-BASECFLAGS+= -I.
-
-CFLAGS += $(BASECFLAGS)
-LDFLAGS += $(PROFILE)
+CFLAGS += -Wp,-MD,.$(@F).d
+DEP = .*.d
CLIENTS := xenstore-exists xenstore-list xenstore-read xenstore-rm
xenstore-chmod
CLIENTS += xenstore-write
@@ -45,22 +40,22 @@ endif
endif
xenstored: $(XENSTORED_OBJS)
- $(CC) $(CFLAGS) $(LDFLAGS) $^ $(LOADLIBES) $(LDLIBS)
$(LDFLAGS_libxenctrl) $(SOCKET_LIBS) -o $@
+ $(CC) $(CFLAGS) $(LDFLAGS) $^ $(LDFLAGS_libxenctrl) $(SOCKET_LIBS) -o $@
$(CLIENTS): xenstore-%: xenstore_%.o libxenstore.so
- $(CC) $(CFLAGS) $(LDFLAGS) $< $(LOADLIBES) $(LDLIBS) -L. -lxenstore
$(SOCKET_LIBS) -o $@
+ $(CC) $(CFLAGS) $(LDFLAGS) $< -L. -lxenstore $(SOCKET_LIBS) -o $@
$(CLIENTS_OBJS): xenstore_%.o: xenstore_client.c
$(COMPILE.c) -DCLIENT_$(*F) -o $@ $<
xenstore-control: xenstore_control.o libxenstore.so
- $(CC) $(CFLAGS) $(LDFLAGS) $< $(LOADLIBES) $(LDLIBS) -L. -lxenstore
$(SOCKET_LIBS) -o $@
+ $(CC) $(CFLAGS) $(LDFLAGS) $< -L. -lxenstore $(SOCKET_LIBS) -o $@
xenstore-ls: xsls.o libxenstore.so
- $(CC) $(CFLAGS) $(LDFLAGS) $< $(LOADLIBES) $(LDLIBS) -L. -lxenstore
$(SOCKET_LIBS) -o $@
+ $(CC) $(CFLAGS) $(LDFLAGS) $< -L. -lxenstore $(SOCKET_LIBS) -o $@
xs_tdb_dump: xs_tdb_dump.o utils.o tdb.o talloc.o
- $(CC) $(CFLAGS) $(LDFLAGS) $^ $(LOADLIBES) $(LDLIBS) -o $@
+ $(CC) $(CFLAGS) $(LDFLAGS) $^ -o $@
libxenstore.so: libxenstore.so.$(MAJOR)
ln -sf $< $@
@@ -71,7 +66,7 @@ libxenstore.so.$(MAJOR).$(MINOR): xs.opi
$(CC) $(CFLAGS) $(LDFLAGS) -Wl,$(SONAME_LDFLAG)
-Wl,libxenstore.so.$(MAJOR) $(SHLIB_CFLAGS) -o $@ $^ $(SOCKET_LIBS) -lpthread
libxenstore.a: xs.o xs_lib.o
- $(AR) rcs libxenstore.a $^
+ $(AR) rcs $@ $^
.PHONY: clean
clean:
@@ -79,7 +74,7 @@ clean:
rm -f xenstored xs_random xs_stress xs_crashme
rm -f xs_tdb_dump xenstore-control xenstore-ls
rm -f $(CLIENTS)
- $(RM) $(PROG_DEP)
+ $(RM) $(DEP)
.PHONY: TAGS
TAGS:
@@ -108,7 +103,7 @@ install: all
$(INSTALL_DATA) xs.h $(DESTDIR)$(INCLUDEDIR)
$(INSTALL_DATA) xs_lib.h $(DESTDIR)$(INCLUDEDIR)
--include $(PROG_DEP)
+-include $(DEP)
# never delete any intermediate files.
.SECONDARY:
diff -r 8c921adf4833 -r 42f6c206c951 xen/arch/ia64/linux-xen/setup.c
--- a/xen/arch/ia64/linux-xen/setup.c Fri Mar 14 15:07:45 2008 -0600
+++ b/xen/arch/ia64/linux-xen/setup.c Thu Mar 20 12:35:40 2008 -0600
@@ -358,7 +358,7 @@ acpi_oem_console_setup(void)
extern struct ns16550_defaults ns16550_com1;
efi_system_table_t *systab;
efi_config_table_t *tables;
- struct acpi20_table_rsdp *rsdp = NULL;
+ struct acpi_table_rsdp *rsdp = NULL;
struct acpi_table_xsdt *xsdt;
struct acpi_table_header *hdr;
int i;
@@ -378,16 +378,17 @@ acpi_oem_console_setup(void)
for (i = 0 ; i < (int)systab->nr_tables && !rsdp ; i++) {
if (efi_guidcmp(tables[i].guid, ACPI_20_TABLE_GUID) == 0)
rsdp =
- (struct acpi20_table_rsdp *)__va(tables[i].table);
- }
-
- if (!rsdp || strncmp(rsdp->signature, RSDP_SIG, sizeof(RSDP_SIG) - 1))
+ (struct acpi_table_rsdp *)__va(tables[i].table);
+ }
+
+ if (!rsdp ||
+ strncmp(rsdp->signature, ACPI_SIG_RSDP, sizeof(ACPI_SIG_RSDP) - 1))
return -ENODEV;
- xsdt = (struct acpi_table_xsdt *)__va(rsdp->xsdt_address);
+ xsdt = (struct acpi_table_xsdt *)__va(rsdp->xsdt_physical_address);
hdr = &xsdt->header;
- if (strncmp(hdr->signature, XSDT_SIG, sizeof(XSDT_SIG) - 1))
+ if (strncmp(hdr->signature, ACPI_SIG_XSDT, sizeof(ACPI_SIG_XSDT) - 1))
return -ENODEV;
/* Looking for Fujitsu PRIMEQUEST systems */
diff -r 8c921adf4833 -r 42f6c206c951 xen/arch/ia64/xen/acpi.c
--- a/xen/arch/ia64/xen/acpi.c Fri Mar 14 15:07:45 2008 -0600
+++ b/xen/arch/ia64/xen/acpi.c Thu Mar 20 12:35:40 2008 -0600
@@ -74,7 +74,7 @@ acpi_get_sysname (void)
{
/* #ifdef CONFIG_IA64_GENERIC */
unsigned long rsdp_phys;
- struct acpi20_table_rsdp *rsdp;
+ struct acpi_table_rsdp *rsdp;
struct acpi_table_xsdt *xsdt;
struct acpi_table_header *hdr;
@@ -84,15 +84,15 @@ acpi_get_sysname (void)
return "dig";
}
- rsdp = (struct acpi20_table_rsdp *) __va(rsdp_phys);
- if (strncmp(rsdp->signature, RSDP_SIG, sizeof(RSDP_SIG) - 1)) {
+ rsdp = (struct acpi_table_rsdp *) __va(rsdp_phys);
+ if (strncmp(rsdp->signature, ACPI_SIG_RSDP, sizeof(ACPI_SIG_RSDP) - 1))
{
printk(KERN_ERR "ACPI 2.0 RSDP signature incorrect, default to
\"dig\"\n");
return "dig";
}
- xsdt = (struct acpi_table_xsdt *) __va(rsdp->xsdt_address);
+ xsdt = (struct acpi_table_xsdt *) __va(rsdp->xsdt_physical_address);
hdr = &xsdt->header;
- if (strncmp(hdr->signature, XSDT_SIG, sizeof(XSDT_SIG) - 1)) {
+ if (strncmp(hdr->signature, ACPI_SIG_XSDT, sizeof(ACPI_SIG_XSDT) - 1)) {
printk(KERN_ERR "ACPI 2.0 XSDT signature incorrect, default to
\"dig\"\n");
return "dig";
}
@@ -356,14 +356,14 @@ acpi_parse_madt (unsigned long phys_addr
#ifdef CONFIG_ITANIUM
has_8259 = 1; /* Firmware on old Itanium systems is broken */
#else
- has_8259 = acpi_madt->flags.pcat_compat;
+ has_8259 = acpi_madt->flags & ACPI_MADT_PCAT_COMPAT;
#endif
iosapic_system_init(has_8259);
/* Get base address of IPI Message Block */
- if (acpi_madt->lapic_address)
- ipi_base_addr = (void __iomem *)
ioremap(acpi_madt->lapic_address, 0);
+ if (acpi_madt->address)
+ ipi_base_addr = (void __iomem *)ioremap(acpi_madt->address, 0);
printk(KERN_INFO PREFIX "Local APIC address %p\n", ipi_base_addr);
@@ -416,7 +416,7 @@ acpi_numa_slit_init (struct acpi_table_s
u32 len;
len = sizeof(struct acpi_table_header) + 8
- + slit->localities * slit->localities;
+ + slit->locality_count * slit->locality_count;
if (slit->header.length != len) {
printk(KERN_ERR "ACPI 2.0 SLIT: size mismatch: %d expected, %d
actual\n",
len, slit->header.length);
@@ -519,21 +519,24 @@ acpi_numa_arch_fixup (void)
for (i = 0; i < srat_num_cpus; i++)
node_cpuid[i].nid = pxm_to_nid_map[node_cpuid[i].nid];
- printk(KERN_INFO "Number of logical nodes in system = %d\n", numnodes);
- printk(KERN_INFO "Number of memory chunks in system = %d\n",
num_node_memblks);
-
- if (!slit_table) return;
+ printk(KERN_INFO "Number of logical nodes in system = %d\n",
+ numnodes);
+ printk(KERN_INFO "Number of memory chunks in system = %d\n",
+ num_node_memblks);
+
+ if (!slit_table)
+ return;
memset(numa_slit, -1, sizeof(numa_slit));
- for (i=0; i<slit_table->localities; i++) {
+ for (i = 0; i < slit_table->locality_count; i++) {
if (!pxm_bit_test(i))
continue;
node_from = pxm_to_nid_map[i];
- for (j=0; j<slit_table->localities; j++) {
+ for (j=0; j < slit_table->locality_count; j++) {
if (!pxm_bit_test(j))
continue;
node_to = pxm_to_nid_map[j];
node_distance(node_from, node_to) =
- slit_table->entry[i*slit_table->localities + j];
+ slit_table->entry[i * slit_table->locality_count +
j];
}
}
@@ -560,7 +563,7 @@ acpi_parse_fadt (unsigned long phys_addr
acpi_parse_fadt (unsigned long phys_addr, unsigned long size)
{
struct acpi_table_header *fadt_header;
- struct fadt_descriptor_rev2 *fadt;
+ struct acpi_table_fadt *fadt;
if (!phys_addr || !size)
return -EINVAL;
@@ -569,16 +572,16 @@ acpi_parse_fadt (unsigned long phys_addr
if (fadt_header->revision != 3)
return -ENODEV; /* Only deal with ACPI 2.0 FADT */
- fadt = (struct fadt_descriptor_rev2 *) fadt_header;
-
- if (!(fadt->iapc_boot_arch & BAF_8042_KEYBOARD_CONTROLLER))
+ fadt = (struct acpi_table_fadt *) fadt_header;
+
+ if (!(fadt->boot_flags & BAF_8042_KEYBOARD_CONTROLLER))
acpi_kbd_controller_present = 0;
- if (fadt->iapc_boot_arch & BAF_LEGACY_DEVICES)
+ if (fadt->boot_flags & BAF_LEGACY_DEVICES)
acpi_legacy_devices = 1;
#if 0
- acpi_register_gsi(fadt->sci_int, ACPI_ACTIVE_LOW, ACPI_LEVEL_SENSITIVE);
+ acpi_register_gsi(fadt->sci_interrupt, ACPI_ACTIVE_LOW,
ACPI_LEVEL_SENSITIVE);
#endif
return 0;
}
diff -r 8c921adf4833 -r 42f6c206c951 xen/arch/ia64/xen/dom_fw_common.c
--- a/xen/arch/ia64/xen/dom_fw_common.c Fri Mar 14 15:07:45 2008 -0600
+++ b/xen/arch/ia64/xen/dom_fw_common.c Thu Mar 20 12:35:40 2008 -0600
@@ -208,18 +208,18 @@ print_md(efi_memory_desc_t *md)
}
struct fake_acpi_tables {
- struct acpi20_table_rsdp rsdp;
- struct xsdt_descriptor_rev2 xsdt;
+ struct acpi_table_rsdp rsdp;
+ struct acpi_table_xsdt xsdt;
uint64_t madt_ptr;
- struct fadt_descriptor_rev2 fadt;
- struct facs_descriptor_rev2 facs;
+ struct acpi_table_fadt fadt;
+ struct acpi_table_facs facs;
struct acpi_table_header dsdt;
uint8_t aml[8 + 11 * MAX_VIRT_CPUS];
struct acpi_table_madt madt;
struct acpi_table_lsapic lsapic[MAX_VIRT_CPUS];
- uint8_t pm1a_evt_blk[4];
- uint8_t pm1a_cnt_blk[1];
- uint8_t pm_tmr_blk[4];
+ uint8_t pm1a_event_block[4];
+ uint8_t pm1a_control_block[1];
+ uint8_t pm_timer_block[4];
};
#define ACPI_TABLE_MPA(field) \
FW_ACPI_BASE_PADDR + offsetof(struct fake_acpi_tables, field);
@@ -228,10 +228,10 @@ void
void
dom_fw_fake_acpi(domain_t *d, struct fake_acpi_tables *tables)
{
- struct acpi20_table_rsdp *rsdp = &tables->rsdp;
- struct xsdt_descriptor_rev2 *xsdt = &tables->xsdt;
- struct fadt_descriptor_rev2 *fadt = &tables->fadt;
- struct facs_descriptor_rev2 *facs = &tables->facs;
+ struct acpi_table_rsdp *rsdp = &tables->rsdp;
+ struct acpi_table_xsdt *xsdt = &tables->xsdt;
+ struct acpi_table_fadt *fadt = &tables->fadt;
+ struct acpi_table_facs *facs = &tables->facs;
struct acpi_table_header *dsdt = &tables->dsdt;
struct acpi_table_madt *madt = &tables->madt;
struct acpi_table_lsapic *lsapic = tables->lsapic;
@@ -245,34 +245,37 @@ dom_fw_fake_acpi(domain_t *d, struct fak
memset(tables, 0, sizeof(struct fake_acpi_tables));
/* setup XSDT (64bit version of RSDT) */
- memcpy(xsdt->signature, XSDT_SIG, sizeof(xsdt->signature));
+ memcpy(xsdt->header.signature, ACPI_SIG_XSDT,
+ sizeof(xsdt->header.signature));
/* XSDT points to both the FADT and the MADT, so add one entry */
- xsdt->length = sizeof(struct xsdt_descriptor_rev2) + sizeof(uint64_t);
- xsdt->revision = 1;
- memcpy(xsdt->oem_id, "XEN", 3);
- memcpy(xsdt->oem_table_id, "Xen/ia64", 8);
- memcpy(xsdt->asl_compiler_id, "XEN", 3);
- xsdt->asl_compiler_revision = xen_ia64_version(d);
+ xsdt->header.length = sizeof(struct acpi_table_xsdt) + sizeof(uint64_t);
+ xsdt->header.revision = 1;
+ memcpy(xsdt->header.oem_id, "XEN", 3);
+ memcpy(xsdt->header.oem_table_id, "Xen/ia64", 8);
+ memcpy(xsdt->header.asl_compiler_id, "XEN", 3);
+ xsdt->header.asl_compiler_revision = xen_ia64_version(d);
xsdt->table_offset_entry[0] = ACPI_TABLE_MPA(fadt);
tables->madt_ptr = ACPI_TABLE_MPA(madt);
- xsdt->checksum = generate_acpi_checksum(xsdt, xsdt->length);
+ xsdt->header.checksum = generate_acpi_checksum(xsdt,
+ xsdt->header.length);
/* setup FADT */
- memcpy(fadt->signature, FADT_SIG, sizeof(fadt->signature));
- fadt->length = sizeof(struct fadt_descriptor_rev2);
- fadt->revision = FADT2_REVISION_ID;
- memcpy(fadt->oem_id, "XEN", 3);
- memcpy(fadt->oem_table_id, "Xen/ia64", 8);
- memcpy(fadt->asl_compiler_id, "XEN", 3);
- fadt->asl_compiler_revision = xen_ia64_version(d);
-
- memcpy(facs->signature, FACS_SIG, sizeof(facs->signature));
+ memcpy(fadt->header.signature, ACPI_SIG_FADT,
+ sizeof(fadt->header.signature));
+ fadt->header.length = sizeof(struct acpi_table_fadt);
+ fadt->header.revision = FADT2_REVISION_ID;
+ memcpy(fadt->header.oem_id, "XEN", 3);
+ memcpy(fadt->header.oem_table_id, "Xen/ia64", 8);
+ memcpy(fadt->header.asl_compiler_id, "XEN", 3);
+ fadt->header.asl_compiler_revision = xen_ia64_version(d);
+
+ memcpy(facs->signature, ACPI_SIG_FACS, sizeof(facs->signature));
facs->version = 1;
- facs->length = sizeof(struct facs_descriptor_rev2);
-
- fadt->xfirmware_ctrl = ACPI_TABLE_MPA(facs);
+ facs->length = sizeof(struct acpi_table_facs);
+
+ fadt->Xfacs = ACPI_TABLE_MPA(facs);
fadt->Xdsdt = ACPI_TABLE_MPA(dsdt);
/*
@@ -280,34 +283,35 @@ dom_fw_fake_acpi(domain_t *d, struct fak
* from sanity checks in the ACPI CA. Emulate required ACPI hardware
* registers in system memory.
*/
- fadt->pm1_evt_len = 4;
- fadt->xpm1a_evt_blk.address_space_id = ACPI_ADR_SPACE_SYSTEM_MEMORY;
- fadt->xpm1a_evt_blk.register_bit_width = 8;
- fadt->xpm1a_evt_blk.address = ACPI_TABLE_MPA(pm1a_evt_blk);
- fadt->pm1_cnt_len = 1;
- fadt->xpm1a_cnt_blk.address_space_id = ACPI_ADR_SPACE_SYSTEM_MEMORY;
- fadt->xpm1a_cnt_blk.register_bit_width = 8;
- fadt->xpm1a_cnt_blk.address = ACPI_TABLE_MPA(pm1a_cnt_blk);
- fadt->pm_tm_len = 4;
- fadt->xpm_tmr_blk.address_space_id = ACPI_ADR_SPACE_SYSTEM_MEMORY;
- fadt->xpm_tmr_blk.register_bit_width = 8;
- fadt->xpm_tmr_blk.address = ACPI_TABLE_MPA(pm_tmr_blk);
-
- fadt->checksum = generate_acpi_checksum(fadt, fadt->length);
+ fadt->pm1_event_length = 4;
+ fadt->xpm1a_event_block.space_id = ACPI_ADR_SPACE_SYSTEM_MEMORY;
+ fadt->xpm1a_event_block.bit_width = 8;
+ fadt->xpm1a_event_block.address = ACPI_TABLE_MPA(pm1a_event_block);
+ fadt->pm1_control_length = 1;
+ fadt->xpm1a_control_block.space_id = ACPI_ADR_SPACE_SYSTEM_MEMORY;
+ fadt->xpm1a_control_block.bit_width = 8;
+ fadt->xpm1a_control_block.address = ACPI_TABLE_MPA(pm1a_control_block);
+ fadt->pm_timer_length = 4;
+ fadt->xpm_timer_block.space_id = ACPI_ADR_SPACE_SYSTEM_MEMORY;
+ fadt->xpm_timer_block.bit_width = 8;
+ fadt->xpm_timer_block.address = ACPI_TABLE_MPA(pm_timer_block);
+
+ fadt->header.checksum = generate_acpi_checksum(fadt,
+ fadt->header.length);
/* setup RSDP */
- memcpy(rsdp->signature, RSDP_SIG, strlen(RSDP_SIG));
+ memcpy(rsdp->signature, ACPI_SIG_RSDP, strlen(ACPI_SIG_RSDP));
memcpy(rsdp->oem_id, "XEN", 3);
rsdp->revision = 2; /* ACPI 2.0 includes XSDT */
- rsdp->length = sizeof(struct acpi20_table_rsdp);
- rsdp->xsdt_address = ACPI_TABLE_MPA(xsdt);
+ rsdp->length = sizeof(struct acpi_table_rsdp);
+ rsdp->xsdt_physical_address = ACPI_TABLE_MPA(xsdt);
rsdp->checksum = generate_acpi_checksum(rsdp,
ACPI_RSDP_CHECKSUM_LENGTH);
- rsdp->ext_checksum = generate_acpi_checksum(rsdp, rsdp->length);
+ rsdp->extended_checksum = generate_acpi_checksum(rsdp, rsdp->length);
/* setup DSDT with trivial namespace. */
- memcpy(dsdt->signature, DSDT_SIG, strlen(DSDT_SIG));
+ memcpy(dsdt->signature, ACPI_SIG_DSDT, strlen(ACPI_SIG_DSDT));
dsdt->revision = 1;
memcpy(dsdt->oem_id, "XEN", 3);
memcpy(dsdt->oem_table_id, "Xen/ia64", 8);
@@ -346,7 +350,8 @@ dom_fw_fake_acpi(domain_t *d, struct fak
dsdt->checksum = generate_acpi_checksum(dsdt, dsdt->length);
/* setup MADT */
- memcpy(madt->header.signature, APIC_SIG,
sizeof(madt->header.signature));
+ memcpy(madt->header.signature, ACPI_SIG_MADT,
+ sizeof(madt->header.signature));
madt->header.revision = 2;
memcpy(madt->header.oem_id, "XEN", 3);
memcpy(madt->header.oem_table_id, "Xen/ia64", 8);
diff -r 8c921adf4833 -r 42f6c206c951 xen/arch/ia64/xen/pcdp.c
--- a/xen/arch/ia64/xen/pcdp.c Fri Mar 14 15:07:45 2008 -0600
+++ b/xen/arch/ia64/xen/pcdp.c Thu Mar 20 12:35:40 2008 -0600
@@ -43,7 +43,7 @@ pcdp_hp_irq_fixup(struct pcdp *pcdp, str
{
efi_system_table_t *systab;
efi_config_table_t *tables;
- struct acpi20_table_rsdp *rsdp = NULL;
+ struct acpi_table_rsdp *rsdp = NULL;
struct acpi_table_xsdt *xsdt;
struct acpi_table_header *hdr;
int i;
@@ -66,16 +66,17 @@ pcdp_hp_irq_fixup(struct pcdp *pcdp, str
for (i = 0 ; i < (int)systab->nr_tables && !rsdp ; i++) {
if (efi_guidcmp(tables[i].guid, ACPI_20_TABLE_GUID) == 0)
rsdp =
- (struct acpi20_table_rsdp *)__va(tables[i].table);
- }
-
- if (!rsdp || strncmp(rsdp->signature, RSDP_SIG, sizeof(RSDP_SIG) - 1))
- return;
-
- xsdt = (struct acpi_table_xsdt *)__va(rsdp->xsdt_address);
+ (struct acpi_table_rsdp *)__va(tables[i].table);
+ }
+
+ if (!rsdp ||
+ strncmp(rsdp->signature, ACPI_SIG_RSDP, sizeof(ACPI_SIG_RSDP) - 1))
+ return;
+
+ xsdt = (struct acpi_table_xsdt *)__va(rsdp->xsdt_physical_address);
hdr = &xsdt->header;
- if (strncmp(hdr->signature, XSDT_SIG, sizeof(XSDT_SIG) - 1))
+ if (strncmp(hdr->signature, ACPI_SIG_XSDT, sizeof(ACPI_SIG_XSDT) - 1))
return;
/* Sanity check; are we still looking at HP firmware tables? */
diff -r 8c921adf4833 -r 42f6c206c951 xen/arch/x86/acpi/boot.c
--- a/xen/arch/x86/acpi/boot.c Fri Mar 14 15:07:45 2008 -0600
+++ b/xen/arch/x86/acpi/boot.c Thu Mar 20 12:35:40 2008 -0600
@@ -63,7 +63,7 @@ int acpi_strict;
int acpi_strict;
EXPORT_SYMBOL(acpi_strict);
-acpi_interrupt_flags acpi_sci_flags __initdata;
+u8 acpi_sci_flags __initdata;
int acpi_sci_override_gsi __initdata;
int acpi_skip_timer_override __initdata;
@@ -148,11 +148,11 @@ static int __init acpi_parse_madt(unsign
return -ENODEV;
}
- if (madt->lapic_address) {
- acpi_lapic_addr = (u64) madt->lapic_address;
+ if (madt->address) {
+ acpi_lapic_addr = (u64) madt->address;
printk(KERN_DEBUG PREFIX "Local APIC address 0x%08x\n",
- madt->lapic_address);
+ madt->address);
}
acpi_madt_oem_check(madt->header.oem_id, madt->header.oem_table_id);
@@ -342,23 +342,22 @@ static int __init acpi_parse_hpet(unsign
return -ENODEV;
}
- if (hpet_tbl->addr.space_id != ACPI_SPACE_MEM) {
+ if (hpet_tbl->address.space_id != ACPI_SPACE_MEM) {
printk(KERN_WARNING PREFIX "HPET timers must be located in "
"memory.\n");
return -1;
}
#if 0/*def CONFIG_X86_64*/
- vxtime.hpet_address = hpet_tbl->addr.addrl |
- ((long) hpet_tbl->addr.addrh << 32);
-
- printk(KERN_INFO PREFIX "HPET id: %#x base: %#lx\n",
- hpet_tbl->id, vxtime.hpet_address);
+ vxtime.hpet_address = hpet_tbl->address.address;
+
+ printk(KERN_INFO PREFIX "HPET id: %#x base: %#lx\n",
+ hpet_tbl->id, vxtime.hpet_address);
#else /* X86 */
{
extern unsigned long hpet_address;
- hpet_address = hpet_tbl->addr.addrl;
+ hpet_address = hpet_tbl->address.address;
printk(KERN_INFO PREFIX "HPET id: %#x base: %#lx\n",
hpet_tbl->id, hpet_address);
}
@@ -377,11 +376,11 @@ extern u32 pmtmr_ioport;
#ifdef CONFIG_ACPI_SLEEP
/* Get pm1x_cnt and pm1x_evt information for ACPI sleep */
static void __init
-acpi_fadt_parse_sleep_info(struct fadt_descriptor_rev2 *fadt)
+acpi_fadt_parse_sleep_info(struct acpi_table_fadt *fadt)
{
struct acpi_table_rsdp *rsdp;
unsigned long rsdp_phys;
- struct facs_descriptor_rev2 *facs = NULL;
+ struct acpi_table_facs *facs = NULL;
uint64_t facs_pa;
rsdp_phys = acpi_find_rsdp();
@@ -389,41 +388,46 @@ acpi_fadt_parse_sleep_info(struct fadt_d
goto bad;
rsdp = __va(rsdp_phys);
- if (fadt->revision >= FADT2_REVISION_ID) {
- /* Sanity check on FADT Rev. 2 */
- if ((fadt->xpm1a_cnt_blk.address_space_id !=
- ACPI_ADR_SPACE_SYSTEM_IO) ||
- (fadt->xpm1b_cnt_blk.address_space_id !=
- ACPI_ADR_SPACE_SYSTEM_IO) ||
- (fadt->xpm1a_evt_blk.address_space_id !=
- ACPI_ADR_SPACE_SYSTEM_IO) ||
- (fadt->xpm1b_evt_blk.address_space_id !=
- ACPI_ADR_SPACE_SYSTEM_IO))
- goto bad;
-
- acpi_sinfo.pm1a_cnt = (uint16_t)fadt->xpm1a_cnt_blk.address;
- acpi_sinfo.pm1b_cnt = (uint16_t)fadt->xpm1b_cnt_blk.address;
- acpi_sinfo.pm1a_evt = (uint16_t)fadt->xpm1a_evt_blk.address;
- acpi_sinfo.pm1b_evt = (uint16_t)fadt->xpm1b_evt_blk.address;
- }
-
- if (!acpi_sinfo.pm1a_cnt)
- acpi_sinfo.pm1a_cnt = (uint16_t)fadt->V1_pm1a_cnt_blk;
- if (!acpi_sinfo.pm1b_cnt)
- acpi_sinfo.pm1b_cnt = (uint16_t)fadt->V1_pm1b_cnt_blk;
- if (!acpi_sinfo.pm1a_evt)
- acpi_sinfo.pm1a_evt = (uint16_t)fadt->V1_pm1a_evt_blk;
- if (!acpi_sinfo.pm1b_evt)
- acpi_sinfo.pm1b_evt = (uint16_t)fadt->V1_pm1b_evt_blk;
+ if (fadt->header.revision >= FADT2_REVISION_ID) {
+ memcpy(&acpi_sinfo.pm1a_cnt_blk, &fadt->xpm1a_control_block,
+ sizeof(struct acpi_generic_address));
+ memcpy(&acpi_sinfo.pm1b_cnt_blk, &fadt->xpm1b_control_block,
+ sizeof(struct acpi_generic_address));
+ memcpy(&acpi_sinfo.pm1a_evt_blk, &fadt->xpm1a_event_block,
+ sizeof(struct acpi_generic_address));
+ memcpy(&acpi_sinfo.pm1b_evt_blk, &fadt->xpm1b_event_block,
+ sizeof(struct acpi_generic_address));
+ } else {
+ acpi_sinfo.pm1a_cnt_blk.address = fadt->pm1a_control_block;
+ acpi_sinfo.pm1b_cnt_blk.address = fadt->pm1b_control_block;
+ acpi_sinfo.pm1a_evt_blk.address = fadt->pm1a_event_block;
+ acpi_sinfo.pm1b_evt_blk.address = fadt->pm1b_event_block;
+ acpi_sinfo.pm1a_cnt_blk.space_id = ACPI_ADR_SPACE_SYSTEM_IO;
+ acpi_sinfo.pm1b_cnt_blk.space_id = ACPI_ADR_SPACE_SYSTEM_IO;
+ acpi_sinfo.pm1a_evt_blk.space_id = ACPI_ADR_SPACE_SYSTEM_IO;
+ acpi_sinfo.pm1b_evt_blk.space_id = ACPI_ADR_SPACE_SYSTEM_IO;
+ acpi_sinfo.pm1a_cnt_blk.bit_width = 16;
+ acpi_sinfo.pm1b_cnt_blk.bit_width = 16;
+ acpi_sinfo.pm1a_evt_blk.bit_width = 16;
+ acpi_sinfo.pm1b_evt_blk.bit_width = 16;
+ acpi_sinfo.pm1a_cnt_blk.bit_offset = 0;
+ acpi_sinfo.pm1b_cnt_blk.bit_offset = 0;
+ acpi_sinfo.pm1a_evt_blk.bit_offset = 0;
+ acpi_sinfo.pm1b_evt_blk.bit_offset = 0;
+ acpi_sinfo.pm1a_cnt_blk.access_width = 0;
+ acpi_sinfo.pm1b_cnt_blk.access_width = 0;
+ acpi_sinfo.pm1a_evt_blk.access_width = 0;
+ acpi_sinfo.pm1b_evt_blk.access_width = 0;
+ }
/* Now FACS... */
- if (fadt->revision >= FADT2_REVISION_ID)
- facs_pa = fadt->xfirmware_ctrl;
+ if (fadt->header.revision >= FADT2_REVISION_ID)
+ facs_pa = fadt->Xfacs;
else
- facs_pa = (uint64_t)fadt->V1_firmware_ctrl;
-
- facs = (struct facs_descriptor_rev2 *)
- __acpi_map_table(facs_pa, sizeof(struct facs_descriptor_rev2));
+ facs_pa = (uint64_t)fadt->facs;
+
+ facs = (struct acpi_table_facs *)
+ __acpi_map_table(facs_pa, sizeof(struct acpi_table_facs));
if (!facs)
goto bad;
@@ -446,20 +450,23 @@ acpi_fadt_parse_sleep_info(struct fadt_d
if ((rsdp->revision < 2) || (facs->length < 32)) {
acpi_sinfo.wakeup_vector = facs_pa +
- offsetof(struct facs_descriptor_rev2,
+ offsetof(struct acpi_table_facs,
firmware_waking_vector);
acpi_sinfo.vector_width = 32;
} else {
acpi_sinfo.wakeup_vector = facs_pa +
- offsetof(struct facs_descriptor_rev2,
+ offsetof(struct acpi_table_facs,
xfirmware_waking_vector);
acpi_sinfo.vector_width = 64;
}
printk(KERN_INFO PREFIX
- "ACPI SLEEP INFO: pm1x_cnt[%x,%x], pm1x_evt[%x,%x]\n",
- acpi_sinfo.pm1a_cnt, acpi_sinfo.pm1b_cnt,
- acpi_sinfo.pm1a_evt, acpi_sinfo.pm1b_cnt);
+ "ACPI SLEEP INFO: pm1x_cnt[%"PRIx64",%"PRIx64"], "
+ "pm1x_evt[%"PRIx64",%"PRIx64"]\n",
+ acpi_sinfo.pm1a_cnt_blk.address,
+ acpi_sinfo.pm1b_cnt_blk.address,
+ acpi_sinfo.pm1a_evt_blk.address,
+ acpi_sinfo.pm1b_evt_blk.address);
printk(KERN_INFO PREFIX
" wakeup_vec[%"PRIx64"], vec_size[%x]\n",
acpi_sinfo.wakeup_vector, acpi_sinfo.vector_width);
@@ -471,9 +478,9 @@ bad:
static int __init acpi_parse_fadt(unsigned long phys, unsigned long size)
{
- struct fadt_descriptor_rev2 *fadt = NULL;
-
- fadt = (struct fadt_descriptor_rev2 *)__acpi_map_table(phys, size);
+ struct acpi_table_fadt *fadt = NULL;
+
+ fadt = (struct acpi_table_fadt *)__acpi_map_table(phys, size);
if (!fadt) {
printk(KERN_WARNING PREFIX "Unable to map FADT\n");
return 0;
@@ -491,28 +498,28 @@ static int __init acpi_parse_fadt(unsign
#ifdef CONFIG_X86_PM_TIMER
/* detect the location of the ACPI PM Timer */
- if (fadt->revision >= FADT2_REVISION_ID) {
+ if (fadt->header.revision >= FADT2_REVISION_ID) {
/* FADT rev. 2 */
- if (fadt->xpm_tmr_blk.address_space_id ==
+ if (fadt->xpm_timer_block.space_id ==
ACPI_ADR_SPACE_SYSTEM_IO)
- pmtmr_ioport = fadt->xpm_tmr_blk.address;
+ pmtmr_ioport = fadt->xpm_timer_block.address;
/*
* "X" fields are optional extensions to the original V1.0
* fields, so we must selectively expand V1.0 fields if the
* corresponding X field is zero.
*/
if (!pmtmr_ioport)
- pmtmr_ioport = fadt->V1_pm_tmr_blk;
+ pmtmr_ioport = fadt->pm_timer_block;
} else {
/* FADT rev. 1 */
- pmtmr_ioport = fadt->V1_pm_tmr_blk;
+ pmtmr_ioport = fadt->pm_timer_block;
}
if (pmtmr_ioport)
printk(KERN_INFO PREFIX "PM-Timer IO Port: %#x\n",
pmtmr_ioport);
#endif
- acpi_smi_cmd = fadt->smi_cmd;
+ acpi_smi_cmd = fadt->smi_command;
acpi_enable_value = fadt->acpi_enable;
acpi_disable_value = fadt->acpi_disable;
diff -r 8c921adf4833 -r 42f6c206c951 xen/arch/x86/acpi/power.c
--- a/xen/arch/x86/acpi/power.c Fri Mar 14 15:07:45 2008 -0600
+++ b/xen/arch/x86/acpi/power.c Thu Mar 20 12:35:40 2008 -0600
@@ -106,7 +106,7 @@ static void acpi_sleep_prepare(u32 state
*(uint64_t *)wakeup_vector_va =
tboot_in_measured_env() ?
(uint64_t)g_tboot_shared->s3_tb_wakeup_entry :
- (uint64_t)bootsym_phys(wakeup_start);
+ (uint64_t)bootsym_phys(wakeup_start);
}
}
@@ -198,7 +198,7 @@ static long enter_state_helper(void *dat
*/
int acpi_enter_sleep(struct xenpf_enter_acpi_sleep *sleep)
{
- if ( !IS_PRIV(current->domain) || !acpi_sinfo.pm1a_cnt )
+ if ( !IS_PRIV(current->domain) || !acpi_sinfo.pm1a_cnt_blk.address )
return -EPERM;
/* Sanity check */
@@ -222,10 +222,14 @@ int acpi_enter_sleep(struct xenpf_enter_
static int acpi_get_wake_status(void)
{
- uint16_t val;
+ uint32_t val;
+ acpi_status status;
/* Wake status is the 15th bit of PM1 status register. (ACPI spec 3.0) */
- val = inw(acpi_sinfo.pm1a_evt) | inw(acpi_sinfo.pm1b_evt);
+ status = acpi_hw_register_read(ACPI_REGISTER_PM1_STATUS, &val);
+ if ( ACPI_FAILURE(status) )
+ return 0;
+
val &= ACPI_BITMASK_WAKE_STATUS;
val >>= ACPI_BITPOSITION_WAKE_STATUS;
return val;
@@ -243,7 +247,7 @@ static void tboot_sleep(u8 sleep_state)
case ACPI_STATE_S3:
shutdown_type = TB_SHUTDOWN_S3;
g_tboot_shared->s3_k_wakeup_entry =
- (uint32_t)bootsym_phys(wakeup_start);
+ (uint32_t)bootsym_phys(wakeup_start);
break;
case ACPI_STATE_S4:
shutdown_type = TB_SHUTDOWN_S4;
@@ -261,6 +265,8 @@ static void tboot_sleep(u8 sleep_state)
/* System is really put into sleep state by this stub */
acpi_status asmlinkage acpi_enter_sleep_state(u8 sleep_state)
{
+ acpi_status status;
+
if ( tboot_in_measured_env() )
{
tboot_sleep(sleep_state);
@@ -270,9 +276,18 @@ acpi_status asmlinkage acpi_enter_sleep_
ACPI_FLUSH_CPU_CACHE();
- outw((u16)acpi_sinfo.pm1a_cnt_val, acpi_sinfo.pm1a_cnt);
- if ( acpi_sinfo.pm1b_cnt )
- outw((u16)acpi_sinfo.pm1b_cnt_val, acpi_sinfo.pm1b_cnt);
+ status = acpi_hw_register_write(ACPI_REGISTER_PM1A_CONTROL,
+ acpi_sinfo.pm1a_cnt_val);
+ if ( ACPI_FAILURE(status) )
+ return_ACPI_STATUS(AE_ERROR);
+
+ if ( acpi_sinfo.pm1b_cnt_blk.address )
+ {
+ status = acpi_hw_register_write(ACPI_REGISTER_PM1B_CONTROL,
+ acpi_sinfo.pm1b_cnt_val);
+ if ( ACPI_FAILURE(status) )
+ return_ACPI_STATUS(AE_ERROR);
+ }
/* Wait until we enter sleep state, and spin until we wake */
while ( !acpi_get_wake_status() )
diff -r 8c921adf4833 -r 42f6c206c951 xen/arch/x86/domain.c
--- a/xen/arch/x86/domain.c Fri Mar 14 15:07:45 2008 -0600
+++ b/xen/arch/x86/domain.c Thu Mar 20 12:35:40 2008 -0600
@@ -46,7 +46,7 @@
#include <asm/debugreg.h>
#include <asm/msr.h>
#include <asm/nmi.h>
-#include <asm/iommu.h>
+#include <xen/iommu.h>
#ifdef CONFIG_COMPAT
#include <compat/vcpu.h>
#endif
@@ -830,7 +830,7 @@ unmap_vcpu_info(struct vcpu *v)
mfn = v->arch.vcpu_info_mfn;
unmap_domain_page_global(v->vcpu_info);
- v->vcpu_info = shared_info_addr(d, vcpu_info[v->vcpu_id]);
+ v->vcpu_info = (void *)&shared_info(d, vcpu_info[v->vcpu_id]);
v->arch.vcpu_info_mfn = INVALID_MFN;
put_page_and_type(mfn_to_page(mfn));
@@ -888,7 +888,7 @@ map_vcpu_info(struct vcpu *v, unsigned l
*/
vcpu_info(v, evtchn_upcall_pending) = 1;
for ( i = 0; i < BITS_PER_GUEST_LONG(d); i++ )
- set_bit(i, vcpu_info_addr(v, evtchn_pending_sel));
+ set_bit(i, &vcpu_info(v, evtchn_pending_sel));
/*
* Only bother to update time for the current vcpu. If we're
@@ -961,8 +961,9 @@ arch_do_vcpu_op(
if ( !v->domain->is_pinned )
break;
- cpu_id.phys_id = (x86_cpu_to_apicid[v->vcpu_id] |
- (acpi_get_processor_id(v->vcpu_id) << 8));
+ cpu_id.phys_id =
+ (uint64_t)x86_cpu_to_apicid[v->vcpu_id] |
+ ((uint64_t)acpi_get_processor_id(v->vcpu_id) << 32);
rc = -EFAULT;
if ( copy_to_guest(arg, &cpu_id, 1) )
diff -r 8c921adf4833 -r 42f6c206c951 xen/arch/x86/domctl.c
--- a/xen/arch/x86/domctl.c Fri Mar 14 15:07:45 2008 -0600
+++ b/xen/arch/x86/domctl.c Thu Mar 20 12:35:40 2008 -0600
@@ -26,7 +26,7 @@
#include <asm/hvm/cacheattr.h>
#include <asm/processor.h>
#include <xsm/xsm.h>
-#include <asm/iommu.h>
+#include <xen/iommu.h>
long arch_do_domctl(
struct xen_domctl *domctl,
diff -r 8c921adf4833 -r 42f6c206c951 xen/arch/x86/hvm/Makefile
--- a/xen/arch/x86/hvm/Makefile Fri Mar 14 15:07:45 2008 -0600
+++ b/xen/arch/x86/hvm/Makefile Thu Mar 20 12:35:40 2008 -0600
@@ -6,7 +6,6 @@ obj-y += i8254.o
obj-y += i8254.o
obj-y += intercept.o
obj-y += io.o
-obj-y += iommu.o
obj-y += irq.o
obj-y += mtrr.o
obj-y += pmtimer.o
diff -r 8c921adf4833 -r 42f6c206c951 xen/arch/x86/hvm/hvm.c
--- a/xen/arch/x86/hvm/hvm.c Fri Mar 14 15:07:45 2008 -0600
+++ b/xen/arch/x86/hvm/hvm.c Thu Mar 20 12:35:40 2008 -0600
@@ -59,8 +59,8 @@ struct hvm_function_table hvm_funcs __re
struct hvm_function_table hvm_funcs __read_mostly;
/* I/O permission bitmap is globally shared by all HVM guests. */
-char __attribute__ ((__section__ (".bss.page_aligned")))
- hvm_io_bitmap[3*PAGE_SIZE];
+unsigned long __attribute__ ((__section__ (".bss.page_aligned")))
+ hvm_io_bitmap[3*PAGE_SIZE/BYTES_PER_LONG];
void hvm_enable(struct hvm_function_table *fns)
{
diff -r 8c921adf4833 -r 42f6c206c951 xen/arch/x86/hvm/intercept.c
--- a/xen/arch/x86/hvm/intercept.c Fri Mar 14 15:07:45 2008 -0600
+++ b/xen/arch/x86/hvm/intercept.c Thu Mar 20 12:35:40 2008 -0600
@@ -30,7 +30,7 @@
#include <asm/current.h>
#include <io_ports.h>
#include <xen/event.h>
-#include <asm/iommu.h>
+#include <xen/iommu.h>
extern struct hvm_mmio_handler hpet_mmio_handler;
extern struct hvm_mmio_handler vlapic_mmio_handler;
diff -r 8c921adf4833 -r 42f6c206c951 xen/arch/x86/hvm/iommu.c
--- a/xen/arch/x86/hvm/iommu.c Fri Mar 14 15:07:45 2008 -0600
+++ /dev/null Thu Jan 01 00:00:00 1970 +0000
@@ -1,145 +0,0 @@
-/*
- * This program is free software; you can redistribute it and/or modify it
- * under the terms and conditions of the GNU General Public License,
- * version 2, as published by the Free Software Foundation.
- *
- * This program is distributed in the hope 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, Inc., 59 Temple
- * Place - Suite 330, Boston, MA 02111-1307 USA.
- */
-
-#include <xen/init.h>
-#include <xen/irq.h>
-#include <xen/spinlock.h>
-#include <xen/sched.h>
-#include <xen/xmalloc.h>
-#include <xen/domain_page.h>
-#include <asm/delay.h>
-#include <asm/string.h>
-#include <asm/mm.h>
-#include <asm/iommu.h>
-#include <asm/hvm/vmx/intel-iommu.h>
-
-extern struct iommu_ops intel_iommu_ops;
-extern struct iommu_ops amd_iommu_ops;
-
-int iommu_domain_init(struct domain *domain)
-{
- struct hvm_iommu *hd = domain_hvm_iommu(domain);
-
- spin_lock_init(&hd->mapping_lock);
- spin_lock_init(&hd->iommu_list_lock);
- INIT_LIST_HEAD(&hd->pdev_list);
- INIT_LIST_HEAD(&hd->g2m_ioport_list);
-
- if ( !iommu_enabled )
- return 0;
-
- switch ( boot_cpu_data.x86_vendor )
- {
- case X86_VENDOR_INTEL:
- hd->platform_ops = &intel_iommu_ops;
- break;
- case X86_VENDOR_AMD:
- hd->platform_ops = &amd_iommu_ops;
- break;
- default:
- BUG();
- }
-
- return hd->platform_ops->init(domain);
-}
-
-int assign_device(struct domain *d, u8 bus, u8 devfn)
-{
- struct hvm_iommu *hd = domain_hvm_iommu(d);
-
- if ( !iommu_enabled || !hd->platform_ops)
- return 0;
-
- return hd->platform_ops->assign_device(d, bus, devfn);
-}
-
-void iommu_domain_destroy(struct domain *d)
-{
- struct hvm_irq_dpci *hvm_irq_dpci = d->arch.hvm_domain.irq.dpci;
- uint32_t i;
- struct hvm_iommu *hd = domain_hvm_iommu(d);
- struct list_head *ioport_list, *digl_list, *tmp;
- struct g2m_ioport *ioport;
- struct dev_intx_gsi_link *digl;
-
- if ( !iommu_enabled || !hd->platform_ops)
- return;
-
- if ( hvm_irq_dpci != NULL )
- {
- for ( i = 0; i < NR_IRQS; i++ )
- {
- if ( !hvm_irq_dpci->mirq[i].valid )
- continue;
-
- pirq_guest_unbind(d, i);
- kill_timer(&hvm_irq_dpci->hvm_timer[irq_to_vector(i)]);
-
- list_for_each_safe ( digl_list, tmp,
- &hvm_irq_dpci->mirq[i].digl_list )
- {
- digl = list_entry(digl_list,
- struct dev_intx_gsi_link, list);
- list_del(&digl->list);
- xfree(digl);
- }
- }
-
- d->arch.hvm_domain.irq.dpci = NULL;
- xfree(hvm_irq_dpci);
- }
-
- if ( hd )
- {
- list_for_each_safe ( ioport_list, tmp, &hd->g2m_ioport_list )
- {
- ioport = list_entry(ioport_list, struct g2m_ioport, list);
- list_del(&ioport->list);
- xfree(ioport);
- }
- }
-
- return hd->platform_ops->teardown(d);
-}
-
-int iommu_map_page(struct domain *d, unsigned long gfn, unsigned long mfn)
-{
- struct hvm_iommu *hd = domain_hvm_iommu(d);
-
- if ( !iommu_enabled || !hd->platform_ops)
- return 0;
-
- return hd->platform_ops->map_page(d, gfn, mfn);
-}
-
-int iommu_unmap_page(struct domain *d, unsigned long gfn)
-{
- struct hvm_iommu *hd = domain_hvm_iommu(d);
-
- if ( !iommu_enabled || !hd->platform_ops)
- return 0;
-
- return hd->platform_ops->unmap_page(d, gfn);
-}
-
-void deassign_device(struct domain *d, u8 bus, u8 devfn)
-{
- struct hvm_iommu *hd = domain_hvm_iommu(d);
-
- if ( !iommu_enabled || !hd->platform_ops)
- return;
-
- return hd->platform_ops->reassign_device(d, dom0, bus, devfn);
-}
diff -r 8c921adf4833 -r 42f6c206c951 xen/arch/x86/hvm/svm/emulate.c
--- a/xen/arch/x86/hvm/svm/emulate.c Fri Mar 14 15:07:45 2008 -0600
+++ b/xen/arch/x86/hvm/svm/emulate.c Thu Mar 20 12:35:40 2008 -0600
@@ -117,7 +117,9 @@ int __get_instruction_length_from_list(s
}
else
{
- inst_copy_from_guest(buffer, svm_rip2pointer(v), MAX_INST_LEN);
+ if ( inst_copy_from_guest(buffer, svm_rip2pointer(v), MAX_INST_LEN)
+ != MAX_INST_LEN )
+ return 0;
buf = buffer;
}
diff -r 8c921adf4833 -r 42f6c206c951 xen/arch/x86/hvm/svm/svm.c
--- a/xen/arch/x86/hvm/svm/svm.c Fri Mar 14 15:07:45 2008 -0600
+++ b/xen/arch/x86/hvm/svm/svm.c Thu Mar 20 12:35:40 2008 -0600
@@ -495,6 +495,7 @@ static void svm_get_segment_register(str
break;
case x86_seg_ss:
memcpy(reg, &vmcb->ss, sizeof(*reg));
+ reg->attr.fields.dpl = vmcb->cpl;
break;
case x86_seg_tr:
svm_sync_vmcb(v);
@@ -943,6 +944,10 @@ static void svm_vmexit_do_cpuid(struct c
{
unsigned int eax, ebx, ecx, edx, inst_len;
+ inst_len = __get_instruction_length(current, INSTR_CPUID, NULL);
+ if ( inst_len == 0 )
+ return;
+
eax = regs->eax;
ebx = regs->ebx;
ecx = regs->ecx;
@@ -955,7 +960,6 @@ static void svm_vmexit_do_cpuid(struct c
regs->ecx = ecx;
regs->edx = edx;
- inst_len = __get_instruction_length(current, INSTR_CPUID, NULL);
__update_guest_eip(regs, inst_len);
}
@@ -1166,6 +1170,8 @@ static void svm_vmexit_do_hlt(struct vmc
unsigned int inst_len;
inst_len = __get_instruction_length(curr, INSTR_HLT, NULL);
+ if ( inst_len == 0 )
+ return;
__update_guest_eip(regs, inst_len);
/* Check for pending exception or new interrupt. */
@@ -1354,6 +1360,8 @@ asmlinkage void svm_vmexit_handler(struc
case VMEXIT_VMMCALL:
inst_len = __get_instruction_length(v, INSTR_VMCALL, NULL);
+ if ( inst_len == 0 )
+ break;
HVMTRACE_1D(VMMCALL, v, regs->eax);
rc = hvm_do_hypercall(regs);
if ( rc != HVM_HCALL_preempted )
diff -r 8c921adf4833 -r 42f6c206c951 xen/arch/x86/hvm/svm/vmcb.c
--- a/xen/arch/x86/hvm/svm/vmcb.c Fri Mar 14 15:07:45 2008 -0600
+++ b/xen/arch/x86/hvm/svm/vmcb.c Thu Mar 20 12:35:40 2008 -0600
@@ -80,27 +80,27 @@ struct host_save_area *alloc_host_save_a
void svm_disable_intercept_for_msr(struct vcpu *v, u32 msr)
{
- char *msr_bitmap = v->arch.hvm_svm.msrpm;
+ unsigned long *msr_bitmap = v->arch.hvm_svm.msrpm;
/*
* See AMD64 Programmers Manual, Vol 2, Section 15.10 (MSR-Bitmap Address).
*/
if ( msr <= 0x1fff )
{
- __clear_bit(msr*2, msr_bitmap + 0x000);
- __clear_bit(msr*2+1, msr_bitmap + 0x000);
+ __clear_bit(msr*2, msr_bitmap + 0x000/BYTES_PER_LONG);
+ __clear_bit(msr*2+1, msr_bitmap + 0x000/BYTES_PER_LONG);
}
else if ( (msr >= 0xc0000000) && (msr <= 0xc0001fff) )
{
msr &= 0x1fff;
- __clear_bit(msr*2, msr_bitmap + 0x800);
- __clear_bit(msr*2+1, msr_bitmap + 0x800);
+ __clear_bit(msr*2, msr_bitmap + 0x800/BYTES_PER_LONG);
+ __clear_bit(msr*2+1, msr_bitmap + 0x800/BYTES_PER_LONG);
}
else if ( (msr >= 0xc001000) && (msr <= 0xc0011fff) )
{
msr &= 0x1fff;
- __clear_bit(msr*2, msr_bitmap + 0x1000);
- __clear_bit(msr*2+1, msr_bitmap + 0x1000);
+ __clear_bit(msr*2, msr_bitmap + 0x1000/BYTES_PER_LONG);
+ __clear_bit(msr*2+1, msr_bitmap + 0x1000/BYTES_PER_LONG);
}
}
diff -r 8c921adf4833 -r 42f6c206c951 xen/arch/x86/hvm/vlapic.c
--- a/xen/arch/x86/hvm/vlapic.c Fri Mar 14 15:07:45 2008 -0600
+++ b/xen/arch/x86/hvm/vlapic.c Thu Mar 20 12:35:40 2008 -0600
@@ -83,15 +83,17 @@ static unsigned int vlapic_lvt_mask[VLAP
*/
#define VEC_POS(v) ((v)%32)
-#define REG_POS(v) (((v)/32)* 0x10)
-#define vlapic_test_and_set_vector(vec, bitmap) \
- test_and_set_bit(VEC_POS(vec), (bitmap) + REG_POS(vec))
-#define vlapic_test_and_clear_vector(vec, bitmap) \
- test_and_clear_bit(VEC_POS(vec), (bitmap) + REG_POS(vec))
-#define vlapic_set_vector(vec, bitmap) \
- set_bit(VEC_POS(vec), (bitmap) + REG_POS(vec))
-#define vlapic_clear_vector(vec, bitmap) \
- clear_bit(VEC_POS(vec), (bitmap) + REG_POS(vec))
+#define REG_POS(v) (((v)/32) * 0x10)
+#define vlapic_test_and_set_vector(vec, bitmap) \
+ test_and_set_bit(VEC_POS(vec), \
+ (unsigned long *)((bitmap) + REG_POS(vec)))
+#define vlapic_test_and_clear_vector(vec, bitmap) \
+ test_and_clear_bit(VEC_POS(vec), \
+ (unsigned long *)((bitmap) + REG_POS(vec)))
+#define vlapic_set_vector(vec, bitmap) \
+ set_bit(VEC_POS(vec), (unsigned long *)((bitmap) + REG_POS(vec)))
+#define vlapic_clear_vector(vec, bitmap) \
+ clear_bit(VEC_POS(vec), (unsigned long *)((bitmap) + REG_POS(vec)))
static int vlapic_find_highest_vector(void *bitmap)
{
diff -r 8c921adf4833 -r 42f6c206c951 xen/arch/x86/hvm/vmx/vmcs.c
--- a/xen/arch/x86/hvm/vmx/vmcs.c Fri Mar 14 15:07:45 2008 -0600
+++ b/xen/arch/x86/hvm/vmx/vmcs.c Thu Mar 20 12:35:40 2008 -0600
@@ -413,7 +413,7 @@ static void vmx_set_host_env(struct vcpu
void vmx_disable_intercept_for_msr(struct vcpu *v, u32 msr)
{
- char *msr_bitmap = v->arch.hvm_vmx.msr_bitmap;
+ unsigned long *msr_bitmap = v->arch.hvm_vmx.msr_bitmap;
/* VMX MSR bitmap supported? */
if ( msr_bitmap == NULL )
@@ -426,14 +426,14 @@ void vmx_disable_intercept_for_msr(struc
*/
if ( msr <= 0x1fff )
{
- __clear_bit(msr, msr_bitmap + 0x000); /* read-low */
- __clear_bit(msr, msr_bitmap + 0x800); /* write-low */
+ __clear_bit(msr, msr_bitmap + 0x000/BYTES_PER_LONG); /* read-low */
+ __clear_bit(msr, msr_bitmap + 0x800/BYTES_PER_LONG); /* write-low */
}
else if ( (msr >= 0xc0000000) && (msr <= 0xc0001fff) )
{
msr &= 0x1fff;
- __clear_bit(msr, msr_bitmap + 0x400); /* read-high */
- __clear_bit(msr, msr_bitmap + 0xc00); /* write-high */
+ __clear_bit(msr, msr_bitmap + 0x400/BYTES_PER_LONG); /* read-high */
+ __clear_bit(msr, msr_bitmap + 0xc00/BYTES_PER_LONG); /* write-high */
}
}
@@ -456,7 +456,7 @@ static int construct_vmcs(struct vcpu *v
/* MSR access bitmap. */
if ( cpu_has_vmx_msr_bitmap )
{
- char *msr_bitmap = alloc_xenheap_page();
+ unsigned long *msr_bitmap = alloc_xenheap_page();
if ( msr_bitmap == NULL )
return -ENOMEM;
@@ -870,7 +870,7 @@ void vmcs_dump_vcpu(struct vcpu *v)
x = (unsigned long long)vmr(TSC_OFFSET_HIGH) << 32;
x |= (uint32_t)vmr(TSC_OFFSET);
printk("TSC Offset = %016llx\n", x);
- x = (unsigned long long)vmr(GUEST_IA32_DEBUGCTL) << 32;
+ x = (unsigned long long)vmr(GUEST_IA32_DEBUGCTL_HIGH) << 32;
x |= (uint32_t)vmr(GUEST_IA32_DEBUGCTL);
printk("DebugCtl=%016llx DebugExceptions=%016llx\n", x,
(unsigned long long)vmr(GUEST_PENDING_DBG_EXCEPTIONS));
diff -r 8c921adf4833 -r 42f6c206c951 xen/arch/x86/hvm/vmx/vmx.c
--- a/xen/arch/x86/hvm/vmx/vmx.c Fri Mar 14 15:07:45 2008 -0600
+++ b/xen/arch/x86/hvm/vmx/vmx.c Thu Mar 20 12:35:40 2008 -0600
@@ -1512,8 +1512,10 @@ static int vmx_msr_read_intercept(struct
msr_content = var_range_base[index];
break;
case MSR_IA32_DEBUGCTLMSR:
- if ( vmx_read_guest_msr(v, ecx, &msr_content) != 0 )
- msr_content = 0;
+ msr_content = __vmread(GUEST_IA32_DEBUGCTL);
+#ifdef __i386__
+ msr_content |= (u64)__vmread(GUEST_IA32_DEBUGCTL_HIGH) << 32;
+#endif
break;
case MSR_IA32_VMX_BASIC...MSR_IA32_VMX_PROCBASED_CTLS2:
goto gp_fault;
@@ -1732,11 +1734,15 @@ static int vmx_msr_write_intercept(struc
}
if ( (rc < 0) ||
- (vmx_add_guest_msr(v, ecx) < 0) ||
(vmx_add_host_load_msr(v, ecx) < 0) )
vmx_inject_hw_exception(v, TRAP_machine_check, 0);
else
- vmx_write_guest_msr(v, ecx, msr_content);
+ {
+ __vmwrite(GUEST_IA32_DEBUGCTL, msr_content);
+#ifdef __i386__
+ __vmwrite(GUEST_IA32_DEBUGCTL_HIGH, msr_content >> 32);
+#endif
+ }
break;
}
diff -r 8c921adf4833 -r 42f6c206c951 xen/arch/x86/hvm/vmx/vpmu_core2.c
--- a/xen/arch/x86/hvm/vmx/vpmu_core2.c Fri Mar 14 15:07:45 2008 -0600
+++ b/xen/arch/x86/hvm/vmx/vpmu_core2.c Thu Mar 20 12:35:40 2008 -0600
@@ -101,7 +101,7 @@ static int is_core2_vpmu_msr(u32 msr_ind
return 0;
}
-static void core2_vpmu_set_msr_bitmap(char *msr_bitmap)
+static void core2_vpmu_set_msr_bitmap(unsigned long *msr_bitmap)
{
int i;
@@ -109,12 +109,14 @@ static void core2_vpmu_set_msr_bitmap(ch
for ( i = 0; i < core2_counters.num; i++ )
{
clear_bit(msraddr_to_bitpos(core2_counters.msr[i]), msr_bitmap);
- clear_bit(msraddr_to_bitpos(core2_counters.msr[i]), msr_bitmap+0x800);
+ clear_bit(msraddr_to_bitpos(core2_counters.msr[i]),
+ msr_bitmap + 0x800/BYTES_PER_LONG);
}
for ( i = 0; i < core2_get_pmc_count(); i++ )
{
clear_bit(msraddr_to_bitpos(MSR_IA32_PERFCTR0+i), msr_bitmap);
- clear_bit(msraddr_to_bitpos(MSR_IA32_PERFCTR0+i), msr_bitmap+0x800);
+ clear_bit(msraddr_to_bitpos(MSR_IA32_PERFCTR0+i),
+ msr_bitmap + 0x800/BYTES_PER_LONG);
}
/* Allow Read PMU Non-global Controls Directly. */
@@ -124,19 +126,21 @@ static void core2_vpmu_set_msr_bitmap(ch
clear_bit(msraddr_to_bitpos(MSR_P6_EVNTSEL0+i), msr_bitmap);
}
-static void core2_vpmu_unset_msr_bitmap(char *msr_bitmap)
+static void core2_vpmu_unset_msr_bitmap(unsigned long *msr_bitmap)
{
int i;
for ( i = 0; i < core2_counters.num; i++ )
{
set_bit(msraddr_to_bitpos(core2_counters.msr[i]), msr_bitmap);
- set_bit(msraddr_to_bitpos(core2_counters.msr[i]), msr_bitmap+0x800);
+ set_bit(msraddr_to_bitpos(core2_counters.msr[i]),
+ msr_bitmap + 0x800/BYTES_PER_LONG);
}
for ( i = 0; i < core2_get_pmc_count(); i++ )
{
set_bit(msraddr_to_bitpos(MSR_IA32_PERFCTR0+i), msr_bitmap);
- set_bit(msraddr_to_bitpos(MSR_IA32_PERFCTR0+i), msr_bitmap+0x800);
+ set_bit(msraddr_to_bitpos(MSR_IA32_PERFCTR0+i),
+ msr_bitmap + 0x800/BYTES_PER_LONG);
}
for ( i = 0; i < core2_ctrls.num; i++ )
set_bit(msraddr_to_bitpos(core2_ctrls.msr[i]), msr_bitmap);
diff -r 8c921adf4833 -r 42f6c206c951 xen/arch/x86/irq.c
--- a/xen/arch/x86/irq.c Fri Mar 14 15:07:45 2008 -0600
+++ b/xen/arch/x86/irq.c Thu Mar 20 12:35:40 2008 -0600
@@ -15,7 +15,7 @@
#include <xen/keyhandler.h>
#include <xen/compat.h>
#include <asm/current.h>
-#include <asm/iommu.h>
+#include <xen/iommu.h>
/* opt_noirqbalance: If true, software IRQ balancing/affinity is disabled. */
int opt_noirqbalance = 0;
@@ -362,13 +362,12 @@ int pirq_guest_unmask(struct domain *d)
int pirq_guest_unmask(struct domain *d)
{
unsigned int irq;
- shared_info_t *s = d->shared_info;
for ( irq = find_first_bit(d->pirq_mask, NR_IRQS);
irq < NR_IRQS;
irq = find_next_bit(d->pirq_mask, NR_IRQS, irq+1) )
{
- if ( !test_bit(d->pirq_to_evtchn[irq], __shared_info_addr(d, s,
evtchn_mask)) )
+ if ( !test_bit(d->pirq_to_evtchn[irq], &shared_info(d, evtchn_mask)) )
__pirq_guest_eoi(d, irq);
}
@@ -660,13 +659,13 @@ static void dump_irqs(unsigned char key)
printk("%u(%c%c%c%c)",
d->domain_id,
(test_bit(d->pirq_to_evtchn[irq],
- shared_info_addr(d, evtchn_pending)) ?
+ &shared_info(d, evtchn_pending)) ?
'P' : '-'),
(test_bit(d->pirq_to_evtchn[irq]/BITS_PER_GUEST_LONG(d),
- vcpu_info_addr(d->vcpu[0],
evtchn_pending_sel)) ?
+ &vcpu_info(d->vcpu[0], evtchn_pending_sel)) ?
'S' : '-'),
(test_bit(d->pirq_to_evtchn[irq],
- shared_info_addr(d, evtchn_mask)) ?
+ &shared_info(d, evtchn_mask)) ?
'M' : '-'),
(test_bit(irq, d->pirq_mask) ?
'M' : '-'));
diff -r 8c921adf4833 -r 42f6c206c951 xen/arch/x86/mm.c
--- a/xen/arch/x86/mm.c Fri Mar 14 15:07:45 2008 -0600
+++ b/xen/arch/x86/mm.c Thu Mar 20 12:35:40 2008 -0600
@@ -645,6 +645,7 @@ get_page_from_l1e(
struct page_info *page = mfn_to_page(mfn);
uint32_t l1f = l1e_get_flags(l1e);
struct vcpu *curr = current;
+ struct domain *owner;
int okay;
if ( !(l1f & _PAGE_PRESENT) )
@@ -672,6 +673,17 @@ get_page_from_l1e(
return 1;
}
+
+ /*
+ * Let privileged domains transfer the right to map their target
+ * domain's pages. This is used to allow stub-domain pvfb export to dom0,
+ * until pvfb supports granted mappings. At that time this minor hack
+ * can go away.
+ */
+ owner = page_get_owner(page);
+ if ( unlikely(d != owner) && (owner != NULL) &&
+ (d != curr->domain) && IS_PRIV_FOR(d, owner) )
+ d = owner;
/* Foreign mappings into guests in shadow external mode don't
* contribute to writeable mapping refcounts. (This allows the
diff -r 8c921adf4833 -r 42f6c206c951 xen/arch/x86/mm/p2m.c
--- a/xen/arch/x86/mm/p2m.c Fri Mar 14 15:07:45 2008 -0600
+++ b/xen/arch/x86/mm/p2m.c Thu Mar 20 12:35:40 2008 -0600
@@ -27,7 +27,7 @@
#include <asm/page.h>
#include <asm/paging.h>
#include <asm/p2m.h>
-#include <asm/iommu.h>
+#include <xen/iommu.h>
/* Debugging and auditing of the P2M code? */
#define P2M_AUDIT 0
@@ -229,7 +229,7 @@ set_p2m_entry(struct domain *d, unsigned
if ( !p2m_next_level(d, &table_mfn, &table, &gfn_remainder, gfn,
L3_PAGETABLE_SHIFT - PAGE_SHIFT,
((CONFIG_PAGING_LEVELS == 3)
- ? (hvm_funcs.hap_supported ? 4 : 8)
+ ? (d->arch.hvm_domain.hap_enabled ? 4 : 8)
: L3_PAGETABLE_ENTRIES),
PGT_l2_page_table) )
goto out;
diff -r 8c921adf4833 -r 42f6c206c951 xen/arch/x86/mm/paging.c
--- a/xen/arch/x86/mm/paging.c Fri Mar 14 15:07:45 2008 -0600
+++ b/xen/arch/x86/mm/paging.c Thu Mar 20 12:35:40 2008 -0600
@@ -114,7 +114,8 @@ static mfn_t paging_new_log_dirty_page(s
return mfn;
}
-static mfn_t paging_new_log_dirty_leaf(struct domain *d, uint8_t **leaf_p)
+static mfn_t paging_new_log_dirty_leaf(
+ struct domain *d, unsigned long **leaf_p)
{
mfn_t mfn = paging_new_log_dirty_page(d, (void **)leaf_p);
if ( mfn_valid(mfn) )
@@ -264,7 +265,7 @@ void paging_mark_dirty(struct domain *d,
mfn_t gmfn;
int changed;
mfn_t mfn, *l4, *l3, *l2;
- uint8_t *l1;
+ unsigned long *l1;
int i1, i2, i3, i4;
gmfn = _mfn(guest_mfn);
@@ -341,7 +342,7 @@ int paging_log_dirty_op(struct domain *d
int rv = 0, clean = 0, peek = 1;
unsigned long pages = 0;
mfn_t *l4, *l3, *l2;
- uint8_t *l1;
+ unsigned long *l1;
int i4, i3, i2;
domain_pause(d);
@@ -399,7 +400,7 @@ int paging_log_dirty_op(struct domain *d
(pages < sc->pages) && (i2 < LOGDIRTY_NODE_ENTRIES);
i2++ )
{
- static uint8_t zeroes[PAGE_SIZE];
+ static unsigned long zeroes[PAGE_SIZE/BYTES_PER_LONG];
unsigned int bytes = PAGE_SIZE;
l1 = ((l2 && mfn_valid(l2[i2])) ?
map_domain_page(mfn_x(l2[i2])) : zeroes);
@@ -408,7 +409,7 @@ int paging_log_dirty_op(struct domain *d
if ( likely(peek) )
{
if ( copy_to_guest_offset(sc->dirty_bitmap, pages >> 3,
- l1, bytes) != 0 )
+ (uint8_t *)l1, bytes) != 0 )
{
rv = -EFAULT;
goto out;
diff -r 8c921adf4833 -r 42f6c206c951 xen/arch/x86/mm/shadow/private.h
--- a/xen/arch/x86/mm/shadow/private.h Fri Mar 14 15:07:45 2008 -0600
+++ b/xen/arch/x86/mm/shadow/private.h Thu Mar 20 12:35:40 2008 -0600
@@ -483,7 +483,7 @@ sh_mfn_is_dirty(struct domain *d, mfn_t
{
unsigned long pfn;
mfn_t mfn, *l4, *l3, *l2;
- uint8_t *l1;
+ unsigned long *l1;
int rv;
ASSERT(shadow_mode_log_dirty(d));
diff -r 8c921adf4833 -r 42f6c206c951 xen/arch/x86/numa.c
--- a/xen/arch/x86/numa.c Fri Mar 14 15:07:45 2008 -0600
+++ b/xen/arch/x86/numa.c Thu Mar 20 12:35:40 2008 -0600
@@ -57,7 +57,7 @@ populate_memnodemap(const struct node *n
{
int i;
int res = -1;
- unsigned long addr, end;
+ paddr_t addr, end;
if (shift >= 64)
return -1;
@@ -286,13 +286,13 @@ static void dump_numa(unsigned char key)
(u32)(now>>32), (u32)now);
for_each_online_node(i) {
- unsigned long pa = (NODE_DATA(i)->node_start_pfn + 1)<<
PAGE_SHIFT;
+ paddr_t pa = (NODE_DATA(i)->node_start_pfn + 1)<< PAGE_SHIFT;
printk("idx%d -> NODE%d start->%lu size->%lu\n",
i, NODE_DATA(i)->node_id,
NODE_DATA(i)->node_start_pfn,
NODE_DATA(i)->node_spanned_pages);
/* sanity check phys_to_nid() */
- printk("phys_to_nid(%lx) -> %d should be %d\n", pa,
phys_to_nid(pa),
+ printk("phys_to_nid(%"PRIpaddr") -> %d should be %d\n", pa,
phys_to_nid(pa),
NODE_DATA(i)->node_id);
}
for_each_online_cpu(i)
diff -r 8c921adf4833 -r 42f6c206c951 xen/arch/x86/srat.c
--- a/xen/arch/x86/srat.c Fri Mar 14 15:07:45 2008 -0600
+++ b/xen/arch/x86/srat.c Thu Mar 20 12:35:40 2008 -0600
@@ -106,7 +106,7 @@ static __init int slit_valid(struct acpi
static __init int slit_valid(struct acpi_table_slit *slit)
{
int i, j;
- int d = slit->localities;
+ int d = slit->locality_count;
for (i = 0; i < d; i++) {
for (j = 0; j < d; j++) {
u8 val = slit->entry[d*i + j];
@@ -308,7 +308,7 @@ int __node_distance(int a, int b)
if (!acpi_slit)
return a == b ? 10 : 20;
- index = acpi_slit->localities * node_to_pxm(a);
+ index = acpi_slit->locality_count * node_to_pxm(a);
return acpi_slit->entry[index + node_to_pxm(b)];
}
diff -r 8c921adf4833 -r 42f6c206c951 xen/arch/x86/traps.c
--- a/xen/arch/x86/traps.c Fri Mar 14 15:07:45 2008 -0600
+++ b/xen/arch/x86/traps.c Thu Mar 20 12:35:40 2008 -0600
@@ -677,32 +677,75 @@ static int emulate_forced_invalid_op(str
: "=a" (a), "=b" (b), "=c" (c), "=d" (d)
: "0" (a), "1" (b), "2" (c), "3" (d) );
- if ( regs->eax == 1 )
+ if ( (regs->eax & 0x7fffffff) == 1 )
{
/* Modify Feature Information. */
__clear_bit(X86_FEATURE_VME, &d);
__clear_bit(X86_FEATURE_PSE, &d);
__clear_bit(X86_FEATURE_PGE, &d);
+ __clear_bit(X86_FEATURE_MCE, &d);
+ __clear_bit(X86_FEATURE_MCA, &d);
+ if ( !IS_PRIV(current->domain) )
+ __clear_bit(X86_FEATURE_MTRR, &d);
+ __clear_bit(X86_FEATURE_PSE36, &d);
+ }
+ switch ( (uint32_t)regs->eax )
+ {
+ case 1:
+ /* Modify Feature Information. */
if ( !cpu_has_sep )
__clear_bit(X86_FEATURE_SEP, &d);
#ifdef __i386__
if ( !supervisor_mode_kernel )
__clear_bit(X86_FEATURE_SEP, &d);
#endif
- if ( !IS_PRIV(current->domain) )
- __clear_bit(X86_FEATURE_MTRR, &d);
- }
- else if ( regs->eax == 0x80000001 )
- {
+ __clear_bit(X86_FEATURE_DS, &d);
+ __clear_bit(X86_FEATURE_ACC, &d);
+ __clear_bit(X86_FEATURE_PBE, &d);
+
+ __clear_bit(X86_FEATURE_DTES64 % 32, &c);
+ __clear_bit(X86_FEATURE_MWAIT % 32, &c);
+ __clear_bit(X86_FEATURE_DSCPL % 32, &c);
+ __clear_bit(X86_FEATURE_VMXE % 32, &c);
+ __clear_bit(X86_FEATURE_SMXE % 32, &c);
+ __clear_bit(X86_FEATURE_EST % 32, &c);
+ __clear_bit(X86_FEATURE_TM2 % 32, &c);
+ if ( is_pv_32bit_vcpu(current) )
+ __clear_bit(X86_FEATURE_CX16 % 32, &c);
+ __clear_bit(X86_FEATURE_XTPR % 32, &c);
+ __clear_bit(X86_FEATURE_PDCM % 32, &c);
+ __clear_bit(X86_FEATURE_DCA % 32, &c);
+ break;
+ case 0x80000001:
/* Modify Feature Information. */
-#ifdef __i386__
- __clear_bit(X86_FEATURE_SYSCALL % 32, &d);
-#endif
+ if ( is_pv_32bit_vcpu(current) )
+ {
+ __clear_bit(X86_FEATURE_LM % 32, &d);
+ __clear_bit(X86_FEATURE_LAHF_LM % 32, &c);
+ }
+#ifndef __i386__
+ if ( is_pv_32on64_vcpu(current) &&
+ boot_cpu_data.x86_vendor != X86_VENDOR_AMD )
+#endif
+ __clear_bit(X86_FEATURE_SYSCALL % 32, &d);
+ __clear_bit(X86_FEATURE_PAGE1GB % 32, &d);
__clear_bit(X86_FEATURE_RDTSCP % 32, &d);
- }
- else
- {
+
+ __clear_bit(X86_FEATURE_SVME % 32, &c);
+ __clear_bit(X86_FEATURE_OSVW % 32, &c);
+ __clear_bit(X86_FEATURE_IBS % 32, &c);
+ __clear_bit(X86_FEATURE_SKINIT % 32, &c);
+ __clear_bit(X86_FEATURE_WDT % 32, &c);
+ break;
+ case 5: /* MONITOR/MWAIT */
+ case 0xa: /* Architectural Performance Monitor Features */
+ case 0x8000000a: /* SVM revision and features */
+ case 0x8000001b: /* Instruction Based Sampling */
+ a = b = c = d = 0;
+ break;
+ default:
(void)cpuid_hypervisor_leaves(regs->eax, &a, &b, &c, &d);
+ break;
}
regs->eax = a;
@@ -1801,6 +1844,8 @@ static int emulate_privileged_op(struct
case 0x20: /* MOV CR?,<reg> */
opcode = insn_fetch(u8, code_base, eip, code_limit);
+ if ( opcode < 0xc0 )
+ goto fail;
modrm_reg += ((opcode >> 3) & 7) + (lock << 3);
modrm_rm |= (opcode >> 0) & 7;
reg = decode_register(modrm_rm, regs, 0);
@@ -1841,6 +1886,8 @@ static int emulate_privileged_op(struct
case 0x21: /* MOV DR?,<reg> */
opcode = insn_fetch(u8, code_base, eip, code_limit);
+ if ( opcode < 0xc0 )
+ goto fail;
modrm_reg += ((opcode >> 3) & 7) + (lock << 3);
modrm_rm |= (opcode >> 0) & 7;
reg = decode_register(modrm_rm, regs, 0);
@@ -1851,6 +1898,8 @@ static int emulate_privileged_op(struct
case 0x22: /* MOV <reg>,CR? */
opcode = insn_fetch(u8, code_base, eip, code_limit);
+ if ( opcode < 0xc0 )
+ goto fail;
modrm_reg += ((opcode >> 3) & 7) + (lock << 3);
modrm_rm |= (opcode >> 0) & 7;
reg = decode_register(modrm_rm, regs, 0);
@@ -1897,6 +1946,8 @@ static int emulate_privileged_op(struct
case 0x23: /* MOV <reg>,DR? */
opcode = insn_fetch(u8, code_base, eip, code_limit);
+ if ( opcode < 0xc0 )
+ goto fail;
modrm_reg += ((opcode >> 3) & 7) + (lock << 3);
modrm_rm |= (opcode >> 0) & 7;
reg = decode_register(modrm_rm, regs, 0);
@@ -2026,6 +2077,15 @@ static int emulate_privileged_op(struct
case MSR_EFER:
if ( rdmsr_safe(regs->ecx, regs->eax, regs->edx) )
goto fail;
+ break;
+ case MSR_IA32_MISC_ENABLE:
+ if ( rdmsr_safe(regs->ecx, regs->eax, regs->edx) )
+ goto fail;
+ regs->eax &= ~(MSR_IA32_MISC_ENABLE_PERF_AVAIL |
+ MSR_IA32_MISC_ENABLE_MONITOR_ENABLE);
+ regs->eax |= MSR_IA32_MISC_ENABLE_BTS_UNAVAIL |
+ MSR_IA32_MISC_ENABLE_PEBS_UNAVAIL |
+ MSR_IA32_MISC_ENABLE_XTPR_DISABLE;
break;
default:
if ( rdmsr_hypervisor_regs(regs->ecx, &l, &h) )
diff -r 8c921adf4833 -r 42f6c206c951 xen/arch/x86/x86_emulate.c
--- a/xen/arch/x86/x86_emulate.c Fri Mar 14 15:07:45 2008 -0600
+++ b/xen/arch/x86/x86_emulate.c Thu Mar 20 12:35:40 2008 -0600
@@ -785,11 +785,21 @@ _mode_iopl(
struct x86_emulate_ops *ops)
{
int cpl = get_cpl(ctxt, ops);
+ if ( cpl == -1 )
+ return -1;
return ((cpl >= 0) && (cpl <= ((ctxt->regs->eflags >> 12) & 3)));
}
-#define mode_ring0() (get_cpl(ctxt, ops) == 0)
-#define mode_iopl() _mode_iopl(ctxt, ops)
+#define mode_ring0() ({ \
+ int _cpl = get_cpl(ctxt, ops); \
+ fail_if(_cpl < 0); \
+ (_cpl == 0); \
+})
+#define mode_iopl() ({ \
+ int _iopl = _mode_iopl(ctxt, ops); \
+ fail_if(_iopl < 0); \
+ _iopl; \
+})
static int
in_realmode(
@@ -2255,7 +2265,6 @@ x86_emulate(
case 0x6c ... 0x6d: /* ins %dx,%es:%edi */ {
unsigned long nr_reps = get_rep_prefix();
- generate_exception_if(!mode_iopl(), EXC_GP, 0);
dst.bytes = !(b & 1) ? 1 : (op_bytes == 8) ? 4 : op_bytes;
dst.mem.seg = x86_seg_es;
dst.mem.off = truncate_ea(_regs.edi);
@@ -2285,7 +2294,6 @@ x86_emulate(
case 0x6e ... 0x6f: /* outs %esi,%dx */ {
unsigned long nr_reps = get_rep_prefix();
- generate_exception_if(!mode_iopl(), EXC_GP, 0);
dst.bytes = !(b & 1) ? 1 : (op_bytes == 8) ? 4 : op_bytes;
if ( (nr_reps > 1) && (ops->rep_outs != NULL) &&
((rc = ops->rep_outs(ea.mem.seg, truncate_ea(_regs.esi),
@@ -2394,8 +2402,10 @@ x86_emulate(
case 0x9d: /* popf */ {
uint32_t mask = EFLG_VIP | EFLG_VIF | EFLG_VM;
+ if ( !mode_ring0() )
+ mask |= EFLG_IOPL;
if ( !mode_iopl() )
- mask |= EFLG_IOPL;
+ mask |= EFLG_IF;
/* 64-bit mode: POP defaults to a 64-bit operand. */
if ( mode_64bit() && (op_bytes == 4) )
op_bytes = 8;
@@ -2640,8 +2650,10 @@ x86_emulate(
case 0xcf: /* iret */ {
unsigned long cs, eip, eflags;
uint32_t mask = EFLG_VIP | EFLG_VIF | EFLG_VM;
+ if ( !mode_ring0() )
+ mask |= EFLG_IOPL;
if ( !mode_iopl() )
- mask |= EFLG_IOPL;
+ mask |= EFLG_IF;
fail_if(!in_realmode(ctxt, ops));
if ( (rc = ops->read(x86_seg_ss, sp_post_inc(op_bytes),
&eip, op_bytes, ctxt)) ||
@@ -2818,7 +2830,6 @@ x86_emulate(
unsigned int port = ((b < 0xe8)
? insn_fetch_type(uint8_t)
: (uint16_t)_regs.edx);
- generate_exception_if(!mode_iopl(), EXC_GP, 0);
op_bytes = !(b & 1) ? 1 : (op_bytes == 8) ? 4 : op_bytes;
if ( b & 2 )
{
@@ -3219,8 +3230,8 @@ x86_emulate(
case 0x21: /* mov dr,reg */
case 0x22: /* mov reg,cr */
case 0x23: /* mov reg,dr */
+ generate_exception_if(ea.type != OP_REG, EXC_UD, -1);
generate_exception_if(!mode_ring0(), EXC_GP, 0);
- modrm_rm |= (rex_prefix & 1) << 3;
modrm_reg |= lock_prefix << 3;
if ( b & 2 )
{
diff -r 8c921adf4833 -r 42f6c206c951 xen/common/domain.c
--- a/xen/common/domain.c Fri Mar 14 15:07:45 2008 -0600
+++ b/xen/common/domain.c Thu Mar 20 12:35:40 2008 -0600
@@ -110,10 +110,6 @@ static void __domain_finalise_shutdown(s
return;
d->is_shut_down = 1;
-
- for_each_vcpu ( d, v )
- vcpu_sleep_nosync(v);
-
send_guest_global_virq(dom0, VIRQ_DOM_EXC);
}
@@ -126,7 +122,7 @@ static void vcpu_check_shutdown(struct v
if ( d->is_shutting_down )
{
if ( !v->paused_for_shutdown )
- atomic_inc(&v->pause_count);
+ vcpu_pause_nosync(v);
v->paused_for_shutdown = 1;
v->defer_shutdown = 0;
__domain_finalise_shutdown(d);
@@ -154,7 +150,7 @@ struct vcpu *alloc_vcpu(
if ( !is_idle_domain(d) )
{
set_bit(_VPF_down, &v->pause_flags);
- v->vcpu_info = shared_info_addr(d, vcpu_info[vcpu_id]);
+ v->vcpu_info = (void *)&shared_info(d, vcpu_info[vcpu_id]);
}
if ( sched_init_vcpu(v, cpu_id) != 0 )
@@ -426,7 +422,7 @@ void domain_shutdown(struct domain *d, u
{
if ( v->defer_shutdown )
continue;
- atomic_inc(&v->pause_count);
+ vcpu_pause_nosync(v);
v->paused_for_shutdown = 1;
}
diff -r 8c921adf4833 -r 42f6c206c951 xen/common/event_channel.c
--- a/xen/common/event_channel.c Fri Mar 14 15:07:45 2008 -0600
+++ b/xen/common/event_channel.c Thu Mar 20 12:35:40 2008 -0600
@@ -539,7 +539,6 @@ void evtchn_set_pending(struct vcpu *v,
void evtchn_set_pending(struct vcpu *v, int port)
{
struct domain *d = v->domain;
- shared_info_t *s = d->shared_info;
/*
* The following bit operations must happen in strict order.
@@ -548,12 +547,12 @@ void evtchn_set_pending(struct vcpu *v,
* others may require explicit memory barriers.
*/
- if ( test_and_set_bit(port, __shared_info_addr(d, s, evtchn_pending)) )
+ if ( test_and_set_bit(port, &shared_info(d, evtchn_pending)) )
return;
- if ( !test_bit (port, __shared_info_addr(d, s, evtchn_mask)) &&
+ if ( !test_bit (port, &shared_info(d, evtchn_mask)) &&
!test_and_set_bit(port / BITS_PER_GUEST_LONG(d),
- vcpu_info_addr(v, evtchn_pending_sel)) )
+ &vcpu_info(v, evtchn_pending_sel)) )
{
vcpu_mark_events_pending(v);
}
@@ -750,7 +749,6 @@ static long evtchn_unmask(evtchn_unmask_
static long evtchn_unmask(evtchn_unmask_t *unmask)
{
struct domain *d = current->domain;
- shared_info_t *s = d->shared_info;
int port = unmask->port;
struct vcpu *v;
@@ -768,10 +766,10 @@ static long evtchn_unmask(evtchn_unmask_
* These operations must happen in strict order. Based on
* include/xen/event.h:evtchn_set_pending().
*/
- if ( test_and_clear_bit(port, __shared_info_addr(d, s, evtchn_mask)) &&
- test_bit (port, __shared_info_addr(d, s, evtchn_pending)) &&
+ if ( test_and_clear_bit(port, &shared_info(d, evtchn_mask)) &&
+ test_bit (port, &shared_info(d, evtchn_pending)) &&
!test_and_set_bit (port / BITS_PER_GUEST_LONG(d),
- vcpu_info_addr(v, evtchn_pending_sel)) )
+ &vcpu_info(v, evtchn_pending_sel)) )
{
vcpu_mark_events_pending(v);
}
diff -r 8c921adf4833 -r 42f6c206c951 xen/common/keyhandler.c
--- a/xen/common/keyhandler.c Fri Mar 14 15:07:45 2008 -0600
+++ b/xen/common/keyhandler.c Thu Mar 20 12:35:40 2008 -0600
@@ -201,12 +201,12 @@ static void dump_domains(unsigned char k
printk(" Notifying guest (virq %d, port %d, stat %d/%d/%d)\n",
VIRQ_DEBUG, v->virq_to_evtchn[VIRQ_DEBUG],
test_bit(v->virq_to_evtchn[VIRQ_DEBUG],
- shared_info_addr(d, evtchn_pending)),
+ &shared_info(d, evtchn_pending)),
test_bit(v->virq_to_evtchn[VIRQ_DEBUG],
- shared_info_addr(d, evtchn_mask)),
+ &shared_info(d, evtchn_mask)),
test_bit(v->virq_to_evtchn[VIRQ_DEBUG] /
BITS_PER_GUEST_LONG(d),
- vcpu_info_addr(v, evtchn_pending_sel)));
+ &vcpu_info(v, evtchn_pending_sel)));
send_guest_vcpu_virq(v, VIRQ_DEBUG);
}
}
diff -r 8c921adf4833 -r 42f6c206c951 xen/common/schedule.c
--- a/xen/common/schedule.c Fri Mar 14 15:07:45 2008 -0600
+++ b/xen/common/schedule.c Thu Mar 20 12:35:40 2008 -0600
@@ -365,7 +365,7 @@ static long do_poll(struct sched_poll *s
goto out;
rc = 0;
- if ( test_bit(port, shared_info_addr(d, evtchn_pending)) )
+ if ( test_bit(port, &shared_info(d, evtchn_pending)) )
goto out;
}
diff -r 8c921adf4833 -r 42f6c206c951 xen/drivers/acpi/Makefile
--- a/xen/drivers/acpi/Makefile Fri Mar 14 15:07:45 2008 -0600
+++ b/xen/drivers/acpi/Makefile Thu Mar 20 12:35:40 2008 -0600
@@ -1,2 +1,5 @@ obj-y += tables.o
obj-y += tables.o
obj-y += numa.o
+obj-y += osl.o
+
+obj-$(x86) += hwregs.o
diff -r 8c921adf4833 -r 42f6c206c951 xen/drivers/acpi/hwregs.c
--- /dev/null Thu Jan 01 00:00:00 1970 +0000
+++ b/xen/drivers/acpi/hwregs.c Thu Mar 20 12:35:40 2008 -0600
@@ -0,0 +1,385 @@
+
+/*******************************************************************************
+ *
+ * Module Name: hwregs - Read/write access functions for the various ACPI
+ * control and status registers.
+ *
+
******************************************************************************/
+
+/*
+ * Copyright (C) 2000 - 2006, R. Byron Moore
+ * All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ * notice, this list of conditions, and the following disclaimer,
+ * without modification.
+ * 2. Redistributions in binary form must reproduce at minimum a disclaimer
+ * substantially similar to the "NO WARRANTY" disclaimer below
+ * ("Disclaimer") and any redistribution must be conditioned upon
+ * including a substantially similar Disclaimer requirement for further
+ * binary redistribution.
+ * 3. Neither the names of the above-listed copyright holders nor the names
+ * of any contributors may be used to endorse or promote products derived
+ * from this software without specific prior written permission.
+ *
+ * Alternatively, this software may be distributed under the terms of the
+ * GNU General Public License ("GPL") version 2 as published by the Free
+ * Software Foundation.
+ *
+ * NO WARRANTY
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTIBILITY AND FITNESS FOR
+ * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+ * HOLDERS OR CONTRIBUTORS BE LIABLE FOR SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
+ * STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING
+ * IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
+ * POSSIBILITY OF SUCH DAMAGES.
+ */
+
+#include <asm/io.h>
+#include <xen/config.h>
+#include <xen/init.h>
+#include <xen/types.h>
+#include <xen/errno.h>
+#include <acpi/acpi.h>
+
+#define _COMPONENT ACPI_HARDWARE
+ACPI_MODULE_NAME("hwregs")
+
+/******************************************************************************
+ *
+ * FUNCTION: acpi_hw_register_read
+ *
+ * PARAMETERS: register_id - ACPI Register ID
+ * return_value - Where the register value is returned
+ *
+ * RETURN: Status and the value read.
+ *
+ * DESCRIPTION: Read from the specified ACPI register
+ *
+
******************************************************************************/
+acpi_status
+acpi_hw_register_read(u32 register_id, u32 * return_value)
+{
+ u32 value1 = 0;
+ u32 value2 = 0;
+ acpi_status status;
+
+ ACPI_FUNCTION_TRACE(hw_register_read);
+
+ switch (register_id) {
+ case ACPI_REGISTER_PM1_STATUS: /* 16-bit access */
+
+ status =
+ acpi_hw_low_level_read(16, &value1,
+ &acpi_sinfo.pm1a_evt_blk);
+ if (ACPI_FAILURE(status)) {
+ goto exit;
+ }
+
+ /* PM1B is optional */
+
+ status =
+ acpi_hw_low_level_read(16, &value2,
+ &acpi_sinfo.pm1b_evt_blk);
+ value1 |= value2;
+ break;
+
+
+ case ACPI_REGISTER_PM1_CONTROL: /* 16-bit access */
+
+ status =
+ acpi_hw_low_level_read(16, &value1,
+ &acpi_sinfo.pm1a_cnt_blk);
+ if (ACPI_FAILURE(status)) {
+ goto exit;
+ }
+
+ status =
+ acpi_hw_low_level_read(16, &value2,
+ &acpi_sinfo.pm1b_cnt_blk);
+ value1 |= value2;
+ break;
+
+
+ default:
+ status = AE_BAD_PARAMETER;
+ break;
+ }
+
+ exit:
+
+ if (ACPI_SUCCESS(status)) {
+ *return_value = value1;
+ }
+
+ return_ACPI_STATUS(status);
+}
+
+/******************************************************************************
+ *
+ * FUNCTION: acpi_hw_register_write
+ *
+ * PARAMETERS: register_id - ACPI Register ID
+ * Value - The value to write
+ *
+ * RETURN: Status
+ *
+ * DESCRIPTION: Write to the specified ACPI register
+ *
+ * NOTE: In accordance with the ACPI specification, this function automatically
+ * preserves the value of the following bits, meaning that these bits cannot be
+ * changed via this interface:
+ *
+ * PM1_CONTROL[0] = SCI_EN
+ * PM1_CONTROL[9]
+ * PM1_STATUS[11]
+ *
+ * ACPI References:
+ * 1) Hardware Ignored Bits: When software writes to a register with ignored
+ * bit fields, it preserves the ignored bit fields
+ * 2) SCI_EN: OSPM always preserves this bit position
+ *
+
******************************************************************************/
+
+acpi_status acpi_hw_register_write(u32 register_id, u32 value)
+{
+ acpi_status status;
+ u32 read_value;
+
+ ACPI_FUNCTION_TRACE(hw_register_write);
+
+ switch (register_id) { //By now we just need handle PM1 status/PM1
control
+ case ACPI_REGISTER_PM1_STATUS: /* 16-bit access */
+
+ /* Perform a read first to preserve certain bits (per ACPI
spec) */
+
+ status = acpi_hw_register_read(ACPI_REGISTER_PM1_STATUS,
+ &read_value);
+ if (ACPI_FAILURE(status)) {
+ goto exit;
+ }
+
+ /* Insert the bits to be preserved */
+
+ ACPI_INSERT_BITS(value, ACPI_PM1_STATUS_PRESERVED_BITS,
+ read_value);
+
+ /* Now we can write the data */
+
+ status =
+ acpi_hw_low_level_write(16, value,
+ &acpi_sinfo.pm1a_evt_blk);
+ if (ACPI_FAILURE(status)) {
+ goto exit;
+ }
+
+ /* PM1B is optional */
+
+ status =
+ acpi_hw_low_level_write(16, value,
+ &acpi_sinfo.pm1b_evt_blk);
+ break;
+
+
+ case ACPI_REGISTER_PM1_CONTROL: /* 16-bit access */
+
+ /*
+ * Perform a read first to preserve certain bits (per ACPI spec)
+ *
+ * Note: This includes SCI_EN, we never want to change this bit
+ */
+ status = acpi_hw_register_read(ACPI_REGISTER_PM1_CONTROL,
+ &read_value);
+ if (ACPI_FAILURE(status)) {
+ goto exit;
+ }
+
+ /* Insert the bits to be preserved */
+
+ ACPI_INSERT_BITS(value, ACPI_PM1_CONTROL_PRESERVED_BITS,
+ read_value);
+
+ /* Now we can write the data */
+
+ status =
+ acpi_hw_low_level_write(16, value,
+ &acpi_sinfo.pm1a_cnt_blk);
+ if (ACPI_FAILURE(status)) {
+ goto exit;
+ }
+
+ status =
+ acpi_hw_low_level_write(16, value,
+ &acpi_sinfo.pm1b_cnt_blk);
+ break;
+
+ case ACPI_REGISTER_PM1A_CONTROL: /* 16-bit access */
+
+ status =
+ acpi_hw_low_level_write(16, value,
+ &acpi_sinfo.pm1a_cnt_blk);
+ break;
+
+ case ACPI_REGISTER_PM1B_CONTROL: /* 16-bit access */
+
+ status =
+ acpi_hw_low_level_write(16, value,
+ &acpi_sinfo.pm1b_cnt_blk);
+ break;
+
+
+ default:
+ status = AE_BAD_PARAMETER;
+ break;
+ }
+
+ exit:
+
+ return_ACPI_STATUS(status);
+}
+
+/******************************************************************************
+ *
+ * FUNCTION: acpi_hw_low_level_read
+ *
+ * PARAMETERS: Width - 8, 16, or 32
+ * Value - Where the value is returned
+ * Reg - GAS register structure
+ *
+ * RETURN: Status
+ *
+ * DESCRIPTION: Read from either memory or IO space.
+ *
+
******************************************************************************/
+
+acpi_status
+acpi_hw_low_level_read(u32 width, u32 * value, struct acpi_generic_address
*reg)
+{
+ u64 address;
+ acpi_status status;
+
+ ACPI_FUNCTION_NAME(hw_low_level_read);
+
+ /*
+ * Must have a valid pointer to a GAS structure, and
+ * a non-zero address within. However, don't return an error
+ * because the PM1A/B code must not fail if B isn't present.
+ */
+ if (!reg) {
+ return (AE_OK);
+ }
+
+ /* Get a local copy of the address. Handles possible alignment issues */
+
+ ACPI_MOVE_64_TO_64(&address, ®->address);
+ if (!address) {
+ return (AE_OK);
+ }
+ *value = 0;
+
+ /*
+ * Two address spaces supported: Memory or IO.
+ * PCI_Config is not supported here because the GAS struct is
insufficient
+ */
+ switch (reg->space_id) {
+ case ACPI_ADR_SPACE_SYSTEM_MEMORY:
+
+ status = acpi_os_read_memory((acpi_physical_address) address,
+ value, width);
+ break;
+
+ case ACPI_ADR_SPACE_SYSTEM_IO:
+
+ status = acpi_os_read_port((acpi_io_address) address,
+ value, width);
+ break;
+
+ default:
+
+ return (AE_BAD_PARAMETER);
+ }
+
+ ACPI_DEBUG_PRINT((ACPI_DB_IO,
+ "Read: %8.8X width %2d from %8.8X%8.8X (%s)\n",
+ *value, width,
+ ACPI_FORMAT_UINT64(address),
+ acpi_ut_get_region_name(reg->address_space_id)));
+
+ return (status);
+}
+
+/******************************************************************************
+ *
+ * FUNCTION: acpi_hw_low_level_write
+ *
+ * PARAMETERS: Width - 8, 16, or 32
+ * Value - To be written
+ * Reg - GAS register structure
+ *
+ * RETURN: Status
+ *
+ * DESCRIPTION: Write to either memory or IO space.
+ *
+
******************************************************************************/
+
+acpi_status
+acpi_hw_low_level_write(u32 width, u32 value, struct acpi_generic_address *
reg)
+{
+ u64 address;
+ acpi_status status;
+
+ ACPI_FUNCTION_NAME(hw_low_level_write);
+
+ /*
+ * Must have a valid pointer to a GAS structure, and
+ * a non-zero address within. However, don't return an error
+ * because the PM1A/B code must not fail if B isn't present.
+ */
+ if (!reg) {
+ return (AE_OK);
+ }
+
+ /* Get a local copy of the address. Handles possible alignment issues */
+
+ ACPI_MOVE_64_TO_64(&address, ®->address);
+ if (!address) {
+ return (AE_OK);
+ }
+
+ /*
+ * Two address spaces supported: Memory or IO.
+ * PCI_Config is not supported here because the GAS struct is
insufficient
+ */
+ switch (reg->space_id) {
+ case ACPI_ADR_SPACE_SYSTEM_MEMORY:
+
+ status = acpi_os_write_memory((acpi_physical_address) address,
+ value, width);
+ break;
+
+ case ACPI_ADR_SPACE_SYSTEM_IO:
+
+ status = acpi_os_write_port((acpi_io_address) address,
+ value, width);
+ break;
+
+ default:
+ return (AE_BAD_PARAMETER);
+ }
+
+ ACPI_DEBUG_PRINT((ACPI_DB_IO,
+ "Wrote: %8.8X width %2d to %8.8X%8.8X (%s)\n",
+ value, width,
+ ACPI_FORMAT_UINT64(address),
+ acpi_ut_get_region_name(reg->address_space_id)));
+
+ return (status);
+}
diff -r 8c921adf4833 -r 42f6c206c951 xen/drivers/acpi/numa.c
--- a/xen/drivers/acpi/numa.c Fri Mar 14 15:07:45 2008 -0600
+++ b/xen/drivers/acpi/numa.c Thu Mar 20 12:35:40 2008 -0600
@@ -104,7 +104,7 @@ static int __init acpi_parse_slit(unsign
slit = (struct acpi_table_slit *)__acpi_map_table(phys_addr, size);
/* downcast just for %llu vs %lu for i386/ia64 */
- localities = (u32) slit->localities;
+ localities = (u32) slit->locality_count;
acpi_numa_slit_init(slit);
diff -r 8c921adf4833 -r 42f6c206c951 xen/drivers/acpi/osl.c
--- /dev/null Thu Jan 01 00:00:00 1970 +0000
+++ b/xen/drivers/acpi/osl.c Thu Mar 20 12:35:40 2008 -0600
@@ -0,0 +1,183 @@
+/*
+ * acpi_osl.c - OS-dependent functions ($Revision: 83 $)
+ *
+ * Copyright (C) 2000 Andrew Henroid
+ * Copyright (C) 2001, 2002 Andy Grover <andrew.grover@xxxxxxxxx>
+ * Copyright (C) 2001, 2002 Paul Diefenbaugh <paul.s.diefenbaugh@xxxxxxxxx>
+ *
+ * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+ *
+ * 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, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ *
+ * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+ *
+ */
+#include <asm/io.h>
+#include <xen/config.h>
+#include <xen/init.h>
+#include <xen/types.h>
+#include <xen/errno.h>
+#include <xen/acpi.h>
+#include <xen/numa.h>
+#include <acpi/acpi_bus.h>
+#include <acpi/acmacros.h>
+#include <acpi/acpiosxf.h>
+#include <acpi/platform/aclinux.h>
+#include <xen/spinlock.h>
+#include <xen/domain_page.h>
+
+#define _COMPONENT ACPI_OS_SERVICES
+ACPI_MODULE_NAME("osl")
+#define PREFIX "ACPI: "
+struct acpi_os_dpc {
+ acpi_osd_exec_callback function;
+ void *context;
+};
+
+#ifdef CONFIG_ACPI_CUSTOM_DSDT
+#include CONFIG_ACPI_CUSTOM_DSDT_FILE
+#endif
+
+#ifdef ENABLE_DEBUGGER
+#include <linux/kdb.h>
+
+/* stuff for debugger support */
+int acpi_in_debugger;
+EXPORT_SYMBOL(acpi_in_debugger);
+
+extern char line_buf[80];
+#endif /*ENABLE_DEBUGGER */
+
+int acpi_specific_hotkey_enabled = TRUE;
+EXPORT_SYMBOL(acpi_specific_hotkey_enabled);
+
+
+acpi_status acpi_os_read_port(acpi_io_address port, u32 * value, u32 width)
+{
+ u32 dummy;
+
+ if (!value)
+ value = &dummy;
+
+ *value = 0;
+ if (width <= 8) {
+ *(u8 *) value = inb(port);
+ } else if (width <= 16) {
+ *(u16 *) value = inw(port);
+ } else if (width <= 32) {
+ *(u32 *) value = inl(port);
+ } else {
+ BUG();
+ }
+
+ return AE_OK;
+}
+
+EXPORT_SYMBOL(acpi_os_read_port);
+
+acpi_status acpi_os_write_port(acpi_io_address port, u32 value, u32 width)
+{
+ if (width <= 8) {
+ outb(value, port);
+ } else if (width <= 16) {
+ outw(value, port);
+ } else if (width <= 32) {
+ outl(value, port);
+ } else {
+ BUG();
+ }
+
+ return AE_OK;
+}
+
+EXPORT_SYMBOL(acpi_os_write_port);
+
+acpi_status
+acpi_os_read_memory(acpi_physical_address phys_addr, u32 * value, u32 width)
+{
+ u32 dummy;
+ void __iomem *virt_addr;
+
+ virt_addr = map_domain_page(phys_addr>>PAGE_SHIFT);
+ if (!value)
+ value = &dummy;
+
+ switch (width) {
+ case 8:
+ *(u8 *) value = readb(virt_addr);
+ break;
+ case 16:
+ *(u16 *) value = readw(virt_addr);
+ break;
+ case 32:
+ *(u32 *) value = readl(virt_addr);
+ break;
+ default:
+ BUG();
+ }
+
+ unmap_domain_page(virt_addr);
+
+ return AE_OK;
+}
+
+acpi_status
+acpi_os_write_memory(acpi_physical_address phys_addr, u32 value, u32 width)
+{
+ void __iomem *virt_addr;
+
+ virt_addr = map_domain_page(phys_addr>>PAGE_SHIFT);
+
+ switch (width) {
+ case 8:
+ writeb(value, virt_addr);
+ break;
+ case 16:
+ writew(value, virt_addr);
+ break;
+ case 32:
+ writel(value, virt_addr);
+ break;
+ default:
+ BUG();
+ }
+
+ unmap_domain_page(virt_addr);
+
+ return AE_OK;
+}
+
+/*
+ * Acquire a spinlock.
+ *
+ * handle is a pointer to the spinlock_t.
+ */
+
+acpi_cpu_flags acpi_os_acquire_lock(acpi_spinlock lockp)
+{
+ acpi_cpu_flags flags;
+ spin_lock_irqsave(lockp, flags);
+ return flags;
+}
+
+/*
+ * Release a spinlock. See above.
+ */
+
+void acpi_os_release_lock(acpi_spinlock lockp, acpi_cpu_flags flags)
+{
+ spin_unlock_irqrestore(lockp, flags);
+}
+
diff -r 8c921adf4833 -r 42f6c206c951 xen/drivers/acpi/tables.c
--- a/xen/drivers/acpi/tables.c Fri Mar 14 15:07:45 2008 -0600
+++ b/xen/drivers/acpi/tables.c Thu Mar 20 12:35:40 2008 -0600
@@ -295,15 +295,15 @@ acpi_get_table_header_early(enum acpi_ta
/* Map the DSDT header via the pointer in the FADT */
if (id == ACPI_DSDT) {
- struct fadt_descriptor_rev2 *fadt =
- (struct fadt_descriptor_rev2 *)*header;
-
- if (fadt->revision == 3 && fadt->Xdsdt) {
+ struct acpi_table_fadt *fadt =
+ (struct acpi_table_fadt *)*header;
+
+ if (fadt->header.revision == 3 && fadt->Xdsdt) {
*header = (void *)__acpi_map_table(fadt->Xdsdt,
sizeof(struct
acpi_table_header));
- } else if (fadt->V1_dsdt) {
- *header = (void *)__acpi_map_table(fadt->V1_dsdt,
+ } else if (fadt->dsdt) {
+ *header = (void *)__acpi_map_table(fadt->dsdt,
sizeof(struct
acpi_table_header));
} else
@@ -424,12 +424,11 @@ static int __init acpi_table_get_sdt(str
/* First check XSDT (but only on ACPI 2.0-compatible systems) */
- if ((rsdp->revision >= 2) &&
- (((struct acpi20_table_rsdp *)rsdp)->xsdt_address)) {
+ if ((rsdp->revision >= 2) && rsdp->xsdt_physical_address) {
struct acpi_table_xsdt *mapped_xsdt = NULL;
- sdt_pa = ((struct acpi20_table_rsdp *)rsdp)->xsdt_address;
+ sdt_pa = rsdp->xsdt_physical_address;
/* map in just the header */
header = (struct acpi_table_header *)
@@ -471,16 +470,16 @@ static int __init acpi_table_get_sdt(str
}
for (i = 0; i < sdt_count; i++)
- sdt_entry[i].pa = (unsigned long)mapped_xsdt->entry[i];
+ sdt_entry[i].pa = (unsigned
long)mapped_xsdt->table_offset_entry[i];
}
/* Then check RSDT */
- else if (rsdp->rsdt_address) {
+ else if (rsdp->rsdt_physical_address) {
struct acpi_table_rsdt *mapped_rsdt = NULL;
- sdt_pa = rsdp->rsdt_address;
+ sdt_pa = rsdp->rsdt_physical_address;
/* map in just the header */
header = (struct acpi_table_header *)
@@ -521,7 +520,7 @@ static int __init acpi_table_get_sdt(str
}
for (i = 0; i < sdt_count; i++)
- sdt_entry[i].pa = (unsigned long)mapped_rsdt->entry[i];
+ sdt_entry[i].pa = (unsigned
long)mapped_rsdt->table_offset_entry[i];
}
else {
@@ -613,13 +612,10 @@ int __init acpi_table_init(void)
if (rsdp->revision < 2)
result =
- acpi_table_compute_checksum(rsdp,
- sizeof(struct acpi_table_rsdp));
+ acpi_table_compute_checksum(rsdp, 20);
else
result =
- acpi_table_compute_checksum(rsdp,
- ((struct acpi20_table_rsdp *)
- rsdp)->length);
+ acpi_table_compute_checksum(rsdp, rsdp->length);
if (result) {
printk(KERN_WARNING " >>> ERROR: Invalid checksum\n");
@@ -663,12 +659,11 @@ acpi_table_disable(enum acpi_table_id ta
/* First check XSDT (but only on ACPI 2.0-compatible systems) */
- if ((rsdp->revision >= 2) &&
- (((struct acpi20_table_rsdp *)rsdp)->xsdt_address)) {
+ if ((rsdp->revision >= 2) && rsdp->xsdt_physical_address) {
struct acpi_table_xsdt *mapped_xsdt = NULL;
- sdt_pa = ((struct acpi20_table_rsdp *)rsdp)->xsdt_address;
+ sdt_pa = rsdp->xsdt_physical_address;
/* map in just the header */
header = (struct acpi_table_header *)
@@ -702,7 +697,7 @@ acpi_table_disable(enum acpi_table_id ta
if (id < sdt_count) {
header = (struct acpi_table_header *)
- __acpi_map_table(mapped_xsdt->entry[id],
sizeof(struct acpi_table_header));
+
__acpi_map_table(mapped_xsdt->table_offset_entry[id], sizeof(struct
acpi_table_header));
} else {
printk(KERN_WARNING PREFIX
"Unable to disable entry %d\n",
@@ -713,11 +708,11 @@ acpi_table_disable(enum acpi_table_id ta
/* Then check RSDT */
- else if (rsdp->rsdt_address) {
+ else if (rsdp->rsdt_physical_address) {
struct acpi_table_rsdt *mapped_rsdt = NULL;
- sdt_pa = rsdp->rsdt_address;
+ sdt_pa = rsdp->rsdt_physical_address;
/* map in just the header */
header = (struct acpi_table_header *)
@@ -749,7 +744,7 @@ acpi_table_disable(enum acpi_table_id ta
}
if (id < sdt_count) {
header = (struct acpi_table_header *)
- __acpi_map_table(mapped_rsdt->entry[id],
sizeof(struct acpi_table_header));
+
__acpi_map_table(mapped_rsdt->table_offset_entry[id], sizeof(struct
acpi_table_header));
} else {
printk(KERN_WARNING PREFIX
"Unable to disable entry %d\n",
diff -r 8c921adf4833 -r 42f6c206c951 xen/drivers/passthrough/Makefile
--- a/xen/drivers/passthrough/Makefile Fri Mar 14 15:07:45 2008 -0600
+++ b/xen/drivers/passthrough/Makefile Thu Mar 20 12:35:40 2008 -0600
@@ -1,2 +1,5 @@ subdir-$(x86) += vtd
subdir-$(x86) += vtd
subdir-$(x86) += amd
+
+obj-y += iommu.o
+obj-y += io.o
diff -r 8c921adf4833 -r 42f6c206c951 xen/drivers/passthrough/amd/iommu_acpi.c
--- a/xen/drivers/passthrough/amd/iommu_acpi.c Fri Mar 14 15:07:45 2008 -0600
+++ b/xen/drivers/passthrough/amd/iommu_acpi.c Thu Mar 20 12:35:40 2008 -0600
@@ -29,12 +29,12 @@ extern struct ivrs_mappings *ivrs_mappin
extern struct ivrs_mappings *ivrs_mappings;
static struct amd_iommu * __init find_iommu_from_bdf_cap(
- u16 bdf, u8 cap_offset)
+ u16 bdf, u8 cap_offset)
{
struct amd_iommu *iommu;
- for_each_amd_iommu( iommu )
- if ( iommu->bdf == bdf && iommu->cap_offset == cap_offset )
+ for_each_amd_iommu ( iommu )
+ if ( (iommu->bdf == bdf) && (iommu->cap_offset == cap_offset) )
return iommu;
return NULL;
@@ -57,15 +57,17 @@ static void __init reserve_iommu_exclusi
iommu->exclusion_limit = limit;
}
-static void __init reserve_iommu_exclusion_range_all(struct amd_iommu *iommu,
- unsigned long base, unsigned long limit)
+static void __init reserve_iommu_exclusion_range_all(
+ struct amd_iommu *iommu,
+ unsigned long base, unsigned long limit)
{
reserve_iommu_exclusion_range(iommu, base, limit);
iommu->exclusion_allow_all = IOMMU_CONTROL_ENABLED;
}
-static void __init reserve_unity_map_for_device(u16 bdf, unsigned long base,
- unsigned long length, u8 iw, u8 ir)
+static void __init reserve_unity_map_for_device(
+ u16 bdf, unsigned long base,
+ unsigned long length, u8 iw, u8 ir)
{
unsigned long old_top, new_top;
@@ -80,7 +82,7 @@ static void __init reserve_unity_map_for
if ( ivrs_mappings[bdf].addr_range_start < base )
base = ivrs_mappings[bdf].addr_range_start;
length = new_top - base;
- }
+ }
/* extend r/w permissioms and keep aggregate */
if ( iw )
@@ -93,7 +95,7 @@ static void __init reserve_unity_map_for
}
static int __init register_exclusion_range_for_all_devices(
- unsigned long base, unsigned long limit, u8 iw, u8 ir)
+ unsigned long base, unsigned long limit, u8 iw, u8 ir)
{
unsigned long range_top, iommu_top, length;
struct amd_iommu *iommu;
@@ -105,7 +107,7 @@ static int __init register_exclusion_ran
iommu_top = max_page * PAGE_SIZE;
if ( base < iommu_top )
{
- if (range_top > iommu_top)
+ if ( range_top > iommu_top )
range_top = iommu_top;
length = range_top - base;
/* reserve r/w unity-mapped page entries for devices */
@@ -116,7 +118,7 @@ static int __init register_exclusion_ran
base = iommu_top;
}
/* register IOMMU exclusion range settings */
- if (limit >= iommu_top)
+ if ( limit >= iommu_top )
{
for_each_amd_iommu( iommu )
reserve_iommu_exclusion_range_all(iommu, base, limit);
@@ -125,8 +127,8 @@ static int __init register_exclusion_ran
return 0;
}
-static int __init register_exclusion_range_for_device(u16 bdf,
- unsigned long base, unsigned long limit, u8 iw, u8 ir)
+static int __init register_exclusion_range_for_device(
+ u16 bdf, unsigned long base, unsigned long limit, u8 iw, u8 ir)
{
unsigned long range_top, iommu_top, length;
struct amd_iommu *iommu;
@@ -147,7 +149,7 @@ static int __init register_exclusion_ran
iommu_top = max_page * PAGE_SIZE;
if ( base < iommu_top )
{
- if (range_top > iommu_top)
+ if ( range_top > iommu_top )
range_top = iommu_top;
length = range_top - base;
/* reserve unity-mapped page entries for device */
@@ -159,8 +161,8 @@ static int __init register_exclusion_ran
base = iommu_top;
}
- /* register IOMMU exclusion range settings for device */
- if ( limit >= iommu_top )
+ /* register IOMMU exclusion range settings for device */
+ if ( limit >= iommu_top )
{
reserve_iommu_exclusion_range(iommu, base, limit);
ivrs_mappings[bdf].dte_allow_exclusion = IOMMU_CONTROL_ENABLED;
@@ -171,8 +173,8 @@ static int __init register_exclusion_ran
}
static int __init register_exclusion_range_for_iommu_devices(
- struct amd_iommu *iommu,
- unsigned long base, unsigned long limit, u8 iw, u8 ir)
+ struct amd_iommu *iommu,
+ unsigned long base, unsigned long limit, u8 iw, u8 ir)
{
unsigned long range_top, iommu_top, length;
u16 bus, devfn, bdf, req;
@@ -183,7 +185,7 @@ static int __init register_exclusion_ran
iommu_top = max_page * PAGE_SIZE;
if ( base < iommu_top )
{
- if (range_top > iommu_top)
+ if ( range_top > iommu_top )
range_top = iommu_top;
length = range_top - base;
/* reserve r/w unity-mapped page entries for devices */
@@ -205,19 +207,19 @@ static int __init register_exclusion_ran
}
/* register IOMMU exclusion range settings */
- if (limit >= iommu_top)
+ if ( limit >= iommu_top )
reserve_iommu_exclusion_range_all(iommu, base, limit);
return 0;
}
static int __init parse_ivmd_device_select(
- struct acpi_ivmd_block_header *ivmd_block,
- unsigned long base, unsigned long limit, u8 iw, u8 ir)
+ struct acpi_ivmd_block_header *ivmd_block,
+ unsigned long base, unsigned long limit, u8 iw, u8 ir)
{
u16 bdf;
bdf = ivmd_block->header.dev_id;
- if (bdf >= ivrs_bdf_entries)
+ if ( bdf >= ivrs_bdf_entries )
{
dprintk(XENLOG_ERR, "IVMD Error: Invalid Dev_Id 0x%x\n", bdf);
return -ENODEV;
@@ -227,44 +229,41 @@ static int __init parse_ivmd_device_sele
}
static int __init parse_ivmd_device_range(
- struct acpi_ivmd_block_header *ivmd_block,
- unsigned long base, unsigned long limit, u8 iw, u8 ir)
+ struct acpi_ivmd_block_header *ivmd_block,
+ unsigned long base, unsigned long limit, u8 iw, u8 ir)
{
u16 first_bdf, last_bdf, bdf;
int error;
first_bdf = ivmd_block->header.dev_id;
- if (first_bdf >= ivrs_bdf_entries)
- {
- dprintk(XENLOG_ERR, "IVMD Error: "
- "Invalid Range_First Dev_Id 0x%x\n", first_bdf);
- return -ENODEV;
+ if ( first_bdf >= ivrs_bdf_entries )
+ {
+ dprintk(XENLOG_ERR, "IVMD Error: "
+ "Invalid Range_First Dev_Id 0x%x\n", first_bdf);
+ return -ENODEV;
}
last_bdf = ivmd_block->last_dev_id;
- if (last_bdf >= ivrs_bdf_entries || last_bdf <= first_bdf)
+ if ( (last_bdf >= ivrs_bdf_entries) || (last_bdf <= first_bdf) )
{
dprintk(XENLOG_ERR, "IVMD Error: "
- "Invalid Range_Last Dev_Id 0x%x\n", last_bdf);
- return -ENODEV;
- }
-
- dprintk(XENLOG_ERR, " Dev_Id Range: 0x%x -> 0x%x\n",
- first_bdf, last_bdf);
-
- for ( bdf = first_bdf, error = 0;
- bdf <= last_bdf && !error; ++bdf )
- {
- error = register_exclusion_range_for_device(
- bdf, base, limit, iw, ir);
- }
-
- return error;
+ "Invalid Range_Last Dev_Id 0x%x\n", last_bdf);
+ return -ENODEV;
+ }
+
+ dprintk(XENLOG_ERR, " Dev_Id Range: 0x%x -> 0x%x\n",
+ first_bdf, last_bdf);
+
+ for ( bdf = first_bdf, error = 0; (bdf <= last_bdf) && !error; bdf++ )
+ error = register_exclusion_range_for_device(
+ bdf, base, limit, iw, ir);
+
+ return error;
}
static int __init parse_ivmd_device_iommu(
- struct acpi_ivmd_block_header *ivmd_block,
- unsigned long base, unsigned long limit, u8 iw, u8 ir)
+ struct acpi_ivmd_block_header *ivmd_block,
+ unsigned long base, unsigned long limit, u8 iw, u8 ir)
{
struct amd_iommu *iommu;
@@ -273,14 +272,14 @@ static int __init parse_ivmd_device_iomm
ivmd_block->cap_offset);
if ( !iommu )
{
- dprintk(XENLOG_ERR,
- "IVMD Error: No IOMMU for Dev_Id 0x%x Cap 0x%x\n",
- ivmd_block->header.dev_id, ivmd_block->cap_offset);
- return -ENODEV;
+ dprintk(XENLOG_ERR,
+ "IVMD Error: No IOMMU for Dev_Id 0x%x Cap 0x%x\n",
+ ivmd_block->header.dev_id, ivmd_block->cap_offset);
+ return -ENODEV;
}
return register_exclusion_range_for_iommu_devices(
- iommu, base, limit, iw, ir);
+ iommu, base, limit, iw, ir);
}
static int __init parse_ivmd_block(struct acpi_ivmd_block_header *ivmd_block)
@@ -288,11 +287,11 @@ static int __init parse_ivmd_block(struc
unsigned long start_addr, mem_length, base, limit;
u8 iw, ir;
- if (ivmd_block->header.length <
- sizeof(struct acpi_ivmd_block_header))
- {
- dprintk(XENLOG_ERR, "IVMD Error: Invalid Block Length!\n");
- return -ENODEV;
+ if ( ivmd_block->header.length <
+ sizeof(struct acpi_ivmd_block_header) )
+ {
+ dprintk(XENLOG_ERR, "IVMD Error: Invalid Block Length!\n");
+ return -ENODEV;
}
start_addr = (unsigned long)ivmd_block->start_addr;
@@ -301,7 +300,7 @@ static int __init parse_ivmd_block(struc
limit = (start_addr + mem_length - 1) & PAGE_MASK;
dprintk(XENLOG_INFO, "IVMD Block: Type 0x%x\n",
- ivmd_block->header.type);
+ ivmd_block->header.type);
dprintk(XENLOG_INFO, " Start_Addr_Phys 0x%lx\n", start_addr);
dprintk(XENLOG_INFO, " Mem_Length 0x%lx\n", mem_length);
@@ -322,27 +321,27 @@ static int __init parse_ivmd_block(struc
}
else
{
- dprintk(KERN_ERR, "IVMD Error: Invalid Flag Field!\n");
- return -ENODEV;
+ dprintk(KERN_ERR, "IVMD Error: Invalid Flag Field!\n");
+ return -ENODEV;
}
switch( ivmd_block->header.type )
{
case AMD_IOMMU_ACPI_IVMD_ALL_TYPE:
return register_exclusion_range_for_all_devices(
- base, limit, iw, ir);
+ base, limit, iw, ir);
case AMD_IOMMU_ACPI_IVMD_ONE_TYPE:
return parse_ivmd_device_select(ivmd_block,
- base, limit, iw, ir);
+ base, limit, iw, ir);
case AMD_IOMMU_ACPI_IVMD_RANGE_TYPE:
return parse_ivmd_device_range(ivmd_block,
- base, limit, iw, ir);
+ base, limit, iw, ir);
case AMD_IOMMU_ACPI_IVMD_IOMMU_TYPE:
return parse_ivmd_device_iommu(ivmd_block,
- base, limit, iw, ir);
+ base, limit, iw, ir);
default:
dprintk(XENLOG_ERR, "IVMD Error: Invalid Block Type!\n");
@@ -350,8 +349,8 @@ static int __init parse_ivmd_block(struc
}
}
-static u16 __init parse_ivhd_device_padding(u16 pad_length,
- u16 header_length, u16 block_length)
+static u16 __init parse_ivhd_device_padding(
+ u16 pad_length, u16 header_length, u16 block_length)
{
if ( header_length < (block_length + pad_length) )
{
@@ -363,7 +362,7 @@ static u16 __init parse_ivhd_device_padd
}
static u16 __init parse_ivhd_device_select(
- union acpi_ivhd_device *ivhd_device)
+ union acpi_ivhd_device *ivhd_device)
{
u16 bdf;
@@ -385,8 +384,8 @@ static u16 __init parse_ivhd_device_sele
}
static u16 __init parse_ivhd_device_range(
- union acpi_ivhd_device *ivhd_device,
- u16 header_length, u16 block_length)
+ union acpi_ivhd_device *ivhd_device,
+ u16 header_length, u16 block_length)
{
u16 dev_length, first_bdf, last_bdf, bdf;
u8 sys_mgt;
@@ -399,7 +398,8 @@ static u16 __init parse_ivhd_device_rang
}
if ( ivhd_device->range.trailer.type !=
- AMD_IOMMU_ACPI_IVHD_DEV_RANGE_END) {
+ AMD_IOMMU_ACPI_IVHD_DEV_RANGE_END )
+ {
dprintk(XENLOG_ERR, "IVHD Error: "
"Invalid Range: End_Type 0x%x\n",
ivhd_device->range.trailer.type);
@@ -409,35 +409,35 @@ static u16 __init parse_ivhd_device_rang
first_bdf = ivhd_device->header.dev_id;
if ( first_bdf >= ivrs_bdf_entries )
{
- dprintk(XENLOG_ERR, "IVHD Error: "
- "Invalid Range: First Dev_Id 0x%x\n", first_bdf);
- return 0;
+ dprintk(XENLOG_ERR, "IVHD Error: "
+ "Invalid Range: First Dev_Id 0x%x\n", first_bdf);
+ return 0;
}
last_bdf = ivhd_device->range.trailer.dev_id;
- if ( last_bdf >= ivrs_bdf_entries || last_bdf <= first_bdf )
- {
- dprintk(XENLOG_ERR, "IVHD Error: "
- "Invalid Range: Last Dev_Id 0x%x\n", last_bdf);
- return 0;
+ if ( (last_bdf >= ivrs_bdf_entries) || (last_bdf <= first_bdf) )
+ {
+ dprintk(XENLOG_ERR, "IVHD Error: "
+ "Invalid Range: Last Dev_Id 0x%x\n", last_bdf);
+ return 0;
}
dprintk(XENLOG_INFO, " Dev_Id Range: 0x%x -> 0x%x\n",
- first_bdf, last_bdf);
+ first_bdf, last_bdf);
/* override flags for range of devices */
sys_mgt = get_field_from_byte(ivhd_device->header.flags,
- AMD_IOMMU_ACPI_SYS_MGT_MASK,
- AMD_IOMMU_ACPI_SYS_MGT_SHIFT);
- for ( bdf = first_bdf; bdf <= last_bdf; ++bdf )
+ AMD_IOMMU_ACPI_SYS_MGT_MASK,
+ AMD_IOMMU_ACPI_SYS_MGT_SHIFT);
+ for ( bdf = first_bdf; bdf <= last_bdf; bdf++ )
ivrs_mappings[bdf].dte_sys_mgt_enable = sys_mgt;
return dev_length;
}
static u16 __init parse_ivhd_device_alias(
- union acpi_ivhd_device *ivhd_device,
- u16 header_length, u16 block_length)
+ union acpi_ivhd_device *ivhd_device,
+ u16 header_length, u16 block_length)
{
u16 dev_length, alias_id, bdf;
@@ -445,7 +445,7 @@ static u16 __init parse_ivhd_device_alia
if ( header_length < (block_length + dev_length) )
{
dprintk(XENLOG_ERR, "IVHD Error: "
- "Invalid Device_Entry Length!\n");
+ "Invalid Device_Entry Length!\n");
return 0;
}
@@ -460,9 +460,9 @@ static u16 __init parse_ivhd_device_alia
alias_id = ivhd_device->alias.dev_id;
if ( alias_id >= ivrs_bdf_entries )
{
- dprintk(XENLOG_ERR, "IVHD Error: "
- "Invalid Alias Dev_Id 0x%x\n", alias_id);
- return 0;
+ dprintk(XENLOG_ERR, "IVHD Error: "
+ "Invalid Alias Dev_Id 0x%x\n", alias_id);
+ return 0;
}
dprintk(XENLOG_INFO, " Dev_Id Alias: 0x%x\n", alias_id);
@@ -470,18 +470,18 @@ static u16 __init parse_ivhd_device_alia
/* override requestor_id and flags for device */
ivrs_mappings[bdf].dte_requestor_id = alias_id;
ivrs_mappings[bdf].dte_sys_mgt_enable =
- get_field_from_byte(ivhd_device->header.flags,
- AMD_IOMMU_ACPI_SYS_MGT_MASK,
- AMD_IOMMU_ACPI_SYS_MGT_SHIFT);
+ get_field_from_byte(ivhd_device->header.flags,
+ AMD_IOMMU_ACPI_SYS_MGT_MASK,
+ AMD_IOMMU_ACPI_SYS_MGT_SHIFT);
ivrs_mappings[alias_id].dte_sys_mgt_enable =
- ivrs_mappings[bdf].dte_sys_mgt_enable;
+ ivrs_mappings[bdf].dte_sys_mgt_enable;
return dev_length;
}
static u16 __init parse_ivhd_device_alias_range(
- union acpi_ivhd_device *ivhd_device,
- u16 header_length, u16 block_length)
+ union acpi_ivhd_device *ivhd_device,
+ u16 header_length, u16 block_length)
{
u16 dev_length, first_bdf, last_bdf, alias_id, bdf;
@@ -496,7 +496,7 @@ static u16 __init parse_ivhd_device_alia
}
if ( ivhd_device->alias_range.trailer.type !=
- AMD_IOMMU_ACPI_IVHD_DEV_RANGE_END )
+ AMD_IOMMU_ACPI_IVHD_DEV_RANGE_END )
{
dprintk(XENLOG_ERR, "IVHD Error: "
"Invalid Range: End_Type 0x%x\n",
@@ -536,7 +536,7 @@ static u16 __init parse_ivhd_device_alia
sys_mgt = get_field_from_byte(ivhd_device->header.flags,
AMD_IOMMU_ACPI_SYS_MGT_MASK,
AMD_IOMMU_ACPI_SYS_MGT_SHIFT);
- for ( bdf = first_bdf; bdf <= last_bdf; ++bdf )
+ for ( bdf = first_bdf; bdf <= last_bdf; bdf++ )
{
ivrs_mappings[bdf].dte_requestor_id = alias_id;
ivrs_mappings[bdf].dte_sys_mgt_enable = sys_mgt;
@@ -547,8 +547,8 @@ static u16 __init parse_ivhd_device_alia
}
static u16 __init parse_ivhd_device_extended(
- union acpi_ivhd_device *ivhd_device,
- u16 header_length, u16 block_length)
+ union acpi_ivhd_device *ivhd_device,
+ u16 header_length, u16 block_length)
{
u16 dev_length, bdf;
@@ -578,8 +578,8 @@ static u16 __init parse_ivhd_device_exte
}
static u16 __init parse_ivhd_device_extended_range(
- union acpi_ivhd_device *ivhd_device,
- u16 header_length, u16 block_length)
+ union acpi_ivhd_device *ivhd_device,
+ u16 header_length, u16 block_length)
{
u16 dev_length, first_bdf, last_bdf, bdf;
u8 sys_mgt;
@@ -593,7 +593,7 @@ static u16 __init parse_ivhd_device_exte
}
if ( ivhd_device->extended_range.trailer.type !=
- AMD_IOMMU_ACPI_IVHD_DEV_RANGE_END )
+ AMD_IOMMU_ACPI_IVHD_DEV_RANGE_END )
{
dprintk(XENLOG_ERR, "IVHD Error: "
"Invalid Range: End_Type 0x%x\n",
@@ -604,13 +604,13 @@ static u16 __init parse_ivhd_device_exte
first_bdf = ivhd_device->header.dev_id;
if ( first_bdf >= ivrs_bdf_entries )
{
- dprintk(XENLOG_ERR, "IVHD Error: "
- "Invalid Range: First Dev_Id 0x%x\n", first_bdf);
- return 0;
+ dprintk(XENLOG_ERR, "IVHD Error: "
+ "Invalid Range: First Dev_Id 0x%x\n", first_bdf);
+ return 0;
}
last_bdf = ivhd_device->extended_range.trailer.dev_id;
- if ( last_bdf >= ivrs_bdf_entries || last_bdf <= first_bdf )
+ if ( (last_bdf >= ivrs_bdf_entries) || (last_bdf <= first_bdf) )
{
dprintk(XENLOG_ERR, "IVHD Error: "
"Invalid Range: Last Dev_Id 0x%x\n", last_bdf);
@@ -624,7 +624,7 @@ static u16 __init parse_ivhd_device_exte
sys_mgt = get_field_from_byte(ivhd_device->header.flags,
AMD_IOMMU_ACPI_SYS_MGT_MASK,
AMD_IOMMU_ACPI_SYS_MGT_SHIFT);
- for ( bdf = first_bdf; bdf <= last_bdf; ++bdf )
+ for ( bdf = first_bdf; bdf <= last_bdf; bdf++ )
ivrs_mappings[bdf].dte_sys_mgt_enable = sys_mgt;
return dev_length;
@@ -637,20 +637,20 @@ static int __init parse_ivhd_block(struc
struct amd_iommu *iommu;
if ( ivhd_block->header.length <
- sizeof(struct acpi_ivhd_block_header) )
+ sizeof(struct acpi_ivhd_block_header) )
{
dprintk(XENLOG_ERR, "IVHD Error: Invalid Block Length!\n");
return -ENODEV;
}
iommu = find_iommu_from_bdf_cap(ivhd_block->header.dev_id,
- ivhd_block->cap_offset);
+ ivhd_block->cap_offset);
if ( !iommu )
{
dprintk(XENLOG_ERR,
"IVHD Error: No IOMMU for Dev_Id 0x%x Cap 0x%x\n",
ivhd_block->header.dev_id, ivhd_block->cap_offset);
- return -ENODEV;
+ return -ENODEV;
}
dprintk(XENLOG_INFO, "IVHD Block:\n");
@@ -668,29 +668,29 @@ static int __init parse_ivhd_block(struc
AMD_IOMMU_ACPI_COHERENT_MASK,
AMD_IOMMU_ACPI_COHERENT_SHIFT);
iommu->iotlb_support = get_field_from_byte(ivhd_block->header.flags,
- AMD_IOMMU_ACPI_IOTLB_SUP_MASK,
- AMD_IOMMU_ACPI_IOTLB_SUP_SHIFT);
+ AMD_IOMMU_ACPI_IOTLB_SUP_MASK,
+ AMD_IOMMU_ACPI_IOTLB_SUP_SHIFT);
iommu->isochronous = get_field_from_byte(ivhd_block->header.flags,
- AMD_IOMMU_ACPI_ISOC_MASK,
- AMD_IOMMU_ACPI_ISOC_SHIFT);
+ AMD_IOMMU_ACPI_ISOC_MASK,
+ AMD_IOMMU_ACPI_ISOC_SHIFT);
iommu->res_pass_pw = get_field_from_byte(ivhd_block->header.flags,
- AMD_IOMMU_ACPI_RES_PASS_PW_MASK,
- AMD_IOMMU_ACPI_RES_PASS_PW_SHIFT);
+ AMD_IOMMU_ACPI_RES_PASS_PW_MASK,
+ AMD_IOMMU_ACPI_RES_PASS_PW_SHIFT);
iommu->pass_pw = get_field_from_byte(ivhd_block->header.flags,
- AMD_IOMMU_ACPI_PASS_PW_MASK,
- AMD_IOMMU_ACPI_PASS_PW_SHIFT);
+ AMD_IOMMU_ACPI_PASS_PW_MASK,
+ AMD_IOMMU_ACPI_PASS_PW_SHIFT);
iommu->ht_tunnel_enable = get_field_from_byte(
- ivhd_block->header.flags,
- AMD_IOMMU_ACPI_HT_TUN_ENB_MASK,
- AMD_IOMMU_ACPI_HT_TUN_ENB_SHIFT);
+ ivhd_block->header.flags,
+ AMD_IOMMU_ACPI_HT_TUN_ENB_MASK,
+ AMD_IOMMU_ACPI_HT_TUN_ENB_SHIFT);
/* parse Device Entries */
block_length = sizeof(struct acpi_ivhd_block_header);
- while( ivhd_block->header.length >=
- (block_length + sizeof(struct acpi_ivhd_device_header)) )
+ while ( ivhd_block->header.length >=
+ (block_length + sizeof(struct acpi_ivhd_device_header)) )
{
ivhd_device = (union acpi_ivhd_device *)
- ((u8 *)ivhd_block + block_length);
+ ((u8 *)ivhd_block + block_length);
dprintk(XENLOG_INFO, "IVHD Device Entry:\n");
dprintk(XENLOG_INFO, " Type 0x%x\n",
@@ -700,7 +700,7 @@ static int __init parse_ivhd_block(struc
dprintk(XENLOG_INFO, " Flags 0x%x\n",
ivhd_device->header.flags);
- switch( ivhd_device->header.type )
+ switch ( ivhd_device->header.type )
{
case AMD_IOMMU_ACPI_IVHD_DEV_U32_PAD:
dev_length = parse_ivhd_device_padding(
@@ -716,7 +716,8 @@ static int __init parse_ivhd_block(struc
dev_length = parse_ivhd_device_select(ivhd_device);
break;
case AMD_IOMMU_ACPI_IVHD_DEV_RANGE_START:
- dev_length = parse_ivhd_device_range(ivhd_device,
+ dev_length = parse_ivhd_device_range(
+ ivhd_device,
ivhd_block->header.length, block_length);
break;
case AMD_IOMMU_ACPI_IVHD_DEV_ALIAS_SELECT:
@@ -741,7 +742,7 @@ static int __init parse_ivhd_block(struc
break;
default:
dprintk(XENLOG_ERR, "IVHD Error: "
- "Invalid Device Type!\n");
+ "Invalid Device Type!\n");
dev_length = 0;
break;
}
@@ -759,7 +760,7 @@ static int __init parse_ivrs_block(struc
struct acpi_ivhd_block_header *ivhd_block;
struct acpi_ivmd_block_header *ivmd_block;
- switch(ivrs_block->type)
+ switch ( ivrs_block->type )
{
case AMD_IOMMU_ACPI_IVHD_TYPE:
ivhd_block = (struct acpi_ivhd_block_header *)ivrs_block;
@@ -786,7 +787,7 @@ void __init dump_acpi_table_header(struc
printk(XENLOG_INFO "AMD IOMMU: ACPI Table:\n");
printk(XENLOG_INFO " Signature ");
- for ( i = 0; i < ACPI_NAME_SIZE; ++i )
+ for ( i = 0; i < ACPI_NAME_SIZE; i++ )
printk("%c", table->signature[i]);
printk("\n");
@@ -795,28 +796,27 @@ void __init dump_acpi_table_header(struc
printk(" CheckSum 0x%x\n", table->checksum);
printk(" OEM_Id ");
- for ( i = 0; i < ACPI_OEM_ID_SIZE; ++i )
+ for ( i = 0; i < ACPI_OEM_ID_SIZE; i++ )
printk("%c", table->oem_id[i]);
printk("\n");
printk(" OEM_Table_Id ");
- for ( i = 0; i < ACPI_OEM_TABLE_ID_SIZE; ++i )
+ for ( i = 0; i < ACPI_OEM_TABLE_ID_SIZE; i++ )
printk("%c", table->oem_table_id[i]);
printk("\n");
printk(" OEM_Revision 0x%x\n", table->oem_revision);
printk(" Creator_Id ");
- for ( i = 0; i < ACPI_NAME_SIZE; ++i )
+ for ( i = 0; i < ACPI_NAME_SIZE; i++ )
printk("%c", table->asl_compiler_id[i]);
printk("\n");
printk(" Creator_Revision 0x%x\n",
- table->asl_compiler_revision);
-}
-
-int __init parse_ivrs_table(unsigned long phys_addr,
- unsigned long size)
+ table->asl_compiler_revision);
+}
+
+int __init parse_ivrs_table(unsigned long phys_addr, unsigned long size)
{
struct acpi_ivrs_block_header *ivrs_block;
unsigned long length, i;
@@ -834,7 +834,7 @@ int __init parse_ivrs_table(unsigned lon
/* validate checksum: sum of entire table == 0 */
checksum = 0;
raw_table = (u8 *)table;
- for ( i = 0; i < table->length; ++i )
+ for ( i = 0; i < table->length; i++ )
checksum += raw_table[i];
if ( checksum )
{
@@ -845,11 +845,10 @@ int __init parse_ivrs_table(unsigned lon
/* parse IVRS blocks */
length = sizeof(struct acpi_ivrs_table_header);
- while( error == 0 && table->length >
- (length + sizeof(struct acpi_ivrs_block_header)) )
+ while ( (error == 0) && (table->length > (length + sizeof(*ivrs_block))) )
{
ivrs_block = (struct acpi_ivrs_block_header *)
- ((u8 *)table + length);
+ ((u8 *)table + length);
dprintk(XENLOG_INFO, "IVRS Block:\n");
dprintk(XENLOG_INFO, " Type 0x%x\n", ivrs_block->type);
@@ -857,16 +856,16 @@ int __init parse_ivrs_table(unsigned lon
dprintk(XENLOG_INFO, " Length 0x%x\n", ivrs_block->length);
dprintk(XENLOG_INFO, " Dev_Id 0x%x\n", ivrs_block->dev_id);
- if (table->length >= (length + ivrs_block->length))
- error = parse_ivrs_block(ivrs_block);
- else
+ if ( table->length < (length + ivrs_block->length) )
{
- dprintk(XENLOG_ERR, "IVRS Error: "
- "Table Length Exceeded: 0x%x -> 0x%lx\n",
- table->length,
- (length + ivrs_block->length));
- return -ENODEV;
+ dprintk(XENLOG_ERR, "IVRS Error: "
+ "Table Length Exceeded: 0x%x -> 0x%lx\n",
+ table->length,
+ (length + ivrs_block->length));
+ return -ENODEV;
}
+
+ error = parse_ivrs_block(ivrs_block);
length += ivrs_block->length;
}
diff -r 8c921adf4833 -r 42f6c206c951 xen/drivers/passthrough/amd/iommu_detect.c
--- a/xen/drivers/passthrough/amd/iommu_detect.c Fri Mar 14 15:07:45
2008 -0600
+++ b/xen/drivers/passthrough/amd/iommu_detect.c Thu Mar 20 12:35:40
2008 -0600
@@ -20,14 +20,14 @@
#include <xen/config.h>
#include <xen/errno.h>
-#include <asm/iommu.h>
+#include <xen/iommu.h>
#include <asm/amd-iommu.h>
#include <asm/hvm/svm/amd-iommu-proto.h>
#include "../pci-direct.h"
#include "../pci_regs.h"
-static int __init valid_bridge_bus_config(int bus, int dev, int func,
- int *sec_bus, int *sub_bus)
+static int __init valid_bridge_bus_config(
+ int bus, int dev, int func, int *sec_bus, int *sub_bus)
{
int pri_bus;
@@ -35,7 +35,7 @@ static int __init valid_bridge_bus_confi
*sec_bus = read_pci_config_byte(bus, dev, func, PCI_SECONDARY_BUS);
*sub_bus = read_pci_config_byte(bus, dev, func, PCI_SUBORDINATE_BUS);
- return ( pri_bus == bus && *sec_bus > bus && *sub_bus >= *sec_bus );
+ return ((pri_bus == bus) && (*sec_bus > bus) && (*sub_bus >= *sec_bus));
}
int __init get_iommu_last_downstream_bus(struct amd_iommu *iommu)
@@ -49,9 +49,11 @@ int __init get_iommu_last_downstream_bus
iommu->downstream_bus_present[bus] = 1;
dev = PCI_SLOT(iommu->first_devfn);
multi_func = PCI_FUNC(iommu->first_devfn) > 0;
- for ( devfn = iommu->first_devfn; devfn <= iommu->last_devfn; ++devfn ) {
+ for ( devfn = iommu->first_devfn; devfn <= iommu->last_devfn; devfn++ )
+ {
/* skipping to next device#? */
- if ( dev != PCI_SLOT(devfn) ) {
+ if ( dev != PCI_SLOT(devfn) )
+ {
dev = PCI_SLOT(devfn);
multi_func = 0;
}
@@ -62,14 +64,15 @@ int __init get_iommu_last_downstream_bus
continue;
hdr_type = read_pci_config_byte(bus, dev, func,
- PCI_HEADER_TYPE);
+ PCI_HEADER_TYPE);
if ( func == 0 )
multi_func = IS_PCI_MULTI_FUNCTION(hdr_type);
if ( (func == 0 || multi_func) &&
- IS_PCI_TYPE1_HEADER(hdr_type) ) {
- if (!valid_bridge_bus_config(bus, dev, func,
- &sec_bus, &sub_bus))
+ IS_PCI_TYPE1_HEADER(hdr_type) )
+ {
+ if ( !valid_bridge_bus_config(bus, dev, func,
+ &sec_bus, &sub_bus) )
return -ENODEV;
if ( sub_bus > iommu->last_downstream_bus )
@@ -84,18 +87,18 @@ int __init get_iommu_last_downstream_bus
}
int __init get_iommu_capabilities(u8 bus, u8 dev, u8 func, u8 cap_ptr,
- struct amd_iommu *iommu)
+ struct amd_iommu *iommu)
{
u32 cap_header, cap_range, misc_info;
u64 mmio_bar;
- mmio_bar = (u64)read_pci_config(bus, dev, func,
- cap_ptr + PCI_CAP_MMIO_BAR_HIGH_OFFSET) << 32;
+ mmio_bar = (u64)read_pci_config(
+ bus, dev, func, cap_ptr + PCI_CAP_MMIO_BAR_HIGH_OFFSET) << 32;
mmio_bar |= read_pci_config(bus, dev, func,
- cap_ptr + PCI_CAP_MMIO_BAR_LOW_OFFSET);
+ cap_ptr + PCI_CAP_MMIO_BAR_LOW_OFFSET);
iommu->mmio_base_phys = mmio_bar & (u64)~0x3FFF;
- if ( (mmio_bar & 0x1) == 0 || iommu->mmio_base_phys == 0 )
+ if ( ((mmio_bar & 0x1) == 0) || (iommu->mmio_base_phys == 0) )
{
dprintk(XENLOG_ERR ,
"AMD IOMMU: Invalid MMIO_BAR = 0x%"PRIx64"\n", mmio_bar);
@@ -106,42 +109,37 @@ int __init get_iommu_capabilities(u8 bus
iommu->cap_offset = cap_ptr;
cap_header = read_pci_config(bus, dev, func, cap_ptr);
- iommu->revision = get_field_from_reg_u32(cap_header,
- PCI_CAP_REV_MASK, PCI_CAP_REV_SHIFT);
- iommu->iotlb_support = get_field_from_reg_u32(cap_header,
- PCI_CAP_IOTLB_MASK, PCI_CAP_IOTLB_SHIFT);
- iommu->ht_tunnel_support = get_field_from_reg_u32(cap_header,
- PCI_CAP_HT_TUNNEL_MASK,
- PCI_CAP_HT_TUNNEL_SHIFT);
- iommu->pte_not_present_cached = get_field_from_reg_u32(cap_header,
- PCI_CAP_NP_CACHE_MASK,
- PCI_CAP_NP_CACHE_SHIFT);
+ iommu->revision = get_field_from_reg_u32(
+ cap_header, PCI_CAP_REV_MASK, PCI_CAP_REV_SHIFT);
+ iommu->iotlb_support = get_field_from_reg_u32(
+ cap_header, PCI_CAP_IOTLB_MASK, PCI_CAP_IOTLB_SHIFT);
+ iommu->ht_tunnel_support = get_field_from_reg_u32(
+ cap_header, PCI_CAP_HT_TUNNEL_MASK, PCI_CAP_HT_TUNNEL_SHIFT);
+ iommu->pte_not_present_cached = get_field_from_reg_u32(
+ cap_header, PCI_CAP_NP_CACHE_MASK, PCI_CAP_NP_CACHE_SHIFT);
cap_range = read_pci_config(bus, dev, func,
- cap_ptr + PCI_CAP_RANGE_OFFSET);
- iommu->unit_id = get_field_from_reg_u32(cap_range,
- PCI_CAP_UNIT_ID_MASK,
- PCI_CAP_UNIT_ID_SHIFT);
- iommu->root_bus = get_field_from_reg_u32(cap_range,
- PCI_CAP_BUS_NUMBER_MASK,
- PCI_CAP_BUS_NUMBER_SHIFT);
- iommu->first_devfn = get_field_from_reg_u32(cap_range,
- PCI_CAP_FIRST_DEVICE_MASK,
- PCI_CAP_FIRST_DEVICE_SHIFT);
- iommu->last_devfn = get_field_from_reg_u32(cap_range,
- PCI_CAP_LAST_DEVICE_MASK,
- PCI_CAP_LAST_DEVICE_SHIFT);
+ cap_ptr + PCI_CAP_RANGE_OFFSET);
+ iommu->unit_id = get_field_from_reg_u32(
+ cap_range, PCI_CAP_UNIT_ID_MASK, PCI_CAP_UNIT_ID_SHIFT);
+ iommu->root_bus = get_field_from_reg_u32(
+ cap_range, PCI_CAP_BUS_NUMBER_MASK, PCI_CAP_BUS_NUMBER_SHIFT);
+ iommu->first_devfn = get_field_from_reg_u32(
+ cap_range, PCI_CAP_FIRST_DEVICE_MASK, PCI_CAP_FIRST_DEVICE_SHIFT);
+ iommu->last_devfn = get_field_from_reg_u32(
+ cap_range, PCI_CAP_LAST_DEVICE_MASK, PCI_CAP_LAST_DEVICE_SHIFT);
misc_info = read_pci_config(bus, dev, func,
- cap_ptr + PCI_MISC_INFO_OFFSET);
- iommu->msi_number = get_field_from_reg_u32(misc_info,
- PCI_CAP_MSI_NUMBER_MASK,
- PCI_CAP_MSI_NUMBER_SHIFT);
+ cap_ptr + PCI_MISC_INFO_OFFSET);
+ iommu->msi_number = get_field_from_reg_u32(
+ misc_info, PCI_CAP_MSI_NUMBER_MASK, PCI_CAP_MSI_NUMBER_SHIFT);
+
return 0;
}
-static int __init scan_caps_for_iommu(int bus, int dev, int func,
- iommu_detect_callback_ptr_t iommu_detect_callback)
+static int __init scan_caps_for_iommu(
+ int bus, int dev, int func,
+ iommu_detect_callback_ptr_t iommu_detect_callback)
{
int cap_ptr, cap_id, cap_type;
u32 cap_header;
@@ -149,32 +147,35 @@ static int __init scan_caps_for_iommu(in
count = 0;
cap_ptr = read_pci_config_byte(bus, dev, func,
- PCI_CAPABILITY_LIST);
- while ( cap_ptr >= PCI_MIN_CAP_OFFSET &&
- count < PCI_MAX_CAP_BLOCKS && !error ) {
+ PCI_CAPABILITY_LIST);
+ while ( (cap_ptr >= PCI_MIN_CAP_OFFSET) &&
+ (count < PCI_MAX_CAP_BLOCKS) &&
+ !error )
+ {
cap_ptr &= PCI_CAP_PTR_MASK;
cap_header = read_pci_config(bus, dev, func, cap_ptr);
- cap_id = get_field_from_reg_u32(cap_header,
- PCI_CAP_ID_MASK, PCI_CAP_ID_SHIFT);
-
- if ( cap_id == PCI_CAP_ID_SECURE_DEVICE ) {
- cap_type = get_field_from_reg_u32(cap_header,
- PCI_CAP_TYPE_MASK, PCI_CAP_TYPE_SHIFT);
- if ( cap_type == PCI_CAP_TYPE_IOMMU ) {
+ cap_id = get_field_from_reg_u32(
+ cap_header, PCI_CAP_ID_MASK, PCI_CAP_ID_SHIFT);
+
+ if ( cap_id == PCI_CAP_ID_SECURE_DEVICE )
+ {
+ cap_type = get_field_from_reg_u32(
+ cap_header, PCI_CAP_TYPE_MASK, PCI_CAP_TYPE_SHIFT);
+ if ( cap_type == PCI_CAP_TYPE_IOMMU )
error = iommu_detect_callback(
- bus, dev, func, cap_ptr);
- }
- }
-
- cap_ptr = get_field_from_reg_u32(cap_header,
- PCI_CAP_NEXT_PTR_MASK, PCI_CAP_NEXT_PTR_SHIFT);
- ++count; }
-
- return error;
-}
-
-static int __init scan_functions_for_iommu(int bus, int dev,
- iommu_detect_callback_ptr_t iommu_detect_callback)
+ bus, dev, func, cap_ptr);
+ }
+
+ cap_ptr = get_field_from_reg_u32(
+ cap_header, PCI_CAP_NEXT_PTR_MASK, PCI_CAP_NEXT_PTR_SHIFT);
+ count++;
+ }
+
+ return error;
+}
+
+static int __init scan_functions_for_iommu(
+ int bus, int dev, iommu_detect_callback_ptr_t iommu_detect_callback)
{
int func, hdr_type;
int count, error = 0;
@@ -182,19 +183,20 @@ static int __init scan_functions_for_iom
func = 0;
count = 1;
while ( VALID_PCI_VENDOR_ID(read_pci_config_16(bus, dev, func,
- PCI_VENDOR_ID)) && !error && func < count ) {
+ PCI_VENDOR_ID)) &&
+ !error && (func < count) )
+ {
hdr_type = read_pci_config_byte(bus, dev, func,
- PCI_HEADER_TYPE);
+ PCI_HEADER_TYPE);
if ( func == 0 && IS_PCI_MULTI_FUNCTION(hdr_type) )
count = PCI_MAX_FUNC_COUNT;
if ( IS_PCI_TYPE0_HEADER(hdr_type) ||
- IS_PCI_TYPE1_HEADER(hdr_type) ) {
- error = scan_caps_for_iommu(bus, dev, func,
- iommu_detect_callback);
- }
- ++func;
+ IS_PCI_TYPE1_HEADER(hdr_type) )
+ error = scan_caps_for_iommu(bus, dev, func,
+ iommu_detect_callback);
+ func++;
}
return error;
@@ -205,13 +207,11 @@ int __init scan_for_iommu(iommu_detect_c
{
int bus, dev, error = 0;
- for ( bus = 0; bus < PCI_MAX_BUS_COUNT && !error; ++bus ) {
- for ( dev = 0; dev < PCI_MAX_DEV_COUNT && !error; ++dev ) {
- error = scan_functions_for_iommu(bus, dev,
- iommu_detect_callback);
- }
- }
-
- return error;
-}
-
+ for ( bus = 0; bus < PCI_MAX_BUS_COUNT && !error; ++bus )
+ for ( dev = 0; dev < PCI_MAX_DEV_COUNT && !error; ++dev )
+ error = scan_functions_for_iommu(bus, dev,
+ iommu_detect_callback);
+
+ return error;
+}
+
diff -r 8c921adf4833 -r 42f6c206c951 xen/drivers/passthrough/amd/iommu_init.c
--- a/xen/drivers/passthrough/amd/iommu_init.c Fri Mar 14 15:07:45 2008 -0600
+++ b/xen/drivers/passthrough/amd/iommu_init.c Thu Mar 20 12:35:40 2008 -0600
@@ -32,26 +32,28 @@ int __init map_iommu_mmio_region(struct
{
unsigned long mfn;
- if ( nr_amd_iommus > MAX_AMD_IOMMUS ) {
+ if ( nr_amd_iommus > MAX_AMD_IOMMUS )
+ {
gdprintk(XENLOG_ERR,
- "IOMMU: nr_amd_iommus %d > MAX_IOMMUS\n", nr_amd_iommus);
+ "IOMMU: nr_amd_iommus %d > MAX_IOMMUS\n", nr_amd_iommus);
return -ENOMEM;
}
- iommu->mmio_base = (void *) fix_to_virt(FIX_IOMMU_MMIO_BASE_0 +
- nr_amd_iommus * MMIO_PAGES_PER_IOMMU);
- mfn = (unsigned long)iommu->mmio_base_phys >> PAGE_SHIFT;
+ iommu->mmio_base = (void *)fix_to_virt(
+ FIX_IOMMU_MMIO_BASE_0 + nr_amd_iommus * MMIO_PAGES_PER_IOMMU);
+ mfn = (unsigned long)(iommu->mmio_base_phys >> PAGE_SHIFT);
map_pages_to_xen((unsigned long)iommu->mmio_base, mfn,
- MMIO_PAGES_PER_IOMMU, PAGE_HYPERVISOR_NOCACHE);
+ MMIO_PAGES_PER_IOMMU, PAGE_HYPERVISOR_NOCACHE);
- memset((u8*)iommu->mmio_base, 0, IOMMU_MMIO_REGION_LENGTH);
+ memset(iommu->mmio_base, 0, IOMMU_MMIO_REGION_LENGTH);
return 0;
}
void __init unmap_iommu_mmio_region(struct amd_iommu *iommu)
{
- if ( iommu->mmio_base ) {
+ if ( iommu->mmio_base )
+ {
iounmap(iommu->mmio_base);
iommu->mmio_base = NULL;
}
@@ -67,16 +69,16 @@ void __init register_iommu_dev_table_in_
addr_hi = addr_64 >> 32;
set_field_in_reg_u32((u32)addr_lo >> PAGE_SHIFT, 0,
- IOMMU_DEV_TABLE_BASE_LOW_MASK,
- IOMMU_DEV_TABLE_BASE_LOW_SHIFT, &entry);
+ IOMMU_DEV_TABLE_BASE_LOW_MASK,
+ IOMMU_DEV_TABLE_BASE_LOW_SHIFT, &entry);
set_field_in_reg_u32((iommu->dev_table.alloc_size / PAGE_SIZE) - 1,
- entry, IOMMU_DEV_TABLE_SIZE_MASK,
- IOMMU_DEV_TABLE_SIZE_SHIFT, &entry);
+ entry, IOMMU_DEV_TABLE_SIZE_MASK,
+ IOMMU_DEV_TABLE_SIZE_SHIFT, &entry);
writel(entry, iommu->mmio_base + IOMMU_DEV_TABLE_BASE_LOW_OFFSET);
set_field_in_reg_u32((u32)addr_hi, 0,
- IOMMU_DEV_TABLE_BASE_HIGH_MASK,
- IOMMU_DEV_TABLE_BASE_HIGH_SHIFT, &entry);
+ IOMMU_DEV_TABLE_BASE_HIGH_MASK,
+ IOMMU_DEV_TABLE_BASE_HIGH_SHIFT, &entry);
writel(entry, iommu->mmio_base + IOMMU_DEV_TABLE_BASE_HIGH_OFFSET);
}
@@ -91,49 +93,49 @@ void __init register_iommu_cmd_buffer_in
addr_hi = addr_64 >> 32;
set_field_in_reg_u32((u32)addr_lo >> PAGE_SHIFT, 0,
- IOMMU_CMD_BUFFER_BASE_LOW_MASK,
- IOMMU_CMD_BUFFER_BASE_LOW_SHIFT, &entry);
+ IOMMU_CMD_BUFFER_BASE_LOW_MASK,
+ IOMMU_CMD_BUFFER_BASE_LOW_SHIFT, &entry);
writel(entry, iommu->mmio_base + IOMMU_CMD_BUFFER_BASE_LOW_OFFSET);
power_of2_entries = get_order_from_bytes(iommu->cmd_buffer.alloc_size) +
IOMMU_CMD_BUFFER_POWER_OF2_ENTRIES_PER_PAGE;
set_field_in_reg_u32((u32)addr_hi, 0,
- IOMMU_CMD_BUFFER_BASE_HIGH_MASK,
- IOMMU_CMD_BUFFER_BASE_HIGH_SHIFT, &entry);
+ IOMMU_CMD_BUFFER_BASE_HIGH_MASK,
+ IOMMU_CMD_BUFFER_BASE_HIGH_SHIFT, &entry);
set_field_in_reg_u32(power_of2_entries, entry,
- IOMMU_CMD_BUFFER_LENGTH_MASK,
- IOMMU_CMD_BUFFER_LENGTH_SHIFT, &entry);
+ IOMMU_CMD_BUFFER_LENGTH_MASK,
+ IOMMU_CMD_BUFFER_LENGTH_SHIFT, &entry);
writel(entry, iommu->mmio_base+IOMMU_CMD_BUFFER_BASE_HIGH_OFFSET);
}
static void __init set_iommu_translation_control(struct amd_iommu *iommu,
- int enable)
+ int enable)
{
u32 entry;
entry = readl(iommu->mmio_base+IOMMU_CONTROL_MMIO_OFFSET);
set_field_in_reg_u32(iommu->ht_tunnel_support ? IOMMU_CONTROL_ENABLED :
- IOMMU_CONTROL_ENABLED, entry,
- IOMMU_CONTROL_HT_TUNNEL_TRANSLATION_MASK,
- IOMMU_CONTROL_HT_TUNNEL_TRANSLATION_SHIFT, &entry);
+ IOMMU_CONTROL_ENABLED, entry,
+ IOMMU_CONTROL_HT_TUNNEL_TRANSLATION_MASK,
+ IOMMU_CONTROL_HT_TUNNEL_TRANSLATION_SHIFT, &entry);
set_field_in_reg_u32(enable ? IOMMU_CONTROL_ENABLED :
- IOMMU_CONTROL_ENABLED, entry,
- IOMMU_CONTROL_TRANSLATION_ENABLE_MASK,
- IOMMU_CONTROL_TRANSLATION_ENABLE_SHIFT, &entry);
+ IOMMU_CONTROL_ENABLED, entry,
+ IOMMU_CONTROL_TRANSLATION_ENABLE_MASK,
+ IOMMU_CONTROL_TRANSLATION_ENABLE_SHIFT, &entry);
writel(entry, iommu->mmio_base+IOMMU_CONTROL_MMIO_OFFSET);
}
static void __init set_iommu_command_buffer_control(struct amd_iommu *iommu,
- int enable)
+ int enable)
{
u32 entry;
entry = readl(iommu->mmio_base+IOMMU_CONTROL_MMIO_OFFSET);
set_field_in_reg_u32(enable ? IOMMU_CONTROL_ENABLED :
- IOMMU_CONTROL_ENABLED, entry,
- IOMMU_CONTROL_COMMAND_BUFFER_ENABLE_MASK,
- IOMMU_CONTROL_COMMAND_BUFFER_ENABLE_SHIFT, &entry);
+ IOMMU_CONTROL_ENABLED, entry,
+ IOMMU_CONTROL_COMMAND_BUFFER_ENABLE_MASK,
+ IOMMU_CONTROL_COMMAND_BUFFER_ENABLE_SHIFT, &entry);
writel(entry, iommu->mmio_base+IOMMU_CONTROL_MMIO_OFFSET);
}
@@ -146,34 +148,34 @@ static void __init register_iommu_exclus
addr_hi = iommu->exclusion_limit >> 32;
set_field_in_reg_u32((u32)addr_hi, 0,
- IOMMU_EXCLUSION_LIMIT_HIGH_MASK,
- IOMMU_EXCLUSION_LIMIT_HIGH_SHIFT, &entry);
+ IOMMU_EXCLUSION_LIMIT_HIGH_MASK,
+ IOMMU_EXCLUSION_LIMIT_HIGH_SHIFT, &entry);
writel(entry, iommu->mmio_base+IOMMU_EXCLUSION_LIMIT_HIGH_OFFSET);
set_field_in_reg_u32((u32)addr_lo >> PAGE_SHIFT, 0,
- IOMMU_EXCLUSION_LIMIT_LOW_MASK,
- IOMMU_EXCLUSION_LIMIT_LOW_SHIFT, &entry);
+ IOMMU_EXCLUSION_LIMIT_LOW_MASK,
+ IOMMU_EXCLUSION_LIMIT_LOW_SHIFT, &entry);
writel(entry, iommu->mmio_base+IOMMU_EXCLUSION_LIMIT_LOW_OFFSET);
addr_lo = iommu->exclusion_base & DMA_32BIT_MASK;
addr_hi = iommu->exclusion_base >> 32;
set_field_in_reg_u32((u32)addr_hi, 0,
- IOMMU_EXCLUSION_BASE_HIGH_MASK,
- IOMMU_EXCLUSION_BASE_HIGH_SHIFT, &entry);
+ IOMMU_EXCLUSION_BASE_HIGH_MASK,
+ IOMMU_EXCLUSION_BASE_HIGH_SHIFT, &entry);
writel(entry, iommu->mmio_base+IOMMU_EXCLUSION_BASE_HIGH_OFFSET);
set_field_in_reg_u32((u32)addr_lo >> PAGE_SHIFT, 0,
- IOMMU_EXCLUSION_BASE_LOW_MASK,
- IOMMU_EXCLUSION_BASE_LOW_SHIFT, &entry);
+ IOMMU_EXCLUSION_BASE_LOW_MASK,
+ IOMMU_EXCLUSION_BASE_LOW_SHIFT, &entry);
set_field_in_reg_u32(iommu->exclusion_allow_all, entry,
- IOMMU_EXCLUSION_ALLOW_ALL_MASK,
- IOMMU_EXCLUSION_ALLOW_ALL_SHIFT, &entry);
+ IOMMU_EXCLUSION_ALLOW_ALL_MASK,
+ IOMMU_EXCLUSION_ALLOW_ALL_SHIFT, &entry);
set_field_in_reg_u32(iommu->exclusion_enable, entry,
- IOMMU_EXCLUSION_RANGE_ENABLE_MASK,
- IOMMU_EXCLUSION_RANGE_ENABLE_SHIFT, &entry);
+ IOMMU_EXCLUSION_RANGE_ENABLE_MASK,
+ IOMMU_EXCLUSION_RANGE_ENABLE_SHIFT, &entry);
writel(entry, iommu->mmio_base+IOMMU_EXCLUSION_BASE_LOW_OFFSET);
}
@@ -184,5 +186,3 @@ void __init enable_iommu(struct amd_iomm
set_iommu_translation_control(iommu, IOMMU_CONTROL_ENABLED);
printk("AMD IOMMU %d: Enabled\n", nr_amd_iommus);
}
-
-
diff -r 8c921adf4833 -r 42f6c206c951 xen/drivers/passthrough/amd/iommu_map.c
--- a/xen/drivers/passthrough/amd/iommu_map.c Fri Mar 14 15:07:45 2008 -0600
+++ b/xen/drivers/passthrough/amd/iommu_map.c Thu Mar 20 12:35:40 2008 -0600
@@ -19,7 +19,7 @@
*/
#include <xen/sched.h>
-#include <asm/hvm/iommu.h>
+#include <xen/hvm/iommu.h>
#include <asm/amd-iommu.h>
#include <asm/hvm/svm/amd-iommu-proto.h>
@@ -132,7 +132,8 @@ void flush_command_buffer(struct amd_iom
send_iommu_command(iommu, cmd);
/* wait for 'ComWaitInt' to signal comp#endifletion? */
- if ( amd_iommu_poll_comp_wait ) {
+ if ( amd_iommu_poll_comp_wait )
+ {
loop_count = amd_iommu_poll_comp_wait;
do {
status = readl(iommu->mmio_base +
@@ -152,8 +153,10 @@ void flush_command_buffer(struct amd_iom
IOMMU_STATUS_MMIO_OFFSET);
}
else
+ {
dprintk(XENLOG_WARNING, "AMD IOMMU: Warning:"
" ComWaitInt bit did not assert!\n");
+ }
}
}
@@ -234,7 +237,7 @@ static void amd_iommu_set_page_directory
}
void amd_iommu_set_dev_table_entry(u32 *dte, u64 root_ptr, u16 domain_id,
- u8 sys_mgt, u8 dev_ex, u8 paging_mode)
+ u8 sys_mgt, u8 dev_ex, u8 paging_mode)
{
u64 addr_hi, addr_lo;
u32 entry;
@@ -397,7 +400,7 @@ int amd_iommu_map_page(struct domain *d,
spin_lock_irqsave(&hd->mapping_lock, flags);
pte = get_pte_from_page_tables(hd->root_table, hd->paging_mode, gfn);
- if ( pte == 0 )
+ if ( pte == NULL )
{
dprintk(XENLOG_ERR,
"AMD IOMMU: Invalid IO pagetable entry gfn = %lx\n", gfn);
@@ -428,7 +431,7 @@ int amd_iommu_unmap_page(struct domain *
spin_lock_irqsave(&hd->mapping_lock, flags);
pte = get_pte_from_page_tables(hd->root_table, hd->paging_mode, gfn);
- if ( pte == 0 )
+ if ( pte == NULL )
{
dprintk(XENLOG_ERR,
"AMD IOMMU: Invalid IO pagetable entry gfn = %lx\n", gfn);
@@ -441,7 +444,7 @@ int amd_iommu_unmap_page(struct domain *
spin_unlock_irqrestore(&hd->mapping_lock, flags);
/* send INVALIDATE_IOMMU_PAGES command */
- for_each_amd_iommu(iommu)
+ for_each_amd_iommu ( iommu )
{
spin_lock_irqsave(&iommu->lock, flags);
invalidate_iommu_page(iommu, io_addr, requestor_id);
@@ -453,9 +456,9 @@ int amd_iommu_unmap_page(struct domain *
}
int amd_iommu_reserve_domain_unity_map(
- struct domain *domain,
- unsigned long phys_addr,
- unsigned long size, int iw, int ir)
+ struct domain *domain,
+ unsigned long phys_addr,
+ unsigned long size, int iw, int ir)
{
unsigned long flags, npages, i;
void *pte;
@@ -466,17 +469,18 @@ int amd_iommu_reserve_domain_unity_map(
spin_lock_irqsave(&hd->mapping_lock, flags);
for ( i = 0; i < npages; ++i )
{
- pte = get_pte_from_page_tables(hd->root_table,
- hd->paging_mode, phys_addr>>PAGE_SHIFT);
- if ( pte == 0 )
+ pte = get_pte_from_page_tables(
+ hd->root_table, hd->paging_mode, phys_addr >> PAGE_SHIFT);
+ if ( pte == NULL )
{
dprintk(XENLOG_ERR,
- "AMD IOMMU: Invalid IO pagetable entry phys_addr = %lx\n",
phys_addr);
+ "AMD IOMMU: Invalid IO pagetable entry "
+ "phys_addr = %lx\n", phys_addr);
spin_unlock_irqrestore(&hd->mapping_lock, flags);
return -EFAULT;
}
set_page_table_entry_present((u32 *)pte,
- phys_addr, iw, ir);
+ phys_addr, iw, ir);
phys_addr += PAGE_SIZE;
}
spin_unlock_irqrestore(&hd->mapping_lock, flags);
diff -r 8c921adf4833 -r 42f6c206c951 xen/drivers/passthrough/amd/pci_amd_iommu.c
--- a/xen/drivers/passthrough/amd/pci_amd_iommu.c Fri Mar 14 15:07:45
2008 -0600
+++ b/xen/drivers/passthrough/amd/pci_amd_iommu.c Thu Mar 20 12:35:40
2008 -0600
@@ -168,7 +168,7 @@ int iommu_detect_callback(u8 bus, u8 dev
list_add_tail(&iommu->list, &amd_iommu_head);
/* allocate resources for this IOMMU */
- if (allocate_iommu_resources(iommu) != 0)
+ if ( allocate_iommu_resources(iommu) != 0 )
goto error_out;
return 0;
@@ -208,7 +208,7 @@ static int __init amd_iommu_init(void)
}
/* assign default values for device entries */
- for ( bdf = 0; bdf < ivrs_bdf_entries; ++bdf )
+ for ( bdf = 0; bdf < ivrs_bdf_entries; bdf++ )
{
ivrs_mappings[bdf].dte_requestor_id = bdf;
ivrs_mappings[bdf].dte_sys_mgt_enable =
@@ -288,7 +288,8 @@ void amd_iommu_setup_domain_device(
sys_mgt = ivrs_mappings[req_id].dte_sys_mgt_enable;
dev_ex = ivrs_mappings[req_id].dte_allow_exclusion;
amd_iommu_set_dev_table_entry((u32 *)dte, root_ptr,
- req_id, sys_mgt, dev_ex, hd->paging_mode);
+ req_id, sys_mgt, dev_ex,
+ hd->paging_mode);
invalidate_dev_table_entry(iommu, req_id);
flush_command_buffer(iommu);
@@ -317,8 +318,8 @@ void __init amd_iommu_setup_dom0_devices
{
l = read_pci_config(bus, dev, func, PCI_VENDOR_ID);
/* some broken boards return 0 or ~0 if a slot is empty: */
- if ( l == 0xffffffff || l == 0x00000000 ||
- l == 0x0000ffff || l == 0xffff0000 )
+ if ( (l == 0xffffffff) || (l == 0x00000000) ||
+ (l == 0x0000ffff) || (l == 0xffff0000) )
continue;
pdev = xmalloc(struct pci_dev);
@@ -368,22 +369,22 @@ int amd_iommu_detect(void)
/* allocate 'ivrs mappings' table */
/* note: the table has entries to accomodate all IOMMUs */
last_bus = 0;
- for_each_amd_iommu (iommu)
- if (iommu->last_downstream_bus > last_bus)
- last_bus = iommu->last_downstream_bus;
+ for_each_amd_iommu ( iommu )
+ if ( iommu->last_downstream_bus > last_bus )
+ last_bus = iommu->last_downstream_bus;
ivrs_bdf_entries = (last_bus + 1) *
- IOMMU_DEV_TABLE_ENTRIES_PER_BUS;
+ IOMMU_DEV_TABLE_ENTRIES_PER_BUS;
ivrs_mappings = xmalloc_array( struct ivrs_mappings, ivrs_bdf_entries);
if ( !ivrs_mappings )
{
dprintk(XENLOG_ERR, "AMD IOMMU:"
- " Error allocating IVRS DevMappings table\n");
+ " Error allocating IVRS DevMappings table\n");
goto error_out;
}
memset(ivrs_mappings, 0,
- ivrs_bdf_entries * sizeof(struct ivrs_mappings));
+ ivrs_bdf_entries * sizeof(struct ivrs_mappings));
}
if ( amd_iommu_init() != 0 )
@@ -424,6 +425,7 @@ static int allocate_domain_resources(str
spin_unlock_irqrestore(&hd->mapping_lock, flags);
return 0;
+
error_out:
spin_unlock_irqrestore(&hd->mapping_lock, flags);
return -ENOMEM;
@@ -433,7 +435,7 @@ static int get_paging_mode(unsigned long
{
int level = 1;
- BUG_ON ( !max_page );
+ BUG_ON(!max_page);
if ( entries > max_page )
entries = max_page;
@@ -441,8 +443,7 @@ static int get_paging_mode(unsigned long
while ( entries > PTE_PER_TABLE_SIZE )
{
entries = PTE_PER_TABLE_ALIGN(entries) >> PTE_PER_TABLE_SHIFT;
- ++level;
- if ( level > 6 )
+ if ( ++level > 6 )
return -ENOMEM;
}
@@ -509,7 +510,7 @@ static int reassign_device( struct domai
int bdf;
unsigned long flags;
- for_each_pdev( source, pdev )
+ for_each_pdev ( source, pdev )
{
if ( (pdev->bus != bus) || (pdev->devfn != devfn) )
continue;
@@ -522,23 +523,7 @@ static int reassign_device( struct domai
iommu = (bdf < ivrs_bdf_entries) ?
find_iommu_for_device(bus, pdev->devfn) : NULL;
- if ( iommu )
- {
- amd_iommu_disable_domain_device(source, iommu, bdf);
- /* Move pci device from the source domain to target domain. */
- spin_lock_irqsave(&source_hd->iommu_list_lock, flags);
- spin_lock_irqsave(&target_hd->iommu_list_lock, flags);
- list_move(&pdev->list, &target_hd->pdev_list);
- spin_unlock_irqrestore(&target_hd->iommu_list_lock, flags);
- spin_unlock_irqrestore(&source_hd->iommu_list_lock, flags);
-
- amd_iommu_setup_domain_device(target, iommu, bdf);
- gdprintk(XENLOG_INFO ,
- "AMD IOMMU: reassign %x:%x.%x domain %d -> domain %d\n",
- bus, PCI_SLOT(devfn), PCI_FUNC(devfn),
- source->domain_id, target->domain_id);
- }
- else
+ if ( !iommu )
{
gdprintk(XENLOG_ERR , "AMD IOMMU: fail to find iommu."
" %x:%x.%x cannot be assigned to domain %d\n",
@@ -546,6 +531,20 @@ static int reassign_device( struct domai
return -ENODEV;
}
+ amd_iommu_disable_domain_device(source, iommu, bdf);
+ /* Move pci device from the source domain to target domain. */
+ spin_lock_irqsave(&source_hd->iommu_list_lock, flags);
+ spin_lock_irqsave(&target_hd->iommu_list_lock, flags);
+ list_move(&pdev->list, &target_hd->pdev_list);
+ spin_unlock_irqrestore(&target_hd->iommu_list_lock, flags);
+ spin_unlock_irqrestore(&source_hd->iommu_list_lock, flags);
+
+ amd_iommu_setup_domain_device(target, iommu, bdf);
+ gdprintk(XENLOG_INFO ,
+ "AMD IOMMU: reassign %x:%x.%x domain %d -> domain %d\n",
+ bus, PCI_SLOT(devfn), PCI_FUNC(devfn),
+ source->domain_id, target->domain_id);
+
break;
}
return 0;
@@ -557,9 +556,10 @@ int amd_iommu_assign_device(struct domai
int req_id;
req_id = ivrs_mappings[bdf].dte_requestor_id;
- if (ivrs_mappings[req_id].unity_map_enable)
- {
- amd_iommu_reserve_domain_unity_map(d,
+ if ( ivrs_mappings[req_id].unity_map_enable )
+ {
+ amd_iommu_reserve_domain_unity_map(
+ d,
ivrs_mappings[req_id].addr_range_start,
ivrs_mappings[req_id].addr_range_length,
ivrs_mappings[req_id].write_permission,
@@ -606,7 +606,7 @@ static void deallocate_next_page_table(v
{
deallocate_next_page_table(next_table,
next_index, next_level);
- ++next_index;
+ next_index++;
} while (next_index < PTE_PER_TABLE_SIZE);
}
@@ -622,11 +622,12 @@ static void deallocate_iommu_page_tables
if ( hd ->root_table )
{
index = 0;
+
do
{
deallocate_next_page_table(hd->root_table,
index, hd->paging_mode);
- ++index;
+ index++;
} while ( index < PTE_PER_TABLE_SIZE );
free_xenheap_page(hd ->root_table);
@@ -644,7 +645,8 @@ void amd_iommu_domain_destroy(struct dom
release_domain_devices(d);
}
-void amd_iommu_return_device(struct domain *s, struct domain *t, u8 bus, u8
devfn)
+void amd_iommu_return_device(
+ struct domain *s, struct domain *t, u8 bus, u8 devfn)
{
pdev_flr(bus, devfn);
reassign_device(s, t, bus, devfn);
diff -r 8c921adf4833 -r 42f6c206c951 xen/drivers/passthrough/io.c
--- /dev/null Thu Jan 01 00:00:00 1970 +0000
+++ b/xen/drivers/passthrough/io.c Thu Mar 20 12:35:40 2008 -0600
@@ -0,0 +1,272 @@
+/*
+ * Copyright (c) 2006, Intel Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope 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, Inc., 59 Temple
+ * Place - Suite 330, Boston, MA 02111-1307 USA.
+ *
+ * Copyright (C) Allen Kay <allen.m.kay@xxxxxxxxx>
+ * Copyright (C) Xiaohui Xin <xiaohui.xin@xxxxxxxxx>
+ */
+
+#include <xen/event.h>
+#include <xen/iommu.h>
+
+static void pt_irq_time_out(void *data)
+{
+ struct hvm_mirq_dpci_mapping *irq_map = data;
+ unsigned int guest_gsi, machine_gsi = 0;
+ struct hvm_irq_dpci *dpci = irq_map->dom->arch.hvm_domain.irq.dpci;
+ struct dev_intx_gsi_link *digl;
+ uint32_t device, intx;
+
+ list_for_each_entry ( digl, &irq_map->digl_list, list )
+ {
+ guest_gsi = digl->gsi;
+ machine_gsi = dpci->girq[guest_gsi].machine_gsi;
+ device = digl->device;
+ intx = digl->intx;
+ hvm_pci_intx_deassert(irq_map->dom, device, intx);
+ }
+
+ clear_bit(machine_gsi, dpci->dirq_mask);
+ stop_timer(&dpci->hvm_timer[irq_to_vector(machine_gsi)]);
+ spin_lock(&dpci->dirq_lock);
+ dpci->mirq[machine_gsi].pending = 0;
+ spin_unlock(&dpci->dirq_lock);
+ pirq_guest_eoi(irq_map->dom, machine_gsi);
+}
+
+int pt_irq_create_bind_vtd(
+ struct domain *d, xen_domctl_bind_pt_irq_t *pt_irq_bind)
+{
+ struct hvm_irq_dpci *hvm_irq_dpci = d->arch.hvm_domain.irq.dpci;
+ uint32_t machine_gsi, guest_gsi;
+ uint32_t device, intx, link;
+ struct dev_intx_gsi_link *digl;
+
+ if ( hvm_irq_dpci == NULL )
+ {
+ hvm_irq_dpci = xmalloc(struct hvm_irq_dpci);
+ if ( hvm_irq_dpci == NULL )
+ return -ENOMEM;
+
+ memset(hvm_irq_dpci, 0, sizeof(*hvm_irq_dpci));
+ spin_lock_init(&hvm_irq_dpci->dirq_lock);
+ for ( int i = 0; i < NR_IRQS; i++ )
+ INIT_LIST_HEAD(&hvm_irq_dpci->mirq[i].digl_list);
+
+ if ( cmpxchg((unsigned long *)&d->arch.hvm_domain.irq.dpci,
+ 0, (unsigned long)hvm_irq_dpci) != 0 )
+ xfree(hvm_irq_dpci);
+
+ hvm_irq_dpci = d->arch.hvm_domain.irq.dpci;
+ }
+
+ machine_gsi = pt_irq_bind->machine_irq;
+ device = pt_irq_bind->u.pci.device;
+ intx = pt_irq_bind->u.pci.intx;
+ guest_gsi = hvm_pci_intx_gsi(device, intx);
+ link = hvm_pci_intx_link(device, intx);
+ hvm_irq_dpci->link_cnt[link]++;
+
+ digl = xmalloc(struct dev_intx_gsi_link);
+ if ( !digl )
+ return -ENOMEM;
+
+ digl->device = device;
+ digl->intx = intx;
+ digl->gsi = guest_gsi;
+ digl->link = link;
+ list_add_tail(&digl->list,
+ &hvm_irq_dpci->mirq[machine_gsi].digl_list);
+
+ hvm_irq_dpci->girq[guest_gsi].valid = 1;
+ hvm_irq_dpci->girq[guest_gsi].device = device;
+ hvm_irq_dpci->girq[guest_gsi].intx = intx;
+ hvm_irq_dpci->girq[guest_gsi].machine_gsi = machine_gsi;
+
+ /* Bind the same mirq once in the same domain */
+ if ( !hvm_irq_dpci->mirq[machine_gsi].valid )
+ {
+ hvm_irq_dpci->mirq[machine_gsi].valid = 1;
+ hvm_irq_dpci->mirq[machine_gsi].dom = d;
+
+ init_timer(&hvm_irq_dpci->hvm_timer[irq_to_vector(machine_gsi)],
+ pt_irq_time_out, &hvm_irq_dpci->mirq[machine_gsi], 0);
+ /* Deal with gsi for legacy devices */
+ pirq_guest_bind(d->vcpu[0], machine_gsi, BIND_PIRQ__WILL_SHARE);
+ }
+
+ gdprintk(XENLOG_INFO VTDPREFIX,
+ "VT-d irq bind: m_irq = %x device = %x intx = %x\n",
+ machine_gsi, device, intx);
+ return 0;
+}
+
+int pt_irq_destroy_bind_vtd(
+ struct domain *d, xen_domctl_bind_pt_irq_t *pt_irq_bind)
+{
+ struct hvm_irq_dpci *hvm_irq_dpci = d->arch.hvm_domain.irq.dpci;
+ uint32_t machine_gsi, guest_gsi;
+ uint32_t device, intx, link;
+ struct list_head *digl_list, *tmp;
+ struct dev_intx_gsi_link *digl;
+
+ if ( hvm_irq_dpci == NULL )
+ return 0;
+
+ machine_gsi = pt_irq_bind->machine_irq;
+ device = pt_irq_bind->u.pci.device;
+ intx = pt_irq_bind->u.pci.intx;
+ guest_gsi = hvm_pci_intx_gsi(device, intx);
+ link = hvm_pci_intx_link(device, intx);
+ hvm_irq_dpci->link_cnt[link]--;
+
+ gdprintk(XENLOG_INFO,
+ "pt_irq_destroy_bind_vtd: machine_gsi=%d, guest_gsi=%d, device=%d,
intx=%d.\n",
+ machine_gsi, guest_gsi, device, intx);
+ memset(&hvm_irq_dpci->girq[guest_gsi], 0, sizeof(struct
hvm_girq_dpci_mapping));
+
+ /* clear the mirq info */
+ if ( hvm_irq_dpci->mirq[machine_gsi].valid )
+ {
+
+ list_for_each_safe ( digl_list, tmp,
+ &hvm_irq_dpci->mirq[machine_gsi].digl_list )
+ {
+ digl = list_entry(digl_list,
+ struct dev_intx_gsi_link, list);
+ if ( digl->device == device &&
+ digl->intx == intx &&
+ digl->link == link &&
+ digl->gsi == guest_gsi )
+ {
+ list_del(&digl->list);
+ xfree(digl);
+ }
+ }
+
+ if ( list_empty(&hvm_irq_dpci->mirq[machine_gsi].digl_list) )
+ {
+ pirq_guest_unbind(d, machine_gsi);
+ kill_timer(&hvm_irq_dpci->hvm_timer[irq_to_vector(machine_gsi)]);
+ hvm_irq_dpci->mirq[machine_gsi].dom = NULL;
+ hvm_irq_dpci->mirq[machine_gsi].valid = 0;
+ }
+ }
+
+ gdprintk(XENLOG_INFO,
+ "XEN_DOMCTL_irq_unmapping: m_irq = %x device = %x intx = %x\n",
+ machine_gsi, device, intx);
+
+ return 0;
+}
+
+int hvm_do_IRQ_dpci(struct domain *d, unsigned int mirq)
+{
+ struct hvm_irq *hvm_irq = &d->arch.hvm_domain.irq;
+
+ if ( !iommu_enabled || (d == dom0) || (hvm_irq->dpci == NULL) ||
+ !hvm_irq->dpci->mirq[mirq].valid )
+ return 0;
+
+ /*
+ * Set a timer here to avoid situations where the IRQ line is shared, and
+ * the device belonging to the pass-through guest is not yet active. In
+ * this case the guest may not pick up the interrupt (e.g., masked at the
+ * PIC) and we need to detect that.
+ */
+ set_bit(mirq, hvm_irq->dpci->dirq_mask);
+ set_timer(&hvm_irq->dpci->hvm_timer[irq_to_vector(mirq)],
+ NOW() + PT_IRQ_TIME_OUT);
+ vcpu_kick(d->vcpu[0]);
+
+ return 1;
+}
+
+static void hvm_dpci_isairq_eoi(struct domain *d, unsigned int isairq)
+{
+ struct hvm_irq *hvm_irq = &d->arch.hvm_domain.irq;
+ struct hvm_irq_dpci *dpci = hvm_irq->dpci;
+ struct dev_intx_gsi_link *digl, *tmp;
+ int i;
+
+ ASSERT(isairq < NR_ISAIRQS);
+ if ( !iommu_enabled || !dpci ||
+ !test_bit(isairq, dpci->isairq_map) )
+ return;
+
+ /* Multiple mirq may be mapped to one isa irq */
+ for ( i = 0; i < NR_IRQS; i++ )
+ {
+ if ( !dpci->mirq[i].valid )
+ continue;
+
+ list_for_each_entry_safe ( digl, tmp,
+ &dpci->mirq[i].digl_list, list )
+ {
+ if ( hvm_irq->pci_link.route[digl->link] == isairq )
+ {
+ hvm_pci_intx_deassert(d, digl->device, digl->intx);
+ spin_lock(&dpci->dirq_lock);
+ if ( --dpci->mirq[i].pending == 0 )
+ {
+ spin_unlock(&dpci->dirq_lock);
+ gdprintk(XENLOG_INFO VTDPREFIX,
+ "hvm_dpci_isairq_eoi:: mirq = %x\n", i);
+ stop_timer(&dpci->hvm_timer[irq_to_vector(i)]);
+ pirq_guest_eoi(d, i);
+ }
+ else
+ spin_unlock(&dpci->dirq_lock);
+ }
+ }
+ }
+}
+
+void hvm_dpci_eoi(struct domain *d, unsigned int guest_gsi,
+ union vioapic_redir_entry *ent)
+{
+ struct hvm_irq_dpci *hvm_irq_dpci = d->arch.hvm_domain.irq.dpci;
+ uint32_t device, intx, machine_gsi;
+
+ if ( !iommu_enabled || (hvm_irq_dpci == NULL) ||
+ (guest_gsi >= NR_ISAIRQS &&
+ !hvm_irq_dpci->girq[guest_gsi].valid) )
+ return;
+
+ if ( guest_gsi < NR_ISAIRQS )
+ {
+ hvm_dpci_isairq_eoi(d, guest_gsi);
+ return;
+ }
+
+ machine_gsi = hvm_irq_dpci->girq[guest_gsi].machine_gsi;
+ device = hvm_irq_dpci->girq[guest_gsi].device;
+ intx = hvm_irq_dpci->girq[guest_gsi].intx;
+ hvm_pci_intx_deassert(d, device, intx);
+
+ spin_lock(&hvm_irq_dpci->dirq_lock);
+ if ( --hvm_irq_dpci->mirq[machine_gsi].pending == 0 )
+ {
+ spin_unlock(&hvm_irq_dpci->dirq_lock);
+
+ gdprintk(XENLOG_INFO VTDPREFIX,
+ "hvm_dpci_eoi:: mirq = %x\n", machine_gsi);
+ stop_timer(&hvm_irq_dpci->hvm_timer[irq_to_vector(machine_gsi)]);
+ if ( (ent == NULL) || !ent->fields.mask )
+ pirq_guest_eoi(d, machine_gsi);
+ }
+ else
+ spin_unlock(&hvm_irq_dpci->dirq_lock);
+}
diff -r 8c921adf4833 -r 42f6c206c951 xen/drivers/passthrough/iommu.c
--- /dev/null Thu Jan 01 00:00:00 1970 +0000
+++ b/xen/drivers/passthrough/iommu.c Thu Mar 20 12:35:40 2008 -0600
@@ -0,0 +1,136 @@
+/*
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope 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, Inc., 59 Temple
+ * Place - Suite 330, Boston, MA 02111-1307 USA.
+ */
+
+#include <xen/sched.h>
+#include <xen/iommu.h>
+
+extern struct iommu_ops intel_iommu_ops;
+extern struct iommu_ops amd_iommu_ops;
+
+int iommu_domain_init(struct domain *domain)
+{
+ struct hvm_iommu *hd = domain_hvm_iommu(domain);
+
+ spin_lock_init(&hd->mapping_lock);
+ spin_lock_init(&hd->iommu_list_lock);
+ INIT_LIST_HEAD(&hd->pdev_list);
+ INIT_LIST_HEAD(&hd->g2m_ioport_list);
+
+ if ( !iommu_enabled )
+ return 0;
+
+ switch ( boot_cpu_data.x86_vendor )
+ {
+ case X86_VENDOR_INTEL:
+ hd->platform_ops = &intel_iommu_ops;
+ break;
+ case X86_VENDOR_AMD:
+ hd->platform_ops = &amd_iommu_ops;
+ break;
+ default:
+ BUG();
+ }
+
+ return hd->platform_ops->init(domain);
+}
+
+int assign_device(struct domain *d, u8 bus, u8 devfn)
+{
+ struct hvm_iommu *hd = domain_hvm_iommu(d);
+
+ if ( !iommu_enabled || !hd->platform_ops )
+ return 0;
+
+ return hd->platform_ops->assign_device(d, bus, devfn);
+}
+
+void iommu_domain_destroy(struct domain *d)
+{
+ struct hvm_irq_dpci *hvm_irq_dpci = d->arch.hvm_domain.irq.dpci;
+ uint32_t i;
+ struct hvm_iommu *hd = domain_hvm_iommu(d);
+ struct list_head *ioport_list, *digl_list, *tmp;
+ struct g2m_ioport *ioport;
+ struct dev_intx_gsi_link *digl;
+
+ if ( !iommu_enabled || !hd->platform_ops )
+ return;
+
+ if ( hvm_irq_dpci != NULL )
+ {
+ for ( i = 0; i < NR_IRQS; i++ )
+ {
+ if ( !hvm_irq_dpci->mirq[i].valid )
+ continue;
+
+ pirq_guest_unbind(d, i);
+ kill_timer(&hvm_irq_dpci->hvm_timer[irq_to_vector(i)]);
+
+ list_for_each_safe ( digl_list, tmp,
+ &hvm_irq_dpci->mirq[i].digl_list )
+ {
+ digl = list_entry(digl_list,
+ struct dev_intx_gsi_link, list);
+ list_del(&digl->list);
+ xfree(digl);
+ }
+ }
+
+ d->arch.hvm_domain.irq.dpci = NULL;
+ xfree(hvm_irq_dpci);
+ }
+
+ if ( hd )
+ {
+ list_for_each_safe ( ioport_list, tmp, &hd->g2m_ioport_list )
+ {
+ ioport = list_entry(ioport_list, struct g2m_ioport, list);
+ list_del(&ioport->list);
+ xfree(ioport);
+ }
+ }
+
+ return hd->platform_ops->teardown(d);
+}
+
+int iommu_map_page(struct domain *d, unsigned long gfn, unsigned long mfn)
+{
+ struct hvm_iommu *hd = domain_hvm_iommu(d);
+
+ if ( !iommu_enabled || !hd->platform_ops )
+ return 0;
+
+ return hd->platform_ops->map_page(d, gfn, mfn);
+}
+
+int iommu_unmap_page(struct domain *d, unsigned long gfn)
+{
+ struct hvm_iommu *hd = domain_hvm_iommu(d);
+
+ if ( !iommu_enabled || !hd->platform_ops )
+ return 0;
+
+ return hd->platform_ops->unmap_page(d, gfn);
+}
+
+void deassign_device(struct domain *d, u8 bus, u8 devfn)
+{
+ struct hvm_iommu *hd = domain_hvm_iommu(d);
+
+ if ( !iommu_enabled || !hd->platform_ops )
+ return;
+
+ return hd->platform_ops->reassign_device(d, dom0, bus, devfn);
+}
diff -r 8c921adf4833 -r 42f6c206c951 xen/drivers/passthrough/vtd/Makefile
--- a/xen/drivers/passthrough/vtd/Makefile Fri Mar 14 15:07:45 2008 -0600
+++ b/xen/drivers/passthrough/vtd/Makefile Thu Mar 20 12:35:40 2008 -0600
@@ -1,6 +1,5 @@ obj-y += iommu.o
obj-y += iommu.o
obj-y += dmar.o
obj-y += utils.o
-obj-y += io.o
obj-y += qinval.o
obj-y += intremap.o
diff -r 8c921adf4833 -r 42f6c206c951 xen/drivers/passthrough/vtd/dmar.c
--- a/xen/drivers/passthrough/vtd/dmar.c Fri Mar 14 15:07:45 2008 -0600
+++ b/xen/drivers/passthrough/vtd/dmar.c Thu Mar 20 12:35:40 2008 -0600
@@ -555,13 +555,13 @@ static int __init acpi_parse_dmar(unsign
return -ENODEV;
}
- if ( !dmar->haw )
- {
- dprintk(XENLOG_WARNING VTDPREFIX, "Zero: Invalid DMAR haw\n");
+ if ( !dmar->width )
+ {
+ dprintk(XENLOG_WARNING VTDPREFIX, "Zero: Invalid DMAR width\n");
return -EINVAL;
}
- dmar_host_address_width = dmar->haw;
+ dmar_host_address_width = dmar->width;
dprintk(XENLOG_INFO VTDPREFIX, "Host address width %d\n",
dmar_host_address_width);
diff -r 8c921adf4833 -r 42f6c206c951 xen/drivers/passthrough/vtd/dmar.h
--- a/xen/drivers/passthrough/vtd/dmar.h Fri Mar 14 15:07:45 2008 -0600
+++ b/xen/drivers/passthrough/vtd/dmar.h Thu Mar 20 12:35:40 2008 -0600
@@ -22,7 +22,7 @@
#define _DMAR_H_
#include <xen/list.h>
-#include <asm/iommu.h>
+#include <xen/iommu.h>
extern u8 dmar_host_address_width;
@@ -100,5 +100,6 @@ struct acpi_rmrr_unit * acpi_find_matche
int vtd_hw_check(void);
void disable_pmr(struct iommu *iommu);
+int is_usb_device(struct pci_dev *pdev);
#endif // _DMAR_H_
diff -r 8c921adf4833 -r 42f6c206c951 xen/drivers/passthrough/vtd/extern.h
--- a/xen/drivers/passthrough/vtd/extern.h Fri Mar 14 15:07:45 2008 -0600
+++ b/xen/drivers/passthrough/vtd/extern.h Thu Mar 20 12:35:40 2008 -0600
@@ -42,8 +42,6 @@ int invalidate_sync(struct iommu *iommu)
int invalidate_sync(struct iommu *iommu);
int iommu_flush_iec_global(struct iommu *iommu);
int iommu_flush_iec_index(struct iommu *iommu, u8 im, u16 iidx);
-void print_iommu_regs(struct acpi_drhd_unit *drhd);
-int vtd_hw_check(void);
struct iommu * ioapic_to_iommu(unsigned int apic_id);
struct acpi_drhd_unit * ioapic_to_drhd(unsigned int apic_id);
void clear_fault_bits(struct iommu *iommu);
diff -r 8c921adf4833 -r 42f6c206c951 xen/drivers/passthrough/vtd/intremap.c
--- a/xen/drivers/passthrough/vtd/intremap.c Fri Mar 14 15:07:45 2008 -0600
+++ b/xen/drivers/passthrough/vtd/intremap.c Thu Mar 20 12:35:40 2008 -0600
@@ -18,28 +18,10 @@
* Copyright (C) Xiaohui Xin <xiaohui.xin@xxxxxxxxx>
*/
-#include <xen/config.h>
-#include <xen/lib.h>
-#include <xen/init.h>
#include <xen/irq.h>
-#include <xen/delay.h>
#include <xen/sched.h>
-#include <xen/acpi.h>
-#include <xen/keyhandler.h>
-#include <xen/spinlock.h>
-#include <asm/io.h>
-#include <asm/mc146818rtc.h>
-#include <asm/smp.h>
-#include <asm/desc.h>
-#include <mach_apic.h>
-#include <io_ports.h>
-
-#include <xen/spinlock.h>
-#include <xen/xmalloc.h>
-#include <xen/domain_page.h>
-#include <asm/delay.h>
-#include <asm/string.h>
-#include <asm/iommu.h>
+#include <xen/iommu.h>
+#include "iommu.h"
#include "dmar.h"
#include "vtd.h"
#include "../pci-direct.h"
@@ -172,7 +154,7 @@ io_apic_read_remap_rte(
struct iommu *iommu = ioapic_to_iommu(mp_ioapics[apic].mpc_apicid);
struct ir_ctrl *ir_ctrl = iommu_ir_ctrl(iommu);
- if ( !iommu || !(ir_ctrl->iremap) )
+ if ( !iommu || !ir_ctrl || !(ir_ctrl->iremap) )
{
*IO_APIC_BASE(apic) = reg;
return *(IO_APIC_BASE(apic)+4);
@@ -218,7 +200,7 @@ io_apic_write_remap_rte(
struct iommu *iommu = ioapic_to_iommu(mp_ioapics[apic].mpc_apicid);
struct ir_ctrl *ir_ctrl = iommu_ir_ctrl(iommu);
- if ( !iommu || !(ir_ctrl->iremap) )
+ if ( !iommu || !ir_ctrl || !(ir_ctrl->iremap) )
{
*IO_APIC_BASE(apic) = reg;
*(IO_APIC_BASE(apic)+4) = value;
diff -r 8c921adf4833 -r 42f6c206c951 xen/drivers/passthrough/vtd/io.c
--- a/xen/drivers/passthrough/vtd/io.c Fri Mar 14 15:07:45 2008 -0600
+++ /dev/null Thu Jan 01 00:00:00 1970 +0000
@@ -1,296 +0,0 @@
-/*
- * Copyright (c) 2006, Intel Corporation.
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms and conditions of the GNU General Public License,
- * version 2, as published by the Free Software Foundation.
- *
- * This program is distributed in the hope 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, Inc., 59 Temple
- * Place - Suite 330, Boston, MA 02111-1307 USA.
- *
- * Copyright (C) Allen Kay <allen.m.kay@xxxxxxxxx>
- * Copyright (C) Xiaohui Xin <xiaohui.xin@xxxxxxxxx>
- */
-
-#include <xen/init.h>
-#include <xen/config.h>
-#include <xen/init.h>
-#include <xen/mm.h>
-#include <xen/lib.h>
-#include <xen/errno.h>
-#include <xen/trace.h>
-#include <xen/event.h>
-#include <xen/hypercall.h>
-#include <asm/current.h>
-#include <asm/cpufeature.h>
-#include <asm/processor.h>
-#include <asm/msr.h>
-#include <asm/apic.h>
-#include <asm/paging.h>
-#include <asm/shadow.h>
-#include <asm/p2m.h>
-#include <asm/hvm/hvm.h>
-#include <asm/hvm/support.h>
-#include <asm/hvm/vpt.h>
-#include <asm/hvm/vpic.h>
-#include <asm/hvm/vlapic.h>
-#include <public/sched.h>
-#include <xen/iocap.h>
-#include <public/hvm/ioreq.h>
-#include <public/domctl.h>
-
-static void pt_irq_time_out(void *data)
-{
- struct hvm_mirq_dpci_mapping *irq_map = data;
- unsigned int guest_gsi, machine_gsi = 0;
- struct hvm_irq_dpci *dpci = irq_map->dom->arch.hvm_domain.irq.dpci;
- struct dev_intx_gsi_link *digl;
- uint32_t device, intx;
-
- list_for_each_entry ( digl, &irq_map->digl_list, list )
- {
- guest_gsi = digl->gsi;
- machine_gsi = dpci->girq[guest_gsi].machine_gsi;
- device = digl->device;
- intx = digl->intx;
- hvm_pci_intx_deassert(irq_map->dom, device, intx);
- }
-
- clear_bit(machine_gsi, dpci->dirq_mask);
- stop_timer(&dpci->hvm_timer[irq_to_vector(machine_gsi)]);
- spin_lock(&dpci->dirq_lock);
- dpci->mirq[machine_gsi].pending = 0;
- spin_unlock(&dpci->dirq_lock);
- pirq_guest_eoi(irq_map->dom, machine_gsi);
-}
-
-int pt_irq_create_bind_vtd(
- struct domain *d, xen_domctl_bind_pt_irq_t *pt_irq_bind)
-{
- struct hvm_irq_dpci *hvm_irq_dpci = d->arch.hvm_domain.irq.dpci;
- uint32_t machine_gsi, guest_gsi;
- uint32_t device, intx, link;
- struct dev_intx_gsi_link *digl;
-
- if ( hvm_irq_dpci == NULL )
- {
- hvm_irq_dpci = xmalloc(struct hvm_irq_dpci);
- if ( hvm_irq_dpci == NULL )
- return -ENOMEM;
-
- memset(hvm_irq_dpci, 0, sizeof(*hvm_irq_dpci));
- spin_lock_init(&hvm_irq_dpci->dirq_lock);
- for ( int i = 0; i < NR_IRQS; i++ )
- INIT_LIST_HEAD(&hvm_irq_dpci->mirq[i].digl_list);
-
- if ( cmpxchg((unsigned long *)&d->arch.hvm_domain.irq.dpci,
- 0, (unsigned long)hvm_irq_dpci) != 0 )
- xfree(hvm_irq_dpci);
-
- hvm_irq_dpci = d->arch.hvm_domain.irq.dpci;
- }
-
- machine_gsi = pt_irq_bind->machine_irq;
- device = pt_irq_bind->u.pci.device;
- intx = pt_irq_bind->u.pci.intx;
- guest_gsi = hvm_pci_intx_gsi(device, intx);
- link = hvm_pci_intx_link(device, intx);
- hvm_irq_dpci->link_cnt[link]++;
-
- digl = xmalloc(struct dev_intx_gsi_link);
- if ( !digl )
- return -ENOMEM;
-
- digl->device = device;
- digl->intx = intx;
- digl->gsi = guest_gsi;
- digl->link = link;
- list_add_tail(&digl->list,
- &hvm_irq_dpci->mirq[machine_gsi].digl_list);
-
- hvm_irq_dpci->girq[guest_gsi].valid = 1;
- hvm_irq_dpci->girq[guest_gsi].device = device;
- hvm_irq_dpci->girq[guest_gsi].intx = intx;
- hvm_irq_dpci->girq[guest_gsi].machine_gsi = machine_gsi;
-
- /* Bind the same mirq once in the same domain */
- if ( !hvm_irq_dpci->mirq[machine_gsi].valid )
- {
- hvm_irq_dpci->mirq[machine_gsi].valid = 1;
- hvm_irq_dpci->mirq[machine_gsi].dom = d;
-
- init_timer(&hvm_irq_dpci->hvm_timer[irq_to_vector(machine_gsi)],
- pt_irq_time_out, &hvm_irq_dpci->mirq[machine_gsi], 0);
- /* Deal with gsi for legacy devices */
- pirq_guest_bind(d->vcpu[0], machine_gsi, BIND_PIRQ__WILL_SHARE);
- }
-
- gdprintk(XENLOG_INFO VTDPREFIX,
- "VT-d irq bind: m_irq = %x device = %x intx = %x\n",
- machine_gsi, device, intx);
- return 0;
-}
-
-int pt_irq_destroy_bind_vtd(
- struct domain *d, xen_domctl_bind_pt_irq_t *pt_irq_bind)
-{
- struct hvm_irq_dpci *hvm_irq_dpci = d->arch.hvm_domain.irq.dpci;
- uint32_t machine_gsi, guest_gsi;
- uint32_t device, intx, link;
- struct list_head *digl_list, *tmp;
- struct dev_intx_gsi_link *digl;
-
- if ( hvm_irq_dpci == NULL )
- return 0;
-
- machine_gsi = pt_irq_bind->machine_irq;
- device = pt_irq_bind->u.pci.device;
- intx = pt_irq_bind->u.pci.intx;
- guest_gsi = hvm_pci_intx_gsi(device, intx);
- link = hvm_pci_intx_link(device, intx);
- hvm_irq_dpci->link_cnt[link]--;
-
- gdprintk(XENLOG_INFO,
- "pt_irq_destroy_bind_vtd: machine_gsi=%d, guest_gsi=%d, device=%d,
intx=%d.\n",
- machine_gsi, guest_gsi, device, intx);
- memset(&hvm_irq_dpci->girq[guest_gsi], 0, sizeof(struct
hvm_girq_dpci_mapping));
-
- /* clear the mirq info */
- if ( hvm_irq_dpci->mirq[machine_gsi].valid )
- {
-
- list_for_each_safe ( digl_list, tmp,
- &hvm_irq_dpci->mirq[machine_gsi].digl_list )
- {
- digl = list_entry(digl_list,
- struct dev_intx_gsi_link, list);
- if ( digl->device == device &&
- digl->intx == intx &&
- digl->link == link &&
- digl->gsi == guest_gsi )
- {
- list_del(&digl->list);
- xfree(digl);
- }
- }
-
- if ( list_empty(&hvm_irq_dpci->mirq[machine_gsi].digl_list) )
- {
- pirq_guest_unbind(d, machine_gsi);
- kill_timer(&hvm_irq_dpci->hvm_timer[irq_to_vector(machine_gsi)]);
- hvm_irq_dpci->mirq[machine_gsi].dom = NULL;
- hvm_irq_dpci->mirq[machine_gsi].valid = 0;
- }
- }
-
- gdprintk(XENLOG_INFO,
- "XEN_DOMCTL_irq_unmapping: m_irq = %x device = %x intx = %x\n",
- machine_gsi, device, intx);
-
- return 0;
-}
-
-int hvm_do_IRQ_dpci(struct domain *d, unsigned int mirq)
-{
- struct hvm_irq *hvm_irq = &d->arch.hvm_domain.irq;
-
- if ( !iommu_enabled || (d == dom0) || (hvm_irq->dpci == NULL) ||
- !hvm_irq->dpci->mirq[mirq].valid )
- return 0;
-
- /*
- * Set a timer here to avoid situations where the IRQ line is shared, and
- * the device belonging to the pass-through guest is not yet active. In
- * this case the guest may not pick up the interrupt (e.g., masked at the
- * PIC) and we need to detect that.
- */
- set_bit(mirq, hvm_irq->dpci->dirq_mask);
- set_timer(&hvm_irq->dpci->hvm_timer[irq_to_vector(mirq)],
- NOW() + PT_IRQ_TIME_OUT);
- vcpu_kick(d->vcpu[0]);
-
- return 1;
-}
-
-static void hvm_dpci_isairq_eoi(struct domain *d, unsigned int isairq)
-{
- struct hvm_irq *hvm_irq = &d->arch.hvm_domain.irq;
- struct hvm_irq_dpci *dpci = hvm_irq->dpci;
- struct dev_intx_gsi_link *digl, *tmp;
- int i;
-
- ASSERT(isairq < NR_ISAIRQS);
- if ( !iommu_enabled || !dpci ||
- !test_bit(isairq, dpci->isairq_map) )
- return;
-
- /* Multiple mirq may be mapped to one isa irq */
- for ( i = 0; i < NR_IRQS; i++ )
- {
- if ( !dpci->mirq[i].valid )
- continue;
-
- list_for_each_entry_safe ( digl, tmp,
- &dpci->mirq[i].digl_list, list )
- {
- if ( hvm_irq->pci_link.route[digl->link] == isairq )
- {
- hvm_pci_intx_deassert(d, digl->device, digl->intx);
- spin_lock(&dpci->dirq_lock);
- if ( --dpci->mirq[i].pending == 0 )
- {
- spin_unlock(&dpci->dirq_lock);
- gdprintk(XENLOG_INFO VTDPREFIX,
- "hvm_dpci_isairq_eoi:: mirq = %x\n", i);
- stop_timer(&dpci->hvm_timer[irq_to_vector(i)]);
- pirq_guest_eoi(d, i);
- }
- else
- spin_unlock(&dpci->dirq_lock);
- }
- }
- }
-}
-
-void hvm_dpci_eoi(struct domain *d, unsigned int guest_gsi,
- union vioapic_redir_entry *ent)
-{
- struct hvm_irq_dpci *hvm_irq_dpci = d->arch.hvm_domain.irq.dpci;
- uint32_t device, intx, machine_gsi;
-
- if ( !iommu_enabled || (hvm_irq_dpci == NULL) ||
- (guest_gsi >= NR_ISAIRQS &&
- !hvm_irq_dpci->girq[guest_gsi].valid) )
- return;
-
- if ( guest_gsi < NR_ISAIRQS )
- {
- hvm_dpci_isairq_eoi(d, guest_gsi);
- return;
- }
-
- machine_gsi = hvm_irq_dpci->girq[guest_gsi].machine_gsi;
- device = hvm_irq_dpci->girq[guest_gsi].device;
- intx = hvm_irq_dpci->girq[guest_gsi].intx;
- hvm_pci_intx_deassert(d, device, intx);
-
- spin_lock(&hvm_irq_dpci->dirq_lock);
- if ( --hvm_irq_dpci->mirq[machine_gsi].pending == 0 )
- {
- spin_unlock(&hvm_irq_dpci->dirq_lock);
-
- gdprintk(XENLOG_INFO VTDPREFIX,
- "hvm_dpci_eoi:: mirq = %x\n", machine_gsi);
- stop_timer(&hvm_irq_dpci->hvm_timer[irq_to_vector(machine_gsi)]);
- if ( (ent == NULL) || !ent->fields.mask )
- pirq_guest_eoi(d, machine_gsi);
- }
- else
- spin_unlock(&hvm_irq_dpci->dirq_lock);
-}
diff -r 8c921adf4833 -r 42f6c206c951 xen/drivers/passthrough/vtd/iommu.c
--- a/xen/drivers/passthrough/vtd/iommu.c Fri Mar 14 15:07:45 2008 -0600
+++ b/xen/drivers/passthrough/vtd/iommu.c Thu Mar 20 12:35:40 2008 -0600
@@ -19,17 +19,12 @@
* Copyright (C) Allen Kay <allen.m.kay@xxxxxxxxx> - adapted to xen
*/
-#include <xen/init.h>
#include <xen/irq.h>
-#include <xen/spinlock.h>
#include <xen/sched.h>
#include <xen/xmalloc.h>
#include <xen/domain_page.h>
-#include <asm/delay.h>
-#include <asm/string.h>
-#include <asm/mm.h>
-#include <asm/iommu.h>
-#include <asm/hvm/vmx/intel-iommu.h>
+#include <xen/iommu.h>
+#include "iommu.h"
#include "dmar.h"
#include "../pci-direct.h"
#include "../pci_regs.h"
@@ -39,8 +34,8 @@
#define domain_iommu_domid(d) ((d)->arch.hvm_domain.hvm_iommu.iommu_domid)
static spinlock_t domid_bitmap_lock; /* protect domain id bitmap */
-static int domid_bitmap_size; /* domain id bitmap size in bit */
-static void *domid_bitmap; /* iommu domain id bitmap */
+static int domid_bitmap_size; /* domain id bitmap size in bits */
+static unsigned long *domid_bitmap; /* iommu domain id bitmap */
#define DID_FIELD_WIDTH 16
#define DID_HIGH_OFFSET 8
@@ -72,6 +67,93 @@ static void iommu_domid_release(struct d
d->arch.hvm_domain.hvm_iommu.iommu_domid = 0;
clear_bit(iommu_domid, domid_bitmap);
}
+}
+
+static struct intel_iommu *alloc_intel_iommu(void)
+{
+ struct intel_iommu *intel;
+
+ intel = xmalloc(struct intel_iommu);
+ if ( !intel )
+ {
+ gdprintk(XENLOG_ERR VTDPREFIX,
+ "Allocate intel_iommu failed.\n");
+ return NULL;
+ }
+ memset(intel, 0, sizeof(struct intel_iommu));
+
+ spin_lock_init(&intel->qi_ctrl.qinval_lock);
+ spin_lock_init(&intel->qi_ctrl.qinval_poll_lock);
+
+ spin_lock_init(&intel->ir_ctrl.iremap_lock);
+
+ return intel;
+}
+
+static void free_intel_iommu(struct intel_iommu *intel)
+{
+ if ( intel )
+ {
+ xfree(intel);
+ intel = NULL;
+ }
+}
+
+struct qi_ctrl *iommu_qi_ctrl(struct iommu *iommu)
+{
+ if ( !iommu )
+ return NULL;
+
+ if ( !iommu->intel )
+ {
+ iommu->intel = alloc_intel_iommu();
+ if ( !iommu->intel )
+ {
+ dprintk(XENLOG_ERR VTDPREFIX,
+ "iommu_qi_ctrl: Allocate iommu->intel failed.\n");
+ return NULL;
+ }
+ }
+
+ return &(iommu->intel->qi_ctrl);
+}
+
+struct ir_ctrl *iommu_ir_ctrl(struct iommu *iommu)
+{
+ if ( !iommu )
+ return NULL;
+
+ if ( !iommu->intel )
+ {
+ iommu->intel = alloc_intel_iommu();
+ if ( !iommu->intel )
+ {
+ dprintk(XENLOG_ERR VTDPREFIX,
+ "iommu_ir_ctrl: Allocate iommu->intel failed.\n");
+ return NULL;
+ }
+ }
+
+ return &(iommu->intel->ir_ctrl);
+}
+
+struct iommu_flush *iommu_get_flush(struct iommu *iommu)
+{
+ if ( !iommu )
+ return NULL;
+
+ if ( !iommu->intel )
+ {
+ iommu->intel = alloc_intel_iommu();
+ if ( !iommu->intel )
+ {
+ dprintk(XENLOG_ERR VTDPREFIX,
+ "iommu_get_flush: Allocate iommu->intel failed.\n");
+ return NULL;
+ }
+ }
+
+ return &(iommu->intel->flush);
}
unsigned int x86_clflush_size;
@@ -756,40 +838,34 @@ static int iommu_page_fault_do_one(struc
PCI_SLOT(source_id & 0xFF), PCI_FUNC(source_id & 0xFF), addr,
fault_reason, iommu->reg);
- if (fault_reason < 0x20)
+ if ( fault_reason < 0x20 )
print_vtd_entries(current->domain, iommu, (source_id >> 8),
- (source_id & 0xff), (addr >> PAGE_SHIFT));
+ (source_id & 0xff), (addr >> PAGE_SHIFT));
return 0;
}
static void iommu_fault_status(u32 fault_status)
{
- if (fault_status & DMA_FSTS_PFO)
+ if ( fault_status & DMA_FSTS_PFO )
dprintk(XENLOG_ERR VTDPREFIX,
"iommu_fault_status: Fault Overflow\n");
- else
- if (fault_status & DMA_FSTS_PPF)
+ else if ( fault_status & DMA_FSTS_PPF )
dprintk(XENLOG_ERR VTDPREFIX,
"iommu_fault_status: Primary Pending Fault\n");
- else
- if (fault_status & DMA_FSTS_AFO)
+ else if ( fault_status & DMA_FSTS_AFO )
dprintk(XENLOG_ERR VTDPREFIX,
"iommu_fault_status: Advanced Fault Overflow\n");
- else
- if (fault_status & DMA_FSTS_APF)
+ else if ( fault_status & DMA_FSTS_APF )
dprintk(XENLOG_ERR VTDPREFIX,
"iommu_fault_status: Advanced Pending Fault\n");
- else
- if (fault_status & DMA_FSTS_IQE)
+ else if ( fault_status & DMA_FSTS_IQE )
dprintk(XENLOG_ERR VTDPREFIX,
"iommu_fault_status: Invalidation Queue Error\n");
- else
- if (fault_status & DMA_FSTS_ICE)
+ else if ( fault_status & DMA_FSTS_ICE )
dprintk(XENLOG_ERR VTDPREFIX,
"iommu_fault_status: Invalidation Completion Error\n");
- else
- if (fault_status & DMA_FSTS_ITE)
+ else if ( fault_status & DMA_FSTS_ITE )
dprintk(XENLOG_ERR VTDPREFIX,
"iommu_fault_status: Invalidation Time-out Error\n");
}
@@ -976,8 +1052,6 @@ struct iommu *iommu_alloc(void *hw_data)
{
struct acpi_drhd_unit *drhd = (struct acpi_drhd_unit *) hw_data;
struct iommu *iommu;
- struct qi_ctrl *qi_ctrl;
- struct ir_ctrl *ir_ctrl;
if ( nr_iommus > MAX_IOMMUS )
{
@@ -1014,12 +1088,7 @@ struct iommu *iommu_alloc(void *hw_data)
spin_lock_init(&iommu->lock);
spin_lock_init(&iommu->register_lock);
- qi_ctrl = iommu_qi_ctrl(iommu);
- spin_lock_init(&qi_ctrl->qinval_lock);
- spin_lock_init(&qi_ctrl->qinval_poll_lock);
-
- ir_ctrl = iommu_ir_ctrl(iommu);
- spin_lock_init(&ir_ctrl->iremap_lock);
+ iommu->intel = alloc_intel_iommu();
drhd->iommu = iommu;
return iommu;
@@ -1036,6 +1105,7 @@ static void free_iommu(struct iommu *iom
free_xenheap_page((void *)iommu->root_entry);
if ( iommu->reg )
iounmap(iommu->reg);
+ free_intel_iommu(iommu->intel);
free_irq(iommu->vector);
xfree(iommu);
}
@@ -1063,7 +1133,7 @@ int intel_iommu_domain_init(struct domai
iommu = drhd->iommu ? : iommu_alloc(drhd);
/* calculate AGAW */
- if (guest_width > cap_mgaw(iommu->cap))
+ if ( guest_width > cap_mgaw(iommu->cap) )
guest_width = cap_mgaw(iommu->cap);
adjust_width = guestwidth_to_adjustwidth(guest_width);
agaw = width_to_agaw(adjust_width);
@@ -1885,7 +1955,8 @@ int iommu_setup(void)
/* Allocate domain id bitmap, and set bit 0 as reserved */
domid_bitmap_size = cap_ndoms(iommu->cap);
- domid_bitmap = xmalloc_bytes(domid_bitmap_size / 8);
+ domid_bitmap = xmalloc_array(unsigned long,
+ BITS_TO_LONGS(domid_bitmap_size));
if ( domid_bitmap == NULL )
goto error;
memset(domid_bitmap, 0, domid_bitmap_size / 8);
@@ -1948,6 +2019,12 @@ int intel_iommu_assign_device(struct dom
for_each_rmrr_device( rmrr, pdev )
if ( pdev->bus == bus && pdev->devfn == devfn )
{
+ /* FIXME: Because USB RMRR conflicts with guest bios region,
+ * ignore USB RMRR temporarily.
+ */
+ if ( is_usb_device(pdev) )
+ return 0;
+
ret = iommu_prepare_rmrr_dev(d, rmrr, pdev);
if ( ret )
{
diff -r 8c921adf4833 -r 42f6c206c951 xen/drivers/passthrough/vtd/iommu.h
--- /dev/null Thu Jan 01 00:00:00 1970 +0000
+++ b/xen/drivers/passthrough/vtd/iommu.h Thu Mar 20 12:35:40 2008 -0600
@@ -0,0 +1,454 @@
+/*
+ * Copyright (c) 2006, Intel Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope 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, Inc., 59 Temple
+ * Place - Suite 330, Boston, MA 02111-1307 USA.
+ *
+ * Copyright (C) Ashok Raj <ashok.raj@xxxxxxxxx>
+ */
+
+#ifndef _INTEL_IOMMU_H_
+#define _INTEL_IOMMU_H_
+
+#include <xen/types.h>
+
+/*
+ * Intel IOMMU register specification per version 1.0 public spec.
+ */
+
+#define DMAR_VER_REG 0x0 /* Arch version supported by this IOMMU */
+#define DMAR_CAP_REG 0x8 /* Hardware supported capabilities */
+#define DMAR_ECAP_REG 0x10 /* Extended capabilities supported */
+#define DMAR_GCMD_REG 0x18 /* Global command register */
+#define DMAR_GSTS_REG 0x1c /* Global status register */
+#define DMAR_RTADDR_REG 0x20 /* Root entry table */
+#define DMAR_CCMD_REG 0x28 /* Context command reg */
+#define DMAR_FSTS_REG 0x34 /* Fault Status register */
+#define DMAR_FECTL_REG 0x38 /* Fault control register */
+#define DMAR_FEDATA_REG 0x3c /* Fault event interrupt data register */
+#define DMAR_FEADDR_REG 0x40 /* Fault event interrupt addr register */
+#define DMAR_FEUADDR_REG 0x44 /* Upper address register */
+#define DMAR_AFLOG_REG 0x58 /* Advanced Fault control */
+#define DMAR_PMEN_REG 0x64 /* Enable Protected Memory Region */
+#define DMAR_PLMBASE_REG 0x68 /* PMRR Low addr */
+#define DMAR_PLMLIMIT_REG 0x6c /* PMRR low limit */
+#define DMAR_PHMBASE_REG 0x70 /* pmrr high base addr */
+#define DMAR_PHMLIMIT_REG 0x78 /* pmrr high limit */
+#define DMAR_IQH_REG 0x80 /* invalidation queue head */
+#define DMAR_IQT_REG 0x88 /* invalidation queue tail */
+#define DMAR_IQA_REG 0x90 /* invalidation queue addr */
+#define DMAR_IRTA_REG 0xB8 /* intr remap */
+
+#define OFFSET_STRIDE (9)
+#define dmar_readl(dmar, reg) readl(dmar + reg)
+#define dmar_writel(dmar, reg, val) writel(val, dmar + reg)
+#define dmar_readq(dmar, reg) ({ \
+ u32 lo, hi; \
+ lo = dmar_readl(dmar, reg); \
+ hi = dmar_readl(dmar, reg + 4); \
+ (((u64) hi) << 32) + lo; })
+#define dmar_writeq(dmar, reg, val) do {\
+ dmar_writel(dmar, reg, (u32)val); \
+ dmar_writel(dmar, reg + 4, (u32)((u64) val >> 32)); \
+ } while (0)
+
+#define VER_MAJOR(v) (((v) & 0xf0) >> 4)
+#define VER_MINOR(v) ((v) & 0x0f)
+
+/*
+ * Decoding Capability Register
+ */
+#define cap_read_drain(c) (((c) >> 55) & 1)
+#define cap_write_drain(c) (((c) >> 54) & 1)
+#define cap_max_amask_val(c) (((c) >> 48) & 0x3f)
+#define cap_num_fault_regs(c) ((((c) >> 40) & 0xff) + 1)
+#define cap_pgsel_inv(c) (((c) >> 39) & 1)
+
+#define cap_super_page_val(c) (((c) >> 34) & 0xf)
+#define cap_super_offset(c) (((find_first_bit(&cap_super_page_val(c), 4)) \
+ * OFFSET_STRIDE) + 21)
+
+#define cap_fault_reg_offset(c) ((((c) >> 24) & 0x3ff) * 16)
+
+#define cap_isoch(c) (((c) >> 23) & 1)
+#define cap_qos(c) (((c) >> 22) & 1)
+#define cap_mgaw(c) ((((c) >> 16) & 0x3f) + 1)
+#define cap_sagaw(c) (((c) >> 8) & 0x1f)
+#define cap_caching_mode(c) (((c) >> 7) & 1)
+#define cap_phmr(c) (((c) >> 6) & 1)
+#define cap_plmr(c) (((c) >> 5) & 1)
+#define cap_rwbf(c) (((c) >> 4) & 1)
+#define cap_afl(c) (((c) >> 3) & 1)
+#define cap_ndoms(c) (1 << (4 + 2 * ((c) & 0x7)))
+
+/*
+ * Extended Capability Register
+ */
+
+#define ecap_niotlb_iunits(e) ((((e) >> 24) & 0xff) + 1)
+#define ecap_iotlb_offset(e) ((((e) >> 8) & 0x3ff) * 16)
+#define ecap_coherent(e) ((e >> 0) & 0x1)
+#define ecap_queued_inval(e) ((e >> 1) & 0x1)
+#define ecap_dev_iotlb(e) ((e >> 2) & 0x1)
+#define ecap_intr_remap(e) ((e >> 3) & 0x1)
+#define ecap_ext_intr(e) ((e >> 4) & 0x1)
+#define ecap_cache_hints(e) ((e >> 5) & 0x1)
+#define ecap_pass_thru(e) ((e >> 6) & 0x1)
+
+/* IOTLB_REG */
+#define DMA_TLB_FLUSH_GRANU_OFFSET 60
+#define DMA_TLB_GLOBAL_FLUSH (((u64)1) << 60)
+#define DMA_TLB_DSI_FLUSH (((u64)2) << 60)
+#define DMA_TLB_PSI_FLUSH (((u64)3) << 60)
+#define DMA_TLB_IIRG(x) (((x) >> 60) & 7)
+#define DMA_TLB_IAIG(val) (((val) >> 57) & 7)
+#define DMA_TLB_DID(x) (((u64)(x & 0xffff)) << 32)
+
+#define DMA_TLB_READ_DRAIN (((u64)1) << 49)
+#define DMA_TLB_WRITE_DRAIN (((u64)1) << 48)
+#define DMA_TLB_IVT (((u64)1) << 63)
+
+#define DMA_TLB_IVA_ADDR(x) ((((u64)x) >> 12) << 12)
+#define DMA_TLB_IVA_HINT(x) ((((u64)x) & 1) << 6)
+
+/* GCMD_REG */
+#define DMA_GCMD_TE (((u64)1) << 31)
+#define DMA_GCMD_SRTP (((u64)1) << 30)
+#define DMA_GCMD_SFL (((u64)1) << 29)
+#define DMA_GCMD_EAFL (((u64)1) << 28)
+#define DMA_GCMD_WBF (((u64)1) << 27)
+#define DMA_GCMD_QIE (((u64)1) << 26)
+#define DMA_GCMD_IRE (((u64)1) << 25)
+#define DMA_GCMD_SIRTP (((u64)1) << 24)
+#define DMA_GCMD_CFI (((u64)1) << 23)
+
+/* GSTS_REG */
+#define DMA_GSTS_TES (((u64)1) << 31)
+#define DMA_GSTS_RTPS (((u64)1) << 30)
+#define DMA_GSTS_FLS (((u64)1) << 29)
+#define DMA_GSTS_AFLS (((u64)1) << 28)
+#define DMA_GSTS_WBFS (((u64)1) << 27)
+#define DMA_GSTS_QIES (((u64)1) <<26)
+#define DMA_GSTS_IRES (((u64)1) <<25)
+#define DMA_GSTS_SIRTPS (((u64)1) << 24)
+#define DMA_GSTS_CFIS (((u64)1) <<23)
+
+/* PMEN_REG */
+#define DMA_PMEN_EPM (((u32)1) << 31)
+#define DMA_PMEN_PRS (((u32)1) << 0)
+
+/* CCMD_REG */
+#define DMA_CCMD_INVL_GRANU_OFFSET 61
+#define DMA_CCMD_ICC (((u64)1) << 63)
+#define DMA_CCMD_GLOBAL_INVL (((u64)1) << 61)
+#define DMA_CCMD_DOMAIN_INVL (((u64)2) << 61)
+#define DMA_CCMD_DEVICE_INVL (((u64)3) << 61)
+#define DMA_CCMD_FM(m) (((u64)((m) & 0x3)) << 32)
+#define DMA_CCMD_CIRG(x) ((((u64)3) << 61) & x)
+#define DMA_CCMD_MASK_NOBIT 0
+#define DMA_CCMD_MASK_1BIT 1
+#define DMA_CCMD_MASK_2BIT 2
+#define DMA_CCMD_MASK_3BIT 3
+#define DMA_CCMD_SID(s) (((u64)((s) & 0xffff)) << 16)
+#define DMA_CCMD_DID(d) ((u64)((d) & 0xffff))
+
+#define DMA_CCMD_CAIG_MASK(x) (((u64)x) & ((u64) 0x3 << 59))
+
+/* FECTL_REG */
+#define DMA_FECTL_IM (((u64)1) << 31)
+
+/* FSTS_REG */
+#define DMA_FSTS_PFO ((u64)1 << 0)
+#define DMA_FSTS_PPF ((u64)1 << 1)
+#define DMA_FSTS_AFO ((u64)1 << 2)
+#define DMA_FSTS_APF ((u64)1 << 3)
+#define DMA_FSTS_IQE ((u64)1 << 4)
+#define DMA_FSTS_ICE ((u64)1 << 5)
+#define DMA_FSTS_ITE ((u64)1 << 6)
+#define DMA_FSTS_FAULTS DMA_FSTS_PFO | DMA_FSTS_PPF | DMA_FSTS_AFO |
DMA_FSTS_APF | DMA_FSTS_IQE | DMA_FSTS_ICE | DMA_FSTS_ITE
+#define dma_fsts_fault_record_index(s) (((s) >> 8) & 0xff)
+
+/* FRCD_REG, 32 bits access */
+#define DMA_FRCD_F (((u64)1) << 31)
+#define dma_frcd_type(d) ((d >> 30) & 1)
+#define dma_frcd_fault_reason(c) (c & 0xff)
+#define dma_frcd_source_id(c) (c & 0xffff)
+#define dma_frcd_page_addr(d) (d & (((u64)-1) << 12)) /* low 64 bit */
+
+/*
+ * 0: Present
+ * 1-11: Reserved
+ * 12-63: Context Ptr (12 - (haw-1))
+ * 64-127: Reserved
+ */
+struct root_entry {
+ u64 val;
+ u64 rsvd1;
+};
+#define root_present(root) ((root).val & 1)
+#define set_root_present(root) do {(root).val |= 1;} while(0)
+#define get_context_addr(root) ((root).val & PAGE_MASK_4K)
+#define set_root_value(root, value) \
+ do {(root).val |= ((value) & PAGE_MASK_4K);} while(0)
+
+struct context_entry {
+ u64 lo;
+ u64 hi;
+};
+#define ROOT_ENTRY_NR (PAGE_SIZE_4K/sizeof(struct root_entry))
+#define context_present(c) ((c).lo & 1)
+#define context_fault_disable(c) (((c).lo >> 1) & 1)
+#define context_translation_type(c) (((c).lo >> 2) & 3)
+#define context_address_root(c) ((c).lo & PAGE_MASK_4K)
+#define context_address_width(c) ((c).hi & 7)
+#define context_domain_id(c) (((c).hi >> 8) & ((1 << 16) - 1))
+
+#define context_set_present(c) do {(c).lo |= 1;} while(0)
+#define context_clear_present(c) do {(c).lo &= ~1;} while(0)
+#define context_set_fault_enable(c) \
+ do {(c).lo &= (((u64)-1) << 2) | 1;} while(0)
+
+#define context_set_translation_type(c, val) do { \
+ (c).lo &= (((u64)-1) << 4) | 3; \
+ (c).lo |= (val & 3) << 2; \
+ } while(0)
+#define CONTEXT_TT_MULTI_LEVEL 0
+#define CONTEXT_TT_DEV_IOTLB 1
+#define CONTEXT_TT_PASS_THRU 2
+
+#define context_set_address_root(c, val) \
+ do {(c).lo &= 0xfff; (c).lo |= (val) & PAGE_MASK_4K ;} while(0)
+#define context_set_address_width(c, val) \
+ do {(c).hi &= 0xfffffff8; (c).hi |= (val) & 7;} while(0)
+#define context_clear_entry(c) do {(c).lo = 0; (c).hi = 0;} while(0)
+
+/* page table handling */
+#define LEVEL_STRIDE (9)
+#define LEVEL_MASK ((1 << LEVEL_STRIDE) - 1)
+#define agaw_to_level(val) ((val) + 2)
+#define agaw_to_width(val) (30 + val * LEVEL_STRIDE)
+#define width_to_agaw(w) ((w - 30)/LEVEL_STRIDE)
+#define level_to_offset_bits(l) (12 + (l - 1) * LEVEL_STRIDE)
+#define address_level_offset(addr, level) \
+ ((addr >> level_to_offset_bits(level)) & LEVEL_MASK)
+#define level_mask(l) (((u64)(-1)) << level_to_offset_bits(l))
+#define level_size(l) (1 << level_to_offset_bits(l))
+#define align_to_level(addr, l) ((addr + level_size(l) - 1) & level_mask(l))
+
+/*
+ * 0: readable
+ * 1: writable
+ * 2-6: reserved
+ * 7: super page
+ * 8-11: available
+ * 12-63: Host physcial address
+ */
+struct dma_pte {
+ u64 val;
+};
+#define dma_clear_pte(p) do {(p).val = 0;} while(0)
+#define dma_set_pte_readable(p) do {(p).val |= 1;} while(0)
+#define dma_set_pte_writable(p) do {(p).val |= 2;} while(0)
+#define dma_set_pte_superpage(p) do {(p).val |= 8;} while(0)
+#define dma_set_pte_prot(p, prot) do { (p).val = (((p).val >> 2) << 2) |
((prot) & 3);} while (0)
+#define dma_pte_addr(p) ((p).val & PAGE_MASK_4K)
+#define dma_set_pte_addr(p, addr) do {(p).val |= ((addr) >> PAGE_SHIFT_4K) <<
PAGE_SHIFT_4K;} while(0)
+#define DMA_PTE_READ (1)
+#define DMA_PTE_WRITE (2)
+#define dma_pte_present(p) (((p).val & 3) != 0)
+
+/* interrupt remap entry */
+struct iremap_entry {
+ union {
+ u64 lo_val;
+ struct {
+ u64 p : 1,
+ fpd : 1,
+ dm : 1,
+ rh : 1,
+ tm : 1,
+ dlm : 3,
+ avail : 4,
+ res_1 : 4,
+ vector : 8,
+ res_2 : 8,
+ dst : 32;
+ }lo;
+ };
+ union {
+ u64 hi_val;
+ struct {
+ u64 sid : 16,
+ sq : 2,
+ svt : 2,
+ res_1 : 44;
+ }hi;
+ };
+};
+#define IREMAP_ENTRY_NR (PAGE_SIZE_4K/sizeof(struct iremap_entry))
+#define iremap_present(v) ((v).lo & 1)
+#define iremap_fault_disable(v) (((v).lo >> 1) & 1)
+
+#define iremap_set_present(v) do {(v).lo |= 1;} while(0)
+#define iremap_clear_present(v) do {(v).lo &= ~1;} while(0)
+
+/* queue invalidation entry */
+struct qinval_entry {
+ union {
+ struct {
+ struct {
+ u64 type : 4,
+ granu : 2,
+ res_1 : 10,
+ did : 16,
+ sid : 16,
+ fm : 2,
+ res_2 : 14;
+ }lo;
+ struct {
+ u64 res;
+ }hi;
+ }cc_inv_dsc;
+ struct {
+ struct {
+ u64 type : 4,
+ granu : 2,
+ dw : 1,
+ dr : 1,
+ res_1 : 8,
+ did : 16,
+ res_2 : 32;
+ }lo;
+ struct {
+ u64 am : 6,
+ ih : 1,
+ res_1 : 5,
+ addr : 52;
+ }hi;
+ }iotlb_inv_dsc;
+ struct {
+ struct {
+ u64 type : 4,
+ res_1 : 12,
+ max_invs_pend: 5,
+ res_2 : 11,
+ sid : 16,
+ res_3 : 16;
+ }lo;
+ struct {
+ u64 size : 1,
+ res_1 : 11,
+ addr : 52;
+ }hi;
+ }dev_iotlb_inv_dsc;
+ struct {
+ struct {
+ u64 type : 4,
+ granu : 1,
+ res_1 : 22,
+ im : 5,
+ iidx : 16,
+ res_2 : 16;
+ }lo;
+ struct {
+ u64 res;
+ }hi;
+ }iec_inv_dsc;
+ struct {
+ struct {
+ u64 type : 4,
+ iflag : 1,
+ sw : 1,
+ fn : 1,
+ res_1 : 25,
+ sdata : 32;
+ }lo;
+ struct {
+ u64 res_1 : 2,
+ saddr : 62;
+ }hi;
+ }inv_wait_dsc;
+ }q;
+};
+
+struct poll_info {
+ u64 saddr;
+ u32 udata;
+};
+
+#define QINVAL_ENTRY_NR (PAGE_SIZE_4K/sizeof(struct qinval_entry))
+#define qinval_present(v) ((v).lo & 1)
+#define qinval_fault_disable(v) (((v).lo >> 1) & 1)
+
+#define qinval_set_present(v) do {(v).lo |= 1;} while(0)
+#define qinval_clear_present(v) do {(v).lo &= ~1;} while(0)
+
+#define RESERVED_VAL 0
+
+#define TYPE_INVAL_CONTEXT 0x1
+#define TYPE_INVAL_IOTLB 0x2
+#define TYPE_INVAL_DEVICE_IOTLB 0x3
+#define TYPE_INVAL_IEC 0x4
+#define TYPE_INVAL_WAIT 0x5
+
+#define NOTIFY_TYPE_POLL 1
+#define NOTIFY_TYPE_INTR 1
+#define INTERRUTP_FLAG 1
+#define STATUS_WRITE 1
+#define FENCE_FLAG 1
+
+#define IEC_GLOBAL_INVL 0
+#define IEC_INDEX_INVL 1
+#define IRTA_REG_EIME_SHIFT 11
+#define IRTA_REG_TABLE_SIZE 7 // 4k page = 256 * 16 byte entries
+ // 2^^(IRTA_REG_TABLE_SIZE + 1) = 256
+ // IRTA_REG_TABLE_SIZE = 7
+
+#define VTD_PAGE_TABLE_LEVEL_3 3
+#define VTD_PAGE_TABLE_LEVEL_4 4
+
+#define DEFAULT_DOMAIN_ADDRESS_WIDTH 48
+#define MAX_IOMMU_REGS 0xc0
+
+extern struct list_head acpi_drhd_units;
+extern struct list_head acpi_rmrr_units;
+extern struct list_head acpi_ioapic_units;
+
+struct qi_ctrl {
+ struct qinval_entry *qinval; /* queue invalidation page */
+ int qinval_index; /* queue invalidation index */
+ spinlock_t qinval_lock; /* lock for queue invalidation page */
+ spinlock_t qinval_poll_lock; /* lock for queue invalidation poll addr */
+ volatile u32 qinval_poll_status; /* used by poll methord to sync */
+};
+
+struct ir_ctrl {
+ struct iremap_entry *iremap; /* interrupt remap table */
+ int iremap_index; /* interrupt remap index */
+ spinlock_t iremap_lock; /* lock for irq remappping table */
+};
+
+struct iommu_flush {
+ int (*context)(void *iommu, u16 did, u16 source_id,
+ u8 function_mask, u64 type, int non_present_entry_flush);
+ int (*iotlb)(void *iommu, u16 did, u64 addr, unsigned int size_order,
+ u64 type, int non_present_entry_flush);
+};
+
+struct intel_iommu {
+ struct qi_ctrl qi_ctrl;
+ struct ir_ctrl ir_ctrl;
+ struct iommu_flush flush;
+};
+
+#endif
diff -r 8c921adf4833 -r 42f6c206c951 xen/drivers/passthrough/vtd/qinval.c
--- a/xen/drivers/passthrough/vtd/qinval.c Fri Mar 14 15:07:45 2008 -0600
+++ b/xen/drivers/passthrough/vtd/qinval.c Thu Mar 20 12:35:40 2008 -0600
@@ -19,15 +19,9 @@
*/
-#include <xen/init.h>
-#include <xen/irq.h>
-#include <xen/spinlock.h>
#include <xen/sched.h>
-#include <xen/xmalloc.h>
-#include <xen/domain_page.h>
-#include <asm/delay.h>
-#include <asm/string.h>
-#include <asm/iommu.h>
+#include <xen/iommu.h>
+#include "iommu.h"
#include "dmar.h"
#include "vtd.h"
#include "../pci-direct.h"
diff -r 8c921adf4833 -r 42f6c206c951 xen/drivers/passthrough/vtd/utils.c
_______________________________________________
Xen-changelog mailing list
Xen-changelog@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-changelog
|
Home