* [Qemu-devel] [PATCH] Xen PV Device
@ 2013-07-03 11:19 Paul Durrant
2013-07-03 16:37 ` Stefano Stabellini
2013-07-03 16:37 ` [Qemu-devel] " Stefano Stabellini
0 siblings, 2 replies; 10+ messages in thread
From: Paul Durrant @ 2013-07-03 11:19 UTC (permalink / raw)
To: qemu-devel, xen-devel; +Cc: Paul Durrant, Stefano Stabellini
This patch introduces a new Xen PV PCI device which will act as a new
binding point for PV drivers for Xen.
The device has parameterized vendor-id, device-id and revision to allow to
be configured as a binding point for any vendor's PV drivers.
Signed-off-by: Paul Durrant <paul.durrant@citrix.com>
Cc: Stefano Stabellini <stefano.stabellini@citrix.com>
---
hw/xen/Makefile.objs | 1 +
hw/xen/xen_pvdevice.c | 131 ++++++++++++++++++++++++++++++++++++++++++++++
include/hw/pci/pci_ids.h | 5 +-
trace-events | 4 ++
4 files changed, 139 insertions(+), 2 deletions(-)
create mode 100644 hw/xen/xen_pvdevice.c
diff --git a/hw/xen/Makefile.objs b/hw/xen/Makefile.objs
index 2017560..fd88003 100644
--- a/hw/xen/Makefile.objs
+++ b/hw/xen/Makefile.objs
@@ -4,3 +4,4 @@ common-obj-$(CONFIG_XEN_BACKEND) += xen_backend.o xen_devconfig.o
obj-$(CONFIG_XEN_I386) += xen_platform.o xen_apic.o
obj-$(CONFIG_XEN_PCI_PASSTHROUGH) += xen-host-pci-device.o
obj-$(CONFIG_XEN_PCI_PASSTHROUGH) += xen_pt.o xen_pt_config_init.o xen_pt_msi.o
+obj-$(CONFIG_XEN) += xen_pvdevice.o
diff --git a/hw/xen/xen_pvdevice.c b/hw/xen/xen_pvdevice.c
new file mode 100644
index 0000000..dbc4bf5
--- /dev/null
+++ b/hw/xen/xen_pvdevice.c
@@ -0,0 +1,131 @@
+/* Copyright (c) Citrix Systems Inc.
+ * All rights reserved.
+ *
+ * Redistribution and use in source and binary forms,
+ * with or without modification, are permitted provided
+ * that the following conditions are met:
+ *
+ * * Redistributions of source code must retain the above
+ * copyright notice, this list of conditions and the
+ * following disclaimer.
+ * * Redistributions in binary form must reproduce the above
+ * copyright notice, this list of conditions and the
+ * following disclaimer in the documentation and/or other
+ * materials provided with the distribution.
+ *
+ * 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
+ * MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
+ * DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR
+ * CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ * 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 DAMAGE.
+ */
+
+#include "hw/hw.h"
+#include "hw/pci/pci.h"
+#include "trace.h"
+
+#define TYPE_XEN_PV_DEVICE "xen-pvdevice"
+
+#define XEN_PV_DEVICE(obj) \
+ OBJECT_CHECK(XenPVDevice, (obj), TYPE_XEN_PV_DEVICE)
+
+typedef struct XenPVDevice {
+ /*< private >*/
+ PCIDevice parent_obj;
+ /*< public >*/
+ uint16_t vendor_id;
+ uint16_t device_id;
+ uint8_t revision;
+ uint32_t size;
+ MemoryRegion mmio;
+} XenPVDevice;
+
+static uint64_t xen_pv_mmio_read(void *opaque, hwaddr addr,
+ unsigned size)
+{
+ trace_xen_pv_mmio_read(addr);
+
+ return ~(uint64_t)0;
+}
+
+static void xen_pv_mmio_write(void *opaque, hwaddr addr,
+ uint64_t val, unsigned size)
+{
+ trace_xen_pv_mmio_write(addr);
+}
+
+static const MemoryRegionOps xen_pv_mmio_ops = {
+ .read = &xen_pv_mmio_read,
+ .write = &xen_pv_mmio_write,
+ .endianness = DEVICE_LITTLE_ENDIAN,
+};
+
+static int xen_pv_init(PCIDevice *pci_dev)
+{
+ XenPVDevice *d = XEN_PV_DEVICE(pci_dev);
+ uint8_t *pci_conf;
+
+ pci_conf = pci_dev->config;
+
+ pci_set_word(pci_conf + PCI_VENDOR_ID, d->vendor_id);
+ pci_set_word(pci_conf + PCI_SUBSYSTEM_VENDOR_ID, d->vendor_id);
+ pci_set_word(pci_conf + PCI_DEVICE_ID, d->device_id);
+ pci_set_word(pci_conf + PCI_SUBSYSTEM_ID, d->device_id);
+ pci_set_byte(pci_conf + PCI_REVISION_ID, d->revision);
+
+ pci_set_word(pci_conf + PCI_COMMAND, PCI_COMMAND_MEMORY);
+
+ pci_config_set_prog_interface(pci_conf, 0);
+
+ pci_conf[PCI_INTERRUPT_PIN] = 1;
+
+ memory_region_init_io(&d->mmio, &xen_pv_mmio_ops, d,
+ "xen-pvdevice-mmio", d->size);
+
+ pci_register_bar(pci_dev, 1, PCI_BASE_ADDRESS_MEM_PREFETCH,
+ &d->mmio);
+
+ return 0;
+}
+
+static Property xen_pv_props[] = {
+ DEFINE_PROP_UINT16("vendor-id", XenPVDevice, vendor_id, PCI_VENDOR_ID_XEN),
+ DEFINE_PROP_UINT16("device-id", XenPVDevice, device_id, PCI_DEVICE_ID_XEN_PVDEVICE),
+ DEFINE_PROP_UINT8("revision", XenPVDevice, revision, 0x01),
+ DEFINE_PROP_UINT32("size", XenPVDevice, size, 0x400000),
+ DEFINE_PROP_END_OF_LIST()
+};
+
+static void xen_pv_class_init(ObjectClass *klass, void *data)
+{
+ DeviceClass *dc = DEVICE_CLASS(klass);
+ PCIDeviceClass *k = PCI_DEVICE_CLASS(klass);
+
+ k->init = xen_pv_init;
+ k->class_id = PCI_CLASS_SYSTEM_OTHER;
+ dc->desc = "Xen PV Device";
+ dc->props = xen_pv_props;
+}
+
+static const TypeInfo xen_pv_type_info = {
+ .name = TYPE_XEN_PV_DEVICE,
+ .parent = TYPE_PCI_DEVICE,
+ .instance_size = sizeof(XenPVDevice),
+ .class_init = xen_pv_class_init,
+};
+
+static void xen_pv_register_types(void)
+{
+ type_register_static(&xen_pv_type_info);
+}
+
+type_init(xen_pv_register_types)
diff --git a/include/hw/pci/pci_ids.h b/include/hw/pci/pci_ids.h
index d8dc2f1..263bca3 100644
--- a/include/hw/pci/pci_ids.h
+++ b/include/hw/pci/pci_ids.h
@@ -142,8 +142,9 @@
#define PCI_DEVICE_ID_INTEL_Q35_MCH 0x29c0
-#define PCI_VENDOR_ID_XEN 0x5853
-#define PCI_DEVICE_ID_XEN_PLATFORM 0x0001
+#define PCI_VENDOR_ID_XEN 0x5853
+#define PCI_DEVICE_ID_XEN_PLATFORM 0x0001
+#define PCI_DEVICE_ID_XEN_PVDEVICE 0x0002
#define PCI_VENDOR_ID_NEC 0x1033
#define PCI_DEVICE_ID_NEC_UPD720200 0x0194
diff --git a/trace-events b/trace-events
index c5f1ccb..0445853 100644
--- a/trace-events
+++ b/trace-events
@@ -1161,3 +1161,7 @@ kvm_run_exit(int cpu_index, uint32_t reason) "cpu_index %d, reason %d"
# qom/object.c
object_dynamic_cast_assert(const char *type, const char *target, const char *file, int line, const char *func) "%s->%s (%s:%d:%s)"
object_class_dynamic_cast_assert(const char *type, const char *target, const char *file, int line, const char *func) "%s->%s (%s:%d:%s)"
+
+# hw/xen/xen_pvdevice.c
+xen_pv_mmio_read(uint64_t addr) "WARNING: read from Xen PV Device MMIO space (address %"PRIx64")"
+xen_pv_mmio_write(uint64_t addr) "WARNING: write to Xen PV Device MMIO space (address %"PRIx64")"
--
1.7.10.4
^ permalink raw reply related [flat|nested] 10+ messages in thread
* Re: [Qemu-devel] [PATCH] Xen PV Device
2013-07-03 11:19 [Qemu-devel] [PATCH] Xen PV Device Paul Durrant
2013-07-03 16:37 ` Stefano Stabellini
@ 2013-07-03 16:37 ` Stefano Stabellini
2013-07-04 7:15 ` Andreas Färber
` (3 more replies)
1 sibling, 4 replies; 10+ messages in thread
From: Stefano Stabellini @ 2013-07-03 16:37 UTC (permalink / raw)
To: Paul Durrant; +Cc: Stefano Stabellini, qemu-devel, afaerber, xen-devel
On Wed, 3 Jul 2013, Paul Durrant wrote:
> This patch introduces a new Xen PV PCI device which will act as a new
> binding point for PV drivers for Xen.
> The device has parameterized vendor-id, device-id and revision to allow to
> be configured as a binding point for any vendor's PV drivers.
>
> Signed-off-by: Paul Durrant <paul.durrant@citrix.com>
> Cc: Stefano Stabellini <stefano.stabellini@citrix.com>
> ---
> hw/xen/Makefile.objs | 1 +
> hw/xen/xen_pvdevice.c | 131 ++++++++++++++++++++++++++++++++++++++++++++++
> include/hw/pci/pci_ids.h | 5 +-
> trace-events | 4 ++
> 4 files changed, 139 insertions(+), 2 deletions(-)
> create mode 100644 hw/xen/xen_pvdevice.c
>
> diff --git a/hw/xen/Makefile.objs b/hw/xen/Makefile.objs
> index 2017560..fd88003 100644
> --- a/hw/xen/Makefile.objs
> +++ b/hw/xen/Makefile.objs
> @@ -4,3 +4,4 @@ common-obj-$(CONFIG_XEN_BACKEND) += xen_backend.o xen_devconfig.o
> obj-$(CONFIG_XEN_I386) += xen_platform.o xen_apic.o
> obj-$(CONFIG_XEN_PCI_PASSTHROUGH) += xen-host-pci-device.o
> obj-$(CONFIG_XEN_PCI_PASSTHROUGH) += xen_pt.o xen_pt_config_init.o xen_pt_msi.o
> +obj-$(CONFIG_XEN) += xen_pvdevice.o
> diff --git a/hw/xen/xen_pvdevice.c b/hw/xen/xen_pvdevice.c
> new file mode 100644
> index 0000000..dbc4bf5
> --- /dev/null
> +++ b/hw/xen/xen_pvdevice.c
> @@ -0,0 +1,131 @@
> +/* Copyright (c) Citrix Systems Inc.
> + * All rights reserved.
Like Anthony wrote before, All rights reserved contradicts what's
written below.
Aside from this, it looks OK to me.
I would like to see the libxl side patch.
Also it would be nice to have an ack from Andreas or another QOM expert.
> + * Redistribution and use in source and binary forms,
> + * with or without modification, are permitted provided
> + * that the following conditions are met:
> + *
> + * * Redistributions of source code must retain the above
> + * copyright notice, this list of conditions and the
> + * following disclaimer.
> + * * Redistributions in binary form must reproduce the above
> + * copyright notice, this list of conditions and the
> + * following disclaimer in the documentation and/or other
> + * materials provided with the distribution.
> + *
> + * 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
> + * MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
> + * DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR
> + * CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
> + * 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 DAMAGE.
> + */
> +
> +#include "hw/hw.h"
> +#include "hw/pci/pci.h"
> +#include "trace.h"
> +
> +#define TYPE_XEN_PV_DEVICE "xen-pvdevice"
> +
> +#define XEN_PV_DEVICE(obj) \
> + OBJECT_CHECK(XenPVDevice, (obj), TYPE_XEN_PV_DEVICE)
> +
> +typedef struct XenPVDevice {
> + /*< private >*/
> + PCIDevice parent_obj;
> + /*< public >*/
> + uint16_t vendor_id;
> + uint16_t device_id;
> + uint8_t revision;
> + uint32_t size;
> + MemoryRegion mmio;
> +} XenPVDevice;
> +
> +static uint64_t xen_pv_mmio_read(void *opaque, hwaddr addr,
> + unsigned size)
> +{
> + trace_xen_pv_mmio_read(addr);
> +
> + return ~(uint64_t)0;
> +}
> +
> +static void xen_pv_mmio_write(void *opaque, hwaddr addr,
> + uint64_t val, unsigned size)
> +{
> + trace_xen_pv_mmio_write(addr);
> +}
> +
> +static const MemoryRegionOps xen_pv_mmio_ops = {
> + .read = &xen_pv_mmio_read,
> + .write = &xen_pv_mmio_write,
> + .endianness = DEVICE_LITTLE_ENDIAN,
> +};
> +
> +static int xen_pv_init(PCIDevice *pci_dev)
> +{
> + XenPVDevice *d = XEN_PV_DEVICE(pci_dev);
> + uint8_t *pci_conf;
> +
> + pci_conf = pci_dev->config;
> +
> + pci_set_word(pci_conf + PCI_VENDOR_ID, d->vendor_id);
> + pci_set_word(pci_conf + PCI_SUBSYSTEM_VENDOR_ID, d->vendor_id);
> + pci_set_word(pci_conf + PCI_DEVICE_ID, d->device_id);
> + pci_set_word(pci_conf + PCI_SUBSYSTEM_ID, d->device_id);
> + pci_set_byte(pci_conf + PCI_REVISION_ID, d->revision);
> +
> + pci_set_word(pci_conf + PCI_COMMAND, PCI_COMMAND_MEMORY);
> +
> + pci_config_set_prog_interface(pci_conf, 0);
> +
> + pci_conf[PCI_INTERRUPT_PIN] = 1;
> +
> + memory_region_init_io(&d->mmio, &xen_pv_mmio_ops, d,
> + "xen-pvdevice-mmio", d->size);
> +
> + pci_register_bar(pci_dev, 1, PCI_BASE_ADDRESS_MEM_PREFETCH,
> + &d->mmio);
> +
> + return 0;
> +}
> +
> +static Property xen_pv_props[] = {
> + DEFINE_PROP_UINT16("vendor-id", XenPVDevice, vendor_id, PCI_VENDOR_ID_XEN),
> + DEFINE_PROP_UINT16("device-id", XenPVDevice, device_id, PCI_DEVICE_ID_XEN_PVDEVICE),
> + DEFINE_PROP_UINT8("revision", XenPVDevice, revision, 0x01),
> + DEFINE_PROP_UINT32("size", XenPVDevice, size, 0x400000),
> + DEFINE_PROP_END_OF_LIST()
> +};
> +
> +static void xen_pv_class_init(ObjectClass *klass, void *data)
> +{
> + DeviceClass *dc = DEVICE_CLASS(klass);
> + PCIDeviceClass *k = PCI_DEVICE_CLASS(klass);
> +
> + k->init = xen_pv_init;
> + k->class_id = PCI_CLASS_SYSTEM_OTHER;
> + dc->desc = "Xen PV Device";
> + dc->props = xen_pv_props;
> +}
> +
> +static const TypeInfo xen_pv_type_info = {
> + .name = TYPE_XEN_PV_DEVICE,
> + .parent = TYPE_PCI_DEVICE,
> + .instance_size = sizeof(XenPVDevice),
> + .class_init = xen_pv_class_init,
> +};
> +
> +static void xen_pv_register_types(void)
> +{
> + type_register_static(&xen_pv_type_info);
> +}
> +
> +type_init(xen_pv_register_types)
> diff --git a/include/hw/pci/pci_ids.h b/include/hw/pci/pci_ids.h
> index d8dc2f1..263bca3 100644
> --- a/include/hw/pci/pci_ids.h
> +++ b/include/hw/pci/pci_ids.h
> @@ -142,8 +142,9 @@
>
> #define PCI_DEVICE_ID_INTEL_Q35_MCH 0x29c0
>
> -#define PCI_VENDOR_ID_XEN 0x5853
> -#define PCI_DEVICE_ID_XEN_PLATFORM 0x0001
> +#define PCI_VENDOR_ID_XEN 0x5853
> +#define PCI_DEVICE_ID_XEN_PLATFORM 0x0001
> +#define PCI_DEVICE_ID_XEN_PVDEVICE 0x0002
>
> #define PCI_VENDOR_ID_NEC 0x1033
> #define PCI_DEVICE_ID_NEC_UPD720200 0x0194
> diff --git a/trace-events b/trace-events
> index c5f1ccb..0445853 100644
> --- a/trace-events
> +++ b/trace-events
> @@ -1161,3 +1161,7 @@ kvm_run_exit(int cpu_index, uint32_t reason) "cpu_index %d, reason %d"
> # qom/object.c
> object_dynamic_cast_assert(const char *type, const char *target, const char *file, int line, const char *func) "%s->%s (%s:%d:%s)"
> object_class_dynamic_cast_assert(const char *type, const char *target, const char *file, int line, const char *func) "%s->%s (%s:%d:%s)"
> +
> +# hw/xen/xen_pvdevice.c
> +xen_pv_mmio_read(uint64_t addr) "WARNING: read from Xen PV Device MMIO space (address %"PRIx64")"
> +xen_pv_mmio_write(uint64_t addr) "WARNING: write to Xen PV Device MMIO space (address %"PRIx64")"
> --
> 1.7.10.4
>
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [PATCH] Xen PV Device
2013-07-03 11:19 [Qemu-devel] [PATCH] Xen PV Device Paul Durrant
@ 2013-07-03 16:37 ` Stefano Stabellini
2013-07-03 16:37 ` [Qemu-devel] " Stefano Stabellini
1 sibling, 0 replies; 10+ messages in thread
From: Stefano Stabellini @ 2013-07-03 16:37 UTC (permalink / raw)
To: Paul Durrant; +Cc: Stefano Stabellini, qemu-devel, afaerber, xen-devel
On Wed, 3 Jul 2013, Paul Durrant wrote:
> This patch introduces a new Xen PV PCI device which will act as a new
> binding point for PV drivers for Xen.
> The device has parameterized vendor-id, device-id and revision to allow to
> be configured as a binding point for any vendor's PV drivers.
>
> Signed-off-by: Paul Durrant <paul.durrant@citrix.com>
> Cc: Stefano Stabellini <stefano.stabellini@citrix.com>
> ---
> hw/xen/Makefile.objs | 1 +
> hw/xen/xen_pvdevice.c | 131 ++++++++++++++++++++++++++++++++++++++++++++++
> include/hw/pci/pci_ids.h | 5 +-
> trace-events | 4 ++
> 4 files changed, 139 insertions(+), 2 deletions(-)
> create mode 100644 hw/xen/xen_pvdevice.c
>
> diff --git a/hw/xen/Makefile.objs b/hw/xen/Makefile.objs
> index 2017560..fd88003 100644
> --- a/hw/xen/Makefile.objs
> +++ b/hw/xen/Makefile.objs
> @@ -4,3 +4,4 @@ common-obj-$(CONFIG_XEN_BACKEND) += xen_backend.o xen_devconfig.o
> obj-$(CONFIG_XEN_I386) += xen_platform.o xen_apic.o
> obj-$(CONFIG_XEN_PCI_PASSTHROUGH) += xen-host-pci-device.o
> obj-$(CONFIG_XEN_PCI_PASSTHROUGH) += xen_pt.o xen_pt_config_init.o xen_pt_msi.o
> +obj-$(CONFIG_XEN) += xen_pvdevice.o
> diff --git a/hw/xen/xen_pvdevice.c b/hw/xen/xen_pvdevice.c
> new file mode 100644
> index 0000000..dbc4bf5
> --- /dev/null
> +++ b/hw/xen/xen_pvdevice.c
> @@ -0,0 +1,131 @@
> +/* Copyright (c) Citrix Systems Inc.
> + * All rights reserved.
Like Anthony wrote before, All rights reserved contradicts what's
written below.
Aside from this, it looks OK to me.
I would like to see the libxl side patch.
Also it would be nice to have an ack from Andreas or another QOM expert.
> + * Redistribution and use in source and binary forms,
> + * with or without modification, are permitted provided
> + * that the following conditions are met:
> + *
> + * * Redistributions of source code must retain the above
> + * copyright notice, this list of conditions and the
> + * following disclaimer.
> + * * Redistributions in binary form must reproduce the above
> + * copyright notice, this list of conditions and the
> + * following disclaimer in the documentation and/or other
> + * materials provided with the distribution.
> + *
> + * 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
> + * MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
> + * DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR
> + * CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
> + * 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 DAMAGE.
> + */
> +
> +#include "hw/hw.h"
> +#include "hw/pci/pci.h"
> +#include "trace.h"
> +
> +#define TYPE_XEN_PV_DEVICE "xen-pvdevice"
> +
> +#define XEN_PV_DEVICE(obj) \
> + OBJECT_CHECK(XenPVDevice, (obj), TYPE_XEN_PV_DEVICE)
> +
> +typedef struct XenPVDevice {
> + /*< private >*/
> + PCIDevice parent_obj;
> + /*< public >*/
> + uint16_t vendor_id;
> + uint16_t device_id;
> + uint8_t revision;
> + uint32_t size;
> + MemoryRegion mmio;
> +} XenPVDevice;
> +
> +static uint64_t xen_pv_mmio_read(void *opaque, hwaddr addr,
> + unsigned size)
> +{
> + trace_xen_pv_mmio_read(addr);
> +
> + return ~(uint64_t)0;
> +}
> +
> +static void xen_pv_mmio_write(void *opaque, hwaddr addr,
> + uint64_t val, unsigned size)
> +{
> + trace_xen_pv_mmio_write(addr);
> +}
> +
> +static const MemoryRegionOps xen_pv_mmio_ops = {
> + .read = &xen_pv_mmio_read,
> + .write = &xen_pv_mmio_write,
> + .endianness = DEVICE_LITTLE_ENDIAN,
> +};
> +
> +static int xen_pv_init(PCIDevice *pci_dev)
> +{
> + XenPVDevice *d = XEN_PV_DEVICE(pci_dev);
> + uint8_t *pci_conf;
> +
> + pci_conf = pci_dev->config;
> +
> + pci_set_word(pci_conf + PCI_VENDOR_ID, d->vendor_id);
> + pci_set_word(pci_conf + PCI_SUBSYSTEM_VENDOR_ID, d->vendor_id);
> + pci_set_word(pci_conf + PCI_DEVICE_ID, d->device_id);
> + pci_set_word(pci_conf + PCI_SUBSYSTEM_ID, d->device_id);
> + pci_set_byte(pci_conf + PCI_REVISION_ID, d->revision);
> +
> + pci_set_word(pci_conf + PCI_COMMAND, PCI_COMMAND_MEMORY);
> +
> + pci_config_set_prog_interface(pci_conf, 0);
> +
> + pci_conf[PCI_INTERRUPT_PIN] = 1;
> +
> + memory_region_init_io(&d->mmio, &xen_pv_mmio_ops, d,
> + "xen-pvdevice-mmio", d->size);
> +
> + pci_register_bar(pci_dev, 1, PCI_BASE_ADDRESS_MEM_PREFETCH,
> + &d->mmio);
> +
> + return 0;
> +}
> +
> +static Property xen_pv_props[] = {
> + DEFINE_PROP_UINT16("vendor-id", XenPVDevice, vendor_id, PCI_VENDOR_ID_XEN),
> + DEFINE_PROP_UINT16("device-id", XenPVDevice, device_id, PCI_DEVICE_ID_XEN_PVDEVICE),
> + DEFINE_PROP_UINT8("revision", XenPVDevice, revision, 0x01),
> + DEFINE_PROP_UINT32("size", XenPVDevice, size, 0x400000),
> + DEFINE_PROP_END_OF_LIST()
> +};
> +
> +static void xen_pv_class_init(ObjectClass *klass, void *data)
> +{
> + DeviceClass *dc = DEVICE_CLASS(klass);
> + PCIDeviceClass *k = PCI_DEVICE_CLASS(klass);
> +
> + k->init = xen_pv_init;
> + k->class_id = PCI_CLASS_SYSTEM_OTHER;
> + dc->desc = "Xen PV Device";
> + dc->props = xen_pv_props;
> +}
> +
> +static const TypeInfo xen_pv_type_info = {
> + .name = TYPE_XEN_PV_DEVICE,
> + .parent = TYPE_PCI_DEVICE,
> + .instance_size = sizeof(XenPVDevice),
> + .class_init = xen_pv_class_init,
> +};
> +
> +static void xen_pv_register_types(void)
> +{
> + type_register_static(&xen_pv_type_info);
> +}
> +
> +type_init(xen_pv_register_types)
> diff --git a/include/hw/pci/pci_ids.h b/include/hw/pci/pci_ids.h
> index d8dc2f1..263bca3 100644
> --- a/include/hw/pci/pci_ids.h
> +++ b/include/hw/pci/pci_ids.h
> @@ -142,8 +142,9 @@
>
> #define PCI_DEVICE_ID_INTEL_Q35_MCH 0x29c0
>
> -#define PCI_VENDOR_ID_XEN 0x5853
> -#define PCI_DEVICE_ID_XEN_PLATFORM 0x0001
> +#define PCI_VENDOR_ID_XEN 0x5853
> +#define PCI_DEVICE_ID_XEN_PLATFORM 0x0001
> +#define PCI_DEVICE_ID_XEN_PVDEVICE 0x0002
>
> #define PCI_VENDOR_ID_NEC 0x1033
> #define PCI_DEVICE_ID_NEC_UPD720200 0x0194
> diff --git a/trace-events b/trace-events
> index c5f1ccb..0445853 100644
> --- a/trace-events
> +++ b/trace-events
> @@ -1161,3 +1161,7 @@ kvm_run_exit(int cpu_index, uint32_t reason) "cpu_index %d, reason %d"
> # qom/object.c
> object_dynamic_cast_assert(const char *type, const char *target, const char *file, int line, const char *func) "%s->%s (%s:%d:%s)"
> object_class_dynamic_cast_assert(const char *type, const char *target, const char *file, int line, const char *func) "%s->%s (%s:%d:%s)"
> +
> +# hw/xen/xen_pvdevice.c
> +xen_pv_mmio_read(uint64_t addr) "WARNING: read from Xen PV Device MMIO space (address %"PRIx64")"
> +xen_pv_mmio_write(uint64_t addr) "WARNING: write to Xen PV Device MMIO space (address %"PRIx64")"
> --
> 1.7.10.4
>
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [Qemu-devel] [PATCH] Xen PV Device
2013-07-03 16:37 ` [Qemu-devel] " Stefano Stabellini
2013-07-04 7:15 ` Andreas Färber
@ 2013-07-04 7:15 ` Andreas Färber
2013-07-04 8:14 ` Paul Durrant
2013-07-04 8:17 ` Paul Durrant
2013-07-04 8:17 ` [Qemu-devel] " Paul Durrant
3 siblings, 1 reply; 10+ messages in thread
From: Andreas Färber @ 2013-07-04 7:15 UTC (permalink / raw)
To: Stefano Stabellini
Cc: Paul Durrant, Stefano Stabellini, qemu-devel, xen-devel
Am 03.07.2013 18:37, schrieb Stefano Stabellini:
> On Wed, 3 Jul 2013, Paul Durrant wrote:
>> This patch introduces a new Xen PV PCI device which will act as a new
>> binding point for PV drivers for Xen.
>> The device has parameterized vendor-id, device-id and revision to allow to
>> be configured as a binding point for any vendor's PV drivers.
>>
>> Signed-off-by: Paul Durrant <paul.durrant@citrix.com>
>> Cc: Stefano Stabellini <stefano.stabellini@citrix.com>
>> ---
>> hw/xen/Makefile.objs | 1 +
>> hw/xen/xen_pvdevice.c | 131 ++++++++++++++++++++++++++++++++++++++++++++++
>> include/hw/pci/pci_ids.h | 5 +-
>> trace-events | 4 ++
>> 4 files changed, 139 insertions(+), 2 deletions(-)
>> create mode 100644 hw/xen/xen_pvdevice.c
>>
>> diff --git a/hw/xen/Makefile.objs b/hw/xen/Makefile.objs
>> index 2017560..fd88003 100644
>> --- a/hw/xen/Makefile.objs
>> +++ b/hw/xen/Makefile.objs
>> @@ -4,3 +4,4 @@ common-obj-$(CONFIG_XEN_BACKEND) += xen_backend.o xen_devconfig.o
>> obj-$(CONFIG_XEN_I386) += xen_platform.o xen_apic.o
>> obj-$(CONFIG_XEN_PCI_PASSTHROUGH) += xen-host-pci-device.o
>> obj-$(CONFIG_XEN_PCI_PASSTHROUGH) += xen_pt.o xen_pt_config_init.o xen_pt_msi.o
>> +obj-$(CONFIG_XEN) += xen_pvdevice.o
>> diff --git a/hw/xen/xen_pvdevice.c b/hw/xen/xen_pvdevice.c
>> new file mode 100644
>> index 0000000..dbc4bf5
>> --- /dev/null
>> +++ b/hw/xen/xen_pvdevice.c
>> @@ -0,0 +1,131 @@
>> +/* Copyright (c) Citrix Systems Inc.
>> + * All rights reserved.
>
> Like Anthony wrote before, All rights reserved contradicts what's
> written below.
> Aside from this, it looks OK to me.
>
> I would like to see the libxl side patch.
> Also it would be nice to have an ack from Andreas or another QOM expert.
>From a QOM view it looks fine now. :) Thanks for inquiring.
Some other comments though:
* Now that it no longer depends on TARGET_PAGE_SIZE, is it possible to
use common-obj-$(CONFIG_XEN)? Then it would build only once rather than
separately for i386 and x86_64 and any future Xen platforms (e.g., arm).
* It looks as if the MMIO functions were renamed - the arguments no
longer align. That could be edited before you apply the patch to your
queue if there's nothing else - then feel free to add my Reviewed-by
independent of the other issue.
* Paolo had asked for new MemoryRegions not to include the device name -
can be renamed once they get the owner field though (not merged yet).
Don't have a better suggestion handy.
Also Paul, by my count this is [PATCH v4] - please use
--subject-prefix="PATCH v5" if you respin and include the change log
either below "---" or in a cover letter. We prefer to see it for patch
review but not in Git commit history.
Similarly, "Introduce a new Xen PV device..." would elegantly avoid
reading "This patch..." after it's been committed. ;)
Regards,
Andreas
--
SUSE LINUX Products GmbH, Maxfeldstr. 5, 90409 Nürnberg, Germany
GF: Jeff Hawn, Jennifer Guild, Felix Imendörffer; HRB 16746 AG Nürnberg
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [PATCH] Xen PV Device
2013-07-03 16:37 ` [Qemu-devel] " Stefano Stabellini
@ 2013-07-04 7:15 ` Andreas Färber
2013-07-04 7:15 ` [Qemu-devel] " Andreas Färber
` (2 subsequent siblings)
3 siblings, 0 replies; 10+ messages in thread
From: Andreas Färber @ 2013-07-04 7:15 UTC (permalink / raw)
To: Stefano Stabellini
Cc: Paul Durrant, Stefano Stabellini, qemu-devel, xen-devel
Am 03.07.2013 18:37, schrieb Stefano Stabellini:
> On Wed, 3 Jul 2013, Paul Durrant wrote:
>> This patch introduces a new Xen PV PCI device which will act as a new
>> binding point for PV drivers for Xen.
>> The device has parameterized vendor-id, device-id and revision to allow to
>> be configured as a binding point for any vendor's PV drivers.
>>
>> Signed-off-by: Paul Durrant <paul.durrant@citrix.com>
>> Cc: Stefano Stabellini <stefano.stabellini@citrix.com>
>> ---
>> hw/xen/Makefile.objs | 1 +
>> hw/xen/xen_pvdevice.c | 131 ++++++++++++++++++++++++++++++++++++++++++++++
>> include/hw/pci/pci_ids.h | 5 +-
>> trace-events | 4 ++
>> 4 files changed, 139 insertions(+), 2 deletions(-)
>> create mode 100644 hw/xen/xen_pvdevice.c
>>
>> diff --git a/hw/xen/Makefile.objs b/hw/xen/Makefile.objs
>> index 2017560..fd88003 100644
>> --- a/hw/xen/Makefile.objs
>> +++ b/hw/xen/Makefile.objs
>> @@ -4,3 +4,4 @@ common-obj-$(CONFIG_XEN_BACKEND) += xen_backend.o xen_devconfig.o
>> obj-$(CONFIG_XEN_I386) += xen_platform.o xen_apic.o
>> obj-$(CONFIG_XEN_PCI_PASSTHROUGH) += xen-host-pci-device.o
>> obj-$(CONFIG_XEN_PCI_PASSTHROUGH) += xen_pt.o xen_pt_config_init.o xen_pt_msi.o
>> +obj-$(CONFIG_XEN) += xen_pvdevice.o
>> diff --git a/hw/xen/xen_pvdevice.c b/hw/xen/xen_pvdevice.c
>> new file mode 100644
>> index 0000000..dbc4bf5
>> --- /dev/null
>> +++ b/hw/xen/xen_pvdevice.c
>> @@ -0,0 +1,131 @@
>> +/* Copyright (c) Citrix Systems Inc.
>> + * All rights reserved.
>
> Like Anthony wrote before, All rights reserved contradicts what's
> written below.
> Aside from this, it looks OK to me.
>
> I would like to see the libxl side patch.
> Also it would be nice to have an ack from Andreas or another QOM expert.
>From a QOM view it looks fine now. :) Thanks for inquiring.
Some other comments though:
* Now that it no longer depends on TARGET_PAGE_SIZE, is it possible to
use common-obj-$(CONFIG_XEN)? Then it would build only once rather than
separately for i386 and x86_64 and any future Xen platforms (e.g., arm).
* It looks as if the MMIO functions were renamed - the arguments no
longer align. That could be edited before you apply the patch to your
queue if there's nothing else - then feel free to add my Reviewed-by
independent of the other issue.
* Paolo had asked for new MemoryRegions not to include the device name -
can be renamed once they get the owner field though (not merged yet).
Don't have a better suggestion handy.
Also Paul, by my count this is [PATCH v4] - please use
--subject-prefix="PATCH v5" if you respin and include the change log
either below "---" or in a cover letter. We prefer to see it for patch
review but not in Git commit history.
Similarly, "Introduce a new Xen PV device..." would elegantly avoid
reading "This patch..." after it's been committed. ;)
Regards,
Andreas
--
SUSE LINUX Products GmbH, Maxfeldstr. 5, 90409 Nürnberg, Germany
GF: Jeff Hawn, Jennifer Guild, Felix Imendörffer; HRB 16746 AG Nürnberg
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [Qemu-devel] [PATCH] Xen PV Device
2013-07-04 7:15 ` [Qemu-devel] " Andreas Färber
@ 2013-07-04 8:14 ` Paul Durrant
0 siblings, 0 replies; 10+ messages in thread
From: Paul Durrant @ 2013-07-04 8:14 UTC (permalink / raw)
To: Andreas Färber, Stefano Stabellini; +Cc: qemu-devel, xen-devel
> -----Original Message-----
> >
> > Like Anthony wrote before, All rights reserved contradicts what's
> > written below.
Like I said, it's part of all BSD licenses that I can find. It's certainly in the template on the OSI website and the FreeBSD license for instance.
> > Aside from this, it looks OK to me.
> >
> > I would like to see the libxl side patch.
> > Also it would be nice to have an ack from Andreas or another QOM expert.
>
> From a QOM view it looks fine now. :) Thanks for inquiring.
>
> Some other comments though:
> * Now that it no longer depends on TARGET_PAGE_SIZE, is it possible to
> use common-obj-$(CONFIG_XEN)? Then it would build only once rather than
> separately for i386 and x86_64 and any future Xen platforms (e.g., arm).
Sure, that sounds sensible.
> * It looks as if the MMIO functions were renamed - the arguments no
> longer align. That could be edited before you apply the patch to your
> queue if there's nothing else - then feel free to add my Reviewed-by
> independent of the other issue.
Thanks.
> * Paolo had asked for new MemoryRegions not to include the device name -
> can be renamed once they get the owner field though (not merged yet).
> Don't have a better suggestion handy.
>
I guess this can be fixed up later.
> Also Paul, by my count this is [PATCH v4] - please use
> --subject-prefix="PATCH v5" if you respin and include the change log
> either below "---" or in a cover letter. We prefer to see it for patch
> review but not in Git commit history.
Ok. I was unsure what to do since this device was under a different name so I opted to reset the version back to 1. I'll call the next one v5 as you suggest. I'm still finding my way with git so thanks for the tips.
> Similarly, "Introduce a new Xen PV device..." would elegantly avoid
> reading "This patch..." after it's been committed. ;)
>
Sure. Good point.
Paul
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [PATCH] Xen PV Device
@ 2013-07-04 8:14 ` Paul Durrant
0 siblings, 0 replies; 10+ messages in thread
From: Paul Durrant @ 2013-07-04 8:14 UTC (permalink / raw)
To: Andreas Färber; +Cc: Stefano Stabellini, qemu-devel, xen-devel
> -----Original Message-----
> >
> > Like Anthony wrote before, All rights reserved contradicts what's
> > written below.
Like I said, it's part of all BSD licenses that I can find. It's certainly in the template on the OSI website and the FreeBSD license for instance.
> > Aside from this, it looks OK to me.
> >
> > I would like to see the libxl side patch.
> > Also it would be nice to have an ack from Andreas or another QOM expert.
>
> From a QOM view it looks fine now. :) Thanks for inquiring.
>
> Some other comments though:
> * Now that it no longer depends on TARGET_PAGE_SIZE, is it possible to
> use common-obj-$(CONFIG_XEN)? Then it would build only once rather than
> separately for i386 and x86_64 and any future Xen platforms (e.g., arm).
Sure, that sounds sensible.
> * It looks as if the MMIO functions were renamed - the arguments no
> longer align. That could be edited before you apply the patch to your
> queue if there's nothing else - then feel free to add my Reviewed-by
> independent of the other issue.
Thanks.
> * Paolo had asked for new MemoryRegions not to include the device name -
> can be renamed once they get the owner field though (not merged yet).
> Don't have a better suggestion handy.
>
I guess this can be fixed up later.
> Also Paul, by my count this is [PATCH v4] - please use
> --subject-prefix="PATCH v5" if you respin and include the change log
> either below "---" or in a cover letter. We prefer to see it for patch
> review but not in Git commit history.
Ok. I was unsure what to do since this device was under a different name so I opted to reset the version back to 1. I'll call the next one v5 as you suggest. I'm still finding my way with git so thanks for the tips.
> Similarly, "Introduce a new Xen PV device..." would elegantly avoid
> reading "This patch..." after it's been committed. ;)
>
Sure. Good point.
Paul
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [Qemu-devel] [PATCH] Xen PV Device
2013-07-03 16:37 ` [Qemu-devel] " Stefano Stabellini
` (2 preceding siblings ...)
2013-07-04 8:17 ` Paul Durrant
@ 2013-07-04 8:17 ` Paul Durrant
3 siblings, 0 replies; 10+ messages in thread
From: Paul Durrant @ 2013-07-04 8:17 UTC (permalink / raw)
To: Stefano Stabellini; +Cc: qemu-devel, afaerber, xen-devel
> -----Original Message-----
> From: Stefano Stabellini [mailto:stefano.stabellini@eu.citrix.com]
> Sent: 03 July 2013 17:38
> To: Paul Durrant
> Cc: qemu-devel@nongnu.org; xen-devel@lists.xen.org; Stefano Stabellini;
> afaerber@suse.de
> Subject: Re: [PATCH] Xen PV Device
>
> On Wed, 3 Jul 2013, Paul Durrant wrote:
> > This patch introduces a new Xen PV PCI device which will act as a new
> > binding point for PV drivers for Xen.
> > The device has parameterized vendor-id, device-id and revision to allow to
> > be configured as a binding point for any vendor's PV drivers.
> >
> > Signed-off-by: Paul Durrant <paul.durrant@citrix.com>
> > Cc: Stefano Stabellini <stefano.stabellini@citrix.com>
> > ---
> > hw/xen/Makefile.objs | 1 +
> > hw/xen/xen_pvdevice.c | 131
> ++++++++++++++++++++++++++++++++++++++++++++++
> > include/hw/pci/pci_ids.h | 5 +-
> > trace-events | 4 ++
> > 4 files changed, 139 insertions(+), 2 deletions(-)
> > create mode 100644 hw/xen/xen_pvdevice.c
> >
> > diff --git a/hw/xen/Makefile.objs b/hw/xen/Makefile.objs
> > index 2017560..fd88003 100644
> > --- a/hw/xen/Makefile.objs
> > +++ b/hw/xen/Makefile.objs
> > @@ -4,3 +4,4 @@ common-obj-$(CONFIG_XEN_BACKEND) +=
> xen_backend.o xen_devconfig.o
> > obj-$(CONFIG_XEN_I386) += xen_platform.o xen_apic.o
> > obj-$(CONFIG_XEN_PCI_PASSTHROUGH) += xen-host-pci-device.o
> > obj-$(CONFIG_XEN_PCI_PASSTHROUGH) += xen_pt.o
> xen_pt_config_init.o xen_pt_msi.o
> > +obj-$(CONFIG_XEN) += xen_pvdevice.o
> > diff --git a/hw/xen/xen_pvdevice.c b/hw/xen/xen_pvdevice.c
> > new file mode 100644
> > index 0000000..dbc4bf5
> > --- /dev/null
> > +++ b/hw/xen/xen_pvdevice.c
> > @@ -0,0 +1,131 @@
> > +/* Copyright (c) Citrix Systems Inc.
> > + * All rights reserved.
>
> Like Anthony wrote before, All rights reserved contradicts what's
> written below.
> Aside from this, it looks OK to me.
>
> I would like to see the libxl side patch.
Working on it, but it's not required to use the new device so I don't think the QEMU patch need be predicated on it.
Paul
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [PATCH] Xen PV Device
2013-07-03 16:37 ` [Qemu-devel] " Stefano Stabellini
2013-07-04 7:15 ` Andreas Färber
2013-07-04 7:15 ` [Qemu-devel] " Andreas Färber
@ 2013-07-04 8:17 ` Paul Durrant
2013-07-04 8:17 ` [Qemu-devel] " Paul Durrant
3 siblings, 0 replies; 10+ messages in thread
From: Paul Durrant @ 2013-07-04 8:17 UTC (permalink / raw)
Cc: Stefano Stabellini, qemu-devel, afaerber, xen-devel
> -----Original Message-----
> From: Stefano Stabellini [mailto:stefano.stabellini@eu.citrix.com]
> Sent: 03 July 2013 17:38
> To: Paul Durrant
> Cc: qemu-devel@nongnu.org; xen-devel@lists.xen.org; Stefano Stabellini;
> afaerber@suse.de
> Subject: Re: [PATCH] Xen PV Device
>
> On Wed, 3 Jul 2013, Paul Durrant wrote:
> > This patch introduces a new Xen PV PCI device which will act as a new
> > binding point for PV drivers for Xen.
> > The device has parameterized vendor-id, device-id and revision to allow to
> > be configured as a binding point for any vendor's PV drivers.
> >
> > Signed-off-by: Paul Durrant <paul.durrant@citrix.com>
> > Cc: Stefano Stabellini <stefano.stabellini@citrix.com>
> > ---
> > hw/xen/Makefile.objs | 1 +
> > hw/xen/xen_pvdevice.c | 131
> ++++++++++++++++++++++++++++++++++++++++++++++
> > include/hw/pci/pci_ids.h | 5 +-
> > trace-events | 4 ++
> > 4 files changed, 139 insertions(+), 2 deletions(-)
> > create mode 100644 hw/xen/xen_pvdevice.c
> >
> > diff --git a/hw/xen/Makefile.objs b/hw/xen/Makefile.objs
> > index 2017560..fd88003 100644
> > --- a/hw/xen/Makefile.objs
> > +++ b/hw/xen/Makefile.objs
> > @@ -4,3 +4,4 @@ common-obj-$(CONFIG_XEN_BACKEND) +=
> xen_backend.o xen_devconfig.o
> > obj-$(CONFIG_XEN_I386) += xen_platform.o xen_apic.o
> > obj-$(CONFIG_XEN_PCI_PASSTHROUGH) += xen-host-pci-device.o
> > obj-$(CONFIG_XEN_PCI_PASSTHROUGH) += xen_pt.o
> xen_pt_config_init.o xen_pt_msi.o
> > +obj-$(CONFIG_XEN) += xen_pvdevice.o
> > diff --git a/hw/xen/xen_pvdevice.c b/hw/xen/xen_pvdevice.c
> > new file mode 100644
> > index 0000000..dbc4bf5
> > --- /dev/null
> > +++ b/hw/xen/xen_pvdevice.c
> > @@ -0,0 +1,131 @@
> > +/* Copyright (c) Citrix Systems Inc.
> > + * All rights reserved.
>
> Like Anthony wrote before, All rights reserved contradicts what's
> written below.
> Aside from this, it looks OK to me.
>
> I would like to see the libxl side patch.
Working on it, but it's not required to use the new device so I don't think the QEMU patch need be predicated on it.
Paul
^ permalink raw reply [flat|nested] 10+ messages in thread
* [PATCH] Xen PV Device
@ 2013-07-03 11:19 Paul Durrant
0 siblings, 0 replies; 10+ messages in thread
From: Paul Durrant @ 2013-07-03 11:19 UTC (permalink / raw)
To: qemu-devel, xen-devel; +Cc: Paul Durrant, Stefano Stabellini
This patch introduces a new Xen PV PCI device which will act as a new
binding point for PV drivers for Xen.
The device has parameterized vendor-id, device-id and revision to allow to
be configured as a binding point for any vendor's PV drivers.
Signed-off-by: Paul Durrant <paul.durrant@citrix.com>
Cc: Stefano Stabellini <stefano.stabellini@citrix.com>
---
hw/xen/Makefile.objs | 1 +
hw/xen/xen_pvdevice.c | 131 ++++++++++++++++++++++++++++++++++++++++++++++
include/hw/pci/pci_ids.h | 5 +-
trace-events | 4 ++
4 files changed, 139 insertions(+), 2 deletions(-)
create mode 100644 hw/xen/xen_pvdevice.c
diff --git a/hw/xen/Makefile.objs b/hw/xen/Makefile.objs
index 2017560..fd88003 100644
--- a/hw/xen/Makefile.objs
+++ b/hw/xen/Makefile.objs
@@ -4,3 +4,4 @@ common-obj-$(CONFIG_XEN_BACKEND) += xen_backend.o xen_devconfig.o
obj-$(CONFIG_XEN_I386) += xen_platform.o xen_apic.o
obj-$(CONFIG_XEN_PCI_PASSTHROUGH) += xen-host-pci-device.o
obj-$(CONFIG_XEN_PCI_PASSTHROUGH) += xen_pt.o xen_pt_config_init.o xen_pt_msi.o
+obj-$(CONFIG_XEN) += xen_pvdevice.o
diff --git a/hw/xen/xen_pvdevice.c b/hw/xen/xen_pvdevice.c
new file mode 100644
index 0000000..dbc4bf5
--- /dev/null
+++ b/hw/xen/xen_pvdevice.c
@@ -0,0 +1,131 @@
+/* Copyright (c) Citrix Systems Inc.
+ * All rights reserved.
+ *
+ * Redistribution and use in source and binary forms,
+ * with or without modification, are permitted provided
+ * that the following conditions are met:
+ *
+ * * Redistributions of source code must retain the above
+ * copyright notice, this list of conditions and the
+ * following disclaimer.
+ * * Redistributions in binary form must reproduce the above
+ * copyright notice, this list of conditions and the
+ * following disclaimer in the documentation and/or other
+ * materials provided with the distribution.
+ *
+ * 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
+ * MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
+ * DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR
+ * CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ * 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 DAMAGE.
+ */
+
+#include "hw/hw.h"
+#include "hw/pci/pci.h"
+#include "trace.h"
+
+#define TYPE_XEN_PV_DEVICE "xen-pvdevice"
+
+#define XEN_PV_DEVICE(obj) \
+ OBJECT_CHECK(XenPVDevice, (obj), TYPE_XEN_PV_DEVICE)
+
+typedef struct XenPVDevice {
+ /*< private >*/
+ PCIDevice parent_obj;
+ /*< public >*/
+ uint16_t vendor_id;
+ uint16_t device_id;
+ uint8_t revision;
+ uint32_t size;
+ MemoryRegion mmio;
+} XenPVDevice;
+
+static uint64_t xen_pv_mmio_read(void *opaque, hwaddr addr,
+ unsigned size)
+{
+ trace_xen_pv_mmio_read(addr);
+
+ return ~(uint64_t)0;
+}
+
+static void xen_pv_mmio_write(void *opaque, hwaddr addr,
+ uint64_t val, unsigned size)
+{
+ trace_xen_pv_mmio_write(addr);
+}
+
+static const MemoryRegionOps xen_pv_mmio_ops = {
+ .read = &xen_pv_mmio_read,
+ .write = &xen_pv_mmio_write,
+ .endianness = DEVICE_LITTLE_ENDIAN,
+};
+
+static int xen_pv_init(PCIDevice *pci_dev)
+{
+ XenPVDevice *d = XEN_PV_DEVICE(pci_dev);
+ uint8_t *pci_conf;
+
+ pci_conf = pci_dev->config;
+
+ pci_set_word(pci_conf + PCI_VENDOR_ID, d->vendor_id);
+ pci_set_word(pci_conf + PCI_SUBSYSTEM_VENDOR_ID, d->vendor_id);
+ pci_set_word(pci_conf + PCI_DEVICE_ID, d->device_id);
+ pci_set_word(pci_conf + PCI_SUBSYSTEM_ID, d->device_id);
+ pci_set_byte(pci_conf + PCI_REVISION_ID, d->revision);
+
+ pci_set_word(pci_conf + PCI_COMMAND, PCI_COMMAND_MEMORY);
+
+ pci_config_set_prog_interface(pci_conf, 0);
+
+ pci_conf[PCI_INTERRUPT_PIN] = 1;
+
+ memory_region_init_io(&d->mmio, &xen_pv_mmio_ops, d,
+ "xen-pvdevice-mmio", d->size);
+
+ pci_register_bar(pci_dev, 1, PCI_BASE_ADDRESS_MEM_PREFETCH,
+ &d->mmio);
+
+ return 0;
+}
+
+static Property xen_pv_props[] = {
+ DEFINE_PROP_UINT16("vendor-id", XenPVDevice, vendor_id, PCI_VENDOR_ID_XEN),
+ DEFINE_PROP_UINT16("device-id", XenPVDevice, device_id, PCI_DEVICE_ID_XEN_PVDEVICE),
+ DEFINE_PROP_UINT8("revision", XenPVDevice, revision, 0x01),
+ DEFINE_PROP_UINT32("size", XenPVDevice, size, 0x400000),
+ DEFINE_PROP_END_OF_LIST()
+};
+
+static void xen_pv_class_init(ObjectClass *klass, void *data)
+{
+ DeviceClass *dc = DEVICE_CLASS(klass);
+ PCIDeviceClass *k = PCI_DEVICE_CLASS(klass);
+
+ k->init = xen_pv_init;
+ k->class_id = PCI_CLASS_SYSTEM_OTHER;
+ dc->desc = "Xen PV Device";
+ dc->props = xen_pv_props;
+}
+
+static const TypeInfo xen_pv_type_info = {
+ .name = TYPE_XEN_PV_DEVICE,
+ .parent = TYPE_PCI_DEVICE,
+ .instance_size = sizeof(XenPVDevice),
+ .class_init = xen_pv_class_init,
+};
+
+static void xen_pv_register_types(void)
+{
+ type_register_static(&xen_pv_type_info);
+}
+
+type_init(xen_pv_register_types)
diff --git a/include/hw/pci/pci_ids.h b/include/hw/pci/pci_ids.h
index d8dc2f1..263bca3 100644
--- a/include/hw/pci/pci_ids.h
+++ b/include/hw/pci/pci_ids.h
@@ -142,8 +142,9 @@
#define PCI_DEVICE_ID_INTEL_Q35_MCH 0x29c0
-#define PCI_VENDOR_ID_XEN 0x5853
-#define PCI_DEVICE_ID_XEN_PLATFORM 0x0001
+#define PCI_VENDOR_ID_XEN 0x5853
+#define PCI_DEVICE_ID_XEN_PLATFORM 0x0001
+#define PCI_DEVICE_ID_XEN_PVDEVICE 0x0002
#define PCI_VENDOR_ID_NEC 0x1033
#define PCI_DEVICE_ID_NEC_UPD720200 0x0194
diff --git a/trace-events b/trace-events
index c5f1ccb..0445853 100644
--- a/trace-events
+++ b/trace-events
@@ -1161,3 +1161,7 @@ kvm_run_exit(int cpu_index, uint32_t reason) "cpu_index %d, reason %d"
# qom/object.c
object_dynamic_cast_assert(const char *type, const char *target, const char *file, int line, const char *func) "%s->%s (%s:%d:%s)"
object_class_dynamic_cast_assert(const char *type, const char *target, const char *file, int line, const char *func) "%s->%s (%s:%d:%s)"
+
+# hw/xen/xen_pvdevice.c
+xen_pv_mmio_read(uint64_t addr) "WARNING: read from Xen PV Device MMIO space (address %"PRIx64")"
+xen_pv_mmio_write(uint64_t addr) "WARNING: write to Xen PV Device MMIO space (address %"PRIx64")"
--
1.7.10.4
^ permalink raw reply related [flat|nested] 10+ messages in thread
end of thread, other threads:[~2013-07-04 8:21 UTC | newest]
Thread overview: 10+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2013-07-03 11:19 [Qemu-devel] [PATCH] Xen PV Device Paul Durrant
2013-07-03 16:37 ` Stefano Stabellini
2013-07-03 16:37 ` [Qemu-devel] " Stefano Stabellini
2013-07-04 7:15 ` Andreas Färber
2013-07-04 7:15 ` [Qemu-devel] " Andreas Färber
2013-07-04 8:14 ` Paul Durrant
2013-07-04 8:14 ` Paul Durrant
2013-07-04 8:17 ` Paul Durrant
2013-07-04 8:17 ` [Qemu-devel] " Paul Durrant
2013-07-03 11:19 Paul Durrant
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.