[PULL 00/20 1.2] kvm updates

All of lore.kernel.org
 help / color / mirror / Atom feed

* [PULL 00/20 1.2] kvm updates
@ 2012-05-21 16:37 ` Avi Kivity
  0 siblings, 0 replies; 64+ messages in thread
From: Avi Kivity @ 2012-05-21 16:37 UTC (permalink / raw)
  To: Anthony Liguori; +Cc: qemu-devel, kvm, Jan Kiszka, Marcelo Tosatti

Please pull from:

  git://git.kernel.org/pub/scm/virt/kvm/qemu-kvm.git uq/master

This is mostly Jan's MSI for KVM patches.

----------------------------------------------------------------
Avi Kivity (1):
      kvm: update vmxcap for EPT A/D, INVPCID, RDRAND, VMFUNC

Jan Kiszka (19):
      kvm: Refactor KVMState::max_gsi to gsi_count
      Introduce MSIMessage structure
      kvm: Introduce basic MSI support for in-kernel irqchips
      pc: Enable MSI support at APIC level
      kvm: x86: Wire up MSI support for in-kernel irqchip
      kvm: Update kernel headers
      kvm: Add support for direct MSI injections
      kvm: Enable in-kernel irqchip support by default
      msix: Factor out msix_get_message
      msix: Invoke msix_handle_mask_update on msix_mask_all
      msix: Introduce vector notifiers
      kvm: Rename kvm_irqchip_add_route to kvm_irqchip_add_irq_route
      kvm: Introduce kvm_irqchip_add_msi_route
      kvm: Publicize kvm_irqchip_release_virq
      kvm: Make kvm_irqchip_commit_routes an internal service
      kvm: Introduce kvm_irqchip_add/remove_irqfd
      kvm: Enable use of kvm_irqchip_in_kernel in hwlib code
      msix: Add msix_nr_vectors_allocated
      virtio/vhost: Add support for KVM in-kernel MSI injection

 hw/apic.c                 |    3 +
 hw/kvm/apic.c             |   34 ++++++-
 hw/msi.h                  |    5 +
 hw/msix.c                 |  121 +++++++++++++++++++++--
 hw/msix.h                 |    6 ++
 hw/pc.c                   |    9 --
 hw/pc_piix.c              |   14 +--
 hw/pci.h                  |    8 ++
 hw/virtio-pci.c           |  126 ++++++++++++++++++++++++
 hw/virtio-pci.h           |    6 ++
 hw/xen.h                  |   10 --
 hw/xen_apic.c             |    5 +
 kvm-all.c                 |  236 +++++++++++++++++++++++++++++++++++++++++++--
 kvm-stub.c                |   23 +++++
 kvm.h                     |   18 ++--
 linux-headers/linux/kvm.h |   38 ++++++++
 qemu-common.h             |    1 +
 scripts/kvm/vmxcap        |   13 +++
 18 files changed, 624 insertions(+), 52 deletions(-)

^ permalink raw reply	[flat|nested] 64+ messages in thread

* [Qemu-devel] [PULL 00/20 1.2] kvm updates
@ 2012-05-21 16:37 ` Avi Kivity
  0 siblings, 0 replies; 64+ messages in thread
From: Avi Kivity @ 2012-05-21 16:37 UTC (permalink / raw)
  To: Anthony Liguori; +Cc: Jan Kiszka, Marcelo Tosatti, qemu-devel, kvm

Please pull from:

  git://git.kernel.org/pub/scm/virt/kvm/qemu-kvm.git uq/master

This is mostly Jan's MSI for KVM patches.

----------------------------------------------------------------
Avi Kivity (1):
      kvm: update vmxcap for EPT A/D, INVPCID, RDRAND, VMFUNC

Jan Kiszka (19):
      kvm: Refactor KVMState::max_gsi to gsi_count
      Introduce MSIMessage structure
      kvm: Introduce basic MSI support for in-kernel irqchips
      pc: Enable MSI support at APIC level
      kvm: x86: Wire up MSI support for in-kernel irqchip
      kvm: Update kernel headers
      kvm: Add support for direct MSI injections
      kvm: Enable in-kernel irqchip support by default
      msix: Factor out msix_get_message
      msix: Invoke msix_handle_mask_update on msix_mask_all
      msix: Introduce vector notifiers
      kvm: Rename kvm_irqchip_add_route to kvm_irqchip_add_irq_route
      kvm: Introduce kvm_irqchip_add_msi_route
      kvm: Publicize kvm_irqchip_release_virq
      kvm: Make kvm_irqchip_commit_routes an internal service
      kvm: Introduce kvm_irqchip_add/remove_irqfd
      kvm: Enable use of kvm_irqchip_in_kernel in hwlib code
      msix: Add msix_nr_vectors_allocated
      virtio/vhost: Add support for KVM in-kernel MSI injection

 hw/apic.c                 |    3 +
 hw/kvm/apic.c             |   34 ++++++-
 hw/msi.h                  |    5 +
 hw/msix.c                 |  121 +++++++++++++++++++++--
 hw/msix.h                 |    6 ++
 hw/pc.c                   |    9 --
 hw/pc_piix.c              |   14 +--
 hw/pci.h                  |    8 ++
 hw/virtio-pci.c           |  126 ++++++++++++++++++++++++
 hw/virtio-pci.h           |    6 ++
 hw/xen.h                  |   10 --
 hw/xen_apic.c             |    5 +
 kvm-all.c                 |  236 +++++++++++++++++++++++++++++++++++++++++++--
 kvm-stub.c                |   23 +++++
 kvm.h                     |   18 ++--
 linux-headers/linux/kvm.h |   38 ++++++++
 qemu-common.h             |    1 +
 scripts/kvm/vmxcap        |   13 +++
 18 files changed, 624 insertions(+), 52 deletions(-)

^ permalink raw reply	[flat|nested] 64+ messages in thread

* [PATCH 01/19] kvm: Refactor KVMState::max_gsi to gsi_count
  2012-05-21 16:37 ` [Qemu-devel] " Avi Kivity
@ 2012-05-21 16:37   ` Avi Kivity
  -1 siblings, 0 replies; 64+ messages in thread
From: Avi Kivity @ 2012-05-21 16:37 UTC (permalink / raw)
  To: Anthony Liguori; +Cc: qemu-devel, kvm, Jan Kiszka, Marcelo Tosatti

From: Jan Kiszka <jan.kiszka@siemens.com>

Instead of the bitmap size, store the maximum of GSIs the kernel
support. Move the GSI limit assertion to the API function
kvm_irqchip_add_route and make it stricter.

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
---
 kvm-all.c |    8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/kvm-all.c b/kvm-all.c
index 9b73ccf..2d82d54 100644
--- a/kvm-all.c
+++ b/kvm-all.c
@@ -86,7 +86,7 @@ struct KVMState
     struct kvm_irq_routing *irq_routes;
     int nr_allocated_irq_routes;
     uint32_t *used_gsi_bitmap;
-    unsigned int max_gsi;
+    unsigned int gsi_count;
 #endif
 };
 
@@ -859,8 +859,6 @@ int kvm_irqchip_set_irq(KVMState *s, int irq, int level)
 #ifdef KVM_CAP_IRQ_ROUTING
 static void set_gsi(KVMState *s, unsigned int gsi)
 {
-    assert(gsi < s->max_gsi);
-
     s->used_gsi_bitmap[gsi / 32] |= 1U << (gsi % 32);
 }
 
@@ -875,7 +873,7 @@ static void kvm_init_irq_routing(KVMState *s)
         /* Round up so we can search ints using ffs */
         gsi_bits = ALIGN(gsi_count, 32);
         s->used_gsi_bitmap = g_malloc0(gsi_bits / 8);
-        s->max_gsi = gsi_bits;
+        s->gsi_count = gsi_count;
 
         /* Mark any over-allocated bits as already in use */
         for (i = gsi_count; i < gsi_bits; i++) {
@@ -920,6 +918,8 @@ void kvm_irqchip_add_route(KVMState *s, int irq, int irqchip, int pin)
 {
     struct kvm_irq_routing_entry e;
 
+    assert(pin < s->gsi_count);
+
     e.gsi = irq;
     e.type = KVM_IRQ_ROUTING_IRQCHIP;
     e.flags = 0;
-- 
1.7.10.1


^ permalink raw reply related	[flat|nested] 64+ messages in thread

* [Qemu-devel] [PATCH 01/19] kvm: Refactor KVMState::max_gsi to gsi_count
@ 2012-05-21 16:37   ` Avi Kivity
  0 siblings, 0 replies; 64+ messages in thread
From: Avi Kivity @ 2012-05-21 16:37 UTC (permalink / raw)
  To: Anthony Liguori; +Cc: Jan Kiszka, Marcelo Tosatti, qemu-devel, kvm

From: Jan Kiszka <jan.kiszka@siemens.com>

Instead of the bitmap size, store the maximum of GSIs the kernel
support. Move the GSI limit assertion to the API function
kvm_irqchip_add_route and make it stricter.

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
---
 kvm-all.c |    8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/kvm-all.c b/kvm-all.c
index 9b73ccf..2d82d54 100644
--- a/kvm-all.c
+++ b/kvm-all.c
@@ -86,7 +86,7 @@ struct KVMState
     struct kvm_irq_routing *irq_routes;
     int nr_allocated_irq_routes;
     uint32_t *used_gsi_bitmap;
-    unsigned int max_gsi;
+    unsigned int gsi_count;
 #endif
 };
 
@@ -859,8 +859,6 @@ int kvm_irqchip_set_irq(KVMState *s, int irq, int level)
 #ifdef KVM_CAP_IRQ_ROUTING
 static void set_gsi(KVMState *s, unsigned int gsi)
 {
-    assert(gsi < s->max_gsi);
-
     s->used_gsi_bitmap[gsi / 32] |= 1U << (gsi % 32);
 }
 
@@ -875,7 +873,7 @@ static void kvm_init_irq_routing(KVMState *s)
         /* Round up so we can search ints using ffs */
         gsi_bits = ALIGN(gsi_count, 32);
         s->used_gsi_bitmap = g_malloc0(gsi_bits / 8);
-        s->max_gsi = gsi_bits;
+        s->gsi_count = gsi_count;
 
         /* Mark any over-allocated bits as already in use */
         for (i = gsi_count; i < gsi_bits; i++) {
@@ -920,6 +918,8 @@ void kvm_irqchip_add_route(KVMState *s, int irq, int irqchip, int pin)
 {
     struct kvm_irq_routing_entry e;
 
+    assert(pin < s->gsi_count);
+
     e.gsi = irq;
     e.type = KVM_IRQ_ROUTING_IRQCHIP;
     e.flags = 0;
-- 
1.7.10.1

^ permalink raw reply related	[flat|nested] 64+ messages in thread

* [PATCH 02/19] Introduce MSIMessage structure
  2012-05-21 16:37 ` [Qemu-devel] " Avi Kivity
@ 2012-05-21 16:37   ` Avi Kivity
  -1 siblings, 0 replies; 64+ messages in thread
From: Avi Kivity @ 2012-05-21 16:37 UTC (permalink / raw)
  To: Anthony Liguori; +Cc: qemu-devel, kvm, Jan Kiszka, Marcelo Tosatti

From: Jan Kiszka <jan.kiszka@siemens.com>

Will be used for generating and distributing MSI messages, both in
emulation mode and under KVM.

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
---
 hw/msi.h      |    5 +++++
 qemu-common.h |    1 +
 2 files changed, 6 insertions(+)

diff --git a/hw/msi.h b/hw/msi.h
index 3040bb0..75747ab 100644
--- a/hw/msi.h
+++ b/hw/msi.h
@@ -24,6 +24,11 @@
 #include "qemu-common.h"
 #include "pci.h"
 
+struct MSIMessage {
+    uint64_t address;
+    uint32_t data;
+};
+
 extern bool msi_supported;
 
 bool msi_enabled(const PCIDevice *dev);
diff --git a/qemu-common.h b/qemu-common.h
index 50f659a..57fe28f 100644
--- a/qemu-common.h
+++ b/qemu-common.h
@@ -248,6 +248,7 @@ typedef struct PCIEAERLog PCIEAERLog;
 typedef struct PCIEAERErr PCIEAERErr;
 typedef struct PCIEPort PCIEPort;
 typedef struct PCIESlot PCIESlot;
+typedef struct MSIMessage MSIMessage;
 typedef struct SerialState SerialState;
 typedef struct IRQState *qemu_irq;
 typedef struct PCMCIACardState PCMCIACardState;
-- 
1.7.10.1


^ permalink raw reply related	[flat|nested] 64+ messages in thread

* [Qemu-devel] [PATCH 02/19] Introduce MSIMessage structure
@ 2012-05-21 16:37   ` Avi Kivity
  0 siblings, 0 replies; 64+ messages in thread
From: Avi Kivity @ 2012-05-21 16:37 UTC (permalink / raw)
  To: Anthony Liguori; +Cc: Jan Kiszka, Marcelo Tosatti, qemu-devel, kvm

From: Jan Kiszka <jan.kiszka@siemens.com>

Will be used for generating and distributing MSI messages, both in
emulation mode and under KVM.

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
---
 hw/msi.h      |    5 +++++
 qemu-common.h |    1 +
 2 files changed, 6 insertions(+)

diff --git a/hw/msi.h b/hw/msi.h
index 3040bb0..75747ab 100644
--- a/hw/msi.h
+++ b/hw/msi.h
@@ -24,6 +24,11 @@
 #include "qemu-common.h"
 #include "pci.h"
 
+struct MSIMessage {
+    uint64_t address;
+    uint32_t data;
+};
+
 extern bool msi_supported;
 
 bool msi_enabled(const PCIDevice *dev);
diff --git a/qemu-common.h b/qemu-common.h
index 50f659a..57fe28f 100644
--- a/qemu-common.h
+++ b/qemu-common.h
@@ -248,6 +248,7 @@ typedef struct PCIEAERLog PCIEAERLog;
 typedef struct PCIEAERErr PCIEAERErr;
 typedef struct PCIEPort PCIEPort;
 typedef struct PCIESlot PCIESlot;
+typedef struct MSIMessage MSIMessage;
 typedef struct SerialState SerialState;
 typedef struct IRQState *qemu_irq;
 typedef struct PCMCIACardState PCMCIACardState;
-- 
1.7.10.1

^ permalink raw reply related	[flat|nested] 64+ messages in thread

* [PATCH 03/19] kvm: Introduce basic MSI support for in-kernel irqchips
  2012-05-21 16:37 ` [Qemu-devel] " Avi Kivity
@ 2012-05-21 16:37   ` Avi Kivity
  -1 siblings, 0 replies; 64+ messages in thread
From: Avi Kivity @ 2012-05-21 16:37 UTC (permalink / raw)
  To: Anthony Liguori; +Cc: qemu-devel, kvm, Jan Kiszka, Marcelo Tosatti

From: Jan Kiszka <jan.kiszka@siemens.com>

This patch basically adds kvm_irqchip_send_msi, a service for sending
arbitrary MSI messages to KVM's in-kernel irqchip models.

As the original KVM API requires us to establish a static route from a
pseudo GSI to the target MSI message and inject the MSI via toggling
that virtual IRQ, we need to play some tricks to make this interface
transparent. We create those routes on demand and keep them in a hash
table. Succeeding messages can then search for an existing route in the
table first and reuse it whenever possible. If we should run out of
limited GSIs, we simply flush the table and rebuild it as messages are
sent.

This approach is rather simple and could be optimized further. However,
latest kernels contains a more efficient MSI injection interface that
will obsolete the GSI-based dynamic injection.

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
---
 kvm-all.c |  139 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++-
 kvm.h     |    1 +
 2 files changed, 139 insertions(+), 1 deletion(-)

diff --git a/kvm-all.c b/kvm-all.c
index 2d82d54..ff0534b 100644
--- a/kvm-all.c
+++ b/kvm-all.c
@@ -24,6 +24,7 @@
 #include "qemu-barrier.h"
 #include "sysemu.h"
 #include "hw/hw.h"
+#include "hw/msi.h"
 #include "gdbstub.h"
 #include "kvm.h"
 #include "bswap.h"
@@ -48,6 +49,8 @@
     do { } while (0)
 #endif
 
+#define KVM_MSI_HASHTAB_SIZE    256
+
 typedef struct KVMSlot
 {
     target_phys_addr_t start_addr;
@@ -59,6 +62,11 @@
 
 typedef struct kvm_dirty_log KVMDirtyLog;
 
+typedef struct KVMMSIRoute {
+    struct kvm_irq_routing_entry kroute;
+    QTAILQ_ENTRY(KVMMSIRoute) entry;
+} KVMMSIRoute;
+
 struct KVMState
 {
     KVMSlot slots[32];
@@ -87,6 +95,7 @@ struct KVMState
     int nr_allocated_irq_routes;
     uint32_t *used_gsi_bitmap;
     unsigned int gsi_count;
+    QTAILQ_HEAD(msi_hashtab, KVMMSIRoute) msi_hashtab[KVM_MSI_HASHTAB_SIZE];
 #endif
 };
 
@@ -862,9 +871,14 @@ static void set_gsi(KVMState *s, unsigned int gsi)
     s->used_gsi_bitmap[gsi / 32] |= 1U << (gsi % 32);
 }
 
+static void clear_gsi(KVMState *s, unsigned int gsi)
+{
+    s->used_gsi_bitmap[gsi / 32] &= ~(1U << (gsi % 32));
+}
+
 static void kvm_init_irq_routing(KVMState *s)
 {
-    int gsi_count;
+    int gsi_count, i;
 
     gsi_count = kvm_check_extension(s, KVM_CAP_IRQ_ROUTING);
     if (gsi_count > 0) {
@@ -884,6 +898,10 @@ static void kvm_init_irq_routing(KVMState *s)
     s->irq_routes = g_malloc0(sizeof(*s->irq_routes));
     s->nr_allocated_irq_routes = 0;
 
+    for (i = 0; i < KVM_MSI_HASHTAB_SIZE; i++) {
+        QTAILQ_INIT(&s->msi_hashtab[i]);
+    }
+
     kvm_arch_init_irq_routing(s);
 }
 
@@ -934,11 +952,130 @@ int kvm_irqchip_commit_routes(KVMState *s)
     return kvm_vm_ioctl(s, KVM_SET_GSI_ROUTING, s->irq_routes);
 }
 
+static void kvm_irqchip_release_virq(KVMState *s, int virq)
+{
+    struct kvm_irq_routing_entry *e;
+    int i;
+
+    for (i = 0; i < s->irq_routes->nr; i++) {
+        e = &s->irq_routes->entries[i];
+        if (e->gsi == virq) {
+            s->irq_routes->nr--;
+            *e = s->irq_routes->entries[s->irq_routes->nr];
+        }
+    }
+    clear_gsi(s, virq);
+}
+
+static unsigned int kvm_hash_msi(uint32_t data)
+{
+    /* This is optimized for IA32 MSI layout. However, no other arch shall
+     * repeat the mistake of not providing a direct MSI injection API. */
+    return data & 0xff;
+}
+
+static void kvm_flush_dynamic_msi_routes(KVMState *s)
+{
+    KVMMSIRoute *route, *next;
+    unsigned int hash;
+
+    for (hash = 0; hash < KVM_MSI_HASHTAB_SIZE; hash++) {
+        QTAILQ_FOREACH_SAFE(route, &s->msi_hashtab[hash], entry, next) {
+            kvm_irqchip_release_virq(s, route->kroute.gsi);
+            QTAILQ_REMOVE(&s->msi_hashtab[hash], route, entry);
+            g_free(route);
+        }
+    }
+}
+
+static int kvm_irqchip_get_virq(KVMState *s)
+{
+    uint32_t *word = s->used_gsi_bitmap;
+    int max_words = ALIGN(s->gsi_count, 32) / 32;
+    int i, bit;
+    bool retry = true;
+
+again:
+    /* Return the lowest unused GSI in the bitmap */
+    for (i = 0; i < max_words; i++) {
+        bit = ffs(~word[i]);
+        if (!bit) {
+            continue;
+        }
+
+        return bit - 1 + i * 32;
+    }
+    if (retry) {
+        retry = false;
+        kvm_flush_dynamic_msi_routes(s);
+        goto again;
+    }
+    return -ENOSPC;
+
+}
+
+static KVMMSIRoute *kvm_lookup_msi_route(KVMState *s, MSIMessage msg)
+{
+    unsigned int hash = kvm_hash_msi(msg.data);
+    KVMMSIRoute *route;
+
+    QTAILQ_FOREACH(route, &s->msi_hashtab[hash], entry) {
+        if (route->kroute.u.msi.address_lo == (uint32_t)msg.address &&
+            route->kroute.u.msi.address_hi == (msg.address >> 32) &&
+            route->kroute.u.msi.data == msg.data) {
+            return route;
+        }
+    }
+    return NULL;
+}
+
+int kvm_irqchip_send_msi(KVMState *s, MSIMessage msg)
+{
+    KVMMSIRoute *route;
+
+    route = kvm_lookup_msi_route(s, msg);
+    if (!route) {
+        int virq, ret;
+
+        virq = kvm_irqchip_get_virq(s);
+        if (virq < 0) {
+            return virq;
+        }
+
+        route = g_malloc(sizeof(KVMMSIRoute));
+        route->kroute.gsi = virq;
+        route->kroute.type = KVM_IRQ_ROUTING_MSI;
+        route->kroute.flags = 0;
+        route->kroute.u.msi.address_lo = (uint32_t)msg.address;
+        route->kroute.u.msi.address_hi = msg.address >> 32;
+        route->kroute.u.msi.data = msg.data;
+
+        kvm_add_routing_entry(s, &route->kroute);
+
+        QTAILQ_INSERT_TAIL(&s->msi_hashtab[kvm_hash_msi(msg.data)], route,
+                           entry);
+
+        ret = kvm_irqchip_commit_routes(s);
+        if (ret < 0) {
+            return ret;
+        }
+    }
+
+    assert(route->kroute.type == KVM_IRQ_ROUTING_MSI);
+
+    return kvm_irqchip_set_irq(s, route->kroute.gsi, 1);
+}
+
 #else /* !KVM_CAP_IRQ_ROUTING */
 
 static void kvm_init_irq_routing(KVMState *s)
 {
 }
+
+int kvm_irqchip_send_msi(KVMState *s, MSIMessage msg)
+{
+    abort();
+}
 #endif /* !KVM_CAP_IRQ_ROUTING */
 
 static int kvm_irqchip_create(KVMState *s)
diff --git a/kvm.h b/kvm.h
index 4ccae8c..7857dbf 100644
--- a/kvm.h
+++ b/kvm.h
@@ -132,6 +132,7 @@ int kvm_arch_on_sigbus(int code, void *addr);
 void kvm_arch_init_irq_routing(KVMState *s);
 
 int kvm_irqchip_set_irq(KVMState *s, int irq, int level);
+int kvm_irqchip_send_msi(KVMState *s, MSIMessage msg);
 
 void kvm_irqchip_add_route(KVMState *s, int gsi, int irqchip, int pin);
 int kvm_irqchip_commit_routes(KVMState *s);
-- 
1.7.10.1


^ permalink raw reply related	[flat|nested] 64+ messages in thread

* [Qemu-devel] [PATCH 03/19] kvm: Introduce basic MSI support for in-kernel irqchips
@ 2012-05-21 16:37   ` Avi Kivity
  0 siblings, 0 replies; 64+ messages in thread
From: Avi Kivity @ 2012-05-21 16:37 UTC (permalink / raw)
  To: Anthony Liguori; +Cc: Jan Kiszka, Marcelo Tosatti, qemu-devel, kvm

From: Jan Kiszka <jan.kiszka@siemens.com>

This patch basically adds kvm_irqchip_send_msi, a service for sending
arbitrary MSI messages to KVM's in-kernel irqchip models.

As the original KVM API requires us to establish a static route from a
pseudo GSI to the target MSI message and inject the MSI via toggling
that virtual IRQ, we need to play some tricks to make this interface
transparent. We create those routes on demand and keep them in a hash
table. Succeeding messages can then search for an existing route in the
table first and reuse it whenever possible. If we should run out of
limited GSIs, we simply flush the table and rebuild it as messages are
sent.

This approach is rather simple and could be optimized further. However,
latest kernels contains a more efficient MSI injection interface that
will obsolete the GSI-based dynamic injection.

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
---
 kvm-all.c |  139 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++-
 kvm.h     |    1 +
 2 files changed, 139 insertions(+), 1 deletion(-)

diff --git a/kvm-all.c b/kvm-all.c
index 2d82d54..ff0534b 100644
--- a/kvm-all.c
+++ b/kvm-all.c
@@ -24,6 +24,7 @@
 #include "qemu-barrier.h"
 #include "sysemu.h"
 #include "hw/hw.h"
+#include "hw/msi.h"
 #include "gdbstub.h"
 #include "kvm.h"
 #include "bswap.h"
@@ -48,6 +49,8 @@
     do { } while (0)
 #endif
 
+#define KVM_MSI_HASHTAB_SIZE    256
+
 typedef struct KVMSlot
 {
     target_phys_addr_t start_addr;
@@ -59,6 +62,11 @@
 
 typedef struct kvm_dirty_log KVMDirtyLog;
 
+typedef struct KVMMSIRoute {
+    struct kvm_irq_routing_entry kroute;
+    QTAILQ_ENTRY(KVMMSIRoute) entry;
+} KVMMSIRoute;
+
 struct KVMState
 {
     KVMSlot slots[32];
@@ -87,6 +95,7 @@ struct KVMState
     int nr_allocated_irq_routes;
     uint32_t *used_gsi_bitmap;
     unsigned int gsi_count;
+    QTAILQ_HEAD(msi_hashtab, KVMMSIRoute) msi_hashtab[KVM_MSI_HASHTAB_SIZE];
 #endif
 };
 
@@ -862,9 +871,14 @@ static void set_gsi(KVMState *s, unsigned int gsi)
     s->used_gsi_bitmap[gsi / 32] |= 1U << (gsi % 32);
 }
 
+static void clear_gsi(KVMState *s, unsigned int gsi)
+{
+    s->used_gsi_bitmap[gsi / 32] &= ~(1U << (gsi % 32));
+}
+
 static void kvm_init_irq_routing(KVMState *s)
 {
-    int gsi_count;
+    int gsi_count, i;
 
     gsi_count = kvm_check_extension(s, KVM_CAP_IRQ_ROUTING);
     if (gsi_count > 0) {
@@ -884,6 +898,10 @@ static void kvm_init_irq_routing(KVMState *s)
     s->irq_routes = g_malloc0(sizeof(*s->irq_routes));
     s->nr_allocated_irq_routes = 0;
 
+    for (i = 0; i < KVM_MSI_HASHTAB_SIZE; i++) {
+        QTAILQ_INIT(&s->msi_hashtab[i]);
+    }
+
     kvm_arch_init_irq_routing(s);
 }
 
@@ -934,11 +952,130 @@ int kvm_irqchip_commit_routes(KVMState *s)
     return kvm_vm_ioctl(s, KVM_SET_GSI_ROUTING, s->irq_routes);
 }
 
+static void kvm_irqchip_release_virq(KVMState *s, int virq)
+{
+    struct kvm_irq_routing_entry *e;
+    int i;
+
+    for (i = 0; i < s->irq_routes->nr; i++) {
+        e = &s->irq_routes->entries[i];
+        if (e->gsi == virq) {
+            s->irq_routes->nr--;
+            *e = s->irq_routes->entries[s->irq_routes->nr];
+        }
+    }
+    clear_gsi(s, virq);
+}
+
+static unsigned int kvm_hash_msi(uint32_t data)
+{
+    /* This is optimized for IA32 MSI layout. However, no other arch shall
+     * repeat the mistake of not providing a direct MSI injection API. */
+    return data & 0xff;
+}
+
+static void kvm_flush_dynamic_msi_routes(KVMState *s)
+{
+    KVMMSIRoute *route, *next;
+    unsigned int hash;
+
+    for (hash = 0; hash < KVM_MSI_HASHTAB_SIZE; hash++) {
+        QTAILQ_FOREACH_SAFE(route, &s->msi_hashtab[hash], entry, next) {
+            kvm_irqchip_release_virq(s, route->kroute.gsi);
+            QTAILQ_REMOVE(&s->msi_hashtab[hash], route, entry);
+            g_free(route);
+        }
+    }
+}
+
+static int kvm_irqchip_get_virq(KVMState *s)
+{
+    uint32_t *word = s->used_gsi_bitmap;
+    int max_words = ALIGN(s->gsi_count, 32) / 32;
+    int i, bit;
+    bool retry = true;
+
+again:
+    /* Return the lowest unused GSI in the bitmap */
+    for (i = 0; i < max_words; i++) {
+        bit = ffs(~word[i]);
+        if (!bit) {
+            continue;
+        }
+
+        return bit - 1 + i * 32;
+    }
+    if (retry) {
+        retry = false;
+        kvm_flush_dynamic_msi_routes(s);
+        goto again;
+    }
+    return -ENOSPC;
+
+}
+
+static KVMMSIRoute *kvm_lookup_msi_route(KVMState *s, MSIMessage msg)
+{
+    unsigned int hash = kvm_hash_msi(msg.data);
+    KVMMSIRoute *route;
+
+    QTAILQ_FOREACH(route, &s->msi_hashtab[hash], entry) {
+        if (route->kroute.u.msi.address_lo == (uint32_t)msg.address &&
+            route->kroute.u.msi.address_hi == (msg.address >> 32) &&
+            route->kroute.u.msi.data == msg.data) {
+            return route;
+        }
+    }
+    return NULL;
+}
+
+int kvm_irqchip_send_msi(KVMState *s, MSIMessage msg)
+{
+    KVMMSIRoute *route;
+
+    route = kvm_lookup_msi_route(s, msg);
+    if (!route) {
+        int virq, ret;
+
+        virq = kvm_irqchip_get_virq(s);
+        if (virq < 0) {
+            return virq;
+        }
+
+        route = g_malloc(sizeof(KVMMSIRoute));
+        route->kroute.gsi = virq;
+        route->kroute.type = KVM_IRQ_ROUTING_MSI;
+        route->kroute.flags = 0;
+        route->kroute.u.msi.address_lo = (uint32_t)msg.address;
+        route->kroute.u.msi.address_hi = msg.address >> 32;
+        route->kroute.u.msi.data = msg.data;
+
+        kvm_add_routing_entry(s, &route->kroute);
+
+        QTAILQ_INSERT_TAIL(&s->msi_hashtab[kvm_hash_msi(msg.data)], route,
+                           entry);
+
+        ret = kvm_irqchip_commit_routes(s);
+        if (ret < 0) {
+            return ret;
+        }
+    }
+
+    assert(route->kroute.type == KVM_IRQ_ROUTING_MSI);
+
+    return kvm_irqchip_set_irq(s, route->kroute.gsi, 1);
+}
+
 #else /* !KVM_CAP_IRQ_ROUTING */
 
 static void kvm_init_irq_routing(KVMState *s)
 {
 }
+
+int kvm_irqchip_send_msi(KVMState *s, MSIMessage msg)
+{
+    abort();
+}
 #endif /* !KVM_CAP_IRQ_ROUTING */
 
 static int kvm_irqchip_create(KVMState *s)
diff --git a/kvm.h b/kvm.h
index 4ccae8c..7857dbf 100644
--- a/kvm.h
+++ b/kvm.h
@@ -132,6 +132,7 @@ int kvm_arch_on_sigbus(int code, void *addr);
 void kvm_arch_init_irq_routing(KVMState *s);
 
 int kvm_irqchip_set_irq(KVMState *s, int irq, int level);
+int kvm_irqchip_send_msi(KVMState *s, MSIMessage msg);
 
 void kvm_irqchip_add_route(KVMState *s, int gsi, int irqchip, int pin);
 int kvm_irqchip_commit_routes(KVMState *s);
-- 
1.7.10.1

^ permalink raw reply related	[flat|nested] 64+ messages in thread

* [PATCH 04/19] pc: Enable MSI support at APIC level
  2012-05-21 16:37 ` [Qemu-devel] " Avi Kivity
@ 2012-05-21 16:37   ` Avi Kivity
  -1 siblings, 0 replies; 64+ messages in thread
From: Avi Kivity @ 2012-05-21 16:37 UTC (permalink / raw)
  To: Anthony Liguori; +Cc: qemu-devel, kvm, Jan Kiszka, Marcelo Tosatti

From: Jan Kiszka <jan.kiszka@siemens.com>

Push msi_supported enabling to the APIC implementations where we can
encapsulate the decision more cleanly, hiding the details from the
generic code.

Acked-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
---
 hw/apic.c     |    3 +++
 hw/pc.c       |    9 ---------
 hw/xen.h      |   10 ----------
 hw/xen_apic.c |    5 +++++
 4 files changed, 8 insertions(+), 19 deletions(-)

diff --git a/hw/apic.c b/hw/apic.c
index 4eeaf88..5fbf01c 100644
--- a/hw/apic.c
+++ b/hw/apic.c
@@ -19,6 +19,7 @@
 #include "apic_internal.h"
 #include "apic.h"
 #include "ioapic.h"
+#include "msi.h"
 #include "host-utils.h"
 #include "trace.h"
 #include "pc.h"
@@ -862,6 +863,8 @@ static void apic_init(APICCommonState *s)
 
     s->timer = qemu_new_timer_ns(vm_clock, apic_timer, s);
     local_apics[s->idx] = s;
+
+    msi_supported = true;
 }
 
 static void apic_class_init(ObjectClass *klass, void *data)
diff --git a/hw/pc.c b/hw/pc.c
index 4d34a33..6691b18 100644
--- a/hw/pc.c
+++ b/hw/pc.c
@@ -911,15 +911,6 @@ static DeviceState *apic_init(void *env, uint8_t apic_id)
         apic_mapped = 1;
     }
 
-    /* KVM does not support MSI yet. */
-    if (!kvm_irqchip_in_kernel()) {
-        msi_supported = true;
-    }
-
-    if (xen_msi_support()) {
-        msi_supported = true;
-    }
-
     return dev;
 }
 
diff --git a/hw/xen.h b/hw/xen.h
index 3ae4cd0..e5926b7 100644
--- a/hw/xen.h
+++ b/hw/xen.h
@@ -57,14 +57,4 @@ void xen_register_framebuffer(struct MemoryRegion *mr);
 #  define HVM_MAX_VCPUS 32
 #endif
 
-static inline int xen_msi_support(void)
-{
-#if defined(CONFIG_XEN_CTRL_INTERFACE_VERSION) \
-    && CONFIG_XEN_CTRL_INTERFACE_VERSION >= 420
-    return xen_enabled();
-#else
-    return 0;
-#endif
-}
-
 #endif /* QEMU_HW_XEN_H */
diff --git a/hw/xen_apic.c b/hw/xen_apic.c
index 1725ff6..a9e101f 100644
--- a/hw/xen_apic.c
+++ b/hw/xen_apic.c
@@ -40,6 +40,11 @@ static void xen_apic_init(APICCommonState *s)
 {
     memory_region_init_io(&s->io_memory, &xen_apic_io_ops, s, "xen-apic-msi",
                           MSI_SPACE_SIZE);
+
+#if defined(CONFIG_XEN_CTRL_INTERFACE_VERSION) \
+    && CONFIG_XEN_CTRL_INTERFACE_VERSION >= 420
+    msi_supported = true;
+#endif
 }
 
 static void xen_apic_set_base(APICCommonState *s, uint64_t val)
-- 
1.7.10.1


^ permalink raw reply related	[flat|nested] 64+ messages in thread

* [Qemu-devel] [PATCH 04/19] pc: Enable MSI support at APIC level
@ 2012-05-21 16:37   ` Avi Kivity
  0 siblings, 0 replies; 64+ messages in thread
From: Avi Kivity @ 2012-05-21 16:37 UTC (permalink / raw)
  To: Anthony Liguori; +Cc: Jan Kiszka, Marcelo Tosatti, qemu-devel, kvm

From: Jan Kiszka <jan.kiszka@siemens.com>

Push msi_supported enabling to the APIC implementations where we can
encapsulate the decision more cleanly, hiding the details from the
generic code.

Acked-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
---
 hw/apic.c     |    3 +++
 hw/pc.c       |    9 ---------
 hw/xen.h      |   10 ----------
 hw/xen_apic.c |    5 +++++
 4 files changed, 8 insertions(+), 19 deletions(-)

diff --git a/hw/apic.c b/hw/apic.c
index 4eeaf88..5fbf01c 100644
--- a/hw/apic.c
+++ b/hw/apic.c
@@ -19,6 +19,7 @@
 #include "apic_internal.h"
 #include "apic.h"
 #include "ioapic.h"
+#include "msi.h"
 #include "host-utils.h"
 #include "trace.h"
 #include "pc.h"
@@ -862,6 +863,8 @@ static void apic_init(APICCommonState *s)
 
     s->timer = qemu_new_timer_ns(vm_clock, apic_timer, s);
     local_apics[s->idx] = s;
+
+    msi_supported = true;
 }
 
 static void apic_class_init(ObjectClass *klass, void *data)
diff --git a/hw/pc.c b/hw/pc.c
index 4d34a33..6691b18 100644
--- a/hw/pc.c
+++ b/hw/pc.c
@@ -911,15 +911,6 @@ static DeviceState *apic_init(void *env, uint8_t apic_id)
         apic_mapped = 1;
     }
 
-    /* KVM does not support MSI yet. */
-    if (!kvm_irqchip_in_kernel()) {
-        msi_supported = true;
-    }
-
-    if (xen_msi_support()) {
-        msi_supported = true;
-    }
-
     return dev;
 }
 
diff --git a/hw/xen.h b/hw/xen.h
index 3ae4cd0..e5926b7 100644
--- a/hw/xen.h
+++ b/hw/xen.h
@@ -57,14 +57,4 @@ void xen_register_framebuffer(struct MemoryRegion *mr);
 #  define HVM_MAX_VCPUS 32
 #endif
 
-static inline int xen_msi_support(void)
-{
-#if defined(CONFIG_XEN_CTRL_INTERFACE_VERSION) \
-    && CONFIG_XEN_CTRL_INTERFACE_VERSION >= 420
-    return xen_enabled();
-#else
-    return 0;
-#endif
-}
-
 #endif /* QEMU_HW_XEN_H */
diff --git a/hw/xen_apic.c b/hw/xen_apic.c
index 1725ff6..a9e101f 100644
--- a/hw/xen_apic.c
+++ b/hw/xen_apic.c
@@ -40,6 +40,11 @@ static void xen_apic_init(APICCommonState *s)
 {
     memory_region_init_io(&s->io_memory, &xen_apic_io_ops, s, "xen-apic-msi",
                           MSI_SPACE_SIZE);
+
+#if defined(CONFIG_XEN_CTRL_INTERFACE_VERSION) \
+    && CONFIG_XEN_CTRL_INTERFACE_VERSION >= 420
+    msi_supported = true;
+#endif
 }
 
 static void xen_apic_set_base(APICCommonState *s, uint64_t val)
-- 
1.7.10.1

^ permalink raw reply related	[flat|nested] 64+ messages in thread

* [PATCH 05/19] kvm: x86: Wire up MSI support for in-kernel irqchip
  2012-05-21 16:37 ` [Qemu-devel] " Avi Kivity
@ 2012-05-21 16:37   ` Avi Kivity
  -1 siblings, 0 replies; 64+ messages in thread
From: Avi Kivity @ 2012-05-21 16:37 UTC (permalink / raw)
  To: Anthony Liguori; +Cc: qemu-devel, kvm, Jan Kiszka, Marcelo Tosatti

From: Jan Kiszka <jan.kiszka@siemens.com>

Catch writes to the MSI MMIO region in the KVM APIC and forward them to
the kernel. Provide the kernel support GSI routing, this allows to
enable MSI support also for in-kernel irqchip mode.

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
---
 hw/kvm/apic.c |   34 ++++++++++++++++++++++++++++++++--
 1 file changed, 32 insertions(+), 2 deletions(-)

diff --git a/hw/kvm/apic.c b/hw/kvm/apic.c
index ffe7a52..8ba4079 100644
--- a/hw/kvm/apic.c
+++ b/hw/kvm/apic.c
@@ -10,6 +10,7 @@
  * See the COPYING file in the top-level directory.
  */
 #include "hw/apic_internal.h"
+#include "hw/msi.h"
 #include "kvm.h"
 
 static inline void kvm_apic_set_reg(struct kvm_lapic_state *kapic,
@@ -145,10 +146,39 @@ static void kvm_apic_external_nmi(APICCommonState *s)
     run_on_cpu(s->cpu_env, do_inject_external_nmi, s);
 }
 
+static uint64_t kvm_apic_mem_read(void *opaque, target_phys_addr_t addr,
+                                  unsigned size)
+{
+    return ~(uint64_t)0;
+}
+
+static void kvm_apic_mem_write(void *opaque, target_phys_addr_t addr,
+                               uint64_t data, unsigned size)
+{
+    MSIMessage msg = { .address = addr, .data = data };
+    int ret;
+
+    ret = kvm_irqchip_send_msi(kvm_state, msg);
+    if (ret < 0) {
+        fprintf(stderr, "KVM: injection failed, MSI lost (%s)\n",
+                strerror(-ret));
+    }
+}
+
+static const MemoryRegionOps kvm_apic_io_ops = {
+    .read = kvm_apic_mem_read,
+    .write = kvm_apic_mem_write,
+    .endianness = DEVICE_NATIVE_ENDIAN,
+};
+
 static void kvm_apic_init(APICCommonState *s)
 {
-    memory_region_init_reservation(&s->io_memory, "kvm-apic-msi",
-                                   MSI_SPACE_SIZE);
+    memory_region_init_io(&s->io_memory, &kvm_apic_io_ops, s, "kvm-apic-msi",
+                          MSI_SPACE_SIZE);
+
+    if (kvm_has_gsi_routing()) {
+        msi_supported = true;
+    }
 }
 
 static void kvm_apic_class_init(ObjectClass *klass, void *data)
-- 
1.7.10.1


^ permalink raw reply related	[flat|nested] 64+ messages in thread

* [Qemu-devel] [PATCH 05/19] kvm: x86: Wire up MSI support for in-kernel irqchip
@ 2012-05-21 16:37   ` Avi Kivity
  0 siblings, 0 replies; 64+ messages in thread
From: Avi Kivity @ 2012-05-21 16:37 UTC (permalink / raw)
  To: Anthony Liguori; +Cc: Jan Kiszka, Marcelo Tosatti, qemu-devel, kvm

From: Jan Kiszka <jan.kiszka@siemens.com>

Catch writes to the MSI MMIO region in the KVM APIC and forward them to
the kernel. Provide the kernel support GSI routing, this allows to
enable MSI support also for in-kernel irqchip mode.

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
---
 hw/kvm/apic.c |   34 ++++++++++++++++++++++++++++++++--
 1 file changed, 32 insertions(+), 2 deletions(-)

diff --git a/hw/kvm/apic.c b/hw/kvm/apic.c
index ffe7a52..8ba4079 100644
--- a/hw/kvm/apic.c
+++ b/hw/kvm/apic.c
@@ -10,6 +10,7 @@
  * See the COPYING file in the top-level directory.
  */
 #include "hw/apic_internal.h"
+#include "hw/msi.h"
 #include "kvm.h"
 
 static inline void kvm_apic_set_reg(struct kvm_lapic_state *kapic,
@@ -145,10 +146,39 @@ static void kvm_apic_external_nmi(APICCommonState *s)
     run_on_cpu(s->cpu_env, do_inject_external_nmi, s);
 }
 
+static uint64_t kvm_apic_mem_read(void *opaque, target_phys_addr_t addr,
+                                  unsigned size)
+{
+    return ~(uint64_t)0;
+}
+
+static void kvm_apic_mem_write(void *opaque, target_phys_addr_t addr,
+                               uint64_t data, unsigned size)
+{
+    MSIMessage msg = { .address = addr, .data = data };
+    int ret;
+
+    ret = kvm_irqchip_send_msi(kvm_state, msg);
+    if (ret < 0) {
+        fprintf(stderr, "KVM: injection failed, MSI lost (%s)\n",
+                strerror(-ret));
+    }
+}
+
+static const MemoryRegionOps kvm_apic_io_ops = {
+    .read = kvm_apic_mem_read,
+    .write = kvm_apic_mem_write,
+    .endianness = DEVICE_NATIVE_ENDIAN,
+};
+
 static void kvm_apic_init(APICCommonState *s)
 {
-    memory_region_init_reservation(&s->io_memory, "kvm-apic-msi",
-                                   MSI_SPACE_SIZE);
+    memory_region_init_io(&s->io_memory, &kvm_apic_io_ops, s, "kvm-apic-msi",
+                          MSI_SPACE_SIZE);
+
+    if (kvm_has_gsi_routing()) {
+        msi_supported = true;
+    }
 }
 
 static void kvm_apic_class_init(ObjectClass *klass, void *data)
-- 
1.7.10.1

^ permalink raw reply related	[flat|nested] 64+ messages in thread

* [PATCH 06/19] kvm: Update kernel headers
  2012-05-21 16:37 ` [Qemu-devel] " Avi Kivity
@ 2012-05-21 16:37   ` Avi Kivity
  -1 siblings, 0 replies; 64+ messages in thread
From: Avi Kivity @ 2012-05-21 16:37 UTC (permalink / raw)
  To: Anthony Liguori; +Cc: qemu-devel, kvm, Jan Kiszka, Marcelo Tosatti

From: Jan Kiszka <jan.kiszka@siemens.com>

Corresponding kvm.git hash: f2569053e0

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
---
 linux-headers/linux/kvm.h |   38 ++++++++++++++++++++++++++++++++++++++
 1 file changed, 38 insertions(+)

diff --git a/linux-headers/linux/kvm.h b/linux-headers/linux/kvm.h
index ee7bd9c..c4426ec 100644
--- a/linux-headers/linux/kvm.h
+++ b/linux-headers/linux/kvm.h
@@ -449,6 +449,30 @@ struct kvm_ppc_pvinfo {
 	__u8  pad[108];
 };
 
+/* for KVM_PPC_GET_SMMU_INFO */
+#define KVM_PPC_PAGE_SIZES_MAX_SZ	8
+
+struct kvm_ppc_one_page_size {
+	__u32 page_shift;	/* Page shift (or 0) */
+	__u32 pte_enc;		/* Encoding in the HPTE (>>12) */
+};
+
+struct kvm_ppc_one_seg_page_size {
+	__u32 page_shift;	/* Base page shift of segment (or 0) */
+	__u32 slb_enc;		/* SLB encoding for BookS */
+	struct kvm_ppc_one_page_size enc[KVM_PPC_PAGE_SIZES_MAX_SZ];
+};
+
+#define KVM_PPC_PAGE_SIZES_REAL		0x00000001
+#define KVM_PPC_1T_SEGMENTS		0x00000002
+
+struct kvm_ppc_smmu_info {
+	__u64 flags;
+	__u32 slb_size;
+	__u32 pad;
+	struct kvm_ppc_one_seg_page_size sps[KVM_PPC_PAGE_SIZES_MAX_SZ];
+};
+
 #define KVMIO 0xAE
 
 /* machine type bits, to be used as argument to KVM_CREATE_VM */
@@ -590,6 +614,8 @@ struct kvm_ppc_pvinfo {
 #define KVM_CAP_SYNC_REGS 74
 #define KVM_CAP_PCI_2_3 75
 #define KVM_CAP_KVMCLOCK_CTRL 76
+#define KVM_CAP_SIGNAL_MSI 77
+#define KVM_CAP_PPC_GET_SMMU_INFO 78
 
 #ifdef KVM_CAP_IRQ_ROUTING
 
@@ -715,6 +741,14 @@ struct kvm_one_reg {
 	__u64 addr;
 };
 
+struct kvm_msi {
+	__u32 address_lo;
+	__u32 address_hi;
+	__u32 data;
+	__u32 flags;
+	__u8  pad[16];
+};
+
 /*
  * ioctls for VM fds
  */
@@ -789,6 +823,10 @@ struct kvm_s390_ucas_mapping {
 /* Available with KVM_CAP_PCI_2_3 */
 #define KVM_ASSIGN_SET_INTX_MASK  _IOW(KVMIO,  0xa4, \
 				       struct kvm_assigned_pci_dev)
+/* Available with KVM_CAP_SIGNAL_MSI */
+#define KVM_SIGNAL_MSI            _IOW(KVMIO,  0xa5, struct kvm_msi)
+/* Available with KVM_CAP_PPC_GET_SMMU_INFO */
+#define KVM_PPC_GET_SMMU_INFO	  _IOR(KVMIO,  0xa6, struct kvm_ppc_smmu_info)
 
 /*
  * ioctls for vcpu fds
-- 
1.7.10.1


^ permalink raw reply related	[flat|nested] 64+ messages in thread

* [Qemu-devel] [PATCH 06/19] kvm: Update kernel headers
@ 2012-05-21 16:37   ` Avi Kivity
  0 siblings, 0 replies; 64+ messages in thread
From: Avi Kivity @ 2012-05-21 16:37 UTC (permalink / raw)
  To: Anthony Liguori; +Cc: Jan Kiszka, Marcelo Tosatti, qemu-devel, kvm

From: Jan Kiszka <jan.kiszka@siemens.com>

Corresponding kvm.git hash: f2569053e0

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
---
 linux-headers/linux/kvm.h |   38 ++++++++++++++++++++++++++++++++++++++
 1 file changed, 38 insertions(+)

diff --git a/linux-headers/linux/kvm.h b/linux-headers/linux/kvm.h
index ee7bd9c..c4426ec 100644
--- a/linux-headers/linux/kvm.h
+++ b/linux-headers/linux/kvm.h
@@ -449,6 +449,30 @@ struct kvm_ppc_pvinfo {
 	__u8  pad[108];
 };
 
+/* for KVM_PPC_GET_SMMU_INFO */
+#define KVM_PPC_PAGE_SIZES_MAX_SZ	8
+
+struct kvm_ppc_one_page_size {
+	__u32 page_shift;	/* Page shift (or 0) */
+	__u32 pte_enc;		/* Encoding in the HPTE (>>12) */
+};
+
+struct kvm_ppc_one_seg_page_size {
+	__u32 page_shift;	/* Base page shift of segment (or 0) */
+	__u32 slb_enc;		/* SLB encoding for BookS */
+	struct kvm_ppc_one_page_size enc[KVM_PPC_PAGE_SIZES_MAX_SZ];
+};
+
+#define KVM_PPC_PAGE_SIZES_REAL		0x00000001
+#define KVM_PPC_1T_SEGMENTS		0x00000002
+
+struct kvm_ppc_smmu_info {
+	__u64 flags;
+	__u32 slb_size;
+	__u32 pad;
+	struct kvm_ppc_one_seg_page_size sps[KVM_PPC_PAGE_SIZES_MAX_SZ];
+};
+
 #define KVMIO 0xAE
 
 /* machine type bits, to be used as argument to KVM_CREATE_VM */
@@ -590,6 +614,8 @@ struct kvm_ppc_pvinfo {
 #define KVM_CAP_SYNC_REGS 74
 #define KVM_CAP_PCI_2_3 75
 #define KVM_CAP_KVMCLOCK_CTRL 76
+#define KVM_CAP_SIGNAL_MSI 77
+#define KVM_CAP_PPC_GET_SMMU_INFO 78
 
 #ifdef KVM_CAP_IRQ_ROUTING
 
@@ -715,6 +741,14 @@ struct kvm_one_reg {
 	__u64 addr;
 };
 
+struct kvm_msi {
+	__u32 address_lo;
+	__u32 address_hi;
+	__u32 data;
+	__u32 flags;
+	__u8  pad[16];
+};
+
 /*
  * ioctls for VM fds
  */
@@ -789,6 +823,10 @@ struct kvm_s390_ucas_mapping {
 /* Available with KVM_CAP_PCI_2_3 */
 #define KVM_ASSIGN_SET_INTX_MASK  _IOW(KVMIO,  0xa4, \
 				       struct kvm_assigned_pci_dev)
+/* Available with KVM_CAP_SIGNAL_MSI */
+#define KVM_SIGNAL_MSI            _IOW(KVMIO,  0xa5, struct kvm_msi)
+/* Available with KVM_CAP_PPC_GET_SMMU_INFO */
+#define KVM_PPC_GET_SMMU_INFO	  _IOR(KVMIO,  0xa6, struct kvm_ppc_smmu_info)
 
 /*
  * ioctls for vcpu fds
-- 
1.7.10.1

^ permalink raw reply related	[flat|nested] 64+ messages in thread

* [PATCH 07/19] kvm: Add support for direct MSI injections
  2012-05-21 16:37 ` [Qemu-devel] " Avi Kivity
@ 2012-05-21 16:37   ` Avi Kivity
  -1 siblings, 0 replies; 64+ messages in thread
From: Avi Kivity @ 2012-05-21 16:37 UTC (permalink / raw)
  To: Anthony Liguori; +Cc: qemu-devel, kvm, Jan Kiszka, Marcelo Tosatti

From: Jan Kiszka <jan.kiszka@siemens.com>

If the kernel supports KVM_SIGNAL_MSI, we can avoid the route-based
MSI injection mechanism.

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
---
 kvm-all.c |   22 +++++++++++++++++++---
 1 file changed, 19 insertions(+), 3 deletions(-)

diff --git a/kvm-all.c b/kvm-all.c
index ff0534b..d222012 100644
--- a/kvm-all.c
+++ b/kvm-all.c
@@ -96,6 +96,7 @@ struct KVMState
     uint32_t *used_gsi_bitmap;
     unsigned int gsi_count;
     QTAILQ_HEAD(msi_hashtab, KVMMSIRoute) msi_hashtab[KVM_MSI_HASHTAB_SIZE];
+    bool direct_msi;
 #endif
 };
 
@@ -898,8 +899,10 @@ static void kvm_init_irq_routing(KVMState *s)
     s->irq_routes = g_malloc0(sizeof(*s->irq_routes));
     s->nr_allocated_irq_routes = 0;
 
-    for (i = 0; i < KVM_MSI_HASHTAB_SIZE; i++) {
-        QTAILQ_INIT(&s->msi_hashtab[i]);
+    if (!s->direct_msi) {
+        for (i = 0; i < KVM_MSI_HASHTAB_SIZE; i++) {
+            QTAILQ_INIT(&s->msi_hashtab[i]);
+        }
     }
 
     kvm_arch_init_irq_routing(s);
@@ -1005,7 +1008,7 @@ static int kvm_irqchip_get_virq(KVMState *s)
 
         return bit - 1 + i * 32;
     }
-    if (retry) {
+    if (!s->direct_msi && retry) {
         retry = false;
         kvm_flush_dynamic_msi_routes(s);
         goto again;
@@ -1031,8 +1034,19 @@ static KVMMSIRoute *kvm_lookup_msi_route(KVMState *s, MSIMessage msg)
 
 int kvm_irqchip_send_msi(KVMState *s, MSIMessage msg)
 {
+    struct kvm_msi msi;
     KVMMSIRoute *route;
 
+    if (s->direct_msi) {
+        msi.address_lo = (uint32_t)msg.address;
+        msi.address_hi = msg.address >> 32;
+        msi.data = msg.data;
+        msi.flags = 0;
+        memset(msi.pad, 0, sizeof(msi.pad));
+
+        return kvm_vm_ioctl(s, KVM_SIGNAL_MSI, &msi);
+    }
+
     route = kvm_lookup_msi_route(s, msg);
     if (!route) {
         int virq, ret;
@@ -1209,6 +1223,8 @@ int kvm_init(void)
     s->pit_state2 = kvm_check_extension(s, KVM_CAP_PIT_STATE2);
 #endif
 
+    s->direct_msi = (kvm_check_extension(s, KVM_CAP_SIGNAL_MSI) > 0);
+
     ret = kvm_arch_init(s);
     if (ret < 0) {
         goto err;
-- 
1.7.10.1


^ permalink raw reply related	[flat|nested] 64+ messages in thread

* [Qemu-devel] [PATCH 07/19] kvm: Add support for direct MSI injections
@ 2012-05-21 16:37   ` Avi Kivity
  0 siblings, 0 replies; 64+ messages in thread
From: Avi Kivity @ 2012-05-21 16:37 UTC (permalink / raw)
  To: Anthony Liguori; +Cc: Jan Kiszka, Marcelo Tosatti, qemu-devel, kvm

From: Jan Kiszka <jan.kiszka@siemens.com>

If the kernel supports KVM_SIGNAL_MSI, we can avoid the route-based
MSI injection mechanism.

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
---
 kvm-all.c |   22 +++++++++++++++++++---
 1 file changed, 19 insertions(+), 3 deletions(-)

diff --git a/kvm-all.c b/kvm-all.c
index ff0534b..d222012 100644
--- a/kvm-all.c
+++ b/kvm-all.c
@@ -96,6 +96,7 @@ struct KVMState
     uint32_t *used_gsi_bitmap;
     unsigned int gsi_count;
     QTAILQ_HEAD(msi_hashtab, KVMMSIRoute) msi_hashtab[KVM_MSI_HASHTAB_SIZE];
+    bool direct_msi;
 #endif
 };
 
@@ -898,8 +899,10 @@ static void kvm_init_irq_routing(KVMState *s)
     s->irq_routes = g_malloc0(sizeof(*s->irq_routes));
     s->nr_allocated_irq_routes = 0;
 
-    for (i = 0; i < KVM_MSI_HASHTAB_SIZE; i++) {
-        QTAILQ_INIT(&s->msi_hashtab[i]);
+    if (!s->direct_msi) {
+        for (i = 0; i < KVM_MSI_HASHTAB_SIZE; i++) {
+            QTAILQ_INIT(&s->msi_hashtab[i]);
+        }
     }
 
     kvm_arch_init_irq_routing(s);
@@ -1005,7 +1008,7 @@ static int kvm_irqchip_get_virq(KVMState *s)
 
         return bit - 1 + i * 32;
     }
-    if (retry) {
+    if (!s->direct_msi && retry) {
         retry = false;
         kvm_flush_dynamic_msi_routes(s);
         goto again;
@@ -1031,8 +1034,19 @@ static KVMMSIRoute *kvm_lookup_msi_route(KVMState *s, MSIMessage msg)
 
 int kvm_irqchip_send_msi(KVMState *s, MSIMessage msg)
 {
+    struct kvm_msi msi;
     KVMMSIRoute *route;
 
+    if (s->direct_msi) {
+        msi.address_lo = (uint32_t)msg.address;
+        msi.address_hi = msg.address >> 32;
+        msi.data = msg.data;
+        msi.flags = 0;
+        memset(msi.pad, 0, sizeof(msi.pad));
+
+        return kvm_vm_ioctl(s, KVM_SIGNAL_MSI, &msi);
+    }
+
     route = kvm_lookup_msi_route(s, msg);
     if (!route) {
         int virq, ret;
@@ -1209,6 +1223,8 @@ int kvm_init(void)
     s->pit_state2 = kvm_check_extension(s, KVM_CAP_PIT_STATE2);
 #endif
 
+    s->direct_msi = (kvm_check_extension(s, KVM_CAP_SIGNAL_MSI) > 0);
+
     ret = kvm_arch_init(s);
     if (ret < 0) {
         goto err;
-- 
1.7.10.1

^ permalink raw reply related	[flat|nested] 64+ messages in thread

* [PATCH 08/19] kvm: Enable in-kernel irqchip support by default
  2012-05-21 16:37 ` [Qemu-devel] " Avi Kivity
@ 2012-05-21 16:37   ` Avi Kivity
  -1 siblings, 0 replies; 64+ messages in thread
From: Avi Kivity @ 2012-05-21 16:37 UTC (permalink / raw)
  To: Anthony Liguori; +Cc: qemu-devel, kvm, Jan Kiszka, Marcelo Tosatti

From: Jan Kiszka <jan.kiszka@siemens.com>

As MSI is now fully supported by KVM (/wrt available features in
upstream), we can finally enable the in-kernel irqchip by default.

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
---
 kvm-all.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/kvm-all.c b/kvm-all.c
index d222012..1913d6a 100644
--- a/kvm-all.c
+++ b/kvm-all.c
@@ -1099,7 +1099,7 @@ static int kvm_irqchip_create(KVMState *s)
 
     if (QTAILQ_EMPTY(&list->head) ||
         !qemu_opt_get_bool(QTAILQ_FIRST(&list->head),
-                           "kernel_irqchip", false) ||
+                           "kernel_irqchip", true) ||
         !kvm_check_extension(s, KVM_CAP_IRQCHIP)) {
         return 0;
     }
-- 
1.7.10.1


^ permalink raw reply related	[flat|nested] 64+ messages in thread

* [Qemu-devel] [PATCH 08/19] kvm: Enable in-kernel irqchip support by default
@ 2012-05-21 16:37   ` Avi Kivity
  0 siblings, 0 replies; 64+ messages in thread
From: Avi Kivity @ 2012-05-21 16:37 UTC (permalink / raw)
  To: Anthony Liguori; +Cc: Jan Kiszka, Marcelo Tosatti, qemu-devel, kvm

From: Jan Kiszka <jan.kiszka@siemens.com>

As MSI is now fully supported by KVM (/wrt available features in
upstream), we can finally enable the in-kernel irqchip by default.

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
---
 kvm-all.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/kvm-all.c b/kvm-all.c
index d222012..1913d6a 100644
--- a/kvm-all.c
+++ b/kvm-all.c
@@ -1099,7 +1099,7 @@ static int kvm_irqchip_create(KVMState *s)
 
     if (QTAILQ_EMPTY(&list->head) ||
         !qemu_opt_get_bool(QTAILQ_FIRST(&list->head),
-                           "kernel_irqchip", false) ||
+                           "kernel_irqchip", true) ||
         !kvm_check_extension(s, KVM_CAP_IRQCHIP)) {
         return 0;
     }
-- 
1.7.10.1

^ permalink raw reply related	[flat|nested] 64+ messages in thread

* [PATCH 09/19] msix: Factor out msix_get_message
  2012-05-21 16:37 ` [Qemu-devel] " Avi Kivity
@ 2012-05-21 16:37   ` Avi Kivity
  -1 siblings, 0 replies; 64+ messages in thread
From: Avi Kivity @ 2012-05-21 16:37 UTC (permalink / raw)
  To: Anthony Liguori; +Cc: qemu-devel, kvm, Jan Kiszka, Marcelo Tosatti

From: Jan Kiszka <jan.kiszka@siemens.com>

This helper will also be used by the upcoming config notifier.

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
---
 hw/msix.c |   19 +++++++++++++------
 1 file changed, 13 insertions(+), 6 deletions(-)

diff --git a/hw/msix.c b/hw/msix.c
index 3835eaa..3197465 100644
--- a/hw/msix.c
+++ b/hw/msix.c
@@ -35,6 +35,15 @@
 #define MSIX_PAGE_PENDING (MSIX_PAGE_SIZE / 2)
 #define MSIX_MAX_ENTRIES 32
 
+static MSIMessage msix_get_message(PCIDevice *dev, unsigned vector)
+{
+    uint8_t *table_entry = dev->msix_table_page + vector * PCI_MSIX_ENTRY_SIZE;
+    MSIMessage msg;
+
+    msg.address = pci_get_quad(table_entry + PCI_MSIX_ENTRY_LOWER_ADDR);
+    msg.data = pci_get_long(table_entry + PCI_MSIX_ENTRY_DATA);
+    return msg;
+}
 
 /* Add MSI-X capability to the config space for the device. */
 /* Given a bar and its size, add MSI-X table on top of it
@@ -352,9 +361,7 @@ uint32_t msix_bar_size(PCIDevice *dev)
 /* Send an MSI-X message */
 void msix_notify(PCIDevice *dev, unsigned vector)
 {
-    uint8_t *table_entry = dev->msix_table_page + vector * PCI_MSIX_ENTRY_SIZE;
-    uint64_t address;
-    uint32_t data;
+    MSIMessage msg;
 
     if (vector >= dev->msix_entries_nr || !dev->msix_entry_used[vector])
         return;
@@ -363,9 +370,9 @@ void msix_notify(PCIDevice *dev, unsigned vector)
         return;
     }
 
-    address = pci_get_quad(table_entry + PCI_MSIX_ENTRY_LOWER_ADDR);
-    data = pci_get_long(table_entry + PCI_MSIX_ENTRY_DATA);
-    stl_le_phys(address, data);
+    msg = msix_get_message(dev, vector);
+
+    stl_le_phys(msg.address, msg.data);
 }
 
 void msix_reset(PCIDevice *dev)
-- 
1.7.10.1


^ permalink raw reply related	[flat|nested] 64+ messages in thread

* [Qemu-devel] [PATCH 09/19] msix: Factor out msix_get_message
@ 2012-05-21 16:37   ` Avi Kivity
  0 siblings, 0 replies; 64+ messages in thread
From: Avi Kivity @ 2012-05-21 16:37 UTC (permalink / raw)
  To: Anthony Liguori; +Cc: Jan Kiszka, Marcelo Tosatti, qemu-devel, kvm

From: Jan Kiszka <jan.kiszka@siemens.com>

This helper will also be used by the upcoming config notifier.

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
---
 hw/msix.c |   19 +++++++++++++------
 1 file changed, 13 insertions(+), 6 deletions(-)

diff --git a/hw/msix.c b/hw/msix.c
index 3835eaa..3197465 100644
--- a/hw/msix.c
+++ b/hw/msix.c
@@ -35,6 +35,15 @@
 #define MSIX_PAGE_PENDING (MSIX_PAGE_SIZE / 2)
 #define MSIX_MAX_ENTRIES 32
 
+static MSIMessage msix_get_message(PCIDevice *dev, unsigned vector)
+{
+    uint8_t *table_entry = dev->msix_table_page + vector * PCI_MSIX_ENTRY_SIZE;
+    MSIMessage msg;
+
+    msg.address = pci_get_quad(table_entry + PCI_MSIX_ENTRY_LOWER_ADDR);
+    msg.data = pci_get_long(table_entry + PCI_MSIX_ENTRY_DATA);
+    return msg;
+}
 
 /* Add MSI-X capability to the config space for the device. */
 /* Given a bar and its size, add MSI-X table on top of it
@@ -352,9 +361,7 @@ uint32_t msix_bar_size(PCIDevice *dev)
 /* Send an MSI-X message */
 void msix_notify(PCIDevice *dev, unsigned vector)
 {
-    uint8_t *table_entry = dev->msix_table_page + vector * PCI_MSIX_ENTRY_SIZE;
-    uint64_t address;
-    uint32_t data;
+    MSIMessage msg;
 
     if (vector >= dev->msix_entries_nr || !dev->msix_entry_used[vector])
         return;
@@ -363,9 +370,9 @@ void msix_notify(PCIDevice *dev, unsigned vector)
         return;
     }
 
-    address = pci_get_quad(table_entry + PCI_MSIX_ENTRY_LOWER_ADDR);
-    data = pci_get_long(table_entry + PCI_MSIX_ENTRY_DATA);
-    stl_le_phys(address, data);
+    msg = msix_get_message(dev, vector);
+
+    stl_le_phys(msg.address, msg.data);
 }
 
 void msix_reset(PCIDevice *dev)
-- 
1.7.10.1

^ permalink raw reply related	[flat|nested] 64+ messages in thread

* [PATCH 10/19] msix: Invoke msix_handle_mask_update on msix_mask_all
  2012-05-21 16:37 ` [Qemu-devel] " Avi Kivity
@ 2012-05-21 16:37   ` Avi Kivity
  -1 siblings, 0 replies; 64+ messages in thread
From: Avi Kivity @ 2012-05-21 16:37 UTC (permalink / raw)
  To: Anthony Liguori; +Cc: qemu-devel, kvm, Jan Kiszka, Marcelo Tosatti

From: Jan Kiszka <jan.kiszka@siemens.com>

In preparation of firing vector notifiers on mask changes, call
msix_handle_mask_update also from msix_mask_all. So far, this will have
no real effect.

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
---
 hw/msix.c |    4 ++++
 1 file changed, 4 insertions(+)

diff --git a/hw/msix.c b/hw/msix.c
index 3197465..e1a7d92 100644
--- a/hw/msix.c
+++ b/hw/msix.c
@@ -231,10 +231,14 @@ static void msix_mmio_setup(PCIDevice *d, MemoryRegion *bar)
 static void msix_mask_all(struct PCIDevice *dev, unsigned nentries)
 {
     int vector;
+
     for (vector = 0; vector < nentries; ++vector) {
         unsigned offset =
             vector * PCI_MSIX_ENTRY_SIZE + PCI_MSIX_ENTRY_VECTOR_CTRL;
+        bool was_masked = msix_is_masked(dev, vector);
+
         dev->msix_table_page[offset] |= PCI_MSIX_ENTRY_CTRL_MASKBIT;
+        msix_handle_mask_update(dev, vector, was_masked);
     }
 }
 
-- 
1.7.10.1


^ permalink raw reply related	[flat|nested] 64+ messages in thread

* [Qemu-devel] [PATCH 10/19] msix: Invoke msix_handle_mask_update on msix_mask_all
@ 2012-05-21 16:37   ` Avi Kivity
  0 siblings, 0 replies; 64+ messages in thread
From: Avi Kivity @ 2012-05-21 16:37 UTC (permalink / raw)
  To: Anthony Liguori; +Cc: Jan Kiszka, Marcelo Tosatti, qemu-devel, kvm

From: Jan Kiszka <jan.kiszka@siemens.com>

In preparation of firing vector notifiers on mask changes, call
msix_handle_mask_update also from msix_mask_all. So far, this will have
no real effect.

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
---
 hw/msix.c |    4 ++++
 1 file changed, 4 insertions(+)

diff --git a/hw/msix.c b/hw/msix.c
index 3197465..e1a7d92 100644
--- a/hw/msix.c
+++ b/hw/msix.c
@@ -231,10 +231,14 @@ static void msix_mmio_setup(PCIDevice *d, MemoryRegion *bar)
 static void msix_mask_all(struct PCIDevice *dev, unsigned nentries)
 {
     int vector;
+
     for (vector = 0; vector < nentries; ++vector) {
         unsigned offset =
             vector * PCI_MSIX_ENTRY_SIZE + PCI_MSIX_ENTRY_VECTOR_CTRL;
+        bool was_masked = msix_is_masked(dev, vector);
+
         dev->msix_table_page[offset] |= PCI_MSIX_ENTRY_CTRL_MASKBIT;
+        msix_handle_mask_update(dev, vector, was_masked);
     }
 }
 
-- 
1.7.10.1

^ permalink raw reply related	[flat|nested] 64+ messages in thread

* [PATCH 11/19] msix: Introduce vector notifiers
  2012-05-21 16:37 ` [Qemu-devel] " Avi Kivity
@ 2012-05-21 16:37   ` Avi Kivity
  -1 siblings, 0 replies; 64+ messages in thread
From: Avi Kivity @ 2012-05-21 16:37 UTC (permalink / raw)
  To: Anthony Liguori; +Cc: qemu-devel, kvm, Jan Kiszka, Marcelo Tosatti

From: Jan Kiszka <jan.kiszka@siemens.com>

Vector notifiers shall be triggered by the MSI/MSI-X core whenever a
relevant configuration change is programmed by the guest. In case of
MSI-X, changes are reported when the effective mask (global &&
per-vector) alters its state. On unmask, the current vector
configuration is included in the event report. This allows users - e.g.
virtio-pci layer - to transfer this information to external MSI-X
routing subsystems - like vhost + KVM in-kernel irqchip.

This implementation only provides MSI-X support, but extension to MSI is
feasible and will be provided later on when adding support for KVM PCI
device assignment.

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
---
 hw/msix.c |   93 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
 hw/msix.h |    4 +++
 hw/pci.h  |    8 ++++++
 3 files changed, 105 insertions(+)

diff --git a/hw/msix.c b/hw/msix.c
index e1a7d92..1622e16 100644
--- a/hw/msix.c
+++ b/hw/msix.c
@@ -139,13 +139,34 @@ static bool msix_is_masked(PCIDevice *dev, int vector)
     return msix_vector_masked(dev, vector, dev->msix_function_masked);
 }
 
+static void msix_fire_vector_notifier(PCIDevice *dev,
+                                      unsigned int vector, bool is_masked)
+{
+    MSIMessage msg;
+    int ret;
+
+    if (!dev->msix_vector_use_notifier) {
+        return;
+    }
+    if (is_masked) {
+        dev->msix_vector_release_notifier(dev, vector);
+    } else {
+        msg = msix_get_message(dev, vector);
+        ret = dev->msix_vector_use_notifier(dev, vector, msg);
+        assert(ret >= 0);
+    }
+}
+
 static void msix_handle_mask_update(PCIDevice *dev, int vector, bool was_masked)
 {
     bool is_masked = msix_is_masked(dev, vector);
+
     if (is_masked == was_masked) {
         return;
     }
 
+    msix_fire_vector_notifier(dev, vector, is_masked);
+
     if (!is_masked && msix_is_pending(dev, vector)) {
         msix_clr_pending(dev, vector);
         msix_notify(dev, vector);
@@ -330,6 +351,7 @@ void msix_save(PCIDevice *dev, QEMUFile *f)
 void msix_load(PCIDevice *dev, QEMUFile *f)
 {
     unsigned n = dev->msix_entries_nr;
+    unsigned int vector;
 
     if (!(dev->cap_present & QEMU_PCI_CAP_MSIX)) {
         return;
@@ -339,6 +361,10 @@ void msix_load(PCIDevice *dev, QEMUFile *f)
     qemu_get_buffer(f, dev->msix_table_page, n * PCI_MSIX_ENTRY_SIZE);
     qemu_get_buffer(f, dev->msix_table_page + MSIX_PAGE_PENDING, (n + 7) / 8);
     msix_update_function_masked(dev);
+
+    for (vector = 0; vector < n; vector++) {
+        msix_handle_mask_update(dev, vector, true);
+    }
 }
 
 /* Does device support MSI-X? */
@@ -425,3 +451,70 @@ void msix_unuse_all_vectors(PCIDevice *dev)
         return;
     msix_free_irq_entries(dev);
 }
+
+static int msix_set_notifier_for_vector(PCIDevice *dev, unsigned int vector)
+{
+    MSIMessage msg;
+
+    if (msix_is_masked(dev, vector)) {
+        return 0;
+    }
+    msg = msix_get_message(dev, vector);
+    return dev->msix_vector_use_notifier(dev, vector, msg);
+}
+
+static void msix_unset_notifier_for_vector(PCIDevice *dev, unsigned int vector)
+{
+    if (msix_is_masked(dev, vector)) {
+        return;
+    }
+    dev->msix_vector_release_notifier(dev, vector);
+}
+
+int msix_set_vector_notifiers(PCIDevice *dev,
+                              MSIVectorUseNotifier use_notifier,
+                              MSIVectorReleaseNotifier release_notifier)
+{
+    int vector, ret;
+
+    assert(use_notifier && release_notifier);
+
+    dev->msix_vector_use_notifier = use_notifier;
+    dev->msix_vector_release_notifier = release_notifier;
+
+    if ((dev->config[dev->msix_cap + MSIX_CONTROL_OFFSET] &
+        (MSIX_ENABLE_MASK | MSIX_MASKALL_MASK)) == MSIX_ENABLE_MASK) {
+        for (vector = 0; vector < dev->msix_entries_nr; vector++) {
+            ret = msix_set_notifier_for_vector(dev, vector);
+            if (ret < 0) {
+                goto undo;
+            }
+        }
+    }
+    return 0;
+
+undo:
+    while (--vector >= 0) {
+        msix_unset_notifier_for_vector(dev, vector);
+    }
+    dev->msix_vector_use_notifier = NULL;
+    dev->msix_vector_release_notifier = NULL;
+    return ret;
+}
+
+void msix_unset_vector_notifiers(PCIDevice *dev)
+{
+    int vector;
+
+    assert(dev->msix_vector_use_notifier &&
+           dev->msix_vector_release_notifier);
+
+    if ((dev->config[dev->msix_cap + MSIX_CONTROL_OFFSET] &
+        (MSIX_ENABLE_MASK | MSIX_MASKALL_MASK)) == MSIX_ENABLE_MASK) {
+        for (vector = 0; vector < dev->msix_entries_nr; vector++) {
+            msix_unset_notifier_for_vector(dev, vector);
+        }
+    }
+    dev->msix_vector_use_notifier = NULL;
+    dev->msix_vector_release_notifier = NULL;
+}
diff --git a/hw/msix.h b/hw/msix.h
index 5aba22b..f33f18b 100644
--- a/hw/msix.h
+++ b/hw/msix.h
@@ -29,4 +29,8 @@ void msix_notify(PCIDevice *dev, unsigned vector);
 
 void msix_reset(PCIDevice *dev);
 
+int msix_set_vector_notifiers(PCIDevice *dev,
+                              MSIVectorUseNotifier use_notifier,
+                              MSIVectorReleaseNotifier release_notifier);
+void msix_unset_vector_notifiers(PCIDevice *dev);
 #endif
diff --git a/hw/pci.h b/hw/pci.h
index 8d0aa49..c3cacce 100644
--- a/hw/pci.h
+++ b/hw/pci.h
@@ -173,6 +173,10 @@ typedef struct PCIDeviceClass {
     const char *romfile;
 } PCIDeviceClass;
 
+typedef int (*MSIVectorUseNotifier)(PCIDevice *dev, unsigned int vector,
+                                      MSIMessage msg);
+typedef void (*MSIVectorReleaseNotifier)(PCIDevice *dev, unsigned int vector);
+
 struct PCIDevice {
     DeviceState qdev;
     /* PCI config space */
@@ -243,6 +247,10 @@ struct PCIDevice {
     bool has_rom;
     MemoryRegion rom;
     uint32_t rom_bar;
+
+    /* MSI-X notifiers */
+    MSIVectorUseNotifier msix_vector_use_notifier;
+    MSIVectorReleaseNotifier msix_vector_release_notifier;
 };
 
 void pci_register_bar(PCIDevice *pci_dev, int region_num,
-- 
1.7.10.1


^ permalink raw reply related	[flat|nested] 64+ messages in thread

* [Qemu-devel] [PATCH 11/19] msix: Introduce vector notifiers
@ 2012-05-21 16:37   ` Avi Kivity
  0 siblings, 0 replies; 64+ messages in thread
From: Avi Kivity @ 2012-05-21 16:37 UTC (permalink / raw)
  To: Anthony Liguori; +Cc: Jan Kiszka, Marcelo Tosatti, qemu-devel, kvm

From: Jan Kiszka <jan.kiszka@siemens.com>

Vector notifiers shall be triggered by the MSI/MSI-X core whenever a
relevant configuration change is programmed by the guest. In case of
MSI-X, changes are reported when the effective mask (global &&
per-vector) alters its state. On unmask, the current vector
configuration is included in the event report. This allows users - e.g.
virtio-pci layer - to transfer this information to external MSI-X
routing subsystems - like vhost + KVM in-kernel irqchip.

This implementation only provides MSI-X support, but extension to MSI is
feasible and will be provided later on when adding support for KVM PCI
device assignment.

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
---
 hw/msix.c |   93 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
 hw/msix.h |    4 +++
 hw/pci.h  |    8 ++++++
 3 files changed, 105 insertions(+)

diff --git a/hw/msix.c b/hw/msix.c
index e1a7d92..1622e16 100644
--- a/hw/msix.c
+++ b/hw/msix.c
@@ -139,13 +139,34 @@ static bool msix_is_masked(PCIDevice *dev, int vector)
     return msix_vector_masked(dev, vector, dev->msix_function_masked);
 }
 
+static void msix_fire_vector_notifier(PCIDevice *dev,
+                                      unsigned int vector, bool is_masked)
+{
+    MSIMessage msg;
+    int ret;
+
+    if (!dev->msix_vector_use_notifier) {
+        return;
+    }
+    if (is_masked) {
+        dev->msix_vector_release_notifier(dev, vector);
+    } else {
+        msg = msix_get_message(dev, vector);
+        ret = dev->msix_vector_use_notifier(dev, vector, msg);
+        assert(ret >= 0);
+    }
+}
+
 static void msix_handle_mask_update(PCIDevice *dev, int vector, bool was_masked)
 {
     bool is_masked = msix_is_masked(dev, vector);
+
     if (is_masked == was_masked) {
         return;
     }
 
+    msix_fire_vector_notifier(dev, vector, is_masked);
+
     if (!is_masked && msix_is_pending(dev, vector)) {
         msix_clr_pending(dev, vector);
         msix_notify(dev, vector);
@@ -330,6 +351,7 @@ void msix_save(PCIDevice *dev, QEMUFile *f)
 void msix_load(PCIDevice *dev, QEMUFile *f)
 {
     unsigned n = dev->msix_entries_nr;
+    unsigned int vector;
 
     if (!(dev->cap_present & QEMU_PCI_CAP_MSIX)) {
         return;
@@ -339,6 +361,10 @@ void msix_load(PCIDevice *dev, QEMUFile *f)
     qemu_get_buffer(f, dev->msix_table_page, n * PCI_MSIX_ENTRY_SIZE);
     qemu_get_buffer(f, dev->msix_table_page + MSIX_PAGE_PENDING, (n + 7) / 8);
     msix_update_function_masked(dev);
+
+    for (vector = 0; vector < n; vector++) {
+        msix_handle_mask_update(dev, vector, true);
+    }
 }
 
 /* Does device support MSI-X? */
@@ -425,3 +451,70 @@ void msix_unuse_all_vectors(PCIDevice *dev)
         return;
     msix_free_irq_entries(dev);
 }
+
+static int msix_set_notifier_for_vector(PCIDevice *dev, unsigned int vector)
+{
+    MSIMessage msg;
+
+    if (msix_is_masked(dev, vector)) {
+        return 0;
+    }
+    msg = msix_get_message(dev, vector);
+    return dev->msix_vector_use_notifier(dev, vector, msg);
+}
+
+static void msix_unset_notifier_for_vector(PCIDevice *dev, unsigned int vector)
+{
+    if (msix_is_masked(dev, vector)) {
+        return;
+    }
+    dev->msix_vector_release_notifier(dev, vector);
+}
+
+int msix_set_vector_notifiers(PCIDevice *dev,
+                              MSIVectorUseNotifier use_notifier,
+                              MSIVectorReleaseNotifier release_notifier)
+{
+    int vector, ret;
+
+    assert(use_notifier && release_notifier);
+
+    dev->msix_vector_use_notifier = use_notifier;
+    dev->msix_vector_release_notifier = release_notifier;
+
+    if ((dev->config[dev->msix_cap + MSIX_CONTROL_OFFSET] &
+        (MSIX_ENABLE_MASK | MSIX_MASKALL_MASK)) == MSIX_ENABLE_MASK) {
+        for (vector = 0; vector < dev->msix_entries_nr; vector++) {
+            ret = msix_set_notifier_for_vector(dev, vector);
+            if (ret < 0) {
+                goto undo;
+            }
+        }
+    }
+    return 0;
+
+undo:
+    while (--vector >= 0) {
+        msix_unset_notifier_for_vector(dev, vector);
+    }
+    dev->msix_vector_use_notifier = NULL;
+    dev->msix_vector_release_notifier = NULL;
+    return ret;
+}
+
+void msix_unset_vector_notifiers(PCIDevice *dev)
+{
+    int vector;
+
+    assert(dev->msix_vector_use_notifier &&
+           dev->msix_vector_release_notifier);
+
+    if ((dev->config[dev->msix_cap + MSIX_CONTROL_OFFSET] &
+        (MSIX_ENABLE_MASK | MSIX_MASKALL_MASK)) == MSIX_ENABLE_MASK) {
+        for (vector = 0; vector < dev->msix_entries_nr; vector++) {
+            msix_unset_notifier_for_vector(dev, vector);
+        }
+    }
+    dev->msix_vector_use_notifier = NULL;
+    dev->msix_vector_release_notifier = NULL;
+}
diff --git a/hw/msix.h b/hw/msix.h
index 5aba22b..f33f18b 100644
--- a/hw/msix.h
+++ b/hw/msix.h
@@ -29,4 +29,8 @@ void msix_notify(PCIDevice *dev, unsigned vector);
 
 void msix_reset(PCIDevice *dev);
 
+int msix_set_vector_notifiers(PCIDevice *dev,
+                              MSIVectorUseNotifier use_notifier,
+                              MSIVectorReleaseNotifier release_notifier);
+void msix_unset_vector_notifiers(PCIDevice *dev);
 #endif
diff --git a/hw/pci.h b/hw/pci.h
index 8d0aa49..c3cacce 100644
--- a/hw/pci.h
+++ b/hw/pci.h
@@ -173,6 +173,10 @@ typedef struct PCIDeviceClass {
     const char *romfile;
 } PCIDeviceClass;
 
+typedef int (*MSIVectorUseNotifier)(PCIDevice *dev, unsigned int vector,
+                                      MSIMessage msg);
+typedef void (*MSIVectorReleaseNotifier)(PCIDevice *dev, unsigned int vector);
+
 struct PCIDevice {
     DeviceState qdev;
     /* PCI config space */
@@ -243,6 +247,10 @@ struct PCIDevice {
     bool has_rom;
     MemoryRegion rom;
     uint32_t rom_bar;
+
+    /* MSI-X notifiers */
+    MSIVectorUseNotifier msix_vector_use_notifier;
+    MSIVectorReleaseNotifier msix_vector_release_notifier;
 };
 
 void pci_register_bar(PCIDevice *pci_dev, int region_num,
-- 
1.7.10.1

^ permalink raw reply related	[flat|nested] 64+ messages in thread

* [PATCH 12/19] kvm: Rename kvm_irqchip_add_route to kvm_irqchip_add_irq_route
  2012-05-21 16:37 ` [Qemu-devel] " Avi Kivity
@ 2012-05-21 16:37   ` Avi Kivity
  -1 siblings, 0 replies; 64+ messages in thread
From: Avi Kivity @ 2012-05-21 16:37 UTC (permalink / raw)
  To: Anthony Liguori; +Cc: qemu-devel, kvm, Jan Kiszka, Marcelo Tosatti

From: Jan Kiszka <jan.kiszka@siemens.com>

We will add kvm_irqchip_add_msi_route, so let's make the difference
clearer.

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
---
 hw/pc_piix.c |    8 ++++----
 kvm-all.c    |    2 +-
 kvm.h        |    2 +-
 3 files changed, 6 insertions(+), 6 deletions(-)

diff --git a/hw/pc_piix.c b/hw/pc_piix.c
index 6a75718..c17f906 100644
--- a/hw/pc_piix.c
+++ b/hw/pc_piix.c
@@ -63,17 +63,17 @@ static void kvm_piix3_setup_irq_routing(bool pci_enabled)
             if (i == 2) {
                 continue;
             }
-            kvm_irqchip_add_route(s, i, KVM_IRQCHIP_PIC_MASTER, i);
+            kvm_irqchip_add_irq_route(s, i, KVM_IRQCHIP_PIC_MASTER, i);
         }
         for (i = 8; i < 16; ++i) {
-            kvm_irqchip_add_route(s, i, KVM_IRQCHIP_PIC_SLAVE, i - 8);
+            kvm_irqchip_add_irq_route(s, i, KVM_IRQCHIP_PIC_SLAVE, i - 8);
         }
         if (pci_enabled) {
             for (i = 0; i < 24; ++i) {
                 if (i == 0) {
-                    kvm_irqchip_add_route(s, i, KVM_IRQCHIP_IOAPIC, 2);
+                    kvm_irqchip_add_irq_route(s, i, KVM_IRQCHIP_IOAPIC, 2);
                 } else if (i != 2) {
-                    kvm_irqchip_add_route(s, i, KVM_IRQCHIP_IOAPIC, i);
+                    kvm_irqchip_add_irq_route(s, i, KVM_IRQCHIP_IOAPIC, i);
                 }
             }
         }
diff --git a/kvm-all.c b/kvm-all.c
index 1913d6a..0117837 100644
--- a/kvm-all.c
+++ b/kvm-all.c
@@ -935,7 +935,7 @@ static void kvm_add_routing_entry(KVMState *s,
     set_gsi(s, entry->gsi);
 }
 
-void kvm_irqchip_add_route(KVMState *s, int irq, int irqchip, int pin)
+void kvm_irqchip_add_irq_route(KVMState *s, int irq, int irqchip, int pin)
 {
     struct kvm_irq_routing_entry e;
 
diff --git a/kvm.h b/kvm.h
index 7857dbf..8b061bd 100644
--- a/kvm.h
+++ b/kvm.h
@@ -134,7 +134,7 @@ void kvm_arch_init_irq_routing(KVMState *s);
 int kvm_irqchip_set_irq(KVMState *s, int irq, int level);
 int kvm_irqchip_send_msi(KVMState *s, MSIMessage msg);
 
-void kvm_irqchip_add_route(KVMState *s, int gsi, int irqchip, int pin);
+void kvm_irqchip_add_irq_route(KVMState *s, int gsi, int irqchip, int pin);
 int kvm_irqchip_commit_routes(KVMState *s);
 
 void kvm_put_apic_state(DeviceState *d, struct kvm_lapic_state *kapic);
-- 
1.7.10.1


^ permalink raw reply related	[flat|nested] 64+ messages in thread

* [Qemu-devel] [PATCH 12/19] kvm: Rename kvm_irqchip_add_route to kvm_irqchip_add_irq_route
@ 2012-05-21 16:37   ` Avi Kivity
  0 siblings, 0 replies; 64+ messages in thread
From: Avi Kivity @ 2012-05-21 16:37 UTC (permalink / raw)
  To: Anthony Liguori; +Cc: Jan Kiszka, Marcelo Tosatti, qemu-devel, kvm

From: Jan Kiszka <jan.kiszka@siemens.com>

We will add kvm_irqchip_add_msi_route, so let's make the difference
clearer.

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
---
 hw/pc_piix.c |    8 ++++----
 kvm-all.c    |    2 +-
 kvm.h        |    2 +-
 3 files changed, 6 insertions(+), 6 deletions(-)

diff --git a/hw/pc_piix.c b/hw/pc_piix.c
index 6a75718..c17f906 100644
--- a/hw/pc_piix.c
+++ b/hw/pc_piix.c
@@ -63,17 +63,17 @@ static void kvm_piix3_setup_irq_routing(bool pci_enabled)
             if (i == 2) {
                 continue;
             }
-            kvm_irqchip_add_route(s, i, KVM_IRQCHIP_PIC_MASTER, i);
+            kvm_irqchip_add_irq_route(s, i, KVM_IRQCHIP_PIC_MASTER, i);
         }
         for (i = 8; i < 16; ++i) {
-            kvm_irqchip_add_route(s, i, KVM_IRQCHIP_PIC_SLAVE, i - 8);
+            kvm_irqchip_add_irq_route(s, i, KVM_IRQCHIP_PIC_SLAVE, i - 8);
         }
         if (pci_enabled) {
             for (i = 0; i < 24; ++i) {
                 if (i == 0) {
-                    kvm_irqchip_add_route(s, i, KVM_IRQCHIP_IOAPIC, 2);
+                    kvm_irqchip_add_irq_route(s, i, KVM_IRQCHIP_IOAPIC, 2);
                 } else if (i != 2) {
-                    kvm_irqchip_add_route(s, i, KVM_IRQCHIP_IOAPIC, i);
+                    kvm_irqchip_add_irq_route(s, i, KVM_IRQCHIP_IOAPIC, i);
                 }
             }
         }
diff --git a/kvm-all.c b/kvm-all.c
index 1913d6a..0117837 100644
--- a/kvm-all.c
+++ b/kvm-all.c
@@ -935,7 +935,7 @@ static void kvm_add_routing_entry(KVMState *s,
     set_gsi(s, entry->gsi);
 }
 
-void kvm_irqchip_add_route(KVMState *s, int irq, int irqchip, int pin)
+void kvm_irqchip_add_irq_route(KVMState *s, int irq, int irqchip, int pin)
 {
     struct kvm_irq_routing_entry e;
 
diff --git a/kvm.h b/kvm.h
index 7857dbf..8b061bd 100644
--- a/kvm.h
+++ b/kvm.h
@@ -134,7 +134,7 @@ void kvm_arch_init_irq_routing(KVMState *s);
 int kvm_irqchip_set_irq(KVMState *s, int irq, int level);
 int kvm_irqchip_send_msi(KVMState *s, MSIMessage msg);
 
-void kvm_irqchip_add_route(KVMState *s, int gsi, int irqchip, int pin);
+void kvm_irqchip_add_irq_route(KVMState *s, int gsi, int irqchip, int pin);
 int kvm_irqchip_commit_routes(KVMState *s);
 
 void kvm_put_apic_state(DeviceState *d, struct kvm_lapic_state *kapic);
-- 
1.7.10.1

^ permalink raw reply related	[flat|nested] 64+ messages in thread

* [PATCH 13/19] kvm: Introduce kvm_irqchip_add_msi_route
  2012-05-21 16:37 ` [Qemu-devel] " Avi Kivity
@ 2012-05-21 16:37   ` Avi Kivity
  -1 siblings, 0 replies; 64+ messages in thread
From: Avi Kivity @ 2012-05-21 16:37 UTC (permalink / raw)
  To: Anthony Liguori; +Cc: qemu-devel, kvm, Jan Kiszka, Marcelo Tosatti

From: Jan Kiszka <jan.kiszka@siemens.com>

Add a service that establishes a static route from a virtual IRQ line to
an MSI message. Will be used for IRQFD and device assignment. As we will
use this service outside of CONFIG_KVM protected code, stub it properly.

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
---
 kvm-all.c  |   31 +++++++++++++++++++++++++++++++
 kvm-stub.c |    8 ++++++++
 kvm.h      |   10 ++++++----
 3 files changed, 45 insertions(+), 4 deletions(-)

diff --git a/kvm-all.c b/kvm-all.c
index 0117837..7f906ca 100644
--- a/kvm-all.c
+++ b/kvm-all.c
@@ -1080,6 +1080,32 @@ int kvm_irqchip_send_msi(KVMState *s, MSIMessage msg)
     return kvm_irqchip_set_irq(s, route->kroute.gsi, 1);
 }
 
+int kvm_irqchip_add_msi_route(KVMState *s, MSIMessage msg)
+{
+    struct kvm_irq_routing_entry kroute;
+    int virq;
+
+    if (!kvm_irqchip_in_kernel()) {
+        return -ENOSYS;
+    }
+
+    virq = kvm_irqchip_get_virq(s);
+    if (virq < 0) {
+        return virq;
+    }
+
+    kroute.gsi = virq;
+    kroute.type = KVM_IRQ_ROUTING_MSI;
+    kroute.flags = 0;
+    kroute.u.msi.address_lo = (uint32_t)msg.address;
+    kroute.u.msi.address_hi = msg.address >> 32;
+    kroute.u.msi.data = msg.data;
+
+    kvm_add_routing_entry(s, &kroute);
+
+    return virq;
+}
+
 #else /* !KVM_CAP_IRQ_ROUTING */
 
 static void kvm_init_irq_routing(KVMState *s)
@@ -1090,6 +1116,11 @@ int kvm_irqchip_send_msi(KVMState *s, MSIMessage msg)
 {
     abort();
 }
+
+int kvm_irqchip_add_msi_route(KVMState *s, MSIMessage msg)
+{
+    abort();
+}
 #endif /* !KVM_CAP_IRQ_ROUTING */
 
 static int kvm_irqchip_create(KVMState *s)
diff --git a/kvm-stub.c b/kvm-stub.c
index 47c573d..db3a7dc 100644
--- a/kvm-stub.c
+++ b/kvm-stub.c
@@ -12,10 +12,13 @@
 
 #include "qemu-common.h"
 #include "hw/hw.h"
+#include "hw/msi.h"
 #include "cpu.h"
 #include "gdbstub.h"
 #include "kvm.h"
 
+KVMState *kvm_state;
+
 int kvm_init_vcpu(CPUArchState *env)
 {
     return -ENOSYS;
@@ -128,3 +131,8 @@ int kvm_on_sigbus(int code, void *addr)
 {
     return 1;
 }
+
+int kvm_irqchip_add_msi_route(KVMState *s, MSIMessage msg)
+{
+    return -ENOSYS;
+}
diff --git a/kvm.h b/kvm.h
index 8b061bd..67df1f1 100644
--- a/kvm.h
+++ b/kvm.h
@@ -44,6 +44,10 @@ typedef struct KVMCapabilityInfo {
 #define KVM_CAP_INFO(CAP) { "KVM_CAP_" stringify(CAP), KVM_CAP_##CAP }
 #define KVM_CAP_LAST_INFO { NULL, 0 }
 
+struct KVMState;
+typedef struct KVMState KVMState;
+extern KVMState *kvm_state;
+
 /* external API */
 
 int kvm_init(void);
@@ -88,10 +92,6 @@ int kvm_on_sigbus(int code, void *addr);
 
 /* internal API */
 
-struct KVMState;
-typedef struct KVMState KVMState;
-extern KVMState *kvm_state;
-
 int kvm_ioctl(KVMState *s, int type, ...);
 
 int kvm_vm_ioctl(KVMState *s, int type, ...);
@@ -213,4 +213,6 @@ int kvm_set_ioeventfd_mmio(int fd, uint32_t adr, uint32_t val, bool assign,
                            uint32_t size);
 
 int kvm_set_ioeventfd_pio_word(int fd, uint16_t adr, uint16_t val, bool assign);
+
+int kvm_irqchip_add_msi_route(KVMState *s, MSIMessage msg);
 #endif
-- 
1.7.10.1


^ permalink raw reply related	[flat|nested] 64+ messages in thread

* [Qemu-devel] [PATCH 13/19] kvm: Introduce kvm_irqchip_add_msi_route
@ 2012-05-21 16:37   ` Avi Kivity
  0 siblings, 0 replies; 64+ messages in thread
From: Avi Kivity @ 2012-05-21 16:37 UTC (permalink / raw)
  To: Anthony Liguori; +Cc: Jan Kiszka, Marcelo Tosatti, qemu-devel, kvm

From: Jan Kiszka <jan.kiszka@siemens.com>

Add a service that establishes a static route from a virtual IRQ line to
an MSI message. Will be used for IRQFD and device assignment. As we will
use this service outside of CONFIG_KVM protected code, stub it properly.

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
---
 kvm-all.c  |   31 +++++++++++++++++++++++++++++++
 kvm-stub.c |    8 ++++++++
 kvm.h      |   10 ++++++----
 3 files changed, 45 insertions(+), 4 deletions(-)

diff --git a/kvm-all.c b/kvm-all.c
index 0117837..7f906ca 100644
--- a/kvm-all.c
+++ b/kvm-all.c
@@ -1080,6 +1080,32 @@ int kvm_irqchip_send_msi(KVMState *s, MSIMessage msg)
     return kvm_irqchip_set_irq(s, route->kroute.gsi, 1);
 }
 
+int kvm_irqchip_add_msi_route(KVMState *s, MSIMessage msg)
+{
+    struct kvm_irq_routing_entry kroute;
+    int virq;
+
+    if (!kvm_irqchip_in_kernel()) {
+        return -ENOSYS;
+    }
+
+    virq = kvm_irqchip_get_virq(s);
+    if (virq < 0) {
+        return virq;
+    }
+
+    kroute.gsi = virq;
+    kroute.type = KVM_IRQ_ROUTING_MSI;
+    kroute.flags = 0;
+    kroute.u.msi.address_lo = (uint32_t)msg.address;
+    kroute.u.msi.address_hi = msg.address >> 32;
+    kroute.u.msi.data = msg.data;
+
+    kvm_add_routing_entry(s, &kroute);
+
+    return virq;
+}
+
 #else /* !KVM_CAP_IRQ_ROUTING */
 
 static void kvm_init_irq_routing(KVMState *s)
@@ -1090,6 +1116,11 @@ int kvm_irqchip_send_msi(KVMState *s, MSIMessage msg)
 {
     abort();
 }
+
+int kvm_irqchip_add_msi_route(KVMState *s, MSIMessage msg)
+{
+    abort();
+}
 #endif /* !KVM_CAP_IRQ_ROUTING */
 
 static int kvm_irqchip_create(KVMState *s)
diff --git a/kvm-stub.c b/kvm-stub.c
index 47c573d..db3a7dc 100644
--- a/kvm-stub.c
+++ b/kvm-stub.c
@@ -12,10 +12,13 @@
 
 #include "qemu-common.h"
 #include "hw/hw.h"
+#include "hw/msi.h"
 #include "cpu.h"
 #include "gdbstub.h"
 #include "kvm.h"
 
+KVMState *kvm_state;
+
 int kvm_init_vcpu(CPUArchState *env)
 {
     return -ENOSYS;
@@ -128,3 +131,8 @@ int kvm_on_sigbus(int code, void *addr)
 {
     return 1;
 }
+
+int kvm_irqchip_add_msi_route(KVMState *s, MSIMessage msg)
+{
+    return -ENOSYS;
+}
diff --git a/kvm.h b/kvm.h
index 8b061bd..67df1f1 100644
--- a/kvm.h
+++ b/kvm.h
@@ -44,6 +44,10 @@ typedef struct KVMCapabilityInfo {
 #define KVM_CAP_INFO(CAP) { "KVM_CAP_" stringify(CAP), KVM_CAP_##CAP }
 #define KVM_CAP_LAST_INFO { NULL, 0 }
 
+struct KVMState;
+typedef struct KVMState KVMState;
+extern KVMState *kvm_state;
+
 /* external API */
 
 int kvm_init(void);
@@ -88,10 +92,6 @@ int kvm_on_sigbus(int code, void *addr);
 
 /* internal API */
 
-struct KVMState;
-typedef struct KVMState KVMState;
-extern KVMState *kvm_state;
-
 int kvm_ioctl(KVMState *s, int type, ...);
 
 int kvm_vm_ioctl(KVMState *s, int type, ...);
@@ -213,4 +213,6 @@ int kvm_set_ioeventfd_mmio(int fd, uint32_t adr, uint32_t val, bool assign,
                            uint32_t size);
 
 int kvm_set_ioeventfd_pio_word(int fd, uint16_t adr, uint16_t val, bool assign);
+
+int kvm_irqchip_add_msi_route(KVMState *s, MSIMessage msg);
 #endif
-- 
1.7.10.1

^ permalink raw reply related	[flat|nested] 64+ messages in thread

* [PATCH 14/19] kvm: Publicize kvm_irqchip_release_virq
  2012-05-21 16:37 ` [Qemu-devel] " Avi Kivity
@ 2012-05-21 16:37   ` Avi Kivity
  -1 siblings, 0 replies; 64+ messages in thread
From: Avi Kivity @ 2012-05-21 16:37 UTC (permalink / raw)
  To: Anthony Liguori; +Cc: qemu-devel, kvm, Jan Kiszka, Marcelo Tosatti

From: Jan Kiszka <jan.kiszka@siemens.com>

This allows to drop routes created by kvm_irqchip_add_irq/msi_route
again.

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
---
 kvm-all.c  |    2 +-
 kvm-stub.c |    4 ++++
 kvm.h      |    1 +
 3 files changed, 6 insertions(+), 1 deletion(-)

diff --git a/kvm-all.c b/kvm-all.c
index 7f906ca..ca6cec6 100644
--- a/kvm-all.c
+++ b/kvm-all.c
@@ -955,7 +955,7 @@ int kvm_irqchip_commit_routes(KVMState *s)
     return kvm_vm_ioctl(s, KVM_SET_GSI_ROUTING, s->irq_routes);
 }
 
-static void kvm_irqchip_release_virq(KVMState *s, int virq)
+void kvm_irqchip_release_virq(KVMState *s, int virq)
 {
     struct kvm_irq_routing_entry *e;
     int i;
diff --git a/kvm-stub.c b/kvm-stub.c
index db3a7dc..ec351d9 100644
--- a/kvm-stub.c
+++ b/kvm-stub.c
@@ -136,3 +136,7 @@ int kvm_irqchip_add_msi_route(KVMState *s, MSIMessage msg)
 {
     return -ENOSYS;
 }
+
+void kvm_irqchip_release_virq(KVMState *s, int virq)
+{
+}
diff --git a/kvm.h b/kvm.h
index 67df1f1..1779e73 100644
--- a/kvm.h
+++ b/kvm.h
@@ -215,4 +215,5 @@ int kvm_set_ioeventfd_mmio(int fd, uint32_t adr, uint32_t val, bool assign,
 int kvm_set_ioeventfd_pio_word(int fd, uint16_t adr, uint16_t val, bool assign);
 
 int kvm_irqchip_add_msi_route(KVMState *s, MSIMessage msg);
+void kvm_irqchip_release_virq(KVMState *s, int virq);
 #endif
-- 
1.7.10.1


^ permalink raw reply related	[flat|nested] 64+ messages in thread

* [Qemu-devel] [PATCH 14/19] kvm: Publicize kvm_irqchip_release_virq
@ 2012-05-21 16:37   ` Avi Kivity
  0 siblings, 0 replies; 64+ messages in thread
From: Avi Kivity @ 2012-05-21 16:37 UTC (permalink / raw)
  To: Anthony Liguori; +Cc: Jan Kiszka, Marcelo Tosatti, qemu-devel, kvm

From: Jan Kiszka <jan.kiszka@siemens.com>

This allows to drop routes created by kvm_irqchip_add_irq/msi_route
again.

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
---
 kvm-all.c  |    2 +-
 kvm-stub.c |    4 ++++
 kvm.h      |    1 +
 3 files changed, 6 insertions(+), 1 deletion(-)

diff --git a/kvm-all.c b/kvm-all.c
index 7f906ca..ca6cec6 100644
--- a/kvm-all.c
+++ b/kvm-all.c
@@ -955,7 +955,7 @@ int kvm_irqchip_commit_routes(KVMState *s)
     return kvm_vm_ioctl(s, KVM_SET_GSI_ROUTING, s->irq_routes);
 }
 
-static void kvm_irqchip_release_virq(KVMState *s, int virq)
+void kvm_irqchip_release_virq(KVMState *s, int virq)
 {
     struct kvm_irq_routing_entry *e;
     int i;
diff --git a/kvm-stub.c b/kvm-stub.c
index db3a7dc..ec351d9 100644
--- a/kvm-stub.c
+++ b/kvm-stub.c
@@ -136,3 +136,7 @@ int kvm_irqchip_add_msi_route(KVMState *s, MSIMessage msg)
 {
     return -ENOSYS;
 }
+
+void kvm_irqchip_release_virq(KVMState *s, int virq)
+{
+}
diff --git a/kvm.h b/kvm.h
index 67df1f1..1779e73 100644
--- a/kvm.h
+++ b/kvm.h
@@ -215,4 +215,5 @@ int kvm_set_ioeventfd_mmio(int fd, uint32_t adr, uint32_t val, bool assign,
 int kvm_set_ioeventfd_pio_word(int fd, uint16_t adr, uint16_t val, bool assign);
 
 int kvm_irqchip_add_msi_route(KVMState *s, MSIMessage msg);
+void kvm_irqchip_release_virq(KVMState *s, int virq);
 #endif
-- 
1.7.10.1

^ permalink raw reply related	[flat|nested] 64+ messages in thread

* [PATCH 15/19] kvm: Make kvm_irqchip_commit_routes an internal service
  2012-05-21 16:37 ` [Qemu-devel] " Avi Kivity
@ 2012-05-21 16:37   ` Avi Kivity
  -1 siblings, 0 replies; 64+ messages in thread
From: Avi Kivity @ 2012-05-21 16:37 UTC (permalink / raw)
  To: Anthony Liguori; +Cc: qemu-devel, kvm, Jan Kiszka, Marcelo Tosatti

From: Jan Kiszka <jan.kiszka@siemens.com>

Automatically commit route changes after kvm_add_routing_entry and
kvm_irqchip_release_virq. There is no performance relevant use case for
which collecting multiple route changes is beneficial. This makes
kvm_irqchip_commit_routes an internal service which assert()s that the
corresponding IOCTL will always succeed.

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
---
 hw/pc_piix.c |    6 +-----
 kvm-all.c    |   26 ++++++++++++++------------
 kvm.h        |    1 -
 3 files changed, 15 insertions(+), 18 deletions(-)

diff --git a/hw/pc_piix.c b/hw/pc_piix.c
index c17f906..044dfcb 100644
--- a/hw/pc_piix.c
+++ b/hw/pc_piix.c
@@ -56,7 +56,7 @@ static void kvm_piix3_setup_irq_routing(bool pci_enabled)
 {
 #ifdef CONFIG_KVM
     KVMState *s = kvm_state;
-    int ret, i;
+    int i;
 
     if (kvm_check_extension(s, KVM_CAP_IRQ_ROUTING)) {
         for (i = 0; i < 8; ++i) {
@@ -77,10 +77,6 @@ static void kvm_piix3_setup_irq_routing(bool pci_enabled)
                 }
             }
         }
-        ret = kvm_irqchip_commit_routes(s);
-        if (ret < 0) {
-            hw_error("KVM IRQ routing setup failed");
-        }
     }
 #endif /* CONFIG_KVM */
 }
diff --git a/kvm-all.c b/kvm-all.c
index ca6cec6..e96f092 100644
--- a/kvm-all.c
+++ b/kvm-all.c
@@ -908,6 +908,15 @@ static void kvm_init_irq_routing(KVMState *s)
     kvm_arch_init_irq_routing(s);
 }
 
+static void kvm_irqchip_commit_routes(KVMState *s)
+{
+    int ret;
+
+    s->irq_routes->flags = 0;
+    ret = kvm_vm_ioctl(s, KVM_SET_GSI_ROUTING, s->irq_routes);
+    assert(ret == 0);
+}
+
 static void kvm_add_routing_entry(KVMState *s,
                                   struct kvm_irq_routing_entry *entry)
 {
@@ -933,6 +942,8 @@ static void kvm_add_routing_entry(KVMState *s,
     new->u = entry->u;
 
     set_gsi(s, entry->gsi);
+
+    kvm_irqchip_commit_routes(s);
 }
 
 void kvm_irqchip_add_irq_route(KVMState *s, int irq, int irqchip, int pin)
@@ -949,12 +960,6 @@ void kvm_irqchip_add_irq_route(KVMState *s, int irq, int irqchip, int pin)
     kvm_add_routing_entry(s, &e);
 }
 
-int kvm_irqchip_commit_routes(KVMState *s)
-{
-    s->irq_routes->flags = 0;
-    return kvm_vm_ioctl(s, KVM_SET_GSI_ROUTING, s->irq_routes);
-}
-
 void kvm_irqchip_release_virq(KVMState *s, int virq)
 {
     struct kvm_irq_routing_entry *e;
@@ -968,6 +973,8 @@ void kvm_irqchip_release_virq(KVMState *s, int virq)
         }
     }
     clear_gsi(s, virq);
+
+    kvm_irqchip_commit_routes(s);
 }
 
 static unsigned int kvm_hash_msi(uint32_t data)
@@ -1049,7 +1056,7 @@ int kvm_irqchip_send_msi(KVMState *s, MSIMessage msg)
 
     route = kvm_lookup_msi_route(s, msg);
     if (!route) {
-        int virq, ret;
+        int virq;
 
         virq = kvm_irqchip_get_virq(s);
         if (virq < 0) {
@@ -1068,11 +1075,6 @@ int kvm_irqchip_send_msi(KVMState *s, MSIMessage msg)
 
         QTAILQ_INSERT_TAIL(&s->msi_hashtab[kvm_hash_msi(msg.data)], route,
                            entry);
-
-        ret = kvm_irqchip_commit_routes(s);
-        if (ret < 0) {
-            return ret;
-        }
     }
 
     assert(route->kroute.type == KVM_IRQ_ROUTING_MSI);
diff --git a/kvm.h b/kvm.h
index 1779e73..f0d0c53 100644
--- a/kvm.h
+++ b/kvm.h
@@ -135,7 +135,6 @@ int kvm_irqchip_set_irq(KVMState *s, int irq, int level);
 int kvm_irqchip_send_msi(KVMState *s, MSIMessage msg);
 
 void kvm_irqchip_add_irq_route(KVMState *s, int gsi, int irqchip, int pin);
-int kvm_irqchip_commit_routes(KVMState *s);
 
 void kvm_put_apic_state(DeviceState *d, struct kvm_lapic_state *kapic);
 void kvm_get_apic_state(DeviceState *d, struct kvm_lapic_state *kapic);
-- 
1.7.10.1


^ permalink raw reply related	[flat|nested] 64+ messages in thread

* [Qemu-devel] [PATCH 15/19] kvm: Make kvm_irqchip_commit_routes an internal service
@ 2012-05-21 16:37   ` Avi Kivity
  0 siblings, 0 replies; 64+ messages in thread
From: Avi Kivity @ 2012-05-21 16:37 UTC (permalink / raw)
  To: Anthony Liguori; +Cc: Jan Kiszka, Marcelo Tosatti, qemu-devel, kvm

From: Jan Kiszka <jan.kiszka@siemens.com>

Automatically commit route changes after kvm_add_routing_entry and
kvm_irqchip_release_virq. There is no performance relevant use case for
which collecting multiple route changes is beneficial. This makes
kvm_irqchip_commit_routes an internal service which assert()s that the
corresponding IOCTL will always succeed.

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
---
 hw/pc_piix.c |    6 +-----
 kvm-all.c    |   26 ++++++++++++++------------
 kvm.h        |    1 -
 3 files changed, 15 insertions(+), 18 deletions(-)

diff --git a/hw/pc_piix.c b/hw/pc_piix.c
index c17f906..044dfcb 100644
--- a/hw/pc_piix.c
+++ b/hw/pc_piix.c
@@ -56,7 +56,7 @@ static void kvm_piix3_setup_irq_routing(bool pci_enabled)
 {
 #ifdef CONFIG_KVM
     KVMState *s = kvm_state;
-    int ret, i;
+    int i;
 
     if (kvm_check_extension(s, KVM_CAP_IRQ_ROUTING)) {
         for (i = 0; i < 8; ++i) {
@@ -77,10 +77,6 @@ static void kvm_piix3_setup_irq_routing(bool pci_enabled)
                 }
             }
         }
-        ret = kvm_irqchip_commit_routes(s);
-        if (ret < 0) {
-            hw_error("KVM IRQ routing setup failed");
-        }
     }
 #endif /* CONFIG_KVM */
 }
diff --git a/kvm-all.c b/kvm-all.c
index ca6cec6..e96f092 100644
--- a/kvm-all.c
+++ b/kvm-all.c
@@ -908,6 +908,15 @@ static void kvm_init_irq_routing(KVMState *s)
     kvm_arch_init_irq_routing(s);
 }
 
+static void kvm_irqchip_commit_routes(KVMState *s)
+{
+    int ret;
+
+    s->irq_routes->flags = 0;
+    ret = kvm_vm_ioctl(s, KVM_SET_GSI_ROUTING, s->irq_routes);
+    assert(ret == 0);
+}
+
 static void kvm_add_routing_entry(KVMState *s,
                                   struct kvm_irq_routing_entry *entry)
 {
@@ -933,6 +942,8 @@ static void kvm_add_routing_entry(KVMState *s,
     new->u = entry->u;
 
     set_gsi(s, entry->gsi);
+
+    kvm_irqchip_commit_routes(s);
 }
 
 void kvm_irqchip_add_irq_route(KVMState *s, int irq, int irqchip, int pin)
@@ -949,12 +960,6 @@ void kvm_irqchip_add_irq_route(KVMState *s, int irq, int irqchip, int pin)
     kvm_add_routing_entry(s, &e);
 }
 
-int kvm_irqchip_commit_routes(KVMState *s)
-{
-    s->irq_routes->flags = 0;
-    return kvm_vm_ioctl(s, KVM_SET_GSI_ROUTING, s->irq_routes);
-}
-
 void kvm_irqchip_release_virq(KVMState *s, int virq)
 {
     struct kvm_irq_routing_entry *e;
@@ -968,6 +973,8 @@ void kvm_irqchip_release_virq(KVMState *s, int virq)
         }
     }
     clear_gsi(s, virq);
+
+    kvm_irqchip_commit_routes(s);
 }
 
 static unsigned int kvm_hash_msi(uint32_t data)
@@ -1049,7 +1056,7 @@ int kvm_irqchip_send_msi(KVMState *s, MSIMessage msg)
 
     route = kvm_lookup_msi_route(s, msg);
     if (!route) {
-        int virq, ret;
+        int virq;
 
         virq = kvm_irqchip_get_virq(s);
         if (virq < 0) {
@@ -1068,11 +1075,6 @@ int kvm_irqchip_send_msi(KVMState *s, MSIMessage msg)
 
         QTAILQ_INSERT_TAIL(&s->msi_hashtab[kvm_hash_msi(msg.data)], route,
                            entry);
-
-        ret = kvm_irqchip_commit_routes(s);
-        if (ret < 0) {
-            return ret;
-        }
     }
 
     assert(route->kroute.type == KVM_IRQ_ROUTING_MSI);
diff --git a/kvm.h b/kvm.h
index 1779e73..f0d0c53 100644
--- a/kvm.h
+++ b/kvm.h
@@ -135,7 +135,6 @@ int kvm_irqchip_set_irq(KVMState *s, int irq, int level);
 int kvm_irqchip_send_msi(KVMState *s, MSIMessage msg);
 
 void kvm_irqchip_add_irq_route(KVMState *s, int gsi, int irqchip, int pin);
-int kvm_irqchip_commit_routes(KVMState *s);
 
 void kvm_put_apic_state(DeviceState *d, struct kvm_lapic_state *kapic);
 void kvm_get_apic_state(DeviceState *d, struct kvm_lapic_state *kapic);
-- 
1.7.10.1

^ permalink raw reply related	[flat|nested] 64+ messages in thread

* [PATCH 16/19] kvm: Introduce kvm_irqchip_add/remove_irqfd
  2012-05-21 16:37 ` [Qemu-devel] " Avi Kivity
@ 2012-05-21 16:37   ` Avi Kivity
  -1 siblings, 0 replies; 64+ messages in thread
From: Avi Kivity @ 2012-05-21 16:37 UTC (permalink / raw)
  To: Anthony Liguori; +Cc: qemu-devel, kvm, Jan Kiszka, Marcelo Tosatti

From: Jan Kiszka <jan.kiszka@siemens.com>

Add services to associate an eventfd file descriptor as input with an
IRQ line as output. Such a line can be an input pin of an in-kernel
irqchip or a virtual line returned by kvm_irqchip_add_route.

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
---
 kvm-all.c  |   30 ++++++++++++++++++++++++++++++
 kvm-stub.c |   10 ++++++++++
 kvm.h      |    3 +++
 3 files changed, 43 insertions(+)

diff --git a/kvm-all.c b/kvm-all.c
index e96f092..489ee53 100644
--- a/kvm-all.c
+++ b/kvm-all.c
@@ -1108,6 +1108,21 @@ int kvm_irqchip_add_msi_route(KVMState *s, MSIMessage msg)
     return virq;
 }
 
+static int kvm_irqchip_assign_irqfd(KVMState *s, int fd, int virq, bool assign)
+{
+    struct kvm_irqfd irqfd = {
+        .fd = fd,
+        .gsi = virq,
+        .flags = assign ? 0 : KVM_IRQFD_FLAG_DEASSIGN,
+    };
+
+    if (!kvm_irqchip_in_kernel()) {
+        return -ENOSYS;
+    }
+
+    return kvm_vm_ioctl(s, KVM_IRQFD, &irqfd);
+}
+
 #else /* !KVM_CAP_IRQ_ROUTING */
 
 static void kvm_init_irq_routing(KVMState *s)
@@ -1123,8 +1138,23 @@ int kvm_irqchip_add_msi_route(KVMState *s, MSIMessage msg)
 {
     abort();
 }
+
+static int kvm_irqchip_assign_irqfd(KVMState *s, int fd, int virq, bool assign)
+{
+    abort();
+}
 #endif /* !KVM_CAP_IRQ_ROUTING */
 
+int kvm_irqchip_add_irqfd(KVMState *s, int fd, int virq)
+{
+    return kvm_irqchip_assign_irqfd(s, fd, virq, true);
+}
+
+int kvm_irqchip_remove_irqfd(KVMState *s, int fd, int virq)
+{
+    return kvm_irqchip_assign_irqfd(s, fd, virq, false);
+}
+
 static int kvm_irqchip_create(KVMState *s)
 {
     QemuOptsList *list = qemu_find_opts("machine");
diff --git a/kvm-stub.c b/kvm-stub.c
index ec351d9..b4cf03f 100644
--- a/kvm-stub.c
+++ b/kvm-stub.c
@@ -140,3 +140,13 @@ int kvm_irqchip_add_msi_route(KVMState *s, MSIMessage msg)
 void kvm_irqchip_release_virq(KVMState *s, int virq)
 {
 }
+
+int kvm_irqchip_add_irqfd(KVMState *s, int fd, int virq)
+{
+    return -ENOSYS;
+}
+
+int kvm_irqchip_remove_irqfd(KVMState *s, int fd, int virq)
+{
+    return -ENOSYS;
+}
diff --git a/kvm.h b/kvm.h
index f0d0c53..9c7b0ea 100644
--- a/kvm.h
+++ b/kvm.h
@@ -215,4 +215,7 @@ int kvm_set_ioeventfd_pio_word(int fd, uint16_t adr, uint16_t val, bool assign);
 
 int kvm_irqchip_add_msi_route(KVMState *s, MSIMessage msg);
 void kvm_irqchip_release_virq(KVMState *s, int virq);
+
+int kvm_irqchip_add_irqfd(KVMState *s, int fd, int virq);
+int kvm_irqchip_remove_irqfd(KVMState *s, int fd, int virq);
 #endif
-- 
1.7.10.1


^ permalink raw reply related	[flat|nested] 64+ messages in thread

* [Qemu-devel] [PATCH 16/19] kvm: Introduce kvm_irqchip_add/remove_irqfd
@ 2012-05-21 16:37   ` Avi Kivity
  0 siblings, 0 replies; 64+ messages in thread
From: Avi Kivity @ 2012-05-21 16:37 UTC (permalink / raw)
  To: Anthony Liguori; +Cc: Jan Kiszka, Marcelo Tosatti, qemu-devel, kvm

From: Jan Kiszka <jan.kiszka@siemens.com>

Add services to associate an eventfd file descriptor as input with an
IRQ line as output. Such a line can be an input pin of an in-kernel
irqchip or a virtual line returned by kvm_irqchip_add_route.

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
---
 kvm-all.c  |   30 ++++++++++++++++++++++++++++++
 kvm-stub.c |   10 ++++++++++
 kvm.h      |    3 +++
 3 files changed, 43 insertions(+)

diff --git a/kvm-all.c b/kvm-all.c
index e96f092..489ee53 100644
--- a/kvm-all.c
+++ b/kvm-all.c
@@ -1108,6 +1108,21 @@ int kvm_irqchip_add_msi_route(KVMState *s, MSIMessage msg)
     return virq;
 }
 
+static int kvm_irqchip_assign_irqfd(KVMState *s, int fd, int virq, bool assign)
+{
+    struct kvm_irqfd irqfd = {
+        .fd = fd,
+        .gsi = virq,
+        .flags = assign ? 0 : KVM_IRQFD_FLAG_DEASSIGN,
+    };
+
+    if (!kvm_irqchip_in_kernel()) {
+        return -ENOSYS;
+    }
+
+    return kvm_vm_ioctl(s, KVM_IRQFD, &irqfd);
+}
+
 #else /* !KVM_CAP_IRQ_ROUTING */
 
 static void kvm_init_irq_routing(KVMState *s)
@@ -1123,8 +1138,23 @@ int kvm_irqchip_add_msi_route(KVMState *s, MSIMessage msg)
 {
     abort();
 }
+
+static int kvm_irqchip_assign_irqfd(KVMState *s, int fd, int virq, bool assign)
+{
+    abort();
+}
 #endif /* !KVM_CAP_IRQ_ROUTING */
 
+int kvm_irqchip_add_irqfd(KVMState *s, int fd, int virq)
+{
+    return kvm_irqchip_assign_irqfd(s, fd, virq, true);
+}
+
+int kvm_irqchip_remove_irqfd(KVMState *s, int fd, int virq)
+{
+    return kvm_irqchip_assign_irqfd(s, fd, virq, false);
+}
+
 static int kvm_irqchip_create(KVMState *s)
 {
     QemuOptsList *list = qemu_find_opts("machine");
diff --git a/kvm-stub.c b/kvm-stub.c
index ec351d9..b4cf03f 100644
--- a/kvm-stub.c
+++ b/kvm-stub.c
@@ -140,3 +140,13 @@ int kvm_irqchip_add_msi_route(KVMState *s, MSIMessage msg)
 void kvm_irqchip_release_virq(KVMState *s, int virq)
 {
 }
+
+int kvm_irqchip_add_irqfd(KVMState *s, int fd, int virq)
+{
+    return -ENOSYS;
+}
+
+int kvm_irqchip_remove_irqfd(KVMState *s, int fd, int virq)
+{
+    return -ENOSYS;
+}
diff --git a/kvm.h b/kvm.h
index f0d0c53..9c7b0ea 100644
--- a/kvm.h
+++ b/kvm.h
@@ -215,4 +215,7 @@ int kvm_set_ioeventfd_pio_word(int fd, uint16_t adr, uint16_t val, bool assign);
 
 int kvm_irqchip_add_msi_route(KVMState *s, MSIMessage msg);
 void kvm_irqchip_release_virq(KVMState *s, int virq);
+
+int kvm_irqchip_add_irqfd(KVMState *s, int fd, int virq);
+int kvm_irqchip_remove_irqfd(KVMState *s, int fd, int virq);
 #endif
-- 
1.7.10.1

^ permalink raw reply related	[flat|nested] 64+ messages in thread

* [PATCH 17/19] kvm: Enable use of kvm_irqchip_in_kernel in hwlib code
  2012-05-21 16:37 ` [Qemu-devel] " Avi Kivity
@ 2012-05-21 16:37   ` Avi Kivity
  -1 siblings, 0 replies; 64+ messages in thread
From: Avi Kivity @ 2012-05-21 16:37 UTC (permalink / raw)
  To: Anthony Liguori; +Cc: Jan Kiszka, Marcelo Tosatti, qemu-devel, kvm

From: Jan Kiszka <jan.kiszka@siemens.com>

Provide a dummy kvm_kernel_irqchip so that kvm_irqchip_in_kernel can be
used by code that is not under CONFIG_KVM protection.

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
---
 kvm-stub.c |    1 +
 1 file changed, 1 insertion(+)

diff --git a/kvm-stub.c b/kvm-stub.c
index b4cf03f..ec9a364 100644
--- a/kvm-stub.c
+++ b/kvm-stub.c
@@ -18,6 +18,7 @@
 #include "kvm.h"
 
 KVMState *kvm_state;
+bool kvm_kernel_irqchip;
 
 int kvm_init_vcpu(CPUArchState *env)
 {
-- 
1.7.10.1

^ permalink raw reply related	[flat|nested] 64+ messages in thread

* [Qemu-devel] [PATCH 17/19] kvm: Enable use of kvm_irqchip_in_kernel in hwlib code
@ 2012-05-21 16:37   ` Avi Kivity
  0 siblings, 0 replies; 64+ messages in thread
From: Avi Kivity @ 2012-05-21 16:37 UTC (permalink / raw)
  To: Anthony Liguori; +Cc: Jan Kiszka, Marcelo Tosatti, qemu-devel, kvm

From: Jan Kiszka <jan.kiszka@siemens.com>

Provide a dummy kvm_kernel_irqchip so that kvm_irqchip_in_kernel can be
used by code that is not under CONFIG_KVM protection.

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
---
 kvm-stub.c |    1 +
 1 file changed, 1 insertion(+)

diff --git a/kvm-stub.c b/kvm-stub.c
index b4cf03f..ec9a364 100644
--- a/kvm-stub.c
+++ b/kvm-stub.c
@@ -18,6 +18,7 @@
 #include "kvm.h"
 
 KVMState *kvm_state;
+bool kvm_kernel_irqchip;
 
 int kvm_init_vcpu(CPUArchState *env)
 {
-- 
1.7.10.1

^ permalink raw reply related	[flat|nested] 64+ messages in thread

* [PATCH 18/19] msix: Add msix_nr_vectors_allocated
  2012-05-21 16:37 ` [Qemu-devel] " Avi Kivity
@ 2012-05-21 16:37   ` Avi Kivity
  -1 siblings, 0 replies; 64+ messages in thread
From: Avi Kivity @ 2012-05-21 16:37 UTC (permalink / raw)
  To: Anthony Liguori; +Cc: qemu-devel, kvm, Jan Kiszka, Marcelo Tosatti

From: Jan Kiszka <jan.kiszka@siemens.com>

Analogously to msi_nr_vectors_allocated, add a service for MSI-X. Will
be used by the virtio-pci layer.

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
---
 hw/msix.c |    5 +++++
 hw/msix.h |    2 ++
 2 files changed, 7 insertions(+)

diff --git a/hw/msix.c b/hw/msix.c
index 1622e16..59c7a83 100644
--- a/hw/msix.c
+++ b/hw/msix.c
@@ -452,6 +452,11 @@ void msix_unuse_all_vectors(PCIDevice *dev)
     msix_free_irq_entries(dev);
 }
 
+unsigned int msix_nr_vectors_allocated(const PCIDevice *dev)
+{
+    return dev->msix_entries_nr;
+}
+
 static int msix_set_notifier_for_vector(PCIDevice *dev, unsigned int vector)
 {
     MSIMessage msg;
diff --git a/hw/msix.h b/hw/msix.h
index f33f18b..50aee82 100644
--- a/hw/msix.h
+++ b/hw/msix.h
@@ -13,6 +13,8 @@ void msix_write_config(PCIDevice *pci_dev, uint32_t address,
 
 int msix_uninit(PCIDevice *d, MemoryRegion *bar);
 
+unsigned int msix_nr_vectors_allocated(const PCIDevice *dev);
+
 void msix_save(PCIDevice *dev, QEMUFile *f);
 void msix_load(PCIDevice *dev, QEMUFile *f);
 
-- 
1.7.10.1


^ permalink raw reply related	[flat|nested] 64+ messages in thread

* [Qemu-devel] [PATCH 18/19] msix: Add msix_nr_vectors_allocated
@ 2012-05-21 16:37   ` Avi Kivity
  0 siblings, 0 replies; 64+ messages in thread
From: Avi Kivity @ 2012-05-21 16:37 UTC (permalink / raw)
  To: Anthony Liguori; +Cc: Jan Kiszka, Marcelo Tosatti, qemu-devel, kvm

From: Jan Kiszka <jan.kiszka@siemens.com>

Analogously to msi_nr_vectors_allocated, add a service for MSI-X. Will
be used by the virtio-pci layer.

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
---
 hw/msix.c |    5 +++++
 hw/msix.h |    2 ++
 2 files changed, 7 insertions(+)

diff --git a/hw/msix.c b/hw/msix.c
index 1622e16..59c7a83 100644
--- a/hw/msix.c
+++ b/hw/msix.c
@@ -452,6 +452,11 @@ void msix_unuse_all_vectors(PCIDevice *dev)
     msix_free_irq_entries(dev);
 }
 
+unsigned int msix_nr_vectors_allocated(const PCIDevice *dev)
+{
+    return dev->msix_entries_nr;
+}
+
 static int msix_set_notifier_for_vector(PCIDevice *dev, unsigned int vector)
 {
     MSIMessage msg;
diff --git a/hw/msix.h b/hw/msix.h
index f33f18b..50aee82 100644
--- a/hw/msix.h
+++ b/hw/msix.h
@@ -13,6 +13,8 @@ void msix_write_config(PCIDevice *pci_dev, uint32_t address,
 
 int msix_uninit(PCIDevice *d, MemoryRegion *bar);
 
+unsigned int msix_nr_vectors_allocated(const PCIDevice *dev);
+
 void msix_save(PCIDevice *dev, QEMUFile *f);
 void msix_load(PCIDevice *dev, QEMUFile *f);
 
-- 
1.7.10.1

^ permalink raw reply related	[flat|nested] 64+ messages in thread

* [PATCH 19/19] virtio/vhost: Add support for KVM in-kernel MSI injection
  2012-05-21 16:37 ` [Qemu-devel] " Avi Kivity
@ 2012-05-21 16:37   ` Avi Kivity
  -1 siblings, 0 replies; 64+ messages in thread
From: Avi Kivity @ 2012-05-21 16:37 UTC (permalink / raw)
  To: Anthony Liguori; +Cc: qemu-devel, kvm, Jan Kiszka, Marcelo Tosatti

From: Jan Kiszka <jan.kiszka@siemens.com>

Make use of the new vector notifier to track changes of the MSI-X
configuration of virtio PCI devices. On enabling events, we establish
the required virtual IRQ to MSI-X message route and link the signaling
eventfd file descriptor to this vIRQ line. That way, vhost-generated
interrupts can be directly delivered to an in-kernel MSI-X consumer like
the x86 APIC.

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
---
 hw/virtio-pci.c |  126 +++++++++++++++++++++++++++++++++++++++++++++++++++++++
 hw/virtio-pci.h |    6 +++
 2 files changed, 132 insertions(+)

diff --git a/hw/virtio-pci.c b/hw/virtio-pci.c
index 4a4413d..01f5b92 100644
--- a/hw/virtio-pci.c
+++ b/hw/virtio-pci.c
@@ -24,6 +24,7 @@
 #include "virtio-scsi.h"
 #include "pci.h"
 #include "qemu-error.h"
+#include "msi.h"
 #include "msix.h"
 #include "net.h"
 #include "loader.h"
@@ -539,6 +540,107 @@ static void virtio_pci_guest_notifier_read(void *opaque)
     }
 }
 
+static int kvm_virtio_pci_vq_vector_use(VirtIOPCIProxy *proxy,
+                                        unsigned int queue_no,
+                                        unsigned int vector,
+                                        MSIMessage msg)
+{
+    VirtQueue *vq = virtio_get_queue(proxy->vdev, queue_no);
+    VirtIOIRQFD *irqfd = &proxy->vector_irqfd[vector];
+    int fd, ret;
+
+    fd = event_notifier_get_fd(virtio_queue_get_guest_notifier(vq));
+
+    if (irqfd->users == 0) {
+        ret = kvm_irqchip_add_msi_route(kvm_state, msg);
+        if (ret < 0) {
+            return ret;
+        }
+        irqfd->virq = ret;
+    }
+    irqfd->users++;
+
+    ret = kvm_irqchip_add_irqfd(kvm_state, fd, irqfd->virq);
+    if (ret < 0) {
+        if (--irqfd->users == 0) {
+            kvm_irqchip_release_virq(kvm_state, irqfd->virq);
+        }
+        return ret;
+    }
+
+    qemu_set_fd_handler(fd, NULL, NULL, NULL);
+
+    return 0;
+}
+
+static void kvm_virtio_pci_vq_vector_release(VirtIOPCIProxy *proxy,
+                                             unsigned int queue_no,
+                                             unsigned int vector)
+{
+    VirtQueue *vq = virtio_get_queue(proxy->vdev, queue_no);
+    VirtIOIRQFD *irqfd = &proxy->vector_irqfd[vector];
+    int fd, ret;
+
+    fd = event_notifier_get_fd(virtio_queue_get_guest_notifier(vq));
+
+    ret = kvm_irqchip_remove_irqfd(kvm_state, fd, irqfd->virq);
+    assert(ret == 0);
+
+    if (--irqfd->users == 0) {
+        kvm_irqchip_release_virq(kvm_state, irqfd->virq);
+    }
+
+    qemu_set_fd_handler(fd, virtio_pci_guest_notifier_read, NULL, vq);
+}
+
+static int kvm_virtio_pci_vector_use(PCIDevice *dev, unsigned vector,
+                                     MSIMessage msg)
+{
+    VirtIOPCIProxy *proxy = container_of(dev, VirtIOPCIProxy, pci_dev);
+    VirtIODevice *vdev = proxy->vdev;
+    int ret, queue_no;
+
+    for (queue_no = 0; queue_no < VIRTIO_PCI_QUEUE_MAX; queue_no++) {
+        if (!virtio_queue_get_num(vdev, queue_no)) {
+            break;
+        }
+        if (virtio_queue_vector(vdev, queue_no) != vector) {
+            continue;
+        }
+        ret = kvm_virtio_pci_vq_vector_use(proxy, queue_no, vector, msg);
+        if (ret < 0) {
+            goto undo;
+        }
+    }
+    return 0;
+
+undo:
+    while (--queue_no >= 0) {
+        if (virtio_queue_vector(vdev, queue_no) != vector) {
+            continue;
+        }
+        kvm_virtio_pci_vq_vector_release(proxy, queue_no, vector);
+    }
+    return ret;
+}
+
+static void kvm_virtio_pci_vector_release(PCIDevice *dev, unsigned vector)
+{
+    VirtIOPCIProxy *proxy = container_of(dev, VirtIOPCIProxy, pci_dev);
+    VirtIODevice *vdev = proxy->vdev;
+    int queue_no;
+
+    for (queue_no = 0; queue_no < VIRTIO_PCI_QUEUE_MAX; queue_no++) {
+        if (!virtio_queue_get_num(vdev, queue_no)) {
+            break;
+        }
+        if (virtio_queue_vector(vdev, queue_no) != vector) {
+            continue;
+        }
+        kvm_virtio_pci_vq_vector_release(proxy, queue_no, vector);
+    }
+}
+
 static int virtio_pci_set_guest_notifier(void *opaque, int n, bool assign)
 {
     VirtIOPCIProxy *proxy = opaque;
@@ -555,6 +657,9 @@ static int virtio_pci_set_guest_notifier(void *opaque, int n, bool assign)
     } else {
         qemu_set_fd_handler(event_notifier_get_fd(notifier),
                             NULL, NULL, NULL);
+        /* Test and clear notifier before closing it,
+         * in case poll callback didn't have time to run. */
+        virtio_pci_guest_notifier_read(vq);
         event_notifier_cleanup(notifier);
     }
 
@@ -573,6 +678,13 @@ static int virtio_pci_set_guest_notifiers(void *opaque, bool assign)
     VirtIODevice *vdev = proxy->vdev;
     int r, n;
 
+    /* Must unset vector notifier while guest notifier is still assigned */
+    if (kvm_irqchip_in_kernel() && !assign) {
+        msix_unset_vector_notifiers(&proxy->pci_dev);
+        g_free(proxy->vector_irqfd);
+        proxy->vector_irqfd = NULL;
+    }
+
     for (n = 0; n < VIRTIO_PCI_QUEUE_MAX; n++) {
         if (!virtio_queue_get_num(vdev, n)) {
             break;
@@ -584,10 +696,24 @@ static int virtio_pci_set_guest_notifiers(void *opaque, bool assign)
         }
     }
 
+    /* Must set vector notifier after guest notifier has been assigned */
+    if (kvm_irqchip_in_kernel() && assign) {
+        proxy->vector_irqfd =
+            g_malloc0(sizeof(*proxy->vector_irqfd) *
+                      msix_nr_vectors_allocated(&proxy->pci_dev));
+        r = msix_set_vector_notifiers(&proxy->pci_dev,
+                                      kvm_virtio_pci_vector_use,
+                                      kvm_virtio_pci_vector_release);
+        if (r < 0) {
+            goto assign_error;
+        }
+    }
+
     return 0;
 
 assign_error:
     /* We get here on assignment failure. Recover by undoing for VQs 0 .. n. */
+    assert(assign);
     while (--n >= 0) {
         virtio_pci_set_guest_notifier(opaque, n, !assign);
     }
diff --git a/hw/virtio-pci.h b/hw/virtio-pci.h
index e560428..8d28d4b 100644
--- a/hw/virtio-pci.h
+++ b/hw/virtio-pci.h
@@ -25,6 +25,11 @@
 #define VIRTIO_PCI_FLAG_USE_IOEVENTFD   (1 << VIRTIO_PCI_FLAG_USE_IOEVENTFD_BIT)
 
 typedef struct {
+    int virq;
+    unsigned int users;
+} VirtIOIRQFD;
+
+typedef struct {
     PCIDevice pci_dev;
     VirtIODevice *vdev;
     MemoryRegion bar;
@@ -44,6 +49,7 @@ typedef struct {
     VirtIOSCSIConf scsi;
     bool ioeventfd_disabled;
     bool ioeventfd_started;
+    VirtIOIRQFD *vector_irqfd;
 } VirtIOPCIProxy;
 
 void virtio_init_pci(VirtIOPCIProxy *proxy, VirtIODevice *vdev);
-- 
1.7.10.1


^ permalink raw reply related	[flat|nested] 64+ messages in thread

* [Qemu-devel] [PATCH 19/19] virtio/vhost: Add support for KVM in-kernel MSI injection
@ 2012-05-21 16:37   ` Avi Kivity
  0 siblings, 0 replies; 64+ messages in thread
From: Avi Kivity @ 2012-05-21 16:37 UTC (permalink / raw)
  To: Anthony Liguori; +Cc: Jan Kiszka, Marcelo Tosatti, qemu-devel, kvm

From: Jan Kiszka <jan.kiszka@siemens.com>

Make use of the new vector notifier to track changes of the MSI-X
configuration of virtio PCI devices. On enabling events, we establish
the required virtual IRQ to MSI-X message route and link the signaling
eventfd file descriptor to this vIRQ line. That way, vhost-generated
interrupts can be directly delivered to an in-kernel MSI-X consumer like
the x86 APIC.

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
---
 hw/virtio-pci.c |  126 +++++++++++++++++++++++++++++++++++++++++++++++++++++++
 hw/virtio-pci.h |    6 +++
 2 files changed, 132 insertions(+)

diff --git a/hw/virtio-pci.c b/hw/virtio-pci.c
index 4a4413d..01f5b92 100644
--- a/hw/virtio-pci.c
+++ b/hw/virtio-pci.c
@@ -24,6 +24,7 @@
 #include "virtio-scsi.h"
 #include "pci.h"
 #include "qemu-error.h"
+#include "msi.h"
 #include "msix.h"
 #include "net.h"
 #include "loader.h"
@@ -539,6 +540,107 @@ static void virtio_pci_guest_notifier_read(void *opaque)
     }
 }
 
+static int kvm_virtio_pci_vq_vector_use(VirtIOPCIProxy *proxy,
+                                        unsigned int queue_no,
+                                        unsigned int vector,
+                                        MSIMessage msg)
+{
+    VirtQueue *vq = virtio_get_queue(proxy->vdev, queue_no);
+    VirtIOIRQFD *irqfd = &proxy->vector_irqfd[vector];
+    int fd, ret;
+
+    fd = event_notifier_get_fd(virtio_queue_get_guest_notifier(vq));
+
+    if (irqfd->users == 0) {
+        ret = kvm_irqchip_add_msi_route(kvm_state, msg);
+        if (ret < 0) {
+            return ret;
+        }
+        irqfd->virq = ret;
+    }
+    irqfd->users++;
+
+    ret = kvm_irqchip_add_irqfd(kvm_state, fd, irqfd->virq);
+    if (ret < 0) {
+        if (--irqfd->users == 0) {
+            kvm_irqchip_release_virq(kvm_state, irqfd->virq);
+        }
+        return ret;
+    }
+
+    qemu_set_fd_handler(fd, NULL, NULL, NULL);
+
+    return 0;
+}
+
+static void kvm_virtio_pci_vq_vector_release(VirtIOPCIProxy *proxy,
+                                             unsigned int queue_no,
+                                             unsigned int vector)
+{
+    VirtQueue *vq = virtio_get_queue(proxy->vdev, queue_no);
+    VirtIOIRQFD *irqfd = &proxy->vector_irqfd[vector];
+    int fd, ret;
+
+    fd = event_notifier_get_fd(virtio_queue_get_guest_notifier(vq));
+
+    ret = kvm_irqchip_remove_irqfd(kvm_state, fd, irqfd->virq);
+    assert(ret == 0);
+
+    if (--irqfd->users == 0) {
+        kvm_irqchip_release_virq(kvm_state, irqfd->virq);
+    }
+
+    qemu_set_fd_handler(fd, virtio_pci_guest_notifier_read, NULL, vq);
+}
+
+static int kvm_virtio_pci_vector_use(PCIDevice *dev, unsigned vector,
+                                     MSIMessage msg)
+{
+    VirtIOPCIProxy *proxy = container_of(dev, VirtIOPCIProxy, pci_dev);
+    VirtIODevice *vdev = proxy->vdev;
+    int ret, queue_no;
+
+    for (queue_no = 0; queue_no < VIRTIO_PCI_QUEUE_MAX; queue_no++) {
+        if (!virtio_queue_get_num(vdev, queue_no)) {
+            break;
+        }
+        if (virtio_queue_vector(vdev, queue_no) != vector) {
+            continue;
+        }
+        ret = kvm_virtio_pci_vq_vector_use(proxy, queue_no, vector, msg);
+        if (ret < 0) {
+            goto undo;
+        }
+    }
+    return 0;
+
+undo:
+    while (--queue_no >= 0) {
+        if (virtio_queue_vector(vdev, queue_no) != vector) {
+            continue;
+        }
+        kvm_virtio_pci_vq_vector_release(proxy, queue_no, vector);
+    }
+    return ret;
+}
+
+static void kvm_virtio_pci_vector_release(PCIDevice *dev, unsigned vector)
+{
+    VirtIOPCIProxy *proxy = container_of(dev, VirtIOPCIProxy, pci_dev);
+    VirtIODevice *vdev = proxy->vdev;
+    int queue_no;
+
+    for (queue_no = 0; queue_no < VIRTIO_PCI_QUEUE_MAX; queue_no++) {
+        if (!virtio_queue_get_num(vdev, queue_no)) {
+            break;
+        }
+        if (virtio_queue_vector(vdev, queue_no) != vector) {
+            continue;
+        }
+        kvm_virtio_pci_vq_vector_release(proxy, queue_no, vector);
+    }
+}
+
 static int virtio_pci_set_guest_notifier(void *opaque, int n, bool assign)
 {
     VirtIOPCIProxy *proxy = opaque;
@@ -555,6 +657,9 @@ static int virtio_pci_set_guest_notifier(void *opaque, int n, bool assign)
     } else {
         qemu_set_fd_handler(event_notifier_get_fd(notifier),
                             NULL, NULL, NULL);
+        /* Test and clear notifier before closing it,
+         * in case poll callback didn't have time to run. */
+        virtio_pci_guest_notifier_read(vq);
         event_notifier_cleanup(notifier);
     }
 
@@ -573,6 +678,13 @@ static int virtio_pci_set_guest_notifiers(void *opaque, bool assign)
     VirtIODevice *vdev = proxy->vdev;
     int r, n;
 
+    /* Must unset vector notifier while guest notifier is still assigned */
+    if (kvm_irqchip_in_kernel() && !assign) {
+        msix_unset_vector_notifiers(&proxy->pci_dev);
+        g_free(proxy->vector_irqfd);
+        proxy->vector_irqfd = NULL;
+    }
+
     for (n = 0; n < VIRTIO_PCI_QUEUE_MAX; n++) {
         if (!virtio_queue_get_num(vdev, n)) {
             break;
@@ -584,10 +696,24 @@ static int virtio_pci_set_guest_notifiers(void *opaque, bool assign)
         }
     }
 
+    /* Must set vector notifier after guest notifier has been assigned */
+    if (kvm_irqchip_in_kernel() && assign) {
+        proxy->vector_irqfd =
+            g_malloc0(sizeof(*proxy->vector_irqfd) *
+                      msix_nr_vectors_allocated(&proxy->pci_dev));
+        r = msix_set_vector_notifiers(&proxy->pci_dev,
+                                      kvm_virtio_pci_vector_use,
+                                      kvm_virtio_pci_vector_release);
+        if (r < 0) {
+            goto assign_error;
+        }
+    }
+
     return 0;
 
 assign_error:
     /* We get here on assignment failure. Recover by undoing for VQs 0 .. n. */
+    assert(assign);
     while (--n >= 0) {
         virtio_pci_set_guest_notifier(opaque, n, !assign);
     }
diff --git a/hw/virtio-pci.h b/hw/virtio-pci.h
index e560428..8d28d4b 100644
--- a/hw/virtio-pci.h
+++ b/hw/virtio-pci.h
@@ -25,6 +25,11 @@
 #define VIRTIO_PCI_FLAG_USE_IOEVENTFD   (1 << VIRTIO_PCI_FLAG_USE_IOEVENTFD_BIT)
 
 typedef struct {
+    int virq;
+    unsigned int users;
+} VirtIOIRQFD;
+
+typedef struct {
     PCIDevice pci_dev;
     VirtIODevice *vdev;
     MemoryRegion bar;
@@ -44,6 +49,7 @@ typedef struct {
     VirtIOSCSIConf scsi;
     bool ioeventfd_disabled;
     bool ioeventfd_started;
+    VirtIOIRQFD *vector_irqfd;
 } VirtIOPCIProxy;
 
 void virtio_init_pci(VirtIOPCIProxy *proxy, VirtIODevice *vdev);
-- 
1.7.10.1

^ permalink raw reply related	[flat|nested] 64+ messages in thread

* Re: [Qemu-devel] [PULL 00/20 1.2] kvm updates
  2012-05-21 16:37 ` [Qemu-devel] " Avi Kivity
                   ` (19 preceding siblings ...)
  (?)
@ 2012-06-04  5:46 ` Anthony Liguori
  2012-06-05  0:52     ` Andreas Färber
  -1 siblings, 1 reply; 64+ messages in thread
From: Anthony Liguori @ 2012-06-04  5:46 UTC (permalink / raw)
  To: Avi Kivity; +Cc: Jan Kiszka, Marcelo Tosatti, qemu-devel, kvm

On 05/22/2012 12:37 AM, Avi Kivity wrote:
> Please pull from:
>
>    git://git.kernel.org/pub/scm/virt/kvm/qemu-kvm.git uq/master

Pulled.  Thanks.

Regards,

Anthony Liguori

>
> This is mostly Jan's MSI for KVM patches.
>
> ----------------------------------------------------------------
> Avi Kivity (1):
>        kvm: update vmxcap for EPT A/D, INVPCID, RDRAND, VMFUNC
>
> Jan Kiszka (19):
>        kvm: Refactor KVMState::max_gsi to gsi_count
>        Introduce MSIMessage structure
>        kvm: Introduce basic MSI support for in-kernel irqchips
>        pc: Enable MSI support at APIC level
>        kvm: x86: Wire up MSI support for in-kernel irqchip
>        kvm: Update kernel headers
>        kvm: Add support for direct MSI injections
>        kvm: Enable in-kernel irqchip support by default
>        msix: Factor out msix_get_message
>        msix: Invoke msix_handle_mask_update on msix_mask_all
>        msix: Introduce vector notifiers
>        kvm: Rename kvm_irqchip_add_route to kvm_irqchip_add_irq_route
>        kvm: Introduce kvm_irqchip_add_msi_route
>        kvm: Publicize kvm_irqchip_release_virq
>        kvm: Make kvm_irqchip_commit_routes an internal service
>        kvm: Introduce kvm_irqchip_add/remove_irqfd
>        kvm: Enable use of kvm_irqchip_in_kernel in hwlib code
>        msix: Add msix_nr_vectors_allocated
>        virtio/vhost: Add support for KVM in-kernel MSI injection
>
>   hw/apic.c                 |    3 +
>   hw/kvm/apic.c             |   34 ++++++-
>   hw/msi.h                  |    5 +
>   hw/msix.c                 |  121 +++++++++++++++++++++--
>   hw/msix.h                 |    6 ++
>   hw/pc.c                   |    9 --
>   hw/pc_piix.c              |   14 +--
>   hw/pci.h                  |    8 ++
>   hw/virtio-pci.c           |  126 ++++++++++++++++++++++++
>   hw/virtio-pci.h           |    6 ++
>   hw/xen.h                  |   10 --
>   hw/xen_apic.c             |    5 +
>   kvm-all.c                 |  236 +++++++++++++++++++++++++++++++++++++++++++--
>   kvm-stub.c                |   23 +++++
>   kvm.h                     |   18 ++--
>   linux-headers/linux/kvm.h |   38 ++++++++
>   qemu-common.h             |    1 +
>   scripts/kvm/vmxcap        |   13 +++
>   18 files changed, 624 insertions(+), 52 deletions(-)
>
>


^ permalink raw reply	[flat|nested] 64+ messages in thread

* Re: [Qemu-devel] [PULL 00/20 1.2] kvm updates
  2012-06-04  5:46 ` [Qemu-devel] [PULL 00/20 1.2] kvm updates Anthony Liguori
  2012-06-05  0:52     ` Andreas Färber
@ 2012-06-05  0:52     ` Andreas Färber
  0 siblings, 0 replies; 64+ messages in thread
From: Andreas Färber @ 2012-06-05  0:52 UTC (permalink / raw)
  To: Anthony Liguori, Avi Kivity
  Cc: Jan Kiszka, Marcelo Tosatti, qemu-devel, kvm, Alexander Graf, kvm-ppc

Am 04.06.2012 07:46, schrieb Anthony Liguori:
> On 05/22/2012 12:37 AM, Avi Kivity wrote:
>> Please pull from:
>>
>>    git://git.kernel.org/pub/scm/virt/kvm/qemu-kvm.git uq/master
> 
> Pulled.  Thanks.

This broke the ppc build. Guys, why wasn't this tested? There's only
three KVM targets to test compared to the 14 I'm struggling with...

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] 64+ messages in thread

* Re: [Qemu-devel] [PULL 00/20 1.2] kvm updates
@ 2012-06-05  0:52     ` Andreas Färber
  0 siblings, 0 replies; 64+ messages in thread
From: Andreas Färber @ 2012-06-05  0:52 UTC (permalink / raw)
  To: Anthony Liguori, Avi Kivity
  Cc: kvm, Jan Kiszka, Marcelo Tosatti, qemu-devel, kvm-ppc, Alexander Graf

Am 04.06.2012 07:46, schrieb Anthony Liguori:
> On 05/22/2012 12:37 AM, Avi Kivity wrote:
>> Please pull from:
>>
>>    git://git.kernel.org/pub/scm/virt/kvm/qemu-kvm.git uq/master
> 
> Pulled.  Thanks.

This broke the ppc build. Guys, why wasn't this tested? There's only
three KVM targets to test compared to the 14 I'm struggling with...

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] 64+ messages in thread

* Re: [Qemu-devel] [PULL 00/20 1.2] kvm updates
@ 2012-06-05  0:52     ` Andreas Färber
  0 siblings, 0 replies; 64+ messages in thread
From: Andreas Färber @ 2012-06-05  0:52 UTC (permalink / raw)
  To: Anthony Liguori, Avi Kivity
  Cc: Jan Kiszka, Marcelo Tosatti, qemu-devel, kvm, Alexander Graf, kvm-ppc

Am 04.06.2012 07:46, schrieb Anthony Liguori:
> On 05/22/2012 12:37 AM, Avi Kivity wrote:
>> Please pull from:
>>
>>    git://git.kernel.org/pub/scm/virt/kvm/qemu-kvm.git uq/master
> 
> Pulled.  Thanks.

This broke the ppc build. Guys, why wasn't this tested? There's only
three KVM targets to test compared to the 14 I'm struggling with...

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] 64+ messages in thread

* Re: [Qemu-devel] [PULL 00/20 1.2] kvm updates
  2012-06-05  0:52     ` Andreas Färber
  (?)
@ 2012-06-05  1:58       ` Anthony Liguori
  -1 siblings, 0 replies; 64+ messages in thread
From: Anthony Liguori @ 2012-06-05  1:58 UTC (permalink / raw)
  To: Andreas Färber
  Cc: Avi Kivity, kvm, Jan Kiszka, Marcelo Tosatti, qemu-devel,
	kvm-ppc, Alexander Graf

On 06/05/2012 08:52 AM, Andreas Färber wrote:
> Am 04.06.2012 07:46, schrieb Anthony Liguori:
>> On 05/22/2012 12:37 AM, Avi Kivity wrote:
>>> Please pull from:
>>>
>>>     git://git.kernel.org/pub/scm/virt/kvm/qemu-kvm.git uq/master
>>
>> Pulled.  Thanks.
>
> This broke the ppc build. Guys, why wasn't this tested? There's only
> three KVM targets to test compared to the 14 I'm struggling with...

Is build bot running against uq/master?  If it's not, maybe we should add it to 
build bot to catch this sort of thing.

Regards,

Anthony Liguori

> Andreas
>

^ permalink raw reply	[flat|nested] 64+ messages in thread

* Re: [Qemu-devel] [PULL 00/20 1.2] kvm updates
@ 2012-06-05  1:58       ` Anthony Liguori
  0 siblings, 0 replies; 64+ messages in thread
From: Anthony Liguori @ 2012-06-05  1:58 UTC (permalink / raw)
  To: Andreas Färber
  Cc: kvm, Jan Kiszka, Marcelo Tosatti, qemu-devel, kvm-ppc,
	Alexander Graf, Avi Kivity

On 06/05/2012 08:52 AM, Andreas Färber wrote:
> Am 04.06.2012 07:46, schrieb Anthony Liguori:
>> On 05/22/2012 12:37 AM, Avi Kivity wrote:
>>> Please pull from:
>>>
>>>     git://git.kernel.org/pub/scm/virt/kvm/qemu-kvm.git uq/master
>>
>> Pulled.  Thanks.
>
> This broke the ppc build. Guys, why wasn't this tested? There's only
> three KVM targets to test compared to the 14 I'm struggling with...

Is build bot running against uq/master?  If it's not, maybe we should add it to 
build bot to catch this sort of thing.

Regards,

Anthony Liguori

> Andreas
>

^ permalink raw reply	[flat|nested] 64+ messages in thread

* Re: [Qemu-devel] [PULL 00/20 1.2] kvm updates
@ 2012-06-05  1:58       ` Anthony Liguori
  0 siblings, 0 replies; 64+ messages in thread
From: Anthony Liguori @ 2012-06-05  1:58 UTC (permalink / raw)
  To: Andreas Färber
  Cc: Avi Kivity, kvm, Jan Kiszka, Marcelo Tosatti, qemu-devel,
	kvm-ppc, Alexander Graf

On 06/05/2012 08:52 AM, Andreas Färber wrote:
> Am 04.06.2012 07:46, schrieb Anthony Liguori:
>> On 05/22/2012 12:37 AM, Avi Kivity wrote:
>>> Please pull from:
>>>
>>>     git://git.kernel.org/pub/scm/virt/kvm/qemu-kvm.git uq/master
>>
>> Pulled.  Thanks.
>
> This broke the ppc build. Guys, why wasn't this tested? There's only
> three KVM targets to test compared to the 14 I'm struggling with...

Is build bot running against uq/master?  If it's not, maybe we should add it to 
build bot to catch this sort of thing.

Regards,

Anthony Liguori

> Andreas
>


^ permalink raw reply	[flat|nested] 64+ messages in thread

* Re: [Qemu-devel] [PULL 00/20 1.2] kvm updates
  2012-06-05  1:58       ` Anthony Liguori
  (?)
@ 2012-06-05  2:26         ` Andreas Färber
  -1 siblings, 0 replies; 64+ messages in thread
From: Andreas Färber @ 2012-06-05  2:26 UTC (permalink / raw)
  To: Anthony Liguori, Stefan Hajnoczi, Daniel Gollub
  Cc: Avi Kivity, kvm, Jan Kiszka, Marcelo Tosatti, qemu-devel,
	kvm-ppc, Alexander Graf

Am 05.06.2012 03:58, schrieb Anthony Liguori:
> Is build bot running against uq/master?  If it's not, maybe we should
> add it to build bot to catch this sort of thing.

That's a question for Stefan and Daniel to answer.

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] 64+ messages in thread

* Re: [Qemu-devel] [PULL 00/20 1.2] kvm updates
@ 2012-06-05  2:26         ` Andreas Färber
  0 siblings, 0 replies; 64+ messages in thread
From: Andreas Färber @ 2012-06-05  2:26 UTC (permalink / raw)
  To: Anthony Liguori, Stefan Hajnoczi, Daniel Gollub
  Cc: kvm, Jan Kiszka, Marcelo Tosatti, qemu-devel, kvm-ppc,
	Alexander Graf, Avi Kivity

Am 05.06.2012 03:58, schrieb Anthony Liguori:
> Is build bot running against uq/master?  If it's not, maybe we should
> add it to build bot to catch this sort of thing.

That's a question for Stefan and Daniel to answer.

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] 64+ messages in thread

* Re: [Qemu-devel] [PULL 00/20 1.2] kvm updates
@ 2012-06-05  2:26         ` Andreas Färber
  0 siblings, 0 replies; 64+ messages in thread
From: Andreas Färber @ 2012-06-05  2:26 UTC (permalink / raw)
  To: Anthony Liguori, Stefan Hajnoczi, Daniel Gollub
  Cc: Avi Kivity, kvm, Jan Kiszka, Marcelo Tosatti, qemu-devel,
	kvm-ppc, Alexander Graf

Am 05.06.2012 03:58, schrieb Anthony Liguori:
> Is build bot running against uq/master?  If it's not, maybe we should
> add it to build bot to catch this sort of thing.

That's a question for Stefan and Daniel to answer.

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] 64+ messages in thread

* Re: [Qemu-devel] [PULL 00/20 1.2] kvm updates
  2012-06-05  1:58       ` Anthony Liguori
  (?)
@ 2012-06-05  7:42         ` Avi Kivity
  -1 siblings, 0 replies; 64+ messages in thread
From: Avi Kivity @ 2012-06-05  7:42 UTC (permalink / raw)
  To: Anthony Liguori
  Cc: Andreas Färber, kvm, Jan Kiszka, Marcelo Tosatti,
	qemu-devel, kvm-ppc, Alexander Graf

On 06/05/2012 04:58 AM, Anthony Liguori wrote:
> On 06/05/2012 08:52 AM, Andreas Färber wrote:
>> Am 04.06.2012 07:46, schrieb Anthony Liguori:
>>> On 05/22/2012 12:37 AM, Avi Kivity wrote:
>>>> Please pull from:
>>>>
>>>>     git://git.kernel.org/pub/scm/virt/kvm/qemu-kvm.git uq/master
>>>
>>> Pulled.  Thanks.
>>
>> This broke the ppc build. Guys, why wasn't this tested? There's only
>> three KVM targets to test compared to the 14 I'm struggling with...
> 
> Is build bot running against uq/master?  If it's not, maybe we should
> add it to build bot to catch this sort of thing.

A build bot is overkill.  I usually do a full build (and autotest),
don't know how this was missed.


-- 
error compiling committee.c: too many arguments to function

^ permalink raw reply	[flat|nested] 64+ messages in thread

* Re: [Qemu-devel] [PULL 00/20 1.2] kvm updates
@ 2012-06-05  7:42         ` Avi Kivity
  0 siblings, 0 replies; 64+ messages in thread
From: Avi Kivity @ 2012-06-05  7:42 UTC (permalink / raw)
  To: Anthony Liguori
  Cc: kvm, Jan Kiszka, Marcelo Tosatti, qemu-devel, kvm-ppc,
	Alexander Graf, Andreas Färber

On 06/05/2012 04:58 AM, Anthony Liguori wrote:
> On 06/05/2012 08:52 AM, Andreas Färber wrote:
>> Am 04.06.2012 07:46, schrieb Anthony Liguori:
>>> On 05/22/2012 12:37 AM, Avi Kivity wrote:
>>>> Please pull from:
>>>>
>>>>     git://git.kernel.org/pub/scm/virt/kvm/qemu-kvm.git uq/master
>>>
>>> Pulled.  Thanks.
>>
>> This broke the ppc build. Guys, why wasn't this tested? There's only
>> three KVM targets to test compared to the 14 I'm struggling with...
> 
> Is build bot running against uq/master?  If it's not, maybe we should
> add it to build bot to catch this sort of thing.

A build bot is overkill.  I usually do a full build (and autotest),
don't know how this was missed.


-- 
error compiling committee.c: too many arguments to function

^ permalink raw reply	[flat|nested] 64+ messages in thread

* Re: [Qemu-devel] [PULL 00/20 1.2] kvm updates
@ 2012-06-05  7:42         ` Avi Kivity
  0 siblings, 0 replies; 64+ messages in thread
From: Avi Kivity @ 2012-06-05  7:42 UTC (permalink / raw)
  To: Anthony Liguori
  Cc: Andreas Färber, kvm, Jan Kiszka, Marcelo Tosatti,
	qemu-devel, kvm-ppc, Alexander Graf

On 06/05/2012 04:58 AM, Anthony Liguori wrote:
> On 06/05/2012 08:52 AM, Andreas Färber wrote:
>> Am 04.06.2012 07:46, schrieb Anthony Liguori:
>>> On 05/22/2012 12:37 AM, Avi Kivity wrote:
>>>> Please pull from:
>>>>
>>>>     git://git.kernel.org/pub/scm/virt/kvm/qemu-kvm.git uq/master
>>>
>>> Pulled.  Thanks.
>>
>> This broke the ppc build. Guys, why wasn't this tested? There's only
>> three KVM targets to test compared to the 14 I'm struggling with...
> 
> Is build bot running against uq/master?  If it's not, maybe we should
> add it to build bot to catch this sort of thing.

A build bot is overkill.  I usually do a full build (and autotest),
don't know how this was missed.


-- 
error compiling committee.c: too many arguments to function

^ permalink raw reply	[flat|nested] 64+ messages in thread

* Re: [PULL 00/20 1.2] kvm updates
  2012-06-05  7:42         ` Avi Kivity
  (?)
@ 2012-06-05  8:10           ` Alexander Graf
  -1 siblings, 0 replies; 64+ messages in thread
From: Alexander Graf @ 2012-06-05  8:10 UTC (permalink / raw)
  To: Avi Kivity
  Cc: kvm, Jan Kiszka, Marcelo Tosatti, qemu-devel, kvm-ppc,
	Anthony Liguori, Andreas Färber



On 05.06.2012, at 09:42, Avi Kivity <avi@redhat.com> wrote:

> On 06/05/2012 04:58 AM, Anthony Liguori wrote:
>> On 06/05/2012 08:52 AM, Andreas Färber wrote:
>>> Am 04.06.2012 07:46, schrieb Anthony Liguori:
>>>> On 05/22/2012 12:37 AM, Avi Kivity wrote:
>>>>> Please pull from:
>>>>> 
>>>>>    git://git.kernel.org/pub/scm/virt/kvm/qemu-kvm.git uq/master
>>>> 
>>>> Pulled.  Thanks.
>>> 
>>> This broke the ppc build. Guys, why wasn't this tested? There's only
>>> three KVM targets to test compared to the 14 I'm struggling with...
>> 
>> Is build bot running against uq/master?  If it's not, maybe we should
>> add it to build bot to catch this sort of thing.
> 
> A build bot is overkill.  I usually do a full build (and autotest),
> don't know how this was missed.

Not sure why it would be overkill. We have build bots running on more exotic hw (ppc and s390) with kvm available for that exact purpose. They're running either way - all we need to do is enable uq/master and we catch these before the next pull request :).


Alex

^ permalink raw reply	[flat|nested] 64+ messages in thread

* Re: [Qemu-devel] [PULL 00/20 1.2] kvm updates
@ 2012-06-05  8:10           ` Alexander Graf
  0 siblings, 0 replies; 64+ messages in thread
From: Alexander Graf @ 2012-06-05  8:10 UTC (permalink / raw)
  To: Avi Kivity
  Cc: kvm, Jan Kiszka, Marcelo Tosatti, qemu-devel, kvm-ppc,
	Anthony Liguori, Andreas Färber



On 05.06.2012, at 09:42, Avi Kivity <avi@redhat.com> wrote:

> On 06/05/2012 04:58 AM, Anthony Liguori wrote:
>> On 06/05/2012 08:52 AM, Andreas Färber wrote:
>>> Am 04.06.2012 07:46, schrieb Anthony Liguori:
>>>> On 05/22/2012 12:37 AM, Avi Kivity wrote:
>>>>> Please pull from:
>>>>> 
>>>>>    git://git.kernel.org/pub/scm/virt/kvm/qemu-kvm.git uq/master
>>>> 
>>>> Pulled.  Thanks.
>>> 
>>> This broke the ppc build. Guys, why wasn't this tested? There's only
>>> three KVM targets to test compared to the 14 I'm struggling with...
>> 
>> Is build bot running against uq/master?  If it's not, maybe we should
>> add it to build bot to catch this sort of thing.
> 
> A build bot is overkill.  I usually do a full build (and autotest),
> don't know how this was missed.

Not sure why it would be overkill. We have build bots running on more exotic hw (ppc and s390) with kvm available for that exact purpose. They're running either way - all we need to do is enable uq/master and we catch these before the next pull request :).


Alex

^ permalink raw reply	[flat|nested] 64+ messages in thread

* Re: [Qemu-devel] [PULL 00/20 1.2] kvm updates
@ 2012-06-05  8:10           ` Alexander Graf
  0 siblings, 0 replies; 64+ messages in thread
From: Alexander Graf @ 2012-06-05  8:10 UTC (permalink / raw)
  To: Avi Kivity
  Cc: kvm, Jan Kiszka, Marcelo Tosatti, qemu-devel, kvm-ppc,
	Anthony Liguori, Andreas Färber



On 05.06.2012, at 09:42, Avi Kivity <avi@redhat.com> wrote:

> On 06/05/2012 04:58 AM, Anthony Liguori wrote:
>> On 06/05/2012 08:52 AM, Andreas FÃ¤rber wrote:
>>> Am 04.06.2012 07:46, schrieb Anthony Liguori:
>>>> On 05/22/2012 12:37 AM, Avi Kivity wrote:
>>>>> Please pull from:
>>>>> 
>>>>>    git://git.kernel.org/pub/scm/virt/kvm/qemu-kvm.git uq/master
>>>> 
>>>> Pulled.  Thanks.
>>> 
>>> This broke the ppc build. Guys, why wasn't this tested? There's only
>>> three KVM targets to test compared to the 14 I'm struggling with...
>> 
>> Is build bot running against uq/master?  If it's not, maybe we should
>> add it to build bot to catch this sort of thing.
> 
> A build bot is overkill.  I usually do a full build (and autotest),
> don't know how this was missed.

Not sure why it would be overkill. We have build bots running on more exotic hw (ppc and s390) with kvm available for that exact purpose. They're running either way - all we need to do is enable uq/master and we catch these before the next pull request :).


Alex


^ permalink raw reply	[flat|nested] 64+ messages in thread

* Re: [Qemu-devel] [PULL 00/20 1.2] kvm updates
  2012-06-05  8:10           ` Alexander Graf
  (?)
@ 2012-06-05  8:32             ` Jan Kiszka
  -1 siblings, 0 replies; 64+ messages in thread
From: Jan Kiszka @ 2012-06-05  8:32 UTC (permalink / raw)
  To: Alexander Graf
  Cc: Avi Kivity, Anthony Liguori, Andreas Färber, kvm,
	Marcelo Tosatti, qemu-devel, kvm-ppc

On 2012-06-05 10:10, Alexander Graf wrote:
> 
> 
> On 05.06.2012, at 09:42, Avi Kivity <avi@redhat.com> wrote:
> 
>> On 06/05/2012 04:58 AM, Anthony Liguori wrote:
>>> On 06/05/2012 08:52 AM, Andreas Färber wrote:
>>>> Am 04.06.2012 07:46, schrieb Anthony Liguori:
>>>>> On 05/22/2012 12:37 AM, Avi Kivity wrote:
>>>>>> Please pull from:
>>>>>>
>>>>>>    git://git.kernel.org/pub/scm/virt/kvm/qemu-kvm.git uq/master
>>>>>
>>>>> Pulled.  Thanks.
>>>>
>>>> This broke the ppc build. Guys, why wasn't this tested? There's only
>>>> three KVM targets to test compared to the 14 I'm struggling with...
>>>
>>> Is build bot running against uq/master?  If it's not, maybe we should
>>> add it to build bot to catch this sort of thing.
>>
>> A build bot is overkill.  I usually do a full build (and autotest),
>> don't know how this was missed.
> 
> Not sure why it would be overkill. We have build bots running on more exotic hw (ppc and s390) with kvm available for that exact purpose. They're running either way - all we need to do is enable uq/master and we catch these before the next pull request :).

I think this makes some sense. uq/master is supposed to be the staging
branch for generic and (at least) x86 kvm changes in QEMU. So it should
cover at least their builds.

Jan

-- 
Siemens AG, Corporate Technology, CT T DE IT 1
Corporate Competence Center Embedded Linux

^ permalink raw reply	[flat|nested] 64+ messages in thread

* Re: [Qemu-devel] [PULL 00/20 1.2] kvm updates
@ 2012-06-05  8:32             ` Jan Kiszka
  0 siblings, 0 replies; 64+ messages in thread
From: Jan Kiszka @ 2012-06-05  8:32 UTC (permalink / raw)
  To: Alexander Graf
  Cc: kvm, Marcelo Tosatti, qemu-devel, kvm-ppc, Avi Kivity,
	Anthony Liguori, Andreas Färber

On 2012-06-05 10:10, Alexander Graf wrote:
> 
> 
> On 05.06.2012, at 09:42, Avi Kivity <avi@redhat.com> wrote:
> 
>> On 06/05/2012 04:58 AM, Anthony Liguori wrote:
>>> On 06/05/2012 08:52 AM, Andreas Färber wrote:
>>>> Am 04.06.2012 07:46, schrieb Anthony Liguori:
>>>>> On 05/22/2012 12:37 AM, Avi Kivity wrote:
>>>>>> Please pull from:
>>>>>>
>>>>>>    git://git.kernel.org/pub/scm/virt/kvm/qemu-kvm.git uq/master
>>>>>
>>>>> Pulled.  Thanks.
>>>>
>>>> This broke the ppc build. Guys, why wasn't this tested? There's only
>>>> three KVM targets to test compared to the 14 I'm struggling with...
>>>
>>> Is build bot running against uq/master?  If it's not, maybe we should
>>> add it to build bot to catch this sort of thing.
>>
>> A build bot is overkill.  I usually do a full build (and autotest),
>> don't know how this was missed.
> 
> Not sure why it would be overkill. We have build bots running on more exotic hw (ppc and s390) with kvm available for that exact purpose. They're running either way - all we need to do is enable uq/master and we catch these before the next pull request :).

I think this makes some sense. uq/master is supposed to be the staging
branch for generic and (at least) x86 kvm changes in QEMU. So it should
cover at least their builds.

Jan

-- 
Siemens AG, Corporate Technology, CT T DE IT 1
Corporate Competence Center Embedded Linux

^ permalink raw reply	[flat|nested] 64+ messages in thread

* Re: [Qemu-devel] [PULL 00/20 1.2] kvm updates
@ 2012-06-05  8:32             ` Jan Kiszka
  0 siblings, 0 replies; 64+ messages in thread
From: Jan Kiszka @ 2012-06-05  8:32 UTC (permalink / raw)
  To: Alexander Graf
  Cc: Avi Kivity, Anthony Liguori, Andreas Färber, kvm,
	Marcelo Tosatti, qemu-devel, kvm-ppc

On 2012-06-05 10:10, Alexander Graf wrote:
> 
> 
> On 05.06.2012, at 09:42, Avi Kivity <avi@redhat.com> wrote:
> 
>> On 06/05/2012 04:58 AM, Anthony Liguori wrote:
>>> On 06/05/2012 08:52 AM, Andreas FÃ¤rber wrote:
>>>> Am 04.06.2012 07:46, schrieb Anthony Liguori:
>>>>> On 05/22/2012 12:37 AM, Avi Kivity wrote:
>>>>>> Please pull from:
>>>>>>
>>>>>>    git://git.kernel.org/pub/scm/virt/kvm/qemu-kvm.git uq/master
>>>>>
>>>>> Pulled.  Thanks.
>>>>
>>>> This broke the ppc build. Guys, why wasn't this tested? There's only
>>>> three KVM targets to test compared to the 14 I'm struggling with...
>>>
>>> Is build bot running against uq/master?  If it's not, maybe we should
>>> add it to build bot to catch this sort of thing.
>>
>> A build bot is overkill.  I usually do a full build (and autotest),
>> don't know how this was missed.
> 
> Not sure why it would be overkill. We have build bots running on more exotic hw (ppc and s390) with kvm available for that exact purpose. They're running either way - all we need to do is enable uq/master and we catch these before the next pull request :).

I think this makes some sense. uq/master is supposed to be the staging
branch for generic and (at least) x86 kvm changes in QEMU. So it should
cover at least their builds.

Jan

-- 
Siemens AG, Corporate Technology, CT T DE IT 1
Corporate Competence Center Embedded Linux

^ permalink raw reply	[flat|nested] 64+ messages in thread

* Re: [Qemu-devel] [PULL 00/20 1.2] kvm updates
  2012-06-05  7:42         ` Avi Kivity
@ 2012-06-05  9:02           ` Anthony Liguori
  -1 siblings, 0 replies; 64+ messages in thread
From: Anthony Liguori @ 2012-06-05  9:02 UTC (permalink / raw)
  To: Avi Kivity
  Cc: kvm, Jan Kiszka, Marcelo Tosatti, qemu-devel, kvm-ppc,
	Alexander Graf, Andreas Färber

On 06/05/2012 03:42 PM, Avi Kivity wrote:
> On 06/05/2012 04:58 AM, Anthony Liguori wrote:
>> On 06/05/2012 08:52 AM, Andreas Färber wrote:
>>> Am 04.06.2012 07:46, schrieb Anthony Liguori:
>>>> On 05/22/2012 12:37 AM, Avi Kivity wrote:
>>>>> Please pull from:
>>>>>
>>>>>      git://git.kernel.org/pub/scm/virt/kvm/qemu-kvm.git uq/master
>>>>
>>>> Pulled.  Thanks.
>>>
>>> This broke the ppc build. Guys, why wasn't this tested? There's only
>>> three KVM targets to test compared to the 14 I'm struggling with...
>>
>> Is build bot running against uq/master?  If it's not, maybe we should
>> add it to build bot to catch this sort of thing.
>
> A build bot is overkill.  I usually do a full build (and autotest),
> don't know how this was missed.

Note that this error is specific to KVM on PowerPC.  It doesn't affect 
qemu-system-ppc64 built on an x86_64 host.

Regards,

Anthony Liguori

>
>

^ permalink raw reply	[flat|nested] 64+ messages in thread

* Re: [Qemu-devel] [PULL 00/20 1.2] kvm updates
@ 2012-06-05  9:02           ` Anthony Liguori
  0 siblings, 0 replies; 64+ messages in thread
From: Anthony Liguori @ 2012-06-05  9:02 UTC (permalink / raw)
  To: Avi Kivity
  Cc: kvm, Jan Kiszka, Marcelo Tosatti, qemu-devel, kvm-ppc,
	Alexander Graf, Andreas Färber

On 06/05/2012 03:42 PM, Avi Kivity wrote:
> On 06/05/2012 04:58 AM, Anthony Liguori wrote:
>> On 06/05/2012 08:52 AM, Andreas Färber wrote:
>>> Am 04.06.2012 07:46, schrieb Anthony Liguori:
>>>> On 05/22/2012 12:37 AM, Avi Kivity wrote:
>>>>> Please pull from:
>>>>>
>>>>>      git://git.kernel.org/pub/scm/virt/kvm/qemu-kvm.git uq/master
>>>>
>>>> Pulled.  Thanks.
>>>
>>> This broke the ppc build. Guys, why wasn't this tested? There's only
>>> three KVM targets to test compared to the 14 I'm struggling with...
>>
>> Is build bot running against uq/master?  If it's not, maybe we should
>> add it to build bot to catch this sort of thing.
>
> A build bot is overkill.  I usually do a full build (and autotest),
> don't know how this was missed.

Note that this error is specific to KVM on PowerPC.  It doesn't affect 
qemu-system-ppc64 built on an x86_64 host.

Regards,

Anthony Liguori

>
>


^ permalink raw reply	[flat|nested] 64+ messages in thread

* Re: [PULL 00/20 1.2] kvm updates
  2012-06-05  9:02           ` Anthony Liguori
  (?)
@ 2012-06-05  9:11             ` Avi Kivity
  -1 siblings, 0 replies; 64+ messages in thread
From: Avi Kivity @ 2012-06-05  9:11 UTC (permalink / raw)
  To: Anthony Liguori
  Cc: kvm, Jan Kiszka, Marcelo Tosatti, qemu-devel, kvm-ppc,
	Alexander Graf, Andreas Färber

On 06/05/2012 12:02 PM, Anthony Liguori wrote:
> On 06/05/2012 03:42 PM, Avi Kivity wrote:
>> On 06/05/2012 04:58 AM, Anthony Liguori wrote:
>>> On 06/05/2012 08:52 AM, Andreas Färber wrote:
>>>> Am 04.06.2012 07:46, schrieb Anthony Liguori:
>>>>> On 05/22/2012 12:37 AM, Avi Kivity wrote:
>>>>>> Please pull from:
>>>>>>
>>>>>>      git://git.kernel.org/pub/scm/virt/kvm/qemu-kvm.git uq/master
>>>>>
>>>>> Pulled.  Thanks.
>>>>
>>>> This broke the ppc build. Guys, why wasn't this tested? There's only
>>>> three KVM targets to test compared to the 14 I'm struggling with...
>>>
>>> Is build bot running against uq/master?  If it's not, maybe we should
>>> add it to build bot to catch this sort of thing.
>>
>> A build bot is overkill.  I usually do a full build (and autotest),
>> don't know how this was missed.
> 
> Note that this error is specific to KVM on PowerPC.  It doesn't affect
> qemu-system-ppc64 built on an x86_64 host.

Ah, okay.

In any case I take back my comment about overkill.  An extra buildbot
run is cheap and overkill is better than underkill.

-- 
error compiling committee.c: too many arguments to function

^ permalink raw reply	[flat|nested] 64+ messages in thread

* Re: [Qemu-devel] [PULL 00/20 1.2] kvm updates
@ 2012-06-05  9:11             ` Avi Kivity
  0 siblings, 0 replies; 64+ messages in thread
From: Avi Kivity @ 2012-06-05  9:11 UTC (permalink / raw)
  To: Anthony Liguori
  Cc: kvm, Jan Kiszka, Marcelo Tosatti, qemu-devel, kvm-ppc,
	Alexander Graf, Andreas Färber

On 06/05/2012 12:02 PM, Anthony Liguori wrote:
> On 06/05/2012 03:42 PM, Avi Kivity wrote:
>> On 06/05/2012 04:58 AM, Anthony Liguori wrote:
>>> On 06/05/2012 08:52 AM, Andreas Färber wrote:
>>>> Am 04.06.2012 07:46, schrieb Anthony Liguori:
>>>>> On 05/22/2012 12:37 AM, Avi Kivity wrote:
>>>>>> Please pull from:
>>>>>>
>>>>>>      git://git.kernel.org/pub/scm/virt/kvm/qemu-kvm.git uq/master
>>>>>
>>>>> Pulled.  Thanks.
>>>>
>>>> This broke the ppc build. Guys, why wasn't this tested? There's only
>>>> three KVM targets to test compared to the 14 I'm struggling with...
>>>
>>> Is build bot running against uq/master?  If it's not, maybe we should
>>> add it to build bot to catch this sort of thing.
>>
>> A build bot is overkill.  I usually do a full build (and autotest),
>> don't know how this was missed.
> 
> Note that this error is specific to KVM on PowerPC.  It doesn't affect
> qemu-system-ppc64 built on an x86_64 host.

Ah, okay.

In any case I take back my comment about overkill.  An extra buildbot
run is cheap and overkill is better than underkill.

-- 
error compiling committee.c: too many arguments to function

^ permalink raw reply	[flat|nested] 64+ messages in thread

* Re: [Qemu-devel] [PULL 00/20 1.2] kvm updates
@ 2012-06-05  9:11             ` Avi Kivity
  0 siblings, 0 replies; 64+ messages in thread
From: Avi Kivity @ 2012-06-05  9:11 UTC (permalink / raw)
  To: Anthony Liguori
  Cc: kvm, Jan Kiszka, Marcelo Tosatti, qemu-devel, kvm-ppc,
	Alexander Graf, Andreas Färber

On 06/05/2012 12:02 PM, Anthony Liguori wrote:
> On 06/05/2012 03:42 PM, Avi Kivity wrote:
>> On 06/05/2012 04:58 AM, Anthony Liguori wrote:
>>> On 06/05/2012 08:52 AM, Andreas Färber wrote:
>>>> Am 04.06.2012 07:46, schrieb Anthony Liguori:
>>>>> On 05/22/2012 12:37 AM, Avi Kivity wrote:
>>>>>> Please pull from:
>>>>>>
>>>>>>      git://git.kernel.org/pub/scm/virt/kvm/qemu-kvm.git uq/master
>>>>>
>>>>> Pulled.  Thanks.
>>>>
>>>> This broke the ppc build. Guys, why wasn't this tested? There's only
>>>> three KVM targets to test compared to the 14 I'm struggling with...
>>>
>>> Is build bot running against uq/master?  If it's not, maybe we should
>>> add it to build bot to catch this sort of thing.
>>
>> A build bot is overkill.  I usually do a full build (and autotest),
>> don't know how this was missed.
> 
> Note that this error is specific to KVM on PowerPC.  It doesn't affect
> qemu-system-ppc64 built on an x86_64 host.

Ah, okay.

In any case I take back my comment about overkill.  An extra buildbot
run is cheap and overkill is better than underkill.

-- 
error compiling committee.c: too many arguments to function

^ permalink raw reply	[flat|nested] 64+ messages in thread

end of thread, other threads:[~2012-06-05  9:12 UTC | newest]

Thread overview: 64+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2012-05-21 16:37 [PULL 00/20 1.2] kvm updates Avi Kivity
2012-05-21 16:37 ` [Qemu-devel] " Avi Kivity
2012-05-21 16:37 ` [PATCH 01/19] kvm: Refactor KVMState::max_gsi to gsi_count Avi Kivity
2012-05-21 16:37   ` [Qemu-devel] " Avi Kivity
2012-05-21 16:37 ` [PATCH 02/19] Introduce MSIMessage structure Avi Kivity
2012-05-21 16:37   ` [Qemu-devel] " Avi Kivity
2012-05-21 16:37 ` [PATCH 03/19] kvm: Introduce basic MSI support for in-kernel irqchips Avi Kivity
2012-05-21 16:37   ` [Qemu-devel] " Avi Kivity
2012-05-21 16:37 ` [PATCH 04/19] pc: Enable MSI support at APIC level Avi Kivity
2012-05-21 16:37   ` [Qemu-devel] " Avi Kivity
2012-05-21 16:37 ` [PATCH 05/19] kvm: x86: Wire up MSI support for in-kernel irqchip Avi Kivity
2012-05-21 16:37   ` [Qemu-devel] " Avi Kivity
2012-05-21 16:37 ` [PATCH 06/19] kvm: Update kernel headers Avi Kivity
2012-05-21 16:37   ` [Qemu-devel] " Avi Kivity
2012-05-21 16:37 ` [PATCH 07/19] kvm: Add support for direct MSI injections Avi Kivity
2012-05-21 16:37   ` [Qemu-devel] " Avi Kivity
2012-05-21 16:37 ` [PATCH 08/19] kvm: Enable in-kernel irqchip support by default Avi Kivity
2012-05-21 16:37   ` [Qemu-devel] " Avi Kivity
2012-05-21 16:37 ` [PATCH 09/19] msix: Factor out msix_get_message Avi Kivity
2012-05-21 16:37   ` [Qemu-devel] " Avi Kivity
2012-05-21 16:37 ` [PATCH 10/19] msix: Invoke msix_handle_mask_update on msix_mask_all Avi Kivity
2012-05-21 16:37   ` [Qemu-devel] " Avi Kivity
2012-05-21 16:37 ` [PATCH 11/19] msix: Introduce vector notifiers Avi Kivity
2012-05-21 16:37   ` [Qemu-devel] " Avi Kivity
2012-05-21 16:37 ` [PATCH 12/19] kvm: Rename kvm_irqchip_add_route to kvm_irqchip_add_irq_route Avi Kivity
2012-05-21 16:37   ` [Qemu-devel] " Avi Kivity
2012-05-21 16:37 ` [PATCH 13/19] kvm: Introduce kvm_irqchip_add_msi_route Avi Kivity
2012-05-21 16:37   ` [Qemu-devel] " Avi Kivity
2012-05-21 16:37 ` [PATCH 14/19] kvm: Publicize kvm_irqchip_release_virq Avi Kivity
2012-05-21 16:37   ` [Qemu-devel] " Avi Kivity
2012-05-21 16:37 ` [PATCH 15/19] kvm: Make kvm_irqchip_commit_routes an internal service Avi Kivity
2012-05-21 16:37   ` [Qemu-devel] " Avi Kivity
2012-05-21 16:37 ` [PATCH 16/19] kvm: Introduce kvm_irqchip_add/remove_irqfd Avi Kivity
2012-05-21 16:37   ` [Qemu-devel] " Avi Kivity
2012-05-21 16:37 ` [PATCH 17/19] kvm: Enable use of kvm_irqchip_in_kernel in hwlib code Avi Kivity
2012-05-21 16:37   ` [Qemu-devel] " Avi Kivity
2012-05-21 16:37 ` [PATCH 18/19] msix: Add msix_nr_vectors_allocated Avi Kivity
2012-05-21 16:37   ` [Qemu-devel] " Avi Kivity
2012-05-21 16:37 ` [PATCH 19/19] virtio/vhost: Add support for KVM in-kernel MSI injection Avi Kivity
2012-05-21 16:37   ` [Qemu-devel] " Avi Kivity
2012-06-04  5:46 ` [Qemu-devel] [PULL 00/20 1.2] kvm updates Anthony Liguori
2012-06-05  0:52   ` Andreas Färber
2012-06-05  0:52     ` Andreas Färber
2012-06-05  0:52     ` Andreas Färber
2012-06-05  1:58     ` Anthony Liguori
2012-06-05  1:58       ` Anthony Liguori
2012-06-05  1:58       ` Anthony Liguori
2012-06-05  2:26       ` Andreas Färber
2012-06-05  2:26         ` Andreas Färber
2012-06-05  2:26         ` Andreas Färber
2012-06-05  7:42       ` Avi Kivity
2012-06-05  7:42         ` Avi Kivity
2012-06-05  7:42         ` Avi Kivity
2012-06-05  8:10         ` Alexander Graf
2012-06-05  8:10           ` [Qemu-devel] " Alexander Graf
2012-06-05  8:10           ` Alexander Graf
2012-06-05  8:32           ` Jan Kiszka
2012-06-05  8:32             ` Jan Kiszka
2012-06-05  8:32             ` Jan Kiszka
2012-06-05  9:02         ` Anthony Liguori
2012-06-05  9:02           ` Anthony Liguori
2012-06-05  9:11           ` Avi Kivity
2012-06-05  9:11             ` [Qemu-devel] " Avi Kivity
2012-06-05  9:11             ` Avi Kivity

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.