[PATCH v2 1/3] hw/nvme: support irq(de)assertion with eventfd

[Jinhao Fan]

When the new option 'irq-eventfd' is turned on, the IO emulation code
signals an eventfd when it want to (de)assert an irq. The main loop
eventfd handler does the actual irq (de)assertion.  This paves the way
for iothread support since QEMU's interrupt emulation is not thread
safe.

Asserting and deasseting irq with eventfd has some performance
implications. For small queue depth it increases request latency but
for large queue depth it effectively coalesces irqs.

Comparision (KIOPS):

QD            1   4  16  64
QEMU         38 123 210 329
irq-eventfd  32 106 240 364

Signed-off-by: Jinhao Fan <fanjinhao21s@ict.ac.cn>
---
 hw/nvme/ctrl.c | 136 ++++++++++++++++++++++++++++++++++++++++++-------
 hw/nvme/nvme.h |   4 ++
 2 files changed, 123 insertions(+), 17 deletions(-)

diff --git a/hw/nvme/ctrl.c b/hw/nvme/ctrl.c
index 87aeba0564..6ecf6fafd9 100644
--- a/hw/nvme/ctrl.c
+++ b/hw/nvme/ctrl.c
@@ -526,34 +526,57 @@ static void nvme_irq_check(NvmeCtrl *n)
     }
 }
 
+static void nvme_irq_do_assert(NvmeCtrl *n, NvmeCQueue *cq)
+{
+    if (msix_enabled(&(n->parent_obj))) {
+        trace_pci_nvme_irq_msix(cq->vector);
+        msix_notify(&(n->parent_obj), cq->vector);
+    } else {
+        trace_pci_nvme_irq_pin();
+        assert(cq->vector < 32);
+        n->irq_status |= 1 << cq->vector;
+        nvme_irq_check(n);
+    }
+}
+
 static void nvme_irq_assert(NvmeCtrl *n, NvmeCQueue *cq)
 {
     if (cq->irq_enabled) {
-        if (msix_enabled(&(n->parent_obj))) {
-            trace_pci_nvme_irq_msix(cq->vector);
-            msix_notify(&(n->parent_obj), cq->vector);
+        if (cq->assert_notifier.initialized) {
+            event_notifier_set(&cq->assert_notifier);
         } else {
-            trace_pci_nvme_irq_pin();
-            assert(cq->vector < 32);
-            n->irq_status |= 1 << cq->vector;
-            nvme_irq_check(n);
+            nvme_irq_do_assert(n, cq);
         }
     } else {
         trace_pci_nvme_irq_masked();
     }
 }
 
+static void nvme_irq_do_deassert(NvmeCtrl *n, NvmeCQueue *cq)
+{
+    if (msix_enabled(&(n->parent_obj))) {
+        return;
+    } else {
+        assert(cq->vector < 32);
+        if (!n->cq_pending) {
+            n->irq_status &= ~(1 << cq->vector);
+        }
+        nvme_irq_check(n);
+    }
+}
+
 static void nvme_irq_deassert(NvmeCtrl *n, NvmeCQueue *cq)
 {
     if (cq->irq_enabled) {
-        if (msix_enabled(&(n->parent_obj))) {
-            return;
+        if (cq->deassert_notifier.initialized) {
+            /* 
+             * Event notifier will only be initilized when MSI-X is in use,
+             * therefore no need to worry about extra eventfd syscall for
+             * pin-based interrupts.
+             */
+            event_notifier_set(&cq->deassert_notifier);
         } else {
-            assert(cq->vector < 32);
-            if (!n->cq_pending) {
-                n->irq_status &= ~(1 << cq->vector);
-            }
-            nvme_irq_check(n);
+            nvme_irq_do_deassert(n, cq);
         }
     }
 }
@@ -1338,6 +1361,54 @@ static void nvme_update_cq_head(NvmeCQueue *cq)
     trace_pci_nvme_shadow_doorbell_cq(cq->cqid, cq->head);
 }
 
+static void nvme_assert_notifier_read(EventNotifier *e)
+{
+    NvmeCQueue *cq = container_of(e, NvmeCQueue, assert_notifier);
+    if (event_notifier_test_and_clear(e)) {
+        nvme_irq_do_assert(cq->ctrl, cq);
+    }
+}
+
+static void nvme_deassert_notifier_read(EventNotifier *e)
+{
+    NvmeCQueue *cq = container_of(e, NvmeCQueue, deassert_notifier);
+    if (event_notifier_test_and_clear(e)) {
+        nvme_irq_do_deassert(cq->ctrl, cq);
+    }
+}
+
+static void nvme_init_irq_notifier(NvmeCtrl *n, NvmeCQueue *cq)
+{
+    int ret;
+
+    ret = event_notifier_init(&cq->assert_notifier, 0);
+    if (ret < 0) {
+        goto fail_assert_handler;
+    }
+
+    event_notifier_set_handler(&cq->assert_notifier,
+                                nvme_assert_notifier_read);
+
+    if (!msix_enabled(&n->parent_obj)) {
+        ret = event_notifier_init(&cq->deassert_notifier, 0);
+        if (ret < 0) {
+            goto fail_deassert_handler;
+        }
+
+        event_notifier_set_handler(&cq->deassert_notifier,
+                                   nvme_deassert_notifier_read);
+    }
+
+    return;
+
+fail_deassert_handler:
+    event_notifier_set_handler(&cq->deassert_notifier, NULL);
+    event_notifier_cleanup(&cq->deassert_notifier);
+fail_assert_handler:
+    event_notifier_set_handler(&cq->assert_notifier, NULL);
+    event_notifier_cleanup(&cq->assert_notifier);
+}
+
 static void nvme_post_cqes(void *opaque)
 {
     NvmeCQueue *cq = opaque;
@@ -1377,8 +1448,25 @@ static void nvme_post_cqes(void *opaque)
         QTAILQ_INSERT_TAIL(&sq->req_list, req, entry);
     }
     if (cq->tail != cq->head) {
-        if (cq->irq_enabled && !pending) {
-            n->cq_pending++;
+        if (cq->irq_enabled) {
+            if (!pending) {
+                n->cq_pending++;
+            }
+
+            if (unlikely(cq->first_io_cqe)) {
+                /*
+                 * Initilize event notifier when first cqe is posted. For 
irqfd 
+                 * support we need to register the MSI message in KVM. We
+                 * can not do this registration at CQ creation time because
+                 * Linux's NVMe driver changes the MSI message after CQ 
creation.
+                 */
+                cq->first_io_cqe = false;
+
+                if (n->params.irq_eventfd) {
+                    nvme_init_irq_notifier(n, cq);
+                }
+            }
+
         }
 
         nvme_irq_assert(n, cq);
@@ -4705,6 +4793,14 @@ static void nvme_free_cq(NvmeCQueue *cq, NvmeCtrl *n)
         event_notifier_set_handler(&cq->notifier, NULL);
         event_notifier_cleanup(&cq->notifier);
     }
+    if (cq->assert_notifier.initialized) {
+        event_notifier_set_handler(&cq->assert_notifier, NULL);
+        event_notifier_cleanup(&cq->assert_notifier);
+    }
+    if (cq->deassert_notifier.initialized) {
+        event_notifier_set_handler(&cq->deassert_notifier, NULL);
+        event_notifier_cleanup(&cq->deassert_notifier);
+    }
     if (msix_enabled(&n->parent_obj)) {
         msix_vector_unuse(&n->parent_obj, cq->vector);
     }
@@ -4734,7 +4830,7 @@ static uint16_t nvme_del_cq(NvmeCtrl *n, NvmeRequest *req)
         n->cq_pending--;
     }
 
-    nvme_irq_deassert(n, cq);
+    nvme_irq_do_deassert(n, cq);
     trace_pci_nvme_del_cq(qid);
     nvme_free_cq(cq, n);
     return NVME_SUCCESS;
@@ -4772,6 +4868,11 @@ static void nvme_init_cq(NvmeCQueue *cq, NvmeCtrl *n, 
uint64_t dma_addr,
     }
     n->cq[cqid] = cq;
     cq->timer = timer_new_ns(QEMU_CLOCK_VIRTUAL, nvme_post_cqes, cq);
+    /* 
+     * Only enable irqfd for IO queues since we always emulate admin queue 
+     * in main loop thread 
+     */
+    cq->first_io_cqe = cqid != 0;
 }
 
 static uint16_t nvme_create_cq(NvmeCtrl *n, NvmeRequest *req)
@@ -7671,6 +7772,7 @@ static Property nvme_props[] = {
     DEFINE_PROP_BOOL("use-intel-id", NvmeCtrl, params.use_intel_id, false),
     DEFINE_PROP_BOOL("legacy-cmb", NvmeCtrl, params.legacy_cmb, false),
     DEFINE_PROP_BOOL("ioeventfd", NvmeCtrl, params.ioeventfd, false),
+    DEFINE_PROP_BOOL("irq-eventfd", NvmeCtrl, params.irq_eventfd, false),
     DEFINE_PROP_UINT8("zoned.zasl", NvmeCtrl, params.zasl, 0),
     DEFINE_PROP_BOOL("zoned.auto_transition", NvmeCtrl,
                      params.auto_transition_zones, true),
diff --git a/hw/nvme/nvme.h b/hw/nvme/nvme.h
index 79f5c281c2..759d0ecd7c 100644
--- a/hw/nvme/nvme.h
+++ b/hw/nvme/nvme.h
@@ -398,6 +398,9 @@ typedef struct NvmeCQueue {
     uint64_t    ei_addr;
     QEMUTimer   *timer;
     EventNotifier notifier;
+    EventNotifier assert_notifier;
+    EventNotifier deassert_notifier;
+    bool        first_io_cqe;
     bool        ioeventfd_enabled;
     QTAILQ_HEAD(, NvmeSQueue) sq_list;
     QTAILQ_HEAD(, NvmeRequest) req_list;
@@ -422,6 +425,7 @@ typedef struct NvmeParams {
     bool     auto_transition_zones;
     bool     legacy_cmb;
     bool     ioeventfd;
+    bool     irq_eventfd;
     uint8_t  sriov_max_vfs;
     uint16_t sriov_vq_flexible;
     uint16_t sriov_vi_flexible;
-- 
2.25.1