hw/nvme: fix pin-based interrupt behavior (again)

Jakub noticed[1] that, when using pin-based interrupts, the device will
unconditionally deasssert when any CQEs are acknowledged. However, the
pin should not be deasserted if other completion queues still holds
unacknowledged CQEs.

The bug is an artifact of commit ca247d35098d ("hw/block/nvme: fix
pin-based interrupt behavior") which fixed one bug but introduced
another. This is the third time someone tries to fix pin-based
interrupts (see commit 5e9aa92eb1a5 ("hw/block: Fix pin-based interrupt
behaviour of NVMe"))...

Third time's the charm, so fix it, again, by keeping track of how many
CQs have unacknowledged CQEs and only deassert when all are cleared.

  [1]: <20210610114624.304681-1-jakub.jermar@kernkonzept.com>

Cc: qemu-stable@nongnu.org
Fixes: ca247d35098d ("hw/block/nvme: fix pin-based interrupt behavior")
Reported-by: Jakub Jermář <jakub.jermar@kernkonzept.com>
Signed-off-by: Klaus Jensen <k.jensen@samsung.com>
Reviewed-by: Keith Busch <kbusch@kernel.org>

diff --git a/hw/nvme/ctrl.c b/hw/nvme/ctrl.c
index 5c6c7d3455c3959e3589867064a952ca2c52487e..629b0d38c2a2e5807a7144d3ed7a8ecfed7e98a5 100644 (file)
--- a/hw/nvme/ctrl.c
+++ b/hw/nvme/ctrl.c
@@ -473,7 +473,9 @@ static void nvme_irq_deassert(NvmeCtrl *n, NvmeCQueue *cq)
             return;
         } else {
             assert(cq->vector < 32);
-            n->irq_status &= ~(1 << cq->vector);
+            if (!n->cq_pending) {
+                n->irq_status &= ~(1 << cq->vector);
+            }
             nvme_irq_check(n);
         }
     }
@@ -1253,6 +1255,7 @@ static void nvme_post_cqes(void *opaque)
     NvmeCQueue *cq = opaque;
     NvmeCtrl *n = cq->ctrl;
     NvmeRequest *req, *next;
+    bool pending = cq->head != cq->tail;
     int ret;
 
     QTAILQ_FOREACH_SAFE(req, &cq->req_list, entry, next) {
@@ -1282,6 +1285,10 @@ 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++;
+        }
+
         nvme_irq_assert(n, cq);
     }
 }
@@ -4297,6 +4304,11 @@ static uint16_t nvme_del_cq(NvmeCtrl *n, NvmeRequest *req)
         trace_pci_nvme_err_invalid_del_cq_notempty(qid);
         return NVME_INVALID_QUEUE_DEL;
     }
+
+    if (cq->irq_enabled && cq->tail != cq->head) {
+        n->cq_pending--;
+    }
+
     nvme_irq_deassert(n, cq);
     trace_pci_nvme_del_cq(qid);
     nvme_free_cq(cq, n);
@@ -6039,6 +6051,10 @@ static void nvme_process_db(NvmeCtrl *n, hwaddr addr, int val)
         }
 
         if (cq->tail == cq->head) {
+            if (cq->irq_enabled) {
+                n->cq_pending--;
+            }
+
             nvme_irq_deassert(n, cq);
         }
     } else {

diff --git a/hw/nvme/nvme.h b/hw/nvme/nvme.h
index 2509b8b039ef432f3c20e1e455a8c24737957d67..56f8eceed2ad08375897057f593e6bddf16884c0 100644 (file)
--- a/hw/nvme/nvme.h
+++ b/hw/nvme/nvme.h
@@ -410,6 +410,7 @@ typedef struct NvmeCtrl {
     uint32_t    max_q_ents;
     uint8_t     outstanding_aers;
     uint32_t    irq_status;
+    int         cq_pending;
     uint64_t    host_timestamp;                 /* Timestamp sent by the host */
     uint64_t    timestamp_set_qemu_clock_ms;    /* QEMU clock time */
     uint64_t    starttime_ms;