Skip to contentSkip to conversation

hw/block/nvme: fixes for interrupt behavior

PATCH In review Klaus Jensen 2 participants 5 messages May 5, 2020 – May 11, 2020

lore
Share & tools
lore
#0 lore raw

cover hw/block/nvme: fixes for interrupt behavior

From: Klaus Jensen <[email protected]>

Klaus Jensen (2):
hw/block/nvme: fix pin-based interrupt behavior hw/block/nvme: allow use of any valid msix vector

 hw/block/nvme.c | 14 +++++++++-----
 hw/block/nvme.h |  2 +-
 2 files changed, 10 insertions(+), 6 deletions(-)
Signature · 1 line
2.26.2
#1 lore raw

PATCH 1/2 hw/block/nvme: fix pin-based interrupt behavior

From: Klaus Jensen <[email protected]>

First, since the device only supports MSI-X or pin-based interrupt, if MSI-X is not enabled, it should not accept interrupt vectors different from 0 when creating completion queues.

Secondly, the irq_status NvmeCtrl member is meant to be compared to the INTMS register, so it should only be 32 bits wide. And it is really only useful when used with multi-message MSI.

Third, since we do not force a 1-to-1 correspondence between cqid and interrupt vector, the irq_status register should not have bits set according to cqid, but according to the associated interrupt vector.

Fix these issues, but keep irq_status available so we can easily support multi-message MSI down the line.

Fixes 5e9aa92eb1a5 ("hw/block: Fix pin-based interrupt behaviour of NVMe")
---
 hw/block/nvme.c | 12 ++++++++----
 hw/block/nvme.h |  2 +-
 2 files changed, 9 insertions(+), 5 deletions(-)
hw/block/nvme.c +8 -4
index 9b453423cf2c..2b072bbb21e7 100644
@@ -127,8 +127,8 @@ static void nvme_irq_assert(NvmeCtrl *n, NvmeCQueue *cq)
             msix_notify(&(n->parent_obj), cq->vector);
         } else {
             trace_nvme_irq_pin();
-            assert(cq->cqid < 64);
-            n->irq_status |= 1 << cq->cqid;
+            assert(cq->vector < 32);
+            n->irq_status |= 1 << cq->vector;
             nvme_irq_check(n);
         }
     } else {
@@ -142,8 +142,8 @@ static void nvme_irq_deassert(NvmeCtrl *n, NvmeCQueue *cq)
         if (msix_enabled(&(n->parent_obj))) {
             return;
         } else {
-            assert(cq->cqid < 64);
-            n->irq_status &= ~(1 << cq->cqid);
+            assert(cq->vector < 32);
+            n->irq_status &= ~(1 << cq->vector);
             nvme_irq_check(n);
         }
     }
@@ -642,6 +642,10 @@ static uint16_t nvme_create_cq(NvmeCtrl *n, NvmeCmd *cmd)
         trace_nvme_err_invalid_create_cq_addr(prp1);
         return NVME_INVALID_FIELD | NVME_DNR;
     }
+    if (unlikely(!msix_enabled(&n->parent_obj) && vector)) {
+        trace_nvme_err_invalid_create_cq_vector(vector);
+        return NVME_INVALID_IRQ_VECTOR | NVME_DNR;
+    }
     if (unlikely(vector > n->num_queues)) {
         trace_nvme_err_invalid_create_cq_vector(vector);
         return NVME_INVALID_IRQ_VECTOR | NVME_DNR;
hw/block/nvme.h +1 -1
index 6520a9f0bead..db62589247d7 100644
@@ -78,7 +78,7 @@ typedef struct NvmeCtrl {
     uint32_t    cmbsz;
     uint32_t    cmbloc;
     uint8_t     *cmbuf;
-    uint64_t    irq_status;
+    uint32_t    irq_status;
     uint64_t    host_timestamp;                 /* Timestamp sent by the host */
     uint64_t    timestamp_set_qemu_clock_ms;    /* QEMU clock time */
 
Signature · 1 line
2.26.2
#2 lore raw

PATCH 2/2 hw/block/nvme: allow use of any valid msix vector

In reply to Klaus Jensen

From: Klaus Jensen <[email protected]>

If the device uses MSI-X, any of the 2048 MSI-X interrupt vectors are valid. If the device is not using MSI-X, vector will and can only be zero at this point.

---
 hw/block/nvme.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
hw/block/nvme.c +1 -1
index 2b072bbb21e7..755ced8b03fb 100644
@@ -646,7 +646,7 @@ static uint16_t nvme_create_cq(NvmeCtrl *n, NvmeCmd *cmd)
         trace_nvme_err_invalid_create_cq_vector(vector);
         return NVME_INVALID_IRQ_VECTOR | NVME_DNR;
     }
-    if (unlikely(vector > n->num_queues)) {
+    if (unlikely(vector > PCI_MSIX_FLAGS_QSIZE)) {
         trace_nvme_err_invalid_create_cq_vector(vector);
         return NVME_INVALID_IRQ_VECTOR | NVME_DNR;
     }
Signature · 1 line
2.26.2
#3 lore raw
In reply to Klaus Jensen

+ Michael & Marcel for MSIX,
and ping to Keith :)

Quoted 9 lines · Klaus Jensen

From: Klaus Jensen <[email protected]>

Klaus Jensen (2):
hw/block/nvme: fix pin-based interrupt behavior hw/block/nvme: allow use of any valid msix vector

  hw/block/nvme.c | 14 +++++++++-----
  hw/block/nvme.h |  2 +-
  2 files changed, 10 insertions(+), 6 deletions(-)
#4 lore raw
Quoted 3 lines · Philippe Mathieu-Daudé

+ Michael & Marcel for MSIX,
and ping to Keith :)

I'll await some feedback here before moving these fixes into my "nvme:
refactoring and cleanups" series. Not sure if this should go directly to master or block-next.