accel/qaic: Enable 1 MSI fallback mode

Several virtualization use-cases either don't support 32 MultiMSIs
(Xen/VMware) or have significant drawbacks to their use (KVM's vIOMMU,
which is required to support 32 MSI, needs to allocate an alternate
system memory space for each device using vIOMMU (e.g. 8GB VM mem and
2 cards => 8 + 2 * 8 = 24GB host memory required)). Support these
cases by enabling a 1 MSI fallback mode.

Whenever all 32 MSIs requested are not available, a second request for
a single MSI is made. Its success is the initiator of single MSI mode.
This mode causes all interrupts generated by the device to be directed
to the 0th MSI (firmware >=v1.10 will do this as a response to the PCIe
MSI capability configuration). Likewise, all interrupt handlers for the
device are registered to the 0th MSI.

Since the DBC interrupt handler checks if the DBC is in use or if
there is any pending changes, the 'spurious' interrupts are
disregarded. If there is work to be done, the standard threaded IRQ
handler is dispatched.

On every interrupt, the MHI handler wakes up its threaded interrupt
handler, and attempts to wake any waiters for MHI state events.

Performance is within +-0.6% for test cases that typify real world
use. Larger differences ([-4,+132]%, avg +47%) exist for very simple
tasks (e.g. addition) compiled for single NSPs. It is assumed that the
small work and many interrupts typically cause contention (e.g. 16 NSPs
vs 4 CPUs), as evidenced by the standard deviation between runs also
decreasing (r=-0.48 between delta(Performace_test) and
delta(StdDev_test/Avg_test))

Signed-off-by: Carl Vanderlip <quic_carlv@quicinc.com>
Reviewed-by: Pranjal Ramajor Asha Kanojiya <quic_pkanojiy@quicinc.com>
Reviewed-by: Jeffrey Hugo <quic_jhugo@quicinc.com>
Signed-off-by: Jeffrey Hugo <quic_jhugo@quicinc.com>
Reviewed-by: Stanislaw Gruszka <stanislaw.gruszka@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20231016170036.5409-1-quic_jhugo@quicinc.com

1 files changed, 24 insertions, 11 deletions

diff --git a/drivers/accel/qaic/qaic_drv.c b/drivers/accel/qaic/qaic_drv.c
index 6f58095767df..5eba5f03d3bf 100644
--- a/drivers/accel/qaic/qaic_drv.c
+++ b/drivers/accel/qaic/qaic_drv.c
@@ -424,14 +424,24 @@ static int init_msi(struct qaic_device *qdev, struct pci_dev *pdev)
 	int i;
 
 	/* Managed release since we use pcim_enable_device */
-	ret = pci_alloc_irq_vectors(pdev, 1, 32, PCI_IRQ_MSI);
-	if (ret < 0)
-		return ret;
+	ret = pci_alloc_irq_vectors(pdev, 32, 32, PCI_IRQ_MSI);
+	if (ret == -ENOSPC) {
+		ret = pci_alloc_irq_vectors(pdev, 1, 1, PCI_IRQ_MSI);
+		if (ret < 0)
+			return ret;
 
-	if (ret < 32) {
-		pci_err(pdev, "%s: Requested 32 MSIs. Obtained %d MSIs which is less than the 32 required.\n",
-			__func__, ret);
-		return -ENODEV;
+		/*
+		 * Operate in one MSI mode. All interrupts will be directed to
+		 * MSI0; every interrupt will wake up all the interrupt handlers
+		 * (MHI and DBC[0-15]). Since the interrupt is now shared, it is
+		 * not disabled during DBC threaded handler, but only one thread
+		 * will be allowed to run per DBC, so while it can be
+		 * interrupted, it shouldn't race with itself.
+		 */
+		qdev->single_msi = true;
+		pci_info(pdev, "Allocating 32 MSIs failed, operating in 1 MSI mode. Performance may be impacted.\n");
+	} else if (ret < 0) {
+		return ret;
 	}
 
 	mhi_irq = pci_irq_vector(pdev, 0);
@@ -439,15 +449,17 @@ static int init_msi(struct qaic_device *qdev, struct pci_dev *pdev)
 		return mhi_irq;
 
 	for (i = 0; i < qdev->num_dbc; ++i) {
-		ret = devm_request_threaded_irq(&pdev->dev, pci_irq_vector(pdev, i + 1),
+		ret = devm_request_threaded_irq(&pdev->dev,
+						pci_irq_vector(pdev, qdev->single_msi ? 0 : i + 1),
 						dbc_irq_handler, dbc_irq_threaded_fn, IRQF_SHARED,
 						"qaic_dbc", &qdev->dbc[i]);
 		if (ret)
 			return ret;
 
 		if (datapath_polling) {
-			qdev->dbc[i].irq = pci_irq_vector(pdev, i + 1);
-			disable_irq_nosync(qdev->dbc[i].irq);
+			qdev->dbc[i].irq = pci_irq_vector(pdev, qdev->single_msi ? 0 : i + 1);
+			if (!qdev->single_msi)
+				disable_irq_nosync(qdev->dbc[i].irq);
 			INIT_WORK(&qdev->dbc[i].poll_work, irq_polling_work);
 		}
 	}
@@ -479,7 +491,8 @@ static int qaic_pci_probe(struct pci_dev *pdev, const struct pci_device_id *id)
 		goto cleanup_qdev;
 	}
 
-	qdev->mhi_cntrl = qaic_mhi_register_controller(pdev, qdev->bar_0, mhi_irq);
+	qdev->mhi_cntrl = qaic_mhi_register_controller(pdev, qdev->bar_0, mhi_irq,
+						       qdev->single_msi);
 	if (IS_ERR(qdev->mhi_cntrl)) {
 		ret = PTR_ERR(qdev->mhi_cntrl);
 		goto cleanup_qdev;