Received-SPF: pass (google.com: domain of linux-kernel+bounces-175873-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:4601:e00::3 as permitted sender) client-ip=2604:1380:4601:e00::3;
From: Keith Busch <kbusch@meta.com>
To: <linux-nvme@lists.infradead.org>, <linux-kernel@vger.kernel.org>
CC: <hch@lst.de>, <tglx@linutronix.de>, <ming.lei@redhat.com>,
        Keith Busch
	<kbusch@kernel.org>
Subject: [PATCH 2/2] nvme-pci: allow unmanaged interrupts
Date: Fri, 10 May 2024 07:14:59 -0700
Message-ID: <20240510141459.3207725-2-kbusch@meta.com>
In-Reply-To: <20240510141459.3207725-1-kbusch@meta.com>
References: <20240510141459.3207725-1-kbusch@meta.com>
Precedence: bulk
MIME-Version: 1.0
Content-Transfer-Encoding: quoted-printable
Content-Type: text/plain

From: Keith Busch <kbusch@kernel.org>

Some people _really_ want to control their interrupt affinity.

Signed-off-by: Keith Busch <kbusch@kernel.org>
---
 drivers/nvme/host/pci.c | 13 ++++++++++++-
 1 file changed, 12 insertions(+), 1 deletion(-)

diff --git a/drivers/nvme/host/pci.c b/drivers/nvme/host/pci.c
index 8e0bb9692685d..4c2799c3f45f5 100644
--- a/drivers/nvme/host/pci.c
+++ b/drivers/nvme/host/pci.c
@@ -63,6 +63,11 @@ MODULE_PARM_DESC(sgl_threshold,
 		"Use SGLs when average request segment size is larger or equal to "
 		"this size. Use 0 to disable SGLs.");
=20
+static bool managed_irqs =3D true;
+module_param(managed_irqs, bool, 0444);
+MODULE_PARM_DESC(managed_irqs,
+		 "set to false for user controlled irq affinity");
+
 #define NVME_PCI_MIN_QUEUE_SIZE 2
 #define NVME_PCI_MAX_QUEUE_SIZE 4095
 static int io_queue_depth_set(const char *val, const struct kernel_param=
 *kp);
@@ -456,7 +461,7 @@ static void nvme_pci_map_queues(struct blk_mq_tag_set=
 *set)
 		 * affinity), so use the regular blk-mq cpu mapping
 		 */
 		map->queue_offset =3D qoff;
-		if (i !=3D HCTX_TYPE_POLL && offset)
+		if (managed_irqs && i !=3D HCTX_TYPE_POLL && offset)
 			blk_mq_pci_map_queues(map, to_pci_dev(dev->dev), offset);
 		else
 			blk_mq_map_queues(map);
@@ -2180,6 +2185,9 @@ static void nvme_calc_irq_sets(struct irq_affinity =
*affd, unsigned int nrirqs)
 	struct nvme_dev *dev =3D affd->priv;
 	unsigned int nr_read_queues, nr_write_queues =3D dev->nr_write_queues;
=20
+	if (!nrirqs)
+		nrirqs =3D affd->post_vectors;
+
 	/*
 	 * If there is no interrupt available for queues, ensure that
 	 * the default queue is set to 1. The affinity set size is
@@ -2226,6 +2234,9 @@ static int nvme_setup_irqs(struct nvme_dev *dev, un=
signed int nr_io_queues)
 	poll_queues =3D min(dev->nr_poll_queues, nr_io_queues - 1);
 	dev->io_queues[HCTX_TYPE_POLL] =3D poll_queues;
=20
+	if (!managed_irqs)
+		affd.post_vectors =3D nr_io_queues - poll_queues;
+
 	/*
 	 * Initialize for the single interrupt case, will be updated in
 	 * nvme_calc_irq_sets().
--=20
2.43.0