[PATCH 2/2] UBUNTU: SAUCE: nvme: improve performance for virtual NVMe devices

Tim Gardner tim.gardner at canonical.com
Tue Nov 8 14:23:04 UTC 2016


From: Rob Nelson <rlnelson at google.com>

BugLink: http://bugs.launchpad.net/bugs/1637565

This change provides a mechanism to reduce the number of MMIO doorbell
writes for the NVMe driver. When running in a virtualized environment
like QEMU, the cost of an MMIO is quite hefy here. The main idea for
the patch is provide the device two memory location locations:
 1) to store the doorbell values so they can be lookup without the doorbell
    MMIO write
 2) to store an event index.
I believe the doorbell value is obvious, the event index not so much.
Similar to the virtio specificaiton, the virtual device can tell the
driver (guest OS) not to write MMIO unless you are writing past this
value.

FYI: doorbell values are written by the nvme driver (guest OS) and the
event index is written by the virtual device (host OS).

The patch implements a new admin command that will communicate where
these two memory locations reside. If the command fails, the nvme
driver will work as before without any optimizations.

Contributions:
  Eric Northup <digitaleric at google.com>
  Frank Swiderski <fes at google.com>
  Ted Tso <tytso at mit.edu>
  Keith Busch <keith.busch at intel.com>

Just to give an idea on the performance boost with the vendor
extension: Running fio [1], a stock NVMe driver I get about 200K read
IOPs with my vendor patch I get about 1000K read IOPs. This was
running with a null device i.e. the backing device simply returned
success on every read IO request.

[1] Running on a 4 core machine:
  fio --time_based --name=benchmark --runtime=30
  --filename=/dev/nvme0n1 --nrfiles=1 --ioengine=libaio --iodepth=32
  --direct=1 --invalidate=1 --verify=0 --verify_fatal=0 --numjobs=4
  --rw=randread --blocksize=4k --randrepeat=false

Signed-off-by: Rob Nelson <rlnelson at google.com>
[mlin: port for upstream]
Signed-off-by: Ming Lin <mlin at kernel.org>
[koike: updated for current APIs]
Signed-off-by: Helen Mae Koike Fornazier <helen.koike at collabora.co.uk>

Signed-off-by: Tim Gardner <tim.gardner at canonical.com>

Conflicts:
	drivers/nvme/host/Kconfig
	drivers/nvme/host/pci.c
---
 drivers/nvme/host/Kconfig |   7 +++
 drivers/nvme/host/pci.c   | 149 ++++++++++++++++++++++++++++++++++++++++++++++
 include/linux/nvme.h      |  21 +++++++
 3 files changed, 177 insertions(+)

diff --git a/drivers/nvme/host/Kconfig b/drivers/nvme/host/Kconfig
index f7d37a6..ac9d9b0 100644
--- a/drivers/nvme/host/Kconfig
+++ b/drivers/nvme/host/Kconfig
@@ -25,6 +25,13 @@ config BLK_DEV_NVME_SCSI
 	  emulation to provide stable device names for mount by id, like
 	  some OpenSuSE and SLES versions.
 
+config NVME_VENDOR_EXT_GOOGLE
+	bool "NVMe Vendor Extension for Improved Virtualization"
+	depends on BLK_DEV_NVME
+	---help---
+	  Google extension to reduce the number of MMIO doorbell
+	  writes for the NVMe driver
+
 config NVME_FABRICS
 	tristate
 
diff --git a/drivers/nvme/host/pci.c b/drivers/nvme/host/pci.c
index a8361f5..8265c4a 100644
--- a/drivers/nvme/host/pci.c
+++ b/drivers/nvme/host/pci.c
@@ -50,6 +50,9 @@
 #define SQ_SIZE(depth)		(depth * sizeof(struct nvme_command))
 #define CQ_SIZE(depth)		(depth * sizeof(struct nvme_completion))
 		
+/* Google Vendor ID is not in include/linux/pci_ids.h */
+#define PCI_VENDOR_ID_GOOGLE 0x1AE0
+
 /*
  * We handle AEN commands ourselves and don't even let the
  * block layer know about them.
@@ -101,6 +104,12 @@ struct nvme_dev {
 	u32 cmbsz;
 	struct nvme_ctrl ctrl;
 	struct completion ioq_wait;
+#ifdef CONFIG_NVME_VENDOR_EXT_GOOGLE
+	u32 *db_mem;
+	dma_addr_t doorbell;
+	u32 *ei_mem;
+	dma_addr_t eventidx;
+#endif
 };
 
 static inline struct nvme_dev *to_nvme_dev(struct nvme_ctrl *ctrl)
@@ -131,6 +140,12 @@ struct nvme_queue {
 	u16 qid;
 	u8 cq_phase;
 	u8 cqe_seen;
+#ifdef CONFIG_NVME_VENDOR_EXT_GOOGLE
+	u32 *sq_doorbell_addr;
+	u32 *sq_eventidx_addr;
+	u32 *cq_doorbell_addr;
+	u32 *cq_eventidx_addr;
+#endif
 };
 
 /*
@@ -168,6 +183,9 @@ static inline void _nvme_check_size(void)
 	BUILD_BUG_ON(sizeof(struct nvme_id_ns) != 4096);
 	BUILD_BUG_ON(sizeof(struct nvme_lba_range_type) != 64);
 	BUILD_BUG_ON(sizeof(struct nvme_smart_log) != 512);
+#ifdef CONFIG_NVME_VENDOR_EXT_GOOGLE
+	BUILD_BUG_ON(sizeof(struct nvme_doorbell_memory) != 64);
+#endif
 }
 
 /*
@@ -263,6 +281,51 @@ static int nvme_init_request(void *data, struct request *req,
 	return 0;
 }
 
+#ifdef CONFIG_NVME_VENDOR_EXT_GOOGLE
+static int nvme_vendor_memory_size(struct nvme_dev *dev)
+{
+	return ((num_possible_cpus() + 1) * 8 * dev->db_stride);
+}
+
+static int nvme_set_doorbell_memory(struct nvme_dev *dev)
+{
+	struct nvme_command c;
+
+	memset(&c, 0, sizeof(c));
+	c.doorbell_memory.opcode = nvme_admin_doorbell_memory;
+	c.doorbell_memory.prp1 = cpu_to_le64(dev->doorbell);
+	c.doorbell_memory.prp2 = cpu_to_le64(dev->eventidx);
+
+	return nvme_submit_sync_cmd(dev->ctrl.admin_q, &c, NULL, 0);
+}
+
+static inline int nvme_ext_need_event(u16 event_idx, u16 new_idx, u16 old)
+{
+	/* Borrowed from vring_need_event */
+	return (u16)(new_idx - event_idx - 1) < (u16)(new_idx - old);
+}
+
+static void nvme_ext_write_doorbell(u16 value, u32 __iomem* q_db,
+			   u32* db_addr, volatile u32* event_idx)
+{
+	u16 old_value;
+	if (!db_addr)
+		goto ring_doorbell;
+
+	old_value = *db_addr;
+	*db_addr = value;
+
+	rmb();
+	if (!nvme_ext_need_event(*event_idx, value, old_value))
+		goto no_doorbell;
+
+ring_doorbell:
+	writel(value, q_db);
+no_doorbell:
+	return;
+}
+#endif
+
 /**
  * __nvme_submit_cmd() - Copy a command into a queue and ring the doorbell
  * @nvmeq: The queue to use
@@ -280,9 +343,19 @@ static void __nvme_submit_cmd(struct nvme_queue *nvmeq,
 	else
 		memcpy(&nvmeq->sq_cmds[tail], cmd, sizeof(*cmd));
 
+#ifdef CONFIG_NVME_VENDOR_EXT_GOOGLE
+	if (nvmeq->sq_doorbell_addr)
+		wmb();
+#endif
+
 	if (++tail == nvmeq->q_depth)
 		tail = 0;
+#ifdef CONFIG_NVME_VENDOR_EXT_GOOGLE
+	nvme_ext_write_doorbell(tail, nvmeq->q_db,
+		nvmeq->sq_doorbell_addr, nvmeq->sq_eventidx_addr);
+#else
 	writel(tail, nvmeq->q_db);
+#endif
 	nvmeq->sq_tail = tail;
 }
 
@@ -667,6 +740,10 @@ static void __nvme_process_cq(struct nvme_queue *nvmeq, unsigned int *tag)
 		struct nvme_completion cqe = nvmeq->cqes[head];
 		struct request *req;
 
+#ifdef CONFIG_NVME_VENDOR_EXT_GOOGLE
+		if (to_pci_dev(nvmeq->dev->dev)->vendor == PCI_VENDOR_ID_GOOGLE)
+			rmb();
+#endif
 		if (++head == nvmeq->q_depth) {
 			head = 0;
 			phase = !phase;
@@ -711,7 +788,14 @@ static void __nvme_process_cq(struct nvme_queue *nvmeq, unsigned int *tag)
 		return;
 
 	if (likely(nvmeq->cq_vector >= 0))
+#ifdef CONFIG_NVME_VENDOR_EXT_GOOGLE
+		nvme_ext_write_doorbell(head,
+			nvmeq->q_db + nvmeq->dev->db_stride,
+			nvmeq->cq_doorbell_addr,
+			nvmeq->cq_eventidx_addr);
+#else
 		writel(head, nvmeq->q_db + nvmeq->dev->db_stride);
+#endif
 	nvmeq->cq_head = head;
 	nvmeq->cq_phase = phase;
 
@@ -1066,6 +1150,17 @@ static struct nvme_queue *nvme_alloc_queue(struct nvme_dev *dev, int qid,
 	dev->queues[qid] = nvmeq;
 	dev->queue_count++;
 
+#ifdef CONFIG_NVME_VENDOR_EXT_GOOGLE
+	if (dev->db_mem && dev->ei_mem && qid != 0) {
+		nvmeq->sq_doorbell_addr = &dev->db_mem[qid * 2 * dev->db_stride];
+		nvmeq->cq_doorbell_addr =
+			&dev->db_mem[(qid * 2 + 1) * dev->db_stride];
+		nvmeq->sq_eventidx_addr = &dev->ei_mem[qid * 2 * dev->db_stride];
+		nvmeq->cq_eventidx_addr =
+			&dev->ei_mem[(qid * 2 + 1) * dev->db_stride];
+	}
+#endif
+
 	return nvmeq;
 
  free_cqdma:
@@ -1096,6 +1191,16 @@ static void nvme_init_queue(struct nvme_queue *nvmeq, u16 qid)
 	nvmeq->cq_head = 0;
 	nvmeq->cq_phase = 1;
 	nvmeq->q_db = &dev->dbs[qid * 2 * dev->db_stride];
+#ifdef CONFIG_NVME_VENDOR_EXT_GOOGLE
+	if (to_pci_dev(dev->dev)->vendor == PCI_VENDOR_ID_GOOGLE && qid != 0) {
+		nvmeq->sq_doorbell_addr = &dev->db_mem[qid * 2 * dev->db_stride];
+		nvmeq->cq_doorbell_addr =
+			&dev->db_mem[(qid * 2 + 1) * dev->db_stride];
+		nvmeq->sq_eventidx_addr = &dev->ei_mem[qid * 2 * dev->db_stride];
+		nvmeq->cq_eventidx_addr =
+			&dev->ei_mem[(qid * 2 + 1) * dev->db_stride];
+	}
+#endif
 	memset((void *)nvmeq->cqes, 0, CQ_SIZE(nvmeq->q_depth));
 	dev->online_queues++;
 	spin_unlock_irq(&nvmeq->q_lock);
@@ -1581,6 +1686,19 @@ static int nvme_dev_add(struct nvme_dev *dev)
 		if (blk_mq_alloc_tag_set(&dev->tagset))
 			return 0;
 		dev->ctrl.tagset = &dev->tagset;
+
+#ifdef CONFIG_NVME_VENDOR_EXT_GOOGLE
+		if (to_pci_dev(dev->dev)->vendor == PCI_VENDOR_ID_GOOGLE) {
+			int res = nvme_set_doorbell_memory(dev);
+			if (res) {
+				// Free memory and continue on.
+				dma_free_coherent(dev->dev, 8192, dev->db_mem, dev->doorbell);
+				dma_free_coherent(dev->dev, 8192, dev->ei_mem, dev->doorbell);
+				dev->db_mem = 0;
+				dev->ei_mem = 0;
+			}
+		}
+#endif
 	} else {
 		blk_mq_update_nr_hw_queues(&dev->tagset, dev->online_queues - 1);
 
@@ -1648,8 +1766,31 @@ static int nvme_pci_enable(struct nvme_dev *dev)
 
 	pci_enable_pcie_error_reporting(pdev);
 	pci_save_state(pdev);
+
+#ifdef CONFIG_NVME_VENDOR_EXT_GOOGLE
+	if (pdev->vendor == PCI_VENDOR_ID_GOOGLE) {
+		int mem_size = nvme_vendor_memory_size(dev);
+		dev->db_mem = dma_alloc_coherent(&pdev->dev, mem_size, &dev->doorbell, GFP_KERNEL);
+		if (!dev->db_mem) {
+			result = -ENOMEM;
+			goto disable;
+		}
+		dev->ei_mem = dma_alloc_coherent(&pdev->dev, mem_size, &dev->eventidx, GFP_KERNEL);
+		if (!dev->ei_mem) {
+			result = -ENOMEM;
+			goto dma_free;
+		}
+	}
+#endif
+
 	return 0;
 
+#ifdef CONFIG_NVME_VENDOR_EXT_GOOGLE
+  dma_free:
+	dma_free_coherent(&pdev->dev, nvme_vendor_memory_size(dev), dev->db_mem, dev->doorbell);
+	dev->db_mem = 0;
+#endif
+
  disable:
 	pci_disable_device(pdev);
 	return result;
@@ -1665,6 +1806,14 @@ static void nvme_dev_unmap(struct nvme_dev *dev)
 static void nvme_pci_disable(struct nvme_dev *dev)
 {
 	struct pci_dev *pdev = to_pci_dev(dev->dev);
+#ifdef CONFIG_NVME_VENDOR_EXT_GOOGLE
+	int mem_size = nvme_vendor_memory_size(dev);
+
+	if (dev->db_mem)
+		dma_free_coherent(&pdev->dev, mem_size, dev->db_mem, dev->doorbell);
+	if (dev->ei_mem)
+		dma_free_coherent(&pdev->dev, mem_size, dev->ei_mem, dev->eventidx);
+#endif
 
 	if (pdev->msi_enabled)
 		pci_disable_msi(pdev);
diff --git a/include/linux/nvme.h b/include/linux/nvme.h
index 7676557..5c10a45 100644
--- a/include/linux/nvme.h
+++ b/include/linux/nvme.h
@@ -562,6 +562,9 @@ enum nvme_admin_opcode {
 	nvme_admin_format_nvm		= 0x80,
 	nvme_admin_security_send	= 0x81,
 	nvme_admin_security_recv	= 0x82,
+#ifdef CONFIG_NVME_VENDOR_EXT_GOOGLE
+	nvme_admin_doorbell_memory	= 0xC0,
+#endif
 };
 
 enum {
@@ -827,6 +830,18 @@ struct nvmf_property_get_command {
 	__u8		resv4[16];
 };
 
+#ifdef CONFIG_NVME_VENDOR_EXT_GOOGLE
+struct nvme_doorbell_memory {
+	__u8			opcode;
+	__u8			flags;
+	__u16			command_id;
+	__u32			rsvd1[5];
+	__le64			prp1;
+	__le64			prp2;
+	__u32			rsvd12[6];
+};
+#endif
+
 struct nvme_command {
 	union {
 		struct nvme_common_command common;
@@ -840,6 +855,9 @@ struct nvme_command {
 		struct nvme_format_cmd format;
 		struct nvme_dsm_cmd dsm;
 		struct nvme_abort_cmd abort;
+#ifdef CONFIG_NVME_VENDOR_EXT_GOOGLE
+		struct nvme_doorbell_memory doorbell_memory;
+#endif
 		struct nvme_get_log_page_command get_log_page;
 		struct nvmf_common_command fabrics;
 		struct nvmf_connect_command connect;
@@ -918,6 +936,9 @@ enum {
 	NVME_SC_BAD_ATTRIBUTES		= 0x180,
 	NVME_SC_INVALID_PI		= 0x181,
 	NVME_SC_READ_ONLY		= 0x182,
+#ifdef CONFIG_NVME_VENDOR_EXT_GOOGLE
+	NVME_SC_DOORBELL_MEMORY_INVALID	= 0x1C0,
+#endif
 
 	/*
 	 * I/O Command Set Specific - Fabrics commands:
-- 
2.7.4





More information about the kernel-team mailing list